From 765f15bc60cc293ced58270455bbe34d3d52dbef Mon Sep 17 00:00:00 2001
From: lilithzhou <80816848+lilith-zy@users.noreply.github.com>
Date: Wed, 18 Aug 2021 17:54:37 +0800
Subject: [PATCH 01/35] Update WeChat QR Code (#1154)

* Update wx_group.png

* Update wx_group.png
---
 docs/images/wx_group.png | Bin 68003 -> 204411 bytes
 1 file changed, 0 insertions(+), 0 deletions(-)

diff --git a/docs/images/wx_group.png b/docs/images/wx_group.png
index d73a7e9254f3bedfba5c6274028cf0e848f94c9c..60259dd4e0a11035b02266b3f67b35748518c5c0 100644
GIT binary patch
literal 204411
zcmd?R30#bA-#>hg!Jsm!Nb4Y!C6SV)I<u7|WsSmIBo#8KG^jaKw2(%&h!&!<CT%E<
zX|*LSN<}l3rBphHOlQr^Iqz?}?(2T8`&#bjey{uazt8`9-hAjZ%Q44soWJG!{r-N(
z$luC^i0X1%ds~D;L6Eob2az`;!AsfRI}pUt5t)i0$S`CGMGc|C6$Snv6m4Y4z%_y_
zr|A6sdNsx9Z?74IAmQG~;J>}b6@DvzU}VL+|Mq)O(x6{nLA{hj{rh#a;-&H~WUjLp
z%a7&j#o9Y=_Kcax+(q_|gB2sf&%pJhflJjIy{LLFa({4p@bC+r#3GKoWssTu;>Aqo
z)lRndE0zu{YS6^3d-v{EyxC`u-~QEhi^gr(=sJ$pjL_hZ3Ni*6x7FQu??Pwi6^b!`
z`8_cH-~SQU4$K{?Hdc)FT>D(eBuv$D?nU)Cm0w1O{QYHrd@*9Xhp#(4C568p?tAz9
z!A5L{Yl8s4y^8AyxYpPYlY;BZifhlGukR_YxBYxwJn)*es~5v-6gxT2-D@k1K_TFJ
z`nF#^?)j_hzkM#e1W{Vz?YnpP)`R1mXDj~uU;E?d!G?7p$YR#sgZsTbz5K>4gkAC&
zXTQgN#`JNfGiO;Kh+@xw{?zZQ2KJIN=H-7p7AimxE1p~~&-;(Zmh%uK1Kv5};eR}4
zG#5ceDIv(yhkLi~-#Rd!!Y&8x0Bil0iXbW*5k&npyvNOV$86Jq=PAmI5M)S`T>gDB
zg3!(&i0qVHF1ad~%aRd<`VK*!?2$h~Mh&AzQ!h~|qmV(PDAZ9Dc`?F(?HxiH`1yI_
z@c%*7!9#}9h7BK~1TV-|MFvr*)Io!(Lxw1flM)2aBZEf`Q5!#N;ZXHeTWJ&aX_y{~
zxiZXPQ3ikX>Pmd#>}~r)hL0GdIaW)1(qu!UDO1hN=UB|0H-GVxrM7m<?3b@uyKcP;
zEZ2Rz#|}>~@14GW{s#is2M>iFJr))o5qbRdnX~84$6mM?cQyIi^_0}K8#nJ~W@YE(
zK6v=3u&B7CwCw4#=T+4;Z))q_);Bb_w0`*b>2q8A7ty!xJwL>~L|?yRT@-}+^RnRI
zpI7#qb-^Z31`Qre9ZXZKi!vxcvG7rYhm4;!RBhoZ+SYyQ6HJc`(^wR9C4)cQVD@T!
z^tSz#BgRZLYn~)hENx(A|JuSr{?f|+zOesTS0l0*y#3$%k;5SW-U#q+`FjXve)0xH
znMwg;qK-mPgd~lSq(Q+WkpB(8<gfpCEvNE-*K(@<cP*#-KbAu!-pP@W4vb&I%PNzq
zeHNF8iuVzY_=)E5lhRoxMc4@9X@|Z^-kat(L&SO1XZv=ml^TXWjks$t@1311BHb)U
zB5Ke*+<ps2uz;8)V#txPNmSCnS@sBRd&i+-!m?sax-FPaoLnq!lOxId2o*UJ$`~a_
zM8^@d#hl)lOr9(gT9C?tc5<ZD2>D+#Ao>~;8xWt<$!hez%M&`cX==5ZOtY9d##*Os
zi)pRdok`b&<`hcS+H3DRJo&=#6~ml&o2oQDOrqaNqY}H|uH?xPeg!W(K&sv$?j{~f
zrc37%Q}8n-32H>3AFK1dZ*lIU8rKqm_+)H@hh(|yqSq=XzoitiJ8FZ`*}UIxEmblM
zMuqL?J94IC{I%%zbTw%Z@lePcPB4mHO^V~0{8Qdr=yBqpISzW(Gez&7);&KR`R?*e
zD(gDJX?jhu8G>+V82<?aUzd17fZJEpcQ<R6RD?DS!2`JFi4mxpiB#vt2KKANZPV{e
z_wi%<4}O@lYw$tku8GHnZO!#Tut}ImJG`q>j@Y*1I&x&FwSgRI)u8mhkRyZFf?=FT
zcRCSM?79R(LA_@Sm(M%4MFo$IkZQKF)HZb_jLxs!Q9J$-UViRQ`R!SgKG;)^?}$IY
zdZiF4L9pE#6p^7+BUF4r<kHEKsBo$TS+HQU&4$HWrV%E1Q*WuOe&pT!%69E@T{%}$
zU5&FPjki|bGu@o0KKA`X+u2W(=t7tO|Jg3R{RFAf2|HCI^@D#OBsQo9dJy}BDEMOG
zHGLhjRE`{-+PXTiOO7<69b}`JsAk2=k?GR=K_VMQwI0IJ!1#eYVVWEng9*>_GBu03
zPbLnt9Mw$M+tp-DoMD8r%mic0gN7V?eB$l(XAh3uUiD#ix6gyExf&E{gd8c}$;h>n
z>2RO0q+<!L9EozKePK#R<&h(I%5Gx9tAbpt2m^cY<YPz4QFv@r7E@!VOT*`5-R2kl
zPk8$+FuLngH8?-=rzz8xt#>S&tLuIT!B}f4B016}lPn_6h@E6nJTG3u0)GF>i|On{
zVQ*28OR4J&?(O^=*S~1isyCGBWq*I7{DOKSY`VM7{h_mL7@vR0kx^-!_l%xBD20c+
zx1i)K8=BPiio_4Ad>|bPe#C{j6D0`ak8B#je!nWpf8U6hYa-5)Bht~HExppwT=8Kd
zPvqN5KN6@XwDN53Ej~dX+Mc7YoX~%GTAt1weV>J2_qfW`7S4Y*)<(mp(S}iAhBzzy
zUxU(jNsbI*%d#!~!DstrN3Ca&HG+|r7D+^A36K8jgzVB4;=IF!ti!ynX@X1#acq{$
z<=xDSb4EV)q7u9Rt?=H((ZTp16LWf_$O#0E-z&)EJ;^(4GMcC>EYQw>5b7)%o}jcj
zU~@***U3lCC@V6S`gM<W&pi|+RJUQYbu`~uxp`Mn0;AS*$?+h!B~{t;lDU%%&Xa{o
zH%1NT8*!(sd-@7_WklWc3#vq;Oo$n%RQHI)3FFU{BYQTxh<_s1;D?)8Cj+N!#RX3|
zrd&6XPgBOoyS)Wdt3T!C<!rwnne=TOJEYL}!HvF52*rAG<oB`ozdX%R#V5#NYvc%R
z3|Y;<*RmStVD)7!6@?~Y`or*Iriyf3>-gL3Mcpo!)><@Z44?Ek!=ddoW5VjUr#|o1
zGonfi2C#Hrg*cBWy2GgeA1-q`S)g0p>F;y6IiQH8MpW2W`%jLUZjzqmrZpoeC(kQ$
z0^cBQ;|P_v?FmLyspyX~P8x0})!!xA%f3{aFbaQP*V%<n)nTf#lWi(z><D)jjXFOq
z$ZcKLy|@jb3$vc-1}L{DSWx0%bL^PZj#zy|oO#sJ-d|(ydv<ss$*7yLHR6nhg?7#v
zZ_I4fn4+<*r<-GjYc(X#El2z{C?s2uEs*L7Bs<6oK~{Ge8p6>fsQ8+7uQ{4hdyz?~
zU%bQWg1RbFT(-2}u-A+kwMHwp9UQfInQB+XihVYWd4Cj${u^EPzcH-P9*{2uI9s#p
zy6BepC@~A?Zk8n$3~PU6QOu3hSHt(*z_wy7Lmy{m)crWI*}CP^8;_&5p2z3aw~jn;
zo;ojy-m~`C+)Q5cE5s?i#}kSeeajyw%1*Ju8U6Og2xfHlZ_(wqzi5(u5N81-eAGrW
zT;L5fIns79joPz57=5Zvj`WoyXUoXyMCJ%N(ta}-MbfCUC8)5R)o07)Vx5oK5UiGr
zK*&b}ED?>O6IM)e#R;*y9J&AfwjB8~%g2WC>7)&#DqX7WBS)riiH&k(dvc3_jLT{O
zLv)>wNMj%?;&SBqIF!YcjO))O6ME*uQ<|=S#63|Sk1%DVcRlRE^$-lpqDT*5{5Hln
zZh8nN-tv<isa~>{Lb|V|$TVPOx6yAG<Vdg_5l;<V49~60MzFg<2#1OBccLFhE7IaA
z)+kupmUtURON=4_|HBqx*DIu2tK`V!rIM9c{lQA!kFkItJITLCFa#BTX7()(@3Q!T
zr;_nrROyj41#h-;7n(?8(NKnXy|o_pWoz<ln&crtBbPW0!QNhPu8_^Hcc!`h@_CGp
z(;&12V$YdUd)OX^u#UVqBaPa(@0Ytw|I$q=e}$WIs|l+gs>tEjVw`FH2?Llh5Z*il
zyz6Wr=PgsbFfd4tq(s6@EWN0&3t(~2))R0)Je2rY|7q_m3i%Zq#`|za3IX^2L1DVg
zw&yuAu;Fw%F@-0+rYYtt?p~`vw%~;gL+R(V^bflNdqk__-J?kpx;wAK$e){2$!q;V
z2=?&(ulA27fAgQ3B}Jx;@ij1EXai3=Mv2eRmb#||22^uYO<)eAy|!qbe<iVbRO{yA
zvuhbyqdDstODAH9LGMzX^OjMVD(x#nx^PqruiuJC$C_1cU;~_I7<aTBsq7*S%aLI|
z4hpmbK}K~6JDudRiUh>soy-Bk6k<?<^&IBt0|?Vutw70uC`ReSRwg&%y`*XbATR~h
z8W$mA%Jc={JUpz1HvmQk+Z_id5lgKAw#|X%|4gbTSh_%1&28?Mj?Xq3Kr+9b*|R#H
z@ELEkSDb$K?Zig6qVFdS>}&jI55ujmDHj&L+Lf$2S(S2X@jW%$*_Wes1fyY0J;Eb@
zbStLCHvhV`gY8hl8IQl%GIeZiSC-zE@5?QwhM!fxw8m)h^KgEf7xiosUEB(u1T^ot
z8MZB5GOmdZ`!L&?)&^)z55?Cy5$3VQ415*6aX|~W%gsZw+Dkw;CMW1kOg%Oue@akV
zc`CM}A?U)_yfDLP=Irjh?Div3_h_HXez(bFDkeO^!yO(kCWoVvSwyHfu!du_QR;HZ
zO+?;kUDcvwJk@7{k>)DLkf)8W(mE?!C$IQP++oAG2Y{5s_`7(*hy6~{QN5dpX~{d8
zRDZ^a&n@J+yI=M>x~DgK71?PUH)qdUmU(WrRz%Z?#MmokB7~!k6|?l=)+SFB(MZih
z0dhn)N!gJlzMc!j=FzwlVryJ*EJYUc74wXwm)?*v@4Qcq`bnhhnx?O(J$?NF{q5(g
zmb?y1di5eZ)TdP|X{6`xZZ0PdQv^fd+s7C={EWnty+(YOCrjz!M&^(@c<c$OX_H14
z{`hOG`3v^ds)qVKOW##a+-b~twc*OVptoaTfpv}uXBh+@Huh1Kkt>j$3UZSpPKO{=
zow$P>vH+__&=M9HkP2sPS=x+qhQ+8J;lj7Sk-uYmwV=4+{rTuzpp8|=8TslOy7yL$
z%%aGi(+N|95;jA0X^Vcl99cAAbWY#VHMPN-zgq%B^jGZ@ZTY?D(7tVj*s??xnvM!D
zCFanhc^*35dX0h09^hNfHOn0HUA$9Yq%1yU-nenb5(n>@YU4v+G}u!Q_#O#1qJDoB
zj9%|13|LZ~$9FKkYXL%3x1NueEcV8RU%|tp1$@UVO~n3gm(!xtm3Ipoa>s}sU5(GP
z%nTL8Au@<!5x;dR#_Is7Y8^Q&{3dKM!>P>RhICIgX#y@RB%h253SacFl6N1?IMkQu
zwd8yL*G9V+?-GtvWVF{bQ9T6MB<#6$!EiZJe=t;zIF53r^__I4^#H0HdJmk&outZ<
zw>L!KHID0RDCEN)N*_#O2+*#xYk{YT7v_U+T|eZHV8@pt2Y4!h`U2Gg12*-^6!xRT
zL~CimX#CJl<*=n0EcIPE^QShMxldl2V{vs`ut$iG7Lebsdx}Wq)1q}hj-6|h<ycyV
ze5-0`8;{Sqb|C1gM(m!z(D*!2UG5<S?D;Q-D|+i1ahvSI40Ky}!{^qz&P@LncA)n@
z+@UOhGa|NRj+pB{GN(4@fD7gmUFDQ81k2g|mcOk2vab!p@HOp+f^k4V3w1XK-Zdq+
zGpL!{kW(bSRAsH@Mw}_(Ovj%StGf$(A6=_0Z$5Er*N%r4dy6aYJ(Ae`6#Q|H@bevn
za|%G-@1iKZy@(uk5(YnC!xP3}RE|*~qo!bbec-0p7DH}l{?53PXs$I^|MK~`O^^27
zXYU!MZR=}zF<%*(b>^}Wt;<K@{EDEFJs)Um?-F9l?VZda+4K{wtwV3!ZI4;(e%Gz7
z%x=DMYD3h0v(AEm_+EsWSI2EZNDOVU-~xE5Jk5rCuJm|;9Eo<Ob=XNq3CNLVm8kGG
zBkO#g#M)bsnQ{Za>ZAE2F7c#Ogh`--_pOro?G;(|N?uwQ%;E!Ig%<dJ3&vKSrSx@!
zQ7n~3?14aW70uw5(Ce4KC>6^Q`V8yw@e@;zf4x1a=1R33$(%t}1-LF!K7EiE5WKIB
zb{)bn^Dnz+M?R#np)TxH{+0!}OH`t`^0kRe;g%V=ATq8zaMTTh&38j@*MCW=-}GdV
zS<CC<v6rU=UjFGdtv91tHoYxwWaQP0OFl2FGJ8M$<<QDm_cFs?KiBv+ILuh6e<`CO
z*u0HOU_>7{$!Wq<7+%Udo=E#Bi#C71$0;mffygi<aRawBM|E;alYY%tvrmtgZ7f=n
zJCr_iyq5FZxh2R!hSXNn%<MOHQpK7BQZdnUge-#nNn~V~<aX6d7=fc?x9(U@LXY^6
zvxs$;j+)12unc&YY+W8b$dQ=E#;Bb+<@18Ir967C0}^z}hVd*>I>vxhxmJe>Z!mM_
zq)OI6<jcAngFg~D<Ma^e7)#@W%_qv&do>mwFs;vS&kLESdPfkaw`N?*Oq+6q<MIc|
zoc`H_R03}w98v>QJB5xWM@nba6qsy~Zoy9I53bGMrJw4ayvgg?wY%IqCCQfzk2kMM
z@s29k>%R$^bEYa79b-i8rIW+JMF#JDg9`6$nM=G8zMtJXDZFqfv9Q@uskmwj(<^uW
z^A4puJ6o5Jsl1-QXX{M$PqBN(?w!TGK<yv<2Qf{%!9uEjRYIz)&66YH=h)#blTy1@
zxd-S-M>ACXeCDu*zrR*s#n|EK^(FS1#^eQbqs)b;t2_q<&*u>*TMKzT+Y`n?0PJ1~
z80IWsnBV0{kc-)Px=hP30>$&HdBjJAZPB$D?sO!3mpDR}U~L+x5-zr?vs_?O;)Jk6
z`Kvlk6IF#RR#UR`=45N84|?dmeo$l7PSf{8ji|(?B>Ex(;@~8=X0Gh4$q>N!_9+T*
zE1e)mPCMa6<g?EqydOH%M%34BfI(4ViX8cNirGjbUsQ0ho^E3#&=~S$g;f0vsj@~y
z2h|JAV}k4!(1o4xI#0&P5sjW(@hj00n~0SX=S$1S^saV0=oFygq+62zxPFab?ARAb
zP|078k<|agDf17&rnM>HG6+BXH2n#(qjJQ(`=lV-N6_VVmG~kMem=%5t|C@97Fvka
zJ4WQL7YlYof8B6??(o@Po5~n7<^rlHJBb7xmLS-(ujF&4cne1x`{I-en;t)a|Nll=
zFnR5Nj<?C<7``2qo)R?rlFxZSjDDy>Vmki|pZQz4Z=MMQdm7owlxQeYUdmwwQyW&n
zWq#X4&W7~4fjA*YZiHpn=sVdky2`*ExR4{g$P!>5`B0*`x+95R`FHjvN31Zys+F`$
ze0zp0QRp@8%^<;>0FRQ8%8*^pG|9k5Iz!szNdH+FvEBb?A<O{cY%>gO!x3QOg+Ev_
zt2RFhIU|eG!|mKBM-06svw@QlPB53!NE^o2m-FdKNkjflnZ5zR2~FjFoFzx}(ayER
zL0HMHIE3^E4m%GMzK|n*c6Fr{5))Wb`&CzjEN)XUH#E9}u%eQye&9~zi<gXkuj9e!
z%bmgK)kMNPoLs}e8Dxo51lGSZCyD-}B8mQ@n=rPMUQJe%U+$~4pt5PzX;kTIBWfdW
z2di%$y<S4T2q;XHIhkq0E(V|gFP_qt$dP-tmw?6XeLeKCf-}*GM;M=tel!sPXpD_#
z%8?ZflMs$Qki1HapTheVPCv#2E@=#NtzMBtZ+BJPGAc0x*4Tu8u#-i=4(<|tCgY5#
zQn3-W|1&%RY5e1)Cd+|&VKO}%#utNM80P@dr4idW0`Shmylj>%HZc<GKl4R{@@+JN
z-RDYmisXpVURfB7dv>uLN!@H4jC%jPL**|>I%K~yS+bK@XDAznSaV`ILQVpm>l%FQ
zcscUnA=!wD?WF*=ZiKw1^-p!CeG6BF9PIhj-Jlx;0$r_wLX7m@z=(>a{8vy_nnVq}
zEJv|h^;ELN5lmwxXyB66e^@rfhH-*{@9mXdDH4Bx%~@qZ#`}GNo1^^&>F*W8#hLfA
zv%JoHshJ$9Npwo0_b&LwsuUG*dEH<4Gt?Lr-blpP(~mH5dx-a#a>2yF1Ni5tK)Yn=
zD$#*kLYW$!=2{lu&r0*2zt%OoICs~$!bPu_-#nSXjM^86NWmc^UPAtg7xXw-&xMPG
z1}OhURglBqgBFw7$|Z+)rNJ8h+Ka3VgrH*URY<5(Pa+kUG+Tsow56U|bpUkp7UKtE
ze1onvUcamz_81$PxXZX&dGwJ<7gt2hH!{2o+xq4`vVV)U9Enjc<%y@TAu6os1;Jrd
zqCW4#L#Ze1b%sb2tVLy8uw_Cp+6Z%naA8y_JI(_;LSg>pHjJxT3bBC^!o$}vh`EMh
zAm<^Q1NT>PsDZn2W-(WlNZx>-43F~Zww$z8rS@`Mki_Tsl!<3z=As7=GM?F>^LQ<@
zQNl1lrdAORzrjyGrjjF#V3=Y~1?;_(8s}95?+I6#-B`3Y(6G|dA*JJ4tIpgz`k#EZ
zN#8YQC&u=^QEge#V@8p_27;gjxa$Yl)f&z)j6W6iN{>RvaOM$*OPVIU4h+Iy74aL^
z<sLP5f0Q5YoL%ctqwjY1%(38N;ZU-7tkqg-KmJ#5hGlGPu2fe?VubO3Do`bQ$>A3z
zD0^CqQ<(K^vVOh>k5-`9L7G=LXW=QaEqcq@<{S3-zqo+o^~20AzIneg;Qq+jHjF5N
zXsD-D!-+i3eFFL|6r@FukRR#68198a^2K?r;f3@_Z9y?FBv%^^E|^3Zl+M3BCa(O&
zd&gA5xA@S3^n;`Ck9MwpHYEA>@?}Gw4h=@DcKb0F+9>_vR*Vl@!MU9<;82#>+Z*_I
zpinNbdmRwI_QTG_LXhF6pG*{Q0UM}Z0-QGTHLbN<I(jjwWG-2N@h8iX=Ji-%FP<A6
z2yUU|E@4WPNY#$HU&(SAm=S}q`@8k0&cF+n5_n^;)E34W{1r0IMo}4DiEAhXC^xi0
z@L53FJTPKvK<7jJyD*ne9~!t6?~ag85zRoy3f{9H73#PRH|xo8bJs9oOb+_8vgtw+
zy%oUSpT&3JEy62|%wFK^JJa9Z6358m162;Ev^UKl^6={Q_<<uqx0h9VpKmzUBR;*<
zxbwvKi6h4O6#D|<`(6omel{w+iuJ7J425yNewHIAU(>Dzj)O31jR}(y1`(M!ugo>I
zxFW>YJ3)zv6ef<@IkU0D!g!x?csN+*F3p%H*Aee)xtTHBR@-h$d-}?YDw|@K=Xt-{
ztYh9VQ`Zd*Zp&Q0#)dS_z0BSAbYGW^X4*g+tH>lrxPaxRlkb>0^se`km7HoOy<mLc
zglbDIQEbcR-l)I@Tkm|y<lk&)@14&{+!@1Ma`X9_NaKkUmG|1|ehRkM{p0C(&^xa{
zYF<k~Dj(V$s9yyugFBoJ;Vh8Zns^eU`}?@H%t>#dhmFbjvPJ#UGUCj1*D&so>>A;+
zI_JgiMvt}}fBB$vZxX{EE8c<OE4%wACF)|h6NtqRQQV<|n9AbQiK)E83O(sYLaVu3
zlO3=Sr=Pf!uXnSNT~NEg?OC?cEn1D5n+G>6c&XNq2YnBEf1+34L`Wr`u$S8hs^>OT
zqAgYA{NoS|t&St)Lzrec)@L8UMIqQEf?5aG8S$D{Cdduwx`cO#sj@iB0Eu(p+NvMr
zp8PEovVwe5EuA-_;+CT<d7qfgDOdM!qrdRXO`RN!UK0E^=29C_PbXr5x>#Lyjhq)~
zLZ}G4!x>|5Tdonglr(sTW=$)1S)R*XnOkFSw>`w|ndaFe%i}xebHJT$JVHJPGFu$b
zp@dV3zDbS2c{ln9BEU^FsmEEld{RPq%DXQCvkK?EOjo&e+g<k|6=00R4~(!3Cx^Eu
z?`%ln`r>)5zVEzUPkz0$pe(sj!~6CgkL+`&L}yLZN6&4KzL=!k{k!?%9c2Q1mthxQ
zvV&+WVP?k`rAIJ_W$`Fv<)N?FMJEci_vGDtxA68`;&554DLd35yQX=a$MJq0hqtGj
zM@MWN9bAeejkF`jK{^VF1dhFm&;h&7OC&997=i$)TCRAVsG`$BLM30J_dytlx3hZ_
z=!0|lvj|^ijhU8f(Bw~DtUY&EoL`;$QnWkXcICVZ>ewXu>*EULpZ?(%ITA=k$QVvF
z4>D#n1A$yQToh2$J<<zeM1eN@1i#^({w&`4sxPh1Z}RPirVq|JF?PIggMrb9Xhu}@
zS&Fn9E7nTHSE2n=Q7VQ9_<~&9M(qdq73$9uMliF|rE07eyDs&XZm#}xoF365M@~RR
zuP)qdtW+n+_(N@7oOQnWg7AE`A;Z-d&ACGD7L{^-qJMcFjH-b>Jx1}BENInO!+U$P
zQG9@yc9-pLBl2UaTN<{GDG+(QXlPHY@7Jxe_f<VT(<z}j_~gE4JxTPI-ap<nL$Z;F
z@3!l@AsN9bXQ-h&>5(K#INc%)o6jc>O}^tz>c263@aFFL2M-*!yDhu8Ze!5{!+EWi
z{t#<x2e`l?%|f(qsz4d*yuvn-BdZ<f1f%aOnl+`Tu`q_%$#S$DNo%@jqaSX=Xa;OO
z!~*bj2>BKjZwWjMAn%|TwRtD_%Qhx4r;c=PmW+dNegiOPZA&of*G9$F$q{QH(5i5r
zVfYHChM^|)jjc{0oJs7^k}AS(iRgF};p7zdd4)FTp^m-JsM9M0dqxD7l<9+()(zG9
zix8j}@Y>D?0)FRPWkYFSD^aSRM@D*<F{J8^3Z?!py(S0-9}LJP^aQq@Nlf8(8F)xc
zrCK;z{GM7G(4;1^h?>9?txr_RISZIhsiDk2NZn!VM5l@SDQcnz3bj4yIU?Nv1BCNF
zj%4H-%1%e}dOA`8m)HGkwT@qP7y_#OqY@nD&$t&Fi4s<Mq~mNc%<aCYl`C7=3Do)B
z-=T4b;4PJO3Q!=S``__uxUE6f!RWiM!T)7Ejs6XiY3Oql=mHf!L<P7#(1oWZzb@x=
zu%xOOxmqR}1gxtH?J2(l40P$Q8K?$@v}NFaz0wr4wM=$Ak&gDf0cr6S6x~uyq*@q2
zy?&w|^7AK5X69hScpHyk2opbC#(c|dZw1K+`tcPE>`zkhZ@W>^XTIdfXJEi(Qn1PU
ztMD@bJ6z8pnE6@?8H?f`cG8QQUGo$LDq8^Z+dc)Onf-qyLkMY7-Hrh^ltLD00D^&n
z$|-Y%a~$JOge?<5<uo>(hPCxUYJO-$eO6$NJyRm!K5f6$VnJT%hvuoHEH3ZEVNts5
zWC76H>XI&GKTPs76L$g{YRag$BcGZSImwpx^dtTKzhUDFc7$h=BbUQc=W^;4#Z2$S
ze*fGgx~O4*m-Lt%#lEd80ClN;5zZv9#rYsuUSgq1-$Xg0jCLll0UND)Y#MAeG8io?
zB1f>G5IaP659`b)dVwK*drd2>p^(Kq+$UYZiEQLZ#v0hs>N_O}r~NgpZ!5-cLBE|l
zj7r7{Mxs5Zfh(^XxKoOm^+$u{ECi-$LPGSvPZz}k>OY@FA@8iE5E&3F=-6{zbdQr{
zk~-F2w-P3OH;JB({Y3Ax64X0@C<M#4hh@tV-$%bH_|oOSYI6Ob=uypda^zabUb9F6
z19xULVujdPbO-a;fdVsp#zp+nBD}om#+--Vsb+^BUp{u%xLEzxtYIj3u2m-YEVZu@
zpx0(dF!y*ptT>DuttfZtKsj<e{0=FETwCA`8vVL<fpibBzheSY30R74Un<~e7m!~c
zAw;Yg;(G{60^Y}0%8@~YTZ=`P9bVeV*B>p4jWEPZLxYB9@48`Lo6hpez2A5tY29|E
zJ^b7R(|u(Ng3*~?1FpCj6Q1GeS?enL6;`F!>$KLDlyB~#hY6I(>FrIEMPZqGdG8An
zuer6yUC)nI^3m7dHt9j+fhn|46?2OaEG1UDR1-q;K`7gW)L{7Onlx$(&Xnptga^-w
zAlXVygiK}W!5Wi5yH<v(xBl4Iw;Kx^qPBIo)R_pY{W^RNR5K#)ew`4Nr)#wPvhX#n
z`@-i$=JzfsZz>Y@iQVP%`O$^0um&y9b3AxO)x7S?wF!yw2xl;nOpf$~9k63BC~lg7
zN9T#UowMsKjqiRTb!&aXkNInk>AxGEw#hp#ic&YVMJI1#vhqlsJ+{6!W0@iis55*f
zqm@>Hvg)uJ_I<($7z&h>O6VIP?02pOp>-98+a>nTHMWx@!;O~*qpw;?rT3Bzfum*F
znDCllj5P`Xa$;-=rYv=YTS43Ct^Klg|69vm?R7`-Emv|bnVQ5Jqz#sJil1muhzc9V
zJAqUOB%2{zSt#831VKh`u^c&;m?cL_lz2nfFQB*8Lu7p}b>WrwUmUDYHTXW6p<Mgn
z=9%$I_k!)VK%lKM{QYkC(ZM~Fp?Q#(@mZ>|n|zTfUPpxbwk8rT1|>ZF+`)dwR~N;M
zE=w7g)qvLaUHt6TTHr<PQyqvbkg|lejJ_4v6HGGNdOQGun+?HeoprF~t=8~T7KKy6
z7#?_7l-m-0v)`~t5LUo=C_+z|dkGJx-m5a!pP10>^E&+Otrel$E?*X1cBVB8em6sK
z{D`ZtUM||R*Og1EK>%m`mHnp^$4$s#-vQPuDt672ZVZ5lrGPW^u{L2gLpm{gDs)5}
z@d$$oneLcmsNSEy4sD2v5H*yn6Hu`|C^5x?%R$Lk<mA{Af$|StI23ivtk8{d&o^6r
ztnr(s)Mh@`F{}R6u<aKS_aaY&u@E2)K;GH}pK^}zZ96mTGl+K(23MT`gkI4`?KAm#
z_b)P_bxLDbq2T`4OOk2qL4q<1=E%T>Eibx9+gaaYr38fu)>cQ2iB9p~YI4!=`J%+w
zfoc;FERhm)Mre{JzeA!ZOb179^`rFj!Fz&bXRQG*@Og<@oQIO1Fur4;1%r<r;Y`No
zq~b4@iLB0j;24~5E$M&JcV7EoM#j2x>(>pwXy5QylnA8=aE;$?J7+Nr=L@$YA3}y>
zdZQsZJX7>SNPB`sxfB=?7hBcAj63zlFzL(<*X!Q;)91HZ%;t2>vHR|Oxkvvdh4h6^
z<0&vTC<301<$xNzB_CFh{1G}6J|xkth}<h*tPP1ZB8wu|B@<fLi|9oHZLdR{hJ+0N
zq32d~=-jchlcE<z&e$`-vG}Jw&!AIKnv@Sga45t(KGWZ6G?BO_Q<m7%XKc<;?Qp5R
zj}@*<Jt?j`bgsOZ&NVq%x$I7KebUIoe;6du96>M--y17cr-?Ta#(1W9KT#sODJc?+
z<_u$N7GcVq@m~9gJ>*v_!;&U?wQ-Nn0sf4uviF6J%mzo*i`E+lebw`#*7AO9LA3wd
zEP4~Cl!3eQ2t(-3{UP!uhjYr&(RZaDq7f+l3&${Mvd@U`D;Hk#Hq!5J?NUE=F>-@y
zuv*jGBzmnjBApKQdRhU2fz8W~$9a<RJ^)1zq){IeZg^b@Z!C0%&TjS$?$lf>isc3_
zOeNflTz#M4Gyar*bH%vKT<;5`qwEbRkHctdpv(BHK)R?KBHA&qf<VAS_${bEy+;n9
zBf-kv@`!1#j)|yG9yKN7RHtyKN~wEu{CaY0%N?uyTgIqcM&6?{UR=W$AzC`ryR({t
z(KnLj+H;4Kq$oZk{D$rL8b20|?Y^9MbmrF4U*oTCXxMB!_NSTl%`pL#&?vz43lxI;
zpENc7xl+vd2(zVuoc?@WCzyL^AZV;r_7D-5{GtKVzbqhCXkg<!f<-<>OPMlLR|o>x
z&a{>?sjk%kP<)7Xo^@2H=JSKm(nrD8k7dE=hhEIXL}G%fpg#<&+o-ZvzX1G!Oe_D@
zDVYC(c;|26oZd<$=`Nt4Pmo=J-OKMl`=i{T(IOcn-Jfd?RcJ<;G(NSne4ud*B~!I0
z0a~5eW*ri28n{r7jMHP0eC=2=2BhF-Ey)V#Uwg`Vjf&k0reR+WraRLj{(*A*+?a+f
zMe$v9;2lB~c!#PS>3IWMQb(c<BUXUBa{HFrDGo(!*D1#OT@<A5+Aq`j+pIuuZimP>
zbS?HQhEokXr5{UXJHv}AL4QtzkVb^x5~1lam<LUd5IdJBY->jUfKE~AL`)2qUdt0t
z1^oL%*ezo~%VqAQpScsGIZ^_|2P$`Ocu={U1ocv<G-uk=wg1>rN0C9G<HrjCWmf>o
z9tA5nJrzD;O~XI&$*+mTWIIK}9!T!@zY9%d3re62#`xPt6Ku!{fXm16a_CUqfkwCX
zCRc=T4MyL_k}BZeKqj6t#8Y=-{r(S<=<nyl$4xgOEbYkE>9E4>I?ZL|X<%v3|F)7L
z2)0T9nz8h{94Uf=9BB$8o`w$WH#ea)jtZN*6@8Gv$F^T($>z5MqcE?b{OInaK&dWI
zGDT6U1I&8E8g9zhfM1kjunCI%*9WHlhjC|u)zU51NEfRSO0AEX&D>H(19Yz!bu+y5
zoy29w@|WIKa&&W&EIN2fc9m}O;k){$Ep7(9p~03FB!dB|vIrt$6=4844~z$3JK8LQ
zFu4Nlw=xMqiPh;pqy9g9Y5!sIVHvz&2Q=vjiY>6r+*0O{K(`$HT5T;k;S#_9RE_$w
z1d-(nukK;%zBPD}+qLG6ZGR=CUxVuF6NvL~picLcw*!@!%&#Dm$G{Eqs-bkCLpVbM
zH_MVO13d9EVR&%hV$rm+hGXZ%$sWxtW!IqG^$#+tS8XqNcwDf5&++<+w&yaxh0$t*
zt!qL5g;>dV!rER~TL1)`4d+?fz^MM^N2&EZPn4MCB-L2l99sr}8yl4$B$*V*#B+~G
zX(E>=)+VCxNPianpf)pHYThhRZ;#8#z5Qi&-{)udfh#zi$h$c6z+$Bw9S{+Jm@@DU
zz>y~F)1*6M`TC*cj&*>Z-1|6dDVOfRJh$<L7ruKn<j>+XFa{@V^m{=hFBMJwdT?Ua
zxm9*wE2bAy%9yS&eQj=o(z@#GZZ1<NyIlXP-eUV&j!EG1H-34N0|M6J)ScF|Ywvd6
z6>=;V`}&xzS@O0ohTm#N`Cvxrh1O$Kcpm5<?>I*p|6GiQ^X{`sy2r4ml{D;jEen&{
zUt?ok+B-5*R=&#EQt$ZL<l^9yIzx6VcbQS%VnvrRVJff3BM|^4ei5>H5At9980EWN
z*Kmn0r&@47R%#%?E15&`_u&A-c|bY}7XYydyC{p7rHfIbPr~SCy!^KBd;ei*ynWe2
z`O1U6*}>#8vo#}s(1&s=1WNfXBDa$H>*M@S+AYD)g*SHJF?&L8d>m)oo<kW|pJ(Mo
z?MNqlxl)Z@GAgqbBh^t^K``nMvGtu3F}b^|LSl{a*I6q|R}z&V->oA2g?fylE%S0~
zq-$C>(>q-qFZVw3$zPx0xbIU=U4qT}Sy>-K6#WtHAO8yiJ^d2q<<tP`ikIx{9jR4-
zWQo)RcPVsWjV||PHP`2Fy`f@dez3Fp<-_Qg=UB6@a6j#fIg*g?gY<>}u`iQbdBP5Q
zW3PCQER`orY*NAvpO`dgHdmCOn)OmIoXqe~Vmocn>_0l+BY)D4eD<Qd<D)Lm_T=u4
zf5P~@RS_cp37#I|dOq1bxxFDQ{l>czkCgTUF0Rp64leKst12xWK4Z_<9eWOx3?5}a
z{bcm0;h*7HQ5od7AyCZ^@!0t8o$DsDUNB_*!|VjNT@lRU<N8yC2aa#Je9)b)d;DpU
zuIeG@!3>;Gcn?FSAIoP*r$A1J$`#ScX|$g7I^RRGg;T|uh8Lvny2B1AW@<^X+g@>x
zqY``)uBR}Gu>O`;;)azL{C!HdV{+!Lr$xP{y{Mt|askQ1kX^l^)F%mgL_D4@-auru
zauFiu=}v1M(aK{<vRn4G(;f1+eR|dUifKBb?L{8-T-i>fo7emb#n%Y>O)E~q7=Q}I
z{RmOk$_0!jx+LBuddX9<UJwYKYDIKCe#+Zj>TG7o9Bk_w_iFQH4V$e+N$-fTFXo<+
znct1d5%!N?giXu^!#x9}*SCBppIQ`SvWdro(He0us>>4=Fo#=D4BQVpwvMa;dZ5Ne
zH;O)U^BtP+SPK^7x#3*}b+>k&DZ8<7T8TI7VPM&Ynhc$AswL?A-yp32LDZvmVuWF?
zRP8W1Vz$r(8YJ9AiBafwr>HFqJU5Ipn=si>m;KnTsK7BgHf7_Neb4+42bf$Ad@wFb
zGbZB+j4OukAQ|ZU_|v|sP6t3SV8kG~ZKq2{o1?P%?*NB`(0X1|st<W)ic2Pd1r@wi
zj@VCv%DCDuVk4T?O_(uA(3qRy6owqp!g`+Vf|ijt|8BR$e||XS|5{}n(5MqJuL8P$
zBrAgy$1GBABGRAl0rX*Ga1e!Df;IY*r91#R?N@`*$Jz)v6BYiz`b<sgx#Wviu-R!P
zpnu1iBiM2t?wt!5wb@g4ntK8|h^j#^_8K4o);_ZP=!YuI8`}1HW3V4X>wx@D=C1i+
zhMXPfhUj7eSO6b{PGs;QXtM%Y`}(1i2q)<mG)%*X0B_0nmqm~XmT=NmEf~!(A!bv_
z)pSuh`P8_Kj0Xx;9XDV<-z20$+cHXnbL3DA<;s!eJzpT6YYbH0el4_9Wiv40b1tl>
z!57v8&X-nFj$ocW6sac0?`C{6Ad4}vIq@B|cMS@87<!q0KF9b^lz_>KM$?HVtxmGj
ztO!AWcn7?3RDq&zq#OE1<|vN0IB8~4o(+uj=e<Z&>2*xl&Fuq|kAzi7wL!9;^ajCZ
znEf(`KL^#n$kzkZ>wk_W{x6)U`O_*>6a)I&K*5P7$8nwuGGwtwEVWt^wVE~r?ibxD
z|B-$yOJxo7VP@_b{`#0QtC#aPXTLnP=-Z&kiO17kjR-yEMU{49&2c<0ZvS|6EQ&X>
zA;_IG14v@dsbusl;>^UFb41~CWU3b!QrJ$}RkZD>^e5m7KCN_e?Q8H5<u{TgSmkhn
z56?H2GNd;^Sk<~E2CjWoT@(suB66e9zBM+CjU=C!Syom)TU;qS|3k%2m|AjJNePX{
zM$TZFu&nioyA4&fO(tq@Ei>P3a=d=qxk!6ZmA%2#sUZ%?aA&$0L?DMU!M6$QI<mwB
zs8o9bMD)Vht#xoBD+;`4z4Zj3VOwuI)6PH%Yad`i2#o*o3Ah21d7Vfq!&+k)V$jz!
z(w|`cUP<(B+kx&u%Ue&jScrN%H%sPtlS6SVQtFO*iz+red~_c#ICi6>?m%r<^R4Y0
zKj&6;-Ss%G*6J1W#yMB#zT!}{${+0#z0HkOIfB!pt(9P|8#&rUM+@Nh{8e}+e?4R-
zqx#K6UW<O{P4?`fIk~Zqb|joXkgNSU!Y4~+s&gWDy@S{GhZ`q*OfcUz3~=!X*!^>|
zcsRq97YG!j=!{T~kZ#UH6Hbb{g;+$kgRwf?*|D6lyW(D5pPN&i6>m_+8C&B&O%kU3
zVo3bSl<Vf@FH6>4uv6aiVanWbdP(}3HjFuBJ>#1<npGz|#XN!bhweH{A=T|}^v5B~
z2RO1g7xE|s(C1Uptr#3l(gjiFF>IIfK$*evdr6|E^=M*bzp+WF98n7#E|e{2FMT;i
z-%zyUSW=VYo|ToB>l?0jWmPRsuq?J+A-epUW()>1O#&db@&JKCzBL51l0vgra~?^D
zP!M$l&8U+aaS4T*$>r)`EUfUB961i{#+qHIs9OgA8I~?i6l00NdpU1D9!xCCgEkRe
z^!G+ik@;c#_{8DXQ`t)5KH|j<qP00UTFOXy7pVLoCj?p>NcFlboq2cB?D+hMyZ#P3
z?CqCW)~?o4-%voO_5)DvpByG}WU;K?7mvxg_rllT*KEDG7PtLYvCr~VMv?NYah|6)
zZuf|uTe<Q5+o>VTG$@_7By%{gJFmnFT}s63Dd)nAu(7JoZ5Iyq9;a$I?nuQJBg+3b
zeyZ=;!WzL&<{!Eprx|5A?=IWTEzBt8(lXJQb3w^F_br|~S!;Vn%}=?HmVHDnyuaEY
zRJ;8|-9}f2V99W*@|MWGIUpE=lPXS5qKaY`l|7&AZcX6&?(%5ij&3_N<bC-gV&_L~
zw($>#M(-U*XN}l?Qj5BGC5STWe{8`tyaY1fHK`-9PekK0GXgn-MZKYco|6+>5_KQ<
zvt5cCEF$!uO{zLjTIX<m%ce{I2bz_yEY_#YD-6a4X;7+(HGFGRQBzXCQ%K;H7F6>~
zOktg=cY#)=zwv?o1%8t6k5exC_|7j(?OwLZIlzq4#>2~*S#~njM2M|E`CGxf<B(zz
zollQy8dY`ITrfItG7;1KSgo4XkR5cGZ#}Wf?}g9j;hzP}cHFh-0ba2E)&ty>okX7j
zO)o5B(&QT{ZP2aK{kCy^t+|r%G>2+(9%1`QvZ`j!mHQc)xzXM$&b;5e>6wP=qR{iF
zH;&TG8v(J_i$!=d>jmVAT!jbBv!+8Z%SH||%a{!d#6b@U15tF7pp{!V@2er<A{sGv
zTzHGg(e3BGujA2oW0q=o-;6zNuNN}GX!^@VX%nH(qv%E1p3_8Y>1x}KiTl~<#Vezu
zwDcG*c^-Zq&+p$qweARi@F?)VDK`2g?h?zuz?#mkY?TFyAT3dP>;5qFaVBxNCGV2k
zt%GYV*EGIcd_8N;;t8si&vZ;pzm1I=qi7WSeI}s)NP!5U66iuO;S9lWbBPwA*<x4B
zq!QH90)4`L^m$%0LyNez7%z*=a^v3+%*%e0Wwk?Gy{7u>AXDC%03VGb8`K_*yGQG&
zlbWS>cFU2ajf%*on@EH1L#SH4I82UwEeby>WdRY|DH`)oSfTEvKfRJEWbq{)Z}Kf}
zz1?+a*QSNZo2TQ(;!9Q~8F6FAdqmk6%q|LE)<BUyAAlnVtW}AC*6xVeTighBzV&ob
zd}*SZBSt$GXmyJn;%2lxDCUK~{<&TI#iBfE#J6MXS!Z=cHhKc-Y?m%bMN14@X`~Ko
zEcPuW7)@mm=CnfnX^OU~A*Okv*wRFmU4r4369j{$SPEAoaCw!9l-~S3ZL@1}POZTP
z(RA$t39K2zCMa#)_j$SPC>_EEt>Tf|F5&=4p<Os8M@Bk99_{n7(b1;C0sW?_Z+!er
zlmp#}T`k6Y`d%JzwT*1$s_c9e-+1J~_3V?;$yPO;$i#@&Z8XE<$d7?g{g(;G1t;hm
z#sgaFqA_7w&Y_ABax5^&l1Da#7XIz+mjrD*jCa?&rDfX2CHHj7O-ozeIchbxnzFWh
zF{1W?GWhR62HGENFP+$t-e5pRIF-_+bDCy?Q#c}+BZhlp{PYtA!@%0I*#VnGiGZgg
ztko;m2w9~{o^SH)h25%NaXa5RCA;2MH5XrZXYHPR@yLc`Q`RXwsa2UmUNigm`WpU?
zQyQ~>s2D7NQ#Chx^K7S@*2ohj`d2)B4#hkiYofiWWNNbN%#Ah^_pY3qq?|}B$_Jxb
z6dQ)&OF^&wHVI+i$JiF+F%xBwTTVED^tT+u#t}G6He!nwpo%fEa4z92f)$Py3i>tj
zXT$awn*hL3d}GwnX?$MK6ou?P*p(?qCO8j{VWWl4%g!urW{oVGzFU9Zwo0?NA5PYM
z+^pxb*kH;B(^CsSj0*OmUj6R_Vn#c)*T>$nJ@w_0PgtskPr=-5r3HtdRzBtXIb68D
zZrj{<NxJ(IYA7)P0$f6|w$*qA39XOhk|g>k2pS{rRr#d@7UMHW!!HSX@Xb=xHQbi;
zLUwYEpWA|<sWOAuH8n?LuD_33-#g+oLcRJd`$-WM8aJfZc+Im#;W8D4($D{T1C)QL
z*OtT18Hrquo1ntGaPq_Pjv&&!<CZtMpnbUc<5wAJ6@#Dp`Rtl9ef7Jn{ZqzyY~FAf
zV^d|9yr`c9jRvHnp*RjWyOIcQfV0njfU;gc&^0#02T%nGvMat6g9=9{+^+39MTz|`
zX|HKtR0d9!^j*OCpY+GoWu+5FntU&_$1IoXtynO=VDZ))TQ(L~$nJ(7wqsc?bJN{3
z4l33_!G1q`8GL=&IfR`>HYDQbxP*PdQ%fbW6R`|y@fG{X5eyWV2Gsi-gdok>+yG7w
zMb-i0;cv}o5SiF4quHNJt?IEdaJ*ROUY^>Xb721?|F>0ceD(XE<FCcoX!i_o-9IbY
z&~iBCMyENIJj5VoT_rguaH^f1^V7Db`L7pGo93}`gPPVdzAp4Fqwg2n;LOnIV^_~K
zqbh2C@n)y}j7=t9ntKp!i+4nJ{cAQ(`JH7@N!9eBa^&^(=0xczZ%i5qUsp7K3s}h&
zha!}iPz8Y(NcS^1Sekz?qZ$)tFf{J?bxd=#e!~GE;E{7bC#-hJj;d_Em-n8;lxjtM
zIQFdjGS!Areh1m*-7n6&Zi}}cs=bp^V!%3WjIPShJo!#}s9#Rpq#a1B@6freaGsS1
zk}OC@ZM#G!B;DVMbVyQuzu2wTf2;PFM+K=jBJ5|WEF!=5EO85)b>gDOuw}w?rH6kc
zDMM>r4kmo%Gu!bECcJLFAoq=#L~r>JVpNN#Dzm|@X_2mDQN&k+80yHEn%1o+pD@~i
z&i~dgo}?4=Qj7Gpt090y!(q&lH{s!a3v|*djs@mjvYK4j;P>k0!g{kgqNfjabuFxA
zpkPF#4ccIs(6HgFamx87_65r|3k^!9ubgkWf4qyW-o<4G*@zzlzep63VLUY$wHNxO
zK0HEpvCB$c-R8C=j(Rlz+;$(|H69*Y!`83yTey1UFoVH+eb3nFwoi)OG4Zth#LK6h
zCZ67JZ+O~X`W+mC@pYDjLZF3fpqcIdwQ9CoQKqBG6D%x!;{CR`-KUpzHJ`N1%*(#v
zxx@1OHhas{`yG)Wke`3MQc*jws}TP#-z5+QpKu!Eww$>ArNQ>1r)WL<j#-z<s{DKR
zvbhq|Cw&LR=g%x=4cdO@zR{cgzDMAAju(W>ic&_8IROEc_9zujO1>UyE>2F{>3iwU
z^8WA652vQ>F}drRm$(g2^LTLDengt&R-V#ZW$6=*0VxQ5h!Upmc2(wBG@A2hXTz5G
zn-kFyfg>s>D`I`dEYYEm$~!MtAKf;0(xTII=T$tXkl_#leiz(H7Z67_FL1(nQnlI_
ztMKw;iQ)cKQC;)9$&eUw!k0as`S@jHi-l|8f$tX*sezmwrC&$Re8Tt$i~o&{(ch&L
z7PMk=5}02|5&85$ZqxWWmhpAbma7wU^WOYWikQf#_r>w@y`l$V4>Spc%Hh8h%b}g~
z|Gm62)ExMnrPWPGalF&Q%50sJSF9_sl#QFqGdN!k=BnCHe$$zEZvjPWqrv;DDePZy
zZ_Vj}7R~Qx?DJ3Y(w-5&fApzM8`=-dIdhbwH8=;i8@FE9^lMQ7jbSwO4Z@j@@rwj-
z?yT_)`D}@x-(PVEa90}jPg{3K{Jqza`Y)X)i-dc;eQkyvzChflX0LmYNEceN{5Ju>
z8aIE>w8SUmhQ46*kx5n~R`HV2D@MtAHnavG-sewriG-o}stjt^9KrYRG@0T!O|&yb
ze2LZ97fO<4a-_@%kvQf`rNACA_kmNyK`D(S=O;PB%K&x#IbvN6aUPwC3EqPA{t;n#
z=g5)73ut#ICfy%Q|Mf@e#&Yla@R|pU()*V%zQGAF(iJ|y=5U3$;+wRMDxFALVb6St
z3aI_Nx0sMA+HiE-3X0_nsTxYIN|ua<Z@0lQGBPe0T?&HJ5%_i;6uQs4=Ka9a$+$Es
zJ|cMHvK^;(%)Pd++9;|Bb-qIe!LfX_vkv{<h06NR3{)xlRWW|dH8+AGZqQ^rFmB;d
z&Z%>{_iEd;^LFg(>i?cal|28=ihIsa89wTT?ot0?Uy2XM-xj1v!H)+mN64L+urnPB
z+a)HFX^i1$Px}pMh@J78_9xxYv~pgKoixs=3oLlMW=jQ`)bJ6ye=EOmQi9cj4ad|O
z#<1qmPAze&AQvZ5|8d%qbqXAo5(XtuX6MY)jv9M-6zN<o{Uz(yh#xzItrmuM+I@zO
zvqafH-kx)tb@$l1FuSJ>Wn6eF!{#IY2zH@Ds*i!%Pa~$lM8*i9&9@!S5kq4zrZ@o1
zZ1imx&PPZfj}-32eNj)_#~LDeQ$vDGi?_?@pN}Klf3OzndA9B``>O}6MsK$@Y@^PS
zp=dL5J{UEDuVi3+JI;kK8GUL2GG0%*qQ7@p5`B;CBJaZo)U#V+EEs|PuzU?l_d6-7
z=<ZO5?hY0iQv}(bLV1{@s}n)dRrOejDj@1V7}n(VAZ~;XfXRN<WoY8Z<A|ahJE?+L
zU<iu5$ccPrSLh@&-?^3|ed9bhQse$0_v7}+msMK9;sxt?Yd~WjMkULA<zT;ZL~sLz
z&ZAi^=PeJ$JkvkKvfU&z)kI@IhCUbs`|uF5&mWx)`jgIYqahI^h8qSW&kWy=<bAFU
zo-8r1kV>572<MujK@ZW{74t0jGUeH1!&W&m{`FuG#o@3v#)l2~@D%f4fXEL`K$_45
z)S*z_wFSi707a2uKUTjX5rp(ZK@d=055A#XrSeSq?qk!KJo?LN^zQ<~Mo+veu}_>x
z%Ip;<Sq51|FMFzHQ0g>p{{1sO?}lQOzLsDu8^(J>sTN(aRpLefj!r@+*@`j2Xyxi)
z^ejqPI*_Z@fS-GGG*^)49BmjM!)+Lkn79uQY7=6ZVn!wufvPw4KudTHa`f2X3}mFe
zAwu7O2uoa!g#UwK41^H^!U*KKF<9(p8+9m0{PU9~bYe~uTb&%%Dj7GeT<sp7l6uZt
z6_~NQTXfR>m_81(m?BC3KNFq)^?*nf?IK6&B>>tBnM=OgnOXdjqv3dLqbkG~_00>C
zl#5Abg0(YUl=-i<vi}Q*Lpl;wIM#ta#JpWMHU=6I>8;)2bK+i#Ht)=7QW4pvHxvFk
z7Yx2#N%&?p=ga#&lcG)DCqLIaXT2)ReJ?EqRNVa_!1nktKv%^-vO(HBzZR7F{}B&T
zba}b7u5ioMi!Hk5iAp!@IN5UEkceqXu{VfWZn-CQ_4so&=R6DstBu?bgh=PstxFx3
zE;T;;_$V(wVb`r&7vEn`X!+D)d;Z|3nEO=@)9+R->kL?s;eOf?S@_Sso#0Pud-NV~
z;ZR#_hai{%UqvB|-!fOESwatGs^nwo3q?i62WEEY8{^R@?tYoIv24VJuGQVnnF$_8
z=+njqop`Gp|Ag_zMLM&hlLhVgeB44&)HufLbF*O-+e!7F`R<eGvWvva?#s@4#)|#2
zMBK1^5C4`tR_K1ThzMNj>v41J9m868Tp&L5MXhnk;JsE)W0IZD?S-TMl^~IPgfd}U
zoAo^4Y0pJ2pj)R=`ytMSf(q`%1yDW~-CfbJz?Qfn!deWAtqfNZjBA1o-c2(_J2KPj
zH#VV>h9~ddx$|S%t%#QR<U*}c>y8W6YyhKGLR(l|BC%i^9QqPDkk88?vd-V>M<E@u
zDE$yb1~U}>j+`fo);O+g_AYp)(vR|;)dd84@)?OERzHNugxcVohv5EZhZ1ijsW=WF
z<LGN^=iJtR_i}WU>FqlQhk5+O9v8BZmIfkYqmy{QERwg~*iL+ipr<j(a&&OFNjj-4
zHF^9v?*ZZ89Jy!4-LLthu2lH?TBfIJp7#C}^YOgxr1<)ak%nfJP61woX7FS>G&s~e
zrwMM=M@VjGOr)d3#qqMM9Do3J*lv?BrjoTqpc8Qz|IWv>vb)p^tw-Qk)Q|VnB7L38
zuC7j3mDx)xvpU{R_v;wzYB#IVZL#jMjS*YCsNbteomlZx=+8<bJ_8(`R>G9thw{l9
z*zO&cL9MJv47X>5Y%wN|wLHd36AT}VoVeTa>!gl1H+N`fzudd?z>5)1K|#}d)!@{0
zlx4WrT*MQ;i=Yp_5OFq81Bmm_7m_1j<8mR|K60js8sXHW00MWN_+396j_U9l;FKqT
zc-{$yI%gCI;2F(aaIDEn<XP0U^LuOH91GX<qvqv{7xErt&R@IRC*@IfYKZqZ<L?J|
zm-xP#4>*PYM<IYB>4hGFmt+laxY%N7!VTa<6aOOO|6%Xl<6>OfzVS&kNCu@SH7K%6
zHf^YMNeD>@mDC`qB(q2wHO-Y$A<3>HOA{fal6EMKX^SM<Z`5o;O6i)^bZKU;^&Zn&
z>weaIFYA8R``pj-`Msad@BL$8wYlcJ&ht2r?f3W|og#;_!YPGVd5lNFTMN5(U(~7k
z(7rI+JTJY(vLUh0ah=(-l08>#8z!f#16%U%S^x4ILz-awRO!tv;#yhy(U|6V@!9Hr
zPP0S18~aWtPm7q-*yz;ETrILt&pifRGxd9Nb8<rs*0@KEp0Vc4oc8Uj*Njb4I79mW
z<;Ms03p_n^CZ1A~2)<b8W>&G~d=cx(eMWpqa|c&@p~Sf%C%<Vx+i_b#u)Fixr-g-2
zjxSv3xo?ewUiw!^g!6$UIh=%E?2u}M)(7HBFmd~m1^Pq+^brsT?kdVABm6hea>e7@
zlu;UM#OLZ+^O!U601;(J4a>AseTQbuS&O;JUk2lj(FOUZb()*ow`3<e?e-@5O<J_7
z_}EfNxD;{(5vV&lWI{Q?feL~F$OIhvbcEm68v;3DLIpPz8($OH6jdom&ElO!HqpYm
zTNpj!Xa=P)d79=HudSl&7N28}*IoOxCjKV}XxKxQ)D&cX;7CHssPAI04mi{zgg6I+
zJ28p!oPsYOZtJv@F~KiPk^58A50E6lEI>>0N`x<ko*~K;wMDZprF?lhK2ua&w8Jbt
z=6-WAz}!sy?B>MP_f-q@B7|c|3hlAlTGODq)twCK{G}4eHokkvgk(suuYgYMcuS6F
z9+s1mxJ!V^VT$2Fz>1-vJMxa9%7xQicfRLhUy3JI(E(Vs!p%uPVogn$&4ML2UuxYO
zmww~KNJxqZ9ow%Yhn2N03RQEu5m{<E{mh(qt8N`GPj$UHbI$YaYc}6c+D#&rjUpYX
z{p&tDq%052IB2eN;yqiuLmw`h8^BMaj1-7O9(KUAyIf)}-659_23m?vBF0D;i+HEm
z?PzI*Mnl1Nek;YLF7Zu`V~(xMjzUgWq97@6>y9ZEJ<*%gNM4FV056P~DvO3nYNI8%
z3W8h=uJU|{Tl&Fp^wGjNV_D-_wVX#Lu~>?L7gnh4{glI{Or;%&vmB2R*8~#RRee;=
zoyC*uKA0VyFsnpVv~j)O8RyNhWTg-!8U&#e;lDt?FGO>BvWw^lv}<=ZKxHl~NN>RT
z3Tb_3u2Ch1wDD;F{R4223(QC|$U%ictS0BY$pip@bV59l&;jB2>?Fz=gr7WApCPDf
zG#5W=II6N?(560{eS%nCSYzyJ^2x<kqWr?>o|{&x?OkV&P{x%*=hij3+E4FqRv;jE
zj=hJtMy3r#1?h4(fU16sB?WMxlI+WCQTif8i=$lF&GdGrc+heM#?+W#E25Sfs=c)n
zYMU+JdqHdKJP}lX%pMlcpKwHlgyv=x-RpP?g`T#qPtN*}pI{nc9gh=t7%r%(6!e&_
z_qfm)cU<qn)z&K)Uzc25Ur7pp*Vdt84(RuFhCN|&+;}->C?GDxQay>1RG=+!Vga)O
zb43bn`<$mkEWmAwH8&XT+<!vQI}MNAP<>E$|NEET?w-La1!JA!=1<$LT4j+y>VyM(
zf*y<+GNYVB@goHgd{sb$(EsIM>#%T<>?CU1R$$^3fuW`p#q)$2<9s%2xw@=ukOl~~
z9oE;TMJX+EYi#A&(GO4baCqd@bi~|!p2D}vzr`W-pLbpU2Cn(%zy8mG6n{k}tSB9j
zCj(=?>!5q>C2G<D&{PqCXacNgRxiFJnXnHMUy~=qLGyHNf%L&`0R0#v*fB@wUKPg6
zVq;FShmO`lx*^_6aUhQv4KdAJmh1%5UC>D#+FuW?7dJou%hbAN8@M~?#ixHE?Flyh
zg=9MeD1eU0uA(1WIG$CKnbfh!fZ|(Vc!H1E0!sz3Y+o$BC55PHIF0njLkwIo1Q&8w
zFK!$L{uHK5lvSOg4IRx%qI?Fd{A*8w0!z9aCRLNa1$c@nM}dPR7c4&Ke~{Z)rxp0J
zpqHYG2fY;V$z8{j0R8d=M~RhY`m#_Xo=i6)#5l$pXm0?Sn~Ok`Mh`eD9v#ARXwqmU
zF<Zf%o(!%abP@n37r25}{cPEaFP9aD;5#26geLC;`{ho68EKk>uLrbT+6M3{@55Hu
zvgPLn#Lz#(8-GhiyB7Pcq&5K(U=$D9zolR&GO4m?mIWy2H6%0<%-Al_Is1W)IKLb|
zrvO6hWU#oay%aG;V!S^rU>U0hrHw^9>w&0R6S74J{~{IaU1KU7LKvr?2`c0LWrYts
zD)J#*>J6W2Cjt|-S+*C-C#C=nT7fqZW%53YPN6tt`8-&oUj_~aLVA@Vw`lt`fYbOU
za2f{(ucWrSz`Ko?zZ-DW+e_xF!f#~&EmUa&{MrsSZlpzQQ55i?wFf|x-#k}RYlBJX
zwRovc>o7|hfKk8*bBGjmP~96glLujIl&qmL<R^nfTK6R!p64)XnADr{kwQ=)6@Gae
zq}rMSaWLe!9iV+?&s*}>c&QeM%G?-v0<ZgEEvLT$L_gD*^nWe=^B)_UzV(W)fSq+D
zp3`y-ON{X}$p+SI=+0s-kXqU4l&_?z;tADVzH5#~8*gZGii{8%_xaYHvdYr6xFnrU
zH3zo>kn}N}{=KkuAQ`Txys!ol?*IGhU+#m#P#V@QfHM3<T(yWdj?nf}cT>kqL$jlN
z$wh>A&Gw7MHq{Hb>h&gfORw$$e9e?gRYnReo5`Q@$}{9GDN@xS@pb%K26G?YE14rT
z!Y_!Hmi$8UBw9s>I358D@7<MMMCS3T80os!A8q>5)ebi-n-yZ=M@)m{`p96ovKlyO
z&rm1_LZOP%9#1IqBx{&n@FinbgQcb{0pVY5v#f{P7IWh1V6iK;Jb_c35i!ZeBZ<ez
zcevz1JTY>-;JSgn(DCDilPA`FPM*4(QXGTB;Y-yOi5PY}l^Ayr`K}aVxmcEp(CwSs
z#clam+F*~pD8Gz4b)j@3J-}PwUEp|lWN)$y@74bNGc_+mC!|_L26b2r0T(m$VJQqy
zx^pC6(DjvRe`cU}8moe*0mwSE)J?y7=3D<(&!Vj2QEkb$tbaOmGxM?c&8r*d{#3Sd
zM91MbHX>4R9mw!-pz$1i_e~1=fg2qOKJ{C2IPs2#fo_-++2e^-Y{pF`)~s4|V&NR7
zQJufG>5f{9Swi!Q_1P8It=C;#fAq_}N%Nh4e(9r^{_cGHs3%oxNYc+q6dQbvh%ST*
z)@ir#I2UFU?Tin#)MF?*#kgUSXZi!y4XmIzM$FxFZm9CZ`ic9UQeqYM?TNIY3P`fm
z5VFb_#@&s4{!V~D$S*4fn+6@MEJc*-%i-Tcrb1*q1Gg0!ip%g+AukeD1s)gu=Wb?L
zZc6aJp@36TU18P4&SvVVqqhCFvplovj8a=t=g02Y;kEy#DXOvX^mT{fq~Y&SWCBIH
zh{G+o0NMWG^G2$^C5_#gBAsAJj8RGHd>SodcQ40|Dd%fx<ySb@*mHO)l{p2MtmdJ&
zWuHH;@gJ1OA|^lxz`=IK;Pdi|5r}#TEnGufiDwF`MoKO5$adODhIw(lhS=6U33n_?
z)T`~MQoQOn)NictzTIh}9%u6E_3IfY6!b#96sr({{Pi&>%8(^`7R47ogxo7MxIV&f
zi>%ApnuRkMdBVI>%A^B2i|`ZeX?uf%N)z?#T=&)br@4LH;#rt~bo-8kHPMQbw^Qe|
zvo9;2NTO6zng)rnoe-E!k;M{Iy{SqBiKwGavYaY4dV7d|sGmKC!Ro6vSn{fU%LR{u
zIJLNH8x@OF55Bv<Xs-)dkwJi00lypqH=_eD2UlH-LRckH<oh*x6gFRADpd~?)9`SK
z6`b3q0@Yir2F?_zs%W0Q0`5|ds<{XC_}E})oSrp%9lPj!&8fcf&}5IX@8aGsecZp9
z(q@SNL<V|f;w+61d=ME>j$RmISth#-XU_%$JJ=169Yd2Y;XLOk$-&YmdEF=$ix$nL
zTTok&-M^4x>~p?XSLDASRPuwrmA$28-c|fWZ;le1$r3oe6eVvOh2zaT*)|n_#&1=1
z)D*`j;djQQJj$x8Up%4k^qz~eRYyPHb#d(+MSj%}ma-y43_2ln4a!~4@jy>MwJEa?
zK_2B5(<9JH{aQgScDkj7I$8OV+cNWNb)7QyhJSc`RQa&x{lab7vfh<cfZhBH1Z7Zw
zyrEQoh6F*I4&e^aXZQ(X)2|&M1m8)Y+2YYPX#Y7)QbQ62GCE*yxnmIiB>?;BJQMO%
zS)kQM=b#fAY^PfPo`CI79~PRLoy$mdtq8u$+chyD%((7-U*(J`qu!raTA4Hs`N7K}
zTUZmBy0|~4CCCUXq)7F!+-v-p2!d3t`<6LFU^hc&#Zv3)8F9|ZnOW;~Y@D__fBb1`
z<Hr`TZaOjlUwcpHFehXp(;JH^cT>mXu`jejEaMz=@&gy2HVu(_!mIcp%Z&Z&a|?3<
zc5F|WRB~fpamyV(rH=LkMQl!wU}LTb4k2O;OU{j6xKNf|XyB8mDoTkim@hz7Kj+1Z
z?Yti_f=U8=B4X0K12*JtsPw(-ofmI%DcocBT&EEfa6m8WOa)m{20H#6NTK4WTRpOu
zdQzh~H}9smS>e546E<^y0br7qgDP6?t$RwUIP7Qi=*3EI0FY5b0Q@7rK!G>{{0wl#
zh_QYWgm?>em%`CZsw~5D4z3~aEQ&b|WZE!VOX+h%{ozixDOgNIPE?wq_1pf>AF={$
zSC|#GX?aa<ZCo3o?hY+PG<i9TP*woKi7FAeWJ)GfDTrL#pmY*&ao&@FzsOyU?B+Dp
zh+oLE5KnXjGe~xuSk9P-s55uqIqfbk*Vw9fO^IQUzgP1e=y`u@J*&c{rFiPxNh)`h
zmdy@9Rziph!{BUqLyOTc6ko_0hDQ+dIsB(c4yOmicx0&T=>Q#kMPdwBWM9I5#O`jr
zBQ?P&F$-8~Oyh+4hT~fBys|*%VBzfg_Rz0d_tt#M@K^J5T^#RY7CQNGI7yC%{IQ`B
z^fsr8TV(f8->QkYdnqT!und_jbr)%b8AUxK7n`M=>4*r5)?HVa@A4&I%VFH2@Ccu5
z=btB(Gp30*Q(~dn&%ViC+$np2?A3ND4eagC6s44Mlm_Tpo%MnCXXwniheuBOVip`;
z)nZ_9==Ix8n#&fVfgz)c1A%(5;jdOD&VqOjNctEc9wkJx1M!NDZH+k%>a52@f>;Ab
z0c%B+Af}{g?=!56kzagX@9FHvGqNM<X}R8g<`ycSdPer8EC=r4hQDxAT;#NdBO$~H
z^<h=ARhh=3OZ>?0t5}7gql80#|BM$Qb-Cf}w&uRK>7=!LFS>YM{P@l}d2h;*Jo1#O
zuR>T?-jZjv^F}c2g=m<C9X41VI4k9}g;u-v%x_krg<-4*;+?q{Q0k@Q?nmywsvIY}
zcusMyE@yBhG6(s=Vdq0_hdUDCU)Z@ec~7c053CsXi=FkxXF0V87v8MtJsH=K>u|d=
ztz=&--(qR$w9{<imX*}>orkV_-14s1+;n}!&dwp%*!1grd?u6}tKYHN#@A=0`y=a-
z>oqA~@AfAtXDfWSYOlg)G)O0<3IoncZAIjAG&8;koy4dp(|q>nnb0n1^TTI8tuBt7
z^A!QXM=Orrud$!8;N!}X{ebTK@g^$%zc2sE{~aIC<zcG;KrxE~j@V>8pp2l%qAVwQ
zxy{4wWbVnpgPb2deCG1Z(6MIWq^`1|ri9OT6PkraaZ9WeF7Gpe=8y-&@HprS9Y)Fh
zv1?x5lJXybk{|sakQMae7Rkg`S3Uym7gHcxTdN?2(Kvr-#D^mFe;+x+UlYIn;Uigp
zHHYc%<tm2pQt;l8=t211YygUsB9d8hlt$gX&E#*}|C-c+!iOdiP$AI21A$IvMmbyF
zzw-@9P`*H84=A6@@eFxK=xM+-0Gi@1G}@6+?i)F=bTj>n0BU1UT~Qp~y}<J4G)U1Z
zA>9v{W<WDnYy-(6Ka3(hl?eIKOj`hCdOk`T580h!Rk7Uv19It96v(A*<>b<vlPKpQ
zv2hvYw^0FQ(hc#_zME8;S(lN5#2!lX^C9!_U!4mzEGNH-%is8m`o}S1!%a>xRrvB$
zLRm`!`4jFU)*ROBHg#ESfi?fS>Ux~bH&pvV%mY@QB~_(M?eBKF%`o%7)@<wVcgWY>
zWlv<ZGMRn_`rxJcXo(AoInjowEEO<lH>%O^9)O%&lLxMFIDjVy!CTeT*8jC+{r~qL
zRUpEt*lKn&ubt8ZYGPz3t}n5XGC1#Ykkg6F%m&2+pHAGezTtfY;&7fUk=H$RaFXq`
zDKWji7xHx0^T(28`;sW-NJ|Sb+C?%F<*qAS=*`iBzS40Z?M|ZlSV1Z>cIPhKz1*=)
z>!It+(ma<ku}S^6wciv6Z@-HBYW|R1%n^;zkWTzUoHRU3!vl=z@)+d<ofr>N0A;in
z9Bd;+jrDqn)}3~P!C9S+vy&Jpn?+mkjM8}49Ph72X{%eBwbnj~pgy`^wy5hyUZP*^
z9CBaQFkD#zPSZ$=><kOCINX&$ja)cN_A}xYZR%QAIQ2{02ob?SJq4AAP|5+{yLe#H
zt9Coaz<kE-3%cuP`|0gjJbjtoZpfg)4V`cS<?3)gufn0}C(>`naCC<VCmG_p4YJX+
zDIDzU0UkJHWa+9k@R5NaXv*yxgmxw5@*oO+fKjEIY4D9W{BS!unrVeqohURY78;IZ
z#IF{)g&S0rx^Df{W<V<*iq5^4vVKLr^Sr9Dt}UypSG~TudeoQ?JtX3M<TQZ9pdH`e
zD%qSuJ}AQD8g&6WQ@#AIqEr|0W&`_wGmO0mD5cBnjk`!9&EGM9vS3*VhyNJKL#3*Y
z0`wGX0`b~%Hl`lU=czJc3mose;3>~@EoVOqcIBJ&=sCul{$#QFyQj{&txI13i^5m1
zB2n~V%0KH~E=Ou)W-o^VfNnXJ3&9?623krX*;#CT3lh_>fFJPR{b5Bh8RQ965ok2-
z+FwX?3O*;VsH&-093r9@`#K2NXZoocv@S7A2lr3;teO1YOFZGd^A!I%hWjtha@H;=
z9UX9PJeaW08h=gHKwcxDs2wOAE%Jnf?G1}9KCB5M^(bMq=6azjS0^>;z>cTsiQiT=
zI^VR-U$=AJ>5@}pr|l0xZkf|?b6ar4J$>oklo;0d(Z0JQ+NSeHXUFZCHu1sXP+qgH
z-n_Fvk4yY%=@}B6nArki3_TxjCGvr-D&sHxTK5qeaPS3&G73kYUVPh$K&>BG|NiZt
zm9CfNqQ3NVG|+bC*FD8CvYW;R%G>d)888R##9cRr-|EQXW*2XQ8sTL0A;vE6qEq_?
z-RmV6$?h|zC$bljUm7V4K@dAq0;_HuCUe}#i%c}c_ed5pE<H~$3}+q|x}cN4JY=Lk
z7rW`_$2;a`1x>k?_w-5riKLS*Pn0HaRc1CNQ8>sC-jGsB$L9n{Cl87%;PGs49>s{$
zoKm8rT;F3k#jR^+WH@A5#%DpB>ETo>aZuK=`aDym&YhMMX<8KBDF8S%oTK>Fuu-2F
zJ6U3ba!pwa574xb{jNafhMe=KtGF??_H~pH+Ib>_FmH{@r;O^dPhJye>wkSso!_YA
zr6`{N7h5c$ib2#ubBRV8whKDXpVP2S)D{j6qo3yNN6(SB=$OBgjuI%_kH8Z<Buc!V
zA+*tAy}Mt4pL=fm)e)=O_DXM^-2>ytQ2ZneZXE!`vqqVr)I0%>VY<8(`EP0uEf$2K
zVVhxU6-!k_jg)S`#ZVRY2j)fOJeyyl=~5|j-yWw|d(;Ov*3ysD-8=F^T4lkSc0Kau
zZ5J*=fp|@gk;GTmuBJbAQHA#%r)F>Q{hnv_&9B<by%e8tk4;>C<4VkxZ)IL121w{~
zRF%W8r{j8qrB$a`#yGp3q9}EKEOG?KiBvqRWNDMsp>9FXth#~xm-LKVxj8PgT^Kus
zo{LK!tXin%4$hnMk6%mHBa4Gg15=yX`<xz0%ZlVi893PEo><;$8U^1}wX^nPz40=m
z4++jiu3Zj#M_ku`+Su`WpTpU4<CxE>E8haGnVdaqBv|jzAW#6SS!S(}-OP>&FPzZ@
z-MVz9X%b^=P?ZWk1WM=Wo@xtI`mEfmnw|CHnzAgtt&8G(pT9aj-5utKm;YEB<B&!T
zy8?VXoJZwSRn_+1?hZ^Mc6!I$eiIK6+81$D8CfMFMrx^srT&v^j@43orFidIX{E0h
z<z&zI)}3FH6oFm=P|?+23Az4iSSI;ngH+vJoPuXb)ZsnMP%m0D!VeZ-NMUZo{gaw#
z_@QsE)i>+=A0N`^oN29mbIs_2&5=Tv+{xE}I{pjksG{`i4{y<3&R!HZqY`5QFd@#B
zC9)>q-^zGnzB08$8p4!ti@8FVU_`J#eov6eh13HRTiR@V81a28vMaSFT$;Do`UM&K
z_-gL|hX?-)$sENOz~~20aFG_sqEP-#PPS>c0J8>LH!iz30v(?tF{h|Ff8nTPx6-#o
zv;|E`pSG*Z!L4{!#nfr%Utd)`XEulY5y&?GJts>+b^_%-;$Yi)n=6R1!C0~Qs>r^_
zA)R6CQdc;I@vW$kilyQXMRA4d5!&WEut9Njb8D&3;hz@XHeEhf_xyIt%}Pt2Q){>X
zPPf83h;rRH-sl-5vsbE`B2JVg6XV=E1JRPI@Iu8)B|2&a_1Q0St}j$sL+>#tcJ0hs
zuNJW7yC3qksa7@Y0?^2Oul$iaNwp$g^01>EKz|8&J;oSS5P8;uj$LdoDzwD!@i}7|
z5z0vTPS@-2pRUbIYj)z<(2qV1GcKs4FFo~X``+jX#S@PeWvhmp(SHl1kUx$u9d?J`
zWDLg)7ds@A&?bEmkGP|<l3Mh~k*9wMZPXM94uTz&m!!q}hVTpkU}r0U692Vn3)~TT
z-7^Ns)CXV49Lk=>o5>h$_?xnMAe(o1^1-#?Z1INSkw2++0N_dfJ<1ywGXZ@;c9G8z
zx(zdg{D9DIEvp9UJ3`H-$l`j>@P-n7Y88Y=|6HO6Xzn|EsV4Z90B52A%X>-*A~Obh
zzM+n{<d1_=^<-kp7RgaLWQa<{${F0|{;L>*1@LihCR@JHrkF3n44E-JNd^GmKLr~K
z#37*h4v^FS`9Of9RR)uue%=X?x4FN*u`4|2zmxk;z6A7Km2?`EBcNdV5a~V$`QL_7
zB$oQ{^3a>?VSu!rshqyyQ3kLEuAZYPx&8YFQ%Ym-WkGWOdq8B*xdvzV6r#Xr|G$Nx
z_(Ol^u-HH%ke^1AhyF10$I)IxI$IQ33WqeoQrnHGi?wA6(q^`B6OLU7QcCuH(B-);
zA*jCg&gqLzA1!w-Sb1;L`ZsGxEGys`m`u$yl}#Hg2Of5YBMe&udRZLwW+=iBqbfj(
z2$Tg&QECe{BOf8P#LixrQG!muO}A%?60apU?d%lqw24fr6FkWZHu2tf{6m6+)`Jt9
zr@wfKfOmc!X0xK0H4oRP7iZve2YbqIv%n`%V?vMr2*yr`+sb1sJZ3)=n7e2_b2eno
zPbm)YKKQVUqWvUP@L~Q^^|quDPd;oWi{X4HY-5%1`u%v?9L!2Ulb8NZKu0iw9cbuI
zk&I^*q1i}}J)aiC(J0t1%s8nt4xQvRq{K;P9OJF4O$YXaMynv>+ZFF^COS9et$xs?
z*p);r8t!LQ;zfXV$e8$9)99<nI@Kp|$5o>XA#TeudeeIL9VL^V>fOGqdiKQ6`+$zq
zh-m_wvy|gb(PWxol(Tfy8$D-|m$(3P3}@N~_$q}9_J_80xW0Yhzn&kS@OkMPwLJT2
z9wB?rj3L2`{OCOtiRpI%v--KC&>Th<7jdNV5MEeqk*K2Ys+h+>hRNZ8)`GrniX2JM
z?Z9W_`sLkE_>?AhLB|<m-W1$XFer`}AL9sZgKS;hzDYK8^emCx(b5m{Z{*GEcE12L
zdT*gp3hX?P>m^fI?+|PQueqw2O@bCeE!?NQm+}Qs>5rLOnTqFcYgmvoCBvWGGAsLV
zXQsvMk<(IDIe>?l_zOvBY~SKeL`U%*2{oMs(31cR)7L>h<RRQ^CYY7bCts*za$=%*
zDi##%EB$2Pq0Bp-6Y(X#khT}vtzEJIbL+;z6o=@kIwt1s6ZcJ6&V{H_lO*eg-y8ui
zI@eNOLviMKAyMqng$B}<qR8S_s$j4<h&r`%C?@Pd|Cg4Qik7s!uFHE4Ic&YSEN)b^
zV#Hcre&}dP61Cp*2fJorEj|fP-z1#{$9AAllLsH*z}X*j%55>;DsmLm-$-E^*Dw`D
z8!9$c7w8_p!`v89lVkqnTBVZnfZEf{GhCf_?sjM2wCy3OJf|jdM4Bl*nE=8r#peG)
z8cR`^4~Lex6JyF@W8GNwNUnWPsl=F7K`e#Q>lEAvLIO>eo)f;*r6GHTPsydOW{=pq
zh#SWmOeVChGWc-a?VZ|!TXT0DvVyJXjR>(wZs)|%33a9_o+nzmAv=##OHiezqLj1!
z60N+WIR(#Ph)9mg-sa|y?MkadkEcyq?yRu;Y^jw6z&WabipN3R_!3x3_TrNph=^a8
zml_0XyG=Cb<pvz1OBPF&x8>(1HZ=iHbmgn8l6OnTwOlwq?ZSn#1C=CHr_EC2-%&!*
z#v95L!m~{R2on+gl+-oY`{3((abBWcy`k~7Yd*)A=GG${GSoJ<A6~PyQ4ey{|D>Je
zogY@vaRlKUKCh!^c{`i|qHYlc_241u$wIA)c5^!it~fBbM^@TJGk%zxuwv~xq0`vt
z(%8I?2H;uza_!HzLT9n)8!=jh`4K8_aG(3bC+SH&ZE1p6i5&T@mF$RqI!S7mS!2W*
z-F2*O*|D08JB0;rKeD%+b8+ga<4;_>{mS_g1!$-2fta=q@r@ZWF+4%TH#4&2^SsGb
z>>RpO`?|yh&M}J(PEk{FDJQFeTSZ~e+bLlg1ydjR1X|_^od@hpUJ>i`E`EBYb$3m{
zr5h(!J-9XF<;!WBBx3Vt#kZ_U_}VHVxs(>+rdYdhf7U}$UI|r+v3BTd)An?OgVOcF
z6{c&SJjrt`GJ0zr*jJK%EPlI6h#e4^Y`O{+A7@snY4agA81w9o^6f$STDB?M-!kXt
z>=cgv=&`G{gY-mkML5Y7CQ;aTZ&E<88lLuJh$qhfrY7D5wdaLAuvBOI;CWchhWy8z
z8cV8lohT*DNb;q0_0xh!Si>oQm6^?-KI*>qvA~*cUA(V-qAB!)!b%-rNotcLG@=NV
zRkBOy!Hh;fv5)Mog{Q=dxd&@F!!nXCdnAg>OZpsW8mSMoBQ4jhH)s3!OifSUDqJ_&
zU}Eg;?aZw4O}i<%sBmmd<2B+OtcDnE0hdJ%togICk!Kg?A$#dV%X`o9@aY(MG_!^y
zsejh?e^X0R@X%!OY$$~%;&8~J6o~iwT!Bs<ge3qhW*eocg&5;3nGGj8Sj0Zdo+L$u
zDn7!s7x*vxV>CJgXC+>r9Cu^t@rCa3n-ZMQer9Z+KBKo5d<yWfs9vbN*UFcsg>o&?
z1BN74E#NemhBu(I7>d~0oUClhq+PBp-Wu%Y9lZVVJxiVE@48d^(IUrUpd1QscObzM
z4Hpzv2lo^FFme=vmjW;S4|EFh16|w9GzBI08adcIRk|9`t-a{}m=oe$)_hS)c!6<%
zpAtuSC$X)?wPC^0CmrE#DiKaMpL*-6^u=%1f)e8AIziedSD)fUYQ6sIGrlgXt8<no
z$IQ}=BZ`#oO;q+cqx)+65x)??kGK58VQHUFC`W?!m_by-U><5q%ywBkp&?bn_q21A
zotVoo>e=Lm0>@OJ`oOF<{C-hTe*XT*O{aULQ30nGEEwN6ruApYr2w+;2X7;X@{QUL
zXOzbTM=dl@x*m5C847!g0Y+e6OH9VW^FI{p&Tr}2G}N&9QQZI|=Xm-do<p?W`u2nw
znhVg`MeF5G#)n~7nLGf;s}z8~+TIJKZA=@;m#u>gxV@L+2Z->+78t$fi)0~llPu06
zSZGHta+{3bxWm2El>lkJ-8oO{x@YD+v5mc0tyOLtaO8#B_p>B2Tr$*<BN}DXZ3orZ
z6`~dhR(0UmpqCsX{}3wX<=Bl%Ea9H*5#<;2vOUY9-FM-x?b=#pJ9ot2^Gck0u80_1
z`^4pnZ~BLbkGI#&@=(@Yd;59kSw)ysJM4LY>)dfevJ2(5p*djMB@#R75qy@Qx2c9-
zn0b@A5c-wumAnJ*+)igC&X-6G)CYQ!p5&h!y>YI^-5a-@-aHT78)E7ILuJ+{lLW9D
zSxaQkkOPLNc{xZ-f$2?#Z7EI7c#g4w!%D1s(9vwo$@4C^pWeRJa3*j3?ah<+ZT9;D
zzJk^8`OAvoR5rl9@gvOOPhE={A|Eaw8dz#l1V8O5@+(O&J?$UFp4w$rdUnYFP4peq
zqj=QC&I@J*TT3M`$K6`QBI&}!b>@$j2f31o_gO8}{)4m;94txBK`{j~W4ha9xya|n
z!cnNe3c6eugCn%|p5!3<E|UotxhSlZE!_)=#~)mxzyK}>H)NOh;tLH+GT?`~3s}pg
za_B^sz_ACtONl64BpM9yU*WSw6cB0j^1^^>@J;6~QF)5}(OHk8N4$SJyZ(^Xh!6zc
z_|KY4p_4NTv`1A;V=oXAWkganj@ma{x4YeYz%&)booHWSx{32vx>!55=j(~PR(D;R
zTP+;z=2;n?pSSK^s@Xlg5i6m6;`kq981#?r*>Z<oUU0%8YZ%tT`A~zn+e-{}mO&{;
z=S>Jw3IF{UXf)@*7`43T2w+m@115Dm+9`YmBo-BXs`NZ3lqX(aI1SLc)a1A12fr_j
z{5=5Wev|`V;e4QFx?JAj@nP@q+9Z~+wt2ud&++C+rawi9+p6FWyFx+C0ThG>OyjWx
zGHxgNtxmDq1cVjGLJ$}5XF>nQ_cp2m$>Rlp?)n}IL7D+S^q&w@{_mk4wJ}~g^#sIR
zFg6R~JoR>noB+DhSqVd+GNjs|lIQ${_d-*MXE9G^221)rW%!N$CVHoS1?4Sp&~79^
zG4~x~XIBBOc@%Z%XXr-z2Ju<8qjW+Lu|+{Xpo(0L_7A4Rq<8<{Qx6cylKTu{2LKQC
z%7_V4;_;;z83L%@_ke$F>jU}MFLq$=&;eJRA=_va2xcz;fj?fH=4Ze-<mIq6LGwso
ze$N#Ak8@D|AcoK8VRQ|7udWuvu9;*aVPGsQ&|}{3)ddhIxs?9$$F5U3bSg9-F58Pj
z4fq!@+MXvW13BHDLc=qTTXBuzLL&f##n_5w$KG~X5GXaA*A$!}igU$$XRx(gau<Ii
zv78_TJ;YK3n+Rr%P#+@8Sp_JcZHduoU>`06o`(NWXAEA`uC2irXJXNo6Bwzhu-D?i
z!mK?HEYpYT&l|6K65j0`H~A#1Gj7xR51wEXdLZ*T36<vt5<I&hI`L>Prw)ZR!J#|>
z5}OR^M94dt-a)wroP*Iz#dqoowYw12o&^>P?X*#B<<5lI^mu>0#08mW-WiUn`!r+B
zfxQbqZ~8=5gj)<x`@wnvYR*qaG5-#b_d}dUPCKXjb_(%|n2W9GC_}~|5xYfpgHc94
zUtG1DzqWPsEZynyB*A0Eh<x1+gT%uTBx1p5`Su!1k$cgqw6_Csr(Qo@n{lUW|K#aY
z*MxU#tetL8dNHGM3}00~F}@H1ql^X8PcSF*p|oUT-&j-Tm?1D6jnj(_8Hq-3w=CpY
zy?c;-=gDf%aZ$^2mfU~%{>H<ftdhp(6S{sFRW_9tjbM3i5jpLGNQ)f;y+*I(aQ5&*
zxMo1y04n>v6i#BTW4pc7pq~P?57c^|RMl483Q&>IQ>+(M74$%>Dc6xhc|gp#?)VDJ
zy%9{m+X~NUil^<j^k~i9bLlm*sFCs64+10zA1k0VK4-WBPJs$|1R^hh>kj9Es7-^?
zsapVM_e=tYwrSKm7#}_-5vIX*g&>PzU(-35eWfii)<>d&XNb2}GlTbF_1{Vo7h;j8
z<4eur*nan#Tl6Cqrl-3#r5Px0tn?R$+<(63*oO{CfS&KgLzkBG`UB-<^CDfEypdjE
zV6b}(cnh(NDzTv~6KbAS%?T25I>tVmW;upIRF&oFXk=k=IUfFLHa;G~UI%aFUthh*
z{Mtie-_^~kQ<lF9L4E&qo9f74H9Cx&{GvsSLE*!m;d#W^HInJLmN>1N1C>+h@+?te
zs!(OIg~2@r&9OH9LCdNqXPRyY<*)JDwP_zPNu@9H1C;s=L^RSfhVywqK4tObT8_L!
zUw%Nk7I67&?(s9EGmQbMWhze^Ol|KST44{?b0YZBa^8R8=qf7uu1h|uZFDp&BYN1k
zUdxaWF%DcBz8dloFzU<9P=zSV1CjC${=m+oo-NqpDcRA;bUg*IHm;>w7Q4vt>FRIP
z9{ZlF89C?UzO7O`RdCXB|8aoo@ctxFEIJBe!T)RzTg+X@+4~EL%6hev)?Fc_N17%s
zt&=XlmvY9Rs^+a3n|9@G`o||;iPi5CJeQs}xu>Xn_GefqStJ@x;PvpbOr<|ZcFW)b
z=L_tKlmzJV%4dV@e^9)MHtK@3R4?;Vbm~W}p*?u1#2tc+X2@8=K#T=w*I^hWSXrwe
z%O;+)v11uwD6FKBCuYjx=o}B~Wc*bxKR(33LgTP73Q_-@yW?qXTwX)N{yr_|rv2el
z+(tG{Z@pYK^?QgV?D(H<1NbE!d}L?|M-vs$lTctYL4>9VSj~5_cnyq#3<iaE4u(93
zgU=}UakBxQjyW2J@+J|Bym({%X(!(@Ekt$f(2v!|zB?-%uWi2AyD@%?<e;Ck`nq=`
zS<W`*r<Z?l_ab-uLv|760$)h>5R}h>>JK+~;6qc&VMj%Y(CrOJU(BdBPF0!KNnVpJ
zaj;!VTXOnj@rNx`-Fs!0yD85N@#w1OXuk&w^l%YxrJROE0Urd#3b9DoS>{ssmz(TL
z0V1)41!k3>VoiSknLWXXqnXROkH^^7rG2tnK2~d1Ue^B07FS>R^_~0P?nUm)TS<k&
z1s9&V*BoM+IcN`x)rhy&SonBgL*In@;*?n7^;Ko<TXJA8vV2+9%nIAhHS0eGzso+{
zz!s2*yuT@!5Y7GR@-{FLSo?y40DguJL%(b*tI2KLvzQ`4S@KRNR$6FetR3`D`eL#(
z)kQZuXiJLvoms~OwjVG^#I|GNxTAK8*6((1S?aB6dG&VotI?`j%5!|jB~6>4x0p^k
z8`8KeNd?$iXR=nH+z)P}@fo5NA)70S0oR~(2$H%g7XCu}Q5C7h?FM5<`_HvN?!EV}
z3{qRNjB%0weBZXiPpo(I-;yu>$VPy7`!`gtVyK8BRaM>ruXt+!%1NPj7&gNFS0#%&
zHKkja8ng%<{pzpME%#lju<Pif{q>vfB}Zm?`8yn)b!GzdMd$>+Y8<KfN7{k<CSIz3
zi5N2`4dqTJ<}<0ddz#dxnq>gh#^+BPqW9GHN9zo(i`o!lQ!jj%@X9G;RYAtmb+gBZ
z+*Kr#dr2%{r|B7#PNWf|8`5ENxCLF5!hNb=rcD)0V?%&{15XI5q^&KiS9P@Wan#(}
zvMO%gPK}dnqipv4@k(%USN%Yb4foMifOc$Uq!42qCC-(S4L;OtTBPONofUk_B;1gg
z_{^|4Fg^B&UV~s=++dI6naqP>hjt<bP3`hZaq26%dqSbeZWk`$0&A!Oo-caUVRNkn
zwr;FcIfGH}JE_g9r=w2yoq=ZOgPV!FY*gQ<?#;h`|M)T(n2Y|cGx?D@{Idf1rp&{Q
zU0_Jo<}dj%EioHqccliH#9nsqBN}bO8dGaA|G@#%HjO+{CQ%y$jo~>v@}4HV+}D@L
zSRei3m1yN;^v}wg+((5Qsv(P}b9NgxON6}c+xHknRpohXWn-7%C3Lk$=c<(EuIR(Q
z+Ev#k-+nXu)STeQhZC2Uc_|9Vl87Xfe;)LXFYgS>zt4X3o|k<mMzRJE5cof^G!tor
z#zYLzRJsz<US3)mtLyX5f4`<dyjI`cpDNDz0hndLoS1%AG$_+a3FTm?!1D&C8|weK
zE&9=wzkLhka@_RbHcUQX!ojw)CrFuZEth?U+Vkr$p~qlRC8c;MF8n@TIEN#2kzMml
zJunYg09BcK-#pD1ojeS*vdKTXl9ZjeGPv#+7`p9f89SU!?tvfz2nRo30>ElRgj7R4
z!}M(STtUq~kx~jh#i09P*~gEze#<UgQ2L~N;?hcLOU@tTfd9$aqQ8ZRzxm)K8XGr}
zlWIto1T!6nTkgR~!h{IG@chNRU9!AQLOzMA7g`*}R!33U27PD_eEJVZ<r;J?e>Y+`
z`@qcDQnt)0=M~8PnB-cS9ck9@Z;7S{NHw{`@fMgOch3?-nHdZz>)-x+`j=YnDb)cV
z5+?HMgVp{-kr}N4W(woqzyJ3TnE$d{^tagXFDGmLB?wa<y~tpj!rT#gCwmsPp_Dds
z04ku8-}eB*|Gz|2%6A^AOCeO0z=>Un2f&C~7|~_{F~tBC-tZ&HQ638J2aFo9DxgEB
zNa<~eK`txEs!=Xn!oFRQIOIt5h>;M~bG?A(;kS5dj-miYOIJyg+r>~Tc%<CkJ3Pc#
z(EjU|ktLBRcQNN9m73WLB<OG?oj2!^B+&0$?n#Va2Wu%O8mVFvZ_bcUXb7SVPqqgF
