From e7b5e33b7ed9992cb375760712f6d860a63587ab Mon Sep 17 00:00:00 2001 From: j3xugit Date: Fri, 7 Apr 2017 14:28:09 +0000 Subject: [PATCH] the first draft for protein structure prediction (#191) This build is based on https://github.com/greenelab/deep-review/commit/b3c72b65796a5103c92d7f5b90d2dfb4231ab6f5. This commit was created by the following Travis CI build and job: https://travis-ci.org/greenelab/deep-review/builds/219705004 https://travis-ci.org/greenelab/deep-review/jobs/219705005 [ci skip] The full commit message that triggered this build is copied below: the first draft for protein structure prediction (#191) * Update 04_study.md * Update 04_study.md * Update 04_study.md * Update 04_study.md * Update 04_study.md Now each line has <80 chars (including space) * Update 04_study.md * Update 04_study.md * Update 04_study.md * Update 04_study.md * Update 04_study.md * Update 04_study.md * Line wrap to trigger CI build * Fix doi tag * Fix arxiv reference --- README.md | 2 +- README.md.ots | Bin 300 -> 300 bytes deep-review.pdf | Bin 255889 -> 283822 bytes deep-review.pdf.ots | Bin 335 -> 335 bytes index.html | 248 +++++++++++++++++++++++++++----------------- index.html.ots | Bin 335 -> 335 bytes 6 files changed, 152 insertions(+), 98 deletions(-) diff --git a/README.md b/README.md index af620f1e..e79a6864 100644 --- a/README.md +++ b/README.md @@ -15,4 +15,4 @@ This directory contains the following files, which are mostly ignored on the `ma ## Source The manuscripts in this directory were built from -[`f55ffb034bf9a0512bb801c930aee191efe3c670`](https://github.com/greenelab/deep-review/commit/f55ffb034bf9a0512bb801c930aee191efe3c670). +[`b3c72b65796a5103c92d7f5b90d2dfb4231ab6f5`](https://github.com/greenelab/deep-review/commit/b3c72b65796a5103c92d7f5b90d2dfb4231ab6f5). diff --git a/README.md.ots b/README.md.ots index 29926b698f81de46127be98c563090ee51c146cb..363a21f1ae871d27bcad4db0ed61d40e66068cf5 100644 GIT binary patch delta 174 zcmV;f08#&}0;~d%A%8TBfrOh;#~3^7zmFwLRiw})O&G?pWv}13QoGYt0s-(4e+dqj z=FCDZj)M<+E5tE=2=O2-Y3uG$AWUpMPLfH%k`;AIRdh!wyu4(0);j+KsV}1l|L_n7 zZ^Kt>!a(@;Fi_eww4qQ4EmAuI7Ba0Jpcdz delta 174 zcmV;f08#&}0;~d%A%A#lM0&R7w?(gFJ}b@ay{J)gN&dn)+Gozh;M4bwm=w?nm;p~BrrG#|L_n7 zFoqBlusjJn)@LNG(A(|^@dQ}sdXex5eT=k<^n>!^0E6G-4UwQTA?@o*+plqF4S2Sx c6Hr39;0W;qSm%0?@CdxOVw|rfyYG?oIQf-SFaQ7m diff --git a/deep-review.pdf b/deep-review.pdf index fdd743aa73ddd5f319dabe4c287b650afd2cbc73..28aae027a78ea5c3141d548375b8de6da7a1354c 100644 GIT binary patch delta 133926 zcmZ6TbwHHM*T-Fyl zs`vih{bQevIcH|hnP=nqe&!@j5LLY(c@~2y2oVN>gPt`!19agdB>_YzTLJ(RN{bLc zjmjbb;2>KHI8Z+xW8l*Xfdqk22nc;+W(LFHK)8?y`o_#G1css}c`@ja%m=elrA!1eN#-K=aQ4VEEq70V*hn zdLIu!VhF)da1yi;6g@%VKV3nD{>B|D2muQI>kO4A_Nyx)6hA3Ij0YkJ1^$XNIzv#f z5b#&D(2pTPkbegb5d;c@kR&9Uw{ie6Zr30K}3Xrpx;A)h`|4j8;Z^> z3^fGZ7a|0WdzNBjR1444QI`aAIu^ddlq@psBFgy1M53V;~u*%<(z z6f6t_3jJ;eM(0criXwgv$ix7{P`4-nlI-YE|4uwu2nPJ+96hU_%j&Nc42C2bMPk69 zs0&Jf7%MvZf0rCs5cYTWV04;)g%5-R81@y_eZQcoH`uw zv|AxZ%yEXB2lp1tfBJm9|+u)n*Md%d-YV7NMGipZ5hoUisB zproV_0i*GT+1b~k$B65P*D@%E3?8YAtD(h&zJ#XLl_z4CUxzYZtX@xyBA7}dX2BD8 zE;hb*bzKHtH8g$YT$Q|D+*!Rkc*b;fa%_%}xEMm@T%CTkZNEaCZdC58J~=~mWnT}Z zRk&zeCoW$aUmqe?t}b&45mIV-kC~dTKWM1wl!M#Q%9zxkU!PzX@aPBZ)eJFOMAK}Yn{XTPjwTF|WDyZk+Cc^%8mi(v3hvO9P z^@Q=^6LcQ_FWTZhglhm3N5SiR+(p>87^h+s z0|lMM;d}UL4OMGCha4J&>zg0n3*`p-d;xd?pA*-#GhH$8X%~t)wiFRMPtdajYSQn! z1!xtMrR*>t@#}BAJSTIK8L~&UR{HuXc&x>RbjihquVRltP)Nk^XHjuoO4|rjvvwYT zrs<;Hz~M!9_slIM#37|7>?_9FTz-pe6Gs5z^k|dJ-%9zoKe&*$H^YJtZdE^8`z@|> z@yjD-X0yTXrKi!y0aN*$+oLDMYV# zPif5Si_JMmrJx)O?x!{-yD1$l$S)NeMq3+_ENk;1bZh(NU0+h9H}ODTh3(*XIls1~ zIQF_)$#PJ`yFsV&HC6dqIUd4S28?rNy;ETEn-8Gtig?xt>oo&9lij861NnrVsDe&k z!9ZDzN-jygRKpE8JL|xy4`0FS0TWE^;#%r`($<!pd;laza_J+iMYm16}5 z%F<pVmjggT%>=+53)OezwN@^OE8k2-ImOmC2+q;_)BS+F!-oY&r z0nhpRK#Fwu6xZ_9i7!mEgvR_y^Y#yR^8xPu2vnMf4@fjwIyhDAqXleFPA#-|)F=R+ zjF|8Y=Rl0aO_j|>XJ-M$#VPGhBiK2faGjKRTZmRv+~?rU;@n@{psSgU%XRUmk$M@! z)>9ocz`Ie>o6oeZ3|YZ$9|nFf31~ee-I#Tslc^lT-<(oqytrpPd}>@1Ya^Z!L=JlC zIxPm)7uenw7*dJVz+)tE+xjBR8`4kT=){p5yQWJ2VF{a@nF7av^(gTn>XMYDY^a5(jU-0rUq{XoPNVX>o1 z2-F?a=>90Ez^FtHqU0tOlcTG&w!xR2!KAGm!+96mK-`HJdo7iLU~hqC)00wjt?@=o zz!5AWIU{aXRxC|?i?yeZg2M56>{RcyBj=3!(lTQq-TAjxWg^&9)n|fjWc@%4u1`a( zY}OOq>oHtS07Wu>FPsGuTGlS}%PHiZ)Hd!^PqhbHm}KDV>TH^6nF)1miz07)k(~(* zi7}2LAzMP+(6D&NPuYXJ&)MANEwOuh-yEU*Axk@#`oUg)se65*rmpHz6ofHN-fQsQ zSg3f7Fgr2-li=?WG;yRKGm|xe7};8v@Av(N_6IIJw~+ew`IM?xYZ*|x)bmh&nlVC4 zhX=HLRExS#jFs8nvSzrY-p4^pIrH>!ECvp$mu82v9wV!_Yx| z1Nqk7?J9`6bn3`Ka)TQp4x6>y^|Y^oyt%An?^W+aqP$z#i|!~4#OC`bD#n^`vJ4A; zmxVDuQlD66Ja{DSy-OSv=h5C2Yx$asI9TU-Y2?dKO^u6Rg1y+>)kO!_@n+5xldz$F zph}e*AEIru`leJzJ9{dTA&xte>kr!2t~(~sBrn?6`!>waU?X=epAslFGfSXLAz==OCoM37+n%}dBqu>08-h=oS)A|6;euH$j~#zr|w}pE`pR^L!>)fRK~_} z)|l%aJ}#_>010fE`y5d3j673?x~UzYFba4e~NcO+pniMdCVTL|}|1>%ew%#fXSY;HFr ztc5%JgtgoAh~YeZ=*PXXiD<99blH@3s^eRzR0D+&RoZ6mSA1eYL$8HwVM{AZfft(= zW7hg6SG2K`PZ}QtThye#NbP$;sR=U439MLj2;PSDe6|k0{CN5P-O;6L?vtwV1F6%+ zMffOfz(UqEd3ff=G9?1$EpeeVI5wmo9Sk=ZxmThYl^Z(t?1%NBlX!;k?o*Fiuv(cp zXK0g;+Y4JQ>fFzE-%0 zlA}ykvv?Fk0RMmx|6U3jfiRWDo($TJ)-kHTBh9=NHEQ}X_dv>D%GWFIb*i`DwM;Df zY1rDfZ^G-G@$Ey-flA4GK{?g^f>bQLHosJRJEgPG2-XUfD%qBDkiRmr*fgRMN}ceb zBU4}PYwXyY@Mq%ia-?U4lY=Bf^hizaecC9BX{U|*u4IS|Zv}WsZx(sB8Jkff`abBK zwHCYRj+2D$0^2NG!*|w(^bmoqBkx+N52WkuAqQ32L-z?4T4pS{TKtV}$B=wJN@;9N zuL%{jW>RA@UAY|~LTw(NY@h5DICS{f>D-m$_>|l~woy45Iq79*rcLJb#VZ^?!vl|f zjN|~s&Wjj>%L=Gsv%lYWX_@tQJLR~tk)~h?HuaU{RADIn!mHcP(y5c8ouaB1p1)U= zinYt#DOX8RycfV&mWCA*SlK`|q8H>dp7#o__nafNRxWfcf|&_HWgLBrB0bFK=J#wy zPGSGvFI#)qm)cUGjQ+4$-+?SLpkI)Y=S?NaId#P4Cgrp4XGTuDvJLDZV-*aPOb4PH z97bkocZx+VW%G@4Fsxd%`^ZwDxh)2SKlV5dZTtxGDK#Ij_e&b= zOX=Br+n==$re?_Ob$LCNTVd{G4c*9LdGY2U!o7c_$4d%hP}y2OJwIe`3sWM0q!<|9 z?Ad{DrnZb<5>3-S0-k;+9-{wqJ$l$4IxT#<3Zcd~yZm_qXZMslwykMkS?vqB`CT`A z&70NXBS)Wbu8>W1#F{PylBLs7&J~>aLA9WEx%GQDaK;MR@8jGjU?}ga4!uZ=RzX-7 z4)uG`wB5~k11G^=e@ZI{=CV|?AXmItXWcshnX;r|$r(k{aOj5s#K^S8!V64BC6wMP zk9t~?^hWxJBQuoV_g9^PZ2N6nEF5WQSWP*ixm|@8GI+ux9(%|ZweN*0q3X6UPS6HU zw)J!~mI2PIH2l&LcitON^f9VWA>{9Sj05Q-q=xf}- zW|F2D3YZl5tcRKYO4GZsEs z9w+Y!`eN3l$JmB_my;29*VDW;BUMWiSn^CP^%G~kZa6#g;F_=j{q@*1wT#^AnnI;w z6WOddipY28zvZK!SeT^X z(r8{y716#r(pI;VR6hyNn%N^&;Sd9B6pg`*b-{z{3H4Sa88d@{5b3AGGMm*%5T|Ihp@Qukdjq>-D7g%`NZ9;_B-F| z{I^rs%`91I*MbEa9(GO?`b){pzSx3TwZ{|c8LVV`7V}4x_;jd?OveLh3AIrhVp0dm zyY~*TtE<&p#LEw)BG;04OF&EX2kI?LHhsC|Vq5Nbbd5$A2)hFdxF= zcjxw18l-n$F5+$RF}CwQrfuPEj%VvO;kuTb@Af4{O+_ka9^j3KaEdHl#zaVOhMLLP z$(Q1ZtW~K+D%Btdfu5+X?ojrf^RbFy8xG?t`fps`csfK| z`i79R)|(iJxsOgzcclBd;j`CWSFI}m3$+&P5I8pzY8!?z`brG}T_8NSJv3o;2M1|JSd#u&U z*YH%CMFOf6j)y^3>5@JpQ*>sYmTvj_Ny+Af@28|fE<7J|_5p%sk7w)$2L81m z5u3ywnk@TEX46JlkOoMFaS^MDvE*GXw`X9lfIsMEmg_ zHHgvB^#Mh7nnW>_2f`|jewm-GaXM-+&iIj?U-HIQI?b9~QZ-yvFT+z%@1y?ca?IzS z5q~YDs@M5-enm2pq?l(Sw8O{kvxKT1<%!};=>D|Yf@^Zs&{;Q;eS&#p)lTd9+hm7j z8p2r1yb;HGObOn2`&M6s8I0@XR`0PZ;;oP!rK;g4RZFqC`J~f(U0!A!I*eY>BwhS2 z5}DFZNq%97G_Iu@-%RQsBjg;816eaSl0Q+8PK9=bBy8`qlXoh-i4P~Y27eVZGP-kD z`>Gh%a!Hy#M4~W{yw(`hI$3WM4=Qu^bSO$m*Wp#lth{5pl=Q685TlH6tSPe|9|4mt zf6rjz(=?tX_BGKretMs+A}wFxDugFWT09vFfdu*X?tgl6@^-g+$^6-o_i_E_-A-#Qe{mmK%F0r-Bjwi(Pd@?t>l&@7! ze;gr`mxB#IP)FnyKC%n5;E{X*LTns*GAs4rrOKq`wR?FbDy|mmTDvmbKZ>y|aV*`w zJC(g?HOuQYA9Xq3m7X*LOI0k^ZLA%vBk<5r#hxWmU^#*Yr@w&(3*$!2y?8mI3HMVk z9Xd;x`VgDir?1XiMyfx_Fky+{-dU z9&$*?uqK&vJfs;5$D#1FJT0x^NR<82@o7;d@KqDFi<@6^v$}5qwj2xMW!hZDXrj_~ z>?_;i%7gjtsoKpT{)r+ESr~rx`KosB#vfT` zf)q*Hqi5&6bVSv7`x~EYk>8>MhA3i0EPRGJHjLTSp{{PMfG_zOw8pSm=RuC$igF{Q z$|}Sx^S-?qM~}dpnq#&I8^ZOov2=@e0|TRthqf+kscVld9<6yFmd_+M`v^WrR2`e@ zSDH{w#nZtLO2E$n=;W8HrTbA~p6PN3ps_nx;$G zd|Lj&P08F2;ayqG=V{^w+_CtSGAy4}siMW~Sdc%~mjkEn^C~K9FnFF9ZrwJ*z>lf# zu^{|jVLdVyMHdHF_m*s5u5Aw7xw%1%zxkF7i)z5N`ppv?{;C9$yGb9_eki1sF>$Dn zzWfx*Ng<>sAx^T(PJs~ehGp>pH^zoos{FL4o*X|p+*ams%J_!G0a)=e!H5hMTotQvZ~fBR*PA-4lFJ8&(n=X8&=_!w=P{%8oc*T)(tSgc{XB;# z3;i;Ot?)VbM8U`!wpmL|)wj`Fu<_KpwLa(N&a$_U*@BS>96MEwgEW+9;a*CAxC3pm zBksVdp3avD;UCroEsXdZdEXX@dB*N)J`iop;HhuSFdk-hP!e2Wp&Vx$>~~#wlm3WO zqV=gx;6Alz%c**Oa!H4lA?r3{{7lnnI%U%0gP!MMa+xFPR-2Q7RN5}Q4BV-!w;Ueg zC>74BNAIx>A=byhitz&pM3HlZ-y}opXrkkZS5>#OCS6OGPG5Eo)ddi&*;0Jvza-%WhlUKRcNJ z5a%#lZYEaCYKI21jnU|g=cBv(zf5RxkO_D3~J1fz4 zIbbL2k5}xEQ+6!MQWjz<%?Y%WOEEB_V~JaDa`z%QaCO{E8B351HnOIp0nJ)H+!uYm zb@;QwI$3O{Ri|3l+48FEEn?JZV@uwwd9d6HCY@UY1}~3Pq^O-*8j4)~!-uzzgTG>1 zhbawcvnpUo*?b_IM_!}a8E8m}%AN#%!I{7Ha!a(cORx}9bX43E%=RW%wzY4X7 z=6MQlN7kjIOsfl!K1r9kpW>JMPdKKzVVm`cUe>vS8Ij9P&4UWKRvq7K_cCFvI(km> zaZcrL1>ZwZv+~1rpMK_@YyM~6SIF2$TdCKG9Y-P9G(rtJX{v-3EA_5Lnj-6}>Y;+4 z8wN$qT_)K&d84CZ!33Ms`^Aq>#bwjIwkl_Rn5SgYne`PpWUT25Y~y~DYoiAKx`_17oxAntf6s80U)Mt5 zCec-Nj!PdwInl?;A|&6#W1AwuR7Dq7Z?{`|^4S=}qJWWn6jb}l9F3cregn5e(oZrA zb$BQvk_?Q;S592V-Zj}BBD85uz*`CTH+gR%ZCmO^x$nXi{%SPc>o09)}se}QK?{F6CTYp3<8YRC>`7M zJ^Qk&bRz}F#)RTDu1?2R+m9E$jHf>3YscT$hK9w3^RBND%@;@KU?%nWmtwg~ZV{_1 zS4_NHq9QP0LMSZ}w8TE?1i+_(L($yDPwo0A8h{qT1%FHQaMaW#fZ-=a@l!ij z6a|RU38EHI&Xum!800Jx!bVvlz{{GBN5c(%1=+|KoAPAM10(gd! zA_fqm*rfmhs6b)>KCK|+Pm0jq(Zqw`@A;q+gy&*tAOj9U;Yk0aIM6})1HV9p{>3c7 z=pO$j0R}~~d`?2~lA^n~EdvlhwiV$sKtca9EMVxrlnE64hhafICjh|#Ga~KkU$B6+JMtXqTZlj<6K2?lS zx0L}NxBopAgi$?a7-ToW!h)!pXE!+zd#0n&n~JH+L;9z!dF{tfx+kAeq)$!8gtLtlvmH8B_hDNug^OCd(0E(smHAKY=(qTaCE>F}wER zL-n=KY323TAG+j$D5LdH139IKvt>&j!x($wx&PJui z5lYjAj_TJ)lZ(oL(=H9fW#3wE3AbSN#-r=Y7ml9_o2@CQBj(ThnBuh%KfWD*RatiF zYwBWn^L0D>`t;kSU{CIw>7|%HMAItKbt0SN*%10&DT*Q1>d`0F6BX?dujVt*BgY*2hzx|n|I8N4MZPEbr_hrv8$Ln3>g#o3W z&dbrKA({d43H478#1lAUu$5A)oqy2SZ~5xBbGmOiIh<5#p-TQxGUYB|^Ug*Qn3aI^`D z|MiYkbbf0pv*Fp*a5*@^W#t;0;zu=+sz!{zo#T{lqS5ebD#w1_VO4=lwPd}8u7j>! zyz|Y(j+dU-j}V0TV~Xm5aMF$fgAU%8pBQ#KO~)EA zr=|$mEX9uXwi6@lTk0R-*kg(n=yHe~T3weGb#+gHye*YkUM+Yu9x9#1eherRU)!^C z<*fZ-7^k2z5s}3)RHj~?v)R?nInkNEo_^b&-SWa>;MO8@8i@K1%QJ)}c@XoP>p9IobC95-)r*fO*Tqmb;DK2#=^jBB`WrhD7}OR+4@!Bkv4&Ir^E6DR@<+r zKg|`|xzb8DpNgBt2>Iiot|G(Ktp;k-YV>jM zME9$}p{n^G-0{h=g@9=HN;>Ao}EEor#{5_y);`{kWMU zbZ0M3qFbGJ-Ab$0c#QJU!N6a@hgWl*<=HzIaM>ZI9X;Wx*eR|gXFIp=44Pz_Nozga zj?nIhy_nLrv{uFTl?c^-DpJ`uMM&`Fgk#887qE-O3@o};HM>&%^~}FaguG;64w%#Q zy=|(kXSZji8Zr9i3pV)kWvkC%J5#(*K-YdY&w)Y^4F9Pg)`ppxE9{&8TpL*Lo#u5n zV<4X@PTtll0tj2HS?^D9DnOv4zk4p#mL6#MHJOf6t$GjAQ$?hBs6k3 zPKj9ia7;65+sZ5Vm0o_Vu$%>>7&JYLu=kPIX*{)LJmPNg>Z_Ik{tzGB1I#tjhEu(x zaLB!AB0yKjJVvUlW|zGB!viPJX?#9}k3fa?M4$dstw!AZY%B)+FtWV$$F;%cZ{K*h zo`hO~pw@-iGGVJ@6zT0}+uU?>Kj5~d7NMUK%cVCY(6yj^P=Xs9=FYm}dgztvVi*hRCq z7%Y0x$lErdj{V&~pi3&H#L(n|S=n3OD2{Ll;f;__H9J1QM zHE>O6M>NS1*ivJXaF%#Z5)ANY(i=vf)Ng~VH!U8~^4Rp8%CgsbyR3@?KEMpIX1?}I znjsX}pcjB?c&DWNS@3DqJ%PNU-b#GN5SqC0VoDWAUZJK^4yqQzhMU7Wk~=lY!U96u zO5W+Io1X6Yfi7v{=JedQcuMdsO)xcebTElW z6ts#Ww;~1(vdjI3ZbYZQTo9>c;XbBvmUu}8mfSEsqfNv5yS8t} z_TQImIPgg&6761}!Hck@+Sj{v@s;{8^GK~iFiF##?uBEk0QO3WIgzg$v$FWtRy3cH zjsZz%tdpKs^%YxHB@})L2Gk|hD$%Gtx|eva$mQ8zgnIgr%amz_ViwkQF9}2&mWRmC z!R>QGmKQkBCqK3NVoI$qFs(;SMHkNVJn}w8o>0+qpIc%NYOsTqjtuj1Wp3UbAQrg8 zNv)8lCX|c;!L+VPBJPAGQaJ13#WLQbCD7S=#ik62G7b93l6=Pr<}%7zXO(1L?z~&c z@B)cG|Kb-CJzRS+3SSq>EsQO!6F4HEyaEVfOh+j1y@(AaN2qhPREu4qTJ4O;2mbFa z(5;dM9Ye+so+&b|0 zHCur`Ug@qGIM>f2Z0S{keR0%x)O`7m-ipOK~;uZGReKT}zbj&?5fE0V7H=}YAlAA%0XjH!L;x?SCc zMjC@T4u8wiaFwxPN1a|8|7rhYHgyPxfM{|9Z5iv+9c{JGi+C>Z+Fk1dU4%`>;tKoc zs&22ELZW7s^t=HfYYu0+a>c#WcOqZfcYnA*+ujT>6(en%0Xo*(B6;}JB03~_@5b0# z8l&ImVJ-l8<5Y*~Z*Ghz;`YWlTOPKxEw_yfI?%=vlH!SZyIAdXdS+Ts2smpPc!UZvvaR^jd5m%-f(BhL@GBzT1S;LBb zbxbF&;bB6n0?IGFcQa!6hF|r>=r@NTrx*t5LP{MhfKK9I>GRm44mE_FV5)Vm*I**m z^6NIBIUn=F37~#(blslE`w`ny_O^OSwx`%0=^{A&(pY((wi~m`P7I3LT-pWvn$&0ju}{oDs+M}$KpQ!hs%X3QekO+8GlnI945Wl;ba_~O8o;K zSLEv?htf`d+GM(R{L1XklFxjWs>aImxT)or7F9Q2vrh%5Q-AhGaD)Ly)r+MP8-)#n zTZUeFuM1KZ9_&{=9NGqj2ldy@Ld=vbvPvhssg;E!(}nwq$hQjRwyuhJQ|x3v82ZHmyw@{ePM@{P`Ig1|SN z6C6V?XT7Zomd*%s5P++AzH+|utfmf{M>QHVom-lAOM}fj8(CcxoP)yt{`}O1_V#1% zdif4#d?b5zBj2sz9;B)4b_o@2t-r0M@*ObH1sAN@-ePU4ub-&I&S?uf9RA#(n%}3A zqqpIFb#e7tm_YN$WcRawmB>;{;8LAu=S?T{4evvo{Sp3>P7i z9Ab9eL=9+Q=F~t}&F@{Y2)%%l53B*PeGd~K_Kx}nMSF2QIjvxoC>g(RdvelQwB?ac z^*G#r%CNQ~L4-IDzM@o{gWa_;xqY~F=u?Ariq$VN!^y4o@~KAJ^6>kmnw4muJte7vFQhDuG{pCIfg3s!qO4u%-&T1~DM%*NuoYRfF%&{N`_9zCMglJ^viD5lq)0Gb5mawd&=fDlZBF=Vr zdW1J7T6Sh=M8H%5AR)_2S|J$n!N=7J*GVwLt&2%*laE38(B(O96X};1JKaSID%xo2 zG(4T1Vy`*Std(!F6b^5E_xu|VHNHk*p=xidAe|Sc0`b0fnVx8xgr)(8Zm?PDLapWk z-L#!d04DX@8&Qf;Px|_m*RW>*Z0uyJJM72|(W4yjjIhp-bN6gd0LzZ^1{K2rVLQ)i zu}pUTR~!^4Lzd_&Kql_GCMt#CGK zrN%tn`Aoot@@=B7R2;3pxqM4G#Xk7HwqgWk2_wIbq4Nht3!4TO$zGQ|;x2&Mb~tI_2O#g?Humh7gk$^4KrF=nDDtlm` zUm*EbPIRr*MhK>q0?nKR>)Y-`s_wc}DJnyfO+nbo z+ol@T*u*w5ttDaYUo&qEv(ZL<&1-nNtPKxh%T%5I$fbMMQJ9qlc8&zDytSc#@in|m z!=%esjisMw$9w z?S4cQyLVU}pj+o8u!-yBJn4*g^iy@+P&*^SwhrnCSt{uUz?*s-WUVt~AtIvM`S0G} zbM}CWhnPY&yV6>z#d;^F@9}HPn&dWq=VMfgm+g=mkSWy z;|CXZq5MjYWe^di#lxJ7@%2Is`5u2tMGo7-NdnycxOPD3a_aXEj>Asw8R_I5}!d$}mczATrqAeH)<<#U=Hq^)RvniogJ6>`66 z`^ErFf^l8+zN(wSe2QJK&KJV!WCNnRXyjgu&t3ex-S}pmC5)J<^fPPS<}{bD3M(T1 zKspMw*dX>cmCI3;r=VV{!0k{{;lA?QRZU0fSf@2A)X*18taZDvkJbxp9118q<>kXS zwFgN@(uy5RshO0^aWW|gh(zn5e;TC#>9a;@Z;Iqa4@u+f9k%nM%_^2#>hAt!V|L>> z->V*Xu@j2KCmpG})V)r!CK5+^LZT29Z~Q&fda}%McjziN*VOoxnsuWSh(|;=-ZRH) zGvTqPv|I$0o%Q=s7$vk*%DuXw`2a+7gbRty#(l^;7lcW6uiRQ?%J!`DEi77ObhG=orSoKLvQ+OlEC(D_A5Rk#Iq^S~C3c z{j70dZt?tyE`u6{WWqO>&eKlB&#MCT(FfDwEf+Ge>*ua`vk@B|r;@K@NbXof9~hsi zU$ndR&Q^mX>-`9l0%;w!%UGNg!&=$7dt>i5Zgad0o(rJQu7WJC@4L*-sD)yUGm^d3Ay?br*s3ORHk^?^u zXkX!ni?X~?Ivs50;G$QF7!@IIGy#>HXtVeG#;|jHq1mO1WGo|Y)okJ7SDH6Ou_&@- z_Os!>1$PN)1d@;0gxQiLtY?wId>7Yu;k`Q)#{%0~<^YA%5#tNK1=NOt2I zw@Vftys!t>)L$^0VLMcgv1O@D&pZq7n1r2}F_me7K5WA_j8*Fp@Ag*=WUSTI8p2tu zEyK2Aj)IKH{UVavZI{f92nVP+eI<@a-aJm@UwL=Yh+DYeR8%^{FdPcQpY^(p^)@;& zl}}W&+Dr4_hfin(s=gKXbHm+*@w?njX0f-C)NBX23(E8 zTk8pF`6-5o2*YQW$^xNf#wh|b%h9zN(CRkt0&{LRJsQ)1-C|;c$K^ahXD}Ci6+s)1 zNPr8wzo1`@cx(W+4az68*`V7CzVq_-biKIB3IzKlbZxo!e5SRd{`H*r&~&{zf3e!l z;(kp)yyiW`u#Flm{ugQP`J}P_1pN5o0iRUrP3IjHd$fQi!>6=Xt>WN&BunE?r5)f~ z)dTo3lnb|0f=9dEHhj{s6r8J~r|dqye{B8f6UJJxqq|J7@My>GYJ9R}RERLSv}Yjv z5Gy1<0Bl_s%eek>uOkg?wefnW^q!47ose42ENY4dQBmP3K@z~atg_evrA`xg$tb73 zlktvBiO7fp8wuW#Q(B)>esspDftNgS6==$35)oKfB2Q?vf_B|Xkulh&H zvjg}B5yQ|;U8D?-JBdm;uF~FZYahrL&GG~duH<>Ao3MsFSC#1H-N>ioN9qR(^0iNi z8y*Q36oV!sM|UjzbL%U0`zYnl_~g0YIOiL>**O~}7rc<=8E4`AoSU}5m(#iU81~rt zrGY(4D)Nn8?YKg~T9?g>DuvHsO1|AYWdX12jS$Y(Jo4-I#CJ6m4P{o{l%!N0S&~B1 z@NxIYaH|9iT%stPJn+-SXAG1I=9tf)vXdOe6xaC08rJp$i7-S`i}XYhTWQUFmSbYN zv6`%^POEz_p5l*%^H81i5%b>xrB#^`LadD#vCAy;&NkK!otIhi$g8&ubyFQ8*ql`W zn1~N8?+P}QN_npwUMn{6vRCDV@GE|I!Q`Z3NSI@Jo_6t0(_(UW;_Znmz2N-kBq_6u zYB%bKvXA_LZxUz9K0I5h^I?nrx;;Des2%%<@$D@^zsINQkslsACtUKXtPEpsb%;u( z(O<`L`iaKQ*r@qbY(Og$=6T=mZ?TJHa=$x5h`-$7Uh>7BcrKM|T<|!&U`s-8>x0rN zJS0cKO~Nmzpl8`5Hr)O>L!2SeOSjW(jgLyv`qA9CRxqPiJQdwOoFf_N`g{)kL2>H` z;(aaPlF8mS(_SB(XrHC`Se1h1$>N1nQB5nBzBskBG)p(JJJ9}?EVk>T0(iEqe5=zD zNTes#Co%NpT_3_+qNS!z?N{=8@rHdg4nD~lfKEfm-73+CBedF^E5Udnb;)XN#PuCc zOS#jV!mYw@eKRHVI4_G=Y5D1ZKWI)q1_KBl*p+2aBpgEbE^wHny3a53xY8S$%9deaUtEw=_90r_Vqvb+CE%awA?#{hur@(_dwzOx?*OKu5=tV`)FD@KWyl5 zWg^v7YjMbgsd6$XKdFu;?d{w=LOc2h68I&eLakav{@c%gYYq3}TT%;JWHCP8^o8ea zZg^JxSTERbTekn2cUqd+Ge8tb$<*`u7MRvpo%`<#X> zvjO-TxV8-%Y_JVO3q|`cXpv$~`Q@ZJ7lDe?q)2MK?XC&%f)xTvyZPoEq4MJWm>K%V7H>U(ohj``r8**WK42)zlzwj_{gG(mMp(=q{DToZi4%AH99eQvc{7 ztRaSeMQSi4(rw!$A5zuU=%+&6rOAAt!6T+?k0a_fvb&);y76VwU37Ed`Piayo+WkG z;HpPxt5hH7<8(Wr&wHnHZV1dCOX)V$-HeTH{fbvB@f%RjaQu^!dFiZ!X^q4$(W^bR z8S|gcj8~P`D%OyOd|{syqH(srq?(9V0VQ~VCTHB_Ssh5Fy?j9eSk&5w`sq(mOb@8jy5StKKoYT$u@$#idrid#(!aK zmSSG`>@7I-hK{)C)Zig=+AOK9wU3~@NoYZn5)Or$41Tx((qhzEMnmpI!G01*&>f)) zX3~)!eP2!Na6@;(0N)R5ulWKc2Ce@zZ+EgbfhXwV2k6 zDR`yGz^?#Oh>EJdsfZ93eVIs2fa`{*^G)UX_Un}vx~3dij!5&XN=-s;5Jj=+3v&9w zHy&{SgoQ(j%(Pze;r5&!Rb?(oLAu7mB`A=d$)t4MQ`pFS%U&91Q7!l>hm5#uD}Ax) z%@ztQ%ly($exl~%gC#cbAgV97DQtTD`i*$+u;E-sAA~y`E$Sgar+xT?)sJR!0_<~h z=jbnDrzW~{XcnwM&-21L9xe<<^~Mi064*}SHgx6+!379H>pn=o=-0r{#*!l;3u$pa zT!RKji<}UJ*i=(o_AIU83s}}ze+cHU9y!=q?MpzW254KgnSZ=K-D~5P`owzjX|}Ig z;~EPy#*q1^JkRm}G=AZqd$oUQ{17PouYNC#L~k2M>)J4&$nQ<;aJ10>x0(QYpFdhn z?tlZbh@d}<5BzKQv#J0P$@CGQ=D+ee1diUj{#&Gl|B}y9?$m&Al)o#045e9(!FcCi z5&6G%Q1s^gKWh0uc0Xl()G7^t5IOlAgj&_Wz^8=%QqhHhkUzRT8c~3OkU<*Hh@t;e zMM31J0t`f{Gh^IlfPuk(s}X=<|LV~&w0JJ~vlKug%1{`s#EYP72mn#8S{MZMB51e2 z5dg5D2wK?xhbMrc_tv8eH>98~G(`XesI_ANh(i$cSK$^CM(@4r@0CN0i%f8T0 z(fs!yA!w0XN)JL;t@zU%&EovS-~6n+fes70c_Ql7 zU34Id^8w^YVoU}Obm5JEo1;(R->455&At4YEkx*7(GgUK!mp5^Jm6@+<6Hq9L@*rw zpCTClDQ+PQ|5H-Jhw#746N2Vl(4Q`-gL?pcN-&z6`BOqe5dHX{(jurbV+>~Wy7EDl zDgp>GgyF~K;ojo%_`+4*iX7a>*L0X zioE=90eN{f;UrnUx*t3m%CyQy!p9;bAJ4Ku50iT@&kq+dhCUDLd+Y2VpSxp5&*3!N z+jW%Qa}S|DrSF<&b^y<(Yg>eGe8 zQ^?E3rNhhVW8(#mLe=`p%L}^x;-`L?+7utHHqfp2uI|g_@&2IUvJYX$)1r$o%i8H+ zqsRpIBZ!i{^y1$8;o?Fh3taP1wC4bF9~lJ$*bj^=*t5B+vNH@6^cOD4+d=0ShxGHZ z>J6@+ZazN;J}azsJM?Zp!iq3%(BHZ%RAnc!yUu4Uw>BsBHnzW5n}fPihqpRsa~aJI zB~?di&q&U;BSG8_o7-IotOB>6PXo5hd*AHhgG>xJoKF3>z`lEuV_*1kI#qk<&4o)M z8tF-YUF;vSA60=YnorkD!n`L42;0o7wF#o#o-A<5yj1uUvaoHU>8BcUO_GU=56H zfTEx*qrfRGyVd4Bzbu^x!dC1&hzWdQc`Qubwp=ffn=D6Bz2nhc9`05M7ZLjAT(f#< z2U=0dpQ~iL1XpDZ8sa$#+<2LfvaS2G@`+trBi69Moa=DaYl{ zB=xwvCoY=fjk-~Et|Qn%I>88De&aRMUI&sF($Qgm z$c#ZMcJ?JDj|Ad@Qae#o#OEhc-)#>^wB8Gi_epHrNNcxi+ zB#Z^{88{ot8GAf{@A*g&GRNP?#^K`+|JrJGq_C~gVgFrn%?Bw}a|+O?n*^9rR>WS! z_5xmeq_XAdE$YyW;-(L)zgTh&@oFGWdi2?Fb%joLT%Hz-n+&$*m=V<3Y7gw_4eq{5 zFf4RdPz!T|3HPLhj1hNdT7`wkN#?w5IG|zvsA$8MJ3SXyOWxmyhyPg)c^4Z$9LO%V zW!BHGY_G3(Q21iY7X(rW6F&G!ig|68QpmKJmLo|BeXuNeR1^W@h6|&8Qj(dJE&a*m zk&0{dlj+PB>b?}yjO6OA&WGmejvG|LZelWW{@{+W*c|%nw}?3_-;)u!$;kT4kV&cYT4}D@3*AgF#wb zYB}PkrsrRl=X@R!;moSovEVe+r}%{ic&N>FO-cZqlQn+CFRxb9U#)tc5v0mwTs_0H zA*X($I2DvfNj#f2MK#$tsy2R0Xx)hZzM@1i<`pDnyWMZZr+2sxUCEWJXHv2fb|K-{s_vs-FdQ`y!?+v|5E(IxP)r z_cl%j0zY&B4WX1+m@H@+OuO9+2ls3DYljKq4gt^I11r;f~1L6 zx{zc$sOdIAAQ)x{(3b%`cACS6&7dc_mm-x7J5a>OGBL z`wgS|X*EgHv#%HD<8V!ILQ-9=;x}1)D9^h#J}h6|^V7J<46 z{c~wDzIg&BZebx8)CSn#aaz77dAPQ1hft|sj>l(mto`^g5JWgn zfPq^-Dm0g0Q4hi{un)% z&wY>adsR?C0jr4opCh_kCSi}-Lzd-GPlVEGTHXk_+U7cE7VinlWzsi!k8Umv)!cl}gge|Jf zNu~G%4EO{%QnSw4Pv6i5dNWqg877_kO4iv%cmmVdP}x>2xYVKH!Jh;mUnr5IeKdGY zZTE)&0*!{p$S5|Qi5Nqos8L&X%-Ny9&UtZLMh=!$U*Sm{&Ni9ra!jR}{tn||8)&G?`|Zg) zt8&(mciyTra#qA(8W+@DwaYy}|8eoVCV2gsLgu7wRLTZZ_{Np7q51WD++$DmInv<_ z?wGIL83vA%yK`{On%qvrtS8T0eL3=5m?OMFG|n=uesL8>CR)pc917GHHeI}$MhLNp zRe0m~tV~H7-{(yAj3%U{+FYB-T9~hjNE&vf{MrhBW{gV;^wWX1`b$&JwduNYaOKP_ z#u1}2#unbUP59zD>Ab!X8_1S}ALzc!Z#G^|vkmx8ZEt_k|4gvMp^XjFQ4RMdISY(# zB0jK~r~vn+Y@w>S@+$OK(`0B7t)p8&d`18axjpdap4>S-2x?}-pwBZ|*yj}W7h48%Pxb~W zquqOouT9uy_Gi5K?T919rhPJ~RGun2bm}n$xZ8ruR8G~Ph@TG8>kV3!xa$GSCYS>% ztl#ot)%DHYL9Y3-aDtsP(`tlF8;jH7Y^qSdFJx6}T?a|TOa1xK8ZfLiF8R<|pO2qn zkuaWv+wbFw;h{V&<&@FJA?c!?BWPL}p$)PR=N`-p`NMDgxCtZlSacJ7zQ3`VQIof| znNjTimc|8g{Uq`u68jiegQsu#3@jZVLof_fXocMd`c!w1Tj+_f5-Kq;k$s;A6)}U- ztbU;~T6o7Nlopv-FVPvi*&f?qTSLXzII!kuvEg zG+${DFB%|pWnCdjIzVmvV>hyQ(nZr#^&K}_{37IT5fAS^Zm?P@amzJigLi;uEdYmH zq)BBeP9W1-r}V-BW{jhy-LY(z-S!1AU+l)Yx{r}CdZyNfeDR*4Xt|>|5fv#p!NdveoN==@aLVUf}~0*W$t6EQdw_B5#*mQ zCf5x2knpXX zJ#0MGb89VUrBtd)XK~BPv^howU@&_>i48Ns)HAFtDBRuN*y7g8SWTt-@i-y1FeVc8 zo~CCQxTWX@zZ~3(wK7DZ{$1Pj>FiN36AkO5-dCFOO{IyB>gv}!bhRb>9#P$Qo)HGg z$JlasONCr*35hX$R$+r@Ee7@D!l_K?!OtX}(Qhfg|G4hSbva!=kBiXv*l9I>3r}Y& z6Q-CZC$u5$$mbtd*dPXoQNwZ?n_9@In9PCPqI%I=TV;&TKgEETPK+ zCzM(Jl1ieHFE8S0AZcJQ4nFLK1a5KX{^ij!EfRD3KQSe!}ypaxIVbp0v9zBj#FiNO!uF1!&U1e!f=| zDK-{ixbazPQ3$43Xaoj2NpCsr>rmUgrwAByZErn}?TE6UO50UV`@0=CL4@BayhVWO z6oEyYXmW({Cu93&b$BV4HoE~SRfz*1feP-NX^aVu_aEvqMv;mhx!3139PSEhTAzLq z>=)J$vx?u|ub%seXD@u@foYf5tBBgLpQR3eeOAw*6z_5OO%wT1ewd2eO?lX!(e`k5-t0U-_i`n$OdPR_}FFPS&9LntIGw+P61y zA&PWDKh<;bQYL=TE))p`&h)Y&eV;7WpeMkjNfVZTaN{)Yx7y#&Lsg!Qmo0k1SWn|}8bHb$Tj)0D6%^{0ay?WOfc*&^bL@(C;zvP!M!0N;uvFk%=d9}dNk+{IPw zND&aq7Yg^5cv<9$qey}t*g66+u@}wlJ&yUt;C5z>N*K~PZkQ)DH;A+LHHSYMnvcnO z&6C{v6h`9&7!kU^7qfAy+J%C@9I)jLVk=HBUAgzvOAwDF+gP_FFiCLoXQi#P7|C=- zoTtI!h|jIoa@zH1;%)(-pHDx0?MY>Gt1dInX!Bq^pVu-V-bDm`p9(FZ|7emTM;?1> zQ`nlipJCr5!z=F3bZ~74FJ~R;w}(tCUU5(oYzDJ#=k7nCL(F(lnqumU{rpDfy+eRR z-x@_$Jf$|4#vBD#9heD^W{P$TahDEfEJ_2EyhR%_K6QKk^}A%m8umOUm90J&XBUoT%aAQt%9Gc&ala?y@F}8_^+uWP4h8O>rTW$u-l7L*9Y}3D#CtuW z&25D;hEG2N;OcLSHFjS+x3Q(g@16y3_$L5NM&?pM!xa-?L3pVqO}3dxO!I0g!=6@a ziGs|{5<0IL)PUV&b7)oqSNG{Hv^hd}sPY(2+6{ei_I7rgTfumZLY4aqRQLs%3t8EP z{v>WJz7qKkEKl{7SJsj|mndIj30fI-vXya|daBNl1?-ShIob6~^$Qt=(9&1)*(7^a zLVOKD-F_Eu7fgc6m*U}YE=cyhI%bU{T))K`txDsiH28OBVbEYwgB@$Kr zM9x)P91YOq6)$Ond|`2N)z+dGs05WuA(v}2FPJfn`X#-Jd7bRDT4$Np#l7U72l$TxJK#3Z(2f+V9z#?6v5@C<8q#r!`mg+=N#-5L zA>7zo%C|AAWs?>7IW)G(4)CeRLfy^dGk1g*w44FCgzPlfjFtl2fFrg6A9g2`0L#_u zg^%IWqF*u4Qx~h+PsoOAHd@Cp8j`|Yb$X`V>hsaC8pcgfoe^L|?$kS)?18j_x79SD zVtykd1GjIx)XHAI$2}%qXU3C0;__Tyyh={eRWqIEai+xS z#z?>Gl946^f2E7}d9z<)yWiOZQbPE= zp0aU)7z3vfG~KX&-ZP`ax|GN*((Y_5w@z-WDZ>8R!)P{G^2}93$oU|1;XZb!HYbL4 z2YQPUk+Y-zZBZuWycMk(N5PUPf-99xDD{F==TC0tQ5+SanHbTuPiM}q3<IvA$flHV(?W{_sDpopue^uXof328jkuTtUb=Lk% zeAc_+I_K`)XlpPB%Y>95pH~KsE3pao0AdK|dSS&r-4+{8JBjdqq#Y=ESI9EIkLLcC z5nO6j(0PN&9z&k7h6HDHGkhr@ypYa4*F^o`WtP3zEU3ds3||AxVxYuli)H}8;RbUt zVUFO42;_0+pj+13V1 z-A|s}bAOY@q$=`qVI=`R-245XPtgJH(p`P^nL-lI`_Io4!*_;J(NCzQYmX&orm#2cG<~RsZm!g=rO)=IK%-) zFy&uXw74+p3YHag!<20>XFn1yPB^k#U+AKyJ9@)tJ3$p{Hp}(k2#3QAPmvScn0z3I zML77y0+Oq$=rN;13)WkS0D2IqXG|?xFG{2Aj-ET?)$CBQi@owp=m-l68vF>L_C?NH zOc70#v}sL=>eh>by_&{)6_@XkNn`=;<#Ru~p1w*z^44r!C7_9b`Ih(ylU1J}zh)Eb zJv|g?%eNFAtJy+}Jzhc}pTv+_JWv2PN{ju}1bV7oj1Ln=bGry2fV{>zS)_6s7Jq}Y z*oMx?78$zEH#KFM_`U~tJu9nVwttR(a@T$#+dh?`0Xs_{&ba&eLy+j{GY<)iolo3; zYmYDR=!!#hv|5jYs`|xVLW)u69dmh70(Jb7pLSqkU@$lX2n@JI{nh0$x}1w_hd?X_ zY8YbX?&Z=kz5H=XT1uO~4NGl7_Rfd$q(*yp$bCLd5 z2%z`RhyB9)NO`4;KTTfNa7%wB7!SK zuj-9Pedr4EOlmr^195AaX{2yTg+u`~&{-wNz#6BC(HS$=qLe(F12tbFTl;iLli$T! z{@XU1xq@L+`3D9``P8Z2iF1_Gv|EtL^;lucM-I(r<3`pVXY`<-goLz(7E_{XfWGJI z_tPp5XtnER>+~q5swhr*0fuM&LtS&MS%yyg1vvcLdi+$*Cv!QEYI`>@;Z!CeATzuW zYObK?(SC0oIQ(3_aTVv`4A%i-(?kT^nl<#KA?pA_1;QWlxGeaFN9bA{s__&>nN#72 z#6q*Lo~EEghBU279R~twS#;?vJPExJD*z<5=JtJ_lP`iGk$?lB#R+KheDD`2MFz#b z6wk-|`D*_KDj9uh;kWMz;-{kd4>Se9`4^h<+YIs77zKco$-oSUA|>R6lpXo&YWY7R z3V@U6&)bj^D?H$h6DS;LP6$8+(jJo&Vi5QbSOvfd#t4TZPq>o9g#kdYA-~%|^YDSY z`k`;;kgp23@94p-XT>lt-MEcj}*tED-)j5Ev;9@-bYHx+#BsEC5Kqza|wf zh?VB=$NXzjiDyRz$45coBojeFye)t~){Fl_Kmfphog{y4BS21wZG;hgk^zMZv5h1o zU;s$K-!q|*p>)A=`B03g|D%9F1_O8JLt#S$`2SKsCabq!AwMlW53;?yo>%rAUf%rdHG~jA4mZhP1fR|>I1FyO!Su6C-uI&p zPlkz16Bv?@Cxxg8}lXov?HccOP zSR33ns6VT|OL@YCjsW{1!No(c2a!Xc`>c{hIoI$aVFb~GiJ}WRYzf}nqOFk!bAiqT zVW;7%#f-SRp!+87#{3YvVMDbcjMl-LvtEw04|6(8b0PERZ(bc;cqj)I)G;W-ImW$L zf-Odgk<3tg-jAX`nJyS{LRlQ5^xtrW{S0XGoh!KKm5G|Fe2YwDqu)9b9WCMqLt*S} zx`dEN5J4{KU<6B1OJ>qroMck3l?hsZyxlj`$)7t|*@e$kRfm1S&?Qjvq_Ad&<3)C$ z3>L+-X5>`dsMB_!Pr_3evr4{2+hqyHKXu?xNpTm(klJ1`e(fb=>b*oQKY;E#9TdYP z`OWC)4AC4WTR-wkm#ih8^wMxvMK2Pn`f=q1Wej*l08jUTbwv0*GSc>TL50yImZy!5JF`lpe(Qeiwc7Fpp2 z1~C;@cLB{XVZpO_D9oa%Zg&vGukHth$e~iLN%jJtEsZKk3F}wUr(}u<(Lx%~mEd}t z2J-69e#jZ)ywh(Ie*FTqfXqp=Mv+pEw52Ya_2~%Sp#LzIO;Q3h?u9DrrX$E>$5C&b z*3;FY7aerv18fqlSTqg8)IYr8qhLS1HhOL8YyPR1HlVAI78b9=SRAzUrJD}l-Vco$ zgRl!K3!&p|*rPo$!JGrR7Bd$mU=Cw$EF*;sVI#9|FriWC7RT&$j*g5m4}w0WcQe!F zPcpK1uB8GkCFlpF9>8TVeMEa&eao1fZ+Vb|$6e|C*CSSx&%L`892oximlk~M>83V| zVrXB1r*?BzdG$m=%8wv;Qv|V7vPzuE0H;-;{TurXdh=bQ-jBYZ0LHi+UP1H$;zH6$ z80b7dcfcmAXdZ@tcP*Y&I|0I&Rmy9Qt9Na1juay+M7{7J>XQ=ZAA;r zBkWC`-k~^dY;a9U<7Jr1wJ|h~(mS{X?9HSlmaYIsIK{}Xefdb(kt7%sw=~9j+%~jo`-8t; zxAy*A)qfu;=j0szUA!k==_4%8%N4y)o-}NJYM+HFBT{5ABk?INl9HXndrYoTTa-MQ zP}@-qneZCSz*R0x9z0>2FAip>CgaB&q$!R#<5+swd~vfNi)BP|*+{1G>ZU$zY>cBw zbQ^DIbaN`BSb{RFZ4$)dzPrUFz!k%)a3t3p&dywWG|=2hl5L`9;M*km1~FBy)(fBX zAasT_*$@Kt6ZCrus2!#$>s59YH^9#6?rmzeG!5y%sDr$wj9uQ+RK!j77gJ+&Qsyd0 zJ`XUMLmipP9D8Z~beMS;Gc;mGg>?Ajny9HdcY#5{M1|e0*%VfsD>6(G z#OaYb4WKmSo7CZ|Ccv~S5RG>|aGUGTxFwgIpJWDMD-eqz#de$U26AJY;=kTxNAj)f zn`WHfD6}V)xV>|o82&QYK=OR{AZpOAi$F=hdFK#W9 z+-W&b<=b#gDu4F>wUwpW43Q|0R?pLoM;`_d5$s~W3!iD!F zGna+F-z~aD1djtMiQ9_Y$l8h`F{|AL9#`Ju4dZys4@8e#VlZ6l8g6#!Q2daD4K?fa z;}ga4O+yOm#xVKvgpc}Tq%sjH2+qHgPiuq=G}FV}PbpqFjF0Mzpb^d)4ihBTg!hZl zfbmE9D@iA;S8w0=xhEUR+Lwa+879qEB9#X!Et%XKJjJmiu8>?3W%`PTw2r$7ETD&0 zZr~zLF>TkLg00W!h|x`YP*Oc4lEhaNFMam<|{5VYlQJR@71Hal!c9 zpgwhDnASLr%B_3-5pRh)vD|a*Fg~+8|5$P(mY-wBEe7xm@lFb1KMXoWb%LCw!DqJN z!>A4s1H*T2LQ4;Hhx~D&E+g$}`EBR8BD30J%*PP27%sX#A14iC2PH zYnlTW4A{hR6AIP=_2^{2lGQ^eKTE%4w(+HOZQr+3xp$*up_*GSXn6h>1%yC?L zROu6kkl5p}+Qobda1d=Fn3C9|B)GA=7VrD#T}?lZ_v?|vc&c>O>Qb+b6KTTv5TX`<1}tOe>WN2YDpy!$ca?e*WM39KE6se zo_}G27^-GkS_HO@p`8wVd_~G}*geM_KHW9G(DVi`Ln?wc2DbjTluTk^>c*(>?&pT8 zaNIKmogx0U%)kj6-YF#h&Ez;M#FLr^XDqTPhfD5&wPZ!YJkY%_>3Q`ME(~QC*)rcQR6Q9+zbbE?Zd|CA#K_by>u~$z z(TNowA}M6wyclMJK05m7wO}W%vKNYGbZ{n6j(zku%vu9|rmp4o{Z^W_vK$fWDT0jh zi_blbR~lYtup7DD*|aou7UW&MAGFt9&1ygiJGwZcuY31@<{U0^D1f|%T<54A#T z{lFRiHTk>fkx}=zXnLON&}~GFDg-g#oPoYN9F{0*`ROav!LRyab#Afbn|FGAy{Zp^ zsnevw)lQCX=y1K8p)UEBuUmY-^kcKT<5i#0(<+=f@PMuTlCYc+WkJzU&`@Xuq$95S z@`^h)1)k~l$^>dsep6bs1M>*>Rz}U*YP=4-6-4!I%L|L>Dm)SXO?-YF{Sv7eDz`uxp;f|P_@SVOQM^l5forRumePt$+ z!av)AggtWf(kIKMqod~AazA`RPN;iK&`4mkSb&W&rVXI3!ExZ6Y*^m~B24yq3JV>- zWV`XvOhKP-S%GPcT_0f&qPyVNB2kkc93RPk_j0U>H)B#?ytjREPDZn12;6}@mS_B2M7?@;!MKIA&UkM8)e~v*@ z0>zsVQmRoSRNKf~vyURdCnwrJLr6YGcA0aGy6u6?S@-q68fv#^ZE_4e<40xn#ika6 zl4Z2C8om(LZOR+@D`@qGin<13M^JiwOQE1%qJCFj6TUkGJzKUaU*GmtwxVL|lC`{N z0~i_mGqNM779JXDjxB6McgS(%{VSSCoZvSEo$p{Mn=Emku1_^|<$851(oh`g6--=j zy3iazuIt6a7zZByoNeYS&8NxfLQ4@=<61ncf(d&cn)OG#w3gUj^moumr{>;31;3^%rNnD$6Lwmcuke= zcZMjB974CA^T@4$#P>P5edGoX!Vd$9+MEp0M(-pMjI8wt7s29p*tG$i zS~9wuEE%r!>5C(LZvnLA`5KW_I`2aV4rjoEn)Rt8w@=A}nuU&{<}z6o0nq;}7a+mp zpDs5bWw(iEk{+h@26xC5E0^P@YSGzA5lPQBEP(h5qx%|IK3Y&%dIu*F**zILi>Mkf z^vk%M{2(Jr8j2b@&m{cNOx5yJlTcGMPwqR#I=(pmo;21AWQ8EoK2%w$Vi#1`FacRYa>0VV9hr`;-tGOQgTf%19xVXsJU}mn zN62t#vRb&9cICS=-L|}$rtq$P8* zt-JZ|9_~ppUl(y_v$u^s4d-hB4yZw|#NbWkh`B&YIYWsqxuV>vnnJFIGM~&{HDdDQ z)Q=Gp)gx(%^$xKe51wv?W#3)sUz#XpmYZ+B#fQYGlP7o=#61PaB$bdTrJjCGcc~{4 zK7sreYfm&Q*3cJOqVg)g*E+#v``uPou_)1c^cw?^@zV!2pz6^KLlIs9$a4dyHKS8G zfpYE&R_v>6-a)Dfc%Y*2!@YGHBgVY3@7E5dPt9o*?vS?zv-+EARlmGTR%VtkZ-{gy zHYNydh{~II|INogJ^PEA6lWWX`1rMV!NIVkM9QNJl^)&7pzlFcuiHh6t;4(1VLpJ1 z>*WB78PSnAY+)y7w^Y9{D8&?)8svh_mrjpVrIz0n@*(zGQb()kv$WbhUTIWQDa@Na z9tTUgRikbt3QQ>sB&PfCLv?mNNW8jt5e!cu#Zoz2vT~-mp;=jxSoIy->8JG!H9=Zg z&b{9cM!xG`MU~+798s5kR61rR1enr1T6EwYEyd#EEaA1!$l5z|ffha;Se38vOXVp? zlb>F8xpIB__ku7zt;OpSJ+RD&ok7+VJ#a$&&z1a_6_+W@r6~Tpum}}`+&w7^m%_Nzalmm6EuYwk7bx-BAx-soc6FocKne-PmMpwFi!~s|n zs3IHWI4x6cFg|%{$9Ia-@r@;}Sp&ChACrXXO97OyshVKH`o}v;*?6MP^ zX_g&ci)jnBfA2nM9wjaRG&9|bpg?f&_p;Zcq%o;~Vze6QQPws}IAt`s#)FYd8+K1gyJ!P>kY9D+2lEHY(rUtDHkjmRE!1k+)Bo<8! z@Vfjsp&EEsMz;CO?BnP4O$TiQMXzRFWb zId+y&yxvz%UGCT{iIq4ZrW7(3GkrCw4@3InjEf$Thu?UEYNjh1&t$Z^YglrUl30GOO=L^z6EtvOkkF=Q4rj2;(u^ zUy@`}u+hYhHlA~i-z-LauWx<6M@z-xB)LnFw~KE_CLB@!8VIK=kBw(51g8OHQ}m_{fvYn-gW4(hz?Oz&9e-bm!gJjw9E6ppzU9kFPXK?N z|MMCNdxiWKGukTd6HX*85vKae_x)SEN*(L_hX@1Q#V3RMIRdG{fsV16{+06}nbn`G z8(_fDK-wXKu3qt+ABVovIZ9GmLy7`p84+A10^rab2YcUp!VD@Id1O?|#X{x|WiH9w zdPUA%frHV6o7ag+A5{3RT;H%PJjdevBlbG)gc_%1OZ$!S51G;Clt9aAqN0?|Ghw^R zO3Sh@X4bALC_8PZoTejy_cOUck(^|cauo)TCqTIY_%)SAc*er|tcZ#y1GEeSaOEI# zD$W2m9$QW3@u{HP=Ylu6rLERmG&U6~OgWA6xdoYi(&-q7qD1k{E~R zrs1Ham#Te+^Q(AvSZ$K+L zHS;o9$BIENro1B9H{VnE3!A>X?8~fyfYRI{MrlfdY>hr*YcK=*%*6Z813LE10oQnj zG5Xg8>isE-uEcAhOVHH7dadWR_b~Q*m-7nx6%o#DOU4*QrNYa16QZOuE6&Ir?;POI z1t{Cs(Hw@alszx%PNRY@+5#MyxO9vP2o8K)(9~-7d{o7jI3$sdP_aJ1`}?tik_h*0 z;69ytoU+0Gd^G_ym7+DMY|{( zBFrca3=pLc!w6^F1GCNe*O!xeP+}cs{ga$3lX{O(f9vsnj+-AWE;^Q8s4ns(7j-dn zwP+;mq_i2&+r&JW_}*Tu{LZd$YFK!>UpH*a*8In+no7~k$9%Ht+YX=JWbpWKi#t)q zj!a`4Oc8Vc-a~3;=B=%7T&LN3RH@7Q9a~7z562 zP)>_i6o@?~e4>;zz-Th!8t>NR$|F}z>Hh(L=nCTuJ&MP`0T1I6m7y;C!3F3r64AfU zKgFS96s^<6c2S1`y`;wkvfofjQpb|uBl-nj`A+>b1(&~{P*oSNHcLW%5Jq-iBhrb) zc!}eq0t0nT4MaQ}TwY!#;H@j{0V&T~Jb zJINgHulg0EVK%GqwS-Gio5=GWxF2J`dGfsV=-J$lJr*W-zI~)EC7f1&eW)0co%M_! zrZ_hIulDy}=@0Nbr2Cttz{AbO{Xgjs2=4BMB7^V{fNWg<#=id}xdFMrIsH(iU_~@2 zJQx6w`_C&ZZXg8x|KC5sZ~1z%h{!2m$b)g!XtOB%uw25Cr)Cr7Cbj zH-UYd+wABkSNzT4;9g@4Pz^5K^7sf7AO<-W|vb7FC2M`G*?d9Sy-H z&`v=D&5rSqBhV5Jg+dDChENv%RT2N6IsrU?C<{YSe+ogM1G)eE!CYX@F(`IANY?$Y z=m-SGEjj)@S&*^7PPmYq zy-^(G_;CG^NHgQ0&|X8p4}XycKp+QLp$?kjH=w``tP~HyAwU{J0{oHjP^e@82s+|_ z5+x*s|7ZU=AWA0(1QYRxQ~{Pu_&xSN00>B${r}SQkQxpD2J(RruEqa&4v>lkId33y z5=?}|+_`uXzMimyJrkjj@qX7p__Gtg$LEHK#{ft<2fhSHwpcK8(x3UiWo83gB|)LS z1^^)tgum;CFJZJW`)?Bn`~z%|3{imr5H0raT!4_Y|KD>800RDS3&3BJp@Jdv{*%jx z928VYs{j8;2SCUN{FBRvlw5$+Kln{L_~*!Dq(F`kcLD_j$03*U|LesCK|A~v7z`$|J#NOvJn5W zA^uGB59b4p^EV^}Jk}4z_h;>razfS--@i-nzpRGen2>+O<=@~6@P0Z($j%sm)S)1v zf<__Ygp{iQ{O!g4-{)Y(cTfr7^$bXJok56#O_B*wusI>i;O|@@>N6YXzn~6I2w4IM zHco>uCnhr?J@|(pJs{<`{@)&4kkj(lMex5KvEYuZ-vaZkxQM?; z=s%~A698G>fA$_4;s|IThDiAK@{r~b{KMZphY$(=w`MsZ_=o@P`Dbw17l;@?o&|;V zk51102Tkz*=;VJ2$3W_5{ME^M0Kf4H&^!>ie9AW=R7U^=;Q38W_D&2CZNigHbWf5u7T4P#pQq8o zm*>N)!AZlHyZgh<$w@=Q@Wx!$>laVbm%GjLfE>5uwDYUQb%TyhAJE;-4tVti1adll z0Idr?etkH<9eculzB&{!@cABdUGwtea8Km;`FOo*=aBsK%T?pz%l)raa*&Xq!y`&( zdtCtY=a=)Fr%pc2%cm2O@CXob=VK>W{nVdt>1v*WpgLGOmVYW?|oE3MI_xMkP7 z{svOf+U=zfbQqG$%HHWjdtXSf)>*QCaAT?OGn?xmsCc;E`C*Mob0X^+=@~wHJC)c= z)jGraB|RN)=9pZio!fQg^GZi|g6nIaS?sE{F++^1e#KMnjYL7kUe$Yd9$E|Y2#}7) zMwsT!5qyU*>(_wW^vOQd6zunU9+Pol^$QQ_N2S|8COW(KqL#=f!--DFD1V)()v)8i zLF!!I1}z1Ll>R&d``o!%Y@hmD&y)E?9q&!LoSvEc;g7i_r;I9wHaou@^y)71)UF5W z5A{!q(5hvw$<+q?_&@@WUb+AQL*EYsktS(9LZ)_sb^+b?lOC@TshRK*U{|iSKsE#{uS~9q z=R$sn0$QNR*`Q~QUcrtoY~lA6zV;D3JBU2WkIuwt?`}+-;a|eFTrg@~m!y;pOz3Oc zN*;|-#21dc2x+2=9@odcnm#E*Bs7^i99tczNojJKJJYdc^TSrc5685L^TM&FEa+J> z#gjcVi!f$f^_evL8VsN?11XkYjG&^PSPTqw<-KiKQHF}Ch0ZiT-mK9=3g_j46)$-( zPvOcNrpZN>&C#+3i(&}cbhy2xvoL=3%_OWKh^gN*(lTbzKa_*KJn59@)3>g9!y}%& znUrmos6(cD6u+?)KtipGb5M3j8A{f%bhvN+@Q4c>N27WoPmJ=#f6#x8-8dEQuZ zj-_F_Fpd;L{z_VHaZ~g65~sbh=p}q?#sco|E>WkX_)c-I6tK3LslFY#dS;B`H!a=H zZ3XoWTWVfqH#;l5sn4t3OWd+$oCwbqBObuY8{(o$}bwCu~8}`jE zA>FZbr^Le2B`w_@(p^%LvUG#=f;1?d3P^WJs(>INje&HCmeutkuXPUs!xpx5F{CJH7Vk*?WjO8tg`LR<@ka{G5tO0n7DMURjsD zJ-6=t@A{yWU|Xsw_f_Hi)}{e>v=-5Oy++?XPSnk4zSmO}+PWjTkk4pi1>k#_k5Uf4 zs{6>@+XVlZ9&yqy=Z=n~_pNYKwU;&B=5&kAj?{t~S;;!