From b50f1872ea6dbb386615c08789cef8c7a2c585ca Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Wed, 9 Aug 2023 17:36:17 +0800
Subject: [PATCH 01/45] Update wechat.jpg

Former-commit-id: ef5b299b18040e660eec264e3433774bf0748243
---
 assets/wechat.jpg | Bin 142464 -> 145701 bytes
 1 file changed, 0 insertions(+), 0 deletions(-)

diff --git a/assets/wechat.jpg b/assets/wechat.jpg
index 741cef3ce5bbc52b605c66c02e6b967000b8f0db..35932dd202142e5d6b88acacc360cde9e055dec3 100644
GIT binary patch
literal 145701
zcmeFZ2UJvFw<mZBl7r+-As`@8a!w*iL_h>2D>(=Vh!jvz1c{On5JWPPDan#^R3zt|
zBtsRkfTEcCf8XnV-P66+yqQ^V_4GIY%Q~#8drzHv_qqG*{oCo>&fKm5cb}+fssUJ7
z0Dy)018(O56#x(S4&EJHJiI%2`1p7PM0bgZ2nmU($;e6W(oxgX(^1pXGP2wUF*0#5
z)6%l>vvF{7Kj3}901|j8!1M4v&jX&n2EoF|$0s5nq9P)q;$fs^<oUmS+_nQ01b3!!
z@o})&0c;8^915)4ZUBUN&O2EDIDmgUu&{A3&x=n$NJNb3P<t1^#=^nD#>KgF2NxI9
z+aGfsz@@lDc~4jwk4o<~KD#@$NKir!0mq}t4w~nqa86Mxk6=O~S~_|L#`|2{JP&xq
z#3dx9q-9hdtE#DMXg<+@VPI(V(%8h>#@5c>!O_X{o!5JBu#a!ZhtRO_h{&kKq)*8y
zsh`u*bMx{Gz7-Y~msC~P)YjGiXlU&0>h9_7>mL{#8=sh*nx2`RTU}e<*!;7#y|arr
zJ~=%*zd&AI{Y4iRfb&nXF#rEa*#C(x3JhJ?xVSjD_<zxbh3$=LI25>d?g`^jD(m6D
zcBf(&2_m3=l#o-|LC7Kc98P29F-k<sDYkka@fT_TpzJ?JSn&TN%Kj^1|4!E&K!SsX
zd3iV#fFf}9Es7@_`|tMm9Q@4#fAhfKJn%OU{LKS@^T6Lc@V|}+@*LOn!;<kQLv1XV
z<vx_DQ(Rq51?UuKlYKdO2px84JE~NldvyzBjzX8M8Zxp>MEo6azFj{Boy-Yc@9Uj^
zR@|zwyqKx5=bnnrG68m4$MI3;AJNy;{@=(i59n^r`O*f?^0=o0W!!2yZUOn+TcEWE
z<QRhXQ@sWJqyBjKI}Dutue#YYW9tCQ$JYDp(?U_Y8gk6K_y(#rFD>g3%v+oLlQ5+_
z)O!DV?ZqVL@d5d@I*n^}?J)98=i(N~$tH2)$^9JsA6SO~<MzId$edbV1z$$_4If(0
z3I$EeCVr=<dN1@}ZMN*4p@{9Qqqq%cwd&e7<f+a@-?120j4HMMT|HH5;D2bE=_2wD
zSPuF9X9fShp6gHBkcf{6wB2_;sMakIN8bke<AA+Lt*7x?m0B0;U#S_-Eu(z7uh%ab
zN6QY}(DGC_2viJUjhUxzYCTne8qir}5HoOYc`hJes5wzp*@u_(p3&&N071TDmv-_~
zy_!UT!Z$rc2XUuhKRV*ccLs)|7bv4LRX10cyMG+RmcKcAG9FXFBE`ZY`h}=oNpa&^
zpBBY^6k`+dq7S?vMbm7h_FF(`=BZT<Zvk-TEl{90AvQzYJ1%(BPjE0-m{F^?RTM|F
z?RFtf4@Dfo3&|<BZ-J8=(8ltrg_4Whr$ww;qtj7^^F`K!QgvyI_S4chdIyQB&s!6@
zYRoHPw*XNbnhHvK3yheMUvIW`HmmgNa~{Wx7JYu}zSJZXHif30wld|r5s|9pWKOY?
zjB|DLOyYgOa)jbSGu;B&HK2>1&`_6I^A3L(mY)2N{FJ2q>pv>g&EZ+$VmpcAqlV^-
zPpUnyruFmYdtAQFtGbaV5M>jB%=5-dL>iJp>x4!%wG8e4noOFuf?J?3+A{7$JjnFQ
z5(Vx--U6aZwA;B+DnvZFf&WtN_tY)0KB~A3>Ys}FuZz+FhcSxh8(VG^4^UE{?(!|)
zT5Hnog*3fX>5CxqAmNHNd4xWD4uaKZ-IQL;r^^J-SRz<Mp6_dA+mFVE^tB7`&jc?o
zSZDUk?YqGykL^`Xx_|&^T!MvriBpHzSw`HF%zGB{Cmq#Q!ZHzDA9lv%yX?$ot|qrv
zL-fbvzsZd}6tyttf{9x3Eb}N>`S0bA9vzmu#B||oyw7*`W1(hrD#@yCl?VPras`=B
zv8%1tYOJT{-M@Y&$@wMZ-H`2WstzvCE#MV&HTJ`7v-U%Bt#SFrWoOo>q&>stdnM*C
zm*u{wy<)>-F)Y&dy9KCv-b>5T(hqqn5JYuaQmJ8csC$I}T6jo(Uc&XQH@0By@`z38
zL@H^5%hTfHs^ik0a#gy1szV1iz4Vol{8~=o&eM2xXIBO~&dXQRTB}VTnnvGj#Sz4@
zs{cH6+G%;?@j`4nZ%6fw)!)qZm&E^Xn5#^O%E>6*=ApJN<BtVMdzYNyqMEI;obHd|
zbPXDjnIhM>-jU_bQ)AjJ*@@EZGZlG&=PfV=<3ivHU@sg0nzxv-M(o8+wdRCP@xANj
zEolujih`KKkLPu~z|m0`F&VF}tn0_B++|)2i@;Ywu0|;miT9uX!mqytDw<%Vdor%D
zZ{|C@RfJlp=8IwALX=2DO>Ns3or(P!O<tEJr4IbD7H4I+-Iqd!(1J!z&%2)WEM$%^
z?{jojl)Jv!6joc*!Om`;<`+rve%(CHoTXy<*b;22)6f*W^0-u>$rYEZ&7^&@nI`{R
z+1xL9cR6TyyoHZA<8lhFmOeqqM8#MYGc;Lk?j7>3DQ)gZuim=xf$p8(Z<{wJOb*hW
zWJ1vej*i#yJ=VIyC+~*0hP+z|zNbEhcb}L)eq3GjlcHnT3C|9kiGbA_#t;f{gj#;{
zf$5D2gr+3Bya~Y;BMv%+iNQm&=20(V*cQHH|7m%J2r|-ivNg>nGT|wYe)Jb6`CEtn
z7afXI{vsK=W^0YUhCZo(>5FEc5c^pztcTNZjOb3oi9XLC;V6R~aSG_QjthtbF+Oh0
zzhZjfQmE22HwH;~d&o_eY{reSZd*+B1vyQ#dwxkfbF1HrnMn>Noj5Jah(zNQ-Azxb
zyAP!|{LWz}x|JotzP?6b;B}oJ%@g})QWca82bxkkYxoQUD83OC-5PhpINi0z>o13j
zMP~{b&nvaR`bcxL?BSTS*4riyZzhBfoEXd>ICb{oCI2DAmwKk{v4eXx>vtV@s=7z%
z#l0|5ng42sQA+<_%qKO%VH})HC#{Q#0+BhNDlx1Vaq+Ln2kT7c_u=_EBD}>*W+bj}
z;Cw7zz2nB=Gp!T*3n=6s^2H}llXtK@i>C-DSx&c4!E`9Ti5ltjttYi<^L{cm_2nKz
z<xjg);A2Tgx4>>iwPsr3_)7B1wXp4Q^3AYQqhFfAws~$8pZ-5IN%@V?KOt=y(efBZ
z>GTTqU8V!|s?^Wv{|C7DKZAh(8=W(JkC8B}k!U1Zt;Si_6a!5k-vSUQhZ22mO8oQj
ze?VNLdkD}kb6&2X09z|(ScP+MSt$+d`e}Qd1rJW2muWVPbS>WGDPCW>a_JM=W2SkJ
z<z`(FYs&+OHt|bMezr~equ46a=`V%Bg>5=28dV2}1^!|mncwuiowFP<`VJ-{msQr_
zmd#fsMxTN6t3HKq{$3sOP3`nI7Q}sHn%G2m+FtOw_r32|X5rbPvOX&gVlGcMzF2ua
zBeP~xgG7~Xjb^gR0p&G97@9L*;@oLuZq?+Kg5A;ZmU(J~n9rH4X|nfAYcE0J=^Z_5
zs6*Pi1<$mr>rtZo0b}f=*j(6-NqelRqs{wxJghG?-A*S1j`9OPT*gv`i%)T)v`F>e
z;2`Hm{S1y58BTWnHqx8xRZT|IU#3O&?n7mks=X5zS(Q(cHyGiT<j65CpJdljJ*rrl
zTEE!v_o-L%1!M109(I1a>z>^A+)1xWD9S{6ETu-e^KXHMF$%0Ov!V4=`7i6|<rO+A
z{YU+>Y}ZER>l%CMv+^hxz+Gm#>ss58t;#~10bkkxzt*Y?9^A=uPWpU1dh5nyy6E|c
z?81pUQriv#;HjIKp~K>TU|{HKw+wyO{#^NhvF9ZI&-5ctO(o`|Fganj?B${Nsg!2g
zY#oBaalShnX_*WhD)n!BeJ5T#|2|&1#W>yhh4lI9qC5)mWTC+T-%B08)09BpVWJ||
z-`eD~@j=9ism_PScXSq>%fk=QEZYxjw;63eSAGwhG)}WJ&(3U1Hpva7epiqih5I*x
z`X_>-em+KIV)Y*~ME+gYw*RX;6?CV<-k_t&PfAY*)~O!%FA`*vnK7Dpb!+1mm{xux
zvZDPDIAO3GNyS{MBK)CgWu_X8KEbfXBXdR`1EM^Dga|O<$&LSqkok`c^uJLb0#|{&
zuW$E;%k^CDc^(g{PkNes(#116SHF?OSi*`|r%)T{I245e-!PM!|D7*vpC*F>o-f0h
z4Zl)jF+a@btEDDN#8|n1S-O9>zvtlp+dL3_emvJpC4SH?fgmIPwMus)h9yrE`Fd@c
z%!!U`sjHh&b+TO+QY2*ke)t=XUETKoHA|OjLLFJ!gH21{WyIsUu#sr)^jSC!0n~bW
zW(&&Bpfpdhw|yC|3Bl1f+JKg-2x0DzNZa~^`!S6&4`2CtQC|T(`R`w-*}p2u<4g*?
zqvGu6?A^AX`f#|KskxrCelPYb54P`L9y4C1(k5qB!=@i$cU-X&Izuie5;PKfs~+!e
zckvE-5S)tvyMJ}#8$`Jwf`_{-@17o!bXi@6oerJG|5$vZF^CvhRkp*Y$7Mb*fzr0}
z$1trPWJSu6dRiB8HYU!BCHDUMe7s==3}$Q^on@21%57+Jqu=h*A3Jp(dPH#cbMIZI
z>{FKCL`ey~!pHRilRbNFlxMPTu_pFGr}L%{-fTsZc31JG=|$DV1T)bex39z&N*^e(
zR;to;LaY&seO*Vyq|fJ{B<Xs`KO#G*e59jtZ!cuzmDf54d!TG1Uk1GJ<LNRLSO!Gl
z_-n{%q$W}Ap@oEAiprs)1-Xia<t;!ezOsCU2M$4_R4@|N0cT+wMsf>;eSy2(;K*T4
z1>D0}L^aSZZKNbRC=>Ig?d~mrm;)|k_=GyI7V9JB<X@M~<d4oRb#uko0aExA*lJKJ
zlTR2=xsT2P@wz1in5nBVAZK;5=J?Oo>`cvE!Tu+kL?=NPrqcNJ&sJE=RbL}P<~meh
z6^$utz`tsd(-vRWlKwUFAqTvC5a7xEG0Y0lgdYXDy09TgPC73%hUrp?HRH+1b&9gd
zen%baARTmoTVVL$2#wW-)WeWOxB|MoX}Ja3F-n36IBVn^P+ZNC!?^_%>)Vicx4^2&
z5*4y<m7K=;yip>->)pk)EUGqfHsRTXpoh*g>Jj|h(Z6gu<JS9M^y=W~9EC_~(Oq|?
zBRA^y<b?Kq{a|u=nG2ZkO_l<WB^XU=v20E|P;8=6)%*RPs>4g`9PT`v-_QtM{O|uT
zAI*llq4XwTx=wGC#Zw)b?nmlG%nUYejSU@rQ+Y`L^&#@Ube;KkvKft7Tl1RT->GK9
z&lvCmWlDHp<j6LJjwKw_ZO+(nu&*9=_}F7WI3~L6%Xz<O69Z1>*B-zxlsckgL?r%2
zbg=B$2$t$w5G9(|A&Xpa-{q{-T+O>#4Lp?Z(0+%F`OxdELI}<1+|hPQf*`*^hTj6Q
zBlr!Y7LM?}b?)$Hf0#l--t6vovoca?=c2q9bNSdJ6@7{i4_?-TsB2xg*VB<Y-G=Y$
ze8mR|_WwLhNyuU~3yrf<;R06U^rJ`ct41wk5qoO^IV(*qRGRQkW%mk;C!7wpIuCu!
zN{WGy6zg8aM(3$_<L};G+f5kWZxMd*)#`UU@-ld<&QpojAvX%(DROZ=sF+M2TVt4f
zMi|)zK#9^Zi%-&Mx+aFfpf<*^vt3ff>urKP`ii|^?95!&sQ%=%^jd2hx>Zo~lNarz
zkxg9lSJdG@xxIxl1h93$PdZ~D6x<UP^gZ@CtPF^stk*9*pjf@0ma<^Q)#2y>kp&P1
zILk0`{#nBBfp2$adwuTq#bclURvLzh=UwY)OGX3;Zd<SzzwfS8A@ZI((Hub41sI$?
zOS5+S=6F6Op~r7nH(I8rmWO4a24s_GoDI+VQWU0)pj7U)81!ouDaESnBuiT?td>l-
zvp_Li^wG7MY0X>#eKk0L@9@5~gpG$H7;NL(#d*PuW3gHcZtF5J-%s()ig@QI`vcKR
z{Bb7Fs~^|aIEVoklv7lWa$l8ATMu6DVWb{^(m^S8Mv(S3vK_!1;5^SJi;7xNkawD%
z19kd)>y5b<)U<^6AODH0tNH_;Sh)^XU$?+og+*B?BFMw|QR1trt!>k@`H4rmuUJNj
zAthyfP9)ZUq&5(-HCSh7G^>iF<#PJ)X*GmS1e$AC#H^hyGS^s0yXz1me0Y8<@f$03
z99j6hE=4^@4i{T!=U5Kp@7&d_A*f(nB51i=MS3hjz#Xy|6wiaRm(I7L{oFF@7QpAO
zoA4sluGPMu!EyM*?~8{YhZ|Os)JmX2OX*G<7Wmt6Zvai@)G4aWi1?(}w~tNAM)483
zd@s2#u+0Pes<<kEz(<J-oztDyE}u(g#Z|f8EGhhKqQ3?F(ys(DeE0{35gXkP1I9-k
zbe%3D^S=CiArY*z7(b%cZ9S5xvu+MMP=-0RvcO40m%^d)F+X?RyT=~QYpPJ^!^U|P
zd4Y2xz72EYZV+`VuVdYneb`}P_4mq9tHxGQ`sk{4&gM37(#_;H!LZq+`EzJg7Cx-p
zi&R}N^WtejXWjL;hWhw39&e8iLDM)h8H~U+Wz<#x2wCy7x0SkT>J+ZWZKg@(VHN+j
zUkAc^(@P3&>y{T<Gh1f3-!PJ8$;wD4QJ~tg&d-GFewoGgRycl^e&bJ|LKFsB;A|{M
zwH<qz(A_{C+pd_s&XOTt5Ju={rH?SU`T!$DRbpGKvVrNCfXU_KCvi8>E>wEUr~mG2
zpoF5`@Fn{a^Oe~xu=BKbr}mF!wLL1qjm%#Vdiu@ebPni4q|L9~yo=54#K86Vxz>9~
z(e(y7oOK7G0&f}f-g~Xlo}UM<XAWPST*jWAAH^tcJezLY>xACWP9o|%+b|-&(|cu@
zN!=lP_Z(#VqgsAp&3_al<O4wwCct&TlEWcetOJWpm?~Lp;k{4M!Q|f_{2q)l?l;g&
z)NIFY4Y)~ehD=|XEs;-duwOsBQ%;gW_kq=Vs}VGE3sf7mIq5{GbS*KlvW(l|=ueDX
z_UH&JW3DWl#bvuW0r0k1w7eS&cuc+Sy^Lzuo}c(IiRg*uEr87yfaSc`_95%|soaBG
z;8nlkDV8NK>Jc(PjtU;9)HK14V=@7Q%pNN^B>lW^jl$<%$Wpd6fN<UdiLY06uJM#H
zF+w6>&-nwgZ~fXWe2Lm`3fo+5UD)`Muh{47zMJG-e=y5lQ&}91W4d5r-eBwfy4oM(
z<^2uN8bLh?$b1_~fH4ax06Bs;bqPCODb%jTF)Vh<=~wJrc<jK<ei%pFl3hw<K!y?m
zWrQYzHoaRp$2O1@)~hu?xL<xWlZ+uxnW6A`)>LopX2@(WzKy1Bsznwe?9CB|)dXL<
z=1B8~?j(^}h?8*8?+pLi%4!K3OS=V#IaZc$s4i~-II9p^TImiZBI`yGFtS8CaoSXV
zZkX;6=E%bMFgpI%OGx`$EauY+@6N`)7%y0)?=27mp~>wl{&k|sYN!2^QgMNPq4|TZ
zLb7KMt03GmqQ>VG7H&b3-q$V3`;)m7mxO)pY0i@dX&(A_8<cznHyN(>aIfL+7;Xav
z^WHSz>nB$Hq(6AC_dS|#cyMx%_YktZXlaC~YSGRW^ql!#%G#ef_9o7`CnTxz?mHIi
z<&N)tmy*1{?Y`wdH`4oD5YtqQRv$SRUu&x({nE2S`edi~PH*nK&n~SP=vR6n;;3s$
zQCr;*!4<Nm^^p{NZq`NuzAO5;%PDY1_0mJ91u#Jx!PmRdSj~tJl5~}lVfkUILG*_{
zzdH?e@d=cqq<7V`@qXw8$}ERMgjZNB4H2RFl<jjF?<<?;B4>mnzO#_w4?0f<+}TYJ
zgTWP?hb+$jTpfgtI9Av?ngzw<{o!0O3PtfC`oiVzuYCLnr2+eU{q$Q@&5z=alBSi2
z6=RAB5RDX0I8tIt&*~l~Y3mu1IsVN<*YWZ0^;5;!%HwN#|FNc4#FHgt?V9BsaQ&)b
ztzyW2`FdvWp9xL2s=FdbcD-a+(le8POtNAi9G-~hU6yt5mS$F#V1dLPXSch?OwQ5!
zIIlN9o$Qzk#VW9ab-I_Zr)si@kQVZkmJ8<EoVu!a<2g&RSpse19<IwvV{mR>01<S6
z8(x0{F7Syv_;?Gjr-a@DmgToV?vSr{ht4G@nieBls4*a_61wqxq;2o4?bJA%%t-wv
zT~J(rz4sELM`E>$`<Z(ta=t;i*ZX?Rsjd10qkR3GmmBlWxx{5(n&%r1vn$(t8o#yJ
z(lVs19@2SD*7<sB6y1ma#o@jiw=eJErDt@Ngd&bb(zV7;rSmer+o8*#*0~A|ieDNR
z8;b)C{BpqgCD3LkEqIXoqYmNUJn}SHV?&GGU;-mP()!sk;r1}o228F@1qcNwqC8rx
z&%Gyq!{Dip1w93FYrkA$LZ{GWCg&ltLjSe4xiZt277IOvJN7?6D~s|rK5F2d;oumu
z%4d~k8^LcyDyk{4!YdN;(mwp@oRanAF>g9AV(5J)V>Gl}r-KTIh~EMh<Irs_%NIt2
zw?LiIzn^*(D4!2ug_9|6m{enOsg`enDA*lDUq}rx`5d<r{N~#iY7_as+3u9VXHyHN
z@+ZIfY#@xt+PF(r3}El5#q_7Ja9$I_2RG>;i@!NAhSen@!?u1NyM-=PP`!VThF2h>
zS%U)=%5H&B*%k6DS{KEm21R5E+zlVIT!aqnX}=V-*PR<)&=FEU(Iisn-(tRTHbqnS
z^}w!JRx!6|!Q6tW!g7@>x!9qtruA!~bG_AxetqI9cgB-3H@B*O&xSy5m@jY@vKAIv
z0#0pj%uLEQFGwkH-}KRMj-l7C%9H$+osewEOm?=S2+u|HKu4DCZ-Ed#7+TU&e>w=|
zgj{D$9KzfSM`qbMN%GV5!#LsBR_?XH?~7s9z^lwaR!=whc<;2u{YLm%eK)C~w%wF?
zinCscxvPQV`Eb%Wn}>+#gub{?(2@{5_N!0%A~;obd^K2PY?x)Uvyy??omP{Z*;@*x
za+85(nR#isw0CuwgIBQjV}{`sWu!@WKjE)%?u+sh`|*Tn@qM%id?{(I1$$j<dWCXU
z@%OWXZxSplI#!A|5MBIVzLK3n5hj1WlI7$h*Pk|3eYWNL_@eO@^cZAmJ>(4{ZlwmF
zbXmxZi@QL(4eJ|S+P?g0Ppzk<U8E|yGzSz#aRylUe!B>IEq|apC`U>f)FbdiTy3$I
zB$t_>l{}!`jXta<w6moaL>d{A3%=02wn&hek7&PE&Hq67T4DXi>FbWG4PAWY@~^_D
z7xhy^IaQ79i+PRnRnHf2@BNV8i9*~+t~_O=>;B^6-IZQ*E+b`RUt{R98nKGw9eshH
zDg4_*1{PyUPTp-vrs8}DmEv6&J2t<Ob$=$>N=g0)i8O<}55@o`Lg>Vlp)|c#pfo$x
z3ttF~dnO7=vEv#Uu4E*!GzPAM+VSbu+dj|*5Wx?6VPz1dtiia5^_L*XZ(sH}(c&9?
z?gCFJNdP6r=6!eRCA;{{lto6gYRT_|WAsplMEgrf9zC}_cnhBz;!**w?_=dNSMO$|
z&GUZkEUY@79jh88j%)X@o%NZtYvFvfZ*o@8Fst$GyzzD$33u=meYgAL6z@9*N-F@h
zdlaw^5r&_ScUy3^!O=+wcEi@{l?QcR#+6+@_CuG!>sh~SJS|S(-mA5@K%o7k&4l>H
zWY1b}38RFikmH*(>uK*DOG#wmGh}!LG8<X9+>zSIl|ymzG)2Pmv6hhRU;}5JOgx|v
zo_?jjV%cqo3|W1jay+NurNw@)@iha5#>{F$qlYx#gg*f%4)>Z{kIo&iTla82_=|Gm
zgqzE{AtkXCmu{IPK>S0o0AnhWvw)2{)7v5hAzFD_Maq!vsR5R<=a(<-xx-)KleEb5
zEh~tC>)S1;rm}Ttz=kaP`c3sG&!PL*KCxJHIHK{3El3y^Qg$7LjkMO8$U|g>$5wjE
zm5c~i2Ct}PGe`HgM<>k%iphZx9FIm7MD_x>>c(MVY5v4rj?TCgmr=j0=ZNr$@Lrs9
z7Zo6b%1}y(nBK2N;E$`Q(wo$lDUF(+GitvH$vRGg>ZNfhhUL=dkPlm#5Z-+*c7KLd
z>&$7jG`VZ58)`nj;S^K48^W9K8SOjmvId)@>-2Q=?M+N!oLkT&naYV4b|HCD;rYj&
z;XIq1cFgM*VBAB?E67yA7a9*XL|z;4+h@+<5n&e*cn}2>+==7n=1#vVTsexA(^>N;
zI^XLx5jt1h_vVtyz2^V!7U%SJCO*-J)5g6AatWtfI}CwebkSY8Ry!9jXv?quFszAW
zc(kwDv9S01#q>BD7hxV3BPZ@szN2T><NVf)*Q}&8rtHPW1CAp3ZJZof48&kfCsL96
zU5w82mv-Ob1^MK1k+fDp3Cf(ba)bU23|rW(gB7xCgPe*)C>>^vV#b}Ml9Ffun;GE2
z`PpV2cS-Gabai8SL(4a!AdMV@;vmVK_ee<9b`5IktVAegC{0-h%hX%>hHn9DD5mrB
zKRY)g&sOHM7?F6U=HlMF5}pE!%|1B4cp0A9T$M8W1u3yI!G~AMFS@mPka;~Oo85ZL
z)ir+VX>v1COTYSnw{IVM9=uTCL`hBskSx4JrgqM49Gj#;vP4}6?pYBM&GdV@6n<xD
zXKx%1!A$q*G7-|1oCG{IcRskTEkNSw!#}H}Gq`t=occ|8#L%+WGKvuKXlcBY6?go|
z4YHGB%qg7gNU*{Svt<~t>lNM>RRXC(BoHFVuJ$XnMha`cAJy!FZ>#ud-ipq9WcXP%
znrybxfz6PT?Y?SZhn(qi>~^n5`uEK%cd@MSztc=$d=~$?DUta7Z8gNCbmxJiwWjeu
zUC@#5%uYKsLt+(YU=`HCSh#n6aE)Tn;~?@1h~?@Re=YThpCvnfpC9G@N#}?elz;i^
zu0mchO+ShWc~UsLeLhDX%jj(S<sn<t^=mILW-K=5FlLgI%tDrSnYM^5to>%^wY?i(
ze~-1vLsWBT%5KMdwc=^T{VfCxcS;UK7o;HNU~<1@ezA|`?8{|Ej^@I)FVU6TOXT#5
zwnYDL@gIi^6^2YNXBXoNpJMC@a1>mTyr&fP8o7uc!Cw~oH6{Z-%vbkBm09(DvFqWy
zFS-7VE`D>sSljVql=>~;x_qTiV-+xPWm^mDMEFrZ5rheDI@%o0Kbr(8hg@cv%-<A1
zHzJX7=*70vZ@hWe0=K}xy-Exg!+`ozQ6-GATYI+f9U}2Pn9Eix=81R`ewZF}Ob9X%
zO0*w<^-{SI#R*?C<<S$vTkU7AZldKxgRijkAW(QdS`g{z?`#c9gF5W_xrUUA*xZ#L
ztPVPgIswhDuj76E2DSi^AS?DAPiARNog;&emqaw@bnI{whUxo!fpeW`iMvi9;%0rC
zLB3qa@n={JmwAtNp?JE^n5Fcw*ajZk<g^*ByHuJjk7dKqcXRxpjkQ&HTLi3jq6F_b
z^^G~q&U{g=lEf@bbplx=<lw-O8sM(;U7juN^k$_)dNQXXYP=gWz9I75gCkzxK9_wf
zx_6?BpLz=nZ9wOT(X-PNXUFP2mlbk2*N;uXnJ-zjG}HO=iF;2eBJy=Tr^EJkWsREJ
zdWEnBW5ub~n6yK^eK@#W)A~|lW>GviU2?{^0Cx!f4RjW|BWJL!Sci{F(w0PbB5r~0
zw9;NNAu4F+>uU{1-|f&Gt(IpT%q|znG>^F4mz|=or>oI3pkr;Z<!j<T3`DleCi}U<
znuYfjp41ijCeR0y3F(l~RNcU@i9(!W!%$Y=>n8%}z`1-yUHQI{@d`gV2Xmd37e+Aw
z!&&rVF`%x8Navnuu1>8-+6BK#p2%sR+_A#PwOPp|U@t_TAYdxZ@Y?0@=KJr24uw8-
zxqH&HOXW!|U~&B6ms$@L9d-g5u9+5op(&@!m%S5cb{nb2`LiB^E(XzFrz}2}`-`g&
zP!Hfs<n_IymQEr}1J<up*w}ENldIHVt6OOlBJ<Xtg|9BNN`?yVi$4%bd1-ZG^+NcM
zsv3R@K1gbqB?j<~!DNktxKIYSKvD??`(JBgvTN62Wu+TJF}|U#N!L$y=3-7h)ztj>
z{v$0og*QU)^NI%;Kk7UOak6@-xFIZw#=`uhdj!8Y9RZtbLTh5YzAz*-$uJcqG#uV`
zO8?~+0KVM<YtDS<g=vBc6smIS!asd*ihD%}taxzT&fj3HZ-Ld}f?@QhpB7&%rBT$#
zrj=!4l*njJjPvsp;%hf4E&95I_C?Vj8xQcqXcnNIXr%Wopq$nQCv!t{1<JNDLi@kM
zpwl02fh&R1ty^HGsW2LPfDz@+(-qt?H9|~C^6qH`<|e<-0tehtY7S`+d%2{#urm4j
zZvm^rOyOIgUhxK|#E=DQ7WIK&tlbHZlM`uuN=L_P^F{q9ORvH!^$Q!OxSNYz-nI@G
z%8NBv4~t4~07taKSI38`uLsmqK}`|%B}E;Is%PS7<dN?LWV;*h|4u5K|M-z95H}%F
z{Sg6MtTHZ6f*w-8E8Qqm>}5?k-lWSLEB?WtkCd7NoEeD$HCSuVIt`3DA*+moh7O}Y
zI-Oa5fF5h1xc}+OxC7(6!3c*9y%Fdh3d6{CFaspU>A~66nRQNrQ6EawSW_53&}iM#
zgQ){Ca#EzL(5bw|YHm})GlgxEoH(0yJB23qegBjw;L`70@4KlYJ$PFCbJ-5IKohM6
zGk<Jh$jo|ydNIVND#Un#pC8SzlWJ*`wkkkx_<M?jo5DTA;Apnhp}ds?qL#TjC1vI9
z{a%VF*O-tSUJjboB>j`O{`HNZz>(|v@xa-yimu~#*`!B70TSa7pQ7CYx%=O?FBlzS
zl2;Av?_Y%c?%|7REB65%U}1z_NWcSRu-|=R@`#!aDq(z9(>un%nJlJKr7zS5%wdLj
zY$1p+PYr5LPJP}aG6dx&bi>x<pyLM{+-tC&FXS|6i%PGkM1*(GeH{*e^Ylt5T);Sl
z8}aQZt*`hEC0)hDPR0oSm?icpHyq=)YQdg*6Il>`@Em7?K@$SXex*oMi6}*DDn>!~
zdiG>jyt|p8EJ!sfzFcPs-tSB8Qd`;a<cD>l`yR3{mO<jg{<5pw5hj~;mvH~ghxkSw
z0BCD0Q0gd<PS+HIr2r`j4c*bz5eV#F@ovYR&OotD3pVq?IeLB3S)?`f@9W|GkN(K-
zq&Os3iIA(K4Ss<^s#6CL!V+XF`uNYd+IX2O|LOJs$8gGunEI<zQZ6muBJ>dpEC?l9
zv>fo0m92j4z_zJ*(rwD&UZlsbVlf`~y|@QMz{_zoJIbI9-UN;s?~-9|6zSpov!MhK
zGQ8TtGeguiz0XH@M;^)9UQ1Xy%M`SHl7!A_k_lMNG#R9_Xv*I>Wrjl$qUuOT?htl3
z%deC;XLl!i;kM7XJp)p|MI<x)c_ziQb-HeWs>y?tT3h+#<@4OOWRU}xf>iv|Y;qqY
zMm=(7buOQ+C2ua*J>6|(hVSDa$xp;e?-yZV1WKrH+<)U}Ai622Xj|n+R<AA-&eUB5
zati7vCO&IVh_fMpf}DU;x9Cg;^ehsK5|`AgV73GNfh^Mm<8nl<Li=B2;S2S}2H<>^
z@Ze0BHLgihmV<}t-vZ1w6?NBrnRimA*0mcEf>G=abA;@Y4g_X2ygn5is~lkb)!(V!
zE^g~LGN?t7=1kiOji_CMK_1R%L`Kp`02{OA{=iy)4LJawS$f-OY<g3eNBigp>qd7T
zf~NGE)1mrVD+j_BZLxOlR?QjxgWt1NZ8*oy;_Jke&;jw<Fo6kDM1Z(GCa-!o729-r
z{E<h53hAbA;=GO=>RXHAYF1=3`Kl0(BjZDaMfe}ncDc3sgq6W%(ag3bu0VOiR^yxR
zwvlprt5E63MJdOkeNC>LOH#vcR!%In9T~sIRu=a77npbW*b71$R^DzSIhiS@<JDI!
zMNo#RXe#nK`>sZTJpP0k`DT%5#BpW1W#8+*39CQ>XqP^8A_EGyzXj%;f1xxHPxmZo
zPBG%rnqHx66Pj~V`2HphEg!-#e4*{%an9#<c@d6A(wyjYYjEhy6?X9^@D^SPVS#rS
zi?DT&PSC3)9W==;aINj3NZ5?2A0xCs5Pe<YpSqvrlK4l5<JXA<G`C;42}fT3j(^a(
zEM~ZmnBk6(!bgM-KC*WQvM(&$jCNwQsyyH?U+xHZn$!OC6%!1;&=V7?jH;u5wGtcF
zJie;l;ZN0GC4V-K5#Y}JFNxFQ??md@JbUXtadz6i)?!kLQeQKt>N5PcWicth&wj9_
zf#qgvQRW}W6{ui4bpTVqfmmoi?|onkugf>Ra31OI698TJzf~{lE3Rl;zXf`l=-@G}
zk*l(Cbr#O?*kekbo9zoOedY%j@6T`)CWlb3kPX%d?4M`~MZ)dGU8^<Kd3Q1UfYtto
z*<12HQtNV7KTLH<!PTp@^#$NB#_$azL&R!(a&!NhbuYswRIIE$hrDPMaHY5s*Yy!f
z3??o2^3$-L5ZApwLHUBT1H`XB!NZLZFtGrTdVhIQkl`;TYeGHmFAq>^p7^ILltt)p
zQpM^>0sbD%REY78C3cP6#wK)M>+4^Vg--UZoY(pO8VZxUhSPT!A63n3F;`078x6rZ
zWe?=+0WSxlA6SB6YrRdYGh*}l-;5a~mPpFJaJ<TsSFs)Fdx~Rz)IEGRju=dCI(cEG
z<!5({ajK=n92cVHep;wrvpYbo;dQVzzK%$t!n=8YU2_lCvLC6w3th;@F`E+CXmdK?
zdm8t=#aw#*`Bd(E3dt*3R(+{m(X(Q?*RVZ|E3xJl2(S4pZi=jFu!<OPNG1E;z<W7}
zQ*+8vO5RTOSLonc*C~^|ipyM7Tm1pgCIQ9egSd-GeuWR`r3#n`M^Zg}_{Ia{s8mV}
z6mQln-PJ=+n&73UhtBN`zagG0`^9d(`NTtCN-k)=*6&-EwPT^utM7GTbCd`7G~#&=
zt&C&wJv7Bnjv_J2@(3b}@#N**0*{BxL18F^vPJY*COaIJ{$VK3bk3*P;jP|Fb%OqN
zjuh+Z;%q$hMEw<4+KQoM`ui-;3QK6Nv6g(JT&19LLUKfZ`>7l)&4wi@3JTxqYQ0-M
zj#62<1rGf6h-;>w&zS!Fk-`c5o^7Y6nS=PkX)D)249<;9IdWO>xE&J?bCdO;KcQ}m
z<o<)^aXC2Sn0c~~qerT%BJCaRD|Y8xYOjWc9kyOKV=vf?VZq>SYdT#+`rRH~7zQDp
z-=-sU<{Vp=FQ~1`*9xBhN~W~!FIBEcBiX4br&uOm3FojBFB@V~KUcXIpVL9V^&ZMR
zo$l~(M?v%^7iHk$`0oCgxPXKM?s>GdKDASiCa%5ofw^LYqCJSf@rta#pDd@MW@>8e
zY0y?gvF&p;ce-ahIH@;Z6jyO>fe-2!ME9W|X?f7*qF{LrOZejxVDj!ndOqXYUi{SU
zgIlw(e9|0PVNWe_Y$kYQ$K&yHf$mI01v*aQiw;-{8VABLqQHi9)146iQvQsv{O}h+
zU1DyrdDN_ILaaw$i$6KMZ8aqZzICm0O}3?*b$~Yda^65uxp~yPSk$&G{o79$-^(n#
z=jif2n)ep?j=_spSQV`e<R6zY;0S|TU(I67B^)JI<7=i{pg*VL7MR|};G;@YXiZx;
zQyV_gWBA$VmV>zb)3&4d)!az;AI1CK-+}Y3Hu1=7W@vv720H%Z2YLb1nY1kv@BF#D
zY4yY^9hYp3o@6IRTx&qHd6hBtpDu0|X)EARy-hop6=U+}TE8XI4ovG2Np|R-EN!=y
zlwp4&wM?%-Y>ZTYoYa$v-PruSahxDs1G1)Jw$dnl8N@Jvn-P6OZVRtp4-XR;BAJ~>
zhM@)c4C-R6lI?NTxd5igM$c^@=CeJsi@vxRR@O$i|9Q~rm*ff8qol!<cett~K1>aY
z;Dq)*=CgHbEKKzrHpE>h7DY&UaIRQ$U^cfH|Bz(@8-&Iskn<k$y$`+x_FGo{mdPM;
zK4((tOfk<^7Wtnw8C%cMc$4b~2YfL4{-0k0X+|F1N`aK^sn_u?cGRNTOC0ue>gD$E
zP3a@?@8@dT<-zGaJ(?YEpGsADv;Lw*`}%|aLSP`&d&58q<vdxV4SF_Gv%N3o$HQN?
z{(B}zsf2Jhpl}MNr^LW206)xHYa=(`ZtRpxku-($sPjIXCX;5}<yFK7TWZ6@!dVk?
z`#9h0HOZ94ng#c9Cz-y}#j<c;w+vw=MR=X2PS;VIEI-jqTpQF8$XG9?=hcC%D>T5=
za|7l2Iu&tHbgLh?VJ`cBTADxNNxS8rNPL$jp#~a$OZU;QwTQpqX)nkPd<d~!C8g>$
zv7dn`TeD0&(Tr)`>-IMY=eU~Abc9y-1?_uLy+iEMMaKVhbC(hhR$K(Ka=2Q;qJWk8
zRh`hbRpF5C(b)mA0-VRA1U3&SeH@Ci*;arOJif@A?}{7gj&0|3aeqlLi%m1hUh=<!
zuf6c0x`dm8o39&-<<nm_d}cdk4th7yXwS?EtUj>HYE~G;6JFh!<UGZ>r2d*wU_#J3
z)h&l345aIx6tw^kZ@&fZAg!+WX5MVIHARFlrydjj8)f}_v3%8VLrjE-pf6Och3-Jn
z>6(O?{A;xx%6mR3?$^{JfD_d77PFwWR?Yo-V%D_PmNskE5~aP%yLb<yc`Q>4RcV&>
zln6>_9>oS)rC=i3!_C(4@Xi3CY1_%OCet(_8v`|mLk2wRS8wG@>Cg;E?TROxez!o7
zn?=1U;KPj)T@Hk94PIb)d#Ndo1}v(9&<Vdcau!jL*vTGUl1TF=yM4n*9n%$r5}giI
z-0D5~t7{UPVXGZ<vi{GJnp6RIG!@c28?66$qUI|j!+O;O4&mgc;4>q~4tc3U3?4iT
zT2ZXW%%v>u$W8svljecwwAm(C(eb17y`q1V50hE;wc?G+dD%NVuF>12FgcP`Mr`t^
zP{iiV)z!!g&OaV4t7!s$I@%L;+_hVHRMb9j)MN3_%1;H?@m3;m&Ul}mc9HTc@zfK6
zXnmU}xY29UZ$+p#>{-9|DxVLBw^CHWbwXXlOmUhYOKJ<7F%&5J^dDq)eFxrGV63#`
zxTsrz#BK`m!0Yu&SlyVbkeh1HG4H-<yN>J(2`c}v?^&}w%@j%+uJ&ccl2j#sC&bhL
z`eT?bs}H%MVn11DV+jvVKSUhv3nGP>r*>bdj@FGX(Y>~3=)Lb59F9)A1@5v$uqN-C
zRD5+Tb*z^OC~!@c@FuWM;bs=eob<pvfn}^B*%_1uDf@J^nL{A?Yr)4uiDx8ED}ob5
z>NpO=xSpALG}{W~Bd_7ZVeeWPC+k&hUE>e?mf0Xf$c(JsI)`i$DE*ERCg57DEz^0h
z4-;gw30q}}@4b1k>4U>&&r;fMi^N$wc$v@2j_5Ko#wXFsbi6zx+8|k%n;x6^gr(Be
zs9hJ5^U_g5wspuJTbaQ+@lmwRb!P2GCLvPZKM(S1qS>h2N!ZD73o@1XkX=XOk@A8J
zX=y?q{@Y&Rpo4e1_z{Z%Lz#CQ8q<Bmh3?`d|BAlTErpwr5wMFfFvD!%_BXiA7~u}Z
z=-G9HLrniLXqO0B6di2AhDor)Sd54g#nm)928_c&=q_{_2}Mr&Ve*29k&2yDOLyU+
zbwPP%&qn6?gBdCi9Gi<-0yR8Rc5)={m(%GGI;2)HSgTiND?(1&`_38a*H1UG$6v31
zZ^Z}W(VzBRFZH%mL;-vQ&HC^zy{Q1<9afQW-P+jS{avk#T|P12GuETMVa6C)#J9H9
zxy*>n>eYA3q(|IoiTUuox16LtsavI}gN;2Q8f&@2sjmYTd|^S1;J-pQmGnRZMyuE#
z(+3V+d`)ZMF4|d`uG3RuhV_tg8&5S;6dNP$uKZYD6*>FdOZ~~OjQj0SI3Xb;+J+HB
zFe!Q{b-J2(nb~>&MEYl!!_(Nl*s8XG<##vEI~a~igS!RENw3U9g}ScHBo2+7e=o~b
zh-Rhz*4(OdKQL!TsLSb4_gAT{lyXFe;TN%;QD0%$P|hAs+{wyjm!%VHPPu>5ILxYd
z)e8o}S!#bnM38)ILU+LE-xt?Db|ZO9#{xBR#82im48T_uiRAiYt--m0Qi?IGgcEZ9
zt2Cphi%*y+xsCHrbgKrbnIEuiki`44CjSDLuEHX8I=*suD$q^j)MgS)3Rztjdo`uu
zp=?Nd;-0CqPVRrw#ixUpuA;Lr!3gKV+4?MRLDt>vB2v*?K$SrXG?Csp-Acdc1Z5C|
zF>w`;yfx;1P%E4`RP*q9w*7j!0OKsfTDPPNAhQ7f>@jR*AE_NLa1wsRE^WA*e_(4X
zt^b?!B~liixZWXy&Vf1m4ouei6+vRIO*e7Ftc)Z1HFZV%fUD_>Q6+i>nKABZ6jS+*
zp*E%2((F$+rY9Oozw)s)aMv^2r$2$|&VArFu<*6G-uDTX&R8?eQ>mT=gIk;Hn_H}t
zhc&wRq^sni78LNX%A4YB$6*)0%^Hi>KN8S}y=n{*`+UU+n*4e+>%{&+yQrDTPfD8=
zR~xo~Z#6vG<h(bTDAF=8dFNGF7K|8LqYbuK!Tr)<9wm?e)_IZ^BY1ej+6v~KeZ>#c
zAZ1gj?qZZBbz=r$WaO@-mbrn_ZUB-=6)q4nSs1>S&f{uB+0XfD-~=lcJvpvLsZG_#
zTiT0x^(2^A&z^-5(n1k!81oZ*Pf|$keX6fntcP0G!^gu_4{bq8Kxr!#60{OWx}@$+
zn|1$Y0{q}UD%0V)Y$zb;A)4VnPqR~_`;UG2a034&w*vWB+pyzrdj9pctNHt#{{Q=(
z(jfW{Xb?>=n-Jn#VJHD_i7$nYV>MI@rVW~|_w~ar!+Eo#0ck1DOeK(vtZ1bjXVL&y
zpyG>aHRt1MF+0qslmE%5ZP?%-56;=MC{$T4rpq?$Cm-L>FT_78%3QhX8X^v39nA7_
z&Pb&3T7TSpRuHvB#@+&G3L%=`(=tKoiGye1Lp*v7c@S@al02L3Bs$Qa@4pLNL#h#q
z=bxq_itAsE)ZplZZ%o<>cD0T|$xE7?k5nGc>|$>PO3T?~zLikGLys`l$DrKF!yzgK
zGu_h(4Rz_=GW#}a@5xx^*-A1fMq<OI0g$>LCXw{XO~Kg#aqm@YLbfC0>#FfX){&N7
zjbpilN+Y-A39R_HGV7U07V>szTskZkh0%Hhw?Kqg6TQV6SK^%W-Cde^+E=TyRRr70
z^c*?AmGDQ!j0P3&-U5zsceh)`r^L#s)W=$TUVEJ9z$n!d<|KHBlDz;*GCUb0132+|
zT{!}$r+V82on}qkjOXlAmOLT2+{`R1;X%+r-%D-=$2z@Fb8HlY8SjX~u~lI5jhHq_
zla{ug6xZ&6)5rAnoF?yzhccUWpt{MSqWy9#V_~-r<<fTg`4?+98PhMH?^lzRhQ4BW
zDJrS>umis`o814uRywuHO({ApQ?NYW-kuYTPxt=(nD~97Sd%Or9~;9Jnf%psNx-lo
z!x8$vrBk|H{kOS!OOswf>y+`gF?s#Z3tlGDQ)O)7vW74R`uS78F!KE9Y+U%t*h<ZV
zt{AY70#c%o^LjU>a?huy@wG>5G!95&Hp+JRemsTKM*`fen72C%wTFK6UZxT_1ABcW
zb~e4nj%%HM8#8*0PojxSBfZ5O1)M+m$F4hNM-;o24hpQ3;(ifJ8e-SRt{WPNh7>V=
zrD2BlY*I$Pb-t{*>YgF7Fj6KZjgOR3h+n@-XDB%t6p$cgivf7}x?HOu(tpj|`{g(#
z(=d7G$-?~{1+#66#rJre#hrJRIlYIa0M>C;h_&K@(K@+wb2+GMjsI)O=u-{Wtsw?f
zsGm?5DCRjhY1KPIPOHo@BDImR$xPkpNuSi5@q1TJ7&CKH|1bm~cg$^ls90+mUFDvK
z1d|{>V=AfFYfpvREhA1)Pv9TsVs0`FQ_t)1mfXo8g;FF8%AB$E*o1OXz-h@pD6Ozw
zbNk2vmkjc<q|kaZ{b+#h4ez{nQ)6zhAL%a<f6<N|B^G%a#M_&aX2)es_{GiT7fhAO
z^jH=y3bIVmOiEVmZdd1>FgP_$ku%6Z{eVAh;{gd7IY&ZLbUfMlKMGU%qL4l7_DJZ8
z?4k#;vbRUs_YS5E?$29TUmEl#f(Cj;MB}mMMlI=}T?WWubPy{m#$pnMBIeFptk;cZ
z(05)U4@vo%=Kn~HnlzygxJkB+BQ~;yE891sNuoCZKM4oe$F^0><6&IAS!9^}?zA@K
zfgh#{K_W(W<$dycsWwmTpF-~0dv(dx%$^}^R^7~>@VGhXxi^_@;;{~>je0OS&y$%L
z@Sd$6-gfkB8TG!k9O=FGx!HN8oUi+#seFm0npyikeQ!t9_XU~g;*R$IKAZyJ-MfF(
zq*#gvOn?~Nnq?u7`>V!F^SUoIh1{sZA|8@d7B)O5RQQ0k41k`DTBaD)drMwCU$ivS
z_Vna)O*GDEYB=LO)>c*mm89i~UmB-l@@`7r8Ex89V_g-h1{h)j>i_Lm{p6_l7xC(B
zu@hKVd8!ciYZ-{klnX@1ezlEeS*O@>?!!l1Iyy27Q=gZs0oFesaO1M3;oil<-}l8-
zTww}v{u3Lj%#Oe3b7J>K(o$j+gqPJ}vO4S)nL-V~`CMj`<2{QAme>)y+vO(r&9U?8
za;F9JxEx=OsiTZQmFb*PHlzecy8X`Q*fRHl@+biH{g?~0PvLXuGFRMx`OTTa|Eu$#
z5slWqB3s|o_|Bd*kDq^lS0t|~hF+as#r%*o@%mJsQcnORO2%ldP?8V6Ux{7G495im
z{ER%|;!!9|cr2M{QMaPWYUvLo=oYx|*Qt0;8GZ{q$|fHj{6`_-A5gSnjau8PBCC@O
zZAih6wwsk9SK=9G!XhG3txuN@m&5}scM`<i0iC0=K*YtBK@?$*z9uRV7VDG&&X|Nr
zt<8E_o7WoaEUJLNxGJO;VR;^hskv$PDRD^0ODbT}B9sgJYrGBj$*^f%d?h@>b?Ge}
z^_B30I~CyRzNt(W1YYX5-qd35B=M;JFp0<ZvLQ`2ck|t&K_&oPDAt#&;2Xpkq10NH
z394Bh@*=d6N_@av)tLF=E)<(UE*O)7|Eb1Y?P#4V+5b+)kbPmFT&k!`XE!4Qg@=yE
zFM!22Gce)8&XEV_Bx~I;uGq1Srr9uc%CpmcqCzlta6|6-lV7m}E5QUe>bk1$m#ZY=
zigXaJOK3jnx!;3J`tAyyl=9mK$_n|XEYAqH$rRQw%32C7>?jkKg;42v;Abx<Bv{}s
z<EnC^b}Y()vv5}#_V^}Ftn@N~=O-j3iAmN}8h4zgGbezWYWMA1`M832%Xixxp_C3w
z;#2R+=`==rl0zp_O58XSHCghk%DT&<9%@L@?<_SA2P(37#eu-BF2vo8u3w4e8`25i
z7>qkwHq*=2c@XppN3%Ap;A|d6rmLdU5cu5^^;1b*Qmtv{A|3`ARrMpqxFFxNteN2_
zZEg`4S8=m;v=1GRNRnteOPg{GhKq@olbw;{=U)*DU)P3A`qan73@U3I{J`jGX4t)%
z_lZyX^AZVnFroi)Z&3A7?~U~uD9TOD9Qtg0H9;Wem#8scHkPppAD-QBZ<4#}Dtk&G
zrGML<2J6zV%ppDVWAVbyRYCOmblq=^H|?20__lfuGetivSJBF!>jt6)+l_M++36fH
z_0LHquw5$uPBjrI;y%tH)jz~ydR~T*bu!}C9IbhIdb8DUTD0l&=>3AB65$omna=Fi
zx>^AOBWMlAy;>v&mv+ZMwdw+&ZPlKJ@*1SYdU$4>?hE(hHyW55N`BF5in^D^N@u{D
zA7TmGwXC&|nnSPbG)jcCWS(r~Bc2@JIcWG0NOg+c;(h)e*;5SB$}*~&=I1{+;-9qB
z<nQ8t^I`pt@#jPQf79L&ZU2I0$LZ=hOXJ}?6O>}ld$*O<@Nq%CtywpbVofuq-B;05
zV+Egnk_i9$8S4GKC-AJ@7PYtQTUxRt__bikqQ4+<jxOyJGbN2?-;NS(5iD9r(FsP+
zG>|^5F+w@h<fF+iD$8rXN%A<GD%vJcA-E-ECAIhCFcVn6$0W9;(6q-hyHn6|-!z#k
zIzJj9=?=RYWpVG%CU>iVuCc`YE?yz0ay+MCfC--dEMkkeOWgE$ur#TCfX&mA$1Gqn
zK-f9bWL^V~dYhtPqDlRGuwjMrK9J$vKAa76zxj&hTY@|r7k4jxIX}0{=-k&*UQdv(
zDS6OtrNQxfn7i^E01=wKT<nlb&~cXY=<P9Q6=8V!RvAABR?ub-B6Ai&(IFEyccwuI
ziRzT4K97TItq9vkRPq6K#F=URzm@x@{tx!vGpfn1-5L!B2?BzkNG}443W)ShK&8op
zNRbXv0cj#lAV4690@5J}C{hIkM2vLl9mGQKNK5E7K}vuS&*FLCcYk~DZ+~N)Gsbzx
z`E&j_Oa@uGbFX!;HLrQiYhD#^aDnDPo8vs*RwCWEyo;GEcXqeZ4E@aX7#9qe$SbRk
z)YNfZ&@#N26dyvnLW|9#KuK<9UqE>)zmSD@t7;*4lwF?t6S@xLH5m41=OFVSXKTuC
z5i$Sulu6=j?;E9?>I(cm)cZLyYiN3k^e@ueS>Ff%LzeFrZ-jl?>U6%XB~NF5V?)#!
z{~Qw-Fatr|7#^W-Rlh74LdSUs-<2K^e;mZUBFC284+ISR$&+ceKhz@g!YkGJLjAPJ
zGFmadEYes;+t0GUGmnaG`evA7OgR@@PW!gkyiYk(*B2%9c|t;a;qevg+K8*F_ZfAT
zf?iRrzcOui*!@7PT#FHJ&5Wu010nm@7d1C>It&G!a#58VEdSClk#~D|YocjhFj;?c
z{py|DaZ{hXFxUNiVHKunL4AsPyHx~-wq!-U@)Xk;Zpn`l(Pgv?F%!`0@sI8b@=aaX
zw9fT=c=Q2~a{qx)EdnStw2f}UzBB^m%IZU#qWJRqkon?kECQ%XZ)4Xkj5f9WOyctl
zJxGv$l;0o7Vsz$S7q%L17YL59+oZqp4a-vpC#Sl+AW1$cFF%|_c{h)HWaLo0gPOum
zM!5Vat|kiA5cx)O%ZjeFUc&jNN4~~ZhbLw-ogphrSU}s)ejtMjQNy181DR1i{=E&U
zBr$wM5I>1-z^bM~*Ye#u^)kmr`cIPI)8ZNZS+sXl9HBMX3r8QyzfT&?>}J(?msc^l
zDVs`)+FrCZ)Rk3%%pKZ|fTR~+GUCw}aL!7ir)e7A62(I+N)CA)Wi!W8G&lO^rh6cN
zb#b2-`oWF0a`hJO<xabgU%fihAdB2Uga@5`YhGR+yFtveM6wv(fZf-A)G5TkOCvQP
zSPTjB4a`9A<Ask1F3V`}yxGL<-Xav|Ec*&<v}{jX->ecp$1ozKByNMr7A)xDzmKP<
za29od%qtyXngZ<sVy&XFIxS+UCDSv1tyov-f}fOpxiNvNq^?*6^;c4|*KbqiUqWFC
z1E~-adTjs{iJwAmK1XCvL!K5Lot=e+t;x8UyNcU}j?KJseL5-fg+Yn?`plLhH|IVU
zd{ZaifBK^FI!ogFK&hG7ID8=dcJ$OrnOM2W?`lU};@;agB_ZDr-*;U1<GBnSFZjn?
z)cv=)C?}zD8btK-r|EuTKR-9Txcxw^_CXi5L!&^jA{}|2w4jSZ5S@fFNw2>t9E>at
z1$m-8fJ(6zw|PkWoxB7M%NI}%r~TO3%P`=wEx3y!WF*b@gefu{c$XG;=h(B(53W&R
zv=VN{YeVO`KEv)SrD3S{4|km!|6XS~YXj&W4)|swqHypoY(8%6{XeGjQe2tdABeIY
zs4CJ6%0HfI_@lQJWEXG+OI7_TM<ymx_GJ#zH6TG|N2zYdQ?=!IC_VkVL;rAYWX1a&
z)xHf;hy8m_!5+1-_dx)zhtVlWr1#U)m?vY`L>CROtBJ;hKnn%KIA%r`q?ZIUZgI4f
zK;agUhb1NqxnVH=*r9`bM3VPG;lG{m;7jrrt0JS?f%lm=tFG%G2%iMH0iF&C%J0*-
zH}r&a<geNcWw%H+@1DOxh3-;$7mQgne1BVu%;Xr~x|$YvssLWO-E%-|4nl?<pqnRu
z8`@E-w>9|EF!+Qwpzzn%O+gkNz(X7iU4<OopiB(Fmo9*496dsI{Le33wzeW6M>i=G
zz-Vsi;y?btxjk>Ec=mylH^Y3;UJC!?Ko3N6(iE9tcyn7na*EwmaTdpB@#1`9mhj_*
zue0peH(ObxdS`lJOZA=zw#K<vq)tjR8~Wf%QSm2T+#VT|2}InZ0-GScAOi_{Z0bRP
zQ*Q6QK(Ae7qJ&zh`0@dL#T*u?MQOPUuECh<KmqU=ux7q;-*E4jLE@{wE8qG&s7Y`L
z9TxbuWXupE^^2KgJY6OiUJ;=xvE+MGug1~^E^9uI{PxlLp*MBnMpKZ?9!OH`9}*8+
zQ-5OHdzkCS?MGu*=6>B+Wwyu2E;U7m$P12!FbPzm?XirjOvDte2FlMGOQmxex)B+Y
zQJm#00$Uy!^H=x1rn6+GQRYYQ>hxx(`JBhEg+6$H@qOl%Dlu=auUpqw#u&1o%j?7C
zsxVX_5v2I}$B+SS=M{aE{4su><AR!frJDU-$&W+M#7-7Kmp@!_%z_?!0R(xItq4<P
zPDG`4dXM)u&|?6}%zVGEF0>DnzMs+Xhpk1-9+?MSZpglr>BW;&oE9b8uW4Z1NT;xJ
zhY8v{Jl!TfZq%-*6;n|xt~pTts<fmAGZZv%_UoGM!@LMan2-ber>fl`(jpeQ*WtNj
z!(TEms8XwUt}?Q3?aJWwM~@c;2iP~&_+(MqgyE_OKo53C-8;|bW7vTcY$NK#$N5(w
zlC=EKZLo(tj*1`Q%oV3K163D9E<fE1T}Yt2?DO<$-HwpR(`{_&aVH(pVmqRIFG`H8
zQE0hr@g$BlTB_Vbl&Fk><a2*^KPBIDhlJGtWo*)`2xev7LdWr=q-akKM>m1gK641&
zQmDC(LeFbPuhCvuuk`WVv2|`ZH?m!PdjZJgv>GB;CVzMeV}4NazOZ9nQs>`c-DX;|
zpFOgeXbZY9_^EvLwWwao*o%P+YW-{`<5Gh&cKG+29toMBK%VD^o>r^kRC>X$(WzzA
z5^Kw^wh0WPKA+{4W0?j$5L%S>)boW7MaVdV+LA9JT{E$h_6c-h@v^ua75D3br7A8f
z-!4NwPM;nB9xz*PF+3*Mqmo;qSq=4)JVT?;qvv$b+JfLTes6vC(oSXQf=Jqv@mxGS
z6cS`_jSMiatjRdF{dPN*&2jGLlYJ8yHLv1AL(EZ%vUoANg^DPo`c_%XxXjWLE=@~b
zHo4La^-=nomZ7lP+^OJwXh<5{uy=9cVZ2PA$SKVtb#*aV|FxvY^ojb;T{lu;AzAnH
z@Q77KMc8jWr2d=7bU7d25@a#jg5OmRdtq%}OPoLJwr;8>7HEk+Hs~w<#mYYuA~D=h
zcog}R$LOyC)V%^{S*h`HoBDE}P04jhcJd?%mUeoz`}!TVC6Dd;N635=YiO%t!)GE#
zqJCNGS{s*>2PfM^##o1x8mD(RN~@%Lkiv&dZE3lby0!XOO$hs7u=V(*dj=T_LUjAr
z(Bml`>r&;i7A&`P#@|ofP&?R98!1IyAoUdh#o%V-PO*bt6<Y(}vkfatmI%ccz7QxT
ze@?J)1=Niru0?!kn->}ozt`MMLqpB?ts#kYcGlX10wEntA6}bx7-G>VjTy21>z4hU
zUUz@uhD#!PJhAiQXkK#2xeThR_?bY|w_Uf;X^sjeX0j+zJVGW$k*RX&!C3$3E1j^(
zRtUzlcgqoqKl^UwLmQ9eA^*%DNKPZkv>#2YeC5;kLmA|FpY-?43g;)APN%GtzkLW!
zGU~ba*dHG~a(TK!IM$aZ*`JngXI&U5Zn5tg9U5PzOf!>tyb7F$d({J7p1RImx6=f!
z5%<-E5?XBwm(}=Af_f_F%<a>_0;Ul-gl|{1#+`EM5Bg>!HZY=Fedv?zvu9Fy=pj^6
z@cg3j@61V6m{Ba%*Pfx{lW}@iwh(}*iRnYa^en<9=&Qx0ymw@gw3SNFajr>C-GcK5
zVaR9?{>R=Nmq)uLAu6=O6F|CAH@zFxzgdy_Tve#Fe?7OR7WfcEN+jFs;C-4lZWM8Q
z%i~kGBj<udNu>xVS#^kf)9rDUl*LtAoL8vmmgxP=`&KL}1<>C)hg?R0TJqjJBDoQ;
z)qf!Nfdb&6B;B;<&36&|n8L>Ls{<1)5%Y3<bVZUzGr-FR29oDG_{ROMUO0ghK~b<7
za3a1t&>y3qxf|ZXc5}sdjp!2RIeb63by_#d`YfuHsqx}9P{5Y{v@RYCK87Z6>VuCB
z$#uBQ9>P2EF9xzWbK`amn1#QwY+h;P@gb&H@BRcvGsx-n2RY7)oh?pmx64!4*WCoF
zlLV}Plms_{+DADCygY{rmCP)hFqonA2a>4*ju|R|EAUC{ik%%%6Nq3%-`M*D{FH%)
z#I+*&lC7M0rEHY9$YJS?MdesWD6VHM(1`M^s&NmfB3c^ai22WzZ)BFrg+7i=^4OhJ
z-JYIQgGg#XE1DTh6_zDgV1mK`5T*h>MGkR<eBcNtofbb{Aw@x~$YQb^&=hDB87f<b
zE!zSKVZWe4(J!7a-NvX;WifT5*Ie}}&qgx$W*b`?l1ZVnJVD}PN-aL(@2#R%>Dy~-
zg0I;{UiwYA4G#L`!h}JG`0E}!NWeo}`vWQ2fgiwZ00;;J-a<!=B;h=prAE3snFk*%
zRf^YSbAE=B?1V5>Jh8O+f+d$qAk87qk-&a`(YJkimpNsBWNGq$Lw4AH^b5q+?K&Qy
zD-6&-VyW0KLF57^zi}=JsYGeo%{;j12rd4w=Q~^(5H#Y&pwzJWde&?Erx~_8GuD40
zyBKwks6Kd&SKYOs$B8ak$u?H*a<(RNR$YEYN1E#3^ifqF(8Fu?P>!o)`u@wSk@-nL
zAV&O`_x`89NOUb6{na%;F=hGvl2nC|iTnZ7WOZv=Yj9IK@b=Unh~5{>)CSC45F>k4
z9VVEtjUAX2@U7|pxN))bjp*g|*f&>CWzl=@KL9l}d`9So>F!3O)vcCIiYmaNfsN@9
z4aA3}eN)7TW8bL%q07Smd+$;0n;K4_JAWWH;}#8!E2oLS*+BDxUf7DSb2g4|AL5Jv
zI}Iho`48m3ymo|cZT-uS?^Dq%oSX!w0VwMch(WaSJS&D!V0Xgg&rBaOi~&#u$`krO
z%r%IkyRH!9jzYk)*)}q@v`RVFld6KspU)c8lt^a?dGr&022D=d1UZ}0+k8iKU;H;z
z!Smc@Lx?+ikqzdo0J`n@PrF8@XAR0or8W<;Igz{d*5JjkmHAH(FzlQf$cgNeo+`-I
z_Qyd#cv*<l2=?v38+grzs?t)OxeTXS$E2G#1;nR1bQ}zxB?5f4cN*MagaFw9pe*xT
z<eo);`DN)t7arC5^>ez+NohH(*Y0RaYW?N0Ia~>Zv`C_ORKo=efV1wJ>z&-vng4#m
zPmf;6^^!;T%#DOt#pfAq(B%V>Iq`_hazs=)me(cKCW0~3)i!_f`=cUJ8yAI`Z*0PX
zg&ZC7%RLs$st}YXQ8=3H$NLO-Su`(ffekqy0vCH{XZR~RErdqNCyR-mR5sJYqE3`*
zTRS+P+ohVL(&olftBvLQ`XiT>+LQLcG>g)u$~xj2B&<_ty)0KbY_jVqgy}gIO@lfK
z^^v^8q<_1N$hI3p3`)VLyXp014-~I`d@^6Ik$hHAE_q(}Y{FnPp4)+SG^mG<3#EmR
z<I>GWMsi&d#EiX<G&|E_du5P0G2UxE`Yx>hsc8pA?kE>!?|@sk8PJKkQ(013jiJrK
z@JW1*HoaIXc}HrX-Z!%>vm^5i>34fos_@I~GIueHg)3S5qVI?A8pnF2;|JKYm{LK4
zAoHvlvdXmTORu-M#FJbFT?cRe68^Be>eOdEPY1}Yxy=T8=@X~09}vsFKFRd&>#N>+
zFvm$smt4QgMNNoCE|Vg^5fSTRc=<Z5Lan^k_Y>w?N?PaiPbDN7<CDL3$@i!a*9-5u
zlDe><Vp}sW2N_APDPjKhOa3!deXaGGLL0C2nMFbN%?wbf&ht0NkByXZ8xQCxU14Kc
zCO9QmM$ffFTkPy?nkqH*8-U5Y13*8W#bmW1nLzP2+1)Tp7j%V&O3iE1uXEWpULy(z
zJ8M*%gQjh%aqB90Kk<CbH?eC|lWDJUjtrq}pTz(wJQ*OoHbiu@Bt`gs90eDIQ0!wt
zE>;=7G4x3i>G4Pd_^Ec-Euz@IfiP9<^tg60qu(sY>RGZKKV#rer+a!2_YReG>Km>s
z&^z+FzOCE<kWZ!NE^MJ<<T+W##Y?PioaRGm+dOw5=4J>Zlk-9N?sdWjHfI$1e(&DE
z2A~yr$M~h*w=BTr(BJ%;rt0wAhn&B4vEu=q`d(5^HO6<#FbPMmE8BaZO4UtxnT^IW
z{dO?~P9wIKxBg`5uVw5h^Q@D%qOMWsb_Kt$W_YcCg_Vy0mn;41p*kqeTCoK#yY>68
znyN6BX5X(w$%3vKyM(S!A*FSuMd<rTgxN3y+>z5eZ6xT<dV@o9=<XtnqeJSs8sFOi
z61*+`FQRUi^S7mTvp~!pT6vYCrN*(%<|+R0nZ-88C(^aAFAa;%g>)!{ClAPn88@8{
zp;>?mL$_pFdk*zT`|o{TE*oS&yr=p-J!e3v8tX!mAtss-wOTXTRxRd}%2OJ6pKVaR
zcs)Fes&dcuDks1Gy2I`971p_h)N1pn6<p#hu$Si_*BRNu<^;jO=0Z<6+c}JJGL^7)
zE)+0ty*&DxnZqd9T}#<3dW2bgT@H^IdEJfqU>5&=FpWvS-e!P+#a}DJ@x7p2B{G)T
zKHfWEQKHPf(xDixNOb>tx|s^ury^g(U3{3YK2)j{7^_dfen#6K!HoM<`Bub=MtdA~
zYw@C26dylQi_D$bQlR!)w&9;H?}^)EiOIVxoIfTNH(XykyV&7sYDyBviL*NUobO+;
zkWzGweidsQdR2QpW>kI^<hVA^nYJPb12%BVZW1tX$Uinnc|tJl2sAPzo&H4XxjV3)
zHzfULeDED03f(5xJr#qM`{J3D{G$m}8{4L>4!dtj!Cu{o;Io-WNr4iFswlMvQp}9j
zkc_2cR((ZnfP#hK*0+1id!ZCwZU(#U<7DbzSjD64r%^|~Gn+5Bj|9dxu+|rzhQaHa
zFU3so<~x^k8pVqAxx|than)pI!lF6xWO|L2)wlpHK7Q6$If`9))U8_y&+VPX)Pp8z
zC~E-N$zur$SLXn-I8bbDZnyxz&8@lt#w4$!9I>zUy4H<#5~o<psy_J?%CD|liIb&!
z+L$(%;iOZ4AY-P3=o4fYKm2s~aB6oR_1!7IIO+z2ed`py>pW!{pc=86j}SqMq6BkR
zQK-5AD)t_Eu6zq_=Ms3`$yC@tH_?AM+v{@Fk4N28)y>_Cho(c`O=Nds-gIyS@05$;
zV`e?S1tDg?+;vT_m^R<e{Bu+UVX{CFxW+8TVsPI)(L|YT^h;$n+@+(0Z62|wZ<7a$
zuUYK6+?Bf8X8-27>&hLTJBrcUGb6`{RfY&VJkY0{K=0mLKN#t7y)&}wKcMV-GPG*9
ze!J0*_QziaUwJhoZND6iDS35l<Z}n3fXJFQ|0OtGsfqx>$(>2o3tT-^T^c!nHUEJy
zE|1rjjZbx8YFkLZvKJCY<d;C$xuVGHGQUZ3YJ3Tw>PHjKf?7gsaS#i#6v2&B?jx&v
zJxLy6bHr))mDXZ3qZ2xB+)eZ8R#;Wz2|Z`}hSD3P+DP)9F{@A!nBu@babxM0eS7iY
zH>KnWvX!Q$q~c;FE$qDk;xsVFM5$4B|Fw9Y13M@;+$MBse0zA+Rc!l&R)nRy5r&cJ
zy$0lI(ta1QlX81As&P$F{5Q6*=D(uDUfAjXj{IDHKWx;YFr&M!5`2%It;4r(_rfd@
zsnh4tbA^EUX_ad{P+8r6-zd`{3Xp>lz#Cz1=<R@B*p+|DQxF9iyufjd5v1a6`oR4Y
zXuP(%bGCJQi)sM9Iu1XgzDpAtWM&Q6)WPY^s^8eb1M=ru&>EnyaoTNBjoEZ1FX-OL
zO5VTSRWaz^;J@(L;Jtrwbx)}d(#H#s;S@E_l%oJ0u-Z}^|J!Q_x%OY8k^e8C*}fFs
zp)-yR>5IEol;g-^VPS42*#f;$S{_p_@zBg;ujJXXA(@`Ej+<s71MUcZMx}T%orG*Q
zGr1HIA{BO03+~IV+Av79;|*kQd?!$kfD)vgl#Nr!ChX>Wx%mc&4N8trD1Ys%O*X1l
z=h)%Vc-mOT7ECs9BD(!G<hHGHqobuoA;am(jH6Z${kza#H27q0#xa?EqUG*$VsZ)*
zYT3WdR#x(PV8f=5Im_Jq$7AE_M?5;zGf&yMx9&_gF~!<kB_i5%^_fWV<C7E9TEZy;
zngWhIf=*IR`_JgqVT0TRb;wp`)HG9_HDBB2+3$Se-j=VgRd7WWJgTVhOTc3^y_H@u
zr!glCa;~B6Gohr^PFx!2`<(2R^RnlYxAsdXn|h;bSnt*b|MFVyZ}<rXgyZ5!yM8$3
z8X|)CA_mUrJaXoj^VwhRV)aGeVlmo3awS6)nLC{UiD}Ptl9ZxHO1YF@xRe?inpB_m
za?84@cRrKqMi`{(=79XB0a^PfTiGQaiMRUPf3du@w6rd{Uq1TO6#?1HW@QX>@pqIQ
zSVt(`XijAVygt{@?Bkz#Q>;wyTuGaN;~PZ+4LkSVHt#-!F#WQTN%;XRc4tOeuBbaM
zK(tV-J{OZyBw>Hy)fuEfJ$zT2;P7(0@ul4O`{EqL@~**X-NlU=M_(nyGv}^T6Y}Mk
zJB#f|tGkiJ_Nr)^09OGQB`&UjH~f)jpOnTkN6wWoU_3HA>Y=-jNu4+ZH>u|ECQ-`L
zDu0Y~;Y!ZSTVL<JV&!Kz<4XZ}F%$KGu-GBnaCzu~8^bC~f|uYa{&I}jD2w*hZ?uBc
zzSsvxabzx>7+<GFWl_=6-gsdAK<cQ<FR@cjoMxjC_7h0Wxu^KyYV7%-z9|I+%Zg53
zkAs6P_a^uCz;wIO5AKpbB=tFj2MG{dz1;32QZo+DL&`DgL)~`WP_C8oV<2`2+Kn%T
zYD1}cW)>@__GiG3CV~?S*nLI(F+Q}qJYa4%w*T=$<@r_d<lvbTYm%LV#BO&mn$fN!
zsTIe^LrOI2Q@$;>!Krl9@AMCCfh!e#bgajtJ@1gssAh-k$K4~j;q{aZzLb3-bbQ`w
zn&-$W#~2Iwn%=$1<nA&Bw_~Cqg=-UO>OVLiXv~M^j2$JN{-t@z_gZmF<tYf`m-avl
zJ&6fAoi(6^d7B#En#nPS(07pw-Fc4@zWkH*73PBA7kg^v4Ur}N(ajeB-S7{{@VmoL
z-~Ln`H#4NA44u8z-CV^j0?}?>?XO!jJjz1_c%VJRUhT>zf0}EDjbZe~yjnH{g`xEU
zh_zOtr48|@yCR<yZ~0T^oe<ILR)|zZY=5X0EFo=YZU+zHJv5&ZDo_3cX|2gi|3c#u
zW&Tj@4%MR@uoV|gAW0dnA^GA(0^U48)2)W$J);d<OgLaL>T9%n+@=HQlB!Ud9wsx=
z`%UZqjdBv!W{ltR^^Ye2wt9@atL=}ZdG+K4@_-(2q4;&hMKgl<%kA&fOnw3He@)!T
z3kma$NKzMm*2O)%4v6@C55i%`^&TO&S-k*H38NTQLrImEnUP-LbGi2nrQ<}1!|~+p
zCV4n`eH`m@vuFLb?<V-t{*LiX!LOMU{=X51!}?J>KE2Aq#V)64)5{O9d}`CE4424u
ztzJ(9Xr((sm5<cUlT>#Fi^?q&)PEfv=U%a`)eJvD!iYapRT|hRFN$)Rv=*|IzXnSY
zdZB5W^q6(=S=ytzam2V5PVn|xYAEJ}4xMOkunLW)%LL^t`Qh+R;|rTBM8=Af72P@6
zA~mH<Vh&vNm+#ZiD0&RqHVQ%Kb_r}zLoZd&kHpI3^<RS;!-{Q0OXdk=0^9-4;gq;+
zG_j(X7I!hP9REF`$(Xz7KK80dngbCl4L}rz6az|W=*7t=;>aSu8hB6-hGh^TK^KX4
z3SHbhJ}iA)q9aErT+P$Hyc|Q4Ux7J8_T@?R?JMpm5$qzRdEgk@n6@@jk&@8%o9CAL
zos8N)teCobyE@5j%su7Ry6X{w%5RciU$t>WI|BXXuef!QYv|=}i6fC;sI$a4^9It%
zPo&nn<&|HC%HO08dYc`CP#=Q<UiYuC^m<JsWdA1Rj6Rz<tfhto-#tF5Bd{fCyUo)p
z6b_Pu2V=dxSW^2!_QhZB->&1NxSi?6?<r``{O)f+!pnWzjN)ZP0ITo><O@Rs;fG4d
zU0{8}77XU`LWKDGlc{cVf!RmnFK5Bqr`m#<2lz9jf~i(m7c<MxM}VVs7WVS#B=6)W
zw_NS?3pO8F?0leu>n8Zd+uMrxpzEaftIcRene?C*_J!IC?cs}w8)1IuuXhff-tR=&
zMnXvRlrwNPz*Q7t0W7Nl<|qv#M{lH0>$hgoJua51RS@0C%dmXLki*FQa^j1(QRmH{
zD8QPW79?^}LS)gTlYAu8q0AG?PJ#HIa3BAM;&TSmqH%6lqC(^IiFlG%Lc)N;0fgxm
z_Uh65r#$*?a60kgr<OY6*B+)SJ!iOi%|K_e+*^3m%cclqGQiDmf#E>0E^EamnN1IP
z#EkUW@Dtb~_;&Srw~hN90xi0iv5(5t-`skn^EPzFk}=<L5zD%B(XX)+I2cZoV6)pb
zl?e=z2wg>swgJzQ^VBI5<Xb1@c_h^ppw!S=G)Y|<0rXs-8vj5-mK#Z_hgCC-p#?~{
zS+?-<`bWPe5N(*c1PSUazg%zFv<pM*&{hgrI=s#8fEEx#Gw0B=2n33sXx$mpz`fMK
zUmv0r@b=5E+X)y`f9*cNPF}vF(BY8?+_Y_N#1GApZl*->R$a%ZB3G~u<%700n9G`r
zTWy|&SL@B@JE5sSF$w*xOF6^Gjs~|MJ=1^ycb`m>%dsJ$XT!sv)|>DN988Uhmdx?I
zz<6{H*Iob{$p4Q(|Bp!a|Aau(6pcBPO6mqST%SLdy4APQ;gJwMcvR>J`1&?DE!pec
zD%<(fv*|W#x7tK^ljD><uiasnJfW}}J?H!J4}@egB?ml2;caM=Xy);g-R}>k&6dJi
z1$Lz+eIg%)UKl*EVC{yw0XfOpV;`>&kbx(q9ti`|S*CH;3OP-AG0lRW`Glris+|r&
zoFXH9iF`Nf8+&)eu6F9xh3z_R7TcV;^@RIsIHcsFM~6pR>R~A;^TUup5=1D!1HMZO
zAbp>nwd;S)8H1F=!`irKe;`|?b%K8&4=$s(ug<Ri&*DW@?xU;*9%Kz3*H9h!V0O~<
zZdtu|2FJ%@8}F;e3@pJj^ZKxi@N8E2+B{*ey%K%?X`q%NVYe{f+}tlwS7yDa>5D`u
zBrmw@J0q+<uv$lrUIXPvOu;wClxM!I=}5_%zWILCY3uF17}s~Ns%<{~Sh;h@p8*2F
zD+ufF29hSgbA1+dyQCMT<99yWAqZ1&i5BN*F6ab39VwOm;!L2|L_CuUsgKapaTAAr
z`AG~V_hH&%%0}0hVb`QopGzp#4E2qXN%{HLgZmlbyY?h++;_ex&L<@w92H~V<}WWx
z>5A>tHrJxWIbSdU49%lR)$_QA=!IX5*#>)h&D5K$spfpAgTU`&xw>xLdEX=q*Sd4(
zGFl-FOoqN_2W(7(7*V_}05k?m#+|+~Oi?%i$HbR2B$U75I`@S73N_zD+Gc9#%D4z#
zJhG7w<?(<BwI0}3B0XJKZP~KCJZtHgCpKSwV?JG{OHh7Eo_8_-IEPEBYDb(ot4<Uf
zR1B^7b*<~dGXW&Wxk}A;NC*wBU~h;5HMEN&f|4Pok~N7!@kj@=x~eaO!qs5`x(dQE
z{aV*@B?~#4Ie64R{oXYs@I`E6BUqekI>HN_Dou7{qxUK|X%>PYcblFpdBvt5L>|%r
zs6rw|0WZ~f(y9$J31MCnt2<_|Q~WqLRnFA&+szO0NWmqfa^w!vm|)Y?r9Lt`u*gO6
zBIT_;PgHZwE-bjV_6pTah~*1~&O0hDsl?57D1HLX=^}fenRhxgrhYJ9<4I+SPPrg!
z_oAVWeAi+lJ5X@GX%HtBb!3P)^r7^<SV3POm)Gyl<mz#l+%=tpobmMqlsYXjvz3W!
z99d-Y0%>YdZQ!|9W-+ECednVmT}>e@L^Yj}R-hiedy^m+@l>C4WxS&*2jP@_Q{P2H
z?z6P$mUP1B@sCve)Cww!rhIt5aW1gq6J9yn{xx`iU=1^i4;vGE7_a%kC+^K)i7<z-
zicS+0;_nM^pH~}2j(8{k9Vr9pW}+nMcW#vqAoV`F2M}tW{T=o_fYp^X2-r?VpV|Ot
z3hVF_2E)(4T}AUUgsJCPy$xvrJ;BW@GTvyXe}a^9lNeT9?BcJo@Ya5prVLTp(p&qr
zAEYm@UZ)N|Gq>HvOPWDx9cq%*sz=1_2V~f#C!{VgTXedd3z6Zbs_rf0Gg?i9$n#bm
zUPoCI#bcE<J~#$lPUh-}?VW1%Ecy7?TkA8~#^8oM&!BBy1w3!;xoAn+9ct4S|DJWA
z{M|IJ`7*F`x%^N^=#q$WMVLCvuJw->B|E}}Yl^0&l<$8a=aqHv>6WJb%KD;mF>bUq
zkxREq6l0oip9#Lg9l}V(J(m^@LG=`(t`WO43#=@3W>q{*L%(r|t=T8bHm`j8a17m-
z#7BWCaK?pR88Nis2msu&Q;s=RF>V5mc0Q-S%(6qHnS1U`#stFePV4VlibMs!B^QkA
zRd=j0;DRr`YdAd!ckFpNFs(0Bi9Pv|MZvC0SKi##5P8@96#vsMp^-`u%bl-N$4#DX
zcf~ylJHTs)K6#g9@bFq5;_{7P#e6GYAho#Bhd4P!=|eCt)@GQ=ymrO;ihj;}b<vi3
z?y_iZyrQQLHVbXv7-6v&x4j<3bXJG8y*rgqYL)B4;hpkh|8@a1PP||TxRQDua|<7t
zUMjkFm(A!iJ0;T<`v&A&S>E_Kp$<^DXMm@MTb~+V>J~7;Y|;LT;=u_R$KPHqyH2LH
z;wuJ5Bc`*95ycc3a{5lgz{S$WVT0lB@g+Y)30l(ylX*{n2i|&GnVr-3Nqbg_t*Z3H
z6$KY`=0BHXW(7TOlwHM8bw2KeEtZfx@ZJyen<;`q+#c_Lpm!|}1ZTBcr_bIzR7SRH
z*Yqz`4DR^)R#a5;8b_ZfXce(_y2vOMJ5@2)AJjX|GS+1sBn|sqQq@H6ugA|;wIBH!
zk+^XBXIFYBtigs<=REHtO1@(F?F+YCGkx?kpbYueIXe<l*7LmXKMtb<@!7@J#<NUg
zaqXN82X@H^9r;g3e1;P<uW1_0&0K%%YCA7W)oCywGlwAbPM?J@-Pb2m0kZ>~$co!o
zYo|qi9EGo|V0Fp}F+Qd8_9y3y-@&)s+qEE+-`pPU5R^G!Jr_GV^Pksgqc_;#V6R6!
z4EmOFZdh5E(C>&+f3$0ATD9pDf1B=W$!okpz5Z3t@aNqUsk@4J%5~(2#b!98OmR>L
z?P5*EEi=r`tuTy6w8Rdi!DE{JD`50u`~bf-NpC<DZh@a!_Lru?QH&$-5&jKb314~h
z>D-l4_t`1Ejb{}AbsTaGre%SGQX^m_yy5z46{0Mx2vFZ*b|cdJ8p&i=9qu~Ei&y2T
z{j!p{S?ZrPLH124tewZFxgJrNLbrJAJh%=Y%V|x`4^w3wvv5g{bMTYyx7ufMUy6&#
z4kuKV%=tEBQ$SmqhKm3?H?vV7mZ0jNbtk_G&f(0fNOH9;auH#h7@C?g7i4F7)ih=7
zp1Z^Ek@1=v!lZG$BC!{b^2t&~$o$!-l{KexrmQEdwVw8BeXqIt*)@k0_p{eCz>a8B
zVdHXm137I#U~4U`&||$I_xnzAiO%I;0;23G6TR_l!N_i@gk}gS{J&qm$VO&x2Hzk`
zoRHFbpjp9`fTp{>)t>fhTIws+9#u!+Zv>DfwB9_H5oCwK|DI>O_a^TZ()Z$VO7{U@
zwCH+q_Hc#y&j@r~yyBbZx6j3T&0)?F=)G#$`T5lP5K#0#KOKrD&`g7-03&0Wr^p5#
zczAf~wZ=edlReV}iDdbRQ4Z_)gbcNdUa=l&b~{;&mB;DOMg>)QKz3RaeUDz)Q@ju4
z5NVm*=($g%T1t!LABZAkwrD$ZmW_<T!qB6<1Hi|m7zp8(;Vt6(7i}0p{#7NP5@7+F
zF3iGX`2d!3Ao5w(m-NZ-n6w(zrS6_ubIrI<9cxuY-DOB`o94j5x(CW^_KU#^#dm;Q
z?j;i7dQW<R36n;*d?9uoooWyQ`yJ?fLDma;b-Wo&6Fw<}XzYet1e{=cyq2}dB5~a4
z-nTe?GAoFF+D4Z3z=(ab(=Tikhj#)3SP322M?&{Gaa-Nm0x_PF?P*k<R^BQ_+i<-~
zS^vO_64wRxWR=klcgRn(V@<*TB4`nXJO3%I9xQ+zx=k(u|NL8Pc#d<2<f8mr6GB=Y
zKJgxg(-u7s0z91j&bE5F32KNwx&C#<U44(5`R-lVl70w98Y$9d!&^0YO=ewAxBC8h
z7t4vzP<$dJ9?D$pHyfsY5JIy_7M(r;9tHgDCYik*_@$XH`~=Db5W>;RH8pO_lLG!g
zXe7`?V|2?;g7^`mDxkERqY>nLL65-=&z&Zeq3A)QlDP0jf-Q(fS+)M>@$|uX;!V^4
z9Pc`qLNMTWf*cqSs4tqM;l~348WzYe2&LrHe~*YJ*fN0;ZCr_)c)n0qWvUNOin4|F
zA6FEM%adOwdwq;zsabdMv3eGa<k8vf!oGCOI`O|_>ITwjSDBFx@l%~{T-vKB%$b~b
zpUmTWA^NjweD4O2i^EI-AmOISktS}El8TmqkwVVEM^TOk#sC*y^9d8xE#m$i&3GMF
zuR}PCcZl%@kwogFD4f`B70nK^4u?r)q#%c}0|VXFKM-Gb1#XZ4AdbvtMk8>^_1?C#
z>skP<r%m2eNviD*<tn+#$bLp0mT^QYi~ww%8*YLrO7Ri)!;<fTYgi_Ck7!lpCgVDx
z&bjc*>CS6Ry?l?4vNa&63&fmRbo2<$_vI5;Wu-#IxXt8uzmx4rwLQn^`h>B2Yqn7j
zPc!v4yiB|`?z|bwxMbKHc$eyZ9>__`4H&iHna3fA?q1)IwG{pgncF_8dJW}E913p}
z^U~`DsAPhkzEDgsQpqfDa=yfRt^1>=NAG<``3)anT$UNbD*@njlnS)1s?XUZh4{T9
zbxJ<<y!QBC*yN(L^=25M)13)UY9=&xL;>wU;yB66sIj8HF?EYRO2i>mR>#hPyI?lV
zU42RG{i)SP`<`$c(Se{#?H`>Fcy{09kE!$fiDx_NmSx{FW}vbiOmm0Kg{Z-84b)Du
z67}0aK15ZejJQy5pW_UQD9-zqj<=t%XccCz)@x5cC4?7VwT@dGAU<!O5+(i`7mnD>
z8FYPL>169g?IA6i@s`z-(XZQ|!8R?}kEzWqbV~Mlr^RN<Y`V$D74eIy3hQDvj_Ev6
zR~T}5Xuqx@WtOK)O@r(7-;p-{K+J}>1AfYWQ#xPub58H9Qf)TZy<j0oxNsHgP9E4p
zmVpA&M!t;bgM|<}il4rlA9E<Lz&PfbC_OX1t3DIw(K{0&GV%vv=SGE0Q6nZ7e_Tz~
zSa|EOOH+=hyw#^vtHVY9(8L0{nYbYmPV`-e5vfUUjVM2dat79hXYrBaO4{6iqvGBY
zWjl<2Vl<=3Al!a4O7#k{Dso4Kt)0zdx+7W7@7&alTT3RlZhX;zJYR4~$hZt`Qr04*
zbXq*B4}63YFS7PC{CsxYLPy$=F*|ASehBRz4=pvb&CsC$8BUaaj$6010!Xn~^81>Y
zU6db}h7d+Nq2V68gCnh{zg@6Yx9W%Eq`{>U!doz8k@rK&`JoMqQ*?gAl^ES~8PTyt
zA>bj!!Xo$L#+aIdgb0-SVi(TcMDO?G^`9$Z+yo_b1UF2C1R2&Lfg5k+Ihs3@S2D{@
z+boOkEZ{oLB3s6_#CSgXxGT2Rzmm$f-MLhsvui~v#-X`g6dMw8Z(qCBy*xQ+%6t3P
z>hG(sS|GVzX?PVs1$M}RY>~5jeMh=C!>v<75eZ%cLd+)4^!q<6Vt%aDrd}RvIfy^i
z4hiy_2HMD>^rQA7lD=iqY(kEty1h_BC8OlWtT0#!j|(^g?We=1g!-TIaddrLE^`Vu
zsyy%0g^W3yOt=3q*G{}!;Q*iKrvl9Xn2DEgyMDzxqwDt*zF(~KF<C1KLEg2H)aK1i
zwAncnA?p&^Vn=Msi5hLDn9+U1cJ~X~ufu=Q#Rp%b`Sm;+Xffwk8$&acDrz&v$bz4g
zTGqLwO$@bPJdZBAd?7rW`uTW63dk*RU`Yr)qS)fgr(@gmjR8+XL|#dJLp92IT-H=O
zn(6SL8R=tviwo2#cRRf@o~l_YlayO^fqh2emRJay4$}Xwfu{Bdz?+{Qq6wsD{HC5Q
zm)Bgsb)1^joD+uT3uPL*EFs1ma3+oTBPr`DqH>AVq*L4}-MkoEr}GG@uzOU<t%G32
zL%ZOBymKb-=n5xkK3bR5Hi?g=%ezvOLT>3F`#OG5bw1R9JT{dAKC-n$SnKAQWkq+B
zy1LS2_M6-%=I_u1n6?{IdNa>*YT6Uo)$aHdhgPzUK9WRBA#PKCz|T}P@*zzP(JZ>G
z?en2c(N-T_IFiiHb0w+W)w%=0Ph+UI=*iN=ZzV>vr&00l9{Jynfw%CiTx;f8jYAbu
zt!C}O_=xe)<XlB%wW?)2D_{a9g|a(5M$_mHoU5@BLHwC54JU?oYUOj1ao>J*aq6V!
zl|s6eywvz0nPEWy!vIzYK@$6$@BIw7ccWY7Jx6j9<ve;jW;Hq8^*7orFqGzcbu{ow
znD>S0k48c-;w*Z}9!mKoh9*r08XU?Z@utMZ)_phf4dQpS3vxF`l3!n%T8qILLbXq9
zf<{7iEM`>|NxlX|L`w`>tIyWl`&7Shq^xW04=<rm^_>f}92;7<H$|dBO6qOnNtDcR
z#gXFWyt-qqH)SIXV+>$)2N~ZWp~gp(8P*`*Lynlb&~x-0XbN#IOO1>m?omdRR@Tn$
zKW4BPBI77ra`K7yF>K?lWeFTDJ9E(I@yd>rUijI@rAGj?(IY@d7qg&litLcGA7Md7
z-+Jrq-LTnpp=6n{h4p(EzQs~0Rd;f`!_=sV`eYF}Hvk21aeV(|bH@>7h%j$`I`Wm0
z(h;$r#d8BX-g9!(F$k`kj6a`Y6r*Qm=1dDsPZGbkr%>ofE+1SJ)#BFsUgIKit()n;
zSdT7gcdIFb60blF6cDJMh7px0VF+@31rVy<19H!%Og02*Vm2e?;|dO5RgJVqcJy7)
z6sd&<-;q?qt6YXQVlN%#D~l8XUG)n9yi00pE30+OPGqL{-K3#~M>IUf7k1Fxq1s9P
z4Ik&~7|#r*aR2<^`0xT+dfA1$^xEwCnNFlpv!ENt*Pgpt=ctHs|NM#R|M3%hDUu{x
zT-Z@ESs_dP;icj7+8hvP-<rt1{8M2xhfk{Djb~0@Nrn^<l@fZ<XF<^Mh>t|&@(0qf
zx`usjbG4R;ZM9}Y48DsL`JU@>IVw~XBW)kNp8n=Il5CzP6r%ugz2W;3O|TX}j<S4F
ztf0r`EFyHACgx*Dg7f*_s&r2I`=PlG6A8`y*I^tS-KSQq>-`Yxg7_HACIpl1pfoI2
z<gjS4%sd^sua+BYc7^)vVV$SMPl$gWD}jkD&DS=4z=7Vn0gj!Rk@0B;WepEp55zHb
ze%nC9JK%{>l0bLw7(ZSa9|>6z`JIT5BPUI>p+_|li-;WW{UPGYxC<g$RYdpYxSY#k
zJk3UNm=P(239rU|GO!+iAHL9Q6uxDDlZp@Kj^nY3t~@!Ne)2K)#?~K}eg=EO1PfG0
zWr(K%di-B#&ljM_|3_#K;KxQN8o<$MGJEUaaJ2p3F(4~mKSq^~u<*<S?S@@WP47K<
zzoN9-JVWQYHuVK1(^#{aAR*{JQDh&Zgy)Ko2IUdQ_`bBHBI96ps=xbE>H<z9^75cv
zZAlaK7e-`9Hkj^jqlSN20{%Pxuj+i1r3$e|#+hWcWzcO+-2LJ@;=}Uw%<W0Jj-O4c
zS@BBGGnjhx2yvY&+Q7qA*HnJH>QbX-0mvKblZ-P!FqExj1|BpoE%H@;KJakod_>PE
z03{6I#1kp3!f{|h8w|q<+<hS2N(7L2`|>IuXq>YB1cV$w8WpMo7XHCG?>I6HL|?hY
zKae9j#ZC$|^@t0Q<qGgwLnQER3FWZ`^7Eg3ghFLBSnaVTK%unIsJZli=PKS0%xt!a
z6uq8_Yw>?q7w#z8IbU#4b_TL{so^6iilxtQ{W3pPQGKDB$kX;nw59!BRdl@1C$;a~
z43&Vk?NS~3M*~un0r7Aa5shuQvg)6ll5M?mUaq^WUF;nPjch3Y!S$=i1OLIsvZ|I%
z_T_2T@^n*^-i`7@dL~J8yV{%V3Ni5~cN89UX36X(9m{n&$d055e7ZQ!q#|!A6&!)*
zrMMh%kyCP$RYelswu$=zCYft{#JG-(3#6R)OE&z2BL*dnI!Q0$HNzwml%KtN7q?-X
zMsS2Ckh%zo9sMX>;_0EVXTAmIQ&YNr47-kYH!u?SnQ|Crp8xVl2=Qkc!_K-zZ8t{x
za1NK{SUIHzWpe)%oGiR~0hu9V%+r00-+!k;jc}Wtd_1G5n<Gz;_)M%7J6bk$wX-9q
zWWQ>Pd`bD6-Q^vsI#fd05mR0^Tryj-xiDyIP}-oQwRz!|W29y`-7Dy)U)+nukU%#L
zls?fEpAsiZ;U|T(Y7dR*my|T5c3*y26%#f(ndosXcMdO{W~zTE2MQ&_BZ34v-y!1+
z^)$OzPO)=<(cYY^G|uAj=UyaaF{zUB0rWgg<ZAP>K>8dU46jgJ3i2zWyFi>`<+TF6
zA9vTK<|sLc3t4uP6m_lSljDe@Yf^nqhsE-3HisrESkQ}1{WeoMA{<<ht8xH#Y}ul>
zq@=7qb(sBORYb7auPg<7p~UEnfD!vcDU>DAHxh-u7KPKh<t$rfQTM^-oMuEK4;|Hr
z@}dgtHss(#B*~?7npYXk7hO>xJ?XM0>u%jeeOJTn?9~qv2H&p;rP=+Ssz1AXpHzbD
zIZqni(;&)b)x0?UXhfp+#>>7^$><Zo?!LY~>T0|yc=!umrAnen(YYY*hT6DwNkx6<
zFzq>$hw4V#g1?^13$m`lOC10=NOI}$|5SoqHd-DIYc4)x^{F#qTzh)nW`?#~|I2E(
zYT>(Js{Tllunl;i1*}|5F%~YmwomNuC-+X=`pyo6ocN}5kZ4nTC<e+4^rP$UeQ<0L
zuG1y;f1T2hx=~WYijkh!g3ziy^5ei4Ly)S41c%P?pnNX}pwpY&8Ntj`@0Pp~i1)F7
zGPAQ5sHSXM6V!!ZB<szlqXA5~`Cj+^gwA6zTis`sI$ZuG7yQ?${L?&<-HfVfP?|8;
z567ww`g=R<HjDZ-A(GIpvjT|#TK`0G6a@96nLlCkGkAR0613-&X^S6RVW%vKzL*I}
z1mT1f%9tn@1)sfGNfhgMrs){9ZWXyz&{}ZW2SSTiks1M-)UCal#kh5G;VNL-m?xnm
z{N~)nK7vOQX!{fs$b2Y>WNNcH%PY`8JfD6kBPHms>+SLrg6DqeOf3Iq462wNX)K4e
zMQki2i=W{_&Ut=k@cEn+eRfatZW@!P%QW0xKSEWA*tvU@8L`p7^YMjYyF!cs;qaN^
zIjUz>DpEfoflzZ^@qzcl$|4_^lYesd1lX&TKe}-v3^}GQbiyNX*#e*XooMpz03Ny%
zb4Lg`Pj$N_nI3*Lpoav00aDT0xEJs<{2AK0gp3c??1fhtcZB?}e--@2ZQ4IwD*g=6
z0<_B+f^|fmL^B>uZ5G;#I4g~XCy#8J_JbKceKH`;s!$K&A?~jf79GBqKd!EsX(q?J
zs@-vyd@=KME_HYM8G;R^`-ueW^gSTQudp@nmdky<O?)A_ppa?%qNz!OQ4O}BI!Lvh
z{s8$~_scMu8F%{X*Q}a`feo>T*W7ETWbbvYNot_v<T`sI^hIh&Nq7bw+RD!l0>g~t
zvA!f8qHP-}+Gzn)hqM1${M0llXlUbpdJe7q(O@iV%Z$4bH&t^M)7v?c4t_f!o}_Og
znh$#{6xS4cI#BQM5>m>Fkn*C!fykO@_C4?^_8be5|DJWV{d(d+Bgb+y87~gLR$Ne$
zl+??jeXsmEbQF~YVrbeRQDmGcNe+M7%M3HMXWZL2#gW0E;r7+cjs96PK3G1B>D?-Z
zRF-W6%nLdibCw@U1^7PhV);68Qx3+lo)##AO9Z{GEzd*S`cF@W=JJs>0yPe=lNK$A
zBO&5Xl(iIrpXL-7TA5!n!Kdg%v%A!9+hE1Ju&?>E;&GkiZx+o5mZ=AeyXJ(9vRS6^
zvlGa$-m%GJg#WpEzmHBc`1+dX3w&=}UI42n9~(-?7yJ?DLiPJxeLp9b*Dtb?7`h8F
zFsmm$vuZqn?wQ<mas|hoJR9FFedn6)PN>sfFIR{)jxSKny#Q6Vi;h>o#sEKnAe8PR
z<*FZ=zyhMnn318(GSgTYG%rdy88DEixvPe&u#X<yqO-rJ{Y#3fBgj*}2^XxY3_pbm
zLNEM%1doiu%iu36&@hyguJ{?tD3^)1h1c3nmJ)NW2J`hb7<x<y$}2#J0V7+txEn$k
z&cpSD`d6mByLGAIk^)9_rmLitm6rrNQP25>DR%1WpfU&H`@Lal&nPrU?B(<k_Zu5*
zK1|fXo>GAS2Tbxmj{7Sv$H)CT-`Hxaf65CAtCQlRD%z04r-{nuU1}R<G~4omf)%I;
zK;*6B5J#s$R{UZk&~9M;tHI_ebo3^RcppQB)s-8K=}f-PxRvo}>g{K#0gm-?(Cip{
z`RG0BN)fVg_QkOO$;NT4>nGcqEvnY9k?hy?O<Emz2p);SL52Wj>=4B%VG(G;c_xaS
z&3-X(#NmdQX(r%CZN^o64!rDs@86^kPf~6->c9!Y-b?E96hIfINyZJT|H3yu*s2B^
zq}yFWTuVJnbJ$RNkUv%{iMfheT;%Jy%)8vsSV_iH+#{WQga*oS4xmSzasp%u+f2tY
zPuA%N;Kjo%F4cbcs2kvk5pB5Jg2B`UluRf|k$PcnHt<73q?Q>myw%LL4*vMb#@Z&n
z*6|#9<C)NnGrkJ(-2TvQ)n*`v@kF=e{RcqeJS>~#eI76y7tTqO@Zlrq)%Ezio#rpr
z2%+dbL3w$+3PAaR=OCt0LA{wYXK_LjD>b4CmdMyCi8E3XW}gJ9(~E=MTZK=@`tx4b
zDhH%%x+wyt%0FS%1tl37`9!aaoqMPO*ECcoUlk`AqtC#X42J$dHk+}19{{vVp|J9`
zF#+6$J|r9~9khs00MZZ?GUpNJF>XW88j#hF@0}?q+N6p5HxJ;X1sZfx2P+E}{a&SF
zJ&*$2<`4n8i>@IoU||u;F7uzVpG*{s3dGAt)VL*tQy)OIH}oHD7yo-Z)%~Lr9;)lm
z1a*g~^`g!o&vA-t5MH)E_0_sdNSgs8rS!i4j(gT6BoyOMjN|HUP$b_d8(J4mX=9n}
zz+4Fkcx4i+K4}Tz4y2W)+B-V7(PZ74Ue_nW+ilp0#bGhCKlREFAJ|o0>*JqV>t5|r
zn4xCxEW}C?_&R#{&@B${I{mP^ey(D_+`FiWg<7Wf^W`CX_>6L@z0iLQhFBLTTDDVo
zNwEjw;X5V+AB$b__xUv@1r@_CMnTh_vbTF8=o+Mv8+e>t_~QyOPs>B4iKWN$Avl?<
zCR)Zqu8<HpJFh`sP^IhMcU;wn;1B9UpC{XAlMM#TRp?Dz6<eAw7@IMB&f|3GCGLBA
zlGvM?c<1oI*SZTIwHe`mKK;Y0r}KH!5qcL<IojL@aZ~J&Fwj&iu;ib5cNpMmIVnuY
zsY9GkZGn@8P)%sk_QJxI;@7O>of~NhVLa{5LEVj~>g?a4jQe-$y|feWeEw{np_LwH
zRj1E>52~Qx9^lT@-_Z6hSoH#_p5W3c2WL@M4}8B=x&pozW8ErKcB0nqVH5XE7)LYT
z4=>B<vXmA#_NB~o(noGq{Rl8ztl_$Q69;|U(g~`xw9mrw?7UxKYjn3EBrjZJGnPjm
zAXC%Spa`*}^quXI_?{;##<>`&U{7@hsSt@lkA$Dz@G{euMm8iG_m|>SeV%H%y#sad
zn_o{7(r*W6IjDGgMb1;MlAVZye9R~fhxe;)5A1CNWzw!sbY$P?SPM>X!gK-&ga3UG
zYtyxn;~5hmC7h4IuH+k()`n;bAV;1)jx=MvN+Z>~C{5K#zH{7*y@4MYw$Q4cdeFUb
z;8NgFWq;LB>l#exj2b`ls&(gDGoiM<ftwV1sSJ1e%J!<fn2q@K=Se}hT9$)+r$VPQ
zPF0xf050fE-A$K#m(nrwafF@=!*zjBdahTn_&awb@A!{+97>>!h;D!w>&ihsH+H?z
z?PB`M?`v(2T?bt}%<4BNQ^`c`&0youv(m#E1L_&OQ!drQu>n2|v7g$*B-Oi_l)bS!
z0JyW0u&ulb^~0!}?`1O`EA^hdd&D*Q<`p^1A^6;){EA37@Q0}F{Ek<7!X@T!DdOmM
zF-)F6=ET(;R`##LS&JDJ%Y3ChV4<8x=6AA<!(Uct4!T-&eLD9el2caGU@HmwyfJMt
zD~qYH;5gF-(nPSP&6nqECbzFIowd;{lbwk21&>-<NN&8x4rJtW(6>y6I-^+xn{(>O
z%kAI!6Eaceo!p-oFiJXHb5xnHXuEX;F%lUbbE^ELc_PiCWqk(I8C{37-zNC0Voc-*
zxgk{46uVSFK#dvUKrCU7`)R5{yp(;Eb7Hh&&huNA4wq=c`(90&e(uDCK?d7Y$U#s-
zm~oszB>8BNqKvxLFnH~nQ5(y+&?hs?iH;EY&OPHZcVN|*>T7omNhvrq+iFdBBF_5e
z;f>41p@W9n!P?>eg&gercl?<L10A6)J50}}&#Z0cS0Wg#tX`&NrP3J2#jfbU`0k2K
z{DC}T3Y!upd?l)kycC6(wXYYHhbg4FJ(}BS>iPb4q2bM@t{PuPN+Wg9g^B-*w)YHb
zDs0<DgQ%#KNC&A>1f@wY0s)mK0s=~C5*4LN7YH>dg4EDN6e%JiC_?BpGzm?lNEHwW
z2~Bz^AwtM^9^P;7GyBY$^Y45=7{?h|?OCqpzV7=f_xsLljgEX*m0G;nZN3%!Wk7g*
z5LJNyEfiJ?j0W5-*{u2KkQd`5KRtMSO_l98FC?_R#SwAyo<J~ZcO81CDM3oB8k;8R
z`SNS3;(7<f;TwkQZ^f!M=h(B}0KH7|5BJBeXV38xubk~PRdC1mV$d*F=lVD11oyWu
z(RRIPweO{--lu2MfsSvWLx|ly=L~6jtkE|!u|LX9cD>-;)t|S0Yr3WqN*@WnN*lNh
zK`IZl$;Bl1iJ`XQXVmqj>d$eiZLkD!ZlEAEPGqhj;A|gixlPJwZ{k$+e+*roUDRj2
z-8k!_X7NNo0jc?H;MJc{Z%Vutt^yRbW`TlanF^(ZEAzYbUMHFc@Twgs_EjOMW>`qc
z083ur+^T4ZyY#QQthb9ENmUVdd0yQXG}F3DH>XHtr^NDOgs39>CDm~o7BBe&88qYe
zbD!rjV$JfYS4cD2#Zv|k!!VOg)!*1*KMQ}PsA?-#u@Oh`?C5t}vCXQXdT-z;$}KN;
zKlW<2s|;m)!SJD$ii{+H2QDmYj5vrla#y<U=!;}0ODB}_B;&k&YSD^&px%3ufiJ`v
zE<6Uk8ry#{Voz?J4_X~-+Qe^uA7FO{V{WW-{>q#3-z&D08dJK8)ZUj@7>UA<cQFOP
z?xTO7B1K}g$52Jafvl0w2iUIVPB^y(17#U=7<WubRKe)_T=oN*483#3pddCK3f{#y
zm%<63n*u=@5=^W7{14?(S>r0AAiKbvOefl63t1^nq;$<qZE{kT#;fj&-SpbXLx?~2
z_}XzF;TghSdtBY({Sd6E?`y}L?i=^_kG7AMV03u7xMuPCqbyA$Li|*Wr+@a9xsr!<
zsd89odh`P|Y0csF)^5#t`lT_74Cy@T=a(WBF_(HT%>7WFqH4Wm*8c7(R+mjTC@@SM
z|59Z!5tQVr9<vwKc=70e^z)mP`?;*JW%p)alO%fF_MXuc1vqAy*5@EXGWj{Svl@K=
zcX`|#;jk?eA`k&iM?^<1*hlv3Xutm0qS-PR5UK}io3A@G3TSV|VoFg?Ej(^B%9N)W
z2h<*|^`?<G(N9c1>)iu%gTz@C1^L0io)6Hw7){FIT()t?ETg}N)XhZk7DxW_&c4J$
zetGi^9YPZucAvNur!2%kl^v}P;~cxTE8oGN&{*T;T@bMcA?XP5fLAh|5)b-cpjW>L
zZZ~9o_WLZm&_p<tspOkUsR3*%u_$eE3ftTo?A7*nSPL+{=lqZ9r+J_E|JS94)>8@z
z=7(212>$gk?x`kZQNM-l)7$S`|F+O$ZUcT)E&v+Fp*O+rSl$5?8;}}*K!2ZNM*#26
zlOw+GTT|B$))N-V#t$XLc*Q;9g1JxVcA1_zNy?=IRp11Pn-*QX*#iSaYGb@}tTL24
zRC;hP`^m4(B^wI^y-B{u7@kL3hIh=GAfa{uAy^V5j?UH2W}*_qinngG!)fO!c6lzD
zb+KOSWAJoGkG9*X*yn!e4{8iGx^1GF8}2<Vu1UNrv@;Zcfu&cAhH{>p76biK_4yX5
z>X(13cAqxXWS>0=2_EJA2il5^7=qjpGki7`5dO(aZ3Ao$l*YN7s8Y9P^X7(F>@eH9
z)sukR#F~J4at{%BeO3H#-PwQtxrLZ)AwXj^**{al$J^n?PcD7hdFRPVZ}<@=cdi~!
zzsOI@VKDWj+q?^3`Ud0I7QLbAE5tsn-N{@|OelV^E9!OYaQM5%>xO|SrTZ2O=FUwO
z(F<c&L2ID>yn<1uSqa4bzugf+7Y?ndH%Sta8@mFq_Q1t9+5FU?CyBZ#amjhUzA8I8
z*)GRoogJbX<u8e)pL})-2j^yFe|7a|oP`c_`YFUaECw@Z$b)5Vbo7&y5V{2kJ1+13
zhkt~WXd_?=sg^`J_P@5{(bB0j$+eEGvTt8gFTw+-cg~-=NNVfXLvF+#!l*YIM%RVw
z2m4+_h1u)}H1$Ls=A#~&4W8RWP4*ICahi&R5{a8YAb07#+t~GOgW7+OQzi{=bJ3R1
zw(`tpEhQkEBT4yP2jP{@&q9qt0?L1UYJ0Y5_=$JHCgg?I)vz;~z(336xHx`7;_DE+
zKtiW|q@eC*m{uL0n7L+0#fVm=Ug9~R<7wTKP$vrN&A}3Kw5mxe8PcG)(wJZ`g5{QJ
zR_Dk%-#&cu5S@QILhBVdk8+PNz>OU0wcQ%aQD#Jy%T#{66+^0MtT81hoC%-!>)?S$
zJV%`8?zB~Eh>m&caj^mI>Uk+g1cbOIox`R2JnynZJp1*Zbr6ya2mUzJhXh0n2G|Rm
zTu96wVg0hbc--c^%*2mw(pRLw#kXqls3aLsqy__T&R;E+Yv1JZuy0%o6B5fC>hAz#
zN2EvZwH%rOI1vy}y1eVbZe2J0)I)~&#Llv;Q6pM<seYN^0mJQ7GtsRK!aiDzk~Y_9
zcND%s>nGFrj6O7YfLPF%5&|*P$zeJx3TGImpTWCORka;*+<%2Q<er&uTj!b#uZhIy
zzbSm^ODe7ec3F1xWe}gNv_7g3rHL3wNuOIaUJqUc<ddejeizx_-BDuRpOW->yS4MP
z5?;`a(*|c5G@XIh>4i6qb4T+2e2lwMue%gP^)18_iWx*3oD&i*bWv2djJq_va3$;(
zV^t}I*1t__6^x}xWe*BBiuA?O{=p}eUPG$jov13a4k4buD9*k65{6?NRjZ7296kna
zQ}mvR_Lad|<y6SmT)Ji{%FrPpC%(|FVSOymF-%lfTGgb<`)(E{{{({)9k~h6J-P<s
zW)#aPUrJmXtQ!3be@`gra;TJ)Il4T}wGvN)ncd1IBPj($8B0P<N`c$K!dTwJuthy-
zxCxd?zn?dZ<*whd2_!7{|3)Oa9JEuE?M;GEj9gW=up5`^c%_4$l4WQ9Q}jSp^aEeT
zNKmKGZbfzq@=_AH>Iquv&sMKli`@}>Cs5#Q@j=oe*jYji^UdW0^c78<%0${5PonAu
z6)d-SRbZPkS1uFkVs&l!>rN`3IMWXmD6Ro?YIbDf_hwj{E$LF0(yF0{gE-r5^iU;n
z<E6xB^?nU>^%8;un2r-*F*BmI6(r{M^pX<lSXtZCLskn_NwvL;z|7YqP^e-F*TUsK
z%PRJ3Z`vX3)(^t3d6D@|mkIS#(5LfD<69gc`fDE#$DP%5E$J&<&h@Oy*)Ss1MI*mg
zG;;AP(x+qzYQ*iLCIaWo;x$Ey?$ge=2JhkZs5Cnf&7w_F#mn92%;=rfyUZQTNeo0N
zdaCn)V26w&_n*Olie2XYW0#^19r|VUJ#~!n(d%Y0WFgK$qXkH}i_N~+G{xJ?O`0wy
z{u*H}EW6lmQ1$^@>;md_iYS6|io`Z+jg6bSe)yD8>S<l{qHrqH{CVEzLNOQzfFC$1
zg>5sJN0nRHCobi(Pbj~>Q}uOQ-BmPf_OdWtC4biZcntCt(JTqGS3DI(5?J`yBgOe*
zqDGHr;9UM7)WYnGL?#(Fp$fi4p(cbQ_)XI3!b|xs(zbho#Klxq8`0ZRsAsV(x&y+F
z{PXlb{LO&)hamGN-^ar%FaEcq?-w8DpSt1#Z%;qa(Z<88Xg@&cmZllLFVK1vYvS}q
zO1Z2$j{Qnw`JZog3u1KJ#L^(l$<b!zP@NI~YwrgrMsgk}`|WkzX!hojzwU+54QJ(A
zn1sAqdq;xeW+cX|OC)d_Sq+KUfbT1+b^ksk)_zUqJNzX0O7nTb<9YhOlgA^%e2k5D
zLEegrd`0n}DWh|h&rZFLFk6Bbffn;&ZUl<$Vf`DiE#mE>g_|jV+N)dD{wUz_d-Lb^
z<tFy$;IW}v_+0)8%8N5qha~~OZFSkZ`9F)k3sm`&;AB3S)`rPHKF(@VL@wLI9K-Q;
zFG>9?V(3MIlE{%KZEXkey3n%}Bf@>?+-8!r@^aqrrly|L(-J};%Oer>jF9y1HU;#{
z9Of`l>t?l4=RqttW`*p|_vg3kKDW8_hu3oaPg}2^<bTYxElEEY6iy#%gof+ReX+kR
ztYP0Cy)UE{cJWX;RBIf|nCEVRd^0U0t9&+Jerl6txjdTp(I<4E51_c%y(fTf+u2J&
zb{&}?oW;=-0q8IY9{Lbr8s3wmK8}4jcjO6WovGgWyJIjvAbIqc?X<W}V+1^5TS4C^
z0;08YOhm2sinBd69v3dS+8~?rs|2vt(4rT=Z*&W@Qyw*B&eNn;Yjb!hewft@Hi=#y
zGB$Ir+}>8*yj`Kr+OO$Ia#TsLSk$gm9|)qmgChBFI-I5$6O7F!hH*I@Eyf`p?)U25
zi}HdL*ab~l-#$JOnm;U>f&dKV?=zH0Q2ct90E5ZZWwTfR1>w*%gRPAc5UkS<;+@q6
zf2s?bMt*I06!x@VUH;kZ?kLJHtGgsDcp;pgdP`{MKWDrO&bTqdPPsv_^O|VHX|p%l
zxhVHpEhF^kjgsHnvWYtJduki~P1#`y9`9cw?BnQjDaHO?herr|TJOx&QL6LSq%r2a
zsjQdu=BG;blvgPy^zkjl;adp_-VC9DJ2OC)ME#$)x0ky%hwrm%on;<29UoG|l*;C;
z=js1>#Psz<Ilt$>piTqjmFqyWG|;R`u`?%Sv@^pEep#tD`|+3O*GwPss|*G|Cemxd
zKa6M51|7sluo7DstC3x!m)xyg=WfBp>FX?dH(z+W(9c6DpcG|8EyLe~6sQN_zuxMM
zY&bh5nf7y#V(u8R`bBk$x#caa8PPjE9KfPM)LLCKqu8~8mZj%t%O(j^1#=!?cMF%H
zzzbF~=Zm#St`~!@<LeBroZWAo2YQd>mbvTHTcG51LQv0wryq~0pq)WJ0t8Y}Lsn_A
z!+*=+Z3OS#NRBpgFsa|rc>t|bo(2V0kiohPKGCiV#X>CKIxvvgD_C(jV^j0)p2f{8
zpGdh`f`<GT_KwA2OEeBM9Oi&?2@J~L!hlTN62Bjf(G=R6x-nR-5{vN=@lwM1=DnNP
zf`)o_h80&}gpZ1_h+>NB99GpqsQyr9dYT%LsF9R?{uBCEPM7@i-0L9heRUVU1f$Aq
z#t7}IxXtb#G{2y{C&6$)7g{;~a&PJyu4=S8h)e4ITXuKg5NKJEddFF?5NOd0iZK1p
zNqz)~a1z%FHMrj-jGpw0;1#<DWDSZ|!x*bxi?MX((T;}ohcEvGpaLZ@3$6eCTQ4b%
zBs!j}B+rs$zn{3I?(tb|$f&3CEIl&Bs(FIkxUS9kF0?#qhE&YY(tasIwfH;Ct>qOc
zXVB03P7vEw{swAi?wpmOwxYkA$uSj_sZI2o1@9j}A~H0hh0Q5PZ6JC+;o2$L{j2Nt
zIfc+H_;fS-@Z(>hGWo8S8a*<JuT}d0{)!W}FsS7$MrhN<1vN+vFO>lEXA<dqv<9mZ
z%U;Zs6+0TwsMcMBzV%yuzL^aD0B*M!m4G_`pdrqOP+~Hv%ED!&D*9mQl-bFU3;VN^
z2!cUul^F|Cayf9xB+PcAbmMW&j~aaW+}%_|#EhWUD<!(!j?nc(E2;`fE>X6M4}9Jy
zWcDX448@}!-}V1<<Mh869-@HXaNNJi0iqm^&#v29@bbu$?JQV^()(X!e$fJ5)!{E)
z{hDUpx%x=n*~X<_b$BN%C`pfRg5eo(A;rHhX0$Jksq;&j$^5i)Jb6TnmZ1NA;c$cV
z^0|vEUr%W_KRGH1zleO_eqk~G6Eegi&9wH-@9A4$rKQnR?{s921{rk!>|`+=X9$EN
z{~|7g3%fQXR{ti+><vciEHY-uaYUdP<-Ef`9IzttM*hcL*>(cgZ*0Bq5L5vrb_vp9
zOll83Xw&PpBxV|Q|BtqPgzhZ5o>*9zyo{)9!P#X$8j@NUl+hmjB<L|xn+!Oc{Ug3z
z%M1%+&6}WU>wZ)e&W<eXY`!wIhhd!P3Xgp(z3^%PEs;>%7~NZTQb>`XKOtm1onlE2
zp~TR{hEY||kX47u*$#H^yo6I6@2N|;sXs!tLi|P5_{XRJchf&K9yI24Nv0{Ja{@y8
zV;H3=zL1-*WwJ_G1AKk$zx0OvN-_;I0XmcK199QXNd$>kkLpN*g&`h3Va!!!_y1`A
zYA~GRyV6z4qPyloKLT@wMuy7)YBAS%sG=iRzPtYkQIS6e?BBaj;QuAmGZcwx?VuHV
zs4^5r;*?N~gV2a__9ZQi@88LMJx-}Byb7Y;hu8Dd(h{C^xQ34)Sg|$*p|`?>497N=
zdKrz86Kp@MUV<9m$+u8Cm@Wsx&tjyB`VIwGhKLR4t8E|&7NFciP^Yp69l(`U7>nf1
zC7q;CHn^?@>im!4gU78OkoFSG9lRg??a_#bTz>qOT2|a(h5nc59bUie`jgJH0{=Fn
zgy=#83#s6<B;a>D-lVLoF>c!^XN5N~$dm8U*K7P)j#+dEye+ORu@17s&6gdPCrxXE
zJ_En}o8?_kq^{dz-PZbbd--bs451eTsDTHR)w(3r_ObCCkr97J<Eid%i?vO4K`E&Q
z=M0&U53^oeLlcD~!e&7j&Y>}gm-oyiIa>bO%?>y;8Q!JL8Rl2K%bRW%OL_Mr_w7PA
z84|L@jYoArFTgqkPW12KG@|pI!@O(K+Cr|~j^f|eacpNfBbp;Wn5?4#DBiQDLfr{+
zv6DHqbA8Ine0+O}ZFX-z%BLsZDQ0*wFx{^oJZkm=aZ=OgW2n)PGU~i*%HJ9;7H8*>
zw*|7^I(n>?{5x4$O!Nb~h)3o+LV{E;%0*&WAi)ar4H;q*Z(i2?<lTMGM=^Pm9A8*k
zmlpaB_>+Qj@e*xje8>=j6L|CEO_EH@%Xh@2$K`K57Uw<joKF>|;_Ize>PL>GWDpq<
zaR+0$pH~l}v!6WZJ$ZG)bgNV)y7v@Jk*zCjP2*hop*r;z$$!5iQS|o6-ueymiWasT
z((tsZACSZ>g$<b069ds27y5NKsW}W9#ZG-9)FBhx--Xja=Q=y8S4T5mp5!=F_~l&C
zBO|e#kY^BPuZd!ULA)1D+BH+Zq;z)0HvB`HSozG)VUG#+?r-mte)vX&ya+3!Bmbv?
zz`5nsu^H!}>=m0MR$EQtDtPGr1f`sH-o{>2{;Qs-u$pBr|Kqfg+}i;6^|pXF_6yY{
zUVluvaoY4d`7C3>T8X^Ap7ggU^xZuM{_{b+kO(GB>@1yTSP+RG(fn)o?H(WF)PyG?
z=UR~?!9cF6xK20$PeQAe_LAKwFi_8(MQ-2s!jW13jtx#}8YXJ`R{f;q*r`Cke8N0o
zCrOr4LNe>bkl1nNcE~i#irSi%%rQyn4MCn!QQ!FOFRRacHr(rv(!rOr7*0u)O)7DR
zH>4-_<#+c6N!>C}OTJV7V^5eEIs2iNL6rysL(V&qa;mj6^1Dz^d;)<bh(X}F)9^>L
z{(r85!MK2dk3;OCJXMV39Eo)fD?xtj)Y7JN8vE7hjJ8V2em(!x3BufTq!*@jEdE$(
zNd)e&i#r{`RbL2UR-@rVD`mF!Sb4y`kGf|W0ipl1Czfgwytx)n++=sBDUMcoguoj{
z*1+C9Nn-w!DAl<t=DJ<34vA$kDEkXYuA4Za%37RVnA^4#v!KlN5yRVXa|fMxksRI0
z4W5t}8V7qAyQ2)?hLzu@Ejg+@z%kkJpu(p5wH3|x)@J>$-UXhB(?1>6)yYeLZ-I@c
z&yM_zlIZ%yerm#ngqYG_aBM(tpR|Ejo(g2rehIG(@0~ptYK9MUhL2(`JvftcJ!Kx%
zvWu0@DwO-)NvBWnA8=sc{WoPzsVDYX-fd_?ak$lV2FlIb-Z$bfNB@Xpfe2`Y8EK9;
zhIcn>lH_9ir@aiu_w2Gw$F4mL<hx&=YB=lrg>K-Qp%#liaNqK!+dIX<l72c*(azew
zKY}~%k~j2rgqaeZ7U#d-qEI+#dVQ~Bm{MsI`#I&4d|LpivtIpsk=mVWbO=3Kx3($?
z8v81y3ixolS+2-Wo^gsc&QLHc*gwq|+an}Ntktff+x^cg{TSJ06o&eKVKIrRn5iKu
zfgCT&^io0p$5}FONidGo|3&*19&_Gp?2$?XFj$iP$j0>LS>iiiI$mCS(6WTV*7iv9
zm!2V5zJ2|slOk7AQ?|#O%rE54942rm7pguUO8;RMN^tpD(>UUk_-6F6%>`whcIh9t
z<~RE3H2s&JZ$AUIJvK;bK#J0GpJ~ndWz#eJv>cA}5HdaI9`{-z(Il|L0l2<|-ayHI
zm<;{(Pq9DHxW;6@7-j{*@XkB)wn88w3&)K;KOuH3%!h!1xMzyLPU-LqdFq(>3D9s)
zIt$(Ud(3?)b)dM<wZteVUIy0*pi1DKHalTwhRB4a9Pl(qUR1>;+~xrnYC9azdeLw!
zg?mmY<|)j%<vh~ZLS?L>zr=goymN@bI%c~jzheIJzx(<=jg!rwDvrt}Vh;F~{|iEe
zrcvEj!H~<Y!?r6v{6e1(fw8u!p|U0Qmqgk-fs;OUGpLxw>x#>#Afdnjf$5N><i9lr
zTc;hLaeNe38Tfq9j4o6=&2$|5sigrM(iN@3JRf~iCn{PO);PfLH`8?Y7qAroJGMQv
z%k1ndWq2;9IN$X&tQM9bZMa1DL(QXp`D~{Gvh?6T&LzD21g|4T20#wxYN7fzV=1h2
zGB?^_OuK!2*4rHiyqD5NJ-#cf*?AhAOu(mf1Rq<c>VYa=nNVW)9mcrWIOu6ou+aB2
zAlhF@)1@t*L+kxQXK+I5Rc{@J;$MjOxm%S~FoC%d4HzoVw*d(!Y+y?Y7n~HY&EayP
zzgncX$TU@~;>!ArBcMqcd^vM=JrxQ1%2>?hqXJFlV#@y9nl0+pkh<5pmj&rz;Mv89
z)Qfj;5zv7gLw<NM*kT~4c5SLDKqAin1(8@sQ5Hf5a9%I9MMlEuDbDR0HZ_f7O3pTq
zV;`TXNaaz;S7G4WNlPGZVe$YzkLX%EB&j7sqa*fJ$$wDuxAwi_VP~*<#Z=y{oagyr
zIs5kpM^7eydm*m^cm3V3LsXwnc)r>Z#_Yg*Xp3yL+p3LJ!fhSx*03S&zg@^VeOdC}
z{Z=U!xCC#PdDnGcx}Apbfn0W*QM2gpC%mm^ryhi48D$<uwjSJU7TWTR%C|R6?JNA<
z=C*w)pi<<Nm~)L@c$fj{Mwd@_Y4g?bSc@~$VWLnRdnDq65VYDZJFvA`aMX*;*EkI@
zL!2ykNar{2YfPDt<zq3T)b|!QuV|VuiJUw4k1BsMFz|w3e}$b=F`q)u0=2cCu`8on
zqA8>VxE9LPu8~Lz?r%S4=%CL`f3K=Q*a7}t@A*<w1_wIri19kwFW6bC5di~5UWYO3
zF(36Od3s!W)n^_5qDb|N?`cJ$+e;~ko8$ek8ivGB-MP1+s*ZC<w}__btv=}?J9V`F
z<ub`XEIR)N1SlV%oE0GBN8Bvvq~-Eu4<_{2ou<7;=?6L1QOT*`bjthT?>s$}R_&Y~
z^f4LpiZ%m#mhmSfyRLx1GW)`~y?3^}Tsmp@u`$2Z+b^p<T!Gpiu@99BZr-k}do7u2
zZK2ON$o3)Y&Z)2e#>x4h{S4Gf7xatgLR>J|ct3ZUmaG&F>d&_A`OEyoQa?0Q?chrD
zgQHSv5Wx)Uo++5}VWn2%N!6Rh4TCB_{#5Nn9mdG)5A>}>2|!qYY?sK9{V7R-MN@4-
z`34U?M59s{jk^UNIjr5|FF&9|1jCoKQ%HVY!A^x9!8w;lOK-C=Y<rgGSbyr;YDha)
zV-=rn(U*k$KW8NW1AlM6dxQxMB$zOEH0USht?Xw$$^C6bO3W%@N)j~^3GM&H_Cc8V
zIS9-I(9cQmF57_GC&(PBu~H_B$oXl99wk%7?#E8TFJK41kB0(%oZG1NKq7|SC0t=h
z$%EadY4|DFW5-C%9g5dC!2tz+eqw{@J|9kVxP}K}aR;?>l75C7$*R)#)OV>Hfgxff
zjkFiMyfLE3gMpLuK;3y4-h-&dps~jJ$W5~$r7$3aF{v;<RX6qIT5JE>##twevva2z
zbo+vwBeXA3CP5FBPc=^YebLN@S`K65gmdiZ@5@XHIKi_AAkXUq)b*&K8H%jzps7aM
zZ#8|geaSW@t)REm#eTi#&*p<O*Xp|R2A*yKh#L-}kcYc6IXB8b-zc=6n9=i>-t_0l
z(r+V28AwR*{GDkY2WSa?luu(u)^<2()26rjKG_0?jx(wH5@?C2<Wox_5MBNkT4#j4
z&7`)9@j~de=8MZ0(Dk1?KL0%kn}=82-j_FS?Dd{Ik)QUaQ*qFrtOAU)Q89v7fU>eV
z<<n|I^!ta-UVH`u3BYE=pQ+==dAl0L;Ti?Mu25mhJEAh>-CU8Zrg8BM@F6&FYj$e=
zPg>`%9+CXeZ-xvEks}u%yO>D{d|^D$w6QJW1bM6CEvlQ*TFVk>uIDM7R|CsV3Ve8>
z)kY7=LCp!Zn(<qQUjL;m6Z~C3(QJ6b5gP=2nB)IF<q`QRuP){&0b$D3F`En|$i;mN
zzk+Z%Z~*+E$*`9`XY$MIkgD9V0`cj><vm0#2(dXRDH6f1`8byytM27TFtIuqBa@}n
zKQuL@A}7TBEu$WJ>-(Nr+Ugb-j6J*#JiY)ST^!*~yHFW#UNyod@<RfBDJoSt$#te(
zeInVQ^qOD#mOI`caTgOasZxyjtqY!C==Z<#?r#@l?{38v;pTIez)p=9Dla+?_{b%v
zP@)J^u=wEScb_S*EQh~6y=i`fIeoab&LzFa<E5v%W6yeAA()ftlR{@Xca8ZrS=}<D
zvRA9IgNXY^8k|>5b$q=i%)6J0z#!sT3qNSojN4N(zOD)imc;{QfRvhl9a@xkk|Uqt
z;XrRzLIUgVmu9+9Hd5}Z=1Z;(voUL;!{4e%V(;8z78EK)h3~MUwaLCn;QRyL7!At6
zmPF?P2It{;Z|mr#Ms{@G<pvurf@H#zB*`CZEczM=FV1WxE1{e>Uo^{a&EkrgLvS_b
zu44gbr68d8eV3z>AbF;XXExRB^EJ|5fjC5lbIZ9C1+rkgVre3^HKa?$j_yx_WImm@
zo;Et^mQruNoyY5#kQxNLP4|oT6UC|t4<RAUe(jlrn2&r@KC@`!*jyJxejPh~i%Y`w
z2YCgmRR+oBY(9Zp*(Y!{6}i|x4bI6eaxf0z@A2cX3+<A<Q(UKR;gj<!a!n&?q8C3N
zkCh?rjbqK6xw1W689(Yf8#x5{OLcY%L)66%AyBa}OG6M*fQfB^P=i8<N5T6yhebx7
z9ZvOl60$uX#>etE50&EQodvVtu7}E@{3NfJ*g%zfLcC$sAyOt&lDI=m|LjamOwfU_
z_^lihfpWRD#VLI>F=VO~irt;diahy}{7ct4NR0KPfvCvby@-AWiF#n!V@_I%^GC$8
zx!N{27P(oesw9bu@3r^eOWSxYZRAj>wt9?3ek;M3#&Kq`Gs<<OPON3eg^o`?Grep#
zT!X~qMMOF8uv*i#3R(8dj~9mbI9wxXmlF@yMQz76_pV)j>JV02iCse{eBcDf<C$l{
z2H#;2Ckjmv2`4CpcQqW|vE)DAsY?FS<<InR6usH7!2^sw(I4<BFT#w;4xo#VI*H7`
zJ4QA4;*1GVbY`wg7kB=%ys(|1L*Lt}k#*>|dn{nl0~;?IpwrMBm7}sAYf5$gjD~+=
zO7k0|<*$%EO6OkHX&nWRAA^XqRAYRY9ek8(&ZJ_Zd18p~Uiy;&zV!Iimqq-L(?*)}
zOia{?V(LB8*q1MVW_k!sZqtRni&AzI4`a2`IU^wO23mKsI!U&^)IHOx{+okaK>1|Y
zSH&PH6A|qN2Z2~M;oW`(1^KjO=$EJ)VGNT`ONE)MdigG`IrH-Lh5NZ7IH>+bZrt3y
z$V&UDMc>71A%7nwDg<AVf_w-*cfd&2CShW*x<&2>o{hy0ZdO~k4A$!T5~Ud2y6`o&
z06dFg!nOubd7A)-!c~&s%mIPV;{souXiAE#Xw!46dkoQZSyvfK&zx>rBHIC-pAcPF
zsRJ^Hr$nN8n}D8pY{PTK^IGp^ss1%bb^LhK@l@W*AVK+TRqk2$;Ka|fRW`o|e7Hy{
z^`cI^0-bsgci`1n6^Wr(+Wva;klszja`No0jkAmTJ&nA)oqloD?MCpKjgNo#IjvLh
zB0RmfEa}c0rIT4%SIW+5K_D!no3oVzI28L#s4^v9yYP!~(H|GC)7wAXI~BV5JFY!^
zn^mzWjz35l2K7L1(!>(M-m>R{i~Q}c9XDf95IE#$uD-2pA94u))dC1m?O|SX{{?xq
z=S&i5Ked>D<!ZG78?dct<$5CS8|VzlV)9GX`NxKe4Oh<$5=ZeMUn-VAVmD9c7F&v9
zf(W)gYA!fR)_hn>$!%{~$HXY}B^n%RyI^BW?RzBI7{nBFte+tJeX^dlLV$w^AdkY8
zNmA{@$ja8tnJ8B!Z+nE@+x%3WfMRcbpMIdN-5WQyPwHs$+;7ii#@;CYYatCEt0|ZZ
z_}{n&H1ZRY)^<pWw5WN|XFv3!sUv{!edg3?DBae92@J8FN?Rdq3iA9`O+6=aN?PIU
z(F~H91@Cx8P)+;?w*Fq+OVn>u6zxiLEG2}*javs~qFh9-WhxOqw2?VBx#6>D$nj52
z+V0xhppqAUgU&!^P*{ZNLr|=n!9JPSh0CQV0sc%J=3^)?Nv;EWD1*#Eyf=5PPE^>k
zwwwN;YE+(B_Xl^hbi%1!NH=RAau6Ly9fz@EJyDB3Bp|}Q2?hiNU>eUHYDm&-sVHs4
zm=$3&`=VOQqe=A{4^uh+bTvMg3Y|utoS#pI7P?^iP!ww#csBpXxvODLC^okoNs8r@
zpwjzV3-gvn9!uT68($I&l1Ti5g=w>ykiA=y7#RRb489@}b0$X{ubz=Je2oQKvV~aU
z=A*BmqRI!+2Ii4K0Gb|zSwD+UH129)aUR+=h(El}JG3FWaP4LXJ@fO1%NOe;67Xzv
z-N12%%ODX*{d|C74ztAZU2LEtG{|`b#LIA@QRR1i9n>$<zPG0lkNYEXEu(m~`XI-g
zf2KW>DlKaCc)3tuxt@dyY?TbW2*0qLd5Ze0FVLN@s)Xm?4b8-h=BqC#1EJ0dgvKk!
zB`2u`1aMzyi3;a!(~b<0N$*?C6UZN6<V?BKC7S(B-SRa*+9wm_c}wls(YRp$&kI`~
zh|6WwG!t6;JIDNJ$})BQ-srsHMrlcl|3x}`YJpExYNT2h$`*1EfDI&PQo_Db4!^6g
zQhXdU556z;arMtJ{b+Zc{CWL_#<sqW))utOkpN3<JBjSVWd*PJ<sM}MS{ViL?LGj;
zEYxeoZl)CuGd2d&{a-m0)93NeK0xF1k8;8<DaDn{Q3l!QoSJLDA7*L0uWz1Q9zuq+
z9b5wGw@Ni!cP6nhS9w035)0wC8bw0ag9z(^?b;W9$xvZm>|JNQa&$Gv`W;`tJf{Qa
zFX~z1>6afCjtJS~{8tObVfZ_)80N^6SRX%V)h<wEO1firGDJ{nSEw~OyR#af#Ill6
zU@}V1_VAZCf8Io}ddku+Hh6mdxUYbeTQ}JnzzJ&A4wh2oIpCJRc^y+HK3Pwsle632
zMiHHFFJ<n31<dCw?Z<G!C&-JlGm-h5f+2uHYfR;r)ugZ2i?P%}fRm<at;!p&v4z?o
zdFhby1iYj2BQVA&!vb~)bea-kI#})Ar*d}4IM%k`Dr)+56r>_K0H34byevuox8dI`
z`6I=WV8-v3sg0RKK5TMFGJhNEycav->BPdA{K@ZyVoDc-gplIj77nbxBg!ivM`rv9
z(q0WspxHP+6--FLT_r1T!m7K_p++B91%-x{OdK|ES7#2d`@X+y%N;1)`(ovWN9$pI
zmme6L1eGR%&Fm(2RHkHex9HiS5&v1wPmRD%ZQrA4A!#r?+kNu_l`s^3g|d2kqy!^E
za;UwgGlzr2hx*uwm7?+|FB}PgLJy3P{|A4d*(le6NSX*~Hb$w*t6j*s9$BK7HCCY(
z`<fw;g%^G1TDRzns~Ypf9wniP+OD)!74wxlHotA&!Tak~&&s`d*K_gi0$PJb!6%hL
zbi?)*<lg{zzhTEl(eI7Ns0uP0p^fs+%GgwCJ(b2f@1NefJ7>B)ns&M}C~&eDX@Vq}
zIr*9^K*dM{JAQgj?Dtn+@2|dQJoF>mI*ZNIppf)9Dxk4Y-ggPgl#OdDyxG4Z-SMqV
z^W&xKr9XMN)Py&w@|zC+c+|vH*BxY|LEUIciS*M4fHdW-diml#eddU-zDAu4W6k5A
zKF|wD7JvgHjy^V@r4)A1v^4gkQzy4)7Ec<91X*1IP=+QNebO&w<IJ*|+*U0)lvMhj
zc*JNiE?iw+R)z&Li<5t)RA0}1KV>-@_dI7pBI`kJwJI8GQW$EMVLa^Lr~kckXhxLZ
zDv<fKagmes-4|V}!__{pv)Urab{dyg^ur8XL-lCWSpKPRgVXnWl)U&;uUx%?OTcf1
zT!8F(O)}yG<FNNgEK7*3$tedrgq|Q}=3Z>xW4+q~T})^0=nCS|bZSE9P|mZwe`n7>
z1GWh0q0V-*1~h5jX2{8p@AMf?Gb>@H97xX+<N~Z}DM|C5BgxTx=-GP6owBlWG-osj
z(5BnLp*w0)>K2kDK9o9x3=uCL3EU>Lu4>GlU88$1MMffe<L{yY+KAI`;gP;K5;UH}
z_Vq$74{rm_bcYIpTiHX(Y`j{EkC7wlVs!u9oRFq`BYRb1m!F{7Q!BYvV1pSh>jmfz
z)nKqy9Ox1B?9Fg{;Io~4QGP-GWN04$AVXd#lEO>aWDhkO@!w?W@Cb4aFRy1-!2MGA
zm}kMWv#;B|*X35QWrsINcm|C*u&N}S@p$^}DQ}^;vyrWS8y&>5(=RLMDs-rRbMi(P
z?Q#{RzQncW0<G|gqmxLSq=&2nCj^J~<w(dXpC1*9(1wxhXStN5t__@D$*9RlRbtl#
z*K^V0(-u!Mi_SbS+VD3=Foh3-fA<*1r98z4V|9CD72sLc&Rj73J+4uA`O{<Yv@&za
zKBUdKjb>SwhA?oK%Eqfw!(x^<z_Bi$P<f#9$lEtR1xQwo>sZok2Wn!!GdLF~;B|NK
zQB_OXh^zxiO!3?8Xnkq(^rSmRyu2qLZ26v^1-<#LYaM`Vl;})=vWZuxSHJXqohWW&
z8;G3L3@u!M$cyFbK;}~xVZ6;E$a<P2GWATT-q*>gNrbUGlS+o!IWd<7_=3f0<oup8
zz&#FHB$K6%Y*nKX<!Swc?=<x{E*J^dS3T?v>3>Xz(*-o*Acq!l6e!?oAh=s_z#Whq
zEWEXhKkxd*=&`?ZH___693B>@z16i$0)YMjjWt7&OlNP6RQG|yZOm96Pm0<lqh6O1
zZe<<#Z|X4Mawk?h4_+lHPv6jgF+~1;ux(ejpn8_|#PRTjAK)EIUjVE!sSU+JQS1U_
zdvQB=!$&nPkB)EBD!k<x9>Rb!-%R4Ad?7(Q)`93HrHf$VFvb^lJ;dc>bz>^WP4xAz
zC%({s;NLGjB;O%*#a6>&eO-e;*B1e~JeRNGuNiwK%Hz(Vg4n5%-PM^0#n3?DZ@ozD
z)adI^^KuPw^zj!XnUG&5B(K?>vYuzn3g=uFszkI6uu~*Dxf^C<EoR1G70OlSV#QL1
zHm}pt)gVu)eh#c8C_yg4i8O=lC3rlVW+S_#O1bxlBy!q+jrgj=Pv=l@C`y^aW!B;t
z*X7`2m_5eJ(t^GjX;oAA1t3?fQ?>v79gEOz_}Fqy)0rgE85k}#TBt2K_MmShPP7J_
z%Jb_k{?t8Tx(KP?z{LInGRcP2UsB+%(MQX7dN$U;7<BEnmEIfWQ^ssM?X`e?d(a<8
zB?9i&z=he>jBy1@zD?|>cj(>_a797*D_<bcf)MAmNTL_(s<zVTZkV%m)5ud)<>Z>o
zpAto}crmwgmEb^G=xj;)xX~6TjNs%-;8j;LIdIJN;`u*Ec_KTOl}{zE_>2T-4cL;Q
z6TliT27C7F>V?AC=04AHF2Wce@)KJ*UnIAfHv@#OM{~Dz;Ae0lZuY{8>Ns~AGT#Pm
zG%stxI3begH(_oK(TWPG1($<hb=ST){;ba#jvet}8(24#U2uKJuYh5~C0`A8=AdhJ
zLscShj2_nsj8ewNy^noG>bNCe=nXJeWhJ1y3XVPXw+eBBbPZ$GP_`K1H-FjAln{L`
z3JY}fvOty~{`g%YOlC!hUDFE{q&3nkHO{)|yH+-9bm2tE@iTdd(4}>S%acwwex*d3
z0294F{Yf4R0<8MSu!g)xArmg^!Lxoz`j!8fVDytNy*sL&o3=FG2~DcC?e$n(pRNq%
zoK8p}s(0$Sj2R?&6%K2PmLev~?B3h)F2%-lvy@lTqp8nzgxLSk`es@4>(bdxE8cO&
zu!*mQ*gF^XFfja~Uw4?;x`ywsoNJ*&4x5lTM^rL$#Hv+vMNW8CN}}(wv7KQudY%0N
z&EYp7Bx4_{NBXqScAF+W9@*V7<&<|NA;%Q*6#9YCoPC52y<bX6G9@YJG=2HABL7>f
zu}%DEv=R5uoxl@IVoAvl1hWPi5_^I9fxgjwZqtgCdoe7v<#Vt8`#KOe!++*7E9bt-
zJpFA(1j)8<{*K4)<`lI+r<Lnrn}iSk83aHta`15}ywH5vtu|Pw9mBd3{IvPrdPzb>
z`z5-`HY0}95gM6`pd}H2;6ifS4wft#2W^&2N-iX`eVqo7N&$8hCbfeHK11(;>kOw?
z%;{jF`CI6(;?>*<QQTx#tmdgAV4OKSfD+FpMPDxE$G*QM^`Lxy+sqcSB}8$Si-q_U
zOrr2l5kv<fJ~uLREcXz$N`sDgmU!%iIRrEyh-fZBs}UywzXhhUe${7cOS7xdl(~{w
zir*YO#{a$us&cmN+&_!;lX<S3Mi5e#{0C^ttF<{2|L)K_Y)eIsgz??Nw0S%lAS1^$
zc|$@xSSpLn$ZK>BYWHvJJ(7m1Z|hsLsAip>8u)?OD${%dCNo~Z2j__@xE+LJcn|H;
zDAjDMpWqQpwp)!ixo&K@_;$mO2d%>QzL>v$eg_1?_2-klk3}3hjb>ElB&z$g5;QWn
z53{yZOi@_em{4xgH><(>@tX>&TE{zQ{nO#%KhecYFEQ&rfO28>yR)_Q9m?t!tlVyg
zxS1m6CpV+BW5CC{98li27W4u78a9{JEC&4-Wh-<T7Eg6w*^HPu*Ssav{$5Me>WxIR
zt03WCe0oB&_H7`i;^Ea&%yRxhLD0CM2>?lX0xbp8;(zd9ypRy`Ev=_Ec-XBn+d(R+
zTxJf|#QKg?d;k8FHFY`!X#VyDg7&*mrxD1_xejvDr(QOMml44RX>*Q$ZQv<cOEKXf
z=_OEhJV5r+qG!*e@ISl&%#Cj#?Z+Cdz4ItRxtS`v$}*?T&>s#9z3tg=*l48>M1&49
zWPmP!TByN)K|b6sq@C`dDevvKVa`n>rRGZGKe#D$$t--mcFXZFFGoMQNQ>0MyO}WF
zCbU(C{qMDC*>e8RTKLiiG8gj*u$Q){3%>nP?%R!*_EHR6I;;3GZN)o3XP0hy#N0vo
z(29}^->Pj_lNj`RQeKB00i=Bw(}|N#>ef-7Uwz}dfv-r2Oi<uDH(tw(z}-&Soo1fn
zCfCLW@%v_fv=`00Kl|P1l5d#&tSA9Su@+PSxVRPvqbLB{Ai(6*FcJf<oov5VQ!DsI
zQ==vyS(Bma4eW09b`GvCo4u|V6*zLC{|VMi8;F7iWrzO@LJy{$7xv9!S{VFGEb(5i
zG-V2=?d*ZWu2_LtcUsofCstbWSbR?1dgeda@v^lndyt?srqBObHB4NNCh1Cezc%ye
zgHh+&k!mA0sp2ANhG;}7@=2S;EhZud-J&Sjmy(QH<aE)lEe9f)0?_BzO=}lC`%B4a
z(P*)(3eAv0ckU+$q^@z_>HUA5A~%7iMUsZmg)$dSy47-J98}~r*421jF~Eh?J!!K~
zR!ne;(9XwPIQksU=MFoM)ha-J*BI2=$h}ryn_d(26@TJh5+s2>GT}u~H-ze)dmihQ
zDwV`ViS*4%SWy%jtPYO~)fg$Q&V^9CEJpZ^f`UFz{ke^&tDjL8JGPgAP=3fwv=bD6
z#7%m<cEg=m4-w=D&SLY~J>5L9g-(73WptTTe?Oq%|C%e@<vz&EdmiX0!3z=*Y)9_U
z>2dDlt|t=u0^1u<(bBsz7OSS{f29rq)ybWk3;*o&=Uy=M|Dhk?pxb@I`LD!-#NGL3
zpVdq#Zf}TG=MP0meGy(tm(~@;Js4kZiS}G**_ufY^1fl><H}L0(`l2;-Ze6KIvptl
zz9hv65_MAiIDL)1&k1v~X4_EonsqclOi<C2{`3Azg8vX#X}@!>j_=CcXrBy-Ep~7n
z+tCxp<vvdT82BI{p(5Gm1rqE%eM$op7{aN}JTK<;?Ck8yIvZK$5wC^PAde68#6fsD
zZD{sP=nZ1_jhiD>1&<Afq1m-GyB~y0^x}~(OwY`N)O2*c+P_z3fcX=`_gglP<WRTC
z`0A|kZiVRiV1>%_2IZJgFwS!)$ne>0(ca~|b2s>Zt8rQ}tGM;hIjke-Nt-DBXP<$g
z2sHr-vMXtCtiXdM&&%1-Wt&c>r`|}jU}Ne(0FRbm64r-!pXrR7<#U~kYMR%+JoYJ-
zo8^nvT59L-weFwbslopW{D0;qa2(s0{{Q&{?-r^C{|Bq5pfOe#o<x@fxjG8wq{Wmh
z$iG<IU}US4h!O~s`T=5SmO50L)doMsBSwHu*1V$Edq_&^s?{m$WZ$#v$G)oY&X(%h
zc>hCKtdHv(ji$YV#HK7?F{>S=8-8hZ*>xZ=6T35qNW_{!aY@W{7&p&Kx2w7$#=hH9
z`Q2Yzi`mpq4l*<V!x?(4F%iaJ(=<2n?Q=AN!$>64?DEIQcbbBjpbpct!nYLXLk&v%
zEz(BRS9c3*+uwipeH^YC$Yu5|)OkPG*GNeq_Xqt&JOC4WxzI2QJHd-PRJVZyedSb{
z!}y!HBl3>=IVy!ye%AY<U(U4zpR~xW7P3DnJe@eres5g7tg2$TG#~4PMkr@>w!MJJ
zC!{1Sq0O{d<kt0wa_8Gg?iRx`d-1F8d^~g37i-^rWNhXLknmJ6a_kv+I9YEN>7YVV
zf^mc@v#mC~Ou0YX=C0&tI+Wif7Wb{dI%RNg1u7Dug(7yfCb3g$3CjC1p)inww7<%V
zTul1`A{MMkv)<R}1_XT23Vz|Nw1IG$L7MRLL1&|QRP~$OtM!-EL&tmzUqT#t_trOu
znrk{_Afd&iUZBauT9Ww8$E(gtd&+p>JS}WOKgXtXp4m*<j?4|ux(hoU4nbntbI<QK
zxg%0LXZqU`e#aq{KRtaQtJY=Vw{GtY!cL**J4hk6<&m}{GHYv8;NUtF5PR>oVONx>
zgmjSd4RRPMHx@SM>B2SS9-KfP_u~utm=lnG|BM9BBb|8`10}j;1TP3Puf(BPl})^K
z-+U+ZQ@_={L0r!F$h$>kfCZ}We`#i*o+2TlXd-_b_G84#Ypcp;g{=#AY^@$YX6%&T
zj~%FO3R+rcCCag4I;MmB1-$OO7#q@y<|$cnNryjB`<{|xy&d$W#c6;C8HY%?ApZJb
zEDzWrFLr3c*{i<(5OLy{S6k2pv)lS*%nNNz$}ld8<czy>slKTFR6$Q|w4}33I%_b5
z<N4}<rJYbEG(q!*YNvpxY5ka2=~GPVvmCzI%zF>Yx>pl%_-OO)4=ibfjI~X|I-+ZG
zrud^t;k5M%w_HGpbhG8f0O>nl2Cn;L?Tn)%z3_6ax$H;-0%u`N=nbvcTx9v43dzSi
zwob?i`6(B=*a0m-jT5>9UP~eB?W1hXhhHf>_9MGC&kl<Pu5a7nT*bHF9Z$<JT4y``
z`W3r$hiC&W^r0sB0m08Lgv;OZq9?fx3DIWCK&8}#JO@!9(2Bm6#7Su<da>Jw-XUEV
z2sB4adFT_yc;@O{eu<*>_3nJtS!U2Ov%H_N_60EZ&fACHs~z?PFYkN4-k{KB&r7bl
z0D&e>v@z6uV1~xbR^}#D!}wP-(dYyJzQn!xS|4m$&~435*}J9C>pnS@KESW3Y-y8Y
zLyC3|tu~9k>6at`1RBS%b35uF6*+lYYbW!-0<e7sx`ZlGiE^3`i>F=93hjUY`P1XW
zfl+;J7$EZkiwC3_fIz(|@&vgP$cBA85`T>kX5zw_<O9^zdM69(oU5nXR<jP_12EQ5
zFG4&<V0H6S%a<a%iTrzU3&eZ(d;yGwJU9l9!j$6s&6L=%&xn=ri~x(@ntcMe61jf#
zb3>U_HK)I;IvkpC5PCmm#=4^5>g!q?V9~$bp=h$e#pL%R8H_GQ#y`>ea!*rl;Gyg*
z8sjkwRb~o(-v}=lb{IXM`F7NGYGNImLmYEIN`8RG?g27YTIb&qheQ<<mF=HDIxkOp
zNj9V<-{Qv~K0Y!D&RCn+@&ai<1Q}y+OXQ;MNj|lb4_=Di&IYn}362kV+WO6I`vGQB
z3^s&NV(id@ts}8~S?pDm-`;Rg=Ep<6f{A4-yf(ZmTxNVvj?WE*cog5Cu@=xO#WucU
zdjHr)^iFGy;oR>B{DNAiAbz%mYh6U7+{VS>`C-|0*6z4~+#d>ZkM2ZPfy*(o-D{a>
z#FHEG%y6qwO+BrTy?$O~>k_<|Vq#nT8$8R~F0F&WgEQcDJddFL%u$a%Kz^Gtsjq>H
zs3*{DQbO21`Rw*!oQ^WX+1zuXa>|J}aF%w`<38V5r&_0IcLKd+_J}-~pkwpHZ<}oQ
zjR7ymbY;4xkNI1L{%;1RB11K8LWGh#3>WN09(Yx4BG<PaKIJY``@S}yZy6q-pqQ}4
z%lWU~pJ>KmZ{Opi8dL2j{q24XHu-U?5uTsLeK`p{YzN``Ee4!Z<EX4B!l=@&Shdbe
z`D03cR@$3>eMeI$e@G@|ZUg0n5E945vi<e#{C2%twv3<U)yc<BO{-*Aid}K2>Tq?v
z|6zOpc4UT?jAWT(>u5RI;yd$@$p66Cud8!hTZyE(&t@VA&M%g5KWR!#NM2oncMns9
za+G6t*O<vB#0xl@*EQFchd9^nOod!b>zmyX^bU~YZCw%o!b4F#&Vf+95!q+C5M;)k
zva}!As$p5@=-c$=W0$UYBk|951eihKgf5DmqH6O?G}~i_HfHRfp5cAHomaFe+~1fe
z#D-10NMb2(LK|^cFwv@a;jNpzl&EHUBzgP3)uMZoxLAn#XD@leC51m+J(@fHK05sD
z;eA5qPv12tNXcn&r$2xX303B{_ZiR4MsBQCf5E4cH<te>d1v>YpX6Yk9_Y>OY!=d%
zBT2CV`f>e#L24&}bo`K>tUJDU-*1lgQ90!Arbz)Qjyt=y;fvlQRY5bcy}b|60AL@G
z5w$?djZn^oPT3b!?U$cI+i+AwTQMo3y>Ah<DzgOBN8ya5Uh9@YZFWkM#gOcUu>&3s
z0|iH4&y>3<%idi%)4&2*^iYwiE41Bf3}#8?AP#qYZ6D7#`^tenOzvKNocrqgT5S3w
zA!)r7_6J@B!d1WL;{p}ooWLca5tqAr)eVen_`ukT6LAx%+Ld!js3IeF54G+(WjTy_
z(>eLKY{-CJxlYt)KI<06vAeqs|A2nwfCw^fERGWdSd=l)2{V3fxGQh$Jn#J;k#4$O
z1}DWS^`(MvKsH4X5C=T|C@OOX)qNSvyeP9cuR}4)D=QD%h3ack1r@aoj_)i)#m?yy
zd(=`|B~nrbPDE(_7lg)}OW>xGhCtg<8Wk1dfoUtTm7AC$wpk$Tm>;=nk0+l^L<F~F
zG}7o3e}jGcI=Gt_T+8lNV6e$2K^gxwCa6r}3-8`<$akw{u!vh)yX>DkbLMtT-SoKH
zBgOHV2sJs1Q&No3PCF1kwYJczLp-|LnrwhOy9<`Bb`02Po7Ffo^t!5MY}baXO&`F!
zVOtrL3|h2|Bfv%g;(;k2Vxrh?1bnk!YM3mR`-@KxVaxti@R#59El&G`_tzX`=!0$j
z1p7kEQuI4OnvA4Mts)|4*y^@m?Iof*&2?zP%IITP!K>Fdf-YPK9f$zMSzbb;1~5id
z1Sz0&VI=NWqk_Mx=i9^F12R&i+ZSGc)!>k!Y8sk)2Rk+!d<OnjV>u$Dq3T=BXO-+G
zo-SZV9IM_KqQ~XmF8I>>+rM_uXHn2#*Ab)#Pri@DNJIWbQMh2=n;JWFPwm+OAJI&q
zNTv1bV;+`9b=)Bnw+s}8RDJ2E1bx~C8-tlh+YD6k88Lt)BjMl~O`yO_y@R)^E<#4m
z!CqC%JX$Ele2&7UUFFMtnh=rYuVy|F4EbmIzm{$o_MeqQ;p8H1Yf83^+nYIt<w`z}
z?80q)SIi~U9<=gP>t?*o>;w8jI*{=75X?#@cl)}UEBElk`$iJe0OQjDbG{mS7K1Cm
zeJ&bcaYZUv<s1zJ?KE@?<4MWea)*zHpQj+&gZ1{R8qm?_g|6}pq0-zR?9>Vbt*#?3
z0{O~Af){xIlBmpS$2fQ}_vJysV)fQ|d+?h=`5f|Mr~0MW5sWqJ5*&2fBD9}0m@NIJ
z*Tq@vWGrzPQ;FrxuYH9T|F}GKTf#&82cM<n{k$K-_QV<=JLR2$^(2B8;^x*&r28i~
zZZ7kgfy-H~hrHyIS5#Sadiv>>;vfeX;;6kzoemOcy%bHo;#U?tLNz^YLW$vV{}*p>
z85ZT+w~G!ANP~b#hk}3%A>E}?A|Xgg4ALPW-9t$?2uP!Vpa>(~B`MtlLkuvK^bo^H
z{rBzjKJR+h+Uq#>w`YIw&@Wsw19D&I@BE!{B|q)r6^!TWq&v-d;wmUQ<P?LlHyg!3
zggRHO(<VbCc6);zf+LXqJptqtP|A_F0=0;b^xQJ|aXea2y*Y-Mt@4&sq_E%q)L_Rj
zTveavd=19brYqkTwgzMt>>bn355o)=e{6vR+;TmDx~a!?*I9u<4|~)lCs$)~3DfpX
z>Tmf;f^+>1k<BboSo63yrGN5W=bh2v!wU`GGo-oJXvyEuoO{cn80ESiz3w#@W1E=n
zPZI9TcibhpZ+EB#3i^lvGo@$TPzJeNZ?5n_Ctij%elA6BHfRb?2&}@WcMI+uhz0=9
zhY`mq<D#yIt@|~+{W9S&`aT&%tKm>^Jv8R54f68nprB;O$G|(am-C?hn0`poxiIyy
z7M`qUAq9=r^~WnOv306t7V+8nMPe>GreUYWeQ-eu;Hlfce-vPb!G0>*F=xr2*r;o`
z<g7nw$n0C|>UO6zwMtwH2rPJmhSM6EcX>Il@f&H9#@wk}XdF1hPX;E<$vGe|qULX2
z1ZyLJJVZO@<N9fK?qM7$OJBd13FX_B$hI<HXYWJrr<~DZ&;Hvr^`HJbF#OU67iCz7
zSxzc>y(971;rEw{fk<v1vIM?wM(DJkJX>ER`?La0x|Sh`QxvU(gtve1sYkzij(XNd
zi`+X(lI&qhgyBG-QXDoK$6a})h|ve!KbJ$a5Xbt-ZqMP5$l<O~(3VURvu7o^tDr_T
z?rorc@v`GN@`boK;*o3xKAdA!Ymb1Jkkbn8Z?_+}#$t|IW6k@rup^;k)LY6qM5cRJ
zXYYWaap&DPsZ%&7&TlQqyTAW6*=%#&jyiTOjmxsL{ILN(a~KgPu&;U!Oq=*(hpf0k
z7LV@p43JoJ7oUXI#2!+Mh@E+hdy2mDDWI|0L#&NJ`kDnw%25t4=Z%5xpG{(ORe6)!
zBzMbp6j^+~!;!l^(me_^pQl!)D?e2?R41%OJ|f|*No&N!r-dRWTMt(iV#OA;V_@|5
zVgak0JC7Atzu1x`Jt%)f$R0)vWL@(EfxMSSxEv$x*b|4}U(NdM?40j}(Qoj4ZF?%z
z*{3cyCW|*tvL;XATWVH|`??C{FlZxn`)acAyyB6U?|RD1>2P2QzwlxXRb<6r7FM`A
z(AunvxbW`t+1oM$=o8Ck@A$rf*l@sf_9T3v?M<K%klv&H+4oRo=4|0w7{V{VDO1f~
z?7)H)Co2yt1gWiWB|X2Gj0N$)BZD0kXm+{LtZ)l4ncA8vHI9Zzspoue3KJGW<5P#V
zwB-mDA(Sf;F=s=;app{FbxtUO&3*nasvnzwF|lQ-NwF3@!DEgEVHFZkpg`5N!)UsT
zoA977IRO()vOWS6wQtL4z$EkZIv)Mzx%%4S9W|(!Rv`?b210CMM97)XXqccifuqF=
z@14Hebnc}m{F{9tWk$O&Y%HG(F!YgYn~lRg`|CO&k2JCv#S`iyAU}U_gaC)73w9}l
zuyP(H_3$e|5r?8P^;HzVSZWhE)P^G-yBC4Uv;dai?>P48%IG?;ReMd4&GBM3BH)X<
zL)mi|@x_6O*0=tZw{k;w4_`xX-QeNkx=?q%w?17XU6Ooind;~Ex<zeLq;Fzn!|=bR
zyL0%L?SE+uC{tk%VIwz>W@Wu9cA8nRnfw`_d>*b{e>O-MALt!SzMDwQA+?2U77PKp
zSd{>CKQP}hjmWAxlNP@O4z;$^E#-{p1Z}lD0d64B=ML!`eq51R3od6Z+^yWafcs@5
z5M-wNt*SWwTOk<*=vEsk_9=iLxqFF4|9C;ZzO8(_?(in{$xiR%S0}ktX#;1Xc%RVF
z`GVet8@?I1XWTxQ6Qn#dUp(J1;l`U!NvHUoI!k~du<W~{LA0~s>Q78C<V$h3GZmjB
zXONl;S?Zk0LJ36(Dfp*<f4+Dz)VxP8+NkAbuwzny??uF?R0nej8`*?Gz{yHuUH~}b
z?#;qjeZybQwkrAAPHDb%eh@aD$)r(~$*0&gh!^i8c1my;L;^Ws)5o|*i<v!NyZHxX
z+@7(sI-Dwz`-(c~t_9v62wWFLd`Smju@;O#+ZS_3I+3}cc++T`<VIugw*!XlckkQ~
z@uzq3fNmCjv)6{<8m-w}3!Lv-zV^L)lFD;yMsJo+rKH5%STs_|gY>5<+5Td&9dNnX
zoBQUNwzp*;>>9MEuGtB4WX-u)x@+HVkG=jNtNpz_6y5k`+Adp$Nvyenx*=<0(BQ?7
zrTiAd#9SRh5wA;r+`Nm8iPI9b8f~<y!;*%8FXvlH$^C4sU?|{i5fcP*1$vR*ZNbji
z3K-Q}*ifdC4reSA&~i8;YvH*&66of$X6C)KX?!4-@I?KK6-QHJU=2;GKVkw3sQO~-
z#&Cum)EYZ_gGjUNrFp*|_4Pr^+*AoW4xv)QL}_<-ykQ^`AYalX#(;E$+9>=^q6>=#
z*^%0O|705i>x$F=8YntdmpOZ|zu7*W&)4gmp*AarlZnPp+hkp|@Pfw($SPNuYIC0K
zT)a<H>s>-_{Ib+i=azGK+J?{9^j=(C3F2~1UgaB1oc;r<H)5!|AEtMAPmPa4#p-G8
zx;3km6zF2MI?${*_@~mUvz6*eV8K4ARE6xKo!!R>AFFl|Xbv6ggIU}3^)!QNMpet<
zzSZNN+*Ch!5rAA>@9mbZpnCi`X9TR%)M``Lde@!z!5?ibmEK~S@Vd}a=(J=CE$w-q
zex*58uTRzMD><1AY?|qNC{lT{9eL)zt_$*A>4?U5g#JYKBAwdppMcKAWcmn#Hf~k^
zKExrjgV^|(s@@ow8gK@$?6A^w)tw1ssxWahtXWdF^Ztw{;vy7j`UC$2FerNLDza*!
zst*&p?z?yji*#LT0Gpk4=?g)uKh7)o2;Frq&e%3hv%IvJ6LrVBIG9HV%uMrht8mx6
z2x#^l4t#Oc?K{?)-5-T+GBFmkBuTA4An6JxOL|epmf(4+L)Zg78d?S>g@OdYo^cEM
zHx;<X*L5ZREUMGlu8Xd@82oaE`_gKJ@m}I1yy$OaD=C-hzdA4b6Q%xo?tvwV4GO5q
z8XJuQkhyL!rWxBty3%%?*-5U~*}$0BfzAKK)9jE=94iIxTa|l$vrRt7qRnjBh{%Oa
zNLOIvj+}Zjk@~|K4UWluH7f50;j;E~a@F6v<y3q(L*#+Put+n^*433g6n7Uei@CH_
zPObD_Db!r|hoW>tcG5bK!=JQ2R392rZB?Q^kk1|Fdc)%t8oYupNKinz^3%Wms)ujZ
z_JeO2ZUdCYrSAH2?>j!-m3r$tSued^cF)xrb`wwWeg$_6;0|wKo(fmQ!8a06-9Uz;
zw>P(NL8bPFnU6d?sQiG|kGp_0wP>0Qh=w>>xby*_d|?SNVBvS*5K9F$DiuNBuUIAy
z#5BITxNtaFP)iNHRVOLqZU=$6LgWl$pJPzSme<ps+V79s)BVJXGbK_?o*&iE+Rq8u
zjOkEJ<a#l%5EoKf!6k5)M;E(ms-cGJQ<ZD=Z4J#D1A!~vYJ-(IMv1T9{2HZXcLS>O
zTy}sOgu5NUJOO;;881(;ezRia0V`WkV&s+^2XHg4gv~$~i!`;kTrU}Y75`l8^%F%l
zQQ%|xp4`A{*{(<cAfJGkEg%czTP8i?N?H9Qpgf)7R^2+-@>E1ds`Hc}`U+A8&?YC8
zNXpfj^PVI^6Nd}U>3iq(nNK!_<e#V?S0}6CCo)NO{WdQE!gAGt*{BKK@#kW^<A_Te
z_#Q%WbV|TXXVQg~kJ`E}oHOCPZPwzGEG4;bl74c&pH|`c8XD#q5G0OuGO9cz8GYbf
z^0RUSKhZ7UO_5Czk#aKuB@<(dd3`U7HNnT309zm;-S3v13W*@x1SPA8jBmCxYv0-B
z?aDNmrVSl<uzIZU`kEQ%9PIV&O7^JlQ`rmK*5)bx(rjIjz&>>bH?zBXA!RDZb!MQu
zjehFk(^s6g-?-iHJaF4Mfm-?MsWB2poCYyJw`B~<6%W&G@%dv<=PhH$(CcV8Qc$1Y
z@$I+2b|jzZ?w@^ULF>kV_Yw73FJK14|1}%XgaM_*LEPpNmui8K>Gp@ek&x|7@A2i|
zZ~=~{hq}sXwlE6{qT_xScc(QY;K61x#yswbeDB`ZH8H!<rAwj2jF>2>-HJU!;S8bV
z3|^1coDBw+mOrW`jiy_Mgn6UE!cv{+d_ZLAjzv|Y5?R+CHnvPyzW+h7`)*I);}_Us
zl<;lrrHpbFqj{2*MWwZg5l4s=Js<6pEw%`lor!OzNnKOy8p>$eDxaI0qoO<PJMfwp
zhzW%{XslStQ9}qvoNi-`;9kPgc=X<h4e>qt{r>ZX#9UzJsep8MIIz9<@9**R3>dk!
zGb$7J&^Mncb%4PnuPa10Oeo%G7s@z~D=@2WiQGgOb|m_msx!q!{V`$=pKXu2hRbSv
ze@Yj=&woZ5?R>ybK3z`p7R5a%!E9Bf@FE+uRSVFSPJQmniVX;)=)LbN8k`<7)s=qt
zy~%RD2jI16vy^XVNFnIQi{RM0ed!C*Avwz%nP_Hprayb{x|82&QH~?V3+e$mTb#3N
zcXm)^k%z1GEY!jNlY3~1fT3Hz`R&_4M$A{>gOtIC^F(^v*(c8+0lO@g7c%8O)GC{{
z7C;v+FwD@4T{!>8^X~rU>9LYtKE0xuY3IO2Y0Y}Fr~v&}b@x)c{mtkcTbiq@%b7O@
z*b}?6PYTFkL$F5;h&}F^vjl1xqA$h|#j7<qKkDW3);dbDQ)n}c$vrBl&!{OYO%iqH
z7v$2EmlJa_z;`a+tgS8=>eqY!a_oiM^j1$RL#1Q3VVr2(l7dcC*bd|aa`!%(8=p-_
zi9NgvKU|T`=<hb`EGfF3yLz^kMO8cORo%Usr~aeiB*TM)`X*A#O869EjnLb$VOE~0
zR{fIYUB>p<dT|7?Y=KRxF{g~y3dnxiU6DB-Z%hud3q*Tw1t+0y)xp_t?JPcu)1KRL
zCf{j*^K!heCre(|<mPVC_XlEr`GT>w81W>u9Mh}WI%j%gqa{_(pOfm=DY1_&3PpjB
zc*_I&dJ@oAKv`~XIl*KwHBL8d)2G|-kBU8>`kRU01>Gv4ttL7H(;wzbp`zGoRAWRH
znC+X}o5~&7EdOrDAa<OoYWqyL^}AIHZz=v7d%mvVlxO!cbFH<W+I$QPk+E+dBMRgT
zm=?K1=<c823Jfo+H>A6OF$iP$v%)4wRwd~9GSXD){#vTOsxBq|9qp4H0uU??a2nv9
zr~<6r{G}PDG^z6HX`!RP<HH@*EoJ}F2o@?Urso{icM4AuJpdvM5O7jb`8xIc8#KTj
zk=pZo59#pumepBRq0Jq_JmBf|(>|$zi~Ly7bI1w`8D~80A}``(P_sNbJNtyb0!<l_
z@PcO@B&R`xn^BnWcs>30M#vW?6Yy(N<}2f}&9YjmaB({O83X8AQLYSpxC2?yMrypC
zK4=%hg*`f~WP})M_|o{1GQYti0i9dIil>3y2t`fVLZI~o|EIP*i`J7jjq&XU&!}W9
z!fz(OUh#krU;l4?LGmwfp_d}QImVdEns3|4`uMkp2gNNfeK@NIzNqlMib#mW;{<2#
z+<>l#XK&bWea=2Tj2SrsL*9|}-?LC|E+uiF(I7G_sj;^nr~eA~&S9ewYtz=ng|0Ci
z!VkDI_oq~wa*w`kyuup?0d_aT?>BwfeTDKIsNWq<ALkYw4*8O6Y&41r4YnG|x-o^6
zHGkum|N47z2>t61t*@FBIw!sIs7*jyzXGxzChE2|@b-1GeG=P20V|*cz18}nZ;t*4
zML-yX7g2p8M0hwhi_EBoH+S8N{z}XD9GG0nT-oUiJQk{WtNSKWr9)JB9eKcL$l<(p
z?~tGH+%6+o#;36lCmO;s0=|5oAK>*{X!@f%JHpOyz!R<N6fIw0k9d85!e3?Alq&i2
z$5Mv=!*-88p-#Aa5g-grK<#}+Q`QdhizVRgcc;*3@D1!`tyD)x6S63Bvwmp{SSjdo
zLgC}|di}jyzM5!w!GH4S&L3DuK2GIful~w^IlG@o5D>`&v>}+eUGn<=vio2HDA7~z
z^l-@5a0p{LiJ}P?xObcoHxz+1%}2pufhNVf*&l?>aIuTX=d%{I?^LUw5oi$bq1r6n
zK`Vw$z#vJd2^=)8$sT2Rw5!~>XP0fmX&0vY;ayNQtZpxu=!Vn`r?I*f>&fr*DokVS
z&J*l$+vki#0-@$xC1Dwo?plhc=kiOyjsN-|P<259_(u|O4wu&C)C*!(^q?dh1{5MO
zm|uw=EuB@C2e^6#1xRKnIXCvxgGul#h|y`Cz?XmhKg9vGgMs=-Ilkyl#{MX)rhC=q
z5B%J#*CEAm57zyuL)fg$%z$qt^h&t&uL}dR+Hplx@DC^+QDYK#uQ_go`Gz1!m*(pW
zl!H@9>7a2#6jE|@xQwYPAGdG56ni{|bkb{^gpoUW{B967d9lgEfR=OyNDVUV04Gco
z@8tYS82(@9!vB3PpxPn4V&V>IB`DoCq*M2Dqu9!LhA(h#*(2cB$qmZ28#mliMILl)
zZMt6WA*LkExn`60ya>jyd1FX(ARD>_b@EYm7Ah#`R!eZ<HW?ycT}g-z4%x)Tt*F<U
z$TaQ?-m_^6-M}xB4!cz^1v-!oaoYy0)0qnb9b0SvIugc#(P|8~TDrZr<D=l%(j+cH
zL)iTblxjm{QYpk3zJo`|4qTlItLd$7fY?Wy9Duqn0C;TM{uO7uDzf4V`pwGXR8S|E
zE&nHMMWm&sp0H3r5qqmc0*3JT2c$p~jI!-e>aa62xO|NhS#8-j*wiSmPqarGQy8an
zr_gaXk$e-GM(5juC$|6n+Xfj5wik@d?4Z69vU4agUqGUa$F@`ie25mQYQqMJncXGj
z^UrqlPZ0Zg<xnGE4wbr^<_xBQp~(hfCg+nUe}UaQO35?uWQgo9pb&@gQ0Jl{>ef~&
zcW$zLzJ5@^svJw8+-eYqihBHRDN@<~AtY?|j)2>OBz(A}ttC+hvp8B=$~$x)pK6iu
z0*Z~qh}S4|u|;uds;saW%t{@`ich`UC3{PnGVNkH_d6Fz0t5xwnJZsD3w+du%PaY<
zFi(e_ZMg9GRprt^RBSBnGQ`74^}{NBxt%6TrzP@L22;z&4HdG))Z|x*N`zJL;$R2p
zT_?ROZk#2X&PQ^~PBYZrPru`63kwr^w6IYBi<Evai2evR2orIN?(M5u-CpGbS8Iy=
zg)T1?X%gqzUYd7dN>?fl6J$A7<wTxL@2@o<U2)BmZ*Hh37dxu+(nns8%Mq<3dha3g
zu4^&c&H)~GCm#E*t3R)PEQL=sz<v&&?6?Irb%YTfon4>iJ1@=Vxyp`(NXR_<-VY8<
z)dtKL0w}B=Ce8GjfL9WwD&9Lz-;Jz!L)qjn?8N<V<1b>*4ike3;8z}>d~Fw$rYqhw
zO}r5=G<SKGA^c6kMW;t%K(bFt?BJCSVV^j(x-@QN@xi&_I7hvJEF#u8Z${*oz7R-D
zhFbn}n-hAgD@dasiv({=C*(@JwpN)>J$Ena4?Vf9Z`HyJ9JAXm{~P)UwjkYmC2_+v
zb-y`7=3cBuWCRmlF+Pvgj^Y4o77!US2DI3g#xY(jI!gCNv1GYz$x^>K#vhY;)q^K_
z_MpuQNHh}s7O31ZDR|1FGDw;u2^+?kXS}lhWzm&xw8c%|C)f8FCOqZUrut%<|M0tD
zrA13aicPF?@&<r5p`}OE<0`}My-M6D;Ca!PdHD!eQOD#46h1N=JnYY%og6$SIIMdu
zhJURFul_sR{BQT%f9l_|Y}iN~_tz^`gy+7U()8PvmasYFA@Wt5C!tH-5Z<a$(=eib
zt)Rf&pQ!4s80S{i(y@wNX<>));p!L8oEhqB=8F1%!4(5{{1zd-n1Zvj3Z^xn$RWYw
z{uFwOdc*nTnE18vt9H3}u?i%B^6$2zM7H){tS7!{V$xWtXI&*e%OJ_It;C+VlA_2~
zde^WQRiALZaHTTkdES^j*Imjv(Or;kmajOHq)V&1YOZjE)@hl%3Me%Dji5GJ<llSF
z8Nagn5kmP1KTM7s1p7DvrH0KlC#(Z0H215r*HmljsuNb<7o>^@9<cl8k`=u9ZSHy$
z_p)UQ2rjU6B8_2BKV$M`NJjG<^HvvF0yOd|qQ4*{yCX@~64-rE?qpDjj(_*LgQ8O`
z0g>o@u2*6hV7Fp9>6bV{00On!5$FUqL&TXg&%DDqu}}4{bw3ug#UOy3K*_=$Qzy(4
z`#xXI7Xyv6`_^`h`S$rgAK&LcRx!h~B>XC;ZBAHN$+S}c$*(lf;HYTMbjW4O(R+6C
z!&}!xb$v6+GT<qC&GxS}igxnraxaacvG|MLlHTEJ6Qg3Pg`WP6nAb#0VM*SWnr~oZ
z*$JC@nBZ=xWm5)a?Gk@$7t<}Zx#RCLiBW`&gk^W}4i%Ub>_%MsU~+93iqLNk6aD^W
zwy|e<B+{Qthw7NvBM5UncJ*PDH=`Hl5wbz0PVMY7qJ8y0xQx)%K0c{0EWk~73A(st
z^Mw-KWec#y)0Kw+67)t$P!AClq|-Kw*jPTL79x=e9M(p^>abl>Ip{{SMNLkwsp;<M
z;qOG<QX5RwZ&OiJ2q~nr$Ms&yI4$!7s~xyKV`WDD`K5KL3?hY90_*^Gd<+Sp(4yqD
ziqC?P-i*E83U3XvStAdAHDw*yBcn{4CR5Y$=!N*YDVdP0^%m`EFSkGFHRNOn8!;^?
zMj6k_dFeb{6X%rC@LtARmGM-WB~o3I<(J&96ZYjaoL^pa8=HZkR(1AI>-S(R$j~?9
z^uiAi4#(*bE_?n}1R8E}_f<SUe}#juHHndCxiFGHUqe3Fl6jD5@JeYLM7sbRK&J1*
zdxrub0bi|ZS{pu?KB+j?sABl=<Bagrj&KklXBYwkeF%TCN?v~qRR8MRpx-eemca5|
z;DB{ap<eyL^On&yBBQx?>Y%NZ#5_nNfJg`(Q9#k0U&Zc4Li~wXQ+*hO%g0oRAS;eH
z8TK+nPBY*du%tt?<fW$%y<F*?B$^Y38zgj>)SK0rI71c`FF+t5#M-d_rspC8iMxC~
z710ycrzup)NOl#eobix6$y9{0FGT*==h&^Q;I$km0ZvsH0Z$u_oqRBK$N7t)EH8M0
zGc=bGBx{G}2X|%rFZ-tHMh;30Bp7XMaFi)CH@uvhHcVZ*nQp%w22^bKn>Q*;Qun*k
z8=Ny06Q0~^e}c)36@$_{olCbjxlb((FL&@c<x{EJu%<^SvN;BK>*Kg@Ld`jVrldQ;
zm>q@HZ&x&cA{|ck4+s^lfIAB`1bkgzfdCN0m4XWb-=v!>fSc(DwE2<cN7~o7T2=o6
z!PtSq+ZhZ1wXmgEV$iVQZQ3=3OAO&ZAl9NAMtGOPDG7${`fti?l3}l>4k5DOXa+_M
zuwaJKK&Mj=+m(M3&and@kUc|sg3s3|Z{(n~6aV@5_O(L65en9Rmoj7nU*KS7uxW_?
zc9~aA^9Ky2244P4J{cw>6s<wndbC_ZzNa4g;M0Bds9TdZB$u_Q>a;SfL+FZOEzhtE
zdEH+1$+VX4UA~IeJYDnSgVM<kWniNDhUm+unyTBSJzi@`@=$H@$v{Uwc_$J^G)e&S
z6n$y#da2^*)2z^yL~ZWiBaB@7_LOmms++eJQ)Gc1tRXww43T>fd-Z+n&q$EeDwJv$
zjFral_hZu$1$N%`2FHJ*JariOr~4>%#9Q#TDQ-XJ0q^>wflZ1e?xa(?0AOwJ)tD0a
zmb^))h-Y{jKn;@I%D!6U=H^a=`oRXJBbJMf{0C5`>QW7j8Osm09&(rxDD6p8D)oYV
zUH=!b?YIO$A3zhr9ST-z$Bt-aiJCtL+uy&`Y2*zol3%;@x3kq+nj4kscf<nI``Q@J
zMw=;k+@!nU{3iHD%p6gEbuXYJ1S<~?*MAvo37euFDm_?sAKftWB6<-oGEK(JY;maq
zXwQI-#dKL$0Fgo|f>)mX3y_xMtX4&jBtk_CC`yLRf?}9Fd%X{RyuCa3B!v3wA!WUP
zlOTMd`$=WXMu9x_8jvgtjW#Wa-%oMbfnnV$z5Ne*z=NatB{V(#Lh4BMW$R2k4bJvl
zw<%-i*hJAf+hB22iG%Z85}Xa%-yQk~WKS8H38lUgba41$-tAOoti9EZjgKxKKz2pr
zCraUi2l@k(bl`a%&Sp;z7CMn)C2c(^CT7c2Zj9pipA6k2&p;<hI+Tu}6Ttpxqs@?I
zI6QMDU6m?9n{BCh7`j1rW7(FohUhYGH9Xqt_P-tXO_lZ_l=`U)!tt3$nX=0AE9z1b
zB6e0^KtoB^&!Mt|Nv$lVwc)&@>IT(>O_<-Gse}<-Nu0nBm@aiErQnfQK_#<1AL-8i
z#JE;c>hOT?QO+CoaoP7=j76nB=#=4{65$!+1k?Y*4qU`ywc{1=H!w+|t^7sX{@G9c
zR_E@`F2RL4^5s&BTzQKldmNQOiABuofl_4EU7{XJ9Q}N6Sy<@*wZD~SycWcH1h{<-
zlt1b{+p7>;8qIW)ooq1lj?oZ0=S%|@jIVD7VUL|YX6zb|b2JFxCo!kV@NT8v?ZcDY
zPbqi_5W2xH;+~CrF7YlctG#vpT&Ei7Vxvq#+#g8N@pqns^1&7^A{I4Po#ka>?hU??
zoqR$;yY!rZ)2Gw+Y8$-ckI_a#R<K1F{?nWhN}byJwrX<HeqUoN`-MCJ-W}lM#ao0=
zTNOkB^>uuL&WC(oQc27G?irQDucl0|KbDJr(4KN#CFb<I@W?rBanPBoMnYFP7c$hj
ztm+Zw_B#!bQF@~0HF`D6OcJz>$I(>+vYQPVhRIxhNXEy+7b#HdLS$@vIPD+TFDTRB
z{t<210@-?m_W@_D{Q_A$B|hVY2zGKt5Ndq1oP92wLu2g{kM|?eY`A6Mz^(K2SHGSj
zoB0GZ<3g*%uV}{`8JXqruwuU^=0UbdqzcRK*@OQ$o}~K7GjxZr4R89?=4uxWUx4bg
z$V*RnqCEuX^fXN(!hHrf9c*-QXFtov>U{5omn}KG9?8Szx$)PnQ<wLPzR>@$RaHn?
zt6<L-mI<Xg8<8=dtTS<vDdTVd#ph!@H^7khJRSdFsNMsS_)+SlgA6v*T-Olo`RB~_
z&hG+?NSll|PHiPNdbt7{AewmV)6sWs;aQiCC^Ll~e#WlS<ayPS;$4cvWVOk`?{jDO
zU)dpB6@JnoUc1J6=CqK;{OwhYFRl_YV6dR%n5LG0KTbFYWosvd2C86aGwO<V>Sr6M
z7UCQ}H<Q<XtKuDeMN{x`oy~Feise#ZSK235bZFnoY-%K-;-~flWR$S@R3l!KA_F_w
zg>Czf&mpp4^jK#(iVasu?x8xf&u`=8hL16e1aDlHkc1<+*)IbAML62c)CSZ4X3Vdu
z{5|Gm<({b`qq(VSEt3wj0g;8K{mw@J1A42m`?dX1uN{96+=(Dw^R9CHEB+QY@bn7o
zNh@Mid4D9TXN+{nMmG65GQCx~=q+`{gL1*~Fr|xe!f6GTb}-cUq>lqSQ5Pp;%pGBs
zPukgeD+Dy~ij`=GXeF^!lKH$lc)s-2_?2_9<Hcx8>5x;YX~|sTn46F9oG@jUB55Pp
zd7t8L$`-0_>3XTZcQ;4sz$U=N#6B?ExZUWSoj-JeyW$<{Hr(aAJj7Q^0}^B5C#GS~
z+&8zxH27^WNC`iT2Yed%1OJvH_ZNFIV)i@Qg<?m+`zB>(OMx~s&MpEEQ<0;R)XU<Q
z2JGWS^%DzxJ=q&D5=T0JpWa}SwuIsOeeoiK)=2R_+_JL4(;Xr?9`H6Qp`hmA6WW!<
z%r{PXX5aZ{o;lb)c28n(J2vE#jxTR{*I~Y2=cgk$)@4Rk1f3^hI(ydgh`;ua{CLW1
zNM@?ixe_0UG+%uklPHpMx&fF0mN1!E-{~s{bj#SC?dINGQUEMR?mgz&`6RDfxPGSy
z9yrZL1T16a^nSu*rq1?VjjJ<AZ`mY56npbuSllvJp-T<&?QaVb`Y$Reuoq*3b)Rn4
zL|15=PMj?eIUl0N?BQU-gt28%8GEH4Vfd(2pH@_t;5@Ly?~(`VlJ8Z0pEnDNmBPO>
ze;|`Lk(|j#h-ci$f*GS%kcW-aodX*_df6AQ0-QMmE%pzV=a12PKa&)t(2KLEz0v?L
zXH3nrhK9yeNrA~VDYU`Q!Oz6>w^@z`DW?IT8{g%hEdM2>^OXIX9Oe<|HZ5>Q8E4Mx
zv>W5X$f6S98y~MVFai9h2%(snx*3!>&p<JX#+f+9ST(dJdNw8)MWh?hd6CJiegNXm
z03|l6Mx`9?yI?dcs_ED#qp7lbFzoZ@yl9{|*TxXvg`0Jb8O>dYV>zM#_JJ_r8x&xD
zHHiwlm$=NN{8mJ=OTdT%Gzwx)%Z<0VCjKAlEii2jKJQ@&<USX<75j#3-V=E-*&}fN
z{Tuxjd7auys<1R#lA-eysCQT?L3Jni=+26n86b~$0$i(=S-yDl-ZQZ54UrlY5IRK2
z26~}Qb}LkSk-an5Jq;4VW&m<Fs$muar)`7Sl*|_{@1o2Ed*m#8?UIe<Gfryp+Mhlc
zp^oe?BkPtw#lzYFnIU;@d~76dbsv!wU(~)FC@K<6hgjV3mL9?V;h^if;Png#D3&4a
zAJ8*xMK)T%cK`pGKyJ7It%~TW0oWgPYg;ujW2*%1XHWsI&MV_2G?W^F?5KiXrS9w9
zbPEHN;zu4ZYBvigrjcxQ1Cpk?YZm3Im-{~1DiW)z8a`3lk>Fi-OGzf$H4_7(Z0?tt
z{VpH~)bJ`b9ddT3Ch7qmX(T6o@p!D?dj6YFj8WLO9RCL-1(aA}ohetUjRQ%i1BnH`
zoA2tl)(&~c9P=%teBeEsldUC`E%=B@Cj7~ns?N=cSnQ~#JzF~b`B+fECf5X|fiG|6
za*QT<&yK~<rSu0Ki(K;HO6yxb!`y-XLE6*Ot}!r>&W2wzzTR?a_OnJ?_S@rptK%o<
zNJ$3S{8l?~{a=y;&U#&bSos}Bj>?f*e?Q`Om-@q7mxLIy(z)}yc)QDv!F~>F9T4IX
zSQu=CE>i+X`IeQE+dI}4pD#jjxX9@G(W(U93Q-^donHJUOium*C9+YPtZ9cuCXuhy
zG>2^ntfchmG?xpoF|U+2XyJCpF9-nsrQu3ADK>ap?lT})fZjZ4|9-?B#c`Arewufi
zG^$C12;bbJOn2+6B#RVxqT7!D>^_j*^zkYX=fK_5L5iG5Li=q3>=g27UWm}#694hm
zU2>gj5m4)v8jcp!*PYFDgn7!?bEnD#<jm5jRD`nYFMeuMl)Y9gq#ywJMJhV88;%Kp
zUzGi+c;O6xgDXEW0*bp1BTvaW^e*N4!Wo}B^G|MYDaIyr*bhkRC^2lSNDbhO8B%6a
zVykn*D2T)^^p4sO)-G2&B3uqO{3-=RTO4dsMvWJHic^Es0d0b19FW8&l!D)UlUTLx
zVnba8ZUWl_I1OGQD^v3HlW*{gN&b9wcA&OpQX)BQxT;V`3k3W&Rj}Fv>MO{`;X>ra
z4up+pURgHA;oZ5e26a8{ylh@^B5rvOxJsrj5t55drz3up`@~17=rynQ6t0}Ir*=W5
zHDxqmR-S=&g`9U@Sv7FAQ`P62I7OG%r>y03Q8Gs_7}Eo?!EGh@R1%4)hS#by<J#(Q
z+K3HMJn3c2L(C1VP(FO`uwe;YRuuHQ=6Ci<68%|CN1X2+RtWS40&5eN8?q7fWe|xb
z_p8X!trzhNmWMK|!g!;v?B20^cL95x{kmi5eoK&NGd%J;{sAA7*!J<(h5$Wi?yQSL
zelvu}0aaaDi;QTT4pc576=6R~Xg}c^KB=VzOa9J2g93r-2ko4uza=hIO36!>a#iNh
z&AL{;CykGJh3T0wVMKCOV?72EhWF@_m*%w}whargN?AF7kC(oq`V?9Qv=!_;_y^<}
z%nSNMqkjFttLUpqu+#3vyQ!;AyGtIAM_Q{s0>b{*r@P^k?H&K-uK#X}{!@Q2&?G26
zYRw#8x`Tv$&yc7;3d(qvC!sPCX{#$V@V3RSjDtmJEN8o;I^xv7?t6Aa41ac)HM#J@
zlb^ap)fp>n?i6yA&o2gJK^Va#1$-M^dd2UH_JZ<<3G4TMPCm;cRT*H&QtQ88y`SG>
z^+jq$R^k$ZdhIkzCk>C@&mUySP&Yu{qkgQH`^e?C8kk%1RP4;=E+|^O4BT+H7MLe~
z%1Z)U4WLYm9U;183whKUia%khVyuN~ew7%IyV(Q<-waDM3&z#v8R}ukYiDOAh~2Fk
zBEnnllNIrIGJxgh;bnl5@fL33=yot0uH!OTWA<}Rwta@wYGZu$28iSpyljUuxROu-
zARn=|E#tCQ#ZHV3(fpg?+Kn1jw6)5{OZ39=4fF#HKv~1<3va+MKUTmHY&YNPg_df4
zTH@q}RQHT-xU<-2bX%E-gep<IY(Q#D_9Qgz=n!k$baWnH+h%8o!58T&9nDmy{jO}h
zu0j@)F+iFuf?{^dP5NSANP>{4umApeXOl?79Y2*|Y)iTBx@77@eRG%{_@Py@5Ev1Q
zQ^kI5Yiz<HDxZoTtJ7ASaPkbRC1|$3eUjMiA#{;u1O**b^1}puO_5`qTqe`8n(->9
zR*`x;lqDco<f8fOg~uj2AKzFC1Lp%<qYT7srj{;phd7rkJyEfT(j2nVfebrdL!-5$
zIkFvC*dG@^2hI=tAo+66I_-?B>}E96=UxYctSEnJ>V45osarx!<K;HmWh-uqY=u5_
z_4T9FYnw#5@ezKva{W(~+cqkOXd9X$4VaTtZ*v^OLlhiHzaGFuaLJ~coSzvoc)F`(
z?mS2?8n@1u+v(ry_mE{{ypm(4QD@YeQsphb!wG#jO7Ft}LQj)`u!6C`%6A-b<yN*U
zF3(d{rK!|Deq^4MWt!H4PnC{N8p^e;l_w4)WTnK6bVR#csoRyc#VVwcf4t#Vd19K2
zsmm!1^ys4;Q($uy5NLnwTePn?BovGJoHXpdK;=s!GEVt?;X1wY;684tu$ydBmHg)-
z=Kx@F`n;@y7xj)o=>kyuc3{}Bc1;+z1Tjq~H4MJ^kt2=RP_GxUtvYjRCv7Jrtsl0C
z<~CXB-#eEGm31PFuG4gp`NdDw+bTYBmmu#s^+rvnfh3q2o&OjX_BjMH7UbZ{_3NN8
z29pP8*Zafh1q3aP9@c2zU2p@kXXh@~IwWs|%}vb%^)VhxBb<k<<z6JB@kZY`zke$N
z_RQcPw2QC-gwLHo%bJErKaQ|9<Qq}P9=?$K9PbW+%(|d%x$C@^y~dV?n14Xc#}Nt@
z(zD0NL!Wt<`#CNWK#v8(SugBepm9rMGcIqfp*dSU&r9QDmp6Zev9Z2N0Pq6ihDEUc
zhD@qJBXP`&-}tk-Syl%3p)NuxC<Y+%uWqxWChRHmxi@A&x_khSEmZU}pgbVC59#}b
z6al>dtMU_*HM!|?Aol()!b3xLOD9ibpf?f!Jstq3Xm{DF(wBY)>a^7*I^08=Nvo!^
z_dC&gc5yT@B?3#986Ydxu%jJAa6Oy8CiAK~-!?<EDIz|WDV5lqf><{8upJzDCv){n
zQ}B03#%8G5*js0gM54h5pKNToWz#;Gdb0R$CAk1M{SoDY;f=a&`ahsIA3IK;ogJR7
zu+_t_Qb!;+-9o|JG27s^Z~t1Mz@Osejv5Sc>E*v@(?~UiQ8)6g#k%^GHZQNVf<!vk
z#|gpC_z1mX@d<U&(RWpSC0R*7Se&L?AxGf;8}<g1@RBc?F5SGy_<TJH$^OitaMY#_
zWUn9!N-it1LW&>t^v6@5q-E3#@Yth1V*a@RGRontnnKE?mFu*#&6wA&RKH{H?|i1S
zP+;Y)ECrU^pNYDr&MLTmU&cuE5Te7OHUJTrD=&gGMk^5B3-Wl<tnsqov<#F@EfSyk
zB4Vt_;?j@?`c3!2Kzu5XwZ+p%4EX+xM;-ABHKPe@y*au^)!jwAc6@_{aRGotBB?nT
z`4W{3WWO@<V(Utv_Wl*4IprqmPC;)srQRf4d}SNkkrY$;A2#695DWPSlxp@Zc+M7j
zO+yf<4|Tlh4L;VW)4n7FD67vFg}?*U6^OQ2!SsiDX;UnjYB;m(@X)B>@eJ(jyz@^O
z6qW%ZL=o!1(mf>DdjBokABNh!U<4ujQ#*w(i9(b-Y1A{5{<f?Vr<e^$-HUk8eN`Jg
zRiOMRe#J)=NJ#(;7p8W-D15;lHt@1N*M7yrlnf)Ag+@ZXWO^Sx6f?pOI2j;z$#N<H
z`gRD&o<K#xOJGdCx|DdnG@9eo`;{D3s)Di~O>0DC*xqDFVh_zUeKYri{M;OlVj^mf
ze{#3SGCXrJ^rv-yg)SRl4$vX|0i-_v{s#o0MF*ggEc427g*`-RK#mALIGlZb`#Kvg
z;Ce&sYrBT^bK!a<?V3I$KGT@;${I^jNNK?Xp0Wq->_^o@%tEzk(*2Cb#)+UShZ#UN
zSmNd8dW6=9WhhR3boNmFrBUdE=%o<Xm&uf?imY4LaNxa<KN$l?JU9+Ph(WBT;Br=n
zG<XDi6E<>hJOy@}#Rs}?gY)x*mOJEXpv7#rT}s<<daEWmO~v_Zn(_~h)q1lMrBnUK
z2}jE>xt&ZR4{02_JbvAgk_4KJJIEdGY_qaVo(4Fus)=42?Y~^_+kB-{nCq4gc@Dha
zv=Q1?74>zsA{OI09=4cL1n|>sAUGQqm#3;pTzgg981VIBglR-d+ApM$4@ZF26jzA+
z2Ten5pH~`}Dz=iK7sY9R5Fc;G3rLG_DJ8$PnrmO%%w`|I_3J;#vR-JE2)v3*4D#Jd
z3bRT<r(4Hlaq2P{YY}mLy4{|5UFIe8`-bU?_vBE3lP){J)HE`=Kw62miDJ3pq=Eag
ze>Q}p`ugLS)MNadzHAYdN!}k3>h8o__G8?W?X6RMMN^(bGV}~cCL>*Z0VPX7v5)Sc
z{o`?W@+q$dSrT(}SNcpeuj&&@D`+W<I$xuJWP56OpR;=Y$-#lU(Fgogz~VXsXSn_Z
zqb}{y-Yeg(D-&%gmmA;QhW(ysx`-UOsxcpda^5U7_;45>jTGi9IXZ8A=xy8+Y$ONP
z^+3?(*Am#Uwm`|-%zWX=;(7<dj~U&<)#s#X@d$*TCP6Nb3s)k2-(AtfyMIyYA2&M>
zx%6p6aSwHZ9zFmdTK*d#s=$zjnfz-C<gaCq>l)_B`uMkOkJs<@zgVHL2k-iNK^TgS
zR@>@QRaSThiwnBNTsrvqTm?ig=Un#eGM0A;Deq258LF@laU?sP3H;Bm@PBww%!}GF
zkajO}q^qC(bW%IU`(@9oPg{&2zpWUm^~2b)o{mOhDG~DCTMh@>Y?r>SwaaSfaX1pK
z7Nz>_)~sTbZs5R)A39D`Yqt&QzanrGgC~j-SCphaGPog=eXAd73Le-3U`)>6s5(8J
z={?bvDB$j>rOe9@ENA`uQj!B#I}jER^=a<ow_+MhU#s$x7dh(X-jN4tUYahB2=*}z
z6lbf(G8x?ywdG{zAqIK>;>0W>2#;WX`!~EerTs6dbk475g(s?X^-J_^&{g<hGGWRf
zSUhy68J1Ub`Q&020ek-Egx96t)cRyv@2yw`<mJcS(j9L@-O$Ejnj+w}D4*l(u;tB-
zIcW15vXdcB1I$7(&w&~se(T}clMrF1SkFHj3}k3-ti@HW($!@fnVo6)W)-=Yq!jDh
zNbbE#R(+>dNLkh{cr(V)4-;<0kVm#*i7QhGdUU%s!v<nQDv3IgMdA65QQg|}-*I|<
z{@zVxxM=f|xR|0yQdhp(@PqXX*#IjHya>R%1u*`bQPfec@9b{RJ;tLSPIcPyk1B>c
zgv(67;#)Ys?5|Kh>c+kU^vyDT-i&AG(aJ~F57#0gP>5jsEp@9tV)XBkAWGn=I)jN@
zG00nEV_x31>$~^*S!S0_>N4h4UZ?Y5-tO^Ld!EOL1`vG+%UUg9M()O~UGm#H-SvrG
z9_Hz?dDPVIjV23{>IKLV1~<`vKnz#fTSJ`rrn`);mfmH`KjTkXXPmHBjoEAo?aW3t
zsbcA0)NcvlA%1y;?+D}d1LdclxU5UDq<g+oFA<)oJ8_)t!O}*y^(W5<VAx6ez&6Q?
znM2^YUo&vTmf+l~90Vu*<4n~<rR=&vJ{m!6X|(Qa)!8q0IQ_3Xgm9>Uro(OtqKB#c
z`b`yW!XkM2#I7~Kq}Ra9kG{s@5Sqz}P8w~d3;`Gf&qE^LYE3^`J?Dx=lV?_mGGh8h
zLt(_hSuNniyxmR8ctgFVKm$y`BG8Y)RpCWl8*lVp{ol|D*s}uOX45gcJ`?)S8ER|U
zNUwCNXlgtL_QwZ!pp!)s(3t`YY?vR8>(HM$J*2AmPIpe}IX>dFDJ)ZE%RnY9tR-2Q
zxdQKed34xw;^gCx!oI&zioBah`RJ9zGJl5Dw$M~tKptRKO-cj^#yqaA-sX~NnCw*T
zs+Wu@lS!{0r_GHe@p7wmuG<u_jUEH!V6}(%PUpyJR_;2402!O@OZ&T=jy*IjWQ%!4
zEHa3J4k`hzCd^@UNmGWpjr9AK)_9{_HO2Ky5M;K)3KEmHmas=E8i${4U5Sa5F^WpU
zPh7XY@x;T59kdkBYmJ*bW<1vVxWwE*^j%n3n0V}|-fv`jMsear&{Esv{ccg=@{k}3
za+ocy+0Ug_zkI}lFYU#-dq%1X->^@<x}4KOuw{A7)S4FjZH*c~O5Kl<z2E9__8t1f
z_aVc0=Kv(cM)(r$m;VFetN~owKZc<f5B$$B{$KF!6c3{h$dj31q(>WDae;bYw*4J9
z*RXHsT?X2W$EwRXFG;Rhj_e};{%9ThpG`g>c5=*7>LlC0(poOgSBGaPSSC;jv>I0K
zQiQg5zfI<_OLs8JFL)+~WdRvX3-X`tlIAST8$e8ak;c|8vEPKX*C9QvOqXi70Dw1(
zdy@8|wJK~=kGx_pkzU4bW4X{qBFF?k9+N6pWCldpRNTANXw#i-NS13YWTyUJ%(exU
z_l01sqMl$%D$C1uKc8z?f7B*#Uf6S>cmFhs4xy0t0=npnog{jYO9fG~3IN)Dv1?~T
zs>#)r!R`&7K4p772`EmtA#8op1P!sRf?w@@#yGtZ;~&k?3c_2T7|we^es<k#NJ?eQ
zSRCCGX|vfJ*~|7{oYm#uivbX3<z(T>z~Xm?ZT@q<s=c?}PKqh1?<m{t*14pzt&fCQ
z-7~6`5aRQ<K9XzNH;1pmquZ&EK6xU%79bAE)Z<Yo_C8^80WCjf%MjUeYz@xsUdO5(
z#yp4cO0%}?#O;TNSflLLlCX(<PFwj}^X1?Py8k_VC7VT11H;wg^x&=^pJ=^j7d`tS
z<f?p4@Xl{TM%z!1Q-9u!&KJAhcV)Ln*=j9rwex^^|2s~&edZdN1lKbi3$P5C0FBxO
zr8WG&dC>nB{5`53NKrX{zLQeOH{6e>xT9Wbn0hlqGeX#<Y2U?{a?h74R>*A7M&}i}
zibWaZ&vwO$q`ba;I(9`0MBu3{cJgcf$`-7_KapKR+b-I9`&%(@0+NN_q!mxjZ|Dc7
zu6aG0uR2k!vVB}{xuGqc=u*HSXLt3Gk?>!KkN}6|<2qK#1*6Rg3N)Ih<+~YEd}!yo
zp4kwR&F?gF%O;s3wzyf-Pcvi0Ac*VuD{=FKl`Pkh*7pPRzOxS4Ph06shTuY^90AAx
zP&vGt>ew#cE;tAJkoJSLk9L2~wv!wgWUgKkuQxW-Bvgr@1%`X)8~2P83BN#2q&t*?
zZJo;F#zJrCGY%Z=K5$~9QifQ{-5z{f$CCV2g6I&I0wQpbAs(h^br{$;Ls&nYA6m)M
zRP>18FMkT*eJY@$LVwPT2nMbIAhUoh-46&B^y;_)u>ld~TD}@j8;-^zJ(uo{tfwyb
z5rncnk$C)adabUkkf5ycz@nzX?_evIh=!>+{b$*)&R?=#4@%(K3eLMRMMsB`?CQlv
z51X>`ANN4g^?{*}LGq#ih#JR})%?^N$LPf9A;a>m-1B|fwp1k>y7Xr>jOTZ@Be`<z
z<-Xiv_oJlYC-4Sen*oK+M)n)8Ke*zr1VFAB6|zG^p59c`=V;+)CY_$TTu-xRajFK#
zJv%EB>S1kwod&|zcM110-MNy+FmJh7x_vbsFuC}33C^}@uL0>%+AO|S-P!v2r2++9
z5p47O#VPooKJ&&t#fz`!Fjr#1Z6%ojVaz%W*Qr}{LP15nRib$PA33Uddp$;p-SPS>
zf;7JvE0;Rd*Nb1a=GbT4U3B*@1KAzX+5$yRzhKgVEZQ;04w+Pr{hBS+<gBmR#8{Zp
z`jiOSJqmudImQBU3EJ0bX0C-!K>kK?eP~zK{TiOVw$6tGgs9tX5Oh?R(apof-#!@^
zEe?qKmqEFiQ6s8JoTtF*N6y2%8~7oWe}Vmq5M%P0zS6BRCcJIGQVV?j&niniAgl0O
z*<Tv{|DQjyg4j@;W+86u_0(COVQ>91bF-2BNKT20db;|XT9GeA%O1aWAQ4AwoeCt%
zYx@E~H6^1<?d%D0Z^Gwi)z9@d=^xlh(*OFS4LaR5?b!b$TY*vec{_dhQAyumMB)sW
zy0GGYF6Dq>5{|@y|H^gsk41)>HQmu0v}G|26A%l!7LzR~xgDakuG8a@A6^0SrR*;V
zzhT8d0RmBsjUI_}+1V6)l!1Af5flQ_^NfkVg||EHc{q9{1Qew?%k*8Cm(cYvk$XE4
z8yv_$o+Ys4i<1?nK=KMh)cWNC+M}*^atM=g#+kk>Sz%!`^D#N3P!jy(7x#`P0H%nz
zvlrGH!k3qgA9$1_`;6(o@=kuMUiaoz<7nj6A)IBS>#!SAh-T$agVrNTCJ*eCP2M{}
zYa!`6--b>qmqCP;LFJIYR`_3bt4VYY$fbWkF*ef%KlK}9mt_PwgufCHy9!m~G4$OI
za@vmN0ZSMlPJ<ZMG@%Ald#pMyoV?sq=)jd=s%+czHBCyxCMRcZ<0;olTCxP%x=FPz
zB~TTKdY>A@IqeZ@U(-#rgE>%%d9tPW(c6@Aq5F3i`g#Nl7l++;)DMjmZk1-B(1h=^
ziCM*+aO3Hm)D_+TQc!M#PiYJ=6mdl(%+l@fQX`EA2hW<D)Uu*pg!EYJdBUl-v)hJD
zeoK$$1idS7`&<uf(F7EK{e9W3<@USRrKK6J3CkSR0|!;~RA8a|h6C{{{$#0c2aZ7a
zm+X%~+oS$8jz07CB4<m9M=Ar7fBDNG7K{SctqxNf@9K4_P4(+b-6pF&>)04mXBYdQ
zxX=*Eu0sHWmSer^j|Af@0+rm>9*Idg(_Vcv7#qOLdcK~wpAzP~*0vc7`fp>1ycv)h
z!toPvBQYb!(0<>@Uc`~QZau@WoT<4r-uNR8FH3P+Xri3cXB_#FPxmq{<^}Q~4-J7*
zo0%wYd^Y~Z+tqqUV?X62nTSQ2wXNM0*X1GB->Xe8*E9Kiw|q&3VXlg#y86GLwT8}@
zoMWQFjC|BmX4EE5@Bil9wvbv}#D~BNbqJSg0Zn1x{2FTB5%C%$Q1o>U_f3OGVAJ7d
z>|s>y%a0pC<cW|w011EZ&NidM0FUcFjB&TYzh&3mG1I-%8Qt5ET=kW`UIn|THFMfW
z`5Ft@Kh_T!VPiLUS$U@O+BttCl`^aZ=inSMB~ig6VJ`-?O_M&nQ#CIgSFl4xyrv90
z)6_7U3^a6I)5yAic|UdfK|jeo1NI1?A?or`RdA^oAl=(cMfWoH)C?RCo~MZbDAm$*
zJ=H%OdpP&SLA9sCk_r+HRr9g$X8ArO%1$j!4Hl<#1}1j-aoYcraoa`p{$D^2g!=~E
z@RZ*a-Yk4H(58Ex$r=#nKr5#o*r~nQeF86s02B<?1yd|fhRd|$ziKzgfqfZW9ecK@
z(kI^)GFHfmg^!c`Yd^vp+4kMYo>6(Uw0yCoE}@z()le4iTD!3P?BqO->>m(ayWp?Y
zXCDe~;~X-~bU57HJW}zWCHhCw@7G<1D$+bH$05+P3kp${2Z4|k&2`dKCRbN|Es9Z?
zqLv?P-p#&`JPA%|WsI1}=Ep=~V|#ogT|+GQ$H(iB`F|F8-D)X^*gNxo4ny*+HpJk&
z^)g1h(s>L5hk($u8y(a)GQO^z@(*viJ8Gw01*vE^#mjQ6$b@ejM`}F|2IBN16lGp=
zI`H02p&gH>eUt3?CFa{}Cx7Mq($eP-Yg%G#W3Kfhm2Ch~S0AHcXrPEkrBfmxI;FCT
zzIEKR9ZSr>oqvr7_@`nqW1ZSy+!qO7&$&~l$;UOn2Ooeg`{bYg0Utu|2O8n@;a3tJ
zeP7}{zG&RnaKb4aucO)Tz;-QyP*!1zBSD{zGS^09FhZ_V>J6>6O*KgC$PL4XCXyDw
zd`dg-oAz|-O8t|$17mOVqLEo?Ca3eCzg22?0`BmDkFl{kH{3Yk!%moJ$2gp6#K^>x
z1}~IPbFf|p$mwQJbox+ut9|-9wqreG=id-c2}+AV7E~Etr7rz@l^qR;HU8IKwF%?I
z$s_t<6dgy`d&efK?svaW#DDO*{{h`Nz`$%bZK(P8<^$jf;D6C;xX}&1^zyImxGmZw
zPf;2YS@e!Wd%5DEt3bicc|+vf9d_XF26#L^{6D<C2T+q;+btXfX;MT$Iza(ZiYQ$O
z1VunVib#<f1w=8@rI%0y6zK>kEg&i&CDMyP=%Dl>T_6OdOF|70;@Q66^S<-V`Dgw)
z|IGRR8KX0f-XuHsz4yA-wXU^{=)PeI-E+GtsGnkh-!d;;L($8blcZR~T;wfEo>mfE
z%<m*Wo)o&gAsnVYnP#V@MduzfUNjy>nYi1>+A4qR{WBdedSg>#buY#ZP^f<T`xqyI
z)JoD_O7!mtphUF;Fg5xxAGZ|wv_W9LPx|`sFTV{bVHHQu<a3Uf`?`v*!efIS$Csh&
zU)|*gn!yKXY{y7E&%V;%@xrt=#ndC<B!p@L0*33J@bwf0NLXBUeH1-Ad$sOoO6z%6
zW5W|4G1P!XNtKT9_tzADr(bk4>7*SsCPfx(Dh2-<Qc-oKu%j4Jg9Eu>w5C;>B3_Nb
zNL3!`T*mx<P+cAK@Ymbc^D+2T7z=W%zbT$n5#S@eUPjpOt9#opIV94P+-NE~#ah74
zE$|trXI9}PeKJ@o4K?uL!s)T3Pm`n4C34f(F7NUcNJ-gu9^@xzy-I@g2-50M6EQso
z63d&h`!%w`{Xnn_`}z|lfBGBG*SQz91Et_%4n$0Z&3E?pdSO|9i1dO8`e?Nd8V_fc
zLqmNXCy>0u%c#k%gTy(#%Jz(IV}99-*XBWreOx2!Qn$$I>n#HGnD_9<89i@-4)G$p
z^?(@KA*rYF>o&htSEQ@4qM}%~VPOufC1t>tBXyM<_&q;;JmTWkWPRtRz?Q%C=73lY
z-3r3oLi+<rug%P?{D;QFk4^+lB!A6a8ph~c4p{|SU(SOlgjVpd&wdS*p^5F)t1mA#
z`_nIr%?T@J2!+kD9KdA*H@`0Z0{R%Wq8eg+q|2kYQz*9O;XVnPTf>JJ9@)7;sLFLz
zi@dQCp%UGg5olC$pVvGq(~OPowzb|zD;<%ftB&4RFG=bwCi(S2p8a3-nf<F!|37^B
zZgyfr$I$Fv8B(V5Z{><6&K<#`)+zqoAxX7-<4z|tz@+O`Er2z_J!a`*FDo|;?+|aM
zM>M`~^?aJDc0DmIg^%3X{3FbFk;#3mDd}B^Ocl^VWkXK`%k(u~mp5WReg>7VoZG${
zvoWLPo|x$$UGl-VtBto~)xt?lc*wPM=FDdh8zS+UKU2Ky0w4xI|MVs`+{ua-eXe4X
z&81Qiue<FUQK3WEj$v#tAlB@a6wKt)(VEf(3ni*U@00CGbPqxBb>f=$;Y__g{^t2w
z+UX63%QG}Wt~T$P!oSkR(XzNTP#?)y_&(~Ga;YisP14tSdco>V$?!mTZ-K@+sb7ka
z5cjzM0#jnv#LE#mauOJ%rQ;h%ns#ZIyr|%-SzliC%P9MgDl5JsE0)@1*l@M;s^vX9
zT^HHEcPaCt5PQ|VOpj~(n3wP;bsy63-nB{Hy=B@@6C*{QBhK+r4?cWztW#<@zc}1N
zv~Ss9isuusoMg}v*RA|ithD}jsAZ7uJ2;t8kG{a6q<X|@%e8VRb94Gv9<j*1h{Ti|
zMXXug<WE%mOFa0@A(HcvyoIlJ$DZ;A5LJB`?MwFid|e=avMZf_s&{%?$a5>k0iSt-
zb<o6!uI~>-K<%`zp$)Mzy+0rSL#0izrc9#8HPGJv(w)!Bzf<VBjg$p7%}7!N&sJsa
zxRNR)>kD4jm!<1QeVkb@V=eg31$ItLB;|Qu4eenGfN~w@KTI-=ofSEwaz!h`ggq<o
zU(hZ2;5&5kmv%f}I)Yr|@WaDoL#g}XBnW^w8n+@MqMq@+t{Io#5Yw-HwtZHjryJ*J
zomfjNl$f9u6npt42&7#bKXW~It={Z|9I_E#(Gcfo+Anh>NqxYdB)j)2RVP@^H?P+_
zsD`nqb_t?YOFdN_+&lc0t8<f?{>N*T-WaM*+#;543z`B64*N?Fzbq!zwD}86RE?3Z
zzjqqfqzU8uGJ*4WpTLK_qBOs~VZ0nHI#{*$_L%_k_UP1Urgz^~uSGvn`cawrhz@j&
z0?3cjb(*An8-MwD^L}VO(W^66YO1Xo0-_mg6xT6wWr1#k#+L-n-urE=Uaao5CGr${
zC*^Lu5!fV0DUyL41m~r=Dv2ZXi+b~@%hEGHdL9JYJz`l*b+30&xwnWHY`}*OQ2-yu
z5GaMip80`HH;k2Rxa)`Hjy?G@MzosHoW>x$_?d<&LCJ-c`@nNyy?lUZ)yWKt7iPBI
zpz;>ku&^)*g8Y3qHS2FB6Uj~^W{M9jq)IuPm6%}VZb5z!3J2Bxzf!&d6$TX{t<4*T
zrFBfbb$vNy5IL;(D@$$vo6~(s>eD)JSy((|>RJT+(KnB?)wq3KTlNpLG4Fh<N80y2
zH21X9dvoGwMDG<cZf>6e5l7cYN&=l0-oDRhPp?e+0xR*k<Y=E{u>Dnj166jrob}~Y
z!}W=qQ?R(N*$nJAsDwvd;BSGoWo1z_+!=8onr?m^2@b^}_HpcZNbnSPOHU4S#MA+L
zhhjOb(FLGaW_buu-V@RfmO|%8IOy2{4&|8p8V`Jzp3Gc?NE#RlKDzJwqxonq0DqX?
zPgnmXa+d9_sN1VuVAyw<rD<346CKeY;d=QCin*#WrCbEe9EH>|(uwSfQ2a3P5%^=*
zD8+FP^{MviF9<swcjxo(Gk0mILn(WpQ&*Mo6Ewuu4mip!TSR_pyM3~Pe(;@Jk9QFz
zx%w3uY)b8YDAbql9ai|27)h%`{sTFKZ1^7NJRfVX-2|G1&)eFjFPF@o6*i6%QcgrS
zZd%i=2L&<Bo%$_lgjV=QF-br*Msnw{dSLiSg#<2b;vdMXlHC<~y+$>&iL}Jd(Dvm9
z=I7|dvQD?Q(5>Bp@*yAtml7Y*7%Hx8REA&f9~O-i?RU$u-!6`Sk<By$3%~`$vRZ7P
zW#Dnpu@izBOKVw&$$sVEyHIa=k-Arnk$mnm3Bi2SQP5ZTgzJl$&J_I8jZ31Ycf|z!
zM1b2b;C&#0=p#3$YT_)(%GgM=mpDLu$a&*=F#H-Ob!`<s9eyWLZ<BGQsF%>?u5u*t
zni>q^$Z!3GaPH-=n*t8D#^f&ti#<HgNGko`@f>hI`M0vs#7DNn*;R61ri+Q~lAlOo
z*r$s;BFW7VpF)Xl&|9q&iJ#9dxAVR+JY6{Dp778}dpJZznsjd#h-P!U+E64pK<r1C
zK3_en_fbHnbfYu-+|l)V8O-bR(_ezgc$aJa?s3Si;imZ7FK7yP>-e+b1TUz>Cc18^
zY<dR}r}=Q49hfPKf-(aHG|Wm7FAOv%l*Z!cqV~-!OW##{^POC@4C`*P7BxwEo#b79
zKXXflVbM){i8<)pf>qu0-c`bo(+8DZc0P~*djZWD6#2b*Pw+KW7H~vR0g!FSf()dF
zfh~U^U*&gw!k^hdn{8>zN1R>Hs6q{%*L|sbeNrfCQ~2uq35$%uN5fdTqUCt*S!>{5
z$?q#t)!z4OwdB+<jqUS|v)|Qno?5enKmKd61tm)&{*EBw!H_sF7;)+_19tDUD}vVu
zB4f(G3eruh8eaJb9DQb>4l$8+2r*}Aqls%oZhvfwBkq203+|ewmC=`Q9eu99jeg6u
zp<;3vTIHtV4Boajg~K%rzFa}_q#S^SO+qe9^9J${ku?bew+UX28-1^pADC6tyD!On
z?iL*v=R21XcMWFGjTx9aW+zceCX@LHdHoy0$H{z|i}SLOYda7|2=U=@Vy9N%(_5Rd
z5`TeoexL~nfon9xd1vI>uIc4WJbri8xxhyG^{aQm+nyIyLc8vKg=(z5;oOees1kWO
zzsahVq};jRYe9u8<LT56G@?UWk(me@GdJ_`Vo~pznxDp~^VBWFhgspF7!)a);HgVQ
zhgQ0>v<1263kaN1;frnZZZ|bOH!kdLs_k~YWtVPyU;faLkZO*`xjujRc1j1TQ|0J=
zP4c#0lA))%AOqccB5xlLy3>UadGfnlb8%0$`h~mUaU{5gZ6Q%(ek>UuPyg6oge%!8
z{+-t6dow4u!)|hN<Xw93<s4UL82w@RDi|E=F~gSMmhfb&C%#UyilV(4uTS4ww@53K
zz@jpb9~>X?*N<pQ`Czd!rfWPl%u(}FN94ICW1d)YGbbSW{_6O|BdbcB{5jwKK>DYo
zW<2c|w!nzpGAg5qU}|TYXOFixUt2^?dpp5v$X8kRwls?|6)=}sTK6^Jq&Pv^Y*pNM
zx11mqU9YQuTdmEc9CP{iY^a7q`p6hC(!5@!m*si4Y{dCTJ5uUr>y4NRvB#3?VCFx7
zm)smXb&V76sQsU&`-|hs4~Y#{xPn=22OGK1#_b`U8p-i@Lk}1shiQhyf#+2c6U955
z9FNU?JO#>HXmgq3mIg0AP2MSu#!8+`z2&I$z$fnBgp4x3Od{}niNxMr{-2LV2eg;0
z4UslBN0TG{cS-(A**D$J@iYc{59QF>7z!sk-*>4s{lmJMzw66Nl`nM$!eOUK?-ToF
zTO+^D^B!pIefNfQ$i0{Jz8ueUDw|4HKRLb`3`S_QxeSq_^n52%##Ybs7|^S9K!S6a
z_IEzK=ckHudQ9CFe1*(7q9%geJ8b#xBG=O%)t%vR!}K-vR^}On#3WVH&QcZuvEKEb
z6wsmUQs#yEf%p*9<5ms&b8NI;*Q`I=xzW=J5D_=-s6*!M{_-7XNmT?oT~hdhC+0$l
z`-j;c-VfdPp7&fb;=FePWNSi{=I7m5biR>LM{+BIoJ1FNsm)rN^4qcmL*}}W8%<9#
z^R6=KK(^y#)CYLz<_dC{f;>Capo5Na0-iR*m6HxO-Dzp=t-Z<)w8H1#4K+N#HM$U{
z9?uF5%b!a4&})`f)_bH4GQZuse>Fjm3qJUe@C)+ZB0|-S3^?t{cS?epNhtPvC~ASU
z1#reiml=D}kIv)6+_ZnHL$w24Nrv-T0i8)mUTo}opkjgT%@X;>(X!y3f++!w`h+Vv
z?p;vJ62c=KSLjYWztxj;MBI9ycF;pUedZ6vWe=@>Z{Ck<5B-wQpSYtYe5BvKNyN72
zPQEY;HDMq0V<U*8*9q%ro`%b8lRnl7ZtjU?z@GIZU{CtfL9m6D52mBUdZof9LiVQ6
znIa`_Nrne8AxgUgC3=L4J--nc!@^onu>KSL-P&rc(^~gWe^O!o5XZgrQxaq^;{O@9
z_)0|1Z-xMGfda&w@9(a*w6v}bSqmF|HkFC2e#P!$YC-pbYrylve%`PZebN`zF;Mcj
z>1qX0A=in4O&clcTQkd;20p}3ojcjrVC-RP_etHGVjb~Zl`*|-rghfI*+A?@f}He%
zmN&$q`2*Do)<3Oy6jL=n*z;py#Rx%LLsVTC&@EWVtnbp@rbgm%R8;AvgW}zLGd*lT
zr~T_elcvyB(X!v}osea<6J$xkqo}|;?#0W&SE|69aAs+@vap|afeKa7cz+cR&17wC
zQ4;=%e&qqfa>DF}RaBVmHm=)e)AU2gkP}~ACzSz2SIcC3!RqYBP2I*SnF(!Ygd8>f
z(_3TBB<meXuJC?rM4=um@D-g3cx!1|t<Z{)tP!D=WZ>xsKJpIP`4yzAguD8rhlO*i
zcb?o`HQ&E-oaNw4{t&0y^5!JXVp7pp%BH~f8jx(+QXM@3C!^#rc3qZ>@*=yD(m=fj
z3Y{d24?smGeaxl`2{iqy>oS*f*23TVzFsA9K5Vt$q4q_6q|I_m<n{TOxD<nE$V!7N
z9?j-EXN1JEts4Ijn7pOk!(rPU9!fs;K)wZgsRZx?dW3#Sjhl;Qs`))Pksd}bt$wol
z{!dYemqhYfn*y!7tp$*B<WE_*_#rv_-VYhNibuuWdm~ey{hY0ZQ!0>+;EE5fArNV^
z+|X1049;^86F(EOe<`f9x_PKx(Ggoi(U9*G%wx_Ou5UQ<1$<m{^1mq)z5`SIuERmK
z%yIWz>a_(6p+u3`NDbi=tP7XY%)Q6lqrTKe(uh%gaLe14qmq9}%0r9<O6S_F9{%_?
zeQ)}lkLT#2c-*=_)y!zRTFr+w9f=j**sQJgY5wWn@{_I_;Vqa~a3tb1(g(}FDQ9!z
zk&T$)y{T^}Y0k0omL(s5{Q4k!qIpwY(31%7G!IrKe$2(#3cWCBITxXG{^IKg=E>vK
zVxviJve7>;>5Kk<1#afjM3pv6#48x<Cnx2Nfr)j8_+DfZj^-rGPY`uVbHcO3i*pfC
zWmO)?F{V()lRCrC@4f_p_yU~pzDWQzgcw3?{shi`f*8LV(jD9n?fDLDxRkU*hQ4n*
zVU<nEV>r3t_7yji<gB_*7OG|B_FmKO@#`$~d}(sy+cLR%hb8J>my6V(Ou{7<gE%R`
zKs28Ov)w7H_2`yk#v`0)O-RGho8k?1-^-!Buhf3opA(>$>`PgR6tWLf6$i~U;Ffq{
zfD(BWPlg~U*KCVQ$pQpQA>rLXbHV;;i-`p*tAvjCae-wriJ1xuTMzvMm$;jShnI6$
zL0cJ@R03ai{pV*ctc}1t2{o-FfzWC~*6(*yW{Dr3rt=-WjQahauR=KWoWbVREhVTT
zcXQV#^H9pv8JWP#BPWum*VrWP<AuMznWqQ^n^64(?)Cp#nfpWVe^#cXK{R)Q<l1KG
z68}C;LBXx~j%?H3gX4vOBoAMp5ti6y#NNa{IuHaoHylnpZ=1Slbq1(jfcs5;2`bRQ
zmE$|mvcmOJ{>9F%k_#4gqoRpsN^i*Tuh>!1@O#lIb>;5(A7J;U6USTC_+#%q4_KO!
zZE1A}(O!|^^YLvT)8*k@&-`SXpXx}pDXP+K8UlP(x+5hsesk!lT;ny}$;swaLeW(_
zHWTqJe}cr;GCzM@6~I*~vxX5(r~U3#WL+3dGSE5xwwuFJ{&x#0SmW({`_3|`2-th=
z9NDzM*6PMYX8f6Z)%$EQsVEoi2hgxVSsHu)4$l_h?v}0vl&reou`BUZON}Dib?1cv
z);?JA872H^uA{dY1i~khJXH^d&gBn%=EtOQ%*?OOUg$ibb4_d?IToTI(FMk&4Wnkc
zi4mnK0Sks~Ex4~GHo4Ki7jBv)-gD@RddSn=EwrZ^5O6Z^)sd?T^fbv%wXb4*M4%3@
zl9uT|c;YdF8kjE@PioC^$Ti-Q`^22du3m09Fk>ttCYDbvzlyD7ZHs3lm8(h^5xHIp
z5q@+uIp@{YrqFog_B?GXk8R$f2SIhj_~+k2IOLh)e4+3#<{-AC$6wH#q=%=jUjO(T
zk}cyPgdPN6ejW}RI>&NZ{)j$f!S-!YWQyklS2^*0==zEBB-wS@mAI<izF>1r;g0h<
z?B%zf%amtw0zo9!<njM6P^umK*bw3?+@5ymShz69mLG*NlYN>Di>4ul-Mc)dG+)6V
zzv-EMbt(81Nt5u&kaHz$x@+jkHbc0!U2<onH-E&p%w2)HLvxiQr!pu<WtGVMV<JCl
z+Hb8r5ay0qdCt$~#U*@9RENHA5+*u^;?TmBGt9PK2C_@vxMXSO;%W!19ZZ>p7o@|h
zva|%?y2Q&VJ^@$HE7{c=_p5`mOX4T%v>cjbxu;RHX97b_>ma)WnK}ZIIh-QU4SHqh
z;p0lG!)pt0F5iGQ8TD+Tv8T_LGS9wraiq6qatH~2SGYWPI#9F(q)$2to*&+n7VYnT
zP;2Jh{IwaA07hjmh%FqQFgW>~8<<HH^#${sQ4=BErHLnHSU!XAOg^UG(QXgE0wfC$
z!=z6m&3K<BhYfw`{T30js;*l1sa`CeELysH=|P_FE4{`>H{CYo)`}|vZYtxQe;`%4
z1D!$YkA$q$)zCtve7I{8M8lQ|5mfsxu&-$YHRJ2QbQK=X5e#{3t4h3GzXR-9&;iQg
zGxQ-|uRleYwFQk-Uc|?#dZq_Mjx(RJ(vv>*SUZ~?8Kr)0=zfrBCVz${Mt947K(KZo
zcV8TM78=GunbibWXXTSkh^#TfV8in{1N#gs2>g-vIPhvWP8}bgwq_K_LwZ11G9Z30
z&CB)emqHm=YNhj;q|lBu&qZ8Qq9Kl;pAh~)rjQi9G%z|a3ra;D{3mz?o|E(yXYErD
zECYE%)HaB8ae%>^#yu)sJ9Lhiye^guTg6l^(FwS>d?0MblCLi4)l?07{ZKKgyE!{#
z)yorhh#ziZKYqso2)%qt6W1SzW{JgdI()H)WKXHr2SkuV4r{F!;4DezZ8bo!KMuS8
zal-Z~>*9cN(MeSnIAd~Kwj5#nrOW&3LHRzoQEjYGgZDPwLu<T}K+$}B9CqtoATTRL
zY@w8=j??bN$7S;}-s-NSn+Zw>+7jU|bcIh;gXu!XhKqpNkwMa)hTLH1J58Cgo-0z_
zUzY^p+(Sd;g3kt?7$j3IJRG+lb(QCv$aARK4$*w^S-daesJ&1a;JimP0iwV<k*-5J
z0?zas=3Y4aBwOj#4eXbMnHdXMS(yd&m*YX4)`Dkzuu^l$W+<EmMx9H_#+wobqDR^-
zQyNrEjZIISd>YnavYprkV=cY9)zy(IO*HHj$@~Lp|3!E--R}>$U5730&e$s_)D`x5
z_K!!5=q8oN+ga(C{hy~QV_)oiQRyZ&@syg{c*w}ec32YTVqmHZeq!-$hkcRn2#{f^
z%5OiZ`WdLf(b}`{_6Je1qwX4fYTsThum7p}J4*M>_?W992Fx)BfHn>4`-{T<zW$FP
z+rR(AKYyNVA9cNR;={(bf@PXw$j80_FXfHwRX_8gMUmhtvuMZ!Vc<m->eOoDsxob^
z`vdRi4{%Ric>1}YJy`c(eoEFO0+mC%faAF7_l=ErjnUtov*u|S6jD`Qh{ZP!nnN`q
zv`Nt*3Z)8ehdB_f&p(cQ2#Cbf<{e)D4>}5R_@C(cx=kp_pska_(KP6F_x$usL0kKT
zgPTXOs++nJjCUC*WYe*@<VwRso+{}hBCutDUMawZ!-VFJysJCy<R9n^C`4m>k*xl8
zLu!$1WfqXX6B9hnWIq36mVu`s#C`5p4hzVl#D-v(U4~GEOQ_Czzp5++0glye#Po&K
z^V_i_w`c-X6n?+K^wb}Uwa3j7I&HZQu)jolNPgo`oH|qoq>@Y!xQ>~pq$ww*N#lk-
zAQU`Z_Q|;U-jEZL`}Gf4i1ueZ^Ouh-+DnKuuN`Pi*1)>!d|-uVQ9Xrf^9ua>ff3yZ
zf>1@K2Aw(YSHQknk0K9dPia=*>@cBpo87jOFSpXX$IUjfim9l&98X$A!R{{&j{iDX
z9A8#kpYQ?5w{GpIP@AHa6at&0@TdUCliW08V8O#SJ}AgJSln<Gx5mt+Yi=cR@%f!7
zyIvro+Z7WdmgrujLcRdyEB-bqB}XIwiRS=PwBg<3%jXtzm)o4?!<6<)X8{T;gF9@{
zD5nuG0Q;a&Cj0Y^llg;WW?X<$>88}q55d-p8+J*n&bcO#PhZ|_xk-38*&o+>9>2m}
zVd$E=%mSakq%Jtb)nRx-UtAOV>X~EWgm~z)hUAe(@9@NqU^DX-!`4V?!k3oyE|KLu
zhvYvH>oXzBpMY2ZEXAp3n3q#Q&y`a;IzIHhs<`gXDyo6^Pl28G#?*F=tAzpzbk6zc
z;$sF+_JIW{-Fv1jzU{X@jKgB~x;GJ;7M5-H4_%#}WJt}&+QrHke7_SS;|F6*2h~Ow
zI4`M58o=|;bn{i!?x1LmcQjvGqIv>bQifIdAb%Wk0=bBo4n&h~1zwiznsfwQ-dG6x
zbrPKL4sTihigH$xH}b$#^IZ=w&%x1KI$>3%UqBwY-_i(OxVy5?mFA<s1)|<8a@%NQ
zlDt!3ZU)!+cl&+2mbS|7CL#*QHtMG?>>8nTAqRNf;}qC6=Yr|Jpb{XwAF6ed`X*^-
z1b_bd8sx}x2lH<Ee_rr)g>1U8LEY*$-Ls%-Hg2F7DZ!OF0pvc-{Fx38d)>RmL1GK&
zF!><eX~U5Qz5D?`w*{qlOxL|tA>FG1=DmpO8uIHO!ju-^^0d5C)La6`|Ay|59}Lnh
zvv#1vE$P~ej`WTV7Pk5iXQhSvs~@Dri-sjM$DO*`y_9Sn9DXECx~n%Px%qK)S8uc?
zL1)iTSNs}(h;E2`rXX{`1_u>xu97Q^BDm)NF{8kDyj?06ad3hAhlr75tbv<J;rv-x
zj8`c;Zokd-IWu11Qsn@j-ly5FD@lCU1dMdozQsh)74*yYU!ARkPG7+<bfmIq@0q}(
zHbpV#2!NW69tXc+NbnuILAvoj5NHD%VbgMa=UBVbRwVhxs6$B3)*??92SGQ%Jti?*
z7X9R>8ZTLqqyvrJDcnKv^b|cY(!EeqnWSm7CEc4JK?CyoU<(bP3j)aHq^3WRT^uW=
z&lXAOq3CKyz}Ka~7*bwph#HuVBdD~^otFPTChdYL#zfQqL^>LaQvY7Oq9o+rDFJPn
zt3XANWf!cNom_V%CW}8<EJS-7W!q&I3FHXU)A4Nl<v>KGI6%t6LcAc!s<kH3(l^Sg
z;&WXJtUcQNAbN?Xd838=3Cc{?B1FCPx|;vk@;LVA_BpRGMU^)!KzY$+%`&N4v@kan
zPtn7?#BNLjkYHE*H<w|{yg?Kk-rU$MU?)!e(hO(vg7T^5EnqmBqKa&8mPlUxmfs6;
zW-%%3Vx+^FHZhaoB=d*(4CEX$klTBjRcoc5@)I@U*trZb{R+ue-HwxJgFB%Qq$n4F
ztb+AeEUu6;fZDPkUjj!x>PXm#4jbr5LJ`XMJNFf5#faAFK#b3;CtvANIO9(zRCM|l
zb^KG1g~{N{iVH^L?`@K$7T(JD*-l;X%}Fv+aBv1cXMiT|7`Zez#YfRIjK;1lqn1yZ
z!5z^DmjwYzDNP1*f)}Y*HU)zR1i6V(FMscT#dt~Zy($-cyfLX;vG4j;hX)^*g2Z8(
z4Vise+soEcebOU#cP{qOO<8!iTj*9mavU~9JlAX+H6b3s!lZZ{EAtcQd}OvzSw)j3
z%ix!!Z>+*X)G={x0c>k`N#<R_+{A~YW3vKd>%ejRZf}xO#xK);+3;y4a#u+pH8Gx|
z+Z6+D2Y{9GEbCE|4tP8|3iYGz4`dkdl;gqTfhr_@#W;D1Sf7sJ$LU#Z$M1d(Uo#C!
zq&Pn16gk@uBrt=4CkbM0HWKao%4y+b-NU^iSC$iB#luf1Xk0PMyEgBz$l7lHZH^PD
zSPFf+vRTUf1~8kmfJBl*b*v3z*CM(o(NF&Rdc2%tX=3Za$C}|Sh_CClKEOUAGn5w1
zP6?BH{fxDA`ei$jpGeg6=qO(87p_%MTHn~+AG#NBKxMAKVSD+N5z_BQYv3O5f1tSV
z0&Lr93$v(S(mvp4WjC_!frl%pgZ0>VfE)n{axNho5e%X}8RkD}68T)+6cUko)=roI
z5gj!~J`ZQeOLFZfacklqlai;afqeU1BhIQ3J{?V7B~F6pbgmt?M7f{{J(Mp3(+uFc
zTu+uLEuN@lcO$xK0K*}i^K;c+l@;WR<w+uvsfU#^x3md`(WYO?7TS2*ZP2obCrdfu
z4e3`cnYW)t|CTAb{G(HhzbG^@lb_$gi0+pSXB}Sp)M_^C`0r;y9y4t|r`XWAM9;Ru
z-}z<5%KxvU`w|=IiO$)3Y8<m)#W{6$o!-AdFCfH4L4zz*g0;2j2EjFA7r5r_feeDD
zz%|cL>A>FhcUO75<?&e(3Wr3{`fBJ8KFtX9PTX)(*x2UZ8i;dG(&>OaGx<jY!fACE
zTQ$6M^lC$EK?SMuOh!*KHbPTCN8Yg^d=CTGShN^5jzk^vS5g7W=2Ql5>6kg)=hqz0
zEcl!S$*L2}EX`**YKLIyqp#&6_MYgZlMqs6peyS)<i_;t9`deHZnut5TBATwqY9BG
zOSp!ic8p6=n*|_E@JA4h8V|`3K?13jdge(sG+GswHW~^D_a498`ZIcEomcy>BC=EV
zH`l8`TT!tMZG}F*lp3N0{M%Shp#}r2{y@4BWKDmAGD32jGP=#mkfz~o$A`&8(aOZ`
z!0w+t>=z7ak~R~x)a;w_g5ls$UjtIA4>;5z%NpX~6p%<@9Ls7zInx$@^1Aa<SN>!F
zrw?e7yA5Q1=)jWhq1%ZzqW{PdF?#==D}ihR4#+bQ2CxvufILl>C?11pynj3;z1Ngq
zFm1JxVH(k+bg`vrg~qGjw<8z<0%bU{i(_j1e<0)FYa6_f!}<<f!6Q)hKUerP4%5zC
zR=)DRDoA%~Twld|Mui;v<@bDC-QGrD@B{O~KM+3D;tw@0)VMsrwkA|NPXg^ofxkaB
zNsN>m&{{<8e1Z!E%Fh3Ru#>c=TzOXWZdX>nQLWn2<pQe4R;q62W`99i(BV1QL}-BN
z%dMMDY^VuU)Xq&g$PpqEd(!qdC!Fo0TR&~<a&Fy9R%Q@#HRIT@jdGIEhu(|$(Ug)m
z_1-&y#~P<@eV%DCIbvUlX8f0F5Z#eB>gy#Ecu)6H$GJ$*erh42aOCYSj$rW-pY9f}
z%O(z)Oi5SNK7RQ|F6I}M*lELI$jNBNKadG=a3tO=fjoQm_xV~voTZ#uMzNm^U#<8t
zWn-}@vezqPv}?@G7Bjz8;x7eCs$b|M36ky<n0WKDvHi_GBip<lT~0+7`4k4GjCRfx
zn3~!`xPc?Yze<(=tpfSK{daU(fp<w7IP6PB1s}`F#+rs|Ln#-@MHBCMk#*;9%^^C%
zF<XAWYLB6N4+v^mh10gGV^#j2l~5Hk13a<a%iNMrZUGw=um8eduz>|s3x+vRd$a_z
zn})4!2!mhLe<Ys!?CUuzkI(bTF;t`DD!75kq87`ubq_y@_E-6dz7I2rT=ngJ+Ln92
ztWZfwHNdlrb7Rzq?s*`Jkl)6vZT-24_^Zw<?}^6^Q|7#%A7c$qru76ubXxGR-7bD~
zn;{LE9l_@$^Vn8=R9@Bc(#`6zabsug#NdbU)npO21Vt}_wE^r?)&mJpYucU$g7fdc
z(9H0~JfL7~t;QZ+1I)*NS_79z@dSxhKr*u<{WI&cvdAwAWi4(;i%wnP3D5I@P{+8P
zyi0(Cs8KLxMoOBtsz94HeyLK)1v$bF+Uum7#<w<-<NTjt?7;lQGtwLm`aG>k4mrR>
zieA$mD?O)`Dxp2_hVENwTTdWAM8$z6g|!7l1q;;>8yg3n!}J1*k!Hk(gzMf%H})j+
zzUlW;!5oOmaWXL3y>zgB;TD2DqyCGfKvm$tDKdc5$V7dx{{!KF0y*qiplg<9@SSc?
z<NZmV`jEBJj8L-ZjgD)!cV~+2zoLYX%L1^91tjGdHqQ#bvKsmL{qkmx+3Ae0iK#ML
zryvCl^NIcb<0W_0`CX%fFZ#;1)>k;CQ?SlX!3Cv*8$J$nA2&X^-CCl{JnheR7o<|d
zpEo=Rl1ht`{}QlPJJR<!#fHztJ%O7kPTSxuCn3K}<FD4MqXC>pXU4|Yzx;uS4zww)
zQ9$+bz8sk51j`?1k@*$|&e*J(%RU)XG<W`Gkm{*b@Y}yhfIBQ;9zTT%R3dTW;8aMw
z#fZo#MtX?x=6-n|y$8KrE}NfFV!!W7^e)hWbT&Ky>FtUgg6A8@TrSH8NMw9vs?dHU
zXQz;`H4qj1LS)OYtLjOh9O)d+HEIgKdT5d!`(r=VK6dk=)>i1z^o{c|iJb;PWqtm!
zfY@)6I?fE_157Id$!?82>p83In5trdXK0MG2yw8J@$la-K}Ui%$yLJky$q9K(NPzS
zL;N=jX>hIAyHlJdo}HfAnf957;96Ulb_7izo1px6@F<uUtfXh$w*56cra#mr?BRZt
z)*`YPJ99=`K%)8>-5^cq(i@ZKn}Z<Jlsn&je(@7oNK48SBboOYFi?xp<t#oz$HRR7
zK!mX`Q0v0LNj{Ia(REK2+!a-xV9%<<?l?T7+NG@{27vvs#OvdeV3DfGy-9Li9)s|@
z%c^c6j{yAuozRwqeaq}M!gx%_09jbMk~g$^dP$tt?Di5(BxD|F#t@md0DUILn*z9B
zhuV<^W?-DCuXn%0*H1}P;=@AJMu7t}bv7|ST#l93Y3<R1z5?GwvQKL76(aj+9*cR-
z%Cf}QQtU(Yi9UCc#M!mI<*xG=OYOw7+nza49L5`gPz#uwwuzI3DPc__pxFToZ~LHx
zZ%y>ToE8WRH_1@4V^autd{AH#%#@&geDZHv5ehtTc#w$npZ|v?por;!0)ZU-0lB&6
zEzliKqS!G|?1HR;A#gAPn(UB~l>FB|4#>!nKxR$71WxIP2pZI&c{}h*Wu?e!1J6|;
z6OmXjJOYn1nApYSfL=8k9B3(lph~-rID?ua{oj9kqt)p0@zDFVaXQjoN6+sjbJVa!
zxY}1=L7kVFUCv<%Xvt<8gz?-T3+5YnQ2vtmOD3kFw*S=CpJmCT`=wDp>UL5i$j9};
zLwCObIXBDc&fy-(*s41m4rQt#)bjFvhz(5uFeKXN_5eWmaL4E*nq-S*{z7^;M<E05
zvEAG+XN?#2LhhCGU9B(C+lF;EL_@Z3pZi${!{~*!z%Pld)u!3wSomVn3^?12U5`&F
zL-wEiQWIVAr!AY0H5=&|E>^zuDsY=?!j92{8*1bp$6`Rss=Kn*4F`SDgY%YM$KEyI
zZ0}JQ6a#`<^JdmVXce{>wWS(RH4qUDMTgE!#{06j*iCPoAnR{4Z2?PWZ0+%nMgxUC
zOo+ttE6joHDZ?X1=rR8&1l7t;pr%A^xB>{i=zmy*`}OwSV+zx7afA5Ug%=5t)rkny
zn_|7Q10H&3f(D+D`!HDS!wPh0How|O@Qv#}%waYAYZ_i^zn~Z<7Y`*lSK9bF9x0|*
zXp69%z4aav^6i$ek|TfEHTjXD|8{d`|NG|j7mHToBt2?Zqb(<7bu^4k*`$s<ab9ds
zNX*u_a!rIiGkkR155O^S^MSh#4eq*lFt;KE?pOKcCK-}B4$bK+)7GY#Fk`uL7(DcN
zuiWmILX6wJE7$n>!^g+dH^DszF`PuSFs^ppC9*?2OlCB~H`+~?ZLg6`acCZ2bNxYi
zdKak@x$h?BNyGZ(>YsAG!<xewLs+hp4@&o<Aq0BtLllh&(7^ddY8o?W)kZO|yPsA%
zWU~#&Z$^Mht|m&ZkLylYfR*#r8x}f_I6)f7<qk=1=W{0a!w1cSG$Fnm(DJTOe70hc
zcDt<jXM4Vxo1D$6Ea&OyCc-yuH}a>fR`$-0ncb08s2(c5s@2HG@c_7Ni~C>vSKp=-
z)u|=}kY~7*P#saBkOj@O`Lo?xmg}hl&V)U522U+KYJMGYK6rX=OUBop(g>>N!462k
zSQaq@x0i=<efVf0K)bHio&U*Dcj%i;s#GJo1>_)d>+|@k7ls_rL7AtRP)Ttx{?||Y
z`(^&myqF=Lf(pdV$Us)k%Ir(u?aH_$%|y3mwFAA*8J%t3?YzdnC4nANQw)TS?ILYt
z4TvI!ejXHS9jVH}Nc%yTXN(NXG9Y0%Nn!~xjiBhdB!Yd}7l`MceV;c&6Zh$k)_}9_
zP8w*KoQ6rydsVAl`kq&YObUFzHEaIVO9j}Y{MM3Sy+Il$?Av_bc-E)b?ykDg^Yo=?
zbvmbGr+eezMKNwjwgF@@6rn_!$|T<%jyk^n4t^l$PdKJYE?t?$I)Pj722hDc0nkVn
zMj4!ik}QHsSH^AOW50J|O>-Y=JV^7ir?Kz7!~X`793IhcXzkt|c$t80CgMvj?XYq@
z@z2KaXKXmTe$R`2Pc80o|Mq8LA(vHUUa<cHCmTtrzYq$gfRQIX0>?RDD|+uNC}+Tv
z4n>j17tEhGuzR^87i{b#R5G5+hyW?3a}HiatJp-a9j~O+&wPha(+NG{n{OGVGbB%1
z8ofZ*^K#=?RItj4<s9_WUrqYQCGf?nCkWsm3c4g6ZgULS#X^!IrA+~K2xP~bC|<HY
z$jxzm0JJGJoLMtpe@RMkB-J9!*orvms@*gBAJKmDy}Pg<+F)MX_2IGQjydo1XiFtq
zh6mDdxwl{BWO{~phWO5pcO$B_$!Suk;~rN)v(y<rhi{oSKmnQa%J6S-xE~*@c9Epm
zV)@)#O>_)j*^mfpVQdrW<tjVJS-n*4o{qsmUp3JZy+7BRg<>?_gO_d2B{4j>OlR~_
z)2OS7PKgE&0l5iPP;#0pJ8Bw3N$R+Vrp7K}eDx@;Y^Y^W^tg7Rn8|{*#Ikx~%?s(u
zSdYi1ALE+2wjtbRZs(JkL+GA{C~cZj=t-{gGlH-%e5>I3QTZU3i&-1<qa2-|d{2&C
z>8fiMi^*nC!V9;foUjL%)&4^p1q0g15+?C)+GsiAuRdBf@Cl)-#VOFw3;qhGQ{coq
zGgO?WaOmd5Q8v&|iqtRFl95qr>AI#2rGhCEBF8&HE)f*bG$vmwWY3ETixh=R$ldP{
zG+>}BrP@e8Xv1U3sZcJ|#0_95mJ0UfdS4`E7Q2&b2{}5|f#wQ4g;Ra7QxbrPo->No
zfL)@g;22Qvl)KLw$ewxHJxqKctVzrl-Z?%3<`pcXw|;YM^RhD~RL|grZWYzCkbnqh
z)VyU5^ePvNecLFl;|RcL<R7P^7VAi!D**AlA0DKe`%7qWfR`^O8t%ABLyAH)et(a?
zSC;zQP##y2E&t-`Q9Etyi?3f{yJFlcHe9V>lM3c0xfj?VmP=4x0q)$d1{xD>@@_AF
z+AxKjq4i*xH-8UO--H+7lld*Yvvq!j2ZtS@^$Fc$d&!;W%~ZW3|6Z-rO}3L3vI<&g
zLLgQhS=7_WG&?hEBlNOV+QIx;ywGc#PawifEUs<1@oY3?JIfRA+SbQzRzKneJ##ZH
zb>Qc>GAn(VA|Jj)#w2BaL6+-_yt4%D_=c<a7rzm$Q<dg_AdW)1*;bt^+0Tm#l+<dw
zXt9c3tfW5>z<iaKkoy9<*1^9K)dnc-GYosE3^a@q79<>DQ)~41GXQ4JXw{hY+o#x9
zXPcS{Xf58z7woKu()jMTsVUXeX0#p_oA<cjM2+J4a)pInyYbPLb+XXYg~#{Z@Ea%r
zy?yo*fdQr9>{kTRWiXPOq6u);UoGd~*Cl~5xkuiN(K7ckA*%Y9O3fDr$Vme{s0rR}
zPz0+&Jh`CICY}~p0`}!#!&8&j6}4U>{BAuwGy!4POM6<Iu7ZOidPiYQj@kHB<F!+g
zNn6<#wz=O!E_K1L@SpZSjocP*dU>1$zV}yww&3F-|3Fq?qg0zRWEANkM_lZJ7trZT
zJoIQ29(+O2t}O*mhWHTdL0)MguJMM-+F3{a)#4BL_)mXpmdPdwxm{n@47N8P{{tyw
z8;EGo8z46Ubd<vjIW+!94R+!)+1>u#s_mKYQZVClly+a8dRc|<rci~AA20sS`bl^+
z6z$~!n9T{&p1c9^SWrOS49A^4&c-hL1IS52Q8Y0HO>9otXNMc;@-^lyhL>)7Z5z+4
zExqkml#7Byb+^zk*LKogfh!_6<|am$aPkB2ZH^2eG>;2Wovn+1nU0Fug;TGN&0a}7
zKg1-OpMrLZ=a0O5Jg+Wve{uf)PHCoj0_FSM*3GK;!@nCNV8&r$6gJf48YKz~fa+Wt
zSS`y#4IKhJ9Pk&<y_@)kQNs;a-Cda2+y6isB>lE)>|ZK+cwEg4tL?`0A07?t8GuhD
z@c}g+5P!@&N)`Ga?2`O%?6RhiK=Q$%L+v7#Ug-}$tG%^-;px5z+RWR|ms%u~Dmi*(
ztrhze!0oef(nkz!U{Q6{1?b3f&^G8H4|eMY0H3krA6II|>a!vmf1}+w`#!WXoKL^N
zRGJSIAChZ3)O$AEpt?gJ`GZ9*d`$iQVh(_|4!%Y0>Dm!;eqM;vbJ`E>c^Dns?j;#O
z7;{q9n%OOPx3L=)0N(E31h+;GIC+eX=p5l7>`F64P;@cV)7}HVm*pSZK(wCjQmrhM
zl)3mUkz1xb`#Oym=K%tdhoXTmCirSlT82=|a#S;aZ6VB{F)*uKlA<ne{7WzHb*WP?
zzQ47YoY*rpvWO3+gE-bm{B)guGw&-wnM3W(hN#+2<2T#%UKNz!iE<Bqx<xxw*2Z2J
z<FC--)UFd^*<GbyJ1>z(YsLct;U%`pvA5ND`N>LzLa-iZG7OK;yrvp9SH&Ch`UbQp
z2bwos9L@V&lWKTb=FZ}yr;j4k?;T~{H`kbIGB~cq;*VOPyZvhi9qNK#9}wqoRmZ!H
zA8KCaZ7(I-8OjVV$j<=2ERQj7CgJ2m%HW}5TZ!TA;Zn8ps})WUMJ15#k5b>Z(3CsV
z9~e-Hxzh2+%#(Q#Hz^a)wT4pwnSf<1A3TX+BAM?6r?vAwbMyUjuQ4UxD0RGGs7>Ng
zgtx*<1bmaVBuE~Q5b%A}TwOiln`au`BIbIT^Wv7uyhFl^Q}=z8n!9-x&3|Phmr>I&
z`;J+@CCUTT(VYV7WmPCXzAX!Jw|ZnfP>;{Xx;9QUY0UR^*|MVYNuGWA75M{;F|5h*
z{}7h@4@RU8BV`cs+Td*OD>4v~3o2iX?p$_uVCy#8E_+GC`L(+nuu=|7Nkb^d0V*f|
zsuH5U?*94~0}YpFQTyUEmRwpKKoxY{*LXNLOn_#ed>iEL#lNVSoIws*clkW|C|6sV
zp!JiXzf*)CaAkwZaUbi@AT>ykW~eGdRr9WKLx2TM5%bLIkSt5a>x<oVbUfgU<>6ch
z43{Y0^~Ed);pP~Dgf@@+hK!Krx4R^2mlw04X-8LA97Kmlv~2{KSC7aIu!*GANR^9#
z1q@gGxTIdES`LEd<%1`&TV|4g(c)yK#H2Qax4^fQAH(-f*MJA(l{)kRWsujfHJ~xT
zFk|6tkwV6&H;lOzEW<k7FjvO-5EeHB+2nD44;0f5G*ki~bcw8rnMJIwTv)p$J*FA^
zp1<(mo%%gN^#N-)eZQp*@T7}IVAon%kJEv(KAQYGo)I-+IRdC~QAp5NBH#Q#1PoQA
zQMh);rY$h>KwmyHm+6z*Z^Ct&ICWJwH_y-*EFEQFmXT2N;^}UN{OIM<629IDm!IaE
z?n49)59N4`@KMIy&3s*J;5j=d_{=qY-&VY)miP;!@U|2NQ7U5@5NEA_$iO!r+AIdr
zK^U*TA|&4+2|(j24`9-zGpm*}4pr}Y!@3JZ1bBSfLu%J${1%x)v=f3yH_&l{pdRy>
z{;TgPy^EuRayAG6uor+h<^^_3HSLCBwrs@Tb;o5zU4E_wa(355D^l#Yy$K&$?k0gm
z+Igx8yP7by6C~Ld%DI}cVq?nc)&x6!!&@6+Px6BsF`t)sx`Aq#O0Y~5;^)&24W+OV
z`<tfvKc!XgZ=HXSc_f~b#!q|S<Y!qZT3uCKnu<(M@ChSpoUO{2I{v;ceiQyQ#m3&A
zeE%d=K}IIXFc|s$aSjVt2lgzP6(?Y`gFTHLnnt{fdTh+#D9(EOwHD+^(BFTg0Scxp
z)vkQUR=`wCySxI8>RSi(815@?Y;$@9MadJ*H(0@Ume+yU-Gc8}evh9lqUc3M{DG_$
z&mErEh3remq;s`gYVh%qe&yBEsA!$YS(^L#jB@JIub?s|b73AHftNf04W6iU4NuL!
z>O?+OZ2iQ4wRP5bPNGOFKk{m1EY;I9EHZ9}>iU7>#1V=f>nqgyIkD{$^rR=~l}4gA
zz@DcA>hFi7##}OMa#z|G32hBPP~nw*H>U+{jdQDg)nbfGwcWeIBFPv6RGS2MUiexH
zV<^%uTzko|)j{y`O(1j%5kKnipOI1To@Ym=HC+KpYja}MaYQL<2WfuiND}qcldw4q
zN`0LVIjp5XVw2Xa#xo{1T!lW}b&^bXrag6A|MTIQ7wT%0-9@vEB=r0yA42~jF3Uxt
zr#3m@#*ogbl-)b$3-7DomGG*Xf$YRb6kT)-c3nOI9JRAw;wNP(dS-7y;j<4tya*;(
zDZ-?_c~%xA7cVf1y7;cCM&)6gir+{2pi}4B(04MUhY05F1N6SLTLFel$UCz(>Y=m3
z0NIYrFmU>?T0i1{RDcN$z6d^sKGEkmQX0PFIsPuV8@_c9FuE8Xa4v&v3d{XYY*al9
z#uM;(=i^kS9Vf-Long6FBOd3BwF3oNIZNizF9pvK8#_9%k@7D#AJrHQYuP1TU<?RZ
zWe{Azs*-&GRoM04sY1E`MinY8Q?J+{@jh*hT&8s+UaSC>djCbG(}v>zLZz8LE%U^&
z8W<aqy|qax;5DJTjB_WoS+4BB<K`-BQ0b>z`_!BVBMXQF?p?&;5x^fB@*7xSZIB{m
zPH7$n3J20f>UYHaC@d%wNxD_=YEs-@=UTtLd{cHgj15bqkvJ9e2u9lsW3s0nMm_2i
z!^OT@tq8m}iZE_EXS{&SX?+%7V9NC3#u=Gg{0TA!v}*iV(3)IBA1=@JH!kggmUvR$
zCje1-5y#MzkN`Ezkv!>?@%PKFobnP^Ro*h4t@xzLqk5xIBXncX8y~$M%o8X<7-&OR
ztA#RcdijKk_Kk+=D6qVshnyUMF(r>?GWk5448{jBL;Ol}iPN|ybb#uI0dUsldh#83
zS9I42xP6N-9A&qW4ukmt`@a@X>4>U(n`4t<-`*+4CL|{4)`y3$#kd7YY~SAUFqGfP
zU)MmBbOZHP_P|%YqYOEu`~{pDk}o-2MM?{KPxSV>aj?8k_wt-$LoH^L(3BP*#Nnj^
z>UcDO7fNDDVnpS$V07AIMmdQ^=>dSx;o&xhQDA@xAwcpzT+#*B8kMma=7|)>0^@v-
z?j*>Klz~}pVfljs<Q$k1U_B5;^TOAe;5(Y%A+`+wbg!@1fFr-dy!fVK%~!_T-3>ct
zTP)hSU%P)Jf2f^$j9`fm#T1iIN&0_G_8kx)CnpW}-V!%9z5kv$AL2|+2U|=AR4c*)
zn0joT0ynNV821@;$YJ?pIEiX@mH3=tr74n7^z6-@Wtry`ujSd3V1`4D$M;EqjYv>-
zpGE0qm(uqqr;qbpXd&}%5bDt+|HFk~1r+e6^f?Z91lQ7s+(%E*;SxlkXLfeV>fo3(
zoAKujci)k)w`>gC(ui{wvK-r-7PlUZa^B~L7H}TKBOHR;=eEuvv3u-5#i^qX)iib1
zrHtJEoPBhsHhQX`Mn+GrnDdUdu}+cvFLY%*IT6YM8keA^-0?6vX-IkukGuxB?zH$r
zLLH#%yv94=k--;8;twOOf)MJXM(re;M|8DH4qi&aN<tyN3t+5q!*Bva(XlN8r^j2+
za^#?Lp!CdA|AAy5ZsMMy$u5o3>O*8j0=7lU^~qg>p{GWDa@z_P@x)sj-`@~NCa7W<
z*O>hbH#7i*6$8+G545BJMN5sv<urk}H8=onYc4q1rT$Anpdu08Jbbgfqj1^swY3h9
zP<UpT?6>*^7W!`CL_~VZzTrC9%SAj`50B9Y(U5$A;2Y=R2b-X7n>Rs$ccSoZt_^XD
z%$VqxxKe*M<E*6)gPozb63-+1Wf`4jfuc_s&f`+-V(bsp_#*U>69TB)w(I1pppgqo
z6OqeU(hnqAy^f$59n5VQuz);rf=Re@zjfv87N2V!ns=FH<x>~gde5JT$dcdrn<}^1
z3zG)Rq7AUeI0s~I9XdHqiTvS1IR>P2_wH^~Xi23CoPNc~*KmDF<CgZMC3T&Jf9imT
zfsR;`&cHQHum2HP5FkKE`dbwE{ZkYmg*$9ThGCcT^gfXy4^{Z+yUkKsu8HL^Qb`7d
zT&P<6uaRqi#6Kg~gLHR+4F?^eixTNd&IkN8v3?b1k8=faf`m-7^hB#Z1-qKF`wHbu
zh3m^Rao2^hp!wSu{4#i8ZcD2H>?*eCBUvN5v)J<ekWiW#WWGS`u;(90!WyusMz#Tk
ztJKq}ly+W&S5FF^Pn>jYUjFncrvjRz`t>(5ACm!ho;%<mgL?{r0rb@}^&mSQEUK8-
zorMQrKE>XH?w9?-7ikXP@wRNPLEM=956jtU1GH=M#olZeTw<BgnHhcpEB_9r_uHY^
z-gB@LAcBL^vI58s!XRCP_K|v^Y?4WA;O*SEVPRwD;xa)x@sNOYpDsP=sQRww#rdR&
zAc3lc1l5#9rrVSEWq)p%NxruiI-x|fLlYcqJ_xTwk(03D@I6}axSXw|@^5RyR#YIY
z3W>HQmmTi>?5_}U9?)_<=Uw>Q<BAH<>hFeGY1_LBgtex=<Smw<r#eBk7171#9%Ek>
zY>^HU%E0C>f(INRLnZjqCDa8(-4X1-h#IvZmGS}m!Y)(3C@nI;hvB?KU{4bLl$*@e
z33DVMYrN9dd*&sO-D0jA$cUHUdu6Q)wOjhfsHj?&-83Hu1`h}G{%}e+BunazeDu3)
z+^=Ue>pkO<VamNe@7*lqAsfCfzXjcB1P2^6623O(dd&QoYL@@+7PoWwGpudFS@+jn
zne{N@!&?s5aMo{3oOgSuY>$W%xJQvq=eiucx*hdi-F{IjWi67yY9bQGxRk9eP?n7D
zk3Kd4WM=#=)C}UMx}fa);J)A!BfAn-UHS&w-3dC2HU{w=&poIMt|x9}3X;vM{~@3L
z5>WqBFL#P6DG`UFs`})ccIWK2^<o#(<KbH-C^l1?hyfnewT$6lcp&V%yh;GK{Kq(_
zne%dcTQ}dT%Jf2@vA(Q-T|`A?Z-G+z25R}Bs6`Ts-(o#n#`cH6?>+`x4q<bOaU5qo
z^)Y?!^inGG-EZfG9l5^^ErlFiKk&?abX;tHYW|dT2%7K)L^7MFm9O5o?~KP<z>+Vh
z*-XD?sqHkhHa~adi27;{Zh+6X5r+zvTdXwTabE(Z`8XMRc=a*$Iy8O^%kX}@*nt%4
zF2{>aa(mkLmKR3+^^|UoXTQ`CmzCcfig%VcVqD#Fw(6%Hl5P)^RwM;)g?>@$48S*n
zjo@FkM<{k3O84J2f)dDoF(Gihg7=7ewbzwYOSW**HNaEzubujY^foS`wD9#gi@knG
z{gHA_$c@$d{xfPG%4rhYJn9s3Y^pILrF}oO>lyq>M(@43iil1{mi8T3oVxH|?0u?p
z?kt&azT_V0e;3#{O^7^x7xD8VtN!%$7U4`m_dYX1d?%XP7i0BGXIS3D^&{OkogClr
z#EG*I_6>w^@F;4_CJ=0UE^~)^oDay%;7P{fphxr_fBAnGfCE@Fn&hU)oG5<lnwZO_
z_s4HPb}?(ebzSw<y0IdjXEVkGOY^^IdlP7=|Njki7z~9$+1H{}NJ#b>QkEpGlx-?X
zl0Ew{mh4+7MHr=$#3aTR#=ced2w?^x>x^}nrTf<RxBTxt_uO;Ox%YogPUrhQr}3H3
z=ktEQpV#wxp68`h$1R`u3d8>Mrv*;YV$(BthTC&ZWNUX+de)Ts+RA^yJz+=be%s=A
z@TfmGNpJMIGx!fdf|6Usk>T*E&w~`ZUth&Y*ECG?WB0gk1i#luJ0PP1zTHD}<Ea|L
z31(EO`GAe@ku@nH5~{@rZ=nsS@6TC@rS1nq->{OH>5u*u%tisr?nQ7y9vW(r-rV@7
zdtZvS!BfS?wF&M5<F*Y&0z4zm<F_o-s$AprF)I!Pd($FJn9u~wf$|`k=#t^4-pjRG
zH$e!*>C_a>BOW-(c~|Jg;-1y1U^=e<1@nY;2P+F2sAU0~pQTkGcg&R_!62+MlnJ9C
zV~*qF7ZhULmS3PJ?!4%<^8VgAC7ygaC?_zpml{mQG$Nqi?+mQ#Xk~6stq!_moPVpX
z=Nq`MC#Q0%s{57LImjw@KOIU|j#Xc%wcqTVQTSZ3c37zYd2iN+YvsI!{J``eu4&?t
z)IPG%8^p`p5wp!<hgJGQtZwVsWnn$F){v)jp!V-1@GUY~v10f0Z`9UCUs+2Yarh`8
zat7Oy_}<7ni@5O(u!E%=6^+mF!KO^fpJs@{Un~b|#t&Uz>#8SuW;sa>F9Gbf!ZgP4
z0pMx`js_$3z==gyaAW_i(d&T`TnMETp+4ESS1#qW&U3jrs@C`pacDM4eYTQtTuTm|
ztCK`4k}H}P;y+_~hubTxj>rbgrhWCu=<2c1Hvv(_a09!s{VdbkghdM25w8s_yTATe
zT^<yzdeXB0`Q+umC0lgf4-cK2Rv=iF<Q`ckZRV#D{UF<|Sl_rZo8H07>&vlN1$nIf
zSS^d&cQXBNmZq-DB}P`Aw7vaLyWpr*xcZlE%dJDQ=Ql4U!SKG)ya+LW7t*4iLsNND
znjPPr9KSWWW7<dDzD!8H{cL>>V*TKgoQiI0_3L|k)l}*)YNrP6@o@y4{#i+HW(_z(
zdR_IU{CsGJbj|XrbdJHh-r`S`Bwv5Xhl#Q&%Z>vV4m%aRc>HDYLp{Z>29)-lux(I0
z!zUsoDH1#=I)B0=cwx=H?mZ(4IRhR*g)#%!r%zWgULd5pJr%=;F2?Aj^SAl>z766I
z_dwu(`pc6DQM#fp$A4>|$`&oFYyV8s7r-|@HX7!VRwbsCQP!liv=+e<6@0eoObwDN
zo}9Rxpdx?UE=&1i`)~V5GOa!k7W%N@fv(2ATS-o9htrKF<cdbUrRqIk2HjrX1AxS|
z8~~D4bHq&B#c-W?otUqeTMquRGihMEmU#r)VzJmu22bWo-CN|+;vYz)-?$KYBC1Z*
z|FZwdU&$7IxnrK5o;uTjY1Q+ur$dTA)zGH4=qcSon4t^+5hEct0j4lQYXHWzwwnq0
z0$Tb*SJCD~f;d{2Fi?zBI!;%Tb9?6M`Q$28l$&?Y$Ig_ALngC=+l>ly$<dhAdk6&B
ztTULFWsu;s&4=J)qQKflf7_J?^kHDP$&U~mWvJ+)y>l%Y@wlDVXNj%0^ifK3ZP%&I
zmJrX~X>7X@c2@sPV*?R(6AL-i2&pjn@QsI**wA~2_ig5BK#m)6u6*_#tDN;!`S$Q2
zpwNL%5K!nKv<?8f=kK76b!{|l@O6xB>2Lo)CKY~nNAgqcm05om5G`ZN^fD5kt*Bx6
z(Q5=fZ>HdODeB>Fg=`F19Nh39$n#wIgAeY)R2MUfU-WXBtM3on8@Ajhx<pOpI)hH=
z@jU3&mpTLh@i#J!8E_v<6~Z)rrcFfKi`0-MLg3HNC*1Rtx&G=(RAQL?SIc>6xvZ^@
zE-#m2FqqPUDT*a+(J9}NWB_K1?4yH&NXb!Bdt$l<cP>QUe03zn__CKru2)wZh})UY
znW@esOOIXw6#;lOf9I~OvjAMeBTk-Rh7X~r?BGN8*`W3-)7f$K6U-ip>#rK7Xys8w
zVg6IsJG%OXRQ;wr>Sx8nrqs7*QHE5YWjVi9kJ$ue%5*ivq6Ea+uk#OKR$>X%xShLG
zE%v-*WM~=Vth>?!`l*!HzIP$E+Bsp64u^ppx<Js-EPF7VV#72?k=hvOWCrVf$YwY7
zUU$^Qx(Im0^+yR*uwQ;bHZ7s*g@$r8KF6-#$d)UusyOuet{(SU19@4TNSif%8}};Y
z6xfKX=}Lbfo|yG(5X4^O@Q-s5G&39b=N=0uU<<1caj)rSL}@Png+?*A3?C=%>NA|l
zw+`h!P&wRN@#Pl%kcUc7LLe(rQ(+xT5~inOVVECM{eM-G9x5T<knp^x((MCT3Q+uV
zm-r9j!{XW22_Cx2loH1mlfjeILg|`9Sf^%lw_FxH=UgSk^>_VLSW+vPyLAtFmk`}+
zArUot@v$t&ku7ul$3O|+Vvzc%3d;Lzu|9LJc_qZR6816`Bhv%=B4;jT9ceQNYs35a
ztX3?6#iDA#0Dgm|T1a4U#S)x6IXQF0sPveqgp9M&p=-8Tj;WQ<@!Gpb1@5JqWZP@7
zJbf|^no_|Et9676XyHGQgCU^mvKRUT*-^R&-IWzTDyMp?r>9Tw&-mXFIDvO@9S;!M
zj_%|d?#GF?1TxUEG<QsU3n_VL1C&*U))2QaCz>vR2G_U!b`nwpzTlP5ykhmPPdmmj
zVVh;u%haXD0=#GJ405v$zbgOcjXidx_q7*{;(F1<e)xALeH8bS;0oUsuTcjaK7sLW
zfMt2asqGIWIxy*_yAyBOZ@FcDYZqr(eT(y@+$_s3-wRjEJQJ+upZ>35)PG}N&v69R
zg@lpPE6b7Vf3!9lUh+-+`Se~!-<NPJUE(_OitTtB84C0}1vV(dJl#+AlV@$>qn~m~
zYA8tXFYa=ltN<GEDzr`Pr5~O8@*)V!|GU-Y9bapjv8)j-aOsdqJ6jfX-es==m50a!
zzg{-5Se79+|Ghb7dkcP$Lz3&Vhvz_0nEe}2Vpz2ehwOV>dy~L-Ak)rP-<J`Oc`n;6
zb+=Ko%M%Mmc8mQ)?*W~%iF0B+wbRxWS?qbJ?WkSYxyJt&cGf9$1Hl7)ThZdL-~mWR
z34bM{R?Nmh?2guv|4K%Ap>0S`wCI$p?fuNnbkMQ9X_P5yEh{6zGivfve@Z^<rtkZ4
zKI%~-hHK&R{EpF+3AabmM`yEt@u$)CRb-?MM`{m(9x+A|eUI!Oj&JQ#fAKq$A!u!~
zL{be15M?RpnSx0@x*D>FI!45axKQqmTwDJgtePh9v+=e7KmC}6W$OjKM07IkJXo@x
z6l2E2h+UfdJ*ISpo{r%HIDCTHKESdLYmH^@QBHERYUQ*2QXJF2bhy49q8G6w<KC(^
zAyN>+*-f^iyNXed?~TFt@7v~}S9MFDNcNQVRa;4*4RvG6G{DD!eQBr@uzUDIWT`Pd
z9cbvm|1bsPrk=1YBSqw?GP*v2y4T!&PL;F9zwM`g3uB1if`+JIn{;Q&5FF-YXZoFr
zdUCf=-7R<drO_?^h|itl7T+qg+H|}})z_Tbl(!X_VAJ{dh|pC|xYNCTd|V71={UwA
z-~}qDxHlkKsf9LJ%kK{(Hix&L5;G=wC&mw$i0X9*;)$q7Av|QG_dDrfZ=Td0xpX(v
z@@XEdc&vFz4ro>-c2!RQeo+JbVnU!)sQut^yw2Fcv(;4?Rn$S!)aAB<5nhf!hKozR
z^j@?)QK2v%$LriGUN9hQu2`fHxz3&p*=n=hvo!bjMTN@_VnEc|h-HDd!{IznkIiKV
zXQiUADJ7D=A4$=bQ+?H_Xv}`J5{1L&9a;z7?fn2+Tl4!@`$bC4(fO*~;9#`@r?tY1
zyR>HV&iwx^t(j9uJBvp2<Hh5R29CR(KYsCxD@*Dr+3MY2Kpe6nFHHJHA!HauXsy6*
zPnU>8HSG>*&T}M-Yu%!fa9c;GweY$)Ml4}Ec!;nl3Vhm^LlK>?Q$PbhCHz(Rjpsy`
zdFT7-pZ`E)%ng~f(|<|00Q0RSN%Cgj@;$^@kj9f=y6vv<vJTZbQ{&YuP+*d(Xy%c>
zgf*(UAydX?*Yfbio&&I%o0vzT0lyn?U1B@3yPf67(#-@KJB}Hl4!J(CjC8^*DZFy^
zIfZbU9?f{<G=ubTlUB~h2l*{B<eO%D@X}Pvj$f^ZQRQ-@HDnK8;h(0XTjiUyA1f$G
z+)m*pxbTco=1;=rWGJHGZ{j{ab7uJ-u{=`D^v=`p*Y`NgRCFN_xwOkeO5pS}^-I7k
z@&o_^Mk-VjFgni4(JANVl$EFVG;`P(H3cPX7SjOJp0o~^cD{_Fx~+95$*=xO9CA$2
zaP16$BpJ*FtZUNyaYdM0-P>%`3g+ig%$zE)pKbtd8sD#YTp7<bRZ&)v@I~>H#4!-_
zmRVwJ)!Mx+mvbDvxeq3GxN>=w2+r=nz_VlY+QYNu<?Kdr+XV_PqQ-?NX?Q*8a{yhN
zf2vm4EKEewuwc$J9FZru-G(|t2aqMrhMM*VV%N<K6NJcPcM(!|aqTw1W-V5qsfR71
zu;lz=tg(<Yu|;5SuNSYMXQdhT1?FcPrYh+>nLT(Pd4*}dw+#@Y%=9t{_8udfFgpbH
zq`F;<RXB%=)}C!*z)7Cp(F)!!cSu3z-MR@LzvUsZ-fN<~7a;*!7ghHR&0_Cw{ef^`
zI$nTcr3ny%bG{E*mgF~|QiHm8EMJABC))YGoWy}!L-Jwb-;Irq`Cq<nQoUxtZX!E<
zXI_rm3R-+_>Ye&_FX{|1yq|At0leP+WY8(J?qS))0$LD_@21VWGmq6}Q7<rsF|5Cf
zh?m6Xf-t_qIuRJYG~JbMKm#gT&3q5;r5(U@h(|C9`OMp)dY|Ih{~07~2?9!0Ui%Xa
zUzk$ZGwp^=3|-y!oS5rudg@k_d|7Ak8%P_NAa|j7k#mB6z+UtT5P{b!D25hfp-6JO
zGgGOoqJsRb%A{y%%2u_bt5-;CBC$<2%S}FED9`F&nT>@1WHzeOj29RFKoEf|vjBAq
zqRf8!TQt<2BC|(j2xh+-ipo5h`qf7AeVq1w*{6C2D;==f5D%-lXN!Q(lyhptZU6$W
zR2|q#nl$N}0B+k&Fgd<3)kmrM;5Cdwo0DDyL$CIcwC&0#N?U}Nzv=`&Q0SyIv|X|@
z1&A~pIowTZJmJ3hA}u0IeAoj+ncpZ>xUN`q0ACh+pdymZvrUY2NxvTA3%OcW_IJ#(
zReDH|02*P=+Qbw(=N{8J;cz`Zfc8=2$!OQa=g$>m-S;=Nw{13O!Mc+FN7SHF)&Y73
zx87M}Ud?x&DOj@=O>jv0vbGiM{~KlZ1OM4tf6)F?nM%Q<Mx(~(+rCo?4kB$LxG%>T
zT4W?WHW0>$SOTQFbq~jw7m^PlNp)!}vCfZuNbwl_YJHS{M8kGN{7FQz`$dJ~+#Xo!
zb!In)cX36gD%mL2IJW6g=8W-agDz3i2C3Hf%pM+cmuA@;YNBZR0=D!}@rtgzR#zkM
z*Bweu_QJ5j6k*hfY3vGRHUP|HgooIXTe=rFbq(^jPYv~cnLXZ0P-SG>=mxxaUi@>T
zTdKxtNO0Q<_5R8K<NWD=r#Sw9_?x8xC<6*f*l5)ohPZD!V`JgHqt>~p`VonXTq3q)
zu6cu256{lIiL&TWA3e*#=5OzYU&ONA4yt55^JrbMA&J97t;b0odbsAw-{h$5*8ndg
z8Pe+hJQZ~GStb<>Alkw1syeZBQ!1W>iGG$|9=KQm#%3OPdJ~^u(wR|zw?l+&8g45#
zi&m$dp^Br19<0AZUBYM18|?QQ(=QNl=n)W^l23ZxF)IjIvT{HA(|9C}tqHJbcW^Rf
z6g|lr#=wk%<=&DD|HwqnWA-Tdj3K0B#2j~$il);q-Ko*_oJ~HFmk-NgwJ_}0U6tm>
zW5X<_vX7qR;oWJoa`qjuK(C%OJ4z!2T8hUK1+-JfYrxX42bEv_#3mfzlDj}S>J0b}
z@y3ZwK|Cn^JM^XBtpAHY4Zt=lCx-BmrK%XoAI$yKsR^cXzA0R~xd9iqZ-pCcE`-B9
zuc7%V3k@MBD1vqI`Co!vgFI8<BP7n$6HJJq^z*fZDb2RaEa2VF1HW?#2)1^s&Fh8x
zi0<^KD_oc#M#_vmp$HP?StEws|BEHFpCT0VOk<Sou4AT&8B4!!t0)jwv={^14gy9_
z`jlB5@$$9s+?&fc<kMY4H;zfhz<zoU4GaaHg?NM<ry3J6!uMC+yX@@u(^!5fb~WdE
zJWKxpQ3J$mPR^4$IMs}Zh0RZlgh=-ajL;99&6rKUoVzhvn*@K5<nkFf99*bvAJF@v
z+x$}eON~Dd__TkwoVS4<eCNJgH_Rst71h)W(_#Y%ixux4@eHAIPnYNQPv@P5`td8M
zXLo&|in2<wq>@z6h|nM?>J%D~AE^zK6Esd!8|0@I`ZE;wAIRtsGj&&+3v^B_({(DK
zz9eGzP~%pNYEqf=%e^&KnO%Y@{Is`|O*4}4m+kYMMat1g(V9b3jN5j~Mw!cd*OR}Q
zxF6)txEj3lNPtU1KvQ$t5t<ww0=n*3D9;#C?#T6^A$OWQro-hwURhe#F2gBB(OBXq
zy2BSLGOw<@8#8x45NGNtW0mYN4;2yx-Q@vGKx3fd>x8J$!9I6hn)ygBRR_G?`>y%H
z?<^r(0Zz6A2I$34VYek~3_xuyp3D&WzJM*k(VrG3)7Zg@__9r3+6}B&mYK}9dP1~y
zUB34_d#Vqj-($v<+e+=so8-%a5=Xi}7Dv=iz;%^~s50!{-zvj;UAn{E$qFAOIcDX;
z7I_s4!5-Uxl`2eLxH90TTwJ4_6Wo*`w_5x*)=%p7vC2zS7CYfErZm-|zQe0dxG(RM
zhMXuWsC_9Z!aLP-Da+~#z?H-26fg7(@<$$ue+aD$F8FR{2Qbj#m!Nc5#ms@Rkrsm+
zE5jsw(1b=)b#_xgcZQPvo!G7bN7kFHIrS{-f{RFqdx-eFwo!p}(+lX`aD8Q^WdrBu
z{E_nBDn|mKSw6(Ob)0$&;2ink$2=VZa7OW|iJkpVSQa%f5YQ~KGxL;>h6LDtW~<2-
zmYa>_g-F#y6h4xT!_j)gc+n4Z<+pDV_8~J4OJ-+neo;~M{NWTg2T*L`{AKI2`F5rS
zWn4E?bZpfdSXD$|o@h#iC~$jTFSePipr@{YzU05Po#>9B;*hST04;>b;2rjkiM{9F
z1<V)Da>EV6)^UMUIEJXRvZo6nKB|<R<(=Z~F5|A}sK*@{dz*94%ZLT~61p~6hi61%
zro|^$c9vYhSqRp@#dn?%J8d3brrm@3YHR%^FDFq8W$9C--*E(6i`JEGx!v#k53Yh|
zP7h(6+*Ok42X|)a9HyaQ2iaK0@cJO5`shb%E6E$Rc-$kBK>VEBladRIhq$ko$cju)
z^TFPrxlo35S%9xlRG`ZL2VY5#VoPz1JtMOW1}M#whk8KRZ5~-ZCPXaWg7domxY*Oz
zQ7&4KOW+N6o~X!`@5xH_KW5OXuvQ0VbHyXUXfDF02motm4D=<f=xVb~$~!{cOxSME
zv2$)^-GYOvO!~GFx%Bkuvq&Y9gZKVCuaAhT4nKK_&utHwW^rc??k_1O{9j#ZBc>c2
z!xlS2>fJONfEv=cS4=4zfAJRSy`&OyZFWtaP-WuFAfwBoHq><e5ZCR74!Il=0Z#Sp
z?<n_>)`?Zq<CsNx4Q6~?DEQZ;9HuCyr*eWS=?xN}6nb^FUOMN9_b!zc@pkFxA;x^$
z_TQw;19NatTB^(WhyO~(xgMfUu}>iq;I=OIGF7XJFB=nhx@ED|F||LKS!tnM;!jDK
za$|a`335>$&I~&QktPoySV0F|dxFWD4;+8?ZKN?53yM^%Q{mjuQJ?&=R=K#VdSDvq
zOEx}(Y4FGD*RPP!kS{<ZTBEWKc&WY&ApTT{qIL(6_A0a5c2GFjD|<7~7{C#ZscU7%
z;k=8iUYta!;o_MZ*yaG(<!-zIujD_uiUH3yXLqsL>2grTV1SHZYGb1!QoipvTI20E
zQmNZRr!e#U;i3Bh{zR5mMD$}XxcYV&>J_qdmY#y$gBi+ab`7x1Jp?C^U?=)4Wix_I
zGfuul5o~((P5vqOiMQK#-16`5uyn|UYKnG1Hhk9LOW?qi`CC^5=`8yZ=!f{t9P-}I
zxQ-&D5BkRk3xS{5%X6Kh_EXpV-cilti!(_<&QmvTr=Q~fB+DJe=3&4lG4v}j<SANy
z_nQzWXwVzr$bAHa6g3r78!MrTbmUEh+(_^zuER*jaHa8pn;esoH%u<7R8I3=m3mcs
zuFjGHZraQ_P#8}zDss8)G;A5!&o1L-n=;GJoQwWi>6~+N+G}AZ>)rARFC8#+%_*6A
z$lr;?OvxXc?=Ni%XnAdUU(T5i;&-H7TwkFpD2_gfdxg9|u`R^hUF$~)>zqfRLmvpv
zSjd=N0~Bb$M`P_Seo_DQC|O?od6Ybqz)@Qw>6mDLiaSA7L?T?zWwo4e35^8zMM;BM
z3y9vHvk-fn6eD*-IjY>l*%NJC5`6ksy!wYzpC8TFipDD(-=`@epVlKG5hNT)UDzA}
zB8*Xeu*f~ZdM?rzu2CpvcDqbU>MZW?FtXcn>TI``WDUff!f9{rjOI-9X=AN-#t8~M
zAq74Ef4iaoub&4~nZiJ*4sHCS@^KHJ<LbV)yz*KRFAhv>H*^@)BnDa`^U(ZMzr6e9
zBS~0N<>lpJdV8i>?3-(DZ4OnJjW4A%G|VgX&gH`OsCg9cCJfKA;*w8EfM<|b)HN{5
z&hERazVFXYh#a3*N5gvYl_qRiMc78ITeta%1tA_E`Sc254<leZ&=5?C>8S5g>DCmL
zZQTu-KW1&k2m^PzcWo%fKZt2K;?6(=K>RWeLNnfs96|^W-`BoY(JU#nU^Re#o4Wr~
zXUBfWspE|%%*zJ#a_jP}m0%FHJe41?jCHBi#*see8kf~v%RdwD`a3)h7nbRok|RRP
zh>!hSjStblvg)`|2A<DV<c_NKsxmw1@czyP%nyUWOt6M9_D~2-7%u<GAZNNoUR#h}
zrGmt(8wYRlY910b<JE*|cC=l&5h6iVCo1zSoZ2Z)Ce6JhROq})Sw8&YR;#Mu&*ETS
zRUNCBz;|(nqBez!43U~A{|@{#F^;MUaeVxvIq8=Dsg>jNEN&hg7NVN%T^8XW6A{bd
zk84g3RI1U&Cw?D|A*F7L3~D?k7d%agkt3_B<f!%4c`<;SsBHd!xQPI*{@(vTcB1vS
zfBD8%>|V6^l%0_dP-xGkjEx@*xP2$n<T~!=`3Ij%MzvZD0UQkBASV@PzKzH|>;CTG
z6Vn{$yGL{0oJnY~c#z~_(#2K090ys&#SkO`ZDhe$7hlj;V2?72Ta7=Gl76u1!1Yp`
zrVJ&A^IeEEJ(wJwSYyf?drPX|PLOsQS8b%=p<8rFUd6YB91l(1SqQy|W{&wkGkFt$
zdIOu)Um|aV&@7!C!cRuV{OVQb+|U{QaaFXgX}?>E%B+Ks&vr>ftFdg>^p!eEx<C0@
z68%i(;cTgVr{H61r<GrwxnVN>{)z9bsP0s#%QSXGmT89kCj|yyC|&AAe+J*>^KSg*
z^Ogu=HKt@sWhpy@$!?x`vGR)j_n)c#ZppjYgp{4U*97L)H4wvcPJq@11>U$9CsK<G
zE89CJZJWu#UzdC1>wB#`2^o$J;ALxv6)VibcSGxRZ2*$vQgGS?_$YPZU(){@_n-jS
zNQCs%ag(ECmLJ-@+c2YjzEbX25~q=*u_CItkS<B)6m9k56lRLiJ;>N*9?!BwC7-=|
zhPP6W;PTF_)w6v(U0nZ|!vvsHCm`;n-vl3DGzF*awZqH}7UyqX2md`wn%Hu~tnUWo
z(OQtO8rWpfT~{019@)x)!w}PBw&uKa4=&DmuC3PlE8!uc{>TR5v@xo~nYga`hx|3)
z6w?sB!{-xP3|11Hpj!3CyS)gU2(veYn@osVDYJ=D4;qcv;=W?)#O*5Lx%EuxV5$L;
zk`OmSy+z4u`~%_lvB}mjHb33BBhp+bxNZ`4jd}2eUD#-=cT%3{tTY@AqR3KTOs>$l
z?J;w+xU-DM$mx$@MceNKvX(NCl^+wb7mc9smUzFvUniMlvyeWanDX-*pKq%gUgEfg
zcanvC<`CJq3Cl&5vdKUz=e0Q~)$f&yx^c)0rY|}iBh@4d7|D?%&}N_`I!Bc?98x~n
z#8DbD_$k2E)0yq=;afw_&ik_@5~f69q9XDIR2xcA6H34hr}R)I$UWeGj*HEB>$@c>
z{qfH?!XS^~TGN3;^$0nzbH`u%w^e+MCJ1(lE8hYA_CLJhP59&y@A_)D0`I$z7c{;=
zpU}>J;6HiUq2O{Kt8tyyQj0HUZ=IsQ%jD_t;zvUscGkDvl-DExX^ye@jVHza3y+Ou
z0Fn6?n1aACEq@A(5#G&AJO|luy|RfVD@EmbJ4=>V_<wA_XlrA@A9o%jkp5UekE~Eh
z1-|ActN^i=*+jdlHuCroz9!MqJUTD*Lw$jkq<2cYgwxuqx^zYiX4;ve%m^?B+h-4~
za)J({x?+O##=4^%)Xp!-Zu|N17n^o8ehAZ_&4pN5kfdS2454i|U@z5A>UR~%nDttm
zCi*ye*+6(R>SwC&w^NeoaoQ6Me^8^m`Udg;wX$SNbP4JpC4I54GMRR~F<BAs7%Jgr
z*9VOlaD>jY2P%X2Bi0(;oPBD3rD@4i+`7_qY5B2Y$*hhkBqbsu9PaQNm1T?%zC`23
zd|d*)BP7jJU;5g$V57;wev9hL0l7GxBe$L913XI{An8guDhV?mtp2Z5D8hC?_Yw_Y
z^C-~wkVQQa71@4lGMCtW4ONzy*K^aM$EpTMpQW1y$vjK5VMH;bw43%eb8ZIv^z{|(
zEfrfn<sGQdfJ(T^gkXkxk);*%6ihrQ?s+f_`J7O_)2NXb^}l70{bn9`%&PBG5zVb<
zbO*L3_}iX^LWl-VzxzS8hwN=`Tkm~gRym;Ugj6U#_qZvCO9W8`UGu7|XT%VuuTwM_
zA!k8RAO=~7QCY~Hr_KDG*mh91p~%%kK74Z_0%Ym16(#oUH_=f=lYUi67yDisZdEM*
z?9Q6XN~>QN9JIoQ^dMJ_=vH9eH~~@Qb_G))8Mve#xdSKv*6!xNBTLGUqq94a3u#VG
zIuEgXf2^$aU2|83u?Y!yeeLp7_LY|v1@h^_Igm58dQVCZoO=1|_*}R-5^og_qc_q`
z|3E@P_Q3oGl8c(wMA%m&GL*ijBJ<h^zP1J5$kXnjVSI2p)Wg`jflbN)_w7FI%H}b!
zw5{F)XmuaO;Ae;UV94DBZp<9ml{#oRrk35q6;gLp4lMm@l#;T3-((kPjCzMHfBmZ8
zp#`EVigW%|DC`>0cKHloKzj;^!vZpr0oDC`IiTd2-;`#f7MfA=qn6ZbdGDrnkwO-V
zhZ3X5jU=B-@p4H$N)V&B;{b4URNj^a@`E8DJ@|SNRz4`5SaanKRjtV{aJ)*v*vYMG
zB*MFWqA^CNM3?)qo{6Y>mAwMZ9+0r$-{O8CGjh2J)caVFqM*=>rNUvm$=6opX-(LD
zx;zp^+}1e}?+vlXueG81=ZimZwrjC@_<swzZ#9AWG(>;g%^M81;J4Ng3|(yn_{4-T
zt04fqCs|#YBmfjeN{$d(-M0L4uIySxMR~bF)hO(NXE7M)kdqf`3Q~Hk?4h*#7?86L
zw#2_d>!EP!eH^$H<^NpDjT|6)6{ryv#!86Qg}uv7QL8e9_{)$x>%(>U3?!^(5<?<F
zzM)pn>}pkQ0W03SdXY`~pXUK^_&?7>BMFTntwbUA#cf?Sbic)K<Kkkb`OI6wBc=Fs
z0>t7XL~zKe3ovNi^fU&X+2{cBT}}0co^2M8wPYlfn<IXXHncH~aK;pD3!Iy3(J+{E
zsJ1$F!J>9}^K6-ZpY8JOH1|~x)e9!XK%#V)_Kxh7JN-p|<HXJkVdq$=9s)U%SHE(w
z(78@l--JDLL_)c$xvDsC3KheY!Zp!!y2VEmQ6_|Yze=LUOfzqLG)&&?wA$l!;TLf#
z`v(%cv2L*h9*dkA3C@V<2lc6;K6eSlcyYa=e3r?V9inKGkM7fXwf6nllDv$ch?ZpX
zxt%L*sC_j`L<?@`YCL^O{mp7YNmR?V@rU&~j|Dh-KICgbQ+|eGv|Q$-M>LqGRt?EC
zs5FD$MPI7ag?5=L<nsJgYFjg(Y^7zo*}b?{E?V1z*F+!UzB!&iY+Q{J2^V86_&RVb
zORPAP=H~j-<V%AnchVz1PQB!VKvNBZl(6(P0C0XB-o4x+h*vs>B}8oGXM?A3l&tER
zg@dujm*p-Sw0D+8UJv3kSg~_YK<^_>h&uONr>2?3ZH~Yu1eWi8brK;xj)F>-Mo_KJ
z0wYdha=m-hphl0~DAieaPEJ5R!%sK9QX-*H%TlVD&!s;K0^!~RB>7dz68`Wed(Dc9
z=oYc7R*hrvXGX_0&L{9maz2on(v*WHgxTrsvP{PH6VcWe-ne4!kSZg;e~y)Pc5`}a
z$or_U>IyY(y&8NJfRq25KX6^@8;V~Oil<m{3EvRp8T7F3w#^Nr;`t2O08m7<%NG0V
zS{=w^_UCX6;LTjLl<=Pgvm1k7&5x@{IIb&|8_76T!d7uI3_M{v-&uMSw%e02w`X_#
zHlW4a`2*30Fpg64!_XR2*O#6(@0<r5!d<z>8yTW%1FWwlBzV=56yRro#8PjcC4OMN
zo7cJ}=mFDFRjotDXmQjorhT4QQ&aY_LJts7+O&qSklPZahUw<74MH}FG1(a@Po(mW
z8OSC@eX2o{G4+8QIM;&TArj?LJHr(%vZit`#6{Y<o1JoaHFc9*>HsU(J?RdhIbG*F
zv0puAL%eWUcFkJnNJ<TU@p|cz*(?V=9ZTUz<h@;eSX#4?^_f#UEm>QsoVU}LMF#a|
zBiTMA7+ru}*dSuc;q-K{B5<6@&HHPpDD3K`-AKh-Ho)uugk^_oN;x`YKPCHHZFg+h
zb>(6iLkX|+haRtIosSsE>aldGdM0G`lDiw-p;QN}=&QZDdhV_m&VyPBsU}xs*+Lif
zHCF~dXa4mYLObhZ={{s^XwB)J!}(Hg{4in)o#~fkIYCaJNGVQ^&%Ybw6Nqm<BKsxi
zn!3EQHhe#9HzxTJzj)-`%CKru^S)6TUhP@Y?TEk3(87NRY&a3H^(&m<2>FM>PNDvq
zBs@Bmqwazfqxx=LA6z1xO}9~RtvC8*iMa1bjF^ZRETjukFby4aANGnzx*EZG3s;&c
z4SMr#JU?Gy)i+BZYhm@Wx<sn`k~$#d+ghYo8i2Bs@?Cj&jcKM4iCLHfv;+^JSV5Z)
z6M?pQk8$>oFsS%kccdhvqUJ%Twvz*Qj%Qle37tK<hd{6ZlW)KpFuK)}wNIxGBzN(y
zp13jqVo0b~_nePFXK)E2%2W+lQt=Y~2I<8Z$|lTG;l+z1o5iHw>)QoYq4iyG5z#5A
zA<-%xAz<&{@&V0HTBwz_u8e%-Wq<hm`#kSs8qcYkwmBKlI)#~u|4~f1Y=Zvg-(n&F
zoQ<HE7zV$ojacnJzazbmdVkAmFt=I5>eNxZA5MJtS3L{80M`fh{kD1wA@pHj6y8K|
z^c0ub-JfB7KNOpu4g1=jID=5B@RV?Z7He*MPr>O=F(Sa~IJ?_#0uxTU>CY^9gOC4H
zN;tHl%B&)E6($wN#3p&{H)pyG_xIgs5I6b+jXM9iK#o76Rc*YCDN1)I<C>K`*Aoxx
z8THYPfMM)M3c3fe*_YHL;~kOwfH*aYiASzprGvxd){Ab2)W*+{4(zJ+I$>w45}Z=z
z`chbmaVv@~IthQBZb))9zWSY%)>&O$Y97<4%ne^jQ{-&WTpAd<LPPpvNq1jP@|^Cj
z*A^U1eBS|4+HC&vO+@Z7_+AiAIY*~o3rUFcR6wE~(r5t?P6RY*db{mocU!Huqk$xo
zc{&s^2O?IWp+h0O{xG}m%XmGDCf~mCpM;E)@ro(YZcF&<+&QnO(ozEsm!n0ANN(y8
z(6Mb)d2MiWNOSGRjclwW$Bf)mPPW3^e}dr<K&h7$!ur!c?Q+gqqjq|px7=di*vR4b
zeDvzP?#6sTS8q=3n*B?f9`zZiwUD>SnY?ED%sBhpHy`$|{^@5W6qd)gFPe&sHOPm3
zmM87oa+yp1QOGn#f$Tpuz)&56peWLbOqOpr)E8s~Ugx*HuTvG?uH@YCuYrJt23<z4
zeFgaJG}3aB)B6#+Zr|HJJGmQfuFF>hDq97&+|js;9Lv1^q8(HXy25Ydk7TA1Rsf)+
zCzu^cNHE3#k|1?f=42Vm%n$}lQy3fj=BW{-6w%r8L$2@DbXL}}N{Dw4g#96KA<RbW
zLUYHAUkw|)`CYPHl@;EW`$bJn-~wc0$xhOn0L)%k|Nof1I*O6^G0?NRG3inVG6dsd
zWy@m}LPN-7M>G*c>H8}+P>NL}Q;R%PQ24vw+|~8%(V)4NiwgD69je09#Z4@=*w)pz
z`@#FaOBQ!1QM-Zq=oY9tj?b}ftRxVeW<F<+){lIReC<$u=*M5Ezf0988`zV`ek~~Y
z@jB^wKPh+_v$OXh5XQOCf?_^W=x;KnwOqHa=`OnoOMXPK0BX<)vwaW{y8v~G0o|L0
z9z=Vp6<bQPZJ!{AyO^iTA30Sp`;n$!zgds-rJkMq1A)M!HL8;zDUF)+l+C1?RUS@=
zJ)m3S)Zp42_SjGAgk~#jmJ!TPHe{O5PGT8e?hJu@(Eb;j@5!3CIWi_yVV>@A<hD?)
zPQ8?R4=`lic3C>6CHLvBi%Yyq`xDhHucWGD25f#Z5vUhFn^ucvK$XV}=-*pN#_K06
z<M1#fj}d0_`6JBe*FO-t-9ruE96_>tsPcizQv4@!?Q7-P`reF7DM=!#S<imjh-h`W
ze9&wO$X_}HK$BsM^hPU^At?#*htUsXO1BuC)+^h9YR%lSA}*JbT;1!_0pNjun(Gk?
zwSF=}*4)mpb)-+6xb5ZrSxXDorTDKA&)@z#;(21eJpY!<d*>}7?Zg_tQ-1u<-$WTa
z^^s-cx4f4NS4FZg#d7L|qd}j~38*v{2tDiZJbO*?{U_tEFDWWLHr}t2z%D(Ivul%*
z&#VPZbqD>*ceKuHf8lGDlqX2%UTfrW7mpX2w*wC*BtOr(Ien8j?Vvtr6(2H$S~XfG
zEUdyo*Hae^Nmd@FF(2s#t^I*4XYQ1emfyU(&)nOmj6N1G6-Q<-PrTxk{{Gw(*%CIs
z;0D~+dD_}0*wRx)N!>-=%qJkttH>7B!f<IEkq>jfL-~4e!6@s4l1r)gh1r>_1}SUk
z*AXd+RqZXlPr)4Yd>|240O}EtC_D%KN+78+u!K3oAwoB!W~g`MSu=){>Ogvno%BaZ
zNu>ygA1OKjsbEtyiH7j;NAZG@3~M;`b&Y2Dw%Q5{=EoIq6?4#j4C#FhU@Y}<0vMvM
zG(GUMbSzVaEH|<D<@XwkefP&D1`;q%QAfHn)O^Ift`>6d8+eF3N2lRl)+t~*KJ5Ck
zaG-Dkp`}pfKIHDLQ1bhGE4EN^_6pQscXDInO7o_}74q2DlLtkvubMhS_dIX?HH;4)
zN&sW?SmFRu_oKh4Xk6KXG2?P!hEo>9)N|2HK66)(re~0_4e5JY(vd;fR`@=;UMV%`
z?N9&eulK8^UjF)W%hg@T{ygMbo|fJ7D??u7b6X8mj=C5IU{R-I>0otz3<Sq*%so~c
z3oy)n!Dla=)#YgPmFjEb5jh_S_q3$)V2G$%!W1%&DK$bBKL4WC5<4<-;nn4!y+QX@
zS3`4S`IhZ+6%w+WqT0|_^-*<Uj6BSA8eJ58GLqSY&)u7eNklrW*Oj|d4e<V(VBh^c
zV^p(ohi2$@>F3R9Up<8d+TKd#Mb{O@UpI=1WpCV03DiE(F8c$7Saj}&SYSG|$oY(L
zSLFJ+9+;`Ea*J}D7_7<0=BN+F?_@<nqll8MRZ94jH~MQXOUgv8-?df0rM5qi18`2v
zkOFR25aM?VgO?90KK?Rt(TC=f(wK3lJl^qUmYeIRw6LL23%nn4A41XJoAj#~d#Ysf
zOFn@;H?C9N?+9!{Cqv&}e<3H(2iNs}M`5$>55$IuqOoIVp0LdOO5=sYidS+gQoHT&
zrnU!ig8a*kv6HT!_Ju9-_)hFcC{f)>Z9lARge!!FLu)*4v=)<Kl<cp$+>{~2nCJFL
z;83r0c``joN*vrOAja!W1n>FVkO3+7+5x4dJLN-)saBkB$CV~mr@DO=WF>-Ga6Y)s
zuJ1ef`s}+|eE*9-A_zrtbTr+Lrv_Lci}3p^ZubUsdOb>z;P2Lye!Tv{PM%~OR+e~S
z^kRRF`YoPcrLhN|xu_UPia4gE3`n(QF8{Y2;q`9G;UIU%zo~fEoT>iyrQ#k-ji(oW
zu}vq0^Xk2iiXb;N1acJezR9Edb`I`HJfFE}dK%<GS!k*c|Mxpe5aQ~jsa7q1z4rdm
zXUgiTV|DTA`iATsn(eZ(a?HE0Ko}ARMs>O>mL5r>Dd_yV9It&;F?XZYwAC1d2}mv!
zy$aPr?WgOJ^W$)Rv}Q5MHya}|31%_}%EL`W*#a}2qL>o?#vuK8Gb$o?c-VeXB{s$M
zX8t=-dFe=>vWu`p`e{aPjS7*l&vx$Fslca8-#@b6Ek8SwVC4Jjxno$en={BaMIim1
z+7O4MPnVSqJaBiiJR$zX?vo(fQ}*jb;=XPPt4>sz1wB1^SIXKBb_1ag%x_SpG6pya
zU;oWTpxHD}@YG08CSIJ^-VAdJK(~W5VZ#%WYT;^08M;%@j|J$XWLjg6KyCUOfV*!Z
z2MT_jED66}?46M6p$hwDXATm6ja#~7P+T&U76u_kSN?cKp$$L#x1T58CoVLcHY>PY
z+3GneI*AWqqppzB-%M7OO~9yL<}j=3M~|Fui)LFXwu`7aK-}6OQ$MV}{DFux08-mt
z1nhnLY|^?J5AXl}1Gxi2+wCR)Kzso%HJ#3d+k@H=M&b_99v}|U4<YdHM+w8rZj%W|
zROLIb{u(;unJ`sx#ZI^$B~DlUu500w@2}RBde6Fy?0B-5`)mA{BR9V#c$k0zGoGkp
z=o{=INBx`mjOiD^NxU((P|ipLxop3&8!Mni2T09e46rkhtzy(~;FTPj(lxC59lmBf
zi9&O;dVreCx|W25`a3nM*_x4R#riGMNvF?IpjDyM+^PPv^vUXTN1%;#bD9~|;`9)%
zjIhYCCXbx!sUaU^yK67P^A!%0&A<OUL=Xs~x0-`37tH&WRz9rUjPK$fJC>}X$|6%B
z8*YrU23<MKs;+-Lv+>_)=g4gZ4b<kI7XYZA)t73Ugpc?&Y8&AL^4s#~hYDQ}Broy3
zV)N1_szHaqZ&*O=qu&8cSaZ?o)uwY5wWYP?@!cwR$LN&lqc^XnG~5wE;K?W;o^|++
zSy)CfxQEJ_E&2fE_I*#{MV7poWsWug0;DKp00HmE+Z5^qBdzL^xVUdm?$=LsD6sp=
zt2WQ>i{xUeP%)F&e*H>+#`7ziQ5ND^QGMxIvWzghse}+zwEG|d>E0j~*?!YNpxj_5
zbBplf0|@QgDQKwgod&yic~1|=Y+T4;v|trQC{MjR{J6T->x_(*gR$3~9;_0DVQJk4
zIrw?$(%A%r^g_h7Pvqk<Ba)r2Ig?xmS>71b2iKSD%Ztb-tWOU$$`=L0X@U`;=hq#>
z%ggb^2($G5BFD3s5IMR4)}D~*4z)$Kuc^4adv(n;DZm3HG~f*EOKAvdMSbQ*d7%^0
zTyXuyL7()=ag%yy77=)*Q2V?}j-D1Sf`%;$K136MQ{SV{1qH_;8N8tg!n@-9xV}M`
z_-HQfBaXhv##~2i=NrMa$(3zEb==M>7SLvRh<|qX(iIYC@tq_+dF#xRUWEr=-_(S?
zV{d@a*6hQ`6O%*H=zy2^N8TilJecZ>qO~ul@d&uyg|DV^9uUt*Duwtt0y_HB7Cmhx
zp0K$WFdAp=F0ze8H)FyD=7m);jkVP;9wwmMr_NYSuL^$^f%QYB!fX49(s4DSlSp2B
z8*rFvpFiczQTYiZbC%BXn#i3ZMrWn1P9^N2(-#%y2|pYq>(n(hk+;FqAgr@2siDQG
zK__vtRI3X=r%M_?iX!TXmW=dgq!lRDn)2X5QQ4T#pq5*)fF{%KrKzvmn@TA<L}$;O
z!V_vKK$xYiz<<44al!SaRmRX~sZ#x58zL7k|4i3?bQYFVxu%H=rb<qNHUqZdatraB
z(nE$em$gnj*Fi|3hL)R&Ytv6k+XJiFqF@Y-AJbkmiQ1uqudSa82QVp887P1w?U1YP
zlp>OFbe*LC=^2|K|FSWG1~EUAi;uNBH0Mo`ebw3k_#_injdiVdsq(*}k+ag|?>YNs
znR1yVKyib@ux04w9ASTr(xgRHU~=5-jeTb-;5>`Md@FY}A+1*(PZ1Xu{eoTAn!){t
zW2Sr1K>&g)|8~*m0=58QVlrc2`~>Y6lDZT$5ngwYdTf*=PK`ekS+>IKe6#7$yLdQT
zhV<w9cE#4S(*j%}7S;f$<)TFrfo;okituFZgORah#qiqBMHnbln=Gjo%+SnP#u}(X
zfWm~l)|zRRj8@~Vcv(hp@>QwvK0Xv@kWlE`E&JGHhHz2=1g-TmO_kahk%z&p_UwUu
z1`;^PH9d#(p9agmWb0fjkDG7(#qHH}?3XPjhmmLt_@*`H{)#5Tj-VnER8y0Ms8)nr
z53uDaFAaG-Klw$!d~<WKWVKD?>ZN)K=s;MD&r=e%90`OlgvCQ}X2c=L9Spe4EC1s%
zTi)6OFT7RQ%PPLMclWGcqfTDQhw+5J_C3}a_)y_-pcuUu2U1P$^cdF@cvv}+zvZzz
z+v%Yv38Zs$ZP1On+yo>*;8NbB;xpSG(GQU+ZE<DORXrU^eF~hf*tDQW;xzY$K1wn=
zH5Bv@HLmpU5u|`UV+tuZz<1<@30s{Q%E(Ss-)Iv;hpe93=Hwnk`qF;sWw=RP?pa!{
zGBEUe_6u&~P(ZNuUtP@C0mup!G$nZfF1^|hIt(k6;zAU8FmaM+A5HptziV+>Lxt<-
z!-BqyvC7{`S-w+mT!8H^5X;V#k}Xt@JV`HT?^@dtnyDu6QC*BEQjbs8dYhNm_BSag
zs^)jT2^4HQt?l_ErDw2qKg%qIE=CQIh+80;>v%mjGUQKLyufpjx;2U6@b_yHN@r^K
z^A@`MopSwfS?}xD=U>82_LiLxZAYHc>cCmx(vfZ#(Kz-j2@a4oPQS>(lME7__$O>i
zr~v^&!~3_B#k6sH=eDC|0&mXLg;SF$?oB!DJtwm**ME+;6)I=kk$jKtYr|hBd<%u}
z)&^`xOS-Jpq@fH$`s0k7k{?0BIs39CHWB7UkU(=B*sYGb9Za&5wu#d4D3X<Pl#x%D
zL3n6W>@?aeM5*xe<juHo)xNGO*N~82CpOPt<x6bsTnb@P(Bcb$Rt-h~^cNDQZK!uZ
ztq#;)Ju(2ATQp<C`R3ET<U`$xG7cN<t2y_DpBr&ox9PF=6zol~-lQ_Oe^6^?@F$mQ
z*Wx$#UYo<7BbJRX8ga6PpaZRDTti-h6heJ5cSg`z&_0WU{5rYv$ey(yx3r>2SGgtm
zl~*p7+~xv(>~TRMs3nK5N7v4kO-inB3C~HqO$o^T#U-LCo02p267<kM18M7sRU?Yk
zSv#_JBRcBA!Os<*a+hyC&*L=^m2<T62FnPDF8?=wSk_Q&E6H2-lHc>m6X9t5fYPYz
z7`-H#U&Ah?(8HpT&-kWV`}Vv=d>!(xJX2uI)^RqrU9fe+t~)r$(=R~5+}l&uOfkXB
z?#Fc6vZ*>d1Sn(4<uw>!zw<2tW&n<4s*ODotgp|Y16Lnm3nZkY;8-7_{DFW0Wo=A0
zF(k>=A6PM?mH->a%Hs)C*Y03Xa{=Ydx`vB1mG2(dXz^2k3lYaB(>c|fFX>AX=t?-S
zR(r-iV2B2JOrasLCC|AHBFQpC{Ad4PT*0_C`ZhJf#an_xExqDo`1nxT*DD*AY|_nE
z-?61Dtxu@Y5}6k&Bu<UXhH+O$nj}r0(EnK_&#kIjHT!N#2CZzjQg4<~6mo)kX=(c2
zFwRods`}%xra1#yN-n(D`Un8tIDY@VjdgMZ?3<r+owv5a*g_tXqg?YeMd{|s(P$nL
zhR0rSJnW_OPH=yIdS3&lrNWTt`d_ft#FPO-2WfJ2ocl?Yx2XAjH?YT@Zy$@3Th8;7
zm+)TL{*H?TJ5`yq6!onVhm!V|MQyfd(bs0}&TD3>zXB!STj+W&B+#=dBvHf(>POY5
z-)Jc*NxWJhD9rUnofh@M`_%om=`s(c?>=#4%E4Ph_0lH0K^Bj1ohG65J6)uD`Q^Qk
zZ+fIy;o}+X2)7I48JknG9C!o=+LRjn!5_bl<MF?S)_ZT)e8#Qh=vrg=;)`c*S<`1D
zl5L@D(yeD7QcEVu^PL#H8!u-Y$7ut4BA{_UsnI=q_7vQV^|}rQN|&~)Vgz7Q`qKAY
zFz_0A+&mmpL=8F*23<lePO@Z&iiqk_3q!=nPEmB_ll}fdXFm47%{XncNshdG(K|d?
zA>^NM)ai}nXaw&d{wSWus>#~!`(m@`;dIE2E|pZwo*wehBQ%oatwr8UzLPIXnR($K
zF5GsW=~D&5!}F%F@u)qhmTK$>oi($z@#|%?nJ9I&kHm7go-sHa*QJNo+Cbv);g|27
zbNI+V$c*{s<Bl<Ba?M2!n7P?&D(1mxKmFQT>W*0gCZv;CxVOrH?Pf&HP<cqpu?cdS
z`rSRgJW~2L60CQlq3cL63emg}eTsY+_pTdLKy^A-Rg+y|rZ!sOgvcKl#73c&Ui#Y~
zog#0hIpxbOid}LSF`s_er9G870%YqZ6pdE+qY8T&LN-&n?R{xy^TbJ&f=c8>lb701
zmh?r10fJMvBo!pkk7W8q&=J+_U$0kYj@`a^_`-g~FCsDPCh&52lCxrZKg$*>6f`Mt
z^94nq&N7?h+S;UKZsy<X50#f-y<>}!1w;S(s-v5v+2%F+$L*;O=Q3w}mF1G%R2Ku}
zXoKaEq$OD6DPe_IV3y#Wg2T>{H8zHP+F4%yf)!Uc-YDgv-p#vCOH+S?O#c<k{cc`E
zKV9Dup?uFhT34S19^QaOHB3A8482CnP;sx(<`o0HFI>Hyd>p^bG$7!e!y^sRFggCD
z8mrmZ%dex@_#0sOU_%;jUNE7hyKHf_C~DC&s{_QCZ*(_yvyRz*@$@8_OU^v~n$$Y%
z_4R@&{o6w+DKayK^ktt@?(f0s;DXyT3jU3(ZB=botRG&433`(Y>z}E0oE+cRCw4=m
zRAYF+F)h`<=!K0Etez=W6Qp3~s%%`EWL7meu6XQy@oeyW)eFKu$?R@cDil*OiztM$
z$v|m&vcKR}$FykzYpCvB%t%F3Z#zj>@4Yt1f$JS2KlMCw6b3@}qtEp+<?jvoX*&uF
zt&UX2&5V$|tbBbRH$|~Qfksr@X)QOL3l^@h;Qk}v`n_?*+xgBpKcPund$ySlYaqTa
zjbt2}W8_QDNxHsjF`QoQ3O(Dei_BMwioWH3%kjf?L$@$1x#=w4jW?$HdoTSo5d11&
z=qc909e1ra$`JZv;K@Z8?*vy3>qJnaH6Y<CJnazKubcz#CRLk9M&j!--?b~Ho{$~!
zPN{ealSf+jz6#Y$re0;Tf-Tz21Z>ehg>{ENlUG9Pq$ezboOmjtKivE=__3i%)OLyg
z5ftWqLV~j9_*m8DJx+n~?$R_>bn;r_jpQ92zC7K%F|o){wjS6cOydu#G3YLrOVhRl
zgO$nA%_Fe58u*<6l9|QF-#s%Ei?{W#PCk$)QxRKX?F(4ju<eF8n%}0n5$aKC`eo?6
zx1gMlxU8sjj4U}9?pus$1M`aL`$#ThewJ%I_>WLWvvgmF$V(e?UjD`5z!zR@g$d8B
zYq{GzI8AhdezU&^1#+1vT#q+ktH<+!SUC@8o;JA^okqG%U2yIA=vm@QFP#znWc6j8
zR!<*O>Bfin5A{}D(qzOhaT!fr{;9b%TY)?WSwfONJOt8-4J3^r8H9@32!>Bugt@Nt
z4ZE!eUpNJK_YpE^61;OWSk$Du{A|emAt1e&{DkHq&-onl_V6&$)DYm`xyS*S@na|m
zrYA2CLsT7da%HnD@3k2<N-+>%OF9WTc0>fi(!+8Ya6cXdG4v(U${v3y7o5-!jI;BW
zK+&h+=S|zI5+CjTB@`sEAOy(A9!7JP)o83e`94+ndQ1s4cU69xDR&X}9Md@X&wW>>
z%^ch5Z`5vmfulN^e?N&h`Q^DF5v6?h`|k?(2M!{?+XT0xQLj*?^PkY7<lLKA8EKlb
zk~b0|fr6e$L8h!ThNtX^I*+C3vRvh}&hT;mFFE;X^$_oGzTemUmMZg^{BYhAFxKtW
zV5FZugz@1d4^bmpU0<zZ&0(_cQLnm=or`hF_mf*rrL|$l58S={CS^c1OUjsoaJ&#i
z6dZd>wIQ+JG~_wh!)L^a02HT{J&Ll{YQK++i&i93l2X;5)IE6-1Qog2qEfY`vA4ow
zCGB&PTJpu+iQiWpU=b1u4|x3dI)s}1v%h+8qJ%>D&jz{6(-ro|JI0OQOgcOqF(gU(
zZQo15O~S*=Fr4=)5VC#IF^U^#+v}qzb+y^A%%G=!)%y^M^fB!zVkQUy%Z|EA^(H5y
z-RkV0w{KrQRgvb>u&yhmmk7DhU#CPr7=01{j4AlaX7Q;_*Y#HpQk>hND@m2O0(CuU
z14??L3!^B1<bGyQWM!!B*-EdP6Kswi7;Ab5uAjhBV8MCsetqFy5f<1x1hDQY;Tl(<
z)<2s(-AU$BS}#52+?1XjtDlzc`c$V(x9LNO7Cj1#Ec&)IHZx-))3{{tJzh$ic)M1r
z&WKTQi9k9)ZXzzy6!SEvbXp)jJD4RaQq{`MxRo;lMXqCwHMH-|H#Z*eUeK#5rcKVa
z>Isv|GJ|V>6@6zc&2JZSi`bQTX{5Vp$wGo8rU<i>1$Xy^g4Ab;G}P(;gT1#5ifilE
zMjO`zcWE?efZ*Opun<CUhad^=u8jl;1W$lKf<u7dE<uA!@Ss5(*KV|dhJ9D|e$RK#
zsav=1pYNPs?+=QiYE?6P_3Amtc*Z05i<~|EjC59oT+^Zv34&}+i8CQyXuHO?M|$g@
z;}up2mZ`LB8<T$wxsBH1MqkBNu@SjA?q;H@p4zkN#;Z%wbC(RUT~gfS{f}a`e{f)M
zYDTd_woOS!|2Bs3aN6{}EYMK}L$g33;eJAERpCKWxl8k=c9K4&9<uh8C(daG$#67Q
zAD=3E&b44Q4&Qr<oUqPaO#qX+Cjh2s8;%v^^=JS%@S_ecc87Nll$YH>7AOrcy!?-)
zlmavwsB0~NhM3IEoBV&_I{&7D&WH|*RwVSBL#t@|>jDQ!jrw$Z_Alrnsan2*l9gzK
z^o+UpEbTysXdL+TnKV#6zj|MFM?Qmg-DK9;wR+970?bsa<1EIS9gahQEojIbo__b%
z>K@QGiv26*44blH?S$~4DaUcPO{UO+Dp0|HPTzg_Rz<!gOnh^?9)b^qpPvoMM&{WE
zI8E^lq)bhT8>LHhb7Lzzx<tr<1%r%{-D`AUdI#E&yfrC)60-%dg|F=`u}tb!gb@Lx
z3fV#fOt3hzDmd##Rk-&F<xu`^2vGxktiYG#iA<Z;2J98DRxA@!BvzE)+UY~I>-R8W
z4<&?c-vPJalV5z)*wo^Dst^TS*7-TjTOK909C|Ox1hLogd}hPp3jE1TL+3P1!K%4L
zL5T4q{sKSSzJRnC=X|sMM?dtN(%c1No_}-ON7GitB*{jK!!Qp~Y{-j3lp=!14!U4q
z{QBVO>n47tRQ9D2J~=v7`UC+te;g1BitylKL6*TNQfa6UsGKke>1jSioHK>X#QSvC
z+PM4$q7~kL`i5l5RESup-rvUNhA;_2A_w;SSD%J<8XbQxjFOP$o$2j-Snge_ZB4iO
zTamPFOL?kSkSsw!&@)uhi!tI>7Nr7Dk3G52;OQ}^YJx9sJmt%#^Go90#ntZ3PtdK6
z1E09Bf~ow97{O%wlY`|6O`h2L)J!MRB<9Pvlq$@~jPhqBagGIpMTo9Pc5ZkV;*ey(
zi)jwxP(eCi(AF@Gg7^nJ5`Gmwe!jsk{7?*wTNQJMW`}wZC+s(h2KlQfNTaiJI=8xc
z+MCX>V}dyu)y->!Zzuplr};<>ZF>R~jpoJVN&8}2&*SM!dCCgnD45?dCG`LCgN;%|
z@Ec@Ud*a(ajAznmFe-i<s#%IZ&p0p@DjKAN_`cpS_l|!X=-XT^%c4GR+}l|iv;Ynt
zQE}P9*B9Lz?JAc+$oK0Iig=S3@VRIf_*P#BpG{*F8_8%iL8k{NU04@W!dq);*!uf-
z*)*H6Hg=owu|Qmgwq&>z{oKTAjV{n3;^n+6#O)WE_NlWDef}Q!|J|r~kWXgfkwt+#
z)SV(dnxC&iTcKvm1@ny(PSm0<OvcT>N&zX4NLNE9W0!3$nA$f<dc@eoDxYCIRaqs(
zvBG6#{50+w${xRw3QMJ7Q_xF|!cU$dnFk=Zni*}G@%Ri@ADypnG0~MbOfU}QC>yHc
za~@QRI1WYkJdnH%s$8|tq^8*3L>&^^ga=&6IL?J&pp@V_F~W~sZ*df&dt$!k)SH<o
zlD_5Cy@xw#eK;jn7V5gFddsF9S1Xz7EzS*`g{kM7haWJ<^MjP_Hj*RuMre}?kDY5#
z1*^%xK7VD=knk~GhrKECggwmk-DY5e!a+73-3vU1+K`5QD0T?9{VKWek%A+SM7hf$
zn^N|Q%4(#WkH9OC#9|MF@~XQ<_H+CcYGM^k#$Oi}heCAk+U%K?Nowj|&aWl!Y+~#S
zLY~NhbtdGP!UZ1yeK?m`4;1q!&j|lpFP!9|H*cml*gnkUq&3w$v%Z+*)gMu*X{fH#
zHhR?<V=bpP#H@?<(+nV>j3cnC>~rmWZyNJQ8+~5ZpQf4(IK+K<xLwZ^i~L150?+A-
zAr<;=KfSJNlaq9rOGFK6_?p!m2`^fpfr)(+2MiP(7dl8$iq!yvr1+QDt$hVslr`q%
zX;vRKNSUqx9arK8P*Jhs;8X->+>#em^&GukFyz@+v1g5VH{mV^R-9FNk5aT=hdgkg
z-7dGG)YrOB!A<vUV#3neR|DCEjk7&;G;dqIjEmnw@@z!L)66T<5f^Zv>$2T}Fz?CH
zww!N`4AFqfF>CxDy9!VfnCBD*nZjpV;OSbxFjMH!XFFPgB?;D*5&mIJGl8YH-tOB@
z@iBpU)Q$AqL^ghXGjtv{K=<lnhKO#ez}sV`K(kR9xy1z;?CD)Q;C11x9l-c=l4S~V
z@*9kDXy569kfZoIu?rC)S=Q_c$;BKgH$hrh0>k(vF$$<=-n}S#OJF-D8sEQ>`q4NI
zo}cbhvP?Re(O3GBlks%++HxOzcIZ;}Bib*qBirJ{>r}o@xCj(rNw#$G_<3p}#XMm=
zOI)v~7KJb`N&D&R3lE=w*O^jeT|P<3^Z{0-ldVBy;Q3+HwYFy|9?r18g@oYhi?#_&
zi$4p6PO?rxPnJ7@eXr^(F!YcGBUXDx=qex@F~W`|&mUOw@@(~!r`pBB>|e?=2q*z&
z3@gyukFMZuXVy_=v#GI>W^1P5dDg{HDZqK;&pDI_=~+ec0mo@f<=4d(IupB!qkPp@
zF$ya54itZ^6&{8c{VQL5y+a7uqzLy7tK<hVzWGXRrOTYs_^G{{leWUMx<Wz9PC3Un
zYX+X?Shw~ntSIj343Z2T!<a8&Nd-fn{<SeP0_P$@!FtmjDFf^M9RNKczWphTXLA6U
zZ?Cl@j>hY@nNn5ngQe15eU;Ud1aVkgEUIUyY7!F4i9pf8GY88nE(P3raIj)<PIHOm
zz8Uv_C_eBZvMx7VvM^GWx8t)SWhM3yEKa~JtAB~ZlJlUAjIIn2wG)+b4YF#!BZi)2
z9%$*2bA!HNgBU=6-r6ZJ4ak;+&qYW}*i*#{`!$R=4L_v~swrbV?6zeO(`xd#P3d76
zyuiP9yxfXabX^;I>tdzSrtJ1`qE|tVW9G<p5a=ne68p!al2(?0i6wrqMg&mwI%&z7
zOVb}~Z6&B^4c^4wd~=i`>ZipjTUju6Xw{oi@x(WiW1ulo<Z*WI?miZ>`1I?>6233N
zj0{|V_f<JTcj`!QvL8PgKOFba4xPX?S{A>tId)^|M*yKtZ#OKNmqlM0hbf#Y{DHUl
z07fWRp!Q2Nqt@eyLdVO=rGr+pv?5H`Z8l#w-ftCgHrO~{$c%Mg;ykeh3asu~F}GzS
zq=7<HOJu|VuRaM6>%+IiE!fTci$ojXNNG0V3aQ=>y69eCq2uO+MZcs%PFb_tEY%@t
z5RRN@g3w_q;_9e@E|D4;pioXyPNQfY{xq5u@($GpGKR1A1D)RN&E4DL^~#*W$PDQ8
zS~z7O^s>vahd$7AFFW<3FT|jS{KIInLLysbw-&KY4F?GA_uPKu|FJ0g538d8+rLBf
zfb>Dc1rZQMw$J@vE7W3Lu`94mbxX@4-B5gCxA+*Eh{#I>73%#>&RqcTfdZlo)HQ;t
zZ=BxIxUX*yAzNiv`(y#DXsMgtF`Jy5So}gd`;WsJ0o?%jT>MW29`*;|$@U(aF2Mk#
z50UR-i8u5?KnEV71~|gELV&Ipb{`EU0UQaYp!a+&#+-j&7XSBU>i@dz+lef0kRc*C
zr#JrL$jdnq2K^qLwy0q{{$ul3m<hy!>1?G~b^Stdw1_WrI`9I(wm$-TC}aybT1<sP
zo;fBXyMcjQ0WfgGZ3T8W4Kr^K2sLE$qu;=NH`U)}>t}Di110x=2D!Wa`u(?CYx_F?
ze|2lcLKfXB8zDLR#f-6XqPKpr)#GzyRH8y_NwU6)ippVn4zMBmp96YgyZ8JIejTel
zSH^^`rdQ8aZIjIW_s4Q;v_<(=KQjIu*)f@{PSl@6mSOF+D%pChEyfk^rcwf^?}W8S
z5Z`;0IhT7~g<AAD9Ip8UgRxb2RzhXpB0t=!8Tu*qUA0mqMsLsjOg=5ns_Yn(Ufs~6
zqhoR(r{9SKX{^4@sb^12Y%`&XKK7FEDl2>I)9CKGQy%xt=p*WV)0VdU&R205gb(Ex
zr3%2rHE-f!)<MS@rI=-}{?w90c@<R)4m7Wku=Sl-F|QaGb7eg?mb+{vprCzCbHU`B
zOVo6c1H_S;ey1{?i#`ul-sO}RS^lUauJz@?C_n`rXjjWa+R9Hb8cft{@IUttB^Ajk
zbW6v`m697bzo=3ISvuSk!N(f>TrLtH?p|HJR>-I<&rcl$h@#r=`ojDV?Lxr7P_o*~
zS?u}Q<#nf3gKE&vmj0t$TN?54(hB2$2-Z79)FCrzk{mG&4E!(Pqpfc#J%z#cfvara
z`5(MJUBX}r8)wayG%$giz*2L1+g;Aof@#gTZ>U2isc%)XYnhnDqzvq_(&OlN7CU$m
zC_#Bx74d+yEBX5p%~Wr0l@scO2Y@B(Yw`+&{;^*CsM^<OG(S?v(-T423*D#^jIW5*
z0Zoji_YDH8g!NhO1?j_*@%uk&$8@oV9uUwIzOvZ?dF{%OUHT%k)~l#i9Ws@p*Lfg>
z<|@>!c22*fTPBE$VxDS#qSOkPWF#_1_VgvI;lzr(`>K-gv@OGk!KXHBC!FZn<;nb4
zF3sKwtJm?PTm^%1>o0y;3$vjItLhI-0Q1mU+MIA{u5YXAfu%@m#kRJO5o}kqLE5UD
zgoJfm9HBVt#<Y(j3+)8dup^5M?W+fuR-~l7B8_U|l5%Ze?RJxLU}nDp*gzPc_wAk)
z(xTu17;jlDm-Dc9(YrGJ=(YX5Lp>&G@a=vaOp20#QOulI5cN>5&@?we92`xQG`aPp
z3`U2!Q43v6RT|A4NwT!x!%fs<oR3*jXf`}jR9I4NEdOYq0JxOomQ;}k2qY$Z_xnet
zn{pD@V)<)rG?|!x>T=n<4mkL&>&c_<9q`5oPj4Zr6#IIl*>a*spW35(;h!8T@UbjO
zds^qTj5h?DY(74Fq$xPOdZ3{$`inLFBAwaP9F@+e+M^SyAPlywuc5NabN7-+N|;f5
zMhLlGYk1q3((-sp?fCer+bbusnbVMlLok{wyPsq4J>zUb^7<0=0bg3kq`SStH*3ye
z%}@9&TM7M`)(}pBDv?sq$2!1WEBQsRx*O9Q(_j)>t5h-CO4a)&3h#?raiOXn-tSxX
zYn9SX_qj}vOBG_StBJN@vtXSs#*Zyy^+*gIjd#JyxA~NgFd*mCX+U<vF}Q9u@f=KN
zXABTECYa#2_NP@@)-K;z&^4CKo_z~{(LYW@T@^-6yeYsG5&j_9<n}=icQ;HTQ;hh#
zQU0Lr(HnFtIoVvn&A>!J!kYFUQj<EAQrY4zEyaCBNjl}9()Q5oX9K{&@ieCB4|)@8
zbWe_~<ndcEGST?L&638kDFezS`UHH-*Q|MRkUAlZK`Fju*qT5qq*7H7f`yz`LR5vP
z!Fq0<2k?r15NXQ{dX_J4auW1ePaac128Zp=@vJKx5A#**tq{<2sosMf1*Kg|9~C}(
zqK$z@ixIJ!&55xaphjY&tVS*<(2Z$B3vBEk0N;F)ag^3Q-S}Mu698-g23)`DT|r3F
zEA5tg)%svr;?CENkqSw@Ed3<{Al|M3&vA;GjkfQXU&0*|!&V-t+ZCq%kY_fWsljg|
zsT0L{m75X>()c@al>qR5Iadw>VC5tr2(u#tslL%pH_(CB&^{%+Ra>Jc>an-|WUFg9
zm%1OM|C}TaElF8U0^=;DWBo3HIgS6VwscJ^+v3M@I>96cpZ*p=qS7Oqn=pg62Z2`x
z0jHCKZ!`pMgad#qT0shP!mem2nU41M>4Fbx4WTAQZ?!n%dxG9m{&vRqz<G`~e_q+q
z_WrUw)+hwu5uJK@cm$*I7Y9DO>4!rZJB%Nl>wZYO`*-8fkW=6VS?6X}38Xc@B0IZU
z%`5U3gJXZfP9rNPFtKEHXxGr{NkZl5t;41|MGj6}7oGtC08b&Fwbx_e<L@8iN|UQq
zY5>j`K+^v^KdX`mW)bHF#qy&J1j0q_W6DfSpEZ_Rvy`zNcU9~E#>K!LU}MbUX{*A#
zt7lHJh?ERep=<3c`*BL4t-$|5aPx++n&yIaHdMOco+>&Fx*Lb|Jc|Idy$8Dxnjovm
zHVOEM1|qT{>*pqIT5nwgr1gz;ie3V)DVB?}KfSF))%dp?m5rZ8!yzMJA_9}q4&B}H
zkE~X;>h(Et9#0OE<>(?Vn>Bi530_I%9BJN9o7cdAnC1YLNLjGGELT%qbElahb8REX
zFdy!3-^Pc(`2y*`Jp7V4UT2|#@@Qy&f9Z|1vI*BcEs%e-$!aM8C;DoK5dc|SVdajI
zraoX)-ZpzqisOOvQRq7M!Zij$eV&Nz2UbCl>!;HuUXI#7j%=3Izu+!cufcgk5pAW`
zb-{Xe+>vmXcf@CjWQ7-uHDx||kg3GC$vWbR>8F1V;c<hu26JNMMPD|({YjnEox-LR
z>Z9>o2y2`TgLvo?sASKfajWEQPu5S+cbje!VzZeW=JYt~YMSG4HxqGhR|Bd|aZCxx
zQe*~^k2>Dp6&|`WAvNId#vR?M8>cbJPT~@|6|p)<>Lxo>*3M3I6y@-WQND0q+SKw@
zc!`&~C4C9ta1b#l*cblroKaybTI=kgiq>pL$5hqPg_+>^%UX}iCbvR(4JM`X%V9Tv
zdJ>DYi<{T*V;bi+wDnJXtbWDspEs(pCO@kch>`z$F9vxj{Mj<vzQbJm3HT%Til5#%
zWoF&&y)ndGc~<WD>-~|7ha3iDF)RH|d{F$Ys}d6eC+gAZ>ab`XQxHST?}tJM;hY$*
z&+p0)zs@*s%mi^guK3(+WSj68=%?M>cg)jt9~nU)$UHBE9clYRz~0!WW&N`&nWp!B
zDnrFoaVjZDS1n!nI%W{Wud%=y8bTGtk*OO?ZqM$6|1N?(W(_3u2kxE<e~Pe;dhM5G
z<t&`*QHVamJcu*nLc<7~4dwN-fOUo*1<6|qq4*}x-x?`<6UY;&yDkMpMyd_e;LsdQ
zb#zvZ(D|{qHpi|ka%B1&s?fWazVYDzfh5?nf8AvLFW|r4kBd>kihpbAVv7LBjKmV?
zs>O~Rxby;O?yjjWAU7B3098vQ&*&cDd?!P707|+M^c+JDoSwBS?3+7*{s*%&F!)o;
zrvIN>w&mIWJvnqp8}51+<pgjL0$$?mY5ez6wyd&J1%`u4Nu&leVk`#$3r++S;xy+8
z7rTFU^|tenmFldL_P^J<ppbZg6NUDn%+FC_)`c?IT4|bpM-R8w%y@tYp!8yei4RE!
zPVF($O`IELJJTsodtGPwNG&c7kO}UHv&#+sbO0&|L6ik@dEJ?3z@XJ^CzMxO(Y-j#
znSJX(xtyBx8Q<=gBbtNcA!%PpFh+==pKWi5um*u9Tq!;3gx}hg+fgun`=j=jiIbh&
zj}*zdu^h&V?FblAY?@!;yLOmr-c03|ICE4W$)HHnEw9N>{sDfXAwq3ww~En}+ENyg
zTj^K=#vV`3p<}rdIdCv==g`iY5~23rH3^a5btavHg!uB)$VBSXNzJcmlJN}Cf*RL|
zl|9FcT!I*pceV(kz6W`xcG`Oet|TeTk>iV`8o6mWn`XuI2`xLtG}k;ZthU)sindqj
z0z(9i72Zrn<fYtGbv&;wx#e-1ndTgC*}H{1Y0BARxfsQri~VVIlS1=e8eY;cdH#+C
zU4rTEr!PAg5c<Uh3w~DW=haz9nZ6yTzLE?Tk*7;>;X>Jt`A8AO_VflM=F#MviOs3X
zw~y=$&HN9TEZXqqzpD-on4V43qaqMic+MRh$Xbbw6ps5}e+WP?W>aIo02)@65|(zd
zOGvoAEMIv|A8Q}Q$9@$(b{%mHjz$f@+r1R2(i2btTiyo@=uTW#>9(6|6I7i3DH5@a
zC+i=xDjFFmO_d98eT-nLZ&FoEo|cju6LeGtg!4lEtKcAJ0Ma!7TD<->qrBeZTKOK4
zZ}es!eKUgKi+fm_G}?6TIsd}J#@8^IhNkx9c_Q8-(qUgbWY)T#b8<8+6^~s@_LW&c
zME8LO4YC!lB1o^S+mk$LG~J_YfF<>{)Ho(%Dv%lWluQh$bl0ttmcAw->cd|AdXq`|
z4Tv((X9@&A1Ks}FwTO~~)NbXGvM`gjG)R=>n^HPeg_Y%cHQqHhny|=sd&}_wfY~NO
zQyx(|bKcI;BY<?-m=7+kIBFEuW?n2hEy)pe!{wQ#&yH9x8}UCp-fnD*G6{TVI>07a
zY7igRt&fu=55w+=yo+_-aivDau3IK8Res*8&wDZPlO?I}d<ef6uwam5P%gMNHtvJE
zGv=(*@HvwV_X%f4f1Kj;v&z<5%6Xo74vny&b?RVKD#H^l^(w27kiJSBJNuP?4|gsb
zlZP>Wxd|fkBy2nzdeG}iK3Ss@KU2no5soW^3%Cs=&hO7|m{ifK)04J27vn~JBPB?2
zYq_}0#q&;F#OXfUZ~6fJ@okDfV)W-_S@b(KF-UM&SoJ{Z=AA|$|KSya^(kU3=bQ0v
z2VF!@t>Iu5dqv|k84mSB{QRH+dra%<o$0k9SPD(yAK+S<d$0#fRU81$J96)$b3d6P
zTi5!S!U~$?A5GU+3z7;=i!;G~1JnyKSps;&uhAwdS=z}*%^9O9>g;cykM(ZWuzRRU
zZ#M|H3-+irP`2u&R-{(5ztx}Qwf``)E=N;lUG(4hVQ6DwVbUz4>V78jY^Q3|>pOpe
z7%JLjhv#0@CroQLhR@DQzm3BE!Lqrzu7Z)GVF=(+eTd>1UKA@Jb2OisHK?ym3@lgV
zTVR%B+)UzeXxq7r33tp1zvoO?;bKpcOVAhL^(ikIb&~?&aa04t{sizMj6s_0FHoul
zP-_(D2iZ-EK;Igt#~<_ZSvJ;KyL#L9MvqB><mS=B&d9dA4hGley37FE`yi697HYQ)
ze)H_9BQy={E$>O5oUy1htg;}6tJ1D5yXF^m?a(1ei)><m1uv5eU7d$sGq%t50G!lw
z30Xo-m{c&gqtpWtrMf+Y^_2(uyP7=EgBCB)+AGEgptdjDJt45aioc~r7CtWO7F>;i
zp^IL_j4H0gVX1Urj@(5cea;ymyP8600PG~y6<B<stbpCzD%Nq3t7p;ZY}j)$`(s6y
zk4aH7^WEBp4|=_1qzt-J&m-gGC+^}zz~}b&f$Ng@hiBATg0Q=6A{lXEG@W~@leduF
zBs<8}DfnImh$^P?`_|JRf-~(M|7c>tcrc3DG3?zlh9Ybbma+gYMnX?fh@-&@5Dxo=
z%mhU~*^5vk@wZJ@EHxXpv7)vuk*!BNsQ;H^Ck56;R>NIkW791B_tG=bA;Lih`Tz=P
z+8p7RIDo1c_mAc0uFl`{%)CGIOdMEZf^Zc&QegQ&QsLl<*=XbB{7}<KYRA<WFy+DG
z$lJs}m%Ki^oCyc~qV@z;$P<zs9L9zV-(rh_J}jkRf&9M#{z-nXpzAN~<RZSrSOUY*
zp@5rC)z+yX*L4D(x%z~YXES+fy;N>xQ8<JcT!U=H5E6?Fy&DMJ7}!P0k@|GN_(n(b
zy@zU%q5E%I&}3p)2U+|t7`*>&Wg*4u={!uZ*6HkQLGi-AkG9die%e)<2W&Ka;|Yr_
zE-QCe#2^m540v-AL}4f{%u-%P>WO?HUs%u{fnMl=%j_pIfC9!y`A7n}+V{p2(`LcX
zlt~On$)ElZEXo++L5m+r{CKuE+y0`%?5_E!IFKL5v{102UfWFh^esEdr>A;qyLh<0
z4{tT|avyqX-sMU&W8KxS!oLYSy7K9{O5!mK0&aL)jB?k}T7x^y2j<HN6PW1dHOK=y
z$#Th{#m@)!JI0FJ{;b3JaYmmVtz*dId@rlQo_q+K?TEiCIplwQ&k)3YcI3ixE1Er_
z*&*1W8^z<qY7<m(Du$Eu@rD+(>)h#RXS~hc0{0w1b7S+nczjrvt*S#jgwd!RtQ6qd
zcM^9ACh`g<J$)=_rtNG*IH9zC4N^u=8C^2L**1T2V1b^W{^}c^l^bYZBbesaXkl4#
zy(;QRrXEpl9N5F<H@NU<utWLWg$O5`s0iexZ_!XsYA%<a{V;{@jW_8CW6`!CpFeX@
zUUS$It89Kpa$EHjO698w;Nop)yt>e_082Q=N|k4gTI>pS)0Exr6kv(pPGyau+4P{$
zQ&uON_=BQ8dYi}?Q(PE3c;HG~3$*W?gDjy!N74-IpQSz>JE0^B<lV_!5)8|&^cDAU
zO(TqHJ3bUUijKIJ%^O=RmAR6bDb$uYkt9)*d?6J&b~qwOi?jwt(tB50W`=&g%8jk>
zjuY&htW*^Dsri&#CjZk`{{J>s`ya3O|BczsLIK+xiy0u&b=KkzG#jb+t77HgS$ora
zMw^K?$LtJlKol_Ia!%V)Zl9hhTkLLGYBJ*N>#{8YbOuTC1B&P0-h`6=X$x9ABf+P-
zZFGOc=a4(b)M0}_xY*Oq*?S&8Gp_^RkH-7NW@Bp%H3M|fRs&7=a^pH-a^U=n!M~lh
zv|!*{Ou%XDYh~;Id+=XpHva!V_{SNyAn&xBfm_78))Rn<f;$fbwT+5ePir@ALeWK)
zad%`J_<tlWa!I?xq%?_E{Ghio$35>}C+z&B`=xFfrq}f&+rKKS)9%*(=w1l<@wXzI
zg3+w%NGaM({(5xD+L=D&jg<Wka&CPtoQ6%6)`;B(XnRv&91}KSIRT){#P=4A!og$%
zVmI<|y|mg(oWoNybICGF%#?4UV#9bB7aR2e4(&V@qyw|7$8BS^&--nme6Ks}RL+pB
zo*Klm=Bz(p?zrQcINL<oNgbrkXPE*PMHI>H8+eE*)9e#fIEM^3Hj$xRCjj)wCTiJ*
z(N>S$TE~?oe4v`Q>rb_Ky8(O(giCiADl;z9cZ$!dTR+nU4OL7oF=nt7O$*11Jv)Qc
z3E!$igy`<a@dv)s^t)lZf4Key_<3#RyjumIn?d?BE3>8XiuWsCOBuKSg0T!9(QyCv
zckY)Tkl`ss;vo{BA<|<i>(!z7#=ByIiaV?ewY9@!5=3uJt);#90}O^cDV>WOkU1Om
ztx>8r&f42@wOrbIMQ?P*t#FxcXwO~eFBkno9$zDRSi2v@t=NRA>5hShjy_BQe#!t0
zfgyO-v-tR8<hS`At#@A!%&T^ZweHm`o8%8&=;ii{ucJOCWICr-<V%b^Zd0eW-gm=x
zSpK_8Kuh;sU#2|u%{)<=F3uP;R&!3yNNiz!RAIm9ot8nM_;q1v@#)XDLXvU^ldo<#
z<~UpWCAoy3a~{|k=(JrK^JNGhdz`A-(+PpKRfozbFEc}8EG25|n(7<nrX@{!OYks0
zgkrw{p}t+{&XN>O@fQG;Ai4c-*lO%Aus}hnWg(M`!Z-1w$ta1bN-KY(G{<Opwg`xS
zr;oqPwaB4G-A?b3)WBsxRz+F_WsJys-bAbx>_-NJgt&S;TN1FT4x-jr6UUfW&G8$V
zqTGnqq~WepT$u;K^0!1LFZkLFyeyMU1Wo#DQ>E0=H%nv$(dYw#2;pBkhBM4lUZ0nh
zcO{zY!j9Q+1(&qc>4{A7n+!vwnlewi8JQa!Vz!$<zi3VnpLm@W_Z)=L0yqh~wA@uZ
zA{)AN#qEDZ=tIN*&Z<2IO>(XZ^#ECg)5jpC13lcCKJmJe7dE&36lvk}qasxEw^N+^
zfa%=MW(h(SF<hPy2e1fmSxZDNpodbno($>lk2jlAooxeEH!_mm`e;hI=}aWIc<wa^
z06E0E|2paaV=uA-*hpOKCpVxy3nUI0sRv>3i+BOpa#dY?elIvn3!sII%jgj;sT81H
zY(pYTGZFv9BYu0&K)8Bu^@9B4-lbPiJ~rMZ{S8s;uz60l8vw=gqe2Or1M{M%Jqm!p
zD&{XxzvCYGEu--CJu{N@!L#ufw}h(wQzi;%sta<%mEt^6)N4S<?x#3<#)d1Z!$3!O
zM|g64FKKVxKIu;qZ=P*t)&8w<tyET7fJbhMH0S{|TT@W@LkPS=vv%n&@h~4ZvT~ef
zXkOH*U-W1%&OP=s`Acju6iWxdugawa3a?OiU|!|eN2^GxM9S-)AE~Pqn(GDfA>Zc&
zRSIpaCIY9ZpQM?`_LO@686~cz@=~iZX8g~2HxD<*x+;>psv{S^5E9D9a3g&J<m-?5
zY4c&GlulkWOtwHxoek|VF@sLz;oK7v{RN_{GMMWSwQsfUJ`x*nx1STNEZ|r;I-&{B
z5vL-gY59@Jb=^INLqFHmaPAxL6|4ra9>->#H{)4y;P)2%^!drtaqmWDSfs4gt87`z
z13B7G`0Sj;IR`NHiogKw<H2l@CZhPo<UNUfwnT(RbGFPgrSO%>d1g}7HR10yqQ}>u
z9gNuyjH)78=DmQokXowng0$f+wx$`(GoGhi19ZW~j@5_70Uppc89VeVlt{ze9X>I1
ztqU0Jwccp{xar7&C1TGz<WI;%72u{HUU62e0y2?qa8bx>gWb-qK>NdcBBU`vvrYKx
zI;?`^XVoL5f<!gsXp@Qfe%`7)x%Ny6@|1t>wlfGYV*<QaK?a{uuE+wMo(H7EyL;m$
z=H|oJ;rv1)n(0{YMe*oK;xNpUErdgs8k<Owfjv}Dx_Op*JOxL%OSvOcMWbY}-Q3S}
z3H>@9O(Ixgf|!XDS=~)eRT+Cmw%dg^uKMo4J+j#F%da>5m#u!`>@5xTE%>EdM=I>6
zbeR4zVG-jYJnQV*qZIPIFF*`xRryeR4Zi!;b;u8Enbc9M5e>>K_6^HT)mKeu;KbB_
zkh^Cf@AM<)LM4C9c=(cPjJA?Bo8Y61?)}bI`h?$ev!brWexG|T4Lpno+${q!eV_Ds
zk~R$w(|8oKii=!FlL{x4($@XWP~?~}*h>$6yj>3mM;%q>`?)YrI~Sx#ke9LIexzVJ
zXF|k+vWW=oD9==cX&KYWC#$`@>f?<J>BDj9i5<r}?f7GEpgji+d@EP-?@coAmZ$Ir
zNjRUQ_oljpSSA3(AD%79Qq-bP`VU_o!l(tx4bhLlFy^@pdVtN^y)Ihn*ZavEUgP-s
zlpu+zqvY4x)AB4dtTbC6{S}jEJY*XVp|l;s-SF*IWy(bz2Lt02nqQ0-T-ws!Y#3vz
zumcMmMe+|n=>?Qe)mbs0%Bc&Hg@PC~f+=oaEtn}wlSwz8ma7+kGFEuYndB^`b}`@p
zTB^P)UpUDH3OoFA(a5gUjds4_md!{`L6xsc0T#ggJf0XOoTFYJq2{c7c_}12mp=S7
zk<GiCZKsiE?bgJ<^4O}WAy%z8HJlB=_}<8XFiNZP{{lJln4}^&qAD}-lxn6?d0aYm
zem?A;fBM;#P_g>X#a?wNx`R0du~A(ZZ}p0Q!u#KmoI>Y=Ri$O8W5ZJ$_os9wob1gl
z89!9WpBXJD%7HdB@1FmUdU~P%!8e9m|8o?tq#4=ksl1b44XFTyluQ~ik|7u~Hrny$
z3?^N)H>SKFoL2M3$+1&o^bp3y&gJ6I(FSLYX21Sf)K(Sek|JWASW1C6c|LxFWAqnj
z6_DlNcdFW<r(eou0J!4;<QCq!uJT|1Feu5zhs@vD-eHvg1v>cwd>1txZ5{0l7nc<A
zT-&$$sUW7h=P%Jr6Z?TgamFXWfL3YlKGxx&>?hPtlYGyd0clp%U22NeGUHTK*AS*E
zsCvpb8mfA&eE4A=V`hz*Xx#X}Q9ZqT4E`%935b(g0dW$K>Acahcwm{~qgMii$x-4c
zJAlSjZ#CQ42p1wGqqkH8M;hgaaxdG@8hGIGRXpBy6QEV?p%3s>ceJ#%8=Y;eRu~m{
z?jAY5^~X{rQ<g1cnhfdLZ~XOrkza_^JEupEwwc>jai`48@9-Z|u_oLh_l$F}p>`5M
zo-kEunbO_&w(L_9O=&CIFb^&ddiqz+I^jZN;ybL{ISHa-K440eZqw{(*R%uKIMt^j
z-~Qug%@>%ym7g!QYreN2CkuZsEse0Z7a9<`vsCJ{+nMRs9HqU&0tJls5-oZ_iI4;O
zll4_ZO77S>ZkH90NBd}&smq9|Abj!zK>=PBC*hsoO~hZI_x)j|b}`n%lfp^rn0<5H
zc(1x()oTf`Ro1P}xJxm8Fh!ejdgE!=>3+IUKby9g@@X9_(Mx$u+9RmVd%Cj3I7;7_
zCL`<(6>a_gdTNGO%zD2AZQz_36CEtTm~$^%@K^na1+M*7jJ4b4*T8<d<)l>WE@>Nj
z&uNNZL72!nD!>ZTJ@4oGk_NXwlaLa7#<hv<%WI4pAd%MMG0MEQxfUo7OODr5Xy|lX
zQWYx+SXD>MW!^DGoR~-7G8&owM3_}>gsU_*)qd3Av`7=2$HzPRLJxo%z8(B`rF>ph
zWcvXj8S8^@8tSvwjQBXX4|R{&A~7HSDN$>YRERN)7)G}5ldk7pHCc~;j%q7%c{3Yg
zKlnl%^15H<U=ti8YCy>-iy)^;Qp4<dnboNJ{RKfv@vb~(AnirfeE-<i{i9WLHd2vt
zW3S{PPK#8dEc{a0V#g5WP{|Y0kY7}y#us;$8SBk#b>VMf#hSmA$C0n4fS7C|`h212
zA!41q6@XGuwS9{6$3B5;gED-mod$!uTA~|=ulZp0s|0eOU*AB|_r3RflU=r^mV{$D
z#Nc<s*D>_LxRCfj=%L<gk5yvt&+C{`lfq4rUl)CF?DJo>TreBL7DL4v#ur=n9~d7$
z=IME<W|Kk@Q^~=@$_e8IF@c?O`rC=tRQk^l{6?Kq{Pkg$!I;?lqOlr?ddpNp`n9N$
zN^z#R^8pB%<0VzVo0eePk|m&xBOI+&;!b@nar_`~?oh9ISuF7){;A0`C+lg(<>IAN
zJX*s~e8Fzz?niQ9o2e&@h1sER(mpedIsw{&pwA`uvtmCk-HZL+{9ZVbRo8Uu9`^YS
ztR=5X8|W31?;rjAX0&2&)qnX@|D{zhqTUA9$y}G?Wbota(pS=Z=F)dU&D9$-O2R7H
zr&eGs{EuZ;)bP%&Yj4xa(@yJKX}6e0ROFSdOR#%HEa>K|2J4FyWq;ns4D(wVdD?0h
z928cwvjnQcCW_>oFP`;>DF?f88~<Z8LY|{GUFhRGeMAGuH}O+6OE=06s2W44;j=}h
zA{|SQB^OOvSeleDzj662iqkMLZZtT4L4L4zFA0)FXk>lV6#3Pk>P@J{G$K{cOCfi`
z1~jU|BezET*^bxdRJEt4KF6(2utvWF^E!AFOqcNqod4a$!4bE*T0M1kU`d|a0Sl%r
z9bWesaU3IZCtLPmnpmuI!c&`MJbi%boRFJnd}DCwKG{HXPk_?2bE>3wpybEO!-a-y
z;Dza4NW7-^6$b6xm3uoM)%8$HdG@<lh@;A8fzV!uh+aTU4WMFVtM=L^>iInFy~NFy
zO!!7VWC_3N=eozr<MhJxaWF}M**_N?3^ale+s=0@O7T6Pce9**`k^6DZ5VP^x?u_D
zAb>&T_}HQ;@(==N2MY4`r8Q`87((br-M*&fC0&_ggsZT_QqmV}EKSo_3GD}Gz5k5t
zA(xNqS{jkQ-^SfBlY2Z%s3I5K*j~mKR+S90?mqwp;}`<pE@8$8@Q~uF<!<&~eXZ!z
z0w$r9Ze5#m8p{v!wh>cJ2jfsXI=Tv#wUd~N^2b{(Ep{YnoFD51jI<I!n``J$d=Omc
z*2s~10L4m4b8)K(94G;+(H)`-4msE&n)JW-bov*aI$t-}D`N`x_ER~6*8so5<m59q
zHzBw?mKq^0YdDd^jbb<L!wQbWU>L6T^l_%+PkEQ431xM4{PxBjt(nB-PRL;<s0zqg
z72&wUT?F4<07I|220I9ilS>G$IdapURPpYMQh6F@kL%|a5{0HKAr_xOnl)HX1TIsO
zGVcC^V4?M3G9B0X)!s?=$f3+_gv+`n?{Nyr;}$`*Ww!*+-3K7Cd)g5LfEsEApu6Z6
zD4yM%w;>XV8{qH?@A)IfF=d}9Lv?JlxytF^!ETOcL4OPAjsX+LvF-7-i{i!H6tLvJ
zeZfleN;I|QVIVE&(iqn9*lz_Ay@L0B&Zfb!1-Nf{59)>E+{7414iX$|s_5^^%a2`U
zA6gKovVH1tC!fO9_Q}u5NUT1#!x-vZFK>X}Q1brk6ZbVl{<L+4D^Cw_sHW+JXnZqL
z9clXcdw1#cZAScen>aV2fo{F?@fcw<HK?@((bHKf7RSK;)>P@bR`IyHq)i`F$uBn~
zax?(-cN%C!6hx!{C{->+E~o%YV`Jn=SL>UW2L+1G!DQG?A2I67Y6THV%3I<W)PoE>
zXWE_efHUY4U4_(Iag=k_(N<G4p46u;?Q)7^w%6ja-@5LF(N_NgS?}eXKdb|KI)Zx&
z7rTP&M+tVbs`HkqQSyu%IZ7bgzd%-d&~w7tf3D~cazVxcNz&SnwRl68Jl)+M=fyS7
zJP7ZtIf&!hTcz0bJ_BM~^K3mm;Y%8sD|l?NOIk*cyiMQD^zRb{{H=JPr@kw1h&x-4
zhk!2t(0ScL$4|vHx_5<oTI9Q7VHClwR~G9Bx{m^555=oSY42E(Qugka@VXzfP1R~x
zqsBq%Y}#D<ihu~Q2;?KO1t^+jA1T6NJ+mI7A6v7=^0*Q>i<q9GuhqB|H+K!d24wRr
zL_(I*40QZldkox`hA$hhgz$<0WQVte3P=W`e~Ckbb0oSa?XrS9*`<&9(Sxqax+%iK
z$gZg|<Pm%}t5gO1fI)eYr!aOMjoD0*({gYSGhl9;*Y0$%rnmf50{>e9+lL7>iy|mt
z1s|_0J<@pWk}7<ERlOnB@QY>gEIDDJ4-+tyH}i@YF{G%j$D|KuHitYr%Sp7L0N%a9
z-i=n%bmZM-{;9ts>ruSE3}9LN1$7X1alSM2GpO^{b-{Gx^*Y_)TIvy#WAj$3ZmP2T
z!wF5NBPZa<wlN8rmCRYRy48&0ZHtGEo*vWr{633|v!%GuC``-TPPX@&;B9QuV*kOm
z`w)YZl{NVbEvhq|#)Tt+4xk$qzKWscdh-UDy#sOI27BXgU!~3c2zcDy>r6jsBu`gB
zNMT7EPSjzC{=@FZ$)z&WJG1pm6fq~$&3P5qUG6<^AN<f>UoF4q+p@XKXDT+_WP?Nx
z6OPlHxmtrEXyZju_b=2F|4mhAOwu23fP<D7*YswMEYVty+^ciZ+iIqw@H^Z*gB)GX
zCh*POex*NG61MwvvW3}CtoCzUKIT}v1bW^QYOjNcL&wwov8j!lQ#y4{o9t~{k^yqZ
zmaN3T*|-|beQk=9a(A5Q<&rP52+P<gDU%0_nNuh*Lzwgy|I=IRPYM4&@k2D+k7f9e
z@R&WA?)E8)c-C23-G4(9-$=96e(`OEOkW$mxud_~R)^LI7Y9?qmx%eCFAL9V^f3j*
znJ#m23Tu@GQX~_a36=O4fp1&(6m(TNKR=ZHPgLyLqUP=H<Rt*N?9}I6O_r48;)P5?
zP?sO>Z-j&Mq1Q(ocjpe=oalgA)?Xl*LO=l#54Ov>4+Pk|!L+flp!6d$=(s)Pq~N(f
z1EPJQX5Ui-nA;t9V55id`>BI+ke{CV-OSOvw)cekQdGpZ&4~$S)%4&~`O1c=u+7bZ
z>u|b11LvUyaD16;G#f8z;UX4{#8%!3EKbQvC8-kZs!78@;Aejr3TGC5or^+R@Wh0M
z2RO=;&K$7=P9Ew$KU+5Y)zCBv1{vFE)7N&6h)Vp%2Sfv~>k5pPiZ+^K9rc~czNKA9
zFn-dR?%??WwO1!7<B#iL2J$zV)id<{jONgva$|2@&|rpOmxqj8QW-#L&Lt{<q&qqb
zQKMz$r*>A}Sc#S`p(s0y`FpOADdY4gY0uK#0KOR8$L$f<M30w5_t?OGp@pFKXq|q)
zF|5Bp;j8t)ft(=i?YO>dtJ2VTYvbWhW-Js``OY+oFn~gT(|4c>EOS4R8N8i?i0JKf
zZc0OQ_nIS4H^pPM?x`wSi@OPz(w{plDqdbjIZRdpt4UcI;$v%0Rv96!CuK<qUih<|
znR||qDX|-(_zN?z3j8kaaxv(q!_*2T-|^dMCax~i9flX!FNDIX;3}}Ax%f6$qUeTT
z(y5<SES|0+X>JVUEf1;Bxhn=w5+nw%M_>tA!9^%Tgu?2Qx{C|{cFX$!*l5h}RBxce
zxk<=xO<aIzw6`ID7MYuvsIynZdA5&!>*STsOQzyjeg_H$_KE<_Ue`s<mlJKyWxF@Z
zVU{wDS#Fn6CO`X@VOEuLI-%dg27+E8zC4Uxvl#eXMUNcZrR|8ZeRD-zpLS-H0<ytO
zT2Ew(TT6KB1+1q4W}X%J6<~cEJ5Ww7wcTX(jmr0n^|6%?!e)q-0KuDRmcK8;XYrE2
z-lo2;>ErL`RGaTT1S>U5WpIMPigJJ@S}$bLKx?wiwrtPXAo_f-qMXlsZnW*UjgK<-
zz?L|GhyymhX!jD!ez$!eb^|Ko?QMDF-h`V-!?td_dl=W^NcK(!MpmiI`2Yq_tT|Wg
z&59}zF58?zDIY(0TE|0Vg)6cPty#(Obw_BOST-deJ$Bk88^sx+3Btm>`9l9UR{$Ch
zwQpB{fp`N<!~J|UQpRT2*Hpgq+BiqsbJ!2C(Fa<6cMXNmRfaJ8rJX6woClh4&Hnz~
z!=5Lls*e_dEfZoH4zxU-F~-5O83>GC{U*2?=EjZf0?XQwE$^S62doG<+EUM(jUXP~
z-%I(^4nL4pE&N23YWAn;DdY*!Mf<+dxfre8h<H@s`ezA&)Nkza9c%mb@!iuF-N%H&
zroaB}D{%|c8|V%8)S7H~%(4iarLGNq?4?pb=Oggyg=f0<lO-7eA;eCC@NGL!Kkd)#
z0$28i8D9^V1f^~eU|2@{Hx~wQMxza03X+=@VQ>04ukuyqyJpiSsc?<J^P8kU9SAKW
z(_N{E$ZQpv8vHoZiBh4uVZoD!bB%J*4eucAi`P)U;}q0(HC`u3X;y@RPG>5&wmqEa
z&q8EPlxegguq7U(n_2@+P}g|BEc>*paInu!k4|IS>?xj+O{`uI*BF#VKpwa&25ljI
zw672R4vc<bmFFxBeQmEZwcaGH2Uc<=t}i>kQ<6(m%V{0Sc22Hi^A|{eRZ$KM-tTMx
za-&m6>{sZf?)y!zAMN<3zwuj+h`ZzY6@ni0CVc=7M0aQEiW*3ISsjSxxWjgpsds!-
zCL)&TQ7WbLX7;{-w`*iu+FaTx^;VR8lrn=n#ks+ud(rCiAZvey`U*$(^aX8GOJnmF
zUBPbhmxC9%m>qXz7t{U{)$sdIji1LqKC&l>9H)EtN+|K!CDFpw#?e6UuD6{xgCvjC
z0ea#myeF-)a{FqS6tcf#fy~_5jUShRQiC?(+c-}~$lo_4hjnB6NK}<Tt>_2}UCH-}
zTv--|nW7}#T^r4`yvAukONVFyi6Osa9JSu7PM%!wq$cOqg{#L<3J_2z@yZQhh%+{L
zRHM8dWv;>-Zq>y{o5W~?d$p9?jTS{gM5QdDvPL<V)d;+7Mc4IZE@oaIo*`BNAk`Gp
z2G)TyVNii$K`w*i07tc83xjFe*`1z=&t-)TQYnUwQt_V?e67SV8D;mz*@&X}m@;Sm
zD~|I4s<<b8vTBhLH9ZB>DXz2?Frdy1f_z&J;zkz24PP95+xt2@2t(h{qUhpoIf!71
zJ0H3fM#BU$tY5TF4^9G(QY2~Yis@E3ErUZ>T~(Gybwpl}$r=%<V`lb;OOr&R{<Ggo
zP97=!Bww7ynXBV)rM)C}FxHLoQ}7C7{Z=Hbtr8LhhE*_$188roFWn|C)Dy#V$OuKn
z2u0Mnm};+|Tt*WQ7jjgMb_M$Z+M~BQ4a#k1v3`zEan-t91-=A+IWtGVi8ic*p23%w
zfAwdliKXdW3$bxL5pd0Twv#-734VUng}P5><YJW2N_om|WcaEScq%T%4h@Qzm;88)
z+p<?wk1CI6x>*VF6iW4Z_R(Onxd6L=qJIT^y#?OLFT^Lel*5?7HwF4)&sE|eEusi%
zns3PP?G^QsL#x%z{%-Xe=6^b~G3*ya&vqTh?;-Y0nVkUqPiADFHnnUvgkGyidhD~r
z@}!I51=`w6oUt7E%stc2Jy!+6QVaz7?S_9Xx)%$XycSm0kT_rYs@YNf-N$Xn6Fn2o
zfFtgMxg-yXe$5@*hT!g7`n~j4-yiRzI~voLT}8O~oKbxP86>gn4`pJM2P!!+mfGGy
zcjBN|gO#ouwy_0pvCs=w;`Y)OO0CL4-LRJBU*ur<mZ;!5GmSr1%Rm$z6iZsexp-;F
z-A$VY(><<6H(SgKr?wOl2@1N0Is`+TqUs^Tn)lotACXcpcLYXPW57cFEqz}mUY^cJ
zQhv<5`@OZeQ%TRboH0RMHj;432n)ZEVcOu89JtTG5rEdYaHMT8v#hl_DbF(YdJwt)
z1O)=-mefH`vS5IDZc3Tsl<};G@ag*V1UCZTMZh;E!2L_YY#a2la{Bm!C?YeowkgFE
z6Jh#{)%0V$8t=gTL(SY=B62|6RScvM5d?s@IR2zTYohoOHX)L5#?&f6LesH8hZFiG
z5;$u>SEi){jXY=Oe|JmmvIozRJnJLD3M=Y?@pbB3g91zD{tPc<euV)C`c%j`5$ZJY
z{+i+k$_W`RON(SXz<Yf;Qd&tmwaDTYuB`9WvQN$}_Cl=-Cc$vw=Xw-fw6br3@4WJ3
z(Im2Nrl6(ybWYY(flIgL5-VLcWBeKD6o_d79t7wFhB8pR6#?{{vU!vd;+qW5TH^=U
z=f_+tFx_P5^@fc8`K{|K7t0~7Em_=d2tNh~7{_25O424%wiGgWPu$o%U1{B1(@tfe
zB}hEcyK2`Xu#a;>w)Z^~Q~-3~78|ZNAiLEsohZFvy){=7#f`r}ID5v2q9?8tD0GWQ
z3e<VK$|_n({xH_+J&O#A8!<RU^ZQf>8S>4Dc2mmYF^^UUje+A=!zA28Hx+WU#cLAi
zI41x_C!4z>b~bEs2vI`z{`OYs>2n{h1<EVl&0+AG;BX{>7{(ikz8Z|JHa0~58V{>C
zX@0z!vN&~~v8gxsPVh*3CG-p7P4a;pJ<I*Cd)9~FQ8I-T<4n?_$rnEXr{z9q+}^ro
zbT<85{ZF1F16q{5NSlEQ9x1tj6x`zlH3+9=oEn0soX8fNTqafXAB~ts$fWVZQfIiz
zdTHj5b!Z(gEb0nq6BB&}`g0-0qqsMS-!HL)APvf`wFPSHStLmUyu^criPj$?P9n^E
z5S}nAKe6r*>RI97CzUSJinuL}>IDrf@2w<O>vw$E<Up{}aS%8%FIKHcHBULk;8n>x
zLBnid?psv|wRa`jZ|uaor>egsW7#wISX`yIJv{yN>Saa3mjc4`!7Fk!dPoEG9i`pq
zJ>yh1Q%^u)3KwfL4hPOE&YpP_;xkMVDfA<RW~RP^*Ic7&;PMmcNn`@ky8%C)t4_ZN
z5Ufg?5XJ9|*(63H?@azdlLs?Fe5OtcnB&l=Xgb_zUKRolDi%?zgxQW<slo@Wj(+ol
z;W`w}_`P03@dN8}6crGMEPIzc#~i}lBaoX7iBMVUxRe?jMnW`pNBh)?^-TSU`Yy#R
z&nIo0hoD#GxpA;gk!yc@SCQpW(b*a`k=`-Z1kS(zW9~yWHW<Z?`?`t0v-FdKCDXhF
zKd1GCr!sosAeC)S4DB|@%eTmONRR&1W8phKf`gIe)3g?jk*DSYAPHWz8<5ma;_Ipp
zoz6>i1tmzwmz1Y;6j6qRYeD%6idY;op_*&Koaj-XZ*?8YN=&j8dyEa{dn3kfCEOp4
zo;7jM3_PC$NM^`MclhFgTFq2_a}__ol^EX@?^An5HxRyu;g=f_P7nfqpJnY`y1KMr
z4_9;db=wfni}x7oT?6s#t}bb6w3~VMWX3e6A404|`-5BLIo~+Hy%gVJfCJSsL>)N!
z6OVXYk1UAxb~Mj$axULM#cVx3f2O&V39#|5sysMdRhzC734XSuIsLe(wY|li{$;=5
z*g$pF%s)0`ocaI${qKA5?`PoO&%nQ*fqy>(|2xmXjtfKm2}M`m=fkVg>P<L3H2Hc?
z$*_aBUW`_K2z+zkxx>uhcW?vcHfcBHk6OILB}Bv*OeTW!#<%1s&Z8ZVeD1G&Z=@m3
z?|cULp_;P|2D&x=K}Fxg3is;8#~czq7xhp@Q%;?1sc1}PExBn*=eXoVzHzXC$KQpk
zT8j>^KkS!3vuzSatK0CgnQ*1SBP0u#%}*0AXohb-`Tfi~{%AtzxF36ROyl*C&~Lv7
zg~{(BAL;=?Ld_lb@;EA{^Abqc1LZ!D5#?N~iQ=?5V^R9i;hB7p>C~F~QO94<tgH5X
z(FPRT-`pMu##R(oar%m>=c?fMQ{yW#V3>~0=xECR%NvytHS4)OIjk}<)FZ00md4<$
z7eWILE0qRM8$?4mJ0m;=*_&|<HzQ#f@I9n~uH;J$O1nk<7#l^FxLRx)jNe+qlU<*;
zlUX%Sq(Bhko@zh@*7Nc&P$eEJk{__X-n2-+y8!I48iWxFVcRL<6F>!Djd50DJHval
zi6=Aqr1Q$COC%Adc}+EU-^y;5s8}t!VmiCWBXC{lo^JhnrvhXzTvFp7x6j(|OBR-B
zI!pPXT=^!Uu*snMnv3O~x@1^F0OMV02buOw@X^P=Ky9%!E5vj;>C@N0G|1Fz({+Er
z(iTM%B;@4$7bOPC(HeZ69o<V8OxuTyUXsz`fW+@RBXv4_82P8E+!w8a;^2P5KeZH)
z>56*PFDpSPuEl6n0<FcJ$N$sbd44sq{cAjc6cIs0id2P2FF|?@(h)hJNRZxJ=n#q&
z1JXO83J64yCQWJ}AiV?y0!TS@1q4HruC(}@^Wv`i7o4^3x<6mc+gUSvKfBEA=lgup
zPo+Qz#-x?D*{a4a_vBfJRCTWNRy?%CZCXFBv^nLG^KOw#`}$ROxYzJv9pM;F?fTW#
z3e4c`<*`3}eu}StwpX$?6BLr*Gym-+EVA=wU9zj3o1(hpZ;;r|c(U@k=AM#_ZIhzd
zy}Hlp()*2KR|SOKe7f7xg4e@%_Cr7QmOM8QXIPMp6Q3#Q+*qozm|xbi0{}`mRQ*>@
zt+gn!@}E`b_@f`proGb&{KHkyIVI?c>G$tsIeSGTOl6g7QoW<&_18y(E_bIIs+LRh
z%7(lyP(5NlUmY=KqLH}X8e~V|xs>4y(<*Eq><^m|c6IKsQ=p0BV_$7uGoToZ-u{NE
zm#}D)n(96TF9ZeAO*d8MbYP|s<)#N54|da_x!NU`RL6{+c0SozD059nJ-;>G;~f=-
zDLE-!mTw0IyNmWjLaH~+Q6r5{>%59eR(7QoqfgUIX4YVem@`wzJKR#?5civJ<6#-~
z9T05HK0&E8==xIkE&KgU<SMZZgu~K%W$tu*WzJ#cI#Iz95zEvr{+|gq%q<2mw6oOr
z*j|DL!{}C!fa||{m%BA(7=r>Q9@)#?GGEYer%!u+Q+5z$7jAx%M@r}rE|{Jd@Ra2A
z%q}3odrs%wdq3T-=gl!~Ix+NA`1zN*T_oN{h1qWlR}-QXd@r*{r%7vsjVxKK#Kq-p
zzWi8$en#OLRTbT>?<adMzUF@q+(A*WYjyeIa6AXlC{IG+!SzO95CCTjKzMbQNl_#^
z>#6DneAilr>w&%_b;+wRw0-&*44UAcG?2tIh!Vc?DAI4WmXI4^(s`~GUF1ynL(E@2
zLh&0mWkZGS{U`SEm}qqrou*+blGV9DLLgQ<+TgwcL>K^3GxR=m6RkZO09Ofv6ALzb
z(tGbLTEgY?d^dEy4IuxvI)Dm-I*Vq_W)t4HdwurIXmIFtSAJ~-rIe5^^s&fEnX^js
zCFt#?=;;$Xxy}sRffLj_=Y?lyNP-GwL{6T5tLX6Hj@abJ1g4U#yYPSx2p*{|BdTJ)
zlb}Y}6Svr7w_|7+!v#a>N^XwaQc#?wm&vbe4I4UQjQQGmdeznHc%LX4SbOOR_)x9l
zRM^t=*dHZ;W1~i=Xu3R<0jS?lg^3n4{obUIsCQ$_i0!YBa=dvD<C*z?D3c)(`y-Z~
zw1gx#8|g}1cls}4EBbm+rFXfh8IGr=HChJRGfGk=!fcQ=qc+I|yFX-&XH8$`5Y(r0
zZmuWnFIo<Wii%Rc8o7hFd>ICA%P3?bd}ZHuX8FQNw^Uu35bIvN>@BEXP}4=dZd=F1
zDb^Uz>)F7vSgh*95Rt0XaIs|t!%XyzFKjvrhdoxWJ}B({05WfCZ*IFXug~7HRTtiG
zzcq&j65v<k*Ci>u8brVL?9WWp@A7>vnn4g61L7>!4!%9h232$sPNxbP{x;*#*G=uo
zTW`H~m$-)m)DC98u|R5_E2pwm7h!i$?l`$^lQAf=aP7`!sS4HE8O*UU^*dH-I5$L>
ztb>_fd?t}tT8}s{yxwh((y!Nv9ZueBxVd=R5Auhr@jC#4VpJtHPTe!&?U<A4i5z!s
z>#_9LVZJE*1Qrgm_toN&ALLfJ+e>c!uuzSl-k?E?msZ#NN0j$N$!&gl5><m{E<=z>
zoI<}rCcydF%!&cUTDk@U_YgOaTu-$1HLeF+Mr)H*#f3v@n~d4AbLNh9A+{;Pb5FhH
zbT~sOh(G><Zv_TCa`n7dZCrLA$u%QMQfK7hcTE>|&D)p8Em$JV^_pf!!Mr-qSBLlB
zJ8AH>kfltQ?42aJ^5!v4Z3}q&dMOS(r&z`R)kxBLtKvRo+|<~SgB-;x4>ai$#{_z0
z)#>2S@ln+^FspvtNCr5;to))sMUjV2A4klj&9URy?X1j0t!v$ipfAMg;K2cd(0mZW
zAJaVJ5Y^Q8By%m&OS_Ws(^i%j0UMq7+>TXc$E!{r-DMrx2_418gYYvV*6r_|JvI)w
z!8gxet{zst=hx;ja{u@a<hH@X;mS3DJP=z5H9n0`={p?a%yxD6m}gyAj|o8Y<bpoC
zKk0Tj$lWF3D|;n4`3@X6nZ}s$@<G4D;MX09Y3z%k+bjbI%APerQ~K9+xy|5A_gj4(
zzcHp*w6C7Bg2u+(tE&)SIhqFZslcaE-JBQ6f_=aY*wvske2pP=S`_jk=7TNJNg7N5
zA;)SR4yAU1HT*O8siwD)GLNvm4EvwCZ)YQZUHo2qg|JgGkG&CBJzz=o2(8?q+);)t
z^92&NlAE$i%}VoDJ|(!im9&pN@36A=oj<PVTn_+}jG>pgVba8Lx>w^l2HM5G0ye2i
zaG_Am(G^2>W&@o&8*o|P*3+tq^8txL0cGuQi*!XO4~QrWjS`D*oOe}H@^o9#4UXk_
zCi#$APIhBDOg^_9o0&dpW7a~`xp<Ri+A6-rfz_cAA`WXq*$OL=!MOKMa<hsU+!9r~
zQ!lsA*2AQDas^DL<>+T=S%gTRo}U?R<?cm3dGNflUnqxz7gV<ybZ8;+ug6O906F?N
zsh+v{M}C2gGWgl*T~^-$!X(|yam{W#Xyd}Zo<o_m9PiN}^r&L8oR|r-P}9xraYiKA
zXW8BUY9q*y{1!EbeFx{RT{Ms&sIY&iS1cgmuWTK5^MWqRRTQ_!A9nk^(or_*%=nUM
z^D=QpM+4*w_($YpBCRpzo{+odXUT9jF@|{!x71DUXEm#Ax211BOVO`JOh!FTeF)Io
z(6q)6a$PR;Lq+=B8QKMVq&F*3iRUX=0NYDzWzZro6PL8}v9+N!_AW}(pFX)W(`X-7
z%lW1l*Gzq&T2rbMpZORMa;g~dS0{yNYTPl92Lb7J_`uIZ6@lL%i<{WH6F?12;a>I<
z01MjRY^px)K-I$IUaT|bXQ}&?UGXIx*OcagU3eAyV1s(Sk|g2pAjq4}Ze@Z!s&wV^
zk6nLWwOi4X-u*)FQHO0Hd3}?LScJ8P@ZX_-Ny2#qg3-lSJ0a^VO$Qu;brgG9HhUv3
zg@WQiU7W(7L)z2D@=}g1O%@CFAL`R|%ZIi-ZdbdC4PBAA5R@3Y*b27wSZoRiTFF;o
z2hi`63nIHkmma0}d}p;lNbz6B4}c@~(eGRbkgdWWWyF!k%MH#;chU_4^S5_X!HE$H
zRD;kG#H0bGSf%Ux<=Ib8CBhhHhMLhEVd!H%skXOsc}nt23$I^|2`HC+FnRj){2fIo
zxqR4tMOgMtXJa8aODoykpiCT|V3Ng}oY*|xYv+6qD+r2JJ7n1Hd(m3+E>-ZQw!zh1
za*z*^nLI?UoNQ7@6!`;Xs!R`Tj{QG>Up}L+{-<q`ryQS&AD}08kF6A(bJs8vJRg$A
zNL7m!SP;;SX)KU4gW9KP%*{$yG=<~#!)WR=1>Z@0E?CFvNe~}27ak8mkLY2lQRl_C
z5a=tJYHw)nQ>}3x4><3AJe!oaN54Xxf>Rk5xY{lbf6cgR@{?ukyAG8tX@r8-oh@hU
z{w|2QID_pbISV&dnuUbmA_684kWb*pw@C26v`Wi>)jH);kc)DWTWTjF+AJeyO;BrZ
zPr^d#7RgAU@>lv=VJn}7D*W9i(3o%wky}8JCF6Dw(}x#gy(J&!ui9k@@F)#(Sxpqp
zNYR%sy6CwjyE74>L0vVg1zm@OQZrbgMEPjWNXHBi1tlqfa$sq!lun!eay#-=CAa`0
zXrgH7h`KR8a`L7d!>>khG@`r0=29ajnH4SVm;6Nnb-@!qM=cAc8Ka;`DZwGk8rx5u
zJ_zrM{|rF9!@^oQGd+Z*9MW#1Pu1MPkex(hBRz!i`A1rLdjm>7Pd||#i6VT?1^hIR
zxQ$taTcF~c8<N>iOy>iOHZ4392R?Q>1*mRPD9Np3_aFkdjiOmdmHI*6GeEuMX0Dpa
z{Hq<<7ZTo)0vp_S`7c4|)7c3+k%SosT{TayYF-68s|hSM3&j=46;AXR1uuG<cxSxE
zwhqXUO5{{ofT*m$d-f}OfXy$#0ETIcbw@bI4RH(z6(>t3(z6OiT#m;-iSJ`F(EhHC
zrZ!dDveE#TPN7r$c|@iLYZQ#mfWiGlXmYG+=Yi+c#1bL0?o%VcIZ3G{Ja!FAgUgtj
znSrCHJPdzUwW&!5fu0}Nd>A*knS>niNTNXy*;^#TzlKUo8q+-_Acg#Nz60d?eDQ9@
z3D?<Iw#sHd-k3eYKg{l9`fP4V5TS9~V(bdwd5>FfHZh6?(<DcSK_rtz?TVJP0uIbA
zvjeymMYfENs(m=_kGuI<@0qJe+rPORV=Hc<AvJErmcfp|p=kqj@5AR-;V%qJ-fG)L
zwsgN}4<`&&gg^m|Vzn<3TYFr+2LWCQ*}uFAIQ-A`w=Bqk+9kfke?A;r&T0>D6>vA?
zZs|trKE|#rrrBGW`!O2!8Cq;UIBGMyJ+MJi>K)m&Ay<>BIYl;R8wO$Mhmq2B<~D>6
zv(av!1sPr_xhnb_V~lmb;r(|6^mmiJFJ0Y=vaUCJaL@`Ik&OIuC2Einw+n}-@KnD+
zxHhLF&r7Q5eqaa`o^lLc1|hd6k;6ZL^%Hq6g#a`T1&Y~us+TW=Qr4UJlAvprz%vpf
zg?!>ceKW#u+sDz;t1>qjbHz8c;4hc^Zz!rtH@7mQ?QmDq%(l!}v*(CF9mp)vYK~H9
zofa@DTGne~f^HN$&p58p+@QoEzgT9FFh@QbI5d`#JxoM*&}u$k!J2;D+ew`EpBVCg
zfNe?r#H--c1Rnz`iwyqHUz$K21oeOWH9FOl;11!>3;y^U<ikJkY&yv|*Uc3g4>+qy
z3dq${l)`~E45-=757C8lr!J^@dvc5Z5PjMUIXSMx%)OO14GBoAisqXSLo?tX7OztA
z<f-Pbfmq&u4~|HMRvYf~UBpWlYIkOW;Q1bZZwfh44@pKfH*kM5IM0>R^H}O}h_J^n
zgPlG`1wc57nL*P_LS~5QD^Jx~K-4{Z@8UfC%qldUWN31f!oE|U?x-*y?RYF5tZr|o
z%`ez-%VoPjAHF|})Tz+0?Y<z~z@{pj(*Oo&`@jz{g<v7TU@jE(cW-hh@y!!UByPh7
zko3-_Gnwt#H{l(JmJD(2830XEZuU3|QTGC-4;}yyD7&#b#Uj_P2LA?`CpiL$ZCoOB
zYZh7p(wB*If69j<AL$mn9jTL_bGuu)3Hfkx=UA4ec{otk-s86Z8%C}ebn>!B!?O&~
zc;)K=uzl`G5_xBmg^fj}QpYKX(53|Zh6|II|8LS<_Y`6a*L1hSMT6#>|0jk$FxeT<
zL7uR`D$1605E#?YRW*K#Qkm0Y&)z1G?XNI+w&|mT<+uK+>SWKIyAztdBzBCC6Xbp<
zO1Y3X2CwzmCx!!Qgm2x2o=o0Lg#Uzu=CFB3gI@UV>LFN;2n3aD<9fagbIUH+%>L;d
z(uHoH0G+GuZ)JRhiUA^1w+ph$a{g2Ke0BBP>+4h)4=0b=O4gmEd5bwXYoKup1}BbT
zn>;7_VSwFkP|Qh3rH!PO9g`~td|OR2ran0=R8OZ|^@K>{?7(KP>j&eEw8c_aRGLbi
z<SW%fG*MnGN9<!1O!>UEHl_1mri$QenpH{b<N~PI?8p6<V@Ky1*g}KJ_{$obM((4e
zP-rIHMAd!!OH?ec<Wq1!SHIrBo|yG-(sTdZ`+v#3f7;?tTl~?$9}WD`z#k3#(ZC-K
j{D08^*YDY}a0N;rJp?pPU3SA{^l0v=)Fd_}(naz=zS8e&