z3EuZLc7rfJ80H{s2i@y<`40gfHHwP;G$^kie}E6%VoF>;+6B+m>>>>fis2uCB)!Z5
zBq^BpS~m2f4x0bR{y_Zq$No$FKHSOg;0iqZg_LV3RcjD8=gDpmN{lc;`c?jm!l?{f
zZVz4fEcj@vy~Nf!blumVT`nV!_ZC+$rEOtpB-RYHL<^AiRYTJ(A>0tqK;3CUM-;Eo
zRJbr2<n`uf9ShtRbl}Ky9i^<9vQ+<QbD?4JcNWD4gmX=7dXzxf5MP@!W#IIs-8oNv
z_&Z0xFc*G*{2ZW39N`iKpPeBa!-I$xyKrgPBP*ijq2-41{7^lMg4uM#xs4=3!8s~4
zmR+&1y33@ATC2)NqXLcUuLaA66{Ds4DaE(mFGsW6?%X)oYvlBO#*BQZAjoa3DZ;<a
z?q9~F$bTLA)dPP5YlX_#jbc8G*A@NdgsW8Hb3J9_sfw_UjF{iO@Ixrq7Ww=W4(}Y!
zM29EthWQa=oqr)&M?MF?e17|Iu+l-(K>5`uwES^1!rKr$Lr7O=PQ6yl8(*NGh8OQV
zQ(o$A{2}2`n7hf1V{z|Bf79z3H%@nxsfiK{ou$4SAbELf9vpg;ddFM6bj_R8yUyOZ
zAKy%{5B8g!y{5`z(uUhHd#VnQh_r%Pu-_<bj13G-q>aP#?L-0G)inhU0-09Hu`x*0
zmf5xjes;OFJv%)52<IKjmnRu|jy#q$K8%<So>6EV`c{A;V0Q7}EQJzGhRq@(DWDE;
z$MWC9mOpJ8wT&*FD4)*}FB!>tj*!v)hEeee+HhLZk^WUZLG1Bu%008~d$f4j)?%Lv
zCuXbm?*1uwj^cp&Uj>y45|E{l1+b=tL=79;-fAd}NRwFlSdOYrBh<Q}eL*;}UEAYm
zN}TWBmL8L`*Gt&OtEMfd9zA;)@|EsCI$7%5NU0VFo~kQ*z~RS2X@*{A7)6EotC3Gr
zA`61BSH;|+)ov3SG`1DQY|X9rsMQNP73)|L`SK^l#Iy@?U%Y#)JdY+<BYz|Nr9eF4
zVLoKIq<kA$G<yQ4zZ0hByc<TmkU!9*lPQGqEQvA7oijZA$jd;hTr)#zK<}aBa4`gw
zYOQU3a+<~Vj-G2y$DPYxCKmQ9e4h$g*9&mM#^~owI1i%9vuvPRw45t@%<+kdM6sg*
z#nEK(Vp(YS09zRd?;V8`MPVfxZ1o=dh@*3vrsH(&ybktGdmJ}dwQBV;Nu(vTd2ENu
z{NIygDF6KVt}Fg+;Zb~7NI|(bsMr=@t#QF!+M``cd877K7fcmtadqa`R`W{l4pLI^
zfRg6@8G+~Z<G;PGuRn4s9{>478zd73H{@Qd*jRJ$BR8(P=uPvglzIC?+vc#nznaeF
z_7A-s)&5oniZa0wGRQ`b#8^X-c{{s|dTK}`iltn7lkrVRKWDM$mQ8@L<XqPZ(-M|m
zt-p`n<FhyVI$VMtm)v8=a;FJHEMk86I_Oef4%*FYr#C5!qwqY5k<{i%b+5%?%&#=w
zk|HvAbNaCO**xD3&CT_lZDk!XX4HjUi>Qh}pF12!`rb91e*Le^%b{{O*g;QuO3!@=
zPXz+NZGav+`VUJ$#`m&72>OCJIdn*Qo<@f*b*U-zO(vo1*)Zcmiwp&XW?oi6QB`w*
z06kr(>I^8ws>p^CH|^Ypf(>yyvv=4XHh%0<vG-)@LfqwJ)kdAAE2(xrI%P@&yAdYD
zTVW1_%01lB7cNq@EU%xZ#<<jeE>fuNCzQ3_#<J`!b+J^2pOD#k-)m9%y%{STGFw`f
z<wsJVGROBT48%*<f^d(l1NCYSbKWK`k;mP{S`J^>i_a?u5CQ03XZ)iCk=npK-;62V
zI}M8|5pKE|>lyW|QP%DE755Ib>4ojc3tIQmeZkDOCE9%^h2-VtE0ODFf60o2k)>)d
zS!*ulNsNib#v3+kwI^N|`r31|7+Gtvz!u!W-Yo8pvB4p0?c{4eS@kJ6e6nQGm~Y8^
z&X1Nmlp>$!8FmfX!*01IzJU8l7BlB!Ri!>uZRW<>*)`eb850<XyR)y{2_--F^FOld
zRKwnBla^Z?hTA{zFGv}FcMNTKjqSvE(6*sL(vh+_%Mr{eB6_&x;`fYe0(P<CI0mwd
zy$Q>za$k5mxzWb=h&YQpVx8^1M*M@{8!yF9d;Dw+F$QF1RX`b-AiaIV83QSFOu8a+
z6_ID!BN2Up{(T4UG$$CP;P%A(?&o8L)3Z%>X8*+Xn4BM#x_5na-o58lx``kR3H=JP
z`zUt=Wbycred&cQluUY0R8gl%5s|)w=rTvOz@>fEW9&|7bKF(gwRF2_ad!_Xp6CdP
ze7%vG^5g|Dn+5$)X(6yOemon4K)H!WjLjF&&*&_Wv5&?~?KBH0ikaYTsVy48DBj?^
z(3=iEgJeU~gSso<GZ#!VA0JZ=_=36(!-3;T=oR2#n+?OjGTYfVReux5;k*?5(pGFm
zOq7Kw|2wc}?e^E+*5x+mNnAa*zH##Gkr!t#wPbC!YDX54`N)sX7^`Y_Ae7T2b0BY1
zI0kZUW1=LByl4}S(#~;m63?@=nHIIufSV>U0&}|q&dIgvcAi`J@A3Be(6nLPN$NGl
z$4QhD6sD>F*%#=gKI)Zf+Jl?}0!y-%IT5@8+Hc+fF&jt9G;4QQ&R?IQS98>0e?8G-
zuy-QScPv7EVv5!^MP2Am>`(aLe7Qf2NFDZX{|bwU`U1UcjGVd!Oy52&E>uTcq58A-
z9~F)l2)3E>eF?QWAEJj6j%>|v@nGCe(if)UFC<-FD5u*O>mU+=Sy1-A5b|U%m2m@j
z)HNUry)m?c6j=F>5F<W7s<Q=(JqsaEeV#rJ{oDn2e2$_F(us1~s)^u7S4lMG+37oH
zNJPKauaf`YVpy?J;Kl#FS!aLvRPc#O8uqv$S$bO{mVitB4en3pwbe<Km&3&qN<TP-
z^3vH0j&~JIMKvXoHH;Kw?6+jhRsMPCix>&KRbxYYmCOi=SKXWCK*RqG2Kkp0fE9>W
z2u7#K{d@3xA8nx!*Zl!wl>*Tr99~!<^3jmcgKtnUpI=BD<F_eD9}gcRdXx<mzR>g5
zLML7#Cy;?fa_)pR8vxS+J1;VXh$03eib!N=yD`lD(<ZSxQEo2Sk1MctNI;iW+JcGk
z1M(l$&Ez5u=D_}LQ{EE}hpbV~8}f>S*sAPz59l{XNWODx5FjTa*3o-llRRiJ@#+*b
z;@HZAA6OX5<sp5is1xN=evKi_z4i)zW9|WxOrE}EaO7d!L!K=56w==S(~l){hj(tc
zs4yHQlK<Fyrr1|3E&;vxns|;pj4a)U8%Q=t_vK>xB~y*GO-oXfZBts`XI2~UGxYHL
zWTkay)^>gR$N*4^A1V0hd}1`5tj%I@4nu>%2mxB?+da631(@S04qw32hN}qi%E<Ig
zB-=0?@nnzrQaBb*x!d%(+OfE8hG_Of|JIo=tMSz}TGx$E7f;_!9?q#jma1e44Rf*|
zGI8NawL5v>jv@D`xN#OS#$0j;T(uR57urk%U7Ujy1;s=bYKsPo?3Ecl#TqrV7B>ST
zBZYClV@rOnvcld7a~tO+jpSuh%OKZz;P-n@`58G#9f4mG26Sgitfc0c|LI!k_IBFI
z#AN^_%l_!G_F3IPuu|WfvPSbg+bS&XhGaf(eFLhZ4+5rB`#AiE?EZs!9NZjctbZi=
ze5PW)(ArT|Z1y!uXy-0I_$;TIiKKQuZM%2G|N4sX!t1x%tkK}*n?JDUHMco0Qrh%>
z9N-=Y{rFUTSvxUi1y=JMcIL1U#yc#muoHQPSj++*4xGKoceboD@>Ie6<R-73br<|Y
zQm?sgDEc(w)v?TtXG7=6+n`Qe<H*T+{L~R#zZM-v&Zq{+sVUCw1%o$o2qG&vh+rn{
zTPRek-@SGM!DWLkX;%Z^(~gHh%lq;us#LWL8HN%3=v4c9;8ZUI4+r?xL5O!SS^OBD
zwx6+`Siu~JHH7qYBA#-hd>78FPR<n_J4bjB3A>*yepNYkLO{53qv!J7louR40vd-v
zU9GJE+ZFcA{MS;zTv(!YeLn<4rcu^-s{nv;XN$#wbsR5WN2_{5-Nnswhvkf#0%K3H
zw%bvk$(JG8IqyDmqiStH5~Yp(BN+reI?)v6Qqj+!7ND5@T?jYNKp|pwFRt4wn^p>z
ze1JL(qFCJuA07-f)+?DHQ<ljPj13q0{&_<KdCyVYa1avE;)^(kPd$N-v9xSvToT%s
zR0dL27|1HSg<_4`f#ZTip}9xgx(gGscl70cTK4^L*w1&#J=!E@i)0`A@)l7;6I-z$
z4(?KcYB_GFlhWm-Fsz~xAC^4$E@Wfd&}Op0{sha2xlqPFB(=gMLd!5;A#Kc;^{Gxp
zPvb5e>(kSisedU@$?Ip|Lu)}IK@a^&AW>H;wgL8ZPH#6G-qBNb&SH#2iN)hh+zl4e
zR%%jLV4a5NJr6L6*F5|55L*~^H^ay~p{rUyytJ-hA%FfPlWzd;`kW#C(H<%GcR+=(
z0NJ0>GAIz|iLAwYMDN`!F?wXqC#*j8zO|3>{g$pTV}5QqczxoPQ(WKY?s>OK!BH!z
zFDO!N&{!j1$R2X|H_F9;b&lA9dmtqU?ig0e!D5e_=8DlKUNcS9dC7d?FC<ghh4MTR
z37t+8je|_RDiIwAbsBPDz;Hdh)0|KN@eZ|OwelWn^hAlQg^@igu+aABIw~Sc(fR`8
z+g%sC)bo4}r{>=L7Ug?>=@RyWu#1bUSjR$8|7^L>uR{~1wrhlNJbS|}&F~tLwUetN
zKx?CwI!X!`)VrC+9_^X=PPqQ-M(=Nr*TyQ<eE`08IS<{xdFwcTymkA~#RJ-2ii4iq
zLW6e}K?3SH2Cvw3FZA~m8ooWsp7h!A#)ha$U6;(TEj8(@61UlVA6C0zuX~JQy;9qN
zghr)EHNuFkG6{%s)A2g!qOQETnfxJE?mLi&Aj8IurS|OuIu2R<b+CO~q7q0%72w%t
z!Zp0Z;V1BN8l-CGMVv?V*Cq3r#@B?}yyB_}jNHM}EyY1}PZ^tjuk((w&Fo{Y#6I>y
zwSb<OD~h<_2U5q3l6at8ZIn)L)EC^2EUjwx;rEn>6sX+$QeUW<Rrj&hcVWG&n<&7_
z_v7=A+@F@_PZ`h!ue0J6NydU>R3mCoGG`wJpG$k55ku3KZcLnjd-A^tDLt)$>-Igi
zMRGd{yTZMvom-e}L&b9iec%95|DBud<3+`eu%(ksOP+p>5>)DnFFtxwS74Ru-8?(~
zpi%Tnt!SI=1rcShNE=Ct29+gSKK!iGv@D6z+G;t(b9nfu_5ts8cCCRyM%q4VUv=A8
zwM23Ux#P^5LpH2h1;F&YvK2#iB+L)tKcK2|4$zyGh14*Pr){MJB-Ze}_Lxa{&=$?o
zlY0l()#ZGtsL(opR&d#lKF_j#>Qopb{Vhgn4e6&*CqYGdJcDy=yDrBeD9Q{`el#^}
zu$0o|BWFLE#~g)M2?nvelB!6AlKjrXG3n*JP0hz{J`MlyNpZn(o020pUj6LHQ&SiU
z{!iN6iF7ay^)!4=M>l-dND*+yZN+26ckyUSajU+NTFe{cl{ooHLbb5_fZ4fgLaX2a
z(gfYhi(M#77m!vgFNI;Xpl<(ce<-i0_yYT$NKA`RekPe`CDDM_F=p=3D5p1%<&_ss
z!(;VBdv?cfyv(pEy7<y0?_ThoftiuMw~owLd%6D&Igms+A;PJK_yT>ImO(EuzQ0$F
zzBmd`F@Y7w@^ax6>Ed!6ksK#lcw;+odx;TqY+brP=d_N+V@@t7YE{kGo_$%No;kUO
zMmN6BT3s~P0><V@)P8W6nKvZ<@Z=I6wt*c5<&=gVU5xXboxLJ4quo$jswYAtK((Y_
zVnqk1b_E_i@7zN6+P<wqchvh5g`ed-h4Y8wC`gU}L`;DP<s;a36dpujQbpF!*;&s%
zW%7C2pG|aTVfA;PmY(!MZyosQbobZQzOz3}D&vOcj;@-%`wZzT=TC1;{-H#`C;rI<
zP=BB>6^R*8gbgi1-#u!BUF+R69LLHgqudEnUqLUJ;O9K8LT!}0jjkaM6nXZj2rJ2D
z9A%>~&;9p{D3JxmU?cY&7`!W<Kd;CjXZr5>vz2vhNWhoK{-L3euj`+UDfLIvQGw#~
zdwVz%Ii%7Ld}9?GF*t#tbVjq`Ia+pHT02jPxjqv!NGoXG_XxLn?wEBzXLfC29pl0(
z^`d&lhq)K4el`Rd+4|=lqo#1MwQPJkTdGziw>P1&z+$7rS}Gltj{)&nByGaYdZ}ZE
zWvYjnck_`(w^jSSl8=0~c)8`LWQ#!v4GVvF8Hd4$P)+({tD2$|?3jr9eZw<0L}iiG
zq4l7T=Uo2F8QcHx@Q1DbaziTm;&eL=pl;wi-wA4l{Tf{rL+jf`ECp5WgoDW3i^yE5
zWB+T_18}ijGaw!9g94@woG6bkCA^{Oo5=(305F^JYtZx!`sEr^+Bq@{2uDWs|CZMJ
zFVf2Yir4uCrg-5qIntZSVz{-BR0LJB`L$38RRT89Vq$!vyxJmVz%_9YbPx&YUW(uA
zVc$N2;xh#9Vk$H#cNbK_`<?cJX`+AJO=%$<M=c6tY7fN#Pl~R5Os!c6QaV!}5GmMf
z8GydNZcZi#M!a-a$Rj3!DU5j0CI!y|%IZ!~jC}ZL4vUA1-TYD6Qm@Mj--kOc6o`0e
z1F(@^)#zCcyVS83N;Z2x0io@$9jhN`1O?pA%U-xL@RvrQ{Xn_U*ZcdzC{FU~=d3Dz
zIdI!X(uS^h^(zoxS>RM&Q03CZa~U9?8^zlc>W9NaxrzM_o)(u!?Td%@TmVDLL$W)z
z_esRA--B?)p?ug3bI!+kgKV)?A-rd&9wZ%8vK4S=IJH@T5=DGqRPJP>Dw$aefG&If
zBVwa|)doEbk<Mi%s~n1G^2EWNQ+E}=-v2dgE)N#XIQWWo!j|2=9yW<xjt*UcvsXSU
zKYMB|lqM+<9ikXdb{;*;8QKTLxC8%)gMJfg`Cq^Mp_lIeRIV*_+c`bUF^bszT|Y8`
z$;49x)NlYo8!Jc8UscO=y>ETH<v?I`+SAruJ9oZZJYxOLJdfCSn^sNO1jU#|;OI1i
zI=Tj)3xhTa!O|ZM>!mM<5<*8Cd&QBWTeR`SSk_y%N}(R3IoMLNJ*t?c(a@8ly`{g$
zWU;|Ob4$tVPg-7$78&oXrdDAg!N*rl1Hkj?L`8fB|2ir}N_cACmg6uaT+Uo~%EB()
zbyi~Rt)@t)%F2%w=QCGp*9EB7o-24uE@5M04n7<Bymd6V_R{$e<nf&^1F}vK_>KH7
zI`%9XcP~bBBDp=Dw1xpKSN*ls5189M4L;qt+xF_=&Ssa@=^HQE&wp?@Y3lyn%YX-Z
zaOJe}=(JDym}gX)Z?bp#LgwO(7Ve~R1zC2tUCY(H%(L?6-Tk`B9L~L*c>r$2R+PJX
zfPNNL&#^NX+u#+)Q@UuX7RT>8wH6D^%Okqpd@D6eD@`lew9G3dbakA`v?Ysl(@4kJ
zQZvZVBM@MrqEsjYYBdAFb~OhBXLShFNi+>#*CxlgjgN)Th=D+mErm}jS}P8*T!g1$
zmxMOYq0D%IGv0Cm198Fh&lw^55|?7j8QDSFJBoY8e6kitd0To(OTNc1e<8iB_y&?z
zQxJZwJ!EXx$>8w#Y&T=fvD`2Q9ZyVUr1!Qvp7QtcRW9AxQ+VyJU5fqaiLSRMJ<5!*
zE-G9U%sHhv6b4-eIdFGO0ceR8jztBo9>e|(j>C1i(G3F}Q&{t4@dNga>TUMT$@aKA
zc%73w3CnzZbxP?3e8=GSTD!9AzNtAk_V^gQCs&5=`grO5_#;Wwmz+Ou1_l8qJJY6T
zyJYESGsn|}1ye-?@86@itCqID+MlNG6|d=Wn@FBE^SFV*$N0#rGxp4|df|}Ruh7>@
z3XI|QhH@rbB8=u5K{=PB2={e@^f2mq92ivCDo!u&9>`M<3|tVOx46aTo6(LJUDXfY
zye)2+ry!>Oc_}h;ID$ne*O+sN6OLfpG8(VJ1QGkDNFnEBfvNzBHgiiA{x8<vJrJsW
z{U0AGDwm;LLX1l3##EFN7P}%zqZFY`H@PL1WXffUqLPHL(`J-PB_y|!OzxFLh+O9)
zxieFmR%T}H-!t~v=XXA*_W7LeIiLT^WvunQpZ9ZpJ+B9u<!z?+EWdF)*w|FpF?w|D
zs7r5za@_8QXU7w_cS2;}+#nr_zc1Fm&q3Ov_z_$@^5F6NL&nT9-;2fD;xnA*#SKEV
z-)y4!(7CPMrMl$R>Z;1zDs_3M570Ki4Xp4P%fN<`UP0Xk!QMCjPI!|ZXg=d-3PQwA
z<_MKg{4(52M|_r1K9B+5L0Rc=VIka8XK-Vm|6%dN+(MX0ZvAkRU<&@4DG_24(!h`_
zRTb{XT)1>gZbl_3vU@piTH$pLWd?TXHHUALwP5$7>xnKdA7*(ZZ~jfnZjj<k5M<(7
zt#I3*4U=9j-l=uoA8ae&pFdB4VqC>n2CWnW+o7vb&m*yk(I?3lw4~pctY+;t9Ggu$
z!Cx<0I_hn@hO_=x_S2ZfBQjUBCB9pi)e%AzpgMa36b?S^b4*S&o$m^d-?TAbLUi%(
z);t*&_||Z1SV)*Z@dA`EfgS~z6(@GKv`oZYEU^mIj7acY-IIUnU7g<|Wp(!*_hx(y
zZwxg9qyJM~{*sB$g!h=XN)&Dev3wc)+!4A2jmS^OY&i5ev<z0^^E<S_igwDy5&Z%^
zm+i@)%5LT76<O&me|xXjPWEcXU1-3?MvIfVDOb{AeXpnuJ%|D%r7a1H%H_Tyi+p3k
z#1?RyO}I=MSo#nVxYgkjIQxmi?-omb-V$n7fti9h!KOXNq4(d!LCP3yOb)dYkz~5M
zdh$aSLCtH_Uv4Sq3&r22gEo4wJioAh{=)}m14sM-B?J|epB%R`oP`x~xc&+w4;ely
z*nl1Esi`0-3-w#SGNO&QW!@feUEZ{)Nq=m-U%39bWLHNOdDSnEA6<U(QucV~c><o-
zKUrLlMKO~MSVzYq3&^f&V3r`Ps4W#o0FrmOzTqVlo^+M5YFLTtEha6dDr8Z!cH>D+
znXTtKz0~ifpV%DM*r;4<V|Y&Ug48y#wFO!KXy8mGf$nuy6FyUpO}Rvu720#iZAg!N
zESM9yAam|fM)qK9sd7m~z0JFPU&VpqD;f8?UbpSediXFa<7w(ID`Gdk0lYTQkB`kj
z%N)ScrJSLhH!-;5MJ=c=yJ2dzmCKLHV75~@wT8(p{94<>!s5c~ve%vVr5at@g!ngK
zX_jdCxs!v7^b^<j^RG(qnH5yAORHd|NOF)0GPpDYE^=vk73iMO_GZ_@PYIb$Ngh>o
zDX(WeuH35@Ala<-o1L`(|B~}TEK2-K77w<&aBhMa+9b*Wt=idyj1Tz&Nqd30CT!u>
z<B&A4m`&W6bK^Ja0xL`BH|vdO&9-dfc#ak5S&j&wdE4xE{1kWkxAPn4#@C!72%2!s
zNt*)~8%z$g+M`DZB<=v@o1mO<_|-W=9bXTSbmSJeD-aa-*F(vXy&w=^+&_4?-w8V?
zAAIqdD+LNd1u;>(JNX<{1@ZJfpE?yw;102k&Xp|bL8Y+B%2S2L%!rarw+G&J3_4rh
zKRl|{&VTE+vQ9?sWrZ|UARYiZ0MNR&aA!=&VRX#VX%f8rZ+>l}gOyQJA5PDAKj%`z
zm})kU=FOpow{9<Pbg48`X`bJl|L$H}N|8eZEsC5XId}8*mu8G|iJ@@`p*FlFiv;=Q
zIsql<py;YUC6qd?#GhwR;nq|jFx@R;o;&IJb@R#0+8m{Vs)FYoRh2vHq{p^X7OoqI
z*>(4y$dhJOgMO&t-xmEk9z45}q=iq%zOiA?m+ddX%QiqPcB|9#RC{W=H=+1Retw*9
zqSLLS{mX8BRsbI7PfH1SG|FljyF#B1;*7zTx}N3D?xDt%tmiVzRtRCz;EoN4qC~C9
z@btVl=)5W4VNJB3zLKrq{Dg}JiP~o`tHarC8N&jp!r3SuBg4wTm0PG#zF$v|U~qR3
zKX~^@N9PkGNwR?TKksVDtGPs$qQf<m?RV8FD0%PiU4EI<{kBal*=fy!H4Vl~OnCW{
zBd`AqbtkhXh*Vo~{BQpRI>P_A3=Fx5V5ZM-TLZo!<TAXT7?H&=Y?kQkmz=l)2F(I8
z7Ba3OZUphbOKeTdm%y`j{^vW$f4){zY~22`b0>+tfCqug0iChK2z18k@pP4D3EZPu
z;(uo=nT+TE#i9)HUH{)wGh&}3E>DMyV{K3XWOaG$F+4>dSk^9k{H?wBpQdJ{p-qsz
zCc{YBTr)`TA;B#YkHxJ5>L{;2j~Is^VST25w~Ps83_quVf8Hq8Dg8fEw|5jwCldBl
zh|vO%zp}<eB*?dm77|Vy)zp8VhpMB)@j>Ej580hY<VN$p31dQ5!z&b3q7$^|Y4Tep
z-OB$x0i^U_TFJ=@)heV{n~@8^L%IYy(7hC1=h)=yh=G-W|L_H;Jx%lo9WD{iAZ+{G
zAYI-(nY#_1263GNaV_-)S%_b(@g5Lel}x6p;%+62Kla_Q{zDU9TXO~ZDh&$~zxYjl
z^WW|S{^Qu%U)`i|Cf6;gGA6Qn8t@}}l=V<FfDhh}8D!60-eQMExi%S|*5|~OyXeGq
zkw>%I4_SS8>XB7`vKG3zC1}Mj;-2)aT5b)@f1vUxP~;&|Ow4pP$Gsv)GvqZR_bWY2
z=+7pr>}?A=!ON=ZEsFFPk$0!I%rH@QSGPT`e7<u`cH8$&r<Ini(4Gqu+kK9Ioz)Wm
zl1s)@Am!@p31!0Jf?(8dT1rAj4HBuU>KL%&wn__k6USQ$cHSi<uDum);U3_&OHGIO
zITS5Y6&k`qfQjV{S(IDE?fV<?U&MKg2rb$qb_}k9&z;=QMB+9uB*0xg7OsSxf>}YI
zq_k641+p6zDlD0G5pNes+$k|weDqGnkcvw2-S}Y7`<_J&Syc(5=t@V206U~JxtpWn
zara{G=aM6@1GUWd0bej~PUrg3#bBl1CNRfIuo5<HvJEVnK+6)Tib0Yr>QuaaNe@LG
zb1C;lIrJ&d<7iQPnogd2wtM`y%`F^!u=CA5+QB*dh+<DS8ioJ(RKiHHA8LOa%VY43
zM8D-D;b&4}s%*WtU|pGm>WT5(i9wR8*BX^ow4>cRB_(+cw;4~03Oy~N-4j<|ezH`<
zPy!D?xb<`>RrE&-p{iI;szilZFt8Ya%bn&@E{7<>A8-y)r!{ft=SwWOedQ{2S(s+n
zOQTm$Kx`UeC=-}jXF##B@htE;AhT>v-&23LFx}rw9^Rh+oA~~3r(+(}P{=o&LZG0^
z_62~1lbzI+Sl%<%C&qoJ{&rFo5;;+wb#JTQuwk81Rk8Aj=;KF&={`O_V;@sbZ^=}M
zO)pfsDlbPkGZ{)x@+L{(5O#2#(eGP|Mw%weY4i!33_Iyw5#h|5v)phF<I?y`f4XeY
z8eaW{lTi^T!*{+5(t9}QFx$|5)^f*Fr|e%0p9?KJw2(4Y{+dJqc(Lz5RCR$_IVPMZ
zw1RspF6eE^!3^kSFqY>yEXSi$c##pS$7HDWytjcDj+9v3W_qgbC=(@Ue@PFUn65*6
zeEhjxHmx46PJo@XCWmn$M3M`!D%+4pk1)~YLmL*_Lg%7fnognx-_aqnI=<iry8f^k
zN<P(%rOB`V$>BE9sGOt<s)lQA25o&!H<7pP=2XNZxwZnI-&o*%FYT?#fOc;2lh}jx
z8g0A$I$?V4b10HR#f(w$dP*j&Q;K9voGgayYvrFcvT2%D&>a*!#-T3t*%Qel0|xsX
zb0Xa1q{Mi_zi3FIS*k*H$edv?*#+c%ki*1N3||xQgnMGovW}=MH0P4f?szBk?18qb
zE%|!mE8(WBfp+7r{Db<1)i|ef>&tb=HCAqKhJ!p#|H+*r_6Fkeojh`+c*M4BeTgDF
z=A8Pc1m1>9u%j9!>+Vs~FsX@-s^pS2D}ULT6gMz2x2V<HM{Vut_Ic{x_5)-c9Pf|E
zSkqc6FLdITUN<Y?P$E*-HyB3kWq7_{9W!&t+5dd$m-An;M~82Qnbw@|?$N#LAz9Zi
zIZnc7Y`X6`fk5KSU;Tcy_mHiTZ7w#zmWk>R3|+`NGP3WAr%m1Q;|9y^D$~bIE<aeG
z8e&J}Lmc}P`*IqsA;wUr`V+&Ls>91e^n=!xl$K)D9>Mzc>wVQ|A>KDKGwCx18#+2%
zDzwrXZA`uxZ&KQxX%c7r;Qj$g0(wdHN{B?zi6^|A%evt#z*js?qgR<}C#&7!QZBfz
zI}q>ijP^0Qb=%VZ{<7F>32v}CR4R0xFXmnkQPH?1u8XpX90XPkf|2`CJP4a^1|^Mm
z8e-m~bI1kQ1L0y$8afV1rxKRJ?PJ#nXSwiKa;ahp9jh`kazp8yO`cgmbZ@5Jn6HOg
z7voI~G|$U;TSkCSu=`p|hn0IQG)DHFx-5J_%3=QOoIihnAgRE_ksYIkmbsX0dWhLn
z8XH~AkH9pzAzl8aGCT6$FIT4-df!P_^mjAlsvS5m^&;bh)*RUe=`ZspEockP@kPiQ
z-kTyj8oAs|hj)m-@^;Z9EWomP^+mn;v4uKGGxqx)Ye;9j&e683d!fTK*US(a^g(@f
z0Xmz(jpUIg7Z6T>mk!?sxezs+%C+}qs0jBotfohpsUlNLwLN<(T}_&f3YXl1MY_8a
zyWe;@cRH*&{OQE>OT_hWu8Q76H*x&=Pmerg9%yhCLk(66-N0u^3tT*)OP0bi4cJ}6
zElf^Kc8%v`&9-Aa-T%OROPQy4@bb@(2yu#YdQxkSNq?W!Al<>_+IkB#$^7r4*sq#f
zKDU>b*Z0uR9@QW7)xFZ*(lY()?qut1-Rp^qZ<(*EU$FGvNBU?eT2S+U?ecH~I(K_c
zI91hTyWkZ59-qf$R*<#4C*<>Z&uY{o;<G$(rIxRnWc%DF?+?+)%gH0l4;MZwye{9L
z5#|`(-0Kv%RD$RJqu{}|Mg)oLKpTBI0k{Ib035LoWRpK}Lt+VsZv$naha=prH|l2y
z_wb}*xHVOByk{J#G=L9%O7~>jdVJW^S?Mz-bv>)@Ak8O(@N33rlc$95PPx!9Pes``
z#q;>@+^~aCkYYq4lf>HpW_W<AGPq}Sp*`zrEJ{ER4k6R9p>o&d&mLpFJuXqTzVbs8
zuDPZUXVI3vHB9C#lX==bC*|{{&x;-QnXiL-$lL76z?E-<%L!WzedRzgT)`9gcvq@t
z({Fk3Ny^&E-*&z~%6jxDKT5Dlcsj`_VAZCnOX8#kbhgJyorbpTBi$d*bxIw*h92e|
z-Ss>gn)Q*YMMZuKpJ(04xm<#nR#kp}w3{6^=c`nl)1d>(iptl24NcemMEIKgXQVDp
zCH|e+u4lJt8I|2y*69r<R@wc>J6DzP2GmP$-n~7V?p0Fjc6{c`*Nmlk5!%OybL`~k
zm=mhV=ue|{z{o^aAdKDIlU(4)qh2zxP_pyx)@l9vWZ$W{>Hfh7w>RI4u+(tgoc{dD
z^@FFj7hIo`D>;=#criywLIm`<Y2cZXpu7WVhBQ_y?#MWU!Z<u22zP@nL^68iNkzJ%
zxQmSOA@KNEtKHN`6w~~7dt7Ywo>%HWt}vNpV!vtr+}p7>8q2B&EqA?3xb|7z>a_%z
zo$^{SsBM#EH51$A%@n3UZDKj&0(0EX#0NCP0d(zEGXec~eo6vwp>wToWZqz6i|U8>
z2L$?=cD=88L7#48TG!mTXAfn;hBX6=k{vvC{JOfX$K>pNLAnW<;9tgsPjA!cm8Duv
zN#;EHr#G|89G>i5;J|2dIJNz}fy@&2tc1Bsf-B{vDc2;uaTRK%@5ExQHSj)}piCbJ
z33w%gRp&j#B;oDU{M3}5+S(iqPd;7JFzSX4NtI%BarwwmjmoC<JuV&I<kM3Pu4!z|
zo&VCYS1)uLkaX1xmmv}0D<~xp>z%tB4lEwPNB<F2wK<j45Tp#gAOcIt#*8*Lnk>xT
zz@xCYDh!#aH60n(bXe@({8iC#<)HD-HAS9EFC><GudtpHT9zxWxh>84iAsEp!KXX+
zjA?p_NghAw*KFryM3INHL)LpL?rH1BVs;l7Jh|N8{%&Qxuj$9wE$2dEDRhQpxKNuk
z%=ojH3J-k=5vtJf9a(~tpeAQ{x_BJ$aDGR@b=gAZ^^`^`J5w)I9AmlRxoo1$U{ND2
zhcok4;C&jprSw^><IaMaTAaZar{5L}Zqg-94q~CmO$O$Gg3|TVX<(0L*h)_lpx<gd
zsYf(FJ91J_W6L|A+vpmbVBKwbP5Vx-j9JOfi<r~KtL9I8bi9h-1?X3Bq|%RQH!4H4
zkX#C1`oikx#1Y9F)B~@1?>o*!j-L9?JdjE<^5u?oUFkBjBf?14PtKBovg&+QX*pEr
zG=v6u6to6WIGSGV$l2Monx$D;Ue1?i=<=V%teju+G<kK<@e=*v<FP_uj{%U@zhsa3
zFUrtJ3lOM5Iu3OOJxJekd64*SOLilmTmDFQ%dLc3WgLnkEwGG~YvZAg7u+nUxm0qh
zNG^RasogWhar45B)vsx?Hz`Z!7817}2u13fe)j4)kQg`U+1vAA(kigV*y_q^jWM->
z?Lt|e-kUS-mqtfTHf7Wti<gVPlbExx_wgG>#nNicz!jWOq{#qRrdOa7y4_IzDz}GX
zka&xkJQ$4YCUUG45<dN#z=k<sM&uSI1<&B-Bm`#AQ0vFfqE0>7a5_MxC3frextiGq
z%;!+C4<=B&{rwLfm+MM+!w+y6)d~C5;;A>P9ec)F_T4ILjVo_2KSZ=&XrezF4~ao^
z3O-+G{uJxrp6c$Yab5YO=n0m=<*Tu7RJM0Ms*^3%Yn-+F&c(4y)=C%N#96Y^)0OvG
z%nwC6*S~-ssiqApM(UMrXppK-9L=$!UMGg_BiC&eJijibE4ean{mSLyRlhoMe&(8v
zKe@GFyYetNn~j1p61K$ynI!iA$>kQ?p<~ifCiZMVZ`6<E-lVx#mTGfHBQt{RQ{@oL
zhJq9;>&A!s)!oi2v<<We*FC@U&^#{a$^MP6bfp2f2+6}wu6FYK)zQ>F#YXKdnfHcN
z@60kd{eZe>P}Op6^(=$q0m~m>vnXE@{9rA?PRf50;Py=wt||rd13WRt7I@X<^gFM4
zs>vwNnex2kOm&jm{oC^0`n%lE`|iJf;1u=R<7qm_wyoVq90$X9-kr<DFMbO{xYruc
z@8D+>>Dc;ot40$UFjmhVw*)EP8R(X4A=XeL6Ki&Hv!aV43l{k5y(xF`3%<R{^W^@k
z$8{{JpvxC+3FxnmKRHN9g=<{by!To7=)K87PS%#{<As56zZC|*CB`paH^n6|>goQg
zlHrlR7{IV$ug4Vcz#CEPIK}y?-Y&mtx`lDN9V2Lm>=WXa9B_*>vVUE?>XAvfs6oEs
zCu)zd-Gr7<X*t{DD_mMIhHnGR?~iuj$Dbndv%y4oW3$WDYlh_nI$p^|;X%tDcm@1U
zkTjX$7kqEfOskRXo1^d6c)(L%KYP?i=gE`Q4Yjq)o;~fo2pyMa;Hm4`4AJbdsVIAi
z*zd|g6!uVDM15z*wZdm56aU5o|HF@og{E1bIoLf79^sZ#f9dr1t4Y!xeJtOK+W6a;
z1HZ4L-SbIei@jdiqsE2#1z$dzyjl|V#;<lIL^OH`_c6WzeP_-q!j+xB0aK10BquUt
zj*&=-sDZL~<)EMa;Yt4ahomfXI1D9P*y6O-66>6yV=qQ@6b2`%)9-JRy}Twr)++G$
zs_Nl%?^V+dNX>X55xN56oGGDt9|lz?_I4UgxQ%qD1Rw1f8C!nzaQ?Hiyl8HxONrmc
z4{JNGWJs&c9VI~D?T3>fW+VI13v>m2m>6ku?e3er(KmXEDm?Cnp|ibDR4n)9coTTF
zN=a2ZmeR6UW<Fwt$wP^_d@?VFYVl;J?0?J{3-b{cH^pEFNI&d;TdijXb1zg*XOh^j
zzbEqJ-`qHV`>96H#nuH+fY2CSf_`xB6G;?Mg{pfe(a`t(+yW2n^3~JNQY%}mzb|nL
z>)4iy+l>e`9m%mPi>G7}oC@<wN)Fqsx4BVh@wBCVhWnu%$7Ggo?zJ#E6l&09(lD)l
zSkDg+ZO_4FKz#soj@ltykAkcgKuJss;P!TB`W(O89&?sDr^m?n+$yJz$Y5xMsAa1x
zuE=HYEOaS?$*sW~qGvZ@H-x*m*Atxqj5uHu01r>u-ZTa#Jcd*#x^w<(%19@3=AGR>
zKKeJGKDo{)09T#3cz>d`&JkJm_Ly8;fyk@J%Uh9tIcVaris3ge+euP4wH$OfZJ1)K
zXmL2;QSq)_bK~ALgc3BZ*J_PRTsNO7p+lIV!;)yGTeQZH3;QLvdYow@)TR%fQ<%Ph
z@=QDYrS^sC+E*;3VUIVNuksXFE`1!~#adB7$Rp6O!)K5m44k5KZP5><?L#{r6+Ve%
zQO?cs&?oD$lXVs9uEi}`=%iTV@3v!o%xGu`l-G{-qvL9T#yIE+CIK1_=MmPYS1kDC
zuqwRC&!tYea%<4C!rh0g4<7NPmiIeRSd;;CQs}tEDW+|+ufi+OjP5Q=`9?lJ{_WZ4
zGFBS*%-W=f%vF#_1@#k>UJOhh6hRE=-|tNymwJzK_rm=JH2XhGW|A{nYWG0<dHTpx
z!Pe)!=X)O*z3|c8HR|Jj)``K^eX@SnooI=*eS~<XIQrTGVIi>W4ClbUq$;vt84311
z4crIDFyK8Iq6ih@cEg1admi%QbDhqyDpF1Kyi}*X2cuO58qU34uOGymFc*Su<hAlg
zK?W8n(EO8HLl=gG&Moe~G<+|)^w&3)OCK=X^9*K(DsTK4M<GbWopCpm$bcZt4b~6d
zU)o-p`0#e-GB1_TOv$k624T||OjRuFxkR9hbP@4z=n2aDAeab3xq6|>mEE7jw*l{L
zte9`3MAr)1{7PrrW6MFGlVjKST02O;-)6cJhSbXe0L1{G?(}u9ek(qGfV)&9WlMSY
zE}nbk`~#_<yVVHBcd}<#FLY2FiT!>OikR_&fDOMB(%vS(FKJM4#9Y-UdxQ*U;_rL%
z(y6=-+qwN&{m#0_h`HJ#k16oJ1R!9qp?-)N=z-#v1b|Fu!f&q99|7n;D~+yx`&n^O
zL3F9)p2dOJ)6dGF-i+7kj^^9ey;#RL-6<(rgBG1cCS;*2x-uUQ721R&a6y{Xfqf19
z(n)McuJF$O<I8Ay>;XlK)4qo1<aP|{=$y!#8}LFn0dO>JJIS6))+q-fj1nE!?^ydT
zr7Om5)3jd&67)L+&wo-735|^{*HnBK?A&GNbUk12NZI_A&aVVG1#6aH6~V3HKZP71
zEfieN<h@}n6<d%~-hZ@UnT)4Z6&kg%%Cxs+6~W)uXzJ~_bVA@Phyw;T{2*Vw96hpf
z>f6Eg0fiCw!qg@2cYlfSxL*F7Uzp!Ocj$76lA*RFUG$;58(Y9%sB$X6(aprMr<?<`
zHLoRaA9Hza)sfmJX__Z)7u1N(XQK@oyFiBiOz|uXap)i}+gC*249aqiHrnS9=A++L
zIu*OF5^QPo_}tSQ=Vo<2T$yB=bc(VMJtrwLW&G25V5;!P4XL{asa&g=_5)4ZN(vay
zu40b01&^OZ405)5l?+GBSQyJ)x84qd%9lw3835s@t;vMHH^hL+6WMsnmid-(dzbCF
zWT=s9z5M8QCHdjHnc~}$CfiQ<iiTE_YslftD+y-HgzI(ULkYZ)E7m$Jg0xQjcQ8O*
z=nLad7ffVWY5viR71?C@Lz=%d#bEJmXwz{a0|*oUS%L0eiCbE7B>RfwvUDj2`-V(?
zi>W8IS%L;h=<0c+FsSu?D>?!=Ah!dkX!XFj1op@Ns1qHOX7Mppe;%<LIaOGY>~Fb1
z+hCK%c^$tni<+%p%vI3;X?`R`FG0%jTiT6kNF95>p7&X}2(xGHI{#Jsh=T)F_5QNi
zd#7)_oF1o;D;c-`p7ey)4|Y+Er3^1GFkJ&fu}%k-@7_G<PZpBtk+!`ZU1f$hHZ`~T
z#J(Om+u0+@5a;uH>VH-v>}F^ataS<U*1BnuPZxz6R!3%Co&N6YFVm+gW^`S(=(kW%
zu+W^2q<TpF^E2aJP!sBM*-{9{gND3Z^UXMw?rVZfF_|{^QC*;q(mUe;MJem~rxV^a
zb9aUA?+!KV#k&B`(Qg0{sA}jN#572qGpIPOUgv9QUnzX}=w)0rbF|oCS%~zPA#*6i
zecz9kagfgmfv^maDUXcDITP^hi>GuI($O<&4~Mfy%N6E!xY(}!<+Q}*&1~zrq2^}S
zKZWZ3(@mXH@Rokh>A~~f=-yjT+oS!w4QCEZZFWdBlr;xqedP%!bjjZ_Os$MZ(!$iS
z;X7jVD?0<fJrPVmpZ??W4C#N|Xee23wm`gqAdgWsLZfAHBFpHJAtpTa_6-MC=P1$K
zx4kcz=YDkMv`59oVNOB6S}E3CeLz%_cma?WsgP=5t3{;{Zp1tE`>aqT6gD4bup-Mz
zVL$?RFpzYPA#XB|J6i7y#cJ>8&K>2>da2R!>3-6e+UK}Gv)O!Ww)NZ%wU)=K9M8sZ
z2>3F<2!mFH8{slSU93s4BcxTGQbmtUosC7_^#Jo1q=DY~faO`{xA^1>UX{4}Pc2%L
zySHj-h2hn+8ZSdtEb_;dB=B<#ZW%o{QlvpDixJMtr{XYn+82+@`(fL;t$fh2ume!2
zWgMnTV%{ujWe0khE@QaZ#GhMhQ^?J@$a(iQ+F?-B-ZsXeX@1zikulDuD}xf+*0JVC
ze}x|FC8#`dMR{m{pP2zfdwyA{m`oe!zZ6i9`WRBav`~2ge*?CGSHxk)m@N^Kf$99$
zlR2t1nQDAKqqT(0F<qTM)MlHQbDL>vnl+LeXXvN&?(yg*Q;5~>X9!og5AO#UXAKuB
zjnjR=*u*ZD>?L<(@E3?;4T&FG`Sg}?YFjN^<$C4-b*kBXre|?<QrI|6e>g|4#0XvZ
zye|Khm->>MvAcR=xAnSBJtX$zD#?VwAUP#_tfP#(HEOv5Eu(^GqY?c|XNOnI#rejm
zGmCQ_r0<EBKaa$75x%PE3P5Xw+NIA$$))mLuC79T3=IhIGYT{GT=eAW6K-m};g{DV
zUQWG<j-EOvPQ0!B?Ng|!6+Xv+4|v}6gV;)_0<DRHD$q3vd<&JkW2~LQPr$Ydf<oQ{
zM9WNFcpOV&p%OHs9`dD=<e4<93fHC0<5|)b6Q;M{$?P@S(v+3GpeQddbk_VZot=BQ
z=};+;MA#+B5g1pj2P7i4XNU?*AxI_|vnJI5OW>>@Q^laOG0}S2Wc&ToXuJBi3*?OE
z0-O3Jr)v07&m!ZT){tL6yK(kK%@)GPdy*nul=W_`Er}n5Re)j*^J*^OOM>QN!962q
zm_6u)VeR)qd-n77j*Lrh*IHa@(?6P)xamA`>2<=X5p&Xeqn{lU$T>n*bQ#E+aaoit
zcMFoCg^+^$po?2DcY%>`8fNf->P?+#rijU~`C0d!^{(8!huK#Ys+)H8Ca+7-%U+Zp
zp&O@sHEDWfXy=u?p{CiPNHXN8VvZ}G4bbM6Q$)sICna#3e2E5{V^xR_7rAhfu7V+Z
znoCx5e56Kv^L+xWiW6(>o5+ezDxY4DsyOYtzP)SXg|g5C>)tMw96&H%Vuula#jgj!
zRgGK4JoC~G^v8}cj6wA#73dE4-SML5M^?*-jCNpM+|p{idZ@a%vZXaKz}4{G&8LM=
zE@oePCCgiHrE_FWvB8oTi$akO!by^ax!Reb{4-jQ`>XaEW^H<Rjg_hQTh(r@N`>iW
zj4dFi*+@H?-y?dI&ul056sGJKSPS>(J$i(_;+egE-N>wNuRP}5?u>Qp32Sek-qOGP
zSedM+l1KW?OG^7@5_pWCJa(PdLY~nD6So|%)Wtl9+<Ua}O6RT)rq@JErr*L(>GxeX
zD_Q%U_Fv+ud%ZlgaowvI1261k0Xs>2K`%f@mT@a85=#}Z)4Uq!<d~2T>NK|X16_hc
z#$C@-jeAyjc5B#pU*K3KgnCZjldHCF$Xqj0XcRee0|ubi@#v6wrg%^8i4kc+l-Cdw
z#h8t+3@IngDfpdmDb<8$74tQd9Q|c?yI1Pc_t^ze9f2$FZ05A9jed-6RHz6@%qlWr
zDSjcKPR$ZR3AAiX6Fq=r!~pyC)HLEI%!e1iLCd;ahp3oiPrVXN()oG5K(6~fd$04*
z=I2lH^h!ExOmD|NE8ji(!R6@bGDwS#{>(}p$JdJ>4)LU3#OD@sBP)?B$Sk2dJ8)oL
zB5X`ny;NXIv3VNT6=nX#c~7VH?EI}ZmRowQU)DGizC%~o8ezqhssWJtGJ};t%B7wc
zI;ML<Lud3Uk5SQLTpIPD!|HT~m%|Ac>U9X=4m(_;?YXC?ao;}Q9kxKg=n&l6ef7Y3
z5*@>;!4=xczZ2%$Nk@mgqd%%2K7BW&9@YG+hwzVJrtWBlP#vQOx&`>MQq*Cu$%u(!
zHpS<-Y?PP!y0m1&L?h{?gAZ4V4clxVe-%?-r)Z_6i!$%sT6%fugRa-g=5q-U#Wu_&
z)c%L9>OkPKN4#>{t=QB^ZEj=*8qJ&)G@Bjh5*3vhEL{<@q<Q?gzQ&D0IX{zAiLHv2
zkGH+fS}?6vVa8hnahvd`iZL4$qaL|TRl--{)y(`;1w5A;d?^;530p(+qcoB2JlC9n
zhkM)F7!jijy}eIoe~j%pe_P{?*RkD)`a;u4tpx;_<%88msI=MYO?S$8-t3I<()k;1
zy;@$pHAJsr`0KT%+bO5E^20`#Y}8vS%(l~Km(^>Sn?(N9EDP(<vL-6ggeV~4FK}5b
z2DThJZop>rbie_Zk{WVM_Hql$i8H)gjZNR^bb3O6c8ALz)8A^YYr8M*YVI?Xz|YXR
z-K0Wa(cHdjRH*NtBi=vMPQi@EdTOdsp!=8^FF=CkVIto8YWf^ARoXS|GA;atE7Iy4
zWW^3XI9%*|!&h|k_Ta2e=g#k|f_UF!g;_5bq`V*vv_YR)O)v*I?uB!p7OMfpjgMOh
z7!jXI7pUN`N<o%kr$Ros>Tp|!zi@ACXdU+rGh1HT=pSrQR&1Mv(@!g{`0ltSNzp~;
zYMs`$SBeQq_r~`Tc~8=0RFWLm7&)}I_|<x!c0VrnKI@JC!Q}2w-J6nUaQrKDa<@e3
z9F^bItO#rd&0M?-E!?QMX)2OmQ@#6UgyE&d@ipo%H*=zYu~L^fN!Yn-W|}+!s8eSA
z%r%Xy6wDl)NPR9)4|${BQV{Haf3Q8Lj2P+Io|qoEj1mxd+K>HA?Wq3I!zWyw?#5pK
zcxv6@1KWTh#sP*wS|F6JV2F7nI^yxyaO@M_P5qL*!+&7dW<Y%(nij1rN;SRBkN@<2
z$a3K#U~Mq9vBJ7owVF!>RjFE!T8<uB-p6(8Y;tj~4N%IGUA5(w)y8(U7dmq#%}F1=
z{mg|)f)9lNtZ+QVh_8oj5G?ejK6Xne8s=qUO&qJs`3#joWX1uiR5vYZTQ0ZkS>EIK
zcN;h5b-TFi!SOkF4)09rBPhra#7&SNbVFW-Lmn{|Y=92A)p=Z3co@6bQ!k&@X;iU^
zXLYea(`B1#V&95AQ7-pC#=N}{w_~fDU$tKXf&2HK<CvLpM_B+5VOM3nu{M2PA7e7l
z_xRKYTI<qoR2rW=Q8iynO(*)gCPnt4#h{0z=y9nsUQW+_-Uoz7-vX$4R%K5|kCseh
z38F6caxE`^wl%@Enr<gL&1rW_wN^{{@OgLKnFeXN`ae-UFjZi^786JO0IIu21Xdj?
zIRs3yr>2T1jV&vus+qXoq#Z1ukd9hyyM9wcexTg8%@4=BR#kU%*6J-Exg{|#OlM;>
z&2AbY2!+NZTW}Z(q;G@VLA(2$LA#4V>`n~KD5DxDV-Pq^jTc=DqV_Omd28wMEOXe=
zmU}6#tvQFA{cQ$j+ovdx9vxNdO$ZB3%UeUJc^n$@9#a3GW*U+L9vtXuhCAad3~Q9i
z;B7_>l~QFIiL<a9M&`>@if!woPReK=_AzcpXP(Zp@LbrrBY$Nrc}#P*2l<tFn+P8=
ziJ!;>UyHIn6U9nc<_k(_`95TJ5GY|qoy~5nt7pBYgP=S_>(RSx-(eQl?RG=QUK>v(
zTf>#Ft2bJ8D!<Z}jfeES;Lm;>$rXP=0@|&|NdT0Lpijg@B8vB&{M!WMBp~u1iAyF+
zqW{m#Jkp=pkh=!~HcLAQOp03{s})c6T@0HA?n&}*;Bp*jPu1WRDBnRFCPQ3rnv;g_
zn3Le&us`|ZOVn#jpv{;COzf)7@H;C#B(q@mc#N34A-AZ4K0$j7ypSg_>E8)DWLA+V
z&KGE5BKyZBhQL%VK*g*~->m~i!2h}pRU;E-O9{dus4dTggWpbsU05&$(Zhr(h((|n
zQ$msKOMfP}e^r7Lqu7EM2<D0zHs3IT;2W&s;BgrK7v#x@_b-;ztt5Sx06Q|Qcpn{@
z+bn@A!X5|U4j++(M|j350`Bmy1lr#LacBlvs0O+TG^JpE-9Dj{oSPp~ekf>(WiQ4s
zf0e|i=K^}wL3CaW*7g0JU=5@?FfV$Ez#{@>s_ip=ICT2v@u<HO>`X&Kk#Zw=A}#R~
zk>QCBHA}=9$J_&%xrw6T1~j6FsUY7OjIUA+{gMMq;gkL&1|hLnP`It!(-dL4)e9=$
zwFL6CTQEnH6#Q2VLehPXLgflL0f85oRViL1ela2xsXQTmPXIc6GZt3|ptFA!c!Yq&
zLmeF?XEaE+f=pL4xW~rC2+fb;rY3?Xk53SUPJcVs&GpzSD>2%<7QV@2p(2YhC<h)%
zBeK2!*l(#JCOK+z<n`gVlq7KfYyY6`TIabj_?Dsyl3=l`41k*BVK3!3KMd4Xht7@^
z#n(?m2aWF9N#6u#K|{-I0so@SC?vw|fed<}+nInlhaw-6a5=D5*Vcx-B=MJep%ea3
zfWN>PM7x}fnONaFa=-u`(qPx%aW=q<xsdRe>l%q13N8*R<F|^UVWOAT17{j2{vadc
zKlSwAVoH7gkKlcXC0Tt8Y$-DGUZ9PZZSo0G33<!Rmj;w7O%9jGH1_%uJ>o4BeI5+1
zX*BI+%RUTx^<hel>67~P=E!e37#br~j>luf*tDsT#Z?20{{a}g0FxzYXxWjFxp)I^
z1SfaHlNa({YcaN@$5)Q#$cBOPcYU2SNAuQ`<dI;<ZI3*{Mk2o0f%;xmto=N_Ojbff
zLMy@%Ry-;maFb_nwY&(xQ3n?|?085uOc7oX+H&gKZm?n^%v8Mbm5q8Q$%=!<vPa{+
z&g@EPa_w+>yPH*`+?4-d<+qq1VH)We>_hoKwLDXxms@qdE9>j+Xu0<ZZ&Xh;k-IN(
z+IJ<ZsfupLDo4U8w{($knIdRg>})uhLf+8G@XooAc*Zq4!^zXw=XQYN^pPWmLmm%@
zM|w2mXUWhGTs$p!VB8=TBSdQOnNSV1nnbuwfuJ3hLwtAPX>drF&6Z<BRmiLmZEr`{
z1-zW-)qyDimBk|iQbRl6`%X79ijNIFbnV`9bK#RT`K8cMr{L3PfGfKQf;$t^MCIBr
zL#cC0*Wh*CGCXFb?`#rizYlg}SA=CLL$D~l*P?HQ4oM%mcCsfejo4O?9did-2l3n?
zyA_v&%{IXyZ_vjSSP3;3Q~AQq8SdrwSpr2!rpPt;%-Du^ig7JI4j&H$R?Awi+Pv0;
zSRHB>1BB)=xI6?)VQ2AHBo~PSeIFvV#rGIp3{i#zvm#_gG0)I>>V5UWA{F%;8@v}*
z_a4b#Da^dm<yYN6?8(GQpz6~OK~<vyi-G7H>KnXbn61STt^;MlL=1PtW)sB}y^=rw
zQPDB@<2oU2%zU~~*^kfXx(}Zg1PQmdGCX41nySoZz!*#cGa_VV?4oz%uK*iq17I||
zJz>Ev-}T#4IiB;hxqG-UDBV#bTx)>O0EM<hxkJAbBtE4P>tbM&x=;}if<;_iF_EB~
z&=qqOtiUwNsS3hr+`t{JLAo4UGc|{AR__nIKY8I`N#Vg42?tYmO^sm3%uR;-yX*|1
zlUsk`D_xov^9#>Cl6<FWH3TSL6Q`cv@>((1cK?F`qvO=wdn->@CVsQqv~+IpM_{Y@
zn)CCQz2b|Jfr&MYqqgq`{{?oLN11!|b;PH+CMV`u*$YZ9KZI%fSDFDX%|e@OZ?gq0
zqaPV9K^6}Njk|M#$*a1(&SZUTZT)s4_4A&+du+gL=YI-a^=BJYa*Ws(j?X+O@WHE7
z<=%&o@H(cJi4(V?+_)v%e}rP$aCc*2@q1^oPS+E+gP-QS+S8Yj1JmG;XA&U>3VV#c
zV~c^=FMdUVwLR#xU{?r}RFEWIT{%5RXvnN4%?jEDbt~NwIu~@#_hlUF7R<oSO6v<8
zwE}o3g+stbCp{k@dVq2qQK}g+q!~z?$l*-s5MylFR_rXdZ!fnmI=g@>JJ|19J=#NQ
zN=tlet$ZxNVu6bNhu^+50_$o8K35HH1(Xh849*`i%`bM+!CYWdD}b*yVJmqu<y2)e
zb-cE88kZ4cB9~c~LvyKAQP@H@OmS#CwA)wj)r(n{kJqZ<-$Gzt?VSec3vjtU`!MrK
zqF5r7CnwIrbQ!K#stxcXKA3?0!8X4WY!;=-#x@5-jqw?I7<~1jc)=?0J<~Ao9QW2l
znv<*n`rcPnk1ZezXZ!XPRnq4K?Zxhkb*YM>APmfWl?|)Wlh3Ux7|nf9t7P};^obW&
z4|%XI&d|K-|6;jBAHrqTV>(RnAYVm}P&yfQbk*YW=As)RGGg|<o#YFsbch5^V+$6*
z?c0L+aa(uga^GIQoM=ZKZ1*28KL0v%R#5Hr1(sKNap{u4gDns9x_`-Egbje9&tn`N
zYtQ8Ca+8K{1PfLL?d7SKXXa?oVyf4qF1leJyTHexyJG8?tb<QFHI~WJSMgd~F>~Y+
z<tL8gn=U?GL$DzT>>c!ETU{55ZXm35KyM(P^ssi-3fqG<XB^L#92+uAaOY&LvQ0ED
z@m9~2xuAM(T}?J}K~g}1u|+Ag45BxfTGT{YCe0GVR0&vTk8mlJwi}$I;k+nT(lC?7
zm=a_OY?w%{iS3;Y^Ibm#KR}kEi;GAlF}-HZ5)%}2QVQVOh%f^sjul~pGihRxIIL29
zO&CYRlXpct=LN7`BOi1u{FL-iXvH;4xbJL@XAf3v=b0tBpZ3ue%<iaLZj0B7e`y~@
z3ZmC(;AgJ^N|t-)D|0xKJL;Xul<N-&SXEH+?9P*?jU$JXH?$_jw#ArI&JYF=b{W7D
z8DiFeNGe7IGKt@)BkREiCV*&SgQkdXnXMMi#%u)?Y^aPWTdE9u0RqmNNVE5@p(R#S
zbSuN2W&gV##ha3^E%X^a;=>8krc4jdFqB}U;O|M;GF_3DctB)Eq?q(u`V*x(25#Fd
zh6m+h>T;}t7g(i=8)DHt!5arl7xUKKEz2qJ{-VC|=+QX@nO}Q0u}->vH<al812pD~
zzlK0>dwmC)zl1w>QK&002=e7o%AZSC8$ZAMjAOdq{Q81O`}aq()T>)YkG*b-Tjt<#
zE&Lm09jt75Px{&Oj0Z!PdKAlqpf&6RkmlYLf^na=ra43Oz=X`TFEdl;Q7iP8)!e4|
zviG8UTP<TB4601gzH_*wslM&D&c2MT`$h@)1I7<hTu0^lwF+l3`O4yHRaS}vT^Vur
z4b@VFUBj=;R?;*qd=>g1y-bSS(tLEGONZ$j^Ur;|B(QZgb3OsqO(hg@MOfzmmJ^~H
z@(x~lC%G-2^{v`>#sIz+^MlMqk?Bl}tT4WnnA@BdLrLCwWnanF$BXxmLNuJg_`!8Z
z{%C|GAl8wEw+4d%><|$0zBqkPGLcs=R4Bq{xUtDov5Kk?k|+)*dA5Ix0LVL0p+EWD
z3Y)&&UK9H90Wi;$ds5V!HIhHOroP((EL~kj4T-y#ge~<wSK{46oQa|QLhStAels;5
z<$?)v%d?@g>f;Ab^)*=m8J>mP51V{<D0FDB&_d5nBb;FV>~MO?F(!T-`o+{$JaQ#b
zWAK~$x2O_oX190u{Fb9Hy;E#<F|O4G1gx7QWqo_$uRS?;2q$(*0J4?qlOyK-G#As<
z*h6ZURRv6)4bpvMBj6mS7KzTIM~D|}c=r5@m;{@2(bdFue$6jbg`w@6zlCpUx^^ca
zwYR7Cto8Jj%-?v>bNC@T#CM@A4q4e7&Sm2mL<D&?XW@vi8()LCpK%^vjp<gFF2ioz
zp-sP_C)@=fwSh*%tY3s%=H$1Wp$QBR$xa>mEjo?#miW`?OC$$(xP+=$;s%=d5M%Pb
zuwkyGc2BYBzEFbe$Igfh3gGrsl$y>DL}Y21F1)$zi!w4lTnI~e@~n|_rSX>ylIo!=
z-?E59Mn9gq^p`Hki4D-QWyn5en<@km*+H(NREB%~^dKWF(w7&-J{3;9Xy|ds<->TE
z`_8A4JMtV35UvdB?1~S+ntm5D;eHQsN@(_66bh>lQ8yUnN92&Nx=^nYz3u;7wm6xv
z|2HHv;*9e<gi_(9Xs?19<adw1h{;+hM%N_p9nBH~&{`=T`uZTaJBDDH0y>ammqz68
zNF!cFS!uw4hZF?BoDvXvM**QXJ}3Y0sh0GF5X75H{TxYbiyz`JM8*x$W1>lP;rAi<
z&KT*l30_I(FA+~0P~^->Wx<o*90%(xbp!q)t^%9u#lx(3v(|(nEmuMj9h6l99IeET
zG5n=j7EHN5mj{N1mOrw|$>jMTr1D3K*$-PZ!1n-`lC)S!IV3ze3%XQ(lk$Uw-N}%8
z9TlZP?tDSK+@zUEot$x%7Dg5@1Turv38&SBL;s2)LM%UF62cm$f8UB$A%cZsFwgt@
zz}dPBfWLn5F0cnBHlzt?F;4o5EcKxki=_SK-Vovt9hYr|=S`x0fz@>o;fr_I-J2{X
z{ADt2dXPniM+hhJ9n3V*!TqpiQ}7g%#y7%twhwoJSjwQyP8vps*cS!jWP22U&w#4N
zLa1s`fXVjd&c7cSFdKF03e~_cFRbT7z4|fWr9@l7M84Y*id;i56B~THJ!pd45KIat
z*a8Cr#5b^g+lL{})DaMH-VU$=dmu`xoOC2wr+ERmHBt#QOuDJ7LA<{9Al}Q^Q}_!m
zoB+`^GNk<0Agt8ME<2h_7HWX<0WZQ;WO!#Yy!{YB+KNw3)~Wj7e1O6Dm7nM$G8`|~
z8ha^7`GFCzkRh%IY-{3!*Lp0+iV<zohdZ(WIJ5pzu_r(A->8lK7en!4&FJy%$mFYv
zp)J$oF=c<jdi<UGshkHv>q~h0XE{%^PTFqqS%eLCYdi@|SzXnMtc})*jmnr;GgS+A
zr-6nmB%);tk^U`MEC6e}(<gKBUPcFOtjrMr!5jofhbd98rXc5oz?|oBUSJhc+pXzK
z3etZcWXoBWf^Nub?Rgqe=GJT&bxpyw;^W>oCimr6S@c06>{lihWrfcKW!u20phG`@
zkT6aUr*ltGg-YrC0w6E?!as;Xvt5!Jv;zCa5y`MAF%QSbraSFsc4GF`Mcn=~oofPO
zKiRlkTz>S^?hK1!qj?c<TrIxBU!1`FgOR?ujAMnXo{9M*jhnX|yx<nJA%V958hg(n
zW%y%7;E(SJ5!MA8gcV00Ul_C*O1Wd(C(~uz9Vr6K0a-ykqcCP<3D=%Itm;rY1M}oL
zR=O&2>z<EP?A^QJ_Q2!n#8;-Z!luoVnr}Ok2^b!VwD#e05W_>s<_(>y45gfwPz0T{
z0*F4!x<%q{hK&jM%z*nrCJ>vB5G&iCw^nHA*-*F@9_pgEmmhTu=80|%?^!;x+s`m^
zsmth<b<hs1L$GtepS%P}>>Vp(g!;gND_)NH@4vq^l<jhLg&N}az*JlmT59US$oG9g
zl%v(#Vlip`$)C>d_kW1p{jw`}FH)&cC6$)rxoWDr2iyG0Y1iYId(Dxf{gR)paG46Y
zHldRwHQ|HIi9jgz3HN$PwxRA+DG!EXshLn2Lphs!lPd5v*nJjxruT7+c@dt5DQM-y
z;iP*yJF=bEX|)XQvL?J<DQh6^0PcXeMM``$D-g~lEVc=c%z_iUfV>d_nb6HlwCpZ~
zvr)Fo?4L!Jp^BA-#lgLm8?hlfZELyw$MY8Kd)b-3z(fG$b1$J6w>pVMzf4ytQM;uR
z+h3}PUGJ%v^;CTQ$?0I@I8kY^cvqOsyie9DKsys)@@z`Y{WPKhjPTt9eK`ULp2y^F
zWwyKU6-7y<ra?B`q?n_^K%NcTH5xF-ogZk`6K~)8IDZa&`6)1P=P!?{9hAzKyx^x1
zix`|)kUM3FW>9A#+-Jb!+yz_3W<d4B9zU0Jk%a99cIKn7<oUlm!_TM1t{xtAA>$Ao
zTQpB}J#x6>7A+oB_k0Re9vp=_&~KXfTAkO2LdBydqfPjgs_V=1L5zY>1?kVHVc`N%
zr~$@#YQN|=h;VKL2X|2=;~X{rZbG(Kw*t4m{C<$*U5lGuO&;HtWG+f;VSjuQszMV!
zk}6-BN_&+?97)1|cD<ylU`HGUTbMDN2Oq8@x-Q;-K22S8&Jb0={P<Te$9w(7(&zoD
zszU)<iyByuc)^v4#^s%!s$P%Yezur0Jtb!J?dIx`eZ)Q>yN38U43g=FG4as;bRXkT
zvu^l`4_fwy#I+~G0gZ@~LL`I4c?;QO8qbE49nGdFb?l~@Rloi0vxqNKl%hA9UQsd6
z!(P55CIljkgq;!`QhuxOM!*R8=R+Gf3_R@&08h{PUO+yYf&u(M1&`AB>Yk*IKWV!;
zX|z@VR|QC)-!PzFJR2R9gF0>Fqz5BipcgX#evud<rry+i-x?^n(CUF~HOC`a>n!S@
zPP_YMORf)6!L5}_z3(|w$#M6mW25Vi8vN!KY6i#kGpmE`0+HE>ZcdJl=2MkRqy;O{
zGK1`4;X-ajPry0-kY%*)rkYdtdo+LPsXB8iEm<?aB`N8JX~E+UZw4C{x9=HU_X7BM
z$9}N4qB9^q`ynvFT)(?HJjZhQhkXrIbpXv<lyVy!p{)IjWgzJzi2?Z1r|PF3w2mE^
z(|tSf<ArxMFFPT-n=U=v_fum-tb_8|lKNqPt{*N}jwz?(?F_E1K-0dQt^t@%8RB{4
z?3+E&zfe>?DQsIG*`kJ;4A%!$nzbu5bV;kO{EBM>e&fgIg-9xF5ef}d?sif~06udW
zm%=}Ru{rc<UWmNOQGe!0t?jb%#4|qQ^nH&68?|gZooGDW?wano$1g+G0k(K21RclL
zv&PzqY}HCq7+oQ)_$ilCSuaP$>?>RNj&hy>`tkQ3WW9Fjyz8tseS6`ikAePxGK>1D
zc1tlMP*yVoTSz)@V#WTQpfq46i(N|2i70r!jM}6{o$EErGC3j|l}bO{XlW<C=}F<`
z;@J^)7O=8#!n(;&5Ugw4G@fZ*-nun6ziA{~^NrqcR!`wMgX2%n9bc?FH<gwq1GteM
zG512a3tY=Y=<v;_wQ>TYN=Q5@bT;;_vY2t{>-Ve@Z=h$^&&qPO2|mLuzvr;H{q%3Q
zB+t+F1H#0gpU;<65alzx4kCI8sIbKu(7|07q?OOe#g22I)jaOkQYmO2BiWTkQd~5w
z%RBeJlzf99$__9>wq26={{Of4(ivO|5nJjwVhi`XibtwMltSJZ*GG;A+jCA`yxXs(
z8yvW)sO8`dN@EU`{d{Ht->|C7*+e%rw;otP1xee5e3X@ij;|N46C(n)r4cWozC=vF
zb%Y6OuY&23Uih{fRHiDv51OWW@nnwjox+Rk8@7wMUhM|gm(xe=f^g?N&38R=K!l9d
znv<;XSM&mM74o5?r}ZrA!3_JB92{ARMg}})VyDJ-K8~@=y~Lq8OB=M?eBNqOv{tU+
z%LhB@F`{r2L^snw=WjA0xKwEOCj@U2%m$lUKbe$AGYg6Uk`Pm)wUR4Wf6i=v&`xf~
zd6P}kGQyYZWxF-{`&j6WHqL1MbZLWzW_HPa*aI-(XD8+GEWHhj+}L6$XW^7D^4rpE
z7}#ssKH`4A-Nx<2y^4f`vZKn88^k-wd<yEZ)kJ)T1xM3+ck5S@RFI|5hCT0B%ej6B
zY(m%qS1-pC-JOrMe^K1}W!<i+=jRZ9I~)p?z8^gd`YUrJms?cMn2LH++cns#XW37R
z0PMoJl)99g5xMVKO7_tMjVDunu?gIm%6`<eIrpC2f)h3IsreG0M+hNY5Fq!syHM#M
ze+4GP0!Si@aY;BkL%0?Ta9bT2>9d+Lqs>e8olE`t&b6tddrIm)t^3Fyhq(rD&y~(l
zK4(m5iV0NMaxjDY4H{cgV<BLtNf6~QA0@Qb(Ay2{-+iTORq-6F&6I3XX4^zs=i#s3
zWn5eyKm16TdLUDCtn4WMP(lE$fxiyAHz2H37|o})`{6U<dr%VXcLM)li=if|lB6hH
zok4S~j2mh<RDZK-z}Pi?(S5J0n~+-ButCei{_B>V1(4j=_@6lx5nPKSXqh>~gN&{9
z=3mQ&|3e#Y!&h64DkQHkk4>B{-1WRssCs7u+x7^S!+!qoOj1Ojt#kdiT3_Y5$_N@^
z*{Z7=b1Dev2IRkS_3}gP7@<1cfvE+U4R3ZCb#}=D7TlH<H2MqIg^j=lZtK_MB8Z|D
z_^5ndHl&-ZV*a=~ul#(&Z!2Js0>85OA5u}|e=M^9D@LG8u=fTLUS)Rp;jbB?&<<As
zVwuE0lUqauVX?<xU_=ZlKXl7iWDHZa`08efuVcdysp&owED>)m@dR>XGdeiRfIkob
z(&DM=GB94f3z>yEuv29Y6jHH!l3xJy_}3xvWbXV&H~Oz*%K_4d9&yqIhzqD0&><a*
zo&_-aUngOTX4+6BmxpM5U$o8dgw4=X;a(QqML&K-JV*jPrZVHJN$WkyN)Jg9IC&NX
z=ASRce4s+g8;OkPY)>OL{~d#l6Bw*-R~WgZrkZOK!YOvre8^m-_MxvWaL6dzragg-
z^BFi7KS;SVAr7Me3}LRAIO9Vb_Z!T@!qyqw(okaJEF9%-NB}ZJ%9+r+Hu{}#yEvMF
z7E}>vkt6U)i@F3B@OcUt*vWfZ3`mzXIK=NnZxohLFvcHy5nlaE(3{d9Su~nV7HYvK
zz=HWzXxp?F<m=+FPy`s<@Nq=EQb-Gh&74ZY8Bn}94-xL-G~!<nCli@WY>feKPZF#Y
ziQ6Z4A3#h_E&j79JgHttbrCMFDXwndopio~t{7YaXCGzw4;lTRD@;#<8vLJAW)Z~_
z5AQ}kJ{4DTtc3xfec^V}+#<2^{cFa6f4m=mlxqG5?ZJ0Q<-_ivWlvFfZPtBiKE%S2
zmGo#t(PSEDdWl*Zw!$M-HMjAoVUK58<vWWLu3u%#JvYqOens61O}!fkZ<^6?B(z%*
zF^Y7W88q*q-})IH(ctMqG8w+fpDZ6l5+y#&VtAx1<3G)P3Rve98_v7nHDeXUk>N*)
z=SP;GOlT;yT3Ar`M6j`Lt^Pvxv0I^LUSI=S5b#qIYsg{Xxea;5NfuRRJ*KTHRLy`v
zQRRvrs~Tzqs#H1)o6Y`|!p&sYM+)8VQfXVh*gt*Rtp62NuuZS^K3VkU<)a#<Uv@6}
z;=2}BLP}2Do&-32H{$n-;)7NSPK1CI<{)mU8ZB}cHz%TW1<gme`oYkI&N15uo`oZR
z)9UH&nlD$B+)UekaczZGi4+0f#o)~JZ6o4WP{1C~3^JW88Aw+lTrcu)oj{4^yGN8=
zGEZO;w3BCb2|e7U|E0$$Qn)_LvslM@TTv!$iYia;%Jus;N5g2tW4gJUkE+koQGBK}
zFVs-tI~Zfdhas&{mN`P525#9C%4zE$-L=dWP@%Y;8>9+KmT>{adLhkHuNgt1r&`5@
zZ%L)j1NmQr%{YL522x55)iPA5PJ{D-#r6I5XA#ziv9=C;rZ)>}<v1)<F*AlV&*%!`
zZMurtV|2-69y<%MdeUHGnu*<3JDOj&$h*>LhtrF960^4jl@r&c$>-tPcSTphw)No#
zpfDK~Oi1jIdlXE!Xm~rDi0LwhbGQP&BbR(>b<P~joBfTS^uZqrESr4g@1GXk@#9}N
zd#8C%LBIOD<Gz7blLqN_7!iYgxc*Rq29hA|mil%mlMu|*gc?|2&j`^8^Sh0`^`TZ7
z6=#*+)a*_Z94g%S*zH1j_-LHbnI2#*P(nDXW7?sm86Cx9z}(gb$ov*dPJo3DM-f&Y
zRTW`nmug|Z6EcPNm?w)Fc|th9C1~B<f$i9qA-kuxmZn}cBRifmJED|5ZY)i8B80!C
zwM&RJ{yvcMhEYfzu4OYjmhoL-s)J0+69d%Q_0eW>Pq{~Pay)ny`Wbhf_jI*&71Uh|
zP&<FuA#c|d^PLf<YnMt0>1d@A^D#*TkpjR>m9Ri@IY!Fr0>Ww}wgvEaHx$s@4_2W0
zR#i|VK=SxW!MqzA8ot^Z{Ssr_vKoyjY{V+IHLd>tXnXf~DA#v?e56c>2}OvhB%L%J
z9Li}bMUr$R#8i?~vWha*n8zVWLMn7vOeLumDaS%4=SoG0Q05>xXGV;NnR(Xl8f&e+
zKcAh?-fRCp`}<G5ih1UK?)$p0<NJDFvhz2rc4E2qD7ow}wtl<s@QE7-a;3(=ms5I#
zHFw0JM@V6!<G<tP+s}mRZvHhXhf$2MU7Y2A^l?8+a2Q?rFv$0~Pk&+hg%gj=>{M2Y
zyLQZtA6!qY0$+xp=1N=7AB+*54)#+>9*ik=<=U&cIlCmTbpArU>hz-NnEq}3#FeG`
zeU<9&&+dkuu(_>h6++G=wisiIAH)xk!>kst9sMj0(lW>)U!akBL8Xb>GwM%fy6kM$
zY0O(@E9)7*!e1FbGvmGIfwEP()8;_yHQeX_dnpGKJ;_6fJL&J&!=3qJU4if#(=qi=
z*sHQOOvxYJ!{mc=c`h1LA)`X93H!q86yL)3jm@d$Lnr45QMAEM2zTm3t;iETZy(8o
z)CN^;KwhXtb;J}l^ByM)kI~<z=zjbi=S7?&a&D|HP2LUh<?ICo4Y_uvTW+p;n4j*w
z;({_KYg1NdL_lZ1cCHj3#z_8rmCsa{+M~<K5)!~#uFMTEqfbyFucH*MH8h+Y4jZhv
zSrAvE^_kU<rfz%AGjrdWef~YmfpnPfYV*r$@^#~BxJ+Dd8{D<s^jFAm!0$M9Nt_D=
zn8xdgme^}1&)c)VRJaJMC*l!r!VG#HYqrReO9(Gn^Ms#r^)9gcMXDx0bp2#)cD&8a
zW{RwXs>+qBHAS5`cW8^qgtilldFaMwE+M4YpS`T_Mn-?JKZl@<ZXEjlc}YrBb48)y
z_n6}WxT6nmPWN})SFVuOC>;;vL&p9H8}w_7szC641c0ZkOK(Pl9M!fzN~2OEe50Hl
zcF>+)mcP8V+jPgYgYdGj`qM=IRapKw<Up4Fp<$lZMz~GMMjd%IFyF}B=zFlGJsQle
zRiLh}8t>V8%wONkFtw#y&}=MseS-ntDdWmHP0OcMGK;1EV*&=@2e{}H%%r2tBlg01
zh1FQv%pG`s_<CY{<U}%Fr*@zKh=~}X<Y8W}6c#Z{Y7CsZX?wv!y3<VH&);ac4|E*l
zU7DkcrS5c>tZ5K=Dn8|SE&opXTGY^R_*AV{Z?2nd_z$HYiY<_}><XGpTbdPg0OrM(
z1DPle6bZps3QHHsqjx&Um&nRY&vLv#ErdSL*}KV2t%=N8i1fEWK)J#o&!urlXRKrB
zx<}#H89VT$jL9^=)Cj&G){8$uZ)D~+4^Vh2Z2ScY8Sc44?Fhan*CLYrZntBDq5dg5
zOZ(i9hXb#C`>-fL#ZJ3!&laO`2`D~fmQ<x^3tzuF4AZ8LRvBHk4hZ4^<5PM?etv`U
z{AbVVa!m(z&Cy-<V|wS6kM5VV=I?_57sfcSl#$YIW?p+VS0AyC6dXa<EH~Mn_ghc=
zmhtopmanxUh6NA4PrW{)Z|0F|b+a8Zq{;5U|HhoDmBC;XYhd|UK`w{%6H>YxTTOH0
zca<@tDEZSaz<4|+4qZ@VTf}2V8mb=-UfbfWrg1T^Ke5bo;V%tWgx9+Or3J>~!V>0T
z@;F50kmx~|NPJld7HSeRM?k}F3E%^%473&wjX^*FNrgEgY$k&zNkOuBTLxo@Q%cmK
z+boJ(d*LQhNr|+W$4^NQ=GORX`X*ZpZsE@`F>`2obZEAQ`hv@^I;#dcaG~WmI=I`z
zumTwjl#CWO3;uMa&!-_gq>>b6K|)XZ^5RNUwD>EM0yxWD<Zt9@#(27D?0@1K-RZH9
zdui-~dsh&C9#mi(36m)-;$%)5I0--7C8<fZ2Y#}Lo>l7~gB72n*AR2Mss<xinz^Lb
zHQe@`4B85QN_2<0F=PSW$(b$_zEJHvb}+w?n)*Xk-RabG`|FCy52VM(ZY#Q1wSH}%
z1EVF1ZT*z(k8=iR_>OlLD_>b~(@$?JWPFy$iAsT&b?2)t@3jzjce!-<qh>z@&v~_%
zid{HG{w)mlXF)rWZ}H+zTP;iykXl~nZIb!%+^$!uc6V)b3LJk~JEMA*)Z0njp&=C*
zop=aF^^s#(v*^aYg?=e<8MK}Jl)MgeTib?K>+B~*O^=<w`T9C0tu|P-#&6^AxK&!m
zU)?5XO?ZzITpwhrc*m2fS*UJSKa14Q;x)6Y%bD6#y1#l~`x=wxi=;f?;7zW5^$V}M
zOz*#OXz1XQ3$HaV^g*+z7^1659o(`(bXmGcmB?j^W;aw3r;82(+Y~b#)A}OP;Wg!O
zGMBcMIU0|y>@^yH|KOyJXOZ#Sx7nNLueu!<^WwzuTxmWxgiME}g&=<lr6Vzhs6|&7
z?UV>*<Qb$aE7F3<Pq`7yG4<8FmArA#kw2r=EWg>7rmi}NdudCA%s|)F__7dW4`njB
zk>dW2(?-}g5CV9U1SyC*r)?1bS{%To#IlqP^mp5%Yw)2Ki`rUxusKOK=p!9(lazGb
zs!PhlZasYGjKm9s#P=08FK0>d)&5NlP(Xi&@Tm3!F8nM(2x`oW-Y5Q*KC|bHV+G?v
z`z+p(I}ciXoPT_2arPa5k)Le-F8j4!V#CI(NteiP_TeBRp7G$HatHPQ3GJPz_W!Tp
z`4hCPFjD8$zZ#tTM=;q2GaQLfMfxDV$m|cNl7XuT5<quzCrYKAUqE|bL*CdH<0as!
za^kp22svI-C)$XhFjRFD0%?mXCfw7WyCLKk{}H{zqY1Z=qj6&f41lh`B2_ZP8}i=3
z$7E9*#dbJdNy-1u^-TcG5lIq+(iYIeHo<|?@sdXS!8+EFJBgI3P^2#|g-93%^@ytA
zrY?b>>_a+N0cI0^v<HQ7;{GEac=rDBe+rZT%w~|^sbK_1sQ0&vZ;<=L`iSG#Hh`lE
z0QNIPN#-SvoX-+pq{G)VltDgW*2KE#IHdGF6EF;|{(T(i3x&pWX$V=;Maq`gE?20*
zww&PwVR4DjpLr$3>{dsY=rS{gEU+ojL3)W9z=3B4?wd7J1JV&~T(a2Atq%X^G$p|9
zO*T3AkIn^Sl7vqkk}HOiigpJg-mVW#i`O6_bQPNc?|G$5=m)Ng9-ve2%daPPKl~%x
zGqICWT?{qk>s66GeDrh&MQk-<0(&yOHkvWCw?q$V*~o|9^_vN}(*yr?^Z(W4<^Kj{
zEs@B-gHZXq)1T?xU}NMFc9t9DHNrO1C}<1cw1Og=7Z<%l<cK1-(G{$1A4a0hwOv=(
z>A2n6clLE!v4hjDy_YvWi2UW?)yu&4+LMmyoInpk@aftCHCY3)#L>Ku@NBZsb(B&H
zZx5rjVY@gnm)KHW!J6SeC?bIB&>*PvmMm&8blJk?CI9Mp!s(XF9_1S=&J`(@w%$7o
zy6Zt94()&#Mw~eehv_KTgf<Hud`e~&^Q(YOx~!J#%aGsaW_d8^(cS%}egdy!-=!|9
z>aX>CzbXQBz33cLaf~v9++9-)(NIJcoMBb>l$1Mgt2}9BalujO0XVEc671y%h2;@j
zo!cp4O>Qen_@}N0dvb~x7Z%e@6?0B{PNko0R3*?>EDT$2b}?d47{GpKfWFo#S%Q4%
zrfR}52ZMhdE8%RKX-$LDS!x~67@IlXCCO2M^AL-sK~Cfj+g~()q;?{9-=GCO&vBM7
zS*T?nLoyP$FzeGG+nfa+If<qSV=6~#5ao874vO4GS9DPF3a&J<9Fn$LTNR&g-BrKw
z%HUL;D^}3*4A*rVnSGDN3oL+eK82Xas>o`w-zUCayq7P%1EmHsO)E`;F6L&MH0;3Y
z^MqT+P9-h&h^vwI{(Rbe&3fYWN%lv23}iSAD&(a*0!ir7!GR4N80AY?k9z(_Tb}cQ
zPVAw}`$;=|^gHPrmDa#GUaY2i@#$|ib@Rs}X~&;Q^P$9F2Y4FT>v*EzGZfmhfOg#L
zHenE=L@*WE6rKYD(O#(zU0!5_mU4*^&dkba{|oeW#k~C{R`HKox_@{NwR*?*E*i4Q
zntf%@LA^vJivw1t)mxz~1QF^_Cn`Pcj>&@L;nz}$n4GuJ6?GTvt^<??a~930(aeV9
zKbv~PZ1+Pn)%K&lenYuxQ}b#U^R#SPr^gN*J}cL5e7+8fxxF_-EOv{ZF}?eUpO^w$
zv`@IJ*G;349}`(hmi~g)_m!(U?sRhP+v(iy<gB@NMY_VQ)vJmXwKwEWo!6z~h37-C
zrSzP(5{`cbzZz510JW}po3I*ffR)`48KPM{1lX&mligVw-n(g*fD8a@Y5B0G!%o#J
zcg)Q0Hp}bv*@d>+J<2UvQ{PAbSbZPA5Yl>!PA?p1{OFOeq2wL(@z2W02%l=4bV_ii
zcD_Qz>zRjSY~(0<(gkO57|Q}=u{4RL0^Xu2Pc80__^Fx7K|-vV8pnUm=7V}+c@Gz6
zM_06J#pY?<yRs}Iejssp<NMSZul60#zzvYO(+4k7hR<jN$hM5QUy?eK68X_Tv2rp8
z`{RW>Xnx|GBgC0))W<?LEA}LD+I}?leXFgh-yv6vWv4X+nod!Pv#!>@jCs!~`1SA(
z{JV*SD6yU*QtAVMIX_7>o53bsDp`x36@&n7dJ$zh+9^_76!i!V*!i)mC_!JNZy?G%
z_1vcUr}_^T>6JjWqWtH}L#!JW&17KHW4H`-AznnHpf(l;UKCX(#~;>B79jn8ibAM|
zVU<;brURGSgOb}kzYH=sI4lxcP6IIclMqWk;sdb-LlDR_emO{>?Wm{EcBNW!r_v3s
zx_lhI^}<d5*_C&G^s39wt6#-#{oa1U^f(AeXVTBB(Nc?m0j4F3z$p3~-a^o&wJ~%O
zfJwH8@(v#6y{j!7oYCE!qVX{=%IpNQ<>~(NM?Tw{E~mv8tln~E$!eK0HdJUIPbSk~
zJT81Lfu`XkNq3Tb4BVIr+|JJurr5w?ilD)&fvZ+wFGZChe(I_!q?N4Yr$-bMXNyt{
zKOByIW3SwoXtLC`*mpv@Hoc9mSvng!CNNh2iinYjgzjBoL-noQ^Vcn2G2iCYm9@G-
z9XjK6(^6mR)gcT@=U5${uK=Jj!Wz_%m(I<qh&yjgd^#9o=!9vbU6mH{w{}NY-CVxP
zM;Y~We(~XJaqObFK<^VxM-E$!CAi=(dr7~Cnw>G}JaO2geJvbQhvh_OSR6$!XXcK?
z4crmxisZx(16c~LRFm=!&(HxgZ(mMy+k>oQ!}}g*k^H9pI=W-pMekWHNu)WgmLaRc
zpGo%uA(aEdeh6uefXGnrac>}btSG?AnFk1b`PLQ5mgRI+URD&N(!}$k!HX6395Z{r
zfB>y*=lypQoLl?^MJkc{@t5RqBGA)6?=}7=wOO#gk<9+Std9Z-MNU^#CfP&FRB*b0
zJT1#ihbwfVQBG%l?#%+XO1aOd`KEP}Lax$6na`z*x|X-)T$+Z1`#s5>0-mRGn2UE?
znL@4R8kW(Uo?^|fiRT0R$oM6^0MtgLisb1&FAjD;)2`2rTcWyrPOaWL1F3EbUo)wk
zFur<lF2Xr}=a0snpZ633xu5kTgAeSmUFAgHFIWs%s;T!OCFd6)kR+A8{1}JOG_LuH
z>WVYlLudv_E@QxC`>&6J0t5*byT*<hX>UKX1AiH*P-4wub`P5=Kz_{P`9xB+hy3XW
zrXT~&_A9YQqBPTM8Q6*1(stb()S8dq%+IU}g`vrZnz|gKw+5ZcInQ-6dv-YKmnqh7
z9<Sa25!{!(|GQx)LT9i`zaXW}bR1xu4>3bn*(r~N%h8-_HPEnT3sTgd^X(#|18>RO
zxos|b4xDgCSMMawUJzU~_i93|dRa(t>7*LEcNr!N9^W+04@~w$a`q91a91~(U`RS|
zq=v2KBVpAMBd%N9_7=rbWZO4;>vD?n<dtl5Rt#@Dwti*6#}Kj|V0j@y71jgFcZwM#
z-lv^KOL@DwL6H`wRD@$dl(QiAmB>A7I2lLi)u<@UxLNf=F}vVN&5<`x?k!V_h%Y^&
zr-P%tBu23KQ-dl=<K_fu;HZBrncMWeSczBPfhpjgm7vvp&7IMLynJ!ioj6WPaw=-F
zh!@zd%qRuLC%dgM;cIKvaPh^VRf0ViXJ%@O2H@I#K>nVD{C#aB{j-;+-6k}GSLZMl
z5icYTk)`9EuvhJ?dE(|u^4xAW`GThJG>_EJ8`5erzv%zwW;gZt4c%Dfwi_zn4!#67
z4h#mc-r^^o!xr-*FBY$v+vlfuq1gDbm(7V@Cw-P<!;0LI<ZCBtRQAiNTQ6LJ2RzbC
z`%92SWJY0&`9T+pZGq*79lcMqnQI=f-`112J%8ZX&HVI}^e1a%SH4sJJiSIyc7N>M
zkae|am?<|&tQ{E&jQ7oCsFxtB<lLmvM~s$^j^@g&(B_*J7V`>Gr;g;H(jbj5#RNZ@
zhby8_HD#TirL4?fSQ)f*-Dmv%9r&off+zeisO70};$bt<983;$5aSZIipQ$flIAQ&
zrkm})gC6-->r_V?iTucLu1E5ezp3ST>~okcZqT{(^y`6OHlfY_r&ayojg#4-+9Y`L
zM%n?i8N6{stC-9`BI$@C43nn0^c0dKM-;SOm)a5CZaMk3q!k}=y5ezd<AadZ@qhT^
z*&*b`u%c`xZ@ZnZguX(<hmzeC&L`NR?g+w{E_E~1*83RPmh4c!R-_lK-`1jOyDK}X
zNV(q{@1t}z<&u;rZjx&+5tYB^`4SF`Q^@T1geb~Xx`{{!Wu6qep(N*g0FFwRVFnDF
z@Y0)aReW|*k-hniZts_**u|Tdy>xdxHY?wLD-IKQ{Zl~R_%{P`2yO8r5XQ)aLrIz<
zG6oNy*g?W>=tGR}{6L!TjwwP}5)i3Syxj_RyWw#%41sf9|3CHY<Hx2zP>M)MELsEl
zI25tr4m^AI1bBX21;ii9U_&}s8$)1-f>78&A+{T=CfFrBRBHi?{uY7Jz6`OIKE%-<
zAdZH_Y`7g$5=YCE`>)YlU?mQ>fUetmJ@E>e=SCdcgdZ4{EZzZr(hiVsZr?<GH;V(U
z_XID7K&=58&X%-A1(9HWBJ`)zAoQHiOO^DePYE$E0gyg4&Du}&v=<`K(+)byCa@PL
zVoqeNph!<aq1u7%9O36+HE*G|>IDJyQ<J}<cnpegpwa;RP>l#RCAHk1`d;7wteWT=
zY-1krA6rLg5n?q#Wel;s!Nf|)@anJdn|{YVvAPAN{$J{3o_5pEL$kc6pc--jxd2A!
zFP6~OH+Hx{isFH^_%4$j&VufOPb$z|04q#?0nv~P?9u-jx?d98i^Z_p?qtbs=MbAK
zDdRpsoGF~xhavws#?VI000R3(64=HA`cnxCp<{4NTLcAXcYny35A)BipPTSk;Ep`9
z7ioYUTpY?9gbsBWM0^^?RrOj-><bD;N>ZCYk5wRoy`t0bzz7_Y?D5x!gBw6&yaE9X
z!>R!hL4%5JIu-Vzy!#JnhwDv1moP>5)C6FyvVU~2#HIl7lK&5tBT2wa2%9D1S1VkK
zNyBDtgaFd{uXCHoEBz;y1X4l&H0GOk7Lx;OhX=Ylgtx6W?l@LU)*AMY3c?JM%5EA*
zH~Gt=ej~1HgvR?Xz4p4_>9L_JV)wU27Q07o6GF+{RfFgvd$BxuE|YfwjyMi_-be#u
zyg`T;!ZV94*~DmQ5tPZ}jYc9K=9)XvAig5`(iKTWarSs7wPYAg+7_&#PHQ;rldpGl
z=saZw<tl#N4nnOz8Xk?w1pqFd^s-i@4?+&efr*NX2`It2^d=|<$3aY~fa%cd#<R-5
zSZ+cY+!wpcG^WxOn;Z5O#_g&Pd^X~nSQNFFc3q+PH>IH<3+p*!A!M~~X~9)&I_T1A
zI26>u)F$zZz_?bIjx<J;f3dtfuZBxFryE{&)7Pi)VPo;m^bNiTYJ4)zz09_154<s^
z23~UeWbYl0K@+$uxn&t#GM*`yAW|U;oa-&-_aNF#Uw!owEzvqAe=w4@k(W?Xu>6D1
z^zo+Oh6-C+$BcG`FWLQLwpO<^uK}9}PGZ=vP}-{DtpQa)m7t|hu|T@m&Hilw?lR`X
z#xi|aQLUq8nv6&!hj}44D;hH_s4I4vEyyuGUo0D>&3sgy)^bkf+gN{pnv0i|_TeH&
zoq2WmVNLLN;*iqSNVgN(369$)7Vy;)&`W0lJO(+$+G>($0jR)^VelH=0q)HhSEa8*
zI7+ZfL`Yodz?}~}3~S5b*FcgZiV5trME;ynk%}G`!HA`xo9w_5?3P%BY9{*zd3d?#
z1f28&f8doRMlte|L`v0yWO&Ai=s1i$nbjd6Hx?`0qfJ9c1Y~s6atob4Zw^bfjAl*{
zoE#LyUg5N|)*Prz^sA}O4xzCin{pf}-Hvp7pz4slDgFaB{<E0C;^zV;jxnxImq9rG
zpn@Rmt{?nFsBl1dKHC``;2y!9v23Eu?4{&+d`sq%B4c=d2xnzjdv)1N<Vdfj6ZIzP
z>6?bI1G|gfZl_kil^O?M_UEm$P!ShH&C*^P*rrl&Jcfjoa_V@&M;e*0!X{lvN10q<