#bK0tfU$b^#7Lw!O_SjYL z-k1Zoe$)p_JasoRF|34FY9Vx1`#D7D8vGv-9j0E^WdAD`mZg z4@!kW12!_-z4LDqAbgq*v0mn6kF4cnv46ZUdtUQ{Y;E_(tzX3OjvMh^V=GT_dPK37 z0@DZhrD`iNdyT0I88NayX3*S*=uk11lcQyVf0@J}8jD0Q`Xd&c%Iw!jY0qW|1bW3~ zqCP}WYTvMXAUKTalWusW@3K%cJ!~clwSoEQFc93Q@U1p7ZYkN^jr+cp&l44s>j5>b z)r8^_ym8*dR$UZM7jb%0)wL${K~uR_%Yz{WPL;FWvfrXf>{IxB>0v{QjUgIsRz9Z> z=mobJXP(IgKiv@CoC3eZzI9>ivqw)#NDki)y(IrMnta_)ml@^vgr8MGXNlXMlgS!S zsFQBIFbqR0aIP;8~bWK5&-TR%h!aUoi4y6*JrR!@UTJE{#w{K2Y%=c-=KrGls!_l9{p zjD4vH>_@e5=O4vKDmT8a=e$2(?h$p^@BN&Hysw$m-hfzacK#CUwR8-phfvrPoe#{U zoJ21~n{gs~cOgykkF95KTFOxR6>%5A{3Y;YGDkA-l$~OEqO&}!Y6{x+10#MGFGr{= zoStO6ExLFghD-8fHc}|M(Chi{jO2;Uo`a8UilqUd^U>r3kxs$a|oHw57+c4#7B_*zS7;Y@Jq^ocgk=Lg^=% z96i4yq{ zarw3P;+d}o;g5GdKI`hRKs2)<-Dh@5CiD2p)l7G5>J+u?ASQ{g z()%vK_D+1$>!H0O*4gwaU%j=?9#ylsGtFM=*7tJG8Fsw0FX5ifiL&MoczGb^>q6u| zl={kkZKD}8n$e4S>Ixs?b!zfT!=&4hN+T!6zWFwwx!ZAB{4#tv^i$8vdwyp zm=bo6UI;Fw=RV_P&_gb&#DF&8(;$l_Wu|emhs616J|(vxpeMvj{(|mi*qqRW=QDWN zJKTYutJ{Z^WPFNW5gR1P&avhSKdtTDnxKi-AImb(t??PUW7&M?k`9~Xo#s*Q7t5GC zxR(^fpiAzZQ=LRJGKvT9&bjZs3rylQ#(18N!S&I;s{VfethJu2Wzx0;v(6Ca3nyv9 zj^PGB!iA51#!l_@a?$3?_&$kp_xGaQu3Z%3##zEK`I(3BVJDZVO%)$?0F*IuH`_;ha1Ku0LkJ9~wFc7Lzy`oBF+B}uE%iEGYNSPkyJ#7g(^0o^YD5N+y5QSD& z2vgNJ<4s|jK+`Q$U*_VW`DR#7uj}c@)2jWB#sKmMiZ8aGe2>jR!TqXy>CTmsL>Wdy z`dPM(yaBt@Ayg;!bd1ffQ~=&ooQb~0L73?hFd5x4d@KDvo5afLjbp8Nu+K&5(!E{f0CThcVbzo{Sw4#8hQ5N8nSfNeXnHa5UK=lcuD5&DdbsPS(_F!sJCx zBa-ux9b@L1750<-H4QjK`y^{v#k%TkT=;YQEDCXrA!LgO(A>zGAK1@&sjJL0-bjfq zk$zLRdO|I`c^%blyebx|!88-Bqrhk}nI_S^z-hMbt>>hH@; zn3trYwp|Ye>D4`e?>+~$wQJq>G??ooLkEBj};XAJVWm&s`A&=VEH~2`CxcQ%Q@J3D{2roN)8g~0DX0U z(;`o9fov9b%_J1o;Qz_!QnqRKu4Z;`llhfiP+oHhQhM(RJpH4Hk~?_KO%#v5#Y>0Xr)c7KwbZ3| zYC0TR`24pQHK}ie{4cKBU5QZejy0O9N_A?st6(`%VyY#=8rDHW=P_ry%gKupjXXT} z&M(Idrj*eZ;9<+er6zRFzE7fE?AM>iGfqWW4^NCrmpnZ%;dEljuo9r-r1{Xhx`tBy z^~VH-k5d{#WluEJe#bzO_Zcv%z0*3dsYlcNDu&g@w3h+@cBPOJ?AJA1Ss1 zDSMY%s0wf)6zv0AoisZZ`<008;vyk8QI;2qCA7X2_|&$^NFAbRQm(s)p}QN+!U*Gf z5IdY}Elz*SMP3LAsoi7PatDUA%6(4}8L8!iu^bZ8(=W*+EWY(IiOQgq*-{i=GVu{sjxmJ7 z9Ms-E`oK)`hGu;+b(6zxb#pD*{z0~Df9f4UiLhjklEX<}_9yXRQZ&a=sZUXI1-oC_ z;nd#?=p^`tC+qJ*oJ7-oTeQuq7F4dBPn#r*7GGv*=SIA2Zn<-00f=9VpZb8Vkr|dW zlN4^V-3)OMw|RXS#JvGieE96RK6}3VRVhq zZ()beB0W}ov2^gmT3xl;Nn^o*opC-}Rmbt7r2W9P4>B5P|5dk2xZX>VTx`6;*m;J? z=>^T6aQLGYTM?E~{b^FxzHMKzmCY+`$x#EG)OQ-oNUsHqS5pYx1v6#owx)HKNa1~z zVvP@8*!B`sF^a&Ap?Rqj#zIP8<6PKCD6P&w6Y0$^Wl_pqy9mU z=cSg8X&q?j%==k0PTN-UN3=AVJ9RaZR#N*7K^7f*Jx!cVU7yeNjKQwhP@217nk*G1 z^dx2tx(pl{kK$}K^zn}-ixF=WI=!wvQnf$DFX%*t^~x#X7}6r!m2z4cw!**EZpSOO zR%WXS5E6I@?FoN-IC|e{Yz!}Tld);bg{$uqW7F=po8fgC#kM`@sRnkeY^pKz^)wcn zH#_#Uipf2-MT((=p?6kiXO${OAD(>AjxX6gjId4~uO{C{*9zy+1Uxzv88|7ZX?Y7F zE-94=l{eU#Py&||?_%hin67>Q)67n6#f}xQc`I(-O*Iej$5wT{%4pz?Y zWuTMZ9SBCWm5h+>9eK(A#mwzZ9ftT=qdMxIkiKe;_6p%KvZv*^Y@?2qBr`qH%Iu^_ zRzVkgYd#*zk=4;ZXCm{y-^D8o8DCFFE&5qf+k2uJQrwv#w1$f43^=bWms$a?;~TjJ zaOS1^8@=IB#c%bi7b%=)AE}>A zO11E)ma97p;`gM4lDi1|regafvUAitwr=fKeUL95P=+@@6z0^_d9f#^US-DIv4f40 zhgE;vvaH>A2fX8!<=E$f9O`2}Q5q*QdIm7Y)5Fn-VJQopmN7VDBm_OdU3`Xx$G z7Q;(;rz?x2A@>?y5+JD8l0kB_jkP9{D6utKEb6a5X%?J3;N|JoOu8n)I;R?)fMSKP zx8i)%4r$4+h;%?dc1n~keUjX}5^p2KvZF~>^AX89$C4}R37a5amat*9Btz5=QN_T; zQR=E!St6+u7hEa7oxar}u)!8bkbZEsyjejtZzZ83dOaPk@cKu8SRRHjW#NEfo&qFa zFkEPfHK3>v&j~Y}_wCeaz^-#c?`q~rY>(P3(K2oyr!*FZM=Gy;thg`U+vL~q1vFRGkCoEYu+QmDa!%pceV#|&= z6yaJuo7ZdbGG1;*S;1+U!vs!Vxe$mVw~_gdy~{{nob3od?$^W~oZRy;pI)w0m;?rv z>ClmNp$uW>A*R z=H7kT*Nw4j?_f6UZoS5A=23182GG%h8SyND{*hEOit5`Stfef5Xn1;NzGR5SvD1Z+jOa zUuUGDpjtCGNoPS|-G);a&MRpt=G3XJ#4a7hU* z%yt!4{Ly=LDD*7_4reR-;^U`K(_WS;js_~M+@mohNEMkM*VXTOue+QHd<%_B z2i_n4InuU++U)g1U?gX!pCg5t=$vY*MUOAQQb|n=<5XCBQwt+BiB-ACLqRWM*bpo? z@|?l=d>;YqClU&&x^W0#OoV&TGWo>HrfXpAg*7KGs}zPA~dUcD$jt_)Ko?;K$G(_u?atu>qmxY8J~HhL_PV( z08aC}1z$;wyRwQPjIy;h*!Bt?97ingNb(uQE#evLTxp5Q1Qcdp?M%b%)e)KMgCoI1 zO$hyjXCrS^ai$G3PSV;>Gzubg^Nl&uXCN`N{G$3>7>Pp`fJjbf$oc`>SC~Nr^d(mh z&b0V_6J+{_lCl@FIUV>RrJVi41VMyLhJo5yNA?YIxr1T#IOV9at*v-3e{erA{i(5D}h9je!knmssAVC0{_j19l^IOMfH z7tn4r(k^OQ8k=>_ZXJ(OV(?x#!cdi|C?e+!?Qo`pml9L=GX`FqX<$8WHKqShw+z+u zBR_t2=_IHZ#%eY-rsm?fD9VyOi&$v`3B$+sf5f?s81BVY?0t^nZUX-0#5eLiiOgx8 z^{Dc?GmItP=*=>licj!VjtH&J3S#^Ps~b(p3>Q1Mbbn2Dwu@O;EVA}ncdK^0K)75` zAI+)ZTa*ZW3Y4^G^Y0P2qh1GmOq69-T~V-BEgz)RkQ$@7HKu5EE=U@2Jv`kLt}?qZ79BtofQ}%JY_RJc&XPg z&VoQ=;4JYnmmpsa|MMd=p88s|+cOp^s&`*B%nQ>q$BQ2hdrui4RRj_YVkRgin_}_BHpe(l6{UJB>mL8ykHwc7*aSraF3l~G9U`?5X;qx}NGa;F&nY<3G9Lku0T?xJEGL{YJWLv+HI z-1Lr{4pypu$g2qVU6GX@3=F?jl~1MjCCwM>;NmFRZlhdG?`nVkO|9`d($?%jaB|-n zWf{~gezn=$ADuKW(t5I0R2B8Ko(sU{qp;O;0W`ixs|AP(@ z{F%l4<+K44`i~9~M0D?gXd!~If4Tkt!9D~95cs1Xe8tl0t_(z z_idq{2T_yr|17-zlac=4X5nAW5I|b~UtRS7HX|BOP>2rrJoM~=_%pwO zXc4ClAX0?DHxM1Npojp1z7|Y`a0a%(01O~q4JJku0UL-bH83e+?&nT?0>8`75jeXb zItY;E{bytFKjR6dN#Er`aK8p&;R(V1^Kswsza{2;BG5EmT}lM>2M8$D9|2(^6ox^> zh)^9M=iWUExHSOe%U{R_1^^2Ghz%hC#5n^zJ`RRKN?>82(teK{Q4E!X3gy0kSX;Y$wm>UPtBNz`rJPbc4{}&Dt244PC z4*&w^_d+0&Za}(-ya|xRZ#Ln#XT$%+Cj7Q-_>WBh;3t4hKqMUfvVMRgES`X&aK2xr z4u9|jp+6Xb&>xII=nqC9{0B=A{zDN6|G^T3|6mEftaSjaLHG}rAp8eQ5dPEOA7hFC zL;`-D?ax7fJ3Ig;FZIdDEbv`3m8y*ql8=`Le$|yDG`EK<8&kAhfMAU4`Efr& z_R*;6`&+MXH`iY-q|>OY`0fQ?2A_Led7N(@s9dhl-r1P!rn+%3PqK!eFx0DBE+Macv}vH9=hIS^gxIEzbA9uFMjFRaE@}=u=@4@Emo*6xS7b2_60U4be0}p zZl0HgIXMbrCb5@T?}k_g6&l0a0}Z>MA8UmK9IvijmPY;kB7e^kXj=*9XuJ%$&&IT`4G0xmtVj0#M|8 zdcZtVSmNojOITW8|CaYTcbN_#40)sGRcR_B>!Ugck&L!gQ>g5Hht?|^bXGdb%jOYB zjx+rdNtW1-1KS3JMA^Uxufc77O>gSGIblgVs-Zr*rD#vEDme;dA_4yZQ#M8{InIiL zC8!_5qN3Nr3uh5prkqITBoTqEm;}uozCU!0o!kjzuKmyh!??@!JQ`-2qa_I*|7vKz z=ob4zz?5-Av1g>QeQ+q6Qvq3FwI9l=XoWO&^8I-w4w>PI-5bjI!GsU02nl@tcR?Y( z&{YQOFEm~XL)}@5#ml>;Uq%YEssm)!DWW&m2_T@fL-qq!Mmx|#ZS_WnZ=L1_k$p~t zK9N1fzW=#UO~%o+&p`krBtC_&xuw7v=Pel}Qh=Rw#(lMk7ND8O*}|>j99~aT3-|Wu znjFfp@%X&xl&h8eIqbvuQ!Q_zPvKXtv)<7Ybo26e<$~SmyFd)syfJiM>*3bVw8rru z*HwvuujEFAHb@-n0)f-I-ln|@c)wkaJK{b_E)tLL?A%qd8;R1JtJ66grx|f4B zmIr5%x)!+`D_?LtBxTtItsr%PWmxfvnO%B%@5g>Xxl(A!^DFb4g}5Ey^oL}SFzJ0| zhst#{Yj`m$bc3U(j$&UV9V zvE@zG1eX|ypiHzoJPEkbyis)dLT@t9k9-|hf57N)-gtCS4k`9VIC-VP{l$6*I~3$Q z;==%MS6PLmRb(Rf6KmK~QoWTXm1nyz15yP9)xN1sZe9rDbAZ{rv)Y%6{y2(2A;Mj^ zSgjdrnOINFhxy?n2uuE={}co-a1ox-Q-4tXEg@WU_F!t{e!4`ohmd5tk^G#6{HnXuK;^^LhT2vtj8v zi5_wu`E7Om*OB)`L!Z@5X+g>7$qf$i7k(}`=%n$GbH7o8a2n}Uzs=zKl7Xc8 zqqPZ@P$_ts@Z1IO6tR5*Mw?6p#&tK)+kp_H#qBofPY3T-^; z2Ll1@!N_Cu7d~o5%;;o|@Bw!e+Ily0O4Ad3EO}?ax6 zOue9+s7e3y(;HpuwI*ui7bw)Zk&Q-dZ!kUq5;RVa>z8nha|`8G?h%u!O`b}~`r%i!5_H<&7p zPib+d+w!Sy-a}|)OLwi4%qMl#I-*lPqQd^hoxv(Qjv(tKtYS$2LlWdjf^6;6rwmDa ztZi-=eM4u0U`iB{SqAv!=(R$&+e<&x2)pukE-@V28k*tBgHF1W_QXlo{!*1JEdHpL zJ>GTb4m0(66dieEJdK~|M0q0{fj@9@$C)u4%6rihA#~)w&!ie_iQ-XP)$X! zUxf%J%8`kfk639&i}(y>^)wDGxse_W55@Rmhq_5`v$(@sv1JeL$5(W z2*`uz{Zgl=ikh@&Iw%P?&xS+Zd)Mi@%HSTFFw%4AwCyOyob6}OV!V}fWeLUnbZ1&I z=H_R8a%+C!lL)xO?SqRH{M~6QNon5rh|lm17TK0BC=I}o8$+0CsEmG=M|~00-&VL_ zQf#L2PS{q+zHhQGnTbc!M#z!xNE1Kq>eqZYl0ERNNMv$+}$e2@lEr~`+%|4c$c--|M;+h1w0!s-e^@wy&ch+b^@{PV zZN3L{1~S|}@}y`33D6P>_d;TKv+UA*?TAvNLUez-QnKbiXbJkKpGAt>0bwZ&{Ii7e z4omO^hDN+?;-Wvdi(DDHzo3Huhwqx>1CSoYz$|@>OJM-y7^{vu#GZ~rDk6zTYd4W$ z^OvBbBN+?z*tlc0*ky3c9R3k%z$2B!;u0!ZXTV3bRhpfAB0iO=7?3}u7;tH!{;vJc zxS`KJl|G>R(pq6h)QD%tbPd0v6e90aV~x>x$I~T#rv~Yquzbjr#cTOX6u&g_lmuZY z4_ncO4ydF3k{lI=8HJna=l<{ohRi+5#MZ3I&jY%;IQ^g2DF#qR2m{ccEA0*CWkb-0 z+M~B<5{k!3?9tQQP#{F!UDWY)T#Bz6bSm(rGrDbTum^05)Ujfxr;A|moJC5N`ovY-2j2~Xr(FztTev3? zMItx_E3o!YFu~Xd3t4Yzj_fd6!taOQh2&nTKzr_n9HaY&hEjT{WTg?+ zbC9sCzdg9N^n>vLQ?K`V$8haYtX2-)5D-%A^!=`-Bpxo){!* zWZUH1LyD(~q|){*XF z|6THZ7sW%u(l7$hbDO#QvmS`?t50HIIGA`xm;KaA3orad`7uPiC;)o2oJ~at5vUPzTUAdC-zvR zX+KRU=&AxgS0yEhqGbd_^$3wg+4s!b6Jw1A=k6#s(#t={`sTq?nMt3?YzCoBmUmi*&X!bVX8!w$nztDWx~au%!`Nx4)mj5| zMv=1HRyZQ)^}LVpV{j&0m@d39s**}F$~aUjwUjklIY2r+E8ztQ<B_^WT4ZhIGwK+CD?=OuX9-7pzPU%YM*Bq5ve}`M~Q_l7si|TpnRP@L-G|gvT zG``|fjhyyT+<9j)`$(1k!&zIt%q)dr#lW!5m#GLej`Sl9?_qN#I?@f9B%5PvwmWtm zpy0_s?&jpLwfk;v_`33iW%tL;>r5pxB%>#r2nLr&1z@2DB!k{p8`#|g2~Q79FK(jw z6Q5vOzc+wiCwES#Z=fJkB%owH&en#uS#ENSCliz(3>bO`sp^-#b5Bg!glIXxdU~Bg ziGW3|wF}_NrQ>5dmj=;yMQF~8+zp^^W*SBQ(Bu9w7hUcUwymgVwl*3m;B)y}=38H$ zU3_b%6%JVW2B;S;+))Z?mp71^O=L3I%z0}YoKpwyU%eS&w?!VZ^5!@$tZlA(7oVV1 zqk}``U-OYMZsWUqW?Mk{Knd)SL=JO~dMU9CLacFQ1u@x#xiA1UKb z$cM*_P0&ZrKX7fPi3nH4NT{)4^3j? zTx4<$il05&TIjBz%B@sOrcGOxz=<-rtEBU_DhR`2oYxBax+u9ufhI@v1U|X_BJk6Bv6bY1Q>RN`{apjf|a z-N#G*5uLF^l*T|do51iXS-=>5rmv9kg)rO$McZR%Rfv|Jk1*Z*94&QCx-@ZwYjl*; z6DyLP05@Gvq#vaC#`Wr2=usO)Ic>-%m-kf!mHuiczSej;%7r zGZI40z@~=RqZ*0SH`(7TZ`>#$Jr)d7F0e@SaA8A%-jc24QR8wl%n>vLi?N4oil}hx zDzshnQ5xMo?GqCr$dG4YwO*ie)yI+qdBnVsu|c zynf#3L(w*0T@d$z9Bj?fMM+`nz@Es}m1T^W_VOcoggiYim^bPcg2AF-n1 zm{w`8Us+F8tjnYKiB8WFQaulgf>-Is6I9EbxvlgQC5TeO=l1Hj zAhzK)Q0?dCBs#g)f%jhNv2&ItucOJ{V)fGuOrx3(5JuB=SAXF>FCD4|XOqNhE6WAy zWlTKt#NsM9v~MmbmK~am&>u?Gv$;!YsExcP#N2_H5!YN+wW;$&7~5$C!u4R0iX4F% zg%%bnqzde+-Q;50#WTJN8H6qZ2RXTo>n-|2G}XeTymNSG_3Pd$S%ppvcas7g-zV_G z^xVQcaetiQn`MkDge}J*?(Qd-;;wO!4G_%8s5%```ppPCfru1Cy`AVLspC$1!}Kp%FjVbbIQh;)={r-clW-bC^VGg5nVoSLvU9 zR0NIibuFX!)E&kYRezHZrfWn3oDKf7HfGg%9ge0dnlw=T`;izDR2D$0D{9NNBGEs%lo@oqlQ6krvN3RWqVSi?#5WzO2G*p>(99 z?V1r;wgXhld^;PhWXzVfY=OhrQ=u9KH)ZINfw*la30{m_{1k4sg4VmbWf4zy!X zC&9Sk;@ySMLvA{_?Jc`^c&|{>XuQJanTm|>$sqxWsR_6+wRpIsb)r&wbM=n;u=}H? zR^6eA0U60x%U)KI4GU8btu)TVU*q?QRJD>l%pWFC9aIU%(}GVx^<^D8DdtI=!?@Sj zOPi}gyZd4-BMua0u9ngz`PMzP57pdQd|{g*!`b(RM4lFBaYqpxR4{3t;|NysK0;lV zdsSx1#bXagPva~woZT#?ve`t{gr4L(z7D|w?>uKdq_?Oiu=`3`E*oQ8Cp(VIV}2AA z|KjIkFxxtlzVsQ^7vE-iLI#$dbvKx08S1hw_??Tbk~!q{E?67&>Ua;Qo45&I3T+r` z=vQB4-_I6*qWwxU$1T<~V-z#R#A(k7%&x|}Q2>R1vzaZWEX{}cbU)nUwVM?-L^H8| zubcCnGJWbTYg_8Ri=@QtxpLP_wK8}1ZMuktN##mgZ z_Yak?N8m^V_Ki-R#$0H4!sd>*97Ym4$;)cWB>dmQ;%f`mcIJ>d0zA z&VcT<{dyIWCX1dxVdry&Zk}yuy^jXIR6VYIjFgN6Z7!>_os-{Qwzk^ zv}lfJiSC7s-bYUh)k0Ef6YzSC57FJoa;LCxIEFhvSixiYP1*E)8?Z*%6JqV^Mbz|^ ztb^eCEg7M<{LoOcsBztr&hBQ;HD$Y*{JWEr^YfidU)HCIk952*8qqI}7}ZTER53kr z>q>nRU|Ym>gy+r|6k7)631S|DE8R7v@o!J((ECu=UKec}#Vf~g`wO`d^A+Uh+F_=M z!4Ec>M$<{xdm#5PbMO1TB&8(ZOL3?~5N2})b_LX{#EwhL_(|8w(ZrJKS1Ay3G?61N zhNBo1()xxVea9*N+ddUvsa@;a(d4u$jmf>OjLFTph;}O}3NETa%w6;@q^eu5h8vAO zqg7(EMq7qFo@jdDHiTgnEWnD2y&3c?E)2HHXjLyMRbk{~fe%l{}xAXAz z_1SQ@L#>bR4no-FCy76+lMwgFVe*k+9;zcVDEjo9Wxs6} zdICJ7UK~2&ngQ_~TMb!aDL$)znv?-Iz=<>v@jY&cck7MK2J4QdkS~^SqPXMoSD~vM ztwz7aUG(bcGghVa^Ft3eV11<4*k-om*Wx}WS{^_5g!H5Jq$_SMlwNv5KPnjgi!cpYvA{QU5^vst$JmImHv{i@*6J=&Y|GYN+b z1Elk<^RcyCP}=vA|3^Ojmyi0-Y!{Hy>H5uxABd?Nh+biV($S@>5ogdeC&|C>bp-)4k6HJFB+ zA4rh?&D#Fg3e@WVIv5#9RsKFR7?k(dJBS<_Fd2}|B%lG)Lim7`?4OeQ|2-D=b59zu z(C@4{7*GrNHA_Hp0I~KCm?aGQdzL7GJUkzQUKLCPhW?%^!hjk~28RBg7=#bdGWeV8 z{@?zAY&Z{MMFWh5XbA-4p#Cf(7eR~$f{hVwlc2kZfLoxHJZ}`lh!~v&Qm#VK|Fqrt z_(iy3zl0s==!5{B_%{n+DFk#11Vsk)*%1wqU=a#Iq5m}Nd4bn|Q@nugJbeL(>USC$ z{SEMniVIAfhL1{(z>EcBkqQD`|5I2O7WgA2fuIZq>mhFUL5#o%ztZ4rfab&B1h^m| z!odGWa)MtF`cLmd5D;|uleMnb0t+HW4?wJ7Kz_Z!fSVNtXhZxB1_9y!1PlIe+4^6C z1)7MgLl83vmPUZggV;R;lqvwFj(;KuFsR5sy#fJ&wBB+mMBydq9>VPiP}v7mAbRhE zxDo9@qL>N_tjIsjc_=sUp9K^E3gv|m4f`N<1m7``hle3TpMapu00!_^27zA)SnfZ9 z4uC%YzsxtkFk*iaq=l$|3}oX$CjfW>Ozx*r13n8FhX7E1_@_(({i%+RXgC8IAv#V# zOh3=dLkPr^rooLx$j=MJ{jaPzp!)y>0pJ7x{|+)n#GHaCf9?pt5TXi@CjV9f2#DS@ z&_jR+0Yn;j5c#uUK+psB-$neXckypDeqh-UhsI!aL`W1t1ko*s2@)P0sy&9?ezxK8Q0sED3 zryYC)_>Q0W%O7ckpSA%1ks-qW@f|?t#0a$u5HX;l^6TO!U;KaQ2tGjW;%|)qqa!pB z4;aC?2+~UsIT+wFcrQUBl)OI~g@0lj{{_K6Hp1-ZAsd$<3g9GvA~9T7AZ+rVH|pPL z0q%ysvrT0eJHdo@4^(lKdtefcxQ>P6F^r@aOk$(gC<{|7`yz9bmk_1sGtwzexue?{Crp z#`~Le06?2R`~T^W?@xbxfBNI&Q~$jK-=7Zo{&dLqr$fF!9rFF|@En;_J6n1`cuWiUDTv?tU0i?petmoEH<~jC@Akj4zCAd} zi8enUriH^VejGHH^|PAayIs5~xY=*@yP3Q0zCQhN@5bwPxZp|m+->B(#soLd+3mrr z{(GDL?|m-eueon3;OF-=Zf`WE;464Vf!E77xA4&_r`v8{Z0n19zrgeU^I46R{NS5+ zXFCVGJX1d`PjLa+hA!GQxQ0I>WA^yfy?eJ_nIwY+`M_WmfnOdO>CDu{QwOBJs?(@M zRJ!bJpi$@BkVvo}&$0Ud(B8g&Vixp?sFQVZ&DYQ6!`kab^XssCE1i<;=3gtnEcB+C zABc$?=3jj3=YN4Howy|H#A%K(9}z7xhiy-zJ8O1HR%kfib_tA#!KVaPNhX2^8?K}F zwKkE4^@9#`9ep6Dd(U0JzC|~z^?)Cl508ch*tpz--$7KB+jskkD#({d5LJG5P-Ny7 zC`hs0s$eoo>>koECUz!1@1Ww5+-WP;2>Yyx=WTjVw%k*i2A633cv^g$s_ZJ_du&Dl zB7UWxa)@GLIQJp{HoW~j`+}zT2Tvc#Yv&1l^gWr$cKo8i(c*(-nfvSRD`!#rmoAgV zpput0UyD8%ag>_H?hp`Oy#7LPdMq-g@ap^>^|#YD$vou(&1NG;*^O>bmRxU#HtC@{ z`6Zeli9JnqIL>?}8BcBJS>#f=cwH3bMwvVlhQDMcs<`8rMm&6~+a)SotJ*}cppx@O zDYjdGqoO<^WTk&a3;P|%{VkhL%!&(=UG_yGOLE2jv0CvPdrV_f@f=p;1MAXc5;c)j zzvfqO6+S&T2u|!okxlnI{-8%m>qsOvkI%m5DbRu1Y>G-YQ!nE`b8PS$nt*B(g2h`_}!6gY#yq?rs34O|7yaMJcIdTC?Qm!9+T@S4jJ?Z=xF$pJE z2rL%X4>mEEK4VXFd`G68J_hWSb)R9pfDL_y1b(i57q@T#FMm>N<`HmG+wc5%-urQ( z`a(fF%2#YUJq+f9K$fVyu&G0R4$I=%qKDQfoFN?p)lQ@3{wP74FOY)s_rEbo@w~Ki zfJl-C_^a_!c8Cji{BRVXS9?*8E6^%(Os z90P0ro|hF-?r{EtNpmt^ocJ|ECNBcBGs1nwn@Kcwj%2F`LX1VqsO*>)VMo-+E(0dc zaa>0d{W{-$P*BKv{RpjE2SaQ-(N76=^Lk#~>^}&qM>BknXX$6~95i+$+L3`8NY^DN zV?YSHJR^LfQsjx=C80z6ObZ5meXK=AX-g(pZPm~5EgOE38W*%|Yl{=qK!&PK8j|FT z%7^&ATNNPGF^tXnWZCx7txU2S@zfR@NGoY=2J8OH1OJ9l;+BB1ck_O+I=sbu;o~#9 zKQu9jt$Pp1&oLfEE7a%KTqW0av9D5 z6{p-aN>1%1xN&adL_t2(ii@ifnzGVmOf@b}m^vb-ZrQ_+PscCP&CWcm#$njag^st&Vd{ao;mUp8OUK?I0oS z%~m7&xOsXvTFu3O<%G|>FKEY<;60Etkdyt`kPG=r;V=Fk%sQ|X`joqI~Qze8Ryj_YC=P?pULx4O_fmx$^S)IK^4-?SBNvMx7&VJ31g zEWBA=&TKA}`gy~bgC>V=DC%OR9&#)_42KCQ^qi3NwrFCb+PVy-OiogK1@BX-bTG^f zfg*-qTAr`gd6?x|oLQxxykO!Wkl(t{)t^BZxI5J9S#j`%0b)B;P7-U+qe~uIc<-G= zSysVSUb6M87LkW`ycEilf)Mk_j*z%$R>SgCse`8N9TJXKXN=UYMHe=r$qZ)(VfeD? zw=*w`(5%vyph)^8H>=Hd{9^H~7Tpp!$zU~g9xf(Y6e4Wkvy5a$I`8BKRj-zoAr@wD zT!ja>|8p54%kW1v_5X*iw}6W3`@?=^2q~qzdkBG{8>LISySqUehVE`DK|(sE8>G7g z=|%}DDM7pk>;HS-d+%DVC1>XBefHjG&oIOD`94pP$WWS7zWQe-WD{-6wD%#BF(;NY zsA$<2z4;Q8ipBi4Vv8Z<#j_Z)U(N>VGm=+>c~q#vBG`$Ru8-TGupV*9PF%iD+qG5` z6yuexcZMnG7Tj##u|jytY$Xz$=C_r$DmM;v$qDCMPdUq|JHYyOZ^?6NblZ$Dh;8mY zX$VH$_hmBhm&8SeAX~dvN4#{JleTy_64NG3uV|EFo>}T+`^6cvvgkb|lsta!qP9*9ETm>urDj7iI^zPEJ=o_17)JG%}l*}4FW2spM zyDxL`&` zy$)23G%9cPt~k_8j5pN#iOTIxPQ2sPVii0ns-AHUTWUdggG4E%UO{DHcIhWFH#mJX zMhdPdii{ULBT(bwhJ(g4r88?4_~Ivl?-T17g7aeT^nBtlTGXA>oo1&APZA~q94BHs z+Y!DYPiCR`e*~(c%dj6+NTykp=DKN6ckq|>R_SdewSv_9JuBgm^(dA3_^>Uo7}CDL z-8*;p@*D*zJS}8=Jb_{rhCnezqrS6L6S|fkyp409gb1aj7Z;=oe9esxzT-JcqY7t> z2rU3tPooYt7Qe&jKcSgFX{jJpO!869Q-6c4f;*^?WMf3AC*q>n((l=SXCs5XQ2(Ly ze5z~`1A9gUQbcPWXuQa<9j`dtKYcIeOFJaP$)bU#C7Zu0@)4!rRP9JH2f1IDw+Snu zpROrNMhvQpTrD!AA5x2rXS30SB;V`N^Tashkx_&a&CbxXFmQ^HPYGpW$^M?qlya&E zjbizS5Gr|s7lkR$Pcby2U4$2?>O2pjyqAgA9Ug2rRrn@+n(GPE4<<{4&YMY>8C+hUc-A08+Qg9rM!JF@dj}P3b z&&(a1#&9$Z?}hqI+rU%Eg>&(nrpWPHA7)?U8O;h|BIgpC*(@Y4c#wi7^2^JnoLapo zK>3xt+SsG#;k;!bZob2_F;*v71=jdf=zN6?^%#N25qWywK%YiTeBbKLs(Au$o$sUk z5MiM-(qgT}nRg^T)ZIS115yydV8`qP*rN<3JI!gCr&R~VF?<3dPhg6srA$nWpN3s4 z4?V>zR5KrTBOMtRq$xX(V8$sx7rnz`YNN8stGQqy|`1^{|YYLPtJgzdtyM5Bjon+S++HaKE&PJ+~}Hr?9$M^r6CHw zpqXC4YfPSiGb=f`)pI^Vjo5>os2`q2Xa>j2>c@#)rc&Vbo_@dp4^hdO@ZFD3(Z_a^ zoXsD`wlFF*iJmSN9o=;kHiqlj2<^QaQmG8|Crh}M9NWv9F(i{WyH_y5 zX+o$VFw~fKmOVb}1aV$<$V9QW6Ss;-sWHEGRTCZtocuhJ#<>Ysx2K3nQuf+Nk*e1) zYCQn)QBrdCh^zov9mKqxoIMMS7ZCa-gbml7qSRgb-sJi8i#ms)*>L;!fQi_dWZCwj z_ykT)digL7;%hQL)~A&rU>r)enRj%rxiBKAOQ6;Pp41^=)H5CChAT@XjE?q^XbePg zh{EH`Xsd!!*e}cbFQqowoaKXOL4}be)Sr_?ZBN(bSIl~KMvpfOyz9P}PW44p-UP4i z9vhkVIkIA^7Ome~&N}u9(wrG4SIDTOSM$0YHVmpZ6IBjAdt5}#%o=2Xb%vRN(~p(Y z!U#=ebP7lT6XIo}h_6k*8MqllF=y2=HP0ZEV%<&8XqD&=YGRQ4LX&Y``8Jj63g`$xRWy0vC-R_sdca@m>j)!8^O zxNM|lvn|{H{{FdyQMEiDUXw8#HJV^{2clc(b7>Sh4QZ)goTL!B>u$!9b$w|x)9ri~ zl<*@4XM^BvCZSy9P2_H*a)*Av`$)(odyxO?h0l4zpnSTz7B(=a0ILg8d{S&HR%k_o zec)FwR_AJ;;d+0g>eD*Ew>(!vDeK^`EoIn6ICrRKvl4jb0WZd`(_A31d9tm6$3leb z=Bnf(naQtWBSs6W?Z;#I)kU>^Zc)IJVGer8aj0*@6Lq*1Lk+A*bGKi&^eDuAbP2+0 zC{%~Tf>D)!DsGBjxv};9%kp{|@hDW}Np?19SSlkT8j5GwXy zE;!8Dd8C|l&D8}q!*D_`x0L?NuDHQmrClpG!U#}B{R*F{HLrmB?6S2DnYm_4ge|3U zqO+f~#-67}>#VprrTCZ-5jEe@MudL;P1TETPepT|nbju$h;p*mT%hKohbP_ zlit7*r*&l)-N0Rb6JJW3ZX^oZ8^z?(%2t%J?w^k~C0==c#rQhHZ5A|`md^);l9s&D<60F6k-<%{W%_2T+@eB| z{W=-D&rp1VKs51;>zHjtYu^OguhUHE@5G>ee<=e}h_Azp) zpO%a}1@bIhY;pJmgO!;s?8Lg`n-Y*OE7A{TP{zfqTp{$c(VuEM3$|q{<92SDPxIMJ z3rZOCs$+JDaooKWQ(a1-9831@*0=WSZmmwad+YoSgY^tmU847rphNkm0;{_fJsAk)N(bE z)+K2&wHya44rEq^X;)!T0zTT$0!w`Rf7D`UaVOsq2W?(#PRfcxUx}QS!bZA(H++I! z#rM`@>i}wn?Vyp<7_#CvkLfcv#Ris-GY^+5-Sp`=ng8x;X<%KagsD*CmtR0Cj8AM< zuwwje)*}%)c9D&G=4M9b@w)duYYJ+`>5~R<@C$?RDUH$OQRNLwZZQ=tl032k_Ege< zF?y?Z3;~~c`O_gv=u_kk{+DiNl|pL0Ik=9hm6&UQjuI4CgwGOT)v)78-*`{+sr?(v z(13Ru2x2R_8CdE>-OUn-|1sa>d5QOjQ7}bJW{ubv!4A(MGEy_9x`ZwP!eL8Q8zst( zNUMVbCNR=;OSf@>yQ~efVJ|Su!mY`@B0n=JhIT|KZG@*})HVoQjT}RkX^JkG*wN+1Vk8I-)g` zbk;4bnz!V$P^Gm*8j3JSBNgAqPO&K_&*)uEceRMQ>)6^|5mhQdor3>9d8`B`eaBi2EXu1pxipc!YgX!U9`0@^ek;L|PXQH5L=E=wqP_d2BB z#TQQ6@@cIfuW~;AY0&9O!&y>Y?97}|n55Mi5$0-N#ZA$Zs^Y}PFN85M-InmZQuO;6 z`9(5ox@(c?9YS%GpUb^5>gDvBsU2*PFv!&0 zcX_1St_zKr9px4(qN@_=-&t{w8E#Si2ANf>d2*AXmn5>Bd@)F_L<}u_3bGZOosivu zX;K9~)R_rNi`&Gq`XtMAHGYWss?TVoKe&Ci{F#~T(++&j*ocP8XZ_Y{1JqqGZj?8T z%b=?a9@FT)sx=Z3)w&P`h{}V?M5kh_A5_OjpaZ2BvPgB3KR+&inL)zS0j_CO`hvC{d!NND=({0M5u0=4#dl8c1DhB?P&$_-p4-YOPZp%Hw-a1+;3h}jnV~8_vtfqmPY&6?fUX( z+g|IfK7}{;y|ZU5GF*HWr!5yS#@p{AIv|D(?P_>o-dVkO-?3Dz-aEmFuX)5VdDwto5RKJP6fFxUW4AXpB$R%R#~*AIn&{W5H;BGV6EO9< zzm#0$Q8#=#)y#Y`C6Ye(PLUuHy^&)z2t^AiY9j5C1Bwlyuo3WO44<<1$!t8H<@%^W zD%Y1qW7}DJJFE~?&mli0!V#ndQ7R$uw-_1KoDvTd0T;{!<rpZEjP_kDG zJv}m-O+#TlZGM^y9)X}<3Yl1a_#4W3^GGq({F|T?lC9+}OX7}!L4|cxBzta#U@K4C zwvVqEm+?;PFMj`03kzz!JPT6m>DJ?KF`f<%iYh|1FLW?o!F|f5VS)14VPm}&$#i%y3S1v5;+EKBU$|N#Fc2trs~@w6QWwVf zrcV~sq!@rg$IGP(FnwW#CIc-Fkt%u1ww^tHg*q4)pivma+#x9~m1GWW=`(9+if(pk zBm%xwI(rn@I=vo<8j->&gj;^rgaA2B)MxYVE|)(|tR(iC9FU#4Tbn6 zTLHIFb9L2GE~-iEsudqhf8x+BWvwpEc)$9~o7bSw=$vhlS!CH!+$}oF#gFTC|W{9p%W4?+226gRH zcEcMAjQK}{*&>vgBhQ+bL(Pb$;x)r%UshR$;0-~&4;GJ45&R<_Z-{?y*zi%^CRfW` zl{^6%bexX&#Rk(RlX%6gC7Pw_oo9_;F+@?>mNUEuHOB8 zeC3Ly|NrJL0Cfz|+J12tylmWmxeLIm|Ied3X`f_4I6%4n|7zF&LmB|F19BkJw9geF zTtGkMOq-GgkzoE`A@%G4yYcXN8m=sWE&!1azYYLj0N{Y<_d#~qulBsZkK};!fDnO6 zSrA#8WCaKZ7Kl1X3y=qq;sdn7Uv>3AgWzEM-EMXIp%)JJ-$#_QAY9NxoCYvJ5TIH= zSP}rw{vIO%1PBQDt3Cg>E3RJ`N*jL(LIVMn_YYSE0aOMg5(c0$09F3i*h%Dov2*+x zt-t5PbvPe}OA61y_D~V8Bn!fVfjpc`ns!JGJPj1QKQw#_xHkODkzmzJlam2~5!l%u zO8C>fWI(LQ5Ayl%D*3ecG9V`SholI$U)`l);sUV2uM&Q4j{jK-KmenN-`aSZ136HJ zpX&?4NwfR`g98L3Jm3P8H6T=6fa751{DUL_HUEDkj6cMs$Q zaB=<5BmfY5{|VJVh5i3QUHsCUKi>m-ID1r`JJI(e0Wc)!1 zH~}OEh(=L*5a;QVsx?^XfcyLJOilna_y;ur6Z*?ra6*3hHb_61L|~-}QG;+`fV7Q# zW_orAke=~}wcy}>U@d-+3}6cWF&G>?9DgAPKL!v&+D~d=hZ{`*HVZFlK&Y4;0C(_r zbnF2617x?nRt0ILeV_r6fVj9Gro;CX*b#pdIM{$)O`#6bOH*V7@Qi3$5G^p4hr9lg z9E3{3&h|fd1UA5Z;n#Q`Y%U)5TUx>(j7XXs9f&^7ksF4U7Pw7+2mFBa{99sP0PXm* zUJxGce@hHp=$DfQfOGu2)es=3USJ=m73l!C@e3G41H%nud$9c49)Q6E+dt9* zph^C%GK7uoFKI!aR!k4#1~%1(E{K(y`xh1PU(*J5?q6KM{|E-|L}V~5@aTW|1?~rA z;Nfiz;1@Xm-~ya~_yf*A{K4<23UXlY{l*P`=Trd4e|QG2KllOHAN+vp4}QS)r$4Sg z{c-*2kLwT5z|BhgtHb|s3BTs>e{%_d1;~S;!*4EOxgR5B?Mma&A=eKr}Tg6ICGD*%kZp>rW)LdyKUcxS7Py?7A4y>a#jjHRg9_OSqx;-zILM zKmC60{&@Fu70Ua==jRt@@4<}2>j~ehW3)7~!@AoB=C7ST?YC|}kG^iL4D>=fS9Zp> z)&zd=Km`Ok&#yz*I`7=BcCYQ+2NZ+YK&F3`zSZWDG2!pmD@XPquri;Q)pttn<#n)+cpa5 zpKAsQAPbFqF@>|O!_u%)+uO212bGR6=(`?_lciEj$u5&yj5j`AbaUsc221k-$v=9F zT%q2H!mcj7#J#;nEqCkltqMx-hPT#TML#3SG!x0$6xC@r?yW$hoC6dz+3F&5wg(z_ zK8eoku0Z_mvk?@8Y*3QZ%W3Soy*}W>_$l1yxaz+0sm*-5=osNC<;msKUcEb)ld~OC z-FR_o*LO2@ew_SXra$8x%cvO$VOLTJPqgfiRK}9u^o^^~z4@j;E9WWH-;zyT zAmFeP?DD~)2Os9mw)TQua^u~=M$Lv599=XTl?`|~CUP#|Ez(W!sg=>oK|Uy%u!_2m zCn+?x7VR2tBB21!Pef~{HH4;s7#T|;s8PB3g=R+(MUAxm+Pp$c?gG}U+g+xFGe7Ex29NCT_Lb$?WDz@K6xc@OKYVc~ zNqk$1U1l8x**2R=o1MopHJqPC6m|r|bnbnlrX8F-l4+#BbxF=Ejv7Hr4QB&Z^WZ&) zQYc^505{Y5EOMW;);!v$j>7|*kI2wUE{D?qTeQA9bK#Yy_n-Ek5)*L^V#}IK2EBU9 zfH)pslVZCCryfxawvA$lQa;>z8S6N&YH;VXONE2;CKg)_fIml%I)C3 zs4-u%y~B{wm!U*5N-8%tv11?NN|NJDDSEf~gbc2Nr1z<_G~H=ipl-1xz&t-IpTt`xX9& zm^-bF+s>)TUHWRieb-)zBUvK5x`i7Mwg$cD)nsmkOUe6Lx|JD+XF)WH(2*5%uf#-f z^YXSr)ILWNG-&DNb4Tr`)}Qlm8l8U<#a zX>OZlog>$^_EMpXg!CgOL&ciWgW?j{%p!uq5julI(pT10jl}axyCzj5bJjT(m)BJD z&5!!&Jq_)j1(P6#Y|@0YiY4|+2^T2(vRc`WOp0L&uQf)f9QN(|nkx&wwA zO#CYDXgi8WcGwgxa3J6AB%kSE+uO5dd%sPc3Yxj}4PRA`q$B1%=&VcWGObROjBe%| z@YK^neI^WLCCob2Azm7S^X$|TnRjNVY-;r5+zzj;4c|NYTYFZ}BE1FKXbY`W^6W<} z;~O}Xk&jrcW5v`9k<15I(RjG)%k3@eqx3XdJR%6Pmwvw`KN98`97}jh273pOxk}ABU)-T&_Sm$)wDSBGE ze$`T6i-?zrBLF*0F8Q*ac(jN%HR%vW`1He*+PDdA9p34lVj4=ca-{@g(>D7%>f(Al zgo9D5Ua!n|6wurk#))1cMHJKoO0~w7oD(wpGny_mDU`&`{?b~Qja!CQb3Uh)bqyQG zFBT_>UXF}sJ=35TFEQ8di2YQbK_5M}N}a`6^wn-Gq`%{C7o)W@eqveq$O^SbU5SPn z{E0)eqQO$1Rb_80AN!Qqz%%wtPmGv}!fsz=KDWFt;~aYFl4SVo&7cCj|4IEy%CXXF zKaT*1*XyKcC>=p(9J;=oa;T>*!SS;#-@ezYY^#EE)n;u5)MGmR{L?)Saa{!{ zM9|A<_g1nd99i==ch4CzIia5#LVg=6FwqI!n0&YlU!2FI2yv%UnKh7q#9c@3jjc`Z ze3Y>N;(QEsZKiNANyJy%Pj~-v*07j8vtycp4H-!|~n6?4084@st$bOo%Ot%+=*7x%a!_1hy#gtqb9-qC+2dWUFAea2p6| zuVp5QB9&;Oe5!bvjA=RE9>*1|*L++h#o(U}HzQ)i-cnxuc2nL!uD`pBJI?~)XY5-K zXw2+Wx{hE@R&`Y^Zq*;Jh@g6ia-3nXlS4mMq8i?o0Dmz6dCot{a|dHl9AAEF`hXuV zN2A6NtVp{}2KOCqyolH0C>a&8Rq)QVMU=qQFZ;&CgL?rTF=JiqdH|JvD_d8fZqPjd zyGDsp)b=w2Z{m)LT?YZrQ=aBT4Cy1BKxl8?`)U?)&nI17Q)Ly7@qb)Bom**x zTN;ix25Vn0eCJ9)y|ii3ye{IY%VKtl5CA94A9%mwOQQG!5^k;k4tJQTlLD^6rjSA> zeTpZQgZW8$A4D13fO#gDJjOA6Ro((XbRj2(^@Kv83|2R zo6$&uw+>Whm6Q#}a;n|Mwr>X=35fGl9uCrraneQjDn0p{bR&Hj)s4di-MBaEyEUM< zGOA3KG%Z(pHgS<%vZ+udbZ?LbD=nvP#PSWo*|W(5%0FAdqdi_i*0Qpyf|KrdcA8gL zm0l@OLxUjse}!{>oXcbJXtAZ^&JY@b`kXs>wL~rjrr3Q;_HfwA7<-g~xzd=wZihh? zP7GU<_su-wtY9UtOS`Q2PtBpN(cZFeT6k?_L0i73cLA!eKjgvaa|fjMgG*M8QyIwx z6y^kPxukU(JH}XaK2-7j=ysWEpOhav7@K&!2F23ra$$eG9K$>cXOBYRVa;*J-+9W{2q3*c#7yl zSNAw!#cU7JN996DlM=M{Bo7KjDG2Yv%MNi70ucwy+kFPNy*FK}Ys>`-0N=jj^mhXV z0fxb9Kh-*p3ew>kr(dcOb>HC9w0)5OAX`e?6u3=W2g|s=e@5e}c;-Vq-*c>_-Ar(& zDnqvVM6j}$*TVP-&AxzdV5BKC*;djtZ{(|7R%iMiYz#v+sQ{IorY3u+2>lNkRgxN0 zqT>7t{f4cxhRw|3uwn>?pjv5N7e_=D%i4&*w9WVx!`1cZ&8mlfxfUaAq+4!>=I*TA-HCjKC)_izEe_)C{spyRZ?47b+)DV8urZy zC2r)5w?GLwV5)_~n)d`6OU@CG4U4NMM%CZz$xACHs-L4m&5Mi;U0AfYaG#*KxjPFo z6FGxjNoJs4Q~wdNO>>Dn#AwHXUA}wvgbNe?`pM&$C5Trqp@{mCF6(-yp20=*{e7P^ zs@Ge5_Etk=B42!A3+MWhs$I>jX`(>Yv%;KruwM23^252>wm_giG{Klp^yy8Pi@}4lK*#V7~B}-P#EKinY zSp_K&zFNffsMD!6gF=||dyK(H4wTM%jO4Gd>|?9xa~uNbm%fsA<>Bv6tNNCke>G-a zkcYU}vDkmEhoemnUmJrQseS6G%27~U2m6|+iaNW!HgPa{7TcD1Dss!Pb|wqk8gKms zR5e@E;EZhAZOQYd!2y@ZV%{%o>0NW=xE4GCBY#7^%O&hmBNTSV-mObKbLU4DlTyXF z!TRDZ>glo9f!iznUtjQW9;xd_nZO=QT}3_7GGAy<_0Z z{unGr*CYWLKgcKHn+AiX#lDl|F?4?>*~+ALIRcYmymk4cQt@icXUCOVj3!T6W>}&(t{(F)yA$udeW=9BBvWz zvX+RMa~xeFAvlEjG79u*)LY-N28JLB;+*5>!h}tWP#HqXltKq-*d?73CLx0ys$OZu zEME2X4#82(it&sn^yt`;D4q85wU4$JXGr!-l?JrtHYpt0*tSIO@<#%rR{(2kZmxbRB! zvsHi?x0Y6x5|xCRkhx}|+pFd2=thG_O7CP=2xco~yft?du(UP8itOKtYjrpd^F~gf zOBa@lBE4sDHHx$vJXcB#?MTtK`HK1GvLh3_;R34XQ}~#Lr6ZAhJw!DyT85|}aZ{M0 z2D?M(E$y_%FgkhRGP%)|Q>kA>L}`Aq^Y!;y3BBW3wbNqj>mhX^)mu{rtah;QER z`_7P8K5k?lcL;iZc6eLoU0bJ{(5Ug3Lk<0?@C8(G11y=~9uK!{9KXz>U;{G^$wW zmL8SX1-v-ans6=x93{WsCS`M zBqhTVz7W|;J#h_O+M%fK2IsGlFdql&ELmCEUSy%eDe+*%tc$(mqIn#g>?ndR#)f}| z?Ij)?$Ne#;4;6Uv@rhw?%4^@-Hh2#Fe= zp*X)lUf5Jli~*@6bc$viv(VZTV*Dk5zcG|MJfv&^^&}?vg}+N+0149PD8|!JTO_Sr3=SD$zk3hCr+7g z%puCKQTf2p4!mrY({`=GHMaNCDGRbS|hTn zntt(u(kvr3i(9|P>7AIRw&*CfT$b_}S$7gS`8TXQ)Pu{2Ti0lPTIT(p-4on`ma{&Y zYu7#kZ36^t6I&4ywWcBDb23KgS$HC{K8i-Xy&Saws5W`u7O{KbIAu$mPcX`*P0DQ; zDD*jq9GZWZem7&L3$OI8hM$K)>P=~TJH(p8^cno(KzWSKexi3P`=Gvb3M)vUnAb^s z#pQ5~)mp2wNlrALc~5nS*z(%O!Gu1iBc3kFOg`Jy6)oratPi*++8`KuhD_U!^R$DJ z&n?&7we3<1YPC0^aa4BCP55m1%4OnS_3`YXDHsUXU zsl1-B#zJ8s7y*=@Ja#@!q?;wuk}$Odh3AdYjBR7w325*=A9OYn%P+pvwq z;swt@|I3CHTNKeJDLVDzr*>Ik8;3+m8ip_U*}Yg5FseA?o>BhXLqTE1bQLKb(PYMb z-XTbENl6~>J4M&V*H5vX^9G#Zvk-H`$|eC$lnPng5VbrkIZ#T1det)6bX`2+A(LJK zlGKJa^&FM-$FuY*1G^Op_?>w!)MmV!J?|QDE~2x}@)OWtm`i|-=@m3HvqlSnU@|k(DZik4bmy%gEG5&PfXR%anE`01H zcs+}pIO>R=7$aC4YYfecKqA?;B@^rkmz1iV1s5c+2KB_KXxkW?qWYCQh9; zL=>k}XA#~y%R|4o07M3(I}v`ej|n*N;WCekn|vI zQKxotY;uTfxX~Sb>k&gf0-_$PFiziCZLdjE7M;G$PNFLG&H&y!3B<$MB$#!B`z~k_ z)e;@=IlHT@KL>s;GpauRI-!_#j;fVvLt92Rac#*#0Lg}|Ow}W2_{vB0F6-T?38==$ zFhQJ}-uRv6C$$g~^^DZkF_+DhcjbyS^eKALENhO>K$Sa)ur}&&HXG3`d7qr%P%{m$ zwIbZ3PPsauad3(en4dLv^P?;5LuyWwAf8b$Rm8!~Njr3h!yBpwe${Ypg#w@_nX?Ni zZXh?aWD?#KCsK{p!da*2X*eEs*OP2+H=@du*ujxD7G;T>Cz&w7pXc}%Xm&?9>>t6L z;cExjio3k^?@Q2O1gF60dVa_{Mt-Ua8Bf_1VMa88hV;?+_5%6JD~JO)clNmM%Y=Am9$wvvGSlL~&J560J@lc5-L|>EFmuX9O z(H4;cLyIO2ASa=O)y(5I7}SqE^liJ5MHvEHWJ4sM=LSY;(uBCJrR^$sWs2mXJJyd) zcYe|k#3w1cS!iiEM|oF?E>FE5XtllA?5t1&Pnk$t8MYIOeVkwDcYSrg)J-j{xxC_# z<>zxlxTB!k#6Wf|y!wMo9!|LXpB_$W_fPr;{PX_^;((6#zel;!oV@;jC1`H$v~CX& z2>~1DUjg^O9{lXwKqWNrbW&(8AW(AyR{FpJ@WXFEd|1Gx{&(NJ|DPH=4^Z9f0Vv7l zq=1sl!}jacUjj9dP~QbSLBIi2aQ`jy{*RW;!|}Ua)~|M)zb;Aul&1eHj{onuTz|&I z%9})y3rO97*!M8Vv@XEoA2{OqJwTxF{I7}oe;favP}+kM4$BVMls~k@=K#vW|LU&) zIhi-j(F;Tdf&k*)FQ)@upp+Uo9mK(!l>1Qf&HHCgfNJ}z92-z9|J(Z4oJe5VAb+hB zc++a~009+^0dN@juQJX7xB&dNZTPK>qcedp(g1UNhBP955K3&Q$kVo=KXAZ`O|&Hcxo9tbq}SL)9ND5&gdMj3x?^eLD?j4-^v zd>9@ib$BjdHF(k>qA*zST!4c<$AbW?@LRaOVgk{oOqQWYKkBdfUGe34#Psv z257f`m1GXUAmi_>*x7(Z_#@wQ0tM&)oz<^!05T;I8E{P?j4vkeWl+w8E|O$Wnw@GNW_0H|38X>oSXxIENU8ou{!zz3i*3w z!1@B%8a4SWY4~C=s2Gs{u?1}0Jizj&jmH7u2>){_uvGzJ11)~n=U1zMP>I+94~V~b z0kD(*3Ek|#*83GhK&J|#1}4OnM8V9QB*21#$qslW@csoRI011Gkh5@Dfo5 zDWu73fG9y6kYBnxuyawd9&#K0Zd`VtA)xc8O)>(m0dKBhFz|sZa{w3U32aRLTOb1Cm%#t02aX3YNL zq{6TebN@3_fdGgF>z_ap4#)#&kVKD2okkuClEwjc?EgI!b|4z0oQ0MH@<;nuHgj=! zo{v=TaLILFdf(@|S=v~zT#DT;h zmPErE{SOPG_kP+l`)CnAA~#(U4s#Noc-RglUMo?T>8rJSo$pQjT>f!;wUYU6;^&HQ zwx5u%UomM}1K-_ZOeeIZja9X^vZl3ajkM#qL)gdb`u)%I+aRkqKli?wK%wW?aX7tqM$Nj&LPE%7`2x>0|C6XLI>c-UnhRZF1PP{q-&sSRXb`?k3qC zuyRWBy1HN(Ro9Lk)eAB}~k%Mzb(saY> zfHFMo`BR^XduxeiOPe4~=IUC2dyTZEvbpW#pmz_PB)JujpUEl=t@%E?H{Bgyx^f!i z5nf|Ib~yK;gVEA{p(w~9iZ%BzFQ}h*&AHuG=NQux}(7cWdNefqO##*voEV9hu!S@e}kZ0r2AbHkCzLR+8MCq@<+N{QiX z9rwx|JT^v)VzsZ~fNm-EBu2nMmO*WxlXR?!^#G>GwOLVqVQU))c%!`F(ld_d^V>*- zw=ILQlhy*u*LJXNK_S7NieY_*-xxsh{?(m*+&cqwRNI>8 zcA>9X^&nX5QsyNPK{NAKY5OPX)*aG?l3%=3CWNgv?qe#3te`?Fv2u31V;n;iy+KwV zKe`ZcBTIMV!iQcLAEOs{BW!P!aL@1wyu6o%JrhWIZG{<1UA>%1N3C;hhmVb+9kbyG zUWtBoyk7Fb5aIeM!YRBVV%$Yo-!4J_K|c+oGJH3DzPY8NO%YT@l*zi5f4=@o)GW}r z|9Mqlgtww#v>=EUf4&H z8{PazF7Y#_>=#ZBTQoC|t5o}P`3_HdP3F0x^j!V*EKZDjb{Gz~@QZq%rEj--mLtqp zB=h6QK~&_DY&H_35XX9muBKp0>#-d;%X^km^J<8{w)=mgXWoQA2TIw$K-m;i)$LFa&Xd zNKrQl#6^lpY-spMkO)sAQlmd2M^ud(lIc*=c|R%i-E7oZ7e7!aOlYKuLS{a_7rxNw z{35;dB*wX#j+hy3FdAPt2v*Ut)A1XVepJhnwU{F{mKuISfoD)ejb_t{u@P-7fiP_% z56Dsd$#*&8W5RPU^I{fse=?N{3S_UzirWyqxUX)Lj=1 z1lg%U$AF#KwY2@C(Y(f|1RrK)vJzdH_=*c;{;5k}uX=uNOk^`EzdeEqg`fn>7$(A! zySuIbvN2{69(ZWF>0)k0vP$YUdh0Veim~X^+_sv zB;&{QQxt?67BPQ)KAxh=CnzcWJDOauI3nfNP)XP9aFnyRG3lRivW_ZK?Xh5f%Yk8 zH*}|Iwjax?q7p5qGw{)71W??7^=-!4K72dHx8o4e_aL8yN%?z87{n z1S#idAU(Qpt7nyo7AWTNsgWm3SRAs=frB2!^-)rcTO(H#zi}tCv3tZsEr^ z;V8FAWX!0Z=#pmm^_g)P9xq>{p)Uf`?xx1vywAY(eo^A$u%-_!R%?}kr>3*Hr_^=$^T&pU5tOs9@P>yag zqXK(dNN(F-5>z9E5M=T>(FmpZ-jFTk0;5LAaWonpD;@tuOc%aW>BfCeDFtmMM5_Gt z&ejCEiHf*?v{=#fvs>bmaoIv9h!rMO3_s9Bs@ZYW$#{JM)>GuLeBKSl2iXk7p7N;{ zsf93Q6RSzh)9vs>sU(en()wM0_OK6nbWO5bBvMSefNqaf ze)-#W)|8|SGRnySn7xoihRoQ8FP1~D90AcUI7wJD`v-Wm2v$<90>ZdihY2^Jdglrr zYJ%Yrp{l1iDn{wi{;U|X&Da`qU`LXaR3rW9lY+He%n52g3#sf>oNZ>G=U-C!(67y$ z&Qy371t=?BeW}AHJ{`9+QEI`_ITc*BXL>JJz|!auBVk9aM7{P)gT^z2hL^lH zk^dr>aGyMfo|jwvmL$2ROAbY!_%#FD8e2Sp3u{HAeEPRcJdh>BUM^fziP)>^c-hsD zjKi&^2+_#)YJrVfh3R;gP4qU*O+=_;toiS7HLzHHUVic}ReBxho@8V=KelL2O5Mwbk@$Qw^$K5^uPmeC zwGHxY3mqmoWbR~ao8N2vl*fU1>eXGg^hP(l`DP?Ho1MCeb5ag|WXh?;>24?n#&b#fiE;e7~0&#q!m<(HJ+U?Z{WeLjSCJrylfj*|S=l@znX z5FU697Z34W99iG(eFGyz)YA7>&Vaw#dhbmE6Fz0g;~>*!{7EkG=+h|Cb%X_6HJk57 ze$eIp^zZuzCXWd8EmxF!o*CwH@7i^*zXg*!u(alX{zymeP|VDN4RaRS7ma!o~h5nJM}K#g>OYw#@= zGf$oh_M#M13@K#Fd9L(BRCdX1-~jnsL~&XIOiD8(TYXI|Ir+yQbJf@f?0MOh!D3!- zihY?^r=L+pFgO-uX>^%>ZEHy1+LrNh=ou+&qO$5$LL@fBhwfr!p-31NyTFyl`h%?> zcf#6{_e=Es=zeBMj#nyNjfyz_8k5?T074Kd5?T`*dT@k+RP&g6+anJS)0mEW`#m)q z1x5XrUS)*6C=ro#Fm@XYn;&}-DxpM>Avj=SuX{nCL0oPl{Xm-<6xs8cy~WQSRO4ke z+JTLQlJ5f~|=*d$R4`mSP3x;QXC3#%C9HdwwNu{G-J+3-6mwwdJ1S2=p zSa}n%7CJAssYsi}S=EXyK;`1w_fGj6VLd;t|5CXQ_o!q++bh|GA9+-@l)(^6Dt^VElH zET|M8;87u9pqI!ttMx^cP$OpU6XQViJ3aeo9)3bO{P=oPKpn#;tBD8$fz?kil^-Es zC?za`aUSOCv5-l7^BfHl>U{d3I9BC|P~Rkrgr-%qy4z`nys8rjM?|eCN4q3@*gzt> zHVNb|T?rjS)6-OTp7Pt|59X#VLXhxGRb2*&ykdZoe-Bca#rwEHZy`k;A!>-SldEjp zV&8$%PWnD7LLGgRhPAX|G+!s-u;xvQlk#_} zQ=VK?s=|U6M+Tcy*cXw!)#dQ%z48v$yEX*Wx1Ys-V%#c5!}NVi*BZpIC0~7OD#zQ8 zHV9`#!))XiG#4lGYBI@YcPzr{>N0-K){+owa{$cm;Yul!&*sRZB2w$4RGKI!@e%d> z!U=U52>B{Skd!eL?(q>n;g?f`iQFKK|JKkz zguIxitJ$h6sPp`>IQMW;@iurtwe*ZUd1|!pQT-0{>Z7hWi_AoHIlCAVsuNZ$=M1V` z!J>H97mZ@wy-ju=M21~`cnDl5)KmDU%7x$JBDPM^ZD_5G1zRIo3*q+a?e|%7n(!r2 zp*be~Evj=O{|{Sd9T!E^zI|D`8>IuK5AsDVKAzHHKe%P>565!gGBJ{Pe+3GM6rfC)*ywa)rH^8BDUqgF47# zoJ+lrJMn8jpovU_6+}qz%4wGZ%#*@{u=rk(xsnOQoDn2le-Fm2cCXB~hk4InpJZSZ z{^;W%TKG!cJkJL_GnFx0b33%AUEUqfeU~E?`V>-Vb^y1{;;O4Y`luOebl~gV#Go2Kfopkn!h6Ay>&m3KD$3@G z(f+ItH%&Nv5GRqeAd57+vlHq3P+>>ybBD2#lGZrP*8pI+^_^LkEw~@;J z(L_LFDi6*2A-1aKS>PxPTWOgdGn&-pC?5~y8LK1H+?|5&m}Q0K50+6y2yh2%-inP0 zW<+EbM?$77cE?e(RjFq%T{@BZ=9%}}%hqD^OBm5BK|xd`c62)K;4#qLmIQ@{Ri-1C zRbwFirGQv(u)Y}qcKH^MNM_~rJ)NLJ6`?M#Z<|kYMwDZPpP=i(_P`~h3_D zxJ8sN&zEG_^Ug$0Q4g(HL*|IORh100O?e|q3I8sZ*Fd+_*KTx?J8HD%He3@8OOKD> zDBc=N(J`s-dJZA)U@X!J`GkhAL{E*<3Z27NF_L=N0~Z*y?Fam)!s|~X%-onnO&{QS zoJG4#A3RymHmKFo-M%=Nn~hj8aevEAropJa`Z)jE`K%-A8RNtE<4vREQbOA?&T6)CG)Y4A zT)SvP@R`u#$@gpn0oGwQbR9UmSc3aq$U=U5!NqIi=frA2|z{Lq;rgq8kW1>4^x#D#1ZXg1|132mA3_d{)7TD$6P#g{J| z^cRFYGVxlswn&WpqQOMa=fl+?18tF|sSKg+5rL%;>C+5#xl1-E4&@hQDnuYaQkHib z48Qy6el~rVUhb+ybAp91B^Huq;qOd{N1w=RyD&smS{cq(#xHLR(F*O339Xm5mtk3o zs`s0l6yO-aGO6FPDJq~?o)e+RRZro%US(!x;-bmZLI{1T_oh?VT>T9J-%#q#?#?oE_a@jdfgFLDB%5a4FkkJItVy4<9k5lnCKv4{gz=^uGfS zRHyE?TIziCr1^8HRg`t=Xf7d-8Ce|sQOF7$Zw%+e;Dr_ou?&$?!F@C^>a;m@*{@n? zfM*6o_l_;t9zs~37EcC)-zj&#zv7Q0`$PY+@|2EQ!+m6m==eRxBcx@b2s{_1tt}*2 zot=8h$`To65>wRO>`o!sd)$>36h%>t&(fCl2*rEPB337G?Kiwh`tHPSD3tgwi>ZmZ zb@Y4Ib-OcNo;c=)~=nzuvGmMjY_Dfx_0?8bqg}PrbNnvo_{TXvLY6(U{DFsN zF_h;9EJ&omIx_F|mdWUHiKh{Q7w|8UoEeB*&7k`cvW__<#RfCm5T!MiwBrrWjk%q! z4bOM=YB&nkG?i>~?m@z_WX}S2oMELII~jv8(t~EcLzj17J)um-yax4BAJ;tbVc#~5 z9pMyAjCyJ#Qz=w)lRcz+m-J0^lQzc|7-ADfq`i+8Elj2fEg^DW9VAz)C8k0c9==NbTSl&T{hINeRciAHvdFVT zq@(DzQ_iLI(7yR>$f)F*<_7PzglQ zqk$+o5ROW%1W~8b08xA(W1X%Xh}_?VBcZ<&h5r-f<^vMWfnwC)e*$Fx{t9Fu1C=83 z1G?