literal 142464
zcmeFZ2UJt<)+c<B5_;%OASfsxAib9WN*588-c(SUbb(NVAXPd7iu5X7y7W#2=~a5~
z0tr$AgfQ{mckX>>)~xyFn{Vcu_1^g())SJPlXK4A&px~T_I|E@UM&GsnyMPA03IFy
z;Nkv&t6AV4Ktw=DL`XnHL`Vbz5fPJ8k&=>-kkV0t$*CCWn3x#p7#LXCZb4YCbFwlp
zaPV_*-sa}L!^;d25E9@Ky2ZoG^XDXZAP|U@n3R^3l$M8ufraP)^>)<;P!khQ5rFXV
zZUWb+@$jkfuDSpS?mh|e{<Z=Cw&7jF$K5Z8n1qxJcR&pla19S1{~7^4At3<)?r1;U
zet>|QkmiQ45)rNLbI?r}I+3?=S;U;m747toN03~iFI)pjNEsNJm|1S!=H}tOBPK2(
zDJ3m)@4kwvn!1MOlc#$628Kq)mR2ucS=-p!xxI1s@Pv7J2fYgp2@MO6h)?*Cn3VkS
zQ%ZJDZeD)Dm%^gTs_L5Bx^MLj9i3g>J&5nUeWPRJ6O&Uvr)QQ|R@c@yHn+BSP=~*d
zj!#a}XXk(D!UORCP8RO<?}YtVbW!8zx<){NPXPKu7v42b+`y+MAiN<=M5Ck&dhSAd
zQ{*i%opM}OMLP+n=wl@P3)c}+1}?GXTc|&z{f)B!F~S1>rzraeVgICS1|Y}B!%ZGO
zH2?+9^CNgNul+Ut;=x}s@RtnyB?EuSz+W=(mkj(R1OMk_Ajfv)Nk}4SBG}4eQSM!d
z8uj`4q`yvKCgtZnA^4C@>p_Lu%(E*XeFVPvqCPdlSj5i;KmXzZ<oAr=#V_5HkI>C(
zi_@QF*4&ek8OFeN%NPiA@*aCZ=a&yY+he>u;rrBkoWngCAmdcseg(*9UjZ%M5ZfTE
zkIEI`7qQ{$XVZK9f9hfnjV%3X?pyw9n-Yw8q%Oz$2&AWCWoS{0V%=Q-HBqieNT>U+
zH=a&#9qxfI)aV_nYKG9qI;U4aRwlU}Pxi;a|AFTaVA=X*B{HLmDC0{lz2w8ny}-af
zWs<$oRk;!Tuf{EFdpK$<;~-|uUbU*G6@8?0iZ~R*i&CL`LZz!h2mEu~w!yQ>jsSRa
zbHsq^$FGtCNO`jVJ{Cu>fYlyB#92yP>80TR&D?r~2Yq8TqE~h-_2z5k!`xz4XIshV
zd7dw(lCFR`v1w5{2d*GH!6%QlXO8nQC2_TrY$DzhP<p$OS(EE8pKtVi`@X03aVj=k
zP9s;$fqT}Hm{K%o>hMcOw>Qfd?#J)&cjL^xyz{jp9lCVEjhBZ$*hDSev1iudc?A24
z-j+35Y41-ItiNzaRI6R2L<*c;0l`o#jRPkZo`d;};rq@q_G`NVueRJa+wYf=r-M1K
zkKzgOQE>*7jl>w!=6IGWz7z-UH)>Y^$xtYCOTg(2f-I*(eaK$!q}NOHKW+BNpb%&h
zF1ugLc>Y?~jz^k*;9{sijCbyFZruD8(BZb+dQN77Rm1Rk^zB@qHQaYF)gTCRmD=j4
zr?Bs6VXa6K)24q3YT38~Sa!s0vV$~hBPJoY=V?hGnMDtJyKi371tygQWJmA-{}7}S
z(tnB<P^UuonCYJ)`R~Q^Kkj>dkGd;>JsgY1s#e>}np^=f_pbn7IHw}hxUP!u-_)~S
z)0l%eR6vmG>yyNGgt*<Oii~}a!nj9szpnsh!}gp9`CTGag2o%wd&(OtD;q1nJyVG9
zbALk^@3wGt{Vs0u(vRRkAFmmF-bv~C!L!LpI29{v8MAFzRYolnc9%oOH%9iA*<zZ7
z`i=4G<*zYfl;F{(S~Fwo=QMTSn<`Gl$7~rr^=tE&Hoxy(+KQU6(aAj1L7)38ZSj1I
zr;Ar&#0wd=moM+%ZIBlIW~KkC(4Fg14sZLD)KN0l+OJg#h4RI<7}mm}QWK<qp}VE`
zASdYiGknVIs?EkLAiPi3w`(a~cFKLp@YgMrjKkBz{U?;fODr}ipBaxW7<D{>dA_sO
zh3P9`s$2zGGUT>=L9))(%jG4#TEAy{1w=3)G+}$AGZcN24AenpkkGp{yZ5WpZ;?iS
z0zBwTCEZe9W#;A5)no<yksN(&GLGBA{?PtOLZ@vwYL#&nGXR)mNAQyK{84ZJLWA$$
z36rsSTIm&l$cDE-<Nj#~3B*HJ%pM+SMv#<pQhtv4l-^MX)_msvJ10Cq;d$G@gNW!&
zGT(-UM#sbV;&JBkS;;$m-!b>d6zggeSmnQR#3*}VYmVt_sy$Gz1BJqL=DT6#-B$qR
zt<vs^xz`a>k(AO$_^kS-1y5gE{=i7>I3=!!uLrAc0Otmt%gxr>lW#r`$TR55N?XYX
z47!Zwp`XjAwdCDRD>T~k;+$<NUynF0#c%g^Ok>@)bqq@o8e$I>bA4B&5E?RXVrDc{
zDWo{`<KT#%exd$l{p<aAA^OV|!~XO~@>f8}>${E)Weft9CUvZ?_URLHjJXtgHJ`&c
z4PrL87e^Dr^^ATu|BMiDycPblht>C&%(;QdQnSd8XH#p<3%SDB09o=wI|0`LcQ)lK
z;89iG{I#@-Ot$k^jf;faV_OQ0_3<0zT0$U!#pn-)E%u)lc$)1mA7#^MebdA2jvc*M
zsT^}C{qRinEnM@r4vAr!TDQ6tvx&16-*LHtx$vHWPGOYN_~XZu`FJXHkMK;+EikZd
z^c5g#NZ*==o25vLnBQV}juAY$LVtAOztSk#bHd8gUZOBjHk1F<djY~<vLjyn&7ml~
za|LvD{J7jG)(c&ss94xe))wV>JKt1s1qev7(#|B@kxjFen_#=ARW+{p`G-|pTxC{|
zB7J{(<Y!L_6!_#pJ+v>|>3a3=YJJpA&4P>!`66v{6XK&9boi}LFDV&)A5$Z~VfA(u
z?{FxYGp?2@Y#cnP8eeHpzw4lM%Mt=(wT32o3kLbB-5PjjF=OH5^nD>`YBmB-eYegn
zqvLDM`wPnr<JhuIj*wq=6WAI{TQbeVV!ZEq(zokB)Mc0mewhPTAR6CjjK>}5_8V*e
zdS=Taq{|zv$VnJ3_uNkFbUcs7_-z0AQ~Y;Kh;<X4*Rl7%{8s&<;Nkc#y^}ubvUbKQ
zk}$!vQCLb4Aus!wd0=}?A;ydM{o`p)-&~5;W%I18!Bb<slKt5qJ(LH=8<h3Z(oW=s
z0&`cud2#V*uBa_i(*fIW=C%&^oGHtABJ-IM=d}%Y{n2ZDa4np1VQPhJ*k7wVYDJ^o
zqp+_61V&WoLg;i=fPbYy|4(gK$}%{K^G`>hi;(ZbL_9c4P~G*1)EnETp<Qkw+teDM
z1${Yx`l<B90oXuor~Q7v@NQeC{XK-cY+M6bHL0<(q@SIDUZ7dKA0DUoUIJT;&bnH}
ziwm+|n#i7q=*!|;j$KK_8atgF6NbfebzX%#j5N6}qHLYEPD<wOOH6W?G*Z@Qx#7SS
z0FOYu@-Xz*X_DS-_KxXrcz#!p<h7{(uPdM`T2xa@8;^~PYR1(5(a(=<3DV96Aoj$<
zGV{_V$AP!3_Oa0Oi9V$jviAq!e1y#@O%mVk@j!%k;HB;}*qFNoPOfCcInU#w)lb9&
zBg-oqyxFB!N<B9A&!y@<wn-C%IXsHb&C714FWBbVUi(UJ|9<i=S53;=;A|ehuFE(c
zLhJ|I6Nhile?zlRO4&H?`nk*Ro0)5c7zt=;F};BK>aWKQlvwTenqnrFu<WSMvxC`t
z$~~!+^-iTe{=pu}=h~&?a96`l(vRKwW*XddaoN1KAWh@p&A&a=|8(L0FO3i)cx?#<
z{^rXOd+R@J%(-1A&^d_m;x}|)b2B~)%>!Iy7CSGr^*Q3r2iPfHzmSxu`&0ewW*^!!
zsbwA{w>QsHk&XOqfY!0vy8>=#WjuIo+_-!Ngi_g<!)hn)hJ9csjY(=;o)Qar|8@{<
z(Mh#s{I1Scc20`}NPZ;bXP=*9hU{g*G=_+dkH7iQ-9IEf=}&KHmmIQvuz&MY{fi>$
zFPUhKmn<St54Y+HBVRS&O?n^p+eGJGiZ3I({o<TfOhZ`^nfcp%qxaW|9rWt>wd%M_
zO?q>4uSI{T)>VK0ml^xxz6S`a2yo=(bNn&bxnk8J0Sy1WWWzsH@K-*#=D#Usn``rQ
z|HjTgEd5s*PsOu7Lv!#kJC`D`t`QMQWgh%wqNFqzS|winU$%iCga5}k)<7ZVP#NY2
z|5Re4&f*nVK7U}w!lOr;1CWyf#yr`v9sfOc{|DVi>?n|v=Da&ps_SsW?XXX6!p-=@
zBO-$nwM$8yrwBbif?L7I;25_npxn6nKk|;PqeO`R<3%K^ei|Jft57aq4IOFxznJ&?
zujI-9Ma#dB%mPmiXL@MG_qrrdlw|$OjJWtso<98f$|9v5<L!mcE*6!EHd)^<f|edb
z`S`DDxBj1lx~zuA+G<kAvKs-C_c()xHXrxwa@nCO8-&P93OIu1Sa2FI*W;<G^|PWJ
zu7m(WK!BdGbgy*b3aB*gROoj+oRe5AlX+A4=SpRDCk}2WFB|G4nK}DtD9+emgFY-k
z#ow%ph*wUaDA!<IPqWca080CyrlGD;d9tcoBq-3*xefA==Nc|+9ze=B)FR@mQHyrZ
zGBNWa@%&=;fQC%puW_8pEc_R+s!+mvedSf!7qqN-^GGiBZi-`~I3|gaSZlVj$B_Ty
zYMB)1FcYkt-BiB3*)40|8gin*iR#`<N<Ec)Y*KtvB<s#Y03tkyCw2uyx6(LbAvWGB
z3vq;b0|Uo{^T{{Dgx=kX*n8#{pL!_0t;n9EVlbAzqOCA-pi8rzt)KWr_Fk+cUb;)_
zc(RC?#Nqcix&AOi)sJ&)ltEnN8Lm9nH2M1cX)!p+W3ai{<}^tl&rae{)WoE3Q7%r9
zoDv@OP=h`6XG#T%Y*i=Jmt=)D$L_?qX-fIlXZI4_%I8(l#wyaQqXF_V5y)1w?GO=s
zShNd@27|8vwj3;rTYJV4ISP!$@5Je28~k~mGW_M_ZZq)8t@-1YaQi~V6PrqP$5j;q
zVGc|Cx;)RLzvtA=>5-pTGQz}LBV>kxeGO|}eh?>enAk)<)1X}r<w{8GUCU@R-^C#_
z`tUGuA8mY<rV01kR#nxJKltt>-B2?J83V)FO$k^Q+BV4VORp;+AVv)%#vK6PJjW#k
z@z?|89U~5P&^9`_uw&?W6MIb_f*elAy8<$HhcE=*z;VN!{FpB$JLlR-c~(ZHP8?cc
z@9&Q4j*F~$EJf7eu4@p6aYSdV=!9YgN1lgF_8yUCRyKv)7Wt-s?djmJYj^~#aH>qO
zV|Gq$I$yC(nZTR|`4=|UH!nFkVmfkhP#<6aU91?|4za?v80^jvKi;Mft<&gIHQ)02
z!s<IuGW6h@qP#-HYsbo-^6NyZ%ml1~eB3<-4M}Os+WM#WLas+?&P;Cv^1i+ED@Tfh
z>Y=kw1n@Jv>GO#{I>8HtuQq-K)CM=~Bt3~NA=dLPfbh0+H+;hsPuc+4*^S>!OpMC7
z2y)d2+r49nVCB@fj8F>F=@MT~39a;J-ZH^%*-fHY7W?+rKa+Qf4#i2{)XP<;>vj^j
z?DyyH1?sORNj_GMhWkC_&R=_7iymqQ#fwH>b3<^j2dOw)ko$AYz>1gq{cp4+@mN(y
zCF@Fe^lXXjMYlx?orw2iNYSAyf?+act@6jE)!u~Fp}$6n`})X)`Ix2KJKUvq)XD*N
z#NQskd+wf2HgER}A_%zga-B0X9@_5l+mj@w5h!^wa<Ai;xos<gv%kO{Y~e>Vl~+LP
zHdgGv_<-jhQEm^n_UNxo*8g1@8Atav3=?Jeza49(1g{EqF)6Uabjt&|&J#PCN|r1;
z6{IZ}?V@#-qz3Iboj@3AR7{Ua(25}K&Zvp)sPwd0Tcgy_O{PP@eEK`{@kTY=b~_&h
z?w%*>EWZ6SiRZb}l>NYs-reo0!}6jxKd$?U>eN<LhQFw(jd~Fw&O}-yL<sn9mihk}
zj*_blU16r!DN{C6CUAv3DC|4gq|k2xcCm7!z9#NVi!|G(r<3~AU>DxrD$clz&Q4Ab
zX6|oGET<G&-B^~<1FNcW%)OlHc|;V_)V%w8nIh+%e&X*OmlP^at1%>VL#zx~p5goy
zed0L8m;3#{9s+=b26qkc=BWW5YFEEy7)!g`5S`#IkoQYLEm=&vCRwzn05nGfWkG@|
zd>^B(fG}TS>*aD%fn#jT09Uei5U?L+vaa_^_K_t-7{MLso(V4hQqG7L&w|yp$D!Zr
zqhkuIzsix^Y&tEK4u9e!E&UyzfSs?u3}cVeN{r!7IL%0?Es>miZ)pGJp~UTnEJkVr
zN$qE9><EHk&}6m2_w$7yl+Uumw<%~{-Z+2uGwZ{$+=g%^Xixmx_CZNJ#Fv$pK(xy0
zEIdqsCgzFyB-zi;y^U{mNI00k56P-?#B_Ox;mzt=$iO>HR={XGi(i6!_N~91pb)FY
zizdHh4`J-7_>mbvXE0#Sjfx3-c3F@;oH=not0tNB#epEB@kVA;7$S=+qo>(qb+Xhi
z3&TC-&USmpMC+$-%w3z$_IItL9)iDd`HSH#j##jvbr29*IiKA2<qYz<`RgA~HGB#`
zIbRs2Wq$Wxc{?oF@9gq$>d5JRE4i-tw)^MC&+6Z4D$V8lwGslj)PDK7^ms=OgML0W
zpg{T{PObnd%<YDVE`ft3`0c}aS7nXJb&Z|eXE>V5B`>qYhQjI=ge&3ry`}jn{M1$#
z6syYVPM>bOW`)XX{yxeCXWA5{$n>RX$Il&zmhf;bxX#u_uANE`m}G*};ICuO@X+?y
zmfKgruzqvGG8~g&K?47IbL<MZp9h=T!=VWrf)wzn{>AfaQ<c6#S3tb=vfu@QF?{Po
zCS`gB=9LE;YR38HYw8uSCQ%yEgDSv;&zxSubgqExnE(a&<@-UkNO1YjE8wKrPt-!x
zzkAUWmqb6=yaH-=w^IPVrNwaX=8a%ftNHj*uWxaqnf7{A#c<H#&!d|yhviWr!L}<F
zGeHI;HFN^@UhaO^EsL&zobUI*&WfCKk3i##U|$34RUML_yZT&&X)gR;8@wR$&@H%d
z&_aXE8Q;=Iy$>>DCG83b@>RJ#FSGxgBo_Jp1|m^tNadL#R5zp1;-2qQR9fp>UpB8|
zP5OP8b>l9b^WK<wh*GaY+b}2;I|-$do6bhQN*4UVY;?@<SV-sTd*pIV%cABs+;ZrQ
zT0G-|1ae8w<^>IsV|FRQKv4xTn0uer;}=})Mu7+M2^YO^{6zNkX1+<NW$PK$am?S3
z2Ye>1VegG9ktd-(#dh&$`ucI7UOp3Xmc^S(2#_DJ8RHI-4dV)djT%}#j*iYs{18c{
zv_rh72;o6Llk5hE(ZYWB9Ckln7La)pF(X?*Ml114jX{UND#UF#Av!d|V1awI8VbRf
z#pr(HaW|1|)KU&J)!+=Iv%(K>8%G(#=?cE0)M9*Pk&PYMd=!*@OI#a}#*NfuglVGL
zTzUyYJZq(u!Iukp5LRAoe+69A5Fd!lu=e5%6e+EX3^Zi{4omI3Qp^x-d*IHBhY0<e
z39kAih(Lg0-#3tlkax7GS!++4&SR8y_dKUMa{l9Z!HB>ruT#Jet^g9QAx=&_C1xz`
zV>4e>w4y^k!vgR@=j+)kEm&U&g1IYA)NqkhY|pY2!qOz@xgwTyJxZ^P47B_yW8ggw
zUVR&RNi4sG|Me@n4)xH2i92Lt_^6`jTB&$L+w2H2xyI!Ns80!_;wL+m5UXCE-U6X6
z#*U|pwCJSp$$AEV9!cnX4?(Z`>+L`pj&!K)v9uSEcj63VHPC9#vt3stTafhigmhP_
zOt3t<5c%1#8T~#7UEb!}Y3=_z2-ZK#AYpz#*OQ!RQg?i{%y3F_wee}gWs#h?(-$?0
zHXoS2idCQV^lkqq>7u$A>Q<CPOwVpBB|4*vB_z{9Adg1n)=^cLAt6TqqFml(_W`=D
zNG%IH+VHAW%!yy0;oA19PE-5M#4zTAZU91u<$<xEcF?pheRNowj104M$;-wxwm3C?
z+csnD%m{0xJUBX_v^h^qEWzHWR~L9GkXOdC{Ix)!4gxza=SL@^XxkQUBC99a*^t^T
zF<+X(c0Jj0slNhiQQt)eUZX`<rUKpwfK?-Q+bXn+4LhlP#%t;*M>8Q!<f^~ctmSR+
z3obwEJz0Uq=+fQci=`t6<Th7Y63RcrUqR1!EvTTog;&7HkcGx0P?tFk+D)*_Da@bu
zYv^NbxtSrc0Z!$Y^B1GSQJrk|O*kjTxB*wFQVYQVP>gFKhQ6wP+}8^wGNe<VG<xFW
z^C58M_MMT|w8L1n7QWic#3ntj!mXLf;F2N!hUhf64Sd<8*qk*gjm`C!(9_{eN}8b)
zuga$J{IL>@U%t*4<-GN}2;lH$ARd4AgJ^Zu$xw{Q{M03d^A)f}f=iJn3dCLk@1Q5O
zDPouN`Ckobr<>Lmak@mmS?wfr>I(R(Kxfw<&EPX!t4qEVe+9t6Xuu8qh={R$F)+bb
z9`&@Lt6|M^dCz>Y&%~1X`lp9(->v}9(MD)g#-T>jlPkbn&<I}DoOWE-(R|lWe`jZ9
z<EZAlBu#ZNRVv5z@V1#EXQ{RG$%3AtGp+?*Ttb5}6U=h?sR@5Mh7QJXe%i$%-+bI>
z_gJ}7-H<AfCybSPK4#MxUJdCiCxN*w=Pl%9+a#TM7RvEd<_L2zw(scJ{RY2-gE1~E
zL*#Q7Fs&|)yq^e<;U;ZAIl^4xceR|4!Q?seuwpo^|8*Sv)@e?cp^Jisk3Ak^xUU*u
z$4L7!{OsYBqLjiB-Yy)}!UPjFMq6j@$dtXcck_`W%uD>#ZdqS1CEn#fwm^hj4EB}I
z(OI6qO~zc%XtQSXgYC!HSZVI&moUcV(5@(_4sXY-ELbI<yu^ygut}QchWz_)kq%Gq
zNQ1J<d{#kkn{1cm?{utopJTl9s>ePAGo09-NSSWZ@ZiU20_Tg%a0rDQ7j%c(7H2pf
zH|4A(cVw``k`XdVrJ1mZ$sz5GzQboe)2qC-T>5=QR3hy$pl0vZjw|5ob~+7i@>B}2
z;hGse>32t(m~&e4%I9;bqN!zi`QYx1h0wK?00k=)q$4Jbopoei+|0h`+&wSqTYr2@
zf#=*3K;Xrd4zP~GJ3>)YVQ(%>s#c<|Kdy?q|5bhJlqQS7{jsIIU368M-QsmQ^*WXb
zj5!3U_hI>7|FUIdD!J8?hA;>T`a0-ejcf@jKwBzTkBZs5I9a}K)b&o7tB~)N=DHXL
z^?F<Z6ipURD@w~YzJ9swMct$g({~Hrd^2UegFRpdtZmlFT_U!wfMr#TF77Gm7`g(8
zJS{Me4bT_}xtRPt<De%a?)GMo?z-2nA{J~hHO()3G~@k^q?tGi_fk`ekyWei3mIP|
zLln6AW-Lrv;|250@^nsLv9&I4N4G9;up^Ll$`f+pOAewMQwN;Q<9nDP3D}w~zGLFF
z|0^Js5AXY)A%s?e73Q?^vUIT5VP#5y<ei052<#*Ind{qeS1GwjROD3d{cV1Gt_N&P
zFAZ@a&3ZW-TC_WZxJk`z6-;D4XR+-j9l!mmKeAr>=5vgF!YP_lK86dwumMe5GLM$+
zoY6)3PM*rZKJ>Ml-<R!Mrrmf<?STXr%DpKEqjW-QWFf4GWgSX%#P2Saa5yg}4OP@T
zFJih0&>%7~l=7V`r(F=n^3EZ!cFs1VjKgF-tzT9Skwz&h^5&AWDQnO;r4U$$53N^j
zk0oMsvt0xqxU=a$yJxON-W>;RpyS4uOW&M&8$gyWgoc7Un;8Vuj8Kin;_cLD5=oH*
zobDXHs~p~`{#K;hNALj8^PbO6Uu;{c*1lDKQ{sAzAd79g0%jk!%oi0zHaVaedm88(
z#7{eU*=JVKd1GZ7s=xZC%}V6$IhUdDn{<%L7S$uZMtM8BHYo|1SPkmCkq;%iWDixJ
zx_w`j`VJazxik41Be2XjpraRU@?KU-Ej5nkl}l_B5eIwofNwae6_HMZii%DK2a6fK
znRdwY{FFY)<QWLyuUV#tmioP4p?$UhX-&|t)*6uKu+6Va;#i(gT_LR_Kn<UkN4DI7
z{b*<BnvjEX2RF%yeU_DGE7yDdJh*_ow}<n^bv%<otY9hNh|xtSFU8QH$+y`d!*}`J
z<xxUuzcM4A>kRznOWNEvNs;U>r_xx8B2z;$2H!n)uc$PU*C-R)=2l(?@E+(4V-{hQ
za|W^~H3Y6QLQZL!<md_@h3Q;~KDFhvs>cr229aLpm*iwvGbq_I7e)?s3<alCw>FKw
z)SoC>5mCr+pz^XZ98ZdkZ;X-m62tr5loAt;k<BzmtHb<P)n(S=s>6m&xlT!$_U;?J
z5Kf3mbcGS594)?UxrL5eh8;OknxN_TjzSt$zSSpdHqbbD;;mWFF>-iv0(d_S#8Fx+
z(V;6tVg3wL)w_PmeDb{6qFkuVxWIv?J-jeE@3;Y<@*o9Sm}Pf5SH+0fF2yUp#>&1Z
z7jn&jV18G9WpCa+MfMtVFBGE3xFFCMLbwrR&o-Fj2}{!DEA-2Yo4uj70KE;IV)%x)
z%fi;RiYlHYa#@22to#YEZ}5=Nt_v_r@peN&%zA?)OITcH#5&eWZ#E;=g-JHF4_7BL
z%P3=Tvf0`DCQdT$<dT2ptWomJXL+{KQPm}Nm2=kuipA9hp)Qwo&ez>AtL3>z;4@&m
z^>`104dlV!v_QWMx&rRSV4pxQ+}+?*EqJpqz7Cm?daspbp<3Pb&dapwfg4g!K?72~
z^3Kj8_PaA@o+l+J^7EnaY72sGqa>9#NrvN>XKQCKbiRO~NMS`5XvJiZ8%0Qcx%T1g
zT%~b`@M-PKkCzJ1p!gQ*lZ<bhWI^vRxAL}<zVTWUt2zrepEvsRYhdg>{#h^ipX(<7
z7ynmKxy&@QX@wj)=&y;)f-B}3_ueIdKv+BrQ(l5=#2kmj<<~Fs{ULMc#!f0bZ#iv=
z;yg#ETSUe-pRWfTKMzpES#<ZC_0XUUw(1K3mRTiKefs1|KQU(u&mfx`pbl^eEm@}?
zOJ!0PtZ=va2(qAY&Wd@4>9-o6Rzab>hOcIEMeaiv(ir=8@fc9|l1?-WL%5U2sSCpI
z71sXCV^10Oz?M&wEs`)YtxEg1Woobs*I478!Gpaq`~p|LHjGq5YY#N)D>|gF%jl=k
z+&#InX<j-P`kB1OM$3ZwUp)8=t+p0tH~l%UfI6p3-iQjQ)uqVH)H19&;ufPpvTJpv
zr$d2VOlCV~sMG;F+In~&gDYipTs(h5uX~w<3%_D<LDp11u2O<Ilk5tp4j!+~0EutD
z{ba60ek=5)ZLH*m2$ZGcGNBqR;LrH^%NXr~dEa53TOBEA81Fpg2pVqArofM6ugbom
zD=<!Q-*Cw*;@)k)x;s;YH~@dBt5^{K3b29c^xW1W?*1Y7_<I3WZ$|^vStK)POnLzD
z(<q!)DzvTBm$c!ks}IkL1FPy2P@l1!#hgesesW)=nO?_))<wL4z7wf{q%&dlwxD$W
zl4yGkWcRo^+xYVMvCqd8S(n$I{PzPTtKO`R6+*+AaOqyP;Jac&ly(Xn%~3frIX6v+
zR)uf-zxJts1_mtzNL!pW#a!UhgR?T{=`HH$R&hiC+~&O#?GOB3TTzuaZ($38!gLXo
z-pF}BHM>(!boY|VNKdx!J=?Ek2V7PIV^W8xMrbeqvyQ+q5cv$}m0bv4az`|OK1sxd
zy1XmU3!L2^{aCyyPuA6n{OCV&{;J`I(5;O^>S&3bFPS&37wsZ1rbaTV%Q0~}xIEl!
z9ri=FB2z}Ce}<DivTo0X=R#$Dt(?y?Bl&M^&M-cnf84Y9bj-0t@R5eSJMFzY?2#ox
zYP%M^*g*(AE)POHZEB;y9k+M1y?&umI->eADyFYbanqKsofW@P-hNbp*Vh%fpRxR-
z(SmHTET?!*)iNrJoQAC}YCWu-^DFBn8xMZ)Wz=S(@4Xc*tN9bp+Gu52M*|LjDNBo|
zQ$g6?bearxWX(ydA9b|0(Fc>_YvA}#ot7ivQYXM^S>1|t4;$n1Ni$Jmk>L&_E<~tP
zDh;G3HJXBm;-_Y1g0-@<D!-XD%KCh4ClRy188m)P^wQ1_I2$|paRtOOhP3W(I9=k|
zPb)CNdqt7Smw^J%BTa)S>@@uF8glB=ci;-h6mWnsu6n)|D<^=#A2v_C(rT2KSy-@Q
zNDwkmPc*c;3+5OGeVdewg04NBYTfOCUouSK`VKm<Xxv0Ag<k<kQYY-q2H6;zZGV2W
zZK+)Y-JXl(t;9aQ>eugLBf;c&nP8^N_YE)v7FRcfOFM{A_<CK4!Y(>nK)47X21}zg
zRdeAq-4yjz?sKhIW6PFQ&+=q*BbM=?v-O-DcZUH_;3zS?oti_ZcYnm<6;3R6hc3C}
zn>TUyLGOEOmExJLH&NED?(L|&pOd2;GtCc;6!-bsn{aMD_gleBxH)`Vt{QI(S_2Az
zE^{DP&@tMf9))>Xzm)|=O^!KtQoo+e@0s5y*kR{HkE}q+(cXua>(ff>y*yZ^_RF=s
zpA~p&_4t!5i!aIZ46o%r3mTr{TZRBIxArN?I(D|Pq*0#^xs8QEXObJLD;|;QL9@e;
zh$OqUMW$5I+-2J{Z<Mdgnu%ve2u~}nqRKc=WUqiYGDDo3oB-eYvOD504)5hbsa?W)
zF~zut`%hM)2A7qP#70C1$nVsH-=`#1<oe1?9yCdOuHnh*qZ0X^8!(ttug4XW=YB*T
z7c?56Ab<pgxI;*N-I0Dt)r%;hRDrp|)Q>-MFeN^9tQS*7zGmpe0`xruMNYxn>iHKd
zhL4x}+wVu4NU#2Q;q5-~5<QE3xY%iN+$VT>P4m-QjM-)R-pu!sI9vVxs;_cHhK7*d
zwMYIT0XLA~ZVJh61?e$4X)ok~##jpKBQu-cCqVV}o^a%*qs(ck;7rlz<^v02SXsN;
zr`&}-ZZ?vc?NpC`D|lORBzmfB(Ppt)yY%GkwS}Fe$qED=o>l^<>Jmjxyz|n^XB2){
z%WV`<cqT^dn~f7jTdeVzH`{qh;jovQGvMj7dRB^@;bD;d1<y~@`5AOhx96(pPI91-
zOMvG3T+EA6ODw+m6<{IfKj1Hzpq_rPyco`DBCpx!MUm+!d8mchSNefJM;o1ZBoFw`
zF7-UUENOZ$s9e;^%asa9s69Zvi?i^bWM{pQEh$`6)?f{kt8?md;FwB-MIiK0QwRu+
z5n5^e3eZm5+xSGHXJ}y8DEzD50#P8{zJtAP<kx!gG5WesU1Qy_fOh>Sgb1-fcdcLL
z8U)eerpQv|tBt}I;s{h=L#793FV=o`bB>52g7ih3y$!boxD&T`O(2vC9O%Z!=%N;C
zlz&{V(StbIAvz)7E#Kzd-=OE;Ik4ir+b0SNNQ>?v26U)0GI%nN{&--{<rAkv)b%+K
zrH~(;zk3~W>;WpI7=^-^tXNRY+v?7BbSPWi6uXJX5}EEb1Ep02Uw&#`%f~2P0c8bO
zfSSA?r-h_{|Kj0eEC+n*9xe(`Z|+=SZ%SuIs@4>LZxgSn`grKlpe&JzG5nGVCcI4k
zyq7^M=itX|!P`k4l+h~8$==Kx246PEP6)7I4<1g1t!CbYW9O><`=-1NLVRUQcl|lW
zOU1U_$v)K7+1C?1$K>(PYV2aa`s+>lss$}?MmH&C@6@r&ja8q$nRwnXvm>zY$#;Xi
z>uq%imNsp^;yZ}%LKkrvgaWm<a7R%d^=O_7X;y*}o0q(c)$^DI9#q{s40E2kQSQqm
z#T;iaAE7dQ{t;qpVppa$Qh|Br3b6LGG)Cfbozohy;QkS<^*Yus(wlD&%&C#dT{5>D
zQN_lnjBs~ZHVNWW@l;ROlKcDYrq8Ud0P<%+2NA$|r*>87!?C@R(x<nYo3zO7WbZ`c
zosq6$B^wGlEa;Lrz!g%Mhd`+VAA+4>;+Xi?n{p~!&*XS(D<RXI#%DCnc)e`|mw08g
zCpnX=qb+pi#*Eke9@^~XupyabA^EhBuuL#1S_M_1zy>qcW|NaWIf-B2H;31G8Sxi=
zn_aJa$+k1pEf|ecw`piq2QZd;=&M(JZaRFnY0g2qXHmfW(tZ(YJwPaS@UDTXDe!js
zI{Q|9R*Mm9r7L`z&j?;aSVHZ4tJx+i{Q4eMWD($h40v$|;9Lj`^nDCxz6~51=~w<s
z_ao+D4=f1pf9P?h#^$GutMp*P3d1|TkNo0?e{HSu{~?_LpOxEN<YnsVMi`%4EXkKP
zWst~;9S}U&___9hAS11K2>x1~<9&xQNwn0+z^>l}5cl;8IR1kFQ-Sdc_;*%_7p1m@
zO~XA8GY<>>{9GL9E1Al&BR7W>U9;ReSQ8S0V*QFs(9iPxS<u>sRGsS81ni6Tn%iuN
zbXBqc219#+dvxBNuuOG+<8Kmv&Q`M@S$Zo{7<~LbCSC4SAi&j&A0(I3hNegPcVoB}
z5>_rMY~k5$>PfDiI;0+(@w*+Xg}b|E$x&Ew^u3NBT(;igCdc0yCb<-tTteUtO~-2)
zc4tMqo$Pne^(z!6T{GOFH>&FrM1-PF)HtL2`u#oSzXR4~qKa(r;Me<DDa^BEv+>e2
zImv=b+Uo&NZ>HvQSMa-<loOmsoLYp#$=-k=_wengYf6u}2pa1Eh5o~+=@1#~0>?8H
z_#K2vpap`9FFLqa=83eW9ZdJriND86uYIy~)WA&>`^S}*NaRptlk5tN!3<<r)jTJ<
zp8GKt-V{pS8kR}ncc)`kbIhTH|E8gXK#pN{l-eH%%h`gh$&Wigj@!x8-pR?`%hR=X
zo1*O&=~6P+HA*4j<D1Q_20!4{tV637QS8@|4pqVta|qUUoq1bzYo`vO`~Z0<g5@L+
zegR$B&Aa|!{H_>!UwO8#t-Jn1q$GheoP4D|2DB@an^8G{%|B1vwX=76^TWg1O(c#F
zxVPyEcuyAW;mSz0LCZllu?=eOg<{V4NK>h4gkDJ<o#Cq8+lq+a7>R+WjqR_$v&y)*
zqp@bd^lVf(0W{}?SXN|Cr<@Xh-#Q&+G`@eVHGRw#z3DKX(5NXi=exEUQ|g{kg>^p&
zUVw~?$Geklai^=7-5vVO<**^MCTbN_9d}el+tzw$YrJ{sifjECE`m_ndt%uy;~8~F
z9j$7bqg^>upiT9S?TB$J98z?Yi$u7Pb;{XtJ?1q~d(_u<Ou(Mqa%&2swM>ka>@AWA
zx5F}2xRDwe8FkUDzM!9bw5)zt%j%@9JSRWF0tNncSoL#W`);dO4u6Xs-iBuPJ&{Pu
zjpvC>`}UgyqMaU#6#ham5eokep3v#|{j+swW7J|lq1%>F&hC&m@T+y%5pxTdY;y-=
zN%x>V&95L-zS5Jv;@K_L=`7V%@zd5-6%*Q(V#4V}KYuQ({fL%Y`DPJ=6-9bfARC`l
zjEFmkY@DmTL(q{+#@>4Dg2)t0$02^p2kdBYhjDF{XHlnf!&4!`nj~?vBEz5%Rq3-B
zZ(!M9B!dtVp+JFi#N4pdV>t(*d)b}6hr*k_3$KQf_mAginrF3J&|3KmDqi{XFox0n
zn-AQSZsoe(-~^FNv(oTwT12#xRqGIA?j!Sq^G$a|jP1&427iB#y1D=H#*8u`bqq>5
zK0v>6BErppRu}XX3&DyjvajRNXwZ+=S~vy{dxEVuy#l0mx1l(&{5wb+=ib>|0Y}sp
z68`^XxQ&70;4d~wY;z9oTq?lb#v&7&6oHdM{|kZF%eHIs$Q&v5%V6L73%PAia~X6+
z5qiwPtk-7GS@b(?Z)}9Ta+&`}<6lsFI3pRDavJt(($v(BockMa6b$YIPxQc-h?+r5
z7Vlb@!D#maV{P^y7jY+zt-73V4>L}m-oeu0ZT<gIwNBp->s;mzYBKPc@{aycVP4OF
z$`JY7le_S>Gh!~#zBxNU&PnGo6+>q4w$ay+H*+w*KID4sQbFZO3=R&xKXSB3_At6r
zttxPE5ZXw`L$4&SGk!j?!-LP+-NH78F<9PhW;9fb@7Z<w_+?#VW*NabucEeeK(^gN
zHM}}1X1EsJCuyx0_aoMD?a<ya0?=G;azj>iq>F!Jlo`yZJpvRa95<wi*}_>RS13H6
zJ_5zVJWt?Ai71a`8WVFn$Iu&28<d|;_4GhnO7y$J*(xl#cJy@5pOOQ-<By2IkxiW0
zupbXDz+wEB4dEKXb@Q;8qOWUk^W)71i(jZEAGZ7Ra^Ar!RwA==56Pd!oGNl}X_P<9
z4v>;SKUsdVN(QIN``lTmh!Rs~CcU3-OrS{R4Cvw0r(NL6T6^hGkjqB+_5;1`ntqHH
zvK(`Jfn?kldNTC&vI-XwsUJ4Yf4*gKQO`U$=YF8bu7Xr+Yhg##wL3^as0_mwu78fX
zWA!5bzLr}}sMLux`^3171t3R}XHhzyROkJkkHPTZ6`19mA`?(N9cr;JMH?**p%pPV
zC_{0$aNU_}Q#w3h-P&bPV&7@g%mE-S3M-1BfBFm%0?Mu<;G`Jg)hKf|tuBUBk$DxM
z?g{I!pl7e1x1G9~P0-=d=TgracM@=fWht>-b8BOQ^yT5&DAs-bmgK<*{gYT&2y<Ta
z#HUs2z9;d3G{QWw2L~c3z(81ECpc1IcZw$y%!LDWe34+x8O{e@GMRE&UL@LPXGi9y
zvyI(M6zgSB;TdIkEp?;Ca>#pY>pve3!l2FJOC}>YPv>+Ae1^m!Lv3~CCOC#e2r;*j
z1wrl=MO*rH)aJx90z<M=Ge3twTX%c}L(Ew#Q6s1~RHhWRx$#eKe2UnA^U^7~KvjHp
z<ZmtB%-9>F0^*kpzct78(g*iN>559L6Xl(K^bG<q^f_pV_P4LYH<S3hHblS0D{s00
z?h0)NT7Pd0ft5#MukW4@^;}AJGdxB(<=-Ri8so~;X8yVAT1a5!Nbssof!z151#q{}
z6YkLvQsE~W($GE>&y>jeJK^;6+-h3ALYfNT#?Fh3Dl0wNyZ4aV5SFARQC&Ad@I3b&
z>nh|EGwV_+wW&qhVlc^)kJ|Wgl2$p*{57rw*)CF@`6qMfB#xIdlzxu!0dg@b<?U!W
zvlTNw^U<=0X<ra7keT(E5ZH&9G#y(;R!Bi+xhZa<I^mSY#&C!EEg=)z^TGVk{w8Y2
z+lsk-Lf^lGQm=`)dZ+wu4S<pzU<f6eeae3j_T+i-L`c7IUFvAs?*-OB9)eF*d?@Gu
zgLml7Wwj{rR)!s!kf>FY_Y@Kvim2J^xIo*3_<8`TOXMbC(D?Wf=o5w!ZN%95QJm!E
zIAzwCMF$#9&YwlasXWZwg<SD|QY-oBy(PaV2!CoV3YJTU%|Hw2_k?%fIH>6AUR!h4
zdY^EP6x!hTIr@{@cozg<WS0jXZyeqljcwGh+R4{Q#pk5yn`0Q1F9u_oKesO)(V{?D
z0$ga&Q#^+Y0myNZ_7Yd)M9_r$>Ma#6D_Ugn4<!%zo02zQR{^|q0u>0+*6j;Cl~dI}
z=3kin(7vEqv6-ekvXUlal4m;IHG;Az(Czv|QDvbX;>1KjL^o3?_0^xIHVVEH2#~cT
zYhRnH)m5Z6F+n&u&^Hkm(y3F^t6mEbB6PU|7@Kamt;P@xWgD4xI*7$1IO3ma1{BEG
zmrde;M{SGcT3pJINNk8?Tc$+fy!$;TSzeNQ&TCQYjYo#gh&tonTJY(qZtuK)&TfS}
zH9oLGxF^>XC0$%vV@JS4=%xFy_>D~B*ACg0m31|5JvvGGZCV1z@ytOegJr=;jO}V;
z4k(Ni_DYxaJzM|2K5e8($-O`&=3c}j(QZFN6#%vv2(@k{*`i?UDc`V9E*^AL{i$?J
z9o?!bVBYCOppVRbq#D;1a3`A_kOvAzhe6V-U;>zDkHm@~)wZ))WhB?n**oA9)8_#v
zgM6J$x)&mMzMryFPo;B1#GEYzRd=AU)R>F4snHLQK(~;^O5+&GB?npUsq&Z2<E!Lj
zy<7u#NnR)y$}N$DJHEA&Vs5vLwusoVi%z)n`6=^awmBXJvva!z2)PP<F;W`J!^s}r
zvIK<4UE5+M?=-B*tLS#|bAq=eIr!{0tVNY?o?7fj7i5^8BciK$<usP|OjH7`0<~s6
zf~{O)b5A++kLmC>uwb<9Wdbt&pxry|aIlU~M!bvE!0Dp~0sb6Z^Y9xk1{0(fs5`Ml
z55Oz54upu;W>b$3PaTF@yr;%+b>{v-`AS(^#vX~b2~IDa*s5PPB&v{-oUSx|UQz4d
z%ee0QWK8_WG=GL-#5_ksovoEdhOOJ-o+%s$I8trd&+y`;83))YVQ?*(nR1J0W?aoW
zz4`l0TZR<M7~8z<ijO6wmek6@@9A}itq$$5IEBfIu1EehMd@_Kkl1@as+Ewj?3!Y?
zYBX8@DBXsTGpHE(yFPNkDxlqPu9Suq_GX}$alcIW`h$^G57AWji|mFUhBPxXOgD#;
zX$VYHQ}NzD)>UK`3~OP4S$6sJ#&od^G<OeXMlqX>)g@iCwA<T`(Ou-8Y|I3h*!~x%
z<o~fm!e8h8-wuEL#p{1tj^i&``%BjTBgz2ZYRnAZsz-(Aiv17;+IZhV^5IyX4fnS?
z*Z4Wj0*c?1dP+=ytFwk&fWjLO>lghR-0|9gf8ndhsr6Sj1Gsk^|BZJe7Ku1xypu+M
zW~72UWW*v#-ym6n(Z$SgPTRqadY*N-UhUovlPiUv%Y~5dt5y`-6@a;E&X-o^el6x!
z;x0JlQI4A`dLQ6|C@O6-1Lykx=p#GxSsB`6D}z}hywk&%FS#Gsn}814wI<>ga2>^*
z_Fu?`@p@OKH}UT+h3?^(pl1<DEo@Uq%?h`gHRQre5{GlLMbp~nD@hBt7a;+lQL#TQ
z*am&AG_AINKFX+;@?0Ah)sL^MFG8yYz2oI#xL)hcuDb(Ptd!q2TR``akkyUi)mCO)
zb_%*eakq!@TX=45N|#Kq72!uIkJo`RRI0>VZn1br+VqDpUC<fpE*8Nk_b9%5qhc8q
zhNitXEGRB7wwxCw(+0YMN^OV-26{-c)!I@Nes|TDG4_)sSv8TpSINCEX00&X`PQfd
zg(KPPm+-eV|Hu~Knht~ic>2w1)a@LHY?~S7?cx5B`+_uTAA+TBMKnaKa|14^R;%2@
zI0qvmDkZrq%IERO@qonO!Va0GCtdCHL{g%sFQ4@zq}@4!cF_So-GA0qqDO61T<D|2
zc9PXpN8Y>}YwF-IsgBE={#kDX3=O*+Og29`@j?&5QFkt$j=7ggOwOE43cIjQetze`
zuAF$ME4D3lQ;Oq=X$^mN>u>9L6z>?=7Hm%$#21Tspyc`)3rR_&PFDi>buE||uoM3L
z_Vp=4D#JOocYC-#-ST|eJHr$XaI>l7D+D$-o_M0my->0bCy9zJ9qi%xO(rR`&e8W1
z{O5?c0wpB8aD~I>*%5;%_F{&h<-I2c-=fxnbZV0J@5fyKWU1Umj!Gp3MzUK(>Mx7s
zOvlupjZ_naRq&Lv_V+cvy{VLN@b-v)f3a$@`+7!r17+6Tyyq18IR$-%ymj1taTYF*
zv%>Ev5>RPur~VGW+d0xTVi%g?L{0_Qm^s=Q!+r^*FLf0Vo<y$7f5A78K;u9+yH;jK
zS+N*!@DR`~EQRY6s5#<p(+T^H(L}yG4?a%ysGF=KT5zHCt&<{WR^p0gBB$}9!`puI
zN9z8IVQSDLcpch_>mga#jWs4~OnezaJhv=|kRqS&i_WEdC>)m>KsgzKZp5;b_rQ;c
zk*AkupZ|1E)a*i7S`f%P2!ylkvxP<B+K7q8Q(Gx(<$kqWcu~<Hu4h)(9Qe~!D!yj8
zN-w=topS=TCl-T46A|`1<ss($I;=284HTWnfc?qOpVj<s6(iG3_;Zd=g0`bS;Emyy
zDfqY5sC=g`bbMd7Aoozn<X~<ehx>Ol{JfjQltR{{<iB$yY$>vC0uW7Ii=^Z_PsXXo
z^A@JsZf<;z@wTar^~d~Pw~JGN%O}~bd^m4&LWb|KGUC7ak+-i~{f;_ktePEgshiy*
zb(y9hu`AKSU4>kXI4Bs$Ew!T(NcKKdP@t*Axm9@mLqZ_<K_D>@bRz#>G4#J?@c;ME
z{`#T}IQ{p+RpUTIMS&hHW;fj_1cDd?G`wuCagCIpfE3R-Nz7pa_DuZTX<3lcHl2RB
zv^tGpT40akk%Asx>e97-HGQeI$!@^k(2Ev6#0x!0Q2WPEGH|N$fAZ@GuzBHTX16g;
z`tv+<+Y4owfhMl3@Y35)b?$npOrOaQlXrl|^`1N=320r`S<PAJULXLKFRn+kbo&dU
z)j-VSFCeAP6LY?}#AAH(GUS2;7tLlp0&T7TO+owy_kd?OyVz+_8+lnaGWtoAleb>j
z3O-izVvXvGD$69N67F64JxGv2h|AAw`}8O%R3Nd<dE2%WNcU(Hu6PMGJ;9{8y|`e`
zNOwBC6-YHh%$MyyVXL|m|7N&jnehhE9T%<#{`ybmoYlsu$601B6UCyGH&<0-yeM)9
z>|Z4OaJxIluQ{y#K4HI~2--F^-!+O`HK;t%|2B3s{l>7bTk+O8Q9|xzj858~)<SN)
z24@cat1~l2f@mV}R4PTr+tzBLnpG#ZS+r-DjCmSUlbVqBi?Ywm)T#TFd$G#tAKQ<I
z0EjHcZlp#)bfqq0^t=zI-IM3y&al3so}2iImAajP`aC)_Vz1T80rDtgWxg4s?^ctR
z*p$EG_m0bFxG2z1V@oELc9_+pl?H0-L~WBUIr%0Eq%o?Jl;L(Oe|J1ygKZV3_l@WP
zbt$Ipg$BX55ccD8HrTigJICnJ7IuVIj@Ga^zayZ6lI=4`^gBzYgLFW_5#63^#{%yb
zq(k^v%jw+Bagiaq_gnQ)loR|&iKv<OM1Hh<aV4^@*H-f9-s@bWe7<01G%1PvASI~c
zNPNs!=`s(ak&T`*8m-pmR=c1N?oW!CitFWyw0u$0RT3fe=U&V#+Mv-5h?ph|iQp9q
z$%pTn>r+Tx>y6o3G%?B6y2V9WGReEOfkx;IzAuPh7Bp!Wz{>S1M&3j*zZP(0Ux3yi
z_~X+;ZMjM!q&ZS816+oT_u<w<(JY8A9l8aZhLf~{_M5ZUKfbIqP`vh)&`?yu?c1}+
zH)42)PSIixTAHZn8tX69rY7@hp0~H=DHZwE6_x^tIRlArHdmeSk2*~uQYP~qZks8{
z3-6_Pm<pzhcJ;Mv5+jhO^w5Y_J002?^n2a9k!nAB{u`gN!^{wTKZ2!x(8sB25Fn)n
zrFYFkFXVh*Pqc8N-#;3u*^}uZS+0?E(Y9h^v}@qw;}d0FZzo_!L;$Dle{%8b`xV)C
z-ggV!3U-o2XU%7&qYcxf^&wg8@0`_cwuE1XAALSXafgYkReyT3Y$MPz{NtU-f&NdD
z9YtoT?YLB{S0SYr`mnSbzJ%PZ8Gdj7-)jQnu1#68j(i*hmxSI0zfh=*$LN?Zb`6g8
z`O~xW#`jJ>Bz+gjxs5ZWIKwPG=<G)%4V-H7t(s}!YT`^R=GY`vhgYi>d_Sb~nkzQf
z3B_LX4=^ah4lXj)2!^}uoEWbMr!}vql=R;E_%bRs^;xkfTifr{rJ!NUSg=FM5S__@
z=kol+nQ5&d@sf(3%vCo!4obfP+X*v$)SJD#hbJm`lZKo6GnZnE9zrz%OncK<`Ro17
z+NSDnAuEhFc~Pr3;yx@DvW0QdA6x%Y$y=D&Zs8lhEATR6aBt5t(13|X?8RgBanJGQ
z;;!k-gFU@fqQ+09Vo`Vk?<>M@g;>9QfXNE`aLpw$7Bp<|)t4GQ<g;o)f-csbs-A^8
zz9Clm@t(g(n^}ViU=F9a$Ulfy0et*$a4Xd>M|T|n_F3m^*npR3zL^q%Vu0DqZ9UU;
zA`SeMGv(EjV$^$6&B+<H(A9SphokEG9(qGS91p>23y~k8Pt30al=<%!k59!>#ytyD
zkH*a9_Z+^d+|+1w#HUfZPU$5d^A$jAuJr7atVQ@3ttLs_*+Rwkbp{g4Q~0@G(Ai3!
zm@}g&F6VQHjm_uo3*^OlgnNJ}4dfNMnYUA;Ggr4+`cVocDWH%VhZlw^Hb)m4jx*v?
zcR7`Rf@)n9S=@VljmP$nEkgWneI3j6dH>@*iDd-W#bEbC;Iuqx{jvySyCIpJi+*$a
z;qdSSb@t5xW=!!<EF;FY0VUaSK#fi?8oQx2Kw*{_J^mpY{-JVj&akituUDZN2`Jo<
zNCsjje_xa0$SSXDwUobm(z!gCBfnn#RmIDHRMFJ3|4~F<Xsr??*z1Y^gn}g3uhAwc
z&@+ckF8G@9`Sm4Gmwjuv0ysl=yp*H1rKVou@G|43q!c-6pa!cD6W1ZGiDk?tyS_CH
z%VR+V`%k>Rc|6p8`}aShMs|tFnk5P;`@U4RBvDyoDtoer!7wQMPAFMJLXv$q_6gay
z?97lX!;E!e%=vqFe(&r1y{_xNfA{ZxJnsAc{o&Ec$!Fg4p3m|;UdQV=4q;;gr3GhM
z`E8D;^tJwMG!!afruNf8ig!dYqsNcnW4yml6I}thtb%HW@jYZ~S%efq2d^-`ErzXX
z4q3qon?~+TjT_eT2_{gxw+TZ*^~TVz5A_C7Sz@k~L`o9^op<}Xp0s}AjaGYKHkrY%
z*UlcCrx+iGu>xWL5pjVWY{dz*c;K=k?)z(&b(cg2QuJeMN0!JVkuALwOYv{)W?A5?
zN|LR(0rzhIA!v<B0(sorcB=^Qf8SDZ+9JF*OY|MvjimA-q4zvN9sw0j?Z~C#CZL?E
zqv(-~${_ggYsn>QV6Q0qE^;ebWlf2H5uQ!zew{M^`C-Z+-#JMN5b1cFq+`kO9?&K5
z;*o1-3)=*G(>G@tX=lmVsF#dXrjQbJlTN<w(d=G5maXpL{wn@+NeFQYTOY_xWOKF~
z8BuENPH|q;)urYRzVx0e;X0gs@NQfc#qYTT@Xl6avJczJhJX;O*D~BfM|A0XRFgD;
zvrg5D(g=IxoA31vmtrfwF7>J(0c<yb<B&(E;s3emo8(KxDVgxrIbpPH`a`74_~m0k
z<DL2&Xc5UXcHL(mw|7D`mS5&DI)7*p1%h=k0d;b!38;cWyFjpBKWbd6jY{iop&mh~
z*K|+f%2OTmwtoN+8+1<rzSHzSu@_3AMZVlMx<yP(S8e)STwwK)6f6S?K9=P#@LF;^
zP#zT_7<vT(U#XQmSw$SjN0VO$n}<u55*;8b%CYz3)a<&kyOrg=)(eSqK8p#^UO6V{
z({&FW=<320IB&8fkqejOD+m9AC<CWSi+*6?usUxluYLA^Y<Rlrpc;wrH(32mNdZ~N
z-&o<39fxD^+3j{40QGl0;5^0er~lZ>p1e)))&@BMV8M#^z+K(5i=VE^C2#iGJ6K_h
z82)D(rkQ3{X9Aam8SoOII`!=bpJI~y(=7Tx>w6&z`5GCbOj7$fXGB?_^y)nEpvJ+9
zy?&d(ir@TGXV!mk@&D0(DSn$6%^=a=VcYSmh7W5N2)h@KvA>6);N3m4G+ruw*;kM4
zCg%D)-uq^x3Atb%n^Lk_-&h+-$SmV8LJy_CI(EGaMBmWt6JY)#lzlISJ!qVe;f%&<
zEv+_WevjL^A(*7AoLy>^!#E~^YnCF1uptjmklpeB^xr56faqgtxMm4*i1FY5_p>?7
ze4dRh?f!ttwy(z-^d#$VN6YQjeiu-UNi1Co8~ws_$?KDbqv9#=9<Dg%R}7>Z0xAsj
zkih6kA%Z_y*B}PD$?OAi`soeR%PkVfo-2g-laOm5xBcKg^k<h3FC{pS%HZ1M_WOT9
zem&}#bjPca?;O5HZVEAyL)Cfn-hfyM0U8U#eav2Ap%KE#cN?!;>QuJz-!`+Pp-9kf
zNF+W4W?;GS>PXcazkICq8!Vg*WX)FKy`P^y^P)=LY7LY*j0K+*>?*?S<U^DG(gN@E
z>1&}+KMH-!zEUgh#rbXL`uY@I4&@qdqDmEt2p}-VwD6e`eP9)%q>|+XP4N%XcCtFt
zcN5HNPcg-~PXsC7!uo{CH1bBx3rnq;Jaus`Xb4S}+z`CCk`l~8#e}3xaMNHN3GCwm
zfii~t{u8$(PV6&(=GNsT$!K=C<%Q&!r0?<^DhBF0xE7JXVY1K=$@8<LB3otw+Ocp}
z<>N-aFOQbi)&7KDh@jgI;smM{72m=~$c}ezQo+#DtMvgrqo|ZQ??#1y!Ys#ow@W&>
zC<s12CXi!dv*ilB5O>jX@xsN}uCcr2Y$j2<9K*pJP`>!wZ^|K})hD2C>@%``D&(YU
zCcmPx=FOmV;+u!_p$~1af>}YHZ7LLHHT^=6z?Y=Q`1wj(f>PI6)ik*BoR%+^pT8~7
zm$u}TBsE{up4t?n`gnNPg)za=FV+>+m5Fxgw?&F!p<MgI(BNNAu9yqBE0|2-?(~7Q
z4{Jwvy^ZCJ%*&<)e7R#@JS@L7=Aj{v=GmVhgR7padYcE{YsuKjW(-kj)IUEFHM}W0
zc3sRfBmS;+iv#6PR3y*EiA@9&!|qa5(wVTJ&v%NywM4;|FOHh;RhJ)X(<^8FD@hYI
zW;eR^tRmb1Zj*ldYVOof@$Oh=>FCI3;^e_wf?YPI*}PBZ>xe)_bv<0b4CBkYETk(a
z&y`?sPs(y~bC=NgnD`1)($50cLEZPl<!Sn`*^TO;tGA@=mBR17Jj`KCB{?s}dfp=h
zmOJ_9&vz+W>X;#tX4wRpw5gahsN6Kl98O}Y=#3Z94J}2Ug|i(`(w$~F3U0po?)dCU
zo^tYQYxG^uF_RybBK?2ItAW<fTh34JmEQRYO2)KXqSoiWdk80er{H;Q!~9EwZ=ZGV
z!lv!giTO-t;H5E_s*Sf|`Uy2G`W)&b7mTN+#uja`AGO>QGXg9Q3JEXMU*kSHHb~2A
z4yJGo>k1Sf(6zM+<5_{_uQI+HWV2+8ir4&3QZgRQ+`ML;J|ua?vh7aFGvA&8P)-O%
zr^knRH7gm`mp6};J!_~<K=ZSk;-0wqa`A#9_V|-%X>p@LV**c}kG1>h<TmbxK7Lku
zKIbkizr`oVHhz=iZ|+l=wFp<I9(MRumsYrD`RhK5^*Hm>hL0{wF@5AiklOWYO=_Pl
zEcL059+5n6<G3Owwo{(CX<)xN{o0=L?6^2gL=yO9GnJ9ksZLJzKZAKQ`QqHv=i<U{
zb0~XGg2$;t4rAZ%J$k!&-%HuLV*;tNBEKfKXux&Gz5#nk?ownZ`smnlbxekn1LrJZ
zHLc|H&ifr*pbLp|;LQFiyD;BN`6H`J<p(9#f|oL3rtp}a!l{1mXY^Hdv9!3s7g&xQ
zdc{>$fgiK$NX91pr%ze*eZ2s7(mhwdhGV(<X^<=FG$Kt6r-$`^H_a)2mR}I}!1WSe
z0Ac{R%9%(GD1v>*0_%qPK&FcXLXC2^uI5>!;-(LTGO2J`plCmbF?wZK#xx^8-FCfB
zp0Bm_)o+s#B&D+G7gQ@FuSM*&75i~E&C1wG>wC9t?+v;65w@ICkkiR%p&tdQA?LFw
zY7-U%5WNSkp`iJ2H6t@gj35yy6RXHr^Xuu<$d5O=VRIdjB$J^X2TCm4@c#WyZrQ!-
zi}9B#)b1Sf9w-d&(@QM<EV!2EC{tm%@b1a3iK7>Fu?4q#iU7Xsk}G%}<zV|9h#xXf
z2Vha6=(H*~$*yIhg?@u9HZHy4iOj3)9#+4jeWN|4RYCMHNX73yv~Kb=0jt|GL1rc<
zb?DY4Z5`;a=7rrz4X6-h%ByeBf(qyKl)XhQ6)A&zzp5XUObPsZXi-p17zAT-RTUue
z=vzFf4w>JX$KiyS?5$I72eh8kS!(SzxC0?_{c6bl3-Z=^`-F(ZmsKNeR9K9L1KI9*
zIVm{PZ{-P%(Ta2^Ysn|bD#pBdVb1~KdCh{Ws)Cqg1P=Q>I;h6mccz+KR;8J|NcdK+
zCHegveD15#(xW&<LYJQkAO&w}=Lh8ATGZhZs=eh+IX)E_<LHpSHqwpRH9Z~@BlM^y
z2jwp3W51JeAWg2a?uUXu@;!+z297JqV*tA|B+;$ysrPqUhRKW$6GsJAsh~djyFkLz
z8s(T?9bAIO>O`|~9W2&LrVh^-Bq7r%>agW9KEiHn<bB37=sQ%9@ovn~5Y*F3;n>t+
zI<6d_t~pYAxzXR^+ZS7%G*7W7*Tv)V={w{H&KRic!wpC>e=M)7eV|;Z)e%Wt@wX}^
zWbI5YGGt;SUVdXJ3=p0&zF@r~7(7RtwSx4%H+x7aO+kPiTq^;P+qZCx2MJ-A4t%n$
z<&wb^XL}9wPW`cWwx8=(Y3muCT<^L|QL8FKF$oGB3Y1VT^uNDafCWCUs|=eRexw~?
zTg_H6FR&eH;YODvZS(5aopY)HVh2y+i@??>9+F9NM3Y|<ttvdnH;I}{&1%NgNh1gH
zmdjxpVfs5tpPv3$3Ii-EuWXWI@}7hsoUC?aQ=&^8Q@_s6KejB$w+i&WWa~4l&X>g$
zV*}Kpweu#R!ey*Q3G~nr2b@4r@TX>aAKfv?oP7Z<sxxUmYaz3D>H(AU%sm|eHA8i7
z)<Aa>AmgM)%)&9X9%*CG10_!eu-juva6Ne~1Pb6fCf6i@c>7z3hv83&ZW%esoHSiF
z0(;H7b4MAZBA!IUPR>_ROxoo%Hdj~D9yGq?qt9%45$qXP*KYvJgmLXRo;YF3qBsrf
zd^Nx1spM1Z&TzTcYtksW<;kymI8g3!9y{p)4ogV`9A-es`<L&^`Wi$i_gVhsXaA$$
zxIAN2(U8_x(^Qjw*Zze^30D9`iCI}+tiXe&;eB?Ay_rboFF9^&f-fEdwIp#%bnoy^
z=2lgsQl$G&JsKtJk9R^GPE9!NbZ?MVKZ2kI5c<izvOdl-;?et)UDB^pS5F?6^){S5
z{J-9`hOxIHb%w0^>`#An#i)WB&~@=GYJqnkNbN!6BSzqv`9^63ypE1)dn_i$|MmlQ
z-n2Jw{@bVwFW6bYiA6*?a)&JPNJTW|OUaTNDMbx9Du37kh|n&81g4n$pCisc3O$0J
zfNh%>hzG@*KQCU+6_Ya&_oPc`9TDxGqZUovKwF7gA3Vj!NAMHTdFyf)W%ip>q6`E<
z`{~I0(&zY3Hf>U$T$l9UV?6|}lHbRKlD0s>8e$&CjPv%GVAT4N(<hrSwnQsAkKWc#
z0=a8rJoGPN72B_sHF&%F<sM{fH-~h$E2Lu_F*k~#p~Ha(l6uz{ZT3|tzP!i8<5)2L
z0_cW`lw6I%8Pl~x35O33O%on559<}!d$@0)QRY)bX2MKoiv%&)acP@r=0`vMo`<hT
zYu1CCJ)F83o^?!hXzn<FCdN-9WeF{p1BU~zjUar#IQj`H22`hJ`-rE1F8BT1EhqB0
zM%zd##(gxadL-}<tTpLk)wn+fV?@LXqJCb0uY^dk-2PS&&$UJ6KEqo8s@A%YXW&VU
zkye+5Z;oYN9&@=S0LM*h!sl=HYTVZb;K(~A+}^NX*10Cb+0TD4(K|nIJINyA?vo4M
zLvq#%ZENYr*P<Y#bz<z|;JInyE-cED)=+G6^iC;vqt53qD{sQ?Yfv#cjUjE+Xg5-M
znYOEya94LGdPSk31z7^@^WQ_&=VI`Mg$3GJ<g|9Tap1Laob=nO87EPvcYJXQmv%X+
zyF4re6uSL<z8oHSfKn%39(c{}Z3369%ju%}C<m6lE7fn9v=s&jYccns!*5Kwd9I!m
zG&x>~lo6?_tD3P8DRjU@F72A#%<!1v0D{nbv+#inv8Rj(+GtOwMq@;r*}Lpq`LpJf
zn;R{eimW`duXC-&_~R+|WLOEtjmP@L@#^;-A3Hvpn?;&xQnuUL7`C3SkPN`9rmVv%
zUArxPFwlr<@e7^C2eE?2A9qBfy>{@d!!)d`3iAveH(pIDLl6Z;M9!{@b_Jf+J~_{w
z++BXkopl+g=ZVP)QEBT6pgiV;!yb<+>wQ`Gi%auyd^Qm>Wm^}6ZqQ$!QGdbGI-uT)
z9WxOsB9D?~-y7AxzZZVDs><iG2oJAxcXRNISo8wrnpNYm9OBueD&KRzmUkFYf34IB
z3a|4A>(*E|qfp<$$EzEYU!kaB7}L61)pqPpR9#ZHisbZo#@k#6+Ba6MHNL9*n+uiN
zU8SxV?up8$7<T8o2N8fYXe+PYn=A0$lF9%?4gU~T^cnax?H%Q6yOi6UDf=tE=Vl&e
z*w@|xlP3_jyp9dvYg@K(mOEjWJ&e>U*?!b7+SJN(*)>ldI1&p=NIK<{){{9k7Ns&v
z{0<a`{Cy}$*)!$J0pCH5ICYhe^`I@3U%)`=BJa7i3FhCmY3a7(O*gh^TeKrnv<xT(
z<Mdy$0yV(Yf#<@6%qzy}Gnw`>3(1YH?^bdH>f5^AV-|;?KR3yY1VI*MqhA**8g3Z#
z+9ZcCFg`yLJ@CW3MqDDQ5O_pmul+n*9W7j29%s8VZO_zONAXpCcF|@rP^<3UUl2zW
zuk^q$_f=`JJB3r8Rett9GC0^7U_+Z1A$np?oE}%yBpr4PskB~CzBnt9x^+IzJ3E<!
zCofI`LL?hd6XWD;0w^ER6;g5PDv}@VxwdXY;lY8Y;723+SAU9@J`&WEx8>w{CUDh_
zfqFI0Jsl%-@{xqF_2d|Nq<8N8baL?8VaYGp6dH4{kiglY`1)w2iFw|1mOADueTm4^
zIz)5C+GLgpB}NOM9g&dY3^o3SJ5W$-Z;2)9EiQmNbcY;qq^;_Ge#=<CuQ*?#w{i|Y
zWssX+iM5J*SHPpktDP_=)N2ys=Z$iU(#}-bpARG7JJ|h_43Ko7-2F^6=)CpiRS9yD
zD|-oM8~E*rtuPmHy<Y!Wuam;H>MPft>83eJ+sVJVVU_pBrEV?jwaIH_hdx4Mb#(-Z
zX92agWiEJwZp#NoM8+@hLBp@rX%9yVotz`O1m2;%s=uv6x3#o<XU6P<7=1ppTm<^o
zdt&ejP8hI)+VRyR`-1>tZ~%d`HNI%QW>_P@+55nwN_+WDvA{MhNm-t~XDy2e;d%FW
zS2)o8*Gm8R{;!6(Mtl;Po`1rk9+Q4$T0YF;Sw=Z!3-qZ~<zpyX?V=T&c4=;_`eWPq
z$>0&{riF%>RnCa*_<JIhi!h?5vL;|IZi~MlkY&(x;OG_%Zy%9;pf&WebHv5MsIoz>
zays~(v4GfjhWUJ4;ytK<io6WKkq&*52KgWv&~f%jiobU`;r+v^ldQc3&W^&FSSzwY
z#c8|Cf(cT&$=o_8Q<pruw+gv}5dNStw=e?rixP0&-}~!F9`utT2jh_)4c=uh4r<HM
zqD>O9y*!TGo^jz$42p5#>A(HS7o9MShkMy1s=pwcB^$1<vqaN?i<(gy$uw(GT6O4v
z(HXC$?-9__!RCtGQ?lm*bF^BT3=Zmts`Y)AyefW8Jh0MrEYkLryxI89nz^^2d~9c9
zRK;`X;z6$n<?aPCvnx1hhG>let%B1Cr4dN$DAyp|zQEvkRN81^)%Kj~YdObFU3vL!
zAV&aa0|`q6J#U13x=;s9X}>3bt^6y){h|#2cRZ{9T@A}eQL8P#fh&>Q*ic8n_4uW%
z9~sTN^RUVGFUZsPgC{)uM~1NUzaVUC%SG@$a)(^@5m2_@{G*3z*_b{hS>4?A#9te*
zS`?}A-+ll=!BXX5gnhZ=)!&orUV)}_Nc7mkcJ`_|v&niv;G@a)g8%zT=SO4zPyxpt
ziPw@4U7+RlONyh+=}+)*viile6Mqm?<)owj?RyBR@gHyVzoR6N>Tv|PI+6+(7mn$s
zfZ_&B6(hy}Cd`>wJWlJ{{w<_>;}0S)kfWYMNI$+{7m=itZE65qcX&@|>io7;p#ZF{
z@bM?FL9{mXiE_{5Z>P?02`mhq%?|g+sFr;A#pq~n`=pPWo~Cz1cBrf8DSEyeTA+uQ
z!q!GN=@A6<hnj94q<Q-+VS8_uOW*zU`1)0a<NPN_$M>P*dw&>=mhgUQH98rCj1&&<
zXejM(p4+RvukZ9?SD<^#?A{8MeU`SNCGTd>qQQA{gxRQ<#R={G@~>ax?9!d;E;8@7
z|H}2Ds9OoAtkHjtFgeK~*%pv)PD(eQ-rRg4eVkT#+kws4lauw@_l_dI3<gN+@^U5$
zYzsCms>y_KZnT3eC|jF4N$6SW;ktP_ctkc?P2_hljRtg#3$Fp$iH}AxBCN7mnhd0^
zxW-*_8UXI(rcaFj)Q;2Tc(Zk1%S{WcE^V`X5y|p{^x)OQu_|$`_U@5+L%~{a9Ccsr
zBicH8v})#6SS09<ARM!|z`HT+5-yaP5*oT(<H)IAQR(Y)zuu19GzMGaJ|OAmI<<;*
zjY3#rmxmBK&ew2Txju2XBW~j>uDHnJTZ1&9KVG#_QLbMFjv+PWmP;gMLI5hRdWDB`
zd{wJC^>*s~h_2yyMnys|wKB6+?y(?Ts?!hFNsD{%Xt=EESY*@c+@YJaQf4MK4f`qh
zgV92s>O?GRS{TjNa$`cg>fPtf7o}VIAtHeT6$#8y%awFV?x6i3<-rRg8Uy2k_a15z
zu2k0<RLFccLNi6?UG1fAW#`-49z9R7dl&<XI=e@^%&R1pGM^b;vw4Cy*A;s~dyb|e
zgu0D;<A<7!1QZ^w4QyD&>qjaF**4M>)r>y0H_f$DFGJ2ivY6|6#{1!w2Qh?i1lf=2
zk0-ok4F?Eat_Lp@J`|zFIs`LpvLDP8__=Nl5|Eu%iUh5~5Ae4$k6%>EB@A*slYDQ<
zlnZ(C>L+E^jQ?s7<9pIA{6e>%cWr=^T$oHG&lS@PAp^SBE}kh~RLooS{`Q;5#X)9A
zU=`nwg=sgBZfd?V7q#phmijiL)q3?SRm`QHwUK&mpMzw=SVRBD+OC<Y*}J#5UWiY?
z#5Gz^|FDfO!U`)>wC=l3Wdqq7;e*jPxJ?kLt#iQ9FU#HFi}*NsTjU`flTQdEd)r20
z5aU~rFKm$>1%kb3Je?Dx7=vLtt%6W&^Y|E3$Lz(!_$f*dzmLeA_&*rmi|(T}sXD&O
zSr>D6%giG5#W@JKpUHUsDEHU=tPW|4-?g6*a)iv#ldg>U3lmnRkMpEujdWhet@lz3
z%$f{kmoL2bI$?ZBIFt8zO)fp4>MPonqapHx{gatE)*$(Ag*7XmBH*o1sT*^-T#?l~
zGCun@=U7rLwK>h5L(Tjkdnz7NgTfw{IvEf|LVTT>h2t=S%o!}TPw!v%v*Bs>BLp!9
ze%?N3u7_NjGa8tsl<E3h&L<41yhOP#llMp_dSMfH`G8pXRy@bI#$~4Qd2-CbnmTWW
z9E^cv>*^^cv1IAi<YE|QdZ&YLaWS-w2OD+ddLWDcSiiKvdwy=C!Ez#-bz?@5){`T4
z7X5Qe=^iz&;%=PpUYTIyNr5U~pO<vP+>bUE1=~nxrq@*XoQ&kSj9sh4;~NNri57mx
zqdirAQ=Sd|BsPnPH<$O{sP#T{Pq0=#oWmti;KoC-IK@>{yhOxa9~Gmo|Hs*xn*||Z
zK9MOJ!ofXU6F3md6}sb(yRQ!lx-RMad)#;st7;^v(!Mx31jN?EV1%wCJ|4qeaFF6|
z@A-ME$JLg#_qVG+gzdeDi=So}Px<Y_j3x}C_q~Udg-e~zQfF2jU-{ChSrdLE*SQ{-
z0s52<M5^)<ySP)Du9PA2iv$eb#>Bf-@5gVxTFSKVm}eMYo|Hb0W~{L+X-}gz{Pt*C
z_rbT-$x`?WR8@SZR?3tNtsqs>d9z%IA6?93ElHTzRl0$sZhC;>eopI@seRML%ZA5%
zK$h)(D;?cS>?d@f4isg=O$gU%o@`+!M&BLDm0EagUeU$GcY*I!a2Cnr8xlXF#*F+W
zU=8@h5R9t}5I<HxXpDm@ecAPGOP#<u`<KqsrfCjuoaq?mgJOSx*3GI{Tq%gn_ZUM#
zF9x+rpg~}qer1PSQ4&N>-YT1X=qWH@DE%~7JzkyP(U&;n%dOFUb|)Kxtt}1EeO%b3
z#-hKt9ey-{a&+bH?|*YCFYe||Hai+x>m-U72{$sV1xg`-fyj0!W{b=^397*@fyBcC
zvnKb<bB;I3;0G|`-n|wa+cIabVZ)wQciPF1u~`*+#I~)n@0VpU;w}-`_r^Y>@HA@p
zvQKLENS+1KO+3rGc_*`WfD=H(rpIDkhFq$%h$bBhZb_+bHwfsPEE-Gfz?5X4M#xW8
zA5&YWua66urrz2KOVW(FvENO?w%q7~ct130=f#ikGS?5}*@Mk;|Fc;Zc!BP;d1a!n
zW665CCinWcgPp3!Vxc#7u{ECG)<{9nT$4BCQ9sb*hxhg5L4$aHt&d?pYg^c*G@A#h
z(~8NP3T4LNb}j`}JW2xcga#D!kZJ#8M+cdRw4Ed5-ivGz=?`%wQ?r~?FUIT)Ig*T*
z{8Brky!SV{o!0N{BIC*OC_J=d&C8OX-3FEVvO02w!!-Jo501=#Dudx2Uj~+naw(gr
zj(GYP<mN}zUyzVB6fym{a`8;)C!*65Z}?vG^Vu0#XHru_In~GC`H33nfG?6tJfbVS
zL{O$0f!brP0JtXttg$UEL^VuFJSRe7xJkqJl*Q~^UZf#|iHFW`Do@axxwt<jRcVT;
zKZD`^e3`8B0{Hwb+9D26<27mkr7gM}j|HHyzy?6$CTvxmOC>Og#0QYYaB&+nc?HN)
zP}X03B1F{q3nG};ws1gQ4ca~GsdIY;I+0Al?-gJrj_Kx+zYfbrqDYg(a=?~)-LM0W
zCh%+9>Rbi?m?s+e!GU+@)F^UeTJ40&WM%GOzQ2F^h*S=CI=%?Cpf-SdQ1Z`O8YBw!
zNOb25>XFF*cT@TQy?F|hJ74}wGy72FjKcnuI7IJ)9PQFb&RA*e&uq^ZoKMp)$9qe~
z#Ml+$^wfDXc`gyPv6&Z4>%O|utsg$Se96A5$`_mH;ntSGJowmNpc$N*?KQv#aF#6a
z7o-j3+)4U@$sp(co)=%*1k}@;$Ov#vnSnxM7BE=yPojS)8>2>IGXMg|<k}wvy$r-Q
z3uxZG%R~^eoCaS6nR&!HqCS=u6>-G9m7BY+_$cnZFa74|f`u|)#H{=y`jE~PG)K3G
z(^jvP`e_f?1=R8zBAVQG0eN(#r1Xm(=8%+>%?>_t{_VdYnlg~U7bgFb2!5>cZ<*qU
z6h{hwJX;+LRTnx?7iJNrlM<WCO2x3L`G!Hd3F9)wl<5mpIr0h*+E#P7uEM4Um$svw
zRmFcz#}xqc@V)YiUE5Tr8Ixc)Yg@<#N?|d1)`OKlSz;c9J#iarDU6XV&nE^R&?|Nr
zyqUQw-sJI)X5ZW1$V2znbqHmP?(j*rsw}*t6Ls$MOU84S9iPk=@{1W0x%6oToi8ty
zYk5|(tJdSbL9@Z(?6E?iue(w-jD8(>f@>@0+Eh~Vle>EaQh6m$QE;?DvpcBvV3ooh
zC^l1)$>@H-1C6nlbE-i7GjlDSr=P*Ryq_C1>(NRWV(_;^Mx%I@T?-MnCieQz9MH=d
zjBZ)=^IO1It{Fz9*?9>IzG`iyr2L{gbCMM(OKj{!G2@igFcxpUXf24#)Z{z2hos%B
z-f^C{VajI;h!CJ+Vw54G@T1+j1i8-2wxwqCgXtZzf1Q0rRT6w!LC_BNP6M*Iqxi}a
zxml0@3({ro(?u;4U;<V+Ln|V5PB*}1;y60*>T3oH>#Vx9IQPWticdgRCN^Sx&<osV
z@LkLD;cE7m$F!n7&itXyACmdyuGCc)Daa4z^n0sS+IJhU!W+>n4uQfAQRK6w>8ANR
z){(h3dq48{Rx^(1tTTe(=uWk;K$&Q?A3nPMsDI4WSFA{;AR}38ebY?G%ZQOGfDxK#
zm3=HkWa&JTi{JKgcDjukeUVoVm$RB_5}KRR`9rg}k`&!H9xueB9C&%!u=>vzw8x%~
zZaV+LS+<EQ-o-Z_r+Nt{$7FkUpZ4S3>?0t%tfUB`TalRPw`Ct>V?EEeQ~gb}zadoI
zb0;E{?M?Rhj_!jjv2hUnyf!hb=XLZpD}9llE^Mq6Z1RW?uRWUE#<~F{4B_A=y$|q7
z;|ZDmi^q2>F8Z67%{*~1TXqz9T5Wnto!TQXrJIO82qhen1dj?2><SpNZPFOIWQ&~W
zLS!CPAqMi3+><=VG9cHXp8~)0UQpJ@BF{fh7mi$w+7o)$6t#FeQPTP4&9|D6KXjYN
zz1s>MwBje4U^CS$o2!vD8zyJ2nEKY#81=J+c+fNyQGuRfag3}u7+4uXFPGS1JW<`u
zv#VcmOlej&^VGo-tPGjL{%2)=Ho+TljGeG5`%c<emm^CxqwsrGl3&9{luT`yMHcC|
zLf;a4RCa@)loLoh-sl!V0xDu(g|X{g2i&?!?&}=&gs}dXi2KD5YNg$*Jm?oV<VdYq
z*6{5QWQAjQVm&*69eTYbpSYJNH!&jyXE&0XuHEl(wH7O{Q7S1;==d8}w_hP{<vi!%
z8}pjGysKZEE`q~ad`rLJ4Ryc7P8q{IcyBH?K;iy^!GVQnbkKX!r>TQg&aYBYkn~ED
zw&Ug8C7jU!o$n-DMf>}>Ge2*hL9)OY)_K7ja@;OVpD_uUILlRGQIH{44!YQiy-rZH
zNphM|vp0y&x<qh^#?%m$ZuzSwLK@?2HD5QRAQM6%*2^0)yXw5x=SY%-{^sR_?Irqv
zEMXgQo@~(~KJ3-r{Ofys2~KO-KepSB4#Qq38~A`~kqv#l)fydvhL|n^-#`~^c@Q6B
zBgAjlO5FEY6QP~8<@jXcX>H_eMSJS~lo*zToQfp%wVun(9S2BaeHk&ktwpMtz!!D2
z0eq0!{KcJi#<~_yX|7*?whkrudr<5NvEm<1;#dinE#BEin}<+&i}A;GpDPauT;DQs
zQ}oc?zuXdfk`{P*1Ibj??6J@pcBM=j{kn97KR3m8A@phv<Y6~pFfR8biY?Y)0;PWa
z1cDYLF4-z-Ilne3I~4Zf)Hci25PuI%*Ad{_=_APy+Lf;0kr9y(LQM4E#<zyH9Ze2p
zpAO`Or!Oob<D@s{t~rU~Y`I_Hw!%{cm_h?>U$FM9W!RV9y?7M$72*FycPabyhWRX0
z{>~3`UAM&gOJ*s$5Nr*Q45wN+iA%b?_;7g4B39t;+xU|+Ep;c!$`YT6_N{V#Wk2+v
z%B-xVi`W|_QPGs7KIq*kEOT*MwL8fr3#AG=5g-N{)~><e3^y6gt~80SM)DFbuWxjo
zxgzm-E4Dt9>C#(gy0iBoR4!<qHxwHwog}BBYQ}RUal)D8yxW(@u3ds%m2C4#@p>i+
z@1b!|$9>9S?6=0RcDF7Pl{%3`v1~9vx2zm#D?z(XW+OT(U);|~HIYl(_D;AX`>p&f
zR-@VAs>kca-58S!K-I6hOjK!z?xju#*U>U_#e9)MYS&g1t?Asy=zpUT)&dGS47~tM
zus>gYBBm_~M@W}mr1V5y)f9Al6+(X8bBS_++D$2g=J$E>5Q;?sjNT2o0U9jIf@ij-
z|3Y}+^^G)GWmqYe^OG|W{PWo~2-QL!gq$~@7P;=6HEa|8!3AMVK`mJjvE<ha=a7pP
zNb=>D@so6fLNVNW=}}5n{qqB#q<*2nL}7-gAJgne+9w{20w|fCL6Wpu=im`LAcR(0
zB$>aZnh3>;-$KLaH@%!~pB?)(&h!adpE7#O*&9cpqQhnX5elIk{(I(S57H6j)7R&(
z5G}=PUO61rdQTe8Ni*+kMSbRcLcaXFGPnnq+~qg!TZV&T1GHP!7=DgRn_}d4@1-W{
zip*XTV^5nIj-YX4<Y-f%q@o0~5lnah-{;D8qKKk@?|wteDI(Gkk7~E1)SbxPk)bv+
zNfn>E(qiw%SVt$do&KAJe326aJAq=s>N&ycn-{u}9Mz2mQ!}>-p()i>iW4PjKYUJ>
zLxbOP+)R9xEtpUkWRZbQV9fbd7K~iVC0|AkwegZ6K*uGp9=Nt6&p(C_F8z!yn{Q)n
z9$&uQ$94P7?3ZtMRuj%X#%9=a$oE6Ms(@O>`WsC^k?(9rBR4Ne?3NAp{;Re*1ZDg;
z-O_3IIZ8K>3ADb$`+&ZL$UUPWs3nXB;X*XP92-O&t*nGaep;`&n6f<ib<q0xE=15{
z#gtD-NBj>*o6Nlg7bA)!8qb9sdXF(6u?_-48wPBK{ZBUY{KIBP4?uZ|6SBRIfId0f
zA_$=G@gL~p#eY!-5z0;1NHx64AKHTzNZ*I9YKI3h!#6C(ahoO`7n|QE202CdDey(f
z42qexa(SU2Orvg99V07DEiBmPpGI2G#n?mW$SHQ9Ny@(v^Ttm5e`TNle<kKjmKTY<
z9Mc+IdI3%<p}4MsWdr*Drr7i4bsYjlZO~aJO$~M4Tvb})@M0M=+;8E2-DK6=wUr<+
zlgEYaLL3;AL#MT}8pZ)#r{+ZAD-pDR8rp`V@YWYVEj{c#*rPQ$lGws2z{S|5Z4!GI
zpd!P+AhdF=s6BZiQ033uKoTDPt^}K&jBM}6OPnyM@_<c`fsr2Z{=Xg%_1_&&;(vI&
zzJEEMZ!NkkCXlrnf2Od&p`jqP^3(dnh&jL6K!w<yjq1!Pk292Ripj)uVicH0TSw#t
z!H|Eu+QGSS@VS5CY1|H)7_~$^!)JICs^V&UF3~7nU3q%~p#pt=j>P{?iy)8$G&VeT
z&mhvgCVmNbM@6>1_4Cz8=kp(LRo(EgG+o~6hJu>0cVr4MIHfEi@UmmNaKygIJ|`zZ
z&_<@pkbTN9L-Or7vN47$&Ve$4xPkBMHX!I(HGv@$c}V*JB})yJl*YTE?0Z5Q9dt|@
ztEii2!3_!+rz;JJ<sY3cehEbunmVpY^$Be<GT;cx&r9kM=xdBayw8tyz}3<QEKy{3
z_#m0hC&HyvS*Zw&5Vq3ur8+R)R+Q9P2@=w#lL}!};CU<j;p81?I$zO&DGjSE{^c|6
zl3l1RswcEZ&#(iI`ib2(&O(?AqM#gPL2$yN;`SCI8n)i$K6iIa4+tGgP3++QY_#>G
z)737Keffxv0?n<vaFV4eP3!|9z>7Ub7MNy{*#cU`*a?nt;1ICyeLL=%S;x0p&W`5e
zH3g+~6nKfQ_%I$BgGe61)P)Ye(*5Rs(xjn*y_4W@h0e<4Ps5nSn7Ho7!705ZaG#eh
zix_==G!Pl^(ageJ*|zM7gI*>6)EP>jIR*4RsWE$A3=&<&Gez9Bsu-?z*xR60nsR;?
z0<aA82d{I18hYYXagwfj9gga1QGS1$pt9!|TQ$*tRUt$^oy|Z$svfGms&eMWfVwi6
z7u3GMsez64>BCQy_?%Bm>f<^{Et7Wn()A|HFz&UDxAMik?J3b#E8$^;5Aw5<xy;Y^
zw%o}UKX)#NX3o8dE$$G6gbUZQ?w5kta22Z-Y9(Jua4B_BmMWeTsrNR+iQA-KVU1gS
zH1-6qUH40!FKT~#N}x4RXh`{K$%f)qqR&U5x!YHNp#7FE?0ZpX*4Y@6aEdG2{jpjM
z(m^6ZxN|!Dz^;rZ+YVr{@$j@d=5H(xjl>EI7s<C`n$-@hh?SV^i^P%#!-xkytRzPp
zSN@FlL`Bw@H|zdjv3(-QaJGVGNZDvb@8$pp#Ep_UU!$)qQ`5T2MyEVDK!piYU|udH
zM06Ty0|S*O8U*js(v1eqiTAb#Y?Yc#ksp$DFX}8@jgfpvrJy3IB{iU0#P#w6m^Vt4
zGd}r1=C@ook(zo=_+g>5l2;Qq05wgk2zx0Om_J!b?#pIv(#OWyn~Wd~-B}M4Aq`Qh
zy*inJwnne{Xn@f`YBISSq8eR%^hUzH(DD}ZDP4Y>$9p|gkD@}zp3c_pieG+4$-1@F
zadbffJr_u#gz>Cg_vd7mi50kw!}kO;T|wtpF(o3Gf;Y9z4?%B7<kD)SG>VZpyq_uS
zEOl%&W;RoRk-ldtx)PO`fRbO`h)Y?yuu!Ge`98YmXYHhQb(pqTg0{6wAG>hq={&*U
zbzPZlSOvOskeNu3*)TZ&80{+_pCV>*N-#1aptXql)S~k@`N7LUs+LVeGi;+?4=x`c
z*RV}nW&5E#>Gm@Q$<H}qTy7eC2W2^GcH7}CLZb*M_E)u~tfj28%b2;jHoTN2)%2&)
z_k3Fu$ujxXGcDPl7g*Ny^Wd(=<L6`&KbH#CQ==)OF1=s@=G4{D!S-$Lnb|(7ET&1l
z&MtPewP1vfyu9scQ~wYR98915jpytnNwk0%ci5?jXz*K?a7no3Rv^@A<?_=;;Of}~
zimZD;V^`(yIlhI^OZsUhK%jCHt$%vnTXw~o3ZLh}5<Hq@P$aV_w^>P$up*G(eVHMq
z>wIykLccB;=`OE|mUy>Q@riclH-hPm#_C5MX#KFJHF3~5Ssfo%|5@W)|FvvunGvnO
zAYM_Io-io6t8=pUBlJeD0bMGb%s29KG1k}DR_q4UIj8(Vkim1}N7ytA##-llivac6
zf$^_3?Z<T820xwFU{m23wp#2q&AI_b_gP)j*E(j}lvr!Yl2aWXpJ=GpTDkmp)KzaL
zFQvtu&mi7{$;)F6-hdl=St-(H`|O4)8=n5x<mPgD*1>*Z+}U6=)>FkRZ+7ahx-q)=
zPu5`Svm>+k>hWbk)%`BrA}jS3HICN0QdPAY8Q!bX6r+YP3S@@{=^h}hH^{})K;q#n
zD^{}`<a$1@LiE3ui-E>j-kjLSU9?fpa-fc5{{*4j#Rg9YsTkKu#I<mZDVyL6VZ(b4
zBl<TTu<$&)nWp@rOZNkGZk~aI!hq@plM95jE;%b&qEqR{^h-lM-{*;+^(5x#b_AbB
zY{vo^eioz(EP0{!*z6+~(yjldK^{qTWYZ%IE^L;5fd@J(ca^TQL=#`kmGfyPU7+cx
zOo{#4><d|3W&|QtSO_r_JgMc#IzS{OG#qstcoS-hu!G?@*8R5MOuxRs=qOnJ$xwFO
zi0jiLOJl|p8#gF0n--E;%i>VK?ty`V|DmelB&Nzb6bnPaxaKLR)SnYOj`<BcyxZMn
zvAN-Ptv%dYZf+i;W1zlg1=9!AFNDmvOeUY^!LRS`w@0omhR$Bwd?{$p#{(*znm@Eq
zfOWn4$GR+#U|AqnZxZcFa>kG9O?U-gn8|1|s&cAQ&=PIrrjYNUjM>X!AU9<nWZ-K`
zF@vF_=D*lVeSDogxu=3vNfEiPsCYXUWWD$0TYFkn%dPB>Z^0)Y;DtNhRjbhUTHbVe
zT^(n3!7beF+2bpViM2uSo@Jiblo(f{aUcjDa+Sao#cOGQ0+<iV*7%Mt#^Y;n&knrV
z+3MzNn+vn%w`cN2`W1fU1gxI2{l*g;L&)X=>klTr|EGODmRKB&0#nlPqqirkMX+%@
zYl3FC5m#&yv0~5{n*&OAL8^@Z_!$2*#tK<O0x7w1F+1Fgk%n|_NZQ#>I{JQ1Xw$FZ
zj@RM>s<UxtDT3s`4Y(PY<2HIxjuPO6J4N(o%UwE%Q#vjgt2E1`{H>l>|0JKuV2kh#
z4VqwAnY60mhhQwoUv?k^O$#H-!JH5h1XgskGQ|g{v~q3Maa>)}!)PT+n^!SO?p+YC
z{~pYL5ed#RW!=22Xm?I6oZkqfqxdUbA%TjwJ4pWOvCpD#w;GKlLT#if)Ba(_@ILp|
zX(jErou?PSeUdH7lco4Z$>+gXD??xfpkx5pPqY<Ow8+E%8~i;)w)f)IP8fLpVBc$i
zeSt4P+`Ohn5-Xzw)VTDA8i7&dZ)*IfzhwS{jpcu^5h%3(dH`O0MHAT8Tj791pZ*7j
z0`c>I{fMFPL9AK`m}2o5HUm=3MIL!CWK&IOrFw~ilvSA}XfZeFSP)WW*?FsfUH2ns
zUKdzwvXT<}`Eau4Ak|%ZQtq6Xp;n&ITpWaFIT2P=^?$t>YS#oufj_m;=~qP-#G9l}
zhjzF$eoa*mF<xttb)nd&3ZCV{2l~0b4pcMJCh?osqYLU~j%2{zc_7>K34cLOXcW<8
z%5=3QBRI&HgmPO0OHhAcEReY~vhhXANWi~w<se1#GoJV`3j{y8+8}q(>IusOG1aOt
z>8b%h<)OUK$VG)bGOKVrNIMu#sNuOLf6W2L=>aiTG8%m1A{)F!29N;>)y0A#2j{)w
zNl=+1x&M__WXfhk120)VMpFf{jI%B)X{4%&YmdJTg4WA`Qrq8nOI2?4wbIfa`oV}T
zfg5&g;;c<Juc|%Hm@e9uqcys94~qXzK2T!f<Fjd%ZT$%k%*S4q(rn)^Z}H0FD81i+
zbzoPGp~+Iss+QZ_#iP0}x^`I#KOEmvDpl<6_?~L!lJ?|9_czvnd#3Kc=mYXJAwuE^
zTf%#+Yg`p7QgQy>hxL*|Q`39Sc~PS91@8rn&{#%xZ#C%V7Lz3nMo7}j-6Sx4Y>16L
z>dn`sau|LSxlC7Z9_h)oS_<({V<!XCJ^3;`Y^P775mnjTl(fYeeOvO9F8f#LaG0CM
zmi)IiC6F$4?clxO2s^)b;$%AfMS4!nAkHpvYpnZmlflJehE?(89}wRGib8@bHal)J
zkbh&^JHGg1_S2_6LizW*Tv?X2x_r*9-@7O2?{9p#i6bO;tMU`yTj#=~4fVBp$qL$o
zRPI%G6RU!XG_4xlG{8Uz=RYZfHq?3YIpR`_06fmgfMEe<J_qXu*2eO80yc?#aRKTt
znAhn+@z-RBE1Lt7*#^kAJgfQ@wN>yw%PTiGi4eZ+q2x7U)Dh-S8-lPVe*zXwAfnB~
z^(R_7>`%`obWs@-B#ZM{xhj%$yA4z}hSh*4DT*vc@QIap>3VJ<R*TYZ@<!v$h~bKg
zm{UP+K0bpQ>O3k~iC1Lh8p6Gj^MRZ#jqx~1MS~Y%I?Kiir!7XI>xx_#_p_(umbN2F
zGFWj7Gs2k^*-XcSGgX@PoIKi&0*K-x3XYyb_Cfh0(1^uw1Xytx2XGdPgYi{3qxC*A
zbK^#tQ14lswWKEBX760v#*d>Qc$w!(QXmiUeTO2sx`Y^hH%-4L>4$lD8s&>51+Ql}
z_opH2?mbv}`6GYIk;I5cf(INqr;CqbBs`ybQr(6oUDg!N5oD5romn6c!I-zoR<jnY
zjs+zZY(N}hX3hPE?dCm|o-(JCcjPPZ>@E~O$%2@=rK0iKxob?aN{^Ev&+!KY4cQws
zhx`VGN{fe}Zri@bm3H0eZARfcI?Z9cd&<)>p6-^2lSCj#^3pIe!{E{Pk`46A>F#&Q
z&j_5K)ty<}WeBWZszvdV1s~7rD+lr&E4!?uvg4kL0-EKx`A7UL3phyu6({lCSXdQ7
zW!kDaEk*3X?>q6~vox7GbcmcdsU9$r(n+|h;bj9lv^rCAD$lI!#m5@0goNBI$()Q+
z?#aKL%z-I+wT^fTOS~o+NfItj)w+7;R=qB?oQru&US_X$GkVby*{*0GQKqbEGr>D8
zaY}=#NiU+Q|9jsL-s=<taAk^+z>v>Kj^`r|s*ICf5dylGM>mBVZzFJaak2m#ZSJ4x
zdO|$ZtD7J(W9YAb%V}sC-4cIx&8=}>_EFEKq$Wa10=*q6A=*HU#nN?#T9{sM>9{h|
zJhPbH70(Rv474QIVyB8We`Jn@%6As_b`CW!fl#+P_R}|7xh_I}w}Dcf6U#I*v$$!3
zPH(FYsVhML_%c!AF+Tpampy?s`A1*COCAv+jCygSM@NmpJ(7Jc$M0@~c+|OcEn9+U
zj$meq;~D7JUXxL^itTu>L2mei2d?J^(&W75_+{#tUpumfZNI$+<(8)ezUe9xVFK({
zGeM#wp2^|JRI8=2+9uSpCd`FFYDONd>}hH10R*T0jWSvy<>QH}z1XcyT)NKdSE7wU
zXSt7l2C0JPn*V%L-Jjg7UQ)4d=cWHxLpYsV+0ckeT6^NT`{8`-nWY~qXVJO5cYgbj
zFRXT=jyUqcom>w|mb2PlREt&>Y(B|S*2I)*;I_O<uP#n^Nok$e%@$=?ujAlxTnq9B
zEufDs*Qv#|aE>Wk;7hEAt88PpS9RtEy{Ea*S9Kop8ZiZDNowhEfzb!}7P8PTP=S=Q
ziDz9wAJC2U!Ksq&BUHR9j~)^bvc-nQWTh1UY02Vz7mgsTngSODjBEteJK!-yM1C(U
z0jIf=$q!_u#Or+|lq9Jq9!TNr>+U1;<CV9$wC#P$Be}i|&zKlMv?gB!u&~IoCSH7p
z<)n0(1a6||Vy6uUjZKQcT->tjx5M)kvnCnjA*+jHO=kEKlaW20*OQ%SNzDunS@~yq
z3y?*>$!{<`N>vCsG$=tnW$`x@ZyX>9frVJtuvEW~b;CV@6+NzkpMnwnujxA(Bs~YN
zgYHI{?SFkDItC7^UK}G`!I#|`^{TslVJgWkT;WEzN1Mg)S1s-5dr`Z{Y1*#o?2wj=
zR_p1U+T6!_+zofu8vEwmOewpyUo89Ihvox|!wC8OcoPiY?WA&Q*3!yi7VWi68J>9l
z`Xv>Mp94yZSYZgM2%N>+K)9vbK`siR{zM?~NmX%BPB4|NL-1;x`PdksQ0Ys`x%o`@
z!S_eXF8tU1tOW_`ymEh@`G^1Qk$>#J2n*Y1Jd3z8ByDNSECUK;R&G1D(Iqtn>+Vyu
z%47K!6Z;X);H5S&CjazOm4P(Gtix^?m~7|s$h2#}SEu5)HBFZg{Ls3_+kV}FvaXEx
zqzEDK8MLPUEZf3UE$8FyLzt_g3G`7<x<sRt@r^-+V!Q2t2Do68$*dCp#SC*O>OIi~
zI+EOwbigBPy$!u5m153yqUUruwqvkyQllzPr9eq&GC6n!(hJPi(Z<8JQ8S(1goxmG
z5mA`yx->rq*lm`|2XA##gU3GQZt?cwYNx2hJ3b`Ma~DDeThtBy?7S;{<TD?@MWaoi
z{4d-w1h|93e*zON{*z^RI~o50Du3e=K+Qa2@IT`al1c3!urm%|2Lw4Z@fhfT0+k^N
z6r%knE<FUeBtg2W7Qy>pF!VoNLJoqT7^Fklu830o$_&c`f8P&-sM#NvF+9h^le=5Y
z*V$IR3SJvQndS6%B=H70RB_7(ymyBG_nSEVsp<KjewlpFuD0*oJ`dOHx@a7`)O_n$
zb#60sHtBXykAPHzv*0StDf=J=;WBnYGcC5PQ&h!j#229<KI_PzR$H&>u*^OFl>$BV
zN}U(^ck3*eIC%`*<VcT}kQaYA%yW&+7}T)odEm~`Z++itO3s34#DIp#Y80dwAoF|K
zPs}HMuQqjL4o#O}aPf_oH;pH-(J5@G^TPi&%fN)lBjC={1Y6pU*#1C7K5|y<otT$B
zMP)f6_34@ER|T|2j0xz0i`9<XQWb!E_ZdwJ0rF}I8!){lU{qk^byI7ITEJUdMLX~f
z1t>}S)BSA+in~B??KbD_;?}!0RBbo<S?nCEX7Gh;R)PtYY*1l|;21iJt$T89$e1ko
z7Ghei<Mfx>%el4zZZ#DhZc$q7Crd%pPk#(SXa2FT2m?aN;$iG|%jJ!Ru(S{TpUhzQ
zT%M&*1#?;u@9A{Vd5BsZcmbn`_yx?zv{I*Z<#eymkh?o79T32L?2jNB8Feak(CL=;
z4?(~Lhr&Cd<#^8_H2e!4lGfQwCX?oGsZ<sjRor;OpmQnAO-uD1y9^4L(S!+M;e=ZT
zlQMHcRU7mwLUwWtaynMRF;^)vGHUL*U1vm49)P~m|FMzui2KA*>^ERv3!A(*X<JcQ
z)sjBJ&KG;CxO;_SnI*YaIIDUFB-+n6Sz-(#B!A6c8n-a4%o<<e@woPDO40c$^zm-S
z0hq)~*lu&>$-cPZYM*1^?<_N0sjbp@Ma$$_2UX&a-R<<f>?!-Dv?!7UHX3K$sc5k_
z5p6TEF8s7X%kIHWa2DiN2KG~#0%5y*fmb3{rtWj~+KbYsO{wfTwXvu8^H{kxRG`ZH
zW!fjHEyBc-E>-!K;U>+Zn%w6H`feW|-v9QkSJf!T%_n13<uoL~;6KFr(-?`k1*R|i
zjZ--Usv39c(h6GmkTyFMBEpw7*CPWcp`s9+;KbUl^e>SZ{aeR3Pggj7TS<W=$(jl1
z=fu$GtNKC{Zp49<^1iI*Y>bKq>+j=Ygby(?;M3#Bg%P$@x{tqx%u6$Ot2N|=xT*BW
z;7}Fuon<V<MyrXJ+2@`M-wzXg>UBOT=bV7cBR=N%H><ha+Mr{Ck|B|sz|c$fx5dx(
ze~avm%&~P?rX;>el2=_csS5=67<z=^QXa+$OpE~%X8XaI`7N&P-AWIiC-y7Lo<}C5
z0NhtgSKD|aywg0vv#edl>w0YO?{disRmMjO%NLJm5ANcj5hM_{ojtu8DMlaE#H@MV
z=(;5L`^Xmzkl*2ffFp|)_py%CJh}cCOLU!G1$)`u?i2w9;{NG%(cS4)Y<#ELm!EIp
z=|3+Cq)u_Uhd$0&q;I0I+CEk$84|8FF2-}uIZf(*dhB4_=i@Ba6ZUNQ`;3K)WPOAm
zkAcc{ioKL)#DzQWt?*W;d~;<-_fL+8Y>CpZLyePU?^{4JY<9a4m)r?xjM!+81WEAF
zJ?+_yfKL0Ddv0guPx}|P8bpzWCt7&m<XwxpF!FZC1nfn1x&wZoTj9MEj*s=srBP_|
z;9Uq@8i-D+{F(yY2noL#_SoM0ueKO`S+~_FRoUTb+CY(QRx+)L>OO{5`#YK1CI5V|
zq7$6$prmn4#Q+-j4qwzFkI+hT<Rzc;ubS_w$#)N`-vkQ?xN>}jcxXka2`}x!SQIq9
zH%Pz1wHc(-yA(+u#~@2Ud#dwF_23OK{Edu6&9ZxEIgDQO7WTvL2llp5Pa-&$R1Cln
zGg$dxfQMk*nI|@n?wg1~DTNDVaAWr=9+!3L)9iMl7-ew0cX#@6bCudmi8Vzo3}<m8
z4aF;gS2k(`If%YxPKvH+2Yq^W@R;cOTQ(4#Z7_4bLk(XL$Pa??BINJFKh3NGLaBvj
z30I$FDC;goH^RmDOO02)<kTiTW~J<DFaO2<TR+bITMrkub+CH@SQDk^kqbBeq(QgI
zl*AH@UfEPfXO>7)7H6!}ojNB|`y{m{#+0xp@?y%PxYlpV%ap6zL?aZl+OJ1|K+wqU
zma3DtO&p(z!b{(4toTF4x36vIH!+==9nRGmf^Kq(+@nxtUf@H#13*!WkvpL`K^=Za
zlJxz-Ku~dbI&7+>Q&prPeLlcduaWJ>V`UgzN?q<N^mKTJ`w*fe!S@?<#H)?;5<j1d
zSqL>hh|McjA+2$C#?9P%yP`|8I`*2+yN6qGvTFg~7|4$t8ubOGv|t}ttWK5|qOva6
zoxH6cJU{43m;N(bz9MS5tM3$n&zqYz_{wS_CEsoWV>W<KJrYoxa*d8b{U&hYmTam`
z&9tW%z9Ms&txa$0o!(2<bsGxniBY{?I^)9?=Y(A#TB3mA*Qo{c;}Q7d-CfuWyb1^7
z2kkhs&&zb`f4L?1BXzZJId=X1|3lk*2Q}HX-J?NJlq$V9X(~;63r(eofTDB(6_DOr
zfY7BwXbK7l77!3H(gi{fy-Dvap(s6o5FqfJ&GWwBncvKrIp?4A{S{`I+4s)Av-iHP
zb*;6ou0Kr_{vDFjp~^`}$+ORJp>H5O3OdaCh_0~WE1MPO|6y}g`*uT30C61t=$y$8
z?RLVt)3gI{YcT3}<y-joA-2Eak818D`f4|#T~Upy`($p!UkOw{J3!nx2#ARJ=NE6k
zAOui406w;Jver6C*MBPZd!8w7o@qkZm2xw^Sdjl=YISn}(<Qaehv7s}JAlS5`N07s
z|LNP&`7wyZVv&`>S6~^Ba3QU8)Mn$XvW*{Xv^kjUU#8Yw(gY@@M;D5D!Y!+q+=JT-
zfRB-j1&|{3?)axA>c~vS=#s1H=TrO+Ml!SX-9lYVa_hvQfi~>4^554{S%;apugJxa
zXz_V`@YC?a?Z3i=LyLaeN>6#)t@>qRKF(`95vrUO>cRdYty}n50OA=Qo$Yg7)Z1O)
zY&ia-_(qoB3aceIPJKpoSKm!ImKod3A*N<>ALfXwa(wt7y6V55#^LAVT-Tv$i=|?1
zhx4p;jQWM$k~*H>DB^DH?G^*iz!!roi-F6;^M%Mkm@?StQLtsePW!xdc*h0FZfj2Z
ziOrwkDC4#Ct*Try`|13i8~TB^lzs7>zQqbxxb3oSeqE{h&w~bXtNhSbG<fCxyCVeI
zU}xG@*&I`ARhySri9Ef+x|I}ukq&tquk}ZF>Vp&c$1~ZOZBZjxsuv%w=Y?|BO+pq=
zuR9%3T@fR3ot*8t);S!?_3sS@RCfQ1pQQaNRX}n^4~nuUMpaAkd`)qR;&#G9xAS5Q
z=Z@g$OIP9UUMZG}pD(1$*SD(zwM4xoZHc=`G9(9Sv?gkKOwtZ6HR3Pqc2}CFdoi+c
zP-!fOp!)84kthILE~9*Fjx9AE_#TQMToMmk*YoOt#<zyBKm(mHE~2JTO@BY;pXxAC
zbGYX%IiBdR*N!+eC?SE=TlaR0+jB&5`RP?lrBFw}Y<_8OcmFL7*4EGY1KQ^LS;a7F
zrn%`-Bez&WrGpRU^~vdbKJ3G@x=?k)#}S+iv3yu!#Mk<r=?aWW99$6_Psye_27<Nm
zc2V=qe;`6y+uKywSx5x&hNbnQN&=^%KC69r2#w|ZJ^on_?aRG8iDQ$u084UBl?x6c
zQ=+}-Ey3NS4o-dmVl6~_4;4<$sj=pY+Gxb4+)Zxj+8DJ~iS)E04<^O$jbON<k$)}N
zPL+FFPuRL&Wx~2<*oxm0nWXC7k=p_HaODwthsnN)HdkXvIzCYbXq#jVs}#vA3SSNc
z=7jn}yu11KRIUY!!7%!<%4}6l6WcqRJ8nyrndD#R6H?{wdQ2uxk&aS*;SLR&P>1`S
z>c9;^(?8tZqx(ZXu?K(zd-Jt}qc4*fYrSMn-pUU}2q{Z9bfM)RNc_BPNUSAiW#h#6
zDssh{$&%IZt%Mi@O){;67ugZ(@?Blr93fkjyIvi^JoryVPK{h6OpkvwX_-Fsx*=tI
zL#yY9r{Y&wS4&AqL<kAW>nq{GkhF-CHmgtK>kTiy@<FyMWWnpNUuvto4f>+3O0EvK
zJG~FW07M4P%5c~_(RD-b#}67!df(C?GQY=qWs7>bn#qcG*)UCsVvV<c*ZCHIpL1lC
z9hRbP^PZWwtle{RhZ|g>PGJ23$@n_K4PSvZcZ*})ipoM-{5dp|%rbdV=_Ya)>9yWj
zBzOQJ8#~JaUQWORuk@?^^gxwY;iTg=39Ss>rWaZ{s$GHX9|5IdGL*X9{$JnaI0!H=
z0@7eE3%Dxo@@mHCtPP>wKi@xRvS1-4h-a5r!bt2$5n~{%ZMufCSaRd)Erq4&viju7
zAZcUyH62X<fz(~kdABCFE;JB)lVE{s#zzB<KjEIgwUbkjbjc&L+WSuvd#IGOo~5np
z`l*z!p^=PEO*DMKWBMi5p!{c*|J6E|EjB^!Pg|qFUz!YpMFsBiVn}i%$G?tL-V1V2
znoU#ppupX(t7l}FymWPdnnOK6Y_kcNi{#MY+P5n(7E0cywiQk`nr}Js>}=)Pw3&(g
z2Mdw6@e#P_P8Y0;hYL#G-;TsMwT+#zzg+6+%a23{>?he1dFMSLG=#?(E)FaElk-xK
zA~S893G&1A`Jr-!u*yRc`7Yon_}2#<Yh1)20ERT)FyAmPGy5Il5@zd?aeekatvWQE
z_zo}9!?NL27?jubBT$CloPq_)EtF-;WjyW<7uTn!%8KlzLO3nu<yTRfXcbm?NZ@y?
zXJRosA~dTnf0)%|&|AxqefdQ+o3^gK+K_grhfl#I(ZT8shoJY0x(lbt8Bv-)kMk18
zwMNJm=6H@tp<xML$<%@%LA2Y6sm@I7Z)l7>oP6n7<VA^A<L7vH<+&<|jH`otKv;p&
zU+niX86)W1RFg~psEWJt)?#+s=wd=x4lm2gz)lh$R_Mqio4cE8Wfq$!%3bZ5qM>*j
z(d@na?{wM@b->#Up?8`dPr-2L>1xaiJ@#Rng%9&a8*<AYX(>_pvs(*tbLkDKf}`&X
z(o{MBvm>Lp#FyEbZn6sR%gKd;`^BRm{@$8sR*Ahn*G~)c)PBoen64P(`;)Y=HzHzN
zXtv32CsLUoFvP=AUUnFASmo&l-zgVvgf{UTwCDH3nUsVc1@>L1KNs-G9(GX>{mw2`
zdUvLWCT<6@%266IYTMF+^uW{XrlF^)+jcljHp<jq1Y(3#%4X1eq~bK}-3>j<&);i8
z(pr#t#Y`R3B_~V~(r1=Kyn$Ji%9wTjH?)!1#CKO95eZ@0G^$+Wy1Pu6yAOXuWqG&w
zTy@^L2X#C%9tOOIw*J88OwKC7*yo6G?4xu^j-SbjpFet}3#ANf-_eEFdf=ilB<1zF
zjE-@A?J=glh#<%!5d6Cx>fkz>FhD82;SNI_BNy~DRRsJscRaW~4YTl}LLL6g30Mdu
zu>a!z^jx6#CP&i4l_$I{l)jkP_gb@S+>4w`Ru!yyu<rB*U;5n==Cm~4RO0YTx}mb#
zF7kulW#_PLzYecKU@DEzIH%V{BTpDs0g(Al1z&fRdx;=`<H+6I!03}%<u_DUeVeTs
zeS~^{ElKEVpM(Vb%Ls3Zl|o4C`~{M8&$aX)NJ#N-6(YDEPO7QGDi)>b<;O8lm2V-P
z-_X}x8<#6c(e~xezBvmv@*?1EYg)Poe*l6dOE<!`2ynC_eRh$p)U@Wo{pZ}f`HR&N
z=3}Dm2kw>cX}0d9u-(CW?CC!Y`{^~_=m!9&1^_rkZCRcF12LYkDnCk)PGTX*q;H^Q
zFRkwloa;zuCqQFpWiOvHo-MI3FH#-z5wC>cAN>QNZTfAT^$+CIa}Ww{%o7xMv(V|F
zB&1xmBowoSo$z6aTy|$6x)vQvsD9CbxuGQ2%|0;%14<}mFkQmMbYK3@bp6Zmk1?Qr
z_sTq%4COA`AiUthIR9MkcnnoOC;P%<f@Q}_<6o>)U=+o_i!>UOGQH2#+)4Jpn4kEk
zLGMnuyFKwT_5Vx?M9lx4)c0UIeDZl;3QX5Jbfc{QaKx35ov=dqF(>;R`(1Lr5wI)&
z^Y!P<!PlRd0@GiJtx{nm2oF~&dA?A>Iq*2#UF@RiiE=Y)UIvC2$tp|t|FhB||6S<{
zaLa-!Hn`BpE0ySsY2r7$Q&yh3wCMEg)F^+vLpp2h19+t?T$MH9J|P?gMMPWMANHE@
zj3OA1@3gI<_VsUR9U=ZXTA(}>0Cv{eIlapN`O7&qUObB0fdV7@yTATw#3EgIUWmQ%
z{Pccmmm2Qcxnz;(Pr9Xv*|Kg~p!`&~dp0=bx+8|Ia2SQ-K54VRQ9@V~E5ANkh6J23
z9QT7M`dt(~Wk9j!h%hl1a!q<;fexUW(jT67OicK;tlCf5zaqJiD(A=@fU5iGe*rg!
zC9W!~wx;W_jfcMkcl+xs`U{|~Ixb5buX8NP(wnh%@=6<XzduqBb}17Rw9@UdZj`7O
z?{4|kMLRD4mGktl^2KZFrfKYcPs|_q!&fcyXhyL>!$vhP$vSQvnH%Di%K?jMW%Y~N
zg37}Wxu&Vz_jQ?QgLitTKh+bjvY%W8#J_I~?SGrWGNIW5-+kr0J(}II)Y>hu;>^An
z(ZV^l6H<vp=hE3Xvh5ll^5aJG#!7YP`qG<=)5%%L3IxgbbCjY%bIdXL&Z!~bz6&Ob
zdMn*)lW~H(q%5XiPI)}xK_c6y3LMS{9~_SlVg=D`=09CY<9t{i>T&g^>vm1bw^3oI
z8x{`p@QE0UOJZa*_6qTTJ|9lb>+;w1Y)(Q%ceeLngp)_4iIZ>$KG7?Pgw9^YHcpdl
zq}S(~R?B*)Q(D;OHDuEZgOnRN-#!|zAwJ*=FqCi48U_h`6Gxmn(R+45xJK%^7eFd(
zHC>Mpi-k*nvsZW(QKVgPPpPQ>?N*~HM-2m{=17%;)N8;UpcQir$4mJHM|@bpy3B+%
zx@O`o3-K|4{%%3e897@1jB?$j437TwzZtIo&)B5>AyrJRZuIzxb~DSGr9iEnA>qjE
zAO78R9foslwv(5r-y^6rls0-$LZ`}vdm*$AP1CH<fS}I<NRhr8_qvX6p21H&(YUa%
zQA9-M&OZ>RT&@cWA0q9Gy;;L}mK})Py&QP&?<s2EfcH^p5nUmz-ROXZv@|h7!Ch4<
zV2yDx_;a~ONX^G{U;ReewBcLC)fQ!JVGHCJY0dda?lq_}R*ZRa!;+~y=~4WqlM~-v
z!1C;=VE?fBF`}s^-+wSH7y?&|?@*9B`I^L;YkhgttJi~nR^sy0=L<gUWrEDi9Qygb
zDpeV1YC3yM`H{*XI%Cv`^KiSX{^W}N`?<N%?dP7VSMMlStuzwhi?J}M%V()Z{xMNZ
z;7<%pblcMYgGG3<A%dFZjOyw~>R-2omaRZGKpiLL7US-~Dq7+Hb;^92?~`*rpoylX
zw(IGuCF!FlatR{%0_;<K?9${KPM|2z%_E4i@`vR!)KHz)65H+0@6@DaAE_Um<tbC*
zo7z>bn_KMmCzvIcj$TcO_h8p2=O&L>m00Hdm~b>8Hz>GUvnZD29Lxg?c{J|*4p2_X
zC*AlnbqZ~|65F=O5PEvzBMgT%wHv^-A@}qiff1J!+sJ3JG0$2S|G2UdA^dGtlkRRe
z>sPiu+X3HYnq0WwS|dIDGHyN>M}2qHr(Z%urnPljs9PYg&~|R+r%PPnrl$wOo%FYJ
z@JKE#+`Is8&@fb)=z8>UWb;OhCdZr7Yo55`3VBjt&q<C@c8XMe$z3N*ZXy%LFWLnO
zy;r!-m{F5rI_MVYzOT{E8_9PCdS^h5qtjiBo6B#=iunRhzZjNaGI|gsXwf+0eWomD
zaC7sky`Nhw-z(38Re4XPP|o9aJQwa>M-om`wE~3$T?S1To{8k`S`LixXhM^jcY$sA
zT7oy<b*d)pQHvETJ|5#+P>T1`)z#=YvvxOHy_sNZ{q%Orl-H40N>0_2#f%necKcjf
z??fv!t4)AQWW@Wa`PExG+1F|AJ9k3Y1c@R0zl<k=r6v#tovIQZ;(Th_Eu8d6s?Id^
z%gcPinVq*e8D=VFw4U_RcBXAAkYycu;nOf13-JmX?~9yHqP|!M^TnCC+4;XAVw1C-
zg*g{4`}RXD@PLWlsenW$IvM`*?y3?ZKThm<kn#Bb6Thb~cyD!eK28ajU%!$IxF_Zv
zIjs1sf{l5;5zI(Q39Zrb_m^+;?P9THpJkk_gFQWl8>>N}_-|X9#f1bedV>7|ElS(X
z2_y2LbsT`aa?i_u!mOVnG}Kyiv^%twv=5p&+&1fRT<xGS&2MiPEAxiYO|g&3b(TFN
z^fq^Ux=dE|A9=KprX*j5^yA;|(sdr?C`}-}g9McBw(jkj7ZxD4nei#TTer8b^2+_0
zu2-3=SR#!_@nXBBo9HGx3(W614{1bU<jr0M%k545xFSEGkYJbXZ`Ym_7cA`-tIQWQ
zgU<Z$wyKmc{-edtbGM_2?Tyu2{b!=u+q!`2QG@29-{7JV7?v(|JTfTzS?NzU{}9Zq
z<uW-t(tt_<mlFYDUrV~!<5?YMeJ7SF9A|Hi4D=b$e(oJC7V*|jKc8#3!?r-X0dc|&
zoLvOU!M_ytxkpKv=G8c^!wzBe_Whe>=T{KQ4`-SULGtPlk&ATXDn1(YyotY38^jTn
z#Q4dc-<o06IQd3T-ZIDcrRbl$NTk_c<dol%U>uYV6ZFI!r(0q~&i}q5Zec*%;z&Ur
z{^dGkD#1=rdG_oU*mP{g!2`K8rP6F!fg-Fzr}nck+h0&&jE0*n7H!m@7&aHm7g}@a
z@LZ;2f8FS3?pd$R#G90&yM=1{N}j#pR|UDmL|8|#&$zNEZ2#$)DAKB(wY2O*?=fSB
zfjYf)`R9mGQhh4}ed+Doa+M^M4ss)4H4W)Fq41%^x5+8-`)oI#E6dZ+wGVlm(Fq9&
zHp{?=L})+j3BC&3fA3LkeF%fob*W*dz0&W?aZ=r<-##u9{jtOW)ES6Y>Lv*MjArvi
z9z~rFj$%B4of2J1UA!Lh)9!MFvb1FCLVM(Z(mhNKV#n>fs8xGFpRQCQ4}!d#H__88
z@{#+OUt}j@hiX842!8-oe}P<ovMz3R^;MishJ;rB(Ir{BROQE=uJ*)#SZ}feL3T+J
z^B*zbIc+|)^!8=iqq3YKn-|GZT2pGuZPG}vkk?c;|Kh|2QkdF$Zx>(Z5G&oC{mYWf
z&gD)S#BnMHSD1hkecYhL*nh668|6<ZE!QfSatw<yee<?^?iXF>vQqUA&^hv%eK0AG
zs@S$-%Sgeey-%^jIN*L?Ek}pyN6KlT{b2n#OT|=8FU*!V(@V6-W}HM2M2<_%%J95f
zS!|r}WO2(#<+_zl2!G@+ExQl#J%$q=6rJfU1Cm|r-oy~jjc0@C1?pK?akIi|tSz6%
zv-z2KSO>{U88wzpSD~-p^)VeaflE!%oZMpv1@{}Ps_Y`??p<~crJ^Qf8H5Je6Jv%>
zmEjj$U%P)Gf?FZ@t-!&sMjBY$;9&r0!(K(_<W#agLFt`RnKga`TEbhtY;$a(p*pu^
zB_7f`h?_X0Vg;-^(r&<EG@e)}uZjDc#0Zzp#G;v!0)6a4_&@9O^dZ7{Jy1vT$q6S!
zs$3Gl<4XBoaxr+|ZN?^oXXCfCECS^>Jn7PAW^m&TozwbG6wLbhxt2}hB}R6-QxvXK
zdfofcr3=8sm<Tt~YJ4t(xBF&4t(LSk%)An7qj>L)`;6Bwx&#}X#Tk%SP!lBA&^srL
zA$!3S7xKS}rb}?aLYUA}PR5rFPLrmDZsEejpFDmdHN4>(?poO4hAV);^<VTv2ry4o
zh}SQu&WyQB2MF_1^HG$EMXUcuzL=-t+{CASd#acQBLK6Ip0gZb{LaW#j#Zp4EN!>k
zBF{N$cXQj$sP8qV-*PP}MSE77PMh-q7dyG{{~iA3XB6FGEiR{i$%@jx`kb>VLto+b
z-nD|zt&FSmk;d$WeESGJ@od_wqzyw>S-XTMxJ`q+Sj?Gs`JLySe<y7M!nJrOaaL-<
zzd~pUIxltym$o`G;L|>xF~X5s%j2&*1(xVVs+iforQI3~h73qtB^?=CQmN8x3%L$E
zX=~fZzr4D=?Y)+*RF-0GX9{h4CwU#^4-XGKHSYNVe}Y?S==0J9PeZwK+|T&Ft|h->
zlH4{KP91P|ZPzznrfi9W?n(XC8m8(y&$@U?ZrUZ9EVAQL?TME*+@`RYXA|O@qM_-(
zw;A$NPi9M==a@t$XH%gr1rZ@#fhv<0pVqF-wfj1lEspIpk#lxhm+4)Z<jNzr%U#UJ
z3U<)AliQ6RXGnq*67MKRWn$0zw&MfG!Uw708F_8yugzB$JxK>(AT*AfIv?bsEkH%F
zRMJ;R&Ka(y9?qrT(HFj3%Xjgx;u#eO{YXv$oVPOT&)s&E@t(tB$FMzgIkt{MP80;^
z*7D1N*ws_9cur^qDtrlQTGI8a?x*(m`TNf7OK-C8{3VUB{0n+oTpUMAO|Q4SEewLh
z#Y7U9cV4Xu`ZY7pQuz`(lLqCOA$tvqhraj}5T(m1u=~1eEm>l8CRwaGCL=vNIYU_w
zlcxQajAIEdg}(zbBDY>P%8W1AWKzrD@LB%h{w$I=+h<UZ+U<kXLOiG9k5Q3rV2Xs=
zB9EqdW&urUq}DKSS35xZo|O;+F}mbWcMmf@t!J>hpvD)R-<Gm>@-||Aic<N8DtLFV
zT})+84Aj#TeF|q<{J*GMyYWV*u9|moytLdjWG_1*vI>}3(@4X4b!J)?xWIC5j=4S*
z<?61h^M9Gf)!L(XRpq>d^^EC%-39hi-0_pc2HZldlo;b5R+;GHd583n7v>!@o|O{N
zM84R4h2#gJ5s3F|IpP=v(snBjhv~Vu7#Py*r}zWvG@n4eA-O@Gly~7*|AA~g2ioiV
z!J}N)p!(Plx<3I6kz1I?Yp;Y?Y9#U=W?9TCTZD2SzqT&ygyU8$Y0q*{!EpVrA&w3|
z9LBZX%XK}E+`D)U6_a`pCTR&oh&-FHRK=v}{Z3_I!tapx`$V7uD6T^6N@vJ#Wit7*
zYSVAqawlJ+p!(p{3FiS};GyNM;WbcFDKM?Qw8ZeK9+e9Fpxr`|fabhF`>$RJRl*)-
z;njf{3D(&5Y`3Ar7<ZOz)=Qn8B4um(P6kYFXKRG>@0v2nSuQF#=g!dT7;qKcU=~ev
zaOJCuQvDX~H>WAA-9gHsHE>1OhpVavfzthq-`5{H&~*gv0f=@{aa+Z^1KZnq6}Lzd
zZZ<=yoO+eH;16+CHHil@0OXT>+_3BcmGvvvr7Qknz`_y(r8|ILbk;H664?!G4;I5U
zG}bP~n;@6cr(UFrDF8Xxy%xpEOH|wQ(MlC<1Zg}mc1EAm-ng#gJyq@tJAa1JiiS}G
z4>r<i2KH1C<g-sqA4FX!pRk8t0n<9Jawis<=b*6t>FcEFOvg~CwW3<>hkN3ld5lWp
zIUmZ>auk=S(St0VILS8mSy+ZqV|9H~_^`d0>vE0^LYskY<rOzOC-ozHVi1qp0~(na
zaH23z3~7@|UN@>4(rh5XweCvX_)2vM38%lV`IGxO-5~rX{u(x(I>t=k3(}<R>!icc
z<9TSQ7kxgVnu5lVIXhK6E%b{1M-@8|WeJ@B3b_n@fpLqK1%@tBC74kcX_A=oSZwS^
zn?&nrzaIm>HZ&WE*%bmSo(3Bvgpo=7wsCmfp)NT0PMtsNs=Ury=#6jU@`AO^5KcVl
zrXwI;HwvERfkdc!5HpVi#`8R*2ODfO^pDzXy>~@f`8~~FupvzL0$&!LgE|up%1X99
zh_>StZT9t(UBH%`6ljsB7x^G9nLS(Psi*w90-Y*{Sf^N6qx6I_zvJ{^bP3O>vs=$V
zNsah@YsI^7cqTn%)_4;@ky(G2d@(73poEjuoq%e-VEkZR5X;y7GY*88ESEwPJdP^T
z{#cn0gQOiUgqy(VC(vQio9yf`N=B{2JxisZ!h8NKs;fh7cvzs>X@iRp)`bdLv<jmY
z+<f@pgTB>h6&v}%dx5x<l{I1;u^Vq^^+w*XBhbG{r{4f12X2ltUgSdT1oLc6Bci{y
z7yUVjT&^kW%Me$u>GND6tB-gq$nqPe+*Wv&-ADmzYF)Vg#0saBr{GhXEn39#f!}{>
z^!t?#ZlcjtO=se6E<^Bc7~p#N%h+CEdpB_+`kwJi!GfiZAN__8{WGnV!?%8<ymAL}
z-v<vJ@rsy*Yp`_}uXe^1hw<GRDe=y=PN9c`f68JY*4ZZ)vA$TlTWpSG8b&3SHv&D~
z{<fitF#fB>)}N2&v&IS&SsukAMNZy?|LHB*emiRcI#%j~mBpx!5L=dJO6tkd+H`l%
z8)0qB0=bQj018M@*#A+bByCmb++4v?4L+@{`7yjSD#Gh|qYfk_=;EpB`Y6uDqrV%L
z-KTd5f;h2oX<ch0-2FRK(r7N5s4-j9XNf%>cOSha4uuTp%M%l;$T|WF1Kj7Eqn~U!
zINxd<W^3cd4^3X`+p0f@Fl;xK%Oc=9-<>j7-+p&u$^?;hu4BUdgAuKyTI1-KdRLVs
z5fb7c9+n&f78Ms|12kj6ec5qbt;x1^edI-`Z`ww2LSs2}Dh<Dh<BCW_8(Rv-39+~>
zGNzcvG8%P0;ko$$n@yCx8{m<f>7x>7DL*Pr9^VZ&Ep6SG_u^<cDATA_Vqge;q;xsx
z@+lW=1Z~MZ3RnBOQCBse??R#N!Db0&lb|<DTqfuW!G87U&{KuxK0f=<$UCzzd(Y`}
zyFZDiL_XT@$L*evyUmhOEbNP+PV-H+$S7r=U3Rmrwb&)ZaNv&U&8^q*yq9%5{cZXE
zfrK1t+<-|BC2Ia0*M*DvB$Kuz;`X?5<}hB7W&6gwge;$P8}HAqkCXvI`<`nUH-xBQ
zx<A;`4C%(A_XA#+eH?_|{z0Y`3Nda99=O<X`hmL>7~zwg0?;(?<78IvEhKZ#Qs0wm
zme^j=m(MQ2fJb&Hb&(TWN5K6Y?%k-GFKxHmL_b+8eeS~YP^6TH2SqvU4`VS_iL05Q
zcOoYb#im8IqvY-7lbg>Whr|qj1veuwXMX5&jtLi(TG2B{mFGF)Bom^RkJ!bzYgeL-
z6V_NdA3}&0%R$IOM{8ubOYpqOM{2$Sm;KoraU6GI1^T}SmcmF|&B9@cyx}M3YyGDZ
z3bnCMUD)1h8|Xi~<SstoNm>SUI9^t?a3z&E!nIO+-!#q3?OU~R^85q8rbz7Av4GcL
z;a$~5Fds+46?bRQtkrLCv<EpkTC$KuO->QZLD}sZAaHBo#_$&>c^)|istFrz=kiuh
z&il_VkHLv{Pjpo*Y532^u=CKHEvV$j*Ka-75<K$9B0Q!Wg;i(>u3ot;u%z8laN%K!
ze<P!tXyAGa6<o1@kN?VxN(YJlDRue2#%U9#pBCO;Q5&5(DX4VCD-Lqv8%G#dp@BOG
z;5*T|cqss=$00w%S#Uz_w86c>d=p<s&0Oprq8J&EQWJ-rU)8X^%SA0doV5NenX}0C
zXviwF6C;~ANElEC<dc@{1%hD^+fr^!qhQwI(+QtHRf1`zKYlXanZMIlPm%n6RcZDb
z+v-NMpdeo;NUbl74myCl6_w?M(E*@>6D)#?Po-dKhB(dEEVh2%KifMznL2LDdKP1M
z=*!LpT9$M4pQNR!au%sGtTa_3WB%$>z;BL<c`FuFqEbz^2zb1=vE0i7ygA3!Wglkf
zFVZ65EW5B<cHxzEFA10@IH-V4ZKVNfY_hQ<hfjv2@l}FJW{p2(BrQLp8%<JeW~uFN
zNFPVb^CfVCB(z6p-}szSHW6rwJaK~Ol!6zlyWJGp2zJ_&-Hf)A#;QBHkKXfZ+jfv|
zzQ>ZBCVx>r2CBd~p$^b>c98DEcf^+D<nm^<tZe3M-z2E8k$OTW?}rlkC*ZjD$b=j|
z5Zbr;e>`C)={4)zQOr3DKG0|^&lI$Oh|uMJR6j{88P+_Y_=xny#gS(za;i^|8$aiU
z`8{x>ERmf5GoqMf?7{N>*-b^$m`56t;lOXX3v#LpBi#zboO!izvj5s5s1(jREV$*U
zLB|KI-@GPM1qn^}8uNho2zMw<n5&`_-7S#FrHWv9MP?JY;aPZX-F<m65QOZrjq(JJ
z3E;%J-=Yu_ox=vovy{`NS*<4OhoRQ7<i+BU=k1`cW{&5b;OEg8E5wUT`6oc*Bb>lY
zI{MYNOKG(F>LKWRO1G$_hc(|-iLNroL~Wny)zoG^Z{2~7WfA}A2D~1QCeRO1)T!`u
znpzx*AK8d@5Rpj>v}}5L(PxnkXrKC*f`otPunv2Fee-Z!CVBe(rLBdsOWU2!Sh-S#
zn-Jucv!Y-YSVU*_3CBXzOT;>+xtBh@|3*h2vT~aOO(o+!7;uyc?4OIdrY0~-?VHzU
z?SD<T@kaVDH-x223#RitOC@~<(FO6BAgFN#->KQ1>x{Y)bz-4F>>zx_FgB)H`MKy_
z>`DvEATLtxM6}WN><8QyLwRYxZQ3=m*wW#LK6?gjLv}v}qUUa?g_febK8NJsi%@ET
zkq|qpm>RQ46jQ)`Bi}taQ&G!pCX)<k8QF4n01>za^Z+v!G@+)Xd%BCdGv83!G$l|i
zQKEN6hxBrnfqV-~7nWs>OWL7{9twJ1G9G#0E1tbyatBsre{2(f1V~6;Z=0NGIG^B@
z#|gxK!-wb$`2;pdOgmXs^#G*VV!`7XlL4DOS6jASS~dm5KPTl>o^ZcmOyqgf@TZY2
zK6n1Q)NK|Low@4bbQQG|A0*0YKmGx}2>U_Opt>QxJ#%)8IGV=7N=Hf`(U*V*gdyMm
zpo;&C&D52B&I**N@$sF(+{1sHD1vfUFjb6RW?dM8M;x<qujIIQKwGKijkF}bUm>p+
zIQ2`WG>g>iUc9okQdpklYc#Xti+qwU_gk*JLyH};aM*~ei8-)h{(Yj6SYr8S-0Hb`
zHDj8T@SaratzHnGb7SQ;EO|bII&5`riRoNBrqt0m&d?=;sa%)MYbYg_BuE_+#Cxg>
z=f-g%Q2Jk;f?3Q1b4B}1``!5BX>`ADvop$FWHZ(C$JkyJu-(LOoYN{vSIG&NIH!91
zf2ZuxE_=GX^mIAr4rvh!x0i+0RtVOL)_z_VP*m7V8%Ih)DpG%!=Jk_HB?^Djkx15R
zrBd)gvh60Jb6HLCSr6YrAD8sUzkb5ezMQ7foPE6u9eyY1Ch>2vY4b(EnuB{_HblJI
zblf~Y54qTeYcmSmE$L$uc2`@#ThF78nwu(W5T9_tNgl=KW|NaGo@u@EKefreU7sW?
z;LEngA*%{MLNXuVj}Cv}nqs{dVrd-AjA{xU4)2(D-srSbyq6ANE;Eqz2JW@Fa1neg
z2H$uC@4|-@MS4*j?V9&8^v|Yg09+x=h|gQ_k1A(3py1HsjN?%OK0B2BRRFfQntFe)
z1QaS}2$~dETSx-K73g~4;vlth`U-GGwb80B5oc#%u^XsVHH8P}z0a2}6|JL*a;5~c
zpPyv_Ie>9{Zsk!D>WLlB`q6OZR^a0wBb%QiUq%fqed2;@ZJIk=1oB>StIAYxbsRi|
z4kb;J!#n1>wY~GHGKa0cAV(IVY3TMg7P7b-0EquVkFDr~M=YT@-Q4y$bDfdDXAe4B
z+qtz~RMx_bSEJ;aKjgV8zS|3md`Y0fM`C~p?$U;N63u!>dQFD5i<GRNzJi4K+KQ*D
z@@p?u&PW_~(Jh{!ixV`MKq*m{u6#J@z&oT1#HLH%%@beSUgMpZE+R03tc8OZqx`l6
zA5QMX#&*RamM*W9_jzdOE$nyCAfb)An?$#qmU5@5F{JS#=fWG~GL}4J!X=x_CNH9f
z|17CngdP}3cy^}s>z>NtPuu;hn7l2N0SdyJ|M%_Ad-`8<bI=~u#CO-3v)K!mACao#
zq8FJGs&ZrM21-A+I|RKbeBf%>w?W6_TG-QBsMJYf{i%yJ1QL7&U(>ok>kL=J$ZeHu
z3DX*UOA{`Ygt~AJRqjGOJ%a$t9J&K!6e>)}cKz#S)A%Gq&_}M;$^Lmo9gOC&K06O2
z*TBCtWGbt3{&oYtA@5HBqN&Jg(u(holx&8W-0=&4A`A4EVzTFj=6v#V<5he=j;?(W
zcMqL>4<lkUK0kb74oFpXu*bZbx0`M!OLhtteNQ@kiQ9>DJr{APGy3#>o#~JX*H@wL
zuc)W2^<nfo#y7ig8FU^x%xBo;Y6s4y=i(r%5QyTUT|Z1e%uH<0yEARkP2s@qDN#5L
zQVP$2G>hoX3hW^LIv4%N;VRw0$D_icTPqUk{_Z1-pBWAEyKj5_nP`A;(W=-X*%t6S
z7EMDRD;s_^pwhSHbzARy@G?ysB0MFN;IHbrraBj!;-GBU_dW}>`Biy$K$>$uwWL%?
z#j4I7Lflk%)R_PW#XRgaWj<KC97?BnM$N5I-;G{o!t-uSEbr~mZ>qmZrRUFkEM;*r
zpK!Mh`ky=l;XPYCXg`hm>F}ew=dG~bd_jTbSt0SJnb1BU_o3Zbbg>^mz3cBgUY)6z
zTBSRDaZ-jaKVCL~ERr1nCO%h3Bi~pfkCj6}V^v6@srK{zH>eL^Z~y*HWPna-<NOdp
z207XL*Y}-(!GScNb)9u?Yqas^_zi!4>+<<MEo)2nOP`Y=1fl7FAU4!&uq0F5_tGyL
z3hmopY`hQaFKxeipQ@fJ&1M?Q{bE2<?+sB0NRcY6l3aWg3#Cp2bvC4uV_el+HCGE>
z-PC)EEnEY&Tg#?XNjNW#Da^;h+7Kr@wKb2KHEVNY57*#zf4Pi+EJEZ3v$4b~;_D%F
zaHqnBL4VmepLZ*7d*6f~YJFkZEX=#i*$T?>YQVARZ|-%cY0HWdOGlYlyTd#t&-d^7
zsEXTUw3wBEiQu9%aD%{HZv5my`5`}lrCgq7R-}hJ#jDN5IwfWD0O@@8E|?x4(JX5Q
z_}jMO^X8RWH);@SJUn`sl{br2?xdE~dXjtNUvOf8A+bNFgK6;Kv=cQD`eUs=Xln?P
ziAxO|hmS5?6Rq$%B0fQuJCEbW`z+cHyUku{%d<<;dNgC@&g_ZTGlljYPcfQ_qT65V
zfP30{w@^h22chtbpd48VgtIN~MD|b~=i7vAiWT<hH`;j|K6y`A*H7rd?|#(9D0_&T
z#DC%ja02hos+Q)S=<LIjYRr0FRsj`1fMG5n0W063Za=P)u(LKy;QGGK=*+k}-0F{c
zDM6g)+>CAFeni@C%vIrrHreJwD3Y_JFn{aHh+$~nmQZUsf=pDyDe;mec&FxIsS&uk
z4`u#ZtVrTJ1h>Gc#MKqm<#1wJTS5CyGWj*)mH6D4dFdQ~fwcBIygzfj^C-VsY9OSU
zXj2YZpkkfN<kZMJg3`4`I{GOj_)@#>D)mKaGxhEJU0O?3R&N#r9#WkGXIbaGd;`S=
zrdpq3sA|eW=dpb3L8{URnw#Cb6Q+3or85(J>_?qB4VxD(bZ&bo#1PYHk`L%bZ-`f_
z>S4fzaDL_dmSar-S6eR4dVq6|;Z|uvc`JTkoV?}2EMfqzxqARQ#fnyaGS^GrUofwH
z+|A8zSeJm!wy~c1b$}f7!$)H^1(ktmAoU|@mQcV?FW$(V;zCak#zHR_e(*gHJQm(b
z5Z7^UU!Hz!emO~`m$Q;+9|wRDUPPTC^88Bsx%74pDPFO}gaDa9aTLx?&#a%Iw!Epu
z*oC2`g?$PUD-=Ke)PL#(2k|r6PD)o$KWQRE-TU^3E-2@Go?$N^LK~T9$Y<Bcj0siN
z;HpSMdxk7CI3!tBgJY}RxEA(gv9Zkd6uJ>i+0alOZ#rgo>I<1;7yWvpjFB|^m{(ZA
zO^&iA_st-hnkD4d42C^q@`2m=?Z(%;ZfF&H+08Ockf*MZsZWs1UuTyczT*{|3iE_?
zC6jBp6#jp}RSaq$6j&Trbq4q&tPBwIx6iro@+H&31DQ1uqOMyRW}`Nzu1u>4L29-V
z_4>6Q$d5%vx!EiFpk}oD18xT_E*Sh`fS951%|(94u!M}^6i7SN$6cuvErw<`BzBdD
z;|6SL-g8@;7ddM!h_4moL;U>+l1XTUN4^;lfjC=Xt&9m*b}f!V&)KJbpP<T=j#IYz
zO=?Ygb8B;!Arb+0a)VEHkp$uYd_yh*i|)nJ4Ed^Dhdv0g*(i0WGMzDwTQ(_OQfZ($
zj;o!U@V&-XcQoMtk-D&Wuu%~&aoQO|B~}?EV3K+opikS|+VyDdRiItjpPa_3-Q52S
zJcs$8foH(E<=apMOIi|03o5nH_JeVnJRJ7yD_k-5*mhUQ(S^W)CD6sHorf7U;n}$G
z6ZK9Zq|(K?Q@uL777b&?PQJ8fHGl!jn~^fjJr_fp%=YQFWT>aoA*ee^p0o5~RL(AK
zsqD&#;gCNXsqqHb7B0#Ied_r^j`p!U*I|@3dBd^;S<)kNEYyyrc|fQP$$C}{*Drxt
zExn=p%``kYYq3r<tt}<!94^UTa<|4%UVYgMXl%+SGR}vBiANi`2`WjtL!3MzPOyxW
z&Rf&`(}+kv?<cX6X@3yMTr@y-)r!kKqnrndjQ`~24A`9*WdW}8g~+O(Ri>ZD{tC5M
z*4Dp1_LLCI?Rs%!a;%*yz5w)JP3J6~|Hterlv!cEZr5Pc>c3p=5STzV5B5}zt47YL
z9yuOq2}hl3!%fN|OV9Cx=TSQ}A&UbSzsy*CK6Zz;{ZtO7XcS@ivt~?M+d!zekUE1m
z^gGP9h2#xOhvZ->BuG;|*fHck^UUD}^ZfVsMT0|j=I})?s)SXsSr`PGpIJPbn>Ek9
zt=a0jlyTWpQ?3J?u?o1{X{V*?3LJUc^yA$P)uTG^fsSdTsnsu`*NmW-WFjI)=>}9u
zv3X$kOFeIEZ{+ThXBmCPyEkWb541Dfi4|$2A}H6FjkyYM|IK=ubLjMHFGqt$V^(7C
zi?H66o5X#e&$5+ywQut}j6W{EPTsHyv+@epQx2V&$)O!qSptTPiC@5$us<Itj~W@6
zs_5JMCAw8BHz1jiNm!4@FeQ9`%V)*tbebH+R&~t;R0Sth?p+iG=T&eb$W~xCMK|eK
z!)VO95?bGoK9-=jeIx9rZOh~LY++lfTu%XN@e4$;C6bza!6Fk!yjT7mgnf_vd6Ko&
z-QeM1MI3)d8yLmY0W?~9UDuyxVX1Au%H(IV_#*1rbxB@#rG}ft>mYD2o56@BifrG=
zQDNA~$GWd`o*Nse4c;~hSBhyt4uw$6^@c1p+}>M^%ES!cdKQ*Y!7Z3n^pW!Rt_rA(
z88u=599Z0IyL)Z5LYq5%UYo?u$&+1NPDwn;V<!m?gr5u}O|J%HWSI_k=cdfe%<Cgl
zed`j@PC4d|+1XUB4H<{;K!N3g?p>D~rQ6V#UcEYO?$3U|+2)74b_;v5CW%MFGWNIG
zP2!~>-a|*+-|>7G%CoBc#4lYWH3BPqdn5HOq{1FEO@$j)-v3{k;Qt@~4?jB9Ca_I3
za>Jq@j-%w2zMy12512d5^xh_}kEZOBpVzn}9yS5>H;O+M$4^_B4N0?jZzxJ~@7tN8
zhT~^nGtUN<D}QdB0xJVM%R8q8P{g^wh!2uV*M3z9$Xn=Y(v6+KNCRH0Gzk>ejj(+c
z^5A}DX?+@p!J_W3)25mds0!h82g4^h^wxq2J{(SNhiIH-eF1Du8{|Aj+XKxGMP4g2
zJ=T37Nz@h!hp96|snHV&FwIWRTR(8h86IvJ$|rs1Z<te8=ClM!7+$K8(B1(qH%WH+
zBHSF34{m}c&C`O5xiJJ6td`FLRj-|dw(}1iioq4;9&hgbpX{gB3>^gMSvJA80Jfk$
zDjhwtRZ{XDQ<pOACY_YtQXR>>kq~GneJD*2Aqu^G!Hp3k1mWp0R!nf!IvjQMtBM@*
zN$mcO_k^#JDGcHZ?mTLP99o?$aw_B@c5X}#me3(-3r}-49TDYq#SZcA)rxnMcU0xp
zH6(Ks_J1t}kLF0=D6E*IIb>!`8>`F8;uyl&{4@D2hIq>qJyLZz0bg#qHY9>m7AM6;
zkCJ2t>|>s~rl@iI)t?TJu_<FF3@<hBO%$%*B%U~j60YGi<AZN3;G^4yr}SNdQ<xAA
z(66zn(n%agm#(wpCYZ5;l1p5h$Cxc@%$1}Pr*adsDU*ky?>faY2aG>M#1l>qJA>c;
zZTMG^4AeLsNukQ1r7zmT*!-8xkDnDywpp~jD^f0%w+!S5=LUk-tpKZ=2x_?G7BlVI
zfb6!uoFD3csZA_%xPq74a`F_tHZzafIgG|-#B{dz2M=(#JqjNmR*z!-_~FY_-B2u<
zrZ4nw5uis8GQ(oe#eSztZWtdiW<+Tf#Q9IF`_NQS{hQi(sd6N>FHG&XPw${PtWAF9
z{irP;6_Dlb6?-$3zW$nhjkDpb&M2ulB=r2|ZU)+k`Z_+!{H90=twC9vAec+X5Vrc4
z>>Bi-6P{FL|B*UH`Bw12x#$=nsC2!KZ#rlU@yYYZakzIyg6U;DS@wY-%b*hMx6Jt^
zu%NQ9gL%wg(P+sjH8*SatQ%4K(qHWo(>SU_nX~tQQFx_b|ADZ=T`&&kxA2*g-`B;}
zFH|YwZHo6}S>+9ho`(JMSr44hJQV|ovjM?!_@FoO)yZndnV?RLtAK-H_}%S1*MY0H
z><OG-8X&HOhi4z4FTUaTEdUeT`?Q$j?w&c?fq(Bka29=pRJzvTH6RJ@p!2~dVz}<&
zAm`U6%&n47QPWIX7%;JXJA(!o8DtH5$h@tb1$LzZKK@hz*83PYP*eDax?py4!J^lX
zty}g@Ti?s}2Nz#-C*joC3do4*aDw3kC(&BWE$1bKkZ;t@Pig%UOpB=+GGR&Z)y8x%
zm;|nAa{;T<#N>Ds@C2A|8_P=QvnR=>qZJujp0QiYZA{niOtmdG#SpGxPW#=^p3@pl
zHcs}RA_=W4@3mjTI%JZbtxY#3U5MEB<2yhjB;BKUX@_m7+P5oCMOSXr!p2lzLdfzU
z5aN$Wo3kXi@pnf~g|W(pdqybK4;Sg}1V4%Q;x~F}qb`?gQq9S!gnpy|ZH}yQ<xWiT
zU;Dq#O&z~(SL|P5deX%&Kkfe~?XNUwTWQzwA#e!*M$#f~Xg|i7y=gYN(vNPE?sl4@
z<_Xr%nOg?*AA0Q_#i0B#r8<^Ps9GFV(VC8=q`#jCE5aRO$W4^semO8?(+ATZXn?Cm
z=B=Am&8_Bmuhs2s+eo;)wT-=ZX^J=cMW{RMfS#z&aiM|<A@Y|k#`{=mV@Cbujh~YK
z8tR)5*te9YRGxs|mC^*R2^8@+$CHAX)hcb#=cmHHyRz>*eu{g}5{)pgH390AD#D?h
zmL_JM2tdS)Urr@pt8B-Zau=EPDfW^l?+Fz*4*A_Z4ptii0Y^{iJB$N*PbH?H3gMzb
z<iP%jOVaW0ygtW(_||x%!6?)RQ+oF#us-+t`d9z#`ZsmnEQ00X5YA@frt@x2p$Rwh
z_Ueq2{JB}i!<Xy)waGmOdKzlZIiVgSaS*Iq(<>{Eb{}D*0^Ofs>?NLOUtawkq^x1L
z@kP1;>m&vzE2;_eIvmA#E9SH}kvnu}y6ES&$WHxN#lVR!<3>50?#k?uBtG6~Ki@Yc
zZm#PK-|`}H*uh2ncqHa5V54nbQK#Qvb<@PObgry6M5>d=R6@-;8y?>YdKRuv|G(#i
z!EP3?3G|Vy!K^NZ!=zV;$x{?aT9<dJE(cMACF`}y1#cbghZ++VHP@~%FP8b~Ce_Y^
z@YFMxXGxTkXAjJTEUlfrhr-jkQod*KIiiK)OI%qN7ZVaD8iM;($fez)PJC-bn><`6
zGP7Dqu5+~V#|K{vUOtXa@TderQ(V<x0zu!yLwFb;Z{ooBax-{}yIW8&)akd%)I=vI
zU)EAIx?8l$DgM2;1$1swsfD&ug1O+O=M>*7X*qe0AYDk1m3r_iU91>gzhtjqOjY=H
z`@6GhwbTHRe*`g|D&a@a5(I91ywX6imzPYUEq{ALTFtv(wNXavQ+%$f?w%@=AapWf
zLL3O<R3fBHX>X0s>zta3j2l_BkW*Yz5BAo<a?$AsX`S<o9hG&fxPZh46uW}2f`N5+
zxjJjS9BU-!9JIFVr21E(_d84o{i*(IId6zsRY0RM5^-|IcdEaG$K}t2{!oJ+C9C<n
z8!a_PX0yK<wpNCI+ApoVWV;(y-!wO<_ZLSD`i%@|EAx!YuPaXps2torIRy~Lm0~1}
z)WG89N#KYb{6*wfm%hTs=4?rVsR6%8>nDV8I<0&7!DiOM=&kS8)j5aUWxY4jANBM!
zvwMXbvutHA9(@1y(+Xh*<Q>uH^sx9R!=kLyC(KVZCBoj!-6-N!+U?94CE80l%T(r0
zI=24Gl|Km#VGf%p;Phj{A$DheEykH|%N^@Vn;Yh1%;YQ{s#^=nb071HAY?hI<d3P(
zX?JKtezg@}WNf)0)BAAB4~sU$F}0q?O4m=h*;b6&R7Gj)J=5TgZ5A1;^<7>*+yq{|
zS%>!c*D5RN?}9Tz&KwWV$QK8S?M|WV=U0D@ibNp`d&FdKr?I<o3ev1K!atH*Cw4|4
zf-tC!!^m;5O0I~;8>3QL8C9r6p8J_9s84H&Y@E?fG&xfrsWU>B-Z%1u3~~qonGZ1S
z8{0a}y;5ZYba?SeBn^Lmj0>b~y(fcLrFk}GWP^+@asEsQXoGYJ+MwEZ1qrg<a$qK4
zcn}$#2nVWqiX{JUGNqo=Z;Vx5ntz^pbKl0WW4pG6S1IT>F(7{|y*~#z$RLLpcvvUi
zc&zC%l*4CHKibSSx=^1TctE5-VLmqe5YIr`D4B4C3~gEx!-<_y0PjLl@Vs{8*@X(D
zibAKRtKRVqmg_q8`Tj5VZRB5qlv1z}!k%MKJbI!36zW&87lT;A)puDP0(ATCSeaYo
zFYDYnw!MwWA`hXarOW1gx0CY62eVLxjPbFehKq$Am`^#cphm4LGqn)x9oBDSD98NO
zm!BVQ+Y>iagI^}O%DQ(Gfr4wvatX^7!TSBvF8SSEi>Oa7qQ!XwHGxwPIwiSeJ&AK5
z`v_))?*R)IT+nzO)W3@ad=<mW0VHs=(41)Z2uIqgPvscU-u7%km%JqH4qL7>A~f4a
z?(tx7CDC31K~e=Rc8tvjzYp*d$PNMM8gj+Q#e6X%8h$H(VMe#;_dx2(=x06mmQa<p
z#O3e+Y_-$8I?QS-5V@Yo)Yh+Z%6SqfP~T+o3dUwF;q2S|qH9I-0c3X&|0*fQYQI+p
zl)cN&8Mc7yAC=?DQL7j^-b<^YI_vSTR^nA+R{gy#XHH)%QYol-S>?a4nX2Ny-{8Qm
z1IdbyRVIQ_?THwz3N;vBpVz0E+)K`i<|Xlz?2uGtXASOKxNbTsWoA|+Cvy11V`sPl
ziKT8SwL&*hLU}*`&Zr(<zz37m2;C7cZQdIqCmxt}9-w}t5B1u+4eLdBwPc_bem?h9
z83E;L)-zhXN1+7(K&esEh3mASV{w2BYF|54R#77#L0Tjbs;m@$_zE97{JG9S>vgEf
z+?xI04xy=YNdog2;qfBZ6$dkhzm+wO>{X1mub!~Z-a_z+Qt1stKl)~`d8{FVE`S8z
zg+m+M=^$&DRN>IOD_%ayU!d7UIbaZZ0{b}Bd~t!);pFmM<wP)7jZ8vC(oMa5UW3Ku
zE#4k3F_l0h`{^8f>eLRFW-{aq6|u{6l$Cj8snk0lvMgTlj!;_rV&KXlR&EU-MAlBH
z68ICHm{gDaHaa<K!RZ=jl0A1)xdUXCw<7gs17n;1K6Igcrjzu2jqBDoCQwrI7;cgg
zK{;Rooni{2(F*0gbkTQRm*u#Z-SsGglksNT0BoZ8!;~S+8W=BUn-WSa+;vTkLtcug
z;9t59!(PEw8Ya<8XX1cE^5K~4=H~R+cQZwQioo1JQSE1Roy-HkwqVTLcop#;JILTr
zb9j{A$TN0-SI&E1^uq-2m`So`88dQt{r{!~{|`O*f8%qPdRlC;x(a~i0lrmBr!#!%
zY-)DGOzO#y@7pXU#HsLB@vsr3^lBpu@E@^aijg|@G2ap(XI?esPi3*xB+`Q!e5Cb2
zG^{UdWnefgCprU^-#Qh}wiJBLthvFQ{ZT@yyvQ1%fAvewJ1mPl&W#j?v>JgM=FLrQ
zFbxfI^tcGTdJ*ofMzR=wjf9iRp7<c=g3;HGqe8mjsP#r%{<0MczZ^-mjk-m4)7Ot;
z)<@2e{lqvG^b}QYu*Q~NaniLs<~w!P?(1tVzGFnf9)6V%2;B3ejN6;4CSs~$f(6wJ
z)QW(9%gvUpox*_L&`!_;%OXgO%=cda9W|V)b7={8emPEj=^uR_*7mTEg;TX8k3Aw;
z>m&7$qYy@ax)&WVOfb&U*EQHy6jHQD#aJhQ{`!}qlbD6?2)ax9;+ZjU^4zG-Wq=AT
zJ0AV~Ia<19Yy7SbWuR5KuGY0ocd&j(Rf=f)9|#A~R7G?IE7gKb5?Q&=vFs81qV82f
zM^{_~$(1f1ubGAypk|m2ef$}}p)po9vn3&<q`azFF0$;vF)c&EbXgUW(mlu-)mWYP
z@$zTzK8m0^rCbS_4WVPA!vpQyh<WDan`KeqvTpSWsePS-f;Fl_bc-sne`nm{E!niH
zoaQo&bzCH%eH(nP?oT6<1B0%Tc-|yFxDtogJj}w0RxdDD6{~FNR8^H<Zzs{_QCAn|
z&AFly8{55>S#+uh%QBD0fZKV|!eFz_Pe={w#^aMhma0ao)T|N-3CQRWo7zBwX5taw
zoaz|1GskP%ZH2seHM%Crj3edQiF1SB%C_hOZ7Mp479Ux#adNhuSv}4(XE3dilK1Lv
zsAqB5K3z_L@hCkY+gm~4fjH$_FPs}XrTg1_OsT8cG`;o%VGp2-cpi}Jsl8(kwEwZv
zNc^)prFh2-Pt*B4-0VHR$g#2KDH9Rd*Lw<hbLOsnq#DG8Y+^BqZk?F}$WIFnsiuxh
zV!p~Fyc)j;+T#|5>`4BW=7xn30V|V*7)H=*>GEODjdE>hVJSzRM8y|DzHeI)aixtv
z7cRF=P-O}H7mIp)0VYMc6lSzw@Utq?$RI97Q|R_hVqyr<-*r!Iz!KEM36uv5oB!D<
zn(Mapce|Ox!Y7KXBEQu#a8>E*S}LGfa=AYK*M1=zkgA>_E%8Ta+S&D^KFo=Ihkz}S
zVNW6CIs`&wS#-()i#E>1NFQc-6wdZr%jE2#B)Xn4X{we;{tE3M84snVKku^?%S8ty
zoERRaDJ6_lmT5~N*^)cuZqqL&V2l1(CY>t48OF~!@o5I*a5aCMVO@J?Ow`!_#oK#D
zHPyb~qM?a&1?g3Zg3_e7&_qf^Ku`!BlqS82^e)m{5Ky|Jbcpm$=)Fk`1nD5Xgc=~=
zUVPtgkN+O~=A4W3JC2bHFL<-oO7gz*na_OYj0$G_)lbJ^&ZfuE?Pkh*J@I&wv$XT(
z&!Jv3+`2xV`hg{jAPDpr1pHek@=@6abRY(6-QGiA-1C1Ao1yKDK3m4u9@6Q3dOWNc
z_PX2XaO4FMh0loz#&c4DG9}=<k=_@QkNQ76Gd>+2a=#40rXEvG_b+r+>>qY8C08tw
z^^2$4y|lhVS8U{9t-?DX-`a;4?Eyg{wMRfwz{A1ASmj*sVTYtQCfhEQGsh(lzQe6!
zzPPpW{9L%OK$Z9p17Il*FBd5nF5p0Z9*sY7uj+mq!gw{DWo4ORXR_%JaKaQ`oETsV
zO`cbKpm`@3dty~Ge*L?`a<#hmUW-$R6e3t8fTkTWZ;%%A2G5>uB;|HKxwu78;$f!X
zOo>mY6*hU(P=LxdMR%v%)4>uh)*@l{Ata<5;DuObuloJP)e4d)50#QZE6O|41(w5v
zF=2$U)1up^1Xtv?{*?Q1QI)g)=QBG}IQ;R;6#k#t&^uzPjB-s}46MWu45o89!o9yJ
z!n$^1ZzUBbBwf`GAJBkqnJWADUp<-xK${5;Do~*LaI|=GcX;TY<^#n$D+vvfgk538
zu~M$Ccmuxy6`KTfMNdFP@*eq<y4nOa@_m<|%Nr)G`*sVtUl-*FJ=>uucjka9Z0yu@
z&9s7hjXXTtOhcS~E-{)iB-3s=E$g<OxG)rQ-ux9jTIcb;&Q(967rZu(`o&)15e<=-
zrT5jan&3*RbdcLgzG0+l_Nh~vnv5Xf^T>`QBXO*Tyks5Ha-OBQw9I%tjseV*7G@M@
zWWUsuwn*GsqS%I{#OZ&tGl)UX_$m_22-awCIOdAWD@gnXhWCO-n?(W4F+SQ@zIh(f
zv#tN7)CjJr?S&|oGw1VpH35-cB9nP`<x}MIa7k=S*hz4YQV-v&^%13T^xF6e!GOhO
zN~0^-EXA9d8+=2YnjCmvgcTMDy4Tu2Z9~#{y1~XBfQM)VQ+|7=Ug03=Wn%DL6Oj*)
z4T50={`5UVXx&H;D;3%TP!--q{W7gUwF5Q)AFpoJmrayrXP)e?{8s&G@jZb11py09
z+gkRP$q!%xe!y3Kza6M!QoDj^+h){!ZoNuMan{u(7m^XmW-obn+=_cYfjE-;(9Ah5
z5>>z_@GVqVGG{UIUF7$Wsl_<XgmJSG;{N&l{$|KuP*w&R0LLu=n<w`pM2kQe%9H{}
ziq^<+rybns9GiRon%Kg=C#nj%5_pSc!5so?nV&&-UtUKd?B4}UUz=n1VKI=W$Z}X@
z6<(1M^A_2nxr4K+kP=x{ttGj#xB3H;6$`UyKt*hA6xN9(>0gcu^X?8SDWyjHR4TH>
z3Vnqu#S?5lGnc6_%>b+w;&7c#$bVN(+OLU*C;%WQY>8E&$}qRNyq@l{k+9X%(C<Od
zK0=NXT4dxo8CqJ10sAQOhGpd0*+{_tBTc~S7G=F}btN}_zT7futNRy(PLxAO7v}Zp
zdk~XkHuxJkb~j7Iy4Yvxk(WdQ$o>1*w~;v3Xbjv_`(2ZHKv!s#wq8mUX+$>=MUq5s
zwTpb0kXZ}FrN-fe_G15n*4CZ?Fgox0ksXV-B5n*X$c3TbhkCiD&<oPTVbQM@rRg5Y
z$PWq6A!X!h+~dwkfi)KZ=pEyC-+gH(;EM6a+{K0V+780)guZt6qMi72e}#u>J!shI
zI_Vv@x_7b*T#Ewm7r2`N7{m`>rDdvH0qu9+N(D|hDjm=MD$86nq=xW-`)<{kHZTQy
z$5TRkdx$5h*4Wu)*=K={?V}AT_i<fR^(oYgDxI>IQD@BVc3-rlWb4LgGvU=NpnhHF
z2>@_C)s^V!+BT-xmg%$=NomW}G@jy7ny7pTjTxY}x$WwcQJBM(7@eM;3OdWVV)W<^
zI6V9su<$Ai=h$3{HD7$c4+}gBpn|&ce=iR&8Yrq178XvYdR|7bVXS@!M0!Ytrdpgm
z?FhIrvvICIe#-=H%s24H_e0{U<s;Qz7BRPw(I#17kBqxsI_nVgpQW4fmI7M+6g%Wd
zA)<>V?AkSUJLDX&TI~plnW{QQo=pkaq+R#Jy`7|zO#@hPrKiZubjh2BMjnZX>M!`E
ztAw#Y;|wU9dlOHLFu%;0_7ur7RqNsx0_N*4XxgO2ah!9szmA-whp&b!B2j}gHU1rB
z>F<Yl4qrnQi*tA$l=v4@D$rBjl*{ms@f98CuW}0WXB-zLd5%gkaA+J;l_#3b(bi=9
zK0*wQ6a<dYH;r~D@mQS#oZ8qCY-D%mYvY)S)2tD^;9hS9moqV;mwAn(=yuhi%1#l&
zHq|O?QH@XP+{ezYrVjotqWUa`J*fih7SvR`G5oWIIMdL-KU2adJb7tLDU!NHx_Gw}
z28kCVUwGexsHil&{bT)v$-6GW!7LWEV)j#qEZv{>EX!_5-4XqVCtNf}S#rtcp44wy
zouD4hEQ52-lR}JgC4_H*FitD;4zfZV_Jniwto%dxM~Uv!nb`EBY^oA&#DJO{%~mzc
z`8Rs=6f;Z~!Rg`*7-<p2dtaw`SFJ_3hrikP`*AJ%XzKi0pY+BdZRJdPIr=@<Mxs2s
z0qleSufHH-_qcF$Y*N8n`ec`3yW2HzvTr><T=gECQAtX4$uR*lI~d)jJW~ZP>tfSx
z7lH`R0GU_@S4xuBbEb2elO7CRnO!N?O{03;(X8A~pi1OLo&0FB<!={MP*CO<3BYie
z?D=J~uQ~r$$JFGnIjvSkUZb7DJ)2E^1@a^~CBHn=&*!F5z6w8V#ZV_wokuX;U2x1m
z=g;<W7N8Huaf$%Q#s5a#@_(<nZ;EDL$SlFIcSyaj$lerttVxkynan)8-k>u(k(8vj
z#;-v5o&}YTwL1YOn46KWWf8@16Q-1edLLH7^}%iFi#j=4y&zZ5A+L?+mz7l+N)WFc
z8yhbn8_Y(L>rYMxQ_&wu`ms{!_hP-xWAkM117=9CMjQ3Nn*akS1N2%(xhQ=qMhTAd
z>pJEpm<a?Vzl31fi-A_BOmbEBrj#%BkNm1s=E-*qx>mZ{h0{X^H#y$Xwqb=&%rMFs
zOC|iLdJ{)FW!1INl?z^Y?_IP9h(I|94gy$CuMJaU0CY&>{IRqpo&X`Uzg2%DC-14j
ztl9N~Gc~(Ok7CkS)+D;5*J9OL*ek>%DgzkT(YaMcZ`=C1Hl=gv;A-G#tq>x->wBb0
zJ)eXJ%DQ1O`}78w`IvwjoLs@iFR%S#w|3v5G*S?rcH}gn`;cJl*>XD8<yy6_i0O(d
z+DfgCJK>Ysly0kb0>h=1Ta%oB<%8md+{UV0fUoONMS&W|%8<S(Yy18CGBQua_=LR|
z1M;lBU79ZgrSM`o7n<t|ZLSATxF<bMZ!;eA_}DnFghfJij8!4G6y&~c5`p}qfI(Zl
zL(^_5pi(f#AmVm$N{g48&sF!d<)nB}3FMC9Widar%5T2#>Xs}X4mGU^3vZNMZsIXO
zRT=Yar}}<d)b3qLSaH87^tcx&qbUQ(`(*QgeugD>;zDNYN(dX|w?skwYIKx&RT<8M
zf{VpW&^{K+v3RYp#%haf1GLi}vUe~iX{B4}T=?W9uchkt9F%pQ0ko?Ga@BD4euv1v
z&K2g)rb=$+@gI`CixG$uRd<q~c$Z)kaJSZ?$&1tT_vI?HRuEelXQ^P9MqDe^^o8+0
zEN&a8gxK`9NSFp^KG#*}Oz}{T`@<U*NI=u&)#(Qe&fmEnZu*MhI<dj9)m`UwjT}3G
zr#I<^jCUHnI1`wZdHX|^3GqE8aXeqKRQogr`vNg3rl?~XDb0$GUP3Wtl{q8Y`%#4p
zcw?Y7nX=&_fXH#5efa_~8_jJ*)_!U&Ff!;Mi>KyZu~Y&%gFyb@51^%}6P3whiDo_p
zvlJBjo2KedMo)}BCBC84R@kLdL;=N0JXP&mfZ98w2SooB^pT<hbhb5Uv638_n2Gk^
zDklZ^V-*^AIYI&Cn}AB#7;T7~-dpeoV1B1~>Krm(Po2|{`R(g+YiGq<`k%{71uM-!
z-l=D9OIB6+n<g?Ciq7NXfw9Zxx2M9Hw0ecnlz|}PMU#>XCOf7;Y+OK|U#YpdOI~C7
zlffiw75S23XGWwH2)PtD?F3{v-5qy)7h|pj@3TSc#SFl^A26(u(}^(>CbX_GF8}c6
z+Q#f_!mut&6l!{?DrUyD%<<_?ibUzsN%!|wyoGL`Kp^SVGoEgylY$~oq+Zjw!0v-O
z_@?^MkT*dfaW+u-EP#l7the=xZ7G&#t)Kh!W_GCymKxC6BR-dRe63q|-j>+P^d<#e
z(ETGk<o<Vlwa|7_t-!|Py~Zl_68(ptEP;K>Dr>sy*;G*g1z@SkP%>U*SF}!T7$*A?
z%G<a7W#Q-TTa;3Pn&T)0?haPaX8}QiMW_R)-OJG(t5lIDGY}8d4Y%|j@?DJ0u#lcI
zaT6?E#7)2Kdg4Vl1KU4}zj2ohnyU;{ANk=6%hlKG__AQ{yD$6&5n^jpQKAyyMhtCT
zWm0kL+l4rbaMEN_HUHy?9)ci~Rp1L}Dz{y<ESl6Ig$y&{=r)IpJ>-N?6RHr;aBZ3F
zT*&zHH({;6*=lW~CQo7KAMy|Qs+aHBrUeHM;DL9+`!doN=M(=iw7wpY7ss%-6Lk6C
zoM@DFR4-@>e>7P0zKfiJS~@y^@ZBdsy%(_G*DU@Ql#K|J+rt4M%c?1{J%BK8teMD;
zjksarjre)`yd8WA{D6D)#=Sz}VAv^7VxSiUR(tK6bzX831?a+3&^iOjHFhjo#lQCZ
z5F&4<WFmB5C-e*fCYQcI8;HDQIpCZW>jYqHV5DVPbccz()n>byz}BCzcH)iH$FE%I
z%*9)u;by;nYGpqAbDyyVk_Z$d9CmfXP=Nkz9`x(JxNm#@5wV{nSGl`dFP2hn&@!6Q
zc~9p_LFRXgZpQV;#6Wqzlm*<11aK>P9|1sqFW^epWYA{cV3R%2gN?{fm!RPJUyzMB
z9u>U8k*We^1!FGD#O|2Pn&@As+j`V%wl+B~tJlHNR)SB48Qbl>X!WeC6*w{3R9VQy
zBwyg-6-+4r==$`ZXOz6WP%?%ih-)b*Y`^@$e3@@VE31{};*i-O<kG;n7dwHWLhl9u
z6KNOQs4<xA{mFpEShM=8(Ygz@-+n<qW)D9@Mw?B%m*0{dEd@&MVJ$V80j=Q6_5kc{
zIUu+a@pht_%*Yr=G}cH$SwOleCgfMZ&#3w7kna5>&z|Q!%Fddgg(F1R)r|XJkeu*e
z(0sdTwp@jOo=*{uVSjZl&+Yz1ul_fo9@kE7YbC<dwtV(bknaxgQ4}o$5d42Wiet^T
zVG=;SMX|uj*<Rvyl2@TJ|Jdk;$NPkzGD0g<W;{jz@pHMJyf_#JAnXqyNrPNpe*;#1
zQZor-d-rC|Ii{6ak2*V{n=@pL^*~0W+3T`W)NqTPVEv@pL%Wq|b;IyOOm)PQ(pBPd
zf{?fuGJxJ`?Bo@u`k1Q7)A@ndJ+EGZja3+``OdGg_Bn=8k+*+r5;ZF$CoEqFFW%x;
z-ARZ2dF6p}=_yxnld6yJ>*vJ3xvnti561~1>nCJcJ>UEVz4(&XSU2YICslWI_g4Y<
z5}Hk!e7y%W2Ao7<wE{3w?aERz$0a4ZsRui1C2f*)>R<Y~!M9hVU4El}0tY#Rp?{yh
zkX?B4n%@4AGabH4tbRt4TQBJwQ2P5Gz--bOySfi(3%jN0KRQ&hKAYspprf%T`m+ky
ztWnH-5~R2&1H>f=SKwF^EWb>Gm^m)e*Eff-mbv_C6dKqq%6wK3jqfs*a7?>p$}UT`
zX(CrW!mcoGzgRzZPi!FNM-m%^>a$Q0>$Fyv;IN70DIlC#=931Vk-4d1jEdgKas@wR
zEf|_9|7Fegb08PncUZAM5UYKx8IDR-{`%Ec?1iy@>xs?X#^3q9@Yaw=WkVW*(LR`v
z-Kqo49YnG1V?dIEwudElzdq5eJGdQLkmzOHL77KDAS2`gl<op*m!ELD=b^O|?lF`+
z)XPJ5HkMvbPSGU{gqNy_@}h&kASiH}bG@t{7uk98hpgX*FyUD#RF|zg<Op)R6){fw
z{|}%HKr=67=3&^4$Eq@jwJL&l$zEIYiiakV5_^yr7j{d!-sEJJJU&E_0d>RuY7ASJ
zY)$ccZ=Rx1uV5uy|NdPsynhg-MW%2%i`n>PRTXm=W;Z5RbRbMK?zvi0&2Z0s8%u5M
zaT_;|rvpNtKW8J$J1X#Q=8J>lQk^=yW9oHcu<Qrwo4X`Hx|@@dukpp(`Pp6!v-;TF
zlA<|fI=q|GU+29SftZY{4tpp73)^F!?5SaPgFNz#pia&;W!!M#kdJ1?*J~V!;GLT*
z`9d$u=jSvr%4CeCU>|8a?l}hNnK{445^(`q$S>@NS4|tM^4@c7+G0d8KR6g*P!_Bd
zqjP!+`~1SeFNnti*!=Og*dQ4A))m|M5jOR$u5`FlRQ8iVB5o~1j@ay=Xm`owG^`zf
z7{dU>7*n)$UY!Ua#y`?c$=a$1(p{}d{%^-Ys*I8wXR+)j4dVz$e?Cof7kA#|_sPG!
zZdLdu?8zGv<N|qBFN4;3!mvCACN^2LxQ7q2-0n9F7R=13zF79vepa~dr$9JDyuNd7
za!I@Iv}2e*XN<F%bl`Wigrj>oV$l!ec%O|ZZ{P8$ptd7hp#0&Ji!<5p{aKehzawtr
z@Hpn@{fN(w%B-y|5;6dn0a%%x^DG3Q;a9tR%Nlh=gB|epw|66F88A$rwZaK%20E@p
zRZd)x4Q1Hg#fgeqIBngh*2juLU`yyDKV^Bj7XXXS^nZW=s!z@LuLfoBq988CwpZ*E
zm54H1ajN&)=TK_ZIRM8UD7WN;7uF=sEsjj_{PFGB2r*wU#Nkykvbw=^v2$c}9+XD6
zI;SU`LF>Q>L4Ft2uXq0#RulKKiC=#9g%)_$=pcWI`mt=BdQiEr&wvx|GY>7@)SGgK
zg4PmZGk9t*vt)^FkyH?Muad3vr07G@y6Rx&2k(spiM~WV%35R~7)SM)(zr91yj-FR
zFLm0rtWDb)<M3kUvaQo$m(RxA=9xsAvp(Auh@d}z=Rr-z93!)u%sS5_B}HWK9XL4L
z1hVz4017F-;RsRY7#h7_^BSpE$uwTzggydSMWSms81DsQ6AoI>s_1T&dAZo!l)oi)
zl4I%OE?QO@$^5`dJ~V0i%a>Pc+#rzm75iP!Wv#{u-`B-+o8m&hXQA#7Ho*39JT&nL
z2*etQ-&Aw*aNH5qT|b%5fB%9!)%<rQbF<>LI;gvp^&`;Z-Y(=W4d|YGSh;Wgp^DEN
z=M*_A)X%n)s{(mMFar)sX97X2m-z@<Q@!Ur6SY6~iTA|lt*&zBE3wpQjg7<LvsdHI
zH-Z<JBfCv%l_aRUofvdanZFYs;x{W`QX)l78SGA_x*A-^W(U08sztXWrMb6xC<A2y
z-`tNy`<j>djVgaZ{y>X;c{~8{O6Ix>GU5CSLf#6%Y+_3zK^HHtm=LXUN2D0`OKiwr
zP@EPpVG3KwIw$gsyTSoPTX)e?JvYIQCe{{72%OA39Ra6HR97-8I8oE!fOXm>xeE;O
zUl1E262OsFa*KsH8bCONxDA}d;d0nx_WcQe;8@Luh5rTZb+F^Sn|Xkv;lF%sFAe;G
zN0zzBncQz$tt=tb?{r(|RN8{;;70R(z#UZ`!U#O?u4F2&#inP>Q;GOWPZYb;mt8Yq
z5;7|08^%Qb=2+oQDMU!sg`fG8y0`p`km!tZfp@QGHGlX=;p8wnh!_ZGp4}wRYNrR=
zpSbsXeBI<97OV+xr;~!#?LGs0lnACf$la8(*L4!<pfn>?x)U2)`aO2pH9GG3QE4VX
zZXvE1VO>BS8IK4fzO2dm3yMaK&J+6L{{>O{l2Jyvc-Evi@faIH8QMq~xF3C4$w%MN
zAlV8^r#x>&ox~#>#$-u7lh-tB;d2XM!3a(XGX&&q3%mOpIVR9i;c_#-r(~no0;{Hc
z?25slGwo8+$5&F1Eh*-0IEh&fP1C-^p<H0pSgKRFsLKVb$&e^>5X9q2D|r5p5BWK2
zHRN*kJ7O&M6JFRqILYap%+n2UZyg3G(IT{?ELq<?XsjiwAnzoYy)r5|uTOYapax~7
z@f+xBvlcDdU)t><r{7=gX<>zP)cjoTcO+E)EK+rS!kMHqoDISQ=p_Q+yZH)Jb4lbY
z3Hb1?T|;32e<4W%+Q=68xY~2qb~N8co4lwc2QtrsHj+7f-qhD$&!qg%3&DSx?aw&t
z`+UFYNA4FNqSCLHav4-;;Z<ltRmd$87T24^39)BvIi}M7==#vvnxINMHJ^xk)eVjJ
zmPG7H00>60iMPg<%{Gqr2jkzDDI&z30^8sG%ITt?3#P=fy%_$(vo%k;{tN5TE~ODe
zULxY{SFBsa8BKrltE);g&mpNKlQ=9o$(^I<AxE%u<hk8RB!;rYZC_;(nGxVTk4^>>
zw8WQp_{*NQAOSAT&)GXjp55@5Y4WXpUSSmxA&6^N>#vc;Pa*LNtQ7vqIEkj(aUj~n
z<orZ8b#qOvD%FO$sNw04&lhk2!A1VFf%B|p%>=w?h=B0CX6=Zw6h}XR-*PV9`uR|{
z_V?8J2=Vy%bs*_%XUsPJ`<GCnaZ14BaK?_4AYNfx0Wr?X1j|w>OVynP$188oea{Dv
z5cTZ?G&s)!Fnjgr@Y(n~KIKvL$*8asSUfFREczexL{7!fbyG(9v0oeTon2N2z>x+z
z<L9J%2Azzw=OKC&YGRymawD4jQj_hHXWRQfb>{zZ++5QU|6jkT$brwE)EK9N64|kS
zHsMgbn_YbYE}JDNix-wkdPFjtV!wu9eeAa8&v2RIZs&As<8X8C?wD*Vc|UI(%I8^M
zo`i89h>M|jBYk0+SombJsm>|fTP8~LZbh6Lbdq(RBO&*bu=#kXR+cUBgl?J|+cZBM
zb=>B^UtJfwDFpB6)UynP#Jx~G2l}P2ef_blei-HsU4zNE!Wo0#Dvo@)k?Ufkg&jo?
zUugK7$NLq9?R#uD6mQINj|o)8cxZ}=)sR4<Y*~6{G!|~Gf-C}sH&V}m2fR-eXNtcE
zh~P`-jYdU%)b!PBgWo!~;~gYc7QFCYSz*`omlOc`51d!WEwggIoD`jv+_87mUhlSP
z?AZ=;MGR!)#S8ff8wKn{__>RBN))fwCheG&MSKgd=?krK<nT_X>Vc)=xSQ&JV!teU
z+2_GY47F>*D;<J$zUx??-%^z`0G&rIXSK*Eb~g($T}2cb%s@qWSUbdD|0E@U_#}H@
z<%J5NyJX8+Gpqjqx;k{(m|btOs)4?8Y*%|*<t<A?qdg-ra22)Wr(N<}3CG+(zg(Dz
zC0;f+UMe#hnc;i)LeEw_#gwtiPA%<=MRi3iVNJy6qzHpMR;1FF5fAF*fTG$vpfih9
zCT5^|M=Xb^@uOH5c1}rGar36cy+GLn?DU14t*kg^VmQ1~-C%TnC!zZ!ytLi7G(4Y_
zmN>Yuk?m|PoqR7etKxmpB9djpd#<i{8tRbO!%xj4zq*p(zkWXu&#f1dDcI#IoBku4
zDqU9iYS<tp2IB6u#7YmV2y)UPADR6~Wc-S&?-fh%9X6jog}D2d8KS17o`)JJ_mVe#
zdc0(6EV@Dckk2i|&L9CmX<^>1GA8m53cB{|1#Wp${(O|D$ZwQt1huBp&G%&Q1-5J{
zxw$U+eVA6}oG>R%xky5A_7UiPY1+w)asC~D5gWSXQ=x7LFP04#>S{V2J+M2u^x`Kb
z&_&Nr0yyzxsM$1H;hYJ#h#d}a&-c~En9u!IlU6D;YGq7_m}Yv`7byb*k;5D=?#&y|
z;7qHkf|vAmN=tHSI*On4R*~^9jw~D<0hy5-z#dAl0@T%415%|qOC`D&&oDRAlRaQm
zQujw(?(Ip3ODju&7tY@_m<E~>AY|D*Qx6R%+<gptQ-9AW(q>D`zagXJ#T)JujPpJd
zUXt1A1P}vGN#<zO=(*FH`hnoTAPvMG>OH@0?dL|{QFI~kC&9sT>^odlk+rxFvyiZf
z2e}v_RsfhXYrSi?Shw$X=bk?^t&;D)D+RykmT)Lk;D8HBElaTSjL3p~Z#QBcN;eIf
zO9~BVXN?T~;M_YwQ~CP(QFNfhPe$hRrJkpc3_v`sNc*iXW>_t277ZP)i8rzGauL3D
zhQt^hxlo@$z>1^jF9=)#^v-?`n4UlI20HNnPa9J+`ZAjTe%hk?rW&V^LSa9Pt@ph%
z7ct_cfdD3Ka`VJx<X_NAifz^jW6EF9(^h%5zo2W0l@YRIj|345At~~7-vxLkrRm3v
zQ|p^s3BMh+vg}dUM5=v&CHtKd)H}@~)2n+;@UexO#C?8KeAz<vm6DU(dXkGW*@Z><
zt%^rCyfZafd8Y09<ZPL@(x`&Z$CKUq31de(2AOH^GbxeRv0j%xp+I%@2r$I|$Vzqy
zTuKBY_lm~zu_tj$s(;W)jt7`IQ{N@fc&fsIeU@;+Iu-XD$FmXD<n~gotg7<k`9crM
zYB0U7H)PQA;PCP{NN}SQ<md7y?=(3JjpU*56JAnx(s7!aZyyu;>S|fnXa_YNkpBKb
zO)9*1w^oVahyiF8;l{5H#tu!B>JZ;?c1si{no}2TG`$7JBB<05{okq<(mz5C)^AQ7
z^MsF>Una=2n|KvmuI8au-TnA5>_!t0E)#s2A?VQlK`&pvQm_q)jk#&%Ylc(y2F;6=
z7sMzwTkE|NStJiJ7`EQs(r{6twkYt?5wQkeTHU#^Vew{Q>dg)4R?8i{S=48oFV=>u
z14)LSMw@n9`dG_l&^#BW!58~!?<%o!`#11Q&4UePRaYHPAcNmYzvY5U`=m|MD3=6r
z==}cLBoDci&!!}jv(&szZ+yl13mB<|w8@`4AYNmO_Vut0i+dlBC^%nU$aIA;DLU|t
zu0D{Dj6uHbk$9oNxUMYOgVkqDm`I2$PY<TNDSG~Lzj=4*Vi6hYxV!3A$~C(4BUwe#
z+Kn0xJ>~*hr;!5FDji)<skhFhx)bU?J?o%cp~&Hyxd#kR?j}-(EM5w$RC=+^YgR>$
zipK|Umt_MHY2*uTfjm$C?OO5l1iY2|kRnG@^6rbstTMCoj?DKLHK5P4?2Ie;3Pj4%
z?g$eAYeUK}KS2~vdSqu18$5Q6Ub(k9C7vHv(t^_8^&$Yb``Do&GI#UUwkW1#Nc@c(
z0gqCfpI_D>9APi_JL5Ao%PNHiMs}siK7zz#oK=D1%FlvTQ`vMV8Og`%<oNDlihtV*
zKKF;1T8Y=%j+0X122yJWYWG$@7Ag<LapOPxFB+BnW3T+5^&x7@!>x?|_w7>s%knmR
z&Ao_zRwi&g`Bi?yN?WkUzQMYTVI$U)isGv8bTaigbf{=a+PeIsC0wUE^h5NWHB!SM
zIH-3>hQd!@33QL55VX_DcZU?T9~Uc!&jq*^g^Lb~>qquEaVUtuZYRG4np&KhYm%j~
z7NGM$u^Y9t!=*EhQ*;omifJ?Yyt8Pw+oQ<c?|egkxY}-$i5o-!y1rIfKkV>@tokyN
z378ZRFpM2i`>x<ZF<{wZ@gnf6kxP}b!muPT0M5LK+U|E>++(sa7(yW0Wty|3jKR8P
zwAKj>LimOp{PM%D5)4BsWA9!N(u*fgekV%{e+{<qilvv!5SzOiT|F_FRS*t>bH)o2
zpswX}g$=c}*y&P5$U!efz2SH6foEO0b7FT@%=ABFey6r#NjzL^fY7h5Nw3#JbKkYS
z>|x%5tuJG<0YlMs_s*S?GW<V(7@&;!Q2Z5tkr}%!!#o}bH^PA%s<HCjY>lwifGUcK
z&h9ScEfvTplogrC@@!i*#`lvw_qGu9Jn<FMTDl?1G`g9IZmjoKs@t6)hM0A&^1N|p
z2_PVHgePjo(2C5(jqtg$JX}7HJ%+Dpr(}$_x-R%l4p@c5-e2eJ@<P>E(+!i=5xit`
z|JFd6GHb+C(4A1mbxDZNMBH@bu8&iUKbxu*TqM1pr7&9F{lI9kNm>4%<^Or?CfKhR
zQZaU2=**JWOHs_+$0P<j*-h+H4P|`6f=NRz-8B-uC@lLvt1s*F+eSs7C23=2L$C{>
zm}1OO^oKghWjhwUSb#%3^yr%DPaBSs<*}QI$n&gXTvX&tMn|qZiar97;~9%@=lQrf
zVGv3+HMObSj_^gU7^Bdnr__(hr33q@0A}zQ5L;K=pAUyv7uhkXgmVPXzo_-8j@-OA
z09Wly$FC)F)sS8eNw?U;(O?PW0wE@%moE(>5A@!=-C3%RE`+xf(eq6aw6G4FJoY>P
zcMmv|yY79tG-9lZyeyYy#n5C0=*8EK3`t}e9V{&FTx?UNs0I2~_kH$oL{(9YKk8Mj
zotp#8d-IhS9dx`J8va^wUCmIy9o5BadH&Y0he+ap(;Csrn{Iq%ki|KO$Q!k7?ctn7
z<5Z#zGxT~l&ALqfzJe!rRK-<XOg1-Ab|6iwtmadv%xa~pfR2}rDvqSUp|oW8Q&lDA
zd7<-jM+O2c?2l=5=~^?!s?5a;Qt=z=@emkN8JHJjNjyS0Tt+(X(l7CsSylA#nLhI}
zVC^=S8oWiW@;q=r7|_{oqxY$1U{_Ep;9PtC8F}>lc<*?Dy&8do4+>nh3kIxzStnR3
z{AUY5C}L&yYhc)<qF&lWGIgQ(D_N)S+B#I0S~v|q;g*#VVgd(3)XVQ;zn~&R&ShOc
zzQp}xwHt5b*(d0};;1nU`H`FA*!~C|o%vE+qC2HOWd6?!p6#qKQwoUi0vTbL;!j17
zxVib~1j7%zNcg6F$f*X_Dzd5K7A_Ny|3nx#-QI|}yZMdYOpfhQY2L*n&u=$d$Bs)L
zY+OW$cMva!m|Fe<f`(;<f$&K=Vmd#!C-rJtSx5PxRBGYql;@$tIc66y8q-4Udp*%`
zZVNNThm3^YO#8@axr(8frOq_%{qpvlEo<Js>?og*tI@dn0JPS8{9A%0rbj1LKTO|x
zmkgy98A9#+?Egr<jyEe{AKsJ@RsJ8Vs0-K$&tDMSxG-SK%Jh<k$XCbI_Np`BK&?vi
zg1rAP=$lzKRlxtQuS?aI5FszSn~BFoNcdAHO7SaQ^M_En0Nm)3AFgt*d_mkJx8GZj
zx+F(?Tw(z!l3~_DrNF)d<VJmQrnfknXj-9ixVBpCcp9!^bb$gq`q%;bsr|uLz>%=c
z{3*}}>uU1R!wTBD?aTJJ<iOx!W4(2;$0H6xyB-exmjgZRrSjce#Img>_b*Ile;eSy
zos|b|74z-n-^6(S+=bmYWL1`5Mh8Not{5ZNuELC|uFm9^O@1MKX)4FBNNr4RR8Lr`
zj9FY5H_hZfI;6Sl&`11R)@^|x;o!j8ZQtEL#D5T01GM@G8--<#J8Ed`Gv$PId1thj
zl(Z739tFw-W>ZlD42332ZryTBZf_uI>xvi4<4}ZS9&MzsE*xaGaV_bfcg?=La(3#d
zh2t9Z1-Z(aXi)n31KntXD@JLcLGm?C1CWDx0FCbRfL3oJL@<s$cm3C_NHW?NUqIIW
zp`$BwTaqho&6vpd^1{a0$Pgzny7bY8Gtv*L{YsZZ^_#dE^SA{8_62brkk+w*Ji1i@
zuy$KlqOCZ0tcCvrtm|kW8OGcTKsaReyD{aDTLahfo_ipM?NJ-Vt!V{&fa!|vp934N
z+C<IpM!2ArdF=O0z+t2Cs<-~vYh(x7+2}y2AyZE39MNv@REjw4T@xEx1x_J5DgDYz
zurBEiHy4X_<S5YWvN%+7x()ZD=I+*gPfEOGa||>=c3S=1DB?cYFV`mV@w070U&HQN
zSRA}hX6>1qI_k})`a$z*Y{mU)bEUO;;5SQfmcpPBq7S?Bb_EtQ&8ndj)JlvJJ-6Jy
zTCzjpVIBW$!cvzU#L&?D@v9lp&(vssCk}#~UbziZ)H~HCOC(DC-o9gvDd1T(oeAE2
zm)NFTBod5Ps!Lb`HX|xtd|<ce&T^Ay*UbH4&RKYWr$U;Ah`IoLPHER0@YeK05$2AK
zc!J$y8xLK>uY)}U8ZpD;Vk)btSlu}<hrC!jFQ$h%X6x(DeeD2IVz)j&LwdFObpI8#
z(esg}h>Z9;^zcDUhxdV3Hi6n@%AB$Cv;f5N2h`DsgnG*VIHlGo-Kx0}{+4%95%Q}e
zgp_fLn~PQQ9Jbo@W8lO?%LuIw7Qg-2b0sFS{O3AA^#BufHYP^M2w!nq6m{C`yt5^~
zZl}-2LH6&4K#ahNSo&t%-ZH^F@?d(!slMnD+&@Ir7RkHqxN9%)x}hx7&?nyr%JNGR
zrp$ViGv4l4=zo6W|3#$5W4cLa;LS5XJA=+);ZeVr%{5{%VW;k++E}Z{Co0XjX1^25
z^WnV&qXcLz`$6-Q%PPk_IqG*&x~Yoxsn1l)=%r~>ZGz02@b12%nI{-+RBS^Dqp>N^
zGAC{Be&PRS0T+AJ=Rv8!5(&B#L5|y39xEdotE*rVFCm3#&0ioBKI%Tu#yUeEb?q*%
z8ENy#Mb_9G8X!qToSgl}pMXfQTaa>*+Zzex%6duKqfj`pyv7ax;Kr~P7>xmTUOpAA
ze=o^tp}nh9^>TqYVfY#7Qusf69e(cDLOq)qxk`Z&2ah$UETQfOR+q$KEiFAeZ`~Zh
zFwOC8$6T`7@ANl!;*1E!Qn5}aCH0w3G4%t&n!E{yosJ2dbv)ONm>1p@0~*Axn%=9j
z<g=PK81ps(JIK$5&E#Lg%BxV71f5R_Qw1Ec3#r!(3-R_*GX#^=9d7Tw_f!2{IRCxm
z`9Y!tT0HHWE$AFT{4BRvvkj;*Z{|f;?zm}TZZ)aVALe{ycH*rJ*<DGN6R74{M{^Ub
z%|``Pv|zfUcX0~l2xb!gHpZ{XGlR@YPlCTcZfT39^csX=jV~m!h&?CUO^_Jk8Kne)
zJCD(4UcP43YHW`+*`J+10(PvTW%`y3@eUKorfX%lVEHH;mbcu{<X&O}vnx^vm*<pc
zbg1o1k7drqT6fc-*D-#{6?~DxolaJTH?4>~#0P%Aj?0rk5MSn$K1f?uiTH5Pm}#`d
z<n-o!zEWG9!p_U90TfL;2GL$bUankTS-_#pA$7*@40NX?j`0I!50R7anxrhFUW^x)
zSX$0edrdGzC#2Jig`B;tyfjoFHy+a*=iQyXp>OuZ;*kt6=H%z=Q=huF1Wb`eA3`6h
zp<>j^910tW7`KT*y2IOLZV*!zk(-ODs>>OU+B(-;>6|iNm`l>jklzI8KpCf{JRlMD
zhwYr$c-32+jpg@@C&~>zh7;PYvk`Cdym8;xSo|MU`QO0Z8iPDs7Qx_SOHqcIePXvd
zE$$Fn>pn3hpJrAwa~U3Ow#QDp$)Gy3=$gh41~v+YslTw4-7M94W&IP7?|^uITqg0C
z-o?&-?Ib_w269kHggE2DT7=OAlnX)yt-E-D;fT+*qAbg=)JB(v!O5M6$19Mqn|D2$
zcqBDZW5QlkPWVFh1ijS{V7z-eesx2_A>E2+ur-0sh(C4MFH61rmT>(WMel+#qHcZD
zC6(L;fhb#GU$44rXp}I{N-X8P+EW8yr8l&g%IH|E2;Pk6)``bxX(Oy)wtKlrPQ&mE
zI-tV%+ZRrLk6RkY%i!hM&!uAvRIQiNtxT6E6~qxcRsA8v(@75aA7b=^Jzl3JR+{Je
zR~^W8Fh@B0x-F5ET6q1iJ8Sh~D3fp{=Gurn&YBusvc!J%!B^cP+b9?ADt=N%#YVzP
zDdcjTO^ND2mRHy^U0(kAvgBS<S)24jHLp@g%2ATzrx_LagazilFT4<m6P`pI+hv$v
zC?72SMrW(~et*haJWUAK&<Zd(N!OAGZJ6L)0ONnXyrX|X+rso$5t&U_jRYsa&jqdz
zXk@0W1TJZ0K)yNv8FOPk7$@nqG=FEKo-MVmw&}AnbGd-7+jb#|Y#UMqNaW6tZewFA
z`y~<{Y=F)kaVC%sps6X~g4?aeP31gQL(p#3w`V*3oB`jQUNz9=%KJs(7bZ_zMb_^p
z)%NG)L??1`PAplU`CDOh;u$GM;KtRcEE|SB=HAHnF#%q)x=-zPKAayZr_Vg{8P(@?
zTR9;Xac@!Y`u*&bnn(9uRh6tzIdKyc7Cj_BLk(w=UVdI2X|%B`+>}~x8r#mLX?Lu?
zU{YqucdK0>h|TcJ2eMfPF=92=i^Ej&TTF>k)6-AJpg%3cSp$z{h$kxpK_Ds8ZcJ>D
z!xDmQ(}4A`m$mBX0lnw&K;62sA4Z-{f4}=A<QMR{b&kZQxaSVb?v+D%3%4FJN*gku
zr&w>OnE3gXHv3^;UPuW9<7^xdn^ODZvK{&*Un8eFlX+qW)Y!g2s2qJ@MSyHi;6o#~
zvFPmkKC}g4<=xXB!RVN%#7q@{4SIB0e8r-K!0O{*xvfBiv%@3EZ+n(OM=HtF=gGsl
zd$!op=7UBoW|0T`AWUr(2Q5YSgtDno@MSz!zA4=YWS?N8I;3>EzT`g#|3p?)+c5VS
z_ArT0C)P;rMS}YNH%LWP1-wC?-ACtLezg1x`t=yrsdIwZh5;;J!>enS4|)E86m)ui
zeJHB9$1u(Ye~V7$0#UtGGcmnT|E+GfFWqrqs0P-i4I)$S)R?S|(6e~|NR_h|Ak+ZC
zD2Z|qAafcN0EAl00EC<PsJ9kzKXg(xLy_lekC_)C+Y*vzSoG9sJHbV?yR}~dQVv};
z9%LeZ^PZcNRCvT~G*BirO{A=yreC<&<zBLfTVS>;R3=oa(fXvSE^~fibs-oQt>#Yd
zYX3(3p(Ehgx8Ner26O@cecvWw;C*v|_g!li@Z|4bq7Skj<FBIn*;QQQVjO)(`Gxaf
zW>K(P&+#BItg>MW1=_|<0aL^6mZiHS$+(%G=P@ZUR#~=eURq_&QWQDB{F5funos&F
zQUhq{{S*Cy8b?y$#LX83@NLjw`j2YTH@~F*u_9%t5U!vZAusFB?&d1-w}Jig7oHH9
zKatvh?K6`ryy@WdP^f0_9;_Z6GyDir*#e;`TIVL%62Lwap*pA7$imXHV<$5@JH$pr
zPH9x`JQ5m}CVC+Qg}_#W5Z`6*+KQK7bzhNq@$WL`juTST+?8H7t=v)E%z0zRcH<YY
z(&jK0TFHJ+VUvZW2duvYFhz?4rqnntwoj-*a~CIxx?qR2@yA5Y{awloN3uz0DAR99
zJpKoDSeZqAKwV)&@mc#o&^)jMFL51Uph2FRI3XZNoT7Rji$4!|TnjGF3bnU3Q;!|B
zzN2o}2;!jmU5_9D(LpFR4@>&ax=*HSQ{`@~-|FU#BCkRGQ7(ScOapA3JMwJX5Y4Z`
zp~lwgyc!^SC_F%Ftf4M=Xm+h3lbhqYBOFGW!$knbj7Q?i$Q1~kRlLttPQu0oxByO}
zVyK}upb`yE+>Or??YO;|Cg+J8>^+vP#5*?dOjGot6PXnG^Ilp<cY}m$#LY;qTQl?)
zPHRkLoGHrY6zx;@@g@7`P*cbK{`i0(X9e8WAqJ{Vntr>7v&ui&_i=6N-eOHCKYuRV
zC*oZi6r8NgJJfaO{^v|6I1ykNC+AO|)zRNhi=%6MO)mbS{BGo0x+Ny{JW+Y<>!^qi
ztJp|gO+~P@XMcM2_=_Aa6n-h7S@D8(JAHs+lqvXfCMYw!%#1iiOx_5|isK7Wx)z0G
z85{$yQLD^y)BKhVALcyE#j7dFi)a&IJel(Mdx%;0?Z%o#aI|4deAO{0#`=|88A8JN
zA_-oFn9L4*%I%;bKTm`iK!v>=1H3M4)Yh2*#W<8N+u2gJxKEnj3zWja?#pz9jAbW!
zgH3YFH8Hdu%rlwpJaIJ<5M_fTj@(kj_x{uL5kM_4waY|u@6d)#lZX1nNv+8e_I=U_
zXNEe^#R6*L6>uZh0-XQyi%kwwu@IWqv18KxifJI#__&eX%IZQc6RynonW7<uklxX4
zt~P+mzb`q}zs}qz-?P-P-_YwR!ihEvK58B>eij2vrBJbD*lhr9nKr=IjOqiLD!QtO
z!cGUswEMXGRL*j6N*S62eO)P>d-Jzj?<UX3eSsEDQpKemp?BDX+${m8#~Jnx)c4`q
zS&v(31Y0<{Rv`pqC9cS7-_e>*o|^`+w~1;o31rlT&*2t9GZKvFqz6=Z!=gYE3W>ui
z{Gsj@ImG@|+?;@mJbUgq(4DZhJ^z#V?M!Zl=|^Z-Mi$QwjD!wnF70tU-QwLbE#;Rb
zKVqGR3;r)98!0K$YF@ghE<IRQ4LmMd3~a{*QjW81iiNU{^61ftT7fwprsGXnBpyGg
zbqYD`cfuI9?hztrGE64OIBz|xEG(ATeig6c3`{fU%S!t)?axOv8j0>Gb|p76H#9}_
zCd?7KfBshZQGN8mWBd?QjRqD;dPRga)WCG56(}rUu=?8()4NPh^S<r;6{P#7ybet%
zB)Bo?rMaUI8O1qFYTH;Oa73adl0L^rt>ieED%5>m@Bp8dzFe5-n4G7B+`Zpe{&ekC
z&6A!^sIGHy9cv`~Up*UZk=?m$3>aFpiS`TWBZ+Y=)xNjj-ttEH`Di2b>*gt+)AubY
zhvT-d3E!&gp?cgL7NOu}F!@C!P#xsfrkcjea>UvgkKU<Po%kFd>9G8Ue#FsA;uw5d
z%pUX-O4UV0QYi$nM`6A`Qtc~iC)ifti1P+r5==BIep^3Op+gjHOj7qWB=DDNKX^UH
zA;@?q6&8M^{$_ROD8pQ%IfLFu#laQ+(<lgjCowu<xXJh7g~l*1k{PF3-m}a3_}+~-
zPXrPeTaFmPR}TuVh7Q%}y8_70w56QtxM!e@+A;BdaX&zQ%ow>7=V}Zh+DY`n^mcD<
z)1S#yy7I%W{l}`ZZx+lbANv59pXV}z&zEAE+4PnXHd0-~P{--s{fYySmG<|txoHNj
zSe`jR^phpac{Qk;$&U%WYOr2{zMHSEN#mum;JRGg)M@N&fw+XsXQ?4c<gC+7LD;VO
z4skY)#8<V|Du$kuV?8?7+M#J{4)Z*>W2I0HTJr3IzA!BPqMuq9`eU98@FC;-DGdHj
z_9$W&TjO$jDD)Mr;n})3oLkd^43TBoAC8u*DTY|vEU7hcPbsl}R?<@|x@$edx%Nnp
zGUNQ@ENfY(1I13)XmU#3%M+vR*)C0Y7Z_sh-M#}NHUpLBoV=_}WshZIaUKNQKs*Hd
z@;gZAd)}klYkwCy`{c{y6YP0k6ZdX_<5audk^rr3YG+C$_-E{uN1dG)ukU5)n%nEf
z&Smg|dE~l3g)dPRpy`KJ)m2T48s?CEq@`<q(nWF>aW>;30kj{yPfh?O)v6pjj%dI5
zN??K?jE&1(A9fMUD0Uou0KfMaB-ktbu$7B*mE|_7<UJA~eA<kOThTjLRX?!xsE!mi
z&FPA_`30n9<Q_x`2O7U(0mrqP9@AUa$;;jT_uyU@$%^@!$?thJz-E%NZB0%mLdmbN
zMPAgZTQnQ+6!~tuTbO0eW}yHn(D`4G7REH}-Apf^D60>&vObP^G4nuT))kLY+g0eM
zT#VtUXodjL#1Zj()^{Z_gtF70e33Vj#6^3Sv(yiuf32c^7hZVJDZBy_t^YYi5}cU}
z2;m;3>R`VVwUkQ)9Yy>@74wdSf;~8d%{PHSy&cm3mt?MN{$-fsYki2Zttm}ybpmVx
zHb8h~4=`2bNI(*9wa28mFdUKWYTCbQQy-ir9VSELa>j6RgFrSg5sj4pS0hx&bW-*@
z+lF2$6XBXXDb`~GHQFaWmJLRrhUa*NaVA(KS17f%FY{uqzPo%(XXlXWZEDF)2J>pv
zE)htb60vL=4YR*uYpB~t1c}EBj5stk=}1URp5aNZczaw_{l@Yv5{E#*caz(dgO|>Q
z7)<84GtJuLTWU!C#z}XJOTJcCN0PnG`8Wr@-F?TFR!hDGL<u5@h9%?Fuse%LAi(^B
z@@h=0vGFPFGy0oD^y3~zD@KV{+*^W4TF}nZgzx#<Phn3U>wHuwOnp#bQ?fzloteR%
zJhXd1mFrlSkv0i{r3^MEMEiq7b*0i#Z^M;hImr6fzHx&sGJTE+PqHa_y5;a7R7;;{
z7st$hAynIj^SHwa_888nR;ZeMLow!n%jEtXfH70~hQ9EI@*x(9`xB!B6)sAGPW@s@
z2t-f9+DxOQxH>wK>JR1)M*j@kz?qYblI$0Qv=vxjmP_hXnwSTnUf=ELwp?Dn&ooM?
z`E&D#BT!)*kgWEjd5W-E3;h7O2UKZK8l6O6JB&QPL0BgKXlkchYbsc>R6CIL;^oOB
zplUPxmz2Iu<O?&k_3t!2P;>teDb0~hm3=)+$CH5#T{JXh#)`k(KRy~&All?UG+#|f
zniyYhW>jI_`Ud<&fr;LaDt|!*fW_=Xxy$;d!$qfHK7V#kS(HLWUGlo}Ah@EZ!MB%r
z?^fduTL>3hUI{>C;~8ajvJ;|l^84;{Azp6x^FM{fzB8UPis~yfifJFWupWm2{Xkzp
zJ3t=u7xc2`N;}}f-8-{pKR_uUc*OI8pSGbPAA~t++@iaegMWuJSbiqujj}Sazpx3y
zg#F_6Ft6#&3C@IH_s3sDY4rj(X{QJ_g|Zh;K>*Rm30!c%e##~74%rnBkWusk#uVMH
zCCqwQxm+w8h4K=Q9UVO#Dw|l=g<mbHhdgi4Uq~vNjc37x=ZN6MvNm;XA>bn$V097?
zUIzxSo|JSH79~z`0~7$BJewwP#c{wDBd)ZOz!ir@U)^384<P1kWEU%Wke0W>%YS#A
z>(P5v@zoW=e2%lYE)}g|+AYxq05%+S?&r4NOZfF*V)6Q#w^pA|$>u!DW6OR)oD6J8
zf%{pPw6p*Fl8>FjL`no&nkmZ1l+dlYHb29!etF!>Cy>#P1@+36_*(^JCLM)!`pnb?
zSjL65seSpZCdIkhJ68zbAs#0ur<`K@gm+8@ApAQ3W1gi5c+q|^#p$@Qv*HMTcpkQC
zDHhv3{6_Zh!>@U@%5FyRLrTU^x7>mbDOI;du6$ffReYVA>QVsrIh$rL)4``gA6-3`
z_=UG3L9y)9hi!3>1aN#Mv<EE?s&D#zsE$GgePEq&AI_=5Wm(tH)#qw1!|bx}sW|vK
z>m<;Sv4MZTVhOC3WLaMKCnXHNDZ%-b_9$$J#=}jZRW6jxec*tk7_RT3s`+^d*g>od
zn9~x62mIdR^Jq?=wz<wKGGNPchMOjQYrloLtz1rAo~<0Qu>bIL&)!Hr=Z|>(WBLYN
zwJv7n3B$?<CDJ;XArPWJ7vPCEoZ87EO0SqbwU@40aN`G~P;f$&a9JqH8%;1=Qv`1e
z@SFhYsK0#mbJU@EtZ?37YG#+@&+mb<2g@Ut;fxCt^=E1)OQ~9NA2_Jx;Qg+DWV&t4
z-hh@-gok_b?5YOX?sj?)K|{1>oO0%pE(wd){L_JLPY`Gte6G9W59=6Jy4Oont{UcA
zo_@;@&<V<rpw($KJ*+(JJbSGq!8<I8Tk`bRb9VptRO^51|AQ#e9d`N1hwih6nqzO0
zs=H289^Q0Afyx*~Ri0xIq3(9;#g#~!w;na{wf<oA)kCl^z6PycVj#&m>6zTgM6x1A
zT=hfGAMmeb_a_=7&&TlPkELM;Y)OlA%*m<{l^^|Njjs2{pO=OaX<&Ld0W5^XAQs&7
zItHzVb=0lYADuAx4|Dou4eQ^chQ(}E1`iHtxa}APd<d{ilcDdZf1j6+{igl|J*p*G
z9oyX_By@n3`#i-AM9>5Q7-sa<zjF8poH>@r5<}lnwv~MTyIy@+nX~rB%nk|v2>vbc
z($O0VD<R#_Sr$DBPRuLzy8$;2rxzG-DqQb*&Fz^VDG6U*eSBg#;D`rWW?3%rqfr8R
z1^|1+^m#XsXs*#@SE}}XPMzt;XjKiGTDy=|k7BEGhe1Z6W!)`3D-2m78W)N)%-iyg
z3tt`Yj1Km67#-eT9@Hik7XLgAie4t*Vp-<6qyQGR?i`-p?C-wfOX;9PAEoL{wB7ri
zV{rM^!s10_`<gNGQJ{*9aJS^w{h1THa=<iX+a1_nZ>trgJvcNA;*%YMj3`f9%}o<z
zeKO}_S-9mN{4Ef|xzxSAXT3bkVB{H$GEG#=+<@)e7R^J`&c5QdltwxAfx2acU*H5T
z)9rDz%Vg@_haEz;ac{{KNDchbNu+L={}GdCT#i*;i)Q!xpP4IJZ@@xScnVw8dH%63
z&ok1P*Hkjo&ZFW&{!kA-Q1XW#Z!097<uDqrU_a!<w+!pI^aHh9)$4B8b31uLsb+fe
zUZ2kAn4Q`kFlC^{@iN=866p07o522Lgx86>b%%j$mPe&rqKD5N&B`Nb{h&_Guby#=
z7~-tUkm8N{tc04u3))fj)!m`l8EBWYOJ=R;*&@c(h6{-5Ax>3E_=tEPSO(tF*MH>)
z0#Us-uGvoEO~Ab!OJYs3sIG;979Jn;wjMKr0C^ErrMnZOr@3;Ios4;hH(w|^aFA%?
zNK1k^{#gaVpx#KM%2#a2@Kx*NyIIYinLn428xIantsQs1FH1-a8!_Rwu|;4wP6K29
zY<KUn=$%kE7rrjYAMoUfx?u0W=WF^@#Zsm0vs6gn6j~WTUE?_O3qG`6Cii%iF+h3Q
zoSY_#03gbT9g-u?cX$?+Pz#q>U~!wQpASY7@X0dSd8AgRj{h(2-ZQGnuIm~NML}uO
zdr?4?BGNkn6cAB^iqeS+N|Poap@t&8gMuOu5EKz1(mSC@Q<~CyAV^0Nqy$Lf+1$_l
zj_*C=eB+Gq{dmuh^CN>Y$iB$F_O<t1bFMX4X354NIY(ZG6p?W11c>n$uA|j>T6fWQ
z;G}gX7;cm}d_qKml<=E`8`a7g@^4Znt%SAmWqSBY9o_w&ojNUQoiJBC44Z&zm$`pr
z>S}w_%%nDI;~T{=IpwQ`MdVChDj0F|XA@9NrCOk_@B{HoI63kWVDc2kFk>OXlbCH|
zMa&T+SoH54a#><O9>ahQmWg7VxBB}k)QdnXO_MdbP^Ji6cCkk;q)biIFFO?HOS2WU
zm9*xfDzx^SA7Nnss!etn<m8iYy(_!_d!TKWvnIm^6FE8iZGz3^ZH$g}1Wl@cDpGR{
zwTh}#X@9$3C*u0Z$_8B~24Lr%TmE5GAipPogKBlMZ+q2WYnK#jA@d;hS$^$3nWUeM
z?t6wcY$TP&kwV|}qdUVNGMAkAy{@&>h*6L+O0&+*%6Nma{5f-E+r#<IjQd**tVLIn
z1+NXv?w8W+X*Wl@N9=Q5fW3W;l)MI=ZD#~JuUP`gl*o`z1eD0=;ZElJS(MaM<8GYJ
zg17C-HdyyKPW55Sf&cP^WA15nu2)WZE2EaS*}@qPPHgRL%c>@F2cw2xVJDq(wD3)x
zQJ)XuXyePp%FYN~<j%jO%e$Buz-}YMLA4}&BKXXBpll?0Vfe{#6~j!i*rxEyJdV^)
zMQMqRskefPhgQH%hLLL(rx06#GDR28Eg_ae$*1%5pu$;W-+gw4H(Nipi|<Exk1QeV
zYO~3&zEOb!559wE{|AWf`0S9>c(YJllOFKKE|UIbW9qfmdmN0ZMp%f(@T)~+m368i
z9%dn@an{LocvH1o{-fxsj$j5!+MO0Th?~cd_;kTEnd$j3hz~Irj6KJ}&X99wc+I}C
zy3SyNJ!=G(RlAf9oZyDjHlS>Q@u3TF15{6XwLdWZXMFQ802GX&@MVGQ$L;BkYh<11
zK<&S2vlP+HfUwnmR4ik5+e(8eY+FXx9)5}K)Aesg8Wxvt`SH?jtBU@KY)&S~Om80X
z5FIOjW*^2Rly|*khec?&J7WV_u0ud^dxS6+c_>W*m8{JVK$d9cA-eTYISDIEGvQQS
z<0O<+sAD~2{f#===y}R(Je4#oci?@n^VlFouAvTEHW21(eIsz@US43`K|bJG3x}EF
zLI+)i!a}qQz|e7^Hb)>3WG{qYgAzY|Cyrn@MZ2_VGP|rs*soXDymG@~o>Qv|>YDVW
zQdw&-vZ?Q)`!f$!y2i|yK(|Vs9>o&O9ad#)Qa8rwc<AELDaW1M(@=bVBdFQY&u>TW
z_X?BJ;CjRX>{8N#Y^0@cxa-!Xr@2GCLjAv@GQYK)F)lS#zM>-V46zzZIzs<|hOik)
zv|>&xmV!@i-)X*xRe(StkX#ln;quGVDv05fX(o_-Qe*akbcbj$-vpZvew#PzB9&*p
z5+>*J(MJ7c+`HgEUKcb%yKeUYUbhaeN||CSM?l-V1;dp7{5C<T-jcsSGw^`Oi;b*J
zVCjUibTd$u%R!zj=l+FU)1;hi8d}(In`3stFe{cwnT&RZI^>CYS~_!I`k<P@AT*)E
z$&?HczN}F;U=mFK$OxT}bTlq1Ad2I&zAGjzO-QG@J@Wk$Cj0s@2d2F|8X{6gWWmO`
zyW*fciu?)Nk_VTgfosgZDLHrBpBZFykMo^>>z?npr~WwhWhJ0b9Zky~v2sX;!E6Df
zOBc%|OA#%k(@1jo_@VCY@8?q-fB&9(^SbRtD(tR~p#Cu<=_kC&gx}XCl%{-`acMqR
zgWpJ(m)8$?cK6#^!ppO7bR`~$=5&?#{|d}I^6ZhAB#CnEZC|%8v&NhA9CV?fB@nw6
zB()~1=jRE%Ij%}A9V7LNmDSt#Ew<Yi^(3Ckt*C0|L^f9r`)CQbtJ|{_J1bj0j@=AX
zlfA?x)~^a|FC<PQ)?3lNv`-8fUld>L7Qu$Y4JhL(Y~8c;2CpWVv-(ZTM9L-x#%4pb
z-<1*OUbz2Ws464e5BKs>c$&!L!(%hkL><|qR9Ld^e?QzgLw%{x?!#0=^r+SM7ZJQ)
zZVj4;+@ez5Q#uIcuY4xe*_&7(53r@79uqJ`cKj}isA!i{C`<6}n>WI;pY7=?Y9U2H
zf^HTPm)k5~ZQS?wja8Its7oBjhMPnhdaZCM8y%YI!_!%f#n{EZG;~Qn<rQS+y|#G9
z@Cv+W2Xc&#{|8@GZ%bV!wL=-bT<S|A-*P0B32o1xHFMW7kI(EMBKgc~UEh<^Qu#V5
z_-nfGsmIhB>$Tth*b9$t1{NN<^qeK8<#*b|8yDnDE{ELsxrm#i+2+Lh2Q>OtZyWR`
zrb<a&Y`sI@dNM?NmYmznN4(oHHu<DvL6*Y`;k>Qd_o>fEQu0<v`e~Z-_^{q*(58Mw
zR}MlJo76tE&?J+#pp7cg^g6{2tIA{*Z-u|UAa(YsXa7kLX8JX?xvAT)NhUu{DtJ1R
z_)iQ#-YoTWVWDSjh3C5*N%|k;<?eIeKBL=8hsb3amIMr!Jm{yjVk>+(A;4nDyIuLl
zqwUqRbMX#t3kvaxIaOxY2{jQ_0mBu5|9m$)Q!D<Ura9K|=`)YK7$w5iX5{bSORmz_
z<LIBq<F$AF3ft-IiD~nC@j!=jo{SGqWc<Vg2~G!6_wkLQN-9MZwjVP7XfJiCzyDS`
zu8wW{x2i?djgYVQrj{!4nkpr;{5Jew`3&-(an!~B+*3n2u6caEr6w@;6<#Hs_!lBT
zGAP~n7B*qZo}It(LxAO$5i|xctDk#7601J%%WOR*RaI;j>wyW~+~ioCrbeE%7~d?W
zV0EZ|%l!A2Xv7<T`;#KL*))BEd*tp^h08lc__bj-y^PC~wp|Jlg11H<d3G<)HuI1_
zOibZ?V`wTa-w!o_=`I?!e1cc>CJ1-BGdz8AXUysrv;1il4yyeRH49tJj%Mektj2b-
z)v+HvIcF9<)$S%UkBZ>J2dela;u_V^)=kY(XeZ0~m4-<T<Jw=3t#M)(B?J;(u%H4<
z(g2wtbaA$K%8k?DmWO85!UuAu;26$5QcsSm;H5Il?ZuuBNN^aKK&pt(Ae^%lT%J%u
zomn{ynjNIv9ND<7Jv6xagpTuk#6mXBS9W8rh>}(Rmt_~6*!q8~#nN<MyW=7CX9_2s
zB+-$?T^ZkNQ0!4Bp?Qj)1`j#C27T-&1WnWyHhq3C#qC~;Dx>zm+5bYS{)D-Wfdb;&
z4*$d|r_$+FjQgun@tSSgmg`-YTH~mx#Im0Ebn(!*%yYNc$j&dv2wS^aE}J%)$KRiS
zcc#Fq_K664ZA#0AJVJstKnYmqYylGGiH+3P9LZsyw>H7cFss$f`Atf{ID@D_9izDb
zml!ZEMuI5;=_vC9(GH^qC7wmIisyNJubwjxJ^C7QYrzAuvs{YUQ~z(3fDm#OaxvWr
zxyKs^3NZ$3|3dyB83G~NlE=Dnai(&<XMyvc<*O839|w8st2S3I?L24~!5bJ}FVu&4
z&#Djxq8r^mD$PajUUW;`C=mS>D{HUYLJ6)@>iq#M_STRU17Ab&Tb+?sK>NL^OUs>0
zJ$3n6YKZ!)p56JkG^%K6P+xj?^U{OHs@SEUj)gU^u5?M?;I)D2(PLxNLH|7!0>+An
z2)|Ve+ti}1-YV}NN5H)JpVQT<3B49e+r@+hUq(oEo+;dWUcuBf<Wt#nj#u|kh<~jh
z7P52tV)AzZf^_n=RD<@jDWP-(j_<_{E~@P_(^sh3Lw?rYKFEY*mXmIySDU#9OLv3Z
zez|^jO>U@L3*t0<t_At+d{fx!=WfBPodkpagGa83u1glwHF>RAkir5?uaz3+V)ipH
z<tdb=6@GSoOdWP6D#5&mGGzZ{;OPV3ToyGV3MYjTA`Y31q0d~Ec#zyykiv2Oa&gak
zXxf^QDsRWLmmgC94<Vpb7BOmGN>%`20BVMZbFSsq-Hx$?D!LU)&Jslgd`lC7&KFMM
zmF%KKuA@j)aLT2oTk!BnF%<n!3*51m>!W5z!$7}8|CWOL4L6BjJF5MD=o<jN!U4Bt
zTn;FudO(vxxc@>D&5E6obURX8FVbhJ*{3@koM~FUne8q16?lPz#*?S$8*(%PtgEIb
zbRZcP-=hrQQL#p1sez68x80jw*rm`S(#3MTypiXL>_qXZx@pU+DCrvo4otN@=gmIw
z;CN{LAA|Kx2lVC)GSe88NR1)c_`?%}xrmnT;d`^}jl)((KNUfHU8Lt_@^H9^yxJp<
z-+zQfL;rZjGy1<raBmiDpib`$er{1NWBWqKD|DlajzF`W%W@eYJ_-CfC6*EcDrIGW
zgZOTc#k)zz;FEaQSDCg~9Nw+sKRbE(dE`C+3-Qas+Nb5T=V(d&k#9~Q7HLN9nG>Je
zOfSuwlDd-FtZ8;ya#=oL=Is#@lqg{);Nj9QOp&ah$OftoQ6vy<OxoMOt<*|>u}@8R
zs%&WSaCS4&4|s-$udys^1t*eNQ+^SR=ap09<c8jX4g<dj>EaEK#LemMO2oNY$g7+v
zGr6{>(JwU|EE*{EBk(acTe!X0%d_x9;>UD_!mZOHChskb4npq<ES}L(>(t!=8Ei|c
z82cvG>ydCSo5}|X->RJp60w5IbaN-0Q-18M+Xx$#4|x+w#Tu%2X6g62ruZ)0<JHW9
zsA-P;TMdj>CJx}=!iafJg)$>MlR>}YR9<O1Em0>Vd^1gkM%)T`)N~#!q0H}lH-F+9
zLw}QQ`)bwKSEk6k%k!ao#ngxsjmIuDz?0xijZzB<8ipua0w*RiAx3%7dir{(ApJl_
z)MHvwZg>_}V7aE;tb4Bd4~8Tr#B;On7dib7x26MPnFG?tf<?B!sVctkb|E;UNa}j-
z#MA)9@8Br*YQf?T_DPe~UMSIh9?oU>-dX4@K}?&!wz{v<SA{J%(O%8<_4C`c0?E27
zRw9I?2vINoNWFW*i)TD^(RqI%u5@+_Ta0;6E-3bm5Aez>!&ab@ke~RmmboZyKJts3
z1lb!S;Wz5O)MRo7B<)dIFnDb!tR`0Nb_AOlY;nYi-RL=Y=G!+9T|rQbl>a}kh1(Fq
zv7K=Q`NK}i_V1;3lj@{yh4w#wDfVCP6o!WN>I!Uo<;Z<?{0k|^L`hc}TgA<|+-|Cl
z)!BO5T^6H5VQWuj3H17{KK6sdgf1p{5XnN~T(H`dy+?3}9%|n&;;l+_NYTl;Y<yy}
zOZeimf8h|&VYu7I2)3_92qg8!wg)|r(hMS>3q*9Qt!X^-KfQ!;y${7#dcW_PRCJ6M
z-kbFY1-&e(N@T?_U2eHEhDQq_(CdLl3w<>0+J+=C^^U|$hk=)up3uVuJ#(7$O@2_$
zA)pq+55N3uG=6V0+j}jMr<9{2%K_rM3YVZv^ujs<&ZdPHw0Ds!(;LG<fiX^own&8B
z%m`vB8?F%TRw5a{)lv@}`=9#2IEWvhEY56tkJ`Vz&gQ`DmEm|rMN3^$iJo}W{7WI_
zpKjh;g6$=?Q#T~Sdc+fMq)PM_{!&vFRo#lkLZ6;P&U;)I!)=^-fSkLxgkdyCe?^@J
zHM^6eP0;o7LAq5R5L)nPixQHeG?hFVg+xYo$m&in8=bu5^B(uhQ*CC3>k+mhl*sx|
z72#)AGsZr@Pv^PBnH4%5I>$+XZ{GPz@+3gSlDWni_k<nd|3WMmDQ7e!I^cA##DqcH
z-;|ZdK%RXNc$-L*N8Im_cNqY#Fk=<(N?KyB#$CL3x+L<b&^feAz<*gWkm^WvMVpOq
zFTED{V&%!a^-JzJ7d?DUT~4Z}OVb~x9W`C#NYl}Lo-$C+o!EF!yYVplH_F%JlZ5%p
zYvuAGdChL93%5hNIO*Ea6JvUKhuBS}xlOAXy?`5)tlh<Y#XHl_w>=cPZn;cf$YpsO
ztU$U<^eG}Pt=}07bl?gtvUz8uRPim!vr}chju+y&c{h-i_!$4^I-$|BFp6X+9Id3G
zn!OOFX5#0^Z=75x5h*&9d2qBBBi^itD3nmGGl6bRg36+Ta*`Xq^AvCD1LRikORD{o
z7~H^VVt|RV`f{}56n{%zuR{N)d6=HQ@%#WE!sTnt9V<t^CDcc2gN9iRzwq)>&n)H<
zJhaRGKw1dw+~@--;O{7{t&VtlQS{uRu;-qWBh3m*Fu1Q-hyaUq2kil^$lAtZ3O5;@
z%P%HpRZvgI=XegXd4`54Hgl0^1{-M>9*@}#%PR4Y-%g$DqgIOkBH0&l^tA|Bp~HAv
zg}U@1Ug}#(jS{)evJXo@E4A__9p1c5A2Sm_UQWCA969`Q%=;W!mi7C5@yP%kPCnSk
z&7#)Xgy+n0nb&kcdD7LOQ{3U&vtF^ANc*3Lo!OVEb(HrZILz-pNYJa2Ks|h7)JLO-
z#l$4zibkLqXU^Zt(2P$Uac>X4jNgh`%z?7F9FGn8SyI{hjb<<z<Y?gtFv@lJIOK*5
zsD;SSX=2~gd8B>$E0qX8@3s~Kz-O(0b9AWCAph-{2&^~z4-v>dHh8ib`46~Th5rZK
z!E}T<|BWO5d(cPasl-35Qxx#*lRA|Y<YI;G%&i88`kZ?^ey4;>?!609)>K%KfwvtN
z_$P*bv^KZN<3U*QM|wpH@&pKkTS3`KzDNR{MHQ%X4sb+2s<LuGP*LA}pSW-ex8=>*
z^+SiM;Fx{fgpC(djNFPqDbgF{&Q(r=O;FJhC$dLo1iGOhONj>jOvo|H{D{8vpu8$`
z>9-@u%`)y0OQ^bSD3&&ETJ;NAbxITqkMXz$il!fSk%RPg_%Zk@(h0yqJqOq|)R>|)
zbiU?;b`-fhW3SL?yAe27#?Za`h100<?g+LW`=QJ{hwtP|>P0S|r?Z{)4m6?ri2?)A
zfB}M!2XFxcDA&9%jr2+TXqh0Uzf4B=37KWY|DGA*Jn)6<+h`!$xhIk1H_-FTN{}Fl
zzYqqy7Wj_x3Wmsm!P#g|j=p7%_whq-rWv#Pb*>grYh}Ony<zriCWv!E#hqv;%NY0<
z@`uf*t0o)d3tdOHjpA9zOoy_R6SBlHnD1w&YJUaiF|*MhPH{Z1XEa0fzu=e}J}?_D
z!!z#f2*)oXiFPO&!M^^|bs}A33z_bhj_^%n+c=(^%rprGksuCZ*hs7eq?`CKC*OVM
z>dkXDQ#Mwr8s^kAiM^*Qn4evqSPzH49XVt=QbzW$;W1>-jK2`W9vvZfK=Nh3Lk@li
zw`VSTeW=k2AKpHDmgrN6&{67ppOba(^OE5-;Y+vJa%A@tkcHNH!e4b2XuBY9nDOm9
zU~KdKBY5Wvw0HOsLyHhwW!S>DpoYPh1eu>!MPE3)P0V-C*Yh3J?9Kdzq!_lNc7a3S
z;iV8QPi-j#)1oPi%dxjt2Vx^3@uzfz*M>7W@;)!HT19{RMm|rxIz{0`XfBJbD%WXK
z?hMk|=+nICfZTV8Rw7HZSD*W1GwkYpbf=ZJhK=@4S1#|CrO$TP=C{6z*6Tc*=XX+a
zP6-)6-{i9RR1@>DF2%2N!#n&n`lMdyb4#T^@}OI`c@=!$Y;`Q`bxpW8mtE>pnI7m~
zpCio^;Q00e4}GC|+v)u*u>pRQ+RE24ug;I-6i58a$u7EFH35QcEBqW~3ituEU77cs
z+NeOTxahF12-631Si9Ue0l)5uGMST10yJ{oTU1^0Glii<mnq7nK!G34;@a6>e!e~r
z1D@pE>0{~TyL9VkVdAN7(VV&g9ukR;P6QiEzFFB*^&+Q%4&c}v#es>t?dg*VjboFu
zd48_^k*U+~I>l*-#uvNkrcm#Lc>ia4od4P`sFt8y!Ay?AZ0JqdU5&W!8taJ0P^AjS
zMn7qeNLmkzL_^In(cIu(ZML6PZ#52f1@9loEV7Yv?&}+h=8BsNIM9VpHf_g3wn6p5
z+=)t-7{2V%{G#dEi8)>Fl%*+o(#k23GcAdy+;XbE13O=P{D&B{`I>8ya}CD+<&7^l
zx>?`&@lNLAJ^c632q2aT<F>-I1=rQdiTOAvE66UBrAoN*?9vl`QQ9Ip9<g{r)_%t`
zBvY^e>1<r*r0F3v{{5)Z1#xyOiyq03e(Xzk+rm5yfU~tN^5{Vp=|Ql+IpD=B%-)cc
za0{{#Uyw^uE@0bD4}spePynyjepyMih-9bG5Z${d=hSCrM#vGQ=pqd(`4r_r>Qo+q
zGY|p0_fLotdmi|jh(_bbB@COlC@w1OS2)`x-*wYVb~)05<_)U5eF2~ad@mxuD~AhT
zvUoq|?3nb%rN=P2bFZ2o)rqPK)<R$D&tsxz&k&S9C~h)*<x<J$jTq&b{{285Cz|Ec
zvY9s7EJFrQ++scP^<T&snv&XK4yWk>(N5F6v=8M++yU@Vn*m3IunNh6h{RVPb(yKm
zN)YTbg^IoSjFc*$6;1ooi4>o?Mw9dumzjnw3~A<GrKrRnwarSVS&x*w^Nr=Tzx`{`
z^%CUX0Bs`KWqEE=m||=gi&<Ypt}xt3K0zJG%!1IcG!3~DbAe_RBSaYx;w2zse(%<~
z`|$I}RWv(pPMogTcYn>M&?P82*A5G5=Erl#<d3;t+GSW@xt-2ZSsoEx{^|*pPLJfZ
z<<Vm?Yq8OgAlgdd8xg)Wrm1V^5Ev_%=PUkF$UujVR{*;wI{F!5)|JCecsx8hD>%Bb
z%Mg9HH&ylR5<`{)U9|_4bsbI>_-J_n6n<{V07uJw%X;L53&j`}g`5F(fIk6Z7^Pe!
zsS%B36MZxid^@&R>a)L--!1)AyBCr|c{DD0Z6*t+wVp?0z`94)o5lS05vD_=-we1{
z6Gdg8Ohc`#`ykIuZ8a^+VcP859G4pC7|fmnt#$`KOx0UJw#**rY?7ih{f6$`HK8jE
zyh-%IC&C1rE7pY+(f~^)t}ty;5Zfw-e3Jyc+SNJ(_DYMfbYa>8t3qd+hXeT+l*fE7
z<k9<fD^Sa+oz*M#3Q_2n+k#63g3Tl~W8pV)N!Bk3Ij032aKm7vOR7QI?b^t*z%P6q
z$)U-ax0b=@R=i-$;IhIz@$iXQUZ;nL`X5VRdW?=FS&lKx_dOai3GiSnW_XsnqpDSl
z?neuvh$W;e{m2D7khPz0+2pdPC`@=>irnB<p+s>)g6fVvIcyn@9HwqU+urh?i!awL
z))y;4o__t_8bxHkKysNaMiIcWE75eA7c|ywcE2Q5`|WdsFAwTdEiZ3tE9iRs=Thu6
z0TodNr*qlo({6Y=Xw!zEY{caV!hBTM&AtC?Yx&=PHJgH#B{U8OUTjJ=l}uY3GGc{(
zeRm;4ltTL%disE&wV4|@%b}+n4+Se=zH*DWUkL#XLf^DpcC1>2xT}C;oZ`&Mo|-Nv
zuF*8NA<-NQLchF`>c56iWq}U!4J_P5pqlE9j7)wbP&ao@e!Ft=KHac<IPv}nU?d7j
z@);4p)?cdxxhP9H-$cDXU-P2U{e?;}$ME@77M3p7^o2Ex>ROL0eg!-ISZ2AzarE^L
zKe|m@=t9rhU&!w4++T=62?T*ANx6q1ckYxTx4=_NoCzqP(Bm%>`4kN>4D(9in^GTR
z4j6B3Y53~-3T$@Dd&Uw^H%B-8g`{1s1l1TInElbGLw7s|>Ua#GK!e5-EF>xDsdZ5F
z;0?N(S@ww4)5qU*@nUR1d5XWtt7kKo15Tnm=mE>4qdhCIe0b*vN(Lw|zv~FVN%Bb%
z$Q6weiYJo5xvD<U24h@9GL>ku4Gu78JPG{nU<7-^fsvE9#~WQbptW1lWR_b|G@<JU
zzPQ@pk<a~IbpV9ruV&@rQ$*xECM>u5PySG)mh$Bs`)&64nmaJDbTRTx?{V}$+~TX%
zf4N1Nj^IA&l6w?tN7-raEi%HV9uw`IAgoUr8{Xx22~u$WQs;1jM`cDN5gdZ(?}vvH
z3pJ$7fp*j(QTUK{m}<w2bO|1X*-=_OezzHMZhng%QQXREgk!qry{XQ7$0|rqO=q&N
zWolS?egm;w73^>J1I`a!oThLi$CSV%nnJGN=df=wF9(ZG*Gg;O>84u#@Z1w_3j0V5
zXlZ6ISIMZJbc=D~cs{6E{`h%B+Xc;ZRUr~}fZ3935c?508hTiX`hOu0cEDbO0MW-v
z^JNk=-82!(xf+oj7T>MYvo9LErAte#b`m4GzJs2Jky3$;KXRN8sHj)PAwEjKP{O2(
z*g4h}^}t!Rk;Ejyf)3ZxQJv2Wp>+_E@a&MC?x|h$UGyuAoBKg3a`Oma`V|`fTOj_r
zV`DHXM>xF@wEgJ-CMgn_vYkGDndQdfqvpnn>d5||3|986t+Xxzj?f6x2f(T7S3+<$
zAgZp;&{10V0NRpqq;YOsUzlsNVOe<aCFp)YgK@Fui=2tb>q4KdP|IZr?4tDn_ZJ8%
z3mzj+S=qKK^%ob<wDU(I1XJ$oKwrEv(b4W~4%4R(HPn$a{*S5ov9bC8<jZSTj)0B|
z<ULW@EHUz6X!KWXn;2u?SqOWC$*q$Hw=T<T=|6Ow;|ZksyjETBmc3wQ;qI1Y$oMq%
z{Js{qbgq34WI!xLP3e#OX31ZOxCRIuzGxy3)t&KDFydhpnLt{h*(vPQR>RMQvrRPK
zEY^Lvm)geZ(K`_DK&7k7E3mUvRtbw2q8?MSa#guO5dU0LBm1j^Uo2(I?uYJU8i`!q
zMJ)3*bSBhQlO@(4P7Vb~_{X2IXkq~63MlRm>;fDr$e+*yaScsxvjm9?-}TkTcJA(|
z*s!bnqquKQ((m>EaBH4qj#pjg>9Y0ec!;<Qu!BP+>55%(tJh~+n1e{>h0>+?uEuAd
z^*Rf%*u(&{s{a&b2o%N#pF`Yjtx2{-WLVae)#X5y&OLT`^rmxxgN;fg1ZFMZtz_2r
z7s3IEplln(CI)zQ{jk`5`*4?hRT>GmU%Vprzzzpy2Nw(;E`$m+UZHH(P|~`rQEjpZ
zlFiKfBb(5jXM=Q;P*%zb1n@IB8{~c`<4y%8h4XyC#oJdKbnZ#>@+Skt6C#-rigZBJ
z5usnx>t{Y6mh8?H&77XS5y5(-&B~FgVUNqnmKrbn*9U)&klzNSzeMu%@?H1~VY7AO
z>-*~XN7wNv+%sfDvu5%dO?)Cmb@@CzOy~?z-vMXZh<AwIQc7^1J&$<T5a{<H;qy1m
z%AQuyB5aP-^1!6%F)H~BSzkr&SdO433Ms}&uw}0o%pINsmC7fMr@l+m<wz5u(vc(9
zI9mE-<g;7pRPP(<<peulEjV+$t`^>N(~QDTV!)$9d<dUc-HU9U=_AesOr*R}qm?s>
zZcM(mtm)n1|DJ;ir+pbTcII6`vCgimu&?czsig9LNbE<9Dg(k{9{1>!W{&5kC5B0x
zj%YT&SAW=?cK*F@(686HfYp@7+~B>z=zKBHG}Kcs+kB8dRuczx5I%`W`>Hbebzi+N
zYd`hP``A&$*MI@(7R9Fqw_vurW%vPXVFX3Uh21&xYJI}UM#_P)_{PWCJ2?${=MoY9
zUPJL-sM`g(EU%i+5k;}lgwhgn)k2fujp4~h`^xF%Zi#CAX-U(~rzFJOu&Ju60HcuS
zk&)`QWglmWG}d*aN}MkYrEIRLbnrsknk_x9@v3UkiE5JkaHs>pMvzhZkRgf8GqMgZ
zLz7+=qmEAW0h}NTN=|$*aTMyoU6&GtyWqJ{sUG;%$5)V4)=(r_VDBxeDG=UWl0=7C
zZnst7hT`Mb2Axk%xo3UkwNS`n<Dm%$!d|ckF^ys%V@7HI1;WR^0N;&!;2DLPf@f43
zbEpl>_wN9YG3J2I9L7OSr@|6cjN-Waqnoo<m)kqC?2DPrR0UZcuG-%8Ig%!F1w^Gv
z%>>WzEb4Ri>YI(b#iM#N_dkIeZ*bT?hJl<Hv#-Dr9b~ESf{r_PTZMNx3BnXwu%UOl
ztepP;>7NyqKp&&1TRl>Z1JKj&vR=Qhi<b%zEDJwiwcQ0&*DM3xHWb@%vJyZ0mg%K?
zlHAM3zdAiEk|c^ipL4wv0QZ>yGrVyNYg^ks0j_F7W%+keS5tjnY1jiu-%Z<lN4JqZ
z44{>419H{Wl*)y|jc#|NT&Mvq8|(m8%&u5=qR^{clOGa%nU+w`8WwWSfBVwUG8jfL
z0=lDsLghMi4aN?BAXX#1$MF{ESBW87lXXcknYQ7d5yJN>3EssgV;{|V$^}J=6rK&;
zrDgPZhG<>2(E{iS=u=`aa}axs&K;^b(tHJF_os+dfW_IXIY^=yqH}?x{Vda%HtQtO
z;fAc&e9;1%t|v!b%4<@a$#^HZOI4H?n>cc?4ELk-%x&u{q7JwIfxjV_p}?UZWHsr+
z0RcTngcAD#48uukg$U#JbakNwenUIQ2s`Gf*Nf@Wup&^D=whe%AKyY?Tjx&e1U;jv
zI;q@P5%ji~T`XC9;#@K{%|(&;`Pa|rm+HYj^zUi}%N;DoeTpze$u@2wDI%Dcz#aOu
zQNq(*sXF6v<>XsGIWcivUdEQi<Xg{}y=-Ox5(Uiy<O_jeunaH^28aSI<kFgj<?u%E
z-RaSR8w8tnCqAysA6+h*cOG4X=@chV#qiR)YT6UxlkTG5LS5$$I2u*{Lf}7<D;!i`
z+p(ySla5^^iTdwspquUK;KB_Y&D^=>SVCdj90PHzjn6j1PhfD|X(g=FN=*l0ashJj
zK}%RJv+gb+kh1R`8snoVMrPrlgq+n0q!^}$Xg14Ez*|pADmdGoO)(9-U3#`XK;SB~
z_6xIy)zSNtZYUQ!JGOQw7q(VL;~%U#r}M`lCt$`pCp?8Yn)w`|Y)_>(s{mbRu;(^F
zyI0C~bta*GiIHBzOieK2`Hl%)s1@bAEq(0LMYze>(p8Vni<iz#VukCUe`RL5MX$ZP
z#g$I5w7TN{G)(oZu7IPuWM~LfGX4L%oa+nk!gYjXj`SU&OiEPmZAkhf4nJj2KB~zX
zkkR}of1Q*5K<CyovoQaJacw&+@)U_0)sboaTw7{HiNE6VwyV+!$J{p{h%$eD#;5b-
zd|GI~hbP9>{orDzAvqLCus{Bog;)YB>0#(r(z1?l9-Jfi3<C708p$55c6UX~q<yoB
zyT~b+-*cA#*%uMEYdB4m<*!B!;MFdy4}xS-?TTR6f<IidzY)E++CN~+id0oq`<d0c
zyS0G;z<d|@Bd}^pVSTG%9L|LC9*<M8P)Q@3khSOU%4l*vlr=&Lh4VmyQHpbwy-whG
z2Dh#rPc?SRV;t{SQ9sesMWdNRr}PK5nGLq@M<DYS#HuSjF5Q}(e<7kdp~oi(lFU;{
zb;yZxj}c<W>vvQNogex6Eg(Wh;1RF5Yra5o?Da&11n;sSlLNv9ClL~Ov&3puT_xHs
zhw|*|ZLP*1?jjCU*^q(iX8~rwk(z~+4pu#y9=hlJ$2}6ceVcR<T*L}a7O14!-%kh@
zM;N!N+}~>!AAM9&m(3V+Cf3NxQMHG)n3Jvyd~Ahq(?D55R7=jq(Y&ADurr3*d6%`V
zW!)7u2;<!Pv_}p~+wPlSQHvMf0atNBlDVWJi8qgrrdG@zZw0s;t2)9@1FsW>=b_Pj
zlWT{?&Ro7n)W$OLA9Fk)CuVth*dQ&U0v=wWKm?)<9?)jQCq-}i6gw#y9~Iwirftqt
zPE@1n_Tc1}*04`j$BD3RfY^!~2a6SA!r#vmOp5WT-INuBUDo^3qi|y5n!an@X4aEj
zd{V~%4Whx^MoMSwT`k7@QgRRNmkgSZch7yao1@91_6%{sVJ6ck#x*gR%`wnt;n8PU
zC~Dh`Bnl>n9+M3T>cPpz<}l!&e5m)0QR?25wV=(p-Y(ty$)9U`q<VR9As-*rfd+jg
z&H3O_RpY<izwvTOgqa&tjX9;5viEzawMgVD5zw0oU@E%6mb=vo9(+-d)3Z!UN1jHG
z%~86CskY|$Nh)|Zq~J|291H8*s`}y&(E;Wh-gY4!hp+OkpsqqRgjua812ibtX0am9
zxCL@A@RQ-zk=%}thN)oEtdWkbv08m$hM*=vV+;gMl~7VT3aq5DZZPpK|IkjCgg^}X
zFYUB};SRI{EUi!E&nj}Tl&h|ZMq9G6oR0rVx#AO=4s>+10uN%}wxT!!`EfT5j9Hjm
zJZEhUavnzHrE^JT*`MO{4Zx|QBPr@2w0lVKCBG0W!A!VQj0+<%n;^6{dYsGhrO|@m
z3xdN@;p<~n#FAb5VTI#M+}L5XGl1;3s<qPbocm0aTJQ}gxdtg^MEU>$V;&}#ce#hb
zwO^h#wjpFI9w(}$A1FvjqFnF#@643LF*Bu|0X@7a0|`1frdq&@E|yrh=Y}_n_Bood
zG){QbX<Vl{7TNDOD$=!ehEG{kLhR2lA+*CA&Ol)T-2yxFe30*eALr)RN3z;`SDz;+
zq!w41f<_G85}~#!fT4dwU}!^8WprkK877$(t=3SUyU8f-Iz5Qhqk8aJ#4~F#r4|6g
zPP#lwe?51*p!)j#dZO=V>$Uw3u8Y4KPdtg0_o_1OL|qhN^YEj|ky@Bz%|gVPY9W8V
z{|C0YADc#q8w2e2Me!flcBo-ZoWXBh9VxAxG5ne${;T5P+2{L1H$KgDF6tE|M$v^Q
z`<<*Iv4VY#5wWC^fv+wB_je+ituUQX?~pgS#>`SG*NX0I0MP9(1h96c38cP&u7A+s
zKbO(O9IAsJYzpSYdHlNJ$R%%ZAhKPpdERgL%C0Ks-hDxXMZH3yt`hHHdcs^}qo=#4
zN<@_-dUU4kdY+Jqo`LQKtTKyTyucY2I^oQ#i#Z2MJJ90~A=-=?AhiJwwYx}1SZ56>
zKoM}tOz2Bkj<Mrb-sV|kog9#iq@z7)DjA%cxf;lhLopF&`xlmhFlP3{3X^eRoKP-c
z|JeFu>jp^r7&wVhjKOj47;gCfg$Oq*gZVM1xatUzKL$Q0x)#n^TTBx(q1vbTRJJ!I
zQhaH`Bh9~k*{6+NYvjbiSrD^rE&~i}o^!3VIi@c}JG;jRB(az2$~sxJBQXk1E^92~
zz7)OSOJH`C=MFhQ6<GGM4k%jR8NYqm^mh6aRnUyf@>j4l(ANk_<WMwp=gJ_RoF`#_
zKD%-IibTv{<5QED6}3KM{_>r=LdRJgbi|4)gv4YeD3h}sd}2#3L+4&fl*zXi`5dB%
z{Tcr=E`NlYtH@D)KS?$Nr0QxMg+?3_1pRl>9xq%fi(rYtN?=*yXqKs6;Z~!}?Nn|)
zz`J+`_@_g{$3tB19D8&VAG^~E0P<+)E9yUh91X|IzVJ?>AwHYCI(XJ`^W1lZYg3??
z!1V8cvaYAhEf7d)7?2NN6Ehy>I@{8Xlr+N6>A&MKw!Yq1*-6#?7&=Q!eUkasNtWxY
zeF$K!b_@>h{x5LYhMDIg=U`R>2ok_-kD`w*4(3^<2>m266-*)2P2XNU6BYK@wbOdr
zB1u?IP3*yg#RuN#KFg!7J+SAeUVM)n3rHgK4O4A!q80?*Z{mL^d3{><O`dSt``&Ey
zUX^B?xgaOhz(^IsZmBentTrU2UchWfuu#IqA>P9HK{z9F<q6na7|@`FYQ~4T$+gU&
zroC)W&c5unQ|B_zPnQ<voy5pUZGlI?dl?IlOd<n~aP>wd)mkMWS2$A^&}>dZ96x@I
z1YQXed=+nhF{SkpxSC{``4Qx%%NWQRuO1+!&9GD2!a@GIk&fVNkPFMp1wmw_diEuN
zU31dM1s4<*TD3(bMo}A%_>e4SB!ezNd56A^_cOmgSL+}m6<DWfC;DfyIfgj(7jk|b
zUqkKyyR9jeASMFpP@#z=SP>&A54U^RKjrwd?>!9NO4&shRl&~?;Vrgom370io8Ftu
z^tiC`Qftl@`m#^IP>I1N2-+Rmv8DHbH`7A`+RX4jwAl|<6S~G$!qI%zpDs2;_LXa`
zx&az5OpLmu?I0}AJftcJ4)KaCryP4e5u@%#_O}_BChGC_zSuea>6DuqR!D6X{^s5e
z_og7Y$@$m;GJqF*5haiy8dt)f`S%^e_X)(z%iFd3@|M;HxoL?O`KR-zlj;0-X?;>S
ze}qdhhB51M9$3IX2r~lhBnx3iX;nrYieI29C<Sq*B7Q&2n}`m)U1a#sT{_3wKEJzv
zeBg9Q*}iH^>5H+Rxf*T~5^;r6H-l9DNHZa=twB=8G}lyC6BL63x227iz(pHkTD~hj
zG!00Floob2tKyf#o^Jkby#2}0j<fy1^RbYDS3LeBgWVhtuh2V;!ukqoiiYePirL^o
zW)9K~@;P+DqWE(1+#fqFT<hy-kH)cC4zQ(bGFFJ(-W4-5#<>d;U0U`={}A)<#dcic
zDY$-n(O1MiIhx^tuWC#8atb_CX$Q8Mi(Dwhk3s)13LnaSq*;*>#wE39!|zrPZ3G(g
z+gR5oil>hvUY4x@_{O`hw5kMV<kSB<4-Ybr|F``7|H1F6T7evLdma`2)1nXEuIc+c
z_nq}*ZK7UOcKkJw(pv_Gv{`l%nmKVt$_NI~mB2mL*$m^5d`+Yw+g3vK!!ErF<6b|@
zm6R0Qq5lv`ep_D$Uf6OIV**?PO1peGe>FZit5p!nJEV4cf)0S-!^~w(r6YBOu+G@i
zD)Y$qzDfmQ<_lo7bM1K#&H!k@+i9q_YWp4VFwjk^f<TFevCo&ZMb;`->+4>8dMmQA
zZNMV*z%-Dm|5WEQe_Z0-`SftSxTgDiXcS2|09jnnP~y6kt{?qPd-dSm?OGuniH`z&
zf3~J*D8>xPu@}NeN`B~e=)pxHaJqddK~oYfAxCKoDy^tf;8-(@yF2=*A&e}#o#|qv
z$7>oigAF?+5aN%;F4OP^vyf4Nob&EyCf1e1O+dn(yF6#ZD;v%(!<O6KYg)`KCQ@S&
z98?{|p@pRIK+cbb4?(M14dWj6-ZP)&Z?pz*u@<`yJjE&91>_%3Bi%y66G16`0<9Ok
z4<&g~W+}=~Fw`4@db&8)jjGacz(qwvr>l<%?e&*LJ?O$3Q6mfab$09@;UWl4V8>>>
z2n;AiK6U;vYqiG2lW6qx7wbXoym45e7$f6u3(c1fcN!8c-XWZ%?_Og5o;>pF%XwBm
z`Few8Vp|07!+wP<2i0y`|0WQ{V!8ymu_FTswn+YUGfTGyv5U*%u?_A#Z50s-El(eK
zg@;I@TeBIkl|rgDZ(ApSjTV*~@+&o(t)DxgHKB4fq?Wg(2b-2Cuc@*J;-%F21>`<R
zcm9c3E8N%xC?vlPwF|zMgWL=s+(4);3jrPdc(h8fxu)rj>5ZpL=A#AzZX!hoWuZjo
zgF+IN$PeC4Lkv~QmXR%$&A(HowD@h|oB_jy(`v1v+Aqe6ZtDo$#-T!T7zs(yrAoy+
zn5xEFntm=XNwIind|2U<C~v1iJWedlVScVvmY&F9LmB@m+E|NWeZ7#neJ`ztF7g6(
zJe$12%C+AvLmzWl#0lsY;S-3OufFOdi@tjqzEH(=2l;o`$kRo4Bee+eIsU0Px1x}=
zBvGu4YgN!WsfL6r-{rbw7Q?R4LlL>BFJHsv)OKUpQ6NiK6C9!#9@2^V>_<UT76e9<
zxfV}pBX`9qmb?ouXg0HAo3NOeS!P1X&ZsTl=t5&dj7p^pM{*tf9-Cx3O|wFL{4OWn
zumg1<nIjKYXx}m=$^SRhe(ONS-;5kEja<ps5jG<x;Z#C2)jnH0`D3Er#U4&?rmJo)
zy*|4m9c>U~@J!0E70#}C78lU7sVP@t=;uRChlrQgn}T_W-J90(%<{~6*L;EqfTFlU
z2WF28K)N3TwlPN{;6SvCqBH{ht0U|MGLoopwr9y%jSMC0zV%sgMV3ZBXJ5@%W#19E
z8*_f8O1lU;<HR$bUTx+jiQ~K4i{F0Frj~jmdnwsqpZm@!!MkeX1ymu|Xb-Xer?-Hk
zN?PFK`MC)q0)6+$J$mc2TjRpx>Y<;8VJZ}nh=&j3mqWhNB+JFK7?X2$P?21BoGVDG
z!|*%JXYVYy<hCJjYn)fkop}+TN+-scb5dJi7>c7TQ|fE}C~XOs!AY+K%_uEDL5NGd
zKr^X?oItl&CcgJU6ct$<Z1ZZyS&Tf2&$*_5!9%MebSic|m@iNgKh~b-Kgp=j)Hlmz
z6<5e+^u=382Xe^|`^+;7YuEIlhf>lU8KP}K;EwtWad<zrO9hRo;x`IhnmdtteXw(4
z@$8u|Qz(OTO)l7zZ}9yCuu<XprQFGs=L~)_OY@SYsy3x@FV?)`*|XB(S;Ap=;a4eL
zNCFuLKRS`L<Vd|#F~2o2?-TGtHfC4Dk9MG*-~P?J_~lo6XCIh8SA*WQ{eA>>!Z2?r
zW*brX=9F{AUg=N&`OO~B^6J{-JKSlBmyI+tiCoPnaujq?VqxnL=L$p(BcM6oeVlSG
zVe!?7sE);(MC#Dtv|CF3Byh~9`Yj|Kov=Z!q~|j0;c#2+8h(qBF!nsec-}<^kF{&6
z^DpdKf1f&8pTt6}Y0J^X|Jv=(_-N^DY0wq9asB;0lcH%cLr8=as`0Xc%OA6<fc*`?
z5R0L4P;zJ#CAuM(iD7^xxWC9(Ra<&gw7sT>?Z!&u&i!p>!@Qn>B6I+erI5(n-j;@J
zt7+h_es=3-UvEU?zUSaIUek{>xsxZ=czY!Nj2>gzDAIM3C33Z@<`}I7L2uhyw?1}J
z>YM)2?tVHW)rjBuU80f1m!g}VP=4fM3bBFG3Q~L&slt$s!jU@3fj7H%zSKdFZZ)9k
ziG})v;N<$kC|O){*^`S)lI6+mA0=DlQ6D)ElHm5iUoqQfT`+s>0FHsF)4%6>oicRe
zOU}{l+Sti{T6tr|&z!dn?-&)IYaH530aGZtfI7qpQxdO3ia<XxI(!#&TOP>lw%kyX
zTZ5}YXRzQfNpe=ad}Pyp@CWbgM%g_b^+>uCurkQDD!RA+kfs&0{bUwIwO@LXhk_vZ
zrvk&Az=J{ncTzIQr^We9S`dfYYjW7!*{7WM2J)nz+;8YPo$Eu(nmmDX;vL}W3w}KI
zLz@Rk(6S&djsPsekI_o1-ypsJA580TjA>C66`On0+I+-3tV6UhS7GkjZGxD65sw<J
z1L*7|z?y6beAb-DW}xNVq?B&x`W|xW0yWUo_d_ttIBQU+0tzy5KC9^-2$y+eXsWEd
zmWtlcql1FvR9LY9V@27<Nh3|>kr6(E`{Wko+ROksSQo54lapwYB6qv+AuncI4+QB6
zWA{p(wE4k5kyJ5V>7N)E9A@qi@x?T$egzp3y1P%y3Sd#y{_+OkZp0MYQC2AO#q^NN
z$oY&wI?(i#ZwqA3{e|$O?eViQg+%u&<v(^@?5l@1MOi!8_5_V2>cmow7E_JZEloj)
zBP2;UpE7@w&&KSl%8&7J1HH=9Q$1P`lT&sZR7&W7D`%v@{&4KA8J!cNjv_ijBcVGA
z$Y0va;C~(2ipy^a79k=sgp`KQ!dbqpDm87e)YYKse{eFICtqu!3qzm>)DJ}$F#nOh
z500g8$S?f<#(FKX!>UxspXb;Yvd4cP7HetN?nL2U-J@Kn3+rmXd~gL;TParWL49C#
zFYd8F9?_#EuT6!17Yvg9>PaA0^hEC5gPb8E_n!aB7sjW%xc^~!g?oK*IZs=sg`^))
z*wH63-_+IKq|Cx=-TE-k>cH-1ZIy&SQ|^Tq7ao1P`>~aWcY_{eCd82Qpo`Ig=fLvv
zN3LorQptP>J<<sJ0Fcg$V)L|Rnx`qfAE$Th78zOYf>x!Qz$V8V6}fLq_Y+4b>1)m|
z(7EkBu|Z5#YT<uK|M@y6Lo!DxbxuoCNb!p7qz1(Z)`~i!0ihg&C8X)${J_Krx}=rx
z1bqo4B=w&M!D0NY-sauu#RtBEF6%Se8jeq8-hDAo?GwEh`Vb<bqis@)<LQ)}#7GzK
zw9liU%z@HgS*!OkhvJR^=SwZPx4W~D^vf%22#j|ltDfNp+~&n@B)xg2%MUbVjnkWs
zz`nek{3h^(s~_zqxx}MPy)M$_)diy3WQ&NZ7BhVbOVT8)-Ks>v2vD{${@w%P-z<8Z
z`TM65O(gRLvzW(enGHOGBAc&}YcPv;_%5=4A6T^|5ZS0e=#iFo-Gj4)OL<;$Z=UNS
zQVZmj5oYH(`C);{QX^1DCfWJ-o6Ro)A6s>D$U*PUsMVd>012PTj|Vrfi)gw${%Gha
zvme6MAnb|7w7aan2in{)Rq;HWnKIzIprvYsNTlo`)dCv=c2XXsRJ7GOn%Zf_K!Dlf
zZWG9IT+99+RqOwICV;1{HW4)sqnfd@0`{8Qm)y>Drx>zdKk|xC&c(84v9GwWxBrC*
zHd2+YE7#j3era3#MXGIF_Y=J{6Z!<gxFZC*uWka2CvN+l(s~>D51|Y+nSTgnI3Bsu
z0f@O~=oIOEnr#^}nwwqsGtaHLGCFxhKV0u5i=Zb8d3}+>J3*ocbpeSH0e;$_#ZQjN
zehrgVe1`GUnaRl^ZpHqYn_{3C(;&ySS(ZkY<`|KS1&F(pwo%NnsI*fbtACWg3IGZ;
z(6^GPPLYlv`FM&q*A6BrY*2kKCsckfgQ>3OcqX>omu+)_g~s7ZG>Oi?mx3JGT-k54
z(UbZ+d_SJ}3u)NuUIeCrllc_mjQ@tT!A>R)Z=KuYt^zx2v5v4eVh-pzn(Utv6yPI`
z4dl>Co`-%GRI3e+8b$&k0bchSMG1v1v(lsUKfM#I->!yk=jYHxt5H^@Au=oKnFqpq
z`QY4G2aTeA03tRoX2H6X(pl*LLdNxEfo9-{*utX216@n2(Vs1?#g9IHPPH=rbx_N5
zqIMAntiSwCgHIEMu^75_f{VIMc5lXD*g2P?_%L~7jDhx^>Ohmf&;Km~r1}*sLuA4t
zIS^s%=9r4Q$lt3qy)sX>f6O(!>YCLu`t7%N(<xA!2%Q(ES}R}k7tHDlBpkRX^)^3D
zfZ1eu!1O>!wX7YX5AidM4W>fyf34qLLI*hcGL1|fq{casDjl?}c-CMN!&##>?U)zZ
z?D&)KJwG}sSMX2RZ8xsJ3G7S=h`_}YQ1AVjfGq<!>bl@jhg@YoAK)%G+)7a2GkR)L
z?=elgndo+1`7Pzrf@CL2;2MhK+?m!MEi9d2CK&Nw{RQKt3h5Rk42h0npqhZF6YYw|
zY>0aPn+vA|X0!tQ@SqN}qq;(~ISERAH3KZj*(9n@&KYI{c0R7+?Qi&`oO(SkYCSL2
zzMaz^Iz?UY9))BGl*3bu>l(G5tb0A9HuBUsNPQ}>LAzs(O;%4<$YS(0HK9&?IRPE1
z85}AqM;5#1<+7wS>N*^0AzkNUwv_1Kvnwx_v%Th6PJVqy`S1bmXOrC?vXc0u+On4D
zBWm5f!L_yduyNrv_F;&I`d+73CQ)-&X{TmGAAAPG7|eRVESYhTu6QoEZ;qvOOQM!5
zW6-Y@{w;Fx(|MRa57x{9e>A@w*~~van{b+jx<BxH+}juxwR-~8UGEZaY4w!$muo%*
zw%Lh*f3*dS|4MZ(i$)zek0M|-{hh%@zAhn_EAi+3JGz~^3}IU9U-0hnBn1MuVPl$J
zecbH|>sBw?vSghH2=9mc<_;fQmuE3b#cSwsJBow*5tz^^P>v~;G=ZMq7-RcZUH?LN
z{7a%nMU}$1J$vYG=|CHlrBNI!%jA4tnd?4bb?m|bQf8iI*|0K9{0TPq&a@CjX3cAI
zOp0JfF>;SbZZP=)DMBbj`%zOlIXnILQ~0`wyXMvN7t|=iY3~E(TFe7!SRzWH^H$f&
zEh{JL{Z9S){->sQ<_}d;L#H|}YizkG$mNIy*)7q5*_{Lwm~Lu_HtN%4c0_4W;k5bM
z8ADG8Pu^Jb0$#H_mvuRx#6k#HjuZnmq1#(UNk;+Dc^hyLG^)&_BDFbknAGfsc4}3u
zt*vhzNQ`7B*<OZ1q&{m&4Kzc+sm(`lAo~fE=}k_ukQ5!-k%tV^pp1|ML%Qi>_?5yy
zsA>OBzP_~mbF1yA!RqrdUVR<8<Q8l(W0BytX`&}QOKHcjec(V*SQt4DJG306+JjAu
zQNs9^$$!A;thSr;0mw)@Y)7y9!=Xj|6@#VCC2R&x<}6g@ZSw1Z-pg;1k<Ab^67NfT
z)0K!hhA2@EI$X_NN9xEyLjcH{0jyh)nyJ(Px*EQ~?d%s9WxR%7=dU@9IQz#aFiL})
zMi{r?iNF)^D?~Wu2Bu?fj{`;)bkU(32eAw6kbWJt-wU*A{U8neJ19d3)<!At^bh5z
z8oBbICHDo{9N*<qWE)1=<u_`AfpTpTx=V)s;h@^|Nm`g=S`C&Nbn^~+KH>JXP*^_a
zbjpvHdk@D)TR@!wqL8Fu*gkhK58yD?0%+d5D5e#EWMrj=mZ(9El?%hw2@aR3tW?-D
zBe9&t%;UZrodoxo$|U-gH0N27#>AiB<hi|`StRM-6R7aP`U_v5KvR*cpo_1ZQ2!P_
z>+sLve_)s}E3#gHAt85>1f?2+_EC!^XuTqbT$uXqM^f<Uk>+NtHt*>lVinlxoDc0+
zp_#>KO0A%Tkv%8`3#BdUzy{nC+2ebnzV%Dx_?D~^PEXCh4LtuP$;QQ;bnczPgS~=O
z7X0hz?_8Z;!{m*9G(bJ5*tu<I(7P6ToLtRr^6za`2e<W~F_sAi!jVb%@3)eG>2}9_
z!8WEMb4a|UWomJgHl+q_tB^@#{$J!Rr4CJr%{?VjLC$Vm`3o7^MK!8=Dh~}H!L#7r
z(SVG67O3$vDsFaG@PuaB2c-|bpEh~6pHpAo?DU}T)H)H!00PKC<n~o<R^+%AbSH-o
zk~%rC=W}GLiW%i!7Pu~P+K(ceR#*~rc3}XRM0e%z4Ly~0J)3s3`f9)_>2BD0t$b=_
z+h|i+nlQ)?4`~7Y6CDBFHf{u*?E@Dm`OHhImewOBl4c@e(Z$9<C!QyNCc7kj5bMA>
zOZ?tx%S-gix9BOgZZEXR(siZ1oqEmWBrAu5bG$c?fSDOwDtP+;;_W@7n*7#p(I8+@
zDbl196cnWjNEHZJs0vm(1Ox;`M5IF?6e&^#6clL+h>8$VdJDZsl@{rQ&}%{ski@%i
z|MuDcbMA+G$GBtP4-CZ--z0BYYd-Utb3T)sB47#GrH=MOKaK|Wg7YRNb>_LV?E8d<
zqlta@@3*(;iU+AJ4}C~D_d?{wApR+jsevZ+J+u!azPvSGkJ=DjKy2fH<>kXrkb-j?
z+)>{(g5A`~g(&C{ep~p2QR(=6Li`(KR(iS$G0hkaXtL5>a*P$Ih0vxzeVTQ2;S?~k
z@65)hpgy&Wrd8e9(h~xK0J2#w+wIK(Cx}p7N*GnB(a!Tg%c(#q+~0d+>;jcBE`Ji5
z_t<WJGZ}N&1&$wr?Q(&Pg#@R}ADNBIOjjn>R}{U^aWS7MFQ4Q~7*64jx2<2dYX1Mu
zyZ*oBZE0|rVvfn<Nk;v~)BMlczwBLWk?8RKAVI`9!KB<lFS{fPR|~U10$~2ULr>*?
z&5GZPHz&LX7{gi5F^`YF5O9mxZ3b(5sRQ}-JHS0$1)Jzu^l5@IInIoQ@?7!{Wa~jM
z^fv-#N7AdLgpjB_IfLY+vK8~D5LM#bW8s*2p4~!y+*gp@j>0MQwS>{n_?4y6XFoc6
z^<)YukSJcI)Z&k?@!eew1<)znaGwDFd<ZPeV9h!UYslO=1$s}vIy7Lp19zmyVR~ke
zscLsZQvHPDAp0Hc7Fw1hq?O8JcJ3wfyV8Uzh&B6YR36SS1Z05LRvbm=F49fZxw9L}
z{S<!E$@zjW!gl1)`#h!UMu@1dRTymfjCOjczZ#+d4smA6W4t)#rBeGmx4@UO=U|jH
zUxVL_N^L?;+30)dPv3dpYj%GiCmD5^7W|EK0Eg&`N5>vE_sjRSiu~meEm0fVuB=HN
zZEZ|(5!^~!!b8~v<Cb-LpGK=_?LPEJ{l!;r66e^tU&iC^_?0TR7NHaVK%CHfQtIdb
zKma)`8uvM{2eV;%sCz)54uo$(e@`foMyJq6m-`m>YqnMc*ur1Cyq9HYr=AkdwuK5#
zr%$0`tr&?P%PY0w17s&vl4GS#%j~_G2^0O?VxUHYj>9;{?@0=RI8pR#a1x!2;1ALN
z*gF!z=KmI?)W`E6-3Z#xJr}nE?ofTDvuL}`9T}tfU^o%Ca&Lhj%SNhD;=B4<r6It-
zdgw2P9h~-=E{|-nhuDSn+<}M+24%F=N|Hb+coqz(-Tng+-7LsF?6jAKqyF7#&nF{@
z{dE(Z*g}t3?OUnlqqe0s&c|Nc1}>N!zDGRt<#h`J$S(r2LK7G9YnYda6{(kFO%G&)
z+lMo&HD=<<+ZUMCVWdXzCbliG9WFq0{3@{2J}mst_BUqgXHw232dRG{I<%sWkZi3g
z`=qyz%~qGVBn70hi)F`s0&n9XTr{=8<OsN7430%$>HA_O=T_#EYn!L|B4ciziK~*s
z3bb0dFC$KnO#nE|#lsGmKw`$iZsF*T0r&?j=obRP<S8Il2PK-inh$%U-|9qC0Lo2r
zJp0YZ1J$F~j0KFJ;)MtS3};FL1<4kKlz7hzTa=?ahNVpz4oPl2FVg*ALkllct|QM9
zwGX|Ym=zg-Z-DocYX=k!iOo=oui;_jkK{Bghzh}$yelx+)dg|ah|ZL@a-V^sC9lxn
zjE|+&v;@Ge(N>mM66VruD$}oA>TByOM`)N3SdgA|g%iMpb<=w%lyEfmtEE%_5|>=Z
z<2l8|k+F6jn}MSOPkq|(Q46&K<hz8jCcfx=SLTVfU86P;AJLeWS-IFc!7Mf!weM=z
z1O5N1IPd;j#p(Jl73Y0G4g5<Eup8^J6HW@qeUzdg`$&<C<;k4F;hXTWVx5_}{o~Zc
zb4lJr?S{#dXR9I_%Q|-z>MoCVXIO->@HJ%QC(;PN7gp0ksg$sgsO6y9T|i!*Ks^AQ
zHE$UuI2T-A!Y$~C$Nzz&J@)_>iC@iMmmR(3u=&p2lm{8&-jfpEGPcz!_<A|>#|(x^
zaPl6EDO<+k$z6}K%eBACXk8~nR=WM%(&kH$+$zR>nMGXWSgW((fdJ3kmOUf@b=z|9
z$dGHv-h)wF-`Xb%k6SPQXpu>q%l#S;HiowB0-#ZencSD?h5qzQ?^#)s7ysrWudVJp
zTR_IK&QN;Zt?j$@)nIl|Nw}@lQy<j%L{m<_4IKKOytx_rt#;aAvZkzT0)T;2wO=UV
zexy<A7;0xaNd5Dd%Fo{|WeH0k$DA6t6wl4OzMgftyt{^fB9}&tnJX}sB$`uoz5fRY
zwyn+F@jkHV+Of2r9i`k(6Gh;778k9E1YScLU7T|~$!MWy<OQZL`Q}6*)ot#{*`oNW
z(L&{}K&0^y=G!_sglrDX_XU_O9^mTQM}J0`EhckujXWogj+nRDCz~>Jf}FXVCWXRy
zZ*rpHDNtZ4xKikPp}Z(3L_}7?PP)$2PO&PrWY~!+PjHPz(tmo?<Lv^ENLJJzthDfb
z1U)^l!F*BMJt2yW2g-{E#Gd^63-~s`!3(SaJjDa>6d*r&u;fY=tC_5qL(0LSg_Am`
zoTlnTxbmMdUv(?jW9dY!*<!|8ACZm#*aIGpbHVuagJXXn`eqPBF||8$bFnZGpg7|F
zwOY;;TnhQZyN~_qb*gNiril(qu$UL_o4yspG33?@WhCdR<A5#A+;~X6{J+>zvjgpw
z#Xk`B%K3Q!t2wP!KXef~$*k#LufHK&d>;J@xyF?6a&NDCeAI0$y?hmsxxHl^1{%tp
z)u0lv0Sm*HX8}ZX1m%F(+r*-jF{Cg1-ngqIOY(`2nMWeSBg%Gd+{+b-$K<-n+dUi2
zfse7{uzx7heL#`Yj*7#YYeo9VqFI*Cyt8d(-^+L&E_E7Szdin9>iBQ;DM}$`CUzW!
zoq_FhLPW`z(4Z#R8vxWSE2gDt850@ET}-$XD3x+A`hfW!6MsjCPE*3{mu=~nnpVG+
zTbRK6^-citE=h)fri3AY1K6S$S_E*e9$EDdB(p#pguaYa!Xi_5gFGk09E^x+T?O-D
zYYoM+v?;#Ib<wbkFKvQp6T-V8@_UP5^}|c=$;BT5H^M_h3iC#gL_qeACpeCp_vSmW
z2%^1-F_VlKoPnwOSBZD^H;SUqmhrF6JA%U`t<RHcss7kH&MkO6IDb9W^x*sf;85lP
z0nF^bj^e(Vd3coELZs-_B%jZbM*X<%zU9W7yt)r1;|4WUID}vNg-#~2_>m{kzhJ|=
zRGpqM_;Np30v<u?L6FTe5&$I{`Dij@3fb8Ze>SgiRlM$0sg*)$lmHu4XJvJD;A>Q4
zR-_aIUb&Tf*(y!NxC>LdPL2xX15f`je9IEzc55uK`@?l$;TZ~UJD>s?)E|hvzjyOu
zEw6NA!Ff`Ji(RjIhsuKy2Vcu<L4$6u)fTrFnN<WW;OdG+$q6ja5j!k~6w&Y?0T;J}
z%O~E%56ipO!iRW%Zm(~8pwS*)MFC=+i|1(bxNnUHq=mv>7e5}3Q&?@0X%XKT!RQ5Y
z02vR#n-csOvvuVb)Y!`SKtrFfuSRPKV>3p&$OR<W^kk+1Q}<@s+S=-_`$&yJtr4Z-
z4`fakoXd4eSr((E?EH?WrB*K`#fVNXQ@%CLmOs8$F<=t$Xn&_^^qmsZ+sKhO*q;~i
zHaCz};KRo}z(_O)rjf*{*Set8>O=Hq)_Zgih_9idjUL5)txb`srqM+u<}MR~GMfQ?
z%-!ie-Qw8g8TcrQs%6NC!hW6H=k2A*>N!wFxtQ#EVHW{XHrx7X16B2X_a?Ts0zXt+
zWx>e@vL<amif??!++Bs^`~HGrx<lp5|0EOt8-E8=S=GgWey)GG89jDm*IT37Yv_~6
zOCh{JV~ms!U53t~LRO7y=Oy9P=X|r@z9V68%U^1VRtDb@5e=>h706cGXU?NdB_c-6
zsd`V6KyBj#&XLF9lmdph$#u{O47f`HK$FEE5f{h@>@#wshSLG(F1Lfb?b$J_bI1MJ
zVvY&6=GrbG=mG^7%xX&WK9YRdHipZZ`YgJ{^1Uuwwu$&hZTW%0CMZXyyEwJNs=P)b
z^k(Mm!RI2MzDuF`?TB+69#bG&XG=}4N0iWEM&c5U-OA~`ba_<3bavEPepBz)F_v-I
zrLVN?bt0avXR6nCFaSU_0AC^wO6{_8A@or7w8MZ^)*0XnU~<-906`6h*`k~9NE9#m
z^wdr71DJ8@T7af4o4!EK6+;ohd&&|Qb>;=0YOcu{*TV&Yd=ntu(S=fWt5GWpyQ@3p
zS43hdb@lO`zAbKPwx1#7f<vExCsC0Cz}TY1{TGDGRK4}*s1;$Z%`BFjS|w-L(E{=#
z0%=maV!(4CjIPJdQtycDmsvXvb8fY6hA)%!+wmbke3mApp$k|)uX5t|QC(!8Y|+~1
zN^KefGc%NjXC4Z^l2iPil%VLp@iDarwP9K<4)8JJ5Fhs<cuV3#2&B1bY3_wxv$*mE
zq4LD0$}qJn6{aUo)eF3+r&YDqC%}zKa`0F-7(Ez#;8jh$7`&dUH6gPpVV`5o$Frq$
zwAPRE9y3$ylTWQR24brMQ%WjorVJ1vm0h5i6v<*KA{!Hckx+cde$cb)sPg(OL|4MS
zgPH7Y9Ydo&;}E=fm!yc(4k7b>QOL%Pvt4$c24DD6l58mU$<Af>B^k?R4x0H1&}x{1
zb1cg}svHXC$8OqoCM(ZI_->!(J`zAoG9q&k(DfN72JA3|xfcQ@%XumDs^4$?tZ7KG
zEdvq)tP7x2*I(2!lAui^o%@A9aHLsG+MKCvd9uL#L6dNh;pdi{Rs`dQb*PW%t>lA7
zGYPMhXQIguo$TI*>|gM5saTJ3sPTGr3!3heG=&aQ)%`m5sIoLD5n&<F*Us5g1a-ah
zsm^X!X|A}MmC)EA2(8K3Y*bM(v%C?a`1)g_(dfd7qET>+n61KYz?7}C2mTT~N}yKx
z<{IqB4h$u7=Op*h`|L7j$=IX`vItyvCCh`!@UD^GSR*rzvfC4*d8yWwF<RYk>+S1>
zf^>QYjO*2vL2ATk6KsbSV8CCoRuu=lfJ-w2zkIWp+~@b8*#a{KF5~9P@-3FkA|h^$
zkJawoFqXTPTb2F^r{$N%Y$#tb<r-ndNe~~ygl)Z9$*q9spH7(dqAi^l37({lS_VCz
zOGE1PWR=0~%js8k8$a*7i;&fMP?368!tW(GMnIT{;8Bt<&h-c-hlJ3LYhQDlYqRsU
z3|dfB7Gbe=v$Xunj?>KXrqrP#RKm7)y#)m{FLhgz^FQM)GRp5ig-(foBuEFT%1@a*
zu_d&;6?j=;-B<Qz<Fakm*i0JB4SbU2O%BLi``ELw*$G#L8BhAL8UVJrbbgAZ(fr+K
zhor_J)vuNzRt8bWlyUd#Xn!*vUna9pWwQ93X>m4WbSnDYD7T)%+}-VO|0Y>!o>tWk
zs}1ynh&zdWk}69#Z)WM!y%@n2GO$M0U5=iQE$?=35o>-n$%&as@oA;jVKetnFVN2E
zkGt=9JGZwJT7I0DG9a1RU+uY%P_p)8WXT2FwErIn&7_Z!8)0Jj@d`dI`kgPW@AByd
zI|+S|CYpGWL4X04QL>F*0Z2+{5?F4XT`5QPAp3zo$XBJ(nIlkA<u;a*LS7YeF?q=}
z<K=!2ip5`$hCx?0x*6#Bn7}@-d5LDzd>Qql2voe&DMRrehtIWA<I(OU@6Tj~B;0*U
zPsD103sRIJWtA}Y`9T`8;a2n87u>C)LP2WcQ+oSAlTS|FZOYtVL2WU7qCu6R99;J~
zw<iw{0Feem`w+Bp3B$E4mX?mc%*=gc&%U&}`Ybm}P*EZ%`Jwykx|y^>xdcRJ$(AAD
z8ZWW@!?FTs^Mu-s*%Eio+}&}d!mT!OmWH@|BzJjbAO3qpy-d@xCtc%NzL=b##FjgH
zQ~W>J-gFzoGhc79gra7;ssR`}oW6Zo1@ePZai9WZ)96Cf<S>Rr3E(2^VCyDCNj6OJ
zVP6Y!2_;P~jL9)MvSmbxAo7H@lCr31I9&`cc?n&HvV{Y)6nuvhfU4`w8Uie2xyeT9
zeJ&O%6fn2MpIdO|1v7{6VrLiet3Tp;49hHwngxBfMKW~N+S}znssMTLGX&=oq@2^#
z%{vHwZT?EZuJ28C$qDhtKgS5nYzf79wzW7H9HL7A#bPc0ftNnJJM$}`%YLO0;0Y9v
zvf{Vbc}qVNh65QIz5O}AW*qh5v@W}M`R#X&YqWE0Ydu0C3=eebg*SI<Ul3vBw-Mp+
z6;ML$tIP$lto97|Q^Q=msZ_9Dwk|KG%V`lDVyJw-O@99#wjvo!(<=~Kj(_>I)xu)|
z420IhJ<8~=73?IS6z0h<(Ll2~ilx5R*I)v$&YxrtJ`^4`6#!fU{DoL!0wO{lbwV^k
zdMn!L(zWA7BcIL}tuEXvJf5(`0Q@iu;1LI(<Nx@7XoLR8l9xKTC5o4Q3S5LV60Wk*
zU%TB4QMsRb9`aS1$4CDvG2^9251Yx^)eNH0oz9rAm-8<$!fD&4%fJCI7!GgZKB#;n
zEVe20)RYTvNOyS!ULa^ksNQ{lHQPD$1{bdf{S6r@+l^9!0*4Tq|EV^=F?cL}El@vp
z8<l2^5ll0m8ELJ?E|N8qpbO3eF1V-(23%R}nQ(cA$k5|olpxFvncy*=orJm8!iGtH
zO-S8h#h1D=2O=le*pneo?8Sa+(DEp^3JHFw(fSAU=>gC;`4{vlT2j)J?v&Wzy3Ed;
zy*g)csfYKOvgc;pJ#TJhfq1=r>1k*ah7k$x^Ejxn!ljCLOFsB<TUaNEwn9teH;tA1
znDM*-5bzqK+R6xx)OtXfW@%~&R@olhARS%WIpIYeDl@e>o$73VOwd44hS>lb$Ye$O
z1Ib4!l=feWCz*J~etK(BljapO;WhT`;PtDO8Si?fDSUhg5WbEa{1v`(0O2b=1!_sj
z2rqkJLs-jpkxr{RSLJKQ@tDxf_*G%PP?K{D))XJ5&O`D_)-UXl>+w$Ftod?G<TpGf
zCWuS*nXj5$XneQczUU5+T_m(NqIORF)xG{M&KxnUQ;ExSd1yc2qPj;H2xbc<Xc<DP
zo+t@Ypn>8e&Krf8`~QhTJxrhaFZ|G^5L<&(W$<qblA2X+7*9SFX*ifrys^#qGQ57P
zTg4dufqatW{b8weBsmjTp1d*KlyE6U?&p`r_rKTfB@Ea-ed>qJx#dXPqkaKt3B+X1
z(U)CZS5vsk^z6)?FByBri@3UUW*yO_&CpQ~uP|f|?3-X1+OAT&RR7h4?=zjgn>XD9
z5!-$@Q|=&xqi(?&Pi3`Z0!K18?rmWXXhBhvJ&$JTh0k362p9~Y7CM>5f~h@^d_d@F
zj9xaEbVOK8jETORG1V$;dt3K#>?cIzr1FY$8liRd>{L3+>U&=z)1gO&PFHP)H8K=d
zdCZhMCXM+``#fzbVSQAQs@I%H33(6p?L=%mNt&uHaDs{fvddr5z{nP;NOHkR^No#K
z_+)BD?f83M;14l9W1Uwd^F5jWDYIKr+{YhCMS^pkI^R&iwhOHBS0n0){&nH8uJc!8
z?8GYoC~I4_dhP_=_0)4nX9Bvh%Su4nh-)`xEyK3amaQVpKw(|tr}8|+4kUlbB~4*Q
zgzE#vi2@nXE~oNs-Mt!=>}ta0V`o&w>@6cj2udwJ=rokG^qwSdJjLG}P!dHJ@CMYL
zSRhz(zICNg<AJ68!l9-7uT%0U341#Z$S^BKyuB0hlFymughlNz+S-rD4R?#N97CQV
zC_XQFKIkEEL_9!~<HL>34&Y3B%_c{$_8kaJmCjZDJWm@*jBm)K_vyH2somtq@9kYr
z7x=74LgJ^&(T!p60E3%y4e(>ttWy+=M;U^>*(DeTThk2Rhhoff#X2iAcpiTI;c>p4
zrwt=|H=yWzFh(kwksjQI=qNg;L|+?;qi$6l_H>&An=#BmcgX^Tym~99L4#ae*YV5x
z)8R5;40rRr?bbJ4@t@|Z9+^Qlqy+<e1z89VFGPPyKNyroOV(|Uv~bwP{(+S1SI!32
zE77^f%nO;3T!}pmf$G>lkfP$^fye8^Cz?`zojFz-VsgRwi3Z&iJ}5vO4`}5L6IE~)
z5gRC#5lRYbrqaiOI)U0s22$vKVD4t&C+4lPPW2r3N!+|%5q-b5qS#f+N#cEMwl0+M
ztrJHg3pB<d-fWQMKrCy>Jh}bCqd?@*DZh}`Ic=sby>NS;@o)JTe-d;=`1xa|@Y9CT
zu#J0Uur1iU>!Ap4j|KvBuh|xQ9|s7452;{M_Thz_g}ty6C9#9pl-jTInI|5lX5@Td
zt*GoD&JNdw(0rp?kY%A;<q^v8%b4Hs!3Lh1e3Y7A);NA@S!W<dZ=+zxOxG65qa~{w
zMi&zkmmNBa@8JjZ-|95W0fPl<p#^PDiN(x(9Y_5ZbEWWr1f8nCX=R5x`8zTVN%}XF
z)_5U=al@Sqf_!+kn`83G_;lmx7z=Yv)~FjOw&a&=YHgfjNGVbtZt}&&2?2)jdN;?$
z&eH1Zx(XR_3O|nI%y@Qd3%O->N9`N$oB<17UK3_}7DkQGfFbTmor{||CYnB^^x)3g
z%2Q5_c9H!DnI;pm;AL<?O`!orwq<?`8b1*@5O_3X!{<Q}cFi_U>w1Q`W2p{J%j~$&
zW5H}8*e&s2>Re>K-awJXGH*Ycir7_tk)2%~Hj~ZtoFJY0`g&*)x?5^{+i*n=+cLR>
z2m1{N&Gt>>fhUqKmk#n{ujpid4?maaox*qO;;{$Ac<$~;|2L>P+jX@W_5)W_?n=5M
zzo(fH|CTN;a%}tlfzF89(W7eg{yp_@R-_sVGBM#JNRzXuKtEd>KE-;8n?Cwy^P6uY
zvV87OWm@{6^V?v90)Vsq6bTf!*5Hop^If3rpQU^zGbUCRW=``!M4;x5PbcuQp?gtV
zm<XyOVYg|WvCBUvzvGc<!J3N{{?;P3qfqaqwkq4jfM?LT*?{1Kli;G}Pb2opfS1z9
zVy4tWkTMVSeH`fa5f;H<%@qG!)=NG+1*aV!DM-(a2w$D{mAZ4w&gkxL^je3ypL?Q@
zU`XYUq3VO3AW(8EPGj~r;fG7>I}GkTB?GmfMTAiRo}%iZvyf0US=lud%e60Geaecd
zOjo$Nv@){dYRlGZ-3Ke)(q59~w(@%ih--)JT#_bnm+~A8U@w4x67+fIjxW%baW!S`
zf}`z;)h3ZFMMG3?6S~D+kC^}>jbR{aXydou+w;Ho`fllVJUb?mHp8Dbhq=u3ukb~Q
zLDWCtiz;Z6!(_M`SuXwn1xj`NYpjJXt|vSYns_f<Kw6^@*WH$)Gy_pH9ggG?70EC#
zaLz+t_!~j?ob--5j@kMm2AKs${4dok;EeUKp{7*fWWSn-dUdhBYA0OPP1h@LbAFT~
zDwj*}HoFNC522wU+V6#rVr0eFP{UBF&K7L@;I9^-q#b!5%+^bvQZsYqZ7AbC9huY1
z*`a4Rer?94y@gqenORF#W#k`CDE1#sC=^71G~7M_XQ($CSmdcdgYX)mvCO;Mm_WF4
z`Jlt!7s)F5EsF1Y^qg%Q>zVm8x9VbMPS%y;Vh0aTEhz$s832p0!-1j*xM^@=aH5A-
z={hit{p%_fs<H%pc%;q>=D%N<iDjBPGH7G@<%`m2T+>NS0ZphpQ2A!TzOVm-QhN?j
zs^I5CQ=CO>lOez31cb0RxV`(XU6vs`ik?ZN5EZ5=<N(Rg;XYdFw};>P4Q>+u52)X_
zPz#^WA8Dg}sb)5z%_YV7zw=Q7cjDlppe&HnWghoY>2E44yJNAq`=05;NS2cntu^_5
zR2)=P#G1u|cCX0~`wPa0nHdiRynH!kUo1#X1wN8j*L5b}sa9A+hkj&)U%H|xhkAdl
z(q*+YH*XYS()nY=qUpy~F%J)$*TPE-=nfT%C<)v;BKbg_#&FnE71sc?<s`ehup^w=
z<C8Q}WAlfFvW{fIE7iV)*k(;3n`g4R0@}}LR&FBE2krqq^r+EwYHt>E>FCZ!f<&3j
z^|SGjLfTIYE+zlsjA8E(_?{Gu4KmD<U#dM%C<_lfy<n8E?Dh0fs{3krnP);u2n}Q0
z^OFp%SKH8CycFrf{;tR*=<gmo{45tV_;d5>T*FJtZjD{Xy^_7QeCCC;l*k>;wx7je
zCjKWi*TlaP<3m)H@s}ASK<IF@zbmb4<P*mcxtOPInzW-zhoR`3;;^7erpK<<Rb_;q
zzP&o%^WJ*9omqnli}i8Zv|eFQTbAAl9}47O^n3E!mB^+4?w-S|nu3xHBRj~+o^ll%
zl_QGJbf)mazhjYM2LKeeb>QOnw^Bd=1qTf*9BFz0*Pdol!x6#@?8UlsrCd#O!q4?E
zE<NEBV>u9Av?@EevYg4}aiAs!IChKF;<Xli6N!5EFV-OkjUU;SxX0V;>qip^u>q_H
zj;OXt9QuHs<xuDO9tS%W*h3_}yXo(GY^f5tEc<2+J)Xf}{zFNT`6A1+JOTfQwbkgK
zm6Ur~N5J;_-~3@wCFK&m8xc%}Ah@C4-_75n_JKLTh)ZvUeE~*n6E5^oH9nn9?aw$m
zP@?awELl>;ncfiaVQE5!-)>8?R7=RA9J+xPqLg4}@|<8hP_WKw-?XB4=^9X`?*D<P
zKq%*ldBLiR_<N-Y_Qi{nls8GoyzK0HZt>_SDe*h-A8+oZGLueR9dz-Olxy@VvKqm>
zWsDU`q<hj;@+KxO3m+Pye5B?`6l9icI|XC(=_+{FG@{D2{+Qv#@LL|uCD|N)e<R}v
z>Pa7##{K5f{2Y?k#EFB4lZ+otVYaqBG;d-y>stl1<tKh=2-*`b;|;^r1(tf9Qmg%9
z$F~yp3OHZsOqlIRSPr;%%VebIvp1lMy+g>l(<da)@C7{V_HU<?(FMR;Z3cJmRYGKY
z-Qq+w?O*z}QIz5X`nW%g-;w3OQy-5rJUC&7TA0aV>6PXos}S;H)(X~;`q`QRZb!KB
zuYP3Hiz=OE5OK&-ZZVKaN)6;98BWDBJ|aF!g+;8cqNnM!WfFCd+gZO9h=kI`Y|mhy
zIsS`_nnkcuRY-<)R;L5l@5bM$!o+gU>Yq<|ZXRUMXo^i_P<05sugY|A#piO1ph(Ft
z$<u?63)kX)o9%Gi^ImzY{ovT@OQ?W*47vV)XK*798Qk5!3@$ZH17<_MMk3;*kNCvh
zWWN)Aue$cd{g0<^WlLNSzf$CF^;8qDcK!O4^ayQ(pgi$xZG<gzzP;~LmPA*Z*gIPM
ztMmdDErvqOD%cKKF_xSH1Go$Wr6#rB!#MQpNlb78%#sa?PI*XR<e^Qd*(V?85K?WN
zU-}unaBA0hEh_$^cTBl};$s`&3iGX_&EixHrSg)l)?+s&eO#}F{9Mx^e|&Sf%B5R}
z{tZ5B5zC4F^*72LCPOwQ#@Amb4&`#aK+2)=mrp7td3xGa-Kr$#8m`}!D6dyqjY%^C
zCIuJ=`Ft%=ay%tyi!YteB$c`BMM|nn;?LMQbx56vy?Zd$6?TDa3a(`ykOMGIpnw_x
zzjFipK9~zis+v}6^d*|z*uyYRapZMkW?ett(!|+{9Sj({d2ZOv`{6s`L=DJe7G+sY
zZc)c}0M>WF?ExrU?nh>&xef%-cB=TbeJ)a0Y&@^$lJ#aub?X$HtgK30O|3Gql^iz#
z>{Q>vTB$LQwtIeKsd`<ll#n^F58|^}$_~kV!npHJ8bRUhmvNJV(lS5E85zolp$R$m
zgn`da90?gDPApl9RD(mmP~=Lk<gO{JE;W=XShv0BbtAaO>&FoqAwl;xd@6Yd7@_uo
z5lZpjMkuC3BNXb;2eqnq9^@9JFQMoI1(*E7nz!e}T~g<8or$vgVn-i5GZ?gns)tm<
zkCLPtCr%WXmlC?)GCenD7D=&r=pbZm|0XH;w*#%{0QM?q`>-a2f(t!|MbOsrk6Ybz
zmkwKfo><|<+9f8X+uo2L1XY7Tj16(?%p~_Z70waIA)|e(cihhJ!s+8RlOF<WG%dEo
z<9;_;2b_QJ>CAtONG#{dDW9vEI5xBeA|lUSd+^~}*>mwAzu3OeQ0z;VZ)h!&4N+8s
z^!lUa*uqg9?_h(4xuuY)@c1F#J^jrs#E+Leh5mxEiQh}dthNF^@I38rQ4$Yf!YM<W
zQ1i{<L6b!P`SvXuz{?sVaPh|-1G#?1C1p7!@0j_^U2cd@5LUd?{M`gko9F}afEa;9
zpSeWjagvKj+&a!bpi^JqM`w1X2JA>>z~vi+$wu_CF{w<Kqpv><i;s1bxm$Za6q$WP
zdG-{t04Yo?&y1bi&^c^1TMuKlcp;m%2lq%qGO*!5VZZ`S2%<J5{UU-?owATO2_P=m
z;+d}9$`1BaS%TE8(yYL<dm(uPAxwcT#>0prMRKIsI^{WLC-D~4Cy<sbKLb#d3<JRP
zZ~s6@l`F-l6AP8_h@)Y7PNhYT2fT8QcacvD((6MVyq>wI&6QmMg*OU!4XL1pl!oHm
z0aO`Y#t^8{A4%TYrF~oaN?(f)JV*Iun7Kqn_7inQlG;Vjq5FDi?w4FXar3T0#kT<$
ze8rFQm}UuC#{AFRLQy{eH<mzRB@E1~b4O(E`vOWh8qkqf`y?(&p2Obgle8`VR;tN=
ztK}jG>h4%lJ2+zgz~4E<H}5v4Y6pAj$Me)ldEAAiNY-|(%BXQ_X?Gf4O$hDQlC!k;
zDO&|ptjLpPDh50<y`N}X;W$|ec#DOrSX2+C8rvVp7upc0Kyq?$%`-6{^D){1L(#Ka
zOYYxW9gWA?+b=O+mD22y->X0bQacbESIElS6`<aD>v*J<d>$2^xg`$hl;FF6+1LP;
zpQAUwm3{vfa@k+sXXyNg7@NZ>)&XLEW)SgIQUu;oxK5ovz`@G*`nLod9o(%xy%{#n
ze}O`N-bLqsO?9mNcbI!G4KQ%spNik7_elaQJRnf;gTKj!#FU^}65?plbYxrsu1lu{
z@Q1Q?Z!2T7UXN%U6G@0+1>ywHEqF6(Pd+{ZzP$n8I`#>I;Gy=6Mv!m$P0fF9fp2{T
z^WQR~Y6a^Ft%g-Z`P4^&{IwggVFkq|4t4a%TH}X<%Rh+;lh7%#ePKL;{I0BFa+eEm
z(6?xD#Jm;iIYUy`cJ5w8C;8L-XPBchEEt`Q1UqKK#4$p6O2LG_L@C!ocmtnOqiWl|
zllRylG&-sYi`XL=6&AC7;hj+MoC*`7sr4P8NI#v$Vi3LaiE#O6<cN1<wvqLTl42!<
z`wQLHUfxBjJgE!^M*Li=isv0kchDb*vM|wn@-8Q37TS5W!Yimy)|;7+f1Z=T6hM2R
zg=)V*k^zs1J2<SQ;}J>(P%t`d)t0)FG(97Kw&fug=d`NhOapsIU;RUdZnP|&SQhuM
z(miGF964l?V-@pNmSH~>k^DTBB1Lircg}JHX6N`ni1_T&Nsd6#1sIdd<6LamSl@)4
zgu+fjep|g;MeW1e!3mQ!-trsXtL>oeKm`?1p)sOc46+8)jgJgzEi4=9lFcaJ+i<*2
z1=GXaqU#F7x^>HSq9ZkTn=?)lCgX~ib-<jl`p@QC6CU)@7cAo5zX}aeiJ1ZpS8@+}
z01m|JP@p$M`+$O;MuXofonVZcOGml-I^o{gmj^c`{J1F|R5v5=bq3aC#I*{YhDVs)
zuYK)6WYK+R{5D$Og}&G~`D_bMnd^-ULIBNbOs%%<2qXBZqlgVvN7xQCd|&<(=zD~z
zx)V-~Hko-0)v^w39YoF<#jUL>HlLtv{I)q7<iD!4j4;Q}TB8S5{_FFN>xmA}RpHYM
zF4M$wuU$UnR-KCOzWYkVdMeJhjvSHuU=e;?l_}?u*GPCeE}y|wGEJAU$~5ZYLZb*k
zgFC8NxBqAg+yf;Dis99kXR9-Wf33RC1aZdBLd95aztk2S%*;Yc5{x6ZP2?qEjYCFd
zy>_j#A$n{w35mY-g?;~t6q_aA2Rm~^t$1zJW|PX<-0cC~*{|P^a7*d7X!7ugZpw23
zcLfRFsFF2MEX^`(Iw|(b^p~zZv{2SSR;g`8XWi0;gXSlodFG#!jRR9APWP@#7yhDq
zU`oSO%%;&MtaeiyhY6>$5?XR8@3R&Zyo{?DAK2Y7xv+Fpw-sjz^270&Cozry*@`#q
z@(Ofi<obI)hP*9ExRTA9j*(+bW9^S~0@CO}xQ|1tx@g3e;vWN#(V|8d-@UT@evwD)
zcu{?wXoeE{73zOUKi`8#rSEHC7i`0vEDbbZ=Rm#1b_=>7e;6%u2w!B#_eo)dC~5ws
zr_K|s?@R|JCs%eA-L8FMe6jIb*;AxYVP2qe2tX!7ex%%>sv(#jpxF4#(X5zYcd3Z6
zGrHrvbytivn^HH_;`rX(>27T>G$Ryr`q<w!0)}d$3N<W8)i(lll+`ZmP)Rp&2?#-u
z1w@u<SbaZU?Y<}CsKuj{MVYwi4emiBQ?Br-%Cr!fG%#-i;{Z}%m4`gj^;+u%VM*5p
zZk#baWzw>keK$0maqLJVm@M5|3s|BAd=uhdUE~6*qW`UnEUN-J@Hz~f6g45;opQ0g
z3*-pWOeTi@&*pUP#WcOtbn9tFKdbe~gHT_n52``NLK{~gHfo?B&ZL-_C#B_<=!+MH
z-CZ`{x&;kd!jy=Hu9WaaI|p1gu89&oU&eZd?iI_lWhb+Gmaz*VbDA1V6w_b0^r>)F
zT9QMDV0penyKdrmqD)eZgZ$po9-}6=aE#y}2qwD6LOw@5tu*Ci=Rh6Fh4?*ue{k1F
zePrL~bie1jPub7%Rh_s^A!{^G-L{WZ^#Eq*RI&7nhx=-Rd{yS}RORYDf@H4#3cjNy
z*J^>CM-h$tCW`>K6_wYNSN0n~>)r+0<&mhRX>oniLMeWci|-Qv6>&T`5ha%V9SI=D
znT!%}Ilz{7=m27^0K|ka&;YdNIN#MOyl;0e1_~iOwP4Mt#%7TfNfo$Bfd!S?&$c`7
z&TqZmloqZuOg+Ez1G@K&;WR7{E^Y97Z(`Di;=B+)9AHAeOsWbp@F=E@%WJR9I<^gq
z1aWC1nDQhl=WJ(Hj_sU?W7M$7jP#p}EFTZ~z`QzX4w7_}6@PVTou5;mP4OG~uB6tI
z*J%QUIwA^F&mBa_<@hKvS(Tr6KYxcCDyX2O1b@H-3W_bjP2L}LT`ew<+PsZP`uU?c
zZ0%eU-RiYFolljgXpQn;kxG&XWi2$CA`82rwA6Anh_3{u1o8a!?{l$*N_0W>XLxpJ
zmwS;gqIskGMS^<%GzYt8-f!{fg(n|6w6fUSco-SBx&nPXfY%m^YOg_p^HnGbo~D}+
zCXk;I4)A=mrw?lC0_b`=kJs|}?AGCN0TVxP9d~mUR7~C7A3tckGU+qJxF#wz!#-dP
z^l*lUy2weOiv$plVS%<ce$^bO;nTa_03VMAM?>opcAGPqoDt0Hs`mV4C!M^mC(eq*
z?<OdXZdf~rtlzN4&vEVzi7&)Qpf()U)j)-iKs73+SqR-JAmjzGfR1sk4QR11|JpK)
zR_7fI<QpdqK1>ym>U>oZ2cWYrKSoNpZYt4ICN%zyI~JDixg{F%3xkUzdv0WSzG=gM
znsn${aHQ>ngFd(@ApCB+Q}_OUFk6x@{-$_qEl-4b?+D^Y759X4igUA2%k3(Mgx-X%
zKu6=z$)YS_+}V1en587v^$l$e!BrD0MWR9aR1f2Xb84fGq~t4mTg%QB15S-?k>DQA
zDl1LUf^26Wf-jd~sIg{$dB%JsIGY-3;B6xg&xa*p2QaAy1To9jBNLAMF)sY2yI?dc
z?sdc`;ZIv_@_Pjc-@tYZ2>ZnBlgh!u79OOC!Re7fH~zg*<0gtKR#l_Wg1yndy7E{k
zZlZ5OKI(in_q|&eAuzon(17hkY7uLbrdXKNOA2{zs6S}rxK%ixeC62jnjc}d?X(cA
zp%v}Hqa%T*oa|^@ez2-fmS+TsqZ*A!hSZ?cb;>3HOk9>d@krFv=UVpb(*n<q4yr1D
zeibKf%WUQ#lKw`iktw!YOJREvkqHufnodpvuRd*^>LXuOuRjpIEPAjNM-$-G7hUU=
zjZCsv?T3;QLTkJ#%R{!GBfE`dFSD<nXe>H{9;fjcE#t<26yLlnzkl`sj%w#2<Wc-k
zD_w_8^Dt>J3NP#?{0~GBShAPp*O?Y7snL`KFn9)QW<6Ru-haF*aBVDL%mdHwW=6^e
zVx$)Y1a9P_J)t06v#lFX?()nyLzsx2Ruevuw&*WFdl2p?^RS1B_ZBMOjfuO=roxMD
z=FVbKi<s(ZY_r#8Zz$zU2xSbnNmwyp!QTs1Bjwd+@DsXn1XcPC@?AX@T)rk=?CgBg
z5u$UVSc#|Y;P?EwnyZ|L^IKR_rm{n0u@&&$>7a{W1BJ;f`lH_9EBJbz8s^xGXH!9|
z>Fq@b`B~4^R`st74Lp`aKc3lidGf&Y!&8Q_7@E7GJwj3B5-<t?gF};(FegAq1`AND
zB*eoLR5``asA_f`Mf7GFaUyP{uKabz$e=J|qhq6zNDRvx=IXf`hZ<z&ft&Qc_j5k#
z+i{j5F-T8epa_{Y`u>juRT}q4?Xnlsr;SY_h3AAjm1zqvIxlzl+LnhXzqcmpydR8s
zs(lnA!4&u&M1gClxKjMJ0gVv_>HP}lBx!39U5ZcoB<Yn72nR<u&r6>NJr({KVnQ!~
zcVnQEX38a>BS^jeG>Z7`^Qb|*%7Lv}i0+k5Z;*kkc*Q>wy~E1>zxA_+#_zTspqZEY
zNv`$kW=p45OV_9-jgjvP+Xl-;2P1{Q@ysB%MVlvKKiEMd?Mb1md(+HuyldfnZB2Pq
z_3Mun<q5gT+R9QZ&vbW1l5X9Cx`I(j&aD8fDmueRCmq3cEWra&2wzD&nL)nkwuS(P
zEQyAJ?P&ZfYDzPkdI}p4wwxzJaDot)^*RJR1BB9Q6et9}ia>>RG@fK`1@7YaRVDyz
zgjxW?U=(1n5#k>Rn1s|yrQ?O;TmUqthOgQJI}wu`IH&)QpI|nxMLJTUOr{4GEwA9A
ze8g0>8Ghed6IdC|57fy{@zWyDoKyzDV-YtYcY*Z+>zEb;Uy5=gV@#p%l2RM&gt0$?
zdCnOT6UgrG%d6--7#p=~L~jx(lt2QbSPBl6pxDUj6F67a$rIz=8vW?C1_tslj&W4$
zsT?U4!3;b^RK6PL8mMOfSE;wEKO|0ux#X3ijxYX)$7Oj_D~7y7vT?=78eokGdd`($
zIpHJI#P`=bUu=9QkEQ3>s*kyaf`%&%m?5b?RDIhm*T=TuGj7#jPN1!WEs$b-Hs#q?
z-FybKnnur002K!Nr}JX&55D>Xd2IQTq4lVP5cU2DwLP+nn1AE6?g`hYy0lpvZSCY{
zNC5sYno{&DQkmdY^7-bcN430tYv_kU?U%kC>2`}rA{j-5R}167Wq#WeQS2|N)DwSc
zwCQB6io<3B=yK0bvIY*Q^YxW<jGhR9-L8omxq7nP=pGwu{b}=;av3F9@@Z242eSUa
zy2mn~copBWW8kmZo^TeysoYR$TKZSk`@{?|<}d1XQq#!ya<PWBTbkTL>+-9aMS5jq
zNhZ^f^ib`Wr&=Lz-`7!CW;F!g6$B1U@DvR`DR{gUptr+dcRuBK2X_kpljCE-+w+T-
z=muEKKCG$gU|hXrXsPjbX>`wqC|`BkcaCrN_>i9$eP?V{@l!6kJn|=Ad5SHIhvZm?
zmJ_Wu4>(%pr@ab<v}6k7&m@EQ!`1cFvdlQrE=d~VGDAxA_tr~T7@dr{+>3neeP|qP
z2QvMz1Z?;vIHP$NG@>|PP36KH)o9m~H)j}+xDAy8ym|Q+$!026Yu~wJ^2DN?MQ8I&
zLIGByHPNufc|l#U7b$G>ELHHW(?_!+xf_zxBiz9Qc=VCdFB#uYIC0e-(GAgn{DFA%
zLVK{764wZu>eS0F&^zV9{9kT}T%V9@wBC@z9<31ud_i-p@U#bS#EwfrSNdQALg5%8
z(5QQeQE4Vb1{+1*@D!UhIoTM(avO1r+88zIO7wTRjLf%5b4vr`!4BLBKk8D9WydUh
z?rQtyK3i)U_I>coaNZIcfmF*TJ!vnv#H4^VEPW_tTNWcE;h*MRpm&}&x}0P@rT=0k
z)%Dmymqd)^yOoZ1)zsf5-+PfPxT<!ERmtK4!ER-#`BTJ`U%phO?|c$TleRYGtgyf|
zUvpCD_Nfsz@^%p<-oIy;XpMxNr)@b$$*)%ZUfGIUy^UwQQj{oD?xoZN#tI*!o&{?P
zd$D;(y6lN)Ul$8)cbBJGugBB#ghLphi=hd<ncayDO2Tx97MJ~teX3@&290XvOmW9f
zoYP){t3Y-h0#^Nkhy&`F<4Wf<!{L+yk5sAPoCQCVever7k?y)vaEI`VpS(TW)>Hd(
zvzTF2PQ@V9!tGyRVd6K|2Pn=cAtzQ^zkjvBV~RT<^?XC?X242Zgz{YR!rxa*2f@Ls
zF$4kUynYWg=^MYWdi`4m<^h+)g-YWt3XhCyO++XwZx`^Mrws?Iqy7|ZpTh%1hH|2s
z$^r!lT#W~d>k_y!!gdTF?+~}_SN>Jj=d@7wi~Ph*yO*-gU(gbiPpmT}o@x77(|4y<
zuEY?e9*9sO>>7eB;F(0=U=+6BnDxl~#rWsyqP?p3r<}!mj++HU7(^)if!J>B1f&lf
zxf{na0TnJhY=LhF)Ew|;OE;K@{L~0Ac#sp9ZD%V?qzmMrSsUyi-@q2TGWm$!8$jy2
z?i5JcZp@Z#?Cy<r_r6(dSDKGn5^qC2nsFeBetk^0nmir{b(Q~DNV_u`2o^qA!YD*e
z|Jjsg*zrXK)TP?o#znF=n&Sb3M0T{ZM9?ZLP46>?@&1}XB=%RzZ6!+d>M<?+oiU}=
z`7*d3WEF+`>H*t7wGBU@2>7Fe$+vCr(kD!uM-1QYTrN3#ExP3O>!|OU*%x>g+z@X)
zcPCfMzh=@RRRr|X#`sy|n3|pkzM@&$ZX!P*YDZLGnDWoZoH9GHs*W2Pe%$F#{dJl?
z#Jvu8!Vb?D@RY^HivGYUyR4x=KEdSq(3yuq4vrfSdEhT-iT@fcTtJAYzIcq-?F2?6
zIa1ivCv|4iW$cL0fU5E*y0VgQ3ZzZGxh8*|hSiPHj7(8FVkR(qV<K<szz_E&dv2G3
zMnM(Yov*Mw;Y|(@IBy1aQ}{wMU=fkm5oSKxCjEf0WL&KL7}4_9{dn<-***VzZe^~I
zZTa0=`HklUowu}i0*nkp)j1fCXMT5}v(1;uV!2W%I^EjjB8@8zWzPQ6sMgUV{hGd7
zc+P!b!w16bT#9~&aVZgPQ0KQkc4TkEt$)`8qJ3cEgRG1otbO{TLtp*78(Jv7V99m2
zWT@C*;I5(j+4PlTZYN%gnF?IJP(SnZdZ0OW5#Atg2WDK*dx~}Bn1wLZLw}k<mYqA*
z)qQ|4h=Z`2rY_T`^}Xx|yQ}pBM(&S_>gKbSRS^~Atvs!H*baD^m~t7xF4&}{z$dq;
zPULEEW*t>4=;GsA39?As#xXd~q?LJ?v${Fz+=b8qX3`egkCZ|T)g(wqc(@puta;K8
zadm!=<2-K{|ETDh2~!r-fCa8-zzkQ*Ux**^;<<~%HrymuuJjAQ#I%MsTzu3cLiczf
z>n@l_fD@@{(uGHC>wI`fGs7HP<lf%m+nVd2oS(g*F89j?Rq|NUQ%(I=*~u69F=dNz
zkf<fUmw?E#I&Nfmbm9Ec{p1D+*nIY!HMmPb&Igl&CsP+X%p-C>tSO;#wD5j#8$8_}
zi6E8bAZ5)NwojP98?EZ<e5n~+0yUEGSq_e?aqbhJL5%wHQQ4#!)tP&~|CEc9tG)$A
z<WWH#ma)T7xOG!+Xnah(r^Z5y#i(uuKbpE;D}}WErOv0GNxxb!NC9e-d`5`W?+U&S
zL1kg?Wo#a*r}<7NCm1OYQ9f52K`H)~(mcWYAm>8ALTKDEHif^7+!mAEPClWl`@JjE
zTI^?2a~+&r`9o7W4G>dse<pafD}WvG`ctg%7p>g%IL>=^NeklJdO^oQ=E;Ba2QGf0
zH}9?3Ik$?SF4<f_uEu{tO(H;IO;!dYNqj1kc|l=^i<de5i$#2@bJeK*>lq|OXFNT>
z^|%u3j_x>ED;=g>_u0t+4u+60q|{}A@Snj6U&)}}xq<?<AprquFLrG(EQ=P^n+XsK
znX4IS_OO=M|NnTMl?!owYCY^pEo>>ZFgADCaU?Fu<?M*A10>e|o@HlTe$X>0^b<T1
z?OmvihmsQy&7{B2EQ3M?^gLa#t^^>U1UjF;ewjkvs?1=_KauX-AksZB7Bofw?eC|V
z$~*~ox(REjmB9Rx^e`{b`Q2un>UJ><`KwVZB<Cdy4_<`d&N^NC`6O&BK2q?4C4F&z
zy3~&2@M?71S<|SH&I*XjWlb=;MNqR2E`oH3(gB9;8hb4*mE1;1Bb2@!x}Ztt&$|fb
zLpj<d;p2jcEo((*J1XvroS|e~Z{OmvR?oQw9W`qWAxjq^<mZZT<7f6)gN#z=x|R8Z
zdrt|=tB2wOXdC-(o(ui3Sn1q(q-`j<o%N{-Z876^0g&d+Px4ax{eXD-_GibPO3tAy
ziHFYKnLVsWJ(a9EJ}ZG?>PTZv0t#Q6LWDK3FEr|v_e&eGU%EWY_BAU2_ws0;EQ^}&
z);!)dDxn&|jx_l$ER1rtRN%Ir>TH5Fopc79wJm9!dQtd`+!>Q@)1WqsRa*5|G`nXH
z!TH4JDZ^ur$~Gn+1@ioIl|HAO=)02b?H*&7DC=_s?WoTG3#rEcYeB^z^P@=9rD%sI
z2~gqpmo>Va0WW%%RE2kK^d;+msv;{D4RPPNH#v6ocuhUO>*<$i{BB8UwnBo?+`yhp
zR`TiTs)D<^01-|R##bDrGA0;aiV#CMYG-cA=DxE21Mva4bDGKE!s&sv>9@SfcZ`j-
zxy7ymjZv(B&&pknh-s^$i?N-$IXW>@$gH>Th1i$8GTZin37UfqBC^GC*ZvRW9w?JE
zU<ym960cZSlpAM!gtSzF?aQz3C2cWnS%gu&G>6_a?0Y1P)SipfiMY@DyXWJ$)6cVs
z7x+4vSYgPEv@bR3Z2NiUVKf$6M7>`>V-hcn{A{FS9RTAMuLZWU>R>Z2*~k7&za(_p
zXFM6dtZ_;{Gd7dyiM;=Ce8Z~B0kU)QQSy1(uAWJ2TM>5b=Vhz*19j4Fqu|AUPkjWV
zkDoM)T3WQo(94!DG5pp_Vi3_;p%{l?1YKqEGud*nrKezd-8UY9ecOQVL`2)2pD>LY
zzG@pYXCSw%0W&$g<!@;Sg~&32g5-i?V?14tc!%P14`{;gAD5koY)#uDs|DkHI{n4^
zoZo+~xIC5^NH&rDVA9L>Oy@qt-T2dgbTQvm5!2$ZV<hKWSb1Ba@89D)UtX(9uYicf
z-_fBJVfGbrH`@lC^y0^^()9=q3CHdR?%!cNN^I{%7a*d9EEaKBfDBQcd7#G>oDbF%
zW1wW=Qk5?xiwqzgJn7m$j)a%h{78h$Jja}8^%xngZmM65;42DmogWNfn#$LYpu|?Y
zmL4;p+<$oE^d3kyc4xVzP-^Is*B2xJP+J}jO<|K1rc&y`oTB`RYB{2Q!!qHKA=l!N
ztjQwpGhgYv=970Bv%W;WlI{rUeq+<Z%^s&grcBqy5Spu7cVvjSP3$gjb0$2^o4YMv
zYvLDN@;x9v7vdRvV7iE>+UujqObF_k8d*dXm46_RX^hmp@zcqp<tp6{TMK`2&Wem+
z6PtEdKU9=IioB1`E?zhXFM7SVku!gXjpfxw?KAt@bl)rn49*h?K;(z(IjG5e6jc#%
z1Js_ESOk%(MV0+qzwc^QJg*yKYfvq3NGxEra1mn={>XstK7P=#b2}DKlFsps?5H#k
z=L?r*EC}7oV(y`G<^^LG3<$f_6Yl{<>HMj*EWB?*+}|;g`^>F`R^G5fR*p97vlA0U
z(s<ypi@f-SuiJc%Qmf}Izh70g`PKUVZhbzuH_iegyT5L7OVr2z4}`uKbY<nQ;<=(O
zvCxr@nrEJLN?1C@R%v!UZ_hfXThsda$$JpiI2|g?;`q9F9meJS3@Xgb{fwNLGppQ7
zSq@)0#!Ssb==9%)w+H#~`)cQa89`Mjs{A2RGyg_}g?o%6v>7iQlpE>XPqMk)?<5e}
zqJ)o=72jmWT%$s?h)7<-ho3RAI~SjV<@8M`DTE2)M}KF?BOPx3&0pX_%gBCvn<Rbb
zm1_@}weAt?IlaI@)Fd1_#MaZ1ER5T>SEh6Nb%bTs(h&lI)Ezcud{M2FAQ8MigReNc
znOD_pe&>0PW?l#HclFLw&rCS^yrpYRY%2ucBP0XIUbYQUncis4iTF1;Kvo&Ai5v+q
z7;5bWYX{th_ybqXHM9XZlrQ}*4LSvL%}iCQKFVX+9X!IHW&*AU7|Un8n5G>0wJx!;
zZ7H>+v}R-ei7s;K=kZc3KR@ioCgKf=bKizAeL;xikMsGc6?d<1_5QDx;|a$@7*-sQ
zl$14Q(vMnPSHnEt;?8>!+Iq(18V@&r2!j=a{9tf!a2+7C#1$=jJ>pNAYr7UMdYQdq
zt=)=SqknzBC?=?GYU(i2NjVxGL1iwg3Sa7JsN}5aln_*Y@a=^uw}((zn|ys>FN#(5
z0**D;Y2qwlOkKDl?Oo(*-izZ`LT4u{24a%xAYRI?JUe=%-JLOQK^ekUn2E57Y<Mf(
z3uY}8>s}P31c!-(F<E0*?7`$$uf$VeB{dA^HeKC2n3Aw2V&ej2eb9yC8<pHS8Fo<|
zPG=KBJ_vQls6rr1&f8l5Rfu#5zu;m^=6`1vZc{xu^+96~b_-DWnC>9>-7kCx@!w4>
z##WhTTKcTh93ZKwpM0XJ*C~e78t^50nIM5kQuOqEoP5E9=`B*Gaj_})N=V=8RGfUz
zaZUz)zYN2}gvzj3eEdE?67ZCvx`*PAnZ*H`K~zICv6aHCF8l{l4{mp|C5%vnbOReq
zd?Y|tP5^XqSrl*_zOLo(>%cSj_jUKZGwIYNd}@29wIb|4Ij6?u+kBq5YJZ7qSNh@+
zZ3Z6w#Csvbc%Bv}pKj;Rq}w2$<$q4s$(3&-gS{~EHqcYifgs_Jg_RYs7}$oMDkz(*
zqtO(O0h+xZmOXWod*W19(A$VXbeBI`3p@&B;0QuI<@p1u2hG0XA!8LnzC+2DAhYB7
z{xA04GAydE-ya^Nq(i!fMq0XCS|pT?K|)GeI))Gsq!Cc0ly2z|>27I<W<WZI7>1hP
ze|z8Ood0>=J#WtQyg4tpE-twC%-(CQz1BBAQT1T5_iR7a4Exd8La3tY!-+Y1dgcdG
z@Kk>|r*4!&gKB1CJH&nCwhG5mAgIW_en_DG(m~IpjGo$kghckDCacqNs3-=m8#`9p
z=AL8e$4&onpoMMLFzI&|w>@fQqkzI~Q{wY-d0Jy|CaruVyZ4DASUFhpzOYi3W%G6g
zf3Wx5`SsBCi-`q@1Cu*~Cm|tOawpMy0K|FHqYEP+99-r9(0e7#&}KN-BV#VvdOoja
z$Tt$Ix^z~?d_KeAnU@(VtHQ286KqnihN$iE!<daQ{H}CEs)Rc=zi`+{Sa;`ndfrdp
z_Ow}$jVF~Xbdq)JhOjRFCRt_*w>sDoHM9PMgkjBT0ZP)aMsBePt{V*rPT_So@MArW
z9X6Dr9omtk>6QUwbO`vX_sI){kZ7G6CPkf`uyJzpr3yFGV+t8_(~*VY(GDF<!4*V;
zog>VlAFKp5H~q8sGl?l@%XlVCb>-eOM=q1P6yK*?XSpZRJ?f`OOdj-RuPoO{2#&EF
z3nCgT;A+ME2?{v}CI^SOYJHO6K%yp~0_CpS+3ta<>tjn%-9<=P&n@Ple7YI~MC%H-
zxN0+RCl%bkAbDTLj4kE-XbR7^gUc5;I{w!|PLe9?fLQ9PN9F0iIvv!UpVj{EEKxzO
z>DhqM5=lTR3Fnf;WPZz(ek*&Yb2|9p7-;RQoc=6{c4vh&esNKK!LW;87Bcc);=#xh
zb>%0zlBi7MIP(y#GlKW4+}}R`g8T3cbpsDJ#0;2@-^U*PagQ|{@9Af0PSWsS)hltX
z*C8|H%4g>#l{u0JguEyJ(oC{DPPyK>pK$)@IBkmb_7R~@b!tFat|WQ&d+9)?FLJ=W
zX2E-fPEBU^pEv{9>YGhIW_$JP5oTjdT46gI>R6gye3?@;&B6s?q&}Zf1hXDM8Sn5s
zE3L*l@h&VCHAR|X$41zx!D;8;p9o=H%I5xbfhjoTy$=N=sx!-Aj!X=Zy{V=eb=nSZ
z%C?_)hOL)B9)>Bwzi%ycS0_B8NHJy9B&>V_h@w$m;vULTL|paP2e@y2I!czGX=t*O
zoUCCDldma<6<yF-efER;Z}Ypp3fQ*BgT)nnTKDqDshlV2duZQ!3XWxNPT0<}6yGKn
zH5fazu?AtLY91+q<yO%1^LswRFn?)N*WMKR(bD%TnX**+kFX_=(vn`x|MXsO?l&ii
zqcglf&KvAq%?ng`zd8AXP4dH*cioP$YLQe4$#p&!AIg{jRNslr$C_3}S{w_h&g8P4
zxL-7We%k+$>0Q`}#!*XP9Ox$Fzhj-hu=2ehW{p;(VUCXk2-H$4r|@ouo$gio;eSA-
zcHlc5h5ojtgmo~c)xLyLELBoT5A|WwR9K-9;9Eir;t0E-{ae5q1q1~?1XKeyo}c+Q
zQ;qSkqHGQNc$U@178xTVG!HQVn)1n=AQ(W%{GvTQ*}(-=Hy36AqLyI+(AyPXXj9A0
z@Wn5lu#_E;ZJ~Yn`UwbsI6^2)_~8&PbEwy5$NGKZs|>++UO<Gz4*NJsFPzNiJJ8Bh
zz|$+ApE!%L2RW^72lh2(KcE(PY7Bh(BlCk2lj2XJv-M4ulB!z-`sDB5`45bL8d&y>
zx21FjP|nZ|3x78ByEP!~U9ToVbG8A(yRQ73&5d!AuA52>o@!qWDF{KJb&$-<hJX1?
z{mZh)0zA-3vx~+*<-0;N&pa|}YDsrB7*I9RTnV?XY9Yk|A|g3y;{YqXTSpc4PW-pX
zL<ye{#!*{5h3_Do_J4Cek$v?lK*RM;kmTQkvGec2m<km}^?yToZ)ALi6S&BH84c04
zF*VCu5fK{Z8N?EJ(99;AC;}*N&V{fcUQcf;*eAtARsDhRymc5Mt2O0->E=t(Tn^<E
zb>CWWZscKqXPiWq0CQ4ff(l0@X`BKgE?oWMNI&X>3M_G>zH8NR`b1>+Qq-hvTKG)L
zHPU5|B#|nU3_Y<%7(B$G%OlzMB{|_;Z<qY%e4@pFKrZj&<JeDaD+B01`#jNL9AeU&
z{KGrf!G32qElax|Lf`k~_p-CXl2Bn2)%vyhY<aCfALAz@#@GqmGB7~v_B$eR@+Rpf
ziIVqe|8*_yMDb%PiGPf2!+h$ls`~;+DW{Zf&3k6lxK7{1l36pRy&&xN9tQOZht$b`
z`w4~%_Jau><*|ObH!@NwkeRy-rTsQpR>Sfdd>3-^(?kICD<;-fl9xPzRXX71!e#5{
zeww#ImOH=6;#%~cbBz|lqk`Q*R)1x_6-uJOeIW{R>SS!k!zpkPLaBU(z27!IUnixt
z>BjXUVe#rMy^dikJ%B*=jFC6RnGU6<QbFb|2L7KS2-r&PC1n!<zK7YpC@P4sX%*vi
zjGC<dg=(=md&*FeDb3UT2omN<SeP{axk5r7Rb>Y?%xS<<+*0f2{>w<X{n<xWxHmOy
z<+mcu4(z9JhyH!b*UTd;vTQZ9hW1BcR~eN@w3zZOxJk~S&R4gl=Gr0dj92eCnj4%b
z?LUiQu&KQ}>z4i65)SY(53JG=X>ftO)71q{?h?)Nm3`(>2381%5wrZYP1S@3$aWzm
zF%(nrOrtbBXL0>8E&C1rjR>-{IfuTRT$3;^QrB@*I2rRntHsT&9`1|c(0Qt~eQ|`d
z{G-(Q@nydH!P0;#x-08ky_KYAaYqV{kl0Y#(|9_goNE8qOMNpd&?0r0w`JQ;!zjt^
ze|6Y+fe2=NHP*pV@I=i&8ei2@)vZh8giX@IK*uh-HA`A}kNT}>(!cp6G)#w&Bopw(
zFoX45?1GZaKzTiYru7}J%O(LHWWzV9kaa`^>iE5(30{0jfS;CRLoBjT6as~0h@Wcv
zcvAjs&=o!Vo+t?iw$A4uke8-h#ijKBpONH&LLcNza36YkAczigQLevJ1pq*3;Jve|
z1G=#|VG4pWQ;sJ}`B=aS2db{wP`^tT1Y)km*OLHwqsS|L?(Q)VwH2+e-4gx=F1W}i
zna<h@eF7bR40w=52y|Y&Ft)Tv8X&pmU=jkNHq`}moHOX|cl8dSRh<ry)qsyu6Wp;7
z^8-qCTQw+kbdL}eAcD~UJy5xfPUoX4m-%i6J`_VH%E|278WW8OQauPBc&7Sa6-~=Z
zeBG^=g^4E=@108@`ZYEta@tRLXoaU_&3y}#w?}m%e)PbM6Kmf7m;{;lN5gd=z_dr_
z6HK14f3*B11ru3dOZ7i~E01$$cRKn0&M}KBZ~cQH(W=$J>dSe1`tip_BTb7_;Zf{9
zH97KfKre3qOv#7PBnuiyQ7N+=@lYW|vL(B?BTbsKAc<CBK;^`L!~8Rcy}?r#WB>(~
zT=XCPwB_*CF!{~7JHdPx3106?$3VmTo>C?Lxk9fNJxxqLyg{y{z)ZDMbi&t$64hzZ
zBgye_mWCz`Ke*(?lt_GTM`H<qhC70<!Q5tg<ppLh%)x}CrA6<?T{-$PJJjry4Z9%x
zEm&mP=X>UPxE%_ncPa=KM`JpQHp8VpyXIHk$aiUu9O2j0PJ?XK0eb_XTBvS!v6_B&
z8jdkx<)aEq4JuV0t%k~xMY*_SrQ?TUKquHkM{vLek2Ov`k=|)l&5)h>@k`)YG#IXc
z_~7X8D)>7!f0D@2bG})xe}>qW_#2onnLC+rv}6FWxRdzl^GW{iBH7+EjM4z?O&cCk
z(|2TbtL;js!*c^8X(+!hnABt=EPsAf5kXKxa{?hlwc%9-8v@?v6v<<mUKN9^>Gm>{
zSWM|howZLNN%U3OY}h0(6(=Ja!k+_pCGv7zMGYh7Ag~q)hVVPWSl_$T425qLXI5fx
zc83K1;;TQx@FRWYC2eh=1bVwG$At_65SM^??y9qyc+!Rol*m#Q3>%~RLHL51Co1-(
z?k6h!2x;3q-B_QXfGV+|rJxqOh)wPs_Bhr52m#rqf1V(@n{qnXTqLv!i(zuyn*kAw
zaYv%6FPs0~iXSY52o&;JiTQ!9lBzi5QuhJ~6^M&&&CGhCvVqWdcCC%ZtlWjO!Nn;-
zLJ;nzof*!b)az>l9~5HAoJTp+1k!y<O0p5x=)eZ6uT3>llQb7QGwY+DIGS4g1=#$H
z6XImufW4XguZ!~EUdQ4wY*RarZj^z%!QnX>?n`!R_T<(}WImo)9t^kV(f&eeP0u(d
zN_L`We`C{>bNKd<kSGGk_+}Ssg7Hu#`TjME3&c4ayTOuedbiJzO-p9NQJSmx--6dz
zux!8o=^)RNh&GP1bKF+r8S2H>*nfEFvOD>kXX1}5=BA8_yXXgj-0uVWc&oP@*SHm4
zk_SWi#L*|nrf8Y6)P?REKKw(r0I?_Wqq%tcWZi5{&*gLi{d$t(Z~T&i#cKBJoo1Yp
zzNzn6_WlWgkMbPSnmcc8O28R^9;~aWsoVJgrNEc*q#?Y!C%>Hsj^P_S{qC&u0iiP`
zg`{udo4+u}!M7|z{#<8<Pmal{f<z4N^JQ7QOD&&=KH9xf(=@$1h-D$yPJO<^q;#G1
zVtDIP&Se?uaPmG&F!6#H_0|$!4*TLag|_X(7v!-Ll=eL3#J)|T`VlghHDTG5>_B2g
zQv+1>UX7Z+Yd|E-bn`4}AyWp0l*xVr;irVMqW#dn!_UOO;b(oYcIA?PQ14A*Z7{od
z(-XPA2%!v~0(Oi8GsA&9`)HS>C@bD*7&ph~i@lygXVXuIvLNXx+;_vb7Cf~C>)+IE
zd@P_xbe9mU;Mddbnx`{Hh^YRk@sEM}_CFep$iFi`cpy&gt0Jf;lsC|+Fdo8k8ai+c
z`3>EqiJSKm(K@1996o<g7!-FC9+~~NJD;fc)*OzutCoW{LT_~f^P)OL_N3Q(s#FHf
z7cF4jB{Hf$&Xw%1n3tlAx(G==WhYDD8h5dib+0e{S|ffV$bO&P?#0&7;+t__9M5N$
znV{OV!{Ak5yLCaY`!%1a2IZ1>lL|en0_<#D=0p_|=s0Umlk^ck`xC~*=lNWZD^ba@
ziCv`PAnV}1>+jM)2kn#zz#?KfWlo0<T8ZP?dkwg1y4mzS3m!^2@*8URs@U!@@emge
zqyr_PsZZW1>?Z9hTp_`Lm$wgqvhKep){T6Gnpo4yn&{Qhe{&OTa=#q*jn7dL)6Ehj
zNrRg|z`O2zBhv#5frk!sa1%IrJZs+c?@H+KpLYqFkla=q9NrbrQ&FUco-Tkd+5Z7;
z10b1(g$i?Eu>YXs1e2cUd8Me{meem31l&d(k|h0TN&3Pl^)G?R<l5RcaBXw}IzeGF
zm<zI!Yx}4ekZ<YUDZvR7SJ<eA?gQnLk><a_X+E0Xzh`DJl(=@`@Kx%H048tG4yVM6
zXbVJ2N_W>}e$6<~?HclBgo!-9<ECS-@Yy8miGc)n`ggU!$OAxE<zIYZNPj2zHz0bE
z=HfSBw^BdWG>=3ytqB_uc2RZ1o>Y2A&U~$&VQmK^#~NE&Q7>$er`mkJHw^}3Ls9RX
zwx9j@lo;v#c=))>FP~TfO$SF8*4$M<ky$t3p0@dP1ZKCmS*34;tDwRAG*>^mW;LFX
z5iep|-Db+`9LQH_eL_336FXRtaZFf+C0c!nHC@Kc*k1DA2ZAWZG_(g3fRfx${Hxuz
z3q@shqu}ra-mVTpiNJ~EwlQOG{sv!R*RyT+pXV8bAPmjjfCoQ`JA&^UnKK|U($7-*
z*$U0F>g&o_FcP}hkz3)4OwOI)U?v(c>Lnd1^A4+tm%^D<N6oarMGZYlZRE0wt!=vT
zF}bI^A9>q%$ycq<)s!$1l-KEyf;DXGjHL~UD3+Eu4SP#@=3&<KSA!mon3}S98#F7H
zvi$iKWkat@lpo6HA5~Rp^-R93OPzet(A?gTNx$BbV5i=VWv{_}hH=L;PJuDeyAH)y
z;?c&9ek)n1Mi`$J%v{F!ki}nY@Gb-u`wc~+Sogtf3cdi*PkCn~;!5@;zU;NHy|qxQ
z=UJ-mj#UvvAaprn=4WQ)u>q|aj9UPns`u`j$V06%ISdw)@HHpk6qwipmK^mv*OQ|n
zpxCBcuoU1`5pb%Bmnuu@_xKEVOgd3rZqV$CByNrNVhsc3pfvY&=FjY!xvumySY((<
z@1@Pyk~xD$w)G{lCQOk}5d=`ljHkhk98ueyy){uxAw_+V*E2A@s{!MY<X-o8!Aa>I
zYU}4`f0sRSt+AcRn|v811rKW>mkrE78cwovzqbotVw3;Yk({xhjSx?8zM2<bK#F{`
zP^X;bWj$-fb$K8Xd9&~bpI4^*+h0dn!c>=wlkQ)dB<~tDO4JkW>Y^+eNJn~}JY)tQ
z54$E(R?x|tvcD$!!bE`88ZpWkuQk-QMXJ0-8Z@6+Q@GbQUbl2W7m{<|){<h=7G<y3
zMb?6Y<$x7`oApB#;Ig44dXN(OmZtvhs?Dvj&FXwcLTq0zu>_Q=7$F(}*xkN)Ui4Px
z*2j~z-v@u>YZ&tr!iN)hfPp1YChz@z3XOhy+j3kfD^p;$^3)qJukTBYnSEjrxBNc{
zl>AQsCZO^LNV@NXuNJOn4f#v%LR9{P!>kJse#oFhon-(Kw6gOCqYA&G!u;oadXPJn
z?C@9r1v%lD;fmmlm0dA!lNmwLkmsAd2F)3;<CQTlb=Gv|$i$evm*b$n4N7buD89m0
znZE?OR~&jH$NRDRYTYdsT6Jf1+O{`;^0;vpaLVkGu}zd9?oV=~2+u8f5T>Adkb(l5
zN9+Xy3do0TK(Ru{(#K1IMq%qCp!wEK4Y}bi#XNug|LiJz%q{=zou_tgfNvl6-|b>v
zxB&+oNp1&s=&!Q%PP#QmDtV?`=pwSPUfXBeerOnk^(LNdBE%UI1#U3#hY$GXg4=;j
z8yU0`LpsVhnI%!4(fgnxJWBQ)A`|)xIkR-<fZ%S4)_7N6nUZOMS*F@cqTl_eGIY#$
zW)e};TZ1L6kt|xLH3XCxI0kg{hcV?ZFpj`f9kPx$T#4=%k?R%w_J<_Dx7yl$jQH?4
zUcPbv_~_uE-!YgJwN!{I8TjQ$=C3xR($bM|S62DkM20n!2c_Btt4eHNm{<c>x$%cS
zK^0o`vdg=|wZdnsx7^}IX2E8#`*c?~F4S87I3f8kQQNZYM7h>L@v16zlC^2rPAdh*
zRX;Ebw9Pj-Nu|pHH__>@e8k1C)65H+<b+kVi}&c<GXFqdJ0UYdUrXHQQYQI|#IF;=
zxy~yZ^l-;RCGSfXdeQuIZ$lJ%A#b94Hjd)YR<2c7XewSQOVHvnENGsZ!)Kvn!8+ox
zK3{XhcZNlj5z7H1-Jz#%%Hd7YS9=ilV3VfYh0U0_=|8kAzH<V0hACR9RlJn1asdg7
zW1>KX!w?rw!TztF48pRcraFIy%C(9Qv4@YcJr%(o`T#^WGi2k43cY>o4j|e#HxO;6
zt|=?up9pJADjA8c2=qQv*AX9q{j%83c$o1-w+tW0d7A>`Og;7=kmC{F<t-ey_g+Eo
zy@sJ%4I`DQBAidQY4G`HS(~3+$Pm<zBGie&>{}NIQTJENUTRp6S-Rtvq9GnT&qlcY
zq5lbl!X+b<if;U}>X3YCxnHC(MXSj4Ay)*6)>YnO@yK=us|)q0J9mAOmcN-tjI}Qi
z*qaw?7)c^UF_89uhuv77Vd}Uf;P|bpxKs9aVd@<1r<+tY9bN3I=nt=?>gpq=S1_OL
zxs&I!15VI59=O}a<;B$roauq@i*oWk0KsbCTx?KqdU@zC7++h$Z0()ow44664QK9C
zat)EIFIVbmvu7@g0O>sT#;=b+&Hy&37^9NapP?SH9QmkX3unl1XIlym#KXj`(dJHD
z5t|Zmj*Arq@xtH#M#THzMPHat3&Q?aw9=bBce;_6M|_0uO67zx7=nQMq!pBKS7e-h
zW;e?sb*O)3by|3=Z!1+}E35OBty4+&mJ{;3S$FEH1=%}`Toi*LkWq2E|LcnS-7LHR
zykam_lP%5ou=3~)|NQCzjNXx&T^%1wKUWtsF84o>;LN0e&vrKsa6JPG(z})Vmj7|u
zu!3Du-gLtee|~+l20wG$!$-%1<k*HMT@3gcaiD&A%2UKa{+<E~7mPBhYHX)!+QY6C
z>U}`S?pbQ6zYAJ_`KhQMWDD)#-Z-?q#nl`vS%d>^)gpKesnhC%Q!#X{=WDZG-Po7r
z=J|Sl!r$^4L8BOlGR8VquWp~2S*+fXBlxKFv^~Wa!K1MR;&UwtpVocx@#Mn*U?Nb~
zBzjZ;Y)V2?ohO@6MEQ+65L+bh$+P%V5j`K(y~WZPm^N35NT~}?x9~6~6zaM$-Vvga
zlyf;&SVYcmE^M%g^^y10?~{*{`SAW>(HPjC)K6!V9|)5(Mw98109|=<Ms^2uHgLzA
zF2;vV|A6dnA4dz;W1Y3<M7|!gbzA*3NK;mtPw0+P4VD4eB#c~u^7Y^6TZn$V9>p3(
zC7Gi9jk!zjuL@sAMi@R^o*#9*wy_quy&%BP-VPPAj>dBRv1P3>>Pz!tf@J!KGy)7m
z&u{K*khwT=d{lxG*3ry1_2IAA4l1QeT>%zKfB<15un&KgC%K~m1Fdp36m_33ll-Y%
zbR$tMmD#&zOshUbBlT=R@UdJe*B|Qmr&CGcb&&QK!3cS}=rcU)FK<rj40)E!0%)I;
z*i=45vWCe(P^z+j@1E#C@4gE)h`<44L*C)k%cZRj^Y_5IA9Qq0jEtX)6XTHmV47QA
zhz$MRG_xGcUF*-1`ZUrg&Qyz=RrIXba(q{KL}XtT<cW!F1gOnl<-z~HPYsl<@ywn3
z`T>!i8SeMlSLFA(+mt^pUJXgIb_yR{n6Jw<_UHA036%T@lb<zNf7&vr6dqGkLNZU`
zssn|ck{IYNCeS}zXMk>)vw=W3phIqd09SVIp70Ndf(2z;V6vPeucLg7e8C`uC)+^p
zVN!3M`_2{wR2Dfq9PgW5-Pli0qTOtxjhdtX*rysA7!vTfVqMU74^D>(v)os4k(cK6
z2wexza))~1hSn|=0|30v0+ekRnrU`P4ZWaC`m|ZRshP!4_Gjr<hl+JLtF`s{S{zEr
zPbNywG`0(~Ozgtph_`5M>5dEa%%QF|-qfknp;hn^39AZr%fq|is*i4ei>>+Cb@zp$
zpK6$s?<A&K@K|e8X&(hzUqV)h)!kG>AdWYFv3GK|e*Bfj;?&>OeHi?Nc!?^%X7YH6
zMDU)Iiw4k=i@o-hG}KSkJmdw;3GW-`LqVr=RkH>vo;aUf8B&IRn9|zlDKWlhqsb?D
z;%}?ioF!FtAt%6HV7<HNL|6A+3X4>dc>x<Svv~+1sJdv>;oFzU+#X@c98rBbuxXB)
z{T<HKxNp<(^1h|2#3`7|>-&&vikq#Ia!z^QfFf7%a}uUmkSTh;PHdUiZ$mV35eNtU
z`ac`XCgT%k`;G_t4b&t&36^U971i_+mOft??W3MK@-m506Lg*`d9%>nbgJ#`$v}l@
zuh-}%7qzGtS_k32fT)Eo8X4=8VC(X?SCBjXp-%n^W2b*U9pVWK6R%Wo5f9p>0<Z-Y
zL|;8P{A;x9ssmWG2vhb_tTGXu@TUqyT+>I))8_g=f5uNXZl9&F4KsRZN*%?a<mXpi
zetfDrvPpH?wvJ_RHvOEW2(;UGJD0TPw}Uk|DvBfY-@N??^jPcV;VNUS#l^%!$*X*#
zBSB;>`C3<2X4F%vH|yGb61leDu7Gd67v|k}zN@cmXz6T^o*%(;6L{^B(ode@XFCO;
zCOcpJkElQXcST9ev+DngaMXXX;mPo;b+?iPKsA+%8mYg#Q!n<fQEr^OF#p)nkfRb@
zNx(*mrv^IhWn*pJeb6|N=itrt+1tsT7fxT8U;^J)Rel>Hf)1aDUTr{oYNqKxMfXX?
z&?)Ij^qn9||G)iixc~S;5>oy%ut^*!+@5xByny|N!KlQWrh3dz0z9K87bxW@K*nr{
zNVKLpU?o8YppL&~we6onV2}ExjblQoT~3{iG2DFUx!)XF8xW?-eyC?Am)HNnjs#sw
zVqu=o|Gx<#@fXCY{GK!I&*Oa}#AJV!s6fa6hVO``v8t+_uor8P#z1|;$T1(k*~}+>
zIYPTaJ0P17C-uo7Q}#4$(G(^kR#iz_Pxy%8mG?Cu&I%MAW5Er5Xa+<^*|$9_aN;IC
zUe@nFt)B{H-)q=Bh>RM7cZNzEm2~I9h#?H9X%Z6WK-S;GZxda&n?JTJ$63VS($K$L
zk3Ub88oDllCnT)I^mkJ7U$%c2g}poP|HEyG_>o#Ir^I+reE+f1?Ko7_NkSZXHtg<1
z*}JS!ksfPCZO8{mEn_G$)gv&G*?GNap}7WE-rippN+rv3724iHYJO@web|$u$)0OW
zIlMvZHG|i(&`m_Xvl*+&xTp?0dAE?yNQB7~$wNuDmaCy_ceq@_Tui?^RnbxVciFb0
zF+?my!mswqq25_iC7d6S->d`sa5#3(EY;bttBbo(uAzaE+>oIk({EXP(1pVLyatOv
zNPMHK@JtU~$fn7%Tv;xn!_STbFnj?b%!_^>1q^wyemIn3dDq+DcSbu)gYcK;>rkWL
zv+{~^146l%XKVcjydMN((L9OAv@9u&+Mi!auh)sjkMu4^Je0rO&8uC>8C#=-?LTG@
zatY!#GF473HK$jStB*|jnooS*4_HKTOr&py=<3eMN)gUCrL_}q^le+l<rF68hYv~a
z6!}6Ppru9;Y<-!5{OH6Ck)qaCAa@ri<OOxtN05xQ4E_v`L~xF5_6~NiMx5a>e{z0Q
zt4ZzD?e!8q74Z-$b^xhO$QixU>1fwrcQI+$V@uWR<f)gYGysrHcL^Y$RrdRGfj5l@
zx0WKk9-0nV;}q_mQl>QA-RvO6VcfW5G}fIvG}i10of1^YA2kbTJ0@(T1q$OBT6;$j
zTY4^ZKJ!K-#aC9pJ)t_FYM#<`_hwuUho|^{A0?N(A%)T*fd)l65UnyFNc7+leRFar
z&}ld%o_X5sy{}w5*a&j+&qQX$HvZ_gUmus={B6g}(426)px?y@VrG)I+x!Ve2BU^?
zlP*KlRnrtBHGf27`bSSiKFYT@TA7L-6_Avo-AfD=KoCW(ey&<G-4{Pwvh0<q{2MDj
z4~n!x+*0gaxx()q^v#`g;sec>8VA;rHKzNzMgINmA(<M3&Z&VIS-8)g02~eg13k_H
zUCZd*SJoIIJxfqFE0tGy93?H*+uq&Q7$Cq>QhQx+Kn;~6U*_OfL_n}4W~vmYgJ`#d
zcRYNvvqX8?c%RcBH8!95T@!nj*fy1$+Jy~sh_xf_{{5S3Xw#i`uyxwu>s868OQ8n0
z_094v-GczI;{|6*@zCL!6Y|ZWmubF%=3RMzRRdM+r=wSK6cfG!;cB4~#r+f@Lpc}y
z81-K@moi1ZfWJxVA0-;>xNjNE57Z@@8oj>X1-Ocxgx{&UxX^2R%IA1ec;*lLI(`D-
zmT`v3+qK+RHY&*L)}yQ@hnIcbTV0pD2Rwwbjmv|%P)qJ`1zc2~$?V(KH-AnqIQRD-
z(?}0f4SIf0m4EZJ7Z<Zc7IB)5(8L0$!0rsSwE(`>9v}(H5rxp+iaNICO!m~ojbkVD
zRu>fCXN|p6*du@MhVk^0xs9Yx54#oWBBwhJ<?TUQSv_A_;LrA*q&xr<bciXz!4LeC
zARW_%4PO*vciQ5piR+lvJiljqzhlGgklIO#EK1l+ho{rBD;uUqKV&OqoA3B|pZzvO
z4H}X6XiVIgSd4(1rGbUyWLfDkPi_exF5@^zUKciMC@hh=@qzL4wVy?Y=ShDMw5v0=
zVADOOMv!9$-RtKdkuJ=8qBb;=&Xt?v#e2Rnmon1A)3qM)M5(4$!R7<+Ct2F+<#BgD
z-z$j^TSFtt!;KoQB^@MSGeD8MR@P;c2Efeeys&JG*4(IlH}*?S(cywtPKqsSshS?1
zkUmjxn%Pg)M5^iO&G7{*PHcHRf&{Y@1hCCIPv^}B@9}S-Xsk|@aWzU8ZrD_N3N=Ik
z#SKx{VvQ+X7y1Y$rFwdRp^10?i2Y1z9iA6)u$L6w##!Qd%Z?hLc_Uz#b$nUP|I3!6
zC?Rw*Lw`B+=mz?B3RCyUwjz6{mkNv`2LvUp8i9sQLWWspcg~AJ2|9~i*xOF~++QY7
zoMPhz=DCi+V!|3j*DaE?D7lYiEa6xlN=Fb)e>Ua0*Y`=m5JEGbnQ_0;s+BfAyN2%-
z=59WWU1w`~NS(j$NAxG?4cS}4+ugs|(2Jz*Bu4W)rBhkDpFhKAVBB=`=D`thyA*ft
zpu{w`%(f+O@+Q0<vXezfM?$>NL*f!fkcYjNOQR24J2L!(`+5JU89kv1jCgr|q{y`N
zuTrjGe+~WM>20?};9*9UAFFUb1xr(ON+J&YcZlb>uNXY9F+FE|r+ak62`i1pjWtU#
z?)mJCb0H)!_W=@jVi`361f)Oa-y?hgrmZ|eKZN;j?U9eT4>xLa)B0Zj$f5U!dJ8qb
z0Y20p#nbT_!TF&~@X%#rH>+n$$K8Lypd`ajoh-wPA#pV3MV|8G2Qhb|y#OXK;VVG0
zxK9WHkm&tJXrN7(xk;}b!2hjX?SAajh_DPBMq+_iHu#x`wDF&W(08HZ84`i+-MU2P
z&Qxbh7Fby-3>A$L?Pt>2K^2$)zqQyn@6H2ps6c#2>SOvAQL5?bk&)cBg(WaDo(ahA
zJY2!p0h?XmpC&B{jQly-P%X^(uE}HWT&p9bI+>CCvZCcLQtcLSB&$I=`2lGMwbC8X
zQU*Ll;O+_HKp!iFen8yJqF%G2P93|U8l=YZ!Tr_Zn>V&KjJ)gZSvQc))r7Jo^k^<l
zToA#ht-2h$R-^bx{Pg4py-nJJ5Pd}gAl0-7dO1_IuK=hv37Ybf>z&wQu<k~Jpm}4Q
z#;eX|(|IYORu#f^KtTcsr4%BY{G_2hYXmTe3Olus$<sPj=3eflsW{M~%P3mz?pe&n
zV9eo`E4L5e7NMM?%Kd=(5%vc^IW-XY{h<6;h!}dwSK@`gdvB)4(g?C?;St|-d1`G2
zUbE*>HiI56CGL^<o~+v{?yoQ-tdCywHMY%>5HrW-`B2W&iJUTyjsr$D7;j<R1_q0z
zO6jYiA*{fKKNOB<L$!UCXHxj(1?*{47?7eoB7Olrc?Oi#1d=*pe@&|~&N2H{$~+}u
z$ug-b$Yrq5DP^dJDaS6cP`IuuKSG++)5;sM0~pKm5-Dh%iB2}J?h7&}SR&Obm{$5y
zCPIC>25aOy${>QF%i_ici~ee7kJ(+c{n`>u4ZTVy0k2Wy#QfiU3uybeFzOzWg_QJC
z<v!o|6X3`H!!2MyO(cF{jw;p6B}=tvMzIdI+?V+<Enku_KR!{-!{zK|mDY_bk-&e|
zw9xwfd!PgYK#_og@<GoNZI=xls?|T@iT;)OiBDfeRI=$Q-8nw!3sn2U5Qo^Ccii`?
z{Qj8vCNl0U4@R;%1|c3ux69&_uA2L^SL>x1OX+{bQ3n;Q3gQUPymdFTV1)Ei=RS>|
zQsU!uEwT3Mikhs(_*+U+*S;??;@5&{cVh3QD6-?IH@UBt@1%YPIWS65UM^3WuPH`s
z(iFen=otx*dEO#F&?!)PBp_JB(oh$zdHmWF08#z;ym$&?91ksAiZ}0*qac|M<LNEj
zn;NN8dga6r=S9n+r@o18`Ts@5xzgG&?mKg21c|fOjRZ0d#Zn3L&p7YPrL50;G1G`=
z7I1Vl)joe#I~rWT@i8q~U0H+KLwDj$kx9h^8V1rJJ<MrSHX1;-)dGh#^GM_$(4>>X
zNrAq;BKdtW^!kYF{=)J0=|7-znC|WF>`OhYPRBfS5YXrfp_BimaEDAo--uCCU<ks1
znnUh1bWJU(LH8e!9FXF7B&e!_w2`r$I}LePzlvoOQLBfGmr})m?4ln86GMUejuiwn
z;2fX&-z?DDJ9$I>D610M7bJvpYJ2gi{p7>d*%ubT0N$h<cwdPj&>r*67P=*Ue1xf~
z!s`1cuxW*_%*nW4*^yni(+z*+fff+%$oSb{Hh7ha(ys<{&CS{+hSZgMw*aA3#TPwi
zxHULMD5+bOeI^macy4&*?!51we>H2?+!vnxq7Li9Vv38CxYInCLC^juT=2cs@v4Tw
zAH!TKpRN@A@YWDDC2)8Jxb8k`Zfzih*u^D)BYI&x!%shjjwVORcG1mRfewUJh=Vj_
z6m29i><2P^_+~olZjDwM^&M~1#9d|P1_{2JMdXTvL)V*+@SyM{Ml+P3>8~T#j^7Nq
zPv_&;c)wAzL^4GgwlX3P1}C8noOuBych(5OCD-jR3rn#+j)eyri>C+sl57AraIW!|
z+v&;=O4s00;S#{!T&Eo?z^;GRSB3SyFwghz&0H$Pt~Df}BjF*&>c8$3Y~`BrM9dnR
zqdxrj$})uSK`k}-B0MdA-Q|u9b?XCnr|;c3f;?@gcNxlF8n@5JRTWT)mcy^`l7T0B
z$$O#TB@iblm8@Pfa(d~@hwEya)s}NsKMu<fB}Q~XX>|cocgEdaUc|{bV}5$r#g^g3
zo2O=!)X`3L_{s(Zh5$Sy-v*&pY`B`_fW|nFT5Qhyirf~O5`Z(s8?1a^)JOy#Fzd)H
zL_wctzy`LDf|u$B)PaoUu7qYt64~zqppm64QMdhmbWCLhFw?P9+Wd?enNUbOgU|&V
zRv_3)-U9XWl~2i?9^$3({FxHlzVRcTmVvDR+Vq%OEV>Km;(=~k-ezUy_oX3ZQ<DK+
z5y9C21Vc%_1Q&A=U_O|`p)mIsGwr73i%wQdGj{b2EkSRCeh#QT_D%K8_EPG)==WU%
zg-R^nDI+-k#Fp(M=U-H{wnl3!`zBXn;t+c0efCTWycXGXjavPJAm}q%K%SXT2o{H=
z3J}kGjfB1<Rk{#6P^e9an(ZWaKY=%=RIbmga&uyC8#^eM14Z_=bdbA$4f5-3Rl8>C
z(w-ILjMV^b*5j<h+VqHQ`=)H~`1QAYY4bmEk)fzLPXyRv_N}Y7MpJED#L|(crt|Xw
z#gTZLN7>u6@xGJK2=4;mA@V2Mh|pSx(j{$IAI8})-_m`*W7m-Y>62VH&d+A#<C){7
zO`O^?{`TS?IiVw1NGEejQEY_6Y?Lhk)VrlU+Vsv$p#7;87pN9{MP|Iq4X7(>4B~4Z
zifHR(-kU?)95J_uULecn%=KGbWqJc4i=%m5M;v2cWZYn$c}0OGU6A<cJH-#?F)ydT
z*0s(QP)jvT3?>yn8I-A$f5)`Av&sLw=RkW_fX|0MfWL-+d5}+7fR@Gw%lH?xw&QaU
zHzHxvT+7w2j;c&xExl;#5Vy4(mkB-}Ms<s<Ss+^eGuR7rY*-EcD$*MA8SYCglGsvX
zgknP!-31l;M}$D0nQP}Tv3PU+e!i8;O0$k5wQ?pA?8Un=dMlG4e;g#UXcw`MM<->I
z{3<298!PZnN5Xx|MS?Pv*dOg*#o;J$7*$DbOThlsZUl4rZWN$}xUL;RYQB^ul*BYN
zHm|Olq;K(2%i=JHJr~NxuxdQDg)1Q@qwaLdT<EMD@|5B&o;Q_v4U?(AY7m+@S^3^A
zxi@JMEac}dmNPrwtS>7;Dx=PBz-^?4i$xkpswM={3>~EtDDghs*>vf45D_S_Yno4p
z$15@^!G4+BBFqdNgE<RFFRFizRpCZA_`GX<Tu1GTDf_@nIOJ45{h$-KLu_ypU^n>}
z7|k`*)ah<4>dcM#e&XgcXB-$Gl<D+cj~rZ)+6q5inSEq7$=^6;pdu?^n<_<7!G;;}
zP&VS>z+Z4pLGMk~WKU9@E_VE3R&}=Po5W>;SR?u{L+=IbS%+fX+I;-;&^hUI9X3Go
zjDp4HrBuo%;@?uV@+itf1-wvEt`ET&n>yYvr#nVHH&D4j@2j30nDmGwe3l|Qd%}#E
z+Q}w~`V*(Y>0hj#X!5VNNlG^sQ*=iLkO46oz}K7mdWoWXi3~<G_&Rh^C{Mz2<-of!
z+=aw}Fp;hMa<R=r+tc~?6sc+_PqKru7k&mmKmqVF@^%qFAD@yz+aP;|G@c&sM-dX-
z7H5*{k!#uK;n(-<?i5K3rNgY`l>pG4QHiM$DR}tem7nW+mD};=s>xCH3B}UTXw*lZ
zbeoQRCXs!TB2Wl55Vm7M9b(Y_g8&6Tt{WA;=_pIN{?oYc(URPlk&rv7nsTO%a_R7{
za4%OjVK!6xFCg*_opBd~Mxr_EX84QRexzdCMRP14H9fTrlg`HM$|uQD*eJUMKoe9A
z3M0B`o{kSF76e8pyFat@kv)?4lK|cHCWlngn4!@K(+{n-%PodP&$R2`5RcG7GyCJ5
z({CFmRYlF_<67EZphBL1p4!B!B-!ZK9>jrwas(^*dK-Ly#-sHgDlsiJ@$%wek~_Sh
zY69J+XGO1~Sv`u?C9!`TF+P`)lbld@u~N7-cDf3T=^j@!H6NCW@UP||tJYF0%&<Mg
z<(ksNjf3{SKt-aXev+J}0XlItC1`3?odTmjiPi!@Gg*6+5ZU!yxlV=+6jc2ZzC?Ll
z1IY)}immc|ii0Unw>!SNbNn3mU39iZ<60w|Si-d5g%-uw1OL~Mek6SRm+LjzA$g|+
zz#!b$^N#!vRgcE5xSg<`dxZZst-C*^q%R2$QO{q93jIw*?YD-{Z`aVAYS%Ek*^-?<
znF&tkIhOxz4^999HppV~oj{H>A9D?aCjG|s{#S6>K&~Dl`hFyxenyCOp@+GJxB+K3
ze@i?$&Hzy3k~Z$4yx8f#$24rZ&uk~59LV!#fCZOvz=TPsxJ^L<GF~H6m-&Q~DCYzy
zI;x5U7-1;F@G>v*a<N@!!<E;(Ichx)y&^!x#a!n`+s#(Zdrta517w#QX1vGMLqK08
zvJn%_*Z7dV`A0+#dj>ZSEj=SZfWxe*;YZyglsag4smNh_zt$j-G27VJlCclg>&2zG
zl0h^F*@(z}!*lne+7c@Zf7{-LDRXU(qN2WUn=dUY&3mv{cJZ(rian%L6d9f=G^(}L
z6=-hcP$oa&BONMCvQ3LTiL~lP_`xkgiB63HLmTijmoG`AJp!%<yGI(|78CuBq*h`{
zL6ZC4UAP;LKt0{yv3_CtA`dHJ2G?I$^Bx#xtT^Y|L4GF~aL8pieR0|JQ^t#b>&a57
z@!+$MuW~=r96x>_33Wap<@@#dPOeB+Df&{A-9X)g_&G@j$@EVbxWs_Zsl&T=k+MBA
zllCu=4?v3J*V9k>)V7`bcI~HrnCC(+PAE-Ng!z~E`Au<}CFv3BKt(?%whKh8$<-Lb
zeQV&DmEPIw9xW*LaU{zpw&{m!E2)-EGluz8YN7%@sKVv;<$<;jfd$Gz!x1x*`F&ff
z#d$-jqU5fJ%fX5OQcn9>Lr1vftJWw^#dRF~k%tp9=a5XJrWwl>6;~IEG>d1OlA8(I
zBDIqEpsm#BLV%(JWpl<%&y8H;2+dCP1kIMpoCz~ls7;s{??9)c?uVH>F;~_Z-Rbp?
z_MdCqE~&u{s&Ihc(wCQy+2M~gyOu$&)V4_{MK@x&v(%c{hfUy)u(Llg2+_bo<4%V?
zcIs?YD{qTxos|_U3W?W{J*m$me={lL<(d6<Jg{bPErA+LRNe4}kyktJ><r1~b>h{h
zma+o^m2kvP=7anu!6bY05@tE~KQr>MGO7KD+2Uvx_yC;v1K`MNxLBH705nFvIED5i
z+1@xTeI$>5y{K!O=(&bFHHCT63F|aLPdB0GPSLqQf8<HPKcLqR=e~a2t+jc!3&rYY
zH^-g)7dg`B#<Nuo_rfsqilt1lenMNm!yx>7jsN0RJ^1g$n*aU%f5+v2*TDa-f&X0t
z|GNhMht>dL62;P}cpBT%55Di{3s*VMg^LQ9{nXEy^%lsrV_2T;+GPdd*N@S(xDmX$
zr$6XrDvo$m&_#Dc_kg-X%JtL(RQc;Hb4GD0o-5%8eN4(p=Ql=uR~Z>fwpsq);P{nY
z3`U$OKwX)H3<Al!d;%@#y*!3aR%v?^M>_e*X1@Mil6?G1a6)}!;xVQdZ9%ROyD<Cv
zK;i7TzSG-!7qiWJ@(P~yM>8je5MrG-ZI>6)!QqvwLXR;eiMk~3_6K7+w<G|wf(4+n
z@VoTEWkOE7bkQNgf;#J2K*}+$!-lbd+tHDp+}N+alCQp)VhI5DX=uC(x}>me`1XfM
z<4>a-<aO}kMshF3n@~O6=IaB+N2ua`sRsX|#vIGw3;KdL3#t||aThJF`~_LVvV{SL
zb_tS37m}9;*9#Cu!Ys#py%~*^GD+-3kZy_SO&6l%E|M7#iGPPyYu?33(V<vs|4gn^
z;^J!U5kmBIVZH+I)3VAo4h-TC@FFiOlOEoWYoj~CH->;P+x(TM<)jbImp8|yNh3AS
zvIBD!ctCqcZb!#cjSas>3L^Bw`)yrrYJR(4EPicEz<f^Svn**PePbBl`fWz`g|#bg
zCgXHn8%djooC1$S9A8ypSz&9GpRN~`$~|Jx47BpYMf;ZEegABzzJ*L-<Tg{`9FdV&
z<LSpTR}qa^(%+nislrl2CHqZA^U_rL{45&HPrVy#Wev6SuF0TBa|%TuSEr{rJY%_<
zQ3@AiM&-r8yfS|gyjjn?d06qh8-vdSmwgJCs|X-__B>5p-S^TOZLjo*?#XAGlgoB5
za5**U7p*g0{>qao9$Y8ve^*}h(Ks{Hk2>qo2N@}F8p}}f6^m!TP6^(d#__i9b9VuX
z_ihsQ8sHur93qg9z`ws>9yglK{L@L}11~oYc@wvAb=~tLBU$6Ot_`M5(8@1Yyazy<
zdhoi&$v0x~d2<uZJq)B<b_`J3wl5$1Ja%swJ~_S;&D4_pR7a+!&U88LTo%G7RGh=^
zDZ&7-V;eme1Lqy$0H=p@RqzF}bIe!;&=8F>GHH~v!gFKK9vG64`Ue!)X13D~=rFL{
zd*5FG3$Nm9K8_BbhnkhcOQpF_Vc06JrDm2w<}RO#Ud{48aPDYyFBwbGET<sr)6_c&
z(?p(FOtg9ys_`F-ma%BJ6ycIquZ%7OJqjoGUwuRDDWt8iub+qprvm5vW)BO#45#i>
zW=g^=?C1s%!LFh=&m0fq7eF{KBBTvRt8kyE4roMGI{G4*!IJVn+OmLdc{5#9QG0~U
zhflWNcYNz~1HS~fKJ<W{ynk+}6Xc@UN%quZ8q7`KqK<29si~<k$6hBp)M$ZHj}3y&
zOWt|c=G%t(D+D$le2g11@$Z;^{JCSy);9rf=sB?>a`L>oI*cvRzc2-!_|!qSy}T>^
zCUI6<R_`PED1-e3g$PE=4T)Lz$B>wQgyuBwZXEb(@EV+BfM5`IHcRN~>kElk8>2yA
z1Fr}RPz$s;B-6A3VY27WeW_4&DszgmZ+qFyN0OB&lpGT1Q!YLK4bpQU>Ts3mT$|5S
zDFvP=Kv35GMroKQUB4#tlF5$G0;TnDf1P?X;(KkUNH3c=kOw!m`{l<*`HJU^NEJ?U
zO%1&8A>3%me?ZCL`$jX+E_dok^49=MeOx2kFz(B=%?VkB05s5yoQeWDSMZkFv3!h=
zAXX|K6g7}3^kS0^**3zene{DwyLOFh23=rBE7ytYUejE5UKIe^dEyDPAgyP=Yy>a!
zB(Fm|J}v=7yt*FPCb!m^@)OBw2i+NXCjzHjAl9GtBbZl#S3;6&fdTC!@&01>?b6%X
zhB>Mk^Y2^^7$383DKKX0v1x5U45>kej$-sOvClk~d)XsVhTUSZo0{}`U!Kjty;3Z5
z&PW4zqz2C`X8c3DmarnQqhJ1!N{caKBo531#dIG}mOQuo00wlOFE$%WW-F_<&UV_O
zCC~?wICa5<G*09(x2r)82jM#Y!BX~rR0V&39_DC9BkC`=;y`O6{o4T@HUa(<_sg~m
zVljmWc${C$s>ToHwWO~(auf&UNJ8fzJ)zG=%S{Uc=DE3!d%f_R{O@`ANwjOzjON)=
zR~T7SlyPVYvH?k<qT&RbbJJP%_+zg45NCQ<wTW${k~d+}{hWu6w-sA!6J38%!ojBt
zl%5PvgbGt{QjyWzi6~y>a4j$8GE`KD;aLpZ8U@Dk3yyRa<+HYL#Fc}fmvk0KJ|*VQ
zP^*XvaXHVIccz0eCkiWIz~qt0(B6meop4IB(vaxlpE&im|95%L5q|;psRP}ct-K*4
zH2_=9#?*|yfDTRJ{Avrs)6C*iMc40u1sA5jxCFWQnwVRta$9Mr;Zlz6-MsZfv&@Cg
zoiFt>%NxIT^U$5in<zpE<Wc3gWhG|Cx{DGMV9^j3h7a92VwT!a*d>GZbzUaD{GI{j
zlo_IX9<CCfHH08&yo1?C2x+)YTtAde$CRwd1j^pq^^UepiQ-F9J~Y-L-t%E%_9?8&
zXTtw_Ov8XndKn+JtUnaSv<4PLy|c#BnO-X>p1l%|qW3hTOeI*h7Zo=GDX(uV8!}1N
zWv8E3TB$Avdf*o0GU(wf8?e__rbxFv8M2qc$+6gpw!LL4D_>+2ijYu?f0WaTm2y{6
zkWaW)XAW2@g;X5j9?2W6ary5ZBsK;HY&ORa%f`}?Y}$Y-6|Qss0UgZ%X>6$(Mx}r1
zo<Q`i3kQB}TDR?l?!mPl&QZ~2)duVQCb!1cJ8MtF#tNCBdPg_6XE;(SoD2f+z46!n
zNjecDgGV=jW~t|!?(TTkp^1mgFWS|g<Swaw_9wUK>k49W#LQ21{AOp${?tKIZw$kB
zSI+szzOVlH*02x$Xtt|>Fp;f)hj>N^r>Hl*To*=+{!Ed@k15fe@BT(GF&6)1l5DW$
ziX*0xJ;|d2a?A&KxG7l$1xoJr#efu%h;pmD&?mB!p(8_C5z#c18)tg6G%vG!S29m0
zAoH-L=YGY3Zop{fzc18dpSI|!Rao(Y6J5ydn+I&Ryw}8vWo3#lVQrnn+w+~-J=OIg
zpBPW>a9pNTVFJm?<PGE$n{$5I9R1*8W3=js)d8{oeRMO0tK{uT*`TJmr2%I&O~h6w
zy04sSsI^c=l#V{}48Odn#BH{8>1!5R**<!ct~{7az!Y$ow5Pe5z`9Dr>GFZ5|3^MX
zUofske1rm<BBNIF$U{H{g*C3C7&;_AgKjB7NAS#nD1a{9|Ktzlh=1eOj9}HJ_N((c
zQIeCvQFKU@&s#klqj0-<X+9z7sFCHN?p2aP3Fc{akaGL<q4nE2)`4$~&DXV5ucZz>
zEgM->`DjX)W#Cz$?YCUQsJden+~le^_-cw+0>-pgWeqOxni1k<uHpvfv6yNPZ@V7o
z?pgBF0I0b*P}eVDZeGuK9_5{tr&uJBIA=P&O+xa1mYq#kQ%;+u+b?A%ixB>62rXXP
zmR_k~Yb+killeJZPuE%mg-(Y}fiOYV`ie9i0!<k^>h3Js!;fN>6H6!=Yhn+pS@cQM
zkKg|T(ks<|w3nOyQ|G0s%}6=TC9@u`-1@ga^LHacEN!l97KAf*)WE)R+j(;X-D;cy
z<O<P^z}_Jcdzp5PuS@-!bH*J%ji|9JKxi=#ojx@!9Bn}JfJDy`lir}SX;xODRIc!w
zs>vsE3>wL+^0NZX^|1FvpI|kZHA<FPo`tq#-%2kqDfU5Qrz=|$S#qu6v1chZ9%@gz
z)l{w8V~(iORB}3OgdJwsDER@;8b?@rpji1HtWFavIhC>}!w4&FfP2~kOzMn;*e0L{
zozX61qhW)IxZ832jFbb6DkRDxT7O4@xF<|%#pJ^YxrwPyBFK%MvGeyN{O;fzP!Qf}
zHv}ZD+63=IYR)OAeyM`ISHZwBiT;7UfgaBN{exF1v%fWhqa9Xhe<)j3s+D0Fd`0=9
z&p0D9D^TEzn`~6N-*dJ_`N%}*&D>J*!){s%KqHNY0%H<zr&Xrox(^}QGoStkRPKJI
zBMp3?tcaUHJ2%o-l03%Z#x5&TbAfgGC@Q)dP>=&ZmhqzX9uTrL6#HdP3rd+`lguYJ
zAZQf~hB!#Y!o&J{nBmx&xY<GHmahMvTHdI{ECHK;K<17I!YK=plh?>%e?y)AqW+7~
zm>=nRxOK!Jv=6{>^OUwTkZ&f=WX<F3G3BI{8k!uv`v+w98wk|HIDyze;ey2+pbN=t
zUfOQVv1sBC4Pxe(-syWYeccH6c4^%ZpqJ5LexL+DuoR=fz)qMoZmL~2PqhhNPf3gc
zPgfkwpZFD8s?dzqdB|z}srjYIFrT?8BYDUiy)&*|I;&<wR%@d;$Y9`arD;De7Rwbs
z$R;eLGN&|waVKOcFgM{Lau8<*f>+i&&#O5gA4!}q)?>G)T@2`!Uvak7)lcff&kFri
zy?abpm)@ny@DC`j$Z^;AT(!daZ1fSUHtR?Bkc~;L-w(^K)E;9J^a1}t(%(vsCB6PF
zlvqT9lZ?uOZ8B%z@h!wZ`F_ZTyL{3AmSGwTlnTQ7yWY7rx@G(g?p@E^2!^N`y~C%!
zO<gk5!FnF&60U~FBn|T&u3QgZR*xjJMlVyiUG{^#tsVhf3A<s^*s(Y*l8xKtm%hzY
zPV&ri<oEHczM5ns*<_O?^KIpCQVKtvlbJJ05KiIBR7Tg&xi<Q5ikQ>fwMN9J;}-m%
zCC%6G)QJ@>`7!U^(>^iNp7qZ>fOliOa5!*aNl2*5D&Re}3t0O9Mm;T?!$1A<t(Do=
z-lB!^y{V6^O*S@3FV>1Z*2ohVspBE>SY)1frRue<Gu54CdXiQ7Lti)kOxoqt<yY48
z>(R6Zi`UxEE`NQSl$l(<<+H=o3F(Q?S^E99s;ykPZjZ3vrK@*i8)pUW;9Br(%a-WO
zq^nyyRePsdrUdo4^H`qES+(l#Rpu8O%<M<PK0AFA|29#Qcc%4XSueqJn(7~YXGc#=
zQ=Rs*O-P~LdhhRWXAePVArYqPM-7(ko2vO)@46ZsIkWUmm5@6y*=;we|H%?rf1L%`
z;B<`v9+mW0{h!;~^q2CV)-`pteM-*HRaq@ESyb|-#A*4Z48Y`l<~^{Mp8wm_xm-J=
z>dW^{i)V9hVBXd9Ch_&Z>MOH??%dD*lx1|=Gnb?Nl<S1Ve#><opDY*fMTcEYzV<+>
zY?55s%!Wz6OG8Cuo;+LjQO8N5q+<HJrB|}pFYOX_jo-Dfu4%3VFjaNdKT%#(fBnS2
zsk`leE&mk^JU_kuQ2Zxd{XhIaXZ0Vh|D?YC_avb6ffpn2<g)|2&^uGV<p1TkqW;ev
zxNLpB|KC~Lfm+f7j9%<|va6x<>870lYB!7m+YEp`{jlp>8&|#*UN&E3nE>OoaOn`9
zKHGEI=|MZM<h;A&q1$?OX8cUyze)4stIqA4^*g2~@^Gr@(%l9|r!9`3;;Rs0+zYJu
zte_R&SyK@utrdq>8~_%auzjGTH0>@Wp+?whTZdiK-Ev@2?tAUOCl*U4N12|>+_>hS
z_LZko*eXPYY`(D+|Eqn$RQgeOMVE6S@6kOSyZ7u2<~e-(+PeE$6K-aHyyvw^VtyV|
zgTJj-e~&_6E?4;khNQy_-|;#8+v<O0{wIxFwMzd&7*~~^PP-CzPIOYFd2X8PsfFK7
zvS!`g+j5%iWS*YK`Kes*5^Vo!{av~zY=eryf{?&I_W0uW5q6?X(R0IIInF$LSj|7d
zivQ)00M@_O(cAaz_-(DS?YqQ@zSLX(m)6)V6v_&W*idifqPK*9I<x)5NICt`zI&(M
zvZ;4HEHik#9(c}K$n^Q2EN|3b-(4TYaAV(2Z|lfBRh}TzJ5GIBVpkGlb>gQ^K55u#
z*gu<X<|Fx^OmV<LE9J=P_Mepbe|__R=^%XeQFJKh6UX8LueL;?^fb2u|F!X#wp!Qv
zG3#!Y<CBzb3x471{_utJ_KP{!R2fXNp39z;RNI{FnEyCDB0TGS{yqn*yz}vwzb@bd
zAC5tz;le>TJ++}U*6fq5M^;p@;+%CSlj>NS|I{DazFIxx^1G+IqSxAeTr@R2v(GQt
zzCDZK==EdE%a@#2TQD*5umq2?+nto021dI%SN|rRt3J;g<(Mn?(@;`bp@7{sQ{MAR
zsid;h(KGgG&bv=G$j|vbIhOx&7lZh`?D;!3@4IX&`EJVzv1i=PyF9#eweBr8y#TEG
zAKiK-b@c6?SyOl2igLJ@&|UYe?CqS;-m+tF*S$66wb8y4{Y2t+@2gtiRol0LW&d91
z-|?XvX6bM2HoZNGXW=a8o9+&F%a&YGtgdG>G3wo<x@$GRlVL1t$knw{zy0Rjlu-8A
z!BfST!OZP9Ief$A+)3Oads`QWxqW7ua^f6c)uZ(pRkf#L{a0kU)?GKaC0OOQZ~m#R
zIt$Id6*wIcmeLZkevq_mbR2!8WQ<XNj;8g|G%%V5M$^D(8W>FjqiJA>qyhf_H!n0=
bvjcBiVE9&+FuCQ}<!4#l6L~+f3V#Lw*z80I