z+$GiYG82P#jm80$^<m~FLyp(>Q{-Q%r{H83jO73XZzy%rLl537)PPm$BY?p(w(f|H
z19@F5&d(ANGhxGmQ++9lBew_<<Avz0%B)P0^?7}=<ypU_D`Rvs0`|Q<miE~!t@Z3h
zocPA1rvbi6sU4XOdA?85f?{zP(}&oS&dCqupJ}TG76Ww()v|m|A6@xM&DQ8$&l>~V
z?)mnmu9-ehsnzDj^tFdlfrH`4y9s^S3{^@|WUjqXzK1#P1bN(|B*{d30CHI%faBUJ
z!EuGI*JgpilDtCt1o-O<8NgpQ67bha9GwD^sVBt4iV7$NdZJmHg5$g_7`g85f;xJ#
zNoqbCRmt~Dx&E`P@AuYNM4O!&wK{9Hai{GovmD0t!~FF`_sI^tAJ{?Q^}>$OK<9?J
zN<chH^ch&4TDuApKsP0NHe9g<WxCI!NBdS>7Z8rBdVk!r5IgwIg4$&UiAMSZNkfNR
zFf2I^GS%#B3|<DAb7#1AZZDaa$IL@G;8UD4whVT!!N8QoG4lNahXO=%liM4m#&jkV
z7P1wSeUr&^%WA=6a?$*=6)YJGTn|wx7a-GA$1?v8g^d(7^yKg@nTB2Lw%O-TC6w2G
zBRr3_Hh&;JDi=bAi^<M|^clGlX383iR3Mo2?62k>>RX?~AB@hcmGdz<iY;Pl9J~C)
zs~~5;tL*nrE3KBy8>+H!uh!Zlje-@vuolSe$iu5ekghCi|D#K0LnXZ8Izap;u{|+W
zI7}#8lDUc`0IrK?Fr~@70?@L6Nr4+jW&;)PV&+R^7IgK-MT;zXM<}v14d5rc*uFS=
zJO7fS#<>zLH0VJamFW^&_Ici#_L3W^U-UEPeRFy<uc{YPE}2X=uNGZ&KE}`go-5V2
zBlLOwooKsfOV|&cW(XGNXJ7L-^Vk+|)DUY$P`@q9<gSA8F*Y4Q?i0Ho6yJhPFMh`j
zyucx5C6F};-%DoSVdQpQ5;+R3P<aZv4j+n{QKOwuCTF^#zgng5Z<d^_3xyHY(FR*a
zG~KoR2Cr`an3>os%~*PX|8-!IWD3~N>O)FcAT&ivUlN~Ii<F7%r1Cg)QTl*xD$S;`
z#00CRYejMh6>Wo4(c>doa{JLFyJycn>Rq+`=&-M8yrQi)pQAVZ`7a&P(!bzyC()vW
zD*rZ!FMWj@kkU%l+>>$WlInh2ge_KMlPa_{H={wR`)E5pHkVuI+xc@UR`-l|eD2wW
z_Zm2Sb6mxDMaQ?t9iQs(Z<u_%nrK!RcAgT-KwZiw@MS{KYT>7t>?8_g8rg~DoVY6c
z2y(7OPD!H89bht$=;705#7B>_4&U^69AkJcI$-s0W4p)RPv3r`m|qVgUVLzvJb-oL
z#*_f+Ko)99{FVEX`jR<l63`i8{z$1Yn1oV$3nL%q6uMM9>QG%dPL3cIQvyiQd7JA4
zJMQl@YDrk3zR%#RxN&e$G>`ZkWa{VDLX+0?7I4cOfrjphqiB{in>DbGAMS$g4q}s}
zwETsK2#VDE=GG2?79{1A3w$@PKU^;UPM!M$AQ=D`?HZ8f=aeyXkWz9KL#z1sY!Rk}
zI&ti*E3tXewjp19R(ls7O5L7+szb*6)xoE0YV&Z&O4iTTN?H?8Kc}iiDpx!3k$HpX
zN;G&~RTqk_`PNn+EwsEKD`T{8d1{l$&z*ceH@)ck+f7CfgCD+ZR=!Mo7G9#bYCW-5
z71X{=80fF9BKHR@v}hZjSUS6bw)^Z!uDQ7<|DKJ>r@pHt>0UW{Nkd!bJ-@VtSvIEi
zU`mqcZ{9trf6MZHVtis?a^^fsOd$i+1s8T|qquN6ZTDl^$vV;I$0E1l5_E4`!A06)
znlaykI$}3v@yM2TUEfaw1HZaihId-v0m|t*i75YVhd_GlAxLH{!p<jW7l>y2a|-%9
zIpmg(t$ac9L)sF4brb~as{F*TaWmz8eHsLT-;<p9G{j^ldBt_kx>r?mq{r4x*qU*M
zV1@V!T?c@+xk2U3mN<@nRkSZi_L+07ry|G~{H+-G?(x=-Djxg(YJu*j$f>dRb0PCQ
z>@nE}n#;rFbj?*lDprap7b^*mhz>McsJMz&h;PCa9U6uIr6AI5uXgI=T22=JZDC@%
z_SHkb-Tyvz`_hGZGfLd~@mIX0$Khyx_H{J1!BGk?C9`kR7d8)*6+eNxJkj3QT6N=Y
zOKyo=!XcN;T>^Q{0G-oaM-MsG6}d;e?t=KxfqK?CkH0AN4tlX-OTIG}ci-2+cj>@2
z=idAs6qhJ9^sv(Q(B_=Or*11quH=6c7tjYm7}qh`EHd!l%6amXp#y?#^psa$U3#3F
z<HS2_$*omhlmMsV-F@uomzDX4^<L>#CytH1kbxpl@MPw+PX|)~OA%&6{{RWrt@h_x
zx&G%MT^M6sdggo@uzhOJ;j2a7DlL5DJJZ@DwOhyTxmxURCEM=PciOjaw_5jEO1w-G
z8;E8NVscqRbEI^s1yN*-#t1j|V5=Gp?4z({AI;|9)(g)))wCkDF*kXIzH{eeuA}`T
zY5%3?XvwuhY<B}Gv}iKZ`gdJe!hI;VpzFE>vf@0l7g#iA4zcVDp8F*$4%nYU%}hD3
zM+RkIyWgum!?XKEVW4eW*_`~1EDOR~aL&g!O~~MXhEs%pr{kaeuZDF`pbNa2>~<2g
zviF1M?Qw`8Hk7pWzu6--L;yvJZHtr)w#k;~lYm_>7=pcW)xR1P<UlYe4Ae+Mqbyt<
zEa|3CcgBf-|0|d@M^N()%$o7(DppAb%h1DlAdQl46Z(bNj0Y1{`TJ1C4+F<({2Gw8
z^qgEzd}lAxN|3aJORAP0?i@WV*-%o;)@}VO!u&sR0mr+L(hMTh_(O>TBicqta-?PA
zkj;B0fDbR#MB4mjfzCK}2zb~b83gm(-V(;PkgJ%&c?HDrpfj+b?LL1k=>IiC-|skQ
z5^B&4d(0~Us_IIhRq(zB0zMIxBhy$gB13cmTs5LNhHeILvKfb@RN@ebiFrpLlrO4g
z2sRYkLtzjC{Y?{TEwTxd{g^luLu_+|_KQp8;cbAn)43+qN0dUoK9}&sW<QniV(nJY
zhVG8qm*k-Z)MZ#DNfGHuc@tvtA4dV%9SUmNfkDxk3ou5oOU&WdaELYk!GtKuhkZ{H
zY5K#j#);$Lw2wk(MYA~$d1{EGL6{?P2CvcRpD2P2v|up*QC`v_GcW1767UILJ4i-E
zddQa|<5#=^1^FGB&IFJH)Ht(Gv$Moj!$9LBk&XXGHz5zZMCuU9E`C@7pBmvqOuF?B
zq)UeWNS8ny8s5hI+CX-<7m~EW%WKqr4`eIvCw}@bX)OQjulVyX;_v?kVZgt<n9@Cv
z64Dfyp<}}1BG-IWQ)rAX?W$%f9~;*9thnr@nqxUN>z>!jw5s$#?NeI>Z56#0$6nV;
zJpmf4he*X57H=PU-0%hLyUZE*gLs>ePi6xfk_}UW`8h2-LX#60F^{ssTO-TjT2ci;
z{G)UOG?PDN{Cj<GLpfbLwu$>L>wZ|1lV1Gh%El%621jZF*P5;;=hUK6(b#k$U<x+h
zMc4&!_8qgNenbw0Lt00I7F<kUgw+spiRHC;tp1GALvEG^O^085zSul{QSyUF?Y-kA
z;io;_eKP8%DHsghol=3b*Zsu~hg`G6xje~!-k6n<+qd>@!I_9Jc?lVo;kxe{Ur$vE
zqfYCLcVvKa`A~<j+ZgDwq^Ll1%?fAh{>n^OZxbvb-PhT*zSr5yCE0Xu&iU~-zsz8V
z8s_r1NU2Lbhk1Zd<@|Y1q_O!aU}YfH^&ztlLDw`O6FG(%qasxT7US;%Z<BIMl4Ot;
z65#jT;NRO2_d#@_b-yRC9D37>=!;uun&LD}Ny`OOVMdO}<h}dS9wW{R5_~bUF~4`>
zB^JzlZ?q0VybTC<5ed~96l<wQ!Lz>DO6Du!dj|Uhg#PsRl!Byk__}CHj&bY&eI8ZQ
zuK6*|_;iNz1~a9~M=dc8tqI9Hqvv(abH9ljZv_$~$^`3We3xJ$+<bc&rAw7~)38tr
z5b32r9>eqoWmQPV(x3x(+!Y$7cEkC2AkZfTO=J5|8&K7D+e65&PlcF!vK>oF7nr$7
zB}<B?fZE)#<swl;<&kchF<U$Zzoqk|;c3oI>VW%c^_@<43lg&5;@VhgI7ll0J9|3i
z8v}&12`P2Utr2Q%{UM+<adR66^|KzQnWWs@bgJp<!XK~JGxQGc%sCGAx^|?TfUXV4
zWbJuzoUHQN2n!SOJ%eCL*T4*#0tW_Fuc}|zFoM6jmG2lftUxZO&u?(8SH2`LDY+GN
z^zFSz2Adwv`xau#m`r>{)>s>vw|Nj<>^LA4I*NkCIb@H)(AK01`<78bG#^Xn>v702
z1Eiyw1I4x+Er(@ZM^COj_$+Z<s?EK;<c)7Gc6iQ%w*DG~KZ)>&)yV*X%x<n?d}3j;
zsk|h%H3adp`>A?d&qy=Nj%8mc<)f=B4f6BH(?lvxTMkF0tX>qOex4wqtMclW{xQLN
z3a!Islqbd&3Tm9V0aO0f?a-4#n(4KSYe(8zlv|u)8#Dj@O#M2l&Pln$roU_<18phf
z*B2ZDVFQ`?@WG3L>O<o+SOyCwZ@)j<y+x!5ykw4Xs8<JMPe4hOHE39iKrT6RA&9$*
zZIJOFAn3gnZl_a7r%XY8N{PlNy+(u<61pI(zlZOMP1h1yQdtvew`BB)e@h>`tAc>D
z;kA{uFZq}jiz(Pv*0*bRHYhKXHn{k$<=m|sdr$YZ0Md#y*|7|h`wPHCynt5-ptlg`
z1d>9DI15v6x*>XlYwNmkbPG?Mm(aiZb6ZdR!shJdS2sQlopEf=*B=9=z}oo+_k$2a
zoMSePC(WV6($|P~G`aHz%LgM(6sn`loHF0UEj)89-iLp0bbI{yQ+vEGAKYLD`JHHl
zuc9VW8pUFxA&F6X4E&I)#nR_dkXHOmq;v_kf@X+D^3q>2bI-?RP+8od&|(+fXld)F
zGxXb6d`gz`?}t91`j@x*n3gHI2^||V^%Jf~I9yYmHv`5k9ejY4<}rAVncFept{idF
z$a6nK`I=jOPFCsnLv`~$*P1EgJYTV9UHl9Phz)Q-YGDlrhR7YNBPFTWA&D+uNaz}a
zR_7Y*o%XC|Gg+Z2O^#<DmC<I;S35^n5}EOuiVl2T)6zSx7h#ZTx-;*{{)b*G20AZY
z{$(rw*E)Ow;2F;&rBEqgha$Y)gRPdBoTZRLJAv8=&3xJy!H5JBw>^z(kMHBLK5bTg
zc+dX6_b=8b;$N>k>5$mIw|prh>CSsE=^t5>X-{N7*cejVsNet`+ET2FK3jYaVc%iR
z%(Rff>M=#We{`{TmWbTQe)gg$n>QNsIMd(El-AI)@X2guL2mK~%kMjW4TU!K)e!4k
zXR_m%!-39vfSVz~yeGV96GR15!B0rN`+>}Jd5G{YLi%VDy*dQ)DrpNXwP(8fql+Ok
z;gM`Aa6})u^EC!JgBNJ7XSokD299OEPxgQ4U9$OJ>@l77?MIo5<?I3ZgMNl#W!0F%
zui^rxC+U;1;443<GON|I`V^BOQg}$yX{H%_X1JD`@oOHn+g{h-_%gQi*MKXlzOOou
z!$u}s5JYZDz#u_Hy1R4jMXJ3N(IH5MPi9Et1|SRar^vT4a`dWLEhIr8I*^9;btHk>
z!1a+8Fg4Vo>CL!ioD#LjbxTKs!84DWx)+U~d3}<JPp}RU*AX~flO4{<5?DugAe?*#
zS)n+HopqsvB))kvnN&uX;`>LM`Eq0IMzq#l-Od|~$cxqwa49+IRI|F(J73Ls#&5<-
z+A=nh!Xg#H6s_TOt*O2nKA5e0gf*Pef1Jq?>)3rTv^~<0x4>^iYw3GU`A@Q{Y?aw>
zm8-VQ(WO@6d_i|?gF#=caT9$=Cr&F&nwTN>nF86Bazq7YLQpQa5tw$4WhDl6fh=X}
z*<CSCyM$XlX1W(YE!=!y&homgvC_C-K8MzanDk1VZ^95n3gIQ>P7(%IFq<GDG1~^c
zzoKV#`0lQLFnCk%i3@<UtX+j47lU}jor3^JNo2}4KUG~$UsRG)lvm(Nc5FOK*_oSn
z^4S%mBUe6Z?AWvN<=0CG*IRr90rX_DpCH4&XIR*bVHVUC@FGidSUR{msV;y~<at8;
z!XF0_R3Ai_6U7?spsvEr;alSBV5eu+BBgK*x|>FrBK02TMD^qRSxg=eFx7%gIE7lM
zV8B?q@lsgnX4p}Q`?QtmK}44X^?%{BI*^wNj%CCdA_cBy8*E6}**^8V4FjD|ZqE5i
z<=(h=pnfIg8uV~s*M_Brf(F`nU6lX>c8_bd=fs5?R5#H)I<n5uNj%p1+y^bY>?|D{
zhkNGb<XN4V?d&vt$G*^BYnc%8oAk*fCD3gqyvcDUJJW4C-!nqw%_X<oa(coWJ!iK0
zE;ZGy#LsrSNA`zY1%Re}T+!li;%$DD(bcybFI~Q_ZXl5nGNvdgo!L@#%G}g=6o^l}
zgEguvEN1e)M{$BrRpt(C%1g<)?VVq^{h?3gfu+vvf_iEce(HTI7^bT>**%v`ux`uZ
z8lS^fV4qmINfm=re7rMI<7%>+8HH=F;Pdp+!k}@hv<I)vlk1m-yveSz#JwaQtPioU
zzfEWcZ@L_tz84N}srV82PSzdFQbLFZ6tmRPZ%r;N1vIREylPI?04L$jX(iq3Yo~11
zNuX?-dnClXX0i=J=q#R}f4{46BUT1{95q?t<lL;#@5NGl8T;rIL;iy9pa4s=$Sof|
zqxz4XyKH*DCBa5|-JGwA8B*fuK$jRfkOOP66-P7KDXm>$#iqhNw0%#|fJd|q^;vmY
za&2zTU-lcNrF%D@ulgQn#9zB}_N$>+?!OL8Kf=)@V*FpViKO_=eny@xN!8<J0K5wf
zo4_^s$XaMc+k`4~3<PnlE568Yact;>wEH=l<sD1Tz_L9zMhNp4+9o^5>L<7_FHO>$
z0)vs5=p<4Kb*{hZK-0lema0Q#ox$5pZUM>|7vU0G5Lzp==xtX)J$ZxNxz5cOMH|e;
z=bm~Op7`Xp(NiML2fuBNj+iQH8GuAUyaa6U4_yMpe}_%sU&mPEu^{9dfGB|^8M2#A
z;3t6sNrVfPxkn0p3j0U^f0=yDUZ|^r^fUuc3jgs$R;G_8{X0&F6~^rM<8@&v%OSA-
zEK#km<xzbAP22)DlRM+)xHIo202-4+H;@O==?6++@drNC@v~Hr?$GtbW=ZT0cuGDz
zzP2BH^g|3-;xMmKN~BhY7l7fDPzw&y8xX;UKt#M%5BLNMZxg<yV~S95f_dN&jX~&s
z719$olX1uu;1<b;!vKZTNn$<G6DK9>z#F!H!6CV>lC?>Cf4@q@Dn9AJC+pV%@<9h!
zdI4I|iU(*#rN7wTuouVIfFRo!K^HJZcT)vw@D&QcFt`kmQzJ4Hk+P@?VgCTI#ujoZ
zqygX?$AchiVhX&6LfGK>;{Z&tm#V{$3fB{>Jw@7~@Yds~9)$Ma2x5cH>xthCfei~_
z4ds$7fds&ED7}I#Hgo|mr>hR%3E@~OwsWU25P^9DL@YtVg0W<l)Br2T7-lf^i{U{<
z9DA6-AgKp3B4Gf6^j+kW4k*|9wO-P0e`Tx4uLCf}*RW01qTTQgJ0;s>=bV5;4EOwl
zsDa=~Nl*?`FT-yp(JLuXJJbSw-w70$UvCo@p=02Ikvi;QD8&WVcx2x!2|YA8*zto^
zOo0&^2gH_Yxl*rTbVM=~3t%m;5Dy{@V*naYHK6hA`yVaWxD`@*g*@cV$ft;t;y}=P
z38#Uw+~c1v#5COLu1eMzm^~H9siaUrzwLYh;au4oZkH&*KZ+&TQS1%O42#m%YoDD!
z?)*#Y)`=dRWXgol0SBbSlG=AL#oIJO$PtA&tQFMCYvP!xJrt5&8HbTa))s-##r&7-
z<eyBOgnuG*_!q1v;yMt?yQ~6I|3jgUh$wyxcnG}?@;M9DUbhwKRrWjE^8A!VjXF6=
zd$O~6%$8%v^b_7B#lo2cz9lGB!)R!%oXHsm^SjM{+>{kTM9so~m#_*9S;Ce51@IEt
zlmS1X0rsXuL!68mQH{B=&pDJ!C8Q=pv1z&Uxu7JCBeza-<Xvoii>_^~T#(4o&~?A`
ztlCTtLTCU-{r}VoJ;iDP(~>UH^iH2#Q10gqQ{%pI9XW@>6m~T^9d@sptz+e0y8&YN
zPq3raB2`CBHtnuR8?{gEMe?%BYV2%nV1}aInjN`GE+MwQ!3R0PE1JHyZ}@h6;hM^Y
z+f7%;w}zN6!-(M3POSied?BfJhrf~_35mRM1YM*j)*38}6KR7&t6mH@HnIWsc@?>P
zkUmQ?T?Se|h@GRbLwK4GdQegkgj?W6zKGz9DfiI6mlpFa=3>?S6;URidzx8C8fedp
zqa5cK7mT^25P(8_DO^S(Jn#{)%dtMO1dig6Vox?Pj;U>?#A~P=WvUl%$Y|}RFKE1*
z`{LHNXXmCLtJhmfdq|M`96kV@iaq6%E*+VP@Ph0I5`+fyX6=>Y_(8I#f3(|dG|i7g
zQlP$RtJsAe5tvricNKq}k^kZIwIiPztb*q0#P>-i3akzxKZ-*QY6KZ#usMeZTl)lI
z{2(aR8J9C=m;ov#+{7&=wYNmK*w)l{U^{8%yW_g|6u;sV4tiXEuud+Z7F-q3P~|;v
zBG(ziK}<O^?=&dOxwi=~C5=OvoTG3iy`I6dC0=04)2Cx~?NVq&$AC50vD|!XO}Vq;
z!`V%#$KCJy7!OhQTg&CBXk~ASnO`lH1H;0|?^FRxhOj|4hmr;XVOH}#9uk~>$6)aF
ziV3y3#wF_8hdk<`T|>6_DU}+EkB6MMK1ls?J;VY`kTDGJ00avrw>wL88f|;rL(KNC
zTsENRj27|r9O?%T)OC6J8w}W1Ilt>nXw-l59fnqTrN!jDS!?z!#4H+{TLAtAxL~Q4
z5_X#~xIKXFK{EAV3<PKJl4QV5sWAY|Y9om=12qR{P@7A^xa8_<R-_OcQ&+NjOny}}
z9xzO6w4SEBIo@^&vBCsuTit#*8W8|{a=yYD*n7rILvjQ|9i%7=N5xmS8mK0jFwW<F
z*|p5is$Tx9xBlTo)nE1=*}OdRnR>^<#rUhpWIO$%1H||-;zJH$<#`S~=OsZ#CkJ8$
zhN8%hmw%OaL_u>$lbP|Uq{PI+o4wKVa}yfodrX)2Tli~06_`tXNQHw1D7`EOL|S?X
zWWdC?2^~2YVLj;YU>dw<_76OEa{)j|G+<y00W~)6NGi)}X@CJ}Jbx~u3Q_OIq<D%k
z9OhgXA2;3RW2(*Db5n$N1+h)*JdQ5*PTqREcHvxiaFrX7Nky1`8Yy)mvopXAq(l>W
zV0PLFoa%B`E5a#=5&>IxxQPIsPUtDDSOfh0zm6J_W~XIr{aXC^>GQqPsQrG4d4>AM
zNNvmSIL8J6+IF!MUu3v$cE8unP_xkV8BZG?$k#MCwc{tooIZPUjkfBtvtAxIt<05V
z!CC4Ly^EB3G1=F7Uagx}W6g-%u6=<u)9+b6-f-=T>Av_zQAfgo4IKqOxQ5Ae0$yan
zz+|H!oqJY--UP5c0KdtTs8{fB2<ae6hma5LqQ$V|Uy0B=MnHqfA{7-Zc9dg{5#~WW
zb4?2KV)9NDRM(zEa??e#$z0<~a+t;15A^v|{4QwWsIkiiCfpMR&Vi$aM{WgcjTCOX
zA!8d9ue!PR%|1ma|Ay|75~%eL2&L)Gl)T7DLxvi?mLV<j=99}yh>aZ|@?spz&jcnV
zHkthT<kp1&y+8-~WgG9WmlXmtUP!P!@X5HqI+KwPcSMe`PKc@|7(Y+2M6>#^7y=MZ
zPKOui2*e@fiE#7NDZ^gKY%nLA!CV7)Yr6P0IOU9d>;9)~A%#ezp@fdO^CdHjKhFP<
znOCdieWlgO_TlFl<zCrqcX(Vr@#uZ<a|0=UFz!rJX;v6Zh3Q4SpkZ9b3U8nJvG`;M
zYvyV2uP}`d^L&%7g#E9-*?-;mT)#8@?dM<2y`%>z=uhjC8nqUwkT6+2xWNP%a7xEs
zr&#c-`=9ek<t0iyMwP$3uf_akLj?`5jNBKio4lR5-@jSym7TRd?ClnC;F}kPm^E}*
z-3rd6X`xS3dlH9Sdz}l)?X`LegYQ`e78p0Tt9|tI)9-MD!EH~g!i!cy!BONd@&?Ot
z7<TSPJ6c_tl*q~^S33DU7TIO(Z+lZXtJ5PacvtVfd%K@>x*0jl9X9Wlz5xQI4?Db_
z*<FV%gRQ)XNct}Ob%1op-IeFqA0;$Gtt-a;s^YjL1(7;WCOWgn(aZgL!-zS#yve9J
z@#>vYn*%qEV;YC%9OzQNO?a?a`Uen)gVYG7p;~<D3Z}fpQhGg0u^2G1(<xfU6?Bb{
zBh&+pfsg8<<`)!3+}L(v>Pf$Qfd{CsK1+2WKN~`6(HDf>HP}jIHxHP@hHhvs(G$t>
zzeVyjUnUV{N@jcU(75vEWkKmBX=&cZJ6Nr<yLH~>$A`UozE*E8<jwOS9swOEvjzP7
z`}7t>;ARFM>I3sm9O8FP;y7w+fx9k<NA+Or<9xafXzTFXA|u}J-o%@$(eMgm8}9=r
zO}{Pk_R>8em3U<8f-XIlOo#;>?#~8VJBb&G$;Am}Xs6N8+qC7Vt<Z*M#xIB~-f7zA
zWufsQ&+C|>)>dlWD}GzWW7{X|_l208sl&4xrQV`7RTi3PFbq<tYG3&=Gms=trEDu~
zIyrDiuwnN7TlbyjCw}|RwJh5_<3P^SZanl!8c%l1@dUmpCcg$G;e8{fN!!?P6!AqO
zf{a@uLq$&Rp8nbeQdiDCdLe_!Q+N2UJAEWx$!gsZEhyUgV~beMy@;~Kyk3Vjy6+P2
zbRe@?PM@s;<_ZeVjEDq3=6}CZ3FDD_U)JH3{_sfCh`bH<=u+c;bsmuAAB5)9<!FKA
zg?aXI#j7|kHnzC}%fKT`{|7xeuOHekus*J`I(wCsLYRp*t3parlKFcI3GVqM7XB-w
zD4coUBxcedO%@tb%Q4xu^w0Fh6iw8IO`b*d<u1zPy(8!Q(MXwn+Vf2A-%9WGSL7>H
zehC$B+e#m%@#C&bk3~*?2kN2d>BzQM1u|s|bNUCeGTe?k<N96wMQr3EB2P19vv6o6
zcfYmO6Fw;_Fiyq(QLoAN<mGKsn;i-RAD&$9u6yXfXHOWQ6+PMNr6;vSf7cJ3ROhpJ
z1tL{xOt!p(qU=Ri;xPvtdFA`~YEg!^=9-)8GYubTckO5#{APRBcFW9l`pbW8iN70K
zW3=`ij-K!rK71mU`28Q)3#2EH0sf4){(HJO2~s~ilG&+9EYK5QiDC}-0l6jG{Ey%D
zk2!+B%CVVD9(cfGwxj*B5EypBwQVpefMMKf|CCd}kIexXElbi+5eM_!Xp+MD-gh{}
zq*1aqB8vqyJk5g7<5a~EGP4zeSu%T=FlOCcN*0n<B0YDPg3Vs_+g}iX|AcA$!JsGx
ze(^VPW2;HPlv(pW2Gm6*Jlk`4;@Phtz~3zi{CR2+_yahyI~+W7fUGi^yrZ5Xpw)LD
zM;}3Kx<?9Q2L8EKfBc2S_>1P6PSnz$llwCu6uAGwB805^M^(Rm4K@Q@OqeT|4MAZs
z72Ne5XW!tkwu$Baui5;icv%wA-%h^pAkZ4}N3ol_K<UW;De@zKu>XUgKvQ@OX>sHc
zB@_6>_0WF;Iu2iefYz*HfYv-N(VCEzxRv}@C>wlMh(lmF!5av&C6fT{VG^J)xo(2e
z;LjKS{U2Ba{)I3}^|(z%jrlIzh!^7K@~p@d5<f3L$>7xZB}~0;aD<wXo?!Os?(?&2
zJ!ZaMh?p$9>~wmT6bmWW!5Ub@8!VO~9JR;D#CY*nz-}}uo@sxPg>Ht<_$bKtt|m*Y
z#wHb*@u+mBYhujuFw`JqeicnYTX=n#Tn%SXi8=@credAj=`W##+Z6W*4KKf$aw9$S
z-TN4ye4gf`2RavQ*GzeOf8f0pZtd85a-#{H*}vfS4BFn1HF8#HPO{B;)@Z1;k-N-R
zQ*%qo*;_?7J}+FkK~CnhZv3k~n;nk0hY+^nMw#egr1dOzL9fzZv;g}1Hvw{9ohV3w
zF^#f(aD%452!Py#Bpf_%Ely{4htV9M&_@Dr2uuMcb3k0Y2gm0zW;1!w&Ezp<@aO*4
z^Q5b2dpX2eR85yc9_vz|f@8UfIJP$5Yy9K2aeAuXH-*<-t)K7OYL9dVEq?X&<ZVSL
zGgC^*Q%Y8gZ0f%@!K{_)Ds0sU?z6Kth9pJw?9PuzQ=i8Tz8qD(Y^(8xS#j?CEGdYS
ze>$Vn)v=h0Xd6n=Jx5le$#NM}YkpgJ#rrkaPh)h8?<cNb_KPRQ7j;%weO|Y1z)S=8
zqZc<IQh@Gcd_|iulf)Ser&)K{Syg>GHn^N-oaWlwY}E7?bFMWoI=+=(L0PI5vg@;F
z9_~!sWCu^`Ahe8jM3#${{gtUKj&&Su$?XjWMn8NtM$hNj8E?OMHq2rAy~5{VYZGmp
zw#3+aT$m>fVJcuRK2p&6P)-OW%R#7s9Hc0rNe)lya<z0iViSUX0|)=D!<pcS-@3=&
z<c<dIo_%pn`09s29_j~Huek{)caVYlGFr}J(Sv0Sk-h;1hC|Cckn=qIKR$GG_HIS0
z?IT8$nww{~I<>k`%dHc+-+H;5FHWBybu8LudA7+tuiJ@7aMG6-OLxotp|H$IuHtmE
zXWWe5@$;cvQVdyP*wugQGhaK~t&PP__gX2RKHr|kF3eBtTzG!(ROQ&;$}PS=k+ByU
z*7m>n9k+QDl5MItNf5L^DRC?KfA1|;^U~vh+mN9tvSBloytS9qtq_sev+|k@OuG%W
zOgNnS@2Yz*_O{wAICbgrT$xwq>#=!+u0KY=2_az^z7oo+By>M=w1KZ#4U9QkKH@{G
zlA%g&l;?3Txaf8DErEYtjaqUiF#f#lc!5KLh`eTU_NZ8m6+uRwy_lk;^~Bc%kv4%}
z12t#nClb3UjyKBn4=eWk%~^E({P!+<Ue^T!vBP)ygK1VJv?tDX%kH^^YQB<BSp!L#
zcO@Yv6j`cqCEcuPKxE8j5`3`5&EM;7^mBt1tAed}CePPQqh|VDv2k_ZowT%A@mb$`
z?P|at#0l+@4^t(BJI=X1hyJr`QVEagZyCA9=L2IbDSSdH5@li(mFDuIJ?*z%H<gbA
z%a8h94*BtA|2v<jmdaP8O|u`odaW)KVzxR&FE8(TNnw6=PMFKBqSu_fy!?n6SNHG#
zyyfYJsXHGmAF5fFar|zW(<hwl+S`QNY62T^y6fI%HmTCVK`J3KY$RKK-{F~?A5NWl
zVkKgNqx`*_{BLLm0+AmXMstO7g3GOPx-qXkhhs;de(>2X>*kCV24AceuaNT6z0%aF
zli+@1&-G7_=T5O+zgRj6(uc575|GL(rgyE7WMYh^szD}(F`@YXxASIYj^t$YSZLP|
z=Di5onNuGay!y-c5v{FQV19OXew_I5sIhWF?3}AQ0ng=3>+t80^8Hv1YnVcZsF34*
z7jiSMI6HsgKJ0Bl&`jr65fybg9uQ#QAmDB8&{ywFIsL2ktW|fvsGeRJpP;l}_eKZ~
zz_Y_MPFNcpf2WcW4<pV`%zsmS>`r{+rLc^9E9VU<p1zH*!-x8i^Q`;Z?oryk&!krq
z&zsQG(m&gJ2fGJx#oF?>Y_Ag!*V)T|k%`^+d3N9d^?frprYr`C_GA|flA5I_>9V5z
zDA=88PU5HZg|wxpe+TJ;n-t%^;&mQD0j(}GQM~h*8npkZ$;oSn2Dg1&e&A-)(3I|;
zZ<acbJ9YH%Uihz*sWgb`e->bHm;bA2MCe4GmqkG;_Li8VH$dSGw(9N5L6<zHCUsbc
zkN6XMD9Y*fvXnj(kF4<TnaBKd`~3E|*`D;BR+FLpZsuW|ee=K8;YS9bfi6HuLOuke
z;QvJ2kA{dUP>KcNHK{D_qOqEQsYqWywXAA3*}v+u&q}KA{-+;1v%?&RxQU(mixTd{
zd>xzpEAHLPZz1L@LdZ!Z)HLX&_C(5k>dTxuvuPaHvsD_O&`zG9zsF6x;n+!Aj-SHY
zDyMgOcWp=sbJhj5<(=(V2Z<oSw@hLU#n&MAog$ixX7U^{xg6ng(H69T_XBoa(M+0C
zt3vtg+rEk)8wrF&LDr$eJ*hT%3cIj5`hM=7^QCPUgjFbVL&zU(F^w*P8<<}qFCALy
z?9_m{sau*$L6;B*Y=mq3m`@xBd{etwnkDMJXC9~VjH?}P(itAyfd$;}L%n{B_Fimn
zpZ@(uXQCBHX~S|Vlo#{^rIxR(Id0&<v-)9g_OyJ)&WGN*iH@>!y;60Pzl8l3dS$!J
ziN(?^j`%9#ee1%ZXW?w?%5?`))_yy`Po$Zdd!lk_WbCGcK9)MEG7k7@TXfdVlG2;R
zzWrTcAoBj6pLCVlAUU4nk7A^&OJ+A06Ekm7A6F%CwT!xKON<(3RIbf0id{l`h<jsE
zpZN)p9A6hpi{V;IIJ4(7^Si_e2rQ|0K5#YY;}CTw?=+*FaDkx$UVU7Pygdr|1!c5J
zXoqq-Y+r0_D0jB|mU-OMj}$hr;Nscq8Qg@s7j2|sTwm@g$8|{{>}JG&M<bT=_4?cR
z!qQgV-*Hi9dwD@zy~tdJrNmqx=c+7@pL6y0r{VLr6*xMozMi$}`vGH9sjIPIz4<1%
zzctMH9~#X6m79x2GI)n!%!8IEJ96O3cvaLrpISmCWy*5WL+|{ccgh9lHY`(dwOkUj
zZ)oG}sn8xakN7!BG_Pbjv|E{$F;RPO@`OwKKmU(fAV@CiYx7L)3?(v9->7@pyyp$W
z(ekDtaWStJ6%@p-&sgiBaK2+5Q0El^VhX4|Mn!kZCkA)#G{%yl&l}jadU4}jf@v3*
zz+mFI7H4&Ctd|Em$ZOHPxW`E+UbAAHdN_XW?2`ve9i}Rk;-UEY9WqGMNnzyf?Yt~Z
zr4@pk1rk!L|Cj%T9|NY)w@pSS{=jl5xRKM4F>q9WfraYOz|l-jUO}9R!}eG6HpQ$y
zZMx_14@gWxZZn0+^CX`yUIzhMsgc>~fJa^$iyr>M)Dqcg)IR8OFwR*zU1p|jz|<>K
zwUSPm<G{`7^TcLSrXqF89EwOUmjNz1;RmR~@RmSxH7iK#?I^T}OY7O_FGuZuYw*RQ
zzGAD@<Cp5wGIq*NpF3r&8gQL6A*-zqpIDI_tmh=oXL?^a8BE&#@n9U=DsJI%9|xo6
z4^~|jKf-#n-tPKs-P~h46xZrO15z5{r_(3C%Mxh?ar|d^0}_XEQSFRpq=A;n0DT#4
z(dNP@^#(%I2*;0x`fqYv+U|wA#Jrg!GYFkpufK;ZcmCqK@AKv$kAUNzC#A<O?{9sl
z`=VmSTEeUkt8Fyd@z2$_h_)P11oGl6P3YO};AV*ypY#V$m5|DGXK7O3`?r=0yf6Bz
zQB(Ha12MS2PUkA+quV+qk^kLgI;pO+PafFfaWG6#5$MfE$y)fj9aH{r2cjKCs%NV#
zR^)(kPRBkc=F^w*wK8*wS|aPYMuOIF=~HLk&TDuQZ#nx-wT9xGi@%u?zK77=k+PE=
z$oO0djedk;=7gaM4SqoB4I%_Rbo<h)b^@o}JO0hLC{ljZ@}22xooRD%F0r59FP@Sn
z2~K4m-+J74)A*J+E~B^5J~_@*XuPxN>5o&xKj<~f-Hx=k{fKRp0tt|9j)Cy~INuSQ
zWdeyL-8vb1D%W1gS82IM{)<eX%Q}NwPHOEMA74nn)ns}jZ8K1m;ev%bWI&%;vUt99
z#=(amMhCHJPQ4BeHzV$)%tpi9n#Nl-i-MM=xy;s|+RcBu`<3o?!UC9vSpJh3y``Y~
z`0!#KPe|8giJz~?5G~`>xLmiGN{!+?%dZ=Lr591zo*a2`UuNQ#PoC!;b`JtHjPHA!
za8!IVexW#(j<a5{JpMI5hqKGX<h->tzbPyuHZ#G)W8vy}I0{I`f4^~?s#dD^{}J~t
z;85o8|M*zdRA@>?RHmXR$yTJ0M<<esj*6IC$th+#h%)98A%tS1*o=}?NOH<?8pleK
zbBLLta?VUK9%kn0caPoueLq{fpYMJ@zwiJ0U)Nq-8=iST@Av(_kFWc6Z>}}tc?JaQ
z?~XMCSIbPTrSt`I{Qa9Lx!(yG@qg)G$nQrWd0^YyN{<1rCkRV)zIUjlw<#2e_19{p
z1*EQv?yE_1F}QgwcYBv?*$-u_ENERiQld7fQbKIYx(;J1DOueBcmQeK6AE4QH9Zr(
zF4s`ZFd5&I$br$Cc97Psg}P2vu`hz6oVPWS9+=7zIL%l<bkG5KTZ&JwJ+(1iG@BxL
zlb0Ru>Ux<z$*YJPKTSpH&MDdTULUTcB+htscNVl8U;GPcAg0iW{K_y&R|M-Ws)*gL
z#WiULRwn!@4`06Q(5@}jT%N!9`o%?lKbjJOeO`FaYicx4+pi^94YMe{6nWZAf8$NH
zs7N!zNQ>@RLcfNzNS{efSH@nF+Ux}n0!a#3?7W|_A^y@OAzI%1!!>GY7WBEQVeWo+
z?sE?q6J!ZLPA1SmK);hxqL7HPit&X!oC%WSD-QPX@!VXSon}4CJ-()+??TQ^hX*c-
zr@s<hT%=#+6%_`$#s&XKV6VLLcHWC;FixI|ZKBPB7JknFlowxzPNWU~`+q@RhvK9i
zPoMIrgEm5yR&4Y6LQPjso_0lMdz+WL)f-O!O+z=AVBIqPV>#i@2WFH<ZgV`qnK?~X
zN_6BeuCo}t5LT4kqULNyIFgg^s@##DqJ~VTO@DJQ>46<(<<1UmOY5civ3utK$oXaZ
zc1|~DjJy%}YV;}F-ArK#Y^`=@Gz|I&35{+$Q@y8*2-80tsZ{fB&UAY|V9F;PgDi;*
zCt`edAy}nLf~bke4nsjE=V9n3-jGOMTJW?{1e14_RnnZfZn|k68_Bye>vd&L=%*60
ziw1xB@kmM8g4qi~PF;lz{g@m)?>mh&r$dJy_G<5@Q_vt+GwI>pLnoR&%znB)KhR9o
z`%ZZ3`W4Uo-@#PKYD;2*Xz2lJ2vfGv*MWN3NcP}CLdtT!O6=4{?`N89nshXLhLg0o
z4wv_u&$8epbE6P{(aKkB5>4S%HCL>WxtiAzYrcBj-AJ<+YVYHBpD|Vm`2X6jLhCL^
z-|psJIHUb52hXPISMG!`KVTNG7q3ZfwQ1Q@l(zI8exUGFWnMwfyZHC89^&UDpfv@j
zc?yI)FXY6z$Uuq&473Me8NE)@dV6cLd`_(IY|=^7I}oCvJ?Fypu(aV*8|G;%+iqKt
zIE!^04*9*$8(H)}lg@c4hl@<bEWq#ZGo-hTo>Q5}8@c|DJUz+DJx#}UU2CTT-dK-2
z-3x3{3(KfD0MWuSaB~)n|1$@#@xK^^q(#}#m~Kn5X+Sv-51#2AH+tYh6lp{RE$=?u
zJL<XG-Nx*6@$uFA?hBhYKxJ+WR&;4!IWnx}3PpT%7-<L?{@W2ErEO|-B}`T5Qr@X&
zL8>~#-}fxi#x(6o@7^OO@$#)TKTyugt<a>U)&^R7tzQIlp}=#VMhRf{IJ`l4r;<?s
zt`<0?_A18}UO1O>uTS3O_Iq>9ew*ZN>Qc9kUT@hSFQqz(f6Z^6^o|L%!vC1N=#uX=
zHG-5{lrQ0Zqy<5$fzfL`Z&xQnR&y?OR?mu;JD%jECkGxRPMYcXe39Q&Lw{DnBzV#E
z687+#9x}GRUPAccR;r^nOHJ^((s0dvua5yu+wbnHjoiBOQF6xKO$pOLBLeYgHv)Ql
zaKi|Aji9%AKwi+W1($7t^1l-{idQv5<-j<B)PLvSb<L~>?@G%<3)v+GDRmC^m@m|v
z%H|k_2QM9LP`E!}XtAL+-QZ59{jG$ka4$f!Naj?2vs82)h3A16-KK!ZFD%w*d)sfl
zYp|4Nh3zPj*=6#&^QV-DF5Wk`$)>(vb$;>og%YOH7q5Mn`E__dNX*Tas92kmmYe2R
z>x4C*JY}NPo)Hx=i9cvOAK`^~CF22UpnE2bM@(R*6QK!7&MTZPNEgRGTBRfBK6NqH
z&wF_IUfho25RE$#72~Rgk<WRT!gVYUw_iT6C(;-QBu7`OX!3H1qW$^AO;EcCgEB(I
zn;#<PVn=v}W(o3M1XpDg8B_7qJ8-?xR>^kZ7S#u`si|w=KS)n(B2GHJwwwTY>%XOS
zlJmBewk-~DUzxSb?sI#v%*ID?=TFJZYFPQW?D_Uj(hEE*cMvk*nkR>#J=V=oZnuCv
zgsAUTQGNPwi^9Uq6`xBE+V;;K%xZeF<}2a+-m@l}@f`biyg5JIA~+rD)xVd_6zc72
z&?RZAIKDmkX_csOnEl>FxVG(-(qJqBEZKLDM!0kMKo<n94)`=+0FX2gEQyDB<hHBi
zkc<HC0G%D<^Wf}l&lXvL*70MBWEwn*^o6GNoxJq2xzWP?Sx@T+x*Y4${HCpPuhW>B
z!QHZ6jkmVe7-<T{<(V?*o2mqq)3tj7$UPPWzRDfGdXPKHxz5a8sW6U*o-$<qV7T?#
zmKhK1S3X%?Q2#iiB1-Yx#_J{)C!v~^VM)xTzhZwrP0kz@M-K%c<MB>9(CJ7gOJ->W
zOJ-^5!7MFqg4_i{2|CB?gr3uTzQUzQ)j4N<>{zx+X0oYi>m3|aCvShgfjBQ=3RoE{
zE_Gb#EJ?J$S;Duo?&wp$zImUt90Nag9AHCUHEPR>;r0a?(WXcI^p>lBPMCK8->_YQ
z=9yyMZ*3GOoK3G^?V_I(cT82}S$j*hs5m>@lY*J7GWlTk!Nj3vzHfyWjBsg*5iO-F
zNa#I`VS^?4+sd9F#Np{lE)ErO4|>NhUY1^zQIu1#+W*Z{KU0lqJ%?lk)dVMi?rDQ%
z1&#VrN5V<v&|iv-1BK9-C(I8|7tPo5bajB84Hn;Y*Iwl|sPzU*p^fe{owWRxW?0?J
z=b`Z}2t+B(QQsd(JN%goUpr_Ow8t}ZuKeibzct<&d^nbU$ijmaX2<Pi*Y6KgUpe4C
zS;=Tv3STz-&wF4YEd*1^OI4P}30lMtlR`^d-)Fqwvpdc2UhCES6rR@UW1OOzI{EPS
z_dl*zh2bM(iuf!GNN>E!PKO)^V-H9SBzG#wJnF9Oa)xMDn}D2CkQ>|7sU7fR>f<9@
zdDZh#mp<ajYuo3(=_1NX4Rt^>n3MF1FSc)n$HI#6%%nr_CR_bz=dv!kezN<4EBkhY
z&kb7fw)9-D@m!Rv$;%bX<`eN5Iz!O6K<=If`}?@h`dl)kIxFR!TLYqRFF=1VJQ=z1
z`a4CZa2nNVbAj=v`(>x+URFB;waPh=kgWnBYCb>*Xs5(E<h>0i&yK`!O)n~}D=Ns%
zIy}#-$ZY@G=NF$ZxSF@<3|OQ)<i9xk_V~OzE^Ee==^06Ei!5f8<9+KRRoNGEV`Lu*
zP0v*jFAy3MNF_DchKUw(xRq}o+(Ubt;8iPKCali?pb@w<rmSOrplJwoX3r_}`Ew#S
zc-TyRZ?a(SS?5wsX_o97gRVJ*8f=G13uwMpfxGVin+lv#p52%|u>6$cnaS;izgBL{
zs;!Z^*l;&vt?HScN?xxq^iPf*Ls^~3=j+~xa1nb3G8}0Q#C4FYn#tt6u%(#Q8xip~
z%?B>A=i=)`2YK^Na+`CS!j?o#<@_@HdB8J6`2=4tNDbzp-+#E&cP1J6=I9hsN7_l2
zH*DUz>K_<k9XvHTKGFHi>H{CZl05pb|8(^gZF-he3M66Jyq~kP%ey6l#L3CZjqU5C
z1gu?YX;b!Pmn!Pg+buCF@4D2dPaA@AM<w|$gmI*KU>Y|<zEpS;fBB?}J9t{7OOmkZ
zFw=<}-g<y<6o2?uE8$_n=|jIty?_IMD*mI63lTH%PxJ((_#(U-AsDT6O)O)}7n3mQ
zN|nV5kKcwlmK86kAF(w)+;(Nt$yM$@IZs+Gd%6CUOgy3Hl~f57{N(UT=CB)+2=*bQ
z7Tg{999ck!E=#^o2dXIKc$lKVs_-c$br-AT=3eoznCr}*{((9b<|>3mgece5jUEWv
zF>C3E^6cDCi=!=HT@xir4-tS9jrWd{LliHFs33{7Mj|9w*cP%#DO+ePa>q)Ad+}y{
zN;wij*W=^2?$T!R+yplrI&aa1Ryi4kaTPI^RcGAh)?W{CUh;#Uhmr;X>Mz4u6NT`<
z11S)jnU9I0_#p{NmqRxf==aBc2xwkA;dIA_=<SBHyOeI93qM5D>YVn|wHeLf+lS*S
zO-&XAnn?W;en}b*s4TfFXfQ)~<S>ABnU2oFZK5MNSy@<Om(a$>i{_ki*V2a7S#76r
zEO^q5?VVwBXPHiaNa)d!u3~bZGEb7mH;}K8F`Iq~L=R~($EVq(_uGNx^Yfrr#xLfp
zJn&9)jjd#nrgikiK1wubr%>9f=2Y0kr3A~E>AW_5HDKm0){q{O<nb0+U>j;ALkV~L
zNE-HGD52;F;4hpB7f6@+LD^^{AS_2Zsp6SnN_!x+1*Tm>L3myqj127Kv+;OkU&L>m
zFiZ(T6phqriyG2BA_6)L_=s<|tx4m_^bwLwU$gbL#`ZfKuDK=0HeSCQaC~XWy8Wwb
zh_M7<YIE0alg7hMyNI~Lc%=lJJHwz7%_a`c0ytnjX3<HmKrhFy<yF7p*&cDCFX0;{
zKIn})Vx6g*RcC!<L-5^zn^#Vn?3WPMmjI_Xf=`X$QfGL<<T<8L7TZ_ex(XQC<oX%&
zhf@~XZ*3g3FRJ&@G_BJObakydyv0;!<@3}%tV#EcAoT|>2Co#&7!uE?vH(-KBnt(X
zLPbWn53giAsky{F5z#7LZlr`w=H+e%dJ-d9(T*l?+IRB8uRGuG)h~|W%Gr(XZb)5q
z$JdN{-oaDWO5sa|XSt43zu#m68h%^+6g?@+z+RBAEVK~4o_JZL=1YRyE*Sw^oX$8>
zdkuV3Qs_X$!`_!KX>O@AUp8M2oOIrQk>4gIQ6@lKQiL1qV7_@~^E02#c&AN4KO6-j
zGo-yL`Hb4Ao~PZ^H{DO4#+r=Xjds3!UhUk6TcIku2;RB`Gz1u$8Tj5bAtZiVF7Vay
zNa(Q`LygIOAih&x(W{Wmm@+1&7Q@n3pc%ttqV5WnWkr0`h@T3Y4L5Pa4E8pBQuW`L
zb7-$?%H?>gL%&@0{Z(WA94X-#6qxjcDv)Q%5rIonqG|e@+54Ho^K33b0~u49RI-1*
zVkt=)bA5I}xBD5U{U{3t#h+{4b*r4)ctGyeW9vJr)wv%!Hm&ahy7JGqP!7-&DJAms
zIHMJ7g$`&*1Lnqo;yU?lDT~}f;azu%zQ7cqj0l4Prc0QJu@)0_ZG_})yQ%0$_JB9^
z-nH>ia2C^GJy#?N=D%1Vl)%1MM09WzS>@nR6wjttQ+VduS5O5Q!DYx}fHVYO&<>`}
za=nXnKU3bA94OkAY(weOTisFW{IHi_?jh^B=i%MHu;97IM~y+H-9$usih~ycbgRnu
zhg`dly`+I*iuPsbvl(g)Egx4&<}&8cE?|l27;2sEz0k&_(a`pxo%49uua^{a3UaQ$
zI<xu1rsJow1{eP9B;Bn{5Dc}%;k&m8=x`g`C6%Xr)sWe0uPKEJr&KL6Fbg^rdl%?P
zrCuNd{{>%HoW^w1ZmlvCKR}(TwRehd<163xwt52gI%M6!bq71vHr~n2%F5>7%akk1
znzb^(K5*VO`!K~1`z@ysVEo4gsGxO<A0s{Q=|2MNVX6c)49+SETcRCexUGwVEhP@D
z?Mzqbw~n}6X!eLUt8U06jB1r1Il8yu5&KZno=4PG_W(mSXY1Zi<}(vi{Ctu95Y7D!
z?W)QPUbi3+3VZ&$ifW4$MF;sczFBD|z0~w!Nuk0|xIMd|aO>x$k6vzv51K`<eQY4d
z^5Ou$@z%dsCeA@P)#PxtlF@45Sbq`JnOs(*<b`XIOYmtiJ0i-hKYV-~ON;7z8Ci5c
z|HRC!(9e0BzRZ{*nZwxg?;O<w9OeSAQf~9*lZA&lz<94B&tG@zqmTcQH`IIe*#;K6
z=U%+GZF^FsGQ!P?H??$<W`D1$*XRzlkp%bnyTcxMd6%K&;^qC;DjBXzqaWPuGtzE!
zPo)kn?X+=CIj=UWTGp5L{GG}eL@*!w2qO``%+E(~SqZg+2lV{}MnLYin?oL(WeCXT
zpJIZL;Ti|yg#h2LdWct$pV0FjP+XNO2S#{_dmB9r9;9c>ZT`X<-#Ka(6EmaE;>@YF
ztm4B-*KWQ)_u|CK$TP8o8raGC;s~^-jZT8f`P{^b*lnC#Vn^aJeG50<^D<{$m=VK`
zknMTn6KD1&%VvA(*=i&9DeAJix|tbUR6c5D6O=X#882K&KyM(Tg&xA4_!|naym3<S
z#k}-V^a?r?dVrteQ$<UPcFiuOFKy6C(_Qsx*gH8t&&I}a*0kiGm93sRs;YzW{m8)Q
zhYFrScf&OdE?+<~1eP#pvl3UM9mA6OfzbcfXtW&rN+>r%8qS6uK6$o)J*RKfx~M$h
z*Sg>jIoZ*@H}3?R`ChC2S?0>Gz%8qn0L0X8d`4#_tL@f~Q7)wnQSka<=#ZbL?ZK{+
z{OXQ~Rkx{=;!j$(Z%kPpyZ-%HEhWHR!ZX#6e(zZ{D2kLR!W}far1N63egUthEd7cK
z!mkVhi+fow(F72mx^h+)isu9w^H$NTPk%)8mz<f;c_OITML_QXNhQEWUr;0yh#+z;
z-?bKACkLVvgZ&av91L5Cz<1=9l7$*}GQe09wF_eVGuSq0OJi_@#8EYq(LTL<mc-j{
zh!}WOEK}ymh}M)+MsCoc{cR^o7OC7;-K4Hq59_SjRlm~k*u%ENZ(BD#eEGP|X48Uz
zC4sv0_b3J|kUcAz3je(?(G-QtedH&HN`^hN)X?K!2?}29GX30;!d(wB`#mvxn{}&N
zvgLM<Z)@n@<{g*!BKqL;AN(p4o`IYSo<Oc$1}Dve)J#eXSehx=DPZM{1g#!ytdhrT
z>ta_h{B!cJ$H3e?T!%OLUdE}pO(|5a?S)a(Aj`Z!WQEeB$j0j@BG)s|U3qr3P}cq<
zVYudd4?+5`qyzNRcPa~yMmP<WR!4y!_5hBG!W#_$bcH?U-$j*s!87d~?C*(J<FC5?
zus7Y)IkB}Ut;+gc-sNiP$XX+r#T({9oRnNnI1{hMLjy$K+(C;HJ7`^JLX+!Z|AE$=
zfHdoEKKEP{g#PFC&P7cLaT?Wddo=YqjGiCs!sk&1;5j9_VF+ZB4n|C*BDU-hUqOzG
zNqaMzbkLT5u1Va(-WW@^^LPNmfB43#)FG|*<9;7Ij{DR0ZCjF3(eIa<HwkFUh8zFn
zi2n;6(L~APKN=oJUx~6x*&u>#QTU6k>~W8J$P>8Um+oi3mBuYR1$!lZHN5(-guOmS
zzR0^zz8kd3==0z%if~RL*!X+it+3xyYa%G1l|X($Erm+dQiidL1vo0;$4!D){>2H0
z{>n{-V2SUXNFdc(!{7WbxtqTua>Z`o^Mb;9NnsuZiYI+j{APAOe2P;(VJuhh9Kf<>
z;(N>iKha^hN%M~YzBP8EMFSMJf&d`75xPhzyP%8oIb_%0F;7s&)(y3M!Jh+TlGu27
zIzccFPnv^bT8JpgM$kLq2V01j!cKk0Asu0W%N>PxTu++{6g`=piuf&HfgfK1E;7`e
z{A-{mcM-@Oj3sLW)7%cM6N50OXb7OR{X%yiLn45XC#H4%yAkIT`k-C77Jc!YUc(mH
z6hnTY?|cH(lJYPVH`Z4d`{}U7s)4?`5nG0Jq+qga4kEghJP0S^(A`;yCo8uTXzg$y
zU?e7im-dS>{TV~BX_z72*#C>vIQ-;&9wZLnmcO*eA$(}jkPrvmOWOfm;uGL!(h#It
zVXuIQ2;L&Y;u&MyOLB^vLkKu%B+h5Xm<gCB;D2`n6s<D+WYzc|py_{Xvmm^j0G9s(
zwQD|rpHaTL_$um889$y6b8#jZ;U@$=g^8^#!1tz~hhD(>od_NYt;K!b#^eA7@3fT!
zpTLAA?%N4S?=%1qhELoY;%%BLAav4rmH10VzpsQ5wR*{6<Fn8g>ew-O&=(eEJgoN@
zsh+b^xC)uK4;|aCJ?Jmlze?t~XSA=80yG9Mg#&d%0jd6kC=Q8b;J4g0q(^uE8*TRo
zSiQkhqzvM@9Obd$m4M9scs9gVF5h9SrxX6%0YAv(^Z;pY`e3ebDatZ&qtBy*KrJk^
z;`Kip;+=Vx-{qvQTN&fzv1@ayNnrt;&BkR%WmL{Y9GxRQ20md>KeT@`dj@b2L5?kf
z<|ee1boyr148}dgDAwY;^yRMX+1bbx=g|w4z)#sPXwCI2sgHKGHx?3~n4c0>j3$t-
zLw~m`yto?reh@jCE?m?Bg5J>i$rq_x_a9XZr0hP|$q3C;=HD;N$&LQfdZdhf?$D|w
zPA9^O;`JPiDqy5UHWgdkERq-6HhOHN?ic?elJ<PiuN?3x<$k-FONT=>e*4VQ_I7x|
zT-~y|?d-wT34pK1PU(r{nRu8scNG62n?S!!Gr&+TIlyQtud<XO)dO=oc~t+bI5$#h
ztX4vs+db!VzM<aRiSxpL+`cDVd|w&}_CG^y!v9yev#Vj3x|>Uw5)%JsD+t0&!AVdt
z*YQ@y6K%g$_(FOOx}b0$%t~%<%^LsE6nv-VgLT!zBUg5vtLiW#!c)Q~#!Zu;KH-;)
z*0zW?eI-1DKI;bv=Uvf?Qd~xt%WPc>DQmxLZOMZh9?+DU4c8`#^z!+y-H-Fi!}E>_
ziuQJtMdZhLk_)eTQeju_)qL}YWL_r~TZ)R6DWyZA$u5BNE@Sfms5`F13}R!`8z8(1
zW^%3rvw$;7V)EwRhkF?`8)az%n4u0Qqa003BDfB`ecISEbie|7U|EbT;6k<+{wR{+
zrI(~fuz$dn*nokimhJjMS|Njr%x>+|_xM%$5o58cL)p3JnAMt&7nk?P!h?YXk`6qq
z26%W)CkdYn6Q^iW@OCdGu1WDK+8JykFSC<0sM{4W3!B}El(D5Ta)lkaD4*?^hEd9m
z)atWrJ$K|E32Pb;yy<<$bxrq;566IiNW8LAG&2sLW@n7>UZA8WCTLgV{U4$qfF^@8
z^0RsC92+*0Jiy>gbx%!o|1_9nbFh7Gc~N21mL+h6kxL$r&Vf!v6v}>CC~q$#h^&Aw
zhhUC6xTz~`c8Q8K7S@GM@!DsSo!{nm*Na+$FKZayNZOfe_vVoExpg^@!)wbDvkB*G
zM2HPO{T76Bj(p-*!Vf*t;#mkM9^sj%zjYLpiG$G-%vNtfC`RNW!RP|Iu4qji?G%4)
zW&XP<yKKshG#b<kLWT}7ulfjs6LSL&jZW2&9{Bb^(P?lR*6D2&!3}t|<2ai=8#zvW
z8_^kE#8Vg0ct*io+kK_v8MNiv*AktrTE$m#t+UH!3A|r)nW=4jDYPWMhlItq3k5-=
z2(rZq=t<H%=XgW$9it`vh&;_I(OGp%eYTIg#k$Tm^TUR<PGPO97qv%Fk_a^J_pLEt
zX-Uf1s1OmUiSOa_UD#7`JuidZe0FL3>CV^v`*|N9#tl5`etaz{D{@=+GQ|@PqZ{*q
z0R<JUhts1#JRZXn!aPrFU&ILIq@X-g5~i-ss=?HeK_C7WZp}aqCqfqMVj)D`_A{mW
zBVh$v9oHY)ynB0B?`OUB*A%Qk$g<ngfI)CB!P|47`YS<wu@FI9+$AW}2Opui=2JHp
zI)Vs248c?wTN18JXigx(P%@M?3$))2d=-C2UT6AD{3R~k;4w6-uQf0)wQsg{l<ubo
z58KVXeLR;hJpME<_AmfT&PSvXeu6aRpRbSfa1|FRNtY6lurplpCAO>~;>fiicqy7i
zYyn2|lZtNfQAy=1Gfkb{nWf4vbW({aWb5C!E0@+_2I`Pq<6Gb_OiL`px-R>rM2%3a
z1-m;18hbZDk9I`ce8CO~X1b78Jx+2UeVJEODLQXBjkK*%_vy~4x4mb)=-xNAmZ;u!
zTa$Xp@%|wArTal({;UMXYXr7pKg5CbwxYjWk(o&?A<q!4&g0)IN=gr2`lfSrPV(d{
z@ouSSy6=Z<$~)iRTTbu^Fed+9x6<zxOeF()_skir2%#o<26N=cT9E&DB5A<xsRKRA
zUZ6+0IdES(?GJ<YZz!97@A@H*c*2`|qAw2&Ap_8@0q)oK`_f_yC+U~Px}w>bgK#Tt
z9iLo+Tv1VY%ug<5E7&BZUpCzEV3knIWa@i)wKLj3Ka%zVeliqa0M6T_bnvpxF`usl
z-!x<LB03v3W*d;g=ay36CJN&DLtykBB=<I6lNZBcD)q;c8wcGkKdsL;U%2M5n>$zi
z?$zhg%STiiq;S)-Qauzfa_>`Upkkf`aF}8<omLP(PvjEivl%`QZ7~GwaN-3A!i9wd
zy0up~J{8V!QWJKqi!=-SQh{a?FJ{-w#c~fy@@6URHVSzS9_GEWxmIUpU3szcQYM*V
zmn8*MF5d~5KPQ=m1X41YrPLWf?t^e$Ns~nzdAAB$F|;(+W#CDDLgW_J%IN~x_qk2J
z8xQ^)--G-Ps8$ZZvS6@jBzkh_5+(ayHnd$X<aL*c%~=e2!-Ywhd;gJ3FP6pF`*0(&
z+ijkiZjJ65y@C{B0OKBk%GWYbLspR0OA;A@`8o(qN8jx=@%#I&2-mF;eUko+im5?|
zno<hA33;qtI;bi*kBO5-yD*gFZE-oCw1t|wiHj=sDJswx*F|sn)u!)Yt=hiuzNcMK
z6y!s+@U8VkxZ`Zz{-JRL<T8ROLYenlRoBaWTY$^JOHbD3vj<&;8@$$*Q?kk7BUCxa
zsgr%PHlEq}wBVKExLsnWjHYXqS<r@XCpE8k_E+D21Xuo{#GjG)nD9;dA8?#V99D80
zMWmz(%E%d#lr2+urg*X<_>kAKCVu$$nAuSQ2vk(N_!lM&E7AdviUIspqXvK6pdVle
z(rkRj8F1Ds@p^_NgVOuZ7a`oAaDo4B_UZ4-N7PVGAp!%B+J<1@QJQ4n(F`c*H~AV}
z6Ne+<#+{%FwGl8jBkOB5q{S^kn}+8|MRk(K2>qyFDlqqkyTIb6*0efGa*Y=yxkgEo
z*}m!vzJ?~iYr)2+!gbTQFAjj?_4k5B?if)fhym02`A6`dk|!oksUR@88GQ76DFDm!
z_-)-xCQOu0AIux(L3#+C*dkCcQe$v!C*Oxx-A`{PL!b!SAnAe!<g2&+b4~vTHO*&G
z-~LZfi30~w5)8cp5v)H5kfZ7Y0F|u~DeVD3&o<$1wB>v)b7Wo~&{X3S=RF2u>;Dpk
zQpXRH5Rc0c&vcR&74-Ze4^5JPe{lwb2LoYsGk#BbqlAaZ7k$5#XC)4?TL7Q4@jAid
z|GpKzm&x<26eUSWE(Du!IkY#j5BeFD{}3_!5!{p4Lt{INy2V$im%(o{n@d0+{1y!Y
z+t@zFm}%u8Z1Hi$O7_IutkMc!Aebiu^G_fG%s;T;2g~9lc8?c|n<1qnp?W;XNdt9f
z0%^-1f<*dU0`FCZepn$u1)GW&puK8YQldKmpj-~F*;U9J|J95q(uY!f#+W2h#6t0^
z8dkd4u)~r#W<C+2{gZyq|Gqg`N*sW)GC)mlmr&*4(~t3(f;c`hsLL&~KhkKV>@`{v
zOA0(*ue*HZK^f>gWVb3huSk|(dxXY{SoAYgp<wNoEny3z*uxfuhao*EJ1jM(3zFo3
z>2Zdzb#Yu~2|EJjPxYTKl<R_tYT6b0%#xbR&GE}(gy%X&ONy21mhjizq`JwD9a<W(
z|7oPjwVw~~uG<bpT$mD~nT~0yi|0tL3D>|4l+88|c=mC~VFYmzk-D42o4)*r$w_4l
zTZ+^nyEB*#uD~Gp`NX*pt1(>}7m-;Knj2h~_c4jr-FYsgl@zivb$GqjBlnfcPRYyX
z^nWqwIIiJIO1?l4ElQZt#buO{XNoki08UMCynL(yIvbl^7HbibrSx%=nq^<U)_bd_
zy=F?>TE_sH<(zO5)Te`y|7f9P{@iBK+$cPRd<jnARt{+PCHbK2pP;o{cno}Y4g?Qn
zRxm}`Vq8N)B2#h>ij^~%uIa4>oi&5${I+s-mPJLoN&SBCpDA#PZXIvimZoJfj|M&W
zKRzvBrzxwFLObDW<TN3NVgq<K4Q1INt_*A${T83NQ+(6V^sduEru@)Z2wP{bO1@W4
zcVz7F9}wJAT_k>csA4LSu^UJew*U#At%P}{=Qd^mW%M?PbN+W;nUsXuGqF{~!546@
z&}3%>F(k_fb%~hULtiA0M_p>>wTf0^>50zPJN3`qzK>a1W*0QLYYl4KSGDbMKihUt
zUVG8<1&#or05_~m1ICaKjjMO8gxw0fK;Rl_z~i!DcXyFm+PLWvxDidChXkjJe0Vmj
zVl^z8lTvW&oO9YB?Ept>$xzw&<TCz(;Gcr_*J~Y*T=Qd3((cnv(wGS0dDMvJUO5F7
z5K@WP0`z%-TdVIPa5wx@LG~2SH7jJ&HuG+We4OU?LAo@jWp(v@6OQ+Fdt>rzN+@mD
z1FXKBzKkE9y`(A?%PiAx4$eNm=bTz>e9?m=eKDVxgwM6>>pO8^p%gSAbi+)bQ4T<g
zZo-DF6vq|I(e$yoyyi+H<qRGf)QDUw%oJ*h(VtuRrLFfMRkL(Kre3vb;LR8IFRBR_
zAoTe5ib#*!!oho@ENxJlL)a<`@FGB<t`RX<f7>NqQ+x;4=(Fzd?@V*u%=A&;{q!*}
ztn8rv=4}oN^6SE<wI7Kn+-?YX?@^Lya}ank@o@QkI%`B4jtQhr1AK-VJp4TUHSF<j
z?N+!wC6pGXNyAx-X<kv>opvi421hiTi(_{k7@3^AW@W_7<ml~fjH)TV$SoIwc-3zb
zX@$B<qF>QUGIqjZX0N`YXeD!ijhQjChI`UQU_jEbk<W!SBi`PeFMxbXVt;#aNIl88
z9H0sPn(#K1w};)@-AO5-w3?RdFU#eU<!CEn%B>xW<C3=6uW@;BU-#*?yAjbA+GOo?
z5SzR=5^tDJL0K-yhq>fjQ?LWGz|h;>0;Ksh<AtP4g>Zn{<#Z)}c<7SyRLr7GUpqJ}
z*w*vzg?Rr5HZIZf`teG8`<fISHX2u6AmCwIOisTxra3t&yC@ZzS*Woj$4}GiB5yEf
z>(3P#2}a!gmXn~O^FP1)BrF7);NhZdBbA`o?lwt5_1kMH+AjuiKUot-&P}d#a})$)
zV-F9z!0%;OD{g&59XhGC_jSvF&Qu3~i)|N;EMF{mt;7u7uu(($eIG%5LhQ@rWU_M@
zqS>z8q0nL_98eFnXAxkmD!b~HV&lKJuCqAW_FzYKs{J*!s9!coU&}fb55zjuZ&yYf
zfnak=<?qU7c~A)AoTd$yMdjL~R*}vdLm&dUUa#{uRd#k?Y)D(X(oVEr+T`aoUIGA*
z=J*;d5f>syB_k)NL!ht=TP-wk8V}6o6ls=4$FIoBu1{Ba>Ct#H=A^vRzQ_a*yRU@n
z1UiC#d;5{jIS8gUDxN*0fO5BNgo$?3$s&1gJa`mF7v7GC{`&wDK5Zi;ONHeO%(4c|
z%eN9fgIYo1og|9nzi=u3A}c;p5<iK*jTvIJz>8|j9}4V_v5ViH&@3V+wR;7b#vZC6
zEw~VV>*S}Pzyph*)YZj?<dXkAu#XSp=QQb3WqOgBhMFqZxkWpK#`X3n$!4u-a?j}A
z-X^D4KOc_n=vcbaRL$|qn<>=qZt-6jy6@}<X&Cwl{EYEBL>{aJrmi=@6mAvS73VY*
ztI~G!=3E#J(LTwsadiz9FbjMOEmtY%e&+7EI1VeV1E`6;_$iZfiSZZ}%~9;S-aJSY
zZuYXUYcg8g;YHoV%WBT2r)%9RaJE~R>+QB~_4LTwi6JJk?-csHrxO7DyBLgwAM|G%
z0X6xYwQoQ+XWj<otIq6Ucx8SiOjqiWyu!D>f+!b$=Qa6v|6>{h{Ey>&WS|QCk1dTx
zOR>@N3#8z?4I|67qrK%c!T(SQmS@g-@{U#>qz2V^43^#pdHhe{C8czL>yJf1mFpTf
zR`C<pT7shao3R)Bwl;neA!*Laf_Rqp@UDFpjRUrUAIyu^ebTHt;b1s3bMqzt23b#F
zr>@@$hX|fACq;!^dp3+{f%$XyKd#j&Lbhj^@J2G%9PZ7zcDPMbPst&0&ZMrW8l?m|
zi*IklU)<II%!JZ6fPbEgR;FVb)QUs6K3=O|6qw!hT%?!eb-;bNU+0OxP}9`Je7)$G
zG0%q?vh5c11}T03fqoZd86li>GzSJ)BBjikwAI2jUQwoRvY5dan&Q&Epc=C?1b8dz
zQqjyoyJb-~8r|@{i4VFViyQKtAM#Ih(ZpOa2RfSq?0O4k3wIxa1pq&}{%b}TxuyRQ
z&ohL+yu+zOB{O1M!~Al}j5O7T!VN!b9~~>Rghu>2(_gpGn0k_yK(0;x?p~6rshH-F
zXnqP9q>tijDCbvpHfeA+FNj4~tuJ08TFsjiI~p)V)y;dPl{VCb*AYHHW3LST@T$wX
ziz$h#)<ehSe;fTlCEy=WqTnGY?>Hj5%5FP@zhwGCKsPgi^eNgG2}ZDu1^BMD0v(8D
z_+1g=O#?mBqIZ7;Grz|)fAfvh#KsnZHBS->z*;={au@B5vw`wQC`{8uPezRCFj`&l
z=i22A@uuvT@IuoP8^pT^OaG`&qK}~3!GiekXAF#)y<?fSSl=4PF&+R0*tOstGoZ}0
z37-yRGjIXx^DT+|wh5vr$f2@skUw8V!yr&`M}|KQOX0sZOoY)fY@T(**f!I_T1m~P
z3>~+Ua0K0*h(_l8RyO+M8^AZviLk~H?6F`u2F|O63^bhV`4?UwsrPI$@*e1};q7)w
zhQ(Nle*#ify-N1yjaV~Y!Il`|PJ!=+6_QwuB-jI&&`VPHdCV3^UPg!e@4q1qZv4OR
zD?^dY=DEs#8v%z9fk@8aOX`2B9RBW1`k>B81XaVtsQ7cFf+Ai6dpe$NN$ji?EdWEJ
z@v|7Xf_{NiShB6YFH({3Yg8omg$W_31N6Wlz)PzRytLxi|0)8=*MIFt9)mE<Pog&q
z$eC=>T#_IxR1k*6@wZeO&E(IqlY3LBtivSDGkSJ%SfgXTy3tPxUyjSgypMnpAyn{x
z70DS)v0{42SHi9234*<F)OjdaIYy*x4v+yZY2daim7i=TPCF}14^~-OjNOdRuk|1k
zk1%8pw>mRUyuH)meXQI3>Fkdej#?6{5WX@pcpVJne6FdO?>Pbd=&fl9?J6z+8FKdW
zFJoj<25piLixhd?1DSVQMfSYnG~Syv?0)6S&_@xM+y^|!S(y*A1M6BGLtVP7)zxDO
z!Y%lsr{XZEe-zQIM1u%$Ws>=mD+gfx+klU*8w8?a!d9$8?9p*;%G|F60}Bb)9Smc&
z04>}^ZSfZ@5(gko)Yd2#WgdMAFuy^jgf<UXtJCp0(zZ~4`~%g^?UI$vOud+#bA}7v
zd$JNpT$PCogKMev@e9qDv*c79ow$wJX_`Buw+de@6b~yr>c{I+lY(FX>aaZqi<#n%
zSK@1-^<8_O9qlZZJeeILQsh$Q+c(Y1`g}XD`K-ahy<UkgXMd>Fs$X(k+z8BU;CfyD
zcGcZo!jrTawFn><tGb4sA##0+u{r*{#7>g}PP^vbhIMgmZ5vl@-JYD+`oq{G^2cYD
zkRm!V5wYRL@Vb{!9@Xj@30pxOi0ib7V9dssVQkj8AdIhS#XG4_4cbZmDYC}mis2dN
zjex=?^^e`;9Rx{h&9^HdiZ-1R!ab|KfTfSMq7&0VWg=iqN@#OA`np~+3lxmh%p>#)
zC|21OYvSh^6I_mdT>?Gu`A7JaFYtbnVa*5_)w^&VU73z#Y+3Wb@-A{~6t||FO7==G
z!x!?C{b+jB#QqAlP~QfX@33uqSEU|(vG95HyLn@M-lvSo<$$|^6|&$i{RLSztxb3w
zeQ_7BrwBIDA(HEpdjyJRmr8P<lN6`|$jUJ--YVj9hT`i?m!Y@Kf^*_T+5rF*?GEeI
z4uy=VPP&rIxLvGUG0XnEk8Z>1O|E$|=XWfceEQ6bei;A$SNEib8bRhAB84(|?v-Ue
zz!yvvZUjDk$X^-Fqb<zM<S)C{34qvOi_7VcXjG*J>QDJU&WWhXjcnh$fRw-c!W&{Z
zTITeXkYgpDNev*6ma`@BSrRDM6HK_$CD5<rc*z5Ss?yzNT?_ws9n9h(!~ep;i@dI*
zb=0$7_WH5xBcWG^1c@t7u3z8kb6>iLjeWQCqI#6YLLsN_4{@%l&?q86Wh37)4CT6^
zC2*<13aq{)4i(N9TeK7=QAQZ$QDXqx!JT`gN5IoP0LQiky9Flm0oKx46ULCCE#|W4
z_1mrFxnk=+m6rK!ah@FaE-ul~ZqL4tpA61-soW?fpvIKH&;ld#@u|I_16Jce*}Dg6
zOXpS&<nm>>NT~!RXkVeP7d)a_ZNJ6l;(gY#V8^+ymMh+PRdC#SDlI(z)_!TBTLK9J
zNtdCPA^3xA&w&qof5B<~!DOFAFV8NbKecdvN?)vdEo$%Rqb6@CDDc;gy+@d`X#wDu
z&VMhYhQiKfky5pA5hl&GDr2i0f>LMF86OTKsADbP<5pHh^NXhiJa&+-)`^a3r&MSU
zPCVKEzS3YGB$-}&?#r>!GPlg!d{|5S(cra+gNM{3m#QbLEmDto>AQT!c9mq&IEuXV
zMQU9+Rxn#axJ0xUP`Oc^2b1}goni4wT>y5%W%$>{NqO0ywbBjExPt~(d43Bj87r2%
zx2V<_?1#YyHs9$ybP|)dhuM-0DCe5NQsEBzD?3tSwwz}vGQ~)h?WJeZIAY6eRTP}d
zQTC~;@ywL@7*pV}o^Z_ci2O719801g5f0Mkk6XB0dtf<ds5J@*VBr?-E5n!Ylfw%2
zP9^p%gIlPH%Gp`lw3(AZL%cZ1xw)aw08qNA->Y5eF}Uh}Lb#eMiWT6i2I7QVVoWod
zi&d4yudSmmqUjaGn+^r<H2K*2x%p72#r8#?ubcj2f4XwRSOG+DETB*rMkKT%DySP%
zI7c)MY+(SXRE-IZCaQzS=i#k(;Dh+l=_L&K5ElzVkG@4P&4MTDVIQp>dEC7_53ef}
z>crDa(mrcz`Zw2OzDN)Q3nL3&f+CXRCE}84k-&aCLLLB%Ky<+yq0G!2bl}Nw$iYT3
zBAvRkz?31cqdlcBKRP~liQr|;fyk}V-aoyzsy5C$a>WM{?vpg6UlbdmFh>AX-|@Ik
zJ(wc^aN$eT2<Cfle!4iUSQR{@@<GDs{Zja4B-2!+EGJYGIg1$pHn(Y_>(MM8Hx{dk
z^mvRb+4D8GxC&fts!`MI26aDKIQk9h4m-O)EVNkWyRr29v)_kg5C;;ul(+r@8*GbE
z3)4K@P`HU3ami>tMXp!B;?e#-y{4)&x;1egk5`NZBp-X$L$v55jOYJR96`un8i0P$
z6mp&@?;v|BZ7DCP99`g*$U4ZFMl(yrqRbej%NA=5zRtY2aE;R=TH~<~m5Q&QBHqo2
z29aC&oiX5Px30yf4RJ^qAj8c-RVYJta95hh77O-FIZT^Rm7#5Udv<B9=&W_eI40v0
z>GQDtrztAiUr4xQFCm>eD6rOF)+>n+b`DKY#P={n@^Sc#AaOE6CAFn<JO^@NMhb;w
zxPglXk2f~O0h(mZyD*cw^()^$waV>yRS>#}sCrv6ShyXU>Hjuz`^`B6+!NYD#^sd6
zQK%ESC5l6>s7AXeW$kONvD-X$sok1}VWFz4z}%qvMER-{e(ba7KPR3jas1ViSWEfN
z-H^tPf<UIyEAd%k|E~mhW;pVMeKEUwUm1NKb}Y@>IsZgP>HuC_5bnMHlhUVk$2_N;
zo<g+PK@b6uY`_YiY7hRx0N_!0`;oUzUH{^;A@YX%IJHTb7fT@Rczf4^pDl!CS|6X2
zwQKB3{f)_)jPKS0&iwz5={D39KtaRLSu(@K-=4m(8V=P!97{W#F0#RH-garFvYs0*
z<rAx@B(p@(zH3=&d&kO|G3u?UojF!H@58E}8(I=s<iF7VAGF)N@fkP8_YpS$s!=;f
z;b^I_Z!-!JZbWFue5zdEc*WIR)jb(Gt~cCwHYL4Fzo9&N!Ih%|(74eT{#oetFJ0Ud
zM*n{kTPb;<J{xxW`6wq{IG?rwv{dK{_<v7rU-a{4Tn24fhMfTtO?rf}3bcE4!WK>j
zpAe>0v@&5z1sr!`X9DSU{zPo@cMUp!l*2y$M>9kIdpGDmssT!iH=vv<z-X8@SJVi#
zK@s=fff`0xk2LlJ$VrDjUWQtzSQ2^DehUcafEh?dB;>8WK)t+zwgt4}fouZ$$VU=;
z5iis0DWBcgS!D4w5`-P&7yBV%oK?ICdTIsW{T%Z#|4J}lmLPkr!Dkzk>a-<p{<0_`
zbRD-$#&*FDuM09Jn}L6qkFvaAyc9c^4NM{w$>@}&1Umd+ycT8MV*)xhfGId=qy@{Z
zhtgUvKv=33agr6(Z#%>{0omD49{**NF}XsHfL<U=0`dqXhb<1FL-br<0#VwWi9nT3
zKv}OPPiv+2b1lUB&~PXXT>Hr)`7QWfTj5-^<pR$WkAUdCEz;L8uB%&gl^KQz%!*af
zK5{B7(*zUZrc7n%o^0zx#D;cnh)bdP=K;3(D~M0jV(~91KwKf>Z=gUeIsN``+G@gW
zl@zqcVEg>bm>=e@Lj-okP_FFV-ynq_Ix8iF66KUO3HTGj^{)isd|hp^ZVeE8?=mJ&
zha=dHejt-`6vbC5WT3>A;66#e7cr4{2Kclv2tG0B85r=xz^UFf%aYg*X3GHKoi>%Q
z+&+`QVqv%IkIa$6YyheWPSlih@M$Zk02rm20P^EL0A>p(j5I9v|GD(o^j{PV6PGuR
zfvt)W&5VJA^(*!o<v3nBOAS*%r`K~ymtgZ`tDQX(QI~}WvTq&l3;j7Qk9K>e(cvR!
zny-kSk^PwbNx9)@4^M|7QbUCf1o}smCClVoj9&~-+ajW1jeNaBSi1v%w{{3lcvGmT
zaLeru6gzTmj2jiKd)(}{?Bkb7clONO{kCQUP{Q5=wf!9aG8-y{XNRz*H3L!i=?$^T
zzK+;6*JE7_BDNB9?+^<0&B*1Rk*!PL=53su5qf?nt);>WJdx-cEG&hE3_IrM;&ZEv
zm8{vK1yTtp>#R-z6I;4FfVXaxQ;{Y-RtZVZx~fjwk2$)b)amA$a^}L#wmN8K-Nm2m
zKYz|+|H^ZM7Q)VHRFHyNjx&O7hx_%C@R<^U0PWQkXwFc8{t7zUT43I^7{--Q<h|-_
zmm6xn&$&$w8m<c4mZm^iJ9^@j)D`3;B<WCj9jOfN;7+GlJSa8mghQQf*xG$R`ul4<
zE(+|aAgy1ur0v&JzkUI*@EE`jSv~PGJS*SAS)69^wQQS8Ud$y!U2aT)%F6wUyApbu
ztoI!W=yLkC_2%ZTu(!}W-3&pKn;kfFi5nZ4yhorg=0WRY_#1jXGi$Vx)H<bk8#==V
z?8?P7uR5*iX2oyDtzDj|hMF*N>l-;;=mo-9zAutS#g?Uul)0Uf<mEsBuUU{%k2=dC
z&!QPKnc0R+*&1$5uGukhoQJ?ZzuW7JVd<B%T^+tgpTQ(9Ed+gRdJ*DIt}2Xu(2#yZ
z(-NM8*oyVwAi&EI!^4~hOy^ss6;Zhrm-w%5c^T?W{#9kM%HG~7&!dVs6+82~+m-<F
z=${UcC^Bi}Fvi!m+rVLcfC8+l3(rEV^eGK0*{3Dp@T_bvb^OZcK?vdUoZ!-3ZY`V*
zFgouPlzSvgY7C;y$)0edV+%o#PF^DpEwmD@aAsuY+~$26Dx<IBp}{d_%_@tOFP-b>
zh8Cm*=B_Ipnt$z7mL6|#6gZRu-VooCACm_+xC-atEl4JN(9E0r!qCQ*x36<DBTSw5
z&We0I>X5mOX8#tanDtbRi~_j=a%uRP&%US|n?K*m5ACV(po%o3eGO~6EJP}YIqVSj
zR3i=1>QI4sqs|NZVj99n{kj@#%75DS7@h5;*Q(#1b#x(2MxT_l`pd~maNamxpUp;)
z?)owMe!?#kslGu_(vv{70u}^9U;u-FeN4I#MxX=S{cp*<rNkh>EP3!txeOA@TK1LT
zNL?TzH5N~@6In3I*)!GncU;%Ca_Z$9RjSkI&yOKqA9ATRCr_0VB;~_Di&v-UFatB-
zw4x;{2={u;z$Nk^qt%OVDq2e6#ro7}PNr80c8f$EEe5{pAyBrdM25YTEpT~642>^K
zn4ApAnjql)aH3FxNQqx+g$>&C^+}3ey@m?0x;a7O8-hmyzy84!Zq-i-ZRhNB&UZNx
zza0KIv#}&q!k?LQXetnX7AuLi@nk(WS6D3YQY`Nvg^xwOQlNHHNPWsH+UIn6<+|3c
z-YH_Ou7C=~TwtDgNm@X0QEW~W6xUhW*&S&ITXD(UU~*ZVd9IP8foB)eWejGva~irJ
z9OXXNiHGgJW?GN8S`yP+VNctOXNs?DflQt-Pj7MD#|P@tQCZ65Plnsb!R1*O2B0WB
zi$-4cDEd8Vv4>)9YwxkIgy^4aZ?3N*SbQVG(l19MFg>J<!8=^Ndla93?GDX512apa
zw|>d3@FoY2UQiTdS6*7w@T|_{+N#7ndJ{kAYTa0Y)e_6ql`o~@;N{H$Y8N355kd(B
zA32sp0PNz^8{uBEGGmXRyd8|$qE&g~+auH<`4{|El;CLP8!$RngwD=*uq}C2tl8(b
z-3n9IMInXk_#ZW-z5jgWh?6^+E$4>t8Se`H1;xCG%U$A_lORd=Em+<W)B6r~mDTlU
zy1-UUo7q^`GxxK+t}pBTmq^c3ej3uzf2Acdc#J>>WNp2f^RN;Qy>Xh*ZiSc2>-wF0
zxM#R0c*pRD$WE)3Y!}6&=Qp828@i>Ph_2u4*|*moP#J$xg`zKDOp;JmEe_GFUm#x2
zo`X8<LImy&G@>V8MyOuJ*TGd4Yu}<L+QoN2<&gUgS)T=!ALmJ9?Fpp2D42Utq^~Gh
zqhO|6r}S?_XSIpcu#0!FI@{?RieHc4ao%eXw~c`xR`Rq*A;9iNJJ+Ry4AsW>u`w$x
z!ZtmT%0>=5?8hqJX>yCf{@RaN<{XA>XCtHBd`<6ZFFq+CvB6q<%q3^n!_X))|8fqr
z6f8oqHD4+^bUpcz^k(G)v0mA6WxY-`^}^Oig=qG{x6GSg=1qQkIhXHNbAdn=ZPMyu
zj2p_S!_JfO7HA=JfWZ5>mxVwrGY2Z>%o7XEEA$wd`&Ww;^Z5JimfCDD+@e=HqOvf`
zD!e%&GwI8rn@!}g6J+Z8Y628R|D;F1-B9GuuSfoqo+JGa6gY8^Et=(s%j712bW#kd
z{6tnO!g8$%pz?xB^2x#X3i6sv?(6Pv@;H^0mVf^%VgE7p>h|~xrvTNvX;?~h8~&`5
z@;M}qEYzmVK!+c?fGn)kCkUJIX*Y$(57VD9b0S#jt?8D$UyIjgVKyo6`c7jf=4Uu8
zyQ`iR$*!`AB*kPE5Aa8NfdrxE&&Gx{Neqn$JSq`?!_G{vpk8c(N=STiLJi-73masp
zsPF{Wn0+sHygjZpH0KhUI`dO|W)T6MHld>@11W^3X!d-`%m7eHEt7eqiOEa(_ocBi
z5W}0W^pxV2;$o&7`3f`Jjv?IP+%Yu^OG;^TIpuYNI>RStm(C{Vg}KqqMvg<%mKUtI
zD*|trMft|pLJyI7CV<X}e*h*ST^!jB#+iB<?4kUl6WGO0iZWWViJA<msbaQW^ah3c
zfid@^Eq-HTMf}gc^rtY>1>R*cRKWcPdl0Dd&WB3DxeT85m(E0y9q&<Ir<40&_O+z2
z<vMOhdfnV#qD?n1fAMq-|CMh{zRdX3!9yyF1I!zgjnepZKHsK`SgJ+lgbkIvpaVDl
zHvZk#u-93-+ZyT*HGQZpzf-aL!@O8+9j3v>iYe2ehzq#Q<zNn;ka@d(%<<`A!uc+g
zCmq+jdW`0`Q!v;pI~^6Zr;qX4u-PRsnYKmAEl<|CUn-F9{!no~f%Kk%FN)^Zh*hRU
zf$EL`ghc`=3?7$&@LVXNC=6|o>2e&jLWMhfJJx1l23#xt$_}NVQq(@`;DP+y9HIWZ
zA)`IZeA+TZ8zG#m6chdle(#&^!asneF(k~O8{Z?{DXA$<VGe&e4>K*ajLBJS9#v7|
z?av26%78&!Bhx`#9aa#=i1>6V2)nijt)SqC#X!_meA}4J4gIfj9GYGunm>1<f_2lV
zSQ8=zm_r8yI=~BrvObVMM-IUQ09jj!hr@3C^hX-%_a*(`Z+}0#5_t|_Mo9U>L`W}5
zYZy#Tj(y}LJ=Qc4a^N^@=5snTr@Q{ySt&6N(Bl@+FA3;NqhfG6Th9Cc7SxP|SrUi$
zLG1E)qD@X(LSn-w!Ospt6qPspM-?*Y_PC_~Z5pGEjPI@&z+HIqg%4_WUzoeEkvkL<
zAyV~&fuoBM=u})ZZh^;-yd;R>2mcXp74aLwHg@4#QK&suFjXKt1NyUn-6H9K3|p1L
z=P`1j>Ki{a2p%c)4N`#2B!EB*_^*?i1M@z9E3O^^#2-_rhJz(&U^z*P=l*?o`Nzgg
z#K3>2IDQXRrA2*k7^4wmtI!M>27@ky3e3?4DUlXTMQw(y2QTkME4>AZGlfv1eNf&`
zK<~Ou<O2Sms-SQFD`J!$<HPyHqO3AzCSY$y1wo;LGgtxiE;(K?+^2m>2buRMg==tY
znbwbo`ZKa}cyo&m#?@S2E>F()cD$NEzQy1xs}H~}*h=hT;8QSfNo|Wb2>HlRe7cwi
zL)crm+94g<m==ekj4i3+YG2NZkn6Yfdg(eXxhTI$r}hC$>t@D*J@T8+8R4G8&xFs&
zX$?|i;7m<E3@ghNDZ*EC(FW`oU%kuJk7nQHrrF5|9ibAP1o1OsD>PFK235^Mt>V2i
zX{mukzXMyR1$NBvnwLN-rD9>T1#*zamgd)dB}mhtYFyDH-A!(d!KWYQL86F?9fm1y
z(1pT_2%*oS*);2>sdp{R4aGIh9X+_)hqVJDi&Ih$nJHXREz)bg`tG6i%#t7bXs@nJ
zA&45lZfK)aWLz7SMtQ_C1X#%sW0Vuk<hfc6`U_{#JJ8JJK~p}lOR<D96IvS-!xH(+
zaua*j#lE|cJ@W8yyUmT0@gEOfbZB4m>+__ar%Jj2Ml+376otEKs~d40yoRh?EbpWU
zRf*;@G_ZIT&;7bLZDH!5cD7ih-y-K}{-Kx8kA|2vzMr~fE^$E!+`<hE%x_cxKGL00
zo*oQUmt>GCCknjZE|v68e-ZhG-MWS2c^N&y2xrg1fAGrO^SDuD#!IYFX83#DGT(7{
z_U<BPUHe|=fu^4vW^Oq4ctd-Ch=d><??E|N5#E`Yf$Lz;#A1K?OuULAUpy0tE|5?W
z!6Wg3tIC4B+Qd3{G%KB6uh<gVD_i;#uPwsvMgpn+uTKx_2XPny8?w`AHhkSH(SmX!
z8f-6lu9sfv1-()?RaRzi{Ho~JwkGC-8%NE54m+-(y*Q|SlKV&{pFoFn+aL&f6h8GP
zNIpKj`L5_-BR*YZ%p?09#mw*bD!XTsuOu{_@NV&lJLVzXA`aX-{aW%%+;Ky1{5fAb
z6D?C@<}$>J%!^D+x|1xSA&a3vmj<>dVsJ6M=pAeiM>~|FRIDa4<lhVBOIIk`$<?o<
zk*N>2oOknaJsRu(?nqp(%37)562c8PJmG^e0&I1k07Pawpa_G2r<)0Pd!d3O*nv)q
z8Q8uGl_k8PU_&$4RIfV*56<35qv=v7SHD}cG;Za`!;##rhYU_9p94yEJPhfSVYk)b
zq3maDk(5MLlOf7pR3t8x)c-|r5uTA}{yy0K|F>WKO4x*A>O%u#ozxanfwuTprX!-{
z66>Db$+$$8%StD@#y$#va_^>(kNw7Xkru}@KRR5z)Hi3l=kWy}eT_;1ThNBftOnIs
zgEt|(J=$$Gofd<pJZ5KeNsWvbQlYOMFx<9~T$YnxmY*fx6TP?Xh~bd)3gd$&KijwF
zNk7;CbG@v-KSC(jdkfElo9~T*F7q{lgX|gxU-`SEcBfaq2f0f^csIyxJJAvd2m`Z(
zGkJ$FJv>Mp4A=-y5xVfW>pfeXj+}ZwE20yqP6ye4sfl1Ho($YoA^`{9odc8>f9@#D
z!k;NytctCr@EBz`L=N8fy^6bbM3;Ea$uPI4PH$Q9Tzys2lS+5}uCQn1=eh*?eUzn)
zxFbPK-tL;VDsC<)oGqpKKzBo1%o$WH*LQ;%oj0`-8<S&fo`kfvbG#qFnG}|&wT%0#
zy}Ox#G4d4stEGV|xm=4JWsfgHAY;bUD`ii>6aGS-{?C5%4Up;7SBf;q;WW5><YJW}
z34r#qs7IO~SB~<WlLr%pd%PHBNEX8%*QG726)oqFS~uXUfwjLeb)>RUr(LFV+xw1H
zb*Z|Ffer!bX)Ck67UetnUU?P1?*njFfOt<H8tevm%Mgc(!zell8z`~giiKk9?FGj{
zcj${&@+?HN6=7QHP_UgLOy*N^%AhUecZZ5*pGw-o-G$ozAKJb>9Ljd>dn8HHq(rt;
zC?%3z3K=Cy5^W^JRFYkq3Q;z<^w@_eJSj#=lB8_29eb4|`y$3rA!J63J2P{y?;LAA
z@B6%~*84u+yT0$A9LIFbJ=c9*=XGv>=kIJkbLVIx+bEJNGH9cvTfQOteDnHg{~L?N
z_(tLRzc6=92-O6hb1MvFkZMyRSxqvqgR-Ptdr(aXHL7!1!Y~P)@FI(ux^K=|)gAie
zY0lHn2@|VMPCdNbQHOhp`*b%F$GHV)Tt7?%yyM_m3_zOu2p@(41yMt~P8Bhb_wBYt
zuQHWu8+1KpuYJ>c@nqtOs@G+&+2Bv4O?#bPPcPldDKpvratR+B<Vo~?U^Q^V25tw}
zf*eNW&BxUS7CjnAydA*qtMY59`ETPGcJ#?ueNkeKy?b40VxiLY%=R;8BZYSec~yzX
zNvpL($Zbw-u&!fD6`^T<-uoYYj8FUpfLLId;fTF%tj*=_*SZ2C;6sFeUPeHDM~s)h