{Yl|VF+hxhgiG9Dmp`;Tz`zx(t3Zp01{%I5=Y)Nh-_`R#T~;MW$-uipfJRm95= zr0D)u-T$|uU%deXvAMq~;s5@8(>?J2---c9%D?F_h>s86(F@4P^8q2%n`R_QAm9QU zzk38&Ex_~2fFAMw>JcDdFYs^ppASgqzwNq!pa48y4KPaJ0}}6leFKDzm%ML^{^}7T zuYm9kVE}w_+bL=Q?YYh3hxf;Tptn}@1cE@g{7<<0HZPt(>3WdyuZ}zkNEiwajs+=G z2mpE8f6V29(D+|R0s&#ZB>X{&+Y*d|yxh)?=59A3O@3h>AZ=dYzeAumaqL~Ly8vYH zC!1V=|39H(z8f;*CTAQV9)M);(a#|18{-AM+e~v39w9-2|E&-X9YWwH0gU8n1%T}9 zKj)qY2q^>F`Ta$zU;f4=Kg0SHIfQ13_Z6-F{ZU&?Xzd1-$nTVp9OA~8@qyAPyk537XZ}ke}{v40cPM2jlwU;^BZ-*hX`=n zIat`dIxO%4>H&Xq^nW`n{Ee@Zc=8GI{7+8>ZYT$C${PmY zSB#wxKr;Sxn+Fi_2dp@@VgLf+kNW@-u^XS@{|q+Yt%g4k@5iL#1Jn-w*i!&X`F}68 zTNjdFSjdgf0I?gtu+aZB2fzlnsDb?SzomEqW=QC-K?W}{HQ|1QKyz-R=)m$K{6AIv z;s*Q!@?dWdp%H`zxh}+WK z%^AwW?doc2YVPU!$kf%^&Etu?8<*uH+y6co|KGHKa)7}2JBB=fE5qOE0!(#wV5;97 zkIK^A!P<$}>h>`ITNhv=aTkzi_=5%U0;B)OrhylrH2%15+-QDq09?bZTLXvw%__lv zYb*kH*1zdAFhBn+ZhU|i0{<=209SvHpnG$*V8ajfLF6Dl{u?h5xRX8zn+zaX{#rEv z_8_WXaE*9P) zAj1OAAs}&aV*~PM+VcXwHh=gZ9>68W%pdSdkuw4j0T2&92>@n97y))5 zfN03yoq$^aKNOx{2U3OaA0ptxk&J;Q5NOzqg~aXjg97d&e78dgobDUVir)kV6mYJ9 zz{dps08F>2Pn$7_?xysOGDw9lFnOWCfc;sdZpQ!5Sp-%IKvV+427o7wet@V2paufy zX#lDdf6QB;fR~Ek?>ikIu=AfDLIIJBd{Km7Ybbon4@3u-^8*p2k|R(t0){U)K*hhx z4}8C*8UBNMz@^C%Sm9!?K{!cRA7H;&k6YP>U#tfo?=PkUXvQx#gYP%f!S@@%;QP&V z@cm{w_S8a{@)b>m({QB-xdDW zqTu_jMFCK9zkmN#;a|y!U!A|{#XnXS0Mzy$JY)sHLrz}r3wInx(Y#@#MXNeYxFd|D zKLq>{MxtPpjz{NtM+cMCyU;E)RrKRZtTSrv7%87~(6*}9_dD8}ifh044j*3l*?e}e z0{{6WK=|_PU{Lw+^OMXg$;^WV8ZJ|ChU)a^VeW;8ZBA|NmzUoUciR_ChmOB7!*F3c z^I`#O*M!#>gYBjPUtqI?*D2gIKdUZge+qaWUd;WxaytaHZLd$amk#@APLJPE#mrr; znqIVnCw^S*!ZM=UW1X&7R^!?~U-}A$Tz_-x|9NfJG$MF~-5s$|vo6?ZGqB<5cE^A@zTHD70n?vneZefSR(muXaI$umeYJ=6 zdae#zV!lYS`l9M*&YH(;`uh{Ut69-qp}@OwjwnK`&e)-^MX3<4nEtYq zO4_H8Z)|D&5{UbIi+i0C>1mQ{L@(+DctJZA2QyXmIIm;6zOJD5hz34|VOii}1Yy3t z46)P_U}E~xmD|_gAm-$@p762dUj1}qmHrZKjgd3Kd476MsAF{3N%ukSTJ+)`B@RAi zcu}Z~xS{hQVGHrl9Ajndol$A`k($PWGy!8#abM}RaU2N<>4r8={XooUJHMw07$eYi z10l`J5Q|vaz8f!OA4d!f!1`{oM0L z0@#b)J?Jjo?(%6b`@-)W+RjLEiowMw5s-87*g3M&o9ZnAL!J7mRe00JM)4YD9Mb$K zm7hWt5rt5q8F#wScFio>io|2n1*W6fXotB~%%47IiOHkm&U!0&Fs96jBkN1cN6$^s z-dcG-(gvR(EI62Ec5)7xqSZwg_&$LlovYrJ+)rA@q&pT-`4X4=15ctqqH$BxtRj=H zW)^~HZ6?}xkEa2f2$#zB%L>Z#XSx%eO%8~q>a6Sk}4~)>V zgf7jF_zhx?h~r7H53`)QzgC%GjWT6nc7+s?Bdtw%$~uF!Z@efQD72b`J#1?JzRxej zviaxwd+Om_A*bY5kT{h>lU+z8i?JnxR<2!?r2HoaufTOpLB-JeQP*XuF0u}*C#6NS z<$@F9HtRESwm8gmd|jBqq&e@L!Qh_)`1#J6)c4TQi7su0R$)Q)SH>lN-{wDeV@5lQ z*l_i=9FKg_#dDK=-Por7A&S8m4+gCheNra0a+IIwLB-{yye6Zsb=e~XPpdk^{JL$7 zr$Y2@+9Sv*^ig1VlDB?qR=U`Bad#mt6l`yHO0mAZzi0O2g~sl1HbIX4u0yV`=;-~c zFPUaSa=WS#)R8bZ;UM3xMTf3EGahj@sZRee=0wA)518C4?!qhTkzArr^Oark)d?2k@6r4eeMy{xx`V6GIPRvsq9|1 z?u-&HZ2_iCRm_Ec>XT9A@soh+x5e9x59f5Tv0@NgIKG3aJjH(~JoV`ycZ`kENDL9zyJv zXVu1AQ_JFHTkJBx@C&!ZU{CMjE!vI$BbL1Wzm{q%@JvF?EjJUPSx zxv$04bppT(7oO)SXLL^&&NmEB&2|tdBkT%6 zNbY#W@S1;VzAGj|IKguO>idy~6kHialBC!H?|HLJ2SdJ~@#<(m*ulTt6Y)Kh!QO{a zcKz^ULA;Km_i!5w7UWtDlFW|}>>O_4Dn;W@Kq~H9)PGrOOThA8NJej!aO*2-#*f)n zDJw?zm9(jbNB8(HZLw@PIewxxxFY3@5E*W~;Fe({3@4hIPqSx}?ZWkf1!KLa+Jfke zE8c|!8p81PwJ~37PoWn$ABiUSXr;XFe7c>?CPUu>lRU#+UGId`FqQRz&16D@$$jxG z{4Vn5aHOlM>HRwS9EuNCW=KNuc4!`!^ofUoYDEcZoFy` zLl`!-@8*5Q`wtH4MaEL-qj<9jA*26kzouM78 z^1%mKqyV-I(WsdEQ5I&~InSF1M7}ASYH>o@VxF1em*1L=))sNb_I*>U_&{;Ro!#z}KmAAe=hQQaSK3$&!$kYpk7jxx5!qd~R zTs(@C`w&S?a!_dXKT*=hn$5`VsV+RgU$szo@DM298ks$dy|_eJy>RKKXGwvwzD6CXnm{u%7lePc9XVYpU|tOX)$p53xf+ zlC>e1RT(b%2TPg9S>p;L>7VK*%(rA*PSry9+GOlq!X6Z{y>IREs>8%8lJM(yk1K{@ z6fYih^5HyMlOp|yCm)EQSwt+VnKc>vS@v6)HQhgFTTeF~3v*0BxC&RBdY@AQLgr}vN0)70EyT|bhH z43#S(rx?!`j>XkM&oJ53>S@Z%($i01x*qq9;KS-J8|j?IM$hutRG9ZCk|=tu74G({ z#ZNrGh!VmdDzFsJm%Z1Qa(|*zq9Rxbzva6jswGRDe75P(Y}z+>SIj&9b?6@neMB=` zv-Sc-a!_SEpN7a`yYUkTN?0JNzwHh$&@URVgry~REW%UI*zz16*cz+|S? z?dZwkC9`FP$fqA0eIq-|Y|%hH!^CPM`d%;oK8w&|KiRG9wOk%OE~0H1Q3SzUF&UBj zLuK!}HuLv!1TVf7h3@Hue}K3d7*O7?HD-G$q^;Y?FYQ4u7+axp^#((2{|EL=+Muw! zL)zAEbHz%F%OgT;IxY8k!p9{;SsbuOOj^pA;fy64_?LY|Cf71e-*GkYbHx3=ak_jg zKTGD%{SH1O5--vJBA+e%YH8VscH|8KUCW1x(1;K2j_y;etTIL~!u_bNqlVwQ34@8d z9<8{RWDV;u6d{qe=jbRZXBbR}LlMQ+XNHNJWZ7s~RpLz*p%EYt7b0e=S@1yVPg5p9`q4Z z1LE%%#HZd12H060y;1{>Vso|8!t~^IxV9duB@^;AVGyAgRzi02*C^KqyY9p!xWA=y zu+4eS?-)3CJ)H2o)_z?}lXOIx&HED|BM15|+bE5T>{-5#3%z zkdaTTpcbfvGN3sP=m?1g3%?n}*tLBXP&+sIPR~8gXqkAirmKkv+Wanfq1-R-O9eN+?nHD!N8rPa>k zEb`M^CNEmiw4|cFq{v#dDe8USWqtSzXmoEYeVf-3SI*_!pP#?34wmi6lRu4I7bt9n zb|T$ulU*emx)@FC9zawac)@_K^8jUe-EX_rD=oF%eqw2RCgxjFU#=wkQa1WAMGUFm z@!`~kxd?a-6x~|j>V=r)ODjj}Q6@G0MYb6`xK^tXwFo~p5DTIjPWFjKekzQW^sY!r z&nFQ5H?8ugDmbs>6Jc&(Y*@{OwqSB41(#__f*Uv-(kh(tUKpz)UCrva@O9t-)XQYf zuEU^5S8!0~pp^@hb0MNJM5iH=o*W2*dWkel;&5R|P=@?q9Z7=hL9nZHsE=dX(#8ff zTCXs-(ZhUgA(YVO{27?H(pD7W5{UJ_X>EzBPAuWwuXiHbvpAVASHMmV9&DX|=dc|e~h9ED@U-?hPVJF6o(Fz~P-QQ`(TDUj&c43+MQ-ddx+H0gDUfsdw|GX~MF=mBbT^FLXGaerz0O^Mk)&$k(to#l-A$ z${tv5+32z!>69rIWjY-xIhxCpF?Una-V+L~JSY~KgoSfGe62*aJ?zdl?YgsOK2Y6V z%dYW!u0E)7az|ynVbd%s#d5`qmQ}i*GP4o8BAEpx(Zh-VC#ig0s->H9OR6Pz#^GwV z;=aYGPdJgY_m?{bSGfb#Mti`?Ox!*UTIZIHbybx#eO|F02xYl{6gqiDjdDrX$E3sB z{e6-^6?SDTSR46ihp=sdnV8iqxa~cSnI?d-z zv)#tR%*|Y`nr8;S_cl}ADQISn498nj-PgG{N0x~bm72W5BtKs@7S7fIPgJa%bwQ@w z$@MTuKwBXNV!2K)&c{cVJ+IILaPK{%-{LG8#L218)#yqd$M?U;ic)yTgY+xwV zj?_2*nP|Q!=|UKK<(Z}HI`Wia#PM`i zRdvHB^w zsx&cIbU|$e7fV;W$CTP^G{?H?m}nZ(mYIOkS3LOz=RG*sF?{pM*GG5ug!|W>U_D4? znN%y(L+%{OAVzuq2zd^$k4>8b{7D@stqvV{ObvYh;|G2dmf1;f;V`-i;wVm%d8eD% z#Oi7P6bjn#Q5#0A=7lEy2EA?t2kon{1?ms~PSv3GHwh0#7Pm(TiY z=7?QJLhJ@o2ttxgjZuP?Og{3k)Uz$Dr+*p1ZT8e8sCV1CUmJJ*lJMmezrzR&+6>dV8O5%dlv z(u9d`-tdmnysYEz6MFUCIWP^{7rvOZO?XSy6{k5ub?2wlb!=pRD>HZo4)Aiorm$woXG z#qYU8hA!JisCpc%0~0(o$|=_fr%-wfjY3;9)M1{ZqpsL|F1)jrZ#+R|8NT^Iq(L8_ zAn)g8TP99e(-fWxxzZkfw{QbFKE-s@Pt8+PGl-XlyRz};w2va*lL!3Kc(LZ>vuJ9> z>Mr@dP-=v0YH9g+geU7;H50TS6SCIBwC%I!42|0~5G|;}Auw^J0jcu&bdC#M1Nr#J zp)UdQ?6K*t`f<&;z!w>g<;_Lb=LG4Yr@=DSgaHb~RY#>q+UyAS!Y;ku)ko&8 z=rep!F_iU1Xpp&iWh7r950UT!Ml>CM$$(I0;5U8+XuESrok)rz8}cadqIu!1%#zog z(BcRd9{%_k6xeXq`Y^Rlz32Qqcn(npNj>nGAxDsb*zZ9^n{CGCrKQ}@4e_od7KiRd zN6{GIJq5VuS+DA|HGLdgVJvw zmmxG`h=rj=zms5=DhDxUjCo_MbL2ZXu$aW4{$0YWeM1S>Yr+g7Yil)|b{=FGq`3?q zEt%T;5yY@``S3wvhBV~pyI3>?CZqxkI2{BTVH=+Jo(c~0Z>==2@?mG0cX*R(&%ZJ< z#J}rBID(n;HdD00aTug*&c=rJ<6(d35K-;m;D&<5lqi3jcA#3EWWsw za&DA|o7>GnPyB@w!st#c<@qNF2qpWdL&CBxvxzwhaDktXDLoeII$6`i zU6xLCuh(Dqh%34Dm&mr{y4jp}sB3t6H6*0>ghDj!ePrrk{-j9ZUhz2{o8rc3oshBa z2;_-`t-}uG)u+ZA#H4(Id-51w_Ct*be!xq$Y?S?=dhX=T$Npa3!R8l5JYcYGH^;k; zUFAF&MzS7wLiD~aj+RD>*@oI6dJKAuJexQ|Pzx$DlM7n}E7yG~XFCszVnZTAvvj4s zv*z7rBUT^4P26vj(Pyd`=p@&3OF|WN zWh|z=vcY_JbKCT*p%OEmM!A%?6tj*7pW^E`RJ>32)-JfP=I=GG;8I?nt{Q1!!Cb^o zpB^3#{zRB0^3VAHATqb=>bHmtfHM3=WCZx(E|UNm0ti2J_5vmgfSv%na~xo1j#L0O z2|fWphXUBTVO$_lH{b{WHr#wc#tWbYf1|j?MWx#V{ptkV8c^Z@4*RPI@KWG6xYjrjVFkW2nC3ln*oBeO#oZJ^)HCv<`Wyx`adCVen4>LR!_n)7K9784F<{xeFOnP`ym9t zxB`fI|HVfBk!;}S7lcD10l5o?XCOv6=`=|6{}@*A0xB7Q;_bf;E2M5w3HaPIkPw_| z1_YtG!43YktKfzFhoSjxSD^%N2?4RgkAdA;fPwlmd;j&R;JML0`NO#U@u^UUKMw^W z@{OPVbb}8dH~tJ24-nG-!|Lz>2m_r05XJ|%6BBsQXJB5HhXL9ffCj?f(+b4yxd0mH zcKG?Af;Zw58gM{+0}sv{4$N|$S)h;?0AP>{0n#tOsRUj?XX4NJ-YCl~#R6pjN`nQi z@C7&>PVCs^|FCK|_9}n(7Wiu?ApmcU08zs6=K$~vBN9L?{sq{003PBG6U58Qmz2-; zn?(T1pu?jg0cD1db3pe2QVN?Am_UDYL3sHE|AIF+llnK*0kn+(t``OLCx0F&l^F%X z#`_1$;{|j=fOUcy5EJJNTfrHpD0Ijo+24NEL04h6wc@Dl?MUPm0K#c~F*T8{q zL<77C|BWCD7aS4;!Xo0i6#)T;>@Q&gz;j$|ASpOJ2510=7$O#YPyrxEe z^a%k@BELZp9?0+0cdJw46bqt9SU(8kz5gk^F9aB z!^tQSv65m&(cn^VK`azdzJE?1??2$&FPRWI_;n>n;-(A0asYqw8ki*iI)eO1wg))S ze;h#sZlrsV;hV341rn$SSPHn|CU1dL2?*Ex;(=~~9HGDdDZr8lM>s|hhUZp+6mMt} zCjMXQ8~@TIz}0aBU;L#>_@RJ<2p}B74^8}Vs~ZC01-=4y`3;i@{bo3TEAE!*5c&;; z2>oU_gl-uQ_?0-KizXnQbGs8ivIZuAhG z**`zqZwiB*9-n?&f`^-`{k%LjYg+$#2)pdxTw6^)B)mS_mZWaWT{woAo_*g`FNR&5 z6!>4RY_?w?pGyApJzuFmJS_UTbQ12!cl?7Jmix0e%?7OYbByEKq3LWo9558=hh>F9 z#3TV34=)?==W5?x$a3bf=V9jlBu%@7-%mHm;qGZk$&9DRYm#l(0pv?wAI}23VVCE# zE3llkH$S^qrrX6Em=`={_i~vp{EB_sc*U?SCq7MO(=1V)`1MN)^O~5>pCMe^1TqA+ zLm66lVd@u?O%h<~=-nEW*w2gHwVMkoYt{bOK8aORTjH_1mk&6oqJ1yq5X4l7FeXlG z_4QAm?ymoM)vsN4 zN_u_0e^;^oaY>vl8oQ#4?7qoUKD|em>~bjFJO$iCm+y<7Zu#&!96Ye)kRG*g5Rn{i z7_6Ah&W$)(cERX2TYO5C1%%9y}a2)6TxXzo?n$6!;CC33|@A(@50sL|B&u)FCmR6 zZiqdnjpKAMdf%uSB?n{1u>)owt?qOAQlCEm+*Tr#Fp@d+({^Z`3`dLi{Zcme0O&)l z5>x*Ahn~;Ac+46mL+h);`gwf-mEFoSmmlqo#1TaYV$Bv+#M&puyoiWY=qp^Z%dSB zh-PV>JWfo$(<@@cON{m97Z1d5%6^4Y{n1Lp(pJQTxd~?EpK}xTw*smI%!^n;` zt2U*n~Gh@!YcY=7LBC z-Ay{^J(#G1r!UTzo(sZQ!v!8PUz+G|FN!)l2-Y_Q^FIlVri>-e=V|LmKYo0f6njS?x-;^-45QN?iJK_1=1AZ zJ5YqR$Dklq@q{jgV)I3d%7`C~iW@o7HF z4^ix;m)el#1ksq+cv3AKL_ea(5|JJULsQTknVCkW)X!AR_aMjO0ps@fyYnhnIObXeo>HpdFSS>OuQz$x51gE3 zf~0;d$Q(a}iIfEda@|4PUJ;I9XD`3rwwK=HyYEJ)ifQ_?KFFQ!B!7D79yV(2q++R@TdsexG@r` zbX>5ou?Jz|%eXhq=nzIyv*mK_jGSxcRB&9vC8&OeDMIKM$)tzswmG2IJYY6qbBZFl z67&`LIV{FYPXvNlo@?bt*}&}i_|5G2TvfnQp-57qsaGw7Qy3ylJ&H z)FobrdKyj?l_mj`vp5oXqz(QfADq<$yjMV~Wl=ub9&BhRja&z(X4&j-M@-(5yg_7L zPd$|y@ER7DACN`UE?In|yTBA#4*Ps3>C)85)B{#w^sodWUR_q7M%Mby_v)v*+w60+ zOlpNa2p_L1B9MYB>|Vv6B)Vx(nRa`zoO3LNXfrGNLyC07Lkf}-2})kgI-ok)#x@}L zX7s6dC&d*a;3`)R=&KZK&(T~1!v${WD)vDF3GgVF($wpq>U^p6+Ye(Hwr9dCIXTxgh5TJK6#pVYF9Xx z4HDd%8V0n&%S}siG|-}eH?D&mVxVAFwW(*8(S=G*+Bvpl83m{}Bf2lR zfbmou@1TMG$BUDoyBRdC8=@58_D{{NC#?qU<3Acrh1Iv6H^W`;;5r(Q-iM0dtr(ia zun^c)r!3W{pLmBl>! z7I>}H^CjO;p{U25wX}oj>>*ijr*FVudP{dwN6oO|3=#zmMgJ*|vE9zFSuK|{zPT>{ z7ES=}aRlKd<%rXZ0Y7e;Q;Sq3zwI;IV+iGog)_wA#*uJY_&aaX*%+D=(*N9tsepG zV`q25mQpI7WM}Mnnm{hn(J@E1ZF}c&iD3aGA>xa`)yw+l-7i<_`D`ZD1UBV_%3i@A zW{I!T+P=6~P^_r(8W0lw6Ln3kjsYi*7dmoSW zcaGOdMb(f^u@uRvN&q*-Fvye3lEmX6HX08Z&sPJ!BcxBYKe(?&QK;b1#HBt$m5FWU zJND-G=2KJF=uV}pdi+s4gGSFEHsWF00)-U?*?vA^Poao;KE3l?+K`z06#ee^F2#GvBk{|!VH(aPZu}*y zl9r#YUy;qt>x-+2*x2EfG&Rbs8rYmg5>ZKK>H5=`GP-!2-xUDO`)lInD^82DZ_#n2 zIaXso9@P}PCtsi-zD_{fAF@nT3h}&hbn&7UIWaB1+iWZGF#5+~dAqS~6uKg(${2ZR zeyKch>N)ab=UH_OGp2F75Ljz5fqkc}*06-S=F!uMrqaas-^kWmZ3!RqiK-Bamt8fa zAjvXG?(Nq6;EhR!eV(z>|HeF~S^|1!6o%j#QZ*Kn156Q80ELQMwDi?SZrP%kIN@6j%?aSA~!zkkU3VQcKShkZp+fC3+9C zuf!K2rGfCDI}Gh#(RpFL=^lqix|TUzHvN|DI`kL{BuE|Eioi>)?_jTU=X>{`+wI!& z%{?>usX24>sbYscS9yO>XG@32PTE5iIp2Eu$h($I(u7}nh$;DDx2)2l#pJ^4m%$FF zf|{;Gh^9wuA2N0*L>>r3SU3gjG}_?BxDr4fBdlurbNqt+@(}x+OCDh z9_n!CsD0tAV@7D)B%&l@Z`Mq8!JtX`3sGFjIXhM)E89KywmdW2qHYu%amY(;X8~28 zr`w7;%!;jbs~Z=_g?vY z2DO=rU%$w7(NvmuhP}@qnSVvI(s5ujXSfp#AFUaEwYsi3WF}_vR7#u4B*6gb$sv;z zG7U;Yr{-!s`nBe8(L6yBLy1pVOsCmlQ#g5&Z=Y4Pc|uUaW4iKp0 z;R<|9y2+`tzj41sDY%cdENW>P@7sf@mu|YMqSZF7nrStJEqXAO&yn#xJbA?L)yL{r z8gfc5_21ZkIZt^M!MXAvN8o9mpe`XkSmgCSMbZ3lP#__iw>?Nvi~#X|Zh?=*omC<; z$$PaCe&$Eq}4 z3~k1!Dz?-xS;iGcW|iTRjE?$B-<(y<4uxZ8Z>NX91e}E)oI23^o&b*pRNXnK&4$)N zDL|C^tR5s0m7Yhajfh*{YvD?#lZiapJ)v9RHQelEzQz((0n7iNNyQ z*>ACakc(3*V@M~(7SsBob^i$9e0WBuFcB9d*RA-ZTkXmDg5-nHMr`?d9`V0>fD3^F`Y&C!-EbSw2jh9G@Om7pvv+idpglP_VAs6{&-`<9 zr=}$lfg9{<_s*aeXRh*7=Tb2G6cwSH@<{=?%GBoRfy$=frT0$Xl0g@PMg<9r*>R@M zox6eD`EBG3Ehokh?b?F5u?Z#C)n@8qWyFu1DeQyE*v0Nqjt*#Is3CpBLmGUa6h%hE z8l0NGwH6!U>j;YISrP9dNeI*#yf(N}bMwmhK05=GYIrM~R<{{k8-%+2bi_5F-C_FK z7&6p@U62*dp`)rLnx{0~Bvl!Qw9F8sPuWSKfvcF@?Wsh2JY3tKwwD3iY8$@sg-lg) zgv_3MD#SrSlX}ma+5al0v8WnCo&R7I-UV% zheG4DH7*sq15IshW{aXBxW7|2Hs&hHg=2I3@R*l;*0b;ak5mZDO@eyn8l4wpjGdFB zb~ty3jMKIRz-(nvCA7V#-({A$ctJ6On*y-p)BX4tM^}dL<|!^Tt18G2Llm{gv5|Ns z*vwxzgfEl`_cL^UTg?+0e;INfiLfOK8?@O?CEc>(_3C`8gJQUoL8S^?e6ztb35{A? z7v(lbS4qyYm5D|6Z(k3S)VKrFxj1e~itG_nS&e*x21`PS(E&dV9iMaYC&74U{>2Su zvQKR}f~t6y=peN76s?Z~JSj)2t+gstv}4tUL-25@5YbWtrQ0^7m3;mB3r#2wHNo>t zbKqrKv<@DZ&0D;Rjx>W8c@D&S2XmhDikQ+*s)ybh9UXu6SW%Vt8b{AV{d!1Wxivkm zLG#_*Vh9{yMox=H+)yWi$$k~4RNb=&qib ztYSlvrN#UD^^f`uKCAIm1Jj^R13RxikpxNywBUS)KQ zQg6^!>uQRu{}8nm&o1~*tBwiS7)Bg*=VXN%blMU@?{|`A=jXZ)BLeggk4yw za}V(aDQ2HLn9(k&C45N2GUKJsPFgD10$BKwLw%$<~u5z|!wA zE0}kBgbLwsm|K%?@GOsyz|9SdZpQGcLJS)XJQFm`Zb|usmKFZcy-tE)gEw6#QWQo~7ZG*l!Hmn?y$f!mZh?_|JvrT< z2^R1z&Z`y?c~QM5^>Oxn)23h~d}nbQG7(QLYk0IvDff`Y3Qm*p-QT4K3_AAw2yoLz zBc40(3`z=uS53abwD#VY@5;pG7CjITE-`bfuT=YymJW=dejW3?T>r7uKFR6JLctxZ>^_nI4Ns_~dUOpW*_E%`aSyy{)MJh2Yhz&<@g@;;#mWD?1KgK1~`+ z5AxW(NQZBIw^$w6k|ofd85}y9TIa&(C0aLFgnLDYDyeqC6OxibUdDS~6 zU>oRU&E0!V(~D1$*yG*hjibIt41<(hA@$>x&GuB@yrIb8!8xB4jw2mWURd`U@}zAk zH+Xw_&$B}=rVI{t?Nw;S)j8Koow!jQELRdO@3@n!C-yKkhlIgX^FoCh_a+ZtRPgFM z1X_l^e~`2dhwW59xnpok85jL>sK#Jpj;yRIqy@pu>AJ=6=D4ISzd=-S z=26F+Abjb=JT*-c^APQ=NUr!)QPDzuC7oOhnt~^>?wk7OUAb5)u}$dDjjToE2A$)Q zc%4N{+{1G4g=Ed^r9DYr=kT5$&%EM@G}|3Et17Z^6?j7|8?wZ^oXuDmd10k@8`lgc0!jv&IQ@uR5Ixtzp-WTjQeHiS5B&2ynD?y zr1LD`y=*D`3ghuf64b=+eg?jqw>f9-T$= zJTYGxv{qRWCstW*eZ2RT(yw>yUaZRHbDoob-#ACAcO%*1lg#kqwW^CsmZi_L2aCx# z6)&69J*Q~{4+k~bo43B!vuoy@1RZ(%?=6?U{~2`hX>2xC@mWT&1tZ+=0M68qQTY2r z#iieoLP;mIDkd=T~YVr~}D+t>D3i?elY&(U#3j$7`hO4bt11h!U%Ms*Xax{2&La4K_xl~M z1Ii<%U!zBg38JKb_#?&qzco`t1d+kIUt)r8$T;Pxn8@he*&j$_Q6O>|A>?4(e;F(F z3OXTZ{gelr~_R@ zaV>-$saZSnJj-D%Ds&G34$fvBS+(R>G#eb0NM%_=i6(*Ys)JA<$sTKg2smF4Ixd}{ zC_n@`Gsxe#A^<4_ahn31TL3CaAYk?2Y|te>5HZJ(1E(=S)e8bzVSwFmb{GMq&QX1Sfkg3IaB391wO+ zA)v;uqc8t}NdRaizYO{tCIMx7tr3*zNuXq>k?m7ZgMUP}PjL7CE3iSZx)y%*HfrsmY&r>7=;1y@TPk{_XK+TZT|IgytLTB-8p|g1Q8PWnc zrGn~bhS(H@lH=p@soT2Pd){Og5k8GzBZ!+I1n?KuGaQhsXaP>x-EwHJu(pU7ST!i3#C~9E{q{3yX~O&Y z4>j+d()D@xHTaKLhYcTy>xpKkj(=**)bG|e-JqU-_3`_>)+du$xLM0_?eggSY8!}C z@9wn)9@o RkTJ09;McA1`5TsWycIb3xNfHyclEhz)h`VE^$d1*`|sdu(Fikcqv zY6Olq9j8|4YHj%tX>F}lO)N}(c(rujVL0H4k2^-?QReZT-seq@O=I*@$+z;A%$!Hc zwqJS??2JoybJB^w->YG zoPNTE-e5nJ{_!G7nG00mp2c3MYDTzojpns(})E>})JGyw#(`U*FQy*QZ1vGG`xSchGa*{~BM+ z@Y7C0tnA9fyc7S88`fjH(#z$muJ75Bz6tQcNqKzS2_PCwEB!5@hX-EvP6_r)Ulr3W z9+W>6OwQ%(8pW7Yye0bmqlA{t!w?~z(M?5%{H<*IX%Tl?8=W1}LNyr|cfD?o(fz{{ zm4F6^mDS?$G1Uu?4stKnK71FWbtM^5S}&X&SUMKGCR-q4orI-7^Yit_HHejHv?-_pCC$Gj<@EE#_nI&@59g(ZFOR1S_ znMO-N^mNnK91%rg;?APF=MCRiWYEyKphEbJ+%*+=_IBo5ty(O0#FZ!>!~G=_6-rX4 zHWtBD6Vb`*mpp9Clf3XKXT^AoS$Bt_VIw0aLo35#CGWAhf@KA(44j`i%H3u63c0ho zu+Tk|nCv&)DSUgs&X$WxaHnmj{arwxHCdrap^22G#< zmha`PeurBaGxuuJJC64Vt_Ht|-eml!nraOl9>D)(AF+jC)IiAxmYH2yw#M`eRy&ce7UaZ`Lrr z+In8Q8aR^_R%ayq($p#IBdA;9x>`)E4=*3xNLrsJypGkPz*JP!l1ew!FpuERG`@(X zQ=|AbE~0nkzIZsiFy)QS%CiW~n?%?6g628GR56$&?;X*Mu!&5~_e2f8U(d*i!@S#G zG*~6LD_{6V#Jv2TYE+y|TNjmp!QzT3Y0;BwtBX35OY+p~lUo}WMs;pGP8l}Qd_r*B zOVxcsA!9n+NtCR!iZ3M7=|ei<%Q|!!0+zB^;|A8xNbgd_mxt{i1WnD`=k2ach?{8+sFA45el2DA-7X%BOkSEdoM8U|pNXJ@FK!`Dle%#0 zP}X49*4eLzJ|XqN@uD|NYJ6}g^MpL}MCTpPc+wig?uk6}qRK?(Xu2`K=>u%6^*oHh zJBCX4hKhS|9yI7i zs?|1S3btQc8d=6PR?Y-W$ct8-CgAxNklQpQ&Yv49GQ-xUapgK`<%J=*ao)Lp ze9}rme7l9UMI2VkWLnGMkFKPOmT6+WWVfR}D-<*Nm`r_jI>`vS1be-StyjbZCOOKa zV8=Sy)!(*T`HE_^ysFYQ$Mdgp10LU@3CeLjEPWG^5W5xDL99~t{oNXIlQemw?{f#mYx0No$73k@VN6`Z&HnEiu9FDBZt+KG0@OwZw1G>-X-FaVQy|@ zzpoJb{W#~bsH$aJlyOlRHnGVqsMhM`S$TF6dAA1Yas1qD<_U!xG|wQ`akWy~WAhrt zUbxR#qbaV~lcjF>qK(o8=HJOpdfl&SZ%5ayC>t(lpBny_cM^H{AN>SBV9=1JFuHtwL}Fe#$r;K2f5{ z$Zy5i%SQ-wKq%mc3yxnL{N-+JC_Vm=axmJsWm&|gY@IiM4nyOv=+X)Ag`$dMT*wbL z8AEqJ)r;XbMLwP1+^yW33f)$F4$G{>CEnkZ)vnhkVzXcs0?P z@SCFNK$?Y#_VbFWkA2=eSuoNaV^Wnc-jA0KWJx?X6Go@z;QbG7c<2((Pt~NLiQv#s ze6giDKZ5hsK|bnrLnS)@Q1Z6i;g6)t95{2y3b)O=msh=VC+?NHMtWM`ZWpJUyO1G& zKpPSk9aW*CX9TIc&C<44cVO11#>PJ=aOX`&i<%9!YxXUC>oHnLp$L}4d?npY=y`v8 zv0QpD5_Gyff;Bjw)B+D1$ioqye;9+K-e>eNM?)|dF86x2BVx~a#rqciiAQuc4*}Pb z3-j|EF)kzi*&-{()#5#h3K{#+y106nnU4yZJMWFkh#1;wl2VS|dj`wY#yT!+-cs?9 z98;S_#A-FrvG+7(V8w=CyLqVms6$}*$uPC^;=(;aTysab+C;k3*t44nMwh}J!0H@2kHp8rlxOC%&d~E?Vs4t~3^KOH}3J(#_YQaW;JlfU9O9g6CLW7nn zWN~TdnOzoFRdNRuS(QXH8@_8qZKv$^np=(b8F6Sk+*O{~Jr@=Am7V+X3daruIY%_V zrG@ig)nN5<$`K~>rmRb7s8ws1a-goo#4CnJE7tG-?lrrJAY*OF|m28i~QtBv8 z2b?Ca(aoI0f=b7NT7JD2t|=H2UDVj9<>52pRlUNW6$~YHCs*K{VIyI^aj%zD(w(E= zM9Y7eM8QJBh?ZpdUKF|g1(BgFj9^kD@zmK*gLfTBTo`M>W4@R5+hk(HV$5WD9OW^! zm;4GTb5(G&jbW+t&y zQege|^-)=oQiAiLXPIm1t*{_UjL>cl$`Gz?F2%f>)nz=oc&p7(ImqIIfI<2zaj1^u zr;)dU9kmQ92Gq-{WHwP{tY*FPh& zb422k~yIj1K*#29z0~XQAl_sB5b9HxE%A%k9 zs5}ZxqRgM+|ImNU!%clw*Qna${?o89zPkJU{HS}e+KV66MLTUcjzo`dwurNEc9mejX+l(bd$AH0=v z_uvaqQdrCito~>p)|L5T8~36RWB?}R*keR*wyLz!e&?c(=Q5uIOkuR?jT-%B)(>eT z%T}?mDT>0i%nv@@4mhH&oLr0^y-_k&J&fpiXPGAq&zSW&3Znd=Pw<(f2}3x8Guq=g ze5XvfsWS{qw*0g8T4=q%Lm{bv^Dn<&4S#A+GRTxQBWl~?>i^T)0JgY^J!@!zpDURl z_$|Wi;b4QqdZGqgl$))PYhYFfVt-r5a&|J@TE zj5DfTIaH#$RoW9l9`V_(8N2h&{md&k`o#^=Ij>t+ge-&SIygO_^ zKA!4P8PE8!da2n8qf1ORJ+(%xwWy=2NQuN&6Qi=ktJs`(%jW9r_ZtE+O%j(6QwzT- zGc+Xg?rZNrlV)i>3#d4CVDgDF_FDu|3=Yl@lz19y#S%SsmsT%Zt=ovhd&fkl;Gc-j z_l$|!``lu^Fl+bxnXF9|KMSQu66C@yKWl$s3Biwztt0Fb{sq~7_-|f*-@CP^AkthA z6r~n+#r}!eS2%4RmlZ5JDvWd?xEc>*1HA}64855Z8<9_Yhh{bKsujt^fxi4xwo6M*UqSG$LB2L(|;ZRT=R} zKIqd`)grcV?l=9${RcZnG z+dtxq%#`du-`w8Y(T>ln<-lhy&%#T-M^!{>PNmP}Xv900_Rwc*YNlLgl`}#;rT!A$ z<#RU}qL{mllvN5^HSSY>nyU2-)LrTx_i`hlLEwgT1>^BGH?ulYNfS~*0}p12TBY;^ zvZ#Mx?j&yt`D4&_DAw-AP3aWFn-=3r^a7o%*|35|{X87gcr;3Lp0*ldpm!1;HH@U< z%-D53EQ>eJ(j*WcQhQ&%%#w&b8C&@2hd*;-Q$W}e%{$Lt%P{7PSvRFS8+O~JKV@=H zu4I-MJWRZj5e`1I-pI4!*gt<3Q7+4N3sIsMw!Stt;<<8t+@*M#JVd?;{!z!eBZ-|n zvf4>rsBA7wy!tTm?ts%?f5e@EO(;U``TZxId>V4*MI#qSpRi%wseDFYR?n?e(qs* zMMbq6fk*>8t3a9)d%Pk}XlMo6j?FZM;HK=2L7Xo*M@IUl3i+$$O$?3)LGmP+&oFjl z(PJjR>?n~MXKkg*FD$MIOIS}Tl`?vXE99saqRH8>m|pGV4HO8xcqI;wQR@A@c2j}& z+#w}@&sR~|sMe83t+s|Xl|AEV6s2l?*W_42&P8yu97E6ZPS zc$4gT#?GbKn=C(Aou_RU<>XsJ)Qt;_?D7nc_0AL2DT_LPWc+tYB_4^5V zYc|$KpF5W=o~u+)BD+v~a9!eK4p8`pO#1KdgqT4ydKja5)& zxhU#ZGAk1B!wN;?`EperAMB3r+tpRRaTy#k$4ez4cm0~x=}lP zuNB`f<)P?|?#1Z^dr0cTYZ*~@b*2kNXVb0U=d*K+v{a!7cPQ~=ZNWdNT<5gW(lCQL zdx1T!>>JqQ{4c90xTTIAYnavEdw(VDhF+`u;G_<#mss%fa6`*6o2^YCqHa-pf3PmT zu4eR=gXEp;*3}KBkAj>W4#xb-u^CbCEQtKAE%KRSDvCee+*c-|R*}^$giQ$VNPjQ+ zB5bXK@qs^@E7T}m$N-T!1m10OzR`L>h<509vuJ(6uW)L`guiLzITxdavO7&x|0)if zB>i-w=Y_;=G;?>cz%^?V-R3G&;iX`p1z-70rHn$JUoFM9aYRF2%V~j;>di^;4671%+Z~-7PvEs0(?B z{f&&q6j5Pt7WLR+k%M)bHlah<9sl>yFFta1QO_7_vr|Y6 z<77MFUY1&!H>~jp3{kqTUI2%i>vQpE<-2lc;jmi0WM+(rD@}l6gg+{833!HZ0y09OoP#>f$BbVCk}7yN+dRE?MO5N%qq7c9 zKbyqFEKF!;3`@Fsu$%gQ2`&*zMk4JoP|tGRy^?xpE`{2Or2G*JiIjzgEG_22=s5yD z*UUUh_J|VrR$A%_j66S%xqN$|bG4u5oYiyp80wgNGH#xua)#|Gwap2X8Z=TcB0%q-{5+spgDmp z^7HoL_R3F*k4N_-I%4Dw&i ztN`Nv|HTwhFv)-QLv;9~$p6H{P@}Sn!2To#{uR7Kp(_7PZsC9qxT8Rp|7vE10T}LI zAl848qd=Jd?k6nx3*Yo$mYMu5pb05C+ zZEkd-Ga){M#9=TzGT$13?9U(yQcQn=ZT}tj5=9nnN212dBBM|gh|V#{HKbg(Mb9!M zL5vOH@&8Jv{R<`?1hD3d>5;hg9Ehjm5LqA|K+W;eVrq zb0Lx@AQwOoa7dpIB_53gizxOii@75Y$N(6A?43$Jn!wND`ML;U^uRtV7t>+az zQ$Ym8n!%)KBT!iFEgvAXh}a}30m9`CI@@_oG$>sW9|r0!1hYa&5Gu34Q(&lxCq54{cK=GC21@^- z`;|cLRtGZVDd!+ec(7lKOi~g-vIg{LM@T6$JdODYk^<6YNfse~7%*l~MnLB$#19L2CK8YV5cp#83xV%qC}K}&{L=f@d5DW)0UqujazCrxy`9+xl>H;!8h|6bT zVW@UhQ%+e+*8&_45P+c=GVqHtpRU7y^$-L-PI)tcK&6lnDhT<{L4=tF!E~W?1F=gX zP?+JxGK332C5WRMkwr!B??3_&b1M)G2uv99c?FV0b9z|fU(Ccn&>U$AUrfM9NDTEA z3&cA?=>`POS4aXBSOrnD3+w}epMv5jrhb@0z$Cx;1EYW%FVyRhWF!>Cqjg9UR74mh zhaV==24aZjZ;&M9LPc5#O8{sM3bPDsK#Cw>&d?8lxj_UMlqx`MZi0Sd{D`nEFcyeb zqZ2}XF=5EXg-G2-4k3)32476D^MYC^NTIO+Mm;iDo_QBgh=Vj}Bn6lmP+Wz750Z=y zL)`ud63$2VAorlcK;q3kh%bf!GlU7wa5F)E#O0sBeqvyp2Rle(fe5p(=;=71LVyq`4FWp*U;=y* zAho~|&<0cq5JC_GSH^+~11rK1HJDI8OnyKne)<*J^=z(SBFx})z-hA}@L!bNJ{T|{ z2BI1h8UW^1=yd7;3t(hnl%ryTLIJ?uS0PY;OyDEnu)(PeKrb=GV@%Kr<_Lh1pcn&l z0OS$mSLB)$Kcxaa#R0I$IjAoNKluOD-$a4npiT|Y0m=^+jrLE7KL)c9FgvmzQWD_A zC`Onp1l4Yk2F4HqGjM7mQRG5An>TQQ1T{a99^nrFHIQ2a7PuA) z0jr>#0E{VqIwtVZvlRdX8uOnvf*Am)4aofjToS?mNH+yjEd;z0Ia?SaBEVz3d5fjuqKn_I>bO$4y_QV9m4}h7FYGMn30tmuDJ}XkFKUgDT$S=Maz?TFO#$?b$ za8*K%hy_yb!NnZmKn{(?1QnCO^jd<=2Mc843yGa62j*WGD1Ar{jl&WE9srnGk^*6X z51npMVBG*`6gpibAZH&O(?Oi0MD+&x06$0S19JF5?^Dr$Yk;aOC_S)%KM0ATR>nHfsm)H|osCKbhTP76fbkKfRqnGwS5X zKYP1^=KsU`HdqN~m<-7410r`Df-NNI0b2jh7+zq7{hu*>Kr?DRfOQ4D3@lMUD&%@U zz3xa$|Gw7+_&>??P;vx6J(L~-7HBdPlm=l>52YkOli}~a0Dp!asty4VL8xvAsBWmE zF1UYn0~uXbxk6>i+Ip$O5HgKZWAJ{Vm5SRL4I^MNipmvOuMg zv;wS9KK4`Oj=#uG{rxY^#lU5P4_GjZ721d-Yv%z|>cWP!_UR({yRQ)90WauV#Rk=f zfNKg$69#rDC4dI{x5-Z(_m9a>=j89fP*cZH_`~kZ^G*r=R_$z=EM0(dLy!nuNc;0}0oA~^ z`foFwu8Uh-$n|{IwGeb=7yqr*sT=?);oqdEN+345pf`WhdT}FZ|E0tKVJnPaxdf!@ z;f6Z?rq#NLr2Re1XM9iP_**@22Z(f;`xl{?e$y@iS~l@hSN!eKXL9^ad%74zE z7zUor{^`Vk5m?$32x1|q(Qn#tA(VyxsRdy3AKEiV92SCF{wCHIMoIafMG3S(&1Hfx zv;?{0Q7UPRAkFnJul=(FoSN&G)PKzNZ{I|k3%SR)i9nf=4;)}@J}_JV_9xJCDuXx- zdf}`eEg}$x()OP|3_FEh`fI}kue?r)`(VgrhhjpDAQMFq|I?HIA7^kIh`c0Ni$Ya@ z6ZfEq&oHKbc_iTTZ#xT|-61@dfMOu%#GtajNAMFvvPT{1|Fe+5yQcq)aOTrpVo90Q=VBmbZmLyaF z0=kDvLfOet?@|77c@-8;eSD7hJWSwpA@J$idboPITibdt!%#^SVC<_dF0P))-=ata zF(_zHMSmNd8fA=+NODETNj((6MVy#{5IyJtq(Qw-1o!+X=LerCTZ2!O4&W2zHsBLA z`JlWeN(1nT+FQUUYUzVd)G`L2D8azz*?6b7?cn$6c)yFv!2~FgJ(1l1swxK)5cySA zPR(^1M%?>gclmkJ+S1he!{^7_qvqBr_U9cOyYQ)%H_!LNSSRetXQU-v z^6Is+rv{wA|16DnfAjOncOAm@w(IbDaJaf*z(8uMB*@+8_}2F+xJk=#@zE6G>o<7X z6=~7gMwWX_d-)}QH{+W9%x^E}g_8{y~ z?125w?yKX^D~iu?N%rB+2Jx+jUB}kXI(#NGYmbh%_qs?kUu#YM;G=u~^t>9KlPH|+ z+kK@iGjn#$(yz%xZXFM*Ck`3X*Gx;c+&B1d%tNp~d#{#<^RL@gnBz|Eiev6wH zmgFer9(Jvg*0+DKY$e@;KW1!nZMyVFie==GDweXsvs$G~u+g+AiS&z2dRePO5BmL; zWF36j*Uo+PkB(w_Kpm^7tL*(!((m9hoEtB4coZXZ-Jx+;q@Qxbs@%|d&%U^BWvy{D zH}?ZY$@e!)#g#SN-57V&*{O5)aTLctX1fi4s-9rK6MHvV%Z;%arfl;PCtS%FyB8y@ z4ND2bl#e>kt|2j}T4}o1kcPXr0_Q2Y6h&Gh-BjR)mj_i*e1~OqJ=|Db8m3gu3GaXU zzAJTZ0RPQ!Tnc8}y`l|M1KJHq(Iw+2&?D|~>_ zNA$63sdS!(z{?eZhct~F_g>Z&(@ewP=|$Wg$B{hp#SE2f-pAi znA|jJn*XFZ8uQb+cjATExCLp+bQ>F-JfD`;Kh&hyq%-&JHx;TG{Aj0dvaSnH@H7Z) z>>OgKte5lbD%;z8c)8X6!=|#6-g3eWufnzS0`XSgqh{ML#}@me(Ii|nf}}})gSV#$ z#Eo@sD1`LgF!v|;P9aT7ec34OhG1|}Se>flNCuDPApR1GrvyIN@UxoEi!!~<(B)w9 z7aH1&9H!RvmZnOt(J8!N!G?CZ&hc3GSldz4I|<3@<5|4I&>H^?3f?7p+!_7xL&1-d z`pGL|A!+go^&2@t_{2}X5>ueXZtKCVCfuNXE1s^Fi`N>M`mHPt)V`GE>(Ll#5(qrD zT!;}AB_}^78*1=AmP@1xh<9S+cFYZbs(3$>Bu{3Z;bk;ubJ$fM98cWjahuns@i;L* zT#O)|YdV%P4S5}`n(=dDw&WZiEAW)9a$q{g(HUgAWoe-Jb-NVc{*Ran!UebOn+!aTXXqoc5(T?0a(MQmnp8gu!TW|~!5*lSH3!xPf(}Qp(AT)rqvzGPWKV=)nD-$Gi&} zH&+?WvvletnD`XOTkzR&LH8U>0LP1awnvxaUP0C)wBJsMI?^~?yih+c5*IJu*Kb}x zgQ#cfXBkpA!+}d~lc@+J$mQ5+^&%dt);4etC3MuQ){b2q3YWLrTN(M9u|5N=V}RQ)hGc;)p&Y6?wCQN|I1w& zC#n_0u0bO^3`3&gq(^z&-hs;uYddC7@MxIc={Zo_Pu!MgQCnawlZIY@QYd9vV-AP5 z6KDuq;11ynxZ}WEEqD{vW@%~~-l1MGHZgzm$a2gAfAw?{}N z4tw<{eZ3Xyq((c(@CdN39*x#+LTk)!wn7Szh#9s6;jmF%J1y}Is=f<1@$RX%3WSVE zHEJD7Oq0J{xHsBWI-Y5FuFEfqF2ZdRjj_ERy8hzbfeVgKE=A<9to0<3oeGa|NnihY z(vfGh`cIhZ;LXI-gcq#fEQEo|hH}&Q2(aF>MLvm!d4$&VXUNEVH4^Ll68T8E1jxw_ z`iHY%+^1PhjQV1UCeU+<`4!rI!}FK#IyeMhr;w{<!5#!CPjhwB*dKmqE$2;>vL*ZJ!*oKio)MhH0F}!B+@EXr5^~AIX!jA(#aY`R4 zdQla2-rw?*efs7+rLOM-ohS;D`Bq1}ys~10)xZ%|jt>L(S6UsfmcE`ymbVIjpDZDmQKK%Lw=nB-{TE3(-97$^yX51c z?cTV_wK1-l88XY4)P3Y;X|BPy9&vZU!wq4$Au$)CEzW zDfP2^f^o>_Egq;|Z2sc*g5jb(e7###=F(eRSU_-+JDh~)gIt#VYq-XE;=4Q94mAVQ z#a)f`RBId7RdJ=g1R2xMC*(HBE<|1b5dXF8lE}5>cRv<$Jm{%KrvkdK7C!4|95piN)-uEWxef~Y?6$XQ_o9z<&=W$E^5~y7T-ca577DTJ$A}5owlTF% zpTO+g?wgZXcVL-Z5Ul-VUx7{aD9*vNr0GlV^y`eti~=Psv?ej@33$``c}+{6QWY)3 zusdZmG-N?zdi3R0F5y{ZqRH^OaiK?V>d|=L-{|7e)5>gO`D}By3pYOU;iRekSL}3g z@efs_kwPO-LgVQjx(GrHkNdv%Xiv{Q2z$62kj%(ytkYY|)IYDU?RD>SGHr=F_kz7? zE##Ts^yO>L=;;&%?;|YWcNNAv%&o0w>%QETZGlnAd%XI7Owz~^J4fM9) z>g)=~d-!oQlURQ3x{CC(fprBxuXskNQ7#o}9IPxq}Phm3?5(@%x$zq}xqav2VH2wS*F2*p8%tZ}~z!J)_6fC#Hw z@sruUh{bg1vN}Ilev#<*n<9RVIGi!zOW!bI(GJ~lOp40Gu_<=_{HiJ5rNWOz(`l~6 z6$PZ+TVa0{u%={~<~(Yu-l%A-KJ>c#D1>q?ifoHGGt1TqSKxRuYW<*}!5A-iX;_ja zBs%4Yd!>))ZqIq7WAIO~e_$I*P)~!`Yb`eP89su~9Qpd<8{6WYOl_I4Qa4OZ4gBP^ zepiyZ=}SiaJ2PB}U*Pwd4-m6Q=s1AqU;`tBi3T)@0tUj7|9Z0^Abw91$^xM91(8n* zoDFK4P+STc@bm0xLIhMQ|MgxS1|oJ~JPL$c6qFZHV-2+ck3%5Z^c2uZ*gz@K!7u_1 zdbZF}=mpRh$;u3uPv6zR#lhOu#+DiUi){Pr2@Ql@Q34tpA>a;30a4l9e?Ovu76iK^ zlobpudipF=X^dy8r!?ATzpfF=DjOV3d6ikFSB+6194%&^WM-ZPBJCH|jDdbNnU%Fp znMa|>c(y*Ko>k!%PgDKWYeVM-qA}g<;$zbW^c<PQg?hYrfoBOb-KM%|mxNn2 zKiZh_QV4o2xW8|A`01FkVzvv-Bg?9hH#sCTc;AIvluJXkj14Xv@uc+?p}n$?sOjp- z1mfx;DO?FlEPyw@##ZuiJ@pmA0Br28xuhiGI;tS{c0%9S*QCul106R$wmosJy!NB> z%0qMK7)X?Hz!Qm9i4p7<4M86EEe$@Cs=Nfc_m9hVd#+eSrd2Z4OYZh~AGdDA1gdR& z(_@G=mNL6)`H6XVdNQM%)VN-bs7d86HfLEOo~(rT_NmSD-Yb*vXSq?I+aNe;@!3gg zf7zG!<`+Tl)~VUv8CCTJ63Sk~+*0eZg~CQOr>;CdauZ!u2Kr&^or_8C>81;ojCYOJ zr6m%2XUdk;ViJdMp|h?A^m1FojAMs2r7xeL3B5DyT;bMjY0LAvvD1+$BOmVvU*Ob>#tMN(^brNv2;wo}}0Z3{7+1F)Kyu!YoGIvQIA^K!e;@ zB!`77Q5ceBJpWP1XgJr0|ue5VX)I4+}d;duL%sdWH0lebWmu z$^rruuyVX#*4~KL^<@>1O6|2()!y6L+0*5X#Shy#){G$0dQIY&{n@@G+NKNp=}Vh) zA6a}#o7dNNW6a8obpg=UlOhcN`4@Xh-?=vPLYx))Rz25W3VS=mq#UDh$Hos9eqXUw zv>%0g^tn7XkUtzyi%%e&wkilYrgv~Vq9Ay=)`Ta;ozho`sAQEQ&UniEtcp|d3oB0x zht|0w?pd+)-FR-Ty%5`rhi+@v*_hlm$8Rtf+EQy-DeCK{=Fbt8sziV7qYpPZCa{8O6-rqDmj`y#$<-+M7_^OM-sWQcv) z=2mPP5t(5%rgCpiHa@9Z*-pG?pGA^@PxLXyMYUUM5=t|7$OG=;-LUb6GGntF&sQZq zXvT!=K}KFkY=#|;mFBBoe9tXqpeMH=?XhUiqBl=VC47Vp^Z(Mqs6lM>HBZuEg-x^E zL%zOtQtG{&_|LZy2abzVQp=VUJB0=2m>E3{+M z?AyEa6GZeDFU)Mb&WIt$?1rSs)muLrWP^6VWz$}X;Rte;3Pyg_ku|t+cU+CffHd}# zRBmpu!k|u&%heU7c`+&s+>_LLD>=J+<2U-`(Oe0>563aAX{60L5cS>-3tMZwIk%>A z*`J(iacHKz%H4;Xk-%TW#Xkkx0Qxle1skDe`2&5$g1*!{GM^dfWQJ?88Mg8blgO|X zDd5-DDeid-mUM6FJkDG$ei2X?jbER6{Au;sWp3fQym}T_xHq>$<)}eZf&OlE)g67` z$KEhcX`eD-s{Gc9vQX>SZ(1;jlML+Weu-pM#wj#_mDe3lkPg_7+SAsTTZH8-g1)VZFPs;6Q`mg#fSR+M0NBM@}>c`UbQ zyz^b1pT+L7DPfJHqc37q9f9`@>8Y9EVWZZ~XXvQP@%5&cjt)WB#W#U=W2--V?zKFI z+j~rWs$a-W!~I#L9{lXNw2R+ID8`faD_fG(gDnBMGJ(axt87f}=we;xO%dD%L}-^E zwvsbOtUhNYG(@!4!eTxiCco>g;@=n~M-Rn!i z%8h9`c9YxtGtWfy3&Q@;WFq;N0@n;QN4SQqi=c-q`G=NayJsG5`X6&kAIZ<%t;}b2 zR4OhT8$ahNXKJmbGjR82^@fsd(pztD4Z9_heLO;r*-f^ihOXK^)q8pLlXZb*toeC$ zSJupikBw@juj*`#D}E@^BC^qYI(dD!_gfF8*Ob;b#!s$6+UM#LTkd-90y&kw||Z=yRte^GkaGvI}Lw>1TC@~N*t|J0AAE))Eon&`)mmCZJ5Et1@` zo~koEUgGai@awM*gdv<=SmA!6@UfHIdDwlP@o{diLT!37EhBWA#BnU6B`zsb6_3LQ zG*==V6^dhTJj@>UT(zXVR}kx&P9LQm`6^=XrEV~e#=(iMwc!>0RfJzL$%c+E&--l# ze&SS4pY08|)TP%iE;%0zV$vl(gc6MwU11Q^JJ+e$DB(4^7I8h+%(Ns}AeNYjDABY| zV(fz&VYk&tcjVd!o$pE=cVs3mM5Ys^!r^-^iHFM9uWKFWbRQO8-LHRFBbc1O!8 zV)JL5u07tpQD)uXiBAkNOt!~Bk3IqyeQ$(b8JS{pwdXqH6{`FUS8*cUj2+k}$BkBW zUQ(iBx8MZt*Ct4ZiSt09fNc8qB4c;%!1R8X1NDTO=(}uxALq|2q&-9&V_VG zeeID@DsFuil`ZCJIGlZ*{_^gDwGsVzh3`&A z{;^qPTy<(t*}TYsIr&!j;0i{nQ!xH>w0h?Qy%c9BJGHpGMag*n-Wjtkxr;ZiU`7?& zMiqN9wn)CUN{c%8k>g={*3;D3fnc~*o=mEg4>`w=?fb`#pEMPRwkZ0R0{8H5C;Gkb z9bkFvI?!vFRlr7Fs-xO^fYje9Wf5vV!wa$;`8+V*#(}c4}Nsci=S_c>j;f zdytNSA#aMhHnIn9d1##10|LHBJbN0W7Z{TB^Hr}|>>Y(Q6V5xSUbka<XY!ZrE!JCDC`R^{++3V=6z^ovh#BuZmFv%|_wy7ip7Sq!BUc$Up+a@J zmwu=$S9WffXPzgM@fmn<{Q%6uWUJ(rWDgBSMH=FuLjUEfE`TS`$!W_0`vXNMmu{ z3&ZG4yukfWUyiH{L^040v8IS3HMg#8`kQynu7#7{7;mJFCP_0z8j%scC+$uoSYaFEcnZa*>|Su^L%T4W=mo(oyzpJSM>yMY5HYVYtc6G zi)&Gp<+HY}6~4lZ^{%YO$!OzW-)!A_$wnZ6hHc5J4mG;|@;+;P-DXqzwjG9DbwXm5 z`o4X4w}Cm{@U5q?$~-a4Vb1WtFU;^qMZ!;3g1gBF9eWqJyiM}HUX5eaEg57fO0|CA zf43WpkU$qCK2PLNu3Takln@h)^JYkgBxSs1XP0-2(8lOrPiiqjkKY(+66qK^VbXN} zaZY}vI?>1kb0PLo=dh5QiUqq?NWxbex$P3l4~I19x>ua7-o5nNX3cqN;V9t)KNqxT zOA=iou2d&Iq}-B2H@?kP%kt6~V^{rdU6dZy{3}eFGMnWge@#+Jmd)@+0?2wob{?yNt#9 zoHnX^A>G)TbogNe*EB<@k1c<89>OC=X5ey%^6c7 zzq1XR9kmE4DA-%RY#*(rR9aAE$$fo%q!l^ZJaG>{>0MBe!SUU_j-q3B>x(;wGyEjfEk;)1g$sUP!ct=)*NM#ku%E-#d%19n5 ziLA`*>`G=LJIW{_5sD}~WM^dk&)e7XJb%~qUGn?R^PF?;bDw*i^W1kqX)GUi`N83+ zCx`Is;(5y3BzgVTN>nM6T+clb{bZPx_JUo5Q-AHGqP2(n7MXyyr+9?B%=iiC3 z>{qN5G#cz5$eV+Xcry){?0X}t2o2ug8ydsf*3j97?bBy-fZlj#rS0T@)5&tOSVx2P zsSC$g@aZajmAPF-s?Iixqr=n2l07YQ{0qr@SC_MyFJ5bIbnGj*!h0}B@@WoP&eJy0 zck5(p`#+}Yrmrlg>xP@?+S%R`adt+P?Ts5~j(xo=u`_4mbX&@ioZ=~sFH1v^>N^b` zgFPC?=ieO-TCI@+rDGqt@Vg5}+3>1`S2kPW*&6sRCftlNa;fC7v*kQy`>ldH(Enz{ z$Gv_#E}5xC@*kIUbi6#8UA;-Rmz~ACh&s>j-Ux8>bclw_SpU-^GN z8z#NvA(1rFKe(gqYupam&wlIHZ=OIBcXP!wv{Mzcba$iG@!YN7>2?>hHoZa{-I$%_ z+1Pt%+%%L&qv?|NgGmZ6~>2}Si;t!HH79l5lRr7Qq_mtZQB7+bgvKsP zvq=uzvHq^lX?khfbL0B`LTXm8>nS;EJ&&uX)bSoT@UTB=cz;@kz=a==ySgGd?E>HD z%IfJA88>kCC6=6~8feri+x;b3hyMNCq2)2PgDXQyD(brrQ)S(uy@HF`VJW^%rTcQ~ z<7Jg&OzGyc{qrUF%YtO~)O>I)>wdk<7;^vep*zlloDe;8SAI52e4roW@11{YofZ|k__NT(c2@63>;o|o6BF69&rz5r>D8Mh)g-QfmWt?7jZDEP|e)XOEdS8*l zj#007o8^u4%Ay}%>naEuf2g0f9&GczsR_b7MgkfLE_HqiB_d5BdEzTD5%jucY z%zC0l*F7e*4Xmx=;yW@O1L!W5Z8(qFN+eP5a!#Ty{hFj1@De2?v&*(HNwsZA(EUXR zlwU!3$6hQ?=UKf_dPl)EyKo!1ud_{>jEnFc$rQDivkiy(3Zy&Q+i}t^au1wZ#5}HA zRM_0o{>ZTHbeVbR?fi2|^(-uz3`z>QAEk(@p;IzJy=-iJN1PKBu3e>=K0^r|x8vbS zvJoG+BSfF9AEfkJi(e{g`+9Ke0An@JT}w!Xr^jMJ;KqyWpsajvmd;SrYSBnz#Nmtb zeS6I3Eb9&VrjHLsZg`_-ckr6%-jyEYo4b*JCuJ~Cbusmm+Qq<`Q?JA&qc*aesm__~ zW<06E!+MEwxOB&y^h?~!W8a?3+;6p!GHJ5HmOaa(P2B}4X4Ol)5({xMJ+IG|j}}g`+SGk*Ul@8? z+e3Ht`IY-2K9;uLindGa4lG>5JM!==KkVnA%h%uGgL-y9U#P~>>v=BJ|EN2DUQ7OA zzLrNgk@%T`VRYU~{)_bJ1J=;7gxGVdodR!TkKA&a3{FT^pg%96*P+R3IKrYK|2pcr z&FU+bX#Yo_gS;n>v*W#rcsV6(o<$jSZTFm(jP316(U;7MOEv3EHh6U^LzJJF!uFL_ zb{drJBNFAs(W#;u`r=GX#oNHR`M5yCm6fX^wbt>;qi3!zQ_6lF5f1Zg@`Zur7nkUzM6!hk_`!s z;^zyG>8IsXzm2}v?VaNNeI-kqTP9K3jdsoxsu^~Y44SVw%I4%t=)#uDZc7VcIoeV) z5ntSUUU5am&aibXWMb?H_0qsw{_S-Fxt`Ag?3ma7=T8q}i}ZJORoMya%J|PkWVH4Y3~%z80EWM&39Us>Xih0`rUH ze8`xA=;Aipu%f-==J-{&;q8L;6^f%F zBXj;MY~+P?O$WLaCst$h?I#P;oH;VuZP#y*K|BFwo9iS=vc8n=V|uk8StqPE(%@+Ef>^$N4v2?>-vj z!S9x!_3~b%X#)Cwx_gMh=i0mar^l{SnOyTTcflx~^hn_`P+-%z5M~khNhtZ* zWjjM?@=JJ}ZkP2zjSr9P5?s^MA34}DEwAD|ebuiWci!EL_WmkFF(aDc;fB>2zJ$wrKPhj% zHoiaKWy#^|cWigKe}dythh32)H+IWN=_DB(V0;gClsJ8v(jBhcdB9xy3onNY^-0k# z<-=J!IDZpXDd~K8!VDOc-TjBWO z$8~AJ3DEQIpb~`(f-yovSN0r!J%57*?^4J7rv-n_VPxcMi0V_i8`%p0WVW}J`PFgAF5tc~@?vzP78tDbqQtrGvHmv@(| z&()iH?L~bySw)PUPdq)>+Sk6Wu$!UtZn*_rqY}9evH*wF!wPhu zHc+=pm2zX#k?GFOoe3$Hof2XjRyllKM+T1HDx0oezVF&j!-dy?YS(Jqy1SO+^&pu} zJ@F5sMc9oeK^mK50?u*SCJd^3+7I@2NkrX4f40k{)6Rb8|Mlh7xI)6M+4@VgozC@_ zcJxh1OhW-*w~z3CEjOima&!ULeymCHTo6Z7?d(TCUFOAt?Gqmc=T#TD=A9mQ`L2*x zP;rY@>FM|j(@Zd3Alhv*7^U%j3z9QR=Dl%}O;*a!E~$RPDBS@c-Eb^hLvBYOS83qz z`f?xLq`G>ZN(9#gi>W8)#G#tVlM|-zE+l?)mjOx2#T_kbXUBuSzrJlTwA(0A=-a+_ zHqK&?pm%CL3P&XMrdq$FmE034Sz_tBPm^aV96IKyJY`yDIug5krQ3JfU8^dtAzr>e zV=&!y?5bW%yT@LO`s=UEJ3lWiJ-^N*eV_W+MET2LzKEzH?iGj5>xnh{3jIG!6V+Po zYC9$N_A5d(cWKhf8QRoR-ITf%FCPa?2y*2Y6$KT)De-#elr_8*v0tFll-WwML8Rf) zl|tLOgjX36#_`e;DhG(#^YPF;?tt2Q`b2t9dP(}R;&wYJ z+f_TdEA86@^y``0nYkhqH^zAtux4zZALx8U2NCzY;dQrs&1ivMj-BU-_K%%6t*1oC zT4jbmG#!tPIw`!UUE6`qeZ%uog*nsXBvpi66|Je)?8MUBAM&h%Z}UX8ebhHfmJDNA zIn;w^rf4dld34O<`ZI@%46c_lciRwT18_^)A^lnw!qH4-xm=f`F0kjGlr=Rv*eT92 zL8NVZU0_ZeajlKd&?