From d01c1231ed8eed27fe099e5b79f32a7370ac9f12 Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Wed, 9 Aug 2023 17:52:15 +0800
Subject: [PATCH 02/45] fix tokenizer

Former-commit-id: 572ea3bafb1b495e33b1abd1998972f3a5e6f310
---
 src/llmtuner/extras/template.py | 16 +++++++++++++---
 1 file changed, 13 insertions(+), 3 deletions(-)

diff --git a/src/llmtuner/extras/template.py b/src/llmtuner/extras/template.py
index 5b00af03..91595751 100644
--- a/src/llmtuner/extras/template.py
+++ b/src/llmtuner/extras/template.py
@@ -1,10 +1,15 @@
 from typing import TYPE_CHECKING, Dict, List, Optional, Tuple, Union
 from dataclasses import dataclass
 
+from llmtuner.extras.logging import get_logger
+
 if TYPE_CHECKING:
     from transformers import PreTrainedTokenizer
 
 
+logger = get_logger(__name__)
+
+
 @dataclass
 class Template:
 
@@ -179,11 +184,16 @@ def get_template_and_fix_tokenizer(
     template = templates.get(name, None)
     assert template is not None, "Template {} does not exist.".format(name)
 
-    if tokenizer.eos_token_id is None and len(template.stop_words): # inplace method
-        tokenizer.eos_token = template.stop_words[0]
+    if tokenizer.eos_token_id is None: # inplace method
+        if len(template.stop_words):
+            tokenizer.eos_token = template.stop_words[0]
+        else:
+            tokenizer.eos_token = "<|endoftext|>"
+        logger.info("Add eos token: {}".format(tokenizer.eos_token))
 
-    if tokenizer.pad_token_id is None and tokenizer.eos_token_id is not None:
+    if tokenizer.pad_token_id is None:
         tokenizer.pad_token = tokenizer.eos_token
+        logger.info("Add pad token: {}".format(tokenizer.pad_token))
 
     tokenizer.add_special_tokens(dict(additional_special_tokens=template.stop_words))
     return template

From 6404167ab76070cddae10552b80ea49d1d10eabd Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Wed, 9 Aug 2023 23:00:26 +0800
Subject: [PATCH 03/45] support val set in streaming mode

Former-commit-id: d86ea314a197fd821770d895e988c48d46679047
---
 src/llmtuner/api/app.py            |  2 +-
 src/llmtuner/dsets/preprocess.py   |  7 ++--
 src/llmtuner/dsets/utils.py        | 28 ++++++++++++----
 src/llmtuner/extras/template.py    |  2 +-
 src/llmtuner/hparams/data_args.py  |  4 +--
 src/llmtuner/tuner/core/parser.py  | 52 ++++++++++++++----------------
 src/llmtuner/tuner/ppo/workflow.py |  7 ++--
 src/llmtuner/tuner/pt/workflow.py  |  2 +-
 src/llmtuner/tuner/rm/workflow.py  |  2 +-
 src/llmtuner/tuner/sft/workflow.py |  2 +-
 10 files changed, 58 insertions(+), 50 deletions(-)

diff --git a/src/llmtuner/api/app.py b/src/llmtuner/api/app.py
index 4fc5fc43..47b7661f 100644
--- a/src/llmtuner/api/app.py
+++ b/src/llmtuner/api/app.py
@@ -49,8 +49,8 @@ def create_app(chat_model: ChatModel) -> FastAPI:
     async def create_chat_completion(request: ChatCompletionRequest):
         if request.messages[-1].role != Role.USER:
             raise HTTPException(status_code=400, detail="Invalid request")
-        query = request.messages[-1].content
 
+        query = request.messages[-1].content
         prev_messages = request.messages[:-1]
         if len(prev_messages) > 0 and prev_messages[0].role == Role.SYSTEM:
             prefix = prev_messages.pop(0).content
diff --git a/src/llmtuner/dsets/preprocess.py b/src/llmtuner/dsets/preprocess.py
index d2150dbc..534d77b5 100644
--- a/src/llmtuner/dsets/preprocess.py
+++ b/src/llmtuner/dsets/preprocess.py
@@ -104,9 +104,9 @@ def preprocess_dataset(
             if len(source_ids) > data_args.max_source_length:
                 source_ids = source_ids[:data_args.max_source_length]
             if len(accept_ids) > data_args.max_target_length:
-                accept_ids = accept_ids[:data_args.max_target_length - 1]
+                accept_ids = accept_ids[:data_args.max_target_length]
             if len(reject_ids) > data_args.max_target_length:
-                reject_ids = reject_ids[:data_args.max_target_length - 1]
+                reject_ids = reject_ids[:data_args.max_target_length]
 
             accept_ids = source_ids + accept_ids
             reject_ids = source_ids + reject_ids
@@ -166,8 +166,5 @@ def preprocess_dataset(
             **kwargs
         )
 
-        if data_args.streaming:
-            dataset = dataset.shuffle(buffer_size=data_args.buffer_size)
-
         print_function(next(iter(dataset)))
         return dataset
diff --git a/src/llmtuner/dsets/utils.py b/src/llmtuner/dsets/utils.py
index 31c48222..e1093a95 100644
--- a/src/llmtuner/dsets/utils.py
+++ b/src/llmtuner/dsets/utils.py
@@ -1,15 +1,29 @@
-from typing import TYPE_CHECKING, Dict
+from typing import TYPE_CHECKING, Dict, Union
 
 if TYPE_CHECKING:
-    from datasets import Dataset
+    from datasets import Dataset, IterableDataset
+    from transformers import TrainingArguments
+    from llmtuner.hparams import DataArguments
 
 
-def split_dataset(dataset: "Dataset", dev_ratio: float, do_train: bool) -> Dict[str, "Dataset"]:
-    if do_train:
-        if dev_ratio > 1e-6: # Split the dataset
-            dataset = dataset.train_test_split(test_size=dev_ratio)
-            return {"train_dataset": dataset["train"], "eval_dataset": dataset["test"]}
+def split_dataset(
+    dataset: Union["Dataset", "IterableDataset"],
+    data_args: "DataArguments",
+    training_args: "TrainingArguments"
+) -> Dict[str, "Dataset"]:
+    if training_args.do_train:
+        if data_args.val_size > 1e-6: # Split the dataset
+            if data_args.streaming:
+                val_set = dataset.take(int(data_args.val_size))
+                train_set = dataset.skip(int(data_args.val_size))
+                dataset = dataset.shuffle(buffer_size=data_args.buffer_size, seed=training_args.seed)
+                return {"train_dataset": train_set, "eval_dataset": val_set}
+            else:
+                dataset = dataset.train_test_split(test_size=data_args.val_size, seed=training_args.seed)
+                return {"train_dataset": dataset["train"], "eval_dataset": dataset["test"]}
         else:
+            if data_args.streaming:
+                dataset = dataset.shuffle(buffer_size=data_args.buffer_size, seed=training_args.seed)
             return {"train_dataset": dataset}
     else: # do_eval or do_predict
         return {"eval_dataset": dataset}
diff --git a/src/llmtuner/extras/template.py b/src/llmtuner/extras/template.py
index 91595751..413333ac 100644
--- a/src/llmtuner/extras/template.py
+++ b/src/llmtuner/extras/template.py
@@ -97,7 +97,7 @@ class Template:
         sep_ids = self._convert_inputs_to_ids(tokenizer, context=self.sep)
         encoded_pairs = []
         for turn_idx, (query, resp) in enumerate(history):
-            if turn_idx == 0:
+            if turn_idx == 0 and prefix:
                 prefix_ids = self._convert_inputs_to_ids(tokenizer, context=prefix) + eos_ids + sep_ids
             else:
                 prefix_ids = sep_ids
diff --git a/src/llmtuner/hparams/data_args.py b/src/llmtuner/hparams/data_args.py
index 60945b60..de470ae2 100644
--- a/src/llmtuner/hparams/data_args.py
+++ b/src/llmtuner/hparams/data_args.py
@@ -90,9 +90,9 @@ class DataArguments:
         default=None,
         metadata={"help": "A prefix to add before every source text. Use `|` to separate multiple prefixes in training."}
     )
-    dev_ratio: Optional[float] = field(
+    val_size: Optional[float] = field(
         default=0,
-        metadata={"help": "Proportion of the dataset to include in the development set, should be between 0.0 and 1.0."}
+        metadata={"help": "Size of the development set, should be an integer or a float in range `[0,1)`."}
     )
 
     def init_for_training(self): # support mixing multiple datasets
diff --git a/src/llmtuner/tuner/core/parser.py b/src/llmtuner/tuner/core/parser.py
index d872afcc..692f9b13 100644
--- a/src/llmtuner/tuner/core/parser.py
+++ b/src/llmtuner/tuner/core/parser.py
@@ -67,33 +67,33 @@ def get_train_args(
     # Check arguments (do not check finetuning_args since it may be loaded from checkpoints)
     data_args.init_for_training()
 
-    assert general_args.stage == "sft" or (not training_args.predict_with_generate), \
-        "`predict_with_generate` cannot be set as True at PT, RM and PPO stages."
+    if general_args.stage != "sft" and training_args.predict_with_generate:
+        raise ValueError("`predict_with_generate` cannot be set as True at PT, RM and PPO stages.")
 
-    assert not (training_args.do_train and training_args.predict_with_generate), \
-        "`predict_with_generate` cannot be set as True while training."
+    if training_args.do_train and training_args.predict_with_generate:
+        raise ValueError("`predict_with_generate` cannot be set as True while training.")
 
-    assert general_args.stage != "sft" or (not training_args.do_predict) or training_args.predict_with_generate, \
-        "Please enable `predict_with_generate` to save model predictions."
+    if general_args.stage == "sft" and training_args.do_predict and not training_args.predict_with_generate:
+        raise ValueError("Please enable `predict_with_generate` to save model predictions.")
 
-    assert model_args.quantization_bit is None or finetuning_args.finetuning_type == "lora", \
-        "Quantization is only compatible with the LoRA method."
+    if training_args.max_steps == -1 and data_args.streaming:
+        raise ValueError("Please specify `max_steps` in streaming mode.")
 
-    assert not (training_args.max_steps == -1 and data_args.streaming), \
-        "Please specify `max_steps` in streaming mode."
+    if general_args.stage == "ppo" and data_args.streaming:
+        raise ValueError("Streaming mode does not suppport PPO training currently.")
 
-    assert training_args.evaluation_strategy == "no" or (not data_args.streaming), \
-        "Streaming mode does not support evaluation currently."
+    if data_args.val_size > 1e-6 and data_args.val_size < 1 and data_args.streaming:
+        raise ValueError("Streaming mode should have an integer val size.")
 
-    assert not (general_args.stage == "ppo" and data_args.streaming), \
-        "Streaming mode does not suppport PPO training currently."
+    if model_args.quantization_bit is not None and finetuning_args.finetuning_type != "lora":
+        raise ValueError("Quantization is only compatible with the LoRA method.")
 
     if model_args.checkpoint_dir is not None:
         if finetuning_args.finetuning_type != "lora":
-            assert len(model_args.checkpoint_dir) == 1, "Only LoRA tuning accepts multiple checkpoints."
-        else:
-            assert model_args.quantization_bit is None or len(model_args.checkpoint_dir) == 1, \
-                "Quantized model only accepts a single checkpoint."
+            if len(model_args.checkpoint_dir) != 1:
+                raise ValueError("Only LoRA tuning accepts multiple checkpoints.")
+        elif model_args.quantization_bit is not None and len(model_args.checkpoint_dir) != 1:
+                raise ValueError("Quantized model only accepts a single checkpoint.")
 
     if model_args.quantization_bit is not None and (not training_args.do_train):
         logger.warning("Evaluating model in 4/8-bit mode may cause lower scores.")
@@ -113,10 +113,6 @@ def get_train_args(
         logger.warning("`max_samples` is incompatible with `streaming`. Disabling max_samples.")
         data_args.max_samples = None
 
-    if data_args.dev_ratio > 1e-6 and data_args.streaming:
-        logger.warning("`dev_ratio` is incompatible with `streaming`. Disabling development set.")
-        data_args.dev_ratio = 0
-
     training_args.optim = "adamw_torch" if training_args.optim == "adamw_hf" else training_args.optim # suppress warning
 
     if model_args.quantization_bit is not None:
@@ -145,14 +141,14 @@ def get_infer_args(
 ) -> Tuple[ModelArguments, DataArguments, FinetuningArguments, GeneratingArguments]:
     model_args, data_args, finetuning_args, generating_args = parse_infer_args(args)
 
-    assert model_args.quantization_bit is None or finetuning_args.finetuning_type == "lora", \
-        "Quantization is only compatible with the LoRA method."
+    if model_args.quantization_bit is not None and finetuning_args.finetuning_type != "lora":
+        raise ValueError("Quantization is only compatible with the LoRA method.")
 
     if model_args.checkpoint_dir is not None:
         if finetuning_args.finetuning_type != "lora":
-            assert len(model_args.checkpoint_dir) == 1, "Only LoRA tuning accepts multiple checkpoints."
-        else:
-            assert model_args.quantization_bit is None or len(model_args.checkpoint_dir) == 1, \
-                "Quantized model only accepts a single checkpoint."
+            if len(model_args.checkpoint_dir) != 1:
+                raise ValueError("Only LoRA tuning accepts multiple checkpoints.")
+        elif model_args.quantization_bit is not None and len(model_args.checkpoint_dir) != 1:
+                raise ValueError("Quantized model only accepts a single checkpoint.")
 
     return model_args, data_args, finetuning_args, generating_args
diff --git a/src/llmtuner/tuner/ppo/workflow.py b/src/llmtuner/tuner/ppo/workflow.py
index 0ca8cbd4..aa372671 100644
--- a/src/llmtuner/tuner/ppo/workflow.py
+++ b/src/llmtuner/tuner/ppo/workflow.py
@@ -41,14 +41,15 @@ def run_ppo(
         max_grad_norm=training_args.max_grad_norm
     )
 
-    optimizer = AdamW(filter(lambda p: p.requires_grad, model.parameters()), lr=ppo_config.learning_rate)
+    optimizer = AdamW(filter(lambda p: p.requires_grad, model.parameters()), lr=training_args.learning_rate)
     total_train_batch_size = \
         training_args.per_device_train_batch_size * training_args.gradient_accumulation_steps * training_args.world_size
+    num_training_steps = training_args.num_train_epochs * math.ceil(len(dataset) / total_train_batch_size)
     lr_scheduler = get_scheduler(
         training_args.lr_scheduler_type,
         optimizer=optimizer,
-        num_warmup_steps=training_args.warmup_steps,
-        num_training_steps=(training_args.num_train_epochs * math.ceil(len(dataset) / total_train_batch_size))
+        num_warmup_steps=training_args.get_warmup_steps(num_training_steps),
+        num_training_steps=num_training_steps
     )
 
     # Initialize our Trainer
diff --git a/src/llmtuner/tuner/pt/workflow.py b/src/llmtuner/tuner/pt/workflow.py
index 2a9f8279..b4ea148b 100644
--- a/src/llmtuner/tuner/pt/workflow.py
+++ b/src/llmtuner/tuner/pt/workflow.py
@@ -38,7 +38,7 @@ def run_pt(
         tokenizer=tokenizer,
         data_collator=data_collator,
         callbacks=callbacks,
-        **split_dataset(dataset, data_args.dev_ratio, training_args.do_train)
+        **split_dataset(dataset, data_args, training_args)
     )
 
     # Training
diff --git a/src/llmtuner/tuner/rm/workflow.py b/src/llmtuner/tuner/rm/workflow.py
index 19527ce8..b19a13e6 100644
--- a/src/llmtuner/tuner/rm/workflow.py
+++ b/src/llmtuner/tuner/rm/workflow.py
@@ -39,7 +39,7 @@ def run_rm(
         data_collator=data_collator,
         callbacks=callbacks,
         compute_metrics=compute_accuracy,
-        **split_dataset(dataset, data_args.dev_ratio, training_args.do_train)
+        **split_dataset(dataset, data_args, training_args)
     )
 
     # Training
diff --git a/src/llmtuner/tuner/sft/workflow.py b/src/llmtuner/tuner/sft/workflow.py
index 69d200f3..a5cd2cd3 100644
--- a/src/llmtuner/tuner/sft/workflow.py
+++ b/src/llmtuner/tuner/sft/workflow.py
@@ -46,7 +46,7 @@ def run_sft(
         data_collator=data_collator,
         callbacks=callbacks,
         compute_metrics=ComputeMetrics(tokenizer) if training_args.predict_with_generate else None,
-        **split_dataset(dataset, data_args.dev_ratio, training_args.do_train)
+        **split_dataset(dataset, data_args, training_args)
     )
 
     # Keyword arguments for `model.generate`

From ce9ffca0d98eb3d8050c22473f9e32dbc9c17c4d Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Wed, 9 Aug 2023 23:10:20 +0800
Subject: [PATCH 04/45] fix template

Former-commit-id: ac29f4d5f0d9d514aec9224fd751b9eb49430e7e
---
 src/llmtuner/extras/template.py | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/src/llmtuner/extras/template.py b/src/llmtuner/extras/template.py
index 413333ac..dd88782d 100644
--- a/src/llmtuner/extras/template.py
+++ b/src/llmtuner/extras/template.py
@@ -97,10 +97,10 @@ class Template:
         sep_ids = self._convert_inputs_to_ids(tokenizer, context=self.sep)
         encoded_pairs = []
         for turn_idx, (query, resp) in enumerate(history):
-            if turn_idx == 0 and prefix:
-                prefix_ids = self._convert_inputs_to_ids(tokenizer, context=prefix) + eos_ids + sep_ids
-            else:
+            if turn_idx != 0:
                 prefix_ids = sep_ids
+            elif prefix:
+                prefix_ids = self._convert_inputs_to_ids(tokenizer, context=prefix) + eos_ids + sep_ids
             query_ids = self._convert_inputs_to_ids(tokenizer, context=self.prompt, query=query)
             resp_ids = self._convert_inputs_to_ids(tokenizer, context=[resp])
             encoded_pairs.append((bos_ids + prefix_ids + query_ids, resp_ids + eos_ids))

From 0dc9b41b167565067260163e2857621df51fa4c9 Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Wed, 9 Aug 2023 23:14:27 +0800
Subject: [PATCH 05/45] fix template

Former-commit-id: eb6e571cb7c0a6da6696e8ce4b39cdcdb7f04e36
---
 src/llmtuner/extras/template.py | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/src/llmtuner/extras/template.py b/src/llmtuner/extras/template.py
index dd88782d..d94d5b37 100644
--- a/src/llmtuner/extras/template.py
+++ b/src/llmtuner/extras/template.py
@@ -101,6 +101,9 @@ class Template:
                 prefix_ids = sep_ids
             elif prefix:
                 prefix_ids = self._convert_inputs_to_ids(tokenizer, context=prefix) + eos_ids + sep_ids
+            else:
+                prefix_ids = []
+
             query_ids = self._convert_inputs_to_ids(tokenizer, context=self.prompt, query=query)
             resp_ids = self._convert_inputs_to_ids(tokenizer, context=[resp])
             encoded_pairs.append((bos_ids + prefix_ids + query_ids, resp_ids + eos_ids))

From 7ffd961b8b531f88064d9c6889af984d44f72f51 Mon Sep 17 00:00:00 2001
From: jiongxuc <joyce.chen@nokia-sbell.com>
Date: Thu, 10 Aug 2023 14:57:12 +0800
Subject: [PATCH 06/45] huggingface login for projects must login while running

Former-commit-id: 3e000c2b60c2e29bcafcf8d39c1a5d567ae2491c
---
 requirements.txt                   | 1 +
 src/llmtuner/hparams/model_args.py | 9 +++++++++
 2 files changed, 10 insertions(+)

diff --git a/requirements.txt b/requirements.txt
index 9b74b21d..c71b6c9c 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -16,3 +16,4 @@ pydantic==1.10.11
 fastapi==0.95.1
 sse-starlette
 matplotlib
+huggingface_hub
\ No newline at end of file
diff --git a/src/llmtuner/hparams/model_args.py b/src/llmtuner/hparams/model_args.py
index 253d9839..94e882fe 100644
--- a/src/llmtuner/hparams/model_args.py
+++ b/src/llmtuner/hparams/model_args.py
@@ -1,6 +1,7 @@
 import torch
 from typing import Literal, Optional
 from dataclasses import dataclass, field
+from huggingface_hub.hf_api import HfFolder
 
 
 @dataclass
@@ -63,6 +64,11 @@ class ModelArguments:
         default=False,
         metadata={"help": "Whether to plot the training loss after fine-tuning or not."}
     )
+    hf_hub_token  : Optional[str] = field(
+        default=None,
+        metadata={"help": "Path to the directory containing the checkpoints of the reward model."}
+    )
+
 
     def __post_init__(self):
         if self.checkpoint_dir is not None: # support merging multiple lora weights
@@ -70,3 +76,6 @@ class ModelArguments:
 
         if self.quantization_bit is not None:
             assert self.quantization_bit in [4, 8], "We only accept 4-bit or 8-bit quantization."
+
+        if self.use_auth_token == True and self.hf_hub_token != None:
+            HfFolder.save_token(self.hf_hub_token)

From 58e95776e1f27709beab84a2ba6eafe2488e920a Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Thu, 10 Aug 2023 15:20:44 +0800
Subject: [PATCH 07/45] fix webui val size

Former-commit-id: ad6e7c76c7b5590ea6ca55c4f76db2d3206b5987
---
 src/llmtuner/webui/components/sft.py | 6 +++---
 src/llmtuner/webui/locales.py        | 4 ++--
 src/llmtuner/webui/runner.py         | 6 +++---
 3 files changed, 8 insertions(+), 8 deletions(-)

diff --git a/src/llmtuner/webui/components/sft.py b/src/llmtuner/webui/components/sft.py
index 678693b9..9db57ac9 100644
--- a/src/llmtuner/webui/components/sft.py
+++ b/src/llmtuner/webui/components/sft.py
@@ -38,7 +38,7 @@ def create_sft_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict[
             value="cosine", choices=[scheduler.value for scheduler in SchedulerType]
         )
         max_grad_norm = gr.Textbox(value="1.0")
-        dev_ratio = gr.Slider(value=0, minimum=0, maximum=1, step=0.001)
+        val_size = gr.Slider(value=0, minimum=0, maximum=1, step=0.001)
 
     with gr.Accordion(label="Advanced config", open=False) as advanced_tab:
         with gr.Row():
@@ -88,7 +88,7 @@ def create_sft_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict[
             gradient_accumulation_steps,
             lr_scheduler_type,
             max_grad_norm,
-            dev_ratio,
+            val_size,
             logging_steps,
             save_steps,
             warmup_steps,
@@ -122,7 +122,7 @@ def create_sft_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict[
         gradient_accumulation_steps=gradient_accumulation_steps,
         lr_scheduler_type=lr_scheduler_type,
         max_grad_norm=max_grad_norm,
-        dev_ratio=dev_ratio,
+        val_size=val_size,
         advanced_tab=advanced_tab,
         logging_steps=logging_steps,
         save_steps=save_steps,
diff --git a/src/llmtuner/webui/locales.py b/src/llmtuner/webui/locales.py
index ba9a73bd..9bf3d7a9 100644
--- a/src/llmtuner/webui/locales.py
+++ b/src/llmtuner/webui/locales.py
@@ -227,9 +227,9 @@ LOCALES = {
             "info": "用于梯度裁剪的范数。"
         }
     },
-    "dev_ratio": {
+    "val_size": {
         "en": {
-            "label": "Dev ratio",
+            "label": "Val size",
             "info": "Proportion of data in the dev set."
         },
         "zh": {
diff --git a/src/llmtuner/webui/runner.py b/src/llmtuner/webui/runner.py
index 763ff614..fddc4070 100644
--- a/src/llmtuner/webui/runner.py
+++ b/src/llmtuner/webui/runner.py
@@ -83,7 +83,7 @@ class Runner:
         gradient_accumulation_steps: int,
         lr_scheduler_type: str,
         max_grad_norm: str,
-        dev_ratio: float,
+        val_size: float,
         logging_steps: int,
         save_steps: int,
         warmup_steps: int,
@@ -139,8 +139,8 @@ class Runner:
             output_dir=output_dir
         )
 
-        if dev_ratio > 1e-6:
-            args["dev_ratio"] = dev_ratio
+        if val_size > 1e-6:
+            args["val_size"] = val_size
             args["evaluation_strategy"] = "steps"
             args["eval_steps"] = save_steps
             args["load_best_model_at_end"] = True

From abdfa26d06a5549a70865d356452e572dbe50209 Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Fri, 11 Aug 2023 03:02:53 +0800
Subject: [PATCH 08/45] support DPO training (2305.18290)

Former-commit-id: 3ec4351cfdaf2aefcc7d13345e19d79874ed61d3
---
 README.md                                 | 60 +++++++++++-------
 README_zh.md                              | 68 ++++++++++++--------
 data/dataset_info.json                    | 20 ------
 data/refgpt_zh_50k_p1.json.REMOVED.git-id |  1 -
 data/refgpt_zh_50k_p2.json.REMOVED.git-id |  1 -
 requirements.txt                          |  3 +-
 src/api_demo.py                           |  2 +-
 src/llmtuner/chat/stream_chat.py          |  3 +-
 src/llmtuner/dsets/preprocess.py          | 39 ++++++------
 src/llmtuner/extras/callbacks.py          |  9 +++
 src/llmtuner/extras/constants.py          |  4 +-
 src/llmtuner/extras/misc.py               |  8 ++-
 src/llmtuner/extras/template.py           | 73 +++++++++++++++-------
 src/llmtuner/hparams/data_args.py         |  2 +-
 src/llmtuner/hparams/finetuning_args.py   | 24 +++++---
 src/llmtuner/hparams/general_args.py      |  4 +-
 src/llmtuner/hparams/generating_args.py   |  2 +-
 src/llmtuner/hparams/model_args.py        | 13 ++--
 src/llmtuner/tuner/core/adapter.py        |  2 +-
 src/llmtuner/tuner/core/loader.py         |  7 +--
 src/llmtuner/tuner/core/parser.py         | 61 +++++++++++++++---
 src/llmtuner/tuner/core/trainer.py        | 31 ++++++----
 src/llmtuner/tuner/dpo/__init__.py        |  1 +
 src/llmtuner/tuner/dpo/collator.py        | 51 +++++++++++++++
 src/llmtuner/tuner/dpo/trainer.py         | 75 +++++++++++++++++++++++
 src/llmtuner/tuner/dpo/workflow.py        | 59 ++++++++++++++++++
 src/llmtuner/tuner/ppo/trainer.py         | 19 +++---
 src/llmtuner/tuner/ppo/workflow.py        | 25 ++++----
 src/llmtuner/tuner/pt/workflow.py         |  8 +--
 src/llmtuner/tuner/rm/collator.py         |  9 ++-
 src/llmtuner/tuner/sft/trainer.py         |  2 +-
 src/llmtuner/tuner/sft/workflow.py        | 15 ++---
 src/llmtuner/tuner/tune.py                | 22 +++++--
 src/llmtuner/webui/runner.py              |  2 +-
 34 files changed, 513 insertions(+), 212 deletions(-)
 delete mode 100644 data/refgpt_zh_50k_p1.json.REMOVED.git-id
 delete mode 100644 data/refgpt_zh_50k_p2.json.REMOVED.git-id
 create mode 100644 src/llmtuner/tuner/dpo/__init__.py
 create mode 100644 src/llmtuner/tuner/dpo/collator.py
 create mode 100644 src/llmtuner/tuner/dpo/trainer.py
 create mode 100644 src/llmtuner/tuner/dpo/workflow.py

diff --git a/README.md b/README.md
index cca8fdc3..fcbe0761 100644
--- a/README.md
+++ b/README.md
@@ -12,6 +12,8 @@
 
 ## Changelog
 
+[23/08/11] Now we support **[DPO training](https://arxiv.org/abs/2305.18290)** for instruction-tuned models. See [this example](#dpo-training) to train your models (experimental feature).
+
 [23/08/03] Now we support training the **Qwen-7B** model in this repo. Try `--model_name_or_path Qwen/Qwen-7B-Chat` and `--lora_target c_attn` arguments to train the Qwen-7B model. Remember to use `--template chatml` argument when you are using the Qwen-7B-Chat model.
 
 [23/07/31] Now we support dataset streaming. Try `--streaming` and `--max_steps 100` arguments to stream your dataset.
@@ -54,24 +56,18 @@
 | [Qwen](https://github.com/QwenLM/Qwen-7B)                | 7B                          | c_attn            | chatml   |
 | [XVERSE](https://github.com/xverse-ai/XVERSE-13B)        | 13B                         | q_proj,v_proj     | -        |
 
-> * **Default module** is used for the `--lora_target` argument. Please use `python src/train_bash.py -h` to see all available options.
-> * For the "base" models, the `--template` argument can be chosen from `default`, `alpaca`, `vicuna` etc.
+- **Default module** is used for the `--lora_target` argument. Please use `python src/train_bash.py -h` to see all available options.
+- For the "base" models, the `--template` argument can be chosen from `default`, `alpaca`, `vicuna` etc. But make sure to use the corresponding template for the "chat" models.
 
 ## Supported Training Approaches
 
-- [(Continually) pre-training](https://s3-us-west-2.amazonaws.com/openai-assets/research-covers/language-unsupervised/language_understanding_paper.pdf)
-  - Full-parameter tuning
-  - Partial-parameter tuning
-  - [LoRA](https://arxiv.org/abs/2106.09685)
-  - [QLoRA](https://arxiv.org/abs/2305.14314)
-- [Supervised fine-tuning](https://arxiv.org/abs/2109.01652)
-  - Full-parameter tuning
-  - Partial-parameter tuning
-  - [LoRA](https://arxiv.org/abs/2106.09685)
-  - [QLoRA](https://arxiv.org/abs/2305.14314)
-- [RLHF](https://arxiv.org/abs/2203.02155)
-  - [LoRA](https://arxiv.org/abs/2106.09685)
-  - [QLoRA](https://arxiv.org/abs/2305.14314)
+| Approach               | Full-parameter | Partial-parameter | LoRA | QLoRA |
+| ---------------------- | -------------- | ----------------- | ---- | ----- |
+| Pre-Training           | ✅            | ✅                | ✅   | ✅   |
+| Supervised Fine-Tuning | ✅            | ✅                | ✅   | ✅   |
+| Reward Model Training  |                |                   | ✅   | ✅   |
+| PPO Training           |                |                   | ✅   | ✅   |
+| DPO Training           | ✅            |                    | ✅   | ✅   |
 
 ## Provided Datasets
 
@@ -88,7 +84,6 @@
   - [Open Assistant (multilingual)](https://huggingface.co/datasets/OpenAssistant/oasst1)
   - [Self-cognition (zh)](data/self_cognition.json)
   - [ShareGPT (zh)](https://huggingface.co/datasets/QingyiSi/Alpaca-CoT/tree/main/Chinese-instruction-collection)
-  - [RefGPT (zh)](https://github.com/sufengniu/RefGPT)
   - [Guanaco Dataset (multilingual)](https://huggingface.co/datasets/JosephusCheung/GuanacoDataset)
   - [BELLE 2M (zh)](https://huggingface.co/datasets/BelleGroup/train_2M_CN)
   - [BELLE 1M (zh)](https://huggingface.co/datasets/BelleGroup/train_1M_CN)
@@ -103,7 +98,7 @@
   - [Web QA (zh)](https://huggingface.co/datasets/suolyer/webqa)
   - [UltraChat (en)](https://github.com/thunlp/UltraChat)
   - [WebNovel (zh)](https://huggingface.co/datasets/zxbsmk/webnovel_cn)
-- For reward modelling:
+- For reward modelling or DPO training:
   - [HH-RLHF (en)](https://huggingface.co/datasets/Anthropic/hh-rlhf)
   - [Open Assistant (multilingual)](https://huggingface.co/datasets/OpenAssistant/oasst1)
   - [GPT-4 Generated Data (en&zh)](https://github.com/Instruction-Tuning-with-GPT-4/GPT-4-LLM)
@@ -139,7 +134,6 @@ Note: please update `data/dataset_info.json` to use your custom dataset. About t
 ### Dependence Installation (optional)
 
 ```bash
-git lfs install
 git clone https://github.com/hiyouga/LLaMA-Efficient-Tuning.git
 conda create -n llama_etuning python=3.10
 conda activate llama_etuning
@@ -161,7 +155,7 @@ CUDA_VISIBLE_DEVICES=0 python src/train_web.py
 
 Currently the web UI only supports training on **a single GPU**.
 
-### (Continually) Pre-Training
+### Pre-Training
 
 ```bash
 CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
@@ -222,7 +216,7 @@ CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
     --resume_lora_training False \
     --checkpoint_dir path_to_sft_checkpoint \
     --output_dir path_to_rm_checkpoint \
-    --per_device_train_batch_size 4 \
+    --per_device_train_batch_size 2 \
     --gradient_accumulation_steps 4 \
     --lr_scheduler_type cosine \
     --logging_steps 10 \
@@ -233,7 +227,7 @@ CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
     --fp16
 ```
 
-### PPO Training (RLHF)
+### PPO Training
 
 ```bash
 CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
@@ -257,6 +251,30 @@ CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
     --plot_loss
 ```
 
+### DPO Training
+
+```bash
+CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
+    --stage dpo \
+    --model_name_or_path path_to_your_model \
+    --do_train \
+    --dataset comparison_gpt4_en \
+    --template default \
+    --finetuning_type lora \
+    --resume_lora_training False \
+    --checkpoint_dir path_to_sft_checkpoint \
+    --output_dir path_to_dpo_checkpoint \
+    --per_device_train_batch_size 2 \
+    --gradient_accumulation_steps 4 \
+    --lr_scheduler_type cosine \
+    --logging_steps 10 \
+    --save_steps 1000 \
+    --learning_rate 1e-5 \
+    --num_train_epochs 1.0 \
+    --plot_loss \
+    --fp16
+```
+
 ### Distributed Training
 
 ```bash
diff --git a/README_zh.md b/README_zh.md
index d5eca99d..955fefc1 100644
--- a/README_zh.md
+++ b/README_zh.md
@@ -12,7 +12,9 @@
 
 ## 更新日志
 
-[23/08/03] 现在我们支持了 **Qwen-7B** 模型的训练。请尝试使用 `--model_name_or_path Qwen/Qwen-7B-Chat` 和 `--lora_target c_attn` 参数。请注意使用 Qwen-7B-Chat 模型需要添加 `--template chatml` 参数。
+[23/08/11] 现在我们支持了指令模型的 **[DPO 训练](https://arxiv.org/abs/2305.18290)**。详情请参阅[此示例](#dpo-training)（实验性功能）。
+
+[23/08/03] 现在我们支持了 **Qwen-7B** 模型的训练。请尝试使用 `--model_name_or_path Qwen/Qwen-7B-Chat` 和 `--lora_target c_attn` 参数。使用 Qwen-7B-Chat 模型请添加 `--template chatml` 参数。
 
 [23/07/31] 现在我们支持了训练数据流式加载。请尝试使用 `--streaming` 和 `--max_steps 100` 参数来流式加载数据集。
 
@@ -54,41 +56,34 @@
 | [Qwen](https://github.com/QwenLM/Qwen-7B)                | 7B                          | c_attn            | chatml   |
 | [XVERSE](https://github.com/xverse-ai/XVERSE-13B)        | 13B                         | q_proj,v_proj     | -        |
 
-> * **默认模块**是 `--lora_target` 参数的默认值。请使用 `python src/train_bash.py -h` 查看全部可选项。
-> * 对于所有“基座”模型，`--template` 参数可以是 `default`, `alpaca`, `vicuna` 等值。
+- **默认模块**是 `--lora_target` 参数的部分可选项。请使用 `python src/train_bash.py -h` 查看全部可选项。
+- 对于所有“基座”（Base）模型，`--template` 参数可以是 `default`, `alpaca`, `vicuna` 等任意值。但“对话”（Chat）模型请务必使用对应的模板。
 
-## 微调方法
+## 训练方法
 
-- [二次预训练](https://s3-us-west-2.amazonaws.com/openai-assets/research-covers/language-unsupervised/language_understanding_paper.pdf)
-  - 全参数微调
-  - 部分参数微调
-  - [LoRA](https://arxiv.org/abs/2106.09685)
-  - [QLoRA](https://arxiv.org/abs/2305.14314)
-- [指令监督微调](https://arxiv.org/abs/2109.01652)
-  - 全参数微调
-  - 部分参数微调
-  - [LoRA](https://arxiv.org/abs/2106.09685)
-  - [QLoRA](https://arxiv.org/abs/2305.14314)
-- [人类反馈的强化学习（RLHF）](https://arxiv.org/abs/2203.02155)
-  - [LoRA](https://arxiv.org/abs/2106.09685)
-  - [QLoRA](https://arxiv.org/abs/2305.14314)
+| 方法        | 全参数训练 | 部分参数训练 | LoRA | QLoRA |
+| ---------- | ---------- | ----------- | ---- | ----- |
+| 预训练      | ✅        | ✅         | ✅   | ✅   |
+| 指令监督微调 | ✅        | ✅         | ✅   | ✅   |
+| 奖励模型训练 |           |             | ✅   | ✅   |
+| PPO 训练    |           |             | ✅   | ✅   |
+| DPO 训练    | ✅        |             | ✅   | ✅   |
 
 ## 数据集
 
-- 用于二次预训练:
+- 用于预训练：
   - [Wiki Demo (en)](data/wiki_demo.txt)
   - [RefinedWeb (en)](https://huggingface.co/datasets/tiiuae/falcon-refinedweb)
   - [StarCoder (en)](https://huggingface.co/datasets/bigcode/starcoderdata)
   - [Wikipedia (en)](https://huggingface.co/datasets/olm/olm-wikipedia-20221220)
   - [Wikipedia (zh)](https://huggingface.co/datasets/pleisto/wikipedia-cn-20230720-filtered)
-- 用于指令监督微调:
+- 用于指令监督微调：
   - [Stanford Alpaca (en)](https://github.com/tatsu-lab/stanford_alpaca)
   - [Stanford Alpaca (zh)](https://github.com/ymcui/Chinese-LLaMA-Alpaca)
   - [GPT-4 Generated Data (en&zh)](https://github.com/Instruction-Tuning-with-GPT-4/GPT-4-LLM)
   - [Open Assistant (multilingual)](https://huggingface.co/datasets/OpenAssistant/oasst1)
   - [Self-cognition (zh)](data/self_cognition.json)
   - [ShareGPT (zh)](https://huggingface.co/datasets/QingyiSi/Alpaca-CoT/tree/main/Chinese-instruction-collection)
-  - [RefGPT (zh)](https://github.com/sufengniu/RefGPT)
   - [Guanaco Dataset (multilingual)](https://huggingface.co/datasets/JosephusCheung/GuanacoDataset)
   - [BELLE 2M (zh)](https://huggingface.co/datasets/BelleGroup/train_2M_CN)
   - [BELLE 1M (zh)](https://huggingface.co/datasets/BelleGroup/train_1M_CN)
@@ -103,7 +98,7 @@
   - [Web QA (zh)](https://huggingface.co/datasets/suolyer/webqa)
   - [UltraChat (en)](https://github.com/thunlp/UltraChat)
   - [WebNovel (zh)](https://huggingface.co/datasets/zxbsmk/webnovel_cn)
-- 用于奖励模型训练:
+- 用于奖励模型或 DPO 训练：
   - [HH-RLHF (en)](https://huggingface.co/datasets/Anthropic/hh-rlhf)
   - [Open Assistant (multilingual)](https://huggingface.co/datasets/OpenAssistant/oasst1)
   - [GPT-4 Generated Data (en&zh)](https://github.com/Instruction-Tuning-with-GPT-4/GPT-4-LLM)
@@ -139,7 +134,6 @@ huggingface-cli login
 ### 环境搭建（可跳过）
 
 ```bash
-git lfs install
 git clone https://github.com/hiyouga/LLaMA-Efficient-Tuning.git
 conda create -n llama_etuning python=3.10
 conda activate llama_etuning
@@ -161,7 +155,7 @@ CUDA_VISIBLE_DEVICES=0 python src/train_web.py
 
 目前网页 UI 仅支持**单卡训练**。
 
-### 二次预训练
+### 预训练
 
 ```bash
 CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
@@ -222,7 +216,7 @@ CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
     --resume_lora_training False \
     --checkpoint_dir path_to_sft_checkpoint \
     --output_dir path_to_rm_checkpoint \
-    --per_device_train_batch_size 4 \
+    --per_device_train_batch_size 2 \
     --gradient_accumulation_steps 4 \
     --lr_scheduler_type cosine \
     --logging_steps 10 \
@@ -233,7 +227,7 @@ CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
     --fp16
 ```
 
-### RLHF 训练
+### PPO 训练
 
 ```bash
 CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
@@ -257,6 +251,30 @@ CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
     --plot_loss
 ```
 
+### DPO 训练
+
+```bash
+CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
+    --stage dpo \
+    --model_name_or_path path_to_your_model \
+    --do_train \
+    --dataset comparison_gpt4_zh \
+    --template default \
+    --finetuning_type lora \
+    --resume_lora_training False \
+    --checkpoint_dir path_to_sft_checkpoint \
+    --output_dir path_to_dpo_checkpoint \
+    --per_device_train_batch_size 2 \
+    --gradient_accumulation_steps 4 \
+    --lr_scheduler_type cosine \
+    --logging_steps 10 \
+    --save_steps 1000 \
+    --learning_rate 1e-5 \
+    --num_train_epochs 1.0 \
+    --plot_loss \
+    --fp16
+```
+
 ### 多 GPU 分布式训练
 
 ```bash
diff --git a/data/dataset_info.json b/data/dataset_info.json
index 3a3b4e76..3eaf920e 100644
--- a/data/dataset_info.json
+++ b/data/dataset_info.json
@@ -49,26 +49,6 @@
       "history": "history"
     }
   },
-  "refgpt_zh_p1": {
-    "file_name": "refgpt_zh_50k_p1.json",
-    "file_sha1": "b40f4f4d0ffacd16da7c275b056d5b6670021752",
-    "columns": {
-      "prompt": "instruction",
-      "query": "input",
-      "response": "output",
-      "history": "history"
-    }
-  },
-  "refgpt_zh_p2": {
-    "file_name": "refgpt_zh_50k_p2.json",
-    "file_sha1": "181f32b2c60264a29f81f59d3c76095793eae1b0",
-    "columns": {
-      "prompt": "instruction",
-      "query": "input",
-      "response": "output",
-      "history": "history"
-    }
-  },
   "lima": {
     "file_name": "lima.json",
     "file_sha1": "9db59f6b7007dc4b17529fc63379b9cd61640f37",
diff --git a/data/refgpt_zh_50k_p1.json.REMOVED.git-id b/data/refgpt_zh_50k_p1.json.REMOVED.git-id
deleted file mode 100644
index 3e8a9e41..00000000
--- a/data/refgpt_zh_50k_p1.json.REMOVED.git-id
+++ /dev/null
@@ -1 +0,0 @@
-f967a4f6d04a11308a15524aa9a846a19a8d1e83
\ No newline at end of file
diff --git a/data/refgpt_zh_50k_p2.json.REMOVED.git-id b/data/refgpt_zh_50k_p2.json.REMOVED.git-id
deleted file mode 100644
index a6525b27..00000000
--- a/data/refgpt_zh_50k_p2.json.REMOVED.git-id
+++ /dev/null
@@ -1 +0,0 @@
-0a4f0d74fd1c5cab2eb6d84a3a3fe669847becd8
\ No newline at end of file
diff --git a/requirements.txt b/requirements.txt
index c71b6c9c..fb5fa72f 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -3,7 +3,7 @@ transformers>=4.29.1
 datasets>=2.12.0
 accelerate>=0.21.0
 peft>=0.4.0
-trl>=0.4.7
+trl>=0.5.0
 scipy
 sentencepiece
 tiktoken
@@ -16,4 +16,3 @@ pydantic==1.10.11
 fastapi==0.95.1
 sse-starlette
 matplotlib
-huggingface_hub
\ No newline at end of file
diff --git a/src/api_demo.py b/src/api_demo.py
index c0ca9760..777f9dcf 100644
--- a/src/api_demo.py
+++ b/src/api_demo.py
@@ -7,7 +7,7 @@ def main():
     chat_model = ChatModel()
     app = create_app(chat_model)
     uvicorn.run(app, host="0.0.0.0", port=8000, workers=1)
-    # Visit http://localhost:8000/docs for document.
+    print("Visit http://localhost:8000/docs for API document.")
 
 
 if __name__ == "__main__":
diff --git a/src/llmtuner/chat/stream_chat.py b/src/llmtuner/chat/stream_chat.py
index 79d3b92d..8220c0b3 100644
--- a/src/llmtuner/chat/stream_chat.py
+++ b/src/llmtuner/chat/stream_chat.py
@@ -18,7 +18,6 @@ class ChatModel:
         self.model = self.model.eval() # change to eval mode
         self.template = get_template_and_fix_tokenizer(data_args.template, self.tokenizer)
         self.source_prefix = data_args.source_prefix
-        self.stop_ids = self.tokenizer.convert_tokens_to_ids(self.template.stop_words)
         self.model.generate = MethodType(PreTrainedModel.generate, self.model) # disable custom method (for Qwen)
 
     def process_args(
@@ -53,7 +52,7 @@ class ChatModel:
             top_k=top_k or gen_kwargs["top_k"],
             repetition_penalty=repetition_penalty or gen_kwargs["repetition_penalty"],
             logits_processor=get_logits_processor(),
-            stopping_criteria=get_stopping_criteria(self.stop_ids)
+            stopping_criteria=get_stopping_criteria(self.tokenizer.additional_special_tokens_ids)
         ))
 
         if max_length:
diff --git a/src/llmtuner/dsets/preprocess.py b/src/llmtuner/dsets/preprocess.py
index 534d77b5..4efbcbb6 100644
--- a/src/llmtuner/dsets/preprocess.py
+++ b/src/llmtuner/dsets/preprocess.py
@@ -46,7 +46,6 @@ def preprocess_dataset(
             k: [t[i: i + block_size] for i in range(0, total_length, block_size)]
             for k, t in concatenated_examples.items()
         }
-        result["labels"] = result["input_ids"].copy()
         return result
 
     def preprocess_supervised_dataset(examples: Dict[str, List[Any]]) -> Dict[str, Any]:
@@ -95,24 +94,22 @@ def preprocess_dataset(
         return model_inputs
 
     def preprocess_pairwise_dataset(examples):
-        # build input pairs with format `<bos> X Y1 <eos>` and `<bos> X Y2 <eos>`
-        model_inputs = {"accept_ids": [], "reject_ids": []}
+        # build input pairs with format `<bos> X`, `Y1 <eos>` and `Y2 <eos>`
+        model_inputs = {"prompt_ids": [], "chosen_ids": [], "rejected_ids": []}
         for query, response, history, prefix in construct_example(examples):
-            source_ids, accept_ids = template.encode_oneturn(tokenizer, query, response[0], history, prefix)
-            source_ids, reject_ids = template.encode_oneturn(tokenizer, query, response[1], history, prefix)
+            prompt_ids, chosen_ids = template.encode_oneturn(tokenizer, query, response[0], history, prefix)
+            _, rejected_ids = template.encode_oneturn(tokenizer, query, response[1], history, prefix)
 
-            if len(source_ids) > data_args.max_source_length:
-                source_ids = source_ids[:data_args.max_source_length]
-            if len(accept_ids) > data_args.max_target_length:
-                accept_ids = accept_ids[:data_args.max_target_length]
-            if len(reject_ids) > data_args.max_target_length:
-                reject_ids = reject_ids[:data_args.max_target_length]
+            if len(prompt_ids) > data_args.max_source_length:
+                prompt_ids = prompt_ids[:data_args.max_source_length]
+            if len(chosen_ids) > data_args.max_target_length:
+                chosen_ids = chosen_ids[:data_args.max_target_length]
+            if len(rejected_ids) > data_args.max_target_length:
+                rejected_ids = rejected_ids[:data_args.max_target_length]
 
-            accept_ids = source_ids + accept_ids
-            reject_ids = source_ids + reject_ids
-
-            model_inputs["accept_ids"].append(accept_ids)
-            model_inputs["reject_ids"].append(reject_ids)
+            model_inputs["prompt_ids"].append(prompt_ids)
+            model_inputs["chosen_ids"].append(chosen_ids)
+            model_inputs["rejected_ids"].append(rejected_ids)
         return model_inputs
 
     def print_supervised_dataset_example(example):
@@ -124,10 +121,12 @@ def preprocess_dataset(
         ], skip_special_tokens=False)))
 
     def print_pairwise_dataset_example(example):
-        print("accept_ids:\n{}".format(example["accept_ids"]))
-        print("accepts:\n{}".format(tokenizer.decode(example["accept_ids"], skip_special_tokens=False)))
-        print("reject_ids:\n{}".format(example["reject_ids"]))
-        print("rejects:\n{}".format(tokenizer.decode(example["reject_ids"], skip_special_tokens=False)))
+        print("prompt_ids:\n{}".format(example["prompt_ids"]))
+        print("prompt:\n{}".format(tokenizer.decode(example["prompt_ids"], skip_special_tokens=False)))
+        print("chosen_ids:\n{}".format(example["chosen_ids"]))
+        print("chosen:\n{}".format(tokenizer.decode(example["chosen_ids"], skip_special_tokens=False)))
+        print("rejected_ids:\n{}".format(example["rejected_ids"]))
+        print("rejected:\n{}".format(tokenizer.decode(example["rejected_ids"], skip_special_tokens=False)))
 
     def print_unsupervised_dataset_example(example):
         print("input_ids:\n{}".format(example["input_ids"]))
diff --git a/src/llmtuner/extras/callbacks.py b/src/llmtuner/extras/callbacks.py
index d325b0a8..61deae25 100644
--- a/src/llmtuner/extras/callbacks.py
+++ b/src/llmtuner/extras/callbacks.py
@@ -7,10 +7,16 @@ from datetime import timedelta
 from transformers import TrainerCallback
 from transformers.trainer_utils import has_length
 
+from llmtuner.extras.constants import LOG_FILE_NAME
+from llmtuner.extras.logging import get_logger
+
 if TYPE_CHECKING:
     from transformers import TrainingArguments, TrainerState, TrainerControl
 
 
+logger = get_logger(__name__)
+
+
 class LogCallback(TrainerCallback):
 
     def __init__(self, runner=None):
@@ -38,6 +44,9 @@ class LogCallback(TrainerCallback):
             self.in_training = True
             self.start_time = time.time()
             self.max_steps = state.max_steps
+            if os.path.exists(os.path.join(args.output_dir, LOG_FILE_NAME)):
+                logger.warning("Previous log file in this folder will be deleted.")
+                os.remove(os.path.join(args.output_dir, LOG_FILE_NAME))
 
     def on_train_end(self, args: "TrainingArguments", state: "TrainerState", control: "TrainerControl", **kwargs):
         r"""
diff --git a/src/llmtuner/extras/constants.py b/src/llmtuner/extras/constants.py
index 6f6dbdd7..dc1cb2e7 100644
--- a/src/llmtuner/extras/constants.py
+++ b/src/llmtuner/extras/constants.py
@@ -1,10 +1,12 @@
 IGNORE_INDEX = -100
 
+LOG_FILE_NAME = "trainer_log.jsonl"
+
 VALUE_HEAD_FILE_NAME = "value_head.bin"
 
 FINETUNING_ARGS_NAME = "finetuning_args.json"
 
-LAYERNORM_NAMES = ["norm", "ln_f", "ln_attn", "ln_mlp"] # for LLaMA, BLOOM and Falcon settings
+LAYERNORM_NAMES = ["norm", "ln_f", "ln_attn", "ln_mlp"]
 
 METHODS = ["full", "freeze", "lora"]
 
diff --git a/src/llmtuner/extras/misc.py b/src/llmtuner/extras/misc.py
index e1fbb156..ee918fbb 100644
--- a/src/llmtuner/extras/misc.py
+++ b/src/llmtuner/extras/misc.py
@@ -1,7 +1,11 @@
 import torch
 from typing import TYPE_CHECKING, List, Optional, Tuple
-
-from transformers import LogitsProcessor, LogitsProcessorList, StoppingCriteria, StoppingCriteriaList
+from transformers import (
+    LogitsProcessor,
+    LogitsProcessorList,
+    StoppingCriteria,
+    StoppingCriteriaList
+)
 
 from llmtuner.extras.constants import LAYERNORM_NAMES
 
diff --git a/src/llmtuner/extras/template.py b/src/llmtuner/extras/template.py
index d94d5b37..c3388ab3 100644
--- a/src/llmtuner/extras/template.py
+++ b/src/llmtuner/extras/template.py
@@ -61,7 +61,7 @@ class Template:
         prefix: Optional[str] = None
     ) -> Tuple[List[Union[str, Dict[str, str]]], List[Tuple[str, str]]]:
         r"""
-        Aligns inputs to a special format.
+        Aligns inputs to the standard format.
         """
         prefix = [prefix] if prefix else self.prefix # use prefix if provided
         history = history if (history and self.use_history) else []
@@ -92,28 +92,32 @@ class Template:
     ) -> List[Tuple[List[int], List[int]]]:
         r"""
         Encodes formatted inputs to pairs of token ids.
+        Turn 0: bos + prefix + sep + query    resp + eos
+        Turn t: sep + bos + query             resp + eos
         """
         bos_ids, eos_ids = self._get_special_ids(tokenizer)
         sep_ids = self._convert_inputs_to_ids(tokenizer, context=self.sep)
         encoded_pairs = []
         for turn_idx, (query, resp) in enumerate(history):
-            if turn_idx != 0:
-                prefix_ids = sep_ids
-            elif prefix:
-                prefix_ids = self._convert_inputs_to_ids(tokenizer, context=prefix) + eos_ids + sep_ids
+            if turn_idx == 0:
+                if prefix: # has prefix
+                    prefix_ids = bos_ids + self._convert_inputs_to_ids(tokenizer, context=prefix) + sep_ids
+                else:
+                    prefix_ids = bos_ids
             else:
-                prefix_ids = []
+                prefix_ids = sep_ids + bos_ids
 
-            query_ids = self._convert_inputs_to_ids(tokenizer, context=self.prompt, query=query)
+            query_ids = self._convert_inputs_to_ids(tokenizer, context=self.prompt, query=query, idx=str(turn_idx))
             resp_ids = self._convert_inputs_to_ids(tokenizer, context=[resp])
-            encoded_pairs.append((bos_ids + prefix_ids + query_ids, resp_ids + eos_ids))
+            encoded_pairs.append((prefix_ids + query_ids, resp_ids + eos_ids))
         return encoded_pairs
 
     def _convert_inputs_to_ids(
         self,
         tokenizer: "PreTrainedTokenizer",
         context: List[Union[str, Dict[str, str]]],
-        query: Optional[str] = ""
+        query: Optional[str] = "",
+        idx: Optional[str] = ""
     ) -> List[int]:
         r"""
         Converts context to token ids.
@@ -127,6 +131,7 @@ class Template:
         for elem in context:
             if isinstance(elem, str):
                 elem = elem.replace("{{query}}", query, 1)
+                elem = elem.replace("{{idx}}", idx, 1)
                 token_ids = token_ids + tokenizer.encode(elem, **kwargs)
             elif isinstance(elem, dict):
                 token_ids = token_ids + [tokenizer.convert_tokens_to_ids(elem.get("token"))]
@@ -146,10 +151,12 @@ class Llama2Template(Template):
     ) -> List[Tuple[List[int], List[int]]]:
         r"""
         Encodes formatted inputs to pairs of token ids.
+        Turn 0: bos + prefix + query    resp + eos
+        Turn t: bos + query             resp + eos
         """
         bos_ids, eos_ids = self._get_special_ids(tokenizer)
         encoded_pairs = []
-        assert isinstance(prefix[0], str), "LLaMA-2 template only accepts list containing a single str."
+        assert isinstance(prefix[0], str), "LLaMA-2 template only accepts list containing a single string."
         for turn_idx, (query, resp) in enumerate(history):
             if turn_idx == 0: # llama2 template has not sep_ids
                 query = prefix[0] + query
@@ -187,11 +194,12 @@ def get_template_and_fix_tokenizer(
     template = templates.get(name, None)
     assert template is not None, "Template {} does not exist.".format(name)
 
-    if tokenizer.eos_token_id is None: # inplace method
-        if len(template.stop_words):
-            tokenizer.eos_token = template.stop_words[0]
-        else:
-            tokenizer.eos_token = "<|endoftext|>"
+    if len(template.stop_words): # inplace method
+        tokenizer.eos_token = template.stop_words[0]
+        logger.info("Replace eos token: {}".format(tokenizer.eos_token))
+
+    if tokenizer.eos_token_id is None:
+        tokenizer.eos_token = "<|endoftext|>"
         logger.info("Add eos token: {}".format(tokenizer.eos_token))
 
     if tokenizer.pad_token_id is None:
@@ -422,12 +430,13 @@ register_template(
     name="baichuan",
     prefix=[],
     prompt=[
-        {"token": "<reserved_102>"},
-        "{{query}}",
-        {"token": "<reserved_103>"}
+        {"token": "<reserved_102>"}, # user token (a little difference in position)
+        "{{query}}"
     ],
     sep=[],
-    stop_words=[],
+    stop_words=[
+        "<reserved_103>" # assistant token
+    ],
     use_history=True
 )
 
@@ -440,7 +449,8 @@ register_template(
     name="starchat",
     prefix=[
         {"token": "<|system|>"},
-        "\n"
+        "\n",
+        {"token": "<|end|>"}
     ],
     prompt=[
         {"token": "<|user|>"},
@@ -466,7 +476,8 @@ register_template(
     name="chatml",
     prefix=[
         {"token": "<|im_start|>"},
-        "system\nYou are a helpful assistant."
+        "system\nYou are a helpful assistant.",
+        {"token": "<|im_end|>"}
     ],
     prompt=[
         {"token": "<|im_start|>"},
@@ -484,3 +495,23 @@ register_template(
     ],
     use_history=True
 )
+
+
+r"""
+Supports: https://huggingface.co/THUDM/chatglm2-6b
+"""
+register_template(
+    name="chatglm2",
+    prefix=[
+        {"token": "[gMASK]"},
+        {"token": "sop"}
+    ],
+    prompt=[
+        "[Round {{idx}}]\n\n问：{{query}}\n\n答："
+    ],
+    sep=[
+        "\n\n"
+    ],
+    stop_words=[],
+    use_history=True
+)
diff --git a/src/llmtuner/hparams/data_args.py b/src/llmtuner/hparams/data_args.py
index de470ae2..7d1c982c 100644
--- a/src/llmtuner/hparams/data_args.py
+++ b/src/llmtuner/hparams/data_args.py
@@ -24,7 +24,7 @@ class DatasetAttr:
 
 @dataclass
 class DataArguments:
-    """
+    r"""
     Arguments pertaining to what data we are going to input our model for training and evaluation.
     """
     template: str = field(
diff --git a/src/llmtuner/hparams/finetuning_args.py b/src/llmtuner/hparams/finetuning_args.py
index 277602ae..c4713c5e 100644
--- a/src/llmtuner/hparams/finetuning_args.py
+++ b/src/llmtuner/hparams/finetuning_args.py
@@ -5,7 +5,7 @@ from dataclasses import asdict, dataclass, field
 
 @dataclass
 class FinetuningArguments:
-    """
+    r"""
     Arguments pertaining to which techniques we are going to fine-tuning with.
     """
     finetuning_type: Optional[Literal["none", "freeze", "lora", "full"]] = field(
@@ -14,7 +14,7 @@ class FinetuningArguments:
     )
     num_hidden_layers: Optional[int] = field(
         default=32,
-        metadata={"help": "Number of decoder blocks in the model. \
+        metadata={"help": "Number of decoder blocks in the model for partial-parameter (freeze) fine-tuning. \
                   LLaMA choices: [\"32\", \"40\", \"60\", \"80\"], \
                   LLaMA-2 choices: [\"32\", \"40\", \"80\"], \
                   BLOOM choices: [\"24\", \"30\", \"70\"], \
@@ -25,16 +25,16 @@ class FinetuningArguments:
     )
     num_layer_trainable: Optional[int] = field(
         default=3,
-        metadata={"help": "Number of trainable layers for Freeze fine-tuning."}
+        metadata={"help": "Number of trainable layers for partial-parameter (freeze) fine-tuning."}
     )
     name_module_trainable: Optional[Literal["mlp", "self_attn", "self_attention"]] = field(
         default="mlp",
-        metadata={"help": "Name of trainable modules for Freeze fine-tuning. \
-                  LLaMA & LLaMA-2 choices: [\"mlp\", \"self_attn\"], \
+        metadata={"help": "Name of trainable modules for partial-parameter (freeze) fine-tuning. \
+                  LLaMA choices: [\"mlp\", \"self_attn\"], \
                   BLOOM & Falcon choices: [\"mlp\", \"self_attention\"], \
                   Baichuan choices: [\"mlp\", \"self_attn\"], \
                   Qwen choices: [\"mlp\", \"attn\"], \
-                  InternLM, XVERSE choices: the same as LLaMA."}
+                  LLaMA-2, InternLM, XVERSE choices: the same as LLaMA."}
     )
     lora_rank: Optional[int] = field(
         default=8,
@@ -51,11 +51,15 @@ class FinetuningArguments:
     lora_target: Optional[str] = field(
         default="q_proj,v_proj",
         metadata={"help": "Name(s) of target modules to apply LoRA. Use commas to separate multiple modules. \
-                  LLaMA & LLaMA-2 choices: [\"q_proj\", \"k_proj\", \"v_proj\", \"o_proj\", \"gate_proj\", \"up_proj\", \"down_proj\"], \
+                  LLaMA choices: [\"q_proj\", \"k_proj\", \"v_proj\", \"o_proj\", \"gate_proj\", \"up_proj\", \"down_proj\"], \
                   BLOOM & Falcon choices: [\"query_key_value\", \"self_attention.dense\", \"mlp.dense\"], \
                   Baichuan choices: [\"W_pack\", \"o_proj\", \"gate_proj\", \"up_proj\", \"down_proj\"], \
                   Qwen choices: [\"c_attn\", \"attn.c_proj\", \"w1\", \"w2\", \"mlp.c_proj\"], \
-                  InternLM, XVERSE choices: the same as LLaMA."}
+                  LLaMA-2, InternLM, XVERSE choices: the same as LLaMA."}
+    )
+    dpo_beta: Optional[float] = field(
+        default=0.1,
+        metadata={"help": "The beta parameter for the DPO loss."}
     )
 
     def __post_init__(self):
@@ -72,14 +76,14 @@ class FinetuningArguments:
         assert self.finetuning_type in ["none", "freeze", "lora", "full"], "Invalid fine-tuning method."
 
     def save_to_json(self, json_path: str):
-        """Saves the content of this instance in JSON format inside `json_path`."""
+        r"""Saves the content of this instance in JSON format inside `json_path`."""
         json_string = json.dumps(asdict(self), indent=2, sort_keys=True) + "\n"
         with open(json_path, "w", encoding="utf-8") as f:
             f.write(json_string)
 
     @classmethod
     def load_from_json(cls, json_path: str):
-        """Creates an instance from the content of `json_path`."""
+        r"""Creates an instance from the content of `json_path`."""
         with open(json_path, "r", encoding="utf-8") as f:
             text = f.read()
         return cls(**json.loads(text))
diff --git a/src/llmtuner/hparams/general_args.py b/src/llmtuner/hparams/general_args.py
index 397d3019..c0c1a0de 100644
--- a/src/llmtuner/hparams/general_args.py
+++ b/src/llmtuner/hparams/general_args.py
@@ -4,10 +4,10 @@ from dataclasses import dataclass, field
 
 @dataclass
 class GeneralArguments:
-    """
+    r"""
     Arguments pertaining to which stage we are going to perform.
     """
-    stage: Optional[Literal["pt", "sft", "rm", "ppo"]] = field(
+    stage: Optional[Literal["pt", "sft", "rm", "ppo", "dpo"]] = field(
         default="sft",
         metadata={"help": "Which stage will be performed in training."}
     )
diff --git a/src/llmtuner/hparams/generating_args.py b/src/llmtuner/hparams/generating_args.py
index e25ff4b9..f8b935fb 100644
--- a/src/llmtuner/hparams/generating_args.py
+++ b/src/llmtuner/hparams/generating_args.py
@@ -4,7 +4,7 @@ from dataclasses import asdict, dataclass, field
 
 @dataclass
 class GeneratingArguments:
-    """
+    r"""
     Arguments pertaining to specify the decoding parameters.
     """
     do_sample: Optional[bool] = field(
diff --git a/src/llmtuner/hparams/model_args.py b/src/llmtuner/hparams/model_args.py
index 94e882fe..6c8491da 100644
--- a/src/llmtuner/hparams/model_args.py
+++ b/src/llmtuner/hparams/model_args.py
@@ -1,12 +1,11 @@
 import torch
 from typing import Literal, Optional
 from dataclasses import dataclass, field
-from huggingface_hub.hf_api import HfFolder
 
 
 @dataclass
 class ModelArguments:
-    """
+    r"""
     Arguments pertaining to which model/config/tokenizer we are going to fine-tune.
     """
     model_name_or_path: str = field(
@@ -64,12 +63,11 @@ class ModelArguments:
         default=False,
         metadata={"help": "Whether to plot the training loss after fine-tuning or not."}
     )
-    hf_hub_token  : Optional[str] = field(
+    hf_auth_token: Optional[str] = field(
         default=None,
-        metadata={"help": "Path to the directory containing the checkpoints of the reward model."}
+        metadata={"help": "Auth token to log in with Hugging Face Hub."}
     )
 
-
     def __post_init__(self):
         if self.checkpoint_dir is not None: # support merging multiple lora weights
             self.checkpoint_dir = [cd.strip() for cd in self.checkpoint_dir.split(",")]
@@ -77,5 +75,6 @@ class ModelArguments:
         if self.quantization_bit is not None:
             assert self.quantization_bit in [4, 8], "We only accept 4-bit or 8-bit quantization."
 
-        if self.use_auth_token == True and self.hf_hub_token != None:
-            HfFolder.save_token(self.hf_hub_token)
+        if self.use_auth_token == True and self.hf_auth_token is not None:
+            from huggingface_hub.hf_api import HfFolder # lazy load
+            HfFolder.save_token(self.hf_auth_token)
diff --git a/src/llmtuner/tuner/core/adapter.py b/src/llmtuner/tuner/core/adapter.py
index 4afad13a..a8ac5a84 100644
--- a/src/llmtuner/tuner/core/adapter.py
+++ b/src/llmtuner/tuner/core/adapter.py
@@ -39,7 +39,7 @@ def init_adapter(
     if finetuning_args.finetuning_type == "none" and is_trainable:
         raise ValueError("You cannot use finetuning_type=none while training.")
 
-    if finetuning_args.finetuning_type == "full":
+    if finetuning_args.finetuning_type == "full" and is_trainable:
         logger.info("Fine-tuning method: Full")
         model = model.float()
 
diff --git a/src/llmtuner/tuner/core/loader.py b/src/llmtuner/tuner/core/loader.py
index c06eabfa..39bec1d8 100644
--- a/src/llmtuner/tuner/core/loader.py
+++ b/src/llmtuner/tuner/core/loader.py
@@ -34,7 +34,7 @@ check_min_version("4.29.1")
 require_version("datasets>=2.12.0", "To fix: pip install datasets>=2.12.0")
 require_version("accelerate>=0.21.0", "To fix: pip install accelerate>=0.21.0")
 require_version("peft>=0.4.0", "To fix: pip install peft>=0.4.0")
-require_version("trl>=0.4.7", "To fix: pip install trl>=0.4.7")
+require_version("trl>=0.5.0", "To fix: pip install trl>=0.5.0")
 
 
 def load_model_and_tokenizer(
@@ -52,9 +52,6 @@ def load_model_and_tokenizer(
         logger.warning("Checkpoint is not found at evaluation, load the original model.")
         finetuning_args = FinetuningArguments(finetuning_type="none")
 
-    assert stage in ["pt", "sft"] or finetuning_args.finetuning_type == "lora", \
-        "RM and PPO training can only be performed with the LoRA method."
-
     config_kwargs = {
         "trust_remote_code": True,
         "cache_dir": model_args.cache_dir,
@@ -132,8 +129,6 @@ def load_model_and_tokenizer(
                 })
 
         if stage == "ppo": # load reward model
-            assert is_trainable, "PPO stage cannot be performed at evaluation."
-            assert model_args.reward_model is not None, "Reward model is necessary for PPO training."
             logger.info("Load reward model from {}".format(model_args.reward_model))
             model.pretrained_model.load_adapter(model_args.reward_model, "reward", is_trainable=False)
             assert load_valuehead_params(model, model_args.reward_model), "Reward model is not correctly loaded."
diff --git a/src/llmtuner/tuner/core/parser.py b/src/llmtuner/tuner/core/parser.py
index 692f9b13..c5bdbe16 100644
--- a/src/llmtuner/tuner/core/parser.py
+++ b/src/llmtuner/tuner/core/parser.py
@@ -19,7 +19,7 @@ from llmtuner.hparams import (
 logger = get_logger(__name__)
 
 
-def _parse_args(parser: HfArgumentParser, args: Optional[Dict[str, Any]] = None):
+def _parse_args(parser: HfArgumentParser, args: Optional[Dict[str, Any]] = None) -> Tuple[Any]:
     if args is not None:
         return parser.parse_dict(args)
     elif len(sys.argv) == 2 and sys.argv[1].endswith(".yaml"):
@@ -32,26 +32,53 @@ def _parse_args(parser: HfArgumentParser, args: Optional[Dict[str, Any]] = None)
 
 def parse_train_args(
     args: Optional[Dict[str, Any]] = None
-) -> Tuple[ModelArguments, DataArguments, Seq2SeqTrainingArguments, FinetuningArguments, GeneralArguments]:
+) -> Tuple[
+    ModelArguments,
+    DataArguments,
+    Seq2SeqTrainingArguments,
+    FinetuningArguments,
+    GeneratingArguments,
+    GeneralArguments
+]:
     parser = HfArgumentParser((
-        ModelArguments, DataArguments, Seq2SeqTrainingArguments, FinetuningArguments, GeneralArguments
+        ModelArguments,
+        DataArguments,
+        Seq2SeqTrainingArguments,
+        FinetuningArguments,
+        GeneratingArguments,
+        GeneralArguments
     ))
     return _parse_args(parser, args)
 
 
 def parse_infer_args(
     args: Optional[Dict[str, Any]] = None
-) -> Tuple[ModelArguments, DataArguments, FinetuningArguments, GeneratingArguments]:
+) -> Tuple[
+    ModelArguments,
+    DataArguments,
+    FinetuningArguments,
+    GeneratingArguments
+]:
     parser = HfArgumentParser((
-        ModelArguments, DataArguments, FinetuningArguments, GeneratingArguments
+        ModelArguments,
+        DataArguments,
+        FinetuningArguments,
+        GeneratingArguments
     ))
     return _parse_args(parser, args)
 
 
 def get_train_args(
     args: Optional[Dict[str, Any]] = None
-) -> Tuple[ModelArguments, DataArguments, Seq2SeqTrainingArguments, FinetuningArguments, GeneralArguments]:
-    model_args, data_args, training_args, finetuning_args, general_args = parse_train_args(args)
+) -> Tuple[
+    ModelArguments,
+    DataArguments,
+    Seq2SeqTrainingArguments,
+    FinetuningArguments,
+    GeneratingArguments,
+    GeneralArguments
+]:
+    model_args, data_args, training_args, finetuning_args, generating_args, general_args = parse_train_args(args)
 
     # Setup logging
     if training_args.should_log:
@@ -68,7 +95,7 @@ def get_train_args(
     data_args.init_for_training()
 
     if general_args.stage != "sft" and training_args.predict_with_generate:
-        raise ValueError("`predict_with_generate` cannot be set as True at PT, RM and PPO stages.")
+        raise ValueError("`predict_with_generate` cannot be set as True except SFT.")
 
     if training_args.do_train and training_args.predict_with_generate:
         raise ValueError("`predict_with_generate` cannot be set as True while training.")
@@ -76,6 +103,15 @@ def get_train_args(
     if general_args.stage == "sft" and training_args.do_predict and not training_args.predict_with_generate:
         raise ValueError("Please enable `predict_with_generate` to save model predictions.")
 
+    if general_args.stage in ["rm", "ppo"] and finetuning_args.finetuning_type != "lora":
+        raise ValueError("RM and PPO training can only be performed with the LoRA method.")
+
+    if general_args.stage in ["ppo", "dpo"] and not training_args.do_train:
+        raise ValueError("PPO and DPO stage can only be performed at training.")
+
+    if general_args.stage == "ppo" and model_args.reward_model is None:
+        raise ValueError("Reward model is necessary for PPO training.")
+
     if training_args.max_steps == -1 and data_args.streaming:
         raise ValueError("Please specify `max_steps` in streaming mode.")
 
@@ -133,12 +169,17 @@ def get_train_args(
     # Set seed before initializing model.
     transformers.set_seed(training_args.seed)
 
-    return model_args, data_args, training_args, finetuning_args, general_args
+    return model_args, data_args, training_args, finetuning_args, generating_args, general_args
 
 
 def get_infer_args(
     args: Optional[Dict[str, Any]] = None
-) -> Tuple[ModelArguments, DataArguments, FinetuningArguments, GeneratingArguments]:
+) -> Tuple[
+    ModelArguments,
+    DataArguments,
+    FinetuningArguments,
+    GeneratingArguments
+]:
     model_args, data_args, finetuning_args, generating_args = parse_infer_args(args)
 
     if model_args.quantization_bit is not None and finetuning_args.finetuning_type != "lora":
diff --git a/src/llmtuner/tuner/core/trainer.py b/src/llmtuner/tuner/core/trainer.py
index ae80f32f..058bb740 100644
--- a/src/llmtuner/tuner/core/trainer.py
+++ b/src/llmtuner/tuner/core/trainer.py
@@ -13,26 +13,25 @@ from llmtuner.extras.logging import get_logger
 from llmtuner.extras.save_and_load import get_state_dict, load_trainable_params
 
 if TYPE_CHECKING:
+    from transformers import PreTrainedTokenizer, Seq2SeqTrainingArguments, TrainerState
     from llmtuner.hparams import FinetuningArguments
 
 
 logger = get_logger(__name__)
 
 
-class PeftTrainer(Seq2SeqTrainer):
+class PeftModelMixin:
     r"""
-    Inherits Seq2SeqTrainer to support parameter-efficient checkpoints.
+    Patches the save and load methods in Hugging Face Trainer for PeftModel and ModelWithValueHead.
     """
 
-    def __init__(self, finetuning_args: "FinetuningArguments", **kwargs):
-        super().__init__(**kwargs)
-        self.finetuning_args = finetuning_args
-        self._remove_log()
-
-    def _remove_log(self):
-        if self.is_world_process_zero() and os.path.exists(os.path.join(self.args.output_dir, "trainer_log.jsonl")):
-            logger.warning("Previous log file in this folder will be deleted.")
-            os.remove(os.path.join(self.args.output_dir, "trainer_log.jsonl"))
+    def __init__(self) -> None: # for type checking
+        self.model: PreTrainedModel = None
+        self.tokenizer: "PreTrainedTokenizer" = None
+        self.args: "Seq2SeqTrainingArguments" = None
+        self.finetuning_args: "FinetuningArguments" = None
+        self.state: "TrainerState" = None
+        raise AssertionError("Mixin should not be initialized.")
 
     def _save(self, output_dir: Optional[str] = None, state_dict: Optional[Dict[str, torch.Tensor]] = None) -> None:
         r"""
@@ -96,3 +95,13 @@ class PeftTrainer(Seq2SeqTrainer):
             model.load_adapter(self.state.best_model_checkpoint, model.active_adapter)
         else: # freeze/full-tuning
             load_trainable_params(model, self.state.best_model_checkpoint)
+
+
+class PeftTrainer(PeftModelMixin, Seq2SeqTrainer):
+    r"""
+    Inherits Seq2SeqTrainer to support parameter-efficient checkpoints.
+    """
+
+    def __init__(self, finetuning_args: "FinetuningArguments", **kwargs):
+        Seq2SeqTrainer.__init__(self, **kwargs)
+        self.finetuning_args = finetuning_args
diff --git a/src/llmtuner/tuner/dpo/__init__.py b/src/llmtuner/tuner/dpo/__init__.py
new file mode 100644
index 00000000..f2b5cfb5
--- /dev/null
+++ b/src/llmtuner/tuner/dpo/__init__.py
@@ -0,0 +1 @@
+from llmtuner.tuner.dpo.workflow import run_dpo
diff --git a/src/llmtuner/tuner/dpo/collator.py b/src/llmtuner/tuner/dpo/collator.py
new file mode 100644
index 00000000..2f0f4bdc
--- /dev/null
+++ b/src/llmtuner/tuner/dpo/collator.py
@@ -0,0 +1,51 @@
+import torch
+from dataclasses import dataclass
+from typing import Any, Dict, List, Sequence, Tuple
+from transformers import DataCollatorForSeq2Seq
+
+
+@dataclass
+class DPODataCollatorWithPadding(DataCollatorForSeq2Seq):
+    r"""
+    Data collator for pairwise data.
+    """
+
+    def _pad_labels(self, batch: torch.Tensor, positions: List[Tuple[int, int]]) -> torch.Tensor:
+        padded_labels = []
+        for feature, (prompt_len, answer_len) in zip(batch, positions):
+            if self.tokenizer.padding_side == "left":
+                start, end = feature.size(0) - answer_len, feature.size(0)
+            else:
+                start, end = prompt_len, answer_len
+            padded_tensor = self.label_pad_token_id * torch.ones_like(feature)
+            padded_tensor[start:end] = feature[start:end]
+            padded_labels.append(padded_tensor)
+        return torch.stack(padded_labels, dim=0).contiguous() # in contiguous memory
+
+    def __call__(self, features: Sequence[Dict[str, Any]]) -> Dict[str, torch.Tensor]:
+        r"""
+        Pads batched data to the longest sequence in the batch.
+
+        We generate 2 * n examples where the first n examples represent chosen examples and
+        the last n examples represent rejected examples.
+        """
+        concatenated_features = []
+        label_positions = []
+        for key in ("chosen_ids", "rejected_ids"):
+            for feature in features:
+                prompt_len, answer_len = len(feature["prompt_ids"]), len(feature[key])
+                concatenated_features.append({
+                    "input_ids": feature["prompt_ids"] + feature[key],
+                    "attention_mask": [1] * (prompt_len + answer_len)
+                })
+                label_positions.append((prompt_len, answer_len))
+
+        batch = self.tokenizer.pad(
+            concatenated_features,
+            padding=self.padding,
+            max_length=self.max_length,
+            pad_to_multiple_of=self.pad_to_multiple_of,
+            return_tensors=self.return_tensors,
+        )
+        batch["labels"] = self._pad_labels(batch["input_ids"], label_positions)
+        return batch
diff --git a/src/llmtuner/tuner/dpo/trainer.py b/src/llmtuner/tuner/dpo/trainer.py
new file mode 100644
index 00000000..a94642c1
--- /dev/null
+++ b/src/llmtuner/tuner/dpo/trainer.py
@@ -0,0 +1,75 @@
+import torch
+from collections import defaultdict
+from peft import PeftModel
+from typing import TYPE_CHECKING, Dict, Optional, Tuple, Union
+from transformers import Trainer
+from trl import DPOTrainer
+
+from llmtuner.extras.constants import IGNORE_INDEX
+from llmtuner.tuner.core.trainer import PeftModelMixin
+
+if TYPE_CHECKING:
+    from transformers import PreTrainedModel
+    from llmtuner.hparams import FinetuningArguments, GeneratingArguments
+
+
+class DPOPeftTrainer(PeftModelMixin, DPOTrainer):
+
+    def __init__(
+        self,
+        finetuning_args: "FinetuningArguments",
+        generating_args: "GeneratingArguments",
+        ref_model: Optional[Union["PreTrainedModel", torch.nn.Module]] = None,
+        **kwargs
+    ):
+        self.finetuning_args = finetuning_args
+        self.generating_args = generating_args
+        self.ref_model = ref_model
+        self.use_dpo_data_collator = True # hack to avoid warning
+        self.label_pad_token_id = IGNORE_INDEX
+        self.padding_value = 0
+        self.beta = finetuning_args.dpo_beta
+        self._stored_metrics = defaultdict(lambda: defaultdict(list))
+
+        Trainer.__init__(self, **kwargs)
+        if ref_model is not None:
+            if hasattr(self, "accelerator"):
+                self.ref_model = self.accelerator.prepare_model(self.ref_model, evaluation_mode=True)
+            else:
+                raise AttributeError("Please update `transformers`.")
+
+    def concatenated_forward(
+        self,
+        model: Optional[torch.nn.Module] = None,
+        batch: Optional[Dict[str, torch.Tensor]] = None
+    ) -> Tuple[torch.FloatTensor, torch.FloatTensor, torch.FloatTensor, torch.FloatTensor]:
+        unwrapped_model: "PreTrainedModel" = self.accelerator.unwrap_model(self.model)
+        if not torch.is_grad_enabled():
+            unwrapped_model.gradient_checkpointing_disable()
+
+        if model is None and isinstance(unwrapped_model, PeftModel): # peft model has no ref_model
+            with unwrapped_model.disable_adapter():
+                all_logits: torch.Tensor = self.model(
+                    batch["input_ids"],
+                    attention_mask=batch["attention_mask"],
+                    return_dict=True
+                ).logits.to(torch.float32)
+        else:
+            all_logits: torch.Tensor = model(
+                batch["input_ids"],
+                attention_mask=batch["attention_mask"],
+                return_dict=True
+            ).logits.to(torch.float32)
+
+        if not torch.is_grad_enabled():
+            unwrapped_model.gradient_checkpointing_enable()
+
+        all_logps = self._get_batch_logps(
+            all_logits,
+            batch["labels"],
+            average_log_prob=False
+        )
+        batch_size = batch["input_ids"].size(0) // 2
+        chosen_logps, rejected_logps = all_logps.split(batch_size, dim=0)
+        chosen_logits, rejected_logits = all_logits.split(batch_size, dim=0)
+        return chosen_logps, rejected_logps, chosen_logits, rejected_logits
diff --git a/src/llmtuner/tuner/dpo/workflow.py b/src/llmtuner/tuner/dpo/workflow.py
new file mode 100644
index 00000000..350d64a7
--- /dev/null
+++ b/src/llmtuner/tuner/dpo/workflow.py
@@ -0,0 +1,59 @@
+# Inspired by: https://github.com/huggingface/trl/blob/main/examples/research_projects/stack_llama_2/scripts/dpo_llama2.py
+
+from copy import deepcopy
+from peft import PeftModel
+from typing import TYPE_CHECKING, Optional, List
+
+from llmtuner.dsets import get_dataset, preprocess_dataset, split_dataset
+from llmtuner.extras.constants import IGNORE_INDEX
+from llmtuner.extras.ploting import plot_loss
+from llmtuner.tuner.core import load_model_and_tokenizer
+from llmtuner.tuner.dpo.collator import DPODataCollatorWithPadding
+from llmtuner.tuner.dpo.trainer import DPOPeftTrainer
+
+if TYPE_CHECKING:
+    from transformers import Seq2SeqTrainingArguments, TrainerCallback
+    from llmtuner.hparams import ModelArguments, DataArguments, FinetuningArguments, GeneratingArguments
+
+
+def run_dpo(
+    model_args: "ModelArguments",
+    data_args: "DataArguments",
+    training_args: "Seq2SeqTrainingArguments",
+    finetuning_args: "FinetuningArguments",
+    generating_args: "GeneratingArguments",
+    callbacks: Optional[List["TrainerCallback"]] = None
+):
+    dataset = get_dataset(model_args, data_args)
+    model, tokenizer = load_model_and_tokenizer(model_args, finetuning_args, training_args.do_train, stage="sft")
+    dataset = preprocess_dataset(dataset, tokenizer, data_args, training_args, stage="rm")
+    data_collator = DPODataCollatorWithPadding(
+        tokenizer=tokenizer,
+        label_pad_token_id=IGNORE_INDEX if data_args.ignore_pad_token_for_loss else tokenizer.pad_token_id
+    )
+
+    training_args.remove_unused_columns = False # important for pairwise dataset
+    ref_model = deepcopy(model) if not isinstance(model, PeftModel) else None
+
+    # Initialize our Trainer
+    trainer = DPOPeftTrainer(
+        finetuning_args=finetuning_args,
+        generating_args=generating_args,
+        ref_model=ref_model,
+        model=model,
+        args=training_args,
+        tokenizer=tokenizer,
+        data_collator=data_collator,
+        callbacks=callbacks,
+        **split_dataset(dataset, data_args, training_args)
+    )
+
+    # Training
+    if training_args.do_train:
+        train_result = trainer.train()
+        trainer.log_metrics("train", train_result.metrics)
+        trainer.save_metrics("train", train_result.metrics)
+        trainer.save_state()
+        trainer.save_model()
+        if trainer.is_world_process_zero() and model_args.plot_loss:
+            plot_loss(training_args.output_dir, keys=["loss", "eval_loss"])
diff --git a/src/llmtuner/tuner/ppo/trainer.py b/src/llmtuner/tuner/ppo/trainer.py
index 3392aa4d..26a8db99 100644
--- a/src/llmtuner/tuner/ppo/trainer.py
+++ b/src/llmtuner/tuner/ppo/trainer.py
@@ -10,7 +10,7 @@ from trl import PPOTrainer
 from trl.core import LengthSampler
 
 from llmtuner.extras.logging import get_logger
-from llmtuner.extras.misc import AverageMeter, count_parameters, get_logits_processor
+from llmtuner.extras.misc import AverageMeter, count_parameters, get_logits_processor, get_stopping_criteria
 from llmtuner.tuner.core.trainer import PeftTrainer
 from llmtuner.tuner.ppo.utils import cast_layernorm_dtype, replace_model
 
@@ -18,7 +18,7 @@ if TYPE_CHECKING:
     from transformers import Seq2SeqTrainingArguments
     from trl import AutoModelForCausalLMWithValueHead
     from llmtuner.extras.callbacks import LogCallback
-    from llmtuner.hparams import FinetuningArguments
+    from llmtuner.hparams import FinetuningArguments, GeneratingArguments
 
 
 logger = get_logger(__name__)
@@ -33,16 +33,17 @@ class PPOPeftTrainer(PPOTrainer, PeftTrainer):
         self,
         training_args: "Seq2SeqTrainingArguments",
         finetuning_args: "FinetuningArguments",
+        generating_args: "GeneratingArguments",
         callbacks: List["LogCallback"],
         **kwargs
     ):
         PPOTrainer.__init__(self, **kwargs)
         self.args = training_args
         self.finetuning_args = finetuning_args
+        self.generating_args = generating_args
         self.log_callback = callbacks[0]
         self.state = TrainerState()
         self.control = TrainerControl()
-        self._remove_log()
 
     def ppo_train(self, max_target_length: int) -> None:
         r"""
@@ -72,14 +73,10 @@ class PPOPeftTrainer(PPOTrainer, PeftTrainer):
             logger.info(f"  Number of trainable parameters = {count_parameters(self.model)[0]}")
 
         # Keyword arguments for `model.generate`
-        gen_kwargs = {
-            "top_k": 0.0,
-            "top_p": 1.0,
-            "do_sample": True,
-            "pad_token_id": self.tokenizer.pad_token_id,
-            "eos_token_id": self.tokenizer.eos_token_id,
-            "logits_processor": get_logits_processor()
-        }
+        gen_kwargs = self.generating_args.to_dict()
+        gen_kwargs["logits_processor"] = get_logits_processor()
+        gen_kwargs["stopping_criteria"] = get_stopping_criteria(self.tokenizer.additional_special_tokens_ids)
+
         length_sampler = LengthSampler(max_target_length // 2, max_target_length)
         unwrapped_model: "AutoModelForCausalLMWithValueHead" = self.accelerator.unwrap_model(self.model)
 
diff --git a/src/llmtuner/tuner/ppo/workflow.py b/src/llmtuner/tuner/ppo/workflow.py
index aa372671..6734ab78 100644
--- a/src/llmtuner/tuner/ppo/workflow.py
+++ b/src/llmtuner/tuner/ppo/workflow.py
@@ -1,11 +1,9 @@
-# Inspired by:
-# https://github.com/lvwerra/trl/blob/main/examples/research_projects/stack_llama/scripts/rl_training.py
+# Inspired by: https://github.com/lvwerra/trl/blob/main/examples/research_projects/stack_llama/scripts/rl_training.py
 
 import math
-from typing import TYPE_CHECKING
 from trl import PPOConfig
 from torch.optim import AdamW
-from typing import Optional, List
+from typing import TYPE_CHECKING, Optional, List
 from transformers import DataCollatorForSeq2Seq
 from transformers.optimization import get_scheduler
 
@@ -16,7 +14,7 @@ from llmtuner.tuner.ppo.trainer import PPOPeftTrainer
 
 if TYPE_CHECKING:
     from transformers import Seq2SeqTrainingArguments, TrainerCallback
-    from llmtuner.hparams import ModelArguments, DataArguments, FinetuningArguments
+    from llmtuner.hparams import ModelArguments, DataArguments, FinetuningArguments, GeneratingArguments
 
 
 def run_ppo(
@@ -24,6 +22,7 @@ def run_ppo(
     data_args: "DataArguments",
     training_args: "Seq2SeqTrainingArguments",
     finetuning_args: "FinetuningArguments",
+    generating_args: "GeneratingArguments",
     callbacks: Optional[List["TrainerCallback"]] = None
 ):
     dataset = get_dataset(model_args, data_args)
@@ -42,8 +41,9 @@ def run_ppo(
     )
 
     optimizer = AdamW(filter(lambda p: p.requires_grad, model.parameters()), lr=training_args.learning_rate)
-    total_train_batch_size = \
+    total_train_batch_size = (
         training_args.per_device_train_batch_size * training_args.gradient_accumulation_steps * training_args.world_size
+    )
     num_training_steps = training_args.num_train_epochs * math.ceil(len(dataset) / total_train_batch_size)
     lr_scheduler = get_scheduler(
         training_args.lr_scheduler_type,
@@ -56,6 +56,7 @@ def run_ppo(
     ppo_trainer = PPOPeftTrainer(
         training_args=training_args,
         finetuning_args=finetuning_args,
+        generating_args=generating_args,
         callbacks=callbacks,
         config=ppo_config,
         model=model,
@@ -67,8 +68,10 @@ def run_ppo(
         lr_scheduler=lr_scheduler
     )
 
-    ppo_trainer.ppo_train(max_target_length=data_args.max_target_length)
-    ppo_trainer.save_model()
-    ppo_trainer.save_state() # must be after save_model
-    if ppo_trainer.is_world_process_zero() and model_args.plot_loss:
-        plot_loss(training_args.output_dir, keys=["loss", "reward"])
+    # Training
+    if training_args.do_train:
+        ppo_trainer.ppo_train(max_target_length=data_args.max_target_length)
+        ppo_trainer.save_model()
+        ppo_trainer.save_state() # must be called after save_model to have a folder
+        if ppo_trainer.is_world_process_zero() and model_args.plot_loss:
+            plot_loss(training_args.output_dir, keys=["loss", "reward"])
diff --git a/src/llmtuner/tuner/pt/workflow.py b/src/llmtuner/tuner/pt/workflow.py
index b4ea148b..f7bf6448 100644
--- a/src/llmtuner/tuner/pt/workflow.py
+++ b/src/llmtuner/tuner/pt/workflow.py
@@ -2,10 +2,9 @@
 
 import math
 from typing import TYPE_CHECKING, Optional, List
-from transformers import DataCollatorForSeq2Seq
+from transformers import DataCollatorForLanguageModeling
 
 from llmtuner.dsets import get_dataset, preprocess_dataset, split_dataset
-from llmtuner.extras.constants import IGNORE_INDEX
 from llmtuner.extras.ploting import plot_loss
 from llmtuner.tuner.core import load_model_and_tokenizer
 from llmtuner.tuner.core.trainer import PeftTrainer
@@ -25,10 +24,7 @@ def run_pt(
     dataset = get_dataset(model_args, data_args)
     model, tokenizer = load_model_and_tokenizer(model_args, finetuning_args, training_args.do_train, stage="pt")
     dataset = preprocess_dataset(dataset, tokenizer, data_args, training_args, stage="pt")
-    data_collator = DataCollatorForSeq2Seq(
-        tokenizer=tokenizer,
-        label_pad_token_id=IGNORE_INDEX if data_args.ignore_pad_token_for_loss else tokenizer.pad_token_id
-    )
+    data_collator = DataCollatorForLanguageModeling(tokenizer=tokenizer, mlm=False)
 
     # Initialize our Trainer
     trainer = PeftTrainer(
diff --git a/src/llmtuner/tuner/rm/collator.py b/src/llmtuner/tuner/rm/collator.py
index c0da0579..161f003d 100644
--- a/src/llmtuner/tuner/rm/collator.py
+++ b/src/llmtuner/tuner/rm/collator.py
@@ -1,8 +1,10 @@
 import torch
+from dataclasses import dataclass
 from typing import Any, Dict, Sequence
 from transformers import DataCollatorWithPadding
 
 
+@dataclass
 class PairwiseDataCollatorWithPadding(DataCollatorWithPadding):
     r"""
     Data collator for pairwise data.
@@ -16,7 +18,10 @@ class PairwiseDataCollatorWithPadding(DataCollatorWithPadding):
         the last n examples represent rejected examples.
         """
         features = [
-            {"input_ids": feature[key], "attention_mask": [1] * len(feature[key])}
-            for key in ("accept_ids", "reject_ids") for feature in features
+            {
+                "input_ids": feature["prompt_ids"] + feature[key],
+                "attention_mask": [1] * (len(feature["prompt_ids"]) + len(feature[key]))
+            }
+            for key in ("chosen_ids", "rejected_ids") for feature in features
         ]
         return super().__call__(features)
diff --git a/src/llmtuner/tuner/sft/trainer.py b/src/llmtuner/tuner/sft/trainer.py
index 21739ac1..6243928f 100644
--- a/src/llmtuner/tuner/sft/trainer.py
+++ b/src/llmtuner/tuner/sft/trainer.py
@@ -79,7 +79,7 @@ class Seq2SeqPeftTrainer(PeftTrainer):
 
         padded_tensor = pad_token_id * torch.ones_like(tgt_tensor)
         padded_tensor[:, -src_tensor.shape[-1]:] = src_tensor # adopt left-padding
-        return padded_tensor.contiguous()
+        return padded_tensor.contiguous() # in contiguous memory
 
     def save_predictions(
         self,
diff --git a/src/llmtuner/tuner/sft/workflow.py b/src/llmtuner/tuner/sft/workflow.py
index a5cd2cd3..10f7aafb 100644
--- a/src/llmtuner/tuner/sft/workflow.py
+++ b/src/llmtuner/tuner/sft/workflow.py
@@ -5,7 +5,7 @@ from transformers import DataCollatorForSeq2Seq
 
 from llmtuner.dsets import get_dataset, preprocess_dataset, split_dataset
 from llmtuner.extras.constants import IGNORE_INDEX
-from llmtuner.extras.misc import get_logits_processor
+from llmtuner.extras.misc import get_logits_processor, get_stopping_criteria
 from llmtuner.extras.ploting import plot_loss
 from llmtuner.tuner.core import load_model_and_tokenizer
 from llmtuner.tuner.sft.metric import ComputeMetrics
@@ -13,7 +13,7 @@ from llmtuner.tuner.sft.trainer import Seq2SeqPeftTrainer
 
 if TYPE_CHECKING:
     from transformers import Seq2SeqTrainingArguments, TrainerCallback
-    from llmtuner.hparams import ModelArguments, DataArguments, FinetuningArguments
+    from llmtuner.hparams import ModelArguments, DataArguments, FinetuningArguments, GeneratingArguments
 
 
 def run_sft(
@@ -21,6 +21,7 @@ def run_sft(
     data_args: "DataArguments",
     training_args: "Seq2SeqTrainingArguments",
     finetuning_args: "FinetuningArguments",
+    generating_args: "GeneratingArguments",
     callbacks: Optional[List["TrainerCallback"]] = None
 ):
     dataset = get_dataset(model_args, data_args)
@@ -50,13 +51,9 @@ def run_sft(
     )
 
     # Keyword arguments for `model.generate`
-    gen_kwargs = {
-        "do_sample": True,
-        "top_p": 0.7,
-        "max_new_tokens": data_args.max_target_length + 1,
-        "temperature": 0.95,
-        "logits_processor": get_logits_processor()
-    }
+    gen_kwargs = generating_args.to_dict()
+    gen_kwargs["logits_processor"] = get_logits_processor()
+    gen_kwargs["stopping_criteria"] = get_stopping_criteria(tokenizer.additional_special_tokens_ids)
 
     # Training
     if training_args.do_train:
diff --git a/src/llmtuner/tuner/tune.py b/src/llmtuner/tuner/tune.py
index 99f5d2a9..dee49ef4 100644
--- a/src/llmtuner/tuner/tune.py
+++ b/src/llmtuner/tuner/tune.py
@@ -1,35 +1,47 @@
 from typing import TYPE_CHECKING, Any, Dict, List, Optional
 
 from llmtuner.extras.callbacks import LogCallback
+from llmtuner.extras.logging import get_logger
 from llmtuner.tuner.core import get_train_args, load_model_and_tokenizer
 from llmtuner.tuner.pt import run_pt
 from llmtuner.tuner.sft import run_sft
 from llmtuner.tuner.rm import run_rm
 from llmtuner.tuner.ppo import run_ppo
+from llmtuner.tuner.dpo import run_dpo
 
 if TYPE_CHECKING:
     from transformers import TrainerCallback
 
 
+logger = get_logger(__name__)
+
+
 def run_exp(args: Optional[Dict[str, Any]] = None, callbacks: Optional[List["TrainerCallback"]] = None):
-    model_args, data_args, training_args, finetuning_args, general_args = get_train_args(args)
-    callbacks = [LogCallback()] if callbacks is None else callbacks
+    model_args, data_args, training_args, finetuning_args, generating_args, general_args = get_train_args(args)
+    callbacks = [LogCallback()] if callbacks is None else callbacks + [LogCallback()]
 
     if general_args.stage == "pt":
         run_pt(model_args, data_args, training_args, finetuning_args, callbacks)
     elif general_args.stage == "sft":
-        run_sft(model_args, data_args, training_args, finetuning_args, callbacks)
+        run_sft(model_args, data_args, training_args, finetuning_args, generating_args, callbacks)
     elif general_args.stage == "rm":
         run_rm(model_args, data_args, training_args, finetuning_args, callbacks)
     elif general_args.stage == "ppo":
-        run_ppo(model_args, data_args, training_args, finetuning_args, callbacks)
+        run_ppo(model_args, data_args, training_args, finetuning_args, generating_args, callbacks)
+    elif general_args.stage == "dpo":
+        run_dpo(model_args, data_args, training_args, finetuning_args, callbacks)
+    else:
+        raise ValueError("Unknown task.")
 
 
 def export_model(args: Optional[Dict[str, Any]] = None, max_shard_size: Optional[str] = "10GB"):
     model_args, _, training_args, finetuning_args, _ = get_train_args(args)
     model, tokenizer = load_model_and_tokenizer(model_args, finetuning_args)
     model.save_pretrained(training_args.output_dir, max_shard_size=max_shard_size)
-    tokenizer.save_pretrained(training_args.output_dir)
+    try:
+        tokenizer.save_pretrained(training_args.output_dir)
+    except:
+        logger.warning("Cannot save tokenizer, please copy the files manually.")
 
 
 if __name__ == "__main__":
diff --git a/src/llmtuner/webui/runner.py b/src/llmtuner/webui/runner.py
index fddc4070..36a8bf53 100644
--- a/src/llmtuner/webui/runner.py
+++ b/src/llmtuner/webui/runner.py
@@ -4,7 +4,7 @@ import threading
 import time
 import transformers
 from transformers.trainer import TRAINING_ARGS_NAME
-from typing import Generator, List, Optional, Tuple
+from typing import Generator, List, Tuple
 
 from llmtuner.extras.callbacks import LogCallback
 from llmtuner.extras.constants import DEFAULT_MODULE

From f1485ab92707bace72ab85364e18a3e670765b83 Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Fri, 11 Aug 2023 13:45:47 +0800
Subject: [PATCH 09/45] fix baichuan template

Former-commit-id: 9c6dd1051417c91074daa7dd6ed6cc53448135ad
---
 src/llmtuner/dsets/preprocess.py |  4 +++-
 src/llmtuner/extras/template.py  | 15 +++++++++------
 2 files changed, 12 insertions(+), 7 deletions(-)

diff --git a/src/llmtuner/dsets/preprocess.py b/src/llmtuner/dsets/preprocess.py
index 4efbcbb6..dc882ae0 100644
--- a/src/llmtuner/dsets/preprocess.py
+++ b/src/llmtuner/dsets/preprocess.py
@@ -1,3 +1,4 @@
+import tiktoken
 from typing import TYPE_CHECKING, Any, Dict, Generator, List, Literal
 from itertools import chain
 
@@ -31,10 +32,11 @@ def preprocess_dataset(
 
     def preprocess_pretrain_dataset(examples: Dict[str, List[Any]]) -> Dict[str, Any]:
         # build grouped texts with format `X1 X2 X3 ...` (without <eos>)
-        if hasattr(tokenizer, "tokenizer"): # for tiktoken tokenizer (Qwen)
+        if isinstance(getattr(tokenizer, "tokenizer"), tiktoken.Encoding): # for tiktoken tokenizer (Qwen)
             kwargs = dict(allowed_special="all")
         else:
             kwargs = dict(add_special_tokens=False)
+
         tokenized_examples = tokenizer(examples["prompt"], **kwargs)
         concatenated_examples = {k: list(chain(*tokenized_examples[k])) for k in tokenized_examples.keys()}
         total_length = len(concatenated_examples[list(concatenated_examples.keys())[0]])
diff --git a/src/llmtuner/extras/template.py b/src/llmtuner/extras/template.py
index c3388ab3..4bf490aa 100644
--- a/src/llmtuner/extras/template.py
+++ b/src/llmtuner/extras/template.py
@@ -1,5 +1,6 @@
-from typing import TYPE_CHECKING, Dict, List, Optional, Tuple, Union
+import tiktoken
 from dataclasses import dataclass
+from typing import TYPE_CHECKING, Dict, List, Optional, Tuple, Union
 
 from llmtuner.extras.logging import get_logger
 
@@ -122,7 +123,7 @@ class Template:
         r"""
         Converts context to token ids.
         """
-        if hasattr(tokenizer, "tokenizer"): # for tiktoken tokenizer (Qwen)
+        if isinstance(getattr(tokenizer, "tokenizer"), tiktoken.Encoding): # for tiktoken tokenizer (Qwen)
             kwargs = dict(allowed_special="all")
         else:
             kwargs = dict(add_special_tokens=False)
@@ -428,14 +429,16 @@ Supports: https://huggingface.co/baichuan-inc/Baichuan-13B-Chat
 """
 register_template(
     name="baichuan",
-    prefix=[],
+    prefix=[
+        {"token": "<reserved_102>"} # user token (a little difference in the first turn)
+    ],
     prompt=[
-        {"token": "<reserved_102>"}, # user token (a little difference in position)
-        "{{query}}"
+        "{{query}}",
+        {"token": "<reserved_103>"} # assistant token
     ],
     sep=[],
     stop_words=[
-        "<reserved_103>" # assistant token
+        {"token": "<reserved_102>"} # user token
     ],
     use_history=True
 )

From eb26bfc2bade43d848587ac16a54c704d97aecb6 Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Fri, 11 Aug 2023 13:51:46 +0800
Subject: [PATCH 10/45] fix stop word in baichuan template

Former-commit-id: d59f9389590aab570f68ad4898b035741f9fd1c8
---
 src/llmtuner/extras/template.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/llmtuner/extras/template.py b/src/llmtuner/extras/template.py
index 4bf490aa..64c22285 100644
--- a/src/llmtuner/extras/template.py
+++ b/src/llmtuner/extras/template.py
@@ -438,7 +438,7 @@ register_template(
     ],
     sep=[],
     stop_words=[
-        {"token": "<reserved_102>"} # user token
+        "<reserved_102>" # user token
     ],
     use_history=True
 )

From 21bf79e72b1afb9a12aaae75f1a331b7a918973d Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Fri, 11 Aug 2023 13:56:26 +0800
Subject: [PATCH 11/45] add defaults

Former-commit-id: d3844e97e387b2106a32a576a61318ecec948e23
---
 src/llmtuner/dsets/preprocess.py | 2 +-
 src/llmtuner/extras/template.py  | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/src/llmtuner/dsets/preprocess.py b/src/llmtuner/dsets/preprocess.py
index dc882ae0..64e0d8b1 100644
--- a/src/llmtuner/dsets/preprocess.py
+++ b/src/llmtuner/dsets/preprocess.py
@@ -32,7 +32,7 @@ def preprocess_dataset(
 
     def preprocess_pretrain_dataset(examples: Dict[str, List[Any]]) -> Dict[str, Any]:
         # build grouped texts with format `X1 X2 X3 ...` (without <eos>)
-        if isinstance(getattr(tokenizer, "tokenizer"), tiktoken.Encoding): # for tiktoken tokenizer (Qwen)
+        if isinstance(getattr(tokenizer, "tokenizer", None), tiktoken.Encoding): # for tiktoken tokenizer (Qwen)
             kwargs = dict(allowed_special="all")
         else:
             kwargs = dict(add_special_tokens=False)
diff --git a/src/llmtuner/extras/template.py b/src/llmtuner/extras/template.py
index 64c22285..c6e21d87 100644
--- a/src/llmtuner/extras/template.py
+++ b/src/llmtuner/extras/template.py
@@ -123,7 +123,7 @@ class Template:
         r"""
         Converts context to token ids.
         """
-        if isinstance(getattr(tokenizer, "tokenizer"), tiktoken.Encoding): # for tiktoken tokenizer (Qwen)
+        if isinstance(getattr(tokenizer, "tokenizer", None), tiktoken.Encoding): # for tiktoken tokenizer (Qwen)
             kwargs = dict(allowed_special="all")
         else:
             kwargs = dict(add_special_tokens=False)

From e86c4ef06c57940654f54d7d0860ede1a6527339 Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Fri, 11 Aug 2023 14:06:02 +0800
Subject: [PATCH 12/45] Update README_zh.md

Former-commit-id: 156710a9954b7dd079c6d583c399d7a3d5c3915c
---
 README_zh.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/README_zh.md b/README_zh.md
index 955fefc1..f9d967c5 100644
--- a/README_zh.md
+++ b/README_zh.md
@@ -12,7 +12,7 @@
 
 ## 更新日志
 
-[23/08/11] 现在我们支持了指令模型的 **[DPO 训练](https://arxiv.org/abs/2305.18290)**。详情请参阅[此示例](#dpo-training)（实验性功能）。
+[23/08/11] 现在我们支持了指令模型的 **[DPO 训练](https://arxiv.org/abs/2305.18290)**。详情请参阅[此示例](#dpo-训练)（实验性功能）。
 
 [23/08/03] 现在我们支持了 **Qwen-7B** 模型的训练。请尝试使用 `--model_name_or_path Qwen/Qwen-7B-Chat` 和 `--lora_target c_attn` 参数。使用 Qwen-7B-Chat 模型请添加 `--template chatml` 参数。
 

From 79f4ba0d262c5540b16d15c189f8d34ebbca6ab7 Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Fri, 11 Aug 2023 23:25:57 +0800
Subject: [PATCH 13/45] Release v0.1.6

Former-commit-id: a48cb0d474ef0648a97387daf5f623498b5e3ee6
---
 README.md                               |  3 +++
 src/llmtuner/__init__.py                |  2 +-
 src/llmtuner/dsets/loader.py            |  6 ++++--
 src/llmtuner/dsets/utils.py             |  3 ++-
 src/llmtuner/extras/constants.py        |  8 +++++--
 src/llmtuner/extras/template.py         | 19 ++++++++++++++++-
 src/llmtuner/hparams/finetuning_args.py |  4 ++++
 src/llmtuner/hparams/model_args.py      |  4 ----
 src/llmtuner/tuner/core/adapter.py      |  2 +-
 src/llmtuner/tuner/tune.py              |  2 +-
 src/llmtuner/webui/components/data.py   |  2 +-
 src/llmtuner/webui/components/eval.py   | 15 +++++++++++--
 src/llmtuner/webui/components/sft.py    | 28 ++++++++++++++++++++-----
 src/llmtuner/webui/components/top.py    |  2 +-
 src/llmtuner/webui/interface.py         |  2 +-
 src/llmtuner/webui/locales.py           | 20 ++++++++++++++++++
 src/llmtuner/webui/runner.py            | 27 ++++++++++++++----------
 src/llmtuner/webui/utils.py             | 19 ++++++++++-------
 18 files changed, 127 insertions(+), 41 deletions(-)

diff --git a/README.md b/README.md
index fcbe0761..a54ac30a 100644
--- a/README.md
+++ b/README.md
@@ -55,6 +55,7 @@
 | [InternLM](https://github.com/InternLM/InternLM)         | 7B                          | q_proj,v_proj     | intern   |
 | [Qwen](https://github.com/QwenLM/Qwen-7B)                | 7B                          | c_attn            | chatml   |
 | [XVERSE](https://github.com/xverse-ai/XVERSE-13B)        | 13B                         | q_proj,v_proj     | -        |
+| [ChatGLM2](https://github.com/THUDM/ChatGLM2-6B)         | 6B                          | query_key_value   | chatglm2 |
 
 - **Default module** is used for the `--lora_target` argument. Please use `python src/train_bash.py -h` to see all available options.
 - For the "base" models, the `--template` argument can be chosen from `default`, `alpaca`, `vicuna` etc. But make sure to use the corresponding template for the "chat" models.
@@ -408,6 +409,8 @@ Please follow the model licenses to use the corresponding model weights:
 - [Baichuan](https://huggingface.co/baichuan-inc/baichuan-7B/resolve/main/baichuan-7B%20%E6%A8%A1%E5%9E%8B%E8%AE%B8%E5%8F%AF%E5%8D%8F%E8%AE%AE.pdf)
 - [InternLM](https://github.com/InternLM/InternLM#open-source-license)
 - [Qwen](https://huggingface.co/Qwen/Qwen-7B-Chat/blob/main/LICENSE)
+- [XVERSE](https://github.com/xverse-ai/XVERSE-13B/blob/main/MODEL_LICENSE.pdf)
+- [ChatGLM2](https://github.com/THUDM/ChatGLM2-6B/blob/main/MODEL_LICENSE)
 
 ## Citation
 
diff --git a/src/llmtuner/__init__.py b/src/llmtuner/__init__.py
index e647b92b..bbc1420b 100644
--- a/src/llmtuner/__init__.py
+++ b/src/llmtuner/__init__.py
@@ -6,4 +6,4 @@ from llmtuner.tuner import export_model, run_exp
 from llmtuner.webui import create_ui, create_web_demo
 
 
-__version__ = "0.1.5"
+__version__ = "0.1.6"
diff --git a/src/llmtuner/dsets/loader.py b/src/llmtuner/dsets/loader.py
index 90a4212f..7588443f 100644
--- a/src/llmtuner/dsets/loader.py
+++ b/src/llmtuner/dsets/loader.py
@@ -93,11 +93,13 @@ def get_dataset(
                 dataset = dataset.rename_column(getattr(dataset_attr, column_name), column_name)
 
         if dataset_attr.source_prefix: # add prefix
-            features = None
             if data_args.streaming:
                 features = dataset.features
                 features["prefix"] = Value(dtype="string", id=None)
-            dataset = dataset.map(lambda _: {"prefix": dataset_attr.source_prefix}, features=features)
+                dataset = dataset.map(lambda _: {"prefix": dataset_attr.source_prefix}, features=features)
+            else:
+                prefix_data = [dataset_attr.source_prefix] * len(dataset)
+                dataset = dataset.add_column("prefix", prefix_data)
 
         all_datasets.append(dataset)
 
diff --git a/src/llmtuner/dsets/utils.py b/src/llmtuner/dsets/utils.py
index e1093a95..f816649e 100644
--- a/src/llmtuner/dsets/utils.py
+++ b/src/llmtuner/dsets/utils.py
@@ -19,7 +19,8 @@ def split_dataset(
                 dataset = dataset.shuffle(buffer_size=data_args.buffer_size, seed=training_args.seed)
                 return {"train_dataset": train_set, "eval_dataset": val_set}
             else:
-                dataset = dataset.train_test_split(test_size=data_args.val_size, seed=training_args.seed)
+                val_size = int(data_args.val_size) if data_args.val_size > 1 else data_args.val_size
+                dataset = dataset.train_test_split(test_size=val_size, seed=training_args.seed)
                 return {"train_dataset": dataset["train"], "eval_dataset": dataset["test"]}
         else:
             if data_args.streaming:
diff --git a/src/llmtuner/extras/constants.py b/src/llmtuner/extras/constants.py
index dc1cb2e7..8ee997bb 100644
--- a/src/llmtuner/extras/constants.py
+++ b/src/llmtuner/extras/constants.py
@@ -37,7 +37,9 @@ SUPPORTED_MODELS = {
     "InternLM-7B": "internlm/internlm-7b",
     "InternLM-7B-Chat": "internlm/internlm-chat-7b",
     "Qwen-7B": "Qwen/Qwen-7B",
-    "Qwen-7B-Chat": "Qwen/Qwen-7B-Chat"
+    "Qwen-7B-Chat": "Qwen/Qwen-7B-Chat",
+    "XVERSE-13B": "xverse/XVERSE-13B",
+    "ChatGLM2-6B": "THUDM/chatglm2-6b"
 }
 
 DEFAULT_MODULE = {
@@ -48,5 +50,7 @@ DEFAULT_MODULE = {
     "Falcon": "query_key_value",
     "Baichuan": "W_pack",
     "InternLM": "q_proj,v_proj",
-    "Qwen": "c_attn"
+    "Qwen": "c_attn",
+    "XVERSE": "q_proj,v_proj",
+    "ChatGLM2": "query_key_value"
 }
diff --git a/src/llmtuner/extras/template.py b/src/llmtuner/extras/template.py
index c6e21d87..6724dd46 100644
--- a/src/llmtuner/extras/template.py
+++ b/src/llmtuner/extras/template.py
@@ -178,7 +178,7 @@ def register_template(
     stop_words: List[str],
     use_history: bool
 ) -> None:
-    template_class = Llama2Template if name == "llama2" else Template
+    template_class = Llama2Template if "llama2" in name else Template
     templates[name] = template_class(
         prefix=prefix,
         prompt=prompt,
@@ -272,6 +272,23 @@ register_template(
 )
 
 
+r"""
+Supports: https://github.com/ymcui/Chinese-LLaMA-Alpaca-2
+"""
+register_template(
+    name="llama2_zh",
+    prefix=[
+        "<<SYS>>\nYou are a helpful assistant. 你是一个乐于助人的助手。\n<</SYS>>\n\n"
+    ],
+    prompt=[
+        "[INST] {{query}} [/INST] "
+    ],
+    sep=[],
+    stop_words=[],
+    use_history=True
+)
+
+
 r"""
 Supports: https://huggingface.co/tatsu-lab/alpaca-7b-wdiff
           https://github.com/ymcui/Chinese-LLaMA-Alpaca
diff --git a/src/llmtuner/hparams/finetuning_args.py b/src/llmtuner/hparams/finetuning_args.py
index c4713c5e..d7d651dd 100644
--- a/src/llmtuner/hparams/finetuning_args.py
+++ b/src/llmtuner/hparams/finetuning_args.py
@@ -57,6 +57,10 @@ class FinetuningArguments:
                   Qwen choices: [\"c_attn\", \"attn.c_proj\", \"w1\", \"w2\", \"mlp.c_proj\"], \
                   LLaMA-2, InternLM, XVERSE choices: the same as LLaMA."}
     )
+    resume_lora_training: Optional[bool] = field(
+        default=True,
+        metadata={"help": "Whether to resume training from the last LoRA weights or create new weights after merging them."}
+    )
     dpo_beta: Optional[float] = field(
         default=0.1,
         metadata={"help": "The beta parameter for the DPO loss."}
diff --git a/src/llmtuner/hparams/model_args.py b/src/llmtuner/hparams/model_args.py
index 6c8491da..44454f45 100644
--- a/src/llmtuner/hparams/model_args.py
+++ b/src/llmtuner/hparams/model_args.py
@@ -55,10 +55,6 @@ class ModelArguments:
         default=None,
         metadata={"help": "Path to the directory containing the checkpoints of the reward model."}
     )
-    resume_lora_training: Optional[bool] = field(
-        default=True,
-        metadata={"help": "Whether to resume training from the last LoRA weights or create new weights after merging them."}
-    )
     plot_loss: Optional[bool] = field(
         default=False,
         metadata={"help": "Whether to plot the training loss after fine-tuning or not."}
diff --git a/src/llmtuner/tuner/core/adapter.py b/src/llmtuner/tuner/core/adapter.py
index a8ac5a84..5db56876 100644
--- a/src/llmtuner/tuner/core/adapter.py
+++ b/src/llmtuner/tuner/core/adapter.py
@@ -65,7 +65,7 @@ def init_adapter(
             assert os.path.exists(os.path.join(model_args.checkpoint_dir[0], CONFIG_NAME)), \
                 "The given checkpoint may be not a LoRA checkpoint, please specify `--finetuning_type full/freeze` instead."
 
-            if (is_trainable and model_args.resume_lora_training) or (not is_mergeable): # continually train on the lora weights
+            if (is_trainable and finetuning_args.resume_lora_training) or (not is_mergeable): # continually fine-tuning
                 checkpoints_to_merge, latest_checkpoint = model_args.checkpoint_dir[:-1], model_args.checkpoint_dir[-1]
             else:
                 checkpoints_to_merge = model_args.checkpoint_dir
diff --git a/src/llmtuner/tuner/tune.py b/src/llmtuner/tuner/tune.py
index dee49ef4..7b9446e9 100644
--- a/src/llmtuner/tuner/tune.py
+++ b/src/llmtuner/tuner/tune.py
@@ -18,7 +18,7 @@ logger = get_logger(__name__)
 
 def run_exp(args: Optional[Dict[str, Any]] = None, callbacks: Optional[List["TrainerCallback"]] = None):
     model_args, data_args, training_args, finetuning_args, generating_args, general_args = get_train_args(args)
-    callbacks = [LogCallback()] if callbacks is None else callbacks + [LogCallback()]
+    callbacks = [LogCallback()] if callbacks is None else callbacks
 
     if general_args.stage == "pt":
         run_pt(model_args, data_args, training_args, finetuning_args, callbacks)
diff --git a/src/llmtuner/webui/components/data.py b/src/llmtuner/webui/components/data.py
index 9787b36a..af19cc41 100644
--- a/src/llmtuner/webui/components/data.py
+++ b/src/llmtuner/webui/components/data.py
@@ -16,6 +16,6 @@ def create_preview_box() -> Tuple["Block", "Component", "Component", "Component"
 
         close_btn = gr.Button()
 
-    close_btn.click(lambda: gr.update(visible=False), outputs=[preview_box])
+    close_btn.click(lambda: gr.update(visible=False), outputs=[preview_box], queue=False)
 
     return preview_box, preview_count, preview_samples, close_btn
diff --git a/src/llmtuner/webui/components/eval.py b/src/llmtuner/webui/components/eval.py
index 29b590ae..ce6eddae 100644
--- a/src/llmtuner/webui/components/eval.py
+++ b/src/llmtuner/webui/components/eval.py
@@ -20,7 +20,12 @@ def create_eval_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict
 
     dataset_dir.change(list_dataset, [dataset_dir], [dataset])
     dataset.change(can_preview, [dataset_dir, dataset], [preview_btn])
-    preview_btn.click(get_preview, [dataset_dir, dataset], [preview_count, preview_samples, preview_box])
+    preview_btn.click(
+        get_preview,
+        [dataset_dir, dataset],
+        [preview_count, preview_samples, preview_box],
+        queue=False
+    )
 
     with gr.Row():
         max_source_length = gr.Slider(value=512, minimum=4, maximum=4096, step=1)
@@ -33,6 +38,9 @@ def create_eval_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict
         start_btn = gr.Button()
         stop_btn = gr.Button()
 
+    with gr.Row():
+        process_bar = gr.Slider(visible=False, interactive=False)
+
     with gr.Box():
         output_box = gr.Markdown()
 
@@ -54,7 +62,10 @@ def create_eval_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict
             batch_size,
             predict
         ],
-        [output_box]
+        [
+            output_box,
+            process_bar
+        ]
     )
     stop_btn.click(runner.set_abort, queue=False)
 
diff --git a/src/llmtuner/webui/components/sft.py b/src/llmtuner/webui/components/sft.py
index 9db57ac9..05a6e530 100644
--- a/src/llmtuner/webui/components/sft.py
+++ b/src/llmtuner/webui/components/sft.py
@@ -22,7 +22,12 @@ def create_sft_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict[
 
     dataset_dir.change(list_dataset, [dataset_dir], [dataset])
     dataset.change(can_preview, [dataset_dir, dataset], [preview_btn])
-    preview_btn.click(get_preview, [dataset_dir, dataset], [preview_count, preview_samples, preview_box])
+    preview_btn.click(
+        get_preview,
+        [dataset_dir, dataset],
+        [preview_count, preview_samples, preview_box],
+        queue=False
+    )
 
     with gr.Row():
         max_source_length = gr.Slider(value=512, minimum=4, maximum=4096, step=1)
@@ -46,12 +51,14 @@ def create_sft_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict[
             save_steps = gr.Slider(value=100, minimum=10, maximum=5000, step=10)
             warmup_steps = gr.Slider(value=0, minimum=0, maximum=5000, step=1)
             compute_type = gr.Radio(choices=["fp16", "bf16"], value="fp16")
+            padding_side = gr.Radio(choices=["left", "right"], value="left")
 
     with gr.Accordion(label="LoRA config", open=False) as lora_tab:
         with gr.Row():
             lora_rank = gr.Slider(value=8, minimum=1, maximum=1024, step=1, scale=1)
-            lora_dropout = gr.Slider(value=0, minimum=0, maximum=1, step=0.01, scale=1)
+            lora_dropout = gr.Slider(value=0.1, minimum=0, maximum=1, step=0.01, scale=1)
             lora_target = gr.Textbox(scale=2)
+            resume_lora_training = gr.Checkbox(value=True, scale=1)
 
     with gr.Row():
         start_btn = gr.Button()
@@ -59,7 +66,11 @@ def create_sft_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict[
 
     with gr.Row():
         with gr.Column(scale=3):
-            output_dir = gr.Textbox()
+            with gr.Row():
+                output_dir = gr.Textbox()
+
+            with gr.Row():
+                process_bar = gr.Slider(visible=False, interactive=False)
 
             with gr.Box():
                 output_box = gr.Markdown()
@@ -93,16 +104,21 @@ def create_sft_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict[
             save_steps,
             warmup_steps,
             compute_type,
+            padding_side,
             lora_rank,
             lora_dropout,
             lora_target,
+            resume_lora_training,
             output_dir
         ],
-        [output_box]
+        [
+            output_box,
+            process_bar
+        ]
     )
     stop_btn.click(runner.set_abort, queue=False)
 
-    output_box.change(
+    process_bar.change(
         gen_plot, [top_elems["model_name"], top_elems["finetuning_type"], output_dir], loss_viewer, queue=False
     )
 
@@ -128,10 +144,12 @@ def create_sft_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict[
         save_steps=save_steps,
         warmup_steps=warmup_steps,
         compute_type=compute_type,
+        padding_side=padding_side,
         lora_tab=lora_tab,
         lora_rank=lora_rank,
         lora_dropout=lora_dropout,
         lora_target=lora_target,
+        resume_lora_training=resume_lora_training,
         start_btn=start_btn,
         stop_btn=stop_btn,
         output_dir=output_dir,
diff --git a/src/llmtuner/webui/components/top.py b/src/llmtuner/webui/components/top.py
index 4fc5b506..77d60593 100644
--- a/src/llmtuner/webui/components/top.py
+++ b/src/llmtuner/webui/components/top.py
@@ -43,7 +43,7 @@ def create_top() -> Dict[str, "Component"]:
         can_quantize, [finetuning_type], [quantization_bit]
     )
 
-    refresh_btn.click(list_checkpoint, [model_name, finetuning_type], [checkpoints])
+    refresh_btn.click(list_checkpoint, [model_name, finetuning_type], [checkpoints], queue=False)
 
     return dict(
         lang=lang,
diff --git a/src/llmtuner/webui/interface.py b/src/llmtuner/webui/interface.py
index 2fb61d37..8539e18c 100644
--- a/src/llmtuner/webui/interface.py
+++ b/src/llmtuner/webui/interface.py
@@ -67,7 +67,7 @@ def create_web_demo() -> gr.Blocks:
 
         demo.load(manager.gen_label, [lang], [lang] + list(chat_elems.values()))
 
-        lang.change(manager.gen_label, [lang], [lang] + list(chat_elems.values()))
+        lang.change(manager.gen_label, [lang], [lang] + list(chat_elems.values()), queue=False)
 
     return demo
 
diff --git a/src/llmtuner/webui/locales.py b/src/llmtuner/webui/locales.py
index 9bf3d7a9..61491ece 100644
--- a/src/llmtuner/webui/locales.py
+++ b/src/llmtuner/webui/locales.py
@@ -277,6 +277,16 @@ LOCALES = {
             "info": "是否启用 FP16 或 BF16 混合精度训练。"
         }
     },
+    "padding_side": {
+        "en": {
+            "label": "Padding side",
+            "info": "The side on which the model should have padding applied."
+        },
+        "zh": {
+            "label": "填充位置",
+            "info": "使用左填充或右填充。"
+        }
+    },
     "lora_tab": {
         "en": {
             "label": "LoRA configurations"
@@ -315,6 +325,16 @@ LOCALES = {
             "info": "应用 LoRA 的线性层名称。使用英文逗号分隔多个名称。"
         }
     },
+    "resume_lora_training": {
+        "en": {
+            "label": "Resume LoRA training",
+            "info": "Whether to resume training from the last LoRA weights or create new lora weights."
+        },
+        "zh": {
+            "label": "继续上次的训练",
+            "info": "接着上次的 LoRA 权重训练或创建一个新的 LoRA 权重。"
+        }
+    },
     "start_btn": {
         "en": {
             "value": "Start"
diff --git a/src/llmtuner/webui/runner.py b/src/llmtuner/webui/runner.py
index 36a8bf53..1ae92786 100644
--- a/src/llmtuner/webui/runner.py
+++ b/src/llmtuner/webui/runner.py
@@ -1,3 +1,4 @@
+import gradio as gr
 import logging
 import os
 import threading
@@ -13,7 +14,7 @@ from llmtuner.extras.misc import torch_gc
 from llmtuner.tuner import run_exp
 from llmtuner.webui.common import get_model_path, get_save_dir
 from llmtuner.webui.locales import ALERTS
-from llmtuner.webui.utils import format_info, get_eval_results
+from llmtuner.webui.utils import get_eval_results, update_process_bar
 
 
 class Runner:
@@ -88,14 +89,16 @@ class Runner:
         save_steps: int,
         warmup_steps: int,
         compute_type: str,
+        padding_side: str,
         lora_rank: int,
         lora_dropout: float,
         lora_target: str,
+        resume_lora_training: bool,
         output_dir: str
     ) -> Generator[str, None, None]:
         model_name_or_path, error, logger_handler, trainer_callback = self.initialize(lang, model_name, dataset)
         if error:
-            yield error
+            yield error, gr.update(visible=False)
             return
 
         if checkpoints:
@@ -133,9 +136,11 @@ class Runner:
             warmup_steps=warmup_steps,
             fp16=(compute_type == "fp16"),
             bf16=(compute_type == "bf16"),
+            padding_side=padding_side,
             lora_rank=lora_rank,
             lora_dropout=lora_dropout,
             lora_target=lora_target or DEFAULT_MODULE.get(model_name.split("-")[0], "q_proj,v_proj"),
+            resume_lora_training=resume_lora_training,
             output_dir=output_dir
         )
 
@@ -150,18 +155,18 @@ class Runner:
         thread.start()
 
         while thread.is_alive():
-            time.sleep(1)
+            time.sleep(2)
             if self.aborted:
-                yield ALERTS["info_aborting"][lang]
+                yield ALERTS["info_aborting"][lang], gr.update(visible=False)
             else:
-                yield format_info(logger_handler.log, trainer_callback)
+                yield logger_handler.log, update_process_bar(trainer_callback)
 
         if os.path.exists(os.path.join(output_dir, TRAINING_ARGS_NAME)):
             finish_info = ALERTS["info_finished"][lang]
         else:
             finish_info = ALERTS["err_failed"][lang]
 
-        yield self.finalize(lang, finish_info)
+        yield self.finalize(lang, finish_info), gr.update(visible=False)
 
     def run_eval(
         self,
@@ -182,7 +187,7 @@ class Runner:
     ) -> Generator[str, None, None]:
         model_name_or_path, error, logger_handler, trainer_callback = self.initialize(lang, model_name, dataset)
         if error:
-            yield error
+            yield error, gr.update(visible=False)
             return
 
         if checkpoints:
@@ -223,15 +228,15 @@ class Runner:
         thread.start()
 
         while thread.is_alive():
-            time.sleep(1)
+            time.sleep(2)
             if self.aborted:
-                yield ALERTS["info_aborting"][lang]
+                yield ALERTS["info_aborting"][lang], gr.update(visible=False)
             else:
-                yield format_info(logger_handler.log, trainer_callback)
+                yield logger_handler.log, update_process_bar(trainer_callback)
 
         if os.path.exists(os.path.join(output_dir, "all_results.json")):
             finish_info = get_eval_results(os.path.join(output_dir, "all_results.json"))
         else:
             finish_info = ALERTS["err_failed"][lang]
 
-        yield self.finalize(lang, finish_info)
+        yield self.finalize(lang, finish_info), gr.update(visible=False)
diff --git a/src/llmtuner/webui/utils.py b/src/llmtuner/webui/utils.py
index 7b667c0f..fb22bd0c 100644
--- a/src/llmtuner/webui/utils.py
+++ b/src/llmtuner/webui/utils.py
@@ -15,13 +15,18 @@ if TYPE_CHECKING:
     from llmtuner.extras.callbacks import LogCallback
 
 
-def format_info(log: str, callback: "LogCallback") -> str:
-    info = log
-    if callback.max_steps:
-        info += "Running **{:d}/{:d}**: {} < {}\n".format(
-            callback.cur_steps, callback.max_steps, callback.elapsed_time, callback.remaining_time
-        )
-    return info
+def update_process_bar(callback: "LogCallback") -> Dict[str, Any]:
+    if not callback.max_steps:
+        return gr.update(visible=False)
+
+    percentage = round(100 * callback.cur_steps / callback.max_steps, 0) if callback.max_steps != 0 else 100.0
+    label = "Running {:d}/{:d}: {} < {}".format(
+        callback.cur_steps,
+        callback.max_steps,
+        callback.elapsed_time,
+        callback.remaining_time
+    )
+    return gr.update(label=label, value=percentage, visible=True)
 
 
 def get_time() -> str:

From 7bd4c59b7e0941fe37647a2be0d18732db982059 Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Sat, 12 Aug 2023 00:25:29 +0800
Subject: [PATCH 14/45] fix unusual output of 8bit models #278 #391

Former-commit-id: dd51c242032ce3f878cb191dc144536db4a2bb45
---
 src/llmtuner/extras/misc.py       | 3 +++
 src/llmtuner/tuner/core/loader.py | 2 +-
 2 files changed, 4 insertions(+), 1 deletion(-)

diff --git a/src/llmtuner/extras/misc.py b/src/llmtuner/extras/misc.py
index ee918fbb..b0a7365c 100644
--- a/src/llmtuner/extras/misc.py
+++ b/src/llmtuner/extras/misc.py
@@ -142,6 +142,9 @@ def dispatch_model(model: "PreTrainedModel") -> "PreTrainedModel":
     Dispatches a pre-trained model to GPUs with balanced memory.
     Borrowed from: https://github.com/huggingface/transformers/blob/v4.31.0/src/transformers/modeling_utils.py#L2803
     """
+    if getattr(model, "is_loaded_in_8bit", False) or getattr(model, "is_loaded_in_4bit", False): # do nothing
+        return model
+
     if torch.cuda.device_count() > 1:
         from accelerate import dispatch_model
         from accelerate.utils import infer_auto_device_map, get_balanced_memory
diff --git a/src/llmtuner/tuner/core/loader.py b/src/llmtuner/tuner/core/loader.py
index 39bec1d8..74b4b59f 100644
--- a/src/llmtuner/tuner/core/loader.py
+++ b/src/llmtuner/tuner/core/loader.py
@@ -92,7 +92,7 @@ def load_model_and_tokenizer(
             )
 
         is_mergeable = False
-        config_kwargs["device_map"] = {"": int(os.environ.get("LOCAL_RANK", "0"))}
+        config_kwargs["device_map"] = {"": int(os.environ.get("LOCAL_RANK", "0"))} if is_trainable else "auto"
         logger.info("Quantizing model to {} bit.".format(model_args.quantization_bit))
 
     # Load and prepare pretrained models (without valuehead).

From 3ba1b81105f4b26da948d7ff529632b9fc33d89f Mon Sep 17 00:00:00 2001
From: codemayq <codingma@pku.edu.cn>
Date: Sat, 12 Aug 2023 13:53:55 +0800
Subject: [PATCH 15/45] add sft script preview in webui

Former-commit-id: 6bc8e9866d482c945dd98f4e9ab205a7d7270755
---
 src/llmtuner/extras/constants.py     |   2 +
 src/llmtuner/webui/components/sft.py |  46 +++++++++++-
 src/llmtuner/webui/locales.py        |  16 ++++
 src/llmtuner/webui/runner.py         | 105 +++++++++++++++++++++------
 4 files changed, 145 insertions(+), 24 deletions(-)

diff --git a/src/llmtuner/extras/constants.py b/src/llmtuner/extras/constants.py
index 8ee997bb..4e7101f7 100644
--- a/src/llmtuner/extras/constants.py
+++ b/src/llmtuner/extras/constants.py
@@ -1,5 +1,7 @@
 IGNORE_INDEX = -100
 
+SFT_SCRIPT_PREFIX = "CUDA_VISIBLE_DEVICES=0 python "
+
 LOG_FILE_NAME = "trainer_log.jsonl"
 
 VALUE_HEAD_FILE_NAME = "value_head.bin"
diff --git a/src/llmtuner/webui/components/sft.py b/src/llmtuner/webui/components/sft.py
index 05a6e530..e74ef5cf 100644
--- a/src/llmtuner/webui/components/sft.py
+++ b/src/llmtuner/webui/components/sft.py
@@ -61,11 +61,15 @@ def create_sft_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict[
             resume_lora_training = gr.Checkbox(value=True, scale=1)
 
     with gr.Row():
+        preview_script_btn = gr.Button()
         start_btn = gr.Button()
         stop_btn = gr.Button()
 
     with gr.Row():
         with gr.Column(scale=3):
+            with gr.Box():
+                preview_script_box = gr.Textbox()
+
             with gr.Row():
                 output_dir = gr.Textbox()
 
@@ -78,6 +82,44 @@ def create_sft_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict[
         with gr.Column(scale=1):
             loss_viewer = gr.Plot()
 
+    preview_script_btn.click(
+        runner.preview_sft_script,
+        [
+            top_elems["lang"],
+            top_elems["model_name"],
+            top_elems["checkpoints"],
+            top_elems["finetuning_type"],
+            top_elems["quantization_bit"],
+            top_elems["template"],
+            top_elems["source_prefix"],
+            dataset_dir,
+            dataset,
+            max_source_length,
+            max_target_length,
+            learning_rate,
+            num_train_epochs,
+            max_samples,
+            batch_size,
+            gradient_accumulation_steps,
+            lr_scheduler_type,
+            max_grad_norm,
+            val_size,
+            logging_steps,
+            save_steps,
+            warmup_steps,
+            compute_type,
+            padding_side,
+            lora_rank,
+            lora_dropout,
+            lora_target,
+            resume_lora_training,
+            output_dir
+        ],
+        [
+            preview_script_box
+        ]
+    )
+
     start_btn.click(
         runner.run_train,
         [
@@ -154,5 +196,7 @@ def create_sft_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict[
         stop_btn=stop_btn,
         output_dir=output_dir,
         output_box=output_box,
-        loss_viewer=loss_viewer
+        loss_viewer=loss_viewer,
+        preview_script_btn=preview_script_btn,
+        preview_script_box=preview_script_box
     )
diff --git a/src/llmtuner/webui/locales.py b/src/llmtuner/webui/locales.py
index 61491ece..c4845735 100644
--- a/src/llmtuner/webui/locales.py
+++ b/src/llmtuner/webui/locales.py
@@ -500,6 +500,22 @@ LOCALES = {
         "zh": {
             "value": "开始导出"
         }
+    },
+    "preview_script_btn": {
+        "en": {
+            "value": "preview train script"
+        },
+        "zh": {
+            "value": "预览训练脚本命令"
+        }
+    },
+    "preview_script_box": {
+        "en": {
+            "label": "SFT Script Preview",
+        },
+        "zh": {
+            "label": "训练命令预览",
+        }
     }
 }
 
diff --git a/src/llmtuner/webui/runner.py b/src/llmtuner/webui/runner.py
index 1ae92786..d0d12d14 100644
--- a/src/llmtuner/webui/runner.py
+++ b/src/llmtuner/webui/runner.py
@@ -8,7 +8,7 @@ from transformers.trainer import TRAINING_ARGS_NAME
 from typing import Generator, List, Tuple
 
 from llmtuner.extras.callbacks import LogCallback
-from llmtuner.extras.constants import DEFAULT_MODULE
+from llmtuner.extras.constants import DEFAULT_MODULE, SFT_SCRIPT_PREFIX
 from llmtuner.extras.logging import LoggerHandler
 from llmtuner.extras.misc import torch_gc
 from llmtuner.tuner import run_exp
@@ -100,16 +100,44 @@ class Runner:
         if error:
             yield error, gr.update(visible=False)
             return
+        output_dir = os.path.join(get_save_dir(model_name), finetuning_type, output_dir)
 
+        args = self._build_args(batch_size, checkpoints, compute_type, dataset, dataset_dir, finetuning_type,
+                                gradient_accumulation_steps, learning_rate, logging_steps, lora_dropout, lora_rank,
+                                lora_target, lr_scheduler_type, max_grad_norm, max_samples, max_source_length,
+                                max_target_length, model_name, model_name_or_path, num_train_epochs, output_dir,
+                                padding_side, quantization_bit, resume_lora_training, save_steps, source_prefix,
+                                template, val_size, warmup_steps)
+
+        run_kwargs = dict(args=args, callbacks=[trainer_callback])
+        thread = threading.Thread(target=run_exp, kwargs=run_kwargs)
+        thread.start()
+
+        while thread.is_alive():
+            time.sleep(2)
+            if self.aborted:
+                yield ALERTS["info_aborting"][lang], gr.update(visible=False)
+            else:
+                yield logger_handler.log, update_process_bar(trainer_callback)
+
+        if os.path.exists(os.path.join(output_dir, TRAINING_ARGS_NAME)):
+            finish_info = ALERTS["info_finished"][lang]
+        else:
+            finish_info = ALERTS["err_failed"][lang]
+
+        yield self.finalize(lang, finish_info), gr.update(visible=False)
+
+    def _build_args(self, batch_size, checkpoints, compute_type, dataset, dataset_dir, finetuning_type,
+                    gradient_accumulation_steps, learning_rate, logging_steps, lora_dropout, lora_rank, lora_target,
+                    lr_scheduler_type, max_grad_norm, max_samples, max_source_length, max_target_length, model_name,
+                    model_name_or_path, num_train_epochs, output_dir, padding_side, quantization_bit,
+                    resume_lora_training, save_steps, source_prefix, template, val_size, warmup_steps):
         if checkpoints:
             checkpoint_dir = ",".join(
                 [os.path.join(get_save_dir(model_name), finetuning_type, checkpoint) for checkpoint in checkpoints]
             )
         else:
             checkpoint_dir = None
-
-        output_dir = os.path.join(get_save_dir(model_name), finetuning_type, output_dir)
-
         args = dict(
             stage="sft",
             model_name_or_path=model_name_or_path,
@@ -143,30 +171,12 @@ class Runner:
             resume_lora_training=resume_lora_training,
             output_dir=output_dir
         )
-
         if val_size > 1e-6:
             args["val_size"] = val_size
             args["evaluation_strategy"] = "steps"
             args["eval_steps"] = save_steps
             args["load_best_model_at_end"] = True
-
-        run_kwargs = dict(args=args, callbacks=[trainer_callback])
-        thread = threading.Thread(target=run_exp, kwargs=run_kwargs)
-        thread.start()
-
-        while thread.is_alive():
-            time.sleep(2)
-            if self.aborted:
-                yield ALERTS["info_aborting"][lang], gr.update(visible=False)
-            else:
-                yield logger_handler.log, update_process_bar(trainer_callback)
-
-        if os.path.exists(os.path.join(output_dir, TRAINING_ARGS_NAME)):
-            finish_info = ALERTS["info_finished"][lang]
-        else:
-            finish_info = ALERTS["err_failed"][lang]
-
-        yield self.finalize(lang, finish_info), gr.update(visible=False)
+        return args
 
     def run_eval(
         self,
@@ -240,3 +250,52 @@ class Runner:
             finish_info = ALERTS["err_failed"][lang]
 
         yield self.finalize(lang, finish_info), gr.update(visible=False)
+
+    def preview_sft_script(
+        self,
+        lang: str,
+        model_name: str,
+        checkpoints: List[str],
+        finetuning_type: str,
+        quantization_bit: str,
+        template: str,
+        source_prefix: str,
+        dataset_dir: str,
+        dataset: List[str],
+        max_source_length: int,
+        max_target_length: int,
+        learning_rate: str,
+        num_train_epochs: str,
+        max_samples: str,
+        batch_size: int,
+        gradient_accumulation_steps: int,
+        lr_scheduler_type: str,
+        max_grad_norm: str,
+        val_size: float,
+        logging_steps: int,
+        save_steps: int,
+        warmup_steps: int,
+        compute_type: str,
+        padding_side: str,
+        lora_rank: int,
+        lora_dropout: float,
+        lora_target: str,
+        resume_lora_training: bool,
+        output_dir: str
+    ):
+        model_name_or_path, error, logger_handler, trainer_callback = self.initialize(lang, model_name, dataset)
+        output_dir = os.path.join(get_save_dir(model_name), finetuning_type, output_dir)
+
+        args = self._build_args(batch_size, checkpoints, compute_type, dataset, dataset_dir, finetuning_type,
+                                gradient_accumulation_steps, learning_rate, logging_steps, lora_dropout, lora_rank,
+                                lora_target, lr_scheduler_type, max_grad_norm, max_samples, max_source_length,
+                                max_target_length, model_name, model_name_or_path, num_train_epochs, output_dir,
+                                padding_side, quantization_bit, resume_lora_training, save_steps, source_prefix,
+                                template, val_size, warmup_steps)
+        script_lines = [SFT_SCRIPT_PREFIX]
+        for param_key, param_value in args.items():
+            # filter None
+            if param_value:
+                script_lines.append("    --" + param_key + " " + str(param_value) + " ")
+        script_str = "\\\n".join(script_lines)
+        return gr.update(value=script_str)

From 3f0a2d6adc76bc60dfccd8147e22772d4b746a66 Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Sat, 12 Aug 2023 20:46:27 +0800
Subject: [PATCH 16/45] support rope scaling, fix #475 #476 #478

Former-commit-id: fa940c17b8d3e379af08804003f1a522c1cd6ac4
---
 src/llmtuner/chat/stream_chat.py      |   6 +-
 src/llmtuner/extras/constants.py      |   2 -
 src/llmtuner/extras/logging.py        |   3 +
 src/llmtuner/hparams/model_args.py    |  15 +-
 src/llmtuner/tuner/core/loader.py     |  63 +++++++-
 src/llmtuner/tuner/core/parser.py     |  17 ++-
 src/llmtuner/tuner/tune.py            |   2 +-
 src/llmtuner/webui/components/eval.py |  57 +++----
 src/llmtuner/webui/components/sft.py  | 131 ++++++----------
 src/llmtuner/webui/locales.py         |  30 ++--
 src/llmtuner/webui/runner.py          | 208 +++++++++++---------------
 src/llmtuner/webui/utils.py           |  10 ++
 12 files changed, 267 insertions(+), 277 deletions(-)

diff --git a/src/llmtuner/chat/stream_chat.py b/src/llmtuner/chat/stream_chat.py
index 8220c0b3..e207ee0b 100644
--- a/src/llmtuner/chat/stream_chat.py
+++ b/src/llmtuner/chat/stream_chat.py
@@ -1,8 +1,7 @@
 import torch
-from types import MethodType
 from typing import Any, Dict, Generator, List, Optional, Tuple
 from threading import Thread
-from transformers import PreTrainedModel, TextIteratorStreamer
+from transformers import TextIteratorStreamer
 
 from llmtuner.extras.misc import dispatch_model, get_logits_processor, get_stopping_criteria
 from llmtuner.extras.template import get_template_and_fix_tokenizer
@@ -15,10 +14,9 @@ class ChatModel:
         model_args, data_args, finetuning_args, self.generating_args = get_infer_args(args)
         self.model, self.tokenizer = load_model_and_tokenizer(model_args, finetuning_args)
         self.model = dispatch_model(self.model)
-        self.model = self.model.eval() # change to eval mode
+        self.model = self.model.eval() # enable evaluation mode
         self.template = get_template_and_fix_tokenizer(data_args.template, self.tokenizer)
         self.source_prefix = data_args.source_prefix
-        self.model.generate = MethodType(PreTrainedModel.generate, self.model) # disable custom method (for Qwen)
 
     def process_args(
         self,
diff --git a/src/llmtuner/extras/constants.py b/src/llmtuner/extras/constants.py
index 4e7101f7..8ee997bb 100644
--- a/src/llmtuner/extras/constants.py
+++ b/src/llmtuner/extras/constants.py
@@ -1,7 +1,5 @@
 IGNORE_INDEX = -100
 
-SFT_SCRIPT_PREFIX = "CUDA_VISIBLE_DEVICES=0 python "
-
 LOG_FILE_NAME = "trainer_log.jsonl"
 
 VALUE_HEAD_FILE_NAME = "value_head.bin"
diff --git a/src/llmtuner/extras/logging.py b/src/llmtuner/extras/logging.py
index 0b1a68f6..d6f185e6 100644
--- a/src/llmtuner/extras/logging.py
+++ b/src/llmtuner/extras/logging.py
@@ -8,6 +8,9 @@ class LoggerHandler(logging.Handler):
         super().__init__()
         self.log = ""
 
+    def reset(self):
+        self.log = ""
+
     def emit(self, record):
         if record.name == "httpx":
             return
diff --git a/src/llmtuner/hparams/model_args.py b/src/llmtuner/hparams/model_args.py
index 44454f45..dc515f51 100644
--- a/src/llmtuner/hparams/model_args.py
+++ b/src/llmtuner/hparams/model_args.py
@@ -43,9 +43,9 @@ class ModelArguments:
         default=True,
         metadata={"help": "Whether to use double quantization in int4 training or not."}
     )
-    compute_dtype: Optional[torch.dtype] = field(
+    rope_scaling: Optional[Literal["linear", "dynamic"]] = field(
         default=None,
-        metadata={"help": "Used in quantization configs. Do not specify this argument manually."}
+        metadata={"help": "Adopt scaled rotary positional embeddings."}
     )
     checkpoint_dir: Optional[str] = field(
         default=None,
@@ -63,8 +63,19 @@ class ModelArguments:
         default=None,
         metadata={"help": "Auth token to log in with Hugging Face Hub."}
     )
+    compute_dtype: Optional[torch.dtype] = field(
+        default=None,
+        metadata={"help": "Used in quantization configs. Do not specify this argument manually."}
+    )
+    model_max_length: Optional[int] = field(
+        default=None,
+        metadata={"help": "Used in rope scaling. Do not specify this argument manually."}
+    )
 
     def __post_init__(self):
+        if self.compute_dtype is not None or self.model_max_length is not None:
+            raise ValueError("These arguments cannot be specified.")
+
         if self.checkpoint_dir is not None: # support merging multiple lora weights
             self.checkpoint_dir = [cd.strip() for cd in self.checkpoint_dir.split(",")]
 
diff --git a/src/llmtuner/tuner/core/loader.py b/src/llmtuner/tuner/core/loader.py
index 74b4b59f..c03bdf63 100644
--- a/src/llmtuner/tuner/core/loader.py
+++ b/src/llmtuner/tuner/core/loader.py
@@ -1,5 +1,7 @@
 import os
+import math
 import torch
+from types import MethodType
 from typing import TYPE_CHECKING, Literal, Optional, Tuple
 
 from transformers import (
@@ -66,15 +68,56 @@ def load_model_and_tokenizer(
         **config_kwargs
     )
 
-    if model_args.checkpoint_dir is not None and finetuning_args.finetuning_type == "full":
+    if finetuning_args.finetuning_type == "full" and model_args.checkpoint_dir is not None:
         model_to_load = model_args.checkpoint_dir[0]
     else:
         model_to_load = model_args.model_name_or_path
 
     config = AutoConfig.from_pretrained(model_to_load, **config_kwargs)
-    is_mergeable = True
+
+    if hasattr(config, "fp16") and hasattr(config, "bf16"): # fix Qwen config
+        if model_args.compute_dtype == torch.bfloat16:
+            setattr(config, "bf16", True)
+        else:
+            setattr(config, "fp16", True)
+
+    # Set RoPE scaling
+    if model_args.rope_scaling is not None:
+        if hasattr(config, "use_dynamic_ntk"): # for Qwen models
+            if is_trainable:
+                logger.warning("Qwen model does not support rope scaling in training.")
+            else:
+                setattr(config, "use_dynamic_ntk", True)
+                setattr(config, "use_logn_attn", True)
+                logger.info("Using dynamic NTK scaling.")
+
+        elif hasattr(config, "rope_scaling"): # for LLaMA models
+            if is_trainable:
+                if model_args.rope_scaling == "dynamic":
+                    logger.warning(
+                        "Dynamic NTK may not work well with fine-tuning. "
+                        "See: https://github.com/huggingface/transformers/pull/24653"
+                    )
+
+                current_max_length = getattr(config, "max_position_embeddings", None)
+                if current_max_length and model_args.model_max_length <= current_max_length:
+                    logger.warning("Input length is smaller than max length. Consider increase input length.")
+                    scaling_factor = 1.0
+                else:
+                    scaling_factor = float(math.ceil(model_args.model_max_length / current_max_length))
+            else:
+                scaling_factor = 2.0
+
+            setattr(config, "rope_scaling", {"type": model_args.rope_scaling, "factor": scaling_factor})
+            logger.info("Using {} scaling strategy and setting scaling factor to {}".format(
+                model_args.rope_scaling, scaling_factor
+            ))
+
+        else:
+            logger.warning("Current model does not support RoPE scaling.")
 
     # Quantization configurations (using bitsandbytes library).
+    is_mergeable = True
     if model_args.quantization_bit is not None:
         if model_args.quantization_bit == 8:
             require_version("bitsandbytes>=0.37.0", "To fix: pip install bitsandbytes>=0.37.0")
@@ -95,7 +138,7 @@ def load_model_and_tokenizer(
         config_kwargs["device_map"] = {"": int(os.environ.get("LOCAL_RANK", "0"))} if is_trainable else "auto"
         logger.info("Quantizing model to {} bit.".format(model_args.quantization_bit))
 
-    # Load and prepare pretrained models (without valuehead).
+    # Load and prepare pre-trained models (without valuehead).
     model = AutoModelForCausalLM.from_pretrained(
         model_to_load,
         config=config,
@@ -104,6 +147,10 @@ def load_model_and_tokenizer(
         **config_kwargs
     )
 
+    # Disable custom generate method (for Qwen)
+    if "GenerationMixin" not in str(model.generate.__func__):
+        model.generate = MethodType(PreTrainedModel.generate, model)
+
     # Register auto class to save the custom code files.
     if isinstance(config, PretrainedConfig) and "AutoConfig" in getattr(config, "auto_map", {}):
         config.__class__.register_for_auto_class()
@@ -116,10 +163,10 @@ def load_model_and_tokenizer(
     model = prepare_model_for_training(model, finetuning_args.finetuning_type) if is_trainable else model
     model = init_adapter(model, model_args, finetuning_args, is_trainable, is_mergeable)
 
-    if stage == "rm" or stage == "ppo": # add value head
-        model: "AutoModelForCausalLMWithValueHead" = AutoModelForCausalLMWithValueHead.from_pretrained(model)
+    # Prepare model with valuehead for RLHF
+    if stage == "rm" or stage == "ppo":
+        model: AutoModelForCausalLMWithValueHead = AutoModelForCausalLMWithValueHead.from_pretrained(model)
         reset_logging()
-
         if stage == "rm" and model_args.checkpoint_dir is not None: # load valuehead weights to evaluate reward model
             logger.warning("Only the last checkpoint containing valuehead will be loaded as the valuehead.")
             if load_valuehead_params(model, model_args.checkpoint_dir[-1]):
@@ -133,9 +180,11 @@ def load_model_and_tokenizer(
             model.pretrained_model.load_adapter(model_args.reward_model, "reward", is_trainable=False)
             assert load_valuehead_params(model, model_args.reward_model), "Reward model is not correctly loaded."
 
+    # Prepare model for inference
     if not is_trainable:
         model.requires_grad_(False) # fix all model params
-        model = model.half() if model_args.quantization_bit is None else model # cast from fp32 to fp16
+        infer_dtype = torch.bfloat16 if torch.cuda.is_bf16_supported() else torch.float16 # detect cuda capability
+        model = model.to(infer_dtype) if model_args.quantization_bit is None else model
 
     trainable_params, all_param = count_parameters(model)
     logger.info("trainable params: {:d} || all params: {:d} || trainable%: {:.4f}".format(
diff --git a/src/llmtuner/tuner/core/parser.py b/src/llmtuner/tuner/core/parser.py
index c5bdbe16..7dda0a5c 100644
--- a/src/llmtuner/tuner/core/parser.py
+++ b/src/llmtuner/tuner/core/parser.py
@@ -151,13 +151,16 @@ def get_train_args(
 
     training_args.optim = "adamw_torch" if training_args.optim == "adamw_hf" else training_args.optim # suppress warning
 
-    if model_args.quantization_bit is not None:
-        if training_args.fp16:
-            model_args.compute_dtype = torch.float16
-        elif training_args.bf16:
-            model_args.compute_dtype = torch.bfloat16
-        else:
-            model_args.compute_dtype = torch.float32
+    if training_args.fp16:
+        model_args.compute_dtype = torch.float16
+    elif training_args.bf16:
+        if not torch.cuda.is_bf16_supported():
+            raise ValueError("Current device does not support bf16 training.")
+        model_args.compute_dtype = torch.bfloat16
+    else:
+        model_args.compute_dtype = torch.float32
+
+    model_args.model_max_length = data_args.max_source_length + data_args.max_target_length
 
     # Log on each process the small summary:
     logger.info("Process rank: {}, device: {}, n_gpu: {}\n  distributed training: {}, 16-bits training: {}".format(
diff --git a/src/llmtuner/tuner/tune.py b/src/llmtuner/tuner/tune.py
index 7b9446e9..a4a4c2a1 100644
--- a/src/llmtuner/tuner/tune.py
+++ b/src/llmtuner/tuner/tune.py
@@ -35,7 +35,7 @@ def run_exp(args: Optional[Dict[str, Any]] = None, callbacks: Optional[List["Tra
 
 
 def export_model(args: Optional[Dict[str, Any]] = None, max_shard_size: Optional[str] = "10GB"):
-    model_args, _, training_args, finetuning_args, _ = get_train_args(args)
+    model_args, _, training_args, finetuning_args, _, _ = get_train_args(args)
     model, tokenizer = load_model_and_tokenizer(model_args, finetuning_args)
     model.save_pretrained(training_args.output_dir, max_shard_size=max_shard_size)
     try:
diff --git a/src/llmtuner/webui/components/eval.py b/src/llmtuner/webui/components/eval.py
index ce6eddae..72d08a1f 100644
--- a/src/llmtuner/webui/components/eval.py
+++ b/src/llmtuner/webui/components/eval.py
@@ -14,13 +14,13 @@ def create_eval_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict
     with gr.Row():
         dataset_dir = gr.Textbox(value=DEFAULT_DATA_DIR, scale=2)
         dataset = gr.Dropdown(multiselect=True, scale=4)
-        preview_btn = gr.Button(interactive=False, scale=1)
+        data_preview_btn = gr.Button(interactive=False, scale=1)
 
     preview_box, preview_count, preview_samples, close_btn = create_preview_box()
 
     dataset_dir.change(list_dataset, [dataset_dir], [dataset])
-    dataset.change(can_preview, [dataset_dir, dataset], [preview_btn])
-    preview_btn.click(
+    dataset.change(can_preview, [dataset_dir, dataset], [data_preview_btn])
+    data_preview_btn.click(
         get_preview,
         [dataset_dir, dataset],
         [preview_count, preview_samples, preview_box],
@@ -35,6 +35,7 @@ def create_eval_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict
         predict = gr.Checkbox(value=True)
 
     with gr.Row():
+        cmd_preview_btn = gr.Button()
         start_btn = gr.Button()
         stop_btn = gr.Button()
 
@@ -44,35 +45,36 @@ def create_eval_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict
     with gr.Box():
         output_box = gr.Markdown()
 
-    start_btn.click(
-        runner.run_eval,
-        [
-            top_elems["lang"],
-            top_elems["model_name"],
-            top_elems["checkpoints"],
-            top_elems["finetuning_type"],
-            top_elems["quantization_bit"],
-            top_elems["template"],
-            top_elems["source_prefix"],
-            dataset_dir,
-            dataset,
-            max_source_length,
-            max_target_length,
-            max_samples,
-            batch_size,
-            predict
-        ],
-        [
-            output_box,
-            process_bar
-        ]
-    )
+    input_list = [
+        top_elems["lang"],
+        top_elems["model_name"],
+        top_elems["checkpoints"],
+        top_elems["finetuning_type"],
+        top_elems["quantization_bit"],
+        top_elems["template"],
+        top_elems["source_prefix"],
+        dataset_dir,
+        dataset,
+        max_source_length,
+        max_target_length,
+        max_samples,
+        batch_size,
+        predict
+    ]
+
+    output_list = [
+        output_box,
+        process_bar
+    ]
+
+    cmd_preview_btn.click(runner.preview_eval, input_list, output_list)
+    start_btn.click(runner.run_eval, input_list, output_list)
     stop_btn.click(runner.set_abort, queue=False)
 
     return dict(
         dataset_dir=dataset_dir,
         dataset=dataset,
-        preview_btn=preview_btn,
+        data_preview_btn=data_preview_btn,
         preview_count=preview_count,
         preview_samples=preview_samples,
         close_btn=close_btn,
@@ -81,6 +83,7 @@ def create_eval_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict
         max_samples=max_samples,
         batch_size=batch_size,
         predict=predict,
+        cmd_preview_btn=cmd_preview_btn,
         start_btn=start_btn,
         stop_btn=stop_btn,
         output_box=output_box
diff --git a/src/llmtuner/webui/components/sft.py b/src/llmtuner/webui/components/sft.py
index e74ef5cf..a4101d15 100644
--- a/src/llmtuner/webui/components/sft.py
+++ b/src/llmtuner/webui/components/sft.py
@@ -16,13 +16,13 @@ def create_sft_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict[
     with gr.Row():
         dataset_dir = gr.Textbox(value=DEFAULT_DATA_DIR, scale=2)
         dataset = gr.Dropdown(multiselect=True, scale=4)
-        preview_btn = gr.Button(interactive=False, scale=1)
+        data_preview_btn = gr.Button(interactive=False, scale=1)
 
     preview_box, preview_count, preview_samples, close_btn = create_preview_box()
 
     dataset_dir.change(list_dataset, [dataset_dir], [dataset])
-    dataset.change(can_preview, [dataset_dir, dataset], [preview_btn])
-    preview_btn.click(
+    dataset.change(can_preview, [dataset_dir, dataset], [data_preview_btn])
+    data_preview_btn.click(
         get_preview,
         [dataset_dir, dataset],
         [preview_count, preview_samples, preview_box],
@@ -61,15 +61,12 @@ def create_sft_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict[
             resume_lora_training = gr.Checkbox(value=True, scale=1)
 
     with gr.Row():
-        preview_script_btn = gr.Button()
+        cmd_preview_btn = gr.Button()
         start_btn = gr.Button()
         stop_btn = gr.Button()
 
     with gr.Row():
         with gr.Column(scale=3):
-            with gr.Box():
-                preview_script_box = gr.Textbox()
-
             with gr.Row():
                 output_dir = gr.Textbox()
 
@@ -82,82 +79,45 @@ def create_sft_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict[
         with gr.Column(scale=1):
             loss_viewer = gr.Plot()
 
-    preview_script_btn.click(
-        runner.preview_sft_script,
-        [
-            top_elems["lang"],
-            top_elems["model_name"],
-            top_elems["checkpoints"],
-            top_elems["finetuning_type"],
-            top_elems["quantization_bit"],
-            top_elems["template"],
-            top_elems["source_prefix"],
-            dataset_dir,
-            dataset,
-            max_source_length,
-            max_target_length,
-            learning_rate,
-            num_train_epochs,
-            max_samples,
-            batch_size,
-            gradient_accumulation_steps,
-            lr_scheduler_type,
-            max_grad_norm,
-            val_size,
-            logging_steps,
-            save_steps,
-            warmup_steps,
-            compute_type,
-            padding_side,
-            lora_rank,
-            lora_dropout,
-            lora_target,
-            resume_lora_training,
-            output_dir
-        ],
-        [
-            preview_script_box
-        ]
-    )
+    input_list = [
+        top_elems["lang"],
+        top_elems["model_name"],
+        top_elems["checkpoints"],
+        top_elems["finetuning_type"],
+        top_elems["quantization_bit"],
+        top_elems["template"],
+        top_elems["source_prefix"],
+        dataset_dir,
+        dataset,
+        max_source_length,
+        max_target_length,
+        learning_rate,
+        num_train_epochs,
+        max_samples,
+        batch_size,
+        gradient_accumulation_steps,
+        lr_scheduler_type,
+        max_grad_norm,
+        val_size,
+        logging_steps,
+        save_steps,
+        warmup_steps,
+        compute_type,
+        padding_side,
+        lora_rank,
+        lora_dropout,
+        lora_target,
+        resume_lora_training,
+        output_dir
+    ]
 
-    start_btn.click(
-        runner.run_train,
-        [
-            top_elems["lang"],
-            top_elems["model_name"],
-            top_elems["checkpoints"],
-            top_elems["finetuning_type"],
-            top_elems["quantization_bit"],
-            top_elems["template"],
-            top_elems["source_prefix"],
-            dataset_dir,
-            dataset,
-            max_source_length,
-            max_target_length,
-            learning_rate,
-            num_train_epochs,
-            max_samples,
-            batch_size,
-            gradient_accumulation_steps,
-            lr_scheduler_type,
-            max_grad_norm,
-            val_size,
-            logging_steps,
-            save_steps,
-            warmup_steps,
-            compute_type,
-            padding_side,
-            lora_rank,
-            lora_dropout,
-            lora_target,
-            resume_lora_training,
-            output_dir
-        ],
-        [
-            output_box,
-            process_bar
-        ]
-    )
+    output_list = [
+        output_box,
+        process_bar
+    ]
+
+    cmd_preview_btn.click(runner.preview_train, input_list, output_list)
+    start_btn.click(runner.run_train, input_list, output_list)
     stop_btn.click(runner.set_abort, queue=False)
 
     process_bar.change(
@@ -167,7 +127,7 @@ def create_sft_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict[
     return dict(
         dataset_dir=dataset_dir,
         dataset=dataset,
-        preview_btn=preview_btn,
+        data_preview_btn=data_preview_btn,
         preview_count=preview_count,
         preview_samples=preview_samples,
         close_btn=close_btn,
@@ -192,11 +152,10 @@ def create_sft_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict[
         lora_dropout=lora_dropout,
         lora_target=lora_target,
         resume_lora_training=resume_lora_training,
+        cmd_preview_btn=cmd_preview_btn,
         start_btn=start_btn,
         stop_btn=stop_btn,
         output_dir=output_dir,
         output_box=output_box,
-        loss_viewer=loss_viewer,
-        preview_script_btn=preview_script_btn,
-        preview_script_box=preview_script_box
+        loss_viewer=loss_viewer
     )
diff --git a/src/llmtuner/webui/locales.py b/src/llmtuner/webui/locales.py
index c4845735..aba4acae 100644
--- a/src/llmtuner/webui/locales.py
+++ b/src/llmtuner/webui/locales.py
@@ -105,12 +105,12 @@ LOCALES = {
             "label": "数据集"
         }
     },
-    "preview_btn": {
+    "data_preview_btn": {
         "en": {
-            "value": "Preview"
+            "value": "Preview dataset"
         },
         "zh": {
-            "value": "预览"
+            "value": "预览数据集"
         }
     },
     "preview_count": {
@@ -335,6 +335,14 @@ LOCALES = {
             "info": "接着上次的 LoRA 权重训练或创建一个新的 LoRA 权重。"
         }
     },
+    "cmd_preview_btn": {
+        "en": {
+            "value": "Preview command"
+        },
+        "zh": {
+            "value": "预览命令"
+        }
+    },
     "start_btn": {
         "en": {
             "value": "Start"
@@ -500,22 +508,6 @@ LOCALES = {
         "zh": {
             "value": "开始导出"
         }
-    },
-    "preview_script_btn": {
-        "en": {
-            "value": "preview train script"
-        },
-        "zh": {
-            "value": "预览训练脚本命令"
-        }
-    },
-    "preview_script_box": {
-        "en": {
-            "label": "SFT Script Preview",
-        },
-        "zh": {
-            "label": "训练命令预览",
-        }
     }
 }
 
diff --git a/src/llmtuner/webui/runner.py b/src/llmtuner/webui/runner.py
index d0d12d14..aab31349 100644
--- a/src/llmtuner/webui/runner.py
+++ b/src/llmtuner/webui/runner.py
@@ -5,16 +5,16 @@ import threading
 import time
 import transformers
 from transformers.trainer import TRAINING_ARGS_NAME
-from typing import Generator, List, Tuple
+from typing import Any, Dict, Generator, List, Tuple
 
 from llmtuner.extras.callbacks import LogCallback
-from llmtuner.extras.constants import DEFAULT_MODULE, SFT_SCRIPT_PREFIX
+from llmtuner.extras.constants import DEFAULT_MODULE
 from llmtuner.extras.logging import LoggerHandler
 from llmtuner.extras.misc import torch_gc
 from llmtuner.tuner import run_exp
 from llmtuner.webui.common import get_model_path, get_save_dir
 from llmtuner.webui.locales import ALERTS
-from llmtuner.webui.utils import get_eval_results, update_process_bar
+from llmtuner.webui.utils import gen_cmd, get_eval_results, update_process_bar
 
 
 class Runner:
@@ -22,39 +22,36 @@ class Runner:
     def __init__(self):
         self.aborted = False
         self.running = False
+        self.logger_handler = LoggerHandler()
+        self.logger_handler.setLevel(logging.INFO)
+        logging.root.addHandler(self.logger_handler)
+        transformers.logging.add_handler(self.logger_handler)
 
     def set_abort(self):
         self.aborted = True
         self.running = False
 
-    def initialize(
+    def _initialize(
         self, lang: str, model_name: str, dataset: List[str]
-    ) -> Tuple[str, str, LoggerHandler, LogCallback]:
+    ) -> str:
         if self.running:
-            return None, ALERTS["err_conflict"][lang], None, None
+            return ALERTS["err_conflict"][lang]
 
         if not model_name:
-            return None, ALERTS["err_no_model"][lang], None, None
+            return ALERTS["err_no_model"][lang]
 
-        model_name_or_path = get_model_path(model_name)
-        if not model_name_or_path:
-            return None, ALERTS["err_no_path"][lang], None, None
+        if not get_model_path(model_name):
+            return ALERTS["err_no_path"][lang]
 
         if len(dataset) == 0:
-            return None, ALERTS["err_no_dataset"][lang], None, None
+            return ALERTS["err_no_dataset"][lang]
 
         self.aborted = False
-        self.running = True
+        self.logger_handler.reset()
+        self.trainer_callback = LogCallback(self)
+        return ""
 
-        logger_handler = LoggerHandler()
-        logger_handler.setLevel(logging.INFO)
-        logging.root.addHandler(logger_handler)
-        transformers.logging.add_handler(logger_handler)
-        trainer_callback = LogCallback(self)
-
-        return model_name_or_path, "", logger_handler, trainer_callback
-
-    def finalize(
+    def _finalize(
         self, lang: str, finish_info: str
     ) -> str:
         self.running = False
@@ -64,7 +61,7 @@ class Runner:
         else:
             return finish_info
 
-    def run_train(
+    def _parse_train_args(
         self,
         lang: str,
         model_name: str,
@@ -95,52 +92,19 @@ class Runner:
         lora_target: str,
         resume_lora_training: bool,
         output_dir: str
-    ) -> Generator[str, None, None]:
-        model_name_or_path, error, logger_handler, trainer_callback = self.initialize(lang, model_name, dataset)
-        if error:
-            yield error, gr.update(visible=False)
-            return
-        output_dir = os.path.join(get_save_dir(model_name), finetuning_type, output_dir)
-
-        args = self._build_args(batch_size, checkpoints, compute_type, dataset, dataset_dir, finetuning_type,
-                                gradient_accumulation_steps, learning_rate, logging_steps, lora_dropout, lora_rank,
-                                lora_target, lr_scheduler_type, max_grad_norm, max_samples, max_source_length,
-                                max_target_length, model_name, model_name_or_path, num_train_epochs, output_dir,
-                                padding_side, quantization_bit, resume_lora_training, save_steps, source_prefix,
-                                template, val_size, warmup_steps)
-
-        run_kwargs = dict(args=args, callbacks=[trainer_callback])
-        thread = threading.Thread(target=run_exp, kwargs=run_kwargs)
-        thread.start()
-
-        while thread.is_alive():
-            time.sleep(2)
-            if self.aborted:
-                yield ALERTS["info_aborting"][lang], gr.update(visible=False)
-            else:
-                yield logger_handler.log, update_process_bar(trainer_callback)
-
-        if os.path.exists(os.path.join(output_dir, TRAINING_ARGS_NAME)):
-            finish_info = ALERTS["info_finished"][lang]
-        else:
-            finish_info = ALERTS["err_failed"][lang]
-
-        yield self.finalize(lang, finish_info), gr.update(visible=False)
-
-    def _build_args(self, batch_size, checkpoints, compute_type, dataset, dataset_dir, finetuning_type,
-                    gradient_accumulation_steps, learning_rate, logging_steps, lora_dropout, lora_rank, lora_target,
-                    lr_scheduler_type, max_grad_norm, max_samples, max_source_length, max_target_length, model_name,
-                    model_name_or_path, num_train_epochs, output_dir, padding_side, quantization_bit,
-                    resume_lora_training, save_steps, source_prefix, template, val_size, warmup_steps):
+    ) -> Tuple[str, str, List[str], str, Dict[str, Any]]:
         if checkpoints:
             checkpoint_dir = ",".join(
-                [os.path.join(get_save_dir(model_name), finetuning_type, checkpoint) for checkpoint in checkpoints]
+                [os.path.join(get_save_dir(model_name), finetuning_type, ckpt) for ckpt in checkpoints]
             )
         else:
             checkpoint_dir = None
+
+        output_dir = os.path.join(get_save_dir(model_name), finetuning_type, output_dir)
+
         args = dict(
             stage="sft",
-            model_name_or_path=model_name_or_path,
+            model_name_or_path=get_model_path(model_name),
             do_train=True,
             overwrite_cache=True,
             checkpoint_dir=checkpoint_dir,
@@ -171,14 +135,16 @@ class Runner:
             resume_lora_training=resume_lora_training,
             output_dir=output_dir
         )
+
         if val_size > 1e-6:
             args["val_size"] = val_size
             args["evaluation_strategy"] = "steps"
             args["eval_steps"] = save_steps
             args["load_best_model_at_end"] = True
-        return args
 
-    def run_eval(
+        return lang, model_name, dataset, output_dir, args
+
+    def _parse_eval_args(
         self,
         lang: str,
         model_name: str,
@@ -194,12 +160,7 @@ class Runner:
         max_samples: str,
         batch_size: int,
         predict: bool
-    ) -> Generator[str, None, None]:
-        model_name_or_path, error, logger_handler, trainer_callback = self.initialize(lang, model_name, dataset)
-        if error:
-            yield error, gr.update(visible=False)
-            return
-
+    ) -> Tuple[str, str, List[str], str, Dict[str, Any]]:
         if checkpoints:
             checkpoint_dir = ",".join(
                 [os.path.join(get_save_dir(model_name), finetuning_type, checkpoint) for checkpoint in checkpoints]
@@ -211,7 +172,7 @@ class Runner:
 
         args = dict(
             stage="sft",
-            model_name_or_path=model_name_or_path,
+            model_name_or_path=get_model_path(model_name),
             do_eval=True,
             overwrite_cache=True,
             predict_with_generate=True,
@@ -233,7 +194,33 @@ class Runner:
             args.pop("do_eval", None)
             args["do_predict"] = True
 
-        run_kwargs = dict(args=args, callbacks=[trainer_callback])
+        return lang, model_name, dataset, output_dir, args
+
+    def preview_train(self, *args) -> Generator[Tuple[str, Dict[str, Any]], None, None]:
+        lang, model_name, dataset, _, args = self._parse_train_args(*args)
+        error = self._initialize(lang, model_name, dataset)
+        if error:
+            yield error, gr.update(visible=False)
+        else:
+            yield gen_cmd(args), gr.update(visible=False)
+
+    def preview_eval(self, *args) -> Generator[Tuple[str, Dict[str, Any]], None, None]:
+        lang, model_name, dataset, _, args = self._parse_eval_args(*args)
+        error = self._initialize(lang, model_name, dataset)
+        if error:
+            yield error, gr.update(visible=False)
+        else:
+            yield gen_cmd(args), gr.update(visible=False)
+
+    def run_train(self, *args) -> Generator[Tuple[str, Dict[str, Any]], None, None]:
+        lang, model_name, dataset, output_dir, args = self._parse_train_args(*args)
+        error = self._initialize(lang, model_name, dataset)
+        if error:
+            yield error, gr.update(visible=False)
+            return
+
+        self.running = True
+        run_kwargs = dict(args=args, callbacks=[self.trainer_callback])
         thread = threading.Thread(target=run_exp, kwargs=run_kwargs)
         thread.start()
 
@@ -242,60 +229,37 @@ class Runner:
             if self.aborted:
                 yield ALERTS["info_aborting"][lang], gr.update(visible=False)
             else:
-                yield logger_handler.log, update_process_bar(trainer_callback)
+                yield self.logger_handler.log, update_process_bar(self.trainer_callback)
+
+        if os.path.exists(os.path.join(output_dir, TRAINING_ARGS_NAME)):
+            finish_info = ALERTS["info_finished"][lang]
+        else:
+            finish_info = ALERTS["err_failed"][lang]
+
+        yield self._finalize(lang, finish_info), gr.update(visible=False)
+
+    def run_eval(self, *args) -> Generator[str, None, None]:
+        lang, model_name, dataset, output_dir, args = self._parse_eval_args(*args)
+        error = self._initialize(lang, model_name, dataset)
+        if error:
+            yield error, gr.update(visible=False)
+            return
+
+        self.running = True
+        run_kwargs = dict(args=args, callbacks=[self.trainer_callback])
+        thread = threading.Thread(target=run_exp, kwargs=run_kwargs)
+        thread.start()
+
+        while thread.is_alive():
+            time.sleep(2)
+            if self.aborted:
+                yield ALERTS["info_aborting"][lang], gr.update(visible=False)
+            else:
+                yield self.logger_handler.log, update_process_bar(self.trainer_callback)
 
         if os.path.exists(os.path.join(output_dir, "all_results.json")):
             finish_info = get_eval_results(os.path.join(output_dir, "all_results.json"))
         else:
             finish_info = ALERTS["err_failed"][lang]
 
-        yield self.finalize(lang, finish_info), gr.update(visible=False)
-
-    def preview_sft_script(
-        self,
-        lang: str,
-        model_name: str,
-        checkpoints: List[str],
-        finetuning_type: str,
-        quantization_bit: str,
-        template: str,
-        source_prefix: str,
-        dataset_dir: str,
-        dataset: List[str],
-        max_source_length: int,
-        max_target_length: int,
-        learning_rate: str,
-        num_train_epochs: str,
-        max_samples: str,
-        batch_size: int,
-        gradient_accumulation_steps: int,
-        lr_scheduler_type: str,
-        max_grad_norm: str,
-        val_size: float,
-        logging_steps: int,
-        save_steps: int,
-        warmup_steps: int,
-        compute_type: str,
-        padding_side: str,
-        lora_rank: int,
-        lora_dropout: float,
-        lora_target: str,
-        resume_lora_training: bool,
-        output_dir: str
-    ):
-        model_name_or_path, error, logger_handler, trainer_callback = self.initialize(lang, model_name, dataset)
-        output_dir = os.path.join(get_save_dir(model_name), finetuning_type, output_dir)
-
-        args = self._build_args(batch_size, checkpoints, compute_type, dataset, dataset_dir, finetuning_type,
-                                gradient_accumulation_steps, learning_rate, logging_steps, lora_dropout, lora_rank,
-                                lora_target, lr_scheduler_type, max_grad_norm, max_samples, max_source_length,
-                                max_target_length, model_name, model_name_or_path, num_train_epochs, output_dir,
-                                padding_side, quantization_bit, resume_lora_training, save_steps, source_prefix,
-                                template, val_size, warmup_steps)
-        script_lines = [SFT_SCRIPT_PREFIX]
-        for param_key, param_value in args.items():
-            # filter None
-            if param_value:
-                script_lines.append("    --" + param_key + " " + str(param_value) + " ")
-        script_str = "\\\n".join(script_lines)
-        return gr.update(value=script_str)
+        yield self._finalize(lang, finish_info), gr.update(visible=False)
diff --git a/src/llmtuner/webui/utils.py b/src/llmtuner/webui/utils.py
index fb22bd0c..6ca55e14 100644
--- a/src/llmtuner/webui/utils.py
+++ b/src/llmtuner/webui/utils.py
@@ -62,6 +62,16 @@ def can_quantize(finetuning_type: str) -> Dict[str, Any]:
         return gr.update(interactive=True)
 
 
+def gen_cmd(args: Dict[str, Any]) -> str:
+    cmd_lines = ["CUDA_VISIBLE_DEVICES=0 python "]
+    for k, v in args.items():
+        if v is not None and v is not False and v != "":
+            cmd_lines.append("    --{} {} ".format(k, str(v)))
+    cmd_text = "\\\n".join(cmd_lines)
+    cmd_text = "```bash\n{}\n```".format(cmd_text)
+    return cmd_text
+
+
 def get_eval_results(path: os.PathLike) -> str:
     with open(path, "r", encoding="utf-8") as f:
         result = json.dumps(json.load(f), indent=4)

From ba65dcb15e11d1359f1aec57b0ad30b39b69b05d Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Sat, 12 Aug 2023 21:00:11 +0800
Subject: [PATCH 17/45] update readme

Former-commit-id: 1836c020c514e7a94aaa48abdf19ea8accbc1a2a
---
 README.md                         | 2 ++
 README_zh.md                      | 2 ++
 src/llmtuner/tuner/core/loader.py | 2 ++
 3 files changed, 6 insertions(+)

diff --git a/README.md b/README.md
index a54ac30a..3d617f47 100644
--- a/README.md
+++ b/README.md
@@ -12,6 +12,8 @@
 
 ## Changelog
 
+[23/08/12] Now we support **RoPE scaling** to extend the context length of the LLaMA models. Try `--rope_scaling linear` argument in training and `--rope_scaling dynamic` argument at inference to extrapolate the position embeddings.
+
 [23/08/11] Now we support **[DPO training](https://arxiv.org/abs/2305.18290)** for instruction-tuned models. See [this example](#dpo-training) to train your models (experimental feature).
 