zdg!N&8-r$|FGSZCRBIEf4{c=|aa_^DH2?Fc`pNqo6EtCvwtDoab=Q^e4z};d=IXod
zc?h(dXf!xSZtqmwupsep%m)72e@CI=hhgNT73kyd^DvB;+>Bi68Df;f21a*Oxjo>s
zgwY0bQcYZ6eTH^yq{r#9{Nf`Os#{XUu0N@Jeu(B11Fe=ocX5$8kU(JUBKD|ZOIo0*
z1*wBssG=Cbchf%9W{|5?-BX`=-D76^(_FQ}fc5qc!cj)2nD?t?M8z(w|I)u_+e=G)
zC;3;lmVZ1K5~}ouY|v%5#thWZ3bt#Xo&VWZqPw|96P+479w-u)U(xDK612y!q}h)t
z*=sbg@_W>7;)1dx0$JAIf&ys2I&A+|wiQyn6O#lr^$rm^j0v@{<$XzY6hUmGt@|Z{
zB%z)pu2@*gXmW87IP;?8thlN~QyJU4=CuIKZ@rfC0Vr+0`~1>w{7_<jh72r;=-%?)
zh~<bTu@7?gpj1#*fq2w0b7z*@95HM{`_YX1%gh&MPG%;h32DnDtMwYx$AHN6_LV<6
z!2l9aOC9)~HWeHDeio3GLCAFLDqKLPA>mE#L^Ei_UD!*&Mo(;Ks(~4dV)zZVMCg+k
zwFj@6cg|4o_AXfu9L#o<9^cfX`dYFk<otuZo;NPsaN++H`|l|;emcIFfBr}W-Q5kA
z!;y&F+xf3-C7!O(n59#rRRT2x%PYDHNvrz^E8q8xH11O0llaV5=jOJ%`L_eVNlfDW
zA)&S00qaMm9k4}k|M1}m%EckOs0^|NXL&3u3hwuQ|8wXJbJLJT9XYIDyRC8~OjVt|
z=H+K0X!rGepAE@XPX3N{>L7!*cjQF8&~BHNU;3AwH(&_PL>Uq&G{zw+y1}^?z+lAj
z_5g4dZ5hQ8{v^6IK>~HGZgZ6ri*Hkfvh!hY)ttLuyRY7xTI(C}yW=HVS<Jp5MDN!M
zY<QNlN(oz*$c=?KXY#!cD~g+l5jfIlF`XvCv1c4fm2TBe2^=#HSl=+3>v3*(ah3<+
zZtK~rp$f_uO<>+Hx)BO+$O-(99W<b2Y-*{s;F?a)fl~0CFWdf6o&5_vrhNa`7~lVH
zvi*Mz#=-&SN#)fZ-5}1r!i>dYBZG*#BYy;ATnK-8a2iqAgci0GM-Vg_)NrQZ^{PPZ
zJE24RCFl^~+iAhl#UToI>EPok!!>(2KX5aI!rC`qG|e^#vXk^b>U;lq_y71q3{ry-
zHh0`)n)w-r{JjeBkFImxpY%s_JWu`-3DAP(N-RkI<q0$H0A$x=S&%_IaNK!;2r&>~
z{vlIgB~<Ko0o2&yfBkFtW2i99?`Rol<R-+1k>-APndY%DLw+G{Wi^Slk=M$)pLK+<
zYaYSRt1-_HzM~PQk|67Zd=NynVGqn6<FMlp^ePYsUJ-t1we+fCO^sIpd@Wd>hM#~U
zKu9hxDBnvSH-SK!)JT~NhX-|8{_CE1Mn1M|A0#NY1$Ofa-L8~5?_zkhp8tG{|3#&h
z*IE0oglE1Ud>#(568XH`K%SSRV8B(i(7DBj9qO1b#5iD8gegC2UJ^G6(%5r1M`5dm
zz5XOXgugHW_Fpc5_&WvNK7<M7vk!y7m{Of5KNMFXd5s}j$SO37POB%2`KH%(^)ys?
zq}97Wa0`?uZdW}M>UYaKbx&DzT>8NT((iG1w+J%f#Fl5lBTgbC(I^3MAX{(FlWfhB
zFp-ZS8WQk-yND<7=!VoNwqal!5wEXF&kgNDmINxI4gqHSm;s@MmrAMO1x<b)8lq0*
z6?+{MkDQot7QGt3Y^NoDGVzau(?&OIR?pEC<gtUgnfcb&U$N0n9}PR3s@au!k3L7%
zOD|8)T0Ln<Xwl)-en#=AqH{>O(NomX8Z8@g@g^^G*L`^Hu`!{x&tZ4e!8SGL_DM%m
z@m+cHFn0U#06l0a#4HWqXJ%6eVv%Wp5rQW@c5EJNhMlA^ERm?nCBVON7HcO;R?5(h
zyEzD;aZJa$z>SOwZfE|7kM^!o*Pj;_wA;SAz2?#qPidK$A8B|n+pJ=i1)MZO9e6+~
zN`jj!1X*7`ObWFoq0U1VMaYS-$+6L16UyI)#x*ZX$T?kS6vTPly!ztq%!DBUM7T~F
zVH_jPDFlK7sEvLADl-y#Fjo;#6*aCX6%t|ehBn@`0%&rgaKt6Ba2i|&#t_V*Iu%6p
zg4eNSQNXHqj?h-?Yl3HZ$Ok=~9oUwf8ugU3DXjW+zrMNk0Y*jcd+p0)#kRU#)%QxI
zMb|AoP1wQ<P(jVqS<x6K)HxLc@TF?K=UFH@aLU1Df$s<q#tjTjUZXAJ!CldHUFkkd
zv%%bIM5>Q-klBB8kiF4g*x8XTYjPmgxTw5{+EMr1{y;aQ{7nBAnGO91wx!{R%doYF
z;b~BAY?WpZF&Pgf9%ZQDh#F@l0TX(~wgeGcOavV%f>aw32ky(QYBP`xe1i!xS|as@
za?v}rQWHQ@?2ar)Sv|spWp>8*U73xZD%LvUo=@^OdHl9P^~QU^jSOM(PvLhU?b3RZ
z*@8*uKw2=yQZOG98;lFk_$J7@2g7EESqN&%G^(GY)=?r;<-=+)i2ch&9=o!Vk}K*~
zer#Mio4-A#Z|TuCih2XzanCqILK$c(KEvh5Kbp9DT<}(xmv&kA;?~u8h0v{WFJ(@L
z9J&By1)g9=!d{I~P{puo`iBQef@4$(eUC-O;Nsgbx}GmY@P(Hh!>ryFN~fq<Ms7c9
zo~P(@DPR6&xo!}^#a2Vhmk&pD5{tseAdEiudCK9-2K=0}42cN>!;d}-^GUmBbu()x
zBl=|vnHImz2sfFkY<@Z&)tI<#ee&Bwr;9JSkkm^Q`R?BR613x75K)5YPD9l)IV*3n
z<WTyc+K1sGIBR5;FEZ#HGCtiqLE7K0Fa78%>qhU_cjc`8Nk!k}Z<M>#2jOy!?tTt3
zfbILmZGkBDXK^78Do>&cG_E)S;x~F}7ZC+pqex<Sf}tnz82W+}2p_^QNxKH!;Ycdo
zHg*-n8dltVdQ-FDIJKvQt#p(1kX%5<RMS6C-sPkbUdjH^{sDUro=u3p_G7fAUQU`3
z7_#EHd#XzJG$P{i&U8)W_CkA~gGFuABM%ZPkF}YVJv#V6U;eoxIDq&3$f!0M6Xx?F
zlfCiPw1rwaTk0>YB!rjlUB5l<^e&Sd=l4C^)*?5}w+-&^5jebOB?P9kKE4BK-UBzT
z3PTcb$m67O+Q5V-eR%u_!>Gz(eXt%$)K`se5~Q#CwAWa<WmRQv8fv${U3<0TZ9|(w
zdS$K$Pah29KmQiU70hu+vNdOwGbRF6@L&qo0VxDxPdn=>YQiY4(QT|HM^abK5X4a8
z;R<Hgoy@*|R|%<ex5+Z?AIyYJ&wa6THAx&^gS#vO1s8wh-_aA$5NUx4gK^$X1`UN?
zD4@*GaoCj!`XU7KcRB+=B-8_X3=)}dX}!akkTx59fokmn;1qI03N3=C)t1=$?<?;(
ztFJsqj5s^PDtpm3z&7rAQ`CB|SZMX@{PT@5!0!4Z+mPUb)l#16Oh9&70FP!r;sD$f
zrUawujk=_;bDV>hPiDJsos%tf6*n)s9Mbu^R5NRIwd9U62%JfrU#*D$mkI>DieFhK
zaAv~i&yqU5hl2-3E@@Lx@?s#izfiNIoRy$Cp_%m<CXP1MkXHJW7&ogE16QKng8_n#
zfg8%M4K7N$pq(4$ukg4By=Gey+IK=k|CwuPz?<hanO;xtEfQXW!`>BGpt?qwkR;t+
zVlSieIn$@%-PeBUws9@lLmM!;uTzS)F|n@V6hEoGc^+{qV1|UX%oe^dTU^J#cg238
zGoX8~3|l-48Hun)VErzfCuy%5=ExYc&-pI^DGWlYchBVZ2(r#~&#0~N-&&3yX7tuD
z+-#rTG>E#?Lk;tKHBq?DJfTVJo_s=wVi@1tb!ZVs4;vDClsMAhs|tj1!;yhIFlu9B
z-2kzN(!e904X>2<Kgys+>T7zWqYkt7kz;$?_0>Ed+MBC%Jr$QQD_OPh9G}yV&^-V}
z^Zw;a^Pw#~S!51}Q3pO-EaDJQkiUgFFOmj}qljmyE8QLR&@AVn)ugKPFqV|LCjoL_
zSe&Qxg~viiF`Gh#&-&^DF73l-V@vsXUgWPtgMXHWiM)nA0Faa+aezW7u4ri__Sn&9
zL(4T8a~2_TnvA@!x)xDSw99U0kBKuI9r5oc22B^}ZR}frzG8%*T@Q>`aD7zO#v7SH
zO9qcRAJ|IDZ>bhp-*g<k&Ez-Cd&cOx{m4z!^-_}c7Fw;|gW~o@F?^rzt^Hk&H`-$u
zY~L@J*21gIRTmH$R<Xfg3Clg0j*W;ANBZOdEv#R*qx|roT`xj#^*P+@Ds^Q2;FL>F
z{PORmdmcP-^HCoFjs$cb(+M8+TsT%sV_bwv(+bK!@puSZ*I$VV{+{G9w23}j6D#Cv
zS66hYe?jNDw4%=G99=!Xm#0diYxYOT-%val`c>VKaO77S9y?4ONP}kHmR`^!qHJq_
zxMRVgfpv01@BA;YHL-WPDV1f6-n#xZOl?ZYjEC8Q{30{SPi3uYr~K|H*@fTUwXZG?
zP|=Vk_|-+^gmH_BCrLdLm=Kd0OuR}dC=T;#S<2aYv$m8JX?>v0J1XGZ`%$SHNtH8f
zr&2|~Pgd*F8>}mE{yzWBw5xf1;=ib=^4#~C$#KvkN@eVgKv#phLdna30+3IiKbPeG
zoj%e3D4x!PM7RHSZrAzPf=XnZohRHpJ;$R!K2X5N{u_?BG(PMKIDpU-8B2sRUSuFd
z+3mFe>m>maVRoK@N2!8s&O;8gLe#V&6{sYE87CP~SG@TkOfyKO0Y{pyn-?EuR=_vc
z3-=5>`&-mNU@dj<isfp$!~e9-$0Cw)Yj`7cig_b+%wUAh7vOfi&V7h{`b2aguvKZR
zs5AQAkNE2#T+0Ci3Lel@WbAtq_EHD9vU3ln;X8QMS_M3+>&K&K8^Y=>!YFfYCEym?
z{^9ul=PvR0QqPcoZ0FGfH2Ks}xE5NxZ~G8IjlUG!{WsjRJrJ2bIR9=@#h<(bUZxNc
zpz-ztB7^4yZgikNY=|)qNv+2r$6=?=P&p6E8YxiKv;dUjBYy5~mp@|gzuG7Ar?`xF
z0ROs9`^SUnFsitPlRm@(w>M9CXp*aC24(r7|CB7`xf(|rVnS#vy#Z8CU<szjz>6$}
ztv^O$_{MTF99m!-Yh$^J09n4Z4>rP>h-wfZ?wu&%#l2MsFYcW+o0kaA3#6I`lvD&!
zo+N|*%e?a%zOm759>xAwQYC+GRppEVF=`9Co;uOU20Tc$iAUi2z!p+|Y@J19pn%1-
zK!H~2x>Sj_`Y(62?G1X2%kwmDAD(@uzQDLq5T<ARfPz4n9DMI6GV%)3LI$qRXQk{&
z43FYSEdr`gTx-!|QcYzX>O0WOGGY=N2(pkdq={%DQ5M{&%YzewyWbJ>3e7D{jlY+V
zxvLs+#hvvoH8qld=bv+iXAi=&Hxfr!21}6vE|nJ=F#syOPKhhmSRKnDfV8fLQWPNy
z1RKJs^!GecscS5r8^HICkUz;0YiEgZ9|x|fWn{AgxeqWIjwO2SLdI9!j|Sq5w~xr$
z+qEjzYu`CoacsXS`_A4E4o6#FOK;O(g<}^5>C*-$>nB6OEy83`U^=gF5g&C}1ihra
z{8=ca-u5HSxoz4HH!4|D{`;o0v3~c5Wsi9SMks~!SJw!<qD*Aq7G>Ixu!zVAq(PXs
zr5qwCf#;xdZQWB>Gv%T(0zc3M$=>FUyE>;Uw!W72PIfQj8d<z3$S;iWzk2)J=Jk;q
z8ZKYrzYZ{uNX|j{q6SiZfaAf`98|i_QDw{ele)^)2Pb8H%jz(hE|mi9_?DyZyt#;F
zr&qv%HVb`n5w-o>#J4nm|JHqB@DyPP{ds=;6Ns*Ea$Mj5z?p#AcSL$4m0?Ah(?FsL
z=o%7~CxNjJ(1-Z|rL|zD!wJAQ8%OxLJE#LR(p69x&KBsp%yBM*GGra>of}3K)z|em
z;TF<K%k8*#j<jQ)k0MU6cdxxezD;;jK0PyLNGKG8@*<C1SzrwpuK4K|DO(QkY$7g{
zFxnD-XTOY{Il$jnuX{V6O4;{fTj-2}XywTpjGQ?1(>uYSpLdfOPDAys!Mv$N<SaB9
zztW-Rn=~MT{tkr<pHPmY`=Cwvx<K;by{WTDG<xbU?OWfFtXZ*WnMtM?AIBZq^8&zU
zJc7KM3~WbOJG~I@J|*aa0}|!`T{DQIZ)}~e&Tj=G?)3)p)NA|R>FZ^X#;$Z_diAOn
za5wEQN}R%2^dG$<g-<k^H#z>c>oTBf?IBX<2?_O2LbAb)%CLhN-HDZyb*v|_2U(O^
zTK=rOuqZ_Pi<)+Qeff>r75Fty58^t-oa6l<WftWRE>^G(Qf+Rq7Vb|rgogrHBTd5J
zXkb!?W6Er~?yr<PIw#+B@~N+TUB6PNc8;)h-pv*{r-vCC{XPA?1_r*)d%!#VLSqck
zP~{e`B(Wiuvs{f=3dYaef`?mhwlRxq2&j_8cmShi&?L+oFXTH`O<6Rpy0Q+F?mRW+
zElaQ}E}^<U$(3(9o2(rtyxC$YWO%2QgNSb+ZKnbcvBPMTf=tdTwIM=%>{VL6L|qGr
z!Lk{u2rA!e#<ULeiu2CsZ#kK1vFf8`^OvVzg7;Y6a18c44TUbPU#TL0EA}4jHK{;t
zJZ`XAr;1uYt^xS367$US3=y*VJC5}km0Ysc!<(-qo;AqM?mDffbJ}z4rXgX6rC=w>
z%J>fUQS3d15f6u;#aYSBte-@UDqlh7M~x$YbipiJVS>gSW0k+Wqdz$?GYjxz0M`Gs
zK|uEC2ODk!3Q7qDLme#e_abm@EbzAyr%%J|Zz98x3OEn_u%RT*3L1;UO~z!uQp63m
zVQ)w)4P^ZFT9pclI`ypvRTXVVFCC`6Y_&Dks<E!Mkk3nlDXf-)Rr5w%?k{nh&*!Cy
z+YB}$)%L`JyCi|YHE1rKRIk?42n;y}YfuHNo4c8$kizKG)sJ109~`V#Nwj_zZr5~M
zAz5kD&Dv}ILE8vupc)VC5bHS5h~dGpm>&&;uZOoggtU(PuLixrMtJ!xI@EFYS<LN-
zy+%NZ<oRHf2+kkGY9`Maka-ZC;H4SCg#E1S5jBKy@j%u|*t`fv_jn&DnQE=4j@(>r
z;<cX1NNr-oB}B-M?M7bwe(Vy#sqLG$6Itn1Bx`8a@>yRAItc*#4WysX13wB(Qzu}h
z=doTSk32Cuq!vypuoxeM-e>D;YJno?wcUkG!AP%a=gj8EAI*MOi9fjMi>sjWi{Xxo
ztNFPiz^!Va&AEUc1eaT{t5U_^ommvc_)e#V`NvM=uJR2=@?EbWJ*1{<6DtKgD$i2n
zncdxxAZwxgylaA_0X(V#Xd#Qiq-cE|#`h{>A+3hI5^^AOIh7G~Dt@%|L*n0X)}5PQ
ziL;Kz{l4c%hVrh>mw}~<yEzT7jW%b{#t^~EO(UgN1xcl;xj-{!>u|?;ai4Y{7t$#+
zW@dUWNeW&!KbpxLyna2OFRly5nVB?D7`qWvIbrs?zm1<eTipLla@Y5|o{7$aS7}+M
zX)iiAXM`Rx@O}MCIy<Xz56p}Fg8Vi8&pXJUJOaYchw#@g1s=UH$xE$Vrwo~3iy;jS
zh4)5MLGE|}G+Z~9&W7o$vG;N0ndBN(zxq<Eu3~zhW_{7F*TYBK#=j-L{le<xuS8Fl
zdWZB91D=DC(FR`%q;Gy|t|yOVVf-eK+~lwT==LLj6bkSm$Ur6$>WIC<P)CfVPPeM!
zkg#Z+UHkav88f!3rbVz>Hh0gro|@M^YrzUnt~c{@QII)f`Y=vmXR1vQ=*qbW5Zr}k
zAZM9wiQ={?VdM79#ILmn3yP?DB~iN$%hqRn+<3u5-?>H7^kkf(QH<49_!jvuD~n(4
z%N%6TstE|0`~kQEe!$P8C_h^%5DLPw{&)tS6^2^hb8w|B4_t*#9@O=T)K?3*5TN>I
zYfr}SF$b6C8IH<7{XX>!KvEXJkklU?fR9HCXjQR2W!RGK?o3VA6*Pg~0-hr24^OdF
zC;5D*7u$8k)XUcgi2>*8gd}8~!6|A6u)swc1Kf#;G4@)lk+hiT?jg-eLMH~0-*hS0
zp@yJyn9-aR<-O8rYwk%$dTd^wz3qnR*?NJ-Z+AbPL}8*gR*n3mObBLy%|wPJ@%=EX
zwwq&2M?Pc^(XJY79>Z3)O%zf^T~k+fL{GfEnbCi|Fr{#!ZFT!;O?!o1(KnNhU$|;W
zaC|5DoeDZgKd_l54Q=eP=z8%op05aLAwnc^`8fj(f0Zseqc~L6xIb%s!UMPB!)`j+
zLSApS#yz`iN8GRjFJMS~EBMPp;~OSO5C?9O08MlixT7tYj)TMFSxk#6+A@yx32)*`
zH2ZjKePvSmd%?ZU&7W${d{`ZCe0*T4CWxrF;a^V<`BN12x4+t73%OOn8CekeYk;{W
zadg19A`Gcvw_(Bt*Y*+rznoRf3xT>&LCl0IfV}Uew{Z1sEMZuNKR-Y<Bp7zsE1g;j
z*J%0*ZYKGU8a#}wm_+r{p&B{>;~)FM5VQjD6wCuJdUK{RoJ=^WvlF}(NSFZV^*#^g
z!7Bv1Ap_&NKa9fa)qomXh&Vl(1s?5YNEMPf%Mt7lm%RW2o;Koia}JE_Fn%Z4Of53u
zBwI{C2y|Lvumwx&`3#-^tzQEl25FZQ|At#Tu@bA9BV&n15a@sAHLJcV4s4;Yb{mKx
zUj@@(knDffCa=S-8HR}V1`-UxqX$H2-)`}7E&*(EmZvCYxWxvk{4i~V#fxP5e{@2S
zkqNXX%LuTg<7au;)J%8<2!H3>Iq~PXx2X^}J5NH~yabuie+Rq#kppX5BfAo-mLg+u
zwT4pD^TolzQ7~=;^D#LMVL5msTuql-d_UIx5k3<7!#I)<6)Ut%6vJRD%2nbtqks<!
zfdd7~<RAjWGnY51<IWrkaRVd{FvAYMxs7n~Z@@E?0_#c8=nRWY{|tcXPk$)4|123d
zH(~g5(DGWqjYfcgm<DPvY99_;w}GDx@zk}oP?;_PBgr!5aX@4RLkq6*{6X^MspkI)
zvo2^RC}1i;&VYRadH>-dbkjavg5&&sS|nss@DsV%(zy90=22mZ3cMwj!d~y~nU8qu
zW6^DXknCVzL_Tb|1A9>FjYFcn{}ML;pBr3ThA_M?s>UP_QW~i5-;U9N^jr%4+G~c|
z4fzG^8>*?4?U;FS);Xy@TG6ua)k*jB&rB6h>neuWB*oo;p5>JY<0YZcge%?BK;W$5
zak0Pd0SD*?(JgjPhjN{W8lthlb65az-<rTq!Zew<8ev@J5p-{YY^f$}_;CYS^g9dc
z$-x;egcSw`D6tM%#NiXHMk43-_iLhY4;*}rY7e?4C&sypr)0i*;(p%dSVrG;WNS}V
z`{kP&aZ=AAuRr}5N}ko+1b}DKE0;l?ri#%G*)N!CK;cMO48>NrT!;A<NbgE|W(_$4
zz}2x#;d(@5T{}l}fC{|@rc8;5!m**OV=q;;Ql$64cu`sTVnfW}CSQA2w(NritNA(U
z)W7yta899y+{eU`{=hA;4t|(2bM&2{#Q^1l0xK<_N%`&EH>&gBa8X69qDrNA=Qf{P
z&me@Shb~^z>9H=^^|Z$}WfL*Cx8LdntQ2o7f?JdvxCHAW<Zq5+km;LmKKA!R%`?AV
zA=Rq!jrK%`49Aq~k)gfHiEgjlGxheXklN9`=RBUu$4k=VtcDUKWfm>86a)#YJ%K@T
zclta{5Md1Xx`S{%;w|(SJH`hYTA<<s9aifP*QoX?Jr_Fw83$=WmNnd}P*e}x9sXJp
zCk3M6t@!a=@M=XEJ`l4AhEjj%LVvpgwv2&B52!^@7zZi*V^qApBb_QW@{Y5N(f6P<
zGgRA8^-PW4qc1-u`oGq=-m|%hrmH;Bi_xeeymp!g<g?)BExPD`0QF&%vEFhYXp_P*
z9R~I4bkpS39^(`^<Vf#qkri!UNBzz8qsa?rzZ=QqEBk-ZS{FQZAsOCl4;eiRgC4OM
z>48a11}}n4!41(x7+#GNq6~#WCC&=XL2{_S5EG$yM{$gZ^=cv3TU;t-ndhRi-7Z#>
zv%an$bQvFinvpW}o5pRK^&Lxv6tp0B1s+Z>C>oI7pAw3E->DNmPM5O+U(*UgQ!;xk
z{Ez^K&R{}DLP`NyyScNoF}UziT5+)0(y3;xC(j$DHffd#;NrnRgpWd!%*~&IHd3zd
zu;9FugZuv31LrzI$y1szT~_$OTy31q4BU3VQ|rgNIF7*Jy1heYFK+j}A5Bw!9Q?+#
zieE{L4{L^*o)b1~tq~HgQg9@>r7<={%HessKLK^ka4UsaIL3d6dqAd}EfPF!XEr9>
z>q<Uq=U1lp$f@{6u$7H`vQBh|#xqO&iC;^FV7ZGfD`GI3V8Ij&)3y`wH0sRhn%v(x
zUG|7vz-C)3t^?O@>AM!;RclX2-#)MmPuQ}kaQ-lV{A)|&3dekka{{$yJi$I>4qDU^
z1$`yhM|=rv9hlR-+_jX@%c0CrPh;<1gtbh87u+OLObtjQ^cDX~@br)c93NDjNjyZ^
zgL;F{5vbFd<AHU1W<vgLZcm|~z$3lDR@I7-j<&{S117(}D%L*@G*dHIa9ALucZcx;
zXjvm<RQI+-_e%lJbS78aH*~VzfG@t=n1Q8uKmO_?TATH;AXg;i?qLi0h6fVLzip^C
zxXuTP{1-~_RX5@^*`rf4XK-&2FUnblX51UOVqRwd^-59JK+csWi)V$q_&?hBS)7#F
zvMVFVRnhBU>fUCYFO*h^<3rvaj%*|UMHS_#o9y(6u!X`;pFVYtI~CmDLu;55%Wu(o
zy30gNUBOiG`mXyM&#YMr9)gbxu@H=NUb5WvSoSWC+)?vQ2aX<P@(%Q<-L{<zr3ZE{
zT)tB_K;2M^%6L!AqjF_*s%acGjy5<7LqD)1eADZqXHqkrU=(e)rNeuYT`fqohkUb#
zxBCUDCV{$@Yd5NB?io|OxK??j`qqt|X~IYNupI=39i-!<SP@Jz(1NqRGL=Gr1i%vH
zWCn@RE$mNtGAg&Un>^jOr?VhH^622(ss5V1o7bvbzrQ!7BSAqC5*gXQ)S1y}V->)a
zFF!S>=gOjv_KdJ_4|?cefJB|GQ`gO`dV#7l#mQ!RFYl!|-nP1a&T+|rjI(6gGTWdX
zhpQ<3mFgVJXJwrRGMp{U?ij6J;gEs~DtNVp_HJ!mqwQCETXuxx)Lq#BQaJoWLge>r
zkO%YxMu)5HBS>pG2jZ9-!8(MK1J+mx6X6d=J~%cI3nyq3>y}a$O!yzBQx@5w=Y|F!
zQx+OMZE=14B*}90ko(xF;x@Ym;Wvq&d$d141D68wME|CeCHw}n$t)%m7=n!ykj);W
zS_5IUiP2kcu(r}bf-N4EhSnN&YU|~y9qChwb9VJnoEmqqJ0-L1!1)dQoVy$Nrv-sS
zEyq<3lzH%VtWoz%MV5`r+}0#!Q;kk?OxB~1_0Eot?qY4XH*bD=#wl{KB_Y{V@`#T<
zAQi^j|4n0n|5*<c{v>+a<fOpe`U(!iGsFp*+eW@dMbGQ=#Ib|#n}H<LBuwA=-YB-7
zkyw&}-mL8@AiA_`=ZLdt2iw&K6mwHtFS||c)c3m;%`DrzQ*xW0p^giGBgdV+lOuwX
zbBaSb2$M{oyyiV!96T{iUV-NBX`NYDSYqEbE8-mBEjB3GAs)Hw>hYyn@YU^vpVx{k
zpfG$MMr30h)B@UgI>1b*`zz7mG}L*38Z{lee2lPUB7<3U=l<6u2lb-TqV^fRiUoQj
zD!EEUhxLzs&ofmte3@Nn3ay7am?S^z5!lbmXf3p2;vvWP!wv?3C~+jC7^Fd3O-nRV
z<kVe93GLTrs1{{0imS!vJQ^d*^|D58&@3dP#7$2Y#A983&hgu-efBnK<SOBK*ha%E
zxS?_~Q@DyM;&ovVA8xRjF&jLs>AC9nPL6rD%}CFM(z}Dw7kmO#EPu4!*kwZXl*4K9
zeIos}LjbSsXTO1ehzv9s6Pwru=p;*)o5hhHphgpxat;qtS8&9cMR9cS1x5_i-xcm~
zw&~nFUE)~SKHY54_RZ{0Q&Ys^y&eN5@q6@+tN>6ul=^;ls`zOB%G@}r+a_`(LFAhf
zmr|)_v{24~;ooki6jZ+ZdA*p!+V)gaMaN9sqwtK|5x5PCalzmym0@B=s>L;vp<T10
z<(kz41VJM$jg6-AiE7J6UnyQxIGNDenQ-fR_JU>VhO#p_-(@*$E3DaeS%zP_J4qH3
zbG`EP&V_Xk?`4wEq1vwEB5TW{4<6q<zEs$pOx`NsupQ^5FNJIQ^?ern0Y<yJPELi<
z5)HQGDN>jeQBKGmHBe%(G6qu&%!#(|CUa&~HCWp1YYwrxi}K7CzLpMlyvX+wsG#fS
z{d!)Af;CYG-b_Z4L@@C{sG1di$zUopI;>)-hGsH|)rnWUha(Fsh`Wm_N~0Q*4mFY8
zx8Bni(rT^S9V_E>+wxP85O^tJlz-=Nna~r&gu%Nl4v<|A-V@l9D#=J{poSVKGM8mD
zO5FoAja&Vl@44xwaG>iyQRGvPO;6t|{V9(_m-d7_6B=Fw!)LDiCH~ldY7^ez;Ckgu
z+QHoEG2t)QS4HQlOnlq7;NmXH^owb!dlxPUa9Kd%C>Q_)D{WB(tn3bcKCk*ee?oYJ
z=g2lWI^BZ<k}o$7xZh(tDv>@^T>U<zKZz_(@lpslGp>JXuvPw+LX51=#JS&OpnPBt
z%#?#_bNDe5szv4HuKCzH;H{u;r4Jzt7Wj=De%4!cA3%xoO3<8zKH@&zNjjJEGg<aO
zK4@3e#TzLP3%cKv{Q*VV=O+KFWy?@v{dMS_44r66FWFT#{z`*ekJ>BOtx_;4UlY7$
z;qpB@1=magpn)F1gyFgzM8-g)QLf~Pi43x;a_xP2H=oDh5rY#atwfKWfBZO2z-l$_
z#EB)iDGF+;Zw!rNTnGUcB>|8kFbMzu@FUG{xn#8mEBgET9QG}0RqdGixL3@}^1J3U
z{`PI$4GVI39-%;)8GB67RfT+R{DF7Rr0WIkvb<qEn|ofTN`@E4rTV4`gA*3-8--n)
z>&8UX0FuJRS(kJbM!mld6to#W{L34p?~G*)t_vJhZ`yQk&GfdH6NZxiU9G=+cbtu)
zmDhQz)((xlxZo!_H>zS5D=fV5)`edh6QwLe-tj{LzS=yDS6f}0g48<5N{JWjxq3fX
zD`ju6R`&9p;+`j9(?!P-Wzf08v>xj^scBMybK}VUt{LCKs3B7Q-l3Qs`(_kkw3uC`
zkHVw29ocQU%4$z1KR<2>j%o4p7885IVDgt%KK$+`<YZQ6qUpn%114d%rcRgQO;`7Q
z-R|t8@}{vf0zwYI`6@W(NMu}w1A9Oi$mHIzL%4bwI7GI9?^64r4lgcX7^oF_-_^ea
z#DDekKu4*s;&dE~uJ{3vNB~As{PUR;pT(k@n~rvpL0P3CQuix6W`tjM5A3K_@$o6v
zS)9XtI&|LjfwxGE299`~7g3qJ;ex+(bxkP<d@hi%R2R5c61sixl<rEI$1nFz#5Gd)
z;hMm#IoDFAajitgVQ`;#tuMluunC>Jh=xJQ;$rinhzCYm$O8R6txe1BIE#hXOdj|V
zzxy|hVuvHc^K&3q&Ec%h>^Zte`L)rejcXIGzI0OfL(n5}AK>{n>xk2GT~vlE!mGRT
znMCLtL!M+M*BvbD?-_rCbwX`!)|?xY@_>7jc*f_Fvk!`Aq@?skX=?fscEg6Q`#sfE
zN%8JoD4w>#+^Vft?;U^r{i93i{_ih)A0&HheQR_2+?`aDkd$5-C)j(5pYAT<6AhB5
zCv0s$Al0WTH|})yAiFN$Y()1Rm3DCqUUMMf(WmgiBZd|i1$WE=jf9gSVe~LuHJef0
zqxaY=u=Ycw(d2;f2<Fh9BYtI@+>@-}wN8DSIGU{FBssoJ4;W6y+kcWtJe`2yPg*fR
z?-5TWzO(Y$^MYv|(?u@`(rVvkzb0X&w3^~;v8j8rxGP@36xPy?t*{V9-XXRiUH{k?
z{=f9m;&a*+XTGfu9)sG;34|(A&;8iCRrg$ssg1PU(CA%JLrE__#Y0UCBUCg@mp*^y
z@;RbsyJ*-RlS5CQ=&f2aXSt}DqnIiWIvRsW<eC{Y1@mg=|6e~MjKi&1G;Oh?n<LuI
zl1u&KdQJ&OxTq?&znpqK%(k}t{OLyi^Tvy&oJagD=uvr1!In*SdHcbrlkV_={7&yp
zdj+c~*UqjjtSl`K*K{d%`rSnA(VgFCw!MO0${+A>5IfQ|;XlNVpDRrrcpZL<2~?=;
zQ_Z=?a>d%zM$*E@g{9^0S!PQ;2JRHO(Br4~x2A9T_%h^3T)bv*6`?2hCsC46nXb^Z
z*xV)M9TK&7!ON~omss^x+xgz8-#dX{v%=8e7GEHMV_Hys`KfP>aD|+-6FQsk(ioKn
zvTAqS^^2xFbPhZD)V%DqxNsMMzL+gkZU3v<B>ayb1VF^Qu%%l&n+k~ymQhh@m$um*
zS4_`$+4h`o+QKDi$y+<`i^x1T+hiiYg`UuPD<+*?MVP1tz0wj!9O1i<3_Jvf_PXVQ
zSP|5FuhTm%%bg2FGmZr>{<M6z<U+ZfvK#nM<I<Faa5&r|9DfjQHIA<e7kDfUN>5Bd
zkmuXzH3#S}6kwB##f!7+rTl%FUmslfYW*lY{b<sXNBJR^dIGDb>v8^?P=Zzpr;ZSF
zkbySdRgZnQ==iLJ#G!)O!_A(0cgx#ff8RIU`5^8#lzz$^8KZ{F|HyU=l$q~>1{G&N
z;odT(8v=0?o6i(ETV7n%BeAP$<qe@Va=y(t(og3MKk-i||MMpcHqP{&`>GT?e5~`_
znj0SRk~6}uPI*uJe!3kO>bHfMfpfAH?4~{w1ZIqJxQI@}^h{vJ2*4pFeSboE3I2I<
z^FKgV3lhF%WV(jOyXl4`Qe>_l)u|{A?o>RXVI}2bd<OR^$do@-V*@`gPa=%GYf3`~
zh7QcTDVz$Vy07$G1T-_REjK2gd@^bcL0ru&x(zc<z31w=wX=(?qm!`oDcicFL6BF8
zP%g!BAwzi0cYgF9j44P6Cr)=if}+PpOMb=pZ3jOm?vC8nk^f3TS$@N({IZi)hMy9*
z$ugU9pyg|<>Et;{FBtsfN&{bFyM>%@W-WezEP3BIc9%1GiEo6JwRL-w=GIu>RY!zg
zshHw$8m3iXcsc5UdZ=?<Jm);xq!LKzR*bQTq%Hpxv%8JlCrhaBsZ~gSB4A^)*U4$y
ziJg**g7h;nQPN7Fb}7k41}F#NT><u7vuO~SlhB^zto+Jr7{Q~){uaQn`=tT|E5PI*
z75c@w;4h9G@1`;=z)Z~7=T|X_T;b$dOlXEM??uX2D*X+2&(i_ar9t{XYV*HY*FnN`
zGqf2fb?`b=ocl+wq$XRs>3Yvxy|Hq2MZt>#4Ze=Fl;tl3vn(&r#Q64r|Gr^Cb52U%
zbWMLl*Ow{!Rj1%Cp>yk%slMU&tS<@z^=@DRUp8Ewe_j#S1s1&0*QZXL%CM%qmmeJ7
z)#Wh&^9pWEwHPD?ELGpoT9}(`dOGBwnB2G_s3q>*M|@95rL^L88>Kcn`<!?bnXdAt
zImC)T-n2^QE?}yt^i2#ZcWJU9woGdhl(3{OBu&dVd#SqMSGXvqC>_FyI9e#)bl!sF
zGcZ+3BS3|`R)Tel(H;H;CTkJRmtD<i5~E6C%4TBEU9G1~6fgRY*`JQzF23Mt%mWSG
zlqU~FZy4izGdVwP9Y6OTap3wDRv5@sH6UD0j>($Xv^`Pt2o&(WwCRps-=4Q!U;7XG
zJo>VIVmQGma%bux<D)XCFK!JAyz`Sh;n95@)xY74HeUq+jn?B`dOpwtq!8+e(>o&L
zLM8$*t-s+kKuNY7eAWRlUPRzw`{a|!K~fNFr0v7;T5qXhR;m9JhTfyImy`@@<Esdb
z^F7#tyov0KP9QK&1o90Z1YQ8F5>Wbco|XcB*-d7vZ3kIj```907TM9k?5%k)Eu!)y
zTxjBE3*AtXz(_#tuuyw`a0H|v@eMx58zQPI5uvZ=3EDJ=TXLePd)hAIGb%c$6}ej5
zOwuVXy8q5pql~A-CDuU-_-#?EaNDg?5=3qK-Yee^e)Fbr$>Vgz=Sn!fcwfnvX(H8h
z^2$`e6Aj|mL=!BBp^66jbqPvQz&1Af@qC@XODR7<H60)L@=#+RQn#;)-wn>rJz3d(
zr5Y$<_tGahp+^sG*(tej`GrW_+%BGULOuu?7~=}tK}Zu^$2s)kd269v(<MO3Cmi)N
z*PZ$z?$lN&TV=L%|8H5_)~G8~YDVI)vk;{J*9$`!-2{7g^0*yJW^LqVOq3hx>x#UA
zvX+<Y$>OdD=9@2{l9Mbstay;5rD3}Iobclb!C(t%Lt-BJ=i?M)fR8Fz=+<*Ib0GHp
z@2;h(lt-(}t%qvYJTEUQQ@{4$Jn3Z|%eiPklZP;G{i8^@!gy)^%Nx~VH_a6UVm8D*
zh~Kkgl_=jC$M+kQ-cP5A<o>J)7U+a<_C>zI<hrenh@bILuw3$Zk(v4y9FAY{X3+L+
z^|+HacCHCtylo9*r|eGIa4p;|zG6bCzlpL+oG)S*uVeX2@rpS+J?oQ3X_eKjn4N-x
zMJ^zVj7XG%g*3xUES!s-Z!uwQ2U)$M`c0KM5JP)5Q_-=~?zl1I;ktO?@Knik@`Sxn
zgnXRUyMb$=mVp(}ViyWN2@vQ^yqn1Zil6U;dQl>TBNag$A$03~KXZd`Uu2u^_njkY
ztWMF{_*FBmALf8ke||l|xgBudZuXo@e~7&Tg2UZcqC+Nri}%08Fxrz;l;2bKh+#aS
z5~Zu45qy5{LjFcc0Q%;>6OisZY~H?#aH!*~*UYVi=-hJ?6XJXL*IBm63$D4i>(kX1
zF}{?aE*y50I53$#rna1S!^fX^2{+kaoc6zj`fSW^{U+b>?Ke};tv^I6ZknqlQtLin
z%DqSo=jW9EBu2}RqliO($hXbY<_=x`>bI3j+Q}zExA-<SsVGNkHY!>!GR6gf@mly2
z>Mg+l-eWoN2oo}WA3@LiJfdp}-2LTLbR#9S+?dD_6J}K=%JSPKZZyzCndynD%Gln%
z138-|Wlt)<_FJWRWhIO#kp!#$Pnw<Z(;4#rf3r1Wxn2Qw=DTkmY@3l@{OMF&#YA-T
z<God<A8Rr1=kvW<0%l#)P_l#1t?l)!<(+Ysc#?H~g7b1)yCzR{D|YkeT~FCZ0*0<H
z{cb2l`QI$tpp<)6UvqF9)g$-AQB!|nPNS5+59-#Um~1{Lzkz9xYNlDh2L?X=yxw3C
z@nbMgYy1KG{&UCqAD^M%id}1}ROXnWhuJFF%PUE>-~!}UT?x%%QX3MlL=T5(mvL8o
znd?#$=rLPW(a|L#KB=YtdVhpWNW$@NxE0VwmTJP;YBB*1n27}QCa;bd4#`m`Lzys0
zrMbk$K=*XPRz^!5AqprO^j+!0CBe*T(h4;HSm$(OY2m4^ZV_icKmA26l2OKMujj4U
zcjJHye*-xCcQBDbkeo)W7GWJ-1Nz+o*yF}C3<*kDUr*}2^gu}HFUxt*|8cY@HHD!!
zsQo#Fd#i=)<LpDGXa899q9|29&sgh_a`{DtO`DbmtMlSuF)DgCwug$vs0}GWWl<Cm
z6t9DH$pXCQ>%tww*UQ&A`#|mGL`h*ObL%w^PqI#|zp<ibZ_DGJdM%BHo(SuhV}0?v
zPZ(`lzEcp&c1=_P>S^SI-!XJM<BoMF;tb+>B_``JN0T9!k>w_lx{xtZ-#5dZ4JpWm
z^l8?oJlFLlr7sdqhm-d>$6NbcJa*Gi?l#md+}Ys5w3cwBpkx~WCEJ})vR&QH>)uWG
z<YDGR*-)~*2Emwm?l=W1q{F<PZIM|h*)9UV$&vJdh()&ybYs$!P%Ra~5f}~BJcKJv
zsm|r-^ut3xNkNuA8Uf_yoDT6C2_=`qLtkn^hFFTHZ>G;7i>M5mFFX{;BT%rd*P1ar
zXA^2MM0(#w2qoe52?HAn$-=(<_6LYUBU0x=lctQaHR@*FGkdF)T)ICStv!`~70iw7
zBc5y#?^96i?5b2@9(+LHBI-m(2)=eZnXzH#3?%%Ilu&jv<!-xXk9r=rbG_nc*Ss!a
z{9;wxt~a(i8*GHOHC(JBeD(h2u88C=3v43h`~0TetW(!7j!H7}Yx~!>K0W%~w5tFS
zaM>!AnsVv!>tb>1#I6f1WR2{+2zl=ltAq9&7{L3%^cSLg_E@tT%bJ@$qpJJWZxb7T
zl_r*}lB;^!_ngP(iSkuB=48)>FOA@|@4u1>H&S?XVkhtsqH&a71y(NX)b%Lz6{L2X
z4S664l}G#*ySdK*hE-CSBQJ@##9lvSygBQn*mG(uXZNAKpUNWaWn^~;KY5V7VDXw8
zJKt{L=Oz)ITRy0<PIY5^oaNo#)F6_Ox4pA1x|7j;v~$J`2GyfizOoN>%n%m26erEj
z7HfRaDwgm`I&5~qQ~ybunO3aA^E0Qa<iH2`h1B4uAXEx23~VLJ8Hn>Jc+MhW6p_j;
z{>BWotNIGebW$rvjgDUe*RNGykIq%@@2O5}FNjvIVF~L=Z;L94>F-}Xbm+~lDIk3*
z#-!n;K?LEQd8h#>sO9mX1b^{EydrbBjuuW4^VSP+A0$SXTMVVxw)Ku*>Pd?9zBA#*
z61MCODc6zs-lBN^+mG_X^Cgl%T^u@(`N8TSlaBwfm)JmMuPG+X`S3!W18z7BK?c4|
zO8ZjjlbjVgLp~84;Xw~_je&eV(;~>Ru#2-J%ih&CqukTY$4tY>&sjyYA@31i=kZ1S
zf=`d|b21Tn9<QByZX5MF1!`3$f5SB?<6gq;cZ=;gTyMKIolaZ9@kX<rj_6k8fB2eg
zkUivaW%fXas^Xf8ioPeaZMFKFkBG<Y{cvEr%$XqlB|$sxur2^Tl6x<Z|9yXNy^pgW
z^A0^Aqc5k<T}5JRw}H+(dTPH5WuxNZ-IFqL&w57Q<iHK%mTD2+(^v-s@Oyl?DF!0>
z1__y(n*Kz_2bE@LAr;f-uN31=cbMZ~slU3QBiX{js+(W8o40$Mc-}s)dM`gOuQ1|d
zQ9(gb^r_<F;t<^4d!dTQhN5vtjIAtj3xZVVV`<;qXrePdh$LG00nUPUjj~MNad1=k
z$zqBflkl6DJ?TiY@QI^~PghhFCi=vzI;F7dv!%N5bvnd_KdTsolFN}T9JE*-2*{Me
z+3{IA@k73CN2P$TGYM6DJSNqDmGSf&dEuz;(`<?GwBp!<)aUP<o)?#v?-hzTx6b6o
zo_(RX8&!l?-M=&+au=BVLnigqNaiSmvmd=VWFc4opghkt4$ZC8O?{ELVt<*FpE0%c
z#W5-4ANJ`>j-B_sYAEOQhl5E-1LCNPl_pFjj+i4$7urWOnrwhUD8@iEzoQ^$U(co{
zVYca+Tc%`hca4I>HA7)B#>#z6h@(El{z~!6*)f|~2Hb@4N+Hh4@&ayZAl}a2)?XF9
z=IL<(ALPGX=H2m~bXA9AXCjTCe~dSl&u$p|zG;KP4RG4`=zGhYM`;aP>Y`*0`exUu
zH%j5MWL&fRAt0}^vf?|!7qn>6wzOqOgNQvXEKzPe@r;N&OP!lk=wUY?SYvK>xT8|g
ze|tHVTES2Xd%AgY^EZ9%lZvwvnoXzM%(e2=FDqWO`91#1JHf7B>HS%<Flt@vAoQ*P
zG={IyuC{V*df$3N{%G?02&Y9U8Sf2+$m`o)cPa!go>kXVFg||wI;SB>XN13@oXpnq
z7eteu5+f?taMqwcLl(<lkp=FIIjm=-MnCyVc6TfD$myAy>#}*XW#c9>r=m2vUnFh_
zT*aSF4c}KP9uxD|L=QS4nf{s#LWsWL$;ZbVFF*Q#9yV&9)^i_wls<L;MQNBts937Y
z;CiiiLmSJ4)qoB&gqhEN7+wc0W*X5=s>$!)PQtu&pjBTob`(t?iV8R|RFnMe-o0jI
zp}|UxdXbNb65<p37hZk8zxUXti#UCNWnxI}X&#@%^aVtc2oNeTONmGRPFeFr&VWhf
zfieZwm7+duBQ&3cei$%D)rXo-`mbf`94P~{Wo`AmDB{Ts--9tvhY!L{N7tpjnEEYF
z?qDc5K<VATvdYAj)B!q#rm-v^1~eY%u%0tbt{DiykWe5TX%QaHbM5I58R*0B?^Jo<
zaY(?qsG^`C#`jp3^1f|~GV?Gf#U30x6BB|I`9=Y5EOopb6$Ty1gv?IZ-&13J-}`gf
z>jU3h!7>AK+z}kM(yg>8({tUI^ei{IMGS7|#Kgxpwsv>2Z$C-4c^DVr9KXvZR;dE#
z{41*u`ZZLW3l7}KM@$5vSMmUOYk1-MFRaY}bKeAmYURMx<W<TdObqOMo?Hf_g^pbH
z7epTWlbIQ*8JFo6BCfWw%`;C7gbNZk`DGQbeH~mJ6Rt<d{Pws+Wec<?4#yc1+GaUY
z0StB(IGymoCxLR{c&Z8cA`Pi>HwVJ%#~NGx%`<Gqq#W&SM_9QY=evAb&G?cRt-F8Y
z=_mcC8gQ78)P8lCf6>LNco_JKkcUM-vT^INsKWo2d@eKdc>U#*h|_=t-mdRMpW)!%
zbE=t<uTF=$q!=~IFMnMn`Z5iF?;p|7ysJ+9Gq=uv_~~#i*Lta++KG-9Y&E%U07I60
z3s}9k^J#&9?JBQ<5YLeK=_TSe$zDT+J(-c{0F9*X!k@#yb7BLRKv!yUEoII(4+<|A
zQNh3A6x9Zbx$)#M;#}CsG~bxkpSnv4!{7@DmLo5}vsWPRGl!^LeQjuOot_8ww!wrA
z50KPY)CLk7>4P*Zxu6GR%2szC^!ndSk9GG&FrywNZpbu^s{+w<UlWh&>E<~ug*YDW
zzX4vArvS*1a0Y7~3<2e9E9yB4Tg^cc=Ev{`o~r$q?R&0<YVu=;@XTE7rD`LV0s(h>
zG7Nqvp}GjK#h8XQw1iRTyr3y=N`2nfU`>9wgB>tp0gNoKkC6#*-DW6*_q9M`e2TA?
zI4;VA_{wPU<^Ew@=KGvosEjLA0EUl)Yjj%T7!Mc@0Dvtn|2VRFFW8tk5YJ<=exHps
zbzh@S`xo$GA8-9%%%BC<Fd##Pvf*v~IAk_YOF#npJ|v(60{*-fjw(_;NtiN1P7&G4
z7<3qS?1B-4z($o#MK#N?L%EmXtI~jEjW>oNJbs3R5WpnOw{L?J`TX#Hh^=xea7aw&
z{}`{&OXz>Zlo*`A0BaI>u->-=4^|9uMlc@+;CjLU+<!5$mItYa8uFk1Ab)ORpWn1l
zXa+}E)7%`QV@nhq$$@CL4#yRmWhyiq2qNW<S!jHFt&bU5JUhT?@KRx6Ny*Zj(!C&i
zUJnRmjl>#0P5^jK3}hgSGVMzeqcWVpQ<nxZb~YwLWdd{;7X6GOOLegbf&9-w?z!)*
zM+~JJQmC9Xv$z&I9ZW7D(9I#{Zm^EBr=u`n$MUr?UnloRsIEJ;SN+lI)gT+WE$D?|
z&Wc9t`Xn7#Tp+3%0`xu9_p0C082f4OJ$YE$;FDvNRe^i}#Slb2hS<n;%9825{$^=5
z<r1SAUwd*oY@^d|l*n1VamiX!{c<(L>u<$_b|}#$>S^KB1;BLaIM`T&h`9k??Akyt
zGYfUE>)9WyYGhg|TE1)W!`odiOl3V|u7lYh$>k7%Y#riAlMtpE<n|!H&Md{vwtR5J
zmh^#P!=pp&y?|ld!L(qdqG3b$P$;_^D4}}Q4%-+{vjXE?E7yWq^3FEbvuRUHUo6~1
zv$4=Ff4ywqc7BL*lc0yP1zi`*m88`GcP0pY;|6}=dLqLZkFH-7L5~$3v$Y<NU^KKG
zZqrgF9cj;%Z@+8ij%1F`N?h_$r)Wm3c@etZl64xM<wy}D4&)<@Ly#B(zb;~cCJgko
zE2d*be9amiHZvN-yzbcV6z?^AS0Ia6wiQI{_U&8o%*K^f#CjL(8?#PkzTX#L4g32L
zX{GzabE<D4z{yJNg7F9`Lo{*DLGD#vpc@{j-}RJZGDMKzNHZ#n2X|cJ&i1<{XLKC8
z6V;S7<dt_rSjH1*t}cRm0FqD*MuTBR=pAN3<Qbe1_>LrmEydal3>bvk$z^~taD;}^
z%Dn|_e4%UqS!}<PklbmCre`;>`G`Z850P=CWlmo$l7wo2Z!tgA_7{u9qk+IXs)@x4
z4<=zt;YJU6@o|9%pt@MsK8z!-P)jNZ&4&Y%FS6*#JcRCe>g%{4CHpXC5EoOro$R?P
zd$+_nSMyX``xf@Aoq{U7v9=v}&bIHEkRxOTd>F;><vfJ>8iySQF&`(QN}1!H?6qK)
z@@KCj&cv&M%TT*P3sL{monGe7R*Z@Cot{WWJ$+=eWB%>ichptJPvacF?0`OAWWn@*
zN9l)eh9*Fyo*?JHiYMxUJ%m*Tt9>EcQv*9@Qr&FEBV*KR6s3*K-Mvy2x5U1u<fYBg
z?HPsvGSkq~_IV?W68rfOcObkD=3qFAYLh?}$Vmcjd=NhYle+8r7orhNVpYEoGbzkl
zcsjn;o<WV^m^~cn`4;DUmCw)KWPf={LCQ|)`w5}e+o7ePhrlqRq3aR_IVXo}+Aegn
zw;n*I)?!Es&P;+M3aCo=h0o1-y}u>vLU$+XNS(cHs&HbYcJ^AriN(XkZaA)KtCojO
zZRyrHI#;SqPlT|mT}8`7xj{7i>~x<-*QQO<d&68NCQd}IZgkpdDCNcXBZsMk34sk=
z_yu_1plf$O8pT%i*JWD7=sQ1d&Oz_g=|*3B6_D+gv?NNquPKZ#!#kR9%`&m`>XEoX
z=vJpN%m{Ou)DSAF3idw65}dDPpv{B9xU{D53MtiRDV&Rt@A*mjKprQSIFJbA!4P4D
zVFwmN1_E^%(EAIYL;S9X+QDN!@Tjk#vGJd;GFa&_Iiy+z8F|Z_B{NWDfo%uQFc>gq
z#BiiTd#@4e2zmi%%Y^83SxwdnG-n7O!9<nn0yWUflFvK6ErctA?YG8xZb}kyb6&se
z1-tRlC*bFUclgB->t7q9a8?mmQ``&#*{@U~Y%A7UsgGV|ux(S@#tt*y7GJsPI_vp<
zuTw$%v6=<85~p6ZEF6XbnpN|LL+Z-r=UKi!Hf{bp)9ss1K5?;434NYH+I2xYHRGYZ
zv7e)l=<tf(zMW#Nch8UDp@p@>mpV{PVH^pUk6aAH!y-3|sXsWF){$P*DD|5$p={(a
zG>j?zQh!1t@L#*t#Lir7F`a4BNRa|ASJHE&-Xa21pfY`8%gGE+^Dbb>F*mPEnrp~h
zpOalP>vLd>)X|vLHz>PhL3xh`@GPQWJq_B(SGKdZb94R8@B22guA>qIQ1o+9WPD3l
zACc}>DPQ3F?W|UbvxJ82uIgKhQ@vM+w!G102;ltbzm^kXG{Uo_VlW)gk-8W<->CUM
z)s!Xv$|cO$2z@8_K~sO7vt1L9s~sOodN`J>yK}Kt=}je8K7qp&{&1=IpD_eu{p}cK
z4XILS*uW0A$=XzPMrbxU2lcdnb|NX+A<M<IxV-4%Vw+Pc`|kRHKFy#pfFViQnJXtz
zuf~xyfr9MFg3xSGOiw`wDehMsGSb~Bg)Mbs&6?KF%G<T2O})j`?T^uVU%C5uteN@|
z<9GQ%MC0tSxp#t}>-#!pyRCJK4{3E<C#|$U{)TqN=9{AHnK)|CW!c}^$+%oYLT3L;
zc<>vf+RPtMPZ6x{7w0}PST}-6Gu|dR-FLZS^)xNYBXjhMmyn0E^MY}`Q^AYdv?S?A
zmR&HE75w*3{ulZjuHw(W$0Q66-@tu_xWEiKVU|1hvAzY{gmailsNdmJLz2{2xR3fB
zWXjfjw0P*6R4G~tnylDKI6R}oIN>U`wl^Dx*cR|j%}!{-z_vHEqV>1wDgTGNcaMu{
zZU4uo9Su^WLMJr{A!#FWNGpdVVGyEZgj7<4iiv4fiVjk3$B>#3v6E1VN~803Oa~n?
zhop0@Nli<$X8B!{y`S%X_U`$9_viKezP~?y|Lp7zYwmU5*L@w|*Zb0!p2Y!Ke>&Se
zq`NB<QLAU4`*?qHMOvCS`O5du0p<I3)3T;bTPS;0suW~pJ=7I9l*la6{Q3Odj5_;L
zpx=<9a-br_%1YI4?evHe(`j|N+NrnNMrroEetOvQ?gdKY1Q|?t9IMnM(B+Qi%#$Z_
z4cK8Kq=YuoGhrbfD88fFz{K7@)Qw|QB}NrR7!>Ua-n=+;LZ5E)Sfy|OK}Z+J3<JHU
z4{9e4I!=L#+kib!x(R0`;%Nm}Q*UR9<ojMR{BkOx3_I*`)}@Xer!MJI_AV4gL)YM;
zAKoPOn--8}!Spp?1ZIVykIOoj30W%K_dHEM`Y5(zqw$&Eb;_$ARcV--znNX?a1L5B
z{HL!e9=&VkXlx-+>080(z*Wn4Kw7Z8889~uAuS-&K9ox}gi4?en4tOhc^NgOX7yRX
z8F>_#nuGAL`y6{o0mn^J9iBH)GVIk-tfFJ~WM_NpHtqfmDPw}7smT%wy86Ok>{P@f
z48cvCc15O#*0Cj}WTMnhiS+SP>O#_+J10rfq>lJ0S}`MtYgL`6j}HjeRAneU$+Z*t
z>&zS@T&oTgUUY48S#|oHy71_WdZm8WKV3Xnm?2cORlGbBoHRPNse1yq`a2~(EQD~7
zZR6NiTUv-&XA+dk=f9cRzG2nlD)VQyMGF?xT=GqeeeIBMRMMFJ5PUp-#-Ch#`u9K4
z_Y6At|KpLm)E*EXDhDFJi5wdR3<nAw2`q?XZ{1Nm(CV2y2>INPx+x|TsUdyFm4L8#
z+oF6<(ocjZKw{^ieGxj~{!GCSh4(2)fEy7Y0S@O6T=p8P^tl&eG89`4MXOaY9a|KU
zQa&Vzr-1`74njnb&X8-+(KQEVeK*$>@CWuM(w@&8jA?tp%nml3+5$mXhZjaCuR_a6
zc_OV*`NxRZ$@0t4(39Ja44iQWQ1VTfA_QtsB@TVLha3;hpE!7Hw^J8@>lz(|OhJXY
zz>-93=q7)&N`*S<GxZ>(7)qXJ<<ZFJPGlbo-fkpB@C(bKiclA}tamKs&-?s)U_~!=
zCk#&jchi6(Zgl%Xlbb`(_htUy1FsB6V$5-Jqmm5Tl&KYHZ~y%b)OP<r!n6NKTLs4I
zPvP62|Bv3P3vUM-`QkRW4wde-q`}CirA-vn`@&FvgNB6>V+G_URoN~`lLbh+$$04b
zJvDfwNYz(zbSG5*HVa@jE*=GgAUC|D_5~>3LW!`0%i1aE@#H;BmdD0SK-*~U@QTfU
zydwP{@E8BnHuOM|jEd%i4cpJcr$P0HbEuJD=U%d4r@BPD$c(lPdedqgvu$qQ&pNg$
z?vB+hub*s>j9U4^-)ND)u2JwR=mxF?1oY|Sq$(9ivR;86E-(QR2Dg8J0nXVSc`o9?
z2PV|c?aAbGh>B=q>p5HGahWm0@Embq_bI=qxeW#VG2-s9$A|V?>E`L(zMSIsu-=7X
zlB*QFR$YJM?jR%?7&Jr35i(%VM2hIR6Hrwi4npRk!iNaPViH;w(sf`TSzPFXq382X
z*1rvrn2C~urqpm|r0ig)Y|aaPwZU1oaM&ZKw_j$2#m<YmKj$qQ;&`t_^-57$Os*Z4
zR+>k{&kJQ44<(K#a&2z0Zx+^^TbW+!ozhZs=u_=$jQ?V0spaZo{ZM5pOz@Hek)DG9
z+lCCxcA}CRPG*YuTg`<=s)Jw#KV>I7mlN~ualGHTS8!~9Z5<eY_a}Vbs#S68-*<^B
zyLYSK(H(uMMl*5O;iS)d0Oyd(hKcJ};V&7)MvQsVY8rTSZ*v00W#M_+5OYW8GLF;F
z__K|2B-)V~af~ruj&`FKuHMwRscN>{wqFiJ?U*c7vNo_KmczXs2I0jGx|l|o`ll(p
z+)}en2$+Yrz|*6+`GO4F5*iRQ_~FS(Y-gc+M{{GKvj{EdR{x0aSzFcRhmODltCuc#
z=_R}975Q}Xc74B!^d;`+D-Tjnpnrd5T3aa?1o+T2NZ$(6`Hz@|Jbb!E=SN}S3b9$I
zndEXC4~xgs*H(9Pmb%{ou+@i11LS2zhSb>^ohkTi<rGzV{u(e7`l3a%=$NfwU=leI
zB&fzok#GRo(0}+%C43pY<1aPy>G%SH6TV2q!>4izg<a=mt38iZ=SbEYcvB-AwpCfq
zIa1F(JByaW@|qu7J8W-cnB}zzT@(VuKRlT)e1P`u$e<CEGnil>>Gy${xjTlZi3l6b
z&hRw&`vmyRX{p-9z~F+aa5Ow6E1qAqiuBSxnOvFca%Vr|#)nU@0+KqeiY*>Q7>%*J
zFw95Cc(jrnKqy1JhSIktSh1p%=|{eXp%ZC!^NRS`1{Q7{bCSi?;Z^__qwT3&)kJps
z*7x@bdv|Y9x8aTtwfO0}j8wa;)wbDL=_fXA)xTkK@$G7uD_uiQPJoAj0F)mF;MwVN
z6lQ@0+RMOk%=hB6V(=Nwo!hY`(qZz{yeX~R1@bulex`ms=VzQRvZtxH*3bX)RAeHv
zf3r&bQm{FF%hN|^?`JePYAs(8bW%M-NggoNStvF0$Mb~^aQ}f+>>HLPwkSnIdDOl)
z#$x0Av|;R<;;<9`cDXJoE``bLoafo5%@LfX$4xWt?|u?pv!&8|AA7I<)XmfDhQVCA
zt*Ee7fX~vILW_1%QF;!6fr$<1#{(@&h<HXfK8r5bHdO{_VygD<$p;Mvnh$!l<*5f9
zxAEW)4y_qN4S^8-?WQUZwOQzOJ~3vG^f6v3^Fi}1^2U<yP<o#X)#9v`AmNs-ds6mS
zKmOVC!5dbRAb<Ed(_49yS4{AEevxkBaGE0j7b!n2g73q`jhj2(!})50n@7x$Cv#>5
zh`UvAR*{J~YIe!1E!h=&!yLD*{Api8`NfZ%Ol8+^>`vZg<(@cl61-9{_-o~WJ|@R!
z*Zhcc|M_=mr@+y?jbc6wLbF}!D?gzq{6-_H<QT7Gwng+%bYx%JST@5Ut~fjM!&awf
zw~8+uWK7Mt4;`xeI5*EskV2O3b!gDi6~eI4-ZWi~2fjk2c4YjMa!W^$cSHSZ)Az5v
ztuEeYr`c8Xjo51Q?d+|L>?_J?Go6i~I??tj3C+I;X{OaC&y4M}<Ah=1dS3C1Uy8RH
z&aYn8I_>dk+l89D+%8|EwOZ>@e=t5F@W#b466%WYlzI0vA@fH43P^YYl!Yga+Je&f
z9!L~9411H~#6FtOA8$30&0L-tpy!xXVQx0I?U)Zwb&bG0YFui@YL}<GpF&BQok`&~
zk!2`$$buLvle*%jSi6u`z=+JA9f387SYg+((v#+Q3FBKk*X>Elx9ry_oLcp=;&p-Z
z73bsEh}vfQbn2raq=wcDq8GwN4=|#UEJs4?_;X$cy^|)^XF(#Qnw-d4{w`C^sge=0
zHq>Rwvb={O>%*O#cTe@Wd8P1Y-Q9;Fk@a}95&_;r{u@39!C)LY@{Mc*BrUyVvx$|J
z9J+|9OBh$>ZoFUj3Fj7WJs!VP;BxxqjB2k;_o$WXuO6<aIb2gqq`xsHY?gv>g@B@O
zN<>BDZVV<v<xfCzEZVQe5EIU(a`s?l#JJ_Il1MvwT<<e&7ks;Vcu0ZT%{D{h92dm#
z($R_|^TlrtteI(FV7O{MU|YL~R#QX|n*V}{*)VHO$dT<bO*W&E-th)&+>G&}{TY#Z
z^iy4Z`T~vpdiNMfJBysyHqR=}T`$MvsfZdDezQhSt)a*d!MY|W*d4$@IM9vHpgr%d
z``VZnc1Ua%$~9jMu>P*|UdI>k=9kqTu#^R69cgllIdCmLIpG>Q2FOi@UYCSVj5Q(E
za^Tt9IV$o9#53&-Pd$(3b(I;-oj@2%=s=`&XPsE?^u=T;lMy}baeQZ!?}T4ZMQ<$8
zPX-r$m-!Eu4ZYjl0p6EN8Ja(xJCl4RpaAbt;h_0VgXdtjr!7|mD6H?MX2)tjlO+wU
z6=M&cy%(#QGi+n#vgZJJ0oM7ydnZGAEJJ+{If38{dD7`PRrV7|hK+w(5R#6DcIBF2
zpUOnGg{-m5ZtdJ}p_lY3g&nr`$<wMYvMnz>*Swmw|K&;1vC;0MlyG}!;rWk#sS*L#
zG5;M8a{wwia%2{kCEX>l=tyg{r>nFHG%Zo_8`WY5rbqR$>`$Hp%&>Lxq}8#mmo#G+
z9{z3a^Ie~57nJc2Ls&=ZkTN)ws;0KZ5*kcWeK;J|MY0xRE_z+1l!FT-dthK<AwtK4
zdrFJ0uFE==ku&msr|7{Yzm+*Bwr&eQyX+-!`x8Jh+&$<Oml7-9r}@z+>}HqBB#=0>
z5a=M@wr2Cs(f#YoSuE4NGugY&7wV6*oUp6U97cTz!14-PVsMkZkVa6W9><QBbgGyl
zHj=Q2jJOdo!M*$D)Opsw?~K3ac3QJ|V(9!?>(|fqJ#Re$9rE1(3g>QyTQrj+Q)TX)
zk-#*!l{kWLyoX0N(o4Yw@<7L0O=*9eH{-o|d!S?Srx_W}@|{wB{Q%3zZOu=P_=nD#
ztWJS$lX_G<o`xHBD;$F{sK_f-k<3Q^;rHZ+<Pm(=n7VGD62n3npxx7%p+3Y6Y=Eh;
zpF7=q?$=FE-@7f&yD#camaQxI{OPxsbLcjux^KrScl;ni3gNTN6pnsJI;fCI*wGWj
zNSOoMN>*ZMfFBp@L)ir1S6%9uQ(;)?SanZpT~2&c#;!_}`c-#~xMQQ9Ujr>4E0CyC
z35|X-91TaX%xb<O>-zuVw~jz9LRi<0Po)*`4RJ)w2q8zeKen2QS%mxX$Fh(0Fm&3!
zPQ9OY^JDSxJxdC@HrYgrr|dqn;DHVd!G;>=pTjrw`(GJ_{!?%cf!_ahJ}v-=@^KRs
z7?Gvm9z5!6q|52!C|rNQ^%HPSA2NHwXm#h?#eDhVchqkAVBYjwJNyKSpO$11PdWn!
zJ<U+FrsqSC&i^;eU7CMDY>^tMd4jRNu#htXQV!{2Z89Fn$_8c*)|mfS;gt5O8#lEj
zZrqf%z&omm5xGev81zlk|1rb#!xk=-$qU@2aIb&|>o(evODO8kjn0X*xF6wjbtI`a
ze{e0P@F5Q+EC7yNb)%qbs$f*pe<)ll%ijH-&j^q|h8y638IuX)gOGPOf{?Q)4jnuj
zZwevBvJ<pvITS%(<pm*t#;Yy+wifLu?UOi?SMMmxxh=|HQ^-*w$Uv0KGZhqFg%Fn(
zTGPdge#GZgSv;DLC<aKOZgc=8K*B~F5;pB^gE^I+i?D_~q{&QCTYf}Spj_9+f?x0i
zWMCDVe-44D|1lGyer1>sObY-C26!t~rh;Mn7XeH{e#wtP`o=);dT(&s<rVNa?;q(K
zD1^wPuEvsAn}Jzu+Tf4BmPPV=>_}tCKUe}@;O4P-AY^kdF8el9pvCXF&W6~1)4DZ5
zmQuh{FCj;QA)JoOW6+LZ1yW+21|&l_|J!wfOEN@mMCtpVdgYaIIJIC2nqSDvWDr_3
zX=RAiA73J5gjr3-lK5&I<3=r195*Am>J<4@`GxP8b5@oYW)9nPhV~{18H<YN0eb^g
zuE`4K<FlS7LQ)W)nMI6iY)m7@#TCP-MHfk7!6zD4++TpIw~_kq#8_0uCG4Y<m)_Z`
zt|_^ihbxjw_iGT4key>qe^*23E8^1x+z8a&ZZJAuMOh`?T6j*Oo-I%@0HsoN)FF|6
zcxRUntB<8omc$rWX;--X!K!;f>J!fSLX~hkDp3J0P-qoBEhym=x(yqs0v@(54Yxp~
z8{tj&aVst%88r}j(K*1<b!HaFNQ_cTli4?<y1kDdNY>pu5_>MgGxY1t)%oGlDd(a4
z3AuiZq7dkiFg_60Wx+J)2!3yS@&+nivw@n;eT|OpMbAJw&NiQU^3%kp*%|LhBaS^5
z6O!qbGEMa;%Qekwe0E^f<JSkr?>)WG<3-DJ!-F;e(~2RM#lXudnh$SeNz%ZpIZ~yx
zRYs2PkO$}Lx&<_<A!2@sXBsiX>yX%fdU93jnYkZZHssy)p78we?ts*ddzF9HvZhBk
zH4_E`nLlSjbtZb!iCQ2CsX3-y@XTQYe(1)>ovm4>0lrDNVHb0BYnJwbg<e<g^*-4f
zAEvT(_u_XdOkx#?<8Z|i)?tqDuUK;>B84pH>l3!JH5@<umZZJ7`p9B5BF}`)caTJ9
zx!s<z&Z_=9WwU?4jf_?K+cu~DMq_Z*DC|@UYR7-TA|{l}c5$CG)6Ie)-`5>Vr&;NE
zan&%YG)B7brdVJ1%X@dS{E2KW-{Y*6Lnh8#hHs$O@+G=i#8_oA!XlqZjNbu6wwJ)x
zsKbF>;_QU~-sJxFX&TJti!1!3gEdEFup*A}LsC0(O_Up2yIoxk@mqf9&(G8Psz;H(
zNTihvN}-0W^cehU`hLNvOf-r#la!un#u>@`2YkX4#$=!VIQvHRZ0`FdF5kkw>^;1E
z*TE&yMA~mC9s-k=6lY??YGgvv?~ZbxGSp^aHQMqk;G$d*jLQv>cnOoE@+Pw5c1QqW
zZ<=dxw;q4or60L)mMKfH!OpwdYGsMzOh*5zCGV(RaEa$5*bxQ5MH~x>-jlomzp55x
zBGyB>MNp?NJuf-haa?9c(B-#tsmZNZ)6Q6_v9lXWtmu`Q7d*{#a%5i4DVtAA=YLDk
zjP+Tqvq-rT5*3h)DB35$HqZv92!^BBc3*1)+DX2!gWljPy)I#dA^QZnfNr$AjoUYV
zd3uX5kXT*6<ykZGZfy%)quUjWPCJ^M!?}B$wI_vPnr=AWZTT^nzZ^>}J3A<omXOzJ
zz-!Kz5hpuy-VHZkkfY1xK?gEEF7H7inR-|4?e%jmY^~~i92@+}ZZm!N24yI<eW{Zi
zFF0bQgg=DnGkb43<2G(4Hjd&LC%-x9duV=4n5F&Q7U{XSwkn>c9gP7a+B!;tEW1FZ
zhd}Qn$bS-91lV0>$9CC5Zu!myc$~y*t%L~_^4W|t{tarz8?s{7W%Bb9x)$wDU-e>}
z^@%mX$IjQBrpUC(i3&A_f+cXqAqo#!0?_Oj=<MV@0PGDwmclw-iRg95MLW~)w|`Q@
zxHnVxZ`F$pZ?enW?pEDp^yJC*q~)7`HQzODP6?zMGk;jE%Y1xcB&oVcW{u`gV{~9v
zRNOv)CurwWk1W^tm*MuQZck)pQe+_JQ2=@rl{CW@nDbXjRst^~PBBEK<-7u#kni>e
zysy6qds{N05ksmw%Aq1Dye`N}D`eIk7sLFjP$%sWj_VCPc_Y6nchMfp#rHO~T-xxV
z;ojVZGso30pF*9mCjoN&FnsX;vK3dBD<Q(GkSf?O=x~;~atnCrW}Oy|WfACj^E4ff
z8}7BI(Ewdvo~g>t078!V6&o|y_inPTxQ_HVsr*(^I{kqBO)>Xw2T|*ghvex&1_y?w
zF=-Icjr31NiwxDU{_tE|@e&vNqVy>P_hPwIpJYABjoz-+G^a$`Hd#5*T&aI4jEpiP
z3_)?i43D{J-SBFpZ+dVLaz+m(Ln4ZCk%p)@Up7|3yqSCo5E#>eIf+TqW`HB&fvW-p
zE)?xu1~rxLM4CW=FPNvG(3~UZY=J6>f{T=y4OAQTON3|`>d8F89GmwOp4}s3U^T_8
z5&OC7)oB@}8g1hr>s`I^#`I>E>fDq1b|ifK#lNc%cv$!xI9eV6ezz(bM1^nb0XFeh
zO*7=LA@|Zkc%w7W5c(*tA!jNkFMj_i3~MgX(+<gqOZh?zIOLkRZAEf}{o%sUE8<g5
z%NIX1t_WH>EeJVLhR<XX2D&mg=w8T5>(rH*ahBkRu(WL$s{qlr6FpPiSToBwc1w1K
z{SvPeQ$I)cWaKovo~@-VgK?(3zY6JB@IHo5g<Ez~0dfkyLRK6VWYm<XsTSrLN>25k
z<jfJans=)fV4k}R3#uPEZC~?{|EAe)!pWD%w9Xt&qy@Do_k$*NhNrxjC-H^n`RaMb
z91Gk?Sf<2|TaDimcbb$eep4EmI`zrZ$NUOEn5aHE!vo*uG;;ejLwn&nMQHkWYb8zV
z@RaCpg*%@<pRdcAEYA`?#`j$3zNmMJEkcL65R<9vbF2DTIxaDz1r2SThl(zhZ|5x-
zlN%8LGSokW4Tc<~4oEr!s%|Ez9{kJQ8_KUk*lIB^!-CLImiobVXc1E+?!-7%=>bVh
zbgrl5ywoi-5{B<$%`rYVa<Kb1kD@bMD{OytE|$G8+ID;227WOl($7II=-+k0KTsDc
zW)X-`!f$9S$}_0RT@--b6uv%^FtOZ!oo3tWjE1u_o!cJIt9>gu|NQLkC;OJ8;j}dr
zay7>`^#eP$XmNjdSGiG6O<$^mb9VIX$<d8l@)8F2o-tfL>-?kJ=*dx(Vt`9F3K7EK
zq3jR=1ER1>VOpq_4rkIctiO;x!h=*V@9I-qEb6{L%iPNiYT&IU%4=H^pDLekEv6i+
z`TG&Ue*U|mqW=Hmca*w#>2$*2z08krm*21&MNE@W3%Wz3SLsEp2(AWpB-P*T_6O6C
z&qANwb87A1)G+6{))a?%tL#KUmcEcDY5Jo;s3;=-nX^Hec)%A*=-<{n<H>(vh0zD>
zOeyH}3kqkR-oPYu!8tbel)0gGDwr*qtnH^np3NBiVI&<p*sowS{tUs&Mhi0DuMKpH
zUK?egUkd*?|51JcVEc+e4^=z{Agq@ps@s<edSHk-S!|{OsNw{OEDGi|))27@l1S-R
z#XQzF*LxeN=imJvNWjo6H2<ogCyH(eQOYri1sM({F~2|J4C+@0$Q-(%FDJlogt~3G
z9jM-Q>w}ONM+TAMt@PSPa($(AAFS^y2I|luN)w>Xm;LC=taYs>{@3%9eh+I&3SZbC
z&3{O*EWMya{4}^Y)CR%4XvA$qh{#X#Ae+<s-rJhq&Ql~vFsGPrHnyGRiefvI<{cHj
zeG>3nb6*R34g*qAeyRHuGLYZ_d8#fhU9rjt#gr$oS*t~fYJ1+v)gVqQ>x>yZm$i!O
zG5!$Po=dW<=@!r)VT0yV`AuW!4~*r%1cLj(IMzhrX8l;z|7{EQzXJWAf{>CdQr!To
zUn%*4C*2AwsoV+gA8G!3nD~E~)r6=b!p~_yFy4s*k@jtHhLfEYZEyy^{;R210e|=x
z9Q)57{U0z9fBJpo&;7v4-*$mcEkg6NV55gS$WtcCZ|O_Jp5nV)G3_u*ZKy1D>@<A*
zek>eAfg~`!Q!{0O`^Wez1&w)D8)z@h@ZiN}1D5UGgwZTMRMzN{VSdcsfEoe53gW|T
z$lr}<;kSHE<6~zYDT?2Kx1kfuwPNahO#S&@MlwgVIMzx(_Y`MJ-grYA6t@axuCu+z
zIknz@Q&Ts3_K`%F#$m*BP-Bf2a^1jKa(oul&*Aikg@75l2$wH7iCe43>c|kJU(70;
zgbEQ|97vSdxps$sHAk_Cx)n#I$fMszo%R26z$|}<yX9#LnR-WA4D-X$j$aU;v|2%9
zY{zZRR&oSgCwDrjYc)N(+>E*U<58Bc^kOFGUQvMlk+U%U8+v2G+YL8@n!?SudEm0o
zef4|2ZOj!LnTETRPtUYzK2xS(-qcQF+Yu&SdlQTh8ee>aEcMFRa@>tCqt(5BxndRl
zV5{{=a8;kY0o*Rr4Sd{G#lGon{<Hwze;%px<?Sm{vm3u{Rcq;fbW#Q>_)dv%tY7J_
z^qTJV^@!G2?^4~w;dE@@2c{lh+ykU%J6i<UW4`zcy4URFfiHTFgfUOD8N9rRE2?W|
zp_$y5Q;cU%{@{D{@27zlX_TYo%}i9J3kxY}0U7(-D^h!ggJdk0hXcx5B-_hbFTYSz
za*j=KEETnj(>Ls=R0=kb6&W(df`>t8zoAE?pRr$zQUZ7)x&ci3Dsb%!7?88<C{H8H
zQe;aTh)sHe7=HxpJ?oWtM(^z~nazU0m<|bX&#2^R9}-33W}`(u^U(6xI!FT_g=7Ud
z?KDyqt~=%^r!^jdNP{{FxBpHV6|(i6U)eN+ecntN%vO2r5^(F*cgmK}&HLUhYJ(gF
zw|7vPa`&V8+aW|rf%h<%M_W3`W@N=o!F?=k#bqP#I8rnGo>5hTfYgXgN#UidZimzW
zqkh39qoxMv>8HOqN-Q(bmK*x_lNy(pP)@a4Ezboiqu=0af&x-*?CxBH?2qH?(1W?k
zIg&)|6~49l&3xXYPqp=dQTI1)5vfVv$f`cWv3(0dB7da}*h&`;=I^d3Oa$nLs>YRI
zBK8?j-)$JaZh1{-wRW}h<G@F6ztnK&rFV|Wb$%%DAsPt~!Bsfu0u`o>L{$t$OFCNU
z`l%Im<r-U1%>7wL!2+)ZgY39KFzzIq_{;U#rg;sY!;DSSjTa@V8<7*EBgvS>GPu&O
z%oK(EdO?SyY;g?mn>*mzviUJ%EKY~z^OZdFR3X%k=&XL9i9T58Yu2Ea?dm!=@TNZH
zc|*grAgk+xs)vM^7iNGN?B!v0G4)#cOyyYNaDMSdakr9fBeU}^XNA?d>bS1znhx$t
z)a1muh5;!sfkX==`ey{aZc--h^zj{4NF&B~ld1<5M+@chO;gFyE3jB{994=LuFqPK
zCQp@TwtoQ)-*bFV8n;RbA8t(taEpziv<?f%hq%TA(IeDY5j}*7p$0ZKVLsUqW*@o5
z>sf4QO~*dn6*HBi8YBrH8k;C2pt=5{j(*G^0C@^w?g3rG;?@}|%Tu(am$LD2j`l66
zEWu9fWU4u}z6g-#c#g}u>T`3;-X6<mwbU<f7nSHJ548Mm0>UiZT?`9#<ZD{d*|Lsc
zNR^F2^>K${z<bQzqs_;nysGs;&4pahMt1icjdtYPg8$eBPA<MQ!AoCN-wfYlGQ+CW
z)7y(rYB@jTcVJ1*lD8HI-p(sI^oHVuV8_qYB9hSYE=Y@iP9jZY(MXReD`rK)942FZ
zrCYr4ol7O|QR-Kx%#A`wP4C09Z8Dt;pMNvWtNYw-`S|rU_pwS}dX?li0RMs=xyFBp
zcAUUerB3umsmW(-^AV+b_AGWh;pn@!zNU`Fx=+?tr(8Ul)v(}z!-;fDVB)I=r}|H)
zsqm=(qSgNNQ|eb+xPNd5gVP~zS=xvenw_CxznYz*ikgGcEXT>uk58F>FFPpby-^jn
zc3<q}mj`;Jqm{2oUs1R}!yKIJAo(kRy5Oq}@@TcmLWbNDoVDwa-<u556Tl#D+CFE(
z2#I-+VFNEA7J8)=$)#Z>@+=ymRVXdR_sF)I1y`0fiv-fUhU3rm_4ZaF8L3up-c+IH
zT;`|7R0_)E{rQEc<;+L6gh8KKCnB>a#>sh%jw_y`vl8FZY>QxiR!wn@<GLgO667Q|
z49__<CzAES61r5R&DIlQV{+HQJ##$4@oF?nS1lAlFJerQV2|ko7j#!ke|B13Wb_7x
zlxcEMmXlr|gq(CiN!rf1l{NmU`lL?9OSB`(c~hljH@<Fm&0E><&((Q80?}8?c)au4
z1{KD?nL~LQw@ffLrUW*YKFaE6gP%Q;fkY{HUZX?OtlR#TsAytEXCM7xM9ka~kZPK>
zsOj-%AS+}7Skajr1umA#jZ&FEw-zda{QZTpjUVwT#2(4s1So}_+=-|<eY$-%uliF%
zM?vemjM^RxA7lJsskstFVE?WZkqc?KNi(5!5fxs7%o2G^W@=V6!)jV2O$na|6P(W5
zC)?hJOk;tgo2j6dd^3KGD$`^a$1VLl@OL*A4_0)#pdhM6s=_B@t7#JmAIUk+#4vG<
zqW@uJi<RyIUoXv8dL2@m{ro+*!qBYm(-F(mu&zFXbrf(p|Lxq``U#B&QhmjFa#XTx
z>pI!|td4GPEZ0V8rlD{T7UtR>`ZXQhRdh}s(^0`mWaJE&TrBmOt~|g493m=A;y-*K
z=c9YjGxSGs>}SLXs~N3Z>s=))k9cwhm~~x^PrusSyR!Z0fren(4#|6KBwFwnKZxzv
zxd^s4pjc)JgIsjDi)i)rUyG<BW>6-C2~irkQ$2mMlciJRTmsL$zN0s3!N;8Z-`XLo
zPW!!*fPzYdm`Ar+yOSof;uzweQ7^AD8cc`(>7C_%KO5}gt%6px;fR;Nbk4j<o-3C;
z?lv|~Ggs>3|DQate`53DCXcZNC2JcGcj%;B>|h&Rvm>b*02j)eflr}Eb4TEdxMu4l
zUW(3<lYQM<`cE^${m#R6*xgRGm;Jo@;>v;p0XZ>7$rF4OGg4>`EPS#pp*ao}rlI5W
z<`Q#Qky;!$>r)$rOKBy%`VFl=x0P-tYip*KlO=m%4~DO33TSx#;_agRdL_BhUstW%
z(@2g2Q>Ic?3QNk8|1zIme-ZU~l@@WkurxspV-iN1_~C}Nc8JL2+11KU-u`eCWv?dI
z$lo}^UL~Ms9CW2=|6~Yh1xfzd!3M8e89xsuW5}*t<W4tkPA4I6BLe=4?-X6M_k|Y#
zfZh+5>U(I4)ce{nDWo4CsY=YgQ#J-J2L`wCL5l^TtA6Vjrpfd<AUVmA3c}i>W<hAr
za#9ton}+)Wq?Q|FZJ0jzAkvzL`SKvfl3p9Y!E&zKtT2s=@RmFY*Jm%H7(4s88OF{k
zL+k-M^4PyZk8ehxhD8vYSaV^F&$2$2d}#{=wzD&odSf~c4F*>86!TgcinuBQ1`@qq
zOiKjX$3Mjz-zjr=I4Bcnw#6Kv*_KbnLo;0Tk101D_T=d2@c0nIOYxHM#AL`x^?n(A
z00FLZa25J;CgKwV6PI8q70;lNw|MeU!3?c?1vxtGJLPP%v;hjMjOV-owi9g2a&Y27
zIE6ht0B#5IJD#!Kq0*=aKbbu+SW7Gi2@O*elK~|2s884}r-P-)no0?<C@uRo3DrY8
z&U03Q--T=gWOx*S#bdaTDtwp)5h?pwqWT&r`nCTkO-x%Mn+{IiaNHBNKMz7U=8s_V
z0#IN`py?TvK%|L=FbDQRS8-FxA8+xW`fn8ngryN<o5_t5nJXkxCH>^psjvvikL?A1
zEa)Q=z$~}gewgJJg!Yz?qXZ7h&S2W18IBtIuwVEBum?uLbwgOM0(Rdg2zkuM=h`X)
ze(=d3sR?7`>t_QN-pE7w8z_>HJhW+yBEtZw05}we#RY(YR1H4yA0VbH|6TqO2nB_|
zf{^Vnt7{-fnu!Icc8<Xtgl2|#flOQ$=8tN6oW~d24-0);X<NllH$MHf^X${Pxq`(9
zp4ZiwD}4j0*GS+#M|)RVEkp&KGzHxwwFJb~wv?%k=GPf&V=)~KGA+-6u>9|o@!VI(
zK6aOw>F($0jP~`sninE&Fvl)NT{@R-dv^M>;DPsFF263eytKgz^!L{Wa$=F3kF;bF
zW4=?K2g1vqYf+ZklOw=URQW=_N3pg1hBQ0*ZPJWmk9&%FBeoIqCH)~u6}hJ4ldh&P
z@nlA5^;>@nhZovsRg2a>ReHdF3xRIu`oXNqXFAkip(VT1MbVNiW<pE$DGC~DO7UU4
zZwgx_F%HX&Jmns;rp7#ejUOWqs}(eSTzxmeV(!Z2%}}%LJ&S{+QqOSnKdC`_7_6Z-
z+(oPTWq~5*Ft$VC(ez~tB0s-5dc9?Vp4;mOT_Hy=PN}{x>1uY>{SveU#m{_|!gane
zNm59H--{9o=FETgJ>(PvpGE8B326;7X&;_1+sK*wuCGiTUsBp^L%y!ytbDg1l_igN
zlLl}ryn~CsA|@H?fVKECf6y`yXo0W&3v|UBkqtcQd>qbf9VDY#z;yQ$<4nj=H?ci1
zClgC~4yt`V+qU?VSvf$cWNB2-OWZNXSLQzEXFgcBtof;Ihfk>T*b}QhEa={Nl|syd
zc(V-Ig$zulPe8h*aMuUNYyP9}DT@Fz2gy^0Oov;GK1lAIFPqEO648s_D4@sym=~~@
zSikqbn-Owro8Uv6o_{@qGx6k_T_Zi(R_<N!MaSqoFh_g?g^FpH9%B_iG!G6NcS`$$
ze(ht>a5I%{T(USm_4-EtL#`seMr-}E596CoY_xpfw0PyVai1I~27y*h+5ndeUci#3
zNJC9nO#8+HH&t?)!u*HlGSvAYcVp^U(qN1aV}X`(rpj*++PB-BY)^5`*vm^x7LUJ@
za=F3b(Sy8m?8e~Uf)tJTwQ4!LPkr7418N0>h6;%h;Bd)^F{;uyP`7u;Dul+JPw3zg
zSa-orI+k3xI@jR%9&v3ZHSen2?9__{El=}};iE<$)=h5(Tv-TDqGH_n4GOb-VHl|l
zT?Mh$kd}{-anW)XKIuCpXrKm~X%rM8V|H@|yk2K=1S+%w&*p`nlB^W$F{t8TC748v
zt?8~06Kj|7&%m9Bd0|<FqzUe67IN<%XX^OG*mr_wlBZg)eZE|pM_Cd5)8jz&q_XU%
zMA~Uon22W43i-G(+VM)<T^}h<IFZJV&HqHJ@^gVSi5*tyd_UDZ`(x21Y4^ydj*$nv
zb?5GNH&_e>se=HOD66HB;bEmTnf%9b;O*on$hlBmxW#CuJJi)|$+tjJwnBogmN+V^
zo@pT&rVPhe+r@4d<-O8#wo6+=$Ho4(QCX$-=;1lnT!WCZAB=hvB<*A%pVwqfuND*n
z-vp#?+whxM^K+hN1`A$N#nhZfhp}gQb1F@Jgw<;ApM^!3$GPmRPz}zSd7GV>vod!7
zsZXhApmNrQh(|LTzLJq>K8>JO^S{kj6sokO1l(`|wgP5(_ob2c(9Ao8M*CI?8yrFA
z^ELUxIH(PUBW{97!%0|w1mPy?R>ysdVx7`L$BWFbm|EJimKG;%j7$#T)Sgq(Y&zW(
zx_AS%aL~=9SMUk_XP|=4;A(Jcu@ZporVu`uV^ppywK!g@@E8>E)sIg(zR2&m#DB@o
zz(*FtT^7ddQ$O}42)QE|LdLDO$nT>*^e`CN6;_x-t2L2YyaUq;K8Yq_aNn8O!<W#C
zLs%AJx8L~<@40`#)y3quC$a-4zs!B>U|>Vx9)OAvFRNJ&K4mZ+r=_AjE|44hN76_|
zv2;=`Nq!&7r!c>pE;~jz<JYmAV!9gQW_q}+ksf8aRK)g&c^GMlxXYQWON&>y?0A;_
z>Gjl&8dr3c+Gt~yWK)N-O&XGUBuz&ilIw;wim4}OS&ia&1<Q(ej*u`zJ8ma<YrhEX
z-tmsr`^s})gG%?zMfcBsZg~3~B6KwFXhi08f1V~gY+KgW)$?6^w^k+E%39U><i%C{
z@y5eRi3^sTrWAr@V976|!Y&AGSwenCH|8KR3%63^?8~brCveOG;o|T7N;kvR-4{*t
zkY)$Imgb9QPdv9x_vQzCZ=)$Nip2lF(WLl2*ltK9<uB)I!{BYy=b5uK4gl_aO}3-L
zM5%F<g)O~l<L0NgrJy2jD=+&Owe>e_KCkUjIjQu-Y?JRP3gI*8!3+{2L+&wu5*NJK
zd?(B862}5$7{^^$Mip?TVO?j%>wH)`QyU8dJn!vi*8O6$&vZ(7#HQT_{SWmQtB*{i
z)$@nA1hjeqKCzq7=gS<>{GU<pB{B78jWos39TV&xK&KpqO_i22#^1PlnP0OjdWq+Y
z?1a_1;V-{1VvpXQ2h&&qjr=*SAyS%$S#%i5<`Qe<ck-rqWY>gS8ep#&AsegBd9)0Z
z6c?MN1<t?L$DQ)XQaQL@a^~3QQ54bx`MW;(#+QPN8<K_Yrp16))nAGRC+OgdBr)qH
z)FPl$N|=Tni3z;(=4PC}y?kGBXJSH!E#;Sd{Rgwhbd|2%2ZSH@f4GV;##atc5OxY`
zLxj|hXPA*NIf7K@?5W}`!wN;op&WyIOwURWaS6!As)zQT%^8~3d&jJ9TSHy)P%)=D
zX^7SRvu}|n_`)mT#%(M+1%n^J$?am+n+f|vISdhFRH0bT%71ekPfg`HSPV>vG<f~I
zezMKy-?r)CrqNFlY46%7<QBeAOvfkncY<>;+yjr1S+aA4F)@}@p^SbK*cyyO`<`%I
z4P<Wuj^6W1b_jMx9@%S#pVNBSf5}_^vlT}D3ikf*21Gs@1Sgk=El2BDkRvT5{Kn=Y
zUQ`}}QO}t3V8RY(sjyS8CSM}B_5K!lUOXc+pgBjj<I(jK!e1sIn3NU0>Sw@Sz6Ni+
zkW>d7cEK8tBwv-~r^zipK`W;Hp#L6tTt-1(bdfJ1VZ_x#+Zf630G&9PEdC?@{_mSt
z|ES~r219>o-{#U`67g?lMSQtU-zcz;qn-X*%TbntiBf<mG`>^3%Vc^472pvN>hwb5
zS0Qch70^lkK161=xQx6xz6t=HgFZY26~C920I4f;hTR(nxW{qy2Yje$s-vwtZ(7su
znE&fn2A8Eow88Iz6#-yvG#If3oTU&a-g-|#&qn?6xan<HH=tw~E46@ZwVnJ5pdSf{
z*Q@kbfY{v`Z0LRAORB$6l(`5Hphp|A^WYacT?7=?=HEX~N$v}u18^4LLNbIhsstgU
z6KQ=KKY~G6L+}&yr8x@u`-z`^229b<z_<zY7k{eP{_8;%=%5F>12%^*n?eqQmE?OU
z%g6n2dH!+5Oj*uAgl`4kY}WP5W7EO~16$N6sN;n{K8ARL<_n-3m#(3Rfx&bo$GwEp
znz4T{u>2?Kf`8gziZb}v0V**T9i$9mL044<4~}9#PrHpAgLM~~og}p}e&}q!D3!*Y
zm%1G*ISy90jt+a!@JefWgh!ir)hr!sPRwOx8Tg=zCO!DuUS~^)ctV>Y`|w#&dxQQk
zBfflm8I)Q?Ppf`-7|;qfq8nVh^b$Vq<8KBY2F+I<hX_RoKFzk%O_Kf`TF+sI>!+DD
z+-Zb;wf1r>;bNfhxO_qN*Nq-Ymu=?vW|<jee>=2<<>p$lq{$kAxT~32-~HTb3Vs(l
zc2L|~*uE`|<JibM`3YED!tOQ4dtZP0zAt`M&&v{BADy`c=XR_;T>(QY4=;mZGsJyJ
z%a1q1XEEd&^x?45k_za^)&yth6hOkXAhNg3NAW`$d~zi?Lz8xc%z%n5*T9(M1aEM2
z&rXN)3Eu`R07&mj4p^%&eS3KSH5A}2e1V-W;VY3d*zLmRqT?gX`39Ut61DDAghO(}