~3WHO^lI4~LzvTb>Mfmv(qx9cjxERh5iv28j(w}Gv6y`rv`R6YsY>8tC1t&Q; z3;%h07j_o@^JehBXW_pN>y^&M424oeg0pbMVLkW=GB}Kan*xVHmw?lHy=|=EU%wCP zvA92k0ACvaJ*PK|pfCdy`gu+-N^eMG`0sUj461}Ol7eBI5DriRabEvl#^`^>2)>#8 zd0v0wT#Q{bMI;gT>kGhtn;`EYqqvC@;2VcO8KZIFTUOF#1IT0NQ6{%&x5^`BrSfr^ zepAIeVve!E`L@&1eU~Ao#&yPMXDi=3#*Dcy`KmlZ?;oSGi--y+6 z^*BWs__B$>vh&Yj<6Cfi4t|bHtnv{0gte>+k2v7Krg{a+8 z+e;oRjE6RRpK#-bB#)%mnfK_|$!44w`Dz_FP=YrRY^oj>#>Rz7??HW1*haaR`SGXo z6{F=|XHD-Nl9skHUeBCh6Nv~tw`y~|N105wUnBYH^!wax;T7Z4Y(jzK)a^DYPtKND zFMl3*Ba=Bqsd0v>IM;n1$J+bA&N`RQW5@lfGXppH=tska6HF-}Z5IRM*IkedO;(S{64N;%uH^S~_CwlgK)`pPZ^R#UCyS~dO_qtiu z!MDjdYNC1M;S4L;{Pp2}G(PFe_wJsw^R#`Y_-$Z#Jkx3C?Vb5CqE68s+wjnLQ(V$X}_@Tr&wBaW=vf(s`Vp9a~@m+RbO+xaL9i=whFZ*)wanj_afu{@7W zsd7K!o;Zuf#%dJHG#Ib}de>=b^ChmClrDs#+w)DeSXqaV=$WI(?zJ{`L zq~4{cL(MaL@Gj@@w3+xp=(9D40QCq(cCs_tFbo)=Wi_I#6lx`pRE92grYC1(m?&zWUBj?2nS{maJ z%-8~5IjwGSP2abTlq5*}kdZFa&Y=+3TyJ!GhhI}>D|d$M&0E}jp05%ZTkmqbWITi` zc#{yE)=oPzX*w0${i$!>!~61yn|nO3yf3ckxUOgzSw6>iGPh^f{YbSVS)I1zhC}-g z$Dgzjx*qNMY2=!Jj1GrY!1cYZKAIQ%PI9pdaUIh1KEdvw~WNOYgqk&e&6nhC>PL;*}~B7RnFf z8t}`Gatt@ESR07V!Gr$e@(OwU*TgS#7Pgkke)n7Jua$jyJ$C5<(}t(l4Ut58+p9EH zybfXUcexbp3KA-6Jbt_=frdsXoN@drtrknT_`>sNiq3iq=&dq4sLR}m>5SlMGbzdp zJFDLbg%9#R^L-yJ(w^vBd59|T3}u6u10U@imAlmbqbJ#GzheSKd^J*2Ky6~i>Sk?@ zceG)lJ##b+carbj)5}L!W}S-Fs)?7*xsmLf_Lhn2t*C)PtU*zl9To~!$;tbeG%t83 z`NOL&%mYzLX~DzPpN!1$a);VGD)Jm3^5hL=$kfvn6kc#OzhEET{h_((hLiKN!N>w9 zH5-p8`6irsR?JIf%BP>t2ZSt|XDf~jSg{yJ3hh{Kc2_An!$rfbsrci_su0Do(ek!M z28UA1g*LE5!%o!Ad!6xya<9;~Usp&^QL*Y`7S~D_4}7!`^Mm+NN{T3zvf!_wXy}`9 zKk)+JH+0(I+(y8(UdJ24c<Qs~Dm;kscNPFq(K`lnN+ocyv4RlXj z*+^RTeryUoQO#KlXyZt034hV^sAYI#zVYa+xxJ|R!N=UxQbP-n9(JCQ=+Wd zguE-lA$-N}O-{qZ1^x}lGyTi7G|Qc5=0B*;pT&M;CO178re6?RLL2*h*MUO?DayPk zRz1{#x3(NSp$zQ%sToJeYE&l9sh>Kk5-jUzxayQ4a29_c_0j6QO`nksOL4ilTitog zHrZX%LcPnaUtiyRaQW+w1l>=b3#?#deIL_A$Rg0XJETLzPi1^W!}p zvQlrawxmRE3lJioKc(Br6d!l_p+vJgWmHh#nUB3|f*Ik?q5&mN3bJ%;6`4c2DcVeBbc;tpNi zcJ!?m{%jb_!;)V4x1w^@HybMYhF@eny2Whe%zIlCuPK-*sAAfbkYx9!PGm4*LG-rz zs-C=?z}S!16|KT44_fXSl3!_PesMY=L#%*}@*cJuv*V4~^ZUHy69>*fcRQC)$F|l} z9rJLsPG`PZZyF!%M;`k$jILn$uG!90=S;{bV^7bItX$RK7l9cd^SFQGi2AESzd9u) zulwwuW$J?U71dK?R6SNwchXR$?|ayBFEZ}!O&*iT)9;B>bBdL z%fUQ-`P{bamJ{tbKk-`!s-boMnShg8{-cM)ZtLkqhwrQUVOBFruX<0f2;xC~@hMVS z7_;>Z(Ar(p;xb>~Z+Y7I>Pcm1o9>-4rF=%KmsgmbI4-uo`sA}ykzSE|JEhTK)5zOj z?cGBy6Ewec5+*O*OPbOj-`O-K(*q4;tVqV33jTb0e%UsePxX@TL*2Vncr49I?pKz= zqmKe~UF&0y1~s}*cuehgyIZ~hX}w=5F*@D1=h2{sV%0WP!nMWYL8ZqX#r6b^SIqRT zSfy({Nt)8~wstgZjty$r9V-w$tgFWMtT498-u*2^x42d{masCcw4+qtEYI1&(X#YT zg9go$%3hvA_O1wO58m64yDzk3?8z9ogr!Q^j}w*h;8dlUO&U+i+j0R#XMj&{y|I0MD*-SoY75= zllhpS2y}wTlA-&h18&PTz8|qTNcOOu%@XS~d>x)$RJyB@CD zEH^%26cg#;_|!@Me4RLvvm}>mT^*@uwEdE<*pXEOYf<)CHkkthANS~H1y6F`KG5sB zwmp#k%e`27TeSz&b$WZWy4!vn$$uXN!Bfu-mpPCQV%WJmn zSu(!$a?cL21@6`#-gA)|OfOW#i9Y8tT|5Fb~A+6^x7?pE5Ef%?dgxit<1ezKdl_HU5GjQG|QO4a1Fp0G+T?ny{_Ni968 z?Lv9+4iEM__Tb&X5e5cg*=o=d#C~<^#T~LQHMd@z>UZpanHhd};@NL3xejPrpeMz5O_L&v&CjYv-;N&NtD7P|ib-E7}Kkj90mU&FNx~HcMr88pR z7o9zPVteGH{p)g$E&JbcY1EwIfl3lD`BwSFYj{MIe$QX^dB7#S!tjcC>zU|KqgM!z zd%w&Z?W%oF%;)jJX}P9r<*zFOI+SN*XKaVRtYEhVCcjn{OTXEPx%uvc@o{FpSKmE9 zU05uZi?pT%DJX$M{&Z}5_5S&`! zi;2rVx?L;naa3_u7Pu3jGr=GtDEew4hEGwKZphIsS8?d&j{W+Jy7rZZ;gLJ{8cKzA zEnF!my6|12#q@@BV^XbFz|i5G2d2=$Vn^=m%IIM0KBWo0&^PvvgPy5lWXqhhGQEr= zQp{X81fMn)pHY(6RNV|-Z{pU->}sswwBDuLt5k?q*Y&gl1qGuThQ8UU4QVdLe`tNk z#m;3?rTU?{_gnf9wF=*hS+DBr`%7AK3qK#r^nw)_H5e0Mj6uWZyL>4S?=xIHYc3lz}3%Pk?dF2%U@H|-2O zc6 zGa)K0{M=Hte#{m*`FQ1|vG~_&r>Pnn1G&0)(41XJ(rMB=$>_4UB!8R7^teok1GJ;p z>KhqavM_J&2@y;S_UswOtNf>@LvnUYj)`5nT%LX?_Y^1d?r77OZNtN#qG(NhLZN#d zZw+s|FLsJ>yj=St{7KST(C1L{ih*j*E+`+XaVSShVt#CfJn&1aC`zHI`?UT!+2rTx z3gR}gpDhRj`%E5s+-hXWG1P)ILi(Ce zsaUkWDx+1JtiT@crnV<9jqi6C4oSUZ;oFQEYHSsYFu%qBQX!l(n!adGbBfC2%yaLJ zqTq&g%R;p|p#iz2Y)hw;yC|b`jYnc5p2=DFKvNuXSsmADG+TOYQVTyc$T{kcsa&zx z?6qDHCV%S4g+6(cey76P6SlI;3?l9E?vRGU)dGPyXP?ouC*lU}PKO~b{pPU{TY98&{q)|LO&L4IR|Mi_AO6H5>x6Ue$9#OKpl`CmZ zw`+?F+&D#y>M zEKot289SaUvBkMmq^2qFKHW!SU3lg2jwh*%E=jgVy7CB%Y0M%HTMT)PFNM2h%txax$rdO?rIxy92 z+n4l}bAMbydYg+amnqlfP|Rt=3a#>q(`6zD4+;u?8>rrRTRy8BWvjDsi91-$F1fK= zs;KVS!|Q$oUiQlBft7gu0R|f_rV6V{l-lEZp#cje5&G-zqx;6DTC`ivw2r;+sA%tn zHm0Dc4)ul6(@XZ8Dkg6%^Dam9j`p&o@rLbLSQZtZdD?6kskr9^Pg_}t`{83xP}W{H z*F103><<_d9FLSy2;b)da&m!?R@tvrt`nhk4^G2~J@(5ASIm?bdxkE%l$r zvAR-OnYvwv$(HpO^$&PyZPqHxx-J4jR}#;s4@Bxzlws z_p0LkGXC7X5j(F|H&q?p>#*xKOJQ8#!;2u=?@Xakvnk4$r^eiUy}~>~|NA~;I*hEL zj=*_={o|D(q8D@;AGDb&yd7!Jsb0|R98$1*pLNp0?>zWej^;$hay*@X64#yperG?T zFN@4)uI%-pciSK6E;3gZ$-Y-V=ydwQ8Hn!khN5ZDqp5{)ae*iEks;Ai*KPLDZ?7L! z6c8uhd<3Pe7rbBj5ZHN`J2XypKK!YGp3bsAXF(JlnX2dx3`Q$3@kTcxdDfpdZ=1@~ z(%G~NjUFjRZib8p9)s53T|Iu9nu&e)58hN3l1ygcs-dWZA2zF$!8n{A}! zK1XSK+>)N`Ma96_I5m%y)K!Phy`qluK1rj`=9*r7E4YwcxYyji?<>{R!On{jBWC3v zyN(S%%)3&(r(-tp8!?6KtV3`A*Yfe&&uB8GZ!yW7Ek@TI-S*kaJRMdaYFi`IF{55^ zb^D%_<(420xweffn2kIvs@hP-J&-6W^vYzkcrK&B9(jm%fqZV?O^dwlE9MqXU6m2c z`4+>SybcwHPX}A`bWJ>qUBwU>Oxoe=#?6a0#18BdvwXK%%^UhF7A6^% zav!i8Y18x(3+D&6%N}^KN~~DATR^>8@Nzrx^km&C)8d}Ni}g*EnYR-OBlH-hwcFF9 zeRLN#_bx(D2pflr_XI4}(o_fzzCXBMPg%DJyVmDf-UQ}IFi@tGcB&p8sZlX z8ziUOm$;dy2lE-nFRu|=jDx=IOmHJ_crg2eVKsN)yDXd3DqUZL9d=<*()SwPagsGR z_U3%Zm%ig38~NfKqag#=7*_imR=3L;F{$;|UaOe(hFDgK3A#7d-%C*X5iAl))m`Q$ zP<0jcGlNQp1p9dGkI{Jdn239kL4q-1I~uiz_~&pdV#B8rQw1K#X&JveaU?3q+~A(h z2)A?mihtuNbEdO*7H{|MO=Ri0vtH*oVC$H+oW9(8#s12Pa@KP4^6C#au4S*Yt*Ubl za5`~@NF_jS&%Vy2@qTMOl6j<)V07jD=c4|vatRk(KDCS{RA1S6Rre~RU$W8l?xnlU zrwaote3RNYHkTz7(L5Y)z4lF=-dk+R?UjG=nt*1tNwMYr)j``<9tBE2MVdDG#{!gU(IKP#`2a{1s01c|?=dVh~M(J?;J z6jU)TPbkiT!$3?4_ykv!d^hO?lXTh%QipSc#I|%#DTKfQA?cj)zq~*ok=cJGTL}bW zNq-=PAo(_|Ldlh4ie_>a4F2btCHB8xl9K=ZB_Ct*jAEGuBn|&Rhl9A7%cT@froXZ! zqJQO3LjPA@mIjm;COrc(bpQ7o^9#rP%14F$m1-3GUy36d$zRV)g8aLr=fG*{F$Y(B z7jE#;F}93?0Om=EaVVoWOo^AkV^A@vWfW>`k`g}?g-U=QZhLMT(r+}#|3;K3p(>|v zBL_`E-op~#3JNnSkVP0Ti7y$irMR|@f(SCf+SE}Tfk^q7NI$=ymcS7uNcpf}82bOT zlLU#d;qCBX1>hLSBo4#y5@=FQUGT&6w`05nmXz9(^z&yhq?yKCucu&yNXa#~{x6U( z7OX4`!;Kj`N5tA3EdGG zmXtmX)(($`kPV5XG-X>G0umyqfCf`Rv;%96z=)(gzgypjL6MStl765+EI|ONmJu(+ z;7D0$Nk9LzBa$vjY{kHyLv(`0BKU(KiGnRLpg$3fB_#!by%3KiKbA;J4Y;Kp5lLc0#RzS8RVh9lYVL%!<6^I})I}*bob_5YaL@W~#`!iqi)&T;_{GKrMOSni7 z!E(aan204JL;(?`pa1>vpJzb@ZN#%6!yAGMk_0#vAPFMKW(QwSaQ%Xmes=4qaX8Yg z4H!lsCF|Xa;ZdYq3or~r%Ap6tzzrL2(*FC0xrhV|sUpX}7)k;Ma&&LSB#3w-Njv=8 zPJ&24k?L;zi=jw+8`h5WzQ27y|H*a$BBU+}+bS$1fZ)OXObxlU6C42H9pmsMwF7I1 zCnXBpff0x6s4mQjN&K8Rfc>bt-#G!zEhPJ}NV zsE-1-3_&|0`Zra7>oODug+ew&VScaEA8~^L0vj+ko)#C1^6K!_y=HJM21k+$>+Fp{?f!vM!1F@y{T zwWkoln{b|8H)Lg0b-krK*pp#sq5Fo_I=pd^Y& zN;D2@2#i@chL%960#Lh$l+k$WvuG6Da|ZST!mb7wxP1*5f&fI25EvV97lvejAhxU-}7@UtF(2$4_ z@c=`Zo{|^|gb^o+K_l#CNl@Ykv2++oxF>;<#1KF^JeUxa#DYKxoC=_L&mW=nOW=Xv z48oKJ^`el22F4a#n1Nz{e-+|ygC2ye5E2d)^?|z~D3D0||Fi>j^WcL9Usl7pA6#um zXnP1`+D4KehedEY4v*g|xq#pT9u~qJ1Q=o)gUk5{1q(2eLi|bpKb#3dcnD(#V2Hg9 zuG_Pv*EA^R~Q9!I4B;rbsGZ*5+?NkwXpb~j83Be-w|So$Weom zXK?Pv0h0pJ3E)g*J3KS(T9K19RG<`*mNM|)C1Xr_fuDp>a6t1Q6dVo*3?2By zfm4eZCm!jtliVq|N&$ZEFBbd9L;$`M!a>C02}u4VNWjBqD1rnAp@G5s5VkfRL{YZv zT#|_ZK!jYvO9EpZHe-POl`X$`5twVF9F(M=KPwF|cyJ0th0%z~0gQk!5%567!pQ|N zg!;e(5wv9{q=|qcfp7r}sE;t70QOso|1}ZtApJjljX^9M%nLp{&~B)VgMX-{07i{N{c#(98{l{#!4RE*@gW8SB5yHY`T!pzn%73hX5?)tt&&q2#5g^@OVTdgdl;%LvTe53>3IIf&wvC#3leQ zM65o*kUJbtKr9moIQ|aM8vfaRKwTq*G^zA0oUGu`3*nIvNaqPK@&imVrr--;F;Xp+ zEh|VMAdMFS7$$sG0fsPC2n0kR8N?Mpg-JvoKp_0Y{$VjNZFtBNJPcG8;$a}T5g3qn zNDN`TfaANZMlgxRK+G6nK@*5SPusRq2IwI~Cq!W7z&j>_C@%~HXJhbyGjJk6cpm;R z2#Eg{V}Dkk2pl5#jENFNBu0v6!8=AF12JIl{f>S76_p?YgfPEBtPSDl5J9y?_(lUK z9HcbLTc|}u|4d{I!w4`11iUE_RYlm7L=Xvsi$&5Y1)K?q7(~n-gi98g!JG8 z1VIff+?xV%7^TTsWhFHIIW!F2MS~#p56Qi-f_YBft>Sk%YnJ0>BW` z5n#Z)gn3gW5Uv~nMnudGUj_!rg6v0TU5Mez3vd z4lWu*;gN#@r!6pn1ls+qwY7!cP!QO4h*2Zs2H=FB^x^wfW(8q%czgl0`sFI$^;aYB~=aHIzS*C;pQ}mY=XS>f7|`s ze;_ghY=plc=`CO|VDaD>h&;l#9SEo(b`FSp!;huVz=D8Vx&Ye-%5=__k&dYCIdM{y?ceu-bpXYGLDc$R@Fyk-_kKV-JUr$BFpww!{wyeuh{XOqoc!D6aB&MBBLI5@aW)Gu5K;R@!9R8gC`JiCMFi{d`|#uMjsQg& z5#j+{w1B($;CNaBu}i>^;ijenuoi(Jhpi{L(}IvUI6!@5J8-=uMs}2fodDu!G~%lE x77pB?5vjtVmV<)}s6hy}Irs}?+i-R7q1;sknZkAV1}8Y8>Bm=yBnl)K#=Yl8fgI$P`bMjk(3T8r4a$CJNm2t z{oKnNHv6o#*V=KI!+E}ocNO@_lekH-s2~Us2o6nZNCI?W!Et&4BybaK04+IG5W)?G z3L?Jf=zzwU)-G1A77myo2n2494WJ_D6$EngB0Pb>|NAoZ@5^8yH$U(HdztU|Woo!O zHUJO6|GN!IJ>BJD<^^&?`Tnxz146h#f5+wJ732o}P63ev9|*pV14zXcfMSY?{g)10 zBM87l!plcT2Zcbm`M`e}L%_e=siJ#1NPK*7MM8iuoJR(LMF|Cgxds1>NJqyj2<3+S zZ4c!I!|MV8DeyfR01jM*2*AS*f`GaC|IPyff&SYDL3z3Vcp-9vK%v~=)Ty8(`06u+ zpA0d82Tmyqz{W=mE%5iyATS>{kRNW_gv69eckF`%1;Tkq0Kzn2LB4+%5d1&(5XAfi z{=%j~|3zfcsBO#MJ-w zOHR!^#(|SiAf~Qt55OV;A@=N_C5C{Z+<%r(5MCGrfWuK80Jv~$O2q2&Lio5L|KyL@ zeegf2Ay)LCIsO^xcWTM-I|oE+C{zHRRP1k9)Chxr9l(5ugZB51{_Bte5NtZ80%!x^ zmNbBJ*jU1rY{ts2uAOLzHJZXH07AiLtAq0?7A=m$j3nL!;`9Rxs}1zZWc)ZWjUl~A zigh@^li}rP8g}#Oh!v2A|48)mZVMKCH?-6B&~bfw{r$lunEXEQ{+G5>Y?;9Q2f^E$ zzH8iH7Sl(g6Fe}HptV5Quj~EYJ6M-^%gzNqEK>X@?hq)E@%9D|6TS1jU7S8;H7$5( zeOP*M9-T1#`Ea?k;|g1QIQ-5|a9b@10cdN{i|<@rc#-1JbedEgou`EdTj^r7y0 zJ{WW5dvoN&{inu>V3ymbw~^aReaE31uwOraY$<(kzuQ?7=Ra!p8xrrlt1DQ?4Ki$< zfHj%&!#FT?@BPjXn^<*Sa^}1ak1pUQ1w@D6Q@sr@o@pJ$d|)j|JZ?UGSF<>h$}_|| z(J0(mzfL{hE@t1DE*!WWSa0q!;C(3+%Ix8R8n|0N&)Vkvr6raX8u-%oQ_n@<%eLtS zIZl&x;cK5ry+jYU(ktOif(;shb{YZ42pB=d#d>Phx}7K$_B5qr$IQy0N8ij7A67Ta ztFLs%@Rf04)6T<|JLlX+CmaIwy2YJ-heWFb@1@Nbj*0z?@xtG^T2|+3T#d}Zv(0L* zYKo9WH}+Q4zC_lh^03yd;aA-!INZgNYdzrZ<8E@4#;V9;yd2X5Zhius%>><63fQIl zRAklI#;HkQL?29FWT?Awqi>KhCb2w&;4#xwD6>fpG(X21ZQO+aggS7ThTgv<>~c9^ z{stgTnuLKe7ugl2AM=9=xoM{M5(~pz3befDf>)M53~{^dp7m`MFXa1?;$5*66KRV= zDkgA6?Eq<~86JC@Hj*pER2fp+2QzHfDy&X2St6cIyZ>o6)>ndM!0fs4UP_?RL2hr3 zqy~2_KTt}|C_W!0=|{pR7P5(WGH1H)3wv!y`?WAdTKch%Vm)meVjdx1GQ$?}S!6^g zOoU|Jv0ChH;}cQ8OP~m|DVeZh09nR^Pd-{WqZYg_4>T|7XOoWmyv_bdvxNK#Xyee2 zp^20{g`ZJ13D{9kp6ia$vLA~(jC}7c>iIM|;o@!Yj3jl{l8P5=8gd{Dpc&8lsg1mxm|NlT6Ef8rV3&>LRIC0tO5L$#FuK3r7ix|l3MMHC;$ zNNcKfKrjTrtAftAAg__@AD;IAa$Y*%RwXYG-=Lq`)O^CByuH|)Arb~-h`ppbPT`N5 zD~iN_&xb>g--12oKfOFk0SX%RUzF+kYVF}SG3l8x@_MP0;{FtwWip>Dhn~#I=AE-( zo}+PL9%X*};qm}h9__Lp*&WcgDX1g$X7Xa-8aEjB%OQqxmP*Z38E!Q4)n+~C;z!QN zOW%xh)@U9RN30Ttgp5cS7pMoisiI&$u`q4CqIP+hJw{e_rl<0ux{Mbd%ehe z5*lQ}{fRyrhf2DBvm2|HY@j@WxN&ztn366}kc{9l19Zi%{rXB6t7nUA5Iw>YW2cDR zNw;T9db|Sc8M#K^EMca#bGzGUribQf!Q0l(b#MPvyqUgwYw^Xp6ne*9w)=NfZggIj zCtT*&GDfvM!)69B|B%?tQgD4ZIiU^N=QjMSk1t%D1-)$GJL58CAA+JXa%7a8sC!zz z0(CfBPd0e%711(BhsltLN>aMfLh;Q;eH96u`rNS4w8~pC(FO=<*;(*xv})NhDuXPT z_r7#q6YZJ(lqkdizTqGyW|emJZ6k~8F#?Dn_v+c;vwv}gWs(sEn_<-&A#0*>L{@s6X9)bOmk?fz=k++|>zaA_W~W?`C%bQ$fHw$0F8(9vX=nYFoMp`M&! z&)SMXlxCX?dyW}M6Kdvvf zTSIKNBZ`C}Yv-QC*%LXZ?%KsZmR zm>%}*!y7zC77Mb1UL$+W+$e|h7K(##&GUgiY|AGDW-t3s%GRgIJiD&|%parB$6H?n zk(J*r!rBR^pb3rHwHp(_kh*i>K2rKRg_xUDXt8h263_>z?lzTR}V>Q#0# zvX7|y0~(WtIcme*{Y?qoa0`%Zs9TWevLqPYVImMq2Mga~nMYugr(w%uw81&`J zlX6Sw+k5c`$%N*FKB6+t1e$pb-Pu>+wx6dX&&qU#h|6A}TZZ-n&7*a13SF4BnvgUJ zCt)GcQ^^F?Jjpwlev3>5VU*Z_)N!j*LKepODaM`A^%c7$+mpN8G8)4U&Xo$O2gb<< zWHJa7AT80!)BT;3lWt)N5G$dr$NWQPIq;Pi8Et9NS1pTtm?d6rPk{a7*1i}g21&0N z^*INs&2j6nM?Ibib=Z9?n$9tE2huV^EU>K;%_&p=R~h63(HKE9gDxN$C4qeM-q&W) z09^Av2;mp5_@Z_{CK2Z26UFEY_4+W|6SgP0lkNJT zI;TwRuRPk`JNhvdM7b}c&MDQ;2E0YF9F!y|2od*f*4E&OCJFJNK~2<=O2daQ-(Y8Y zV&iGe`sNoPXFS5N(kyNMc2B1^tEYIHoK4$h*8G<0`vuC0-0d`F1F}|Lk?<7JrD}p!-lRn%10vP89ESr6wf|#2VjS z=<=SV4V*rfNk-S6KRhc%;NxUg^r2!n%4g>_MlM~S6d6kyD)JW&21mge(UKSE0J153 z|Jmi2fY;50^Y7)gcLW|pV+sO~V`9ThUC%l^wmmDQzlU?)$TF?bA`a$i!pP-2j*qY; z#_N0FtNjfbwN4+Mpq|H>UvFq&ovGqSIkgv#L=4VGVNCh56lo486q@f`6rzB|$<-|* zPt9XeJZBeB?#W&{EVLp=gKCB^<@CENY6k zwj0|vo(JcCTGZiB^_!uP*If)RoZ60SSnRPT!&3sU8v3aBi_OXkk){sv}X7 z+Ih^bXwObPqnRna{39%o`gH?uDvL4PdBPmsk7Rx0()gN^w}>WK|LCb zV}lOY>Mu%5aa`}kRUBt579nPe%G>MmUTMlRC6e@NL@8X+%D2Mpwi%h-i>gx^^Z8%9ObjmyMkNVh~C# z)p<8$sTLk!-{!-@xij{52v4ik#+PAVa;)Esmm`QnAs0q%)peap@aluzLL-rW)JPVH zg^ZUXv$k4B@<=;KaGXH2Gl^p!qLJz0B%C%UR^oipUH{QcKs$IcW9T?DPN$kY1ym1e zANxckOzEi1Y)SKUp`U^rmR01D_)O3*B6pNBoQ>PU9!ftFWd1`{md4?D+=c6V)D%1G+856^!uRCjhZSsc;sMw>zWs;Q8pdP)$dkS z$|<6p?mX*UpJcNi$%$i`IWi#};6_NG7iyv{9y! zzZWPxFlXi5JL1~+XIJ6J#7m`*%MfDQzi7TIQ z=QqslC1d(>C8oc>f-u$T$Gmh+rW8WUFKOL8uHJs*UD`kUq!x9*vwHpOdtGXr$80Q$ zP@NIj?#0m7Aljgc#w7Rk7V7yRx3}u>yWt9f&v8F6Q9&gn zGWKtEUMy{-GaqX+`no3Iwr_QvmAu%pGyX!IaB-%*;AcE!tC{wqngROotqC5tayXmPR??!?|?FRm(YLnL=p6=yG>%s z5Ox317dGl_dxc|yTc&ZFXDh(JL*(z<j!@2&4V>B1f!{wXe*SvYLwLx~_4G9GT)v9PUTcPr|+s z5oF`o%=Z}fj?RKdzaH_|5PMp@$QP1|wtxLI+zWK*LltQIvHhh^jpG`*-Dm#v_4Lh_ z_=uP#5&_#opmVSp`E2)Ol1oK4R4fBd2$R;MsYKQl(q1)re_j0L*J#-~ZlBMaw%wN7 zQ!DXe56oON8teiz8*VqW2x~h*IWvez7^$y9dw`Pqc}P?Zwzw@}?-2pzO%Fds&5*Dp zvi0VTQtfT8y4Uw`hl0V9zH zEMk=)@%FIk4;new2}N86^R>h*itYTedxb-kp7yajeDGtLrV%$I3ydv*e;aT6e2M?u ziVSt9HRoeP1iU7&WufJosQCB8I?Ts1Mq|73oVH&X^ioM|-2DAl2huzyj(L;E*it+v z_WB%s>{iv!XgI$*ANSEn?eF9p*&Lu_?fcRWhT)-spWK!9YG^ia)p|aa++|oLO%4V8 zP&(#8tuKTQnVuS5r7B39uS1Gu&<}__=mg8yoZ60DK4~n$N%!6m6d@8Ign7=iFbE3? zjnhMgga~9lgk2`q5pL!Pk|xA`sB-4RAN-X%Jh){QN#P0cc_u81)3n6x^Qa9}gf=j5 z;_*nD6mqZHU3f<>hWIs<HSmVNX1 zsQnH?1hUfnnF-M^8^ugMSG$0{3fGVnefP^AWup7oWs|R%bj>zi2cTi=S>cDWtHKYX zk`XMVsCu`U9Thh-HFq=6V0LEt2t&@6k6>P;H>a2BlAz(#=6E^XP;HiWxZrE`qe$%$ zqOyxO^(b+r0Tf!Bp@lG=GTwI#8cn@rtxg@$WL8HFN9I=oV$Is@U)}2I;XDS^wULT$ zsftO7Iov!g@7Uj_mnF!|Uz~e(=EmPx?Goe17~&0UPT&_mmzBr+nZ7VFan_V?R*3~p z4{Z35u#A3l)o-(In-*NwoK-l0szovLNnLdbMPngop|MaITdEp%Xgd)vPdNMe$xKnd z8kRA#7Cxbt{=8YAEef&6H`+3jj^{sE!I9wh@CrQPU8iX5;DDxen;@gal4f!aY2ow5 zz8Bq6=1>-0=C(66sJ@j1`YJzJjbTqZui7`Q$E$A{*j%QDW;$%X(n#~hc|_{lbBujN zUVclkY-i0P`7uQn#sL7UbCe3Ful8Pn>iJw)RHxdu!adJRLjm${YvlH^gJ<_LTVJ@v zg3>{+YB9Qm%2e<)!o@FuWp{Gb8HdV;sp2w z4q-AJr2Jrv$I12CzNJ~OWqx+kE5T3p(jLu)wGV=g6=siCFV#~8Gy|vjq6&#mBPov2 zK9&}=Hdig>??`;&1VB&0`*l0f7v1-+tDDiBFKgpqEF)9o@c#K% zG}aG2wrL#Erm1H^f(;>gZaOq>BoEWtoEM zq{2Ljw|l!=#B~P4x1-!p6YEZEO!;w0BYW9rIOY2+PQvKbw7-Xr(!K@geiP=+{BdLy zdg_St9Q-yhnm}!4^G=LNu+-Y@zLWuVoY10+YO+T??sVUD+$4X!P=oHsc za~<2fMF&-t_YuPf1~!IUlQ-1nrqyDeSb@@mP6z|BqfxNx5_L2tSYq%w-xKU&`c$?* z$anFMtll2=M}{$-RQZ^N_0d`CN|+(VGw$_)IAOu?*|f;Fq>0+r!g(Bnj2Fzic|(km zQ@COL#f@ z6#JP`CnwC4P6{lYjshqDtt^cxA{ z1dtylx<8edO2OFPK>L+|^Ude**2ZpYJrkD6{Op1ngRNArwy__>{Caj7gR0Co3&iYw zqrlssvKaLG7ZwgUQN zlvP&oR5DZ12zuF{07|~J=pH^_(t7iwxZ?Ea5`#{dPRfh=cXp)pbLkE_71>L=au%>w zlJt;tQdu#{=eZnQ)V76S2gr`kHeOQ05=ju$nto`>m&*{_wzuySX6IXpYjV~j%%yGA2SsltH zK7K}nP_8&O52C9C-Cs2SzdbT$@}ZUa;XS|yq#N+p4AISN`?ZQ(Uh zfe*(!H#Gjo$rl@>5#qtueMFT$B&b*CJGo|Bz*h`bomb}``0>^zAzI$?1}(vqT&dMcf5s!jsHJ} zV9;-dmr7lj3-5A67(pZuMt_hZA0OyHjQBrDk&hqr7Zn8v!f7M{BCKE#V&;FxP|WxD!2lz$ z@4x&S2!)HgAkw$@1K=Qm1X8D7^H3ww`3JE=5NurFH(y7b4PFEr7epXmZ$J**%pbAX zeF}(V^8EqW#0Vt)uN8#%Z!2i8D1mLHmh|>-W zLY#J91o!>t>_8BhS>SK|0YWD5#}h6PjF?WnB7m2Pj}AzuiOHjB?cwV6!qVD<4sj#? z)g|fT;_CVPXF<3IHGmCCKp++HDL)mDmx>b6hp_oAb1avqipSB>uHAw#Sc&JMPh>)Cw?7F4lFR8nnZsy z;K~;o2Cy#`?Z!n5H(sZ~xpYf1ql~cj-9Uf1`%d#dpmijg0W9<_jLINHS9m6Az(VJ_ zjg{5c%ZH2Y2OQY-{Kk!%>Al~T*VXwn9Da}q)qm*f+b%dhf_<`%Y~;r2y1zd=JUQHk z!%aUvJRCB+nmoY9HWJSVzKh-n-+rHmwFj<=AC^9hz|~+v4@+IR0~?9EU3Z5k;tv7$ zTTL6m@Q3wFc-dI&dDm^n16*WMwfx-leH*OJ==S><>|tnwc9%ZawAxYtcV(z*w?$S7qGy)JCW|=vvb%&Rp0y?&O?^9-*sf)d{+lf z+k@m!%ZIS7im~jX;Jn+7QQZD?B$Mmw3q`NDqm49TXKtl0byaJ0M=gz^%SXY0H;>UQQJ(WARraX+75CxU08Y^TSViLwlrQy>j4t3bY}q%^eWylMfkKWlwm7 z4-z-1=P7dvy7Czx6|DMbOyZ{K6(t;p6g5qFw5%#=V}k5Y&i zzz~cB`%rKqq${&H9+qXtr;Q~jY7li|(dr9F*Ct3~layy!u&=KKhUp)3#wB;$jOl>8 zVw&*H92=fJExrf5%$dMy@S}3j zDT;m?Gu0lCmZ|o^6y9d)vPxrmcyjIj%4V>40{dr|`_2^3@9Npfx?+FO=6x&O!;Hp7r+PXz!*Yq^~&T3s(tRwU|?l z=LSa>csD$W`*6)3DM)AMU*De0#zsT?iq($ewGtNUR9Fu3w-0iKvD(qCRD^HJrmbxS zV6IsLQUt;HK&w@8fyo4w9G+M7*s0RMoToTg2{{<>FmvH3LA21^b)7|KC6Y|{XIWv) z>fyjV{RmHJmiXpl`DFRm7aGi=#HpS9RDi+^{#3msiU%*opKL|bcrCev*5I2^n2gT#*SemtZZW>c4%-F5Ng+}gM_nca>e$LR`cU`-?8RNo>#@2@b zEUS)cS=vG2+SP>TqQt*0ZZ<@Jz&B{kmn`Yx%tPoqb4Gjvd>L3HK1)zq@41ZIH)yoK zdzoW>x#R=qmTlLYbE`%B&^^O6$e(;AO#aCx0?g3kOphb&E`<(kY&I#1!$S6Vf)(d0 z_laYovQ`uPFmn((2f+r_u=K3*Xsa@4I>|v|c(u#$ z92qT!QSbHw`&v{wtJby>C0L^jX&ESpCZ|(X(C8r+RQ|G&J_k>6PEIR8vLA+zg;26x z6?_W&e&4fFm+MPoh)dx8ov}e|HLP05RH@dBk{zc}F(AvE)@gRYvexS^px^BKVP80A zql=nV=JT24Q0zyYMa*UPI(FFp5op+2n@J__AH{`a(I50dNLWglH0Y*C7#$f`tBHdN zM`hHge-gw?*}M4u5<=TWa@~WqT%P&5Y3nqRuohL04Vd)%uI_Y1jjhlwM^#Q(-r&RO zQM{Qd8)^IqP@D|4rB7z~8t7y`8-O;AlJpoj!l*fRo!;;doDY5VDrHuwdO`J6m93De zef{f2a`y?%VdKCpov!ghZ6D8Rx9Ozk3yzrXwaibYr7}^yF-|wy@4uhHvWY+{VkCIfyc9)*;>y}h?}B{y;Q^E4UD>G`&uWX+rUn-NGzqwvd6 z!tz-VmX7a8eg5JH)wZScWlb)%PwJpoyLT(A(^O#vwnDqURp8Ri6eOZDIW(IDvR>Ox z&HjK-v+3u3m$CW4eWpNb*rKF*wkg4U9&iY?Pq^m`Ka1_~0VBgj$uo4WosjPHt!NA~ zRbA&)B^LIW74!%;CV;CCtxO9}2Mh3Qh>>%J>>=B+1$$N_44CkWl&OAko~*0N1*7_1 z`-5a!_;cL^v4n^4=ksD>DQ8cmiDM4Qvzcq^u4VFmJer~7GC&vfhy6UnSWGy!s~WF@ z22g8LJ@H#eV*(Tpc1t{8#Ftmne{!{rNwtG-=P>ouv`BNHTc=}ieW%@LA$^)$TEm#s{+!Bv{wmi;H44lm)G{22z|fQViZsFj6c z2+zRJ2k1&gA-DZDybj}G`jAuVcpqaDtemd0iOd^Z*S7=JCO=6oBmk}!Kfev%N1xn_ zko5sj72Al|Ma1F{=k!TBB4M!hrq}JIlGO*2rLc-^*ZNG7!t7+1x$-x#6&tX{d~tJH zCLgpHo}$Qi$l<%{|GHWyH!&8_4X;@zW>4RT>n%syxRH84u)H#4#;tUklt5>Obn06hn&pU%wDGPK5m*fk8!YjULr?eL(1m2Lo+X* zI`3G=qOvOLs)E_Jy1LzhJ6BSXy)gHcH@q9s`Y0P*OJOISf*lw!puh$aI+v(y8psxbi+)bd6(<>s^UI~gPiC^wb!ICR|=24rGpq%RL z%=Mm6r^ktH=P6^m0Ol>0HU{&nx&{~IXF#0vr1^|uP>O>Av8jk_+}c2HfVsS%j(%Ty z7h!)t?1frz-x$hSp@9bBVBpqdmq+Nnhw5eLbJ5|na(fhA&kvF zNi4vNpY(p3E{NpY$5|{CBGp8h!6;ywxAWxw3e1Q_CL=AaVB1pCMs3P?Wt{d`xB$a8 zy5mi?LnwBUYqBfqw)!x(mdFSs+akMT4}lgTk)UG))yD@;tw?6d`HT*@hG%^?sZePpttfsmtbKD&5nH(6K{DLVJ|s{|Bll# zgBezWVgm)tn$Skwcl~_4`5Z1Kb%6oFx%a3TXz%n=Y;f7~ zXvbY8T1K55S~ZG&?rF;8KW2<{_@Jhu&auO=Y2`=dunabL!-Y&5s710ka3k)ix)LPm zmx&I-p6#xCmD}QY?#(0Hq*k;EZ>~`H^2!W*hsWP^*$>hlL9BqL{_UGfRH%Rmh@$5ir`BrdS>Xd=m=Nf&o#A{i5W&W)+ z(&A)ZjBuAQ9Lqik0WSldn4kUW3-iXFr4{BcI28rF37N1NC4`?`+!%U-fF zQ^tYSlxzyk(VSRJ%+F8h+tR*R#ib{Qy8@lT-*ujzWn_jpPPUaXg|IkHmkN1e?PHha zl7751!PygOA>c+g(wf|g4cQ?qwwc43-bhuzjo{EYSGL7uo-1Do|2)ABfE`g+&SEgU zl$kEKr{zI=9+#27I)(@3YNro&3{z`BWUYx>OC%8(S?a-I)(AZ8x<2A)-9ezPV41n{9(Q(ATad|G|O;PB)rt8ZYW^vSnrc0G1$18|^)oAFM`yjY~* z>t2YSL-z9Ib<8>ORk-ah9o`jPaxkGO6g{}Wo^HUmHm<#-u)+7u4giLB9Ix+!Y5d8o z(q(MYU4;N7)ZL{|W|1z?lTBcd@S>cm)qwzmOhH5MX>QAr+#2=R#^6xoBClC?G6oyx}JPUZufU^(xzi^8DIg z5$5LDr^3^L$E9w&b-k1?T727Rbpt#S2vY^seC&T!SQ$ht2(Eu4a`dr?i{{d{Rmb(& zQe4ESc1Qf-O5iMPX}bkJG{;|EP!Tsvu-h67i^qZ>45)e(nD9ug$4VL;fLx665^t^k`ovOS)eS(0u zH|l+(j%e3-{fo#rP4CvjsYOcEitc>W4^f-D+x1>IpDU4Ik*{Vy2S|l3r6BD__gFS3 zkCTzx(Hvt4IK(`d?NxM=-m{DhnM!gcWEPQeacil^UOHI2Fb;h$Vz;MpeME}H&RC|L zmzDc1aHQB3_tJBVK`@SOi}Q@RJ=R8hkv!j_mvsVYr>c);*JGl~X3!og)aUzX3{-|v z6uVb?X^p)GllwfXtqAkJ>NcX2H~Z!pUj@?tn5n1O}h#wBLXmgrSX zl%rIrYSr0zW+L}2_jFqRuL&1T^kVbMpk~um>RtZq&*(o_p6eq|d^%;C$j+2a46Q9U zqu{K_)G3gum>!m5Qcn98gScZq{>W6X4`G`%{@SDF1Cyy}N4)VMD$-t!-{}?CsAQw1 z9qC8QtahktF{o+~*?I;Tz3R9ACLO(=zxkGuOskO1ZakEg-8|`MPGQ0%wC4ux7;TK_ z6$AXZ!L4)o(5vQ)q~lhA3 ze57bkO)l!Jj*sCJ%Anp#03+rUPU&m>CxW=ED==5q;~J!u%eVSruX?GIMeFsdVy+`0 zwTImn7N@}zopLG=-)qg4CJC?6a;+ySR=vwT)f;qGI2HrsxP$06M9_ObkVgN zhNtg^usV|Tv#x|+jaGvF3P;RCXybGh>Ef+1P4rEARp>F47e4yl_@4Rs4C`7q_g$Jb z;=%|Ec|B?4;>g0gr7+T+Al!jPBA=Q()~w4{o&|ga*&ZN)`Oq4Z$&_|-^>m2k-l!*^ zO1;Lf0!H;b)^@njBUI>$|DoOZ!_@puSdOdmiF^2vNL8=hlIN?l(guTCcE{j|(N~}# zSgrA0bMPm^*ts@RQ_snRUlJZ2V`pF56QQteq)qn6pt=Uh@;v%kJ7z0j>4F40w#*zJ z<6J=39y551NKZrCJ-5-NwR9j&# z5d77?q4r~OuO^?~A1-??CGU&Y9QRI)_N|pj96GtDqMzq}3d?Fdxcd5*_r@dB;+AzdWOe0?vT1zdsa?P$d6;7{ z3+I$%u5FTX5$L6L7r{B{B#v_)(&(^{T3e{1KPL0^C~A&yrLmN_;PL7DiMQsAbCe1h zBh}!tis7`%IyX@VRcA#XpVE0g4m&7XEO@k(!(ji|d42n|XFp#35nq9_DQkUaI7O3f z!R5jE`y@F|>%Qg#WhTYpSF!ld%3!a1lVq7+E1y<@4ZsF(c(mEYbMq`_^>P=efcxRq zmX9~RKc=E@0m(nUWs!Rpf|1`w>i^MHzLnoG=o1RX070Gt+bRBm5&Qb&Mpb~%_>#y8 zyf|EgOa;ezsNOC}n0UMuB{qPX7Kb*@Agn*v`#awfT0RMQcFb`J%Oq^KeFXM{MKk#+ zCR$!W^;HqKa;~KWSzV7US&DKwHnO>srEG%Gcd2W*urHjzcPMF|EHVB|_hJ$IK$fT; z&aqgK?prff;W&dEb z_vj#wy+2PT56nMi?UsbeV4yfG!n*x3I*2q;6pL%4$Il=~H3w*PCKCbUBhIO#DF-?G z(&*F2^(ld;u|CXL`UfLvo~(*@>p? zIlU$1T63?KDhz1x4&B`{+?x)o&5PqG-_l_nluAxgH*dNikTDD!c2wMWsoU$)$yB~m zVJ4V{+0uw7?P%Mi@$EX?>--26J{w#wpFk4H@>Kb`@F-SvX7c{g(K<>sCE(QQj5S&! zLj2?U$l~Yvy{YZF=3nNXv9Y(Z3;lTqY()gCd<3l&()dPAwYXdQ%CeIo$ZuaS=|S-7OBX{k2#V(z%Jz8+^EQlTcACie zHqOxAZz7GWZ}VbkU9sUXNyE%~V_Uq|S2SRj<`D)Q$~K?Ph8sEripwuxZBS^PxnM{oY`Aga@GL08qPP)sn z&kw7LEQ@auA7gPT(PMZtyS)J;GE1cM?V_4SWcXd~uL~#HQp0TzY}o{pTl&^~zUsE0 zDoOjP8Q_ltVQgiYYo)|UJ>QpV>*}wcl^tVl);_xAr&mu+GGO5*mC&AXNm^!?6uU~ zO$C8IS0+ANSbOEm?akCX)F(Fb1e5lR^$fkT!Z3>0$3ln_J;iF|kD0aM)?@unJknT& zIuVu`%4DajMUK0sf3?(&A)jyYCayX7TWwlfEsrN@br!-R#T}J6Ol0A>B z3qRH1QD9bkgXUhyLc26r5YdL6+x6_Z?835Ras`9II&bVKh?Mflt1WYI1lAfKk{_n3 zjW(>K9xwHIei$gkAl-P-XBPvjv*Q=9xafw5rUq#$J!{!qA%LayEc!_-vF4++cM7lH zC7tTWVWnJ!@q!-Q%4v4LP@KCyDfO7CwT{QIa3B$X z6FZGVVN8WoVAuNj!NQiB4_{ZOpTF^Sv#=U$53|!S0=5gj&2!Ns8$hVXPFGigLv%TO zn+C97GOruomuX8;8po+}D5ir?fij!a_V7S^s*?HD4gy&ROB-F-bk*Yl2%jhno9yxFPO^tBNb!mkyP?1m z)S~4?=LG?}e;JazX~pDFrqHOU>k zN*x>z4>pbi9$j5uU-{YXS$E~2VVH#OyHjhw>JPM&hUcm;--hY>-xz!ydY|}okoKSh zAL52mFqOU$=$g1{ zOTE6F`CsxqyH%t(5LVFqZ$(d*95_&qeG-b z?E3^ubZNuwTB1U4#!kxoyf%lsbMZ*mT`-TUNn+hbqA84Dfpe{b@8QMO_Vxqd!SELU zcZIlz=L>6d=YMO(=>&cw-QRGQA5qEtZ?_=m?_NW=J0CI;J+ACQ#~Gdl`_uay5C2au@EY*j_BSm4uNRpQf!G}Wr7GW3_BS3n&CqfsHU5v1YAe0{eRAqyC!T$;t zAP7y&X$%r2T-_Dm2Y|wpy%0)_q!NT~0s=;8K>pSUfDj^;KkAR)74r}vx8R?Od4*C$ zh%|2ikRG7{_@{0kjQF>F3YnUP7x;(9 z|1ag|1^xl@m^^<|eqJD7DgcuW?&}A@P7MNi!EdMmLQfFA|L(*948RUBhKKq9NHG8A z^}Ij`{Imj@9ai;qka%^?bjmSosWL@7g$~8$XdK%L9|%r*Hp~#G5Lc&y#&dRA?4l54 z$PvVxSEH6bNG${C;v*P3;TFL;rGeccEBf zz^~&=6)ot`2%RgQ!|!7?p&Dzj;9na9)K2bKx5r`oA@g^q*G=A`Ym|nNi;Z)&NO9{d zcbHh2_97 ztA*rO`9QhlatFW4K45oKsomzH(Y{AZh&ZLJ2_cUc02;o|L!*nCN7FTP5TM9MpB4~# z)ozezgLKvt_H|kf3}At!)q!#ggbnjRQ6-A%<&TW~pT*%SNf2hr?U=`(b8hgUPT!I>`sJ$aSOk5<2San*>d&>)@A^9hO3joj1nd98^LCfogYX3=7K zl;*5KbIYx80|GIcasN7$!Slx`L-}0nNVZvsr#;U{>6W0Bmk_pE|yFExELq z$Xln*bR+smpiP)GOvXm>-a!8;s1hQ*N%5Ygdx4xbBeKYMC@wk#-mjNVwi6=B{${QE z=ZeJPlJ!9MiWruFw)OQw$Exu650*neMaY`7?(}}-+!Lu<)MN>K7w_iJbJwKy=sp{> zgwX~xZ_d#MxCEIjU_HsYp1|YI)CZV&0p!LuT*;y~bu9Jz?$-VMdqDgY%M}f7v*TG7 zQJoUa9g$t;s+13C@95mhe4X9OJM3}|DaD-I`I-y5?JddH!2@^$wu@60?#PH&78-_x zRVGD;d-*uJM!wZymEt6x3r3$#R6wm{FlMAvN(PF7J=Zu*k`SD&`d+b)$XEBZpThnR zS#KQ>)%Lv)OGtMPASn$3LnBCofOMDA-KBuQ(A_n3N;d)`-AIF^bcb|<)O+yi{oLR8 z<&Se1&fa^?UT5unm|4&BEckHjj1cP5%W0)|a%v-DPl|$N;q_K%a{~Lb%}sZcEO5>~ z!Qu66n@GNazipcZ#j2n)qvVuL#5$+HBzq6WdFagg^|e3T&2b(i^8}r3Fzj3xuhAX} zBCRj|29Hz)y}aKv2T7Ryk_HvgM&HYcjIYHFVA1pGI*1G^|T~#&o!JBUd+gr>n1vf)5RgLBR8$U(wt(S zzk{QpB_3Kwu{o^jWjbmLIF71Hq7*zkXjan>l=0WUIji{g9%}m&20|@Ei!JrTM&#K{ z=l0uLk;&S;en!LCYlfG_NU4GP1wL*^$?TGCn>$BK^3_r~qgDxOe(Vc~C}n2jt8Lm8 z>yq%n&YeEI!g!ioxF~*<;$rZQE+6{+J7C-Oo-qfAR0|&w9^wUIg_D^7;Qa z6g#L?pDKr#4jp*ETcd5))M1247L8ejYD=Z0F5D0x(PpbsH%ZEzVeBL*Q&iq{Y>R}~ zm^vOg>+tR59~ShdwN=MT{W6v1mQHR{C*_NmTeKStqk)68zs*yogO#Zs)XUFKEjKqs zN372Oe%_=%uPg zsO_PZfxtUiw3rT1_+gql z$eklmvbR~+w(V`D6d4|?i-ADA#P7HQ=`(so!ggLkvKc$ws+pJ^z0YII=})3?bUJ=>XiHs(f~Dh-U)O`-lXH95#w_4)Ld7HHuG`y=QtD__ znmi7CleZJF)o!KJBnqFRq^Vai&5Ue^D$= z(?*p@NwJ9e4(1(~WPYB05gM1;<#^bdJNoxk5kb-GG`YZnVPxDj=c~5%_LWyyXC2sE zQX@zWPjb6)S=|UKj|;lxkL|d(EPc{(Jdm6sA=b0AQg7*@Cg8;WX_-szg`6X0Iu!r6 z;Z#?4p%&Sjj-Lz1oAztmxXs;6-o6tG{=tmWk@L0#MBwHuTaFc3#V>T%lo$;Hm-9d|`26ZUM=)TcJE9O%Xh`4Ud1O_+%Z<^k47=B%#tK)W(uiC z>l|91g)PY1W%ViP0iy;^lxlBI`u_UZmn%geN?n5?pkRwM*tzG}AL|i^d86==ILX1@ zFad2zIqdog?j{~@`_VD8%SBt?T6H)3iF1)pxU32Gy1)C1JnTOGhhP>Tu$9paUG5Fo z^}~p)(0Y~;hy_lohrhx~x*+uK=TFEvS^)T>)$&Z-*R=Rvi_yT~hp%Z{Af^xUO9OPV zCzd|AVIm$2NdS%gX}0!T2`3A`s>y>UiP;S}ZA%-P@h_DqekP#yF(OY?Gjw>I!UB4n z`hajG3>`GQz?Yy_(We^rBaFYwR>}lE54pH9jhqZ(+SMxbhD>fDt{tggMi4>J4a(`Z zJ9PG!tc02-1fOa(&de6;s({d^kfGZ1$H=RMz_ITR8hOlW^mO8EgYSJ5HtBzGGU0myY+q z(gEYt#41KiceA>&$nb0)-@I^R(dE`+R60phXv}7T3S@4+lG8yKtj#GeC%0BQ@c`sW z2#&4DNH{WLtD{^XYlY#x7Y0w48u6PtPpB6@F-G|2JTWa+h1UoPkIMIb9c)8uc&JJd zB^}%`(fV$n`xT9S6%B@MJTx(R9BtR5sL0Y?k`jDH^q!H8EBKbI=U4W6DaF%p`tVr1 z>7D%Hy(3q9(y}a3g<9SiY=eG$o74hq#bLsHjHr;)-S>855cDvV8n~X;7SETSIld|; zh+whrur879ENW&9Ykmc1O{#qw9@a>}XT$svw?MdOt=g1q<`$=9TEy zm)m|$dl%|K4jUX2z}!Mq72`!ZLuJ@=td zSAlj$nq*XW`Oea3kJ*Dx=j^Cz$FRDBw|eQQ%Awo1NcPGVWo+Z<*kbYh?5c9C0kPAa ze%NE1uNsxXWP;psd>P6`FU)NF9aZ-XWq4Hf4Ev=aFZy@dKh131MiBj831M$#wFdDT zlPtI*c8V=}!oFeN5r>r+LQ9Zy2piLo>VzC;elRFf4iqL73+VJV;DZo~ZPw`dislsT zoseZ*@Hz{GC#;t=iZg6g{riOII!qco(FzxN!-3CC7qp7e85!X?wJmF@C$;nq0h$?(o%28Y=ii?9sG^#j zD5z?Y+LAW3v2T8cHG|3(in z>JVdu@6p}yH-KakGUhotqlkt(G{P_*zWK78TU6_M|3oS|Q5rV) zh0>tb7O_96UgEl4G%^N9-et(yq_F{s>7;L>fUK2)ByDfmed0Xou&c7_vm0?Z`u?qa z9sm4+7NQ}ETGPPqX)cX{9N`4G1c9{>83uuIDe~u8G7o0?kwNxf83#~1x{%Ft2aA)b zip629b?^qh<>G{Pe(S{hf}G-PO?HYr&mmS*+~pC}AQw3HE%^oq-u;qaKWf!&-dat&24H1?{InY6*$Lp{X;5txJzY>7eD>1Z-R3Y z=0WS@t*le%kYBHG_ixQzS?~;Mb+LGLva#r2bwI*c#B!!|TnBy`WEq}_Q^9Cski`3t zF$DjXPC$gZ&(NsW@2Ye*a!;upa#pB6%e`-U<1IGtTIJ|LYipWBtsc;x_;Wby)t1Yr zP$D*dxdg5nv#Anyy>i=wWm1=F=%nLMV(*4mw4iX+Dl!*ON>^#vUGNybg`4_|=T=mL z#P+`8U+ESfe2+TdRZ)Ak*Dp<{5Wg<11jggC5+7Au-5 zN}k_qFdE2np;h_a1l1ktRya-{{MzMlmg%xzSKz6YzGO81rcsMS;FuZ(9CQY$nb)Tx zPLbs?X^~O4>;}z|oL-4!Rz~aa#{mb9^fp;sg5jn5F)UjFku(2*d5q?3a%5ByhoYED zVJ+8U+beG}y`{XAIyVTL|FQeF^GjFpSi3cbL3ED>2a$!VFU2lWwP`XNX%jBx3o0Sm z2LY0d*T{oIW@M1ywC{2*a&+#$7)}vsNutIYop*LH4q+k>t6rp-;7P}OH}rA`EW@sm$mb|k~CzzPNRI%gLx&Dmi>~1q4yG^_7ZP<@sc>V%N(Cy ze-Js)`xQ%rXGhlstlo04y2a~u%^cst*-iOgyOrX+;>Ol(HHU>mwKoZ=+cPszY289+ z;%3sOrg#zc+D5O+?f{trn(z8o}@EN$0S8^i-N8(Y-zD^mK#BRrOPRl3u*FBXGhN`|(mkw=QTIv1sLUXjP%$Z6prEV?n`EG{>h#lK{VbP_?+51Y2_#;@+w zzT;0VRkqQWLgsj~WhY}qei%8w`97~axjqzN7EtyYIM|q?jD&h96TWw6V-Vnn7)gpH z?IQo`_}rn+5n$F`Q(d;rI42p#`*%x39SOa43>nHjV)jpOIIH(eCo;H%zXb?p6}}!c zKK>NN)if^%T}l!-XEPbMdTd#6FXxlAlyxBoU!A!Qi9V%>vB}YsL3JQvi*P+VB_umu zMlZfZh?#{|^bYd}r&xb0Xh63WxAs{{R(}DhnDkEN^Q=|TC|NNK%1BQ+(H!}J@DirE zuXliFTQ2I|i_ymMqdHx(G9CrEr3w3t4@Yu7lBj!l6%G8lhzWaNm*VYhKsoTv?vb33 z&%PPV5hO=ePo0mvju_yNo(zwh@zo7h^$!bslA>oa2xPX&gVmec#*UI#Pks1@m}y6v=e zw^-8BJ>95eHJ<`O1qXO;ZvWP73bev9`4W z80Jpb$hF=#N;)z5>$w=>9+HsgBUD&|@cbV)p{8d>DAk|^{bpNl1^q=>B1t>_@|Dk~ z3^ul#a$Z%uuMmk5PzIGyXG}661ftB`385rd#SJn!C=zSqZEA%Gp<4J1!%4FZ(dW#? z$}O5QUkt;Id<;65`Le^ePx`;jlul#Wi0szk>t0|a==n?q-#vTtb-dc&yyHXO02nqY zxCq`aiLDpzbi08=RSpZLV4?P9X}g8X4t?cJp8QMpl+nr7V}S;49sikz2tw zE~>b3dZTmB&RYD`cq5z=36-OXy)%B4WLJv)1V@5GX+Ofu1x+*X<-Ze+*E-}cKApi03K99!)sgX{%L<0aIadX1)xaV=Co-eT z)?w@dxUiAEpsM@l|o z^?P5unCjh`8N(+#g_~R1E%#|Mg7Hmorj&zN2<)-Tf#WvfkIQ!GF_w?6nK{X+;p8ur z5S8E~+)pykWleGwy&TN!6@W zcWE+V`lW-awn6)MQ@b`s{qQoXTtuV9$$KPC<0Kv5@r)fpNXJR^n?FagpQ{zaC(&S| zaCG*%#4*R3nB-Y*9&vxiEaGKR!44EV3(t*WTP~X5PXOuVP?Z#at%+!x6a`s3CS85%W?ct@`1g}Y$lH3|FB z+Zm_BUt8%>_tI|B<_9XM@^j!H|Oh+34TNtp7*snw^ag>W~Gf zS+4==jd1K>wtrqR;lL5VvV);t|KT_PFR_Qw3xgR~H2}=L7syR=0-Wu?$v-eR_rLnloa|8f zewcXZKV|@~JS#0Hkj4D#8a7}mj{^fKLv}8JR{SR)3yc;=xtICqUI%P|hyJn;ptS?( z#lJjgAb$w(y7?Xn`f&~{QPADK-;?-Fn5XXt@{*6)&c7^s4uA;#m+s5~pdjRv29){CQ$JZ*sd z{cpPXKj~#)lz-C2Ktu*WHS%Cs>He#N{@-q#Kv4eGkN>w@RyT|rH0nE$%01d*{T;}} zk7V}0p80PevvWeTcYw9^n3%@+ui66OJ%ck~VJZm2p#p?*fC~RNOARQG1F3)D=Ql+j z3T^^?{CnwmOaP4i{(Ruz;STd0`gH=xh_i$LC$_-;s9FBUmB5J5V|18sOb+mWqX-a* z&tidAOg=j2KRr2tT>4)}J@$0?1``gom;x*r&A{M7D_vm7ptnOXD8!H25q}&0_uc~g z(@`kwv)>{C`qO}8+V|MvY~@ff!Mxo6s`mp)@wlfN8BnzuU||~4vqIAzU{DFbT>o2) zkAC=P=|Sy@V3MGMH$W6~{9gM|j~gIbz<^x)zwrbtmA|nJ=87|5jRQi97pnCUhWt-n z9?(AcKkR#Uj{lc^4=B(Bd3grt%PttSIJ8lhxM1yyLdeR z!@LK)_viIL%zJk5Z(G3if9U$`;NOOT*N>9~D(J_9YNSRg!zd`8Fc<-S#E^y`l3Z@1~c+GZHHHhOEzwRGmZff&6J751L!= z2fjSG-yK4l?tksfJ+$0=0i6FY(A6^Yu)BwMy%4XjxBB$&8uJ03eH4_uLg<~>$7`F* z2cdVmH}^c45XfGOfS}WZKjilQ5~_EZ+q@@qeX+Cj{^0^rv9a5FaKBqo@o=2kD0J`P zx_AZ1&s!jB9Ncf%-9Q}5w{I38c!Uq!Ue`jb5BCZckR`cn2;|`Ep|$BDLMOlKwpsvk z-FZ3@wWQ~Dr(L%u>hUm8GRL}fA9V=X%fAXM`DJ?vS;*9T_)1^AqENB)@N2dBOMk+U zSMB8z-o_2josJOR{?+7ls_DbfwcnlC&o2!>Qw&d{3B3Hsm!vb*JMDg5+NG?e+7WL( zcTXuX3tL@|pGU0U%v`oBzO*a;l5cxc{)01bLvZEN7w#ij)HG%WViSG0*_rQf`_l4d zI~~r5H^glg(QsjzcXIl!oqln4I3ldl1mVb|k-+Oz$(~I>r|6aJsmdO4qX#{v#xQBr zbmA$+SIxmOAB}c4wnfi@wUTQ@%@h`|#;Ye$VG=YBmkat!N>Q#fo14GZ*%9>UL|NPp zkaESgg8FkAh{hi9-?^+#zS0T7#LeKUGRT;04}uW9t1V7llGzkkPLH)T zcnc!g#9+dXOBTN+kM$v~TA?bwUBL*gJYeZIp#H^R%nyub>fNSvJ}205$jRMpmo{x| zhUUkvt~$w1dW7$d#9_DVE9TT@Qy|#$2jse~UKmq8B;BEjiePp}hQtvm21&nd-`BC7fbgbh*XalO`BI%(G99&@+G(qIW6G+c`%(8Cz16^U{~A4GZfjhLsQ92C z^0nn;2&Ndx#(!!}08{%{*qJZ~F;8N>u_nTt3m6}_qh8FBL~$23*e{a^(X{q`^A?Xi zLO6MOgt_}-udtK!K3Ok&MAxx(&)3;VDW%{%%EivoQiF>w@CY$#-%O_*D^fC779Q(6 zI5gUJ2j;Zk+a}2o22%O6hLMpm zqKe&`^n*QLg>X6F+)^M#MB;O!EKzaoJoBajDHKIN@eajevuf1joB*PlW%FUvXKNPv zZb4#^NmRqj#ImK4*;`Bk;rx~P+)t5(B-oo$jNnPl`js$d-iHOq)EMCO;UThWK_+-R zud->hBum&p9d!g|y{XNfCR`WyqK;TpVbzyJIuYMs$8t2a15G(Kd-O#~kmBI6yZ47) zPdDM83P1BZ#HBAb%dEujZIipE0ZBT}5M&VrG_n1_L(5>w<~AfkKMY2+p*Tp8Fk*)y zj9(Bc8mE$W36(r1qk-H{6u`EK|G@2>fzM;@_HY3?Yc)xJL%yz2d z)TbI1H;n^n!N9?fm1ZjWTEZ7CZ5J&TKlv9c!cmNmyaOyU_qX5_>eW$iPX4~<>VAqJ z7&7=-rsL-+r3evdQM`iJKs?LXV>ozdYK=+a=CrSt$%F;^7N!F?r`tM~Hz6%d+=H={ zWBOeFOUXWTMhu1PU`TNf@;krYBlV2ADQ|Lh;sK@ml73SZdjd#55@ZMCeL zU6rz{b8}lJ#VOQRkd-R9UsA`E2aP9CMBBw;n$#|QYaVFA2xIAF*CMGwBM}WrFwQ&2 zXXr;WDmhM6cGX*!m7^_=R)p}HF>_EWr?d1gM&fTod~(l3;ZE8h4%0|MoB5v8rYNF# zPv|vnA119u^_|@A z<_~fK3<1Q9yLP7=l-72p{ylDwOuv`twvq|mJMUJtB~jGq%Q?$lbl^bhoP;A+o<8lR z@^dHbbJArX9n8uNRLrci_UdutpSW0~>3NO&kszknM{yE9NgFB3hxGM;bRwQ`H$VLl zT4gX+2%^+FiDu#pqkXZr4jxBfR#>_OHt$^ZS8Pwi5BVf$n#$fyeQEk(G-gS38d_XC zd_^&Mp-=9gElMnt!SM$(QH9;4l%j-yy8Yv`FuLshh=RO25>e-?Npz)YDd+WS+o=)LPv5NB3x!o2 zZwp{fYId|#GrmxIWmW2eiqSX9*a9+>=9zl*I{3{s4QaC|kO?e`Aoe(^VxceGCrH<7 zWL91mBT|cLKxm&R#ydqV8*NMa?3CjWJ6Gpl2t_1B3(gaMT{d;qX|~Z=a6X%Mx{nt`w6kVSR8c5 zF6~slU?TVWtEaU-G_-6TY%TWdqLweMEa>9tK)jk|M*|E|iV*jf-Ma**jpLGWo{f78 zjqf)<2{I5B`Pu6?G}4oXGfKNO5tNt%KNhQBrqBxs_5>FD57M&LC)V*;4Zrja35UDtC}?LZvW09i=IKZ z2E)b*8W#Q8grE%hSn8!F3b$jgQ^_Cl>OES_CCR;|&mp{EH|Y89GH>RlO7RK_;i)IP zxa~3g5sp5dO@!4ZC2iZTtvqMtJV!W^i}3q;5$dw?AEFuGm%zhz`b5cviAO zBE7TukyD|R$?r!`1;g8eHhc|3puOm=;Aa)@FB(_uovfK{rid-)*0DZ+xe{OHG1-(1 zU%MV6z8S)8Xx!%lmBe7RlOc0KB4h}*FGx;3zq;2#5%yNu%MF}TWj{zZtc=Ujr;AnI zrl5wK4Vl|PHp|yYDmVR8tqaZTSe&XVhmF;PMl~l&S2Em(NhpyAo{~|3$73xb26#E`^Bp|P1Sa0-&d3*W^GH3L>1O+wlj;@1CJ57B#~Q~jvyrfh z2;(a5Thz_S{=J?VLK!}YAVOuIB6j-T6Pd2=P_b0V$$3#4d!FLae!TXR@Q5M((ScpR zY!Cu_Cm~a%R?T}kI5LiDjB!8pcbvM(0;e{G7!Roep7#rN$i6|>1+!{^MC0jXvRo%= zzEWBNSkTBr{O-c`+df%?QzEP2)!YEjhz!E4?*d-F4u#c+s32>z)x5ChBR|*r%Aau- zBe!-#a!i#O)Zsc-V@T1+)iY&kfO#?#!H=Pn0rTX; zwFvQ(wxKC1*e9d1L8WY8`D5ci1Nm*O4CrJpv7y>BIcHTm`zajRay>DM%*W{|l0xXo z#zo;(F1Z^54<1#LWzTwVy6j14R^uro6vKriNnicUURBeyXWf;JHa;OiTj6&teWlOs zkG9GxmxoFZG3lt5B|W zzL=qqU$#E!NmoSz$@nj{KKx*6t<^x5NbZj#S+pSB_@n&;#*Z%7S097z-o%z9`NRc@ zWw@Etp}!%UIYScdiRj_){UpGGq6zxM0GCdJ<5>qODR7>L_?