 [23/08/03] Now we support training the **Qwen-7B** model in this repo. Try `--model_name_or_path Qwen/Qwen-7B-Chat` and `--lora_target c_attn` arguments to train the Qwen-7B model. Remember to use `--template chatml` argument when you are using the Qwen-7B-Chat model.
diff --git a/README_zh.md b/README_zh.md
index f9d967c5..5aa9a7d6 100644
--- a/README_zh.md
+++ b/README_zh.md
@@ -12,6 +12,8 @@
 
 ## 更新日志
 
+[23/08/12] 现在我们支持了 **RoPE 插值**来扩展 LLaMA 模型的上下文长度。请尝试使用 `--rope_scaling linear` 参数训练模型或使用 `--rope_scaling dynamic` 参数评估模型。
+
 [23/08/11] 现在我们支持了指令模型的 **[DPO 训练](https://arxiv.org/abs/2305.18290)**。详情请参阅[此示例](#dpo-训练)（实验性功能）。
 
 [23/08/03] 现在我们支持了 **Qwen-7B** 模型的训练。请尝试使用 `--model_name_or_path Qwen/Qwen-7B-Chat` 和 `--lora_target c_attn` 参数。使用 Qwen-7B-Chat 模型请添加 `--template chatml` 参数。
diff --git a/src/llmtuner/tuner/core/loader.py b/src/llmtuner/tuner/core/loader.py
index c03bdf63..32d6247a 100644
--- a/src/llmtuner/tuner/core/loader.py
+++ b/src/llmtuner/tuner/core/loader.py
@@ -83,6 +83,8 @@ def load_model_and_tokenizer(
 
     # Set RoPE scaling
     if model_args.rope_scaling is not None:
+        require_version("transformers>=4.31.0", "RoPE scaling requires transformers>=4.31.0")
+
         if hasattr(config, "use_dynamic_ntk"): # for Qwen models
             if is_trainable:
                 logger.warning("Qwen model does not support rope scaling in training.")

From 8686e62dfa75ce864a66c5c92e5288c3dd2e7b4e Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Sat, 12 Aug 2023 21:23:05 +0800
Subject: [PATCH 18/45] update readme

Former-commit-id: 2618e0b5a7ad88f68971f21d0e7eb4560866400f
---
 README.md    | 52 +++++++++++++++++++++++++++++++++++++-----
 README_zh.md | 64 +++++++++++++++++++++++++++++++++++++++++-----------
 2 files changed, 97 insertions(+), 19 deletions(-)

diff --git a/README.md b/README.md
index 3d617f47..fc7ff69e 100644
--- a/README.md
+++ b/README.md
@@ -18,17 +18,17 @@
 
 [23/08/03] Now we support training the **Qwen-7B** model in this repo. Try `--model_name_or_path Qwen/Qwen-7B-Chat` and `--lora_target c_attn` arguments to train the Qwen-7B model. Remember to use `--template chatml` argument when you are using the Qwen-7B-Chat model.
 
-[23/07/31] Now we support dataset streaming. Try `--streaming` and `--max_steps 100` arguments to stream your dataset.
+[23/07/31] Now we support **dataset streaming**. Try `--streaming` and `--max_steps 10000` arguments to load your dataset in streaming mode.
 
 [23/07/29] We release two instruction-tuned 13B models at Hugging Face. See these Hugging Face Repos ([LLaMA-2](https://huggingface.co/hiyouga/Llama-2-Chinese-13b-chat) / [Baichuan](https://huggingface.co/hiyouga/baichuan-13b-sft)) for details.
 
 [23/07/19] Now we support training the **LLaMA-2** models in this repo. Try `--model_name_or_path meta-llama/Llama-2-7b-hf` argument to use the LLaMA-2 model. Remember to use `--template llama2` argument when you are using the LLaMA-2-chat model.
 
-[23/07/18] Now we develop an all-in-one Web UI for training, evaluation and inference. Try `train_web.py` to fine-tune models in your Web browser. Thank [@KanadeSiina](https://github.com/KanadeSiina) and [@codemayq](https://github.com/codemayq) for their efforts in the development.
+[23/07/18] Now we develop an **all-in-one Web UI** for training, evaluation and inference. Try `train_web.py` to fine-tune models in your Web browser. Thank [@KanadeSiina](https://github.com/KanadeSiina) and [@codemayq](https://github.com/codemayq) for their efforts in the development.
 
 [23/07/11] Now we support training the **Baichuan-13B** model in this repo. Try `--model_name_or_path baichuan-inc/Baichuan-13B-Base` and `--lora_target W_pack` arguments to train the Baichuan-13B model. Remember to use `--template baichuan` argument when you are using the Baichuan-13B-Chat model.
 
-[23/07/09] Now we release [FastEdit](https://github.com/hiyouga/FastEdit)⚡🩹, an easy-to-use package for editing the factual knowledge of large language models efficiently. Please follow [FastEdit](https://github.com/hiyouga/FastEdit) if you are interested.
+[23/07/09] Now we release **[FastEdit](https://github.com/hiyouga/FastEdit)**⚡🩹, an easy-to-use package for editing the factual knowledge of large language models efficiently. Please follow [FastEdit](https://github.com/hiyouga/FastEdit) if you are interested.
 
 [23/07/07] Now we support training the **InternLM-7B** model in this repo. Try `--model_name_or_path internlm/internlm-7b` argument to use the InternLM model. Remember to use `--template intern` argument when you are using the InternLM-chat model.
 
@@ -72,6 +72,8 @@
 | PPO Training           |                |                   | ✅   | ✅   |
 | DPO Training           | ✅            |                    | ✅   | ✅   |
 
+- Use `--quantization_bit 4/8` argument to enable QLoRA.
+
 ## Provided Datasets
 
 - For pre-training:
@@ -204,8 +206,6 @@ CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
     --fp16
 ```
 
-Remember to specify `--lora_target W_pack` if you are using Baichuan models.
-
 ### Reward Model Training
 
 ```bash
@@ -280,12 +280,14 @@ CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
 
 ### Distributed Training
 
+#### Use Huggingface Accelerate
+
 ```bash
 accelerate config # configure the environment
 accelerate launch src/train_bash.py # arguments (same as above)
 ```
 
-<details><summary>Example configuration for full-tuning with DeepSpeed ZeRO-2</summary>
+<details><summary>Example config.yaml for training with DeepSpeed ZeRO-2</summary>
 
 ```yaml
 compute_environment: LOCAL_MACHINE
@@ -313,6 +315,44 @@ use_cpu: false
 
 </details>
 
+#### Use DeepSpeed
+
+```bash
+deepspeed --num_gpus 8 --master_port=9901 src/train_bash.py \
+    --deepspeed ds_config.json \
+    ... # arguments (same as above)
+```
+
+<details><summary>Example ds_config.json for training with DeepSpeed ZeRO-2</summary>
+
+```json
+{
+  "train_micro_batch_size_per_gpu": "auto",
+  "gradient_accumulation_steps": "auto",
+  "gradient_clipping": "auto",
+  "zero_allow_untested_optimizer": true,
+  "fp16": {
+    "enabled": "auto",
+    "loss_scale": 0,
+    "initial_scale_power": 16,
+    "loss_scale_window": 1000,
+    "hysteresis": 2,
+    "min_loss_scale": 1
+  },  
+  "zero_optimization": {
+    "stage": 2,
+    "allgather_partitions": true,
+    "allgather_bucket_size": 5e8,
+    "reduce_scatter": true,
+    "reduce_bucket_size": 5e8,
+    "overlap_comm": false,
+    "contiguous_gradients": true
+  }
+}
+```
+
+</details>
+
 ### Evaluation (BLEU and ROUGE_CHINESE)
 
 ```bash
diff --git a/README_zh.md b/README_zh.md
index 5aa9a7d6..c2f83504 100644
--- a/README_zh.md
+++ b/README_zh.md
@@ -16,31 +16,31 @@
 
 [23/08/11] 现在我们支持了指令模型的 **[DPO 训练](https://arxiv.org/abs/2305.18290)**。详情请参阅[此示例](#dpo-训练)（实验性功能）。
 
-[23/08/03] 现在我们支持了 **Qwen-7B** 模型的训练。请尝试使用 `--model_name_or_path Qwen/Qwen-7B-Chat` 和 `--lora_target c_attn` 参数。使用 Qwen-7B-Chat 模型请添加 `--template chatml` 参数。
+[23/08/03] 现在我们支持了 **Qwen-7B** 模型的训练。请尝试使用 `--model_name_or_path Qwen/Qwen-7B-Chat` 和 `--lora_target c_attn` 参数。使用 Qwen-7B-Chat 模型时请添加 `--template chatml` 参数。
 
-[23/07/31] 现在我们支持了训练数据流式加载。请尝试使用 `--streaming` 和 `--max_steps 100` 参数来流式加载数据集。
+[23/07/31] 现在我们支持了**数据流式加载**。请尝试使用 `--streaming` 和 `--max_steps 10000` 参数来流式加载数据集。
 
 [23/07/29] 我们在 Hugging Face 发布了两个 13B 指令微调模型。详细内容请查阅我们的 Hugging Face 项目（[LLaMA-2](https://huggingface.co/hiyouga/Llama-2-Chinese-13b-chat) / [Baichuan](https://huggingface.co/hiyouga/baichuan-13b-sft)）。
 
-[23/07/19] 现在我们支持了 **LLaMA-2** 模型的训练。请尝试使用 `--model_name_or_path meta-llama/Llama-2-7b-hf` 参数。请注意使用 LLaMA-2-chat 模型需要添加 `--template llama2` 参数。
+[23/07/19] 现在我们支持了 **LLaMA-2** 模型的训练。请尝试使用 `--model_name_or_path meta-llama/Llama-2-7b-hf` 参数。使用 LLaMA-2-chat 模型时请添加 `--template llama2` 参数。
 
-[23/07/18] 我们开发了支持训练和测试的浏览器一键微调界面。请尝试使用 `train_web.py` 在您的浏览器中微调模型。感谢 [@KanadeSiina](https://github.com/KanadeSiina) 和 [@codemayq](https://github.com/codemayq) 在该功能开发中付出的努力。
+[23/07/18] 我们开发了支持训练和测试的**一体化浏览器界面**。请尝试使用 `train_web.py` 在您的浏览器中微调模型。感谢 [@KanadeSiina](https://github.com/KanadeSiina) 和 [@codemayq](https://github.com/codemayq) 在该功能开发中付出的努力。
 
-[23/07/11] 现在我们支持了 **Baichuan-13B** 模型的训练。请尝试使用 `--model_name_or_path baichuan-inc/Baichuan-13B-Base` 和 `--lora_target W_pack` 参数。请注意使用 Baichuan-13B-Chat 模型需要添加 `--template baichuan` 参数。
+[23/07/11] 现在我们支持了 **Baichuan-13B** 模型的训练。请尝试使用 `--model_name_or_path baichuan-inc/Baichuan-13B-Base` 和 `--lora_target W_pack` 参数。使用 Baichuan-13B-Chat 模型时请添加 `--template baichuan` 参数。
 
-[23/07/09] 我们开源了 [FastEdit](https://github.com/hiyouga/FastEdit)⚡🩹，一个简单易用的、能迅速编辑大模型事实记忆的工具包。如果您感兴趣请关注我们的 [FastEdit](https://github.com/hiyouga/FastEdit) 项目。
+[23/07/09] 我们开源了 **[FastEdit](https://github.com/hiyouga/FastEdit)**⚡🩹，一个简单易用的、能迅速编辑大模型事实记忆的工具包。如果您感兴趣请关注我们的 [FastEdit](https://github.com/hiyouga/FastEdit) 项目。
 
-[23/07/07] 现在我们支持了 **InternLM-7B** 模型的训练。请尝试使用 `--model_name_or_path internlm/internlm-7b` 参数。请注意使用 InternLM-chat 模型需要添加 `--template intern` 参数。
+[23/07/07] 现在我们支持了 **InternLM-7B** 模型的训练。请尝试使用 `--model_name_or_path internlm/internlm-7b` 参数。使用 InternLM-chat 模型时请添加 `--template intern` 参数。
 
 [23/07/05] 现在我们支持了 **Falcon-7B/40B** 模型的训练。请尝试使用 `--model_name_or_path tiiuae/falcon-7b` 和 `--lora_target query_key_value` 参数。
 
 [23/06/29] 我们提供了一个**可复现的**指令模型微调示例，详细内容请查阅 [Hugging Face 项目](https://huggingface.co/hiyouga/baichuan-7b-sft)。
 
-[23/06/22] 我们对齐了[示例 API](src/api_demo.py) 与 [OpenAI API](https://platform.openai.com/docs/api-reference/chat) 的格式，您可以将微调模型接入任意基于 ChatGPT 的应用中。
+[23/06/22] 我们对齐了[示例 API](src/api_demo.py) 与 [OpenAI API](https://platform.openai.com/docs/api-reference/chat) 的格式，您可以将微调模型接入**任意基于 ChatGPT 的应用**中。
 
 [23/06/15] 现在我们支持了 **Baichuan-7B** 模型的训练。请尝试使用 `--model_name_or_path baichuan-inc/Baichuan-7B` 和 `--lora_target W_pack` 参数。
 
-[23/06/03] 现在我们实现了 4 比特的 LoRA 训练（也称 [QLoRA](https://github.com/artidoro/qlora)）。请尝试使用 `--quantization_bit 4` 参数进行 4 比特量化微调。
+[23/06/03] 现在我们实现了 4 比特的 LoRA 训练（也称 **[QLoRA](https://github.com/artidoro/qlora)**）。请尝试使用 `--quantization_bit 4` 参数进行 4 比特量化微调。
 
 [23/05/31] 现在我们支持了 **BLOOM & BLOOMZ** 模型的训练。请尝试使用 `--model_name_or_path bigscience/bloomz-7b1-mt` 和 `--lora_target query_key_value` 参数。
 
@@ -71,6 +71,8 @@
 | PPO 训练    |           |             | ✅   | ✅   |
 | DPO 训练    | ✅        |             | ✅   | ✅   |
 
+- 使用 `--quantization_bit 4/8` 参数来启用 QLoRA 训练。
+
 ## 数据集
 
 - 用于预训练：
@@ -203,8 +205,6 @@ CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
     --fp16
 ```
 
-使用 Baichuan 模型时请指定 `--lora_target W_pack` 参数。
-
 ### 奖励模型训练
 
 ```bash
@@ -279,6 +279,8 @@ CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
 
 ### 多 GPU 分布式训练
 
+#### 使用 Huggingface Accelerate
+
 ```bash
 accelerate config # 首先配置分布式环境
 accelerate launch src/train_bash.py # 参数同上
@@ -312,7 +314,43 @@ use_cpu: false
 
 </details>
 
-### 指标评估（BLEU分数和汉语ROUGE分数）
+#### 使用 DeepSpeed
+
+```bash
+deepspeed --num_gpus 8 --master_port=9901 src/train_bash.py --deepspeed ds_config.json ... # 参数同上
+```
+
+<details><summary>使用 DeepSpeed ZeRO-2 进行全参数微调的 DeepSpeed 配置示例</summary>
+
+```json
+{
+  "train_micro_batch_size_per_gpu": "auto",
+  "gradient_accumulation_steps": "auto",
+  "gradient_clipping": "auto",
+  "zero_allow_untested_optimizer": true,
+  "fp16": {
+    "enabled": "auto",
+    "loss_scale": 0,
+    "initial_scale_power": 16,
+    "loss_scale_window": 1000,
+    "hysteresis": 2,
+    "min_loss_scale": 1
+  },  
+  "zero_optimization": {
+    "stage": 2,
+    "allgather_partitions": true,
+    "allgather_bucket_size": 5e8,
+    "reduce_scatter": true,
+    "reduce_bucket_size": 5e8,
+    "overlap_comm": false,
+    "contiguous_gradients": true
+  }
+}
+```
+
+</details>
+
+### 指标评估（BLEU 分数和汉语 ROUGE 分数）
 
 ```bash
 CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
@@ -329,7 +367,7 @@ CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
     --predict_with_generate
 ```
 
-我们建议在量化模型的评估中使用 `--per_device_eval_batch_size=1` 和 `--max_target_length 128` 参数。
+我们建议在量化模型的评估中使用 `--per_device_eval_batch_size=1` 和 `--max_target_length 128`。
 
 ### 模型预测
 

From 6af7fe15adfdb12f7f5b350b730e5ac4d9b36385 Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Sat, 12 Aug 2023 21:25:19 +0800
Subject: [PATCH 19/45] update readme

Former-commit-id: 3ea1fa35d1f951ae411248c03a8549b9714d876a
---
 README_zh.md | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/README_zh.md b/README_zh.md
index c2f83504..7455a98c 100644
--- a/README_zh.md
+++ b/README_zh.md
@@ -317,7 +317,9 @@ use_cpu: false
 #### 使用 DeepSpeed
 
 ```bash
-deepspeed --num_gpus 8 --master_port=9901 src/train_bash.py --deepspeed ds_config.json ... # 参数同上
+deepspeed --num_gpus 8 --master_port=9901 src/train_bash.py \
+    --deepspeed ds_config.json \
+    ... # 参数同上
 ```
 
 <details><summary>使用 DeepSpeed ZeRO-2 进行全参数微调的 DeepSpeed 配置示例</summary>

From 2bcf0025d6ccc3bf3e20b8be4d7df0834ffd3868 Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Sat, 12 Aug 2023 21:29:06 +0800
Subject: [PATCH 20/45] update readme

Former-commit-id: 8a79ded55d6e696368c96a6d9958e7c8cdaf977b
---
 README.md    | 2 +-
 README_zh.md | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/README.md b/README.md
index fc7ff69e..975a783f 100644
--- a/README.md
+++ b/README.md
@@ -28,7 +28,7 @@
 
 [23/07/11] Now we support training the **Baichuan-13B** model in this repo. Try `--model_name_or_path baichuan-inc/Baichuan-13B-Base` and `--lora_target W_pack` arguments to train the Baichuan-13B model. Remember to use `--template baichuan` argument when you are using the Baichuan-13B-Chat model.
 
-[23/07/09] Now we release **[FastEdit](https://github.com/hiyouga/FastEdit)**⚡🩹, an easy-to-use package for editing the factual knowledge of large language models efficiently. Please follow [FastEdit](https://github.com/hiyouga/FastEdit) if you are interested.
+[23/07/09] Now we release **[FastEdit](https://github.com/hiyouga/FastEdit)** ⚡🩹, an easy-to-use package for editing the factual knowledge of large language models efficiently. Please follow [FastEdit](https://github.com/hiyouga/FastEdit) if you are interested.
 
 [23/07/07] Now we support training the **InternLM-7B** model in this repo. Try `--model_name_or_path internlm/internlm-7b` argument to use the InternLM model. Remember to use `--template intern` argument when you are using the InternLM-chat model.
 
diff --git a/README_zh.md b/README_zh.md
index 7455a98c..0b8ce714 100644
--- a/README_zh.md
+++ b/README_zh.md
@@ -28,7 +28,7 @@
 
 [23/07/11] 现在我们支持了 **Baichuan-13B** 模型的训练。请尝试使用 `--model_name_or_path baichuan-inc/Baichuan-13B-Base` 和 `--lora_target W_pack` 参数。使用 Baichuan-13B-Chat 模型时请添加 `--template baichuan` 参数。
 
-[23/07/09] 我们开源了 **[FastEdit](https://github.com/hiyouga/FastEdit)**⚡🩹，一个简单易用的、能迅速编辑大模型事实记忆的工具包。如果您感兴趣请关注我们的 [FastEdit](https://github.com/hiyouga/FastEdit) 项目。
+[23/07/09] 我们开源了 **[FastEdit](https://github.com/hiyouga/FastEdit)** ⚡🩹，一个简单易用的、能迅速编辑大模型事实记忆的工具包。如果您感兴趣请关注我们的 [FastEdit](https://github.com/hiyouga/FastEdit) 项目。
 
 [23/07/07] 现在我们支持了 **InternLM-7B** 模型的训练。请尝试使用 `--model_name_or_path internlm/internlm-7b` 参数。使用 InternLM-chat 模型时请添加 `--template intern` 参数。
 

From bd611e0090322fd999c40187219ed55fb0b86b2d Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Sat, 12 Aug 2023 22:00:01 +0800
Subject: [PATCH 21/45] fix rope scaling

Former-commit-id: 8545c11c45906b33c78e144c2338963eaf0406b8
---
 src/llmtuner/tuner/core/loader.py | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/src/llmtuner/tuner/core/loader.py b/src/llmtuner/tuner/core/loader.py
index 32d6247a..47c72de1 100644
--- a/src/llmtuner/tuner/core/loader.py
+++ b/src/llmtuner/tuner/core/loader.py
@@ -102,11 +102,11 @@ def load_model_and_tokenizer(
                     )
 
                 current_max_length = getattr(config, "max_position_embeddings", None)
-                if current_max_length and model_args.model_max_length <= current_max_length:
+                if current_max_length and model_args.model_max_length > current_max_length:
+                    scaling_factor = float(math.ceil(model_args.model_max_length / current_max_length))
+                else:
                     logger.warning("Input length is smaller than max length. Consider increase input length.")
                     scaling_factor = 1.0
-                else:
-                    scaling_factor = float(math.ceil(model_args.model_max_length / current_max_length))
             else:
                 scaling_factor = 2.0
 

From e1b43dfc7fbca053bbeb11e285e9e67366d9c489 Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Sat, 12 Aug 2023 22:02:43 +0800
Subject: [PATCH 22/45] tiny fix

Former-commit-id: 9dc6a296e327c5ff27cbd1697437d9d3145e3d9a
---
 src/llmtuner/tuner/core/loader.py | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/src/llmtuner/tuner/core/loader.py b/src/llmtuner/tuner/core/loader.py
index 47c72de1..8de98387 100644
--- a/src/llmtuner/tuner/core/loader.py
+++ b/src/llmtuner/tuner/core/loader.py
@@ -83,17 +83,17 @@ def load_model_and_tokenizer(
 
     # Set RoPE scaling
     if model_args.rope_scaling is not None:
-        require_version("transformers>=4.31.0", "RoPE scaling requires transformers>=4.31.0")
-
         if hasattr(config, "use_dynamic_ntk"): # for Qwen models
             if is_trainable:
-                logger.warning("Qwen model does not support rope scaling in training.")
+                logger.warning("Qwen model does not support RoPE scaling in training.")
             else:
                 setattr(config, "use_dynamic_ntk", True)
                 setattr(config, "use_logn_attn", True)
                 logger.info("Using dynamic NTK scaling.")
 
         elif hasattr(config, "rope_scaling"): # for LLaMA models
+            require_version("transformers>=4.31.0", "RoPE scaling requires transformers>=4.31.0")
+
             if is_trainable:
                 if model_args.rope_scaling == "dynamic":
                     logger.warning(

From 7984ae8b629cce831214e36090efc31e9739d51d Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Sat, 12 Aug 2023 23:52:07 +0800
Subject: [PATCH 23/45] fix webui

Former-commit-id: d69b1388e61e5e867ec5c9a9a223677c5b5860ce
---
 src/llmtuner/webui/components/top.py | 2 +-
 src/llmtuner/webui/runner.py         | 4 +---
 src/llmtuner/webui/utils.py          | 2 +-
 3 files changed, 3 insertions(+), 5 deletions(-)

diff --git a/src/llmtuner/webui/components/top.py b/src/llmtuner/webui/components/top.py
index 77d60593..97a332c1 100644
--- a/src/llmtuner/webui/components/top.py
+++ b/src/llmtuner/webui/components/top.py
@@ -26,7 +26,7 @@ def create_top() -> Dict[str, "Component"]:
 
     with gr.Accordion(label="Advanced config", open=False) as advanced_tab:
         with gr.Row():
-            quantization_bit = gr.Dropdown([8, 4], scale=1)
+            quantization_bit = gr.Dropdown(["", "8", "4"], scale=1)
             template = gr.Dropdown(value="default", choices=list(templates.keys()), scale=1)
             source_prefix = gr.Textbox(scale=2)
 
diff --git a/src/llmtuner/webui/runner.py b/src/llmtuner/webui/runner.py
index aab31349..3a28dae6 100644
--- a/src/llmtuner/webui/runner.py
+++ b/src/llmtuner/webui/runner.py
@@ -126,8 +126,6 @@ class Runner:
             logging_steps=logging_steps,
             save_steps=save_steps,
             warmup_steps=warmup_steps,
-            fp16=(compute_type == "fp16"),
-            bf16=(compute_type == "bf16"),
             padding_side=padding_side,
             lora_rank=lora_rank,
             lora_dropout=lora_dropout,
@@ -135,7 +133,7 @@ class Runner:
             resume_lora_training=resume_lora_training,
             output_dir=output_dir
         )
-
+        args[compute_type] = True
         if val_size > 1e-6:
             args["val_size"] = val_size
             args["evaluation_strategy"] = "steps"
diff --git a/src/llmtuner/webui/utils.py b/src/llmtuner/webui/utils.py
index 6ca55e14..168fbe43 100644
--- a/src/llmtuner/webui/utils.py
+++ b/src/llmtuner/webui/utils.py
@@ -65,7 +65,7 @@ def can_quantize(finetuning_type: str) -> Dict[str, Any]:
 def gen_cmd(args: Dict[str, Any]) -> str:
     cmd_lines = ["CUDA_VISIBLE_DEVICES=0 python "]
     for k, v in args.items():
-        if v is not None and v is not False and v != "":
+        if v is not None and v != "":
             cmd_lines.append("    --{} {} ".format(k, str(v)))
     cmd_text = "\\\n".join(cmd_lines)
     cmd_text = "```bash\n{}\n```".format(cmd_text)

From e75024fde3b8f96a706a8915dd890354449b1d5e Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Mon, 14 Aug 2023 00:23:56 +0800
Subject: [PATCH 24/45] fix #480

Former-commit-id: 2f2fd55d8175eb3c6ce94bc821ab4e6331f79d8e
---
 src/llmtuner/tuner/dpo/trainer.py | 16 +++++++++-------
 src/llmtuner/webui/utils.py       |  1 +
 2 files changed, 10 insertions(+), 7 deletions(-)

diff --git a/src/llmtuner/tuner/dpo/trainer.py b/src/llmtuner/tuner/dpo/trainer.py
index a94642c1..f0da9eaa 100644
--- a/src/llmtuner/tuner/dpo/trainer.py
+++ b/src/llmtuner/tuner/dpo/trainer.py
@@ -2,7 +2,7 @@ import torch
 from collections import defaultdict
 from peft import PeftModel
 from typing import TYPE_CHECKING, Dict, Optional, Tuple, Union
-from transformers import Trainer
+from transformers import BatchEncoding, Trainer
 from trl import DPOTrainer
 
 from llmtuner.extras.constants import IGNORE_INDEX
@@ -43,21 +43,23 @@ class DPOPeftTrainer(PeftModelMixin, DPOTrainer):
         model: Optional[torch.nn.Module] = None,
         batch: Optional[Dict[str, torch.Tensor]] = None
     ) -> Tuple[torch.FloatTensor, torch.FloatTensor, torch.FloatTensor, torch.FloatTensor]:
+        batch_copied = BatchEncoding({k: v.detach().clone() for k, v in batch.items()}) # avoid error
         unwrapped_model: "PreTrainedModel" = self.accelerator.unwrap_model(self.model)
+
         if not torch.is_grad_enabled():
             unwrapped_model.gradient_checkpointing_disable()
 
         if model is None and isinstance(unwrapped_model, PeftModel): # peft model has no ref_model
             with unwrapped_model.disable_adapter():
-                all_logits: torch.Tensor = self.model(
-                    batch["input_ids"],
-                    attention_mask=batch["attention_mask"],
+                all_logits = self.model(
+                    input_ids=batch_copied["input_ids"],
+                    attention_mask=batch_copied["attention_mask"],
                     return_dict=True
                 ).logits.to(torch.float32)
         else:
-            all_logits: torch.Tensor = model(
-                batch["input_ids"],
-                attention_mask=batch["attention_mask"],
+            all_logits = model(
+                input_ids=batch_copied["input_ids"],
+                attention_mask=batch_copied["attention_mask"],
                 return_dict=True
             ).logits.to(torch.float32)
 
diff --git a/src/llmtuner/webui/utils.py b/src/llmtuner/webui/utils.py
index 168fbe43..152df6e3 100644
--- a/src/llmtuner/webui/utils.py
+++ b/src/llmtuner/webui/utils.py
@@ -63,6 +63,7 @@ def can_quantize(finetuning_type: str) -> Dict[str, Any]:
 
 
 def gen_cmd(args: Dict[str, Any]) -> str:
+    args["plot_loss"] = True
     cmd_lines = ["CUDA_VISIBLE_DEVICES=0 python "]
     for k, v in args.items():
         if v is not None and v != "":

From 6c9b035c0e524be1ba5a2f1d70ec03349b1d2fe7 Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Mon, 14 Aug 2023 10:48:47 +0800
Subject: [PATCH 25/45] web UI integrating RLHF

Former-commit-id: ec94274ca155300aee27621c018dd1bbaf78194b
---
 README.md                                     |  6 +--
 src/llmtuner/tuner/ppo/workflow.py            |  4 +-
 src/llmtuner/webui/common.py                  | 10 +++--
 src/llmtuner/webui/components/__init__.py     |  2 +-
 src/llmtuner/webui/components/top.py          | 15 +++++---
 .../webui/components/{sft.py => train.py}     | 36 ++++++++++++++----
 src/llmtuner/webui/interface.py               | 10 ++---
 src/llmtuner/webui/locales.py                 | 38 +++++++++++++++++++
 src/llmtuner/webui/manager.py                 | 14 +++++--
 src/llmtuner/webui/runner.py                  | 22 ++++++++++-
 src/llmtuner/webui/utils.py                   |  3 +-
 11 files changed, 128 insertions(+), 32 deletions(-)
 rename src/llmtuner/webui/components/{sft.py => train.py} (80%)

diff --git a/README.md b/README.md
index 975a783f..1124ade2 100644
--- a/README.md
+++ b/README.md
@@ -68,7 +68,7 @@
 | ---------------------- | -------------- | ----------------- | ---- | ----- |
 | Pre-Training           | ✅            | ✅                | ✅   | ✅   |
 | Supervised Fine-Tuning | ✅            | ✅                | ✅   | ✅   |
-| Reward Model Training  |                |                   | ✅   | ✅   |
+| Reward Modeling        |                |                   | ✅   | ✅   |
 | PPO Training           |                |                   | ✅   | ✅   |
 | DPO Training           | ✅            |                    | ✅   | ✅   |
 
@@ -103,7 +103,7 @@
   - [Web QA (zh)](https://huggingface.co/datasets/suolyer/webqa)
   - [UltraChat (en)](https://github.com/thunlp/UltraChat)
   - [WebNovel (zh)](https://huggingface.co/datasets/zxbsmk/webnovel_cn)
-- For reward modelling or DPO training:
+- For reward modeling or DPO training:
   - [HH-RLHF (en)](https://huggingface.co/datasets/Anthropic/hh-rlhf)
   - [Open Assistant (multilingual)](https://huggingface.co/datasets/OpenAssistant/oasst1)
   - [GPT-4 Generated Data (en&zh)](https://github.com/Instruction-Tuning-with-GPT-4/GPT-4-LLM)
@@ -206,7 +206,7 @@ CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
     --fp16
 ```
 