zj$2L7Zl@(baK5}d+h%k8m<L6tf2Fy~-_gv$2+`^s$PsXjhEJ2H=Z(Z2r6IT)mQ`pt
z8Fy<$9;3V3>cgTAn5U$~Jb3eYN5G|d&bep3&!4!osJ%YO>hynD4o{wi?BdrW<j78m
zJp*4PThC^$#gd~l88j#MjWR|>=H>aXj#ijCAJB?RVT)~!7+uCzL&-!B5YVs<L&slV
zArV4_3!1+KM2Ch?F9jR2M3mK$B{DTB2J7fv9K|Ts>q|(z@7%JXp;cRN?D&oA42p{m
z-@EU=)j@|6G!G`87ab**L(DWh0$ficA)i=S&-_lAtYD!T^AM)xpe^BwctU+aETpIg
zyHA<eHr~KwD^-B}McsnSJr$ZyIze5jbfWni`1|O=3U}X%aKY_G>$F8_1p?YSF{-ow
z5_fTBpK6Nf<dMS1=R6MF$lN}AhrREZp7$4i@l60!a@bg<O58Yqpt*>ygF8M?(9MeO
z2Cf;2dW<M=RQ2|i8awt6bP{?#@uScGY@%s@)K}B|<d~qYiH#78!eWbNq1d*VKd);d
z$-D^uqpCJB-k1!Jfn{Nho+uFk;mkjOVI}e!4_liS{AsB?9@)jLa~7h-28#u&@l!xm
zXcZ+w$G1L?^+dhhQ|dnK%*6@=bT&V++5St!Sc-yV%6|j3(XrDTb{A0p!H2j-sJ`V2
znHBdPqD3%7Wf6Hg>6NU^K$G>DTUVNwT3)IuWABSVwF28o?DELz8#Y7b`3q0cNB);v
z`+Kh@9DDfzK(On_gV3KFEa3iEx*(J93f-p79R{#q4gBnV=+cZOC$Ynd6Xs#fVR6NJ
zx1v)-_RmI*@H=_=`9Ax&h+Es~uZIQfQWjW>gODwdZvr#*A199Yp(l95bC1hxprTF%
zn|7++KQ9f6neO4<95a#Wg09Y4rfKt!7hM#Fdc9DdnmxDR^KZ{j#2ll@D~G;R`4UG$
zp9X97#aN+G&y^n+0$RBj8RfCKG-7^SI62Al+zXY9*FF37Zr_}{^6~W~cB<(<|1%SE
zrf-D8!`H9Atf{w(r0Htu&g)Ebe^p=pLN((s<+$j=nf}BPVH^7(pW4BH$cN$W9j_EA
zACUzNZ#SR>PPL|o0uWuBn6M54r)ToZTyxG!e7M9{To!50j9{vAR>0T-*IRWs=-!Q)
zkKPM|RU=$OrAux_Hw|;tpwFQUd%Q%CA}Jf<WNBZUU<*$`Qy}964;m`RMwaq2=pCJ6
zxNcgUxiywbSpSO3J{S8nakkm#H@-VywyWxfm+mJ!jos(=j*>n2_FD<urHBhk17~IM
zC;1b;Q+%NNyflc|A~BLW;em<loPvbWlF^3RHEHu%<F>arIJfnWk2*efPQ!;4CTl`J
zZa;I{ni{-FCCExV^tKaHMG5=@^D&FiI$Zet7_ZY&=&hA6<!PGDSEY~2J1cfx$N99w
z^Pv2OWS_Ig?ggUBP1_@VEJs7<;+H-O3Pv2;$A=7UDNnjAmjRdxXdqeXq1?=<Pm|uk
zm&kMoO^IVCRnqLNz`~iS1jfjYZ>j8U^><D|v(|lf{&wW&Y`fX53m3YcfYv5ZPKst`
zTXeuQ0ZewG?_EXQ$fk8}{MWVngrujH(4oWsFwPUBT7(~@RWwqd&D&WdXXM9b&ks!#
z%L6r+&VAXvB|*IUfa{f;yJ|FzemzPB4{QN{C<8TzfFNDvIhyZ6m?e58Oslq1WiK!=
zF3h${zTbA<xM=BI+pzT&vy!XK+*VgtnMWC9=H9=fzU_Ybpy1u9-O+mNl=KSw`t08J
zS1GDV4o{}ONh)IjZO+fz_v{+crsDBa{9y8}liyJXA0&mN1Yz?s3+5`x6G>HKI`%jR
zdnMi(o!x!1&$)Yo=YZS$FV`yVDVMx-Q?J?|`8xT~!z<A$zGbhKy3iqXARQ8VkI@bX
ziA<`3m-|EI_Z;>mUc`#n$~AmPFsx#4@5kD=rX;aLsZT4XX+N+X>(+F7ftk*Z>pD|F
zD}An~$Svh35#cXL7Pv7X`JXH#f{s+78nu`lY?-5tW+yanmDt=4F}<C3cbnZ^rxjx@
zeqQ5ln-LYCv~@$0cmCVW%3a+<n0Z<ok5mf)b-9LH&&Sr4o#tgyBhdZ)pFFs<3hoqO
z%f1~Z+MW@k<>b{JaHloFNW1^gI)jWk`u_H=&s-utu$7-t(7nA1pN@VJ#hmzcapIaz
zxKYDC$Zve7xR;$6n4iFxSd<P7kBSM+NRcmJI=i~<i$3n|cWGb?TlkatlzApUvAfm;
zq1Vj*2bMxX%<y;9_}|)}fBsbZX<i2OZ@rOtNv-j^Mh3JKheo-J06&yR$vs_Il)WYN
zy5U_fFJE($#j_mCXWA~D^I*xv&A!pkA5epSFR{aK=ldW+9!y;l)atbhQDCc}7jehq
zr_73Zy7+a!k0vmF_+yrIRhlJZ&z#&~_B+edNg2lvKYKxrQJ}m3TT>;=k&c0*=aYq?
zBsS4>aKUrlqpX<BZf!h9)E#!ua-;aTH?yGs*yycSSKZ<ccqf&wJ7k%%<k^P<q4Ozs
z^kY<l5Qc{`7D=kll%M7zTmgcuk9QZ+IvPZ8mJ}nJ_sLnD1(KC|Tkoc<xKR&dZ#&nE
z8#iD6R5syZ7?8vzx%nDfuYyd3aLJI~rg9<TiQpDl<S3SG8lffW2Q_t5WP2*ZER?Bc
zex!!8{nuK&kx|ukuea&y^wAG$3lBWnVR2=q+Ub`EeG{OTPy|zb{u4^{&qcQXL4&3u
zV1jgOelD};vLc5>zAT~-cmd(B{K8<M1~p3{q><>uzGeZLl06>+F0-9;z;^&r>1G!S
znpQ^WgS$KZp!|weo{54MR?#pD{Fq14{3ZpFtr8>%RIpM-^!9l-^pWs>ROaZ?4^TKu
zdyWoxB+|MECut1^hW`zxkrKIzj`{UNakH@$0Hide_dS4k-wsZtSG*{?n~CkC5=rz%
zCd|`}kuQPEo;;s|F6LtZodHlT@DoSzXed|<Yx?Is5d6|Ho0xBFvN}wmPpZ8T9k9Ox
z;r@;d)NpPkv&UPIL6@JUf!988qP<s%)X19%xKk<YlWU;;cZ&VA#L;J~0wLL-W7fd}
zuCDX4xq}r!(8YHagjv(=0K%G$=Ch%7QII8&$Hs&2ytaNSg=;q$$D+5&2wkv4(`;nx
zA*3wl_vt_-sDlhbg1hmhjpRxcnios>hG;7U+hwq|_fdQvlicLeSOzh%F&q&D`Zvs#
za5w?#VmQ1Bh$lbo6r$kL{0t4RHvoP``(}4kL15+9o-Y`%Qj|!2)=+v_F+3<9o5NzK
zDR`9ba38|dS^dWDZ-#`|cgh|XRET>e$E_z^9JDfuI%hucP=(Pzd|s?X2K=jF#!&*y
zIC3P_aftEx_<8a-1o9|zT{cjA8vl3A^}h$xfBizmJK!^3&=oS;piGv$fw*Pe!3|WA
z_a7+-)EmY2#t`wzofik`xab4dTonDE?*;{Y7W7UEV1W(iVS$1He?|AK*B#|{L9H^$
zI4Xsx7wVhfC~gY0f(mZeAH`Tj!T#SGF~w;yd~72R$a<Y5*@T7sI_k*_rH|zJ4Y=LT
z$kUM=N%eL6)nKOFe!F1Ajr4d<<f&mL4L!E5)`pXSHBFH#bwJB|73rCR(Ejg~iy2CT
z6rCT@$MlXunG45Nel|0vF1e8zY&EwvK?C<{G@VzJm8Z(TI-sUUel1q>_wqVc(cpZ!
zDdEGnhe7%?$^bPa5e$~^JLyV%cNQEMDDoIroGC*%<`SeB4Y$zScl}<W2pQ>QdpzaE
zt=5hAvZH$PXT6&d{C0ukmE5w6&Xd+3vZni>g%==;l!El`78oK}!XropU8Hay!p5s3
z!izj?BmD#)+YCwqqVDA4vO0Zm@q)O9Y(%_yG0eVp7ORb~ZdhG@KqWp<H~ZNYvs-@Q
zOKvo6jNW#PcusG-Ik?E-@83MTm|5IZb$57blGct7Ws?e4u!echs(9LQw&YAlpi*xl
z?Ts3PRGkMVH?!Q|bcOAE)cZT-#HGav=L3)4ZrkdB2TR=3`466)tL~|G)ALj{eO9!|
zW+F9g`J%*82sQvrr518J<2g!9XQ7IL>!01hl{1&r00|T+5M=UV2<5a^awg^;$(@I%
zIAMbDjJtZ?7TYTB-wtmOY=8Ok47OHj`s>r{)qz6%REx5F8DIFR8=pxfw7yD@!x1`?
zW&Ba(e8p&hkz9RZ1@=v1EUh`2Yk%)UecJ%5-rsC4Eu+xj%$>aboL6B!i+cgQ>BSfI
z@vt?l0V`e*fb=$eg3$LdBy+!#>UZSHRuu9*Gpj$u5{2OjTQFX6naXyKE=ETAixKK4
z`{TFse|3I!rDN}<Z4+HB_gr#0Npqr*d+sR9K+}fUAh16$9mWO=>-2S(*>jFz)clk*
zZT9*6SS!7dlB;K=X7}>lZiU~!^=M4}hk%G-yeX$n@9O<XOr+Ok5mxEs$aR3%sl))$
zHA-JVZdFw0hLTlascc7}x3V`7f#1NcL;0u4n{o`Q<xoyVJG5nMaPpPN{W)46D{?HR
zhaZcIlHS@QG8@&_Z*le6o6yS<Q{`0~?A%6OPTjQ9)9z5l&z8WX*aV~Wpf5SC(2UM;
zgWcUp)}yMVHrsKlgkMD0<LFiOvPR*I6{lv;_x9a<erv_F=#m0$c7@%i>23$_wT)O&
zqXX?>C?s!YfTV6kTZ;26QPEO(6ijL9PXjvZ5fBs&x5m;s&148Wu88>%DUlmlV^A?>
zg=W9a{PmN^cu&vUt;bkr%p+gP+GF)1#|Br|>+ZW4G-a|T7{X{<1#jL9x9AbK65KV4
zN3&)<viP{VaZ^`k&8aQ3*e%dM`Xw5!$#Bp-!7$*=@x7mcm-GW}knE*5?N(sY{(Aqn
ze0CO$8H)2YjhA06h+23mp}KA4u8YeiS{_gWhJ}3)c?_|&+;CzE5UBDZtwvxqMKPh|
z)Jn@eLZ(jZFYP;ZBvw%-j9cx;wiM1S4BTUq<UK{x+-KbB8B-a`$_J@HZui+0)ILMO
z8)%8v-o?Ix#&m?n;r2n>G8PiP|8X25^=pdaPZbu!3bF*Xbr|TX_0UpJt}<5|09RmG
zoMKrg1>j`Wb_1Rm)$~lDVP+wD^I8_yL=cR7FW-3{n|+bR;-oW~Yfg7g(8vhay{OX^
zssxR{eaUcE)oA_<e#?LYV<cJum+ECbrN^wEN~*#k4tpx0J!eMMt3=9tIO<jY@xCJM
zqIe^_%G69*$y0t+pP7o+g017Xw)qM5j;Gwc5j=IuPd(G8xSK1H7RNq3EW6kCb)|Uq
z?W*pkh?<M+&)FBtt0cd?TsFaP=^bTrfBHt42zE#Yo=W~6R0x1!`4zsqU^E^nt;HuL
zd69<rJgmm1IwR!`S2DZzqvKkPd)6J@b2u{jXS+_H>z8yl&!e;y8tW#~VA(^x2jHN9
zTr0UqKp_<Je4SelG$z{Y6gvsOuv?GV_kMmFj8a;0<-U2;(KWGaDket{6u7UhUv0SB
z&1QGF-6|pI;wyt$WqPYGL~onX7iYPZJn2q9mVDK04`mqw08C?2y<L6=2*>K%3+RvL
z^ExUqq?D=alWQbq=zhR2(T2BT0;{OuU6*%fUg}-v%X;X3{`0z=N5>jL1gt^ahEnx%
z*pnRkPvmU)YJwc-9!r3J365YZ7a@-<^5fk_K3T#yCStQ<+1_%_qCj!{@^#ft6STKJ
z60EMNG5Kn%dvb=sHju0Kq4kFneiOuM1U!5LL<zhMV?vWJ4aOZYM`;>vFEJ9+!xbd+
z4>?mssr$aPJ67erxf^g^epPqY?o~bKzZ}SxRHnUDBAL(<XiI3Bl>&CiO@?w`Bp?zg
zKN)w-Pte56L@r)M36tON_3-;hUz2&&;ZR_}%#=sVMoi3^IY(E)!(TIxq;CfyyA7#>
zRs|g8lv?+9iiwAEFn$;Aa2UMtZg!a{tSE*G$=mn1(S0_rIN>KrSZMB|8@25<=a^o4
z(i~Bzt$lfK&cX)9mgA`_He_dIWG#QG=lRQc-IqaVEb}y9ct(J2PW37_oKo$VRvXe7
zFO5y{0ygK{f@LDvg5aJ4NbWq%PR~9SeM8me5Nnq2s-n?1Lrwrjmr5*RkfYL}XlE*E
z1QXM4Ms6RZ^>%=zfeTR{$zg;4aO{<091ZB?Ax;wgHH;V|5Lpz(fW=-jXt71i;xcMG
zw0%biA<tj#nP-f}gio~hX4>%eF7}32OUItu61`FX=B_0N3mXnn+ow{{|AbUSo*}ro
zFHBpA=EEGjWa+P1S*Hn1z*~Y_q_I;P%}&neskiPluWEjn{k}H`8R4@1gJ65o)+{Aq
zXj8TJ?pI+GcYzej4-T#L=zlli)S-+NHH-KOHslNmq!?yV4I+L=pKLAhGoIO)R5`ov
zHV~5ed&Pa+;JI~_UJqVudQJ7S^Ieryf?ptaCfgchStABySSVe|QYnqI882;g(GoQm
z@>QS5T1}K34NpGqZDbd0tC5uDTRP!wc*Wd1lNWq$j9wGOQ5{;ghqjIX^j-c(2KS#n
zl={sRuG(>{IgqLX{XLu^zr&r$nIO3zA)$!`r!DPWeMQi5#6EoVX?dTAQRVfhri}Q;
zi1@8GGwq@RbV?nd<>E&oWE=M_^8vr}sLaJP&!`&N#|xs5;VS&oyv$0D{c+*51*$FP
zUa|%>)2aSbM$OhQi`=v19j+^Wx&{>N6+Em7!Dp7qb+Srh2;+NvnDa=lp%QvaWW&i?
z{v)Jwj%*7%PI^(EkT;pof5&me+zS?4g!wbzI&G`pKX8i|*~OZD;J_!7oZzp8`(K%#
z-eSF6y%DBl$*O)71LHsrK69qwXC<%5kxbbxjtOonl)IM-G%Q`QdKZgAXzw6K`lhry
zHg9_Pi&Ne5o|TW*?6Ezdy>qOr6GEPu;E^2$f_ZoHjB{ftsa4H{nRPTUQ@eS%kpM84
z`AB9>$47DZII@_h?`NjQW;I#_&5u92o&WAY<D-h+9NL(=9Y8I&IvZSvPb2E>bWFzl
zx*l43Arky^v6l9i4pMNpB`RR1hi^?+3a4Y&^Ct8flYvuvzf+J@=qAl+kad2`q)eL9
zH$`rTlo90eph4;&<pJ>WLzYumxJiF#J|za<0H`&#kr8QqDKA5bOf}qreqhK%9!kq6
zE*g;tI79LRrA-mX57kM!q_HxKFjKUWE<s;x!y`ZPTMkxFSANl-)!*=SKgDy==bpe8
zN+MOowd9N8b6%^pij<Uby{$*sEww{-@S6L+Qx=Y;T%cSJZ&^K{iNt;*-cV-$_?6#(
z^Y<@w#ke1z*FZ7*ouaC=&xjmnTbwr%=^a%W=%Fn4@Ssf82%byZ5HOZfHXG?OmSY)|
zfJ5IYhmh`ziPU#0w&BQU<y6pRMi8x4EP?>ei0>K8!aZsA)MnN<Q^B~LEII9JAB7$L
z<J0M#UZCb37lCkWCJR3IprRFQ-!yq(8~89G|F_&DigL6qwLZw;89z@Dw?XL{KNr3z
zd%*s&Slt$1swnU!3%*qNzW6K~j7`VFtP})uvn8%*D}euQ&8wh1z55Gr)(udW|0)Pr
zHV}mJz7fR~i&rtSi0>4RDFrn`uz<yYj0*N-k4PMaiUQet-K-tR!~w{5G&TMBY|8c+
z{r`u(caMv4ZU2TR#UM2j(srl_Nj4Gfu8200vdE^?AS9^~rD)8RT1tCm6GanksU+=4
z8toD76{TjTP*Ivmrb{z(t>>85TKDf>YgzZt{oKFzeV@<sdH<<TQ_Y;$c^>Dn{T|=r
zDhyr9Caex1cs*qc>?DhRJ<_ky0e4y45_}r%C1CMfIBh3k5_*LNG?IC8r@@YwPhsc8
zAn$(S_HuJX5(9mb*?^euMp@FP#5nS(JdQaKj2`?#c)%DfZ;%Wi>?IJB=8#hfu&%$e
z7&DivP9d+i7zVk%mNc=8#iG8<ApGH*<XAC?wBeg*;VcL4Y)!O07>jISd>aXG6XKU7
z@bpU(;&Bv2%~+VQOfL8)ZsMGrcJ1jegnQY{R2W0l^*xFEqx^mMCmAiHMBB%|CK1Se
zso-vHSX}gk)p6oDe1Zby$r)4R%R+GU7%w2nY4N^B1O?>M*#R@va~9tr-9Ia2)_*dU
z9oItZCFabP>od3bWGeIfb`c_p^QMyzIuORkHT(umApW6;R#=I!KO=#RIVBLmOf<`H
z70PhayY(aT_i^<(?{B{<(P8JveHuIbFt~NA&cnw|`JWP~1}nTYojpg4R@^n&OM0J&
zEBf*w@<={|K|mVJtY!(*pxXUE|01bB8JAgsY4GBnP`Eo7vY+rZm@QAEtX+vFi!qOk
z)rT3$ciRxH9TOszUU>|*1N~5QcJ_wmW!f{%?S|E+%U8jqv{%39YmS#1TEvy9t%%Zd
z)4W(-I?u9a<C4!qDb9;;Er#Lfm5Vc^Vvk7x_)2I}x)y)I$UjqGT_3S?HH%QXa*yL%
z+BPi6NWOwwTYbF3Xa!cZGj`Ff%X-ISE!0<(9G|ziuk6wVAPrdp`{5)?Mf+IvRVXKi
zR|IH`4A99x_@d-9p%qMJ_tW3`5Q4?<660));*{6Hf2>?-;Pg52PWzCeZNA&s&#Oo6
zhE#H{&G@Vmy@OCTE(}8@irma@-%9Vtah%Z#rZV=ycx{vMdfUm^&OJP(T_wrhZYSF}
zUVdHZrto3o)tmdCz9jgPxwXvv9Fg)+b&hZq4p7zis|Y{(8fuPmk3b+^AL%>^)Iteu
zIgc1=A}6%wMx7&;v%(r|vaq-(T?-mUGSVJ2jl2(3TqR>`Q&eDeXzzxt`8Tfz&QuXM
zL2Msyf|%MCcP0Kz>PwLbay6H)&v@I}<@8eZaY1(e<OA{D-KTb|2hVZY5xe_<k7c9v
z@~J>86(U@vJJ1M81fsYXQx**b7=|~>=rvU|7#*|_=#_?K-NG9v6R#y$wb<L!wfvS{
zIh1#mKxjQ~WkfKDhfgN4+Sh;R`ocetef4!S(%(@;Q8HQhwqsAn(~FL89qW4{PwsLJ
zJX~wBWTknn=5g~EQ4v`ww<iS{Im7%*7Ct!#63`i<b5Owu@`IRZ(^ue_0s$Ca21+HJ
zDJJ4qLk6~EP#8+f1W=z(hGY|s%%yhsR%%Zvph(bkv7jEO747K#w?|B7a(&O!%v>fu
zRgNqva@fM~kvieg``nil<yDp>TlAk@EYc?!;l#!Cc_%nSgG0PHd)^QP0_RPpHZL`q
zWb01Nz;epE=5fD9jB44RoXzb#(taju#*3XFhGS}buJ46><`$?tLwH3r^0TcnSp`SE
zfp!dpGpJ<kT`b4xA#Z1ZW?^H=+7y<%Lw5Pb+nstjvW)a~UOtPfQxp}|_a+5;U5C`X
z8Mxj{@h0X7m8rz!wqYHRq-~M7mW9``$R&oUU7sMSQ$`LqSP+@^q|G2IJCkO}^SqJv
z+Hs!5XI<U9BN-mqH-CDfmuGqYp`r$|j2LAn#cyp>h+EA+dBMiV*VkxAYhf#|q)kCV
zzPM`Ibh9~b^W)uAU83}M_@1=4+XX{c-!kuy4laDIh`J6%F3FPA)~z5ZT|THWH&Ac?
zWDTk5$1F7T2+aTBLj35(IprEgBX`$J9vQ}9;G<%jP~&dJ%xy1|{o*W{4PSR9U7sE3
zrYG7W?Vl30H*5dN^V$KME`||$AuF^LhJHkUTcO*20<WzDe=MWwCGba%NL2VWccT?(
zFz<~uPr3$9TkrJ6lGC)rrM%UVoR(6*;Qkqn1^%_`jyK-j-8!SnLI;0J1pi;gTA+tE
zmI-IU(J<Idd`|;$xDyWgAoi%731|WN<}wo!Cf98v-GhYd&{zBUZhy$FDSy19re;i3
znlMe%I>ROz+i_-_U09t)Wt^_(y>#9h(74z2g)4>zwIP}m@G{|J&`9Pt8`C+?nwT`N
zqO5ph{dprNu5Wk9xe;w0=FwK#IyN%%@R<GGFFy;+Z)Wsg;otjgq<d}gXQOHCsf6KQ
zv243AY4Jo1b`21vg;2(>9v%^MZoost$d^(0i^Ml@Gk?28YXdQqf;X}93NQQnq<utH
zaZ>--xrkwhuGd{cOYXVs_o*+I{N(PR;8x%rKjqf3pDa^0dhazOtV^28;6nS4g-C@A
zOmk~-__X&g#fioc;YMI@x4h3Ip8NV(tMasLssbb3-9hc9+nxTM4ez22>hS8_gPty$
zav#jJm#$On`9OFBTHrlgkUaWi;X&Sza8`Py1EiC2d7&jQ3bGB7rPCeK@;ImzcG0=1
z{R7Lq*fHFP*);MgVVmBP^|jM4mAHlwc&6kR%<t`^+NmBS1etE4(z*M!(3FdWmyrB3
z*3Ex)E<L=&FcUlef#0qrXXWKKDaXxSr*iqz4b&x@p59#_J0(eWMSvl75^WU>C{y$A
zpoVlw0Vk2CUSI5)^`yy|`!1eJ>#|Mwo)b<rDZ4u#=O4Ek)_5^%H*R;Lf^eGny`_*_
z7(%7B+-*Gy0SG(8_cSYtK2vDU^^BlT5}Fk#6@}4ebsJ0C=&y9BR4-B}zI(&Nqu0`0
zBm8Ao@3|fIkXN}eB7sGN#Le_`uXNT0)nbOXCX%^nL0NurPE2lT{1t<xL1|q_#<F|v
z-j-b9(9fU#)@jc&yIomZmlNcxWdH!0^Sy=5EYN+!<p125we+hKh3#53sJr=IpmpY;
z1#lmB3$&FD4+r;qF1>qK?>xIGcvY>@{o-3u*3)26wnKns@FLIoVUsmt-9E&C(TPd9
zvirc^m>L!LE2mVSZr`YU?YyLB7=ho%Jfz#c-tSf-DF9LY5PZ04*P(Dubfjj~t`s}o
zGv&Kc_Kgj}7C+fv3Y;RPM7a2$sqi0ntw_sx)U!fYwyNFX#I@YGP$T86s~QiLY{R^b
ztxa1Z^DCC`kyhQ&-ukX$;fJ-f$I(K~2BI!-4ng24ZdZRM1E{}&OmT{8D}c%SA&B^=
z8ZSh1|M=+-<6k9Bdaas<4t&Yc`=osPn8J$(`GKcaM}?PK-neHtjZN2+px|<Yl6XKd
z6j`zntO|UL=)>fZiA^Zi2@(e5T~t5&oBl#bT6z(L;FB%B958Ph?L?7;H$$l4fI^QF
z!cl8L`lM4Jel2bmH{neT_t4mscd;(7`KZnzLQVqJO=^XQ=VX?O<!g<MdgN@9%y=DH
zyyEo|QgRo}CmWZ2TqvoDq3ieW&i7b!Ep$nZZNBEkXPQ^1XeL@|F-{X~F^hcC*o1xr
zTbd92FEa^fxp5NOpkM{2$p^M*4mi1gd=T_Pkx=Cgbz+Wiwl+ULxP4AG$Bg3?vrcKF
z5}<rCv8@v_>hH~zz~sg&QFss(46ocsXZex&Zv-BZsjRqero@22rex(n-G!A!FC+QA
z9lRRl0HlQSZ!g7`n%uwoc<7*Bt+ZRfTx+T;by|R0l987N!9U{5H@ivWNNFO3Uxh1I
z(4R5#E0R|f!_@41&!1Z!JjgXEG+Q)hZqU=Yy&4fv)-5l=f}EZK6r(k>1*#e%7Vx4b
zN#U$Ev~T4dID#|dFHE~~=+KhNyyZ(?R9-oxr5StNO8muEOOF8YQ!CVKHlo0vGyoU-
zfN}+6Y$p`5KZ@^ch)~sdd<HKM@*U|F^`cc^dcI^skv0e9@Ie3RgQHOR>e(JweCG1o
za9+31&W8nozS5$LS`w*|PgU1Qk+KQsKIGpQqCEFhw<}u?<TVduYwjyZY;T>ck~3<(
zamEbe(3;$jNeX=vl*8TqA4KTenR!W1cUaV8<#N>fxuxQkm2Z^@uyubHQu0*_*TI;y
z5qK(h3r^U`UQFA#1eFk)H$IycdEYXU|Jd-u&RzDGqV2#j<n2i!d5Gd?>JO?4kW9jJ
zmNI>2rZ&M$vgfhX^mXjOHFLQZkSp#yQGu&9p!d9OvDkiOcycrPfed%V<QTvlB!#~a
z3<zz1UB`cWS-?<E=AZqAVCtj=F<&cVyczS4!Y+Z^Fu5mXvI`g>%iWh|U*UGFXHKQh
zxE$7{8NAGKQ%m^tm~(bkACIm4W=H(u0*3`Q_q&w>b}vryWR~V~k%7%ebCzaLOVsAE
zPk0#JqgKaM4=vu%8TF>W!$_P3GqJ@iSFy%qjxaH@FAR<7lN<xF2{Q21YGtxhaKnOD
zt+?%ll(QxaVm!Vos`00H?+cOJlb@k-a_^%MPe@oa*Z=84(k0Qdm&GPAuw_y{qO?N=
zO(TL+m#}R|@9plWQt!3zbx*OZFa|)FFnC;@8N0A`fz>RC45VhN=*NA@fVFst5%CSS
z6VcbHoO;p9H+Dp*=yLJ|bob&Gw6JY(r^@NlwDSx#llhC)DNc*;M9Y69EWT4mARHsi
zQ3Kl_gK*};Jz4c%_k_ep;Hgga#`ofrKM6=SUP~^%)GEJfU8|iRyB2@dGHLhQyyZ`-
zt+b#L-M*Mm|FB6f^O@17m#fP^?w|0jX!Zr01+UmL!y5(z45sn0Faq3?Kc6O22gAG$
z<lfH}Wieg6@8=ihTVQfMhr9~|7tXrs*WEIqcE;TOL`C%gBovxqKA`Mp+S1BHQApwQ
zt~c$ScW4W_8tQM$W^(3WkJl}pZ_#eqGnYgIAEYn(ub;KV(apf=Qld}d4Hx&NQ@>r|
zUs$)G`IT?YjMXHs>*VqPy@eAL2ie7;Hy;LTpQj8}>A7oq8Y`--xW}K{Fp2sny^g``
zU-z4bW6&}LasQ^?r1$ilZ2)e3Al6WNyZ647O=j9l7H^NLoDL_pE<>_`K7&ACEL<Tq
zfmBJJG=VfXsM}b7ZWKEVgvp2XGo9R!z_G2E?V-GB35!IVZDnxgrY1wCKP8qH6l<%+
zs0%Qb{*%D`PnaS|d!uEX*;fiFTzkVcM+dV9_d|_`#C>1#$j>FvxVmXi;nP<NS`Vvj
zIi}?22MOpchES0NL(KKXfTa65Ks~nsQp>6TWnTYbx_+~{|2%zqLP{HdcK!Re6bHv=
z0Vi#|rr8EH490F>V;r?`0(tO^79JSyi%;r6d1C~mV1zL?)s2NIdx&or0dH$U`xd36
zY!Apwkdq_@6fkW`X*a+-Eo2Gl0i^$f|9!%kW$dkI%2_cFo)<6k-uE%<p_G5nw=a+i
z``hVvMHA0`eI8eM|Iq`}W6sj2Qti%2TPQ~e4v@16HlmnJv_W?(BrvoY;BJunz#>hu
zXm@JI$ozJJeRfqYItwGYGN#@8Tn(oA#fCaVHpn(o5>E*9&vdt!IpH<&er|dGaLtU|
zkx3S&TMQ2cKvLhm9~qh1%;c_y_%^yh>fT4CGZp6thX#5t1a3I?woK!K!G%!mT**@6
zOo`(GikdN%k(&24w<gs-(^#w?z2oG(?fw^|vk6=ObWo%O=24HGo2US}xMtrn!asz`
z@b)gwexG|t>&e(_g2Zx7AY#1O0eRh77PrUZ37f_dbYfIkAcSGSU$GAbc$SKozrs8K
zv4gu5fw23BfDykYS}@B5;9Pe=ncg@8w`dDYh6b*TtzUV0c<l>-nH|0ofaDs8-%#vI
zg}L1<GSeKcl5!zD(+gFaHd6c0UU4vXrRhxI(fNb0Nq&LSy7le;Sa6<TV>}Ocw;VkX
zgG*u&>K4A9pKAtB`%cYVIpYi<Qh-v!A;9>NnEvtN1rdWeP()h<8R;)zHAhcCszDgw
zZ7Yy+V~$zK5ZoIk`5!6%DR)Q9$k&6KD+=D<JE##jdF6!3#_JSSznD%D<$#OWux$$*
zYzGxbOcj!2*KR!!29uK&tuzjT+}uscBpve_f&4!4+fpEKK*RvAFqFT^WCq%w`BM_<
z!<_)6sTZG?0jp6Lr6U6+eXOwv@KE?DOl&9;s&U+Wec0&hxPihVh51H1dtTxjR=g}S
zSXVuyLslHTfQZvtCS>45!q1+L#wGTQ_t@iw;5OA+9%@F}sV0VupMsrCB)J-uPZvCn
z9NfR;5NE^1EX!Rm9DX%=832MTekpju>ud}mTJ&IIA_9E^*tM}J1|pQJ=f{Uh3l^og
za5h`adNE^db>^EHJ6teQJtU<s{0hR(qD_SUie8{OcwSSHBigKY8kXdXz?Y|?K-x01
z1hW2V{)-8pKoazT*{)N2JK90zXB}v!c0xI-@>W)%@5EhX`Hyqvp10I^ORiN9Fxh}V
zeZ}lkfL&qFsu%mhB>o1mRoj1sjA<|i3}-ezn;FVYo6dV45tT03X~&5PyocB=l;lSR
zEVjwm<%q<2opGZskQn~(Fj%Xs@1#-2YvZV8{@!{Lq(48bID?>88hwY}isV+t?>9+S
zp-eD38f&#TZBXMxkN^5v?Qb{VsXRXG{F-;J3j&ZZ-9PX3d$4Z@J#>4b2k(u=7skiJ
zXqCm4ZS3|X5k}4_Ssq^RLiVnmXFYu_G7x~?`pd`L$cWC>emgnZ8aHF_jSJ9Z!1gh8
zxm-s)5k@E)_v8PD6os@Nvw$RpBPUvPNf!>_qg9&Nn-HqQS=e$b@Mpb&%&*GL3nd%+
zCIlTRc35uFk|o!2;c43y&ZVfNpcNM^j+u&N2>O54%1f)3Y`gVrTJB5bGjG&=Vw5|b
ztQQHV;M*nU6VNLPLNyKWrFZ6mA(7$s0NVgCw7duQ)JAQ5T69TX0j4VOkj{MM4kl?<
z$)xSJMs`;u%O&S;h+SZ6>@^0~oj7Xv=Sw4B=6}%p1#tKlElFN-&vs^3a@sBnjSEwH
zv?M>7&dw040*x!XZkorsX#}rQber$9Ve7}bs_mrP<hp=G=pS{UWSo6dXJA_T-WQb<
z?MsT5RE`j)glZ-RkU*$(e=SG-3cW=CxrtR$SdFj;VQM0fInh<2Ps>A(90KJLd;z9m
zhDesGb>1MCI~(xyL%j2RR%NBR2klGwc=AfxDa#kJ-wwVVt`R&T*G3Chr4QObj(8pq
zoMzdYDD;~@1eX8*i+|V<@&F{Oq|#e`(CMJ6*CH;FROeT#WZZYR;6%SoLt9@#H-Dtr
zR%UVRqKG*CxFXy6RWCkU)Gw*~+zHUC1_RJ4bD=@7UQCm|<qQoB4$!0vff!??1oXrD
z(n+-0SPXo^dtO};Pnx4uK~ed%On96(KgA&@yhIM@N{%dVubI5)Xk2mWHUDGx=B?dz
z`IW-GwJ>-rGa5UP^TrjAWU?u^QjR40#n@CWoJ~DPmd58;oKdJiL!2jbrA{Z*EJ!5>
z(R6P4x#@PzE7W@Iq*Pwz>AdQji`&oVh*M3w0}RIqxG~BeCWRt$CW?49InOhl+gn3l
z##Ku_!>ieq|Cq`34PAxGICr+^1ZQ`8eD?JAf3!CJ?4Gz?DJM-U?*aJg%O$Ae<cn@I
z``(!t0Ep~05A>ie^l$6JbM7`lggy|@3TL%(V4#5yFQJ<eQGy`lLt#boee^}V(J&v8
zy4&E$W8<WFC&TTpmS$^SnXOr+QSPw9YW);F2|S}k;@dWe&dPa{9&>;bs(!GoTX&kf
zLr#c~NQ+I+^abSId{Jv$rdyG9R^OeX)duG--E(T+a%bSyt_4x1a{}H%J2DroaG<xa
zMvjwJQSN|W?=OV7d<ln9e!3{FK-(}Iu@l`aUCr7>aeV3Kf6$j_QC+mU=wy*x{x+z4
z)Mn>zTc@|Vv36$IGxg;=eo74@`04ysryQh;3)_&h-%f|>0klDZ1xY;(?59a!M(3bq
z%`CAQjVALK(~KZewTHyCK@0REGPws~Pm(|NxQgPg0ff6z2N3bzzhSK<ghG(Gx<ql(
z3Sq}kV7Lf7S{@53SDTy9vkh=5Xfs-wMd@>RTyo-wulC|~8U1my+q{<CA;h>`GdmmH
z4RCKWT&bC#2BpXg*e6KVy3`=T$4S$&sr;*=WFwmtR-cIm_JtSi-0;=$WPZ!Y8JF&z
z5vz1`?Fx=dcR0Uxc3<hxX<D`JaC&58*s$!R29jJ7$rI3d@7^J<LjoPvRP?J+13(4d
z3L%1<t16F|DN*s53R9(;n`2@#FNd^$67Ow3t390`-8}LOp<SWJBl28a4LNMbm6JQr
znKX9~H#c``kJFKbI@#Hr*4BkEqik1q*!*2R8>fcNBqWi+#((F`#v4k^Skv9R;6ep%
z#>*4cc_sySuyD$&Qeywk(`iN7tj;6xg{j`{b@|yR&SW^(8h5@O$hdPb(M@xnsZJGv
zo@w{L;o+zNS>8-2PBBy?C^Jk{{WNNL{n487N9ilLg`wTs@0~J#<NVAgLHd2+DTVhj
zH|63~mTQC&zU2J%qLOQwLe(Jl&=h(dnQI$8(XWH=&#gYo)vwk^l}e0LI{k7uY8UrG
z^SH0U9`41}H%tmR+D_!HSE_$nIY0AlzASyj6^1x|tNorx*ZU>&L$Gf>k+VNlB(B8X
z@LZ}A$&cUgUjom@^pc&~{R;Hmo#b}!LF*k4@)<72OONX24=+2FL{c4hV*!uA9NvRZ
z6LXRK0Cik&v?=*>FPr>L&Iqa@`{P28k@|zCsvZPClbZq^{H0RC*RJ$Ccs*lcfzBqp
z+wsvS!BDV?ba6>Z@r@T3_jK;5ziF-a<@tcK)xns~*H;c{#uL(kHQ}-V>cNdgLTNZ+
zdo~q!&5&$2g1NS`e#pjq+}@u@RLK=4YE9k4dYvTG=A$E#8n!|6OA>Sp%BB6ScLmL$
zJ=nM5nWn@=Q)Nh~IRWjdMNLi=LPFSCP_Y{dpf3~Ja<>F9B=PB;X`viM6=PJiZ|dCB
z*Jk5&yqJFZjzjyD-n~k08`Q02rStYe+D0GBF>XLh()g4Hk6n@5!@!*zh@)U=C!IZ9
zj11}vmgB8zb4AG`kCC1ak2uUo6WOK`<1FEUydv!KoxamfZ9#YM-t`{Tu;WEIbhh4b
z@j9OAwJJO1!aB_-oBHa2*8J<}_a7W<0xenQEg-)MXZH@Ecjzx$PcjJw&{({gmmLzi
ziWIbLBe!XklAF)XeNd3s>RF<!7H*R6?%^fCbcP%b-%1#b$)InS#NfVsR||uy5M*g$
zDD=R>L5r??)Igt0qhJX=XQlyI;M_?)bT;;>vP-?Gs%3TV9M6ft*|$5}L%X$>*S-%i
zd(QYAfTkOZl^wC1$=$~IW>GVEPTWIV!y1o(DykuMz#$7SQ-cw{U5Nt7%z^~5%s}Us
z6;eSGXeCaCkcGNAi3EfbyaRA&W9cpA{KQbfX}mSwJQ7SpkNL$^H<k+Rgj;%lF|~3?
z@y)JX?G<;P=Q=t#IJz$@qe{6C*Ua2{Yav?#m=d}vn7eIz0v=`7lY^^9s{nc*Soqr{
zahVRDZD@%&!4)RDXk^htn*IsQi)=FI%qAPNJv`al{#hSKo0i@-$n5-R@bjpGuk6f;
zGE3@A=gd*DBMw65^lyDeWAYQ$oKhXyvZB%gjLI|YCI;vfSS3Gb$rdX1$e<O=DD_iM
zdD)zr9htE^^+j&Yu;xYW;`HZ}2~!C}2qzmkLLFO6ieh3XjUY8x+fKkCeq_JMi4gm3
zTJ%>)!Dj|QZsRHu@-8=~WQn40`fjYY+Ib#tRd!@csa|oIb%Nm;SG$t1_EYL=j<r*h
z@5`iyMc5I?=8Xf$$4PonBkOKmBA5+zXbY3WY@J(;R(8wia=eR_4n0gdvNPU1C%s;D
zd+#lG9rJQ`|3|ym=c*l8sj)W(z<a`T%<rsntcE{bbQ^Iy<H)ZU#d*Kn2>{_&x7~_>
zB8{%nPT55f+VN65FTc;~nt#tK`*U`BYf5nU6uV2Yi?%kl4DMYAru+3ccKSDJ6Y?+r
zM5^&GW;XmY6Uy&E9f21-8Tj+>m5|(>K!^el+IKBOa%4vU@>Cv|g>40r{$>URWUU{<
zR6+(cA|vdVRG^lGd2f&X5`cgL3OB{$(K-@A!i*{S=5W4?xY;NRB9i4ErsNNirsKFc
zNC_>9#&&~{CwH^(`22RTXATAgVGPJF1l`Z)B2*v+9V@~UN5np!!hfL(B}LhAuwfvP
zVbQNC_)~pQmpNOLNG)QMs6oOE{J|IG_25;uA&jvj50gl}4v?w^7y=X4e};Bcyc%Sj
zN^)SFsZbnGAU7zO0gBskcr3iSK7={uT?~nx<>T6#KlCO6-(toGX!!k3E-2P*FDF5!
zyuJcfAkl~n$bkdUuc9ZBb((w==qay>a)DA~5OHG&bdlM}*EOwuk{>?8(hlcfb4P`@
zB>7VCMkL4+^bbK&zE%9I`SqBgmUz{+_%n-gThXc`+a-jrK=k&&fm|jJAE4y-LbseU
z6vn~<%~9VszUIG~H~3rj?gw+rzl+U#F!V$k`kW$!q}%~AyA8j#G(+;67dX;zgMu{C
zcg@IAra)H>grY@l7+5r)6L6T^JsSOO9?Y~5T}IW=kMn>gaJMP>6-%fHrTj%ZOW;Xm
zcw`6*)u9c-czipto%mFcB$yF+j|npZ1<<8216)VH2I5T`$X;wVbWT232>gZcRkK;_
z``~iuGc9Z%Eb2=cWh}D)l7yIoRgKKuhJLfu9Sjm92y2*QzE=WFW<yqBJ}lM)eJ@MC
zPZWj4Gw>TgxiY#11~E$GBsma(%s1lhObjGFB^D3|q2nR7r~vjE#wMO*6pH0u3jOHi
z3nd2S@n4GR3W+2b9)ft<3g*kB{cT%7$KFl)Awt}LYE1o`%P1kLMaz-_*Jm+sM&Lr5
zC(8cHXm@98oi$nhF;NzaVRx>q&@a`!|KL%A`Mpx^lj27OdCNK-e-87H?s~Adk~ueO
zBLUsZ;`T&in!Tbq0NoYJvp_VT11pSIjc~>ZJmLhKflnC%h|)obKiBl&Q^0*TUn`y^
zBn9RYpudv13T%kfb?{_WFfF)&{*Z+G6@$M9e%gXIM$zP9iy%u0wAf-qS%?Z+riJv?
zm`s7Ym?(aE9GV}6urD%3To@qYTw8R!BMITukhxa%W6SFUQA`W&AoSRgzq8u4IWTZ*
zI+K0riis@m;9&ZI8Fz<tCL-frw4*sKnx>{4erD;%T?g0us7qR?Yu~l7A2KEPB=h2|
z?=_m}3U@SmXYR`Keyq`L8|D3Y^>f08)hB1GX}!`}GPqAFR=ubwY+<rgtQ<iTfcPIE
zG<b`0En!2<qs6_T2b9Yo)wtjYK{V^{)42j)J~%Cb7wB)1`;?%irE2s#G(V)$_UEpt
zrLSCDKbIJC90l%9CLTN9c_e=8iLX1k^7)$mWEfOHWTF*3COZomp&?M1<$8fIL;DZ@
zmdS1$>Tu=X7Uh=A6F55K%@k!LS6<ksvMx=oM^5k|-kEm_)a0@=X;VDA^M^gM%a;t(
zhAzA@xg;UF3|){=Q6pwT**u2+0_8jVFhu4$8Ir=r49PKgrXZVf6qqzJaEay!G0_h2
z7d}~-st~Jm&oN2(YMKd;Ql1l*IJ?=%o%^LmX-~XvxRG{N^>F{`50{tZ+RTp}*jnPr
zExA|u-oD|zLQEJz*!`a#`yqD<*NlbDmr`1tZG_w#Ewg+dvYB~B%Sr7Q!f`9rth-T2
z_rX`YCKy)%z%{x?MBdCDszD>^<i?|%=6nz7ZdPR^*{XT4%+X(U<z`+~-obsY6Jj@2
zxVuZ)ZJ45|zEI5WrwK{`tmpq&>KR`GfK_rwG(T4q6+I9G?qw~nrkX`WIm^``SN(GW
zE+<{<Z#x*Bi`Jxf7=3b&#LZ1?ZhEnd?1(%U0QyzKV(x(>&K0*3msN#`dvC-ys0S_E
zh4x|eRd9p4E5yCA&{OdpIT<ELQMlW4u!RbP+T5Y-m<_L)pMu@sX>(c&yxe^5V78T~
zFQ-;t_BRg7FN}RT^Q7Iw!<vT<sILn=RWA`>GU4AcVmUB*C{Z0&3=PCOwB0$MoEP%Q
z1G~A@9c{vmcUYfJcir|)VHvk3-}&gfqKm@Xv#e#i-ssM4)_Pg9*7<5mfS;rMePu%I
z8!BgsMQ@7GgqvQTbB+$8H;en_$=IILx<3oGlg)W<r`0acy#G=)^3qR6Q5RD3HgR^p
zecy30o-ou0gDDaA00U|URfu3d8Bp}EZbL+2|1S!`J=<_uj;Ihl!g$>aEn_xA1GT)o
zYLw_%;sPyW4p=;Ovh%vbW0kWF_FZYxW*W4!4a&6Yu1W_%w|FI0h$!Pz;Z!Ut#6wVt
zv&q8jBvKblY|ZRpRYCiNw#8s{ck~-zR$tlE>C4Z<bZ*gfTWKDtG$Sn2=rki_pu|4i
z>VA!5u=JX&0}6$;!_^=5Cy|Sxoo(m$?~w#1{vgntWs8~YL}ngCIHQ?=7R%#W^X)Mw
zZg014h;1-*iy7_a46gnxvvZ4JN9uc9{;30{mla+{zdxD`Dm<EiUUedzsRd<T&m{%n
zi|8!MeYB$fywb#Iz(e(rB1&gB8d-BXZIo;fXaN3j+4Jb#Mzz(44-$w|Kl;tlxbcMi
zi3-`u@2UL>_+@zC%^)21eer1t*>03v{jzY?ZI;ZuS9%?<^};!;ZivoR`kwT)yO?od
ze$j)lMJ}4tRVC&^VY$W8+R<Ox(Oui}inODlpNGfIEi;t;{aP39U-!l(a+bntyNTi|
zD!2+A+hDk*IE3Qr4^jmdm|;0uNUc^SRhZ1dp7jjMuWKAAaM#Rh*2^}m)Gut^vHEGd
z|NRQ8l-ta`d8&6-_9c;*fA<<89E`=?K^`+9N21soxRB7wOXdoprGz9Rld73Qm0hT~
zL-`>Yn`=ApP=LTeY==os#v1%P@W?;YbhtHx+XU-bLObjnJE1;|Ia4RCrDUrYhce~J
zG6%mC8)82+`7gA&662wFp?z+^isTC~-aHQgzU7MVSVe&{-%SKD<016*&^jI&g1tl3
zpe{MgFv+iIdqHc7MJi_A>h9>|Xn()Z-O%w<N_;mvpWFUueT1T^xdw4<5-Bx95_14O
z--Wyj#w$_2z3>!ZGS3mv8wUyg%z-b0EwodjOGe%-dBhbBEuD{T?I~)?(KF}ipS!DG
z*pjfRWv0&i(AdMSlm#5q88X+_uz_pmE#t2{Lc-|!!dW4BbO?-|m0f~Bi#hRow?mdv
z&E)RZ#TM09Ql|KYJSkl*x^JS3g_k>PH{Bhvl{@SZoOy75RAws0eE*}|;*HI(;uU-5
zt7tEpdtJ@Ef%M`zVMO<5blARh4tQwmfwC?o-UFB<80EU(G(qbk1Brg%qi%h?BEVQS
z09hf{rc&U{2kxhB<hh=w`Seg<cW{X3g^t~icMLi5WFKYp@4QbrQk*>!?D(i7-D#bD
zTH^ZMPwf@^kYWkAA>TU|(KJk-OX2tqeB%yPu%?tw!*ngMGceflf^kkriB{5juA75R
zNz_5V?ESG;PWCDLYrG?$X#t&=@%n$^Vv^q$36&6m6upL3s9#QQXYCfHl}_-J?@_zW
z9tu^HX*OEv55U0ihxzW(L(`yK;<!@g)dS1)2cl2G)2xRlzk?Q2nZLP|lKqr7nfR1Z
zffBvuG$5->wh30!>$@gkhB;W#rou)ei#vYntGio!cQ$2xb{KB29!x8KATv#W%OTTQ
zmM5-X?+!4jfe;DBW&K1crEBTWS=`M!zTG(kqdcuj%2a%wqwoN>^Ojk6j`Z7&9(OEK
zkF;mEI9v{^U9vJM__N;j%O?*8ew_oLu{vp-90~UpoSS6M__*6}613J66NC`N1m>e<
z=5!-KRU_rpaNs1UqNvdfcsI^A3Ab{LPi)Ie*wlJdIxDz!c;Uc)-}BFuOlEIfyR{z5
ze09%%i~;7Kr-y%kQlSJM1^_rjcmbMS@aN0`h%jIV`sXW9SN^TMJpOI1E*X6q+Czal
zrG>aoDGp30*GE$zVYd<l6B@?E8M|ClHCt2iU=7G%G(;7L@hS?Wr9pk&I+R3ebb$Z)
zZdF2D#D<{^D7JzOY$mV{pj?DN@KylwSc!lyL)c$YKv=(t284B<7-3xi(_~k{M0zII
ztw@-La9l;P(c$RW5piU6csvS(#Y<{bl7}S#x-=kn8|_yC?^ceHUoX5KGAIeb(WCAn
zJYC!+7D+%0m_j)UzICgZ>kKGrdhyHuU2hn1>^SJbIEW{rb0VQ7+Z9~N4fztH&;}xR
zBABZZAJKU{(FhoT#vAp1Ez>WQ>|h?Eep^d=Bfx8Y%NZhDJ#lRAi;$oMOxysmj(Gya
zU&k6I{nm&yk$}d+yzq<TIe~O>PJp9@#}_On;71_G@OQv-ArLSRk%4C)wBIW13j~rg
zC2)V>H#`ALj)x`xiAR<6goLeagCJjEBNCVR)5nhqy#ARWpq405{p`;HsvlYK)rO&T
zdVD<X18)f|+Ylus4h@1IeXtiM8KgjmA?@6FaENCk+{609Tf_<`(BpzX*#7|l1sUUQ
zB_xhJPV$40R#_ki`@NIbi{FB`^ZDm=joLYGG`^{*f-G3*42RNk(MF6qz%~LrhrtOE
zU96voe$-F1Bi{VIR!RJ}4=w9PhA;agg1IbNWF*!B@Dd9p#SbAS#GS{4Q6`sBkScLB
z-Z@@DAr5oksA%G>3Ju=*U@t`XiB!l)rDli^>_^5RHvHFMi@%Fw|L%GrSL?C0&G^)9
zVnXUwC`izoM>Wj*HDP*GX)GL?>!~yS94p(D>Oy26&Z>(^^K<UlmLE3f(3N@8sc$0&
z=Pudp*F#=L<eh<nhXY!+7m`vrehQFE9t?rZiI}_N|FM^e4UoXF5UTWoUZY?+0$Z9S
z=GH?BE0uo4&4~hjey#GZc?CgFPVQ*l<EB*i(Cy%vY1^`!^;TF>weLKg6*gls5meY$
zo$vOcet)7VQ2=h-zm<d=&jtM*$&|?q(PM;totX#Ibi$vg&A{9QTm4*j@sM+Xj$bR>
zcGID+{q2|v_sHmKcb6si_pDRNUUEpvUhcA_Q7_~^-y_2HaQUSmA8^a82iQ(egSJz*
zBfg^KrVFP#;!`>VlZ3|`#S=E(usB0uCC0p!W{qEgn6ssTqR{0&JEp%~xO`nkb?Y|A
z>WE>BsZz(^Y@V_5x*c(#^Y_NM1AalcsR2tAxJp-~5icI~%D*-a#YsW|inBuu#c@s&
zm&Upn0W9uuL^w;CPsNP6qihNk1Azg&*@R46#vRN#SE7~P<)`+1>|?WK-l+q@il4js
z9Zt<kJAPoU`~}0Qgs)62tQnUk0;8}06DCGIiohHJB!h%40KhMazXh~F5B)iP8hj6-
zv+$X%!ZU)+w5=^)UlOZI=X0x>UL0ffp)K6Yl{*adjvX$Hv7JVeDfxVVO-#hk??z@y
zjH2J)tS?m{K8Hl<r4Um+L{&;O1PkDZmwk2A4Mr#7n%Gh#jJ~K*xNmRbYUV?pcafQ`
z{{tW8*237yGLKe>Zfw7Mswa~`hmEM*!Qk#f#<aUY^H%up`*nr{c<WLD09CjW4IhJ@
zjutxUVux<tg09v~h)!3UOvG#9dpf=z$e;zJc6w(oT8pft#&sa&5@K1Ko(hYB0KD%4
zHUvI~TJmqiwTn=$1#?VO%!g7D0=OF>t|jxPQiO6tc&L~Mo6Gr#06Q!>3gtSJu(^<R
zy2w|7t@EPUi*5pOAOtzi+6^WP{3w-b)9*F;=-5{4#}{1RG4}Xtr9$TXORZ016{1#6
zwO-o%bYw5-%j@wlmBnR*`cNif=@rh)8&WM-F{gYQ(+~AN^)A9%#o+F_jE*DmwEG*5
zX<B|B+;=kKy20&j5s(iJ8Dz3Q)??bD)Pl!ZhyrbzdSS4(-&xFmuVvKCT@CX5QcHVD
zq{mbq{mu>oW#C`8^*?-30;`4j=5J6Hl&TwLz=W}68{AL(o~JKxk6#_0^ik<bTxwN{
zgH`(B>9vy*l-_I@RyXo`!FjuS&-LrxNu<k%CBM4x#bWr9{uVeLSp>g4PWr&3o@iOj
zrMz?%PepA}n}<*IG)MP~wwWCct1fNZXw+QUIQQ(E&3mWY0$Yfya3fS`CI-Q+eaxl^
z^$W=`mV}(o<XWow+d<55rGOFM1#Wg=##83Ze8!v1&5Z>^=VJVpqnwEsty^f@y!AxR
z*3XvEo7$e~<Sq_f(5Pq_vWXyi4i4r|U}vJy5lfmcoG@wYuk_KS@LqTEr|3i`yOa4w
zutAiEvaiK2P3AWDFv4i=>B1$izA_b#9LL&h63PvlPQ|r`rgqIQ*;+SX;us)^D?0Xs
zTo?V_MeEaBQSLE)&|W}IqS{^mrl*HlNuBD>oF;VRBZGtbHe8RiIA7%Tp_Luqd~@Gg
z@;olQ<D63QY7Ic-7fQl?r#bV|pXi2xOZ<}iVVJdrS6_k3y~fhZN83);p7L<d?0@mn
zL$CGj^SoPE)<#}4125WJa<mOs&xeO9182aQtBWglkbwRhTmudaIn;r@J<hYO(hV-1
z^9tATQ}@uSq$AvuL3cJJ5^^KoHy5CCZtZ6V+~yX~R~@`N-~QN&l)#nZ_|~2>h4S`e
z=`3#DZNIY5Uv)`+ZKGR~j}*N8kl^8EII;Jz+kyR){8Ubus)9=ttTiZ3tZF?Hz}*Li
zbx_2&O>UZ!=RsirZnOB(!fs;=TnFJAOLdFlki%#M*hM2smT)tcz@GDUXx2vuG?MDw
zVV%7(=QDeS*+JF36zTWnLHqM$jk!tWRMz)y^vj*XS!6yYrvGHKnC)>OF@bcnOuuAf
zs&K6+!5f|AXWI>pa7Me6t9o=C;KFUli~EVUs>5CXxZ{h4J)P_1%{*N?rbcM1z}%wu
zAT!wr`yM*7vjjYxU_K;OekEW2(N|$qm3Y!i;3K@9!u^Rw@{_C}7uZG!k?tE>oV2`*
zPhGRZ&AFYWKl6*DZXXmePH8ro>7{x5Q!m?{n@abDK{;Rk;GEiSd^O}P<7(CB-Mj0n
zjQqBHuLd-IWOKXvi?%vDJetRO<$7+;ZB9TL<bfLaq2M`zWdPz@B-u?#_6Cpogx5c|
z;M%^FUFMM78*oa#-|VqJufY6-xAgt8X-9TNUyb0!6Po|yr-qaN;eC@FmgciWkI=&u
zY~jx8Ik*A6sdR3W(26$+k?p6?b~3YSQg6&Qk4e56^_+a)Of&O?dalfZb)G<;!K?av
zKsf2cN?g_(cK#CH!s6Oc-T?N-lQ%Sw+a0M~j;mwgJjW`v*}PTNc{eNWZak6VmZ(tJ
zGN>M`n0kJFzpsX&rSxp)H8n3Jgp}{;Scx&nd`dkKl_H0k*cy3$D(o4sByoQt%ekR)
z%tM_nWr<6$)H9s=DSbRY{T&ypSh54%EhQ^Y7Qal1lV38sZQJu^$MgIb>C-Bpf|p7G
z+Esk=DAXWBYEbOhSBS8HT91q^(Gn^b!6&^C#lQz#rxd{C9<B=~J$!^3%9vInx!^I!
z#@ddT{W;amJtJF`GQlb<rEtRy+QaKd4{fb4oFggh`a$Bv#{a-c4n;8IN4U*Oz8<)o
zyq;*$I^dxYzWqmx_qWWA|KG){<Kgl@g6F3^qA<XRDUdos8MA<x^=H??m7u)l2~0tG
zd1Be1ygocORnz!2Y<kGzx)Xtalx+e0qtlrq{qCmZMtjI!)(h282m^^21ZW3lvj7s_
zZ%VF#L=Y+t-YFd?`q>5q930MZxA$L^ZU4Bs@t=^#Byr$v!2oP_6*2+s=Rg=eb)n=p
z1+a<F(f3Lw6JYbdau_0U!^fk(-=o=9@$1h{V9OzP0`?F*P03iCm<mAOZGb7;Vb0+8
zG#g(g_~ie6tVkXJziJW}LNFC%A5$=YIz)<XP_%A84flKz1vBgs&%x}r70iV{V~#un
zI@a3pp1I#@^MCX0?|Uczx8`n=!(mAEC>#s_9c;u_ef);ExIC@&@yBv7|1|yc-wqQ0
z&y{}>IKP2n6W~#51Yb41qr;Uf$`~P~6L}@b?60cipIQ&_g1R!_SSLG5@0w9Rkntod
zOy8vE`Up|cj`W&>MQZT-!4s3@n&J2-GE7A${m@IuZctZY6z=i3L2Gb07KloWlm$A%
zt(Xn8!PDh<3V}juw{^B_#f&35nHzmL3EGYcyId3&st>2Ry|#{(B~+gXCg5)tN_HdI
zB85TS?i`?WR}_J}@5YjHGjM*?T^~f287gtP>88|p!ENVgYkM%`@CC}Lbt%JBogS?&
zyPol>u7Sih!!;~L#b{rEcupj<xJEqSR`Lg5o7%?ULRxyT-rv>NRqKgvptfpfKEJMZ
zY3{?H{1O*vjGi$)ZULOcb8%yyM{l(7M31fN$7N>j8Gdb5i5ff)Q9-HA<Hx<BtHNLI
z|2SECU&L}-KVH=X*ZmRf2BH@9&B()WqOx=}mo-Qc%)y}|ON}tm6FN20Y<WSP!pl86
z)J;wDn+*;|ejRD^dNwiUa%JNBa35{vhu-T@E4+txOQWwH@$jMQA;D-0v`RcD2%E&=
z3;P~f3BGosVFvZW87%%GAbT`fMAO6!sV?0%^3});=Ru>Ls85eyR5Cj2<7;il$=UCU
zwg<V$=U?whCHSF2^B4h`&PECG5J1!?T0BJlEQ7n9i7nQu%E1;FRYy;y9mSYz;2`x|
z>C9yWHJ25hMt_cfpXHis9dlgkYUe~0@vD)XGXkXa-@;AF_gP#k7N(<B1L)jnp-wcm
z&4j?UfRro811O4N1OkM7dq>0hK)w8x>>^lPNP9sOFsH#ZcM*7w=bI2u#@Jj$3K{yG
z3ti8BlB2HpWKev>3G_yY=Ip8M{@uhfBpg!hrK;G<bzDY;8||QLhM(0?{Fwi{QunP=
z<^2;*8JZ%C>VUs$Y#6e}KzGc5?B^PBERdVW8NzP>W<(fqSaN{$nh1Eywm~O>E7ZyJ
z;v!brSZw6hRFgstY=eBVcUleuZSTNQDyKNsI$T$Olz#fEf+xOKPXa$n;Z0<U>8Cf2
z_V|hfO(^yr0?!|L2{t+i{%;?Rj4_$Y#pG|Uzn&=Yb+Mvw$>!48!lT$jZky%92PI7<
z5<IQyQkmwvpZVdL31k^8+^$K-^?~$dd25+lMvttNU$cvVZspeR*<mCb#q>w<$;tu)
zPy)LU*PL+XY<!t;6BfjNEar``Odw6_FLai3T##3sTip1*Ma`d)f4?a4zWH>IZzmIL
zt`J0D;26XUq>kw_m6=0b7oja~p=1xEy}9BfWO(PN|6D>5C|PS&cd*)9KYXd6uL$ZP
ze&#gl&{P%pH!c+;d}scO`h`>U6VB$P9hU5l{;M8A#MR^s3inviAQ;ezHW(>cD2YAd
zGCB1@G{a=S>WXT+BXoIxev4*lb3@9pHO2PPPxo#aP#JSUmJz>v6<cR9K3bI=RAS}B
zW-0hNR`HirdHHi}<@)HFd4aTVD?1N6+_-z2I<4b)*Y?=<J630JM0i0IUaCjMj5_%5
z8oeX6ErQruWsBxEd3UI;R#bCx@vqwFJ}b);i;#H{GwG;odI7-=t#l(7*@~u<1IeSK
zy8xf)EipF8W@4*s`Rhgb!ucF#dj}AFmg*c<TM39zcTz;+!bUy4<}J~w?ujNv(Uk^~
z!B-D?ExKR$@UdUEu#8!2-x*}8iU`+H1=48a84M6e^brQW)|A|$J76PNh#~yBqN^q|
z=(Ti-)&g`oX2Bta?5sGwH23cL;OsL;isbi%Ql&I&J`CLmdF~YelOJD+*<;Z%(7x=q
zwr@fJw<OriqBpZ<;fsN2C?s50NamnKt7L5V64^_ICbROdG}UozDngCE#jZc}wA!>9
z%5?QPSRji(8}z?`_XrpDBf{gL|9|8q<T~VY*1)__Dwl{)(cpgK`}5GUob!oByn>-B
zrV@6s;`9r(Wsa+YpSPJez;$9@sUOIytkq0*8R~J}-1{CNJU~l+5ua=e-Y*NIfIOIq
z^3p}uOLhEmD!WMBN{UR0eb4+?w8c7gY1I6MKUw=texOZ`nYUjPUwXA>UI22R`q!l*
zYXQqb=DITnt_g_rG5{*VI3>W^&7m*GE><z8={ee1TQz<iX%4=8N~h}Gju&y}ibhXv
z?up4?UVnsu%V$XT)xVKv<DmjmdNXqZJ_U;%8{OPM*AiNB+k$*N%(hIXW)1ixyemxf
zUpH^EYozj`2WJyg#B8@AtRGSYKc+5z?61%%Os@Snp=cT9p6?)wSI>9EPTl=vqREY`
zknwD`Y4VAlJ~=r1)J(qD+m)S@t)h>)&h;=R9tFTmFOb*Ajc@;?PxuR<N40%j-NTQ>
z3V7Dt7C}N=QYB-SpPrlZEXPUBGEwLBygJ^J^H)6UeX8>9#^dEMT{fTj*XN~o0`dxF
zLmqY_%8oUef#q^YLC$M=agkZ2I#{#?Z$TasG5gh_`^Am<^#f1U^90(<hEvW{FQSHH
zu9zv}^e;!UuS41KWUd`66sA8@+Yi<76S1uvRMwAKbw!QDvj|FVGWx0c?xj|f#RXAK
z!zWiiZ#vbGvhUUtNw4?8%bt)MExu>>wDAs@QqVSdLO{h|kl`*IWRhvhm@Q{$ve22^
z<41L9^<K1UW#+E-f)@%VK3A_vnL9Ns{8qQjAK3H1nv(y$!Q1~Axt_npF!*JZ-Ao1S
zOL;wDUts@6UO{i)SYre6U9?ak1m7wlfa1s-;w$FJ=uN;<7XMx>l7JGWI6u4-jP$?L
zI`eynj3^W>yGkCufJUPHb)Lw`=MiYCY5A7}5$Q>2*=*$fV%EJPm>DPoVb0QIScTWP
zVF~%QZ)FOjgDjrE4r%`TU&I)xZIPY<CcjQag-0fhJOoR=7=a0nZuKz1Q6EoM0Qy`8
zlpLMJB}epWw78)5x~->Pqz7Fl1HTsPep|Ho8~L*jJheD^c5$BQl6V*+MAa=6Tfl5)
z{1()WhmYrS#A78asRI@ulaw09*m-!nR_G!Ce`b;|`rLj983>}mFPki(4k@0Ll84J~
z5_>%KHpqo6c4LhB6ag5tJ&9Bd2~XH5%4;Zj)(0?0hGr4a2WNh-am<v!4H52f3#j3W
zr=jNN*+9s+2dG&bP01g@L7sFHBE!X0=D=ebWR;pSz`rUPPl)51V0`#P-}LHVH{-Q6
zQOH+f9`u5p82_bsuYP3AV$1)Pc?t2;zp=&Cp*A~X|0q87Juk1DQYO$1yGPSPLlUL(
zc$pu+^y_3UO)5HlJ>K{A&x_|&+ofM)s`RRWPmEtgyC-*|c(etCcIAty;GZjuurSTb
zLqcVplMJr?4ae>ha{gsTg%29+!7Z$}L}hF<n3X(HeEfx12PWr6F6+l#%x#B!dkP5N
zES~a}AOO`;GEj~i2~47xh=Yn^hTF_V`G^a%oygxE#NS@Z^=2v4Bz@Trgj;a!FrMCy
z;?mql0(Xay1;+wkq?NVN{hol&@sk}fS*XU9r0u|Rxy_s~O5`G#D>qc0IMvA}`O8jQ
zs=WJ2hc%CvPrEiTx8zG0>2V%a@g)I0VTMnOY~vz5dfr1q^)#_d1$7L$;$H|NwWu3V
z1lrCT4?Y9+!4*hu1fGDf1Q_KmuAojg{{&j)w1F=!w+-bL6F%u6?Dw6z;C6VBg!0a#
z{9})>93O%5TucY4JZxm8c2OZhvjJ&x+#PM>Dsw1L#5^1jakuVUk~H$yop`5!Pt6th
zc7We($vnzZ!(SDa65}{;P4t>*hG`^eYuYM>JcC&&h1F)>4@te>5(4ULBTkr-uX~T*
z<x(<t19(E}t8aqI3jk`1Ha<CWHERmOu|xL`eHpQM1<n6JEL7~#p8f8d@i5)<O<B8y
zpy#hoB?5;#a50SjF#v3x3qOR%tkGt*PiX7TqA%r2!S2o=kxXJ+BSwQi4F;dImtEVD
zXPj3}-%I)k9nEW)Fe&8TnR;<0Vg;{vAl<F!YN@JCk1?g9_$qI6t4>Dh)rPyN50@v{
zKRbb3%U7X2s0aTK9)4~lD2u+4e;kU${$^a})+;7ER$&?J$tGR&u)sEq&&<sa>e&n>
zxVhYpetExgV@c7RzQ&JR&K`|FL07IfT3YvQLVy7ai>zG-$EVwZaPvc;?q^5z;5OqV
zvj%(~&74aN*vU6w-lMAcb(EKq9KZC;Ntta?Sv(jVw)BO69q%ZW)_wI`q04n>yE04f
zJX2i$eAKhP(81B^=%>8F=_!v+)SVovaqjG!(AY;0hFYvEAL9Ls`Jyy*phO5>dP|RE
z5-H(#B`P@vL1!SSsIs!tRN)MLwr>?FXp~9#T&m54KAYWsjXr*>Mw)ioI@&C?F&sEF
zd+~uK2fpZkc~YYat@)X_#_<NUul6fRs51|w6)10K!4bm;27YjZkaU-fH2Y-qLGGqr
z&OB@#kjW>!G%1XK@G7{#mA_T#oX)$ewlA*>JJ-T2t=`k(I8AaW8e8ByVBrPtLK598
zRm7j~=+cYkjIbbcQE(DVb6n6OZ;QXo7uc>2oAd00sO!2<FStZ^8Q2fp3FOcELM1&K
zVb>b_mpM=3F(MN9Pn(B!mOIbnDTKzq@LRF#<?9CT@R!>=h{L$qnu1|a53)jj5{ZL4
zH_L|J{p7GbuFk+bJEK2%@A3}`($BItKN$RQ%8=KwX3$g-%w0S5`*BZA3gw!yspZt5
zA;CP#CtGgw6O%dIZAWT*Cf{_jDltxVbgLd&zh_QWS<s8g7PXI5_6Hb;{CQhJ33xh?
zM2d8{K<83^4M6@ZL=u3acz75YIXD)A&%Q982tm5_b%=Nb(+>1w6B*(tqR{+(5~MWZ
zK9Vs9GjZ$+2KM~heDc@|5&S4^DGt@?u+?R70PII&zQWm(PZ*LVst36Sy&MhNT%ISh
z+qHz@WV7c;ZT4aCSoXi!x;tWlyr5H513|=O;+JSsau%Gtc|EqsCr3CVkAE+Lf2jnc
z>Q_xp<&r{;6yUGk8N~K`DpSs>D=-cpoiMi}_L|~gA{fk^zb<B53j*svg$f^=DA@ef
zH((uX<KDr|#Nd5c>WRAx9utFQ6U|dRDVxj>70;M#^!Vl$eb<OX0c#=b`IC~AYK3(j
zDa{4Hf)}sb5%r0G^FDT#Dp;}xYt|3k&&`jgC2B>NQFwYDin`afI~1_PpUcmf08tBE
z{y*=(0ci?a_OVzY%*dX@kRiQpvnw!B=S_;dDSfeLO~OF$j+X2!BhjO}A?-o6LvK$M
zseA}Pipf*lldIHba_d8Xp48$OUeUg^SSQ?bb<fUbj{*q^YWH&KXkFhk4~I-uWeBJo
zE5AmlrhE&pN65G?O*W2iUdoeVNE`X)`00Gj=Xvk=bZ2$w$=0;5`<6~y@A0Ty?sA>1
zm#LcSU*!N_jj&T0h1<zRLxb0N;ry&ZE-g_qq&qRk<97PF)DnfER8eB9A$3XTgbj}h
zJ)a_1WC*l*|8Mv;|9z+1-(vod|8w)rV!viU@GApMmk7-GKwuxVr(6QOp?FH$OIxAp
zW$2$ehCwD5$`M-lja3G45$Cs$)9$38F45l%ZJ}kYw!Rz{ZPD8VzdO0F;mkzfD~Hq$
zQ9P5R4Ad*Q=#YMi%K_e*r8fZeKk^T!Uj%<}L4<nLrw8;&3Pk^KHEV}39Sh;DWIinC
zzDzeF(iTq@`{~Egez`wlQ~udMM^X0cdWd@lYQFMz;_={v)BD7~H%jQ~6<%OQB7Cb-
z4fG>93m(fH6W|t1C5Q_;U};9sSUmstWY!;h9m&NMt_M@>i6df99Lg@udmyf<Y?1iS
znN=i(MhF{@3<7};BAfsKC5Ci>!Ct%)Aa!SOU8%xbX%Hf`RZ>NZ!N6udA3wEjTOh3i
zL%UfN35-8}J--m_{Kbc={}1Y;v43+6;=dgQ-(X;CpiOKebWTov%thG5is%4bor{L}
zFwU{$q0?99I69|ZZ+4LqJ)84X@2W=r;oa6-isWV=JHA>)0IhUZ;BWyGnK{yzE|dik
ziWJ9RDo`qz+XNOp0&TR#K)VDfLu;AXmaGHuod_5d7|Le|0gyhp@#4%_kTpJaIV67(
zwl6*nV%PEP@1!{4Oc%fxJ{c0uVT~K7QS=Rp%czQ&rPJ_6RH}4|Uv#%cSwglXr>(Mn
zvR}onqFiRF`?HZj;_iZ$rMC)xdZ*a6P|-!o#Y?eXPXdqGF7eDy06cqQc}WF%S=j|~
zvpMZFU=w<I3?MG5q=+`J3L^}Q>vNt6riluZE%=MYg3ktjT7DIQ9tPYKz>8Fein%A+
zPkiD~M62wG>{$dl51w9+$!6uW*Wy#pagm)Ek84(q%8Zbt592zeT%VVWaL><4Esk67
zISoG#mtK5zj$D3^$pd^Y2{!3FJ22jR{1cNMP0806&hWLz98+kgdHR90jb1z#+gAmu
z6H8RRjQTxfqfhYN&y{=$_c%m%DSLMs;vC_nBvPLiF4GHVkSc_DJkXxrh|Zxe5&CdP
z!+~BH6L4;X1m$N^iDlD9FMM2G{G#~h*tc1n%X^ovu`{h3a$X&PI>VkT`U>Tf@lf9y
zU7>=TIFK+oG~S4NhXTb>b`K-SAV(-q<QsB_B8*79qVCx_DKzt@Re;dru5;t8@kWD(
zYct%J?Rj~4#y%gTosBvf#EF;lOvy3u-hVw`1iGqW4zn25dX|MPXNuiJ@MDxH+^x}L
zklB`Ga$V5z^m!_mDpZOT_|adeErkw^c|GWP1~Esd%>7^OU3)xKd%I_=rV<kh36o1`
z7ZRJwWxA^r%H9c^sZ?%hYO73}X1Sz<sFX^ZkxEIj<(9}`+)6dMhcGiLxvxo$6>HW$
z&)EAtpZ0mZ@7d?P=l1T;`G?PEhP8g{x1Qf~`##V2QK8i{mtuFTH<dg-v3-QoB~MQg
z9?bT?t+T@UaWZF9vgLQXbD!17Z}65GjVE*rVP}rWQ~A!pKXk`^Ci#fjgGo{@IUul4
z5MEMX!Lj_qxG^TsEh-FaTh_*8<yw8Xx7l9jh0Jn&jXN@9Rj5Vu600B{i^S;P9P53A
zt;TTU5K9_%Ka~1@{ceG^9(QL>?4jjK3m?DnGf*eV#WqSeg2eqTT`Vo0LzH%*{>UKY
zNT$k)6LI}<2V6(kQr@9brfJ;xO_HTU18#RbP&so>n_<PVQNxw%o`$-=nPw$#YXyU&
zTe~LNQ_1ATutbnelM-|2?y^~m(Rl1?L!_L!K&T&u>n59IKSE;~96XhO8r6Qa-gJx0
z$?+G5<W2hV1(a*wN{j!kT_8Wt!WI?abNYD@PH}mo5rkRr4==4;1c0VwRc7$UWogqn
zvqpt|h6f6?T@sG&n|)x{vK8E{`Kze1RKc=gOy5?lI2=pEj2hTj2Y@_(C1PYOrihqP
z2PGXv-Iod3++m2)HQj^-guIdX9F=#MHUkd?hinK_;UQ-_HL=Iz-@mt@kV+{1Oyk^~
zaL!74<k9LVL36)r!c-T}2kWvgZmc_3`k*^Kg<KIg*+DzV@Fe8E7MGNKg<a-!slya-
z3a_arf|%u(=ujD_=<3rlnt8|Fn<Aezi0{7V_9|n=j$)QW-H>+Pj`>>#RaZ=TNgB3h
z<C-fZtE%(o)eSLKgXw-}^;C0P+pMV$1`Rt?+n&~JId0&gw>Y`NamsCiY$TDqUB5_G
z-pAfH5TR}GMDroV<A#lQXBrETYDx@6(QM;E!;f@LAgf@<nB6gIG;CZM7>V27>1#gx
z{oS|gdg*~`Pf5~cRRCICa3p=~8rMV0?y(9i29fdQxVAWrsfn+~)~8{MyY)|I;=B3n
za=v39tar3H6h11--I#S=UHgS@?hk81%+DQyI<+yKNp&`b3uSovA_a-#F{N943(j_H
zc(-yEe=wkm&_qX12ipcB{NwwsZ3?0)x^WMCsnd<VuT@)B?r1NqGcSc42`5dKm=O;D
ztszcl^X{@`gO3V5dCt*<CL?GyN{;p~rrd33l=QumeUc}eL2A@VXx+WfN$SkZ3&%UE
zHFnB;Y?*vv7XVnz!nE0g3*p*H^(D%EF#n*~iRYl;hVN|r#@})^jr9F%dr^b8sLSV}
zUzzFBnNQALJI?)Hm@6ZEF%deBjuLV)klCvyo*}fkKuHK}7ZYPqWlm#B{93LmZjMFp
zHC70s9fi)JxC_5_p1$r{jdz79UA+lh6ZOaoK{tJ-_v~5@<tp!P31Bd7JTY@2456r4
zSy@og6}Fguw;%X&zIt+pc6uj0Yse_oS<_2i@jJO8IjP+cA^f#w1H%&uR^$fdqZ2_w
zcvTW}8Y`KL_77WGDvOI5bm2<<f)-q9ttj@f_-cunpU3=`Vn0v5@hq&|4M6GIC2iXp
za3{Zo9Wk3n6p7%&Rpq;XnpP_1X(o-jPk54jUASTkF+*j16?C!|I}7c~snWrlcx?GT
z%1kTkn^>Gje3gY#ncVZ_>~y(h9ZT>Fw{E$-=y}4oHXV@yRQMl#W^$l7k<E*gL^`tz
zx-G)lyxU!Qq$qr<U^tAH^UZ50AV>>E$KDBgDldxy>ojZ(KHfR#a{c(aoW)Cz>Ua;9
zLH^Kq5)DH!XTD;g$KnPQgrg!?>=Wu%&XTuT){4ao=p3pNX1V@R<AEiL<z3sF4`{Y^
zy5;X2T{3e(_EHZdHSjlgWbUFydH(#Y`3%bn=XBkKA4yOapXq*c?Xt)Dmoi*w_UFUP
z&ZmZ!%i!yNge?9Jh6+Je0v*7>{K_=t(YM1K%Gr7k22x(4S!dK~NOdr=esSrh%c~c{
z3n#C%xgA>+pCkdpnvjZpU8sHB*upM#HtK#`lD7LrY6YcTNickxuwgc`G?`|M?QJ41
z2T@|`F<4b<hWAtNUKWiV%MKnj7rNi&?>M?ax;(}!%|PZ|n-p;Y*+aw3O+~Mu*keEY
zG$j|}emF+BLA{lN(A0sotq$rLW_xtnX1H;ta@rqs+&nmY#SvY9)$02OGJV67sZ+#<
zimo;WJzoVy_7(%%0>&-z1I@Te7NdlAreujQsuSDpv%i6wllSDtg=Xy$WxDlOl%-x$
zV{k4-f{AnaxfM+C8SpSJ(ln{Oj4l~MDtI|Y1{M}9&;@fBnpQBEw&)9EMppYBzTMZU
z9=<8eazn($-2TG~UL>8}TqrKlWldrX|G0zk&bS@?ryy|wdk@l@CCD123Kv6;YQ|C8
zlwe(Sgb_#k`6ZpM+jRdxy<mQT#)XZ~&MrFM7{XB{WuYO6S+4Y_;Q=_cP{z9mhFW6~
zdZpX20HO%|B0Q1)<)5Q}jr<AHJK^83p;XQqeBn<5x_GvM%-DOdS1s@b^r9XYpP>#o
z0gXzs8jS7!wGXP-nUFJ==vI6Q{L?lexb<6Ip`<TUV?vKSXXo4Ei4KEcPxh(^xI{|Z
zTPPo_>tm*cd{YcDn%1BaQ9ec|64t`5y_Jj<e^S8YvPA1_<{*8f6h~6Q-}9u1f4tOC
zMly^${3Z1*g$AI~KyexL1x7;XTQ!jfr8-i1(UcFD=uk*h+j0u>=$E0ePxLT_JX@T`
ze$&qOrjBb<=0dxnOv`&1;ugGG@<?=eq{~FY)>ZQ+>MbaP@CQ?%qWj`1W+_Xw`Xfth
z-G5PftOMq@6$0qB7w{oZ0vz8~18fuAK67&jHSsPScsmxh?1qG82dV`gjNCx_OdxaK
zDwaZ~cA-JlcqD|l6!BT66-2q~rI6&xi5(vXkYk>t*-m_gD1o4+0V*%E{<R@zVnGuD
z>vo_etj59ISHO}8Ii&YOUkLTG+C<1Y6hAT6{zA{VKdee0)hE(qp&%=LBGi=-`;Zn&
zI8*$7ks6<C#ucw@hnF<?JERU`ZZ_~-NXiXY!7`nkU?G{n9M}8u_tlOhK{c)p;%0D2
zTnKa)TcLB$ePsx>>d}8y=bBjJ6lVg?-}MSwX}M^Y=my#|=q%J8Eqzo_L5s9eY{nPU
zeXhIv3dZdltX7C*3zG)ltez3u^JZpgLZ3;L6aiUFfdLiMDG;k0Wdqc*12zpPufc?;
zfbNRz`grJ3HKE{o`4=f5z9zK62MV)9>dN~m*m_@@g!>wcA`{aGfp0+e9>SlSKEa>;
zKXOHxWMLq#4)D(uq{(WL;KVYnD6Cnxs4>_&b7xysuu8i0!sngW0_a~UdX;u8i7K=<
zO>UGX9H8>?xn)JyT&5ltZ+qpnkBt^gl(G)6`zd{`O;QFUZae$ImFsV3CA284aC$ma
z_TF2D3+bDOuo)2WeK<>jovyKfZTPsQQF?4K@CHF{U@Ca}VSt{MZDJpwVT)|59pc#1
zgj{duB5Yr`QVO!G-w8{twom9Ea7E4^epK(BxKXsJv{wC7_wp2S4b*=_YC%>3V7N~=
z5xyU!VPJ9x0q24jN6ACV9Iyo{1MOMN8Y)l+DLSBT-M7!+f$8|xI@}~}r=8%Ak(JiH
z1+g4Q|21W)XEdEvmtK1Pcx-<N)s;HQi;%%!%ffH6d1<T%6v$3g(jL6KHU2e&7^#L)
zam7?YEoJVeqp7cZr&aYhHO7b{Qw6rB$Il11dL+qHBV|4>o=B^Si{K(bY(H)&>*rB?
z&W%zkh3%cC#O0Gux$1L<BNO$mXS(%{`2J9Nr~lE3i)u%-o2PA3-KuLO>m^&DpDBf0
zr{n8M@a_OVz7f$(MWY%ee9b?OmQu0hYl#`h$L(-WXke%CRjN9iDkzF2`u!;8RW@c;
zqQrg>=h)N}78E(y+2<0qX4Z}m`6AN%!{I>TA$l5OWhc6X#uj!DL&FU#wAj?QphQO)
zR}P0k_2q%K)ruW{YCRE0-7V}3eB2hRCJwzbs68Jc7Z06%pI1{bQ04F@0B%$Lbn_^6
z#L|b2?+Gz8LooCC(0~UmH-XTkp16iXzw?T_7%BHlOUmX1%!xhxCLm%v)MW3mnT7>?
zhazRv@J_{+UCNJoBA|HB27Q-v;J&Wg3Fmca!ud4cr8B0<L>?SWi&?I^t5CHjg$(rB
zBan;;O9F@wfry|@MMQ`*nC8Jo7$waQAv1Mx*Jj*-Z|F)>?iF32y&pa2qN_so*Ij}p
zc>i=@&&KM<QW6fT*^q81+e3ZdnG1E6N4g{g?ac_TI84k00x0-O8SxyJ0X2;}8f&`a
z2GGo}=3BC$#mb#mB&XS3ajaP_FzWuE$94O+?szT;^9-QN7Jj!SZD|1OqRdCrNe(|X
z5xNvE53~!t2zgly<Q1B3h#5#qr7(xr!r^{pi`G`!Xg3b)pg%R+7)PNro8){8fzSTc
z2X5hSwHVa^Gu9YSD*a;6C}9RH?DSef1ypia22u$@&Z#J(oHCRB0CZLs(E-}Zi&#U@
zMM%7enj;#qpbx@?uHnGR1PigCpaHhHvatv<d|~0tYPKw)ENL*DlZE?t8_VOa(G_j?
zlaBKHA1_=`uVTMZt~6QT?5w?NL#b((&hlMk{;xpYP#30WTtx#aGg`suz->V86EgHp
zqoe$S%1ot-PQ#|uf}p}BX3@)y6@yOZ$7DQvTAwogrV`5+G^zu%asHLA6fhSeWz{hI
znH)ov4@S`VQkF)ZyHhn~PO*|8>$HtrgD%RvaWFA13a4MEF}yg<#+F-lOet#mWNCTq
zYUs|2h#IYp0M1F8ox(j4Ue?$x@z8Crahu*{uD%w57R?VRYCd=O<`-QZTJS^KSF62m
z?kKbJQWu1LjZ_#*u*$eGi%bw5S$7M`cpyoX_E$ng!q-S*xT#>1Z~RzKMe`aEm_~%`
zqy1cfD+h{XU`M@~HC@u|UBhiXg3Mo}La07ac37Ivo<%Ih%o{ibiO&m87eh7Wi96r-
zqJ0VVH&yN=JUW~4xlTKM#3W>O3i%!~nX1A+vUz+;+X&BVkOTKO5RO(4x3*M_U~GOu
z0Xoj_JafC>SuMD2?}NUT%B-MiH)iA|C9j|TWSLYe=U@#4w2>G8&dgR~Gs5eoO4`r@
z;o>T2)<9tz@w7xLSJ>3W$(qf`AFtfj;b-1v`^x{5wPj7t5x<j%Ev_0rT@zZ6{=!=Z
zgIGP1211`r`AB!QC6ph5dnlQ8q=mqSR3LSaeE@Fkl_(JMq%uC<PRzItDv&MChFPiX
zd@{$OeOM&Kf_%^d(LT&ld?)yLx1EM>dxgbHny4e|6$CcMn{#3p4O#Waof|k@0NEgz
zQuv>@$ucz9RWyK#D^e=PL}&PqDcwZY;S9!U=Y}KJM$^#!H&j{MnzFQhesHqL{<|rc
zEflnVSbF_IK$Mhtuwghe&rVju{`s?n#Is3xpYEO9^*HbKqQf%hj!#WJ??y^H+!kUq
z;WaUhm~||aHjL>}2hxSCYD9`@7VOiAuMk@BOYXPnN1N>oUWA;Odnz@1y-9>o>Q&Dx
z)8DRryozK6B=-WEa1I4mXeVM-c0yRU7(kEikJ1FfLJ!{_OhQQ_6!tpS579*Scg=mx
zsOD#OmD5ip&kJ`cY)Gqr<4@y^xmT9lri)xU-h4H0?|0;%%o!jT;onYFXf(kfwsv^;
zdg2T49>L(wMSQ7>%qFIF5o#1KRN;@F{79Y2^}F$X-@EZ`4OSbBhk!Bj$kNEc&_mBQ
zdX`qf_z|!elekLcOvKXxYEEDvlL*5=VS)d>KQ3zQbPS<oo8s%C=dahpxt#c|=f-+R
z>rd5fUYB0x`mR>!HhdMcJA}I9z08MrLe^{C9CsDpr}Dtf%MC0w4jr52p$F8gT=J_|
znBV&350=xj4r(=OC9m@<Tx+YWv*EzcD|$OasL$Dx$l!8n6a_tKi_fL=hB@4M!P|vB
zrJ2TdC_N-sZf!;9!bYdcu@jtDAGN)DLA4Q~T01UQt0la;snq^himBo_EjI3!NB^Ue
z%c(xuXYw?>wq@7f(=yDp3%c{V(#LO}=eg@WSLbA_djX>i^~aB3E**%OqxXAPdK9D)
z+_5cf$Fgs`_~t7~bkV*CjfSa3dFRr@zxJ4MJ}P;`p&uQm#Z9VWC_-64SCn;#v;7<4
z?YDpjEg3I1nc^aS8i7rI1JaPitfWdf@?MSeRR(5^AHp|2#=OS229xUVA8|P2A)Ed3
ze%`qM{B6ZW7ev!NoBVZSWmB3z+!$IX<)!XOdOexI!7Ux3tDHWz+}#s^{7gH^loC*t
z3ai1WYOvq)kH+EJdB!j|gEX1(p-dvf{|%Rv{(Kf>86-Ij8aK*Jptu=&Y}tcOD2d0e
zN^+=xY>IBMP5e4EIhq<lAK1fgw7>4PJYwk9)f*>O1p$Sj>bs!KW5n@~)#m;CJ%hjD
zvj6rM>2DK7NJ?SwA~x?s+~-Y!ArmuVVZUvmnAYUWoWmHL*(nqVIoXns)8j~-*wpWc
zG*Nftcena`1Jl6pPXpEc)CK^B*5j;SsXRr{ms1LZA%ra=oiPS5g@$bYf?(&55uAUV
z&?|yocoDpb1$H68aEORay)R0Cp1^hgjzIzdUBa`2gRKoul2C!>cD0m4r6zDYPMi4V
z5A$zBBuv7f@S}{wmVkWi@x;>#fa3ic-pdI0OVU%nfz>i&5<lLO{r2lo2$V~7pz2lk
zTflpBCmv%aT)-ZAOQGbeh|kd|NN=huERC%8AK|;2c%{b>-kUgp!v{m5;b;FfNCP}H
zkU{gmSjb=2^EbOizgY7xziZd)1|yz@>0chq8@Gq`ZX9H4<5J?p5`EkmT^X#3YQ}cz
zp{JrXh=fPC&gNzu-?QKE;q>H=NJXQD@CIl``FtPf08EiGuq?m|3h4&E2}E+ga)ti?
z$A4%~WW>j)yekArNaxTbEUW{UdB&RFYH<<8o_vRJP31a#J-@?MYuc?FeF=4&jh4`N
z&njOb&?UJ)kNHHPq>hPmOX=}~)QdAxW&?a5)V641rZkVM3AJQ7NJUq)hkOoZ!XMP^
zukaw(#XC0aYBGHLta;g%mO~0wo(b-^HFKr$3WRf&%}Yms##IwRublpk+Vy|sGwCsZ
z&{5u?i=9Cz*V#yt1KX&3vd*xT*n4R)%o#v($TaQ!u{tW)`98*6?60%7UQnOMxLLyf
ziMH7@R5!tFi9)28IymaoX7k@26!HjU3c_Qd-+^6lc~A~CjT81<xR!T$AV=sapm7F0
zIUb~fdO&(k$Xj-3ih>Ny8AF0FUT3Kb?|fLz%tH<yt5V42Sy%wD;~4|VZa0@ug6t>|
z?^1z{N6uP&1@R`gdWy8L0to0Rd}=hIz{uEvam};S-g}t*>~2Gh*sa`qKKlJ0m)mmE
z=T3I8vvxzX5+pyBAnz78uy5Rn*=#FX2LT%$o8QfiLC`aCkhUv_1mN@?R}$ky7VV}Y
zTK-UB!3P}3Y`wx-hhw3;Bl-XONt9_pyt*)*9(+#EDu7pP-J6j;aj~w)&Zpw;n16=v
zqX_P)R69Qn3wvR5^zjqN4?%r}{$Az$pKGl$!}HBN!-1J-e3AG0P4dw<*M5C5!O*CD
zk-|e5+PCpi)9F3;rrQWYzL`+X)F-usEQDuMvOt0&gLlEtRAw7niK$A|uvKw?9%Wvz
zSJO9fvW%=XDDJ#+L*cF3TRsZ5Cw8wWiF_jSWKN`N>V+DHDkLTM0u9c8`Upr9AvDy-
zVIX^aE%B7~K!0!spW0^3r^GO4Gva9&_m-cU`Ppu3YqRkVTf1mZL;I5x{)!GZnBp;o
zge5kf7<43r|HX1s&ak1ouZ$IiP}santXv9!G~qVWT1p_X+`>J@&Msca&-kvm`5Q`3
zp9idZe}RuS`K|2;`^7``QJ88|TSTZo2WbcA3&@_q+dEl7{gndx4}VIw!z7!^fbC3v
z55ZcAQs3C^fvv*&>Miv^G^w<`k?w^)TQv37*|;dPjB7>d*T&c9SJLToF4VNUO+U%&
zo+C{>^_ZETTw7M6nDgii?@=!LBHTY<N;r8I>a*kgR;AcO!kU;E(=w;#kd?j@>E;iq
zaJRQ``~fcKr2s@Of&cSWD#<4y?@g;BYXYHhU`1h`u^c0ZGd9KfN1L2INBRvao6eS%
zoP45iUgHPHAJ~p5lz1xoj*fjz1E%b17OtoD8z1%m&S%JGD(Yv9E7gkH#8+VEXdq32
zNhO}sl!!SD4Pq8!#FnPSV6Ttm*!gcCYg?F;tz`K64fjxKWQ+Hc{uVQ7BKF$}4?`XH
zWlgY_fKDMQD0%u}#q(0T7;%+MeaxiWox&MYzNQMX&yFVJBZYhNvU4LZDLPo4)b-u?
zq#IarYGfxB2>EV)9E^J-W`?KXo3NleT|l#^;eP-yKy`+_NYqE4RW2DF(ci4}+F>!`
z4GFwibe*5nzm5;D523c&@gpvEUPW8@@2(+J5(?Uo(Bg&rg8U3B@>`B2PGe1drWkuz
z*K~vMA$QhI2!w~cC7uhlR)>(~gJf~wU55l8g`ernMr4l7Z01vn3UdK&BeXbkz~~5B
znGvu$cG$LQSDgJRi-dY!*BbZl%C>Kow|!|q>Sp09t`RdK)4pMaum^CuNMwTd{Ll6w
z#gSP7X@`waDoqck`mtMcDw7w9W0)4+9`n(jDn!ap?OhJSui5jU!K{cp`_*d?n!BAi
zWqd(FsM@Cae!IWzZdyfs75jxvl=*Vb=Wm|mKl86IC%XMJKgVxT#+QRV{tS=v#b|^-
V!{dAb;QtvO=L^EmKf~j+{tE~bBF+E+