~@x#@|`4U)b0FOMCL& zs$NOOIrc(EcRUqM8rP*T8M7G&O_O&6+ea8%V?u7T1_LhL`wu8|S;VHO*z<^mB|E`} zEa9SE%Iit{4)CaQ>FqQ`Z!-lkI+wqq-J(aevoFR8P#9HqL4!hl9757KM#d|z+p(}> zQ!9BO8xGR0)w9ltntWQ?Cr5GN#k9Gb;WgOWEBUu-aA%Z#j-6XkeBi3$R~ET>;reFm za(O#Z2OVFKjpz^4i&V;1+lDDk_Sgz6%ajxf``qi-&}KcM8%gv!ayn^9sZ0AFai<(t z^m<%UGD0qJ)YIlZuVtEkXkz%pD=0YVAe8{BXzqQ9PhXctuHRa;JIioeEl?LRRcEVO z2me6iRGF<#CuY3PX|Yio%$I3vQRCAYbD8<@?8CD_{vcB4{46-5 zBgonk=!nRI>MnOenMU~Q(+|9!-7ZyquP3_3?T3!!I=|Dpp<~G`f{P{xyF=H71emAw zbUeGf7mcplTo-(-v@AWx_DgEuXF>V`>hJ<=&{(WzZ!&}#N&HvEIa0Bbn!nT`w@3V{ zdnK>@(?jWFzam=_eG$yAQQKg=1R4|1!X(54n?Zx?bCDz!)j*fv;YtIqz{9Y)ZJ0houB>&ub{)DpdzWr5A%V)?< z4sn(k35@U`Ug?l+hKwcvcvJZm5FEWE2R|P7x&io)uE_)X`3QVH%bvTt0sN@DmS0UI0oF ziVJ+TU3+xw^lg1FZ_@r#CyQmmSsv@K`$P?S$wK}5Y>=gqw^ZP!f3Lx?d8u-4(YDLT z@(^tdJ&Y#N&6{dHZALWQ%x`S?S_3naDQUAvDqd6g5;MW7&6(eezb}^~60yfaboweZ zW!X7ryg#cPcGs9E7_$@mT37w##R?%M&=rd9euvSU%Yk)(id#IMf(~zQtdBr4oK7wF z)+kiQYDx2eH4bVPy|@EcSCyr0Q_*r0Ic}+eeE1ShWD9mr!1hc0kxyF#iD?SkoUs0a z#^<6An({4}&*#F+S7!^IC95DLh&2kZ;e`q>^9JmYAY#IEOK_0JW=pFE$@-f_ik7bK z_Vo|Zy?OUVlQEEGS$nwqZ50k7X3P}Mxb``qS|R>cpG8nQ;DN z;j##dWIyXA4RWCy>H7-odam-rz$V5m=Apd@C$#{8bt=4fKXWz@0d>#B`l!BDkLB%9zx zJC9Q`$P^d(;sv)zGF{9`U*aNsoUDv<^QC6QcjE&7G7%#-l%ysh$Sr4&iSaj6ZTgHD znm{BhxIUU)tGjS?EzK&lCUA=I-PTSnfB&VNIr9Z3Y#9Qm9>OGI^#Ij zudp8j#6T3hQu#sx1SSq^iNohwq(c*~A?I(UVcYuq9qawlzP?qlvswKSHtUv-Rk1g( z4db5EA0MzO31@BOA=l!=Xy9 z6BOq>N&KY=8R;t2hvBs`#1kB`jiNhS7F%ITpj?_&l!s!6Vv-MIaIW+wN+enhVy0E5KFE@ZS^r8_v^CADb)pFm-|$i-@_m~ zO;yjOhB*m%>eX|Ls(`$DV;p8=86=7JiCAn9MOd(-#j0M{DkA0DLZFW}l3&%&sC)>| zchIGXGL%>RI)0SJ{E)E`Dh$&ZD~gNRu{b$fJyIz)Wq?S#zNfv-BgN#U18X~TXRz0Z z7N!ffEzbTKCjP|IN`b>tfZL43ijrC9gvmwTq{aNBK1b8JW{bMuBvXwCSPU!~sM`k` z4tG^sm;1yd@Y=PF;PtyXL0WmM@D@n%biqg(%Zv2Tn7eQ*)-sTEOm;GHeih>m-G`k_ zLF6KVT2=ZYFjk3$b7?;=aYTD?ZQ4he?iHmovvMQ@Qyp)Te7DFt1kq)Wz_TE3H_9h+ zT7j9bkEHF5_N4)J?QTX#0@M!fECRF*TCu(u-GvY?E`y zqZR{nY{+?mrj_s~#%bO7^vjm-;FE9`7&`3VdI#O2^?_(F8mJif}p82Vew1+YA2^}#kOk5kE- z`W8k6AsU(X%Q#EvU?jO5WIqisTG?5Ep5G@HLYY=m&fBfd`;iHoshpAgU!`s4@>`_$7fqtv;%1w7J=f%9Mm(Yu2anIX^rcV( zYor2ybmB?$_7IAl8HFcsVY2el+e2iu`!u*49Jy6QIdSB6=QZ6-3sZ89=il_NK?<&Wv&esj3Aqk!s zL$~L2%KRi^=+F=ITm{KMG9gt7kD`j&57ATtyH_8Xn#1+ruI7unTCaYkN)hBukMlvj zNF&0Bij+-)RP?zC|LCz9iNJo&t7BP(8-Ew=w`1}=97cL>BVU2aZ!65WF4_zH3-;!- zxnK5*4^I5>4o-mQr#Ad0^zDAWm~ue>1?&nj+)^7wF9IR%PmDLLS>cwFjZzD1Gw;=Z z$rit^&7seq#!X=(1uz-fdoSLyNu|~qAmUC-sn}D^l^;u`SJ3F#Xg^_FW6WaXL8dm#QZ_8Y|Wy`vYZhzoG zW6xm}ARP;ZM?88%ro%ua>@l3@lL7G>Tfpj-aD#0zy;i?gS+L$w&B7`q zWpERrl2k5XKUd*_N$?h>H_gc<3FW7O&QG_WXhbQ6GTM3Uh8(iz=bC#XW>l64a_N~^ zE~3FHzWC48`YtPb(YC8UnKhrW%Sgp!^uA``AH&;%4>5bu{~`xoK|PkVPqE)fM8)Y8 zLTyC?-YOUA`z5pHP}!o!sX9)jH^6{xO@rFO#N{Qng)Nn>6m`;rZQD~#gQ2#Ml=?)w zE}Tmix$UmBVj{il;^bVWE_*K1PJ@j6!+m~GYfR+ zuQ;gph38KpT;n*-$}4H>u4U7!WcDFmNMPU|irA`AX|*ugtbP5LAsw*l7o9rzbjJ5o zqvOZ*W5>}1uKh8a6ai(2q2v_h^xf2=GB%5;DKg?GprX(b==Hbu~S_)zoRB2DqcI))ByU4|Om zE#1~~4kAgCwIB}dR0lz{$_8KiZ(Yva&6G{1_;?g11*V~YGlk{Coj>IFfpMdxZ&8Y%Ud zi*~|(s3J3mY*X__@?-F)p=jeTY#>TMo~&PV@MtY;1!M$G;MXwfRLFficv1!3*N86k zBm5B1^Wh0*Mc1@I>va>=6E;fZxG;Nu0^R75&9#oNZW)qCvqcWe#PwW>m$gL9-=8N3 zbhR4f`x6GZrT7E{D?ru}&7T`g28}Q}kxf;v%keNzG4$YcIvEqtIac}ceyh1ETXD6D z`f0e##oV}5&0l32^)_-c)=Z}y_o5RCE6!a6jZ)no?z5{4ZLh@(t(5lds@5RnTUaJz zGk>%hB)-@Cg5=$n8clEdu0_b7_71}H?p-9dp?8L z)h`!6h4vxbh$pdV4gn;&R8H~S{`4KbyyJxqY`{{>y!z0DVUTli8LA&0Y`okYH4j$g$W;BGA zz_-k?)J$ldn-c_{c-uY6fLXzHeMW<}yj<30347n;#DAQ+s=G$$A`(@1FrUfl3a4fi(Xa04PZR+nC)J0Ad0v zO+3CJ;sE6T+5Z3Y;QBpFU>*Q);jjAse~0>A9{_r&28)XM-L=P5^*X{vsbZfcgw$d$5WLIIuKOo9RC@ z!es}bTEFc)qDa7x2i~|W#%Ivz7Fhm-c~~mo&HaX-JOC4V1_&Sg#jXIJVEJ7-;Sulz z{0{zSKBR(hh|p9RSdq9fRaWSM8Z3Yx0VERs0(XES1^;>kfM)kMQ*4B94sX?;|G8%vlN7*r2}9U|AsS| z2MC|PM|*s1u=vnWYFGd<$Ng9#0y;_!i$)BfY5rX%?7&?I?(lCY7?>UE#R#heeN6*P z@fh&GOG7+*g&2T*as9Oh=HdCbHxPJZwXhmcJX%;nC{545(h=PMe~8LIg&zJV9l`ww zQK5(Beyk7S{RmJ&W#|4EpaNx^g#l}^bNvC1aQ^|0aQ^|0aQ^|0aQ^|0aQ^|0aQ|Q6 z$nW>@e}N;9x*d=8B6KR|>#)6qS`9Pdj91SvBEJ8|Y15X>9elpFR+R>@7BX+#eoTU2Y9*6Funwt zx$^>vx>$d?K8)%~nbpuhGy3&CL(hv)+w=b9>>%zOowDWjqnY+eEg*IA9<&;jU%7bD za@%w>?vr!~d1$&h*;T2n_&GSYDz$$w#Nlql?6DSUH{bHjQw>AAB+Bl@^B`plYIlS* zJ5`KNYBqT+SM{MMGNpI?3MpuPjAU&E>1bvdsoIcJWFE}Xz8642_)bUQ+~XX@lpG8A zPnO%awS5nvwLRQCW4Z!iUCT#8C^EGdwfBfT!vVPf;z z7xPOr3Eve8mH6@`xg%I*Fr`7;M_*W$#Xw^fdwFAercY%ocqeEpfXF|oiajCtS~h7~ zA3u<&D#0>;#;mqcn+a(X`Q04Y4=Fy~T0@e`(1ml%A}K5UTTmG|KZQaBX~iwT`ub{P z8D6{=(#nK5ok_VPo>c~gVuTCNLPqjV+8SK; z<&}SPWGyMTmxEAvU{^Trk5{|Np$_)&+-=N_r% zT^5nX?rc@%XFhR@L+!mJ806mkMwdZ#Fjyb<-ci`$NwLh_Xom3HNz(N6wlk<_PkZbmm6F=ro;5^S{! z*sjo?rz7WTR|ZvK=IkmIrd_bJkT@|DR6_XSG&v_JK@RHa%<7_-uJ5Vrj}19~r7Pvb zKmSREeMw|NP#uLNk4-=%uXs7iM$r`LtF1ah?ZV?}=0UG{ifGi*)6>LygE6<9j<< zo09&sRFvqcZD)dED?CWvEB?l9 zYT)jG;)DF8K`UM>Lzd`(WW&4&H9g%Tl35%6#DFVt5Mv_+^Sb)&r_vvgQn1sjwIQcd>^CF zeDimMg$_caIH+LmHl-cn63YgLD(B8U?+vm=_pZUG>NKC|*NEhY>`WmKINu{;5Gefu zZWyaOtC}&g;u$_qzLA16G?Kd%ScKWlYoGgZJEKRaD%&RCiR14>nk45NBBbTgW8S6O zfsmz5kz(FSI8-(DoytgzL4UxYp}F*nZ{?Mb6+c%sXv#o69R~LZDR@d9#0H}2LV@(T zh7W%ZIDF~+VmGPuGIbiljJg(6d}Wd32951axbjKXXMfS_UmzbGr4%$z7{vE^KQ0Dp{RZxzm|iN+N%(Gf=7bDq1FAxKk{sNmXaxTpqzU0Z!u@Wh)esw*@0^E2&IzCZC zig}Q&rw2lDV5Xegw=;t$Sc1qMAQaqJB)1YlT70|N{w<&Pz6ULZ4NB_F9WYW%l-j`~ zMDA?w%2v$~*?R0pdt2$Z%^kqwX+gZ=js3=cQ`~r!-6(k=t)T_Sp~$M5joLJvXwns~ z%+>0!I$PY{RGx~LXSNdH$A^sG{xj|wPCI1E38!5>7Xt&c2PgabpGDXOMJkUvaU2;P z?0VLw`B76{-y=;)5{zj_)DR=J84#N$Qr|FQMffpqM?H{(~pPR*t z3X79QLw{@w{oZr)g6#2Qae4SdX3f}UwWnUi6qATX%Np{CD>)(Y50MUC0>)UoSkpx3 z*W1?<)OoIs9AZYj`dNzd97rWdNN@>QA6(F-ur#MXwY4Hmo04!kw;;A@5N1g8V(3cC zIXN})kEAs%*65Q(yiG~`hz9G;c(vfJx&y6UL-fk5z?|B4KNir@&3V|Eu2S3o6ZZtK zKP-9H(nDNME7unuqB9yogR)TWCE@2CVbWYZU_8=(0GoJ6mgCU$rF;cWksX@-A*II0v0PLNiXzr5HZ$j7IZ*~4rxzZ&>5a~vXT z!jLsv3LYT^A^SS1Ir=oeMO^8w?8fs7zwzYnVhH?+>KBwtQTre*m7obs#Iinncx&m! zlTYF9m^N#ey)FT>L!2+`7LA_7@GXLl%u_n4&#dI0W$rm(0G~zHE~N)q+L|9y#S}`t zEiwc{#G^`I8edDhiJuvguFxrt?pmlGe|hxLFdtGbZHV0?6hjrS6l?H$A-0A5P2x~Y z)xh$w9cEGecV~2tRNnaY^mU<-Cn>Dg)f`tbdP_(_Bg^gYEqc{RHM=S|FG}#Ca%Z?a zjRDKG`1yjCol2bQ#~M%MqF_8k zF{ogsCCM@rpTRhjbUYrBEMP0@KVFZ-+MKc1K0qujEzcCt&QE2?SaAy?oe|a#3P(j_ z4Ou8LjiFuIo6&P%$#F9SOB6sU_9<|4NIt*B?XS`(ZTSAxri9Sln5OkKBDNc!Zg6y?FXSi?tCn{8@BE9oAnd)az}TBJh1 z>Ki8tWaCNBwaVuy`1EOB$!$7Y(q*r*W=AVO)+_2J<(0f7W5zJdkC3TD4l>LCzo3dJ5&r&;9`}vR_S{mq<;7VN^jir-zv!W`ryka$7O~I*a_mzh2 zA$AeqXj>5F^5#)%n{mw=bX-?Lzw)@~r5|rrti#%~H{h016N|k0M@|Kj->Wb^*Sv5F zylrasyw&Iy&RSSP%LHn))8hvxv(a#LS+}z7;yaV1Ikr(n@V5t)u`7HL#zb73=Vi;< zJ4BXgo1(fN#;0|fJr{c?+4|a%A)~dhS_!!t9Q`FT9&&}x?y$b;0eya(u)UrS_8HFa z7;c?P*+4|#B1H>)k*>|t0wm83CGeKG4a|GBJ=81qSS7a}26qU6m~G1Dq|yWl$H!=Dd{{my>J%bmT@B4K_^D*B?ti%b2igItaOkC$y2|vL zCo=B!m{Vse8*&r;$EGe%Nd& zr0+JbCnwDEngOv*v%NUIW#?(BzqHf!9Z!xUqEp$P4mizo84P)?pv^cns@8|bI>THa zLtVV%gF^vHu_1&Ky;Aa>j!8UuYI*p$)&o<)ySIl-8pRK!CRNRW$8wg)_=+^Tw0*ih zx5aq>yRtpApH?Hjwsbe1ZOzCR`0hZ1_r>z()K1M#;qeBRkNT1(c6wasz;$TuW|2vY zsg8F=qMvo!5iccYJ|Xb5k*Z{9o1e1fe6;GfNOEuW(1aX8D`K!>(LU5HToGW3cZ2$~ z77BAPLj~%oIm&9EHdXZ8c2ReRn?$C+msz5FP8`B^V>d~ZMtfLi;Ee&YOgQBF5jm zbw9hCvj1Q<5uTqJ|NV?y6i>`cQ-bA;Q0-ZnPE89zVJOHTWLN=gVJKjy#e@3d|TOwqJHyB zRPb5)r-9uRcZU2_nYv>qOOU4v?M+djYUmzH%WJ6pAvpVuW4gp)*>mMYUEkEt78Z;c zZwX(mV#cMgp1yG}M65s#uofZQ=^}a-zn->eA~_H0vk&93n9*g`s2@o$^&m*`{4f{3 zTc#_B@5a7T)h3g^7*=k=6Ze_eO6%onS7E;vhM_cip3X~mmx%%rIK1^S?_2@q3BveC z`P5a=nt68vxh)qD%OCQ2w$fVYTMs;|N}Z(??mh_DX7XR2+}=}ZMS|3IRCq*>Jq-9GB*DRn86!{}+r<10LLD5oZLL6>PA z!eRPh9GK$?!(pMK33hQ6Y$|}j%lLH3s103M39Ff+K9atm&F8`8KMCWLUJ;L44g;C< zj36iyJ4eg&)VYbuAtpVY@Qk>gI#lQ-);qXOX=wAW3B8`ewA7eW@PS78Ctk$-z{>RC7_y`&eq_^da`2#%OlIj?um;GnZk zyjE$CFW;9Xy!J@=$QT%$cs`PfNs4R8#TiKgpBx)lTAhGR_ra_;4a+?1*#9IC*hi$g zIZR8|y(wD1*-Sq4!|wuRM@e*v25K)dm~?nQq~?Kt60GcPTCER!Y+)T)8Tt6U&T1xJ zd}#AoN=xXX*MpRDUHu6E!tEklWlQLB;%$f}OLUq`n%~&;&z{r);GVM}i|OZR7W5SR z5Vye6l6>%ns>;EiWmQbdim{?1Uupcp>ix_#8#Si(g!dNG5)y@%{d#DnUxED@qWu*A zIAL?p1v&GxL{@fYR#sEb2XmM`@ASQSNJHSx-=?KgO)t`G5@seim`bv;KMeq!B#dgK z)zi^kSPi0gm1N5*9 z!U8{RL^r>^-O;G$%}9w)m9LPtbz@RNw1Lithaa1Z71G9&+#Y#$7vK(DI#k6S6_Fn6 z1)pRzC|)A5Z@g|=A74%F?cU9rL+wX$Pm()*OPEUd>9|49^|7Re#~z&m)a5fHi&oJ) zT$KLI{H}lyp$D~zKt|WtY%zX1%Bla6MKXT+n@~5aDtX{wHdi3MMnZ}99Vx4oI2&d^ zE(;B@nHj8Xk(Fv5l`ysKEc}wEL|df>?ycx zfv+nY6tpw5*`Z%dJcuZv7Ssa1eAVGCu%YMgI~)S9V}4=)fHdAPYk`I+t5+>TJV`vM zvH2%!UFGN8Ppi?WMkK)y;>2Yzkw9yBb7o zvE7eIKRgt2Jkj!0?-IC)uuOrMpcfAvGSCVo3@t~Ds9me}4 z-r@5+ek+twsg-#gS=aqi9r*n*Ks{7We>+N{h%eS+cGxO6ou}0`xCgfcNiQNxxz1|( zS)xpILP-^)e{WZ>{3kFNAEUUEHzL<1FvravLMgbw#djKtvqVp zu$mUlWh}W(HE8W!V5?_3KZoA3-`jFwYTn>XyJbm5%_jXQOLeDE9+)lSH;FLT_ov_3 zxMDs;=Pq2f(6;Xy+9jSMU6#Bi)(Q{Wd_D z)YLAi`Hp`^0k0{0;8Vq-TM>o(&Gu0w=q7w2n2p>^T^1^pYl7q&)T|R+D|=KjrxOXz zRw-~p_XYQJM7ljF@<^tlU~;Va$PEAP!!jC$_x5o&^Lz=UcQhN7g43kEDc}15-72a@ zr%ySpJ-jj2#hz)VMuIY@1Vf0pu&i;EcI|=!R8(IE-_$bGhsaBnKp(T!;GoDu@R-%f z-gXK1XbyUg_Syw!S^_;GvWSVmr&5buU!=z0lhx$>m@>Br4WNU8I6Qa}R&-lD7P2O@ zv3lo1&BBlv*EdGDu1z9?pU4tclJ_?TwqH#xuX^zmjfYJuYdUb!oUwj>LzCAU_Eea7 zE80Cra^gcjlu)vevFumb;&3Oz2HUjB^N&{x@yV4_SKLfUmU55=fp4SvhQ1FI8s5#P(=UbZql&!H+(Dm% z;hn+VhJ9f-2=}Jy?iT>^%**$`oc#YFwn8ws3wRn3 z0O848!4ts)iWER)GC{xy9`Z*X;y;yz0fmeE(}1GDKgsz2o(4o2?oadb02t^Wfc3wp z`F@>Fo`eUcB@FphfN|4-NbX0!ZwR}(fCmZ!zYCu$K>+wIEcmNL`d{q#zkLG25nM2r z?;uQkApny6HRwHpDlEVSj6lH%GJ;JqgBV~o2p}G62%wMgH;G<|2gv;YgDUb1-iyvB zk25jCCfEQ>@(l}!1;i%|X!FB!F+%|9`@ev`AfR6%@UPK-b@r2S4Jdy%c47rF!t=qj zwm>{ofQZFkuwM|ctN%MTgy-J69=1>m%udo6gbtHo15p7=4ih;%uz0q6F!O&1F9cv4 z{MGyigbIEkyuc(4VXjCZ^kh~p7MKwjgh4C-sD=F9az4JlSSZZpEl3YG4hE6I;vpbH zm~=IWl}Z2zguikbfP(%%jR^q=0Pn&Q-@}vLFGK`38UrBW+3X-fU}iP>+{qeTfT}+r zL-0#3;U3rqFx|iCwEz!*^}@7i>$B@QVP73HM+yaFu_HmDCa;JcLOZgRo$m zoFFowgL^Fz7#9}^1D_v|TKKC}0pZ~R28F?MfF8jLxjXc2 zAArgK6$$yJ#qo=2AP6&$1L45-ctHS5d>;_ISdl>hG(33!cT@8Nk~DuqlMe#J3kN`T zAQZ7-4SXPSctG;Rfcq5$goy+^ZvZ+89t2pg$i0_@)_pV~!Tf+|0`wt*e1Qf1A|>zw zx6ohD3|@c&^LI;m0fUHpmc;L7*7^V|0brZlM@0k?AdJKN-+Pu9aEAC5TmU5`8E2RZ zhT9Gjg)tcXY6AoXH`~3=#-FDWz=aX~d$&V?mVY+~2vA@@V2K7GTj1HplU#!a+`>G+ zv?2cM69L?U_eu}{z%7xl%xES=EX#-E)(S3Qqzs;}>^zMepD zc5dUPXjbR$LJuB}a&e#C9XIxNZa$m!zP|gwbz2F&ka`G6Qj9}aadQ1$xZd4KZP}#H zjEaTvzA$W)xUpOuJq=^S@aYi9-p16oNA{!zc9(^O|Hme+1i<7d3HAW?pY)B;_Qw2u20?Z zsmf37?Y={2g%=lNOXkk&<~=VfCF9M^Z-eut+BSz9+h^NcyQ7fxovQ771t5$Ky{Y!Q zDskiqP#SrIj%mHAhga(=-vu1Cr`o`Z$^A26(G2NJfm;>5JU63eBZ+{%#+eKGiLINt zqp)4K?{y1)EKd}yjVM=u&)4SqVf>#~hq`;j%+q<}J1B9^F!p7a9fx@PT!&9C2t@BT z*_0*oi>?ON*Ckf-ULBzapi_E1Fl7mQ%uuk|0)2p`AkRit-aUui8Ef!HA;xdQa4C|O zL9p&?WVS&tNB^q`4`(%bT>j6m_@69t(5!sZQ}Vz>09IA-IAi=WUjNC`GemzZw3Tc3R#P$4 zg^SN8xUQe;UQDG1@wI&$+PEva80>rdECY!f;$6YImzzWu9UMKLCq%qD*V>1da_av! zVq7{a=PUmZcNYEG3*(?7)`P(fwbii?i5fdl9j_XL5ABi-3&USe&wrKZ#*~*60Y4cZC88R6ZdM6=!rrz zTWn*k_t*)VWx9d9xc3*6_65;iv{^fE^xz8L=?YR`#=0g#dHF<&UdlcaFI&WBvIv?D z1Sp5rS^K3J2i4PYx&5_zt??(UF{W&ku0zh1JSzS4PYU>u$ATP~;YHA;`;65!-};ld z$|&VX_iNZ*%oBA<22-}e8cLsZ#uktH#ru#(e}%+KYpa{7D%gwl>$*M`W z3yVFkc^)n3+euC?l3rR@J5Q3%$x<02~3H8@fa-^scT6 zym|P%V#t8dkH)`gX&*Nu-ij{sKzhCdgtp_2K>Min!7Q^w!niD+^A>HHOXYq zM^~~X$B9W?<8Zpq#BNkxfOV=nWTogjOil}xmJaOpT(N6IC!p|g_~9W>HTEFrmWYk*N;AeagC}v~ zAS8gni-kVdxqKLBmKOht4)frZyu7SVXd_PEGX7*3E-J%wWH>x{QgFm6_m5t2#{Lna zrwVyzcytyhl^5jP{x(JZ^Mnq_N?lseW^o+_uG& znNJ#0n7+XO&J~crQMBHqdHSG%9p#o(XKbip^Q)rviyIA2CqsQ}>+97T@)MZ>C`3T9 zYmGhFJ(m=HXg-GF%^QJbj5d#oqbM^96kO`@u(g-2XMq@{ngL zNOy(`dy1niQpoc4`Lo(JZ4i5I6Zb{>ZZ;1>C|e6hgZ3%=?3_Frx-Y6Z#jFxVYhLW(1AWi5_FI-7fSerl=g_ zcm<}1PONXAdWo`6Wiwdmrj#Htbw$Z-m>jE$H0TYo>*W>64#M2? z^?Qs`BJ6M8#8Rj3ek>xR7X3Pvk8B@N3;|xIxDs;umo7B$Y=@Q!paZ^8im~Yrmg^&m z%M+?$2AZN;gioek8=1~#v5i%=-dGSty`>=g8Ob_U1@1Lo6nHk>sDn}5!>Y(r$M(Bcqu#%yM}20^dF=TOjUHP zF)~7XI`c_KhF>KbXV-5>LrsUg^@3Z;>>^60AyP1v!PtbPdit`!+IbTKqg4MWPIkN5 z#j9DA6V`MVqBk4H$|{uUn~m%_2n{}Jf28AMqOFMzG3?h0W3YUu~5@?rbETPTJ zr{qNZsr)8%bC1XcHniH%LHJ^9DP-nfzXu2>=8=9c)fVB9&1jJAfa;O8cz*Pc-;kH( zQq$#{AFzx`sSuJI-Pf6wom@HMSoi3ga~f1}SlL8aDN61Ht0_GaUg^pt{sP|HcmZM0 z*CI=j=B4O-87#hjs4#(2HRtoQ-!lk|HTpl;Yf& zV*DU{(<4qg<7ul!HCJnU~r zqO9^McyuEfr$95mD!r?F>ajb{TBSDY%U;!`L{vNdob}@B^HVYWXKx!>J&}mhIyHF7 zxY3Ps-;rx)s=@O`r-=?Ikk-eCfJzh>$OV}igmLJ`@+-31LaqVJR+pr!SmF| z_Wt`sr-5 zW;|XySxGaO@-TGmkk>&M$AVSZ)hTVI@8#ZFOV~p>zZkQr(XqN z78McNSt1H;3?DOJV|59 zIQ)?I+<9wEJ%m3CM|=+afMPRGn(LW__JVPk3FZKSQVwH0YwVW`F{NT*WS?DwLi3d4 z5HUiw7q1%{ab~-yqUHphNiveqeM)=|+=P<~Rnd=}F$~B(Tq^qId;L%R@L@%AAXaz> z*ti^s7@80aN2p5Mhc9A+5mm!BoQNug>F}7;3D2Dt)Wyj3tuA86TChKc=y4dq7X1WI zyJr77m;5=`C=^p!)b5Nko&00$ES+A=%|~$a(~q8^EZF*6_#u=u0X*1MS}CfueBlot z@;TGvt7y-FPL$EL=sy5B{YpNQ{KXj;M3I)T4=PSI4=GA4!!GbF7ov+C{vd*SR^<=pm@UVb!z>4}Q(>(1~*qTX(0&Gg{we63!hbHeRzR^!wH z9?RF+*53qe%a*a2Jci$=5Mqpco@4Yy8%5thd-#cHIMphMh&VQSG&HTz5u(scWrIt^ z)CTq2C#n}t_O|i4>;^wJ6%l=ipXhyg>+{lav2TDO!W3_H!&N#Pmq-J~q;1~wbWtHFouwE%h~u%ieyR6hNMK+NxvUZ8`iMFE zck`0|W85cZf(G^OgGTzNhb)cr?rcwQ;50fT-G4Sh!93o5oCWVM@mf?$x~@c;5GMhr2Cj{4iwhiKYkUtzRY3( z4bE@+TBV()e+Ws@qA*PhMRn-!%#0>OHC-g~KmQ0Pln=sNENI&iPG~0F++J?77AWXz z>6-v zO0+@~Xa_RVj!gQIy59zfxlR(|jGGYUSQDTYh8y_Si^|nt)+Fd5dZ7kW^aFPp=r#&r zVcFOHOowsNO>J7}X;YRDL^3jd*TPnkifb@J*HR0gqx_VGgOmUaR#3+ zn5)B-hOY4ZRonH$He%gG{3%dY$-K*+d7`f-l2;rapdPV@dHKE&CLtT>cP)1>N2z{A zx|hES{}b6xV_M0;28aNe7eC}k7eCpT#Q)qwRH!KMqhLA2KF~nIf4nT+1=T3vdWMc! z?S&a*hZ}*~NVqYGQB<(SA?Dprp81m3ea|LVCmxPTBoESywr(-EV}F6F_0Di_DI$ul zPZ2!dXjR9~DNIoKv9CZAVmok~l*HaF`RaguT`9gKsi;cR(m2JF@Zd*|)>!GJli!ZD zMZfuy6lQ@eL#yz{6a()o4A>)O5Ek^A#LEuVF|Kz9rhJk0mEn7-w0e!7u%LAo!#?i9 z;;Z$*-R>s5$B<8ksmSiS6!-wCiKC^KxeD8akN*i_$wuE?>8vAsV&-}bard<3Mt~9e z7=>y;oXpX4@7?-b(Olmd_9&lQ3oOEF(H!68uTXcwY6)(P*6&qJ7kk!~jqA`Kc40#W z;ISPHXeABycZx-)l=nJitE{C1ILJz5_KhF1$RBEVYZ-?<7KJ|g0_XpbswgQ}pyP2g zKF=-&?^RiPk`b8ldzb@w67Nbz)cOnFd}h)t1zz~UBAX=HMh>EX@t_8mPqD-N6ew;; zZ~Xy+r^v-F?gkU@Xai-G$5(S`(x(S^N&an5*>HQG7H~BQ#K{~+(}$hzl4|ZzyFX4z z=pQxA`ytlxY>syVj2+eNnT zgct6=TV(oWPG17qy#jEv0jRyLOjuoyh}NiCI{f3OH$vcv0_r$;EH*Paz=j;*Ap$3r ze&ReSvihpPn+c&-scmV&68Oy$L)qXb9&iO=G7tV@>aw#g3{MqA2<_8f^*BH=@cp`$ z4}8)OMKzYIonTq=vaC6ikyU!};8h4^PoMhf-twBI)hlsD&x+Qv_YVF_iOpS6Din^v zo#TxRMFR;*W$=~AB&!mpO-DhlTb>&FTg6mki&Z457!dgWPG1$}7>Q=B9h|6O zGHJ0s&e0||r<4$~P5TwbAP|KoWk|WAAK)?T-Xi+#kX1cEZm9L0fcX<=!#vfMB6bRj zmmkr?61ArpY+r+4c)nn_y1EjN$~#$!F0t7ozqT_euWFLZu&RWb5i8JtP;h83)(!rA zdALM>6D4mu$l2ifghbxbVViNzUMY^O{i*=b{3JGaq`)JS(H3Hpy+e?fs6al$*dzo)@Qfp zRxmC95yyV&2wD+Tr^UijzgYXhJSqG+8XkRcqlHts_p+g-K2E=W)I$m;nr*Q@UN#xk zXvC-13u!GfyV8*o>lf7fmjn8UB=bb*wCu%iG}ZD~?S4uwy|HQ^Hub#hCa+!jiHrPN zN~^->)3nSQ#c7^S-}CF|IUP*Q^Uny6H3~1KFWyHj;~Ha8K(VQn))}K%*Lm1WMd1wC zbnD(dLwx?&Z26#RHrtdlffu>MhUyVUVX_81tV9`x&?_{N;ONb+$Q}1Q!+vuqS1+>J zyEgO1OS|o_t1X3Cbx-Y1oa?dbTDuSW9bur(VWmBM2e!AHH_@NmcuEgjJ!5#XF;z+Y zN^i>eKw}x_-=SW0Yd+7!bkk0cS}W6n?dK#Mc##BmNXN0L2(;jq0$6@fD1aL84fVj8qr8&MifP20uMKzMtDk%%$c zSY7Xx$be~Hg-Kb#?$WST!aj=3d)|mD$Kw^kq3B($LB(+Ga1qPZKh*Yt#eV>5cSc75o#NFX(b{#fSfZM@) z$A?X9pBdLEu2jgh!VWUx-nq2}rRRAPoqm)}n<>^LOX@NvzyIDv-|+W)7rB(>o2^w| z-7tz0OOu=+2-I}3z%ja3tN5KdA>Ss`wF1Xtg3ES=`)5bkw|5>2RqTJ zO|KsDsb*Kji#ollk;M?Y)J&f%4vl3g;Zf7#^COBOX%^JMlo!a5LS5w@ld#=5^ohAa z98388G#{isY;A2c*EI9;h*1l0pfG_i}qp)Lr;PyZ%%e=@X;)@!;piWBsbH+ zW7{Vwv$>^8&Y3A(cSd} zRy6xbH6b(G4V#hFR#UU0(xJ7)x+A{$fnjB$Ld0cG)|aFm1#=^2yyXPulc_Ipo_8>Q zk`!qY+Ob17L$XcA{2u#|>zkubz21wY(EtmQyLRMMT|GIpLaU^f8>-hT&5!~OPC5T6 zG@_n@SMvUjpGD>LDqE;iV+BeTGJNqr#*kNxG>HxWt`<3`{^ z2R`tUuOp7dIGNfybqrx#R}EZdVXSLj&KPz!Y{0{cL^5-DuE6dNO>AqZ`bq5ym*$($ zq`WGa86QPi;4pP3nq~h&b8-?oQlNx@U=w8wx z_g%aB-bHTJPUg$O?xC+&z*`+H{f(-?6GlfCjv1L9(N8WW1g-5EC^i)>Q#dXJzc49i=#*W!KBk|%?ZULBHENBqK7M=Jn;QP$af2h# zxZirm3I8^MpK+{_V^K$`a;f$tI`#mGN1AmG5|g@#;U+4{fnf#*wxi|*_9`lz<+M&N zYR=1Pcn+d(-Cy1fomsYTXr0hMR4bZT=zJH{Haz~*_yw=>Dm09dBKF#yH>}MmEaS>u z8aaOus}~28Ddbo<&w(-Z1;(Cfz`lyvO8ogBegGJEs>7d$?ovD2Z zD-B)a92{%^Yj2g^HAg|L;f|%z60yAuPVm}H=>@5|oa|?c*r~8ljzWYiP4pP5t`8LQ_cnXcdNQ}#u~Hg21Vj# z8#?|GSZMjFRoaQ#UyxWnYw`{xaBoy4__y^Dm9s=@efoZ$=*F)|>*6G-DDPPto6kyQ zV4~S3j-;$HJ!OQh6W$YTrena@i+n!3>D(`=;$n~TYG;bblX-XMG6K;ZSdum1j z?-xV4#3gT4bFMIX7VW{sY9O$%g;K7o?pT*wWRf&|4b6gnTW!B{xSNZI-W*9dq>z9v zzF#&L-=V0|oVMKO&U<=!*jPIM&7uczs6UXJpfK;h+!t@$|4nEoK)BDnzON4x5C*I^fYW$%AZl3nD!_Mu@CyDuHK7X-3wVWo70CF7{$Ktb z5J0v5S2vw_2=^^Tew9cC0KbX9FysHI1`s|N4KNAVk>Gt@z<&a0Nlhy9zl8ArR{gac zY9Ij{UKoK81T?*$Kg=Wmm=b_N{~B|N;C|{r_WAwPe{mgv#sBIQW>gL$0*3fKcYFar z0iX82o&TOR89#t}{za?*d;0e(V8fq*oN5T<*Q5~y0J?!RJpiu*d-ZGdhuT0obh9Oh z0LC)_cnlmgfza`VfRz8gN!t7X0R(^y$AuAOVR42aH^5hb8^&n_a1(?8O#L4R;vPl* z!-o*!OLhnR7P`!VWct8I07l$o1R_Yb1l$HRq!BQQ1p$)E-%b4=a{#tNh@b~MH3z_I zpbhD7%K@%mkTR7ZKx6pFQvjgrzia?Oz<}frPvKVvIRO2$01bb$1^^3@zk4D8Fa`c{ z74De=f9*~L?>P^@xe61;Ak$=NfGsfM4af@!0FnZK-5&DOCL++g#{ty%d_=R9}k3gEQ-hxnJMWz6d4f&n@f1i?1&M(aO zZ?b?*!9OwuAlHz44NnQP*Z_&ao|u7{l4*bi?>7Ni{QSKCi#I+X_y6De_yC;#-o4}( zcY_VaYz|_B=l`Al4>bp2-~pNI|J!~bXdy7F1Aqtt>^w^{JU~rQvj9jW_r#HZ*9h{L zuK_!31L?y6S_>L%1aPYXy3IWRCd1?V>-uwPUWr%|ApQEe|QA~zu^1- z@d^ZBxkd1-2mp=eKfD5gdzFCSyaK=n<~Pj$$6^BTr1?|+trP(H?WF-&wEX5DKzfCDpv8FaGct_}Ppm!IQc>I)O^uj!RKPZ+~3gQA0a#F29_}9qGR7yOfGAzPt7E z%jKp9%qwnwe(h}>aK-7AxOsZB(+CZpkh~imrWtrW_arO(Fn;y!_VDD=8E1z*Tk7Y) z(dzA8t^dvW#@+STjqF?hn+r?)WNsNS^;Xw++{*r%J-@Vgt> zKqpPOn{UAWmu#uf7RlQKtlkLMSJgtDWDkE{Y_08>wU%cMC)Hu-%_4x6cxU*Pzm6r?E+~kFS2SHSE14j79C>aCqe&*`QFf9Kmw(V_c zHP>CFz}M{+ae?6QbabLJ9^F#ty1`9^xn_(%N^wh%N@=KAnCgr7VmB07%DNMTl^X({ zz8yLe0^WYwEg~<~ecem?q2B(M!}x+f@}xPhL-$iqT6!O1WKW16y?>+rq@+qJgRXG` z!^668ef4w87I&;}(eB~YB?61HAO=68uGXzOzul}+gN^ur%~b^{zGOTtG~{dac1)mB z?AbCY`~xK*&)AE}E}ec|Y6ObPenc}nODJdHJ{?x5{S9$AkgDNSX>>N$fOH%uz#hfAZVVYV_j=$Mt|ld~a`x9`Gz^L@>8Qe!eNj%V{gI zsD!uW)sGN#?l(+h2WVTT&{@n4Csy$?R_<&ts)q98;Ab^AxDBATNzDWSXp2XQ0!~d2 zH>W$I|A1x;@poKff$pl~0NsYSKe?>4ZG2LB;kqr;#m!eCr{6e_i~55%#n2bK!M&4O z*l)K(pf%D(u_E%45_ zgEpEstI*E#CSkWH<@#wFukRBrwaO5r#{Mgw70ORfkrzq#Hd>LZIouIizmZ`CKnbGO zEVy)DhrQz!f2rSAwX4zF?KpSN`dN=^GF~`Ye|IdY#6^0+nU@5*=uyBE)5V)LkY|t} z85(CTzx<5;Sv(Qkn=%z2?^IF^uC|%L?jwmwDw)-xTBZT^TBViXFxpH)-P zFdm@3E}8(F*wk({wMX+Ne~@VgT{ffQc678JQ*YUi^vV?r(RzI79NaQtZ*m9=^f4vu z9kq`4J-|B<3;!aqF2RH4_x@Uau}!IsK;mX!O|9kVh$PEK7HN)G~ zac+N=Kwf%7e<%t(A(P3NyH+p5V(*q}-d(!+k?5sg({jcK36p&)E}i)RV0laFTdLSYo2As14m5&>B9?uW8GE@O2lXyuNa)VG3x#e=V24V8(q;(YaOBHwPt%W;5E>G!h0_ zimkb?`r&xKOozJ&(tp__Vw^QH{p!1NG~(`n;h0H=kx?Y8EOr-!JGfLIUD=vn9LPb- zOYu9YRQd{S_2f%QD0iTvafE~-Vt>+$4Dbyzoxx#Bind_+_)c>nhKt>)0L$B`?fnydK!^s&3`1N zjuUo+%fo9aPke@v$h9P7gTM z8Ys=jWlm!YV=`^kNMf^w4t2)-%ZzH`4!SY=31%)LoFT8h(@&yrqUpY%cQ!R%$$m3l zOkp*IM#15~kM0dkDyB77HIwd{)TbR+lQCtu$RTuoOvd`1H)VugrV^|3_&O(wf;zU2 z;!w+bG?-rR6ZYui$X7Y|%OBCd$WvA{cUAh7ydeqM7>hBXEn|wFhn?C<8FfUw5LxVb z^>A5A)&D|u5ItYQ+O{GQNQ-*Yd{|6Fedk)12xYb%t|7TfBd-yT>-yk-UO0H&Agljq zk)3u>mt#LmK6OIf>?8pnPPj5~3S?J}7g%NQ!d-y}FN|eb1NKya!`^xqVd|JkiuQI0 z&*13If+rCl2p_*9bIuX;mD*;kexEP z5vtv8{fdIsH>=$5?3!b-<=i>o_2cu7QMQ+(`V!frU&LabaD6A4QwVJha;(ZsSs4az zfNHiYQEIUb$)h}k$Nf%iShPE9zhb1Ooa4Wz;E-v{IjtKSC+8-{KWe#m*2sO%{s4mG4-l2#PxU+RQ(0KlkOL@R2 zAT#mSz3nM^kkprNDmGu)L94SbF&t?Ri|H_Zg2Yk%O*=W@6+e5^h(>aAUY`3u(%dV3 z41VZ{`#l6*|K^>^PlIQT$vkK-g|FF#f;TwP+unyqG(J{$Lmq66!FI~ghq)%AFpUjB zn}|DLrc7VsP!f<&+PoYQ6-u?}=4L&+hj#Y6SPO=|1$BnMlflg4r9;OmQ?qm>^mhS9dj zDW5+0*N%-J96dpF7RJt#r!(|IOT>tSy0bOYWgo`jj`4=n`L-3h^k@q3ZOUdgS0xH^ zU8Xuq_-|1=CU~GkIk5I)y(bqg55!pfjB16V;%<>z_kNKwICjONye;YIaRZ#B!eBP>%1SFa$=oGyH< z>w)X4xJ5j;iKSDqV7oAYauSy%cwhr^Vvt$ zMtV3wV;0}1QbPtV3b~a;nBVd0LuDPl7PEtW1U2_H2df^SK5}hJ)@lgM=MJ^_dCSZ7 zLeoRNE{UuCIfl6X^fTknocn`<-`~Zd_h>GzGQMvS)SOnI7g;az?LaH~*6N2Ft__B5 z?^)Nok$86ds4l#I@k~w%ohO$z;Gk;K%=CxciGMDLWr2c$A}N13rd_>W_!x@bkMX_S ze~KGxUzUqZzF2!;c}OG$i9wFmb(Q1k*#6J3rQ+tYu^B|Nq&FFhOxHoDpFY$ZXS#hP zvb-2VAKg1ft1=~B&|nJ)oexQpXF6pg&^Gs6Tm7{1eGi=A>2O{yFDPy3t{xZLrnE{z z(K2FlP-MrZa=_~#o998n@L3js-@ST?brR{9mZ3euAlxzZj!i^<5~ zv%?b8*B>fUaX%9M{IPuiee`O^uB=qk3|85kVpdLvHL9!jL0!$)7`|RJG*^|(uRe)9 zQin>Z#qE`_NxzT(62p6!4tzyTa8i!J6Mxaye4%R9zAWgir>3)V7+=yai-~+p+?0lo zInt|f^n{vf6yNXjhXfIQs860Tn2ktGkH^z5HKl6g-}pkJJl!no168y)D3susOC)$? zFi_>foLiACycM$!QP5WTsu`$XXp**0etga|e0-nu#(?EE6M`p0LRHFkFd z;sOvK*qM|7{SGat@ZtzfB(dXB7$TbuN@Q}J=jnry6sp34ULhJyWYsv12TCFudkSGB zK88y!9#n<2{!yjpSkRYET+3IX5jfJvlP8|6RfN(nCpdXtGMQw@W5ql?pstg_dj+RD z2Legzb(_BJK7T$S1of`=$qeBFdwY?RW>;|*rcCsnwk{gUM}|4&yzkrNq%UAi>WS0W zK1rSWWVlAOXyF;zz6#GsMyILvY5=7QI?FpUu6udR{}#fO(g&4Ts7p5!>^REwnQ?lq zgLL_S=sL@QsJ3|R!_eK`2uOEIBi-GNAfS|hG(&fXLnB?%4bt5L(j_e*-GcgVjB`Bq zy>~u1&dgqGueI0S!^}L-|M&C=Ji4|Thvnfx9+p9b*TGGp)*KstwfoN5AC5j5MYSQq zkSwb?3Hpw)-4Y*vqKFCOn`#bZ!j|oBzhb2ofs1*SycjtT?r^C=fU~&yQ~Z@YiIsaV zp(~v`O~JYvNHUY*bcjEY@tMvWDG6oN3L0>Bh^s`MFiF?^r5C>ltItHAoOM=2ds_Ge zf6;x8QJ zs>gHH{Nqu2ktoB|xhA0=6Q3~Ic@WYgu3#`amwdq$`52f$ zGgR0y(MsZPk9;gr+mYYRyC;H{H-<>wgPk^ShLX0YRA zAuF0ox;r=ziG zTsWf@!(L>&XSZ9BpJbas`Mh|nJ?$dzd}4Tp8t z!T#XtPQaL4|669o^%E{ZqvMsy%f(!B#*4EmM}y_-jH660d}*hy zSEAuZC`@aweSAHquNTsl@`NjYM!K&Zv}UkiTo4$bnPem4yy}j)shm1r2M$SN$l#|< z7tK))==?Cz%lYLFvRI_CZ8@b%@H*5jZht%y@5ZGEOMeZyu75sAsMI|nc^Le15J#O+ z7r%mL1(!z1zb5I09ir)%YyC|rec`e>_xSrH%NOroq9i)7rreKMd0$rU9(T!6`czz9 z^%uhp>Vp=G-bZ`wBzU!Q4S|i3lG$N|h6S7~SGcYm`0HPq7>Ydoz`CnCg#GfyOT~gE zi)}D+q$XeK2^q%M+zew}r(U)ecFUGW!*x2wA#oO1_`6%zeX4aM17%-5Z>002JUTAs zvd`bz^C)z&V^wk-A?l&_PdkpBvQ!Vj(r;1@%thTIrzdeka11cBfNh zMeE%Na>Jr9%P|XuFz}DtDNqztCgwh8Zy)kt>*Hx-Soj*YyPyIKW{IQQ@Y12kl^b%a zendI!W?jMeh|vPBMz-oI+EVZt7=CMyQc2$MR4`}35o~T7){ba#+{Dz$#EAM#Z?`Qg z-!~9k4rfThJ(J=~c0GL>2!`*nYS74R&~Ut->Hhm<5m57JXJnWAHdY_y2eOPbGO_S|Y%EosT1VYhtp zWeiDo4IQ8(%hH`4rYp$+;~goXZ+x`AcC2XR&knLu4djPTD=`2SbWyrextqnkh%52n zD3F^xQ$VyEVxBJ~y2Waj+QXBmP-PN}Aa&Oq8I-QRe%nP1*78c}c=yqA^Odfi{?S!V z#_5oJ2@P#+#Ckr*-%-#UyXYWIF4R3bVDE8VOZ4zsXW2(~EVS2@lT` zS6Ds9G^R_JEQ-^K8TOvqCO8^)(podt4Qy4<3=g5=bE6&oRH3LXGrpZsv~ADeL7Hd@ z%3yLi$>OUg0oyyqB5DzqCdK>q*h8&#mBd(uA6OC#`iXf&wTWHj3aEDO50>L4jY5qJ zZKNF|V`w^!CvQn8BU%LZG5Bo834vRnCy+YSu4WC7Ijm~~emt*xBnBQ*b>T(O_!g|; z9h|5pxKP)FGI`0~RD}%u%US0UVQ`sKS(FLaXGQA>4hlL`Vi@X_$8%fTzcySoDRJ4? zN51ya?${NS-@_?+$9w!lXz6`x`RbIQ&I;}$gxYhPZFN5#y%e`&OI+m~H| zn=$wi4m=%J@4S<>fkFeTWJoX~$oXEuYv0nE?m4^t%eP7R>e^`=71HMyS;QaYS<^R0 zk;+vHX7$n;WpTdx7N1MkPFY3@(g%x2b&XRbI5F?+QzOk1%7`wAY^)MgYE4S*%eJ`7 zUF?5|R=nDv@=Jcq?M8}>al_X z9dxAOzY{U7t3CecR@i^0{haQ#M$mZ(^#1v~rrcopCHPHYYwx5Y+Gh^Rp+d?pw(B$( zUkv6QqycXUK|;z;36FF=Zm=vhj7o+BIVL zxDzAxeUN2|og$ME_AZ!g!cs}Z5?(dobR|Rwb8X4-F}B*e_bW-L-$-_0W?zucS;Y#* z5wISdSe|w8QL?>C^(kq1cP$;PuS|rE-55n(QoJ&(*WPU;Cog7>pA+}QYS7cL`x(5H zK*m|>aE0&C^U{MQl(H?Znae@x-L8XD_Hrxa3tLjB=7?BF+$h0ShG*fQKbAjRVGNxL z0?XHDP*nu|D14PYeI;c=J?fovbJe!>iiF`)av8Cs7EK<1T>t?Jl}=9#$~%#@eN=`N zb2{Zj$siSW#bN!}+3EXxAqtVy19J|<^m}fCg##9`=S1jIOHH{!H6wIYA~=**_~U6O z`)a07CkDlC3awT9t!_QL2#TY~E9>OUv{FbT)Nh|5q~liYDL(PYOuVe7gL}y?Yf&K<=3qo1 zHxlgax(4H)@N%C?ZB1GO{-j^)P5;WLBC2=K1N5|#D-pF_z0Am9%ZQ1#6cM$(QjEbQ z@@=r|s`fW063-n(vYLMqm`iQEcJlZv!A`bd&e)0Z~)h`i3*ZLTx1`i`co@JtsWW6!(b&|U1 z!-LqhdU=uP( zxf);`0Cdlv4<4j@5D)*)2OvOT7W)by{<(M{9O@wQ{{!3rRC@nv`^#s*3m7IeLEIo- z&Y#Z`akBp_eE;Pn;Qcv3G9n(J*7}dg`}fP-KU)+2Z3y`LWr!&-J4nvTgDHde*F^p= z&jP?Zf0)_>(fyZUf)|3R38H`ytAp?X6UD>KAVV4;Ob}2_4vd@}vZV%?8~8XLT9HHG z{T?3to(m7a7Wg9o|1}!EKXU;J&i|;&f4}@|E)b7ZU;)%)LAZ&SaAdH6?hia)3BpAJ z>;ph~KX8NiRsv!;Mlhg*bG?xSYS|x9_gK6@8p1!-^uW>LgUIUv>iSeLFefg)pF9G} z5J0}?5%?>#;DN{S+b6&aEC9zZbs7=^qIqa@y2}bV2mx^aJP$JeZ7K-2uH$3Fo!2XXNyPQGQM0-*c@!znHqrq2Dm*?9eDpxE_fks5uis8 zdjog`iH=PFH9!EB@BcV3xc@ak{GhR_6J7v#gy4)pMi2}- zz^CHC2Dne~qd+*oe*O&mI;IjS%~`C`-A6%jSTQ90B#ZB z1|T5*a#{QWcRNmY=;puu8}QLH{3%6DRkd^9qCWwi|_YPA!27hoWci25G+K#?STUE z3tQ3#z(RCQL8wT;A^7F*faLf9YyY7QqzuCayum%3T>w%74ej4t5nurmSr927f*=e5 zhXjBbd4M(i0w?^(RPl?Q@UN*tfcF8`0AcV4p(pxqfPd2(ev%b_yDIp8yBz?_$xo^R z-*5K>-%rN|B^Tds_XPiM_XPiM_XPiM_XPiM_XPiM_XPj%{(xinv%lXo2;fZr{a8Tc z*MENFAh>?xAb^?weEwgT#n0*ePdbADkoyAA86X`NAY|~UGgPHJ89MYMp!yPKi^`7* z+5_0@DRHI0_?RI(q{PO$d+E{u~ z?2iR+k?%l0s;UtF(#FYz)y;M22*y|Q>H3+Sb4X}qdL@Ra=>7X2$8({t?hYNu!Cxq{ zRz2K%z@lKlXt8r*hY_kGBJ$1hN6`ula0^H-MjKYAQO zMUyP6mkxgvHQh{~oUCDhzxC`~`{2A@TNS*&4y~~(Yxx>u)r3*D8)_o#;B&pVKBlt5 zj-LI!nE@lu?dEiEq&-+P$w%Yx$Nk98KyGomsJ#gI+t(_u<;%VX+YYp=(;=08xdk7u zDIs3hIr7J!Px7rlChqK=H0*Y?XjQVU4zsbM8s{$Q9<>WIL=M>L99>~#XTMKYDLtt! z9_d-yA7~*>dwnXHEH#HQts&Yl_S*VI^_E-4-efsg-D2|dY@*te7!Oy^i+jQHFVf!* z<-CaJ!BR$Gf_C-Np2Y9fOAVWAC7#VKorIdACDm7ooGHG~UF(lE5rkn%iq0~ZMat8j zZLUwDSQLen***^3h)QT>*hZ(^Rr4X-uMw(0yt_)w3i4b;^SB@o=5Hyibc3IDrb$Qd zVlkBX>R#U#WN*fN5xh~Ch3ZePI^lFm?xMQN4u%%&s`c>~&_3)QYnPwsHzV@CspHGE z&@WdK@cQA2|5l`>PB>@6xefSrF2}*DP5N?as}HY+Y5U-*F|~n-cbrP+%?wM`%Zyef zzHmk0cV8%r5w1Js1AqT@Jqf2+?+Jnfa)@c)XYold<4V0(#GltDuY*OwL7-Zybu^Sxecu9x9L#Vlalg+z*x7`RMAaptWk*h@M zak!Jsv>b_SR&t37$%@9tK>CrZ3-)4`YOhV)+NS58^HsI~ioRX&`eOZ3^`8FmM`-NZ<4Y7lkYwdTqumZ@+8egnyQ>Lnv1rx z5Y)GviQXNzk1g&~pWVJ;ip7*2HRkQRVN#RLOOj3yA-}{<7_e9o#uzB>i1qQgA3$R? z5;EX*-?02{_T}r4G-Tz`B+Xfa{3_uk!e#zsV&!2R)P!O>+*!DD_{=lrd^Bh8v-dq< zx{(5xy!akN42?FEw@;sT2z_(~>K+|1C;~y5_)G5A-8Wk)S4`NLGuO3+1;+2U-j$%( z@K^`xwr1havjtSismfURzy`v6n`9CDCID(aw3wU&if`ela5?e6nj-tb*=6#Rc~Cpa zw7ux;tq<0peRkKxCF@-Z6{i;aMX8?@d@*!j;}2hXfx7?LFGMxaX}T_HzOx$U>@&o= zN@7hSyi;+>FKXB8LVi-1R*XMf1M@=!)ZIuW`=$+5v12D)eC3=)j$iTotc`)gZn=gB z{-Xm;V(%=2&=lGd#X23USGl*)k_e#j13{1aNo^9AtY~f5Gp`GpKn1$gKAC`b;NAvW z7RM8K^LPHA)4Y1p7sZ3nY_OT6xfN@~5G@#z*KLPEjT|=qGC0_B9TJbo-->3@(Q?MM zdajDiN~3ec(MIc~>!Hl=K);{bRz%-@u?&HuiTq3xb5ng?ld5?AWjl9_FY`kV*$9MJ z^&aY|BF*Iq3ljh6{B)0UmCd2AHTYanqPM{j-Tcnx1N6m5XA8l{(MZ})^*0Pdpf7|D zt5&h6wo>(d^7ak*@%oAVBynU1xtGwO37eze*m;Q5D!cCvOzt{>0Ca9YI+SNU200MEpz>&%(#nm z9gsEg?%U7ypV#$$T?;r@)9tVBXHiXD4n9|vqE|tCi%?Dj8+VgNoF(Ui@ zD-MnYh=tgcMhyGQwU!D4r``*bsjpNR&Z7ydWGzqzN~DN5%O$|NqZ^cgM>N2(1M(Ly z_YYt5HB3jx`f?60b$>9N41C%1N|`mytjog$df=)LR0X{`$!#z{tbEVfmv@+$#faz4H2Pt9+k zwsjHnC0ek77FU@@`Z#r>Fa9(XUkD-w)u5>SVnw`Lv1XeB3!hOOoOyZdhbb-FOTJ}< zi9S?)FsanJpRrJm5pjoA(#DuGP($CZvE{$FbA(RD{xRt-<;|BMa{d<(zz+NbYm_|upzZA6Z1ZIn#Q zzBShHr&lqn)2wH_w0fMZe?bU~n zzR_h&Rst?2>Ks{iXS~fRCXZyV)ylW^83s)`KXRwqS6`v75>s62%+0QSD9Oo>XI}Xj zlHQst6e}Q5^V{!(oRm+V57$Z*Mh$t;=Tq!kyR=>f1!*GCY+Z<>)8s#)rrKqRM?_lC z+P-H@JP(e;?3Mse?ox-R$8%Z}qa|p9bS46CjOnBlw^NopX~=Pd=eZi3?EQ4xcr8ux zdnKaSa!KF1&A%E_kPPKC;w(nQ%Yln)$>Y;I>e475pf*_D%uLNKqZ)91jX+rLU~sL` zp>!NRs7aT}xOGPp%Kp;+z*mU+w#@)cLx%#IrLh>;^cYb>>0Z11 zbbgrTeAruK$0j->T7G~TqOFC6k)aGgzF6-C$~n2RpY?8 zh1-o?_qpjkYpBWoiWJr#^s#fWGNIG}O0npkAXo6_ZF^O;vH(^*ETmI@Ta1Cjh)*Q= z2;q?WB^-u{M4s)mPz&4;`JEmSU4c-Sk*kS`Zyd}KxD?~oe}ah?Yk6kw`Sx=qIcSAx zez<^)gCO4*=~HHDVdhM331EB|hG&}OA=YRI4Mx>2yg9tf7=Q@#V&ex^Br~MeR)b9? zmI8WWV?(lE=;X!ga>dJDwU2(35$Jaq8BeLuarZbD2BB>!#|vpHNW)>Uv9-6bcXV{W zSy!NYf?Y=53o{)WQ=FxvDm}s#HUIK?5{W)q{*$A+1SeRnG(%omYa#Y}AiU`=pfrt? zA@*zhEK|bT(^6#0U@ze#u|(qoDk2?i#|BlG=7N^Yckwr`6fVV)zTPr1c~p?=8FAco zFbzcOof-tJ@oI6TPT${@kMdC`NXJ00+aySHxPTL(J|3u1@D%LC9a9O>4deB=x^nN(OLdC&7CX;QHOs(y#n7`n6aIY~Jti(NUb!M5ejkPW)ul`NGRWn<7M zvlTF}4=SePX|n$qdAxPBd#e1`&=Ysk-ZUB|W*wbVMTQUW86``18+NB_kOXQQKSP$T*#a~ z9_Mz5)p@S}d?qm5UwweSvo=IXe6aQ%9SR+IFj11+bx^}V{s>K|l1K7rd>f6@RZZOv zJiG9@z}G<5&{;RO=hbp%qU%18P|ZB_#FcCHk=#bw9ov_2Q1O$BlAf_f+&M+FFr|JM z-kTD)TZNd~RsNPR<9kjQOP;+8JcRl*x2Z_v;#MW@q>b5}Xwqe5wF7qAj_1xN>TvIb zI`VY%puAlZk@IhBPjx)fD6tfUO&Z)y!G=hQn)?bRoVlt6t$sMY+Tz{3VD{wqMp3Ue zChfjIXWd+pSs^(5eiscfaYmeRuGuZBC>Nu`dL`1&F}|ze`<6X5fvb3RvxJfhLW7mR z^*%eoj}R+m~zS+h88tD5{<=189rG0`T2%&)-7Ey&NlWp zL3D1L2=Dz;E^AeqN-{OuAxD22Fb#=?LB+g_jqNQZ4ZSlTr_?P{Hd9h>O7KEtQ;RBM zmuXF_oX*?g`(Z|grR1A<|K7mrh1o2a=Ar)h^V-*7J^$R&!~RHSJk|WQ!lLY5{ve2qLmoWFhn*fX62-#hUG zo;(&@j9>)2pErLj+FRgL-Y6|G-BNnTGRaadJao#}tjQn=WHo z{wnl+RFoZqj->#&7a?v;?sfa2MBNPAtkGPpkAmgQnP&M~W$O;}Qt>To!vX{rS58x4 zQ3K2`Se=`!r{AbYEc-kYOf?|2TNN+%NLR4pAa!j3+Lho~hG#-)hk8TpHaU?SqmNg> zT14eZr-)pENO6Ic%T5G2sAjkhRIe=!y8$Mb6w+GuGvuXHEwEXSW}2T|AY#m)xou3V z9^r74mgXC4{C1z9C`s6oy%>;nkc9hfW4E>>DYsGyy4Y+nXtg zx`Jtcr$@Q8CHs~h`ftH9k?Z^*Pst%EjIHHu5}zjRRcpydFV_N2S@5&@S`J5s`WRSJ zZr)E;d>)7vCoHC02Sa!9jiJ*%6ODE+0w|0scs<`ool(hBCV-)dtBUUKfz)N?xWT=BHx@B4EHJPM4G z3+j(~4%WB_IG%xvXpyxS!`|)d&E@4kU%A?%&3G0^BA>#Zj1w@@yI$^gp8Vl?^*D(| ztnK5$6f_0f#(?8yC3FV^-4-eG0c{Gt#*%cYux8>m%Af^5MOY(fHp&9}PCD2#8>-oj zm3$C!D+_d;!cr+D2n{9U*!%43+u~-9y~F(ob0aQ6{OzycqV2*E@y>% zOSZFtnnky)6@Ms#7k5*t!S=M8!MOegKf1X=zR*k{mb%zdV8ILq^`Jvgt!}Zf{hks* z>ILvOjaVN1bka7hc87xAS_0h^4S8|+Ht(3~69yI^lvcDljn&MGyVA0hz;Z=~%hw1Z zoEN2!-^bZ$2+OIatCkL4+JgcfGrI3hCemO%Gw@N9cr#YORa5Fs_+pkaK}qcP_{&Em z6_I{A%ofMi5!wn&lwo%UkX#l#zAl8tJdG=F^?ZM`DZ{&O-?GMO6*-462HEY8 z8w`O2!lE)0m0q|sb$#(f(dk(dveLZ zgjP|lsg21ue5GO7FMjALcwF5bBH;B3x-IObRSiE7-i_s6S zH(qFwL$Xv9Yup3wQg859nBYx>apn8+(&~6=^XteUi{0gxsx9tlj-yVbGJ|phOiCk# zn+-lB-HDypiLiwn%BfBI;*}_iFyG3m~*3D(YG*XDBnG zZi{QbTvI%La{h=R4htOZ7PIY!Iwt zGT8+w!Y#wMGSw-(&gW|LTa$LXUp0C5>WB2*6_M?@Y}mvd2wv?9T7{J}6794e#Tv<=DN z$J|#j!q@x=e#tD<&kasl{j|XT=!e^l<#xfsDgj~T1W=`zdky6`S&eqLr^0QeVI+8_ zQv2(iUC9pEvMp+OpC#!fjfS;xsGxmVG2c)$*F#s}=&^RDprR8F>!|v%gVj2Ero81! zG7R+!U)5MrGuMj}pbn!~R4(#qQ!X;iwCm;2NVdGV8&WpOp~=7FS4ae-Q1=`@w@gWL zZY%Yv1pbgogd8jagNLPcbrlvg%Sy6a`Y?_hEW~zzhpd+qN71G__`>;z|0?=veg(TI zw}BFYrg{%Iml}$(yKXA9rq(MGOu6D#fAlD>H+D?DDf0C3a4CnqgXvMtKir=wcs{$7 zHK+Z+eern`8V%iucHj~CRwf8p>0rz|G|c>hX0}2YWjTGeTKrW+j1;?0*ptD~ruwNT zmK>=$H{tb`dP?}g8O#G~jmJN7vhJKM;&fncIAv$l0(qb8en=v&$@SLa(qq?&(?`kR zRTq56i3?cNJQVitr%lfmZuYCr+i7tYX`g)G3AUFdvt zB|;Ng006|R;Dlkwv`M5!bn%{%rfFQAY1c)*1 z0C4$#Cnx}Q$-semNKnv+By|8CD30Kl2m-Ih*bw zV&k{||6hKkD2qdnLll)TKhj%~} zKtjP!(*6UW|G%(-zfvFmzy<&p0S_d07eoUw%Li-?57qrB+(2qQP-y;e)FJocFmezw z5f}uB>Squ?aQGf_Hy#-N*xZ0P{b$ace1Q7`Sj2rm#HQQ>F(;nvvO&D|KxkO(|1QPP zYJLs~+X0}!U+sa&fpI?Q(~t)!0@=d?{J9`Gft-wgGHDyn0^7=46R4N0J*t-uKb&05E|Lvga!`4 zV(|NV_(^E^X#>&tb!J8#1E&zs@Fz;)!!Zcu7q5X6IF^4r9)I&1cz;E1Xr2I`0q&m+ zjW`hi^YOQ3;NL(Di1-;u3UYq}Vh8r{^F6`v6oiKPz-0I{9R3HWM}`kH?LX%)3S{FH z$Sk?51PS5(jc4ElQW+@!<}#o`AZI|tOK$+M^#hlIhWi&2;cp88_b)8MKNbRR$SEld z14OR|a1fw=_~jsgkk$g~y8$IP$1jneo8y;h0Dx}%)bP1EewhZifwYeY696~oFVg@w z=P!+)oAZ}xfSdD|X@HyamuY~T^LKxozx(6--5=-g{UAX13Ns(tkAHEm0hS1yOUnypK+Ph^7`gT`~0NuGK-( zMT3gk?>w(}Gz16N4?2#<=-WZ|rL^=~as12rGQA1atg8SjFpy z`+z9F#ctl!=Jf@J-6ny`9j>VD!GIu%$cne_cP)B7zVSp3jCOmg1Rh@BEL$Ox8}v7D zW|L~YGY^6t?|U$^NxVvMg`x0F@69y1ruj6|Ac zd3Fc3Ah@qV<@>k#bc~Qw>?$&5-NBQXP_NJv@|SI!crxxt^+@Gq#b|is(-m!s_b;s| zJl*csh!;|qmkQ}VndNr#cc>DYMXWamNlgeTIevJn|Ircl1=BOvrgi5yHQq18rVL%j z*zNOuBbCO)0w{P~F9@AJ#EJ{IgbqCYa+g`TxEBmIIc}GvN+M2QW)HfJS=#1UO?%SR zcW|U)wJaDdywY+HFG#EK&A2rmcHEZgO)6Sw_wJ0$vthH0I3gb3G}CD_i!8_en)N13 zSKJHq5vWI}OIp)yQVP&LUp~^1ST|YB^h$KOeY%iA>?5JU-I*?-M`;hNPw7OmD=S%p zmqwxotA_h6qb}EKHXL}mA_^&>8tF#OkLnY7w-~U)yyA)z$g{76FE>)*Tn%+3h0fC_ zH0T`%WyuSf-N5KjEr*?hR9;&*D%(rAiM(TQ*Vp9m%w%~c_?pJuuu!h-s9F zEqJE)@HDHYsnss+wncPGw<6xWN@TqaWIpy7OyC@;ZaxCLc?2Ip!W52@0_{Wsm8DGz z@)M8!WEG}k@bS6{UVoKhLxsf*_J;36sW34$(nROtT^~b=kdWG)?