-### Reward Model Training
+### Reward Modeling
 
 ```bash
 CUDA_VISIBLE_DEVICES=0 python src/train_bash.py \
diff --git a/src/llmtuner/tuner/ppo/workflow.py b/src/llmtuner/tuner/ppo/workflow.py
index 6734ab78..243468cc 100644
--- a/src/llmtuner/tuner/ppo/workflow.py
+++ b/src/llmtuner/tuner/ppo/workflow.py
@@ -37,7 +37,9 @@ def run_ppo(
         batch_size=training_args.per_device_train_batch_size * training_args.gradient_accumulation_steps,
         gradient_accumulation_steps=training_args.gradient_accumulation_steps,
         ppo_epochs=1,
-        max_grad_norm=training_args.max_grad_norm
+        max_grad_norm=training_args.max_grad_norm,
+        seed=training_args.seed,
+        optimize_cuda_cache=True
     )
 
     optimizer = AdamW(filter(lambda p: p.requires_grad, model.parameters()), lr=training_args.learning_rate)
diff --git a/src/llmtuner/webui/common.py b/src/llmtuner/webui/common.py
index bf1d18fb..04b0774b 100644
--- a/src/llmtuner/webui/common.py
+++ b/src/llmtuner/webui/common.py
@@ -29,14 +29,16 @@ def load_config() -> Dict[str, Any]:
         with open(get_config_path(), "r", encoding="utf-8") as f:
             return json.load(f)
     except:
-        return {"last_model": "", "path_dict": {}}
+        return {"lang": "", "last_model": "", "path_dict": {}}
 
 
-def save_config(model_name: str, model_path: str) -> None:
+def save_config(lang: str, model_name: str, model_path: str) -> None:
     os.makedirs(DEFAULT_CACHE_DIR, exist_ok=True)
     user_config = load_config()
-    user_config["last_model"] = model_name
-    user_config["path_dict"][model_name] = model_path
+    user_config["lang"] = lang or user_config["lang"]
+    if model_name:
+        user_config["last_model"] = model_name
+        user_config["path_dict"][model_name] = model_path
     with open(get_config_path(), "w", encoding="utf-8") as f:
         json.dump(user_config, f, indent=2, ensure_ascii=False)
 
diff --git a/src/llmtuner/webui/components/__init__.py b/src/llmtuner/webui/components/__init__.py
index 5b86f396..32228b8e 100644
--- a/src/llmtuner/webui/components/__init__.py
+++ b/src/llmtuner/webui/components/__init__.py
@@ -1,5 +1,5 @@
 from llmtuner.webui.components.top import create_top
-from llmtuner.webui.components.sft import create_sft_tab
+from llmtuner.webui.components.train import create_train_tab
 from llmtuner.webui.components.eval import create_eval_tab
 from llmtuner.webui.components.infer import create_infer_tab
 from llmtuner.webui.components.export import create_export_tab
diff --git a/src/llmtuner/webui/components/top.py b/src/llmtuner/webui/components/top.py
index 97a332c1..8611e280 100644
--- a/src/llmtuner/webui/components/top.py
+++ b/src/llmtuner/webui/components/top.py
@@ -20,22 +20,25 @@ def create_top() -> Dict[str, "Component"]:
         model_path = gr.Textbox(scale=3)
 
     with gr.Row():
-        finetuning_type = gr.Dropdown(value="lora", choices=METHODS, scale=1)
+        finetuning_type = gr.Dropdown(choices=METHODS, value="lora", scale=1)
         checkpoints = gr.Dropdown(multiselect=True, scale=5)
         refresh_btn = gr.Button(scale=1)
 
     with gr.Accordion(label="Advanced config", open=False) as advanced_tab:
         with gr.Row():
-            quantization_bit = gr.Dropdown(["", "8", "4"], scale=1)
-            template = gr.Dropdown(value="default", choices=list(templates.keys()), scale=1)
+            quantization_bit = gr.Dropdown(choices=["None", "8", "4"], value="None", scale=1)
+            template = gr.Dropdown(choices=list(templates.keys()), value="default", scale=1)
             source_prefix = gr.Textbox(scale=2)
 
+    lang.change(save_config, [lang, model_name, model_path])
+
     model_name.change(
         list_checkpoint, [model_name, finetuning_type], [checkpoints]
     ).then(
         get_model_path, [model_name], [model_path]
     ) # do not save config since the below line will save
-    model_path.change(save_config, [model_name, model_path])
+
+    model_path.change(save_config, [lang, model_name, model_path])
 
     finetuning_type.change(
         list_checkpoint, [model_name, finetuning_type], [checkpoints]
@@ -43,7 +46,9 @@ def create_top() -> Dict[str, "Component"]:
         can_quantize, [finetuning_type], [quantization_bit]
     )
 
-    refresh_btn.click(list_checkpoint, [model_name, finetuning_type], [checkpoints], queue=False)
+    refresh_btn.click(
+        list_checkpoint, [model_name, finetuning_type], [checkpoints], queue=False
+    )
 
     return dict(
         lang=lang,
diff --git a/src/llmtuner/webui/components/sft.py b/src/llmtuner/webui/components/train.py
similarity index 80%
rename from src/llmtuner/webui/components/sft.py
rename to src/llmtuner/webui/components/train.py
index a4101d15..a4d593c8 100644
--- a/src/llmtuner/webui/components/sft.py
+++ b/src/llmtuner/webui/components/train.py
@@ -3,7 +3,7 @@ from transformers.trainer_utils import SchedulerType
 
 import gradio as gr
 
-from llmtuner.webui.common import list_dataset, DEFAULT_DATA_DIR
+from llmtuner.webui.common import list_checkpoint, list_dataset, DEFAULT_DATA_DIR
 from llmtuner.webui.components.data import create_preview_box
 from llmtuner.webui.utils import can_preview, get_preview, gen_plot
 
@@ -12,7 +12,7 @@ if TYPE_CHECKING:
     from llmtuner.webui.runner import Runner
 
 
-def create_sft_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict[str, "Component"]:
+def create_train_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict[str, "Component"]:
     with gr.Row():
         dataset_dir = gr.Textbox(value=DEFAULT_DATA_DIR, scale=2)
         dataset = gr.Dropdown(multiselect=True, scale=4)
@@ -40,7 +40,7 @@ def create_sft_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict[
         batch_size = gr.Slider(value=4, minimum=1, maximum=512, step=1)
         gradient_accumulation_steps = gr.Slider(value=4, minimum=1, maximum=512, step=1)
         lr_scheduler_type = gr.Dropdown(
-            value="cosine", choices=[scheduler.value for scheduler in SchedulerType]
+            choices=[scheduler.value for scheduler in SchedulerType], value="cosine"
         )
         max_grad_norm = gr.Textbox(value="1.0")
         val_size = gr.Slider(value=0, minimum=0, maximum=1, step=0.001)
@@ -60,6 +60,20 @@ def create_sft_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict[
             lora_target = gr.Textbox(scale=2)
             resume_lora_training = gr.Checkbox(value=True, scale=1)
 
+    with gr.Accordion(label="RLHF config", open=False) as rlhf_tab:
+        with gr.Row():
+            rlhf_method = gr.Dropdown(choices=["None", "Reward Modeling", "PPO", "DPO"], value="None", scale=1)
+            dpo_beta = gr.Slider(value=0.1, minimum=0, maximum=1, step=0.01, scale=2)
+            reward_model = gr.Dropdown(scale=2)
+            refresh_btn = gr.Button(scale=1)
+
+    refresh_btn.click(
+        list_checkpoint,
+        [top_elems["model_name"], top_elems["finetuning_type"]],
+        [reward_model],
+        queue=False
+    )
+
     with gr.Row():
         cmd_preview_btn = gr.Button()
         start_btn = gr.Button()
@@ -79,7 +93,7 @@ def create_sft_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict[
         with gr.Column(scale=1):
             loss_viewer = gr.Plot()
 
-    input_list = [
+    input_components = [
         top_elems["lang"],
         top_elems["model_name"],
         top_elems["checkpoints"],
@@ -108,16 +122,19 @@ def create_sft_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict[
         lora_dropout,
         lora_target,
         resume_lora_training,
+        rlhf_method,
+        dpo_beta,
+        reward_model,
         output_dir
     ]
 
-    output_list = [
+    output_components = [
         output_box,
         process_bar
     ]
 
-    cmd_preview_btn.click(runner.preview_train, input_list, output_list)
-    start_btn.click(runner.run_train, input_list, output_list)
+    cmd_preview_btn.click(runner.preview_train, input_components, output_components)
+    start_btn.click(runner.run_train, input_components, output_components)
     stop_btn.click(runner.set_abort, queue=False)
 
     process_bar.change(
@@ -152,6 +169,11 @@ def create_sft_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict[
         lora_dropout=lora_dropout,
         lora_target=lora_target,
         resume_lora_training=resume_lora_training,
+        rlhf_tab=rlhf_tab,
+        rlhf_method=rlhf_method,
+        dpo_beta=dpo_beta,
+        reward_model=reward_model,
+        refresh_btn=refresh_btn,
         cmd_preview_btn=cmd_preview_btn,
         start_btn=start_btn,
         stop_btn=stop_btn,
diff --git a/src/llmtuner/webui/interface.py b/src/llmtuner/webui/interface.py
index 8539e18c..0ae59224 100644
--- a/src/llmtuner/webui/interface.py
+++ b/src/llmtuner/webui/interface.py
@@ -3,7 +3,7 @@ from transformers.utils.versions import require_version
 
 from llmtuner.webui.components import (
     create_top,
-    create_sft_tab,
+    create_train_tab,
     create_eval_tab,
     create_infer_tab,
     create_export_tab,
@@ -24,8 +24,8 @@ def create_ui() -> gr.Blocks:
     with gr.Blocks(title="Web Tuner", css=CSS) as demo:
         top_elems = create_top()
 
-        with gr.Tab("SFT"):
-            sft_elems = create_sft_tab(top_elems, runner)
+        with gr.Tab("Train"):
+            train_elems = create_train_tab(top_elems, runner)
 
         with gr.Tab("Evaluate"):
             eval_elems = create_eval_tab(top_elems, runner)
@@ -36,7 +36,7 @@ def create_ui() -> gr.Blocks:
         with gr.Tab("Export"):
             export_elems = create_export_tab(top_elems)
 
-        elem_list = [top_elems, sft_elems, eval_elems, infer_elems, export_elems]
+        elem_list = [top_elems, train_elems, eval_elems, infer_elems, export_elems]
         manager = Manager(elem_list)
 
         demo.load(
@@ -59,7 +59,7 @@ def create_web_demo() -> gr.Blocks:
     chat_model = WebChatModel(lazy_init=False)
 
     with gr.Blocks(title="Web Demo", css=CSS) as demo:
-        lang = gr.Dropdown(choices=["en", "zh"], value="en")
+        lang = gr.Dropdown(choices=["en", "zh"], value="")
 
         _, _, _, chat_elems = create_chat_box(chat_model, visible=True)
 
diff --git a/src/llmtuner/webui/locales.py b/src/llmtuner/webui/locales.py
index aba4acae..0c7abae0 100644
--- a/src/llmtuner/webui/locales.py
+++ b/src/llmtuner/webui/locales.py
@@ -335,6 +335,44 @@ LOCALES = {
             "info": "接着上次的 LoRA 权重训练或创建一个新的 LoRA 权重。"
         }
     },
+    "rlhf_tab": {
+        "en": {
+            "label": "RLHF configurations"
+        },
+        "zh": {
+            "label": "RLHF 参数设置"
+        }
+    },
+    "rlhf_method": {
+        "en": {
+            "label": "RLHF method",
+            "info": "The RLHF algorithm to adopt."
+        },
+        "zh": {
+            "label": "RLHF 方法",
+            "info": "RLHF 阶段使用的算法。"
+        }
+    },
+    "dpo_beta": {
+        "en": {
+            "label": "DPO beta",
+            "info": "Value of the beta parameter in the DPO loss."
+        },
+        "zh": {
+            "label": "DPO beta 参数",
+            "info": "DPO 损失函数中 beta 超参数大小。"
+        }
+    },
+    "reward_model": {
+        "en": {
+            "label": "Reward model",
+            "info": "Checkpoint of the reward model for PPO training."
+        },
+        "zh": {
+            "label": "奖励模型",
+            "info": "PPO 训练中奖励模型的断点路径。"
+        }
+    },
     "cmd_preview_btn": {
         "en": {
             "value": "Preview command"
diff --git a/src/llmtuner/webui/manager.py b/src/llmtuner/webui/manager.py
index c8f797a4..2d5a0a39 100644
--- a/src/llmtuner/webui/manager.py
+++ b/src/llmtuner/webui/manager.py
@@ -12,12 +12,18 @@ class Manager:
     def __init__(self, elem_list: List[Dict[str, Component]]):
         self.elem_list = elem_list
 
-    def gen_refresh(self) -> Dict[str, Any]:
+    def gen_refresh(self, lang: str) -> Dict[str, Any]:
         refresh_dict = {
             "dataset": {"choices": list_dataset()["choices"]},
             "output_dir": {"value": get_time()}
         }
+
         user_config = load_config()
+        if lang:
+            refresh_dict["lang"] = {"value": lang}
+        else:
+            refresh_dict["lang"] = {"value": user_config["lang"] if user_config["lang"] else "en"}
+
         if user_config["last_model"]:
             refresh_dict["model_name"] = {"value": user_config["last_model"]}
             refresh_dict["model_path"] = {"value": get_model_path(user_config["last_model"])}
@@ -26,10 +32,12 @@ class Manager:
 
     def gen_label(self, lang: str) -> Dict[Component, Dict[str, Any]]: # cannot use TYPE_CHECKING
         update_dict = {}
-        refresh_dict = self.gen_refresh()
+        refresh_dict = self.gen_refresh(lang)
 
         for elems in self.elem_list:
             for name, component in elems.items():
-                update_dict[component] = gr.update(**LOCALES[name][lang], **refresh_dict.get(name, {}))
+                update_dict[component] = gr.update(
+                    **LOCALES[name][refresh_dict["lang"]["value"]], **refresh_dict.get(name, {})
+                )
 
         return update_dict
diff --git a/src/llmtuner/webui/runner.py b/src/llmtuner/webui/runner.py
index 3a28dae6..afef4736 100644
--- a/src/llmtuner/webui/runner.py
+++ b/src/llmtuner/webui/runner.py
@@ -91,6 +91,9 @@ class Runner:
         lora_dropout: float,
         lora_target: str,
         resume_lora_training: bool,
+        rlhf_method: str,
+        dpo_beta: float,
+        reward_model: str,
         output_dir: str
     ) -> Tuple[str, str, List[str], str, Dict[str, Any]]:
         if checkpoints:
@@ -109,7 +112,7 @@ class Runner:
             overwrite_cache=True,
             checkpoint_dir=checkpoint_dir,
             finetuning_type=finetuning_type,
-            quantization_bit=int(quantization_bit) if quantization_bit else None,
+            quantization_bit=int(quantization_bit) if quantization_bit != "None" else None,
             template=template,
             source_prefix=source_prefix,
             dataset_dir=dataset_dir,
@@ -134,6 +137,21 @@ class Runner:
             output_dir=output_dir
         )
         args[compute_type] = True
+
+        if rlhf_method == "Reward Modeling":
+            args["stage"] = "rm"
+            args["resume_lora_training"] = False
+        elif rlhf_method == "PPO":
+            args["stage"] = "ppo"
+            args["resume_lora_training"] = False
+            args["reward_model"] = reward_model
+            args["padding_side"] = "left"
+            val_size = 0
+        elif rlhf_method == "DPO":
+            args["stage"] = "dpo"
+            args["resume_lora_training"] = False
+            args["dpo_beta"] = dpo_beta
+
         if val_size > 1e-6:
             args["val_size"] = val_size
             args["evaluation_strategy"] = "steps"
@@ -176,7 +194,7 @@ class Runner:
             predict_with_generate=True,
             checkpoint_dir=checkpoint_dir,
             finetuning_type=finetuning_type,
-            quantization_bit=int(quantization_bit) if quantization_bit else None,
+            quantization_bit=int(quantization_bit) if quantization_bit != "None" else None,
             template=template,
             source_prefix=source_prefix,
             dataset_dir=dataset_dir,
diff --git a/src/llmtuner/webui/utils.py b/src/llmtuner/webui/utils.py
index 152df6e3..362fa008 100644
--- a/src/llmtuner/webui/utils.py
+++ b/src/llmtuner/webui/utils.py
@@ -63,7 +63,8 @@ def can_quantize(finetuning_type: str) -> Dict[str, Any]:
 
 
 def gen_cmd(args: Dict[str, Any]) -> str:
-    args["plot_loss"] = True
+    if args.get("do_train", None):
+        args["plot_loss"] = True
     cmd_lines = ["CUDA_VISIBLE_DEVICES=0 python "]
     for k, v in args.items():
         if v is not None and v != "":

From 5feb2eda8914736198323e1858fb642251131206 Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Mon, 14 Aug 2023 11:13:25 +0800
Subject: [PATCH 26/45] update readme_zh

Former-commit-id: 2391a84e26d7acb3afc18617fefdfb1b3ab6f45d
---
 README_zh.md | 1 +
 1 file changed, 1 insertion(+)

diff --git a/README_zh.md b/README_zh.md
index 0b8ce714..5664c4ec 100644
--- a/README_zh.md
+++ b/README_zh.md
@@ -57,6 +57,7 @@
 | [InternLM](https://github.com/InternLM/InternLM)         | 7B                          | q_proj,v_proj     | intern   |
 | [Qwen](https://github.com/QwenLM/Qwen-7B)                | 7B                          | c_attn            | chatml   |
 | [XVERSE](https://github.com/xverse-ai/XVERSE-13B)        | 13B                         | q_proj,v_proj     | -        |
+| [ChatGLM2](https://github.com/THUDM/ChatGLM2-6B)         | 6B                          | query_key_value   | chatglm2 |
 
 - **默认模块**是 `--lora_target` 参数的部分可选项。请使用 `python src/train_bash.py -h` 查看全部可选项。
 - 对于所有“基座”（Base）模型，`--template` 参数可以是 `default`, `alpaca`, `vicuna` 等任意值。但“对话”（Chat）模型请务必使用对应的模板。

From 307c814f388cd400a8a994ac522dfd6ab1860032 Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Mon, 14 Aug 2023 11:37:01 +0800
Subject: [PATCH 27/45] fix webui cache

Former-commit-id: ca08e5efd3ccf155c6a0baa0a08bb7879f6b9b9b
---
 src/llmtuner/webui/components/top.py | 2 +-
 src/llmtuner/webui/interface.py      | 4 ++--
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/src/llmtuner/webui/components/top.py b/src/llmtuner/webui/components/top.py
index 8611e280..1f5c47a5 100644
--- a/src/llmtuner/webui/components/top.py
+++ b/src/llmtuner/webui/components/top.py
@@ -15,7 +15,7 @@ def create_top() -> Dict[str, "Component"]:
     available_models = list(SUPPORTED_MODELS.keys()) + ["Custom"]
 
     with gr.Row():
-        lang = gr.Dropdown(choices=["en", "zh"], value="en", scale=1)
+        lang = gr.Dropdown(choices=["en", "zh"], scale=1)
         model_name = gr.Dropdown(choices=available_models, scale=3)
         model_path = gr.Textbox(scale=3)
 
diff --git a/src/llmtuner/webui/interface.py b/src/llmtuner/webui/interface.py
index 0ae59224..1a5c0c19 100644
--- a/src/llmtuner/webui/interface.py
+++ b/src/llmtuner/webui/interface.py
@@ -59,7 +59,7 @@ def create_web_demo() -> gr.Blocks:
     chat_model = WebChatModel(lazy_init=False)
 
     with gr.Blocks(title="Web Demo", css=CSS) as demo:
-        lang = gr.Dropdown(choices=["en", "zh"], value="")
+        lang = gr.Dropdown(choices=["en", "zh"])
 
         _, _, _, chat_elems = create_chat_box(chat_model, visible=True)
 
@@ -67,7 +67,7 @@ def create_web_demo() -> gr.Blocks:
 
         demo.load(manager.gen_label, [lang], [lang] + list(chat_elems.values()))
 
-        lang.change(manager.gen_label, [lang], [lang] + list(chat_elems.values()), queue=False)
+        lang.select(manager.gen_label, [lang], [lang] + list(chat_elems.values()), queue=False)
 
     return demo
 

From 2f70f964636595de9370f68391e8d2a6f303f22c Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Mon, 14 Aug 2023 11:38:42 +0800
Subject: [PATCH 28/45] fix bug in webui

Former-commit-id: 20a29297b1c55678e7a56b2610b35edec063253b
---
 src/llmtuner/webui/chat.py | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/src/llmtuner/webui/chat.py b/src/llmtuner/webui/chat.py
index d0eb61df..e73ea402 100644
--- a/src/llmtuner/webui/chat.py
+++ b/src/llmtuner/webui/chat.py
@@ -53,7 +53,7 @@ class WebChatModel(ChatModel):
             model_name_or_path=model_name_or_path,
             checkpoint_dir=checkpoint_dir,
             finetuning_type=finetuning_type,
-            quantization_bit=int(quantization_bit) if quantization_bit else None,
+            quantization_bit=int(quantization_bit) if quantization_bit != "None" else None,
             template=template,
             source_prefix=source_prefix
         )

From 2aa2c363adfcfdb4141a07ac928ead38e09faaf1 Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Mon, 14 Aug 2023 14:14:48 +0800
Subject: [PATCH 29/45] fix ChatGLM lm_head #494

Former-commit-id: d0199568081f46e5d338ea511266eb420dd43594
---
 src/llmtuner/tuner/core/loader.py     | 4 ++++
 src/llmtuner/tuner/dpo/trainer.py     | 8 ++++----
 src/llmtuner/webui/components/eval.py | 8 ++++----
 3 files changed, 12 insertions(+), 8 deletions(-)

diff --git a/src/llmtuner/tuner/core/loader.py b/src/llmtuner/tuner/core/loader.py
index 8de98387..e480d44c 100644
--- a/src/llmtuner/tuner/core/loader.py
+++ b/src/llmtuner/tuner/core/loader.py
@@ -153,6 +153,10 @@ def load_model_and_tokenizer(
     if "GenerationMixin" not in str(model.generate.__func__):
         model.generate = MethodType(PreTrainedModel.generate, model)
 
+    # Fix LM head (for ChatGLM2)
+    if not hasattr(model, "lm_head"):
+        setattr(model, "lm_head", model.transformer.output_layer)
+
     # Register auto class to save the custom code files.
     if isinstance(config, PretrainedConfig) and "AutoConfig" in getattr(config, "auto_map", {}):
         config.__class__.register_for_auto_class()
diff --git a/src/llmtuner/tuner/dpo/trainer.py b/src/llmtuner/tuner/dpo/trainer.py
index f0da9eaa..458e99db 100644
--- a/src/llmtuner/tuner/dpo/trainer.py
+++ b/src/llmtuner/tuner/dpo/trainer.py
@@ -32,11 +32,11 @@ class DPOPeftTrainer(PeftModelMixin, DPOTrainer):
         self._stored_metrics = defaultdict(lambda: defaultdict(list))
 
         Trainer.__init__(self, **kwargs)
+        if not hasattr(self, "accelerator"):
+            raise AttributeError("Please update `transformers`.")
+
         if ref_model is not None:
-            if hasattr(self, "accelerator"):
-                self.ref_model = self.accelerator.prepare_model(self.ref_model, evaluation_mode=True)
-            else:
-                raise AttributeError("Please update `transformers`.")
+            self.ref_model = self.accelerator.prepare_model(self.ref_model, evaluation_mode=True)
 
     def concatenated_forward(
         self,
diff --git a/src/llmtuner/webui/components/eval.py b/src/llmtuner/webui/components/eval.py
index 72d08a1f..48372b4c 100644
--- a/src/llmtuner/webui/components/eval.py
+++ b/src/llmtuner/webui/components/eval.py
@@ -45,7 +45,7 @@ def create_eval_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict
     with gr.Box():
         output_box = gr.Markdown()
 
-    input_list = [
+    input_components = [
         top_elems["lang"],
         top_elems["model_name"],
         top_elems["checkpoints"],
@@ -62,13 +62,13 @@ def create_eval_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict
         predict
     ]
 
-    output_list = [
+    output_components = [
         output_box,
         process_bar
     ]
 
-    cmd_preview_btn.click(runner.preview_eval, input_list, output_list)
-    start_btn.click(runner.run_eval, input_list, output_list)
+    cmd_preview_btn.click(runner.preview_eval, input_components, output_components)
+    start_btn.click(runner.run_eval, input_components, output_components)
     stop_btn.click(runner.set_abort, queue=False)
 
     return dict(

From ee7da14f8169eb4c7958825dd6591264483c3940 Mon Sep 17 00:00:00 2001
From: codemayq <codingma@pku.edu.cn>
Date: Mon, 14 Aug 2023 20:42:59 +0800
Subject: [PATCH 30/45] add template match and stage in webui

Former-commit-id: 79c68e552722079faf2ab0858870b481844d66ae
---
 src/llmtuner/extras/constants.py       | 30 ++++++++++++++++++++++++++
 src/llmtuner/extras/template.py        |  3 ++-
 src/llmtuner/webui/common.py           | 21 +++++++++++++++++-
 src/llmtuner/webui/components/train.py |  9 +++++---
 src/llmtuner/webui/locales.py          | 10 ++++++++-
 src/llmtuner/webui/runner.py           | 18 +++++++++-------
 6 files changed, 77 insertions(+), 14 deletions(-)

diff --git a/src/llmtuner/extras/constants.py b/src/llmtuner/extras/constants.py
index 8ee997bb..ffdc8827 100644
--- a/src/llmtuner/extras/constants.py
+++ b/src/llmtuner/extras/constants.py
@@ -10,6 +10,8 @@ LAYERNORM_NAMES = ["norm", "ln_f", "ln_attn", "ln_mlp"]
 
 METHODS = ["full", "freeze", "lora"]
 
+STAGES = ["Supervised Finetuning", "Reward Modeling", "PPO", "DPO", "Pretraining"]
+
 SUPPORTED_MODELS = {
     "LLaMA-7B": "huggyllama/llama-7b",
     "LLaMA-13B": "huggyllama/llama-13b",
@@ -54,3 +56,31 @@ DEFAULT_MODULE = {
     "XVERSE": "q_proj,v_proj",
     "ChatGLM2": "query_key_value"
 }
+
+DEFAULT_TEMPLATE = {
+    "LLaMA2": "llama2",
+    "Baichuan": "baichuan",
+    "InternLM": "intern",
+    "Qwen": "chatml",
+    "ChatGLM2": "chatglm2"
+}
+
+# huggingface model name prefix 2 template
+DEFAULT_TEMPLATE_WITH_CUSTOM_MODEL = {
+    "Llama-2": "llama2",
+    "chinese-alpaca-2": "llama2_zh",
+    "alpaca-7b-wdiff": "alpaca",
+    "vicuna": "vicuna",
+    "BELLE": "belle",
+    "Chinese-LLaMA-2": "linly",
+    "BiLLa": "billa",
+    "Ziya": "ziya",
+    "aquilachat": "aquila",
+    "internlm": "intern",
+    "aquilachat": "aquila",
+    "internlm": "intern",
+    "Baichuan":"baichuan",
+    "starchat":"starchat",
+    "Qwen":"chatml",
+    "chatglm2":"chatglm2"
+}
\ No newline at end of file
diff --git a/src/llmtuner/extras/template.py b/src/llmtuner/extras/template.py
index 6724dd46..e8d228a4 100644
--- a/src/llmtuner/extras/template.py
+++ b/src/llmtuner/extras/template.py
@@ -273,7 +273,8 @@ register_template(
 
 
 r"""
-Supports: https://github.com/ymcui/Chinese-LLaMA-Alpaca-2
+Supports: https://huggingface.co/ziqingyang/chinese-alpaca-2-7b
+          https://github.com/ymcui/Chinese-LLaMA-Alpaca-2
 """
 register_template(
     name="llama2_zh",
diff --git a/src/llmtuner/webui/common.py b/src/llmtuner/webui/common.py
index 04b0774b..4e1ea0ae 100644
--- a/src/llmtuner/webui/common.py
+++ b/src/llmtuner/webui/common.py
@@ -6,7 +6,7 @@ import gradio as gr
 from peft.utils import WEIGHTS_NAME as PEFT_WEIGHTS_NAME
 from transformers.trainer import WEIGHTS_NAME, WEIGHTS_INDEX_NAME
 
-from llmtuner.extras.constants import SUPPORTED_MODELS
+from llmtuner.extras.constants import SUPPORTED_MODELS, DEFAULT_TEMPLATE_WITH_CUSTOM_MODEL, DEFAULT_TEMPLATE
 
 
 DEFAULT_CACHE_DIR = "cache"
@@ -48,6 +48,25 @@ def get_model_path(model_name: str) -> str:
     return user_config["path_dict"].get(model_name, SUPPORTED_MODELS.get(model_name, ""))
 
 
+def get_template(
+        template: str,
+        model_name: str,
+) -> str:
+    if template and template != "default":
+        return template
+    if model_name == "Custom":
+        model_name_or_path = get_model_path(model_name)
+        # get last dir
+        basename = os.path.basename(model_name_or_path)
+        # prefix match
+        for k, v in DEFAULT_TEMPLATE_WITH_CUSTOM_MODEL.items():
+            if basename.startswith(k):
+                return v
+        return "default"
+
+    return DEFAULT_TEMPLATE.get(model_name.split("-")[0], "default")
+
+
 def list_checkpoint(model_name: str, finetuning_type: str) -> Dict[str, Any]:
     checkpoints = []
     save_dir = os.path.join(get_save_dir(model_name), finetuning_type)
diff --git a/src/llmtuner/webui/components/train.py b/src/llmtuner/webui/components/train.py
index a4d593c8..76f8ff94 100644
--- a/src/llmtuner/webui/components/train.py
+++ b/src/llmtuner/webui/components/train.py
@@ -6,6 +6,7 @@ import gradio as gr
 from llmtuner.webui.common import list_checkpoint, list_dataset, DEFAULT_DATA_DIR
 from llmtuner.webui.components.data import create_preview_box
 from llmtuner.webui.utils import can_preview, get_preview, gen_plot
+from llmtuner.extras.constants import STAGES
 
 if TYPE_CHECKING:
     from gradio.components import Component
@@ -14,6 +15,9 @@ if TYPE_CHECKING:
 
 def create_train_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict[str, "Component"]:
     with gr.Row():
+        stage = gr.Dropdown(choices=STAGES,
+                            value="Supervised Finetuning", scale=2)
+
         dataset_dir = gr.Textbox(value=DEFAULT_DATA_DIR, scale=2)
         dataset = gr.Dropdown(multiselect=True, scale=4)
         data_preview_btn = gr.Button(interactive=False, scale=1)
@@ -62,7 +66,6 @@ def create_train_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dic
 
     with gr.Accordion(label="RLHF config", open=False) as rlhf_tab:
         with gr.Row():
-            rlhf_method = gr.Dropdown(choices=["None", "Reward Modeling", "PPO", "DPO"], value="None", scale=1)
             dpo_beta = gr.Slider(value=0.1, minimum=0, maximum=1, step=0.01, scale=2)
             reward_model = gr.Dropdown(scale=2)
             refresh_btn = gr.Button(scale=1)
@@ -101,6 +104,7 @@ def create_train_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dic
         top_elems["quantization_bit"],
         top_elems["template"],
         top_elems["source_prefix"],
+        stage,
         dataset_dir,
         dataset,
         max_source_length,
@@ -122,7 +126,6 @@ def create_train_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dic
         lora_dropout,
         lora_target,
         resume_lora_training,
-        rlhf_method,
         dpo_beta,
         reward_model,
         output_dir
@@ -142,6 +145,7 @@ def create_train_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dic
     )
 
     return dict(
+        stage=stage,
         dataset_dir=dataset_dir,
         dataset=dataset,
         data_preview_btn=data_preview_btn,
@@ -170,7 +174,6 @@ def create_train_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dic
         lora_target=lora_target,
         resume_lora_training=resume_lora_training,
         rlhf_tab=rlhf_tab,
-        rlhf_method=rlhf_method,
         dpo_beta=dpo_beta,
         reward_model=reward_model,
         refresh_btn=refresh_btn,
diff --git a/src/llmtuner/webui/locales.py b/src/llmtuner/webui/locales.py
index 0c7abae0..d95fde29 100644
--- a/src/llmtuner/webui/locales.py
+++ b/src/llmtuner/webui/locales.py
@@ -546,7 +546,15 @@ LOCALES = {
         "zh": {
             "value": "开始导出"
         }
-    }
+    },
+    "stage": {
+        "en": {
+            "label": "train stage"
+        },
+        "zh": {
+            "label": "训练阶段"
+        }
+    },
 }
 
 
diff --git a/src/llmtuner/webui/runner.py b/src/llmtuner/webui/runner.py
index afef4736..00918297 100644
--- a/src/llmtuner/webui/runner.py
+++ b/src/llmtuner/webui/runner.py
@@ -8,11 +8,11 @@ from transformers.trainer import TRAINING_ARGS_NAME
 from typing import Any, Dict, Generator, List, Tuple
 
 from llmtuner.extras.callbacks import LogCallback
-from llmtuner.extras.constants import DEFAULT_MODULE
+from llmtuner.extras.constants import DEFAULT_MODULE, DEFAULT_TEMPLATE, DEFAULT_TEMPLATE_WITH_CUSTOM_MODEL
 from llmtuner.extras.logging import LoggerHandler
 from llmtuner.extras.misc import torch_gc
 from llmtuner.tuner import run_exp
-from llmtuner.webui.common import get_model_path, get_save_dir
+from llmtuner.webui.common import get_model_path, get_save_dir, get_template
 from llmtuner.webui.locales import ALERTS
 from llmtuner.webui.utils import gen_cmd, get_eval_results, update_process_bar
 
@@ -70,6 +70,7 @@ class Runner:
         quantization_bit: str,
         template: str,
         source_prefix: str,
+        stage: str,
         dataset_dir: str,
         dataset: List[str],
         max_source_length: int,
@@ -91,7 +92,6 @@ class Runner:
         lora_dropout: float,
         lora_target: str,
         resume_lora_training: bool,
-        rlhf_method: str,
         dpo_beta: float,
         reward_model: str,
         output_dir: str
@@ -113,7 +113,7 @@ class Runner:
             checkpoint_dir=checkpoint_dir,
             finetuning_type=finetuning_type,
             quantization_bit=int(quantization_bit) if quantization_bit != "None" else None,
-            template=template,
+            template=get_template(template, model_name),
             source_prefix=source_prefix,
             dataset_dir=dataset_dir,
             dataset=",".join(dataset),
@@ -138,16 +138,18 @@ class Runner:
         )
         args[compute_type] = True
 
-        if rlhf_method == "Reward Modeling":
+        if stage == "Pretraining":
+            args["stage"] = "pt"
+        if stage == "Reward Modeling":
             args["stage"] = "rm"
             args["resume_lora_training"] = False
-        elif rlhf_method == "PPO":
+        elif stage == "PPO":
             args["stage"] = "ppo"
             args["resume_lora_training"] = False
             args["reward_model"] = reward_model
             args["padding_side"] = "left"
             val_size = 0
-        elif rlhf_method == "DPO":
+        elif stage == "DPO":
             args["stage"] = "dpo"
             args["resume_lora_training"] = False
             args["dpo_beta"] = dpo_beta
@@ -195,7 +197,7 @@ class Runner:
             checkpoint_dir=checkpoint_dir,
             finetuning_type=finetuning_type,
             quantization_bit=int(quantization_bit) if quantization_bit != "None" else None,
-            template=template,
+            template=get_template(template, model_name),
             source_prefix=source_prefix,
             dataset_dir=dataset_dir,
             dataset=",".join(dataset),

From 4a4623cf2d1c87c52ce8ef8e9c4823814f6e367a Mon Sep 17 00:00:00 2001
From: codemayq <codingma@pku.edu.cn>
Date: Mon, 14 Aug 2023 20:56:05 +0800
Subject: [PATCH 31/45] auto match template when change model_name

Former-commit-id: 0bf892ff1a4b9253ee569a7b0dc5270762af13d3
---
 src/llmtuner/webui/common.py         | 3 ---
 src/llmtuner/webui/components/top.py | 3 ++-
 src/llmtuner/webui/runner.py         | 4 ++--
 3 files changed, 4 insertions(+), 6 deletions(-)

diff --git a/src/llmtuner/webui/common.py b/src/llmtuner/webui/common.py
index 4e1ea0ae..728fbd6b 100644
--- a/src/llmtuner/webui/common.py
+++ b/src/llmtuner/webui/common.py
@@ -49,11 +49,8 @@ def get_model_path(model_name: str) -> str:
 
 
 def get_template(
-        template: str,
         model_name: str,
 ) -> str:
-    if template and template != "default":
-        return template
     if model_name == "Custom":
         model_name_or_path = get_model_path(model_name)
         # get last dir
diff --git a/src/llmtuner/webui/components/top.py b/src/llmtuner/webui/components/top.py
index 8611e280..caceb1e5 100644
--- a/src/llmtuner/webui/components/top.py
+++ b/src/llmtuner/webui/components/top.py
@@ -4,7 +4,7 @@ import gradio as gr
 
 from llmtuner.extras.constants import METHODS, SUPPORTED_MODELS
 from llmtuner.extras.template import templates
-from llmtuner.webui.common import list_checkpoint, get_model_path, save_config
+from llmtuner.webui.common import list_checkpoint, get_model_path, save_config, get_template
 from llmtuner.webui.utils import can_quantize
 
 if TYPE_CHECKING:
@@ -39,6 +39,7 @@ def create_top() -> Dict[str, "Component"]:
     ) # do not save config since the below line will save
 
     model_path.change(save_config, [lang, model_name, model_path])
+    model_path.change(get_template, [model_name], [template])
 
     finetuning_type.change(
         list_checkpoint, [model_name, finetuning_type], [checkpoints]
diff --git a/src/llmtuner/webui/runner.py b/src/llmtuner/webui/runner.py
index 00918297..07811258 100644
--- a/src/llmtuner/webui/runner.py
+++ b/src/llmtuner/webui/runner.py
@@ -113,7 +113,7 @@ class Runner:
             checkpoint_dir=checkpoint_dir,
             finetuning_type=finetuning_type,
             quantization_bit=int(quantization_bit) if quantization_bit != "None" else None,
-            template=get_template(template, model_name),
+            template=template,
             source_prefix=source_prefix,
             dataset_dir=dataset_dir,
             dataset=",".join(dataset),
@@ -197,7 +197,7 @@ class Runner:
             checkpoint_dir=checkpoint_dir,
             finetuning_type=finetuning_type,
             quantization_bit=int(quantization_bit) if quantization_bit != "None" else None,
-            template=get_template(template, model_name),
+            template=template,
             source_prefix=source_prefix,
             dataset_dir=dataset_dir,
             dataset=",".join(dataset),

From 02a61b08b10a6c573be7fe0f069266419e12e224 Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Mon, 14 Aug 2023 22:45:26 +0800
Subject: [PATCH 32/45] update webui

Former-commit-id: 9d0f6214b68a653c0a67632437b227ab8f589bed
---
 src/llmtuner/extras/constants.py       | 36 ++++++++++----------------
 src/llmtuner/extras/misc.py            |  4 ---
 src/llmtuner/extras/template.py        |  4 +--
 src/llmtuner/webui/common.py           | 20 ++++----------
 src/llmtuner/webui/components/top.py   |  5 ++--
 src/llmtuner/webui/components/train.py | 10 +++----
 src/llmtuner/webui/locales.py          | 30 ++++++++-------------
 src/llmtuner/webui/runner.py           | 16 ++++++------
 8 files changed, 47 insertions(+), 78 deletions(-)

diff --git a/src/llmtuner/extras/constants.py b/src/llmtuner/extras/constants.py
index ffdc8827..cd22943f 100644
--- a/src/llmtuner/extras/constants.py
+++ b/src/llmtuner/extras/constants.py
@@ -10,7 +10,13 @@ LAYERNORM_NAMES = ["norm", "ln_f", "ln_attn", "ln_mlp"]
 
 METHODS = ["full", "freeze", "lora"]
 
-STAGES = ["Supervised Finetuning", "Reward Modeling", "PPO", "DPO", "Pretraining"]
+STAGES = [
+    "SFT",
+    "Reward Modeling",
+    "PPO",
+    "DPO",
+    "Pre-Training"
+]
 
 SUPPORTED_MODELS = {
     "LLaMA-7B": "huggyllama/llama-7b",
@@ -23,6 +29,10 @@ SUPPORTED_MODELS = {
     "LLaMA2-7B-Chat": "meta-llama/Llama-2-7b-chat-hf",
     "LLaMA2-13B-Chat": "meta-llama/Llama-2-13b-chat-hf",
     "LLaMA2-70B-Chat": "meta-llama/Llama-2-70b-chat-hf",
+    "ChineseLLaMA2-7B": "ziqingyang/chinese-llama-2-7b",
+    "ChineseLLaMA2-13B": "ziqingyang/chinese-llama-2-13b",
+    "ChineseLLaMA2-7B-Chat": "ziqingyang/chinese-alpaca-2-7b",
+    "ChineseLLaMA2-13B-Chat": "ziqingyang/chinese-alpaca-2-13b",
     "BLOOM-560M": "bigscience/bloom-560m",
     "BLOOM-3B": "bigscience/bloom-3b",
     "BLOOM-7B1": "bigscience/bloom-7b1",
@@ -41,12 +51,13 @@ SUPPORTED_MODELS = {
     "Qwen-7B": "Qwen/Qwen-7B",
     "Qwen-7B-Chat": "Qwen/Qwen-7B-Chat",
     "XVERSE-13B": "xverse/XVERSE-13B",
-    "ChatGLM2-6B": "THUDM/chatglm2-6b"
+    "ChatGLM2-6B-Chat": "THUDM/chatglm2-6b"
 }
 
 DEFAULT_MODULE = {
     "LLaMA": "q_proj,v_proj",
     "LLaMA2": "q_proj,v_proj",
+    "ChineseLLaMA2": "q_proj,v_proj",
     "BLOOM": "query_key_value",
     "BLOOMZ": "query_key_value",
     "Falcon": "query_key_value",
@@ -59,28 +70,9 @@ DEFAULT_MODULE = {
 
 DEFAULT_TEMPLATE = {
     "LLaMA2": "llama2",
+    "ChineseLLaMA2": "llama2_zh",
     "Baichuan": "baichuan",
     "InternLM": "intern",
     "Qwen": "chatml",
     "ChatGLM2": "chatglm2"
 }
-
-# huggingface model name prefix 2 template
-DEFAULT_TEMPLATE_WITH_CUSTOM_MODEL = {
-    "Llama-2": "llama2",
-    "chinese-alpaca-2": "llama2_zh",
-    "alpaca-7b-wdiff": "alpaca",
-    "vicuna": "vicuna",
-    "BELLE": "belle",
-    "Chinese-LLaMA-2": "linly",
-    "BiLLa": "billa",
-    "Ziya": "ziya",
-    "aquilachat": "aquila",
-    "internlm": "intern",
-    "aquilachat": "aquila",
-    "internlm": "intern",
-    "Baichuan":"baichuan",
-    "starchat":"starchat",
-    "Qwen":"chatml",
-    "chatglm2":"chatglm2"
-}
\ No newline at end of file
diff --git a/src/llmtuner/extras/misc.py b/src/llmtuner/extras/misc.py
index b0a7365c..b3e0e4b1 100644
--- a/src/llmtuner/extras/misc.py
+++ b/src/llmtuner/extras/misc.py
@@ -95,7 +95,6 @@ def prepare_model_for_training(
     use_gradient_checkpointing: Optional[bool] = True,
     layer_norm_names: Optional[List[str]] = LAYERNORM_NAMES
 ) -> "PreTrainedModel":
-
     for name, param in model.named_parameters():
         if param.ndim == 1 and any(layer_norm_name in name for layer_norm_name in layer_norm_names):
             param.data = param.data.to(torch.float32)
@@ -112,9 +111,6 @@ def prepare_model_for_training(
         model.config.use_cache = False # turn off when gradient checkpointing is enabled
 
     if finetuning_type != "full" and hasattr(model, output_layer_name):
-        if hasattr(model, "config") and hasattr(model.config, "pretraining_tp"):
-            model.config.pretraining_tp = 1 # disable TP for LoRA (https://github.com/huggingface/peft/pull/728)
-
         output_layer: torch.nn.Linear = getattr(model, output_layer_name)
         input_dtype = output_layer.weight.dtype
 
diff --git a/src/llmtuner/extras/template.py b/src/llmtuner/extras/template.py
index e8d228a4..bb2886d5 100644
--- a/src/llmtuner/extras/template.py
+++ b/src/llmtuner/extras/template.py
@@ -273,8 +273,8 @@ register_template(
 
 
 r"""
-Supports: https://huggingface.co/ziqingyang/chinese-alpaca-2-7b
-          https://github.com/ymcui/Chinese-LLaMA-Alpaca-2
+Supports: https://github.com/ymcui/Chinese-LLaMA-Alpaca-2
+          https://huggingface.co/ziqingyang/chinese-alpaca-2-7b
 """
 register_template(
     name="llama2_zh",
diff --git a/src/llmtuner/webui/common.py b/src/llmtuner/webui/common.py
index 728fbd6b..965a690b 100644
--- a/src/llmtuner/webui/common.py
+++ b/src/llmtuner/webui/common.py
@@ -6,7 +6,7 @@ import gradio as gr
 from peft.utils import WEIGHTS_NAME as PEFT_WEIGHTS_NAME
 from transformers.trainer import WEIGHTS_NAME, WEIGHTS_INDEX_NAME
 
-from llmtuner.extras.constants import SUPPORTED_MODELS, DEFAULT_TEMPLATE_WITH_CUSTOM_MODEL, DEFAULT_TEMPLATE
+from llmtuner.extras.constants import DEFAULT_TEMPLATE, SUPPORTED_MODELS
 
 
 DEFAULT_CACHE_DIR = "cache"
@@ -48,20 +48,10 @@ def get_model_path(model_name: str) -> str:
     return user_config["path_dict"].get(model_name, SUPPORTED_MODELS.get(model_name, ""))
 
 
-def get_template(
-        model_name: str,
-) -> str:
-    if model_name == "Custom":
-        model_name_or_path = get_model_path(model_name)
-        # get last dir
-        basename = os.path.basename(model_name_or_path)
-        # prefix match
-        for k, v in DEFAULT_TEMPLATE_WITH_CUSTOM_MODEL.items():
-            if basename.startswith(k):
-                return v
-        return "default"
-
-    return DEFAULT_TEMPLATE.get(model_name.split("-")[0], "default")
+def get_template(model_name: str) -> str:
+    if model_name.endswith("Chat") and model_name.split("-")[0] in DEFAULT_TEMPLATE:
+        return DEFAULT_TEMPLATE[model_name.split("-")[0]]
+    return "default"
 
 
 def list_checkpoint(model_name: str, finetuning_type: str) -> Dict[str, Any]:
diff --git a/src/llmtuner/webui/components/top.py b/src/llmtuner/webui/components/top.py
index 76ead7ee..7f3c6faa 100644
--- a/src/llmtuner/webui/components/top.py
+++ b/src/llmtuner/webui/components/top.py
@@ -4,7 +4,7 @@ import gradio as gr
 
 from llmtuner.extras.constants import METHODS, SUPPORTED_MODELS
 from llmtuner.extras.template import templates
-from llmtuner.webui.common import list_checkpoint, get_model_path, save_config, get_template
+from llmtuner.webui.common import list_checkpoint, get_model_path, get_template, save_config
 from llmtuner.webui.utils import can_quantize
 
 if TYPE_CHECKING:
@@ -36,10 +36,11 @@ def create_top() -> Dict[str, "Component"]:
         list_checkpoint, [model_name, finetuning_type], [checkpoints]
     ).then(
         get_model_path, [model_name], [model_path]
+    ).then(
+        get_template, [model_name], [template]
     ) # do not save config since the below line will save
 
     model_path.change(save_config, [lang, model_name, model_path])
-    model_path.change(get_template, [model_name], [template])
 
     finetuning_type.change(
         list_checkpoint, [model_name, finetuning_type], [checkpoints]
diff --git a/src/llmtuner/webui/components/train.py b/src/llmtuner/webui/components/train.py
index 76f8ff94..6aaeecbb 100644
--- a/src/llmtuner/webui/components/train.py
+++ b/src/llmtuner/webui/components/train.py
@@ -3,10 +3,10 @@ from transformers.trainer_utils import SchedulerType
 
 import gradio as gr
 
+from llmtuner.extras.constants import STAGES
 from llmtuner.webui.common import list_checkpoint, list_dataset, DEFAULT_DATA_DIR
 from llmtuner.webui.components.data import create_preview_box
 from llmtuner.webui.utils import can_preview, get_preview, gen_plot
-from llmtuner.extras.constants import STAGES
 
 if TYPE_CHECKING:
     from gradio.components import Component
@@ -15,9 +15,7 @@ if TYPE_CHECKING:
 
 def create_train_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict[str, "Component"]:
     with gr.Row():
-        stage = gr.Dropdown(choices=STAGES,
-                            value="Supervised Finetuning", scale=2)
-
+        training_stage = gr.Dropdown(choices=STAGES, value=STAGES[0], scale=2)
         dataset_dir = gr.Textbox(value=DEFAULT_DATA_DIR, scale=2)
         dataset = gr.Dropdown(multiselect=True, scale=4)
         data_preview_btn = gr.Button(interactive=False, scale=1)
@@ -104,7 +102,7 @@ def create_train_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dic
         top_elems["quantization_bit"],
         top_elems["template"],
         top_elems["source_prefix"],
-        stage,
+        training_stage,
         dataset_dir,
         dataset,
         max_source_length,
@@ -145,7 +143,7 @@ def create_train_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dic
     )
 
     return dict(
-        stage=stage,
+        training_stage=training_stage,
         dataset_dir=dataset_dir,
         dataset=dataset,
         data_preview_btn=data_preview_btn,
diff --git a/src/llmtuner/webui/locales.py b/src/llmtuner/webui/locales.py
index d95fde29..7a58c4c7 100644
--- a/src/llmtuner/webui/locales.py
+++ b/src/llmtuner/webui/locales.py
@@ -87,6 +87,16 @@ LOCALES = {
             "info": "默认使用的系统提示词"
         }
     },
+    "training_stage": {
+        "en": {
+            "label": "Stage",
+            "info": "The stage to perform in training."
+        },
+        "zh": {
+            "label": "训练阶段",
+            "info": "目前采用的训练方式。"
+        }
+    },
     "dataset_dir": {
         "en": {
             "label": "Data dir",
@@ -343,16 +353,6 @@ LOCALES = {
             "label": "RLHF 参数设置"
         }
     },
-    "rlhf_method": {
-        "en": {
-            "label": "RLHF method",
-            "info": "The RLHF algorithm to adopt."
-        },
-        "zh": {
-            "label": "RLHF 方法",
-            "info": "RLHF 阶段使用的算法。"
-        }
-    },
     "dpo_beta": {
         "en": {
             "label": "DPO beta",
@@ -546,15 +546,7 @@ LOCALES = {
         "zh": {
             "value": "开始导出"
         }
-    },
-    "stage": {
-        "en": {
-            "label": "train stage"
-        },
-        "zh": {
-            "label": "训练阶段"
-        }
-    },
+    }
 }
 
 
diff --git a/src/llmtuner/webui/runner.py b/src/llmtuner/webui/runner.py
index 07811258..49fed19b 100644
--- a/src/llmtuner/webui/runner.py
+++ b/src/llmtuner/webui/runner.py
@@ -8,11 +8,11 @@ from transformers.trainer import TRAINING_ARGS_NAME
 from typing import Any, Dict, Generator, List, Tuple
 
 from llmtuner.extras.callbacks import LogCallback
-from llmtuner.extras.constants import DEFAULT_MODULE, DEFAULT_TEMPLATE, DEFAULT_TEMPLATE_WITH_CUSTOM_MODEL
+from llmtuner.extras.constants import DEFAULT_MODULE
 from llmtuner.extras.logging import LoggerHandler
 from llmtuner.extras.misc import torch_gc
 from llmtuner.tuner import run_exp
-from llmtuner.webui.common import get_model_path, get_save_dir, get_template
+from llmtuner.webui.common import get_model_path, get_save_dir
 from llmtuner.webui.locales import ALERTS
 from llmtuner.webui.utils import gen_cmd, get_eval_results, update_process_bar
 
@@ -70,7 +70,7 @@ class Runner:
         quantization_bit: str,
         template: str,
         source_prefix: str,
-        stage: str,
+        training_stage: str,
         dataset_dir: str,
         dataset: List[str],
         max_source_length: int,
@@ -138,21 +138,21 @@ class Runner:
         )
         args[compute_type] = True
 
-        if stage == "Pretraining":
-            args["stage"] = "pt"
-        if stage == "Reward Modeling":
+        if training_stage == "Reward Modeling":
             args["stage"] = "rm"
             args["resume_lora_training"] = False
-        elif stage == "PPO":
+        elif training_stage == "PPO":
             args["stage"] = "ppo"
             args["resume_lora_training"] = False
             args["reward_model"] = reward_model
             args["padding_side"] = "left"
             val_size = 0
-        elif stage == "DPO":
+        elif training_stage == "DPO":
             args["stage"] = "dpo"
             args["resume_lora_training"] = False
             args["dpo_beta"] = dpo_beta
+        elif training_stage == "Pre-Training":
+            args["stage"] = "pt"
 
         if val_size > 1e-6:
             args["val_size"] = val_size

From d15fe288dff4c5523eda1286aa0ff0c8e8f32d9e Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Tue, 15 Aug 2023 00:07:56 +0800
Subject: [PATCH 33/45] alert pad_token source

Former-commit-id: 80b4053602c02aec724ecf980f8a279ffdf9f975
---
 src/llmtuner/extras/template.py   | 5 ++++-
 src/llmtuner/tuner/core/loader.py | 2 +-
 2 files changed, 5 insertions(+), 2 deletions(-)

diff --git a/src/llmtuner/extras/template.py b/src/llmtuner/extras/template.py
index bb2886d5..62153b0d 100644
--- a/src/llmtuner/extras/template.py
+++ b/src/llmtuner/extras/template.py
@@ -204,7 +204,10 @@ def get_template_and_fix_tokenizer(
         logger.info("Add eos token: {}".format(tokenizer.eos_token))
 
     if tokenizer.pad_token_id is None:
-        tokenizer.pad_token = tokenizer.eos_token
+        if tokenizer.unk_token_id is not None:
+            tokenizer.pad_token = tokenizer.unk_token
+        else:
+            tokenizer.pad_token = tokenizer.eos_token
         logger.info("Add pad token: {}".format(tokenizer.pad_token))
 
     tokenizer.add_special_tokens(dict(additional_special_tokens=template.stop_words))
diff --git a/src/llmtuner/tuner/core/loader.py b/src/llmtuner/tuner/core/loader.py
index e480d44c..4bf767a6 100644
--- a/src/llmtuner/tuner/core/loader.py
+++ b/src/llmtuner/tuner/core/loader.py
@@ -154,7 +154,7 @@ def load_model_and_tokenizer(
         model.generate = MethodType(PreTrainedModel.generate, model)
 
     # Fix LM head (for ChatGLM2)
-    if not hasattr(model, "lm_head"):
+    if not hasattr(model, "lm_head") and hasattr(model, "transformer"):
         setattr(model, "lm_head", model.transformer.output_layer)
 
     # Register auto class to save the custom code files.

From d37be06aa03720e34aeaaba934940276def401cc Mon Sep 17 00:00:00 2001
From: Yanyi Liu <wolfsonliu@163.com>
Date: Tue, 15 Aug 2023 11:13:45 +0800
Subject: [PATCH 34/45] Add .gitignore

Former-commit-id: 448478f9384a65f48895d5e0983a39a1296ebc36
---
 .gitignore | 160 +++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 160 insertions(+)
 create mode 100644 .gitignore

diff --git a/.gitignore b/.gitignore
new file mode 100644
index 00000000..51e9d59e
--- /dev/null
+++ b/.gitignore
@@ -0,0 +1,160 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[cod]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py,cover
+.hypothesis/
+.pytest_cache/
+cover/
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+db.sqlite3-journal
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+.pybuilder/
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+#   For a library or package, you might want to ignore these files since the code is
+#   intended to run in multiple environments; otherwise, check them in:
+# .python-version
+
+# pipenv
+#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
+#   However, in case of collaboration, if having platform-specific dependencies or dependencies
+#   having no cross-platform support, pipenv may install dependencies that don't work, or not
+#   install all needed dependencies.
+#Pipfile.lock
+
+# poetry
+#   Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+#   https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
+#poetry.lock
+
+# pdm
+#   Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
+#pdm.lock
+#   pdm stores project-wide configurations in .pdm.toml, but it is recommended to not include it
+#   in version control.
+#   https://pdm.fming.dev/#use-with-ide
+.pdm.toml
+
+# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
+__pypackages__/
+
+# Celery stuff
+celerybeat-schedule
+celerybeat.pid
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre/
+
+# pytype static type analyzer
+.pytype/
+
+# Cython debug symbols
+cython_debug/
+
+# PyCharm
+#  JetBrains specific template is maintained in a separate JetBrains.gitignore that can
+#  be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore
+#  and can be added to the global gitignore or merged into this file.  For a more nuclear
+#  option (not recommended) you can uncomment the following to ignore the entire idea folder.
+#.idea/

From 603cb30887c4486fafddbe77df674062680567e7 Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Tue, 15 Aug 2023 11:13:46 +0800
Subject: [PATCH 35/45] Update wechat.jpg

Former-commit-id: a7dd9611dbd53f8cc4bb471c4f335558a349ea52
---
 assets/wechat.jpg | Bin 145701 -> 145075 bytes
 1 file changed, 0 insertions(+), 0 deletions(-)

diff --git a/assets/wechat.jpg b/assets/wechat.jpg
index 35932dd202142e5d6b88acacc360cde9e055dec3..da80faaefc155f603d5cde05430135b87ed9e6ae 100644
GIT binary patch
literal 145075
zcmeFZ2RK}ByEnS@PV{b)AW<WF)RCe^^f0<4IzdDmb(CnM1wlx3LI@&y?>)MZAQ&UM
zm{A8aocaIX{l4$D_xY}C?{mKEJKx#w_g}7uS+kyHJ>`Dx-~GGq`&pOYFPDKU57ji)
z00II4Ai)0sm-E1VfRvbol!Ta+l!TOwjFg=63MC~41tk+L9rYDfCN?%!CKeVBZhjC4
zCodNZ%T<x9yf<zN3JJ1<L~n}<+~yY$6!_zx5Rj3PQIb<KQc^MsaIkO){69V}I{<oe
zl38LhB7*AxAw2;RJ;7xU0K&JEgy7Eu__u?AkO<#iGI9z^Dtv|7D*zz@5fLFV5eW$~
zF}`*X{yIQRPr`6b{4Ob@{!_B+9!wIUNqOYF_o_OX4JJ^0lFwd-QBbn5vaxgU-?%9t
zC?q8<BYQ_q{{911HFXWmhlY=hj3FkbW;V8V_709t&YoW0KE6;t|L|84kx|hx@Z^*?
zscGqNGcxlF3X47zmwYV!Qe9J9SKrXs)YaY7+t)uZI5asm{cUFU`;R%q3Uc+=+WN-k
z7JC2S@aXsib9(kizX$-LzjX`$`)?il5B;LY_luC2n24C{kA4vl`r-=_Ju%5OaZ-l6
z`eaW%7_Uo&k~7^)%B$+6;FUB$F+Y1bLCL}=h2Tg3(X~H&_FwB**#E1Z{i9?5?AIJX
zO+<j7JR*7k44i#{3*-|1RsI@-zhvMq8Td;E{*r;eWZ*9u_)7-<_sKwk(~4naD%rOP
zTk9poSD)1B&(3CobwB3PzT3SG8*^;mt5TnPatUNlz?PmhzRxj}2y!I)aQ+B%Fei4t
zqksGsyk28{^1Z_0=1g3U8L-(lMTR|&y*OtI`apNO%ZfV|&KNu@xH%Id?_Se+2`J@X
z0&Tq@r|^pa)k`1<zV<T6aq#GW*TZy7ZGssd*z9!7iox|X6uI=sj8tt+tn1KR>%Vrs
zRjQCM>Hq7M$KUw&cj?a6ncb>u$1q2_Czn88F152j{@bws!a4*vHg;?!=G6Ktgx{Cr
zgfA4IVPP4$R9^b3*CPJaM$5qkhTh28OI&qPtFCRw9O|C*?@JNHt1=m0(N|>x{$0_`
zlvJToz2P{JOOl^*JuQS%ApdV=Y4Q?4_KEeMW_Fb0#QsknFFgXVm+gd6#nGLbTGN{b
zL;YMF+TWclzt>!UeF@OW7>VCFz0KsUoJyLrLZvXsu3z_=9XKh?7X4-uM)aI+&(<Qa
zwOZ@#?CWS~qK0?|$<}i+;^jEdP`*2hql)d;?a<$xSm80TH^Uw^yYvqfVC+g`zl@}5
zWQh2*ubvOkZ5piGsNTcr>gY0WB%VmKnWNlM#g{;rRzc2baN<4~VWoVWtv+&P{-~Tr
zZ1jLr{P!tz?4i{-3}QL<@Llg2HQg6i9j2$F<nE}FcElajI*w%b65zdHUg42UwSU8^
zUA;~=eg<(sEiQGPEj;VG1hnBws(!<gbhm@O1Q#+UH1@#w+m`-*TeFQf%~7GR%qU;j
z-9|32rxwAr2wj@*P5F+)H&>RBmc}vEGG~@8GM`Vc<Rf@TUkIf%zR!Dh#1xkHX+R(!
z4hX}vE`db0cF>v&Vf|q{1|5sOun!TPP-Ti_(pLrkl?roK`6bYw4{HM_m2-^F;Mdax
zcnLHxPAbi$|LCqNlOBPQyVSh4MqhgA*OHeFch<@cTz!c9nxjv8AHkz~{PyBMJ(*Z1
z9hYbu*P-#u48rxC7%nWPEs~p@aIM-*HjTyT)kIGUdr}iz0*kAV_giKS9R6QB+`M+g
zYYr<s>%`@u4zIosoY0*A{Bf4q^1Ab*)4=0IX*GgM(G*D109;0al;Uhwv%t#IY_7I*
zAWG3dWb@i&Dj2!jWH8JpKD5m~pW5;1RBWVUA^Dg2>5TD5Up~D8B+Yxq-A4|hWvB5a
z!_RfF43F1mGwYXe#w%U8@42{6BHofur+YHn8hZl+BI47te#_A3{OEhG4;+6pMJ?Ph
z-qoh(b98RI)zDmjS{Q_H`H##)UCPlDWuDKOY9hOAeA#km5V@6zsl&d6pQ5gP-v*zA
ztTnmQ>eQQwW_1Y|K@UR27i+>B$MmL{YQoZL<+}b*|NjuB&;JxKN5Y>>`WjFF80dec
zVmh~k^}l(JJmr}9wCjGM#E3&<;{hAITJQN|^)N@R1nKN)!57W%U^`5rc)m62XI#<Z
zdPHzQEB7(8Nr>}3H)_Uv-jrYKE}uKkH*X}@3AStyt3@9w_WakA6!W|@LchwAt99@>
zrL^tZAQEVC?;dSY>$i~=`51TF7e97@9i=i_!;c#Kc<q-eAL3-=l>b-+p<CqJ#x)}~
z;_#?-YQj$QB<=op>UC}*>quV~pUk+-;9{5BU)s$>ar9P|xll3fsTrf%ddli6I%H2g
zRU-tmL4GdY4q|EV8qaH)M;mP#U2hqE{UlRg8$=U;ct`cF<fgcgbR#mMc5JJSdJ<wB
zWDmW**be4xmE|O>aTrZsG;m?rq^f?v#XYjvOh>v2k;GN%)rmC;j_>}Qwq47c6`3`h
zc0Ur<dg9v69Iw)+TkhLN^*Q}6>g)bOcT6#m=Jr-9YQz)VD?0)!leos$9ATZO2tT6X
z`1bs8IKZ2YJ{MHvm;Op5A#bHtg8CDY-&y97M?YajMymdyzUuQQe-3<m5q`PHStlMy
zhnk;RW1+p6OF&YEWg{Pc{~xgWcX<8p{`a5iY$pC6_J$)72VCI8)qJ=BP-<9t7L<NI
zg84SrGh4-zSM)REVXUv`FC@-h1*ueGqi&Rhn=1$?nscV**eA|-6W`qDlHtzIGc$3v
zD@h0wc`Sys@qRI3-WMZOF9&l@i%+SFFIFuo?$P|D*h9_grK8fZo`fx%nL2r@o=eyw
z+0PMpGFOycOfzIY=9*F~-*9Jazw!afJ=(TmiuGK}(f&L!yI}h4@qWZ_@BTIE&<gAM
zB>oTd-_au$)6&hHY#gTcB<d$}(!+&}+t1y9Ujh?`pERd)JPn2(uJNQ4kQ_XEDG;MM
z`hswEI`iF@Cu@zQf(rbOfDCfKJVSJ5ktu#H?uKtQt%PSs@3VNnnjhvk;)Uw>xJ$q)
zt5nU?-)pR3@e$hDoX`DDNC8%O+KYW;G|5G>?$~P(X65o$etPHSBAbDEBN*qrb@=ho
z%oPnhT-b{=?zcDZLLLp-b!)0Je|)<okRM9oz6nDkS&`V0C&m})mjI7=F!<{@Ns~bS
z>pvCYf58?e{O?ldAJ(gM!9SHA(EEql|C`a@2#QgDwt?0pTcp<AufT(zOJ71hr>gsk
zJsVcQMm^Elm%was*IZ=P?d;4b$f#@aZV(au8ntGknKfh2jN}Nb$--XT8OJc|Jxe9>
zJ}Tcrg|vzz&eJuM6`#`!gI`UQN0$IW=enaN@g*RYRAPOy-TNUanLe>i!e#n`$+_%h
zRdxMW$kYDkzBmCPiK_8_?)PWYJ~wXjk$Hx-7dgAghRw+F+4VR!>W>Err^fckawPV=
zsjsazPFSL}B}&_#X{LdxzcZ+9T)227BH6!4Hd+0vS=NX-kppeq?CBlE^k|KLi-`x4
zIX-#xlq>B(lUIWX=-;%ks`B7}KYIgFuVa#LGyXy_O+Et-45tS^A(%GWUl<uPTwMKT
z+xV0}&&g&%NA_90D20p0-~t8LS_`OqQ387Nag$2vduj*Kqe}O2H@u{2)HdlhN{}*N
zi4uu2aMstD;1+Zu5q0+blN$XcaQ_fEVkKda^@O3xxy<RJ7=LVFf!xUR2^B82dep^~
zaWr3girU?QC%KuuQ3&QZ{T+^?3HDKH7pnqQ0?!c;LN(<Q_-e3StSjog9@WOxoA+sD
z-l*}q#gU54`egH&0M+}w$Ut%02tVnsz6`n|xi6QgL1vrw?^~Y7z7i2ndXyi!a1i8x
z(2gvIx{c;}Q19rhM3XC?alq+brRfz2e7mI+r0;3<PL)t`bPA!IMKi3;kL&q<o|p8Z
zc0=Co?bOb-TD8ThTMJ0`lV0=uw|FOk#-Ezt|9I2s8Qkoy*53^B|IAMOC$5_?{{!x&
z2;ATTg}48Jy?>=r$}v8YGZA{>9h{-~%rS#AkBMM7TuY}RCihL5k5uQQdw9<F@0|(%
zgLBB;4{@fdhv3mFk!7c~xpv=Hujsrh0&Kvu5SD+%LBk6GA!n@Px`q6~_G5Dm3+$Jq
zBa4|LCW2I45}sOL0_d@eR?rX7;nF3*bqQ?Gl>ZA|l2_taUwTuIC1WmD1=OhpDR)%~
z3-O%)58nS*`D+aR|1AT#tu1v`uSNnqXj<-v9o$jHjsUu<&rG28-I)=C^`FtCN!~AL
zmxQRk|E%5&`BC2VpNn+qEaW&Xq0HNxSfu-^=2ZiAq_HC&BH(K)R3_NrjDko-q*ol1
z!<HkVnc!<}<{m8(2EMAg7aN_SyVHgnKz`B>efF%d(zs+DfSIOzw|wyu_+rtm{MBuL
zL1w8!-Yb^^&n*Fg61R%oa%QvND{RaSr0SG(>jtXVwKxVD%L=H=1P6G(j}w*rQLR!K
z-QGt9rU|@1)gnm~yTv&rJ9!W3UBxThe?NlcE!UGPo^v0uW|%Kh?ooBYd?@LWgxusU
zE!#;gP6?agq-Y``&s6J~5T!8q+HQ+Y0gbNKM5Q!2@2pt)By{(-?;Rl>b+-PYpXU(P
z3ar-jZkwD_HrbTx^P=i(K3Rr=3OhBSNg49zHp4_=^6%Gc0smemh%}nBXCJE2d2+U3
z_WUfSPA~AS2HP0t+apSa*Aztc__&N?g%|>yC~|QXYApJ2K?IVNBeee=bYx&te}1zB
zJk7D<4VqKvk`K=Fc96-csHxj;xn)59Fu5>1Zg>g~9G41r!w??=xi0~gu`95pXc%S}
zgql*Oy##U{qQQGTV60Dw(hz)Ik!a<djtlB0>Y^FB#c68!nt!OYLFmOx=4hNt0Y1nj
z9EO=+v2^KDCTg4sR7|T>Wa#5Re_+?u*YBo6{F1HGBF+I0>^W9|miuFi(DmKBVB=<6
z*;7ZiI)LnLF?+{mQVnvk`40`yn8*V~s-i^WYnP(!G6in&iC;^&ehJ)W-%yGGybd9F
zl<2;PA-aH7!ca6Dmq26$iVTaVk(a<S?>Je7zLh%O?LxIocY4u!33RBJWrF*4iJi}C
z)85(n`OHF6iduwbDhPa^lv3;tG7)V3=}$Ghc|a}UM?yTW)3uyRlOUP!ZN-{YS;WQa
zCVHk8(azFo=Mn(8Usv!FVn#mua&3?=JL-bak!a!1^;+Su0*(_^q(Uf^yPe(4dUlOB
zf|mM=>`#kj<hDLm8NC*T9J#sM#64;W9PAdZm{9C=pT&i%a2qi}qOjJO;z|tN%B}b(
z$H%<gzSre!-}dzxYoFV)|BMSA(^&~o>ViLbvcaw)q;{8k6Tdhb_F`aHYAT~LU5kF#
zx(6vjB;M+oU9TczU-uMaUe7^IWz+n-xtoOVTayO*q55M3wQce?>*T*eeB*{W$EY8M
zM2Nh6@aoWTt`w_{DJX+-u2_W^Z^;->K+SfD_{WD8!uPAj2z<nidEbZ0t%WEh!w={Z
zt?8{hf*I=HJxpf#wpT>q%tfXb8|lld822>zlkgXm^zZVIP%W)iG))W3Tu~2|$aoj|
z<D0K<YhFdNb0;_w`TWVO!>ESMV{GA6Z|#!7ZP5rgPG!?@oIurnYwkFyK)}`d@#oKm
z1F{eb7T8|ClB#Br3x``JeY)nf$X-6T)tp_b`byzA37+f!GZ^!~=<^+LT>F(L7(aws
z8+pnWg!)-^Xd*S{S7K1995@6iSD6(ELaq0K*(>Ox1GUQrT|a2@`g!V5rL8AX`vGEX
zqbkUD>4J0J5!q<vTN_r~rZPQM_bC&at(@ePf@J)?DoYQWIfEp1>Z-oPKdY;Ye-^{^
zqO<_eBO5i_+iAWsl~E&hWx)N7E&sw*^=pSi?<?J2d~;bQZN|$}?}-&P`c@7sz{Z)}
z@0UVbpGT_l)ANCaCk}1u0Q`i;u#bmhvXZZ68hgLs$It$Fl!a9<XN3x{kO}?kyJs9>
zOZ~z0jSX-r7?Y#Hm@sj|og?y$te6gfbdc{jmo^!`tgPfbI|u3t^3|VoE2?RY8rWZp
zsr#}9onAf<Q%716AS&TjU^HE%2v!=Q+SWe%qcC|-?+Nz=Rbc6-erIZ%wL7cmgc^dQ
zBj%M$fV4)81tS5bcV41Dl1U>A<%47&6^jeGJKU55o`m=Mr7pz>UkiMO7G29({`g_?
zFxn6IwN0&Ouv<&u^vid!0P_2CCU7<bKLweO2a5%Y)^n^MA2vn2c|?6-G@Y(5a1%4B
zcSo*o`OXm+WZas(CBPj`)}_DY#L*eR=qLL^5Wz<z$G+}FOpi(E3sG<uUf0EF1>VK$
z2`yX%uG|+FzmD8_cM13=e)-)8UR-O7#cQ<FvYrcZG&QK^(2R+rb;WUQ{~Jwi%89KM
zeHGgwUkYZ{<y_BYt}-f+=&5hs(wNQ5v-ciSuj?lDdIg0KhZ2YWz=2cNduXy=F?gnH
zTLz-}<IaxAQ<I9yxISKyhCwm0MZQbG+n-Q!RvPj+XQjAvd)xB9{<K5u6yJ;Tw#^FH
zRi%gA3wc+)JC1LXhA8*sH!ek82!BCWMu^n5Dl%!#GCWv#TgWBBxJ1_O5pE@hrqjZt
z%l2uydEK1tkBk`ksP7IPChg=3b{OEIwv|znlgp*c{{!ewWsmzVfuwGgOQ4bk@5qa6
z`%;+#d+vV9B}I3JcE_=~YzqN2(AizfOMtKB5}3t;2g`nF|NAdWh#ZbxFM(uH#Mn9X
z5}3?znG*hrcgZcCUSGu9?e&Q`!L(M15D7)QOW=A!BVIvEUIH6+mC?b}n0;LW%na2J
z%h7bXD^`6Hp6uWN?xAe;6@}UPMG1V%)GhEWW4{Cr%^`g`6fLBR-pflI{H5V#-*IUx
zUr#(^PVhRO5D32JNVJQ&k+4F(7CNI*sJF58l3ioq>X~6yDpU3A(gMDe5E3ZePBpTB
zT52%xTA3VclYe@R*hwbW{`(UH81Kc1Ij#f?!-XR{5v>xgP1z+;W47*@oWW7(V8~`u
zh@u#>J<+`1N-AkLD<SD=)8dF{H#K$=))5T~F+1Bs+{C8PQNq4U(BS~cQ2Ui&5I*Y2
zj7sc2@qcyKyY=D8NW;KszOKk_B5EsW>~m#l8v|_cHmVpGe)SR<Cjq=`y8>;{xccrr
zN~LXrTLm6(KjkTLJPJZy3>_)A&>sDqQHR%zuo_PMvSye!TX33LwP>GX6?muYx)cc(
zNr((l_QhBt-qzqiW@z=e-v}PTS+Da(FU&0qkv@s2z;`;z_5&T<tXnGk%}WeUF9XS|
zP;(ZHpHwS2`=W-QleZLG&R5l!(Mo0>!l*3gu!flG3QU>lD{TahbYiXk%hMC2UYB^n
zd+ey+cIafrB|ycwypCf$y#!D^Vi&hUs6Z@^i`Q{^Eu0@Wea&~ur6#B+<)xmoUwd1>
zNEAZ8UgX)vJQL{f3HxZa#bvpbURrL^;nC2$G$zkiYzxH-te`AQG&oj~b%mQJNEk62
zexnLaGxVjfi@ck>E=9ngLgyhVw;VBi6YQWB6Ugm}20i>{KAfW!zDt{nQp}N3psRr7
z6aij2#BDdB^Zh(G9FI7ui1P1M)U&co&{rKNhx3maC<!k*z!jqTBE3wnCq0zEJz&R6
zOuqn$xEM%tXYI!6Y|auQ4W2}OGnQh<4r0q7$l{1V&+?_YEgqWLUz0oKqM%*+BJ9lh
zcZMHb)I3U#L^l!mIc|k@Tmr0>dtM49Pf>9n8<W*VY}E$oUAG&_5sB^=Nfw|k1ss_g
z8qyPF-ODnrTWtD}JuYqKS<!EHQUdN_%0ti(EwgX6W0c!3xBat@^Z0z>Nv%J-zYfXJ
z5<JsjUTvd7yCov%u0VgU_xS4`w`KYI*Ofcls2jV>KuA(CSJ^HBMLp8u9&pSBYYW;v
z@&hXWK4~ITsW;!ni0WeYvy@JYoOIuZyMfwBP}JgWlq`H#QS#|?Y0jWq=H*L2x~{{E
z9>4y+oz9HRRzkPGI$C0a!t};l#N0nP9ZQh5)S5f$yFHe#Be$Xh5`tXnrY{7cyb#oJ
zV-eK7Gij3};yBBLbuO<iOe|Gz;PIQ)@XsFGaNx-Lhr@*?Ng^(A44-$BX+4*Cp~=06
z3&iIJ<xK_;gZ4E<VB2dj9QEYIG;CjE4}U_3qYY7X=~$}%g*Q(4Ke&8cxl?X@EP3_S
z9IM~~ecT@V=BC-~B~X2PYS(J*66mWlt{Xfl%AR^AHl1->C?~dX=CD2F5}=d*$*SK{
zS6|0}>uRJiebucIPKogFutHB9!fJD+y&4~6S}r;~9;j$aNIxz~so<mJGdhdXQ_c@j
zglOY3A&<Yn546#>sXq%d_SeFu{ADP}sQEfq=`lu?7`x`2GtMV?JCYF}xf1@2OZN5*
zLTzp9yjjzhM}Je#ze5d1?;ec?I!sr#+(je1F4!l<Dn#3ct8N51&B<Edj!H?r8mCt3
zNHnL+0@ZAxTP_nN)%w=v;*-?!EAffz?~$pSv)nF2u;+=VOkSX~3wIog%f_{N>4Uvp
zyq$YJA^Z}s9=QbaM*y!j0AskErPUG4+183WDx8JLyVgt&d_DDbOnsa4uJJ}o;&BYh
zEfi77g3+0MFm<iEzIiM=Dec-N04;c*@;&oiH;5wm=9U%bWM$15Tfgg5Pba1sRF=-B
zV#5ytqd7(*AT+e2*8{|iUy@7I+y0`(aPOY!iap6=Q<8-#^0p`PO<{6Z(JCNgSXUN`
z4o4+pjfYIM(K3%B>NpjVGpF9=9~|Mpo%Ob5YC)C9qhVCSuY3<<XpM)M4c{od5ikqG
zs=hG5_*$XH`gJKxpk?}&TKX3gRVibpi%)C5HgLWt>Z5Z8&TsY_%bMjx9L3To+l3u(
zjXGXUeOsc*PFcnvp+-T*n|WG&!Q{|$37m00Z%5gA<0WHz%OAZLcL{KKEs>g!d~p17
zKWOC!i4Reeg7t9@x9&MOTOkCBTk6*3n1iNGb)P6JPw~#&Dv;=YK36nuuYU<VN2iRE
z8HOiB7B%8>T0Dz~Wvak2Qd)jbjnlOH+8T`ShBwBZnr>zvwWGv3oXVlCSWRT}VH27~
z_0a73B$52SxZ?1)7aq0Hel9QO!^rv~Q<qHbnT*o;^4|C%wqG&Tks$qnGd%Q)b+uCm
z`mU4<N^_I-X9oQ&d+~6~_bEP4#`{fws&Ze`5@jF<Cg}sTK)roB(8xt8wR}}_jA2Vw
z#7g?b*IXuGmGHKXaLSC7k~$(PS`4d->b8h|Jn3pDZZ%&RrdBuiU3@fT^@C>=l6`mj
ze7!3Xo(b-u5>1krbD_Vr`Ki$T-B0|p7|xHlJ61vzET3;C7Pq$5xzeOgu<UZr{7g<;
zzDCnI_6U-=xkOlgC>A3#w;m<!wBr?$Fy<IH-)-}XW>Ip12ih+;u&pc|D1Qk=#3H`C
zK5b7PJ)ZbFb1E0H5cjlyDx^CpWtI;@f|Bgzk*!Aa&xT?lAGBXSzbYX*T^<>sPR@QN
zU)5nC5uE#ipck`;AejwoQSdC=23>JeY``R3BihRtd!WB|1b)3e*VrBjrtn^hkq>Oj
z`duuVkj4JPU8zf|&wiBiiHNd(9AkEiXE<D0dIoEbOk}=QutGzn-&|GqHh*)x|9U@1
zm?Ya1x|>|}hVxuD)1NMGLH9>T=f+eqPZe1{_wmB{svCq#$pD}rPLGTekr-8%eH|@*
z2}B*{#4ywz<q&9BmyRTG4lBEVW#ziC;mthB32&#V#1bi&-N>tKs}N(6)|>cp8ldG+
z#?##X@U*>G-nU1|x0kH9?t;2?LYh1=)3&NQ^SY_uqR}9K)pQ8Kq2lah4&lZ1`W_Hf
zO?}$r{ZdnUnkn)DxLQAT{8ZAQRl&yHtJIiUG;dcMZQGZr7NbH-XKg996>;twL!r3T
z3mG;=ALavCM4&H1NpfYeenm-LBRhx^ZO3x#;=(9pm+=rA5f%sd^Bf%}_PJldNOny)
z%`Qr!R$S;_n6Yv0++L6%SxsD|4;^zpEkK;awXhaK=X)}XV-)3%cZJT>U4*5p!i++s
zB0|3JFu!PEvlu6hSn5fnuI$4J_s-d8_&Md9sn_lz>*M)Fe>EjD$t->X!)ulKq3`;0
z=v@tFV1UJNp}ObsJ3+N{FMjcZdQaikzJl!iJXxhQgn01Gmd54Qkd9-)?k;%`;~Bp%
znD!(!lNa*6olaF;UY~1??E!IP`4N`R2lpDLq;q6-Vbw|Q?sRWt(67FRCwAbu5#zR%
zFghnjYWsVj!i{2F1lMi#;!YPuR8MrU(QCI`FYgZ85B4cQrop6&)<|o{vC_);7BDU$
zZJWIw{4{&%D_`B-QJiO-n64S#mRU~h6=m$mT#6Bsj<r;S9^Ht~W2q`&o*UKK9c?)8
z{@}7D7TKmed2!A0L_RInJ%E_y;GK_ojnT7Cqc1Ym)K8=Qf$i;IV_cZ@74;qvt4lTr
z)xt36*yOq}lRV<q$M=*o^@gUd+@15>gY3w*g3|hlLjR!gcq0P!XG}JHVVY+tDxllJ
zeccx~FZl6BuVee6q*WvOr2hcQ*hN0)q+hm0`=s_6!F~w4({1P-VuIh@B`Yl{$9Nfi
zjduiWFN)!)XDW9y>|V@dv$pGLe{%OA${yOGPd^}EIvD|;1j=K$OQ3r?){M}0v&L6j
zzz7@B*Gv_Eh3Xd)WJKBGQ$>$4iKsF&jLv6FaaZO@Pu=((Q+^}*VfTg**(ZqC?}?6p
zaBB^$=!{5hd(;{8;=AiD*Cz!s?yku`pd1M5WF8Bi;P{r1gsL`<L{E<ehs`Zs_w(hm
z4ZRo`9Q%C4L{Nze=!2oMpd4eysYYUt-v!j@gc<c|91=4>_xTCmE$f2WaysQ>f<EdB
zed$-uU$pWFBX*<!C4wRXYX6t#x&_9Z-@QE%4BnV4AF!(f**4wsf6CX1w^%|Tvl@5x
z?h^C76z*$xN{p&kzKt4bD=O%*`{vd9^_u>HHlxkUL}NB$x{@u`m)2y;N>0aSN_FAL
z57{$Y6UO9~a7jFsCk~;G<f1;vF6Qo#AB)Z**lBTD*~a+?y<GYB8{5*Aw$<sQ_HjDr
z*b_4w>cT@8)v?4?!_6?a`x^<r7{t!%75?^8A_<cF6!;v)^Ul_d)gl26|8=c4zR>4*
zGZasNuD6hNqn$4SHf7oF6;9!64t7o+Bs~c~1gZ5UQVo9otQz;v?g~*_KTgH-G{F_{
zxeuPFy)9!r)}0b94`z-1p0Dri=PRsr(`U}7<cu2h^=xKoR~XOv40~Y5yO%(bxDn`w
z`-LXW-;Fw;{anLR%L(Oa1!YfNz0zgkW1K-0Y{QHG1pWcm4Vy*y&aFJs_45-R>Ursd
zP&uOx{?JOb0*TU%RV4eg5&c-pn}D9>4$rgKgE_84wFmV(XtdB+$GRS>IL=Iqymt^<
zl%#oO@We{C|Eet^-J^9Wf}>l-cBDlgewzHuO<r8P5&1xg-Q<e;s>>Sq`{HlA*+4Ca
z^nIR1!RU)CLp{Mzn^Q<$pg!gYqFk(-^#vxtFv{@T?`Mv==CF)sd^Q2~cjFGXbXQzB
zk2O$5eHn-m^&Zx$8i&WWT2B`2`Nl>PeQu)k0>;VkLm-Q{zTCm*4%UXVgCN$m1{B>c
zz3YnK6F<=#juo%PILFnlC*2*h;;by$wZiAZu(gmk-I>;nc9miNxf++z<iKMk>wwCo
zqefd}v#CKxmjIpx7os7#7vDkq+EPpBRQ-6P&c2duyak`^NfEUR`rwk!NNDAp#D;V$
zg%jQgIxZP&4FK=rr<rd``x5xv0z13m4D7TKIH8=D$gy|Q2AD?Nk7H=%9A28FoFX?!
ztTDY$-Ak~y&<`N`Q5uJ*#TLiV@ur-Pl1ffoPVntb4)w@f)NE{pS(o`ET-uruKab35
zSQSy2#0y=vUS0Gy(m2K!;nGeA)!DFKuojG5`9(y&5{G<2ocQ)LG65J<Dj{6jarczO
zXj#<<Ek1yhZs<Px64(J-5<Z<B_x=T|_I*{1D(;3VFL9~{l?=G^$A?mfgwSg!(T-Da
z$6U`@NsN~VMFm0^nVLUO7q*!`jU9>6>(5|j<1yh1L@N475o}4}lk)=Ko!h_$5V^3~
zSlC9lwW)oZHm<6D^~E@BySM%DX(iqz5rIP;T!c%)mbsR&+JOJZO)Keh2y9RYj2~vY
zedqE2P~*wPq*-ltGwR{g<0(FgAC{J#B07&NeqH<g!VeS=c;WelJ2uIk7B=61@n+a6
z!&(l@jA>mSqryr~)C9R0Cez%TA3d-!cLH&KNa`0V<Y9xj<GJuTVhKlTk0%C#?U+np
zFYrHjIda;cT>_`?gX>`DR;@lDPDd+QlvvEI7CE2U+w*hUToHND<ErTQPeaC-SFNjH
z*io=K7+0ynzQltS_e0^i#xajPYUu5ghK2)c$J1vGTOEnN2V{d4T2+lPqIuSVk1<&=
zB$(mGIhEY2W;!8!D*Eu9sB?kaFp{In5=E^%r$zn=H?J(`?-t(|0KT~7R@%G7FUhZR
z!4tc(IVZ7#`MpJs#$9tm5?>iTo@vVQr3@;s&rgOZ8Xn^XGfDj}9*XdcYr?)2^6$JH
zslnebfqj7u@DX<DOk}nci}zq9KSk)Clj85lr3AKJ`q~}m-I=z1GRY#bvaDx3l`keN
z2Z}*nq<Z~do=Ac$8;+HS@w41|F7RD6Cr;7j!&hUkP?n#uVzeykhMhYelX=r^5<kpc
zAjd2)>cg^}FJDn<Dur2}Ad0Ge9%i8GzzOp}HB7ux_oo*<zklz%V1Dq0K^I{L{*o#E
zI}aKE{2I$ix<~Zj_WOYyy5&t3&P#wu(X|+Qp#O~$qoqAjAvM7%^18G-`Mh)eS!&<X
z3?55E1NO)QAEQZn;Su3I{!e1whR4S-+Y(K$iD99}vSHF=jH*aF=I@v*U6mg#Gwj*}
z$#<9If5xUO3cX&yvj(mtNU)G|Q+pqX(HUaJ{YBDAnY-4Wy1lQt&4H)r^TSHZvMIdh
ztuUBmLmROd;ZIW75wjY#Yd|#3t{E6i5P}u1#FN}rl0O`6H92P#_z;P2kFYxr^dT>A
zz6;C4Dyw^U*4Y>ca5MeOFNMp|MOY5&SKjaD>*k4NlB>BKFuI_al@LYsV{rDJK=$&C
z!pAfv*KZ-?s#F2Gmo9HH(t*eFiM_4%a{?a%giU|vA3$ER8cJ5ETX^hR=yqn&nY1L;
zt~X^4cUy72*{aqY9lQ~Lmz)eOPD!vG0-FO)K-G(}ekFmoFj>Bonx38|bD9p6A*QQu
z?q1_n8k8spOi{%>y6|j~Dcd&GUPO)Xw#Kj9im`wm7ogGokWLNn<{5ACF2Niu&-PHY
zTFvT2nk%>_1tl|rs0-h(<~;1jn8VPrh{aHk!p#ff?B9YFnHw$WJ5NlFJUTb`=|*#w
z!89()92j2o4JFv4rm59TU?OJ94aJloKs1HdtBI$4Ea0O9@I|CEc<zH;EJAUh$KHH^
z-fu5?X;2${Y=W0iunnF@|KWbLZ^)S^v!KNrX+PV?6cnt6Vuxv|J<xj@U?E^TP#RH*
z(P+2V66OA8<#brhr-Z(GL*o+AYcF@W7!1C8F@<7Yj8%wAdDL+2@%xLdN1u_nJ&9%*
zEi|Vqo8b=*j8pINH>zFBaH9D%q2&eN+6dtagUT`YVioEz0n0E37qiy%M6XR)P1azY
ztiFhT-s`3~+vLZAjbn0({=iFcwRevdD<%V3Nw)w_C8pw)6kn$Dd$@|%G8Q?bg}AKc
z^gb}Fq9M{H^#<FI9);OdCrwHMyw6oy7z6)-nMLZ7I9*W8Sa(Toq;nj92_BJ}u`++8
zzi;bcw;G?Dple)C7s9a~-U31`_RXF5%t^^d%yTxsajBt>``(oey>>55foq(sads@H
zPD=ol6E&XHuvkSjO)Piutaa?(tW<gM#P1rVd$W3V>f%y$lQB$pfK)!;PBJu}kL=p{
z&n?Abg%OrYslWS7Dxgb}rB{`seu2aFNZjdyCqL$IaM-7_zVhw2ab^Bef1E8#tdffy
zQqKO+0D><ugAw4_8YQiB&^h>A65M<adAAlg^wZ~-yg$de^Sn-NT}>oOeRQEli9~oV
z8}othu0FLUGLg9SL3FVkWiHYAbS?!L@$t?W_S#Z*@g*?lGFGAJibqzwCQYQ+wtm6^
zm5+#1R?k_({YcN6hs;j4O|ztR7uDRq0EB<F!_fk`k`|%QY5=Re$6R>vUg%BGL*T4@
zT3$j{sGGDUs>7PF21JcXJ27ibs#OSoF=?uPXUP1coW*L^Z*U7t<iM)^v2?kE_V~`+
zo@sk+Rh`uri-+_us~Um>^*(JnJ&ZM4VywT7GiEf)SNS};@198c$wc%8JE_LWSZ>AA
ziflhO&n%*~Ee|mqk#6+unQa&s_PDs9I-ysWXnHv=ZWfGoK&Hi%ZOR67qxv$vKeQMU
zjd*a1sAQA&o`3z($$k5UnoIvLOaPLO)XZs7a5_+?`ul?w-!QuaLJ~Qimx4o;u=Pz9
zo=f0dGEm8tta}O*ZdTv-vFrmgzjda9=KvY*f`TVBV6CZ@3_%xJBgUh(5+uPAsTLQ#
zOLkUb>&EfmaDdMycd-At=OPI%7kjhvKDGRmOjK~NLA4#?=A6nM&>jJ1w$jAf%$a4|
z<Eb~vG|kW;b?Ed3FC3_{Qd$PVWfz4pwG~kRrAIZcw+4n-jx-sb$A`sb-zQw>EhcSl
zkIBBu!-8U||K!$_c0BUSGjXwgC<k*O>Id41%EP5}&5cF7sm-)D?A!Q)_eXDQK0mL%
z0$a!>buk+)%&CohS5#5p-{h%QRBr4;4?6Tw{NQucxoLoDxB9l*?$l1VX?#|M^|!6+
zv8~pMGO-pq<tOOeP%s8{pT9yc?Sf08@PV2px6EX=BBE6x-hHNZT)t`4lv3$zeEaHd
zpzI}Jk%#x%^X1aO7@W`Ez-nc}Kd6^1-pJu^O0PA2`=Ed5&3;Db=BCxJ;?3jC=b|(b
zH8N<)nuzNvxz!9N%f5Um)?iG|h0b30($P*k-j>@?lw2u)d2sv&eIryRPuMZ-xb3;m
zE5g#QJ8J6eTvbZ`nZV|S{NhFvr28Cq-)!ped^>s~|Fik9_dCXhu2gEe>2tx#Ol3ys
za7Q-FcttMW4-BrWTx*aJNpv9Yl9~hbY~{9-gKP12JeMNHCn&Nnn6sX7Ae%Fi;zN^g
zN6ecr-R9fZ)cl_&mcM{iv%(d{b1@Dku|1io5%2dc%kmdqQ&@)B2H1zl>jHt=7?)CL
z|D&q$7DkqLA9b&pxb@|@vcB<opEdOI&H!?zxom#ymG4RNPix&}g?Aa|fhpSj1<g4I
z{mCwa-n$SrH0!Q3!nB5SIG~J|uxH<#J_M7_yB!98xC2IHwBSMduGp1JU<m`md;|O{
zcksNNTo=p7BYFwEYQCUfs&-G@fz4&BEx4yBTW9o%*B93Y2`+E%k*-;SJL|2u7d#cy
z<{rI17kpWT*Oh9)sr&t%Vi9Lz2}d~$X3n$nae^P3$hAL37lk^I`K&(2O-ylbZ@?m#
z$PPhKXRBuwiGHQJ9@5R`5<fLL1q_CILQ0C|?k?doIk+BkL}8%|u-O|IMmDW@mjbeH
z{i0EJKbJ2360os8rNyLQwBqBN<0C<wNHDci+ild*XyT^_R&!5o4|n%<7YOdCpRTmM
z#Kpr_6;`azPGT>C`_e~v1E_R~IQTk7*C4fpv9j`q#kv>7{rSk4x=s|x(8?6)%GlH2
zH{obD^I>kLW$n)ATUMT&jVnqcOG&Tqu%FHTux3?sr3=Bb`<BY|wZayD^epuYZuaN?
zO8i{Oj!Va~D}VgB2;H;zDKf?v>%9NcHYw)d#@YVkfVAUWI$wM^`)wd-`KQ=}S7n?g
zA<3$22Yian!6LedP)p|j54HkX(8aaUV-$Mpb$vfNxiBAOjd?ov-Spi0j!0}Io()n3
zpo=-@@=%V+^^v#E=`qbO-~qX;oBKlVgQd6TZX^~vntAzfD7|4Z_E>+emup0Pq`Z3e
zZA*ZNx3`UZ3JXDOx3Ze3C--f;zb*NtIfv-y{i&+HyhpX!c4ZtnDP^*nFeK|>lr;R>
zWOd)g4Gug+w0NSb@qVcWR{7DKx<f$uo7<WkId=?~><W$VV?DAYjOTGMN+#YXox$A7
z*)2+_{($V!d1-0*;Z^ui$;9<|18GY5?NU;J3Ev2^J}deOMY|6{Q9Uf;KOYpno&VjU
z{>|hixYGaNr`RzRrIe8GmaTLTEy3>k`zGma_48cDL%!M1`YK$Ufr6-y=!o_hgAv4Z
zE;RERb;NaDQNpX_cL1SK_zLVf-RX@?cU%M#Z<tF$8X+|3@D4DlszuVZ5atAP@RN1%
zUOVOSd6CfU(Fkk*BZJ6e@e0NPk;sJ=MZG^eijhtFew}}s)+WV0k2jJhgN$pXUhx_V
z+-b$u9(hb9xVLxhT0z>y-5Y&m*TO<rl`$*LU=l}bR2<Y$Ya)2+xbkzy=hJx^_x_;9
zJJ=D|aRLjvXk`Wr#0qT#P0_D!bk#3Ru-twz)LcNx`fGDS)puHt@L77JlrLf2I2j|d
zJyMwyJ%X!Uk9PgdpScJY0^jn9BN<d`9q|OA#;kx#AV+u!OXYtwC$7SyehE+?EYqFt
z|B=H}Us3-U<D2UoO!);>z93nb<WxO$SDadAo!zbYV?{d?eu`k-N>HMUKzA5Q^pBJv
ze|$<12A>SJY(86oPd|@(b-|5+bv5x66>gx$xYKgUv#g%2{_=UxI0d%`uu9{K<nCip
zPcp^o{epcnBsH_%)ghIX&moq7+iv;tLJ1`ki!9@G(Y>N7^?H85E>8X0{f0Q7@>a%2
zbrW6zq7t2^q9O>B>&C7<LUGh7vqhGF%S};sn`)zk=1ui`442$3j<>@O+ruq~1sEiw
ze{~EPCcXFfWha{t=E0~9cPcYgPk0@)Jdo8Ew_q;<`wi@7cYg+aDw5&Aeqt<$+J2@*
z(D&(bTK%M(ySwVhzQFhTkZj-2?5COwWI8Rj@a<qO%uQ{SSR`Vyrgvc}c*Td!Y4vyi
z8EscnsQ}SHpft)~j3&_XUC*4-ytxy@y&aKpBttt1-6vQN*)w26cxDYQhe#GovW2S(
zY;<1>WxJaAFo=m@ZHb``H1-jDJEg0QLdO5G`I}R|+oo3-dSz+viiT;9dW>1jw0DV7
zfJmTZk0$x&k9!S8Ck$WPxtTaLa0w`fH46z4Qel+8*?G9PA`c22PYRgr2Kw%D(*z>G
z9q!!D;<a_<$R9Vp1iGtd`i1VDn1uL4Ca3kt<WL7)K5Z1J)PyKI-bZQ6=^1_02|a1g
zeO3<-d;R)8qQcG%F-z_%ty+2sCu-EU6f;P9Zm1A@5R6ys$OJ|6-n=<il1q0L4c@r~
z5Z3gVuuH(Ct$k*4?btq3HPSJRFaWT*2{LdIzF<%3%sHe-(_IkX$48RNehTE$QQ<|%
z2IYVwZoz*Qibp1NDH!n<%l>aTwfq2#b#9$vdZI@rh??)nbm-ot4H8z(-VgArFx$-B
z{q|Pe$bi5|Kkn-8wv{&`=T&BpHOF3y^@_{3%l3<159bL&ikC8_Uvir0{a^ZvDY-1*
z_h+`GBSBrb`~AV!$E7(r&$Ubr7xpy687=|&dN!VK+ju>ch;_sGuH>*g&$+HgNH-jQ
zizJ16C2yJB&-!7YSSz>Xe-2SXO)aYx<zYt0da?y)R@xeC-NkKQHK)J6#Z;3R9(uSV
zyul$rB-zUX$xMur7FLqj-KqRg?-hoir8c17uV06scHhHgVO1+EsrI;`G0AG@197vG
zpARL(JFv2lX?!DCm)#?+SWqUAiJEEI+$W0`!TUD-cWh2PHyW7d@ctV=SbQ59Ub&NM
zCV$<8j16XY5K{-4ir1XKCQz<jm#5pot<cdx$Fy1|TONQOn!gJ85VJVv$x-LC%)mz9
zZ-o=ybQT~A#Y`<{-S1zdw*)<!&ij1bJ5yIiVpu@$MPY)<AX(mq@LRNQw>3SiD`QfT
zAL06t$~+c2A(nss(YHB+^di?y#3MnSY_c^C4j}PiSHf-tem-Y)YuuH88}KzZ35QGK
z6a-cYaGIY)eG#cM#puLG=u+p~{Un*!-x;!t%-^SLeGg;Zw93On(NgbHczw1w+Nd(R
zM$~K5EKV;v!Kmq@f~l#7kY8sRh^&oa0V>?X5!qV4ebe|`Mee@ruKOE?q8NL=D*nd~
z31qGQznBPU7VKS|^r80m3ee2UIkYf`Aek|pW?k-H*AgqK+^%DYp-`z!9>Q;IrY#g7
zmOkiSBE>QzaL!Z4lZS)P+G>N)a!sjW?``s)^U(Pp%Hu_p3qy}OF-I9UuLP7oHz%(p
zy-${M>mX!Q2a4RIJjT}`%#d;s(dCRQBP>Q2Pf`L%MH19}cN<Ala|7y-v6p|6_|2er
z%2lp}cC%f4_S~I4Zmit6qNj(8Sad<-yPZ;FKFuG@Fa7RtkCG{OHCb{#s+p)MvsBQD
zzVGD1(aS(PL`pWh4Z%O>zsp{^zuNiNT>T{jf62hVj0|9q$8gNgPnMX>gYICK=6Bjt
zHJfG>zeN`c>lr?F2X?3GsR3N?8Fa7GWpQVGuV&C(#mAaUO$($?P4Vxv0hRjw4?cN8
ze%Ub*oE57E8|&k%;7?Oie~m;eK#XM>%`JL-)unFfkDo<yg<D$$e+98uV%1@IuAUBG
zcga#I7_{^6UV9_3Fwv$ELP0iu#jj6AIrYyCe{D9H-t!0B*$n8lyxa{A<4lyvE?m>y
z=x6pL+oY;#l1?J7Y>K^5u}<qbR&#WE;`0`8<)*)v$CB_TgCD|q>PBJ*WIhAeG>~L>
za_NdTD%0CUE+*#IBLjUuZe)Wdz4~fD!>+BinsTqc4qB0E3TXwnvL&-$NGX#CkIRq6
zxA?gre!{e-R?MeU1~LaNDdl<P7P#lHmS$(6F+dRSpPL#3%h1Bd>9aLroLgsVp@mM7
z!@3Ot_f9_Y4(zM-TDC#-Nbh!RCj=~efe9{@o!b=9IauC{tCKrV1EMpvoIc$zF;+1R
zkFp=&%_Dx25G4nI9!yvZVDIBTDBjP<%$iKrXoDV}Ge->1MYd@2s5^cylQEhGH14o%
zT0%O+66WfA=vc<`T+Zy`ZlT!A*S<jXnuAd<c76=-65FsTd3KPEzkv)|$>NGzR6k(q
zjXqCE1@W1fElAP7t4aFh8s)@S7CJ~Ann1wq(twZm-Eq#LSJVo~cbRjIZ`pKndy@85
zzJfW#?#b5$Y&S7aUx?E2Kh^<e&jeSds4-VGI8kGHT}i?3is<x9#pVDv;j8R>9mrxO
zmv1nK#JCfD-caR{wxs!#t7~Ps4XgNwQ=^UJKGR3zXZHy`-!F%c+q|Z8$8X-*va74j
z^d<hJVG%kVBB{bKVa@3bu62ONAFk_8cu7^?T6{Q~+!89;x55X!_a`{4w8Ldx+_DbT
z4u^!BOzoDtf1*@Ax9pNtA0VREq^8WquIiC}(N{H?l@_Q8V$_@dO~1$~?03u3kTXMH
z>xyLG3jxAiW_0=NuIWl&)0?@CCTgEwgYkT(mZ_c{BKSx6c8uj#y5q9jge|Z-lePM#
zCY??noxZ5AaXvql;Cpv`g@TX1DwOPd_$WYDKA~!ld1sucf?6#5q|G_ZlH(*g-<RXp
z<7g84+e3xpF{DrO;#=_TgLgv{Smo=3bLC7up}zwsat0u6!DCB|6~Q;4_P@6czC=#x
zSG^m2w{yB*NKIjoo2<@tMRV0baT)MW{m0w1AmV7*-XsiXM^j~VQJqVkZEPL0YxB>a
z`gfHg*VvUamJf4jv*DPA;BQt^dxe;Q3+xEqpY1Yt2#(i=p>FD<7tRdIe`hb*XZ`@4
zxXan<4e4H^NYbI>O-hM-O=q9qCJL^#j;Hk~#CXBx+1oL@0Y*7}8T=S}aKh^3>@F&&
zTDrmJm(+O}Dc8Ea372=Dnywje3>U^bv@p(N?67g(?xmA8_X`E9DcQal)@8+D1xrC#
z4IkLv{Mk|y<8Q%+gRBf<WW}9<Z55s|)j)h~mIi_ZpAkd<-qbBvM4Ko+lF5_0LjT;=
zY?n<@ZUCZg`yBM>Bky-Ou=VG*tNfM&WO0vRFemh?^KqS%ub<M8nMyL_ud*-IZh==c
zx@3W)UVJ{2V`VH1QKsmG->H;{-=4ITf$va|(=wR8e`Sws)9lcA*SQ`Q6PWg-m+qc~
z{63vx5DDO}uNwH{TsP3Z74>Tql@nnf;Hq+6Gr+&y|7!QkzU<X2nuk|uivf`4gms$l
ztgr0J%b(U3S{|OlZpo$=&5cKV``UN^ou>T1>&L&VOT$$7>&LIvuO>_roE4}BdYsE+
zU1wYabsZ36%uBi-o#tM}60@?>TABNqAcAd%1#c4bWDwJb(S1MsPqzr;UtIG)Q-|Qx
z<g00{IDgFG%~pOU%wxrgG6(7ot{_K1qtJUuZAadg&}s3((!r4K`ReRu`>PhY4^YUx
zx=Al{+xnVNmD9(ho*`7?+$P8rb0&h*r>old^y5PO+y1n9{(jRI-cj&BcR?KK)puw&
z_9zarmFic#OM$Qy?txJqZ60yHaI4uCbQg8ILcr(Fk>0@vdb8Wf-Kj)N4q%Tf)|XwF
zG3^L0E(iVS;M#782MT=g@Z=S2ho9~uLIqUt2G2-?VHcDBI(ZK0lpY3)xkpJqHBvf$
zL6apyNXZ(8l8#6}T@0UT<7WFTX;&5}9m2j2OZSAns49N5MFHNENbGC1L6Yq<R$_Gf
zGt;7So%{{WO;TeDTQ&LoK%6n0mMUylZBjbdKc_6W?^0D_hWlsaN!vd-Iniu*Cus2{
z%iy6aZDEJJD=9(l`X52su!$pX@XARH&9C=p(Qm^R9lbhyF^NM%0SV-1Ve;YT)?Yp(
zl~`aMC-$z@WRyQQ@N=`9kBBj&pnn<kBcvb0ao+aFM#a?N8yAz=*Tw@q%QntPQwnj3
zx>>tgi-pOWyamkmr<N+j38Zwh?`d*E+iOT`k?!2{7~dXwt1Pw__1oIdD&DcvzyD3U
z`}$t?+s>mg0HlC*o~RX-T&agoo((~@`-(igS$?f(6sBfy(RUKlpC#nNciU}UL3%kW
zKU=ZJp0!!2Egx2bUhxHtmxcvtZpgo99Ov?BX8=37(>rF%&UnR>X-=x9<#_Uc*q%<-
z<VHUG^uv_tPq15T5(@?K9942dryY5&P9APtOfV`knv{L$3aO&yes?wDl?~foHlXZ=
z=`Gyifc4Tb745!9wKS#nY&O!3a1MNpqkD0ew3YE#akBCkra*Mx?9ubx=c|m5Ae0dp
zJ>8F;9mmS4Q-SwzMOe*$gWa3zc-Z}#RyvzB$QNhx>{HJt`0YP&*P#QHFgKQnwU&uk
zp;6F@$X*yBt`pi|s$8Vzo|nVJZXcZmieUewXXN@N@CXyW!fC#(iFj}Q=$u_nXw=kv
z<8c^WtN+-JLy>iTPXJDRr<#gQ>LgMVjxhQ;Ub;RV;Xc$hhIa%CuY#B^rmX~O(L4pN
z1)Cf~PN>m|2e%5+G&Kz4s!W+)-c_WPOz@x04;}+^N?S)?T$}cC4bZ6yTj4bHrJi}X
z(QNdM-9%D+)Z)tB9Y><YlFEdmy}5#3icgC-k_sHFHn#;$jJM=$(;bl_!`I`oTWIAy
zXIdqX{V+dD1J!4}xjC(-K9&E@tCRK6wq@7LwY~T0(^t=Q;wwSWBLHujOn57S9yd+Z
zDhTdYx)zELGgb<9Z+4y~kgr1U$uw1JNhLR9j&}8T)-CUt#=6%?T!}Q;m?u!unMg^<
zVBeBF`pL_kn{zi4Fe@*8aIjhL@jzCS?^dZ~Tyh~L=EMM3iZyZnP5Z83L;pv*^1}sH
z;Q`h<JoM+}GH<t67x#(ku2`J4rZqoxaQCw>lhCqfx3Rs+`LJ4iflh9eT!~Z;z`WC&
z+{4GeVimUKmYYJqzU<k!&r1|W6LkNa-U%{J$FwCmH_ZK}i`*t-mhyMeFg|f%IHx>F
z8~>`WSY&|Ukr=7Vv5d!QIz&FGNBoA$!Y(HXyCSW|wmIIP^{aL!$Ir!szjdUjN3y3+
zjyGAxSdENCgZFCUeQU@?kH2!BLlLytpt&ZEQ+T#6VG+W{@D}^?Jm9r#$=IV+Z!(a!
zv3pFd;~|LFxv6KOrpoC!`Dcss3k$V~X!buW2BXnHjW3PUZ9JIlzOEaYzqhrT!r0Qw
znjZ|6Di?-Y3+kRf#(e#{Cd%7g#{Nv7hy88I4h2J8P9R-T0H_Nb&#HV26R;BRYc_Qa
ze$b6-?HP7_*e_4)tCtre^x_p)%J-8F025}e@V3x4ZPC`7-_(@+YC`Vp_na!X^Z<wD
zEc^YfIgRp4x9Rx-B%vPM2!0mwQgVxY$7Q4tSK8v#8t1PYWkjLye=+x-QB7@q_h>kv
zLFq`95(E?l0qLCpN)wSLy+jXPq)7`kfP#P&0YwiWT@Wde-U(6#l-@x~0v2i_5F&(l
z7w37~^StjE_do8v?-=((J{Z|MYp=cbUUSXgoO4~*R1yfJ`Beyi6DraLOHy#SNV3Pm
zSvSr*iG}8uwN}lgRSo@-Vc+tUU2%Ce*MS7#pVtlp(h7htp7J#tJ%Q}W@8%v?AhgI-
zq|^X{Sf}Or*g|K9H~VisG`!L<%J@bVWvnk0efo@Q0%Y%R=jLw8oQQt=6osCY)~8>e
zwoWLFHIJ3bxW_x%ayLlDJQ*s&#Vz-Je~@fMu)^y`$$2wpIr>kzWnK8v;HUCV!l>Bl
z+$IAg;=$A4-;i&Kxd;8|64fg!;SVT%IYM>kormYpUNQTv7q9))sMhG|-0z)PNHLa{
zrX5xRAjKS)JZBkuVq`*5)DSnrm<W3qmM}Vwnjm0bz}}h~FS_w+cH+v_zgDRmee{JU
z_VbvA&v~k&Ci;Iv!ZHuNy1JCdBl;A8oN{&==9adMD6)+RY+w}&jtuA>v0d;RYp<zP
z!R{nF;bD<-;t_P_%{h>Gw%a~$zC@RLk!V(<rup7d@5asFR_?kC#HQekJ6$3D$I0S%
z&j$?82{kHg@h`p7{3;XtY9pygVLa!>5b!+TAuUm&96+tK16(m_KU^D+>XpO|Lw?Nd
zO$~ag9*tRztOb>^|JjDu`v^_l(ST)RTpB^XuQY1uG)4pCvL(N(Pm)|2w7<D%;E|MW
zu=erV2M5;x;lP3S@T3rX9|VtTkYpF;kI&_EHo8H`v5M9+S1wS$7nX5MJIzGR{UdcM
zQe{pIiy_@dgGP)1-#Cd04&t?duZ@FLLphgA<<2iCzvLX+?TP<(!JaSmeYkW{%6-Us
zSr`0)p~QhH$;^w$F*bB{|8k|s&)%E5YXfiM_6E`49KL}`+B!TF^=n1)81J!XyPT61
zS`>B}|2etT^J2xSe55}8Bl^c_8bVbccmvqyM>S4GJS6)8UC#&X(h;pbd1LQBMkzw^
zH>4jy2s(c3O!24c8c*fF{@B9t`)T*OLxv+YlBGU=r;G1&xcscehDmko&-Il^X}zD9
zAD)4P6$U~!p?+x^aCrM5oQzMWkYvL%Kx0HeWA^0}Amr*J*`?i1fR2E&5ayCaLF$8H
zkc&r8%w4AbIk-8F8U|1yox&EL>3wq4PgEzy%q@>;P_pI(DglXHyAjER=#KgJ7!1?j
zkh8~z1(EQp{8xmykzaHh1^+^nL|nyx{R`W6_M6_+RvLCJK#1m#k23<wX^Qu97$O{p
z3Y1ngkQM^28yYNS>~{zB?mnd5Lc9z3WsP{}|NjOV{+}F6^~=(99@X~=v6J%6<WBt(
zVV4smD~><`<}<gqSgG(IY{K=D@`LR5KALJnIBk+%lx;S(G{h5LqXkR*Mlv5AdEWxs
zz?j@)pur-bKegGKwM+%jJ>9w)EFn?sDD-jDDotM`egB;-)nUwWLosSZYnTG$hosT}
z`j6?gEkH|v|9=C9-pzmf!Q@ke4cpGnv32f&;~fc&nL{i_+@qDmU9t2%m%a7cTtHX1
zL>;ZTh{o+%p)`#|uAr58?(aTs)O!w1l24a@-VAg64Y~UibIb<zZn;PZrRW<aqBdEA
zfG{^R5-u;+NCJM&i=mxxBI}di5Uoo7mXs(Ueza^A`VF~v8MP-0IpziA2K6#(PNl24
z8^#r;DdXPYLRZRx<z^Ke_-;%eE%QZ>Eb2X*)pgiHb{v(^$?_N9Qk^YH#a*sK1AHlR
zTx^ouI=y(u7@~ZVe+XBLP(V7`R*@>Oe(Em*Ildm+zO{CgsJ^jned5}N;+3E@EcPOs
zRV$y}+a-TAM~zF24Rl^v#JUSzu6hzP8m!oVwJR!pVvi=J2^7D;OCvFYdup@&3bJJD
zy!|rg`J|qSTU8vE3HsclQQT00l>CpX(PDMS>0xT+sLt8w<J$S6>YDoJ!?LN*?<~gL
zal#4bMEZSDqpHlp>TX@cCh_zsPrb;Ward2dKQOgUsqOu#Qv1OC)w6-qdOwoI3Q1=O
zOAZ7~vZV%rAvuQ?%ZYTU)k*CANXM+{@=+l4&q(!9QqxMs{b4LBfg@Z3AKaXHz>>;x
zceFKepd&GnSre^woT{o;CfUO|Aw0hASW{Q?%;0J&@C|e(honF1*i(PeK%fJPncFEi
zPY;7p|E)_ZVmzhdmRnM)$-VNTR?n6n>i8yXpj>y7Lv~WaoF!tgJbY@`=W9m>-IDk_
z+peE>8G5KDmApr_=gj^SnuKys;jlv1SX7T)m3>w2>E%E>g%wHDVtyeE&ZMa()_j~3
zA-|E!tZErtfQ%jbKwGD9ZSa2LtGR)*&#b<wJSn*e`Q_K!&NVH&!|B1%&}_JA_GTjM
zWxx1`KQ(E5G-;T1Y0eW-GlLqFp5)N7^_Lu+%rloVr+VGZ?(Uwq`s<EsdZ!ewSZQ5`
zDFNla!!b`vF@2;zUW<Hi?{z`ENc`$E(d@8Kyno!58E*^G-TXl)u_46+zmE>#)QbcI
zhWyrqP;DoOSOL59pR*!JfoNrz4dj<fcMivNZAGYYHD*jQzd6>$(nz==A}=wT$yie$
zSv&OQrt8WVhowH{UbJ7yf@X<siCk&sof-OytGC#w`+u3l_uZCBOdOp?-8bS$a@cgY
z2~1&#dsZ&>`P|19U1~olB)7w|Bw`44zAh!M9RO7Xv(;>$s(kU!D_hipTb4IO2lV>~
zqZvc_FH<j-R?`=TGlV(5qf9JHrkgZ|OB#(XS4o7u`Wmwu;4uG6k%sa|n<MZ)^p}yf
zCj$({5Y;uw*g05jkK9`))0XHLqWa%Nl!jL{V0`NipqJ;5o{DW<Iuv`$+b76nwRLa$
z*|ms*_LSp<JvGyD3xKYJ?)jcd*W_y+6lG%&@a)kER^$=^PQUhwx85R0a@o5t76g-W
zy?nFjv5}Vw+Qr(AOIuwI($#Wu3$!2Dy~3q&^PO-;yiU0GQ=z(>D+CX7((~rY_7+!n
z_Q@w18yDX?>*y|AyOZ|tY>wYj2^-bc$pVvon5;DpvN))aFAc}ByJ$(-!7`-K0%ust
zInl&4NWV~;ul*@!Qi+8<b(eDcdjfa5QBCG%H;=pTd5-y<*<QfKW=ek#sg+SR{SnkJ
zDrI32^}$q8AklIRUW18|n<B{(1Z9ft3fHF|+RhX`eLpM^tKIo1S>YP@kL*Oe?<d0}
zQ{+l<VC<llUcAbk`Hy-qgiI#U`_XR*_YT<-f7FvX)lkwQ83w1eyH<aZHDG2N0=;PT
zQFC=4qzXuSG66!yQ-Um=eNBGnS-7(080wE!u}VybBuqng@I}_0<QvsV(~h@qINu)e
zT}pXdX@i)1*>`H*IU~9(2Rgf|$p6=_5*h_4yo>i?tYLDffO4Dp524OJ7XlkT>QnJV
zoz)kjPA-mAXIH-T?hHdQf}WRm9z(L6g;fnDy>&-0$z2KH(f3UINgjm0l7DHOP=u?L
zrX`ClmgSP7u#%)3WbU$!f(1Ld#M9<+^&{W%XpRjb{<4<c82m(p3ZJJejC?x%Hze7z
z4k?4vCYItQZ3*@H#kTeNZ*EO}roUyOgg6Mb&3*xEvrEiv274SVhz()fJcW#>)fth=
zUApz5Wz)%5IJgK+8A^qz?^RyE>!VIjq)$5)Hio(n=L@VEfgZPj3BONlmMlZ4+%+aK
zl_M8-4IZy6J7GWPCw)w|J`*RF_>=QsFf}d0Y*?0x5I<UX9R$|QFCF2RRuG<){!EGd
z!Y~odce9_}4lL6ay3>-%M_*|0@M&B;dDi$JpIbrUCmevAGFO}$<Y_vAAnX@=!yRaT
zJMs-YuT9T!xy5wq5V3|fer7ZO_8RmV<H`pSz`4aZ*dN_;SC||A@cI6{L4DfSm0M<9
zIBw5)!5U_j<v&no_d&J)=k5`wIhcVx4=B{2-xpyum(s%v$$X<*`v%tHu7Mc=$(pe<
z+t#;LoS1{rW|f>cqa-g?>Sa1$Uz=0Os`HO!fBrh*lMi>~YSfQZuoc91S?_O%2Lq4E
zl=?ZnX7i$HZ;xR1Po=Ee&WWc_RZT=Sx^j2)e?Nmh1FG(C2utK&-(e&MNgzW+zoW1s
z>k+(UwJ}vqU$6S{FGBG`W-sWIKk5-L>i0bufqt_j%9jFy@1*NabQ>tj@Of=Y4Bopd
z@O)c&#>;r#YFTN=hfgwf?UMUEu6$T^g+c|}I9fMwecg1qPxztswnnuH>qP922P@~!
zRdK(V_oIr|VaC(0g6x@5H*`E|r8Er8@ai^D3N>)v^oh-pZ3#ICw@94%q^T~$M@QGq
z|MIlpepTxW-%(KD5{~%3>=eENl(S3PVc_o>3~-oVyCZP?c?3lXVM|sfTp*yL{gZ+#
zdTJ0F7u~LV(LU7V-ualvnEcO(G~iGUk&WR+t)fXSlunRwp$u8s*h^Xd4I!tYp}!$B
zg5$p-H!32$EPud00O!a>2PR<p(JrCH+dwwf>B<H@G>X@;T-2`ojxj^ZI61Wl|1kfl
zoqcq|;w4X=l{Jp$$HZs(A`g6f*Vry^bw-EZHiSfgH-MKreDB<eDDp8dN5kv)fWKe{
z$fC4RNfGv=8nCyy^C3pZG@B^=1$b}DR+|VVIuK4`&cz{K`Wr*n%`4wH28q1%2=$bG
zH8itkp?FAbQTB65Z(>I_{3eiB-D=F3+lI303O~3L&8RE%15q+}XUDJDdRRiW|DNsx
z|9)dbhz)v>2SYl{Iwn#A4qqcOI8lUsg`+L<_di6B@1O^okMI1imn4}6nlt7p>VE>%
zj#5r|QxD?xUtyMDLy>T^7IP*b@A*h=3J~W>S6LqY%NH83mqoa!Ut^~KwqZyB3JmMj
zQj!X#OCEKoCic7%y#gn{fJ0F;F2lbex^Ey6&;J!oHqWiLOQQO(5K@k#B|xf)nKATh
zPY?_|>Z@p^E~V2LI1|jfrvE4MlFf1e!HscaVk3ED`AZ7im}joL-W}v{dHUI%$L(TG
zYI0*=ogw}@ZYd$moN&%)B!KkbQe8V^RDNRbXmz2Y+U<yKshuqinAs@eBH8vhU)4H4
zOmlKC{4}5X;-Jy<c_w8cF}=ab{*S%kDvoXLVP{ALtTzYwMiqGc;_g*vF8;a*sw2NV
z&k1{83f-o;z=f+k!Eprz?b@WmBW+}j9$~xgjhnGMaY{LSaPD%3keHH=s9koVQ}Qd^
zQiUh^G&muGGfb-%e-M{fok_QvKA1uCGm}Ga?iTKKM~o0}08}jw`jmVR?;0=Oei@m+
zbbW$f<+ToWoipKds)LfSiw=EWvL<f;ACufn#K7XY>19%xcd6YE&U&JJ;Ll&<Ml4Bp
zn%RvN9EEeb<Z1{q#6&`1I_|Td!KlFaRyhbnc`)hx^s}o}lh*2LfRJJzpVEds3>4fP
zgb5|d>6!$34$~fAy4HsuxZ?|)O#2dFcO<KX95FM|lWLZCSUwPBy4Pyew9~?MLgR8&
z0t1k%sjscrftL!cah#?qUO69jJKdyZ%-gfPqI^~VVvZ0OJ?ZS}O{Z-BoELsrku6Rf
zgJkdFVpd^KiLJHOmqE9u!s;Fh89j6yA2$+lbpgPB*4MklLR#t_`|co^Frc^A;9ihH
zORnL*otGR$o%;avb{HvXd6;z$dDP7^n1BD44bhstu))>B;?IHWmjzatJti<tl5`#!
z2o}jYmE%i4r+UTJR@}V=rII$k-(i>;KKn&fk|$zJU8a^0h;}dPDW12Qx9g`dX}zS=
zDP@=Ov0ric_WokTQ0{qEZQL-&qFat&?B|YJk5|P}^v?3t(X`)`7SyCbb^*HA;VeNu
z7u#kthtr&B?P+@U74(Qeb~HDDx|-OwETR&=x1ofzYdI;$An|jo$A7wh7qsK6PR6CP
z)OT}OQuw5XO%Nw9p3Frd5KAK4=0zOX)tJaYI^UfapPktKK%W<__5tm4oJ1C@v_dXf
zP3|*DY>$2L|5~@}CG?*sO4jJWscGNcz|>AB*;=(KKT!63lbhLj*#Rb%s@ZX>S~?sr
zHi2;&PHf&uA?uccD{8297pzrCwX;RPtQ9L@6C3ht1zSWwE0K_2ws-Jw&=ehP?}5g|
z6c&1UMSONzX!JR=ZlkazCZdm@Yie(YbtAL5@XePwt1Qh2)eAa~UoHfA=JB5E7y>Ps
zYBIDW<`HTSle#k>oOgbcr6i?+|I3>5y?AQl(Q^2ykZ+swgeP{Sn6mrLmT`Ht((|1X
z1&s}6u))SFTCLwI{Zg0EcbQP{=CX~{vx3;J`=@EG7f!qoFwcx7qN-=@?AX7o>5O>}
z$wXzlM8aj!s_a+~mHw?tSq5|Gm``4(3g6T4@@sq`i~T^%^mS#nr;Eg+1Jsn)hvovJ
zxiYab+<~TRk<c7!NyR?74hPD3j$&wk&ZKOk_-uc&kaugrw@6#nV~=^b3p!zLxcgqk
zJTjWcyVffBp)g*eW8f%uti8uC`v6GCeNT${l_)Gnu3s66uDjVQ7nT(qTx@UK#lPHA
zg#7D{6JA<xl>pE#SX7;d#?&R<?IK2%%zh49-Jbs%>}Kob<D2cMC`<LLMd7ikJ`I{T
zld-%q(Il!iJz6>W#U5gFAIyl%+W1hctl1CCYoE(z+u9QSsCaH+xh)~`v^=lqvza0o
zd|4^u5vuOIy|T|uyWbr@e_c4PQa0DlNOqbGGr}+R%uO{oJ~&w9kJ^?wBURD)%WFmN
zoBJ=RoRDP2Rq9plunSnd*Qt99R?^ulL?va4`96BnFYkV{<fFR3>OAP43b(dcMBdKg
z?yMXO%<;`pV#kev-&1#Wgb{CE5ux5}>exG)+=A1oOW$t>fd(mT$fs6S^)HP_CYKI`
zdux(np0f-s`0r@&@7j6y8eTfEBO$PA3OA!FC0*;p*lub_XhNygNa*w|Viw_7py-@2
z4PJc^O#l4o!ru^A@3@eZZUofcVZ-=3-`4BVhTsc47KrE9R6@fy1HhY{QQ<LJyPwiI
z*@&a40!HfW;q{fi=++xc)eTQ&3yT&i-@ILvo2;mbtTy0ce#rKdX?qFs)3?tgxAc1+
z(<c;3TU8qXEC$4H2m~CB_4uB9iUjL^LNp@WetDekllkHp-(FuIb~Wurg_r)^ZOv}e
zdqyJEkX6_9!^_mWFbWIMU5-F40i;lZlQNGiiN+%VYSb&nDc04Q1GZ(e5(UZk>OMVq
zOU=G6Q;YZU9rktTvqSYPu@5h&Jh2&$^AyK5HrC!?>ATRm9n5jf=)GiiP_N>K24508
zewr-1)XlW>15N^y?5yR)5_OnI5N_#7eaO%o8so*_6q<IEU-s!RMZTJx=TuA6=@Apk
z@^31`gfNfvYBRQx#R)68IOqB+gUrg=A&eeVfQ0*GMA8ns4z5@nH2LI;&mIjE{l~m(
z;g53ot0pUo>RPokbrUpWjdxWlbsv;n>3c9;1m=1qddqX=*b=04<k@o!*!ExD=i2_7
zd?AB{cFpudNIe~zS_GrvaiTH)gdl03=!F0C7Y?suIitxpFx9z__p095)fbj5icd~f
zL&kJ@f=L&3g(>^eP$AW1M)h}S>Eryc&~W^2hyFYFYuClMKF?YAjz6r*(0$mA;-uA4
z+{l^9V|t8PVJCo)u{s2qI-M-G$GKde8k!#_j{o}loo!v+CP$vZIow0m?&aki9}jSe
zBzJTiY-tRGKL_Ga(eER8kqJf|Bp#h<mj+vR?>tu3oAc4x&wRoaL8{P;UxS1`%FW-9
zr<hG4$zAm5JMtO{2Z%3l%ZMy50nx{VC_({*y7~eMr+nbZJ-q0#`w*i?0r<3E6puQN
z7KiWM9DyPbnZF?%nAK8bAEiqn_fTzPuK*)Q5l8i#0@xGl4c`OCdXaxm4@bP6gOVd_
zcMgduKvT{B%kQ7RV517bTm?u8x9)u~|7&_ZJi-cS3;HIOb{cl`Xv_9ZF`^s$rvK~d
zk8sg-TW`{@r(pY-b<zJskdWisTX$5!9T#9QptKbMLedcNuU|j_y#BYUy`?2SmvR&2
z#%#*C{)QBeHQijNfE&|qh7UT9eE{HKh%xks;X?!`X7xF#kMe;Nb$F$$qS63+Ku*i$
zI$0X*{Dx@Dfl`O%KOsarwSR$-cc>1PKnFuRmuf{KTg^p!Dgue)8WY=_JHpm0nNzc0
zbxiXDjC~$5K{Kn46p)-f97~-3o?2kiE8i?{ops6Fo?S7fzg2iDvVV~ARCpI~De)1e
zJJiLz6BaKt)#Hh!0Ukj<)6Abe?(Yf-`LAh0j`%$siOygm7gv}3asD;sQK{9;Owo?)
zbl}d}$dTsDJdq1Q*Oo&wk6aQUBu*T>Yl)+zhl$mY>4}_eS!m0wSAOQy<X4@fh3eM>
zv_ninZ!6P@f8s?F#3RIiWLD<q-<Ox?VEMFb5aczIlEbM@*T~R!RgoK8OA(`tL`aXd
zU+@TVcsvy-e*!%PW$FHgo;8fCuGEp&lvZC4N5Xhp&9Iobj2wP~$gJ1a4laE#O|jdM
zy2$eq5_au^-iLjtDzk=aE)cwQEpdu;@0=Tifu_BZsN;G17O0qlQ*miJ&KI7ujQmti
z*n~T$P$P5tu&>)o-+m8?Ns1|A{Z;*?V<Zv4&z8sqzaL}eVa5PryB3pvL+r<8b!)1t
z#&<5(B+JJ*yZl_p2ahxh4X$3}8@HMQNWN(3QI{Ybx~U7xH&`4PtLO^*(q?jKi{~ib
z)ev~J>Gy$#y02X<Or7vBOpkCimE>*F7>hH``Fh2(!okTunEQFGv96Zlmi^9V{xX`?
z9f-4*%qI41kLbia^jTJ0qVfmfXE}X%n)zRgV!Bi2ID6-UlIuRhwR6UXItR85tZ>(^
z7X5l~r^<!9Ur2qnuPqQ&^RgGkh8*wRb&I$d>ha3C-n}qG@v-4K5k&Jf&0&-#DH7@3
zE%_uzvVkzvjKL?}-z{4l6xH{37k)eN$y(2uhM#uFA^%92q|j|+*-f`)W<HAEyju7k
zNe4P<?aN3cHN=QE#J@b3FJhEmp`MUj<yrLE)!ETAK$zt^G2nS*qUFU`PU(phxh2YH
z3>Vn|hvG!I<y`deczz^guPqDuM>;Vt@$v^K^3ZWHQa$~IecyfBjbI3KFWG3AswO#F
zy}<ny6&k8^F{$aJ=sK-XPC;7@5stTvt3{n(KkTh4Ty!T28F^?-`^sMZDVzHKY&n$>
zZE&x$<pnHk&aBVOL!qL0(D|P0w`;L8w&giC)RA=W2VrJ?O3RGUvWQO@uCVJk49j{Z
zEIy{KtIy5*UQ;}cvsUv3t-QOBTA7hTf8)t?Gm;0+P{E|JI=Ab5*C6$Cm*^pVleWnT
z=B|bF2bh_I5A7F5!}KfGn}<?^OJ?ix>aGW|#J?Ua^Qnl|psticn44%o-Zhs?`WD-N
z;5xM$Gpe~SkWgp5WJI(Tdm5Sb$)q;RFLBuwXl|Oq@f*DsLFw~`E8*@%F#}S`1Nv;K
z-v?-g7I*=DSOw_AFQ-}!!YkZ^Lc%!3=uPO}Cf@!O)8F|X3cL30O>jj1(M=@9V|J~s
zrKUb+;rU~mLfUCn3BCTlQnV{j%t?B*OO1xzSrksD`)e=m)R+M?3pZQY>&H|nz<oIu
zsUF35865*&)b5gT(sl;I3W3)^twPZD#@)yqwsUE|<2fF4lt04mO|_aY?{FhX$y^d%
zVQ@%`<x}Qg1Ev-;8Z`95x?lSU{<vpMd1BDMInSY*WdzHQ3{iAYXz{|lR{6YUQ2Yx`
zkrH9cUQ0x6Znu(zFcSUXp3a{Za<y9TWJX*?s29>5fUQ|N%y>#&@JGl3|Al0ZwsC$z
zi9e}mN+rlwG_IX}H7xY@UPrGU^y9fOHA13ktLKVm&_gM+r)Ts&?p^Rfojv4ShDNFC
zd*D;Bd~b*^eO>)3tlQe2D~PSBrmWq}F1!17p~DH=UJOb!vbn%al4$amQB1(ton@TS
z^OWP$?N3M|%Lvx%c(J}fg^)YMhRjy4dLKI`30*<y*0Q^3HJQZeiOCpTn=e^r`HP2b
z0(*`DoxTLeCC~hXN)3^*NVrN?hx#=O(J(QzlMf1uko*lfhp?x7r0`CX)xb*J`25;q
zqGV8*OM=ow;TCOKTFHwwBOAQJagn;9ye={38I+QZ!|kE{NZy1eHk0jVkm)_jbrp5J
zrG+xRZz68ITDh5FGS1WMJHa?Xwc~HuDMU;mC7wprme+t>RuF#|6Iv1-A+^B?AX42D
z*E&Z=&HCJ}&-ay!{1mo{EtkJTv!x3w7YkJ(p_A<Rc<{4As1cn(@YjVW02D<GCJ^%p
zOdaB-9?ab@i4ESaS5tjz*l+k8p`OG;`cxAM*$^fpwCZu5zu|QNZpqrk--%m(;et1Q
zoIU<_M%Jq%<n=q`&R82GsJ4y*(=SsGm={Gi0F;r;`VEOk*TY@ZsLk*@ouOj(_lIm%
zN?p^IuV3zZ{J?WT`BC5X&mna_7#>yopk>sP^TiFtJ-iK7n+cLK+l4Bp=yk(f6?YOK
z+uzZT;VVTEK+5x<O+Qe{O|%)H45cCmRBZ26=!?uWz#3t}Mhf0&?)h;M5RiT_AO@Bs
zBzLB`r9D>ATOFWa-`U<uQoy#LdBPM%_qVjTR$7`Ytu5fq*1}F#)npnJx3Z0)?k7;k
zLGs8N>1RO@>7qmWt6V{s)J`}yr*T>O{rXes&I^-4-e$ZK+6{~<89;VY6Hv!z!GeN9
zSA2rYcXl3Ssu82`U0ux%6-=q&-uF4I8mQl=-9E$Xnz{Kl?~nfv3C4|G1a_MILh=M{
zxa<NJt1aTW<yTA(`j@DK6yPC0f1{*?Kvut>gye&d*glU2$RFc)4F(UmTi*pS5+48h
zRivclTFeqT+>mmFQa(0957PY`3FuG79`Vj{%|7DY|N5qb&^3Dl&)*Q36xd+$wf28h
zu{B_&U2il?ZFxScf1Q!<k~QStbmivyY&4e<30GFVj!)@RmZ<W*yU0Dc-F-`I*xhZn
zPjGeRk$_N6<B)F1FnC%#aj3&{WZmD8=zcIZGLJ>TEtH3^*dmuGkAFj00Kr;~qFhQL
zSx^)}`YF5{MY@`MGK|MUz=UN>Ved-<y=F)dp`878a;sZOLipTXP87?Y%KR~t=~%h{
z{6}w>oWsc#8l9ESw?1^Nmb|Xq{pc3A&k*+hFz~udGsD4-XaFYuUR+uKFV*Y`s$W(1
zeI0;A7Fz^OdH)=K$nLYh5XJ-!Yg>UGodL)--x&IR?LTzLK<R%v%#GxZlOkO24w7Fy
zKZ&mJmb};!9V(_Tl2jOURo#i^4`t?H*sXITlux;gWNT-Fqv7PfA+^c4|FPPoVCVJ*
zq=I)m&N!-L;o+BNs^JI6Zvc8^BA`}s6Q{A>@vaT{DRii3&|9M*dZI;|5s$8L;>NCk
z0589;u?PA%g90O=Z<F#*qs!csle6x*(iG@RRxC2~jJ%6sq;kw@{ISH%eRK(cNNOy)
z%ZR(A8^x**jSYA@7@GBK`yPNa`i1@{NbMO*HovZ;{lRA8vJsm*g7xBLKsb}h_=Acq
ze)-V(phtdijr@);Bj5N`_SuQ2zy!KaU|K0jx<D9e6~edI1n|C08%VdY)|d6F?0Npg
zTmY{-f)8f@0hUpSs8+*5(&3TCl)vq<^pFbYqkv7OGeJRGf5bv^m{*B_1j4`~M$}cG
z^xAZvzx{?W*U=l`8rwJJrl)9o{F+A6Zc9hBTINT50@1;(COA0fl3Tk-Q^Fy<TJDn>
zt*J<2*<~TxJm#RFy*#F3H}n(oy`$?$@HSu#H+Ii_t{k3br|L8Jrd*nK^7iPia+-dx
z{;;VJq=P^A=o;CcppZm10al31IVvL=E<eShSZ|DO`D?w>hCKR8om0rS+W{SR5F%OP
zlHm63b>3$;EKgst3a+m={d_KQl|8_e7StSg`Nif?PVhgFFSQ8C73CY5YG1RQiKlB2
z^_?RvAIEflLqy-xL^7}Lty<Dy;d7E$n!_)f&4F(iU%QN@GL#3t|0rnE!Sk9<YG0(?
z2eKT}x7kGq?aomk@R#J13aq^817nFi`K;YP>PFA9YfwQW{3h=1f9!yc?pb!)3~#{*
zr+3VobB!jj<6FYdUR9V*anl&g21&E_P2`lk|KZRNuG5A9OO{h}y-84gzgSlb^GI0x
z(<+<c!%gPME4<N+(n>=QuTkw8T1>2jCdsdJRejYzC;DyTQ3A=){J1be(Eh%{x!Lt+
z);1s43hNE@>~6F@m$~4WV(KvqB6QGC9Rvvpq=f129rn!(X1c}_tFcRc0=-(1QoWkJ
z!D`43=%;B6y{eyUxLSD*%XAUaE+1^$n#u#^wlKf&Bp-UZ?TCoL!{YthsCz$`xhvSX
zPgI<GQ|-_@($Z-8%>$xE#$`LA-vlS;AL)c?H?}us&RaxJK_8BGWcXBf7vb}&SvB^B
zGl`hlY3x!Q;OR*ft+z%N7u=bjHWuHpvYsegqcuLQe!CG=7<c)Wx+Hn$?FPLl%+s@2
zKOWhTu<8o}IMIQkYkLCEJifzbYdZLa$H}J%XZb7*8pq#Ed7i2$3W~cI=$z~<Tt*-B
zsjQ#BZml7=7SSDF`7MLbg$`Mi`ualh0Eg8F)yr3lo+&@n-@nv`S!U|_g6gKw5nk-X
z(^yXDz#{GYMBjD0Sm@h2p5o&A*&Ic-2c+8CMs945QbS_~w<-+#2cbzv>c`LKG)6^d
zQ?Afx%`L=jpvq&IZj~tp5jx%HJ$thocD$ZxKa5o}diizxBJD?5E7!2yd4dYi1Y&PZ
zvkvmt(VX43{xf-iwJI8EzEHI6&n<kI?y!WH!Ev-;B_YUp69Sv@5(4>fW>NSZ4Z_Wg
zw~{5g@-+PoFX}=@#eN9b^o=BlBgT=^g?EHQR<FhWdBfe2%A978TgwMr8bC@SgeMLJ
zrY<9{^-QwGLFSVuo;hom%vr+of*<O!PwG&el6nTE@qMNbIb>~0i9a--YQgVz+c;!L
z?MQ4o$Fhr1y()#c)`i(iZm|%=R|D+>FHgo_;OPB>bH@1$`VHL0J1%p$n~CneBvy#6
zp`P;($=U)M=`x0{hmjDW5{Q#y;_5S|<dm*lo@ib21&Yuzar_eeNIaYqpS5xjUvQkG
ziYT&oPn%m0sZF_9Q8Bz-beElzz9OmUx#Exte{6eBr3cn1t{9&i6;SbTjwUBmEHF)R
zRai_xGutmQ6GA4#8N(hZsxuNqVEc4`Pd)|h{9&quYsat1CV581{odvGUm6B@y2H0u
zMQIf2d3daT8AkffD}D}UInIx`#E)9JU__>B0P~aBI%;;XPq`$e9oGMLqw^A`t~FFQ
zQ;zN%1Z>@rOx69W1DAM^$CQ9su>%st$2+X3<w`L9wRauA2oo<yTCNn0Y!Dw>+btsc
z1E1ZL66|$M(%D}xxr|iZ`y49nyTqkx?zfD7a()WqNN;98S3sL)_Va`q?Bml|0^Fdq
zOeSR04IP!TMMsdyCA}4b5pmwXA>Edfn40NiIT$QV<*C`!nB*3%1pk<|AkU2+t!HZ?
zS1!bA-~>`5eIRPmt@FgV5jja#@}<eWV~#JCXg{{Wv*F)pS6*G-_F;sYjkR5BG)kge
zZNnRNd0N_xjmftq3j1=OHs_v^*bM6RO9ptZ6NRrJ`vM)9cRU<#4L<?sHx|HZLhAUr
z=bqG?PXVvC)fcC%jt^dbmBD%29~phwnCAvbt(5aHtm`rT2_|5`0Bqo}?@M@OhnTI0
zab30OvQMS-e8sIXwU)fD;Ti^zQ+>f0I<SPVR0UTEehP)R1V6=s@QWBe<j7yaJkC7*
zSNVLF$M-2~;h2;w?XCgPk*#~FGX#V%pfIkVl&JHg;eX#YT)>X<QZK{9C8F=Rs<Y>*
zcOEat!sD9kioE!5NXL9P+U|$o&+ut{fE@mjK;ogC=fNWWvza_-Btf^=^Zqn3XjB#9
z>gmbY@NVdV!SQO$#K3vuBEMYWXC;wVu7OIB!_)tHv>4E`%7cowSKGH<3|GLs(d>>U
zAtKk?)1!QSyrNJ#=?5K?g5SG(aQHGWF32odLp`#f{WOpX#FAcuV4So6)=50*+IDVa
zvNtD^?#8s&ax&f1d0*kw?%N%+0yx#E)NEdy8ZavT*w}rEOCi4i^C@aZ2k{M2X#US!
zW*#H|zw(sRU)Lg1$uNL~%mJ!DqOn1Jf04)7-p)o^!Dsalb6M&J=NZnKOg@kk>qp<s
z>76Zx=&cRBzx6_(Ua8kNb?d+$V1^2?=J^IB5PKrX^xu$Eejpw~7S-8Az#N}x2e;t^
z*t~*l81dW%q@H}mki~LO)UsJvBQiUHj@_gGe()vIzaK1jXmaAme67w$(#2JHTQ*c7
z=<gH%)S)VAtwjnZJ(TXrrs}f`P%GQJZq#seF{ZA{^D9?|+DxwxWcY`f%S&0v|1Y4n
z@}*fpP-A?Qka0K_N;m)10Ax*N1)sm-=o_nR7R}TskPsaS*9zk#6n=0J#P7G~`=Z^O
z?FS(>Hj4{aC3#=bL~<n!4%16BO)x8<R$NG&lqWDI)N-#RMHC|snEWwsl6CtjFu?n|
zmPk@he?t-qDIDHWsPDe`0}4GT?qkb#0nd1R2>|r0WyB$GnA&g1BQ6XHWOw|}XN&mP
zvmyTHv%UPMv+a7PAh^h49?J|yH%9{u#9d856`}S<&1)yC7t<ap)vNTUmjgryas+lZ
z$bVyOm#8X=n@K$@id#5$KJPRwHh=R$sYA#a@@48ho4*~Iq$6A|7+gHxQLOu(u&%@B
zAptRuu8KbOWP39-($gQ{nBKz0dzUHX<+4cc@OfZL@%pQP^wc!P<UNK}$GoB-uGgHq
z>q?Sk?n^EukmfgiqEKVA1i69gs>lr^QcR^gh{55=x_2+&Vv(<|2U|;Py;xi34afB4
zM-bmWzosMAFn7Yscy-~~@9G!ZpPJTSk~O$1aPj$0Nx2Vy-ESINVD?*=0*`L>Kb*Z+
zNV$ZKzWNB&=?Ar*HfrjfLquDvJOe8ytiqMXntb7GC*s!@*lY=o>1miXuBq0D4WavX
ze)z#y?^jx^o3)y}=qhvYPzb$ac4RP9w|7jtd;_7kv5Mqs-V$rj?~)v9V3{eRds6D|
z(~?%<edLgJbP3cR=Rmj@gQM-}j_kZ*xy=I4$4ri(KvVfLh#nQAk(Zjs28a>a9qvV)
zCHuN#Kig2PyteqEJy<>Let%b6uxjCKb0R<O0Us?5iv#9J2no)|l(_HdYBwgaH!rzK
ztVPt{7*zqF7kB?}%zKouqKY`uAZr0`qU6qb7c@t%OM0uXt+p%OJJHvLU;SN#aUvq6
ziX9zjv*VW$S4m~P?V^*i2sZm36mOu;&nx8T@grpg?46fs_<pTA{TZoB-HkcZ8dU5t
zP4rv|ci0(t5t%bKreE+ZfWER2=tdjAYX~F|b7wg__%DsL+Lw8R5K{w+t;1u=N4<sI
ziqdq_6b1_Ex4^aDN=@8Nc?W8HLx#31@Z{PrreAx=;$;i9jAlc1wP{wMKH+X}Ndw+_
z4ovLZU?rmv*=mrt5%RU7@`r-5b5RDyDao6;M>wU)>e*vgH|iaJX&;$KDc2@3Zs@p|
zA+JC8WK{9R)F#`gxRoyl#D`vX)#_U}1nS5pu)oQo!jxRNNRNWl5|;p{iQTU&=g^<t
z89%1d`zDmQze<Q?%Bqu$;a7P-QTbr0v-#5&^>gWY>;9`o$?NOaH22Y=$In}-v4-5L
z=D`&|g3dn(3hW%BX-z5!iqKcl8(P%;Jgx?s@>Z1R*XO#V?)8?;%d?$_e-od>;dpeP
zMbI<{dPU94GG>4?dO%<30k5;QCcPRhP$Ryi@FUZ2j;Oj{swpun<};nJp1+8dy=^77
zmXP`kqxfwj`NbMwyJ~g2|JrfQoVQONgFfu13AEU(Dh2R2CXP=@c)BfO5Oyw17uP-g
zfNbK+w>qVMr)nWCP5#0GLy-VoRD)7)uq@TjwVlpkzBvaYPm#v~(@6V5arVSCA!gqf
z&NSC>(PyA_;Z#TRWHo|B)-oK{Rpws(qsAk!U~EjK$1S_MPddX{NVMriaG#>w{wWty
zuBj(aS4e$vYI%~gyZA>PML^iz&FzeYt9nInlGf+GS_$K0n;k?h_o`eG*3$d{bja@k
z-VZ&0NDa(ho>*I0>=JXuzQMT?J7o07S!<x0Q=k|6D@Zx@g{d>w&IiR_nbm1jtwMKG
z<;65Ov+&t%D{a1{rW44&@OasOHGSPMobKpq*v;uyRI$et<x+-NR)y5c75}PoPr<>j
zJ_C7@{0L(x;=wbf=anZd;$Dq!y@2M>5xr@pxAE`1^9Tk);YDV4t%#(E%c&`CO3H2M
zi<d7qA1^$PdvmKf$C$%6aTm1Z%;Cc-vU8b!9_hN><Z_ZzMrUtmk<Rx|sWlMac|o9t
zcOp6Bic&{gVd%8PJ6rElV|DYXX+jeY(L{ZGY{QB-?fq&}bf+y8)5Ju7u2f$uPt@ST
z+QSgn^u#@Z+?+SZ`RbQs69=(!DZ;PI?s>-WKaxuSm3e4M{FC&kf^^WCnV54iL+t51
zPbzbh)s&3j`vnEj$R<8T>O|KZJ!lP3ad}*C1Kb^o+5Nn!(9##1YOUlC*X1>Pji?p-
zRKvw8Djl)jabBudmc_m8suPsF9#mvnjoEF9cX0Qa47pJ+Q{$)m-~c6nx$0C6iXcbi
z+^ksfbObwqtUW^1qz$b9q!`pj8{deFRUGj)JJ6R(Ptg+ldhbXx>^6K?h6<?nIF`s>
z?ZwAgt$t<5gAUdupoq<&Ozm%-chbmlqfO?Diews2hVU9g%Pdb3rn@qqBv9Bk4%L&+
z>%TSR^=5l8AN2f%uqab0t;R67Q0i369F`)DEZ+ElVhCtM-A+*Q7B`iPcRtF`xsfQv
zN5ClVfNd}U6@V{4vDav!;!t>dz+jz2I8i>Svrm#m@E^qMj?CPH0s=11%{TqR>D)QX
zi=qo_nxAlTg$jhV0EOf!0*K=ECO@kK<=cS$C!0GsPqV&UM#qW$r7LwO5MAHVzQFTc
zk&;P=NRVR3$7ROnR6pvzf@e|TNq{69S!7D~3oBDC!bK^AIl`58B@SU0>7%lT)~kf#
zauasQocLk$DgP+W`NE&r_#&QYPM?;lk<{!+Brs{kqE@U)8(;|nG=0m>2(Si8Xad4H
zm*W=~Mq+dNY?8gVsGZ<bn~MlH*wS3Yj$L8^w$+bvo(UUtOpp0>!Q&(VYy(7osY_r8
zm@YL*>&b4E_e1;uA*nzs(J}%G$##nB4FQrY5Cs<FipxDbcT)L}{sA(;4`sSx3}LsI
zsw-1gNSpMTUlMyMwz5j;_tGrqph0v3zjT#o=(DN!J_CC*9nR3GBbU)-QdaV!_u6Z5
z#|-lY!3D0Jvq)i2oT}?vwae641j!;o380<p-7u2)-z0cG*%9B>!*nEp^wW&)8=9(Z
zsd;N>-5sQ7lId5a0h5W6Uehgf6yYWSPSOn?zqBWVT8%*iD1okSXF$H&>BTVk+oMdA
zFYidCb{9ynRqKsrP}c{#^3?$T53t&GgyG{ei+~^h@Bjp>hw~7qS4!gNK|yvZ>eQ1t
zxeC!WZwXHUZz*D~ifH4}C)LT1{7+~Ixy2p>V2WYly}w9J3g^hrll4V>ad?yY={s!T
z6yM+}foN1vvW`+76L9a_kiLPN!-86__XhXofTW%80P+dzql)fV6$KINKe*iRS+Pp<
zP5sat>bs9lM6KktivaLE2=F=w20RIqq<dJgTn;4k#<XNXf>_z0&(9+3IUZqNS4Cd`
zJjR0mPKm!p@oyE^k$SM*jY)-pZNU()^%-Y)oB&0Zs$+T{1))eGEp|GT9n#C!GeXqg
zys_1F`w=YNT`ejx>F?5^L@gWuou?t|1A|%`=%GPSa$!{G13i=oC!53Y)no>+Y)N^7
z!aFkkhOi^DtPhF+P(_fe9sU6=F(j!c{{X6%e?!Y13jMu*0;(_mhL$QPKoxZYRGpD0
zu$K`G-(J{%!M*?ObpJrA|8%-D1avR@6w<HjfRlJg%z9@Jv7j^HdGXh3I!&b3xjN)p
zGUV_X)3;CPuX;O)Uc%!bx(j12_H&?jDs*E$)Ic%QBP*Wca-m7ub{_~1*8}dh|NH%W
z|E(Z!QWW@q^7&Opj3iM7VAX6;tVppzPpy>(`xyfPO)>u^9tFr@w8l0g60WAu35S&m
zQCP(7aAOOZ$L9grt^{8(Jt0S9_#A;={xb|;s)_<HtmG=N0{x27IYC24q&qoq@cign
zfufC{E?LwFK%kMra@4XCdbAmU!*mL}ND5%kO{U;@-jiindLLNRUEY9^7(MntE+}~w
zJmdN}u+0=(KnfWOmUu<r$vQr(+()6#Jm&d7YekmPBMft8rDHVw%zDC4zPv$LM&bG=
z4eEu=!x~~3Ng|gK_tIi&HTGL~5X-9?I*x=zY0bGN>Z?A|Gq^mQ@R9l7$q!Tnc;8M;
zjtquWTCr~9@F?i5)nfB7+tgbpJ5OrpXp~vs_*C(6jyfl@?>4ET=zeDXlb{kStN4o<
zwzX1^Ud{f2E*)f4qbkegrN|+dv`K%JyPxNH*Bl=6wEPNhAfX8a)^#_hr?>p*C~b$3
zT?myJ0^DFUO!Hl3{=G2aqzpYFtr(MR>E|Xxmb|NXN1!uoU{Mh!Nk~}{Nm|0X#(9<d
z(|y(&I{az-(Y5xvAY1An#K?vF1GEt*M6-veG3wmm{DENG1PR$XjpMS%>Pj}cEvB(r
zgx33SRkG`i?oxeIJpHm~IN3p^iu45Sekfnw9FP>Vkdb<wz12IFhke1I^kf}TF+4ku
z6G?2<Cdl_`5jYAmlU^L&R7bMP#}#=Ch#E7t_-GE>Kn@oZ0JD$x?twpUSM-is6v`a$
zYM7pCu-vuRRx5YWn-zYD%Qfbq33+5cy;+B!@5zL9TgxT+={EgcCy<`LY22a?p;g%v
zK8z4-!o=tDl0#hy)ZH957S;9g@<HP^J@giYYjix%p;wF20#|nzTM)A)SP*`0j{7Bx
zV7)K)cJA~2pQT8a^YXX+FGKBiv-jAj_SHbS1(w?`LsG*HqS)_O^{ljPG*ZSk)8~4f
zlC_1O4J^JATJTqt+dx#ic463%-n};URagemaQTrLxdOV>s@1?W`2lnuOLngBbawE!
zr|s%wP49E-x%Z~pyWCzf@mdgUZ<w>+hJa|1;$Styke8bc1j`-=8A3!!Au{uq_lf~W
zOR`Qp%V@?8!@|3uFyNXxV+<XaJ&DK&T3<RvMwQ&Rl+VQ{A-)7lvz-IYO3+Z2Ztw{q
zs^{^GE&PT!p`Wy0!cSo;?k`>#`#SEKGh@^HQs_aN`f_*7MlIC@F1^>scW+CG@16dp
zj0T<gK9>W|s3}^IdhncJ`Uop6H~Nt`#tj~Kpinb>y1718Z!L{>>w*~9<_YzEaH*|~
z=t3&NqBu~U`W{%@+KKZ63-(=NH{N!=y&MRU>QIyE{~D<tf+4Bkj#Os1vN{i+7mHl^
zdY#2TjMv#J09u7EfIAueL3-MW!KWgSPjA08GN__A(S{ESu`TG&|JY@WsKavR)<U~O
zpZ0%!+I@PKZ}6(rT!+OqqqEYFmqXJ$fTO|LJDy*N&_~=aiQgXE-)t=v%pJHBroMb9
z26ESv3#STpp*>x#iYPGzyr%WIc&1zqalpfEPD?8mIjgBigA*C6+mP(BdHS8gHMy^t
z%5Na1RB4XbuDWrRiinMbt8c3_lQp7)1=Bl2O8x&VDpHw9{oW;z;^T`ji1^rkYMLxm
zj7&c04m7-XTqt~ohLs>8xT_!|c{i*Bs-$8BxtmMXNIE85GKd#W>2_fcyeed%C!Z&x
z_&2L<7|wDe8m2>lrpN*jtt}}BtJ1XhEvm4I<(<lzuuo;DsE|KN9Xp>Pp*{CVJGUCA
z$Sh!deavfaAKw=LQ^KcZQQjnd#wlT;90WF1e|KP*t_$_I%<1e5eLG3nkk<J`Mf{k@
z6e$c?s=XH@BqNKfCC~@LDILiS;f6;dNCi85%40bl0(<(*V7R|<!WrZ<(aMEtV*ii|
zA?MZb5NlXCDQ!%KiB;K4c*|Drn%a|o2Lk>Hwl8iU^(f-nv|PME+oY^)P^@d5NY+md
zp3Y?-<TB%B>hJGXDOcSrb1sp5u?#Q+TJOp&>kF2i`B?%uwVK2bKLlHD3W{~%S^K4D
zuJ3Cf0p9{TnB+)s?G8!lNMT2&n5EHucGPjzdxPwZS2q`Eu7vWmIKF{|efj&94{ATD
zMpkg*=H!0gX2S}fV<(JJq=9g%w6>)MleTpi)gI@V1o55uD((AScdQ-MUSkdjGL+b0
z6rPodqJS`u_{=Xqr5b?$3BV)s$MhweC4D#7J7=AZQyYaDWsW}iKs7b^tL!BEU<^y7
zmKBG#g6|tVOGBxxdRw(h$#<T;RWG_gt?HEcOsR6c5uHf6+y>lEYTou^Q<vK6^rj?b
zoi&3^)w+Xv$+)>lb-*RY7~sG>B%S;owY4xkgQ!t>T!(^^rILrcIl%aBZE8u|)VS>+
z>ugMYs4$ovNOJ<rOTmscyI2IAU<L}-kr{xX)DB&G0(gqF4WiprnC-)<oys4(PlX$$
zB4N@uqO}T>I&OyeK2rpG<LlGelo1D>iB?N|ndP{w)#HP{!M_BD=%aY0%=iDK>e%*J
zg_97uteEevfWxH{2jKHdc)*c5vc(K+^KPJKX6Lqwv36=myP?tU(s>#S<yQIz0lPes
z8Yt|N|CoPKc<!<-`0TeXo7~3~c6S?>2HJ-9FTy(<4~<nasM;85KW;gsIQSlEkT9lz
zqBr<rcEc(8xX1vx8t`RpNAW5ZPN@E&S`s8gJJ@JTYQ@kNs7L58)r<%n8LqW`3X0HB
z3}(6h$GOS&2e@`dX&<Hm1iqbe5ukx#Ny;CZC)-CZ$zh(V3qK-yjH%kv&0lIkU8xBs
zzt>6?e(PlZqlM+sIJd}&|KUGc499a#&!229qwp4Ku;TEj&_nF}eP&&ufw=i-&KIj^
zjova+^>IUKl09&_|D|&TPO2$iT1`M*G;paVFbX?p=E>1i`h|yE#OkNeoXcr#v((04
zLY2CokUunFI+iJviFVd0RRn&D$D7HPr=>-Ov^__2v>hd^#T=ycj5maO>Cw*}vgB|)
zYES0h@2poFt`E-9s_%NFObY4=WbpQ^xZfb=E?mmL!mql}T;}BT>w2Zef*bYM9nzh}
z6Nq%(8bG8#bc;I6cZEeRlK09n%))BwaBAl2$D~FvR$K0E9UtFbT^isvL#>4WU1Z2X
zkWc;nsMS+AkkK1G2qu4VZ)3f&?Vk(!mlN*=IYsMEbLTO>Yqcdt1C+zUs%ER5A?D38
z-z#|9MNlm6gNVmn-fR6`jx&U4wPFCa)>}wwgws>&2EnRVN^5$3Y%Y_#;X0<8WQObD
zIepo;1dth=L_c|`0H&*lsLnTp0|3Q_Db#WR{Dp;|fWI0rv_B&XfpP7d2w4;XLjkl0
zfV7Os|Mm%}0n^bBLF!ud?>7kkjXpEmArZp|4!o1g1d*iHx(V_5fH#j#p)3l9Kc(?T
z1<|?>Lz&(H;m?Wd|9|x(_SI;v1{;*13f%i*I{3x><6JMI*0#Q&hEwWaU*yyLVyI*~
zc+;m7k&7{eX)JpV+Fi9>AY)rL)nI<1p!Ol-`V0j3T{Djf^SAg14PG7hQSlT}JSORZ
ztT*F*&ysn%i%o&?rw?3$9Q21F0=;U6!MTwXdE^o0oEnG9=MdXLcQ0J5o=DP7<(5-Q
z9Z(D}Sg5yNMnA{6gOyQ)IDR-~Wwcf8o&0Ne_JW&X&+W%#zNy3y8@4*~Z9F156H5Bh
z0n*o~!%V+vt8HmZn6vU?xmo>jMSqz(H9I+E1p4I!!LV6=MAf2tFkEhkb7+4<sJW+d
zZ|~*ka@wY6t@qcm-z&l8PhY#HUfi_w7}%Tw0tm1!qD@Zy0nu_g)%r6-_kGI;$!w{$
ze{i`9P=W!UGCyGnRs*$`J;U5?D7HHBb6S&Gx8J|lGqIqt`{*G2tcC-c-LBXNv4DR-
zCFUr$lV0Y#&s8>D%6AS|o*QK7yu#&p5DU5R#8CtGzy!fZ89{NXx)y|)P5OtQvkF)5
z?Y@!PN8_=}kz|)~mGQO}1!4}5LUE~jzY8-N_5bW>lWttq9vH1xqTO*yf1yRA>kmw6
z9yM;bQ<CR?E?Yo_nck@L#8vOtt_ik*G*q<`Q#6UpcPK9uHjnV>y_vc_Y9f|XweJ2K
zbWGCL!i{$oT;w(0Q7I|ijfk9{z4N#33fcwgW`Gu0$6<BjB>lHo?X?E}#_#pcV)eRL
z*;b>uTxlY#x#UJ7hTvyf9rMY?@2B=ZU)<ZEqg`|Lh=qJ~5J?oJ%FeF8?Q@MusRMr3
zMB|JBf+@^`G>I!i#Vra}j|Vhey;_bgu8HMAPF%M7^6XqAs})Z88~PTbSlMxA`&A^S
z8^$U)8gOi0$HUVo{HI7#kj}?9QKos0BVg;lUa}EM4p$m+0U7(vXvC$xwxvCDoqKEc
z=JV7+5OvASNW0HIPx%C+BipYG$C-L4Q=T7sKIcGl?#y_l=5n1q?U%^Aw}r3*aWN;*
zn;KcOoYGC(YWVIdKP>0xIiKzH`Rdw-V!Y931onjLfLx^nKiuQMM#v~FN2VBhs!sse
zeYISnr8c!*gL|2F$f$8+NVTjN#v37^k4-Y3+xR)#(IAof!TUUowprw9O*N>WL&eb*
zvLV6psjB0<?-b#Q<opNq41p=4kne-S?vXl*g4eSjjre~PfEk}e`k6U#Cq1b`4{urG
zFx@I&v_jvKHRTF!u07`XUZB|LY^<rsznV!Lp4G)#CaBt@J(P{&<*46YXMf+EY;6qH
zStzSzNoMr>xTV1ldmF|`iUG=$W>O$_?aznNZq_&TsE!8oq0oi1L|~<Pw9*hymI3=A
z$xUJI9p^}TY8SD5!i6e`1fKS3O-9{3{IhIbCd-!YuZcfKb%v;Z{9m+vc{G&&`}fEi
zA^X0>NLiD8S7Jyg6(OdwC&m&6W6!=53ehHHnPlJhJ^Pk*m{F2#CfmfA&-3>A{+{JL
ze?8|s=l6%xaZc|2e&5&ix?W2Pm`CQ<=2K~1n71{hC5c+xjaewnVUQGE-oKl74gD50
z13S?`!WTk<a<xFeC@~;K;x2uB{8Jz&_6ZFLG~M|=GcwaEpe6Ef$yAM)gr!a$jCp^9
z?zJX6%cw;?c?izdWQ>(&>(QP+j4v`Qn3tYvN*88KL;rbGH$aNOqz^%YBrUaMCLE?U
zBum#bhj@nq9TdMRyofwU$e>zgOUT!S(=4GpiLshpU(vDU@0WfY=*Mt)d50?Dse{<S
zfCx-lwG<~vFo21V)GFm2Y@OSFYj7ja@m^{><sdh%{w;HEc`zO6MmW|WIZBrQ1IHqd
z{@sX4fdxRwIxr>%VVS92mM%09q~CrG(p!GuHd`so;{V!_GL-q1+%OW@we7sGK)IC|
zza=Vzd?7RCx-VbywiQ!Y3qB;iqJcytBzfpT+{}fN_8Jqt+ttYFj(xStM?u}Cdc*AU
zJp)&RuIiygmBE-V=BvTB#$9B=tC}f;Hl9tICF&r_<IBiHIH5eOtQTut5h#<d@I*4|
zHq^nRE2@{44~f&re+v1VX&Ro-!ZI7O!rAhS^=hMmc3I$BJ0`M4e1c3*95I;Odh&I$
zy2_^}+^_R8-91DPh)vjl8_uW_M6K%aYoYsFPuld|H%ouy$;x>(W*G7a{GwdBNR%d=
zm!s`y2<!E%`Uhfl2x?6X1IpA6&)l8Ocv5AaUGrr`thwpyqjC-%CKx0wn*Wfe8_9-x
zMvVDpO=itm4$71R1E!fN?XQ@`YBVpA_W;52FutTwz4Oel1=H{Q)<1ix61$%B0|50N
z*tj%-iyR5CuT!u&4qY1J$V}_1#X$7d%#SJSF8i@a%6h^$>pc3=nbI!Qn@5FbIk_O<
zgdLL6LGr*40T#9vmutht<R5&;FB0Ed@;Te_7hiaIl=6aURuszQ+6a)i5oV%UD@(|)
zd#$92(_Mt%4?9NK+K;bn*mBj?MoD{-Jp(x^4F@9A#Ob;`_+yf2JkzeiRF3rPOo2YQ
z6`$PE9B4Jp!%C(~6!EiZXBbG`?k!?6IHW}1gb>kW_S!D^5r=FmVT3FJ*e-9HE`LU9
zuO#NZDew9MH&n=Cq<i+N6Gy*T>)H3SMGKtpEI8q)63p>WZSuwDgAr_DZ1QHOCpNl1
zvbwMEn04XRU&)^m2+>a@!De|-?56;E*n1M#Xy;1g_M3M%DlyQC;V+r(DVO_!p@ZwF
zCmcXp9h<~N07BkBJlO@mvDuz@5=^^Fj%7Qsk@fM%o<ENnZ2VC;f8*unn6Me{5%BCB
zpB}jB5eEVg7r>D>gJ>(FH5Po09R>J0@bkJC8|_q^)4m(_--V(pOs_BBR9R2a5l<j;
zkz_D2`FjeOU_<bF&9rV{*qe-)H@iO~Kk>!b&BzuinR-?Hd21+vxHGECk`&<OLCX#K
z3s5@t!qe-7ZUt>A?_oYC=_26<rV<v7l5=Bhnp=A9i}+MysOmAK`n`}_M&n1{GO8>b
zIRdK7jtye$e@>0?@2MSPTjPT-MjW`{596eN)J9uR$s{J!o)uc8p9udgAG+2Lo{(3|
zOjjtD;rN1oiQdH_e+vOH3%WL3Yox7s-ARsy;!2J+(p|AxV%>eMChk@y?eM_mUl06C
z@)r2ndQc2xE+=rXyogr7<iNUONInGXt{NcH_KViI<D#H6H(lZ-eKAjkP{ph+^%l?l
zS9ks=7#KlNC!ol*#>$@1gC;0DgmC{*yd#ljx8olO({lm7dw*KyU7oStVROGdD-JpA
z@q`0}7u>k`|D4^Qd;qt1<3GTg2(ZpKl$MjE%0Rzof_A)HW^E}30+oPBVyVZ04XK7D
zDUMJRVYjpd8aTZ%HryZISjc=XX{0v=dQ#v~F{>Bsk3L;gKDU9*-$8nKhZkKtB2^#y
zh5EQA2DbCklm&=#SxldQI3UVxGs0WZw%7lb@n1K-OK6IdQc?(3h<cqYP=lY^`3}Dw
z7?u6}w*47xS^Bfflt1|i-=A+&>`Yp$mDTn#$9ypbclFS0(_<cbL*lz?`(J;WmLvSu
zW_D)SCyWV)Lw!w>6WtXG2P)irHe~dD+~0|Va$-DIn|8_BDkN%WDEwj1?x2+OxrA#F
z850jx@j?LH$;6TVej%<An8uG7q6h-r16?cM<lnP<FfTrWsAyY(57}Hk)+v@GS?sKa
zh?WI+1y@&n{P63I(8d~UqnGy|!su<<@jZ*Rr5pkVE0A&0{4MW7_PU725EbEF(zTQ6
zHd^5F41fFb|Hb$0|L6Zu?3!+w;?dn&R|#4fWymM}ve=|=mV@*I#UdPjK^QqmxDY@8
zszuVT{7Ud9sh8P~GrT!BtMi9K?Avaanc}vAIrIdAZaEFt_=`CBN=(EiGa`Ep$57gz
zjS8)T!Vm&wWQ)`}L$-54kc3A8>A8et5x^if6`UjH7-gEWEZOwPu-GzWV))}s=pL_!
z2ld{G&=KrBzD<%KIK?-)IFun>O?e|*zh};liQPcX+RTf4@=$3%+e0hiW!um(3^)N^
zH{BabN-2xI3bkwx0j~H!ZskGx1V7WKmi70cTIMI8TG)_{-NBmCR`{Q}iVT)+W<aSI
zH};k7Fd}uvk9$P-A*|ChEKjqBa1aJe+Iu<%#t$TmoidSq41Cd|^j)hsU<%YQGM9xl
zjoTB7@%VMbS>gF<_L_4V_WR~l`8BcR;0VF|L=X76U>PEg%*h=6wt>L-W2<^<Pief}
z>v${^Mu*(*v<H?~$31a>nGfC&bFf=n*t@^xE>?Pe`aWmfGf-3b*w^Qr<%stvQybH3
zXdY}uxqFANwJV=-4&!?0caT+!pB<u}K<FFtUr^c9-;DT8dnZLlU3N}(c_#s-OSpRY
z5Py){keIRODYW=4W{f@Yb`%rLrk;DmYK_c>I!{>42V%n8Q+r{B-@h;E`AJQf@|_X&
znAg!47w6`GQmyR*kF%8-*Sc)$JSS731voeCM!a)xA_sfQzpd2|KJtYy$*1{az)Hu3
zLWp;vnXTX(Gggsrwl>YZuldf8egrMa#5wsCnCGzmHOTLOAkHBXE%F5BZbl;;Jo!~a
zV&u{Fw%)535<VL8m@~mtZxQks?NPkM%eX{ktVwjStD3<?sg$^Nu~;T@{(|Wu1$S=#
zK+8C|e>$K*!u$0A$!<5}U8_-rcSYYZF>kCcr`~d~yC>1!w<ei)^~zIp+fF!HWsEHD
zJXc=-kx{Lq{o5ZSDMHOQ9sN7*+}sltjhR^m&`qH35_Yk*37c9Hu2$|D<%=!{N>X3+
z7)zdU&+&;*#oCQbj2D9G#AW2*Jusbcv1+OJDS80+Pc`D0Kv1$&M!|J6_>b6hgq+6U
z=Rw9%1<o1~$Ge~O25qx`V6IFv%`MG0c)MibsA4rOL<ho)sr2h+=6;Id*c;-roG$m_
zW7vBQO<IrNyv)|e!)*7G@|S4^M93bL*c1fuZtxIlG!mn+3@73Sypf6fn=`4#?1Z9u
z(aOjRulN&|XYL%Upe_)iqk={C%0c08qc%K#mi_ESn2Y6P5&ikR*ZsqiF~n+ucP|4$
zyeqTD>SQspAcHjR(eC_H;a;4+j^|?1Rny|lE3}r!S4k!W_e#Q50WRxlcST8t?@*~M
zn7_WBRB3dP#-jt_y*?^P5bDy~|9Hs=(GnY%WxK=(TEz#_=a%LbJDdE@FWlhNvrEx&
zr1II8oN9H*j|`T;`^EbhyLL@W7Iqv+JMo)JG&U-vwyad8Kv;0;3l<Zn4P7t_nT%6T
zMpw@tXAs1;-uKJ9g-?C+)Rzza+#E{TV(Ml1vpEYCw#nEKR${IBn1*igV6jy^n?K)z
zz_|W~$VHCU2HihQ{*>Hn!h%MhQ4nIROOYu8@$RICJNNTLUC;A|Fh`14{6Zd*bFB^`
zQGyroj<uyvmo8RXQ)e$&byFIYZ~IAd{xyVnCe{CJVtYWGz@5O%>KX+_o2%IBJ*g`<
z<F1%4-uce#Wm+treSn$O#tOxfjLTiKEn<8wD{_l~h1}G0*C0vYUMOZ!cu%tz9Es5F
z`$77YK99ZKOF0S>Q`qK3IsWbxqnrpQ`&-@p1f_TG4qjyArX0~*;Om>z20DvDZ5EU<
z8Xf1RxOLOmP-;tWmDzh&tyZ{vXFHckTgrZvHff0jAtodI6y%UW<($eAVo6x(B#Prs
z87AgBn&#F6526J&<eYPhjG#zOvT4@+8qO~_{*-*V@X%e|5Q4(BAT=yTW-vUc5FdZd
z5fb$1g|uYltCjR8;4Ajwi(m#|SXIh)<paAAp@gW$;JI{{%t1ed*sx^;2dY?=2cF`&
z|1(s>c4Jw*E5_X+Rzqz<LiLV_jc#?DO5b2a#7rt&mpq`&tf9N?%FHr?b-h=absNcf
zNu%5EjJCYg&9p11_9X9i^y_W197?CYA=8&&{wWG3W2UO~eszrk1_=ay5C<^ANQ?&j
zG1eN+f&t~-cfrM#Y+L)2wb%8A+uFWz>*-se%x}BxkQ0Vkw)o)2kJb$^rFV}t-)p>{
zD}z3Jo9i8F{4B1UP*b86xjYnNQDw07v(vP$G1h_V9TMkA5h{!mcA^5D=m``RF`Yjt
zqSMTPBW5Y<%F363gceV@0eMh}893XVaaNlkW-qB5?w|c2Hy=ycAVI_Pq@I5pM>R8;
zCBz*pY%Ez%a4cewd%|h+Azz0Ea)AB>I!g@y2KmaUVx0yo0FiH+t$op-c89)?JHu|_
zOq1Vg0uBOHh~D`uRYu7yL@t}L%{+zr=E^j>(;<w5uoK7Fj50y=K?9IYhWebqP5+Q*
z92}&^{Nw9&siF3MZ<(W$+YqrAV&La)w(;9PY#$+O!_9)fY0*qKF)ruKhp^~m&SYgd
zF6;2C>%Y72(dC(5tHIWePP4{>Pc_Xj$^tl^U+fNoA4?EBtq38ZI|T&B*`~V6<}1)`
zE#K>L-|aBe$tya7V|pAtC<g){2t)?+xXulyj1`<zNv<}<)?IMAw;0hb>Z|W{mF612
znw<24z4PNS=b1vV<x{OY`v5(cMfN@*<afHrG+pzMK}hX<tYk}#PCcW)+*%ThPaw1j
z{4t<Kg8<0JPUWR5M4R$o0R{<}g^zBmee=5P8PNH>t{1|A+1t*b1<EJh5dz@61*l8U
zJOVZC#!{IUJ2EBsTP4xVVA<H$vSm)1^bNOW@n`V)Vr1{9z?rd(HSH>{*lO9cD9Nd6
zW7hlOMD7*Co2xwn9hq>Cv+Zx00c^i=g9J<?-aSE(=)w0Z9vzUs0*k>M*vEIcu+!Qm
zA!>3$u6XZe{W*NT7?D&r67Y9~%z4<we8lI2I6i?pkw=0rPLeF~wMLWs^_hWK`^X;*
z5yCJDo@P9aF$5I}-e(~|NT=ToS-vGivI+dzh|ljzCx4rRy;wGY&OCEcM!Q4Xp7l$<
z>U_RH6)`Wo3)i#A2p$JU8%+ZE0qd<Z%ht9F4|`qLA_w(Lxzp;Gl?M|J`g=fZz+ZJT
zJFE+R#IXiY@0bu`!(j+JNthruvbnedc578uHTN^i7x6L7zAfF`<kr((N{`*>hs-Z`
z$6LTN9obBvBu5QdkpF>LjA{X*CpJdgQI-jr#!&Qbdf%w>vF11|ioQ%`scBjDe7T9-
zt`Y<X5d^~qa6YCd^nU;#{qG6QBx$iB({>KLKxVp>`O=!zc)9tuf!VFI6uV#JOlV-a
zYRdwJb1k0#*B<lKiSMw;R>yosBCuXl4|2WNV67hiD&3V@(AN^CXSZ4&==B}+BS=5e
zrI1+op5zE%GhGQG_#9w5(9DZOi<O@d@1IJu*;>w1z3lR8{A&1bJ90LpV-XCMP8hRm
zPsr$oV3v);mP<s$TUWK>M09ys=dH2w(QjvYdY|DmJV7W)3xM@M;05b(`Da~<R~nx;
zG<<EiT3zpv@VgT+U`?4+2=%m`u)T0>ft;|$M@Q_nuWa;u%|80o3ndVQ@7sL8J1>q7
znGV%*g@@*|4nk1!BH^CtJ~Xu8RrE%(8RYBWZ6iP0G$A}6+fkP$@`xFUP(z(+Iq+9e
zkBFg3r5e_sbNN>a`p_*X>dy<VEwuQ9vY^T>i@BiYZ!Hfxc92s^w2#xADShK4+q7Mi
z1zgGJ$5GlYD}&Qt-1H6STx059(vg}Z^4k7JYdD4V@la7>HriXCjzk9w6Q-i0%QfVm
z)uy|QsP(%G-1%<wV<qef4Q;0Ph$mdGoe+<ul81>9(f5Hb8RcNT3A2Yj-GSoaPQ<=y
zHTtY`L&0267FW&fhkX^{0O={!p-NB4zpEoeKTo10rYaC_Eb%*MEL=`AD)Rkga@AvX
zIgk?Wj|}MFTVyc#lA5tJJ*UEx^-kiQp-=`K{ifA5Sr8fB;j%I>#PplhWy62XNZ||N
z!t57n#_y<J?MO%v2V9*z6~c%z#2LL|zKtKs#_-uA#lr0G`7B&x+&xc+698u9K3|DL
zgfr9I3}M(e`ghr)|NKY~P-60E+o!EVkG_%dZJ2PAfkHIKpb1=Wp_ILAIzm+6K?RWL
z9Yn?EtnCm7qn{Y`Icw1VyVo^6Z1XJo*-23gH!`RDC@;E*=r|-hJvXzN){!Qm_NF@m
z)pbU-_u;X{I2gmU1;95?2)uvDOG~)x$Lv>CEb6kcBB@N9*DF*6+h#B?|I1Skz_X_i
z?SI#gkRo{!ezxf@4$k2cfQj4^ZN2f8)2n?AYRY5b$C&kqPRmh6mHQAafk$s16BtQ`
z;3MShOjtjNS8J-hy4bP7TMT@02ibaPXwB4^Tw71dQ9)QY+8#BN#CJ~*t=#gza}tI=
zCbIQyL1TBXs0veY3$CHL!RtgrdO#S;kE?o}99wPlk?T?!_0-@Hdt9%Jxw(oc)p61X
zDoP!pccO*sMj(!2XF3b8iPyHI@zLE&GK5vNnJ6jINw<P(0~_itPla2rH=47A*AkrG
z4Xcc>;c=jTU@#SQRt7u@!F)R>e~!t@e!p{}bPHN%K2);$E_-6KF()Nj!o)%YVkA@Y
z2Um-1x5Vc6H#vvRV;WUeW$yiG`PEyf^{ELg<xbt&yuAdr&WeX%THah!*Ag}NKxSLe
zc@gT7m|-_VD}10p@msqpq<w@NS%{@{ZQ-qmzf)Zw#xmwE3UmRtoBk%$(-w(p?k36i
zlVpg@xIv*9TcJsxH!{*{9UTXJ{f>ck_I-(+Tl7i2;@s<TW8nD_MOGRSUmj_tmMyu|
zaQYIyioGm5d)loH<W~HmK|(xKq2LPA0q^Qjlo5Wh852`2xnVs>l=>O|XXU59A@m;j
zDnnLT7MNzKC`VhJ5#eNzrqmVK9N1|t_f#ostT)3t2Ivb!?{%K#@qSZJ<%AKy2eYAQ
z2#e8OnhbYl6#A;>85ru&(AV4VvivevYN%nLYg-FD;t6qSjn$H#2)?}ZYtE!L@YDJM
z%MH<_0rMC2$f?=f;z9lz+DEjM(DWe`AF&X3hoIb9Gh-_<iP(}wJgLlEI`GR@GUUIk
zV%K>N9G)LoU3Tatam;hpSkw4YEy>>jS{daSmm1!H>!~mYdEG{op{RU@w%lOqCggBf
z86Y=tf(jvVYSeGmVtg0@scOZ|Dg40qHNEO;Xm1q%d6<UXJK@kE;4XBVwim<SLWjgW
zBxaRt2ntQuO+J&Ib+oX4_5PA?q{NMcLQ!`wmDRvWFqYS%N6vR**c6%(T~^Nmd=isi
zUvR9CyL;hMA3NP_{ol&;ux{`CDz2_pZYz@cbk&J^7jII&!;?X(E3?C+<w{Zg^!`sC
z{0WM_c9%RH!is_i66_2n6g2~HPd-C*r3^A|Ze4-s5A*m5@=2%yJ}FnrRTKkZx(vY0
zQwjoCrEhkS?Yz<oChxE>`0HEpn80daqIn?Dc_g2@$!P^0+YsE0HP(_=8)z3pWMeN|
znIrSAd=_==QC<@lkW3`56V!X|knYTsZld&xzFW>dR4hGjX)^j&Z~J;`CN#}(tyTvz
zee#+NMK<0hR(^6mDsK!bb~y^amDU?5Z(M6HEAo|7`|<$A_Su6r+PQzpYnBAlW!Yo|
zun;<miZAFcegIz@@@F4g)D@&kf<LzKQRV5qWpU?NmBd3xjwFldl@l|2q}9g9c{uKu
zi)Q?EGON48cLm}D#ns3UYnA-h6jyQ=6j-3;<Qu`gdESk6GaJ6NBNnA1En~}_dM7TF
z!$gfDePMOx6S;p$o1kBd8{87Io<=B2UY4@Oaivw2CMCReb5!M7@=_=02w9H7^T-NB
zG%ja^^XaZ^?7sC4Wtn2?g27Jqr-jb+pFLD47xq=5+50=QLcL2DXGl&Z&U4#nzhXdX
zajKrnspqvwoHm#M;~)xF{^Z9yDa@J#X5c%-$>?>Y&L}8-ee`CJWD#0eFV(P5-KLYN
znQmiay4m%>mS}`crXuFwohIFN;|T9}oc6T-+b>z~1e1OF1i~kmu(~#%|FP&TbA%#d
zb^@!I_d$cxO2Oh_tp36Flcq4v`<v}Hoy_kzHe0^}6f&QUDEVMo5zao>{rDgr@m0_L
zqM=QT!;{r_613bKAgTtqO_-_9ea!1rpupP}fYdBl&7{nQ>M^^fT9~9jmtTs+K=Uga
zk4~)@7FEGR@UtZ2<$X}B#cpB01RjI!F=waMmjAh(n&g?V7SVf)^rM($2~q}&78Yyc
zs$<TL^JtW~){7>}&)(yQNG^pqm<l1Y9++}ixds&U6big4YKlY)6g&^U#uqz05!U-7
z0mzUv4>$J-A`F8uWqX&l-x>0lA$Q!S>>>Cz`EZJjJ{n*t<U_oJ)uwmOlM=ibC3$w=
znY#4K-+0MOK)|Sa$)I;+LUnl$ovX+kx1`?Ol=st=V?(ZKoQGc5m+*cqS0W-*6M}x<
zY6sA8Vg#_UPJMukxLY^*r0V?8xhxa=RHJPwcX9RLJH76}t-2dDluoU&cWjK=LFx<3
zhYU-zqXnr@ej198w6thpp*!&FK&HY-y0$E>-o-}DquY{E<O;K^u{6`aAOG!d`~}I8
z*KH$D4{fV%pNZ=uB&8i4zVI0m<T1aj$iTD3sn0`apZJn(Q{HMn7MExi#N2KxJXPA|
z_#5<l4KW5+3&xck`f#Gcw`n^k$Rltz(d-+h%+SxJ_tY@+p-Z@{lo8w1K|;r}DAgGh
z<=fvo*az6!#z}|7%vq=<ucC3M)WFTv?O_UyK)jZ5Jy<R4KlwAq@3(DtCk@VAOaDay
zVSd~lbQTW`1!B<vQ9RhaKgoByan#MezA|s)FwI@==CMsfL)5i|-F}`P2;|7U9>CkY
z&?P8dzy5<&U+MzfvY!fZ(f73T6DZj|GL5rU_qQ2|uO5Avc>!ThDLzUWw#Tuj+zyWw
zyx2e#!Wo0E<+kgh?hRPlICkSOd4RK%?J!hRe(x6qBB+L`T<aP&5aeTwNQggn5BVb&
zq<t)pWVXa}h0AQAVEi6%TbrEZ3s9yfeAh$2eJs|NL5i218cH*0_vY^@1L*)!NSrt`
z-TE~oz8cLzWIAcO7AUrNVDX7=?PX?a%I6xMWa;qU{F{&i;0Y{6e7eyi#I)bcV48WU
zd>ea*%_t=&W2z1MoYq}_UU7O%M+g$qsYNmtKq)34hny-Zs`pJd=a+--w!xY0C#rWY
z%Wk7>2`h;a{uffuUwE8iy1m=We~32!uk~nION|K5w+C}>-<Z!p?xaDNzd|b7*(-_9
zJ>ZOILY`zRTKVK%lVLFsX`7&nohwiBj^#QkDRzbZ=x-kJvqgX*lmU?E1`%@r!DbZj
z-cR)e{AhGA71HQ7-lEe+9yWb0#Qpl8fz+Di$nP}^dr&7H{s4<k48D>CE{kqdE9gJ|
zgd_%49%~@u%?PlLL{YQp!ehMy`^cZ7KT||AsGF@J_J_dCJj2%E_y$t)9s$!CTQ$Tm
z;C7wwMMKhMKIbgwN4Mpc*%}iFzhPE;b0BepndpQ8`UWZBN_g_`0r?Uv+DFbWl84op
zUq~%YyLsQNR~|YaRC3u)POgL>0(&!w_dOU2VgsyY3cw9JCa9Am^GW%qS2SbCR*85G
z%hx)K=_H?@Seo9j8h6*5??G2Phz}1)D$#&HUKVJsgkL|s9YAHRjWPj^uhTye@vT<k
zkDy|FZD*|(8!_bFFTZh^^wgU=Lo2DR1w-h$eM*%f9{|qTDuM(;?HP^06=^kDMm>)1
zABp`h=d0rRZk8&&DeaPh*%~Yhk}ST?L<LU=%T6~*E-y+eFswg=wlvBOYMe8_{kbvJ
zoWI;eKFI+0P8ITZT6=Z}>#A0Qd?pM%GZf+mf+;V$ZfjGmvl2tv=Qxu+MU0g4?Wju{
zoLsL8o9PF7!t>S%uwECD24btNn2G+zIvwenisBJ1fJpAu&uE-oSu`iUYiya=+IGBj
zrC>?XF6_JgD9s5Hv1NpxhZ7Dm-i@rhdf!k#Vf*NyOr&RADi1F4?K<JdpR)qU^@#0z
zpwiC-u-8miy;G8^hSrx4U+zj`tTS{>msC=6nVQZ7B?b+y7|ZR!&$;kE{3?aEb7Cwp
z_&_lrYzC|7b_Ki<+q1`Jz+9gzS#;E$B!>`KsL^`nW&PSr>B7OAwt9xJO73BQj-8C%
z<gFka{5<jrGt&cP$lbb$wB6xM7wOzyQTFfb?)+i9xk)5orWiL%9tL)JPEFxFd+aq)
zu>1b2&9#!;;+Yrhp<y$jrqw|!h1NO7TGNZK<f6n9QZbNoGix@<-5~<AOYL5HlG9?&
z4^<4ymu@`cUwN6!@|!VmX6r5vZ1L9tnZ0_<M|H2OsYW@yIp@_I48+bub$(Y+TVweV
z<!=ee-7I5bC9Z_=`qGxh{Zf<BI!})7iGIH2JqK|rp_?aMVc$~p*Lc8CY+Z->U->NX
z)BSJ$7ts9XB38s>NeFC^tTn#5$)a7VJGbWNz1Vcu#^}K>R6<u0d_d%$Mo9;}awd5s
zc5h_MScy@sUxw4L+;*FX>S>`T=zdr|L**tDQm(>0k8P1_T@B(1dvEw|Oe3BTUU*cS
zcyKwJ%b(Zq{I-K|%9(Rax!D8^9m<r@oE7&REZ8fTr`>hx^qk#ypn2kE9=f7(w5KDf
zaEiCgqeqgf=(+h`+lyl;|HmJGWltTvw$Hrm8Gm_>)6dVt=xT~VM*?Nr&!)<i!>$oN
z;IO3YM9{DF;ah^Sb@~ND#{jA9Lh7rCc5U5oRglM_T~*t@!-;I|!Mo;hRmI;9w|0m!
zAQ3?$p#El)g4U0FPjxF#bOqfeBshoMz1}>n&<9vj>V4$}ueaZ)$9&fsEm~d038+u^
zZdv0-*j=<)9*sW_^1gj44Pe!?mW&Q<)be`pj=JSnPk`@l*AX%o%9v0Z!lI$hza&1j
zBMf?{r3W(LJe6S{rSuh|D&n*iZ<#GlK9CuawcWN_<VF0@N2|o<NxC$B@g(uc6za7=
z$Tj|OO(vb?yt^w-ZKeVbNY_Sv1iVe+jlX<piu+>-EFb;+-Gv_A^YZdc*b@j!#DM$_
zvc!azxK?_B>s-<sJ0}EBk~&{hgtuDMlOJ?a0d^WpizdjC7jQNWXRV9KrwsY5zh`P@
zvR`nfU-S2a5~{omiZGI_6z$7NfZ1ns@?MLc*w_VpCS*|8R^lI2OIrW>fZ@(o$)mY(
zuYIZQ%7trGf#Eb1DnfMRGsvHvSV1sannPJR<`*vX`a^-6MnO}}IhOm}xoL9gp){s%
zwH?ENGvg{|j%_(V0cA*#IELe=a;*~c-loc5%2UavF__jY3_Q2yculZwYZ%^+4ii)!
zCyRn9Q|n`<oWk{8hQicsyO(|U@Nq?5t9fex@Ld+F0G4M!AkUnlrAlHVl-Cm^?@w!J
zKIZ&j7i|7$(34$XzUuOIfhg0#+=K*{Z3#qZ2o=d`WhXYky%wuZx3|GZ0e!8^GT`8-
zuk%)vKWKB6&>iFnz6U;p56c#b`-qsip5m1gEooEJNik(Zk%!$Z3AOhed}&1m^siBQ
z=Y|y@Yogc)$x$t0-(AUKU*k<GC&tYoBm8CZ8gAUTS(4;hq4~G`aM8paf>4hf=^kiS
zNQR$5vtr`*Hhf@xYwS+Yb(TzMZbkprC`7T<F(k%Ts)blz;G$L-qbRN4adOe{eM`N6
z+blz0fnR#n5143;Gv(Gz$z@Cepmk{2W7tS~2v@AJdC&B~A8P~QiiR}0DN(NvKdxWx
z<$?YQent82<aLV(@#`VBC+EZO?1Ojzc0lQRK_0tXvuq7tImKNDlHOcftAS{Zy>Fo-
zwi4uf5WwF4i#0I|C|fF9^M<*l%2&e`lT>-Qeq5(e;ZN^Z{2n|Qa$e(p5y>>(V5)i%
zX7;72Hl<%+zjLJOeWqDtzeJGdq+q}~NKmpKik4u9bxN4WTEsr7B`8%mA>4{gUC(j)
zrCsBNdR>*_cyMmnD<IuwZL<J}VSn1nDmLzu=y!Xv)ZI-`-)JK`w?tgxUcFdAE)LD}
z>k36ASaCro)B6WGVVQ3*{(kbAjLgS8$(IdLSK7bV&^JZ_VtdkpV>9!CJrRaQ(-1q3
ztqFR$0jTR^Wly4q4Xw;JKjx>#K%VVv*j^^EVG$tI#Je1r-;~bj39MGYs0&t8Zp0<r
z*U@`K3kfp4*sOhEM8L%E=L=$7c&BSaif0X<9P%NYTxJCXAQImU`BnP_kO2FH^$1mB
zx>&GpZm$R2vr3E~9@p~`Sdfc0SC1ap=4Q!dNqW1SIt0`=t@x~N!LtPY;xgi=IZ2K{
zc_+rY*J3WpYO1IY=OKRx5&a{zT9i21I6>uRf_s@ZNlb01NLi+Yv0yW$D-LvcacxYH
zBhcZ^d?EJ%#)(WV&c+%xy`QsXGI*?6-q!r)({rCRMLx?|;m5-;lb00xK^%W4u(EN%
zs*^tR$2A1=YIpnBF;U|O51w4hrM_nlfk4Pk;-@+rU2vj4Biv8tWa>+8pZdZdxppa6
zDvZNK`gu{MT`6?lsT0DdG)42%87khf@K_C{Up-m=sB#S3T$6OYVxeGx?kn1newc3L
z9mke-rA0>w3sFm!Wr6tqQYmy&nA1}SYM3K%>%|9-UKL8v4B@o(?z@>d)l9grCG%k;
z7riGg7=kT-`T+Vp{o3X{J~v*zx;q0O2}?R9A`zi&@==iecF0T<c>2yYU0c4iKR?3O
zDe&=d>EWGusd!Op<6eIUG5^H`Ck@`ej-cm`PC_6G8`8Vy_7L)~>2*V-Jj+cyK9`<J
z){*>$`F`AE1-gc;u1qy@u}?9~Wp#*QabUc!;4q5sdwTLq&v_7Pb9DRp`>^q^D0qlc
z#)_uv?W*~Mc%#ePq6=4*q0c%+CqzOOHzd0o3};A}%bYi^gjF+KE~6^&quLd<Urdru
zSmXG+;!Tk4<&q!N$h$%~(=VDOqIvz^ZNm}zQixz1MBQQ{zhWn0AH2%zdC&?1O@U`o
zdEUVe)r^e&OaB)4A2KEeC*OMl><@Y%8frM8=FbxNR`Dr}f1p$+Rkp07j<yI)2e&B>
z(7r@*e1t71-8zN1*R77*9g2`~i(6^u!Ph%k#-LNx>|PpN#SBTAn?^f&iouSPM}%0%
zE$)xnrQS;sK)j`B?s>=`<ht0xg1RtMg?J$-HCkO*`*lk3M=GmG$2E5)cJSgv<h9Cp
z$YF$oXdm`=ZP{+<fDBRr3lop5%uKT(hKzF@;b~0FRuE#vzs+e}7c=t*NI?7sN{AHP
z^5!I4gR+E3*{*E9bb+NiB=EUA?;^K$no=a>3zXG&E!m1V)dEE>#@gY8SnZFVdV`Am
zoyC|Cz19U}Hhgxi`)FvXKy0b-qk6BakN5E`MB<~n^qVd&mEk}j72iAI_*)(<id<MD
zOA4{zT1mIS7zbPOYb3grrO5<S$rZc5UM*fXt<d<Xs-`j5KaE<;aFBEOr=aVj+&uwL
z9U+rD@a#1#EU_XC>>lj;F(9)MKOgb6x3ZFMW4?~W`uyGJ%D0@5rH*V~{ea5hDeL(u
zzN5mX#C>#bt&Vo`p2g34Vj>Pu7Uh#;{{90g%EgfG2fYR(yFv>BykoCnW_#|>_Vl)q
zUY?1~*nK*gkl2<dSN&CZ`ul|i>mPFU%F<zwn3eM4_v@y-7{jj*GIc{Y6Zjkhow)mP
zd9KZ9#gq3<y2VJl6)kkpuUcu>zy=j_`{O==j<mvxb`{Mc%GDhrKm_vlWSB|JfSQKO
z*xzZxoP{IBS3YID|2@J6zxdbT!SHpiC-d?P2{_T@n?wsTr#2u7{_+BrlZ@i0*K2EJ
zcA|?nE4GErMJ=k<YsRASW8|lhrWO9rnVQ3p{@(%FkiCIRXt9&RCesqo9X_kIVw|5{
zTv&uTJ0(ZoaCqXIzoIR3Rh0)vOB1U3G>$Y4235(p<Rktja6Kdby{5<LW#t$}A$F1&
zo{TQLB|dr4(9F^|tab1;jtW;#C%=;o*(rM8gg7WA=%ZKQQ68pmmj>;;{fl7EyB*#3
z_vHH4T$%3upcTz^4Ao5gpSuVF#y<ah7aJ=~wLw_zs64@`Y8OuHjgs5D2JA9FsS=hw
z<Utbq%uEt_B81r|Ux;|ULz_v;8~)(oUyP$4or5`_$@CL;mYKAx6g#0oad8kLhqhCG
zY{Kud72+r>g=0%?+o1|-W<iU74AZ&9Jf`4MVEJ)k^If$I`y<sG^ag^_9F?||$F{^q
z8~jLAl~D-uZwp?}*1T&P)5CTzU(o9?soqH+mdmyKbPaO&TXOfQP^^|}+`Uu{JFG}w
zduY&2M;^hqHFO6W7NhVg@h(%=S<)4vk7wc2yDo=nTQN<|*+U6e<zwDHlcdZrvws%0
zspbU0o-ufi`Dr!aiST3}Ber%?;~xEnPKx<J`Tvx5T{#erSZdX65-0+<OjmcZ{NUu%
zpB{!--*b$oT))k$&Xl%ITnC?s4p?Tv_P2|~C*0+QU#|9wRo<+qu$c3sdBT<Ws>gdM
za7qk<VzWC|C&B8PYol=banom^_x+dt&^&mV_aM^CfHD_i9O$$q21c^BvNia{-YC1I
zBO5hJ7X=!<*osiy@J$-6hb*~&<nlcC?t^9C{WZW*^&47iqc#kphUZ`mVuMRM8?1*<
zKL1DpP*~FiQyKPKYfr=S3)JEuR{!0SnA7b`{t%iw?pYEYL)m>;=T)i*EVK*?bN_C2
zYqYX&+@MhEBkPrj1Y8+1DjFcqbHNo(enN3R#d(+SH#Nr3UQDcCy(%Z1Tm}geJvK%O
z15?T-(@(C!Je-GGHm@qPmkc>Ya%>h<994w>zObuV`imPRT|QPt;)&_TYHzCFB*vK}
zEB4nLDoNz#O5~>ZK=x1h%3A}MD1VO3!IyVpn-JIy0q@Ssi0AqC)|QX_%`YaD$kPn+
zpZ6n!F_i{hdvgm{qyhi1@N{}QQ|ZdXUWLk>mnlX1jG>>s40~v&Tc!cU%K`Kp=ZGJA
zY?UJ(;wOC%JqDkB8Dtx%P`LGCnbb>lKJa<_!3Y<A2Y5F(Cnr5u-+IDeB*~{<w_T96
zb@oEnGa={us{K8hm2js!IzqC+V{mHJeL#T_6bS4zbn{1epMj=ol(2R5o!PBh3F8({
zMkT%YgJEcO8sS~XoL3MqvJRi_=`?MMJ20eLzl@OXd_iBR`a~{oc0`k&JV{m~Xyp_1
zZDenObaJhp@g-4yb3aCi*rYt?oIWn!Hlv#)3ZgFeiEmb7J`paK`A|=Ppe`L<wCTU9