literal 68003
zcmdRW2|SeR+rL()1tn6lq>|{M#Z+X;mXLMK$RHtMG|5`F>ZAx!$x=gOH)h7p493!t
zWGNx@SjJWeO{gs6NSy!u%uwg|dw;+G`|rG;|NDO4dFGgz=eh6e-mmL>eXpzYXXl@9
zc{UlG)IZ6yVg(P+DRA-p`Httb?p4<dJUm85JaRldJiqX)2zBLI4ZeXQJfh$~9-dXn
zD|l9c&lT{!<dw@$S4StWTK>Lj3@-TRGoH=AtQuP#y=uiSo|T(dtlGTd&*wanpyrwt
zoEKaOyjHDTy<!c|THasQt=|9&)NJBeu?pN<wR-Jp-nFZLS?jf8CC{qOs|BUj9Ne;N
z(^^flLqhf*yLo@x`9MoJ#r$x=9)|;4t36w_qm^#{@(y`I+EE1Y`Jq>;sZL?=m%SE8
z&c}E^JNM>R+P2FBGP*~dicmi3u{F0R+fMr47}P^wC|<XIyXZbyxns&Ir}Pah&sbrc
zU0ktOuKM|hgx(1YzaJO>=rKMyBa@g#BIi?SC8cHb-=EjMsH<;iY+*p{9i3gww?prT
zM@B!erlx1uvwzHeg+~eIh&xE|morLu5Ubbx80BW3l`FwGSFK#NX62fdt5=7v0CTiz
zlc3b<od-3ytU0jjkeSelwTE}xYYBUpEB$7==Rt~g!B$=e1oCsM^to!Ccb;EFZaN;h
zyjRyES|;@kn7nPAaU2B$IjIMRf<nWCIq!4<9*>u|k1sr?8^O13bI0{CCYC$2wDhIH
z6&4jQjqh2_@(A0Ohd3~}G|0*CqwMC<<`Lt(4v2{z&^d76fR2q0c-eT{z%TE+-kjIn
zrB@c`^&Ecne1H83gsj-LXU{REW9O8V&J8EtNF4s@*Pl$DjjK4*w;61%ylymnJ;HZm
z|EI6kgVwmeUKj^A9EEe)qrTRluH{2^*$B&zeE-XSIBRpKI4t}hC%A3P(qL0Y=%WAl
z=)dC^qeF?&&zWz`-u+AS%$;q@X(JpJzAl9lMwgdh>CrzH{QvU{Y14)b;&i~}Q4Z}N
zu7d~~1YdNCfE3PyLfNbSVWC-%{`x1??Y!<5X%r1AMC#Prf`R;!YR}p$%JS7!<g>`3
z;vO2*FixxKFy$`KY|&}wd~ZO(<GmQudSGVD##l}px{3m3;1|DZ0JdA#SRxfi0YW5T
zNk9H<=S<ws^w-40<TM*9W<{!`8E3VTl~|EfyKQ^PCtq7bMeQbTE{qrCcJdQj^B%@1
zgAkCM49X~G<?Wj;Q7UFKGux9!p<nF|lAll=f1w86PAwa!dhZuy-Py~BQ^}Fczixe!
z)Jm$9h7czoWq7O3Y9os+Mj@pRrRnC=;EZ2w?X_`%cI_mGIJl*lqip`DS7G{`d}2nl
zaD_yDK$U4w1>Tbw{4hRYoU`IOweZCJxLm`Q)}v*rmk+*;4G8ylcCvOhnWW0>H}{19
z*n8MaTHvb=80j}cTd3%cGnN)FpiG2uq)t`xsH-yOp?-k2RwU&#HbJyouFzpF(xU7l
zEB6rfyu5HTj_eKeP^ouUug*9Mq9Ld%2$j;z@?<kR(>L&<lail~kdAs^iFqf-z`w-M
z^R7L6mW@|)La6}xKw&J6i+))|C^1LBk-_=HjGl~&|K4FYJl37NtF{PpchmCU%)K-w
ze3fW7dn4eRqzTy%Po6kEIa3g1hx!BFW4A8RM;#o>hDeCOb9fq)+*Rg;HnFlo;cz)9
zcsdl}!yLZ9&p+l4s854pp6K{i&zxyM;HWN7v}&Q(87`EEduB{Kn_M3G$2_6n$Y?cV
z`48@+LbDY!3Iq3&voJu)c}3HhN-H_WPE~h{0BqY9+x5}pkr$fKV910VDJ{es?A{v2
zS0;Wqp|#KK&DZ6_VB}?F#O5@c>cAbz^55d;?OY`^>NgPr-=wP5svR!ME~1Feum4Jc
zpJ7C7H=edsjoq7fcMH1P)O5r&f_i{VCBy5@*%@Zd-l{WunVDZ7B(iXctcx%DXcW7H
zLFjJmvwc{~5yr?7=pa?xh$_DW6`cQ+9Crj&%&Lv7Jo4J-s0UU1ib{fI&lKZgH!^Yg
z9goP9dsiMZbFx@4XX=UBaZB|>{CQ$WJuJrh4#TYyuYj%F7xiCVyN*sv=qwm=Vas2Z
zj-qpfF?$HSA%_d?jF)QgyOv5+GqHM+{UMh~6SB1U5W&LVZ0|DDb9j*;S^(Lut|p6v
zwc%cvr{@-Lh|noFS*e6!Ot$<!#SaK{#f+*KvITi@7AZH+#JYkZWPeF$EZvSc6c6?a
zjal~PzR%Tn<>AB&U!K5Xj+bSYWlmp-@vd4{v_Gp##y<Yasx6O1G3!vipo>RDi<g%W
zwcA;q^sb&w+Hs}>`4O9d-5?_bx!P+7>DxP!%g?ZK^G01<6t}YqfZ^8;RVfmAh0SQQ
zM0synn{`1Ly~Q43o9$!iCT=0+-JW9yGz;#88>e1QXfeXxj_N#yIKvu8%dY}j;+VU=
zVZ>ewLDVry6K8t726wW))Lap2Iif<LMv=!INeNA)O7ZvNoYMgtx1TXS$R1?d(YrGk
zRt}m@LFJMmH!>AddzKVQ{-YXUe#DcR(^nXPifu=L6OqZpgDEu2%W^0EO7McP^NFLb
z-bWcqB1V@;kAEe*jHDG(_9{Mt3^Q}dqkH*GNpiU%4z736sdV8RGX_u_hI*}f;poc!
z*bk6yC0`0ju0=kNm2pqE*+R{!1?6TO5`N=DPA$Sc$P&dCW(&JCe3Z+&+*^BAS}T)I
zc1w*R!%KV)^Vgn&ZVS(w|E)9X6GgW5A;;TdGdeINpY|U=LX61yBA5haCS`xN8)=4W
zL{$PiDV2%DmSi!hl_x3q6jnQmp6k|)ZQjl*<8x$fADcP%2%(iurf&};d-q~D`;yVQ
z__kRKCZUi?=w?{A>{p;#sLJ!%9Ky9ghdU}}tWgo@f?UY6IycAO&K`Z2P)H8t>p{J3
zAje(u0m^wd9Wu<Nnt#l(pnpUuEdM}lod>$48x6XXInP0q{}$E328xS-&YY(Z=@F=j
zR45>edX^<``Td8t!NuWQcs9xb>WxKJx}(sqM0_kTv#4MECH&zNRiNvG!q1{MeHCTq
z__m?OTIiHyc$ln81}4vEE2USL^pXQoNaK~dl{o5p!YoR>hsB8Z&!SF{<paofg7|DL
zJ1sj=X)@X5<Z94?Fx$02FpU{3Y{Z&1!;Y@Ab%cAVFJuONm|Iewh$h5Qu0H)wL0tn%
z4bx%S1zj8U@oH^7O!f<InblGmVdK$HQC{6CjIP&^DcD7Yg*k<ph$sBy0Q5A}G7*59
z;pZswT-h1#;{>Rs!;;k|KuU{<E57^&+bu`*-eS{-kwtp@kCbRr5<_2MExx`cUt}^Y
zwadnB`9}aO81WZ<0#4(31$+)=8>_h!JBnebW_G7Q&4RP&F2s(c@X;%3q!tAmOEtPU
zioF9~Hd9%Daf_vq{lp4l2m^%2UF}xTUBmT(Z!+ch`E0k1QujJqYUKu4>fXY1W3$fB
z#9E!1Dze=OVcHBXMIxFng<7iRjdJi4&ck9JbPL>3{o3?I9cLoE1KVgoNqK}HdC;o6
zk?~t@s27uPkm`qGqo5gAbM~9E@fH>sI5{}wHaaYsGTw|Gb?wGLCr906qnbYizqHE0
zQ7)?3czsPhl9Q*e>0z*bwCHFo4cZ>*#F@>B084Epkr|GO&vhkLaTdH7#z=p?mVKgL
z0kxKP_yMt0+>3pw3-u&G1Z9^;HU0P&Qh%B5Hg2IZk4Eg0n>S@0m)=Xu4??kPq+$KO
zn=tCfg4JDzMTx~noHh1qzT?^9N1O_%^?8V`zBx4Vrv6&fkK;y3&=F!RO%Y4)InO~@
z5A`g2CTq>8waA`rG30UsdR~5BKE()+tklAmZq~KfVRW(*AN?HK5%5Q{n$|}gxhgM3
z7YMRY86f;EtV7G8$mJ`OyLoleE;18KvZW8%R<u1yC7Rc6L<&<Mvw#|McRJITJF}Nk
zsZai7(QU3g5lR4krRJb)FZkd{zsmOF?vsC;dS(XZRanFvTh744r7L!|Ly@D|r&J06
z-pIjsp!(9GD<p)?)a#3K@wUh}?)!RB!{8Qrh3VM-F}c=<tsUwE_1=G=R1!`bdCM74
zu)jfY6+9g@_BfwXY3*b*`f1;Xpyg%xJ3BRsS#<O+EAg$Zw@M1iZHsX8-fomJMsED2
zGoolK!{FmCl(By!?fBQcstc13n!Qi6h}KwA)ur{+3~Q`7(6{Q`jk4~64VwlbbEsf>
z^;ASGaxR6URJu*GKk6Ej#g2mh#yg{eHte0jHLEzAQ6nwgaNse@=(kAYBTzyq0Cx%x
zj?f8mnX>r|V24WrP8pOm*`k7MGi<roAZQgPd(=vevm<^U-T7Ezickp$);}Jn$gj_h
zKx2y_{NpfvN{M_Tqz&|4W8;HbLGnvv0J)Q*TD<^oS?ms$D1WnSY<_Pm8)xj}7=9%`
zi(Z5xWJAsrgw+t4c21?J8iAvt+k0u`p79JEb?f5I#fLC7JlQU6ceMi*m+hz5jqai<
zL-_3OV(~*eAzMK=s%32)xz~wYvdO2Y#-g*mRw1VibHz(*B75cnMUAUrf6nU&5k@5)
zO&-JY2g>0JKWv|LB`|3LN_FatnI?Zy(S9MM*gTqcp3Q{=W%F`DkJWODM1!>LtxnN1
z@+qflYei{|Fo5;2j64~h-Ap~md0XYkkvBa;cur(Z&##7PjK<WyRN8Jb?z)A)Uo&Rq
zqiJE9H+?@-ZcXj+uhds2dbtI;WG?7&4a>c)kGj|>8xgR&EuQ)^xuY}n(Fos1hq6p%
zcIFd4TQFB(et*LIpskU}X~X8`uVXdg6Hp$%`&Vw~wfXiUH|#WIT!^xWOr%k+WE0w}
z#Z{VobP=&Qj`wj%wpq>H#H=XX{mIG)IHWx~Hg^Yf$%jPAHlQ_KNE{bd&LvvqlJhHd
z5ifON6ieGg65bk#RbY+pQvygS7eG<0ubXjJHY>{-+aBvqW|oe#N>p@LIhlLq7&*PB
z+*ZPqL|ZMRc9fEReQ~6wNVjl4+XQR^zFV$~hiFA1{~dsLT;-)o0Z?(|N~+@xC{QVt
zbdbus166$slYVhp*kLA(o}F9lsQFy{VydZ!uC=472X*aN!f=G45Arnl+K}+|dZ8@t
z)cUL=@-`_hxdAxqt=OLN4Jh2%n@La4vng&DDX!y{h$lOKDHa7=+!FTi0RGT7D!6S{
zu<t4}_#D)0Re0PW`*P1rKyX%rxf>z4miIJjFhnoyyaM^-*nZZtH>ocoZ(-c^0?ba0
zx+;8xamCN!b+gmiAG5J9DnH6mIG(AELf#}oaku%j(Zxp@L)fzhq}wQ}?J+vBG}-KI
z>krIW=xQYz|A2JSwunqLvVr<^-rO&?REs4qxY(NB{vF%fZHX#$c!cikK({Fc(5?M-
z`NoC!r;PZoKYKEx)fU<27M_3!ctG-Legs`Wr@zPQ3mQ~)qw67^L;N&=o(ry^piiBk
z?(i_79kk=O;b-*?@F#)FvV3Z9k4^yQu4d=QB%mKrUv6U4FTMx6-m0^`3pTB$`|?~!
z-}QC|JtMOm6o;=D>&Wxp$v+g}IDq>{o7p<ejGl;w(HLmqz4-fMu&PhMwgc!5dso1}
zirrf_{c722Bj+*Y2F`$DvBv3)D*rrIgN`t!+jpZmhL(CxywW_|kTO<z>Lp+?U_XP#
z%+97<q~J(7YLy7t<DB)$0(=DJtZBa)=@Q8`C&QL{{3YXV<*AEba?Hm`2BI*CN!Cft
zGhPK9x8%b@gGXnqY@Mu;vJ6%qVIrZm-;~kcGOdJ{2qmP;s>E)lShZu2?MD!h?L(54
zO3aC9Iv;hnBJ3E$Q;?lhP>@c+-pVsSG$R$j8TC)tzB9UvjX}VRvpap+XIqW{y1`Q=
zqndG1{A4wB*`h4jk(KQ-LC?-6<4C*IGrebvEnu${3mf_L7|dNI)^mw84RRqxo_Zk{
z4=I4|L<W<`b!?Fv*5*VV^}aohqSAoJ|GKL0mTlH|O`?T6QO8(cPRF`vvJPzqCk98^
zbUrs2EHz9&bz_UXMT!sMPE(Rlt@%ExrKy#itw+oG31i3E96`5(P~#L?l_%ObYgdY&
zC)xE0izIdgs5;8-n~ju&<7Duan73b%l(6HP{}XlMsB6^~Qsxy>+6FS-!3^uo%9*j2
z^2hW2feo=oxj!+|88?dlWZFQrr0y2oM@!ufYfYj&jZ3S+JFoXe=HtE}@92-{Fe3|X
z7vPq@dg-CPmr?te!m6!tGXNdAII^_Mf5z=S7-VAb%17PS=eE$+SF9w?;V^NEB6dZk
z3IK7+ovebL@}!XK<O;=6(nom;>`obWqx14JQUz(ulne-mj=>~}=5UCM?U5!$;@{tB
z!{!g5VoGeK?$+ujz3BG-C33ABom-HaR+UytgpYAk&!NmNxIyrSmxbu&NMbvZ8(a&>
zt@zxe3o789E&n&p*{eQmcDp}C>^O8I2z7_%Zh<xUz%GKUvS-k4YBqoLz-=k)ZKQaz
z$i3t85Z`SAnZ5|HCQRWe;xKWnIDEH7H<HyY$ILk<X+D#HxfCK={EW0%h==7O2;>zi
zZs@#DjT~C?VIWt*RjzS+KZLu>g4kj+8sQ%sT|R#;((~~q3ekjJR<W|h%l*^f7)XA*
z9J$GvTx7*c+3>>0+0oX~h-t+FGRbHDCeK#ntTa73jrMWmY?&JWG!D&%JVr6{Hdse~
zP!tC+#(1wW<v<e^_1GKooCx^};HK(~8{iOL&wXSe00$6!hS3%hiE!qdvEG{l$sd&>
z3rnL~UKt$<e{QLjFG0f4-Ars4tO^e+XH9p6u*-^G$S_n0acK%1Uu=$#g_Q{U;xPV6
zA&#QrM(!-WC--(QWSZ~*wfcDnAOs#~7W=ep=8{sXK1g!61*_c4$V)4y1VXt1-TiZz
zg>sQ<u3ShV>ZAZ?g?eFP2P6v9Za(I<eJKx@tbtI}MRHCHnf$s4i<se?S@M$FA(h;F
z&9+E#0v7L}*@Zk0c<zV6s1}xw6KAx@y`M~tsOD7f_Hey>A0oy{vAUJI60hS?H7oid
z`QRx3TpO>-HCs63H6GS8suw16H_@hJES3mVqHbTZB$ms)Z6X$v<HLxQ97FPW^OAS>
zGph7Fk+K)n?p@oP5v0OU$qP22$cyGt`6=VtWZ(r}?gyub<I$r^zGfj7$!<?bVzb)K
z9F0xSpc7F?Af*?Lw8rem6_BR*DojQ<+T6zuD{jdQMX^&D?`hEa$xurz`b!mOtLx-o
zJ61ZCfe6qOrS|Tn40x5q43IleYo9_V2qAQe#9=>6z8-9H#{hX<q-7DO162%_CkUXS
z37su0S}vl=Yp^2izryZQQXS8HeGI$o;+y2|0S!y)VC8xVp=i)y7dB-K#OcUVR5jSO
z4$@Uy185DXStcztx!1fGNv531^?y-gF-}2vx08Gk0^YI!of&kz%nk5X;rQ|PQ=+hE
zm=zL>fQgUyyFO6}Cfdym*`qVMo&(Xj#3v;{9YO5~=*)O#ZiE+;zPS|}3HIgbzUZ=7
z?<!9-E0$$t(cH`|EfRhpr=Kt*RJCXO3B7B44lMcEwCNX_-N{gM(>HUdxptp5x`SV~
zh(V_1=H-T>m`DQv<U2c1x-B8mbaB8xxR~N2$kt>ZDhe~<|18@V!qeR>GB<Xd3>P*n
z(Ii`*zyrF4k^;OtLx~XC+fUrq7iR@I*({DC%<W1tDT{trd^B6In5tA!tb#0LyA@D4
z2<Q{`sW+b-1!wJGI;$zHnJM=zRo<**)FD9)c*X6)@O0%Ie`zo>1Ihm_Y86L8G5O2p
zc_YcSPdq+^jiXL>qr;%rnNui<oviVujnG6YgH;+#YzN5bO+t#WMgM9>R6yDqvyY7!
zV<r`n#V)3=O1DA-n4(Lb$+0KJRX`^xS;Q^6SuS*x*+6)%ZHpY>;fOxz;%X%$hA^>L
zxh~iMjzSO#qnZ{N&{9QqX7?FXQ1A=Lp|N7R?kqM<`g!Y{^bWNx!q?2U(z5sfBWUb4
z)&oi8=Mk@{e!#Zht^Qk^8Al=4A+cq77qPj%(=mhRTB(P~{VpWN7fr#3)>fjVDOeeQ
zW!`|iNK_jZ1=EEP_GngiGXn&NTcpL;cejtgPQN$g)55}8?jzxya}mlmMy9u_fpJVy
zAqURH!rJm;IfgW9RYZAWL>@s}wUPk8l#yXY{#sU=rMSI<5*d7w!OTUP2{DgQCf_tk
zRgzgr$K27rO43AxuM9cJ-p(3l_&rMV6GS6pR%p7}+564!VH9}_|G-|@*6gJpqc9qd
z;K?(YX=R(SDG%FGMMRWk4gxju%B7=cs<Ch+DULjo;iL=n@Qw>Yi<y=vdTS1%tM+L7
z&CF#Mt{a=OF4!7}&ilY;m{W*CeKbP(w;?yMO<YxHb)Es)<nkAT6axoA3L?z_SA@U#
z<vA+C7m51Xim)zbxN%B{3VWQoU<Q1G+ZnbQhl!=crJxc_t^TCrk21S(8x<R<i!T2a
zO7Lw((Q8nFPoeYVakNS~DvCQwCnsws>!Y}nnSjKxW;(UB%VjHn6lF%RcCZlgkUWA$
z^0O2!Wa6tFzUJWRe5M~)w{dMMj%pgx;$s&s*VRhC`0+xvk1s{6Vf<nU!jVFS10G5y
z+e{~9`=Clp8>)Rf>U_NiDp$<NDh!XMgWx0TA{3R&Xf5Lf!%|jrGpIFl9eW7U6_5%c
zfNTH?nfen8+Kq*L?~+r0zeJik8p*o*Ufyb)qEqYD5Fpeni|~D4gwJxqG9%A7z2`*f
z?hFJ6S-PGYp#x(4^@ofZH~e)2Ha!^d0ED$S=OS3NPVW;pe9ojdd7HmP_QZmf*l&FR
z89(zp1zq86jWfP9deLS2NRun!>T6Ix(~m^(BRxsG!eG&BGmxsU<*LhXK>U6}Z7>c!
zj(3g=@2cL%mKTSDY}-U5@I`GsKx)(bpM?t?rtpaEj`QYIPG;<IPG}Hy>W)ge*I9qB
z1gzIBj6KQ?Wt>a8L^*#mve~8sc`XFbiq#G>J6#lMH2$cwjo!T%Vg82<1zotF6UOYe
z)GAhhr<#1Gt2pzIU~o2Lu(}I~+meK`%&bx6`Eb7kiP(IXt^${Y^%0<n1gxi2CJ9Jm
zL)Xx5<QKzy)|EM4#$Gt^K1D3d`I{RRz@}x=_N2Qg>R^k18DMV~_ofVBCE%=vc%`q6
z)Nus4I4{i>+4aUHmyA^r-GNuwn)(i_Ao#%=Ac$9Xhsl9P+Rqtk)Qd)z6s?c;wIb4}
z!Y!vM7az5JL2RACzex?@y2rpdu2sz}CTG%OePiuy9iEX35CU6F8^SEZ0Fx0G-iZ&*
z+>fI;$%TexUB!pSh1-1(asLGW-K9h>mtz8!<~q7eB^DrBa6Kg7IVpVd3A))zjokQA
zjun7Ixf#jkx@L7Sj>d}3YHFH+h7lqb9ZG-zcOf+}vY!~N3$_5XE^v*zRI;93hH7ZJ
z<cx@6grl<+*oj6Yg%4PzT3lw@X*3gjaWg1aDd1G&{4ow6d1q!>yBQl<9ycTEV|AuO
zBjY>^ZJgZN`{oT@bi-GeP52o<%fh;??G)HA8Qn*Hyx<W5GB^`y`FuM`8I$UpbsHPT
zi7BEVF$`J!ArG6UK@glB{x$3{fSBkmG~^p;{+>dne;?CGOFHV^E&njE0jnI*M||RZ
zXwasd6RA}x8Fnq+2qAF+wNvHk6=@bA%SAsDnFca6;M@hIEPhl*j>dc#3)&@jYUGhx
zK`MlYkA8qY8Un`zf(^<4YoJ1FW`l@nfzLWa=!Wl%NoZEIETs!WmcbG)p+V*gI!>5|
zqA_AbkQi&G9A_tl7-{Wc!3rl<36g~fhl9|EjLv}g=s3u%$+O6ScE*VAmB^66!ayz4
z$Ja;bG&gJEJTt!`OXrEz28P^6EIlvq#j~UV)>efSso%*rIms9njx``<rp&it90lU=
zZ6CL8jrrA9doO21KR2ds^tr{{gbE7mSDdFHy^o7~&0|$dpq5RX^ckSgVq<Z!hw}nZ
z#)pGW@<TH2q!;d_+R8OEJ8)<AGWw4(dQSq*-3^t`XQhv_X&5G%??F}XnK{|phRMIm
zJA;8BoQD6izJw9c8$l$c5}O&xqm03$jDgqD1B74b3ABRT%*R#4?N!80eAYO<xUZQ<
zD990^+v8f=ky;82X5lkJ;)SnIY`vE9Vt$6`g~;P_tcp~|)M3VFY8{0^L8?xHw>j!B
zyi=AGtdq&~LUwp84%N~IN*vt}Q7)2Ptq8=&1Y#662~}iigpCPCSUKZq%*P{mviYV~
zUMfuTF7cnbl`G2@q&29?Pt&JBRLhK(<80-pt?YxNo=f01`PaYZB2>$gS09F*$z3zH
z8NUc14mTr|t^kPhb1vayHDc<KwSX}>S590V08WigExJYs3dw_<6P3w3)trpQELr)`
z*PJBtUem5KtaE?zfFleJafwB72QW8hs-bg7{0m#jHJoDEh-(>X;4vpsk8>nbWwMBO
z*{S_4c8ml05S+E3wl}+W1|h;NEnqsCyBa;+yj+*UX9o*9;7h!Oh(!2+vG7;E7q_zd
z8_V{n3WA{&fI}C8z_7r;SlJOU+Ch02<Mh@t-aX4EKLFrS1-N`UwDf|2W>h8xa+xJm
zrJ47dw^_>PWT~{F&M0(3`fFKtBQ8T5pOMD8nz;B1#q$I~jFOwdU@Ou3x|K+!@^tPP
zgPKtUAZ044g1FHROK}6K2qSB5JE}vsY*DT>VN;ya*Nd@mC(af_CuHujUcaEffbHR}
zXKg4PK)64tN-hl$2-qM?(InIhBIH4uvHAX^b>CkP9t=X#I@?&KjITLl-|TUsHtI+^
z*q&76u>ulNH}zlTU8JLSE$3Y*i))Z^UgVtT7O2=ujB-p6dS(aOyCa!QIShCxB(h!J
z#~FvP5sZ9qgQNVCN?yl_wkU-cbZF5g<entF@Sf>ar?f##bhr5q!KhqYKo$M$MuoGz
z9`ib%p5bH$M4V-%cF?n^W0uM>a55VlZvnhibJh&zkQ&u={V+IV={2L+sD@@W%abGS
zS=9)6t3&p<)8bv|w#*s#EPi>PA1N0)eXMmfIOPHgZd9|x-$SMCwuXMW^kF}|Z*1DI
z6%7Bp{X01O{<Ssd*JdM-hY7-;<VRQNH-gCU5OpgG$n$)pmEvmF7UO^0NqGIb@O8F1
z`Kw>qg6-Oo8SfoQw%!L1E~V_jqs%Z$W5Ut(+%~K$aG=dI`cZt0DX7JKLkf9E5tcD4
zTJ}+9266rct?`jsF%+FyY;TTjG|mn7`by_(W{n=z07(}l<oQL|y(bmo%q)fG(U`oC
zgkJt!KXDngQy4|L$7<pTCD@(fUXP&bs*B44eA}=^1-S+38Bc7jq0+KkKau}phSZCy
zXH{f2*!Xd`JG6m$m)55Yu7*x^y~LLIH;kXm&CLzR=D22P=nCum$C@iIg7gG$L2@^?
zRSEL2XnaE+ySSfK=S8RQEuJ9df;^7y6eD1Jo&?fIJc5xqy+|48qFX8me>s`Cn_1&@
zt48+H<zKT#vo;4=9KC?Y2i)pV6Bl+m0HUJ3h~5qU8J#haEwx@u2jbai4g=yvi-oyH
z)QoSe_rzc#o}T_+$i)O83XNb3&sc)pnq90w-}WfiPtIs5$rM%IZh|eOzn~{sTfj44
zoxz=S%>aXLbOs%yH@tRZwNoQlIUQfK#4#nr@+5?<l!wjmQ$|MXu9)rNWmlSwR2<Tw
zNRSnID61qZC(#;OOV7y3WM}#RLZtL^mLF#f+q0)Qt3iaNPG915OS}tlT?%TwVrGGZ
z8`gLF-IF4`v%_uRJ7n0xhxBi87jRnGQq)k~>+2`f-g4+vd8G08<V?~IWgPkw2bGGl
zf^f}oa4v#@%stjaTUPq0wyh=RA;?51A_^~aaWp{L=z6`N;yu93flcR6LiCYOzpKQZ
zo4|Lq(}>VoKIQTe6Lhh{m|Q{%&ek&UOSW7B5LRj}N2Bm$1*m5Fe?sk!IM~)|QG4_e
z)G8N+{bY^}VHsIho@djT6xP=ja4+?9us>p8KS4S!W!XsN`JAK#nUVV~uqBXa&OoLt
z`GOoL6OOLyoGOpDJR3qrRFNa??M_yW*g{oV5U%;SGcff&2E(~5`^xn&?+C2G8cA3H
zNcF@rGv)e!p&GP*5Q?7(PHpeRlMfFg3bL`Zg<yZ`2tk=l_^k+20&TUx089hoZuV!f
zRS>`I{FPO?*R4}I2aB4@N)tj|nt90VdI*Y&=FFTVPG7>m`CDmkYBdI#TpFLt#V>LC
zP0pmQ`*au?z?Z90;2W6f|ED74!bTh96A+@hZ6Vr>a>%}r^crg3^+vY!g+Vu#<mf4C
zl;Et?m^zb#C_m2*)@~^vPlQxG16-6FEB8wgQyTG7g~)BYE);RR8@<`ghRumgsLF|0
z;Ku88{8vG?K`nA1M6lGYMNt3eUHj!>{C9l*j5h?EqyN6^|0#U#Kae=g{9p5KJA-cV
z<)Ey8n}fY<^!_J4|82e^%<KFQUH|sqpTM@^f9m@2Pfzl|KbZOj{Ij$bYgYcrL#!WC
z*ZyDhHhuml&r8Ah?mv0rq!mm4<oVUW&VBJrOjZB40@<Lavkj&a`hL<w@zH1h`9r2s
zyFtUZ*iV`f)Dt~v*r2#)Hp@@C{i!Fo1<Kp%1x8Pr1O|iq711B<o^k~wJ^dlihj#*$
z3mE!$4G26{ELoh}%l?yxpjY^w-J{;@w-7H-((`%#4qL3MJGkqUMp3Usx9V|zeiQc!
ziRRw;&WU3SA1a@g3zYPpZSJw8eq#mYRsIVhB3vOh|92tO{~)ACY@hh5WN-f8T2y#X
zjEb*P#i$V&;v+s9o*4{!+%e8w(FJl+nfcpsjlu!;=y8HL+1BK0WjOy7J>HyUbFFuk
zU16O>qY9&RA&S9&O@UvD-~9rp9B)U@bFCXQ<;vwWp{y2?<o3yM{7;@L{mK$vf#G{w
z?>SF2fBSuKR%Cpk^SH@aX=T+&XkTdOCykznZa4Os!LGAy_rNa^UXh~(-=0+xMGC&X
zBGi~rXEVSg_V4NI4i0{F|1h%21fj~9LK?)Lt&7kr>T}wZ%@|-^m43w|>FqIQK_1Sl
zS|@-wMvq!|^|#r}hSk4*aWU|yR?srf29=slM_(J^#p5mXs}+SRVy2MmgsL^XZ=79s
zZbfH!h2YSsb@Pqv#IrTAI{P_|fVPgs81(#8FOougO{kq;T-X6k<Hx{A7Fz2H2EV<M
z=$PV-xYy|waQ#&p(>e6~yhhW@v7q+>#)|e$V8}rwGWl)eK_B34d6f3{?Z~@X<hU=q
z2=CV+-%DjksS(-NMh82;xJF8=rmLGv$_m7UHfqO8sU@?^DLPZ=IZ7Z@g`E!*l0;mL
z&TW;F6H8Y%-@@E;dp}(6z0_pO@yYvhr!MHOYO)^ASuWp?aL1Q4S%+cG6RT=WB!F0P
zX#7}Z=+Tl{d-lH_*|l$r&Anm)67mvty-CFq9X|7tO`2n6^V;Lr59GPcJqjvOga=dm
zpAIgr%p%q+X09>ty}D<TvXKmaPBt*6%D_%qF*?5S6nwWl%z{_hh<8uf9$D4*N(+aF
zGP2;3g0ZrBor_ojC(;V{=(v<+0f({(J+LTQ=}B`~RN0pBsxkHM-BJyPo6>`z|Fl2x
ziY2q>*i|3i&8D)&7Tbc07hl4=PXiAQ<WX=S+f<LsfN${R{K@mkvUq?k*4N(~+U*&@
zr{2*AE1JNE0hjpsT$&~&rQr9rkkEpG*0Hi6+MOw}kYanLfNRc~sDRUuTA9H8If46(
zeHf?bUcJiA_gn6Ac0KK>V+p4ws4UtRb~!+;J+Y2+qabj$d17z!NL9cHm2$ZEm!?Sx
zRbsPi{O;LUuwo&lMXI6of#4l-CttzDp2(bkGk^V4v^%c4W-hHZt;=)%lB82JXuoOf
z*uv9_r?6H5I!A=oF?O)-lhJsoT2j?h=ehc0WtE{KB^;UehR?G_ziXKu&(-Qz13%Bb
zo2nTJZ5AL%s)C<idL>@QcVC`YxYiP2>~3E-&IZrKJI0HjTPUvyZV3F`l3&#}ub_y_
z>vX(#>2l!b>MXB-O9D<g_5uXa^Bd$1o_o~{-&WTQ4DQh~9NL<q)wbDmXe*c}eg${?
z+VP@fSR6F2)F?>!=eDsivhqJB{d)86e;gw1s;`*nLWxNJ2s=@n$hJA#ZI@psQO7xg
zpaacJN!OQSN&%-ok0m)ZjF$$5a#TzH0GnBvR0=jRW8Ti+&V6B0Uo<AypkLZOx}?E+
zC}$Eqw?8np>aRE2lHDQ=ETu0g#e{0Os1c9mx47hbZjTijlh#$P$0CqXIvJ`%$`(R4
zhU)@{5E5=fId!7ZnNRJb)b1MQN#j<AfeDe$*qs<E_pDrd9C7zygY{;TOk!x0_IvqB
z^mCFxVn`9{*pLD8wRWtgZgnkKze-w;9_Ufkurw(AFFOVIvOw@7S2?~~NM<MY)qTAw
z;6`5vSu!Kk-fJ(#0`@cKKYMru@GH76Si;66{HEPOfk&aQw1=gpHqG4$e6H>w`PUog
zItt9evxfFLH;-uv2hn<rpIacc3ezutJy1$_65I}?7Df|VHwhfvC@)h8TKfkvQfusd
zCEe(i1W+NUQTQzQaE~i_Y5hJY<^gBPU%2EtNVv?ucU|hl?Nzg3X~R;(_c->FIQVV)
zY)u05vTst?<@umyfl85rzUyW19=@(7K>arT%|^_oug;sZbQNOiV?R{0w06BR)m!0q
z3teJbA+&Ns3iw_(rd&znbuBe5Ek<h&NoYpP&Zh3(7+kL!oH(1}F`6B0wT+$Ptu|^4
zN~?{I+F<tYoD_iHEcFE-uC(;nkQhA^h}?SLbKDeu#DBsd3zWWmdA=|}NPcI>4->~f
z6~M27qjZfO3+e)m_9S%v$z!THt{fCKc+au6ZtO}D7*#^oeA>OfgzowL`Rh%(W7o63
z+qO`@%1}ufG{y#l>T3(0Fai&y*ojqj9(4r+gY4tYz|<JQrUsm3U`(D4r?E8{=1U5o
zy7?J#!h%kiQOHOG+KbA}sHvj@Tfgx)lgdq_OE&*V7|BQDvd8#|QDEg^d#)1Bdn78O
zhMcto3$B<UNInV~wd8$bV47XjS(W&BWbaPL6GLp44ZtsdSimsA)=(euUGr2o*UY5{
zs138rtH*=dYCh?X1$8YX(T^=Os8xLq06Jsd?$v(gCxQHtfwzORsIi6T^9pvj8Za5K
z4drfLDFef5g|ngfL|f<86@s4)G3J78b)0Y2@&=~Lq+ufd@gDZAm7fjQt@<P(>96Qm
z@f?&6;+C#fTwSpRHuT_A@cpPVskWU{Py|oFPTWc-uFmHfhY!B2lrrf>hA<*PE#I(O
z?Th|Z=^MN~J4J4uU3XPFM|PbnyK1_se||VABKV#kj9n%ane+MA4_plzt&Eyt2QwG_
zue(-%tA01T8>~g`(jm=rUeNqu)#VtmUg*~8dodv~(rW3_vcowUSSw(cAN0c(c#wJQ
z4U%K0?{`ddY%<qKyFAc6{dqnNHu{Kt$y4Y7GgIs-00?eLYj~Ut`>P?xttENN>suml
z26E%Pt`_D5?2>Wk6gm}{ez-PFh?i^c&BMzzYwruqox+|t0YARR$~dD6;6p#+(jg1N
zWLCNtdJq#s{Ly&;N#6Cr`hB6MuHtN+-{SPOZdLUGwNORcTlTfD3->A|^M~2<)l7Hi
zi72($@t{AB&Bk{ExKs|WX7GQokmAGaZAz*$2M>jfHy92{XPK5(hVHq|OUP7zJSjMI
zcc+q!w5Q<n6B0eQs<iB+6z|8K-xoN8*uM)jHYY5$bThslRQ>G)w8?BpdQxz1o%5)j
znvuD1)CXbJTU7?yBeGLa@4=M>jsQ;ClogVmcJ`4bQUG4OMvjo4TpKtqD>iei-I95Y
zUlRA=K9FOs{tc}3P(t_R%b$)G^s0CC*m2A}2MmOEFWO)0zp9orv*%A9KL7{RhpyF4
z9hwK$f6DZ-*XPUN+%BP7TG#<#umjm;Ug9r833Hzab7S~;5im)3k>C;fRFc%Y#z6LC
z^0mnY+MPnNkXz?OJAy}w^&ex=yE&!d3ceN$Rrc-k$8)w<Y1>@oA7wv&*Qsq*ANbmO
zq<H_qUbpJ0ezhrPM*RU7KTw<tC=SYRKCdh=t4QeB8+1QgKn)ZLBVU7eFBlWhIl4gU
z<$mTG>xK98m%w>}%|lgPpO_e-JG67YC_t@s{BTWSP`nK79oWq^a}7Zy;D`>+F5}ZZ
zPI~t<i-ml(*7r*pnZ<;xF_Aj`ETuo1SdSc%ofLZZ+?^1IYzn-su9YNZe5`RpLZ_>e
zrJ7<Q;oDG^KPJ*pCq;GyWCa0p3H>&b5gTSFWpaP1?%aCE;D_BVG198j#o3BL%@yCh
zCuI_2KKCJTD82!Rk6iVz+oN8P;E&lV!n`Kgv^`ot?bqb`x%Kuo#AZW{({t;wchd@y
z58#f>t#=S!#nmE89LrF1uN%t_;1}mOT)hdvrDCgI8<TjMG@%Pzk%KZNeFDxEeH_Sl
zW^i^7*w3&A%izMs(qqbh90Qt_qm!jD=5twYUo9K0^;*@jVi5+>g8(=XdKl{1Z5_o`
z-JxCBGls({Q>JrSo~L4EN(Y;?hf@f@o=lxHJTDY4_~t>Aw%wa5^L|<P7+Oq>1a{j4
znWqQ=JTUshDWvsg@8$2C+eo}LziEfyJAX>q&?QpS2sYPVK%N`I@@kx*GPm`8<^d}}
zCje52n==dqt&36>6z%J8%i3fV2W=Tn$@^d@dr3)F?tBYeb$;@@MX0H`)@@ldyBR>!
z5`n^zMIVWp`0h`<J@K6`>@HZ1i2;wQ%ZnY#oOtTAbfP)P@Y{26{QoFDHv6!1Nh@;B
zir43fNMZbT%&BVkH5FSwTbmRPX@%!o-`Ny@NdKyQ%!AIAXE$8kcV+3RUg3}dcnW@%
zn(H~7@Jq?QEwiRr%Ia_pZT(gGj1w|S$}7zCt#54-TS(oxU+F+G;U#bZJ5x~A?xE*m
zMQ-+OxGG=UGNm1>FYAHducX@6zotT@50NgTbU+j6^L0GKAwmzk{E2JwZ(IJb!Mqv*
z^dz^r2H<JahS=K%zYW3gegXhR&d&etYJ!#TQ3rzz0vOQaY{ylA4JAH*KiQ$E4WXK*
z!wgfph!%hbrw~~_{3(np+`kF8H{1~eev{a>8EwuTmC%D9JOX_9(+b{cjh=bi&hM?D
z=t$L0s@!monT6fG^%z4X)RZMg=-m*ZV*$SYkgM#Q<--SBKlLJ|xsJLaLYGsLqn$W*
zsCQ}ksy+vOP=iq&*qdfx8zVHReK5AYevqq@>lSojpBZQ}u7lF{<+8^-nSE_Q4R||!
z0+llPb&`Nc_|~ZQbeIVAHNYhIhT-+^zk2LXo~-17VfOK5Z^FTo^odtuGC){<PdiK$
zX!hB;R{)*rkilS!1qiyC!>&%rFWZ(lgqV%TE%J@v`-5z;^bzxS3{&NKM(GS{mFc#E
z7r(V2JbV<06}t^!paFOSa~4MbB>*N^cGcJ_17724n2cCvO6IHUDwg2feD&;J_VFk+
zKpNP;gdLZ1#eyD>v8$yEp}<bDZT{r(1P(`+<2|PUJ70eOYmz=di(67>&%OJ`(Sx%3
zfyL?=NqK(b-(Fb{%TH~6lObzUeCXEFuepf8+A0U8pb!C+5r_+w(H@q6mLXWbX0oOI
z)~x~nxTcW9^3f*eGI(EE-_}`Z2)b7W%el2Tu^1t%Z+Q^MBd=O~000Y@8*=YcZtd}f
zd67=($(A{nz+&YS>+N;a9lFdk{5ZYp+mlX=3lZdX4O{i71QwHmm0NTGp2xJw(iu$!
zN|-6&21XBDZjEaKWr_u}p}bt>%miq9A&vv-hmo+yZOc*F71#Qw1i@KPy(V6rG9o`^
z*&h?ypgoi_Y04-J2G&D<Dz0@pFFK?EIL=j`{mK{R4OL^*&9%V;^xP@M=%vDQ`#jV4
z6lZ5l#SLI$tX618hh#+y)&Z74TKtu(Omqm(DbE~z)%Vn2DfwoGu+X`e7f=#4_xcWI
z=>3op7^7|=DJ>8k+IZqq6KMO(*qt}X2JLn0#^oCAlyQ;<bNk>T+ah-Y9Soe1uQKnx
z{r(Q#S-@`#s2eY;$$vNN3}eVEz=e|>mSI1AEOwpoB|sd8jliY=(-y7uuo)QiDMs~(
zx|Yp@Tf|RZFX5kaEUdD`sFn5zDfo3D7dFIgnqvhH_FSGQGe9*XS0&D^^W3?+Q24hI
zQmT0dVqUzA6*!~rpQL6&AaX_i4n)EjBz}+misbXTT9_1`gLM;f=CNLkLA3sS;r`9s
zazHGoSjBPzkOCHPI_U6_^n-@KNZTT~Y=}R>M{+SanE6SjV3wDEavpX`XLqrg3-^YT
z0doZSFB;Ih8lENq4g-(j-bgZ+65=pK^`(WdvmbKj^H}k)GN0+dHpUX)ASp#UMi@Qm
ze*Z%QKo_3!T50#^#Os#icm48eQbLT_jK!z@;20h*a}4iZ2F^A$OiJ!9!vnk$SXb|2
zydw9sq+06fd3j}iZHqJa$~gQ2+;H)LjhJX_0!#%XwC^n0fw5(aa1~}Uf&hPKUsD0o
zkX`^Fd(MjkH3J{dYhzV4dem9r@nZ{7Q$QDeQY`q!0LCV?ce(!jki~B@gt*w-3qY|Y
zxs@m@65{x@fq4>rCaET`hX=-f@9t(p!SuMRz9Z68Z>qE0?vQM5D-n<*8TI*9tyw##
z-ZVZBzMtt}y3u3&wc$ucu<0-x@E80rYUnapCpKMJuoEDJa61^43=40mN1jN<q_0z*
zF59}*#_gBD>Ee1J%nSV?*+&-YA0A6hy?Nuw?~Z$sAKDCrO}%N`Y33#jK#x}P@_KSS
zK7axpIZVrSUoOo;yBhQy>{6i4{>nZCy&XRa2vp|Mxm{*-zmrLeSeSP(;M(vV!zmfA
zmOb#)lzRdj@7k(h=U%Fhd@rBu*^l7Y&CvIlzX%+J2U4nx1DgETT{EOjVrZu6?7KH|
zv7+hvhGGHR;?J5h*9Z=?JGElw3_W%di~C;QZM9(>D$W8D-C)!JoJyF+f>8pr%XL;*
zl+tK@BGsm^-#eW&8P~GWnTt%OHqX9?*;W`$1vHE)asBM&QwrQW<ptnRb3D8cfYvSQ
zb9bhzWnT3)0pHvwYR1ceWyu7n3UC-#zWaBk0!n56ed|5E#okOAHpyf44NW-T6qPE#
zjfbe&;NssO+$5~ODw`tLKmZP&J?!kBUT3`iYF1En9j8<iT=+c#dHcl1H`Tnw$uVz<
zpoXhHuTo+0?0TcahJ#Ft=xm|S$m-KKk86*YAx^Hpy1&ZRFJm|!e-?Z+YFb`B#j8~I
zi}___t#;uL&Pz<}n+9VO*U<~rD1JL1Julg;IU7H=FuHKBT!WXx#!t1(U-}2e?O94w
z7ECa-O|;M@E4ALvQvpIaq`QwpB*2^@v>|xHaLjj^TmZDhY;Yw2418ysHl%#g8nwAr
ztJe$MCMJv_vhH4czaeEl1|}~=v<BELvCx!dgY52J0k1?*b#)%_ql;!Q0K2H+ajiR|
z6JR|)P5?pD7r3UtFXOI6;Sd+wTvy`KUiZrX0u3Gp1ic;XGjW{Y=eq`APZD_6<1rt>
z?m1i6&jp5^m*<P-jrrUDgO8BT_1uA~iu%BHmWPtBh8{?{?BO}Dy!EC*fxxCO0-IQ8
z^(xNU-}!V<YRG0!^e|ggEn9mha@VBdu1Pg8FVAnEw^^M8>|6J)XOSKPn;H=H6%qBo
zH9T=vswu6$l!}%bkdVsS=&6taKBL_pd@_ftDu|zzdbL)qZQMa(H$5o+#24|AJz^1{
zY6Qwb?8dnzx21_o@C~vT-wpOHc|Co=BXADwi|F*4H-_yhj3Q>?po2?L?nh40p`$-I
z0yql5<Cp20FZ%KcMvlvzXt-7q7ei?7x0K%~wR-}#@L2XEOfU>8<|3SQZIjuLd8#lH
z>%?Kxf4Hs`egu$1YCIQsw4@fj)PA4BK?VSGbZVV?pOUA0fC_*GETS>Id?f%Muq;1H
z{$O`pD7W-`9@%Fx4_<0NOqtr+V$ckO6!0Kzg4pGTxmm`6FFir-aiNFISFH|xWxd~Q
zY&qHz#&KapfPb!*H!L7IJiwkOhyyTD@@;f*Hj~{0_PdyR|I|9RO%;HD${YyHr9pQ^
z=K`SMo=AJx6d0rQA||d#XDaSFIqwn-AMk7)S98C8o_qGp8id<{QYEjgPm|@GL&4Xo
z4`I>(TmT5buX8Jj&MyN$gn@<~zywi~0YL+ia$+&crdwvC<^1*qrmJ(B0cAAFHhtX&
zFhnriKQ2IPcOfi2vAc$?8IjLe;TRpdOUvR-Vd!gX{oyJDu?HElV4;@re${7>PcVVu
zs~S)XeBQjLlC)-l(tC~-iyc$D@AW(c-cI4jd_C}~Io=M3#-Epb1^dUKs#k#cvb%ef
zHhAx7{l+N6k+?8}h=-kydd-RiyFDWH0{10%DiJpgvl9W|ziNoLp3h@-MzD6XVxl4V
zA>KNMXf*eG|J&=p8$&0>(~O<(iw>*<R<KN_q*qW#-Ouqu(oXdgfj8z>Os?8=Bft}u
z07M>mG57w~l-*9{&A)l1fp&>8bML1%-n+m3){1V>Ag4yi(PnvrSF=l7b9uxiyu3go
zfeotu_8rg!iISe~j-?=E<iZUpemBFQ1GDEAePL*!3_Lm5tpn5|>_C+NgQpM?vAbRo
zFbN-D2j2YhUOiygOLAwGX5h^^fhT&+fZT9JYr-Ke`K$Z^Hf3ucAnO$$g8Pf3aC173
z_yZpJ0mJWRtbN}q<Ts#|Oc|3K%su2c4=b-(e9vJnoazf*lmT<Wc{4oa1>Y)9B#v;}
zoE!!%RauV|A@nW<aAF=}_q_EIhD3}FM=Hwl?%RQ$IEccI%i<hy;4N@c#IE)CaF-gk
zL4+@`S%ibB{k_4!hqaH=@CUJwv9iS>V2g{sF*rNDJHB(#6L<*xlJ0XP5PECcwLS?j
z`9kr#3;?sPW3YZcF;?B6O9=SX4cpCza$ejdw9U)Q)=#3_1aCfm*Bwo4uohe$<th5w
z!1XR^63w_tXbCU^L5=&6oWYsEp(;;x)}WlFG)^XFuHN(4{S!L}Ky2WF92`O@pS1W(
zEJ4&j`@tFzWDvs|;A@e9)z1p*Gl&?<c@XA$H_H<}sHx{FXra8{RsYV^{ki(UA$8Ad
zFi$lFlT+$ri!Vbteyr#MmsbriwsZW%4>Z5qr=?)yYf9IA7&z5jhG%(izp@aV%d*$E
zbkMRxY&`X_f62_7LgYlZQzl$OV*Pp$?|sMaQVL`o-em|g4Rvd?7$`33Fra_RlLHZy
z0YupK200rlP%bBw)@Us&JAVT;8TU-+A)qvNmp%|0P6<9+>neZ`5fbjql7@*97<fn`
zHs<VsQOLyOEyh^UCW`?KATA}BsgTBE5>V?L%==w^gZRvFGN((hpM5<sscZ4v<@u?F
zdVrXnKk0$v3!ekn(_he8)&{QdHrGXlC%P&lF4vj}t6F@q>`hS=K1M(5>N&H~VsMRI
zll*z1flNX25+VQ}c9N9`q>qsXh~Wg8D!HqqD1n_*hJ&5=B|sG*09k8%K&!~m{NcU6
zNNq6~eSiqG;6Wr;JkkIa-vhn|wXD0!5dj1Y1julsa(ECN*Jx4%7Q1j{oVOU>yMRrd
zK3g}nAJ7sU`k@BI1A>5OIkNNJw}BwoQwPiW9b|9?>}RH>rG<8R(T~Alr$NBmn3U^*
zxOgKR<f)za_6PuYuzFUSEj#NvD+>fPwGz8L1Hk`jmm8pP9b68c!^&^IF2bbf+p9iK
zNTGFXp^I};I59Il5M=>^;>QZ=3QJ)xyJRsv>0a->ZjPfm$qhMFTrCdJcm<H%a$p#c
zYSj6S9KSqj3fR{QE|GNXKh!rB2|p`)HDe$?+?HotrXZSN;3qpo1ouAein0RVq7(Nd
zmDLW!72iEG*ez6noV~DTzqHCj=4O;b{E6d=o#UW}{jNtD+4r+!CXo%s>FnU*7hw%1
z!0fz9voZf-5(xhDU-`xG5s%HzEz50Tq7$zT#AFvDLBV4_hDi~(R(crLg22BXAat*P
zDuiJWU<v`*kPdT2&lfwcssT`-{`Tz<dvQ!WlO2%8T*SgF>BC9L=##Vu%zqtVOMxDd
z3%5gDl9tsg69&AysBHHmK=B)sJm&%SmkZb9lJJN2%`Wrz#aM2bACA!5uQPF7w+=iJ
z3gCw${o5aCPNCo2%JRHF#3TDW2LAkdw8>g?|56%2+;9D`^kq?{VX{uLAVARY*C;|@
z?JR&1mc75|o5*hDA=Cm&HAFCm9dn->B%tlF<v23`#4CX)+oG!k%x{bTW8~Cr0^o2-
z$BJ(ni2!I$eK(a&8e3z9U_JrBgfI(K_c90Ydon!ovGm(qB7n;xa5;en?UldfqNHs&
zGz14XMCtA;3krFF@-&^3{RDyoDUXE!WnordsumAmS!PkV-%Hn6ouZnFU3+3CP2b#q
z(0gkAcZz{aa&QR+m~#kOE7h=!TuMY6Bh0lzVI~3;BYt`br=Pg~vCz5jbXa<MoxzD%
z9Fu5+(eC0j0Z{#LqeR7gm=d4R8@o>03rcbqxxUOpb+Wx;abU~ip|y{+OnOG*VPK$I
zp?vSYSlI8yioBek#fA^C!eTe(R)Gc=7>cW!<;{mc>|+vG^_QF)m`IbWYH`IE=2i?g
z1WrN<nrX~&4z^G~;kqrP?DF@b@`6K{DA}b}s{8_%MaUb}1A=*SmCn_mS}|nvuc!5w
zB5<(JwJ-<*HXvroi7CRd3DQoOUzpXSetXW%U6_}Atywd%?9MGrDjZ6ySL7Jjz?-d`
z_2+wT&aIi;FeH*BKUMR<D(=>Wxiz_}z>8XVy%dsPoAPqJSntu02&je`{9TE-yZ+Z_
zC&s@EnOj3T7BBXQpHn>6D`_WH$6Qn`w}+cc5c1HpW3w3?^zQ;W4I6Zvq_fTwG%pD9
zUV@<pr{#Ki{pY2}f}o8L>wneK<P5HzOApzo_j}b$?Cu2F4S>3b>0CEhgJyz7ROc)r
z2i?ITxw;8T*P;U_9RdKI*?N-T!`Ip)Qp?s<QduA|^shIzA;w{)3z8QOhXdjR7~JiT
zg~mY$A}8HU-h*TSJ|~lu`Sx$%WJ585%4GvC;y#?;fICPs;ie0uZ@Z_==P9l4s-y}8
z9EUBwi;$s3nEr@VV{+<%so{uJrcTx@U~-k=I6ptYQ>Af&bM>xgEPM6_gASSWFC~?5
z^BaKr&?o}r7@))cMN$KZT{Hve4)Tsb2)K#`C|-~z#R<!Or%t)4QtRZnG5e+1J2z}!
zg_(!5M1tO#8V*a<`vu3*OjBk#(S0}t;GRP<$O^cB!OzWOVF!5bB}Of6N#>?8xaG;?
zM5RI3V`8GX=>k$4#HfY01*-PwDsyrO#L|U>I`6FKCKX^G2+fB_`n=CcHdAoN-HKJm
z^ytg}dSCJ&pBwP~|B}!ZI;9r@XUD+@k{hhSa08AuAYxSxXOwVb<pAs~SaTzkj`Y9y
z?&2<-)V&Ij!B>9;q*BEz(nbwA=K0-fK)-bhAK>#eZ1^?3Ls%(5!PKkPSm8mmO|OaY
zG0pYk(U|($DCHl?f#0hRv|35mpA;|P*m+dn?SZ_JLO4MXZlys1XYuy!N4;npI#F0f
zqG8DqhGU6dpXW*Ja*#2_Nte;^g=48p#3sP$c8+uPYJ@4+tA?|tG(buid{EzxynXlY
zv#?+=*EL`G-OYFB*m{k`V2~_87-d%tESu`!;C~T;<<jC@z6xXpy=DPQ$26#sS98`0
zj^{V6O45U)`4Qc4e!zX%&J_lLavoss6YIg+0ONP<y&;?p5G5q(`CD^cg?jHQn4-!O
z=k<=lgZx1_dK8=#iX|+Q0snCh;FKY|*M@+<kB4*ozK3CeD-QStjaSS5_IudhE=>ZP
zn<MgjVgM)Q?|ZTxXEI<9XX2?dH#zV0<xh7MSC{!4gMUY=3K9pZ6;22yGFmd6x<M*H
zW2_C-ze@sp{vi`Irl4kd6TpEpgCO4Tbg~T2DM&m?^7=BtCxLa@Ls?DTgG8vtd;ysO
zLuMPAj1NX%Hss-NgIBa_DX{v_3k(m=mg|UtqY6iz)4JzF@AU@*?hV#aa6Tg^uq0W&
z<a?C}6!dW-6+hC*Hm1wMfB?9}aJpYy)ClJ^^r*i~0(3jq-n(LA{FVzWqkzFZ8Gs>E
zf5%g0zAchNN^nYu-H<lkBdr}1@-!_FDx~huR|<sksrx_+E^y{xXXGAFPGgAZ;3jL3
zAfR)f$jwJ!#3Xy6=OO@i&u#O3N3nXH=kn7IQ7z10&YT9Gd(hs&VHkf6_X!D@*a4|@
zb8`TfFFD*@AYBA@vN_bRdR_ky;y0n&6_E9^Q{Qs}O%jP9QZG9d3lhyN^v#AcGNh#7
zoWFR03^H;L#N@m9EhYci`5YqcPRz^@_6?r+b$|e}1mU`tUWX{Jk@{pzs=+;Jc|9Q4
z^{aD(Vf2CfEr%?ne|;{<Nj`)#4)MuzjV=hogzkyhm@3aiPvNa6wWH(gPf1?ZidL3&
z6GDb|dj6X7qi3fZGuAHq7y6Y?M;?s*$#WRwrcT4h8Fsrp<^lDa2+j(edj@C5vJ<b>
zaFTn1fVaJr>Z1I|!t_$6nZmEl%b9=N<iVvBLLtdlBtXFbi@a;4-2$cQL&|XA|3Ux#
zIlf)N^ubq?qMJjVhzILkBLwTShSc@E_)MhI9YhCSErkgF4|{JO*TmVriw}!}fCLqb
zdl0qY@-|qt?hJ~G;)05bwn|)UYxTulRKlRBsHkzL)+MO5;8xVS7ZU;^TC2Dric8o8
zWK9f^Z8G^?&jhs9zURE(^Vj*D&krKW%slff_j29Kb5CsJm8$c?wgB#OacZ1&VbG=x
z1#RFyydM)ai+o1mKUYrIu4)L-KE!eLo7ewOpSZmsnN1P!74i5FYG@k^^}JI!e~ACG
z=C6k~RAZX2moL!EMRET({Yhlc_d+537!=00px@oh!Xan4`tc;`WzyG$I&!*D*UwzV
zzsc!kSP8c1GBlB5VR1kH9$b^Y{AhiJtkOb7#h6FXhlG(9>U7pCI!hpDR0$Spg0#dz
zMVD!H!P@j{neIMwSQx<!h|-WNDN%Y;%z18Mo$y|Vf4KOz!Y{9aIXD<j!8ucjS$iZ+
zY9(x;F7V|;jpS>&Ah<X}PBVcSwSL02G5IgQ(+_aKT*DNr;*LfN@v$<!s!Tt4-~v74
zyVf<=&&rsvOlx=P)q?AOm`J3ZL8%MW##SFe-Rv`omNR6{(#5_KE(^<L1_)k1u5O`w
zo11v~#>L8HZExYd@pW|M`x&nlr1+KcYw^=?%bBm3LxW>qeV8#BVqi%11o0Zh%EPDD
zj$c{!u5wb{(F*2p&gD_2y{1d1y~003E=9eqpnn=1tD>t9)d?dc54n?c1UBL`UkS^4
zJ!5*WCA92cO}OBZp)nn=nyfKhifrFiMac2R;UxW_tA2E(uAV8^C1qrx?`k52uB^~O
zsa|192@8%%R1J<@z;jUk#SQ5!U6!eYzAsJjQ=3W*>D+PpzU-=>p<x;iB>B2n)!@^)
zv4$93Ih!#s^s=8K>hNiu;AQ*My5I#&pK*j?rMMgKDN!0zVKt#dMN}UDddo!+Ixe7P
zi3@f-865lCp*rW1V9~%Ia<^25JTt5Pif>#r?QMU7UKx8Sa6yOMmGolfw7_H(mhHh*
zmRl&_s5}cbMX1X<{gJ2@=x>0X^21_>$=~U(dxt93kwSqY0o+!WzX;7JUE6^rL;|G}
zwZqks192H7i0p7f`Z8M+8O5F8FTEz84Pbqtc?9!hPPMRhaYYm;Q6)^_iQO;JxXpZz
zkwFGB4Q58xUb@IH&t#OO>!ez5GoD2sC@ju0<q5$hxeaDqeaM{u%tC#|s3=pWSuLn-
zjlN6HaSNGKHQYk;y|qp&a}2N6x@J}juMZTa(1&?W-;?V8j8fxz{xdUVdX6skHBle=
z46(!G;|0b`Wa@{lD|kRyT3fL|KbU^vC5=1AUn5KZoe-eDc?TI2R|q8-JkT}svNW#n
znJ!PiDo;PyuQ)j-Pv0w%XL@_-ONgjdzKq)})Bq?MxW+v3CR*|C!OHSV@a&nactKQ2
z;A^|;m>2GKIS_u+2D2IhoO8Z1QM;Vb<&t3G9`;r6%g2`lbF5nPT1W&+&n-4<K&SZ-
z9f6AeQ*hNPzou+c4S0S%<@fJ%AxWAJGfuPFe&&IpRm%-^*z7XZnsO!Sxw4gtr7@Rp
z#J)`ADP8Qq&vgKk-Kkn2Ct9MNe~yEM$ae^0^4{tq{P0$;(+^l285SOM!p~3{n(<iS
zFfQB5ccE*|aa@kLZdU3d<RU?!GN+0VjK9u{2|mr}m#wb0N|Kx@@)t&$^VzrDV#&j2
z<bx^3unyYtCVA1??{jaQWbb=l4ygJ;*Iiehd-&8q*X+UPg?}D-QknQVap4UNKQ4Er
z@ZRIg0hhh61jjF08)Ep4Gi1)uIn3dH72cPXfFg^59!JuvI;0yObB6EME?;@-_wg&+
zOSNmve+Y_y$c?>x<GNpQP-U=UXJvq{Pvx+>lNC|sz?V5!0v8;4BK&hlcKhq534>3q
zon)cD1TGMij4Cyy8y@PUkRNES^cDT5@G5;kxEOj>$BWdmXOdLCTm&)jb6NSoPY{+S
zjmfzzjmcNNKNXXAZ|ws8bqD2kziJ2N1Nj#xp*69K#2!a0qN*Lz<x<k}tk5G7r!o~d
zaYD0ZW&2B!VTa@L^bM$K++d+jVHy!;Es^HsARtk*;u{<3Cne@d(6P+wBe`Z}aEjks
zQ+m}8>~o@o=S`Vaum-$Qs2vXt8bOrNtldHM#ZN;{(Afdd($)lSbdRF)i3ALwpUkQH
z;hE&#S}5ZzzQ;_1mNN!Jz#)APx#PgnLS67{;16a}$j{`u9P(mZPLTA0-1Q=lsQ5FP
zh!IYQ-MAFl`f)XE$c>AU(~`A=#2!lZazWgwOOYK8gBT78gct-+4qt?wiI6<j_2ECG
zzwOKcy835{k^lj*2ACmPuX%*uTW+RA$=<P0l1^2jhC?B`0jA0HvKA$|ia+1J7?)dj
zv{I}Hd(s{TW`XotaPb1>NuR+<1HaGRYT76K6YHg)cvth7A~{~=nn}dwd!5hPP{Vce
zy@BJ+f5ygZdsl|M7?)i+F4J&bb}j4)`Wu&R%I5wpTPFS@8*-<6#5^e>5&~S)a*2c@
z1Zlp8K`4AJjXS&6YFi>__*Hv^iDD8L_=%+2IhTYU2lL*JW9tz$_(otxv3`(*=fAR~
zOn-e<K5kBa)nIv64u*`OGh(m4ub>aoWj(o*!sT_mAQEzDD*T8zTpWDS1S5ROlv{Uf
z?cB8?Fb-=c%(#d(G|N>)DPVY4o~+zlxyw{j89?ZhP>I!CtQ22Mewk1iTpSG36;UT#
z&h%>lo7Z$%FJ15~g2eleUkP30mygjOUz(g7bIz2vGRgGQ%A8BVv;FkKW#e+qe?W`M
z+Q^E(i>f<(Ayhk`$QjYb-*mCsMDs+yQf{$@^5Z4&T^Vm`vX5hhJ<54|OFX0CMZ~Dg
z2`(WY-Y=gRrW*sl(?gU_Mv7sFIv9ym)tXXZh#QR3%RZ4an(O`A$6AejSqsb=bC;@!
zta3-|od9=~?aGJ|nE>OehtLaaoA9KQcvlJK%#-;BC4ect3JM}9&YZ*ML0f8MSBXiR
zAS-duCKA5`cIF^pC5E_N1BF++vjij_bI08)lV8&Je=_Im0$&p#C%8{YAOPM&RBBs6
z8)d};=nTOU-e_d2B?J7v*qW_gVF&me_<&g0FvYS^<<rA@KjF$~xhRbHu1XVmbA+qH
z#(@iXUkrLg)(#?oWXpjU!V&N7m+AN=8*;UlIB78FEr=a41`?JCYx{$Rys@emRl&94
zbxD;I0K18o99}-IveA|6dYiI^OJ&PpV87R$)RjdQAHHPT$Ib7uG66<RbSdy9FXel5
zf?3>AVI-531!R(F*oWL)00Uvc@wlYO_5(Yx#972lf`7-`jJ1>10EiBSmxK<xBG*3D
ztl;LFUkk+Ee3OhE1;6@EchsxnGpm-Yc2E|GJ;HbfN5sOs2vhusjF+4<#Tp)4X%1_c
z2*t&VI7<ZD-eqf+7ltYyZxnj$$y3+zBwBtP0PCt)tjl|w3zZZ*q+d1djr6M&dVFm*
zqJ4FSG5fvk;^`drB(i<P1kDQcRG?2$IRwGlL*yWl<+qY!_~~Lqp8Ie)$jSknl0x=*
zg&#iD5nRYhTm-A0tV(>D1hGf}aBO*3IU)D>WyR9OOZ0)H%Hb;?S52zB$@EQLG;mT~
z;sSFn{_M!Ddwe;FTWFya0TAE_=2YM-VHWXih__!Y3(a_&dqT6kJ%NM5*Bh-q*Xe>%
z2>u;nn9E-TDzk&*61Y`?%5t5L&|#vU1RF_}L#^%|anxan)TTo-0vA}_xr_%AC2?9I
z5)kV7)k0k=EJGm_7tq+sD#9>m?Fj+$b;~1#Q>1ZkBioNIxGjkNx7G2NsvZ)T=^E*G
ztyQz(9gE@ynAI4z156@MzBZEAg4kE2DP~Z`GqVs;3iKtGpeHaCMs5H!ssYaa(^YOR
z=rg3=3}wEiBg`5uQ5eC%YV(m%3Ryt4Fon4PPvAfU*oyfX<i7;n06BtK5^42-%Q7Eo
z=T{|-%ekB!yG#{V_#(#BkM{`Ieg_&$|1F4(LgT!Qfo<e~92j}?BXWMpGa@l~*w$1@
z-<PETICuf%7Ouo(9_wP~ux}x%1?abGJVuKWNmYY+eSORe`Bc+Sq`R8R@p|)pLCguU
z$F9is4zPl7Ppzu{9@s)&geq&uT*$m(1_qaj9TL$paX~DRLx7g5Uj^cX<JH2FK;t$^
zKFD&Pj}5C2LBc|+A6It(G@Fep6HHI&r81`3o7}`I!L<pwvZxzLe#MIW4{+tDx}!ws
z<Vy%t0~En`5_jk=)FfF+u<6b5d>vrK>f$MjVV^(>R|H6M`hK>0c_r`TfvGD@Y*+xa
z@0>NJ8ux@}HGp8UmFe||Ye|?TgfT1nR#?^ofXhQARF_v*bV7hF!q4>v#_!21r1)AW
zRy$)YbTT>0Ds*v)Q6Qm45SKiQ_qAMaoB}bwxRHc;@<c`_kI$Q)SZVYSL1aG?kw~us
zznOz{mP~gyfHi}1`{+L(BUtKwSr%b7n(z6&fpxnSWga*lst5|@n`!~(xspG4uP}jk
zq99)8q4W_NsvcoZ=bdy!g&;uvAJ@X_Hu5jF0M_YMkQ({mQnAORGQHmop~tW?{eUBx
zkh)gFW&E02oX#tYvQR`BX9SY?30?p$w->y8s9sZ-t(j|Q4Fa^s;qVZ*M=vWTSa1l}
zczVHH5m~&Ee#Im6O5=g*HS!)4por7n$F+wrvoc=L;7{HcW-Z}wi(GpoSN_GidPM}P
zrdL61ctcIphQKUuz>WY1t|R{d12z)N9=En3?x?_QjuVXF-~+-85YItU_zboUXu`XO
zR$+vC93)2gnFMYU5&%{Mjc6Q>EYC`MZOq$9@swmxYCvqp_O)1zN2VOQbKaEF8Zn2G
zUuQWSQlZCC{mOd%ElX6s{rh0V8dM8Pf?v~ckD-G)-YOwn&RRrf0BxAf95`u;tr}m4
zghTsFl@DKpWw%d?stkUaG58Wt`VRf)6LV!$ekXq6GGexoQz23V3j9dJdr3kD1FL&%
zc$j-5Y2XjKZUaLNKOt0+zD(l2y>V*oguw{ICap|d82bu{UI`hp0{hFy)gh!)qo8&a
zu@nd;;7iBC&!`gc+}xi`^t)cNt)Jeq*+L=}L>U_XWL7;SpEDugkE=A*ff+e)CHbRE
z_07cB0{P}7{2KV-zG|6a_Iz&Y0epYM=ggZe<j5@k(C_~xpC(wS`1SaIoO`*E|Loe9
z{NVp{rYKuyn`TtUo<k<gTC#4}(eq06L$`d5r?3C`nTw-$CY(!skjJME@!cA9M+>FD
zZJ}Bz3O(4_7HXSV6UX&rZi)06|NqJVOps}YKN(ipd{qC>jPXz8Cu&6}C8Nv46F(ee
zvB94i_XbKWlx_c>k{~~Ug^IN{T>Y5yj*|qrwzW`sfj#(p=Ob$=-&jkR^}9ma`1C`b
zv|#*GzPB2h{r9|v5kl~sf?hJZ6j9Ab<JAxT*g|E;=@qPndSsyvskCqa*>*x&8*8DI
zSqN+M&B2B&3)SC`A*VUPOSF;+b<#pED$_#6>?yHO8TT-;yhI@<DzH!*#sjBo>HIt6
z1O}6gTUEgr&odt8P3g1FJzMxy+oqe+Efh1FYdobb#;hBy)LN*175hhbnalMcSC5xn
z2n*L=(!v2r5z=*S#J_P%Bbldx1$US^T)?ap8C^HJ%kH;R7t-4GhnPD(WHt0gGS{ZJ
z*l9iX2W1`S+U-RGPIk>gb!-?ABQqt|d9Zd}YQ)1oyr#H{yjioz;1E}uvcy6)_m$u6
zo@K5R&#Rk~xmF%miq}amtwdHA5?66@1*Yo9l(ejwI*7R+EHzcgvy3Cf!>K~9b4N-1
zbIHD>wxtOL<=o;BseaGv>fgA{IQ9Y&sXG#n*7xLhE97@Yw7MwKfi-98$2pVU-@g*b
ztj}AhvJ6(t&5cNrX=@bhbZ{e92>QqBXf`gUgu@*5{rzY*lIC1#&cTDsT{(`%>hy`_
zjNV(z=ywDyX0pI0?Orp4X$oPAdepkMlGByq1*Xv4Qv*u4L1DoLqnLB_+9uB~DV6VQ
z>!^Z#6lR(CXV;xNEwWH?dxzJZn*LDIty_SFN`EU3(neaS%wCU-Q)Vmqj^1rSuTDS9
zvz>PLeRR^bJKeo<%0AHve@B{a_htBrh+#}>v(s-UCoy{0`st{BPGPjV<V(1J>ajz(
zJs9bR+Swe2vo}{`{0b&B4w(jeG}y^9S}D62gX*txOtp8Gh1z!qge=ms7AorjWKh9{
zRx<a4AU!mu_@G(6xHSUt_!a*zI#0bBSC-P3`%CZacOKc2Q1Y5{oX9%OHnlY+U&rgr
zZ_vk>jk?f*bfLRyaz7~>Ek0#TKDBKp6fmyr<R0fLdHILeg%}H!8of(?8kuzkL@-Yb
zP%N-6QQE+a;MY)-s7$t=$2$ZvEAfk<d$N;Q<aC*k6Fjj{?+lPOp`N2Hly@D?CKFDI
zUg8f2ty@PFjc9I*I@&x{5eK5tYoUxLp{C#+lLc+w+(=)Y88uL&+HqVXKEmos6oU<A
zQ^$#Suq)=t^gb)12IO0)(J~<`II^WGJTr(%uccXeu*}dbKWdHl1$I@TqJ8t@OyLIA
z4t9AD6Yj3g0<ANISd(kus*<Fk%OD)k@&7&1xn+$08pt}`LJ2uP+Mon4^RANBvF2b4
zB_c4Q1Zz1#E_l1Igo}bs$6#@UentObxnA}+r#@NGr_FZ9!E|9p5r}sJ%v0o-#<jWk
z8?D__$gLiHxTB7_xvK|hHA>H9X3Y=mwQ1_p$^#e$ASdOb64@rRZyleIgpSaImY>KU
zUgV56xch*Kfo7lfJ<3)LP=g2dH-C(4Q(iq~d0_{JWe9hN9OHW)5rcK<y{!mip$>Gg
zVhUi9mT{obj$o<bE)M2Ehd?gZU@+XlA&V`v@*9}KMk@HBZ<|F(;4pnH)NTcCydPra
zEk|wXCHYs)l}D=4AKos_lR=5K2RVjjLB4h}p*9|%%ufB>_|4nz?$BBtU1Y2Je%P{m
zHJ2wkN)0^HTBwpT=Dt8?Jm{txJv)oRElVRU)a_B9ho3mV4zfO@=2BYs+)1PMaNECJ
zRJU(Nt%dR$VG7#Z0Z7!spkSMf%EuHGIvt&VYsrVnSf{plTm#NxW&bp{62u!RLLsIG
z&gbDkJ-vYqd651T5J_7@mxx%R3x}9Ov?%Gznv)H8!kCg$M&DP<WIZL&QA|_;DiYhJ
zA6-(#7*qr~vpYSO^!?=3+v~c8lJ*lNavAT*WY^B)#U!cGQ@p5idr01)_!q+!8t)Rx
z#mv=`-r<1fV4hr)4!%vw^k&~!sI75z&uF26Wi*d>8^616Yg%x#{qy8sVqc0EHO^y7
z7K<+LVA$5IinDL&{PyR)g<Sv9Z{-I+e*NNXp(>mjW&zk*4V@4|<*g#Xi69Z`sCX8}
zR11YaiT=_~+?#C-eI(yklYBzb9Z3BCpquS-CS%s{nq=YRx|83ukaQn(f`?}11;ES3
zFFkSz+P0MHE0;*R?Hvi&1i21*yzgn+5de}0^ic1#PI7AGf<sK^#ZdkpAoPBSs**m)
zQ7NJmBW9w%-P=WE_^@!A^K6l3oV3uvLM2ZEAjvW@d&o7!8EB$j#xi;^jg2F!g{v&X
zG4*zdhCn}f!kf68vHz-z5BsJ({H*-)Q9!;|g#5BmW#W&Gx#NRbjy7a+PCsfTi(IaX
zF3KO7M$_;we4B)m&q1z2y>mwG#bbH$N6G(IRG*b!;qECVi)ybteKL;8JK|ja{kONo
z-%C!^#pmJ;<lgT3O=b61Da_60;_s(*xNLlZky4hFaHhMJr1zu4;-9L%kt}NB4bc44
z<LR@=)eV-(oeQ^6jm-l6vIv?LTi2>@r{H3UHR^Bmm2|RtyoxC#+AhU_8n^(2x{RIx
z^G$V5yoJiyLJ)f^gF)6u#j^cV;4p#ThZxpXY@vn}gKURXoU1@)%F<e`nv%d|e<l-$
z`m0?rc$^-S03sGd`~W;Dzs;k5^Wnv&w%u_o*<Qz}=GFCu-ylTeM*ZZsE-fYcyIQ{9
z=v><4KKz$L`%#icH69PdLnMuq>EY)uEJfdtzUpH}rw6gWSA?IJ->Q6b<-zo;qVp+D
z=he5Y@px#V=B~0{-aU3vjk|pqo!45FaIPJ(RPtLUy4tPG=(N+hV!;e(uDWyFy}G{7
zRC6$;uN+{N{$l1u_0!8M!h$so@&Xv(z95EB08a+a=3mU7kl^Syz$juH;f?~N2vZ!Y
zL0pJw_Le3sBtIl96R$bp+Pz%-wJ{sP$@UhnYLm}Oekh;Z)kacYEnaiH8<VddTl?#e
zge^3vBgF^hufuhxLmOFJ;!ZKwvh@#7<(Fzt`KOt3_M9^2j0ptMW2PCzYruncKh-9=
zb>gC~fPrJ4EYc<ILN^41ddF_8O&TH8%)L}TekSgS**KQ;xuA&2giTNe1F<y3A5n8T
z=nez#Bsi!6bM?Ol<B+xr&bxN|O2{(L!m`MSPc6oRiTh-<=RLWuLnJjBD?D<>@3Bw=
z??`H#3MI=M2XKR?1cu8`!SqF^hyQA<{#BzsSS>e^*XdZHf%oAf+Nov~=7}XEEYu70
ziY+wP*zR<F8@X{CMBvVyc-2Ii0mC<U)~O`q^{~}9rSj{uV%a^d#N@;{hIO1bc*U<x
zB(?S;wtrVdAZ{__0<3nsMs0P^@5s0&CLwo*7f>?)f-DcrR0(;_L&pU$O=Q_>E|Hk4
zLo%8REMssYfFm*pU22eXmaNu)v(-%oqv6xiYdNAk>q*RUny6Kxmbuj_yJqh?2lKQ8
z;g>FSN;Y7epf#gkZLPa_9K!#V>FbC=%+1}ROY*0tZz>nOYWcR?DC@98zO2HTaO;Kd
z-O9O9@wuuWBxkF#*92Rrrj9LN)m=I8#ncPVZ@b;4)!Nm32c&t@4Fx+g{f*>o+`RL1
zJswT=hSH@MKAPx}$eODStXREkmE>%*FD(>_c!;X^g)9Ka=}W*!MG5$QpVshZm|<lv
zK?-7ejUu6iYEh>s=>pKHaa1hh2K9LZ-!HM6$pXICuIEICW?u6_$WS}b#daOK$v^eu
zJFG>l1^J3A4j`t_31?HbSg5_iRgg(oR}s$n$_xjQin!GY{Jk-`s{=QxmbnYIv8^w6
z;-(ictPu0+-#v@nu$H--D$}D{h*G-ECPPiCz+`rI5Wagl<qAC7Q>!w$rQe}@i03i;
zVW{uL<R?3>Y*6eW^@f1g=|c))yNG|(4`YRTgY87vStjcl=1g-U9IB&o*rH@&>)Ez&
zDsC?gi2+P14qB)`tTB+F@3TpESJ)u~#IxpSa{W70pF1{2Rx^99Jk8`>;dAJYr0M)L
z@hr*4x|ViOOoe*O3Q5z}o0+mP0g|SBVAAhgly3r>M0Y${`dzyp(9Yu--`w4w#7cM0
z$v5pn_yncSh>>jc@b3QLb?1TKJ{O%khVGOn9vR!%)_?~`5mo5dyNUxQm4%=CdG(9Z
z@84{#b6;0!q1tredmQ#<y?*ohsyYgN4R-+3iBkquk4I-js#n9qn+}2R4hG&K((OPZ
znivDTn}!IE-ELBGlX$fvhBCF@q36Z^OpCp`haI1YXE-f6Sp8cJ!?vxV^G9x%8`pWW
zaD-&&%P*LyVqH);0t5eWxv2`8*LdB%+Sh6(qs!WA2Os(G<?`Qi(oAI!z}Z8Kh=!{q
zp=qia)egL~CHq2d`s1?k?aVCVR3fy#+3K6>1Xh6lIz|~2+QgOr!YF0)YE?68pMBM%
zbzjwZg2qz^uNcx3pPSA$RxHobw+BLv!20=&LJOUO{RRxC0ZsiCB_P9nhWnwGE&*kB
zVTmM0J%mU{$cJ-qe@SpFEhNJ6HxH;YZ$1AAsa3G`?l7oILY>r%zE95Bjn#>+N$yrn
zjD`U9h<|y$qfpcDb5s17Kz3*!$sb$A`eIZ3OZgj}k3Unme{_|2sVUw?YB+VwxbNLv
z$sg{aEU`ns)ydxsQpZ0zcWQLi<e5@9AKpQtljFD0C2yJgPMBAlhvKEjz~1$@5At0M
zFP5L!4hls7Yt7#24s67km&w;A_QY|}n!T+<fOuO^csx0kCeu&wczL3m;_m?(7jE^I
zWazLT$FQ#jGD9ZrAu~x~i0fAZy2nUCdw?1}nj{n~)Mv!%b&G=bkYoU{iUl%UF1Oz?
zF8QRqqa;H+R{o3eU|IM{`7fy@t6qwiNHTQ7I@fQiL|5enn*kfdp&IMC2f-Cv2-(vp
z@1lhV->`FliXH#CDXxO>#N&x7B!Z($+^UQ}!5JI>MqY5VqlL1GkpHr`d;H6=c{yy}
z&(q`i8=cnmee!h3c6os*u1fTyI7BkGwlBFoCd*8^CMahYzUt*9UnVt7bn?WW*480s
zVd9NQq@Q7g1Uw?6l7xnLDRZBO-U=Zj`w%LL9YB{Dkfp>MfJOr({**Rd&9KQZsj@pD
z8BwOGU5zSE1SwK;@kAfg;FpFD;FhqkD_AF+j9I+X6IUC(^;fXhkR+JPyHIbUuCTl_
zL1H&MQRwSz#2N5cdp6QtY8+O5Mt;F~x16pfIF5m$9}d@D9*4=0Oi9kc>Tn1LqtEbO
z^;A4fGOsrN>3AQy`o_6ji2D2$#ikd@F<i3%NoCI6@G}XHLUl-7^$){BB=ZUbh1&JI
z)Z%HYqnM(0eaV>lz1pj$EzOhHnO;<f-;md7V`Zi?(HW9LC}=#?_=jB#=c(m~5KpL2
z5eXW+3JvUIKdcm9sVglr-NEQ|2ysO3lQ`1@l#D^*=tJV9UenaDNcF-^0@Dn@&oFZq
z9Iz_JSk`6k*c#DMiPO;4$>$|;6(NF75~l~>$usIi<WM|&Q`hjL66ew|zw(Mnb7Xq^
z@8lbU=t5};_bw88IkNk+q;WER4<C8P<M~il^qP-5C%TJ=UX5_Bn3~bObb5#o37!!f
zR6~#Y!VD-%8es`>Wf;@BbZRT9;ZA|cV-N0&gQ(M>Lf+vm@{M_rVTW=26ec1ZxFEg(
z@(|E1B$*Mc7=bog@G`{Hp?Er+6F*vNFbVmDW+q3Yfzk?aylehP{gU-GY>yH&xO^9|
z)uhF4(}E2m)<+pn!pSxL#Ea@wi;CWOJoaCF7N6=K59;$aDPf7^ipoD|S$)jeyEWw2
zTb_Fc-;;b!JoY<Rx_F&r)#IU#SjDchHLuSx%HgJPC1&eRx_Iz27}byW>aTkIEodD!
zlF8h9?_OQ@B!ybnrREWxci@fWQrD-VUyR|u)FDk+yPM?8Jd4$El@)9|rZA?2OkSv!
z6>{oGMt>Ee-VVqrAQ2xN>tLcq`;aCyH_1;H!HgqSXuLBtgZJ(uM~Kd!J53^k&7uSZ
z3QiZGG0k@t8=da%5Fr{cwb{QYc}S?2z{DqQ;zpNtttl|IZ4Dpgvqm+aEuxW7P;Lq*
zw>p>dJs{NhH)0P{+lqyqF)s(n_R-!n&tUZ+Z;uC)K<R?1IeZAh&j8ZVYMgcK)!*LO
zu5JCkt8JbQ9)?3XlLghT=}^yMeFRx!eZnRK3QwY2k}qS0S-i7VFWLnvf|EqWhscPu
z5WJZ3#iGwv3^+Vy+5vFNVgAIyVzvM3-%dV3n&<FXDAK^oUialD`b|b5H`?(0u?<rm
z1FJ-&xIg;$@QoYr?sNK$_j5R7Q$b8}+qe`JpGVR6zd0@Ea8G2k+Vq?DurX6M{w+Rx
zbz4!yyqh^Ji~XhZFy{V@&sGe?>>J_#z#f+v^6LzmYvT5~*ZsthlO2*FsoZ6lIW`iZ
z;Cp8g3l-}AHVLa&rRH;(2Bw6#6LAN*s2msv+)Qj`63GKwWqc2^NMEO{`No&{G)PlD
z(T1NmXQwcLG<k>K$@j0*h&R>lFn;~rDlCzsOJ^>sxwsjh+c>Kq4o3E<S337|&6R75
z^b1Y^`3=~l>v`qTqmC+)7}@Srb2%iPE3LVbvPl0OIQO+*xOwIq{etG@oWAB_%wx{?
zM)C<HOItdu%1;RDWlhg-4q9sb+IJv892U%FTn<TWiOlM40~Jj@A<M+YH%xq=A}s;_
z#vz@z6l{J5kp?w5A^45uL)d}Gd`hc6Z2oIBabHyL%<yy|iFKk9^*JMbLwNqw$xS?V
z$(Va}PhjNqiAY0lz5S-<0UV0^kQ)b~{HJ?2!!O^5Xb|5N1Y$)Fag{gV4<AA*D^@ku
zE130AN-pLRh4y30T6jF3FfN#6{2cd4`uswZJ(WCJqq>~!XKNlD7+oRH{c;q}kgPDK
zkXk0Xd=`d{1bP`4@gJ`sc}w8TyXD)IE~3k(@XOm$5hZ+xBJ1%U85bZsTm@@1i*z=l
z=ObtjazSd%>*<LxRhq=w-6^0QF_8^f<B<rjKvlhh6`qB`;L#7k+o+*&Z)Z5HH%;<+
z$ETWp31nxos`p8!sar5{C<B#0kk2(8IHCgRAan!3^n*;hd5%*pNZ5r&c^@)X8<2n7
z@}{k%``tutn0+PL0yx#%Q*>hH)nr;-(oP(-88ESDLVvzTb*WV-Y>|}Obee1TsKfh|
zRnHJam=jt{y7%fL&mJXW-v@)<)r?Uf#D)(Eb_+2b@P+G+m1jYg3ZmlCqc>tkIs|9_
zo{QhyS$?<ppv2>$<g+lE^^3#c>NJhSy(4(z&<@EI@w@}ClaCoi$BcBfknPqfqh|b1
zXaifL8eYTTpLH{75)z-ZetfY>bivfU)e+o5Bcsxd3cmoEDA?M!fLsvymLpy)l}k;|
zN-`u~?mr*8TR}2l!~_$sx#CJl#JQ<Ggj}m-L!29#lQ9kLGnyT=|C@j9zf6DS+-SX$
zSw<XXauV>T>*3y9Z=rscW~^7vJU3P3z(-rcta%rnhgVMZ=@Q1|`NcV~%~eQlYwj$(
z{0$biK{h;wt1iYcpE)<;M*mAjK!*?8^UUE$7qxqhoja%sh^K+d>4}R<Y;Zpqh8+wx
z^+;+R@f@);ST)CBl7HEYQIaq&izI9E#6TDxd>x2?)%A{~i;q2@z}@~iqUv2;ML)}T
z8-vo7(_V#Nf`Dc=68)Uej=ANTQFm$E4d%XBKUdcGQOGaO^}#xyG2hU!%=Asdd@CUL
zlRs7dAn8)G_q+XkNPl)Hdf6!+wIpBu^zy${qsXPr9=VT}eI1~bbQ###^!4dv*10j?
z_WC>3sNs7Qn0%FZ)YJ&X!RBDS{6x)OJE?){DLGqrCHG0@1!GWeP(#R_1E!Y1)B25U
zG^m%r@?s>QFd(o26(Sie!RQPK2cJ@-v!I}{nCRc!4uW9skiO8<^^rcI)SF<U@+3!8
zyR_gpJVek0#%}kLbZ*`?A=h&=%;35_nr%Mr2b$2{t_@w>zA1Bm8TS^VlP9pTP@V9#
zn9HW3k5Qd3%DAQ%Y-P0x8JL@|J(_nGa!tJZ$n<BDNCOX`+zt}cxX;7kR}U@-^e20v
z3~2K^=FY%;^!I{7uL8K)z}3vF7&f3~9=<_-;vBAfH@c|pI;*rkb=D!J!5Bz#qX=8H
zhCv8$=R*r)tkOZ&sjbgiDoT*7BPw{L%{#~jG1cz%TYkn%BlXD{QQhv9D*99evLlK_
z8DS1wABWSLNfWENc0~$cf9&kv#4XY4n|8EjXgC^Q$YOi^uIOV(zBg9Jb*l3yP_)!1
z-B}eiT04F{zZcr9uXZX;BSYm>#b$ery4!djKgh=07#)<3$y}=uueW~Qk$zAk-pYnz
ziAOyZEh`>Jg_}ART`4&8I!+sQ6Kty2Odcl-2(PidgD&DY28-RrdtFd#BjOi<)IxP9
zzF0a0#uXcBnTScPn-^>dkd|b|icb(LM8X$heR+38gc(J*oGrQ$xheeYw~!vCdu&GS
znQPAX_p8>Ja$X;8i=o)oqnq)W%$vz<++gw7HMekWa!2gr%HG+D&ITU!czxXQTPE-J
zY|~RahHWRLIUzWpwz28^^hW<=(vQ4Wm-Icvw)seTwl4XGr0iJN?oW?Jh;I1X9KZ~A
zVFvzM@z<NTHa2A!!N?-7xv##Rb9v}4**{EMP4+U04KIl88w1g2v26=kh3U&95oU^v
z){=mq&oHQ`%iy76btD6pPgYK1lfeqOEmsg&lVsUCY{&5_kT;^gQRwqm!gqo`oq`1r
z%#K;#-rBWJ$l326@kHX>W_uv=i5Qh#IPBSj`BDQxt<L7L=aU}Ytsi?(i#Y*~UUZ2!
zxDl<99k*`$PB1BopoAC3Ft=BI-)8#=Db4<<K2Np*+iVZQ)YCG*>aAAf+x@gFQF<6@
zuYa|UYttLiacu8qWV3J;bUSt3h)!I<zLAeJJ7HCw$FKQIF6cw<la>;=-a&HPK*9SA
z)-8a)!eIGiMktji`HR86U%#mjuMu%b3%$bJsDZ@go%J<}mXUQ9>X=`*CdlRI&p4!L
zIcx@VUu0TV@LQBvi!iT?VCxZ_VWh^c{iC|S*lIYL$T}xx%w`o>KkFpy6A7nlYcuTS
zCf$LA2wU2fDEdg%5ERFd;$wAv3r<!s4@ceQH3C|-d?n0<!)Oij&?)_QXyi6DT)$zF
z{`Bo~?wS5%lNVf@0phJSTiKjqMW2@20bjY0cOnnOO+>QKm67ss1IcMANobmPjTj9$
zC?wO9K|+e654pf<;?!}85CkF(amWk`p>JhxvgU9-GHLaFlvV6u3X~!XwXyS$TR_lX
z0JY7h*D)X=*?x-dqJ++mT#e_^XA0E`jK2BEG2mTk7CSaors+Nw294OQ&A%`Oh3Ggh
z@VSL*_9&hE{PQDB{$_>72KIl)9ff95e@q^J^r+6V&lOzD#>25MB#Yrj3ZOirx;9a8
z0)o`*Nv59noD>0gL=o$O)@frK(kt2sL<>hqz}X4^teG67|A@Ab?A2}AH9!f$LnE<{
zynnad^|EPi51BT;z?!q_HiG8t79Xsd{?nSt>lc{zdnqN|ehyMCjrRHAk_~Il=%a_O
z<ZlQJ;qaa!7HU5G%s6QXb}n~Gy6*}`%IoE^Th4E%5EHK1u-`c8#8V~<n!C$3oaD}s
z6WiMcOD3t6qF2e+CbR;8cr{$w*F1zHuGSfM;vI<Uz?3Yw!T(_XxKQ2^=QnZg3;9ic
zG}E_JapBdBA=qkB41e%uZuagw#Yds`(ye!jq1A<Z?(MlJdNr@$Zbp0@3n}P6Q4=7t
zZkx}(Q+#1<c6^*bYHAuE2lD0Zys(=8Wr!(3gj*N<RdDY!9Tuj9p&AAv#gDo_tcM?6
z;nyRDdA1Ps!@@pfh`;|%Y@d|%S(vDdNqZF5BV}5|)jd6qEUf=s(5gqu_)KU^?d<UU
zc|>yK?*&7~3L;MKi@)ccJySz|E!1pf0feqUQz8FhjuxkYd1~3@7!YjZB>6)8{xAy+
z58+YgX;$&w;ytD3BCNaMfP{?>TLq`N#pAt{1@Exzw21-_JzaEJ{MARly{9D%A*0Pa
z7eRinuD^?(tiLCA^E2S<hKMQA)K~X}ihwTlDH-16{z}7G5Rhod1(Em(Ia%+&q=A%l
z9CB<t2AAL2bFHCDZSPB=#><1GQM>l-=0DpQzm3kH+Lu3W9t`2OX<FNL>HzS9p>-%o
zb4&+1dQFIbncR_9x7aV9xML^G#7S&X-mRLl(Yb80mn|=^tDE2!09of-Dk7sNz22ZM
z4y~Na#@h<?Rin$Mu4@g4)LvCzC+Fh}0Vd&Fe(%+oi|ATN_Y&ZWmp!>p66D+wS;Z6e
z8F-ur4{n+T29kj<QX@R-j|fGG{__j<R<Er9cN1aGBaXHRLGu=)tiC)8-~#jDjYP@A
z!9AFpdBCqcn}>jlHjI8kZx2$?g1IPLVvKd2-LZ&G;RKa-$71C-w?Ke9|F}!25qh&D
zk21xv=24v&14CM&#5v<>mi|pIpufW^c6t`gHtE=$Nl%}qV2ACVeugQ|m+3b+?7_r$
zj~XgO%C^fCu|Bdd*Qt|GyCaMJW)ofFjMKKGHZbOyB+M5dCB{_5$MRML;>2Z)VLB3D
zAGVT7OpI(G(%J-$GEhvx64(4~&-xuI^UUMF<+7mwe3#!5t99-RG>3CF)B)usgpOR(
ztvHNY@C;VRemxKNh9qp9`6gyWbtkq$TsNO)J!hzXT7MsL%(NHUN$vV*CWABt@<FFe
zh;C+3<fH{K5h(8gjasjY0>K<N|8|1({DW5fleJs^AVe*!2_#*IW(FHOk;x?W08PrX
z`@$@cOzajT(ue3KGYbky(&k_Qib`1KJlOOPwIk-wN!%M#t1!>~sVDu;%Y8FAv{1|W
z4;FIntq8Z)JqOcv&IDsV8xNbI@4RJ@)F8qTo-R0g337Ih4rZk?Zg>QvFWNndfephh
z4%=#UGXT8aft8}-A%`8+3Ep4lopM@p6caAkJ@PR|+H-{FrtcorS?Vj*+f3=i`3Lnv
z&L>!Cuy0?#GeA-k-lX^~4+<>H#GUyhon@iS9)By1ddEOLN#h+q^RA=M^Ek5Qor6v5
z1J9xQNq>sSuJ`x@Er>fP)nC};Wz4oe1MU-{QT8Q1VDgRGDGK;xQ(gEwPj{5*cO{>l
zwwy`b&)>|SJxeoro_lzQGj^!)=_;nAlF3*feu+Pmegq0NE~EC8=QNpqOJgtN)0OkN
zaUH79PFl{WhVrw){&(277Ru0ku}-p}<0i&fYgL%P)`>75veXW+-;iDc5y+sR8p>K4
zq0-jxi@t**8d_MZO8Kmb@pVTb5z&t&NBmQCEK@JE=qazpgGkS*W{R4LOWDcmAio_J
zi?{w*9eyv$nI8*vNcF`&6Xt~9^T*BcoLyhu|M_`UadLk;&#8>f9@>}NI?AIUY4jd$
zvmvP@k)5$t1<gq7WSx@!dh5H|kWw1)kn-^j!zoP;Bz)mj(Y@g_W%X5m(1X$SpaB{{
zCWxIX_}`U4sfn2F!4`>t0J~L&-<Q@ls*fc2+5%ExVaS|sq3*p{bom7m>O@nfA(HJQ
z)Y{o299hOyPy%*{y-c&ZFE^|}tX;a2x!vwAH+z<h_~;H<=AlesBjhne>|$uFdQ|@o
z2>K$R5VwZXoc*4Cv2YwG9S0nA4iM<A`~Q!*UO+g0U~c{vseU1Xs6A`Q29|Zh>rbf%
zcANN384W_j`(o$}_D&iK9bu;V%;kEb!D{&a#A(#PBho9Rpp~&dwqJQ}lkNTwfmpM4
zyWZC_Kd!H3MHyVXp(c-K@{DDj=aZm(T|e>cLq0GZpN(!?F}+10_pgqUtr@)}qfz2A
zX;co2V(EQz2Q&AbOB;%?ySvNO1ugjPjr|K%6K~kTs#|wFq(;Aub=^Lvxmou-^Baav
z+6xBeZ7zx{o3!j~UAHx|_pAQ64zs?hdb$EFkgYQlrfW#L?pMY__5GWEt2yZYdcx=9
z#6KgxVjc-OdE?#$+Y_Q^aMRusRR%$t3txgSlU~M7UD%f-Ae3EpR|Lj2y>$jv)sNy(
za193p70?x5wT<;#DAR$)*!iD(VOo#9y$t7&NfM@|Llj9}_AGLUB5hE8Q<RS7oV(0t
z-^agXKUa*o{H?mN^DN1}?x+WUNSqU7v|x-DCSRx=S6LIb2}RW6c|nay@~p=uXx{8)
zr%M&+pow$owVnrXpUA)hSB0cm-Ne8FNQ~J@NM#Y*#wZ@{jVipmrDt42<2et9P?9ZM
zQUfU=(a<6Bym(}5r2kw{i-JLpF=TPX+P`KD!~zvu#hTkSl&zCUfn;uFY=d@@Xp_&3
zu8S&R`Hex6D^x<i8fgBP0;Ugj%8=2l65jR{?g8CXA4;u7cb6h)*6D=k#Kdo%E5oAV
zNYUQpk?E3C32Pbn?UP2fBt_P2BPUbn&R|j;L0k|=J4v^>4%isnsg~~>=)@;UAC>c)
zk<G@5E(QM_|LEv<i|S6&-2&5$q1)(!)bLBAf8;BziKL*OxV|z5$xzfqeNc<{)uEi=
z0ai);Amr`XP#QJp@YHGi*_Fl7GHwZ9G<?XtzOvt-C;{d=H(+jaew&o<yMBM)Eh%|&
zt(d%;?EF|Lq1ovnQ7DWqYdSZa)Jq!|j1hRelH|M^{to4*;GBE-@Y+d;(58GVBG+s8
za6QM?hWWLSJlQPP2<PcSnr^S1w8gnXUe>CQ{B)W8blM}+DKMpJPjYmssH5omnW4zS
z&eMrk=q5>IS_F%YhOfa$CzlXK`vic9u3%bvqE7h)fyg#7Kb0ZZ$Oa0rKLsB`Q6W^X
zbP4f~_BCigktUa~iEgN>#=)=EqyXPjCF4rBsrdqFXhBccWNX<qDIU*nq{(5}VQ+(X
zp!xBmvKq&`nx|b<W*hC3s2?=1w?<-9D{sAGu;84ceSEKlrYj|<=Va^4npay3p5|3X
z(MEOAs;D(6ZPW*ye4%|fD+gQFdi6?0d+S7=3KZ`XxFuXXCXDn;=_(kGzwY~5eSHlS
zP8&xcxxT~{83sNrl}GC8keASVXc<EYAz#ceYjO1mYjHGEB*zA!2Hm4RNAmvkzn{y%
zmBI;04;nLXo%^(j)qu1kdVSXC_ivdJpSU(HWbBXc&L_Z&dT>1V*3n2N?}s>ac6>s@
zLpVW#?&S0M4np;m$Y(iU$PB*U^2OIA59785r4bWHb%y8k-@Y7RfDB3o&nmgt*-*C4
z?3z)O8%0sttU>>?S1N9zD8c{muB<`)Z5YnS`r*I%YZE+3o60v740;#+-`rts<Bk<s
zKmBz$Pn?S8PY<Cy@xhk=O#V2MKep7z!}*ib3DMMgybZCZZ0sjS``AWPo)p#0#@?-w
zdjig*0rk;v?l|f1<Y`0srczX@3pJ#XPnr`|Ncm7>-E4;VDuei0Z2fpR8-L$4e=5Pg
zUjjuHP}ERID!{ST97ihM1^sy1QW(xR-O1;lhI6N!C|f6ra&n_EO|le6Pt3ue!p{(X
zamW5)ILgMKI!0kRo;LQRC4Ou-dr~gseEpe!8qObLL&kDEM!EUpMtdL1*2k9e8A4Gu
zLx@D-XZ<I`k+b%o6?JT@=Q(>{e~cCFh|eJ8gE8*oZQMQ4fb#r!b@skkemWV^$KT#J
zz#gRVNyRc<TG@@Hl*6e2%8hUEF)=9&fFDO6uV_!agL~E=TO#xB6L9E=$vA#I9OX`7
z8KCxNM_atxyA7cn*GKzAlbhZB@n>kguzVycHp4oHP`(u9dpy{aYBcEG9m;mQTl$DW
z+S~5w!R~6xGv>ador~`$!|~*Dq&z`S{Mg_HMuWvvPYR-9Ki=QdD;lK!cy-NEsaceZ
zXLJwWG_ctP69tW$<8PY1f4ckl5$*{Aj#TPDh>FcH7q>LRwuE#dAyk0NmrY#UqFrp1
zE<K|un=O<rMB<N+hqJM3XJhB$Np)WDPm!N!7Yqaez_|XY_P)w?(YDmiEoks}MZj|W
z?cJgY7os6g_zb3RB^`JUcWdTP*#tS-q7%nYhNB=4=-d&ycFv!)Cme~UyrS(PCzQW?
zLNt}&i2470IPk8Kbyk=m*bEKz@!AR^_}JLk+cct@jijipl-K5uhijJR(lgz~Gr(~-
z;ew0rI>(8rJ)-}x|I*fFR(d-S)|cv$FzS=lS>*w@U@~Ne)*m18*2tY`5JUy?{0XYq
zQB(_m+~i{u4NL$@+|oAq%-rg>ew#aVJlzGCC{NPs-&Y3?H}y<+aZ7h{^K^2HCbaie
z?x2oSK9oyu^obw4&2DL6;wP9nglJ8x1XoJy;^|NM21L89-%Q%uK(k!zy<Dh{Slma$
z0c!jq9h9A`?`#2O2N@xX)-TP)hqCES+4!abof-p<J{}G&caDKnwwk?xQezAkUz=^d
z$D-ZR?d>1icMhQJ=eT$tb#YAiXgK@cAleXH%C2!gPkY~?j&AEIpR}#MF)savINGPW
zxIGMVi?;Xe_=%_ljG-p@gNCj0FcU{=sJ*Z42G8Ts(05<L0vGqAxW)BTz3Uu9aFw#%
z=1T}csO_8X=w|2A%BCrm=I^!(0#5lBZ2fq3HbW?X$}!r>hl=*Gfq;R1>&eZw>mlCm
z<00Net3FvBW)_GZT|9^SxIr-kD5`M}-?SdS0WQ9q+<T@|RGLro7?*RO5Yxu4)l9%~
z0CnuPy(c<JbMZ=d+<DvO%V;;>b*&bsxgU#h*%|i<X5vZYb@m`VG|SVG7$G<N9=6dy
z2{6N**E>(&t@gm2k3}WGaRR0U?cU<zi^f9%(LT1LaF@%H_4b}X;dDn}(Z{^AY1J{)
z#W%pwHpa(y6i7h1;Vvk9nv3r(_nxtivz%Q#?Oe|N@$qmrc5aJpU~GI6g3vRx%%%BI
z>cU8>`4GxC&E40QpvOPy9feE2sV=@-ecbL-m@OzyQ9UpOz}YjvW#;<MUI8u%pRCTt
z-+d+o+NQ^ayTqbUc8+dC?L9}kd`VFgqTMlK!cZavpRSHdr6{F~XOPc1v<F!zpD%Ai
z3mo0nK^t9q#`(D2b@}r4$E&lm?@Uo{b0|04ZEl|6A%ty1zxB~n+J@c}{e9D2dgr<L
z7P|C{{zwt9^hS949F+I<Brrgw+EY+!bOmm%8{@kx2p|Rs|9Cjar48ljN%<rI8DKj|
z4?ZyZkOGhqXQFZ5_7mO##B88Gu=}{==8sSIRCF{t@uc|kH~}<@{$w~C$~~Hhq$jnF
zZ^Hk9>QR*MI{*4JvxyV(5r~mXxE{S84g)3(^9ttyT0{9aqWp(Z&4P#<LoR&+Vz?2j
zJ4ShaS?@gf9N$LS+PkHqyGHKE(C4RmN8r>Ku9sb7ADhO$c8<Q$Kouv#d>crOz3;jp
zH_DZo{STr-T%ZIO-}RfQW%i!7pdUrWY@=dadIP!%FDNRZ@h74J1B4eW!cflykUiC9
zCMW=8pj>)xwD*T?!b-PNUvB+)br_2bhd!YWxFy|jqSEsMhKYeP!k22b+s2+;{fL<F
zhjWBM#*E01r!8JS&2Y(|8WIGHNle~H#Dr{s^a=RuL=n48WX16q3~`LTjlUaiggfW?
zsV9c^_`&MHGW?KUJRxI1!k)vhaB4i|11Um-kA@?S|Er(Bi^_j>3u*MX;lS_z_SNn$
zF~a{amPzF&_kSJ%{m+LJkng2vW@AIXnZm}lQ6qe?v1wvMHS5^Z4PQ**xkSBd-O=+8
zmHC=azL$c0!>t=~Je@v#!|h)V$+uSEP><>l5tnRW@E8n|C7jqSMy|08h2<71lvNlN
zNE>GV|4shiq>Iz`%|%F5EM-zT!T<Vg;w4AP6%+$8MvwpEultbCQODlO6mFr0%6bHg
zIoqFOv4{2`DP3%b<n8z_3Ntnu9$5b#Ytor1@XdU|7~N0Kti|PSQwOWKjY!vsxsB}`
zi5bnz{rFN05&GJ^o4L7jOBJ$1<OpwFCkEtT3$eB5yZ3h-B=s<bL#=~<Xi09k)I>sa
zjclh|$Ceu7ns}30{dF?`>GQF3W05i~<Q&wDF<vj@m@=9r)hjoXA26&CwVaxqG9fp+
zOO#2UEMm=}7Aj!{{S1|dsMv@{YPEoykJOV+rq#t6Gg)(qg*t4MBFU*Hk8eCt+(em1
zN()KTn!XQIcv!39jn>r)HM_Rg^gbrjkAD+a(XG`qB%zx}grCDBUrzVNg<4zp-gGo9
z?R=aPC8!7RC|2MmJeBC2QPaDXLSw2fNKt5iJR~}IUCP<#g&#fIr4#4ge>NT%6P-(H
zbr)mN+B#>l?=pB}-DmTq`b@htW1B5^gldoV;YUxzw<+eErrZV&&=HC=D3M^4z1x<C
zAC+%1zJFWTPRg1Ck?_<DF^e*B0dqTgz%q#-qlWf)z={JjfB(2MvxBGtl}JL3B%l_T
z>}=_x)G-nA%;J>wNHLE<^}PWHCgD%#LtU$n>gBNoMqylORM$uT$#b-mr-2Hwp7GDv
zSL-4`$<*jk-fc^xd=|!F6|{s!a(u1uU`5md<dfHWmS<xQ(I^4k*(BNH8LOJ3l}PW+
zb62YZG$W#oc>a!*hZa}Oz=f}WzlNQaAduuUsxh+TJE_IRM&Z^be;&Oh=5MQ`znUJ3
z=bAOlRO~~fV7+tuTW2QvFv*6GTUXf}!E*pKYt^oac%YGsf^5E78Y#wCui%3C7!q^B
zQ`pojH!^G!A`me;RR$jA5uXDx)AJx!^vn7wnjtm}wu&m~l2Vx#A}D$-D#w=w;44v(
zfrJ*wlF-~j+39KXXg1hj@@A(D`?X$A{_?glsguHRN5jmr-l4V|saBYsyv$7$tW#Cj
zw-8lxC>wFiZWmFrZk_j4V{g^Fa|(2E@v+a_g4*YA`2}O&v`JH_s+em;&9{Z{DrfY?
z_R{2jF|L0ZN_j_sY-VAT=Zvvs%CTk`pvJaobHN>G`l?e-Dz1^cZ+k}H(k&3XhN-8M
zMrX3?DIKGCKf8Gu`ZgDE=fTg__!q#9-5l%j5%2_b9xXH&k-DuFa&#W!fchq83mRkZ
zM+Wj~qK@O~gL+E0V>Bjj<n+ZLwzR2P2VE}x-job-My&%!qGGtHnALtfQ}}}L4;&_s
zqtWQLAm6}b{UeplvC#2ap>|IJmyWg*Fm+XlR9|%48~EU*n2gP#wx;S5F8JOkh8=rZ
zZx{sL?s~qVGg88rbG_cE^nM*2*rm-w*?qP>EmYfz7Fq0=iNnyS?c6!P$TaqW*MFzq
z7259XpkRM_vFOUtVE`C`;W*i#YSm-E+Yk(DB@)vDyh?}pmSJUNBJc4^TL*y<K)B0L
z@|R&oIo5IH`IEDm%uSrh00Hc7V$4muDIjr&u=_MxC*(r45H!>!GkWI#pm(m>+z#Qt
z-dItdJY&t<DK@)Big$86$7(_?Xf>Oxcz3$Z(BRX!==H^OAX+n9=ki~}WJ2{~w>?*K
zroUVh_<J>*)ZX_+c|vJyCe|})u`G?;dw19^)r>XeIRQoPOr}ri$&)S1j|JU{uX^9_
z^#z`*NHIW`K63{iZQ19vJA{;iln{!s!61O(E*GWv2&=N5qdpeu2nE7(esz#4DYiv&
zsuU~(6wtS5P|k$xB{&o&U_uG6Q8_}IIf%dpjE8Td;Z?CjgD-TU@rcDg3TAZG7I{{I
ziH@t_M8E0r@sAc=@d!tQm^Qt;c(#e&x=_g31m&eQlZfuEvYkL{U;QkeuaU?y>KKpb
z?9@dEARDRF`BycH-YX*-)iLhixZ0k;uX2Llm<g`p_r3E2`CMI0sx^`_^6v2+ps(>6
zHKG%SB+p0vPj^Gn-OyR>RHL;RtwzzOsGTj%Honyab4$2aCc*g`8_>Uo(~=Utov*#D
z^KMweIlHt#;1W-YzZf&w1ytL-C#78pwC*9pwlq0Fhikov@Z}}vFuk!cq2U!Uw2Ed`
z4^b+>7b@e~V90L(<mJ<zw{;r@T6g~)Kk>~UcNjw|ry0GFW~1Urr_b;J)HtIzr}K9O
z)T$?_?Mk@ivyc1?dwc4JjI&Zn{44jp(n1MSegwm3dO~Yxe>5K7<c3iDQ@^3BIIkuG
zJi`*!&x2b!u34XsQoVbT%y4_dDY{5#e+!BebgKnmHy9FyO(ylT2f=Lq_~PT9P$Xg$
zwrynC#8vp|v8&lfdRXr_aP)AD7g~V|1al7CUsSGuBsGdE#0r9ez-sR@nsYL*(voir
zORfj0yyWK$sS%CFQ(NeDND3(<2L3Z8M5NQXD1v?D-yALqyn1h<w2JpYu%FsubGC-=
z`+!viXeEMs|BH{WR}G`pPF<>u!k<{119<#dge`c*OwjA%Zq*#kSZAL@O!`mfrcnhZ
zVO;8|JzU3ygT-?+6Z_cq<-+ega;}K#I%DFrKjpE-(Q~~OWO2V6@C{$R?YohI^Lqe+
z)>){QKTlaK&n!4Hv2G7}HZ!x&=+da~y_|cHt|s*Ia?ys4Y)hulsMW0dcPpYq^erZ}
zYWH-}dHor$br9;T29dcxVYhQ>)OSNS`F$5XSg79Aaqf%KsPLl(jBgYk{7)ah8{{vg
z>utxciUyYK*k`ORJ|x7$v;0>xaSl}wN)LP+1>w(aa|@1v9S11=!VptC?wre2;Jn`{
z{wHbRC>6V7MLwQ@n_@;e@c2U8Mp_HiRk^Pf&9<DhZh{bx?T>>1H<}zU`U`JPNTH1S
zk99?*VERP+h~~Q(J^ERyM?Kv&L~gc0lhU~f)9!p>Q_l6+2}#&`>eMc`Bi`?s-J;Z0
z@ou9Hssk3>b`2)o+8%YC1Gn^W!cufR9u~*yeS~mOk5ZA8Bj!;ZT1s<*SB&050npS;
zIQ))UA>@kGUJNz~3`W$ic$1Sdj(&y~#BC|~yDpwe6<mgcaB@akueOXK6i~|;CQb9N
zW0XG4ui$867gns}&(>et7En!UzB4OiYL-0v_GUtHYWYUg+lje-_&d4DpVg?UiL?{_
z>;1Xuy|tNan3Eg}?<O1RdKE3boaAEdAq9IHt2@~WI5;uMXgyM}r&8RZfrk#ELSgdk
zo)vH2_hTJFOulGjdql@N6}K;P&aZUQHJAPhb9c{7m#<&|$ryM`+O@Skqv8<p*3WBF
zza5{)zhB3!KfcdGp&iu))coVvsx9d8gUbsu!nRXDWffYTP#aUsLBb0lGg6!vo@0z{
zq97M}AJn?jX+Q`BsUS)jf#QSd3IXhs7Ix^2>NSFjqhQ!au4Cu^7gnE{9>HX(@Lp6&
zYPA;hAJF3n!wzE9>w~-Oxwos3`=O|ZpY@k!7wgZx|GjpwTX3I4ukgfz6?r%2SI*Lm
zZ#$8%8onO8K<M%1x1R;~o&D8O&8+(00OfPctf;>D_uskyM!(+lJC8WQ))rx6uJxd&
z%a&uVMHsKUZRE5;uOlEVBbUw9e)#>!3Gr+(UCg~1>v%+^_qm<UnD&o>KW&o10epE9
zIHm4~_WQjvEL2s<MTtzm&Ng_}d*Y9Zix+cWB}`i_vhu2)W#ny_ZCGo@w9nov5|)mh
z!9I@R)(8oYPS13;V@zP+HO92p`SIOwG&U(C?vn8)#RihY56mY$8D&zVoyQl|2!}0{
zPK39eHlt@i#3;Mz%x$XQGKl%|b$H0QdCxcK_$7<<$d~<AFoww*{|M6j&xz&01N-Q)
zOB%+tj=rz)ukA58c*(4gJHdIg<~@LIl*kjqP0Ba_2(5QG3|p)K6dGEgAkT)}A;6>L
z{W|!yvSTP^Sxuap`Kb0W5YGcK0q?wM>b-v_@<m9kuckh$=Dz%wjVDHJ{bRW!V1c^4
z88rT_SFgSPaJ*apG(KKoa2!8Urq2o>j}Ut_cszaGf0@Jeis5fPaNx(U#?b7)wvmUV
z*B#h<tw{jaYgdhTvi@#Bn1T&z`*?1OO0zGh4Am`XPA0S8P3iy@;4x%xm!@=K*P<}G
zq?#_iZAk78zP6&=gc{|(YOZ%jfdi~rf+yCZ?$z_DTPE+6X0n1jl$X<`3T+K4{GebY
zu2B=$I)iuNi9r-r0ly&rHNBG`(ya7BxHqT}gWLi1jHggm0sqA+#2U*|#7}9qy1t_Z
z62%Nd_|ha`?&1|p+LQ4fLjBv>LzuK_cLsH_P|g7h`^xl3dlxMUOV$Tv9?s!L?Gmf6
zumC3KiJ8B>!i`LtSRK!zZIih_?}T%S-Zw>#j1Sj;{bs;$oSkxV1YOko^L~o_AClr^
zoXde>*47{HTizLW$919<`97cTg?;sAT^{_|&?#xQ_p9NtK6xo-mHJ}8b9e%%MT|&4
znrn_nPmX;S!jo5eVHuP5pzs%!df%qNcJ(VQc4+LP$mg48J)eh?tD^2?rNKc6tg}8M
z0&!mm%##EKy$UjB4GZ51OBGyTR5`#}oCMekC8jzLh7a(qmz@g}NDU@POQ)J(h#tOw
zt|2X;rw@>JvFv{emFkD;i?7OPbjDez>Ady)_X@S$4j&=`lo!54B6XnJel4WpM7Q$_
z^|6FFncCy$lUQ=r1yoW~P9qn4bGA!1f{Aaun+&DdnH1ITUyeZa86LpVXv#Ix&QFmy
zf@%zBBDLL;K-PZL9`V-et>&{1tNzGh!-rzEHz~?_y;{XjJ$=qbp}JXm(^L7InAFMr
zLooeyX=AEc8@DbwtaTxO-t7s!xwCUVs9_aN5phywoU1sN$ea*Ql)?WLvcyK_Asb{K
zKFTBlYioqsnzsDIJtS=JjxX>l$E&Xn3f+m1#tcNG@oK<>%5`6f&C{6Lc|a7M55`mG
zetA&DolzJsAf*D0KE3H?BGv^6Dd=}hY!*BGf#)H&i8mBQK4)qp3_j@j^MoZwlJyyT
zih5b7h|##ESDH6BCT-5?zxBf($nSY`%ic}Em{X^IPhcU?o!K@uY#Eg90h2{kiY;{a
zCf{P*JQ&kRZd0@i3#?xE+51W6XT`I3KMx`EH7zVQU)D5y-|_V##L6p<;meQBg*S)C
zBay6t*Kxm((u=HuA3bIEUhwFBe9CRYs@NcN4DhhboQbuY6>!#}^qIzF;sE0lF{NsH
z$j81f+7f#2GG&<Okk3M`0lv8yRQ`|`m>J>Km-C{04<T3R2LHaZsPir6*6fx-joVdQ
zCU3O6{{s-u)~iL|-T025U2Z7S);IhA+WYRmsIIQxp*eua&=d<Y2ng7MiV;+F08xqs
zL<m^Nf*O;UQKPXWaHxW!f`X!=AeKZ0#g@m;0TDseV2w#M8LC|wnw){d$^Gs#!vIg-
z_x=I*bAMnsWtX-0UVHVmkLg=eJFK+0F(h76Z!oYgzJ$}qh`Qyz`e#l*C{t49uG_<v
zqh#zxLj>1OwSS$y9oozhTU)=0T9~LR<5~03+pUzoa=<TUdxlBTiort>lW6B4aZP+o
zRgJ$MN0oz`>^^q>3No~`{ib4z{z0+H#IRF8%Ql%wn*X^li%BW0NtLrxmrr-8<LnAX
zqnq7qU&xILTf075fLpD1Aq5WgA0VJOfo+##i{i9K$^$S>qR3fVB7pIht}db?SXiA5
zFu$TFRJ$-I87M{6Tz+Gt49q?ZKg(FxhK{XGj3QaL@OK_`mHJhTHBDFThT-i#na)?3
z(=mNlIG<Er**$XVhZafiANE6mf@<<nQWdw-FKnRnyL#{XtGu3OV>ljn5#_}w7Mpg6
zVhoIRWmlNo`E+l=>Kp*DZOp1>$>)Ls;P$~`t}j^W-!AD-So!S6h=BVRruV>dXT3ex
z55ba#70*9LqVN1yRc%?@40@Kf=KuWimXnw>G?~8+52L^A_u}r`$dwm`hy?`A(jzg(
zN&6kvpz}Y&%fT#_hIep1!u0Mqw4Ho-O*v~FbQz7bW@WsM{IXKt^*~?$>1aIZ=34A-
zn4x@j*^I;cy@@qD(7{K)yW;@b!>Kkvtdf0eF)p9c^MR(_C`IEZY$+^r7A9H29mRAj
zf%ybOC((g;L<NA3v|xGkc;hiZq_Jwy4teHZ&|bo}#~?tp`~{^HAHgvQJZDoxULa-}
zMI<3YO;Tkm>I>gt{h<tZk`%RwQP5m~G^Dcrr8X4v)v2xelLHvJ&L2=qs?u6OGiK_5
z#5vV3ZbV*3&~<UcJ~v}2W2l2QE(!_mb+bQJ^4jDvwpPfzp|i}9eG@kkjg2fXzy@|2
zp!JMX*}Qe_1K1;voP6?>eord=O4wshAY8OLvtJIfq8nmOCjuYDYD=@VM<_SDD6R&G
zZ0w%J8J;p;XOjt<?lC@9#H4$;qxmpTs&cROu>RjsW!`a)Y|j`gs%-mW)W`yomoym2
z*%QKrB4&NrKB{b-05jO#_y8njh~E3^88jVZVvf%wVg#eI4o3pb3gwU#x|>+ZG7xQA
zO;M7d{fPetd)tqtpMtqw&C)EbGBhqfky1gk%P$|?`m2;X7dS)jww!aJnEO~*SKuvf
zyx$7eQ?13w&#^?7APRZ=xpGcX;#HkbHmzo<n<Miv&)a{q*WV;z2NGJ>KT!)B%DHZJ
z{iB<x$|0U2w4FO^pEfpd`CngzGLV{Ye>+8i`m-F5qw6tIcFvN<z9P6wnv*^gE)+9y
zdI)7*o^@ivYPx;NbpaAy-}mWRT!R}@<R8IrL#uk0&LCA|48v+WL7P8QuUZBf5|u$f
zNI&H6=!&h++SP@Ehi{@7Tbt3fXsax`Bv~0{S%snpJWNNdRk!VoQ>czsDkU8T@WMrx
zV~dtuOT9@BWf*Rewd5{2km0SQS~g<uT?!oHY?yiu21Le?ci;R++BQPg9vCKh-$&H?
zHl1q>wjA}xH}of*`*1g)`t1D63KY5&w2_?I5L37a*pd5N`+ZDk(TH@8iz!@NeG~2c
z)-KZEM+cE)=ru%y_JK0TVQbfHcuTVsgq<0o5`3$G%7gtc65EzYm%qvCBpfv6SH#j+
zNNj#;Tst>Lf3=)T-L#pk#{-J#A8#b(QoSFVVzbvm2UYzFY)}oP6XR3Z*oR}1nhqSW
zAjTVMePw>_1iIbfmlaZ`eKKdK^IlU+Rd=mcYAk0sdMV&)9`YA5{JdXoz_mZp-^@H>
zw~(f$>^}rOydig7e+6}GnH{9?rURFB9s?a}gAND$4BAn!@j6g&j|VxNNZU&Y9DVb-
zfqxsMrnu_esCr`HuNV5ERXhiN9m9xl<LeG-tPa>n3TV62Y+(4CJJ9o|i()w~K^LXW
zGaTx67xQ~XyO4-Q<op>4dx;k>yR#l<STmMm5c#ZyLS+;p(sOJF&Ep(3q?!=SdSaT#
zj;v~50iRC`ydF@d!Coo!a&qY*xJq=rC*ScpguUEv|BkDskYwZRqp<Vqq2=_ptR*zk
z)O=Qp9nF*-6H*PmGliUK529=2aPF=x@^W@eg`1?>_v8es=1xKhyXu1eRstB>sjvtK
zy~(}1?{FULV0J<Wa+UX~Y2~b43LHl5>70YfPBG(|I54Y|u=NQ7Ro60g8yPENHal&C
zGcBeF&DBR<07|x%LoKu;=jF%w>t0D_K5qS9!W!MHO?w0UQ)zCqiH)9cU_$I=ghkk1
zBWObBa}Wzqjt8uo(6d;2F}cOmjDzGPQhSNOJcuw~tLab(djipk6p>iWR29n(fVwC=
zR{^KmReO7E1KsR}d$RxfByr`vWAohB7{fQlJM1j<Hjjdv?cue=6>D3G0}j5e9FvP9
z*TaSO?MXA)GI+|qNvKK;-@4&`M}55dBpLH{v?DjecLF?rL4&k6LJ%?z`6T)LhHolW
zVN&U}Yc27^c?h+{_0@i#;c}_2h=~`Z7B@*OY$iYE0)Ahj@uH;#4k}mk?)oXeRCe!2
z1+uEkyY}I7CfjjJf30LdIxKTpIwvf<^89_|Hw$7`aJ|oN!Kl`Rhi}Id8Sh@qw6BoE
zY_D=zX8lent;DWP1UGq#5E}W8vr2E4m%x>48~TnufFoyn^Ol<Sp3NC#F=OF)6N4RL
z?<;BEgC11W;Ej2XIkoG`_de{&wR3u5@r_)(RulO;Bq*C`E)^C!AHaG&i-4LHr+&&^
zC*)?goD(Z*=rbJDHQj_PrPqw%o=_Qrq%!906koa0OP3^v=y@1}6E%DV17UQZIp|+6
z#3CqH#s)9?kt!QF8b(3hjHErtf+|a}DZ_r*CQ2Gbb3;#^3Y_`X`MfOXF(U&&15QO<
zQ~1zIj~Q1`rZ^;IM%Z2rjfSL1h@41y7i^{JK=p3!8srH@WHx{e0Ne^<yMO)pe2UZQ
z5KI`!29B^sDTrJnPv^3LaB#nA<e;HbsuI;Y>9s3%-15xHuI%7d-VTyizGWRst@{0Y
za!x5@btGAlR+DepN8J9{^qdPm!1y{2^vPh}Y4ef}bGZ=yc3n=+B#v~zjyn4MxOl9%
z{ox%i2dc&=)nXP-JsgZ0YK8mIOalzRx6{FX(6^hK-b>ofmP6a#GdVlc$wRDI7qs&m
zNn1|Q6>ETVb<=Q6!3O)^sTS|;m<?-Z`}=4q^gC><3bTX*CnnzLs;BRHshAEN_(M{i
zaYUxp&`+O;1Bvj_7zgE`a76xSkVA1aBowl+Est3%eMW)DJVCm1hqhDLX?i|I9maHX
zFgw#gwNKC{v344Do1sBhh%L9iyEiM$B6lWL;AViOQW<QSvok!Wx^P-+v?QFvte&7D
zf<DH7<d>x_aTzCPv>y)@!NtIQ7q&$Jv9Z7;i`k!|%rBerAE27f&;S`zVc}cO==X9z
zxa`yHI1l)|h;6*S(S8)($7egT?K+%+f9@+*FDgK$G_wdRXAHKwtw?%*#A*CF&_Y4~
zve5x-|Fa?aWwjfTT1lYON-?Ds_rgF3we{@G>BfWOKLrIxWhsdU#|<`cf+jD7p&A1;
zE{4TVAc!u#RmawDlr9?)rVs4sD&GQuRi3sk?$_G!M2{?#Uj;tCk<7v@s<v-AEZ!5y
z1{cDCp4?sSvDCQngmscvb(YQ07hshU6G!`J;&hF=^Vtp)OZN1LLwBgM;2?9UM(Dn8
zdv52~RN0FXlAmrqE_|m<Ik)$%*q3kw<P5e42peEq8@1Hil|8O|o@Nb39D`*0_@Piz
zZKo|$P8&N08(w9%G#Q;kl|8zxEmpP)*=07P!MabL4M1i$XRdF_9#1=4@B>}@W$w`t
zy22fxBOaHrVH9gwSOD(<-*yYp!?U-8&>9KWxEGn5&YqrV^BwLnpA~_l56Bz-)4wY9
z16}K^oW=!hd-O_!rPM>})oYNnJQT*P)zV<K+v+fBDmQ^6G7Xgmod_CIf(E;K&SCrJ
z4cM<bY(hMeLLgvk8HmwnQl;<jwF*t(2Cd0H_l&*SM_Uiz+YZ_rT+I#o7C1;W=-~|$
zvpCTolFB5zpMm=>=4bmum83P^*s7?7;zmaccbe;=uN;C2PxSVl`d)tw=)m$!dHrD9
zgk)w_Pl#QG`*+lu<+0wT1#s6g%VR{$7Hc1xxp4&k6_=5oD?x&@Fl$@fPZr`-D0_6j
z-}!2gyO^{0_THh@3tv3(HZ5nT=<ZvDtl#>A4XTOJSS>od>YnYX#506Vp4<Xxv^5%?
zgzQ;M_%?|^Sp}Ja0xc^fyhD-zX-p#RYVw3#sivUUIJm6NTJ_ukw-`KY?<0rWpMY;*
z61m&9H5psOKAwU9>#WuoEL_3iJ$fBZyg-{;@Nh)9z;_mrT<pCVW4jCzd^jH*e15bW
ziWjyeuzHL5mhqjQNHo*Y3ofr~A^rgy!&9j73(nM%4&#gcBRIdBM$;itCY69M(SAiW
zY0a4(68yEsyLLGGE{&AU1J1>tB)v`J61W~|k?;d{>1D{ExL@wRj{MrAr!^mH(Z#!A
zxLUsS$0AGq{7bDYDUNB#C@{v+nGjWUgZI345`x(8+Qc8dIpE%Aa!@}`$)x^lFX3z&
zq@EsbKq4Zuef?me!&(GjRXJ*o-E`8Ys+$6s2UC6+fGp>eo~XQ0tB3@9n7Bzj=k~oV
zUpjYx1E~*LN>6n}rwpV^QEwODWH}Nj61DDo&EcCVEBEE!zEk&K>&B9p00mE|Rc;>%
zX1@Y6UM~UWD2TpUBE@1T#g{2&7$4&E3JS+IwN15aL;>+05!|R~2hq9PY2)V!R9EdI
z=Z^Dv5V+h~z8kU9=zr(azHm7_{nPGfF{E!P2m1OdFwb(^+R+Q^{t;|Gb^IyL;y50G
zW)B6#2=kO9gRGtLeCq}93Ib-_=&7{w$#ma|Qmz@Yzv?$6I5&^QEE|p7==zV~#yShT
zEFz(b6(M#8=*;Hq&+8J1);fMyqUE)J1b@@*@9=wK8#&_7)d1<I@9KK*!^Hl5;=rh)
zOX!+XKEtn%+m=D7Ui1FYlVr;z*alD7>&3_bO}2d=YaF<lT*Cc=hBse)G02=$46hkk
zfw|NY`eJAq^%ED$@L`t=+z`D;XLO&eE0Q(@9&^{g?sjkH1WNftkVs|+thZhzVF$D#
zc!noP#PWu-<`E#oCTP6dJej7jQhd+{SuGV&CdTkWBAmy5k}Asy-2-1H;ub7iY%=eK
zX2rSaiMZ}y=;cGZ7&jb0g{dT(klG8IYiI1xdhyxr_zBMR@<+WB?nKNr5&&-%-|hK+
z_NAvsuInVQW>Ih3TIBidTY{_tAWI%~J<@tHF&fm&#IQ2|^x|p{tpb`eC^OEZmlax5
zH5YfkbfzQP>K=U-IkolD)BMv8F^0i?54F6$9XRY`>&3z-H1Akfh!XP1qMpVO%g;AR
zBhf?N!Q_t6KMSpSZ{=UK{$g!Y#Ob7sbA6h9(P`<my1#ArC){tA<wVTY*Iol8nL}TP
zr1KkfTOsU;*4moJ#_Kjm%ry>rPu<Vl<V?@Xo}bkjjdhBIm|+Sy=o{_YnIr}Zr|1x=
zs8AJ^*evCu<;0x;8tGhw2rORHBj%Q>_T?N9NHPb1OG$ai841vBIxV{4TTRAGWt9<h
z{3p(GQLO#4Chm)Vmv`V38MpB`^oWHU#9zH!&zsFC;!}DY6FJ*?!((O8qgzmH8L`ih
z0jj#snb6$bR$w7MHkY0=s5{GYm_)HGeA=bQN@+lCgk(>Ko1Yv<-2Cd8@qu-|q{x@w
zjPkM=_p<yPdGC>+CFVZsvA^%_R%+-3Q5wyC=4j#iSQ#>9#DJ<(vB{x)o9L>Vk@GTj
za!UO0^ykcMEo&1T-ix&F7ahas?P<C2R9POSNHQ^5_KCCj(%F3!x_>lEhgRJ3NkgRt
zF-E)6A1lK?zW$t0m7^K~8?2*)1VQY?;q5dea3%}9C=L9(b)X}!z9<(+K`|KG6SGRf
zM|khrK^j;MuQQmoQj(1~L`?V>%SOm+eY0Obib+d+d>4e^<dSa&LHNwdgBf<=V}!*1
z-p?!G9Jc#3*g9y|-}Bf6ECp-*{2|4p!^T1}d(H=YiJ6MQvn^43>h$fL49A9n%uZ>x
zg-9G}F#-2MTZh-{Vf7cZzXvM~(O!#fJFMyCV*(KghJj5#?Z9-9tGA(f_C6pt=t|Wg
z`TeO+FnpyV+oA7Ctb_X18H1>Wq~)^SSA;oudb+=XoM}CL5EX}p-kMa0SGyg0j|0d?
zUk0#dq7i|cM1T+g6R_u)>tC01$G2A{mJ_M*-zFduxT~>=vwVD}@P4}vq|%aq1mg1`
zA1eB<vNezpnn0|a<0nF3h{;!|UYJ(Le*_t7%1?NQ#2f9BwlPBzupnNT;}82NR_kV6
zmw&k78q)}IaofuMrV;f0ib<2N%YEuYUhWrO`@ns<`-{PETO~=aIp?Fyjis(he_LBI
z`SO=-R0dv}*n-R+B*BraVDl(c{@!XA&RQDfl!5X?txweBQE;vD!ldbd7=|0Bc$^C!
zJw<?;>16Z2nyB6leeqEN_3CIbX8X|5up_Hc{I+f(uAN$c3ZH`Ex;|&+vPQ2S0aU{b
z2Xi!`d8WD^v&D9{xWQ}l9Dvp^DRBeo|MB$FB*57B)gR3m-Bhg?me1LTRfDiui?tLF
zW5PCRhOy5F)FJADpXEb#q0hq}YtW|o)%W>%N16QODh%Fg`p?N&4^^;rhY3*SC&3z!
zfJmSQ1Nrn*05Q3=NV)OtlR>Qf)dRD9Er!|#jKGqB5Y}>?q>6;8-RJG>yb`Ro4_9p%
z;L7eu4l~Va)SA9i-Y|Ax<+qqYGw;(rj`xAan@nZ_sh3tJ?x1T{{xr!HH%IEHL0i~$
z&e?DQ)*ZUb(H>@-@OEXz7Uvwr^@)?Hs*Q`5%9I0c=ab^ztWVNgc6h9BtfXSdqcO?J
zH(O~#T{l0V*Tiimq@^=gtae>`$^=bke+CIHnN@`XKlv`o;HJi%v3@&QthjPB06+p7
zs#M-Uz@0NYt6|vs3smWg64QP#EYW!lpYPOi2jIRH<^=4yAj-E#Zw1ibeDThX^T|E=
zoSRyqZ<<+`qsY%A5KMowR5R1R5j#M`p~zUWe>GulwyAB1b#B9qCxZd3@2Wu5!P-vo
zG@logi-`@QE1plNJetf54DUtN#P1X<^mc<!N4@9)@I179c7-Ee``kc^5tx5#OtHON
zMeNjbcVOJQ_#ojf6w$$Y#vZVMJ??`Q76`8HhRkth95l%{_dH+<jrsN~7f@3o$yn%X
zi4{|>f`2Fm8Z4j+O+McP9<Ebb<LcM)k6=uYkktwb6|qG$)8njAqS*DY1k6LAd)iA!
z&Km4~G+D&Pz7;?;Z}6{#((nX*R|y2?=VQEGrMK=#$otcs3G(UjvsL$$Y8_3a#?nrt
zt4914nGhoCW$|>J%oNX(TZNqx2BCe#eb+xGJ01tChP>1%XAPI1qFDfPPxX5uB@Tld
zV80o+8Fy~g!k`oEztm|JiqA40^b8F=IGSi8+yHH8dZKCAzf&U4-wtrZpm$l49;Dnd
zQZX(19jAR(brZOyLBw3PUGIu$&lQQ80Zgv;kI2usU-5D*Bvxm9(Hj4TF8VB#balof
zU&@VoIj9(5d~rg5r3!pKvz~aMZd}!BsIFjr$A}BeY+3=m;o|WaiQh>=viDQUbj@$5
zy>U0Wt7BGbNgPRGIcNdwcFZu|GR+4%v?ZgiIxe)T4l-eHLXr>DFzxZ-Sj2gck8u4D
z7MI{rUaj&B6PILmh|!W(`2jm^2`8D?7$LjAR(sEfNatUe{AX2hUOue1c6CHvvF;Ik
z|4Ll1^UJ%)uWwmh!@D*)j#7%{OF0t0?jGlSTA6?8CE9X9Tk1H_cagHw%G^OyZ{!s_
z`SmBj6WvlUdu*FD$vUSVC#=*Jf1<@r5k1<Z+qP~MH!PZL`5P<e0coaPyIub6Uiq>7
zU3$)k$j_DoqmmabAYe$70BxDiI6&T|f~{9c=%4Ju$jWCZs4|rXE}ky^p)q9B3eIqK
zaSQ!<TnOGJrB*eR>}``DxaIlvZ>)4)eT1x7P?5|e?3SHo4zCun?XjHE%80uAi*jzZ
z%D3IFWN2*PZR>{p#s(U}4!(G&?1nPG<t18(2_>6Fy%^)(mEln_X*AFXd$drB_$QM!
z`InHRT?=$rboTc(Rju-*wmK$Za?-oVAwBvfGeScEh`w5A*S}eLx67r?HP<&*<X@`h
zPM%Of*-TsBT!U4}sEgEq;E|T`F47qSJtkJ2Pys24xCV<mQzsfoJS>#E4(7f~`kYWY
zpHAQ<K-ik~Br><E!&+JnOFgEG7rR_YG7u;=Q9qJeC^k#w)|kH~zfQGB@SJQF^6!e#
zk6HC5afn!)MiPeF{iED7zuv#gI1lflY@*=JdGFqfq~37H@jK^1%y)ogr|m`=-Y4vG
zM<NRSb~_z|)|>=35>`*&9jyj69s_m186_G9;&5)uB=EAr$wt^lC#lvwU#3}-j?d_N
z3{*9W-zBqt0QEUS6`aWH$%8?F$!(u*M=AIAlHjvrdwrf_+Xi}SJ!9AQSCY<Sb<@R)
z2R2r6o=&K?GJY2sSV~pNPy&6+4SbKJ-(}lyW_M%~2zHE4hK6Eg$2hLhI7#NU^2a!4
zv<gA-d3$ri1aOk=TS;<9!vs}KB@9zUeNJH8&9P=M<?+VF$VLp1g<X#*fsnXV{*ZB>
zVrA!+jx@qzIvUJRe{ke5=;y&zy3xx%4NV@3G4L)z`+|A+#!Fl8@lU~3%%bn`XQH+B
zp0>voZC<QuwS15t|0gDW;6Bq!zn@;1EKvwH2BoaVH()$cCR+Y}JbZ2byw|}?O<Xwt
zDAxyvo&QRDA3Du+q%uDqAC>X#J5y4<WqkiJ(=tdW<j6MNs}s<ey|Hul<S$oizA<xZ
z2^Oqyi+?}u{NgD@$u=sxBFkQUT~b{goAf$#4#GvYt{clYuwMUX&6ZW0_J|FF4?%z&
z1Af`GIqwA&2URpY!PSa7#60(gUFQ`sUGO>rZ~TPT3lR?Pj-Ap3=)j$7802@LZuTxC
zd4nYz57)`!$Ud)F*D7DS29gRl707+~G7j2vgn%h>H^PSHg^jTrYW!%$zDh5=SC_2s
zVS}4xLNMkOy2e@Alf%oD6WLJ*p<|01qt^qu?WXoJ!jNZ=7~V*g{{HBIfwR&ue+bU*
z8m7|9t!`M$ls%K<jaBhWVC$hZ-$JxmbL&?S%wDOLmR-f2PQgcgH73uJUG5hBr~?A)
zvyPIN)msm$Ax!w)rM&BeveIinbz&slNgQ&?8n<TSyBxKZK=Pmi65<hUw}dfS)afiL
zKI1U&P2|eMF)k1nMyJH@M(T%_|4$OOr}k5mrJ9tCF@6Cjk^EtH$DCF?u)g~yQg&RK
zpZ@ZY{*)}nXrz9|dv3_MbD1zPoS>K#ZSPe`;;^-8!MV#p%SQ1z9j=d9O?pRPh}qmw
zd&eW_%WI|kfH#p}g?qxBg>#%YnFcTKOdW|bigDzo%<KUORsFR}Gu}jgGq(zOv(wXg
zTDQktO5pOwOj{9i%6>aiNa!#-iZwEI+6RCydpZHFIzwYfnhc0)r}WJRAeWMUB58L3
znzCe~+`A~JG_)zYo56JWw-BFM(4hj&ueq(62)omF&x%2rXj~iUPX3jfe?<P8Nh$me
zP1bi&;pj967YrI0pAli+HK%oYkQO}iA|_?zQdwvHIHP^#jA!-_<Pq&<kO&O1o<6(t
zy}BFx8yc4~J*&Rz&cP0;$xzp2>^$^%LOVf<h}MS51qV{a=dt$y7cjiZ$s`J=t?xH6
zK|0*nl$lgfq3#EO74{dAzz4=n%tJ5dbquOG-Rze;MUFXLM~mwVmuEqW<;4<PU+)>H
z8M@7XqlRy-Q^g%1dV!Dkf5+pfIf?aGC1hapHi@tZ5=hdWc_qAIK%velscL`4s#5W)
z_@HH2fK*wqA@^olvc>pD0(DQ?_`hFyDhwoS0598$GIcu2TkJYEQ}VhnE>p&2lwMBY
zrarlTiJcMftKR_X-kcAc{4kJ!CqxS4-A$t(l%RuZBRnRVI@_zFcc32U0MdNpt!W0T
z_@}3GYut|0FNf_a-At02ZYrTYRmlcG6S}`*l$*4%PspOll7`SxyQq5$LPw2<EVl`*
z5hqE>HH$UWy`bymB%jCbK&i8GTdSWLsC>nzFQknY`Ncas>~}%Dd$gRec%CQJ&bmV_
zU@1h?NxG*NX^1ItlGKn$qzd8#q4(oLGoL&FlaWpWir$J<_;>tPuOKV=_idN{%Ky+J
zyB6{Mqx}2g^Cz}X#0lRNvrjPCG-7i~X_~wA`*L|IaKTVGqUB4YvQ$q&WM1C%>&YT^
z2|V?itttiuxc)_-a=iBLL7k4%#{3U+7aanss+?j)w(+uSGy3*G%%q4}9wcAb+Vbqk
z-X4qUBmOM5se`&_tK0J(o;I6R$bc=*DIc2a@**~idiRAOKtijl+@~yQyLc=b^F?e&
zI>8C9f8~WcP<O=d^KTd%jzqCe^Q$T`6)vqVQq-ZyLNY5zI+fMYSp+k)aZjAh4Q4b6
z<p%>Nygmu<9tEgGrX(f{&KFGCf5+sg4Kg>64aUNE>@jEcR?9X25o%vRQk#f(&PMMU
zo|y-IA045D?eVh>fb}eO%fnyJO&MS_6YspvQOo>|hC!yO35tT&6hAd(wUnE(Wcl@R
zpc)n5jk*M;HMg%jU2(qfBiYT-p?5hqpF6&Emo2XDeyFXz6?OA?0=94|%_iaZ)V5J~
z)1f2HnA=KK4DAInp?(3^Kw?oj+sA(*_DfA@rH>riNpVA%Tnk3)*C1t@-#JU{-2nA@
zN<HFXH4Y=6qn#snbP&M|JnfdA^6xB1lA5P-Qouo3*@?}f7n1fKYBJ1F5We86f|Ie`
zvNLp<82>t*daw|7xWg>FCYh38$I8ij@exsXusOrRn*1%h(OmLh(#d{CGB`o>*QWm-
zl(R7XfqYhpEB`&xMQP`ysw8+EWr>UGEXGkMqE}JLFMMwMD>2meR+7I?lHe1Bdz!|0
z=1Z!rL_J@B<!Dba_#)$$Y@i&firUqg^xhCE34$(00z$ur{F_I{RR5O?2O-@gP}BJd
z)dfKsVewJrrI#*)gT%~Cmd#|J7nQUlrThU)F;l(vZ9wpe&e#T|n>m;Myq!;4&v+mC
z_2^@69YHjfTd%BNC{^B?KOPY`Q7)Wy>3EX!zLNV1N#)D6EBQk<6?$Vj>LNHLZ{2Xd
zu*nK`ffLg8!UX-RvB7=K)}!JCLAkimn?Y#J#?*wIwsQ7syM)k=Bzt=;vM>24-NNAQ
z=peLMLUr2Y|Mv`OQ5K(+*vPX8ZB0s;u#Vn_ZBhOduJccrcs-)b0OQznm>i^MK}IMr
zQi$0r*iM^$1<xq>?3Us#@p7HUQa(0w+$7Z*0312_nb*fpTE!K0Fk_3<wqe0PiaGsi
z=?IX6P{SpQz*1Y6$U3=h+>#wohYYuqpf3|&^@|||+R+HqG4lu1cd|u#*eM~)OtJhb
zxGR?JdV@T43j;Lc*$^fYVjmWqysQUFp@UjDtqcBz5CV+NFMcyR$VuH*g&U_jEn958
zz`mCE1Lv5L2Z91g^)xNKYZ@Nx;70)167MWlqY=|JOrc{gx+dOFqL_9|gu0$l=YI~8
z{0v6WmH_#RhAdvwG1J0s0dXV>mEXqafX*Ep43=TFYgq(lt}zA$fnD)<Z(y_~i*tCs
z<YES_djCD`xBYr%aXntVv~tV?`Cg^HDs7f0q}7U57e)i1f~?25o;sSY7WaaKU6DI)
z10Fv28mYtiM+4WkNB=<8*VUJ^lTswrdt+{Hh9>&OgWX56@qyxkxI7w7ejB-b6sY-j
zvodVIUNy0BtLOg*3$P>YAsmaN1AQfIpXX2Zl(XClh#@JZD@o@Cy8Zr5UJw-3CCeVf
zyq>7ed&*TA{T1ycYIt99(P$$NvGmfzSjz6n+J(fjLt?k+my}YBF?S<kim0q%Bx(KK
zvxQfipOFmq<B8jXLVr*dGtF^Ms$<V2#)E<Ga9}&L+J`pxGtW|%&E4I~SA1mqA&Z1t
z;rpcPjQi=AD<L2lFHCkNjXpMcgyV#beGR1xzRMg8JKFyA;ubFWi|D<wGie9;)c2E3
z%a_7`B12!Nf;ai)U+q7-FZ%$yAA=3I4{y!CieMVpuYqD4CBxpGxb8iZy6wKcj?FO`
zB;op)dCBO8>j4n98zq$omR{=^&uCu|-mU@Sm>aS`c{`Y-m14U)JWJoI{nH{%<hDe^
z3HWdm1yQJrTA^M4Pgxh?u8Bz<dF~zg;@z{~JDy(s=B9)_8vJJe|NQ#DpKe3wvDVNf
zKXg1FVlXOmNC~@Q?1kb_R3H82rV>uzLlV`0dI&>8QXx^1Oc8DJc!?<#DJ=(=Vtrt*
z>5wX-1i9#EDQ(<!yJXnoa*=By=B39&ldc1}INgq$(616rxLr?`D8dSnjo-Z*UZpwA
z1{SKg?#BS$VD^9e_Xc<>Zl($(2;7aLAZG~3Axv82>?Iv8j}f*@6{1%9{}=gxb6!r-
P{)v%5w1bxa{PBMP{#akc