|1o7A}dD}fyFxp z+G&<6q+z#ElvE@Y#DK*>%u;d0HP)ye5=4nll;lE%j+ouIWB4d^WE`h3Ml!-W+yg9b z;6BF;bJZO2z%-4F-12W165;%Pz)H-fKUSVJAo*x6G^CjU+O$cTY&vs9`XW}9R9)l` z4$P@i`3zG6B?!yKus`*h(;QYX9$7-H=_s{3mj6=^4YK4X?x3@ZIh?1G7#{laar296 z&)+cNcnwPW=5IVhb%mr3Bhtrr7Fp1L0LOWK%U5FAVzbu|(DjaLFEd?kxnDS>ai;Q% z91BC#3IIiSsW!swbbsVLAOaJ>a7cd|8eNEXg-l*Ap>bc-qg47mC-PpRf`>O}PtGor zVh!{6!1x{w#bG*2q1b16QbYL+d$S&+mQzvQ=WG$-`_+~NQsp;Ust#*;*2*>mL@+0E zg67pCkvbI4kN#>8QvbR70L{Fs7UqJpg#lbdjBL6nv@H6AcW%u0d5uRDrK@8r#6~rb zEq-iT=A|h|>LpCz_;(q6B<441R&Yl#%Ww|4$r}+deLsR&+Kckk)ch0gPSc~=@;F=L z;W>h)^Axq1LwKRaoBZc~Qi?kIlVM;&$+jDHU8c&4&5ADl$ehb_1-9x1J9IYWD;#w{wj_sxLzTQ#9VEvu%*_ch8kyL zcsnOq(w=HqqDiEaZ*q>6Pl}34cBW*rr=~xm+;R~KA8U*h@SUrv zzo95(HI)QqneopcbjT2REts)Ej~3%2Ylt9)gD{25b*fIaUSFyYc~zj0XJml_=n5NE)M@)c0<+^PY)PG@p;V(P*@`J(i;kQ<*b9eL z7QI0$JtqwI-C(DeCl1<(82%~y2}vo~!k{qt6^xdxsT?rO6_=qwi8RjBcaGJ$3tBM#Fu1oL@BUVDh5a)$bh=Fc~6azDlIR!y}BExGJ17V3TpBIg_mnQ4`vas>V zz6TV`B5T|#hFP$e%Y-q4>3Sm+o6K4~t1Sx$HZvQ~br!e85%@Sez;pcq7a{zKKYriq4hTh1#KcVvJ27^(d?Y%kT@K zqCvgmC-ko?P#l`7f{2VrE&1xPOu1G?_R}cs7!t1z>MqUt{Ptfq$}|;MAkQ-`K0=)^ zubnIxE;*p#x~aCGj;q^OO1|om2-k(C?XuTtJoiz8^hJXM6JfxxQ@)5Dt*^Da(3f%D zV=b$u>b3~o+A>;6?Px|Nj407x(`Xdke=p3&kZvB~d42s!=HPYnEG_3zShE8GvIaTF zfW|0y3MM6Ie~?+J-{&iK3-c6Wq4XIlr6E#;_ek=(?9+S|D$=5`go_H-z&^r(v;HHYh=e5C zituba|EwDB5^rYei-Q@<>T5qXwtV@#xf$VeA&V5A^U2k#4RlAZiphW0%U}NWB+j_7ndx5nYtFsKl1!*9Ba9 zu{nzD?yY8M`(XP{{!1(bg+@LQ_FzqEpRlp~cQGeiw6&b#s-QBZOoDAHh#?;8m58+b zhXVUblK1&Mxj48m0JR2%A~JH4l^HkM(2vX~Di0m9mXnfu6h-M!P7Bj9BpF8*mka}sI!G|~)#bB6dtq-+sNuDP6l93|G8eB8xe z&vW#k2yAYVnWKy!x_M&!PGMtZPoK8BC$mB;qk;#?ivv%<7;#jV^a-k-ut8B}y8O9a zzg0JZpiJYF{0|c-z5o&fh3$N znWB7LH5Hk_B`&s7f8LjX^7^y#SX7Wz8zxhpkYr(}b43+6jL)uaRS0JFT>I$07*}48-V^0QrJsMWFa`NYO1!1KPZl4=G*4+@ z@0|t*b;49gM0*Sbm29qGBc`Ln&zx9A9JBLl6&9j*?>}B_MlMvA#mEhywzsbT+;U|Y zzBI~;TheQXFX%<*P}=YqzD70NAXzyAMV?)(#B>@*#zN6U+nbyN%OhaDg`ByTVRJ0l zy0%YqaF&Jr)yQIB)yRlU7-4tA*I1_r&bzuifW9X#B z)%i3@zkWH<%(e3QMgStgv*dErb4KzwS~*fl`iSct4!D0aYTNa0E^bEj_`3~ zfnk|cUy%vTHA8Dlh?!n=Mu1pDZpA1PJWRP0-Kw5x5Oh|BhDSX?bDgceGWU$S$D<{N6$mb-C zMCTSQ^CoeQQHaKOlBpnj_wZewm7Mii1R<840^Vty*OSw~aa>m%|?etkUpCr}w1{ zu2SC@INPtWVfEl*&qf}^7!BK0Fshfw5R-$H+@9lxN{5ByVK3o#V?65>q^CV8Qw!)&}YhAdq)ZWZh7jWFN*M z&v#uLPZsr|@6j2#)UL{{rXbRVbg>)rHh_w$t4bXk#N5!#5{C{OQQA&tLLK_`sjLV2^x%{t)$Lm1fKr59BYFW}+}k{c9)Wcht5JHh*kU!kVge1lor zS~f_$q|W$<566?hEQ+^V4Hv2DwC*O1T7;=~;;zbY1(vyI?hQl^hCB-+(n+K}Pa{kX zlT0d`C2Vi_$|N7do633HOG;XX+rH_$$hFoSt+#KF7s*Mbdu<3f(63&n$OS$|AsfOJ z;yk=hmN>Cqs8arMmkBl@B=Oou#=1MPYfG6Xc+*m2|3&)ZC2mM>l0S6|ANyPU8Cnnm zVHp3LI3gUA_a3SkDDeJs!XHzFIO)z&%21>DP`aL4H6RsF@hb3) zkbvavKVJh?5?smji;2a#P>#c;=7fBb&halx^rSRo-=(tQ(gnj#;3Y~>)8xBYOHWN7 z+O18%C6jZ!PEPLtpCRVc-VgUsn|w(xm4id&f$UWlDDbX09W63*J3YQl6p?XC5Z%T# z6_u)Q^4dA1f)}CC?yQ1p!SP;gvh~b?^%g-j^A*|hpD?&iGR7W-r*bc-f+5vt@@L$G z65O3BL&j^4GlWmIBDL%mAJ(Qpj!lC`X%G}V@Ki?~hZR=#Vaf7%>jDYqD@cI*%ddLcYBWG{UPwm_Wf z%hrp)h)z?ua71h?M%7hMQNGSfy}Xpsj^UR6VEz*;BJqk|B{x%)@yn;|0(cV)>`@h& zwWt>@ose;#+>_*> z6z(I!c5k5Dac(OtOANgdPmg*|&28V8I3X5uGKgQ?{efH8t5;Z zbb(m0JI*XULwSBEp{Cb`NRghK&T5wZ&x;jti=_?a1h`i@XX&L^xUv-(~ z5B6^&kzka}L9Gm(PISc2jB|*uKS%giyt{O!UCLn^qEYp)aI8~)TioUH6x2feFrTl> z164Lf1Q8b;FBkj9(-B(s?=@WTo+9C5q4kh5ZN}SZgs`k^_@KgJ<4+5Pi>sw*+c|zD z4DL1TAnB~pjNDQu&tR06aBb<0hHl6sROcy{O>DCS z56nP7hw4hWlrzl##^?~=v+XhFF;edaVpi7`rj)?DP}q$M;cFBUVm@Woq)t2FldHBR z-Eu=^%j1_Y5?vn+k$SP3<58dbmlPGG}OpAg>7AKZAW9u(mh71D3OxY zLSE>kqfB=P%dxD!u-%s&O}edND+ikuE^n-cw9bE7v6Lj*I%c&f?OV2#T_!`hR{99n zm@B(FI35*T!Wx)UQ(E;RES{trLV)d-ie>s9DNSB5U$&aUM^L8*((jA=@=B}rA3UCc|zs3tXiFPL*-bZkRt7?y>1Oaq@i1D`RS zDA(Hey)&AEY3`O{na#V;OTv8j{-B$@#m3!WMC^0+WH>#n?{qMPCG2OM1+AN3nQ`NK zns!@03HN8XC7dxVRV|%iM(ME^hqHnU^M{?H^oD}SXx^vDA@sjR4Vgyta?QMR_QD;$ zjB{K@z59P`U3VbW@At>s7TL4x8R6nu*_m09RklLNCMzz=-o!=r-es4}jO?r;vKq2C zNy_iJs865o=l7R;-}jv7oaa2}dEW20bI$8zPTkB4bb&<#-Nous4hyW)HRu$X^5Wpd zB(X^0s9%2RIv!?H$RmqgEjqV(;`vhdA6ngL|;3W*|b1 z;C8miTe{fvqejRH1oaTjF5$P3ExC^FsU+wtLXWi!hk!M9#!iK{`Bn>OH zon#eF6bxDRptT>0RZ3|3+(Z*O%c(H!#}Ar|cdOOv<>mWz9UKk>ke9 zoHxFj!cCRdni;+FhF|Tz@e1Z_fFHB#;El`aPEL1j)?2sV^fq|k{%Zd#qPoVhfGRvK z-&b(fes_X(`{tws+fO+$f5HHJ%UyKo%NUsUOfr{RjQ20Q8=vzET%y1U|CXdrvhIdC8>Mk?FO3U@zUD1SeQt@PqXt^&>D97W9qUMjfv_BEp$ z^PU`HaBl(XUUf#&Ecar2AD&*lJ-JdBT_N1s+Gv+)8rDn7CQ-Dt4y>X0SaT%A)&80L zk1XjFaHQ@oWF@Z$db*DiSQ0<|5|_^5eW(9M735ahihG@;>>@Yo$M(W`#K($P7gzn; zHnyIvk+>JYMgrXQjOfo*OK$mK-ovlIW-(FP8OM7h)HO`DOSnQ*ksJAyycx$1*Qy79 z7P7yGzQ!^cY)<3ZNV)m#L)AI{t#kud&-xCJN6At=2d&*`F|;e_e&+?J0^!F*wzy0p zR{B=g5d1sKYv}ERFSxf*m;wEK5Wfba@KAX`Y?mw>wgrl@bKsD zkKgX_H`GdFdgv!Iyw2md0u4f4-5iwWjir&wUc@GxvJs`SW^x+>68$yR7M@RZqAtF5 zzu(6_-L-T5e5yF-@{6OVdsg6n%<1_0?y=>bn(X&>>2K0RZ={ag4pv@EH~YrWe);@E zdKS_ALpoZJ$LuUYg!QjR5JD$-d#XK>n-nyS09HY1PW}Hw9e`h% z7C>-1b{Cb=@mEU+R3KkO40%ouLlLRFgCT|d)`~%dbdZ9W*-whQ&%)4u!#Vs7)_>OD z0id4snry^7Y@# zBLEi!5C{ny-k<;-HLMr~bTaHo)&eTGfbt}V;;aJwpB4#7=LL}daZCYSKq+w&uvVjl z2b~uK>_-Mgr5`01BR;*7VAMCE*gji5(1_(ZvqRMevE#O26Trw>xCc)0{VZN zGzjn~km#R5KrGWhOaKq$B(G&g79?iCK!1tf>mVeI8Ub0>1W+HE#UTphOIk=D$T-Pn zAwq!yoMhI3z&dC<()tN3mpH|!d6I>TL zsn16#;|z;9tc_$^?9bLsBDTN$diqlOr=xM%an+A0zLl2a+ux?%gnV;9p8H`t zy3o0PzGWTB#{f|wVm8)(9p9J{(yePUqn>a7_+|e5r*ekoiWy0vY4Ha?$Hix#&itx> zI_tr5RLr?uXMp(W_473kcer$U%de`*8T+YwH^yQjnh@Gj-5i-?35{M4w5N3~Z!pw3 z9J7{=CFCSZ@0j?;_!;W;WpXiOj(J#`?D|=;d`(!(q-OA^K_VFFGe5tdF6gO$8N#Ay zA@LsZrts;N7m{&dFY6?ZsCqvPS36^{n^=Y4GO;Xtm825aE7fB;rukV9=cP`saY-u~ zvr4d0FPXtT@6qnM7>36l?zKywt;R0(jMI(A&fa*p9Og_c)l0Gn@hZj6_GyV^w)v<*vSBqDb?~EPP*du_eUGj8+7`cj z15?@uA{;NzlyRXtO|2TL2`Swr+aIe4_BWZVPLWxv?zU+YP0@%X8FVE1NJ~S~IRLR@y-1m|V44MgHLUn)AZY%h=ZN z#Zu-@45XJTZ2RVCf2=(Xe6M8j{Qd5cagmM}y75<;iIOscVYZitx9t_U-n=xi_$hg{ zYhw7L9a5zintXP6bkWn{@{DSN@_8=Lrskh}i?3T)pM2P=SZ?t6{-XG5K4EFYnoYZ& zZJK7no0g^G1w?yuC)+@t1Z`V`nq+Z)OJR!=jwrq#TgGUKS0mCg?RI*fSa`jO{>NuC z`x`FA#cx$LXRMAQ$*I$+VwkF7(Cp0@y=FBPw^@HAhPc)& zr0#pvcA*F9fXl&A_MyNT&BG4%yrUw-7d_)b`Ro)%Mk4~(@X}nE;NYW29JO8JJrMEc zWy?o$vmss6?BzGexmjM^;I}1}9DaE*ZIgM^e77Q@J6TV4GVSfW2dO#>taVGJ)J(Qi zs@06gDQChVbAA2N(-`UGPwVT!PC3)0=p4N{4Rh=7G`rQ4T5v>CDD86mWVBySmS~a1 zOc+4%jw8K@pwDEx;en6B^sXdk6xWrzWkg$R{+` z4hBNuCwDkitj6J=Y#Rv;ig7oY#u-h(0JRsy(#bQX!0-}cz=*~9L0I*0Hx$0u}g7jMs3HR<49WN}rC*SwS|L)^-)wv1>0Rdf3W#)ROjq_pXvWk3t+C7)Dne|=RWeBMnj%4q> zSVlT#1z}D(hW4+%i@HzQXLR$@o$0_%5sio>Wy!r(gqeFSqDCAZy>ujtKbvZ5|HAfa z5UYE@gO$&pJ7!|yyvDkOoWDu7Qg^kb<`CJr*wu^(?v?-8Dd%Hl&V}9h#O3mUkLc?5 z4TWv15DPW}S^w6sEy;>|I^9$8%M19Qtb@0aEc{)R*%j!C0j|#X_*wOR;8vF(T3?dM^K_(%9|X@vWLA){vkc9CQssKCbN;LQd$NOk z?}=BD=wt$|70rV==GqB?ia&=&OIh$^?#q4XaoW=H)Y1FCh2u(2?|xNO%pfd0xb*&& z=NxHL;0}=PSW&SMqg;Nt1a(sjz(NYAxMoKlcI4xJfkM9V_2gnKt)69Ma6( zrA9|@4^Xv7nU=rH zi~ZL|=p~Z|f-~3?-wAbaQqH4af(m<|ak7|+VzLNaA-Z~%Ic|B#u=ut`Uc%@uUY+<; z4ZfkEM1z$?*Hpqh(x|eOg+0PDs_`!_1;+H@>1kx=UT|V-R&rYl&Tq0u&V7%?mMm;> z(@}`=*Fo@-=($PWsK>2ONh{ABMh_x_sYlt4KV%%LxtSh-U<{5b|4JxodcRioP*B*n|Q9nNSe#lwNFP2(vy2i-nKCUrC zc=wC0n2M}G!l6}a4RNPZt?jw+6fWhK((A?j;NQ!&0?Ng)&xwjtnDNjsn#a*-X7L%c z{bf9zRiq*?cS6L>9Ur%HlGJvyw>+Ad94{5Bx+72AHf>=h=;jc3hbqkSGkzp&WM9TO zL?qD;;i!M7)3iEF#Zv-pB}(dgvqratRiR)t;T;wxxffPX!qU9sC)rr2XD7u9)<^Q& zw+|e8^vx1U8>pRpL#W7=F<92$(6Oj0TPD3jZaS?v@n7;mD0#y#aTfhhG?n~oZ3EG zi*?}P^&G$!#b+mJR-7|qX0{t4Pm-}W>VE2b*>8M%1UJ(4+^nGU1}RO~+XamF31r#8 z-6+{xo-5ds{$VpT8XQW-@~y$X&=1w~UVRAM&_#On=OxS-qXJ2u1MDfh^D>>Y9l`td z%lGZh{Te3~F{ITtc^R-8&efp1vR0xs?meI)eCbuA=40*T%iWEaL+ovw$`1ux#yy3l$u4>OclS2iLBwH%5)G>j(3Lwe?@d#Rd8wS3>!lPrNoBe` zmfpih&ABy(UpF3h-DUWDo;&uIX%`1?ntso?!Wb2H+?BP=1T`~eF5dNo561Q;rY;_YvTvGl z7CU@c<>dRdZw7`8apxT;(7VeIBZ-m}F>BeFJ87<|*RZ2!Vd@R8etYg}e(TW_r7};K z8`IQotR1>GvaBQ+BYW}uYH_U3?eynD=e>H?wp&Bxy<94ngnIUnvyK*3tXm8EUe&|i04DpK&W`0Yl&Tjox@55J$ zw+$NiVr9g<)O5di4$6?f@f55%2eVhwrci7c!K7g>;j7;Y(uBbj;?W}K^e<{&q9H=i zP9$FPBJ)Tfbib5YWxqb7W}mTAFT|=TEOF83xfU|IEsZM0t=Dy2SHg4bj-enn{UXuT8=azmimF!yd zo(<+pvm#E(KwLuHF~X(YvMc;_ugKMIBSJm6u5l$OaH+Joj0_saOny5Yz9pDIiszOj zWqyl4vNnmONJvNjTlYZI>P3pgsZ}3}7HWyX)eI_w6q37j-?db`J}%yp);_;r_##&o zp;FCN_3o<4$Im27b>fouoV7+mgDF|vB(N@rMC|vjfaF4-;P#Gk_?__PHPHv`ZclJY`@|= z`eUQi3vI18jmRhJh0TXZ-PlE# zdmiWJI$6@T^lAh}JrYbK1Tz7_FOoCCHq88Bm0^hPf{*%cAx$J!g4WXX_5F&eX!eLov07;3>OT zkZ_{7Ul^%F{(WCliHdEWj~%VO>8eM8VhU#NF7g>S#EgC)Im!*mAXw0z6f0(O29WhK zpld6}UPUuGm)r6TQ=LxJEv9W8myWyI-bXyi)uZ}c2dzjG_U(s{Jh)_)KKEdd(K@pz zMwtcRNv^!H3&gW2GZCF&;^=dF&r8-}#wwdWex0Gg75}QM;0$!zue+ZCt7K$Ke(>Xv zJ8eBDEq1<&h4xa|`#D5^fF}+^o?MJ%Q5b1t@;m%nGP`6=k~kWZTybXZzP9aIXjI)Y zhhs$vR+L#B)|cv&H|M)1et9qJPfR4ZmaAX<_-4m$hClFjT3&CD!39%o3;mJYPyvoW zz0n43J}R%Otc?ei)C$IX(Z5y~uV}xv&+C&U|B3s#9y@)pQ{^!sVwh#f<|br;ll1`m zp0g>vzNugX4~M;MN`gbar?Bo&5FYM%&;wS(Gw_Y+IIE5SM>dl1PD4 zE(=TailKT=E=NPk>jR!N#}+G7op!qAkmo`NuIy89e-+NBkXj{#igL;l9W=~=bXitCPquJ~(-ZeszZ!S$yyiSZ?>W9!b$pN5b1O=$Nn#w+(*F>44(lo0yNzv8@vwd6 zqxJZ9e6mx4;~3)T-SXOl9q4KG+e$~QnWKQ<{Fc74pT;H>?_Xv=uKi5LbssA?I32%R zT_IJxsq2oSRt)3op+(L)o&1mNLlU7PI7hL!ebV>qhGRp}Zq_(LyHB!9Wfb4vz2@dS zK#FG*6qm_*HyKxcb@lCju-l8t8pUe1s8L9MfH!Ex@wHeoBBo-loY;d=kzIQfo5@an`mqOEV@~3`l_BKFun`G4U7zW%(`rUSVQeNI-i^(n9rg6$J5vV5}M#T5#Ty1T;dpp>|rB$%o}SQ({GjpndDP^?ua zlghMr)Cf1$HsWMV>d78ZI-0ac&R zjds+~PeM#_P~|MqzHfdo^n!S)zqE<%_#B(3uG2O(q?oyweZi=I=6+d_<+Rk=y(#)* zU!vJ(_L|>BNGclct*9DZ9k{Nke9kfnkxLSo#mJ2&N#oJ!v-C0`Sr9`ULz%L@S|b;P8#nZ#R7?MZ_Q@>cm)5> z=Y11nqDfKAuEY|ub5+hVqL+pA+oHbE!H4}AQyUQPq)DP&aV7L$?w8m6Fs8fDTz5am zn{4>IjZ4gX44z3z4i^zqA>3rP8&6l62FMzgvU^8vkJ$#uM>zJIirnJprVblyew*TL zAw)HbZdv@Y_nZV1wU%#9n6a97l0r?aZ=$e6)izVkj12o_U7qUV;6gQ8+c<~G4#|Cm zoCK`|A)yhYLRdj{SeK%_)5gL5W|z0p*o;ghqs4KHCbvp(?b%a|d#l8C?_&Gl%~w<3 z`ZlMg(1dIefM^h7->hoOeMG)~F4rirpGvTw{_xFw`TO`yS8?iDJq|@x`=!2Ejyxl` zVhZ-51@#%SWS!bq@;s3r>6^Y+TzG=b##OxTPKd@T6$p#n(C8Is@^M%msd)=tsk542 zGNmoA2~o7UOXU#JN(x1+PWUH*F1 z^FY{ld|tZ4Qsdf4jae)uj3mE>ez&VxMwVN3YmK9q2&aR!sc)Rf{cgiwjlUWXrS?G7jDGVBG{|>ByQg~1a4G^18+LlWpO@5&>Aa}$deq_~G42g@vpr-S8 z9q4K7>9jsp7!{90%5sD9S`Z_$LxwD#4NeCa5xbt^p0lAUlCL7mR!+&z(FuXV-QUHrpjwjm zBxVLM#o+)fz^NW|ray^&5laIm@6|s{BETG|`2a!z5Xbyo%_}4-n8399&*U(X>(USm zDZK_5HDw?GMh@V&f16wYln?*@45GlNme)W&`-zT)?3ICNkgbOSXU0v zXKeuOP?glAp!4@1cyv%@i?Zy+a8S(;oQ)Z2zX{x~{VwE!Dar`swOK=6SYnJI^ohJG z4;4THq5{ZlCCCFCl(-6l-)$72LR7?@R1u0t2LbeJkS8XnY86EC>qFkyz}rCJ6UFcV z9l#gpL!Q{6)Kv&X+EHU*fpS+t!4n>2tO|sQ8V6kWg5Zoo$k=O;4D1Sh!G}y zBFY0B6x0fUEj^$HaRH2hOfW9eNgcvO2QUaIZ9#o4s+4+19g0Qm=Xo`V4A=!GhRjj| zfg4bD3u2*Q&R$qB0iYntT7U;kA2d)vYC_Rif}v3!IR~|2F>Wf#zMLq0v;g$#2|1%AYTX>Yy`!l^aTq6CJY0SQ!!++ zF%*BY>{!64V3AOBzygKJz%D3jVZ%|{2qN=M0V_}$8W)4g! z1kf{t%%NB;l)s#~5D+8^E{h?RETC9SU>g8|0oDjh7_2hDyFe`$2pIi8^8#hI1woR76Ql^Cpv>96F~v^DJ2}N@tFjac~PbAr~y#Rc;cv@*g}GU8YC2; zE{lMVr>ll3Bm^A2)f)1_0*)YXBFqyg27Cg{!W&Zpm&BR>SsZ4Hw*}bJRx9)Q$Gdp`QVg1{TX-z9C^zNiaAM%L2-Z` z4t@&;9_Q^1{1jm4qZ)zi`#_P{06$+yRP2NiB@C#C?gvF;iGuY$IgLEA0gyflvjim% zXbpxwS$cpggUYc%T?I1`0=MN80$^;h6RXmJLmcEfpwbBbbzZ=n8Nr=R9wNgQFb5kO zj${ggD5+1<3jReWa2xz8g47IxxcOnH82zeL z$OaJnKbP&mWg+Bd5JUlBJpQA#fFmQo!1(WFXGQ_!oEk)q>B&v~ofr4Y1)!GQf?GAPhx%YU;m9 zMNp*VNR=>1{x@k6AZ0z(4WE&b@4AcSeMT_mi2?p*6+LC$4}*Arvr32m|5D(mq`u+M zgFniXM?jQt*h#JlN{ph7i>uooIpBht`N{Xw_4rFJnEZ*#t`Q)o=eMz{z+ehzi}Xil za3e!`;9rCPk#;ute@rI;us=}tV2Xs;!KAiAf$1zFAxg4Sbcw&M3=X<5a0keWgtUKC z5=KFkET{Mr|5)x6pW@$~=rMrPFbb0T%~=CDSx*rw{^C4!rN8D5POTI5FruM`->l=& z5Ea>((7#Aw$YGQaN{|%*?l>nC35fv~KZVWs%P431{jvC&7Jm%|y!%8d@I8FwW(?$n zg8DgeK%ZEMiu@E<<6jdzaX@lpLoAf^ht@6*Mf+ci18gRg#er|IBZuO^9pp?~%XoHJT1L8?>yK3s5+5bqzB z>qr#KsS5&&oU)wifFz~_EGuE)l^`1l-TK2qnG9xl3YYSanrBk~G9tJMqh@HC4AK8# z%>b;drwA*5vHoWR2%c?#wPZ-*53hU*iuaFWoN0P`&i^eIj?_wqsFBE&KXPf2!zmC2 z8~hZ?|#zeWemhfX%x z$272c0h}3dTUw+;Iz&MZ8dRNV|2Ofe1D2&jcmB}wWT4D+hUoK;H~hx~!E@1xAny#w z;1BcqDf9nK_|YTj*a3f(e|^9b`fUCp$a~;uirg1^;sHe>VG> zl}_P#{;|?Y>a#ZT{KZpOCO~p!LyQ>UY45~1mf0xdoIOMQB^x}Cp?o(t8{+&!`xVdv z8AXn?mw=5wn%42`{T;stqQfOf6wR4mDBPeA=F#h$!e|UL)74Y zc(OFK1rYNesRjk8hm1e2@mtEN-~J=zKYk3JE5O7lkuM6MyQmko42TXXSqM>5{_{+6 z_6Tt@BP7!nc!u*Ygs7SLPfc>R6;HqYt-cuYMIi*k05909g%B$_|NoveLLwq39k}4& z)dDqjK5a`^M|YRomadF&_~}IWWE~tF-B8~JQ60E2;PCjN2qfP=1`Hp(GM@gc&JY>D zkKs)4r{fiPTt5M!0)J0HsQRL(x+$htxNC^aQgk~xh|dihkhs|B`s z#&+>+|6o$M#(dM4S62zCQ1V20&LM&HW{lxC=k=9Orp-U?jyI(D85mlAdK@?X_)h^~Y;#zn<%~9|gWAmkUB7nzqbK_=I|`o5g)k^Uw>vMbWCRMziCm=62IUziEVrBvK3z9@TNx#Uw*-vb^P!$D1+#Ef%c*#0{CVzkQScOxl$ zs>~GcLs9QWd+A#j!zJj}E>}}-+VmWBhFvk~CK{X={57wb#Z<;;L@bq+j6?5} zsMoo74nENrErhJ5pegRP0KwY# zpJ!fq;jy&sqJ3p^B5JJaA4TBOuuBkXhquHYFjYpMH{pFiJ~t6bS1Le%xi;iWpKJO2 zn@O8w1jlE@v2sa0OxErq;EF9O?3^D>DZcig%)r8AiI7rIGteD&W4u=4oQ@pl z7urmrvPg@;HqOs|r5c_Eo|p#bV8Q3~()mq0-J)h6#t;@!zGt~-i;vg6_W`2|84k_Nl zi-vO-u^!pDBnxkjZeqf|b+Qu8-j-rfd}!0H>q;=xA0=Nj!z5?Tsk*kAnC^X;I+lU> z{UF^(=;Va96WE#tCETa6t^j)g0xYA&lhad$z?oBn6%Lg$v$!Qy-eF2<@6_2P4fEwaE|2;<2~$T3+l^c zfe)MLoqllDOE6ILb9n2y9WE+JKAN*rx@>-L0wUQ`oXdX_-sK(T&s2i2+{DxEc)9SQ zI#0laFp5UQkBp3yCvutzs3ZD392 zpsxF|+Rd=Y_*$K!L_|Q?_4Z)2%;8$On`R?-l<&!j64p6wT`a?GRd4%1H=A-rF$R}^ zk3|{x<|X;N0uGo*LKuolbSm#~tG4s5ETk-%8pwIv3+9ilh{K4D z5NPV8X2E=0aHZLFdQ0tIrQ2)&3*^^QI)u7JwS|siF=z(44KN-{Y`jXnww1Xl;~c|9rM=8p|7qd7YPKT%hg@X z!7y`{`jW(Uu1w$x4tZf6!aUCZqpf;_@7a0o=-HR3Ts9{f$Gi+B427X zUfs;qxJ;81N=T6zR7lj$Ssc{C&b?s!=*EW-^RVLE@&Wjrd#seswwQA|a3pP}ySWzS zc>>Gu=Xm5j)G^}^8!$APi0>AN2VZ9hHD^*2b;%>;tUMyzc$$&sK-_Tr(d&&F&t={x z6|6jWog06Tl9-|n{Of_1rQroU&pAB;@y&^!9O>oXgBxq>#8xFY2wU&M z&u#jO98>_jl zJ<0jZb&blxhegXLu(Eii>NSb6eB8s78*Gmj6VVKl+B~RRG1h})hww@$Ilr5#dlKE| z9$t<~XHD&^{uRW=nrcUenCG*WN|Ns@T&aM4`6(P$ry^vUlOATDaMw{k2bMo&kM=D5 zlUoXrb0k0WHWc*5!nOn!44Zu$L0qBW!b@z{_bW&NA9EtEEP};o)H~CGq@tv}L3sYK zF32usVx3(7wHc2qj`exV*isF)mmlkJp9VDB1;>Vk=QH{KNZxgyMz9vapqqk2^!x@- zn*-3Ul64~Ggsv9k5AKor7Z+9Uk9|v3pt2@zQt@puGQgCVE@=MtS&^rl)4l$#LP`Dt z=EVRNoXEG7ep|tx8`2emHidX7Hh&SDPl;yQzbzk6`IT+T)QR(aJ*?SYuE**vraXn) zZH3EU!UnO_s0)Us_2CGiOq<6{t7P*fSuP0ose2{+h<8!4dD^WsS@;O0`rIGJ1=IZR z*ByDoLTq@hC^S`b*L!*QJ=RaCC-0%;@40+)vcW!6_|dLn^RxSpcIEClcKm!Uw`j}# zkm52v5>F_vO0@14A5&1c+grbqM{)XmvfuhYF?CkSyRFGi!&{$3BJ^=$2uDSoymRqq9;#p*1okj6yIx>Ulhd=Hxw5v0zIo$MlQ1Wmk*vaesusQDSO9w=wp9=^s za>&<~4rT}KS^H2?;vp`(Idu11ge_Z$TbdRnl--gxadN=Pdqx~0l~=;V+lQUL)LU^c zr$5b%N06O3?&2WmtHl=?Vlv{7bRrX1h_-_TfZiTX_#N5o%3V}zH3cH zWR|M=+J#HtD8#^G`qGY?zKanXN8 zlpAjC4`1W|K;q%oHJxKlu&MHnOi^kk_|uIjIw>-RgGU>4+Yy;o`tb?2%KAGZNxV^X zaRQIAmnIH|Gg}1P+mgSVdOa|A4`=Uc3!C8V2oD{4n(7r}+f!9;S_|uO+PS?~O6hj% z!bKHUMcjUJRca#_T10fJt=S`ea_?dX(KOl6C^2T?w_nNak6|1GNM45g| zBz-??$~&#(MfhdvdxgZZrO*51G4rck+6ld=5+m(lsk^BOixNd#(Tls?kOfG?->^DNNcgxv*X2)&+zhZ(p8c<>b79ut8*CXBmyX4XZ&U z{zh9Yadhwq=M;h|%Ul$n;&rM;zv^30%b9kxSv8++2TWPzbo=GFiPaC{?{1E_JX2QJ zSy3eABGuMc=U~HEPInlE6|8aS9q1T-n`@;25Bj6ANy?r&X#B+53AeXMv3@nyDQ97^ z1h`$%T&6AU=jHF_tA5u|7{r;-Hx$iKNuO3GP&l0O-j);Ls(;(|c6f#)+TwcT7a;}* z3rssPRmBF{&sV3@#lB|7kSmpx&}Aa|bt>@(Nu8=`%)sr(dQe51f2RP24(?dr&{ zWs@V6KMA75XdLxpLaGavy7w`KhiJr!`+YxN5#fJAEYA6D;8iAFIrXPQsw6ZDHM)GS z>|o`rK-EIJas+NxphK$#Rw6~tNN{G5!bFFO`E7<JI2k1)R6$?7ExWwpo%+pHBJJ4cR?tN)6#qDJt}Sop|TUgN8S z&=Q4DNzpldO%rpm9K&%#5o#}2f$0Jke&72^ZWo{1c0`Gb*vCHTbU|AN&=wOWY|c)}|o*em1F%0~SO)&eGx7CI)_(`%_>)0`^85Aq+@ z7JoQ$V7(GwW39tRLtmL`sHKtcF>kloM>C={n7gVmH^|2Ia*ZSh^I?&<{dLNvR3q-> zH)B3>7s%k>;bZz4WC_M2z$+}ORD+drHd2a?*G-qL$4%{}W_}+0V!8Xc>AUpXH>8DI zHHgWd3`3m$Ye-oVh>RKpH&5S6LF7sBPnZIPI{_>y1yW`NqQX$oL&EEzSPD4&KmUXP zudI`G5EBH4gShM~4EB1~dWe9627Emm75oPd_RpvwzOn@}3nKad6Nn zN_6m5r2g_C^bX>L!%tq*jq&+(9CaORZadzw1h3O4|7`f{WgP^>{?SlF4|#J0rQum% z>n}QC(1CsgVus)#NQcr8QSMD(bVF(XSkl#zl$-(svj&$GBX6E5hFT7-a}#%ClNdcy zLf$0MKPxm;`LebuN%6b-@|<$z;ad?EQ)`E1Yjwu@<+yB;87wk zFXtt{OT$zT5=XuS&bVCppczgk3Jq9~_*- z+2xLNiIV$#?_!X_yqG|AN%JD|nko@|5j}_2yRXzF>H`D?ThgGT-4IObq`5SzQ&MwH zDizZzKhU`^Q}0$j z+OEHNT*HaOV?x1+KG`f4hgX=+Z*P!8nW4<{CJ#S_cC4w%*Y|NB^C+5cON4nxf|LBE zz9Y5j*wyX3ddWPp666_=$BliZoOgr>RQ*k^QYFbTtWMF2bHLlK6g7FOG_~DG?s718 z4wshE4SR3TYFa(9ist{h0*{Ws$`BzzhMfl=4YNZ($SJW2Led{%h)jQ;4Mlkd6rsS{9jI~ zV-wB&H=pMtNK4KS*62UJbXnA##Me^()(gxlKF%_XMh~nzH);@48F`LspL!b<{g<4X zc5KfD@FM(M_w^~8gx|1p+kF(Ga7_JZyf5d>CL~3cUb^GXYtg2uuq@($l}uB73BJJh z=q49>_9dKQ(Iw0)PsYwC=#^nB52hw!kQx?kN5oV052V!&+4OrFUNIC`x+Im_)C9#B`D=GX~v62~&!0FvaqwS607SlWWSc(?#o-j=7yGr)pwXX;KpN#l|Mt7fy z#9TF7PpvtOlDdYJH`XSS(V!ezi1NS9*I+fD$F4*7aW$|gY!ktppjGe@Il=Id;`SW} z2R|}>sp^I)z?GV|c9M6c^3iH&)4qLO+DOBXhUl-*l!w3LfFRvU;6R7|BgPo@&;nYJZG3 zGpVfG7CScEyCae!>@+SL8nd5zM(U4mr`p%4mMYHisneX*zC{12S zI=1a!6R+%-dw%s?-px(s>%YPRGI5Vt2c9ema)iIS%h7R{(vJm|D)5;HQ>7YRwk8>vR{Qfcc@$On)sPc;*<|E6 zLYy|=6pyaRRBx#&N>JiV_)HbH=$UNU=K3l6Ix4(vR-PfR-le)Xm#Z-AZZ&}3J^9^v zrGeWpp2;=(yYaAgXwhV1ixanH2hp955#kxAQp7u1j?63Yk*C;S&;Nk8As;3>eSOTCRrvL5Y0a#6S?A5lM}eV9M1(k; z8*9E;y-&AQo*g+pc%Joh8mHee{>JsbmCdHRpYhV)!YJRP(cDMB_hZPK_4S8WNvRRa zWK>75h+AB_`-c6%a`lm@E=e4EX18;akaei)a9{pfL~MwTXcrHl^jT-VhBqgDJYM%Q zrBgMDH|6K|oeshMwJn#gS-%qdOdO-bY$Y1qW-83S2yt=z8bAMir{ueIre{KnftZId zT90!1wy8n#-pD`{UCRVAD&fo58yQR!_b)sp*rt(mcYHC|Wy8SP+_{w-oprxMfl8I9 zIV+9MdO<#8bO&K^+-NT%bEM|$;r!Ys?pi&}Db3wQyX*DcG!LN3&}` zUnlX?vfd-NpC5eV#4pPHYTQhHG5HAN=_2c{B`H-%w{j&-9y+Z(fcHG*Ne7{E7GzgYc5sdc8OoR_;XE0VQ9Q}?CV zWG_csPXdRzPjiSO*~Od;L7xim^@4bQ|BaTgzMXgKIGuqAm#eCc8#*FXJ}r?$%R6J& zlBzkB2?zZ{mqT;=wbqKX2#&v)FHI6o?(%ERUj9Yr)M)wrwFq&8{CfTzl(~*Z!t~HY*l?IjS<#^mynz zV#)didRidvI!4v;fGqTGCJdI@*L`HJD;9|6`%n@KCwfPEOL3-kb|Zv-{aqFv{@08D z*Vc1@Q`x?6Rw#RAWRDWgIXJ^8dqq}8L^jzoyVEd}6&Wodqa-9`Cwp%xvP-4NCglH| z`o7->nYPN)`%;hO9}Oykn(JT#ySotH1rFQb5h@@T3frW3^uEL!vN*!|+j&!ySXM zoebT&v3D%ERE7LYu2U(Iw>Nz;PP!>TO?@t75*cofZ|&9{e^~J~B*dkN^Ohe?Qp8?u z#;m%xtbFIqh3j#~@hKa#1)fLkhYgQESv&sVR0;2g9t@{>>ew8mlT&UTQ#^{p)AuY=E=9TBU0B`(XS!FK`2S>9sFMy9HH71>!Kp za}173p{0kurcpZ632D8FphD}lp)%Jvg!~n6%_yDJ#mbm`QuI1uXXwVQs#}q-|8(?3 zHBmLarj5T%pYe_0nG;t&&rkPk#%{;tFYc}Sl7FYOd6=jer~i}zCAFTXAjHPVyH_nbzLt+N5CuP4uUMl+`Z{fo8OjX=*R*;5IHHcMs~k`I%o*i5 z)G@l{FthwZ-SVu>*LepMxAuwgND?R2o12TbscPm+9L@y;SIR=F9OWPc^%S#Ukr77QbZDZwgMoGpCC=ed(n0QiBu?|5?jS61LY#AC4a7RH?0T z8^6pZjdl^T-9FGybho#aQ^44rKIu5_s2W-l?fU4LNZAv`nunH4nvJ2fl2O!oukSml zH)y6ymu9F3G`sfYti)v2+SJWWOcgHSae=J#meQSAy!^}!+u6^yE()<|^|*ptWv7km z2AZzZ8_iQOn~4bDGLoiTRTaE=>@JGQwZHRy%@%o3N{`23^&{JuM!7Xry8t|u*L}Yy zW;*I-uIB9WX6Bv!hw(&v&FiVzUutWQ8rrrhaiu(Idv9`GEcCgpO|taBqZG`T=qcU8 zw{N+Ge_q4ODzSB(VFSK|CVNzXc%HsyR%S=?wG$OuT!wcuUu1qD<48>MWHJiuTk|UR z^PuqcSmhBuMt{~xC&}Vj4XIwrvEinch7W~%OMLPdf*Yr-b$whvo{kXHL_I#-&9RKGqbfroI25U?C4YJ587-`l-*TAdK90;%b6rzE@~bzdQ0nel--+z#EUIP z%F}Mi!eYtJlhjA|S{D{jgDWu$oWkh^Q%4RRvdi|uG+R{LjZW%)ompLwcD=aV;*W>~oS1Rj%J5byjj zu1|Mcf{WYpL-F!$rI^0>yIp&)C&KS~t#gFV4zf=_MeX7zKe9|aIUJle;SE37n9eg4 z8~K6wbCg4zX&skc#j%4brtUs%jInV~w6$9cpEj&FN|Cc^D|p|w(mwSwA(4!$7_-5DSVE&JaRD#A z(s8EQ{7bc`-%meMiPN~Xsyr)I@%Np^3M?0gx)+`12Bw6y@!yMHaV}aRr7acf~!`pfiJL8vG&-M%J_C>Q0O| zRv9&%*1Z=QnfpF^lUMW29GZsL?#bN;%GdY09ac0hCgu&``J-Gh`K_;ey;dlmO~kC( z6r3uMT7G=yeEc3?+K-BWgKc{$0n<+mY`?|XzY!DK`s6vQX2sI*`eMpA)vAEdeIE;Z=*9T%gw&9%rc(3_o+C=TK}uVwlB3(^JuBxQ(x`e{)?aLD)kR` zrC*=x=po%WD=P1=kTrhsU>U>NWy%S%HCxLEveld<$)9%;UoPC;TD$*B)Isd^^$@Oe z4lb{CMaEOz`0Klpjfbw}_P3%cvc0Uf7OmoG_N zQ;BeYZtjsMb!4!8iOn4)!7`>a6Ow-NI z@ZMgqMxNW3=Za$K{S~phM>C6?iU$*ov&z~&pPH)VJ=EVR-pHcu?cJo>7Cw}sdxG+v z$Yb>)nZZCu##25)3MUiWxa<{bBgrl?Gt=s;lSommyc5a!R)NMhBYV&n0^{qGZx&@Q zCnT$`rxe$YbD-1tCL#~}-?AX*!qbd|R~7sC1nc6LLyx6+?Iv~Nla zAl=ys8(l5^8NTve1Fyk3$8K~#NQ?=EYK<=EsiPSaA9^k9&M=tK5^3gI(!zk4wkDZf zs4KP__`c;Mcr&8zWB)sv&+0_Bt1a0yxRUk^osq&wlEBKXtevP3ZsmtEQ5c7wGVmUZ z;)y1o8_G+q%c0SBZ{-_~U5!Ufv{pU0dfqg2o6^X|hGo{i&=bEikoZoAd6@W~qzG}O zU&I4amkEA+|Jeo7No>wX3S-BM$8+WME@t`JwglO7-VCz)!ONCy|J*(=&6Pea@Bn*I z;49{vmK&4QZ&y9E6P?*yCf}cQzT05D^?l6ZA^wNU;-kA&tt*F+_w5#SeQr}I2C?_> z@40{DYIh{cOqiK)$B$~v9@G+4N2gsGh@C$AeDkZJl?v@O3r3fwS-t!R&o~zL9=Lxp zb8GPXR8?*z-=ek_KHtC&3!^*bQssW?X5(T*Y2rpv<&Wj48I%Q*z`t) z$EO~ZJ=tiI@>#a;!Y{0lgl?H-8{a+0-968a7&$NhoM@jHO7aN2{JBSo<(i^?tM{CG zFO3;LVvv!s8{^NlO!z=%HWVEe=kjtlSG(ed`-opm%E;CB0SsB=jG*+0t#sAhDl_zhz@@24BC`sxi;pdLL;{ zn(_2EB+qVu%r0%_HC47H!g4gP zKSiK%_@|dxrIF9MFUMo<;ML#f%twOv0F&!Z83f6eW<7H4izUS}XU-bdi_m{grU)e; z$WzBoaaZ2>Wq4J7uCNCidMJW}KMpr@b(}S?;2>vUxu{Gn8eNI6z`v`0;4b`=9=v7o zU4+BukY-`8nrOyB&Q7`06XBE1E%*HnUHLU~aq8_H=Sc2m2jgk8C2`hE!5%NJ+3C4; z&~Prl>OEBdW-44}$L`S2t{GyN4Ls^t-ea|59GHB~;L*C70c#n8*%1a zReYBBr|{%cpBZ=19x|E%8|(?Lxp&f4TAyXL3aNdHYDt&rH^r`Z%&0j`SiBAK_`cQh zQ1QLx$75D`q+=qDcaztQ)Jn689uAGN#~+jLnMgJ87p-+LeB<3~(%sl);d&VpSU+>XX|I4Q zIAT{0Pc>iI#O5qe6_%`uXJ&9pcSj0UmJ);TpwYP57PDOEHCOs7CaakLw z#+Efkofqq!x%;Oh?AB5F1C>-@Z8EyNQi#o=;z$*}*KO+&+3+R-Sa0>t^DgNpQ5DA(aAg`X~Zolnob6Phc*U#l1;M} zj^qfl73*fs7pbf{H=o$ErP^s7T{mbSL>}u5Zd5;#^G-`J(WZvVe&@&h-pJ4T&$<1D z2Mr>1e&uh-rJ8W)6#T$1=vS{4CqH|Aj_2itsEX_nC$dvf#p?#wCQ$Zt$w`*aiJk9Q zvmetUwzi-f77y(8MkgCuw$XR8hhjR4E=!eFTQiy?X&jd=6%j0<6-3Tw8(5bfXu7yp z^tPOI@q7}0roOMY!X&nUncIA?V*ImE@-m~##Lx9T@J%kV6hxv^l<7!mS%2qKCZtWk zppm%$#i!BWOWGyUOc0C45kk{^8FIX8-*n`(GUrk!B3^NNz0naP@!rprz3NME9RA}ncIV>U)?UpA7G0wcw#Yu-;Ja%1pa3jkTCT2qw<0;VPQjW@{Tk zG~bpZJ38oTrJnanZPXDh4km-mp7=W;;ieB08hbWL%k+wr^PP;{27cK8Da$ z#@f^GEvoF~64YMOhPv}6hYSyuVNm954ks~|+h@pVOUQU$iM&!Vu|EGa@#YeV)FqW} zWOJTid;0nKB4e?{+_Uu$4B|-ini--oU+`Hz)Ki*dmNXf<{Eg`)iP6!~@6~df($8k- zo=_GOqhENSp3$pp`M~H5F7|Y5aub=lPgj<}BmYs1U@H^1>8Mvh2@SWT5AyQ?$4~~^Z`!1 za15ngU)r1zLMQw7t5=`jRDAI@Y4dQ$QItwE_Gl=V$IBCkZijsre!m^^kXKBdo`VDA z03Tgr4D0?tyKbsBYD5$iN3;>cHTazLcJ5h$+Zs|V}DL|cNrXCJw$*BxXq+vO=X7v0WS-jy`*vzR2(;qFq zLX-Ta1*+Fh3)DF|@Hi=6?@0c^Hu+SGVftsWCTDly#f)<&lIX^2^R#8Aj?>HgU(#`E z!D|I49jjjbj*Dn`xo)&h^pKvtu7JXo{qs>O;@4wA>|%#=h#xO_TGn2Xe!>wy) zSAX2Oeros~(_0B8+H=X1XrB5tgm5zdm`bJu{*r!UpxOf zIs6P3R#@*zbUw&1%=#iioTqn<=qIN8uy~Sg?yDNECgiq9+_&PZq?3<0WP?4`Ieo6! zd^WnWlsxusfCpPxbv?r{&ON@KWrF3t;PH6=0=cl=IcE3UG-LvyEJ6%lqZ4lo94(KK z=3KxR6rR81v|PV+C;ZOL!|mMAp{Dg0!k8!dkxcj$k$P~C#=T{m@JWzKZwH=RTIB!F5B23@lyM8l6=Zc53*JXjs%TSWovDBIl~@gEQHT`$a`nGKxs1X77gz%#mft%)}P20 zBDm<58cqF2cHUgjI)pqQt$K}W=x*2>qt&au%sg~epTepiU+X&J6%d?v@6=g|3#M1R z9*=j?;)VvR7|0YWumx@d?d#u#g~BOVHbx@q_M_w1WmDRW6 zQ6{k({_WD@7OF>_r;$^%%FDrR+VcL1F}*AF_88ZL@n@PIc(eahV5+><#XZA*Gb7ug zG-P6`bD2~$VX|;__{8DSHqsj^4Nc2E8n$CBbsqKqB4ApI&=``ftuNh<;tsVL z>9&8u`S#oq>p;FQp4Yy|2e+*s7oXi_Tj!xU87wd0@pgv&TpLN;ciiT~Y6Um7fcR1s z{pADm)8zPxgi)H4RyQZ)Bo?|-Q9LK!yu?e7Y#xcoeK?f1_E^`SwQsiWIex|A-9q*8 z)S`0bc_7bShv zLwT*nEUoX>MmZAh+G07)I&BH!%SS#-U0%P=(b#wPE?zaq$oN$H?!m0w`sRkFauFm} z<~>W7g=4X-u;JV4LZm^m+0L@X_hjLybkv}bkh$&r-FAwpu~6*RJ_ovEiHDZJzHVZmj5Ew$-EzE)2Pz zt6LtisdLSD_|_-K3MS1~R_>po79D5I^HilUMMr2qMB~FAw}Cg6xXLgxwquqIOH_m5 z7DEeV*9rxr8`=*puPQAmII(=uSf4-QC}_p=fd##;=xDjhQ!hI6v6BA9O8GY$gOA(y zd^NL!Y$x@;p2U@vEMMrZi=|=i`t;I^{tespqStqR(I`{>Qtci`jf-!VEM>n=#Vg+^04O7{Ir$N7@JtD!VDRg`7F zoL|VIWkMAnhR^G!k@nL}{$qIkAA{Ku%7<(H{e7p#4~Wiw z6gkaB5!yLaM%`iVTJ{o^&yae%`-bme_jh&sr#azX0ovVOw@d;>#6As|d(VH9pT|X1 z%_@bi+%KFlmlD(7KG~3F#6ut5l!yjq>*F9#Vew@j`*OHR~Nkw zh82oDL60l+i^!xmds?`!nM7q;9kF)fz~qoPrAnL!58qJj>%B zR8}<5(>|{tpHpbN6Z3HPoQpvqt^AdzT`%k`+?t1GueX|gwaMx3`2g;W zr=@XUiKAWj*e_nP0HWhzlmZ8`r9)|Z`8&?l)cF;ZXfc1b%GFi>pd|aY_PNmBO|5OZ z(yvQn8ker-3lNFo?-$jZjUKqh=R_Td&@W?_<(>_#Yr`-9oRqwAK+XGp?k9?#FBGFu z{P$D*KigKwV&%Smrefv%(k`_4f=2P|HU}qVM^y6ld}(p%y5s!={jL)(t=k^2RGG8R z)4kk^?$Q{&Zrq@Ax_SxqDxz8BMv26Vl2NiCKY8I^Qh9(ay#k&wQ(f64N=`Q__9dYC2^~0f$Uw@B2U&*>_6gb`Ft20k;(OA<8ccTPy?hID~57 z(E?IB)Jy@(i1RlU{Ca=Kow)hX77k|@wffP18J#s%|5|Efd-RQWxUmtZt-E$jeQm=-JMm*p<|g0l zj1xX!AEw5!D>zOcD0%2fLhG&du1(D33~P)c|K*i3pCXa>$y|6hnrT$xsL?@{pCjB_ z&A7d3dOZ$tkE;wTuNK^9onpJ5Y0HW=BjRRUUL5$|qwBR@rSq=PY9V$>O}HsWfy*7rA7oXuy}O94e%XXVVGrbRA%`CU-6V~%eoKi~66o*5v+62y0b8=7$twain1}F z8fk8qEo*z!%C)x<^ytg_Mf2KRSDfqGro%O!I;5!+=4$^l)tFtPm%?es`)wG&J z+HyzK?}wcsj$-V>htz-Xs@a$lxN3R2Ybs~)ox4r9=*q^S(TlZZ;n!+*%MbLpwXbJf z2|@2#>c4g8;Vl+Ad4lqa{kz6M6~zNadpGwWkBbND;~l;mRy}LpS{}ruz6a-CHr?V_Yl<-9pvf9;91Es* z6BE#Q6TGHu?JDwGjU^@mk)6H}+9#SCAe6r`9544MwsWWayMg=bEO|}M_wokXW=&7A zugwiJcUwd=h?lNX;&Y})=H3Z@Ua6&E!jP9AyUay9oW*}J3#*pkx5b(kt>-^q72(;Q zzG)I>)Mb;O|9S5@<+$o4B`mu?dc`B)E3t~shDKXNXpTha?WXcKD>KSYT^2=kpbu$HZcF$RqPOJ6!|ITXFbZ z^>wG1!MC^HoQtv*2e0_qazFJ%)YK$Lxr}S)v*r%Eb>153F*sG>otxuwF)_y`WZa?j z{O+A8T@zUk8$9!-`rtKsj5FH@rl5ppt%i3vOu1-$?Gh@oTs=AFuT*Uc4B33*yM;L( z9ai-qiIF0Tv1qJvLOJF_gKq?Vete2{MzUCNTb*ct???UBU$%HYb)3_I3(4GFsVuIK zEL9G^#e8{`~#sDjQv<+yyx=jLjVj!nLvU9=m+C0T1UH+%!M;5o~vWS|Gnx{6ZD4x#ODn4gv z_0qcTG~NB=RJO4yDgh@h@voeZCT<|Uf9jZAI(ApakzMuX!208u1Z|RrH~7s~scJGVM|&Il5P`AsfSn;CnN)YzG8)O*~hENB*d9TK4ac=X9Y zcGL2*QkTkuYi-7tgSdk@kaGY8HRk_))TD^ zTEZ?7eBZ9Vl?}3f*hPDUYjaLy^LDD~1;+dJ$=#vibe+Xdqo!U`x)tEZZuY5XxB5)U zmhhRUIu0ub8O%o2R$Qn##w{Rc?ZMT`NEc)lW00WTi|al3>qWfz+PB(WPLHfo($>b_ zS`y*s>bQY(7RRR8Fm*ytPsnGX*XzGe_kTv)`6F1P#bVBv{m8zLH=TENdBke!YdO_7MVfR~rcjxt&-#?B-79~R8{X@2~7%$O$-;&Gg zgJ^8u#Z&8QqZYKiSfW$jUJvDQ73ax!o(c2g!eZ0zE=vV^ds5&tui}G5PD*ThB|28S z4Cc<(FPa1mR{tL(5#KD`(DdJ;JnW)qjyiH+mM(3Mf*#n;qWSwnJ zUqH|)H!lC`IqdU`|6C@1*z<9yKa%|JnNa3y@lzjcuxjJoX9XwYnSXGtITN2$IeV6w zb(iXw%g1RKzRRU+hwL(q1O3A~nmnc$Y^bGmlnw@`nS+-GN|$aQSbWaor)czZnW?|i znu#xgEHF%36@?1vP!La#oRl|C=Pv)4_EGAou~2Wtf!niQ;!zf#D)5<-2>p&M#{*kV z2W;Ma%PzU#P2DtC!%<~zEi&!xm!X(D*`UO5q^kAdM+cn4M&+TB(tGO)9{u`c#ZNcO zE93U=&th zok!bwh3-)SO*7q~unhE^H+T1R|fL%LmV*Sh*? z^SvS63;ZHa3aFPm44qcb#;n)=aTNb7)URziG5L0q=)wNSxPOAdJhf44H zg-OXD(iBWBuE*T##I5_Hl|1$ipEysS&Wu}LK4;YCA9O1&*D+1@jlX%X^1_FwPr1%GyM;>~ zS-Ms#rGnDBq1QbjKF)YHGh8;3HEAFMm9D-sy;;vV8-)##^FBP)@%HZNYnJly^$-11 z2a_{GI~mV^AHA`COWSS+FPiY#^<{sj`$gloG+&mHw<)4F@;Ch@OSiVpm0u+a3AVIs zS*NQfUH5E%5^e0{H=$wF*lh10WWH0*6;xhpTK{;ihPt~i>bu18EWX|Mu15|lJEukN zeiR@378hHf5%0;GuPW|EU(k4CMd8dv@+3+hZr&BmO#@e?@7{Bp$T<89fl4CtwCp)M z8kxwxpMH-q)obT}kj`drht1nv%LvQQUe0*%{!!>d?&dQxYi5<<2I=gonTDUAaA>&* z^CPkhoH;owKG2;L&AT)BV&L0pytnL<#9GBbjm!ZCvhWh)k@M?k*7MsuTS)_dZbk>% z#y1Ghi*mH2M+)|Dxohd1#*2S>srg;s_RY+=t#Hm`pD(xOH5x%_sN|icDCPyN8s^ep zj-0x=YsZP)D@+5ogDOY1cuwxx#ndg-rasXjI`fw3_-d)kf%qV%W9}bllgfEK9|a{T zN8oOA8h$vO(4{o1_;W+eyX;Pcg+@^MsN@nY$w{zsn2vdg@W4~JBT z$t|bcdOmhF>G7jknqkQcqCfE3(Pf`A?vMAPYZ=q?BEB(xNazn`QT6*A_!BwZ_tQ|| zo!N+FO_?Y8>x&ZOTc1mQvN$+}*vt%`C_XFI(dMqab&uBU7KK@GQnty@=BD$Mh4-_? zsjTbUsya}ZW5Q<_k5iOYZZBzPlnNXv>g<@1`?Z{Wt;(VD!vb$_ZP%DSEB;VFlS<`P z#Cuns5jixf)%0#W?MDvIl##vh1MiPGJioq3{w6nfl4NSe%{?&nG=%>FK8YrsbA|@m5z9NB~HG6HTcpK#2IXV5U1K}VUtI*dMrx(_iA1Ab1XG4y?2gk z3>7;y{dSrCtS3fd@su1jLsSYio;Uv47loF*q8_oXt6r`br#kL2$k{25j!=F~C+bd* zbT)hY;KJIFi|X^S_x>$vU--$=Q@`nku=yLtev5nB-VtAVnHXoqedH3mZ@l4;Dsw%i zBScHi5e5eIH(f%~95=p1%oGx1BR)22T^HPL3LIK{ol2C(j}>9#^w)IJHMPNKdZM%n zeL}l)q?O1IkSNbhSO2&*sivr*{HSLt)-8LgyKXg2`q#H}3|5!Zsv8~;)q?$15>v$Fvb*Vwpd)Plf<|0)#*9{1udJ7XYLb=y zQ+~?kE7AZe64dZF*%yea?eR9Rk)q2j37Cn3v9%V=2_#1sN5x>mTK{6*^E-IO=GKGk zg0E+6&x=mGQQE{_%Qm6d^yXQZBz2QNoU|ijraHbIV8zE`9de6IUc6{f^+mUOB2^qo z1$E8Hy^@sNi;=?Nnwv35hrk6i@5HrV3C=C&nen^zwdeF*xwieT7jggEj*}rSK|B(> z#@@BlaB5I?`-p;qjiSnz&nvek>}T@dJ&* z<8wt<4vUjoTt=uJwz{Ww9XBl$T({~*JFiv*3cWiN2dhg`I{$dylCOl-zBMCrFpzKc z(9KQsgOLq0ufQdtRI_hQejiNyI9sZ9CT=KK>om{~T5VtYR>#&MF|tTfriY(ilNvT} znpZ+OEtys8+_PhS?aVRnW<7eGL*m#rXH$jHt2HZg#QKBA?UTd!$c>}VYY;bIg}FL= zh+$50H7}b7_e^^;bt({--D4gTlCD6KtK1*{F{gQsVUf0+hj{P372~Pvb%~c|iI*7z zqN=s8s8#6Y<{kaIG5Gf7{j-l9@Z^`T8`Ue+yVjo$rB*B`(tKsSne(H%Rex6CCbL~6 zn~IPM{-sAzyqU7;Vbh%<`jcj&JtkM&m~8IY_;!A+9@5TFe|E{aS-P*X`R9u+%6q%$ zy-OnH+NCuvLZ$gHoJ)en7t+@m))z-F&Pm~%bf1Kc7jncX-3VhCD#}0bhuUn(K8T^FHn{n zmYgx5DRZhv21lA^67R7sncl51I5hvh$X@hR;isT7>-jb8@+XH`O@Cg7jZ(qGd*@Px z(jH1GGp}E2S*qN`EqgZl?id_+oGuM!?R+2c=MNhdAW5wUqJ zYTgnU_Au%}YoShWaNZ8Awh4=l7p_yp*8)@$cUf|6~o5!2QXk zCizc}IFM-U_x}fjE4ll+`EY~W{(8J*SP%o1Ou5BNCPpp^5drtXw1IOzx(0{ zsZ;+OkN`uW3EAWJ!*IWkdp`_?!r@^9f~Yn)|9uUyAl369bp3}d3{HYj{$XD`3{H}e zauNyy(R?Tj^KU!+|IiO|=feiXB0yPjNM|fUk`S`rHwhNt5V{yF0!R2De}6mVZ>IPQ z2Jnvssp?^!kr>z+V?l~&C>F3NB*>c$nFOSIh8_tPg$4ou+77S-9txB|5?)gHYlUEe z5+Gwb42CA8Ie@goAt1bnz)GSCSwHu+!-027{%wcc0th(I0triIFNuCGBFYy6d`#8WIP-pZRdWNB=Wz-;Qz=> zBp_fY^%Dk!@;iXx{_Fg2Ujz=iVX{SEvL1LdAzFbwQGfn5PLUch)* znD9XXu7k=2AiR-yD1m@lB~TrW1mYVeXiyj|95aApVLQg6U~~kE*Ziv^{%8LYbQ8fb z`42S&EMQ^A1`-AQ1E_RF;r>0(|F{VSjM06lCX9yz=echVp#M-f?0%aZG$be|=p+)r zvw>2z1kf;0nJ$4q!z3oCfCM`d2_RUdAod?<25}Gwi$Q`D*tcqc0YI(L^8^FJ+5s3A zid8_q|6|2}?LQ;{!eqY$RuXExfOc?oEP(|Y2C_g2EZ|Q#49c>2FdzoTvj7I;SqY$s z;FDlsEDN>)BTNAMOXvSl`JiAIc%}Nc{`rq+0FdDQ?ju4VfT6A!62P#q-2qD$%El64 zEl@U=050YKXy#x04>;j4eg{>%pf)~I5`lziHc4O(L1_UL>LT1%?AwSW0s}K>@W7Ra z?N}11ll^Q02o5Rn|Fn~W5tk$qL%4s~KQSn91{)7}cHlUF7#`$H!NFqr)( zg#dmW1PcHLJ0>YS5Fqf~Ny2s~1uRe4PQV#M2~G-uhI&Lu0E5|>QV0y8&cS~B4*+rS zNdUvcb^>!W!6Z0Bn!J7IE(MSZ+Z?c$|H9z!^G8Y{frx{3hIzuE9qc}U0E2l>pnMP< z)nGW-ks$F{ctdbU03|{x6wvz61%ld5f9-^@bEFi|?s(BO97?udr9Ct z|6?Ff9141>z+54eDcMIRXkhrl+95Hp(*ia)Oy?2Yr@uExpeJCg1BV|r3EVvdDio&C z0SvY~G_aZfp5*r&;1UHk2MUEJT&nKh9SRr(a2U*x2Ie-*t^qKZ#72Y5RrnlI|LC{B z>{&D*S*RoiFqkk!11k$QAP~AxItHc_Or@YDB=>vv0RO-*Lc_E-u$c*;4DH8^BoIt6 z7~BX)@AKIIXm5~%0wyB?`C{>~a{}}RhXLadIx(oL0mByXRiTUqY;`yRqk&NSGv0sp zEa1_<9bLlCfdm2~7CJGg1PfaX8VBbHa7_%uAm(>z&A+e;>{(#_!+a6o*uyM2GzJY* zjsU4Jbp${#*8z;T|J=dmFof%Us67jSFron>047ub3`TGmDY(lBU@$WkBL&Ju{)1Ju ze{S-S00`q_U}C}e7{FkB3|uNujgX8H4`W<#O#x$E4EPuhdVl~1BW(=0MuhSis7VWP zNB|yzcmGFV{Kd@RtYG(f;8G2~9W=NwfKV6)ggc<54X%k`!XASGRY(8sOJHUUKn7HN zVKA7#OsGE$3GN@@RDwkl>h}Cap5O`wRHTG5Jq9==usN_&K$!0Dj6#CjHz*H)0bxf1 zOcvOY02u5@uz(U_3j{3s&tm>#W-K5@C<*}#CQHGhVY>q`nDvVVlnC81uo55=2rx4M z!i)#tZNZFCED%-DbH{?4F&MT0aY2a{if{8)dSZF(NGYP>KSPXD5AVa!c1|{PV%>-0apF5 zVF+Nbpe2|XIGB*}fH4a`RfYh8aZac>y&nr0P;nRrg9SE#VGr~StQ`(|&kT<9w>wBM zCjVMB7Iyyw8WM`RLly&g77l~iUBJ!+6_25Pagu)txIfM*Xo!ZoRsaUX52P~)i$L`O zAVNu~9SvaMh8xxw-0?wSAUFpf5O``(6e58xff6BT1@8;)cA+~4)*@6n0Wl9s9fQRX z@c%!e0nvW=PJpQiTQxuybTPndgkC(MP$-!81h9Rk)9>>KZx8(bZhRjtfS4upHUuR> za77^~1o2L&xCQMbp_U1FWdhs^LHkO`0=p6l1kM~3h2YeoCJcBjBZXFvjL`sQVT0X( zU?^l%a0daCo8ayQ8g~OxSr8FUF*`@b4)l#ArMQNhy*oD`!{6oT+}%N$UIIn`smtf> ZW98-J?`3CC2|5AKmQp}KMMstL{{UBdsKNjM diff --git a/deep-review.pdf.ots b/deep-review.pdf.ots index 6e72049a919295efe478dea7feff8b5faa643a66..52e230dddf901186797f93af84132e5ade17cc96 100644 GIT binary patch delta 206 zcmV;<05Sj10?z`FA%EbKCWG&f8v=!$^&JJ5joP2kAWtXh$tJd3cO>XQZ^iHscgI0o zKo02zZC~e|{+u_C2=O2a>3;Z7Qq)fulu@XHq_!a(@;Fi_ewdC%pQkLI{=w>|jhtS2;`qP38#y@DK(th8_?TusjJn)@LNG(A(|^ z@dQ}sdXex5eT=k<^n>!^k-j(~?dwY0uW@G$c($n%P(rxi2=N41=X#Ow2)wsqoUbLj I?~@k+I2jROY5)KL diff --git a/index.html b/index.html index cf51aaef..5e8d31db 100644 --- a/index.html +++ b/index.html @@ -111,38 +111,41 @@

Micro-RNA binding

miRNAs are important biologically, but have neural networks produced anything particularly notable in this area?

Protein secondary and tertiary structure

-

Jinbo Xu is writing this

+

Proteins play fundamental roles in all biological processes including the maintenance of cellular integrity, metabolism, transcription/translation, and cell-cell communication. Complete description of protein structures and functions is a fundamental step towards understanding biological life and also highly relevant in the development of therapeutics and drugs. UnitProt currently has about 94 millions of protein sequences. Even if we remove redundancy at 50% sequence identity level, UnitProt still has about 20 millions of protein sequences. However, fewer than 100,000 proteins have experimentally-solved structures in Protein Data Bank (PDB). As a result, computational structure prediction is essential for a majority number of protein sequences. However, predicting protein 3D structures from sequence alone is very challenging, especially when similar solved structures (called templates) are not available in PDB. In the past decades, various computational methods have been developed to predict protein structure from different aspects, including prediction of secondary structure, torsion angles, solvent accessibility, inter-residue contact map, disorder regions and side-chain packing.

+

Machine learning is extensively applied to predict protein structures and some success has been achieved. For example, secondary structure can be predicted with about 80% of 3-state (i.e., Q3) accuracy by a neural network method PSIPRED [77]. Starting from 2012, deep learning has been gradually introduced to protein structure prediction. The adopted deep learning models include deep belief network, LSTM(long short-term memory), deep convolutional neural networks (DCNN) and deep convolutional neural fields[78]. Here we focus on deep learning methods for two representative subproblems: secondary structure prediction and contact map prediction. Secondary structure refers to local conformation of a sequence segment while a contact map contains information of global conformation. Secondary structure prediction is a basic problem and almost an essential module of any protein structure prediction package. It has also been used as sequence labeling benchmark in the machine learning community. Contact prediction is much more challenging than secondary structure prediction, but it has a much larger impact on tertiary structure prediction. In recent years, contact prediction has made good progress and its accuracy has been significantly improved [79].

+

Protein secondary structure can exhibit three different states (alpha helix, beta strand and loop regions) or eight finer-grained states. More methods are developed to predict 3-state secondary structure than 8-state. A predictor is typically evaluated by 3-state (i.e., Q3) and 8-state (i.e., Q8) accuracy, respectively. Qi et al. developed a multi-task deep learning method to simultaneously predict several local structure properties including secondary structures [83]. Spencer, Eickholt and Cheng predicted secondary structure using deep belief networks [20]. Heffernan and Zhou et al. developed an iterative deep learning framework to simultaneously predict secondary structure, backbone torsion angles and solvent accessibility [84]. However, none of these deep learning methods achieved significant improvement over PSIPRED [85] in terms of Q3 accuracy. In 2014, Zhou and Troyanskaya demonstrated that they could improve Q8 accuracy over a shallow learning architecture conditional neural fields [86] by using a deep supervised and convolutional generative stochastic network[87], but did not report any results in terms of Q3 accuracy. In 2016 Wang and Xu et al. developed a deep convolutional neural fields (DeepCNF) model that can significantly improve secondary structure prediction in terms of both Q3 and Q8 accuracy[21]. DeepCNF possibly is the first that reports Q3 accuracy of 84-85%, much higher than the 80% accuracy maintained by PSIPRED for more than 10 years. It is also reported that DeepCNF can improve prediction of solvent accessibility and disorder regions [78]. This improvement may be mainly due to the introduction of convolutional neural fields to capture long-range sequential information, which is important for beta strand prediction. Nevertheless, improving secondary structure prediction from 80% to 84-85% is unlikely to result in a similar amount of improvement in tertiary structure prediction since secondary structure mainly reflects coarse-grained local conformation of a protein structure.

+

Protein contact prediction and contact-assisted folding (i.e., folding proteins using predicted contacts as restraints) represents a promising new direction for ab initio folding of proteins without good templates in PDB. Evolutionary coupling analysis (ECA) is an effective contact prediction method for some proteins with a very large number (>1000) of sequence homologs [82], but ECA fares poorly for proteins without many sequence homologs. Since (soluble) proteins with many sequence homologs are likely to have a good template in PDB, to make contact-assisted folding practically useful for ab initio folding, it is essential to predict accurate contacts for proteins without many sequence homologs. By combining ECA with a few other protein features, shallow neural network-based methods such as MetaPSICOV [80] and CoinDCA-NN [88] have shown some advantage over ECA for proteins with a small number of sequence homologs, but their accuracy is still not very good. In recent years, deep learning methods have been explored for contact prediction. For example, Di Lena et al. introduced a deep spatio-temporal neural network (up to 100 layers) that utilizes both spatial and temporal features to predict protein contacts[89]. Eickholt and Cheng combined deep belief networks and boosting techniques to predict protein contacts [90] and trained deep networks by layer-wise unsupervised learning followed by fine-tuning of the entire network. Skwark and Elofsson et al. developed an iterative deep learning technique for contact prediction by stacking a series of Random Forests [91]. However, blindly tested in the well-known CASP competitions, these methods did not show any advantage over MetaPSICOV [80], a method using two cascaded neural networks. Very recently, Wang and Xu et al. proposed a novel deep learning method RaptorX-Contact [79] that can significantly improve contact prediction over MetaPSICOV especially for proteins without many sequence homologs. RaptorX-Contact employs a network architecture formed by one 1D residual neural network and one 2D residual neural network. Blindly tested in the latest CASP competition (i.e., CASP12 [92]), RaptorX-Contact is ranked first in terms of the total F1 score (a widely-used performance metric) on free-modeling targets as well as the whole set of targets. In the CASP12 test, the group ranked second also employed a deep learning method. Even MetaPSICOV, which ranked third in CASP12, employed more and wider hidden layers than its old version. Wang and Xu et al. have also demonstrated in another blind test CAMEO (which can be interpreted as a fully-automated CASP) [93] that their predicted contacts can help fold quite a few proteins with a novel fold and only 65-330 sequence homologs and that their method also works well on membrane protein contact prediction even if trained mostly by non-membrane proteins. In fact, most of the top 10 contact prediction groups in CASP12 employed some kind of deep learning techniques. The RaptorX-Contact method performed better mainly due to introduction of residual neural networks and exploiting contact occurrence patterns by simultaneous prediction of all the contacts in a single protein. It is still possible to further improve contact prediction by studying new deep network architectures. However, current methods fail when proteins in question have almost no sequence homologs. It is unclear if there is an effective way to deal with this type of proteins or not except waiting for more sequence homologs. Finally, the deep learning methods summarized above also apply to interfacial contact prediction of a protein complex, but may be less effective since on average protein complexes have fewer sequence homologs.

Signaling

-

There is not much content here. Can [77] be covered elsewhere?

+

There is not much content here. Can [94] be covered elsewhere?

Morphological phenotypes

-

A field poised for dramatic revolution by deep learning is bioimage analysis. Thus far, the primary use of deep learning for biological images has been for segmentation - that is, for the identification of biologically relevant structures in images such as nuclei, infected cells, or vasculature, in fluorescence or even brightfield channels [78]. Once so-called regions of interest have been identified, it is often straightforward to measure biological properties of interest, such as fluorescence intensities, textures, and sizes. Given the dramatic successes of deep learning in biological imaging, we simply refer to articles that review recent advancements [10]. We believe deep learning will become a commonplace tool for biological image segmentation once user-friendly tools exist.

-

We anticipate an additional kind of paradigm shift in bioimaging that will be brought about by deep learning: what if images of biological samples, from simple cell cultures to three-dimensional organoids and tissue samples, could be mined for much more extensive biologically meaningful information than is currently standard? For example, a recent study demonstrated the ability to predict lineage fate in hematopoietic cells up to three generations in advance of differentiation [80]. In biomedical research, by far the most common paradigm is for biologists to decide in advance what feature to measure in images from their assay system. But images of cells contain a wide variety of quantitative information, and deep learning may just be the tool to extract it. Although classical methods of segmentation and feature extraction can produce hundreds of metrics per cell in an image, deep learning is unconstrained by human intuition and can in theory extract more subtle features. Already, there is evidence deep learning can surpass the efficacy of classical methods [81], even using generic deep convolutional networks trained on natural images [82], known as transfer learning.

-

The impact of further improvements on biomedicine could be enormous. Comparing cell population morphologies using conventional methods of segmentation and feature extraction has already proven useful for functionally annotating genes and alleles, identifying the cellular target of small molecules, and identifying disease-specific phenotypes suitable for drug screening [83]. Deep learning would bring to these new kinds of experiments - known as image-based profiling or morphological profiling - a higher degree of accuracy, stemming from the freedom from human-tuned feature extraction strategies.

+

A field poised for dramatic revolution by deep learning is bioimage analysis. Thus far, the primary use of deep learning for biological images has been for segmentation - that is, for the identification of biologically relevant structures in images such as nuclei, infected cells, or vasculature, in fluorescence or even brightfield channels [95]. Once so-called regions of interest have been identified, it is often straightforward to measure biological properties of interest, such as fluorescence intensities, textures, and sizes. Given the dramatic successes of deep learning in biological imaging, we simply refer to articles that review recent advancements [10]. We believe deep learning will become a commonplace tool for biological image segmentation once user-friendly tools exist.

+

We anticipate an additional kind of paradigm shift in bioimaging that will be brought about by deep learning: what if images of biological samples, from simple cell cultures to three-dimensional organoids and tissue samples, could be mined for much more extensive biologically meaningful information than is currently standard? For example, a recent study demonstrated the ability to predict lineage fate in hematopoietic cells up to three generations in advance of differentiation [97]. In biomedical research, by far the most common paradigm is for biologists to decide in advance what feature to measure in images from their assay system. But images of cells contain a wide variety of quantitative information, and deep learning may just be the tool to extract it. Although classical methods of segmentation and feature extraction can produce hundreds of metrics per cell in an image, deep learning is unconstrained by human intuition and can in theory extract more subtle features. Already, there is evidence deep learning can surpass the efficacy of classical methods [98], even using generic deep convolutional networks trained on natural images [99], known as transfer learning.

+

The impact of further improvements on biomedicine could be enormous. Comparing cell population morphologies using conventional methods of segmentation and feature extraction has already proven useful for functionally annotating genes and alleles, identifying the cellular target of small molecules, and identifying disease-specific phenotypes suitable for drug screening [100]. Deep learning would bring to these new kinds of experiments - known as image-based profiling or morphological profiling - a higher degree of accuracy, stemming from the freedom from human-tuned feature extraction strategies.

TODO: Make sure that at the end we clearly emphasize our excitement around unsupervised uses.

Single-cell

-

There are not many neural network papers in this area (yet), unless we count imaging applications. But there is still plenty to discuss. The existing methods [86] use interesting network architectures to approach single-cell data. [76] could fit here.

+

There are not many neural network papers in this area (yet), unless we count imaging applications. But there is still plenty to discuss. The existing methods [103] use interesting network architectures to approach single-cell data. [76] could fit here.

Metagenomics

-

TODO: Add reference tags to this section Metagenomics (which refers to the study of genetic material, 16S rRNA and/or whole-genome shotgun DNA, from microbial communities) has revolutionized the study of micro-scale ecosystems within us and around us. There is increasing literature of applying machine learning in general to metagenomic analysis. In the late 2000’s, a plethora of machine learning methods were applied to classifying DNA sequencing reads to the thousands of species within a sample. An important problem is genome assembly from these mixed-organism samples. And to do that, the organisms should be “binned” before assembling. Binning methods began with many k-mer techniques [refs] and then delved into other clustering algorithms, such as self-organizing maps (SOM). Then came the taxonomic classification problem, with researchers naturally using BLAST [88], followed by other machine learning techniques such as SVMs [89], naive Bayesian classifiers [90], etc. to classify each read. Then, researchers began to use techniques that could be used to estimate relative abundances of an entire sample, instead of the precise but painstakingly slow read-by-read classification. Relative abundance estimators (a.k.a diversity profilers) are MetaPhlan[ref], (WGS)Quikr[ref], and some configurations of tools like OneCodex[ref] and LMAT[ref]. While one cannot identify which reads were mapped back to an organism using relative abundance estimators, they can be useful for faster comparative and other downstream analyses. Newer methods hope to classify reads and estimate relative abundances at faster rates [91] and as of this writing, there are more than 70 metagenomic taxonomic classifiers in existence. Besides binning and classification of species, there is functional identification and annotation of sequence reads [92]. However, the focus on taxonomic/functional annotation is just the first step. Once organisms are identified, there is the interest in understanding the interrelationship between these organisms and host/environment phenotypes [94]. One of the first attempts was a survey of supervised classification methods for microbes->phenotype classification [95], followed by similar studies that are more massive in scale [96]. There have been techniques that bypass the taxonomic classification step altogether [98], (sequence composition to phenotype classification). Also, researchers have looked into how feature selection can improve classification [99], and techniques have been proposed that are classifier-independent [Ditzler,Ditzler].

-

So, how have neural networks (NNs) been of use? Most neural networks are being used for short sequence->taxa/function classification, where there is a lot of data for training (and thus suitable for NNs). Neural networks have been applied successfully to gene annotation (e.g. Orphelia [100] and FragGeneScan [101]), which usually has plenty of training examples. Representations (similar to Word2Vec [ref] in natural language processing) for protein family classification has been introduced and classified with a skip-gram neural network [102]. Recurrent neural networks show good performance for homology and protein family identification [103]. Interestingly, Hochreiter, who invented Long Short Term Memory, delved into homology/protein family classification in 2007, and therefore, deep learning is deeply rooted in functional classification methods.

-

One of the first techniques of “de novo” genome binning used self-organizing maps, a type of NN [105]. Essinger et al. use ART, a neural network algorithm called Adaptive Resonance Theory, to cluster similar genomic fragments and showed that it has better performance than K-means. However, other methods based on interpolated Markov models [106] have performed better than these early genome binners. Also, neural networks can be slow, and therefore, have had limited use for reference-based taxonomic classification, with TAC-ELM [107] being the only NN-based algorithm to taxonomically classify massive amounts of metagenomic data. Also, neural networks can fail to perform if there are not enough training examples, which is the case with taxonomic classification (since only ~10% of estimated species have been sequenced). An initial study shows that deep neural networks have been successfully applied to taxonomic classification of 16S rRNA genes, with convolutional networks provide about 10% accuracy genus-level improvement over RNNs and even random forests [108]. However, this study performed 10-fold cross-validation on 3000 sequences in total.

-

Due to the traditionally small numbers of metagenomic samples in studies, neural network uses for classifying phenotype from microbial composition are just beginning. A standard MLP was able to classify wound severity from microbial species present in the wound [109]. Recently, multi-layer, recurrent networks (and convolutional networks) have been applied to microbiome genotype-phenotype, with Ditzler et al. being the first to associate soil samples with pH level using multi-layer perceptrons, deep-belief networks, and recursive neural networks (RNNs) [Ditzler] . Besides classifying the samples appropriately, Ditzler shows that internal phylogenetic tree nodes inferred by the networks are appropriate features representing low/high pH, which can provide additional useful information and new features for future metagenomic sample comparison. Also, an initial study has show promise of these networks for diagnosing disease [110].

-

There are still a lot of challenges with applying deep neural networks to metagenomics problems. They are not ideal for microbial/functional composition->phenotype classification because most studies contain tens of samples (~20->40) and hundreds/thousands of features (aka species). Such underdetermined/ill-conditioned problems are still a challenge for deep neural networks that require many more training examples than features to sufficiently converge the weights on the hidden layers. Also, due to convergence issues (slowness and instability due to large neural networks modeling very large datasets [111]), taxonomic classification of reads from whole genome sequencing seems out of reach at the moment for deep neural networks -- due to only thousands of full-sequenced genomes as compared to hundreds of thousands of 16S rRNA sequences available for training.

-

However, because recurrent neural networks are showing success for base-calling (and thus removing the large error in the measurement of a pore's current signal) for the relatively new Oxford Nanopore sequencer [112], there is hope that the process of denoising->organism/function classification can be combined into one step in using powerful LSTM's. LSTM's are working miracles in raw speech signal->meaning translation [ref], and combining steps in metagenomics are not out of the question. For example, metagenomic assembly usually requires binning then assembly, but could deep neural nets accomplish both tasks in one network? Does functional/taxonomic classification need to be separate processes? The largest potential in deep learning is to learn "everything" in one complex network, with a plethora of labeled (reference) data and unlabeled (microbiome experiments) examples.

+

TODO: Add reference tags to this section Metagenomics (which refers to the study of genetic material, 16S rRNA and/or whole-genome shotgun DNA, from microbial communities) has revolutionized the study of micro-scale ecosystems within us and around us. There is increasing literature of applying machine learning in general to metagenomic analysis. In the late 2000’s, a plethora of machine learning methods were applied to classifying DNA sequencing reads to the thousands of species within a sample. An important problem is genome assembly from these mixed-organism samples. And to do that, the organisms should be “binned” before assembling. Binning methods began with many k-mer techniques [refs] and then delved into other clustering algorithms, such as self-organizing maps (SOM). Then came the taxonomic classification problem, with researchers naturally using BLAST [105], followed by other machine learning techniques such as SVMs [106], naive Bayesian classifiers [107], etc. to classify each read. Then, researchers began to use techniques that could be used to estimate relative abundances of an entire sample, instead of the precise but painstakingly slow read-by-read classification. Relative abundance estimators (a.k.a diversity profilers) are MetaPhlan[ref], (WGS)Quikr[ref], and some configurations of tools like OneCodex[ref] and LMAT[ref]. While one cannot identify which reads were mapped back to an organism using relative abundance estimators, they can be useful for faster comparative and other downstream analyses. Newer methods hope to classify reads and estimate relative abundances at faster rates [108] and as of this writing, there are more than 70 metagenomic taxonomic classifiers in existence. Besides binning and classification of species, there is functional identification and annotation of sequence reads [109]. However, the focus on taxonomic/functional annotation is just the first step. Once organisms are identified, there is the interest in understanding the interrelationship between these organisms and host/environment phenotypes [111]. One of the first attempts was a survey of supervised classification methods for microbes->phenotype classification [112], followed by similar studies that are more massive in scale [113]. There have been techniques that bypass the taxonomic classification step altogether [115], (sequence composition to phenotype classification). Also, researchers have looked into how feature selection can improve classification [116], and techniques have been proposed that are classifier-independent [Ditzler,Ditzler].

+

So, how have neural networks (NNs) been of use? Most neural networks are being used for short sequence->taxa/function classification, where there is a lot of data for training (and thus suitable for NNs). Neural networks have been applied successfully to gene annotation (e.g. Orphelia [117] and FragGeneScan [118]), which usually has plenty of training examples. Representations (similar to Word2Vec [ref] in natural language processing) for protein family classification has been introduced and classified with a skip-gram neural network [119]. Recurrent neural networks show good performance for homology and protein family identification [120]. Interestingly, Hochreiter, who invented Long Short Term Memory, delved into homology/protein family classification in 2007, and therefore, deep learning is deeply rooted in functional classification methods.

+

One of the first techniques of “de novo” genome binning used self-organizing maps, a type of NN [122]. Essinger et al. use ART, a neural network algorithm called Adaptive Resonance Theory, to cluster similar genomic fragments and showed that it has better performance than K-means. However, other methods based on interpolated Markov models [123] have performed better than these early genome binners. Also, neural networks can be slow, and therefore, have had limited use for reference-based taxonomic classification, with TAC-ELM [124] being the only NN-based algorithm to taxonomically classify massive amounts of metagenomic data. Also, neural networks can fail to perform if there are not enough training examples, which is the case with taxonomic classification (since only ~10% of estimated species have been sequenced). An initial study shows that deep neural networks have been successfully applied to taxonomic classification of 16S rRNA genes, with convolutional networks provide about 10% accuracy genus-level improvement over RNNs and even random forests [125]. However, this study performed 10-fold cross-validation on 3000 sequences in total.

+

Due to the traditionally small numbers of metagenomic samples in studies, neural network uses for classifying phenotype from microbial composition are just beginning. A standard MLP was able to classify wound severity from microbial species present in the wound [126]. Recently, multi-layer, recurrent networks (and convolutional networks) have been applied to microbiome genotype-phenotype, with Ditzler et al. being the first to associate soil samples with pH level using multi-layer perceptrons, deep-belief networks, and recursive neural networks (RNNs) [Ditzler] . Besides classifying the samples appropriately, Ditzler shows that internal phylogenetic tree nodes inferred by the networks are appropriate features representing low/high pH, which can provide additional useful information and new features for future metagenomic sample comparison. Also, an initial study has show promise of these networks for diagnosing disease [127].

+

There are still a lot of challenges with applying deep neural networks to metagenomics problems. They are not ideal for microbial/functional composition->phenotype classification because most studies contain tens of samples (~20->40) and hundreds/thousands of features (aka species). Such underdetermined/ill-conditioned problems are still a challenge for deep neural networks that require many more training examples than features to sufficiently converge the weights on the hidden layers. Also, due to convergence issues (slowness and instability due to large neural networks modeling very large datasets [128]), taxonomic classification of reads from whole genome sequencing seems out of reach at the moment for deep neural networks -- due to only thousands of full-sequenced genomes as compared to hundreds of thousands of 16S rRNA sequences available for training.

+

However, because recurrent neural networks are showing success for base-calling (and thus removing the large error in the measurement of a pore's current signal) for the relatively new Oxford Nanopore sequencer [129], there is hope that the process of denoising->organism/function classification can be combined into one step in using powerful LSTM's. LSTM's are working miracles in raw speech signal->meaning translation [ref], and combining steps in metagenomics are not out of the question. For example, metagenomic assembly usually requires binning then assembly, but could deep neural nets accomplish both tasks in one network? Does functional/taxonomic classification need to be separate processes? The largest potential in deep learning is to learn "everything" in one complex network, with a plethora of labeled (reference) data and unlabeled (microbiome experiments) examples.

Sequencing and variant calling

We have one nanopore paper in the issues and very recent work on variant calling that looks worthy of inclusion.

The impact of deep learning in treating disease and developing new treatments

There will be some overlap with the Categorize section, and we may have to determine which methods categorize individuals and which more directly match patients with treatments. The sub-section titles are merely placeholders.

Categorizing patients for clinical decision making

-

How can deep learning match patients with clinical trails, therapies, or other interventions? As an example, [113] predicts individuals who are most likely to decline during a clinical trial and benefit from the treatment.

+

How can deep learning match patients with clinical trails, therapies, or other interventions? As an example, [130] predicts individuals who are most likely to decline during a clinical trial and benefit from the treatment.

Effects of drugs on transcriptomic responses

We discussed a few papers that operate on Library of Network-Based Cellular Signatures (LINCS) gene expression data. We could briefly introduce the goals of that resource and comment on the deep learning applications. In the Issues, we had reservations about whether the improvements in expression prediction are good enough to make a practical difference in the domain and feature selection and construction.

Ligand-Based Prediction of Bioactivity

TODO: expand outline

    -
  • Short introduction to problem, related reviews, use vHTS definition from [114] (vHTS doesn't fit neatly into classic classification, regression, or ranking)
    • +
  • Short introduction to problem, related reviews, use vHTS definition from [131] (vHTS doesn't fit neatly into classic classification, regression, or ranking)
  • Introduce ligand-based approaches, hype and excitement surrounding performance of a "high-parameter" network on the Merck Kaggle challenge, cover other neural networks trained on fingerprints or descriptors as features that followed, Tox21 Data Challenge
  • Multitask networks related to the above point
    • -
  • Realistic view of where things stand today, high-parameter networks struggle with overfitting, cross validation needs to be done carefully because of temporal structure [115], low parameter networks based on chemical similarity (IRV) work very well, especially well-suited for the domain in which training data can be limited and contains few positive instances, may touch on BACE example here and other discussions of training data limitations (e.g. [116])
    • +
  • Realistic view of where things stand today, high-parameter networks struggle with overfitting, cross validation needs to be done carefully because of temporal structure [132], low parameter networks based on chemical similarity (IRV) work very well, especially well-suited for the domain in which training data can be limited and contains few positive instances, may touch on BACE example here and other discussions of training data limitations (e.g. [133])
  • "Creative experimentation" phase of the field, new ideas for representation learning and novel approaches including graph convolutions, autoencoders, one shot learning, and generative models
  • These "creative" approaches are definitely interesting but aren't necessarily outperforming existing methods, improvements on the software and reusability side could be important to help establish more rigorous benchmarking, DeepChem as example of this
  • Future outlook, what would need to happen for the "creative" approaches to overtake the current state of the art, can representation learning be improved by incorporating more information about chemical properties or even more "tasks" during training, how much will future growth depend on data versus algorithms
    • @@ -164,10 +167,10 @@

    Hardware limitations and scaling

    Several papers have stated that memory or other hardware limitations artificially restricted the number of training instances, model inputs/outputs, hidden layers, etc. Is this a general problem worth discussing or will it be solved naturally as hardware improves and/or groups move to distributed deep learning frameworks? Does hardware limit what types of problems are accessible to the average computational group, and if so, will that limit future progress? For instance, some hyperparameter search strategies are not feasible for a lab with only a couple GPUs.

    Some of this is also outlined in the Categorize section. We can decide where it best fits.

    Efficiently scaling deep learning is challenging, and there is a high computational cost (e.g., time, memory, energy) associated with training neural networks and using them for classification. As such, neural networks have only recently found widespread use [6].

    -

    Many have sought to curb the costs of deep learning, with methods ranging from the very applied (e.g., reduced numerical precision [118]) to the exotic and theoretic (e.g., training small networks to mimic large networks and ensembles [122]). The largest gains in efficiency have come from computation with graphics processing units (GPUs) [124], which excel at the matrix and vector operations so central to deep learning. The massively parallel nature of GPUs allows additional optimizations, such as accelerated mini-batch gradient descent [125]. However, GPUs also have a limited quantity of memory, making it difficult to implement networks of significant size and complexity on a single GPU or machine [124]. This restriction has sometimes forced computational biologists to use workarounds or limit the size of an analysis. For example, Chen et al. [73] aimed to infer the expression level of all genes with a single neural network, but due to memory restrictions they randomly partitioned genes into two halves and analyzed each separately. In other cases, researchers limited the size of their neural network [19]. Some have also chosen to use slower CPU implementations rather than sacrifice network size or performance [133].

    -

    Steady improvements in GPU hardware may alleviate this issue somewhat, but it is not clear whether they can occur quickly enough to keep up with the growing amount of available biological data or increasing network sizes. Much has been done to minimize the memory requirements of neural networks [134], but there is also growing interest in specialized hardware, such as field-programmable gate arrays (FPGAs) [128] and application-specific integrated circuits (ASICs). Specialized hardware promises improvements in deep learning at reduced time, energy, and memory [128]. Logically, there is less software for highly specialized hardware [136], and it could be a difficult investment for those not solely interested in deep learning. However, it is likely that such options will find increased support as they become a more popular platform for deep learning and general computation.

    -

    Distributed computing is a general solution to intense computational requirements, and has enabled many large-scale deep learning efforts. Early approaches to distributed computation [137] were not suitable for deep learning [139], but significant progress has been made. There now exist a number of algorithms [139], tools [141], and high-level libraries [144] for deep learning in a distributed environment, and it is possible to train very complex networks with limited infrastructure [146]. Besides handling very large networks, distributed or parallelized approaches offer other advantages, such as improved ensembling [147] or accelerated hyperparameter optimization [148].

    -

    Cloud computing, which has already seen adoption in genomics [150], could facilitate easier sharing of the large datasets common to biology [151], and may be key to scaling deep learning. Cloud computing affords researchers significant flexibility, and enables the use of specialized hardware (e.g., FPGAs, ASICs, GPUs) without significant investment. With such flexibility, it could be easier to address the different challenges associated with the multitudinous layers and architectures available [153]. Though many are reluctant to store sensitive data (e.g., patient electronic health records) in the cloud, secure/regulation-compliant cloud services do exist [154].

    +

    Many have sought to curb the costs of deep learning, with methods ranging from the very applied (e.g., reduced numerical precision [135]) to the exotic and theoretic (e.g., training small networks to mimic large networks and ensembles [139]). The largest gains in efficiency have come from computation with graphics processing units (GPUs) [141], which excel at the matrix and vector operations so central to deep learning. The massively parallel nature of GPUs allows additional optimizations, such as accelerated mini-batch gradient descent [142]. However, GPUs also have a limited quantity of memory, making it difficult to implement networks of significant size and complexity on a single GPU or machine [141]. This restriction has sometimes forced computational biologists to use workarounds or limit the size of an analysis. For example, Chen et al. [73] aimed to infer the expression level of all genes with a single neural network, but due to memory restrictions they randomly partitioned genes into two halves and analyzed each separately. In other cases, researchers limited the size of their neural network [19]. Some have also chosen to use slower CPU implementations rather than sacrifice network size or performance [150].

    +

    Steady improvements in GPU hardware may alleviate this issue somewhat, but it is not clear whether they can occur quickly enough to keep up with the growing amount of available biological data or increasing network sizes. Much has been done to minimize the memory requirements of neural networks [151], but there is also growing interest in specialized hardware, such as field-programmable gate arrays (FPGAs) [145] and application-specific integrated circuits (ASICs). Specialized hardware promises improvements in deep learning at reduced time, energy, and memory [145]. Logically, there is less software for highly specialized hardware [153], and it could be a difficult investment for those not solely interested in deep learning. However, it is likely that such options will find increased support as they become a more popular platform for deep learning and general computation.

    +

    Distributed computing is a general solution to intense computational requirements, and has enabled many large-scale deep learning efforts. Early approaches to distributed computation [154] were not suitable for deep learning [156], but significant progress has been made. There now exist a number of algorithms [156], tools [158], and high-level libraries [161] for deep learning in a distributed environment, and it is possible to train very complex networks with limited infrastructure [163]. Besides handling very large networks, distributed or parallelized approaches offer other advantages, such as improved ensembling [164] or accelerated hyperparameter optimization [165].

    +

    Cloud computing, which has already seen adoption in genomics [167], could facilitate easier sharing of the large datasets common to biology [168], and may be key to scaling deep learning. Cloud computing affords researchers significant flexibility, and enables the use of specialized hardware (e.g., FPGAs, ASICs, GPUs) without significant investment. With such flexibility, it could be easier to address the different challenges associated with the multitudinous layers and architectures available [170]. Though many are reluctant to store sensitive data (e.g., patient electronic health records) in the cloud, secure/regulation-compliant cloud services do exist [171].

    TODO: Write the transition once more of the Discussion section has been fleshed out.

    Code, data, and model sharing

    Reproducibiliy is important for science to progress. In the context of deep learning applied to advance human healthcare, does reproducibility have different requirements or alternative connotations? With vast hyperparameter spaces, massively heterogeneous and noisy biological data sets, and black box interpretability problems, how can we best ensure reproducible models? What might a clinician, or policy maker, need to see in a deep model in order to influence healthcare decisions? Or, is deep learning a hypothesis generation machine that requires manual validation?

    @@ -415,239 +418,290 @@

    Author contributions

    76. Shaham U, Stanton KP, Zhao J, Li H, Raddassi K, Montgomery R, Kluger Y. 2016 Removal of batch effects using distribution-matching residual networks.

    +
    +

    77. McGuffin LJ, Bryson K, Jones DT. 2000 The PSIPRED protein structure prediction server. Bioinformatics 16, 404–405. (doi:10.1093/bioinformatics/16.4.404)

    +
    +
    +

    78. Wang S, Sun S, Xu J. 2016 AUC-Maximized Deep Convolutional Neural Fields for Protein Sequence Labeling. In Machine Learning and Knowledge Discovery in Databases, pp. 1–16. Springer Nature. (doi:10.1007/978-3-319-46227-1_1)

    +
    +
    +

    79. Wang S, Sun S, Li Z, Zhang R, Xu J. 2017 Accurate De Novo Prediction of Protein Contact Map by Ultra-Deep Learning Model. PLOS Computational Biology 13, e1005324. (doi:10.1371/journal.pcbi.1005324)

    +
    +
    +

    80. Jones DT, Singh T, Kosciolek T, Tetchner S. 2014 MetaPSICOV: combining coevolution methods for accurate prediction of contacts and long range hydrogen bonding in proteins. Bioinformatics 31, 999–1006. (doi:10.1093/bioinformatics/btu791)

    +
    +
    +

    81. Weigt M, White RA, Szurmant H, Hoch JA, Hwa T. 2008 Identification of direct residue contacts in protein-protein interaction by message passing. Proceedings of the National Academy of Sciences 106, 67–72. (doi:10.1073/pnas.0805923106)

    +
    +
    +

    82. Marks DS, Colwell LJ, Sheridan R, Hopf TA, Pagnani A, Zecchina R, Sander C. 2011 Protein 3D Structure Computed from Evolutionary Sequence Variation. PLoS ONE 6, e28766. (doi:10.1371/journal.pone.0028766)

    +
    +
    +

    83. Qi Y, Oja M, Weston J, Noble WS. 2012 A Unified Multitask Architecture for Predicting Local Protein Properties. PLoS ONE 7, e32235. (doi:10.1371/journal.pone.0032235)

    +
    +
    +

    84. Heffernan R, Paliwal K, Lyons J, Dehzangi A, Sharma A, Wang J, Sattar A, Yang Y, Zhou Y. 2015 Improving prediction of secondary structure, local backbone angles, and solvent accessible surface area of proteins by iterative deep learning. Scientific Reports 5, 11476. (doi:10.1038/srep11476)

    +
    +
    +

    85. Jones DT. 1999 Protein secondary structure prediction based on position-specific scoring matrices. Journal of Molecular Biology 292, 195–202. (doi:10.1006/jmbi.1999.3091)

    +
    +
    +

    86. Wang Z, Zhao F, Peng J, Xu J. 2011 Protein 8-class secondary structure prediction using conditional neural fields. PROTEOMICS 11, 3786–3792. (doi:10.1002/pmic.201100196)

    +
    +
    +

    87. Zhou J, Troyanskaya OG. 2014 Deep supervised and convolutional generative stochastic network for protein secondary structure prediction.

    +
    +
    +

    88. Ma J, Wang S, Wang Z, Xu J. 2015 Protein contact prediction by integrating joint evolutionary coupling analysis and supervised learning. Bioinformatics 31, 3506–3513. (doi:10.1093/bioinformatics/btv472)

    +
    +
    +

    89. Di Lena P, Nagata K, Baldi P. 2012 Deep architectures for protein contact map prediction. Bioinformatics 28, 2449–2457. (doi:10.1093/bioinformatics/bts475)

    +
    +
    +

    90. Eickholt J, Cheng J. 2012 Predicting protein residue-residue contacts using deep networks and boosting. Bioinformatics 28, 3066–3072. (doi:10.1093/bioinformatics/bts598)

    +
    +
    +

    91. Skwark MJ, Raimondi D, Michel M, Elofsson A. 2014 Improved Contact Predictions Using the Recognition of Protein Like Contact Patterns. PLoS Computational Biology 10, e1003889. (doi:10.1371/journal.pcbi.1003889)

    +
    + +
    +

    93. In press. See http://www.cameo3d.org/.

    +
    -

    77. Chen L, Cai C, Chen V, Lu X. 2015 Trans-species learning of cellular signaling systems with bimodal deep belief networks. Bioinformatics 31, 3008–3015. (doi:10.1093/bioinformatics/btv315)

    +

    94. Chen L, Cai C, Chen V, Lu X. 2015 Trans-species learning of cellular signaling systems with bimodal deep belief networks. Bioinformatics 31, 3008–3015. (doi:10.1093/bioinformatics/btv315)

    -

    78. Van Valen DA et al. 2016 Deep Learning Automates the Quantitative Analysis of Individual Cells in Live-Cell Imaging Experiments. PLOS Computational Biology 12, e1005177. (doi:10.1371/journal.pcbi.1005177)

    +

    95. Van Valen DA et al. 2016 Deep Learning Automates the Quantitative Analysis of Individual Cells in Live-Cell Imaging Experiments. PLOS Computational Biology 12, e1005177. (doi:10.1371/journal.pcbi.1005177)

    -

    79. Ronneberger O, Fischer P, Brox T. 2015 U-Net: Convolutional Networks for Biomedical Image Segmentation. In Lecture Notes in Computer Science, pp. 234–241. Springer Nature. (doi:10.1007/978-3-319-24574-4_28)

    +

    96. Ronneberger O, Fischer P, Brox T. 2015 U-Net: Convolutional Networks for Biomedical Image Segmentation. In Lecture Notes in Computer Science, pp. 234–241. Springer Nature. (doi:10.1007/978-3-319-24574-4_28)

    -

    80. Buggenthin F et al. 2017 Prospective identification of hematopoietic lineage choice by deep learning. Nature Methods 14, 403–406. (doi:10.1038/nmeth.4182)

    +

    97. Buggenthin F et al. 2017 Prospective identification of hematopoietic lineage choice by deep learning. Nature Methods 14, 403–406. (doi:10.1038/nmeth.4182)

    -

    81. Eulenberg P, Koehler N, Blasi T, Filby A, Carpenter AE, Rees P, Theis FJ, Wolf FA. 2016 Deep Learning for Imaging Flow Cytometry: Cell Cycle Analysis of Jurkat Cells. (doi:10.1101/081364)

    +

    98. Eulenberg P, Koehler N, Blasi T, Filby A, Carpenter AE, Rees P, Theis FJ, Wolf FA. 2016 Deep Learning for Imaging Flow Cytometry: Cell Cycle Analysis of Jurkat Cells. (doi:10.1101/081364)

    -

    82. Pawlowski N, Caicedo JC, Singh S, Carpenter AE, Storkey A. 2016 Automating Morphological Profiling with Generic Deep Convolutional Networks. (doi:10.1101/085118)

    +

    99. Pawlowski N, Caicedo JC, Singh S, Carpenter AE, Storkey A. 2016 Automating Morphological Profiling with Generic Deep Convolutional Networks. (doi:10.1101/085118)

    -

    83. Caicedo JC, Singh S, Carpenter AE. 2016 Applications in image-based profiling of perturbations. Current Opinion in Biotechnology 39, 134–142. (doi:10.1016/j.copbio.2016.04.003)

    +

    100. Caicedo JC, Singh S, Carpenter AE. 2016 Applications in image-based profiling of perturbations. Current Opinion in Biotechnology 39, 134–142. (doi:10.1016/j.copbio.2016.04.003)

    -

    84. Bougen-Zhukov N, Loh SY, Lee HK, Loo L-H. 2016 Large-scale image-based screening and profiling of cellular phenotypes. Cytometry Part A 91, 115–125. (doi:10.1002/cyto.a.22909)

    +

    101. Bougen-Zhukov N, Loh SY, Lee HK, Loo L-H. 2016 Large-scale image-based screening and profiling of cellular phenotypes. Cytometry Part A 91, 115–125. (doi:10.1002/cyto.a.22909)

    -

    85. Grys BT, Lo DS, Sahin N, Kraus OZ, Morris Q, Boone C, Andrews BJ. 2016 Machine learning and computer vision approaches for phenotypic profiling. The Journal of Cell Biology 216, 65–71. (doi:10.1083/jcb.201610026)

    +

    102. Grys BT, Lo DS, Sahin N, Kraus OZ, Morris Q, Boone C, Andrews BJ. 2016 Machine learning and computer vision approaches for phenotypic profiling. The Journal of Cell Biology 216, 65–71. (doi:10.1083/jcb.201610026)

    -

    86. Arvaniti E, Claassen M. 2016 Sensitive detection of rare disease-associated cell subsets via representation learning. (doi:10.1101/046508)

    +

    103. Arvaniti E, Claassen M. 2016 Sensitive detection of rare disease-associated cell subsets via representation learning. (doi:10.1101/046508)

    -

    87. Angermueller C, Lee H, Reik W, Stegle O. 2016 Accurate prediction of single-cell DNA methylation states using deep learning. (doi:10.1101/055715)

    +

    104. Angermueller C, Lee H, Reik W, Stegle O. 2016 Accurate prediction of single-cell DNA methylation states using deep learning. (doi:10.1101/055715)

    -

    88. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. 1990 Basic local alignment search tool. Journal of Molecular Biology 215, 403–410. (doi:10.1016/s0022-2836(05)80360-2)

    +

    105. Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ. 1990 Basic local alignment search tool. Journal of Molecular Biology 215, 403–410. (doi:10.1016/s0022-2836(05)80360-2)

    -

    89. McHardy AC, Martín HG, Tsirigos A, Hugenholtz P, Rigoutsos I. 2006 Accurate phylogenetic classification of variable-length DNA fragments. Nature Methods 4, 63–72. (doi:10.1038/nmeth976)

    +

    106. McHardy AC, Martín HG, Tsirigos A, Hugenholtz P, Rigoutsos I. 2006 Accurate phylogenetic classification of variable-length DNA fragments. Nature Methods 4, 63–72. (doi:10.1038/nmeth976)

    -

    90. Rosen GL, Reichenberger ER, Rosenfeld AM. 2010 NBC: the Naive Bayes Classification tool webserver for taxonomic classification of metagenomic reads. Bioinformatics 27, 127–129. (doi:10.1093/bioinformatics/btq619)

    +

    107. Rosen GL, Reichenberger ER, Rosenfeld AM. 2010 NBC: the Naive Bayes Classification tool webserver for taxonomic classification of metagenomic reads. Bioinformatics 27, 127–129. (doi:10.1093/bioinformatics/btq619)

    -

    91. Vervier K, Mahé P, Tournoud M, Veyrieras J-B, Vert J-P. 2015 Large-scale machine learning for metagenomics sequence classification. Bioinformatics 32, 1023–1032. (doi:10.1093/bioinformatics/btv683)

    +

    108. Vervier K, Mahé P, Tournoud M, Veyrieras J-B, Vert J-P. 2015 Large-scale machine learning for metagenomics sequence classification. Bioinformatics 32, 1023–1032. (doi:10.1093/bioinformatics/btv683)

    -

    92. Yok NG, Rosen GL. 2011 Combining gene prediction methods to improve metagenomic gene annotation. BMC Bioinformatics 12, 20. (doi:10.1186/1471-2105-12-20)

    +

    109. Yok NG, Rosen GL. 2011 Combining gene prediction methods to improve metagenomic gene annotation. BMC Bioinformatics 12, 20. (doi:10.1186/1471-2105-12-20)

    -

    93. Soueidan H, Nikolski M. 2017 Machine learning for metagenomics: methods and tools. Metagenomics 1. (doi:10.1515/metgen-2016-0001)

    +

    110. Soueidan H, Nikolski M. 2017 Machine learning for metagenomics: methods and tools. Metagenomics 1. (doi:10.1515/metgen-2016-0001)

    -

    95. Knights D, Costello EK, Knight R. 2011 Supervised classification of human microbiota. FEMS Microbiology Reviews 35, 343–359. (doi:10.1111/j.1574-6976.2010.00251.x)

    +

    112. Knights D, Costello EK, Knight R. 2011 Supervised classification of human microbiota. FEMS Microbiology Reviews 35, 343–359. (doi:10.1111/j.1574-6976.2010.00251.x)

    -

    96. Statnikov A et al. 2013 A comprehensive evaluation of multicategory classification methods for microbiomic data. Microbiome 1, 11. (doi:10.1186/2049-2618-1-11)

    +

    113. Statnikov A et al. 2013 A comprehensive evaluation of multicategory classification methods for microbiomic data. Microbiome 1, 11. (doi:10.1186/2049-2618-1-11)

    -

    97. Pasolli E, Truong DT, Malik F, Waldron L, Segata N. 2016 Machine Learning Meta-analysis of Large Metagenomic Datasets: Tools and Biological Insights. PLOS Computational Biology 12, e1004977. (doi:10.1371/journal.pcbi.1004977)

    +

    114. Pasolli E, Truong DT, Malik F, Waldron L, Segata N. 2016 Machine Learning Meta-analysis of Large Metagenomic Datasets: Tools and Biological Insights. PLOS Computational Biology 12, e1004977. (doi:10.1371/journal.pcbi.1004977)

    -

    98. Ding X, Cheng F, Cao C, Sun X. 2015 DectICO: an alignment-free supervised metagenomic classification method based on feature extraction and dynamic selection. BMC Bioinformatics 16. (doi:10.1186/s12859-015-0753-3)

    +

    115. Ding X, Cheng F, Cao C, Sun X. 2015 DectICO: an alignment-free supervised metagenomic classification method based on feature extraction and dynamic selection. BMC Bioinformatics 16. (doi:10.1186/s12859-015-0753-3)

    -

    99. Liu Z, Chen D, Sheng L, Liu AY. 2014 Correction: Class Prediction and Feature Selection with Linear Optimization for Metagenomic Count Data. PLoS ONE 9, e97958. (doi:10.1371/journal.pone.0097958)

    +

    116. Liu Z, Chen D, Sheng L, Liu AY. 2014 Correction: Class Prediction and Feature Selection with Linear Optimization for Metagenomic Count Data. PLoS ONE 9, e97958. (doi:10.1371/journal.pone.0097958)

    -

    100. Hoff KJ, Lingner T, Meinicke P, Tech M. 2009 Orphelia: predicting genes in metagenomic sequencing reads. Nucleic Acids Research 37, W101–W105. (doi:10.1093/nar/gkp327)

    +

    117. Hoff KJ, Lingner T, Meinicke P, Tech M. 2009 Orphelia: predicting genes in metagenomic sequencing reads. Nucleic Acids Research 37, W101–W105. (doi:10.1093/nar/gkp327)

    -

    101. Rho M, Tang H, Ye Y. 2010 FragGeneScan: predicting genes in short and error-prone reads. Nucleic Acids Research 38, e191–e191. (doi:10.1093/nar/gkq747)

    +

    118. Rho M, Tang H, Ye Y. 2010 FragGeneScan: predicting genes in short and error-prone reads. Nucleic Acids Research 38, e191–e191. (doi:10.1093/nar/gkq747)

    -

    102. Asgari E, Mofrad MRK. 2015 Continuous Distributed Representation of Biological Sequences for Deep Proteomics and Genomics. PLOS ONE 10, e0141287. (doi:10.1371/journal.pone.0141287)

    +

    119. Asgari E, Mofrad MRK. 2015 Continuous Distributed Representation of Biological Sequences for Deep Proteomics and Genomics. PLOS ONE 10, e0141287. (doi:10.1371/journal.pone.0141287)

    -

    103. Hochreiter S, Heusel M, Obermayer K. 2007 Fast model-based protein homology detection without alignment. Bioinformatics 23, 1728–1736. (doi:10.1093/bioinformatics/btm247)

    +

    120. Hochreiter S, Heusel M, Obermayer K. 2007 Fast model-based protein homology detection without alignment. Bioinformatics 23, 1728–1736. (doi:10.1093/bioinformatics/btm247)

    -

    104. Sønderby SK, Sønderby CK, Nielsen H, Winther O. 2015 Convolutional lstm networks for subcellular localization of proteins. (doi:10.1007/978-3-319-21233-3_6)

    +

    121. Sønderby SK, Sønderby CK, Nielsen H, Winther O. 2015 Convolutional lstm networks for subcellular localization of proteins. (doi:10.1007/978-3-319-21233-3_6)

    -

    106. Kelley DR, Salzberg SL. 2010 Clustering metagenomic sequences with interpolated Markov models. BMC Bioinformatics 11, 544. (doi:10.1186/1471-2105-11-544)

    +

    123. Kelley DR, Salzberg SL. 2010 Clustering metagenomic sequences with interpolated Markov models. BMC Bioinformatics 11, 544. (doi:10.1186/1471-2105-11-544)

    -

    107. RASHEED Z, RANGWALA H. 2012 METAGENOMIC TAXONOMIC CLASSIFICATION USING EXTREME LEARNING MACHINES. Journal of Bioinformatics and Computational Biology 10, 1250015. (doi:10.1142/s0219720012500151)

    +

    124. RASHEED Z, RANGWALA H. 2012 METAGENOMIC TAXONOMIC CLASSIFICATION USING EXTREME LEARNING MACHINES. Journal of Bioinformatics and Computational Biology 10, 1250015. (doi:10.1142/s0219720012500151)

    -

    109. Chudobova D et al. 2015 Influence of microbiome species in hard-to-heal wounds on disease severity and treatment duration. The Brazilian Journal of Infectious Diseases 19, 604–613. (doi:10.1016/j.bjid.2015.08.013)

    +

    126. Chudobova D et al. 2015 Influence of microbiome species in hard-to-heal wounds on disease severity and treatment duration. The Brazilian Journal of Infectious Diseases 19, 604–613. (doi:10.1016/j.bjid.2015.08.013)

    -

    111. Bengio Y, Boulanger-Lewandowski N, Pascanu R. 2012 Advances in optimizing recurrent networks.

    +

    128. Bengio Y, Boulanger-Lewandowski N, Pascanu R. 2012 Advances in optimizing recurrent networks.

    -

    112. Boža V, Brejová B, Vinař T. 2016 DeepNano: Deep recurrent neural networks for base calling in minion nanopore reads.

    +

    129. Boža V, Brejová B, Vinař T. 2016 DeepNano: Deep recurrent neural networks for base calling in minion nanopore reads.

    -

    113. Ithapu VK, Singh V, Okonkwo OC, Chappell RJ, Dowling NM, Johnson SC. 2015 Imaging-based enrichment criteria using deep learning algorithms for efficient clinical trials in mild cognitive impairment. Alzheimer’s & Dementia 11, 1489–1499. (doi:10.1016/j.jalz.2015.01.010)

    +

    130. Ithapu VK, Singh V, Okonkwo OC, Chappell RJ, Dowling NM, Johnson SC. 2015 Imaging-based enrichment criteria using deep learning algorithms for efficient clinical trials in mild cognitive impairment. Alzheimer’s & Dementia 11, 1489–1499. (doi:10.1016/j.jalz.2015.01.010)

    -

    114. Swamidass SJ, Azencott C-A, Lin T-W, Gramajo H, Tsai S-C, Baldi P. 2009 Influence Relevance Voting: An Accurate And Interpretable Virtual High Throughput Screening Method. Journal of Chemical Information and Modeling 49, 756–766. (doi:10.1021/ci8004379)

    +

    131. Swamidass SJ, Azencott C-A, Lin T-W, Gramajo H, Tsai S-C, Baldi P. 2009 Influence Relevance Voting: An Accurate And Interpretable Virtual High Throughput Screening Method. Journal of Chemical Information and Modeling 49, 756–766. (doi:10.1021/ci8004379)

    -

    115. Kearnes S, Goldman B, Pande V. 2016 Modeling industrial admet data with multitask networks.

    +

    132. Kearnes S, Goldman B, Pande V. 2016 Modeling industrial admet data with multitask networks.

    -

    116. Altae-Tran H, Ramsundar B, Pappu AS, Pande V. 2016 Low data drug discovery with one-shot learning.

    +

    133. Altae-Tran H, Ramsundar B, Pappu AS, Pande V. 2016 Low data drug discovery with one-shot learning.

    -

    117. Ragoza M, Hochuli J, Idrobo E, Sunseri J, Koes DR. 2016 Protein-ligand scoring with convolutional neural networks.

    +

    134. Ragoza M, Hochuli J, Idrobo E, Sunseri J, Koes DR. 2016 Protein-ligand scoring with convolutional neural networks.

    -

    118. Gupta S, Agrawal A, Gopalakrishnan K, Narayanan P. 2015 Deep learning with limited numerical precision.

    +

    135. Gupta S, Agrawal A, Gopalakrishnan K, Narayanan P. 2015 Deep learning with limited numerical precision.

    -

    119. Courbariaux M, Bengio Y, David J-P. 2014 Training deep neural networks with low precision multiplications.

    +

    136. Courbariaux M, Bengio Y, David J-P. 2014 Training deep neural networks with low precision multiplications.

    -

    120. Sa CD, Zhang C, Olukotun K, Ré C. 2015 Taming the wild: A unified analysis of hogwild!-style algorithms.

    +

    137. Sa CD, Zhang C, Olukotun K, Ré C. 2015 Taming the wild: A unified analysis of hogwild!-style algorithms.

    -

    121. Hubara I, Courbariaux M, Soudry D, El-Yaniv R, Bengio Y. 2016 Quantized neural networks: Training neural networks with low precision weights and activations.

    +

    138. Hubara I, Courbariaux M, Soudry D, El-Yaniv R, Bengio Y. 2016 Quantized neural networks: Training neural networks with low precision weights and activations.

    -

    122. Ba LJ, Caruana R. 2013 Do deep nets really need to be deep?

    +

    139. Ba LJ, Caruana R. 2013 Do deep nets really need to be deep?

    -

    123. Hinton G, Vinyals O, Dean J. 2015 Distilling the knowledge in a neural network.

    +

    140. Hinton G, Vinyals O, Dean J. 2015 Distilling the knowledge in a neural network.

    -

    124. Raina R, Madhavan A, Ng AY. 2009 Large-scale deep unsupervised learning using graphics processors. In Proceedings of the 26th Annual International Conference on Machine Learning - ICML ’09, Association for Computing Machinery (ACM). (doi:10.1145/1553374.1553486)

    +

    141. Raina R, Madhavan A, Ng AY. 2009 Large-scale deep unsupervised learning using graphics processors. In Proceedings of the 26th Annual International Conference on Machine Learning - ICML ’09, Association for Computing Machinery (ACM). (doi:10.1145/1553374.1553486)

    -

    126. Seide F, Fu H, Droppo J, Li G, Yu D. 2014 On parallelizability of stochastic gradient descent for speech DNNS. In 2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Institute of Electrical and Electronics Engineers (IEEE). (doi:10.1109/icassp.2014.6853593)

    +

    143. Seide F, Fu H, Droppo J, Li G, Yu D. 2014 On parallelizability of stochastic gradient descent for speech DNNS. In 2014 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Institute of Electrical and Electronics Engineers (IEEE). (doi:10.1109/icassp.2014.6853593)

    -

    127. Hadjis S, Abuzaid F, Zhang C, Ré C. 2015 Caffe con troll: Shallow ideas to speed up deep learning.

    +

    144. Hadjis S, Abuzaid F, Zhang C, Ré C. 2015 Caffe con troll: Shallow ideas to speed up deep learning.

    -

    128. Edwards C. 2015 Growing pains for deep learning. Communications of the ACM 58, 14–16. (doi:10.1145/2771283)

    +

    145. Edwards C. 2015 Growing pains for deep learning. Communications of the ACM 58, 14–16. (doi:10.1145/2771283)

    -

    129. Su H, Chen H. 2015 Experiments on parallel training of deep neural network using model averaging.

    +

    146. Su H, Chen H. 2015 Experiments on parallel training of deep neural network using model averaging.

    -

    130. Li M, Zhang T, Chen Y, Smola AJ. 2014 Efficient mini-batch training for stochastic optimization. In Proceedings of the 20th ACM SIGKDD international conference on Knowledge discovery and data mining - KDD ’14, Association for Computing Machinery (ACM). (doi:10.1145/2623330.2623612)

    +

    147. Li M, Zhang T, Chen Y, Smola AJ. 2014 Efficient mini-batch training for stochastic optimization. In Proceedings of the 20th ACM SIGKDD international conference on Knowledge discovery and data mining - KDD ’14, Association for Computing Machinery (ACM). (doi:10.1145/2623330.2623612)

    -

    132. Gómez-Bombarelli R, Duvenaud D, Hernández-Lobato JM, Aguilera-Iparraguirre J, Hirzel TD, Adams RP, Aspuru-Guzik A. 2016 Automatic chemical design using a data-driven continuous representation of molecules.

    +

    149. Gómez-Bombarelli R, Duvenaud D, Hernández-Lobato JM, Aguilera-Iparraguirre J, Hirzel TD, Adams RP, Aspuru-Guzik A. 2016 Automatic chemical design using a data-driven continuous representation of molecules.

    -

    133. Hamanaka M, Taneishi K, Iwata H, Ye J, Pei J, Hou J, Okuno Y. 2016 CGBVS-DNN: Prediction of Compound-protein Interactions Based on Deep Learning. Molecular Informatics 36, 1600045. (doi:10.1002/minf.201600045)

    +

    150. Hamanaka M, Taneishi K, Iwata H, Ye J, Pei J, Hou J, Okuno Y. 2016 CGBVS-DNN: Prediction of Compound-protein Interactions Based on Deep Learning. Molecular Informatics 36, 1600045. (doi:10.1002/minf.201600045)

    -

    134. Chetlur S, Woolley C, Vandermersch P, Cohen J, Tran J, Catanzaro B, Shelhamer E. 2014 CuDNN: Efficient primitives for deep learning.

    +

    151. Chetlur S, Woolley C, Vandermersch P, Cohen J, Tran J, Catanzaro B, Shelhamer E. 2014 CuDNN: Efficient primitives for deep learning.

    -

    135. Chen W, Wilson JT, Tyree S, Weinberger KQ, Chen Y. 2015 Compressing neural networks with the hashing trick.

    +

    152. Chen W, Wilson JT, Tyree S, Weinberger KQ, Chen Y. 2015 Compressing neural networks with the hashing trick.

    -

    136. Lacey G, Taylor GW, Areibi S. 2016 Deep learning on fpgas: Past, present, and future.

    +

    153. Lacey G, Taylor GW, Areibi S. 2016 Deep learning on fpgas: Past, present, and future.

    -

    137. Dean J, Ghemawat S. 2008 MapReduce. Communications of the ACM 51, 107. (doi:10.1145/1327452.1327492)

    +

    154. Dean J, Ghemawat S. 2008 MapReduce. Communications of the ACM 51, 107. (doi:10.1145/1327452.1327492)

    -

    138. Low Y, Bickson D, Gonzalez J, Guestrin C, Kyrola A, Hellerstein JM. 2012 Distributed GraphLab. Proceedings of the VLDB Endowment 5, 716–727. (doi:10.14778/2212351.2212354)

    +

    155. Low Y, Bickson D, Gonzalez J, Guestrin C, Kyrola A, Hellerstein JM. 2012 Distributed GraphLab. Proceedings of the VLDB Endowment 5, 716–727. (doi:10.14778/2212351.2212354)

    -

    141. Moritz P, Nishihara R, Stoica I, Jordan MI. 2015 SparkNet: Training deep networks in spark.

    +

    158. Moritz P, Nishihara R, Stoica I, Jordan MI. 2015 SparkNet: Training deep networks in spark.

    -

    142. Meng X et al. 2015 MLlib: Machine learning in apache spark.

    +

    159. Meng X et al. 2015 MLlib: Machine learning in apache spark.

    -

    144. In press. See https://github.com/fchollet/keras.

    +

    161. In press. See https://github.com/fchollet/keras.

    -

    147. Sun S, Chen W, Liu T-Y. 2016 Ensemble-compression: A new method for parallel training of deep neural networks.

    +

    164. Sun S, Chen W, Liu T-Y. 2016 Ensemble-compression: A new method for parallel training of deep neural networks.

    -

    150. Schatz MC, Langmead B, Salzberg SL. 2010 Cloud computing and the DNA data race. Nature Biotechnology 28, 691–693. (doi:10.1038/nbt0710-691)

    +

    167. Schatz MC, Langmead B, Salzberg SL. 2010 Cloud computing and the DNA data race. Nature Biotechnology 28, 691–693. (doi:10.1038/nbt0710-691)

    -

    151. Muir P et al. 2016 The real cost of sequencing: scaling computation to keep pace with data generation. Genome Biology 17. (doi:10.1186/s13059-016-0917-0)

    +

    168. Muir P et al. 2016 The real cost of sequencing: scaling computation to keep pace with data generation. Genome Biology 17. (doi:10.1186/s13059-016-0917-0)

    -

    152. Stein LD. 2010 The case for cloud computing in genome informatics. Genome Biology 11, 207. (doi:10.1186/gb-2010-11-5-207)

    +

    169. Stein LD. 2010 The case for cloud computing in genome informatics. Genome Biology 11, 207. (doi:10.1186/gb-2010-11-5-207)

    -

    153. Krizhevsky A. 2014 One weird trick for parallelizing convolutional neural networks.

    +

    170. Krizhevsky A. 2014 One weird trick for parallelizing convolutional neural networks.

    -

    154. Armbrust M et al. 2010 A view of cloud computing. Communications of the ACM 53, 50. (doi:10.1145/1721654.1721672)

    +

    171. Armbrust M et al. 2010 A view of cloud computing. Communications of the ACM 53, 50. (doi:10.1145/1721654.1721672)

    diff --git a/index.html.ots b/index.html.ots index d878a06fac06955f53462c2519c2e3cc486c3bdd..807bd983d6d8fc7c68199ae9c105446d0e480879 100644 GIT binary patch delta 206 zcmV;<05Sj10?z`FA%7TUuoJ<-|36_6(;XQk^#oHI6NPn*!d;l+c|;P+ojULk@rlY! zy=CDlW1f)tkGL-72=E|lp8?JXG?2qL^$-Mn`wJ3y9d5)bj!PYsuZ#HhNW41;@F006 zA~Rxy^GP_!a(@;Fi_eweg#*tC1$_UZS*ONz&0z&porv>`ve26&in%V6*jMN2P9 z9Aovmho2|BLevhE2=E{t65tLVo!+SZQlzZ6-8Zno)vfv-u!TE`DYd@F0Ma zTsEnMGGCWbscf>dC%pQkLI{=w>|jhtS2;`qP38#y@DK(th8_?TusjJn)@LNG(A(|^ z@dQ}sdXex5eT=k<^n>!^k-j(~?dwY0uW@G$c($n%P(rxi2=N41=X#Ow2)wsqoUbLj I?~@k+I0ztQ@&Et;