z0h*UiS62COk3e3?@xgyQvpyVT6V~FMk!60*+2*5vn=5f%KBLQ6Fxb(mAlKKmCF;4*
zo4zT@wd~g=%T?R0&1|)ROupAt-PB;@@~*MLe=IdwTQZGePc-|KOn6GWBq9!80JD`o
zEh*PbCEE*=o28qp4<1L2|5(v!ytaCoM^!w1<cAq~SWts-AJe_aW!g$$Dh_lA{MoCZ
z>?eCQMZ}N$oC(uCuS=BgiT!xy9;=xyV>^q{Csi#Ktyx?2JwCB1wf7$N7HRfe<NoDE
z6Wzufwj_;j?J8jm>GHI;oO=M1lqemPyfwn7&+ex2NL5wq)lTZZ8(_4pbt7gEdjj&i
z1Q2&eKUB4RnUt|5@PDhg%EMiuZ?1U_YVV=X%pX>Zr(Jf6BgqnYN4f-|oYTH@C@=Oz
zF-t2y<eUbo@CQT=ny(|*FZvsFYPgU%?aH-8MP{Lmxx848g8is>{?e;nJkuB9a)U1=
zdOxS>2QCZBhn<!Z)J=C+goI?8u{Wx#YOBM>RmL4MJ~^4Sc6hu!muC8fG7a_;1t%us
zXo*=^XJK&bToEosBd+I^YQ1!MVf+c7m5M1<`11hubJn{z_|l}-CEbFj0o(729xToe
z*vI`@DQv$XRba$cwKjncfM`L<BY8@bD1Y=*H3d$;Oy!xh#oh~b1~IIvK?M)iL{$!Z
zTq{p<wZMbk-lCH8kRW1h)yY_?%Wfz!FkZQTTu2WQIMG+f^?)P=JwgOXQc$=}%(AEn
z$K1|ay5v^A8l`X}<#Xs~9q`TDvB7}OB$@4L2y6bWDW4(T+h6m>Va_|^E?@c+IHXLL
z3JY!bC}6$t%GxA%4ZG5Z>q-Y%HOgFOrtBDJgMOy&FZYwK@$|0Ph8-)C_zBwKsK<e2
z#0=>&yJJynJ^Gtpt0G?;ssg3ZNU!2DXtuRPh&>1a*KaKsr=`ba9b!dln$j+f-5)5T
z|47ls@muICEV0R`1a+->ygJDJAIP1rY29m={nc;XA4@q?5Gxv`sWwWhiN(xz_J-io
zpCSt8XEBzB>ivvu^#+F}{0Sxe;qu{{Fd#t%(g#OS=Rxt}vHQk|u0e#`Ps<z?H(bh1
zxc7%FF%yX31MZ#18r36^bxjQH5~}n|zA(L$ZX7;*G&p3(*jHyj>Jk^uu*=Z+qq?R7
z^$LO?I1sh5vS*0N_hvM1xoyoE={hi0nTW~oj16YeLKs6S`c*Zr$nP1mL#lu_kAr%@
zb*|K1`taHA2j@dOmA)R&y8uf)3sN5P&%bTXlXR6-f5kHJanLsodK>wSEj3@)b{)9=
zO3$<ieX94vnu_r3G}WaoDYKq0uG^MkQk6~!v}gjbq2-e3L`V%v^Rr7ZBK=_A(yeK+
zu?BIyu_86sNu{q$;u7TwU;b<>R_bm)dqB|3UkN&y>Rt59f*+v#f=!`8kwKKla;PiQ
z!2%#QfGmPE+r5j_sfmsFYsg7rU0%+-DoCqCtFavD_856+RzF5!z-pRwXUJ?;&F8$b
ze0hZ_RsOQ@{wh6W5d__AgS$EGVyjO;#3;*is{8P~v_FD;7rJ@GwoIFhGWNlzV+Dra
zeO|C#_%)62c2Jz)vDcToxlrlVd5MlzI|CAA_&`f@^dmt#La~gMXL`x{*id@f;$fuJ
z$1|{k-~b~6bB~$2I<gYp56yilPtm`}7zV#i;vlTWI)Px7vg<qMbq&qqqJC9%^$4XK
zVhmnGY2P&x)&$f%bp-7(l_6ao2D}dM1Y6ybNv?}pd?bPCDs&hbimxg0c60l@aqAC6
z>v`zP^{B@)>J(u7&>{!FlpkX&4_)hS5pD;w|4$GyU&#h~R<!G=HCBVND8wB;5+wG9
z5Ovwf>yO)|vhd#YLBT2tF#jaUXLW^v^wSbk5@S5rnFb5q+v>|nK4^VtpDO#-9JQd;
z7n#uf(#K^*-_#;j6TS>9KUZ19xJDBf%P>Um6?lw~jbKP6rky^f0Kyvi@jle4g&P~H
zrKepP-Srr3dGJ45UeJ)lJ6Tkp<oS%y*xJr_mOZUJse>rP7>Z*`^}^P)3<jBUhAG|F
zes~iod!V@h;o{#1As?V{Pqr8hW+6smeNQq-rsam8hh!=W;S79%{aXGvo=xM#lBL>i
z5$#4yZ$frWiB3W>z!ruc2%_4)xAwFYpX8$$Lh#2cn*rt(3RmGH9*15ZAC`)Rhvu}#
zgnoc}rs?$YXfuU$v=cH-N%vbWtdZZ}1$vMJ#AYSd*t!eD5ZTih2!fufS2K$h3T5o>
zUsqN(*Opt|)e&rcDBiO}2jJC}mg4s=X}toQIb!kOj?4Sn!_((gdIBx{0A3{k<`*jj
zA##{+2$7HLbk;<ralwBX5q0Dk``L&+t0F5CE}Nn6D_tc>9G18F;Sb-E#WI?9hS1CI
zC|K7&+g{DSr12|B;C>RH%kM&gUeRhuK!Jbj7Ui;=E|#{lzN(D(ff<=+{5SN~ih~3<
z`yvS%r6QgX%0$sebR~1rCba%?BQTTC0g?>1WIKO$=iwwu4<D^FS)CKr(BhM+{JLfL
z#p^M{eU&>S0g%5iGEXN~`{c})HeN}V0K%w%Bn#Ir#JFe5yfngpXr9`5laO%9AL>G%
ze4Alwgr~<tK5o;sr}Zg@?Khc|=tOnIxCs{7>=(Xi#%Je+6M7}jmra@6@jq1&JD_EN
zK8m)Obe$L?nDeelggi}<6-~NZW%S~?WFyf8I&At$eH8@*<2lM6_?{9m8gOz>Z^mI(
z+oIoXf_O(h^V0?Zd^)%0aj_2QP!II=o$W<k8iJtG|JPN{|Mx1-<};#Hm0vW`zIV2e
z@AcG7tV=Q!rIJ)KR2fm}?+H9&$`~K}4}l3evn7w8e?W|0S(GHR{0!T9{&#k$sjPCy
z0>a9Tr)yKu)c5a$97}_`_yruim`)iSdmERgVyNd1G%oF{K*{&FFnnZ2S1l+pXt`Ow
z^YDtzsm--zl^4bNy1V=a4_Q4va)-?Y%B>{`y#)l%7mrSIf<^i@oR5MM48F1V&BWZN
z&@?#<2v*t0ew|iM2jbBRWs<|Y<W3h3bWW}NNeq|oRnJ@}9>ouERr2>GsTq`dA@Rw`
z^(mREUpv`9)?6*mxs%};RqouvsI4{)0x%)~E%P*n1?2*|4gwKvBg+?k`|#A?+uy8<
zWmBIKtfQt#OqZ`94_;)wp?WH8vUF%_Voo*kO$Ur*6iy4$@YaE!!M!{198Ne3$;HKo
zvCCFsh6n5M`z!qBZv)kK^Cxr%q8vMzYQMxLy@JjaNjGuxrU`6gmGWIa!&_22Z9iT{
zB$*0aqWql=TB%N~Q(`INQA2!p0Io@fA38JM_af<ADL9r>3m4Z`USd3K`Z=Sz{?v}b
zhRd^|Z{g$Sk#7z;e!L9Vr@aN1lF#p3l@&%}Ta+fM83OB2_sa~SG6H6Svto%Av#GIE
zfh_Qk&;Nh)KmRX(OR<aD62~KKL6s>WZ-rkyNW#y>tGX2;q>7lf`IUv&{JGnDXde=O
zSjleo;#_=OZ%jH&lx}{D7<%KldsEiyyq)Jk-%;FYwALkb>OYWYfMw8tNP7N8(qzhg
zb5lP%OUtI|lSbe8z{j*Y@fUX@+DOL8-WMouYz{Nqx*sR`!jy8^z&S&&?T>8X^cBM4
zXSsprBJ5w8Kd|%qn*)p+00ozaxCN%9OoI*z%I8H4*~k5j*@PKpQreON{Yd^HQTdhV
z1cg>X=YUVu7xP^m3YYYdTl_7q+L<dI5HK%ts_`DtJz)HgiPaxWZ+@88B?6+$dgnYb
z!20qQdR&AsB>r;rt9}KpW!aZ9PYuU!{}Uy0lCNR&j-)fTpXg=8U@dtza*+GSuI0d<
zj!I8Zpk3=tVDSV_I;zzrgc}(rl)yWc*iv0&J!cv7BwW&ePE1cuG><(Yw?b88%5-+4
zCm4!|&!#S&`}sk$dg0UA_jxM?7S})G&Y7rErVc>@n0wHnT9QE9p?UY=sKb}K%2tOg
z*1;bNZlUzcYnt-;z-brJ2IoLpV`bi;k>N&sfkqUfk#TUv0x(KwvJ*n)9tX}$or*9}
zEQF1o7APW)EK-~NaPo2Z`09KoI~ODCds&?`De8Fv+0!SlP<GhMEppCBKEu@^-gc<;
zkN5?X@e=BA7`(}Z&Vg1mZ;=rKve!=8U9D`|P6+O&d73jmnjD)lr9KSx2$t2X0f$}w
z`s+=5;ygPuNg1Fh$55h-mv;m&He)Y8dmA;!Wq^$&O=rBhxsP+E%-#JSn9N1&1cj|B
zLA#p;m`eV>19K|^nhkzsMpme2U*B7a6g7&G;7^UrvlcyqB6T2N_tWBr@{Bo`htZm@
zr~FN(E=MzEws5-Nn!hMvSlP`DoV01L+RuWITKV1%L4MK;2<VbEvs`?JWW;E$Q;WP!
z$R+`A+Ld>fP};B-Q&1VFQZ~o^K9baVA-GV(tw(b=`^XHGnlrsZl_80;8a5@1>sMW7
z%q^779>PS2ekcjVKPslBauNu%@IO`|aS^z}G~j;a$Rul(nawgc_AA=29#&q4=E)^~
zm{C1Ie<*rO{k~O2!wHwnzC=Wu=Y^bq8aDUyTFkbsB18^CZ3f|{e2aEGNkZK(|E;Aq
zwO0na-_%s$rdZz87<=ngszdBG>aQ@S1t#X@OOzYg{|kNqaja;X96tpZD~zon*NN(k
z9>nr}tNb<?q{a`{ySr=Xmhd=wZJTfFt>)%Yv;oh5J8jktwMJrg=l#iYv|HIU&wDOb
z9(s8$#wd;DE&f1Yw)OI~9L~Zviynt`%=IL$o#Y?u54)PG0}NaC+x489RVSX|t6J5?
z;5)CZy*Y^6%e=<B?j`($BXU}qsh1n=LR7{hEbud(AzZ)aer#1)kT28*KDuyMWFDRQ
zi#t!5tyM)JbS+PF8(iL}@Yx*fnMSH)BOh9dmZ^l}J@iwe1>W0yG?<lG*xAs%U#b?>
zkQ6aUb0gx1#90mZ5nDBj<{3%CU9-lq7Jg*<oe764OV$sJq^?5F>eGApgV`JWz=;^n
z9me#kbJFG?$bQMX(CpyK&qDl8`;2eE!f5l>kFh%El^=i|6V_P&*PA%gCElYY2Hhqw
zees?Qs<{=Sc1(G29kj}5l=h^BMROj07Q_A6B_~jV-Mk0aznWJ?v0Xigx`Xd(k;gMt
zzO{bX>!!JS%`Q_<>Rm=go5~thx%|9p`#NAlQQ+GF<4-vGuU!G))IR_O`-4^n62}xE
zg*a^?+ASI~ngYs8JY<a|IFq>UvOL($e<cKUXmr=$_uhF4=uATRZ@{oAWGxd2Bf>&X
zo*qRkbA`#a2old|ir=q`a$PSM{C$T`{fQL&GpdtKYW`)fpszuKr#^>God;>t`;3$;
zOcnkg7Y$vXTU|nZALf?q<M<ND&w3SDU7D6t;_}0H0PAV{SZ_i+j&FQj!`At)P0H#8
z&B~1-6nG2I`o)I)dJL=wmEZ*QhjI-!tmv`m{O{JeT1vCP{4dp`+Raz~MtU}udq5Uv
zmO@pSiC|&~-|f5w>d*=@$3D7!=j;MeXk|Ec<#;1zTgJ^?KNSZ=y`ly>`#qcBCF~9k
z3hA)HnjUcj_d-Tc&%~iWfEt`1nHpkW#$e7{-SOMcPk+F*K=|fp?8V#4OxHRt^{wL2
z7elp*(DES0kRph{)n-_*u6({T$kjJ1RjTvH%g<`;(j*N76ORD978r;`$U>%*qfR{-
zDeX_o!5!fEZ#}7@hY#tT%d7;b@;J1j=^MoHto-{m0h1~IeNEF-jx$MKxl8sh?tk95
zqn#Ct$WM59=Luiup@rPSh!fjcu|xy2Z=0*<c0;Xsa4N$zo;D81{4LM&-vP{>3>ITM
zj+JRUW=V4vku?>kB_Z>j0a?-UP|dG#q<&!0-Pjefcg0nR$Oo`)4)|L*^}g$t-wGmH
zQQB&y98p+JWBKY_{W*OT^<{d<7Pz6b2b~0(lJ`hiShQhlUOz&X-sF*6!azwKV`u8A
zzs8APuzg_y%9uYaz00ai!b^JsoG$MP8_}(3@##MhW|U&lZ`QxHSpjZMyN034jrt#R
znN%-^s;QkWCIt8#e8sndAlRv!)wyfUP^S2k=In60$N*c)V$2^PbMw2C6!aOqu)Z?~
z4urid7AVdr>Y|6BX|?<Jv#ZNYp6j-Qgq`J|)i_4ljuc|>?>W~*&V@M*Z}P^FS&?!?
zdUE{)2F<Zky{o*HgSByj!OwkPw@Mt!XdcOzdcUO0<7*t|1ykWm7TlxwSrbwzo9xgK
zrJgY%9GOd-wWVGo<F7)Q%QUqKe4*PHy?f6iec{Xxt@^ccD5}#gk7U)GFyR*txqDl!
z{;JLNkS|p?L}8g(o?;*58oXHSYdh}+C=S0|f62~GbZ74;XffAROu3hd)K@CsW$^nG
z-~T5etxj_URBlOMK=O!>;wH<IfYRJz?%9i6d;_w+bC6+8$r(_AgP}pgVKSo1&8V|-
zuBis_i5xPQeE$#^9;moq+eipfJNT>L*&8hRr8#_izANBM9v4$&`C<?3bJ{h*t|zTG
zw-)*lY_;aMO@HVJv95<yawW;gZeb^(6VJAI+A@9ezsaG!*#E0!3i$75-9X;A+&HO3
z8J1chCqQ(!BHt^89~0A?x@<3L%+vIw3|8@RG~R%+noHCV!(z^_TTNx`M^%S#OtO}I
z_`BK0e@H7FF&a)M8Uc8~!DIO}is%?y=tMP|dC9UgWsX7hQN5$inZ~+DZTz54?0szO
zx{^N|-%W5mLd_K%q*K52mP%_fYf3tki!abO>`|MU)%id46&wWZEhns|L$KsvVQ=wb
zw}XotSL!)4_Ipe3UM<@}^Vaepo073aRqPwz5o}iE3*PEq)`>Tl=I*=1m-IqEmw)7s
zNHLu#`WdRNTXgFD3M#RurmPz<+(q;wAb_#_{*h(AlRuL4%BcR<1drIO+}ylHNU3a?
zp!ke*i!{<*#m9Ov%fis@sz&Js^qI%NwdLeh9nn-B!EQqBGHt8|oIvvr1U@SJ{AX<o
zOJ%Ru7N7n=+^ZE6eRh7rOd*H$^3LWyHiUt6eSCk&_wbgH5sf@`P3lJ%J}T<=`)#Ya
z#};|P6M16Orad<9-=1d0x;kEnqx<mkea|Q869_l?qFws`pzW)|qI&dp2SG+cN>Wl%
zDW$tqN~EP@R2rnaOHx7x0YORuX$GVlq`PbAP#R{4VSr&g%m3c{I~U(`o^!P?c(_}$
z)|xdl@9+J+F(#sh@&PvLG`-BEKby+;pLy~qNJ<yXstiJ3g=eoZ{|u46{MJTrlP9fS
zK7jOx)H~`RU)LtRjuig%?b~fU5DnviT)gJB5IW5m5@*Z^LltbAnA5$<=K9ps|BdVQ
z9f_md?Q#HG_P`bvdze^}!2ZkDpvf}T`{X8%hC<kxo~bmXki~Di{lx^D20_$qeit2O
zwZTkgZnr^C`KgbpES^nej9xiBg%%_sv7HMmv;{b6%msRNqulLu0~RyuXnf&Vdsa98
zZBW}HhM}Yg7X$)txM`?zHfb(Ey074vH&eo>8UC#w?Fp6*kw!$N5fRHlD3@msC1_=o
zpDKX%gV?bnZ4YOqUCa02{9g-(Y>Lc=e7>!(h&^LV<CfQFpipujtM!23$l~o|?;w9V
ztUPF~TyVaZuLZAGv9!@CWG`6*xvz?>tpefwuoa9J^2O5LMZU?N81)3}3vC_oElWkh
z1_qF9D{xLg{exV$-vGMGxv;I!n<M2{%Qqj(k4<geK!Bt;%LT)U@>4+S0{>?*a~icF
z7wfsW%7hMc$TB|9r<qa`5PC>hp(hUt0f9&WPqall+1@EeseQ^xgub}UdTPZ0xV~af
z55g{)JH*Q}wjiYDLC?l@BKe-suEEuFl#s^c^V1i|MtzB{hdMZPeL&q0Gs+FkSB{ZF
z^5>QNgq;j*WsEQM+$|Lf0|-oN77Kb%FR3ns*Jr9n^62Zvi22^BK^xS&=_3Uv_Y)Ej
z$ORBmfC0FXI6u}cSv!~L8PC71)tKFR*4y}G@Vf8wkM5E#=>*0oM)xs=5Yozjf{|gv
zS&*I;mYG_fFV{ELSd-EyK^3g~L-s425O5`kWabQfV7I|%zTX|hGI>->pMD_QgZfH5
zYRda{#+&$TZEdpx-;WS<1^?O_ekkhR+3fs<p7Wp9!XRsVbDp_cp50?-kq?*rHty_+
z+rXYDxkH>;0rC{v&%`#%Df_slVcN;2E^OFLnW4Oey(PX&;s9LLkxSxriZVf#n7>R-
zYZ9K_+frs5-WhIOOPP03IZQU6xfTN&liKL{tKBP>;cA}2zj~`oDE3OeB&d-|NHtm1
z2D}N}2=qds97@aIVpV^;vo&T#8fGSUp2NPkzist?HBH|`O*3~vBl__UXM%)jTc6{4
zG*|Xh7^8SDX_4(enbCkcFxoGEG--{QVo#LfQY<yYYNVUO@$Qf3G3gp43Z1wxN}O?R
z$Y))jWaA7wj%59IJH`(}^ic!sTgfAVZ?0(_uD+M9ACoQE8%!;m7^xG~8-4Np;N#YD
zZ<IjGbxJ}8NRpyA%oZDOU*|+5z`pqc=uiyx!{oBr``wa2rgu4z1ic}4lNR8`5Yae3
zZxfD6qx%Eu_dL@)I7RFM(#5{_PGutgpSfC_8^uUi22`+Gxwl+2iL%4X{jywh=xU{w
z>DQfO9-lQw)wM=W#JMY9e1zgF9Jpu#i~8zs?T>p~OOwM(R@VUNr4o#U)dX8I+Gx+p
z(wGI?o2z*^iby!F+nKnkU>d$S`6V^ZpYA7wXo2e|6n*Co@0P9ZFoOe>i|oW6JNU#<
zItcqLeN)X->XadweUsEbs&offy;H|x+OW$r0P6V&;3+Jg9C~gtn+Ew;cTF6z!KhGc
z^1ZFPiCXl6cmXPIY)DyTF}9r4AtTR5+tC4@sqMyCAw00sX#Zkc{WJn+chLb>Z#o@-
zaLkc+jYGl4bSr?Ha7NWGX993%_phF5AiY}Jf80I1KoyKFzHPWB)NH=^K#*i$ebnxe
zl%4CE{SX%jN{=*09%C@J!Dvc=4|ZxT*<9=W7HqSFA<l@enxW9Fjd_vtaG<3<Ej!me
zejWpuqYjrL*f0Qyqsah01H=pP`L<%K5y2`AGyM5YSxkodsv0J5CG_wvIzX1=;L%mY
zP@D80P)<B4U~10!1B#!k17O(g<0CA{ZUsL|w;nrHm3*is^ynn}>CrEtY0?!kdd4WK
zBfZU>TCduoog9K+f!rq!s{GEZbt&tPWfPTehq7}S^ThWACGW{I5drtEd9(fUa=abE
zt%((b7zS`5lF{_*sKoW%3jIBY=kDw!%?dV2`LAU{-EfX$F?-mh89-mb6$s1{q*#%&
z;5dxK2Imd5g{Mfg-0g8K(WZr`OFl33XSvI*R9qR>I0^}I2`(-cW}~?;0YT_sUnG6U
z#ICW~u;@kIoSiRC)<-CnflhwhW)zpW49}YL;j#&Ee^&o>dtLz1BvH}KC~ZV)F1&Vb
z@tmlqrm2A#58B-pL|F+uzNxCo)0<E?H=O?9E`IbWHpt^Kh7gzl;^*cdBrT!=57x4n
z6&ym{P*3i_UpGHvfV#*CI-G`Idjr~n-12`vMLW)&W;z$ffMdA)nO>uR=|Z^z)K2#F
zqsqX!IBxdM4sI>%;mx-odV6LH>Bnyz^2|)S0<B!PSxqh3L*nzPK}n%w&&**ES?q%G
zN<o1Sk8-$rYsM^3^Uvz&FLocfkF->_>9(1EhDfK&S+z6LXk8dn6iX+$Kj^sf^+a)n
zpE;wI19<XwR=jg=a{h<81gdone*g4|GE&l4H@-1s2;*lU3zh^T+rp97EkQ&W@lW<?
z0t2PSeKM!ri^NB)@u6>}tYSJg#l`>Xv#)qgaipJ>^>7a`!HRf_iOcv+$&{bXwku<q
zR+skLa1gU!gCAv_9KWo~_PvAa-|}MEMVyWu(ENq{-3hlCppwwqB^cFkrJSctHL?4e
zf;@_DFG@=fC5Umhs2D6-{Q<E9Ql#*pdDEsT!U-m9TfWc%eV)FY;yyrBX}MXa=np=`
zMK~BD58I>i#~(XVZ>qnCw|GC#wd}a;0{QkcShO2oWL!c}6LYmDe9yWvJX6k4D5xU)
zA%ljldw30nG8e8~dM?R}K9mvi%SuQ6+!e2AK}V{}cW*1xlc%5PWEZGVnX13KV=@mx
z*6o5o-WKIP+!I5Sn$d&tej*MPJDp#TVp+>%{`xn<?_DtY9tyr4CA{dU^qWY?3iUmf
z=HdOP!vzVFh$li*T#V|VKu2Km+0Euh;-?gwtQtDdOqiK80EbDhR4(EIcv3)bqF(f0
zSFm>i^FeS}MqzinyT+_A-;2SOJt9{-Ua^Q2LVO$IvXVXtw<pS*`@io$J}BRam?_sN
zFSkuH<PH7qV8;Q~%DXO^DB{`+mQ9g6EG*ws-ISqGQ5#5BSC?^$XsVkLtbsmqPbjnp
z2~A~MaLIz#m;Lo7YKIt~skJxMx74SYOHhr0WuwCGtc7>tg?HwXP7VHtBI4PAEfN__
zn{2!n#XoViaj>F)cIO>84VNp(TS{5BYqDi?owL4p7ZB_O2=HZ{cWYe5kx7tfZ23}|
zsYsJ4u!hJcp^*rB6Xa^VMtiQk-@CWAIuc)7ZPX(`pklpuVKT09-eU}_U5N{Y-5Ff*
zXJ!!7nVL|m!ewPI{ZUoRB)BMXX=J=SUTAX8&xT|8rt!2jx|;99${0$D$M!v0)KBHs
zX96H^E;<<>++5lXS)%=n?^#HbpYb-quH-p7^|fgvDnRY0_B3#1_pZqM@k8VkcVH1f
z#OsZ>P6N8LXkIPQ<z3m^2)=dF_i@rHhapm_TwE}PabmX|Jds0Pbu?_<-6Bx>=?0eG
zDCAuXYtqa=7lt9ZvxO_q1jNgAp@QcE95%1p$pOE{6ou4u;%Myo)-(EVycTw^-i1*z
zxsEyrts5MyZvYVrr>a1+je+VlXU@Y00?QMX6Um&EAWJQ9@%DFtQRV4rV7>Ub7q)N=
zL8f&BU(sYc_p?(!uJSsv8^G8d=}YiB;PFWj;Cf>i;seSHNg5#?_<g?#Rh(a2`+-iI
zcSV(Pl@@fQ2y*is*qv;tN9E0m5aq;OE!WJs%d~z5GD<JrSm;tnfc&ckF!nhIn`L{V
z?X)d2RILZdp1<P(em_(&KZUKKAG`+`7ha<~ucds#d)j2DYY)MClc{iYSv=!re?nr&
z9eV9h27t>x+vJ3V{q@B}OEYfn-wigO5iSht^Y5ejGQcJ!DEJyUmp3tEGCaFYfC;2s
zR?b%4LAxE@x*^ik+K}Yvq+Y{qx*<OIt!M|#exoX*TAsL|Yc%bM71}{*t-R!O&dcWc
z{K3jnFHG4qd}olv4-2T|0Gac_%m!m{pJZG0SNki?`E!Dmor3)VhNNbwQP8W`@?7Cb
zXPaC$Yu*Pp<vwhEUMs9f{Nh3t-idCq1)9Kvoqlg>y$+njms?lf2-1WMRAF&|5mKV1
zEgMn|QFpk*FQ4#mspMN|2Z*#As>z<Q+irE&UAGS1_ucTBzj*aDYqr#9H7BE6PmdbG
z4OV#MW7^}HA>KI@a@IK%4%y<fwSh$XlXp||+&Sh=`k)fz`EQ3C{yD4ctkSP&as;Uf
z;&FT$F#i=6s_<=igweB}IP()ids19LyCLq`^V>Un{}^U)x18@|hh;f|RPvHlTX<XC
zETFn^WrlC1O5mS1gi(wBonQLTDmH<R4aPD6=&S#(hwT{$jL!DLY2T$Ukqr|GQS0@J
z#(@B!FAs<jn9+p*5>}jx_9TjDHk>g3X_sLbKMn2Ur_U5DdneAGKq@}Q2>$lhScF#_
z9L2;6xZC};E3<ve)ibpgXguMg3B$vKg%a{0+{WK+FOZ}4CHC2mXZg?R_hXJMo08qD
zvqPhC=RqLcN{~S<wkM<Y5_wcz(7H78hBM=W4_DM+*yyd+(wVpy@UaQWC9V1kkI{>I
zH=i)CGEzDwJ+mlGK->NP=^P0<t@Yf!9n4=v_HwGnIvsV_d-F>6jF0lHWR|CytrB<y
zhhx);4Xe2`w=8He8a@}sfWSva4W)~GyGKLx-Qhk%s=^rbg5m{~YK%U`;x*}b#(820
z^X+rFr~#N7Y){mEY46DgD>c^P=qzt)E8S*xWSr|<w#+1OKE2ndS4h8=mvxeD;T%P7
ztiD$1jA_!;+T@_kBD@=ig?1ac_Zewj;^*n-&HHL!hW>!~@Vp{2gr6@dECFjMLuA6t
zA>}{u!#GCS4=3++Gu@x5@wKwGnC9cCv7@xiO(KMXi$?5#H-tw0n%if@+LZ*hJ{Ntj
zHrwt~()>qOaoGitBj6*|_j710(wP5yM(s`O?AM|f@vMtZ0!qaV4~aV{G&VGGr<SfJ
z=yHiY{;`bQssyZe6B}drdvNjw(N(9AZcd`+H+v*fCp<yGMU`O(+eF7O5dji^K*8Gp
z{uDWL?$C^?hum!;yHfh<dw1qn&(!kZIPR}f!Ec8;^6YQG>?Ge?0V<g16S{EyF<K+w
zcBzkt=}iiZTvWZ@#X#A>H1TCFcs^rq2WZ6|JAnNa3kztH1c}`Pz7jJ~@%PrsP!OD2
zxm%|8RrJ@FveOrLq6Sr;DRnCO@_hbe13J+Fzd!+Uz>Rja6$D89A2;X<Mw83^0nx#%
zYSBENq#ddlBF<7UDsBw*z2QzkL7GosTY+H1a@<9qJ?iRY4DPJkI@xx2Uw5}&&_w+~
zEWXbuy+waXm&^|W=6F!RAHc>)T)gRgMdafQ0BC?D1+MwHa}{50?J>v>ez#}R%=B54
zs-^tXV=vyP+fW>!*g`hEAjsq9$2K+y*j2QR1Rh=s*n2m@I$Q4p;6W|@o84=Jk6Z7R
zx$|T@aL|AI$Z(MD9qxcKfE~`4wd9rdL2#+#BuqO?0mp<P#G6T8{7i<Y{Tr}&1EvLi
zWZ%tgX)yph3;lD0G$}A-Ia+9%JoNoB{lvp7(^ZFrjaNLB(BubJ?`}QCURKI;h4%PI
z_(UBB0y7do{eLwR-7|R;Sk!0dVj(X0+#)5`rU(z<oR$7^9mt&jbsO!F!JQ`Y25Uj_
zYaj0FKL{*%Sue2kGQuKNbHxcv$KBrgj^SPhEDE~_ymN9Ce?Y6&Z5e<bvDIx8^#YM{
z*jwm!krA;|)AYH1Ojr#S$0MeTEf)04bkop~ml*qBuU0P0?Bn92-07zjj!LZgLW`&5
z*zBF1zgamT0yH2k=3JuFuva2ij}V$VEi+h3wDq-aBXsY?MB%skqoGE;xa6fZ5>2jF
zkk|RIoEqjXwXv8W+8@=EFvmmZvI^;)p3!ysFwsg51-1lbFd%sF%L+KXwzaRHl?6{6
zzB?)iBO02lC!lwvh%g%@Edqm?asiWMdf`FwrsE#7K4R_V$@5bvcxn)4iLOQ%ANS;t
z1n}kB!ZXQNcF6L{Mdo=J_Q<tAP~jC{D$s(bmI9O+&zIbIla@v+@&$kNcZ5ljxkK{X
ziD5^zno;Jo?bDog$NV0PzTZ=tnrrM;<#cRue~GtSY2Ftu;+(-qCZ*D+Tc<}e_A_Y>
zp!#7jj6nYP@bz*o<yid;8QDo5#iU50=<$c_UZXrO+H7xX&$E>JR1j@hHon5t20u(y
z-6<XGM+7cS#K;G5{TVz)Nwt8^_g1P)SNsBhM^^^aT-uY7d*FoUDJe;z{Vd@h6^<dQ
zLDSjnGr!TUh-n$jS5FfmZ@+tYxKiGRgmvJI_iy7S__;j6IWMQWbayCOg)jBEI{14M
zBgLT-Z2vw;C%82z-A8B7VU=j-!a%ca!(ZGg34#CH5-*cPG6OUXyLa=gjSLVDT2DUq
zkh+%gZ|1Pg=J`1o7nK>QXYo-<AmK(y;_oMVQ$`N>27Q-?2)1*$PpNO!mpQH`)L)$K
zRUCKAbx+Z+?~n*#qJ#e6pTBpXB+tg#(}r}xoM(A3E}fTFu#|B@y!VOwv;}w>@WI7<
zv@dA(XcjHT8yHToR`N^#u7etU;tKo+>?#JP00HBFL&pD$e>m%q8o5q&<qaiRp~HRZ
zJUiz=XYc|?B;rT!M5Og!T<&EUcBd21qfhqDI$-89=`Jlds;e*CLGG5a*uh6tnauJs
zAKFe^o()Lbh>LeZ79qNAvO*Ic?X_w*nHhddA6!<sGA3UX^dGt8x=-ckn&Yf5Yg{PL
zz4N@wzh`1H&tZ^Y`-Qj=bBCRLJPdF$PIL2Y1}{Q%TAUeo;BOk!4#;<lT`Fz1D1?6$
z2l%bO3Zh!a-kF98=g{n2G+7PT)oU*+*BO3s0BPL0B2!^&VAlfg_D~^%5A%wa2axw(
zPOWU{T2W^vv(vXJdO_t!5_KqvuMHA0V(l$c(!RN-yK@vKrhEq~?D3Ue)wkaOO`OVK
zb4g2s<N}Chuq-}5R)Lhxf;YtRps)@4-yXQm%MZ0pBnJA$|H?Z0F|s%5Ey{W5G36PN
z5fhc`yYYi&f(tE)Y(Jh@Y0*)>{wU}^*{He5tks;e@t#`9R0WkftKIS>`6N$5Dw?U#
zc5K|>O8{GY^HdNg+T4LCh9%{k_JQ@YFzqcpPuU(-@V-2=I>rZ0REj?C<$?M5zp9M+
z9%IoEd6{*K67=Y4>_Yz@$rEMWY>BZ*AdER$SdL-ms;#xu4)SKI!lIeT@<ciTS!rb|
zY!6~5&OF2P&D3atNB)2q<!o*sy;$)t-Lj0+2W(=lxsAXfMFm;EeR^2_@o8Lm9QpJ%
zF6nyI!5nNyR$|gKk!%lcGk>-_d8pH~<;$K#D_Fb|duz+hSK5MM%Xj<Gro?<Dz+$?r
z!pRCc+53<JP7!1n{xe~osUU{^T-uLt3wy`j8*bB+NF<u&$=7UZPi8z)xu$xrzhtMY
zf#nZ~ZJH2tbH4x<E@!{2baK3@QElh$&YL=`isv;UaQRk0Ozd^u^5<m3;A(7vZ()Er
zVs75aD*RB)&hd#-2b7b|?a2`dPUjKM-b37FtO$0@Z#%BVA8<S2zM8T~IM*(SJ<lSP
zEYeJ~7gZo66_*%MKzO4(=UN_Z9?F>kOT&WN)<~-aDVn6^tzMASguE$*`@kBpd(znr
z2Z--)?*Jdh=q6L-_@2+H&MZc*)b8!<P0K~X<yQRW$D3dD<e}y`;P6p_UX8?#<An=j
zgk~aNi{b{OuEs|r_8gnYo~(=Jyay|0G$0<P^nt#Y?q^w)I4AWz%nPOFa~RK32U(_>
zv-YYzA(1JLQvl}oAE-T?k0adLVj>mqqmz=m^>-U?(a9tc)oso?BYdfz>BRT7%H}W9
zS4&guAnXI#u%r6qRLC;~{;szF@82b9PtK3+7=Dau(b7)SXu_4plyK#j`V<bFgo7uB
z^Oh7!3fpuYqu{UDF*&?ixv2Ztys~{DDz|5Nws-Pj*^nTfwZ8RR01+#KaP}cAKAYMt
zGXZbPR`$dwmX=jLId|FLL2=;VaS(_x<W_4lifh_ws_$wvqn?6@K)A2LR_KI|+Azh0
zDOZZLH$-;uTkJL_aN~<6zK@#D9IO{z&c<7H1h;hmnJe={g|0Ifw>L;|T$--1d^<+u
z;k&$m=$Qm_9m)6FL!OK>ioSfShiYrK{7m9yp8M~0LwUWB5zekn^1)AdwwURB^7H_@
z)V<81o=;|}uTaXuyg%<p!q3Vb+(S*YeLEh2zj(TXFE&WAGRhbcNLcU=)rQ<P5)61D
zV6gzgtCd?X!2rlk9tN!SSEJ?7=gf$iAOe6&GXRvM$KA{rzcwBK=>RcIk);UDn_Df)
ze?URNle;+`+@*)+!iJi&{Q<R;1ta&+6|taeUF>~Wr`#DKlKBQ54v1B?fl^h2wVbQl
zK8aY20^o=(-2h1I|Mt1fh!h^Guh|_@NW81PwzLT6twT~l-0PkOvp8%36UEt7XEt!d
zZC+nw(@%eGjd{ZUO5A5)RP50&_wcy$&Wf;h0Sx^{yg}#HO<mPj)6e|ta(YSrr6hZS
zJZF#&rbji%!uK_Yew|*5-n7<<UYqw4=m<8&Spdt{IKX|kAJUYp?6l(tsQQ059dht0
zCU)ReCN4ScjSpYs9u}Gd%xSSaC(XwchFi%4oNZZ%`sj0yn}il+x2E_pI3K_MAb~e#
zqGMA2`xC!I_yIUqobfg&(9YDmXACTsD1E!w7N{jK+9pJ*ba6A%3Pv7>DG@NlUF<@>
z<Y#8PLo^M2q6{?W_fW1yrk4{-M3ino-_~L}c|c{LC{#AnBbS&NZKX*c4B+yde{lJI
zSRt0oN1)|45W=I7x?`qEvdfvIMB`s&O|)vu1<KYC>U)7%+gxe8zDYYLDpy+@{lIbF
zokr0+`Lq1Nz#S-+-W8BO^D{)|?LS20#b#mG96EW;J>SHfui$`&$ot(1>2ghMSu+Im
zF<;6}0t^Pz%0>VpA%lN_eDD<^`hS4@VYUF88(!GS$JKFVUGi=JMT*m##s>9b-kg>t
zTH+Ml-*p1eCz7Y;6XrD67JFi;%!GW<!QX}y$GWjAF7b(iLX>5%uB|AslsH?=h{0@8
z;5n|3k~_u#Oul_vErL0U4UrgrR*ugT*{F8sxw$+4<TiV9g6n{^<&Aof<=oD3C!^A-
zK3aSI%i)XPCZz-AI8ZwicZq+YSX<nBK(m`0M7JF9646@lJ>CCy#aAo2EK<18FV&>l
za54f0w>#jZgiEG8;Ofk10F3&KHO4cTq=Z6@D&wvEap*5%?Ax|hL`6JvRDe+a&iLSz
zav{9O{)c#{ODqk$|3>X!rZx=~f!6gzN~ul)bri?zHPbcQ$%7`D(fLNAPQ<S$?#$m|
zVcKj{E;kr%uRALAch2`Q(Fpr(cmQK>^pAeEZuJ*l`fa*1_?-d`i8PKtAxCS)Ta$7J
z`Glng3Td~O2c_96y0mzkbxMJX{-nL_VRK9(h^UCO5h9&PqVijhtDiaxQrHeE<yYyk
ziyeJJ2n=9knd=%;XEz=l7C-^rD{_F|(ekw$EqOo7X~vBJf^X>pdk1m;Ybny?OT6TY
zou!NsoP!<9o&ma)9Rfyo*Om}-2T44y?~149k2t+wjd8r{{hA;YqB2&r4|!v!uljx0
z%S%8=*sC(1aAZ0{3r7~CS8{2CTAAkGWTKjy^raqjmbH5T{^e~Js>;|;yqX0HR96f#
zHZn6cT<P?-lKyDg$vQ+?Ri&>Ceh_+~0LugDVRiE>!vK2AIh`-6Wh2u`HiRc@celJ*
z1aPOgfb7neZ|!^@i~FrYTJ^0{N9=Vi^{(u?O7p-~-4VHpf7esK!Rz+omgc%V#eWfk
z{s-RTY%te=JJoVxjO)IX+DYb%de<QWQw@(41Zv)-e=64g*imgGfk!^YaQQvz`jG=c
znNb^T<F2&bJ~yE2w=xh+0kiY}>frjW`EnafD`%!H&qN2ba8ov+lgfADy(uM&M#~D$
zC{>Flumy{?C-<v+nawsbRlwm^0k+ht)qJko*t55EgHd?COLFL8Ey-l1ZYKX4zJ@^Z
zltgDQiOnG_&;xMAIpDfJ_>rilI?K4ql2njlRJHfCXsp&JIQ(%%%_{ZIg<oKI-JHO(
z@(;t$rNous92b=<z~e@`dC{?d9F1kGnym@chfn$(MwCsM;luR|f242|j14c0)w*Es
z-lWxL3H3;tp<orehZYW!W32EfLD8uOoJRSr5TSX$Ggv9`V}%0{&a8k`izfE)e1u-V
zw)WFvG*|4d3G*YV7b)Nh4TZCd%})rLgT6A`Hi1bIbqmv}mC@c`{ONqp(iuNSYKc9h
zWhcJ22K)&j<ewjIg`NwCPM=v}_Sa$XW3A=}bfc@COgGHY(UCOip1>m0*6m6uOJMc6
zZ$N~w#_nsP%h$zzG^B#lUe$#?qqp#I<}Q77jz@7Pm$U#YiXEH7N+4_69!!XQaPV%5
zr~Xaf^Nxr0{cXbVPCYvCSG=w8&_u)(%61wfw;9M~G$+#hh)?)(cK2=d-8NOSNJWlM
zphi%3AucW*AdChi%u}0?gUceH%{~KJR^XT>?tV!hqMf<|Dx!-BslZ_+Elii&nZ&CR
ztJO}W>p6^e#w)eu3u=yq1E`4D3BwlX<7K=`5Wg>1a<BNP@>I>gh$}#ui)%C!bc!SI
z=@du03CWgYKCU=!djiPQ(kMo@%xcJ;-(9-Gmi5j67HW3&pw!>f5n27LK5WBqty%4T
zw#Zv*Jwx`_wBB70{|3131IN21RFWwK3LGKqtgZ<^6S?d}W@C@F(6EiUwuV%%53d^Q
z<NCV~;yJ7onL<GLJoPt8&PV0g$8flUE3em_{wcwR{sT*R+)#suxHJGDax3v}$#Z?T
z#d~t8!Uk^_3*uifT?JxlUe#9*Nyn6!{wzC$lCxXs0kH%i*z=o&<BD8##q7s1R^z3G
z7uL}Niw{hx*uB}fn1S3{CmOxIAr1Yz7b+r`1xVJbZ7t+QwA177zT0A_{Fk)J5pE^2
zAs{)zbJ+ag)t+2T+N}2IzMW_mJM<~(;Bdtk!$j@mvBuoHz$OYilu<huT}z+E(=BOR
zm0_LA(L}<b)hF(l8wa@5D*$_8D=sPs{m2usYCi25ud83z+LXO`kfdx+((`n1c-Rm(
z^V=rVjeB2Twn$s!bc<Qp^2(B`t1HLSm;9s{E@JuZd5|2A`NB07WH~X~xF@6`?Lf!Q
zSxTXKOOSVhyG};li|!(qw3?^0{oQ{0YqYivgN7d0%-aNcxtM@P6EIRwF2cVb2D+&Y
z-b%F#M$vuv1A0Ct&rJ3I`~EUml&gy9p2BdqCD;+=-!H4`CPlzf?!6A;?L)$1N)I*X
z3$dzBr9jj`dhpqEN}RLX?s60uF4BYm(h{Jr>$jR-Yy8vU5Q>C~iud8uPI`ay_Q-r0
zpX{33U90Sg`5AOVD9r3GC>BlMM+&g3*M_wq^FV<!(HL9Y4!QM|7PZB0N#<>U#k*%d
zlovp)_Bb>nD&+S}`jIF8^~-wO9M)kv#;3*GuTRDqPJE?paMu~K1NV1fg!>sK*gjc~
z&-P+Z9ZY&^Lc60pT+$t<dkc$9rHK6)wD9e2QUcUQrJsGxDNg<aih)_zp!usXglQd6
z`~2nMpz%5ToaTE+Bp^mFEBSLau7vRQdjkXe5bvOYi#hLAa9RA;iRPB+5T4aMG&=u{
za^)CqEy!mZBbXK&yd(21cngAMZA(L|t#z(FzGVj|XcBYlp!=0Zdwb!7=)}#q2EK2_
z&;R(rVMI4owe&-z5!3@#?z}|a#eL4=@pYF;p9|hb&8O%l+Ji3m_HTY+_CwH{h)W<|
zg6o=<e!|A{_MllOTijPqeN(vrDzDw~&7EK=_d|$!fYO|&TH*qI=hRrRJNsPPPxyw?
z$JA7gYYnPI*Xx@7n+bWkRm9-mme5>m#3jIFchAiP;Umt>$q+%wxm_+7k1Rk@qn?DO
z=Uj}Ds+*a9*|nGHu~#M=zw1jhfiMVaqvIncwFRhDsvUt4ciDS<8HAqS6(Hf+9@adD
zZEL~?J3IULA^8xnvy&%t#0pJwd$!t#^fFdEj-+~_tX9<{Tihz3HV1}+w>Q@#=AwAl
z8Z)F!N@HFk*~oqz-Z5}F4pZH31uW>l3om9{U_g<&-sA2Y9yRLwVMg|n9#CO(7tTi%
z^3CC|Et!*4*p@6Z6GOixeM_>8NoRjbXQXk5?JIt}n(Y1MB?v|mQTyv)OlGDx#X<N-
z<B(Jhw^7VFd$K$;zryuHbb4W&CTwSb;5FmN7sCMZxeW+sEdvz;-M{UQ+ct4TesMaQ
zMd>3Kl_ozLmORul48eDixor*u<Qz`)`$c5|Y*tj=#da<iI_eZvL~LBiHA$=VUx410
zD<8@1xb`ex(nA77#^#1qTasv2QJ0l}VNLAq4q21D!!dakmZH8*2!Y86Uv0Mm^~nGw
z@`=xu(u2RuUc&qPS5m0z8RtSEfUm*2IsIExl|#&UJ%tJduaI9(db^N3xlzy{e7!Zq
z9G$34{aumbvv(RnwpUp0y@Yg;eP+fWqA?`l>a##L-Pp<fmq2jQ4n-mzZf4QwLD6A2
zuxnc1SBojDSZ}OJS#OBwPU?>$jw4Y3SK+s7o&nSqy(U!uG23GAaQ7DU6(C9(XbL_k
zs1W@MMkPiI*Zx?M-UPXG4xiRg;fkNbU>16@BBls@iNd-5;JTTcaimz(f$Y@d`L<r0
zYuV%E;gz9wriM*nob&M{f(|RL`Ki*AIraAD$#WwAL&*&jhaqnYMPc>Yq^PHNg9`rV
zhX@Dw@%V52;Th08F?r~ao)@9oA^1c~b5pCU*hrm4<uPA@`-F}jJC+XivE9*sF?n-@
zx_j(VwV8mmp*Lgb@Q<+hT_}}OV1y5TyuW+l-mJHm(2%n`aS3whUW(;wAC?3~P+&E3
zd+k#W#d)io!PL2BRZG~Caes2J5go43Bib#7W}7+z09OICa=X=sMVX+V`>&9U)e8!O
zL#&JPKa~jGLrg-UV6JmmX;JPSC#UkmurKKw?*Y|f#MriWXVHQtci-(_HBd1{)mO=@
zxgyB=!>^uF0WYV@Gwg`T?UI_F8y@Q0u@lSh5WpL+2$6q4Q5R{V<`Uyv)!8c6COh&z
z^TnVBJews7V8*TceFUF=)YK43w(33X(s*T^7q0!I;iEF}0za5yp{4iGxDYDrd|L7V
z(#$bGTR)7RHqJGbL#ys!&1=NB@Mmf12f3IaYUnJG&RHvEJEYX3s-N<VL<-;<S3+b<
z(F+LtU}kiK1KhfnD)YzBDov_x#Yn9;FR0EL0UXw`jb#oR_Ravu)x=xqBX75$t^x<m
z%Je4$tDBx!;njmIiLY%6wa?Beaqf;Rw^ppyEIlJR*aS|r&%9>H;aaNjk)x-xP-s?F
zKkepRID`;8bCG7ieBaaGKjs&d$hLg)&O(gg^v}4b2@7vE8AsfLZ^QV@u)IF%Wk(#L
z`O}B%?(|g4^!+7+uG7F7w}OJ=OjqFWkNi##-X|sf7B|IuEIM4P74kT`;|=4xGK40`
z_`u!=+ka{PkQm=|Af(CragndZB~O6o{uiJ52*yxru7s^w<_@*e50E$}YSHQ;0_Zs_
z@ES{TE`xyzN5Bmi+(6CFhq}8Hv&V(9hq}q#S991m`S%N`B*^zM0pW#^+PgD7eZB2K
z5cZ@hV|y~givyMTfKTrp9tf&lCBF)`kt+S*9Jg|xdP<chOCnZ2PU(!>t#H}}&>W68
z0m=<o3v~Il%-!HV<J4-ajSoDO)AWa=xA!o6+sCiN1Y2A$-Qxf$S6B|tTqjnd4taYu
zhEIf!7plogFOZl%@iA@G2j2HxqE9Aq48nX=!Qb`-oxe23yP~(L$o#pdjuk=ahqG$}
z`BA>y%}ihO@Ii*kdZXI=bgF2IhxJJzZ<g@2e2)Mw0{PeAH5^P9Y^??(^BFRj^?;En
z+;02o&atTIPHwH7^x?^E7s9Pf?>>fesZ~5IXl#v{Ha=K)cB3^FNalM`A6+yQR%V*r
z!2~*9jbjdCld<gq1+;IrFC<)>A-&2=c7s^zB3H1#)nNP^ocwVIMv(?#f*LYaW?~)3
zA#=>60S5fd#rp#0FDq$Z$F2yLo(698w1)KZtW*JmckKp1*SlT3Ldq5v*=4iUCrloB
zXX$L<^A{v+5SWv)Z^D}&heB}wHnOX=I%v(ZwKe#=J5stRX+N#H{lNwrq}MC6zXXgM
z&;X-rVLCv<sP<;#P?FxxS}@-r{GzxyQ&jwTE-v<rbm3wRj**)aVYCkPTGegJSUrfY
z%On3rZNO1Idhl&?QN};Of0L*2IKZIHbY+b9g2HUjSe4#gY*?B7h6nx!^tQA)6iIJb
zty(}0E0BhfY*uAiNAr+Ru?P`g17RM*1|dQ4_7gSX_G{vvKcH_mu;BkE!1H77JdS8;
z?=Cl2Df?_}2RIOrr-7`{B7g404fow_FlUH3y)9=`)`PlO)$i^R2_n41fg$_m1?!Ly
zCOPEorX@x`;bt#7{b|dS^u}K4HdA1eFBl53#7Ls;)^k9H?2?RO^esWCtJh4v8c$T6
zy5?U8*~nyp0)oK@IWXZDQ$C3=RY~>8nt>Tj&r@zew}(uLB&Xa?Omj)KjYW>({xv_)
zlf912`8h_17O&pVpqey}`vWSF#~V+|v~05E52Odf8!~QmlBl0sCOu=jMW@owHa-`D
z0SaF|oT+)?X%VT6%q9}ndUz&F%&1R+Q`pG_3}7%In=q&?=e~Y6<K(ey3|BMRw-Fd3
z4!^B}4c6`~5&qAJ4Kje|RB2+M%LdpkK|t#m&9l*Tt&4i}2Lv@K0Xl}PAvff}gAI~2
z5CV+VT0YyIk~vXM6&!U5NP~MyI1l^x_FVkp2L+|#;u7+{Gx7|#+o|nm$*twY*!;*g
z@vT!}LT8pb`flpeF~(K_%X1LoG1^)RcE#u+cfAE@7yXq!Ae=_W)||esdCeL>3ksUy
zJ1Ls~ainaPcby(DnVFQWq$Eo^RvXB%?r5R~nV;(+sjq*q)@EY+`Gs4OaNYVx21LFT
zvBxEl+w%(ed9iE&V#gMrViEztY&$sM9!*F9fz`JjKj$xe(#avWs@n(_uwyTSJTH6x
z0$Oix8VI;tc1jp}DW<E)gyMR8>q*`?DUjoA{BVR4q$5>kJ9<DxX;~!>^2;zrK89nD
zHs;pv5&8a#uxX2>d;ZQ~_-Tb#j`l}!R4(Zo?99|3&^C4o3)ue9bIiz`^)m<shUO*_
zHCvtiX!Z#HT8!&+$sle*fr{5|cqWr;_~k^#&Fz~EXMSfafseKSVvG~#f}}6B(HdS)
z!5x>~sV9%R6(TK#rIRuNE-MXiiM3N0RQkZyYU|1o?k^A`*`>mc1%v!7%_%sfkB`Ni
zv@u*G0U$S7a}AQf;9GH8;C^Pr0DSd^818(7_B%gj2%ru(N2}-gNcQPL>4YPV_>QdQ
zpU2_cZ6OIcdS_9{=rHf91cZD(Mq3^_U?a_d?L7|gHr>Da@bgL$_CA2?`!0{0@tkHT
z%e~R3skm|3U7U6M+j5KMdp<2Rxt+u(`9hE<sUmJvh;;g>04Bg(m`kHK^?BX-kN}k)
z+}m=o+5U4CZFI%)p3GkEb!$~hengg7?|KL^Es1*vOX5&w;#kK)93hmJA&gYyN*ht}
z+kZ7Oo2s!n<#$7xS}m|>`XE#lqspG055QQ_o@>rM?AyyWu1fMPr$wi&%{B>G@tGQt
zCNbJUeTu@3_iW#=oGo=dghTu3Ba4oKj<esjJikd7!>)eQlipeAbLBSHxX|R~-r|kR
z)|l*Sr2oQ{E{t=QVc+E6^bG<R66m;P#hx?TGPk+?GaM<~r9uslu4<ji*5vJP&ht!m
zT9e<FV_n;xo3>ap+_KF#V)ati`40wv8N_U=?z{H*uIciNQ&aWtCaUZS3pN%M(C!QN
z%yIB=-Rh5^VpMe7)Q0~9eB^K1!=<{V)3{6<AR>lAmZWo2ov3RZzOVXyAWs=(o8R22
z%Eq1dn~hnlzg2v{$6r05^xAoO6pGfEu>NS=w4{Q>e_#|EcT3p~NBk-S)C0J(3R16@
zxscf(5jC-lukrc~q$39Q6H?C4vzp+>3eI#K58^F!mQ_F`!Q2>t#c*(l-mXvC$Pgt-
zh<nC^`%&S*;T-pwSF%#2FKu;g%e!F>RSsog9KRNu&F#(@S<V%^jEKmRI<dN>&ve%6
z6D^jh{e2I91ikmO$gu`~jb8v1t>xl+kx7<8cU^C{_-RB9FFViDBPW~-mYchQ4*%U1
z_TMizoQ>cb@J~LezNW}8rDGek9%>D&*AM(Z-j}$?+W|6i>8SSJ3gj7X5xW-MIoY38
zb>Cc;y&~t_uB(22PX)~WXk~tP4jeZdd<gS&0XROBXVTwiglkjY@O$y3TM>8q-FjNY
z)i$vUk=r#8TCmqo(|1?9JuFz-$E%_D&m5z0L5Xvf{1*?z2DqT_%d+pI{CX;?Z*n}T
z8tRfPqehgpx{GO9!c*U!ADY^{2l&Rq$rjZ@dmz0pbpt-@ne{DfI!{9PqZqaL7D->+
z+|;xdI27ZD7Ao|a>z-{|YI|5W!OujFTL3TR4JEzpHq^g7X1>e=Mf+dN>>3Bhot0i7
zdvl6!Vm+HoOcb@*evox#9rntPG4jOTR5@`5$L<r2usx%CFU7|D;=7s+ZtjS0+RuUS
z_2`s!JH7(wBe|NxG&cd=aPBZc0NKj+hRD5$@-tT4RRe_d$V5~CP`J>hNoKfTsF}S^
z<;0dsDx#`KSm4sd{E!YTl7zTGSux#5My8kBo3mY8dfnq2c2RKrMGfqUe|-=eO6u>E
zojHrA#94#TwA&OQx8o|LpIN|yZALtz`B_s}N9dENaCB&2J8E~1i72nb4(5Q;n8LWT
zLm=LRW>KdHv$Gz#ILOCx-N=#1G!Od*V3N!rVGQw9i}vP+cn@^bm~zIYc0m|Y?}y$~
zD_N^eLK)vN5y(2YBT%SoMe4__a$H$$3@`4>)LGk-#im=UQUNX5%PPNHAgJ7iROk2=
zR@DNiVpP9MSJeZW{<t+iVn9r<u5*)B!5O|O7bzWMnK;}a(ALZ5_UT=5pY|cp_u~P0
zzyo%XLa29uQ?<I+FS5DNC*mFS60`dzV{w@0!S(!uC`R$pc3{D114AoFJ2-Ce9C|>8
zY@v5pV^;_1vj{1=tNP)62@<fNf93oO9oM#U{|x55)~%}hLf_I@K~B%qz?naFt}dr@
zDpw6FcfzzdpKuz9Zi|nS(sg1=PxChKcF)OCwU}S%stHw;d)}`}Fnn)hq^IP68-z0t
z0w_i(Kj(){kvMI2Rz;DTeQq8yb{qnnC|96Fq$8KK`R}H9v~|wo3mu$(M5C7=jXp1R
zohnjENkO96<4~hdhk{Y|1iqsj(o3g@+Wzk10~<+<)*W%*93xi)tcio31fAY*yEl|@
zlMj3RRu3#(1p{<iLphROC%;U|e$vfDhFnV;PfETV2di%a7V{U$-SD_dpJUC+SfiKU
z;K)})OvhjFj!n-uLwk9~<!n~pO9qFW$159!cc;Rahbsd=27ba}ur)H>CSDbI*s$fA
zfu8H7xo756YMA}oh%^1@7ax0^(dT0s)qV5DYsw{!cJlKNRv8NKuGIRG_)>JKwqaF_
z1UR2@7hQpvnd<!xIY-DPM+W33^6ekcgwR~OK?oL}^DkJb_!m|JI=>>|mMkJ5QIDlf
zHnPN?O&N&MRNv|HQrsI2Oi|#PH`qB2mj*tP&${#_Ig`lLOkTlBKsL@BUgzv5>H$^P
zl(sRR8?D}k9(PgnmZ9|0r7ybAt{%WDcW%{v^>r>ZG+kL4tTJBgog8BG(9?znKQQN5
zTZJokt*{~UjVUv%e#OxUes*-Wbm?4_<3GoJ)Vw#(ZWEVuc>gHfx!krMZnl?N+Y);C
z3$&JE1?-+BFHbum)eMp}@qC_U%4Rf8fuDZ*IL@wGsC1=}cr$=pr(m@Jw`2r_lO13r
z)~|0YbZktd9va(;5?A3?G8N`t&3*)2a0-T4fpNYTz4bt~XKlW0b#3o|9@AZ>$=n%h
zdx$$EON$-%dz=tp*hD^gRI{+Xj8f%DF|%Uha<HH`!OiX8AA{A-jkOa7+bAL+@qD#6
z%I5c@=0Awk2$<0R=y7d~v6$itJsh(?52gZyglQ2(-<vBIOkRehzL&kk7f7mdbMN1z
z#Q6oO&FPA+{oBD2dp0k(;aH&jB|bd8ymMQ27rJn^d<dbMTQSIClI94;8{hSV#E2{!
zk>w4%DX9K_rxWZ8`2`{OiF(m6B|`8%l?LTC!sbili>q`k=(R0_l&5<Id|E00rZ%7o
z4|M;4*z5KFfGUUEuUHAu6>ZZuu>e=R*N&nC$gbxK^33uiDWkxtmc!y(Xo}H6SYN)w
zV=`+)<R`1wdrRjypDexUcFkya%`L9gmdC6R&$_3KHF+r<!vv8M`tUjOQhw#P`C(FF
ziXa~vj8H02x}nJR41jWWz}DXZ4utjIH_PdiI=&t3Fd)*3zJApd7P0u-iqTA2J*Qbi
zsDDfviEs>ju<n07a}zJEHRLV+WR;gZiOq`5uKa>!J3?xa#l;(7^{D_~Tc*d@)gxo(
z8$>(Ct1ShszRvR)<byx;vVNNSXqj?pz^9vA;tg~&^LC$xCgc8TbmrXTeb4jxo4DFr
z{6{!yNGCZTpSa<41?(7CLS_c&3WlW!1Y^T9AlTb|Sdju$(e2_!Atk;r-)tvjwY5;Q
zj6i}rBvtP$;`djdU<;n7PDca|e?SkadECj;+ces}L;Ey<be~nUY7O(%-Q65C4KsR9
z3K`z14<TzIa4lH2XMPiJ<h0RDYpJB=GLl^y=H7ok6688J8N@#AY=RO(tUEM^S+ze=
zd0BqiTdi43DqRF`V8ox;m8S{@E_M~TSelz_2v!QBi;+Oo!oo0A1v}LD23&7vI>zL)
z2?{k!i7|c<=r|CVZ5P@<qd{3WWgo8vM{+c082uuUezWD)XTPLU(kK%a0y+Ru;_RL?
z4kH%jPfdpyp5zWm$5FlVPoB(&-M*>b&R~uZ<str|lx#l%=wZni1B$;m_{EiEGOYo+
z%POD~Nj(?PlsEtwy#ts*dTWlH9@K_zH)V*w(J)q}<%UgXU@BqxZM=IliPyXVa)M?{
z0vR5v1Rrw*^f=)3(|7Z5>%aw|K;S_n5CsGx3gtq8h<6jf-WynHjC3_Rd|id5WhPn0
zpROCZtQJl-Hux+|S<7e39dHn1_IDpA&|D-3jRxP1oAag!{7N>`q$F!c?qZVdx+sie
z{&=D9rJWXY%zYlKke^D8vRxOo=Ls<uR_mHXc2~ZQh6f(|w&ke~QA8&t?{)e@lS9!N
zm>w`5`A20xz=jny{g@N|08y##?aBpzH)&L_qM7mY4q_yyCCGZ~>Tk?;m9+ZTmJjzG
z?z>AdmXvzDPu`KJVBReJ86EU+x5@sl;@S0~%u?3_O+i6ey^qz}_LeD`&vu2wDblbM
zUq=DVk}z{+g!sAP&>p!$+0c}#dc-m5$PP84t(KzIN`33g4%M*5-8Knc;NH@tYe2r^
zWbE=$qoA$u?S;_63g5k%!U*LWE{F`8%r7V~mbsa#l%AACm?np25>CN@iQh>oJ+BVo
z*rdi1rrhiuEpcb~8(DtB`}mRY#lJi6gb&>ryi@E?_TT}~=G)KJK5@jXfrJKN5;wVR
zH1n9JKPHd(Yv*f>4j3mmGjZ?nY;r5G<$SWpF-AY4OPZOe`;i$&a!2IM`_RJQQ!KKL
zK7UTAKiTe3_-j79%JqZGFVnz&=EN2R42Ras+>T1{N3sugh4T7KCgh{pT{M>=D@>YB
zcLQywjBGPJh<v^8*0$JJmZec}<3_AmuGxU*eqnT4gxD*am06!ydcLJo*#W@$&e$U1
z(XYi$0h#Wxdv=FBY!WQ(L#b@nD`!Ape0_9{Nime((Lz1LK>St0F5TxV5UEHpBvpTE
z$N>=3P7V+wpLjq;x%<05{%xsnIOklwc^-U}_zPI;%m2G&=s$L-|H6N1T_A-@Hx~^*
z)tD7=rc_uTDzdm{$=-KR<i0gZiSN;MYuJrcW62=h#{%Vti0_N$)xxMvd0L%{(ent$
zSbfHUmOLL;)^I6|yKxXg%{64TJ=m%1wY%Hm2XseKOX+3UJ$u54^WfFJRogSFHA$=C
z*`}<5o|?%nX(9Hc0eXP5?uf*5yCV%m9q9aZWc<ZF2uR}UoikDzpV%?uX^e%#dUN`N
z6T%g4!O?rPkD8u$(l_o;dmIl0h@;8Z9UkAeGpNKUPSrNP`)%^+N44RkDjmssfBGm6
zC|lCc;O_NX(-(h0)bBNU;~BijE^Q4zrt@lm-({R?M!2hiyT=HnDWsimO6{AIqQwP^
z*b=%aWFrB9z4PyL8JVfQVOD~&&zw8TcK*Watv7f??jpG+K=<9woPGS<KeY~Uj)-2Z
zYxPf<n6bxEk`a;RXIh5lxI*q?r|0geA^W<?M|hf^rViY%rXW{udH3|{?r<Mde)f3;
zg42!#-MKCrs~jJH_2A(=hbCSiAhcb&54!T+s6W_u0BRac(UFVYI(E=yZeOHGPBz`l
z(@5P`7ts>r80%NG)wLKLEz~P&e6t$8HJvf`nD~*+)?IW5`(=Dmm`}&yi2uz$@?@Np
zQlI2!6Gzl{Ax2a0Y^+GCmX_IH8n9d9vuv`Q9@u##lI=82LzVOPt^zSh@uPM1j~;37
zJxz|q+@MdT>zQc-^r!ky1Xr4F9cWO##yJClWK$Zic}94H83JUJ)L7X|-Jsl6IMXKm
z^Um5Sfcl+j8s~zubi<5Pxq-|cKKjK%LTf|nOMXmjW{68SgNrY!SU50Pc{W3D;@pG?
z6^+YM+o0-Y^W7^sZVjt<xrzL~;gIve{IP$GWy;*EhB%?Bw^XJj00$!v@&Xa=1zSnQ
zd_}K*%@<O=ug*B^Rw8gq@r1Fi@<@z3lWC9C9)wzD!Vc?ab0WP8%jeM<W)hEoJEZe%
z{WwMSZ&iO>d5cixqTQUW!;)Y?HE8dRWSAyQt;9_Ep7JfeqWb4cciyf_QEwK$bf2rQ
zpOB}$%83=e%e@2N%lMsQ<p#2#>REDa0DtWwvkzo5n-qabUDLm+oR-e?&9VA>72?Y^
znKocLKx8n1G`F(>%7nMGl~ewJYAe#*tGI}GKoSQ)+X~?sWd6$N-o{BURI-mFZgE-w
zrwW+vPDemqE3#<tx)J70iRloMO?@K2@$-^M3jF!xbI5NAPAhTttpK1$aJCt66cN;r
zf*tDNvnd<<5b~u<IqQ*c-}1S;?=|qWslo;Ncl=P7xEsVLOFqZlI-0!Deoumv(Vm#@
zI4}T~%dx#WgE>^rtg4J}0|j+<Xb;~qNn3jNYHrIjMQlJ8^rv*^U@$<tLp5aO;L7u&
z?quF~gy~3S^*6tgdxE3|Om{JW0{y@N(P!8;LvNg{*U}h%Fc>tl#p{}?FYWves^>dq
zkGlU5=e7^{5ZKJKH>>@Hb&xAozpGnq!G9$HY5$P`XaW*|34kUzS3Zz(rx@7S*Fk6X
z)Wo)xw7up!1*9)~gBa}1mlivka9K{HfxyQp-|Y$C9T{fdbG@X!#>bTnNrtz}+x}yc
zExC`XdXsBMBbg0zOPv+%H&-=9fX`*aU~^+Dyh0drbURlBg-!td`gbEw`S_lmQ5A9u
zi9K42x;bEHpiTnYgv1%V!!k8DB4D9nA0Ll9wzletOG#YfNNo9eU)L_7dDg+<((DgX
zyOhE<t^{e#R~}}+?)unKPZ&5&cvM#YwWcnX$Tlx>g`K>Q(w<6Nz5_%8!i_UX#i*eV
z*V{qpv}+RVly*zYbA!+DZ;o+K`su9c#5)0@!7LT%@Oe^xk<N32r%&`E6-qKVi|xvH
zsr|CE9;S{Sab$3v>(f!C04Rjk&ipoFWVEqD>Vs{xVgd`%;P$tN?3UU7XLzVw5}rOe
z9BI;x&&%sm785X)pJRm&qj8UEFG;sZTLU7H7VG-$h!p_JWbloq^G5K&*1<!`aUlxV
z<)IgW2?V$zsHmTi50YHnJ?-k8D<_k`$Lyf=DW)m!)<bm^7!0g7)JZjwoDshH4%GXu
zZ@jWilj?upIb#V?IFQdJ9fI?eqI1@U+HueT>GZV4S>%o5_)C1eDlzW)!#?f#Fo_DC
z5W@e(X$KfU?1OsbPSiXh*Th4>w8sK)mEafKZ7k@wL%rsAr_UI53?e5_p4@-MLOWRL
z>EyHiiTNim^0c(0G8!}H*1Ti3>M~R;AK_W$z0EH*RVr%)7;Zdoyo58HwP3^;_8xG|
z)~~^<pW824Tc;Gz&%X5`j5j+);raDmlF$!n(-w=@0|d%OWVe(6I3nmKmHE337Y0B_
zi^kty*6-#2XfBid^v@T#R~}hj<<hW=6x4ql3}|kM<AQg*{(uGz0AvHxuSW}ge}?da
z;4nU2-$=WF`B&nfUf};9?%p%1sjlr74uS@x2}qL?6cD5eh_paJrHKkQij*j*G-)Cd
zAP|ZmARR$Ls?uAK-a{{<bOGtTmxLN1#BcH3&-=Y+yywq%emv*L_k%HX494D@WbL))
zHRrtMHNgUxm{Kv>nN?o>t*{FiIe;qzwEjY}0h4kCaa>jVz`^gr^JM75`kQ(chW-1w
z9)*_OW5#ZGj>p?J7Be|F&GG@CkqvGZhFNj}H`6`2nc)DL-r5xhT_aJ+(tb`YQF&$H
zqvoLGEcLo_UPlRC$<1w9>;5k*J4qUGd~fmqoPz7WcXYks;?`4G=yDxz|6G)@hrkPF
z?ft~K55=%M#VldJdq9YhBjy5$C4;zKzqMRdmvJr2Zo#;u_SGx@0qJ=kKGvqf6~*0x
zts0iSo+QIMR_o>SVL{ii%vw3kP^8GqhFwlQrxgI6-CRikGy^VesbC4SzaUB0!nzW|
z8xkd(k#yIvXBx6hjM~y;Jd88$i=Kh_8ojP;e>uez5}uBx3<jxRr>_lxTn<seW~^AU
z1#RVUw0AQ_^b!Q7rphY{K8NtX5RgFe0CVKx?$8YGa>(`_-*mf2!ln3|(UPx}1ID>D
zMp0-}`Vr5cw)-8ho2#YU@7GJ=K~tNIH;NRS1h+q_HlMj96B0)3NMT@u{n>5|mzE@n
z=d*6r+i971dNA(z{qjg|PWF3(45?vCMZM+G@RVQro~G}5nrc`zCrxqQ^||mOKXeQ0
z-OxbA&D0ZoEJS?BKSsKW|2V`9FII`R={})9K7WkT26Gpo>}$~+F?4mZFGH9r4Rl7d
zHYm2k80+XgWm1-74<$Z4OCR1i5|62Q+j~agBGZimU1x8RjoxJqVG7%yWI41@DKY(7
zhNeX3<u9(8H36KlFxlI6<MkQy-D4))gy5YY!k}@|;m;+(il74Y%Pm!IPNr-1A)QI8
z7FQbD(A@49c3@$Mvu|LmM9hsaV6(YTbi3w%BwPEdYcMn-i;Mc@o$~gp&cg(=4&)WD
zv4=k-(1P{?n2S;KKRv?w>RTz=3TPH5yk)Lo|135-ue^<vYY`X0!?Q9*Cgk5mKM#-8
zO>%5Ync7{l{R3tavlSD=9lizbTED4Il%=~O6uFpt3-hjEev!{P*-Tl5CUoQqn^s(u
z`o)-!5o$aQ*a2mlk-IFq|6;aZ#D!@qX@Yn>+rs0h^{=%)n>9$mmYfRd7T&~p8LT7M
zFT2X!25c77+pk$`Qab<81}D_PfAYsosCCje7Rj=iCrgp5lWueU`@tEcP`CD_&Z`f;
ze4>(JiSUR}o8`r%dUfnwcW+7W%AbjI$TN?Cm-QMR&<n$wa6j$7$%EKsC%^VX`+orL
z&ms8DuS+y6t5tQ083N6LBU|$q88r3em=d19$#FLbg^TA-YQ9KEeY#aw6RpHoh6^_0
zq7*utr(I7)X5B)pCkuoUZnsLWS(u7kek|mDCu+`T)91h-0>x}}kx>4A;<E6po*!K%
zl@|f*Qh&f^uR>)+(jHnb{p^x`?aRRX_Al2bt&ARdj?u4{ge02~`nk7NsE<Fb!<A=h
zWejlyc>Ftk8S%!^Hj6nn7lk|JpU2KXX1Qlscffdrm;z$>O)wlO^0+o{xymxR`p1!7
z-^~TP*7fSZE@_#tI98-47~)o%+g)Teak00Q_G|h(Ff9Bw5yTUkJ0Us9gInz<eYP><
zaTcZJkErH^|0FQp(?*0k;;83iVZn6_RsyK8mAybCZK&;?l<-t4zcnk$omk-8XQU%m
zA1VJq{e`sj5q*bDqi<JL+%o=Xi!2cf$ze4_ua?j2Q7_7Ob1%Gaa`=<e*{C8}G~_^1
zm&kY({)LLhde3Pg#O|ftdW6D(hO|=`1$0S7m=b18(0#wd`R%>w(3|}0*6i+r^K*{H
z5>62uAtC!2R!QRcjq101RhMyTHU=UA8o^zA3JJn|Y$E?g`E=y|q7D!$&P-#PY@wQ6
zv1oI^l)$LgL0U;(BV3C_zrYs;{h>99k(X~5{sSEWfo>;N&cV?Pa_x`hEF*{6YMIKV
zXdSuCpGnZ-Bh_t>qVmrb(6Fm+-ST|2IBJ|_F5*2|;zKjL;!b@2uNwQZ_A;EM;h&GW
zQU{~1Kbk4pOb!jIKogWCz&1ittDD@DHU@}uC6n|fHqiA4e~_QS*3Q#aoqpqd0>Q7#
z(W(9qkrw~Hv5tpeyIKW8p3(O`UT`uA28HU6RTafL7AYt^sJEoVJ`6#36Ot^y%(Q(N
zF|c&XF}b{;prn7}ye9<A6E;Px<->bPjy(^q@Gw8B)IRl{VyHc`&{N|^g1Zlhp+=u*
zKA?q)(btT=iqMto;Y1L+jX!Rg%UCg{?Pt*nzc|*m5XaD`&BrvE?~+`+*odV=A)Dn6
z2g7G|VCN^)sr}W1Lc|Nva}`D~!7@U^8~fV2N5ou#X!JdT#qCXgke@M5_r0S&#vA4h
z@nT)VeJ2$Owd~w_`^baSFH4?2G|h6*TL}T0gD6StlA)!xL0*@wFVK~<aJ}!}ymMDf
zY}Tp6`r}k|WMofRt!>3|k{3J5w$Z)o+leQ;h@d|CQM}yfJdJ(c>zk(|x3uqvME2}r
zBjKK6(O@vAq*6%~&y88JHUToBdMMwkHChYDDlud5jB7b6)&%beh;<@4!&{~DH{5pG
z^zxyxxIy!XV9c7~w)lp_OxUe<suxB+oK&#X+!0>fw^f3UgnzE}VS>}SBIU#%cg`hw
z#|WLl$-*Ckn|(2eva}ZPsMJwPh*jzq7%C&TJZM+Byh4CJW3k$j;S-_?N`*wO$iaRd
z_CQBix4{Yj{|#zG42EIF2CYjjI<^REVi{_Bz~U4aZJUz*v?_)9Yt*(9W$OlX;%E{!
zFzBfKd+a(sq(k%_G8;I223AB_?bk@+)5dNtqeQ#>wfxA-l@%XPx9cegWeaWpawEqb
z-9#(YRY#o}6f7M%V%GU!%_^{%#mCLK|6P?gF>g}h)3-@vdwSF$jgW|v|AXOU(N-W(
zdN@fodHj0f>gG@?e%rd>Wn}rakgL+Vi^OQth1^*r)#?THU2g*BeXyx<jhMlM_k4<x
z$D8g`2#eLFmqetK`MZNMgRL~b<F%_T5RVwtWQ(qM6(XEu>VVbGF<ss&Ypw~q#vr6V
z^-6#g8YUCK&G@OnGd39VE!A2-o;UPG`f~5V+qaQoQWDPol(59-=v(-gakG=OV^;3n
zp9{4*FX=E*%tX3N&oL!wEq<2;QS`je@CMWL?;D8@6XF_nI(oGuMfZ9SZZm0%<iyFZ
zb_gtv)T#sn`Tj3t)bDP2*nZ<kl&^$k7z);tbDp1Y@C#jzyiXpx5F(948Qe9q3{VQ1
z{9}>=JEXvV5{6()5$4GUf>qe#fEPdyIcLz~^1rd;kVD9uJQxqktgY3HWdM!HiDT>A
zJym?mvHeFw%=_M$dy3r7#!czqO~jI<<xUWw_6Y*qOg>AZwZLySp0jdCZ;eg=>>ty*
z`Xa^~=%~U|5QTc~sMortfQ{#ACKwxcD*w!;W4Z#ge%~BnPJ!rph-<cJJI<zqSJF4W
zA@mPT?D)O@5I>PQPhrYpFrO@R3d|=%%Ymp-Ed()VIX^RCWkmFYkwyH(BW6vpSDdC=
z?N2qSn7VAYnq>Dg*S?8Gky!iCLzbhmjRiCJATfndFe~90G<BZ7>EUYk%v$#MZu9lQ
z!fdz8agCRRSB%0T9`W&D!d6$3mKUp`*lWeKsv7M}x&LB-p#v`-P3=Q~`vdgz`F{ZY
zeE<C<oy6Z$>-=rJRPk*lzgcdVl}Fmo;@(p5xtzXvUF)>t+EuO9L7bxK(k9H*L~hL!
zM0EC*9c$g1R5*zA>MzyJ;xeEt14Q?E+_WYUXgj~G@RP^eKt2>08meRI5d*J7E#g)`
zyvg##xnLPldvm2IPi+A_1p5<2*d8aRxp~o}?$5t{dvsa4O^D(%1rc%@;!f{Y-j+;h
zH6e$88g4ph-V^oXMZB`c>$*n*5oz)-^=?{HE^tTgs6->>y+)<0fUK3p(<r;Wp@%Z(
zpXDD-?(&9{C#MX2%}{_M!;^eI(+3uwv3Cd1wR;M9`fl~Xxu$J(+&1|N`>u_m@?$+Z
zezP*Es*kPvMAz?M#z}pSJ)ucyTx3WrMBRm7{0C*io0UNmjT2%lD{gRL?K<>=1@h7-
zSxtio9$vuzUb<wX7yD8O0^y1_eo~g5xwI$gkvsWh*eI1EvntL!R9l|&B_CZpslr?v
z`xRp?S{+w+9?iFHf7wap)1ppg$HWZns}tLBC7jd-nBKH9qR-T7ZeU6&!Sa1Zk^xg2
zxBEx2uI4F_bR6ZOKZwm)9+8c!9+X@c6Q?-)65K&mSFJLT6WbZle0@smS^K5B;N#4V
z0zsk!KB5y|#91;J`=CqrGx=`$&xy<hd);gxEM|+NJ*JMTKilQQ4N`QSWCCY{l{MI0
z%^g`)UqxxxpQ=QJCR4Ag2#;D%yiC6w=VAV9zU$)l$8Q#|KaSzX&76VBZhKp35O<Hn
z#osy2!=P=Nr1{+^?(<xfg-O&24bJ#SWI}Y+O%i<qRNM2#q?RX)wY;}dYx+wfTOFN1
zZ11t=qh69shH+VL)*tIw5&elPk`inJLe`A|=F*OtVKo=}v5`2tgXiQ)e6#WBk0Y`C
zVz?!05sZ4-0aL?j1Tk*(>EXd|xV~h67I19Jy0Nk5ep>I3(6D}NY5wvP)*O@7_{6X;
zuQuoN%0JPZm3f+&@|_Z-i3|_T?>y+B%NDrTap~NTAC8Ys+}uKYy|MZP%K&pXo^KcI
z*VDt~I2rCK--FI617*k#(Wb&YfHf}$%6+Qo_|La(R2SY!D<!L@23@WaAz=fI_~Hcp
zCloB6wl9H1S>O*3z&-j`R?vSuo=nDsa3(ZG^L+e9e&%xIkK3(7pRQ29R`fOu`Asj^
zrgHwb<kzBp_}#B&X1ATV+4c!q8%_zQ^#$-WB+--o_lYKgW#@CJl-jq}2Iu<AZbeQ+
z#&(1EM7t5hIqOg^X~pqdH?K56Q7Md-pn_rQ_Yz!XbI*%UZ~fv)g|K|9)tr0j)a}r;
zkaI?rJkBPc{1GqKh13bW$=MS`-xV^Kb}Xit)`@M;%s`AILYhK;wLkdqaxkQ02(6t;
zL^sN=Ti)%b`?J+D-**$o(KX&|$;qQ^QxG$Yq_x@4woFGwtJhXL-fnR4JRdeZH+_@Y
zb`}T<nhqaYbeKOpJJXv4oxsWxpROe{*9hMaPF^>@AK}6I5p=Eb&+OORNzMob@ej7<
zJ7<`2ynE;$cqeN8D35~Qn$mB0%8^04885lkokp~5iMKCzEf`xtJdLxzRDLA?j`Ptx
zN1?eRHI#?Y%?VvP`%mZ-%;eu-*yIEZ8?nC@oF8*6W{9H*Y|koj;6OPT*IciX3%nMp
zF?Yq8HGR!lrFUT3FyXZVHV|iH#6wmfhMN9~IbD<`q@?-ht2}Q)6o$nRL1`2X?5^*E
zWv(c`vvG84kmuh@Va!nx@#=O*hQmz}f&7C7*G4~lyzj)}eW&2ZsQ}Jr;U9(DN^j!3
z_ull8SZ9vzy4%^$K8Wg(QKRv`!FJ;+M?{-E6=5|zb^p}^b8ohjp5BxCfRYF%gDtCw
z5bRS6p4fg;d{*att%sVb=@-K7V=UDvw>Ei!*$|3!{FkXSF;k4^3u-UtJLi}+aS0Y$
z)x^9j6#1!YQL!#ShnGfy=sgfM=$D7;Y`P#m`^1*_Qg5Ebz!@+pMxDjmohPq;V1zyR
z8tfcAJJt5d%QU?Ab7Q571!2Y;Cmi-Jwj1{0^|fSHxVdRu;Go2iz@m=~<X>G}>zL<5
zF2vpp|Cf4u`)BK6HSi)w3?Jq!WM?yI@=OiXb=FgVZ+nI}F=cH0hy9E1mGPIBHn*h?
z(&&A)<@s^xE(nMU2NBVva;f%r35<ON6;GC@Qhn0N$+prwSE%_aI)rXEb#)Z1`rv|Y
zf>KNR@OScHHK#=q)lUd1w-@o8hR5q_b566UVRr>(!;~Hriz?qk_PKO3+bE_d2THuU
zR1=qRg!uEWnzQni)Kjev0rxixRan1d%96=|k>7)P@o9%KJ3R-~-RvK~r}C9kglS)N
zQ<mB^s|6I*BH)c}wSLR4t{yR-8#f#s`_Y&Gy^Q(bMJ3<KrR3YYq@vuDx~IAu?#fNa
zW_o6;XX|Do$C~zBB4@W^<@e9<2@46tnCWJ6EUFX(2=ae`XCMFKx?j;J>mh!-J*BER
zPU4b8>L>TB&mvI#{OJ$BdMi3`5f_`y<OVKmHoJc;*L?ecX{&|3h;IsmFo|EV0y5h!
z;?LrYD$#ZJrub3$-sKcu+8zqK&kEYAJh$f#@xOc$3}Qi*mA&`rs@ankXX|SZM}MTw
z#$Ki?$V{bjRM$aLp}B~Y`6RenrisvAlCQDrpXaaOLY1M_lw;iCcbK&gCmRqPaMoGs
zhet|A0TatrBW96wDaVZTyARYM?x)YUzX(IcY6)obKZ|Iu_ag?s_&r{l@N8q_q7O~I
zwTUq47v!#ym+;UI$j32ouPc9D2FPvGv0atmZ`HisuCkZI^OxQGZXarVvqXQt-gbRk
z%*VKA3U52pJu=jMVSKas=oh<Q+>mzv;PoOOHR_wKdC&hstcwgS%dRT*Kody1gqX>2
zv~`y4P~at&Lak>Hw1NdW9?$L8ckBuz(9KldeyFW_+B5s;{9j0i_;1Xbj3hbGw@&nb
z;#56o4`up!hL8XB%$VSsH|dwX#PBM)X{+Uk=Y>%SsSQ{EbLW-(qEPjGEJ;$P>Tc}o
z#{8OiU8CWj>FMPU{Z7E79yAEk7)5jj<_rul|KQ3%XZ%|b59RT7$?P=I&B8j<=4xHI
zEk$A}%}b<&>ZH<bW2GAj(L)e}g$3KC+&BQ41Efn;EM||iGAESb3$vgmy$4(Y?Rdm7
zg9`UwNJ9dQ^!SbRioTR}XjZ3(9dFY2gVyEr`{(XhI-zIn6}p^|vAa7Cff*6xRxeI*
z{)ofqi=Dix@w1R4Iyz6_!IlMH82);Y+be(alxG~l#4E<;O)6^)3;e>!-JP|_YVqIs
zV^G%^J}Ob4gsUPyuc9tiG&I1z>>g4aUDet`)8RX{*4977baE%q#;JlTRuB3a(<Jx=
z(=(1BL&C`twK8o8+8S}Wh9d`>2^;aWrNjG1DY%NyUMp(V(hBD1H@0E?KHE+qbnf{m
z*!nl{5f<M)`;*d^2{1MM9iOMjd|r<z=Kn(QX=ENuFR~Z0S&bnpBaKlyH^1X214_nW
z+7ho`O&fj%NhM83zz91>XoEjBSYWf=(b>a9&JpFN^&>yX{eg5SOi7*cRz6dq8v>)x
zQ@M<K&2xQk1B@10U{Q0vusQ60RWyQwY2&IRXfOfZzt3Nfqs4Aw<7PSoR5hl`L}+;1
z+*;g)TDk+w*aqu(t)J-U<`(!_d<~o_sY`l1$~^thVw3N>Dy0IH&v^i5gpI@4Epxs>
zp1r&<Jj6~_fD#aSO3vI)VYoLsg=xx`_!>@T>x#a<NITymTj<)cgsX)ojlt4A+L|~0
z;$fkrD_@s#uN^+^)w}uZn%k$&4>s8$xv<aYf;tp_gt_}2>@-Is(~-0SZeAQjwR!U`
zmGGA*-DS`2X#;Yp{5bkr1N`M|ivgDUsp*5>$8pZ>RNn!G3QH{jSt7_LBn|OH;|0Ma
zY+loJwH?<vBU9b0)Pav5wW3=oikKX=yE}Kg3k*78<)Rhg4OX1KQnlulQ;~E3`UB|h
zDuF*Bu9V9#yv0;v37%7TOJ+1ku!GVj`6>y}r(5n94$HRndeNe3#uDE!pdiKPj5hDP
z(q0Im6WMU?sNd~oJ@7+nc>Uhin&2W@SB{r2hy038NQ~I>JOE+ec6EFx#g!FNTO{ie
z%h>H1`I+$h{&B+z5Xh@2U7Kl~74rQAC&54IliQ3gQ^RzJok<Jpcyl1L`m<<Wo{#pk
ze15}t5A=n+9}OR4K)`@J-%u~2j4H2>$Sng1a{o%@jyv&?>Qpp6(OKpD3r=Ek#>(xj
ztRn#yRV)JG$aIk}DS>XN82y4Mn?cfvzFQHkmwdOb=qB^epoQG#dv9X|UNiFfDkdz3
zpsB+MMMYBjVEjVJh2y*B3IQ@JKoha=L%g8BH2R<{`2m6F_j=mvGqHc{-kI5iPmgg&
zFuqo3yW@gQKiR1N10IO~c}=+Po*y3Si`Mo`U#Sx(FcZB97T0@!ZV9ur(%4lS&We~M
zgl{@0NiCY4C92~KbXk{uC);{$wrTtXgYRD89DdEtx0towu_T&!T*?9Vo!-7rLRP(B
z`7onyt4$;ot)ZPh8Y+`!HP~h_<oM=Sz!Ef&RX~gAW!wxc&%+j#E4qv}%#9WY^T^4Z
zBx-#)80fhgKj+0o5+=TYhVJBs%C$wT8kF^Y2o#Gl9O$@&{^9j7=>*A{Fs}Rl^pY$T
z@_EL9AgpX71MV2iApGA(A!clC<o`knI)OWu%ufgh$sS@-Amq8QKkYMHsKjMzLk1ei
zUd|yboFs^~p|SXe`k+%=<9NG?zOlIM%sr@==xq2%Pc$L<KGhtpt)p_Cs%-#KCz7Dk
z(9DdOaejGnZLrCvs+?RKE(^a(^fNB(&xq4${Ujs8rUy)|Hhwn<oBp!L>aVUL&n<r%
z%1e*skYG*l8St58B_P-J=Ii8En}pt#kr2JJ1-es5MQDAz+XFSl+N{|LfW-EoEii9I
zSUCFO<(k9Lg+vcUkumCzT17E>Il~TBZ?m|a+6tUC>tV&9MOu>8$TmD@(uVm@c^usX
z;m})t-<WF>oZAXymOOvN{%Zv(bOX;*Bbld;?a2ZzW7zZ~vzyGA$o2jNZdhtMem@K@
zHKWv4(o;IFNS-%v3!ZR!OS|fWJnPkq*m?p7@*NVWP$}Yu$8molbLB(}a@Ad+FbaVF
zw<lnED&%{#C9&d<j|^A6`1-{hSZ{JzRy=lOyrCv5_a6RYx|>D!nhI<%PnI)PZ%eRb
zWi98PwCZ|1Fk7_FlSPR*fhbwy$d12|j8Id_wW1a;1(XMsxBeUqvtvyUao=n~_`~QH
z30>K4Ij9-&v*fLt&dSk0-{pmG%b?r%odmy+!kv6Tj|v5u8wI^dQ0@ZB3#^%erJ|^}
zIcXh%tVOh}54$wE@hTN<oiH{NFszv)Rsk^z#Pvo?5cCKT>VgB&^ohA&lkpfO$Wa>c
zR+gKwY)lS#NSu9KIY4q{`+#x2Qc+PHWjLG`4zOvMhIW^dWnUG}8^J)LNkN6}Kf>dJ
zzY*9{&tHNI1O>7Nb%<mpOfk4yOw`>Yrn!Qk1naKp<NUWW>x;7PLTm4u(dI#F_19&>
zd-R}VJO5<|{Xal|fN%bPo2X8>XD3;%0o81e*Exz5qL3@aX5;s&Evu3q@3_&x@*jWo
zz7Cg}5l1#2rn!gtWSJP6rgET+xdNO6csM+`{ft3!fGdUoH7flL`Z%qy=B37#D2h{e
zPa5-PH{Y<f^{P^&0SaIAf20nu5={RmKJr%a<`GVL^Y2Y-o-BmIXcFXnN=@aOosU1=
zLDh6f$gOSNPHfoj_oZl<8oF<G^GF}ebu$Dxsj-89A&0V7l+C#52P8qTy0RTWc<M!<
z<?MIr`cJU^_M(46ywSR1v}4WB(%^BvIxbDf@gwhhf#~uKd2Dyi(a@eAXdDww`wJQJ
z2tU3sK+*gc(sX$#c}K<uxj!x1i`K@sH3I1x3VpXLS}FEMSphv?7iE`E7xz*+3Z)?N
z=V99~pD?Kv*6HLjC03#)Tc3uDuv(i_(QU|<1Mv}#2Dg+0R@tN@=p6Gjo)V|Ljc)^~
zMf^?cIGyqjS34Hml3vc3)NLHaeQ*NE0zeBh+s(~C9%L<t0NnbVmw|K-G-rsPpji(z
zTqG?IyMC0_{g=jpF9S)vS}EScOQ@F|Ds>eFRX{Dlk~!sWoOE411ZYrdsE3!WC$C0o
zdX=92wt=yz@Vom^^Pj34y4ebdP@8{G;c>*UdvRJLS+3a{E8m-$>GRO0i`~Dv^-!Mf
zY!Wpeqoe9#mh$|0itSej&?!@N@l&(=hZMe|Ks{z9Fq9#+B=q{IR$pGpxC%m~tBb14
zSGlFEHJDXHTVRT!W1#>{MmF;#_)XhK6O?od-#@IZD5?DZ)=J{yTo&5lL8dRm?|qKL
z#U85^#xLu*7mSSQ&ANy{7X@?u)3Z!pH%zu<Mpa6eaI9X8OUdWnBwp&W-nh2VH`G}c
z%Yu7_f~3+pdaY~^1@C|mvV|kbKboq+8=Mj~YFT5d1)MWmk#bq^0!@DYkfBtxAyMHg
zNxi%;c41{zH~Ah?<jS*yZ8nu@`ZhfV_g(BgaIiY}gMJoWh6WF+@+TYVF7{y;wCHSn
z)uAzC3Y^Q4zdW#kaP6-s^AS#$^#|Q|12nc~ZR2j7;j^SvM&-K`SNA{*Xyj<qU&tP#
z!^tYz>H0ztYH&~7VbkXo?vs#Md>Yxp^CR}HGK6#0)Xqyk_gua2O$IgA!}NcKwVrUf
z|DhiaFA>69pJ`$71;XWA+f2cx^B=livGYBja*>}E8gn01i6rROAGW4DJj)<{(oNc$
zZg#rG?Yrhq5jn88E4vLl(MklBfDxk*jyY7c!7J$!58of}lHXA5L)U4haPJS-7d~a9
zCe1t<b~Nm!;}sw>8fj(YX}ta&?Ub9IqFp2=eD~B`omgMXr_hDp_V4OjUI^rTkO_PI
z`W$bHBcrm?klc|CH|arNA&T7k;V(q9+l&VoliOWEF@pUJs)viJu(Uc6Lgst;WxKY(
zT@O>`OseK$ISJ?k$YSkFtK{=&)&iRd>k)&%H3<hxCG#Ld-$l;)c5VVvjQl4P#=0&w
zdHeleh=z`t^3*qA)e7?mM-v{i)N;aPiJ+PAo6V*+^u~Hsbw#yW?<DqqRz#{DyyoCd
z*lBxGk5673nnp4ZqnZrJJp-F&U5}IGUp}X)z4s_Fg_j!2iGBUm8=gG#zP_6Qtv1we
zR99X1y>6@db!oALAf>;+Re3(AR#E;2TrtL<Br{`7|J>$6;!fFvJba51pW3B_T%Hsb
zUy4?zHre$ZJ!ll^E|F!r8;0^P&8@CJ;#y{UVDajQzUjBKT2CESAx>ew#c-A*d7^RS
z#O0zv|GjjK@|U30u(yxZBhEeU>Y-@>`Z?$%QHMC22@w|kYee0^ktSQM#FfqR(=Z+L
zxxgub;SY36mhrs`Kk8<4)#v^a$#aGiNa;_87HAoYA;_%!s?VrO48dXjmKt$+=6=HL
z+T`I~)@Nv7;)$xA=EE%Uvu~#`)__WwvN4Ih&KPmdwoG5{PSVq04W+U@0gATUmH~B_
zh?>oZpZ2``u)`A*Ytqyf&J|fJs(hCDXx*bbWnOZ`<OC9)4p5FYE;@U_pZaTB#fcmP
zBGN#shXH|c4~_2*Mc;QS7&BYBE@!~U88!Yfw~%9eo6;}j&1^8HDcPI%0v-`GJ!Iz~
zOS@XZ`cCxq4>swh({*~%j#`Uhw!>K|47$xYrkab>iCW8t8#oI~_Z*&7*G^$&X^8yp
zVwtjyE-eZ`&N`8rsQ#vr`NZ9y)(h9}%l%{rjY+Q}4~7~*z^EIsVWvO%AW$20(NDLv
zAx|E-n;Zqep%!s=?=`+Wm~_`8Uv@A3neuGh%dqj^k!eC=lUD^S3l#Okn<Tv_F5nph
zKN*));p%XqXH|#9Qf?TszNAp?5YbvxwjBaQ+-I*VBn6_5WSEOmqI1iZRaIi=@z<F?
z1y<i`QcI;iUzw|g^08<tk$wspP-PIb;UWn|#gvZgPjH`hGq}o+&TCNR>~cy{?XJMa
zsEA=Y9uhurV_r>Qmi7xq_W;(48cGM}T5TIS*ydl%83UPFXEssa)1vjk@`aCEaRqk#
zk%j`H+fwt|RM82RwJw-qkT{^iV@1A*8-R1o$PKv&r&6Ok&OFr+5)i(hFh#c=!uaGs
zZpLSvAk&hm(3OwEY{?9L^=e0K8BT+@l@qjqLp*K4_l(yM%I?mzspBs!uk~J!y=;D)
z+Y(BNEwv&Vv?g&AEXIW1%YL}4wq7rtC;o#nftgW$2B(Mywk>{KAZCr<=`ZBlg<9J^
zz#v4Gg8w|N#O=s0QLf}$%!K>CvAZ{`J3!LTZ1Auj6%4Et<f@RTdhW`T`>?T23cvq>
zZHnluUpk=DrTX#4ScY?mU$;CW#><`<P1X=MAlHF%zvq{S!x}K73F?9l>_Rczp~_jv
z(F35Eb(h6r{MjDd@{gK)9H322W|jI;iTEH)vYp6HA^^QCA8tL(qoam%kPxl-kyq`?
zk?^|UjE1A5Q)E1Z^z&;u+;ZB}pkacqvUj)B`Sh3&7T;buJ+S|9)$q;#3cdcfe;bS<
zzlYWoqOa6OJ$j)Ns&CCNMEt19Jm)3SgvUkzBc`W8>xJX8@KoctjRgdrW2pN3+P6w<
zuLFemWT+vDss9ZZ+azn6+LIen3kvheJVn~_DGY&cvI;gX7MI2^tULl;3f2O#H~rNq
z1%uWgoY}TNfIEO!<qYam|1tYt$nCEw^fH7`(^G#Tam9BL8i;9AJ8a;a4_L;~*OT|l
zqP3(r$%6vChH3-RqZ|r5EnZIzdU@IYLZYPzuE+isKztnpN{j(X!r^@a@SPZC0LXKf
z%=&#sng0ecZptXnY|7N?@Z+;@6=l&q^G3Q^{HYsP)xB&8NM~}pCRvQcy?A^bvmi`K
zN#@jS1baRY0Crp8!C1^%rhmb15Ur=I)@Cc%;x!Vvw8dMR_JzdOJhhdf1*ctPXC_M*
zNs-mbhsS;g@sJhFv@;qGP!RSWX_O?Y(^|*TQzt|eU1N8S+gVL1XWQ2=dYvC0Gq9tq
z3k|c~ww2vqTq_3}rA(8b!op&>?oON)WeA|g?&Hgkp)R;CN%E!Iwnko_{u;sQY92=Q
zps$6+H@*j$H%IBnFnh!bBx7xczy3q67ndG8+{N#G0@K_knqF8iQ7pt1QV2oxgRqHz
zA$H*hLL;zae<(RP%3GPA%r4i|eRyGK^B1nMdg(=&Rg|IW^PmA~MeYD-gxZGaRuxX8
z4$4j9!-pUHeI-Al80-D7K8JRKT+kZ_+F|j{Ko>4gmEQ>(MR$hB+7cM%@U~`k`p1<)
zlntu9jH8%6e~5=mdCMlmWK5mPOr-x#RH6FmAHLf4b0Lh`1uHmFQINKe+`7Dk*qw%w
zKRkg9=jom&=HA4o=gFyD8dxs@vtA{x$9=E;o&Zr;Hg?&PT(1Rhs1P_mSNj}}=>U(Z
z4`V~qWSKQsTh4eY>I98xUU1-d=7n@{6g&j%fUA!zai@Bd+t%)@iN>9)CITtfatAW{
z7*vH6C_{GV3>qKr$gO{M2C@e|_b%|l?tRlclN_MJPU5H{FeFo{D13Stx$xLnLGY2$
zDE#6@ip5N-3x_i}(bHbpJ^zWkq5t)o&_wh@;>~Gn5WBwXlRvfPHD$V2?8VX^y$u(5
zXZNGtTU%MCW1*6sf6p9$Ju!RKOu4Jly-*%gJT|5|-x2n!OjiDo$!r;P(0y8`n8pMg
zCJx4Uf_kZ<ck}wSt2(d#LTci#I+ql*9jLupq>Nqt<+kYyr^ib>uv^<^pvwM2?$^ms
zo=zO|xfH9m|0r4TwQ~r4lMrRX)HbpVONtQ}CD-?2mJU{%;Y#2uIRi>;<bBMNxwlI1
zaSB`#-~VV}r?VTj>U!aNin-K{!9Mg>Yo@%2g%^c|pU=Q<0c?8ThfFOOvXi`qVVhiu
zkNU|w(YM~HDc~1Fbs*ZMp*kk*sIY8;SqHs;g67@i22ds~b=bC_j11<_@B<lg)QBDP
z9*|QGU@LcKh}T7wK|}fVy$Qj={55-_n*vVK#ozUuD*UN@61?xp_=N>zqiko&ebG9(
zP3cQXd}UR*&n(<ZJhW5KXJ32YlXe;kKH%R2M*67!EO2I(;>aIZ8%aXlBykwj0=xGs
znH+u_a<ttFV}f6q8m$mV>*N~`k9EW1Bf)*{%sQQ=fkdbZ6)>jeOt2Hvr?BT&2+Qtl
zI6Idtt-OviZ8xSh?v4MDR_&-?ge0}rDIRBox|z1K8G8K1;aDiF68TPuERp#Ru_g&x
z9KXeSt8@{=@uH1MDg>5Q_XN<3Lki3mHM$)vI08BC_w6<Csym*3gx6f#oo1v)g^fNb
zPS&rfE>%i3?ud+p6;Cs1^OtFJ?<|`&gHv9R7=&kU^im`UR4shFo^9zVE%NhKhPu(&
zkTVy1E9=3fL=X@(1Rw5OqkX2*LM7MD&7Ltz^0J23Frk7<y`xoA=;TsBIyy0MqJI-U
zHQ=Cvg!%Ed69DpURpx7_lY`~uc`@gJXY(q9{7_Err?UxPCQp(0@$sFl;+|9C=f+nR
zf0xj8@QhA1XU<!y&1c->f)-GE4E>tq!ORt+s>!&yeC!_wvJN%?v%dNlqWvO;5tjw(
zcS1OaS631}K{IgEC2m93L-a!jUB9OFHS_Bc3B}KP0|mC+T5S!cGaDGsl2isK4f@gs
z>^FbDl%K1O;ram$7-2g9JWSeoJ<N6>D}|AbV7WFw)y<3wI5JFHG%T*DiTEW35*zp>
z#)=`m-#Hb8l6Ff$v5@Ic=Cx&6iN949`g<+%Ku2&bi~Jy|`o`DYS%vO<QeB=ZpxQN<
zVfXc-4!^L{W=webEq%p7;^>2p<~aB3bF>!wlD^s<Twyf)VKHM~-F56li}aO|n9Zi+
z*Na92`WLzL-~5HZrJ_lq^O<01m0!;rY*@_+_9zF~<KjQq!(s0lSd8EQ8+(k@s2yEH
z2^kJzWnm?&`Y#u5imKT<afH&WmcM!Brl72Ac2&M0cNz(xBRxs3uBNt^mprqzAVquE
zL!?^3jVfpTF^9kv1uDE~-#<DmgGXx;T@TzkGe0@#+OUK9-3jBXO)Uw}a~TZec^IE9
zo85Ie5+kyCypIwc_ZsJzQ)m6DZZK>glKeA%KS!8y5lkt3p-TnApPRu@7T~hs{6^+{
z*Z|vMCBVp3YM`Oh7U<Rg|NOP=e8h|7Qw>hsQ4_2C)R~vB*qfi7XnXaPe9NE~Y)AWk
zaLyEzq^>Q3e!1|E263BH!|cwRv9{P&UF4Ic6MFmy0r`FhWPuXhkJJaBZRbg{JP%}G
zq&lazrUnUjdQzeX?Oxn_jMtU_jn)L?U_d5%&L(nt(vPHs&(y=$_I-xS&0i~b?v~(=
z43I8<Gj=w6%EG73W=C-gH}Zk3g$%;1kNyi)G134@UuX(?%s4W%fH>f-i@;N<FoC(`
zI>w==$dW{ji|@uftF&DmV9gtjDzEScl+X}{n1fE_tzk^Zrxn=VdDx+%B4nRH=$<Y=
zrq}-q*`IMBUn(KF!QR2G@r@g^rnsOQDfp?y_1=#1PRfm{1lVWZQxR^n&~&fY2EB5K
ztrvTkz4M4e<tz%s^jRY1U&zgk(t~ia1~$iwce(0v_|z|v&c=cwt88jpYu5dP;7KXF
z@eYCL4c*(ME-Vi70()@Li|cUgFJ$>n3B}6qWbCS<>-&-qi4UJt2Y2h1oA&*>(pYd0
zIuNA_d94l64@EJqX9GM{tn*JN(`9$a8tC(OM_&h^EMt^R0)<Jlfn_f0i+Uei>9{s)
z<!#Rr`N{2akA005;>e?rjV=wTgaw?G?Gh#sv1K^{p12STPN5!X4+3AS0^-j5?Ev9=
z8Keh`ID}Lo{op!ylFqEMx?PDXVN4A}^-bhgW6AR~^bu{}+YzB&u0y>rR?OVsiKICM
zOw~PtDbP$`6$mp0VWa4tM+B4{Jv1P*GQo~#_g%|K+8foyAD1{aIb7aac@)BD^1MzH
zZW)r~J+zSM$wDY6-|e4vJpwev*DT5<VX^{R&h-)RmC{abx+Y;*+?}&K?_45TXP>q=
zUG!<^A267I;T4U(2zHW(Q=3OrPcTa{DGXI7MvR)`J#YNJ&n}k1@D~AkKl+aEh<RDL
zUPqZ~*O@s=+vn%LA+8+N<_@gRJB)rIKTp=(nLI8}p)-}KpM)geFS$RP_`tj`g-MrB
zYcg`H*p1d-nP%*&QdsFOV-WoAboVK%y(rsHc<We5ehq~m_cZG9W2yv-Wn^6Gj_x`j
zBrgp$F&>@9k$r$n%lLwjJ3YTX$5uubDwDahdf`m)qtik4ozlR5HuD{0rOSX;`Mq{6
z7aQoFH=?7?uW9G>iWsvc-BE5a@6P-d@6--PFAVh}OZN{a!Iw2G_7`$3g_g{NN7U<t
zzHJnSzsR1rvn`Td+PM~{-%Z`tZbI3=$(Zkf16=DnFD_#E9{UjedhPG{k6|-ik2Qy#
zw|IWkhivz&>>f-5=Wxa-of-#h4E6)$O5HjBM52Fzac~x&s6<7|J3BbR%}hS8jaooq
zH$=ym{%;6hDLv32uU<vQCh;4yx2uRjxntJsM;Aoh^0xYQI21;H@MAtQYsqs60F!T<
zH(UtMN_vr2kT_HO@Il>n@#&0|{PWE16e64<)QVF+7501#&VJqGk;$z`JJ=&ibpSpV
zvFAYF0O@^6_u*kUcv*p}^M0Kvg(Qg^cvB6&f?tD?SqFEyb<JsY#GWoPd{t1<({A8%
zIT3i9!fDaeD%blkMev6?HtrE$TuSo~OB_4+s0wr={P;qADxP~QbTM8zdWq>U1f_oA
zBgCQflPyLnz_u->V1{o@0+vi1;+1gHW(WY%I&^yT-F_1P_iDmMx$C!;ts9z$XG|B=
zLNnBvIznDkG%iMVO(;yhOJXNpN`m*lz)uF1@)#Hf)r=b3sTuK0OI{3kcBc?T_`I5$
zsp;G+{TI61?i&HOQ4gN3JI{@NfOj|!09bejBoY7$m6=S9dN{*S7GIY&;-8dSs3}$^
z{nc#>!j5|WRErW4;?&S}My7`4NEs#9sKQ@5n4k7raw9*C_~6iRoIm2+P46t+)%7S%
z^z&27Aa{Ef{(^vLn4lfF^qp|IoHcPSXn|UvQ}AhR5L2#Q_p4c1WFwM|#5gnl>`gu;
zGGO62qUr{|lg3dhB12uTEdjK$vxHD0dA$ZOTc&8RYuV!tby%N~=)s?Xn}!`xgHRvJ
z4I~HYA>Pu@MYg|2Gr!H+T~B*1j~Pth%h(h*w)bS9R|h5a2NfDZ0QUIP^!i(+FIxfX
zK~u5J^%hEZf>V;*8JwbfE)Yle;=xttXE6KV*ZWX8$!q&y$4hQZ-rd9OoB@j2AyCZV
z@7t1#_x(tgGdiL6Tc!_MH>M~Rhkph~AzUx^7B)ZSe(IF)JjxC&M`WKyKv3ScPmEU#
zJ}voQ^jN;%tmH%YNJLVAS=zZlrF7-r){~R4r(|OCtdMt&R9<(wtr=e!7m=E|Id)s8
zFfxJd7Qf1$_IyT4k`6vR%--G({F0`JlYQ!3V2i-rlDX$JtU3$|{J^jzEjon=0tNd6
zl8GYN-6u8yi>vG|x#o+RA>5D|{DGMhrmP}<na22F@Ci2E`NYge{VsB!Jbx7=<qKT7
z2q~lXY1N(!ZKH~8QycT@8ewl-rPqB_MRdzET{#R4vWVOCF}f6z6qq4TexTaXu%y64
zRwfLQ!Y07ZS{T|!4zU5b-~Hs<zuS&|7XTR%*#jLd0btp!tC~<gz3@I~Y<VO+aK^aW
za%<i&KO^pOb!fZ4HV*?Mly8gPo(O9ra}a*IN>1D9D{U5^yNnc{ebas8alJec|Db)i
z>zhGx(Tikvxos@!IQXP+(0RnqZM6j|BBC(m(zwO!`=)%@qssnGk~Z`Bs6yD|utB%S
zhAv!@XA?YlOM8j@ZC)1$Ew#-+kNS!;cBK({hZ?8hr26JPtQ>aLQ5DYZWz7<$oA1eT
z#DZx*E(kvaApB7ZW06(LZ)IDu>d;@v?xqmg*NU=xmoaRTzT{~2gHke!=3)cC2H7-i
zOJd5=v6EAucPi3U*AMP6-F;zsdxB{BD0puIx;3xFd;Mm}RskhN7_nioXK2AkK|ukV
z3VI1P4=705;PW%bEXn9W>r6pZm$kK<a>;~HNC$1nD~pnf$kz-EoE9QYVblV=zlVs3
zRxe&c7+^Q1YwwAUYJE-%WBn!-A+s&&t;zcpCnLHP{s%+WUTm2>I-N-lu1aA<R}uH|
zeIpB20tEOgf`W$Bh>2F3pI>RLbe?hp-(EpeqsrADIPzOv_VllNT%+gB!};g}q!k17
zDL#d{foW(43r1MXaROp47HtI{^eXsf*&k8t@U*pU1~}>pNY1>G<YE;R+u+fLa!a?W
z@lHk97U8iWXv9n{eQ7A=Bj3$?H#!AsGGh(0T%MkF>i!9(;(d34^aB{LEing|CfRXo
z7l3}KD+}@XiGz_j)%m}n61DY3ZvQbUY~Q&AXW9_0z`qYJ(dDpO%|SWng&nrUMZ4!-
z(T7q=7t&?cQFEVjXF`7dz55rkcFTF%ff=j~$w<u78E=)3Sp@|bOUhzrHms(~gLFE?
zLRjx>3PhnG?gOQZe<3j3To?*VuIVPfE1`(RSV;!(VcLIzEr1<Kvabd!jQByDa4+;s
zeu`O}N9M1@s1*JPM>VQbpojPZsX-QWIp||%23P$04L^avAHeqfg*@J+9ko+z^%y;v
zEidt*i~1#c`jMWrwopng>pZ5E4KuBQ`qbkI^pv@uk06*PRgz3MDWrloqZK?(=#a5v
z5;yVFl=UU;uP1%P+zvtC$If{^{_QQ#&ljelA-a7hVQ3WhJ-fTns4XK~j_kt&0a!HP
zF<ui)Oa!aw1R}?8>>^GPWtwo*Sh#GqkHd7}Q*kGFpVzkI;xF;A=VKHc99<BOnOf!=
zzdZoxqJ;7=L$;V>?h-J-4gZ1MjoeX2MUzQjzT~B$-a}oV8rB)fcW;xp`reNKgiDc<
zbOqs7vAt;D#%w?5E0y3|F@uBWZo>X3k+m{n{z6PQ5IeSdXP_i$tS@5e%8V|F8t@b$
zL!~I)rbAbMw35YFtDb}FY}Q4xLIQJ9hDmc#+A3*-`Qi}P7WA_|(v=({!y$tAPT9yc
z!;fQ&jC6h%pQzAGc0eG`QCFRV*KO-DcDe_G?w0>f-|B&W1QLSekus)de+GGUa%g`*
zs-YH}CK<j=ja~5i?sPiY&#q6wP9i^D!a_*N>~a&kMP>DMSKiVCz}DPEs{;O184oBy
zUGjL(X$%pZ{+7vyyeBWE7Yyi$wv%K5qQ`V6NwBPFXd|@{pO_VzS70YN(YvP~=zY{d
zwfD+bw-!x-R6m)&6K!~6vP1HN0Z_5tBAI}rv&?TIi(;9Q+&N4bj|H*m;GmjpOuCMj
z4e%3qaqYtk&BFNB`&F!zeq2C}tl{>Ha2}%adSkG(SigB7paO`V!Z%0&x>gW>jO*K1
zer2OV>-l1AKRZseF}<v{G#x_A;y0Vvwrqyp9^Q+z$+kNc@vENVG1GarR}?*z-Zb!2
zwrMB&ojvS;5~WM7lR)g8Phr$~KfBqakP#+8P(1VYEbSkYOTM~qrM#h)@&%nONm4y<
z!LJ$lGd%aWlvOkyhnaA~8^=%O_!@_ypeOsmW6#q^8o*pBOq2nl0OmN=Ru!gZj7AeJ
zot@7!5tFWNRj5lFjcKGrhZ%jN(5&QR*kxa*NJBvl@h$Hr&+Q8JuYdh=%)Buiech1$
zhqvek)#>gZ?vyRyL%iwLkK4LecMdar1>ljp`4HVSl<ponTEFQ2tg`*<{*<kPn3-VR
z&uHLsvoy_dXQ_0Qeys@nUrb~+Ky9&wXZ*1VQm!(@wB-%n%OYezgr)PYN3fSaLng`|
zx=!H^Sl+-6@#SC-vsL9)%OrQ!ohhv;Pty7IcKK4IX-fQzq=X*B+p=Mwh9yU}eb)RV
zwIvvX!<*2Hb$m7tt!AK7MxD=TJ{C;q@oD~Wq)|fgy_WG?d%Pv|4<#Fb_=k}_jiDpz
z?qNPPKY8Pbsy3{R%T$dW%Ia+rwe){`q&N+hmX)XNGVTIcre<eK%fB=qrs>ltz4Ctx
z4<m9<pJS`vaPo=lzJ!yd2d9x$0t(l3lbnh<x(|vS9{$<zrhMeXvi>!+HkavGMVMgJ
zAS&kb4p-u-kfZpWcv4<>{{IDH_JpK`QV`^(<pNarbuFE+EzUgSIiLGx`qa2X-Yr@%
zGpGtgE@`(tOhPt+5xwtyu9?2<zW;b&lXpZ-q{-DWpXKUu1r;Y`YF1u;k)Q8K`8>rU
zj+xAaLe2=T=Py*V4j9A|KEK<GqO!a*C;FV*TbqByHW5FxwmfA=X<GFA*@eWhvM3sh
z;0+hl2Zndve)972UD68OxA$)U(;o1vPJqOLPj2Qh<9TQlwn4iS{=+(No+<8;=lE1M
zRhT16US0uWA=&^QZqXMlKErkQWv>fub*{$tMdOZ`UXwSTsTx`<%8bgYShKI*6ekQT
z?+XIU)Tei7w}Ma$gRrehJz@S>9fez7j+AFjSQYr5TJ}&3Kh~u}Q{s_+Bz1>;Bf;bG
z`HIU-9@14BZNx@@&ZjaVb99TGkZ?m_c5o%mrjod-2;X#P?Q9nB4n6lr&wXbYt>@5>
z5VNOzO3}q2wf>k)YlS;)M{>54^xN&s*D=)^8*T9_;(ItA{YHz{CrSDqjr=S>kZP>4
z2NJwOV9R0zL~1s<J9*291pbS}9P#Txm;6uYVb}*1Nfdjddvm;5HMY$G!$C(7VE$FC
zOd%XXu_vUgY=b<5evC(k>|XD$QOIv|b<cXf^@P*><^6%smCCj<+v&})<Sj$ADY)vE
z{?Lh9BRJZG+ylpMT2T*lnD+{?X-b#YmU-EIGVqcwncFQ=ldAZY(D#$eciD9NlCOaL
zpZJhiUarBJ>Y7{IbmHeDzMIY83a<V>$=QPDCww5^ZVSPziGvOF+mkBHA4{@M=6lR|
zCBWj=R+O7RPQbTc3$A#`7p^<Q^~-$Tg-hp4x%8dbXc3JrukEMV4MNeI#xQy>dZNMf
zLbg-3Rko}0&i8{GwjFFYnfGb@Il4LFNlMPin&8tnAO@w7prG{a*0-~0mK#)zzVl>Z
zqW(5M?cIw){mYzP+7uN6VTuN}M=Hx0T5(ys+nD#DpCC4UUgY-Bm*(%Sy7)7<4zwu+
zD0@5%+LD0Q6*=3C++(T$TT8xK>CpvDhY@)4LX*iuX+_APfiZC$zkTbs#GCb-%})8_
z)d#Qh-uC~<8E>7tk;&VJWi6k=1lF+<I>$2ocTt<q%OIVwSc0No&$`P4%?b`|7nT{*
z-j)NnlYz-YPOSZLFm5i}2Df(xAP>z3RUY<Gw9*kmGT24Nb4GMkv{*{UczKn-f$J~i
z+>OG*R_A(iigu1xXqs2|6Wzw6D=tQ4ZX3pI<hg7o+2Lo3ngb#Mw=U9f3NUDK!s*aT
z7=Vq25WBx|aVJ<CGnch8`CoV9?alN0<#lSnMf6L{-P_&W5gl_+x^mAy-@0+-rq6l;
z!<d07zI)mrppJ*Dt><FmMnR!KWtEe_7w9c2$f?C7s1Rq@8Kn_nNJjeyF!NFP`P0WA
z!>E<B2lSxtC&)jOPF2+=O&c4OF;_a#u~YNkNm$p@0~d{7uj3_2;oF+VCGVH@J&T;J
zU0AUlV(9W_DoLOyIW=Zc`OEcj<SgUC`BK~=AsJM*r67eQ2j#lwYI<)C#%fc@B<jan
z^^=hK)9aDL#iDO13JN^kPxdgr7j&C!9cXNpr;M|9#jn2=&FVxk>~U$)^0}B??i;4E
z0Gbe*K{5Ccw)Yf(C0Bw6xdCk7U+uF_wPsFl{uy`Sict~yJ$*`p*M#<3WTue6!^Lvp
zHCOJi#e5IT20PK|e4MYWVGWbhnQv((Y^Jwu8;&~It+`$@<lHzd?yYLVrrd!z_4eNY
zn`HUl0NZ`L30GV+0V-bBm)~JsURn<<)fc~(N4!>M&>a}6BC++5d6G{}8ll(>ckf;%
zB#AByb;Q2%O&8RE%x5QXdgKfZ|0XjAA-bwN>`Ycmk^rvP!O26qj}*UgYq%+mm^jP9
zUhVeBRd*QeAIdIc7;2!Yz|YEc2frgN_FC&jy!97!(f1|8DF!G=_T){A{te7gFHuSK
zBCAa!?|3h<Eow!v8DmSX4{Q;-bRDqwM@m+^$tnO9a1oHlZpp{gdeE+9(76rx(yGnl
z!!O_kSzTuIV>4qc#EX|888eL&dD1LpqV7c9cI4=Ypne6psyf9W6--*dmSW6tprJVP
zI6x4*P3QEWxItK$_#3538k<<M*b!ontI=z~QvR`7b@8yDFZJ9x>P6N{GlQf1m=0Qk
z+;KH#`OYH>P{QNG$z8Y|xmDc$iUU~_`>l@m?iA)M>8kH`<NjdBv@L4Hver_d7{k)8
zMPHL}ikiwX1p<L3%I+8hVCH_pkw9O&1s-6pgkl4S??KXGmcaMZL5n5JnZ6iPxj;;v
zvJBK=%S^SoPY`(StI*C5&&}qcZs(62!O4nlg#V#hoLb{omb4_Bdy8V4BvG_vFWGL=
zSzWT6S@krG!8(Z<znk>otJ>V~NWpKa`OmvzZPhFtvRL(h&5At@0w`XAp2{hmVG>46
z&no9JX-_`t3Uu5G5AZMtWRCf&+NEkR2KmSaJe4RAZr5bk$$?4U@a`kZJ)o_mR)f77
zCxh8i_UI11mK=b^zaHr}Na`DHxG!L5sazc^JhpDtPWN+BJc&l()#7Cw!geV<5VmDm
zM+uf;HZa;$L6AgEoTbAdSLx)@HUelfVBy{9^WxIM$N%VWW@j5rV)D1R&VD`5DEcVl
z5~U0?O*fsEK+;0pDb^j`7KZ~_hxdrx9mEbz76fdyn17zi1U%J4@Kl~ko-+>AMD_-E
z&$<1*SYBSO*?``!^p}ce&z=(&8*{E#?$|_rVcpSe?LLr0`r~$gVRjxvJo?ZW;`5Ch
z_#FxhpD$by2xg1=Y!$gtG-43VVv7jJ&X)JneGp8XqkCf7Ms|BeXDGc3VG7}`5ATjY
z6y19V-eN&e=QJ#Xa|_=?6`RC+oOt&qKbg7q+18X@jL>6ycNsTI%7HmB(}-sl`JR`R
zuo>`XT}H1|{d@7_>(awTss7i!*{(xE{cnU=K+n2+YcBG5eLOjBql~8}k0^}*Bk?YU
zQQmLwDW2+K8-HD~bngE{+j~Ye)oxwG!GJ+2QIIMS6c7~^5$PqOQWOO&fYPE=Ap#;T
z)KCNjDFOm2NKjBjgn%@up%)QDHA)Ww>5xzZBst&3eV_B5=R41j_s2WVj|@h1jGeu&
zeU-W9nrp7T3bnlvkKmV{z+k_^mn~7^YpAH37-PfmwvrDi_tBCu4U4xwCkAF6io0=V
zvEC}~tFX;Wfq8)y{2~#wjWA{su#2@2+u2}o=9{6LrO+1t&&}V$zkR)uBs&>cIC^#Q
zR`;7G=+}7^YEW8Xurc+S=+GyQI6KI36IG~~op4L=9rIaI7-8~yw;FL9@XbJEFjy!x
z;oF4ZW$1-<XzGYJc0Fir@pwp<hQ!}a;yh`@tI!|hIRu;oj(lK!AKmkXP*H_lw0z4I
zwQ}Z+Kz6eDOLq2zha@MOJb71XCX=?oFyE*3!A$O%`N57Npy!r8V6M5Ms-=*exdPIc
z*QGX`se*6vEN#xs`DR{<)_j)NGx0qB$J@R76~XI^UjIGa<-hctV%UUWX$-2~@wcX7
z9+z{W&I9sgXX%s3XD(qEh$2cq0JDM6mT(z-Qx;KI+SB!D;=J?XWb`?IhnM$)mBvx!
z1jdUlDONoVyxZCQ`qN0t9Ac9M45??96Y6B>mi3PZn<lR7w`Wcn>T`RP7T)&w`YG(f
z4?W|zYlYa$^mXdyodMU<9c$;L383tJ1um^35mLg<z4oUrh*^f5%$vn->;>rzpt?Ll
z9@~P?;?~!NSdqt}>%1UM!7J5<JH(I~x12duqno#-;&~+Te$26yur-db@YoHam6>^z
zwkn;gS$&`|CD&z1{*0yl#Q@*vyVYm@K<w8fa$~mAHvmRVsvp}`m~`2ofz8-eWCU{I
zwDaG2_6-`T#T`aR+=y&yw#ZHPJL;Et`-bM{{9i>T9+|)A?}BXp$T*T8?K5bU_8Cz=
zT;6Vmk&s;ced#PzwrCl3I`}7H#X<?Oy|Eiu*?vWVMYeLA0l>%obR<*a1>5$m-DrV>
zf_;|BO*Q&-z6)!*vb+I%a6e+o9`<1|Yf<_!+hUg)HK7#;#gte^UFlVN?EV4vEgt;J
zH0WRmsmPi`PhtlXKx#aiut)-_aof+xP{PU`O%Q3{=Olmzrr8uiSRD_oylSwu*!fY~
z(x)$1T;<dLjNcQafI`!4w%HpPxp8b51A%Y(JnQ3oZtP;tWzsDXc6&g0#|AC^-rQ9n
zYpw!kSWW(Mrpt9%%Yhh_-rJmbVH@jLGr}KLwP;HxSX=<);ziI<!|pOWnVl0}F%N8x
zj(c-`<<9U7{Uu6o+jt7bxn}~uy@g^5mf3B9d6hIq#%|B5u$}_JQ5%H^*8ro9P%won
ze|R2!O3&$Qj!@aI8<5miz%^<9Db?l}3_Sd-Rf7Td*4$oc95o(liZpE)wjIrK8_J-(
zsa$-$g{R;dsc=C61sN%2J7=&R-qbc`Eq3-)DcibxtTsDcswG&m@~0A)##8w-!OnG^
zDlfhuedS`^w1}P^lSytE=r}2y!Sz`=|K0Cvn>*YdUEZ><ZDC=zq^5watd+f&tacnA
zZ7X7emFe&%(UI2VPiWc2pVJRJ*!iv=tj=@TRrvuV8X9~GgKSfOQH6~vL#c0^U)6kN
zi+M$=G^uxM3-rQ2(FPf)j@z^}W~4iEEr8^)Ba7`o_Y&qKfD-aN+h*KuKy`fr`5PD?
z`?D(iK3U(jzU1Nc=dR$Or>L8jft5#c$L6sLZE=PyHMePJD4hr&yecnie5%Vs_T{%%
z9j02Ls-WwA01#tjQ=Z9(AgRE5dr-x<*k)^*cwN)r9v@b4<deSDG)$z+AHJ}Krk2Yw
zZ?MF=K<k?q?|iCxra8Z;TGt0w7iahmWe(6lws_KR06Ra2Vseuzf?KdFx66!opdf#m
zJ)Q;k`iXUNkJ>KG^$grl@gY9?qZjjYyiGKj<U)Zs6^;AUPk>=Sas{5IZW!j(XgE&H
zO>zW0i&~Z5N4lq#CMC_A!#xcZZ~43Q<EF;R6HZ4=^{&W5_TRc)M%;n|s6nd1ZSTv1
zE7mgK28#{aTzA{<Pdaa8QtI=iDv9@;@EOicfdPzSTb&?{l@_G@49hc8@b!eZSo<xP
z=fNvq5hp|@X$|<UlF_=S%w+5&j3&owK!d`ct_y-n8uGt~UCi%|&E@ToCWf!sA{5BH
zeJN`jco#bVed-@$?zy5s?GwIE8e6^-Q-+3;Cog^ut(Q4#35k6?x=xjM+0NY%29m@f
z)-ZNs0+d%$Hqi8dQ!C$2dvk^t-xu?>KWV=v!ag2^E~6%ZV1FM5-ei#yk8^MrOn)3p
zTCowps-Mj{D5t`)Cv}|4=6)-)w-WUXv0@nL0O%t$k$GT^ZB<|qx3Ez07>LmBLj5L=
zDCiKWj5l(ltUXQR!hxrO0A&B5ZtB2?uD*8TOjL$ndfh#RO$@Y9tt&Wy^mD9enGsfF
z3E0O1!kWl&=<Hk=Yh}#%OXkY<gT=*rm!!Y651Ag~PBMOPY-QX^*pjXTaJcB<7#MI~
zD^?Gd7kZG1_XzR86?YO@8rFpo1$3tgrzcAijd#4SOUx+M(9{pEo81b_eJieU5cN7`
z!+ZhY@}q(vz`HNc%z439hjmPds$o%XgUMXN5}45z>#k9vwl%B_{d!tDB0=(QUcJUh
z@~^J&tVe`EiwZy0^lc=9twCl2w@pE?;;%wFk)=*e;4u{fa11sO)I}G5SZxm8Ps23@
zFt2WJd|PcSx!5@ShkQ$z96Dtsp2K@s!(duqs|j_1uE@GZX+v&^?B*#;vbYd-%r`(G
zUWgr2#L@(ow#5leNcv6l1Ha1BF=CCaQ^RjszMLcC(z`|~#2dKc!wAUv8QgxbR$tPq
zS&abpWfwev?8q+612b`P&lLdNpW@%u2z<sIZY1n4DAU@rGCI~+-s8Ty;;ZO8s!e;I
zsSrC20dk4o|CCE0cjXe=>W~d0{Q)r&(0IaNoa#>y=&#p^Jo8oV(Ozg*bKNS#{KFB$
zNwRaC3;LjqA<BFd^##jFw3h#iqjb)K+k#?4NTC!(%;rFPeFd7v_SRbX-1B8tvBR|U
zGlVk3X=ycXq{O-wc(%R=N{4)~Qj5MLfoCgZRlYbF0DnFbcKH>79<!MsG8Cxj9`57e
z(&d=RWg7citFnXV2lr+BSK2K9<TRo`eQ#YPINBMnmO}^&C$Sq#A@DD|S2#!_{Cx%Y
zr)Fb3hsjE@7W0GpWPyX}urlfA*eA_;S01_%Z*@A=20~Gn8Q|?FoBtpF_jL#y+{PHV
z4dh=J**3PI%7-&N$AeU^qVLFbXB$iw1*mXMeMFpyi?UVKF57BDWn-a)$wZ7Hs{w$h
zne%LIvI~9G^*e}?x$EH>hf9c{HM3|bG0LZoqv#8*QP(MFNL|;qd_3AtlC4}aTXk@+
zo<gishWW=ob#~0J*hwBTa+?kqiSC;rY5|7wzlRpjhGReDGOz{w9DuP9>g@YjmqJ6t
z1<H;ixo1gfn?K86Uw)c+=lz#uxbTgN&f7n3pHUNIOSpkbF9X@cQg=oYE*!bB_aBf!
zK@b#mvwD!cGMTE8#p*)$sxx~{ZzT7>se${aQwR(>FmJeLlNlE)bzFLdieLYZe)WaT
zPNMtDMdj&=kqvxE{tR}|mvw~*2)G?~0xP;Ki_5|lw=fUbWFR=^F(vrdSnYe_l-Tqp
zoNK;)qOP_?aN-A><55tlcll^**UK#*O6*_ImIR<JcJf*`cD`{81pWmO*WD63b1BO5
z(@|k6p9PNJ<>;VcQsBoV<s?05zN?9W_!o2`J|63e-x6Pg@4yhdwL(kkPK#{R`5XPG
z7sx+H)to#;OE-G;LNe1|g#aI?bqmI4F|iott-riGcKtj4TOVGHwlSW-)0CfO$<0*{
zzGnUsh<`BLCV7h?lg3`?*D&KbhfeR+h6HN%uFIik&O4>v5i@wE)ztnZ{?*M#`o(xq
zGN?*GFwRjc5@gELn_v6lJlx((y;c)d8Om6cF_5-TVw<_LUKpTDhk;y)!GLQSm#cVY
zfG^)sqK9<eS1*}6;}pDh%3P_*o}*7pK0%-%9H#+xdlKBuA9&2#A3zKJ^RMflEDBmO
z8YD2Q%86}OPNS8`NOJ@m&=)E95{g`xr3PZNEXuZ$JNf<rfsabN5sd?VTl}UMv4AoJ
z8Ik!$IP-PQ7#YWSkpT9cUlx4;_{u<n?$eeiKpT}hpc!!J%lt)&;HDabV$i@R?Y;z^
zsJhYOiQ_t>y%*c)`2o(`<SX&Tz9rq|FoSTOIqW9w{zG|(A~sRe_8;oiroiK1o6Q4p
z8P8qp%vu(_HwMs%s&FyJQL1}fna3ye;N7I1dvPBk9iM&`S8)X=^*O&*m`ec}9+tV6
znpw$<&v8DzF{87R={I^q1lbJbgZfJh%z`0Fr_WRL!$H-PcH(E7?<Cs0n#7@I`<0b*
zP`kISyjq;=9~lC)>IR#$q<1}Q`Ljve4#%V`pR`I#(QjuK#Ct640$Mc7Fd+sx77hR{
z<rAu8nURX!k0p(<+ClLeyfwORe}dRBK31|a&H7;O;Pt3iApx*o7sg_p>J)AU9>k4T
zV+LiYRzeEs(Ob-nvE}K6^v+>}kqRxfR`fJ+5WzCkk0H!YVb?4MaKCd|hEWlO`JA!s
zqgNrD@Yq|70}>jw5+@_)u0K5a>_gw?lQAN^Gz29~Bxib>v-V!6dek^1KwC8ZX%RKa
z!B)jq;w-a?VAskYYD1a7@#%FyW6*Jr=FWT~&GP;Whk8*-%o-q2+XbS&&w|y#fibxL
zBjMv5zhS^bpYNydpZn}S!o{ls+A@x>aOT(Tp|^}Li_P=ApLzWx44kB1k$<EPzTh?b
z&YXZrd$$^2@BlNk6|r@abrx*xWh7Rd%N!C2;m!n$Z37Nr_B{BVsF*rV41e?X$d7`&
zS#J44OD{b8@&TWvmfNAp!<(7|pch8qF=|xk`U!6y=yiqBDnDE2_s(~PzFIdg5(CZs
z0MJFnXlv4in*66mlY_7xG9%e9-|uMdy=*RS{%8pf>kgAMCvUb8_hB|~W;R`@eLmWp
z>6Gjt68$D11y-oR)CI(~KI<y#8FoS57o44$wTCx!z?5r&G5w4g5s)Cn1TujI?N-XK
zJu|m4hkj<hG8us3PFxbdaGwMVU*d6PXwXZ@#)vViyAOnt4!84vKT8yed*vB+TOv!$
zsqbhV`+sOG2){_It0RAFEV?OKywz>y2nhU{_oY0Y0GZ8<%Zs^N@1U|Y*TD<(8I;&R
zppAN98rcpt7f$}}F7bXUT(x>+TKJ3h(uz^nkZ1gk=V09#nmi*`J(n2-j?lNSHweG*
ztg9aY5V+@zT?;(Kwk%3q<N>eZYX8zh`+he!=?c%oS6tof<DSl!Rh?Z-Z2N%zK=-BT
zH;x^!uZn7__)aow@*Mh+X%&I2JG5@~e4qF}kvYY!0kHfLB*c0TV3qxtF(3v05%kLW
zCxI<7`w1K-aIeLy6J2G8)5E!5C|X*+I%=zt4~>+dgj<_85u!m98s?VVl-4EFQqG;a
zdL;gOF1=pMT+`l5%Oer@<0P;QI|f+B9UCxPoLB=useUZ+g?h^U6e;$b@h?nZi6bl`
zlDBYa`fIYmYvq*A7cp~}bvu3t=+sLzRuMTENGd+WtM;O}V5=M&{oH2GFzTF}iMpZ+
zH8fNGF}mn;0sz77*#!Wq$Bu2E>V|#uf77<G;H~j_aIRp7v{mrEQEC<ORV8T5*kuEB
zOl28I^*&I4N_R;=$}c}8_;72h;WxqKbj7mk;%4z^9Yh>Ghx=v6a&q4Rux}^^vi&K$
zsv>e}&@$j-_Ulo5g^oipz=N>iCY=3X>&dIfq3yEFH^E|^nn`btO&J!u=KD$BNbcsQ
z%2iY4!@S^7XZQQ$eD=~v{V^%QGb<{#-~Ci)wB&Mt7!UvPuBD5z;_opX@s`gswy7uf
zn)b&Y&;>atO$_S?me*)u)?7+UAwV8c;Iq9qf{cp{=FimNB=@cb+#~UKN+qv5+I46_
z7!Uwkfw#yeBatWo2H_RJ6{#U=mv#&Twb&=7Zvc#^$Wr>PPeJU(bdtMRCBEKhWF)t{
z`1D{Jn?h_h-efor;`xn|B)9s)yaxeCP7-T^z_K@<b-n-_L^`3i)<{@20{zpb`~EI!
zOC7hha_T9ehLf@Kg28X;?P<PUf2JK3SDhTk$oLSs83YX5k%5_KH5~*<$;XS$j>vpN
zm?qjV{s2j8{!9X7HW&(8kG8K7xGjdPFlJNSk1yzq5W+?sgYQd9ie>sAY?WKBs~cv<
z5hjz2!F^mItZ~jjEm<N&GLgB*jX+gKu})UW^UY}Qqv9GnyX5YaR+LfS*Qh2bcbqc5
z2p@IkvV#?RSkv&5%+JWlXqU0|Q2;>n3LqFfcpUewRT5$MakizJMvqaE5y6KB=hc;(
z1xE!cqBeEaY=2KF^tW3|##eR)`;dB6P%DmtxLt{`+DGUwB>8=?1eW+C_aBguIF{1b
z2;(xfB{$TL7Fd;o`t>;L!5<q7J?Vr~@;A5G1s20I?veg<1s$9~ZYd(Tu)j<J=JxJ_
z1i%^R@lC|7?S3qkC;uHi3Xg$V)9-k@^A8V>Vr-~#mn7AElb^HS*?JPr%_$bE{D=)t
zRNn*|c5FvFMhG-^#m@M?_~*cOTK_$;BMc{%rX$qHJ!SVDWW%1u-aTD3Nze%V9N1_~
zz)kGz5Mt^e_;;FMtse^*m*3fsV^(Kztb$~{`xfn=<6%Z=1hou=kW=T|KQErCIp33$
zYWpkYHuUYqv$xk_1Zj!&t!DJvrHw$ajVh9jx7Wa88Vos$BxeMRVY^QPAPNO}0nzmW
z^;XigqKYqli<I)x3LP<%=4#VJk01iJ!0@0;TeEXql_3XXobom=Q-qgD2h^-E8MR*@
z+AaIED*lk`o&O_06vfh!02l@TEP=^!9GX&E307F&K4t&<t+sQ1Q{aFT>H-Kkl_nd?
zkf9K{oy&cHnvwjk-Ds0}lG;_IFVR!+&1)vYqkcR@izC=wq6~B%#j&n9g<}`EK&Idw
zUx;cu(g_=-a9s@dTZs2YxNAU%f$D;upMXl62lY>9lZ6%HKmkSY>C~ws<u^;tK)9v$
zg&p9@+SsWZm|I_|_@CsjoihnpQn3+%-#+#k>0dJ}5^I;Z*}7akzlgH_mF`m#^8006
zUfeakQx8uhq`9fKD#V&kr37ct<ye<%VXT_dU`kkY?;SY=%=7SvnQtBF_KCA-nuKyH
zHC4G2UA@Sr)}djbp*TO5>%Vc8KD_~GnZ$W;kQKSJ2K*tl<N)07N%Xcy<miUllg{EY
z=#dJZOz!lTg~RJa(1)8Ouw{<=^x}V2RQ`wA#sBO7vo!`@qvw+oqKusc2_A|gBYS8=
z7FkIb!s8asovj?_^OFQ>g2^@1;H7$t4+ZG^(=LLJ6FS0~a~uZGGxu(A`^0X}sAuq^
zTd=O~TPLcFnV}RwSoWER41$(w-%-o(X<(bSYwF>**-Ur_1hu|?Xs5zXPnoGkk;{U`
zQTEry742j9JcbqWc#euN-v#pE=gTc_pL0=8wr)QfZ}?F_irarrL;Sf%+8=5TZ*+kP
zTns4IdmpiD_GOb2`@ooX?v7~$ffZ{3VZvyHCIlyabc2h)H@nZ*bN`FGX6Di5+kCN#
ziy4&HT~tsP2*LQ?DIYjrz-Cuh;o75_rg&AL)~7l#@fRg#&Ehz;o~7-A<17d?bcQ8(
zb*%F^-|N`nUB^&lr}zZCe)!c(KR<_Icgm#S`pn=;Zd11p7V9~)pp=Jj%^3_$%w9Df
z;i@uY>mVDck3*B>I+1AP)@k(wW&og4W{RzljqY0~gL%OX|6TdT1PzalL%*|*Fwg)}
zDUB@z*`O=eIbyGz@=4sM`P|QBvZz5S!@B|YwF|}0@S~!hF)q2k`*l6Q)z2dlK1k+F
zQrU#=MJh~#0{QqHD45{7AR-_&k8J?1A*a|(7D%4t>cD<p6pQ*1qD`hnHDWn1ZKmqD
zr(UW{_!xG@Q!CAPcwhZ$Lf-09Y(#O^z{#iYD_O%X8V0Q0%#$OY8dB^JS*S;N5G|Hr
z;K&T8`8LW!>D70FZlSufzY@InujH&WT7fIn^Dp_$&YT$Kr5ldpU@l{W!4loQgN!4(
zuclLNv(|@dpTf%Zl}3P#o)W|Er}n-}l<M}HyV=$*q7x?z`m6jBZ>wyg9zNv?jm?m#
z&xqxLhgwuLbmD|dCHD5Zoo9NJ?f6k?j^5ByQu0#d;v|o-3tFN0*DMX}KOu7g?zb<?
zumVU-r~j_=!q?oQde&zi|LHj3U9eZp{ZHGgQ=Xl?t8*96vhPri6=3AapvFSzV+WSp
z#ixO8GnW71Xa6cCAmU7$TL49*aP)7t{w?f8a543yyx3g+ggD>Eruh}0k*<5aryOr=
zMs=PHtEXb-;DMwYROl_bq3^_DdL6<!d3L)pL~zI1$4?4|J5^n$(0ZX2Gas*gm7WdW
zj(iJYL#P_85~5AqjwH6@eiW-Fmvu2jWBF;-bo~vA^&PZ@ICJB}F*#4IaD$@K@W}eR
zx9uhNL6b9U(G^Ubl5vE&e_<}XsX#dhd21!*y=7PHT`Aa`h&L=?wZyrI630xl^ie?Y
zbrGyZJtcqA0$4FJj?=6LPwb3GH_VRG<vtE)$PD5)K12oh1<T55ZtT^Km=sem`SNi-
z<M~-5W3`So#P9iAbZb99P&P|3N@_AWm_IBtN>Oada4^&m51~j(yazb<I8BxjYk~V0
z&IO@oF6(MM0CfKZx;=EFB@>qrQ>GY{H99)t_x;${zCqJ4eb&)X2NONdc3~|&oi2al
zg{?e>e07>>6uI`<FBFIQ*O-OkH<TF>!)m^e@dLbUQj+#g<a|&V#=^0xfLz(1`3z)u
z0M_LIydDs19Ilfi5uD0t$KJVlam`mU{;bOO4eC}TR=;fgmS3S&4_ODTaOY4ZKi^Zo
zt!P=uM{daHj^vqDMb3M?>3dgns_VrMJ+|?6=p$moz?#!86DhR>-q1ZF^V18q4PC&i
zb12{Gy)&qG=-k-`*xN&Mp!ud73&K#Us!E+e@p2@YpJ~+S=y_~pbVwBv0D%<Z8?_JY
zYAcQ|Y3=>SbEf%EK(xe7%#F=Ved%%TbyYEvvDI=jc1t!B#L`8kBv!NwWR)vn(6|1e
zm7yq)t-an!8N&+#^{gY6p$(gm9rJg<O(G&I@KWz|Ep@7?s3N&DbKeW;V!m@+XN+MP
z5>f2l?C`Fb?UD`QojRrM-`GWe$o9Ley6TFe;;%KeuO!PV;y$E;4s^OAj<Vg5>qh8g
zdKqB&@U4J)0tmDRenG-Rb`d!)koZEsKou5gY1*(kwQ3VOz8oDaLIb=lLpOKKPdz9T
zbh2XRJPZiRMyhs<$LlF(Y)`_w=C<X=qZsL9Ey(SOv2E#YSUv$<OGetyqNkx%_Y+Vb
zC6GL`>6?ubid5dW1+~rBvs#m6)_k8T9)PAKs?D9fAo2r9aZ~G_(m}j*tePgvs>U(*
zoP;1NK#pKr0Hn%yR6wYMe-|v!GH!f;p(L0IFAl9yQouNtr1*+6-0oiDGqiK=kPXZ6
zh+!=4`cmSz!2JHFf4T|mCf11<d3U<$A$E}kwug9g5MRkPy>^eVZ1S<t{p^4Et$wLo
zrIp0{=nlu(RX$^ArrbLK=0WA*GC7UO<qC2$n@GmvDjFiPP6TtzDN@+00CqhgfW2cm
zJ~!1zn)ixyB5<=r?QXx)^38;eMc^25!vv;{1xt@DzGxmx-;ZNh`&bZT0}oEPKiTwW
z6`VH@7JSVo8*jR`=_Fy({aAV=<gEWVu*G96;@<(W!ImPAA*&G(8Gy+_4B<-f%>CQT
zOeSCf6G*)h0=*1=3Ou6RKRC|L-Gi+0Qdy>$<EmZQz8_@Chf^gPtC{OtT_`_R2S^<1
zp2u%5gODiOBIGo-Nh|Oi2ueK&RPDb<!gn0cYlG1Zl=sLPQVuJA+S|v@4t>im7D?0y
z{o&W;Wd(B9AQa(Vr*^U;eu9`}2gg#gE(FEMye>BUDDIJ2W#Zd+?1m=QrjVK2@usy6
zMA@HJCyztHPm?q2M6lgeV{`<V1Z{HV12z2QF3ZLaawqfX;B~&U`Qo3#o=J>GTpiC7
zTjA8iS<HZP+7m#49YA*gEu;Ws`uo?+pD~QdW05xeG4nULoIcdw9334TRH#g_*QI2!
zGuaBi)M1$A$jQtZfcFvGC$B;P0T!7>6=JOXBv1vg^r|(a5U{)=M%2hSPiS4->FUEP
z2c<1g0*2j0w)FIMI({f5SnX$w;@{ROV>_w>PyYk*2s`r$vi_c#%h+Re;<|&5>M2)q
z=vWlq5X=FF0!^r?ym($Pb-OU_tYx~NK>K*eC8<1idw&VtVCUCNX*v>2gzSc~4NkDm
zdJHd!eMgooH>tkABGlw}ydflDC=xWv>j=||^CBG3^aGDJc)Fmq;z;C)!yV<N<rnN9
zWxt6|T>02)46*0Uf&SXHm6RW*qL_7{dTBpoah#c}36-X`N&Y-ycJG6JDTfJv|Dx?9
ztx)ZO5*o>z{<!W??~*2Vze}pnO}|twmzJF55X08?Z$9@5jTRo)K9(C-KwS*(!>$-<
z!?C{r_TLo-Tlaa7{qA>B&cQgDMe)5|JtOKArzTdl)s-UN%ByeJTUS^9z-I7R*@MB6
z=fdCZ1?1v6+kg_noHRQI$fil7Zo)1P325vRkeg<+s)*NQBPWQpJ<L4bn_a8i{1O+?
zI-i;S_<+Y^hF9T-7|5!);<$cOoQg?BZ5egKkE)ueIg~eL()7VJBafj2lE;4^zyKos
z)fe)=@pp>hK91nV{E@S#86^vB{*G4A-i+(h(px)H6r>mrdw=f%aBC2ssVXRcpG$X>
zsV?`BzJB`q>dB%RpXwB8ZxJyuiL#lszx-}tvJ`BT%^_OT5HjXG2JcY8t}B^jkOA#?
z)4Qo(JW*Vu1v*Cj0}{3V%=MNbb<JbUQkia_^vC(4vcj9fsmB_iF`^jYHUIL<l3?4>
zL67Y#(7fV+rquE+oBy<XlFY<aN(lehA1!=E!n5N>YMU(7S97cp$<H{5m6+|O`!-p`
zr+of4YW@C4R+n_J-J?T}PHe;r9%CZYxO2?U;02YsVAq%3K@`>U2&5FjKjhI>SWRZw
zZQ%dD25>0o9@mQF3Or1)*3$qf{F%g&QN7GNuN;#E<$fu|rTJ($seaT1N}u{szMuTJ
zyU%kLEcaZ^A!?Q!BHnzbcQ3bGUsc+Podg)b-Gctz59lJ3-;Cq1rA_rqoT84t%}0Ca
z<#}pnSSWETu4o^NRon%8oFbGxV4Z_)U9yVyY><w)P}{TKk#(Pcjw9n+cIusu)N5jU
zjIV|GPt-l416V5q6ssEtlFG3TV_RqHME(J(z%yllwE=iAfTi0sop&x{6dN<8YkJ&l
z2iyx?4|l#cJZ|@(SEuQig|2jX14RU+0)j;tuGHLSEn923E20sZS;bq4O+kEbU+MWi
zJo)`x4c|T(OiERio;=aUNI{BVe>sAHTff$4?29iWwN{Sx)e1ZFAq#$;UOa&cou_n}
zYU+~XZr%DRmE2if+4nn1IT3R|^xU5MSpJHSbLc!~rV=pOQK@2OqsMvpRt*L+1ssJY
ztGk3{k1C(L5^_$3yYO+XN82~#A?)OPOa-fP6T7W85B$2}Tc^OgBE<rl%mn65Q*0jF
z6BVy?zW{@}-Cw4Z=Fs&cvr9}~5~2)wvsXu0)?J!ExgRiPCFRxBa!@)maJo)nrovzE
zgROar`<LF%1DXqV(l1S}PTD`#hgG!g63&2OM{FGUTVq?f14gsMq6%P=+dyP{60&Ux
z+6qxOZ3dRgt{XQM*DiLtr%)<iOLoONxu}{Lsnt(TW?m^Eum@^VD?;NsF_s4LElZDN
zlE3>|hs>yG@WF~y3k!j#foxl_933Y6%v+y8{g9hC@+#qtZ?2*~J2ABK#lCLu9|kv)
zlYcx8Qrz$%)rq*kZkN-&-*@Gp?p^V4$6U!YXocKs89TnO75d0n(w?8XxXB0}#**|{
z-phz$6eC64oK+9BdVdwIb)m+wyJG&ze)$>|erqeWB(wNA%!}eVpX>;c4S^nvY<;E(
zRZ|G4;;+QiTb*^b(r8}}4dJRh9s0`&DoshYRyyat?#i0Yg7XgS9^QL?xtJ>_4xxSX
zxg$7QN)HO>w9!sDvq=qyTDLWh#bk=plX=M9al=A_v6@5Z=G)26GZggMlScYhlaKf6
zVby^-{(GdY{FZ1SfHpSZ9G&rETVLNrl$U~(YHL%PH}PUV#jQle_6xcf&b5>biF956
zU)XR9WlGauSQ~slgl!ZmG~zA{9CF0G3pV~9y%8~(yW5|nXx3((3N@z+=8^+Gvc&F`
zzj5dOT~(Ta&=M~x%Sg4Y4|(K!FhpNiGC0@={|Rt&_^mhC#WqmI0(w8h<mVeZq2G^M
zxy13^xy04+s5dq%_Qa_>{&s5ldcY?NwQ`kV2jKAE!h)44QVh~5+4F=pCyo8pV_$u_
zPnrIq8;PNdp{R?2@LKAz%JCDD)MFW^y+WlEt$q!En7yR@Sf*2%Ss%;SwKDDz3n#h&
zJ>?NF`7E0W>svys&<heFV<nqS$83!w^{x<wNK(MX$^3?36dAb(BeB$d_xCJIv^;vU
z^2(uau~EzJBb7L{HZ|GM8M)7REoK&SGLF7mbhZkrd)*l_h!PT$a}*qE1VC(cn68l;
ztc@EjazhrSsXP_G&tu+XD}U!qI-z<>x<k4`a#~gr*d?XbfnM{3$uuC#7ET7GXO}>)
z`CY2T^g8N#$<OnhZ;-bt;{SXDlxLIVw=b_N0HI;dP+V?8Yztyng2Ap7PjW?s*C2(S
zV9CM>SIe!rI##qh!b|aaY8`GnCDoeKnm?I`*rrfaA`YltEx?Wa1*6PLKGJ<b{pjln
zgVZ+NKHGhV4xG80Kz$eH=cv^&nUUO2GPgOs15~FShrzPMw<Fh1LcRfP^9dBF?b)rU
z3ZY_Yb8Gp_(RIRK=)OX64z-RwjtO3t{O1f93CHC(H}`y%hUj!a_)+QWYsOCrlbsBC
z@cQ)uDj}U`o5fKFY&e0a#XK5Gm4LIfY?40*8d2*~s4RrY^`Jj9r2X1IY^^#JGgyna
z6IZA19586@1Kj8|PY@eLeoN6E?A*T`jhVbFAzJ1vQG7i>68iyshc*DXmp{Znex79=
z>2XzZ6`eX$x>LG(<dqIvRQC#}8oPuJyC|a?-<|Lq&oYRL06EMTz^?2nWbIo!4Ha36
z6@1V;YjdPHX1n#GaO-3qyI9!g90Bd|aEmn)SW3e09rAO@1n$^)zq(4~&dtvZ*Ba`X
zh;jA`+K<;~ys+m~7eAlA^)TW^EN4+?$%78$IYw5zXZWo5s?#Ef>yD;HA?LTity)7K
zDYK3<Gj*G>v`h5pvrT{etY3KcCzNI+JVd+?ipw6pYdx!Mu72zCU|@Gn>f(V$;csE@
z^*<<}QU|uLQ=?c{%)<!_2f;S+&MZWIZWWPGMF>4l<g(PK3AU6ScK*{<mIK*ZnveVb
z<Xj5mVDeXPG0*4l-2Xsx(_Ui&!Ne<h7C8FFvUZP?2xW4nL2}}%deGa7<8yjH+kxjr
zfP0&@FS!dJ?)6I3(}|h?VdQJFaTt+l7m~Fcmy*0+APsp|EWQ#ShMqAVbZ1>@YXG(k
ze`x^roln@k`lsJ#sWugEW?tIfotnk^tMFHSi#(P3H<iov^x#*wRVGwgklYN-p0)WA
z)9J5<%k=fI7J<>r5OK<>><1ivL=V&(pc`Ji3r26c3Gq*h1^@R^i@M_h)$;*<r^<$I
zr|MqPr0Pz!?rZ%sbVOC$h)sK7smCCx^^4xOnR)>SHFp1SRLc5x1bKN2GCQ`uj9`Tw
zhhF?IrvCT*|F3<XZ5nl`?komPuAO?HSzbXBN>kYC<j8PY$V-AM|1n{c4E=#o`t|@>
zN(@s9vLzZT4-O<&KEEtm8HPN)Z}#iw*bA~yRV}~zKQ-u$en3U$z6aA^RGx{@!bC*@
zK&ABo9M<2l+Xq0Ww5z+tHJuZ_tvFErF_6`#gjqjaJu9_pA$L5nrflI(^~dQ9J1q?}
zSs?9vu*^YEZWgq-ti0Wov6;kao3zB)9MI92cS7o5a%kVKAP3O@GLgpHR1gn3;#KHQ
z;47jLw;@;x7IMc0L{{cGu#Xsvj93oS({#hOC)gF*PB_IiCcnvf%O1N1Q|)Ly=aLYq
z&5uDeiPl(rh$8O<py{CpzuPBWd?hUHX3N$frG>8i+h{^Fe*)PC?C!am<=?q@%$fUw
zOnjaL-`Qu6xZgfgU(tcZQeARG2@i~6n5O2`Dw!Phf|QY-AugN7n0RLL@dX{NFAz$h
z2ks@(MIJyjkV33LF=b+kSRlb<1AO0}*n&jYN9Y3bDNaIDX59E-$2*rBE@^2!CrcZ@
z9ddYD)NXcW{s*U~)5niB96J{<ms;zjE4ml%4?M)5VOzPrpzz6{=}lI4g=)tnTbsp}
z-CrgHlz*8F%%v@WwJnq0NLXL{%VeM+5KG%>1YQL)t<hNCe7^uUp&!bRvX3f=3>sFg
z#!Bo}6YFD11o~6c;%lx5M&(}mba(99=Wb@XwCVHEq}|_PUyYBfGi1KQPL^WwS#{>S
zqL@3(*VxGl42o3+4)%BUS@6=pThxjM8@Yk>TckBEboHjLe!6Gt57_mgJNA^=%tw$J
z4ns0LvhE-?h*42@xaagj;iU1254=l5o=I$usy|>xUkmGiw0lgH)y?9j;+`(8S-$i}
z&*$r{{0RU6wx{iOXB9Ak;0{>NALC)EV+`8{={=>T{h!xi7pZMQe3I#q)crSxp0lZ{
zjt7{xfL77S$$WYftC57-R?7d#OvO%CVLWF+B%agT4J!(~L&G(g(tf@diq=%X=FcBc
zN%r=3OgCvKoVc~q-+AV&k%pbDKzAUNX5N^-%y0c0!>?1>Qx=`ILrR?b6X)lkw~7$1
zm$r9;7OHL;{s&M7rl;<V)B<@{EvR0;kp(Z*)Tbubk4pWN9Z<r-Er(IW@Kf)uNR3`-
zH86u}ufH81i>ec#TDN>>iS%R@XHH!-JX%sM08{0NmUQ|k{z$Rg+*B9x2ppS@a5j3l
z`Pcq+cV5e`phv!jM?0khKxY|&5TaOyr$rq%6&Aq|lKGw~+%sO3G|rK#>*#Y&ox4~X
zG_C~h9~0V=%=e6g*=Y0E9_2sr7K4}gjnVER(Z9S0Ep%l8=3i{pSLzQuO>Y1Wc`6mD
z7NV%3QA_>l^FqZZ74@HUo%wn1f;6;UGlmXt#;qv009Cm>D-u0K0EE*?e>^i3><B$n
z6PPZ=SS%2RYw^#iVDa^r9d3OSOY(94aso$xrJwP}=&?_1My6--EVg7Uy6&(`!lz#h
z&0eRTrTV)|l2yi|-@-#0D(Y?vwkkHbWlS=Zz~V=2BL9`=Cro|>b!xD`sZR;QCsrSe
zqp07_)vo;K@=c-RXHI85Z8@fq=X@w}eNQzqVob0`g?0pQ+s*pz?)?DT;T|BG{Q}vl
zT(U>8%__l2533<JgZov&(oX=CF;E!K^k?zF`7o4mvF$}hg#>)S-NE@)=EQ)Ht{i{o
z@l_J2JE@Wb;$5IAW2pCgEjl6~&Ylk@_VQaddA&>|3B#)cru3ym#-~FwaN!9SKtlkc
z$pnyI#Qr*qRnCIntYv;=@VNSE?sLEo=eU_%=c%!jhA^i_j)ScgLh_^zO|(U_B%KFS
z{v?rnvW*TG+SbP1=s9bEO?l}`uAUBW3HAod5r}fZj?3|HRj$Na@10v1QLemqhVTPM
zPp8c8EWR6(ge~**hKKjtL+3br$9P$C7<3zGt)4e~(`x9_2Od(ipRj(!*=DuDJFZ78
z-?I2oZ;sVPF~QJIW(J}_9CQ;qll}<;x>Dd47~2t@yYnx<v&u}{p=!&cLiUBRmQUV-
zax8Dt`vBhtc|nF&CDkks8Ri<F@P76)Nr-B7M*7*5?03(~iaj46`AP_h*$Gf3cXpM3
zC0It%GKlLo_9C{AE`?~>x(}w;(+uX$FCW}(rfvA(QB%L8$)lipr5RwnR(AvwFyZ*j
z+xzXSn8s1lu7Y75`vkSt@qIUiq1`x8*Q_!>kJ_zD3;uUi+MF>}mR$X9X;i3!Bi`Wd
zA1IZLD)*_+Ym7HggZid%cB6aNUnZ%;^GsJ){6cd^38K&lA`TJgA?><BT|qn}EF4zZ
zHRt$oc;h40z!v$1p$Gh6-S@(QZNG1bCOV68jryH5c>D1c`|m$Vf~3SVQ>s(qIwy5*
zYCPA$cYWHC-;f0mp9=_!ig2vntB@sPJvK6nAL!4TrC6h2{!Gm0?gP}<=>AlLi73Nv
zubJ<$#d)q-xaJbRppIiz{;S4=`YeNl80>s3h=-3uV-RKLVT%8NNaFdh<C+MDe=S0f
zo(nu=$y3C*Bh^W#MM-jxWD@wABnLIiO^m-A9y@EG5v8>C4L!D7I&sXMLfGMM1LWs#
zU_KB0@J%sNtk7U-Kt5oQ^mYH`kU+Z}%zV=$v_Jb7t<O)grSr=TGVG2wJ?CtnF_;e1
z4R+l1ioh|FEb%-0yTGnLkp=htj1s}5U#DWD<dte&^bO@4Iv-kpZ_ePjF=qsswKSbX
zltnQTPor4XmSAZOWbtAFOa^Xt|3=9bG+06)e?g!muKD#hrRhldqrWzYg7&_$5@y3x
za-`yOHj!b4*Nj#CJ2x;bzk?OJa2&hl+06z>9NYpE(4-R8EbQW++UYwq^QIQM6mPB{
zTC9%iYv{emmiwJ*?4f*xm`Ck??G08-!d=0^q#kV$UL~KU;+YW$z-m}^L$Ty@OdNLR
zB8h-x&Eh|vzoU7Ak`Sf*=#GV@4aRI`#4hGnTiKo?^;X_DKYyh>H;-Zb!gs@ed$J6S
z|Mpe9`W~pHsqeH!!NFeO4%l*jEX}){KGfn5Ycl3+!{+I^mj2prF*Iax`J55&FN#&j
z*-w~u$o{$cdZa4~R3@7LZ9c)j`&2xpppPgA1}yL0l0(YC%;*X5mJ@g}qQsUo6@IH(
z-)h~D<q7AmS;|5bAP3lFrS0|-)PVXm25{YI1TS`e5<A0h0{f6Km|s5`;zi(td%D9A
zTcTCOXEtZ42kB_1Jj|C7;G}Z%!RSxUTj#EgTsn5w=lQuS!>aAe;afXEc#H-k<V6Q@
zmlX*d_9jsB?gk)p3C!JNrG&Fn)C=3YK|*6)3@x_>&V@lVs@s(LO84DCy*n)Y`5pJ_
z5fdGgt0BOzPO1+tu+@jU%<9IiT&_Ea{iTIm%M^vI(FrbC_A5tE^abjQ7DK8nn<6A~
z`H~Z|lzJIxntRiXCj9*{X0U%=-7KfW<32}~TN3I8LV<T`aL?C{VbzD88Rqq-Yw$&J
zKr`o#w=qK#SWnBcc(~xAfe^!aA5a=7bSL(6jPLhMdo<AuRjNF-F03Y=aCIXgUR###
z=S1c?iaPvT8=pJJD_b1)+gCH>xVG(^&Jbu}@cqr=u0VM5JnACrKL}#WK20=rkCssn
zsX??<#_^&|Ci2|*t*Gyv8<E4=d54X5IpTCck87$?Y9)G4dm4w|EZTBA-6?40fH_$o
z#s>RII6uEO-yp?0-t~VJ6fh|Vsy=xnWm3olW7B_~>+@3!XREz1sgpl+wh!o9(%;i=
zHGw%{`$c`uhN$#g`?jD*eqGN%{2!COkJP7Sb1jz9lg2|>mL8HDOUfZ^3unPE744od
zzCj!{ss4+fu9CS|hg!a9g7eDtc?SF*dSoE{xPxS#2)mckmL)3hXSi#Q+)NDeDrC3I
z_<w{9|CR5&SfItCE<fuX3A~as<brULZgha1$Z%r6H$LvMASbI!zcubEbi2IYi^y|F
zC*a}G#ETQQalb;Yh=?cHb@}xF%{Qw|5TO`!ORr_egqFllmI2$HQIp46>muwPi+8tS
zjbj`btZDeo9)^|!I%w}T&*Rc>r9?zzGej5v{gAw<78INzH&H!<2QBF>P}Y|-EjlHS
zgiU7HA52WbD&7F05+R(S_=(5xB9)o)W+hMe&7o8ie~L6kDmK13e*I_91u=X~umr}q
z#dkJnLY%jEd+~st%<}_7I?~)>iv9h-2G5<0;C4X`it>g#IRq579@CWx%gnOR2o@j#
zEr1Hckv82}jS#cCL(gqYM*G*6I*XAGUe0#Ec_3*%iA`)LH1ni;n6^6oE|&6OA@Bs*
z&r8epkF0^?YbEV_P*2GtL$LkEZYitNi$LWdliPw?Um<Ltc4=Ga0bi>dfyLBMcjiGE
zNO0J$`CB_Cjf@YkG5n;RU5N2-UD@iF`!RRvtQhxmJq$&*2wG-LHfF=18qIH6(LFDh
z6n%ZAS&=Fu6Em&x_=+SjdM|`!6@qjGrAC)ZY^akOu<tJ;!0+0izCrxn7&z9~|KKD-
z;Ee^CTlJYjRKV>?i1j>N+nP2>y2AJ7d)Bnu6>09(P!hD;-1vWx<ZmhXr=X3TJuUbX
ziM1bRTD|)8^@i|(pc5b;&*XlbkN4li>o7BG?5LZTgaMElgg)SV;M=H_`Cv+?>Oucv
z`@OFEQ>yJdV;1z1=CKupIl}tpU!WAb3ra2iCn%MV#&Q&D-pK3l9e=fMnfj)t^Vd7z
zL0uhUZ7i$#T#-}@a`BrK`KZCk?GJCpT@S8l)W4fR`$@5V&~02E=Gt?p{H!s}Ym(Gd
zF&=ku@~mXEr6luu0BOMt&N#)qJwE<mg&Qa@A3T807MPQfq28`nfp*Y@+VzB##%rH?
zF-OR*KSRuZU8#(0%!$D^DsLX3yI(Eq%l>+AWlOH}IcWJUD>m*WqWz)bK^8CnO}g&5
z)>YQT#%hyjV47lOl6ljaKbpxQ7Ydg5(e-txai2F>hkJpYcdm!@I{`HxC{%hf!#@Cb
z%d*r21`D8u*9%o7k66)B#(sZ1rQ;$+Y#$r3)lc_U1R}{}aE#QIYHNPu^1f4g^U;j=
znq)t4b8MFD$rH<4ht~1iE4aT+4|4MZ_31~><a$d5YFB-AOd0xNFd%rXSR-sJ9J1aD
zv)CB>z0)GGpT-;dQD<CBru1pSF|*22uO^3v!O1Y4_KWJYidhuo^>+S-Fib*ORTXI~
z0;tI~wna~d7(GC9rAs#A{K$%e&)|E#D;@PPNC2<gM`~N-8kWnM@S7#kH(pe7#$Ms#
zSuQb$gvS`f^8%OSP~soDB6Y>tMxXitSJ`0~Qg6Qvc_f!lB}I89a65+^Q&~9o`;KBA
zQy0H-B$#P+%$y5=sIvk8SGH@frC-a&*qg^Sn3;*N``?<$f@>i;SQ28>jjyrfY6iN9
zx)lDN?rxZ11S0M)Ut*JrFP}YUbY)$3_-D!?$Zg*Px}NT*gcPK^ufCU#*X1tExfcfX
zw1YqvtGU1sq+=8l_W$x+0_V1MFpkWWGw#-lcFG;n&Eb5F_`=zidZVP$V2RhrlO(H@
z4@gupU7ory%{pPo8#D&GmKKcI?C8*4$L|)hex#~Pq;L-LRC#qmK4n)r-tSm?Y<$aF
z>3=8Lg~0J(3F>n8pb4AbQ4DKe5@?otq<|q+O*49XD0zmj)i(3zgD+DZTvG@8)Fkc2
ze7l1ENj(-=#;f35z=bM+%s5CR@3sx6B3aQb)65kd!>6{_U6jrl$RpZNWJY~v$$aHS
z2Hg;cCKqR`AAvNc+A@mGyvH01G^W@MfN2N2F7SXWL{l%ZqPzaSLR|(m9b?x9`X-t+
z8CRq?38glkFG-aYFg5Cp2oP}4jGuq*FUlCg69^8(ts|KLp;q9SCsIR{kYqy!h?(9R
z8v_=V&|0SHInl*MP7BW_^h=$cr5-9RFDs9Ai0>;uFUA&e(@*j~D86B1ma!qZ&VON}
z+<l63A2z-qKp^J_0swyDx-V<%88B-soNlW_Qg4s#dq?A~gt}yk^?%bBD2_U<m}K-_
z?0#hHwA5k$HT-T#KY+;pv@8I_BeM(1xP->BO#bw-ohc-JF-RAUCMQT>B)-d}lr2ZO
zAJK~xin8TERqdfwxGA(<Gyn4@IGRNytmp-v067obj-m-bn1&2>@JP4*ex$0dfF^%s
z!aJ-zTGX}d=S$_q3j4;wvFZwOg}(S<uY$lEAU)g+=EkD<9gJC%<fM}atL!J{S(Cr~
z)3wm|^RHdYmNG@|-b>LomFq54KiCHjqsTIv9#|O^x)OCoO@P$++Gk2KHd0FJSBB_Y
z+H#9FH+4Gl;UdEuaC9z1BiGZ>hY~y$kDVJ&+Ox>J_c6RH{xQ0bhLU3q*KY-LNs2jc
z%CZ<Hd&Se5v7xg8Zf)eL{tE`V-!X{;eskvE)u4SarA6f7K=im4vSC}3CExd>GI#EY
zu^qd|RE|m3R!CCy1>V|*t-3O*QV@LSI$-cscSQmBKy2BgT*~Ht9XPS;Y3U7+$m_$0
z9WMRp;8SmF--_9alJsunMb8guyu_cDtON;X{-eP1J2uB2(N3C5MXymBAPIXPK-O8r
z3UDR#g~O&%6NG&;+NaARhnoGZ>)vw9|Iw2X%eI5DKUUoS3A$W0I*)PIaO46a<2S~(
zQ6t@#?=s!4EVNx!R$!#(-{;W3lD7Yizkwv$4uwgI@^)sZjN?vwbNLMUVo%*~1|Cf=
z!~Ap22Y+#|X#0bmH{o|Qz^)bMB5D(lIXo&zo?|w{)}vSZtcDGGJEbzxeAJ|OlCFF_
z4h0V%QAPnShlk4&7@X&^jf1p*K$tv(Z7g=t36RPjV+oQHK*Z(t6G@c>Sbz%+X7!p1
z8bLpRe)=v%153BG0k7k2^&JrYh3hj;;}IZs8TpJI*C{+216c&x6<5qZAgstmJCFqA
zK~T$Uu%H^$rx*k`d%uE2^Z)GzR{AJH&4(>tPA9ayPFEKTJ(X#T`&1;!I$HOci3>iT
z+pId->;08nP7~vFXg+z__h8uGsSefqi77-aeGw?xe>r%MRTmlvHWS>X^t$|ulWGS-
z%@7y&A>tsxF}ql1)weYv8p~6<EH*s5{SQcZn$5szDFeq!g{(n8jdQ~pd(<{J^0YY_
zSLvRgFoUNRo9Yf+hu*G^OA=|mdhO-m7zaPG6WST(L{Noz^IV`a&1gz4J>`C4zaP2_
zaGL_@6l|n&n1^)ExR=DW7i?=4&G13_Ce{I(>7U5+oy?19#VO(`-5X{`b1IKL+K{ZO
z3MzrJ5l$I9)RNVbYcso9lC4zvGV1QtJVz`1rLrs2@me0ZWNmdh*;cq<B!}_FlPGo{
z*`pOmhf+uSEcwePgBy?UgmwqLzL88%1U?Ww66(wxO%1a9A=HD!*|hB^sYdr-Kls-A
z&Q#Tbtx{D=WG2A=u~)DlBcNL|Xu`DnT~Tcx{(~OcLs{JMB-gpA;Mx8Aro+R3Jc73)
zMGj`uiz$$~Z$|l#e@Tgn2#3?dc)Sr`8WjUQs9UDC2hZGnDl(`|l<Z&_fbk8Y?Fv}_
z`GLviGgPQ<#UF{-PZ;#os@jaDROi$ECAgabuY0JZi1To+oO$x%(M}`&=MFoXIx3V8
z9+qAvu#U{L^jhlc>E^-M5;~ib^R;FrTIBg19WkLMPa;9PY!+EEmQcepp-}+jAW>1I
zrp-@be+)eiNCe)C7^}o=8M1{FyITF)nzoZP{Ug2^e2)L})!}{^r`>bi0OG=}!?@_d
z(oo^nD$N*6u{};vP;ELg5s=Uiq#;*kj(_N6<*_0|v@!iy7d+3<*zhIicP{2;w8}fv
z+#jW(^svke+6+=wpd`IIn{i!*I=fM!Y@UkojSN(t)ZELhI-eL{wnc47d1ZB={<)5@
zB&ioy%vZsXW*rO4eNYnXB<AFo1;-zkJbNTWlBxp`@x)S{AVzR$&hi;<|195iIUyfq
zicE9|@>OAY(<q`>Q^oe-nz%a2xC}rU%hrR25;+vYo-z=Le3~1KC{KKEGiUK6)AUV-
zyoYAGp=;cs5;5-N5IFR8nN?fMd1q~5dacyWAA?%srZ<v|lNkFUv}l6MGF&>IziBW#
z%T^#ZF*d=<DP}Dp3MW*9q!{l*i>~eSbP$Y>JH2-6h-37g5;u;=!UAUFo^3|27^_k+
z_IokFy2d+Rk1kJhw^bX<ZO;$gO!)m_JYWT>4q0+w=9h9j=5MM>m$&<qB}>H{P*nh-
z4yWTAMDfu#+#W8jWpxE`KfOzV@Nu^BJl6<T+L{ehok7pwLK7>QIPK%lV|_0-JW^*(
zu4VC(2`)@d1uk;!^)ymfe63s~LReE{0vWZP@XTk#;_mGG)FlzAQ?%!75WC2Bxmf?z
z7{=P9c@L74{G8QE;Ny&IcYcZQB<+*%xFBA^P)dyeoX<~iG-I>z_H-0wL}fxYa&Jja
z+xg=qN;Tczv>!}D_!PUT-yUlPJ(gR2L>j~u+ZNGH^8{}~(R@I*Bymj~Db@`mn90Y2
zZd3G2O+wpCMN8#o4@4NHr@k9<m_mF{ebZ8#3{AE)ox+vv(ZoSDuNCJj+KxQEk;(D>
zGOm8l1869+EI3S^?tUA9*&b3chO6bZO7qxW?1MqelHXQt<&|ZX*B0;Xoz!}HFJ7M=
zVHLVniE<Bao$I8=+kU&?YGj)epKWiK&@2jBMv`yRp{xjTumiKD=F)z=1wQW#^_t{c
zeG{zuvL_c#**+0?)aKGElq(%?{Igng1vywC!rG5Me5m|b1<1%rxU61~kOF(W9T~yo
z1AOW7)BIV~9v5v<+E~*wmEN_P$WYIH70Txt_ekC<%a-dN4{zTlO>+cI?;MwsZ@4b|
zIPujIXZxJy&n(_tmH=?Rn7ILt462wZP8^e~OLrK@FLLGk#J3K))|M5NVuqW=Ki0mu
zCt(bMTR}<YQNh2ywUPM**dzMip1Lrj{B@dkG(>Hgjuz))MbsLz0LNw%F#!w+MtRf7
z>^8>-PUCKLdtcKLc=6mJH|_Mxb}2~s;T6o5s;m?em$HV4efuJ~O#;RoimR%w4$O^^
z_YDh**Vmp10g1E-JaAfyeu&y^HeCunPY)mes?GZw6S>{AbE(Nc#4x~BtDt^#Oj&9W
zV_j$I**7(wsYn$M5~A%rfg~pc?t3{l)!&q<4&C_7zu3-bfvB(q+Zv6xWx&&pDVPKA
zsh{UZuubZ2d2eYStQrg4>QHHA`iFfsS{33mJWwx9w`F#;rDPj$`{nY?c8)#LQ$5pc
z0()!DPqUj)PNNR#MBNy;pE)(1_APAm;Hui3pC$yRG-Td2Rz8>v?77aehyH);y=7D!
z?Y1r41b2504Iy}fYd5YzLV^Vi4xPr`C6I*Rnh-R2g1fs04IbQsJ52)(@2$LhpYxsX
z+#mP;ID6bbJ7X|17}O|=>Z<3NYp%KG^uHF{y=?-4?@FW~Ch|zHm%3~ZmfzT-&r)9S
zeJq~ee<dALT-cZ}ioj?H|B=iSlasW1CdS&%yMK|jMpxy}eG=M~Y(2p%i(wwN#S8MN
z-DYN%vR#AXd3%mB)9iu`jedQouXu_@0IZL^XJt#>i$@z#8<YE6SqsSw3rGQ(5=*@n
zow-%AhaF?9L0}mgJu|8{(taQGtca(Ib&3#+ZgPsw{TSp$d9joG5_;b~a}8W>bU}Bb
z>^WaHHK`RzH&@4#0WnB_8sN4(NgT2B69qZxQYk5GGARNUXX1o(I?!k>Qh!xl{ZnE6
zNkLqqLg&x6(LxM{u(4LpU=Z+60@zqU`d?7y$i`fhfC)reZe|;Cc9EjO(sd|1{gFvx
z)uNTPP2sK{XfSsvuSImu!`W&3X_p_PS61A8LkUYCkeckIC6^GbS*16LMR~m;(v-#w
zpuk}vaU~4Y`&!-fjnl?57&Q2({Y5Df2!5$({}Igg6)=AI7k{CjRb}A$Bz(1{|3om!
zW%QqEnm4$sW1(%hKvn7V+9i8;K7Zi4pt-fq&{VktBl_2Ti*s(sC0oV{QZEn-fMZaA
z>qGJb<;sV$8l7=$;HAbJ>sYzqKmy<k^EBub0D9FYqIbCWz*^DiZ-&zFn9}~&9wofJ
z!gNoB)Z1UXFaEYOZ;}#Sv@0@Ul9%F)Go!1xII!1%&Eo7E2AE>_ZU#OH8hm^jdl$;r
ze(^#ZdwU5WV$p*(%K`g|uBSl~Yu_7@K$XYt*9F*Ge*EKgwFC&k01nI+S-olk?x%jG
zSH9|}MQZlztFrL)@N;V;`nm@@svy#_V(6>H)zn@mbXcHptm;n`x$Ro3+}R@)N8Yco
zS9yfJxC%O06j_?wzkx!OPc(KzD^jY*ddaz@IYL0Hwv%ZZuuY2IL@um-eLqaPWy4}?
zcV=vLL!=1ua{_j?kC_i9{XP5&_5v}hPbn|sNHl!h2n-1X)HvPQ!pBU-N5wiNn9(6z
zNU+K2r=K-Yta)#z=PluHUl?^WAJ;uNh<Yx|%$4#xHg%Tvx_zU@i5dWSxTJAku4%An
z2Bg9+ShfKzqIE-FBG6O#O#0c<sK6CVF1A`tAOK5+S?4kO;S#JfkY<)&W4ju_QTGQ$
z^1>tOt`Kb5hN!6~;&6bh*|koU{V6meV)&9L*4`nW8VFX*w$huXKwE`K+~r>Y-oWKL
zKY88NOO+4E0Q2DouSo=oyNfVYJIs~*RyU%^t4|U)W)e?{d!loID>1Q8EhSBPaE7a6
z(o&TGnwQzM4GS9(!`>wRSgA_gYerL2G|+F*Y))goI^4sYgF$)#tGi1Cdm$$as=xyI
zwcStAC6!Z@v<lgX`+}vb``T-k<JZP099~J6g?kj?)21!`KJSsu(4@iS(nNnCxB~Rh
znURP2DCJX(n<o=xLPOJUM~JQ!zCUL6wnD3k3L7&G5JxVo=Q&aLtWfd1A6S{6)rg=G
z6MTw&ia~b~G<Po=6os=FHIO2hVlF@}DKnN2^1b>9bs(fyIAzKycx5N*bvplSoIX}c
zfiX3rE1829!*4@)3SgJk@=<Ctju%B2L(w8PH2%oxO|p_ABKkNi9CmMMH#2hNz397s
zQ&!dI@qM{2$*)79-qZVLmFbQvlt+0ueZrIM1Rc(}-lvI3AfyDc04yNrU#42H*BEtf
zU`y)VbnHu;{qBilmgLP@I=!wMc6Jb^pDof9m39&u5nmWe=oG8#RpjG4*dEiCssUlf
zOJ<1DpKuS!PeK5aGFdU~f2dLxIL>-5bI+3Rinb>>DiJc#3=}d@TKD!3vvlDZZCP%G
z4pi^ipQ>~zCa}sZ(x44+V30@vek}gjE);R9Y=V8KL-Z8x&hp$uI4l>!7g`6Z^jQqv
zp)ygHCSu$WF^nG#7&qz61LN3}H}ATSvG=GGmR7ut<Qp23KMwNLkqSyO<s~#es&Nk>
zS3GHr?%KJ5r+l_kO?z1rVxc3m_5S%_#~t^pZ!kO~l{Q_B%8(B9_8)fWztvafyHz63
zybq-WeWaVj4YpwU@~mOq@OhW`Q32RwAOY&tWE;oA9@MHbORKrk`kW=XJ~TR`ga#xL
zL4&v|AM?_SQPRgW(Zm5`Fst2!A()i!$4}pc(+@p0di=&EeXuXU4t<{v5D?*g|L#CF
zuaxXp>_FzD=jq(qa|Y+F6Ve%W7SIGq5jjk|K`B@hf+8|`BXE{gttR+>lA`epzjg+p
z7XvY@{16d;(S$vF=-P{2Hz!yBLDqVLYiX6UF^)ahd>bh+5D`<qN@(=_=SA|k`yGj@
z@L`g3aQvg0fE^OhFSKU#IPV>;jsv^|8WLp;rs)f(gIy<~EX+1unzeY;0xme#A+fb>
zO!SAe*09(WQw<se(Br>nP<7!3Hm>}Myv3el@72Y{OYqP>GVSUZuvR{~ik4!2c(NQ#
z6^&U^N^s+<^qs-Fs-tY2A_7|)Tl^+#EEI>0G*lwY%>2HmQtYJseV&zhv=5CQI8@FJ
zSYbg3Go;=!_TSsY3X<>7F7Z9B=7u`NpFY6zO?K7KHxGSGmnW8VT^6|r2oeI`NzKX|
zQ}=&bHVeDwR*BDuPGtNh;LVaX5e#twpw#3EFnC234_G_{)jo3CuRa85jcIDV6$K(C
zFAx*%aGsVIbFw(YI)r(wm|Ypm7Um}qN0rY04h1(eC4b4f&+S&kj25&+cZzw&oLeuA
ztlw44|AO>S1WRNgVd}&;6lLuv;z7vBOG1wIm59rgw`s9aN7<i+7PM!s>Rgq#pClbV
z;QIsz&U?`vJ=bk{zNtUnhF{y@G`G=rYo2sGt@pBv`7`CA=#xtzI0-Qy@EBm?PMQMg
zgMPA3K#!N&{{qpM7eGjm)e5JnAtklMgujApe<^5G8{%}~+ZtPg%#xx=WEtPDGK5^W
znmGsdol5wFlQU=)($a_Xq8#ZhX_<0aVmrpMoAmilMSFbp!~IE%_nPkN=MTWKLw@iO
z)cS8bIbs^N)x=&8%FUo!9jPo#kKo8*g}MW0!C-ddG|(wC(gaEvv=eX7e6_doCvc%v
z_SC35EvU0W%GTPNO~OKnC8Fzb7iAJ-=);g;G-kg7ESLFq$o|ek(Ie){@baTP=F_JF
z0h3q$i0ih@3$*E=Zo|2t@j?%PX8HanP^%fKZae{DMN&<^jG4BbxD+$As%wgRvYPB}
z?k#4noYNa<Hw*mYssktLj$<NV+sUeVLhv-sk&WdJXFFIT2&rd^2|O^8-7Eiu6*0%p
zDq1G#HJYmGzGzvI3fII(L}l+ZtMrL6=mL@Jz~c~ZzZ!Hy-XaZT1|tEBt(?=3y>!H(
zQkcB6?}Y|x>2XbXStp0p1_k3(<Z!q^c>Qca#ATRXwScE(640T?tJzktc{#&s%VPET
z%!U>c#zqWa<WE6eJ4iSnEdB^+cH3A0dR+Ii{H;ebF)7kkWBS5*sBo#HO0MJ!Ub#lT
zA8C|JKZBM2Hj_r+2Nare?CQ$f8AWTS$}nJ8;8B4lgM2Y{X52HNMiqXH7Jc(Kg9E&n
z!*U81L`pADJ1IRQL7Rzd<)M}yb|~=LB_O-B|M`G;%ft=<pB;aM=RQIGD~w23XC{3O
z@Rv!w`i*#-CtEl~R|OfSay6i&_u_9euxw<?0bD2-;|6B#2|p=qv-$Pdf({83*d7FE
z&h(GpXk$ZPpH{E6P|utJZ~M1Nc$Z4)b}wPGl(+a|?hq}w=#SZ}Pl4&eo-aKAE-DkA
zKcy-q0OxL`!FmxEDuA;;5@=+uznj|KkQcTF=$S1((}}-Ouhmzk;Zjh;0as=dh5l*A
z0*M-7r@$xD0*2-PLS}VWct^rg?Z4J$3Ry+JG;cFh8Px=#&lU!mmB$CcFI?7{2~=xa
zs$pD50AMB(lp!2%6ZIDfGptHNJ)LwPuPVN^BIz2#hWew?a4>p{mzTwdeXU|Bre#(`
zRm6y<D8{hcdOg&AX_ep$;=UA_gV#s-K4#7bss0x_=s#$o|LOm+w@lOl9Wk>u@OZS3
zel3Rory=T|hkTF=LLL80Mo!D8`Df%&tK=_qiAjS7+KU0TX-WGHS~l@m7OB7SX<u!L
zv1u0h9^xaIG47Q~3u#~Dnp_Tg?2Ex;!}gQP7GTknj|EzdC-zXlh`~5RB3PmvY0xGH
zydV8Q*rHGtfCY2!srX}{J1#)jA|b~giB9s@zmHn}$5HKn9~D`4zWS~=miiY;W&EI)
z(Czv38)3<THwB^hTiYO&HFQo_wKyXwOR>&!+7S;Z9x$u>XVOUIACpD}jm%NZXKGRa
zuWJ6W{ElLJRv&yCcCV@oIMtH^@BP=bV5kV}|A&-k2~B`1;F^0Gu$K6Hk{znF1CUO=
zGcBBJOP*Vf9<AuAQ;w7vFy5_nw<p@Fc=!)o61=a3>;O$6BFy=G$CoIvV06E7v${Hl
z%6imUEm)=Wn;NKR2FN~c^*nXFY$x8Oi=N1i((tim7^NVS3;m%fpvAPEH;#ImZAUi5
z7oM~0Yn2dOC7hAxoQj^!E;CT-3uI8Pyl<~*eHSie81&Wto+<U^EkOisp<jD-tgtG<
zw`Q6+V%iN3DNBr9)Wf^%5(tZTu@HVKnaQIw?_Y-o#sN9(T=}0>=bAOgRrZL8<)ctp
zrv<d7_q^V%e5?r^66r^jQ^j1PhML3{c_e;+^Obqz8mHZI-WCZ85t5(Tt#uS&n-WKs
zbB@-u(gTb;=!dPAo%gge<%G-CYfJRn&0G{CUX~Q)A>)qL9F&%FC(G~F&YzCE!v&uS
z9-asG!WbYniyNNff>G(VQ!}L$XlL!c#L*D^y3@Yzwk{_t>Zh2!=>;?Dx6>a?mLPh}
zLOjkmEM$KAy0;M2%ai^ODEWX~dnUrllh(zG52**Tc#CgG<&&eI=zA);O*eU6Yps{)
zbe>A~#*fkW0Ch>dM(tE_XYzY7>Gv$LmZq_{OucMb)CwQzYh31rO@i6@JVO!}(m=Pw
zFHiKtYyEclJ)8)4>Sq%y{jWrxsvU-eq*$|^fKXi9@79*cWyqrOl_%XhMF>KFOg~(S
z`8<#1n^{{7jGWD`$Ich@1J`~R@p<h-`5`?2DerJ<n`~SR?ys#nEVmf9C?*M_h1IOS
z1^y578TV=KAKXjj=9z1F0)sLW6lgdl>&kF=S}faV+L%tQbY$0~EjZ&h;`txxDVvlO
z`mBacv^FC+BVU|y8+^8O!?QaBi@m5d7{LWZf|QjP5MW;Xj}I%uQO+u0g0Ek->;R!2
zFFV?n%*zab&@zq{TQmyif3@?~410r^XZmbeu+V%RO_|XoS2Bn0f(PON5!~X91HjL^
zXV3{FQO$+H!6Zg77)t|EystKvyYiLO24Hy65;P}en8C2WK#czQE+xyu5DL4dV4w7l
z%~|Z$Sy^qX9djJTn|H8ZfJ;o!dkFJsL2~he9TR>1$!3bq_aR#c@*n0K?=YWZX-?qu
zzaA_PvsR8&j?uO95c|V0@8OgK<l^hOeClPMFGE2r{f?m;h4$uFVq;}E>%ox9Z}S-!
zbT(hSg1mcqOJDnJnd3gBR3{Ia4R#oL?R`V*D22Z+2H)_pNz`am0GM1snol6mp6a0t
zv$RRz%*>FRz`Qra9^M^KKUnnE#`r|-{oDg3vBCc0ebXGEg&-2U5sL>)8=k4QkN&D;
z+a6r`iOt9)>^mlXRb@TR086Y~AP5YzjJ%;oI8?MWAdOCg25K9UjCTVS*ufklak6en
z#zaQHHL;;jE-64jw~eudQ2MZ^a1HC7OPi9DgcEo;k>mKu88L8&&KGAJ{F#usYZ-oP
zh|;3nd5ImWG(1(|rSKNmR?jCNQRT{)B65{rJIU$;txo>*j1|8I;R~~a-}Y{zo#h{~
zxP8(!%ujlBanBiyM$XEMuTw+DhNbNu;Frf_VkT_s13Q;=_DJL4Iop{i6zYAydggZ)
zSbM?TVaBin1Qvz0zq>G~VXI8)<;!|d6v^6C@mTjIiCbdFW>zr@$@Ma>BUFq9QC|KR
z$l|dzpvO<pi=N@+2jgt1=sgtgJdTP5EIktQJoI(S&^cGb)D=Ol?I&d|Kr>p->mBq&
z9{|eaCWPVPnHiRpDOvVTFk7<27#uJz9kEQdIbJnzc-Bdo6p4hlWOo*lJM0hV5;3b;
zRWmu!H8HV_Poc#N--y-|);bDD2l?rKdyJk0B}9?<BL5uD^Y`H?KPHy@Cg{~4v)0#!
zW4OqZVW><IU}}qjh`I`ZXE^d`nZDT`eq<42$F7s0%o?Gu^F$UqlsNy-rT+Qd#K5IG
zMF1PpKdu$sK7}_w<H@fDMzLOm8*Ga|d27RQ<Wt|GOOz$0V<c!RXwBH0`Tp$+;a{ND
zGUEG3P2Q|AIF=DO6JojA@85%TwV!~{vkTY%F)z&h7CF7a2<|k+Y-pdY>6tD)Vl#SK
z885546k=Z5Kthev&&HI&-CT}wSMxAMD_p{_m%d5yd(RoAjvRlYz~&7>1??qxCRi%#
zo;o%pXeSEkbq@T|@%uYkwEnMuwuqk$X`F_aX7p~g<vr&qHM-&%mSs1okXz+;6&%<A
zTyz8UCzz?=d@m3NNT0T4YtrGEYdMkcY|W|{*kJ57MUZ3pp<{l8Kb6dpQgJ5qc*M88
zhd3%tIUK^YYsy(y-7rP2hx4ZN!<qV~qvQg%BeislCUyhrwmdIT^w(nlt%Se8Iec4-
zlSH)<XYQGl?W7P|GxMPfZ9GOBzQ?oZUJscbqD|t*w&#?4D`af~cJ~A)O3~{yr5sb5
zRUdex8z6cYTqTBcVfA?3m27bmR<UDEH<E~V)s62~-*trvxBR@OvZrW{i{%|AT{9X-
zm3IDF+kST58p%v5Jvw>w0oR-{QHP+@Z4lHW5KsmA+Dg5_(aTR@$At0I>5`EOO0obe
zN~O{heOeTwg2+7%*z0s=bbg|#as^&ZyWvM9RW{cnUz~)E)n-0)%D*N*jw(-n7`lS<
z>$da^ik+2oLkAK%j#<VEdtUnq1s*?CaLv6&tfTW7upG;7mKk_3KiTwPge~23`&&%_
zltc!xUgQt>^UJ*Rk~RG^c?C*?=BF>}(I%s8vn<WBYA=Egm2Ew+yR+V|!|Zd<gtjB#
zr}0~Zn+vb$rRldEiJ_c8n<i!qQCsiVKy}=C*8OJmt08zwA?8D-TJd`<7F%>MB)2p`
z^tBh^U3Bf%cc=P>XzS?8sP6<}u{q^KTz66@?mDt-(p0qZxe1bwN#pzQ;^{(h4(!k^
zsxMo~)n-d|7K|r(Jh<J1o?5ijt0o&^IHHBnc1qpOgj1l67<!X;iBP+k7$Eg!>YBpC
z`h_IiAMbQGm^WhYny~Z!!{hqTIO4zX3+-B8XFjNZ3VK|AUi9t1@V8lp_pHEkAM^s4
zKQs)wKz~J<wa$M{eF?p}1XfoWA+&3s<uGMmLJ<3(;=&{758U(qC*0#sDl1=8MkVtD
z-0qwW&}vrB2L&e0{x`hWwD-R;rz2j_|08kAgnlZIh^;-*k8*Oy*CJ9aZp|dKPdgt#
z8yM7<sv7{sxJsV(3;1*0eTDmbY|xt#FCu+41dM8(SOJM$w?wf;-Wy{A^f77fZ2+B!
zEE_<Cyg!YQ_it-fJ-P(DCYgQmjcEaFK2w(M+eExwaw1i!9wl#b<-Ijy@6*a2M2B3U
z-Jk=N0roI6<-HhEUkhkl6&RzS6#&iWN3Y7ypC$|+=yxr$OFzVx+H}=o9G&ebj{*`i
zIX=dO6?*fHb?bW-y%Oh16EXb3{ZxN}&JvF%ruD|_`pbq%%ghQD$~g7OAbwq;REU*M
z-gj|pM?XrIutgGB!Cf-EFl)mzobAaFn=^Y)LKP>g@DquMf8fA5q6@SmCdggE0ZMRv
zL`$*qWU49BKyzloSo%1=NrBxM^K2nlmUax1Tz%rTahKHtrC`L{kV{o-2F!ey!pG*R
zG_q5$H%$uB!DT+XE1Ts}dgt>fz0<c|+C%ttr5>y$kuQ_%4U>bVBjaR`aD?aY2esAt
z`(U`Nk;*FYLLb_faltv-ss{oxH}Nt|iKpBfg>Mf~uy1o?C;J++kE%YGTD7*=B`?Jt
zg0VQnFi6A?=bOrdp;Rcz^>#8-J0SM@4NBNUh}9Y<5LNvTJgIafaVaf{%0D&5g|~5d
zSBbADaNJKf1(fA$&Ll^sGxd-t7qJ2+q0*uD5WbSiUT`mERG+dg`wKC2jyesHZ|cV&
z2Xzzbo~D@3mTUI>*)^ZDZxvFPRXtxsNJ`ap42b<MEL7VM#P@frMW&W5?ULCg*z3p3
zBHo|6{VsWrLzr0U8E5F-IjZ$czYj{1owm^529Dh3WKH(KJxk!!5d@QVB6@vS4s0>T
zPBqW0B9$Vf1$2ELBz#lv57v6Ky0K%2o_AW`zk(CgrC(f(v@d}1GPFH!Z?Q`;tOGT&
zbp2*9w~ht5??~e#e38Z3-jgq;Y68Gro^5nZG0e4}4)KKCSghq#Uuu1#HqAcqo$U+c
z3x%0cI-D5yvI(*;emfn=H`F}u{4*qa@El_V0HzHFc_O=cQgEQ#X;|>~TjDqw<$ET8
zn1YP|qY3op^W5Bim<;oiJ+s>U64CSiLZp)VcSTolnuhWZTL6+>E7M*KmMbHLPrnYW
zOG#8Ds->zmO(u?cqmBXlq+_i4mQALoyJLl&`XuTw7eEdu^`fVjNmQq#1R!G1{$%Vi
z&M&IEZKzs?>PPK%vFyDCG{t-Y0Yzvx{fFqQ*Ox|v?Q6RB)ScTlx4PXuY+th4Iy&Yo
zr_KTkT2m2z$oQ|`zqO`$-OUI)*M9L>=_kxA*nbF%!p3|Kv8)MwI_I`~Vs$!5Frg?Q
z<O-&=dL5d_K%#7ErhVT%{*2Gx#MvTFMrR4*i{?bffZFc@XW9N%P=MK=cU97DH5PAi
zQC)g0qDHkKc{o9p{q3{Sp3O>jS5@h~>COp%@Y?sGPjHsXi5et(Qv$bv9M{*-Be^mx
zZ;s_U?lj~VC*hUuAj*YT@@!Y+_M$HR6b^<<QG5@I7!tHVc%W;?^9_K)-fGFzLcUC6
zyWO-oxAvBA(^_^~WYT&uc$0Qj9^6h+6U5|gvuc;dYA3ImIF!)G(x&6Y(~+aAEqndp
zT3LL5>OLT|_Okr|iZPgj{N@#*7^?oBlATMgnyYP<SWanc;4pq)hUx4c`y@R<aBwXE
zH{*ki&Ey(e;#6W?;WXbxJEpf90wF1O=w2#<%SYInrxyHp=!q^3EeF8-I#KqatV4}r
z!y!Bo%be`V7rUW(4u)U0q(Q(rDS}`jiS?aqQ74)J{WcGq_9s960$AR3Yv}i0Nd$}}
z`5Ia`)#{zAusnDmLhUI})Q!~!g1YNZKyltUZr!D}J@nD5ZcJ0QtLxMHL<wHhpQRD+
ze?|xT1+DD>Xa?@sTRdRqDea2+Ei(x`klkBlUoXYgpvA4W261$YP{}ch=vaKR=N4j?
zZp*TIq8LzlOWi31#dBJ=%Zq+Ay_Fzh%_4EFUAA>et4GV&$ykKG{D%nmV-3Uo8JW0&
zZ}z|&kQw$+kE^;cR%loBIa_%?)ODM)0_HCz>2|*lLLG|OvsN=}9)8&5WFvc`$l>te
zWrE9}MRS`2(fAjr7~nj(R_+SgLXj|F%V6>r>F%M@R&`8ec^5wV%(0N5af5}#*;#Oh
z`&_N<>Moggs~*Xnhnt%qa5l#JPN(t3LzbvZ?3rL`&;OVuXyA_y6q!KbuKIWo+CDRR
zMgKYFCr>(Iq9oRvr!2t@O5&Zn6V3A{W(yG90^|$xQ$@B9MCAktRmBgfe|&cBuNZnf
zcqDTy!(<5P8;-k}FO>fT$p*a!MuB{9s=Mb?k~gxj4IQ{0Qt9;a@dgpFhU5srpGlhx
zq&NE8aDYEV4~02}EmK9VbF#+0!urLU3{Z$-knAwzIM#0zOC|~Msp!i9%~W?NN&~*9
zN9CKJl;WF_-@29eX$7+f#wvmX=90yZBVN-3RCOLw_-l`sOriKYI6qwT?8ifi%a=+^
zmAb}WR3(h*)YiAuq%v8Fl9s$jvP2crNCgMsRVsqqDv<PD+8PYw-ez(!26GY;_9qc~
z*es#cVa7rIfa#rRa1H+B%M2oEpITZPMN5&N<TyW4Y<NnrMES%xpr0<euDuZJbr}c#
z%&u`ifQU8E1T~T*rPrjo9NBE#=3W3Vj-EaV3{msrbOymiQCx7Qg`s4#u~3y~@vC8t
z%5jj_Q?BAi60<X;{dvgm-7{{+h6^k9sd4>o4EoxDkU{&0S_K#o89bSFVc}Di0&!mq
ztsRiRLSlD&UOiVO?2&-ixGJ(HTfHU=%>!%;EGa-nNdB&w`lrUPkWPvb>l|dtJf2Uv
z!RT=v;R7t^@kSQ~3Bgn?<r`jo+*;u)zbfjI9~JxgHO@SQCVZzmF{qpQ&E^to8W5n<
zrhHdlx_!TV(sKVYQ?@PzDrmpZ>@DdwDfXKJt-LUBCS_tpCiR{qW%*}1u1)Oob3<yW
z3NubGOTffBTZt$qX-j1Q+)zFwX=mM#_IJ%4AIfHz|M^q;-xar6G1bSq8OjSqfWuV(
zf6HM8*djmPE8c0{w^{=-OF}zclUGYeHQyJ!E4MZ9`x<u>^6)^Yx?|bzX?>r4a`(Kb
z=#pFNW<8HA6T~S1^9OgRT7ezf->><1<a3~fuHM;>psfZ3u)D_jUhU}a>UU1-_2ILc
z^*)MaIhcz_XbTK34W2xH&Ut{{9tGMBzj16I@ijZsQ-DBJoV|~9*#91n?EK@=1z6zW
zGyx~lkYWBJyodry%Y@c7{#%@0)P6+S?^P#M@-c4EoDx4^WSZ;#$J1!~^y5ZmEI?W_
zEZg;WrbtZ9JhVvJK20K7@pA}`?pWHN%H%Q&1%51)MT%Tskk`UCv`aJtvw2ayDdfqn
zo9W`Gd9gpr-Pvw8WPjKOV9qgucw8GpFkiS=vEp@vZw#?83}IE9`5TBzGsUq=5Ti2$
zs6ILVgbm&u0U4AX`cfPF{RC_cO6x#Jm9e%`&->9wXGC~5`7`!1np`F}!|$1Lf0nzP
z7yX%n^(U7{VSr`igm`@lcdQtd0-wCKKA`>f*>1B%Uc+2qr(wOUd$_L!Ok0Y#IZXx@
zBWU-COX2rF5cT9Jr*WH5T&u_>K2uu-cFPw|tfBoCJU_&bWU`5ygbAtDrfkR^=-|my
zIiA>=n4ai~QSp<B*OM2A9Ga8(84_I+<AcH5IC|1^yXi!}Y0ZitTeF>4O1aRiB6~gR
z7nV++rJE!zC)sTBCfze6)z6{TJiI)pVW(RuZl(iU+78-2t>5}~*#rxR=*sO4f-CkC
z0mGXhtF&K??{DJMQZ**r5BZ1hM=ENn&!4s*oRs)grrUySX`vY=Ko;qX)7eA(+y(FY
zm$k7SzMciV+jfN03*QDGY9(#@!26W2rJtJAcQAjPw~fBsKcrjfLNpPMhaKX<X${$R
zSs3SZHyvsAr)CFMNO{Y6TKv`@JLV+dD&>d+pifMIjEK01<z9H7xw5q7%>Q{ygMk&y
zyXF_=CZ1ZcZ7e9W=WocDfBNo_I+x>v2hBPV)K-+I^C^6KBRj>UoloU#VKPRG2AJu|
zg7DdAN=#q!GR=1D-B3<x2bAFJ6o_jJ<cv}Vnw$7B37XM)cg?I1^-9#XZVJdsn5fZ-
zQ+($i)nfS*s3cOI^?|N+zWZUV2K%~?Tt?)f5F4kofLvvNC6LQdT?Fik*Oo8uO5XZL
z_xRd|r}(kEvZlbsc$Wgy-m(M;z$PHr{&pYN;YYczMdIe+hb0!yR8GQ7QsMi-vV*&k
zNTsuh^}ZvR(HwhyWB3nMif_6<-)YNk(gEc=du=+hy@%R@Ht+|Jf)c#Y+>JG^aX0;e
z3qbSf0kc!`O$|OrHE<F8O>rN#S|uAuUVNgy&>>_<BRpra<nZvJ&9V+oiZE}mV6emq
zl&c!h2XJk4`!WuX@s36p4Dm!6E*(Dq{9#WOoT%gTD(KzCm(TFhhqcuqOkOebEMXxD
zGBmUurF1_4hE{7W`V|dnUW&SEJgJg8*;^2{n+&rMKs6?O{H$31HrLO2Q)$AC(xP!S
zjwIOS2sU|7=5JSfx)OnsFJ9VB*gdmz`EX>ZijkAcb%2*nkhN5gL9}>ohBZju4etF<
zeCaI97#+ZP%M0k5Q4+`^^O{Uq5A!&q<ftqWS<1fW0Y<+~v9I2hf*gzkXA*fP0<yZE
z*GR^um?9&%GX)-)Gw7G0@R8r0PL(yM^6p0iAXi%M<cixq@AWhm@anLy=)nw@%k9tt
z<w!k%hm9wi(W7CB$&S2fM=(@3#tijQyUKFSk0RoG1s~!ZWP1LM0>=v0WSA9I*PMul
zz^Z)j@}*+b1%>l~=2e2rD-k`<8IeIJODDqGCfhgK*^0*L>1Z9BQy1E3vS=X~4~Q}7
z9%mCcAr)HyG$F&w{TB!X;8whkJ@a*0tK~Rmd(p6}Zq2ttn?|hVv^%-M;=--Q9Q{W2
zM?ju*0lkV<noYGuWnccM)lg^b3L&&}pS!OY{=kfbz&N13Tn{?^hB(GVnN2vg5$-{r
znlP!;o3PI>8{r4!;bL_?V8WW*<`AR#9V3op3&q7CQPzg-nnm1LquQ&Oqb(zUY;88Q
z*{|t-_>dq>^~-~1OR0Y|$o)nea|H>5W!;biOK)cg&_M<4zpw*Y7Er1uv51k@^b}WR
z$ya0eb*oie-x*6$@<c7GUhD1RDlBnIIOTu>$x>4Ni&?m6EHJ*r>GEwzh?kX;j3XKE
z$oP^zIYQYPKoQG-Z!k&7{PiI%OmF%ZGne=xt)Y~pcb_cDs^fArgWe{NnM%^_?GFMY
zB*49B!-uKE0jMSg07<rR!<UH{=!r5+)WA3rbzaa<f${dg@5GJYn%2isDW}x{$yXIs
zB2NYi1mS(GR+4IE5=Gz`Y{u}wH?u3uRjnn6b}XUr#t0R|%`^ovPMrZO3mV%>hA<hX
z#{by6pPLxTdO4n$$TOUXb@#JI=rO%zA_T}(3!n$fxpsU5+Jz#qnUQ+qe}slafCb;*
zZi1LT?S03Q+Wt>|z9}=(43IlRD8<kKlb)}g4XVxKJ83$EcPY=64;x`F{O?eLZVY6t
zfOT3))whHq!xwgz<F?q$YM_CSh`LWdGM(OnNss6wR}y57L$)r<^Dh>nC3{WjE80W`
zT?+T8PEGnLhaFb9i#!$!!BUZ^9+}ZjZ7Lfv5J3X$qIxa424O~s;)Zz<#2I(y%`5Zh
zaOx==Xj8IWutZq|^8*%o4`>q%YBZj*D-o2{72(AXuxx}GL;LwZlA!o1CRT2RYw1T6
zH%XPZukF4wE%nslj*T|ZTwauGm8U`-C5=AtN-?{~sm+Yn&Z|C_drsSuk)0WwQ5tJf
zI1$XX5I9ry<$d_r+vnzdQhhPV6#PtQJdmbzZ-9ECsdAD1Onn6wsj)&L%j>Y(U@)ML
z-)ec@GLh#3mWq-4Ry1hg3zAH`lQ>BdSU?&nMXb?@=T?R=GQ3U>1v0&2?9feHyby1Z
zfJ-D6FV~Or?h{y|sh}EbBa1?03)Pm<_F%u}+jG-iGq%751b|iFAT9QPBpxg)RQdqz
zzuS<7%+LK|h-|zY<wN?M&Tp^O`MX=WW`yF^wcZR^N7<&BEyXBL=~wtWAt#BuRWF9r
z10G&-4-KGDLTx&1KHxH34sZdK4yiP{3~T;$cxaJezhxfZPnTk2IS!Ke6Vf_a+o(l`
zblA#HGGdhK-aeG{ddF7gYm334jkWyw*6@QczWWr4ho#!~f#K^G0SSIrifoKqEB$H8
zYinzJmjExq(j>A)<Ka$lBpc6Be>TWfBd}onf4T?$-_Q5odg6s$>*Bd?KMj~?f}t<%
zN@8)5vCmFT!(rcd>N#EPB_gg3sFYa9qR`eNloVm(n6PNb=(d)?_QbM&pQtKbW&c#<
z+npQRJM<D@iRuDcGiG^0XzI7j&j9-~kE}bS*-7qy{pS0WHYW2yfZ%<|OwiR@P*-7X
zb!`m~9G>&Ge)((zKdX$(7YI{R-oQpB{E0D{W8DYFMTI@4x7%wtBHhc0P<*qiQ)Lep
z_e6Al&3EL{viVTk$guh6da53v?K$IM|1SN%pg1YZE^a0J2)*nv4zSP?p7aoF+SOK6
z*SOBqrJPi*e%ip(+T<FcK-1gTs&00^5=l81qnB4Dt3;nnGg8MlQnIP7tAAV3`BaG-
zYOI?ENbs`Wuhg{mzmu~5W;d@CLfIoiE?hn2#;NB+64!${fFIu<h_Xe0u{w5$%O8WX
z$A0IbLCUk$tSoYFq+*PvX#23+67(&dG}EPDoLR^pUeQ#6EML!Wtxv)FqMhd56RCt;
z8q{OT4x0h8<{^E1zc=*a-JG0wQV{Wr(mj!RC2HeMq5XA+!OX8fm$C1mHA`dfP?q0z
zZnWO5WfDTx4w;NTM^+EgQ(6F%ge)6v;v3mZNY2w6GiL7=JG#t|j7ilB6GlT$oE|?y
zhiRC##lQ`_qsl|=v<GD+f!yA;{P3-i{AoU0S=Qtw`X9q3(y|yNu$sFB9=<XN|GbGT
za3wQ!=d@e8?3PRPTu0JX-V|t8Ccr95apc~N|2b-wO0VjNZ*x~k|0!!;O?6%M62i#)
z^-}cmE6lOkKOq>`+`a8JB`)5n+7V-&^ByBDI-gRE`tJ4*uU+RbJa-d&eW?#_Ph*(h
z(=F%bwuLz5Ep|I0YGFEf5?C)EHeic;^qbyEqTDuM+!>O+HzTDNWJZhqUXPHQCM5zr
zF_5>+eXZ39&Z<VI_rASnoe!&H)xBTA+gUu*$i4&J<rwKA^z2!v!+w!4tf8iyP?kMG
zMek8VdP!7nR0SwFpel&&0vf_hKi1P~0|b$^O<;y!kkA*+bBgx;jL<cA3#I34SPd1N
zWI$=G<Kwpq8n;zX!MEy~BWw&=lhvxyudrJ85uB4pU;1zQzVhaaJl&0ToIJe7h~{}l
z4awtX@YBX@bwi}bz+dei_|q1QI2d2+n8g1UK=234KsvID%b(w6-t+gKVqLwM$Qxgs
zC@amSR!8VzaXbqbr@;_Wxg5CYhm!2+iN3N%AznoT^_{bLjDfXINt8q=_fGj-X5S9-
zyV(?GOS=V?96DsOdRPDqAA7cHOM!;KRCbeCcr_yR`4<`W8;=fJVRqf=T>F~ubGeuq
zs~*}C2`_>4pwyMxn6LaCR1n&lTfXxxpUJwr)_QYGsM;kIsK@9M`)~BxAvyW<v=`pt
zJ)zXg<+OTIw%A?+i%b%3W_YGXTDD%owTn9PF8E0a;?yuOdMJ)GAprW=cD*K%Uh5R&
z^Ur=7&*9q{^OLDpX~i-9-ZmT7M(6(vWZ##7@ch^*A>T9cR4IA%Ou*UTaDU%)@)qK+
z^=@6{^LLh?8xv*jyI=ujv|-IE$9ZM>tfe~d3lxN)A=C)=!Mq`QJ?#-Eo8yxo9jl>|
zB(S})Ip5K3Zhq=zRg8}?$$4dpeCM*k-&CNhzRNK{@jj7YPxmHf>`|7@Xg|*gUV>Kf
zdiD(en-;;f5oH~}fRtAOHTABFj*+-8bq**1t)5>DOsBE^gJ&uLPKa8b+gg*4c(~**
z$x}R!s&1C+d`YyKfC*H#&_nBQ*JGPVi~gZS&*KLFjXLjr)9RDzjh9SMPx#;09?GjH
zOqgJCWNwQYNDPF~_H{YCy#-^(q@VY9PV}9s{rpjqz{&Na8-sL;)J!dK^iR3hY4%~8
zp8FkLcv9T_YvSWYS*~QQV}=R_FR^Uw_CxFMv_?ZGt+A<@=*Dd>*ZK|qgvfNrBc$)y
znFjPg)YHvJP^q*L5VjTXPnP-3{%+<5C!6agt%g34Ew2o+_`>~3ineS?%ZbPHO7!FA
zzz>;3hqby)q!b{a-I?z_Eq`x*QW!2Q?YT$G5z$A^t;i$_HOC4VDE+o}4|GxYm~mUQ
zV3ruE)MSQr1Gy%{b1dmZD0CTS5$1Ud9an$ioco}==m63a(f)WKFGwGe_sVs5#K{$r
z(D9an;9F@DqPtvMd~E98TYtf<GCyOau<J{IW2j2XZ`~9-^<pLqAMLWWJ%JKIpb<Dv
z6kT9@FI#7JwFd2~g0LVnCp%2BraXR?(i!>T%)(nheA~7!pYLF8(@IBpOLG^N3*F_F
z@Bxo*CVSoFVSZx+Cq}4PQZ=ROh9?WLqE29FuoU#3ZU_(=<XChAGpboA?pYzHXua7L
zOn#IPUn-ZX`HF4YPh*lvDn)vu+qVOSnKEP83@9pDj8GhKk?=)-guApxyJZ%+LLK!Z
zqc79V;ZoI2DhhR*4UDrmzahbYd-2!;H1K%&Ye0_4e@}TK+GR%HdrIAbnEzEJQVJF#
z?-*wEDkltE(gZ#mQW`NZR@>oS;+wKJZA8BU%5Yzwm)WF2pQ;@zfnmg!b3iaakR_0S
zMpy;RjN#u?dJ3uJzKqr-mBIV=1jHCz_)oa%oJD*8Q1N88b7g_w>Z+dHm^Z(L<?scV
zTg3Ph|E*(i`l{E!(SBF^8$bN!<8Z1mTR6`?thm=jJX+ny<$b=+;f}x$L4YxJ1Aw@7
zNzi{@{eBx4a@IZ%#W>O8VxK?4In7RW@=9&>P2%sR#&;*NoF_7a8zG6Yq19&U{t7>Z
z#h_MN*z&^R!)~w4K7M>ZFF^PB%OQ;o<Gay}pq9sIL2D(ou{TtJmK}u~bW&a<MZ7L1
zWwItMXj&7Y^04i#8f#qWfG>&4{(#i%wsc?=3a_ehvW?W+bmnH+2$obszYeUqTOVk#
z3o*2@z5XElXPGo*tagRb?9olRMdpFolRMdXoVmO2I}<gl^<Km&;Nd`%eSY@`Med%o
zYfRGdimj~d@vT$Z{-P`K^GcC+J1+w<wM4n?37-=mb`~y;6?p61(A%4i{yIvqPgXVf
zg(?k+eRKE*D0HZZ=f`x;bI(z{wwp)v2osUR#^1Dzu&Fiylx6kFySFW(MMicj^U7wi
zha#sYBt`MzN|L8pU#>m`3rL*zL-FJ(64m16qS?wipOaWs0z+PIW-5VGbV@n!lgTxg
z%u#nELaBu%vUel{7WHUG;|bA!f&9}zkz<JTr^}seZ8#HD!tsV_9Nodh?3&AMnU|MG
zpM98e3+7Eb7%dt<Jd)WUpbQ3!9ghSW8+G#zpLROY4+)i!l42PG#3nWXJu{GMxbPP4
zTv^v>hEqei1TX=mc`!*39cUUzV#>bf={-5H*Hiu4{cXwA@e4H@PP~4ss}aZ9&1{wa
zU{Kv<e&;mmsn4bg9D}UOX5JH<2gmqkl|0r`hN7l@-os)Fo_XBqYo)g?W94Uiw^!!F
z9f##I8E<ByfmwTj!^aZoJ<6!}k^r00Ik5KH!xn(YK7{#60$Of@?>8ca7f1VpK}WF5
zM~fCX{a-_XvenML6oN~o@&*o$*OX`D!nq;g+WP<Ii~l?5{h#_Cd&}}ciEao61BsuG
z{?GOr{Z2xg$)5U5wv*EIo6L+NwaGCj$MEg?<QStZTAsZ0Pzf?w^w2;zYpz}cl8)7|
z`Cz7mL7Fa=J&Dz)6z%p|_kMtKIe;!Q5|nx<_!nr*9(tUsJP))>GADvolz_Ah_FIL&
zK<D#Xw>v)(AN~UUuF1OkSyYsb8~787LL}bb2nBo%x<2H%yKvxo{J^Y6<1dhO9w2s&
ze!A(^cV*LGJR$-#ahi}cxX9Kuo(DneuSqxdjjfaONA4E&>O{_s7k_y66FfFqyuoBh
zG&2qyX2x<ti3b5DXikx9Ft`G^-`Z-$m!%t~f;^IC;m=_^j{*g71Md~~NDwvDu2s2<
zNwOUfKT?f)zwzwnseiVRF@SN1r$miG<g=cS6u*_<g9x2&WxczMrkhye*PWq-8&;Kl
zGukH5qR|mMjy%bl*X!Er-w=IodbF&fv!|Ge<<Sw3ht?RNB+IPrFm}-T+RN)SS3YfK
zS=FNE+&W{*h0aH_(Q+XL{WP2SHL~zl)xnhJbhI#MCyhOQ20^?jRsUuoQ}*gTo(%q+
zhm>v9s+~X8UG^n0J#ufhyJ;pC21iUK4y673vgIzfDP}qTtsdMa?R|ZW2%9XsIF4ds
zyM@<O;D1SJc-yq9O3%a?AlvzSigf<ZuoW-l?0iEn+5-*;Ag`(md(StBL2`CHMde|N
zInbSgoO)ZvZL9o@eO48w;G)drU!yQ)l1p*(7%i5<m7&G1e=6ACG8#OV64|(4W0CPs
zGymcZ{`oN7NvRj#4D9JB!wlz*rJIB&p`R9rCNw8wRK%E)=lwB|TRx}sJ*jDYrTIql
z#_Wxa5k`6zdEB7RFfzAuzNg1#Z2sz#M;Bx7nI6_n@ov_qn?1W_qmeWm;=Cb|W$J5%
zA)Zz=C9Z4Sof!?xIPV@)B%+;tug=lsm<o6L5iHuGf2+Nxx6C#LzbGf#Ew>~Xq0sO&
z_|d5;u%Id4=PcnqxNjN;=u;gUkkMs(5+eC|Mg~jra7061Er%tGwGsjDsOz4Rq-}vX
z^1PoeU^#h%&}Qs;eZu^nWTU7u>uNAyXoT~(%wi$PA_H>R#^P|3urTljJjt|FxNwF`
zXYlz2BoEBo6)fa0f2*^{U20kozt$>TLZ(wu`^%F1@ha84Gwy!KWPzqY02mR7oEGhY
zTR#6;%ll}#4}k-jC7&nXYI&V(@We9imvx^2&e$^ol;ygK$d&SJ`GK&cJ)EEAbBc<!
zu|WFoGcX3zYMANiEhR9pwvhqTr2<^;ZIugJtBv8;7^OVV$RQ1S(02_$%ya+Yi(-*N
zclUC_-M;Q&R&=lC5ze``nLaIJq6`7w*`oWgELt06e9$-hw3tN9=&cfqEvBhN`U1H)
z+07?xZUZccekd7?&SN{icyV_=qo%IDZqzb`7K@g~Q}{yvK()}6nF4l`^(+5=jLKr+
z5HnDo?*bltB}NObQ}P9ZHgAA#C(babprc(b@7~K7)4G9sAr}-i+88!CpX-I5r(AMz
zm6Zozeato66IAM39>wKjGe~meiZlo&!g7{&xO|P=oAEowgdtK~K07)J-M;@C?OR>*
z2E?obG+Zj(ZuOBL3gZ@#UaBhij32=R!&u%Jp=e^qAs3YUK{$aT98oQ_>j7r^lXNq`
zyT`wL&#RS8FsPG^<#!o9UJ_$D9siwx-7QL@0h5{mI(h`B;xQ_DK-ocr7w3b)^wdoG
ze)yhw+OdrSrQ^-RVKwc(0kd!kapCHA&exZAwtAlV8PXTbJ#zbEzm)q^i{RH1>kr!7
zN-Um(1A|u`<*rDyQ`!4Rc4==@U%48X*`Iw=t4QzJyXMlD3OkFF9g}ez#qOBwWg}N5
z*!kzU8Yeaei7(yFK7xJHiMK^`o%{Bqy+ut+CS`r{EIwbyTHQ^s3hi!sOkXf0_KD(O
zAgRRrYs%F15gTHBq|wV6n~=QFuB9Wo#})X8v>i7yBd*NUZ^~-<BTPA}CRY>Ij{+c8
zk&pWf=ZZVf8?%Yi%n;?LESB}%HxTGDAlZs6H$S27%b&d`+YjWMa?fQPjWOY<54l%0
z0O{LH1<Qln(%okB{aHJ{ll((~TF^$gBT+hgFt|}vdcbFusvW=llM>f{S{LoH<XMT>
z5g`buH@jcJenU?+?97s~P`ZAiVfTW)-~4H&Ti{-K6sc<Qhg&rv9foZ5B8QggoaZmt
zHe<hlR&Cet0Cp5ua)B&F{yr+ijmK~*^=zygjI2*xi|(U;6;a`x(ZznuV{`!NHuuc^
z1&YH0utWf`6ZPr?OfyE?L*fq*Ji6{*U$`3_=_a;WTr(}c)oEYO>&}+5Y`R5djW*o|
zAtWcA+>gUWi38xgP0J2|^Xq47CM%`KNn!|gdZvbKVxZ>)J^ul{-F*2)X{mu`H}civ
z2M?Wab}53c2Fmo;%cICUn9!#@wrXzG7s6F?upLr8@tR|=T$Nl#JTAA)YVN{ka!$;Q
zBKwwAH<co6RA$*&RZ2!;WzaZwWtf10UqZ8Wvm5G*pxZql|FkU?sj<O+dKKb+=l|-9
zD#v0%Z+ZwTS!vB*AouwAOxU@56q91W<t?9q$?gi4koR%vW_^Nkin}2euCnWo6*(MM
zz$~`GDe7LDnaH0QQkN3+rqNz+d6{+L#qgU|V7&%uF=S%X4jjRH+hG|Z=g*Of4=d|Z
zHV*3J7oUMmdRcXyM*||1v|f*?fzFtb4uD8FXdGP$we=_Db&ok>nfqNo1Z$#2p0Ij>
z{Q%xX&2r%3sSLWJPp<HbvLZ^p&Na(1tB(2NlZT=?St`hf^Olas*c0ePJ<-op)=_@*
zofBwL{tO%=MxDh1Pi?2aD>4k|EUa`g&;Oh&JsiH`w8zkNVM6G=Y4?XK?XECC-_cz&
zt5XbLy6W<?=Q-T`%quKb#_?EitA9(aIkwJ3c|(Tj1#lC!%7IFxHYZRg=v4R8E5u)!
z_+tw=A2xHM>cn`#YQ{dmPAn%NaYiiN|9YZbfcwH})ic8@5B2C|BqiU8w{}r4cQ`_K
zen*`=RFZ+T7vv=bG;{y<IwGMl4_0oa(8jR93BgdL0msQjo@LFGua{&IT#T?8?pB(u
zIOwQWXa>1|07&Q6&BRt`NZTh!t=W^nMY@2+o1nMlYxVI5hnaig1<#YCvj*pg3p_=)
zWEfft54B3!r+8h~G8j4-s7a~yxidWzG>Ie_=x%P+TVr)5dHJ&mDwgCDS^ORGe&CEs
zuFh^vm$uBNIxoEHWt$}zixJD!89vZ_5G;|P`P8fsLSCOT*)=a++eNb-F-o&UtoxcS
z6!0sVE~mT8f$(db%-(WMbgPHnoLxxeEhcV#<)`To=TW@@N&b$tE>CQS@8j1^`3n}|
z`w&L<@(_wKKaPv*R6S6^h|RgkCRh?g6i!oX*TLnxUl~-B5lbE{ybTc?T}-!S6f}Dc
z##Xf6=?1JA-1q5AHSd>)v_jRtG`}P5rGTY)N`Wu~a~Jp)?2;~Vuy<w1{)IZ#f29WY
z-{=0<YK;GVfB$`d|NRX7`x*FGJ_BfD-Zsj0EmJ~+pFeeTe5SQb07h!#uV<DHf4njr
zap72s#%FR#_blP{rzeDIO{Ko|yp>v!2f!qBKnGbnGB#J!5JR{3y#21jHE66Bq`u}M
zawFc{no8*0r)N(OLZ94`ycaHp7a1-+HS+sp{d$T!)FHFjzMzJLQKhe9;Ur`Qf3iSw
zR<g|49*4I~sPF?{PO{2bDlOx=3_XO+hAtwK@)NqN0NVaU>At7umD2d)==im3ghm;)
z%f)svGWA|gg5%CVc2{-%D9!9HkR9Yyjj5vra#cOlExn|d{QS(>M4l}SsoN<WzaDy}
z|0?hk;7w%k8An>zr}tHAHv(%;8H1bP!?0^1LFwG@;%pp$rh^;zt+~ou`C!$<UdLJK
zRr79@B<M1EU0?&{_&EB*H1VVw0k%jYfhe4@eDD~VT6{|+TToNpSzuVJg*X4btPU_|
z>~){y(0(Qam}{v*FUV!+HKbr%JyuaF-CW_pY<_s*!rUBLgX#@*#v69dnlNIM`dj@E
z7jyUmxe5z)>5MI#fzn*?Af+lgsbkC*?IAoTWko>13Jm<07ca8cCY7D#!D|lqI;_^K
zQKA5qTVNDWaP+9zKQzC?!i&L0Ih84YZ6L2jF}ZDQ+q%8AzEt-l<%`X{ZvK!_PS!Sj
z)0JpLhE9YkWqhx`y4QJ3=#zkwF=Jc0t`>2#NNpMSQ!aMn2fhJ&jzIAMh5(72Z-7T4
zw|nJ`el}?TThPH2+PJ=CS<7tNNI_Df)Dx8mRYoFwB5&r@eG^KzKsMIBoB4@BnmyI?
zl5Y$C`Nvt7yD}sWX>7iVQK?PF=dw~@m!Zuk7pzUv5}u6}^%)(6Yz~{mF2-$zw%<$e
zALT!3sfs#Ec^oJwbL73<^j-XG<_sr%;7L_z1Aw~Vdh%ZD(h@&J%;V9e8l?QtuLEmE
zs;&KsOnM&-n^(*~i>7})ccbe>7*yaumy+jW$aZE4Jt)6sPIiBZHt2Ap;LDh3w(F)&
zP)TB_Zf(X!K#H8{`#L+Drof@Oz)+3J)3f?Devs+%OOf~~A+oZ5Y>vfw@`uDzMhg9B
z-zepZx<Y~<qh~@WYe^jqk07&AABNvaws(~x^=^#r9<nt^<|{omnq}PMC)p$J`%;Ly
zs*WzizI&CKcTbs{H&>AHnzF`L5}S9=H&KRauXg&b^gCu_LnNy&wmG3;3I3-VdiCPX
z7LjcqdN*9nGvdiwkfXeiTOSiGnSwN}rs9(&qgX+~9*iLWV`l!@iN-O)p*KyJ7j$kB
z)pk9>1glJFeStM20)5;|L3*Ox+lmxnr1eW|zY`=kvPOd?)L{#-uZp!jI`M-z9qWZQ
zcmh2jaW+JmnMw7#XYG^aJeDf7g`^d-!Lo>5Aj~?ib$U6gfICS&^^u%p(IILF-D2f=
zbwgY~=9GX6G~AfnFq6nLQ6s&X3h16X<BD%YW;zki(L2Y<RH(J&6?bzUxGjJ4rg`#n
zlStJ``aH|@F?ygO85=*Ci)C-v8I#Acip9$OhmtHdOJvL*W@B9nEo3wD<vnXfi=Y=G
zfNEVb@qA7&_YO<zFObnmHsM^^!{jA!im1Doc?qsrc|K2dFRm<4WsX}oL5CwGY^LE3
z&2;Q}Wph%xgTvPOxQG|=u;qCvcL)NQ-S4C}D6u%FzeDWaJ$~>P=oc^G9UM4~2t#CX
zGik0(&N4dWH&>n}PaH`qRy=g;*Dc7pW{kb_0iq+O>Q>GqmVg(!=pH76S>1S={J+|}
z&aS4mt{V(RctDXNMVbT<l7N5^r8fy8B1JlhbTEV}9Yl(PbOkP`NC-$Tp@t5k1`ttt
z3<`+egpPo;ARxK#@p;EP#`^=F`+T{6I_JX~XY8}j+-vW#)|zw9c_f=P$xRoi6Z|K2
ze1Qehym%B@x9zYb%anMF;*m5EZzFWv&ix_Ea$r?hAT!6DGuJH^#<9ULf{7pO`LEh`
z0T~JjlP?EzWTt=-!PFsTiBzFkJC--K-WsWgCLzdz1wm3;r58A0mp?fS@DGy!h-6lb
z9GT?j4hS6cWB{RetG~wKdVr!LTN$e?I8@Hk{UmR7(bC{;F6H%=e_h#|84|e~Dd6K-
zA6%mDlr%p<^*n@+MGV|u#0={c9YN)lDOy;o7u`Khh)W)deL^}=TAg?f$c*}|VquJ$
zJ023UJ+YA9^s~As%u0V)*Ie0H3zoTgGlvC7yz%K=Rvok^vz%l0fmnBp`cg<#l*u!#
zj(aSvHk@JFV{Zr`?Jt&%ZQ1V<*V$s$tK&^=Hu;yDCcKn}Gf`_tAEn&$>7uLnNxYtV
zFy1*Te7%^298D8hB!>`0BW3F}RTRq7Z+gU`QM=4v^FR~h=fAc$H9HY|_qcd+&A+w;
zj_|v#(5EKEz1vbmc4lAl+8@X#;9Bt55YZ0!tbXSuk^G9Gpgco*Hq0!nrtw`EOa#(D
zT+tm%<71k3oG}b|_BB(=%*xql*>dtsIoj?G6&qNh#*Ko(8`hvqphvnoi*<xhD-cNq
ztK93Z2jjdCS2n7Za}P=X*T7%}tBKM}{siEUKyn@)MYZl$&q9C%i}c3PZEsMASCYPr
zGEx$v@s6u)cgR4y$jh%{0`)qDb~}r<AXYrpP#ZI6=yb4awrqJTPr7KrIcuH2Ezj;l
zs_`)EC5z>_K9FCZ<kdK8NQf3kJ~_-*ajGyX*)2I59a|f7HT4kIn~nM26rFpkYfe9?
zeo>ruV2MGDVbyU*^DD_&Lf4d#Bxqp>ZL9D^B4vGc&BW16q%h>;OKScEva!VTGLH_l
zH1^yQEPqnob!=T;^$Z`ktA$8KK1^<xFK<U2yNP%P)_`i7c4GRGCyZ$0u@`SxHb(sX
zqa&`A@ij276}6Y7f|VXy%sLbPXBPi9KMf_OgY#aR&^)j%X%y+~Ey3SN0X0L}A;D6y
zFXN3Iii;dxVVot2`6H*P2J2iu0{d=0KDu$l6%@(!iugSWCHoH9s3^9t!qgsA;o#8b
z`0UAZw@>-t8zguTQcS?sV8BC-wO&)Fc<QZ-=;t}(z@Pxl{s`oHxBB4?l^==OR5vds
z7&b&G)9u_@JFoJ2U11wU==YV$CQ`t-sf^~p-o1J{ZQpt#1jLiAVwZa4HLPt`SL(Vf
zzNNexLm%0<HdXi)B5e<^Y)ht?U-nwh3VJKLIkKoLDoR*u{^a5t$X6P1R)hyy+w9>V
z?#neq(wiR?Qv5Sa{5lqbYI<NwR+)Y$NMXgs@m6uNvVFX(gLBxlAE5)BlX7Aq*ISts
z8N;m0$X}7^G!U7pUzB@@%wBtJUen$`PLrTqOr>GG#@&pRnr2uy)iZF=wb^^+TK8VJ
zj6r<&-N6fN4I?*(K-dqC>H6cddZK;rmD(4vmslH!ruvt^$4sF%vIJ+2MK`6GoxIi3
z_)+@9H<Q3XV1h7et5dNT8Dl4(R4%;f?w?1~;97<zs1h4VQXf6@@(?{Vc}nHn*-L8R
z=^h#y{lCea|4Zup4OjC?AI9fWl&<zZ`t3JTrkxlsrrCtv1F(j2xog;W-<(^W+^+2a
zXzHr;MDbI4h^mBd%y!o*Jx&(fhL$J!%b$ny9wKI<ank6CBswSE0^Purc$_(E6l4(#
z5+gGWhqstDy(>kl<7Ljo0{*$ma+;>l5!Q$BJH?-ogVA~b6PS-F4sc_jm;kcjoQ+dT
z@`4O$ty_RzcHR_KNi@Yoo|Y!%c>L9#mLlx#BWU~;bM|!>P!NN4gsvHvVR!xOQFv`)
ziYz#5Y*#i|PN7Cse$&+^OAV1FWHuUQITSwm0y<uL<+FS}6%h#Z%gDWbQySt^^(7p<
z%SU9211fEhjHiGhj?=#I=w#0vdo*zGxNm(R3qEI*_V&+b?oM5lLnA6o=XE5GXoENi
z9=6;RdelFtH*8#+_rx{5<b<0(t(g15{wV9|hyU*!RzbwxZ`stY^hMY>KCnr1+W-mT
z$2%Mpl~1SvLYYbxPlNH?FUmH%T7vo^nUCE2!{-hIVCBza)w&IOLt7EI^~!2mZF%Jk
zt@V`(l!JE5jXU+f#~BMotO{9Vya$9w;EbG6C^WwcC`hb;Mn>1PNhQ;|NI?a=7-ED@
zWsSeUDl5ByS^9fzXIA?LTvfEzY9;V;U!7ic8T_qW`kJiay4(E2%Dw8T*|wI*%eJAb
z-Q`qyQW^3#7SJT1?ufcwZAhPQ%d<c4xnR%fRaZ!_#hZ}VnlBZYAU&-#t|Qf|iRQ<Q
zap|oaZYfmanIVOKc3!Jg;l%Bp_eq^fBZ6NieXm#>Ggl030N5d(!)RKEE`0g6Liq})
zpCt^^AQZxn7>0rR;!%jxQC{m(;fbiezd%7i+hScW1IQZlI?ljzBh#M?NfCV-Pt=qX
z{4I0^!M&7L#kB@wkIuk^RF?tm#WJ}UY|I!QztzQ-yIr=r{CZNd%)1hnqf98Lq%22Y
zOYE&nrsk4!!JLBNgBW$dBZgDwlw(AszVWf2lxtZM;@<!#m7bi7uLFq<Xm91($<9CO
z*%~uIjF4`C;04ruHDjjnj3M*4w{!2Rx&^k|P+2lAq{a6=_+e`bJfMs+@}L=E(+69*
zYXb8F93d&o6@7dF<P&B|_Vwg6ANrE88b?|VThFS+YUGSShbh@1bU!(5BRauapj|3u
zd0^mG2?v`l0O!XsTaxFGBFlHmmOFM305{8_FdXfMPUCCrpv51j6U<J%-d<`Ad|ism
zW-_|#9jz?i;ZYJKF;oBi5$Tcy2WVhSbIrsSZ<8gHaT0FmCtfK!vi)<VD-1&Fz)hix
zqZ8-fS3=Vgv|iO##WM*)Bx5h(c~+_`r}g1uL(i|4E;w5$ePfN>JAxVX6&Wtzq-XqP
z3Nr4D7EAD^dHp<?yaob;oXvG1%*SNG8I-yj1!8Iv@tMaRg`0+{&V~c)x(m$;E)hQ^
zA;~6c&DUgC?pipeRP}`D9&k8-NZ%Qim<ShtjL3sE_T$L3$`LTzRMXD8HX^TdVnqVC
z+AYsdZR+^)$Q2t|^19szM>3~ypgHna3}Hn+tLm$mZM1j1;0T?Y;h(y&#3VxL1A!C)
z)?QGr0$sQ-vBb@*RoCrN#qbOuo4+C-<QRW=r9JifmkWH-(=HsTU!)k;WbWgy<}6vj
zg~wvmrt$lyY#H4*4y+<?_HgCmeOb#I4WRv<C&n+Q0x}}g`~<3Of$!-RSu0}N6Aq0n
zs4oJJ$;}+Rce^)b(8z#j&Z?c%xS~#+s_q@ro8LEQ6R95t3OCx#d;$>C6mBLgg*r?%
zfLm709gDn&He^1gs0*2mRl^=OZ?b%7E{*kU&d`I8=ST#8^y+r#d9|7Nbah^5OSP-2
zde7=)VQC?UFwlAvy8$={rvjW!vy*^FI?o$M`}r*bhb$A&VN?}UR2xOgrEa|ToXL-f
zo78?PQ-{sFPfwnownw2flR}9mk4|=<Ppp}57%Pmw!>$V_jb>e!1;egRN=(Sxys<gS
z{GBVm=<EYmCCg=b8A5GK8N{8KqNBh=W7#G%$~w*&-to8giA<ZQ0OQAc<~%8k24Emg
zzXsxu{tHxF3Ap^(0H)6);hpQ4o?SYg^0*JDQg`t4>Jso%VtRYFP625Lz^v@*?BgTH
zXfE{>(8AC2ehPQw=ERs^xGqjek%c{*k23w>s^O&>c6Y5jP!B{K&^Hz_A9Z~KOoQH4
zd)ven1Bs4B=j&1}=s}~MLXLWrX&CiTZ9btWxyWD6X~Xx`T*jx`#`@susLpQtU6ZA|
z!_0S4O1!yXGG+H~-!cDPUilk#;(ywE`Ek|Q(^JkCh`k)foJCp#L3mN+BPE7^<@O+0
zlOtIy>X;g^uy&Z#f8>%4&$cB}XD_2?QI&}<Ypb^Y%~yPgKQ^T8In7#`bk7y!81-`-
zArdqOKuRv7du$Rh4&_v{eLi39%or$$tUll5tv|?;g$eWZG#o72Q&}@}XsLhd41@16
zdkrRpDT*<2LMgK_yVHW)lCn!WIF}#x%#)D>a{hZi2hD;aBfW!Bmd_Vat=(;h8b+$k
z!^)Dl(CnRR%P)OmrTp1!+wS@+y88ha82p(uzEcJs!TDyQuh;w*r+4^Py8sr8n_B_i
z%l{}Z7E$MCIX>=~4CYHW;qLZkCA*+yopQ3@DkUpDe8&2GQ~YOaUC(M}&Gf^6N2zR}
z&nmgg#Jfl{Nx_fZ{JrG0Zv(^ZyVVCa<^Gi~JzCsekstws(Z?%(YB7f%s6Ov4WKBoW
z+96UDpFl!uy3kn^oIdLoH=)tNz!N~h!T%u8^53gkZp~aVupEENIUj1m`&~+$#^CR^
zS1-Va$xnU8>?O)shd!BGxbP#ikQ05!J1p|rmNz?pzEbRK6@W@P0ODO9a-zT&yS{%Y
zIX$+KeJV?t4HQJl+`$=H<f>8>Bo3hn8XqWqaVBpm@L_o=h>g;|EPtNZtsg}|u(Y+d
zX_VY=87Hj3fF3fn-e$6D>U2a1ED(eBn|1B$3=Cb%YpApPe5Mm^aGhiF^w-L?zd$_r
ztzT!ag=X=!=$rlF+P(16+K5LQSE6K>*l3n3+}%C*goPgaWKLJr4+kJ&g;jaY#ZH)s
zX$Uh$fa#&$5Xc47=uJ`gZFHQL|HdWUsQ5@u7a0)ky`EY=Lm(d{me{Uka}^0>3Ln@I
z{?PGiis#W&>aWcF0_Io!9W(zyOXHum&;0)0|BfBj?|J?`&%ZJ78w0;F@EZgFX$-{t
e^=ULtgYB0Bzy*@)Ma1Ow41QJln?wcbm;V7bn<+#9

literal 145701
zcmeFZ2UJvFw<mZBl7r+-As`@8a!w*iL_h>2D>(=Vh!jvz1c{On5JWPPDan#^R3zt|
zBtsRkfTEcCf8XnV-P66+yqQ^V_4GIY%Q~#8drzHv_qqG*{oCo>&fKm5cb}+fssUJ7
z0Dy)018(O56#x(S4&EJHJiI%2`1p7PM0bgZ2nmU($;e6W(oxgX(^1pXGP2wUF*0#5
z)6%l>vvF{7Kj3}901|j8!1M4v&jX&n2EoF|$0s5nq9P)q;$fs^<oUmS+_nQ01b3!!
z@o})&0c;8^915)4ZUBUN&O2EDIDmgUu&{A3&x=n$NJNb3P<t1^#=^nD#>KgF2NxI9
z+aGfsz@@lDc~4jwk4o<~KD#@$NKir!0mq}t4w~nqa86Mxk6=O~S~_|L#`|2{JP&xq
z#3dx9q-9hdtE#DMXg<+@VPI(V(%8h>#@5c>!O_X{o!5JBu#a!ZhtRO_h{&kKq)*8y
zsh`u*bMx{Gz7-Y~msC~P)YjGiXlU&0>h9_7>mL{#8=sh*nx2`RTU}e<*!;7#y|arr
zJ~=%*zd&AI{Y4iRfb&nXF#rEa*#C(x3JhJ?xVSjD_<zxbh3$=LI25>d?g`^jD(m6D
zcBf(&2_m3=l#o-|LC7Kc98P29F-k<sDYkka@fT_TpzJ?JSn&TN%Kj^1|4!E&K!SsX
zd3iV#fFf}9Es7@_`|tMm9Q@4#fAhfKJn%OU{LKS@^T6Lc@V|}+@*LOn!;<kQLv1XV
z<vx_DQ(Rq51?UuKlYKdO2px84JE~NldvyzBjzX8M8Zxp>MEo6azFj{Boy-Yc@9Uj^
zR@|zwyqKx5=bnnrG68m4$MI3;AJNy;{@=(i59n^r`O*f?^0=o0W!!2yZUOn+TcEWE
z<QRhXQ@sWJqyBjKI}Dutue#YYW9tCQ$JYDp(?U_Y8gk6K_y(#rFD>g3%v+oLlQ5+_
z)O!DV?ZqVL@d5d@I*n^}?J)98=i(N~$tH2)$^9JsA6SO~<MzId$edbV1z$$_4If(0
z3I$EeCVr=<dN1@}ZMN*4p@{9Qqqq%cwd&e7<f+a@-?120j4HMMT|HH5;D2bE=_2wD
zSPuF9X9fShp6gHBkcf{6wB2_;sMakIN8bke<AA+Lt*7x?m0B0;U#S_-Eu(z7uh%ab
zN6QY}(DGC_2viJUjhUxzYCTne8qir}5HoOYc`hJes5wzp*@u_(p3&&N071TDmv-_~
zy_!UT!Z$rc2XUuhKRV*ccLs)|7bv4LRX10cyMG+RmcKcAG9FXFBE`ZY`h}=oNpa&^
zpBBY^6k`+dq7S?vMbm7h_FF(`=BZT<Zvk-TEl{90AvQzYJ1%(BPjE0-m{F^?RTM|F
z?RFtf4@Dfo3&|<BZ-J8=(8ltrg_4Whr$ww;qtj7^^F`K!QgvyI_S4chdIyQB&s!6@
zYRoHPw*XNbnhHvK3yheMUvIW`HmmgNa~{Wx7JYu}zSJZXHif30wld|r5s|9pWKOY?
zjB|DLOyYgOa)jbSGu;B&HK2>1&`_6I^A3L(mY)2N{FJ2q>pv>g&EZ+$VmpcAqlV^-
zPpUnyruFmYdtAQFtGbaV5M>jB%=5-dL>iJp>x4!%wG8e4noOFuf?J?3+A{7$JjnFQ
z5(Vx--U6aZwA;B+DnvZFf&WtN_tY)0KB~A3>Ys}FuZz+FhcSxh8(VG^4^UE{?(!|)
zT5Hnog*3fX>5CxqAmNHNd4xWD4uaKZ-IQL;r^^J-SRz<Mp6_dA+mFVE^tB7`&jc?o
zSZDUk?YqGykL^`Xx_|&^T!MvriBpHzSw`HF%zGB{Cmq#Q!ZHzDA9lv%yX?$ot|qrv
zL-fbvzsZd}6tyttf{9x3Eb}N>`S0bA9vzmu#B||oyw7*`W1(hrD#@yCl?VPras`=B
zv8%1tYOJT{-M@Y&$@wMZ-H`2WstzvCE#MV&HTJ`7v-U%Bt#SFrWoOo>q&>stdnM*C
zm*u{wy<)>-F)Y&dy9KCv-b>5T(hqqn5JYuaQmJ8csC$I}T6jo(Uc&XQH@0By@`z38
zL@H^5%hTfHs^ik0a#gy1szV1iz4Vol{8~=o&eM2xXIBO~&dXQRTB}VTnnvGj#Sz4@
zs{cH6+G%;?@j`4nZ%6fw)!)qZm&E^Xn5#^O%E>6*=ApJN<BtVMdzYNyqMEI;obHd|
zbPXDjnIhM>-jU_bQ)AjJ*@@EZGZlG&=PfV=<3ivHU@sg0nzxv-M(o8+wdRCP@xANj
zEolujih`KKkLPu~z|m0`F&VF}tn0_B++|)2i@;Ywu0|;miT9uX!mqytDw<%Vdor%D
zZ{|C@RfJlp=8IwALX=2DO>Ns3or(P!O<tEJr4IbD7H4I+-Iqd!(1J!z&%2)WEM$%^
z?{jojl)Jv!6joc*!Om`;<`+rve%(CHoTXy<*b;22)6f*W^0-u>$rYEZ&7^&@nI`{R
z+1xL9cR6TyyoHZA<8lhFmOeqqM8#MYGc;Lk?j7>3DQ)gZuim=xf$p8(Z<{wJOb*hW
zWJ1vej*i#yJ=VIyC+~*0hP+z|zNbEhcb}L)eq3GjlcHnT3C|9kiGbA_#t;f{gj#;{
zf$5D2gr+3Bya~Y;BMv%+iNQm&=20(V*cQHH|7m%J2r|-ivNg>nGT|wYe)Jb6`CEtn
z7afXI{vsK=W^0YUhCZo(>5FEc5c^pztcTNZjOb3oi9XLC;V6R~aSG_QjthtbF+Oh0
zzhZjfQmE22HwH;~d&o_eY{reSZd*+B1vyQ#dwxkfbF1HrnMn>Noj5Jah(zNQ-Azxb
zyAP!|{LWz}x|JotzP?6b;B}oJ%@g})QWca82bxkkYxoQUD83OC-5PhpINi0z>o13j
zMP~{b&nvaR`bcxL?BSTS*4riyZzhBfoEXd>ICb{oCI2DAmwKk{v4eXx>vtV@s=7z%
z#l0|5ng42sQA+<_%qKO%VH})HC#{Q#0+BhNDlx1Vaq+Ln2kT7c_u=_EBD}>*W+bj}
z;Cw7zz2nB=Gp!T*3n=6s^2H}llXtK@i>C-DSx&c4!E`9Ti5ltjttYi<^L{cm_2nKz
z<xjg);A2Tgx4>>iwPsr3_)7B1wXp4Q^3AYQqhFfAws~$8pZ-5IN%@V?KOt=y(efBZ
z>GTTqU8V!|s?^Wv{|C7DKZAh(8=W(JkC8B}k!U1Zt;Si_6a!5k-vSUQhZ22mO8oQj
ze?VNLdkD}kb6&2X09z|(ScP+MSt$+d`e}Qd1rJW2muWVPbS>WGDPCW>a_JM=W2SkJ
z<z`(FYs&+OHt|bMezr~equ46a=`V%Bg>5=28dV2}1^!|mncwuiowFP<`VJ-{msQr_
zmd#fsMxTN6t3HKq{$3sOP3`nI7Q}sHn%G2m+FtOw_r32|X5rbPvOX&gVlGcMzF2ua
zBeP~xgG7~Xjb^gR0p&G97@9L*;@oLuZq?+Kg5A;ZmU(J~n9rH4X|nfAYcE0J=^Z_5
zs6*Pi1<$mr>rtZo0b}f=*j(6-NqelRqs{wxJghG?-A*S1j`9OPT*gv`i%)T)v`F>e
z;2`Hm{S1y58BTWnHqx8xRZT|IU#3O&?n7mks=X5zS(Q(cHyGiT<j65CpJdljJ*rrl
zTEE!v_o-L%1!M109(I1a>z>^A+)1xWD9S{6ETu-e^KXHMF$%0Ov!V4=`7i6|<rO+A
z{YU+>Y}ZER>l%CMv+^hxz+Gm#>ss58t;#~10bkkxzt*Y?9^A=uPWpU1dh5nyy6E|c
z?81pUQriv#;HjIKp~K>TU|{HKw+wyO{#^NhvF9ZI&-5ctO(o`|Fganj?B${Nsg!2g
zY#oBaalShnX_*WhD)n!BeJ5T#|2|&1#W>yhh4lI9qC5)mWTC+T-%B08)09BpVWJ||
z-`eD~@j=9ism_PScXSq>%fk=QEZYxjw;63eSAGwhG)}WJ&(3U1Hpva7epiqih5I*x
z`X_>-em+KIV)Y*~ME+gYw*RX;6?CV<-k_t&PfAY*)~O!%FA`*vnK7Dpb!+1mm{xux
zvZDPDIAO3GNyS{MBK)CgWu_X8KEbfXBXdR`1EM^Dga|O<$&LSqkok`c^uJLb0#|{&
zuW$E;%k^CDc^(g{PkNes(#116SHF?OSi*`|r%)T{I245e-!PM!|D7*vpC*F>o-f0h
z4Zl)jF+a@btEDDN#8|n1S-O9>zvtlp+dL3_emvJpC4SH?fgmIPwMus)h9yrE`Fd@c
z%!!U`sjHh&b+TO+QY2*ke)t=XUETKoHA|OjLLFJ!gH21{WyIsUu#sr)^jSC!0n~bW
zW(&&Bpfpdhw|yC|3Bl1f+JKg-2x0DzNZa~^`!S6&4`2CtQC|T(`R`w-*}p2u<4g*?
zqvGu6?A^AX`f#|KskxrCelPYb54P`L9y4C1(k5qB!=@i$cU-X&Izuie5;PKfs~+!e
zckvE-5S)tvyMJ}#8$`Jwf`_{-@17o!bXi@6oerJG|5$vZF^CvhRkp*Y$7Mb*fzr0}
z$1trPWJSu6dRiB8HYU!BCHDUMe7s==3}$Q^on@21%57+Jqu=h*A3Jp(dPH#cbMIZI
z>{FKCL`ey~!pHRilRbNFlxMPTu_pFGr}L%{-fTsZc31JG=|$DV1T)bex39z&N*^e(
zR;to;LaY&seO*Vyq|fJ{B<Xs`KO#G*e59jtZ!cuzmDf54d!TG1Uk1GJ<LNRLSO!Gl
z_-n{%q$W}Ap@oEAiprs)1-Xia<t;!ezOsCU2M$4_R4@|N0cT+wMsf>;eSy2(;K*T4
z1>D0}L^aSZZKNbRC=>Ig?d~mrm;)|k_=GyI7V9JB<X@M~<d4oRb#uko0aExA*lJKJ
zlTR2=xsT2P@wz1in5nBVAZK;5=J?Oo>`cvE!Tu+kL?=NPrqcNJ&sJE=RbL}P<~meh
z6^$utz`tsd(-vRWlKwUFAqTvC5a7xEG0Y0lgdYXDy09TgPC73%hUrp?HRH+1b&9gd
zen%baARTmoTVVL$2#wW-)WeWOxB|MoX}Ja3F-n36IBVn^P+ZNC!?^_%>)Vicx4^2&
z5*4y<m7K=;yip>->)pk)EUGqfHsRTXpoh*g>Jj|h(Z6gu<JS9M^y=W~9EC_~(Oq|?
zBRA^y<b?Kq{a|u=nG2ZkO_l<WB^XU=v20E|P;8=6)%*RPs>4g`9PT`v-_QtM{O|uT
zAI*llq4XwTx=wGC#Zw)b?nmlG%nUYejSU@rQ+Y`L^&#@Ube;KkvKft7Tl1RT->GK9
z&lvCmWlDHp<j6LJjwKw_ZO+(nu&*9=_}F7WI3~L6%Xz<O69Z1>*B-zxlsckgL?r%2
zbg=B$2$t$w5G9(|A&Xpa-{q{-T+O>#4Lp?Z(0+%F`OxdELI}<1+|hPQf*`*^hTj6Q
zBlr!Y7LM?}b?)$Hf0#l--t6vovoca?=c2q9bNSdJ6@7{i4_?-TsB2xg*VB<Y-G=Y$
ze8mR|_WwLhNyuU~3yrf<;R06U^rJ`ct41wk5qoO^IV(*qRGRQkW%mk;C!7wpIuCu!
zN{WGy6zg8aM(3$_<L};G+f5kWZxMd*)#`UU@-ld<&QpojAvX%(DROZ=sF+M2TVt4f
zMi|)zK#9^Zi%-&Mx+aFfpf<*^vt3ff>urKP`ii|^?95!&sQ%=%^jd2hx>Zo~lNarz
zkxg9lSJdG@xxIxl1h93$PdZ~D6x<UP^gZ@CtPF^stk*9*pjf@0ma<^Q)#2y>kp&P1
zILk0`{#nBBfp2$adwuTq#bclURvLzh=UwY)OGX3;Zd<SzzwfS8A@ZI((Hub41sI$?
zOS5+S=6F6Op~r7nH(I8rmWO4a24s_GoDI+VQWU0)pj7U)81!ouDaESnBuiT?td>l-
zvp_Li^wG7MY0X>#eKk0L@9@5~gpG$H7;NL(#d*PuW3gHcZtF5J-%s()ig@QI`vcKR
z{Bb7Fs~^|aIEVoklv7lWa$l8ATMu6DVWb{^(m^S8Mv(S3vK_!1;5^SJi;7xNkawD%
z19kd)>y5b<)U<^6AODH0tNH_;Sh)^XU$?+og+*B?BFMw|QR1trt!>k@`H4rmuUJNj
zAthyfP9)ZUq&5(-HCSh7G^>iF<#PJ)X*GmS1e$AC#H^hyGS^s0yXz1me0Y8<@f$03
z99j6hE=4^@4i{T!=U5Kp@7&d_A*f(nB51i=MS3hjz#Xy|6wiaRm(I7L{oFF@7QpAO
zoA4sluGPMu!EyM*?~8{YhZ|Os)JmX2OX*G<7Wmt6Zvai@)G4aWi1?(}w~tNAM)483
zd@s2#u+0Pes<<kEz(<J-oztDyE}u(g#Z|f8EGhhKqQ3?F(ys(DeE0{35gXkP1I9-k
zbe%3D^S=CiArY*z7(b%cZ9S5xvu+MMP=-0RvcO40m%^d)F+X?RyT=~QYpPJ^!^U|P
zd4Y2xz72EYZV+`VuVdYneb`}P_4mq9tHxGQ`sk{4&gM37(#_;H!LZq+`EzJg7Cx-p
zi&R}N^WtejXWjL;hWhw39&e8iLDM)h8H~U+Wz<#x2wCy7x0SkT>J+ZWZKg@(VHN+j
zUkAc^(@P3&>y{T<Gh1f3-!PJ8$;wD4QJ~tg&d-GFewoGgRycl^e&bJ|LKFsB;A|{M
zwH<qz(A_{C+pd_s&XOTt5Ju={rH?SU`T!$DRbpGKvVrNCfXU_KCvi8>E>wEUr~mG2
zpoF5`@Fn{a^Oe~xu=BKbr}mF!wLL1qjm%#Vdiu@ebPni4q|L9~yo=54#K86Vxz>9~
z(e(y7oOK7G0&f}f-g~Xlo}UM<XAWPST*jWAAH^tcJezLY>xACWP9o|%+b|-&(|cu@
zN!=lP_Z(#VqgsAp&3_al<O4wwCct&TlEWcetOJWpm?~Lp;k{4M!Q|f_{2q)l?l;g&
z)NIFY4Y)~ehD=|XEs;-duwOsBQ%;gW_kq=Vs}VGE3sf7mIq5{GbS*KlvW(l|=ueDX
z_UH&JW3DWl#bvuW0r0k1w7eS&cuc+Sy^Lzuo}c(IiRg*uEr87yfaSc`_95%|soaBG
z;8nlkDV8NK>Jc(PjtU;9)HK14V=@7Q%pNN^B>lW^jl$<%$Wpd6fN<UdiLY06uJM#H
zF+w6>&-nwgZ~fXWe2Lm`3fo+5UD)`Muh{47zMJG-e=y5lQ&}91W4d5r-eBwfy4oM(
z<^2uN8bLh?$b1_~fH4ax06Bs;bqPCODb%jTF)Vh<=~wJrc<jK<ei%pFl3hw<K!y?m
zWrQYzHoaRp$2O1@)~hu?xL<xWlZ+uxnW6A`)>LopX2@(WzKy1Bsznwe?9CB|)dXL<
z=1B8~?j(^}h?8*8?+pLi%4!K3OS=V#IaZc$s4i~-II9p^TImiZBI`yGFtS8CaoSXV
zZkX;6=E%bMFgpI%OGx`$EauY+@6N`)7%y0)?=27mp~>wl{&k|sYN!2^QgMNPq4|TZ
zLb7KMt03GmqQ>VG7H&b3-q$V3`;)m7mxO)pY0i@dX&(A_8<cznHyN(>aIfL+7;Xav
z^WHSz>nB$Hq(6AC_dS|#cyMx%_YktZXlaC~YSGRW^ql!#%G#ef_9o7`CnTxz?mHIi
z<&N)tmy*1{?Y`wdH`4oD5YtqQRv$SRUu&x({nE2S`edi~PH*nK&n~SP=vR6n;;3s$
zQCr;*!4<Nm^^p{NZq`NuzAO5;%PDY1_0mJ91u#Jx!PmRdSj~tJl5~}lVfkUILG*_{
zzdH?e@d=cqq<7V`@qXw8$}ERMgjZNB4H2RFl<jjF?<<?;B4>mnzO#_w4?0f<+}TYJ
zgTWP?hb+$jTpfgtI9Av?ngzw<{o!0O3PtfC`oiVzuYCLnr2+eU{q$Q@&5z=alBSi2
z6=RAB5RDX0I8tIt&*~l~Y3mu1IsVN<*YWZ0^;5;!%HwN#|FNc4#FHgt?V9BsaQ&)b
ztzyW2`FdvWp9xL2s=FdbcD-a+(le8POtNAi9G-~hU6yt5mS$F#V1dLPXSch?OwQ5!
zIIlN9o$Qzk#VW9ab-I_Zr)si@kQVZkmJ8<EoVu!a<2g&RSpse19<IwvV{mR>01<S6
z8(x0{F7Syv_;?Gjr-a@DmgToV?vSr{ht4G@nieBls4*a_61wqxq;2o4?bJA%%t-wv
zT~J(rz4sELM`E>$`<Z(ta=t;i*ZX?Rsjd10qkR3GmmBlWxx{5(n&%r1vn$(t8o#yJ
z(lVs19@2SD*7<sB6y1ma#o@jiw=eJErDt@Ngd&bb(zV7;rSmer+o8*#*0~A|ieDNR
z8;b)C{BpqgCD3LkEqIXoqYmNUJn}SHV?&GGU;-mP()!sk;r1}o228F@1qcNwqC8rx
z&%Gyq!{Dip1w93FYrkA$LZ{GWCg&ltLjSe4xiZt277IOvJN7?6D~s|rK5F2d;oumu
z%4d~k8^LcyDyk{4!YdN;(mwp@oRanAF>g9AV(5J)V>Gl}r-KTIh~EMh<Irs_%NIt2
zw?LiIzn^*(D4!2ug_9|6m{enOsg`enDA*lDUq}rx`5d<r{N~#iY7_as+3u9VXHyHN
z@+ZIfY#@xt+PF(r3}El5#q_7Ja9$I_2RG>;i@!NAhSen@!?u1NyM-=PP`!VThF2h>
zS%U)=%5H&B*%k6DS{KEm21R5E+zlVIT!aqnX}=V-*PR<)&=FEU(Iisn-(tRTHbqnS
z^}w!JRx!6|!Q6tW!g7@>x!9qtruA!~bG_AxetqI9cgB-3H@B*O&xSy5m@jY@vKAIv
z0#0pj%uLEQFGwkH-}KRMj-l7C%9H$+osewEOm?=S2+u|HKu4DCZ-Ed#7+TU&e>w=|
zgj{D$9KzfSM`qbMN%GV5!#LsBR_?XH?~7s9z^lwaR!=whc<;2u{YLm%eK)C~w%wF?
zinCscxvPQV`Eb%Wn}>+#gub{?(2@{5_N!0%A~;obd^K2PY?x)Uvyy??omP{Z*;@*x
za+85(nR#isw0CuwgIBQjV}{`sWu!@WKjE)%?u+sh`|*Tn@qM%id?{(I1$$j<dWCXU
z@%OWXZxSplI#!A|5MBIVzLK3n5hj1WlI7$h*Pk|3eYWNL_@eO@^cZAmJ>(4{ZlwmF
zbXmxZi@QL(4eJ|S+P?g0Ppzk<U8E|yGzSz#aRylUe!B>IEq|apC`U>f)FbdiTy3$I
zB$t_>l{}!`jXta<w6moaL>d{A3%=02wn&hek7&PE&Hq67T4DXi>FbWG4PAWY@~^_D
z7xhy^IaQ79i+PRnRnHf2@BNV8i9*~+t~_O=>;B^6-IZQ*E+b`RUt{R98nKGw9eshH
zDg4_*1{PyUPTp-vrs8}DmEv6&J2t<Ob$=$>N=g0)i8O<}55@o`Lg>Vlp)|c#pfo$x
z3ttF~dnO7=vEv#Uu4E*!GzPAM+VSbu+dj|*5Wx?6VPz1dtiia5^_L*XZ(sH}(c&9?
z?gCFJNdP6r=6!eRCA;{{lto6gYRT_|WAsplMEgrf9zC}_cnhBz;!**w?_=dNSMO$|
z&GUZkEUY@79jh88j%)X@o%NZtYvFvfZ*o@8Fst$GyzzD$33u=meYgAL6z@9*N-F@h
zdlaw^5r&_ScUy3^!O=+wcEi@{l?QcR#+6+@_CuG!>sh~SJS|S(-mA5@K%o7k&4l>H
zWY1b}38RFikmH*(>uK*DOG#wmGh}!LG8<X9+>zSIl|ymzG)2Pmv6hhRU;}5JOgx|v
zo_?jjV%cqo3|W1jay+NurNw@)@iha5#>{F$qlYx#gg*f%4)>Z{kIo&iTla82_=|Gm
zgqzE{AtkXCmu{IPK>S0o0AnhWvw)2{)7v5hAzFD_Maq!vsR5R<=a(<-xx-)KleEb5
zEh~tC>)S1;rm}Ttz=kaP`c3sG&!PL*KCxJHIHK{3El3y^Qg$7LjkMO8$U|g>$5wjE
zm5c~i2Ct}PGe`HgM<>k%iphZx9FIm7MD_x>>c(MVY5v4rj?TCgmr=j0=ZNr$@Lrs9
z7Zo6b%1}y(nBK2N;E$`Q(wo$lDUF(+GitvH$vRGg>ZNfhhUL=dkPlm#5Z-+*c7KLd
z>&$7jG`VZ58)`nj;S^K48^W9K8SOjmvId)@>-2Q=?M+N!oLkT&naYV4b|HCD;rYj&
z;XIq1cFgM*VBAB?E67yA7a9*XL|z;4+h@+<5n&e*cn}2>+==7n=1#vVTsexA(^>N;
zI^XLx5jt1h_vVtyz2^V!7U%SJCO*-J)5g6AatWtfI}CwebkSY8Ry!9jXv?quFszAW
zc(kwDv9S01#q>BD7hxV3BPZ@szN2T><NVf)*Q}&8rtHPW1CAp3ZJZof48&kfCsL96
zU5w82mv-Ob1^MK1k+fDp3Cf(ba)bU23|rW(gB7xCgPe*)C>>^vV#b}Ml9Ffun;GE2
z`PpV2cS-Gabai8SL(4a!AdMV@;vmVK_ee<9b`5IktVAegC{0-h%hX%>hHn9DD5mrB
zKRY)g&sOHM7?F6U=HlMF5}pE!%|1B4cp0A9T$M8W1u3yI!G~AMFS@mPka;~Oo85ZL
z)ir+VX>v1COTYSnw{IVM9=uTCL`hBskSx4JrgqM49Gj#;vP4}6?pYBM&GdV@6n<xD
zXKx%1!A$q*G7-|1oCG{IcRskTEkNSw!#}H}Gq`t=occ|8#L%+WGKvuKXlcBY6?go|
z4YHGB%qg7gNU*{Svt<~t>lNM>RRXC(BoHFVuJ$XnMha`cAJy!FZ>#ud-ipq9WcXP%
znrybxfz6PT?Y?SZhn(qi>~^n5`uEK%cd@MSztc=$d=~$?DUta7Z8gNCbmxJiwWjeu
zUC@#5%uYKsLt+(YU=`HCSh#n6aE)Tn;~?@1h~?@Re=YThpCvnfpC9G@N#}?elz;i^
zu0mchO+ShWc~UsLeLhDX%jj(S<sn<t^=mILW-K=5FlLgI%tDrSnYM^5to>%^wY?i(
ze~-1vLsWBT%5KMdwc=^T{VfCxcS;UK7o;HNU~<1@ezA|`?8{|Ej^@I)FVU6TOXT#5
zwnYDL@gIi^6^2YNXBXoNpJMC@a1>mTyr&fP8o7uc!Cw~oH6{Z-%vbkBm09(DvFqWy
zFS-7VE`D>sSljVql=>~;x_qTiV-+xPWm^mDMEFrZ5rheDI@%o0Kbr(8hg@cv%-<A1
zHzJX7=*70vZ@hWe0=K}xy-Exg!+`ozQ6-GATYI+f9U}2Pn9Eix=81R`ewZF}Ob9X%
zO0*w<^-{SI#R*?C<<S$vTkU7AZldKxgRijkAW(QdS`g{z?`#c9gF5W_xrUUA*xZ#L
ztPVPgIswhDuj76E2DSi^AS?DAPiARNog;&emqaw@bnI{whUxo!fpeW`iMvi9;%0rC
zLB3qa@n={JmwAtNp?JE^n5Fcw*ajZk<g^*ByHuJjk7dKqcXRxpjkQ&HTLi3jq6F_b
z^^G~q&U{g=lEf@bbplx=<lw-O8sM(;U7juN^k$_)dNQXXYP=gWz9I75gCkzxK9_wf
zx_6?BpLz=nZ9wOT(X-PNXUFP2mlbk2*N;uXnJ-zjG}HO=iF;2eBJy=Tr^EJkWsREJ
zdWEnBW5ub~n6yK^eK@#W)A~|lW>GviU2?{^0Cx!f4RjW|BWJL!Sci{F(w0PbB5r~0