From fb4166419ae48e942be2b43bb042579a444de299 Mon Sep 17 00:00:00 2001
From: lyuwenyu <wenyu.lyu@gmail.com>
Date: Thu, 19 Aug 2021 11:02:30 +0800
Subject: [PATCH 02/35] using ppcls pretrained

---
 hubconf.py | 151 +++++++++++++++++------------------------------------
 1 file changed, 47 insertions(+), 104 deletions(-)

diff --git a/hubconf.py b/hubconf.py
index eb114bc20..d615a905d 100644
--- a/hubconf.py
+++ b/hubconf.py
@@ -63,9 +63,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `AlexNet` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.AlexNet(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'AlexNet')
 
         return model
 
@@ -80,9 +79,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `VGG11` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.VGG11(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'VGG11')
 
         return model
 
@@ -97,9 +95,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `VGG13` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.VGG13(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'VGG13')
 
         return model
 
@@ -114,9 +111,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `VGG16` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.VGG16(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'VGG16')
 
         return model
 
@@ -131,9 +127,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `VGG19` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.VGG19(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'VGG19')
 
         return model
 
@@ -149,9 +144,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `ResNet18` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.ResNet18(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'ResNet18')
 
         return model
 
@@ -167,9 +161,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `ResNet34` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.ResNet34(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'ResNet34')
 
         return model
 
@@ -185,9 +178,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `ResNet50` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.ResNet50(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'ResNet50')
 
         return model
 
@@ -203,9 +195,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `ResNet101` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.ResNet101(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'ResNet101')
 
         return model
 
@@ -221,9 +212,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `ResNet152` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.ResNet152(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'ResNet152')
 
         return model
 
@@ -237,9 +227,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `SqueezeNet1_0` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.SqueezeNet1_0(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'SqueezeNet1_0')
 
         return model
 
@@ -253,9 +242,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `SqueezeNet1_1` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.SqueezeNet1_1(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'SqueezeNet1_1')
 
         return model
 
@@ -271,9 +259,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `DenseNet121` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.DenseNet121(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'DenseNet121')
 
         return model
 
@@ -289,9 +276,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `DenseNet161` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.DenseNet161(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'DenseNet161')
 
         return model
 
@@ -307,9 +293,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `DenseNet169` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.DenseNet169(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'DenseNet169')
 
         return model
 
@@ -325,9 +310,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `DenseNet201` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.DenseNet201(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'DenseNet201')
 
         return model
 
@@ -343,9 +327,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `DenseNet264` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.DenseNet264(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'DenseNet264')
 
         return model
 
@@ -359,9 +342,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `InceptionV3` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.InceptionV3(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'InceptionV3')
 
         return model
 
@@ -375,9 +357,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `InceptionV4` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.InceptionV4(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'InceptionV4')
 
         return model
 
@@ -391,9 +372,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `GoogLeNet` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.GoogLeNet(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'GoogLeNet')
 
         return model
 
@@ -407,9 +387,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `ShuffleNetV2_x0_25` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.ShuffleNetV2_x0_25(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'ShuffleNetV2_x0_25')
 
         return model
 
@@ -423,9 +402,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `MobileNetV1` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.MobileNetV1(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'MobileNetV1')
 
         return model
 
@@ -439,9 +417,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `MobileNetV1_x0_25` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.MobileNetV1_x0_25(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'MobileNetV1_x0_25')
 
         return model
 
@@ -455,9 +432,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `MobileNetV1_x0_5` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.MobileNetV1_x0_5(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'MobileNetV1_x0_5')
 
         return model
 
@@ -471,9 +447,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `MobileNetV1_x0_75` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.MobileNetV1_x0_75(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'MobileNetV1_x0_75')
 
         return model
 
@@ -487,9 +462,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `MobileNetV2_x0_25` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.MobileNetV2_x0_25(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'MobileNetV2_x0_25')
 
         return model
 
@@ -503,9 +477,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `MobileNetV2_x0_5` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.MobileNetV2_x0_5(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'MobileNetV2_x0_5')
 
         return model
 
@@ -519,9 +492,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `MobileNetV2_x0_75` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.MobileNetV2_x0_75(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'MobileNetV2_x0_75')
 
         return model
 
@@ -535,9 +507,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `MobileNetV2_x1_5` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.MobileNetV2_x1_5(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'MobileNetV2_x1_5')
 
         return model
 
@@ -551,9 +522,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `MobileNetV2_x2_0` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.MobileNetV2_x2_0(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'MobileNetV2_x2_0')
 
         return model
 
@@ -567,10 +537,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `MobileNetV3_large_x0_35` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.MobileNetV3_large_x0_35(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model,
-                                                'MobileNetV3_large_x0_35')
 
         return model
 
@@ -584,10 +552,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `MobileNetV3_large_x0_5` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.MobileNetV3_large_x0_5(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model,
-                                                'MobileNetV3_large_x0_5')
 
         return model
 
@@ -601,10 +567,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `MobileNetV3_large_x0_75` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.MobileNetV3_large_x0_75(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model,
-                                                'MobileNetV3_large_x0_75')
 
         return model
 
@@ -618,10 +582,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `MobileNetV3_large_x1_0` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.MobileNetV3_large_x1_0(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model,
-                                                'MobileNetV3_large_x1_0')
 
         return model
 
@@ -635,10 +597,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `MobileNetV3_large_x1_25` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.MobileNetV3_large_x1_25(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model,
-                                                'MobileNetV3_large_x1_25')
 
         return model
 
@@ -652,10 +612,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `MobileNetV3_small_x0_35` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.MobileNetV3_small_x0_35(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model,
-                                                'MobileNetV3_small_x0_35')
 
         return model
 
@@ -669,10 +627,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `MobileNetV3_small_x0_5` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.MobileNetV3_small_x0_5(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model,
-                                                'MobileNetV3_small_x0_5')
 
         return model
 
@@ -686,10 +642,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `MobileNetV3_small_x0_75` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.MobileNetV3_small_x0_75(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model,
-                                                'MobileNetV3_small_x0_75')
 
         return model
 
@@ -703,10 +657,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `MobileNetV3_small_x1_0` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.MobileNetV3_small_x1_0(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model,
-                                                'MobileNetV3_small_x1_0')
 
         return model
 
@@ -720,10 +672,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `MobileNetV3_small_x1_25` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.MobileNetV3_small_x1_25(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model,
-                                                'MobileNetV3_small_x1_25')
 
         return model
 
@@ -737,9 +687,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `ResNeXt101_32x4d` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.ResNeXt101_32x4d(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'ResNeXt101_32x4d')
 
         return model
 
@@ -753,9 +702,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `ResNeXt101_64x4d` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.ResNeXt101_64x4d(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'ResNeXt101_64x4d')
 
         return model
 
@@ -769,9 +717,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `ResNeXt152_32x4d` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.ResNeXt152_32x4d(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'ResNeXt152_32x4d')
 
         return model
 
@@ -785,9 +732,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `ResNeXt152_64x4d` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.ResNeXt152_64x4d(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'ResNeXt152_64x4d')
 
         return model
 
@@ -801,9 +747,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `ResNeXt50_32x4d` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.ResNeXt50_32x4d(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'ResNeXt50_32x4d')
 
         return model
 
@@ -817,9 +762,8 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `ResNeXt50_64x4d` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.ResNeXt50_64x4d(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'ResNeXt50_64x4d')
 
         return model
 
@@ -833,8 +777,7 @@ with _SysPathG(
         Returns:
             model: nn.Layer. Specific `ResNeXt50_64x4d` model depends on args.
         """
+        kwargs.update({'pretrained': pretrained})
         model = backbone.DarkNet53(**kwargs)
-        if pretrained:
-            model = _load_pretrained_parameters(model, 'DarkNet53')
 
         return model

From d21ebb85d01032ece7ec7d89de04a170a14ab0a8 Mon Sep 17 00:00:00 2001
From: cuicheng01 <cuicheng_smile@163.com>
Date: Fri, 20 Aug 2021 09:53:18 +0000
Subject: [PATCH 03/35] fix random_erasing bug

---
 ppcls/data/preprocess/ops/random_erasing.py | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/ppcls/data/preprocess/ops/random_erasing.py b/ppcls/data/preprocess/ops/random_erasing.py
index d96ceda6e..b395d5205 100644
--- a/ppcls/data/preprocess/ops/random_erasing.py
+++ b/ppcls/data/preprocess/ops/random_erasing.py
@@ -42,9 +42,9 @@ class RandomErasing(object):
             h = int(round(math.sqrt(target_area * aspect_ratio)))
             w = int(round(math.sqrt(target_area / aspect_ratio)))
 
-            if w < img.shape[2] and h < img.shape[1]:
-                x1 = random.randint(0, img.shape[1] - h)
-                y1 = random.randint(0, img.shape[2] - w)
+            if w < img.shape[1] and h < img.shape[0]:
+                x1 = random.randint(0, img.shape[0] - h)
+                y1 = random.randint(0, img.shape[1] - w)
                 if img.shape[0] == 3:
                     img[x1:x1 + h, y1:y1 + w, 0] = self.mean[0]
                     img[x1:x1 + h, y1:y1 + w, 1] = self.mean[1]

From fdff2de329373d91ab86cb0a6073d4c69a27c90e Mon Sep 17 00:00:00 2001
From: cuicheng01 <45199522+cuicheng01@users.noreply.github.com>
Date: Sat, 21 Aug 2021 17:49:49 +0800
Subject: [PATCH 04/35] fix random_erasing bug (#1160)

---
 ppcls/data/preprocess/ops/random_erasing.py | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/ppcls/data/preprocess/ops/random_erasing.py b/ppcls/data/preprocess/ops/random_erasing.py
index d96ceda6e..b395d5205 100644
--- a/ppcls/data/preprocess/ops/random_erasing.py
+++ b/ppcls/data/preprocess/ops/random_erasing.py
@@ -42,9 +42,9 @@ class RandomErasing(object):
             h = int(round(math.sqrt(target_area * aspect_ratio)))
             w = int(round(math.sqrt(target_area / aspect_ratio)))
 
-            if w < img.shape[2] and h < img.shape[1]:
-                x1 = random.randint(0, img.shape[1] - h)
-                y1 = random.randint(0, img.shape[2] - w)
+            if w < img.shape[1] and h < img.shape[0]:
+                x1 = random.randint(0, img.shape[0] - h)
+                y1 = random.randint(0, img.shape[1] - w)
                 if img.shape[0] == 3:
                     img[x1:x1 + h, y1:y1 + w, 0] = self.mean[0]
                     img[x1:x1 + h, y1:y1 + w, 1] = self.mean[1]

From 800487e416c201a0f5f07f435d0e417b17af57e9 Mon Sep 17 00:00:00 2001
From: Tingquan Gao <gaotingquan@baidu.com>
Date: Sat, 21 Aug 2021 17:50:09 +0800
Subject: [PATCH 05/35] fix: fix the VDL doc, test=document_fix (#1159)

---
 docs/en/extension/VisualDL_en.md | 17 +++++++++--------
 docs/zh_CN/extension/VisualDL.md | 14 ++++++++------
 2 files changed, 17 insertions(+), 14 deletions(-)

diff --git a/docs/en/extension/VisualDL_en.md b/docs/en/extension/VisualDL_en.md
index cdd99581a..9ffd03e92 100644
--- a/docs/en/extension/VisualDL_en.md
+++ b/docs/en/extension/VisualDL_en.md
@@ -7,29 +7,30 @@ VisualDL, a visualization analysis tool of PaddlePaddle, provides a variety of c
 Now PaddleClas support use VisualDL to visualize the changes of learning rate, loss, accuracy in training.
 
 ### Set config and start training
-You only need to set the `vdl_dir` field in train config:
+You only need to set the field `Global.use_visualdl` to `True` in train config:
 
 ```yaml
 # config.yaml
-vdl_dir: "./vdl.log"
+Global:
+...
+  use_visualdl: True
+...
 ```
 
-`vdl_dir`: Specify the directory where VisualDL stores logs.
-
-Then normal start training:
+PaddleClas will save the VisualDL logs to subdirectory `vdl/` under the output directory specified by `Global.output_dir`. And then you just need to start training normally:
 
 ```shell
 python3 tools/train.py -c config.yaml
 ```
 
 ### Start VisualDL
-After starting the training program, you can start the VisualDL service in the new terminal session:
+After starting the training program, you can start the VisualDL service in a new terminal session:
 
 ```shell
- visualdl --logdir ./vdl.log
+ visualdl --logdir ./output/vdl/
 ```
 
-In the above command, `--logdir` specify the logs directory. VisualDL will traverse and iterate to find the subdirectories of the specified directory to visualize all the experimental results. You can also use the following parameters to set the IP and port number of the VisualDL service:
+In the above command, `--logdir` specify the directory of the VisualDL logs produced in training. VisualDL will traverse and iterate to find the subdirectories of the specified directory to visualize all the experimental results. You can also use the following parameters to set the IP and port number of the VisualDL service:
 
 * `--host`：ip, default is 127.0.0.1
 * `--port`：port, default is 8040
diff --git a/docs/zh_CN/extension/VisualDL.md b/docs/zh_CN/extension/VisualDL.md
index d360c5b15..3c0020063 100644
--- a/docs/zh_CN/extension/VisualDL.md
+++ b/docs/zh_CN/extension/VisualDL.md
@@ -7,15 +7,17 @@ VisualDL是飞桨可视化分析工具，以丰富的图表呈现训练参数变
 现在PaddleClas支持在训练阶段使用VisualDL查看训练过程中学习率（learning rate）、损失值（loss）以及准确率（accuracy）的变化情况。
 
 ### 设置config文件并启动训练
-在PaddleClas中使用VisualDL，只需在训练配置文件（config文件）添加如下字段：
+在PaddleClas中使用VisualDL，只需在训练配置文件（config文件）中设置字段 `Global.use_visualdl` 为 `True`：
 
 ```yaml
 # config.yaml
-vdl_dir: "./vdl.log"
+Global:
+...
+  use_visualdl: True
+...
 ```
-`vdl_dir` 用于指定VisualDL用于保存log信息的目录。
 
-然后正常启动训练即可：
+PaddleClas 会将 VisualDL 的日志保存在 `Global.output_dir` 字段指定目录下的 `vdl/` 子目录下，然后正常启动训练即可：
 
 ```shell
 python3 tools/train.py -c config.yaml
@@ -25,10 +27,10 @@ python3 tools/train.py -c config.yaml
 在启动训练程序后，可以在新的终端session中启动VisualDL服务：
 