zw9;NNAu4F+>uU{1-|f&Gt(IpT%q|znG>^F4mz|=or>oI3pkr;Z<!j<T3`DleCi}U<
znuYfjp41ijCeR0y3F(l~RNcU@i9(!W!%$Y=>n8%}z`1-yUHQI{@d`gV2Xmd37e+Aw
z!&&rVF`%x8Navnuu1>8-+6BK#p2%sR+_A#PwOPp|U@t_TAYdxZ@Y?0@=KJr24uw8-
zxqH&HOXW!|U~&B6ms$@L9d-g5u9+5op(&@!m%S5cb{nb2`LiB^E(XzFrz}2}`-`g&
zP!Hfs<n_IymQEr}1J<up*w}ENldIHVt6OOlBJ<Xtg|9BNN`?yVi$4%bd1-ZG^+NcM
zsv3R@K1gbqB?j<~!DNktxKIYSKvD??`(JBgvTN62Wu+TJF}|U#N!L$y=3-7h)ztj>
z{v$0og*QU)^NI%;Kk7UOak6@-xFIZw#=`uhdj!8Y9RZtbLTh5YzAz*-$uJcqG#uV`
zO8?~+0KVM<YtDS<g=vBc6smIS!asd*ihD%}taxzT&fj3HZ-Ld}f?@QhpB7&%rBT$#
zrj=!4l*njJjPvsp;%hf4E&95I_C?Vj8xQcqXcnNIXr%Wopq$nQCv!t{1<JNDLi@kM
zpwl02fh&R1ty^HGsW2LPfDz@+(-qt?H9|~C^6qH`<|e<-0tehtY7S`+d%2{#urm4j
zZvm^rOyOIgUhxK|#E=DQ7WIK&tlbHZlM`uuN=L_P^F{q9ORvH!^$Q!OxSNYz-nI@G
z%8NBv4~t4~07taKSI38`uLsmqK}`|%B}E;Is%PS7<dN?LWV;*h|4u5K|M-z95H}%F
z{Sg6MtTHZ6f*w-8E8Qqm>}5?k-lWSLEB?WtkCd7NoEeD$HCSuVIt`3DA*+moh7O}Y
zI-Oa5fF5h1xc}+OxC7(6!3c*9y%Fdh3d6{CFaspU>A~66nRQNrQ6EawSW_53&}iM#
zgQ){Ca#EzL(5bw|YHm})GlgxEoH(0yJB23qegBjw;L`70@4KlYJ$PFCbJ-5IKohM6
zGk<Jh$jo|ydNIVND#Un#pC8SzlWJ*`wkkkx_<M?jo5DTA;Apnhp}ds?qL#TjC1vI9
z{a%VF*O-tSUJjboB>j`O{`HNZz>(|v@xa-yimu~#*`!B70TSa7pQ7CYx%=O?FBlzS
zl2;Av?_Y%c?%|7REB65%U}1z_NWcSRu-|=R@`#!aDq(z9(>un%nJlJKr7zS5%wdLj
zY$1p+PYr5LPJP}aG6dx&bi>x<pyLM{+-tC&FXS|6i%PGkM1*(GeH{*e^Ylt5T);Sl
z8}aQZt*`hEC0)hDPR0oSm?icpHyq=)YQdg*6Il>`@Em7?K@$SXex*oMi6}*DDn>!~
zdiG>jyt|p8EJ!sfzFcPs-tSB8Qd`;a<cD>l`yR3{mO<jg{<5pw5hj~;mvH~ghxkSw
z0BCD0Q0gd<PS+HIr2r`j4c*bz5eV#F@ovYR&OotD3pVq?IeLB3S)?`f@9W|GkN(K-
zq&Os3iIA(K4Ss<^s#6CL!V+XF`uNYd+IX2O|LOJs$8gGunEI<zQZ6muBJ>dpEC?l9
zv>fo0m92j4z_zJ*(rwD&UZlsbVlf`~y|@QMz{_zoJIbI9-UN;s?~-9|6zSpov!MhK
zGQ8TtGeguiz0XH@M;^)9UQ1Xy%M`SHl7!A_k_lMNG#R9_Xv*I>Wrjl$qUuOT?htl3
z%deC;XLl!i;kM7XJp)p|MI<x)c_ziQb-HeWs>y?tT3h+#<@4OOWRU}xf>iv|Y;qqY
zMm=(7buOQ+C2ua*J>6|(hVSDa$xp;e?-yZV1WKrH+<)U}Ai622Xj|n+R<AA-&eUB5
zati7vCO&IVh_fMpf}DU;x9Cg;^ehsK5|`AgV73GNfh^Mm<8nl<Li=B2;S2S}2H<>^
z@Ze0BHLgihmV<}t-vZ1w6?NBrnRimA*0mcEf>G=abA;@Y4g_X2ygn5is~lkb)!(V!
zE^g~LGN?t7=1kiOji_CMK_1R%L`Kp`02{OA{=iy)4LJawS$f-OY<g3eNBigp>qd7T
zf~NGE)1mrVD+j_BZLxOlR?QjxgWt1NZ8*oy;_Jke&;jw<Fo6kDM1Z(GCa-!o729-r
z{E<h53hAbA;=GO=>RXHAYF1=3`Kl0(BjZDaMfe}ncDc3sgq6W%(ag3bu0VOiR^yxR
zwvlprt5E63MJdOkeNC>LOH#vcR!%In9T~sIRu=a77npbW*b71$R^DzSIhiS@<JDI!
zMNo#RXe#nK`>sZTJpP0k`DT%5#BpW1W#8+*39CQ>XqP^8A_EGyzXj%;f1xxHPxmZo
zPBG%rnqHx66Pj~V`2HphEg!-#e4*{%an9#<c@d6A(wyjYYjEhy6?X9^@D^SPVS#rS
zi?DT&PSC3)9W==;aINj3NZ5?2A0xCs5Pe<YpSqvrlK4l5<JXA<G`C;42}fT3j(^a(
zEM~ZmnBk6(!bgM-KC*WQvM(&$jCNwQsyyH?U+xHZn$!OC6%!1;&=V7?jH;u5wGtcF
zJie;l;ZN0GC4V-K5#Y}JFNxFQ??md@JbUXtadz6i)?!kLQeQKt>N5PcWicth&wj9_
zf#qgvQRW}W6{ui4bpTVqfmmoi?|onkugf>Ra31OI698TJzf~{lE3Rl;zXf`l=-@G}
zk*l(Cbr#O?*kekbo9zoOedY%j@6T`)CWlb3kPX%d?4M`~MZ)dGU8^<Kd3Q1UfYtto
z*<12HQtNV7KTLH<!PTp@^#$NB#_$azL&R!(a&!NhbuYswRIIE$hrDPMaHY5s*Yy!f
z3??o2^3$-L5ZApwLHUBT1H`XB!NZLZFtGrTdVhIQkl`;TYeGHmFAq>^p7^ILltt)p
zQpM^>0sbD%REY78C3cP6#wK)M>+4^Vg--UZoY(pO8VZxUhSPT!A63n3F;`078x6rZ
zWe?=+0WSxlA6SB6YrRdYGh*}l-;5a~mPpFJaJ<TsSFs)Fdx~Rz)IEGRju=dCI(cEG
z<!5({ajK=n92cVHep;wrvpYbo;dQVzzK%$t!n=8YU2_lCvLC6w3th;@F`E+CXmdK?
zdm8t=#aw#*`Bd(E3dt*3R(+{m(X(Q?*RVZ|E3xJl2(S4pZi=jFu!<OPNG1E;z<W7}
zQ*+8vO5RTOSLonc*C~^|ipyM7Tm1pgCIQ9egSd-GeuWR`r3#n`M^Zg}_{Ia{s8mV}
z6mQln-PJ=+n&73UhtBN`zagG0`^9d(`NTtCN-k)=*6&-EwPT^utM7GTbCd`7G~#&=
zt&C&wJv7Bnjv_J2@(3b}@#N**0*{BxL18F^vPJY*COaIJ{$VK3bk3*P;jP|Fb%OqN
zjuh+Z;%q$hMEw<4+KQoM`ui-;3QK6Nv6g(JT&19LLUKfZ`>7l)&4wi@3JTxqYQ0-M
zj#62<1rGf6h-;>w&zS!Fk-`c5o^7Y6nS=PkX)D)249<;9IdWO>xE&J?bCdO;KcQ}m
z<o<)^aXC2Sn0c~~qerT%BJCaRD|Y8xYOjWc9kyOKV=vf?VZq>SYdT#+`rRH~7zQDp
z-=-sU<{Vp=FQ~1`*9xBhN~W~!FIBEcBiX4br&uOm3FojBFB@V~KUcXIpVL9V^&ZMR
zo$l~(M?v%^7iHk$`0oCgxPXKM?s>GdKDASiCa%5ofw^LYqCJSf@rta#pDd@MW@>8e
zY0y?gvF&p;ce-ahIH@;Z6jyO>fe-2!ME9W|X?f7*qF{LrOZejxVDj!ndOqXYUi{SU
zgIlw(e9|0PVNWe_Y$kYQ$K&yHf$mI01v*aQiw;-{8VABLqQHi9)146iQvQsv{O}h+
zU1DyrdDN_ILaaw$i$6KMZ8aqZzICm0O}3?*b$~Yda^65uxp~yPSk$&G{o79$-^(n#
z=jif2n)ep?j=_spSQV`e<R6zY;0S|TU(I67B^)JI<7=i{pg*VL7MR|};G;@YXiZx;
zQyV_gWBA$VmV>zb)3&4d)!az;AI1CK-+}Y3Hu1=7W@vv720H%Z2YLb1nY1kv@BF#D
zY4yY^9hYp3o@6IRTx&qHd6hBtpDu0|X)EARy-hop6=U+}TE8XI4ovG2Np|R-EN!=y
zlwp4&wM?%-Y>ZTYoYa$v-PruSahxDs1G1)Jw$dnl8N@Jvn-P6OZVRtp4-XR;BAJ~>
zhM@)c4C-R6lI?NTxd5igM$c^@=CeJsi@vxRR@O$i|9Q~rm*ff8qol!<cett~K1>aY
z;Dq)*=CgHbEKKzrHpE>h7DY&UaIRQ$U^cfH|Bz(@8-&Iskn<k$y$`+x_FGo{mdPM;
zK4((tOfk<^7Wtnw8C%cMc$4b~2YfL4{-0k0X+|F1N`aK^sn_u?cGRNTOC0ue>gD$E
zP3a@?@8@dT<-zGaJ(?YEpGsADv;Lw*`}%|aLSP`&d&58q<vdxV4SF_Gv%N3o$HQN?
z{(B}zsf2Jhpl}MNr^LW206)xHYa=(`ZtRpxku-($sPjIXCX;5}<yFK7TWZ6@!dVk?
z`#9h0HOZ94ng#c9Cz-y}#j<c;w+vw=MR=X2PS;VIEI-jqTpQF8$XG9?=hcC%D>T5=
za|7l2Iu&tHbgLh?VJ`cBTADxNNxS8rNPL$jp#~a$OZU;QwTQpqX)nkPd<d~!C8g>$
zv7dn`TeD0&(Tr)`>-IMY=eU~Abc9y-1?_uLy+iEMMaKVhbC(hhR$K(Ka=2Q;qJWk8
zRh`hbRpF5C(b)mA0-VRA1U3&SeH@Ci*;arOJif@A?}{7gj&0|3aeqlLi%m1hUh=<!
zuf6c0x`dm8o39&-<<nm_d}cdk4th7yXwS?EtUj>HYE~G;6JFh!<UGZ>r2d*wU_#J3
z)h&l345aIx6tw^kZ@&fZAg!+WX5MVIHARFlrydjj8)f}_v3%8VLrjE-pf6Och3-Jn
z>6(O?{A;xx%6mR3?$^{JfD_d77PFwWR?Yo-V%D_PmNskE5~aP%yLb<yc`Q>4RcV&>
zln6>_9>oS)rC=i3!_C(4@Xi3CY1_%OCet(_8v`|mLk2wRS8wG@>Cg;E?TROxez!o7
zn?=1U;KPj)T@Hk94PIb)d#Ndo1}v(9&<Vdcau!jL*vTGUl1TF=yM4n*9n%$r5}giI
z-0D5~t7{UPVXGZ<vi{GJnp6RIG!@c28?66$qUI|j!+O;O4&mgc;4>q~4tc3U3?4iT
zT2ZXW%%v>u$W8svljecwwAm(C(eb17y`q1V50hE;wc?G+dD%NVuF>12FgcP`Mr`t^
zP{iiV)z!!g&OaV4t7!s$I@%L;+_hVHRMb9j)MN3_%1;H?@m3;m&Ul}mc9HTc@zfK6
zXnmU}xY29UZ$+p#>{-9|DxVLBw^CHWbwXXlOmUhYOKJ<7F%&5J^dDq)eFxrGV63#`
zxTsrz#BK`m!0Yu&SlyVbkeh1HG4H-<yN>J(2`c}v?^&}w%@j%+uJ&ccl2j#sC&bhL
z`eT?bs}H%MVn11DV+jvVKSUhv3nGP>r*>bdj@FGX(Y>~3=)Lb59F9)A1@5v$uqN-C
zRD5+Tb*z^OC~!@c@FuWM;bs=eob<pvfn}^B*%_1uDf@J^nL{A?Yr)4uiDx8ED}ob5
z>NpO=xSpALG}{W~Bd_7ZVeeWPC+k&hUE>e?mf0Xf$c(JsI)`i$DE*ERCg57DEz^0h
z4-;gw30q}}@4b1k>4U>&&r;fMi^N$wc$v@2j_5Ko#wXFsbi6zx+8|k%n;x6^gr(Be
zs9hJ5^U_g5wspuJTbaQ+@lmwRb!P2GCLvPZKM(S1qS>h2N!ZD73o@1XkX=XOk@A8J
zX=y?q{@Y&Rpo4e1_z{Z%Lz#CQ8q<Bmh3?`d|BAlTErpwr5wMFfFvD!%_BXiA7~u}Z
z=-G9HLrniLXqO0B6di2AhDor)Sd54g#nm)928_c&=q_{_2}Mr&Ve*29k&2yDOLyU+
zbwPP%&qn6?gBdCi9Gi<-0yR8Rc5)={m(%GGI;2)HSgTiND?(1&`_38a*H1UG$6v31
zZ^Z}W(VzBRFZH%mL;-vQ&HC^zy{Q1<9afQW-P+jS{avk#T|P12GuETMVa6C)#J9H9
zxy*>n>eYA3q(|IoiTUuox16LtsavI}gN;2Q8f&@2sjmYTd|^S1;J-pQmGnRZMyuE#
z(+3V+d`)ZMF4|d`uG3RuhV_tg8&5S;6dNP$uKZYD6*>FdOZ~~OjQj0SI3Xb;+J+HB
zFe!Q{b-J2(nb~>&MEYl!!_(Nl*s8XG<##vEI~a~igS!RENw3U9g}ScHBo2+7e=o~b
zh-Rhz*4(OdKQL!TsLSb4_gAT{lyXFe;TN%;QD0%$P|hAs+{wyjm!%VHPPu>5ILxYd
z)e8o}S!#bnM38)ILU+LE-xt?Db|ZO9#{xBR#82im48T_uiRAiYt--m0Qi?IGgcEZ9
zt2Cphi%*y+xsCHrbgKrbnIEuiki`44CjSDLuEHX8I=*suD$q^j)MgS)3Rztjdo`uu
zp=?Nd;-0CqPVRrw#ixUpuA;Lr!3gKV+4?MRLDt>vB2v*?K$SrXG?Csp-Acdc1Z5C|
zF>w`;yfx;1P%E4`RP*q9w*7j!0OKsfTDPPNAhQ7f>@jR*AE_NLa1wsRE^WA*e_(4X
zt^b?!B~liixZWXy&Vf1m4ouei6+vRIO*e7Ftc)Z1HFZV%fUD_>Q6+i>nKABZ6jS+*
zp*E%2((F$+rY9Oozw)s)aMv^2r$2$|&VArFu<*6G-uDTX&R8?eQ>mT=gIk;Hn_H}t
zhc&wRq^sni78LNX%A4YB$6*)0%^Hi>KN8S}y=n{*`+UU+n*4e+>%{&+yQrDTPfD8=
zR~xo~Z#6vG<h(bTDAF=8dFNGF7K|8LqYbuK!Tr)<9wm?e)_IZ^BY1ej+6v~KeZ>#c
zAZ1gj?qZZBbz=r$WaO@-mbrn_ZUB-=6)q4nSs1>S&f{uB+0XfD-~=lcJvpvLsZG_#
zTiT0x^(2^A&z^-5(n1k!81oZ*Pf|$keX6fntcP0G!^gu_4{bq8Kxr!#60{OWx}@$+
zn|1$Y0{q}UD%0V)Y$zb;A)4VnPqR~_`;UG2a034&w*vWB+pyzrdj9pctNHt#{{Q=(
z(jfW{Xb?>=n-Jn#VJHD_i7$nYV>MI@rVW~|_w~ar!+Eo#0ck1DOeK(vtZ1bjXVL&y
zpyG>aHRt1MF+0qslmE%5ZP?%-56;=MC{$T4rpq?$Cm-L>FT_78%3QhX8X^v39nA7_
z&Pb&3T7TSpRuHvB#@+&G3L%=`(=tKoiGye1Lp*v7c@S@al02L3Bs$Qa@4pLNL#h#q
z=bxq_itAsE)ZplZZ%o<>cD0T|$xE7?k5nGc>|$>PO3T?~zLikGLys`l$DrKF!yzgK
zGu_h(4Rz_=GW#}a@5xx^*-A1fMq<OI0g$>LCXw{XO~Kg#aqm@YLbfC0>#FfX){&N7
zjbpilN+Y-A39R_HGV7U07V>szTskZkh0%Hhw?Kqg6TQV6SK^%W-Cde^+E=TyRRr70
z^c*?AmGDQ!j0P3&-U5zsceh)`r^L#s)W=$TUVEJ9z$n!d<|KHBlDz;*GCUb0132+|
zT{!}$r+V82on}qkjOXlAmOLT2+{`R1;X%+r-%D-=$2z@Fb8HlY8SjX~u~lI5jhHq_
zla{ug6xZ&6)5rAnoF?yzhccUWpt{MSqWy9#V_~-r<<fTg`4?+98PhMH?^lzRhQ4BW
zDJrS>umis`o814uRywuHO({ApQ?NYW-kuYTPxt=(nD~97Sd%Or9~;9Jnf%psNx-lo
z!x8$vrBk|H{kOS!OOswf>y+`gF?s#Z3tlGDQ)O)7vW74R`uS78F!KE9Y+U%t*h<ZV
zt{AY70#c%o^LjU>a?huy@wG>5G!95&Hp+JRemsTKM*`fen72C%wTFK6UZxT_1ABcW
zb~e4nj%%HM8#8*0PojxSBfZ5O1)M+m$F4hNM-;o24hpQ3;(ifJ8e-SRt{WPNh7>V=
zrD2BlY*I$Pb-t{*>YgF7Fj6KZjgOR3h+n@-XDB%t6p$cgivf7}x?HOu(tpj|`{g(#
z(=d7G$-?~{1+#66#rJre#hrJRIlYIa0M>C;h_&K@(K@+wb2+GMjsI)O=u-{Wtsw?f
zsGm?5DCRjhY1KPIPOHo@BDImR$xPkpNuSi5@q1TJ7&CKH|1bm~cg$^ls90+mUFDvK
z1d|{>V=AfFYfpvREhA1)Pv9TsVs0`FQ_t)1mfXo8g;FF8%AB$E*o1OXz-h@pD6Ozw
zbNk2vmkjc<q|kaZ{b+#h4ez{nQ)6zhAL%a<f6<N|B^G%a#M_&aX2)es_{GiT7fhAO
z^jH=y3bIVmOiEVmZdd1>FgP_$ku%6Z{eVAh;{gd7IY&ZLbUfMlKMGU%qL4l7_DJZ8
z?4k#;vbRUs_YS5E?$29TUmEl#f(Cj;MB}mMMlI=}T?WWubPy{m#$pnMBIeFptk;cZ
z(05)U4@vo%=Kn~HnlzygxJkB+BQ~;yE891sNuoCZKM4oe$F^0><6&IAS!9^}?zA@K
zfgh#{K_W(W<$dycsWwmTpF-~0dv(dx%$^}^R^7~>@VGhXxi^_@;;{~>je0OS&y$%L
z@Sd$6-gfkB8TG!k9O=FGx!HN8oUi+#seFm0npyikeQ!t9_XU~g;*R$IKAZyJ-MfF(
zq*#gvOn?~Nnq?u7`>V!F^SUoIh1{sZA|8@d7B)O5RQQ0k41k`DTBaD)drMwCU$ivS
z_Vna)O*GDEYB=LO)>c*mm89i~UmB-l@@`7r8Ex89V_g-h1{h)j>i_Lm{p6_l7xC(B
zu@hKVd8!ciYZ-{klnX@1ezlEeS*O@>?!!l1Iyy27Q=gZs0oFesaO1M3;oil<-}l8-
zTww}v{u3Lj%#Oe3b7J>K(o$j+gqPJ}vO4S)nL-V~`CMj`<2{QAme>)y+vO(r&9U?8
za;F9JxEx=OsiTZQmFb*PHlzecy8X`Q*fRHl@+biH{g?~0PvLXuGFRMx`OTTa|Eu$#
z5slWqB3s|o_|Bd*kDq^lS0t|~hF+as#r%*o@%mJsQcnORO2%ldP?8V6Ux{7G495im
z{ER%|;!!9|cr2M{QMaPWYUvLo=oYx|*Qt0;8GZ{q$|fHj{6`_-A5gSnjau8PBCC@O
zZAih6wwsk9SK=9G!XhG3txuN@m&5}scM`<i0iC0=K*YtBK@?$*z9uRV7VDG&&X|Nr
zt<8E_o7WoaEUJLNxGJO;VR;^hskv$PDRD^0ODbT}B9sgJYrGBj$*^f%d?h@>b?Ge}
z^_B30I~CyRzNt(W1YYX5-qd35B=M;JFp0<ZvLQ`2ck|t&K_&oPDAt#&;2Xpkq10NH
z394Bh@*=d6N_@av)tLF=E)<(UE*O)7|Eb1Y?P#4V+5b+)kbPmFT&k!`XE!4Qg@=yE
zFM!22Gce)8&XEV_Bx~I;uGq1Srr9uc%CpmcqCzlta6|6-lV7m}E5QUe>bk1$m#ZY=
zigXaJOK3jnx!;3J`tAyyl=9mK$_n|XEYAqH$rRQw%32C7>?jkKg;42v;Abx<Bv{}s
z<EnC^b}Y()vv5}#_V^}Ftn@N~=O-j3iAmN}8h4zgGbezWYWMA1`M832%Xixxp_C3w
z;#2R+=`==rl0zp_O58XSHCghk%DT&<9%@L@?<_SA2P(37#eu-BF2vo8u3w4e8`25i
z7>qkwHq*=2c@XppN3%Ap;A|d6rmLdU5cu5^^;1b*Qmtv{A|3`ARrMpqxFFxNteN2_
zZEg`4S8=m;v=1GRNRnteOPg{GhKq@olbw;{=U)*DU)P3A`qan73@U3I{J`jGX4t)%
z_lZyX^AZVnFroi)Z&3A7?~U~uD9TOD9Qtg0H9;Wem#8scHkPppAD-QBZ<4#}Dtk&G
zrGML<2J6zV%ppDVWAVbyRYCOmblq=^H|?20__lfuGetivSJBF!>jt6)+l_M++36fH
z_0LHquw5$uPBjrI;y%tH)jz~ydR~T*bu!}C9IbhIdb8DUTD0l&=>3AB65$omna=Fi
zx>^AOBWMlAy;>v&mv+ZMwdw+&ZPlKJ@*1SYdU$4>?hE(hHyW55N`BF5in^D^N@u{D
zA7TmGwXC&|nnSPbG)jcCWS(r~Bc2@JIcWG0NOg+c;(h)e*;5SB$}*~&=I1{+;-9qB
z<nQ8t^I`pt@#jPQf79L&ZU2I0$LZ=hOXJ}?6O>}ld$*O<@Nq%CtywpbVofuq-B;05
zV+Egnk_i9$8S4GKC-AJ@7PYtQTUxRt__bikqQ4+<jxOyJGbN2?-;NS(5iD9r(FsP+
zG>|^5F+w@h<fF+iD$8rXN%A<GD%vJcA-E-ECAIhCFcVn6$0W9;(6q-hyHn6|-!z#k
zIzJj9=?=RYWpVG%CU>iVuCc`YE?yz0ay+MCfC--dEMkkeOWgE$ur#TCfX&mA$1Gqn
zK-f9bWL^V~dYhtPqDlRGuwjMrK9J$vKAa76zxj&hTY@|r7k4jxIX}0{=-k&*UQdv(
zDS6OtrNQxfn7i^E01=wKT<nlb&~cXY=<P9Q6=8V!RvAABR?ub-B6Ai&(IFEyccwuI
ziRzT4K97TItq9vkRPq6K#F=URzm@x@{tx!vGpfn1-5L!B2?BzkNG}443W)ShK&8op
zNRbXv0cj#lAV4690@5J}C{hIkM2vLl9mGQKNK5E7K}vuS&*FLCcYk~DZ+~N)Gsbzx
z`E&j_Oa@uGbFX!;HLrQiYhD#^aDnDPo8vs*RwCWEyo;GEcXqeZ4E@aX7#9qe$SbRk
z)YNfZ&@#N26dyvnLW|9#KuK<9UqE>)zmSD@t7;*4lwF?t6S@xLH5m41=OFVSXKTuC
z5i$Sulu6=j?;E9?>I(cm)cZLyYiN3k^e@ueS>Ff%LzeFrZ-jl?>U6%XB~NF5V?)#!
z{~Qw-Fatr|7#^W-Rlh74LdSUs-<2K^e;mZUBFC284+ISR$&+ceKhz@g!YkGJLjAPJ
zGFmadEYes;+t0GUGmnaG`evA7OgR@@PW!gkyiYk(*B2%9c|t;a;qevg+K8*F_ZfAT
zf?iRrzcOui*!@7PT#FHJ&5Wu010nm@7d1C>It&G!a#58VEdSClk#~D|YocjhFj;?c
z{py|DaZ{hXFxUNiVHKunL4AsPyHx~-wq!-U@)Xk;Zpn`l(Pgv?F%!`0@sI8b@=aaX
zw9fT=c=Q2~a{qx)EdnStw2f}UzBB^m%IZU#qWJRqkon?kECQ%XZ)4Xkj5f9WOyctl
zJxGv$l;0o7Vsz$S7q%L17YL59+oZqp4a-vpC#Sl+AW1$cFF%|_c{h)HWaLo0gPOum
zM!5Vat|kiA5cx)O%ZjeFUc&jNN4~~ZhbLw-ogphrSU}s)ejtMjQNy181DR1i{=E&U
zBr$wM5I>1-z^bM~*Ye#u^)kmr`cIPI)8ZNZS+sXl9HBMX3r8QyzfT&?>}J(?msc^l
zDVs`)+FrCZ)Rk3%%pKZ|fTR~+GUCw}aL!7ir)e7A62(I+N)CA)Wi!W8G&lO^rh6cN
zb#b2-`oWF0a`hJO<xabgU%fihAdB2Uga@5`YhGR+yFtveM6wv(fZf-A)G5TkOCvQP
zSPTjB4a`9A<Ask1F3V`}yxGL<-Xav|Ec*&<v}{jX->ecp$1ozKByNMr7A)xDzmKP<
za29od%qtyXngZ<sVy&XFIxS+UCDSv1tyov-f}fOpxiNvNq^?*6^;c4|*KbqiUqWFC
z1E~-adTjs{iJwAmK1XCvL!K5Lot=e+t;x8UyNcU}j?KJseL5-fg+Yn?`plLhH|IVU
zd{ZaifBK^FI!ogFK&hG7ID8=dcJ$OrnOM2W?`lU};@;agB_ZDr-*;U1<GBnSFZjn?
z)cv=)C?}zD8btK-r|EuTKR-9Txcxw^_CXi5L!&^jA{}|2w4jSZ5S@fFNw2>t9E>at
z1$m-8fJ(6zw|PkWoxB7M%NI}%r~TO3%P`=wEx3y!WF*b@gefu{c$XG;=h(B(53W&R
zv=VN{YeVO`KEv)SrD3S{4|km!|6XS~YXj&W4)|swqHypoY(8%6{XeGjQe2tdABeIY
zs4CJ6%0HfI_@lQJWEXG+OI7_TM<ymx_GJ#zH6TG|N2zYdQ?=!IC_VkVL;rAYWX1a&
z)xHf;hy8m_!5+1-_dx)zhtVlWr1#U)m?vY`L>CROtBJ;hKnn%KIA%r`q?ZIUZgI4f
zK;agUhb1NqxnVH=*r9`bM3VPG;lG{m;7jrrt0JS?f%lm=tFG%G2%iMH0iF&C%J0*-
zH}r&a<geNcWw%H+@1DOxh3-;$7mQgne1BVu%;Xr~x|$YvssLWO-E%-|4nl?<pqnRu
z8`@E-w>9|EF!+Qwpzzn%O+gkNz(X7iU4<OopiB(Fmo9*496dsI{Le33wzeW6M>i=G
zz-Vsi;y?btxjk>Ec=mylH^Y3;UJC!?Ko3N6(iE9tcyn7na*EwmaTdpB@#1`9mhj_*
zue0peH(ObxdS`lJOZA=zw#K<vq)tjR8~Wf%QSm2T+#VT|2}InZ0-GScAOi_{Z0bRP
zQ*Q6QK(Ae7qJ&zh`0@dL#T*u?MQOPUuECh<KmqU=ux7q;-*E4jLE@{wE8qG&s7Y`L
z9TxbuWXupE^^2KgJY6OiUJ;=xvE+MGug1~^E^9uI{PxlLp*MBnMpKZ?9!OH`9}*8+
zQ-5OHdzkCS?MGu*=6>B+Wwyu2E;U7m$P12!FbPzm?XirjOvDte2FlMGOQmxex)B+Y
zQJm#00$Uy!^H=x1rn6+GQRYYQ>hxx(`JBhEg+6$H@qOl%Dlu=auUpqw#u&1o%j?7C
zsxVX_5v2I}$B+SS=M{aE{4su><AR!frJDU-$&W+M#7-7Kmp@!_%z_?!0R(xItq4<P
zPDG`4dXM)u&|?6}%zVGEF0>DnzMs+Xhpk1-9+?MSZpglr>BW;&oE9b8uW4Z1NT;xJ
zhY8v{Jl!TfZq%-*6;n|xt~pTts<fmAGZZv%_UoGM!@LMan2-ber>fl`(jpeQ*WtNj
z!(TEms8XwUt}?Q3?aJWwM~@c;2iP~&_+(MqgyE_OKo53C-8;|bW7vTcY$NK#$N5(w
zlC=EKZLo(tj*1`Q%oV3K163D9E<fE1T}Yt2?DO<$-HwpR(`{_&aVH(pVmqRIFG`H8
zQE0hr@g$BlTB_Vbl&Fk><a2*^KPBIDhlJGtWo*)`2xev7LdWr=q-akKM>m1gK641&
zQmDC(LeFbPuhCvuuk`WVv2|`ZH?m!PdjZJgv>GB;CVzMeV}4NazOZ9nQs>`c-DX;|
zpFOgeXbZY9_^EvLwWwao*o%P+YW-{`<5Gh&cKG+29toMBK%VD^o>r^kRC>X$(WzzA
z5^Kw^wh0WPKA+{4W0?j$5L%S>)boW7MaVdV+LA9JT{E$h_6c-h@v^ua75D3br7A8f
z-!4NwPM;nB9xz*PF+3*Mqmo;qSq=4)JVT?;qvv$b+JfLTes6vC(oSXQf=Jqv@mxGS
z6cS`_jSMiatjRdF{dPN*&2jGLlYJ8yHLv1AL(EZ%vUoANg^DPo`c_%XxXjWLE=@~b
zHo4La^-=nomZ7lP+^OJwXh<5{uy=9cVZ2PA$SKVtb#*aV|FxvY^ojb;T{lu;AzAnH
z@Q77KMc8jWr2d=7bU7d25@a#jg5OmRdtq%}OPoLJwr;8>7HEk+Hs~w<#mYYuA~D=h
zcog}R$LOyC)V%^{S*h`HoBDE}P04jhcJd?%mUeoz`}!TVC6Dd;N635=YiO%t!)GE#
zqJCNGS{s*>2PfM^##o1x8mD(RN~@%Lkiv&dZE3lby0!XOO$hs7u=V(*dj=T_LUjAr
z(Bml`>r&;i7A&`P#@|ofP&?R98!1IyAoUdh#o%V-PO*bt6<Y(}vkfatmI%ccz7QxT
ze@?J)1=Niru0?!kn->}ozt`MMLqpB?ts#kYcGlX10wEntA6}bx7-G>VjTy21>z4hU
zUUz@uhD#!PJhAiQXkK#2xeThR_?bY|w_Uf;X^sjeX0j+zJVGW$k*RX&!C3$3E1j^(
zRtUzlcgqoqKl^UwLmQ9eA^*%DNKPZkv>#2YeC5;kLmA|FpY-?43g;)APN%GtzkLW!
zGU~ba*dHG~a(TK!IM$aZ*`JngXI&U5Zn5tg9U5PzOf!>tyb7F$d({J7p1RImx6=f!
z5%<-E5?XBwm(}=Af_f_F%<a>_0;Ul-gl|{1#+`EM5Bg>!HZY=Fedv?zvu9Fy=pj^6
z@cg3j@61V6m{Ba%*Pfx{lW}@iwh(}*iRnYa^en<9=&Qx0ymw@gw3SNFajr>C-GcK5
zVaR9?{>R=Nmq)uLAu6=O6F|CAH@zFxzgdy_Tve#Fe?7OR7WfcEN+jFs;C-4lZWM8Q
z%i~kGBj<udNu>xVS#^kf)9rDUl*LtAoL8vmmgxP=`&KL}1<>C)hg?R0TJqjJBDoQ;
z)qf!Nfdb&6B;B;<&36&|n8L>Ls{<1)5%Y3<bVZUzGr-FR29oDG_{ROMUO0ghK~b<7
za3a1t&>y3qxf|ZXc5}sdjp!2RIeb63by_#d`YfuHsqx}9P{5Y{v@RYCK87Z6>VuCB
z$#uBQ9>P2EF9xzWbK`amn1#QwY+h;P@gb&H@BRcvGsx-n2RY7)oh?pmx64!4*WCoF
zlLV}Plms_{+DADCygY{rmCP)hFqonA2a>4*ju|R|EAUC{ik%%%6Nq3%-`M*D{FH%)
z#I+*&lC7M0rEHY9$YJS?MdesWD6VHM(1`M^s&NmfB3c^ai22WzZ)BFrg+7i=^4OhJ
z-JYIQgGg#XE1DTh6_zDgV1mK`5T*h>MGkR<eBcNtofbb{Aw@x~$YQb^&=hDB87f<b
zE!zSKVZWe4(J!7a-NvX;WifT5*Ie}}&qgx$W*b`?l1ZVnJVD}PN-aL(@2#R%>Dy~-
zg0I;{UiwYA4G#L`!h}JG`0E}!NWeo}`vWQ2fgiwZ00;;J-a<!=B;h=prAE3snFk*%
zRf^YSbAE=B?1V5>Jh8O+f+d$qAk87qk-&a`(YJkimpNsBWNGq$Lw4AH^b5q+?K&Qy
zD-6&-VyW0KLF57^zi}=JsYGeo%{;j12rd4w=Q~^(5H#Y&pwzJWde&?Erx~_8GuD40
zyBKwks6Kd&SKYOs$B8ak$u?H*a<(RNR$YEYN1E#3^ifqF(8Fu?P>!o)`u@wSk@-nL
zAV&O`_x`89NOUb6{na%;F=hGvl2nC|iTnZ7WOZv=Yj9IK@b=Unh~5{>)CSC45F>k4
z9VVEtjUAX2@U7|pxN))bjp*g|*f&>CWzl=@KL9l}d`9So>F!3O)vcCIiYmaNfsN@9
z4aA3}eN)7TW8bL%q07Smd+$;0n;K4_JAWWH;}#8!E2oLS*+BDxUf7DSb2g4|AL5Jv
zI}Iho`48m3ymo|cZT-uS?^Dq%oSX!w0VwMch(WaSJS&D!V0Xgg&rBaOi~&#u$`krO
z%r%IkyRH!9jzYk)*)}q@v`RVFld6KspU)c8lt^a?dGr&022D=d1UZ}0+k8iKU;H;z
z!Smc@Lx?+ikqzdo0J`n@PrF8@XAR0or8W<;Igz{d*5JjkmHAH(FzlQf$cgNeo+`-I
z_Qyd#cv*<l2=?v38+grzs?t)OxeTXS$E2G#1;nR1bQ}zxB?5f4cN*MagaFw9pe*xT
z<eo);`DN)t7arC5^>ez+NohH(*Y0RaYW?N0Ia~>Zv`C_ORKo=efV1wJ>z&-vng4#m
zPmf;6^^!;T%#DOt#pfAq(B%V>Iq`_hazs=)me(cKCW0~3)i!_f`=cUJ8yAI`Z*0PX
zg&ZC7%RLs$st}YXQ8=3H$NLO-Su`(ffekqy0vCH{XZR~RErdqNCyR-mR5sJYqE3`*
zTRS+P+ohVL(&olftBvLQ`XiT>+LQLcG>g)u$~xj2B&<_ty)0KbY_jVqgy}gIO@lfK
z^^v^8q<_1N$hI3p3`)VLyXp014-~I`d@^6Ik$hHAE_q(}Y{FnPp4)+SG^mG<3#EmR
z<I>GWMsi&d#EiX<G&|E_du5P0G2UxE`Yx>hsc8pA?kE>!?|@sk8PJKkQ(013jiJrK
z@JW1*HoaIXc}HrX-Z!%>vm^5i>34fos_@I~GIueHg)3S5qVI?A8pnF2;|JKYm{LK4
zAoHvlvdXmTORu-M#FJbFT?cRe68^Be>eOdEPY1}Yxy=T8=@X~09}vsFKFRd&>#N>+
zFvm$smt4QgMNNoCE|Vg^5fSTRc=<Z5Lan^k_Y>w?N?PaiPbDN7<CDL3$@i!a*9-5u
zlDe><Vp}sW2N_APDPjKhOa3!deXaGGLL0C2nMFbN%?wbf&ht0NkByXZ8xQCxU14Kc
zCO9QmM$ffFTkPy?nkqH*8-U5Y13*8W#bmW1nLzP2+1)Tp7j%V&O3iE1uXEWpULy(z
zJ8M*%gQjh%aqB90Kk<CbH?eC|lWDJUjtrq}pTz(wJQ*OoHbiu@Bt`gs90eDIQ0!wt
zE>;=7G4x3i>G4Pd_^Ec-Euz@IfiP9<^tg60qu(sY>RGZKKV#rer+a!2_YReG>Km>s
z&^z+FzOCE<kWZ!NE^MJ<<T+W##Y?PioaRGm+dOw5=4J>Zlk-9N?sdWjHfI$1e(&DE
z2A~yr$M~h*w=BTr(BJ%;rt0wAhn&B4vEu=q`d(5^HO6<#FbPMmE8BaZO4UtxnT^IW
z{dO?~P9wIKxBg`5uVw5h^Q@D%qOMWsb_Kt$W_YcCg_Vy0mn;41p*kqeTCoK#yY>68
znyN6BX5X(w$%3vKyM(S!A*FSuMd<rTgxN3y+>z5eZ6xT<dV@o9=<XtnqeJSs8sFOi
z61*+`FQRUi^S7mTvp~!pT6vYCrN*(%<|+R0nZ-88C(^aAFAa;%g>)!{ClAPn88@8{
zp;>?mL$_pFdk*zT`|o{TE*oS&yr=p-J!e3v8tX!mAtss-wOTXTRxRd}%2OJ6pKVaR
zcs)Fes&dcuDks1Gy2I`971p_h)N1pn6<p#hu$Si_*BRNu<^;jO=0Z<6+c}JJGL^7)
zE)+0ty*&DxnZqd9T}#<3dW2bgT@H^IdEJfqU>5&=FpWvS-e!P+#a}DJ@x7p2B{G)T
zKHfWEQKHPf(xDixNOb>tx|s^ury^g(U3{3YK2)j{7^_dfen#6K!HoM<`Bub=MtdA~
zYw@C26dylQi_D$bQlR!)w&9;H?}^)EiOIVxoIfTNH(XykyV&7sYDyBviL*NUobO+;
zkWzGweidsQdR2QpW>kI^<hVA^nYJPb12%BVZW1tX$Uinnc|tJl2sAPzo&H4XxjV3)
zHzfULeDED03f(5xJr#qM`{J3D{G$m}8{4L>4!dtj!Cu{o;Io-WNr4iFswlMvQp}9j
zkc_2cR((ZnfP#hK*0+1id!ZCwZU(#U<7DbzSjD64r%^|~Gn+5Bj|9dxu+|rzhQaHa
zFU3so<~x^k8pVqAxx|than)pI!lF6xWO|L2)wlpHK7Q6$If`9))U8_y&+VPX)Pp8z
zC~E-N$zur$SLXn-I8bbDZnyxz&8@lt#w4$!9I>zUy4H<#5~o<psy_J?%CD|liIb&!
z+L$(%;iOZ4AY-P3=o4fYKm2s~aB6oR_1!7IIO+z2ed`py>pW!{pc=86j}SqMq6BkR
zQK-5AD)t_Eu6zq_=Ms3`$yC@tH_?AM+v{@Fk4N28)y>_Cho(c`O=Nds-gIyS@05$;
zV`e?S1tDg?+;vT_m^R<e{Bu+UVX{CFxW+8TVsPI)(L|YT^h;$n+@+(0Z62|wZ<7a$
zuUYK6+?Bf8X8-27>&hLTJBrcUGb6`{RfY&VJkY0{K=0mLKN#t7y)&}wKcMV-GPG*9
ze!J0*_QziaUwJhoZND6iDS35l<Z}n3fXJFQ|0OtGsfqx>$(>2o3tT-^T^c!nHUEJy
zE|1rjjZbx8YFkLZvKJCY<d;C$xuVGHGQUZ3YJ3Tw>PHjKf?7gsaS#i#6v2&B?jx&v
zJxLy6bHr))mDXZ3qZ2xB+)eZ8R#;Wz2|Z`}hSD3P+DP)9F{@A!nBu@babxM0eS7iY
zH>KnWvX!Q$q~c;FE$qDk;xsVFM5$4B|Fw9Y13M@;+$MBse0zA+Rc!l&R)nRy5r&cJ
zy$0lI(ta1QlX81As&P$F{5Q6*=D(uDUfAjXj{IDHKWx;YFr&M!5`2%It;4r(_rfd@
zsnh4tbA^EUX_ad{P+8r6-zd`{3Xp>lz#Cz1=<R@B*p+|DQxF9iyufjd5v1a6`oR4Y
zXuP(%bGCJQi)sM9Iu1XgzDpAtWM&Q6)WPY^s^8eb1M=ru&>EnyaoTNBjoEZ1FX-OL
zO5VTSRWaz^;J@(L;Jtrwbx)}d(#H#s;S@E_l%oJ0u-Z}^|J!Q_x%OY8k^e8C*}fFs
zp)-yR>5IEol;g-^VPS42*#f;$S{_p_@zBg;ujJXXA(@`Ej+<s71MUcZMx}T%orG*Q
zGr1HIA{BO03+~IV+Av79;|*kQd?!$kfD)vgl#Nr!ChX>Wx%mc&4N8trD1Ys%O*X1l
z=h)%Vc-mOT7ECs9BD(!G<hHGHqobuoA;am(jH6Z${kza#H27q0#xa?EqUG*$VsZ)*
zYT3WdR#x(PV8f=5Im_Jq$7AE_M?5;zGf&yMx9&_gF~!<kB_i5%^_fWV<C7E9TEZy;
zngWhIf=*IR`_JgqVT0TRb;wp`)HG9_HDBB2+3$Se-j=VgRd7WWJgTVhOTc3^y_H@u
zr!glCa;~B6Gohr^PFx!2`<(2R^RnlYxAsdXn|h;bSnt*b|MFVyZ}<rXgyZ5!yM8$3
z8X|)CA_mUrJaXoj^VwhRV)aGeVlmo3awS6)nLC{UiD}Ptl9ZxHO1YF@xRe?inpB_m
za?84@cRrKqMi`{(=79XB0a^PfTiGQaiMRUPf3du@w6rd{Uq1TO6#?1HW@QX>@pqIQ
zSVt(`XijAVygt{@?Bkz#Q>;wyTuGaN;~PZ+4LkSVHt#-!F#WQTN%;XRc4tOeuBbaM
zK(tV-J{OZyBw>Hy)fuEfJ$zT2;P7(0@ul4O`{EqL@~**X-NlU=M_(nyGv}^T6Y}Mk
zJB#f|tGkiJ_Nr)^09OGQB`&UjH~f)jpOnTkN6wWoU_3HA>Y=-jNu4+ZH>u|ECQ-`L
zDu0Y~;Y!ZSTVL<JV&!Kz<4XZ}F%$KGu-GBnaCzu~8^bC~f|uYa{&I}jD2w*hZ?uBc
zzSsvxabzx>7+<GFWl_=6-gsdAK<cQ<FR@cjoMxjC_7h0Wxu^KyYV7%-z9|I+%Zg53
zkAs6P_a^uCz;wIO5AKpbB=tFj2MG{dz1;32QZo+DL&`DgL)~`WP_C8oV<2`2+Kn%T
zYD1}cW)>@__GiG3CV~?S*nLI(F+Q}qJYa4%w*T=$<@r_d<lvbTYm%LV#BO&mn$fN!
zsTIe^LrOI2Q@$;>!Krl9@AMCCfh!e#bgajtJ@1gssAh-k$K4~j;q{aZzLb3-bbQ`w
zn&-$W#~2Iwn%=$1<nA&Bw_~Cqg=-UO>OVLiXv~M^j2$JN{-t@z_gZmF<tYf`m-avl
zJ&6fAoi(6^d7B#En#nPS(07pw-Fc4@zWkH*73PBA7kg^v4Ur}N(ajeB-S7{{@VmoL
z-~Ln`H#4NA44u8z-CV^j0?}?>?XO!jJjz1_c%VJRUhT>zf0}EDjbZe~yjnH{g`xEU
zh_zOtr48|@yCR<yZ~0T^oe<ILR)|zZY=5X0EFo=YZU+zHJv5&ZDo_3cX|2gi|3c#u
zW&Tj@4%MR@uoV|gAW0dnA^GA(0^U48)2)W$J);d<OgLaL>T9%n+@=HQlB!Ud9wsx=
z`%UZqjdBv!W{ltR^^Ye2wt9@atL=}ZdG+K4@_-(2q4;&hMKgl<%kA&fOnw3He@)!T
z3kma$NKzMm*2O)%4v6@C55i%`^&TO&S-k*H38NTQLrImEnUP-LbGi2nrQ<}1!|~+p
zCV4n`eH`m@vuFLb?<V-t{*LiX!LOMU{=X51!}?J>KE2Aq#V)64)5{O9d}`CE4424u
ztzJ(9Xr((sm5<cUlT>#Fi^?q&)PEfv=U%a`)eJvD!iYapRT|hRFN$)Rv=*|IzXnSY
zdZB5W^q6(=S=ytzam2V5PVn|xYAEJ}4xMOkunLW)%LL^t`Qh+R;|rTBM8=Af72P@6
zA~mH<Vh&vNm+#ZiD0&RqHVQ%Kb_r}zLoZd&kHpI3^<RS;!-{Q0OXdk=0^9-4;gq;+
zG_j(X7I!hP9REF`$(Xz7KK80dngbCl4L}rz6az|W=*7t=;>aSu8hB6-hGh^TK^KX4
z3SHbhJ}iA)q9aErT+P$Hyc|Q4Ux7J8_T@?R?JMpm5$qzRdEgk@n6@@jk&@8%o9CAL
zos8N)teCobyE@5j%su7Ry6X{w%5RciU$t>WI|BXXuef!QYv|=}i6fC;sI$a4^9It%
zPo&nn<&|HC%HO08dYc`CP#=Q<UiYuC^m<JsWdA1Rj6Rz<tfhto-#tF5Bd{fCyUo)p
z6b_Pu2V=dxSW^2!_QhZB->&1NxSi?6?<r``{O)f+!pnWzjN)ZP0ITo><O@Rs;fG4d
zU0{8}77XU`LWKDGlc{cVf!RmnFK5Bqr`m#<2lz9jf~i(m7c<MxM}VVs7WVS#B=6)W
zw_NS?3pO8F?0leu>n8Zd+uMrxpzEaftIcRene?C*_J!IC?cs}w8)1IuuXhff-tR=&
zMnXvRlrwNPz*Q7t0W7Nl<|qv#M{lH0>$hgoJua51RS@0C%dmXLki*FQa^j1(QRmH{
zD8QPW79?^}LS)gTlYAu8q0AG?PJ#HIa3BAM;&TSmqH%6lqC(^IiFlG%Lc)N;0fgxm
z_Uh65r#$*?a60kgr<OY6*B+)SJ!iOi%|K_e+*^3m%cclqGQiDmf#E>0E^EamnN1IP
z#EkUW@Dtb~_;&Srw~hN90xi0iv5(5t-`skn^EPzFk}=<L5zD%B(XX)+I2cZoV6)pb
zl?e=z2wg>swgJzQ^VBI5<Xb1@c_h^ppw!S=G)Y|<0rXs-8vj5-mK#Z_hgCC-p#?~{
zS+?-<`bWPe5N(*c1PSUazg%zFv<pM*&{hgrI=s#8fEEx#Gw0B=2n33sXx$mpz`fMK
zUmv0r@b=5E+X)y`f9*cNPF}vF(BY8?+_Y_N#1GApZl*->R$a%ZB3G~u<%700n9G`r
zTWy|&SL@B@JE5sSF$w*xOF6^Gjs~|MJ=1^ycb`m>%dsJ$XT!sv)|>DN988Uhmdx?I
zz<6{H*Iob{$p4Q(|Bp!a|Aau(6pcBPO6mqST%SLdy4APQ;gJwMcvR>J`1&?DE!pec
zD%<(fv*|W#x7tK^ljD><uiasnJfW}}J?H!J4}@egB?ml2;caM=Xy);g-R}>k&6dJi
z1$Lz+eIg%)UKl*EVC{yw0XfOpV;`>&kbx(q9ti`|S*CH;3OP-AG0lRW`Glris+|r&
zoFXH9iF`Nf8+&)eu6F9xh3z_R7TcV;^@RIsIHcsFM~6pR>R~A;^TUup5=1D!1HMZO
zAbp>nwd;S)8H1F=!`irKe;`|?b%K8&4=$s(ug<Ri&*DW@?xU;*9%Kz3*H9h!V0O~<
zZdtu|2FJ%@8}F;e3@pJj^ZKxi@N8E2+B{*ey%K%?X`q%NVYe{f+}tlwS7yDa>5D`u
zBrmw@J0q+<uv$lrUIXPvOu;wClxM!I=}5_%zWILCY3uF17}s~Ns%<{~Sh;h@p8*2F
zD+ufF29hSgbA1+dyQCMT<99yWAqZ1&i5BN*F6ab39VwOm;!L2|L_CuUsgKapaTAAr
z`AG~V_hH&%%0}0hVb`QopGzp#4E2qXN%{HLgZmlbyY?h++;_ex&L<@w92H~V<}WWx
z>5A>tHrJxWIbSdU49%lR)$_QA=!IX5*#>)h&D5K$spfpAgTU`&xw>xLdEX=q*Sd4(
zGFl-FOoqN_2W(7(7*V_}05k?m#+|+~Oi?%i$HbR2B$U75I`@S73N_zD+Gc9#%D4z#
zJhG7w<?(<BwI0}3B0XJKZP~KCJZtHgCpKSwV?JG{OHh7Eo_8_-IEPEBYDb(ot4<Uf
zR1B^7b*<~dGXW&Wxk}A;NC*wBU~h;5HMEN&f|4Pok~N7!@kj@=x~eaO!qs5`x(dQE
z{aV*@B?~#4Ie64R{oXYs@I`E6BUqekI>HN_Dou7{qxUK|X%>PYcblFpdBvt5L>|%r
zs6rw|0WZ~f(y9$J31MCnt2<_|Q~WqLRnFA&+szO0NWmqfa^w!vm|)Y?r9Lt`u*gO6
zBIT_;PgHZwE-bjV_6pTah~*1~&O0hDsl?57D1HLX=^}fenRhxgrhYJ9<4I+SPPrg!
z_oAVWeAi+lJ5X@GX%HtBb!3P)^r7^<SV3POm)Gyl<mz#l+%=tpobmMqlsYXjvz3W!
z99d-Y0%>YdZQ!|9W-+ECednVmT}>e@L^Yj}R-hiedy^m+@l>C4WxS&*2jP@_Q{P2H
z?z6P$mUP1B@sCve)Cww!rhIt5aW1gq6J9yn{xx`iU=1^i4;vGE7_a%kC+^K)i7<z-
zicS+0;_nM^pH~}2j(8{k9Vr9pW}+nMcW#vqAoV`F2M}tW{T=o_fYp^X2-r?VpV|Ot
z3hVF_2E)(4T}AUUgsJCPy$xvrJ;BW@GTvyXe}a^9lNeT9?BcJo@Ya5prVLTp(p&qr
zAEYm@UZ)N|Gq>HvOPWDx9cq%*sz=1_2V~f#C!{VgTXedd3z6Zbs_rf0Gg?i9$n#bm
zUPoCI#bcE<J~#$lPUh-}?VW1%Ecy7?TkA8~#^8oM&!BBy1w3!;xoAn+9ct4S|DJWA
z{M|IJ`7*F`x%^N^=#q$WMVLCvuJw->B|E}}Yl^0&l<$8a=aqHv>6WJb%KD;mF>bUq
zkxREq6l0oip9#Lg9l}V(J(m^@LG=`(t`WO43#=@3W>q{*L%(r|t=T8bHm`j8a17m-
z#7BWCaK?pR88Nis2msu&Q;s=RF>V5mc0Q-S%(6qHnS1U`#stFePV4VlibMs!B^QkA
zRd=j0;DRr`YdAd!ckFpNFs(0Bi9Pv|MZvC0SKi##5P8@96#vsMp^-`u%bl-N$4#DX
zcf~ylJHTs)K6#g9@bFq5;_{7P#e6GYAho#Bhd4P!=|eCt)@GQ=ymrO;ihj;}b<vi3
z?y_iZyrQQLHVbXv7-6v&x4j<3bXJG8y*rgqYL)B4;hpkh|8@a1PP||TxRQDua|<7t
zUMjkFm(A!iJ0;T<`v&A&S>E_Kp$<^DXMm@MTb~+V>J~7;Y|;LT;=u_R$KPHqyH2LH
z;wuJ5Bc`*95ycc3a{5lgz{S$WVT0lB@g+Y)30l(ylX*{n2i|&GnVr-3Nqbg_t*Z3H
z6$KY`=0BHXW(7TOlwHM8bw2KeEtZfx@ZJyen<;`q+#c_Lpm!|}1ZTBcr_bIzR7SRH
z*Yqz`4DR^)R#a5;8b_ZfXce(_y2vOMJ5@2)AJjX|GS+1sBn|sqQq@H6ugA|;wIBH!
zk+^XBXIFYBtigs<=REHtO1@(F?F+YCGkx?kpbYueIXe<l*7LmXKMtb<@!7@J#<NUg
zaqXN82X@H^9r;g3e1;P<uW1_0&0K%%YCA7W)oCywGlwAbPM?J@-Pb2m0kZ>~$co!o
zYo|qi9EGo|V0Fp}F+Qd8_9y3y-@&)s+qEE+-`pPU5R^G!Jr_GV^Pksgqc_;#V6R6!
z4EmOFZdh5E(C>&+f3$0ATD9pDf1B=W$!okpz5Z3t@aNqUsk@4J%5~(2#b!98OmR>L
z?P5*EEi=r`tuTy6w8Rdi!DE{JD`50u`~bf-NpC<DZh@a!_Lru?QH&$-5&jKb314~h
z>D-l4_t`1Ejb{}AbsTaGre%SGQX^m_yy5z46{0Mx2vFZ*b|cdJ8p&i=9qu~Ei&y2T
z{j!p{S?ZrPLH124tewZFxgJrNLbrJAJh%=Y%V|x`4^w3wvv5g{bMTYyx7ufMUy6&#
z4kuKV%=tEBQ$SmqhKm3?H?vV7mZ0jNbtk_G&f(0fNOH9;auH#h7@C?g7i4F7)ih=7
zp1Z^Ek@1=v!lZG$BC!{b^2t&~$o$!-l{KexrmQEdwVw8BeXqIt*)@k0_p{eCz>a8B
zVdHXm137I#U~4U`&||$I_xnzAiO%I;0;23G6TR_l!N_i@gk}gS{J&qm$VO&x2Hzk`
zoRHFbpjp9`fTp{>)t>fhTIws+9#u!+Zv>DfwB9_H5oCwK|DI>O_a^TZ()Z$VO7{U@
zwCH+q_Hc#y&j@r~yyBbZx6j3T&0)?F=)G#$`T5lP5K#0#KOKrD&`g7-03&0Wr^p5#
zczAf~wZ=edlReV}iDdbRQ4Z_)gbcNdUa=l&b~{;&mB;DOMg>)QKz3RaeUDz)Q@ju4
z5NVm*=($g%T1t!LABZAkwrD$ZmW_<T!qB6<1Hi|m7zp8(;Vt6(7i}0p{#7NP5@7+F
zF3iGX`2d!3Ao5w(m-NZ-n6w(zrS6_ubIrI<9cxuY-DOB`o94j5x(CW^_KU#^#dm;Q
z?j;i7dQW<R36n;*d?9uoooWyQ`yJ?fLDma;b-Wo&6Fw<}XzYet1e{=cyq2}dB5~a4
z-nTe?GAoFF+D4Z3z=(ab(=Tikhj#)3SP322M?&{Gaa-Nm0x_PF?P*k<R^BQ_+i<-~
zS^vO_64wRxWR=klcgRn(V@<*TB4`nXJO3%I9xQ+zx=k(u|NL8Pc#d<2<f8mr6GB=Y
zKJgxg(-u7s0z91j&bE5F32KNwx&C#<U44(5`R-lVl70w98Y$9d!&^0YO=ewAxBC8h
z7t4vzP<$dJ9?D$pHyfsY5JIy_7M(r;9tHgDCYik*_@$XH`~=Db5W>;RH8pO_lLG!g
zXe7`?V|2?;g7^`mDxkERqY>nLL65-=&z&Zeq3A)QlDP0jf-Q(fS+)M>@$|uX;!V^4
z9Pc`qLNMTWf*cqSs4tqM;l~348WzYe2&LrHe~*YJ*fN0;ZCr_)c)n0qWvUNOin4|F
zA6FEM%adOwdwq;zsabdMv3eGa<k8vf!oGCOI`O|_>ITwjSDBFx@l%~{T-vKB%$b~b
zpUmTWA^NjweD4O2i^EI-AmOISktS}El8TmqkwVVEM^TOk#sC*y^9d8xE#m$i&3GMF
zuR}PCcZl%@kwogFD4f`B70nK^4u?r)q#%c}0|VXFKM-Gb1#XZ4AdbvtMk8>^_1?C#
z>skP<r%m2eNviD*<tn+#$bLp0mT^QYi~ww%8*YLrO7Ri)!;<fTYgi_Ck7!lpCgVDx
z&bjc*>CS6Ry?l?4vNa&63&fmRbo2<$_vI5;Wu-#IxXt8uzmx4rwLQn^`h>B2Yqn7j
zPc!v4yiB|`?z|bwxMbKHc$eyZ9>__`4H&iHna3fA?q1)IwG{pgncF_8dJW}E913p}
z^U~`DsAPhkzEDgsQpqfDa=yfRt^1>=NAG<``3)anT$UNbD*@njlnS)1s?XUZh4{T9
zbxJ<<y!QBC*yN(L^=25M)13)UY9=&xL;>wU;yB66sIj8HF?EYRO2i>mR>#hPyI?lV
zU42RG{i)SP`<`$c(Se{#?H`>Fcy{09kE!$fiDx_NmSx{FW}vbiOmm0Kg{Z-84b)Du
z67}0aK15ZejJQy5pW_UQD9-zqj<=t%XccCz)@x5cC4?7VwT@dGAU<!O5+(i`7mnD>
z8FYPL>169g?IA6i@s`z-(XZQ|!8R?}kEzWqbV~Mlr^RN<Y`V$D74eIy3hQDvj_Ev6
zR~T}5Xuqx@WtOK)O@r(7-;p-{K+J}>1AfYWQ#xPub58H9Qf)TZy<j0oxNsHgP9E4p
zmVpA&M!t;bgM|<}il4rlA9E<Lz&PfbC_OX1t3DIw(K{0&GV%vv=SGE0Q6nZ7e_Tz~
zSa|EOOH+=hyw#^vtHVY9(8L0{nYbYmPV`-e5vfUUjVM2dat79hXYrBaO4{6iqvGBY
zWjl<2Vl<=3Al!a4O7#k{Dso4Kt)0zdx+7W7@7&alTT3RlZhX;zJYR4~$hZt`Qr04*
zbXq*B4}63YFS7PC{CsxYLPy$=F*|ASehBRz4=pvb&CsC$8BUaaj$6010!Xn~^81>Y
zU6db}h7d+Nq2V68gCnh{zg@6Yx9W%Eq`{>U!doz8k@rK&`JoMqQ*?gAl^ES~8PTyt
zA>bj!!Xo$L#+aIdgb0-SVi(TcMDO?G^`9$Z+yo_b1UF2C1R2&Lfg5k+Ihs3@S2D{@
z+boOkEZ{oLB3s6_#CSgXxGT2Rzmm$f-MLhsvui~v#-X`g6dMw8Z(qCBy*xQ+%6t3P
z>hG(sS|GVzX?PVs1$M}RY>~5jeMh=C!>v<75eZ%cLd+)4^!q<6Vt%aDrd}RvIfy^i
z4hiy_2HMD>^rQA7lD=iqY(kEty1h_BC8OlWtT0#!j|(^g?We=1g!-TIaddrLE^`Vu
zsyy%0g^W3yOt=3q*G{}!;Q*iKrvl9Xn2DEgyMDzxqwDt*zF(~KF<C1KLEg2H)aK1i
zwAncnA?p&^Vn=Msi5hLDn9+U1cJ~X~ufu=Q#Rp%b`Sm;+Xffwk8$&acDrz&v$bz4g
zTGqLwO$@bPJdZBAd?7rW`uTW63dk*RU`Yr)qS)fgr(@gmjR8+XL|#dJLp92IT-H=O
zn(6SL8R=tviwo2#cRRf@o~l_YlayO^fqh2emRJay4$}Xwfu{Bdz?+{Qq6wsD{HC5Q
zm)Bgsb)1^joD+uT3uPL*EFs1ma3+oTBPr`DqH>AVq*L4}-MkoEr}GG@uzOU<t%G32
zL%ZOBymKb-=n5xkK3bR5Hi?g=%ezvOLT>3F`#OG5bw1R9JT{dAKC-n$SnKAQWkq+B
zy1LS2_M6-%=I_u1n6?{IdNa>*YT6Uo)$aHdhgPzUK9WRBA#PKCz|T}P@*zzP(JZ>G
z?en2c(N-T_IFiiHb0w+W)w%=0Ph+UI=*iN=ZzV>vr&00l9{Jynfw%CiTx;f8jYAbu
zt!C}O_=xe)<XlB%wW?)2D_{a9g|a(5M$_mHoU5@BLHwC54JU?oYUOj1ao>J*aq6V!
zl|s6eywvz0nPEWy!vIzYK@$6$@BIw7ccWY7Jx6j9<ve;jW;Hq8^*7orFqGzcbu{ow
znD>S0k48c-;w*Z}9!mKoh9*r08XU?Z@utMZ)_phf4dQpS3vxF`l3!n%T8qILLbXq9
zf<{7iEM`>|NxlX|L`w`>tIyWl`&7Shq^xW04=<rm^_>f}92;7<H$|dBO6qOnNtDcR
z#gXFWyt-qqH)SIXV+>$)2N~ZWp~gp(8P*`*Lynlb&~x-0XbN#IOO1>m?omdRR@Tn$
zKW4BPBI77ra`K7yF>K?lWeFTDJ9E(I@yd>rUijI@rAGj?(IY@d7qg&litLcGA7Md7
z-+Jrq-LTnpp=6n{h4p(EzQs~0Rd;f`!_=sV`eYF}Hvk21aeV(|bH@>7h%j$`I`Wm0
z(h;$r#d8BX-g9!(F$k`kj6a`Y6r*Qm=1dDsPZGbkr%>ofE+1SJ)#BFsUgIKit()n;
zSdT7gcdIFb60blF6cDJMh7px0VF+@31rVy<19H!%Og02*Vm2e?;|dO5RgJVqcJy7)
z6sd&<-;q?qt6YXQVlN%#D~l8XUG)n9yi00pE30+OPGqL{-K3#~M>IUf7k1Fxq1s9P
z4Ik&~7|#r*aR2<^`0xT+dfA1$^xEwCnNFlpv!ENt*Pgpt=ctHs|NM#R|M3%hDUu{x
zT-Z@ESs_dP;icj7+8hvP-<rt1{8M2xhfk{Djb~0@Nrn^<l@fZ<XF<^Mh>t|&@(0qf
zx`usjbG4R;ZM9}Y48DsL`JU@>IVw~XBW)kNp8n=Il5CzP6r%ugz2W;3O|TX}j<S4F
ztf0r`EFyHACgx*Dg7f*_s&r2I`=PlG6A8`y*I^tS-KSQq>-`Yxg7_HACIpl1pfoI2
z<gjS4%sd^sua+BYc7^)vVV$SMPl$gWD}jkD&DS=4z=7Vn0gj!Rk@0B;WepEp55zHb
ze%nC9JK%{>l0bLw7(ZSa9|>6z`JIT5BPUI>p+_|li-;WW{UPGYxC<g$RYdpYxSY#k
zJk3UNm=P(239rU|GO!+iAHL9Q6uxDDlZp@Kj^nY3t~@!Ne)2K)#?~K}eg=EO1PfG0
zWr(K%di-B#&ljM_|3_#K;KxQN8o<$MGJEUaaJ2p3F(4~mKSq^~u<*<S?S@@WP47K<
zzoN9-JVWQYHuVK1(^#{aAR*{JQDh&Zgy)Ko2IUdQ_`bBHBI96ps=xbE>H<z9^75cv
zZAlaK7e-`9Hkj^jqlSN20{%Pxuj+i1r3$e|#+hWcWzcO+-2LJ@;=}Uw%<W0Jj-O4c
zS@BBGGnjhx2yvY&+Q7qA*HnJH>QbX-0mvKblZ-P!FqExj1|BpoE%H@;KJakod_>PE
z03{6I#1kp3!f{|h8w|q<+<hS2N(7L2`|>IuXq>YB1cV$w8WpMo7XHCG?>I6HL|?hY
zKae9j#ZC$|^@t0Q<qGgwLnQER3FWZ`^7Eg3ghFLBSnaVTK%unIsJZli=PKS0%xt!a
z6uq8_Yw>?q7w#z8IbU#4b_TL{so^6iilxtQ{W3pPQGKDB$kX;nw59!BRdl@1C$;a~
z43&Vk?NS~3M*~un0r7Aa5shuQvg)6ll5M?mUaq^WUF;nPjch3Y!S$=i1OLIsvZ|I%
z_T_2T@^n*^-i`7@dL~J8yV{%V3Ni5~cN89UX36X(9m{n&$d055e7ZQ!q#|!A6&!)*
zrMMh%kyCP$RYelswu$=zCYft{#JG-(3#6R)OE&z2BL*dnI!Q0$HNzwml%KtN7q?-X
zMsS2Ckh%zo9sMX>;_0EVXTAmIQ&YNr47-kYH!u?SnQ|Crp8xVl2=Qkc!_K-zZ8t{x
za1NK{SUIHzWpe)%oGiR~0hu9V%+r00-+!k;jc}Wtd_1G5n<Gz;_)M%7J6bk$wX-9q
zWWQ>Pd`bD6-Q^vsI#fd05mR0^Tryj-xiDyIP}-oQwRz!|W29y`-7Dy)U)+nukU%#L
zls?fEpAsiZ;U|T(Y7dR*my|T5c3*y26%#f(ndosXcMdO{W~zTE2MQ&_BZ34v-y!1+
z^)$OzPO)=<(cYY^G|uAj=UyaaF{zUB0rWgg<ZAP>K>8dU46jgJ3i2zWyFi>`<+TF6
zA9vTK<|sLc3t4uP6m_lSljDe@Yf^nqhsE-3HisrESkQ}1{WeoMA{<<ht8xH#Y}ul>
zq@=7qb(sBORYb7auPg<7p~UEnfD!vcDU>DAHxh-u7KPKh<t$rfQTM^-oMuEK4;|Hr
z@}dgtHss(#B*~?7npYXk7hO>xJ?XM0>u%jeeOJTn?9~qv2H&p;rP=+Ssz1AXpHzbD
zIZqni(;&)b)x0?UXhfp+#>>7^$><Zo?!LY~>T0|yc=!umrAnen(YYY*hT6DwNkx6<
zFzq>$hw4V#g1?^13$m`lOC10=NOI}$|5SoqHd-DIYc4)x^{F#qTzh)nW`?#~|I2E(
zYT>(Js{Tllunl;i1*}|5F%~YmwomNuC-+X=`pyo6ocN}5kZ4nTC<e+4^rP$UeQ<0L
zuG1y;f1T2hx=~WYijkh!g3ziy^5ei4Ly)S41c%P?pnNX}pwpY&8Ntj`@0Pp~i1)F7
zGPAQ5sHSXM6V!!ZB<szlqXA5~`Cj+^gwA6zTis`sI$ZuG7yQ?${L?&<-HfVfP?|8;
z567ww`g=R<HjDZ-A(GIpvjT|#TK`0G6a@96nLlCkGkAR0613-&X^S6RVW%vKzL*I}
z1mT1f%9tn@1)sfGNfhgMrs){9ZWXyz&{}ZW2SSTiks1M-)UCal#kh5G;VNL-m?xnm
z{N~)nK7vOQX!{fs$b2Y>WNNcH%PY`8JfD6kBPHms>+SLrg6DqeOf3Iq462wNX)K4e
zMQki2i=W{_&Ut=k@cEn+eRfatZW@!P%QW0xKSEWA*tvU@8L`p7^YMjYyF!cs;qaN^
zIjUz>DpEfoflzZ^@qzcl$|4_^lYesd1lX&TKe}-v3^}GQbiyNX*#e*XooMpz03Ny%
zb4Lg`Pj$N_nI3*Lpoav00aDT0xEJs<{2AK0gp3c??1fhtcZB?}e--@2ZQ4IwD*g=6
z0<_B+f^|fmL^B>uZ5G;#I4g~XCy#8J_JbKceKH`;s!$K&A?~jf79GBqKd!EsX(q?J
zs@-vyd@=KME_HYM8G;R^`-ueW^gSTQudp@nmdky<O?)A_ppa?%qNz!OQ4O}BI!Lvh
z{s8$~_scMu8F%{X*Q}a`feo>T*W7ETWbbvYNot_v<T`sI^hIh&Nq7bw+RD!l0>g~t
zvA!f8qHP-}+Gzn)hqM1${M0llXlUbpdJe7q(O@iV%Z$4bH&t^M)7v?c4t_f!o}_Og
znh$#{6xS4cI#BQM5>m>Fkn*C!fykO@_C4?^_8be5|DJWV{d(d+Bgb+y87~gLR$Ne$
zl+??jeXsmEbQF~YVrbeRQDmGcNe+M7%M3HMXWZL2#gW0E;r7+cjs96PK3G1B>D?-Z
zRF-W6%nLdibCw@U1^7PhV);68Qx3+lo)##AO9Z{GEzd*S`cF@W=JJs>0yPe=lNK$A
zBO&5Xl(iIrpXL-7TA5!n!Kdg%v%A!9+hE1Ju&?>E;&GkiZx+o5mZ=AeyXJ(9vRS6^
zvlGa$-m%GJg#WpEzmHBc`1+dX3w&=}UI42n9~(-?7yJ?DLiPJxeLp9b*Dtb?7`h8F
zFsmm$vuZqn?wQ<mas|hoJR9FFedn6)PN>sfFIR{)jxSKny#Q6Vi;h>o#sEKnAe8PR
z<*FZ=zyhMnn318(GSgTYG%rdy88DEixvPe&u#X<yqO-rJ{Y#3fBgj*}2^XxY3_pbm
zLNEM%1doiu%iu36&@hyguJ{?tD3^)1h1c3nmJ)NW2J`hb7<x<y$}2#J0V7+txEn$k
z&cpSD`d6mByLGAIk^)9_rmLitm6rrNQP25>DR%1WpfU&H`@Lal&nPrU?B(<k_Zu5*
zK1|fXo>GAS2Tbxmj{7Sv$H)CT-`Hxaf65CAtCQlRD%z04r-{nuU1}R<G~4omf)%I;
zK;*6B5J#s$R{UZk&~9M;tHI_ebo3^RcppQB)s-8K=}f-PxRvo}>g{K#0gm-?(Cip{
z`RG0BN)fVg_QkOO$;NT4>nGcqEvnY9k?hy?O<Emz2p);SL52Wj>=4B%VG(G;c_xaS
z&3-X(#NmdQX(r%CZN^o64!rDs@86^kPf~6->c9!Y-b?E96hIfINyZJT|H3yu*s2B^
zq}yFWTuVJnbJ$RNkUv%{iMfheT;%Jy%)8vsSV_iH+#{WQga*oS4xmSzasp%u+f2tY
zPuA%N;Kjo%F4cbcs2kvk5pB5Jg2B`UluRf|k$PcnHt<73q?Q>myw%LL4*vMb#@Z&n
z*6|#9<C)NnGrkJ(-2TvQ)n*`v@kF=e{RcqeJS>~#eI76y7tTqO@Zlrq)%Ezio#rpr
z2%+dbL3w$+3PAaR=OCt0LA{wYXK_LjD>b4CmdMyCi8E3XW}gJ9(~E=MTZK=@`tx4b
zDhH%%x+wyt%0FS%1tl37`9!aaoqMPO*ECcoUlk`AqtC#X42J$dHk+}19{{vVp|J9`
zF#+6$J|r9~9khs00MZZ?GUpNJF>XW88j#hF@0}?q+N6p5HxJ;X1sZfx2P+E}{a&SF
zJ&*$2<`4n8i>@IoU||u;F7uzVpG*{s3dGAt)VL*tQy)OIH}oHD7yo-Z)%~Lr9;)lm
z1a*g~^`g!o&vA-t5MH)E_0_sdNSgs8rS!i4j(gT6BoyOMjN|HUP$b_d8(J4mX=9n}
zz+4Fkcx4i+K4}Tz4y2W)+B-V7(PZ74Ue_nW+ilp0#bGhCKlREFAJ|o0>*JqV>t5|r
zn4xCxEW}C?_&R#{&@B${I{mP^ey(D_+`FiWg<7Wf^W`CX_>6L@z0iLQhFBLTTDDVo
zNwEjw;X5V+AB$b__xUv@1r@_CMnTh_vbTF8=o+Mv8+e>t_~QyOPs>B4iKWN$Avl?<
zCR)Zqu8<HpJFh`sP^IhMcU;wn;1B9UpC{XAlMM#TRp?Dz6<eAw7@IMB&f|3GCGLBA
zlGvM?c<1oI*SZTIwHe`mKK;Y0r}KH!5qcL<IojL@aZ~J&Fwj&iu;ib5cNpMmIVnuY
zsY9GkZGn@8P)%sk_QJxI;@7O>of~NhVLa{5LEVj~>g?a4jQe-$y|feWeEw{np_LwH
zRj1E>52~Qx9^lT@-_Z6hSoH#_p5W3c2WL@M4}8B=x&pozW8ErKcB0nqVH5XE7)LYT
z4=>B<vXmA#_NB~o(noGq{Rl8ztl_$Q69;|U(g~`xw9mrw?7UxKYjn3EBrjZJGnPjm
zAXC%Spa`*}^quXI_?{;##<>`&U{7@hsSt@lkA$Dz@G{euMm8iG_m|>SeV%H%y#sad
zn_o{7(r*W6IjDGgMb1;MlAVZye9R~fhxe;)5A1CNWzw!sbY$P?SPM>X!gK-&ga3UG
zYtyxn;~5hmC7h4IuH+k()`n;bAV;1)jx=MvN+Z>~C{5K#zH{7*y@4MYw$Q4cdeFUb
z;8NgFWq;LB>l#exj2b`ls&(gDGoiM<ftwV1sSJ1e%J!<fn2q@K=Se}hT9$)+r$VPQ
zPF0xf050fE-A$K#m(nrwafF@=!*zjBdahTn_&awb@A!{+97>>!h;D!w>&ihsH+H?z
z?PB`M?`v(2T?bt}%<4BNQ^`c`&0youv(m#E1L_&OQ!drQu>n2|v7g$*B-Oi_l)bS!
z0JyW0u&ulb^~0!}?`1O`EA^hdd&D*Q<`p^1A^6;){EA37@Q0}F{Ek<7!X@T!DdOmM
zF-)F6=ET(;R`##LS&JDJ%Y3ChV4<8x=6AA<!(Uct4!T-&eLD9el2caGU@HmwyfJMt
zD~qYH;5gF-(nPSP&6nqECbzFIowd;{lbwk21&>-<NN&8x4rJtW(6>y6I-^+xn{(>O
z%kAI!6Eaceo!p-oFiJXHb5xnHXuEX;F%lUbbE^ELc_PiCWqk(I8C{37-zNC0Voc-*
zxgk{46uVSFK#dvUKrCU7`)R5{yp(;Eb7Hh&&huNA4wq=c`(90&e(uDCK?d7Y$U#s-
zm~oszB>8BNqKvxLFnH~nQ5(y+&?hs?iH;EY&OPHZcVN|*>T7omNhvrq+iFdBBF_5e
z;f>41p@W9n!P?>eg&gercl?<L10A6)J50}}&#Z0cS0Wg#tX`&NrP3J2#jfbU`0k2K
z{DC}T3Y!upd?l)kycC6(wXYYHhbg4FJ(}BS>iPb4q2bM@t{PuPN+Wg9g^B-*w)YHb
zDs0<DgQ%#KNC&A>1f@wY0s)mK0s=~C5*4LN7YH>dg4EDN6e%JiC_?BpGzm?lNEHwW
z2~Bz^AwtM^9^P;7GyBY$^Y45=7{?h|?OCqpzV7=f_xsLljgEX*m0G;nZN3%!Wk7g*
z5LJNyEfiJ?j0W5-*{u2KkQd`5KRtMSO_l98FC?_R#SwAyo<J~ZcO81CDM3oB8k;8R
z`SNS3;(7<f;TwkQZ^f!M=h(B}0KH7|5BJBeXV38xubk~PRdC1mV$d*F=lVD11oyWu
z(RRIPweO{--lu2MfsSvWLx|ly=L~6jtkE|!u|LX9cD>-;)t|S0Yr3WqN*@WnN*lNh
zK`IZl$;Bl1iJ`XQXVmqj>d$eiZLkD!ZlEAEPGqhj;A|gixlPJwZ{k$+e+*roUDRj2
z-8k!_X7NNo0jc?H;MJc{Z%Vutt^yRbW`TlanF^(ZEAzYbUMHFc@Twgs_EjOMW>`qc
z083ur+^T4ZyY#QQthb9ENmUVdd0yQXG}F3DH>XHtr^NDOgs39>CDm~o7BBe&88qYe
zbD!rjV$JfYS4cD2#Zv|k!!VOg)!*1*KMQ}PsA?-#u@Oh`?C5t}vCXQXdT-z;$}KN;
zKlW<2s|;m)!SJD$ii{+H2QDmYj5vrla#y<U=!;}0ODB}_B;&k&YSD^&px%3ufiJ`v
zE<6Uk8ry#{Voz?J4_X~-+Qe^uA7FO{V{WW-{>q#3-z&D08dJK8)ZUj@7>UA<cQFOP
z?xTO7B1K}g$52Jafvl0w2iUIVPB^y(17#U=7<WubRKe)_T=oN*483#3pddCK3f{#y
zm%<63n*u=@5=^W7{14?(S>r0AAiKbvOefl63t1^nq;$<qZE{kT#;fj&-SpbXLx?~2
z_}XzF;TghSdtBY({Sd6E?`y}L?i=^_kG7AMV03u7xMuPCqbyA$Li|*Wr+@a9xsr!<
zsd89odh`P|Y0csF)^5#t`lT_74Cy@T=a(WBF_(HT%>7WFqH4Wm*8c7(R+mjTC@@SM
z|59Z!5tQVr9<vwKc=70e^z)mP`?;*JW%p)alO%fF_MXuc1vqAy*5@EXGWj{Svl@K=
zcX`|#;jk?eA`k&iM?^<1*hlv3Xutm0qS-PR5UK}io3A@G3TSV|VoFg?Ej(^B%9N)W
z2h<*|^`?<G(N9c1>)iu%gTz@C1^L0io)6Hw7){FIT()t?ETg}N)XhZk7DxW_&c4J$
zetGi^9YPZucAvNur!2%kl^v}P;~cxTE8oGN&{*T;T@bMcA?XP5fLAh|5)b-cpjW>L
zZZ~9o_WLZm&_p<tspOkUsR3*%u_$eE3ftTo?A7*nSPL+{=lqZ9r+J_E|JS94)>8@z
z=7(212>$gk?x`kZQNM-l)7$S`|F+O$ZUcT)E&v+Fp*O+rSl$5?8;}}*K!2ZNM*#26
zlOw+GTT|B$))N-V#t$XLc*Q;9g1JxVcA1_zNy?=IRp11Pn-*QX*#iSaYGb@}tTL24
zRC;hP`^m4(B^wI^y-B{u7@kL3hIh=GAfa{uAy^V5j?UH2W}*_qinngG!)fO!c6lzD
zb+KOSWAJoGkG9*X*yn!e4{8iGx^1GF8}2<Vu1UNrv@;Zcfu&cAhH{>p76biK_4yX5
z>X(13cAqxXWS>0=2_EJA2il5^7=qjpGki7`5dO(aZ3Ao$l*YN7s8Y9P^X7(F>@eH9
z)sukR#F~J4at{%BeO3H#-PwQtxrLZ)AwXj^**{al$J^n?PcD7hdFRPVZ}<@=cdi~!
zzsOI@VKDWj+q?^3`Ud0I7QLbAE5tsn-N{@|OelV^E9!OYaQM5%>xO|SrTZ2O=FUwO
z(F<c&L2ID>yn<1uSqa4bzugf+7Y?ndH%Sta8@mFq_Q1t9+5FU?CyBZ#amjhUzA8I8
z*)GRoogJbX<u8e)pL})-2j^yFe|7a|oP`c_`YFUaECw@Z$b)5Vbo7&y5V{2kJ1+13
zhkt~WXd_?=sg^`J_P@5{(bB0j$+eEGvTt8gFTw+-cg~-=NNVfXLvF+#!l*YIM%RVw
z2m4+_h1u)}H1$Ls=A#~&4W8RWP4*ICahi&R5{a8YAb07#+t~GOgW7+OQzi{=bJ3R1
zw(`tpEhQkEBT4yP2jP{@&q9qt0?L1UYJ0Y5_=$JHCgg?I)vz;~z(336xHx`7;_DE+
zKtiW|q@eC*m{uL0n7L+0#fVm=Ug9~R<7wTKP$vrN&A}3Kw5mxe8PcG)(wJZ`g5{QJ
zR_Dk%-#&cu5S@QILhBVdk8+PNz>OU0wcQ%aQD#Jy%T#{66+^0MtT81hoC%-!>)?S$
zJV%`8?zB~Eh>m&caj^mI>Uk+g1cbOIox`R2JnynZJp1*Zbr6ya2mUzJhXh0n2G|Rm
zTu96wVg0hbc--c^%*2mw(pRLw#kXqls3aLsqy__T&R;E+Yv1JZuy0%o6B5fC>hAz#
zN2EvZwH%rOI1vy}y1eVbZe2J0)I)~&#Llv;Q6pM<seYN^0mJQ7GtsRK!aiDzk~Y_9
zcND%s>nGFrj6O7YfLPF%5&|*P$zeJx3TGImpTWCORka;*+<%2Q<er&uTj!b#uZhIy
zzbSm^ODe7ec3F1xWe}gNv_7g3rHL3wNuOIaUJqUc<ddejeizx_-BDuRpOW->yS4MP
z5?;`a(*|c5G@XIh>4i6qb4T+2e2lwMue%gP^)18_iWx*3oD&i*bWv2djJq_va3$;(
zV^t}I*1t__6^x}xWe*BBiuA?O{=p}eUPG$jov13a4k4buD9*k65{6?NRjZ7296kna
zQ}mvR_Lad|<y6SmT)Ji{%FrPpC%(|FVSOymF-%lfTGgb<`)(E{{{({)9k~h6J-P<s
zW)#aPUrJmXtQ!3be@`gra;TJ)Il4T}wGvN)ncd1IBPj($8B0P<N`c$K!dTwJuthy-
zxCxd?zn?dZ<*whd2_!7{|3)Oa9JEuE?M;GEj9gW=up5`^c%_4$l4WQ9Q}jSp^aEeT
zNKmKGZbfzq@=_AH>Iquv&sMKli`@}>Cs5#Q@j=oe*jYji^UdW0^c78<%0${5PonAu
z6)d-SRbZPkS1uFkVs&l!>rN`3IMWXmD6Ro?YIbDf_hwj{E$LF0(yF0{gE-r5^iU;n
z<E6xB^?nU>^%8;un2r-*F*BmI6(r{M^pX<lSXtZCLskn_NwvL;z|7YqP^e-F*TUsK
z%PRJ3Z`vX3)(^t3d6D@|mkIS#(5LfD<69gc`fDE#$DP%5E$J&<&h@Oy*)Ss1MI*mg
zG;;AP(x+qzYQ*iLCIaWo;x$Ey?$ge=2JhkZs5Cnf&7w_F#mn92%;=rfyUZQTNeo0N
zdaCn)V26w&_n*Olie2XYW0#^19r|VUJ#~!n(d%Y0WFgK$qXkH}i_N~+G{xJ?O`0wy
z{u*H}EW6lmQ1$^@>;md_iYS6|io`Z+jg6bSe)yD8>S<l{qHrqH{CVEzLNOQzfFC$1
zg>5sJN0nRHCobi(Pbj~>Q}uOQ-BmPf_OdWtC4biZcntCt(JTqGS3DI(5?J`yBgOe*
zqDGHr;9UM7)WYnGL?#(Fp$fi4p(cbQ_)XI3!b|xs(zbho#Klxq8`0ZRsAsV(x&y+F
z{PXlb{LO&)hamGN-^ar%FaEcq?-w8DpSt1#Z%;qa(Z<88Xg@&cmZllLFVK1vYvS}q
zO1Z2$j{Qnw`JZog3u1KJ#L^(l$<b!zP@NI~YwrgrMsgk}`|WkzX!hojzwU+54QJ(A
zn1sAqdq;xeW+cX|OC)d_Sq+KUfbT1+b^ksk)_zUqJNzX0O7nTb<9YhOlgA^%e2k5D
zLEegrd`0n}DWh|h&rZFLFk6Bbffn;&ZUl<$Vf`DiE#mE>g_|jV+N)dD{wUz_d-Lb^
z<tFy$;IW}v_+0)8%8N5qha~~OZFSkZ`9F)k3sm`&;AB3S)`rPHKF(@VL@wLI9K-Q;
zFG>9?V(3MIlE{%KZEXkey3n%}Bf@>?+-8!r@^aqrrly|L(-J};%Oer>jF9y1HU;#{
z9Of`l>t?l4=RqttW`*p|_vg3kKDW8_hu3oaPg}2^<bTYxElEEY6iy#%gof+ReX+kR
ztYP0Cy)UE{cJWX;RBIf|nCEVRd^0U0t9&+Jerl6txjdTp(I<4E51_c%y(fTf+u2J&
zb{&}?oW;=-0q8IY9{Lbr8s3wmK8}4jcjO6WovGgWyJIjvAbIqc?X<W}V+1^5TS4C^
z0;08YOhm2sinBd69v3dS+8~?rs|2vt(4rT=Z*&W@Qyw*B&eNn;Yjb!hewft@Hi=#y
zGB$Ir+}>8*yj`Kr+OO$Ia#TsLSk$gm9|)qmgChBFI-I5$6O7F!hH*I@Eyf`p?)U25
zi}HdL*ab~l-#$JOnm;U>f&dKV?=zH0Q2ct90E5ZZWwTfR1>w*%gRPAc5UkS<;+@q6
zf2s?bMt*I06!x@VUH;kZ?kLJHtGgsDcp;pgdP`{MKWDrO&bTqdPPsv_^O|VHX|p%l
zxhVHpEhF^kjgsHnvWYtJduki~P1#`y9`9cw?BnQjDaHO?herr|TJOx&QL6LSq%r2a
zsjQdu=BG;blvgPy^zkjl;adp_-VC9DJ2OC)ME#$)x0ky%hwrm%on;<29UoG|l*;C;
z=js1>#Psz<Ilt$>piTqjmFqyWG|;R`u`?%Sv@^pEep#tD`|+3O*GwPss|*G|Cemxd
zKa6M51|7sluo7DstC3x!m)xyg=WfBp>FX?dH(z+W(9c6DpcG|8EyLe~6sQN_zuxMM
zY&bh5nf7y#V(u8R`bBk$x#caa8PPjE9KfPM)LLCKqu8~8mZj%t%O(j^1#=!?cMF%H
zzzbF~=Zm#St`~!@<LeBroZWAo2YQd>mbvTHTcG51LQv0wryq~0pq)WJ0t8Y}Lsn_A
z!+*=+Z3OS#NRBpgFsa|rc>t|bo(2V0kiohPKGCiV#X>CKIxvvgD_C(jV^j0)p2f{8
zpGdh`f`<GT_KwA2OEeBM9Oi&?2@J~L!hlTN62Bjf(G=R6x-nR-5{vN=@lwM1=DnNP
zf`)o_h80&}gpZ1_h+>NB99GpqsQyr9dYT%LsF9R?{uBCEPM7@i-0L9heRUVU1f$Aq
z#t7}IxXtb#G{2y{C&6$)7g{;~a&PJyu4=S8h)e4ITXuKg5NKJEddFF?5NOd0iZK1p
zNqz)~a1z%FHMrj-jGpw0;1#<DWDSZ|!x*bxi?MX((T;}ohcEvGpaLZ@3$6eCTQ4b%
zBs!j}B+rs$zn{3I?(tb|$f&3CEIl&Bs(FIkxUS9kF0?#qhE&YY(tasIwfH;Ct>qOc
zXVB03P7vEw{swAi?wpmOwxYkA$uSj_sZI2o1@9j}A~H0hh0Q5PZ6JC+;o2$L{j2Nt
zIfc+H_;fS-@Z(>hGWo8S8a*<JuT}d0{)!W}FsS7$MrhN<1vN+vFO>lEXA<dqv<9mZ
z%U;Zs6+0TwsMcMBzV%yuzL^aD0B*M!m4G_`pdrqOP+~Hv%ED!&D*9mQl-bFU3;VN^
z2!cUul^F|Cayf9xB+PcAbmMW&j~aaW+}%_|#EhWUD<!(!j?nc(E2;`fE>X6M4}9Jy
zWcDX448@}!-}V1<<Mh869-@HXaNNJi0iqm^&#v29@bbu$?JQV^()(X!e$fJ5)!{E)
z{hDUpx%x=n*~X<_b$BN%C`pfRg5eo(A;rHhX0$Jksq;&j$^5i)Jb6TnmZ1NA;c$cV
z^0|vEUr%W_KRGH1zleO_eqk~G6Eegi&9wH-@9A4$rKQnR?{s921{rk!>|`+=X9$EN
z{~|7g3%fQXR{ti+><vciEHY-uaYUdP<-Ef`9IzttM*hcL*>(cgZ*0Bq5L5vrb_vp9
zOll83Xw&PpBxV|Q|BtqPgzhZ5o>*9zyo{)9!P#X$8j@NUl+hmjB<L|xn+!Oc{Ug3z
z%M1%+&6}WU>wZ)e&W<eXY`!wIhhd!P3Xgp(z3^%PEs;>%7~NZTQb>`XKOtm1onlE2
zp~TR{hEY||kX47u*$#H^yo6I6@2N|;sXs!tLi|P5_{XRJchf&K9yI24Nv0{Ja{@y8
zV;H3=zL1-*WwJ_G1AKk$zx0OvN-_;I0XmcK199QXNd$>kkLpN*g&`h3Va!!!_y1`A
zYA~GRyV6z4qPyloKLT@wMuy7)YBAS%sG=iRzPtYkQIS6e?BBaj;QuAmGZcwx?VuHV
zs4^5r;*?N~gV2a__9ZQi@88LMJx-}Byb7Y;hu8Dd(h{C^xQ34)Sg|$*p|`?>497N=
zdKrz86Kp@MUV<9m$+u8Cm@Wsx&tjyB`VIwGhKLR4t8E|&7NFciP^Yp69l(`U7>nf1
zC7q;CHn^?@>im!4gU78OkoFSG9lRg??a_#bTz>qOT2|a(h5nc59bUie`jgJH0{=Fn
zgy=#83#s6<B;a>D-lVLoF>c!^XN5N~$dm8U*K7P)j#+dEye+ORu@17s&6gdPCrxXE
zJ_En}o8?_kq^{dz-PZbbd--bs451eTsDTHR)w(3r_ObCCkr97J<Eid%i?vO4K`E&Q
z=M0&U53^oeLlcD~!e&7j&Y>}gm-oyiIa>bO%?>y;8Q!JL8Rl2K%bRW%OL_Mr_w7PA
z84|L@jYoArFTgqkPW12KG@|pI!@O(K+Cr|~j^f|eacpNfBbp;Wn5?4#DBiQDLfr{+
zv6DHqbA8Ine0+O}ZFX-z%BLsZDQ0*wFx{^oJZkm=aZ=OgW2n)PGU~i*%HJ9;7H8*>
zw*|7^I(n>?{5x4$O!Nb~h)3o+LV{E;%0*&WAi)ar4H;q*Z(i2?<lTMGM=^Pm9A8*k
zmlpaB_>+Qj@e*xje8>=j6L|CEO_EH@%Xh@2$K`K57Uw<joKF>|;_Ize>PL>GWDpq<
zaR+0$pH~l}v!6WZJ$ZG)bgNV)y7v@Jk*zCjP2*hop*r;z$$!5iQS|o6-ueymiWasT
z((tsZACSZ>g$<b069ds27y5NKsW}W9#ZG-9)FBhx--Xja=Q=y8S4T5mp5!=F_~l&C
zBO|e#kY^BPuZd!ULA)1D+BH+Zq;z)0HvB`HSozG)VUG#+?r-mte)vX&ya+3!Bmbv?
zz`5nsu^H!}>=m0MR$EQtDtPGr1f`sH-o{>2{;Qs-u$pBr|Kqfg+}i;6^|pXF_6yY{
zUVluvaoY4d`7C3>T8X^Ap7ggU^xZuM{_{b+kO(GB>@1yTSP+RG(fn)o?H(WF)PyG?
z=UR~?!9cF6xK20$PeQAe_LAKwFi_8(MQ-2s!jW13jtx#}8YXJ`R{f;q*r`Cke8N0o
zCrOr4LNe>bkl1nNcE~i#irSi%%rQyn4MCn!QQ!FOFRRacHr(rv(!rOr7*0u)O)7DR
zH>4-_<#+c6N!>C}OTJV7V^5eEIs2iNL6rysL(V&qa;mj6^1Dz^d;)<bh(X}F)9^>L
z{(r85!MK2dk3;OCJXMV39Eo)fD?xtj)Y7JN8vE7hjJ8V2em(!x3BufTq!*@jEdE$(
zNd)e&i#r{`RbL2UR-@rVD`mF!Sb4y`kGf|W0ipl1Czfgwytx)n++=sBDUMcoguoj{
z*1+C9Nn-w!DAl<t=DJ<34vA$kDEkXYuA4Za%37RVnA^4#v!KlN5yRVXa|fMxksRI0
z4W5t}8V7qAyQ2)?hLzu@Ejg+@z%kkJpu(p5wH3|x)@J>$-UXhB(?1>6)yYeLZ-I@c
z&yM_zlIZ%yerm#ngqYG_aBM(tpR|Ejo(g2rehIG(@0~ptYK9MUhL2(`JvftcJ!Kx%
zvWu0@DwO-)NvBWnA8=sc{WoPzsVDYX-fd_?ak$lV2FlIb-Z$bfNB@Xpfe2`Y8EK9;
zhIcn>lH_9ir@aiu_w2Gw$F4mL<hx&=YB=lrg>K-Qp%#liaNqK!+dIX<l72c*(azew
zKY}~%k~j2rgqaeZ7U#d-qEI+#dVQ~Bm{MsI`#I&4d|LpivtIpsk=mVWbO=3Kx3($?
z8v81y3ixolS+2-Wo^gsc&QLHc*gwq|+an}Ntktff+x^cg{TSJ06o&eKVKIrRn5iKu
zfgCT&^io0p$5}FONidGo|3&*19&_Gp?2$?XFj$iP$j0>LS>iiiI$mCS(6WTV*7iv9
zm!2V5zJ2|slOk7AQ?|#O%rE54942rm7pguUO8;RMN^tpD(>UUk_-6F6%>`whcIh9t
z<~RE3H2s&JZ$AUIJvK;bK#J0GpJ~ndWz#eJv>cA}5HdaI9`{-z(Il|L0l2<|-ayHI
zm<;{(Pq9DHxW;6@7-j{*@XkB)wn88w3&)K;KOuH3%!h!1xMzyLPU-LqdFq(>3D9s)
zIt$(Ud(3?)b)dM<wZteVUIy0*pi1DKHalTwhRB4a9Pl(qUR1>;+~xrnYC9azdeLw!
zg?mmY<|)j%<vh~ZLS?L>zr=goymN@bI%c~jzheIJzx(<=jg!rwDvrt}Vh;F~{|iEe
zrcvEj!H~<Y!?r6v{6e1(fw8u!p|U0Qmqgk-fs;OUGpLxw>x#>#Afdnjf$5N><i9lr
zTc;hLaeNe38Tfq9j4o6=&2$|5sigrM(iN@3JRf~iCn{PO);PfLH`8?Y7qAroJGMQv
z%k1ndWq2;9IN$X&tQM9bZMa1DL(QXp`D~{Gvh?6T&LzD21g|4T20#wxYN7fzV=1h2
zGB?^_OuK!2*4rHiyqD5NJ-#cf*?AhAOu(mf1Rq<c>VYa=nNVW)9mcrWIOu6ou+aB2
zAlhF@)1@t*L+kxQXK+I5Rc{@J;$MjOxm%S~FoC%d4HzoVw*d(!Y+y?Y7n~HY&EayP
zzgncX$TU@~;>!ArBcMqcd^vM=JrxQ1%2>?hqXJFlV#@y9nl0+pkh<5pmj&rz;Mv89
z)Qfj;5zv7gLw<NM*kT~4c5SLDKqAin1(8@sQ5Hf5a9%I9MMlEuDbDR0HZ_f7O3pTq
zV;`TXNaaz;S7G4WNlPGZVe$YzkLX%EB&j7sqa*fJ$$wDuxAwi_VP~*<#Z=y{oagyr
zIs5kpM^7eydm*m^cm3V3LsXwnc)r>Z#_Yg*Xp3yL+p3LJ!fhSx*03S&zg@^VeOdC}
z{Z=U!xCC#PdDnGcx}Apbfn0W*QM2gpC%mm^ryhi48D$<uwjSJU7TWTR%C|R6?JNA<
z=C*w)pi<<Nm~)L@c$fj{Mwd@_Y4g?bSc@~$VWLnRdnDq65VYDZJFvA`aMX*;*EkI@
zL!2ykNar{2YfPDt<zq3T)b|!QuV|VuiJUw4k1BsMFz|w3e}$b=F`q)u0=2cCu`8on
zqA8>VxE9LPu8~Lz?r%S4=%CL`f3K=Q*a7}t@A*<w1_wIri19kwFW6bC5di~5UWYO3
zF(36Od3s!W)n^_5qDb|N?`cJ$+e;~ko8$ek8ivGB-MP1+s*ZC<w}__btv=}?J9V`F
z<ub`XEIR)N1SlV%oE0GBN8Bvvq~-Eu4<_{2ou<7;=?6L1QOT*`bjthT?>s$}R_&Y~
z^f4LpiZ%m#mhmSfyRLx1GW)`~y?3^}Tsmp@u`$2Z+b^p<T!Gpiu@99BZr-k}do7u2
zZK2ON$o3)Y&Z)2e#>x4h{S4Gf7xatgLR>J|ct3ZUmaG&F>d&_A`OEyoQa?0Q?chrD
zgQHSv5Wx)Uo++5}VWn2%N!6Rh4TCB_{#5Nn9mdG)5A>}>2|!qYY?sK9{V7R-MN@4-
z`34U?M59s{jk^UNIjr5|FF&9|1jCoKQ%HVY!A^x9!8w;lOK-C=Y<rgGSbyr;YDha)
zV-=rn(U*k$KW8NW1AlM6dxQxMB$zOEH0USht?Xw$$^C6bO3W%@N)j~^3GM&H_Cc8V
zIS9-I(9cQmF57_GC&(PBu~H_B$oXl99wk%7?#E8TFJK41kB0(%oZG1NKq7|SC0t=h
z$%EadY4|DFW5-C%9g5dC!2tz+eqw{@J|9kVxP}K}aR;?>l75C7$*R)#)OV>Hfgxff
zjkFiMyfLE3gMpLuK;3y4-h-&dps~jJ$W5~$r7$3aF{v;<RX6qIT5JE>##twevva2z
zbo+vwBeXA3CP5FBPc=^YebLN@S`K65gmdiZ@5@XHIKi_AAkXUq)b*&K8H%jzps7aM
zZ#8|geaSW@t)REm#eTi#&*p<O*Xp|R2A*yKh#L-}kcYc6IXB8b-zc=6n9=i>-t_0l
z(r+V28AwR*{GDkY2WSa?luu(u)^<2()26rjKG_0?jx(wH5@?C2<Wox_5MBNkT4#j4
z&7`)9@j~de=8MZ0(Dk1?KL0%kn}=82-j_FS?Dd{Ik)QUaQ*qFrtOAU)Q89v7fU>eV
z<<n|I^!ta-UVH`u3BYE=pQ+==dAl0L;Ti?Mu25mhJEAh>-CU8Zrg8BM@F6&FYj$e=
zPg>`%9+CXeZ-xvEks}u%yO>D{d|^D$w6QJW1bM6CEvlQ*TFVk>uIDM7R|CsV3Ve8>
z)kY7=LCp!Zn(<qQUjL;m6Z~C3(QJ6b5gP=2nB)IF<q`QRuP){&0b$D3F`En|$i;mN
zzk+Z%Z~*+E$*`9`XY$MIkgD9V0`cj><vm0#2(dXRDH6f1`8byytM27TFtIuqBa@}n
zKQuL@A}7TBEu$WJ>-(Nr+Ugb-j6J*#JiY)ST^!*~yHFW#UNyod@<RfBDJoSt$#te(
zeInVQ^qOD#mOI`caTgOasZxyjtqY!C==Z<#?r#@l?{38v;pTIez)p=9Dla+?_{b%v
zP@)J^u=wEScb_S*EQh~6y=i`fIeoab&LzFa<E5v%W6yeAA()ftlR{@Xca8ZrS=}<D
zvRA9IgNXY^8k|>5b$q=i%)6J0z#!sT3qNSojN4N(zOD)imc;{QfRvhl9a@xkk|Uqt
z;XrRzLIUgVmu9+9Hd5}Z=1Z;(voUL;!{4e%V(;8z78EK)h3~MUwaLCn;QRyL7!At6
zmPF?P2It{;Z|mr#Ms{@G<pvurf@H#zB*`CZEczM=FV1WxE1{e>Uo^{a&EkrgLvS_b
zu44gbr68d8eV3z>AbF;XXExRB^EJ|5fjC5lbIZ9C1+rkgVre3^HKa?$j_yx_WImm@
zo;Et^mQruNoyY5#kQxNLP4|oT6UC|t4<RAUe(jlrn2&r@KC@`!*jyJxejPh~i%Y`w
z2YCgmRR+oBY(9Zp*(Y!{6}i|x4bI6eaxf0z@A2cX3+<A<Q(UKR;gj<!a!n&?q8C3N
zkCh?rjbqK6xw1W689(Yf8#x5{OLcY%L)66%AyBa}OG6M*fQfB^P=i8<N5T6yhebx7
z9ZvOl60$uX#>etE50&EQodvVtu7}E@{3NfJ*g%zfLcC$sAyOt&lDI=m|LjamOwfU_
z_^lihfpWRD#VLI>F=VO~irt;diahy}{7ct4NR0KPfvCvby@-AWiF#n!V@_I%^GC$8
zx!N{27P(oesw9bu@3r^eOWSxYZRAj>wt9?3ek;M3#&Kq`Gs<<OPON3eg^o`?Grep#
zT!X~qMMOF8uv*i#3R(8dj~9mbI9wxXmlF@yMQz76_pV)j>JV02iCse{eBcDf<C$l{
z2H#;2Ckjmv2`4CpcQqW|vE)DAsY?FS<<InR6usH7!2^sw(I4<BFT#w;4xo#VI*H7`
zJ4QA4;*1GVbY`wg7kB=%ys(|1L*Lt}k#*>|dn{nl0~;?IpwrMBm7}sAYf5$gjD~+=
zO7k0|<*$%EO6OkHX&nWRAA^XqRAYRY9ek8(&ZJ_Zd18p~Uiy;&zV!Iimqq-L(?*)}
zOia{?V(LB8*q1MVW_k!sZqtRni&AzI4`a2`IU^wO23mKsI!U&^)IHOx{+okaK>1|Y
zSH&PH6A|qN2Z2~M;oW`(1^KjO=$EJ)VGNT`ONE)MdigG`IrH-Lh5NZ7IH>+bZrt3y
z$V&UDMc>71A%7nwDg<AVf_w-*cfd&2CShW*x<&2>o{hy0ZdO~k4A$!T5~Ud2y6`o&
z06dFg!nOubd7A)-!c~&s%mIPV;{souXiAE#Xw!46dkoQZSyvfK&zx>rBHIC-pAcPF
zsRJ^Hr$nN8n}D8pY{PTK^IGp^ss1%bb^LhK@l@W*AVK+TRqk2$;Ka|fRW`o|e7Hy{
z^`cI^0-bsgci`1n6^Wr(+Wva;klszja`No0jkAmTJ&nA)oqloD?MCpKjgNo#IjvLh
zB0RmfEa}c0rIT4%SIW+5K_D!no3oVzI28L#s4^v9yYP!~(H|GC)7wAXI~BV5JFY!^
zn^mzWjz35l2K7L1(!>(M-m>R{i~Q}c9XDf95IE#$uD-2pA94u))dC1m?O|SX{{?xq
z=S&i5Ked>D<!ZG78?dct<$5CS8|VzlV)9GX`NxKe4Oh<$5=ZeMUn-VAVmD9c7F&v9
zf(W)gYA!fR)_hn>$!%{~$HXY}B^n%RyI^BW?RzBI7{nBFte+tJeX^dlLV$w^AdkY8
zNmA{@$ja8tnJ8B!Z+nE@+x%3WfMRcbpMIdN-5WQyPwHs$+;7ii#@;CYYatCEt0|ZZ
z_}{n&H1ZRY)^<pWw5WN|XFv3!sUv{!edg3?DBae92@J8FN?Rdq3iA9`O+6=aN?PIU
z(F~H91@Cx8P)+;?w*Fq+OVn>u6zxiLEG2}*javs~qFh9-WhxOqw2?VBx#6>D$nj52
z+V0xhppqAUgU&!^P*{ZNLr|=n!9JPSh0CQV0sc%J=3^)?Nv;EWD1*#Eyf=5PPE^>k
zwwwN;YE+(B_Xl^hbi%1!NH=RAau6Ly9fz@EJyDB3Bp|}Q2?hiNU>eUHYDm&-sVHs4
zm=$3&`=VOQqe=A{4^uh+bTvMg3Y|utoS#pI7P?^iP!ww#csBpXxvODLC^okoNs8r@
zpwjzV3-gvn9!uT68($I&l1Ti5g=w>ykiA=y7#RRb489@}b0$X{ubz=Je2oQKvV~aU
z=A*BmqRI!+2Ii4K0Gb|zSwD+UH129)aUR+=h(El}JG3FWaP4LXJ@fO1%NOe;67Xzv
z-N12%%ODX*{d|C74ztAZU2LEtG{|`b#LIA@QRR1i9n>$<zPG0lkNYEXEu(m~`XI-g
zf2KW>DlKaCc)3tuxt@dyY?TbW2*0qLd5Ze0FVLN@s)Xm?4b8-h=BqC#1EJ0dgvKk!
zB`2u`1aMzyi3;a!(~b<0N$*?C6UZN6<V?BKC7S(B-SRa*+9wm_c}wls(YRp$&kI`~
zh|6WwG!t6;JIDNJ$})BQ-srsHMrlcl|3x}`YJpExYNT2h$`*1EfDI&PQo_Db4!^6g
zQhXdU556z;arMtJ{b+Zc{CWL_#<sqW))utOkpN3<JBjSVWd*PJ<sM}MS{ViL?LGj;
zEYxeoZl)CuGd2d&{a-m0)93NeK0xF1k8;8<DaDn{Q3l!QoSJLDA7*L0uWz1Q9zuq+
z9b5wGw@Ni!cP6nhS9w035)0wC8bw0ag9z(^?b;W9$xvZm>|JNQa&$Gv`W;`tJf{Qa
zFX~z1>6afCjtJS~{8tObVfZ_)80N^6SRX%V)h<wEO1firGDJ{nSEw~OyR#af#Ill6
zU@}V1_VAZCf8Io}ddku+Hh6mdxUYbeTQ}JnzzJ&A4wh2oIpCJRc^y+HK3Pwsle632
zMiHHFFJ<n31<dCw?Z<G!C&-JlGm-h5f+2uHYfR;r)ugZ2i?P%}fRm<at;!p&v4z?o
zdFhby1iYj2BQVA&!vb~)bea-kI#})Ar*d}4IM%k`Dr)+56r>_K0H34byevuox8dI`
z`6I=WV8-v3sg0RKK5TMFGJhNEycav->BPdA{K@ZyVoDc-gplIj77nbxBg!ivM`rv9
z(q0WspxHP+6--FLT_r1T!m7K_p++B91%-x{OdK|ES7#2d`@X+y%N;1)`(ovWN9$pI
zmme6L1eGR%&Fm(2RHkHex9HiS5&v1wPmRD%ZQrA4A!#r?+kNu_l`s^3g|d2kqy!^E
za;UwgGlzr2hx*uwm7?+|FB}PgLJy3P{|A4d*(le6NSX*~Hb$w*t6j*s9$BK7HCCY(
z`<fw;g%^G1TDRzns~Ypf9wniP+OD)!74wxlHotA&!Tak~&&s`d*K_gi0$PJb!6%hL
zbi?)*<lg{zzhTEl(eI7Ns0uP0p^fs+%GgwCJ(b2f@1NefJ7>B)ns&M}C~&eDX@Vq}
zIr*9^K*dM{JAQgj?Dtn+@2|dQJoF>mI*ZNIppf)9Dxk4Y-ggPgl#OdDyxG4Z-SMqV
z^W&xKr9XMN)Py&w@|zC+c+|vH*BxY|LEUIciS*M4fHdW-diml#eddU-zDAu4W6k5A
zKF|wD7JvgHjy^V@r4)A1v^4gkQzy4)7Ec<91X*1IP=+QNebO&w<IJ*|+*U0)lvMhj
zc*JNiE?iw+R)z&Li<5t)RA0}1KV>-@_dI7pBI`kJwJI8GQW$EMVLa^Lr~kckXhxLZ
zDv<fKagmes-4|V}!__{pv)Urab{dyg^ur8XL-lCWSpKPRgVXnWl)U&;uUx%?OTcf1
zT!8F(O)}yG<FNNgEK7*3$tedrgq|Q}=3Z>xW4+q~T})^0=nCS|bZSE9P|mZwe`n7>
z1GWh0q0V-*1~h5jX2{8p@AMf?Gb>@H97xX+<N~Z}DM|C5BgxTx=-GP6owBlWG-osj
z(5BnLp*w0)>K2kDK9o9x3=uCL3EU>Lu4>GlU88$1MMffe<L{yY+KAI`;gP;K5;UH}
z_Vq$74{rm_bcYIpTiHX(Y`j{EkC7wlVs!u9oRFq`BYRb1m!F{7Q!BYvV1pSh>jmfz
z)nKqy9Ox1B?9Fg{;Io~4QGP-GWN04$AVXd#lEO>aWDhkO@!w?W@Cb4aFRy1-!2MGA
zm}kMWv#;B|*X35QWrsINcm|C*u&N}S@p$^}DQ}^;vyrWS8y&>5(=RLMDs-rRbMi(P
z?Q#{RzQncW0<G|gqmxLSq=&2nCj^J~<w(dXpC1*9(1wxhXStN5t__@D$*9RlRbtl#
z*K^V0(-u!Mi_SbS+VD3=Foh3-fA<*1r98z4V|9CD72sLc&Rj73J+4uA`O{<Yv@&za
zKBUdKjb>SwhA?oK%Eqfw!(x^<z_Bi$P<f#9$lEtR1xQwo>sZok2Wn!!GdLF~;B|NK
zQB_OXh^zxiO!3?8Xnkq(^rSmRyu2qLZ26v^1-<#LYaM`Vl;})=vWZuxSHJXqohWW&
z8;G3L3@u!M$cyFbK;}~xVZ6;E$a<P2GWATT-q*>gNrbUGlS+o!IWd<7_=3f0<oup8
zz&#FHB$K6%Y*nKX<!Swc?=<x{E*J^dS3T?v>3>Xz(*-o*Acq!l6e!?oAh=s_z#Whq
zEWEXhKkxd*=&`?ZH___693B>@z16i$0)YMjjWt7&OlNP6RQG|yZOm96Pm0<lqh6O1
zZe<<#Z|X4Mawk?h4_+lHPv6jgF+~1;ux(ejpn8_|#PRTjAK)EIUjVE!sSU+JQS1U_
zdvQB=!$&nPkB)EBD!k<x9>Rb!-%R4Ad?7(Q)`93HrHf$VFvb^lJ;dc>bz>^WP4xAz
zC%({s;NLGjB;O%*#a6>&eO-e;*B1e~JeRNGuNiwK%Hz(Vg4n5%-PM^0#n3?DZ@ozD
z)adI^^KuPw^zj!XnUG&5B(K?>vYuzn3g=uFszkI6uu~*Dxf^C<EoR1G70OlSV#QL1
zHm}pt)gVu)eh#c8C_yg4i8O=lC3rlVW+S_#O1bxlBy!q+jrgj=Pv=l@C`y^aW!B;t
z*X7`2m_5eJ(t^GjX;oAA1t3?fQ?>v79gEOz_}Fqy)0rgE85k}#TBt2K_MmShPP7J_
z%Jb_k{?t8Tx(KP?z{LInGRcP2UsB+%(MQX7dN$U;7<BEnmEIfWQ^ssM?X`e?d(a<8
zB?9i&z=he>jBy1@zD?|>cj(>_a797*D_<bcf)MAmNTL_(s<zVTZkV%m)5ud)<>Z>o
zpAto}crmwgmEb^G=xj;)xX~6TjNs%-;8j;LIdIJN;`u*Ec_KTOl}{zE_>2T-4cL;Q
z6TliT27C7F>V?AC=04AHF2Wce@)KJ*UnIAfHv@#OM{~Dz;Ae0lZuY{8>Ns~AGT#Pm
zG%stxI3begH(_oK(TWPG1($<hb=ST){;ba#jvet}8(24#U2uKJuYh5~C0`A8=AdhJ
zLscShj2_nsj8ewNy^noG>bNCe=nXJeWhJ1y3XVPXw+eBBbPZ$GP_`K1H-FjAln{L`
z3JY}fvOty~{`g%YOlC!hUDFE{q&3nkHO{)|yH+-9bm2tE@iTdd(4}>S%acwwex*d3
z0294F{Yf4R0<8MSu!g)xArmg^!Lxoz`j!8fVDytNy*sL&o3=FG2~DcC?e$n(pRNq%
zoK8p}s(0$Sj2R?&6%K2PmLev~?B3h)F2%-lvy@lTqp8nzgxLSk`es@4>(bdxE8cO&
zu!*mQ*gF^XFfja~Uw4?;x`ywsoNJ*&4x5lTM^rL$#Hv+vMNW8CN}}(wv7KQudY%0N
z&EYp7Bx4_{NBXqScAF+W9@*V7<&<|NA;%Q*6#9YCoPC52y<bX6G9@YJG=2HABL7>f
zu}%DEv=R5uoxl@IVoAvl1hWPi5_^I9fxgjwZqtgCdoe7v<#Vt8`#KOe!++*7E9bt-
zJpFA(1j)8<{*K4)<`lI+r<Lnrn}iSk83aHta`15}ywH5vtu|Pw9mBd3{IvPrdPzb>
z`z5-`HY0}95gM6`pd}H2;6ifS4wft#2W^&2N-iX`eVqo7N&$8hCbfeHK11(;>kOw?
z%;{jF`CI6(;?>*<QQTx#tmdgAV4OKSfD+FpMPDxE$G*QM^`Lxy+sqcSB}8$Si-q_U
zOrr2l5kv<fJ~uLREcXz$N`sDgmU!%iIRrEyh-fZBs}UywzXhhUe${7cOS7xdl(~{w
zir*YO#{a$us&cmN+&_!;lX<S3Mi5e#{0C^ttF<{2|L)K_Y)eIsgz??Nw0S%lAS1^$
zc|$@xSSpLn$ZK>BYWHvJJ(7m1Z|hsLsAip>8u)?OD${%dCNo~Z2j__@xE+LJcn|H;
zDAjDMpWqQpwp)!ixo&K@_;$mO2d%>QzL>v$eg_1?_2-klk3}3hjb>ElB&z$g5;QWn
z53{yZOi@_em{4xgH><(>@tX>&TE{zQ{nO#%KhecYFEQ&rfO28>yR)_Q9m?t!tlVyg
zxS1m6CpV+BW5CC{98li27W4u78a9{JEC&4-Wh-<T7Eg6w*^HPu*Ssav{$5Me>WxIR
zt03WCe0oB&_H7`i;^Ea&%yRxhLD0CM2>?lX0xbp8;(zd9ypRy`Ev=_Ec-XBn+d(R+
zTxJf|#QKg?d;k8FHFY`!X#VyDg7&*mrxD1_xejvDr(QOMml44RX>*Q$ZQv<cOEKXf
z=_OEhJV5r+qG!*e@ISl&%#Cj#?Z+Cdz4ItRxtS`v$}*?T&>s#9z3tg=*l48>M1&49
zWPmP!TByN)K|b6sq@C`dDevvKVa`n>rRGZGKe#D$$t--mcFXZFFGoMQNQ>0MyO}WF
zCbU(C{qMDC*>e8RTKLiiG8gj*u$Q){3%>nP?%R!*_EHR6I;;3GZN)o3XP0hy#N0vo
z(29}^->Pj_lNj`RQeKB00i=Bw(}|N#>ef-7Uwz}dfv-r2Oi<uDH(tw(z}-&Soo1fn
zCfCLW@%v_fv=`00Kl|P1l5d#&tSA9Su@+PSxVRPvqbLB{Ai(6*FcJf<oov5VQ!DsI
zQ==vyS(Bma4eW09b`GvCo4u|V6*zLC{|VMi8;F7iWrzO@LJy{$7xv9!S{VFGEb(5i
zG-V2=?d*ZWu2_LtcUsofCstbWSbR?1dgeda@v^lndyt?srqBObHB4NNCh1Cezc%ye
zgHh+&k!mA0sp2ANhG;}7@=2S;EhZud-J&Sjmy(QH<aE)lEe9f)0?_BzO=}lC`%B4a
z(P*)(3eAv0ckU+$q^@z_>HUA5A~%7iMUsZmg)$dSy47-J98}~r*421jF~Eh?J!!K~
zR!ne;(9XwPIQksU=MFoM)ha-J*BI2=$h}ryn_d(26@TJh5+s2>GT}u~H-ze)dmihQ
zDwV`ViS*4%SWy%jtPYO~)fg$Q&V^9CEJpZ^f`UFz{ke^&tDjL8JGPgAP=3fwv=bD6
z#7%m<cEg=m4-w=D&SLY~J>5L9g-(73WptTTe?Oq%|C%e@<vz&EdmiX0!3z=*Y)9_U
z>2dDlt|t=u0^1u<(bBsz7OSS{f29rq)ybWk3;*o&=Uy=M|Dhk?pxb@I`LD!-#NGL3
zpVdq#Zf}TG=MP0meGy(tm(~@;Js4kZiS}G**_ufY^1fl><H}L0(`l2;-Ze6KIvptl
zz9hv65_MAiIDL)1&k1v~X4_EonsqclOi<C2{`3Azg8vX#X}@!>j_=CcXrBy-Ep~7n
z+tCxp<vvdT82BI{p(5Gm1rqE%eM$op7{aN}JTK<;?Ck8yIvZK$5wC^PAde68#6fsD
zZD{sP=nZ1_jhiD>1&<Afq1m-GyB~y0^x}~(OwY`N)O2*c+P_z3fcX=`_gglP<WRTC
z`0A|kZiVRiV1>%_2IZJgFwS!)$ne>0(ca~|b2s>Zt8rQ}tGM;hIjke-Nt-DBXP<$g
z2sHr-vMXtCtiXdM&&%1-Wt&c>r`|}jU}Ne(0FRbm64r-!pXrR7<#U~kYMR%+JoYJ-
zo8^nvT59L-weFwbslopW{D0;qa2(s0{{Q&{?-r^C{|Bq5pfOe#o<x@fxjG8wq{Wmh
z$iG<IU}US4h!O~s`T=5SmO50L)doMsBSwHu*1V$Edq_&^s?{m$WZ$#v$G)oY&X(%h
zc>hCKtdHv(ji$YV#HK7?F{>S=8-8hZ*>xZ=6T35qNW_{!aY@W{7&p&Kx2w7$#=hH9
z`Q2Yzi`mpq4l*<V!x?(4F%iaJ(=<2n?Q=AN!$>64?DEIQcbbBjpbpct!nYLXLk&v%
zEz(BRS9c3*+uwipeH^YC$Yu5|)OkPG*GNeq_Xqt&JOC4WxzI2QJHd-PRJVZyedSb{
z!}y!HBl3>=IVy!ye%AY<U(U4zpR~xW7P3DnJe@eres5g7tg2$TG#~4PMkr@>w!MJJ
zC!{1Sq0O{d<kt0wa_8Gg?iRx`d-1F8d^~g37i-^rWNhXLknmJ6a_kv+I9YEN>7YVV
zf^mc@v#mC~Ou0YX=C0&tI+Wif7Wb{dI%RNg1u7Dug(7yfCb3g$3CjC1p)inww7<%V
zTul1`A{MMkv)<R}1_XT23Vz|Nw1IG$L7MRLL1&|QRP~$OtM!-EL&tmzUqT#t_trOu
znrk{_Afd&iUZBauT9Ww8$E(gtd&+p>JS}WOKgXtXp4m*<j?4|ux(hoU4nbntbI<QK
zxg%0LXZqU`e#aq{KRtaQtJY=Vw{GtY!cL**J4hk6<&m}{GHYv8;NUtF5PR>oVONx>
zgmjSd4RRPMHx@SM>B2SS9-KfP_u~utm=lnG|BM9BBb|8`10}j;1TP3Puf(BPl})^K
z-+U+ZQ@_={L0r!F$h$>kfCZ}We`#i*o+2TlXd-_b_G84#Ypcp;g{=#AY^@$YX6%&T
zj~%FO3R+rcCCag4I;MmB1-$OO7#q@y<|$cnNryjB`<{|xy&d$W#c6;C8HY%?ApZJb
zEDzWrFLr3c*{i<(5OLy{S6k2pv)lS*%nNNz$}ld8<czy>slKTFR6$Q|w4}33I%_b5
z<N4}<rJYbEG(q!*YNvpxY5ka2=~GPVvmCzI%zF>Yx>pl%_-OO)4=ibfjI~X|I-+ZG
zrud^t;k5M%w_HGpbhG8f0O>nl2Cn;L?Tn)%z3_6ax$H;-0%u`N=nbvcTx9v43dzSi
zwob?i`6(B=*a0m-jT5>9UP~eB?W1hXhhHf>_9MGC&kl<Pu5a7nT*bHF9Z$<JT4y``
z`W3r$hiC&W^r0sB0m08Lgv;OZq9?fx3DIWCK&8}#JO@!9(2Bm6#7Su<da>Jw-XUEV
z2sB4adFT_yc;@O{eu<*>_3nJtS!U2Ov%H_N_60EZ&fACHs~z?PFYkN4-k{KB&r7bl
z0D&e>v@z6uV1~xbR^}#D!}wP-(dYyJzQn!xS|4m$&~435*}J9C>pnS@KESW3Y-y8Y
zLyC3|tu~9k>6at`1RBS%b35uF6*+lYYbW!-0<e7sx`ZlGiE^3`i>F=93hjUY`P1XW
zfl+;J7$EZkiwC3_fIz(|@&vgP$cBA85`T>kX5zw_<O9^zdM69(oU5nXR<jP_12EQ5
zFG4&<V0H6S%a<a%iTrzU3&eZ(d;yGwJU9l9!j$6s&6L=%&xn=ri~x(@ntcMe61jf#
zb3>U_HK)I;IvkpC5PCmm#=4^5>g!q?V9~$bp=h$e#pL%R8H_GQ#y`>ea!*rl;Gyg*
z8sjkwRb~o(-v}=lb{IXM`F7NGYGNImLmYEIN`8RG?g27YTIb&qheQ<<mF=HDIxkOp
zNj9V<-{Qv~K0Y!D&RCn+@&ai<1Q}y+OXQ;MNj|lb4_=Di&IYn}362kV+WO6I`vGQB
z3^s&NV(id@ts}8~S?pDm-`;Rg=Ep<6f{A4-yf(ZmTxNVvj?WE*cog5Cu@=xO#WucU
zdjHr)^iFGy;oR>B{DNAiAbz%mYh6U7+{VS>`C-|0*6z4~+#d>ZkM2ZPfy*(o-D{a>
z#FHEG%y6qwO+BrTy?$O~>k_<|Vq#nT8$8R~F0F&WgEQcDJddFL%u$a%Kz^Gtsjq>H
zs3*{DQbO21`Rw*!oQ^WX+1zuXa>|J}aF%w`<38V5r&_0IcLKd+_J}-~pkwpHZ<}oQ
zjR7ymbY;4xkNI1L{%;1RB11K8LWGh#3>WN09(Yx4BG<PaKIJY``@S}yZy6q-pqQ}4
z%lWU~pJ>KmZ{Opi8dL2j{q24XHu-U?5uTsLeK`p{YzN``Ee4!Z<EX4B!l=@&Shdbe
z`D03cR@$3>eMeI$e@G@|ZUg0n5E945vi<e#{C2%twv3<U)yc<BO{-*Aid}K2>Tq?v
z|6zOpc4UT?jAWT(>u5RI;yd$@$p66Cud8!hTZyE(&t@VA&M%g5KWR!#NM2oncMns9
za+G6t*O<vB#0xl@*EQFchd9^nOod!b>zmyX^bU~YZCw%o!b4F#&Vf+95!q+C5M;)k
zva}!As$p5@=-c$=W0$UYBk|951eihKgf5DmqH6O?G}~i_HfHRfp5cAHomaFe+~1fe
z#D-10NMb2(LK|^cFwv@a;jNpzl&EHUBzgP3)uMZoxLAn#XD@leC51m+J(@fHK05sD
z;eA5qPv12tNXcn&r$2xX303B{_ZiR4MsBQCf5E4cH<te>d1v>YpX6Yk9_Y>OY!=d%
zBT2CV`f>e#L24&}bo`K>tUJDU-*1lgQ90!Arbz)Qjyt=y;fvlQRY5bcy}b|60AL@G
z5w$?djZn^oPT3b!?U$cI+i+AwTQMo3y>Ah<DzgOBN8ya5Uh9@YZFWkM#gOcUu>&3s
z0|iH4&y>3<%idi%)4&2*^iYwiE41Bf3}#8?AP#qYZ6D7#`^tenOzvKNocrqgT5S3w
zA!)r7_6J@B!d1WL;{p}ooWLca5tqAr)eVen_`ukT6LAx%+Ld!js3IeF54G+(WjTy_
z(>eLKY{-CJxlYt)KI<06vAeqs|A2nwfCw^fERGWdSd=l)2{V3fxGQh$Jn#J;k#4$O
z1}DWS^`(MvKsH4X5C=T|C@OOX)qNSvyeP9cuR}4)D=QD%h3ack1r@aoj_)i)#m?yy
zd(=`|B~nrbPDE(_7lg)}OW>xGhCtg<8Wk1dfoUtTm7AC$wpk$Tm>;=nk0+l^L<F~F
zG}7o3e}jGcI=Gt_T+8lNV6e$2K^gxwCa6r}3-8`<$akw{u!vh)yX>DkbLMtT-SoKH
zBgOHV2sJs1Q&No3PCF1kwYJczLp-|LnrwhOy9<`Bb`02Po7Ffo^t!5MY}baXO&`F!
zVOtrL3|h2|Bfv%g;(;k2Vxrh?1bnk!YM3mR`-@KxVaxti@R#59El&G`_tzX`=!0$j
z1p7kEQuI4OnvA4Mts)|4*y^@m?Iof*&2?zP%IITP!K>Fdf-YPK9f$zMSzbb;1~5id
z1Sz0&VI=NWqk_Mx=i9^F12R&i+ZSGc)!>k!Y8sk)2Rk+!d<OnjV>u$Dq3T=BXO-+G
zo-SZV9IM_KqQ~XmF8I>>+rM_uXHn2#*Ab)#Pri@DNJIWbQMh2=n;JWFPwm+OAJI&q
zNTv1bV;+`9b=)Bnw+s}8RDJ2E1bx~C8-tlh+YD6k88Lt)BjMl~O`yO_y@R)^E<#4m
z!CqC%JX$Ele2&7UUFFMtnh=rYuVy|F4EbmIzm{$o_MeqQ;p8H1Yf83^+nYIt<w`z}
z?80q)SIi~U9<=gP>t?*o>;w8jI*{=75X?#@cl)}UEBElk`$iJe0OQjDbG{mS7K1Cm
zeJ&bcaYZUv<s1zJ?KE@?<4MWea)*zHpQj+&gZ1{R8qm?_g|6}pq0-zR?9>Vbt*#?3
z0{O~Af){xIlBmpS$2fQ}_vJysV)fQ|d+?h=`5f|Mr~0MW5sWqJ5*&2fBD9}0m@NIJ
z*Tq@vWGrzPQ;FrxuYH9T|F}GKTf#&82cM<n{k$K-_QV<=JLR2$^(2B8;^x*&r28i~
zZZ7kgfy-H~hrHyIS5#Sadiv>>;vfeX;;6kzoemOcy%bHo;#U?tLNz^YLW$vV{}*p>
z85ZT+w~G!ANP~b#hk}3%A>E}?A|Xgg4ALPW-9t$?2uP!Vpa>(~B`MtlLkuvK^bo^H
z{rBzjKJR+h+Uq#>w`YIw&@Wsw19D&I@BE!{B|q)r6^!TWq&v-d;wmUQ<P?LlHyg!3
zggRHO(<VbCc6);zf+LXqJptqtP|A_F0=0;b^xQJ|aXea2y*Y-Mt@4&sq_E%q)L_Rj
zTveavd=19brYqkTwgzMt>>bn355o)=e{6vR+;TmDx~a!?*I9u<4|~)lCs$)~3DfpX
z>Tmf;f^+>1k<BboSo63yrGN5W=bh2v!wU`GGo-oJXvyEuoO{cn80ESiz3w#@W1E=n
zPZI9TcibhpZ+EB#3i^lvGo@$TPzJeNZ?5n_Ctij%elA6BHfRb?2&}@WcMI+uhz0=9
zhY`mq<D#yIt@|~+{W9S&`aT&%tKm>^Jv8R54f68nprB;O$G|(am-C?hn0`poxiIyy
z7M`qUAq9=r^~WnOv306t7V+8nMPe>GreUYWeQ-eu;Hlfce-vPb!G0>*F=xr2*r;o`
z<g7nw$n0C|>UO6zwMtwH2rPJmhSM6EcX>Il@f&H9#@wk}XdF1hPX;E<$vGe|qULX2
z1ZyLJJVZO@<N9fK?qM7$OJBd13FX_B$hI<HXYWJrr<~DZ&;Hvr^`HJbF#OU67iCz7
zSxzc>y(971;rEw{fk<v1vIM?wM(DJkJX>ER`?La0x|Sh`QxvU(gtve1sYkzij(XNd
zi`+X(lI&qhgyBG-QXDoK$6a})h|ve!KbJ$a5Xbt-ZqMP5$l<O~(3VURvu7o^tDr_T
z?rorc@v`GN@`boK;*o3xKAdA!Ymb1Jkkbn8Z?_+}#$t|IW6k@rup^;k)LY6qM5cRJ
zXYYWaap&DPsZ%&7&TlQqyTAW6*=%#&jyiTOjmxsL{ILN(a~KgPu&;U!Oq=*(hpf0k
z7LV@p43JoJ7oUXI#2!+Mh@E+hdy2mDDWI|0L#&NJ`kDnw%25t4=Z%5xpG{(ORe6)!
zBzMbp6j^+~!;!l^(me_^pQl!)D?e2?R41%OJ|f|*No&N!r-dRWTMt(iV#OA;V_@|5
zVgak0JC7Atzu1x`Jt%)f$R0)vWL@(EfxMSSxEv$x*b|4}U(NdM?40j}(Qoj4ZF?%z
z*{3cyCW|*tvL;XATWVH|`??C{FlZxn`)acAyyB6U?|RD1>2P2QzwlxXRb<6r7FM`A
z(AunvxbW`t+1oM$=o8Ck@A$rf*l@sf_9T3v?M<K%klv&H+4oRo=4|0w7{V{VDO1f~
z?7)H)Co2yt1gWiWB|X2Gj0N$)BZD0kXm+{LtZ)l4ncA8vHI9Zzspoue3KJGW<5P#V
zwB-mDA(Sf;F=s=;app{FbxtUO&3*nasvnzwF|lQ-NwF3@!DEgEVHFZkpg`5N!)UsT
zoA977IRO()vOWS6wQtL4z$EkZIv)Mzx%%4S9W|(!Rv`?b210CMM97)XXqccifuqF=
z@14Hebnc}m{F{9tWk$O&Y%HG(F!YgYn~lRg`|CO&k2JCv#S`iyAU}U_gaC)73w9}l
zuyP(H_3$e|5r?8P^;HzVSZWhE)P^G-yBC4Uv;dai?>P48%IG?;ReMd4&GBM3BH)X<
zL)mi|@x_6O*0=tZw{k;w4_`xX-QeNkx=?q%w?17XU6Ooind;~Ex<zeLq;Fzn!|=bR
zyL0%L?SE+uC{tk%VIwz>W@Wu9cA8nRnfw`_d>*b{e>O-MALt!SzMDwQA+?2U77PKp
zSd{>CKQP}hjmWAxlNP@O4z;$^E#-{p1Z}lD0d64B=ML!`eq51R3od6Z+^yWafcs@5
z5M-wNt*SWwTOk<*=vEsk_9=iLxqFF4|9C;ZzO8(_?(in{$xiR%S0}ktX#;1Xc%RVF
z`GVet8@?I1XWTxQ6Qn#dUp(J1;l`U!NvHUoI!k~du<W~{LA0~s>Q78C<V$h3GZmjB
zXONl;S?Zk0LJ36(Dfp*<f4+Dz)VxP8+NkAbuwzny??uF?R0nej8`*?Gz{yHuUH~}b
z?#;qjeZybQwkrAAPHDb%eh@aD$)r(~$*0&gh!^i8c1my;L;^Ws)5o|*i<v!NyZHxX
z+@7(sI-Dwz`-(c~t_9v62wWFLd`Smju@;O#+ZS_3I+3}cc++T`<VIugw*!XlckkQ~
z@uzq3fNmCjv)6{<8m-w}3!Lv-zV^L)lFD;yMsJo+rKH5%STs_|gY>5<+5Td&9dNnX
zoBQUNwzp*;>>9MEuGtB4WX-u)x@+HVkG=jNtNpz_6y5k`+Adp$Nvyenx*=<0(BQ?7
zrTiAd#9SRh5wA;r+`Nm8iPI9b8f~<y!;*%8FXvlH$^C4sU?|{i5fcP*1$vR*ZNbji
z3K-Q}*ifdC4reSA&~i8;YvH*&66of$X6C)KX?!4-@I?KK6-QHJU=2;GKVkw3sQO~-
z#&Cum)EYZ_gGjUNrFp*|_4Pr^+*AoW4xv)QL}_<-ykQ^`AYalX#(;E$+9>=^q6>=#
z*^%0O|705i>x$F=8YntdmpOZ|zu7*W&)4gmp*AarlZnPp+hkp|@Pfw($SPNuYIC0K
zT)a<H>s>-_{Ib+i=azGK+J?{9^j=(C3F2~1UgaB1oc;r<H)5!|AEtMAPmPa4#p-G8
zx;3km6zF2MI?${*_@~mUvz6*eV8K4ARE6xKo!!R>AFFl|Xbv6ggIU}3^)!QNMpet<
zzSZNN+*Ch!5rAA>@9mbZpnCi`X9TR%)M``Lde@!z!5?ibmEK~S@Vd}a=(J=CE$w-q
zex*58uTRzMD><1AY?|qNC{lT{9eL)zt_$*A>4?U5g#JYKBAwdppMcKAWcmn#Hf~k^
zKExrjgV^|(s@@ow8gK@$?6A^w)tw1ssxWahtXWdF^Ztw{;vy7j`UC$2FerNLDza*!
zst*&p?z?yji*#LT0Gpk4=?g)uKh7)o2;Frq&e%3hv%IvJ6LrVBIG9HV%uMrht8mx6
z2x#^l4t#Oc?K{?)-5-T+GBFmkBuTA4An6JxOL|epmf(4+L)Zg78d?S>g@OdYo^cEM
zHx;<X*L5ZREUMGlu8Xd@82oaE`_gKJ@m}I1yy$OaD=C-hzdA4b6Q%xo?tvwV4GO5q
z8XJuQkhyL!rWxBty3%%?*-5U~*}$0BfzAKK)9jE=94iIxTa|l$vrRt7qRnjBh{%Oa
zNLOIvj+}Zjk@~|K4UWluH7f50;j;E~a@F6v<y3q(L*#+Put+n^*433g6n7Uei@CH_
zPObD_Db!r|hoW>tcG5bK!=JQ2R392rZB?Q^kk1|Fdc)%t8oYupNKinz^3%Wms)ujZ
z_JeO2ZUdCYrSAH2?>j!-m3r$tSued^cF)xrb`wwWeg$_6;0|wKo(fmQ!8a06-9Uz;
zw>P(NL8bPFnU6d?sQiG|kGp_0wP>0Qh=w>>xby*_d|?SNVBvS*5K9F$DiuNBuUIAy
z#5BITxNtaFP)iNHRVOLqZU=$6LgWl$pJPzSme<ps+V79s)BVJXGbK_?o*&iE+Rq8u
zjOkEJ<a#l%5EoKf!6k5)M;E(ms-cGJQ<ZD=Z4J#D1A!~vYJ-(IMv1T9{2HZXcLS>O
zTy}sOgu5NUJOO;;881(;ezRia0V`WkV&s+^2XHg4gv~$~i!`;kTrU}Y75`l8^%F%l
zQQ%|xp4`A{*{(<cAfJGkEg%czTP8i?N?H9Qpgf)7R^2+-@>E1ds`Hc}`U+A8&?YC8
zNXpfj^PVI^6Nd}U>3iq(nNK!_<e#V?S0}6CCo)NO{WdQE!gAGt*{BKK@#kW^<A_Te
z_#Q%WbV|TXXVQg~kJ`E}oHOCPZPwzGEG4;bl74c&pH|`c8XD#q5G0OuGO9cz8GYbf
z^0RUSKhZ7UO_5Czk#aKuB@<(dd3`U7HNnT309zm;-S3v13W*@x1SPA8jBmCxYv0-B
z?aDNmrVSl<uzIZU`kEQ%9PIV&O7^JlQ`rmK*5)bx(rjIjz&>>bH?zBXA!RDZb!MQu
zjehFk(^s6g-?-iHJaF4Mfm-?MsWB2poCYyJw`B~<6%W&G@%dv<=PhH$(CcV8Qc$1Y
z@$I+2b|jzZ?w@^ULF>kV_Yw73FJK14|1}%XgaM_*LEPpNmui8K>Gp@ek&x|7@A2i|
zZ~=~{hq}sXwlE6{qT_xScc(QY;K61x#yswbeDB`ZH8H!<rAwj2jF>2>-HJU!;S8bV
z3|^1coDBw+mOrW`jiy_Mgn6UE!cv{+d_ZLAjzv|Y5?R+CHnvPyzW+h7`)*I);}_Us
zl<;lrrHpbFqj{2*MWwZg5l4s=Js<6pEw%`lor!OzNnKOy8p>$eDxaI0qoO<PJMfwp
zhzW%{XslStQ9}qvoNi-`;9kPgc=X<h4e>qt{r>ZX#9UzJsep8MIIz9<@9**R3>dk!
zGb$7J&^Mncb%4PnuPa10Oeo%G7s@z~D=@2WiQGgOb|m_msx!q!{V`$=pKXu2hRbSv
ze@Yj=&woZ5?R>ybK3z`p7R5a%!E9Bf@FE+uRSVFSPJQmniVX;)=)LbN8k`<7)s=qt
zy~%RD2jI16vy^XVNFnIQi{RM0ed!C*Avwz%nP_Hprayb{x|82&QH~?V3+e$mTb#3N
zcXm)^k%z1GEY!jNlY3~1fT3Hz`R&_4M$A{>gOtIC^F(^v*(c8+0lO@g7c%8O)GC{{
z7C;v+FwD@4T{!>8^X~rU>9LYtKE0xuY3IO2Y0Y}Fr~v&}b@x)c{mtkcTbiq@%b7O@
z*b}?6PYTFkL$F5;h&}F^vjl1xqA$h|#j7<qKkDW3);dbDQ)n}c$vrBl&!{OYO%iqH
z7v$2EmlJa_z;`a+tgS8=>eqY!a_oiM^j1$RL#1Q3VVr2(l7dcC*bd|aa`!%(8=p-_
zi9NgvKU|T`=<hb`EGfF3yLz^kMO8cORo%Usr~aeiB*TM)`X*A#O869EjnLb$VOE~0
zR{fIYUB>p<dT|7?Y=KRxF{g~y3dnxiU6DB-Z%hud3q*Tw1t+0y)xp_t?JPcu)1KRL
zCf{j*^K!heCre(|<mPVC_XlEr`GT>w81W>u9Mh}WI%j%gqa{_(pOfm=DY1_&3PpjB
zc*_I&dJ@oAKv`~XIl*KwHBL8d)2G|-kBU8>`kRU01>Gv4ttL7H(;wzbp`zGoRAWRH
znC+X}o5~&7EdOrDAa<OoYWqyL^}AIHZz=v7d%mvVlxO!cbFH<W+I$QPk+E+dBMRgT
zm=?K1=<c823Jfo+H>A6OF$iP$v%)4wRwd~9GSXD){#vTOsxBq|9qp4H0uU??a2nv9
zr~<6r{G}PDG^z6HX`!RP<HH@*EoJ}F2o@?Urso{icM4AuJpdvM5O7jb`8xIc8#KTj
zk=pZo59#pumepBRq0Jq_JmBf|(>|$zi~Ly7bI1w`8D~80A}``(P_sNbJNtyb0!<l_
z@PcO@B&R`xn^BnWcs>30M#vW?6Yy(N<}2f}&9YjmaB({O83X8AQLYSpxC2?yMrypC
zK4=%hg*`f~WP})M_|o{1GQYti0i9dIil>3y2t`fVLZI~o|EIP*i`J7jjq&XU&!}W9
z!fz(OUh#krU;l4?LGmwfp_d}QImVdEns3|4`uMkp2gNNfeK@NIzNqlMib#mW;{<2#
z+<>l#XK&bWea=2Tj2SrsL*9|}-?LC|E+uiF(I7G_sj;^nr~eA~&S9ewYtz=ng|0Ci
z!VkDI_oq~wa*w`kyuup?0d_aT?>BwfeTDKIsNWq<ALkYw4*8O6Y&41r4YnG|x-o^6
zHGkum|N47z2>t61t*@FBIw!sIs7*jyzXGxzChE2|@b-1GeG=P20V|*cz18}nZ;t*4
zML-yX7g2p8M0hwhi_EBoH+S8N{z}XD9GG0nT-oUiJQk{WtNSKWr9)JB9eKcL$l<(p
z?~tGH+%6+o#;36lCmO;s0=|5oAK>*{X!@f%JHpOyz!R<N6fIw0k9d85!e3?Alq&i2
z$5Mv=!*-88p-#Aa5g-grK<#}+Q`QdhizVRgcc;*3@D1!`tyD)x6S63Bvwmp{SSjdo
zLgC}|di}jyzM5!w!GH4S&L3DuK2GIful~w^IlG@o5D>`&v>}+eUGn<=vio2HDA7~z
z^l-@5a0p{LiJ}P?xObcoHxz+1%}2pufhNVf*&l?>aIuTX=d%{I?^LUw5oi$bq1r6n
zK`Vw$z#vJd2^=)8$sT2Rw5!~>XP0fmX&0vY;ayNQtZpxu=!Vn`r?I*f>&fr*DokVS
z&J*l$+vki#0-@$xC1Dwo?plhc=kiOyjsN-|P<259_(u|O4wu&C)C*!(^q?dh1{5MO
zm|uw=EuB@C2e^6#1xRKnIXCvxgGul#h|y`Cz?XmhKg9vGgMs=-Ilkyl#{MX)rhC=q
z5B%J#*CEAm57zyuL)fg$%z$qt^h&t&uL}dR+Hplx@DC^+QDYK#uQ_go`Gz1!m*(pW
zl!H@9>7a2#6jE|@xQwYPAGdG56ni{|bkb{^gpoUW{B967d9lgEfR=OyNDVUV04Gco
z@8tYS82(@9!vB3PpxPn4V&V>IB`DoCq*M2Dqu9!LhA(h#*(2cB$qmZ28#mliMILl)
zZMt6WA*LkExn`60ya>jyd1FX(ARD>_b@EYm7Ah#`R!eZ<HW?ycT}g-z4%x)Tt*F<U
z$TaQ?-m_^6-M}xB4!cz^1v-!oaoYy0)0qnb9b0SvIugc#(P|8~TDrZr<D=l%(j+cH
zL)iTblxjm{QYpk3zJo`|4qTlItLd$7fY?Wy9Duqn0C;TM{uO7uDzf4V`pwGXR8S|E
zE&nHMMWm&sp0H3r5qqmc0*3JT2c$p~jI!-e>aa62xO|NhS#8-j*wiSmPqarGQy8an
zr_gaXk$e-GM(5juC$|6n+Xfj5wik@d?4Z69vU4agUqGUa$F@`ie25mQYQqMJncXGj
z^UrqlPZ0Zg<xnGE4wbr^<_xBQp~(hfCg+nUe}UaQO35?uWQgo9pb&@gQ0Jl{>ef~&
zcW$zLzJ5@^svJw8+-eYqihBHRDN@<~AtY?|j)2>OBz(A}ttC+hvp8B=$~$x)pK6iu
z0*Z~qh}S4|u|;uds;saW%t{@`ich`UC3{PnGVNkH_d6Fz0t5xwnJZsD3w+du%PaY<
zFi(e_ZMg9GRprt^RBSBnGQ`74^}{NBxt%6TrzP@L22;z&4HdG))Z|x*N`zJL;$R2p
zT_?ROZk#2X&PQ^~PBYZrPru`63kwr^w6IYBi<Evai2evR2orIN?(M5u-CpGbS8Iy=
zg)T1?X%gqzUYd7dN>?fl6J$A7<wTxL@2@o<U2)BmZ*Hh37dxu+(nns8%Mq<3dha3g
zu4^&c&H)~GCm#E*t3R)PEQL=sz<v&&?6?Irb%YTfon4>iJ1@=Vxyp`(NXR_<-VY8<
z)dtKL0w}B=Ce8GjfL9WwD&9Lz-;Jz!L)qjn?8N<V<1b>*4ike3;8z}>d~Fw$rYqhw
zO}r5=G<SKGA^c6kMW;t%K(bFt?BJCSVV^j(x-@QN@xi&_I7hvJEF#u8Z${*oz7R-D
zhFbn}n-hAgD@dasiv({=C*(@JwpN)>J$Ena4?Vf9Z`HyJ9JAXm{~P)UwjkYmC2_+v
zb-y`7=3cBuWCRmlF+Pvgj^Y4o77!US2DI3g#xY(jI!gCNv1GYz$x^>K#vhY;)q^K_
z_MpuQNHh}s7O31ZDR|1FGDw;u2^+?kXS}lhWzm&xw8c%|C)f8FCOqZUrut%<|M0tD
zrA13aicPF?@&<r5p`}OE<0`}My-M6D;Ca!PdHD!eQOD#46h1N=JnYY%og6$SIIMdu
zhJURFul_sR{BQT%f9l_|Y}iN~_tz^`gy+7U()8PvmasYFA@Wt5C!tH-5Z<a$(=eib
zt)Rf&pQ!4s80S{i(y@wNX<>));p!L8oEhqB=8F1%!4(5{{1zd-n1Zvj3Z^xn$RWYw
z{uFwOdc*nTnE18vt9H3}u?i%B^6$2zM7H){tS7!{V$xWtXI&*e%OJ_It;C+VlA_2~
zde^WQRiALZaHTTkdES^j*Imjv(Or;kmajOHq)V&1YOZjE)@hl%3Me%Dji5GJ<llSF
z8Nagn5kmP1KTM7s1p7DvrH0KlC#(Z0H215r*HmljsuNb<7o>^@9<cl8k`=u9ZSHy$
z_p)UQ2rjU6B8_2BKV$M`NJjG<^HvvF0yOd|qQ4*{yCX@~64-rE?qpDjj(_*LgQ8O`
z0g>o@u2*6hV7Fp9>6bV{00On!5$FUqL&TXg&%DDqu}}4{bw3ug#UOy3K*_=$Qzy(4
z`#xXI7Xyv6`_^`h`S$rgAK&LcRx!h~B>XC;ZBAHN$+S}c$*(lf;HYTMbjW4O(R+6C
z!&}!xb$v6+GT<qC&GxS}igxnraxaacvG|MLlHTEJ6Qg3Pg`WP6nAb#0VM*SWnr~oZ
z*$JC@nBZ=xWm5)a?Gk@$7t<}Zx#RCLiBW`&gk^W}4i%Ub>_%MsU~+93iqLNk6aD^W
zwy|e<B+{Qthw7NvBM5UncJ*PDH=`Hl5wbz0PVMY7qJ8y0xQx)%K0c{0EWk~73A(st
z^Mw-KWec#y)0Kw+67)t$P!AClq|-Kw*jPTL79x=e9M(p^>abl>Ip{{SMNLkwsp;<M
z;qOG<QX5RwZ&OiJ2q~nr$Ms&yI4$!7s~xyKV`WDD`K5KL3?hY90_*^Gd<+Sp(4yqD
ziqC?P-i*E83U3XvStAdAHDw*yBcn{4CR5Y$=!N*YDVdP0^%m`EFSkGFHRNOn8!;^?
zMj6k_dFeb{6X%rC@LtARmGM-WB~o3I<(J&96ZYjaoL^pa8=HZkR(1AI>-S(R$j~?9
z^uiAi4#(*bE_?n}1R8E}_f<SUe}#juHHndCxiFGHUqe3Fl6jD5@JeYLM7sbRK&J1*
zdxrub0bi|ZS{pu?KB+j?sABl=<Bagrj&KklXBYwkeF%TCN?v~qRR8MRpx-eemca5|
z;DB{ap<eyL^On&yBBQx?>Y%NZ#5_nNfJg`(Q9#k0U&Zc4Li~wXQ+*hO%g0oRAS;eH
z8TK+nPBY*du%tt?<fW$%y<F*?B$^Y38zgj>)SK0rI71c`FF+t5#M-d_rspC8iMxC~
z710ycrzup)NOl#eobix6$y9{0FGT*==h&^Q;I$km0ZvsH0Z$u_oqRBK$N7t)EH8M0
zGc=bGBx{G}2X|%rFZ-tHMh;30Bp7XMaFi)CH@uvhHcVZ*nQp%w22^bKn>Q*;Qun*k
z8=Ny06Q0~^e}c)36@$_{olCbjxlb((FL&@c<x{EJu%<^SvN;BK>*Kg@Ld`jVrldQ;
zm>q@HZ&x&cA{|ck4+s^lfIAB`1bkgzfdCN0m4XWb-=v!>fSc(DwE2<cN7~o7T2=o6
z!PtSq+ZhZ1wXmgEV$iVQZQ3=3OAO&ZAl9NAMtGOPDG7${`fti?l3}l>4k5DOXa+_M
zuwaJKK&Mj=+m(M3&and@kUc|sg3s3|Z{(n~6aV@5_O(L65en9Rmoj7nU*KS7uxW_?
zc9~aA^9Ky2244P4J{cw>6s<wndbC_ZzNa4g;M0Bds9TdZB$u_Q>a;SfL+FZOEzhtE
zdEH+1$+VX4UA~IeJYDnSgVM<kWniNDhUm+unyTBSJzi@`@=$H@$v{Uwc_$J^G)e&S
z6n$y#da2^*)2z^yL~ZWiBaB@7_LOmms++eJQ)Gc1tRXww43T>fd-Z+n&q$EeDwJv$
zjFral_hZu$1$N%`2FHJ*JariOr~4>%#9Q#TDQ-XJ0q^>wflZ1e?xa(?0AOwJ)tD0a
zmb^))h-Y{jKn;@I%D!6U=H^a=`oRXJBbJMf{0C5`>QW7j8Osm09&(rxDD6p8D)oYV
zUH=!b?YIO$A3zhr9ST-z$Bt-aiJCtL+uy&`Y2*zol3%;@x3kq+nj4kscf<nI``Q@J
zMw=;k+@!nU{3iHD%p6gEbuXYJ1S<~?*MAvo37euFDm_?sAKftWB6<-oGEK(JY;maq
zXwQI-#dKL$0Fgo|f>)mX3y_xMtX4&jBtk_CC`yLRf?}9Fd%X{RyuCa3B!v3wA!WUP
zlOTMd`$=WXMu9x_8jvgtjW#Wa-%oMbfnnV$z5Ne*z=NatB{V(#Lh4BMW$R2k4bJvl
zw<%-i*hJAf+hB22iG%Z85}Xa%-yQk~WKS8H38lUgba41$-tAOoti9EZjgKxKKz2pr
zCraUi2l@k(bl`a%&Sp;z7CMn)C2c(^CT7c2Zj9pipA6k2&p;<hI+Tu}6Ttpxqs@?I
zI6QMDU6m?9n{BCh7`j1rW7(FohUhYGH9Xqt_P-tXO_lZ_l=`U)!tt3$nX=0AE9z1b
zB6e0^KtoB^&!Mt|Nv$lVwc)&@>IT(>O_<-Gse}<-Nu0nBm@aiErQnfQK_#<1AL-8i
z#JE;c>hOT?QO+CoaoP7=j76nB=#=4{65$!+1k?Y*4qU`ywc{1=H!w+|t^7sX{@G9c
zR_E@`F2RL4^5s&BTzQKldmNQOiABuofl_4EU7{XJ9Q}N6Sy<@*wZD~SycWcH1h{<-
zlt1b{+p7>;8qIW)ooq1lj?oZ0=S%|@jIVD7VUL|YX6zb|b2JFxCo!kV@NT8v?ZcDY
zPbqi_5W2xH;+~CrF7YlctG#vpT&Ei7Vxvq#+#g8N@pqns^1&7^A{I4Po#ka>?hU??
zoqR$;yY!rZ)2Gw+Y8$-ckI_a#R<K1F{?nWhN}byJwrX<HeqUoN`-MCJ-W}lM#ao0=
zTNOkB^>uuL&WC(oQc27G?irQDucl0|KbDJr(4KN#CFb<I@W?rBanPBoMnYFP7c$hj
ztm+Zw_B#!bQF@~0HF`D6OcJz>$I(>+vYQPVhRIxhNXEy+7b#HdLS$@vIPD+TFDTRB
z{t<210@-?m_W@_D{Q_A$B|hVY2zGKt5Ndq1oP92wLu2g{kM|?eY`A6Mz^(K2SHGSj
zoB0GZ<3g*%uV}{`8JXqruwuU^=0UbdqzcRK*@OQ$o}~K7GjxZr4R89?=4uxWUx4bg
z$V*RnqCEuX^fXN(!hHrf9c*-QXFtov>U{5omn}KG9?8Szx$)PnQ<wLPzR>@$RaHn?
zt6<L-mI<Xg8<8=dtTS<vDdTVd#ph!@H^7khJRSdFsNMsS_)+SlgA6v*T-Olo`RB~_
z&hG+?NSll|PHiPNdbt7{AewmV)6sWs;aQiCC^Ll~e#WlS<ayPS;$4cvWVOk`?{jDO
zU)dpB6@JnoUc1J6=CqK;{OwhYFRl_YV6dR%n5LG0KTbFYWosvd2C86aGwO<V>Sr6M
z7UCQ}H<Q<XtKuDeMN{x`oy~Feise#ZSK235bZFnoY-%K-;-~flWR$S@R3l!KA_F_w
zg>Czf&mpp4^jK#(iVasu?x8xf&u`=8hL16e1aDlHkc1<+*)IbAML62c)CSZ4X3Vdu
z{5|Gm<({b`qq(VSEt3wj0g;8K{mw@J1A42m`?dX1uN{96+=(Dw^R9CHEB+QY@bn7o
zNh@Mid4D9TXN+{nMmG65GQCx~=q+`{gL1*~Fr|xe!f6GTb}-cUq>lqSQ5Pp;%pGBs
zPukgeD+Dy~ij`=GXeF^!lKH$lc)s-2_?2_9<Hcx8>5x;YX~|sTn46F9oG@jUB55Pp
zd7t8L$`-0_>3XTZcQ;4sz$U=N#6B?ExZUWSoj-JeyW$<{Hr(aAJj7Q^0}^B5C#GS~
z+&8zxH27^WNC`iT2Yed%1OJvH_ZNFIV)i@Qg<?m+`zB>(OMx~s&MpEEQ<0;R)XU<Q
z2JGWS^%DzxJ=q&D5=T0JpWa}SwuIsOeeoiK)=2R_+_JL4(;Xr?9`H6Qp`hmA6WW!<
z%r{PXX5aZ{o;lb)c28n(J2vE#jxTR{*I~Y2=cgk$)@4Rk1f3^hI(ydgh`;ua{CLW1
zNM@?ixe_0UG+%uklPHpMx&fF0mN1!E-{~s{bj#SC?dINGQUEMR?mgz&`6RDfxPGSy
z9yrZL1T16a^nSu*rq1?VjjJ<AZ`mY56npbuSllvJp-T<&?QaVb`Y$Reuoq*3b)Rn4
zL|15=PMj?eIUl0N?BQU-gt28%8GEH4Vfd(2pH@_t;5@Ly?~(`VlJ8Z0pEnDNmBPO>
ze;|`Lk(|j#h-ci$f*GS%kcW-aodX*_df6AQ0-QMmE%pzV=a12PKa&)t(2KLEz0v?L
zXH3nrhK9yeNrA~VDYU`Q!Oz6>w^@z`DW?IT8{g%hEdM2>^OXIX9Oe<|HZ5>Q8E4Mx
zv>W5X$f6S98y~MVFai9h2%(snx*3!>&p<JX#+f+9ST(dJdNw8)MWh?hd6CJiegNXm
z03|l6Mx`9?yI?dcs_ED#qp7lbFzoZ@yl9{|*TxXvg`0Jb8O>dYV>zM#_JJ_r8x&xD
zHHiwlm$=NN{8mJ=OTdT%Gzwx)%Z<0VCjKAlEii2jKJQ@&<USX<75j#3-V=E-*&}fN
z{Tuxjd7auys<1R#lA-eysCQT?L3Jni=+26n86b~$0$i(=S-yDl-ZQZ54UrlY5IRK2
z26~}Qb}LkSk-an5Jq;4VW&m<Fs$muar)`7Sl*|_{@1o2Ed*m#8?UIe<Gfryp+Mhlc
zp^oe?BkPtw#lzYFnIU;@d~76dbsv!wU(~)FC@K<6hgjV3mL9?V;h^if;Png#D3&4a
zAJ8*xMK)T%cK`pGKyJ7It%~TW0oWgPYg;ujW2*%1XHWsI&MV_2G?W^F?5KiXrS9w9
zbPEHN;zu4ZYBvigrjcxQ1Cpk?YZm3Im-{~1DiW)z8a`3lk>Fi-OGzf$H4_7(Z0?tt
z{VpH~)bJ`b9ddT3Ch7qmX(T6o@p!D?dj6YFj8WLO9RCL-1(aA}ohetUjRQ%i1BnH`
zoA2tl)(&~c9P=%teBeEsldUC`E%=B@Cj7~ns?N=cSnQ~#JzF~b`B+fECf5X|fiG|6
za*QT<&yK~<rSu0Ki(K;HO6yxb!`y-XLE6*Ot}!r>&W2wzzTR?a_OnJ?_S@rptK%o<
zNJ$3S{8l?~{a=y;&U#&bSos}Bj>?f*e?Q`Om-@q7mxLIy(z)}yc)QDv!F~>F9T4IX
zSQu=CE>i+X`IeQE+dI}4pD#jjxX9@G(W(U93Q-^donHJUOium*C9+YPtZ9cuCXuhy
zG>2^ntfchmG?xpoF|U+2XyJCpF9-nsrQu3ADK>ap?lT})fZjZ4|9-?B#c`Arewufi
zG^$C12;bbJOn2+6B#RVxqT7!D>^_j*^zkYX=fK_5L5iG5Li=q3>=g27UWm}#694hm
zU2>gj5m4)v8jcp!*PYFDgn7!?bEnD#<jm5jRD`nYFMeuMl)Y9gq#ywJMJhV88;%Kp
zUzGi+c;O6xgDXEW0*bp1BTvaW^e*N4!Wo}B^G|MYDaIyr*bhkRC^2lSNDbhO8B%6a
zVykn*D2T)^^p4sO)-G2&B3uqO{3-=RTO4dsMvWJHic^Es0d0b19FW8&l!D)UlUTLx
zVnba8ZUWl_I1OGQD^v3HlW*{gN&b9wcA&OpQX)BQxT;V`3k3W&Rj}Fv>MO{`;X>ra
z4up+pURgHA;oZ5e26a8{ylh@^B5rvOxJsrj5t55drz3up`@~17=rynQ6t0}Ir*=W5
zHDxqmR-S=&g`9U@Sv7FAQ`P62I7OG%r>y03Q8Gs_7}Eo?!EGh@R1%4)hS#by<J#(Q
z+K3HMJn3c2L(C1VP(FO`uwe;YRuuHQ=6Ci<68%|CN1X2+RtWS40&5eN8?q7fWe|xb
z_p8X!trzhNmWMK|!g!;v?B20^cL95x{kmi5eoK&NGd%J;{sAA7*!J<(h5$Wi?yQSL
zelvu}0aaaDi;QTT4pc576=6R~Xg}c^KB=VzOa9J2g93r-2ko4uza=hIO36!>a#iNh
z&AL{;CykGJh3T0wVMKCOV?72EhWF@_m*%w}whargN?AF7kC(oq`V?9Qv=!_;_y^<}
z%nSNMqkjFttLUpqu+#3vyQ!;AyGtIAM_Q{s0>b{*r@P^k?H&K-uK#X}{!@Q2&?G26
zYRw#8x`Tv$&yc7;3d(qvC!sPCX{#$V@V3RSjDtmJEN8o;I^xv7?t6Aa41ac)HM#J@
zlb^ap)fp>n?i6yA&o2gJK^Va#1$-M^dd2UH_JZ<<3G4TMPCm;cRT*H&QtQ88y`SG>
z^+jq$R^k$ZdhIkzCk>C@&mUySP&Yu{qkgQH`^e?C8kk%1RP4;=E+|^O4BT+H7MLe~
z%1Z)U4WLYm9U;183whKUia%khVyuN~ew7%IyV(Q<-waDM3&z#v8R}ukYiDOAh~2Fk
zBEnnllNIrIGJxgh;bnl5@fL33=yot0uH!OTWA<}Rwta@wYGZu$28iSpyljUuxROu-
zARn=|E#tCQ#ZHV3(fpg?+Kn1jw6)5{OZ39=4fF#HKv~1<3va+MKUTmHY&YNPg_df4
zTH@q}RQHT-xU<-2bX%E-gep<IY(Q#D_9Qgz=n!k$baWnH+h%8o!58T&9nDmy{jO}h
zu0j@)F+iFuf?{^dP5NSANP>{4umApeXOl?79Y2*|Y)iTBx@77@eRG%{_@Py@5Ev1Q
zQ^kI5Yiz<HDxZoTtJ7ASaPkbRC1|$3eUjMiA#{;u1O**b^1}puO_5`qTqe`8n(->9
zR*`x;lqDco<f8fOg~uj2AKzFC1Lp%<qYT7srj{;phd7rkJyEfT(j2nVfebrdL!-5$
zIkFvC*dG@^2hI=tAo+66I_-?B>}E96=UxYctSEnJ>V45osarx!<K;HmWh-uqY=u5_
z_4T9FYnw#5@ezKva{W(~+cqkOXd9X$4VaTtZ*v^OLlhiHzaGFuaLJ~coSzvoc)F`(
z?mS2?8n@1u+v(ry_mE{{ypm(4QD@YeQsphb!wG#jO7Ft}LQj)`u!6C`%6A-b<yN*U
zF3(d{rK!|Deq^4MWt!H4PnC{N8p^e;l_w4)WTnK6bVR#csoRyc#VVwcf4t#Vd19K2
zsmm!1^ys4;Q($uy5NLnwTePn?BovGJoHXpdK;=s!GEVt?;X1wY;684tu$ydBmHg)-
z=Kx@F`n;@y7xj)o=>kyuc3{}Bc1;+z1Tjq~H4MJ^kt2=RP_GxUtvYjRCv7Jrtsl0C
z<~CXB-#eEGm31PFuG4gp`NdDw+bTYBmmu#s^+rvnfh3q2o&OjX_BjMH7UbZ{_3NN8
z29pP8*Zafh1q3aP9@c2zU2p@kXXh@~IwWs|%}vb%^)VhxBb<k<<z6JB@kZY`zke$N
z_RQcPw2QC-gwLHo%bJErKaQ|9<Qq}P9=?$K9PbW+%(|d%x$C@^y~dV?n14Xc#}Nt@
z(zD0NL!Wt<`#CNWK#v8(SugBepm9rMGcIqfp*dSU&r9QDmp6Zev9Z2N0Pq6ihDEUc
zhD@qJBXP`&-}tk-Syl%3p)NuxC<Y+%uWqxWChRHmxi@A&x_khSEmZU}pgbVC59#}b
z6al>dtMU_*HM!|?Aol()!b3xLOD9ibpf?f!Jstq3Xm{DF(wBY)>a^7*I^08=Nvo!^
z_dC&gc5yT@B?3#986Ydxu%jJAa6Oy8CiAK~-!?<EDIz|WDV5lqf><{8upJzDCv){n
zQ}B03#%8G5*js0gM54h5pKNToWz#;Gdb0R$CAk1M{SoDY;f=a&`ahsIA3IK;ogJR7
zu+_t_Qb!;+-9o|JG27s^Z~t1Mz@Osejv5Sc>E*v@(?~UiQ8)6g#k%^GHZQNVf<!vk
z#|gpC_z1mX@d<U&(RWpSC0R*7Se&L?AxGf;8}<g1@RBc?F5SGy_<TJH$^OitaMY#_
zWUn9!N-it1LW&>t^v6@5q-E3#@Yth1V*a@RGRontnnKE?mFu*#&6wA&RKH{H?|i1S
zP+;Y)ECrU^pNYDr&MLTmU&cuE5Te7OHUJTrD=&gGMk^5B3-Wl<tnsqov<#F@EfSyk
zB4Vt_;?j@?`c3!2Kzu5XwZ+p%4EX+xM;-ABHKPe@y*au^)!jwAc6@_{aRGotBB?nT
z`4W{3WWO@<V(Utv_Wl*4IprqmPC;)srQRf4d}SNkkrY$;A2#695DWPSlxp@Zc+M7j
zO+yf<4|Tlh4L;VW)4n7FD67vFg}?*U6^OQ2!SsiDX;UnjYB;m(@X)B>@eJ(jyz@^O
z6qW%ZL=o!1(mf>DdjBokABNh!U<4ujQ#*w(i9(b-Y1A{5{<f?Vr<e^$-HUk8eN`Jg
zRiOMRe#J)=NJ#(;7p8W-D15;lHt@1N*M7yrlnf)Ag+@ZXWO^Sx6f?pOI2j;z$#N<H
z`gRD&o<K#xOJGdCx|DdnG@9eo`;{D3s)Di~O>0DC*xqDFVh_zUeKYri{M;OlVj^mf
ze{#3SGCXrJ^rv-yg)SRl4$vX|0i-_v{s#o0MF*ggEc427g*`-RK#mALIGlZb`#Kvg
z;Ce&sYrBT^bK!a<?V3I$KGT@;${I^jNNK?Xp0Wq->_^o@%tEzk(*2Cb#)+UShZ#UN
zSmNd8dW6=9WhhR3boNmFrBUdE=%o<Xm&uf?imY4LaNxa<KN$l?JU9+Ph(WBT;Br=n
zG<XDi6E<>hJOy@}#Rs}?gY)x*mOJEXpv7#rT}s<<daEWmO~v_Zn(_~h)q1lMrBnUK
z2}jE>xt&ZR4{02_JbvAgk_4KJJIEdGY_qaVo(4Fus)=42?Y~^_+kB-{nCq4gc@Dha
zv=Q1?74>zsA{OI09=4cL1n|>sAUGQqm#3;pTzgg981VIBglR-d+ApM$4@ZF26jzA+
z2Ten5pH~`}Dz=iK7sY9R5Fc;G3rLG_DJ8$PnrmO%%w`|I_3J;#vR-JE2)v3*4D#Jd
z3bRT<r(4Hlaq2P{YY}mLy4{|5UFIe8`-bU?_vBE3lP){J)HE`=Kw62miDJ3pq=Eag
ze>Q}p`ugLS)MNadzHAYdN!}k3>h8o__G8?W?X6RMMN^(bGV}~cCL>*Z0VPX7v5)Sc
z{o`?W@+q$dSrT(}SNcpeuj&&@D`+W<I$xuJWP56OpR;=Y$-#lU(Fgogz~VXsXSn_Z
zqb}{y-Yeg(D-&%gmmA;QhW(ysx`-UOsxcpda^5U7_;45>jTGi9IXZ8A=xy8+Y$ONP
z^+3?(*Am#Uwm`|-%zWX=;(7<dj~U&<)#s#X@d$*TCP6Nb3s)k2-(AtfyMIyYA2&M>
zx%6p6aSwHZ9zFmdTK*d#s=$zjnfz-C<gaCq>l)_B`uMkOkJs<@zgVHL2k-iNK^TgS
zR@>@QRaSThiwnBNTsrvqTm?ig=Un#eGM0A;Deq258LF@laU?sP3H;Bm@PBww%!}GF
zkajO}q^qC(bW%IU`(@9oPg{&2zpWUm^~2b)o{mOhDG~DCTMh@>Y?r>SwaaSfaX1pK
z7Nz>_)~sTbZs5R)A39D`Yqt&QzanrGgC~j-SCphaGPog=eXAd73Le-3U`)>6s5(8J
z={?bvDB$j>rOe9@ENA`uQj!B#I}jER^=a<ow_+MhU#s$x7dh(X-jN4tUYahB2=*}z
z6lbf(G8x?ywdG{zAqIK>;>0W>2#;WX`!~EerTs6dbk475g(s?X^-J_^&{g<hGGWRf
zSUhy68J1Ub`Q&020ek-Egx96t)cRyv@2yw`<mJcS(j9L@-O$Ejnj+w}D4*l(u;tB-
zIcW15vXdcB1I$7(&w&~se(T}clMrF1SkFHj3}k3-ti@HW($!@fnVo6)W)-=Yq!jDh
zNbbE#R(+>dNLkh{cr(V)4-;<0kVm#*i7QhGdUU%s!v<nQDv3IgMdA65QQg|}-*I|<
z{@zVxxM=f|xR|0yQdhp(@PqXX*#IjHya>R%1u*`bQPfec@9b{RJ;tLSPIcPyk1B>c
zgv(67;#)Ys?5|Kh>c+kU^vyDT-i&AG(aJ~F57#0gP>5jsEp@9tV)XBkAWGn=I)jN@
zG00nEV_x31>$~^*S!S0_>N4h4UZ?Y5-tO^Ld!EOL1`vG+%UUg9M()O~UGm#H-SvrG
z9_Hz?dDPVIjV23{>IKLV1~<`vKnz#fTSJ`rrn`);mfmH`KjTkXXPmHBjoEAo?aW3t
zsbcA0)NcvlA%1y;?+D}d1LdclxU5UDq<g+oFA<)oJ8_)t!O}*y^(W5<VAx6ez&6Q?
znM2^YUo&vTmf+l~90Vu*<4n~<rR=&vJ{m!6X|(Qa)!8q0IQ_3Xgm9>Uro(OtqKB#c
z`b`yW!XkM2#I7~Kq}Ra9kG{s@5Sqz}P8w~d3;`Gf&qE^LYE3^`J?Dx=lV?_mGGh8h
zLt(_hSuNniyxmR8ctgFVKm$y`BG8Y)RpCWl8*lVp{ol|D*s}uOX45gcJ`?)S8ER|U
zNUwCNXlgtL_QwZ!pp!)s(3t`YY?vR8>(HM$J*2AmPIpe}IX>dFDJ)ZE%RnY9tR-2Q
zxdQKed34xw;^gCx!oI&zioBah`RJ9zGJl5Dw$M~tKptRKO-cj^#yqaA-sX~NnCw*T
zs+Wu@lS!{0r_GHe@p7wmuG<u_jUEH!V6}(%PUpyJR_;2402!O@OZ&T=jy*IjWQ%!4
zEHa3J4k`hzCd^@UNmGWpjr9AK)_9{_HO2Ky5M;K)3KEmHmas=E8i${4U5Sa5F^WpU
zPh7XY@x;T59kdkBYmJ*bW<1vVxWwE*^j%n3n0V}|-fv`jMsear&{Esv{ccg=@{k}3
za+ocy+0Ug_zkI}lFYU#-dq%1X->^@<x}4KOuw{A7)S4FjZH*c~O5Kl<z2E9__8t1f
z_aVc0=Kv(cM)(r$m;VFetN~owKZc<f5B$$B{$KF!6c3{h$dj31q(>WDae;bYw*4J9
z*RXHsT?X2W$EwRXFG;Rhj_e};{%9ThpG`g>c5=*7>LlC0(poOgSBGaPSSC;jv>I0K
zQiQg5zfI<_OLs8JFL)+~WdRvX3-X`tlIAST8$e8ak;c|8vEPKX*C9QvOqXi70Dw1(
zdy@8|wJK~=kGx_pkzU4bW4X{qBFF?k9+N6pWCldpRNTANXw#i-NS13YWTyUJ%(exU
z_l01sqMl$%D$C1uKc8z?f7B*#Uf6S>cmFhs4xy0t0=npnog{jYO9fG~3IN)Dv1?~T
zs>#)r!R`&7K4p772`EmtA#8op1P!sRf?w@@#yGtZ;~&k?3c_2T7|we^es<k#NJ?eQ
zSRCCGX|vfJ*~|7{oYm#uivbX3<z(T>z~Xm?ZT@q<s=c?}PKqh1?<m{t*14pzt&fCQ
z-7~6`5aRQ<K9XzNH;1pmquZ&EK6xU%79bAE)Z<Yo_C8^80WCjf%MjUeYz@xsUdO5(
z#yp4cO0%}?#O;TNSflLLlCX(<PFwj}^X1?Py8k_VC7VT11H;wg^x&=^pJ=^j7d`tS
z<f?p4@Xl{TM%z!1Q-9u!&KJAhcV)Ln*=j9rwex^^|2s~&edZdN1lKbi3$P5C0FBxO
zr8WG&dC>nB{5`53NKrX{zLQeOH{6e>xT9Wbn0hlqGeX#<Y2U?{a?h74R>*A7M&}i}
zibWaZ&vwO$q`ba;I(9`0MBu3{cJgcf$`-7_KapKR+b-I9`&%(@0+NN_q!mxjZ|Dc7
zu6aG0uR2k!vVB}{xuGqc=u*HSXLt3Gk?>!KkN}6|<2qK#1*6Rg3N)Ih<+~YEd}!yo
zp4kwR&F?gF%O;s3wzyf-Pcvi0Ac*VuD{=FKl`Pkh*7pPRzOxS4Ph06shTuY^90AAx
zP&vGt>ew#cE;tAJkoJSLk9L2~wv!wgWUgKkuQxW-Bvgr@1%`X)8~2P83BN#2q&t*?
zZJo;F#zJrCGY%Z=K5$~9QifQ{-5z{f$CCV2g6I&I0wQpbAs(h^br{$;Ls&nYA6m)M
zRP>18FMkT*eJY@$LVwPT2nMbIAhUoh-46&B^y;_)u>ld~TD}@j8;-^zJ(uo{tfwyb
z5rncnk$C)adabUkkf5ycz@nzX?_evIh=!>+{b$*)&R?=#4@%(K3eLMRMMsB`?CQlv
z51X>`ANN4g^?{*}LGq#ih#JR})%?^N$LPf9A;a>m-1B|fwp1k>y7Xr>jOTZ@Be`<z
z<-Xiv_oJlYC-4Sen*oK+M)n)8Ke*zr1VFAB6|zG^p59c`=V;+)CY_$TTu-xRajFK#
zJv%EB>S1kwod&|zcM110-MNy+FmJh7x_vbsFuC}33C^}@uL0>%+AO|S-P!v2r2++9
z5p47O#VPooKJ&&t#fz`!Fjr#1Z6%ojVaz%W*Qr}{LP15nRib$PA33Uddp$;p-SPS>
zf;7JvE0;Rd*Nb1a=GbT4U3B*@1KAzX+5$yRzhKgVEZQ;04w+Pr{hBS+<gBmR#8{Zp
z`jiOSJqmudImQBU3EJ0bX0C-!K>kK?eP~zK{TiOVw$6tGgs9tX5Oh?R(apof-#!@^
zEe?qKmqEFiQ6s8JoTtF*N6y2%8~7oWe}Vmq5M%P0zS6BRCcJIGQVV?j&niniAgl0O
z*<Tv{|DQjyg4j@;W+86u_0(COVQ>91bF-2BNKT20db;|XT9GeA%O1aWAQ4AwoeCt%
zYx@E~H6^1<?d%D0Z^Gwi)z9@d=^xlh(*OFS4LaR5?b!b$TY*vec{_dhQAyumMB)sW
zy0GGYF6Dq>5{|@y|H^gsk41)>HQmu0v}G|26A%l!7LzR~xgDakuG8a@A6^0SrR*;V
zzhT8d0RmBsjUI_}+1V6)l!1Af5flQ_^NfkVg||EHc{q9{1Qew?%k*8Cm(cYvk$XE4
z8yv_$o+Ys4i<1?nK=KMh)cWNC+M}*^atM=g#+kk>Sz%!`^D#N3P!jy(7x#`P0H%nz
zvlrGH!k3qgA9$1_`;6(o@=kuMUiaoz<7nj6A)IBS>#!SAh-T$agVrNTCJ*eCP2M{}
zYa!`6--b>qmqCP;LFJIYR`_3bt4VYY$fbWkF*ef%KlK}9mt_PwgufCHy9!m~G4$OI
za@vmN0ZSMlPJ<ZMG@%Ald#pMyoV?sq=)jd=s%+czHBCyxCMRcZ<0;olTCxP%x=FPz
zB~TTKdY>A@IqeZ@U(-#rgE>%%d9tPW(c6@Aq5F3i`g#Nl7l++;)DMjmZk1-B(1h=^
ziCM*+aO3Hm)D_+TQc!M#PiYJ=6mdl(%+l@fQX`EA2hW<D)Uu*pg!EYJdBUl-v)hJD
zeoK$$1idS7`&<uf(F7EK{e9W3<@USRrKK6J3CkSR0|!;~RA8a|h6C{{{$#0c2aZ7a
zm+X%~+oS$8jz07CB4<m9M=Ar7fBDNG7K{SctqxNf@9K4_P4(+b-6pF&>)04mXBYdQ
zxX=*Eu0sHWmSer^j|Af@0+rm>9*Idg(_Vcv7#qOLdcK~wpAzP~*0vc7`fp>1ycv)h
z!toPvBQYb!(0<>@Uc`~QZau@WoT<4r-uNR8FH3P+Xri3cXB_#FPxmq{<^}Q~4-J7*
zo0%wYd^Y~Z+tqqUV?X62nTSQ2wXNM0*X1GB->Xe8*E9Kiw|q&3VXlg#y86GLwT8}@
zoMWQFjC|BmX4EE5@Bil9wvbv}#D~BNbqJSg0Zn1x{2FTB5%C%$Q1o>U_f3OGVAJ7d
z>|s>y%a0pC<cW|w011EZ&NidM0FUcFjB&TYzh&3mG1I-%8Qt5ET=kW`UIn|THFMfW
z`5Ft@Kh_T!VPiLUS$U@O+BttCl`^aZ=inSMB~ig6VJ`-?O_M&nQ#CIgSFl4xyrv90
z)6_7U3^a6I)5yAic|UdfK|jeo1NI1?A?or`RdA^oAl=(cMfWoH)C?RCo~MZbDAm$*
zJ=H%OdpP&SLA9sCk_r+HRr9g$X8ArO%1$j!4Hl<#1}1j-aoYcraoa`p{$D^2g!=~E
z@RZ*a-Yk4H(58Ex$r=#nKr5#o*r~nQeF86s02B<?1yd|fhRd|$ziKzgfqfZW9ecK@
z(kI^)GFHfmg^!c`Yd^vp+4kMYo>6(Uw0yCoE}@z()le4iTD!3P?BqO->>m(ayWp?Y
zXCDe~;~X-~bU57HJW}zWCHhCw@7G<1D$+bH$05+P3kp${2Z4|k&2`dKCRbN|Es9Z?
zqLv?P-p#&`JPA%|WsI1}=Ep=~V|#ogT|+GQ$H(iB`F|F8-D)X^*gNxo4ny*+HpJk&
z^)g1h(s>L5hk($u8y(a)GQO^z@(*viJ8Gw01*vE^#mjQ6$b@ejM`}F|2IBN16lGp=
zI`H02p&gH>eUt3?CFa{}Cx7Mq($eP-Yg%G#W3Kfhm2Ch~S0AHcXrPEkrBfmxI;FCT
zzIEKR9ZSr>oqvr7_@`nqW1ZSy+!qO7&$&~l$;UOn2Ooeg`{bYg0Utu|2O8n@;a3tJ
zeP7}{zG&RnaKb4aucO)Tz;-QyP*!1zBSD{zGS^09FhZ_V>J6>6O*KgC$PL4XCXyDw
zd`dg-oAz|-O8t|$17mOVqLEo?Ca3eCzg22?0`BmDkFl{kH{3Yk!%moJ$2gp6#K^>x
z1}~IPbFf|p$mwQJbox+ut9|-9wqreG=id-c2}+AV7E~Etr7rz@l^qR;HU8IKwF%?I
z$s_t<6dgy`d&efK?svaW#DDO*{{h`Nz`$%bZK(P8<^$jf;D6C;xX}&1^zyImxGmZw
zPf;2YS@e!Wd%5DEt3bicc|+vf9d_XF26#L^{6D<C2T+q;+btXfX;MT$Iza(ZiYQ$O
z1VunVib#<f1w=8@rI%0y6zK>kEg&i&CDMyP=%Dl>T_6OdOF|70;@Q66^S<-V`Dgw)
z|IGRR8KX0f-XuHsz4yA-wXU^{=)PeI-E+GtsGnkh-!d;;L($8blcZR~T;wfEo>mfE
z%<m*Wo)o&gAsnVYnP#V@MduzfUNjy>nYi1>+A4qR{WBdedSg>#buY#ZP^f<T`xqyI
z)JoD_O7!mtphUF;Fg5xxAGZ|wv_W9LPx|`sFTV{bVHHQu<a3Uf`?`v*!efIS$Csh&
zU)|*gn!yKXY{y7E&%V;%@xrt=#ndC<B!p@L0*33J@bwf0NLXBUeH1-Ad$sOoO6z%6
zW5W|4G1P!XNtKT9_tzADr(bk4>7*SsCPfx(Dh2-<Qc-oKu%j4Jg9Eu>w5C;>B3_Nb
zNL3!`T*mx<P+cAK@Ymbc^D+2T7z=W%zbT$n5#S@eUPjpOt9#opIV94P+-NE~#ah74
zE$|trXI9}PeKJ@o4K?uL!s)T3Pm`n4C34f(F7NUcNJ-gu9^@xzy-I@g2-50M6EQso
z63d&h`!%w`{Xnn_`}z|lfBGBG*SQz91Et_%4n$0Z&3E?pdSO|9i1dO8`e?Nd8V_fc
zLqmNXCy>0u%c#k%gTy(#%Jz(IV}99-*XBWreOx2!Qn$$I>n#HGnD_9<89i@-4)G$p
z^?(@KA*rYF>o&htSEQ@4qM}%~VPOufC1t>tBXyM<_&q;;JmTWkWPRtRz?Q%C=73lY
z-3r3oLi+<rug%P?{D;QFk4^+lB!A6a8ph~c4p{|SU(SOlgjVpd&wdS*p^5F)t1mA#
z`_nIr%?T@J2!+kD9KdA*H@`0Z0{R%Wq8eg+q|2kYQz*9O;XVnPTf>JJ9@)7;sLFLz
zi@dQCp%UGg5olC$pVvGq(~OPowzb|zD;<%ftB&4RFG=bwCi(S2p8a3-nf<F!|37^B
zZgyfr$I$Fv8B(V5Z{><6&K<#`)+zqoAxX7-<4z|tz@+O`Er2z_J!a`*FDo|;?+|aM
zM>M`~^?aJDc0DmIg^%3X{3FbFk;#3mDd}B^Ocl^VWkXK`%k(u~mp5WReg>7VoZG${
zvoWLPo|x$$UGl-VtBto~)xt?lc*wPM=FDdh8zS+UKU2Ky0w4xI|MVs`+{ua-eXe4X
z&81Qiue<FUQK3WEj$v#tAlB@a6wKt)(VEf(3ni*U@00CGbPqxBb>f=$;Y__g{^t2w
z+UX63%QG}Wt~T$P!oSkR(XzNTP#?)y_&(~Ga;YisP14tSdco>V$?!mTZ-K@+sb7ka
z5cjzM0#jnv#LE#mauOJ%rQ;h%ns#ZIyr|%-SzliC%P9MgDl5JsE0)@1*l@M;s^vX9
zT^HHEcPaCt5PQ|VOpj~(n3wP;bsy63-nB{Hy=B@@6C*{QBhK+r4?cWztW#<@zc}1N
zv~Ss9isuusoMg}v*RA|ithD}jsAZ7uJ2;t8kG{a6q<X|@%e8VRb94Gv9<j*1h{Ti|
zMXXug<WE%mOFa0@A(HcvyoIlJ$DZ;A5LJB`?MwFid|e=avMZf_s&{%?$a5>k0iSt-
zb<o6!uI~>-K<%`zp$)Mzy+0rSL#0izrc9#8HPGJv(w)!Bzf<VBjg$p7%}7!N&sJsa
zxRNR)>kD4jm!<1QeVkb@V=eg31$ItLB;|Qu4eenGfN~w@KTI-=ofSEwaz!h`ggq<o
zU(hZ2;5&5kmv%f}I)Yr|@WaDoL#g}XBnW^w8n+@MqMq@+t{Io#5Yw-HwtZHjryJ*J
zomfjNl$f9u6npt42&7#bKXW~It={Z|9I_E#(Gcfo+Anh>NqxYdB)j)2RVP@^H?P+_
zsD`nqb_t?YOFdN_+&lc0t8<f?{>N*T-WaM*+#;543z`B64*N?Fzbq!zwD}86RE?3Z
zzjqqfqzU8uGJ*4WpTLK_qBOs~VZ0nHI#{*$_L%_k_UP1Urgz^~uSGvn`cawrhz@j&
z0?3cjb(*An8-MwD^L}VO(W^66YO1Xo0-_mg6xT6wWr1#k#+L-n-urE=Uaao5CGr${
zC*^Lu5!fV0DUyL41m~r=Dv2ZXi+b~@%hEGHdL9JYJz`l*b+30&xwnWHY`}*OQ2-yu
z5GaMip80`HH;k2Rxa)`Hjy?G@MzosHoW>x$_?d<&LCJ-c`@nNyy?lUZ)yWKt7iPBI
zpz;>ku&^)*g8Y3qHS2FB6Uj~^W{M9jq)IuPm6%}VZb5z!3J2Bxzf!&d6$TX{t<4*T
zrFBfbb$vNy5IL;(D@$$vo6~(s>eD)JSy((|>RJT+(KnB?)wq3KTlNpLG4Fh<N80y2
zH21X9dvoGwMDG<cZf>6e5l7cYN&=l0-oDRhPp?e+0xR*k<Y=E{u>Dnj166jrob}~Y
z!}W=qQ?R(N*$nJAsDwvd;BSGoWo1z_+!=8onr?m^2@b^}_HpcZNbnSPOHU4S#MA+L
zhhjOb(FLGaW_buu-V@RfmO|%8IOy2{4&|8p8V`Jzp3Gc?NE#RlKDzJwqxonq0DqX?
zPgnmXa+d9_sN1VuVAyw<rD<346CKeY;d=QCin*#WrCbEe9EH>|(uwSfQ2a3P5%^=*
zD8+FP^{MviF9<swcjxo(Gk0mILn(WpQ&*Mo6Ewuu4mip!TSR_pyM3~Pe(;@Jk9QFz
zx%w3uY)b8YDAbql9ai|27)h%`{sTFKZ1^7NJRfVX-2|G1&)eFjFPF@o6*i6%QcgrS
zZd%i=2L&<Bo%$_lgjV=QF-br*Msnw{dSLiSg#<2b;vdMXlHC<~y+$>&iL}Jd(Dvm9
z=I7|dvQD?Q(5>Bp@*yAtml7Y*7%Hx8REA&f9~O-i?RU$u-!6`Sk<By$3%~`$vRZ7P
zW#Dnpu@izBOKVw&$$sVEyHIa=k-Arnk$mnm3Bi2SQP5ZTgzJl$&J_I8jZ31Ycf|z!
zM1b2b;C&#0=p#3$YT_)(%GgM=mpDLu$a&*=F#H-Ob!`<s9eyWLZ<BGQsF%>?u5u*t
zni>q^$Z!3GaPH-=n*t8D#^f&ti#<HgNGko`@f>hI`M0vs#7DNn*;R61ri+Q~lAlOo
z*r$s;BFW7VpF)Xl&|9q&iJ#9dxAVR+JY6{Dp778}dpJZznsjd#h-P!U+E64pK<r1C
zK3_en_fbHnbfYu-+|l)V8O-bR(_ezgc$aJa?s3Si;imZ7FK7yP>-e+b1TUz>Cc18^
zY<dR}r}=Q49hfPKf-(aHG|Wm7FAOv%l*Z!cqV~-!OW##{^POC@4C`*P7BxwEo#b79
zKXXflVbM){i8<)pf>qu0-c`bo(+8DZc0P~*djZWD6#2b*Pw+KW7H~vR0g!FSf()dF
zfh~U^U*&gw!k^hdn{8>zN1R>Hs6q{%*L|sbeNrfCQ~2uq35$%uN5fdTqUCt*S!>{5
z$?q#t)!z4OwdB+<jqUS|v)|Qno?5enKmKd61tm)&{*EBw!H_sF7;)+_19tDUD}vVu
zB4f(G3eruh8eaJb9DQb>4l$8+2r*}Aqls%oZhvfwBkq203+|ewmC=`Q9eu99jeg6u
zp<;3vTIHtV4Boajg~K%rzFa}_q#S^SO+qe9^9J${ku?bew+UX28-1^pADC6tyD!On
z?iL*v=R21XcMWFGjTx9aW+zceCX@LHdHoy0$H{z|i}SLOYda7|2=U=@Vy9N%(_5Rd
z5`TeoexL~nfon9xd1vI>uIc4WJbri8xxhyG^{aQm+nyIyLc8vKg=(z5;oOees1kWO
zzsahVq};jRYe9u8<LT56G@?UWk(me@GdJ_`Vo~pznxDp~^VBWFhgspF7!)a);HgVQ
zhgQ0>v<1263kaN1;frnZZZ|bOH!kdLs_k~YWtVPyU;faLkZO*`xjujRc1j1TQ|0J=
zP4c#0lA))%AOqccB5xlLy3>UadGfnlb8%0$`h~mUaU{5gZ6Q%(ek>UuPyg6oge%!8
z{+-t6dow4u!)|hN<Xw93<s4UL82w@RDi|E=F~gSMmhfb&C%#UyilV(4uTS4ww@53K
zz@jpb9~>X?*N<pQ`Czd!rfWPl%u(}FN94ICW1d)YGbbSW{_6O|BdbcB{5jwKK>DYo
zW<2c|w!nzpGAg5qU}|TYXOFixUt2^?dpp5v$X8kRwls?|6)=}sTK6^Jq&Pv^Y*pNM
zx11mqU9YQuTdmEc9CP{iY^a7q`p6hC(!5@!m*si4Y{dCTJ5uUr>y4NRvB#3?VCFx7
zm)smXb&V76sQsU&`-|hs4~Y#{xPn=22OGK1#_b`U8p-i@Lk}1shiQhyf#+2c6U955
z9FNU?JO#>HXmgq3mIg0AP2MSu#!8+`z2&I$z$fnBgp4x3Od{}niNxMr{-2LV2eg;0
z4UslBN0TG{cS-(A**D$J@iYc{59QF>7z!sk-*>4s{lmJMzw66Nl`nM$!eOUK?-ToF
zTO+^D^B!pIefNfQ$i0{Jz8ueUDw|4HKRLb`3`S_QxeSq_^n52%##Ybs7|^S9K!S6a
z_IEzK=ckHudQ9CFe1*(7q9%geJ8b#xBG=O%)t%vR!}K-vR^}On#3WVH&QcZuvEKEb
z6wsmUQs#yEf%p*9<5ms&b8NI;*Q`I=xzW=J5D_=-s6*!M{_-7XNmT?oT~hdhC+0$l
z`-j;c-VfdPp7&fb;=FePWNSi{=I7m5biR>LM{+BIoJ1FNsm)rN^4qcmL*}}W8%<9#
z^R6=KK(^y#)CYLz<_dC{f;>Capo5Na0-iR*m6HxO-Dzp=t-Z<)w8H1#4K+N#HM$U{
z9?uF5%b!a4&})`f)_bH4GQZuse>Fjm3qJUe@C)+ZB0|-S3^?t{cS?epNhtPvC~ASU
z1#reiml=D}kIv)6+_ZnHL$w24Nrv-T0i8)mUTo}opkjgT%@X;>(X!y3f++!w`h+Vv
z?p;vJ62c=KSLjYWztxj;MBI9ycF;pUedZ6vWe=@>Z{Ck<5B-wQpSYtYe5BvKNyN72
zPQEY;HDMq0V<U*8*9q%ro`%b8lRnl7ZtjU?z@GIZU{CtfL9m6D52mBUdZof9LiVQ6
znIa`_Nrne8AxgUgC3=L4J--nc!@^onu>KSL-P&rc(^~gWe^O!o5XZgrQxaq^;{O@9
z_)0|1Z-xMGfda&w@9(a*w6v}bSqmF|HkFC2e#P!$YC-pbYrylve%`PZebN`zF;Mcj
z>1qX0A=in4O&clcTQkd;20p}3ojcjrVC-RP_etHGVjb~Zl`*|-rghfI*+A?@f}He%
zmN&$q`2*Do)<3Oy6jL=n*z;py#Rx%LLsVTC&@EWVtnbp@rbgm%R8;AvgW}zLGd*lT
zr~T_elcvyB(X!v}osea<6J$xkqo}|;?#0W&SE|69aAs+@vap|afeKa7cz+cR&17wC
zQ4;=%e&qqfa>DF}RaBVmHm=)e)AU2gkP}~ACzSz2SIcC3!RqYBP2I*SnF(!Ygd8>f
z(_3TBB<meXuJC?rM4=um@D-g3cx!1|t<Z{)tP!D=WZ>xsKJpIP`4yzAguD8rhlO*i
zcb?o`HQ&E-oaNw4{t&0y^5!JXVp7pp%BH~f8jx(+QXM@3C!^#rc3qZ>@*=yD(m=fj
z3Y{d24?smGeaxl`2{iqy>oS*f*23TVzFsA9K5Vt$q4q_6q|I_m<n{TOxD<nE$V!7N
z9?j-EXN1JEts4Ijn7pOk!(rPU9!fs;K)wZgsRZx?dW3#Sjhl;Qs`))Pksd}bt$wol
z{!dYemqhYfn*y!7tp$*B<WE_*_#rv_-VYhNibuuWdm~ey{hY0ZQ!0>+;EE5fArNV^
z+|X1049;^86F(EOe<`f9x_PKx(Ggoi(U9*G%wx_Ou5UQ<1$<m{^1mq)z5`SIuERmK
z%yIWz>a_(6p+u3`NDbi=tP7XY%)Q6lqrTKe(uh%gaLe14qmq9}%0r9<O6S_F9{%_?
zeQ)}lkLT#2c-*=_)y!zRTFr+w9f=j**sQJgY5wWn@{_I_;Vqa~a3tb1(g(}FDQ9!z
zk&T$)y{T^}Y0k0omL(s5{Q4k!qIpwY(31%7G!IrKe$2(#3cWCBITxXG{^IKg=E>vK
zVxviJve7>;>5Kk<1#afjM3pv6#48x<Cnx2Nfr)j8_+DfZj^-rGPY`uVbHcO3i*pfC
zWmO)?F{V()lRCrC@4f_p_yU~pzDWQzgcw3?{shi`f*8LV(jD9n?fDLDxRkU*hQ4n*
zVU<nEV>r3t_7yji<gB_*7OG|B_FmKO@#`$~d}(sy+cLR%hb8J>my6V(Ou{7<gE%R`
zKs28Ov)w7H_2`yk#v`0)O-RGho8k?1-^-!Buhf3opA(>$>`PgR6tWLf6$i~U;Ffq{
zfD(BWPlg~U*KCVQ$pQpQA>rLXbHV;;i-`p*tAvjCae-wriJ1xuTMzvMm$;jShnI6$
zL0cJ@R03ai{pV*ctc}1t2{o-FfzWC~*6(*yW{Dr3rt=-WjQahauR=KWoWbVREhVTT
zcXQV#^H9pv8JWP#BPWum*VrWP<AuMznWqQ^n^64(?)Cp#nfpWVe^#cXK{R)Q<l1KG
z68}C;LBXx~j%?H3gX4vOBoAMp5ti6y#NNa{IuHaoHylnpZ=1Slbq1(jfcs5;2`bRQ
zmE$|mvcmOJ{>9F%k_#4gqoRpsN^i*Tuh>!1@O#lIb>;5(A7J;U6USTC_+#%q4_KO!
zZE1A}(O!|^^YLvT)8*k@&-`SXpXx}pDXP+K8UlP(x+5hsesk!lT;ny}$;swaLeW(_
zHWTqJe}cr;GCzM@6~I*~vxX5(r~U3#WL+3dGSE5xwwuFJ{&x#0SmW({`_3|`2-th=
z9NDzM*6PMYX8f6Z)%$EQsVEoi2hgxVSsHu)4$l_h?v}0vl&reou`BUZON}Dib?1cv
z);?JA872H^uA{dY1i~khJXH^d&gBn%=EtOQ%*?OOUg$ibb4_d?IToTI(FMk&4Wnkc
zi4mnK0Sks~Ex4~GHo4Ki7jBv)-gD@RddSn=EwrZ^5O6Z^)sd?T^fbv%wXb4*M4%3@
zl9uT|c;YdF8kjE@PioC^$Ti-Q`^22du3m09Fk>ttCYDbvzlyD7ZHs3lm8(h^5xHIp
z5q@+uIp@{YrqFog_B?GXk8R$f2SIhj_~+k2IOLh)e4+3#<{-AC$6wH#q=%=jUjO(T
zk}cyPgdPN6ejW}RI>&NZ{)j$f!S-!YWQyklS2^*0==zEBB-wS@mAI<izF>1r;g0h<
z?B%zf%amtw0zo9!<njM6P^umK*bw3?+@5ymShz69mLG*NlYN>Di>4ul-Mc)dG+)6V
zzv-EMbt(81Nt5u&kaHz$x@+jkHbc0!U2<onH-E&p%w2)HLvxiQr!pu<WtGVMV<JCl
z+Hb8r5ay0qdCt$~#U*@9RENHA5+*u^;?TmBGt9PK2C_@vxMXSO;%W!19ZZ>p7o@|h
zva|%?y2Q&VJ^@$HE7{c=_p5`mOX4T%v>cjbxu;RHX97b_>ma)WnK}ZIIh-QU4SHqh
z;p0lG!)pt0F5iGQ8TD+Tv8T_LGS9wraiq6qatH~2SGYWPI#9F(q)$2to*&+n7VYnT
zP;2Jh{IwaA07hjmh%FqQFgW>~8<<HH^#${sQ4=BErHLnHSU!XAOg^UG(QXgE0wfC$
z!=z6m&3K<BhYfw`{T30js;*l1sa`CeELysH=|P_FE4{`>H{CYo)`}|vZYtxQe;`%4
z1D!$YkA$q$)zCtve7I{8M8lQ|5mfsxu&-$YHRJ2QbQK=X5e#{3t4h3GzXR-9&;iQg
zGxQ-|uRleYwFQk-Uc|?#dZq_Mjx(RJ(vv>*SUZ~?8Kr)0=zfrBCVz${Mt947K(KZo
zcV8TM78=GunbibWXXTSkh^#TfV8in{1N#gs2>g-vIPhvWP8}bgwq_K_LwZ11G9Z30
z&CB)emqHm=YNhj;q|lBu&qZ8Qq9Kl;pAh~)rjQi9G%z|a3ra;D{3mz?o|E(yXYErD
zECYE%)HaB8ae%>^#yu)sJ9Lhiye^guTg6l^(FwS>d?0MblCLi4)l?07{ZKKgyE!{#
z)yorhh#ziZKYqso2)%qt6W1SzW{JgdI()H)WKXHr2SkuV4r{F!;4DezZ8bo!KMuS8
zal-Z~>*9cN(MeSnIAd~Kwj5#nrOW&3LHRzoQEjYGgZDPwLu<T}K+$}B9CqtoATTRL
zY@w8=j??bN$7S;}-s-NSn+Zw>+7jU|bcIh;gXu!XhKqpNkwMa)hTLH1J58Cgo-0z_
zUzY^p+(Sd;g3kt?7$j3IJRG+lb(QCv$aARK4$*w^S-daesJ&1a;JimP0iwV<k*-5J
z0?zas=3Y4aBwOj#4eXbMnHdXMS(yd&m*YX4)`Dkzuu^l$W+<EmMx9H_#+wobqDR^-
zQyNrEjZIISd>YnavYprkV=cY9)zy(IO*HHj$@~Lp|3!E--R}>$U5730&e$s_)D`x5
z_K!!5=q8oN+ga(C{hy~QV_)oiQRyZ&@syg{c*w}ec32YTVqmHZeq!-$hkcRn2#{f^
z%5OiZ`WdLf(b}`{_6Je1qwX4fYTsThum7p}J4*M>_?W992Fx)BfHn>4`-{T<zW$FP
z+rR(AKYyNVA9cNR;={(bf@PXw$j80_FXfHwRX_8gMUmhtvuMZ!Vc<m->eOoDsxob^
z`vdRi4{%Ric>1}YJy`c(eoEFO0+mC%faAF7_l=ErjnUtov*u|S6jD`Qh{ZP!nnN`q
zv`Nt*3Z)8ehdB_f&p(cQ2#Cbf<{e)D4>}5R_@C(cx=kp_pska_(KP6F_x$usL0kKT
zgPTXOs++nJjCUC*WYe*@<VwRso+{}hBCutDUMawZ!-VFJysJCy<R9n^C`4m>k*xl8
zLu!$1WfqXX6B9hnWIq36mVu`s#C`5p4hzVl#D-v(U4~GEOQ_Czzp5++0glye#Po&K
z^V_i_w`c-X6n?+K^wb}Uwa3j7I&HZQu)jolNPgo`oH|qoq>@Y!xQ>~pq$ww*N#lk-
zAQU`Z_Q|;U-jEZL`}Gf4i1ueZ^Ouh-+DnKuuN`Pi*1)>!d|-uVQ9Xrf^9ua>ff3yZ
zf>1@K2Aw(YSHQknk0K9dPia=*>@cBpo87jOFSpXX$IUjfim9l&98X$A!R{{&j{iDX
z9A8#kpYQ?5w{GpIP@AHa6at&0@TdUCliW08V8O#SJ}AgJSln<Gx5mt+Yi=cR@%f!7
zyIvro+Z7WdmgrujLcRdyEB-bqB}XIwiRS=PwBg<3%jXtzm)o4?!<6<)X8{T;gF9@{
zD5nuG0Q;a&Cj0Y^llg;WW?X<$>88}q55d-p8+J*n&bcO#PhZ|_xk-38*&o+>9>2m}
zVd$E=%mSakq%Jtb)nRx-UtAOV>X~EWgm~z)hUAe(@9@NqU^DX-!`4V?!k3oyE|KLu
zhvYvH>oXzBpMY2ZEXAp3n3q#Q&y`a;IzIHhs<`gXDyo6^Pl28G#?*F=tAzpzbk6zc
z;$sF+_JIW{-Fv1jzU{X@jKgB~x;GJ;7M5-H4_%#}WJt}&+QrHke7_SS;|F6*2h~Ow
zI4`M58o=|;bn{i!?x1LmcQjvGqIv>bQifIdAb%Wk0=bBo4n&h~1zwiznsfwQ-dG6x
zbrPKL4sTihigH$xH}b$#^IZ=w&%x1KI$>3%UqBwY-_i(OxVy5?mFA<s1)|<8a@%NQ
zlDt!3ZU)!+cl&+2mbS|7CL#*QHtMG?>>8nTAqRNf;}qC6=Yr|Jpb{XwAF6ed`X*^-
z1b_bd8sx}x2lH<Ee_rr)g>1U8LEY*$-Ls%-Hg2F7DZ!OF0pvc-{Fx38d)>RmL1GK&
zF!><eX~U5Qz5D?`w*{qlOxL|tA>FG1=DmpO8uIHO!ju-^^0d5C)La6`|Ay|59}Lnh
zvv#1vE$P~ej`WTV7Pk5iXQhSvs~@Dri-sjM$DO*`y_9Sn9DXECx~n%Px%qK)S8uc?
zL1)iTSNs}(h;E2`rXX{`1_u>xu97Q^BDm)NF{8kDyj?06ad3hAhlr75tbv<J;rv-x
zj8`c;Zokd-IWu11Qsn@j-ly5FD@lCU1dMdozQsh)74*yYU!ARkPG7+<bfmIq@0q}(
zHbpV#2!NW69tXc+NbnuILAvoj5NHD%VbgMa=UBVbRwVhxs6$B3)*??92SGQ%Jti?*
z7X9R>8ZTLqqyvrJDcnKv^b|cY(!EeqnWSm7CEc4JK?CyoU<(bP3j)aHq^3WRT^uW=
z&lXAOq3CKyz}Ka~7*bwph#HuVBdD~^otFPTChdYL#zfQqL^>LaQvY7Oq9o+rDFJPn
zt3XANWf!cNom_V%CW}8<EJS-7W!q&I3FHXU)A4Nl<v>KGI6%t6LcAc!s<kH3(l^Sg
z;&WXJtUcQNAbN?Xd838=3Cc{?B1FCPx|;vk@;LVA_BpRGMU^)!KzY$+%`&N4v@kan
zPtn7?#BNLjkYHE*H<w|{yg?Kk-rU$MU?)!e(hO(vg7T^5EnqmBqKa&8mPlUxmfs6;
zW-%%3Vx+^FHZhaoB=d*(4CEX$klTBjRcoc5@)I@U*trZb{R+ue-HwxJgFB%Qq$n4F
ztb+AeEUu6;fZDPkUjj!x>PXm#4jbr5LJ`XMJNFf5#faAFK#b3;CtvANIO9(zRCM|l
zb^KG1g~{N{iVH^L?`@K$7T(JD*-l;X%}Fv+aBv1cXMiT|7`Zez#YfRIjK;1lqn1yZ
z!5z^DmjwYzDNP1*f)}Y*HU)zR1i6V(FMscT#dt~Zy($-cyfLX;vG4j;hX)^*g2Z8(
z4Vise+soEcebOU#cP{qOO<8!iTj*9mavU~9JlAX+H6b3s!lZZ{EAtcQd}OvzSw)j3
z%ix!!Z>+*X)G={x0c>k`N#<R_+{A~YW3vKd>%ejRZf}xO#xK);+3;y4a#u+pH8Gx|
z+Z6+D2Y{9GEbCE|4tP8|3iYGz4`dkdl;gqTfhr_@#W;D1Sf7sJ$LU#Z$M1d(Uo#C!
zq&Pn16gk@uBrt=4CkbM0HWKao%4y+b-NU^iSC$iB#luf1Xk0PMyEgBz$l7lHZH^PD
zSPFf+vRTUf1~8kmfJBl*b*v3z*CM(o(NF&Rdc2%tX=3Za$C}|Sh_CClKEOUAGn5w1
zP6?BH{fxDA`ei$jpGeg6=qO(87p_%MTHn~+AG#NBKxMAKVSD+N5z_BQYv3O5f1tSV
z0&Lr93$v(S(mvp4WjC_!frl%pgZ0>VfE)n{axNho5e%X}8RkD}68T)+6cUko)=roI
z5gj!~J`ZQeOLFZfacklqlai;afqeU1BhIQ3J{?V7B~F6pbgmt?M7f{{J(Mp3(+uFc
zTu+uLEuN@lcO$xK0K*}i^K;c+l@;WR<w+uvsfU#^x3md`(WYO?7TS2*ZP2obCrdfu
z4e3`cnYW)t|CTAb{G(HhzbG^@lb_$gi0+pSXB}Sp)M_^C`0r;y9y4t|r`XWAM9;Ru
z-}z<5%KxvU`w|=IiO$)3Y8<m)#W{6$o!-AdFCfH4L4zz*g0;2j2EjFA7r5r_feeDD
zz%|cL>A>FhcUO75<?&e(3Wr3{`fBJ8KFtX9PTX)(*x2UZ8i;dG(&>OaGx<jY!fACE
zTQ$6M^lC$EK?SMuOh!*KHbPTCN8Yg^d=CTGShN^5jzk^vS5g7W=2Ql5>6kg)=hqz0
zEcl!S$*L2}EX`**YKLIyqp#&6_MYgZlMqs6peyS)<i_;t9`deHZnut5TBATwqY9BG
zOSp!ic8p6=n*|_E@JA4h8V|`3K?13jdge(sG+GswHW~^D_a498`ZIcEomcy>BC=EV
zH`l8`TT!tMZG}F*lp3N0{M%Shp#}r2{y@4BWKDmAGD32jGP=#mkfz~o$A`&8(aOZ`
z!0w+t>=z7ak~R~x)a;w_g5ls$UjtIA4>;5z%NpX~6p%<@9Ls7zInx$@^1Aa<SN>!F
zrw?e7yA5Q1=)jWhq1%ZzqW{PdF?#==D}ihR4#+bQ2CxvufILl>C?11pynj3;z1Ngq
zFm1JxVH(k+bg`vrg~qGjw<8z<0%bU{i(_j1e<0)FYa6_f!}<<f!6Q)hKUerP4%5zC
zR=)DRDoA%~Twld|Mui;v<@bDC-QGrD@B{O~KM+3D;tw@0)VMsrwkA|NPXg^ofxkaB
zNsN>m&{{<8e1Z!E%Fh3Ru#>c=TzOXWZdX>nQLWn2<pQe4R;q62W`99i(BV1QL}-BN
z%dMMDY^VuU)Xq&g$PpqEd(!qdC!Fo0TR&~<a&Fy9R%Q@#HRIT@jdGIEhu(|$(Ug)m
z_1-&y#~P<@eV%DCIbvUlX8f0F5Z#eB>gy#Ecu)6H$GJ$*erh42aOCYSj$rW-pY9f}
z%O(z)Oi5SNK7RQ|F6I}M*lELI$jNBNKadG=a3tO=fjoQm_xV~voTZ#uMzNm^U#<8t
zWn-}@vezqPv}?@G7Bjz8;x7eCs$b|M36ky<n0WKDvHi_GBip<lT~0+7`4k4GjCRfx
zn3~!`xPc?Yze<(=tpfSK{daU(fp<w7IP6PB1s}`F#+rs|Ln#-@MHBCMk#*;9%^^C%
zF<XAWYLB6N4+v^mh10gGV^#j2l~5Hk13a<a%iNMrZUGw=um8eduz>|s3x+vRd$a_z
zn})4!2!mhLe<Ys!?CUuzkI(bTF;t`DD!75kq87`ubq_y@_E-6dz7I2rT=ngJ+Ln92
ztWZfwHNdlrb7Rzq?s*`Jkl)6vZT-24_^Zw<?}^6^Q|7#%A7c$qru76ubXxGR-7bD~
zn;{LE9l_@$^Vn8=R9@Bc(#`6zabsug#NdbU)npO21Vt}_wE^r?)&mJpYucU$g7fdc
z(9H0~JfL7~t;QZ+1I)*NS_79z@dSxhKr*u<{WI&cvdAwAWi4(;i%wnP3D5I@P{+8P
zyi0(Cs8KLxMoOBtsz94HeyLK)1v$bF+Uum7#<w<-<NTjt?7;lQGtwLm`aG>k4mrR>
zieA$mD?O)`Dxp2_hVENwTTdWAM8$z6g|!7l1q;;>8yg3n!}J1*k!Hk(gzMf%H})j+
zzUlW;!5oOmaWXL3y>zgB;TD2DqyCGfKvm$tDKdc5$V7dx{{!KF0y*qiplg<9@SSc?
z<NZmV`jEBJj8L-ZjgD)!cV~+2zoLYX%L1^91tjGdHqQ#bvKsmL{qkmx+3Ae0iK#ML
zryvCl^NIcb<0W_0`CX%fFZ#;1)>k;CQ?SlX!3Cv*8$J$nA2&X^-CCl{JnheR7o<|d
zpEo=Rl1ht`{}QlPJJR<!#fHztJ%O7kPTSxuCn3K}<FD4MqXC>pXU4|Yzx;uS4zww)
zQ9$+bz8sk51j`?1k@*$|&e*J(%RU)XG<W`Gkm{*b@Y}yhfIBQ;9zTT%R3dTW;8aMw
z#fZo#MtX?x=6-n|y$8KrE}NfFV!!W7^e)hWbT&Ky>FtUgg6A8@TrSH8NMw9vs?dHU
zXQz;`H4qj1LS)OYtLjOh9O)d+HEIgKdT5d!`(r=VK6dk=)>i1z^o{c|iJb;PWqtm!
zfY@)6I?fE_157Id$!?82>p83In5trdXK0MG2yw8J@$la-K}Ui%$yLJky$q9K(NPzS
zL;N=jX>hIAyHlJdo}HfAnf957;96Ulb_7izo1px6@F<uUtfXh$w*56cra#mr?BRZt
z)*`YPJ99=`K%)8>-5^cq(i@ZKn}Z<Jlsn&je(@7oNK48SBboOYFi?xp<t#oz$HRR7
zK!mX`Q0v0LNj{Ia(REK2+!a-xV9%<<?l?T7+NG@{27vvs#OvdeV3DfGy-9Li9)s|@
z%c^c6j{yAuozRwqeaq}M!gx%_09jbMk~g$^dP$tt?Di5(BxD|F#t@md0DUILn*z9B
zhuV<^W?-DCuXn%0*H1}P;=@AJMu7t}bv7|ST#l93Y3<R1z5?GwvQKL76(aj+9*cR-
z%Cf}QQtU(Yi9UCc#M!mI<*xG=OYOw7+nza49L5`gPz#uwwuzI3DPc__pxFToZ~LHx
zZ%y>ToE8WRH_1@4V^autd{AH#%#@&geDZHv5ehtTc#w$npZ|v?por;!0)ZU-0lB&6
zEzliKqS!G|?1HR;A#gAPn(UB~l>FB|4#>!nKxR$71WxIP2pZI&c{}h*Wu?e!1J6|;
z6OmXjJOYn1nApYSfL=8k9B3(lph~-rID?ua{oj9kqt)p0@zDFVaXQjoN6+sjbJVa!
zxY}1=L7kVFUCv<%Xvt<8gz?-T3+5YnQ2vtmOD3kFw*S=CpJmCT`=wDp>UL5i$j9};
zLwCObIXBDc&fy-(*s41m4rQt#)bjFvhz(5uFeKXN_5eWmaL4E*nq-S*{z7^;M<E05
zvEAG+XN?#2LhhCGU9B(C+lF;EL_@Z3pZi${!{~*!z%Pld)u!3wSomVn3^?12U5`&F
zL-wEiQWIVAr!AY0H5=&|E>^zuDsY=?!j92{8*1bp$6`Rss=Kn*4F`SDgY%YM$KEyI
zZ0}JQ6a#`<^JdmVXce{>wWS(RH4qUDMTgE!#{06j*iCPoAnR{4Z2?PWZ0+%nMgxUC
zOo+ttE6joHDZ?X1=rR8&1l7t;pr%A^xB>{i=zmy*`}OwSV+zx7afA5Ug%=5t)rkny
zn_|7Q10H&3f(D+D`!HDS!wPh0How|O@Qv#}%waYAYZ_i^zn~Z<7Y`*lSK9bF9x0|*
zXp69%z4aav^6i$ek|TfEHTjXD|8{d`|NG|j7mHToBt2?Zqb(<7bu^4k*`$s<ab9ds
zNX*u_a!rIiGkkR155O^S^MSh#4eq*lFt;KE?pOKcCK-}B4$bK+)7GY#Fk`uL7(DcN
zuiWmILX6wJE7$n>!^g+dH^DszF`PuSFs^ppC9*?2OlCB~H`+~?ZLg6`acCZ2bNxYi
zdKak@x$h?BNyGZ(>YsAG!<xewLs+hp4@&o<Aq0BtLllh&(7^ddY8o?W)kZO|yPsA%
zWU~#&Z$^Mht|m&ZkLylYfR*#r8x}f_I6)f7<qk=1=W{0a!w1cSG$Fnm(DJTOe70hc
zcDt<jXM4Vxo1D$6Ea&OyCc-yuH}a>fR`$-0ncb08s2(c5s@2HG@c_7Ni~C>vSKp=-
z)u|=}kY~7*P#saBkOj@O`Lo?xmg}hl&V)U522U+KYJMGYK6rX=OUBop(g>>N!462k
zSQaq@x0i=<efVf0K)bHio&U*Dcj%i;s#GJo1>_)d>+|@k7ls_rL7AtRP)Ttx{?||Y
z`(^&myqF=Lf(pdV$Us)k%Ir(u?aH_$%|y3mwFAA*8J%t3?YzdnC4nANQw)TS?ILYt
z4TvI!ejXHS9jVH}Nc%yTXN(NXG9Y0%Nn!~xjiBhdB!Yd}7l`MceV;c&6Zh$k)_}9_
zP8w*KoQ6rydsVAl`kq&YObUFzHEaIVO9j}Y{MM3Sy+Il$?Av_bc-E)b?ykDg^Yo=?
zbvmbGr+eezMKNwjwgF@@6rn_!$|T<%jyk^n4t^l$PdKJYE?t?$I)Pj722hDc0nkVn
zMj4!ik}QHsSH^AOW50J|O>-Y=JV^7ir?Kz7!~X`793IhcXzkt|c$t80CgMvj?XYq@
z@z2KaXKXmTe$R`2Pc80o|Mq8LA(vHUUa<cHCmTtrzYq$gfRQIX0>?RDD|+uNC}+Tv
z4n>j17tEhGuzR^87i{b#R5G5+hyW?3a}HiatJp-a9j~O+&wPha(+NG{n{OGVGbB%1
z8ofZ*^K#=?RItj4<s9_WUrqYQCGf?nCkWsm3c4g6ZgULS#X^!IrA+~K2xP~bC|<HY
z$jxzm0JJGJoLMtpe@RMkB-J9!*orvms@*gBAJKmDy}Pg<+F)MX_2IGQjydo1XiFtq
zh6mDdxwl{BWO{~phWO5pcO$B_$!Suk;~rN)v(y<rhi{oSKmnQa%J6S-xE~*@c9Epm
zV)@)#O>_)j*^mfpVQdrW<tjVJS-n*4o{qsmUp3JZy+7BRg<>?_gO_d2B{4j>OlR~_
z)2OS7PKgE&0l5iPP;#0pJ8Bw3N$R+Vrp7K}eDx@;Y^Y^W^tg7Rn8|{*#Ikx~%?s(u
zSdYi1ALE+2wjtbRZs(JkL+GA{C~cZj=t-{gGlH-%e5>I3QTZU3i&-1<qa2-|d{2&C
z>8fiMi^*nC!V9;foUjL%)&4^p1q0g15+?C)+GsiAuRdBf@Cl)-#VOFw3;qhGQ{coq
zGgO?WaOmd5Q8v&|iqtRFl95qr>AI#2rGhCEBF8&HE)f*bG$vmwWY3ETixh=R$ldP{
zG+>}BrP@e8Xv1U3sZcJ|#0_95mJ0UfdS4`E7Q2&b2{}5|f#wQ4g;Ra7QxbrPo->No
zfL)@g;22Qvl)KLw$ewxHJxqKctVzrl-Z?%3<`pcXw|;YM^RhD~RL|grZWYzCkbnqh
z)VyU5^ePvNecLFl;|RcL<R7P^7VAi!D**AlA0DKe`%7qWfR`^O8t%ABLyAH)et(a?
zSC;zQP##y2E&t-`Q9Etyi?3f{yJFlcHe9V>lM3c0xfj?VmP=4x0q)$d1{xD>@@_AF
z+AxKjq4i*xH-8UO--H+7lld*Yvvq!j2ZtS@^$Fc$d&!;W%~ZW3|6Z-rO}3L3vI<&g
zLLgQhS=7_WG&?hEBlNOV+QIx;ywGc#PawifEUs<1@oY3?JIfRA+SbQzRzKneJ##ZH
zb>Qc>GAn(VA|Jj)#w2BaL6+-_yt4%D_=c<a7rzm$Q<dg_AdW)1*;bt^+0Tm#l+<dw
zXt9c3tfW5>z<iaKkoy9<*1^9K)dnc-GYosE3^a@q79<>DQ)~41GXQ4JXw{hY+o#x9
zXPcS{Xf58z7woKu()jMTsVUXeX0#p_oA<cjM2+J4a)pInyYbPLb+XXYg~#{Z@Ea%r
zy?yo*fdQr9>{kTRWiXPOq6u);UoGd~*Cl~5xkuiN(K7ckA*%Y9O3fDr$Vme{s0rR}
zPz0+&Jh`CICY}~p0`}!#!&8&j6}4U>{BAuwGy!4POM6<Iu7ZOidPiYQj@kHB<F!+g
zNn6<#wz=O!E_K1L@SpZSjocP*dU>1$zV}yww&3F-|3Fq?qg0zRWEANkM_lZJ7trZT
zJoIQ29(+O2t}O*mhWHTdL0)MguJMM-+F3{a)#4BL_)mXpmdPdwxm{n@47N8P{{tyw
z8;EGo8z46Ubd<vjIW+!94R+!)+1>u#s_mKYQZVClly+a8dRc|<rci~AA20sS`bl^+
z6z$~!n9T{&p1c9^SWrOS49A^4&c-hL1IS52Q8Y0HO>9otXNMc;@-^lyhL>)7Z5z+4
zExqkml#7Byb+^zk*LKogfh!_6<|am$aPkB2ZH^2eG>;2Wovn+1nU0Fug;TGN&0a}7
zKg1-OpMrLZ=a0O5Jg+Wve{uf)PHCoj0_FSM*3GK;!@nCNV8&r$6gJf48YKz~fa+Wt
zSS`y#4IKhJ9Pk&<y_@)kQNs;a-Cda2+y6isB>lE)>|ZK+cwEg4tL?`0A07?t8GuhD
z@c}g+5P!@&N)`Ga?2`O%?6RhiK=Q$%L+v7#Ug-}$tG%^-;px5z+RWR|ms%u~Dmi*(
ztrhze!0oef(nkz!U{Q6{1?b3f&^G8H4|eMY0H3krA6II|>a!vmf1}+w`#!WXoKL^N
zRGJSIAChZ3)O$AEpt?gJ`GZ9*d`$iQVh(_|4!%Y0>Dm!;eqM;vbJ`E>c^Dns?j;#O
z7;{q9n%OOPx3L=)0N(E31h+;GIC+eX=p5l7>`F64P;@cV)7}HVm*pSZK(wCjQmrhM
zl)3mUkz1xb`#Oym=K%tdhoXTmCirSlT82=|a#S;aZ6VB{F)*uKlA<ne{7WzHb*WP?
zzQ47YoY*rpvWO3+gE-bm{B)guGw&-wnM3W(hN#+2<2T#%UKNz!iE<Bqx<xxw*2Z2J
z<FC--)UFd^*<GbyJ1>z(YsLct;U%`pvA5ND`N>LzLa-iZG7OK;yrvp9SH&Ch`UbQp
z2bwos9L@V&lWKTb=FZ}yr;j4k?;T~{H`kbIGB~cq;*VOPyZvhi9qNK#9}wqoRmZ!H
zA8KCaZ7(I-8OjVV$j<=2ERQj7CgJ2m%HW}5TZ!TA;Zn8ps})WUMJ15#k5b>Z(3CsV
z9~e-Hxzh2+%#(Q#Hz^a)wT4pwnSf<1A3TX+BAM?6r?vAwbMyUjuQ4UxD0RGGs7>Ng
zgtx*<1bmaVBuE~Q5b%A}TwOiln`au`BIbIT^Wv7uyhFl^Q}=z8n!9-x&3|Phmr>I&
z`;J+@CCUTT(VYV7WmPCXzAX!Jw|ZnfP>;{Xx;9QUY0UR^*|MVYNuGWA75M{;F|5h*
z{}7h@4@RU8BV`cs+Td*OD>4v~3o2iX?p$_uVCy#8E_+GC`L(+nuu=|7Nkb^d0V*f|
zsuH5U?*94~0}YpFQTyUEmRwpKKoxY{*LXNLOn_#ed>iEL#lNVSoIws*clkW|C|6sV
zp!JiXzf*)CaAkwZaUbi@AT>ykW~eGdRr9WKLx2TM5%bLIkSt5a>x<oVbUfgU<>6ch
z43{Y0^~Ed);pP~Dgf@@+hK!Krx4R^2mlw04X-8LA97Kmlv~2{KSC7aIu!*GANR^9#
z1q@gGxTIdES`LEd<%1`&TV|4g(c)yK#H2Qax4^fQAH(-f*MJA(l{)kRWsujfHJ~xT
zFk|6tkwV6&H;lOzEW<k7FjvO-5EeHB+2nD44;0f5G*ki~bcw8rnMJIwTv)p$J*FA^
zp1<(mo%%gN^#N-)eZQp*@T7}IVAon%kJEv(KAQYGo)I-+IRdC~QAp5NBH#Q#1PoQA
zQMh);rY$h>KwmyHm+6z*Z^Ct&ICWJwH_y-*EFEQFmXT2N;^}UN{OIM<629IDm!IaE
z?n49)59N4`@KMIy&3s*J;5j=d_{=qY-&VY)miP;!@U|2NQ7U5@5NEA_$iO!r+AIdr
zK^U*TA|&4+2|(j24`9-zGpm*}4pr}Y!@3JZ1bBSfLu%J${1%x)v=f3yH_&l{pdRy>
z{;TgPy^EuRayAG6uor+h<^^_3HSLCBwrs@Tb;o5zU4E_wa(355D^l#Yy$K&$?k0gm
z+Igx8yP7by6C~Ld%DI}cVq?nc)&x6!!&@6+Px6BsF`t)sx`Aq#O0Y~5;^)&24W+OV
z`<tfvKc!XgZ=HXSc_f~b#!q|S<Y!qZT3uCKnu<(M@ChSpoUO{2I{v;ceiQyQ#m3&A
zeE%d=K}IIXFc|s$aSjVt2lgzP6(?Y`gFTHLnnt{fdTh+#D9(EOwHD+^(BFTg0Scxp
z)vkQUR=`wCySxI8>RSi(815@?Y;$@9MadJ*H(0@Ume+yU-Gc8}evh9lqUc3M{DG_$
z&mErEh3remq;s`gYVh%qe&yBEsA!$YS(^L#jB@JIub?s|b73AHftNf04W6iU4NuL!
z>O?+OZ2iQ4wRP5bPNGOFKk{m1EY;I9EHZ9}>iU7>#1V=f>nqgyIkD{$^rR=~l}4gA
zz@DcA>hFi7##}OMa#z|G32hBPP~nw*H>U+{jdQDg)nbfGwcWeIBFPv6RGS2MUiexH
zV<^%uTzko|)j{y`O(1j%5kKnipOI1To@Ym=HC+KpYja}MaYQL<2WfuiND}qcldw4q
zN`0LVIjp5XVw2Xa#xo{1T!lW}b&^bXrag6A|MTIQ7wT%0-9@vEB=r0yA42~jF3Uxt
zr#3m@#*ogbl-)b$3-7DomGG*Xf$YRb6kT)-c3nOI9JRAw;wNP(dS-7y;j<4tya*;(
zDZ-?_c~%xA7cVf1y7;cCM&)6gir+{2pi}4B(04MUhY05F1N6SLTLFel$UCz(>Y=m3
z0NIYrFmU>?T0i1{RDcN$z6d^sKGEkmQX0PFIsPuV8@_c9FuE8Xa4v&v3d{XYY*al9
z#uM;(=i^kS9Vf-Long6FBOd3BwF3oNIZNizF9pvK8#_9%k@7D#AJrHQYuP1TU<?RZ
zWe{Azs*-&GRoM04sY1E`MinY8Q?J+{@jh*hT&8s+UaSC>djCbG(}v>zLZz8LE%U^&
z8W<aqy|qax;5DJTjB_WoS+4BB<K`-BQ0b>z`_!BVBMXQF?p?&;5x^fB@*7xSZIB{m
zPH7$n3J20f>UYHaC@d%wNxD_=YEs-@=UTtLd{cHgj15bqkvJ9e2u9lsW3s0nMm_2i
z!^OT@tq8m}iZE_EXS{&SX?+%7V9NC3#u=Gg{0TA!v}*iV(3)IBA1=@JH!kggmUvR$
zCje1-5y#MzkN`Ezkv!>?@%PKFobnP^Ro*h4t@xzLqk5xIBXncX8y~$M%o8X<7-&OR
ztA#RcdijKk_Kk+=D6qVshnyUMF(r>?GWk5448{jBL;Ol}iPN|ybb#uI0dUsldh#83
zS9I42xP6N-9A&qW4ukmt`@a@X>4>U(n`4t<-`*+4CL|{4)`y3$#kd7YY~SAUFqGfP
zU)MmBbOZHP_P|%YqYOEu`~{pDk}o-2MM?{KPxSV>aj?8k_wt-$LoH^L(3BP*#Nnj^
z>UcDO7fNDDVnpS$V07AIMmdQ^=>dSx;o&xhQDA@xAwcpzT+#*B8kMma=7|)>0^@v-
z?j*>Klz~}pVfljs<Q$k1U_B5;^TOAe;5(Y%A+`+wbg!@1fFr-dy!fVK%~!_T-3>ct
zTP)hSU%P)Jf2f^$j9`fm#T1iIN&0_G_8kx)CnpW}-V!%9z5kv$AL2|+2U|=AR4c*)
zn0joT0ynNV821@;$YJ?pIEiX@mH3=tr74n7^z6-@Wtry`ujSd3V1`4D$M;EqjYv>-
zpGE0qm(uqqr;qbpXd&}%5bDt+|HFk~1r+e6^f?Z91lQ7s+(%E*;SxlkXLfeV>fo3(
zoAKujci)k)w`>gC(ui{wvK-r-7PlUZa^B~L7H}TKBOHR;=eEuvv3u-5#i^qX)iib1
zrHtJEoPBhsHhQX`Mn+GrnDdUdu}+cvFLY%*IT6YM8keA^-0?6vX-IkukGuxB?zH$r
zLLH#%yv94=k--;8;twOOf)MJXM(re;M|8DH4qi&aN<tyN3t+5q!*Bva(XlN8r^j2+
za^#?Lp!CdA|AAy5ZsMMy$u5o3>O*8j0=7lU^~qg>p{GWDa@z_P@x)sj-`@~NCa7W<
z*O>hbH#7i*6$8+G545BJMN5sv<urk}H8=onYc4q1rT$Anpdu08Jbbgfqj1^swY3h9
zP<UpT?6>*^7W!`CL_~VZzTrC9%SAj`50B9Y(U5$A;2Y=R2b-X7n>Rs$ccSoZt_^XD
z%$VqxxKe*M<E*6)gPozb63-+1Wf`4jfuc_s&f`+-V(bsp_#*U>69TB)w(I1pppgqo
z6OqeU(hnqAy^f$59n5VQuz);rf=Re@zjfv87N2V!ns=FH<x>~gde5JT$dcdrn<}^1
z3zG)Rq7AUeI0s~I9XdHqiTvS1IR>P2_wH^~Xi23CoPNc~*KmDF<CgZMC3T&Jf9imT
zfsR;`&cHQHum2HP5FkKE`dbwE{ZkYmg*$9ThGCcT^gfXy4^{Z+yUkKsu8HL^Qb`7d
zT&P<6uaRqi#6Kg~gLHR+4F?^eixTNd&IkN8v3?b1k8=faf`m-7^hB#Z1-qKF`wHbu
zh3m^Rao2^hp!wSu{4#i8ZcD2H>?*eCBUvN5v)J<ekWiW#WWGS`u;(90!WyusMz#Tk
ztJKq}ly+W&S5FF^Pn>jYUjFncrvjRz`t>(5ACm!ho;%<mgL?{r0rb@}^&mSQEUK8-
zorMQrKE>XH?w9?-7ikXP@wRNPLEM=956jtU1GH=M#olZeTw<BgnHhcpEB_9r_uHY^
z-gB@LAcBL^vI58s!XRCP_K|v^Y?4WA;O*SEVPRwD;xa)x@sNOYpDsP=sQRww#rdR&
zAc3lc1l5#9rrVSEWq)p%NxruiI-x|fLlYcqJ_xTwk(03D@I6}axSXw|@^5RyR#YIY
z3W>HQmmTi>?5_}U9?)_<=Uw>Q<BAH<>hFeGY1_LBgtex=<Smw<r#eBk7171#9%Ek>
zY>^HU%E0C>f(INRLnZjqCDa8(-4X1-h#IvZmGS}m!Y)(3C@nI;hvB?KU{4bLl$*@e
z33DVMYrN9dd*&sO-D0jA$cUHUdu6Q)wOjhfsHj?&-83Hu1`h}G{%}e+BunazeDu3)
z+^=Ue>pkO<VamNe@7*lqAsfCfzXjcB1P2^6623O(dd&QoYL@@+7PoWwGpudFS@+jn
zne{N@!&?s5aMo{3oOgSuY>$W%xJQvq=eiucx*hdi-F{IjWi67yY9bQGxRk9eP?n7D
zk3Kd4WM=#=)C}UMx}fa);J)A!BfAn-UHS&w-3dC2HU{w=&poIMt|x9}3X;vM{~@3L
z5>WqBFL#P6DG`UFs`})ccIWK2^<o#(<KbH-C^l1?hyfnewT$6lcp&V%yh;GK{Kq(_
zne%dcTQ}dT%Jf2@vA(Q-T|`A?Z-G+z25R}Bs6`Ts-(o#n#`cH6?>+`x4q<bOaU5qo
z^)Y?!^inGG-EZfG9l5^^ErlFiKk&?abX;tHYW|dT2%7K)L^7MFm9O5o?~KP<z>+Vh
z*-XD?sqHkhHa~adi27;{Zh+6X5r+zvTdXwTabE(Z`8XMRc=a*$Iy8O^%kX}@*nt%4
zF2{>aa(mkLmKR3+^^|UoXTQ`CmzCcfig%VcVqD#Fw(6%Hl5P)^RwM;)g?>@$48S*n
zjo@FkM<{k3O84J2f)dDoF(Gihg7=7ewbzwYOSW**HNaEzubujY^foS`wD9#gi@knG
z{gHA_$c@$d{xfPG%4rhYJn9s3Y^pILrF}oO>lyq>M(@43iil1{mi8T3oVxH|?0u?p
z?kt&azT_V0e;3#{O^7^x7xD8VtN!%$7U4`m_dYX1d?%XP7i0BGXIS3D^&{OkogClr
z#EG*I_6>w^@F;4_CJ=0UE^~)^oDay%;7P{fphxr_fBAnGfCE@Fn&hU)oG5<lnwZO_
z_s4HPb}?(ebzSw<y0IdjXEVkGOY^^IdlP7=|Njki7z~9$+1H{}NJ#b>QkEpGlx-?X
zl0Ew{mh4+7MHr=$#3aTR#=ced2w?^x>x^}nrTf<RxBTxt_uO;Ox%YogPUrhQr}3H3
z=ktEQpV#wxp68`h$1R`u3d8>Mrv*;YV$(BthTC&ZWNUX+de)Ts+RA^yJz+=be%s=A
z@TfmGNpJMIGx!fdf|6Usk>T*E&w~`ZUth&Y*ECG?WB0gk1i#luJ0PP1zTHD}<Ea|L
z31(EO`GAe@ku@nH5~{@rZ=nsS@6TC@rS1nq->{OH>5u*u%tisr?nQ7y9vW(r-rV@7
zdtZvS!BfS?wF&M5<F*Y&0z4zm<F_o-s$AprF)I!Pd($FJn9u~wf$|`k=#t^4-pjRG
zH$e!*>C_a>BOW-(c~|Jg;-1y1U^=e<1@nY;2P+F2sAU0~pQTkGcg&R_!62+MlnJ9C
zV~*qF7ZhULmS3PJ?!4%<^8VgAC7ygaC?_zpml{mQG$Nqi?+mQ#Xk~6stq!_moPVpX
z=Nq`MC#Q0%s{57LImjw@KOIU|j#Xc%wcqTVQTSZ3c37zYd2iN+YvsI!{J``eu4&?t
z)IPG%8^p`p5wp!<hgJGQtZwVsWnn$F){v)jp!V-1@GUY~v10f0Z`9UCUs+2Yarh`8
zat7Oy_}<7ni@5O(u!E%=6^+mF!KO^fpJs@{Un~b|#t&Uz>#8SuW;sa>F9Gbf!ZgP4
z0pMx`js_$3z==gyaAW_i(d&T`TnMETp+4ESS1#qW&U3jrs@C`pacDM4eYTQtTuTm|
ztCK`4k}H}P;y+_~hubTxj>rbgrhWCu=<2c1Hvv(_a09!s{VdbkghdM25w8s_yTATe
zT^<yzdeXB0`Q+umC0lgf4-cK2Rv=iF<Q`ckZRV#D{UF<|Sl_rZo8H07>&vlN1$nIf
zSS^d&cQXBNmZq-DB}P`Aw7vaLyWpr*xcZlE%dJDQ=Ql4U!SKG)ya+LW7t*4iLsNND
znjPPr9KSWWW7<dDzD!8H{cL>>V*TKgoQiI0_3L|k)l}*)YNrP6@o@y4{#i+HW(_z(
zdR_IU{CsGJbj|XrbdJHh-r`S`Bwv5Xhl#Q&%Z>vV4m%aRc>HDYLp{Z>29)-lux(I0
z!zUsoDH1#=I)B0=cwx=H?mZ(4IRhR*g)#%!r%zWgULd5pJr%=;F2?Aj^SAl>z766I
z_dwu(`pc6DQM#fp$A4>|$`&oFYyV8s7r-|@HX7!VRwbsCQP!liv=+e<6@0eoObwDN
zo}9Rxpdx?UE=&1i`)~V5GOa!k7W%N@fv(2ATS-o9htrKF<cdbUrRqIk2HjrX1AxS|
z8~~D4bHq&B#c-W?otUqeTMquRGihMEmU#r)VzJmu22bWo-CN|+;vYz)-?$KYBC1Z*
z|FZwdU&$7IxnrK5o;uTjY1Q+ur$dTA)zGH4=qcSon4t^+5hEct0j4lQYXHWzwwnq0
z0$Tb*SJCD~f;d{2Fi?zBI!;%Tb9?6M`Q$28l$&?Y$Ig_ALngC=+l>ly$<dhAdk6&B
ztTULFWsu;s&4=J)qQKflf7_J?^kHDP$&U~mWvJ+)y>l%Y@wlDVXNj%0^ifK3ZP%&I
zmJrX~X>7X@c2@sPV*?R(6AL-i2&pjn@QsI**wA~2_ig5BK#m)6u6*_#tDN;!`S$Q2
zpwNL%5K!nKv<?8f=kK76b!{|l@O6xB>2Lo)CKY~nNAgqcm05om5G`ZN^fD5kt*Bx6
z(Q5=fZ>HdODeB>Fg=`F19Nh39$n#wIgAeY)R2MUfU-WXBtM3on8@Ajhx<pOpI)hH=
z@jU3&mpTLh@i#J!8E_v<6~Z)rrcFfKi`0-MLg3HNC*1Rtx&G=(RAQL?SIc>6xvZ^@
zE-#m2FqqPUDT*a+(J9}NWB_K1?4yH&NXb!Bdt$l<cP>QUe03zn__CKru2)wZh})UY
znW@esOOIXw6#;lOf9I~OvjAMeBTk-Rh7X~r?BGN8*`W3-)7f$K6U-ip>#rK7Xys8w
zVg6IsJG%OXRQ;wr>Sx8nrqs7*QHE5YWjVi9kJ$ue%5*ivq6Ea+uk#OKR$>X%xShLG
zE%v-*WM~=Vth>?!`l*!HzIP$E+Bsp64u^ppx<Js-EPF7VV#72?k=hvOWCrVf$YwY7
zUU$^Qx(Im0^+yR*uwQ;bHZ7s*g@$r8KF6-#$d)UusyOuet{(SU19@4TNSif%8}};Y
z6xfKX=}Lbfo|yG(5X4^O@Q-s5G&39b=N=0uU<<1caj)rSL}@Png+?*A3?C=%>NA|l
zw+`h!P&wRN@#Pl%kcUc7LLe(rQ(+xT5~inOVVECM{eM-G9x5T<knp^x((MCT3Q+uV
zm-r9j!{XW22_Cx2loH1mlfjeILg|`9Sf^%lw_FxH=UgSk^>_VLSW+vPyLAtFmk`}+
zArUot@v$t&ku7ul$3O|+Vvzc%3d;Lzu|9LJc_qZR6816`Bhv%=B4;jT9ceQNYs35a
ztX3?6#iDA#0Dgm|T1a4U#S)x6IXQF0sPveqgp9M&p=-8Tj;WQ<@!Gpb1@5JqWZP@7
zJbf|^no_|Et9676XyHGQgCU^mvKRUT*-^R&-IWzTDyMp?r>9Tw&-mXFIDvO@9S;!M
zj_%|d?#GF?1TxUEG<QsU3n_VL1C&*U))2QaCz>vR2G_U!b`nwpzTlP5ykhmPPdmmj
zVVh;u%haXD0=#GJ405v$zbgOcjXidx_q7*{;(F1<e)xALeH8bS;0oUsuTcjaK7sLW
zfMt2asqGIWIxy*_yAyBOZ@FcDYZqr(eT(y@+$_s3-wRjEJQJ+upZ>35)PG}N&v69R
zg@lpPE6b7Vf3!9lUh+-+`Se~!-<NPJUE(_OitTtB84C0}1vV(dJl#+AlV@$>qn~m~
zYA8tXFYa=ltN<GEDzr`Pr5~O8@*)V!|GU-Y9bapjv8)j-aOsdqJ6jfX-es==m50a!
zzg{-5Se79+|Ghb7dkcP$Lz3&Vhvz_0nEe}2Vpz2ehwOV>dy~L-Ak)rP-<J`Oc`n;6
zb+=Ko%M%Mmc8mQ)?*W~%iF0B+wbRxWS?qbJ?WkSYxyJt&cGf9$1Hl7)ThZdL-~mWR
z34bM{R?Nmh?2guv|4K%Ap>0S`wCI$p?fuNnbkMQ9X_P5yEh{6zGivfve@Z^<rtkZ4
zKI%~-hHK&R{EpF+3AabmM`yEt@u$)CRb-?MM`{m(9x+A|eUI!Oj&JQ#fAKq$A!u!~
zL{be15M?RpnSx0@x*D>FI!45axKQqmTwDJgtePh9v+=e7KmC}6W$OjKM07IkJXo@x
z6l2E2h+UfdJ*ISpo{r%HIDCTHKESdLYmH^@QBHERYUQ*2QXJF2bhy49q8G6w<KC(^
zAyN>+*-f^iyNXed?~TFt@7v~}S9MFDNcNQVRa;4*4RvG6G{DD!eQBr@uzUDIWT`Pd
z9cbvm|1bsPrk=1YBSqw?GP*v2y4T!&PL;F9zwM`g3uB1if`+JIn{;Q&5FF-YXZoFr
zdUCf=-7R<drO_?^h|itl7T+qg+H|}})z_Tbl(!X_VAJ{dh|pC|xYNCTd|V71={UwA
z-~}qDxHlkKsf9LJ%kK{(Hix&L5;G=wC&mw$i0X9*;)$q7Av|QG_dDrfZ=Td0xpX(v
z@@XEdc&vFz4ro>-c2!RQeo+JbVnU!)sQut^yw2Fcv(;4?Rn$S!)aAB<5nhf!hKozR
z^j@?)QK2v%$LriGUN9hQu2`fHxz3&p*=n=hvo!bjMTN@_VnEc|h-HDd!{IznkIiKV
zXQiUADJ7D=A4$=bQ+?H_Xv}`J5{1L&9a;z7?fn2+Tl4!@`$bC4(fO*~;9#`@r?tY1
zyR>HV&iwx^t(j9uJBvp2<Hh5R29CR(KYsCxD@*Dr+3MY2Kpe6nFHHJHA!HauXsy6*
zPnU>8HSG>*&T}M-Yu%!fa9c;GweY$)Ml4}Ec!;nl3Vhm^LlK>?Q$PbhCHz(Rjpsy`
zdFT7-pZ`E)%ng~f(|<|00Q0RSN%Cgj@;$^@kj9f=y6vv<vJTZbQ{&YuP+*d(Xy%c>
zgf*(UAydX?*Yfbio&&I%o0vzT0lyn?U1B@3yPf67(#-@KJB}Hl4!J(CjC8^*DZFy^
zIfZbU9?f{<G=ubTlUB~h2l*{B<eO%D@X}Pvj$f^ZQRQ-@HDnK8;h(0XTjiUyA1f$G
z+)m*pxbTco=1;=rWGJHGZ{j{ab7uJ-u{=`D^v=`p*Y`NgRCFN_xwOkeO5pS}^-I7k
z@&o_^Mk-VjFgni4(JANVl$EFVG;`P(H3cPX7SjOJp0o~^cD{_Fx~+95$*=xO9CA$2
zaP16$BpJ*FtZUNyaYdM0-P>%`3g+ig%$zE)pKbtd8sD#YTp7<bRZ&)v@I~>H#4!-_
zmRVwJ)!Mx+mvbDvxeq3GxN>=w2+r=nz_VlY+QYNu<?Kdr+XV_PqQ-?NX?Q*8a{yhN
zf2vm4EKEewuwc$J9FZru-G(|t2aqMrhMM*VV%N<K6NJcPcM(!|aqTw1W-V5qsfR71
zu;lz=tg(<Yu|;5SuNSYMXQdhT1?FcPrYh+>nLT(Pd4*}dw+#@Y%=9t{_8udfFgpbH
zq`F;<RXB%=)}C!*z)7Cp(F)!!cSu3z-MR@LzvUsZ-fN<~7a;*!7ghHR&0_Cw{ef^`
zI$nTcr3ny%bG{E*mgF~|QiHm8EMJABC))YGoWy}!L-Jwb-;Irq`Cq<nQoUxtZX!E<
zXI_rm3R-+_>Ye&_FX{|1yq|At0leP+WY8(J?qS))0$LD_@21VWGmq6}Q7<rsF|5Cf
zh?m6Xf-t_qIuRJYG~JbMKm#gT&3q5;r5(U@h(|C9`OMp)dY|Ih{~07~2?9!0Ui%Xa
zUzk$ZGwp^=3|-y!oS5rudg@k_d|7Ak8%P_NAa|j7k#mB6z+UtT5P{b!D25hfp-6JO
zGgGOoqJsRb%A{y%%2u_bt5-;CBC$<2%S}FED9`F&nT>@1WHzeOj29RFKoEf|vjBAq
zqRf8!TQt<2BC|(j2xh+-ipo5h`qf7AeVq1w*{6C2D;==f5D%-lXN!Q(lyhptZU6$W
zR2|q#nl$N}0B+k&Fgd<3)kmrM;5Cdwo0DDyL$CIcwC&0#N?U}Nzv=`&Q0SyIv|X|@
z1&A~pIowTZJmJ3hA}u0IeAoj+ncpZ>xUN`q0ACh+pdymZvrUY2NxvTA3%OcW_IJ#(
zReDH|02*P=+Qbw(=N{8J;cz`Zfc8=2$!OQa=g$>m-S;=Nw{13O!Mc+FN7SHF)&Y73
zx87M}Ud?x&DOj@=O>jv0vbGiM{~KlZ1OM4tf6)F?nM%Q<Mx(~(+rCo?4kB$LxG%>T
zT4W?WHW0>$SOTQFbq~jw7m^PlNp)!}vCfZuNbwl_YJHS{M8kGN{7FQz`$dJ~+#Xo!
zb!In)cX36gD%mL2IJW6g=8W-agDz3i2C3Hf%pM+cmuA@;YNBZR0=D!}@rtgzR#zkM
z*Bweu_QJ5j6k*hfY3vGRHUP|HgooIXTe=rFbq(^jPYv~cnLXZ0P-SG>=mxxaUi@>T
zTdKxtNO0Q<_5R8K<NWD=r#Sw9_?x8xC<6*f*l5)ohPZD!V`JgHqt>~p`VonXTq3q)
zu6cu256{lIiL&TWA3e*#=5OzYU&ONA4yt55^JrbMA&J97t;b0odbsAw-{h$5*8ndg
z8Pe+hJQZ~GStb<>Alkw1syeZBQ!1W>iGG$|9=KQm#%3OPdJ~^u(wR|zw?l+&8g45#
zi&m$dp^Br19<0AZUBYM18|?QQ(=QNl=n)W^l23ZxF)IjIvT{HA(|9C}tqHJbcW^Rf
z6g|lr#=wk%<=&DD|HwqnWA-Tdj3K0B#2j~$il);q-Ko*_oJ~HFmk-NgwJ_}0U6tm>
zW5X<_vX7qR;oWJoa`qjuK(C%OJ4z!2T8hUK1+-JfYrxX42bEv_#3mfzlDj}S>J0b}
z@y3ZwK|Cn^JM^XBtpAHY4Zt=lCx-BmrK%XoAI$yKsR^cXzA0R~xd9iqZ-pCcE`-B9
zuc7%V3k@MBD1vqI`Co!vgFI8<BP7n$6HJJq^z*fZDb2RaEa2VF1HW?#2)1^s&Fh8x
zi0<^KD_oc#M#_vmp$HP?StEws|BEHFpCT0VOk<Sou4AT&8B4!!t0)jwv={^14gy9_
z`jlB5@$$9s+?&fc<kMY4H;zfhz<zoU4GaaHg?NM<ry3J6!uMC+yX@@u(^!5fb~WdE
zJWKxpQ3J$mPR^4$IMs}Zh0RZlgh=-ajL;99&6rKUoVzhvn*@K5<nkFf99*bvAJF@v
z+x$}eON~Dd__TkwoVS4<eCNJgH_Rst71h)W(_#Y%ixux4@eHAIPnYNQPv@P5`td8M
zXLo&|in2<wq>@z6h|nM?>J%D~AE^zK6Esd!8|0@I`ZE;wAIRtsGj&&+3v^B_({(DK
zz9eGzP~%pNYEqf=%e^&KnO%Y@{Is`|O*4}4m+kYMMat1g(V9b3jN5j~Mw!cd*OR}Q
zxF6)txEj3lNPtU1KvQ$t5t<ww0=n*3D9;#C?#T6^A$OWQro-hwURhe#F2gBB(OBXq
zy2BSLGOw<@8#8x45NGNtW0mYN4;2yx-Q@vGKx3fd>x8J$!9I6hn)ygBRR_G?`>y%H
z?<^r(0Zz6A2I$34VYek~3_xuyp3D&WzJM*k(VrG3)7Zg@__9r3+6}B&mYK}9dP1~y
zUB34_d#Vqj-($v<+e+=so8-%a5=Xi}7Dv=iz;%^~s50!{-zvj;UAn{E$qFAOIcDX;
z7I_s4!5-Uxl`2eLxH90TTwJ4_6Wo*`w_5x*)=%p7vC2zS7CYfErZm-|zQe0dxG(RM
zhMXuWsC_9Z!aLP-Da+~#z?H-26fg7(@<$$ue+aD$F8FR{2Qbj#m!Nc5#ms@Rkrsm+
zE5jsw(1b=)b#_xgcZQPvo!G7bN7kFHIrS{-f{RFqdx-eFwo!p}(+lX`aD8Q^WdrBu
z{E_nBDn|mKSw6(Ob)0$&;2ink$2=VZa7OW|iJkpVSQa%f5YQ~KGxL;>h6LDtW~<2-
zmYa>_g-F#y6h4xT!_j)gc+n4Z<+pDV_8~J4OJ-+neo;~M{NWTg2T*L`{AKI2`F5rS
zWn4E?bZpfdSXD$|o@h#iC~$jTFSePipr@{YzU05Po#>9B;*hST04;>b;2rjkiM{9F
z1<V)Da>EV6)^UMUIEJXRvZo6nKB|<R<(=Z~F5|A}sK*@{dz*94%ZLT~61p~6hi61%
zro|^$c9vYhSqRp@#dn?%J8d3brrm@3YHR%^FDFq8W$9C--*E(6i`JEGx!v#k53Yh|
zP7h(6+*Ok42X|)a9HyaQ2iaK0@cJO5`shb%E6E$Rc-$kBK>VEBladRIhq$ko$cju)
z^TFPrxlo35S%9xlRG`ZL2VY5#VoPz1JtMOW1}M#whk8KRZ5~-ZCPXaWg7domxY*Oz
zQ7&4KOW+N6o~X!`@5xH_KW5OXuvQ0VbHyXUXfDF02motm4D=<f=xVb~$~!{cOxSME
zv2$)^-GYOvO!~GFx%Bkuvq&Y9gZKVCuaAhT4nKK_&utHwW^rc??k_1O{9j#ZBc>c2
z!xlS2>fJONfEv=cS4=4zfAJRSy`&OyZFWtaP-WuFAfwBoHq><e5ZCR74!Il=0Z#Sp
z?<n_>)`?Zq<CsNx4Q6~?DEQZ;9HuCyr*eWS=?xN}6nb^FUOMN9_b!zc@pkFxA;x^$
z_TQw;19NatTB^(WhyO~(xgMfUu}>iq;I=OIGF7XJFB=nhx@ED|F||LKS!tnM;!jDK
za$|a`335>$&I~&QktPoySV0F|dxFWD4;+8?ZKN?53yM^%Q{mjuQJ?&=R=K#VdSDvq
zOEx}(Y4FGD*RPP!kS{<ZTBEWKc&WY&ApTT{qIL(6_A0a5c2GFjD|<7~7{C#ZscU7%
z;k=8iUYta!;o_MZ*yaG(<!-zIujD_uiUH3yXLqsL>2grTV1SHZYGb1!QoipvTI20E
zQmNZRr!e#U;i3Bh{zR5mMD$}XxcYV&>J_qdmY#y$gBi+ab`7x1Jp?C^U?=)4Wix_I
zGfuul5o~((P5vqOiMQK#-16`5uyn|UYKnG1Hhk9LOW?qi`CC^5=`8yZ=!f{t9P-}I
zxQ-&D5BkRk3xS{5%X6Kh_EXpV-cilti!(_<&QmvTr=Q~fB+DJe=3&4lG4v}j<SANy
z_nQzWXwVzr$bAHa6g3r78!MrTbmUEh+(_^zuER*jaHa8pn;esoH%u<7R8I3=m3mcs
zuFjGHZraQ_P#8}zDss8)G;A5!&o1L-n=;GJoQwWi>6~+N+G}AZ>)rARFC8#+%_*6A
z$lr;?OvxXc?=Ni%XnAdUU(T5i;&-H7TwkFpD2_gfdxg9|u`R^hUF$~)>zqfRLmvpv
zSjd=N0~Bb$M`P_Seo_DQC|O?od6Ybqz)@Qw>6mDLiaSA7L?T?zWwo4e35^8zMM;BM
z3y9vHvk-fn6eD*-IjY>l*%NJC5`6ksy!wYzpC8TFipDD(-=`@epVlKG5hNT)UDzA}
zB8*Xeu*f~ZdM?rzu2CpvcDqbU>MZW?FtXcn>TI``WDUff!f9{rjOI-9X=AN-#t8~M
zAq74Ef4iaoub&4~nZiJ*4sHCS@^KHJ<LbV)yz*KRFAhv>H*^@)BnDa`^U(ZMzr6e9
zBS~0N<>lpJdV8i>?3-(DZ4OnJjW4A%G|VgX&gH`OsCg9cCJfKA;*w8EfM<|b)HN{5
z&hERazVFXYh#a3*N5gvYl_qRiMc78ITeta%1tA_E`Sc254<leZ&=5?C>8S5g>DCmL
zZQTu-KW1&k2m^PzcWo%fKZt2K;?6(=K>RWeLNnfs96|^W-`BoY(JU#nU^Re#o4Wr~
zXUBfWspE|%%*zJ#a_jP}m0%FHJe41?jCHBi#*see8kf~v%RdwD`a3)h7nbRok|RRP
zh>!hSjStblvg)`|2A<DV<c_NKsxmw1@czyP%nyUWOt6M9_D~2-7%u<GAZNNoUR#h}
zrGmt(8wYRlY910b<JE*|cC=l&5h6iVCo1zSoZ2Z)Ce6JhROq})Sw8&YR;#Mu&*ETS
zRUNCBz;|(nqBez!43U~A{|@{#F^;MUaeVxvIq8=Dsg>jNEN&hg7NVN%T^8XW6A{bd
zk84g3RI1U&Cw?D|A*F7L3~D?k7d%agkt3_B<f!%4c`<;SsBHd!xQPI*{@(vTcB1vS
zfBD8%>|V6^l%0_dP-xGkjEx@*xP2$n<T~!=`3Ij%MzvZD0UQkBASV@PzKzH|>;CTG
z6Vn{$yGL{0oJnY~c#z~_(#2K090ys&#SkO`ZDhe$7hlj;V2?72Ta7=Gl76u1!1Yp`
zrVJ&A^IeEEJ(wJwSYyf?drPX|PLOsQS8b%=p<8rFUd6YB91l(1SqQy|W{&wkGkFt$
zdIOu)Um|aV&@7!C!cRuV{OVQb+|U{QaaFXgX}?>E%B+Ks&vr>ftFdg>^p!eEx<C0@
z68%i(;cTgVr{H61r<GrwxnVN>{)z9bsP0s#%QSXGmT89kCj|yyC|&AAe+J*>^KSg*
z^Ogu=HKt@sWhpy@$!?x`vGR)j_n)c#ZppjYgp{4U*97L)H4wvcPJq@11>U$9CsK<G
zE89CJZJWu#UzdC1>wB#`2^o$J;ALxv6)VibcSGxRZ2*$vQgGS?_$YPZU(){@_n-jS
zNQCs%ag(ECmLJ-@+c2YjzEbX25~q=*u_CItkS<B)6m9k56lRLiJ;>N*9?!BwC7-=|
zhPP6W;PTF_)w6v(U0nZ|!vvsHCm`;n-vl3DGzF*awZqH}7UyqX2md`wn%Hu~tnUWo
z(OQtO8rWpfT~{019@)x)!w}PBw&uKa4=&DmuC3PlE8!uc{>TR5v@xo~nYga`hx|3)
z6w?sB!{-xP3|11Hpj!3CyS)gU2(veYn@osVDYJ=D4;qcv;=W?)#O*5Lx%EuxV5$L;
zk`OmSy+z4u`~%_lvB}mjHb33BBhp+bxNZ`4jd}2eUD#-=cT%3{tTY@AqR3KTOs>$l
z?J;w+xU-DM$mx$@MceNKvX(NCl^+wb7mc9smUzFvUniMlvyeWanDX-*pKq%gUgEfg
zcanvC<`CJq3Cl&5vdKUz=e0Q~)$f&yx^c)0rY|}iBh@4d7|D?%&}N_`I!Bc?98x~n
z#8DbD_$k2E)0yq=;afw_&ik_@5~f69q9XDIR2xcA6H34hr}R)I$UWeGj*HEB>$@c>
z{qfH?!XS^~TGN3;^$0nzbH`u%w^e+MCJ1(lE8hYA_CLJhP59&y@A_)D0`I$z7c{;=
zpU}>J;6HiUq2O{Kt8tyyQj0HUZ=IsQ%jD_t;zvUscGkDvl-DExX^ye@jVHza3y+Ou
z0Fn6?n1aACEq@A(5#G&AJO|luy|RfVD@EmbJ4=>V_<wA_XlrA@A9o%jkp5UekE~Eh
z1-|ActN^i=*+jdlHuCroz9!MqJUTD*Lw$jkq<2cYgwxuqx^zYiX4;ve%m^?B+h-4~
za)J({x?+O##=4^%)Xp!-Zu|N17n^o8ehAZ_&4pN5kfdS2454i|U@z5A>UR~%nDttm
zCi*ye*+6(R>SwC&w^NeoaoQ6Me^8^m`Udg;wX$SNbP4JpC4I54GMRR~F<BAs7%Jgr
z*9VOlaD>jY2P%X2Bi0(;oPBD3rD@4i+`7_qY5B2Y$*hhkBqbsu9PaQNm1T?%zC`23
zd|d*)BP7jJU;5g$V57;wev9hL0l7GxBe$L913XI{An8guDhV?mtp2Z5D8hC?_Yw_Y
z^C-~wkVQQa71@4lGMCtW4ONzy*K^aM$EpTMpQW1y$vjK5VMH;bw43%eb8ZIv^z{|(
zEfrfn<sGQdfJ(T^gkXkxk);*%6ihrQ?s+f_`J7O_)2NXb^}l70{bn9`%&PBG5zVb<
zbO*L3_}iX^LWl-VzxzS8hwN=`Tkm~gRym;Ugj6U#_qZvCO9W8`UGu7|XT%VuuTwM_
zA!k8RAO=~7QCY~Hr_KDG*mh91p~%%kK74Z_0%Ym16(#oUH_=f=lYUi67yDisZdEM*
z?9Q6XN~>QN9JIoQ^dMJ_=vH9eH~~@Qb_G))8Mve#xdSKv*6!xNBTLGUqq94a3u#VG
zIuEgXf2^$aU2|83u?Y!yeeLp7_LY|v1@h^_Igm58dQVCZoO=1|_*}R-5^og_qc_q`
z|3E@P_Q3oGl8c(wMA%m&GL*ijBJ<h^zP1J5$kXnjVSI2p)Wg`jflbN)_w7FI%H}b!
zw5{F)XmuaO;Ae;UV94DBZp<9ml{#oRrk35q6;gLp4lMm@l#;T3-((kPjCzMHfBmZ8
zp#`EVigW%|DC`>0cKHloKzj;^!vZpr0oDC`IiTd2-;`#f7MfA=qn6ZbdGDrnkwO-V
zhZ3X5jU=B-@p4H$N)V&B;{b4URNj^a@`E8DJ@|SNRz4`5SaanKRjtV{aJ)*v*vYMG
zB*MFWqA^CNM3?)qo{6Y>mAwMZ9+0r$-{O8CGjh2J)caVFqM*=>rNUvm$=6opX-(LD
zx;zp^+}1e}?+vlXueG81=ZimZwrjC@_<swzZ#9AWG(>;g%^M81;J4Ng3|(yn_{4-T
zt04fqCs|#YBmfjeN{$d(-M0L4uIySxMR~bF)hO(NXE7M)kdqf`3Q~Hk?4h*#7?86L
zw#2_d>!EP!eH^$H<^NpDjT|6)6{ryv#!86Qg}uv7QL8e9_{)$x>%(>U3?!^(5<?<F
zzM)pn>}pkQ0W03SdXY`~pXUK^_&?7>BMFTntwbUA#cf?Sbic)K<Kkkb`OI6wBc=Fs
z0>t7XL~zKe3ovNi^fU&X+2{cBT}}0co^2M8wPYlfn<IXXHncH~aK;pD3!Iy3(J+{E
zsJ1$F!J>9}^K6-ZpY8JOH1|~x)e9!XK%#V)_Kxh7JN-p|<HXJkVdq$=9s)U%SHE(w
z(78@l--JDLL_)c$xvDsC3KheY!Zp!!y2VEmQ6_|Yze=LUOfzqLG)&&?wA$l!;TLf#
z`v(%cv2L*h9*dkA3C@V<2lc6;K6eSlcyYa=e3r?V9inKGkM7fXwf6nllDv$ch?ZpX
zxt%L*sC_j`L<?@`YCL^O{mp7YNmR?V@rU&~j|Dh-KICgbQ+|eGv|Q$-M>LqGRt?EC
zs5FD$MPI7ag?5=L<nsJgYFjg(Y^7zo*}b?{E?V1z*F+!UzB!&iY+Q{J2^V86_&RVb
zORPAP=H~j-<V%AnchVz1PQB!VKvNBZl(6(P0C0XB-o4x+h*vs>B}8oGXM?A3l&tER
zg@dujm*p-Sw0D+8UJv3kSg~_YK<^_>h&uONr>2?3ZH~Yu1eWi8brK;xj)F>-Mo_KJ
z0wYdha=m-hphl0~DAieaPEJ5R!%sK9QX-*H%TlVD&!s;K0^!~RB>7dz68`Wed(Dc9
z=oYc7R*hrvXGX_0&L{9maz2on(v*WHgxTrsvP{PH6VcWe-ne4!kSZg;e~y)Pc5`}a
z$or_U>IyY(y&8NJfRq25KX6^@8;V~Oil<m{3EvRp8T7F3w#^Nr;`t2O08m7<%NG0V
zS{=w^_UCX6;LTjLl<=Pgvm1k7&5x@{IIb&|8_76T!d7uI3_M{v-&uMSw%e02w`X_#
zHlW4a`2*30Fpg64!_XR2*O#6(@0<r5!d<z>8yTW%1FWwlBzV=56yRro#8PjcC4OMN
zo7cJ}=mFDFRjotDXmQjorhT4QQ&aY_LJts7+O&qSklPZahUw<74MH}FG1(a@Po(mW
z8OSC@eX2o{G4+8QIM;&TArj?LJHr(%vZit`#6{Y<o1JoaHFc9*>HsU(J?RdhIbG*F
zv0puAL%eWUcFkJnNJ<TU@p|cz*(?V=9ZTUz<h@;eSX#4?^_f#UEm>QsoVU}LMF#a|
zBiTMA7+ru}*dSuc;q-K{B5<6@&HHPpDD3K`-AKh-Ho)uugk^_oN;x`YKPCHHZFg+h
zb>(6iLkX|+haRtIosSsE>aldGdM0G`lDiw-p;QN}=&QZDdhV_m&VyPBsU}xs*+Lif
zHCF~dXa4mYLObhZ={{s^XwB)J!}(Hg{4in)o#~fkIYCaJNGVQ^&%Ybw6Nqm<BKsxi
zn!3EQHhe#9HzxTJzj)-`%CKru^S)6TUhP@Y?TEk3(87NRY&a3H^(&m<2>FM>PNDvq
zBs@Bmqwazfqxx=LA6z1xO}9~RtvC8*iMa1bjF^ZRETjukFby4aANGnzx*EZG3s;&c
z4SMr#JU?Gy)i+BZYhm@Wx<sn`k~$#d+ghYo8i2Bs@?Cj&jcKM4iCLHfv;+^JSV5Z)
z6M?pQk8$>oFsS%kccdhvqUJ%Twvz*Qj%Qle37tK<hd{6ZlW)KpFuK)}wNIxGBzN(y
zp13jqVo0b~_nePFXK)E2%2W+lQt=Y~2I<8Z$|lTG;l+z1o5iHw>)QoYq4iyG5z#5A
zA<-%xAz<&{@&V0HTBwz_u8e%-Wq<hm`#kSs8qcYkwmBKlI)#~u|4~f1Y=Zvg-(n&F
zoQ<HE7zV$ojacnJzazbmdVkAmFt=I5>eNxZA5MJtS3L{80M`fh{kD1wA@pHj6y8K|
z^c0ub-JfB7KNOpu4g1=jID=5B@RV?Z7He*MPr>O=F(Sa~IJ?_#0uxTU>CY^9gOC4H
zN;tHl%B&)E6($wN#3p&{H)pyG_xIgs5I6b+jXM9iK#o76Rc*YCDN1)I<C>K`*Aoxx
z8THYPfMM)M3c3fe*_YHL;~kOwfH*aYiASzprGvxd){Ab2)W*+{4(zJ+I$>w45}Z=z
z`chbmaVv@~IthQBZb))9zWSY%)>&O$Y97<4%ne^jQ{-&WTpAd<LPPpvNq1jP@|^Cj
z*A^U1eBS|4+HC&vO+@Z7_+AiAIY*~o3rUFcR6wE~(r5t?P6RY*db{mocU!Huqk$xo
zc{&s^2O?IWp+h0O{xG}m%XmGDCf~mCpM;E)@ro(YZcF&<+&QnO(ozEsm!n0ANN(y8
z(6Mb)d2MiWNOSGRjclwW$Bf)mPPW3^e}dr<K&h7$!ur!c?Q+gqqjq|px7=di*vR4b
zeDvzP?#6sTS8q=3n*B?f9`zZiwUD>SnY?ED%sBhpHy`$|{^@5W6qd)gFPe&sHOPm3
zmM87oa+yp1QOGn#f$Tpuz)&56peWLbOqOpr)E8s~Ugx*HuTvG?uH@YCuYrJt23<z4
zeFgaJG}3aB)B6#+Zr|HJJGmQfuFF>hDq97&+|js;9Lv1^q8(HXy25Ydk7TA1Rsf)+
zCzu^cNHE3#k|1?f=42Vm%n$}lQy3fj=BW{-6w%r8L$2@DbXL}}N{Dw4g#96KA<RbW
zLUYHAUkw|)`CYPHl@;EW`$bJn-~wc0$xhOn0L)%k|Nof1I*O6^G0?NRG3inVG6dsd
zWy@m}LPN-7M>G*c>H8}+P>NL}Q;R%PQ24vw+|~8%(V)4NiwgD69je09#Z4@=*w)pz
z`@#FaOBQ!1QM-Zq=oY9tj?b}ftRxVeW<F<+){lIReC<$u=*M5Ezf0988`zV`ek~~Y
z@jB^wKPh+_v$OXh5XQOCf?_^W=x;KnwOqHa=`OnoOMXPK0BX<)vwaW{y8v~G0o|L0
z9z=Vp6<bQPZJ!{AyO^iTA30Sp`;n$!zgds-rJkMq1A)M!HL8;zDUF)+l+C1?RUS@=
zJ)m3S)Zp42_SjGAgk~#jmJ!TPHe{O5PGT8e?hJu@(Eb;j@5!3CIWi_yVV>@A<hD?)
zPQ8?R4=`lic3C>6CHLvBi%Yyq`xDhHucWGD25f#Z5vUhFn^ucvK$XV}=-*pN#_K06
z<M1#fj}d0_`6JBe*FO-t-9ruE96_>tsPcizQv4@!?Q7-P`reF7DM=!#S<imjh-h`W
ze9&wO$X_}HK$BsM^hPU^At?#*htUsXO1BuC)+^h9YR%lSA}*JbT;1!_0pNjun(Gk?
zwSF=}*4)mpb)-+6xb5ZrSxXDorTDKA&)@z#;(21eJpY!<d*>}7?Zg_tQ-1u<-$WTa
z^^s-cx4f4NS4FZg#d7L|qd}j~38*v{2tDiZJbO*?{U_tEFDWWLHr}t2z%D(Ivul%*
z&#VPZbqD>*ceKuHf8lGDlqX2%UTfrW7mpX2w*wC*BtOr(Ien8j?Vvtr6(2H$S~XfG
zEUdyo*Hae^Nmd@FF(2s#t^I*4XYQ1emfyU(&)nOmj6N1G6-Q<-PrTxk{{Gw(*%CIs
z;0D~+dD_}0*wRx)N!>-=%qJkttH>7B!f<IEkq>jfL-~4e!6@s4l1r)gh1r>_1}SUk
z*AXd+RqZXlPr)4Yd>|240O}EtC_D%KN+78+u!K3oAwoB!W~g`MSu=){>Ogvno%BaZ
zNu>ygA1OKjsbEtyiH7j;NAZG@3~M;`b&Y2Dw%Q5{=EoIq6?4#j4C#FhU@Y}<0vMvM
zG(GUMbSzVaEH|<D<@XwkefP&D1`;q%QAfHn)O^Ift`>6d8+eF3N2lRl)+t~*KJ5Ck
zaG-Dkp`}pfKIHDLQ1bhGE4EN^_6pQscXDInO7o_}74q2DlLtkvubMhS_dIX?HH;4)
zN&sW?SmFRu_oKh4Xk6KXG2?P!hEo>9)N|2HK66)(re~0_4e5JY(vd;fR`@=;UMV%`
z?N9&eulK8^UjF)W%hg@T{ygMbo|fJ7D??u7b6X8mj=C5IU{R-I>0otz3<Sq*%so~c
z3oy)n!Dla=)#YgPmFjEb5jh_S_q3$)V2G$%!W1%&DK$bBKL4WC5<4<-;nn4!y+QX@
zS3`4S`IhZ+6%w+WqT0|_^-*<Uj6BSA8eJ58GLqSY&)u7eNklrW*Oj|d4e<V(VBh^c
zV^p(ohi2$@>F3R9Up<8d+TKd#Mb{O@UpI=1WpCV03DiE(F8c$7Saj}&SYSG|$oY(L
zSLFJ+9+;`Ea*J}D7_7<0=BN+F?_@<nqll8MRZ94jH~MQXOUgv8-?df0rM5qi18`2v
zkOFR25aM?VgO?90KK?Rt(TC=f(wK3lJl^qUmYeIRw6LL23%nn4A41XJoAj#~d#Ysf
zOFn@;H?C9N?+9!{Cqv&}e<3H(2iNs}M`5$>55$IuqOoIVp0LdOO5=sYidS+gQoHT&
zrnU!ig8a*kv6HT!_Ju9-_)hFcC{f)>Z9lARge!!FLu)*4v=)<Kl<cp$+>{~2nCJFL
z;83r0c``joN*vrOAja!W1n>FVkO3+7+5x4dJLN-)saBkB$CV~mr@DO=WF>-Ga6Y)s
zuJ1ef`s}+|eE*9-A_zrtbTr+Lrv_Lci}3p^ZubUsdOb>z;P2Lye!Tv{PM%~OR+e~S
z^kRRF`YoPcrLhN|xu_UPia4gE3`n(QF8{Y2;q`9G;UIU%zo~fEoT>iyrQ#k-ji(oW
zu}vq0^Xk2iiXb;N1acJezR9Edb`I`HJfFE}dK%<GS!k*c|Mxpe5aQ~jsa7q1z4rdm
zXUgiTV|DTA`iATsn(eZ(a?HE0Ko}ARMs>O>mL5r>Dd_yV9It&;F?XZYwAC1d2}mv!
zy$aPr?WgOJ^W$)Rv}Q5MHya}|31%_}%EL`W*#a}2qL>o?#vuK8Gb$o?c-VeXB{s$M
zX8t=-dFe=>vWu`p`e{aPjS7*l&vx$Fslca8-#@b6Ek8SwVC4Jjxno$en={BaMIim1
z+7O4MPnVSqJaBiiJR$zX?vo(fQ}*jb;=XPPt4>sz1wB1^SIXKBb_1ag%x_SpG6pya
zU;oWTpxHD}@YG08CSIJ^-VAdJK(~W5VZ#%WYT;^08M;%@j|J$XWLjg6KyCUOfV*!Z
z2MT_jED66}?46M6p$hwDXATm6ja#~7P+T&U76u_kSN?cKp$$L#x1T58CoVLcHY>PY
z+3GneI*AWqqppzB-%M7OO~9yL<}j=3M~|Fui)LFXwu`7aK-}6OQ$MV}{DFux08-mt
z1nhnLY|^?J5AXl}1Gxi2+wCR)Kzso%HJ#3d+k@H=M&b_99v}|U4<YdHM+w8rZj%W|
zROLIb{u(;unJ`sx#ZI^$B~DlUu500w@2}RBde6Fy?0B-5`)mA{BR9V#c$k0zGoGkp
z=o{=INBx`mjOiD^NxU((P|ipLxop3&8!Mni2T09e46rkhtzy(~;FTPj(lxC59lmBf
zi9&O;dVreCx|W25`a3nM*_x4R#riGMNvF?IpjDyM+^PPv^vUXTN1%;#bD9~|;`9)%
zjIhYCCXbx!sUaU^yK67P^A!%0&A<OUL=Xs~x0-`37tH&WRz9rUjPK$fJC>}X$|6%B
z8*YrU23<MKs;+-Lv+>_)=g4gZ4b<kI7XYZA)t73Ugpc?&Y8&AL^4s#~hYDQ}Broy3
zV)N1_szHaqZ&*O=qu&8cSaZ?o)uwY5wWYP?@!cwR$LN&lqc^XnG~5wE;K?W;o^|++
zSy)CfxQEJ_E&2fE_I*#{MV7poWsWug0;DKp00HmE+Z5^qBdzL^xVUdm?$=LsD6sp=
zt2WQ>i{xUeP%)F&e*H>+#`7ziQ5ND^QGMxIvWzghse}+zwEG|d>E0j~*?!YNpxj_5
zbBplf0|@QgDQKwgod&yic~1|=Y+T4;v|trQC{MjR{J6T->x_(*gR$3~9;_0DVQJk4
zIrw?$(%A%r^g_h7Pvqk<Ba)r2Ig?xmS>71b2iKSD%Ztb-tWOU$$`=L0X@U`;=hq#>
z%ggb^2($G5BFD3s5IMR4)}D~*4z)$Kuc^4adv(n;DZm3HG~f*EOKAvdMSbQ*d7%^0
zTyXuyL7()=ag%yy77=)*Q2V?}j-D1Sf`%;$K136MQ{SV{1qH_;8N8tg!n@-9xV}M`
z_-HQfBaXhv##~2i=NrMa$(3zEb==M>7SLvRh<|qX(iIYC@tq_+dF#xRUWEr=-_(S?
zV{d@a*6hQ`6O%*H=zy2^N8TilJecZ>qO~ul@d&uyg|DV^9uUt*Duwtt0y_HB7Cmhx
zp0K$WFdAp=F0ze8H)FyD=7m);jkVP;9wwmMr_NYSuL^$^f%QYB!fX49(s4DSlSp2B
z8*rFvpFiczQTYiZbC%BXn#i3ZMrWn1P9^N2(-#%y2|pYq>(n(hk+;FqAgr@2siDQG
zK__vtRI3X=r%M_?iX!TXmW=dgq!lRDn)2X5QQ4T#pq5*)fF{%KrKzvmn@TA<L}$;O
z!V_vKK$xYiz<<44al!SaRmRX~sZ#x58zL7k|4i3?bQYFVxu%H=rb<qNHUqZdatraB
z(nE$em$gnj*Fi|3hL)R&Ytv6k+XJiFqF@Y-AJbkmiQ1uqudSa82QVp887P1w?U1YP
zlp>OFbe*LC=^2|K|FSWG1~EUAi;uNBH0Mo`ebw3k_#_injdiVdsq(*}k+ag|?>YNs
znR1yVKyib@ux04w9ASTr(xgRHU~=5-jeTb-;5>`Md@FY}A+1*(PZ1Xu{eoTAn!){t
zW2Sr1K>&g)|8~*m0=58QVlrc2`~>Y6lDZT$5ngwYdTf*=PK`ekS+>IKe6#7$yLdQT
zhV<w9cE#4S(*j%}7S;f$<)TFrfo;okituFZgORah#qiqBMHnbln=Gjo%+SnP#u}(X
zfWm~l)|zRRj8@~Vcv(hp@>QwvK0Xv@kWlE`E&JGHhHz2=1g-TmO_kahk%z&p_UwUu
z1`;^PH9d#(p9agmWb0fjkDG7(#qHH}?3XPjhmmLt_@*`H{)#5Tj-VnER8y0Ms8)nr
z53uDaFAaG-Klw$!d~<WKWVKD?>ZN)K=s;MD&r=e%90`OlgvCQ}X2c=L9Spe4EC1s%
zTi)6OFT7RQ%PPLMclWGcqfTDQhw+5J_C3}a_)y_-pcuUu2U1P$^cdF@cvv}+zvZzz
z+v%Yv38Zs$ZP1On+yo>*;8NbB;xpSG(GQU+ZE<DORXrU^eF~hf*tDQW;xzY$K1wn=
zH5Bv@HLmpU5u|`UV+tuZz<1<@30s{Q%E(Ss-)Iv;hpe93=Hwnk`qF;sWw=RP?pa!{
zGBEUe_6u&~P(ZNuUtP@C0mup!G$nZfF1^|hIt(k6;zAU8FmaM+A5HptziV+>Lxt<-
z!-BqyvC7{`S-w+mT!8H^5X;V#k}Xt@JV`HT?^@dtnyDu6QC*BEQjbs8dYhNm_BSag
zs^)jT2^4HQt?l_ErDw2qKg%qIE=CQIh+80;>v%mjGUQKLyufpjx;2U6@b_yHN@r^K
z^A@`MopSwfS?}xD=U>82_LiLxZAYHc>cCmx(vfZ#(Kz-j2@a4oPQS>(lME7__$O>i
zr~v^&!~3_B#k6sH=eDC|0&mXLg;SF$?oB!DJtwm**ME+;6)I=kk$jKtYr|hBd<%u}
z)&^`xOS-Jpq@fH$`s0k7k{?0BIs39CHWB7UkU(=B*sYGb9Za&5wu#d4D3X<Pl#x%D
zL3n6W>@?aeM5*xe<juHo)xNGO*N~82CpOPt<x6bsTnb@P(Bcb$Rt-h~^cNDQZK!uZ
ztq#;)Ju(2ATQp<C`R3ET<U`$xG7cN<t2y_DpBr&ox9PF=6zol~-lQ_Oe^6^?@F$mQ
z*Wx$#UYo<7BbJRX8ga6PpaZRDTti-h6heJ5cSg`z&_0WU{5rYv$ey(yx3r>2SGgtm
zl~*p7+~xv(>~TRMs3nK5N7v4kO-inB3C~HqO$o^T#U-LCo02p267<kM18M7sRU?Yk
zSv#_JBRcBA!Os<*a+hyC&*L=^m2<T62FnPDF8?=wSk_Q&E6H2-lHc>m6X9t5fYPYz
z7`-H#U&Ah?(8HpT&-kWV`}Vv=d>!(xJX2uI)^RqrU9fe+t~)r$(=R~5+}l&uOfkXB
z?#Fc6vZ*>d1Sn(4<uw>!zw<2tW&n<4s*ODotgp|Y16Lnm3nZkY;8-7_{DFW0Wo=A0
zF(k>=A6PM?mH->a%Hs)C*Y03Xa{=Ydx`vB1mG2(dXz^2k3lYaB(>c|fFX>AX=t?-S
zR(r-iV2B2JOrasLCC|AHBFQpC{Ad4PT*0_C`ZhJf#an_xExqDo`1nxT*DD*AY|_nE
z-?61Dtxu@Y5}6k&Bu<UXhH+O$nj}r0(EnK_&#kIjHT!N#2CZzjQg4<~6mo)kX=(c2
zFwRods`}%xra1#yN-n(D`Un8tIDY@VjdgMZ?3<r+owv5a*g_tXqg?YeMd{|s(P$nL
zhR0rSJnW_OPH=yIdS3&lrNWTt`d_ft#FPO-2WfJ2ocl?Yx2XAjH?YT@Zy$@3Th8;7
zm+)TL{*H?TJ5`yq6!onVhm!V|MQyfd(bs0}&TD3>zXB!STj+W&B+#=dBvHf(>POY5
z-)Jc*NxWJhD9rUnofh@M`_%om=`s(c?>=#4%E4Ph_0lH0K^Bj1ohG65J6)uD`Q^Qk
zZ+fIy;o}+X2)7I48JknG9C!o=+LRjn!5_bl<MF?S)_ZT)e8#Qh=vrg=;)`c*S<`1D
zl5L@D(yeD7QcEVu^PL#H8!u-Y$7ut4BA{_UsnI=q_7vQV^|}rQN|&~)Vgz7Q`qKAY
zFz_0A+&mmpL=8F*23<lePO@Z&iiqk_3q!=nPEmB_ll}fdXFm47%{XncNshdG(K|d?
zA>^NM)ai}nXaw&d{wSWus>#~!`(m@`;dIE2E|pZwo*wehBQ%oatwr8UzLPIXnR($K
zF5GsW=~D&5!}F%F@u)qhmTK$>oi($z@#|%?nJ9I&kHm7go-sHa*QJNo+Cbv);g|27
zbNI+V$c*{s<Bl<Ba?M2!n7P?&D(1mxKmFQT>W*0gCZv;CxVOrH?Pf&HP<cqpu?cdS
z`rSRgJW~2L60CQlq3cL63emg}eTsY+_pTdLKy^A-Rg+y|rZ!sOgvcKl#73c&Ui#Y~
zog#0hIpxbOid}LSF`s_er9G870%YqZ6pdE+qY8T&LN-&n?R{xy^TbJ&f=c8>lb701
zmh?r10fJMvBo!pkk7W8q&=J+_U$0kYj@`a^_`-g~FCsDPCh&52lCxrZKg$*>6f`Mt
z^94nq&N7?h+S;UKZsy<X50#f-y<>}!1w;S(s-v5v+2%F+$L*;O=Q3w}mF1G%R2Ku}
zXoKaEq$OD6DPe_IV3y#Wg2T>{H8zHP+F4%yf)!Uc-YDgv-p#vCOH+S?O#c<k{cc`E
zKV9Dup?uFhT34S19^QaOHB3A8482CnP;sx(<`o0HFI>Hyd>p^bG$7!e!y^sRFggCD
z8mrmZ%dex@_#0sOU_%;jUNE7hyKHf_C~DC&s{_QCZ*(_yvyRz*@$@8_OU^v~n$$Y%
z_4R@&{o6w+DKayK^ktt@?(f0s;DXyT3jU3(ZB=botRG&433`(Y>z}E0oE+cRCw4=m
zRAYF+F)h`<=!K0Etez=W6Qp3~s%%`EWL7meu6XQy@oeyW)eFKu$?R@cDil*OiztM$
z$v|m&vcKR}$FykzYpCvB%t%F3Z#zj>@4Yt1f$JS2KlMCw6b3@}qtEp+<?jvoX*&uF
zt&UX2&5V$|tbBbRH$|~Qfksr@X)QOL3l^@h;Qk}v`n_?*+xgBpKcPund$ySlYaqTa
zjbt2}W8_QDNxHsjF`QoQ3O(Dei_BMwioWH3%kjf?L$@$1x#=w4jW?$HdoTSo5d11&
z=qc909e1ra$`JZv;K@Z8?*vy3>qJnaH6Y<CJnazKubcz#CRLk9M&j!--?b~Ho{$~!
zPN{ealSf+jz6#Y$re0;Tf-Tz21Z>ehg>{ENlUG9Pq$ezboOmjtKivE=__3i%)OLyg
z5ftWqLV~j9_*m8DJx+n~?$R_>bn;r_jpQ92zC7K%F|o){wjS6cOydu#G3YLrOVhRl
zgO$nA%_Fe58u*<6l9|QF-#s%Ei?{W#PCk$)QxRKX?F(4ju<eF8n%}0n5$aKC`eo?6
zx1gMlxU8sjj4U}9?pus$1M`aL`$#ThewJ%I_>WLWvvgmF$V(e?UjD`5z!zR@g$d8B
zYq{GzI8AhdezU&^1#+1vT#q+ktH<+!SUC@8o;JA^okqG%U2yIA=vm@QFP#znWc6j8
zR!<*O>Bfin5A{}D(qzOhaT!fr{;9b%TY)?WSwfONJOt8-4J3^r8H9@32!>Bugt@Nt
z4ZE!eUpNJK_YpE^61;OWSk$Du{A|emAt1e&{DkHq&-onl_V6&$)DYm`xyS*S@na|m
zrYA2CLsT7da%HnD@3k2<N-+>%OF9WTc0>fi(!+8Ya6cXdG4v(U${v3y7o5-!jI;BW
zK+&h+=S|zI5+CjTB@`sEAOy(A9!7JP)o83e`94+ndQ1s4cU69xDR&X}9Md@X&wW>>
z%^ch5Z`5vmfulN^e?N&h`Q^DF5v6?h`|k?(2M!{?+XT0xQLj*?^PkY7<lLKA8EKlb
zk~b0|fr6e$L8h!ThNtX^I*+C3vRvh}&hT;mFFE;X^$_oGzTemUmMZg^{BYhAFxKtW
zV5FZugz@1d4^bmpU0<zZ&0(_cQLnm=or`hF_mf*rrL|$l58S={CS^c1OUjsoaJ&#i
z6dZd>wIQ+JG~_wh!)L^a02HT{J&Ll{YQK++i&i93l2X;5)IE6-1Qog2qEfY`vA4ow
zCGB&PTJpu+iQiWpU=b1u4|x3dI)s}1v%h+8qJ%>D&jz{6(-ro|JI0OQOgcOqF(gU(
zZQo15O~S*=Fr4=)5VC#IF^U^#+v}qzb+y^A%%G=!)%y^M^fB!zVkQUy%Z|EA^(H5y
z-RkV0w{KrQRgvb>u&yhmmk7DhU#CPr7=01{j4AlaX7Q;_*Y#HpQk>hND@m2O0(CuU
z14??L3!^B1<bGyQWM!!B*-EdP6Kswi7;Ab5uAjhBV8MCsetqFy5f<1x1hDQY;Tl(<
z)<2s(-AU$BS}#52+?1XjtDlzc`c$V(x9LNO7Cj1#Ec&)IHZx-))3{{tJzh$ic)M1r
z&WKTQi9k9)ZXzzy6!SEvbXp)jJD4RaQq{`MxRo;lMXqCwHMH-|H#Z*eUeK#5rcKVa
z>Isv|GJ|V>6@6zc&2JZSi`bQTX{5Vp$wGo8rU<i>1$Xy^g4Ab;G}P(;gT1#5ifilE
zMjO`zcWE?efZ*Opun<CUhad^=u8jl;1W$lKf<u7dE<uA!@Ss5(*KV|dhJ9D|e$RK#
zsav=1pYNPs?+=QiYE?6P_3Amtc*Z05i<~|EjC59oT+^Zv34&}+i8CQyXuHO?M|$g@
z;}up2mZ`LB8<T$wxsBH1MqkBNu@SjA?q;H@p4zkN#;Z%wbC(RUT~gfS{f}a`e{f)M
zYDTd_woOS!|2Bs3aN6{}EYMK}L$g33;eJAERpCKWxl8k=c9K4&9<uh8C(daG$#67Q
zAD=3E&b44Q4&Qr<oUqPaO#qX+Cjh2s8;%v^^=JS%@S_ecc87Nll$YH>7AOrcy!?-)
zlmavwsB0~NhM3IEoBV&_I{&7D&WH|*RwVSBL#t@|>jDQ!jrw$Z_Alrnsan2*l9gzK
z^o+UpEbTysXdL+TnKV#6zj|MFM?Qmg-DK9;wR+970?bsa<1EIS9gahQEojIbo__b%
z>K@QGiv26*44blH?S$~4DaUcPO{UO+Dp0|HPTzg_Rz<!gOnh^?9)b^qpPvoMM&{WE
zI8E^lq)bhT8>LHhb7Lzzx<tr<1%r%{-D`AUdI#E&yfrC)60-%dg|F=`u}tb!gb@Lx
z3fV#fOt3hzDmd##Rk-&F<xu`^2vGxktiYG#iA<Z;2J98DRxA@!BvzE)+UY~I>-R8W
z4<&?c-vPJalV5z)*wo^Dst^TS*7-TjTOK909C|Ox1hLogd}hPp3jE1TL+3P1!K%4L
zL5T4q{sKSSzJRnC=X|sMM?dtN(%c1No_}-ON7GitB*{jK!!Qp~Y{-j3lp=!14!U4q
z{QBVO>n47tRQ9D2J~=v7`UC+te;g1BitylKL6*TNQfa6UsGKke>1jSioHK>X#QSvC
z+PM4$q7~kL`i5l5RESup-rvUNhA;_2A_w;SSD%J<8XbQxjFOP$o$2j-Snge_ZB4iO
zTamPFOL?kSkSsw!&@)uhi!tI>7Nr7Dk3G52;OQ}^YJx9sJmt%#^Go90#ntZ3PtdK6
z1E09Bf~ow97{O%wlY`|6O`h2L)J!MRB<9Pvlq$@~jPhqBagGIpMTo9Pc5ZkV;*ey(
zi)jwxP(eCi(AF@Gg7^nJ5`Gmwe!jsk{7?*wTNQJMW`}wZC+s(h2KlQfNTaiJI=8xc
z+MCX>V}dyu)y->!Zzuplr};<>ZF>R~jpoJVN&8}2&*SM!dCCgnD45?dCG`LCgN;%|
z@Ec@Ud*a(ajAznmFe-i<s#%IZ&p0p@DjKAN_`cpS_l|!X=-XT^%c4GR+}l|iv;Ynt
zQE}P9*B9Lz?JAc+$oK0Iig=S3@VRIf_*P#BpG{*F8_8%iL8k{NU04@W!dq);*!uf-
z*)*H6Hg=owu|Qmgwq&>z{oKTAjV{n3;^n+6#O)WE_NlWDef}Q!|J|r~kWXgfkwt+#
z)SV(dnxC&iTcKvm1@ny(PSm0<OvcT>N&zX4NLNE9W0!3$nA$f<dc@eoDxYCIRaqs(
zvBG6#{50+w${xRw3QMJ7Q_xF|!cU$dnFk=Zni*}G@%Ri@ADypnG0~MbOfU}QC>yHc
za~@QRI1WYkJdnH%s$8|tq^8*3L>&^^ga=&6IL?J&pp@V_F~W~sZ*df&dt$!k)SH<o
zlD_5Cy@xw#eK;jn7V5gFddsF9S1Xz7EzS*`g{kM7haWJ<^MjP_Hj*RuMre}?kDY5#
z1*^%xK7VD=knk~GhrKECggwmk-DY5e!a+73-3vU1+K`5QD0T?9{VKWek%A+SM7hf$
zn^N|Q%4(#WkH9OC#9|MF@~XQ<_H+CcYGM^k#$Oi}heCAk+U%K?Nowj|&aWl!Y+~#S
zLY~NhbtdGP!UZ1yeK?m`4;1q!&j|lpFP!9|H*cml*gnkUq&3w$v%Z+*)gMu*X{fH#
zHhR?<V=bpP#H@?<(+nV>j3cnC>~rmWZyNJQ8+~5ZpQf4(IK+K<xLwZ^i~L150?+A-
zAr<;=KfSJNlaq9rOGFK6_?p!m2`^fpfr)(+2MiP(7dl8$iq!yvr1+QDt$hVslr`q%
zX;vRKNSUqx9arK8P*Jhs;8X->+>#em^&GukFyz@+v1g5VH{mV^R-9FNk5aT=hdgkg
z-7dGG)YrOB!A<vUV#3neR|DCEjk7&;G;dqIjEmnw@@z!L)66T<5f^Zv>$2T}Fz?CH
zww!N`4AFqfF>CxDy9!VfnCBD*nZjpV;OSbxFjMH!XFFPgB?;D*5&mIJGl8YH-tOB@
z@iBpU)Q$AqL^ghXGjtv{K=<lnhKO#ez}sV`K(kR9xy1z;?CD)Q;C11x9l-c=l4S~V
z@*9kDXy569kfZoIu?rC)S=Q_c$;BKgH$hrh0>k(vF$$<=-n}S#OJF-D8sEQ>`q4NI
zo}cbhvP?Re(O3GBlks%++HxOzcIZ;}Bib*qBirJ{>r}o@xCj(rNw#$G_<3p}#XMm=
zOI)v~7KJb`N&D&R3lE=w*O^jeT|P<3^Z{0-ldVBy;Q3+HwYFy|9?r18g@oYhi?#_&
zi$4p6PO?rxPnJ7@eXr^(F!YcGBUXDx=qex@F~W`|&mUOw@@(~!r`pBB>|e?=2q*z&
z3@gyukFMZuXVy_=v#GI>W^1P5dDg{HDZqK;&pDI_=~+ec0mo@f<=4d(IupB!qkPp@
zF$ya54itZ^6&{8c{VQL5y+a7uqzLy7tK<hVzWGXRrOTYs_^G{{leWUMx<Wz9PC3Un
zYX+X?Shw~ntSIj343Z2T!<a8&Nd-fn{<SeP0_P$@!FtmjDFf^M9RNKczWphTXLA6U
zZ?Cl@j>hY@nNn5ngQe15eU;Ud1aVkgEUIUyY7!F4i9pf8GY88nE(P3raIj)<PIHOm
zz8Uv_C_eBZvMx7VvM^GWx8t)SWhM3yEKa~JtAB~ZlJlUAjIIn2wG)+b4YF#!BZi)2
z9%$*2bA!HNgBU=6-r6ZJ4ak;+&qYW}*i*#{`!$R=4L_v~swrbV?6zeO(`xd#P3d76
zyuiP9yxfXabX^;I>tdzSrtJ1`qE|tVW9G<p5a=ne68p!al2(?0i6wrqMg&mwI%&z7
zOVb}~Z6&B^4c^4wd~=i`>ZipjTUju6Xw{oi@x(WiW1ulo<Z*WI?miZ>`1I?>6233N
zj0{|V_f<JTcj`!QvL8PgKOFba4xPX?S{A>tId)^|M*yKtZ#OKNmqlM0hbf#Y{DHUl
z07fWRp!Q2Nqt@eyLdVO=rGr+pv?5H`Z8l#w-ftCgHrO~{$c%Mg;ykeh3asu~F}GzS
zq=7<HOJu|VuRaM6>%+IiE!fTci$ojXNNG0V3aQ=>y69eCq2uO+MZcs%PFb_tEY%@t
z5RRN@g3w_q;_9e@E|D4;pioXyPNQfY{xq5u@($GpGKR1A1D)RN&E4DL^~#*W$PDQ8
zS~z7O^s>vahd$7AFFW<3FT|jS{KIInLLysbw-&KY4F?GA_uPKu|FJ0g538d8+rLBf
zfb>Dc1rZQMw$J@vE7W3Lu`94mbxX@4-B5gCxA+*Eh{#I>73%#>&RqcTfdZlo)HQ;t
zZ=BxIxUX*yAzNiv`(y#DXsMgtF`Jy5So}gd`;WsJ0o?%jT>MW29`*;|$@U(aF2Mk#
z50UR-i8u5?KnEV71~|gELV&Ipb{`EU0UQaYp!a+&#+-j&7XSBU>i@dz+lef0kRc*C
zr#JrL$jdnq2K^qLwy0q{{$ul3m<hy!>1?G~b^Stdw1_WrI`9I(wm$-TC}aybT1<sP
zo;fBXyMcjQ0WfgGZ3T8W4Kr^K2sLE$qu;=NH`U)}>t}Di110x=2D!Wa`u(?CYx_F?
ze|2lcLKfXB8zDLR#f-6XqPKpr)#GzyRH8y_NwU6)ippVn4zMBmp96YgyZ8JIejTel
zSH^^`rdQ8aZIjIW_s4Q;v_<(=KQjIu*)f@{PSl@6mSOF+D%pChEyfk^rcwf^?}W8S
z5Z`;0IhT7~g<AAD9Ip8UgRxb2RzhXpB0t=!8Tu*qUA0mqMsLsjOg=5ns_Yn(Ufs~6
zqhoR(r{9SKX{^4@sb^12Y%`&XKK7FEDl2>I)9CKGQy%xt=p*WV)0VdU&R205gb(Ex
zr3%2rHE-f!)<MS@rI=-}{?w90c@<R)4m7Wku=Sl-F|QaGb7eg?mb+{vprCzCbHU`B
zOVo6c1H_S;ey1{?i#`ul-sO}RS^lUauJz@?C_n`rXjjWa+R9Hb8cft{@IUttB^Ajk
zbW6v`m697bzo=3ISvuSk!N(f>TrLtH?p|HJR>-I<&rcl$h@#r=`ojDV?Lxr7P_o*~
zS?u}Q<#nf3gKE&vmj0t$TN?54(hB2$2-Z79)FCrzk{mG&4E!(Pqpfc#J%z#cfvara
z`5(MJUBX}r8)wayG%$giz*2L1+g;Aof@#gTZ>U2isc%)XYnhnDqzvq_(&OlN7CU$m
zC_#Bx74d+yEBX5p%~Wr0l@scO2Y@B(Yw`+&{;^*CsM^<OG(S?v(-T423*D#^jIW5*
z0Zoji_YDH8g!NhO1?j_*@%uk&$8@oV9uUwIzOvZ?dF{%OUHT%k)~l#i9Ws@p*Lfg>
z<|@>!c22*fTPBE$VxDS#qSOkPWF#_1_VgvI;lzr(`>K-gv@OGk!KXHBC!FZn<;nb4
zF3sKwtJm?PTm^%1>o0y;3$vjItLhI-0Q1mU+MIA{u5YXAfu%@m#kRJO5o}kqLE5UD
zgoJfm9HBVt#<Y(j3+)8dup^5M?W+fuR-~l7B8_U|l5%Ze?RJxLU}nDp*gzPc_wAk)
z(xTu17;jlDm-Dc9(YrGJ=(YX5Lp>&G@a=vaOp20#QOulI5cN>5&@?we92`xQG`aPp
z3`U2!Q43v6RT|A4NwT!x!%fs<oR3*jXf`}jR9I4NEdOYq0JxOomQ;}k2qY$Z_xnet
zn{pD@V)<)rG?|!x>T=n<4mkL&>&c_<9q`5oPj4Zr6#IIl*>a*spW35(;h!8T@UbjO
zds^qTj5h?DY(74Fq$xPOdZ3{$`inLFBAwaP9F@+e+M^SyAPlywuc5NabN7-+N|;f5
zMhLlGYk1q3((-sp?fCer+bbusnbVMlLok{wyPsq4J>zUb^7<0=0bg3kq`SStH*3ye
z%}@9&TM7M`)(}pBDv?sq$2!1WEBQsRx*O9Q(_j)>t5h-CO4a)&3h#?raiOXn-tSxX
zYn9SX_qj}vOBG_StBJN@vtXSs#*Zyy^+*gIjd#JyxA~NgFd*mCX+U<vF}Q9u@f=KN
zXABTECYa#2_NP@@)-K;z&^4CKo_z~{(LYW@T@^-6yeYsG5&j_9<n}=icQ;HTQ;hh#
zQU0Lr(HnFtIoVvn&A>!J!kYFUQj<EAQrY4zEyaCBNjl}9()Q5oX9K{&@ieCB4|)@8
zbWe_~<ndcEGST?L&638kDFezS`UHH-*Q|MRkUAlZK`Fju*qT5qq*7H7f`yz`LR5vP
z!Fq0<2k?r15NXQ{dX_J4auW1ePaac128Zp=@vJKx5A#**tq{<2sosMf1*Kg|9~C}(
zqK$z@ixIJ!&55xaphjY&tVS*<(2Z$B3vBEk0N;F)ag^3Q-S}Mu698-g23)`DT|r3F
zEA5tg)%svr;?CENkqSw@Ed3<{Al|M3&vA;GjkfQXU&0*|!&V-t+ZCq%kY_fWsljg|
zsT0L{m75X>()c@al>qR5Iadw>VC5tr2(u#tslL%pH_(CB&^{%+Ra>Jc>an-|WUFg9
zm%1OM|C}TaElF8U0^=;DWBo3HIgS6VwscJ^+v3M@I>96cpZ*p=qS7Oqn=pg62Z2`x
z0jHCKZ!`pMgad#qT0shP!mem2nU41M>4Fbx4WTAQZ?!n%dxG9m{&vRqz<G`~e_q+q
z_WrUw)+hwu5uJK@cm$*I7Y9DO>4!rZJB%Nl>wZYO`*-8fkW=6VS?6X}38Xc@B0IZU
z%`5U3gJXZfP9rNPFtKEHXxGr{NkZl5t;41|MGj6}7oGtC08b&Fwbx_e<L@8iN|UQq
zY5>j`K+^v^KdX`mW)bHF#qy&J1j0q_W6DfSpEZ_Rvy`zNcU9~E#>K!LU}MbUX{*A#
zt7lHJh?ERep=<3c`*BL4t-$|5aPx++n&yIaHdMOco+>&Fx*Lb|Jc|Idy$8Dxnjovm
zHVOEM1|qT{>*pqIT5nwgr1gz;ie3V)DVB?}KfSF))%dp?m5rZ8!yzMJA_9}q4&B}H
zkE~X;>h(Et9#0OE<>(?Vn>Bi530_I%9BJN9o7cdAnC1YLNLjGGELT%qbElahb8REX
zFdy!3-^Pc(`2y*`Jp7V4UT2|#@@Qy&f9Z|1vI*BcEs%e-$!aM8C;DoK5dc|SVdajI
zraoX)-ZpzqisOOvQRq7M!Zij$eV&Nz2UbCl>!;HuUXI#7j%=3Izu+!cufcgk5pAW`
zb-{Xe+>vmXcf@CjWQ7-uHDx||kg3GC$vWbR>8F1V;c<hu26JNMMPD|({YjnEox-LR
z>Z9>o2y2`TgLvo?sASKfajWEQPu5S+cbje!VzZeW=JYt~YMSG4HxqGhR|Bd|aZCxx
zQe*~^k2>Dp6&|`WAvNId#vR?M8>cbJPT~@|6|p)<>Lxo>*3M3I6y@-WQND0q+SKw@
zc!`&~C4C9ta1b#l*cblroKaybTI=kgiq>pL$5hqPg_+>^%UX}iCbvR(4JM`X%V9Tv
zdJ>DYi<{T*V;bi+wDnJXtbWDspEs(pCO@kch>`z$F9vxj{Mj<vzQbJm3HT%Til5#%
zWoF&&y)ndGc~<WD>-~|7ha3iDF)RH|d{F$Ys}d6eC+gAZ>ab`XQxHST?}tJM;hY$*
z&+p0)zs@*s%mi^guK3(+WSj68=%?M>cg)jt9~nU)$UHBE9clYRz~0!WW&N`&nWp!B
zDnrFoaVjZDS1n!nI%W{Wud%=y8bTGtk*OO?ZqM$6|1N?(W(_3u2kxE<e~Pe;dhM5G
z<t&`*QHVamJcu*nLc<7~4dwN-fOUo*1<6|qq4*}x-x?`<6UY;&yDkMpMyd_e;LsdQ
zb#zvZ(D|{qHpi|ka%B1&s?fWazVYDzfh5?nf8AvLFW|r4kBd>kihpbAVv7LBjKmV?
zs>O~Rxby;O?yjjWAU7B3098vQ&*&cDd?!P707|+M^c+JDoSwBS?3+7*{s*%&F!)o;
zrvIN>w&mIWJvnqp8}51+<pgjL0$$?mY5ez6wyd&J1%`u4Nu&leVk`#$3r++S;xy+8
z7rTFU^|tenmFldL_P^J<ppbZg6NUDn%+FC_)`c?IT4|bpM-R8w%y@tYp!8yei4RE!
zPVF($O`IELJJTsodtGPwNG&c7kO}UHv&#+sbO0&|L6ik@dEJ?3z@XJ^CzMxO(Y-j#
znSJX(xtyBx8Q<=gBbtNcA!%PpFh+==pKWi5um*u9Tq!;3gx}hg+fgun`=j=jiIbh&
zj}*zdu^h&V?FblAY?@!;yLOmr-c03|ICE4W$)HHnEw9N>{sDfXAwq3ww~En}+ENyg
zTj^K=#vV`3p<}rdIdCv==g`iY5~23rH3^a5btavHg!uB)$VBSXNzJcmlJN}Cf*RL|
zl|9FcT!I*pceV(kz6W`xcG`Oet|TeTk>iV`8o6mWn`XuI2`xLtG}k;ZthU)sindqj
z0z(9i72Zrn<fYtGbv&;wx#e-1ndTgC*}H{1Y0BARxfsQri~VVIlS1=e8eY;cdH#+C
zU4rTEr!PAg5c<Uh3w~DW=haz9nZ6yTzLE?Tk*7;>;X>Jt`A8AO_VflM=F#MviOs3X
zw~y=$&HN9TEZXqqzpD-on4V43qaqMic+MRh$Xbbw6ps5}e+WP?W>aIo02)@65|(zd
zOGvoAEMIv|A8Q}Q$9@$(b{%mHjz$f@+r1R2(i2btTiyo@=uTW#>9(6|6I7i3DH5@a
zC+i=xDjFFmO_d98eT-nLZ&FoEo|cju6LeGtg!4lEtKcAJ0Ma!7TD<->qrBeZTKOK4
zZ}es!eKUgKi+fm_G}?6TIsd}J#@8^IhNkx9c_Q8-(qUgbWY)T#b8<8+6^~s@_LW&c
zME8LO4YC!lB1o^S+mk$LG~J_YfF<>{)Ho(%Dv%lWluQh$bl0ttmcAw->cd|AdXq`|
z4Tv((X9@&A1Ks}FwTO~~)NbXGvM`gjG)R=>n^HPeg_Y%cHQqHhny|=sd&}_wfY~NO
zQyx(|bKcI;BY<?-m=7+kIBFEuW?n2hEy)pe!{wQ#&yH9x8}UCp-fnD*G6{TVI>07a
zY7igRt&fu=55w+=yo+_-aivDau3IK8Res*8&wDZPlO?I}d<ef6uwam5P%gMNHtvJE
zGv=(*@HvwV_X%f4f1Kj;v&z<5%6Xo74vny&b?RVKD#H^l^(w27kiJSBJNuP?4|gsb
zlZP>Wxd|fkBy2nzdeG}iK3Ss@KU2no5soW^3%Cs=&hO7|m{ifK)04J27vn~JBPB?2
zYq_}0#q&;F#OXfUZ~6fJ@okDfV)W-_S@b(KF-UM&SoJ{Z=AA|$|KSya^(kU3=bQ0v
z2VF!@t>Iu5dqv|k84mSB{QRH+dra%<o$0k9SPD(yAK+S<d$0#fRU81$J96)$b3d6P
zTi5!S!U~$?A5GU+3z7;=i!;G~1JnyKSps;&uhAwdS=z}*%^9O9>g;cykM(ZWuzRRU
zZ#M|H3-+irP`2u&R-{(5ztx}Qwf``)E=N;lUG(4hVQ6DwVbUz4>V78jY^Q3|>pOpe
z7%JLjhv#0@CroQLhR@DQzm3BE!Lqrzu7Z)GVF=(+eTd>1UKA@Jb2OisHK?ym3@lgV
zTVR%B+)UzeXxq7r33tp1zvoO?;bKpcOVAhL^(ikIb&~?&aa04t{sizMj6s_0FHoul
zP-_(D2iZ-EK;Igt#~<_ZSvJ;KyL#L9MvqB><mS=B&d9dA4hGley37FE`yi697HYQ)
ze)H_9BQy={E$>O5oUy1htg;}6tJ1D5yXF^m?a(1ei)><m1uv5eU7d$sGq%t50G!lw
z30Xo-m{c&gqtpWtrMf+Y^_2(uyP7=EgBCB)+AGEgptdjDJt45aioc~r7CtWO7F>;i
zp^IL_j4H0gVX1Urj@(5cea;ymyP8600PG~y6<B<stbpCzD%Nq3t7p;ZY}j)$`(s6y
zk4aH7^WEBp4|=_1qzt-J&m-gGC+^}zz~}b&f$Ng@hiBATg0Q=6A{lXEG@W~@leduF
zBs<8}DfnImh$^P?`_|JRf-~(M|7c>tcrc3DG3?zlh9Ybbma+gYMnX?fh@-&@5Dxo=
z%mhU~*^5vk@wZJ@EHxXpv7)vuk*!BNsQ;H^Ck56;R>NIkW791B_tG=bA;Lih`Tz=P
z+8p7RIDo1c_mAc0uFl`{%)CGIOdMEZf^Zc&QegQ&QsLl<*=XbB{7}<KYRA<WFy+DG
z$lJs}m%Ki^oCyc~qV@z;$P<zs9L9zV-(rh_J}jkRf&9M#{z-nXpzAN~<RZSrSOUY*
zp@5rC)z+yX*L4D(x%z~YXES+fy;N>xQ8<JcT!U=H5E6?Fy&DMJ7}!P0k@|GN_(n(b
zy@zU%q5E%I&}3p)2U+|t7`*>&Wg*4u={!uZ*6HkQLGi-AkG9die%e)<2W&Ka;|Yr_
zE-QCe#2^m540v-AL}4f{%u-%P>WO?HUs%u{fnMl=%j_pIfC9!y`A7n}+V{p2(`LcX
zlt~On$)ElZEXo++L5m+r{CKuE+y0`%?5_E!IFKL5v{102UfWFh^esEdr>A;qyLh<0
z4{tT|avyqX-sMU&W8KxS!oLYSy7K9{O5!mK0&aL)jB?k}T7x^y2j<HN6PW1dHOK=y
z$#Th{#m@)!JI0FJ{;b3JaYmmVtz*dId@rlQo_q+K?TEiCIplwQ&k)3YcI3ixE1Er_
z*&*1W8^z<qY7<m(Du$Eu@rD+(>)h#RXS~hc0{0w1b7S+nczjrvt*S#jgwd!RtQ6qd
zcM^9ACh`g<J$)=_rtNG*IH9zC4N^u=8C^2L**1T2V1b^W{^}c^l^bYZBbesaXkl4#
zy(;QRrXEpl9N5F<H@NU<utWLWg$O5`s0iexZ_!XsYA%<a{V;{@jW_8CW6`!CpFeX@
zUUS$It89Kpa$EHjO698w;Nop)yt>e_082Q=N|k4gTI>pS)0Exr6kv(pPGyau+4P{$
zQ&uON_=BQ8dYi}?Q(PE3c;HG~3$*W?gDjy!N74-IpQSz>JE0^B<lV_!5)8|&^cDAU
zO(TqHJ3bUUijKIJ%^O=RmAR6bDb$uYkt9)*d?6J&b~qwOi?jwt(tB50W`=&g%8jk>
zjuY&htW*^Dsri&#CjZk`{{J>s`ya3O|BczsLIK+xiy0u&b=KkzG#jb+t77HgS$ora
zMw^K?$LtJlKol_Ia!%V)Zl9hhTkLLGYBJ*N>#{8YbOuTC1B&P0-h`6=X$x9ABf+P-
zZFGOc=a4(b)M0}_xY*Oq*?S&8Gp_^RkH-7NW@Bp%H3M|fRs&7=a^pH-a^U=n!M~lh
zv|!*{Ou%XDYh~;Id+=XpHva!V_{SNyAn&xBfm_78))Rn<f;$fbwT+5ePir@ALeWK)
zad%`J_<tlWa!I?xq%?_E{Ghio$35>}C+z&B`=xFfrq}f&+rKKS)9%*(=w1l<@wXzI
zg3+w%NGaM({(5xD+L=D&jg<Wka&CPtoQ6%6)`;B(XnRv&91}KSIRT){#P=4A!og$%
zVmI<|y|mg(oWoNybICGF%#?4UV#9bB7aR2e4(&V@qyw|7$8BS^&--nme6Ks}RL+pB
zo*Klm=Bz(p?zrQcINL<oNgbrkXPE*PMHI>H8+eE*)9e#fIEM^3Hj$xRCjj)wCTiJ*
z(N>S$TE~?oe4v`Q>rb_Ky8(O(giCiADl;z9cZ$!dTR+nU4OL7oF=nt7O$*11Jv)Qc
z3E!$igy`<a@dv)s^t)lZf4Key_<3#RyjumIn?d?BE3>8XiuWsCOBuKSg0T!9(QyCv
zckY)Tkl`ss;vo{BA<|<i>(!z7#=ByIiaV?ewY9@!5=3uJt);#90}O^cDV>WOkU1Om
ztx>8r&f42@wOrbIMQ?P*t#FxcXwO~eFBkno9$zDRSi2v@t=NRA>5hShjy_BQe#!t0
zfgyO-v-tR8<hS`At#@A!%&T^ZweHm`o8%8&=;ii{ucJOCWICr-<V%b^Zd0eW-gm=x
zSpK_8Kuh;sU#2|u%{)<=F3uP;R&!3yNNiz!RAIm9ot8nM_;q1v@#)XDLXvU^ldo<#
z<~UpWCAoy3a~{|k=(JrK^JNGhdz`A-(+PpKRfozbFEc}8EG25|n(7<nrX@{!OYks0
zgkrw{p}t+{&XN>O@fQG;Ai4c-*lO%Aus}hnWg(M`!Z-1w$ta1bN-KY(G{<Opwg`xS
zr;oqPwaB4G-A?b3)WBsxRz+F_WsJys-bAbx>_-NJgt&S;TN1FT4x-jr6UUfW&G8$V
zqTGnqq~WepT$u;K^0!1LFZkLFyeyMU1Wo#DQ>E0=H%nv$(dYw#2;pBkhBM4lUZ0nh
zcO{zY!j9Q+1(&qc>4{A7n+!vwnlewi8JQa!Vz!$<zi3VnpLm@W_Z)=L0yqh~wA@uZ
zA{)AN#qEDZ=tIN*&Z<2IO>(XZ^#ECg)5jpC13lcCKJmJe7dE&36lvk}qasxEw^N+^
zfa%=MW(h(SF<hPy2e1fmSxZDNpodbno($>lk2jlAooxeEH!_mm`e;hI=}aWIc<wa^
z06E0E|2paaV=uA-*hpOKCpVxy3nUI0sRv>3i+BOpa#dY?elIvn3!sII%jgj;sT81H
zY(pYTGZFv9BYu0&K)8Bu^@9B4-lbPiJ~rMZ{S8s;uz60l8vw=gqe2Or1M{M%Jqm!p
zD&{XxzvCYGEu--CJu{N@!L#ufw}h(wQzi;%sta<%mEt^6)N4S<?x#3<#)d1Z!$3!O
zM|g64FKKVxKIu;qZ=P*t)&8w<tyET7fJbhMH0S{|TT@W@LkPS=vv%n&@h~4ZvT~ef
zXkOH*U-W1%&OP=s`Acju6iWxdugawa3a?OiU|!|eN2^GxM9S-)AE~Pqn(GDfA>Zc&
zRSIpaCIY9ZpQM?`_LO@686~cz@=~iZX8g~2HxD<*x+;>psv{S^5E9D9a3g&J<m-?5
zY4c&GlulkWOtwHxoek|VF@sLz;oK7v{RN_{GMMWSwQsfUJ`x*nx1STNEZ|r;I-&{B
z5vL-gY59@Jb=^INLqFHmaPAxL6|4ra9>->#H{)4y;P)2%^!drtaqmWDSfs4gt87`z
z13B7G`0Sj;IR`NHiogKw<H2l@CZhPo<UNUfwnT(RbGFPgrSO%>d1g}7HR10yqQ}>u
z9gNuyjH)78=DmQokXowng0$f+wx$`(GoGhi19ZW~j@5_70Uppc89VeVlt{ze9X>I1
ztqU0Jwccp{xar7&C1TGz<WI;%72u{HUU62e0y2?qa8bx>gWb-qK>NdcBBU`vvrYKx
zI;?`^XVoL5f<!gsXp@Qfe%`7)x%Ny6@|1t>wlfGYV*<QaK?a{uuE+wMo(H7EyL;m$
z=H|oJ;rv1)n(0{YMe*oK;xNpUErdgs8k<Owfjv}Dx_Op*JOxL%OSvOcMWbY}-Q3S}
z3H>@9O(Ixgf|!XDS=~)eRT+Cmw%dg^uKMo4J+j#F%da>5m#u!`>@5xTE%>EdM=I>6
zbeR4zVG-jYJnQV*qZIPIFF*`xRryeR4Zi!;b;u8Enbc9M5e>>K_6^HT)mKeu;KbB_
zkh^Cf@AM<)LM4C9c=(cPjJA?Bo8Y61?)}bI`h?$ev!brWexG|T4Lpno+${q!eV_Ds
zk~R$w(|8oKii=!FlL{x4($@XWP~?~}*h>$6yj>3mM;%q>`?)YrI~Sx#ke9LIexzVJ
zXF|k+vWW=oD9==cX&KYWC#$`@>f?<J>BDj9i5<r}?f7GEpgji+d@EP-?@coAmZ$Ir
zNjRUQ_oljpSSA3(AD%79Qq-bP`VU_o!l(tx4bhLlFy^@pdVtN^y)Ihn*ZavEUgP-s
zlpu+zqvY4x)AB4dtTbC6{S}jEJY*XVp|l;s-SF*IWy(bz2Lt02nqQ0-T-ws!Y#3vz
zumcMmMe+|n=>?Qe)mbs0%Bc&Hg@PC~f+=oaEtn}wlSwz8ma7+kGFEuYndB^`b}`@p
zTB^P)UpUDH3OoFA(a5gUjds4_md!{`L6xsc0T#ggJf0XOoTFYJq2{c7c_}12mp=S7
zk<GiCZKsiE?bgJ<^4O}WAy%z8HJlB=_}<8XFiNZP{{lJln4}^&qAD}-lxn6?d0aYm
zem?A;fBM;#P_g>X#a?wNx`R0du~A(ZZ}p0Q!u#KmoI>Y=Ri$O8W5ZJ$_os9wob1gl
z89!9WpBXJD%7HdB@1FmUdU~P%!8e9m|8o?tq#4=ksl1b44XFTyluQ~ik|7u~Hrny$
z3?^N)H>SKFoL2M3$+1&o^bp3y&gJ6I(FSLYX21Sf)K(Sek|JWASW1C6c|LxFWAqnj
z6_DlNcdFW<r(eou0J!4;<QCq!uJT|1Feu5zhs@vD-eHvg1v>cwd>1txZ5{0l7nc<A
zT-&$$sUW7h=P%Jr6Z?TgamFXWfL3YlKGxx&>?hPtlYGyd0clp%U22NeGUHTK*AS*E
zsCvpb8mfA&eE4A=V`hz*Xx#X}Q9ZqT4E`%935b(g0dW$K>Acahcwm{~qgMii$x-4c
zJAlSjZ#CQ42p1wGqqkH8M;hgaaxdG@8hGIGRXpBy6QEV?p%3s>ceJ#%8=Y;eRu~m{
z?jAY5^~X{rQ<g1cnhfdLZ~XOrkza_^JEupEwwc>jai`48@9-Z|u_oLh_l$F}p>`5M
zo-kEunbO_&w(L_9O=&CIFb^&ddiqz+I^jZN;ybL{ISHa-K440eZqw{(*R%uKIMt^j
z-~Qug%@>%ym7g!QYreN2CkuZsEse0Z7a9<`vsCJ{+nMRs9HqU&0tJls5-oZ_iI4;O
zll4_ZO77S>ZkH90NBd}&smq9|Abj!zK>=PBC*hsoO~hZI_x)j|b}`n%lfp^rn0<5H
zc(1x()oTf`Ro1P}xJxm8Fh!ejdgE!=>3+IUKby9g@@X9_(Mx$u+9RmVd%Cj3I7;7_
zCL`<(6>a_gdTNGO%zD2AZQz_36CEtTm~$^%@K^na1+M*7jJ4b4*T8<d<)l>WE@>Nj
z&uNNZL72!nD!>ZTJ@4oGk_NXwlaLa7#<hv<%WI4pAd%MMG0MEQxfUo7OODr5Xy|lX
zQWYx+SXD>MW!^DGoR~-7G8&owM3_}>gsU_*)qd3Av`7=2$HzPRLJxo%z8(B`rF>ph
zWcvXj8S8^@8tSvwjQBXX4|R{&A~7HSDN$>YRERN)7)G}5ldk7pHCc~;j%q7%c{3Yg
zKlnl%^15H<U=ti8YCy>-iy)^;Qp4<dnboNJ{RKfv@vb~(AnirfeE-<i{i9WLHd2vt
zW3S{PPK#8dEc{a0V#g5WP{|Y0kY7}y#us;$8SBk#b>VMf#hSmA$C0n4fS7C|`h212
zA!41q6@XGuwS9{6$3B5;gED-mod$!uTA~|=ulZp0s|0eOU*AB|_r3RflU=r^mV{$D
z#Nc<s*D>_LxRCfj=%L<gk5yvt&+C{`lfq4rUl)CF?DJo>TreBL7DL4v#ur=n9~d7$
z=IME<W|Kk@Q^~=@$_e8IF@c?O`rC=tRQk^l{6?Kq{Pkg$!I;?lqOlr?ddpNp`n9N$
zN^z#R^8pB%<0VzVo0eePk|m&xBOI+&;!b@nar_`~?oh9ISuF7){;A0`C+lg(<>IAN
zJX*s~e8Fzz?niQ9o2e&@h1sER(mpedIsw{&pwA`uvtmCk-HZL+{9ZVbRo8Uu9`^YS
ztR=5X8|W31?;rjAX0&2&)qnX@|D{zhqTUA9$y}G?Wbota(pS=Z=F)dU&D9$-O2R7H
zr&eGs{EuZ;)bP%&Yj4xa(@yJKX}6e0ROFSdOR#%HEa>K|2J4FyWq;ns4D(wVdD?0h
z928cwvjnQcCW_>oFP`;>DF?f88~<Z8LY|{GUFhRGeMAGuH}O+6OE=06s2W44;j=}h
zA{|SQB^OOvSeleDzj662iqkMLZZtT4L4L4zFA0)FXk>lV6#3Pk>P@J{G$K{cOCfi`
z1~jU|BezET*^bxdRJEt4KF6(2utvWF^E!AFOqcNqod4a$!4bE*T0M1kU`d|a0Sl%r
z9bWesaU3IZCtLPmnpmuI!c&`MJbi%boRFJnd}DCwKG{HXPk_?2bE>3wpybEO!-a-y
z;Dza4NW7-^6$b6xm3uoM)%8$HdG@<lh@;A8fzV!uh+aTU4WMFVtM=L^>iInFy~NFy
zO!!7VWC_3N=eozr<MhJxaWF}M**_N?3^ale+s=0@O7T6Pce9**`k^6DZ5VP^x?u_D
zAb>&T_}HQ;@(==N2MY4`r8Q`87((br-M*&fC0&_ggsZT_QqmV}EKSo_3GD}Gz5k5t
zA(xNqS{jkQ-^SfBlY2Z%s3I5K*j~mKR+S90?mqwp;}`<pE@8$8@Q~uF<!<&~eXZ!z
z0w$r9Ze5#m8p{v!wh>cJ2jfsXI=Tv#wUd~N^2b{(Ep{YnoFD51jI<I!n``J$d=Omc
z*2s~10L4m4b8)K(94G;+(H)`-4msE&n)JW-bov*aI$t-}D`N`x_ER~6*8so5<m59q
zHzBw?mKq^0YdDd^jbb<L!wQbWU>L6T^l_%+PkEQ431xM4{PxBjt(nB-PRL;<s0zqg
z72&wUT?F4<07I|220I9ilS>G$IdapURPpYMQh6F@kL%|a5{0HKAr_xOnl)HX1TIsO
zGVcC^V4?M3G9B0X)!s?=$f3+_gv+`n?{Nyr;}$`*Ww!*+-3K7Cd)g5LfEsEApu6Z6
zD4yM%w;>XV8{qH?@A)IfF=d}9Lv?JlxytF^!ETOcL4OPAjsX+LvF-7-i{i!H6tLvJ
zeZfleN;I|QVIVE&(iqn9*lz_Ay@L0B&Zfb!1-Nf{59)>E+{7414iX$|s_5^^%a2`U
zA6gKovVH1tC!fO9_Q}u5NUT1#!x-vZFK>X}Q1brk6ZbVl{<L+4D^Cw_sHW+JXnZqL
z9clXcdw1#cZAScen>aV2fo{F?@fcw<HK?@((bHKf7RSK;)>P@bR`IyHq)i`F$uBn~
zax?(-cN%C!6hx!{C{->+E~o%YV`Jn=SL>UW2L+1G!DQG?A2I67Y6THV%3I<W)PoE>
zXWE_efHUY4U4_(Iag=k_(N<G4p46u;?Q)7^w%6ja-@5LF(N_NgS?}eXKdb|KI)Zx&
z7rTP&M+tVbs`HkqQSyu%IZ7bgzd%-d&~w7tf3D~cazVxcNz&SnwRl68Jl)+M=fyS7
zJP7ZtIf&!hTcz0bJ_BM~^K3mm;Y%8sD|l?NOIk*cyiMQD^zRb{{H=JPr@kw1h&x-4
zhk!2t(0ScL$4|vHx_5<oTI9Q7VHClwR~G9Bx{m^555=oSY42E(Qugka@VXzfP1R~x
zqsBq%Y}#D<ihu~Q2;?KO1t^+jA1T6NJ+mI7A6v7=^0*Q>i<q9GuhqB|H+K!d24wRr
zL_(I*40QZldkox`hA$hhgz$<0WQVte3P=W`e~Ckbb0oSa?XrS9*`<&9(Sxqax+%iK
z$gZg|<Pm%}t5gO1fI)eYr!aOMjoD0*({gYSGhl9;*Y0$%rnmf50{>e9+lL7>iy|mt
z1s|_0J<@pWk}7<ERlOnB@QY>gEIDDJ4-+tyH}i@YF{G%j$D|KuHitYr%Sp7L0N%a9
z-i=n%bmZM-{;9ts>ruSE3}9LN1$7X1alSM2GpO^{b-{Gx^*Y_)TIvy#WAj$3ZmP2T
z!wF5NBPZa<wlN8rmCRYRy48&0ZHtGEo*vWr{633|v!%GuC``-TPPX@&;B9QuV*kOm
z`w)YZl{NVbEvhq|#)Tt+4xk$qzKWscdh-UDy#sOI27BXgU!~3c2zcDy>r6jsBu`gB
zNMT7EPSjzC{=@FZ$)z&WJG1pm6fq~$&3P5qUG6<^AN<f>UoF4q+p@XKXDT+_WP?Nx
z6OPlHxmtrEXyZju_b=2F|4mhAOwu23fP<D7*YswMEYVty+^ciZ+iIqw@H^Z*gB)GX
zCh*POex*NG61MwvvW3}CtoCzUKIT}v1bW^QYOjNcL&wwov8j!lQ#y4{o9t~{k^yqZ
zmaN3T*|-|beQk=9a(A5Q<&rP52+P<gDU%0_nNuh*Lzwgy|I=IRPYM4&@k2D+k7f9e
z@R&WA?)E8)c-C23-G4(9-$=96e(`OEOkW$mxud_~R)^LI7Y9?qmx%eCFAL9V^f3j*
znJ#m23Tu@GQX~_a36=O4fp1&(6m(TNKR=ZHPgLyLqUP=H<Rt*N?9}I6O_r48;)P5?
zP?sO>Z-j&Mq1Q(ocjpe=oalgA)?Xl*LO=l#54Ov>4+Pk|!L+flp!6d$=(s)Pq~N(f
z1EPJQX5Ui-nA;t9V55id`>BI+ke{CV-OSOvw)cekQdGpZ&4~$S)%4&~`O1c=u+7bZ
z>u|b11LvUyaD16;G#f8z;UX4{#8%!3EKbQvC8-kZs!78@;Aejr3TGC5or^+R@Wh0M
z2RO=;&K$7=P9Ew$KU+5Y)zCBv1{vFE)7N&6h)Vp%2Sfv~>k5pPiZ+^K9rc~czNKA9
zFn-dR?%??WwO1!7<B#iL2J$zV)id<{jONgva$|2@&|rpOmxqj8QW-#L&Lt{<q&qqb
zQKMz$r*>A}Sc#S`p(s0y`FpOADdY4gY0uK#0KOR8$L$f<M30w5_t?OGp@pFKXq|q)
zF|5Bp;j8t)ft(=i?YO>dtJ2VTYvbWhW-Js``OY+oFn~gT(|4c>EOS4R8N8i?i0JKf
zZc0OQ_nIS4H^pPM?x`wSi@OPz(w{plDqdbjIZRdpt4UcI;$v%0Rv96!CuK<qUih<|
znR||qDX|-(_zN?z3j8kaaxv(q!_*2T-|^dMCax~i9flX!FNDIX;3}}Ax%f6$qUeTT
z(y5<SES|0+X>JVUEf1;Bxhn=w5+nw%M_>tA!9^%Tgu?2Qx{C|{cFX$!*l5h}RBxce
zxk<=xO<aIzw6`ID7MYuvsIynZdA5&!>*STsOQzyjeg_H$_KE<_Ue`s<mlJKyWxF@Z
zVU{wDS#Fn6CO`X@VOEuLI-%dg27+E8zC4Uxvl#eXMUNcZrR|8ZeRD-zpLS-H0<ytO
zT2Ew(TT6KB1+1q4W}X%J6<~cEJ5Ww7wcTX(jmr0n^|6%?!e)q-0KuDRmcK8;XYrE2
z-lo2;>ErL`RGaTT1S>U5WpIMPigJJ@S}$bLKx?wiwrtPXAo_f-qMXlsZnW*UjgK<-
zz?L|GhyymhX!jD!ez$!eb^|Ko?QMDF-h`V-!?td_dl=W^NcK(!MpmiI`2Yq_tT|Wg
z&59}zF58?zDIY(0TE|0Vg)6cPty#(Obw_BOST-deJ$Bk88^sx+3Btm>`9l9UR{$Ch
zwQpB{fp`N<!~J|UQpRT2*Hpgq+BiqsbJ!2C(Fa<6cMXNmRfaJ8rJX6woClh4&Hnz~
z!=5Lls*e_dEfZoH4zxU-F~-5O83>GC{U*2?=EjZf0?XQwE$^S62doG<+EUM(jUXP~
z-%I(^4nL4pE&N23YWAn;DdY*!Mf<+dxfre8h<H@s`ezA&)Nkza9c%mb@!iuF-N%H&
zroaB}D{%|c8|V%8)S7H~%(4iarLGNq?4?pb=Oggyg=f0<lO-7eA;eCC@NGL!Kkd)#
z0$28i8D9^V1f^~eU|2@{Hx~wQMxza03X+=@VQ>04ukuyqyJpiSsc?<J^P8kU9SAKW
z(_N{E$ZQpv8vHoZiBh4uVZoD!bB%J*4eucAi`P)U;}q0(HC`u3X;y@RPG>5&wmqEa
z&q8EPlxegguq7U(n_2@+P}g|BEc>*paInu!k4|IS>?xj+O{`uI*BF#VKpwa&25ljI
zw672R4vc<bmFFxBeQmEZwcaGH2Uc<=t}i>kQ<6(m%V{0Sc22Hi^A|{eRZ$KM-tTMx
za-&m6>{sZf?)y!zAMN<3zwuj+h`ZzY6@ni0CVc=7M0aQEiW*3ISsjSxxWjgpsds!-
zCL)&TQ7WbLX7;{-w`*iu+FaTx^;VR8lrn=n#ks+ud(rCiAZvey`U*$(^aX8GOJnmF
zUBPbhmxC9%m>qXz7t{U{)$sdIji1LqKC&l>9H)EtN+|K!CDFpw#?e6UuD6{xgCvjC
z0ea#myeF-)a{FqS6tcf#fy~_5jUShRQiC?(+c-}~$lo_4hjnB6NK}<Tt>_2}UCH-}
zTv--|nW7}#T^r4`yvAukONVFyi6Osa9JSu7PM%!wq$cOqg{#L<3J_2z@yZQhh%+{L
zRHM8dWv;>-Zq>y{o5W~?d$p9?jTS{gM5QdDvPL<V)d;+7Mc4IZE@oaIo*`BNAk`Gp
z2G)TyVNii$K`w*i07tc83xjFe*`1z=&t-)TQYnUwQt_V?e67SV8D;mz*@&X}m@;Sm
zD~|I4s<<b8vTBhLH9ZB>DXz2?Frdy1f_z&J;zkz24PP95+xt2@2t(h{qUhpoIf!71
zJ0H3fM#BU$tY5TF4^9G(QY2~Yis@E3ErUZ>T~(Gybwpl}$r=%<V`lb;OOr&R{<Ggo
zP97=!Bww7ynXBV)rM)C}FxHLoQ}7C7{Z=Hbtr8LhhE*_$188roFWn|C)Dy#V$OuKn
z2u0Mnm};+|Tt*WQ7jjgMb_M$Z+M~BQ4a#k1v3`zEan-t91-=A+IWtGVi8ic*p23%w
zfAwdliKXdW3$bxL5pd0Twv#-734VUng}P5><YJW2N_om|WcaEScq%T%4h@Qzm;88)
z+p<?wk1CI6x>*VF6iW4Z_R(Onxd6L=qJIT^y#?OLFT^Lel*5?7HwF4)&sE|eEusi%
zns3PP?G^QsL#x%z{%-Xe=6^b~G3*ya&vqTh?;-Y0nVkUqPiADFHnnUvgkGyidhD~r
z@}!I51=`w6oUt7E%stc2Jy!+6QVaz7?S_9Xx)%$XycSm0kT_rYs@YNf-N$Xn6Fn2o
zfFtgMxg-yXe$5@*hT!g7`n~j4-yiRzI~voLT}8O~oKbxP86>gn4`pJM2P!!+mfGGy
zcjBN|gO#ouwy_0pvCs=w;`Y)OO0CL4-LRJBU*ur<mZ;!5GmSr1%Rm$z6iZsexp-;F
z-A$VY(><<6H(SgKr?wOl2@1N0Is`+TqUs^Tn)lotACXcpcLYXPW57cFEqz}mUY^cJ
zQhv<5`@OZeQ%TRboH0RMHj;432n)ZEVcOu89JtTG5rEdYaHMT8v#hl_DbF(YdJwt)
z1O)=-mefH`vS5IDZc3Tsl<};G@ag*V1UCZTMZh;E!2L_YY#a2la{Bm!C?YeowkgFE
z6Jh#{)%0V$8t=gTL(SY=B62|6RScvM5d?s@IR2zTYohoOHX)L5#?&f6LesH8hZFiG
z5;$u>SEi){jXY=Oe|JmmvIozRJnJLD3M=Y?@pbB3g91zD{tPc<euV)C`c%j`5$ZJY
z{+i+k$_W`RON(SXz<Yf;Qd&tmwaDTYuB`9WvQN$}_Cl=-Cc$vw=Xw-fw6br3@4WJ3
z(Im2Nrl6(ybWYY(flIgL5-VLcWBeKD6o_d79t7wFhB8pR6#?{{vU!vd;+qW5TH^=U
z=f_+tFx_P5^@fc8`K{|K7t0~7Em_=d2tNh~7{_25O424%wiGgWPu$o%U1{B1(@tfe
zB}hEcyK2`Xu#a;>w)Z^~Q~-3~78|ZNAiLEsohZFvy){=7#f`r}ID5v2q9?8tD0GWQ
z3e<VK$|_n({xH_+J&O#A8!<RU^ZQf>8S>4Dc2mmYF^^UUje+A=!zA28Hx+WU#cLAi
zI41x_C!4z>b~bEs2vI`z{`OYs>2n{h1<EVl&0+AG;BX{>7{(ikz8Z|JHa0~58V{>C
zX@0z!vN&~~v8gxsPVh*3CG-p7P4a;pJ<I*Cd)9~FQ8I-T<4n?_$rnEXr{z9q+}^ro
zbT<85{ZF1F16q{5NSlEQ9x1tj6x`zlH3+9=oEn0soX8fNTqafXAB~ts$fWVZQfIiz
zdTHj5b!Z(gEb0nq6BB&}`g0-0qqsMS-!HL)APvf`wFPSHStLmUyu^criPj$?P9n^E
z5S}nAKe6r*>RI97CzUSJinuL}>IDrf@2w<O>vw$E<Up{}aS%8%FIKHcHBULk;8n>x
zLBnid?psv|wRa`jZ|uaor>egsW7#wISX`yIJv{yN>Saa3mjc4`!7Fk!dPoEG9i`pq
zJ>yh1Q%^u)3KwfL4hPOE&YpP_;xkMVDfA<RW~RP^*Ic7&;PMmcNn`@ky8%C)t4_ZN
z5Ufg?5XJ9|*(63H?@azdlLs?Fe5OtcnB&l=Xgb_zUKRolDi%?zgxQW<slo@Wj(+ol
z;W`w}_`P03@dN8}6crGMEPIzc#~i}lBaoX7iBMVUxRe?jMnW`pNBh)?^-TSU`Yy#R
z&nIo0hoD#GxpA;gk!yc@SCQpW(b*a`k=`-Z1kS(zW9~yWHW<Z?`?`t0v-FdKCDXhF
zKd1GCr!sosAeC)S4DB|@%eTmONRR&1W8phKf`gIe)3g?jk*DSYAPHWz8<5ma;_Ipp
zoz6>i1tmzwmz1Y;6j6qRYeD%6idY;op_*&Koaj-XZ*?8YN=&j8dyEa{dn3kfCEOp4
zo;7jM3_PC$NM^`MclhFgTFq2_a}__ol^EX@?^An5HxRyu;g=f_P7nfqpJnY`y1KMr
z4_9;db=wfni}x7oT?6s#t}bb6w3~VMWX3e6A404|`-5BLIo~+Hy%gVJfCJSsL>)N!
z6OVXYk1UAxb~Mj$axULM#cVx3f2O&V39#|5sysMdRhzC734XSuIsLe(wY|li{$;=5
z*g$pF%s)0`ocaI${qKA5?`PoO&%nQ*fqy>(|2xmXjtfKm2}M`m=fkVg>P<L3H2Hc?
z$*_aBUW`_K2z+zkxx>uhcW?vcHfcBHk6OILB}Bv*OeTW!#<%1s&Z8ZVeD1G&Z=@m3
z?|cULp_;P|2D&x=K}Fxg3is;8#~czq7xhp@Q%;?1sc1}PExBn*=eXoVzHzXC$KQpk
zT8j>^KkS!3vuzSatK0CgnQ*1SBP0u#%}*0AXohb-`Tfi~{%AtzxF36ROyl*C&~Lv7
zg~{(BAL;=?Ld_lb@;EA{^Abqc1LZ!D5#?N~iQ=?5V^R9i;hB7p>C~F~QO94<tgH5X
z(FPRT-`pMu##R(oar%m>=c?fMQ{yW#V3>~0=xECR%NvytHS4)OIjk}<)FZ00md4<$
z7eWILE0qRM8$?4mJ0m;=*_&|<HzQ#f@I9n~uH;J$O1nk<7#l^FxLRx)jNe+qlU<*;
zlUX%Sq(Bhko@zh@*7Nc&P$eEJk{__X-n2-+y8!I48iWxFVcRL<6F>!Djd50DJHval
zi6=Aqr1Q$COC%Adc}+EU-^y;5s8}t!VmiCWBXC{lo^JhnrvhXzTvFp7x6j(|OBR-B
zI!pPXT=^!Uu*snMnv3O~x@1^F0OMV02buOw@X^P=Ky9%!E5vj;>C@N0G|1Fz({+Er
z(iTM%B;@4$7bOPC(HeZ69o<V8OxuTyUXsz`fW+@RBXv4_82P8E+!w8a;^2P5KeZH)
z>56*PFDpSPuEl6n0<FcJ$N$sbd44sq{cAjc6cIs0id2P2FF|?@(h)hJNRZxJ=n#q&
z1JXO83J64yCQWJ}AiV?y0!TS@1q4HruC(}@^Wv`i7o4^3x<6mc+gUSvKfBEA=lgup
zPo+Qz#-x?D*{a4a_vBfJRCTWNRy?%CZCXFBv^nLG^KOw#`}$ROxYzJv9pM;F?fTW#
z3e4c`<*`3}eu}StwpX$?6BLr*Gym-+EVA=wU9zj3o1(hpZ;;r|c(U@k=AM#_ZIhzd
zy}Hlp()*2KR|SOKe7f7xg4e@%_Cr7QmOM8QXIPMp6Q3#Q+*qozm|xbi0{}`mRQ*>@
zt+gn!@}E`b_@f`proGb&{KHkyIVI?c>G$tsIeSGTOl6g7QoW<&_18y(E_bIIs+LRh
z%7(lyP(5NlUmY=KqLH}X8e~V|xs>4y(<*Eq><^m|c6IKsQ=p0BV_$7uGoToZ-u{NE
zm#}D)n(96TF9ZeAO*d8MbYP|s<)#N54|da_x!NU`RL6{+c0SozD059nJ-;>G;~f=-
zDLE-!mTw0IyNmWjLaH~+Q6r5{>%59eR(7QoqfgUIX4YVem@`wzJKR#?5civJ<6#-~
z9T05HK0&E8==xIkE&KgU<SMZZgu~K%W$tu*WzJ#cI#Iz95zEvr{+|gq%q<2mw6oOr
z*j|DL!{}C!fa||{m%BA(7=r>Q9@)#?GGEYer%!u+Q+5z$7jAx%M@r}rE|{Jd@Ra2A
z%q}3odrs%wdq3T-=gl!~Ix+NA`1zN*T_oN{h1qWlR}-QXd@r*{r%7vsjVxKK#Kq-p
zzWi8$en#OLRTbT>?<adMzUF@q+(A*WYjyeIa6AXlC{IG+!SzO95CCTjKzMbQNl_#^
z>#6DneAilr>w&%_b;+wRw0-&*44UAcG?2tIh!Vc?DAI4WmXI4^(s`~GUF1ynL(E@2
zLh&0mWkZGS{U`SEm}qqrou*+blGV9DLLgQ<+TgwcL>K^3GxR=m6RkZO09Ofv6ALzb
z(tGbLTEgY?d^dEy4IuxvI)Dm-I*Vq_W)t4HdwurIXmIFtSAJ~-rIe5^^s&fEnX^js
zCFt#?=;;$Xxy}sRffLj_=Y?lyNP-GwL{6T5tLX6Hj@abJ1g4U#yYPSx2p*{|BdTJ)
zlb}Y}6Svr7w_|7+!v#a>N^XwaQc#?wm&vbe4I4UQjQQGmdeznHc%LX4SbOOR_)x9l
zRM^t=*dHZ;W1~i=Xu3R<0jS?lg^3n4{obUIsCQ$_i0!YBa=dvD<C*z?D3c)(`y-Z~
zw1gx#8|g}1cls}4EBbm+rFXfh8IGr=HChJRGfGk=!fcQ=qc+I|yFX-&XH8$`5Y(r0
zZmuWnFIo<Wii%Rc8o7hFd>ICA%P3?bd}ZHuX8FQNw^Uu35bIvN>@BEXP}4=dZd=F1
zDb^Uz>)F7vSgh*95Rt0XaIs|t!%XyzFKjvrhdoxWJ}B({05WfCZ*IFXug~7HRTtiG
zzcq&j65v<k*Ci>u8brVL?9WWp@A7>vnn4g61L7>!4!%9h232$sPNxbP{x;*#*G=uo
zTW`H~m$-)m)DC98u|R5_E2pwm7h!i$?l`$^lQAf=aP7`!sS4HE8O*UU^*dH-I5$L>
ztb>_fd?t}tT8}s{yxwh((y!Nv9ZueBxVd=R5Auhr@jC#4VpJtHPTe!&?U<A4i5z!s
z>#_9LVZJE*1Qrgm_toN&ALLfJ+e>c!uuzSl-k?E?msZ#NN0j$N$!&gl5><m{E<=z>
zoI<}rCcydF%!&cUTDk@U_YgOaTu-$1HLeF+Mr)H*#f3v@n~d4AbLNh9A+{;Pb5FhH
zbT~sOh(G><Zv_TCa`n7dZCrLA$u%QMQfK7hcTE>|&D)p8Em$JV^_pf!!Mr-qSBLlB
zJ8AH>kfltQ?42aJ^5!v4Z3}q&dMOS(r&z`R)kxBLtKvRo+|<~SgB-;x4>ai$#{_z0
z)#>2S@ln+^FspvtNCr5;to))sMUjV2A4klj&9URy?X1j0t!v$ipfAMg;K2cd(0mZW
zAJaVJ5Y^Q8By%m&OS_Ws(^i%j0UMq7+>TXc$E!{r-DMrx2_418gYYvV*6r_|JvI)w
z!8gxet{zst=hx;ja{u@a<hH@X;mS3DJP=z5H9n0`={p?a%yxD6m}gyAj|o8Y<bpoC
zKk0Tj$lWF3D|;n4`3@X6nZ}s$@<G4D;MX09Y3z%k+bjbI%APerQ~K9+xy|5A_gj4(
zzcHp*w6C7Bg2u+(tE&)SIhqFZslcaE-JBQ6f_=aY*wvske2pP=S`_jk=7TNJNg7N5
zA;)SR4yAU1HT*O8siwD)GLNvm4EvwCZ)YQZUHo2qg|JgGkG&CBJzz=o2(8?q+);)t
z^92&NlAE$i%}VoDJ|(!im9&pN@36A=oj<PVTn_+}jG>pgVba8Lx>w^l2HM5G0ye2i
zaG_Am(G^2>W&@o&8*o|P*3+tq^8txL0cGuQi*!XO4~QrWjS`D*oOe}H@^o9#4UXk_
zCi#$APIhBDOg^_9o0&dpW7a~`xp<Ri+A6-rfz_cAA`WXq*$OL=!MOKMa<hsU+!9r~
zQ!lsA*2AQDas^DL<>+T=S%gTRo}U?R<?cm3dGNflUnqxz7gV<ybZ8;+ug6O906F?N
zsh+v{M}C2gGWgl*T~^-$!X(|yam{W#Xyd}Zo<o_m9PiN}^r&L8oR|r-P}9xraYiKA
zXW8BUY9q*y{1!EbeFx{RT{Ms&sIY&iS1cgmuWTK5^MWqRRTQ_!A9nk^(or_*%=nUM
z^D=QpM+4*w_($YpBCRpzo{+odXUT9jF@|{!x71DUXEm#Ax211BOVO`JOh!FTeF)Io
z(6q)6a$PR;Lq+=B8QKMVq&F*3iRUX=0NYDzWzZro6PL8}v9+N!_AW}(pFX)W(`X-7
z%lW1l*Gzq&T2rbMpZORMa;g~dS0{yNYTPl92Lb7J_`uIZ6@lL%i<{WH6F?12;a>I<
z01MjRY^px)K-I$IUaT|bXQ}&?UGXIx*OcagU3eAyV1s(Sk|g2pAjq4}Ze@Z!s&wV^
zk6nLWwOi4X-u*)FQHO0Hd3}?LScJ8P@ZX_-Ny2#qg3-lSJ0a^VO$Qu;brgG9HhUv3
zg@WQiU7W(7L)z2D@=}g1O%@CFAL`R|%ZIi-ZdbdC4PBAA5R@3Y*b27wSZoRiTFF;o
z2hi`63nIHkmma0}d}p;lNbz6B4}c@~(eGRbkgdWWWyF!k%MH#;chU_4^S5_X!HE$H
zRD;kG#H0bGSf%Ux<=Ib8CBhhHhMLhEVd!H%skXOsc}nt23$I^|2`HC+FnRj){2fIo
zxqR4tMOgMtXJa8aODoykpiCT|V3Ng}oY*|xYv+6qD+r2JJ7n1Hd(m3+E>-ZQw!zh1
za*z*^nLI?UoNQ7@6!`;Xs!R`Tj{QG>Up}L+{-<q`ryQS&AD}08kF6A(bJs8vJRg$A
zNL7m!SP;;SX)KU4gW9KP%*{$yG=<~#!)WR=1>Z@0E?CFvNe~}27ak8mkLY2lQRl_C
z5a=tJYHw)nQ>}3x4><3AJe!oaN54Xxf>Rk5xY{lbf6cgR@{?ukyAG8tX@r8-oh@hU
z{w|2QID_pbISV&dnuUbmA_684kWb*pw@C26v`Wi>)jH);kc)DWTWTjF+AJeyO;BrZ
zPr^d#7RgAU@>lv=VJn}7D*W9i(3o%wky}8JCF6Dw(}x#gy(J&!ui9k@@F)#(Sxpqp
zNYR%sy6CwjyE74>L0vVg1zm@OQZrbgMEPjWNXHBi1tlqfa$sq!lun!eay#-=CAa`0
zXrgH7h`KR8a`L7d!>>khG@`r0=29ajnH4SVm;6Nnb-@!qM=cAc8Ka;`DZwGk8rx5u
zJ_zrM{|rF9!@^oQGd+Z*9MW#1Pu1MPkex(hBRz!i`A1rLdjm>7Pd||#i6VT?1^hIR
zxQ$taTcF~c8<N>iOy>iOHZ4392R?Q>1*mRPD9Np3_aFkdjiOmdmHI*6GeEuMX0Dpa
z{Hq<<7ZTo)0vp_S`7c4|)7c3+k%SosT{TayYF-68s|hSM3&j=46;AXR1uuG<cxSxE
zwhqXUO5{{ofT*m$d-f}OfXy$#0ETIcbw@bI4RH(z6(>t3(z6OiT#m;-iSJ`F(EhHC
zrZ!dDveE#TPN7r$c|@iLYZQ#mfWiGlXmYG+=Yi+c#1bL0?o%VcIZ3G{Ja!FAgUgtj
znSrCHJPdzUwW&!5fu0}Nd>A*knS>niNTNXy*;^#TzlKUo8q+-_Acg#Nz60d?eDQ9@
z3D?<Iw#sHd-k3eYKg{l9`fP4V5TS9~V(bdwd5>FfHZh6?(<DcSK_rtz?TVJP0uIbA
zvjeymMYfENs(m=_kGuI<@0qJe+rPORV=Hc<AvJErmcfp|p=kqj@5AR-;V%qJ-fG)L
zwsgN}4<`&&gg^m|Vzn<3TYFr+2LWCQ*}uFAIQ-A`w=Bqk+9kfke?A;r&T0>D6>vA?
zZs|trKE|#rrrBGW`!O2!8Cq;UIBGMyJ+MJi>K)m&Ay<>BIYl;R8wO$Mhmq2B<~D>6
zv(av!1sPr_xhnb_V~lmb;r(|6^mmiJFJ0Y=vaUCJaL@`Ik&OIuC2Einw+n}-@KnD+
zxHhLF&r7Q5eqaa`o^lLc1|hd6k;6ZL^%Hq6g#a`T1&Y~us+TW=Qr4UJlAvprz%vpf
zg?!>ceKW#u+sDz;t1>qjbHz8c;4hc^Zz!rtH@7mQ?QmDq%(l!}v*(CF9mp)vYK~H9
zofa@DTGne~f^HN$&p58p+@QoEzgT9FFh@QbI5d`#JxoM*&}u$k!J2;D+ew`EpBVCg
zfNe?r#H--c1Rnz`iwyqHUz$K21oeOWH9FOl;11!>3;y^U<ikJkY&yv|*Uc3g4>+qy
z3dq${l)`~E45-=757C8lr!J^@dvc5Z5PjMUIXSMx%)OO14GBoAisqXSLo?tX7OztA
z<f-Pbfmq&u4~|HMRvYf~UBpWlYIkOW;Q1bZZwfh44@pKfH*kM5IM0>R^H}O}h_J^n
zgPlG`1wc57nL*P_LS~5QD^Jx~K-4{Z@8UfC%qldUWN31f!oE|U?x-*y?RYF5tZr|o
z%`ez-%VoPjAHF|})Tz+0?Y<z~z@{pj(*Oo&`@jz{g<v7TU@jE(cW-hh@y!!UByPh7
zko3-_Gnwt#H{l(JmJD(2830XEZuU3|QTGC-4;}yyD7&#b#Uj_P2LA?`CpiL$ZCoOB
zYZh7p(wB*If69j<AL$mn9jTL_bGuu)3Hfkx=UA4ec{otk-s86Z8%C}ebn>!B!?O&~
zc;)K=uzl`G5_xBmg^fj}QpYKX(53|Zh6|II|8LS<_Y`6a*L1hSMT6#>|0jk$FxeT<
zL7uR`D$1605E#?YRW*K#Qkm0Y&)z1G?XNI+w&|mT<+uK+>SWKIyAztdBzBCC6Xbp<
zO1Y3X2CwzmCx!!Qgm2x2o=o0Lg#Uzu=CFB3gI@UV>LFN;2n3aD<9fagbIUH+%>L;d
z(uHoH0G+GuZ)JRhiUA^1w+ph$a{g2Ke0BBP>+4h)4=0b=O4gmEd5bwXYoKup1}BbT
zn>;7_VSwFkP|Qh3rH!PO9g`~td|OR2ran0=R8OZ|^@K>{?7(KP>j&eEw8c_aRGLbi
z<SW%fG*MnGN9<!1O!>UEHl_1mri$QenpH{b<N~PI?8p6<V@Ky1*g}KJ_{$obM((4e
zP-rIHMAd!!OH?ec<Wq1!SHIrBo|yG-(sTdZ`+v#3f7;?tTl~?$9}WD`z#k3#(ZC-K
j{D08^*YDY}a0N;rJp?pPU3SA{^l0v=)Fd_}(naz=zS8e&


From a9ab8f71d7a0d8ab9c41cfad7afa83450f6d41dd Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Tue, 15 Aug 2023 11:19:20 +0800
Subject: [PATCH 36/45] fix ChatGLM RLHF

Former-commit-id: af6c011fcb8ea9e5cf2eb4699da33d8668df04b4
---
 src/llmtuner/tuner/ppo/trainer.py | 2 ++
 src/llmtuner/tuner/rm/trainer.py  | 2 ++
 2 files changed, 4 insertions(+)

diff --git a/src/llmtuner/tuner/ppo/trainer.py b/src/llmtuner/tuner/ppo/trainer.py
index 26a8db99..cc73854f 100644
--- a/src/llmtuner/tuner/ppo/trainer.py
+++ b/src/llmtuner/tuner/ppo/trainer.py
@@ -182,6 +182,8 @@ class PPOPeftTrainer(PPOTrainer, PeftTrainer):
         replace_model(unwrapped_model, target="reward")
         batch = self.prepare_model_inputs(queries, responses)
         _, _, values = self.model(**batch, output_hidden_states=True, return_dict=True)
+        if values.size(0) != batch["input_ids"].size(0):
+            values = torch.transpose(values, 0, 1)
         rewards = [reward for reward in values[:, -1].float().detach().cpu()] # use fp32 type
         replace_model(unwrapped_model, target="default")
         return rewards
diff --git a/src/llmtuner/tuner/rm/trainer.py b/src/llmtuner/tuner/rm/trainer.py
index e69d48a8..55790c07 100644
--- a/src/llmtuner/tuner/rm/trainer.py
+++ b/src/llmtuner/tuner/rm/trainer.py
@@ -42,6 +42,8 @@ class PairwisePeftTrainer(PeftTrainer):
         """
         batch_size = inputs["input_ids"].size(0) // 2
         _, _, values = model(**inputs, output_hidden_states=True, return_dict=True)
+        if values.size(0) != inputs["input_ids"].size(0):
+            values = torch.transpose(values, 0, 1)
         r_accept, r_reject = values[:, -1].split(batch_size, dim=0)
         loss = -torch.log(torch.sigmoid(r_accept - r_reject)).mean()
         return (loss, [loss, r_accept, r_reject]) if return_outputs else loss

From 2ceaecfb422176ede45a3923efb2ac1721079f22 Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Tue, 15 Aug 2023 11:38:21 +0800
Subject: [PATCH 37/45] fix baichuan template #481

Former-commit-id: 273135f59500a36cc30333ef2dd3689c6030e2ef
---
 src/llmtuner/extras/template.py | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)

diff --git a/src/llmtuner/extras/template.py b/src/llmtuner/extras/template.py
index 62153b0d..8ad42ac8 100644
--- a/src/llmtuner/extras/template.py
+++ b/src/llmtuner/extras/template.py
@@ -195,8 +195,14 @@ def get_template_and_fix_tokenizer(
     template = templates.get(name, None)
     assert template is not None, "Template {} does not exist.".format(name)
 
+    additional_special_tokens = template.stop_words
+
     if len(template.stop_words): # inplace method
+        if tokenizer.eos_token_id is not None:
+            additional_special_tokens.append(tokenizer.eos_token)
+
         tokenizer.eos_token = template.stop_words[0]
+        additional_special_tokens.pop(0)
         logger.info("Replace eos token: {}".format(tokenizer.eos_token))
 
     if tokenizer.eos_token_id is None:
@@ -210,7 +216,7 @@ def get_template_and_fix_tokenizer(
             tokenizer.pad_token = tokenizer.eos_token
         logger.info("Add pad token: {}".format(tokenizer.pad_token))
 
-    tokenizer.add_special_tokens(dict(additional_special_tokens=template.stop_words))
+    tokenizer.add_special_tokens(dict(additional_special_tokens=additional_special_tokens))
     return template
 
 

From edc15c62fa798fdd1d9c3aa970f8e0192e61c3a2 Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Wed, 16 Aug 2023 01:35:52 +0800
Subject: [PATCH 38/45] fix system prompt

Former-commit-id: 7407d9daa16bf6b3cd5002e16b2c53e402d2bc39
---
 src/llmtuner/api/app.py                  |  14 +-
 src/llmtuner/chat/stream_chat.py         |  16 +-
 src/llmtuner/dsets/loader.py             |   9 +-
 src/llmtuner/dsets/preprocess.py         |  20 +--
 src/llmtuner/extras/template.py          | 208 +++++++++++++----------
 src/llmtuner/hparams/data_args.py        |  18 +-
 src/llmtuner/tuner/pt/workflow.py        |   1 -
 src/llmtuner/webui/chat.py               |   8 +-
 src/llmtuner/webui/components/chatbot.py |   6 +-
 src/llmtuner/webui/components/eval.py    |   2 +-
 src/llmtuner/webui/components/infer.py   |   2 +-
 src/llmtuner/webui/components/top.py     |   4 +-
 src/llmtuner/webui/components/train.py   |   2 +-
 src/llmtuner/webui/locales.py            |   4 +-
 src/llmtuner/webui/runner.py             |   8 +-
 15 files changed, 170 insertions(+), 152 deletions(-)

diff --git a/src/llmtuner/api/app.py b/src/llmtuner/api/app.py
index 47b7661f..41a7fe9a 100644
--- a/src/llmtuner/api/app.py
+++ b/src/llmtuner/api/app.py
@@ -47,15 +47,15 @@ def create_app(chat_model: ChatModel) -> FastAPI:
 
     @app.post("/v1/chat/completions", response_model=ChatCompletionResponse)
     async def create_chat_completion(request: ChatCompletionRequest):
-        if request.messages[-1].role != Role.USER:
+        if len(request.messages) < 1 or request.messages[-1].role != Role.USER:
             raise HTTPException(status_code=400, detail="Invalid request")
 
         query = request.messages[-1].content
         prev_messages = request.messages[:-1]
         if len(prev_messages) > 0 and prev_messages[0].role == Role.SYSTEM:
-            prefix = prev_messages.pop(0).content
+            system = prev_messages.pop(0).content
         else:
-            prefix = None
+            system = None
 
         history = []
         if len(prev_messages) % 2 == 0:
@@ -64,11 +64,11 @@ def create_app(chat_model: ChatModel) -> FastAPI:
                     history.append([prev_messages[i].content, prev_messages[i+1].content])
 
         if request.stream:
-            generate = predict(query, history, prefix, request)
+            generate = predict(query, history, system, request)
             return EventSourceResponse(generate, media_type="text/event-stream")
 
         response, (prompt_length, response_length) = chat_model.chat(
-            query, history, prefix, temperature=request.temperature, top_p=request.top_p, max_new_tokens=request.max_tokens
+            query, history, system, temperature=request.temperature, top_p=request.top_p, max_new_tokens=request.max_tokens
         )
 
         usage = ChatCompletionResponseUsage(
@@ -85,7 +85,7 @@ def create_app(chat_model: ChatModel) -> FastAPI:
 
         return ChatCompletionResponse(model=request.model, choices=[choice_data], usage=usage)
 
-    async def predict(query: str, history: List[Tuple[str, str]], prefix: str, request: ChatCompletionRequest):
+    async def predict(query: str, history: List[Tuple[str, str]], system: str, request: ChatCompletionRequest):
         choice_data = ChatCompletionResponseStreamChoice(
             index=0,
             delta=DeltaMessage(role=Role.ASSISTANT),
@@ -95,7 +95,7 @@ def create_app(chat_model: ChatModel) -> FastAPI:
         yield chunk.json(exclude_unset=True, ensure_ascii=False)
 
         for new_text in chat_model.stream_chat(
-            query, history, prefix, temperature=request.temperature, top_p=request.top_p, max_new_tokens=request.max_tokens
+            query, history, system, temperature=request.temperature, top_p=request.top_p, max_new_tokens=request.max_tokens
         ):
             if len(new_text) == 0:
                 continue
diff --git a/src/llmtuner/chat/stream_chat.py b/src/llmtuner/chat/stream_chat.py
index e207ee0b..bf602dd5 100644
--- a/src/llmtuner/chat/stream_chat.py
+++ b/src/llmtuner/chat/stream_chat.py
@@ -16,19 +16,19 @@ class ChatModel:
         self.model = dispatch_model(self.model)
         self.model = self.model.eval() # enable evaluation mode
         self.template = get_template_and_fix_tokenizer(data_args.template, self.tokenizer)
-        self.source_prefix = data_args.source_prefix
+        self.system_prompt = data_args.system_prompt
 
     def process_args(
         self,
         query: str,
         history: Optional[List[Tuple[str, str]]] = None,
-        prefix: Optional[str] = None,
+        system: Optional[str] = None,
         **input_kwargs
     ) -> Tuple[Dict[str, Any], int]:
-        prefix = prefix or self.source_prefix
+        system = system or self.system_prompt
 
         prompt, _ = self.template.encode_oneturn(
-            tokenizer=self.tokenizer, query=query, resp="", history=history, prefix=prefix
+            tokenizer=self.tokenizer, query=query, resp="", history=history, system=system
         )
         input_ids = torch.tensor([prompt], device=self.model.device)
         prompt_length = len(input_ids[0])
@@ -68,10 +68,10 @@ class ChatModel:
         self,
         query: str,
         history: Optional[List[Tuple[str, str]]] = None,
-        prefix: Optional[str] = None,
+        system: Optional[str] = None,
         **input_kwargs
     ) -> Tuple[str, Tuple[int, int]]:
-        gen_kwargs, prompt_length = self.process_args(query, history, prefix, **input_kwargs)
+        gen_kwargs, prompt_length = self.process_args(query, history, system, **input_kwargs)
         generation_output = self.model.generate(**gen_kwargs)
         outputs = generation_output.tolist()[0][prompt_length:]
         response = self.tokenizer.decode(outputs, skip_special_tokens=True)
@@ -83,10 +83,10 @@ class ChatModel:
         self,
         query: str,
         history: Optional[List[Tuple[str, str]]] = None,
-        prefix: Optional[str] = None,
+        system: Optional[str] = None,
         **input_kwargs
     ) -> Generator[str, None, None]:
-        gen_kwargs, _ = self.process_args(query, history, prefix, **input_kwargs)
+        gen_kwargs, _ = self.process_args(query, history, system, **input_kwargs)
         streamer = TextIteratorStreamer(self.tokenizer, timeout=60.0, skip_prompt=True, skip_special_tokens=True)
         gen_kwargs["streamer"] = streamer
 
diff --git a/src/llmtuner/dsets/loader.py b/src/llmtuner/dsets/loader.py
index 7588443f..6e293c78 100644
--- a/src/llmtuner/dsets/loader.py
+++ b/src/llmtuner/dsets/loader.py
@@ -92,14 +92,13 @@ def get_dataset(
             if getattr(dataset_attr, column_name) and getattr(dataset_attr, column_name) != column_name:
                 dataset = dataset.rename_column(getattr(dataset_attr, column_name), column_name)
 
-        if dataset_attr.source_prefix: # add prefix
+        if dataset_attr.system_prompt: # add system prompt
             if data_args.streaming:
                 features = dataset.features
-                features["prefix"] = Value(dtype="string", id=None)
-                dataset = dataset.map(lambda _: {"prefix": dataset_attr.source_prefix}, features=features)
+                features["system"] = Value(dtype="string", id=None)
+                dataset = dataset.map(lambda _: {"system": dataset_attr.system_prompt}, features=features)
             else:
-                prefix_data = [dataset_attr.source_prefix] * len(dataset)
-                dataset = dataset.add_column("prefix", prefix_data)
+                dataset = dataset.add_column("system", [dataset_attr.system_prompt] * len(dataset))
 
         all_datasets.append(dataset)
 
diff --git a/src/llmtuner/dsets/preprocess.py b/src/llmtuner/dsets/preprocess.py
index 64e0d8b1..3cc5a483 100644
--- a/src/llmtuner/dsets/preprocess.py
+++ b/src/llmtuner/dsets/preprocess.py
@@ -27,8 +27,8 @@ def preprocess_dataset(
             query, response = examples["prompt"][i], examples["response"][i]
             query = query + "\n" + examples["query"][i] if "query" in examples and examples["query"][i] else query
             history = examples["history"][i] if "history" in examples else None
-            prefix = examples["prefix"][i] if "prefix" in examples else None
-            yield query, response, history, prefix
+            system = examples["system"][i] if "system" in examples else None
+            yield query, response, history, system
 
     def preprocess_pretrain_dataset(examples: Dict[str, List[Any]]) -> Dict[str, Any]:
         # build grouped texts with format `X1 X2 X3 ...` (without <eos>)
@@ -56,10 +56,10 @@ def preprocess_dataset(
         model_inputs = {"input_ids": [], "attention_mask": [], "labels": []}
         max_length = data_args.max_source_length + data_args.max_target_length
 
-        for query, response, history, prefix in construct_example(examples):
+        for query, response, history, system in construct_example(examples):
             input_ids, labels = [], []
 
-            for source_ids, target_ids in template.encode_multiturn(tokenizer, query, response, history, prefix):
+            for source_ids, target_ids in template.encode_multiturn(tokenizer, query, response, history, system):
                 if len(source_ids) > data_args.max_source_length:
                     source_ids = source_ids[:data_args.max_source_length]
                 if len(target_ids) > data_args.max_target_length:
@@ -78,11 +78,11 @@ def preprocess_dataset(
         return model_inputs
 
     def preprocess_unsupervised_dataset(examples: Dict[str, List[Any]]) -> Dict[str, Any]:
-        # build inputs with format `<bos> X` and labels with format `<bos> Y`
+        # build inputs with format `<bos> X` and labels with format `Y <eos>`
         model_inputs = {"input_ids": [], "attention_mask": [], "labels": []}
 
-        for query, response, history, prefix in construct_example(examples):
-            source_ids, target_ids = template.encode_oneturn(tokenizer, query, response, history, prefix)
+        for query, response, history, system in construct_example(examples):
+            source_ids, target_ids = template.encode_oneturn(tokenizer, query, response, history, system)
 
             if len(source_ids) > data_args.max_source_length:
                 source_ids = source_ids[:data_args.max_source_length]
@@ -98,9 +98,9 @@ def preprocess_dataset(
     def preprocess_pairwise_dataset(examples):
         # build input pairs with format `<bos> X`, `Y1 <eos>` and `Y2 <eos>`
         model_inputs = {"prompt_ids": [], "chosen_ids": [], "rejected_ids": []}
-        for query, response, history, prefix in construct_example(examples):
-            prompt_ids, chosen_ids = template.encode_oneturn(tokenizer, query, response[0], history, prefix)
-            _, rejected_ids = template.encode_oneturn(tokenizer, query, response[1], history, prefix)
+        for query, response, history, system in construct_example(examples):
+            prompt_ids, chosen_ids = template.encode_oneturn(tokenizer, query, response[0], history, system)
+            _, rejected_ids = template.encode_oneturn(tokenizer, query, response[1], history, system)
 
             if len(prompt_ids) > data_args.max_source_length:
                 prompt_ids = prompt_ids[:data_args.max_source_length]
diff --git a/src/llmtuner/extras/template.py b/src/llmtuner/extras/template.py
index 8ad42ac8..25907382 100644
--- a/src/llmtuner/extras/template.py
+++ b/src/llmtuner/extras/template.py
@@ -16,9 +16,11 @@ class Template:
 
     prefix: List[Union[str, Dict[str, str]]]
     prompt: List[Union[str, Dict[str, str]]]
+    system: str
     sep: List[Union[str, Dict[str, str]]]
     stop_words: List[str]
     use_history: bool
+    bos_after_prefix: bool
 
     def encode_oneturn(
         self,
@@ -26,18 +28,18 @@ class Template:
         query: str,
         resp: str,
         history: Optional[List[Tuple[str, str]]] = None,
-        prefix: Optional[str] = None
+        system: Optional[str] = None
     ) -> Tuple[List[int], List[int]]:
         r"""
         Returns a single pair of token ids representing prompt and response respectively.
         """
-        prefix, history = self._format(query, resp, history, prefix)
-        encoded_pairs = self._encode(tokenizer, prefix, history)
+        system, history = self._format(query, resp, history, system)
+        encoded_pairs = self._encode(tokenizer, system, history)
         prompt_ids = []
         for query_ids, resp_ids in encoded_pairs[:-1]:
             prompt_ids = prompt_ids + query_ids + resp_ids
-        prompt_ids = prompt_ids + encoded_pairs[-1][0]
-        return prompt_ids, encoded_pairs[-1][1]
+        prompt_ids, answer_ids = prompt_ids + encoded_pairs[-1][0], encoded_pairs[-1][1]
+        return prompt_ids, answer_ids
 
     def encode_multiturn(
         self,
@@ -45,13 +47,13 @@ class Template:
         query: str,
         resp: str,
         history: Optional[List[Tuple[str, str]]] = None,
-        prefix: Optional[str] = None
+        system: Optional[str] = None
     ) -> List[Tuple[List[int], List[int]]]:
         r"""
         Returns multiple pairs of token ids representing prompts and responses respectively.
         """
-        prefix, history = self._format(query, resp, history, prefix)
-        encoded_pairs = self._encode(tokenizer, prefix, history)
+        system, history = self._format(query, resp, history, system)
+        encoded_pairs = self._encode(tokenizer, system, history)
         return encoded_pairs
 
     def _format(
@@ -59,15 +61,15 @@ class Template:
         query: str,
         resp: str,
         history: Optional[List[Tuple[str, str]]] = None,
-        prefix: Optional[str] = None
-    ) -> Tuple[List[Union[str, Dict[str, str]]], List[Tuple[str, str]]]:
+        system: Optional[str] = None
+    ) -> Tuple[str, List[Tuple[str, str]]]:
         r"""
         Aligns inputs to the standard format.
         """
-        prefix = [prefix] if prefix else self.prefix # use prefix if provided
+        system = system or self.system # use system if provided
         history = history if (history and self.use_history) else []
         history = history + [(query, resp)]
-        return prefix, history
+        return system, history
 
     def _get_special_ids(
         self,
@@ -88,7 +90,7 @@ class Template:
     def _encode(
         self,
         tokenizer: "PreTrainedTokenizer",
-        prefix: List[Union[str, Dict[str, str]]],
+        system: str,
         history: List[Tuple[str, str]]
     ) -> List[Tuple[List[int], List[int]]]:
         r"""
@@ -101,8 +103,12 @@ class Template:
         encoded_pairs = []
         for turn_idx, (query, resp) in enumerate(history):
             if turn_idx == 0:
-                if prefix: # has prefix
-                    prefix_ids = bos_ids + self._convert_inputs_to_ids(tokenizer, context=prefix) + sep_ids
+                prefix_ids = self._convert_inputs_to_ids(tokenizer, context=self.prefix, system=system)
+                if len(prefix_ids) != 0: # has prefix
+                    if self.bos_after_prefix:
+                        prefix_ids = prefix_ids + bos_ids + sep_ids
+                    else:
+                        prefix_ids = bos_ids + prefix_ids + sep_ids
                 else:
                     prefix_ids = bos_ids
             else:
@@ -117,8 +123,9 @@ class Template:
         self,
         tokenizer: "PreTrainedTokenizer",
         context: List[Union[str, Dict[str, str]]],
-        query: Optional[str] = "",
-        idx: Optional[str] = ""
+        system: Optional[str] = None,
+        query: Optional[str] = None,
+        idx: Optional[str] = None
     ) -> List[int]:
         r"""
         Converts context to token ids.
@@ -131,13 +138,15 @@ class Template:
         token_ids = []
         for elem in context:
             if isinstance(elem, str):
-                elem = elem.replace("{{query}}", query, 1)
-                elem = elem.replace("{{idx}}", idx, 1)
+                elem = elem.replace("{{system}}", system, 1) if system is not None else elem
+                elem = elem.replace("{{query}}", query, 1) if query is not None else elem
+                elem = elem.replace("{{idx}}", idx, 1) if idx is not None else elem
                 token_ids = token_ids + tokenizer.encode(elem, **kwargs)
             elif isinstance(elem, dict):
                 token_ids = token_ids + [tokenizer.convert_tokens_to_ids(elem.get("token"))]
             else:
                 raise NotImplementedError
+
         return token_ids
 
 
@@ -147,7 +156,7 @@ class Llama2Template(Template):
     def _encode(
         self,
         tokenizer: "PreTrainedTokenizer",
-        prefix: List[Union[str, Dict[str, str]]],
+        system: str,
         history: List[Tuple[str, str]]
     ) -> List[Tuple[List[int], List[int]]]:
         r"""
@@ -157,10 +166,9 @@ class Llama2Template(Template):
         """
         bos_ids, eos_ids = self._get_special_ids(tokenizer)
         encoded_pairs = []
-        assert isinstance(prefix[0], str), "LLaMA-2 template only accepts list containing a single string."
         for turn_idx, (query, resp) in enumerate(history):
-            if turn_idx == 0: # llama2 template has not sep_ids
-                query = prefix[0] + query
+            if turn_idx == 0: # llama2 template has no sep_ids
+                query = self.prefix[0].replace("{{system}}", system) + query
             query_ids = self._convert_inputs_to_ids(tokenizer, context=self.prompt, query=query)
             resp_ids = self._convert_inputs_to_ids(tokenizer, context=[resp])
             encoded_pairs.append((bos_ids + query_ids, resp_ids + eos_ids))
@@ -174,17 +182,21 @@ def register_template(
     name: str,
     prefix: List[Union[str, Dict[str, str]]],
     prompt: List[Union[str, Dict[str, str]]],
+    system: str,
     sep: List[Union[str, Dict[str, str]]],
-    stop_words: List[str],
-    use_history: bool
+    stop_words: Optional[List[str]] = [],
+    use_history: Optional[bool] = True,
+    bos_after_prefix: Optional[bool] = False
 ) -> None:
     template_class = Llama2Template if "llama2" in name else Template
     templates[name] = template_class(
         prefix=prefix,
         prompt=prompt,
+        system=system,
         sep=sep,
         stop_words=stop_words,
-        use_history=use_history
+        use_history=use_history,
+        bos_after_prefix=bos_after_prefix
     )
 
 
@@ -201,7 +213,7 @@ def get_template_and_fix_tokenizer(
         if tokenizer.eos_token_id is not None:
             additional_special_tokens.append(tokenizer.eos_token)
 
-        tokenizer.eos_token = template.stop_words[0]
+        tokenizer.eos_token = additional_special_tokens[0] # use the first stop word as eos token
         additional_special_tokens.pop(0)
         logger.info("Replace eos token: {}".format(tokenizer.eos_token))
 
@@ -229,8 +241,8 @@ register_template(
     prompt=[
         "{{query}}"
     ],
+    system="",
     sep=[],
-    stop_words=[],
     use_history=False
 )
 
@@ -241,17 +253,18 @@ Default template.
 register_template(
     name="default",
     prefix=[
-        "A chat between a curious user and an artificial intelligence assistant. "
-        "The assistant gives helpful, detailed, and polite answers to the user's questions."
+        "{{system}}"
     ],
     prompt=[
         "Human: {{query}}\nAssistant: "
     ],
+    system=(
+        "A chat between a curious user and an artificial intelligence assistant. "
+        "The assistant gives helpful, detailed, and polite answers to the user's questions."
+    ),
     sep=[
         "\n"
-    ],
-    stop_words=[],
-    use_history=True
+    ]
 )
 
 
@@ -263,21 +276,22 @@ Supports: https://huggingface.co/meta-llama/Llama-2-7b-chat-hf
 register_template(
     name="llama2",
     prefix=[
-        "<<SYS>>\nYou are a helpful, respectful and honest assistant. "
+        "<<SYS>>\n{{system}}\n<</SYS>>\n\n"
+    ],
+    prompt=[
+        "[INST] {{query}} [/INST] "
+    ],
+    system=(
+        "You are a helpful, respectful and honest assistant. "
         "Always answer as helpfully as possible, while being safe.  "
         "Your answers should not include any harmful, unethical, "
         "racist, sexist, toxic, dangerous, or illegal content. "
         "Please ensure that your responses are socially unbiased and positive in nature.\n"
         "If a question does not make any sense, or is not factually coherent, "
         "explain why instead of answering something not correct. "
-        "If you don't know the answer to a question, please don't share false information.\n<</SYS>>\n\n"
-    ],
-    prompt=[
-        "[INST] {{query}} [/INST] "
-    ],
-    sep=[],
-    stop_words=[],
-    use_history=True
+        "If you don't know the answer to a question, please don't share false information."
+    ),
+    sep=[]
 )
 
 
@@ -288,14 +302,13 @@ Supports: https://github.com/ymcui/Chinese-LLaMA-Alpaca-2
 register_template(
     name="llama2_zh",
     prefix=[
-        "<<SYS>>\nYou are a helpful assistant. 你是一个乐于助人的助手。\n<</SYS>>\n\n"
+        "<<SYS>>\n{{system}}\n<</SYS>>\n\n"
     ],
     prompt=[
         "[INST] {{query}} [/INST] "
     ],
-    sep=[],
-    stop_words=[],
-    use_history=True
+    system="You are a helpful assistant. 你是一个乐于助人的助手。",
+    sep=[]
 )
 
 
@@ -306,17 +319,18 @@ Supports: https://huggingface.co/tatsu-lab/alpaca-7b-wdiff
 register_template(
     name="alpaca",
     prefix=[
-        "Below is an instruction that describes a task. "
-        "Write a response that appropriately completes the request."
+        "{{system}}"
     ],
     prompt=[
         "### Instruction:\n{{query}}\n\n### Response:\n"
     ],
+    system=(
+        "Below is an instruction that describes a task. "
+        "Write a response that appropriately completes the request."
+    ),
     sep=[
         "\n\n"
-    ],
-    stop_words=[],
-    use_history=True
+    ]
 )
 
 
@@ -327,15 +341,16 @@ Supports: https://huggingface.co/lmsys/vicuna-7b-delta-v1.1
 register_template(
     name="vicuna",
     prefix=[
-        "A chat between a curious user and an artificial intelligence assistant. "
-        "The assistant gives helpful, detailed, and polite answers to the user's questions."
+        "{{system}}"
     ],
     prompt=[
         "USER: {{query}} ASSISTANT: "
     ],
-    sep=[],
-    stop_words=[],
-    use_history=True
+    system=(
+        "A chat between a curious user and an artificial intelligence assistant. "
+        "The assistant gives helpful, detailed, and polite answers to the user's questions."
+    ),
+    sep=[]
 )
 
 
@@ -344,15 +359,16 @@ Supports: https://huggingface.co/BelleGroup/BELLE-LLaMA-EXT-13B
 """
 register_template(
     name="belle",
-    prefix=[],
+    prefix=[
+        "{{system}}"
+    ],
     prompt=[
         "Human: {{query}}\n\nBelle: "
     ],
+    system="",
     sep=[
         "\n\n"
-    ],
-    stop_words=[],
-    use_history=True
+    ]
 )
 
 
@@ -361,15 +377,16 @@ Supports: https://github.com/CVI-SZU/Linly
 """
 register_template(
     name="linly",
-    prefix=[],
+    prefix=[
+        "{{system}}"
+    ],
     prompt=[
         "User: {{query}}\nBot: "
     ],
+    system="",
     sep=[
         "\n"
-    ],
-    stop_words=[],
-    use_history=True
+    ]
 )
 
 
@@ -378,15 +395,16 @@ Supports: https://github.com/Neutralzz/BiLLa
 """
 register_template(
     name="billa",
-    prefix=[],
+    prefix=[
+        "{{system}}"
+    ],
     prompt=[
         "Human: {{query}}\nAssistant: "
     ],
+    system="",
     sep=[
         "\n"
-    ],
-    stop_words=[],
-    use_history=True
+    ]
 )
 
 
@@ -395,18 +413,19 @@ Supports: https://huggingface.co/IDEA-CCNL/Ziya-LLaMA-13B-v1
 """
 register_template(
     name="ziya",
-    prefix=[],
+    prefix=[
+        "{{system}}"
+    ],
     prompt=[
         {"token": "<human>"},
         ":{{query}}\n",
         {"token": "<bot>"},
         ":"
     ],
+    system="",
     sep=[
         "\n"
-    ],
-    stop_words=[],
-    use_history=True
+    ]
 )
 
 
@@ -416,17 +435,18 @@ Supports: https://huggingface.co/qhduan/aquilachat-7b
 register_template(
     name="aquila",
     prefix=[
-        "A chat between a curious human and an artificial intelligence assistant. "
-        "The assistant gives helpful, detailed, and polite answers to the human's questions."
+        "{{system}}"
     ],
     prompt=[
         "Human: {{query}}###Assistant: "
     ],
+    system=(
+        "A chat between a curious human and an artificial intelligence assistant. "
+        "The assistant gives helpful, detailed, and polite answers to the human's questions."
+    ),
     sep=[
         "###"
-    ],
-    stop_words=[],
-    use_history=True
+    ]
 )
 
 
@@ -435,19 +455,21 @@ Supports: https://huggingface.co/internlm/internlm-chat-7b
 """
 register_template(
     name="intern",
-    prefix=[],
+    prefix=[
+        "{{system}}"
+    ],
     prompt=[
         "<|User|>:{{query}}",
         {"token": "<eoh>"},
         "\n<|Bot|>:"
     ],
+    system="",
     sep=[
         "\n"
     ],
     stop_words=[
         "<eoa>"
-    ],
-    use_history=True
+    ]
 )
 
 
@@ -457,17 +479,19 @@ Supports: https://huggingface.co/baichuan-inc/Baichuan-13B-Chat
 register_template(
     name="baichuan",
     prefix=[
-        {"token": "<reserved_102>"} # user token (a little difference in the first turn)
+        "{{system}}",
+        {"token": "<reserved_102>"} # user token
     ],
     prompt=[
         "{{query}}",
         {"token": "<reserved_103>"} # assistant token
     ],
+    system="",
     sep=[],
     stop_words=[
         "<reserved_102>" # user token
     ],
-    use_history=True
+    bos_after_prefix=True
 )
 
 
@@ -479,7 +503,7 @@ register_template(
     name="starchat",
     prefix=[
         {"token": "<|system|>"},
-        "\n",
+        "\n{{system}}",
         {"token": "<|end|>"}
     ],
     prompt=[
@@ -489,13 +513,13 @@ register_template(
         "\n",
         {"token": "<|assistant|>"}
     ],
+    system="",
     sep=[
         "\n"
     ],
     stop_words=[
         "<|end|>"
-    ],
-    use_history=True
+    ]
 )
 
 
@@ -506,7 +530,7 @@ register_template(
     name="chatml",
     prefix=[
         {"token": "<|im_start|>"},
-        "system\nYou are a helpful assistant.",
+        "system\n{{system}}",
         {"token": "<|im_end|>"}
     ],
     prompt=[
@@ -517,13 +541,13 @@ register_template(
         {"token": "<|im_start|>"},
         "assistant\n"
     ],
+    system="You are a helpful assistant.",
     sep=[
         "\n"
     ],
     stop_words=[
         "<|im_end|>"
-    ],
-    use_history=True
+    ]
 )
 
 
@@ -534,14 +558,14 @@ register_template(
     name="chatglm2",
     prefix=[
         {"token": "[gMASK]"},
-        {"token": "sop"}
+        {"token": "sop"},
+        "{{system}}"
     ],
     prompt=[
         "[Round {{idx}}]\n\n问：{{query}}\n\n答："
     ],
+    system="",
     sep=[
         "\n\n"
-    ],
-    stop_words=[],
-    use_history=True
+    ]
 )
diff --git a/src/llmtuner/hparams/data_args.py b/src/llmtuner/hparams/data_args.py
index 7d1c982c..374d03c6 100644
--- a/src/llmtuner/hparams/data_args.py
+++ b/src/llmtuner/hparams/data_args.py
@@ -10,7 +10,7 @@ class DatasetAttr:
     load_from: str
     dataset_name: Optional[str] = None
     dataset_sha1: Optional[str] = None
-    source_prefix: Optional[str] = None
+    system_prompt: Optional[str] = None
 
     def __repr__(self) -> str:
         return self.dataset_name
@@ -86,9 +86,9 @@ class DataArguments:
         default=True,
         metadata={"help": "Whether to ignore the tokens corresponding to padded labels in the loss computation or not."}
     )
-    source_prefix: Optional[str] = field(
+    system_prompt: Optional[str] = field(
         default=None,
-        metadata={"help": "A prefix to add before every source text. Use `|` to separate multiple prefixes in training."}
+        metadata={"help": "System prompt to add before the user query. Use `|` to separate multiple prompts in training."}
     )
     val_size: Optional[float] = field(
         default=0,
@@ -100,12 +100,9 @@ class DataArguments:
         with open(os.path.join(self.dataset_dir, "dataset_info.json"), "r") as f:
             dataset_info = json.load(f)
 
-        if self.source_prefix is not None:
-            prefix_list = self.source_prefix.split("|")
-            prefix_list = prefix_list * len(dataset_names) if len(prefix_list) == 1 else prefix_list
-            assert len(prefix_list) == len(dataset_names), "The number of prefixes should be either identical with datasets or 1."
-        else:
-            prefix_list = [None] * len(dataset_names)
+        prompt_list = self.system_prompt.split("|") if self.system_prompt else [None]
+        prompt_list = prompt_list * (len(dataset_names) // len(prompt_list))
+        assert len(prompt_list) == len(dataset_names), "Number of system prompts should be equal to datasets or 1."
 
         if self.interleave_probs is not None:
             self.interleave_probs = [float(prob.strip()) for prob in self.interleave_probs.split(",")]
@@ -126,12 +123,11 @@ class DataArguments:
                     dataset_sha1=dataset_info[name].get("file_sha1", None)
                 )
 
-            dataset_attr.source_prefix = prefix_list[i]
-
             if "columns" in dataset_info[name]:
                 dataset_attr.prompt = dataset_info[name]["columns"].get("prompt", None)
                 dataset_attr.query = dataset_info[name]["columns"].get("query", None)
                 dataset_attr.response = dataset_info[name]["columns"].get("response", None)
                 dataset_attr.history = dataset_info[name]["columns"].get("history", None)
 
+            dataset_attr.system_prompt = prompt_list[i]
             self.dataset_list.append(dataset_attr)
diff --git a/src/llmtuner/tuner/pt/workflow.py b/src/llmtuner/tuner/pt/workflow.py
index f7bf6448..865ec218 100644
--- a/src/llmtuner/tuner/pt/workflow.py
+++ b/src/llmtuner/tuner/pt/workflow.py
@@ -56,6 +56,5 @@ def run_pt(
             perplexity = float("inf")
 
         metrics["perplexity"] = perplexity
-
         trainer.log_metrics("eval", metrics)
         trainer.save_metrics("eval", metrics)
diff --git a/src/llmtuner/webui/chat.py b/src/llmtuner/webui/chat.py
index e73ea402..154efa5a 100644
--- a/src/llmtuner/webui/chat.py
+++ b/src/llmtuner/webui/chat.py
@@ -26,7 +26,7 @@ class WebChatModel(ChatModel):
         finetuning_type: str,
         quantization_bit: str,
         template: str,
-        source_prefix: str
+        system_prompt: str
     ):
         if self.model is not None:
             yield ALERTS["err_exists"][lang]
@@ -55,7 +55,7 @@ class WebChatModel(ChatModel):
             finetuning_type=finetuning_type,
             quantization_bit=int(quantization_bit) if quantization_bit != "None" else None,
             template=template,
-            source_prefix=source_prefix
+            system_prompt=system_prompt
         )
         super().__init__(args)
 
@@ -73,7 +73,7 @@ class WebChatModel(ChatModel):
         chatbot: List[Tuple[str, str]],
         query: str,
         history: List[Tuple[str, str]],
-        prefix: str,
+        system: str,
         max_new_tokens: int,
         top_p: float,
         temperature: float
@@ -81,7 +81,7 @@ class WebChatModel(ChatModel):
         chatbot.append([query, ""])
         response = ""
         for new_text in self.stream_chat(
-            query, history, prefix, max_new_tokens=max_new_tokens, top_p=top_p, temperature=temperature
+            query, history, system, max_new_tokens=max_new_tokens, top_p=top_p, temperature=temperature
         ):
             response += new_text
             response = self.postprocess(response)
diff --git a/src/llmtuner/webui/components/chatbot.py b/src/llmtuner/webui/components/chatbot.py
index 6fcfc652..928a568c 100644
--- a/src/llmtuner/webui/components/chatbot.py
+++ b/src/llmtuner/webui/components/chatbot.py
@@ -17,7 +17,7 @@ def create_chat_box(
 
         with gr.Row():
             with gr.Column(scale=4):
-                prefix = gr.Textbox(show_label=False)
+                system = gr.Textbox(show_label=False)
                 query = gr.Textbox(show_label=False, lines=8)
                 submit_btn = gr.Button(variant="primary")
 
@@ -31,7 +31,7 @@ def create_chat_box(
 
     submit_btn.click(
         chat_model.predict,
-        [chatbot, query, history, prefix, max_new_tokens, top_p, temperature],
+        [chatbot, query, history, system, max_new_tokens, top_p, temperature],
         [chatbot, history],
         show_progress=True
     ).then(
@@ -41,7 +41,7 @@ def create_chat_box(
     clear_btn.click(lambda: ([], []), outputs=[chatbot, history], show_progress=True)
 
     return chat_box, chatbot, history, dict(
-        prefix=prefix,
+        system=system,
         query=query,
         submit_btn=submit_btn,
         clear_btn=clear_btn,
diff --git a/src/llmtuner/webui/components/eval.py b/src/llmtuner/webui/components/eval.py
index 48372b4c..cbc71daf 100644
--- a/src/llmtuner/webui/components/eval.py
+++ b/src/llmtuner/webui/components/eval.py
@@ -52,7 +52,7 @@ def create_eval_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dict
         top_elems["finetuning_type"],
         top_elems["quantization_bit"],
         top_elems["template"],
-        top_elems["source_prefix"],
+        top_elems["system_prompt"],
         dataset_dir,
         dataset,
         max_source_length,
diff --git a/src/llmtuner/webui/components/infer.py b/src/llmtuner/webui/components/infer.py
index 40e0323e..14aef162 100644
--- a/src/llmtuner/webui/components/infer.py
+++ b/src/llmtuner/webui/components/infer.py
@@ -28,7 +28,7 @@ def create_infer_tab(top_elems: Dict[str, "Component"]) -> Dict[str, "Component"
             top_elems["finetuning_type"],
             top_elems["quantization_bit"],
             top_elems["template"],
-            top_elems["source_prefix"]
+            top_elems["system_prompt"]
         ],
         [info_box]
     ).then(
diff --git a/src/llmtuner/webui/components/top.py b/src/llmtuner/webui/components/top.py
index 7f3c6faa..62c1f9c9 100644
--- a/src/llmtuner/webui/components/top.py
+++ b/src/llmtuner/webui/components/top.py
@@ -28,7 +28,7 @@ def create_top() -> Dict[str, "Component"]:
         with gr.Row():
             quantization_bit = gr.Dropdown(choices=["None", "8", "4"], value="None", scale=1)
             template = gr.Dropdown(choices=list(templates.keys()), value="default", scale=1)
-            source_prefix = gr.Textbox(scale=2)
+            system_prompt = gr.Textbox(scale=2)
 
     lang.change(save_config, [lang, model_name, model_path])
 
@@ -62,5 +62,5 @@ def create_top() -> Dict[str, "Component"]:
         advanced_tab=advanced_tab,
         quantization_bit=quantization_bit,
         template=template,
-        source_prefix=source_prefix
+        system_prompt=system_prompt
     )
diff --git a/src/llmtuner/webui/components/train.py b/src/llmtuner/webui/components/train.py
index 6aaeecbb..aab512ee 100644
--- a/src/llmtuner/webui/components/train.py
+++ b/src/llmtuner/webui/components/train.py
@@ -101,7 +101,7 @@ def create_train_tab(top_elems: Dict[str, "Component"], runner: "Runner") -> Dic
         top_elems["finetuning_type"],
         top_elems["quantization_bit"],
         top_elems["template"],
-        top_elems["source_prefix"],
+        top_elems["system_prompt"],
         training_stage,
         dataset_dir,
         dataset,
diff --git a/src/llmtuner/webui/locales.py b/src/llmtuner/webui/locales.py
index 7a58c4c7..c4032f39 100644
--- a/src/llmtuner/webui/locales.py
+++ b/src/llmtuner/webui/locales.py
@@ -77,7 +77,7 @@ LOCALES = {
             "info": "构建提示词时使用的模板"
         }
     },
-    "source_prefix": {
+    "system_prompt": {
         "en": {
             "label": "System prompt (optional)",
             "info": "A sequence used as the default system prompt."
@@ -455,7 +455,7 @@ LOCALES = {
             "value": "模型未加载，请先加载模型。"
         }
     },
-    "prefix": {
+    "system": {
         "en": {
             "placeholder": "System prompt (optional)"
         },
diff --git a/src/llmtuner/webui/runner.py b/src/llmtuner/webui/runner.py
index 49fed19b..ac74a4c7 100644
--- a/src/llmtuner/webui/runner.py
+++ b/src/llmtuner/webui/runner.py
@@ -69,7 +69,7 @@ class Runner:
         finetuning_type: str,
         quantization_bit: str,
         template: str,
-        source_prefix: str,
+        system_prompt: str,
         training_stage: str,
         dataset_dir: str,
         dataset: List[str],
@@ -114,7 +114,7 @@ class Runner:
             finetuning_type=finetuning_type,
             quantization_bit=int(quantization_bit) if quantization_bit != "None" else None,
             template=template,
-            source_prefix=source_prefix,
+            system_prompt=system_prompt,
             dataset_dir=dataset_dir,
             dataset=",".join(dataset),
             max_source_length=max_source_length,
@@ -170,7 +170,7 @@ class Runner:
         finetuning_type: str,
         quantization_bit: str,
         template: str,
-        source_prefix: str,
+        system_prompt: str,
         dataset_dir: str,
         dataset: List[str],
         max_source_length: int,
@@ -198,7 +198,7 @@ class Runner:
             finetuning_type=finetuning_type,
             quantization_bit=int(quantization_bit) if quantization_bit != "None" else None,
             template=template,
-            source_prefix=source_prefix,
+            system_prompt=system_prompt,
             dataset_dir=dataset_dir,
             dataset=",".join(dataset),
             max_source_length=max_source_length,

From 048f99354faa2a5a26ae2140dd1a7e2591e61733 Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Wed, 16 Aug 2023 22:39:54 +0800
Subject: [PATCH 39/45] fix generation

Former-commit-id: d9e62711a3349d7c6fd3512fb25c709bdfbb311a
---
 src/llmtuner/chat/stream_chat.py   | 2 ++
 src/llmtuner/tuner/ppo/trainer.py  | 2 ++
 src/llmtuner/tuner/sft/workflow.py | 2 ++
 3 files changed, 6 insertions(+)

diff --git a/src/llmtuner/chat/stream_chat.py b/src/llmtuner/chat/stream_chat.py
index bf602dd5..00b25b45 100644
--- a/src/llmtuner/chat/stream_chat.py
+++ b/src/llmtuner/chat/stream_chat.py
@@ -49,6 +49,8 @@ class ChatModel:
             top_p=top_p or gen_kwargs["top_p"],
             top_k=top_k or gen_kwargs["top_k"],
             repetition_penalty=repetition_penalty or gen_kwargs["repetition_penalty"],
+            eos_token_id=self.tokenizer.eos_token_id,
+            pad_token_id=self.tokenizer.pad_token_id,
             logits_processor=get_logits_processor(),
             stopping_criteria=get_stopping_criteria(self.tokenizer.additional_special_tokens_ids)
         ))
diff --git a/src/llmtuner/tuner/ppo/trainer.py b/src/llmtuner/tuner/ppo/trainer.py
index cc73854f..fa4170f6 100644
--- a/src/llmtuner/tuner/ppo/trainer.py
+++ b/src/llmtuner/tuner/ppo/trainer.py
@@ -74,6 +74,8 @@ class PPOPeftTrainer(PPOTrainer, PeftTrainer):
 
         # Keyword arguments for `model.generate`
         gen_kwargs = self.generating_args.to_dict()
+        gen_kwargs["eos_token_id"] = self.tokenizer.eos_token_id
+        gen_kwargs["pad_token_id"] = self.tokenizer.pad_token_id
         gen_kwargs["logits_processor"] = get_logits_processor()
         gen_kwargs["stopping_criteria"] = get_stopping_criteria(self.tokenizer.additional_special_tokens_ids)
 
diff --git a/src/llmtuner/tuner/sft/workflow.py b/src/llmtuner/tuner/sft/workflow.py
index 10f7aafb..f2d72fc6 100644
--- a/src/llmtuner/tuner/sft/workflow.py
+++ b/src/llmtuner/tuner/sft/workflow.py
@@ -52,6 +52,8 @@ def run_sft(
 
     # Keyword arguments for `model.generate`
     gen_kwargs = generating_args.to_dict()
+    gen_kwargs["eos_token_id"] = tokenizer.eos_token_id
+    gen_kwargs["pad_token_id"] = tokenizer.pad_token_id
     gen_kwargs["logits_processor"] = get_logits_processor()
     gen_kwargs["stopping_criteria"] = get_stopping_criteria(tokenizer.additional_special_tokens_ids)
 

From 3021a01b7176aa0e4acb735df9a5540c8a9bcda8 Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Thu, 17 Aug 2023 01:27:20 +0800
Subject: [PATCH 40/45] fix baichuan and intern template

Former-commit-id: 892fd39373b816cf079e0decc9cb57dfb5565242
---
 src/llmtuner/extras/template.py | 24 ++++++++++--------------
 1 file changed, 10 insertions(+), 14 deletions(-)

diff --git a/src/llmtuner/extras/template.py b/src/llmtuner/extras/template.py
index 25907382..5d5a03fb 100644
--- a/src/llmtuner/extras/template.py
+++ b/src/llmtuner/extras/template.py
@@ -20,7 +20,6 @@ class Template:
     sep: List[Union[str, Dict[str, str]]]
     stop_words: List[str]
     use_history: bool
-    bos_after_prefix: bool
 
     def encode_oneturn(
         self,
@@ -75,12 +74,15 @@ class Template:
         self,
         tokenizer: "PreTrainedTokenizer"
     ) -> Tuple[List[int], List[int]]:
-        if tokenizer.bos_token_id:
+        if (
+            tokenizer.bos_token_id is not None
+            and getattr(tokenizer, "add_bos_token", True)
+        ): # baichuan-13b has no bos token
             bos_ids = [tokenizer.bos_token_id]
         else:
             bos_ids = [] # bos token is optional
 
-        if tokenizer.eos_token_id:
+        if tokenizer.eos_token_id is not None:
             eos_ids = [tokenizer.eos_token_id]
         else:
             raise ValueError("EOS token is required.")
@@ -105,10 +107,7 @@ class Template:
             if turn_idx == 0:
                 prefix_ids = self._convert_inputs_to_ids(tokenizer, context=self.prefix, system=system)
                 if len(prefix_ids) != 0: # has prefix
-                    if self.bos_after_prefix:
-                        prefix_ids = prefix_ids + bos_ids + sep_ids
-                    else:
-                        prefix_ids = bos_ids + prefix_ids + sep_ids
+                    prefix_ids = bos_ids + prefix_ids + sep_ids
                 else:
                     prefix_ids = bos_ids
             else:
@@ -185,8 +184,7 @@ def register_template(
     system: str,
     sep: List[Union[str, Dict[str, str]]],
     stop_words: Optional[List[str]] = [],
-    use_history: Optional[bool] = True,
-    bos_after_prefix: Optional[bool] = False
+    use_history: Optional[bool] = True
 ) -> None:
     template_class = Llama2Template if "llama2" in name else Template
     templates[name] = template_class(
@@ -195,8 +193,7 @@ def register_template(
         system=system,
         sep=sep,
         stop_words=stop_words,
-        use_history=use_history,
-        bos_after_prefix=bos_after_prefix
+        use_history=use_history
     )
 
 
@@ -208,7 +205,6 @@ def get_template_and_fix_tokenizer(
     assert template is not None, "Template {} does not exist.".format(name)
 
     additional_special_tokens = template.stop_words
-
     if len(template.stop_words): # inplace method
         if tokenizer.eos_token_id is not None:
             additional_special_tokens.append(tokenizer.eos_token)
@@ -468,6 +464,7 @@ register_template(
         "\n"
     ],
     stop_words=[
+        "</s>", # internlm cannot replace eos token
         "<eoa>"
     ]
 )
@@ -490,8 +487,7 @@ register_template(
     sep=[],
     stop_words=[
         "<reserved_102>" # user token
-    ],
-    bos_after_prefix=True
+    ]
 )
 
 

From 327e14d3ea34b1072e01c4a24ed1f4ca68881fb5 Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Thu, 17 Aug 2023 11:00:22 +0800
Subject: [PATCH 41/45] update readme

Former-commit-id: ff0aa793b6750830b3865c439ef64ed129ec9406
---
 README.md    | 14 +++++++-------
 README_zh.md | 14 +++++++-------
 2 files changed, 14 insertions(+), 14 deletions(-)

diff --git a/README.md b/README.md
index 1124ade2..be43a481 100644
--- a/README.md
+++ b/README.md
@@ -64,13 +64,13 @@
 
 ## Supported Training Approaches
 
-| Approach               | Full-parameter | Partial-parameter | LoRA | QLoRA |
-| ---------------------- | -------------- | ----------------- | ---- | ----- |
-| Pre-Training           | ✅            | ✅                | ✅   | ✅   |
-| Supervised Fine-Tuning | ✅            | ✅                | ✅   | ✅   |
-| Reward Modeling        |                |                   | ✅   | ✅   |
-| PPO Training           |                |                   | ✅   | ✅   |
-| DPO Training           | ✅            |                    | ✅   | ✅   |
+| Approach               |   Full-parameter   | Partial-parameter  |       LoRA         |       QLoRA        |
+| ---------------------- | ------------------ | ------------------ | ------------------ | ------------------ |
+| Pre-Training           | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: |
+| Supervised Fine-Tuning | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: |
+| Reward Modeling        |                    |                    | :white_check_mark: | :white_check_mark: |
+| PPO Training           |                    |                    | :white_check_mark: | :white_check_mark: |
+| DPO Training           | :white_check_mark: |                    | :white_check_mark: | :white_check_mark: |
 
 - Use `--quantization_bit 4/8` argument to enable QLoRA.
 
diff --git a/README_zh.md b/README_zh.md
index 5664c4ec..2a84e697 100644
--- a/README_zh.md
+++ b/README_zh.md
@@ -64,13 +64,13 @@
 
 ## 训练方法
 
-| 方法        | 全参数训练 | 部分参数训练 | LoRA | QLoRA |
-| ---------- | ---------- | ----------- | ---- | ----- |
-| 预训练      | ✅        | ✅         | ✅   | ✅   |
-| 指令监督微调 | ✅        | ✅         | ✅   | ✅   |
-| 奖励模型训练 |           |             | ✅   | ✅   |
-| PPO 训练    |           |             | ✅   | ✅   |
-| DPO 训练    | ✅        |             | ✅   | ✅   |
+| 方法                   |     全参数训练      |    部分参数训练     |       LoRA         |       QLoRA        |
+| ---------------------- | ------------------ | ------------------ | ------------------ | ------------------ |
+| 预训练                 | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: |
+| 指令监督微调            | :white_check_mark: | :white_check_mark: | :white_check_mark: | :white_check_mark: |
+| 奖励模型训练            |                    |                    | :white_check_mark: | :white_check_mark: |
+| PPO 训练               |                    |                    | :white_check_mark: | :white_check_mark: |
+| DPO 训练               | :white_check_mark: |                    | :white_check_mark: | :white_check_mark: |
 
 - 使用 `--quantization_bit 4/8` 参数来启用 QLoRA 训练。
 

From a46f27747772f5baa5b279212a0ae3e01422fdd8 Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Thu, 17 Aug 2023 17:59:26 +0800
Subject: [PATCH 42/45] fix streaming in pt stage #548 #549

Former-commit-id: b0ed0dec5e6788a0344c09a6cc58d1116265fd68
---
 src/llmtuner/dsets/loader.py     | 39 ++++++--------------------------
 src/llmtuner/dsets/preprocess.py | 10 ++++----
 src/llmtuner/dsets/utils.py      | 31 ++++++++++++++++++++++++-
 src/llmtuner/tuner/sft/metric.py |  3 +--
 4 files changed, 43 insertions(+), 40 deletions(-)

diff --git a/src/llmtuner/dsets/loader.py b/src/llmtuner/dsets/loader.py
index 6e293c78..08c35f27 100644
--- a/src/llmtuner/dsets/loader.py
+++ b/src/llmtuner/dsets/loader.py
@@ -1,48 +1,25 @@
 import os
-import hashlib
-from typing import TYPE_CHECKING, List, Optional
+from typing import TYPE_CHECKING, List, Union
 
-from datasets import Value, concatenate_datasets, interleave_datasets, load_dataset
+from datasets import concatenate_datasets, interleave_datasets, load_dataset
 
+from llmtuner.dsets.utils import checksum, EXT2TYPE
 from llmtuner.extras.logging import get_logger
 
 if TYPE_CHECKING:
-    from datasets import Dataset
+    from datasets import Dataset, IterableDataset
     from llmtuner.hparams import ModelArguments, DataArguments
 
 
 logger = get_logger(__name__)
 
 
-EXT2TYPE = {
-    "csv": "csv",
-    "json": "json",
-    "jsonl": "json",
-    "txt": "text"
-}
-
-
-def checksum(data_files: List[str], file_sha1: Optional[str] = None) -> None:
-    if file_sha1 is None:
-        logger.warning("Checksum failed: missing SHA-1 hash value in dataset_info.json.")
-        return
-
-    if len(data_files) != 1:
-        logger.warning("Checksum failed: too many files.")
-        return
-
-    with open(data_files[0], "rb") as f:
-        sha1 = hashlib.sha1(f.read()).hexdigest()
-        if sha1 != file_sha1:
-            logger.warning("Checksum failed: mismatched SHA-1 hash value at {}.".format(data_files[0]))
-
-
 def get_dataset(
     model_args: "ModelArguments",
     data_args: "DataArguments"
-) -> "Dataset":
+) -> Union["Dataset", "IterableDataset"]:
     max_samples = data_args.max_samples
-    all_datasets: List["Dataset"] = [] # support multiple datasets
+    all_datasets: List[Union["Dataset", "IterableDataset"]] = [] # support multiple datasets
 
     for dataset_attr in data_args.dataset_list:
         logger.info("Loading dataset {}...".format(dataset_attr))
@@ -94,9 +71,7 @@ def get_dataset(
 
         if dataset_attr.system_prompt: # add system prompt
             if data_args.streaming:
-                features = dataset.features
-                features["system"] = Value(dtype="string", id=None)
-                dataset = dataset.map(lambda _: {"system": dataset_attr.system_prompt}, features=features)
+                dataset = dataset.map(lambda _: {"system": dataset_attr.system_prompt})
             else:
                 dataset = dataset.add_column("system", [dataset_attr.system_prompt] * len(dataset))
 
diff --git a/src/llmtuner/dsets/preprocess.py b/src/llmtuner/dsets/preprocess.py
index 3cc5a483..1fc146f8 100644
--- a/src/llmtuner/dsets/preprocess.py
+++ b/src/llmtuner/dsets/preprocess.py
@@ -1,25 +1,25 @@
 import tiktoken
-from typing import TYPE_CHECKING, Any, Dict, Generator, List, Literal
+from typing import TYPE_CHECKING, Any, Dict, Generator, List, Literal, Union
 from itertools import chain
 
 from llmtuner.extras.constants import IGNORE_INDEX
 from llmtuner.extras.template import get_template_and_fix_tokenizer
 
 if TYPE_CHECKING:
-    from datasets import Dataset
+    from datasets import Dataset, IterableDataset
     from transformers import Seq2SeqTrainingArguments
     from transformers.tokenization_utils import PreTrainedTokenizer
     from llmtuner.hparams import DataArguments
 
 
 def preprocess_dataset(
-    dataset: "Dataset",
+    dataset: Union["Dataset", "IterableDataset"],
     tokenizer: "PreTrainedTokenizer",
     data_args: "DataArguments",
     training_args: "Seq2SeqTrainingArguments",
     stage: Literal["pt", "sft", "rm", "ppo"]
-) -> "Dataset":
-    column_names = list(dataset.column_names)
+) -> Union["Dataset", "IterableDataset"]:
+    column_names = list(next(iter(dataset)).keys())
     template = get_template_and_fix_tokenizer(data_args.template, tokenizer)
 
     def construct_example(examples: Dict[str, List[Any]]) -> Generator[Any, None, None]:
diff --git a/src/llmtuner/dsets/utils.py b/src/llmtuner/dsets/utils.py
index f816649e..bf337014 100644
--- a/src/llmtuner/dsets/utils.py
+++ b/src/llmtuner/dsets/utils.py
@@ -1,4 +1,7 @@
-from typing import TYPE_CHECKING, Dict, Union
+import hashlib
+from typing import TYPE_CHECKING, Dict, List, Optional, Union
+
+from llmtuner.extras.logging import get_logger
 
 if TYPE_CHECKING:
     from datasets import Dataset, IterableDataset
@@ -6,6 +9,32 @@ if TYPE_CHECKING:
     from llmtuner.hparams import DataArguments
 
 
+logger = get_logger(__name__)
+
+
+EXT2TYPE = {
+    "csv": "csv",
+    "json": "json",
+    "jsonl": "json",
+    "txt": "text"
+}
+
+
+def checksum(data_files: List[str], file_sha1: Optional[str] = None) -> None:
+    if file_sha1 is None:
+        logger.warning("Checksum failed: missing SHA-1 hash value in dataset_info.json.")
+        return
+
+    if len(data_files) != 1:
+        logger.warning("Checksum failed: too many files.")
+        return
+
+    with open(data_files[0], "rb") as f:
+        sha1 = hashlib.sha1(f.read()).hexdigest()
+        if sha1 != file_sha1:
+            logger.warning("Checksum failed: mismatched SHA-1 hash value at {}.".format(data_files[0]))
+
+
 def split_dataset(
     dataset: Union["Dataset", "IterableDataset"],
     data_args: "DataArguments",
diff --git a/src/llmtuner/tuner/sft/metric.py b/src/llmtuner/tuner/sft/metric.py
index 663b037d..812896ee 100644
--- a/src/llmtuner/tuner/sft/metric.py
+++ b/src/llmtuner/tuner/sft/metric.py
@@ -25,7 +25,7 @@ class ComputeMetrics:
         Uses the model predictions to compute metrics.
         """
         preds, labels = eval_preds
-        score_dict = {"accuracy": [], "rouge-1": [], "rouge-2": [], "rouge-l": [], "bleu-4": []}
+        score_dict = {"rouge-1": [], "rouge-2": [], "rouge-l": [], "bleu-4": []}
 
         preds = np.where(preds != IGNORE_INDEX, preds, self.tokenizer.pad_token_id)
         labels = np.where(labels != IGNORE_INDEX, labels, self.tokenizer.pad_token_id)
@@ -49,6 +49,5 @@ class ComputeMetrics:
 
             bleu_score = sentence_bleu([list(label)], list(pred), smoothing_function=SmoothingFunction().method3)
             score_dict["bleu-4"].append(round(bleu_score * 100, 4))
-            score_dict["accuracy"].append(float(len(label) != 0 and pred[:len(label)] == label))
 
         return {k: float(np.mean(v)) for k, v in score_dict.items()}

From 623a34b16f7961c0f742f1fcdf1acdbc27159433 Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Thu, 17 Aug 2023 22:21:34 +0800
Subject: [PATCH 43/45] fix generation bug #532

Former-commit-id: be21fc83f9aed0af1e5a2f83f5d5eeb36f1d283c
---
 src/llmtuner/chat/stream_chat.py   |  7 +++----
 src/llmtuner/extras/misc.py        | 32 +++---------------------------
 src/llmtuner/tuner/ppo/trainer.py  |  5 ++---
 src/llmtuner/tuner/sft/trainer.py  | 12 +++++------
 src/llmtuner/tuner/sft/workflow.py |  5 ++---
 5 files changed, 15 insertions(+), 46 deletions(-)

diff --git a/src/llmtuner/chat/stream_chat.py b/src/llmtuner/chat/stream_chat.py
index 00b25b45..b6d71fcf 100644
--- a/src/llmtuner/chat/stream_chat.py
+++ b/src/llmtuner/chat/stream_chat.py
@@ -3,7 +3,7 @@ from typing import Any, Dict, Generator, List, Optional, Tuple
 from threading import Thread
 from transformers import TextIteratorStreamer
 
-from llmtuner.extras.misc import dispatch_model, get_logits_processor, get_stopping_criteria
+from llmtuner.extras.misc import dispatch_model, get_logits_processor
 from llmtuner.extras.template import get_template_and_fix_tokenizer
 from llmtuner.tuner.core import get_infer_args, load_model_and_tokenizer
 
@@ -49,10 +49,9 @@ class ChatModel:
             top_p=top_p or gen_kwargs["top_p"],
             top_k=top_k or gen_kwargs["top_k"],
             repetition_penalty=repetition_penalty or gen_kwargs["repetition_penalty"],
-            eos_token_id=self.tokenizer.eos_token_id,
+            eos_token_id=[self.tokenizer.eos_token_id] + self.tokenizer.additional_special_tokens_ids,
             pad_token_id=self.tokenizer.pad_token_id,
-            logits_processor=get_logits_processor(),
-            stopping_criteria=get_stopping_criteria(self.tokenizer.additional_special_tokens_ids)
+            logits_processor=get_logits_processor()
         ))
 
         if max_length:
diff --git a/src/llmtuner/extras/misc.py b/src/llmtuner/extras/misc.py
index b3e0e4b1..db91b337 100644
--- a/src/llmtuner/extras/misc.py
+++ b/src/llmtuner/extras/misc.py
@@ -2,9 +2,8 @@ import torch
 from typing import TYPE_CHECKING, List, Optional, Tuple
 from transformers import (
     LogitsProcessor,
-    LogitsProcessorList,
-    StoppingCriteria,
-    StoppingCriteriaList
+    InfNanRemoveLogitsProcessor,
+    LogitsProcessorList
 )
 
 from llmtuner.extras.constants import LAYERNORM_NAMES
@@ -33,37 +32,12 @@ class AverageMeter:
         self.avg = self.sum / self.count
 
 
-class InvalidScoreLogitsProcessor(LogitsProcessor):
-
-    def __call__(self, input_ids: torch.LongTensor, scores: torch.FloatTensor) -> torch.FloatTensor:
-        if torch.isnan(scores).any() or torch.isinf(scores).any():
-            scores.zero_()
-            scores[..., 0] = 1.0
-        return scores
-
-
 def get_logits_processor() -> LogitsProcessorList:
     logits_processor = LogitsProcessorList()
-    logits_processor.append(InvalidScoreLogitsProcessor())
+    logits_processor.append(InfNanRemoveLogitsProcessor())
     return logits_processor
 
 
-class StopWordsCriteria(StoppingCriteria):
-
-    def __init__(self, stop_ids: List[int]) -> None:
-        super().__init__()
-        self.stop_ids = stop_ids
-
-    def __call__(self, input_ids: torch.LongTensor, scores: torch.FloatTensor, **kwargs) -> bool:
-        return any([stop_id in input_ids[:, -1] for stop_id in self.stop_ids])
-
-
-def get_stopping_criteria(stop_ids: List[int]) -> StoppingCriteriaList:
-    stopping_criteria = StoppingCriteriaList()
-    stopping_criteria.append(StopWordsCriteria(stop_ids))
-    return stopping_criteria
-
-
 def count_parameters(model: torch.nn.Module) -> Tuple[int, int]:
     r"""
     Returns the number of trainable parameters and number of all parameters in the model.
diff --git a/src/llmtuner/tuner/ppo/trainer.py b/src/llmtuner/tuner/ppo/trainer.py
index fa4170f6..4474b5bb 100644
--- a/src/llmtuner/tuner/ppo/trainer.py
+++ b/src/llmtuner/tuner/ppo/trainer.py
@@ -10,7 +10,7 @@ from trl import PPOTrainer
 from trl.core import LengthSampler
 
 from llmtuner.extras.logging import get_logger
-from llmtuner.extras.misc import AverageMeter, count_parameters, get_logits_processor, get_stopping_criteria
+from llmtuner.extras.misc import AverageMeter, count_parameters, get_logits_processor
 from llmtuner.tuner.core.trainer import PeftTrainer
 from llmtuner.tuner.ppo.utils import cast_layernorm_dtype, replace_model
 
@@ -74,10 +74,9 @@ class PPOPeftTrainer(PPOTrainer, PeftTrainer):
 
         # Keyword arguments for `model.generate`
         gen_kwargs = self.generating_args.to_dict()
-        gen_kwargs["eos_token_id"] = self.tokenizer.eos_token_id
+        gen_kwargs["eos_token_id"] = [self.tokenizer.eos_token_id] + self.tokenizer.additional_special_tokens_ids
         gen_kwargs["pad_token_id"] = self.tokenizer.pad_token_id
         gen_kwargs["logits_processor"] = get_logits_processor()
-        gen_kwargs["stopping_criteria"] = get_stopping_criteria(self.tokenizer.additional_special_tokens_ids)
 
         length_sampler = LengthSampler(max_target_length // 2, max_target_length)
         unwrapped_model: "AutoModelForCausalLMWithValueHead" = self.accelerator.unwrap_model(self.model)
diff --git a/src/llmtuner/tuner/sft/trainer.py b/src/llmtuner/tuner/sft/trainer.py
index 6243928f..1ddaec1f 100644
--- a/src/llmtuner/tuner/sft/trainer.py
+++ b/src/llmtuner/tuner/sft/trainer.py
@@ -50,9 +50,10 @@ class Seq2SeqPeftTrainer(PeftTrainer):
         loss, generated_tokens, labels = super().prediction_step(
             model, inputs, prediction_loss_only=prediction_loss_only, ignore_keys=ignore_keys
         )
-        generated_tokens = (
-            generated_tokens[:, max(prompt_len, label_len):] if generated_tokens is not None else None
-        )
+        if generated_tokens is not None:
+            generated_tokens[:, :max(prompt_len, label_len)] = (
+                self.tokenizer.pad_token_id * torch.ones_like(generated_tokens[:, :max(prompt_len, label_len)])
+            )
 
         return (loss, generated_tokens, labels)
 
@@ -72,10 +73,7 @@ class Seq2SeqPeftTrainer(PeftTrainer):
                 assert self.tokenizer.padding_side == "left", "This method only accepts left-padded tensor."
                 pad_token_id = self.tokenizer.pad_token_id
             else:
-                if self.model.config.pad_token_id is not None:
-                    pad_token_id = self.model.config.pad_token_id
-                else:
-                    raise ValueError("Pad_token_id must be set in the configuration of the model.")
+                raise ValueError("PAD token is required.")
 
         padded_tensor = pad_token_id * torch.ones_like(tgt_tensor)
         padded_tensor[:, -src_tensor.shape[-1]:] = src_tensor # adopt left-padding
diff --git a/src/llmtuner/tuner/sft/workflow.py b/src/llmtuner/tuner/sft/workflow.py
index f2d72fc6..5b0f836b 100644
--- a/src/llmtuner/tuner/sft/workflow.py
+++ b/src/llmtuner/tuner/sft/workflow.py
@@ -5,7 +5,7 @@ from transformers import DataCollatorForSeq2Seq
 
 from llmtuner.dsets import get_dataset, preprocess_dataset, split_dataset
 from llmtuner.extras.constants import IGNORE_INDEX
-from llmtuner.extras.misc import get_logits_processor, get_stopping_criteria
+from llmtuner.extras.misc import get_logits_processor
 from llmtuner.extras.ploting import plot_loss
 from llmtuner.tuner.core import load_model_and_tokenizer
 from llmtuner.tuner.sft.metric import ComputeMetrics
@@ -52,10 +52,9 @@ def run_sft(
 
     # Keyword arguments for `model.generate`
     gen_kwargs = generating_args.to_dict()
-    gen_kwargs["eos_token_id"] = tokenizer.eos_token_id
+    gen_kwargs["eos_token_id"] = [tokenizer.eos_token_id] + tokenizer.additional_special_tokens_ids
     gen_kwargs["pad_token_id"] = tokenizer.pad_token_id
     gen_kwargs["logits_processor"] = get_logits_processor()
-    gen_kwargs["stopping_criteria"] = get_stopping_criteria(tokenizer.additional_special_tokens_ids)
 
     # Training
     if training_args.do_train:

From caf4a61e21d04bd26564f4f42a9d5a457d8c5bde Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Fri, 18 Aug 2023 00:34:59 +0800
Subject: [PATCH 44/45] fix ChatGLM2 ppo #527 #528

Former-commit-id: 9f4c2adc9a9ca8e458d3868805e077182e0d336a
---
 src/llmtuner/chat/stream_chat.py   |  2 +-
 src/llmtuner/extras/misc.py        |  6 +--
 src/llmtuner/tuner/ppo/trainer.py  | 70 ++++++++++++++++++++++++++++--
 src/llmtuner/tuner/ppo/workflow.py |  1 +
 src/llmtuner/tuner/rm/trainer.py   |  2 +-
 src/llmtuner/tuner/sft/workflow.py |  2 +-
 6 files changed, 72 insertions(+), 11 deletions(-)

diff --git a/src/llmtuner/chat/stream_chat.py b/src/llmtuner/chat/stream_chat.py
index b6d71fcf..0d22bb5a 100644
--- a/src/llmtuner/chat/stream_chat.py
+++ b/src/llmtuner/chat/stream_chat.py
@@ -49,7 +49,7 @@ class ChatModel:
             top_p=top_p or gen_kwargs["top_p"],
             top_k=top_k or gen_kwargs["top_k"],
             repetition_penalty=repetition_penalty or gen_kwargs["repetition_penalty"],
-            eos_token_id=[self.tokenizer.eos_token_id] + self.tokenizer.additional_special_tokens_ids,
+            eos_token_id=list(set([self.tokenizer.eos_token_id] + self.tokenizer.additional_special_tokens_ids)),
             pad_token_id=self.tokenizer.pad_token_id,
             logits_processor=get_logits_processor()
         ))
diff --git a/src/llmtuner/extras/misc.py b/src/llmtuner/extras/misc.py
index db91b337..b57b1c8f 100644
--- a/src/llmtuner/extras/misc.py
+++ b/src/llmtuner/extras/misc.py
@@ -1,10 +1,6 @@
 import torch
 from typing import TYPE_CHECKING, List, Optional, Tuple
-from transformers import (
-    LogitsProcessor,
-    InfNanRemoveLogitsProcessor,
-    LogitsProcessorList
-)
+from transformers import InfNanRemoveLogitsProcessor, LogitsProcessorList
 
 from llmtuner.extras.constants import LAYERNORM_NAMES
 
diff --git a/src/llmtuner/tuner/ppo/trainer.py b/src/llmtuner/tuner/ppo/trainer.py
index 4474b5bb..f929b6ec 100644
--- a/src/llmtuner/tuner/ppo/trainer.py
+++ b/src/llmtuner/tuner/ppo/trainer.py
@@ -7,7 +7,7 @@ from typing import TYPE_CHECKING, Callable, Dict, List, Optional, Tuple
 from transformers import TrainerState, TrainerControl
 
 from trl import PPOTrainer
-from trl.core import LengthSampler
+from trl.core import LengthSampler, PPODecorators, logprobs_from_logits
 
 from llmtuner.extras.logging import get_logger
 from llmtuner.extras.misc import AverageMeter, count_parameters, get_logits_processor
@@ -35,6 +35,7 @@ class PPOPeftTrainer(PPOTrainer, PeftTrainer):
         finetuning_args: "FinetuningArguments",
         generating_args: "GeneratingArguments",
         callbacks: List["LogCallback"],
+        compute_dtype: torch.dtype,
         **kwargs
     ):
         PPOTrainer.__init__(self, **kwargs)
@@ -42,6 +43,7 @@ class PPOPeftTrainer(PPOTrainer, PeftTrainer):
         self.finetuning_args = finetuning_args
         self.generating_args = generating_args
         self.log_callback = callbacks[0]
+        self.compute_dtype = compute_dtype
         self.state = TrainerState()
         self.control = TrainerControl()
 
@@ -74,7 +76,7 @@ class PPOPeftTrainer(PPOTrainer, PeftTrainer):
 
         # Keyword arguments for `model.generate`
         gen_kwargs = self.generating_args.to_dict()
-        gen_kwargs["eos_token_id"] = [self.tokenizer.eos_token_id] + self.tokenizer.additional_special_tokens_ids
+        gen_kwargs["eos_token_id"] = list(set([self.tokenizer.eos_token_id] + self.tokenizer.additional_special_tokens_ids))
         gen_kwargs["pad_token_id"] = self.tokenizer.pad_token_id
         gen_kwargs["logits_processor"] = get_logits_processor()
 
@@ -183,12 +185,74 @@ class PPOPeftTrainer(PPOTrainer, PeftTrainer):
         replace_model(unwrapped_model, target="reward")
         batch = self.prepare_model_inputs(queries, responses)
         _, _, values = self.model(**batch, output_hidden_states=True, return_dict=True)
-        if values.size(0) != batch["input_ids"].size(0):
+        if values.size(0) != batch["input_ids"].size(0): # adapt chatglm2
             values = torch.transpose(values, 0, 1)
         rewards = [reward for reward in values[:, -1].float().detach().cpu()] # use fp32 type
         replace_model(unwrapped_model, target="default")
         return rewards
 
+    @PPODecorators.empty_cuda_cache()
+    def batched_forward_pass(
+        self,
+        model: "AutoModelForCausalLMWithValueHead",
+        queries: torch.Tensor,
+        responses: torch.Tensor,
+        model_inputs: dict,
+        return_logits: Optional[bool] = False
+    ):
+        r"""
+        Calculates model outputs in multiple batches.
+
+        Subclass and override to inject custom behavior.
+        """
+        bs = len(queries)
+        fbs = self.config.mini_batch_size
+        all_logprobs = []
+        all_logits = []
+        all_masks = []
+        all_values = []
+
+        for i in range(math.ceil(bs / fbs)):
+            input_kwargs = {key: value[i * fbs : (i + 1) * fbs] for key, value in model_inputs.items()}
+            query_batch = queries[i * fbs : (i + 1) * fbs]
+            response_batch = responses[i * fbs : (i + 1) * fbs]
+            input_ids = input_kwargs["input_ids"]
+            attention_mask = input_kwargs["attention_mask"]
+
+            with torch.cuda.amp.autocast(dtype=self.compute_dtype): # support bf16
+                logits, _, values = model(**input_kwargs)
+
+            if values.size(0) != input_ids.size(0): # adapt chatglm2
+                values = torch.transpose(values, 0, 1)
+
+            logprobs = logprobs_from_logits(logits[:, :-1, :], input_ids[:, 1:])
+            masks = torch.zeros_like(attention_mask)
+            masks[:, :-1] = attention_mask[:, 1:]
+
+            for j in range(len(query_batch)):
+                start = len(query_batch[j]) - 1
+                if attention_mask[j, 0] == 0:  # offset left padding
+                    start += attention_mask[j, :].nonzero()[0]
+                end = start + len(response_batch[j])
+
+                masks[j, :start] = 0
+                masks[j, end:] = 0
+
+            if return_logits:
+                all_logits.append(logits)
+            else:
+                del logits
+            all_values.append(values)
+            all_logprobs.append(logprobs)
+            all_masks.append(masks)
+
+        return (
+            torch.cat(all_logprobs),
+            torch.cat(all_logits)[:, :-1] if return_logits else None,
+            torch.cat(all_values)[:, :-1],
+            torch.cat(all_masks)[:, :-1],
+        )
+
     def save_model(self, output_dir: Optional[str] = None) -> None:
         r"""
         Saves model checkpoint.
diff --git a/src/llmtuner/tuner/ppo/workflow.py b/src/llmtuner/tuner/ppo/workflow.py
index 243468cc..12fcdef1 100644
--- a/src/llmtuner/tuner/ppo/workflow.py
+++ b/src/llmtuner/tuner/ppo/workflow.py
@@ -60,6 +60,7 @@ def run_ppo(
         finetuning_args=finetuning_args,
         generating_args=generating_args,
         callbacks=callbacks,
+        compute_dtype=model_args.compute_dtype,
         config=ppo_config,
         model=model,
         ref_model=None,
diff --git a/src/llmtuner/tuner/rm/trainer.py b/src/llmtuner/tuner/rm/trainer.py
index 55790c07..99b4b152 100644
--- a/src/llmtuner/tuner/rm/trainer.py
+++ b/src/llmtuner/tuner/rm/trainer.py
@@ -42,7 +42,7 @@ class PairwisePeftTrainer(PeftTrainer):
         """
         batch_size = inputs["input_ids"].size(0) // 2
         _, _, values = model(**inputs, output_hidden_states=True, return_dict=True)
-        if values.size(0) != inputs["input_ids"].size(0):
+        if values.size(0) != inputs["input_ids"].size(0): # adapt chatglm2
             values = torch.transpose(values, 0, 1)
         r_accept, r_reject = values[:, -1].split(batch_size, dim=0)
         loss = -torch.log(torch.sigmoid(r_accept - r_reject)).mean()
diff --git a/src/llmtuner/tuner/sft/workflow.py b/src/llmtuner/tuner/sft/workflow.py
index 5b0f836b..b28fa1dc 100644
--- a/src/llmtuner/tuner/sft/workflow.py
+++ b/src/llmtuner/tuner/sft/workflow.py
@@ -52,7 +52,7 @@ def run_sft(
 
     # Keyword arguments for `model.generate`
     gen_kwargs = generating_args.to_dict()
-    gen_kwargs["eos_token_id"] = [tokenizer.eos_token_id] + tokenizer.additional_special_tokens_ids
+    gen_kwargs["eos_token_id"] = list(set([tokenizer.eos_token_id] + tokenizer.additional_special_tokens_ids))
     gen_kwargs["pad_token_id"] = tokenizer.pad_token_id
     gen_kwargs["logits_processor"] = get_logits_processor()
 

From 66771352bbe4e8d2321a5938808bd5b3abd8dab1 Mon Sep 17 00:00:00 2001
From: hiyouga <hiyouga@buaa.edu.cn>
Date: Fri, 18 Aug 2023 00:40:32 +0800
Subject: [PATCH 45/45] support bf16 ppo #551

Former-commit-id: d125218cde893c7c8527ab27b4d2dfb2474c384d
---
 src/llmtuner/tuner/core/parser.py | 10 ++++------
 1 file changed, 4 insertions(+), 6 deletions(-)

diff --git a/src/llmtuner/tuner/core/parser.py b/src/llmtuner/tuner/core/parser.py
index 7dda0a5c..e039513d 100644
--- a/src/llmtuner/tuner/core/parser.py
+++ b/src/llmtuner/tuner/core/parser.py
@@ -151,21 +151,19 @@ def get_train_args(
 
     training_args.optim = "adamw_torch" if training_args.optim == "adamw_hf" else training_args.optim # suppress warning
 
-    if training_args.fp16:
-        model_args.compute_dtype = torch.float16
-    elif training_args.bf16:
+    if training_args.bf16:
         if not torch.cuda.is_bf16_supported():
             raise ValueError("Current device does not support bf16 training.")
         model_args.compute_dtype = torch.bfloat16
     else:
-        model_args.compute_dtype = torch.float32
+        model_args.compute_dtype = torch.float16
 
     model_args.model_max_length = data_args.max_source_length + data_args.max_target_length
 
     # Log on each process the small summary:
-    logger.info("Process rank: {}, device: {}, n_gpu: {}\n  distributed training: {}, 16-bits training: {}".format(
+    logger.info("Process rank: {}, device: {}, n_gpu: {}\n  distributed training: {}, compute dtype: {}".format(
         training_args.local_rank, training_args.device, training_args.n_gpu,
-        bool(training_args.local_rank != -1), training_args.fp16
+        bool(training_args.local_rank != -1), str(model_args.compute_dtype)
     ))
     logger.info(f"Training/evaluation parameters {training_args}")