 ```shell
- visualdl --logdir ./vdl.log
+ visualdl --logdir ./output/vdl/
  ```
 
-上述命令中，参数`--logdir`用于指定日志目录，VisualDL将遍历并且迭代寻找指定目录的子目录，将所有实验结果进行可视化。也同样可以使用下述参数设定VisualDL服务的ip及端口号：
+上述命令中，参数`--logdir`用于指定保存 VisualDL 日志的目录，VisualDL将遍历并且迭代寻找指定目录的子目录，将所有实验结果进行可视化。也同样可以使用下述参数设定VisualDL服务的ip及端口号：
 * `--host`：设定IP，默认为127.0.0.1
 * `--port`：设定端口，默认为8040
 

From ebde0e13cbf0d027c485b8cc2946087d29218453 Mon Sep 17 00:00:00 2001
From: dongshuilong <dongshuilong@baidu.com>
Date: Sun, 22 Aug 2021 15:10:23 +0000
Subject: [PATCH 06/35] refactor trainer

---
 ppcls/engine/core.py                 | 403 ++++++++++++++++
 ppcls/engine/eval/__init__.py        |  16 +
 ppcls/engine/eval/classification.py  | 118 +++++
 ppcls/engine/eval/retrieval.py       | 160 +++++++
 ppcls/engine/slim/__init__.py        |   0
 ppcls/engine/train/__init__.py       |  15 +
 ppcls/engine/train/classification.py |  89 ++++
 ppcls/engine/train/retrieval.py      |  90 ++++
 ppcls/engine/train/utils.py          |  82 ++++
 ppcls/engine/trainer.py              | 662 ---------------------------
 tools/eval.py                        |   6 +-
 tools/export_model.py                |  77 +---
 tools/infer.py                       |   7 +-
 tools/train.py                       |   4 +-
 14 files changed, 984 insertions(+), 745 deletions(-)
 create mode 100644 ppcls/engine/core.py
 create mode 100644 ppcls/engine/eval/__init__.py
 create mode 100644 ppcls/engine/eval/classification.py
 create mode 100644 ppcls/engine/eval/retrieval.py
 create mode 100644 ppcls/engine/slim/__init__.py
 create mode 100644 ppcls/engine/train/__init__.py
 create mode 100644 ppcls/engine/train/classification.py
 create mode 100644 ppcls/engine/train/retrieval.py
 create mode 100644 ppcls/engine/train/utils.py
 delete mode 100644 ppcls/engine/trainer.py

diff --git a/ppcls/engine/core.py b/ppcls/engine/core.py
new file mode 100644
index 000000000..d0e448d71
--- /dev/null
+++ b/ppcls/engine/core.py
@@ -0,0 +1,403 @@
+# Copyright (c) 2021 PaddlePaddle Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+from __future__ import absolute_import
+from __future__ import division
+from __future__ import print_function
+import os
+import sys
+import numpy as np
+
+__dir__ = os.path.dirname(os.path.abspath(__file__))
+sys.path.append(os.path.abspath(os.path.join(__dir__, '../../')))
+
+import time
+import platform
+import datetime
+import argparse
+import paddle
+import paddle.nn as nn
+import paddle.distributed as dist
+from visualdl import LogWriter
+
+from ppcls.utils.check import check_gpu
+from ppcls.utils.misc import AverageMeter
+from ppcls.utils import logger
+from ppcls.utils.logger import init_logger
+from ppcls.utils.config import print_config
+from ppcls.data import build_dataloader
+from ppcls.arch import build_model, RecModel, DistillationModel
+from ppcls.arch import apply_to_static
+from ppcls.loss import build_loss
+from ppcls.metric import build_metrics
+from ppcls.optimizer import build_optimizer
+from ppcls.utils.save_load import load_dygraph_pretrain, load_dygraph_pretrain_from_url
+from ppcls.utils.save_load import init_model
+from ppcls.utils import save_load
+
+from ppcls.data.utils.get_image_list import get_image_list
+from ppcls.data.postprocess import build_postprocess
+from ppcls.data import create_operators
+from ppcls.engine.train import classification_train, retrieval_train
+from ppcls.engine.eval import classification_eval, retrieval_eval
+from ppcls.arch.gears.identity_head import IdentityHead
+
+
+class Core(object):
+    def __init__(self, config, mode="train"):
+        assert mode in ['train', 'eval', 'infer', 'export']
+        self.mode = mode
+        self.config = config
+        self.eval_mode = self.config["Global"].get("eval_mode",
+                                                   "classification")
+
+        # init logger
+        self.output_dir = self.config['Global']['output_dir']
+        log_file = os.path.join(self.output_dir, self.config["Arch"]["name"],
+                                f"{mode}.log")
+        init_logger(name='root', log_file=log_file)
+        print_config(config)
+
+        # init train_func and eval_func
+        if self.eval_mode == "classification":
+            self.evaler = classification_eval
+            self.trainer = classification_train
+        elif self.eval_mode == "retrieval":
+            self.trainer = retrieval_train
+            self.evaler = retrieval_eval
+        else:
+            logger.warning("Invalid eval mode: {}".format(self.eval_mode))
+        self.use_dali = self.config['Global'].get("use_dali", False)
+
+        # for visualdl
+        self.vdl_writer = None
+        if self.config['Global']['use_visualdl'] and mode == "train":
+            vdl_writer_path = os.path.join(self.output_dir, "vdl")
+            if not os.path.exists(vdl_writer_path):
+                os.makedirs(vdl_writer_path)
+            self.vdl_writer = LogWriter(logdir=vdl_writer_path)
+
+        # set device
+        assert self.config["Global"]["device"] in ["cpu", "gpu", "xpu"]
+        self.device = paddle.set_device(self.config["Global"]["device"])
+        logger.info('train with paddle {} and device {}'.format(
+            paddle.__version__, self.device))
+
+        # AMP training
+        self.amp = True if "AMP" in self.config else False
+        if self.amp and self.config["AMP"] is not None:
+            self.scale_loss = self.config["AMP"].get("scale_loss", 1.0)
+            self.use_dynamic_loss_scaling = self.config["AMP"].get(
+                "use_dynamic_loss_scaling", False)
+        else:
+            self.scale_loss = 1.0
+            self.use_dynamic_loss_scaling = False
+        if self.amp:
+            AMP_RELATED_FLAGS_SETTING = {
+                'FLAGS_cudnn_batchnorm_spatial_persistent': 1,
+                'FLAGS_max_inplace_grad_add': 8,
+            }
+            paddle.fluid.set_flags(AMP_RELATED_FLAGS_SETTING)
+
+        # build dataloader
+        if self.mode == 'train':
+            self.train_dataloader = build_dataloader(
+                self.config["DataLoader"], "Train", self.device, self.use_dali)
+        if self.mode in ["train", "eval"]:
+            if self.eval_mode == "classification":
+                self.eval_dataloader = build_dataloader(
+                    self.config["DataLoader"], "Eval", self.device,
+                    self.use_dali)
+            elif self.eval_mode == "retrieval":
+                self.gallery_dataloader = build_dataloader(
+                    self.config["DataLoader"]["Eval"], "Gallery", self.device,
+                    self.use_dali)
+                self.query_dataloader = build_dataloader(
+                    self.config["DataLoader"]["Eval"], "Query", self.device,
+                    self.use_dali)
+
+        # build loss
+        if self.mode == "train":
+            loss_info = self.config["Loss"]["Train"]
+            self.train_loss_func = build_loss(loss_info)
+        if self.mode in ["train", "eval"]:
+            loss_config = self.config.get("Loss", None)
+            if loss_config is not None:
+                loss_config = loss_config.get("Eval")
+                if loss_config is not None:
+                    self.eval_loss_func = build_loss(loss_config)
+                else:
+                    self.eval_loss_func = None
+            else:
+                self.eval_loss_func = None
+
+        # build metric
+        if self.mode == 'train':
+            metric_config = self.config.get("Metric")
+            if metric_config is not None:
+                metric_config = metric_config.get("Train")
+                if metric_config is not None:
+                    self.train_metric_func = build_metrics(metric_config)
+                else:
+                    self.train_metric_func = None
+        else:
+            self.train_metric_func = None
+
+        if self.mode in ["train", "eval"]:
+            metric_config = self.config.get("Metric")
+            if self.eval_mode == "classification":
+                if metric_config is not None:
+                    metric_config = metric_config.get("Eval")
+                    if metric_config is not None:
+                        self.eval_metric_func = build_metrics(metric_config)
+            elif self.eval_mode == "retrieval":
+                if metric_config is None:
+                    metric_config = [{"name": "Recallk", "topk": (1, 5)}]
+                else:
+                    metric_config = metric_config["Eval"]
+                self.eval_metric_func = build_metrics(metric_config)
+        else:
+            self.eval_metric_func = None
+
+        # build model
+        self.model = build_model(self.config["Arch"])
+        # set @to_static for benchmark, skip this by default.
+        apply_to_static(self.config, self.model)
+        # load_pretrain
+        if self.config["Global"]["pretrained_model"] is not None:
+            if self.config["Global"]["pretrained_model"].startswith("http"):
+                load_dygraph_pretrain_from_url(
+                    self.model, self.config["Global"]["pretrained_model"])
+            else:
+                load_dygraph_pretrain(
+                    self.model, self.config["Global"]["pretrained_model"])
+
+        # for slim
+
+        # build optimizer
+        if self.mode == 'train':
+            self.optimizer, self.lr_sch = build_optimizer(
+                self.config["Optimizer"], self.config["Global"]["epochs"],
+                len(self.train_dataloader), self.model.parameters())
+
+        # for distributed
+        self.config["Global"][
+            "distributed"] = paddle.distributed.get_world_size() != 1
+        if self.config["Global"]["distributed"]:
+            dist.init_parallel_env()
+        if self.config["Global"]["distributed"]:
+            self.model = paddle.DataParallel(self.model)
+
+        # build postprocess for infer
+        if self.mode == 'infer':
+            self.preprocess_func = create_operators(self.config["Infer"][
+                "transforms"])
+            self.postprocess_func = build_postprocess(self.config["Infer"][
+                "PostProcess"])
+
+    def train(self):
+        assert self.mode == "train"
+        print_batch_step = self.config['Global']['print_batch_step']
+        save_interval = self.config["Global"]["save_interval"]
+        best_metric = {
+            "metric": 0.0,
+            "epoch": 0,
+        }
+        # key:
+        # val: metrics list word
+        self.output_info = dict()
+        self.time_info = {
+            "batch_cost": AverageMeter(
+                "batch_cost", '.5f', postfix=" s,"),
+            "reader_cost": AverageMeter(
+                "reader_cost", ".5f", postfix=" s,"),
+        }
+        # global iter counter
+        self.global_step = 0
+
+        if self.config["Global"]["checkpoints"] is not None:
+            metric_info = init_model(self.config["Global"], self.model,
+                                     self.optimizer)
+            if metric_info is not None:
+                best_metric.update(metric_info)
+
+        # for amp training
+        if self.amp:
+            self.scaler = paddle.amp.GradScaler(
+                init_loss_scaling=self.scale_loss,
+                use_dynamic_loss_scaling=self.use_dynamic_loss_scaling)
+
+        self.max_iter = len(self.train_dataloader) - 1 if platform.system(
+        ) == "Windows" else len(self.train_dataloader)
+        for epoch_id in range(best_metric["epoch"] + 1,
+                              self.config["Global"]["epochs"] + 1):
+            acc = 0.0
+            # for one epoch train
+            self.trainer(self, epoch_id, print_batch_step)
+
+            if self.use_dali:
+                self.train_dataloader.reset()
+            metric_msg = ", ".join([
+                "{}: {:.5f}".format(key, self.output_info[key].avg)
+                for key in self.output_info
+            ])
+            logger.info("[Train][Epoch {}/{}][Avg]{}".format(
+                epoch_id, self.config["Global"]["epochs"], metric_msg))
+            self.output_info.clear()
+
+            # eval model and save model if possible
+            if self.config["Global"][
+                    "eval_during_train"] and epoch_id % self.config["Global"][
+                        "eval_interval"] == 0:
+                acc = self.eval(epoch_id)
+                if acc > best_metric["metric"]:
+                    best_metric["metric"] = acc
+                    best_metric["epoch"] = epoch_id
+                    save_load.save_model(
+                        self.model,
+                        self.optimizer,
+                        best_metric,
+                        self.output_dir,
+                        model_name=self.config["Arch"]["name"],
+                        prefix="best_model")
+                logger.info("[Eval][Epoch {}][best metric: {}]".format(
+                    epoch_id, best_metric["metric"]))
+                logger.scaler(
+                    name="eval_acc",
+                    value=acc,
+                    step=epoch_id,
+                    writer=self.vdl_writer)
+
+                self.model.train()
+
+            # save model
+            if epoch_id % save_interval == 0:
+                save_load.save_model(
+                    self.model,
+                    self.optimizer, {"metric": acc,
+                                     "epoch": epoch_id},
+                    self.output_dir,
+                    model_name=self.config["Arch"]["name"],
+                    prefix="epoch_{}".format(epoch_id))
+                # save the latest model
+                save_load.save_model(
+                    self.model,
+                    self.optimizer, {"metric": acc,
+                                     "epoch": epoch_id},
+                    self.output_dir,
+                    model_name=self.config["Arch"]["name"],
+                    prefix="latest")
+
+        if self.vdl_writer is not None:
+            self.vdl_writer.close()
+
+    @paddle.no_grad()
+    def eval(self, epoch_id=0):
+        assert self.mode in ["train", "eval"]
+        self.model.eval()
+        eval_result = self.evaler(self, epoch_id)
+        self.model.train()
+        return eval_result
+
+    @paddle.no_grad()
+    def infer(self):
+        assert self.mode == "infer" and self.eval_mode == "classification"
+        total_trainer = paddle.distributed.get_world_size()
+        local_rank = paddle.distributed.get_rank()
+        image_list = get_image_list(self.config["Infer"]["infer_imgs"])
+        # data split
+        image_list = image_list[local_rank::total_trainer]
+
+        batch_size = self.config["Infer"]["batch_size"]
+        self.model.eval()
+        batch_data = []
+        image_file_list = []
+        for idx, image_file in enumerate(image_list):
+            with open(image_file, 'rb') as f:
+                x = f.read()
+            for process in self.preprocess_func:
+                x = process(x)
+            batch_data.append(x)
+            image_file_list.append(image_file)
+            if len(batch_data) >= batch_size or idx == len(image_list) - 1:
+                batch_tensor = paddle.to_tensor(batch_data)
+                out = self.model(batch_tensor)
+                if isinstance(out, list):
+                    out = out[0]
+                result = self.postprocess_func(out, image_file_list)
+                print(result)
+                batch_data.clear()
+                image_file_list.clear()
+
+    def export(self):
+        assert self.mode == "export"
+        model = ExportModel(self.config["Arch"], self.model)
+        if self.config["Global"]["pretrained_model"] is not None:
+            load_dygraph_pretrain(model.base_model,
+                                  self.config["Global"]["pretrained_model"])
+
+        model.eval()
+
+        model = paddle.jit.to_static(
+            model,
+            input_spec=[
+                paddle.static.InputSpec(
+                    shape=[None] + self.config["Global"]["image_shape"],
+                    dtype='float32')
+            ])
+        paddle.jit.save(
+            model,
+            os.path.join(self.config["Global"]["save_inference_dir"],
+                         "inference"))
+
+
+class ExportModel(nn.Layer):
+    """
+    ExportModel: add softmax onto the model
+    """
+
+    def __init__(self, config, model):
+        super().__init__()
+        self.base_model = model
+
+        # we should choose a final model to export
+        if isinstance(self.base_model, DistillationModel):
+            self.infer_model_name = config["infer_model_name"]
+        else:
+            self.infer_model_name = None
+
+        self.infer_output_key = config.get("infer_output_key", None)
+        if self.infer_output_key == "features" and isinstance(self.base_model,
+                                                              RecModel):
+            self.base_model.head = IdentityHead()
+        if config.get("infer_add_softmax", True):
+            self.softmax = nn.Softmax(axis=-1)
+        else:
+            self.softmax = None
+
+    def eval(self):
+        self.training = False
+        for layer in self.sublayers():
+            layer.training = False
+            layer.eval()
+
+    def forward(self, x):
+        x = self.base_model(x)
+        if isinstance(x, list):
+            x = x[0]
+        if self.infer_model_name is not None:
+            x = x[self.infer_model_name]
+        if self.infer_output_key is not None:
+            x = x[self.infer_output_key]
+        if self.softmax is not None:
+            x = self.softmax(x)
+        return x
diff --git a/ppcls/engine/eval/__init__.py b/ppcls/engine/eval/__init__.py
new file mode 100644
index 000000000..0cc0dc983
--- /dev/null
+++ b/ppcls/engine/eval/__init__.py
@@ -0,0 +1,16 @@
+# Copyright (c) 2021 PaddlePaddle Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+
+from ppcls.engine.eval.classification import classification_eval
+from ppcls.engine.eval.retrieval import retrieval_eval
diff --git a/ppcls/engine/eval/classification.py b/ppcls/engine/eval/classification.py
new file mode 100644
index 000000000..38be7cd3f
--- /dev/null
+++ b/ppcls/engine/eval/classification.py
@@ -0,0 +1,118 @@
+# Copyright (c) 2021 PaddlePaddle Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+from __future__ import absolute_import
+from __future__ import division
+from __future__ import print_function
+import os
+import sys
+import time
+import platform
+import paddle
+
+__dir__ = os.path.dirname(os.path.abspath(__file__))
+sys.path.append(os.path.abspath(os.path.join(__dir__, '../../../')))
+from ppcls.utils.misc import AverageMeter
+from ppcls.utils import logger
+
+
+def classification_eval(evaler, epoch_id=0):
+    output_info = dict()
+    time_info = {
+        "batch_cost": AverageMeter(
+            "batch_cost", '.5f', postfix=" s,"),
+        "reader_cost": AverageMeter(
+            "reader_cost", ".5f", postfix=" s,"),
+    }
+    print_batch_step = evaler.config["Global"]["print_batch_step"]
+
+    metric_key = None
+    tic = time.time()
+    eval_dataloader = evaler.eval_dataloader if evaler.use_dali else evaler.eval_dataloader(
+    )
+    max_iter = len(evaler.eval_dataloader) - 1 if platform.system(
+    ) == "Windows" else len(evaler.eval_dataloader)
+    for iter_id, batch in enumerate(eval_dataloader):
+        if iter_id >= max_iter:
+            break
+        if iter_id == 5:
+            for key in time_info:
+                time_info[key].reset()
+        if evaler.use_dali:
+            batch = [
+                paddle.to_tensor(batch[0]['data']),
+                paddle.to_tensor(batch[0]['label'])
+            ]
+        time_info["reader_cost"].update(time.time() - tic)
+        batch_size = batch[0].shape[0]
+        batch[0] = paddle.to_tensor(batch[0]).astype("float32")
+        batch[1] = batch[1].reshape([-1, 1]).astype("int64")
+        # image input
+        out = evaler.model(batch[0])
+        # calc loss
+        if evaler.eval_loss_func is not None:
+            loss_dict = evaler.eval_loss_func(out, batch[-1])
+            for key in loss_dict:
+                if key not in output_info:
+                    output_info[key] = AverageMeter(key, '7.5f')
+                output_info[key].update(loss_dict[key].numpy()[0], batch_size)
+        # calc metric
+        if evaler.eval_metric_func is not None:
+            metric_dict = evaler.eval_metric_func(out, batch[-1])
+            if paddle.distributed.get_world_size() > 1:
+                for key in metric_dict:
+                    paddle.distributed.all_reduce(
+                        metric_dict[key], op=paddle.distributed.ReduceOp.SUM)
+                    metric_dict[key] = metric_dict[
+                        key] / paddle.distributed.get_world_size()
+            for key in metric_dict:
+                if metric_key is None:
+                    metric_key = key
+                if key not in output_info:
+                    output_info[key] = AverageMeter(key, '7.5f')
+
+                output_info[key].update(metric_dict[key].numpy()[0],
+                                        batch_size)
+
+        time_info["batch_cost"].update(time.time() - tic)
+
+        if iter_id % print_batch_step == 0:
+            time_msg = "s, ".join([
+                "{}: {:.5f}".format(key, time_info[key].avg)
+                for key in time_info
+            ])
+
+            ips_msg = "ips: {:.5f} images/sec".format(
+                batch_size / time_info["batch_cost"].avg)
+
+            metric_msg = ", ".join([
+                "{}: {:.5f}".format(key, output_info[key].val)
+                for key in output_info
+            ])
+            logger.info("[Eval][Epoch {}][Iter: {}/{}]{}, {}, {}".format(
+                epoch_id, iter_id,
+                len(evaler.eval_dataloader), metric_msg, time_msg, ips_msg))
+
+        tic = time.time()
+    if evaler.use_dali:
+        evaler.eval_dataloader.reset()
+    metric_msg = ", ".join([
+        "{}: {:.5f}".format(key, output_info[key].avg) for key in output_info
+    ])
+    logger.info("[Eval][Epoch {}][Avg]{}".format(epoch_id, metric_msg))
+
+    # do not try to save best eval.model
+    if evaler.eval_metric_func is None:
+        return -1
+    # return 1st metric in the dict
+    return output_info[metric_key].avg
diff --git a/ppcls/engine/eval/retrieval.py b/ppcls/engine/eval/retrieval.py
new file mode 100644
index 000000000..ccba31754
--- /dev/null
+++ b/ppcls/engine/eval/retrieval.py
@@ -0,0 +1,160 @@
+# Copyright (c) 2021 PaddlePaddle Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+from __future__ import absolute_import
+from __future__ import division
+from __future__ import print_function
+import os
+import sys
+import time
+import platform
+import paddle
+
+__dir__ = os.path.dirname(os.path.abspath(__file__))
+sys.path.append(os.path.abspath(os.path.join(__dir__, '../../../')))
+from ppcls.utils.misc import AverageMeter
+from ppcls.utils import logger
+
+
+def retrieval_eval(evaler, epoch_id=0):
+    evaler.model.eval()
+    # step1. build gallery
+    gallery_feas, gallery_img_id, gallery_unique_id = cal_feature(
+        evaler, name='gallery')
+    query_feas, query_img_id, query_query_id = cal_feature(
+        evaler, name='query')
+
+    # step2. do evaluation
+    sim_block_size = evaler.config["Global"].get("sim_block_size", 64)
+    sections = [sim_block_size] * (len(query_feas) // sim_block_size)
+    if len(query_feas) % sim_block_size:
+        sections.append(len(query_feas) % sim_block_size)
+    fea_blocks = paddle.split(query_feas, num_or_sections=sections)
+    if query_query_id is not None:
+        query_id_blocks = paddle.split(
+            query_query_id, num_or_sections=sections)
+    image_id_blocks = paddle.split(query_img_id, num_or_sections=sections)
+    metric_key = None
+
+    if evaler.eval_loss_func is None:
+        metric_dict = {metric_key: 0.}
+    else:
+        metric_dict = dict()
+        for block_idx, block_fea in enumerate(fea_blocks):
+            similarity_matrix = paddle.matmul(
+                block_fea, gallery_feas, transpose_y=True)
+            if query_query_id is not None:
+                query_id_block = query_id_blocks[block_idx]
+                query_id_mask = (query_id_block != gallery_unique_id.t())
+
+                image_id_block = image_id_blocks[block_idx]
+                image_id_mask = (image_id_block != gallery_img_id.t())
+
+                keep_mask = paddle.logical_or(query_id_mask, image_id_mask)
+                similarity_matrix = similarity_matrix * keep_mask.astype(
+                    "float32")
+            else:
+                keep_mask = None
+
+            metric_tmp = evaler.eval_metric_func(similarity_matrix,
+                                                 image_id_blocks[block_idx],
+                                                 gallery_img_id, keep_mask)
+
+            for key in metric_tmp:
+                if key not in metric_dict:
+                    metric_dict[key] = metric_tmp[key] * block_fea.shape[
+                        0] / len(query_feas)
+                else:
+                    metric_dict[key] += metric_tmp[key] * block_fea.shape[
+                        0] / len(query_feas)
+
+    metric_info_list = []
+    for key in metric_dict:
+        if metric_key is None:
+            metric_key = key
+        metric_info_list.append("{}: {:.5f}".format(key, metric_dict[key]))
+    metric_msg = ", ".join(metric_info_list)
+    logger.info("[Eval][Epoch {}][Avg]{}".format(epoch_id, metric_msg))
+
+    return metric_dict[metric_key]
+
+
+def cal_feature(evaler, name='gallery'):
+    all_feas = None
+    all_image_id = None
+    all_unique_id = None
+    has_unique_id = False
+
+    if name == 'gallery':
+        dataloader = evaler.gallery_dataloader
+    elif name == 'query':
+        dataloader = evaler.query_dataloader
+    else:
+        raise RuntimeError("Only support gallery or query dataset")
+
+    max_iter = len(dataloader) - 1 if platform.system() == "Windows" else len(
+        dataloader)
+    dataloader_tmp = dataloader if evaler.use_dali else dataloader()
+    for idx, batch in enumerate(dataloader_tmp):  # load is very time-consuming
+        if idx >= max_iter:
+            break
+        if idx % evaler.config["Global"]["print_batch_step"] == 0:
+            logger.info(
+                f"{name} feature calculation process: [{idx}/{len(dataloader)}]"
+            )
+        if evaler.use_dali:
+            batch = [
+                paddle.to_tensor(batch[0]['data']),
+                paddle.to_tensor(batch[0]['label'])
+            ]
+        batch = [paddle.to_tensor(x) for x in batch]
+        batch[1] = batch[1].reshape([-1, 1]).astype("int64")
+        if len(batch) == 3:
+            has_unique_id = True
+            batch[2] = batch[2].reshape([-1, 1]).astype("int64")
+        out = evaler.model(batch[0], batch[1])
+        batch_feas = out["features"]
+
+        # do norm
+        if evaler.config["Global"].get("feature_normalize", True):
+            feas_norm = paddle.sqrt(
+                paddle.sum(paddle.square(batch_feas), axis=1, keepdim=True))
+            batch_feas = paddle.divide(batch_feas, feas_norm)
+
+        if all_feas is None:
+            all_feas = batch_feas
+            if has_unique_id:
+                all_unique_id = batch[2]
+            all_image_id = batch[1]
+        else:
+            all_feas = paddle.concat([all_feas, batch_feas])
+            all_image_id = paddle.concat([all_image_id, batch[1]])
+            if has_unique_id:
+                all_unique_id = paddle.concat([all_unique_id, batch[2]])
+    if evaler.use_dali:
+        dataloader_tmp.reset()
+    if paddle.distributed.get_world_size() > 1:
+        feat_list = []
+        img_id_list = []
+        unique_id_list = []
+        paddle.distributed.all_gather(feat_list, all_feas)
+        paddle.distributed.all_gather(img_id_list, all_image_id)
+        all_feas = paddle.concat(feat_list, axis=0)
+        all_image_id = paddle.concat(img_id_list, axis=0)
+        if has_unique_id:
+            paddle.distributed.all_gather(unique_id_list, all_unique_id)
+            all_unique_id = paddle.concat(unique_id_list, axis=0)
+
+    logger.info("Build {} done, all feat shape: {}, begin to eval..".format(
+        name, all_feas.shape))
+    return all_feas, all_image_id, all_unique_id
diff --git a/ppcls/engine/slim/__init__.py b/ppcls/engine/slim/__init__.py
new file mode 100644
index 000000000..e69de29bb
diff --git a/ppcls/engine/train/__init__.py b/ppcls/engine/train/__init__.py
new file mode 100644
index 000000000..897c5846d
--- /dev/null
+++ b/ppcls/engine/train/__init__.py
@@ -0,0 +1,15 @@
+# Copyright (c) 2021 PaddlePaddle Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+from ppcls.engine.train.classification import classification_train
+from ppcls.engine.train.retrieval import retrieval_train
diff --git a/ppcls/engine/train/classification.py b/ppcls/engine/train/classification.py
new file mode 100644
index 000000000..c3d74f962
--- /dev/null
+++ b/ppcls/engine/train/classification.py
@@ -0,0 +1,89 @@
+# Copyright (c) 2021 PaddlePaddle Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+from __future__ import absolute_import, division, print_function
+
+import datetime
+import os
+import platform
+import sys
+import time
+
+import numpy as np
+import paddle
+
+__dir__ = os.path.dirname(os.path.abspath(__file__))
+sys.path.append(os.path.abspath(os.path.join(__dir__, '../../../')))
+from ppcls.utils import logger
+from ppcls.utils.misc import AverageMeter
+from ppcls.engine.train.utils import update_loss, update_metric, log_info
+
+
+def classification_train(trainer, epoch_id, print_batch_step):
+    tic = time.time()
+
+    train_dataloader = trainer.train_dataloader if trainer.use_dali else trainer.train_dataloader(
+    )
+    for iter_id, batch in enumerate(train_dataloader):
+        if iter_id >= trainer.max_iter:
+            break
+        if iter_id == 5:
+            for key in trainer.time_info:
+                trainer.time_info[key].reset()
+        trainer.time_info["reader_cost"].update(time.time() - tic)
+        if trainer.use_dali:
+            batch = [
+                paddle.to_tensor(batch[0]['data']),
+                paddle.to_tensor(batch[0]['label'])
+            ]
+        batch_size = batch[0].shape[0]
+        batch[1] = batch[1].reshape([-1, 1]).astype("int64")
+
+        trainer.global_step += 1
+        # image input
+        if trainer.amp:
+            with paddle.amp.auto_cast(custom_black_list={
+                    "flatten_contiguous_range", "greater_than"
+            }):
+                out = trainer.model(batch[0])
+                loss_dict = trainer.train_loss_func(out, batch[1])
+        else:
+            out = trainer.model(batch[0])
+
+        # calc loss
+        if trainer.config["DataLoader"]["Train"]["dataset"].get(
+                "batch_transform_ops", None):
+            loss_dict = trainer.train_loss_func(out, batch[1:])
+        else:
+            loss_dict = trainer.train_loss_func(out, batch[1])
+
+        # step opt and lr
+        if trainer.amp:
+            scaled = trainer.scaler.scale(loss_dict["loss"])
+            scaled.backward()
+            trainer.scaler.minimize(trainer.optimizer, scaled)
+        else:
+            loss_dict["loss"].backward()
+            trainer.optimizer.step()
+        trainer.optimizer.clear_grad()
+        trainer.lr_sch.step()
+
+        # below code just for logging
+        # update metric_for_logger
+        update_metric(trainer, out, batch, batch_size)
+        # update_loss_for_logger
+        update_loss(trainer, loss_dict, batch_size)
+        trainer.time_info["batch_cost"].update(time.time() - tic)
+        if iter_id % print_batch_step == 0:
+            log_info(trainer, batch_size, epoch_id, iter_id)
+        tic = time.time()
diff --git a/ppcls/engine/train/retrieval.py b/ppcls/engine/train/retrieval.py
new file mode 100644
index 000000000..d4a994597
--- /dev/null
+++ b/ppcls/engine/train/retrieval.py
@@ -0,0 +1,90 @@
+# Copyright (c) 2021 PaddlePaddle Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+from __future__ import absolute_import, division, print_function
+
+import datetime
+import os
+import platform
+import sys
+import time
+
+import numpy as np
+import paddle
+
+__dir__ = os.path.dirname(os.path.abspath(__file__))
+sys.path.append(os.path.abspath(os.path.join(__dir__, '../../../')))
+from ppcls.utils import logger
+from ppcls.utils.misc import AverageMeter
+from ppcls.engine.train.utils import update_loss, update_metric, log_info
+
+
+def retrieval_train(trainer, epoch_id, print_batch_step):
+    tic = time.time()
+
+    train_dataloader = trainer.train_dataloader if trainer.use_dali else trainer.train_dataloader(
+    )
+    for iter_id, batch in enumerate(train_dataloader):
+        if iter_id >= trainer.max_iter:
+            break
+        if iter_id == 5:
+            for key in trainer.time_info:
+                trainer.time_info[key].reset()
+        trainer.time_info["reader_cost"].update(time.time() - tic)
+        if trainer.use_dali:
+            batch = [
+                paddle.to_tensor(batch[0]['data']),
+                paddle.to_tensor(batch[0]['label'])
+            ]
+        batch_size = batch[0].shape[0]
+        batch[1] = batch[1].reshape([-1, 1]).astype("int64")
+
+        trainer.global_step += 1
+        # image input
+        if trainer.amp:
+            with paddle.amp.auto_cast(custom_black_list={
+                    "flatten_contiguous_range", "greater_than"
+            }):
+                out = trainer.model(batch[0], batch[1])
+                loss_dict = trainer.train_loss_func(out, batch[1])
+        else:
+            out = trainer.model(batch[0], batch[1])
+
+        # calc loss
+        if trainer.config["DataLoader"]["Train"]["dataset"].get(
+                "batch_transform_ops", None):
+            loss_dict = trainer.train_loss_func(out, batch[1:])
+        else:
+            loss_dict = trainer.train_loss_func(out, batch[1])
+
+        # step opt and lr
+        if trainer.amp:
+            scaled = trainer.scaler.scale(loss_dict["loss"])
+            scaled.backward()
+            trainer.scaler.minimize(trainer.optimizer, scaled)
+        else:
+            loss_dict["loss"].backward()
+            trainer.optimizer.step()
+        trainer.optimizer.clear_grad()
+        trainer.lr_sch.step()
+
+        # below code just for logging
+        # update metric_for_logger
+        update_metric(trainer, out, batch, batch_size)
+
+        # update_loss_for_logger
+        update_loss(trainer, loss_dict, batch_size)
+        trainer.time_info["batch_cost"].update(time.time() - tic)
+        if iter_id % print_batch_step == 0:
+            log_info(trainer, batch_size, epoch_id, iter_id)
+        tic = time.time()
diff --git a/ppcls/engine/train/utils.py b/ppcls/engine/train/utils.py
new file mode 100644
index 000000000..f4cbce97e
--- /dev/null
+++ b/ppcls/engine/train/utils.py
@@ -0,0 +1,82 @@
+# Copyright (c) 2021 PaddlePaddle Authors. All Rights Reserved.
+#
+# Licensed under the Apache License, Version 2.0 (the "License");
+# you may not use this file except in compliance with the License.
+# You may obtain a copy of the License at
+#
+#     http://www.apache.org/licenses/LICENSE-2.0
+#
+# Unless required by applicable law or agreed to in writing, software
+# distributed under the License is distributed on an "AS IS" BASIS,
+# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+# See the License for the specific language governing permissions and
+# limitations under the License.
+from __future__ import absolute_import, division, print_function
+
+import datetime
+import os
+import platform
+import sys
+import time
+
+import numpy as np
+import paddle
+
+__dir__ = os.path.dirname(os.path.abspath(__file__))
+sys.path.append(os.path.abspath(os.path.join(__dir__, '../../../')))
+from ppcls.utils import logger
+from ppcls.utils.misc import AverageMeter
+
+
+def update_metric(trainer, out, batch, batch_size):
+    # calc metric
+    if trainer.train_metric_func is not None:
+        metric_dict = trainer.train_metric_func(out, batch[-1])
+        for key in metric_dict:
+            if key not in trainer.output_info:
+                trainer.output_info[key] = AverageMeter(key, '7.5f')
+            trainer.output_info[key].update(metric_dict[key].numpy()[0],
+                                            batch_size)
+
+
+def update_loss(trainer, loss_dict, batch_size):
+    # update_output_info
+    for key in loss_dict:
+        if key not in trainer.output_info:
+            trainer.output_info[key] = AverageMeter(key, '7.5f')
+    trainer.output_info[key].update(loss_dict[key].numpy()[0], batch_size)
+
+
+def log_info(trainer, batch_size, epoch_id, iter_id):
+    lr_msg = "lr: {:.5f}".format(trainer.lr_sch.get_lr())
+    metric_msg = ", ".join([
+        "{}: {:.5f}".format(key, trainer.output_info[key].avg)
+        for key in trainer.output_info
+    ])
+    time_msg = "s, ".join([
+        "{}: {:.5f}".format(key, trainer.time_info[key].avg)
+        for key in trainer.time_info
+    ])
+
+    ips_msg = "ips: {:.5f} images/sec".format(
+        batch_size / trainer.time_info["batch_cost"].avg)
+    eta_sec = ((trainer.config["Global"]["epochs"] - epoch_id + 1
+                ) * len(trainer.train_dataloader) - iter_id
+               ) * trainer.time_info["batch_cost"].avg
+    eta_msg = "eta: {:s}".format(str(datetime.timedelta(seconds=int(eta_sec))))
+    logger.info("[Train][Epoch {}/{}][Iter: {}/{}]{}, {}, {}, {}, {}".format(
+        epoch_id, trainer.config["Global"]["epochs"], iter_id,
+        len(trainer.train_dataloader), lr_msg, metric_msg, time_msg, ips_msg,
+        eta_msg))
+
+    logger.scaler(
+        name="lr",
+        value=trainer.lr_sch.get_lr(),
+        step=trainer.global_step,
+        writer=trainer.vdl_writer)
+    for key in trainer.output_info:
+        logger.scaler(
+            name="train_{}".format(key),
+            value=trainer.output_info[key].avg,
+            step=trainer.global_step,
+            writer=trainer.vdl_writer)
diff --git a/ppcls/engine/trainer.py b/ppcls/engine/trainer.py
deleted file mode 100644
index 14b7547dd..000000000
--- a/ppcls/engine/trainer.py
+++ /dev/null
@@ -1,662 +0,0 @@
-# Copyright (c) 2021 PaddlePaddle Authors. All Rights Reserved.
-#
-# Licensed under the Apache License, Version 2.0 (the "License");
-# you may not use this file except in compliance with the License.
-# You may obtain a copy of the License at
-#
-#     http://www.apache.org/licenses/LICENSE-2.0
-#
-# Unless required by applicable law or agreed to in writing, software
-# distributed under the License is distributed on an "AS IS" BASIS,
-# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-# See the License for the specific language governing permissions and
-# limitations under the License.
-from __future__ import absolute_import
-from __future__ import division
-from __future__ import print_function
-import os
-import sys
-import numpy as np
-
-__dir__ = os.path.dirname(os.path.abspath(__file__))
-sys.path.append(os.path.abspath(os.path.join(__dir__, '../../')))
-
-import time
-import platform
-import datetime
-import argparse
-import paddle
-import paddle.nn as nn
-import paddle.distributed as dist
-from visualdl import LogWriter
-
-from ppcls.utils.check import check_gpu
-from ppcls.utils.misc import AverageMeter
-from ppcls.utils import logger
-from ppcls.utils.logger import init_logger
-from ppcls.utils.config import print_config
-from ppcls.data import build_dataloader
-from ppcls.arch import build_model
-from ppcls.arch import apply_to_static
-from ppcls.loss import build_loss
-from ppcls.metric import build_metrics
-from ppcls.optimizer import build_optimizer
-from ppcls.utils.save_load import load_dygraph_pretrain, load_dygraph_pretrain_from_url
-from ppcls.utils.save_load import init_model
-from ppcls.utils import save_load
-
-from ppcls.data.utils.get_image_list import get_image_list
-from ppcls.data.postprocess import build_postprocess
-from ppcls.data import create_operators
-
-
-class Trainer(object):
-    def __init__(self, config, mode="train"):
-        self.mode = mode
-        self.config = config
-        self.output_dir = self.config['Global']['output_dir']
-
-        log_file = os.path.join(self.output_dir, self.config["Arch"]["name"],
-                                f"{mode}.log")
-        init_logger(name='root', log_file=log_file)
-        print_config(config)
-        # set device
-        assert self.config["Global"]["device"] in ["cpu", "gpu", "xpu"]
-        self.device = paddle.set_device(self.config["Global"]["device"])
-        # set dist
-        self.config["Global"][
-            "distributed"] = paddle.distributed.get_world_size() != 1
-        if self.config["Global"]["distributed"]:
-            dist.init_parallel_env()
-
-        if "Head" in self.config["Arch"]:
-            self.is_rec = True
-        else:
-            self.is_rec = False
-
-        self.model = build_model(self.config["Arch"])
-        # set @to_static for benchmark, skip this by default.
-        apply_to_static(self.config, self.model)
-
-        if self.config["Global"]["pretrained_model"] is not None:
-            if self.config["Global"]["pretrained_model"].startswith("http"):
-                load_dygraph_pretrain_from_url(
-                    self.model, self.config["Global"]["pretrained_model"])
-            else:
-                load_dygraph_pretrain(
-                    self.model, self.config["Global"]["pretrained_model"])
-
-        if self.config["Global"]["distributed"]:
-            self.model = paddle.DataParallel(self.model)
-
-        self.vdl_writer = None
-        if self.config['Global']['use_visualdl'] and mode == "train":
-            vdl_writer_path = os.path.join(self.output_dir, "vdl")
-            if not os.path.exists(vdl_writer_path):
-                os.makedirs(vdl_writer_path)
-            self.vdl_writer = LogWriter(logdir=vdl_writer_path)
-        logger.info('train with paddle {} and device {}'.format(
-            paddle.__version__, self.device))
-        # init members
-        self.train_dataloader = None
-        self.eval_dataloader = None
-        self.gallery_dataloader = None
-        self.query_dataloader = None
-        self.eval_mode = self.config["Global"].get("eval_mode",
-                                                   "classification")
-        self.amp = True if "AMP" in self.config else False
-        if self.amp and self.config["AMP"] is not None:
-            self.scale_loss = self.config["AMP"].get("scale_loss", 1.0)
-            self.use_dynamic_loss_scaling = self.config["AMP"].get(
-                "use_dynamic_loss_scaling", False)
-        else:
-            self.scale_loss = 1.0
-            self.use_dynamic_loss_scaling = False
-        if self.amp:
-            AMP_RELATED_FLAGS_SETTING = {
-                'FLAGS_cudnn_batchnorm_spatial_persistent': 1,
-                'FLAGS_max_inplace_grad_add': 8,
-            }
-            paddle.fluid.set_flags(AMP_RELATED_FLAGS_SETTING)
-        self.train_loss_func = None
-        self.eval_loss_func = None
-        self.train_metric_func = None
-        self.eval_metric_func = None
-        self.use_dali = self.config['Global'].get("use_dali", False)
-
-    def train(self):
-        # build train loss and metric info
-        if self.train_loss_func is None:
-            loss_info = self.config["Loss"]["Train"]
-            self.train_loss_func = build_loss(loss_info)
-        if self.train_metric_func is None:
-            metric_config = self.config.get("Metric")
-            if metric_config is not None:
-                metric_config = metric_config.get("Train")
-                if metric_config is not None:
-                    self.train_metric_func = build_metrics(metric_config)
-
-        if self.train_dataloader is None:
-            self.train_dataloader = build_dataloader(
-                self.config["DataLoader"], "Train", self.device, self.use_dali)
-
-        step_each_epoch = len(self.train_dataloader)
-
-        optimizer, lr_sch = build_optimizer(self.config["Optimizer"],
-                                            self.config["Global"]["epochs"],
-                                            step_each_epoch,
-                                            self.model.parameters())
-
-        print_batch_step = self.config['Global']['print_batch_step']
-        save_interval = self.config["Global"]["save_interval"]
-
-        best_metric = {
-            "metric": 0.0,
-            "epoch": 0,
-        }
-        # key:
-        # val: metrics list word
-        output_info = dict()
-        time_info = {
-            "batch_cost": AverageMeter(
-                "batch_cost", '.5f', postfix=" s,"),
-            "reader_cost": AverageMeter(
-                "reader_cost", ".5f", postfix=" s,"),
-        }
-        # global iter counter
-        global_step = 0
-
-        if self.config["Global"]["checkpoints"] is not None:
-            metric_info = init_model(self.config["Global"], self.model,
-                                     optimizer)
-            if metric_info is not None:
-                best_metric.update(metric_info)
-
-        # for amp training
-        if self.amp:
-            scaler = paddle.amp.GradScaler(
-                init_loss_scaling=self.scale_loss,
-                use_dynamic_loss_scaling=self.use_dynamic_loss_scaling)
-
-        tic = time.time()
-        max_iter = len(self.train_dataloader) - 1 if platform.system(
-        ) == "Windows" else len(self.train_dataloader)
-        for epoch_id in range(best_metric["epoch"] + 1,
-                              self.config["Global"]["epochs"] + 1):
-            acc = 0.0
-            train_dataloader = self.train_dataloader if self.use_dali else self.train_dataloader(
-            )
-            for iter_id, batch in enumerate(train_dataloader):
-                if iter_id >= max_iter:
-                    break
-                if iter_id == 5:
-                    for key in time_info:
-                        time_info[key].reset()
-                time_info["reader_cost"].update(time.time() - tic)
-                if self.use_dali:
-                    batch = [
-                        paddle.to_tensor(batch[0]['data']),
-                        paddle.to_tensor(batch[0]['label'])
-                    ]
-                batch_size = batch[0].shape[0]
-                batch[1] = batch[1].reshape([-1, 1]).astype("int64")
-
-                global_step += 1
-                # image input
-                if self.amp:
-                    with paddle.amp.auto_cast(custom_black_list={
-                            "flatten_contiguous_range", "greater_than"
-                    }):
-                        out = self.forward(batch)
-                        loss_dict = self.train_loss_func(out, batch[1])
-                else:
-                    out = self.forward(batch)
-
-                # calc loss
-                if self.config["DataLoader"]["Train"]["dataset"].get(
-                        "batch_transform_ops", None):
-                    loss_dict = self.train_loss_func(out, batch[1:])
-                else:
-                    loss_dict = self.train_loss_func(out, batch[1])
-
-                for key in loss_dict:
-                    if not key in output_info:
-                        output_info[key] = AverageMeter(key, '7.5f')
-                    output_info[key].update(loss_dict[key].numpy()[0],
-                                            batch_size)
-                # calc metric
-                if self.train_metric_func is not None:
-                    metric_dict = self.train_metric_func(out, batch[-1])
-                    for key in metric_dict:
-                        if not key in output_info:
-                            output_info[key] = AverageMeter(key, '7.5f')
-                        output_info[key].update(metric_dict[key].numpy()[0],
-                                                batch_size)
-
-                # step opt and lr
-                if self.amp:
-                    scaled = scaler.scale(loss_dict["loss"])
-                    scaled.backward()
-                    scaler.minimize(optimizer, scaled)
-                else:
-                    loss_dict["loss"].backward()
-                    optimizer.step()
-                optimizer.clear_grad()
-                lr_sch.step()
-
-                time_info["batch_cost"].update(time.time() - tic)
-
-                if iter_id % print_batch_step == 0:
-                    lr_msg = "lr: {:.5f}".format(lr_sch.get_lr())
-                    metric_msg = ", ".join([
-                        "{}: {:.5f}".format(key, output_info[key].avg)
-                        for key in output_info
-                    ])
-                    time_msg = "s, ".join([
-                        "{}: {:.5f}".format(key, time_info[key].avg)
-                        for key in time_info
-                    ])
-
-                    ips_msg = "ips: {:.5f} images/sec".format(
-                        batch_size / time_info["batch_cost"].avg)
-                    eta_sec = ((self.config["Global"]["epochs"] - epoch_id + 1
-                                ) * len(self.train_dataloader) - iter_id
-                               ) * time_info["batch_cost"].avg
-                    eta_msg = "eta: {:s}".format(
-                        str(datetime.timedelta(seconds=int(eta_sec))))
-                    logger.info(
-                        "[Train][Epoch {}/{}][Iter: {}/{}]{}, {}, {}, {}, {}".
-                        format(epoch_id, self.config["Global"][
-                            "epochs"], iter_id,
-                               len(self.train_dataloader), lr_msg, metric_msg,
-                               time_msg, ips_msg, eta_msg))
-
-                    logger.scaler(
-                        name="lr",
-                        value=lr_sch.get_lr(),
-                        step=global_step,
-                        writer=self.vdl_writer)
-                    for key in output_info:
-                        logger.scaler(
-                            name="train_{}".format(key),
-                            value=output_info[key].avg,
-                            step=global_step,
-                            writer=self.vdl_writer)
-                tic = time.time()
-            if self.use_dali:
-                self.train_dataloader.reset()
-            metric_msg = ", ".join([
-                "{}: {:.5f}".format(key, output_info[key].avg)
-                for key in output_info
-            ])
-            logger.info("[Train][Epoch {}/{}][Avg]{}".format(
-                epoch_id, self.config["Global"]["epochs"], metric_msg))
-            output_info.clear()
-
-            # eval model and save model if possible
-            if self.config["Global"][
-                    "eval_during_train"] and epoch_id % self.config["Global"][
-                        "eval_interval"] == 0:
-                acc = self.eval(epoch_id)
-                if acc > best_metric["metric"]:
-                    best_metric["metric"] = acc
-                    best_metric["epoch"] = epoch_id
-                    save_load.save_model(
-                        self.model,
-                        optimizer,
-                        best_metric,
-                        self.output_dir,
-                        model_name=self.config["Arch"]["name"],
-                        prefix="best_model")
-                logger.info("[Eval][Epoch {}][best metric: {}]".format(
-                    epoch_id, best_metric["metric"]))
-                logger.scaler(
-                    name="eval_acc",
-                    value=acc,
-                    step=epoch_id,
-                    writer=self.vdl_writer)
-
-                self.model.train()
-
-            # save model
-            if epoch_id % save_interval == 0:
-                save_load.save_model(
-                    self.model,
-                    optimizer, {"metric": acc,
-                                "epoch": epoch_id},
-                    self.output_dir,
-                    model_name=self.config["Arch"]["name"],
-                    prefix="epoch_{}".format(epoch_id))
-                # save the latest model
-                save_load.save_model(
-                    self.model,
-                    optimizer, {"metric": acc,
-                                "epoch": epoch_id},
-                    self.output_dir,
-                    model_name=self.config["Arch"]["name"],
-                    prefix="latest")
-
-        if self.vdl_writer is not None:
-            self.vdl_writer.close()
-
-    def build_avg_metrics(self, info_dict):
-        return {key: AverageMeter(key, '7.5f') for key in info_dict}
-
-    @paddle.no_grad()
-    def eval(self, epoch_id=0):
-        self.model.eval()
-        if self.eval_loss_func is None:
-            loss_config = self.config.get("Loss", None)
-            if loss_config is not None:
-                loss_config = loss_config.get("Eval")
-                if loss_config is not None:
-                    self.eval_loss_func = build_loss(loss_config)
-        if self.eval_mode == "classification":
-            if self.eval_dataloader is None:
-                self.eval_dataloader = build_dataloader(
-                    self.config["DataLoader"], "Eval", self.device,
-                    self.use_dali)
-
-            if self.eval_metric_func is None:
-                metric_config = self.config.get("Metric")
-                if metric_config is not None:
-                    metric_config = metric_config.get("Eval")
-                    if metric_config is not None:
-                        self.eval_metric_func = build_metrics(metric_config)
-
-            eval_result = self.eval_cls(epoch_id)
-
-        elif self.eval_mode == "retrieval":
-            if self.gallery_dataloader is None:
-                self.gallery_dataloader = build_dataloader(
-                    self.config["DataLoader"]["Eval"], "Gallery", self.device,
-                    self.use_dali)
-
-            if self.query_dataloader is None:
-                self.query_dataloader = build_dataloader(
-                    self.config["DataLoader"]["Eval"], "Query", self.device,
-                    self.use_dali)
-            # build metric info
-            if self.eval_metric_func is None:
-                metric_config = self.config.get("Metric", None)
-                if metric_config is None:
-                    metric_config = [{"name": "Recallk", "topk": (1, 5)}]
-                else:
-                    metric_config = metric_config["Eval"]
-                self.eval_metric_func = build_metrics(metric_config)
-            eval_result = self.eval_retrieval(epoch_id)
-        else:
-            logger.warning("Invalid eval mode: {}".format(self.eval_mode))
-            eval_result = None
-        self.model.train()
-        return eval_result
-
-    def forward(self, batch):
-        if not self.is_rec:
-            out = self.model(batch[0])
-        else:
-            out = self.model(batch[0], batch[1])
-        return out
-
-    @paddle.no_grad()
-    def eval_cls(self, epoch_id=0):
-        output_info = dict()
-        time_info = {
-            "batch_cost": AverageMeter(
-                "batch_cost", '.5f', postfix=" s,"),
-            "reader_cost": AverageMeter(
-                "reader_cost", ".5f", postfix=" s,"),
-        }
-        print_batch_step = self.config["Global"]["print_batch_step"]
-
-        metric_key = None
-        tic = time.time()
-        eval_dataloader = self.eval_dataloader if self.use_dali else self.eval_dataloader(
-        )
-        max_iter = len(self.eval_dataloader) - 1 if platform.system(
-        ) == "Windows" else len(self.eval_dataloader)
-        for iter_id, batch in enumerate(eval_dataloader):
-            if iter_id >= max_iter:
-                break
-            if iter_id == 5:
-                for key in time_info:
-                    time_info[key].reset()
-            if self.use_dali:
-                batch = [
-                    paddle.to_tensor(batch[0]['data']),
-                    paddle.to_tensor(batch[0]['label'])
-                ]
-            time_info["reader_cost"].update(time.time() - tic)
-            batch_size = batch[0].shape[0]
-            batch[0] = paddle.to_tensor(batch[0]).astype("float32")
-            batch[1] = batch[1].reshape([-1, 1]).astype("int64")
-            # image input
-            out = self.forward(batch)
-            # calc loss
-            if self.eval_loss_func is not None:
-                loss_dict = self.eval_loss_func(out, batch[-1])
-                for key in loss_dict:
-                    if not key in output_info:
-                        output_info[key] = AverageMeter(key, '7.5f')
-                    output_info[key].update(loss_dict[key].numpy()[0],
-                                            batch_size)
-            # calc metric
-            if self.eval_metric_func is not None:
-                metric_dict = self.eval_metric_func(out, batch[-1])
-                if paddle.distributed.get_world_size() > 1:
-                    for key in metric_dict:
-                        paddle.distributed.all_reduce(
-                            metric_dict[key],
-                            op=paddle.distributed.ReduceOp.SUM)
-                        metric_dict[key] = metric_dict[
-                            key] / paddle.distributed.get_world_size()
-                for key in metric_dict:
-                    if metric_key is None:
-                        metric_key = key
-                    if not key in output_info:
-                        output_info[key] = AverageMeter(key, '7.5f')
-
-                    output_info[key].update(metric_dict[key].numpy()[0],
-                                            batch_size)
-
-            time_info["batch_cost"].update(time.time() - tic)
-
-            if iter_id % print_batch_step == 0:
-                time_msg = "s, ".join([
-                    "{}: {:.5f}".format(key, time_info[key].avg)
-                    for key in time_info
-                ])
-
-                ips_msg = "ips: {:.5f} images/sec".format(
-                    batch_size / time_info["batch_cost"].avg)
-
-                metric_msg = ", ".join([
-                    "{}: {:.5f}".format(key, output_info[key].val)
-                    for key in output_info
-                ])
-                logger.info("[Eval][Epoch {}][Iter: {}/{}]{}, {}, {}".format(
-                    epoch_id, iter_id,
-                    len(self.eval_dataloader), metric_msg, time_msg, ips_msg))
-
-            tic = time.time()
-        if self.use_dali:
-            self.eval_dataloader.reset()
-        metric_msg = ", ".join([
-            "{}: {:.5f}".format(key, output_info[key].avg)
-            for key in output_info
-        ])
-        logger.info("[Eval][Epoch {}][Avg]{}".format(epoch_id, metric_msg))
-
-        # do not try to save best model
-        if self.eval_metric_func is None:
-            return -1
-        # return 1st metric in the dict
-        return output_info[metric_key].avg
-
-    def eval_retrieval(self, epoch_id=0):
-        self.model.eval()
-        # step1. build gallery
-        gallery_feas, gallery_img_id, gallery_unique_id = self._cal_feature(
-            name='gallery')
-        query_feas, query_img_id, query_query_id = self._cal_feature(
-            name='query')
-
-        # step2. do evaluation
-        sim_block_size = self.config["Global"].get("sim_block_size", 64)
-        sections = [sim_block_size] * (len(query_feas) // sim_block_size)
-        if len(query_feas) % sim_block_size:
-            sections.append(len(query_feas) % sim_block_size)
-        fea_blocks = paddle.split(query_feas, num_or_sections=sections)
-        if query_query_id is not None:
-            query_id_blocks = paddle.split(
-                query_query_id, num_or_sections=sections)
-        image_id_blocks = paddle.split(query_img_id, num_or_sections=sections)
-        metric_key = None
-
-        if self.eval_metric_func is None:
-            metric_dict = {metric_key: 0.}
-        else:
-            metric_dict = dict()
-            for block_idx, block_fea in enumerate(fea_blocks):
-                similarity_matrix = paddle.matmul(
-                    block_fea, gallery_feas, transpose_y=True)
-                if query_query_id is not None:
-                    query_id_block = query_id_blocks[block_idx]
-                    query_id_mask = (query_id_block != gallery_unique_id.t())
-
-                    image_id_block = image_id_blocks[block_idx]
-                    image_id_mask = (image_id_block != gallery_img_id.t())
-
-                    keep_mask = paddle.logical_or(query_id_mask, image_id_mask)
-                    similarity_matrix = similarity_matrix * keep_mask.astype(
-                        "float32")
-                else:
-                    keep_mask = None
-
-                metric_tmp = self.eval_metric_func(similarity_matrix,
-                                                   image_id_blocks[block_idx],
-                                                   gallery_img_id, keep_mask)
-
-                for key in metric_tmp:
-                    if key not in metric_dict:
-                        metric_dict[key] = metric_tmp[key] * block_fea.shape[
-                            0] / len(query_feas)
-                    else:
-                        metric_dict[key] += metric_tmp[key] * block_fea.shape[
-                            0] / len(query_feas)
-
-        metric_info_list = []
-        for key in metric_dict:
-            if metric_key is None:
-                metric_key = key
-            metric_info_list.append("{}: {:.5f}".format(key, metric_dict[key]))
-        metric_msg = ", ".join(metric_info_list)
-        logger.info("[Eval][Epoch {}][Avg]{}".format(epoch_id, metric_msg))
-
-        return metric_dict[metric_key]
-
-    def _cal_feature(self, name='gallery'):
-        all_feas = None
-        all_image_id = None
-        all_unique_id = None
-        if name == 'gallery':
-            dataloader = self.gallery_dataloader
-        elif name == 'query':
-            dataloader = self.query_dataloader
-        else:
-            raise RuntimeError("Only support gallery or query dataset")
-
-        has_unique_id = False
-        max_iter = len(dataloader) - 1 if platform.system(
-        ) == "Windows" else len(dataloader)
-        dataloader_tmp = dataloader if self.use_dali else dataloader()
-        for idx, batch in enumerate(
-                dataloader_tmp):  # load is very time-consuming
-            if idx >= max_iter:
-                break
-            if idx % self.config["Global"]["print_batch_step"] == 0:
-                logger.info(
-                    f"{name} feature calculation process: [{idx}/{len(dataloader)}]"
-                )
-            if self.use_dali:
-                batch = [
-                    paddle.to_tensor(batch[0]['data']),
-                    paddle.to_tensor(batch[0]['label'])
-                ]
-            batch = [paddle.to_tensor(x) for x in batch]
-            batch[1] = batch[1].reshape([-1, 1]).astype("int64")
-            if len(batch) == 3:
-                has_unique_id = True
-                batch[2] = batch[2].reshape([-1, 1]).astype("int64")
-            out = self.forward(batch)
-            batch_feas = out["features"]
-
-            # do norm
-            if self.config["Global"].get("feature_normalize", True):
-                feas_norm = paddle.sqrt(
-                    paddle.sum(paddle.square(batch_feas), axis=1,
-                               keepdim=True))
-                batch_feas = paddle.divide(batch_feas, feas_norm)
-
-            if all_feas is None:
-                all_feas = batch_feas
-                if has_unique_id:
-                    all_unique_id = batch[2]
-                all_image_id = batch[1]
-            else:
-                all_feas = paddle.concat([all_feas, batch_feas])
-                all_image_id = paddle.concat([all_image_id, batch[1]])
-                if has_unique_id:
-                    all_unique_id = paddle.concat([all_unique_id, batch[2]])
-        if self.use_dali:
-            dataloader_tmp.reset()
-        if paddle.distributed.get_world_size() > 1:
-            feat_list = []
-            img_id_list = []
-            unique_id_list = []
-            paddle.distributed.all_gather(feat_list, all_feas)
-            paddle.distributed.all_gather(img_id_list, all_image_id)
-            all_feas = paddle.concat(feat_list, axis=0)
-            all_image_id = paddle.concat(img_id_list, axis=0)
-            if has_unique_id:
-                paddle.distributed.all_gather(unique_id_list, all_unique_id)
-                all_unique_id = paddle.concat(unique_id_list, axis=0)
-
-        logger.info("Build {} done, all feat shape: {}, begin to eval..".
-                    format(name, all_feas.shape))
-        return all_feas, all_image_id, all_unique_id
-
-    @paddle.no_grad()
-    def infer(self, ):
-        total_trainer = paddle.distributed.get_world_size()
-        local_rank = paddle.distributed.get_rank()
-        image_list = get_image_list(self.config["Infer"]["infer_imgs"])
-        # data split
-        image_list = image_list[local_rank::total_trainer]
-
-        preprocess_func = create_operators(self.config["Infer"]["transforms"])
-        postprocess_func = build_postprocess(self.config["Infer"][
-            "PostProcess"])
-
-        batch_size = self.config["Infer"]["batch_size"]
-
-        self.model.eval()
-
-        batch_data = []
-        image_file_list = []
-        for idx, image_file in enumerate(image_list):
-            with open(image_file, 'rb') as f:
-                x = f.read()
-            for process in preprocess_func:
-                x = process(x)
-            batch_data.append(x)
-            image_file_list.append(image_file)
-            if len(batch_data) >= batch_size or idx == len(image_list) - 1:
-                batch_tensor = paddle.to_tensor(batch_data)
-                out = self.forward([batch_tensor])
-                if isinstance(out, list):
-                    out = out[0]
-                result = postprocess_func(out, image_file_list)
-                print(result)
-                batch_data.clear()
-                image_file_list.clear()
diff --git a/tools/eval.py b/tools/eval.py
index b03030c5a..367f5f8b4 100644
--- a/tools/eval.py
+++ b/tools/eval.py
@@ -21,11 +21,11 @@ __dir__ = os.path.dirname(os.path.abspath(__file__))
 sys.path.append(os.path.abspath(os.path.join(__dir__, '../')))
 
 from ppcls.utils import config
-from ppcls.engine.trainer import Trainer
+from ppcls.engine.core import Core
 
 if __name__ == "__main__":
     args = config.parse_args()
     config = config.get_config(
         args.config, overrides=args.override, show=False)
-    trainer = Trainer(config, mode="eval")
-    trainer.eval()
+    evaler = Core(config, mode="eval")
+    evaler.eval()
diff --git a/tools/export_model.py b/tools/export_model.py
index e019e2574..7d324a613 100644
--- a/tools/export_model.py
+++ b/tools/export_model.py
@@ -24,82 +24,11 @@ import paddle
 import paddle.nn as nn
 
 from ppcls.utils import config
-from ppcls.utils.logger import init_logger
-from ppcls.utils.config import print_config
-from ppcls.arch import build_model, RecModel, DistillationModel
-from ppcls.utils.save_load import load_dygraph_pretrain
-from ppcls.arch.gears.identity_head import IdentityHead
-
-
-class ExportModel(nn.Layer):
-    """
-    ExportModel: add softmax onto the model
-    """
-
-    def __init__(self, config):
-        super().__init__()
-        self.base_model = build_model(config)
-
-        # we should choose a final model to export
-        if isinstance(self.base_model, DistillationModel):
-            self.infer_model_name = config["infer_model_name"]
-        else:
-            self.infer_model_name = None
-
-        self.infer_output_key = config.get("infer_output_key", None)
-        if self.infer_output_key == "features" and isinstance(self.base_model,
-                                                              RecModel):
-            self.base_model.head = IdentityHead()
-        if config.get("infer_add_softmax", True):
-            self.softmax = nn.Softmax(axis=-1)
-        else:
-            self.softmax = None
-
-    def eval(self):
-        self.training = False
-        for layer in self.sublayers():
-            layer.training = False
-            layer.eval()
-
-    def forward(self, x):
-        x = self.base_model(x)
-        if isinstance(x, list):
-            x = x[0]
-        if self.infer_model_name is not None:
-            x = x[self.infer_model_name]
-        if self.infer_output_key is not None:
-            x = x[self.infer_output_key]
-        if self.softmax is not None:
-            x = self.softmax(x)
-        return x
-
+from ppcls.engine.core import Core
 
 if __name__ == "__main__":
     args = config.parse_args()
     config = config.get_config(
         args.config, overrides=args.override, show=False)
-    log_file = os.path.join(config['Global']['output_dir'],
-                            config["Arch"]["name"], "export.log")
-    init_logger(name='root', log_file=log_file)
-    print_config(config)
-
-    # set device
-    assert config["Global"]["device"] in ["cpu", "gpu", "xpu"]
-    device = paddle.set_device(config["Global"]["device"])
-    model = ExportModel(config["Arch"])
-    if config["Global"]["pretrained_model"] is not None:
-        load_dygraph_pretrain(model.base_model,
-                              config["Global"]["pretrained_model"])
-
-    model.eval()
-
-    model = paddle.jit.to_static(
-        model,
-        input_spec=[
-            paddle.static.InputSpec(
-                shape=[None] + config["Global"]["image_shape"],
-                dtype='float32')
-        ])
-    paddle.jit.save(model,
-                    os.path.join(config["Global"]["save_inference_dir"],
-                                 "inference"))
+    exporter = Core(config, mode="export")
+    exporter.export()
diff --git a/tools/infer.py b/tools/infer.py
index da23a3d88..9c00542c1 100644
--- a/tools/infer.py
+++ b/tools/infer.py
@@ -21,12 +21,11 @@ __dir__ = os.path.dirname(os.path.abspath(__file__))
 sys.path.append(os.path.abspath(os.path.join(__dir__, '../')))
 
 from ppcls.utils import config
-from ppcls.engine.trainer import Trainer
+from ppcls.engine.core import Core
 
 if __name__ == "__main__":
     args = config.parse_args()
     config = config.get_config(
         args.config, overrides=args.override, show=False)
-    trainer = Trainer(config, mode="infer")
-
-    trainer.infer()
+    inferer = Core(config, mode="infer")
+    inferer.infer()
diff --git a/tools/train.py b/tools/train.py
index 169678c5a..23a064e7b 100644
--- a/tools/train.py
+++ b/tools/train.py
@@ -21,11 +21,11 @@ __dir__ = os.path.dirname(os.path.abspath(__file__))
 sys.path.append(os.path.abspath(os.path.join(__dir__, '../')))
 
 from ppcls.utils import config
-from ppcls.engine.trainer import Trainer
+from ppcls.engine.core import Core
 
 if __name__ == "__main__":
     args = config.parse_args()
     config = config.get_config(
         args.config, overrides=args.override, show=False)
-    trainer = Trainer(config, mode="train")
+    trainer = Core(config, mode="train")
     trainer.train()

From 23b2332e1f40ebdfb39b0643ac89bd09b484a094 Mon Sep 17 00:00:00 2001
From: Bin Lu <lubing02@gmail.com>
Date: Mon, 23 Aug 2021 18:47:03 +0800
Subject: [PATCH 07/35] Create VGG16_CIFAR10_DeepHash.yaml

---
 .../professional/VGG16_CIFAR10_DeepHash.yaml  | 147 ++++++++++++++++++
 1 file changed, 147 insertions(+)
 create mode 100644 ppcls/configs/quick_start/professional/VGG16_CIFAR10_DeepHash.yaml

diff --git a/ppcls/configs/quick_start/professional/VGG16_CIFAR10_DeepHash.yaml b/ppcls/configs/quick_start/professional/VGG16_CIFAR10_DeepHash.yaml
new file mode 100644
index 000000000..01edcdd27
--- /dev/null
+++ b/ppcls/configs/quick_start/professional/VGG16_CIFAR10_DeepHash.yaml
@@ -0,0 +1,147 @@
+# global configs
+Global:
+  checkpoints: null
+  pretrained_model: null
+  output_dir: ./output
+  device: gpu
+  save_interval: 1
+  eval_during_train: True
+  eval_interval: 1
+  eval_mode: "retrieval"
+  epochs: 128
+  print_batch_step: 10
+  use_visualdl: False
+
+  # used for static mode and model export
+  image_shape: [3, 227, 227]
+  save_inference_dir: ./inference
+
+  #feature postprocess
+  feature_normalize: False
+  feature_binarize: "round"
+
+# model architecture
+Arch:
+  name: "RecModel"
+  Backbone:
+    name: "VGG16"
+    pretrained: True
+    class_num: 48
+  Head:
+    name: "FC"
+    class_num: 10
+    embedding_size: 48
+    
+  infer_output_key:  "features"
+  infer_add_softmax: "false"
+
+# loss function config for train/eval process
+Loss:
+  Train:
+    - CELoss:
+        weight: 1.0
+        epsilon: 0.1
+  Eval:
+    - CELoss:
+        weight: 1.0
+
+Optimizer:
+  name: Momentum
+  momentum: 0.9
+  lr:
+    name: Piecewise
+    learning_rate: 0.01
+    decay_epochs: [200]
+    values: [0.01, 0.001]
+
+# data loader for train and eval
+DataLoader:
+  Train:
+    dataset:
+      name: ImageNetDataset
+      image_root: ./dataset/cifar10/
+      cls_label_path: ./dataset/cifar10/cifar10-2/train.txt
+      transform_ops:
+        - DecodeImage:
+            to_rgb: True
+            channel_first: False
+        - ResizeImage:
+            size: 256
+        - RandCropImage:
+            size: 227
+        - RandFlipImage:
+            flip_code: 1
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.4914, 0.4822, 0.4465]
+            std: [0.2023, 0.1994, 0.2010]
+            order: ''
+    sampler:
+      name: DistributedBatchSampler
+      batch_size: 128
+      drop_last: False
+      shuffle: True
+    loader:
+      num_workers: 4
+      use_shared_memory: True
+
+  Eval:
+    Query:
+      dataset: 
+        name: ImageNetDataset
+        image_root: ./dataset/cifar10/
+        cls_label_path: ./dataset/cifar10/cifar10-2/test.txt
+        transform_ops:
+          - DecodeImage:
+              to_rgb: True
+              channel_first: False
+          - ResizeImage:
+              size: 227
+          - NormalizeImage:
+              scale: 1.0/255.0
+              mean: [0.4914, 0.4822, 0.4465]
+              std: [0.2023, 0.1994, 0.2010]
+              order: ''
+      sampler:
+        name: DistributedBatchSampler
+        batch_size: 512
+        drop_last: False
+        shuffle: False
+      loader:
+        num_workers: 4
+        use_shared_memory: True
+
+    Gallery:
+      dataset: 
+        name: ImageNetDataset
+        image_root: ./dataset/cifar10/
+        cls_label_path: ./dataset/cifar10/cifar10-2/database.txt
+        transform_ops:
+          - DecodeImage:
+              to_rgb: True
+              channel_first: False
+          - ResizeImage:
+              size: 227
+          - NormalizeImage:
+              scale: 1.0/255.0
+              mean: [0.4914, 0.4822, 0.4465]
+              std: [0.2023, 0.1994, 0.2010]
+              order: ''
+      sampler:
+        name: DistributedBatchSampler
+        batch_size: 512
+        drop_last: False
+        shuffle: False
+      loader:
+        num_workers: 4
+        use_shared_memory: True
+
+Metric:
+  Train:
+    - TopkAcc:
+        topk: [1, 5]
+  Eval:
+    - mAP:
+    - Precisionk:
+        topk: [1, 5]
+        

From 865870f5f9fa6b3f71975a297149c576f3d47681 Mon Sep 17 00:00:00 2001
From: Bin Lu <lubing02@gmail.com>
Date: Mon, 23 Aug 2021 18:53:50 +0800
Subject: [PATCH 08/35] Update VGG16_CIFAR10_DeepHash.yaml

---
 .../quick_start/professional/VGG16_CIFAR10_DeepHash.yaml    | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/ppcls/configs/quick_start/professional/VGG16_CIFAR10_DeepHash.yaml b/ppcls/configs/quick_start/professional/VGG16_CIFAR10_DeepHash.yaml
index 01edcdd27..dd27df808 100644
--- a/ppcls/configs/quick_start/professional/VGG16_CIFAR10_DeepHash.yaml
+++ b/ppcls/configs/quick_start/professional/VGG16_CIFAR10_DeepHash.yaml
@@ -13,7 +13,7 @@ Global:
   use_visualdl: False
 
   # used for static mode and model export
-  image_shape: [3, 227, 227]
+  image_shape: [3, 224, 224]
   save_inference_dir: ./inference
 
   #feature postprocess
@@ -68,7 +68,7 @@ DataLoader:
         - ResizeImage:
             size: 256
         - RandCropImage:
-            size: 227
+            size: 224
         - RandFlipImage:
             flip_code: 1
         - NormalizeImage:
@@ -96,7 +96,7 @@ DataLoader:
               to_rgb: True
               channel_first: False
           - ResizeImage:
-              size: 227
+              size: 224
           - NormalizeImage:
               scale: 1.0/255.0
               mean: [0.4914, 0.4822, 0.4465]

From d963dc17daedc3da9e0bfaae17d5f96461b211d3 Mon Sep 17 00:00:00 2001
From: Bin Lu <lubing02@gmail.com>
Date: Mon, 23 Aug 2021 19:06:49 +0800
Subject: [PATCH 09/35] Update VGG16_CIFAR10_DeepHash.yaml

---
 .../quick_start/professional/VGG16_CIFAR10_DeepHash.yaml        | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/ppcls/configs/quick_start/professional/VGG16_CIFAR10_DeepHash.yaml b/ppcls/configs/quick_start/professional/VGG16_CIFAR10_DeepHash.yaml
index dd27df808..6ab045f4f 100644
--- a/ppcls/configs/quick_start/professional/VGG16_CIFAR10_DeepHash.yaml
+++ b/ppcls/configs/quick_start/professional/VGG16_CIFAR10_DeepHash.yaml
@@ -24,7 +24,7 @@ Global:
 Arch:
   name: "RecModel"
   Backbone:
-    name: "VGG16"
+    name: "VGG19"
     pretrained: True
     class_num: 48
   Head:

From 0b9f0ad0014f7912c21f291a07c7d9e0b5ade01c Mon Sep 17 00:00:00 2001
From: Bin Lu <lubing02@gmail.com>
Date: Mon, 23 Aug 2021 19:07:35 +0800
Subject: [PATCH 10/35] Create vgg_variant.py

---
 .../backbone/variant_models/vgg_variant.py    | 28 +++++++++++++++++++
 1 file changed, 28 insertions(+)
 create mode 100644 ppcls/arch/backbone/variant_models/vgg_variant.py

diff --git a/ppcls/arch/backbone/variant_models/vgg_variant.py b/ppcls/arch/backbone/variant_models/vgg_variant.py
new file mode 100644
index 000000000..ac6c0138e
--- /dev/null
+++ b/ppcls/arch/backbone/variant_models/vgg_variant.py
@@ -0,0 +1,28 @@
+import paddle
+from paddle.nn import Sigmoid
+from ppcls.arch.backbone.legendary_models.vgg import VGG19
+ 
+__all__ = ["VGG19Sigmoid"]
+ 
+ 
+class SigmoidSuffix(paddle.nn.Layer):
+    def __init__(self, origin_layer):
+        super(SigmoidSuffix, self).__init__()
+        self.origin_layer = origin_layer
+        self.sigmoid = Sigmoid()
+ 
+    def forward(self, *input, res_dict=None, **kwargs):
+        x = self.origin_layer(input)
+        x = self.sigmoid(x)
+        return x
+ 
+ 
+def VGG19Sigmoid(pretrained=False, use_ssld=False, **kwargs):
+    def replace_function(origin_layer):
+        new_layer = SigmoidSuffix(origin_layer)
+        return new_layer
+ 
+    match_re = "linear_2"
+    model = VGG19(pretrained=pretrained, use_ssld=use_ssld, **kwargs)
+    model.replace_sub(match_re, replace_function, True)
+    return model

From 26eda0efef2a7db993d7abcd7b2a96ed328dff08 Mon Sep 17 00:00:00 2001
From: Bin Lu <lubing02@gmail.com>
Date: Mon, 23 Aug 2021 19:08:06 +0800
Subject: [PATCH 11/35] Update __init__.py

---
 ppcls/arch/backbone/variant_models/__init__.py | 1 +
 1 file changed, 1 insertion(+)

diff --git a/ppcls/arch/backbone/variant_models/__init__.py b/ppcls/arch/backbone/variant_models/__init__.py
index e87a0e098..34fcdeb2c 100644
--- a/ppcls/arch/backbone/variant_models/__init__.py
+++ b/ppcls/arch/backbone/variant_models/__init__.py
@@ -1 +1,2 @@
 from .resnet_variant import ResNet50_last_stage_stride1
+from .vgg_variant import VGG19Sigmoid

From 581af23ac6a5000e1080c6387c2def96424df837 Mon Sep 17 00:00:00 2001
From: Bin Lu <lubing02@gmail.com>
Date: Mon, 23 Aug 2021 19:08:53 +0800
Subject: [PATCH 12/35] Update VGG16_CIFAR10_DeepHash.yaml

---
 .../quick_start/professional/VGG16_CIFAR10_DeepHash.yaml        | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/ppcls/configs/quick_start/professional/VGG16_CIFAR10_DeepHash.yaml b/ppcls/configs/quick_start/professional/VGG16_CIFAR10_DeepHash.yaml
index 6ab045f4f..d6722d5df 100644
--- a/ppcls/configs/quick_start/professional/VGG16_CIFAR10_DeepHash.yaml
+++ b/ppcls/configs/quick_start/professional/VGG16_CIFAR10_DeepHash.yaml
@@ -24,7 +24,7 @@ Global:
 Arch:
   name: "RecModel"
   Backbone:
-    name: "VGG19"
+    name: "VGG19Sigmoid"
     pretrained: True
     class_num: 48
   Head:

From f878fb27e5a1a3f312bf7820dfa783e7f384128e Mon Sep 17 00:00:00 2001
From: Bin Lu <lubing02@gmail.com>
Date: Mon, 23 Aug 2021 19:19:19 +0800
Subject: [PATCH 13/35] Update __init__.py

---
 ppcls/arch/backbone/__init__.py | 1 +
 1 file changed, 1 insertion(+)

diff --git a/ppcls/arch/backbone/__init__.py b/ppcls/arch/backbone/__init__.py
index 1923b0993..a1727613a 100644
--- a/ppcls/arch/backbone/__init__.py
+++ b/ppcls/arch/backbone/__init__.py
@@ -58,6 +58,7 @@ from ppcls.arch.backbone.model_zoo.rednet import RedNet26, RedNet38, RedNet50, R
 from ppcls.arch.backbone.model_zoo.tnt import TNT_small
 from ppcls.arch.backbone.model_zoo.hardnet import HarDNet68, HarDNet85, HarDNet39_ds, HarDNet68_ds
 from ppcls.arch.backbone.variant_models.resnet_variant import ResNet50_last_stage_stride1
+from ppcls.arch.backbone.variant_models.vgg_variant import VGG19Sigmoid
 
 
 def get_apis():

From 2f85378bb61c30ceef8595f66e8380cf4659db8a Mon Sep 17 00:00:00 2001
From: Bin Lu <lubing02@gmail.com>
Date: Mon, 23 Aug 2021 19:23:47 +0800
Subject: [PATCH 14/35] Rename VGG16_CIFAR10_DeepHash.yaml to
 VGG19_CIFAR10_DeepHash.yaml

---
 .../{VGG16_CIFAR10_DeepHash.yaml => VGG19_CIFAR10_DeepHash.yaml}  | 0
 1 file changed, 0 insertions(+), 0 deletions(-)
 rename ppcls/configs/quick_start/professional/{VGG16_CIFAR10_DeepHash.yaml => VGG19_CIFAR10_DeepHash.yaml} (100%)

diff --git a/ppcls/configs/quick_start/professional/VGG16_CIFAR10_DeepHash.yaml b/ppcls/configs/quick_start/professional/VGG19_CIFAR10_DeepHash.yaml
similarity index 100%
rename from ppcls/configs/quick_start/professional/VGG16_CIFAR10_DeepHash.yaml
rename to ppcls/configs/quick_start/professional/VGG19_CIFAR10_DeepHash.yaml

From 0354f4457412db633138773dea82acc6df5c072b Mon Sep 17 00:00:00 2001
From: Bin Lu <lubing02@gmail.com>
Date: Mon, 23 Aug 2021 19:29:39 +0800
Subject: [PATCH 15/35] Update metrics.py

---
 ppcls/metric/metrics.py | 41 +++++++++++++++++++++++++++++++++++++++++
 1 file changed, 41 insertions(+)

diff --git a/ppcls/metric/metrics.py b/ppcls/metric/metrics.py
index 9908b3541..204d2af09 100644
--- a/ppcls/metric/metrics.py
+++ b/ppcls/metric/metrics.py
@@ -168,6 +168,47 @@ class Recallk(nn.Layer):
         return metric_dict
 
 
+class Precisionk(nn.Layer):
+    def __init__(self, topk=(1, 5)):
+        super().__init__()
+        assert isinstance(topk, (int, list, tuple))
+        if isinstance(topk, int):
+            topk = [topk]
+        self.topk = topk
+
+    def forward(self, similarities_matrix, query_img_id, gallery_img_id,
+                keep_mask):
+        metric_dict = dict()
+
+        #get cmc
+        choosen_indices = paddle.argsort(
+            similarities_matrix, axis=1, descending=True)
+        gallery_labels_transpose = paddle.transpose(gallery_img_id, [1, 0])
+        gallery_labels_transpose = paddle.broadcast_to(
+            gallery_labels_transpose,
+            shape=[
+                choosen_indices.shape[0], gallery_labels_transpose.shape[1]
+            ])
+        choosen_label = paddle.index_sample(gallery_labels_transpose,
+                                            choosen_indices)
+        equal_flag = paddle.equal(choosen_label, query_img_id)
+        if keep_mask is not None:
+            keep_mask = paddle.index_sample(
+                keep_mask.astype('float32'), choosen_indices)
+            equal_flag = paddle.logical_and(equal_flag,
+                                            keep_mask.astype('bool'))
+        equal_flag = paddle.cast(equal_flag, 'float32')
+        
+        Ns = paddle.arange(gallery_img_id.shape[0]) + 1
+        equal_flag_cumsum = paddle.cumsum(equal_flag, axis=1)
+        Precision_at_k = (paddle.mean(equal_flag_cumsum, axis=0) / Ns).numpy()
+
+        for k in self.topk:
+            metric_dict["precision@{}".format(k)] = Precision_at_k[k - 1]
+
+        return metric_dict
+
+
 class DistillationTopkAcc(TopkAcc):
     def __init__(self, model_key, feature_key=None, topk=(1, 5)):
         super().__init__(topk=topk)

From a804bb02bd6d55b14be05c65c33843e2bded82b2 Mon Sep 17 00:00:00 2001
From: Bin Lu <lubing02@gmail.com>
Date: Mon, 23 Aug 2021 19:32:35 +0800
Subject: [PATCH 16/35] Update VGG19_CIFAR10_DeepHash.yaml

---
 .../quick_start/professional/VGG19_CIFAR10_DeepHash.yaml        | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/ppcls/configs/quick_start/professional/VGG19_CIFAR10_DeepHash.yaml b/ppcls/configs/quick_start/professional/VGG19_CIFAR10_DeepHash.yaml
index d6722d5df..97228828a 100644
--- a/ppcls/configs/quick_start/professional/VGG19_CIFAR10_DeepHash.yaml
+++ b/ppcls/configs/quick_start/professional/VGG19_CIFAR10_DeepHash.yaml
@@ -121,7 +121,7 @@ DataLoader:
               to_rgb: True
               channel_first: False
           - ResizeImage:
-              size: 227
+              size: 224
           - NormalizeImage:
               scale: 1.0/255.0
               mean: [0.4914, 0.4822, 0.4465]

From 34d9c8593e9865d530955ca46a568419a6c52045 Mon Sep 17 00:00:00 2001
From: Bin Lu <lubing02@gmail.com>
Date: Mon, 23 Aug 2021 19:45:44 +0800
Subject: [PATCH 17/35] Update __init__.py

---
 ppcls/metric/__init__.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/ppcls/metric/__init__.py b/ppcls/metric/__init__.py
index db80bab61..4c817a115 100644
--- a/ppcls/metric/__init__.py
+++ b/ppcls/metric/__init__.py
@@ -16,7 +16,7 @@ from paddle import nn
 import copy
 from collections import OrderedDict
 
-from .metrics import TopkAcc, mAP, mINP, Recallk
+from .metrics import TopkAcc, mAP, mINP, Recallk, Precisionk
 from .metrics import DistillationTopkAcc
 from .metrics import GoogLeNetTopkAcc
 

From c9912f0a04e7706fcdacfdef43ee83454b58f1f7 Mon Sep 17 00:00:00 2001
From: Bin Lu <lubing02@gmail.com>
Date: Tue, 24 Aug 2021 10:53:13 +0800
Subject: [PATCH 18/35] Update vgg_variant.py

---
 ppcls/arch/backbone/variant_models/vgg_variant.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/ppcls/arch/backbone/variant_models/vgg_variant.py b/ppcls/arch/backbone/variant_models/vgg_variant.py
index ac6c0138e..b73ad3592 100644
--- a/ppcls/arch/backbone/variant_models/vgg_variant.py
+++ b/ppcls/arch/backbone/variant_models/vgg_variant.py
@@ -11,7 +11,7 @@ class SigmoidSuffix(paddle.nn.Layer):
         self.origin_layer = origin_layer
         self.sigmoid = Sigmoid()
  
-    def forward(self, *input, res_dict=None, **kwargs):
+    def forward(self, input, res_dict=None, **kwargs):
         x = self.origin_layer(input)
         x = self.sigmoid(x)
         return x

From 15f6f5813914b02644aab39f791dd9cc1940994c Mon Sep 17 00:00:00 2001
From: dongshuilong <dongshuilong@baidu.com>
Date: Tue, 24 Aug 2021 03:02:55 +0000
Subject: [PATCH 19/35] refactor trainer v2

---
 ppcls/engine/{core.py => engine.py}           | 25 +++---
 ppcls/engine/{eval => evaluation}/__init__.py |  4 +-
 .../{eval => evaluation}/classification.py    |  0
 .../engine/{eval => evaluation}/retrieval.py  |  0
 ppcls/engine/train/__init__.py                |  3 +-
 ppcls/engine/train/classification.py          | 89 -------------------
 ppcls/engine/train/{retrieval.py => train.py} | 13 ++-
 tools/eval.py                                 |  6 +-
 tools/export_model.py                         |  6 +-
 tools/infer.py                                |  6 +-
 tools/train.py                                |  6 +-
 11 files changed, 34 insertions(+), 124 deletions(-)
 rename ppcls/engine/{core.py => engine.py} (95%)
 rename ppcls/engine/{eval => evaluation}/__init__.py (82%)
 rename ppcls/engine/{eval => evaluation}/classification.py (100%)
 rename ppcls/engine/{eval => evaluation}/retrieval.py (100%)
 delete mode 100644 ppcls/engine/train/classification.py
 rename ppcls/engine/train/{retrieval.py => train.py} (90%)

diff --git a/ppcls/engine/core.py b/ppcls/engine/engine.py
similarity index 95%
rename from ppcls/engine/core.py
rename to ppcls/engine/engine.py
index d0e448d71..711ec3b70 100644
--- a/ppcls/engine/core.py
+++ b/ppcls/engine/engine.py
@@ -47,19 +47,18 @@ from ppcls.utils import save_load
 from ppcls.data.utils.get_image_list import get_image_list
 from ppcls.data.postprocess import build_postprocess
 from ppcls.data import create_operators
-from ppcls.engine.train import classification_train, retrieval_train
-from ppcls.engine.eval import classification_eval, retrieval_eval
+from ppcls.engine.train import train_epoch
+from ppcls.engine import evaluation
 from ppcls.arch.gears.identity_head import IdentityHead
 
 
-class Core(object):
+class Engine(object):
     def __init__(self, config, mode="train"):
-        assert mode in ['train', 'eval', 'infer', 'export']
+        assert mode in ["train", "eval", "infer", "export"]
         self.mode = mode
         self.config = config
         self.eval_mode = self.config["Global"].get("eval_mode",
                                                    "classification")
-
         # init logger
         self.output_dir = self.config['Global']['output_dir']
         log_file = os.path.join(self.output_dir, self.config["Arch"]["name"],
@@ -68,14 +67,10 @@ class Core(object):
         print_config(config)
 
         # init train_func and eval_func
-        if self.eval_mode == "classification":
-            self.evaler = classification_eval
-            self.trainer = classification_train
-        elif self.eval_mode == "retrieval":
-            self.trainer = retrieval_train
-            self.evaler = retrieval_eval
-        else:
-            logger.warning("Invalid eval mode: {}".format(self.eval_mode))
+        assert self.eval_mode in ["classification", "retrieval"], logger.error("Invalid eval mode: {}".format(self.eval_mode))
+        self.train_epoch_func = train_epoch
+        self.eval_func = getattr(evaluation, self.eval_mode + "_eval")
+
         self.use_dali = self.config['Global'].get("use_dali", False)
 
         # for visualdl
@@ -242,7 +237,7 @@ class Core(object):
                               self.config["Global"]["epochs"] + 1):
             acc = 0.0
             # for one epoch train
-            self.trainer(self, epoch_id, print_batch_step)
+            self.train_epoch_func(self, epoch_id, print_batch_step)
 
             if self.use_dali:
                 self.train_dataloader.reset()
@@ -304,7 +299,7 @@ class Core(object):
     def eval(self, epoch_id=0):
         assert self.mode in ["train", "eval"]
         self.model.eval()
-        eval_result = self.evaler(self, epoch_id)
+        eval_result = self.eval_func(self, epoch_id)
         self.model.train()
         return eval_result
 
diff --git a/ppcls/engine/eval/__init__.py b/ppcls/engine/evaluation/__init__.py
similarity index 82%
rename from ppcls/engine/eval/__init__.py
rename to ppcls/engine/evaluation/__init__.py
index 0cc0dc983..e0cd77888 100644
--- a/ppcls/engine/eval/__init__.py
+++ b/ppcls/engine/evaluation/__init__.py
@@ -12,5 +12,5 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from ppcls.engine.eval.classification import classification_eval
-from ppcls.engine.eval.retrieval import retrieval_eval
+from ppcls.engine.evaluation.classification import classification_eval
+from ppcls.engine.evaluation.retrieval import retrieval_eval
diff --git a/ppcls/engine/eval/classification.py b/ppcls/engine/evaluation/classification.py
similarity index 100%
rename from ppcls/engine/eval/classification.py
rename to ppcls/engine/evaluation/classification.py
diff --git a/ppcls/engine/eval/retrieval.py b/ppcls/engine/evaluation/retrieval.py
similarity index 100%
rename from ppcls/engine/eval/retrieval.py
rename to ppcls/engine/evaluation/retrieval.py
diff --git a/ppcls/engine/train/__init__.py b/ppcls/engine/train/__init__.py
index 897c5846d..800d3a41e 100644
--- a/ppcls/engine/train/__init__.py
+++ b/ppcls/engine/train/__init__.py
@@ -11,5 +11,4 @@
 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 # See the License for the specific language governing permissions and
 # limitations under the License.
-from ppcls.engine.train.classification import classification_train
-from ppcls.engine.train.retrieval import retrieval_train
+from ppcls.engine.train.train import train_epoch
diff --git a/ppcls/engine/train/classification.py b/ppcls/engine/train/classification.py
deleted file mode 100644
index c3d74f962..000000000
--- a/ppcls/engine/train/classification.py
+++ /dev/null
@@ -1,89 +0,0 @@
-# Copyright (c) 2021 PaddlePaddle Authors. All Rights Reserved.
-#
-# Licensed under the Apache License, Version 2.0 (the "License");
-# you may not use this file except in compliance with the License.
-# You may obtain a copy of the License at
-#
-#     http://www.apache.org/licenses/LICENSE-2.0
-#
-# Unless required by applicable law or agreed to in writing, software
-# distributed under the License is distributed on an "AS IS" BASIS,
-# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
-# See the License for the specific language governing permissions and
-# limitations under the License.
-from __future__ import absolute_import, division, print_function
-
-import datetime
-import os
-import platform
-import sys
-import time
-
-import numpy as np
-import paddle
-
-__dir__ = os.path.dirname(os.path.abspath(__file__))
-sys.path.append(os.path.abspath(os.path.join(__dir__, '../../../')))
-from ppcls.utils import logger
-from ppcls.utils.misc import AverageMeter
-from ppcls.engine.train.utils import update_loss, update_metric, log_info
-
-
-def classification_train(trainer, epoch_id, print_batch_step):
-    tic = time.time()
-
-    train_dataloader = trainer.train_dataloader if trainer.use_dali else trainer.train_dataloader(
-    )
-    for iter_id, batch in enumerate(train_dataloader):
-        if iter_id >= trainer.max_iter:
-            break
-        if iter_id == 5:
-            for key in trainer.time_info:
-                trainer.time_info[key].reset()
-        trainer.time_info["reader_cost"].update(time.time() - tic)
-        if trainer.use_dali:
-            batch = [
-                paddle.to_tensor(batch[0]['data']),
-                paddle.to_tensor(batch[0]['label'])
-            ]
-        batch_size = batch[0].shape[0]
-        batch[1] = batch[1].reshape([-1, 1]).astype("int64")
-
-        trainer.global_step += 1
-        # image input
-        if trainer.amp:
-            with paddle.amp.auto_cast(custom_black_list={
-                    "flatten_contiguous_range", "greater_than"
-            }):
-                out = trainer.model(batch[0])
-                loss_dict = trainer.train_loss_func(out, batch[1])
-        else:
-            out = trainer.model(batch[0])
-
-        # calc loss
-        if trainer.config["DataLoader"]["Train"]["dataset"].get(
-                "batch_transform_ops", None):
-            loss_dict = trainer.train_loss_func(out, batch[1:])
-        else:
-            loss_dict = trainer.train_loss_func(out, batch[1])
-
-        # step opt and lr
-        if trainer.amp:
-            scaled = trainer.scaler.scale(loss_dict["loss"])
-            scaled.backward()
-            trainer.scaler.minimize(trainer.optimizer, scaled)
-        else:
-            loss_dict["loss"].backward()
-            trainer.optimizer.step()
-        trainer.optimizer.clear_grad()
-        trainer.lr_sch.step()
-
-        # below code just for logging
-        # update metric_for_logger
-        update_metric(trainer, out, batch, batch_size)
-        # update_loss_for_logger
-        update_loss(trainer, loss_dict, batch_size)
-        trainer.time_info["batch_cost"].update(time.time() - tic)
-        if iter_id % print_batch_step == 0:
-            log_info(trainer, batch_size, epoch_id, iter_id)
-        tic = time.time()
diff --git a/ppcls/engine/train/retrieval.py b/ppcls/engine/train/train.py
similarity index 90%
rename from ppcls/engine/train/retrieval.py
rename to ppcls/engine/train/train.py
index d4a994597..9ea173a9e 100644
--- a/ppcls/engine/train/retrieval.py
+++ b/ppcls/engine/train/train.py
@@ -29,7 +29,7 @@ from ppcls.utils.misc import AverageMeter
 from ppcls.engine.train.utils import update_loss, update_metric, log_info
 
 
-def retrieval_train(trainer, epoch_id, print_batch_step):
+def train_epoch(trainer, epoch_id, print_batch_step):
     tic = time.time()
 
     train_dataloader = trainer.train_dataloader if trainer.use_dali else trainer.train_dataloader(
@@ -55,10 +55,10 @@ def retrieval_train(trainer, epoch_id, print_batch_step):
             with paddle.amp.auto_cast(custom_black_list={
                     "flatten_contiguous_range", "greater_than"
             }):
-                out = trainer.model(batch[0], batch[1])
+                out = forward(trainer, batch)
                 loss_dict = trainer.train_loss_func(out, batch[1])
         else:
-            out = trainer.model(batch[0], batch[1])
+            out = forward(trainer, batch)
 
         # calc loss
         if trainer.config["DataLoader"]["Train"]["dataset"].get(
@@ -81,10 +81,15 @@ def retrieval_train(trainer, epoch_id, print_batch_step):
         # below code just for logging
         # update metric_for_logger
         update_metric(trainer, out, batch, batch_size)
-
         # update_loss_for_logger
         update_loss(trainer, loss_dict, batch_size)
         trainer.time_info["batch_cost"].update(time.time() - tic)
         if iter_id % print_batch_step == 0:
             log_info(trainer, batch_size, epoch_id, iter_id)
         tic = time.time()
+
+def forward(trainer, batch):
+    if trainer.eval_mode == "classification":
+        return trainer.model(batch[0])
+    else:
+        return trainer.model(batch[0], batch[1])
diff --git a/tools/eval.py b/tools/eval.py
index 367f5f8b4..e086da1b5 100644
--- a/tools/eval.py
+++ b/tools/eval.py
@@ -21,11 +21,11 @@ __dir__ = os.path.dirname(os.path.abspath(__file__))
 sys.path.append(os.path.abspath(os.path.join(__dir__, '../')))
 
 from ppcls.utils import config
-from ppcls.engine.core import Core
+from ppcls.engine.engine import Engine
 
 if __name__ == "__main__":
     args = config.parse_args()
     config = config.get_config(
         args.config, overrides=args.override, show=False)
-    evaler = Core(config, mode="eval")
-    evaler.eval()
+    engine = Engine(config, mode="eval")
+    engine.eval()
diff --git a/tools/export_model.py b/tools/export_model.py
index 7d324a613..01aba06c1 100644
--- a/tools/export_model.py
+++ b/tools/export_model.py
@@ -24,11 +24,11 @@ import paddle
 import paddle.nn as nn
 
 from ppcls.utils import config
-from ppcls.engine.core import Core
+from ppcls.engine.engine import Engine
 
 if __name__ == "__main__":
     args = config.parse_args()
     config = config.get_config(
         args.config, overrides=args.override, show=False)
-    exporter = Core(config, mode="export")
-    exporter.export()
+    engine = Engine(config, mode="export")
+    engine.export()
diff --git a/tools/infer.py b/tools/infer.py
index 9c00542c1..4f6bf9272 100644
--- a/tools/infer.py
+++ b/tools/infer.py
@@ -21,11 +21,11 @@ __dir__ = os.path.dirname(os.path.abspath(__file__))
 sys.path.append(os.path.abspath(os.path.join(__dir__, '../')))
 
 from ppcls.utils import config
-from ppcls.engine.core import Core
+from ppcls.engine.engine import Engine
 
 if __name__ == "__main__":
     args = config.parse_args()
     config = config.get_config(
         args.config, overrides=args.override, show=False)
-    inferer = Core(config, mode="infer")
-    inferer.infer()
+    engine = Engine(config, mode="infer")
+    engine.infer()
diff --git a/tools/train.py b/tools/train.py
index 23a064e7b..1d8359036 100644
--- a/tools/train.py
+++ b/tools/train.py
@@ -21,11 +21,11 @@ __dir__ = os.path.dirname(os.path.abspath(__file__))
 sys.path.append(os.path.abspath(os.path.join(__dir__, '../')))
 
 from ppcls.utils import config
-from ppcls.engine.core import Core
+from ppcls.engine.engine import Engine
 
 if __name__ == "__main__":
     args = config.parse_args()
     config = config.get_config(
         args.config, overrides=args.override, show=False)
-    trainer = Core(config, mode="train")
-    trainer.train()
+    engine = Engine(config, mode="train")
+    engine.train()

From 8c8153e97fdc15f1e8f6fd3429373dd8570fba2a Mon Sep 17 00:00:00 2001
From: cuicheng01 <cuicheng_smile@163.com>
Date: Tue, 24 Aug 2021 03:22:03 +0000
Subject: [PATCH 20/35] Update SwinTransformer pretrained url

---
 ppcls/arch/backbone/model_zoo/swin_transformer.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/ppcls/arch/backbone/model_zoo/swin_transformer.py b/ppcls/arch/backbone/model_zoo/swin_transformer.py
index f4348fbae..8ce810c2f 100644
--- a/ppcls/arch/backbone/model_zoo/swin_transformer.py
+++ b/ppcls/arch/backbone/model_zoo/swin_transformer.py
@@ -33,9 +33,9 @@ MODEL_URLS = {
     "SwinTransformer_base_patch4_window12_384":
     "https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/SwinTransformer_base_patch4_window12_384_pretrained.pdparams",
     "SwinTransformer_large_patch4_window7_224":
-    "https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/SwinTransformer_large_patch4_window7_224_pretrained.pdparams",
+    "https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/SwinTransformer_large_patch4_window7_224_22kto1k_pretrained.pdparams",
     "SwinTransformer_large_patch4_window12_384":
-    "https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/SwinTransformer_large_patch4_window12_384_pretrained.pdparams",
+    "https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/SwinTransformer_large_patch4_window12_384_22kto1k_pretrained.pdparams",
 }
 
 __all__ = list(MODEL_URLS.keys())

From 5bf74beb46533152b76cf59e891bafb0ec402cd2 Mon Sep 17 00:00:00 2001
From: dongshuilong <dongshuilong@baidu.com>
Date: Tue, 24 Aug 2021 07:07:17 +0000
Subject: [PATCH 21/35] update according comments

---
 deploy/auto_log.log                       |  0
 ppcls/engine/engine.py                    | 15 ++++-----------
 ppcls/engine/evaluation/classification.py |  8 ++------
 ppcls/engine/evaluation/retrieval.py      |  8 +-------
 ppcls/engine/train/train.py               | 12 +-----------
 ppcls/engine/train/utils.py               | 10 ----------
 6 files changed, 8 insertions(+), 45 deletions(-)
 delete mode 100644 deploy/auto_log.log

diff --git a/deploy/auto_log.log b/deploy/auto_log.log
deleted file mode 100644
index e69de29bb..000000000
diff --git a/ppcls/engine/engine.py b/ppcls/engine/engine.py
index 711ec3b70..ad5c584f0 100644
--- a/ppcls/engine/engine.py
+++ b/ppcls/engine/engine.py
@@ -13,21 +13,13 @@
 from __future__ import absolute_import
 from __future__ import division
 from __future__ import print_function
+
 import os
-import sys
-import numpy as np
-
-__dir__ = os.path.dirname(os.path.abspath(__file__))
-sys.path.append(os.path.abspath(os.path.join(__dir__, '../../')))
-
-import time
 import platform
-import datetime
-import argparse
 import paddle
-import paddle.nn as nn
 import paddle.distributed as dist
 from visualdl import LogWriter
+from paddle import nn
 
 from ppcls.utils.check import check_gpu
 from ppcls.utils.misc import AverageMeter
@@ -67,7 +59,8 @@ class Engine(object):
         print_config(config)
 
         # init train_func and eval_func
-        assert self.eval_mode in ["classification", "retrieval"], logger.error("Invalid eval mode: {}".format(self.eval_mode))
+        assert self.eval_mode in ["classification", "retrieval"], logger.error(
+            "Invalid eval mode: {}".format(self.eval_mode))
         self.train_epoch_func = train_epoch
         self.eval_func = getattr(evaluation, self.eval_mode + "_eval")
 
diff --git a/ppcls/engine/evaluation/classification.py b/ppcls/engine/evaluation/classification.py
index 38be7cd3f..b1ddc4186 100644
--- a/ppcls/engine/evaluation/classification.py
+++ b/ppcls/engine/evaluation/classification.py
@@ -14,14 +14,10 @@
 from __future__ import absolute_import
 from __future__ import division
 from __future__ import print_function
-import os
-import sys
 import time
 import platform
 import paddle
 
-__dir__ = os.path.dirname(os.path.abspath(__file__))
-sys.path.append(os.path.abspath(os.path.join(__dir__, '../../../')))
 from ppcls.utils.misc import AverageMeter
 from ppcls.utils import logger
 
@@ -61,14 +57,14 @@ def classification_eval(evaler, epoch_id=0):
         out = evaler.model(batch[0])
         # calc loss
         if evaler.eval_loss_func is not None:
-            loss_dict = evaler.eval_loss_func(out, batch[-1])
+            loss_dict = evaler.eval_loss_func(out, batch[1])
             for key in loss_dict:
                 if key not in output_info:
                     output_info[key] = AverageMeter(key, '7.5f')
                 output_info[key].update(loss_dict[key].numpy()[0], batch_size)
         # calc metric
         if evaler.eval_metric_func is not None:
-            metric_dict = evaler.eval_metric_func(out, batch[-1])
+            metric_dict = evaler.eval_metric_func(out, batch[1])
             if paddle.distributed.get_world_size() > 1:
                 for key in metric_dict:
                     paddle.distributed.all_reduce(
diff --git a/ppcls/engine/evaluation/retrieval.py b/ppcls/engine/evaluation/retrieval.py
index ccba31754..c0da3ace2 100644
--- a/ppcls/engine/evaluation/retrieval.py
+++ b/ppcls/engine/evaluation/retrieval.py
@@ -14,15 +14,9 @@
 from __future__ import absolute_import
 from __future__ import division
 from __future__ import print_function
-import os
-import sys
-import time
+
 import platform
 import paddle
-
-__dir__ = os.path.dirname(os.path.abspath(__file__))
-sys.path.append(os.path.abspath(os.path.join(__dir__, '../../../')))
-from ppcls.utils.misc import AverageMeter
 from ppcls.utils import logger
 
 
diff --git a/ppcls/engine/train/train.py b/ppcls/engine/train/train.py
index 9ea173a9e..9e36a063e 100644
--- a/ppcls/engine/train/train.py
+++ b/ppcls/engine/train/train.py
@@ -13,19 +13,8 @@
 # limitations under the License.
 from __future__ import absolute_import, division, print_function
 
-import datetime
-import os
-import platform
-import sys
 import time
-
-import numpy as np
 import paddle
-
-__dir__ = os.path.dirname(os.path.abspath(__file__))
-sys.path.append(os.path.abspath(os.path.join(__dir__, '../../../')))
-from ppcls.utils import logger
-from ppcls.utils.misc import AverageMeter
 from ppcls.engine.train.utils import update_loss, update_metric, log_info
 
 
@@ -88,6 +77,7 @@ def train_epoch(trainer, epoch_id, print_batch_step):
             log_info(trainer, batch_size, epoch_id, iter_id)
         tic = time.time()
 
+
 def forward(trainer, batch):
     if trainer.eval_mode == "classification":
         return trainer.model(batch[0])
diff --git a/ppcls/engine/train/utils.py b/ppcls/engine/train/utils.py
index f4cbce97e..b82694f31 100644
--- a/ppcls/engine/train/utils.py
+++ b/ppcls/engine/train/utils.py
@@ -14,16 +14,6 @@
 from __future__ import absolute_import, division, print_function
 
 import datetime
-import os
-import platform
-import sys
-import time
-
-import numpy as np
-import paddle
-
-__dir__ = os.path.dirname(os.path.abspath(__file__))
-sys.path.append(os.path.abspath(os.path.join(__dir__, '../../../')))
 from ppcls.utils import logger
 from ppcls.utils.misc import AverageMeter
 

From 5b70d442f9b26f79fbb46779eaf4e33b6a787cd3 Mon Sep 17 00:00:00 2001
From: Bin Lu <lubing02@gmail.com>
Date: Tue, 24 Aug 2021 16:19:41 +0800
Subject: [PATCH 22/35] Update circlemargin.py

---
 ppcls/arch/gears/circlemargin.py | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)

diff --git a/ppcls/arch/gears/circlemargin.py b/ppcls/arch/gears/circlemargin.py
index 65967d0d7..87baee839 100644
--- a/ppcls/arch/gears/circlemargin.py
+++ b/ppcls/arch/gears/circlemargin.py
@@ -28,7 +28,7 @@ class CircleMargin(nn.Layer):
 
         weight_attr = paddle.ParamAttr(
             initializer=paddle.nn.initializer.XavierNormal())
-        self.fc0 = paddle.nn.Linear(
+        self.fc = paddle.nn.Linear(
             self.embedding_size, self.class_num, weight_attr=weight_attr)
 
     def forward(self, input, label):
@@ -36,19 +36,22 @@ class CircleMargin(nn.Layer):
             paddle.sum(paddle.square(input), axis=1, keepdim=True))
         input = paddle.divide(input, feat_norm)
 
-        weight = self.fc0.weight
+        weight = self.fc.weight
         weight_norm = paddle.sqrt(
             paddle.sum(paddle.square(weight), axis=0, keepdim=True))
         weight = paddle.divide(weight, weight_norm)
 
         logits = paddle.matmul(input, weight)
+        if not self.training or label is None:
+            return logits
 
         alpha_p = paddle.clip(-logits.detach() + 1 + self.margin, min=0.)
         alpha_n = paddle.clip(logits.detach() + self.margin, min=0.)
         delta_p = 1 - self.margin
         delta_n = self.margin
-        index = paddle.fluid.layers.where(label != -1).reshape([-1])
+        
         m_hot = F.one_hot(label.reshape([-1]), num_classes=logits.shape[1])
+        
         logits_p = alpha_p * (logits - delta_p)
         logits_n = alpha_n * (logits - delta_n)
         pre_logits = logits_p * m_hot + logits_n * (1 - m_hot)

From 860e4aa42182acc9fa4eefb054bce063543a2257 Mon Sep 17 00:00:00 2001
From: Bin Lu <lubing02@gmail.com>
Date: Tue, 24 Aug 2021 16:20:21 +0800
Subject: [PATCH 23/35] Update cosmargin.py

---
 ppcls/arch/gears/cosmargin.py | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/ppcls/arch/gears/cosmargin.py b/ppcls/arch/gears/cosmargin.py
index 51db55086..378e102a2 100644
--- a/ppcls/arch/gears/cosmargin.py
+++ b/ppcls/arch/gears/cosmargin.py
@@ -46,6 +46,9 @@ class CosMargin(paddle.nn.Layer):
         weight = paddle.divide(weight, weight_norm)
 
         cos = paddle.matmul(input, weight)
+        if not self.training or label is None:
+            return cos
+        
         cos_m = cos - self.margin
 
         one_hot = paddle.nn.functional.one_hot(label, self.class_num)

From 60ef76b1b15f6051553dc054abf0966fba53ba21 Mon Sep 17 00:00:00 2001
From: lyuwenyu <wenyu.lyu@gmail.com>
Date: Wed, 25 Aug 2021 13:20:40 +0800
Subject: [PATCH 24/35] add logger init

---
 hubconf.py | 13 +++++++++----
 1 file changed, 9 insertions(+), 4 deletions(-)

diff --git a/hubconf.py b/hubconf.py
index d615a905d..b7f76745a 100644
--- a/hubconf.py
+++ b/hubconf.py
@@ -41,10 +41,15 @@ class _SysPathG(object):
             self.path)
 
 
-with _SysPathG(
-        os.path.join(
-            os.path.dirname(os.path.abspath(__file__)), 'ppcls', 'arch')):
-    import backbone
+with _SysPathG(os.path.dirname(os.path.abspath(__file__)), ):
+    import ppcls
+    import ppcls.arch.backbone as backbone
+
+    def ppclas_init():
+        if ppcls.utils.logger._logger is None:
+            ppcls.utils.logger.init_logger()
+
+    ppclas_init()
 
     def _load_pretrained_parameters(model, name):
         url = 'https://paddle-imagenet-models-name.bj.bcebos.com/dygraph/{}_pretrained.pdparams'.format(

From 52663edf603d9f2a1cbe96f2cbb8fabc95d5eaab Mon Sep 17 00:00:00 2001
From: Bin Lu <lubing02@gmail.com>
Date: Wed, 25 Aug 2021 13:57:24 +0800
Subject: [PATCH 25/35] Update retrieval.py

---
 ppcls/engine/evaluation/retrieval.py | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/ppcls/engine/evaluation/retrieval.py b/ppcls/engine/evaluation/retrieval.py
index c0da3ace2..cd8830a55 100644
--- a/ppcls/engine/evaluation/retrieval.py
+++ b/ppcls/engine/evaluation/retrieval.py
@@ -124,6 +124,13 @@ def cal_feature(evaler, name='gallery'):
             feas_norm = paddle.sqrt(
                 paddle.sum(paddle.square(batch_feas), axis=1, keepdim=True))
             batch_feas = paddle.divide(batch_feas, feas_norm)
+            
+        # do binarize
+        if self.config["Global"].get("feature_binarize") == "round":
+            batch_feas = paddle.round(batch_feas).astype("float32") * 2.0 - 1.0
+
+        if self.config["Global"].get("feature_binarize") == "sign":
+            batch_feas = paddle.sign(batch_feas).astype("float32")
 
         if all_feas is None:
             all_feas = batch_feas
@@ -135,8 +142,10 @@ def cal_feature(evaler, name='gallery'):
             all_image_id = paddle.concat([all_image_id, batch[1]])
             if has_unique_id:
                 all_unique_id = paddle.concat([all_unique_id, batch[2]])
+                
     if evaler.use_dali:
         dataloader_tmp.reset()
+        
     if paddle.distributed.get_world_size() > 1:
         feat_list = []
         img_id_list = []

From 9323b1478a07f0356682b447ea19d11285bf04ed Mon Sep 17 00:00:00 2001
From: Bin Lu <lubing02@gmail.com>
Date: Wed, 25 Aug 2021 13:59:54 +0800
Subject: [PATCH 26/35] Create deephashloss.py

---
 ppcls/loss/deephashloss.py | 6 ++++++
 1 file changed, 6 insertions(+)
 create mode 100644 ppcls/loss/deephashloss.py

diff --git a/ppcls/loss/deephashloss.py b/ppcls/loss/deephashloss.py
new file mode 100644
index 000000000..8f02a3ba0
--- /dev/null
+++ b/ppcls/loss/deephashloss.py
@@ -0,0 +1,6 @@
+            # do binarize
+            if self.config["Global"].get("feature_binarize") == "round":
+                batch_feas = paddle.round(batch_feas).astype("float32") * 2.0 - 1.0
+
+            if self.config["Global"].get("feature_binarize") == "sign":
+                batch_feas = paddle.sign(batch_feas).astype("float32")

From e732d40a45e4785ce439f1dc85d4aed93707cee6 Mon Sep 17 00:00:00 2001
From: Bin Lu <lubing02@gmail.com>
Date: Wed, 25 Aug 2021 14:00:21 +0800
Subject: [PATCH 27/35] Update deephashloss.py

---
 ppcls/loss/deephashloss.py | 63 +++++++++++++++++++++++++++++++++++---
 1 file changed, 58 insertions(+), 5 deletions(-)

diff --git a/ppcls/loss/deephashloss.py b/ppcls/loss/deephashloss.py
index 8f02a3ba0..b70936994 100644
--- a/ppcls/loss/deephashloss.py
+++ b/ppcls/loss/deephashloss.py
@@ -1,6 +1,59 @@
-            # do binarize
-            if self.config["Global"].get("feature_binarize") == "round":
-                batch_feas = paddle.round(batch_feas).astype("float32") * 2.0 - 1.0
+#copyright (c) 2021 PaddlePaddle Authors. All Rights Reserve.
+#
+#Licensed under the Apache License, Version 2.0 (the "License");
+#you may not use this file except in compliance with the License.
+#You may obtain a copy of the License at
+#
+#    http://www.apache.org/licenses/LICENSE-2.0
+#
+#Unless required by applicable law or agreed to in writing, software
+#distributed under the License is distributed on an "AS IS" BASIS,
+#WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+#See the License for the specific language governing permissions and
+#limitations under the License.
+
+import paddle
+import paddle.nn as nn
+
+class DSHSDLoss(nn.Layer):
+    """
+    # DSHSD(IEEE ACCESS 2019)
+    # paper [Deep Supervised Hashing Based on Stable Distribution](https://ieeexplore.ieee.org/document/8648432/)
+    # [DSHSD] epoch:70,  bit:48,  dataset:cifar10-1,  MAP:0.809, Best MAP: 0.809
+    # [DSHSD] epoch:250, bit:48,  dataset:nuswide_21, MAP:0.809, Best MAP: 0.815
+    # [DSHSD] epoch:135, bit:48,  dataset:imagenet,   MAP:0.647, Best MAP: 0.647
+    """
+    def __init__(self, n_class, bit, alpha, multi_label=False):
+        super(DSHSDLoss, self).__init__()
+        self.m = 2 * bit     
+        self.alpha = alpha
+        self.multi_label = multi_label
+        self.n_class = n_class
+        self.fc = paddle.nn.Linear(bit, n_class, bias_attr=False)
+
+    def forward(self, input, label):        
+        feature = input["features"]
+        feature = feature.tanh().astype("float32")
+
+        dist = paddle.sum(
+                    paddle.square((paddle.unsqueeze(feature, 1) - paddle.unsqueeze(feature, 0))), 
+                    axis=2)
+        
+        # label to ont-hot
+        label = paddle.flatten(label)
+        label = paddle.nn.functional.one_hot(label,  self.n_class).astype("float32")
+
+        s = (paddle.matmul(label, label, transpose_y=True) == 0).astype("float32")
+        Ld = (1 - s) / 2 * dist + s / 2 * (self.m - dist).clip(min=0)
+        Ld = Ld.mean()
+        
+        logits = self.fc(feature)
+        if self.multi_label:
+            # multiple labels classification loss
+            Lc = (logits - label * logits + ((1 + (-logits).exp()).log())).sum(axis=1).mean()
+        else:
+            # single labels classification loss
+            Lc = (-paddle.nn.functional.softmax(logits).log() * label).sum(axis=1).mean()
+
+        return {"dshsdloss": Lc + Ld * self.alpha}
 
-            if self.config["Global"].get("feature_binarize") == "sign":
-                batch_feas = paddle.sign(batch_feas).astype("float32")

From d9ca31a694db44913efdddcff97cf5097a6cf8c6 Mon Sep 17 00:00:00 2001
From: Bin Lu <lubing02@gmail.com>
Date: Wed, 25 Aug 2021 14:04:50 +0800
Subject: [PATCH 28/35] Update deephashloss.py

---
 ppcls/loss/deephashloss.py | 32 ++++++++++++++++++++++++++++++++
 1 file changed, 32 insertions(+)

diff --git a/ppcls/loss/deephashloss.py b/ppcls/loss/deephashloss.py
index b70936994..49dc3b260 100644
--- a/ppcls/loss/deephashloss.py
+++ b/ppcls/loss/deephashloss.py
@@ -57,3 +57,35 @@ class DSHSDLoss(nn.Layer):
 
         return {"dshsdloss": Lc + Ld * self.alpha}
 
+class LCDSHLoss(paddle.nn.Layer):
+    """
+    # paper [Locality-Constrained Deep Supervised Hashing for Image Retrieval](https://www.ijcai.org/Proceedings/2017/0499.pdf)
+    # [LCDSH] epoch:145, bit:48, dataset:cifar10-1,  MAP:0.798, Best MAP: 0.798
+    # [LCDSH] epoch:183, bit:48, dataset:nuswide_21, MAP:0.833, Best MAP: 0.834
+    """
+    def __init__(self, n_class, _lambda):
+        super(LCDSHLoss, self).__init__()
+        self._lambda = _lambda
+        self.n_class = n_class
+
+    def forward(self, input, label):
+        feature = input["features"]
+        label = label.astype("float32")
+
+        # label to ont-hot
+        label = paddle.flatten(label)
+        label = paddle.nn.functional.one_hot(label,  self.n_class).astype("float32")
+        
+        s = 2 * (paddle.matmul(label, label, transpose_y=True) > 0).astype("float32") - 1
+        inner_product = paddle.matmul(feature, feature, transpose_y=True) * 0.5
+
+        inner_product = inner_product.clip(min=-50, max=50)
+        L1 = paddle.log(1 + paddle.exp(-s * inner_product)).mean()
+
+        b = feature.sign()
+        inner_product_ = paddle.matmul(b, b, transpose_y=True) * 0.5
+        sigmoid = paddle.nn.Sigmoid()
+        L2 = (sigmoid(inner_product) - sigmoid(inner_product_)).pow(2).mean()
+
+        return {"lcdshloss": L1 + self._lambda * L2}
+

From 3de99430c0858d138869a98bd7d66671e86a683e Mon Sep 17 00:00:00 2001
From: Bin Lu <lubing02@gmail.com>
Date: Wed, 25 Aug 2021 14:06:54 +0800
Subject: [PATCH 29/35] Update deephashloss.py

---
 ppcls/loss/deephashloss.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/ppcls/loss/deephashloss.py b/ppcls/loss/deephashloss.py
index 49dc3b260..7d4efad16 100644
--- a/ppcls/loss/deephashloss.py
+++ b/ppcls/loss/deephashloss.py
@@ -57,7 +57,7 @@ class DSHSDLoss(nn.Layer):
 
         return {"dshsdloss": Lc + Ld * self.alpha}
 
-class LCDSHLoss(paddle.nn.Layer):
+class LCDSHLoss(nn.Layer):
     """
     # paper [Locality-Constrained Deep Supervised Hashing for Image Retrieval](https://www.ijcai.org/Proceedings/2017/0499.pdf)
     # [LCDSH] epoch:145, bit:48, dataset:cifar10-1,  MAP:0.798, Best MAP: 0.798

From 9608c078afabd7e9ef8dd8096496ce7a149da048 Mon Sep 17 00:00:00 2001
From: Bin Lu <lubing02@gmail.com>
Date: Wed, 25 Aug 2021 14:09:21 +0800
Subject: [PATCH 30/35] Create MV3_Large_1x_Aliproduct_DLBHC.yaml

---
 .../MV3_Large_1x_Aliproduct_DLBHC.yaml        | 147 ++++++++++++++++++
 1 file changed, 147 insertions(+)
 create mode 100644 ppcls/configs/Products/MV3_Large_1x_Aliproduct_DLBHC.yaml

diff --git a/ppcls/configs/Products/MV3_Large_1x_Aliproduct_DLBHC.yaml b/ppcls/configs/Products/MV3_Large_1x_Aliproduct_DLBHC.yaml
new file mode 100644
index 000000000..1838e3100
--- /dev/null
+++ b/ppcls/configs/Products/MV3_Large_1x_Aliproduct_DLBHC.yaml
@@ -0,0 +1,147 @@
+# global configs
+Global:
+  checkpoints: null
+  pretrained_model: ./output_dlbhc/RecModel/best_model
+  output_dir: ./output_dlbhc/
+  device: gpu
+  save_interval: 1
+  eval_during_train: True
+  eval_interval: 1
+  epochs: 100
+  #eval_mode: "retrieval"
+  print_batch_step: 10
+  use_visualdl: False
+
+  # used for static mode and model export
+  image_shape: [3, 224, 224]
+  save_inference_dir: ./inference
+
+  #feature postprocess
+  feature_normalize: False
+  feature_binarize: "round"
+
+# model architecture
+Arch:  
+  name: "RecModel"
+  Backbone:
+    name: "MobileNetV3_large_x1_0"
+    pretrained: True
+    class_num: 512
+  Head:
+    name: "FC"
+    class_num: 50030
+    embedding_size: 512
+    
+  infer_output_key:  "features"
+  infer_add_softmax: "false"
+ 
+# loss function config for train/eval process
+Loss:
+  Train:
+    - CELoss:
+        weight: 1.0
+        epsilon: 0.1
+  Eval:
+    - CELoss:
+        weight: 1.0
+
+Optimizer:
+  name: Momentum
+  momentum: 0.9
+  lr:
+    name: Piecewise
+    learning_rate: 0.1
+    decay_epochs: [50, 150]
+    values: [0.1, 0.01, 0.001]
+
+# data loader for train and eval
+DataLoader:
+  Train:
+    dataset:
+      name: ImageNetDataset
+      image_root: ./dataset/Aliproduct/
+      cls_label_path: ./dataset/Aliproduct/train_list.txt
+      transform_ops:
+        - DecodeImage:
+            to_rgb: True
+            channel_first: False
+        - ResizeImage:
+            size: 256
+        - RandCropImage:
+            size: 227
+        - RandFlipImage:
+            flip_code: 1
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.4914, 0.4822, 0.4465]
+            std: [0.2023, 0.1994, 0.2010]
+            order: ''
+    sampler:
+      name: DistributedBatchSampler
+      batch_size: 128
+      drop_last: False
+      shuffle: True
+    loader:
+      num_workers: 4
+      use_shared_memory: True
+
+  Eval:
+    dataset: 
+      name: ImageNetDataset
+      image_root: ./dataset/Aliproduct/
+      cls_label_path: ./dataset/Aliproduct/val_list.txt
+      transform_ops:
+        - DecodeImage:
+            to_rgb: True
+            channel_first: False
+        - ResizeImage:
+            size: 227
+        - NormalizeImage:
+            scale: 1.0/255.0
+            mean: [0.4914, 0.4822, 0.4465]
+            std: [0.2023, 0.1994, 0.2010]
+            order: ''
+    sampler:
+      name: DistributedBatchSampler
+      batch_size: 256
+      drop_last: False
+      shuffle: False
+    loader:
+      num_workers: 4
+      use_shared_memory: True
+
+Infer:
+  infer_imgs: docs/images/whl/demo.jpg
+  batch_size: 10
+  transforms:
+    - DecodeImage:
+        to_rgb: True
+        channel_first: False
+    - ResizeImage:
+        resize_short: 256
+    - CropImage:
+        size: 227
+    - NormalizeImage:
+        scale: 1.0/255.0
+        mean: [0.4914, 0.4822, 0.4465]
+        std: [0.2023, 0.1994, 0.2010]
+        order: ''
+    - ToCHWImage:
+  PostProcess:
+    name: Topk
+    topk: 5
+    class_id_map_file: ppcls/utils/imagenet1k_label_list.txt
+
+Metric:
+  Train:
+    - TopkAcc:
+        topk: [1, 5]
+  Eval:
+    - TopkAcc:
+        topk: [1, 5]  
+#     - Recallk:
+#         topk: [1]
+#     - mAP:
+#     - Precisionk:
+#         topk: [1]
+

From 8e0a5513591d07ed761e68ce5ce1447974b13467 Mon Sep 17 00:00:00 2001
From: Bin Lu <lubing02@gmail.com>
Date: Wed, 25 Aug 2021 14:34:43 +0800
Subject: [PATCH 31/35] Update deephashloss.py

---
 ppcls/loss/deephashloss.py | 1 -
 1 file changed, 1 deletion(-)

diff --git a/ppcls/loss/deephashloss.py b/ppcls/loss/deephashloss.py
index 7d4efad16..44c08ef3b 100644
--- a/ppcls/loss/deephashloss.py
+++ b/ppcls/loss/deephashloss.py
@@ -70,7 +70,6 @@ class LCDSHLoss(nn.Layer):
 
     def forward(self, input, label):
         feature = input["features"]
-        label = label.astype("float32")
 
         # label to ont-hot
         label = paddle.flatten(label)

From 3b19dc9c961307e4705e80c282226ac13a9853e3 Mon Sep 17 00:00:00 2001
From: cuicheng01 <cuicheng_smile@163.com>
Date: Thu, 26 Aug 2021 03:37:20 +0000
Subject: [PATCH 32/35] fix invalid link in getting_started

---
 docs/en/tutorials/getting_started_en.md | 2 +-
 docs/zh_CN/tutorials/getting_started.md | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/docs/en/tutorials/getting_started_en.md b/docs/en/tutorials/getting_started_en.md
index 19142b014..1903a04ef 100644
--- a/docs/en/tutorials/getting_started_en.md
+++ b/docs/en/tutorials/getting_started_en.md
@@ -23,7 +23,7 @@ Among them, `-c` is used to specify the path of the configuration file, `-o` is
 `-o use_gpu=True` means to use GPU for training. If you want to use the CPU for training, you need to set `use_gpu` to `False`.
 
 
-Of course, you can also directly modify the configuration file to update the configuration. For specific configuration parameters, please refer to [Configuration Document](config_en.md).
+Of course, you can also directly modify the configuration file to update the configuration. For specific configuration parameters, please refer to [Configuration Document](config_description_en.md).
 
 * The output log examples are as follows:
     * If mixup or cutmix is used in training, top-1 and top-k (default by 5) will not be printed in the log:
diff --git a/docs/zh_CN/tutorials/getting_started.md b/docs/zh_CN/tutorials/getting_started.md
index 7135b7934..3708d3e82 100644
--- a/docs/zh_CN/tutorials/getting_started.md
+++ b/docs/zh_CN/tutorials/getting_started.md
@@ -30,7 +30,7 @@ python3 tools/train.py \
 
 其中，`-c`用于指定配置文件的路径，`-o`用于指定需要修改或者添加的参数，其中`-o Arch.pretrained=False`表示不使用预训练模型，`-o Global.device=gpu`表示使用GPU进行训练。如果希望使用CPU进行训练，则需要将`Global.device`设置为`cpu`。
 
-更详细的训练配置，也可以直接修改模型对应的配置文件。具体配置参数参考[配置文档](config.md)。
+更详细的训练配置，也可以直接修改模型对应的配置文件。具体配置参数参考[配置文档](config_description.md)。
 
 运行上述命令，可以看到输出日志，示例如下：
 

From c9be0fe00a98510c775a75b619ce70847918f5a8 Mon Sep 17 00:00:00 2001
From: dongshuilong <dongshuilong@baidu.com>
Date: Thu, 26 Aug 2021 07:13:11 +0000
Subject: [PATCH 33/35] fix loss_msg bugs

---
 ppcls/engine/train/utils.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/ppcls/engine/train/utils.py b/ppcls/engine/train/utils.py
index b82694f31..92eb35d75 100644
--- a/ppcls/engine/train/utils.py
+++ b/ppcls/engine/train/utils.py
@@ -34,7 +34,7 @@ def update_loss(trainer, loss_dict, batch_size):
     for key in loss_dict:
         if key not in trainer.output_info:
             trainer.output_info[key] = AverageMeter(key, '7.5f')
-    trainer.output_info[key].update(loss_dict[key].numpy()[0], batch_size)
+        trainer.output_info[key].update(loss_dict[key].numpy()[0], batch_size)
 
 
 def log_info(trainer, batch_size, epoch_id, iter_id):

From f61b53104764e7527c57808ae47c1d360b4edfc0 Mon Sep 17 00:00:00 2001
From: Bin Lu <lubing02@gmail.com>
Date: Thu, 26 Aug 2021 16:27:29 +0800
Subject: [PATCH 34/35] Update MV3_Large_1x_Aliproduct_DLBHC.yaml

---
 ppcls/configs/Products/MV3_Large_1x_Aliproduct_DLBHC.yaml | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/ppcls/configs/Products/MV3_Large_1x_Aliproduct_DLBHC.yaml b/ppcls/configs/Products/MV3_Large_1x_Aliproduct_DLBHC.yaml
index 1838e3100..c9a8b7b40 100644
--- a/ppcls/configs/Products/MV3_Large_1x_Aliproduct_DLBHC.yaml
+++ b/ppcls/configs/Products/MV3_Large_1x_Aliproduct_DLBHC.yaml
@@ -1,7 +1,7 @@
 # global configs
 Global:
   checkpoints: null
-  pretrained_model: ./output_dlbhc/RecModel/best_model
+  pretrained_model: null
   output_dir: ./output_dlbhc/
   device: gpu
   save_interval: 1
@@ -138,7 +138,9 @@ Metric:
         topk: [1, 5]
   Eval:
     - TopkAcc:
-        topk: [1, 5]  
+        topk: [1, 5]
+        
+# switch to metric below when eval by retrieval
 #     - Recallk:
 #         topk: [1]
 #     - mAP:

From d388d69a63a3ed0c91b4ce4a31aad725cf01433e Mon Sep 17 00:00:00 2001
From: Bin Lu <lubing02@gmail.com>
Date: Thu, 26 Aug 2021 19:38:40 +0800
Subject: [PATCH 35/35] Update retrieval.py

---
 ppcls/engine/evaluation/retrieval.py | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/ppcls/engine/evaluation/retrieval.py b/ppcls/engine/evaluation/retrieval.py
index cd8830a55..49d9626f6 100644
--- a/ppcls/engine/evaluation/retrieval.py
+++ b/ppcls/engine/evaluation/retrieval.py
@@ -126,10 +126,10 @@ def cal_feature(evaler, name='gallery'):
             batch_feas = paddle.divide(batch_feas, feas_norm)
             
         # do binarize
-        if self.config["Global"].get("feature_binarize") == "round":
+        if evaler.config["Global"].get("feature_binarize") == "round":
             batch_feas = paddle.round(batch_feas).astype("float32") * 2.0 - 1.0
 
-        if self.config["Global"].get("feature_binarize") == "sign":
+        if evaler.config["Global"].get("feature_binarize") == "sign":
             batch_feas = paddle.sign(batch_feas).astype("float32")
 
         if all_feas is None: