From 728424ce5f16b1b45f0c3c1d292f8cc8c4ca4dc0 Mon Sep 17 00:00:00 2001 From: leto-bbq Date: Wed, 15 Jul 2026 17:56:35 +0800 Subject: [PATCH] docs: add Moirai2 and Toto model documentation --- .../public/img/LargeModel-moirai2.png | Bin 0 -> 246433 bytes src/.vuepress/public/img/LargeModel-toto.png | Bin 0 -> 246499 bytes .../AI-capability/AINode_Upgrade_apache.md | 33 +++++++----- .../TimeSeries-Large-Model_Upgrade_apache.md | 49 +++++++++++++++--- .../AI-capability/AINode_Upgrade_apache.md | 31 ++++++----- .../TimeSeries-Large-Model_Upgrade_apache.md | 49 +++++++++++++++--- .../AI-capability/AINode_Upgrade_apache.md | 33 +++++++----- .../TimeSeries-Large-Model_Upgrade_apache.md | 49 +++++++++++++++--- .../AI-capability/AINode_Upgrade_apache.md | 31 ++++++----- .../TimeSeries-Large-Model_Upgrade_apache.md | 49 +++++++++++++++--- .../AI-capability/AINode_Upgrade_apache.md | 29 +++++++---- .../TimeSeries-Large-Model_Upgrade_apache.md | 43 +++++++++++++-- .../AI-capability/AINode_Upgrade_apache.md | 29 +++++++---- .../TimeSeries-Large-Model_Upgrade_apache.md | 43 +++++++++++++-- .../AI-capability/AINode_Upgrade_apache.md | 29 +++++++---- .../TimeSeries-Large-Model_Upgrade_apache.md | 43 +++++++++++++-- .../AI-capability/AINode_Upgrade_apache.md | 29 +++++++---- .../TimeSeries-Large-Model_Upgrade_apache.md | 43 +++++++++++++-- 18 files changed, 466 insertions(+), 146 deletions(-) create mode 100644 src/.vuepress/public/img/LargeModel-moirai2.png create mode 100644 src/.vuepress/public/img/LargeModel-toto.png diff --git a/src/.vuepress/public/img/LargeModel-moirai2.png b/src/.vuepress/public/img/LargeModel-moirai2.png new file mode 100644 index 0000000000000000000000000000000000000000..837a5a693c181b52bb5f252d0eea4725120964e6 GIT binary patch literal 246433 zcmeF&hgTC_+bH~mP^7C!uQrPGBE3XK5s{*ZfK;hU???|SqI3j7Itqx=JA@jQDjlTv z-fJMC1-?D^{har#v)=bx>-+&{eqkkRk_j_=_MX|-wRb*fX{b zk_T^)nyINMpZ&9t>*2@28gi%m`mW%)=$}7x`=m4Qg3L`-{Vv%m5rUYAT5o%p4}2p` zRr%JFSEK6_p7H~PGo|g88EIUx*qz7B*TrvqxcK0X&x`vAzN|Vmi5F3%A=jvL^KadI zR!qz7ruQ?!2TV%IcdZe;6{ZTYiox-CF9h+e^Z)Ph zKa{|Z&-mDB`GoS}&|qNGb4a=7?(0#Ka!MPPQvrt1CbPEoT=hpuUnMPj^8*$lQ0C+< z7bC78Hz%ig_>UgGWZCj|#Sd)ePMDNcRMxGq^~-$QO7iVGNj{WuRmY_!ABpTgU;WKM zz1FEd|G8DvdCCxI`3G>J;^!N-X8+PL=n0+@Hgf{qq}8 zvF+mMz}un~@#VOpX2Pvz#Y4Nllt)Ur>M2KwRpHbShvw!!8+#f zu(mfQX-=xeF(S?RNq4u;t^udLU)g?d<8EqlLNsKj-Vtw`W(uU|jwj`cK?V;i>IA(E zb}QyrM)&6zQSt?N|at4&9QxU9+$c zmOioApGsP1zW}6FYj~ul?CZyTr80V_1}$THxii%7#o%!B3=f})y!@H@k{;}+^F*^d z>yUcLdZ(ZMw&lOtl&6~XE}36_CC#yVN*;EqV#KR>=m+h?!ynP7b`cIBc(VT zhl_?HCxdQAMj!KucK+;e_{5y|l z-l-m;_{e!?szU0fxlQxmu!I2xG#wESvBtTuvHjrWnJ2%0yeoa;spI8jLmY-a>7%OsUtji?;r`$k-`$|q z;iz7JrP+0sa3+3cox>+_C-HKpQ=zG*#>$UA?X2wfkKQgrbA>G?k-BvclQe;b?3(Pf z2iQ-G{GC!u8*LM%GPsfJO$z{zlugh`@sr>Eu?R`q$xFWDgX_IUjK>wQSGL-xcMV8pdvO65^HuU6bK zZvJ*$kN~B=q{j`xj|{4zwNOK1cTyktBsUqNlCf? z{rgB;H*C3WVyAk}y+ti6Do(9EYhhD=Acj?6zT}m~4#irkA(a7;yKolg0^KOA-V1Z6gF#hZZTB1*J9xf?q`LiKffO)VzbQ zQg#+zy=zy~g{Gm~`2<(p&~t!C^KDMe`>$=?&UbW40Iz&ZDxD6tT(z6(+XPPqdA8NN zPLjr31MC?^8rqBw{UVOWj&n(LrsgRc+UiaO;|a&fUrwaZQ(cvm7wg;f{bCHT2AjQ7 z=yXSy)rN(WhBlou#g@~tHSFushHiCj0(bnW!RE9S+J0)j;52UR;CS4QFts%7&$`ty z?dTE#?ZA`_c25dj<>=Co;%Mb?>gPbS1pYz9*27Xq7x9L+z^$Afn@tyS8J=vF$h`cy zZ~JzU-bxh}ZC*2pKOSp%Nlm``oRk?^#h=}2!;?M6?co4#szaMc#*Zl1eeWm}81k#^ z{k-{YReW6MXLC>63hq@+J3P|I6&GnedeFpVX)G{x+BNjOy|dN2EcZj_eefL*JSvHj zeOwnK_w6E@4=lMqP5H+^e==vaDc&{k=Uso>O2ylWH>ia=nT!@2NlDR^gpS>khaMTe zmA@w)jdJtylaq-!hNo)9a->we9+rP;xD+1JA?dT{6&*AtvqdrFu$HBv0gqo#J9>hP zsM&cG?0&3<{a$64q;fqpZ`{Gl6a7K2+sTE(sYr7VJ(k4p_1Dx0*C#b95{!Gb_?6t4 z)nRkDthmhCg`%jUduL=Y$p`;^Tz1aVS;Eodr+LI(vwl1EV9no#H+SVuG%MVGRgk30 zvtAkOmR&tcFPpR_Y-Xs^y05L_?HiN>_C?CCwxrLBX`i$v6_pppiYLUI*_)&mHJBV^ zGaR3~J6!I9@{)l1eJl27cuJDr{sN~Pv1dPQP6mJ}R@SS96vn)nPK zR~A+Y$rn}Maf`R>?$q7!Bq{NN)MxI`+tCcO70FDORProk%mBshA2vwFC@UC?uPex` z5kumKC&v+KWQw#N`4+j5Fu%zpx(bF02F14in3}KtcL^a!1v~3kAjOIb=oPd6rf;EM z##?RIOL>5?&4s8XAoBTITY@yi>nVP5k*8ZT)zNmpP|2|FRey|l0R+o3lvESvD|nc= zK@`CSq%(5Vq3N6X!*pMz{gJ)Ot06N_I8vj32#! zjd}MMn_0u9SOQejrIr$B6K8GYVk_PNb6(n-4$m-UsF(2g4jZey2mT9SY>0+JC+K)l)d|~|HLik_SA5xE5zskf& z17AaZT(nERS*W*Fj~m2Gu}<0fmOB5o7xD*Am&fX4`uVnmNDXsQYI)Xg;1 zbm*uN-D<9_X&Ug&KGeXnx3|@yBWpRmlvDw6D_R=1mDeuVACoYX`W?grC3Hz(#P{Yd5G&>XJ%vlyyYsO0`UzZ;Pgi`na;* zb7ZQUd8*o9_03s*P9Vj{<)r;E9xEJx`!1!zsm`fR8};I3_@H2{sfmB!!#5 zRKahjJIwKSG={GIB0%w}b*KxdxVYA-bK}X4C!`Y_of8*pu6yu55#D+-`J-0*X^gi^ zNF3zWnvbholSjw~Re6UXq^xAU(~{4M&x!;bbcYrovZq(z0aOfpNKO8j0O*@K8a2L! z+@X;EwUi8P=u2Me?{U_iUkwo<+dcr)(Lzd}H+xYMmcq;OsF zZ5lBt+;Uz%`-_HNMqWl&m#&3C?2`Lp`m@*i_g5zCmTpe6*UG)}RDQ<&@c!p0k>_)p zlgW|wfRu`(gUpu8%HX@F9VUdYKi%EM0xFYU35Q}>%f)Z# zI&=qLg0`6O{$Nmc0wA+{S3o?DxqaNNmPg8c;;*~*kkM*BrKS+zFkhn=1W}K!Lofd5 z)J}pn<9P>g@49xJd_EM#O6_-9|IT=)LLF%gZN>$dUYl$B!WeAKR-(P$rp_v6-&%Ir zxXyjXc>xHY_2W2^KkzD=$eA?!=Q;(wHPTiZV-2o$do9b6y__3Dd-H?7Mp*aGA?-RW z+M3nwJW!M}7-Qh$y3tuKrCEP0$tWT7!lh?^qi*TgU2RZMzjA)Q&CxZ`8!{%0pC5EB zSTgMPa`{tVe^QOi!eVTa zw#WKG^y4T`zS8JGd)G4szxzRiAtj8nQM{^YUBlnyem9E|xKOZ9b%R z-$`3i>JFIfn)E7vKKkw@>o=J*@WD7E0}uO=E;R)i?Hs;=*lOFDzFS#u9*|*8(z@l- zycA=aC^+ckxnEI*4or_tk8O{oIqM5Btf|i@wMs<`47wk#?G7#I=b5$vstmc$U4JjW z>m;~LI95-&;oF@_T}fR_^r*cde(1kHerRiF;}%{&gXravPD1?A=092WB8tf+#KCV4@GB zxYscL?L_@BsXoT%@lNKfYQuQliO(UvK4#3gxX#R{!O+~_+<(r~#fBd6Lu<|{h9668 zMxu8UA^vEEG@S1!HpS4%3urKG%zm$VVX93(-XPup-|p_Wdh~ZcCD|7Zm=q)$s~3Jd z6OF$b3;Y$9YVh{oKT%pN>4zLG2t75)t3OB?Us>-T$v$2v$tle#t&d5HR=DbO&!pg9 z4zMaG6JZDVqjhfP+B_wQ#_x-*Sq`@Qqit^v*%bKkEJZCuE%XTMKGS}teUy@H9B2Hr zVaYN7Ht>tnm2*s~BEW7dzBIo)znq}4xzdu&nTBhh_)(|rvm}RR=k)X)?-kZHF)%SW zOG!Rb*^gZHLtC8!{tW!e5zjW@K^+$&-hAmPt@HbuozM5!At4ga@ z3FANQ0>8|H>?lVVO7ex{Kc@nd!r41IbGf9U^ez(fFIs#3(c&MhFLv*ZZT(qQAGo8J zSK~UUQ0n{3jD%lU&or*tOhNS-oxNy$Q;xysTD~2U&;j14q3E$Lvx$Q7?mFFSlTBPA zCBs$yylUsR3_99ky>RW_mich2VuyeUCHt+CDVG(2LAuJD6}hr{d9uZsIgP3UPvuc9n_ z+G~8Yzbzd)a3G{4D;&~PEp8F;TQSuiy%RT(OH-s*9;R)3dw_LDLIPtzU)Tuv*GsJ@9GH{8#kK`aA;74IhqyM zKc)BQeFe^lhTER!`;B;PIa8_p>!uw2^V!(G>X$q12-=&Rc2#!n#41k20&$4;zy*w1 z@npVN<>Ex5|5AU^2~eMgpP%ZT1ddXMlodbtUR*i$>c-i(jzv`xjHeXs56c$%ZDdnxcm$>ZTk}p~doXBlBE4wP=<&>m8NPLj+TdKMnq#UHY zFXy0+g9O+K{}BN4OoX!dvR33;1WRN9^c2?j3HJNqqkufM`K(s8XhOGcL?9BF*hxc?BctFebt=r!J<>``f^`1)DW_$Uk zGN4G$!3E-nPH%5rglK__#(Q~HEBq=}y{I}GY+{0w=v4<`!E(625@yJqSDF&2*Q3mj)$)#r9V`Rg{#0Rbp zA-Kqmc~)#$6p!DY0FFl~w<&ido z<)^**U+WE9<|EqJWAp_oHGr;qoCAA<`^n&rKl&=WiflxYp7q_XkmL=iLdWp>n@3L@ z<_33i=5$*48P@Xt=zl7`;fGFoDb??Q^O^dqx|ZqQr1IKwP)rl=kGNKwPd}V?m0FejcN+ zi|cPUy}Zw*UJNr<0)xS{CiS%e-DZAOJJ(kT`V9Qe2G1X^*H5|3>E|&8*WR6>J6a&B z&#c9p7GI>L=Wb42j~`2v68BYZFl>$gi=PyDI@|VZZ@8V{66dA;YlYx~ug`42n-^bH zpipWi#kY61_mkUh#UI-ziIpEbIZ5`HLCb$Aa$?}uE;Z{jsPji3HZjf(;o7O&=%^^o zip6yJxoZbj;!6eUSsE%$jH-64ehU@rC1misP1@|W+Y8c!cy8Ekk%&iLT>NELhvmb! zR{b;+WUdbJY13Lf{8=Ayk9i&Ox!k7w@Mo5yeg7-#yMimv&fIrJJNBxT!!D8Uc(Je)4WV)ohYrw!g1y6^mh9Mm#bXV7pB*K4)#RMAU$m>pL;T3 z>O<;c^GT-{W*5Kq!g$T48&QQ_`O)klNKbcs<7Or7SPVy0aRU_}nv~QGN783-k^3qP z9p#t{^caoj_Yzo4Ocx8%5Cpm(uQZ4UF$#(CN`5nPUX8l9%s-VAE8A0hGW4m+FYP!sPJKIOxv zvj~#LSC>t%ZQm>@ul;>)nP%;qrxNKf^>EC{E=mZyDJ-CdD(J)&h+g2e=YRG2I*HrV zX8nX8t5(hvfH+h4eqod`TtdjvaI%W+ekTN0-R+sSrAW&g#1A0)tTVgsr@>h&r5qf` zCxt5;d;;+P$hNxxNmRr~fQUs?D%C-rI`dmLfS9nY8S*G3iH*A+` z5|hhAsV^&ER`R8Phrn7}{{Teck06B6{^h~tLD;(y>QYuKvamlKpA>GezL*-WFZR0{ zB_jFU?w4NrKk?2Z)cUm%)sh$(i323v&+~W_H-UVu3Xt~%3Aw?e7NP<8B{zB?&?6h* z;re+6AX@&j1&~>|LSRAa6iJsP;TzUt-4$sh7d?44HO8q$et&-6A%c|D1DLe79|7=D zKL|>>5|Bc91p@G>3H};g58?Ffdo$s`Sko1Xfb;jM9szGwpLNPwWissQP9>8H+lvwG=-yT_q8Mx>bsAcn55Y&rYfIpoj7ogNKKutx3`EF1OKplmXLt4Q*6Vl5(+?qFe_i9TXOZ8W}2 z7_Mvtf$4BSP!}JNXT2pVlsECJ-z6TUhroK>zpg0{!E>%YQTZML^z=oPTqf^bQrQT-7G#FE>2)(gTX?B(Itvu+cdP>hj||neXnyZ9@Mt{17DK<>Uhb zr56B8lye{eXX=y9bidD^Zj}v?X+7Y*9($3iOCDe|%`*oO?`qk1Jz5Qr8=pr-%gtd~ zsgE^(kj-i|rG&_lh)6IeQmLD^PF(g39w+%*@ZIQ94Uv{CUrMY7myj56(Nv4+5-ADO ztXQMvd&=}9<^2z@-I2(Fv_0Oj{H!!z`$_;w+QHd5^lfJhh00!aF#`+>k} z9(T>vRM?5sq2un<`q1RlRe;O>_UPL(a^ysDwYhr>>{R^3`k;HS9j#^bT?mC`4@AX%1c?6L901_H z?;sfIaA%e|3D~LugezYhp!*)050D((!-g-iP6+F&=4U=7TDbv1=KBD07@iJbywZe5 z62j%DbCk2EgWOmh9VpL**LwiCdS|Em&H#L+0D{ba2C!gsJAiTF?6xm$U&u3+Gy_sk z*}iWS0Hg;W4}s$iPlgtO2z{itAE2O>^bDZbR+#p|;L}Zf)}#lW^AII(N`XMJXy5`0 z_YNTK5rx32xLwre6VF*JSph_r*C8;$cMNwz;ck~8Fd<3^SFjYo$m^$fRJ3;)>GKev zTJCXQmUTn5e*nl^6Jo{rtS@S<>H+6^h+YPkA(7cS0R1$@JV5NS4ng&KK?9j^3C={? zPO_fEOswh>9cKUC8ubt>SBn(D$jkmzHqVU-DW?WdE;p_L2t#TJQsWuGqV-G{AnS?f z1BgaF%eG!j!K?0~fxzTVz#)ClYX#f41hc1wr29QO{))Sc^0|Un-u!?pcm}X&%{~Jt zdL;V*qEWJPS{qK{DnSU!upLlKx3Fyc^YIg`^R?BjQdrrLZQ+nHt&^NH07oymbT6bL z9yG^>@i8NM7$C6G(lYCp-9%LgG!QsG4Ctk2mI8F?2|W*AaHA^X00s;p@%Sd*maCEv zAm>%(1K`qbG60;N4}z)*1mL`N2Hur>LQz*CJ$P9WMOWX&gNLX9)-%add*n)XpJrAX zD)|@Sf5szs+I+w|=LkDFg&>*h!i?{cm26o~wbu^`SNfgccE;eRo+n!eQ*cXC2om>Y z{mw}6k&CRh`?uAHB+Kh=9k zW{qu7hN_2dnS76Ux^&UHvpW$v_#;nSKl^B*!wrCW^VfZvr0_Mydrd64Bjdc{yq-3k z84Skt0&wC>Bmmrq88V%|c&qbG;*PgCXP2)ub|zDLS2zMpT4?wdkr7_QY8kk9!+Oko zDe(FA(Qw%|7~c2-^^83%JX9)Wiy0%C_~&Z`Y4xJ#!xz@Dl#qP7Q5=z87~?tkizoFK z^(wF=mQSHMByNHmA~xAXDZdSLM%Q5i%9n`XEJ0iJQ${FzHl6o5WN!JZpNE9bxzX5v zvMnVpf^|F5rMqJlmMDpbxwWqUXhsfIM||426R6QUc-cM_-tn9608yqu8?(rf zAD~(u9WMT2bYVHO(xto;Cc^k-8}Y*4H}cRICV6M|#p>g9Ticch55e@9-11y?S`6EG z3q5M^muF-*!c_c=%_G6Uf(dpSc2XpdfvtJ2ENY|*;fNY+Wo%p`E$aNb`p7=8;0=4W zoE1E!W3W3v1J-f-ygd8{PdrRl^!z24fMdVS3LbOBqOeKpPX)DXkFTf7eqv|mUUT{D zJpAPFQ^Ohad%x)p*=3C0y}K`)j2_2xW{Fj+mL4cLh}@XfXLFGik$XpI=l zg~E^3|CwoX#$Iy%{(0h^0JW{#(@z+8=Co0d2DG;{6lk`)F72Psy|RBew*Sd-)4`vr zu>+^DB2Itawz%wB06kNHYtGJJ+e}61)aPs#=4EcWA|F5b(nqO-)C#OLx*hg+L1jh% z990tQ+4X}vO{e|82i;7RZY;>_wfvSGrfof5eVg7>iaPp!;p1%_?CEc+QH!q^|DF8O zh%7VlY}oXlSXW!B>GDk3nq4ZG%?GZ#Uq1KxK&SWD3ghwsDrZLaQEnEjn>G5`1#(bI z;oh;UZ8PLh$hY$`AdClmbfNh0QT6aJ>UqoE-AlIj75pVOUqw**dcp*iK9SH>mAoY- z_=3nm+WwWxB>4;vf72)BeB~2SvXf~No`mm8f40dVCk45}srgA>7UMHH{pnnj&{okE280RHEB1sBdUQ$UHrU9j@S~pK?Fc7eX!saR>D^WpXI7OoaP~SHOv}U6@~)w^)JQKW7KF9u#-)_r{@Vn_Bv zlq~^Om0cb8&i{~Fk}Q!ce4I6k>)h;6fS>xZmAP1;ZaA7bG8PnN)xKUNHD*PGaZG%S z@qp$CPx;0VcDjcQ4JP>`iY^QtuzB!2@=`4iWuTG<_E+3j+yb&(6_FKw68P@leLP62 z<(s_Z_{YgHxR)k_{a(PxjZXr=!`0iMM&_=1w|#Cr5BnoWH%8Zoil7%9wYa?9@?6#3 zepQU@1n=X&E1jvLenHhY;o#%npnr}z;z9qnu-Q+rvt3bo`@=T@@^Abvi)bt6U-@E4 zxM|B+QEVp~_pq0fA8-6Yy__V9DT-OwIsWgP{}dTiaAxVt^EP!o)kNplzUSpmPp3AC zS(b?x{L@}eR);_G&a5B9H!5aq`fKkioa(;aXbcL{c%nmfbSO4Vwp}lKnRibhUD4)k zW|RVIK;(^NLi)dThFJ8xz5#^Z0-XdQ57pUuy7M%e|IT~Ihqg(BlKIr>TD+GivK;x% z1OaNcW8ZabaW@Qm9_Tr%S0Y!0V6OdEeHTM@T<_Rt1&}JossLyExRb&|6Kn%%rsjjR zSjCkV{Lo&(icx&03 zil~;T*1OjGpBB;}E2;E=k(E_sz3P6d8TF5r$Y%C#=fEwR^o8saQzG5Cp7bFi8`Rs;Adp7gC}MZ<_?w)1XW|GyZ%9V9-3dViZ#hKsP;; z863l*K_sRI1kC1NzEX7+fYb2-KN?P}P+4#_L>!S$Wv+wn1ou*q;{p`8&!&YoQ)bVkIWuN^iTC+lA++vJPrTB7_90+a4d_tRe4B7c-SIU%MkUJ69_6vs^PK%YMj4$Z1=l~=gwPM&m z9sIDf2dAG-_yhf#kHhfHh#esal2{X<6)|59`y&31sM{>9NimdU_0k#;ND|5@U|hsB z?J)zu!{S+lZyHm?h6}TrNOG8lTL8obc2y-ae@Kyki~)|%Ha7=sa_F_Pbby<$(m3HE z5RBmr`j=}^U?*FWjHD~fw-|zo`VR0GiM}-NYrclrh#M2EG{6MU(l&91z!T0vOkF1C ztJ~aBL2m(KEp20frXx2|_hw;EVZC-uwuj~hX$U8yl8`@D=GKF$PxX{NWxc@4ny&Xt zfG@rG3($)vc#&eM7MWr94k0My7vQq#6KjC%Nw6+}5MYg?ZYQl$JUUul%YeVU3Nd|r zn7OOEfQerh9x*8isKXllx2yeS{J9h7VVcM0(qeFDO?N?R!Y->Fjl``)#UGTn&Nmho z@PA1PZzDFoxo9Q-_{#pl%%Hh~BrN;4pK+JmsT8K_o zPTc~$r{+SNAc3~}LnQ^4@Fs5vGVUHg>SVuCo;4Ilbk`VA+DW}SG*M7vaFJ>5}z&^w9IdRhB=iwo>~t;yz=w`5XIgQ@PvKu&HeE;ne;u?m6$%B z>4_vy$H_XlnE?cr>H@)J4gmD$&e%?6PQ*nfOsjvoAuutssymsm8`mMo@_PVjdSO05 zZOBk7@JjzY?=zjUoEP*?e5a33vR-j1LtxTghw}?D@qxPw3+?Ukw3ADBY=crk$u*2|PD;yyCvlQFSmxd@?s{|0%TPby(I1VR5`@Zl{18kBT zfU+k%?D69*FH+hbk`dtrPZ7-_)6#_~EAu<~K%n$9Ikm}EoACCVjUFe~(f>_0QlPlOM83CsR&rZJ7& zAJi{mf**0-^J5~>OD&hT^)UX0tKp45SzzbQBwNWLm=_5ILyD$$VVy^f z?Lq2;FsVV!9Cv3SnDTYoks;Dv`unnj5abJcK#(i?GSG{6J=B08?W;ze%4d*=?{_aO zTt(@f0Rc$ol39zUHq3BqakTl>z~!ah?u;}9Neg8JBioh%>e;0dUW-TPxUNB9d3rTw z8zJRq^#MP~A(*ge@q1lNb1;1-hpDO^#^i_?4lpHzOJ})GFDH9-m!EXy7a*}~J2XV|#%yUEsn7;p0 z6AAc36ajvWz#-6buAYmV#1_9Sfk5ZAmVD02dK16MYwW7BL;|wK%Z*>@2oYN!=y{$d zFXwIrw4=Z9zr99*%*!Y#c>TtQp*^;@tLnkOJxO!5uWw7Y$jZjv<*H-B)y(8V z1sx-U>?0k~<)wC0PR#!hioE`>P{hLC63~fSa99Mm1)eq;G#T72pjvTX>`IWc-h}Bk z7Q8pQfS9j^puGDbN{f{yWB&HXH>cSM0K4?8Pb2=l@cA2Tz{wC^PQXELyhs|^r3Ww- zQB?q>^L7hBfU$x=g@Ik=FpYE(n;wk-X0jWgjq<-rxOddq5FlS3Gpf2lEYb2ZqMwaS zLW_l$I=0@2#={jt^zJ7FHpEw;ZJ3R^I0C8n$axJpC7d4gLr|+20GY0q1(0kjtpxO) zUDG-!-?p6@QVNjA^KEL6XTzYptz%@pKhDmi&Ub%zKn+D;J{HGW!>-0ex3vN* zg9pH?Y1@^dzQU|GcrR0r$@SV1ujzY>;T^0A$hp#H>+o(^o65(?Yn=O?!FS?mjAY%& znH&_yeGg2khIj*Hfg{dH$ENBKclx!X)idh{mTJJ+v2(F41fG%uL4wh<759F4^La@6 zNCq%BBv<%91%od2XT9H5ex$Oc!U|!99EBWE18*#h=#1=ZY9^(U6Z)8~cV~4c!@0%P zaJmNZ;_!#SXUwPT68~0*m%n4M8t0Vs`3kPtT%V1UnI;pKZc`>adZ!sRDv(V1LXYo$ zXdTag1zH^aPk_JP6Cc+BySHK>6r^%}9M7RZ`N+-&g7j1dh%vq`0KTR6JNi5NMZ=Nm z;RAqZVeQY*r@~__p=`rpY@*_LsBo!uFM8!_KfJ!zH{QmkWjRUKSX>**T-$cF_>4Xi4uq)TTy-ifoGHg63v2AivHWR{*nehQXdCdNM=Nwd?N8fGH)My+$M{aRAA;>p9l0n;1 zSC{Lb?888@*z-Rw@EgJse6d2@`v<+t4&!up#{(}|%9pl;=S@ib+c^rW`+~Qr*Nm}v zq#E>ePT+leKF1%v$3#dn^UG@7>r>a0c}0?}DY;xte3_1K!wj?@C3e&{%H_Udeo5Q&P51}>a05;ir#C?dP{0p{tdHs_c*?w-ZJ5Pa|33l0z% z&CGckLh(deomSP+!RlOYBx2Vd0yCVU7*gs6lO|V6+)M@C1>Nl%rd)zbU(y#pO!$U+L1q9ZOgiI5< z@N!dI>$U;(?f^a0Po>uZ#7d@iwgBb(`U`$nk4W!VsseD_@<#7i2HE`zRRFFLnam-h zK$Kb+fIA)Wft5rB2^4iWBut3OA-_EaF1|011jzEz43sMUX*$wG1Da2mVLbmVPPv4q z$N3&ydLqF2D4jTf^qam$X|ff0SY{+ZTkqA$^AmvoGwnhG_L3cfQA5YO2rr=G<^)>| zA`m`vxaN&;L|8C0tvCni-;_K#eb0+8$ZY^35n8NBVc`S`JEbgx8*Fx~X)3gV%q-s= z2*je7T)UTuxM|fn66J|`lP_sd@HibdF3`nm!uY;HFcih~X8esPPn9WlKhH~;M{G@x zOcNHw+AZTyeTXcw2eh!fre}F1apA_2-XkxNE1`XiS;DYFZpd`NR7uS0hW$Ld) zADJPj9mP2i8NseU0(woBr4Y&SUY+>S0-sFy%3-FX)DW;= zlh`FWL40wxPeGg)SzzLgrPxtfE?q~;YHry9>Gne_N~;I+Ls%`x zRy$tC$=2(=&3oOkm|{3KUU+^wHA`FWQNUNlga9b_c;nlHO4DrV>6xS1FK^D93SeUx ziRzpgCXrT867O)s0#@2-Pjrb2)#?{Vn-;e1IYFjLfO zv`fKcBDYc$LBRZQv~-KH+VPowx>6SzzvyAKtx%`CeL>2--84n`&1nec;bB?GTbKQz zXRbUBNoJwN-`~nyBy|X|wbG^ce~G9dAHI~>K=i(lTP z16bCOQj7Zp$)c?KJ9+(cY!Y2OshaX>=Z4otfsfvxKJ1;>g}@%yLSXuLAW)LzR7SX^ zp6%F003})Wh5G$n0kejf&Qm7Aiil0^Q`)v!p04<@3%I5*(=y9P(I7y=}E_dr5aMV9lYVxx7DQ* zN8#TQ(#JKy$_UsJ|ESfdm1)9{l~LwEOs|<-w|X$FIP!|&l|v;*`4q!^#%H=PGxkqB z{L)p{#SfV`!g3o3Q3AxELwQp2zY>VwXCCxFIorR87eD&x-BjJ=O=4JFE7B!(&PmDh zbxR!uh77g@mp}GXCRTAM6TSB8}WhN8A%zD1;#=71#{q+8#<_Er?7|Lhw zx?F`!b@9=N8XCm7*Z3Hk0Pij?@T}#4eF>~gd;90?L*ai%CNRL;K$k8A60 z`@Bwit*x=8Nq`h2Fs{*@9MPn4mFr70_7pviV(`(_DEc=y+&E>PXahVeH27!)gloN% z`JMb+V|6zzWC21%^^S$=CewpeaIi22&6gQv@G@oiJD3turh}pJbTe)1Ucvod>=o{5 z!lK6zJ8WEBhSTW?#+XIxJIQ!1b8Je6b(eU1(dE4o8=3Sw2k&B0ruPTa8%}77b3R^t zz1s9qVGI-3ZV=TbpK{9#Fje_nwo!{6?B*B#9|tWgMya2e+^o0p9e~l0a#ATWounz% zca#EjQKHen;W6R4tQdPj1cnC9?jko2OCtu}{xq1fmhE;+U4cE9`t9&;;-s6iLDb!R zwfPA~tq0gv3h-^2i)dkw3{3Iy*c?HVO^-EJesH8qKX|rNK5IKYaHP9tlFxbI9+t|g zAmuHzy}XyvA9?yFH=RWu0{g;zTl}rn3I5oTc|xq(QlhBG^~=(2O;kSwrngwyVVywR zbFmM&-Nlpl@NJOzlMONt9A;EKMWpaveuIky6*QD_I<%?@aC& zG?lW4T~qU%OXZ!V3FDnhP;ju=`1VuL1RBh-z%lQ)FAfYn9r^28=?-hX4fmh5?x>_J z_YtLl2e_wQUz&i4Oa3-bWJBVjBV(rWN#D8C2(w=x)k=Ldsa;L#?-xWzVuiJXTWp5X z&eh!c-*qYGN(#{Fc}GMI2vz@-sM?un1qLwA0y+vINDE_t!)bh{YR*{&lMR9KnrPR5 z^XXEn3G+oo{nXz>*U7!kJ+P;YcxLb~=23_5oRzn6b7ubGYfPoBJ z$km1J%E`If7O%)-=VQly2Y{4ChZP7EzGd>do){jlE|g>_dAViDU_w{+rrb-U{a{No zEjbLUn1R|^PTOiR9~mwic1N~{N;XssVb*6z>8NmsH8#dtV+?tbPO*Ne{@IRm?Db20 zA(u=={F1k-oH3i<%JXe4P+giGPK&J599Pnmrw+XKE#>Zq&$JRZDh!lM?6o@<$@CDv zoWFXDEbr*YzV=!+(U5szi*Y2ui|``S7C^AwWe4DC^osnxPpM9i6o*eukKb${i+zA& zrG~TMe>!!r@H?!*k~AbX%QyE}WT&sM161oz#(q?2+fNUPx*0B4^*AG+KLbX=M!9Xe zF@@ZS+u}i16rIF2VYWb^RV5%b?eKHrbq&e9Z{Me$&SnduLKatX4#Hbad9(26A4+0T zmnqD2-w`8zyUmxs`$1&;&Qb5hODrP`6Qz3?NmxN=?MsUh{Dx zOygRtQ0f=BWeP@D{4Z~hD{yw6xBQeCmd94$qMMAA_!xbRJ(^?me%WyBKK*L#OZ=E8 zqI0G5LAW5Mw(t7f(;x(%x0~%Tj_Apa7PklkL_Wrnr6i?b0gyn1)PE|6@?uX_z+ zMUb9fHDN={8A*^NbW39 zUjw(H`^l^>&)ft>3Tz%WF?pI+EXEET1)X`x9dXZHDR(=r@4^bi@x>i}AX&6#cFToL zuE#3~ys}++k~4qb z{Jp2%{OnO2f(p;PC5m3`er*foK6Bll;HtP~_lBuxv%gw%O%eXs8G`v7B1&(51BL&( zb$2S1h2nn{@`!4m(wIM z%xJJJC()fWT0g}-ne^C9fVfrxi|ZH@hiOOfwZtN5jyZ8k>$LCxegNPnng?myN^p($ z$>(J*6a96}pL*6nFCB1jGJ_3PSg=k(~SZ@B63V$^58IcdYeqc|W|hJ7<_mIiWT}nEAGb(|-n{chXv4ecZV=M{fGv5ukO-fxyJP zAsG5E{I_-AqBg4`oRw32+24{AT`1lgov-2Y_&J?FH4#z6?kR`riUSz9aL?C&4+mpD ze5MoSF+*Y01CxlP;7gKeW4;dJPOURhzb})e+jtDD=Zs*e(X1{~^QhNlfd5&qzxhgd zmQ)-O^y4`aW5200UfiSg45sOSruv_I_0kJ~>#RZu941~FyZ5edKR(RbRkOr&4grj> zg!T-Vy&B}+D~E@%FyFk$OC13Dk93O}Kb{-|}-&`B=GFClO1Mpyi5XQw+rU=#~9%6l|8@3n+ z@tdmjHzzb7QJ+5$-8P?AnGRZ|9AM1Bcm-mKmPkr+Y+IaJ3s&u&aY-2YLN8-%E$Z(c;5HpSe--3rk-OOWrMB+CkY4%>{(A@- zAqa9}vbxUQ2mX`)wJvcXOjz6aTkS)5#5v9jidU#whyMUX*DpYj(9gc>5YzGy#NKDx$Q5{bwHw;8T@2UF`CqU(*BC0e{dn4XKf*#Aga6XssNwgK6}@W zA3;MMJdNqhP!|BaC~yG5!lWqzxG(W^d-IdD_+TGIDcrc1Ow3!KUvmbM%S30zfzX~B zD1)GUg(yJ*{!;ths=zl{& z-d|Vfd!HKIW0Y9U?Eo=YD*bK#5%aWks`cevVc>G${7uW1RN0hWb^l8SMsX1kMP0Ag zm)*ZXWz>Umo0%S(2r@;EFnv-8!*djbS11TwOs4#i*3Zc^*p$<+(Oe(=mRkeDA|tb6 zfOfXkZhFuS_^TN%37=H$Dd^G0P?rh)ue1yPC6&kq1=O8s2tAzw(!FjyKSTwrSxoht z5LDq*3C!CQn)q(G_yj%TWp>$N!#4oPpGR@6peg>>eCK-J_WyLF(ss2SdB%FCwU;%6 zIHX+`Yj;N^l!15`ZZEFZ_cJghIi*0sy=h-01iWwSZrXPuC+R9_cAGTf`lnrp9|`daSla&X#mao)G&uFGS~l7h#9Ow0nZE5;awgSccYh$E z1B7{=|5s}@3#TmCRc$d@ej%+>;)sWWOd#Vij>U)F)^-t_3frk-?iLm2xmA- zPXgmbr3U7u81xzb{(ldj!N5lBQ zVv358#CYL6K2uUp*SXmT!pO~5f0h=5^RP0;u!|Eu)@8_ImxgB9je3w3u?_6c?&{At`v6}$JQe9%RszD0gRWj#}MSdzMP!-Bk{S1a}Vi> zE;)}4Hv|}m?}H$M%}&14$EQ4ay?C~~uyia*;@1xV|MY90zw2FyfH$CDQKifBt)YRq zHL$jfoi4&dPbB z5{z0fv(tli?Vtqs>|M>A)$hZeM17*Cl7hCgMrzS!ooDoGP=~jFD@XW=oxR#y1=er+ zuF}q6D$#~b3 z=d=V|7Is@kx?Ij6xN|86+`KT(O2;&PLd2Qc%pons`5s^t$d1eGYJ-jcFZ<@|-|@No zrd&h6`^mFevREYTQw2PzOsULH`cAs898jp)R`(-F{?cT`5w$n}4wI5hi2bE5Y39EjLfFLS z6zKLa47s&E3Vu*v?q>ZyQtO`PjkyUZvY~+>=>f(;5afG$z32`0+zZn9Q~SEW}Ax;we~G)iA2rBBl-&;x>v zjg=Ie8snj-bWP#|^ z<+*AsVpw0>fPy`z$DE97cjwRYWa~jxuVrp-j?adJi9?xilt`gWOiE88KcOk7Cra}8 z5Lo;7o+lj=Z%ho9XIWtm=nsM*tnTa}=&i!_`*T7_33q+N3i}0Cv02R}3KCKbGs3I_ zK62eNet7#+W=}aaUX;U)Tyv(h&fHC=%(0__uEz^dHi$1_H${IZ#c=9B;f`b)1ChH< zE>!%mfgL2AaqSz=-8G)%o|lhnPeN{KDh?}7foX}z8`>*OjXT5p-(TsA5p5hjJT8zr z_#~B{?h2u_ePGB!h4Z{FslM0e;^?x&xqq{h3Sd3ha2c)T<~118y~*r;lD;mZY^M(* zA2=vkH2o!!$HUCJn?{aNGo8lNL~bU@WXCh+Ezpi_2fPKi$4tms*#i zy$Av0PmBYuWsC4%a+%7xAdEWKwVz*b?E;%G^)dB8Cjq7{CH|%t4K3q0$pauwxaY5} zccb{opF^QJxGKQ{>~^|+`PU~m2RTQev)!mw)DaRmJDZ#L3*w2UPX=9{bqt#hf*>Y~ z%|HD@DX=_oK`?cRR4>t8pv({D8s=$Y+*0Gu#rW_62r5`&Z!%nm!8>o)((li}+wY%# zA?d!b^R@DQANT`zD?`gckeT7SJU0#47QrZ8p#+-XQxOoX6GU+dtP?@f>zI6Pa&5>^Le>V9n-MPt}#U)G-}cQ*#-zxs4( zbF}4932XsPb2VK)#&z7hptVf8M8N*Lq0yThMip<0A$jcQGAU~7E6K?4ZdV1puCtpvi% zEZ4HiYJY;&PssR!Lby3kpM$Mbd zOSu!6Y>xqm;^5>Igu#C!01E!3MU}<>1cJ_jt4^MZ*5rhOc+&i&xjpDWndHko6vj!u zNxu4dB@3@A6lla(6xnAwAQEIIFDo7nqo3|+{?4FG&3M}|2*$gc% z%Kp^|>q}B0oaal`za2hOF#ZdqnZkDx%8lEdA^vD6h`fx9z`#o?$}H(80iVRng0^pq zuWDnCGziOWEPH+@++6fG%A*AL)uFV`aXT@fAeaAu@HHa}+I6k8_xCu`lG6k;g}1Gi z-=iUXGPIDd1rM(N;8+IO24FaI1umU>2$cn_*nbqr1)PFlA06!}6&UJQez&D6vuJ-uMPm!lED8U1`T&!5 z^Itmu4cTZN2`ad|xAAEQgw}oC0}AH&VGvS^2Gdf>N_fJxYn=a;mrWIiy1t2!~uq6okJ<^=6Rh*u+j0bdgvbDXv4#M2It-D&@i$Y#z zpFG=+MT~kDZR?Li213ZIfc9r7j1~?*$KDnrrB4F3u0|laSUPXnpd)@L$7$aCCavx_ zeV&YO1VAK7EFv1t>9npKd-^lQqHv|2$OJkrq;x@m_Ek18=iapby8q_pQ1ahTq7cn= zlU+t&Flw=ba8kph$pYA^MT^~y)~ou!qeOMPL&P|BuvCMe$Jqw7$gViO+Ls(oWedW9 zOMywvf$~f%xeh2J_OqwZG8R{jg>>1>@Kz8*{!syFvuHUW{EGaH3-KM-kfMSc7(P{- z7I9*LpNwu5jT*p(tU$2Ni=-pI_2noU!d)?xvN7-CeIxS)tSjPErXA$V`q5WKtBCk% z184L4wA76M+#c&$Ikz+AKqh`*Ya-EPi{L*!{;VP|90EVDuL2oE3s43Xvef#-%Ihf}5KAWEn@@9aC z3ABA#%0_J~ftPXnV*kfE=HF6@IpQOwUIsn}tXA4Uv1|V5G}MDn*+B(@IIMmDQfW(Y zz_zP1l_$v9#Rno5<=sVA!G?ibew2l`tJpmx39_ojDCSpE1|JH(YZ+!$to2_FRz`dA zF2R5Qz(UA;0a1ynOM!MB41$e8G?e@=JM-^n0Sfdej2rF)>zVz?iR z{W5)EZ`Ci&+Zd=UrJUiV#4_;xTZz}82e_Q}&?i=%M!q(H#PRNV=Bdzr;z}p4159U&f6O6fPji672KsVCt zTHYahXtUTzwRKcI8{}W)eI4qZF&p4@MwLla48re?Nqr=yau2ohR7DjKm7&L9WUY^# zS)2hE1%5@SSJK)QlwnhrLvA`vQ zFcY~54@B8YZi!`hnayrylU_5fzLa+ilW`T;=zg1$5oC78p_alQW&pkJ>CmlOCA|G5 zN4Ep(jMFH3%4!f)EabgZ!&XBXuAIaaEZHCUJb>wMCA%=}VjOq5aS!UwNA)+G6Sm(Z zE@JOIWP&X00pU*U*Ur1NG7LN#MM+dSRs*3&?nyN4Z%E4y8GtAb#!f(wJRgEbsSg=) zS8q-+cL@V#raXqmrtme>xF!vPv+QnQ$^b^SmvjS|WFzbiNe)3Rs$ zw>H;3ui?gRHDOL)HcnxYqi&KE*U-+d}vciZ{VH`shU==&m%cKv0pN zE^p_tF!j~V#R68n$s;X8tU!0ZLB8bTB@C^B*OzW^y}rU!J+Ywls5lhNx#^}_9~YR( z%NDRsmTDrlp2b~*`QZ6-dAn7_T>#xEE()Yn%FM6n6d>?LL3XQgh!oGD4X6)-AXZ*c zlTx-K(RvvmLjTX_4-OTf&*NLz1N6kp7PsG((zHPTl>lg>pVu5Yt#6-YtT2szdM;pz zkCaOF6Fqa4)z$zJ4kV%gNzSSwxip#;t=2(Zfj|~7PS$O}o#MN2JPIlTT?rI`xda9I zae}0r#q#djjp>)tyC^)&8d68L32mS^MZ{&U$wzi}qhZi9E0~L>)^VnSnCK-52xgR5 zqwbW3(`4^Ae42|t!QBW>a5xz&6s+HO5-dy|I}DeqVY*_ycl{tK(^@TW9omUQ40l?Z z`d$4IIS}*pk%9_vdvSeq(dwR)_%a-%GT%&1XhYvE`FQV#=mwv?X@6rn%s_9ixSuydlIu>Myb5#t>DcJP11BTs~{rsMw03tQm3-4rR`~_lN-n%;tazJ2EzW{0ircUjRTSH}Uol)QtHOsB_nlJlG#2VLjF4wm? zJf3aD3_LZJ$Blh~PlnU;!1ogg(xjv#YdC$zh*d_xV#o?dXs=26Lt0UFbA=4J;*1M8f5E zF0ga~rv!-Q#f}}QRU}CsIHuS-r;IwR+rL89jMk6`-PTE6`29UQHWpp9cB>FIyqUW4 zrd`BRl7&$Z$j&F6uU*GI#({!avgwT$~F8Ga=#gT=Z)8ktqG!OI4|0|fy>7Qq8GGFploeBuW8jthApstVu zQ`6$X83CopAUoB9@8lZ5fLQUe`+M|*ma-L(B3{|;?BZ|unNh-OIo4`CxR`aNmG8a& zMWzzsTb0$TU}~{)AV{Wu9te}Pna9`sxMiPEx{4SDIc{g35fFgkI)Vb&#|bJ^qf-Xh z9j6U?Kc&MCIX}^})dd0Wv}>ZhZouBee0i%CVJCGrDCUw#MyX#OUat=#KY&laaU!gN zO!c9(&dDzbU(9;niGO z9Y2r+Tvi~HY5<|S46@2BF9zY)Wd$nTM2a71Zhak5U_+i1Z0n5YA^t6~Um@xiKyWx) zf+ZV*kJ#!#`(q%o3k4Zo&ePsPIbC~~?uRDMECzyv?x7$z-wtEvks#{T+*8(sA#U5t zV;#kqXv6sxq>MoG| z!l;xEH;90eT3FEW1tsHmPLOJ!0NSBI`IyH?-bMsi^_57a0`!%XFsb6D`k~h}zc8P2 z#-yEhzoM{-^IPDDJF&5xCinZq_Iai#P;XXf?$6#V@bxb8S2xLDg;>p&KjIJlR* z0%Fj&MH~+>1z&p*Zm&gOoXx+SnX0RH<5*BS%eop2E><^As}cX)aLpQ}AtQZrF39+W z3v|aaYH$l6 zI`Nc__|M(A6x7D%ZEmzLp6>B;Urs6VvV9a_BB6rI#(3&XOGphjguwGt~x&eCONf8KhP879C`agC_gPNDSTHgj?-`!(e@If9S))8shpohr& zzd$h6s4w&aQW$k2S93dlm5{~lpW~sWu%bk8PUcJHso0(|*(bBJD1^I@MaN>(xER#77|X=N?8ND1DD)f)6L`_5RBftlwMv9qDu_1IvH*mQCQaZ^drnC-$z5S;Em8K(QE@Viej{T~9pDas^0grtWdxoFd{-|Iq=%POI`pM$FZ4D#99>gDe$6~Q zdDMn@FT&jTqs+nL^wIw)b>G+hq;9|lORXuMwnp4?RUoK))xpUgPjMpI04Xd}(h zWvTuWF|oXDRT;AAWVw6_w@#|#ln?d7;)6j`^<-1`($ z!$m0|jgP%Q?%hM5&?wKaWN{(=xAHBUNBF2s1Lew?<^7%b8U_H%5K+PBbz7@~-(7U$~g z6a~#LcgrLOao?H1-PV`-+S2B+Q<2!I8&@%EX&}0!C7HIMFd1HS9o|(5!q}AiA_~qJ zjF@7b0qg5hxWS$B5KBwz9W8a5BP{FU*e{N+a-Um!5rI&(VXg&WE ztKu40IwDjDT6kmOjPx#S)zyP{>+pR?hAZ4!d_{MZ(uNML+h9o z*)j{FgBy3%#5mc^_y37UfwZxsX!HDglROoDPRmI^UN-OBdJ^b>RO$y2=C>FIR9;G) zT^QQ5b_d2w4ZO}fR9~h;+%ey&-KHRqZ-&n_=4D(vHfgs&g`!ueb=F@yLXz5DC@beA zGAfCVu^d3zJ1@LPql_Q!&}|iqw2II)Stv7iBjw&~$+2G>py*L6mAUYVXuWAwUB>lRE-`V?ac;Z$M-VkI{ zGdF8UK^(7|(!KovqDsjn+i(DQD>bKc4-uYVikDJ1|ncrNSPaycSZCV!`fE*|wpu95R()Qw4AQSs_xmx23lw3xm_YJQ*zJ9m+l4!1eRAWpAh% zJ__RL7m%TOiKTj5YM2;uy&xkmFdawttv(2mc4tPfJ_?e4WU^^OjKs2_A~Td& zu=L6_9UuidZYxg}nXzkdTVRfP+_M!cOENMo1cRkKMkc`Lb#!4LC3Ffe>6aBR8*mnw z5%DYMcZHN1;@=u=VWIcP%+%au6ATACrZAC-nc|`P9)2(pyQTB>q01Ro~&!p&%@kFKP8LkA3 z$r<02`!(5u!4lt+sM3~*X4!x4>$QHx=N3SLsh}q^%$&oq^cmQxltC8#AIXvBu6A9Y z(H6Xr5npRJk-G>Q7neH@c1ZlZ)Cd*;KP=^W$K07r5CKtd^sUbjC;aeUc`fYy?X@`W z^haFWwERNk<_{@UL>2D*xB+?eK-N(1v^ZQG!D#a1`Kt$vw|B`gDWKp;8afei=j#v{ zZ7?w<uX=RtYXlAkRb|UoU0}3T(SEra1s3G8>S**7_>2OKfT3IT=S~t)M?jvVPH?8 zk#HA-JEPhE^sgbpil#$2(g|i~GNzA$Zn&o~`@VsYjGU%JBGQ|dhwpYyQ#z$QFaWN&o?aoDP8W(q;-d=M%*__xHhtv>_G2zF&xFV?#`mF{BsmfY9W} zt&jZZZlUu+O^-mO-2(ub)Z1dNwEVSn6)cP_bMWvg;fxk463unyN5*X=k2I4Yp{h{eEszNq zvWDdgp#O}n%E1B>CDbB}pd05A>0D+wNwb40(NXB48LlJcvFOX@ccvO_4s>r3&pSL0 z`PCS;FHbJ+5oFfa2vsi+xqLMO+BX`_{6wNZoaN_ZvO0r3d(UO340nX*P2m?T%iM}| zvPMeBI2m{a`~5%Hlr}A`VDC&Rx^%unsBfq$B4n2J?Z;hdT97*<=v|fCkkvQr54uSp zcLyrFw@iqKeFhiLo*?axhWIjgV5x2<^r>#hgIXro`&M+@-V|f8H<-_BTu-KR{D467 zONm>^fn4hUx+?Z$4<-?;cni$bOMm#UQEtof)79Gv77m-N?7c82-e^pU-ve)zV9oO> zGc|?`l(-&FdoTCI@yfi*i*CILTOf*CA6`fYJY>xM$#OKC5{svOOLHS~>4FwUNY211 zjWu9447&5>!J<=MhsC@QLF=@>{jN8T)~@wd5TJ65QH1QQE+F7H&ZnmgqBZbAW&jU z3$N(1j_qQG|5!p5PGBT`?bWnpMb^iJ{6U{O=sCfWfXvw)KW7`daji4hz zR9g_9J3rxvAZjbOFJeH|GtxE)z+$@f2ymt9QhzXd{{zFvr4BO{8_aVZn`WC`NUrka zt@d1y^WBU?{i`~9V2nQ8h8!Q=l}Nbk)@?Uj_1+BUc_{~^c7er&xbr_3rOtx9sY`lr zZi;*{9UXaTqV3KZnvoqh3~XdxL%(N)A)?i5Vu=$tQ8UluqlZ~8g=|SCBRd<1deEXE zS#2xXp=p3tZYJv)TBhpPhP`c?1qBa_7fzVb8Rq&kWrVN4Wbo}Cv1D%SCFa}u!HW-6 z69*Z9&*H(&CVkmEt;scHyI*^sGj3RVc<)g#Mo1`Cu@f(ts7x!T5$0QW_0v>PQ?@aa zJj};tO^uDD(8Vx%!Vi_OgE3H$uZ~Lre+#>ZsF|TTO6u5UZhGgd%RPXrv^b zDyNQH^vHvg+r0n#$Qhpyh*2kNH6U~So50>evI;?5jf4O92O_qHhBnBr01fa*QaJ7< zW`uOJ>B>yZgIfy{&VDPid02$#47XJp%y^uns?KWe62@$v6 zvqAD{j-P+bLzFb%N&fD{UR0u@VwR4)5s;k>(e5O;Jv13;$n=!M9k}>*q_~dDE9~Mp zHzD~d6|-omy3HYc92!L~V|4w34_@j7!guX{1p>cbqYdWqXMtcz%e);HkucYe+sGR_ zKiM%&Q1r-q##`(tVX3L(7dm~*%WWOuQ54Gu&0Bt>XN6?Loz2jXLHeF6%Mw8Jn3NqK z(79Jzj1H{oYFKE~E%6Nfayx*g=`;$8k3ngbChdl}qbvN2u9Wj^S@6N_Bi<-?=<5PO zA8J@;-+Ih?vclF;R+%@JBKH0V|EH^o(?XK%uel3483S4#R@%Oy`7jeM{p5-0o9PL~ryL!YtG&=p-e@zMCVW$ec| zfer2^2eENT&&?+$vLXbZ*Dnfek=SM4iGYH9i4kAggW&+}LknU}=|)>s6?8^E;}!Z+ znzGw9$9_I9#5w+Axg_&li_{p-m_RRd>9m=~XbvKo?@M-LPmgSxs?JM+CGC8(m5#&c z4Nq_K0E~AN{)mJ-^4^Sh4;mH4|StnH?4Mb+g zU@HSg;+x;S=LGJ_D5WW@8sC+M&TA|(>)pQ<_`{C(27?1B!nAk!F$+nu=shaZ0i9j@ z?qvEXTUsiKtoTXsJ=w1&qQ(sa)W3=4kZjzQ826RX?Px-z?<;PBWv{R&&(Q z#+uX8gmdwe_#A&a*H70z{Xn=WG)LByynsrnEm9FX2$PT!kPl;KX&^=L^Q?J)g zs5Q&KlNT33BGho)f7igKA0)hf@#$t87a_%3%Q}3BmXw-vs<_1uueA=&nc|X#W0msc zNf4!Ev=lqCFZ7eNUsIxa*p=nVM$!^mM$2TV{RL|k|O~0{t~R91uOmj>vSDk9S_ozOKALpQudt81;hjH zYG_#Wa-={2yf5Eufd$fp1{`1pqhbYzdx>WYYg5}f1FI>X(@Q^=@_gys^g^O1Ld!_WVkR2x(;LudHn{4c(;&G@@*Ml3Uv#&=?5Y% zxp=0c0Fhi7zocDEk1hV)29cO$K}}B=3%3eo#QVQ)^cj(W;t5%#S{8zD`JF+yHIO?> zlK;vcNP#}0y?euvjK+NUWFQ0*o|Ajq9SiIcr&^A%GwyiC*nDP)mML;{$r2S)zvFNW zjN=v-hEKSCmTaR&^=wI`FGJ_)WP3DhexQ|bewC9k$G8F__eHhqWlp=4Cs_*`E$=14 z>_5mMR7V*8b0W#?Wa3~F306{4YMJk-F>1469(X^q@(mGGa=OQ&}z6#6Zvj!MmRvfSP`$-A>A0u%m<{gGn z&$&ba|F7d1)qUt@y8&03Cb*pqdYq;_SRSwV47^=vu9vTqCr+D=XruMx{OXe#p0Wn+ zYktNXF}I%^Yd)4qGrxmaC2E?^TRd0SL-LN~#|d~`%&HRi;e+s^ug#@64gvT1kmR~SIp1;#ra zX40f0jF){QGYVV$1D+3gJG}eRLUOX~*0|A=>x`H^M<5s(g1O2r7*%HM%;ELdu>Sto zIPBLz(!;IaJJAuyhMyBfWPR?;_>@o z*|XKXp4mjm2RS@TzzdG)1w5ERL5}F?XsAh{Pv}mAU^scC>TJ>G&p*gB4I#}1)UrJ;!4G!; zU{5U7a6}$Hr*~7p^b_&1o82wO1?Yb4u{40BGao;7)*HiE*sT0L^pga6v^SDwa{&zJ zhv!`p1F}R;fa3d~iWksNqwyLLv=hv@+1(1tpi(VV_w z_Bgb6RZYMk1?q26Xh4S#cp1)BztZWvcylf_#_n4A7;!xQ$a{$cNZ{M%=poS@|NKR~ zfC?`0nAl3@@z!zifAU2M+gTTwEC`V(eOW7Dk5Jn)cLY?Dlu+al+ii5{wsgdGix-ax9v*91K&L{%=A zu|PoX8msRsN?0XfCF*Qr_KWFXLFe^$MCB(V&y6AsSFH`_(P@uN))jMH#IH9Gk^O0@F#y*Irsk)*K35{RklR-oy(7%{ zlXum%O9F8vnkNc4lWheDYejpM4>1EN_Jxf@FE1yD3GXnKG39 z=k=M!agtA|76Gv3Y3>HoT!()o1wIYsh79><1{^0&f4&Ar8fT}wl%5j@eS7(07+ft3 zzOL325x|8I+qI};i}SLZ_#Zu2(9VzvIMu77u2u#@UoARXFa&$`*^cLWW^>En{md;?}IUXAtL1uTXt8sMaSiN!S&$(;yD2asxwGcF&8RnPYEhwX{|C%+*htn!q zy5D1Q-wUd^3OC3fP&a1$ds44Fr^2nVi+PbeR6T9H*!3KmjaAP;Mt~ zTWqcpc8xJQCAvV6h(Q$D?bI=`Vt0g^bc(<51)yN_t{X2N@|Yo9nmhn+qHr1n0~__g zU6!k(TR*kaBlTZ)jE1-6{&ziF!7`LjUT3ntNy}wq#;DsSOK&gb#Z#KyFEP)%cIp0OMiu?9w{Ss zyYf1-rwSA|lE(N!W}J7k=#?MVLgJmrx)NPxpy|=d=_}+xRYF~_JG`f^jcRKKfvvM% z?uZYcd*N6pidkkPt)$ULvEZ<(ze)uCx!D!0xQI---uc-*3$ye8{#tVp?JWnh9~lMI zWvTKP`#3OU_U+U^8PPnw^ukjCF8Aff;%8;#ckWreSDbK%ix}Ss5yX@9klK9ugJ+vZ zBkzb9e@kCJE@2&T``9%qbrLjrR3m^s3}h+_+X*6RXN5k;;PLiX38b(?wgiKx1}lM6 z68~hTnNCUOmM>=E{1GRxF_ zu9p(8BC6iP0bVvw$2d|#P4hw8E>=u4j~vhVcQMNsl}&jR>iJwi1qG?Zh!pJIE`>~c zDTGq+Mek>NS%Yl%$mP23K1@y`zI^Hnbht#DzPuL3PjhAl&Rj?_38hGqL+9ajO z`2VQcji&j`eb9uZDa~?~&+(zVA+$}Eef3eu$CMF^+YlFhdB+ARijVPsYcoAR1~HWwXz_CtQKEhBAIYzw z*WC+vbg3+uo&HQJ))yn`kE87`z6>kr4u8fi<&pS&jswS@O0NDI*@=;|0t=n{zXU;~ zHM_Gl-qUmQE;!nm@wU@BOduLQj0>ACBXnkI-bbV%da2Imm?Ju!3 zO*|j@!M~;5C*p|OE?^nekZK)K!C%suSIXi~NnoO>Uy9{lUm{_l(Nbf<~TVOxF6(mGn3#6M3`ul+P$= z*4ctzdW51&g#>HGm?}^I_F=#7U(YArJYf=A*sM6;^=5eUwbE9BgrTe%K2|sQ=yS$A zQ`E>z2{F1}tLNqb3+71fX7Z*Qk2;pQ!5pxlrLr~)|6v;%9^$Hx~MI`>hUywUQ3g0EO?@~O5=LdyR3t~e&0=XYlGJ9 z<3Nz=Tz*J=o6Ew=$imiVOgZ{*9#3E5%u1%TluP4l({$PYY;E!~c*^TFn2b)KSpBz~ z;hlFxMGqNmXtW^frILO?r={XnX+sm9FsPTqV$5 zu*i4){^E4w;9rXG!wD#UWai0fyh-PMwbuEC$xi;q_Ygz$N_eF|h<3`%*{^=rCVxq@ z?Q0m!(YV~Ht%eL6r`>9pC^=|Aze0ZCV(?Jn>HNt}Ruh5+5TYb$2FRaa zkACH?0BL^`kvpd*wI>6~ZEg$==49RZN-5bfqy=~gOC}w_f4;%aXr0X&R>@0LfB)xm z3X_PE{1mb(*dC@Bg!9vUkT2WnX-Bk2M-l}ekENNbWNT31-Sw@FOi2fK>Cim&j~f*W zwU`vdq)t|Ba>`d7tM_*C0|wJh+ZZuu`i z)n#BUF<&xCKJM>GOiQZG+90NNa1GG=ZN#e2KFH&QYZ^wku3LeY<>Hx`FLoO!Rd^Gp z)Z?X4yQ|jxT~)5R`DM%y-udDb-LIrMA_<;)otm3G$9hZ6&+M`<*WKe0HbZ|4FQS7) z9UO*qqs?@kbG#vO`_wJ3_$k)y3}eOk=lP3Cx%UFUdYtn%c?4yh@*WJ2n!hstYWxlA zO|BSfeodg9kl4aKhZf;t9Pc1OFL;{&+`g9V=D_cvgiV1m|G<<3u_GUScGRlCiY59g z3lmK#zj+&BVav<8LdE6I+v>CNecd~W&xt&Wx3!_aRE&OpnXLVkB@*1!r2TG$UVK6@ z8Gb?6`5@>6|_41}&neyLML? zOgI{Y&TeO~yfJAsQ1YjmULFrb!Zl~*JTXT@KYS%tV0tJ2`|Ae<;`a;J zRY${p8dnRjdaGQkT&9^BfA^z5e|RmsKd>?7HU-VS`mjpWbKiCScN&SeI~j@`DFM>?>%i|eb_hNxLR zIY3u8H{tYTQtZBH5Cs?*8w=H**Pb5=I-qA=jQ_>)-r!wkt(?o7-#mM0i#+vKfgRV8 z^=oFc>)>@ohKd7+l$P%|-FF647lGkslnLCU+G0a9+~EF{sBX`_scjLQ3V)n>`irlMBn#cMS9wePvTnMc+gQG^cAAnJVlr-@ z{L<09XXb1CmnaM~UNz1-Bqnxm$OsV=*;hlobWhJl-B-P@q&C^rPuuziH*q1?{_)*W zy5@%QP=4mke5uTqq4Mz&I+5*P=G}5aaQgARke8~kBO$8LTyf>EiPR%VRTXvdP!07u|vIf){3at05l(D;5h+3rimGn1@Yw$lEdJ$j%I|FdJfi zP57)AZQ+-{XyW#<#h3~UPHgo8h;(we(F`O{IuR+%HN@)wW@fH_60%-www^P4@8-I` zDJ6lm{0DAg;8QIzw8vcoL&f$ z;gB(<3Mas@`%HiRQzE388k&J?WQp7}NDA76kIO&>)D%Vdd=+(buU>!?06$i?m z7`ZS=2Nfijg|(Ti2wPaD2C|3lhBL=`yfbO_59|2of9-FtIsZ?-rEXh(7_ooWVMMzK zHRH3`>}NxP0e6=X+ae|Nl+L})d}r+@bz*00Gb!#rELwi~zrScbco&HJr?CW_qvnOJ zx(t?uZMxwF7?rIVH#a-`7H`HeO7YEH5B7dEV(#A_ieAoay65%Ny>_@BJ1^!lVQ>rh zeL~!7cI!Mao7pHgu>ZW(>LxO{q!J|HVLlz;bXM&;yV&r{^ZnUsS`5mM%t$%9H|BLL zmQmAE&S!P_WYY$Y=H*hX>l`43@I$WlT1rBpxslg8%lJ>%R6uVm6HzFGLJoxTtFy); zjU4IJ)~J=nPi!}%_Os`VBVeP`7GGJuF*j63HWGR^3B8iDK+IJe^ys=8`!3T7H?yJV!WquA7TkXvO_Qbf;0-vz!JFT{l^-izU7z>vSH^10vX2w*ryVrI0uw>=DFf+pS*Vy-( zE`FiuFT?#?U=^`VrzXJn~9L$216}u#8Fo{?$uX?z%2EIlO8k4!uIZJo1?Ei z+Detmx0(;RN^~!Axan~|Mmn%vUd6gw(_SsMSF>c1@W`pZkyIR1c=2lHG^T)xl6g~D z_YarC;f^ge9g?kWtWyx<)7ZvM<``ajCzEeZ-#d{S`;J#ALd;%coCOKyQo+t92a;0!`)` zzIY03+qxR;mQ0yepH1&JqkO%&I9;F4c#m02KjVAA5*YmGD9)!_(&EQA^l);{kWg}5 zE^;GTXJt@#$Er{o%6+dFI|CD6_iI_|AAy2jvW%)1rv0uXLR*Y-H{WlW3y`Ah6;5*E zr73x6_N*xet5~FFZo-$TYHIAIagmd2!A;I-;*`QBaws?lw&>JYyE|9&X68?Qin~Z_ zsmt79q;uR z^aGxg^(rMr@!l<9lE`R0UaS4%lETA{&5U*T%;#fh#XbBwlhqbUZt^%xcEip~`IKMy zA)FO*n;#sDj!9;&%HBETL|n#x{g(GkZP$vZg{8*y5*^E&-XL1*@TkK_=7HnO3a*Eq zDvfFGXpTLG7c_ezI1~|TRpd7}m!l#@V|qrIf3&wb`G?5s)F%CCTQo&tRdhR%(xSOOB5T zoaoc9x8vVAAV+2gbgWD#BZ`z~aYj$%kz2;Yx8g>bNqL<*qVnp+P2avre4CevkBQMX zZxid>`65wP#{VK!uZTGaQ$6A(4_8p>PJ2#+-bLmPkzjx6bdNuK{wFPIHun&bADdw_ zx!>mRI5)Yv3@7nkL>m;SZ>v3E$Oyl!dv)m>9@}$m$`#-33s|hvZKvj*}0luIEl6DPpX2^FSG;;x&pL|dzF3OkO%!+C$afprhV+eFqxhWn4A zPbNGau9qfE8MFKu)m~DBV+u`NBrcNjB>-bkS+T50L(ylhCXZ#}+2AU1R9oQb;bgBq z9aSg+E(>-`gMJh>kC@l0LXAVkRfUOMx)DN3|fd?(ZI(qwV(DWMtSX&tV(s-~r0^(WW81&K z)t#PyoZ#kU#S_dxSv7cQHtLREvs6V%4(#q(YnU!Qpr7Jj-dYhNIh#GE;(-60&urdx z3QR1|b?kkJ`e=;#UVPg#fd;aVOKzTBK`r3`+e;1$*D$EeEG47+xVj?28?NyZ~OYP+YNtV+!!m@Fcy~YtFMk z%EA=#(iDV2j1d4{R0*;%*^%ud0$BiTx_t1Mzo|4;nxih3QJe#S>Tlz{eW^UlT$(bK zD1PaaO2YfW7Hs(ed5)#jOl|)8{(;I<{yFd7XA(P`QCOdERBBY}#LsN@tz~LgDG1UPrndZMPR^*fFOM2mR3kD zKsfy7@VDvM??i@zC-okEhN3`5Hl27(WV8JH-Am)yxCbg|GoA1Cern047DYE-Ugb=o zX;i=|Wi##{pS~=A1AHX6Pn&QNFcQ^HZ{zWJ>EAgVEn%x(le9kDUz=ACR zIKQN6=0rQNoKpYYis`=mR{NwKS55+iPfZ4J=D|%Zq7Yet@1ta=K5D^&HRC=D*xd0+ zgw-PSo)F3m6Y$bC7;b(#J8f@9rK&<#9fB>FISd3LrjJP+kmzwChw~EdkB%fE5S$`N zG^8C>OUg})Uk7%a_I#_t3i2s-^gV<+=bB3FRp+X`4- zneY)aDVTEy#q(x>a7B`W7Njfcx9?4=Sl{x5X4rqhsNxaEwS9xN;IKYLoy{4bi~%F- zIjw}&v!<=}S;2YWlpyly5p@(1L>U^FlH(*76zDtwm;arqLhs}*N3HO6zIHsV`ZmB?PQpw_VClWcnKimJQ@$8GuDxU_0fEK-fY z*E+TEpnqyMOnNN;8F9rwPBGD)whq@*8yklNJIEoY#}^i7rWx zy_nn65PS}6?CX?{u{p4*_?DimWRLPBMpqFz4$nSRVQsa?Ct5x=G7~yYF;}yG>I{|s@4d`$L`gIIo8-<4g}J7UQ{#qvefoJeRLa{1vyXR51ml2aZK_&L1QNJ5m<-rVrk z#XUqqpO|zl&Is2Do@J84s_&v@5_L@Ykz>tMmjU#@RQG=E5kY15y4EsN3p^nO`48<9 zi$yQ*p zh|f>ARtRM18X245Ix5uIq*$auL-x5%(`!VB0sKEFvFj+|MTf0(qCvM}C^4+$#5M`W zVkjMz7(jnx##i#U`LH`4gWP0uAsYP2KY}icAs_p!oThCq{C2O}D@3CKxKZh`L|@&k zHIj&@37)>q+z3X|d5I%!9dT_jV)Q^@f_;%ntuJzL>ij@ob2Ga`5r|7hMd z&nIlaA8oHS(QW`_-f>!$+mL0(AH7iCdq{32RwY(U*QwuWU@qG+umdrquyxg`Cpy}C z00^MycC5eu!R2!_EZ0US?D8ymk(9zy!?S<>r!;A2>emXybm8{{7^rH@r>mrfSw9oK z)p6(FrL5WEezYW85pB5ebZd>cVVzrsV~QH(V!gsHlq(8{iCoOE7ru zikwr?f4SII6RHh0#qf~}kqe3a1BdRFa#oTIDu0M+0^Q>SCVajNKwe=1IH_w{BA z9SSpo8a|LF3013~1zZ)@BxDd{i82=Z2 z6l%*9I6~HXgN70e#;`4{I_9eqa-rEc0gPyytvy+sNVFzJ9wl<=`GuL0WTpG19=vF*YbW#lg2TrLykNHHU$l zN?GrkW->Kb40AeMT%}i(1u?}0-=QA|QX%v$#mPtdXpmEfw*ogn**T3Xqeno-8 zC&jGbE`=kqy^zfEas5Sv;Mw}PJb^p>ReaxsFX4(t7!GAE2q>S1kQ8ry67nwDXZLiR z1-tthAjyWwAiq##4t>j(o@Mov9aD12D}q3ZDr&~5gB~;&!y0t2pkkgV<-##VvN6>- z3QAiBKl{e5HT39xCwU7$)mMn;U}ga3jSx|~rXJE)!n z#^J=wLFI*7(@QAf%*M1_wZR^ci4eseX?{K?l$_OfI<>mM^^gmMyn#PdPJl@EALCP&M1;%dtV zX50$(_cC?_@$5K5@oujoD;|dM$eBl1Ez0@D(rR@hSmfWh zPpvZ7Jd=2RY}o5M?vS#TAos5Sj_qp4XL*z}hqLLTnz(#B3H`WX+=5@g<$h+zF3Y}@2Vn7c&Ru> z)hH4vd!K{Uu>{b~gcV^a(u-dgpBuP69PSh+mh0SyZPEv}djl);Ywyei0}L^e6pe4k z>-#Hx^`ZL_7>Zv@9;**w*$(VWrYoiUQ+$8OyfCUp*ot64tkwCT5^DK;^;@|V;?`YZ zYsH;4fy6M6$Tx#|zIh~4LP}4w#dhT3)VwSqQSdNE%fFP&;nDXd z#jujNB364Fd7)d1evvYt`Kve4{4-PV_g@{gLcj3`g^Y-6x7oMqu69apS|Z{cljix3 zDaoTCy7Mx5>4vYcm+_(EzwY=XMo%}%?L=*xtXlP`(IW;TL(LEqq*J@cBEV`L`|6O> z!)$J6Th8|7q7xnvEJZ{bwNqe8R41vF-2`9HKS#ApXQY<1s+NCIKwz6GH-s!LR^*(n z#y9>dJ2_cKe0f(Gud?efZ0FtO`B$Ck2x5y&{3A}m$`oFy6Le1N4)|rhtlSqXV=W5G zDMHdU@+eumjYwac` z_}oq|y<+mIB?;&CD{Y9NUA33rQUNj?3O?U>7-%7+HrP zB~y)+l>J^l|I-OpRAjh2UNIv_aV?FDp0Z}#JRJw6L2Qc5B*RvzJpNQT+I8SlJaR^k z;MzqCfvgHe_0L^Roj+u%moN&_2j34}ee7eoBpN@**$+3TIYA)+hw%(@!6Ll? zY0&am?IlhnSEiizjpcEXxh&(;&6V65f$Sg_t0BY1Jh_#-JydUS^s4Sy;CtEk_C5li zhb&V@sZzJBU$Jb&cIrx`I#cOf{>L7Sjmyie4?81Xzhuaef7~DM!mV6qZMC`4;@HRZ zq@$wNI#=Wp{cHO8`PPEm!pqGaw7HA0)O@!e5=EC=Q#0xFEHBPcQGsvrmi2~d;_Kyz?#~zF%Wq$^=QVS1N~aF;-=6OqK6ul;$)9qL?osoW z8$Vpw;DEhd=?Xd3>~hDaFJEqDV}Y>(|7pgqVtii*Uj`!>uNJUz5BmW${|A2=iXuUU zpaK*IWexxW!6;{v6Eo}s2cU^pYg{%ylUh)b*jR6(CnF@fp2Nyv?Qnekb(dq*A?b)O zB9kR>4x9Q>{)4|)+MLDB=n&uZZ-P3-qO?4ESJ&PBW zN%F-7&JCmnYQpFd4iQ|t3C$!eAxxtV6>d#YWWTZBQSvQxoGE~RQwmTB^4;pa7^05$ zBK30D1uv+pI6r>+U^~}FfUbsfIi5d?tO?dcNwoBDBy zbt2UC$@N}fIxODGJ`E~B^-K1ZvY)Wi4%X7)XPGy2)tV*7+T^qHk@Gjjjkjlae$2c& ziIZ#kxa#feH>$fI77FLPBbcM)DSI+^qilWcFJ7Y!PiyL1R#%}y|-ZvIj$`Ss;gDiM| zg9Dm;W%ERv!~p#1!ya^ZjnASUIK!Sv-c-psSrWiq1I2%zKeP1HOr(@xV`Zu_*G?jO zJ`wPCb;JC9W?S~R;KUpM8|({_cTOZ=EWJE`p%8FzH;kjr*rw+WO-_{TZhpxq`#0;G zkU{9a4F5e%8^Og%u$9OU9__6bpbn|J&8A8i(4ah%K4;7yp-j$}c?oxAz6#}x`Q|F4 z2DXBL)EdwJFKS+%L;I)L1+H`n@*1Hk;>W7>J|k_IWQPc1l()*e<=>9QVVIJGe5UV86nm&?sbYr2_g$% zh#zavn`isjQ=qb!CjF5xP}vK6+IwEfD<+Xw;wJJEV^QC_ZNa`>kwtbr|C`bvp)}Sy zvzrH`*4NOV;#N0DPc&teI)p1Hk1Zm$`VI5j0iDS=I^R|{BKtV@n`RF9a{m%*l8D7B zyx#p|t||P@dq9WoCc1~iA$VHhD;OFS{f+MR;>umNWYpts2>}5q>~k(8iG#|neE?#d zQhe>?Z=c7EB*ooR@3=D}1K07ODXUP7#=klxQyU|9*+7+5ybg<6I0gTQAv!B8Kw^%T z7*nhv+326xor+oA;%=oFtj%F1FY$P!Elr(garz%GZqCNIYg-n0!Nx>y7JW@spF{?C zrW!+$txab~X@0KB?*EKa)q|#Go%PbwLQ2Y>$s20_^g;_~FvW9Qi7H?gIEPePsJeSb zs0CzSS3On$A4_hKuK^8Tg$qFIO`n|#G?W@kY;opFG^%Y!#sAE@D_yCM^0!BL;=QmE zC3?BO9z5jlZEEegrJgwi9)~SrRnvQFj7_cp@Si`w05Vk*4K)xW12q#G zcIBsbpqh!?+W0g|pl0Ip<)yOfQleX-0|ZB;9kzTW0=guPx#}gbaUA<1SqFvtP zC=4ZzoM|C8LcnIPlsIg6#&a zNdvZ0F(b4>7&XeiHhd`A7kRL$G(%RHV0+OP>dE%_qxUd z(h-APfV0K^Izg$q)!mw(z{k@mG|3)7sJy?tEwGm9>NIs!1}kE9zkHxXjDsH1(T?)q zzP<6rB`~uunf$q}{DCL;OMfE1wrz`P;`b_$`a{r+eDWHm%R_V0&b+D>CS+@yiS92Z z6mbuX!BDY^{1esUJr-Tz$%HOEHl3vzsXl(h&0iYUZv&$lQz9|2X#dSF7I ztBTsnkREr;PzaA;@|==VS8XB4kaK0?NlC`JaxH?PWKj?UGslq&NdUC4=?Lv2xLUUR`ODL(Y?u>sR5wp?MM#)O@NfAkt$`b{fOO% z0|;q>_my=NTM~~`0h3yqOEW^iwwNl-yT$b_=2HE$k4#VB|0k~<0$FMn^3u5=Mzd0M z9a{}|E?st$;m!d-Cu8M8SKT_<$s_}_OwgZRdGJzu0=~*B2X*W zI4!2Y(!9{os8c>rxyB7JqBt%`nV`B(9u=8z@U z1)-ebvQ))&6c^iEtp{{cUExJ(kNYPzVQe8M`3`X!FfC#+cfMq3-PNK}1U-1xN%z?`HHrOfKcF6qlAYf=C$67+uy zl?bD!ZL7g%omrGM5+o=<*QRQd;|e!q{)hR~_k0{1hxeQ3FZW*<>#J?e9!|xjX6TEQ zRX@qFRd4_y&+gQUBI?)E#c+z1`}s~#)^7#D&u?$VNSaBrSz@aTy4l3FYi4D={3&@;&Q_*fF85ZFJVdvYAL$% z{8A*x(PAm4I-g}4LDJtaotvNL!?g!9cr6f$J;6^TO=I%@suM(xv_6sxpbQzs#Tn#h zN>u7= zf#P&A`8LWhK>{Zd>-E}%o1oQxYvbMST5{)k&Zb!R@6X!4=ieRnmom+Q$_}U`#=d^9 z-VS%&OSuNvha$Tp0e*P;ual={=j0Wk&Pvy4aIQni75Ap0o3E(gSyC>wci#@o{tL~T?XSAP1lchH*nIo0~S}p&o{^ef4m%>E0HD{)) zizs!ckRCXLqq8SF6DcvoeZz^IAVsJjI}{^>&oGu}9(fn!AHSw2FPK46(iK zEhnz$R7i_)p5xiKtqda#oWv0QWQOCQW-C*(6H@%HyreZU(lniI4YqKphNUj|F2O0U=xL z=^5hVzo%GvKm9@aNSrD+E^8^(eo73>1??3-q8WxzahLCALa7|{)-MP(3I|ym>U%5M z_~@zMP1i20(F5dTZ{XkV%#D@iM|vk<%hQVkm{0uU#ip~JPkeIHYC_-Tw&g*OR~wXC z;o)+8sjjSOkmGD*SJM|*Gj3ey{%_^Oq{kf^!^RQ27!~XamcB}12e{f3?>%%B`sb5sJeXNnNuwT9G-=TyRI)m?;&z-3SV3;p0)>NCYvG4SkmKm_7 zed9%Ar!oLVxuGNSztZ1Qe+}K`34!QkAheG@2S1=)k!pLTJl5Y>9jyo2qU_)gedoJ5 zoGn%^f5m`piSju9w59&RA9bzaZ(LGbl6PVefO>muC^eq*Xv-E)+Je1y6a5LNO=954U)^|lo$GcP$CeU)f(cLu79Ldo)9NNw99a^VQzG%jSaAv+?v<+lky9$x<>(}Sj20Tfc;?$uYcz~n<3rn zZ|i3f>v-*KAIk6nNFjY76*s+eSnArrHgs_9JMnKcZ_2Mjgdw*e@qXX5ezgpI*o%;z zGq8g}wN5f{Tes(Ac31F-4&fw&YFkYsfOT_9QFK{D`j%~1jM?_+-(Q=>%X98*(Q{{& zy}99}eAuc6HuUq-jKR&A?)^@_edD>P9uS0FRq)}RPlHT`Rn;0GxtaPlVTcz$;k=s<} z-m3^fqrx`F`$8*Re`Yw74+6@JKi&uMaMdV)Ksppn{3iB}Fc>8a1b$IAR|Dnjb9tjofT?+*bJn!zJz< zcTTxEf5+Q&ykC_)Ozft6+jSgZebP}2Igip2n00%z;0vq$Io;r{Akud{y)aQbrFk_Gt8-f+D!;L9<2BqpmX{jK#!{Ko_v8cx0XQe+$H*k&!@c@~ov&GyH5pUl8KcCDAVLh}XT1 zi%SKU3Lz3Cb_ZAt5C46MJMe`kU$}>28_gmEj^Ngtvn!-H2!0J;QRmiTmi* z-%8oD`Eoh_0|T3mWrv0=HdQ#fFvI4n|I?>a!Scq|-P3q!5Kqu5E9*v1Ox3709r;ON)jy(%#`uuud856Mmu7d$^_^r;@|0ct% zkB=*>`^S7;VQ)|l33^R`e7jmp; z?OPvGsWR^11{e(W#POLr|Di9tPYz~$k4@%RzOd)iU+CU6g+xOoHUCtj7xZF+NwEC3 zKM~t7oomc6{J4)gpO*cR{*n&vO(?evL}wawN_ulA6|bcOZ8fmQ1;giPWv;Q*jwwP^ zLJE+Zi=74B1fYNCJ8Df92X(p2MKh|Os>pvyiI)5ghypZB*Pl0<#~DkAcM!K^S%e8s zYzU_}2aRN$TwN%W?(zHiw5H=n`z$^t(LSQQx{p#~YcMqEFsv-OnjMvv$_|$&Xw&je zz)b_5wr+A-rQDKlF|Uz8JX=!&69AY6%yK3JOaNsP%C^8}!?Y6!-Ty-<^fK_Ej?L`mhmOi(LJ6zNldQxrSHh1l^ z08_+S%px6ABLFOEmhg{=;K!HSA9x@g-4Up4JS%YVFVs(pw~)mMN*E>7eE$-9XhYPA zf0+%;eW?5>ejoLhCKCfB0(ZAm=`FC;eys?8592IYJ8(v9+?TC-@fz(;c_L&9TX3D#@>?2f z|M&n~n+l8z^R7*09_gCXSQ-73H!y6L8hQ@pi#j1(=B{(PEw#-Y3ZTX+n16M6nm^GF zi^O1{Tw6L@r?kw+DgiiRYYS%Aux{93&`5Xo_;(XB@wr{dGt6DqHTzol zAYvf@zpk=_`~?tvrm!RnT?%B@$Guv5TtjX?9`XMR>%j`4@8F&QoR81zqk%xyE3a#L zxUtwTOzBDl(aLVC0^l3@L83Len_V3p2Hu}IT9x-{mY1!^%e-Bw(PV^?U*tq33(M** zRy(cnx=ITr6{@Cz8W8}oAMyeZ=YYWVy_!v0gB?qaZ2xn8FKIgiaSNNiUWv) zfQP4tCl{fOX8p6sR!$g6T?BbPR&fA=s{}|!iqZFR+cZ6Q#5daVXYGgfOJz{T98i87 zQ%MWt%vphKAFm#@U;$);yi}lZs!yhU7K}H1m%a~agYcz&mWj#tUd?P4bP5uP zb^$z;&+olZop2e+GwaZ8UN&+AVegWCgSZ!>wbFh}R1wg15md+?4 z1Xv-gqqq6gGg|4d^>v^&doJ8`mqkTYzfHUfKB0E{{zjChV~4k68(n2~`V>b;r7710 z6*`WUPS5i;e;DfS*?PqpFL>S_4&styeBKMIP8d7Y$!2Kjj&0zk$ZRb8=i-E#X`Zz`9{`REas@(%o zNH8oA*mX~t2}Fl;d65q2qyi0O*WUd9r8|`Bh<*%QBCwx#S*Jp)eyn)Q=H}ICTv2X> zcU&m0csHzv7XGp?01spH*@dFY+46a#_n7LmF!xLSw}q6B{8C^=Vj z><&>J|3Sgj&)n?u%bVeOdTqa=G60>*4|RZ_Py>Md2euHv!yO;X0t8!A-#%iJCBsO3 zJT8q&=eG#}zJ7N-$;<3DeLQGQpf$i1kz|X%&o3Ny;?SjduQcuL)id6m=0WKadb}yk zf&0jJ_+#Q1I*E?$l0o1!?hk4dHPXfe6ZDe+v@Pj-s|!aEvBP*~QWN2+a4J<4b%HvM zV2N#n_bB4?&C6s$QMfo<2`8Xg{s)?n#-mAf!*K9sbwMMspH;Xv<0haZeuzK4-y8W8~6zhCwo7 z?|vIVnQ!Q;U8kSBbR1r$wd!^#vwP{Lcs1&twvp-QCV(Zcm$FP=3DlR5l={xPgGG(s z^&zn3t=vp#KC}#@6GK@8yri-Z*Lg@j&9g#Vu}BW%)ZJ}wnRC5u-8|h2fa=sZyowsC zUaSJ(1D)j%L*SW7@O}Hd6QR>>Ws8u{scD)ZkTG1h#q0SE0*;suSal3a3P5d6Q##NB zppzY0Pt2E)r5_G7ZU08(kCek{?A47~q$ZpEP%{Nk)AoMAcpdILCaU%t@npc&XYMl| z=^sz-PXctEJ9N(@Ys21ekmim4XJ(F~vA7Aaq@?zucM~>Czm`&aUU^>m;Y93S<=aD_8ZK`)|kEoK8-cR%^HX_vJR*M;2xP z=)&tbz*zzW(K87EU;(sq?S<*qET4=b$hO4$XMO*~LsiPJpq7}%jckbmpo}u)9CPt` zD_|fD)Yd%{{P)&0{^q?Q-8^D4-1!r)ujD(am)$EI*7;oepng@);NE+V(X9-&ya2T0 zz?9CbTihAdx_ld~!_X|1az1`Ri;_#}rP{X8hgUI887J=58~d!K)0k$FQ^G#!&B zOEk60zFG#YL$2{co=}_Z#7}a!xyKUG8jOv4wn8ycZ{--J#?aI1QdTUjvrp=1Lc+i+ zG$i;DE(3TuNczg;LctF{dKha+W|%4#%CfJNNe@+MBXtMQJ0f%yQ6 z6m!q2hr!){5o2d2cLl52j!}G#q9EIXl*y3}oO35CY}k!n+pjTCos^>hDPT-&@mC9k z;p;Dzfr<|*gW^h|HE`)FD-jKNHf(WcG(}W^!Ml)fppCWTJat?-VLxZM1o$pf*DG2U z2e-U;P=O}cit2xRq-=`KK#!CRRGFPR@KB91>upr|48zc6=eGHPU>9f?zwx3uzkpiM z%ougjdSOfAWipV58Y<&_D3W@bt$kKMn0>y%{z|6z&-MwazY@7;I4Bszgt*atOxb?iZn8` zK>fXXFItXUv-~bH+r!c$ple1- zo=s{Ms@E?J77I1-yUZR}>J@rlVU#>kJ_;|=gYE#Tpw7SH5PFI>|9L@eR1a=+jnk$T zhepITj8+;~Wq;1~m51FiHU^uAo}EvL-A4%JG%aOdt$o+H(`HB|1htMpbwI{mIe`9> z9$1(bJALzE+@a(KR6H{@}Z}=Iv;Kp22~hMrHQ$b^KW|( zfX6`U=;ueIvO4-&3AQ7@u%dsIrbv-yM~*ZEi6$A1HWbyJZjriT9rZ&iB$X4%5fD%U zV6vxLb8Ol}d;{$V*z*AVox=X7f`RwY=>(dJhy`0%wg=^S{OA*IKywipMpb}I43A}J zkd+=FSio;tMl`dWVs5{`_)@r}2`ONE=|IlUcGpMSm-9K-{wV8fMYurIB$XIfWex6y znVEIWd6$CA?#u7X-*=|6>dW+}MRFrQmVsow{U^8_ftnoK)cB0Y{mbm(C_0Hz+y9mF zfqTm%y8}2yA143;^Zb%fOgZ3=)tBmPtl$G}1sgN*yRy3hnd?KIG*6}ex$_BH>aNIv z$AndG1C|5xBLM%)-kq>vEW!0j50N_pR|oSV7QT8 zyC@d{IOwOeE9wn*^U6?kgl*3E|2AdIv>IJ^4*8{KlZsdrH-T*-Cd@L=hPkbsH_ki2 zaji^S3NOeRER2^XYDn(sFSWPZ;aFJ~W_08sMa%zcF}0E_L@NUL)!YHk{WK8iY;J4E zE=< zo}X-Sx0we_z`pZ?5RbW+3Lbxv3cHY1B-GEJPVZMP8n+Tn{*syOg_}3tqU`!zd-Vn?=sc8Q}V<6QL217FAD!HiW`@qPoXWS#} z+<3OrmiV-x_Y!tPKlgTjUypC~ea+T?Inh$P$pgOYG4KZY2=Pd9uDC?mXXj#gE3seD z#|FSSQplNqn$pDJ`FMY@mERpAK|z<=00o3})AE@>{%n{CTxNu8fg1ox9T@=XdK83* z=T-9k((1?#WJL3moGlY%;sFr%np4jlU;XYC5flK##wHU}8MzZmjit8g2JA1F*Td8F zz0My&P1V~+Cq&h%h{;jzA^-j|&x0M=QKv)yQRF-UFV9+bhlo!U#Aeqv;%|CMnCUaA zcD5J7CUhsNQ#qu0ORZ*?6Y^~!Es?~A0jztsKI)|1uW0XnMS1F!o(6n6ZCtl&-$P02 zW(hJwdE2^JLvF&eXM=luxVt*rd*g5RXnL~0^D#r2$1n4+4oPn!J&G*h{D|a-4Cr+H zeL}u0b?!K`{L@M#QTkfMOl9{q2Gz}kMF_yrW%_gL)B_U9&$4Kc^Qrl`V&8*y#%QpyUz ze`(i_%Bu!2^(aM^5~>r}Pw>Nv+DZT*Ya>kwW;r148v0xBo(kubN=w!7TpbQ(bCt0c zOsb}T<--qr{)JW*tj#W+04&wOY$14Yyk4HnAS?kOC7YR-RW`zebFp7;4EXUGP|H8a z)HIud2HY7@JdZ?mPiafswQk;^5|oSg-@VK2{~J6ktX4MbJ7}|onghiTACXExfO+?F zKO2{=%f21{QXe(QT9^uvN&t9nsyJu7zWRhU1n+x3 z_?{G*DL270DH=plsZjVmp@ie3xQ*6KATzHOFCmFGf`!@SHL7u^MT~|)`OA{fI zI)*PW>cL)5o@K9#5FwPKYf-g8(2;5iH--C)wNl)q&EN_5Muie#pz^c$nnKNjnN%>V z;3A3Jerb%wPf*smxzM_mm zXN%Psrg2^V9yN&U-nY6?@;wA~K6snHvJ9)HRI?jZ0{Jj2F{h>_9AX}`Jtpt0iW-OX zjW9o;dYufK99@#~k1W{lJIOUB7hr!pHc4iHzmGbsB|!s*eToA>=nTkkKmCuBFz=mk z=W@_M0<@5Kh;n7MW&i-^HJRJBH)CzuU5K17upD8*R=0W&J7o60EZV|EVp-eSlm7+! z25TmWF-f1f-`JNQ%a0C5bHl#lzSXA85l z(UbG(w=}SBB&m+Y!pzsB0l;2UzTW2fgnH^=+&=D^{=#F-TlF$<`%jL~_grYf-f?Dr4Zik7`@W5L&}+(UiHf~r0jKSvnk*|qUK1-tX?iqFfN$0}oyw)EUS-M%bP zA7F_XfIbY6U|b9EQvQEOdN~M%0@Udg-#$M#XPLR)fjj@46% z&a-!)vc=G&6Wj@GhrBtzZTLG`%F=gUt-MytEYo%NP_bA!AfO4+0qan}{qFwsVpYfr z{=G8@*TtLppPGr1=TZP{ejKO=oy$KD%y_8p83H?7E_N6|CQiz&^MN2KY+f0&=WfOY zdi{gI`^d|O&^PLYV|D;XP<#Po)M#PSq%I37T;=(lH zPtNQU%0xX@zz)e}=sa@%=K`<)@7Lt7$)HIy?{#A<&KJkWlj}^)bw`io8;^F6e={NS z6rRr27p^negx5I~+`Q)XEORQ5oQs3f72PP!+U$|3=^)rLzUN;S?G2~@N*QIWzUQ@I zTZsV~^!Xmi!xnV)jrcX;#yFGTAdpIs0lai~U!PQOtIgT7+2r{z}z;oGAqC^ zZUY!27o1K0O@si($fylt5TG&gNkZ83f4YBY^BY(v0sATJ3!*_$cxjNdLABUA>=-iI#|CwDOtQ#Tl zhw3lY-v^0LKFWA&aF%E*5V*kL#L661Du|h2${w!e%5Cp%43(~}@qYIo-hNZYk#89y z-*-&_L8QCWb~EWZKg*8k&}OOrLrk?Gv**NOBTUNNta_S*9vns*KvOngS( zV_*xcI?LV`kQsedxWlGj{X{+yN;xxw!vrED;bFrxUfJ&LjQQysMWu2(VwxJwjuoZ$ z!nUt~!jY$IJ~u@nBA=(Mh7{OpNCXSGLQ-PXgS*IV^%MYt$N6#eYVI=T4$PL1e7CHR$N*Z(;lY4_kj3 z7ge|ijM{XUNJxu{Al=<6sE8nv0s}~QHw+CHB_c{A-OT_)BQ4$C-QDos%f0veJI{H} zTp!`Xz|5>!_dhA0lf9jLtG@Ua)Suz#73B1&CFF|M`;dWZpDDrV87I#{>%Mm*0KDG2 z%n2ZSq|F4($9d0gU79cqy~vJKd5F0%un5erw8T-`!UCPiN7ZyR?s#28^p!Reha|H5 z_MSIW66!VV2{um>*rFp_cc_t;>_12N^UdGit=`_QKup52bZcFXCx+)u*1HGd(n+F7 z0W|R?Mq(eSFw@T7?@3RZA9OtE_)Yv6%@Si2r{E5O1oe?wUn;6qQjPb)Nkm6? zSvs!MniY{vFza>N?K#N9wxTO?ZWuw;Uor=9>7I%PI}!K|(PRs$*funH zqVE#cU=Q=>nyU{6YF{9ZY}cO-eam8ctAN5$)8Vlb3+^Ibfb{&%%@Zo6-Pf2bf<2uWeYq2$Ws5H)?VaRpzhYP{t$+v~T;cl11xHZB z0z{9ikWbz;tcUhqk6kc&LAWjhQZg9Zk{u7#u0~4l{T2iP$UN=dS^KpG zcs-vLoB`;I)n7rM56M)|Pp722OtL)-@hu`>8rw$!N-ILhoX6$nb|E6<&n{`b!WGe# z?fL9MJ!S>M+IDgVE8h%K^!`Kn9}8UC7a|c3_c3?2s}QR*^hHlh-}PkBT5*o)6jh<9 zBQ&Q!IOB6W@Hy~Z5cn*gx^Ev&tYXG_*lbi~Gv}=p{ciqD#C@@`mYZ7Ggj$FW!>*B> zv7J~j|B}fOysZ1fRci*b1p2HFFfP3K6Z!F>9naB+wHD=~apTf16UwOqIPzk9!|wU+ z=NWQ|Eb+X9qR!@IXcs$N#?EKsI{P$Jc6VSVy;dP*f2}o@COp_~5fUDLt9!XS;WJ7< zrH8C{@gcl357D1#sPS`0AG^@s49F*LP-p|NJC9jMb9K0Q;6Ypgo9v@n1*N2_47zZq z^M2#(@!mpaP^0XzN#Nc!Gq=Jox2R9EF-}EeqrE@;a=fP zZjFZf)5bmjdCcZp>GLMFpjeGoXF6D@cmK=iUtojM!6R>zLcV z4yTFh{Cpsp58o6V(7bN5k_rLhbeLW<{#G-0lnC z+@{OiGKVE;pxtau&%1`q(qVC!a+u~5tHWYWpX|PySsk6UE$c9g>VK?koj0Lb&Lh0| zm9GeSyK{rQY&Ps|OAMDV`1egvK$1oU;QA?|$6*O#uEzBUCD_XTb_ImWV8b42n`+!u zD@(Cama=9+eRwJ4aEZF4@Y~X<{;E_*`GCA^;0ep|+l|yAY$rpA^mhD&NVy}ZD9{Ur zG7}3nEVCLf$W?+=C6$bxotcuyWsah@4|sdgNE_Fm%b^@kL^FHWrDk-h7-#e1Owm?~ z*zpnU(a%XV(ER1=22yl&kd-d>%A)B`uhz-3wlu~I5A-U*ns}dt8`0}VfpLJ!#;US- zz7Z(RjrLSUr8O$kp~6bc-&^#{e0G>t4QY>!6!Ln@B@$F%C;xxk5+AJr(eC=P15uBI^G{{V}$ZPQ|d`|*Qdj~ zxwE(Ecl{oZzApHs6YyM${lcK1I3o2M@Sul5=8TW^kLt&467&)S2sev|@=>zcJ!(mQ zFEinvGyJvY9(Pr-atKtZaE1qSGS!p<4!FAN*>6cR=;J-?He|9=wdE0EJD0c!yj-@t zf+A`d{Py57X7*{>gZFpxeerX-^6HAHXR_+w`lQ~hFP>L<%a79Ox;oJrId^Zw4wDnv zh*qx8JuWIQ1um~WM2|N5a-VVuue$df#xt{k_jG0raV(3_Tbthmu4OxB)BXNyJ^cx( zVow*}hRO@jW1gzdfHE1Tdt;D>6z7l;y7$<@%mhmJ5`DihGCh42?E3-v0l6R~84;T( zAYti2@xc2bz<6M^)X^Sgw>%uJp#hV*3Od8zr~Gq9UPrK#1F48suh)c9#6qSVUhDv9 zJn?x2mzK5ktqI~$L>-cQ&yS&;{kCp#woqaf8NcgL7tsAF?yeBC?3Y!JM1&iWi72Gc z)G4XoTdw&A_mX)l38f?E8yAzFe?`Kii_Ij}j19OHi%4J2XQ6Vvbw45<)EDR9wx0y-RgTP6(+(Y7=4s(60an7 z5!hTfFp?S+b*SLY*mdvAKHzr+JFh*sZDm^60!uIug}0L7b(CIQXE5KQcLteFZ)6HQ zfy>=d#}5~u)|oQ`G&BkeV7l|!^Y5B7gQfK@22Q=J)%+rlsNR|vINgDaa%B$KfDPB?c%~#eO00VH*frex-l7(VX_DN0mm{ZE zGgC$Ncz;y3W{w($1yQzjt(`u0dHS!CnpoO8dV;lFjavwAb2a5AZ0g*z5l_ArvJTYx z=<*8i0RgQtp#jv`eqC-c`<#m@k?s!<xkX-SF z-gPJE6Z-JstDJMSh_Xn)?+*OomE4sdIhERl>iWBM;-P!$hMJ)U%FsHl=@&)+hGY13 zu162DWjZ=S$vAYFzpC9!Nm_CjXngz!z$hs+)$6Sz>L`rTW3CS_%Qx2^ zgfEgR5wcTj@)!BSl`bUKPcdQxg>Y?2iQ;yx>k#MkX0YL!9Cx*^dLSdyUR33HC;3I9 z3mKV*ML@0#TFcmn+jvu-dY4dR>Ms5te#2h5=jw-yDh~m4ya9YbijxcZ5J2DP9Zt%X z3h~B2;TS`paM~a%w-*H6kbB!r2z)7T3Mc9{{yuY*0;H}hawnMP+U5w^PmI=v6-G#l z|MK6MkU+UFG#oZ%_u`FvculF|r`=sv#}kzEV0o}BA$amTr`k1-erJ|@dfavTUgLC% zpBc?qzA`Q=x?DdKVxi00WB){M`U~84fflyg==u`*F~1^o1MH2{SJm-OBI_j}AjjYw z6*q-3DpKI6vnuUj4m>&1?zeqjPhM0s+@y}L-I%cxS8cmN6d+2xA)dvTM*B+Sy93Tt zE>gp_kaK_q2krknqiKx%JDzzF5~+$S?@K?ww<$K_n#Z*lzbbxPIrZ6`EhGOYR_DLR z-mA;#6E^FtZOzHziicUMZ?k`^V5$UX5-3p-Nys13_B&#f5rMKLQ51Ol8RKFIfm;s5 zH)sXFn=CX*yh2DCXb;>_A0BkXuU8|Z>9*pN$z>U`vZM#EfAn3dKlR=dX z8Dx`7&^BI5fGHUGnP|5;oUeg^$($r?iH*BL#T6lH6uIo`TE3fnqoxJD(&g0_&cho; zIjcE|xgP$ToTIu0$T?83g#Nc$YV#)VvM*P$P_TBQVdi7)9nXjNyLbUoIo@xPK8&-c zJF`fi-5W4rUM??Glct!$Bj^}!lN{?kU74tv^V#vWMrEO$-aL`tA^IwIPsoaE?5sH) zyBgb>c!M5eRWH%U{f3EUR<$nM>uz2GW%$ouaWKLqA_wd$^s=0Z9E;o>Ztu;b{6U25E#GzTkrX;h$|D$g z&}~lJOn5DW=Pg`hoe%P}l|N(uCNh{mZN68Lo0xcb{Y~~{hD}^GVFqt> zHcHg5ze~jQJ|PMsF%F!|4VaWVH_}tX{5j^TgT8D!^LWyae;L7*8we4^yVA70VxTwQ z_DXh0uLTG?9mTuunBnk5Y)zw)y?ecXxebk%S3UBD3a}a;J>FM{$v}2n+cfL$I8Y|z z>#iNx^Yq?Hd?+ptin3m4PYkGkb)BCbhEAEfU>?SH*fbqdO#Re3wOuYU^g&}oC-7=X z{>-YC^*f4+Nx}xved}Zg3SQVV2C;`B{sz6rCwgetMI*Xg9Wt=zw}IaMdQzIm?rlND zBmp9PM$+Sw<*NtZMt7EwUCq?KLz`6PQAQ6*0;A~D2G}S-nV9EQa?tTgn0v3G0UdpQ zc_#*`w}X{l@Jctj+P_;o&Ry^@xhGE4cB976qw&~(6Z7eCS&q5dU~-0@Vtp(;IXVT{ zXRcdwV{vD3$2rcDm(WM28AD-*BhIE5TaT(Q_QXDZ-s(B-wFAtn)!&!(ga6E2K!H4j^*mT{Y_Iz^JOw(Js(+!5S9RvWY@G z)mJ(roQ+m!6jqrWvLYhjwjOI=6lo_-^NuF^5~Xxj^S=Xfx;`-Xn@!3_O_A)^9Z(HL zagT%hJ<1|`tJc?2-}Rfp>&(liL?TNlpb(JE?psuRV2VNrCtWoed!aTAQNsxu%h3K? z_0R=7A@%B~)br}H1C89z|L?MgYg;IDmyeA|hAn!8ZVKF7A2g~3lX=1NzkJ18vwurpsukBpJ?mtzinL!BUP`E!Ij-#~xY7{K&Hpxa{41W!>_@i~}E zt3wY8GkHkE_w9ETO5;SX%{#%cV@P>UJOH!!0O%@Nr4E= zv1SE_a2ZlGUm@r~`2O`9X{l2ubyQ^LW{tmkc%bb3x0tB)Xd5=$QK$0SZJTU%q} zcZ6G>yzv{R&y`U3yzgat*0YxovH1&l(gRw-K@^O+5|n)`i?Yca_bT?U;# z@uixe%xLOBtW5x|{bEH1i^VD}=)H?_P2CIiHMD`9oZ8od{##`i0;S$bJ6X%fm|a+% z812={oDg=spV~xjnTB~l>$}iQY9&T*(~F*?1m$zH<<#i#Cb_myuw8!s-+H9C5o7Gd zS|9i-J^u~uHwYdk$fZ)(m+1HSef+Fa^n7C}H_G5c!YAj)BdA@+KQ=i=1CAeV%8!I6 zclB_JA-4jlc1>#H_KB{HZ+LyE#7}&-ZpRq#h{$MJg!e$6tA7~49EBHL>s-ndX#rFi za+pA4+@IXig-kWl_U%r znnB1oieCUD=fNq%@lGG|=`PVVyef%{uzGX3AO;DO{ptYl22SH4$z3+RtBf`QOBQRj ziJYSV&x!1#x9h0`*ug)Ec^Ezry`uXr%Hr(%&_;eLtWrkZ1bFN;)9)_7O4m1&Tr1;= z&Z4PPR5?dkH`iGM?YBb9qJJy#+XExb64KB<4S=%`1TK;2-v=lw|6%z+#V^6AmGw?O>o=q>YtZ;{Y$5q+T4zVHG)~ajQ%%A-Q|orRpnK2T`{=lTR3`VE zI#d2DJBO(bd*-&ri04l!2(>VRMM8bF8#{su=D0!5HuoZG0G#Et$+{LG#E$ZRO}ZMa;Kw6os^np&1*%0=MD z-8p2#E?PZ);KSXp8nFW+3!Ch!^c9Ns&~6|o-8<$)cY@O8WOCfrl&W%1Wd`wFP{f(h z>4ogfG*fb&%K4`irYv~-l7Y9HeFgY8dp-FoM-|FTjk$={iuo1)Z0=JMoJEq~&TOtv zx=7r{eM;qZGQR(wpUNnpQFMR+7~)m_)7|)gcPrg0Mn|zRYrF}5gm{EVlPcmIs_X+u zTBm7U9j6pEzt^+)yp<%Y;Ih&iqqoAEB&eVwz`VpeA>Rt%X>KK;;UmBjXV~*^y!=5&G60!;95!}#R37y%ZoI5~R_E`b6Df_nT-PQW!GysEs*3D~ipRmtq1h_dq z_;2XqA(1?7n6Py4gQt-xz7M;@`K1PEt+pWz|L%+N=sO|zt!*b}%4)eeopm-URW+UA z!{=(Nv!j!b%M~r+RL8U&O}i5W7|o3O+LI&g3cG6sPbX@)DD_AHl;yS`4BBw^tSy!C z^b&=5c__8Ir(~z$43y`WU9>=@t~DZ-^M=4f&Mb^Y>im2$Lkg(o4>i5k;-Da;J!jpkMBRb>2ueY+(CeCrg)+=vU$2UxK zlkzD*88(VP&I6oaC2C3Yvq#+cLI5`UJBVGcD?n*Z`rk_J|J4NlKZ6xR6di%s8x%ol79dg)Hpjn{9p0 zb##^tRWFE1EiY-nvSYWE*%DdH(S5wn-69=+f$B zd<&iW-$S2&+^nSEsQq0R5YSi%)utR$PtR*pFxYJ(0H$7&hw`t%$jbJIyjh!kdFGo{+hy(W99wd zpz&c-jU-(bdYghlh`ES^y`KdZV9!NH^J^J5|S%}5}52oNwSj&Qd*Ox|9v$ec_J z{<+B029}z3{ad$1KzJ#zHM+;3NyVJR4iE@zTTYCYh0Oe0vPDYa1nYB`F%_n;XJ637 z2sDRC>s(h}*dH5AA3FDrN`#zbPA1v-P=7Od-2FNhmn}ChucQR8?3GpY80(!0`Lb{` z2Gq8v;UV=w?@YE`Oc~@0Uw#v z>^Ir&Z=YArvdP&`V=lQGz3uF6ld~o`F=ZpSYT#9XQ9ctJ0k4uhID4D;<^M96vcY2BI$s=^Cd`1rPmbJQ8 zjwe7)zxvveN4Z|zJas@WJ=9+S)e>z4qam#{s^^U;{(26C`+DgpYEY@kq_;aUPC^F8 z$`u{+xd!)k)F_|>D~eG6SIVZmyWbRHuHO=>%XbJ)?rrgmJw008HFvHaURTdFIQcVv*HkI+2bXr8)%n3@xM$Lc zWWM)%yON0x3+CL$lU%4h;Qh&>hv{C=TmmY)cCl4U3Avr}W%7xP^^1GoxE~qds8;DM z!=o>`)9qf(f0tdG%&?8Cvz{}HN>vOK^fSy4D;FKXJNY`z=h8bCzA5!c0tiaAmtPEP zZG;WhW_T(}ahk3|urIw%&bJQzEMjtWb6m2wR!1*$PStjEBPE-W<=CP1sV#7h0ZC*w3 zM5~2B7c}z+O#Qhbl{{zucphVd7WjQv80T{uQ*q}>#74plKY~OxpTqO?3Ci@FFse(b z=FG*SHNZ?$((wH2GG9Ysyq-wC!+)kpVzarZGUbq};j-tHX7&6QRrY+oO{7-H(@sqclj|9__uDyYq3Dw4MpxwLAJSi4!1NWC-=)zq#`_Ny_Ku>O9$| z5V-AaBX>WvMy|Tp=}%6x(LBS!=y9^ ze3-??U-n0*dWP+YRLcvzonliak^CVfY3nuP*3Qxiphbl~rco693_4EsJY6mn9+7@} zwlGnjCzq@`f^7Iddw>p;xbU8EavnvqiNc&})`V*;&U2P?f859af!XEXxIzE%sZt}*9;I#LqASyKd=9C3 zg^Bz=iDvuA)_79J`b)Y?y$aZWa_NMv^zt%8&_fLNGH}Z*L;&iy_a0o+;#vj%X zi}*D6L8z}Ck+h~PuhP;@BDlhJRO3AcaXUbgi@XN0SB20IdHvi0k2DMY;Fkg zBSxRE8g|TGRo4%pbGg+QZYh7I_~q~XNK<>wVMX%V@3EY|;NWIfIx`VcBdUR_R}cTZ zWWEyn-vP^gF;s|;ZP6c}6Zz}|#yn+~doB<_ZISiK3A~I9D*vyJr~O0Q{T2-zy(j0+ z5biKuz+XYo^>Kylp#Wio$IIF@STc`_VOr}8p0Lih zVGa$LoOY-;w~D{Jxtj?h1fqbu?xcZS9nEPWFNve>d&^z~9^n?5_hv}Mo%q6jz<$hN z>#^J%t@`RZx^#~VU=QtNO}|B=GX|180O3%*!2f^RQS}`VwF}N8lHoOUk;|wmp0J-d zOzxl4S<{u3-=7tvVX_?eC#bUB`@R@TBXoQ2oz#;lR{hpJ^+0ZU{_K2lH9RSV%})cZ z5TmK-0)LY$k61%gXWcFx{re0{?OPA@&;K+S=CrAXXYVO)+P?Q0tllz~XyXEicM7ZP zH*Q$nO13*Hm?s-Rj_$>6BuN>g+0#BO<)tSFWcBvTgxAX#UB@xa6m1+~+O&-=9W5Qp zdRID>lB?Z`%mQ=AO9*^Dew4wgY9Y35>^SCDM{DHgQZP9OL>uNU-)+1H@Sw)X@iLnD zgGKdGUK%B^K=6O9k$<2tHQ#c#ZUa({p#9&b~S2tROIwmZ@kNELB@E;Ac#qkLTOm;3TT zmhCQmg%9rwTI;Oz^~1}}hn*D0UrHL$@Y#JOH@t@~vzqhe5%kj@7;$3OJ7raOrwFR4 zocM4~y0&7*xzN=aA<0lkPyhH7p)fz!)EE>n|9t67J7pwEK0QDbzRb`i^&^D2JBizrsJX)M)B#m64gH<4o?G3X z7o*c=KX@j$JSacMx1;1bwJUGVjlF#N?)|k9E%FxsDY`yW>II&^F1>nr>1#szx|h{7 zC!~KwKU)1^mMNs*XbCp)O{=I39x|G*A-}36VxdzOO-?g|HysA9;`aWchIK&n%i7H% zBkdNS-6z4pKtsK+KFh{E>sYty)te$2u1FszF36~Ars5|JethG{pt(&INQ zJb33PEbS2&RW5DVkNHxJrJrl(T^inkr?J{lvmef{=I*YPzK8a~K87r_aZ+o`FGw&y zPe%%0tPU%EXY3>~_I=Lx#e|}JZsF|9B3D=lQ-;Fpn4kV$l;@lWx^usepKpE{@qyRi zTS4YqvuQGI;{6kHT?!_u@T~G6$>}iYIrlwCeJ=NfKy#;j%Cd^`Pe>~uJ4@-UcjS#~ zO4|D<@nv9qsfej#k8H$ifFMgf8))2#LMnkREZH8AFjp#6th^5)EvVlJZTob^m_%C&wMK3nVeDo+a%zVl0n&(>syY~ca)YBla&Ou-(p zl9J01chkp8@ZYuC@bvx@_?R60%ujkRpYJ6<5IA(4z4>mmdGKTA9i@jGFjwJqyMg&f zc?Fuf#not7y7W%&eam`2wBz$oM)PTb41h{TP63S6COU{LJ9HXH#50K~I8e0jL{GCJ zHOv_N=0@+q&Ck;3V|a~f>1qIK>=h;mp8cJxkSoh+n#EE@sEccR&50nM7qDCiWjzdb&&oiZY zkjMMKTxG=dT>K~Wip-S5lmOz_6Wr#_X@n6maR9!7B&2V&q|kVFcHgXz{TlGbu_PUN zl!t4zQ&qLP0@i4+x2Hdh@rP-H@#EGqy($2yAp2C2U9)U^WG@hMCq$l>-u@ov336=r zgH(HQ2jJ(Kx)WM0gr`FYX{cxd&@3c}yt2+wh^ioc_AlBP=|ZF%bNw|i3+D#>74SJZ zd}(xToT~aUml)cmB79I5&|9!S69+DT>tc#gXo4!v!{ctcHnUe488X)KFmm%QRI2si&vL8)PB@qh2_cS`w>A_bIAqYfV~!!M z3hsElaxpj5ogMd1=pMp&>W96y{#b|sL$J4RMocWs!UJ~(qZZfWF2?D|thSG>MBX=D zCgdxM2svtVx8X&#a19Cl(jk-VZWZTqnr(pz%Q8y#*!TS9W_WQ+LLIi>u#u%MrrSOK zP|jS=c5Wj&SEB&tN8~|wB)aBzL3aFPEr+@6WJ==Hb#eDSG+v!H@c=Px^D#kDj zy$3CZQu*#5u=n6en}k`P@`foG2%ZPLk{sehFK)+*?M{;xPwhba^6YGgx5>!m{tJXn zS=q9znUnIY&w{U1J^bHjiZA|JlBHfI=M7y?xH4A`+{ho}xOF2(aP<&BdK9fT(vpN0|%h+>#fL5jTg z4_dk?Z-Z(0DW5&yx$~3W+`ehx?t%)RVq$`STWm+6?PLGO(u~PT2F&5G10=ucjThW_ zT#Yw9bbjMx>`Nxs)A97rHZ4!|8BK+a1Miy232|C8vc+fm+kV;AoGq?1l)|`LtsmyX z0e5JHp|VEM#pgcBRq{+P&QyMhwlSh}(w#aQ5>FAUx@Y^HIu^hao(~-4n&ql}>Jf58 zq@19apvNIDs&t=Z2>qQ-X5}0EAbF4QBl>ket$2X;#lMHN-eO&=cPQh5$x?TF9QR_= zSA0HB-+s1K4Y@*nqtQrRd6(<;D;lJbi|rQCf42q`c#3cB7q#85(r_qTg4w}$uV&6A zwg|^*N`x@dey9IV4@8yro6~rSjB(-GP8=G1Ukvbc9Lqk#kn)fqvMe{6a3GO-P@n$% z!S|_~%FX7-ES28OZbL-(d3LSGSbeQ>^?iIzOXZ)VJc-6Tr%V9(~L?!|i4%6~6Q@+`bi-g$DPV%WN|G}M%)SCIQ? zqR;YaVPjV=S7BWZo<-YmST0?N+$K%FZ5jR5?)mItJvJMghI&C>pv)>(0*w<<+1GL) zWwZf??nxM!njcmnI)V3$=mB;lI{R3a{;Ge1oncR?h|r331`MFr(Woo%(xnnS9vd=z zM_dH^Q2WQDgA(0WD)qJMW7fNVyRT<|)$ygYh0rs2c=*j-Zb!IE;?7h(ku+xcOF83nR9F9ThVx8)?|UDt9tZ3pG{LwABem1?$R$4+&JLS7Yi9 z{eKu+?6f{@JW4Ilt2b_Pe?o!-um2tPobHQgQ!JllPoBsjjifx?E3xk`II)Cgg303y zd+0xM2-RqE1RFjiw1|Fjo?EL;Jq?@5t>3SmIJOwg)txWtHBRcPH^`qAdF%LlrtW%De|$*uQ^|277q3H4g=TJB zFh1e2W})CcNmKq1JB)Z?V2#`Dh;Uljenb>cmmN_kYWWVDVNQ9a?* z-9W6op5c^qcdXhNqnd}(VqbMrm*Fz{Kv^7OSL8!~`)URx|IXY69 z&tO(DY?V8H@&&!ynA7pL(qUD51@K?B>K#ukZL}_%G(+?!FJ=R{mdB$Tg~K7Jzkvr4 zZV&dKEp!x0^?dFlPhpe2J%bM#$Ax9}zHK)@*Lsjj_Y9%9?P4o@#sIgnHzU)GEt3>u zWhiqZi`&Q#*=p*Ik^I=K78H}&`%}iE&1*D}Ci&zIqf9ObfzG%|O}7k55|atXzOpbE z=4=M~E9~z?EKg#S_QdDin;bKOHA(>9dmOy)jmlnp>_ucW{07k8Oi9{b{;?k8hYy}9 zF5X@^e4pj0*M?IHW?QL@r=iI+(O;gsCJI$qyw zhWWr7&QVvfr42p+50*?I4HD^|uTH<~zr1AHjwuTX5ns}VYvZod4v3`jtw{`f=A0fC zO5GHPw9(Uh;(T_8M8qWl#DbnyCu3WF(vy%Rcry4NID224bVvCAK2bE*_fH)=j`r#j zYs~6-`y5L+4Q!+hzr???x(4POBGKiEXi*neTDwEznuE$IM_`)xj`H1f0|$XUM=u$%d(hc7P4;M2KqsG`NHMzFdMp*|H~*)ngc?`&%N$VL-3g1G zP>Z^RA4B}#A4ecSY(YOHekST|4Lvx%!h}4g*PQ1V)&X2LR&~|9EMLDy^Z;zE+jf*W zhDY;qO`(A_MMfjvv=mJI`b|>1<>DA0Lr{Cj5}$~_!`FiDmVeDIi>O+eiukBm0K$LXg3{C=sLN$y|uzR)2ckhUHo#Tj;pw@Ti_hJieES>5gl zSRw4Di4|(t^P5L+#wLL#%p7rM4P*m;U~qixs7HW99U$4R2c>WCRIu2-o%v0GiH}b~ z!5(;Rd0IGKPeBn6v-4;)A9Hn>?eJ!v?+jogE)YklHTLK_`r&u+uaf`%6dRCRptC$1 z?W1u+_Zs&XDK8U-(esiAT6e6vrhgzgPAP{J;E7k_?t6Xv0O}_=>Gh)vzqBeqgr9cQ zPN&(v$@$gZDk=G=&(EIsJuM9}A@0VAyMQt9q`buRzfeL%*)UJP(HHA*`L_KiOC92C zbqct;jjrg%xmqVrqyVT}M3ZzCVg&zcU-&`Tz)xe%YGna9@hnO3{PxMy- z0?xf>_TSP8AI061jhbm)T~YGyeb@Eid1`W=UJHxw5Ss`aDqcv8d}l3Y0sjfFH#QTO z1uJ#0-YxrQJ$y6CabL@>L2(#zri!x z>K{uj#(pOL*^O);XM`$-B0_$-OH zp1&zv)OYfxv~Zz5&pgL$HVxz>fO6U?v*jA`Wx)uV3yb9{GHNr%|7r%@bLSg=-e~h?JrGC~G3&3lX5vAgcLsh+;!=A3)Mm zAc?%#vS8`q0z#IEI~cx6Nw+3{m7xllo2awAEZDwW@P1dPUZ+mKK)49VuqgoeHs#3S z9d)E0tUmS%@Y=haJ8j%W|6Bgkl4o=wJlt@h|4-^CdB%}P0LGRqL_dA|)ns&qG8GIU z8qiHczHs$}HUWKrN&l;prKGj;EZT0YNUqedFI<5Athc{uJ*P29>XwZ27UX_soYDE3 z{LE(4Rl9x?Z|5;&?#?=Zo12jrTe9&*{&?um2;jf{b~3d>p7BcYEN)L|j&p@`<%Uh} zA?q{p-*|Ukze@Zr{)7rPQkUngg|Ew7ZRFHJQlR876$4@cTn8ahwQo*D$Juym>P9nX zh%rP>#ei8A-W&a^#xwL2SxWM&IF6}myUnKjJ~z|FgixO)su?^}vL2RrKTL9bRti<$ zG1=mv5BNcqxc|Ot&o-e$x*_7R?&AxSqwcl153w!UpAlrXcfjXzHueX#Bpn@5IUPCp zT1=5H$a*M!JhwsMyYTrz-0vR|vjWaXPrX^bDzFo)qwlBi+AFq7mmb_upe9JCxZ;QM zl%||GhbpA=8nOHAxur{ZmAPrVTqvEnG|LL^97?6aOiKcv{j)Z0Z#< z+J%jjQ}1f9?MbXAf3evEqcn&#*quc8msXhvg<+m8f!ab_% zyOiC<*LUh#`_a$A(#`?oOYPGwa<>|MMQ0IIu}T}mu`0C&7+HA%(k1p~AZ4=Ih8QUk zh)GGIqiKg~!k$p<)T+Xwt6rUVM42~F?aZ|S8yr8<08M*c+bDHL%;C zTWD-?UdMh$^pTbX_cH^T6bszfguI(IR;?j(O{qT#-a);uu4#ZtNuz7!RD8#4?KovL z1XeV+=pT+Rd9!a%>+N;4-`}>mAk1O(zP&7>+M~+`>WL%wxI8`C-6sCBK-t3eIV4-| zU1$dNwHIv{VJ9#*UFEtulMn!3ybRqHq7OD9?52*zByM5&EJaqRz(vCj5M0Y%2A$vs z3Fc3fSqyHBo+@R(hv0lB212geP8nu>X#X=j?XgZ0hQD4a_`vL_kAv{V^0s*BE65vs zMbOkaVe2?e*@;GRLtlreS6V2)s1zk#KVwMKHR~3AWxJ}Yy z=pcct<B1Eltr;VyR$T*$q`p#ud&0Gtnt6Wy$?(i0Kh@J4I-O0Qax4Ff@P}&o# zk?B077G*pw%XwS4-Rjs6$M>Fb?eQ>E35!v(pz8b~644U9q@7;3LiFXUjkL9&hflw8JX0pw^cG`T-rq2 z5ybE}uBDT6Yaw#7a2P;O0j}0K`tw&ohm<`bt&*YM)R#lvje5#;$zC!dA4NU^xc1-` zg5v(8ap0hClK8c_Kp`H`86meOhA?va>t*!IA1fx5G&W>vAk}PB9%OGs{KQ=PdTxZR z?8b4pFnF?eGr&O;^5F%4Wszwg{tNcHpFDXYH3Il2YsYk6Coi!ChXJV1v z`Hd6WzSl3`$3Bgz++Rc~oUeN0TW23HI&M{wU11b2ECBt?+5ADJts%X=9^* z04BD~@A_vxpKeI(9?U&!z;)xi*14$S*W85o?&{7W>s$@&_ZJpgT8dN*(_3E=npGCg z*9Bj~eY`o{VFxu!xVd5lJC)TB`4z2WC-})-0(Ws{?+}S0cebiaHyIe@3uPgiGv4l} z%hQc6S=-8!Dct``D>>0m#tezcQ(5KIzWRgR^Qc8rwpFtZQdvI=fsm`zbNR#Ykro~| zi@l+_?7fNO@=Z&PA|3Op0SBhP?6F$>r8rhykoz}kSGi{zK04lMHZP_ls!?4MAw@8+ z+A|OPqTyn|ucId_fmsZHlo{lQBC>YLf9{GvCx2yz0B(pV(QEpj;@nzhnvL~w1wR0a z*T-=iLc_#=kY8LWcZnuAfhNj-@ni5_mXR2(v1jg}B4%V3>R-IL*DiW|(r%pW^aba(~kB>jre$Tl%?7Gew z%zaZR+W0AwSuXJ&E|{2Gp;t8fWtA(j-i;{a-vS`Nh$X3w>+L#X1T96~;>RyZkxurK?l{vXMkh&kx(Ud*2}_Xsgpb z$b|Zze(B1o0D6yNe4Mb*`SCGrozdC1c?Idgcfj4XH?GyU=ljlm5X^PtfNT10)5X6~ zGe#;!_?6Eyr#)BEV69JL0}wqO49Or?#4hIr zN~*6zG6)o@V$oQ)K|jP*xIl2rXNW{@veY~1@uA4efdo-jW zc0eHA@|D5XN^xA94OU6}1APBO$o(tpv&lWg+`qsav_y7L413JA_=|8bSlbQ6U}v%&t1)SzA=f2m(t zFH|3~`a@}2PqW8kpmc`xZz1CN+?2m7KT{)C&m5j!RmDwmCdB5cGEiGXz)H|5s-cd% ze}+t_SCokLAT!cQ(>2nOUXI>ii`J0Uxv^1!4|FjtO1Gw@WH_zr)y-B;F4}}637db1 zJSYAVoPC(tlO(G5akY2dGkZ6^E@lq-e3s+F=8oncZQ%BBbDM1!Af_bXx(DON=$%YN zNCXMkVn`XzZhIq6r-Q}w0gf@_8qj1~OcHa!d82!UJagrKXd^fERe0|m#Mh)Y?$~DJ zvXHBHpYS5L3VfrZqR_PRuu zgRrV=;d_l7bh_5}p(pKV0`}Kb*HbB18*Ptm!~9eqo*RfRN;m0@ymVgG)99($0<4Ng zN&x>A>TZEpOJAiDrIKFCS-o6O)zhpURK%6d`Sd}+b@iq5>1NZRdat&GE#{!$>q%~F z^uO)`tHbdc$rW#`6DCAc7;rFdO`r-%v2}KgQ-U3;sk@y2)~2$I#sj0}KP_8GYa1x%Rcsx%QsVKmI3P&J6c+ueH9b{qI_V zBWXkbQtvS2MNY6(7yxrnGo^r6O*S%G@}y71Umw{U_pk{UT7w0>;~ado0DvcctqezM zKm-Wk6I0!!iGFOQKb1yXz7gNQT_$`atPuO{g5gB{Lngtm%y-81fWHUXnql;RD4;e&Kwm*`Q!MzI^b7PaXD@J#o14f5)NnIO zUx5TT(LaS>;LkAg>c{GFj8;Kazhy;XcHbl0R;A9f;S=UyA{ z5FL}%^4mEpoCjTn8HM@rle-T$+J*V4wk-6>-!fN8Z~mw3*QMX3FQaI}@-SE;_%bi{ z3m>^;j`EFeF2l*9#6}31vAfq8s&_%sfBuou(jObkX;8>th^d>Pd*)5c^?AQ$@ml~I z*Bc~AK>z)>`=Z7fCJu2OzY?5M-q`qF)GvPiQa)C=dX&I)q` zFM}emHW7)0xFgd%V^B23o42=-%LVBRsCvXMb4I+pO)l~iBrc1^{JoGRn}Q<$sOB)_9#mg8^+GBpklWsD#Yz9oU4sK7$-N13w$Fk>S8SJ8JUc-|o zZ!Q7|C$pGq3+MvHj2b9JnckX>-tdL%;7(!Z^PzyjF3022E zTL=MfU{nqg32~65dVoMmVZ3Q2QR7f@^LD>Tq&~xeuq?qZ2{ymhEuuLAT$OTE-taR! z{>AJm{PS>9{{c=8qUZHIBv1Y|^soTkJW(IB zkk_e@7sYxEgUibzF)ET{w^L93(KSQf3tehxf<3bWmes39X7_C%7)Y} z%Y4N_(8iCvK&AP$T!a)8uR?|-g^W(zq?Go<2zY+|R(rdaz2o+*#6s;o^N57^r~{RP z{^@(f>eJ6l*N49cnzFjL%%#SavKTzC3DJY5IExK*>O#{rgmN9_IVLoL_ z0DM2%Q=9GrQuhU=jW+<(+zq)uF<(!-m$$t$w>t)7`9y{faNc;j*36fpj;7gf_dNOkwemS!=Ms3>V=M{bK1G_f_}uOK^K^}?c8voSHFy!yw% z9Z@S8#~v{)e6>$HZtY^3s2nioYI%cx5lJ@(l#(HCITTRnvL*RU?vj*(G^%9 zflnJ5g$N~}^G8a&rD`fJT;?+3fH;V9LSNbLg-T?cxdi)XBbU<{SIH$6S~l?31ZmB%`ldW zadpcB@17q^O2C(*+MZ3RE+BcshvngaPvpq&sKm+88GsCErL5HmEU!|AtX-oAm0cpy z;$C|r>Eak^p*R;zAr9-<18)}cnaFTXkZaDoUtOxXX>k$c-6$*QCNNZak&m8AO6vT?EM}FCrYmc$xUqf>M}LwiQR+{ME81VN-s|FX-kguL2*Vb$UMh^P6~ ztF7bmB;QOr?dr|%+01L<78JIdp_jalg4G8L*;x$&1Oa^hyO7$Gt*fTrS4-}EXl+fc zyEi`>A$<3Pj>nI#9PjdjIAqm}L3TP5EZimMLrS;9*6&cDavf-_p2nNwKrvY zN*w!t-pB-ScLU8@9RF4agPA}E zQrI4y5kC!M-pSP0T+OXjQ@^Y`i1m>s>gHBG<*3J@pepXRo5wwUaI18^7R|l!aT_o& z1^1lxks$fs%FR33Xo1CZW;0yTQUN0PaY39}5%GF6#@_0in(h_0l+6tnToOk@gE-Ym z89x3xIKQ=j6DBfYR>GKu1FA5L<(dUQfpNZA5_2Bg2O1D^x5Sj7ZO6IVB`sr96^9&p zMrFM-kbT<|+IV~8x7RS$iR)P1$m~+U*ZeIijn;rjtt$tdL%;fEJO*8@uYEdXzx+S} zeTKe{jEgGEP|M8Vj=c8R^Uu5}hP}L+y$FKH(4M|xeO~^zSLWOcBl!Lq)l>i0i|a>* z4&K8zazx7df=7^VZdc1b5#2rlx~4)ReGNjQF3pEGYY)h|Id_fd@#|DrXO_Mbu4>e6 znW&+r4Dm8qBuNf7V#};Yv>y=h4U6)BiS1aO4I*c;_Rb9+?z0 zw{1oJN+r}VLiBv`WBGv9tJ0AvRGpKncFl-BhU^{h)4?^)6I1y5ChS}GVaW8tr<&}j zq7Uc^UbFVC9R9hygV!(5*P59lKeU0jWhu4%OtSTj@r6LkQEe{DFPVTAV+IzvH(hoW+|)sQW&CO$L&5Ge&S*@Xg;{U<1lM=15^)|8BPo{(pO7&(t*1 zKC#%bG(M9O6(YGTJ7YR;<}A7R&_ppuLG%o}M|LF0PB`we*)v2k-QaLpwms2^{Viiy z=N_hf@3Jp`0j=S{CnnS|`Dj88;MyKh_f%AOF+Yd`z|Y z5_82`;&hWJ4mbGMbLeCGLY%=*jyMAoL_JFk1aMY#N{IkWZi=t@8%7jsPp^`ky9k9w z?XqSh%cVTzz~!;YqdCA;*Vn6Yv&Vg_XOMj-E{DKSZ^hbUqClnfke^K(Qnh)5kLQ4e zcz%)EkoSdPEjo}d48{c8jGp0rVV7-}5mh$Frp~8Ts-=Wo94GU&w5i=L{fe}FBg?Si zorbsVl0SN^K?cJCQO-z?md4}eYX27TTFQt04-FQ1BWkBCdrR}XemsNotsi%Qd-_z4 z8YHj7&pv(N@Tw6Sh9(Gn@_Vu9^yzWofg`W1He>jA1}}%JnDPM|{pstfxo6gXKgLLV zxGgEglVX;y-8cr{WM;mDp?pdhq}K-%C&?L2G(reqGh#P)#)NLJLeK*<%qSMc1+t9x zjg%_d*+*UG#|8Ou-`d*^SGq^h8dKbBcD@isrG1rulo9&J+)`&=G^9@~qt)MyRv+qQ zGi}>Gv#ELf!$U<3OTT(_ZuTKn&VYGIrhG>pcxKu}7qhF5D3#|OkHYhB*4&kt8Gl|M z`!!-(G8t`%PGHWr3h_Bu5N3~XMWm~R!uRhr9f1qdV{OqIFWXbCS@s$=+e~sT$MQwv z2E@MGvSxFEms-#dup<2HC9Vq@&Yf}>Hz9DAdK-T5n9BusuDbIxlEkr>*4e)w?VW0se5BarI@*Iuk;UlK*^|To6ZUY*z;luRY-n?9_)IdR-nZ6S)3V+ za`3W#9TFQ7;P)!lNvsumr4z|WCdTHYymWcT2Ol|*AnVi#EyXc79dU|#9tOaiMP>n) zs_oSxSJ2=7rdA$F$>0;ij!P8NzqF=F76#k}*<}!O(&+P6V{ql0D7Piw(Up8z|=tH>}z2lIP-YSkzr=-`3?*EoPpMbDNVMbls0t#eY8dc6zB9OtKwN5^uP{ z`N*~)ja)n|U8K@cuS#vGC=WqWe{`NanW}60r+e)g6gghb&Ec!$BqZ>WEtX+M5>} zwS)WdzNnAN6raZ@W97UZ!yXHmCbz#azg9SoPi2PUgkKZyiRCaIl+TZCK}1P1IUS$1 z?(JWE=%N!(+(+^5J6qW5m1ZTk*24650RDWpP3qZZovi7THfWlH?R;$t`N+e^(`fTD zPvwLP*PmLM?|LNDS)=fEx1iVbiELUyW=G8z-;?DJC)BPFms>CrlQ`|G`)NDzrSVzc zL9W-RR5F&)K`O9P`1AJMd3ca7w|c=YR}521pIwc2WJ3*KD1VFJ!Q#_M-&4!meFfe9 z`gfbXY$yk)pXa#56t-9lcQykKf4-uISC84lHs53yb-rRDxn01dLXvW`06dQTD*Gqr zJ&7hIAUrijfX+@(y4D9ET>lEG{qNQbG0FGbzo_Q$UC43n6g=#ZHyVmbFNAmi)}L#i zzt{Wyi^Tf<)bF2PrZPPjZ(Xmq&qg{_;pvwmhpUB)8V$K0nDrj%J$hPleVe5CSqTt= z-$V70o4kuRvGRa;aK$tGSb8OF^m&_}f7NY()A#rh_DnT#;%)g1CU5^oCCWxAjl2|% zy>2nPwCKd%zhxjm!zGMyiR@-mLR-BnD_n{&2ncY>CW#Ap_WyZwpH$1Fq{u8gL1|+! z(_G$>J}B6icMcVaFHJ{0vp))QQlI# zk3NSEfUV&!=_sl3sk`PL+X>GFAhhDn4AFHXV=>U5n7|p}A21<}bX9#a(W!>dV))$f z`7Mc5A8cxA4GV`10Vi1=}&j@#gC%yay#YCW5(am(ad z^Z8gCkt`>hmGRa*=c=V4521yipcDX&o8GL2SyC8}ntR5csJ6lgQsBSs41ouNa}K{2 zOr577KRdKPlcoaj>JnG<{q0`q4pt}Iyr?RpuM=Oqao$J2c)G7ST~OlK6o!0Z5_}hR zK5PHOocRcV!MYWbEcA|%+>oo@(3u8V6t?AsI4^WL5Z8s}d!{Wr_)uXD((|~sEB|K5 z6cP8l9qBU%-`(6oubr)~(>Fa6Wsdb)te_c~iBd>BXM4YkS<8cK!vL7{#%^9?tFogU z^&sbRZvR3L*v?X}XPh*5v@vDNbBN>oQgxFJ$ue`Zlo$F=`IJd<8)fJcQ4@nllcR(h zK9aRx!|7&klG*IEnL&g@m|vV<-0l6T>$rXx0Ml6n;?aJH@&u-JRJ2HjH9{}ydZN4| zT5$e-*cCE=XjfOh6drVmeuzEaRYNhOV+enTwbo9BZbIOPIzww{=Uj?P-b)1t=@O1` zC)$@>t?V=p>&v433BY4-QS-^XC|d|n|C9Aq>M@U`ci*bt^IiNroI-$mA32EHA8w?eu{b#iT0~)48#I2OeMXJw?T{uQ;M)6s_q<>v zQ@rb-e=eblBm0XAYnq&3)`NTi*57G7PK1v{rnmUH-&WAWk5ACRSeG=jR?I&>snDeH z)^Zs+k3cdy#H3u)wgiV9vUs&Ee)en49yhOex1Fmc?Yly@+ed;RJj-UyU0&91*<)dH zR9RsL_;r+yB!81t@2`w4O3eVM(;tztXo>*1am9=vv9e%flEsKYNa7a&DL*c9c8uUs zbxy?Mm`Z4o&mGCk`2aH8QJLi|yZW({Lf0)|3?U242Ft=fUQ(Ro3`Gla&r8D34@c@G zOzQ6OVN^%li`gv}z5|ygekL~oRv5jyeSnZ&*O+bEk!&@?%zAMbQ_c4PjBJo}d8jdJ zsT-2@Lt8-bQV+5w8sH*sU$$h}MB~SJC3@nGGBlZ1B%~SPSk*YRCx}0OoVTJxg_VpM zZ1Ol@pkTt{MTO`3+{8$@bcyTQD)6hP*=XWj7;8nL+hTt#db@M$rBzRbNk+Mw#2?dB zIUcTBZcgm>A5G|J1W7V6x`$=+LC$P13=uIe4L`M&CpW*gj=1@`C42aiR`NHzX*_Bd z;^dt>v!Fn7_>LTp$Hn?SsI;-mdRtuo#+KQPb6HdC3ZIwgJ*&>>v?{FXgS>8q@8UV| z9-EMU+-k;)Zu1){%%iR|C_p^)2rt1Y8{S{(sl`d}RPe|4X9u~HfjhU2rX^0a3i zpGb}P3orc=!rv}yMp^4vi|Af!e_J&&bjM`c!P=*=y_xkwHTR39`Rt{4$hN8zkJ?Vd z*gN^9JFcyT*oN3ZxyN`PVzvV>u}g7>V4b_!#^`9jvi_J_gt5UoEH9G5@flq(sbZKz z3elQa&t|uL*LStcuB|0>8_ZqF-DUS{!G+w%c}o*4B|=$8Mo?SWnLK!-6vO;gOj z$Z4X(XYve6MB*%ZA@$yJKjO(0j6TzHLUE>LOVn3hk~fy~FpO7KIU55-HMv(ANX`oT zpQl+k>v>i?L-EP74E*Y@98Z~_ktIa+XRR;9I*GJ~RiTd%iHeX&qa!hMyZ~rK=1QJe zCI6>+<>*)GKrPpTe!Ud(YH0{tBK__V8S9@BV*ID&tMB8k!K;*0zIC&#zsY44P1{Eq zO}=-xE*6AHW*UETzREs^jQ8cmDKR}!=LXyhBhnB~mu+*z8C{|g-b5kzcvR$EPuf%V zt;#YRh~(E|yBVTFs?BZs#_#%4yW!=BZMFBb)tdGNv#O+PJxE#eRr6Ih^3#tv1-Jp& z80S+EW|t{z_-bb@7QvAi2hxv zIbb!UA3K5g*-7pF!krUbrn4|q12?x_6PX>*@v74G-FCMrVjo&Q(-uDX3n^1#*e4JA zWg+zUAp_6+>n2C>( zoW%}(=AryMR?8#JD^%z@sl@^yyRz_q>xgBRYnvUs7a%n=cI7k166pc&bO4lI^}YZZ zYb^lz^UMmr&NJN4I^O}e`|EeFZe6doovp#m(9`KYmN^Lk)1~qfpZkc`%jqrMG;@d6 z<&pQ_|GbQMIT=69TTl}E?eul!Q5$aa-*ER9uHk%rq!aa<>~A=dt?qH81fZ;^@UcT!Bg&jNK}C-i^ufe0~84>h6dNA;S*&8}78;KG6-> z*hf4vd<A)CH%{BZ6`?pWx7JfAkFdvTlv`r23ae2ahT<6YR% zrJ2IT3BWkV#?8K`M?Evw>F@RW>BU*>D^!ECu=ut61 z=B#fc?hlhxag~l$>2Sr2TlucBX1&k#-ez$U3$nqI7mth0Sh+bF#|hTFbnZdyRdTAk zUWt`jV)HBi53PchDj>WcoB)4zA1a>++Rz4cT-GhDI0oSRR$T%eWWi$lmn92TL}m0% z9}fCo)Q^CV&_P&+G|m<%%^JSBmvV%>##YV4UWC3@N9hfoY+i>a^Q$%Ji%pH5?#Fmf z<&ZkzOD3B@T`dp&&BraHkR~io!i-lt2?HESATDXcz36@$ZFV~%nZVW36Hh1Rg4Mp! zw-!6JGd>;@Q5BF&U7&p}%db(xBxSA*@$w;+hMIK8fnM70d}6b^rOFzH3n_p75SF*x zEYGh_i^#RTXknJ7s$|(ldxTD0Cs-MCy#bkr1VYoIP(>rm7?q6~1&)rKoYT|v?;X`K zV@LV6ZTHT$3=}}CR)nnV`ih%A0mCN)4=o!48qM=>(bz#Nk^2j41=2(3^Wr-=L#&0{ zey2TDA9`E?n|lZmJ7D5@$8Y31<6JD_hB}l6Fy>-<*1i$+M_K|=_iLtB#iI~|EvGv? zQ+39;v=kH()_%b)s2co*pqy3=H}IZ-)e*J-9uA8RYHzh^+17OoY+Y@!yN8Dic%dBD z>d&MmbirYFP*{-Wiru4s^>}i>BjcS(1Yt;fq~6 zHPXFz!3umHc&Qn)>x`ECCYp(31|on+A~wR;LcDHp@2=``H_eLO?CEH4cS$6;I}6u; zdQAH7VdBSxWI>QB`e@N4;C3x+TXg$c?RaKYC&GZi5g1;^xx?FIos$0<_neO+4hgaA zw#tTd;=cj^a@=+m=Az;Z;Q0=}TZC=5MtCE(plLJ2wBZ5-c2^Quu|A?lr}cg2Jh~1Q z^QHpd>0Lf_{~xp9R6oK$(CI7ddp zGbtqX@uDA^itJB(4K8aBIQji1Pc*vZO)c#K1Oie>rq!d?*<;dPnYZHCU|xF?2Y}SS z$uRyA$BZRWx{ERS`+#re^~cUDtF6qZ(raPe`0dzCes}f$LWL_?{vkI(y@D|gv*3*!>W&Q z)vUzQp>C9gT$)DtjRX}PSZR$2|0_`Cmv^i^ay);0zCs__={ElzJ!gTtT&^GVk!a8W zaOGr?B&p3pgP!P;KgSCfYFWPUnZQ5kj^}RWwg%Y(f;-GYy09WTa-w9(qS;K3-T}~J zT^Y88hY{B|gwc$YV$W$GMZ6(OpvQYf_d#&OJ}{D!l=Gp!q(OSEu!H;Fc=Y0H$A?0Z zz7()$M-;CqWC1~?UjU^1yW>BU;<Xev}`T!Kaf-0i~D{F>9qA6r>}x-DNDi=GsJj?@<| zxOZj#u8fmEt*}w*h`;c;Q?u%{oOQXAwvVHu`zh{8o8LKf0f$85Nj`fc&V&v@EWbWc zYTmPTTFU#=zJydxji(+&Y-&tT9H?=xdomLf1Oj#Iy0g87mZh!)#oW3vQsdi z33QJ2LH|~s|Kj_q@UVyxB(H3(5tMZ&JU0oh&BUh`vT-d+XZ#(#!S*aUXdjx+O#}~8 zOp4`btpA^;`w5l#-cb@TcA%qbppp|YDcF4X4VQuJCgD*pj1RC* z37=?a_;@V?4~XU%eu?<|+-n_b9Zs&eu*-G624K7iRsv+7_^%>5;p!i3;w3OAD{@y0J<7cmzeIB z_nphTT8P?2+dxhxEnq064x_%e^(IU5w5R@x4cJTtnz|=2CD8&T31|EmQONvqCmTTJ zNe&3DHJbOZS-?E&LE7urGdBVd%HTzi6J39i&w3i8LtDyEE6phbU>SV|z}onLZ z>h83S24vrt$7OnXo*u0&by$av;pCEj7UFc$U#{lv-&9{$Uq|VJ5v~aNF+F+40SH(s z$NzwIe|s)vC=Ea(s6k?$VFzqSpWS(?SnRV-%a_CjdPMj6OKMb2$j-N)7O4T$0gyjctOG$wLde@jwmYq21A z;Qf~S{5~Fl-dD%4;^9YU8i&%x_>pdORldwX3 zhpJcASlSKgcx{~yuFp}~5zr0_B-`;&{apgwhcoK^^IU8DX|p0Y{)R2#K5Z40s|~=E z?L{K-XS-FtqIp@jy_n(Zn^kFK-qh=%{#+A)z7el>$6fDos((;3Yo{Qf2nfoxk)Mdo zm-3yTxJ&XIsin4f(T_GAR)%jxg?9{ZJHBs~5Y_T3JW)r*5@)*dgShG@Ry;S;>m1nz zA>VB(PET5qz4^k6kqxA-tXR_DWB}Q#2jNw8!{Irnvg^^Ux_09Zu?zHduSlu{Egc5+ z>(#xp&!iuhd@*3HlOsyabnvge*0UBa*!`aG+FqG$F4Gr2I$2Vrr^)9vdMkB~J;YLB zWjfNH5h%Rp_X3jnBn5yQQb3Tw>xD<4$M=5dvjB++89;L;w(Z7Ozsr&^O#ov^7s*-o zxMJB%%$nR5&$)-m0vp{|Jpp*N~MTdE{N%PZC zS4?lDCn6>utp}om>EfO@i@PLz+?K|J%`3}o0)$A}-a6+cTY8YcOV|niema&Snk+qs zj#8yYZ?mqY&i991yKuwF2-pKB?`E#yLo{s7Om`ap9rc4q+?V=G0kJn;4&)Ed9s2p6is}nus1D&S&~OtDQnuuA ze0sJ&p?iwQtI`*n%tuhMj6A?$W$YO+Z}G2b(&C zD_!-9U=G5vKv)+?l`&yB87b(!iA8%3Y~93T9`)nU=7It@*}Mkxl?c!%+TOk&G=p9) z54}EalMpR^O^ri-Te_U_2FLqcY*Ca>gP{zy#!A6b;-YbGs0e0n?J)5(VeBCZMda7l zc`KvXAF;Jfyf9!2h}} z%AX8@W&Z#PQrc}0RcGRd|2>hP(?)laDFOjC;j9KvoEkOIN-(#oyjov^*OY8xdd}tnE*?PDFSePoOWCC zV3eZyms(}EwVP64 zm4P)xlgh_$)jmkU2O2cgNLs1?vMv`(hnuQ_SKqpe7Pki|tb?om%1^O{Pk^=Gy$c597@!tD(`;_N&yjzKkJ7`*| zG)bR(b@h9ME8}obh#1l3%USdr_6xP0`yP~yUF`x;x-ll|CDP{w);f7d;uG7Aso%w< z)-@Bh1ye+9&8a80ogCH)bI+=@*{TceHRf_2pe15F$JIrna^Rjft`3_P(}!6vp|c&N zRXBCt>*kJFRt~*O>oN6OJgHxGxpQjQGpp*f3>$(ZA)y3?pkrebg@4a@CFbfAJmX)E zFotKRxbUr8_Kw;+Ry7Q5b$+pvFJoxHs)H{bc{VYXv^C!C-4s_uso10ad#uEg2glto zLc{*9|G39uGs*y#wOB-o(|~zzHmGKp{+xbYf1F+8XLmh)mPKCxz^+OfG8TkK(^Z~f zZxH`x3e?OlkX8$mZm$lsL7iYW5CBj#qqhL07PJ?%;nC*{94i1ORqS*kVFnCa)=9a- zm_Ss0^lB{YpfSbVcscekm{<@>maPr6S6>llIMyp+lwO>Plc+qI0A1B1#<~?%n62jf zkF%>|j?5p~81IbCK3Qe@-(a0&KiW3@BEIN!J^pn4Wpzrqh`b)^{K3pp)`g<=NZY{v zXV^R@A4F<&VD=NU`u-2MTr-i1%45$JQ{G4hGNY}n74?cS-5h*|ZJGm}u~1mVuTM2U zA-&a4HYagjwq2>Z1XC#EQwLXsmH=6oxa9!sDrLZ+M*Mx%RN3-$!E$*9MJ4My`EH^< z#`JVy5qA$^!zT-pwL$ji??TINglZzJOsCeLkr=Rlpa5YW`i7$m7$bc50<2dwFlVSB z?(MMBZ6)eoU`Y6?xH`O6r*Z3rVSrE9^_~GW?EU07+zC^E8Tlz$4dzSfL6hg3$J_xR znrM6m8OZ#ZYmiXBh@!^%BK~h_W`PJ)SHcf zX^0F>aB(BfU|#QUHgS@Zb|IHYP`|c+k02XtbD&!z_ zo6t1a5}r@+Lb&U(2?45#P(tsA@Cws)-4r5T<`&JfqN~qN1o?dl5M4Tc-#L-S^D`~%+gq*Y~%c>`+s+i2Oy3F6~8vgeAyi&FCc#!_Rd%1%z z&t7Sj{GYvrr%rPjAHAol%wVc*FVw)yo_?$F@yC!Kygxy3lu?t?yak1NU+$~ksad3; zI}`-!4f;=`c(z9>3aQK*PRHHTZ~-)5?){b$7%$xnPA-5IX!$;YDU2yAslEu&92piD zoue2oxmc#0q@9k__uhd=SD0NtOB$LZY3_3x?8zpnST{^;6BMCzN?SwYv|GLY9#HR%w*&-`;Hgy1E$r`yDx8%`?a8s#W)zri;qF8G?B z+K7Y6H*xK7JFL|Qph3uBOIDbFbbEl5dTr)ItuP%6P(#PL>kCB2P;h5ZSUyu`hnd|7 zF}9G*Tm$dY_-30Z5vUE{&f^sd5!2bIC<%J`=9)p94mQ2s?fw03D{|lo7!?+QenH+# zCZ^nJVx!z2pMSqLJMY@!v7AXZGWQ5NP?}5n#cejQ&skpA9By+taZoUakH~IKJM-?} zewrzzCXW;u0f`%%%qvpTq)J904vwt?xVvRe){lUsq_*Q_$;Z>r0VVPo3_1>Lr<(@jT(Csc>{T3b-tnW$AA8H?S}=k0APwO(Pr)&6xH z;1L|rWlZ_LmSP^t?fU7wTju?JaCNwG%~e$8RsOPy(3OpKm_%rw-5D z_m&l+{Q0x^+;(u)@F1N5*Fv34dm(yD`Wy&ya;#s)@*%vS-4XabTzEW1N~W*He1%}N z_Z8hXw%R#uA8bnE+@T-$&lS1cce71>Mm%-YIhuqXA3@cgz7MlwZR}t%kE5PZRA)3= zZdfelLSz>)=Abua-@b0pnnSCdg+7nmis7H+kNGS_VW#@;Sr*)p)FGE zuKPa1pRAzfMg|@9kuvrb`pO9Ux<#wGW3%blYwd{$sW;J+fiST!|G%-eK`6=vX{ti$ z8L(%Qj{vVbT6ul|HZl!Nv|W%Y0&MQb?;iWXcADf@#DecQ;N=Hx54y>F0Ac}J(h3dn2{*6M^|Or%pWycP z8Sp!-!4DG+*0QdMG#(QyEVh(7iYi=Cz$h1RW+=rj)f5BQxGZw|3#8$9bK)l4e@d)T zSHt5>esE#hNGO6&A5j}m=jI)Kc51tI(Amd&S{|&kJb(I8>hnA&{LePS$%eh5aboaj zYp_#V@;3mvVO=A#Iq^-Ck4ff@agYW=4oWw5T-0Fv6)=8ZO8SK4w#I2NNkR?Ik3#D| z!(+gO`&#Wam8?^MaLdo8Jq!WDC(p0k%A7m6C5F`zK}$fpC-iuDsJ?;h<`{?)fV|-` zUg^z2Xedl7OuwSL8#(qEoqv0-8{bHF87XN`QoA^mq4|~1WhOy1p$KPJ*%z_MX4mn9~DS!k9QwS9Kq?CKJ6MZ3j z^AI_S2UFkAO`W+A?bdGl;^R*7%mBnyT3F&MW>PB2c27O|Y-Vq&zas=1*V+a`XBe30 zJgh$3zkNoj-CVwdRWn{=D zz73rM?IQs=1TW~`33e0Eenbkb|2Is8{JNBAx3B#`^>Dj~|93JbW`CyZz_3gbP8Uzj zc9wwRX4H2TL{V(=4{PG%>;UB?gZXu-vdjbjKek>xw-)zjzl_+%oQI;mCXuBf*HJd*i#L6?g+Af`i|pZl9)9d2P`UHNKRrX$;mo8^$Z%2{HX zuhPEwttx-B?p8;q>WZB}D;n|pzTiiDzL-Y`%~j{as@J!}s>qX4HpjnX5-_`Xryr7b z!SurI3BA5+b$T)8K?B%%w~P!@_}+8jDI?L)Me?=}R>p8tVvIiFyHZ($ZrYCa^~9bK zbJO+9L*D&H=|rho6CL3n@AEnRxrsLEIKmUkv6-V#NxpcNua{nq+Nea-vT>Mpooe5u z@wfX=8m`#fsDq1d6to|nBo39JK2JF}*wd`dX)akL_qrGuuM;iyY}7`v&Kwn<6Qy8d zV1zmL8`nSbIQ)b8R9z&;N5etXw9hfS4hLgphbjj9n)r;V9jY=TWVC`oQO;ouYQEc^ z1`&&01S&BmiE?@0o-fpm9EexJ2g5CuQaNpEU+)|Tx6S-6V9kv6vwd5^?`3l`4tb(D zGT%#) zdTWJl-FG!VRcQh75I>94q0tr!YK5yTrn;jxSpH&oyuNP&e(m(SDhg|$Djmk(gsqm? z$1w6XXm<2RPP2fi6nk~3C6%Au-|{T>2GMT@=IyhUQeckJ9r>|3wrEL$8e6XWY(f%3 zKt$vJQ=_H0t?8DJvWtvEi^IkqZ>~1m?-NTCvIt4hHi=TjRR2&_-ME42#v^#Q_=4)q zvd&JDNyKbfd5n3m*;$3lvZhlf$;^uoczUq^zzsjGGhX-bN4+Q~bo(IG(e)IY*CR`i z*6FGaDbi%)qoy)Vr5%@!$?6$pFTP!)J-UD=Nt~Gm;@E|d)$$^F19#SYkK8QJ(X2Yl z^N!uo!R&Jw+L{)k95KCYJmw(7q|1gd3Oe`G-LBkR-=i^|&v&J(I)_aAJH_ z!A_z7>(I-6^E&;4tNlo&*yqNV?k;<4nQBh#jsT*epQ+jc#W$C0i7!~q<$4MJWxn3- z$HM*3;X}tlT|>k|5{tJe6GSQ^{D+@Mo};fY;kcITU&i?QAHI#3*PV)nR6NIgm*8nB=J z6_}Xy+Hx^tJ~@7<&Gt&gklxaX&b!5k`Mv1y#Iv#>HjS8{7MuG{Ew6bO&0k%ZtZeF3 zwVEbQ^p@?uR)=bRf9?yd+crpe3qXq#5j{GoM<<6<0q$LILT246ci&B9yAMor8q{Gx z>Gp&soo;Xam%?ik-j)Yas_}~pyg|8kCA|2NkAA-ADXNtzKx7ZW!xxEN7#*Xm{-#~E zd5k|p*df$UD6}Yf{1I6_>9x6Op+O(%=Nw|*)OrMdtGW z1+LW>P90ELKk+9$;#Lb)#{19 z46r82b43)X6^N(b0O80BMlnsao&u)k1u^M@9>I2o>+>dFx3IKjC!aRlSP9{~ohw%4 zrexfuhGpdFdr{I`?mXkHIPn{auZ|zr$@`wh^2qpxUlKS_cantP5^jo39c#5|A|#hh z08FZ2&%Gwil^{RQ>!Y)!{Sirs&Gp1t!KRe3EMT*4`+R&eZ$OEv!suO7w+F81Dz4gH z`-hw3vz5&@{BA@KAX}i+20^?7b8>8JKx_$qIuxlro^2&j;UAOu$|dv^fJ?+LK6)i- z4lKc8U;V{$71j-ky#81lO!@3LB4qDzlS*UUYpTz(v{7|!EFpr`*@U4fVSpf+j{rP= zWt-%29!zq^XQEnH0unamd5QXxp^*zuT|Ew5&$bC$70nQA5N1dYb z8vx$lYLugU-jWG8qA%-D zAe|xj9rc}nwSC0lcO+D+?=ImL6lNTqI+Ah)pHykmO^qfS3*9x`cTGZDaC}Lz?m`gQ zaAK@ie_zyMZWp6tGTDEs*(OqzL)O>h;v~cw^{s11?Xz*mGLTHGmMoiwJpLAg@-e?h zy&`jE!Pp+te*98sVL<&t$SOTF%N*e&ip`xn#KyO$S4 zPypun;nu=XUC6LedXunchsXc`f3OBgvfwS?f%wr~J$&8qW!25#9$#_xmc`IOY7&pW zd-a~Rwt5060fb=UWBWk7%{VMZ5JZK}jb)61eDvuduW&S7YHvhU_*WOl+eCs?PuLVp zsH$t~0I+iN)_TEhD0JI!o zdH2{og7REr`IcC;G;u+&{f*?`77i>!NF&B3QIA)jTo~6ydub-O zi$_42#CgR^pro>Jd@L#gF6#CshvV9o1+08;q){v1%tLVP%npeSwjBA|u(ZA;W~iLMZqzkVw)v1VQ#(8ywpJnfAm$PE-R) z*I!%jhfFHQ6&qgzFuM68ZAd0i{LGQEHmedx^uf zC3}pt1J6E=Xl0VL51GZPBsAGJ5&U7VFh^2zzk<^fB$n{JkL4%!&_9i#+h%db+1>Q zaiL83adck*=rU8FOux4`xtEG;B5$2@|Kdf_h}v)0XC05+WWnVe>f@c**R^q6Qzx9> z_9%`x$hVOpwF8sN0bCCsgQQoE+gF2NGHzMBvXfzL%?k|kzP4w-h_T+`0D=cxf9&v% zn?|iscJV~qO9!D@X^j2MRY9-w1vx6a0zwtMVo&KeItFuuVwc1T=!nE9-MK?iH;ALX zHpp-Lh{cCZ_zc6o;%6sBSgs`7v?ElS-i7CHcP{$-b^i;ws`oiKtZABFIce#E(=Ugh zTi~QpIpRFh1exq7{L=5X52{)piV+hd*;G#Ba~sn9b1k`blX|pjyE!{h^@DqBB{jr_ zMAl0iKyL81Dc4<**a)|k#Fq{)jl@lKV&@*wEBZKbMygY(4Iowp%U=s*;@Nxrb@j<7 zo+v=5V%N6_RarXQDFK50ZR` zv{p9+`NL~a`?g{ZYjiqRaxX~T-T**Ep~3ajNI9&o$U$qlD7cF`Qh_`+(`LOax=9d) zJSqm08xS$bd&e1f+HH)X_hg#|2{MGSma_}8-pJ*IcSc1}k?}oiN2cFMz^s9%API+xEbmVq6Ti$$|Af7dVCKo~OjI_sT_qB;t|Z zm)1`lvfutsaK;@iMKhm>kU1m`%r&Dg_K|CJpk?ysPm7OPsPw z0(9>xQOCV&-?~@{DaY-U>v$F!*YtkZe6n%e-p;2_6}RYwxTtOc zk&QuTkGB+uzA;)j3@xR$a0^)j*}!b|s9VI@-f)1tJONuK+5$_xien0zu0JTQhSSDy zN>q5w`}f5s_D?&Y_I>>-hrw3t!vs)Vt+1Wb-1{URtL|K?Zz~(HM2Ime)_Btg>Ism3 z8=Fh7DA78Tevo#MMy&W1r0!a|{b5uw$w;)t)%)p!N^6R*5S6`Mea-3=PSO`%=8rE4 zz8A(9#v@;e`TL%nboEehGkuW#6MFB{NGt&QuZO0UJVr_ab$+G6M$Qfh`PI^MNk31+ z>0@-+ypO+10pkic4t?sbvHx>M#>xOdn=70TfPF1fYUJD@@ZSWvVRnRgkSvLKvZgwV zZMgRl`yO}TzLWI0Uf6J9(&7a?Aa^imb~<@^WolRUY50Byvn1?Y-7kBy#g+B$SQ;t& z(CfubkCbI%l1~ixq!>wdR6!WEnH4m0{Ak?6w- zr+sex%eU*4>>ec*0AF4uGMQW;>~lORJFB+r$j?j_nUs~=+Ozo~hmcRO`%q?*>G&hp z{P8qZkKp=afFoP77JsoNxw44h#>{s!$A2|bTEdi`+{Y%2w3a@Z2SIoD6ZO05A{@wI z0~>sa!v-V`g`kQtcOd;)|CnJ#syzVhp&uJU{T!WzIGNvaj1kY301zCP$iyXQ)LCj; z7rsb+d3MsW=G3(wfvfy)TEjt{<~IOlj48-*8}_N_ziGhR97y4F7!t2vICh5h z6R+;n=dY*7V5YnWvDxb^l-1MO`qrNSBT^hkP?<;>pFc}qTb}q3Dn}$e*Yc>_SNle32|kjQ9i1IQPa zJp?7WuyZR*C@pMlb^vOgIJ3|62{g6B5zWLdA};1Eo5y zMWaV!zF-VVhg$yK?ccL+!)lp8(pkRV*4{K<3n=S%31Gi2vAD|0VtnZ5y>e;*L2Vhx zgwqZ{>_E~ia=h6-eZ!$@kXkQ%KLPtW;X86PFZNUIzolqYUZCT}!=^6Kx{m0|cH$?Nw)+T-8Q1mj_>+n^4z z5Oy3L=~7t$M!J>6xc5)&;RoqY#Ffx4%yxSwVuMhWNFIb)yI2PR-4~8HyjUcJF^4M< z#iD(71pS%+R>$fveB=PH`{BU1Y~GnN=pfiOOVs!jG~#U0pT^-K+xaK-3qCWGxz)f} zQl9bA-QjT)?1)6pE9F#h!>%3H1;D<&MFM0@+dBQjkoQ>00N8nwpcRUB0A^2rOd=gu zNQ4|2^cjgvloiMjB{~GE_%GKU7EVXwZaag1o=wrRvfY(K5o;7`*yBz&@3y~(%A_d- z;6opg$kRpyn(f7wO4ZWmn!g_Wo{aA#4{~&!88uaacKn#K(95tQhP@=F5wLyu$GRUN zcbT;>H!D?nxwi1lMQum|pFOrjyz&%p!NA33a8O{RuPNSx-63v|Xdjf{ZwwK_iL7I z+NSBlm+Ad|uIz;M{cCS+MXGHVDxLbf8R6=OSqpNO0d>g5OOFnvDFFG~NDTlJ-12aA zA?Mz`hXAY_%uC#TNT8q&IPG&IzyHT(3V@L67ih(8X|H+OOSlDozBKNW)mzL`p4=QP zS?TxT=1;JmMa__P+MYQEP6^RFs)`5a`MYY983fMIza+hM!IwVsM|N!#&8psM=c4!w zk0f9CehM78|AU!VL0Z}bz*yf1K2w7qP$DvQ!S>jT!8Hd{wfgfU2J9dcubI<|4b!%k z7Wc3Nf_XBWEK)hvaMjQ(K{>ndxj^RJC;@n-<9hius+e|4psNXl#-3U!6E6pV zI6{=n?b$BY`^B!tNH^%`aaPp~VPHa%q=)Yzo&?E-@WeXgoz2Dnqc^5=i@2jM?_Du~ zaEZge?rQNprJ(oBwrS3tpOa;C!^viMN{4OqEOnK@G4IC5KBd*sB#e%_UJ7GhCLM8r zLD^DU)fBJ;U+VV;u4o6k1v=RPpD44S|7TgZEcw50G$VRIDAz`R=ndVW2jv`YHN=uZ zz0Png_qXEbZ`>^=-W}%{aZ+3XMyOi{F14So0RHo|QID6)R4Z4rCspefp1jAs?W~vq z-cnjLS~Qbf^iOnB)d6TVG9bornV%ubxN!_@QtJMUf1=}Q*x_9Ro&N=H`=*tH#l0_? zGw-fZTQ_$R-k)n8*gTmc=vJi_bJ~AV|G}+y_bb^I16=A5$V~?_5`0`8?v0ZILYN(5 znG^B5E1W{cK4;-&Lp0;z699I`1F0weud&JvjhdR> zgJwmm+|2h@z;Oy8C&xQe`T805$W#CSPGn=xQxaPAF0E$?!(S3bQ&OXU^arVv=>qV_ zm~$0g#Ri6B6BAOC2g=3)MjWIZ2>#Ve%b*HxHLkg)=D;ytn(*5l6cRIIUb4>5UTzx4 zkjB*xt?V^O{VMZVW;H|#)tDnH;}*Odn!%N?Z2Y5@*H6Ys3vjVDJk5YpTbisZP-Ts? z(bVS=e-s=sM3%-b8ls#Bc*5`Ku(H%nN3A57VEg=feB(KU0}xgetwmETC>zp4OfTd~ zG#HsoA;1-EH8)MH=S#i@P;AmCjx{JH?D-1nf&Td(W0(=FLMU^Sxnf4h_ql(UOvX^g>0?8epd z=$%FqQ+ZgOH7u!*mOrY`HbfQT)cNFhmGkeRT$};SAEZmA3cswczp)%xe;G(@i)$gd!=W#}M|Tw^zpu!EdDNl1xoJvbiSC@bYO)+4 zP1al2*mj38u5myJ1DU_()f@DotJD*bbxa-$v!U+PuwDk1fTY{iHmP!ja{3nUbArD3 zegKpBa{v+1S7JccT2BN6yMJiJ7|4M^flC&E;Y_i`Ib=u#fk=p-AK!anqX2-N^=-uu zvlpML4D(BjpA!{i=t!0O({ExA5GGNe(XoqZyO&_KAiZxcE_F7PFhN=SMzbxIKFgA^ zF|lHjhgXz`=dI&6st%syXU{iviCkM}uHSD(YBj8v^%%xBb%NizchDi$1HgUufE|#< zpAA@>FK2F3^-sHPEW-b)Kpo3ZZzWx+3er-wDW5VbZOKf}0-%#XX-PzAG2r(PnHLlS z{Zj$71yOIQ@_2Xmrbl-RCf^h@Kdfc5eKUiR9x93N_kf_)#%~%?7PcA>V~6SlYc~Ky z{w@!dRnZe zg~i!{(8BXTj?L32Mm;_q1k+??>;@hAd1^`-3F=PuYmQx}Y>}X=i)p@p$7$$!L9a36 zaXN2<&Txkq@TL+=xO9z+sD+Yq1Qcvy!n2v=DYx6I)S_@nZ$Uv5yFsaIt4WD|b_OrFZ(=cebe6TdC*QS+`9wSfG6B>ao{oGjLces@^FhQl+6I|6>)y}#=k5h-%f+$u0$zS}y|@e%?h<@2do zq?61vS9qD8Wm~_0v-l1``Rr12zi>{UALzy#zJf3!x$*DIU9fhwg~O4KM_9iTe@k6PxlbpQp|8x(0Rho4!0H| zQ!HT9G2R9Xs11cWw|R!pJTKhF%-aOCtsHZMA1g(1yv=ImGp) z7i-+6rwj>e9@)rGMTDk!37y$*2IVyXL=0#g?b>V5M|_>ndIwqM-VW!cD@t@7VFxqf z1s_X%O^64C;WdE>If{IItTo{{!WWLas}syab5~zao~$cPN!{%V8xv`*!e^FTkBUn% zP0wMgW`-Ctn_tece-H>yF&6yY=EU1^ACmN>+pk@$G!1YmxNl_U9({C$7{8HH!TOo* zWfpgeO*?^DMAB;;umu>DX`(2#Zybp9IkAn-Bp8M=F2=uQX*F!nOu_ZX2p%eL2(dNF zvCfTmW$#}9Etw!LFO%y?8m_!9&V)!*q%n5?0{;1onz&2&pr>giG`d2S_PI%S<1bd; z35Wrj9FFqsivfg}wV;4oF9EOFzaPIPs!P`RT3>69*yK*XD!>0us>a_M`B0L=pYzrL z@5sKj>H2W9E_VD~(=T^3J4$`YT3=g{cOP{DG91x*)HVE>kLn)P$-c)hIm~W|+~ma2 zVFjt{KCOM56(khGC*@*1ry~O60L=f6t?MO_m*?~29R%$D zs|mVMlH7>)ZOoO+KL=mtbBdlkj_3c*|NU_p8@~YfV8clc=@!JKrDpvrksp}y6;;2b zvaDR^Y{jNa1tUIQ^zQ;8kOSuKlc?yQLdi>(qEP{YsM#xv5y^T##~Tywr=UP)WbC=r z7z1?z|8uK%7(n|EorIg~WDE3J1UtZVCoWGefh6c>gbG>8Gi{Dl`0`i6L;&<@mWFJ8 zXs1ig*302&6T6YQ%o>>yaLFL&MY-VPc+}C-QIlt*G@YQXP#U90S5xHO^E_IYy6Pcf zG}ixa^+LSd)99ob<6D_>Bkv9Y8;6!&QT0RWIsR%ZOH5QLQvaq7=Nlmlg)(RRz!yG+ zAzFOoAR^^L2yS5? zluf8U&;UXdlwuR;IXYhU0|?LaE&#;|{#riSmoEiLq1PVX4xyU^kF5&}Yc z*;&g|n=5zwx)Hd$3T^pgbuAqL^z$j{25TS8WpS%<*sTM|zyi{aN2>gWFD2$-d_OFB zhd-FK7oz9mqz31H0xa`%>`2W7RNSr9Ng49Bt)N?o^v}wzm6?I>m5jG3nr;r#AHiqi z6Z##Z@2F2Lh495RN}}vYO3S^DD3?O%x^DtKL_gYRPvROi+*If?|0f*fIKR4Edy_BY zl{7}Rp`lK%o~}oaHbic=N8F6hhoeGuG>g^v*j3z;!%5@ltmFQ-VTsW1@WQO#Ko!hA zW{JM`b46Hw^xZ{lRN+AQbnKXcu}~PH^A``;$1GjfZ}Ed0S#n16`!TAF0$SF&6SSTl z>n)pNet1=lmM*n?me&*1#d$=?Y@dFjdSd7M&VGOCWv=5-hc|Rqz9Qr6BbVZ81aJ0x zdenu>*xCxnnp193Z}hIzNP|k5dbr{J`P|#Pqu43k6~!H>vw2(@+@0DQ(IIHT_aq^q zM!E9tMV1RIG~$e;4+4$GlSALB%Jm;RK+k?9#-tMB?oDu>+0w%V3JZ30<^z@X?ia0> z1^RySWeqDS-=pO@e|;&h_Q~P8;&1E2f!ZHN{L>`4bFZmhyM^&-bjqs2L%XhvBgj32 z4Z{wwahc!w+pcC!Mi<9uRJXlqwd9S0yWEL;KjiJh%HP`A&imcJH5xolnRa$s{G(Os0m4_F2!L!| zuD#;o*B6E;z!hz1Zh*CW?1BZ1zA4M;yj zf)tSkEI(Y07>b=vC{E!SZ0j<+SvlC@opB+#=V!8F10eiwm<+C)63BD(SRO_}Y+7`n zIbT>iVxk{nM_k(hb^V*nOWEu}3_+!eGy-Xd~{Jgs9H z2Ee3-Sc?~&WbTfa4-KZKhzNz?laSJ~2nQlrZd5K)etX{KD|&x9f3$-grAC_}EO|fX zUv3t`Mai!2WEEm>chMMXJcTFlGhbm(u|y+d*U zju~fE1NSCIE&|lPeLqJbiXI*63fZ9_3)_@ShFD3yh#KF0rpE+8H(yM1^O!QMpWN53 z(jRu8)=(OOh>q-uBT=%ttZJVsBfT8*B6DxQM;YkkJ=HQ^g-B$jHga znI-u{(9i`*qqH3e`nkJBRYJ8VIjIn>*99E{^qo{G)#{8 z7S2FPNEP#AkI%j1>bx+rdq3QxX^En_1@p+?v;6K-B8n*K-ro`|2eTy(m2a5CRA<&| z`Zy4Kab7iHbX{Z>>xb6Bpkf7c$fwZn=7mqnsRILMK9ef+nJ#|fXwsezA$WHFOtRzU zH#?VUtBZ?bOsDbEX@rU_{L%XBCgZ)T3MH2$kupWjfO5DceG$4vYxd^RR8gxcwgheCr)hMIr1bNs8z#;@ZOZ&sE$B zx-T;*Q@IiFJe3nK;8~9_C{oR3YDibi!Y=E(e-vQgPW9%YB%81L71|awox$6m%~EwT zwMc}l@!t!R=)a3SasGqJYMpVT_u{~Z2;Ni@}MYP{&@2A@QVQus(* zdhfdx%vPBXCrfl#qz#K^xEXIfE7LJ3uP1?OL5k_M!(K+xaM)Koa#4 zgsTlHp?~EYc({HUJ1=Dq6>Zr&FJpG%+95guI9NU$XZ!hvXka?IT)n0@w|0xzf+LyJExwG zCzWh|cr5-8Y#;KgOL0mJrM|5i{nm%xRN)&P^;q5F3?JtjT}bY%yAt3U;p5tYf4asAW19-D_BzAkM)UPx4{r=j5J zE@bqDSZ_~^6lP&=T!6#tLjGN29Qmj1{hJ&`Y;fe+vVIYxHDB&+bkV$v&!MEBPL^Uf-w&OT_l^8~IkLUUO zcfYxLhC=U%8)HCS2OsEr#lL!;w~ceAJrPxMO^}{BsaQAdS?xdL8g3JZ?(5E0Rayi! zRRVZl6)OQEozxfI*ml(?vR)bBU$uzs`AyxF)YU&YF26jJAnS?`3@F=P9Bs<+V4w1z zK)tz9I1qDF21}EWMFo-SEt<7pKP3jQ9Uow|`OfR^3>62YkSY3;vbE{>m*3K-E(Lhs z?I5)nxHGorV+AT2)Gg}an;-@7<U?{kJ)-x=ocr~e^JgQX1~#P$IKI^1#*awVRlUYfX@UQr9xN83 zY!^CFTCF-42)D}#s-jZ({5)%&wyF_I@K5!(FJG;2pAf&+_Ps^}7UJ^p!;Ja-TTVkP z>TEjiN`$d;eO{-$hF}lAtg*I11P|C!rBD|KAl;m$HgL9HE9ZvSw9FiqNQ$2Yobc>0 z{ib5UT8Io{*PbqaiIDD0fs(F6I^w;jA;$D47bkxMM2{BKyeyo%9{dPCuH(+IlpOVa z#P$??WPn;!AnO65+LIO|Mu*$d;~jfJAJAglwCz56kj+#p44@oa90#m!AQJbC%b*GU zdYm@bgoM-=z$Dq_d4|R!cddgW0`h1MYJL9j)&oTC-aJW6H*K^CH71xCm?(+5IMaEn z4ZwlZ911Nsw0D<8>(W$ZW|jc<=Wj|s>yy^8kdzqdx72zQbM>^+-Z(%14j!DucUn!3 z7v*aepJwR$*y%0z!^*>d zP1>FtsYvmS^IkCp02g}!5(5zp*8K$dGkvr5bQ7RmI>7BRH(QBR&wNZ>*v!*k8t?Ock>2sOpcSDh zhmz%Y^~?MIRX2kxCYK%iA+wC@>>z~Jy&`5SUC^hBNI=s__FP?lqfmPxR0EPv?=;oj z3o3aar+LL}2x*%4G5A}$C;k^Fxy$=-A`;UDnsd1=qlhQ+vP3@=tn24ov40E1#76qzWHfG__qP&I_&yR#-NOt+g>7a8z*_90 zmZ${zB~jy8yF8W^z#v{Z4@U!@%#QXJMi)Kv)`93{f1`5aR;m;|NYs*$rKPJul#a}m z*ljca_UzZHCl*SXCk5R()Um-lV`jNMujHcZ1eT_*!XV0&`c!}ppFH^MqxhAnO>j7S zxag+az+RGEX`CKYjsGe{Y=;eW><$M&6Fay?alrn%;0KOiJ2aPCb6-j)%=)!`!-8|e#>l2!N}b(OXKaW+?fE*RXU*B7DW0p3Q9j8pPbs z24#jR6aFCQt(furb@JZqkIz^5tp9J zni0(CD^-MivGU)@yF)jDFT8*Uq(Pkl`Gp?yTf`1BNQ6qt?rvvrR=?|}WC$6yq)2gH zuu=8lb0kC2snYq@pe{X@0)I@KXgz+^e2T^79!}7nT8)%Bexs3S+QmQz!ahIDP$=m7 zuX$2Ax=UMw(f4;W8=bXq^aKym6!R=1+a#CPCDnQ)53#gq#PrzU%oRxRd}2E%f@WF zyDYO=$t&@+%-B_uP$^pgt-J28*DtAVW^g}47m5zo@G(F5g)wO%FqcPo6DWmbs2 z4bwlKcxP#6d^w_|OiJsWywsP<0FCMGTQ3U76iK>ES zg%uIFr$$XigC-v|!EJRdX(f>TfU)M5=cs?>Y9r(+D@{mp{4RHm3!@pUhC(+({YWleqK1&y(EUm}`UkMO|d`!zMb> z*j+}ui(k7JgD(Y8W3J>cXo(tb*PZI&r!R+KA{A8!=E#<4)7~tvH|&}nU5+16SA-Rh zKekQ~K#P#1Ws7sCu9O8HS~U6e&wu&tm)5Eg${;J|$cR-w4SUs}`=jxTOw2Uw@OfG@ z++5~~JSRw}X;voO-PnWk3Zo0ePEl&D_c%;@0yhTVl@-Dym^`KUtEEKFcO^HJEd<>d9LkYO-knbfhY|a1J(LxyWsW4-tno z=(<#GnszmSLZlm{>ziEVJ~Ss{Lzs7N214CVOE_M& zuVO+?3n$(d=Dl6^_tjWei53#@zdKo0@eTn`|B*9-XoG;#t8v>BbwE|-TrpVTd?((GXF2Pi$tCcay#M)7c8X>pg!*0Ds9+9l?P_Zn>OHZLm|gnRfR4x?$zQ{ZY&NninZA zh2Lpx_4@St->1Fm#a^YGHOV@54)rCgmErv+8yE4@I!4e1jfQ;ZCuIvM;-;sey`<87 zAhc8qv0?Xg;z*z!`FT+wr?cVT#y>LSlPWqwhQu+&v^pZMZ*$CZ$6w`tQFVNL@oM*r z=15D9j{=;;a&e-u(2X>PUc&o)`^~n-SbnZrZ{6_n^9jI1TBrz#%WYUnotgo^)97#c ziruxM$YRdmgdYIi$+K}BWg2cfwCB#oe_Q|1@yA%vjtk@Xdx1tK3yU>$SS=qcKWs-G6-Yd4$C$dEcrMH`QEp&^1}N2D!!l2$!D5({_dD^Fb+1o&5qejwH>M?Ft_c z>i@s3+-h42lEWpT6UHjXZKgl=ZqGPw_59HH<*C9-!ZpX4{-!$l8vqlcp|Jt{>0&sE z`Y985p05X#`~53pI=W9J$Y3iub35Nq*Wlo?yxF-_38w57soel4n+p=I@R3EiE>Dz0 z+Kl6S*k|J7-y5I%Nh#jM9%-E)g4<9s8!79%0%Xo*n&?ksutP>`LcU<>JI+($x71kO zZ$S5!rCI}g11xK5Ca1LZ+lGvd9sN2kut!E}v+U&b-+^0DE!|17ASE46!v@zU%5JH< z!jtzgmT_8%eF@hMdKzL}!d5go@`KenX5!6sVhr&m<$S}hh=MxCLh^8xVDGwlFQw8P zfr!HQ6iOEw8MGqY^Q9iTHB1_$Y=S|WNnTTcJ;0Hy&tHcpJYY%^f|~@{js_Ce&HwTgW2@VT=E$xYPEmWu zfB))iy{)d!dZEaP-SW2i1h|p!x*{#A129g+M3Fm91%HLP!y=7NjV+SN>Zh3&rXX0yrftfeNBLE|N6T zxrY2prniuV-z*+^v_Vv=lFbDMgVkCye5I~s1C0R;uXbJoIy1{c8cB8u^roxJtPDfs zilCa1!r-hh0QTeGPXHb2KeBk`K1vrgX?heMI_^#u>f{^!Pu^6=;lBcT#yt>P#-VTH z*$xR70tbMhV)a^sa^}5BBrxl+*3NEzu2y}dA=5?10pbV%0Qp zx6=tlb%*Sw7cKb0F)p-Q!X%I4K`ZfN?0Gt`9W%Fr(s8{z>pwZ&U-vhCU-vc#fLT`E zIJWzm-LoFGih%BA>6_HG+Xvm7e^RStOR?=&1TS35U&+3fKa1StUUF&P@>F{@|D~4M z4%eXB{{+#s$1Q2%+kS<{Hm6aYZwU$24FhyU`g0*#wsGy$swGA*v0Nh$$mco8b6FMD zzNTaScN-nvXtP&V+C+Zn2G-&zSe|5cLCG01p>H5L^e-J}S1Vy3abNvX288;V-gI7u z1^qV%+aC$M=?eoWG3KzV2`zE+yg}T@9qqW741ubIY2>0D{zzdS7LcP|XAS-|OJLWN zNG0;b_MfFH`YE-?tL;!s29kCvC?AOhZKl2QVp1cIdUF#H;A&4<{eC1uhMiHA2XOze zx^aA0m7QBfP(=_Kwybt4wB<0gOoWaAsaMu~u<6xy6y0(i+M(Sp+5yN&BLflW7cx!C z03v#F@#miJ(CC`QJUx!fM2!#;=VUwVE7f;acjG9N>ORZ((Pys)K!Nv56{_h+sjce@ z7k|3b6r{cr`Q`k`K&a8)brJ7ycCo6-LPtQ=9^SCH<<5|_;N;xXOV?&tXF-_3;ahvJ z6B5dRCxs?}nBF$dkrF;|p4+GD8fghSna1iyL)q<>W$K_C^b|oq}O@1oK#2QmwaFr_8S9Kw%vm zVKUoW`RyYs-Y)%wM1dGETrJ6*n|ST@IzZ;=^Ac$bD>tK~^h%&~oc|-_3j}_%x?dJT zmaLN`9O$LP<4yd)*=taQV2p1q1U%I|EM3rVpg1I|=UW*E~H zZj)SURl7;C)KhC=KfFNFqZw03tASo+XFfkr+ku&ZD+%#6d9nEksJzZR$Od4<%YYnh zZc9(TNgO|FypDDM*F{( z$J$-t*Q+3Ii;2`XE-$t0%1_F=o2ZXW{I(Gz7g-g1$TE1-lH6-c02`%Q)UR!OnF8|M zk9h#5PENp~EaB7KxT5ZZ!b{R1Rf$w@;fqgIf~_BXd#}D;47Vi@Q^;unrwHIztJMFu z4k*J%5N>;@+wE`jsNzKtfIz#i^t=93+#6!>j1Mcv>g9OU&yN3w(Permo_*bX(uP9d z%0fcv@xoqB13u$>Fq45OOMogC*q+j1OGJ(3#Q%r&E~mIxomA!O=}hzDLHt*gkk^W<-{AF2S7 znY>wm%u-deh1YPq+vmLp(@!#ua{-z)=gx;I3qlP*;vWov**!!pX0!iXT`I@_C6ef2 z%X4qt^9P()2FDF6nCMlQ=C}jlRL*>V&Hu=3whsjT!0eD-1~A|dGw}P4I10BMo;Qx~ zWfsb4g#5(OI_Uc~1i;yX3^Ali1=8T;9rXn!s@Q#Xb7&VCvUPv;D zwP?Z9f(ebn{I`(Y$vnX7{$W+q@L{+K3?L`md~)0kqscl}Uj5O@XA&@m%0()AmZOZ2 zEusU0Np{@xXsNg&5NCQU)qjs;*>%I_Rx&KRQvPFPUuKG#P>+2KH_R@4o=E7OmW$S) zmOz$L2q$qPf8c64w2dA`ZQVbSP{@I)Nc@eOMIdzl3gDTENwh@4~kRRu?cZfcUEB(&k{A`mI84 z%qu$dv~D-*X)9XWH*?w5^7yh}L9L518HqThAGP~S5~wV*KmYuF7M4aRM&Ta*xsX(Y z^=I&a{f|Z7rXa>(yx^*Ci|Ao$7*#)Cg2i*0tQT1?Rw-H;xVDW(S~3a0xv*}j4q^KU zfrPhRMz%5`CEk_0y+y3*ZK_Doi@m^U{Blww)^^<`06p!LNQuikT&8i^pJB=&R8RO` zn|}|K(}9+e7%~;usn2KDKG;grPSQ@oDH}G(XpqEAKMrMOJm3OAEJ}vnrTq$yQ)K!k z%hGXz(?dh+fab8*T6$B%|Kr-#*~%|%Y!HZ%|9_~l&!<_((XZBhPMfDd4pW7)m zxcZF0%Yh}XEtOe!SE%)ySm^IxpUp^nI%`;`(+zwuWCnwiu?32g@qDk&SI41h9X|&E zM1uZHS6=HD!g+45_r8-SJpr|Ul4M3iZDW!RfimImj(#nfU5)N3lryxLP{JrL8CoIa zi@L5BqjdGg?rr=2RQ$^mz8YsL+`bFPPOzSiFJ}rv3?dL#@6>IlY}{t*TkGTEjuDY! z7wxU*Gyh()1kLe2>~L)oTc5eh>H~aP%R!jVhc@c#tP);lBmcG&ZoJXMcV zS414Grt=IF-8-V{_?Ct|38&IQj_)=q`b^2{|HyM<$b1W7!Bq(lGQ@GGn!)}+1Tt}l z&n{Yc;PxcOaM2mdBC^lrcOqBJCk6{r-Z(~Rq+&dJ4l-Hz;-t_`hu(vXxuJu@LfJ{n zll>}m8dun#RB{lRe4T14@-h0HyaAT&~mM zfjis2`V7;%S+|eV@rGH=GF~C4ILA}S(i?q`cNK0OQ`q`U@5BH!A0+?y#$6xxw4l1{ zqn&8h+s(fXvZSb(jKh9HRUm(;()q+KgzIUYMAn?KNFFkWVLArl%n*G`Ttt|(fAx-FsPYQD5yb%|IVB5X&D#%#qcE_|d`L z7N+kn{p*q>fuA4kERW;5Xn=&be21=6`J%wJ{m{w$`uh2wab~sHPyrsB_fx+E&j_~2 zhS=MemjLv&RcQxzP^UVJoAd*pS~+if0H^qfU*?T8MOm_eK!Y^~6$LNn2ZJl`ws=MM zZ`v%t0RfbRgf8_ zR1C`LJhtL}X9PYwOM5}}8>^hLm8emw9KM$osT1_J}M=07~;Sp!pVr)gJqIK_;Iad z!+inb1%^6JW$tui9{ESA%f!>ehEg6#+E;ep=h$lQ8;Q2M`tDrqt=D8~$rJ}`tFL=S zZwaKsUWI)jU?wxU-q=V?G{(tOeVz7Q6rJbyJxrkQog$VCM~Dhk&wbH3L8xJxgM4`t zeYrxu3P9&7YpqADJrj@iqCVrt;p{eyYNEu&t)p9Tcci&PL41xASGYQi+pF>{l!gb~dyt}@ea1eq0Ng4l3&m*tzX{d1=< z*z|P}zr@ia*v_Akf4rWzbpSpTH|;Xl4J@*kY$@xj9TKXxx`A@*+{0-S8N0GiJ#j_T6 zfXrpZzUeHCQU56bKi>FRo2GvdMjb#1<%h3*kW~YnO1NrQfAjBjN7}kzLcK7HF+dAd zbURs;7o9R;`?g9^MdvH)YvArXv^=0X7ow+e(pL#fk-d^>TRVn5& zM=LOZFwh|hKp8ba^FxRi65$&OVB+Nl+@ls2aS=1Fm`FTB$#RU+o^x=KpoRS&7J zphG+vPkb~by1XMU9570MFG$vYcf%uttK(tIVo8z|vB`(K3J zby!s0zbJ5NB^6YV6a_(Ar5iy&kWfmxyL0G9L|O>}Y3c561*B7&0qO3X0q)xPzUTbz zd2Y;yKh;NOX0NsP`qos(S=132vU@-KM;@SJIW|*NrwtRTC)>07QFjOFhNAHQhfc>$ zJ$O|CAN5orvkP3p@20?PLw-2JJMcRZs=u1QAdrW>n*r%V9&Q?@CXj#TBDFj){zmJ! z_k-vP=fYWQxxU9G`>c4mEuJ8?y^yrdY|~e-2w;5(rZeT8m*|Uh_=aI9A+GAE_=Y0G zye|4j|IIJtHSp@ZjJRmiJGr~_bXEcPLmO`31AL+;7!=}cy$TY#JQX!6>d}g8h~l~` zcXV~Ml&$!W5Qi=zV2ikNI2R_h6E*u(@7ty38#=Y)XBs$}M9f5ZuXrw`9=Z5rcA;Ch{0K2= zaeYZ8%*hLE?`qtyV{%f(WhjBb0XUJ*AOrmc$?8H}Un%(boSc{r-%sn?hX0Jnj0Q09OoY*b z4ssIJ`M`;?yVLpR;l!;Zae+`*QjFL>Enc@ZrXNWm0uS5>m;2*{j4#PMIUwGrV|zsl z+M*>60J}k9k49V#p+gVE8=X+{^uo~!l+=AUz3G1IjBuY8eOWF^$mP=DBkvu@N9*WO zo2(PsrBvnDeIqbmaYA-~>;}yfiB^|xc9k-A&;6YX=V$sxpHx0m6Tman?{1F?4Im9P zBhR&6-59F^@$A9U$LQL(8VTJo^6^zEN_o+Js!**lX8*v6=s5_@XhtNxSzcccsRf%4zaVGM`81!28C!@}gB33D`SK!-mPRE+~)OMt1J7ADXTX4`BJi z(06XmqL>9hu2P%&t_wcV;kO!z+_7q7VXQk^_?OQhI@>1OCYNM|!iY0&V}?G(Z#mvt zU)*H2IUb<)LwWZq(ZiBMfbS%S`B=Ut(fRUG??wVvs00_~j5GT;P9aOrN%9ebTq;GO z)0GA)B)@Fm-NdK73g`mohpWTs1nABfz{UBcb#3F?c_!och?m$lWxc(eTvN@H-z#35 ze|ZQ+u0|&02kX*cudH>L%~cT_9ZE0W0jN2zp+doy0^%yIX;&1p7|880Hc%)iO7crm z{4#JsY)JwT4V=Soq3~@vXjpOUKN)~mHCcOUJJ0X6E(Bm*HP-CMX&JjFu3>nM^sTl*}ToAo9ht5ehXg&vn8 zi3;F%od@aARXb`K`YPB|X1Jg#<0YInHgrtzotRBNp^d++*WT_K%yr*tdp;+0t zZ|;?JbfES-x0f|tuE_iw=z?2#m0k26H;mE9NhL+`zkY*`4Xq+xOIZXC5471Db=;{J zJf8oeQCV1)45oyNh@qePqM^PFsFXOLF7-Ynm&mhLVvcJq7Sa-qEyRG#zG3VRl@j}W zErUvlTNeqw2)((akoIwM+S;6)xcB|~)#zEDW1}#i+3E$CEekWE=;dwwPVvR*%_IsZ zo+YRWL+@Y>(sGc~mX--}=CetG%Tf^9;l6y?x#JWDtEgU;^Hx!@LJ$!Bi>^olnS`SHwR8C^15 zFK*=gZlHYKtF8ML>cZVg%x;yNw+dk8vxBZ3V=MZuQ-_snmNw5|K%f);7WpJk^PMmo z^xf^98F-!0SNvxbdc5e<>(!&Sd1wuXZfhq)fheq)^(4EfncwAJCx6B_YKL5%@b<6s z=56E`MobD+xS7G^halc-qL~D7g6Ixd8~DE|a6|hzc*j=B->@a=@VAgm?wB_fT?mfD zy}Yh2E_5#!&cauOFnc}X5$q=|fyPux9?qB^t>^wnw-Co0{kF*qf*lmkaUjBOpN{Wt zJ>50y%i}4cu_84Xn)zP$8a5zQ;H7rOJ_61KR2z&Vklbgs`wK-z~pX z%1mPV;bJ|&e~$ho{t1VSS?g9tRqU+RIbuU8Po+Ss!c|0B9Kbw09a@kMXKiXsj;sVY zoqKk_&ObvV=i;`qnZDsFaTvBr;geO7{M%EF%&a$uBkT7!YbxXOgf;8TCZBM1L94TB z4EA%A!@*UwTV#ZsG!{3g6o6rEq2#z>BiRHVyiy~>-mZ8IUSo&8)o(s>V}hCphXRH% z;Y^=5Xb-+BvE}I~qu|DKoBT-$N?G_0pof1DAU!rA8CjRl6D`2^nnlbST>Lu!azyR% z3JPqbOlXwV7}-}3ZA;ZLZh3lK)h;{sZp8a(RRi3s&X8u!x;iJ=R?6~je49CbATdH~ z)2(EuDmhj}&Myg5KzFtTDy{PqnZmRrE600+E_?Ia?VFDABc|`NeiczkxMb{0Ek&~# zc0^GqIC-yoj$Y&ad{X-bq+U-N5|<$+fF`Wc6zL(~YxN`@xIEEoJ8Yg#vY$3Se@9Tu>m)YG-=3MnoXc75|7l zWobM{Up*{KMV0&yPK7woeN~eNjaWcYOv{2;AA1cw(IXoO3mXITMJmcV; zhs-Uk0k>v8*U~yl6FQ!G{jBQ%{H6%2Xsx6l?n=+aHnXOvN|(hLF~3b-@2ZKjf3aeu zbmCnKH)#6h`6atsNM)p*ZI!Z(mHKm&Mzy^JK^h_`{AiA^`)u_`3eo5aQZD8j6OSsYS^9{0PAF=^ZOC|B@e5sJRR3fGAu3PGRsqQ6dYPNs z^G18-&KCMc>N&ieJl{n3+)bn?*s?Uu0!5$FYx)Y& zS7fNzogMQQ!}uq9%C4s8m48=2SAegPNAG$3(fsJ3ii9@B#mZ>vcgNQopMUSVj&3Lw zsff(vITrnHU}z6~=pN1hp0^9DaV;;V&Z~}Kq^dJh3R^q>*T?PL-TomxY zTt^HGyq+pz^GDdG+OiLM}lbd%NS*1)NbBh}Q z&Ovye5bqVe2EXIp$WMh7Zh703Kh0N5+TON5?#ROVJq`|KBpD#yR#|n4+fMOU2LfIg zBBw-W3*kZr$6cGz6VfgRa_V_olNR8hcc7~qFwfqBo+Y%$W%FjAHYm7eeY9mjmT_y^ z)h;^RQY$wZOnZfMQ>IJbHEAZka<&myh?sV?l!vo^(iRz>i2b^wX%1L z(|U?&{oTy3hkm#zXNm_o%L;D^<`HQd*}(g0+2zQd$8i$i2pP)c^s=qow^*8xiYAJc zVM);7w0-E}v46gJ1QiMPL2aRRT8D4R(eVek`R(8}a*<3nnj$i{$iS$_F`Yt`)#Y{I za)-UnUH`0oKh$P^Q7s@dnd<ehRr?tDLp&Z1-dTO?^f%ymhPr%2DVurWY@rUzm|DCwSvZvD2iyQ>$ zsVPwIzpTKY{7}{h+Rqyewl0R3&-AjvS|V!(8Gs^HI1fI8aezT9Q}?Twt<;$c(eI+q zCvSA+GUgTL`(C$CPVk#oZ>#!Xt&6Sj^-&i{Kf1Mp4){mO{tks2KfLSd7f1jBUnc#Z zVkywNgY+cZu*|P2nGLi`PnVDg51SoGZ%e{=^yHP5Q;%2`Dnh@v|A}^fZ6H zn#YWND;~EBg8~#oU=#ay^&1uI0zN6FNe6fxvc7A14$M)k0k=59#<;Z>oJSArGnS=# zqfs&NAa7a0Kf%M`zzFDmfj?=I@suC0Eq2s3ANMpsL~cnuEb}tD>bXpl{kuZGJW3nZ z%4(2BJU^M0ZAf(G^FXGhqyvxFA3-bVRVbA9%g7=MrTto7B#JAa94jJ%n#ne-^P5Vv zXzQ|zf!m_b#4u343G#V3o#QgES7uM8v6L%VWR1(-u~>pYAOGn}&GVzK5k`gi$su#; zBIbh0OYT+MfrW?HDG!oNiBG{h`){RV;ep0VO`n2m=6o7}%I5%E-5vo0~y@`)=ig(V5N9#EUH(WQ)rD%f{-m}k}N zRUW}s2U0}?Y3vZNEqDdhr`UU}j9O138Tk~Qvfn)7av$5(Y)c(4fH1(u&x#Lpjf1`& zvutTj%5N$HG?6$0lR zE#5QjsZ4|X!%XtHN4ETRiB5c5DeqGu48-vy@gxj`)(K0T-Ig(a*Ewe+xf)+zL6oe9 zgJoDb8@xAPOc`YsEEjgEo?^#&j_=@^4R|hmm~l+q*Swgf+ht)U3_$RIguzHKq23D>Ubp zU#6eb$k;{Pivv`nai~ZI+1*W)tQLYpSvecM^N&~PzHAe1qdm7=m;ITUZClToHfEqIM>c!vNbU1Z%U^3OqVlMVkw zM`^fWo*iqsmPy0F`MGnrFC~Tm_UF!C=VuE)L$o`2`0bn*A4b+8{%i-=X|Mdv?2?am zn^-kWk0rd|+&QkFIhaI=x!rPTy(b9^ZmWh4$INe(I7awVvjW z+K-MlBd6JWZ}qts?An(?g2nn8rW>Xy9$mj^>@baS`y}hVq7hMmu$U1? zu%>q@p_@z8;QG#+)~?PMh9T`P6idY-q~S^O2ulqABl*KfGID-y=VCFIeExiX+(*Q( zSUzdcMCRSoV62W2+g3g+*jK;(fE>_>=JU8Zycj;AwF_DTZpV=WOp#MIu5D|0gf-zy zRNX?Y*PU-hY3)4ZP6TF7HsGRZ(dj^*ok%W3-ay$vnP%_ufh8579nH75|D&6THC3_> zP=8d#;c9-U95nrlh)(XCe_H%2j$kvXMM?>wFi)mPb!WZ#paA)=+5qd`MJ`b~QTpS^ z4OkjH$W#8S+E+FDJ$rT|v7V1#?-24d5V_JFrmz*V70T<~%KeJ=3`-ARaS2rj-SZ~$ zMnEJ$D3qo5P9LDPyKesP-%1ofhb<>rFZXqtv&QmKq&iXc6C1p{+f37{)e4We){ehV zeW9`I^8w_mOKl8Oe-Nfu$wBRi{lyC3>ZCV|%X8=1D9Vl!v=>_T?4c0EKj&Uuco?M7GyFB{8DR}d2PQe$4XR$DcrW&%E?Nf2>qRwkbK86j zcg6GU!?&*=8ONH$n$Sfm2xWS{p?E%PSl{hT7diF0n$}iYsZVGOAbwtGsL@^hB+@PF zlzuN#Y&-Qz^eQl-L`dJO{W4eMvlCD$#?m9yBZT|B{zd%@n)T1@{HD%BHbK+)R}`0g z5B25SQ*eW7X}07hlgqUqagBKuUTJi)^|SRyWJ|x30zQdZeZ6A4(v3)sN{ym^8*AL& zACmk)dXe`Gw!ec;HS6JgB9g{ZWUsYS-xqqO?8p3q}V| z^pCprl<3_Pdh5b%uRs+(7g)x~d_Rr;MO?dAPZ0ioSpq3T87cG1v4Qi9Y-smIac3=cUJ zp`hTiS&aGQ{Fhk(7UwVP&u|)o0300R`z%P_s*CAk`>Fnxxs(S><7 zyT)w2&jw7+F(`XN$j|greFjVO_T0EMFa*{xXD30@0;h6|fy=tvIN*PS<_Tm3@(`@^5KEap=O^9Ku$ z)f_))xI1u=e1#aOLIc2;6$;qzZ8&7tFYGZnn>9@D?Mp4!|pKlh%;$vck=C8B* zoW8hKwk?DXdODnF8hf```b}Y0wY2^K&_k|YbF+JHV7emKm`Y)hhPF?Oh2I2bS14UTqDcC-^gb=!NPURn=BL zu16cAMJiMlm=n0A4`^&XZ=Dcs(|UJEgojg7rEYK4R&C6F*e0r#TNOkJ;UuwSo4jG` zv2fmggb{u{noVS8!qIpv5pM?Juru5sH@cT3>&1Jq=CBKnh``j|5mIe)_kum0FJ=Hk zWZlK#@$l4K$xr>@j{Di8`O0N#z0K_q;YVY;rfpexE&4SbcHv7YZ?l1$Gk$egFoKv) zW|!K2v*r{6wlMff80oa~5eAYDA1RKQH}Nd3<~RMVonTo<=v)QEP(4OzW2^SD-`{KwfD z?K$mv%bS2W>=*ZhXQTlfSKN(f`5kO**zAD+_i}*!>x7%pPF#b?j2b|SIZa^$Yo5tn zpmliyv?3P{!uy5r*Y_c8!k*s98!kc98Q-N}$LchWxfh*jrNlsPI8|(!O^ssyO}Uk9 zNO0vk0&BX^6%u)nl9CHxdyE}w4#V(BW}ep?I^ev!9$2TS`oYVWyXf8;AD2TrU~X&*{)<>A;{Oc1J&xhZOgg z06Dn_{Z}R$Hu^207+k`h2}sT}OC7hCW8bZS+~dYds5ukr{VtX$O?sStEr?E@A%G#i z1le_M-MCdm*WB8OgJ9Xj+3$nsBtRb8*4Gv7G&bVka4}66+n{w`_Z&yD(4K58_jnpP z>{@W7hI4;b+u3X0BVz0efHp-?No66VxHKE$`Qp6SWVdF?zI)|qbnR=AVU zXno!quWoKtP=jH|x}??BfWGn&NXf0Ok2^7%O>imxZBj6(_Bf)B0j9Lgz^%>$C>^P1 zDgZ?Jj_MngzwsoW#2FOtn7{`c^+)%fU%UWxEdwhkemg~UCRPerdXa``Kpwte2k!oU zny>Ia>tW=JQm#_2Tr&w}HGnjKs^03LwwJJG>v(FbN_x@Ec(gAUZ$qf?s`~QrZe+MS zy4x#A&~$3PsA{01?1-wWv0-8~zF0ws+Jg$h_%~+HbAHj-kgTx{>C(`gNNz_?gX!PA z6Ke)^^rRZ>)h2B0EG8NQ)i=m(no%T=Wq^P$RC1pDXLmk53Iaq7^QqydtPxC(d*c=O-;UnhSN zB&=Zxtz}8Q1N0aE1D zY1N)JnoA@krAgDP?$T-1Eq|w^vE!i-1AvX3JgYb`$8r%`yofkKTZN76WW(>AQ?9#Z z?@)JIC&LAs1WunK02KELR%%5*_3e_6~1GH_T{F z0AcNo#@`+Vsi(4giGTQRzoYvuD@h?4{02Y=7f5=AxC+Hi@H+7*Ywqnw6ab#}c2mvO zvY^SjDd1BO2w=I}BLAHJyo0KYo`3T_%CA~T3QHb&@dxnSIp|>`J_9SKGbDl9g)U@N4z+;OznG6j5O;=ZdA#7!`Hl0xcs|WK6we%O4l&@hu$J1;>09;L)ETi78xT4X~m!VgM}- zooAjAdZ5vBJaLEjywB z2$=ibQsAd9YGDA-bUy6{MDyyET?+P7^dx^qe4@TR%~-8s!x}VgU$bJ@pQm%h-&|FTwI5t+(Y5m^VgscBRUM-e+rNu9+L2*Hc=vd2t7R(Bu9O!t zz)-4WkSCeFsWg#SO8m{!>-=Kv#0Jdy>6;;s@L)w~9SfA#!q+|X5b|qE07lD>9~`SG zkS`MUXh+O<(`DkIbcHqymcRdg>Df(L0-&htHSEjwlNmg&%Keb#46!;is$WoGJly`+Da&S~nLcDv7;(DJaT)`%J{*Qsc ziz{=BR!g9zV;CTU!hq{Bpb?m|Z)4sk!l?%0by@ zTJ|S-jcAIJX#cx5G6p+_TIEW=@b3jZ*uJGaw3CLV534lxf~JiIh6c-q-{ag10OLRT zS|kF>nN@Y=m(|{-Z{l?J%7;X$qpS5EJ~7|)`80%$=qlVTQJ$R5ze!J~_E{wc!|w6D z@Gcn)X7h#&Qo2;%Pt?%@^XZco?v8bZ-YD>Kb>m#ut7>B;#YRP$xHbUX;D$Nv0Ip>;Jb~G9oo?)cSE1G#U z)iGh6qIl7LJT(09Jg6wYnFAl)7t!m4`ZzPIcCOfdkS zW-ql8)%SD&>2d%T_h45=mGDGLJ^?3R?i!%vFz4Y-;a7>!Xb+v?MVT*LGMF_l?2H%t znnB(B6S=Rrkz9|h#b?p6$ROOuyW?&81iEgA$E#a|Y}I>n?S5o_%4v}7tcS=*-h$Qg zv;%C(BQNAa9pS*q^Jl|nXIkQf(kQRbkjY-#Q#I3XJ)FdAFd(jCSz>JPu^)G0EF-A+=wde@=%gcMWX^SBLWc zM%=Ey3s(+Ml}GzI6s1u4#lpY@{E&A5pCR~hFh%%V4~ZaNW#WZjvSbKwjGTZj z4rLog{C7nauk)2sU7reokjEh|ztAai?!d3f>k38y7jQqhz3DW$RX%mgm@3)B{kSB1 z<3x(HB(rk!S#}Q#f!K8V8Npae)kIn?Yl~vbA4R`k|M6Bdjb#}n$9tHy`yYy~1r8BB zP()NHe}5JGPlVhn>ucY$M15Yr4b`!-LSyO`kv_j~SWQ>lHE8x&%lQ6-1mZ#@7dLzf z{eVa;#Mj=E5u9YT`+R%f|Rd3NA`Uw|HO=@kJQH6BHPedWEQwH!$6bO8O_ z+j1nnRoIq7swJ-DWH(g*rT6Xz^&pJ0iq6;Xa-x-*=@ zoefBioVg8}zRJ3nsPd5)shZ^V$@quO4F z1Q9Je$msEh>!h>hE-prRzB}3l<`LP5_3h$_0H%Dwtvp5rmCQn1e&>t1Bfl{n(8@0D zOiYp_ij7|P_6VtpxkiDl9{J7DwtG{Il`0mroFN7@k6(dSw`*;~fmsBBt@i>Lv9EqG z20Yg`ZtQc%`=TPF#i5%2cy}g3n+}YnptxJkVxl=fdfH0dh&;XsT?_$I8sB=PpNk6< zxDKFw--6uW+bkvm??Q;?^Ivv{;$JhCD=oKm2hiE~L3?9*DN#j_yZa(%k+p$or*pyo zetTMXoU-O*TM|)ow2p^2vnJ%o_z>FaMT%2U#8H@$6(~w=gaY}Vx_kw$S%3W; zUh)_7V<$x$PLZZZF*!$2K{Mnrc!#`l`#vS! z<74!#F9cZfw@NF!Ct2i;6oJDH1l69l`{&Gdh6ZB7KE5)lS-wh^3rjn zJYj@si^*Gpf8TI)kgg%04M>*TWdkOCC4X1u|GqsiSO9)J(r$&N>Wa68e0Z=czm(b~ z8lAhPCEXC?${Y|7gTqX$*2);G!cwI1a7C#v@0*H^;bL^8ig+tej{Uop@4`#fUAEB+ zyPj)^jGb8D4wP^-O00)a@pQ=f?Fue4>k+ML`@O!Oe8ls%GaZgpV;EaH5bnzL@m&49 zhaNA_{SR-knOvn@_#KPZz6I_a+HbE+_dk$j{HDx(;Z^-yyf9NkrY8QjAt`_@z~uV= zu-Mxs2a-Ge=RvSBQxeL;aC=s$oS%AWm4WrIR&XW`QtE#=-vQ~dj}K&*6MOgutW)5i zYI(#&K3yafGmXWI3(Qi)x<}BueWy-hlM;oi%PMMy-tRK{f)SzO3s-BZ@HD4B_c%nX zp4_fybXM7AAz=_ubj@aGN-lP9-M>|Ax|`XOfaxcj1_8TO?(ipBM&6>^j;Ce(S9jr0 zGSDe2YL43MOVa%Q`{MtvyNQ36?YxU#rre96lz(#}j%#2>ZW!_b zL2iUInx3$;nI)U~t$-@xS=6zbzN6d~JWGgf&*x{-?MyzF)|Cwt$C#UMASoI4(kn52 zY$~3GNDY^p8YIV7%}Z(+9e5%39Z^0Ni<^qg&NSvZ!A{!}IwVSEGN#*bNOmAS|`AFog385iZ+5$IF3 z3x))sJ-~X7Z{WR*+KUm7)4Y~CESeR#`#HA$ZkzaRf1$Sl5&%|Ib}Y~qqpEIX8He#H z0FovM4?YL6!Xhs!mTFC#x;ab{8Q*wx4j+5E{zQHl;b$RQFR4rPNqmS-x*mYaC*u^q z$&M20NqWJ49>nXpW5zmIH8OsaHYgxiN*{&w_+u{8R?3LDBepVtbtm|Dl-1p%>l+d1 zIyueC1C0-$_HJJK4N{dEbpYQ7>`#z1{eUKk8Wn(88brJ|72=Yu`-&Z;?^3JP8XKMjm>2> zU2VI7=^le%I#I>v?i9gS-ZJLzb&lU-FXku*&|^ng@-&tGjPfDb8`?6q*AqwCK0FIa zUEgD@S^Uhnt~B%i-{4CA^VY=i|GL5bNBoMb&dO;fBv9`owpRvTCfOUVudPxOP+Cdq zPD>*ef-=AMQA^hVP|N_*--0twl0jEtulKxOYsid))T5E#3hHtJ*r6ZKaF?tYKU1eBQtCOK-%?Zyl~%6tNR&kB~S;^ElSTac9HkMWPq!o2rImi)S~ zM2>%M22T)EABwE{4Pe=PUx`eTaD90^yV)?|mijUSSPCy1E}m#JB1phfaL8Jf9mHU* z-$kt)o7Hl2vwRi?zDhfWVZG!DJoh!wAUZFR{-?M0Z07p{_=vY9-`wzQlZMAOvU;?@ zGI!p<;|6-G6KuKkxtDCa7NW2e^K;%Sl!%Y0-oCt98pKgy7%pPy2L?E6`yZ|?ucceg>112M9x@Z!mLDk-5T&(1kBN9*wN)psJN7e>ZCeP0t%?Aigs zG%+XPxxSiE!p4vss{!#1N4J{ zk3jG-(`4CAUe=xZ`U5+XNn?}bRFCF@B6aPm#+0i!S8E%kK8MW>{s0jMvM``rZ(c<9UDnPd!LGpVarR$AaSBp-SHMyM(wm>3}`Rc6+s zv$;xB4a?5Gd+}&ij!@P=N&_~8giw&iU6Q-=BpRxg3t<$w?b&qF3+M8JlZzjbjOr^A zd;|^EL*)HGUED)39!rDPSbm2DLqg316{2JKnm9={kOA&OTaAV4qVNQ=F92HERIe3R z9mVF}%tBoUF@v&Fe(r4;&E+tSTHxBy98k5g$L4*L#{?jUj2JsVXcZU(L+_%`7vS+f2kc>8xR7@(&(=x4X-WvV<7 zVX`C-41CiEIpCw7rt(nM-#6DEs$@{spZl*#L{{$~N}twbGv)UD5Y-35w`ijz`AfVp zm?E5!UY+?-s!z)Z$-5xU)cw@`^FmAkqaIQxPiL}(VYS1W`$0D!eat=|0<#yf76{m4 zYkDpM7CL)k;+S8MwLc59{Vey!n0Gjpmf7^yZjhY2Z`{2Xj7G{ga>odP;=!igyZC&rAmlRYSs%H4Z6D6-MK|_-stmFNxVe4vBD# z1_HzgUn2u980r}T-QjJSIMIMbT?3P^V>L2$qVvid-uLQ9=0COuNQ1? z^fiihIg5l$UbV=z{j6OrFi+a;)$Wfk@E2Ehx|;zd#O{!L^t&@y|7l!kP7Yz zQG}=C%bKs1xxeZ+q}UQa+y19&2YeM7^KgNXuHLi7Fa64R|MuZgK~MVU5l)5ZDddOo zA;vu4-AxkU=6z;nHMMf|Q8oy-&0=#TnnX&vi^M_LGL-FzbRsrj?bM10_<9Htx;kpC z_OiYDgRh=tm00V@ut?BF6(uBj*p{4OE4dKSPF-AWXS?)A>ivYhVSluMuIB(UBztC ztT=jkf(55?_hPeU{&I&i4`2dp;>c{Ta#9LT!zVh{DY})CY;rc?})TSB|Z>XwPFv$watD zb^|B%bNgKa%XzqX$8R6Q`St1~a2TeP@N+CqE)tFZw*!+u7h?bJ_VdYwlXB>@V5cUw8`!xJA5 zEhf{&;zCSs+8+pod`Y0QPO5hy{B}K~X?`XbmN#4A+t7Q$37;ZYh6c|+ExCJ|G$4@b z_Ohj;55Vs#?*^y^`2lkL_f*A#I-1d5R>#wuH4o5daB7Ji2n=Gnn4(qB`?ugF(oa;w zW^Ycc;8b~#pIjvGwCRV-P~!S5QEuSI-ff8;`P`?CY=j>uk*2MuM-X()jM6cX2A>f9 z`^4c7-vCdxzI+((8XPns?U;WvY5(iF78+uq$+ZAl=w2^v7iJ9^M=`(EAHEO(DAd#{ z%FRqM9w&u63;yBZH@*6P@#Ppkd-XGeI95x;msppzRm=avW9#Fm5a8~`nK9#UCcwW< z6dG)K*J5S7uOtzT=;agtH#c*-C>mT2mj@GyyDejjU%hU0!Gi1?9jSx5(uFhv2a0>WY zeOwCPap79Etjyk)W>YHW9@Qyzrh-}vT>qT@&%HlovZ&$_bM5sJn}5?}&t}}ji!u@2 zvR0#jUYq>E67ddFk#1e)zzTBA$E_0n;Wq>6c;y_DRu#8SWD$wSe3taZ3_$Lr;C0&? znrhC=$y4)%`SwsUa2J{F%8xkRTqp|{x&E_3370U6uO$ApH1XC?)TOig96||EvD{0_umjEO>S7n?r1!a{Y4s zvF{0w6_&Glm=gh_K-|~Jh#L*Cb289$Kx-tWV?dZn{k%F8s()$!; z53R95`)_lD>a+^OjhF)UO3a@*pA_z8wN|)}t!w6CjTLIt*==ESCZ~Py2;)9>^P=V$Rd?6tua@StcG-`-GV{iv9v>@Id8bTIPlnk)0 zp9Su9Ph7pLZ)NWJ{QvRU|?Az!QQe@#Pnk5JmVKW|Z#KvaC6GvkB)P2|My_*ZfgU^k0`vLH=_H1J}wIIYD&}P<)X~T9pNM zg@5;4aBQk^G@SnR=3Vf2ymNz^o#wY%rvm&Y6w7jq6u?@4eI^uo#}SzxoegOjbyL~V zY`_BDVF$-}WGEnPUS^jg$fxYJgRA1{WDnu58^5^I(Ko8Pmz_S*Fm=k9h&Gack!z&b zn_U#HP$trg$w4N3-Fi(yr0|OdU5)In@-_T( z?XKx^R$Q`?@2q@7s;g+rWDkIQJ!mlj^T8m*-Tu+8@k~EV2!6rOWT&-Q-}#_6(?Mn} z9yNXqk|t#SKOV{530Cg$teTx=h!&3;WNll2g0n7$j*Th3!kK<3<0r(I9ToaFq+Ywbs$T}+vhb~~Z;rk6EQZ&@zdI)tQQ z42Cz-bbPVX^yu!t&mRAjrjtVQQIGomeJ2=s3;h;=-+8z$wGa5Sp8c( zCmyXt&dV45jnz&@g>gGQ6?L1N9B3E_oP>Lof!<|;e#^X|`P5WxXl!)yY3|8rd`3JT zshs1JT-A?zPyTD|RDZQ721xQcQNcajPxjg}kCXhP>BBykWLrj4G(%bIz7zrP% zPXP9tcZuJ5U*8jFrD||Lv>T{}suc-$KT7a3^jdpPBdgDBvmMiahpKVSwWpBzS>c{< zth0hyp{dg^O^7C+l*?qV%FSO9E8hlgc=R3e46oqkI~V_H`Glo&(!L!!Lv|;bH7bdb zlwk`?HS}{(JtjLz$hNk2c+aAo@r#v~!`GK5?yc0iXLX{t?z4^PdtdBw+vH4nyn-_e zC+^?Ve$g>m#@OMVnmd6nK#+ni^lbg#1?skELrCod#~16*_eGI`rv!USO%`Uu#r2gE z(|72RjB)IorWQjaS{3f2`%2QzKN+=xilf{{^Hhqo8h9nwAs?N=2%EQuDXSubeknWK zjdh6a57a`qXS$sPDbZQL$wv2sTK(0qRKAC4FU0~rV#`M;M=0+SZ%!D(hZ-|Hq>FR( z6*|Q0`)*#fyqy3ZltQ_DqKrwAV_j3}2rF%fDyrjhXNcr}alVuCVs8n!Y8+ z>uwlR#LdTrS-q;Sru8!#B_-y(LdnG4i~m!HeV37Oo#FiBCE>rn>S@*6UIkzwO#CAt zX=D?o06DXit4Y3Tu{k=PS289I+Ccy*$_OuBeqk^jxoPchH-Dd_eB*|XASTzU{nN;= z7c&7*#}ALwE9ILL!2sQy<_a!!{oIg ztIyNAZ+@(Vcr#z;7`NtpS9qQE`SS8KG0xNIJZ`O+#Zhy;R`KQxX)T9?`56Vly%pus zVNLvi{sSZ887%j4q)O6NXZONmf)6hTqtp0a;muRLzT9yVVXWw5zxdpFIu;9Ke8m61 z*t*MrD8sH_^dQ|T0@9!YN=tVtUD6%WIdpfbv~+iOcOyu5cgN6O!<>73p0oEk``!M& z@`;hTuX|l<{nukYe`4}%ks&76Hb;a-`&;Z>Yc)QWtBOd_M(k7JJCBd9P&c~?`}2qQ zuwqNA_TwzK1D%x)N%qANcHzjkl`_{d8Gk+Ad#XuDmSgp=>d)2kncKzdl$?aS91dDj z>i>=D;G~R42HqqUx>o($kUS{ORVIsbRVZ5C0}vY5{b47H4c=^iagT@_l6l8}>)rGh zu8rpsbVAzjZdB*B;L+0w_?TR=Cx61&{*=~!Q7F!T2(fR%$EBxMm+|c!2Pl|bl}^-% zjeTi_78de+bILyS9Z^at;E<<$#|3!1%-I5RUc=NOaMT%hf>-APc=r4hD$s_p?fzL! z=XO+DQ!_fsGfv;E;aFp4OLW4vY|h0+#SQrR{7`epm^jtjj}c66TXwpNe6?{p|L&|M zZ&l4Q-h#u7ZzieGgmM$O{y(!+i6=6#lt-;FoqB%zoif4jpnXdH_%?WJ<7vZ9s=%V?BJqROHhE#fNJBryYxuV%Kvm-5wK zj1P+1&YqQSXo#A<8UL91&G*M?#-IEQS-})NvT!Y8 zP6eIPW_72}SK9Lq24g{IcI*9qWzBVmv;LNnbHB@} z;#ObnKmEOrzI=~*fICT(i^I44WP8~&qTe#?oOH}`Hg^j^F~tSMOuAD$%IqhOzgpC_ zhFu6uK3r8E8qD;dPr4Ue%Ym^9)-r%jGk)a^Z7=)l0e!FDhacCkm?by?A*}tQB)kr# zwafpC`tC32iPbqOt>6YZ<3O})f$g`CtEV310-wATA9AlC*B%cskq-wAnVEa0Bi#l* zko!p4wWaOY&%OPE9yiYrkk7K9^U`FGF-L-p$=a71w=V~mK67cldWqb#={nrVrOMWk zNp3&p2>t>@KK8_tWX|E0=%SH2!BRpFuXO%dOmZ|7LJnuNtRy z4jG+7{v?O86Igc7a~Lj#StBgkz!;;$)(`yP^0p*a;>fp>fq(xqx>w?b{GuT6qR(&U zJbP5SKv*28j8^pe1?xN5_?B;92obr3C#t_n0KmYpX97!{GZ5hq3J~uWpwSHAY2|;q zIa}&Evsr8LI=$(FjXm6!Qf?Xb4TQx|MDkr2A8t7I9{iYkWyaOKu^*gc5IT!r#-JIs z9}7Ul(I#&cuw{@XpJ%Gn(-*9iZUBTBruXwjy7}+1j+v&9>FYtfu5V{hFz74c-Z4`~emP{!|j)=+d?-yQ005`;z*D3;x`2tG_Q2 z$y^OE*U{nqTJL8?FG1LHI3UaX%hZqjSD7{y0Hz0Ek{x*PT!^#F10s2eC}EPixaDAW z8u)j--&Izr8CLY^GL^~t%K$Wwp$C=o@Bc+423BYbINeY0C{=da`fa=rExajTY_T}) zoLL5F_wH59?mDE5SqlV&CIATbU60P-R&bkX9OZ>N+8{SK zOBg84?hFQTl)Mu!9>H#N&~m7b z&wQ=J)Tk`4u$mTtSUXe)U@*6*li}jUn1))LX!LpD`)l<=4S+C00x~0(HpP?I-Q(G< zFH`%SNdF`0sY@%4^j3TyLwEpc8GymH=vfhx$MZ&>Fv@pe+4p~m=goc}Zu1Zxh~ z#aezsAZYsg$4AZ!mxIPiZXOd<-?8v)!b@+bz1i6N>f!$O@^p{sLFqrAN{+GAP;C6C z>P)NZ0h79N!~y6s{%uC-|NX%ZI~r|3CBpOi6b;osJO7!-Go%Emu9miHyu9Mrw;tEp ze=5Ap8I2TL5`*alk8K*P1K>D5!E~g`{Y%Mu{koW1S+)@Ux=7w@9rEM${c0p!oDfVQ zU+cx$V9!SU5<`QQ%}Zl40Cvg0Pn>C(kj->FRe1duUAa%@Dp52{R8%wozJC5Y#)l>< zfZxB{mNvgvVfwulOfJJMpHKT*1CSYrYR$%9{HK4a%Q|w2tbxCi|HSU}kr?Iwd}h;^ zM}P`!rM#7+`QNlcTN}-qfKyzo_`k+@w{`q&03wDv{eD#@N@H>};5mR!;Mcb>e_~e9 zaO-DWG0@vKgDYf(m_o_v59@4m=H*k`6CM6%F7B3(vo~c7GN-I`?|fIR8QWO66T~=R!ZMH;sPtN`7KbE`!(Wr1L1f#$)sZEykK=Nv=L9 z&rSL=ez(iF9HAL|AKhi@G=EJtL~}YD-}c`;JC`f}{q^=-OqWIJjCx5`hbLJOHah7W zp3dZSOZV~vv;j!WBMo$z0No4!Mq}QTa#$q^m~BR)#!2IIyih>(wC(Mrii!;Gm3G?X5cS!j4TJtsryz1G_Uz~o@zQJMyMn}}IYABvaH>u8lyQ0;? za$uLb^bf58SZJoWqtvNLNix(i1J7=eI2Az5=^2mb6?7juLB2>+CvTB~xIhh)tp|ji zT*BTKVegAXIGkiQe4g`J7pv*^g0DVCO8^Il7N0~703kPUfBdhK95gLq@qyhRk480B z-R{LjDNMT@fiiEQvx4=VQ|WY zwSUpelb5B;ItI<%Z!D%uzEz0vS1ubg_1L3aWi0#HLkIE>`0~;fuk345?sB$ z1{h@A3Y>baj$frEwd;D-v1n8k#l#l^&$d#7D_+^dB zM&*SXl$L;YuCV!KD{Ftg%s}v}S4#&N_iC&tBgyc1AD)8cip?F_J z7-_ zBY{*?&EE21Op?4`;$wSZz9Mqz9|>()F0KZ!%yZZ-by_22nY1NeTg=iyv>AbuXn_}F zNy<0Q9LD#8NB>SK#cG2d9&f%Rpo`o)P3`87$PyPi)=w@-nRYGrF84mEJ-s`$+Ftlu z6FqLGnPVQFFC27$?PwD}3O`*!c*XsCv1EZ7ozj-;F$j(1^Tv|i zhY(npXq*M01hz66CwYz#v0F};zGBNE9D(BjGDVHaI;LqP7$q1{N$h%~C;OpSc~4M2 z`S5+B3HIzPWcI>uyNw5yQ*-$hP|#4PA`ldG3j}wO&AO^@@g|~L8I0?T@G^bmOEeks z>g<0l((44|nkY=y7g~SQ`e>#xUY$wxcVQ6l{Cr{5&Z3vZ^rbvpo45%d8=sV(M;(y> zMUTYOj`_p1Okt7&uE5%z=hh`w;c|t2#~`HBptZ*}HBNw40(@abxXc-oJK!=EAN_ z@lw$m?6b_N0683XsH0Q~q#RhtLLMO0~lvYq8<&eOk5MzX=F20=WSdNg+;nJ_6@eEz`6)G}d z^~#8SL6c7I%9zH52-Ft~yH}$JYU^@7dTXCXqt0(ae(yu=3o`vm0CYy2e{G@|PMGl# zH{_>T@_+->8w~~Lh9jGW6;;!?+nZP*ndeMpqoi9+2dp7qp03T$E%uX=_9u=875ljvs4Jr`jt^%`_&e-;FO6)X1BX-(B4voKBJJ>bkmZ z>>$Cm4cIFBK6c#(oT31qdv`|G=CYJ?^H3L8ucz^!taoH|X z>WFpNvpnix3RkO}`J*bra-Zhsg9jR0@>79TFTN|4gOW)#q=#50Vr}s_hqZ#mbN2UR)U$tr@EUoHm9zM@d&42jyZ>E-Y*oc~S zqa=?K_!(zUf?@N`Eu%)>JVqn|5A5;yC$9i zDGR<`^YA=9Z}ahK`cD_(dN8q#P#wYywlXhVPIK5HOEE^`-z0!v4^@aXcET4y;9I`X z_OalveA)U!=yk~~tuR(gM0y0Vuup@rrE~}MH%NQJKYQk-=SKHz=yB)u*96Wq@RLk4 z%GbrOaT}hIYO>$mMLZ($wY2E{3oK{Hv`kUM({|@HT}j@EviXc#P(6sgeE!ke z+*9_@aJW{CLq6-H4mE&e*uEoVG4e%3`@I$gB3_zD#y5x2Uz=?&^8%jGN`Oc37T*88 zZ~`3y0>8ez1I|Dv&wDasN$IRRvHwZ6Ype|NNk}fSEH7^&dPurrJurQ-Q~=6o_j;H+ zFbPZfY}3c3FsKT^w}V}WJ(Cd2gZSFc**@@6@lrt(*ssJ*BSuNm(8LyDpvyN_*o!2! z+^DJ9Fci7+5*bkf>$QK)qo>|6w;Ov7-Noe2nLgT-ZU_Z;wlJ^8;U2!$D|kaQFU z|BN5MC_FQaVvGR>14~q(gx3FeymY?B@HCO^WTVYrgbg=u`Yf_Q`M5N zR@J&}Q4{aU5Jul(rzoyKHTY6~RFE5rI1MhEQ>wUy{NYxjde6@f$C%DHx0#Db#BTO_ ziiJc5Yxt5uE+JoDq zcXF#Ij%DzLbQVYBsKKhgj;(n1gz^kkW>W|gxS>;FKO9X_-AJFFJC)up8-BU1s*3uQ z@hw$={HAod-oES#?5?J5=KAOEgQM5#`tmgRX4Gy@7F}M=%JHwejFV~~uvA|z zvK*X9c}RwIA1(q}wpv|8pn29fv*FH_N%1iI1 zy}(b^fXX%3j&pCr#AesS&7#n(6L50{xE(nsRE!{;@fi>AJ`Oj4c)k;Pe!Zp3C4kBE z46^>%Mc_1J6<5?M`ra=!?Em zNIteKY3F|ZgmummP#Mvqeqrshcv8E9wjC{=zQ4Ct5+odecHEkIrk~7k3r{ZnQaTR9 z|BUIXXtF>mgVs;asRKxl&GH_5;0NC#YNXQ3;iJrd4?SkuiwuYT3mQ6Q>gaePlM@iU zf7*74;$?EtJ$V9EjK#pqBl8i$K^z|sZl;3plvHIjJ|kIt3)M89bkbU|Hg1Tk>^F`a zA|v2(I3NDQN{}VrCUE`m6b`f|C#-6_JABf%Z{PMDAZ*ojZ98-S)_l=(8s~krbvR9~ zzuPo`Al@Rh)sJ00JElK3ofC_7dE-vJ$lwy9GpL1JeiPi;U@7H%<4a3g($0E{C{2#& zczcY~5V@@F-h@zlGTv@NpySF9HSWX_tGph6$i-u1UNG#NzmLlV{~`lFFg)44JIi4` z;%;HEw_cUU^%Kfpj>PqAfi#Tj@S+C~6LPy~BSle5R!AajC|~DTJbpM+{$w!;xn_n= z+H<%i5c-s`(&Vc7*7*%Ny*Lz|-I6C}UC!*v5&NC*ka^S6Jgujc{Oo^_w9=mFcj&6#4N}-LA?_gv52$Ls4j3<@w189gS6I zVe)yrRAeu>t0CW3Y9qG$lWY7F{k;Z@QN@sMLL0U}2x$$y9cg9F(Ps*R^%v(i`3J#$ z(+4|)bm?-&Tt4qiFD6`jW#HFS3|F9cVo}pV>c|S3DKvhW%Gy-5ZlaPH#dCK_L8`7% z&FehI@Wc6M(aRQM-;6dr*O^#;>l;mxT94uT#8IJe@(}kSfoeVUOIR>Rfj8{L^Kg5!iArB?u*f z&A**5jH^{t=<4w|PF#!45$j~Rn_Z6|X8CwKDPO`g=Df-V{3cA@iUlZp_lq{Iz5m~Z zzVEVtKkq`iJG4fS*a@b!;UMw1TB{(pgLA?uQ<;pAmH=&D67Cq_a9{I;f$l}Wpjihz~SezWo^R@`p zp@5;ISMNw^0q6w9x`pmpw55$ro4dbKqB2tfPfq-du)U+X7=ED$1SpKbOWq9-n@7LW z1b~`(GtP{F@}mM4>{ixdKK>}Wz=J)=PY zFf zX}i1@VG7hmZG#ITCb6+?gnzO<2$2@l_E#?e&?G84NR51A7k0@x5nBoG&<2hCt6&KW zLcO0MPa9?_Io1*N_m5i|6*g_n*N5FniL9SZzjsy#7^4hfzkEXgKxx`?=--QgB0#Ps z_L|EnCAvQ_QGJezdV9+?3tzowTn1r7)}6k~x|5_STwE-z_^N8&u4P%#(YF6WeufOk zYajm6L7u*?nLJ-}IJKh?3-LsK#o2?AzNmW9=22$ethpGxX(S={#LQaM1uah)6<@L9 zmwzkhJs7)CW6=FBBdV#1x#r+$L^xr?PqDXEi%M_LLv!bRCgi=aL(}e?vYc_xFJRw@ z30v+b&z&i8Z=Y0jM|A9M0-^9n_Q%-;?4g7*`Bg9Mr(MKSOsT-b=lGjCnT^$E-k)=H zKj-`eWr@slg)@r0@^6yRHE^q`jbwdSi`!H@N<<)I*Mf6UOy1cO^%g^hH}!N@YV+l6 z9yX7;{<$hI7w+!^2=1BL94F%cuv6Mu+3#!%ug_RzZDb!-R&SEcv}H(LhBr=@h*WG; zGJpS0Q{k^yo~{K+Uvkgy5}89&qjvRzwEk+6OT2lna8%MNW{&NtNn&`vsO?xn@kI1B zc8MCBG;3el5{l^w9*h@}H&ER9>$6H)K*b7cjMe}S8@PgCiZXfnKp_!s{T1`n$ z)(eRr&x7_x8Sy{S_wYuvtEf?p&Q)ZepB=q= ziVv=p`}z}IEVv$HV7`jWR%}=0+rpD?q}?}x$7o9k;D5O}y!5{`-_(69W`pD#r^H)p zB2SIU2(m&j`D9!Z&zRwZ(}YfZ+9 z8n4f1`?bg^t!)?AYrEFjM$J>FxEf!%y}uCqV0BW})W1u1#&_s8eaW)7MpDgfKtFR_r8fnUH(_ZS!YG!;izK9I$2bdG`&-R#@{yaI1!G{qB)L`v?;# z**KkN8$BU7?-xV14y1V`Jt)<4<^mufg^aOxyZUMtQ;Q9OhgF~T5)=5w^fxh8TI0I*WV0Q3V ztWfWs7kZlC7UvwXIwef1TPB@c&ksTpX(XdKsArfKFjy`G1Y{iGF4gce=&z+Obv)cZ z9a^#57LQApbbL1BoH;1;xB_34X|Y=wcA(1;scNV@vGS$dxh@6hRT*jJdk335AgkAy zEL6K)>R*@U2|SY;47B-1w9|DxSBoyGrb$gnPvyDfS*{;JVhkfCvJ4x_o7z&dTv{z# z41#R!mXbcv%nD7^y}RvB=x8IK#+_C(FAWjUL=bD@4p>%cD!EYEa9C=+ZgN>y5;igE zLt>sPoox@UZhVH?3$B=7I!;f0lNa|}a&0)b+3I?Vx~IGpt+eW?#v0-fdg||2A;a=d zAv>>bDC?rL&g}Q)uH5t=YcA?B7#;sin|`iGP*EdBC&jw&qqgOf7X!)~*RH}T+_u1) z5`af?Pm(088S=RcRf5?Ny9nPV9sxUs;`%~#-YM9uQHZ|*QV)7?m(A(QYD=3z zzytC%*5SBZPMR1mR;6PDAf&C}CVkNZ{ULSVA$tt1(6wZ<>{o++L$^)P3VKuvjf?yV znXBz>*9wY@-wwU$J=0vrSI^Kp{PW74@pw@!Y~deT0OBL9X3#-I}I;7t+@kgJ1WY+W{1f?NpvYbpRt^JTE>%Ga>y3)(|OG9CVA zcSocrpgH%o4$OMb4o-rgb}(_8jbrl!S}rF{=4I|^43GSa+6gh@Smii?FfeliXFou7 ztJb>8r{I(M45vP%Ioqv006;U7VEaJy`tL!;lwg72hhf>o6lMYr`h`lPp4ZIU7LC)k z2joBIn0lc@mkS})IL72J%Isj7yTY@JrF9yXlqX-8LmLU!@4GL9_Au0!ysU5QRt%zU zKc0T-5w6PKSPZ`PZqmIV-D%s76k2KKAU*P@$nHtIpTnRuHN64T6BK4?vzCv3|G2ET zubKEQnaNTlUrmgEe6axXZEt`np@K{FJmNQQp?jgTB1Xt^e^YZ4#22q~8M;DFbt{ei z2FG9ex|>BbHB*e3E{&(nKx%NfGKd|{dKWI&MY<&E5Nl#6y-$oG_*JkWI?WQWa1(hf z?7NVm|LmvZ$1}6;p2{C8`g6sfN1&ZJkw#S2uNS1bbS`@bMZuZQBkpO3D55~3$=qU+ zXWH|1k}2+Ki)iPLukR)N&h52WQI(=$UZ4J^IRH)ZYf-lHtKVWY;c`S+NUJb;TTQ*0 z0Jv2ZSG5ftq>lnRp3RNJWU-TnFlTa*v@$t#@3|hTHgQ5ev#X-`WxLlqn9; zfsNllStzT~#qM-duS(5K@lWnANxZJT#IesAhA)-M{*(aU!r)%hj8@jnDqToxGBTV8 zTyZaP8;kq3l1YjnrEm1_ncp)Mw?eC8?TEj4Ou0>U5Wj9+Km?RuDgfYwCt#vj_>J8M zuPUZ?I);6cps!bLd|DvK&WWY08m86Qcd*9Zn$x8j@mY1i+nctPTjw<`)H5`IpBhJn zmF|L2&kU9tr#6;WcmolLJ@^3(uBhmE&LCcsyro&ma3n8wES$`eLOC0Sx!c;oFZ-zZ zt@g#T^$NhiCj;Vja|<8x1%~VDXuL*rth2J+z^muM$A{#%*oSZlB6wsJENmhY#Ctvf zsH*M{qf8MZVUghUFEZpJEb`hGa`=n{DjXSav>=QcV~qyknX2xXUyPZM+3NBc8fNCY znwlDfom+SED;Ma^6j39L$Zsvck{&_-k5mAOWdF%lOXR75_k1jQq%LDag&@$MaMsT$ z;{tWahn2MK2S1Mjmz`*^H)M$gR<`r{>Q}9u#Vep-_$zkfCC!*M-0{Tlm zwe;eV>za%mhIVJ6auxff0L+VjfyXFFMgGdd==bF?ofOEQQCTKoq}KM{q|kAwSz(_v zZeg#7u>wVC`Ynw<)^X3Qdyz||@!#>ah%wgGOEUGq78n52f0N!mXuf9niT_bmxwKII zRjs&1}U610181_SUQ$MX@$C8m4IjZwcG5TM&5xUPCl6oKWa0cBi+90 zM}C{u5!aOS`pJ$IJqz!mdn#1T*R>za(5ygJZGUd?nGd%LBXb10j~r^oMg&m)&6|D_ zXEi^N7dR|f^Y>5Bpuum)2aYTzyXqd!J8zkl*_X2R;A7WmvfrT!{%R-iVtODWIdI8c zi}oH*XupEs-Ph8Yt4)?8KH$R$u@`?k=1iS`f#J=#=0g(MOyFfH~Hh;U?}O3TX>wd%}Yf z%&l2BR7iqQZCq_PE}VV`}2cjHg@5m26+cqVc{||$UQk!IFI5+LJevB z1TCwO=H+28LSjLpu9mN8A)EA_+ZK6lD$bCr?G(N8v4lKfnv>)keO-35ss>rheu~6p zCzJ)b^I!kLIO9kv#~+<&^fU#00Fy??)1w-={7Yoc`~Gs!BJ$m|!E z+jD+OG~4;_Eb zf`~?yd{R>?rwNPYVC%bJ#nbn11wp=gut3ejQ(8x3Ci>@rasWq6|T)%26S*hR* zq>M>D;kNbYIQr;=8(3=wgH4sSKB;pPp~YIVVD_>wFetx1$8ryVgYXwPxw~cnpsIE1haj< zkRPGU<5ZHvdGwtSwwTgmUXt$$4wGlkikC%dj&+QFtIM$vpP^t70FYIS8HN8Pg3sE8 zVU8=3L2~X`g4o=)oEB~@ef~nO==ttfx6=C8hjL2ZcS2nX1~J_dg$g~2f22rhtx8k% z%*p@yC|+^k1a`}FvsNEdLN;{Lmi*2%Ej#+ihzswgpTFZ87fzJu)jSJiJ}K&{K*^s1 zBLA-?Naun^pj+x0uey9?4N1YCtpm^3`MnkmoEuUcxz89ydoZk|J{Z18W0_Iv^}2f- zMyRF4S?#!don_g$ambcy{z?BbU?gjFfQJU08mPYn&?=m1Qr*)`+t7yuT;t55q-eu1 zH<{XxZIiWzyAxk?pzE5oj=N^TpH{SYJUr9(R)pyhc2r+sBnS@9b^+l1V*&tO7RfVg zCFZu%TZR4Ka416Qt0}tp03<+ybBCli0!lF1TX+4BR^&}y+c(j+R98~_aSG2M|Jqiw z|8aHP2;93dd6E~YwrCz(Fw>aG4W}k{bO;P!!&KVNb?acb#F42T=1xl%;a*uS)|ijw zGrxBgI}6jVl&b>-{U>XvGlds)(%OVzom^`^nUA(FKbui1Ncg6O&ROjUyCKeoZbkJ) z863Q9*Fg1L~Wn~sjvByv_k6K}+yb-kM3vEU70GT$jh7yYI9_lk&k0t}U3 zj7pzfodf*br4Jd38n)F^*ArG=Un6T<@G$nAZG*#@!zpulF@eLL>4H zbm);2u>~RA7mwb)Wk}nUJ5lp96brEQU#R?pEYE>#{?f*pkB4={QlvTTPsb|=X z7k=}Q*k(A_^7iUb2XhoonZdCw&KgOCzM0Q&*qHtm%6#j#p5xTB{(XOegGNng{3IOX z*2}kC*1hnC?y@JBlU;AGqY*hqJkpxih;chXV6PAMD?_RGhrWhVJxlBV zTa_B`OBJvFJbcIAjryJ<`0aDx-1{BmL9AF}I4pAs(WrM{(Gyc(I!n(qHfp*JG7p7j z{E9QLG!AMeP5%;~tx|qM3hZhPK&_A_1p}P5DNGbGJ*T=2Gp6xJQNJwKkWmsI++SXy07?pe0O@R+ zu8aGM`A1tu0P@;ad%ss?fL=9##pvJCKarip!OuXG3;!qpyWVE8*&Ix*KJF!ai?l-d zt3RNU%6&EA!rYrUPY|1$K7!C-01TE2ll=oNvu zHi_LURo*gZ&Eu<~lmeh zd^wn7*LrvH1n9e1fDy#$I|I!2DZ23Qn#|uo)d9MO9E{j&$3g9Lek;veh?-{(8>7VLa zgMUQ8k_q+-J4&CcABLn7_#Vdk3j<-+bWe`7YCt3HNL{7@W5^jn|IKqjM#4mY`v&gc z-tY}(%d^t18sKUay@7VQ7|#62BGUWoWpUJrE6n>*^@}IvUP-#ZZJziTa=V ztsExH($d5C?~TaKc%ufPBblP}cK?bdo;zJa`=#>6E6%$gC+bZLZMhOlU{waqaXL4_ ziE4}2)1q*r3V3lnlk&gRh_7#MZj_q|M-xG=QQHs?6n3t_n*~%a zR2Y;2`jQP)EAfs2;!0$)Zfr=wk;DQvcU%h1xsH+tisk}4cJAm8p+W`i2PR4w>Xo4N zX!wDsnKPRJ-VuYcfQ*eW=mpPr$NKI~bT0~r)zb#3mDq2@Chk_BKV$oj=f3@6zk~*(FU-mwV&2EqKGW4SM}N_Xs_=uiL5x=*K)$ zLu4O-)dK}O^1OvD4JWJ39tOKZNld2gf%G8&l#i+m03<3y7+1cF^S}M?Hx$#Pofv2Q z>OiyY8jq|l%mk#LcoZEG7eUq?br97?$E(Q+X>o(aHuKy36$E)N7%r__tCbojgX6=x zlDTrm7-&HrHz_TTvLldutJz@zDQ4Fubc5w_**Z+FDn_ z7U<^&vts_^SQ!fEHC@P-4bsK|=wf(qZ+as@J2+ zyh_J1?5Ohul}!5N%5`NHk3JaRAYt)ST5=$6()d+Pcgq8YAFzFH)r@SJ@L0a1;ymOB zo`?cxemKS#*EZuV!l%d`+9#i{4G;UL;hm$#7w@w&Wzk(aIpDxat=jgICA#Ofn-zJA*RKTkL9D zYllbMu3mEphS58;7A8wwdp35aWuQMmlSYc`%lq0{#-MCP^R)Fk{c>;~`*#k-x$>IIJhHZ6Za3zFZs@@4nv>*XIsVt?a8JO`^O7QH4S^@AMXmFbIlXs zHbM<{D}7GHD3ZD`l^dhMyu>F$f|1^25We-JfEt`p0bkWytrd_HdFz~8#icIFw?8mZ zr!jyY=rQSI&>{7VY;j6vNiG0UzzH&at)!-KQn>uB>3wbD_;-;|w6lTpm@OLdzvMB3 z)E&d%`a`psiQ=3trzWud+)k(GE-U$Bx|K)uv7~GD)xac;v|j+p)49i3h5&(*1f}ic zoz&!R1Pj{U$GM6GMm~5Ff$fAi3Qj6OMk-vz<={5#yh;B z4-hay#$9j0h2ZF9ut4+6It+ogx{<8{NJQJ1N$hMmZ)l4~l_2<9TQhkyjG0VDc9WAh z!^OKuJM8}AX=-GihZw7f)b1i_fB z^;XTEV7O8a-e!(s_~FCG>$9}TO3z8*xscHHl^f-l)Q%m3g|FWZTHshXmNgME6`H>7 zpc{?3R@r;Oy}6Opb#~*n4YHWb>u9?|-=YmGVLk!#@)eJk^RygV%>!Fz+920L&I*9D zG0-t`8Ws<+qyBKfXPiDhUn~gc)!I0D-!M6=&+yZl?yB}&Q@Mi#*Bed3s76 zrzL1M_&5K1wg-2 zIx3i#f3<5qHkDq0bsu?2z6V8f(H+`_>8bx={b~sauJ>JptAihpj)49+-G7`t&?*3t zzFji2%Bcz$;#YtPW8GLg9>Ur8X6jkEnUxju|6L|rzoGjB?7TH~&8WXzm0c@>*!oi- z!+Z3KxV2q7?j+78aK2}*wL%W@k2<94A-wZ%OOnH-ZZ>(xUG<;N_t)Zwk<(CDw<%|t zrO(zVVZ>ZE>y3VS{j#blr5qYqCk`un&0`EYM65&a0SMZ>&CZU;sQf3_jdNRMr53Hl z=f!AybEXwz9iGoE&D~*jth1x)7*sP-#c|3Aw@xSE~(WU*rH0xjE;yKq^a52r2Dp1^K_y=}#0F zD02;r&x+y%l&gVOU{q2nxGX?^%ncuVV2UMu_rTZ9W9wa+Yo(i>PQ#{p)Y1o<807$B zFEVzCR4gs(z-c#klnd=k7rX}OUj6A-uRQ2dt&?aZ$0-}TJ7;-rplC{H<*QRg_M4;p z*=bHs_`b!5wrO_C=b4c0qvJ{e7T5i)l(U|B*Rn94PPFpZw`$W|4kKlKaFgB{rXm|N zn{V$0qDmf%XLo7NnoLu7o-KWz=*@AbsPkp%z22_?w?QkEL2jgbPN{RRs^^S>wjE$} zGGPyX3?UswA2NK+ItPv5{QYmO&Y`HV7$CT@9oE$a>e))}XOFauicJKdQq@SCBn$^% z&Wx!E>XFhql}|S*?wvM3%=s0arnV_gR2;io$j&VEjf|F=CHP3EI2Y=M1)s{K`=oJ1 zpW?}Lf65Q#E|E+zH_0YBMorE7 z8Sa^OBU~ZV+)m{;8TyhO(rTG4mE@SFw+2yAre0b9Ju?Y@nQ^hqKhP+{1D<;nj9oC2 z;;A+&VI$d;P9Kvj#;QPmN<2w>2K(aN4&@SSRm(rT+;qwSNm;0BuC_eQ^#}yUOL7){ zP=9#y+JX;9(w;Vkan~snXQ^wr>gH(hvwoh1$MCSIHpu@Bv_3OblAy-)pNj;XVcH}W zmJ-AZSar?k&gsE$DZ-k!9?PBKsIGFZ8>-VzFN9Bl;TwR0L&oR&l7rjEet8>}j<@Wn zwi@?ixtexGqwZAzu*`W|j#gF;J(#CKJjIC>Xb=O zv&PIK8l`u+?TyPhhmn&+Q>y)&L*0uj@&d%yUz_mlSF_36@2bBT!)^j5IU5m{tmBvH z8#G>ie*-`iO$+A1a`Ip3s|hzH7G;9!KDRv?%ugFAg~%dNGy9<_z;?=a(W9OctTMHd z-@UyD2unWxzM(p%^fG$*dlkJ_z5;AP0x;We{SAN5+1>bc3qM>75m+m?y*2AP=!A+IrYsohFC*xYk6>`I8$K(K@Lywsu!=k(J zHk2^Tye`fsN7`@aOte$%to0Q(Gkb-9?zm|0lzSE2Ia5o-*_-Z^`Td$@*P44Et)T~| zRG1b%sF$}}&fCbV-#ebhdzCy(2Z#I?8jvcit~<}~lRNY7bn3CJ0E(>xh5!G-m7V8; z=FJG9jTG0ryPb*HNL$x=Jg}On(5i&JrsAvM1Lp;m$4@dATg?py@51;F*-A%Q&oGE4 z|FpBJlN7zl$9h_S)$qRU)6rX(d_~815^;G3)loijj~5iUf zi7{>jr_;aJ5}J$>7aRM)2df_Q<<6R&9SfLc>d4y;+T@$J!V~242gl@%$`V zQOY)V_v@=eg>7X3A|5FnkE&y!_W@?tK@%6#>!{nirFm)hu`^$&Zud zljpaS)z;CePLIQRElJre24q^yEY^jGW&Zj;f)Ko$Nnq6awPe4*@i0?wYP-t$l;uF# zg6wlCwmWgyt)=8(Zi(VSuyX>-!L$V_XYGqNQZbsk*uhk_ib?addgDp+S@h4PNbV?M zjOOO?k_C$98FEuB2aXnL?4=m)U40vx55|*m%9<4`nKC>O%{{C#;nUJ{Tn5AO+51mA z^#dZuK&vU?bo2LlxU2HF{bosVpH2e?*sN<<4#F%~m}?PMmiowREn+1J3^%i^QK>C0 z5|xjzFQ|TbUoiB#;s*L4jIhDXYj~I*r`2st^orw>%Lc-`BU~Zxccqwu(%al_mWD=J zW87#ivF}E%Q^1@AFM#Ein2l&H-<|&){qjr3uw*Q4Qs~ut`HU5NyU#GI77%jXBv5JRCXJ5Xv;~4_#-2o0vGW8q2zfz0okNTOP*P0 zO{HyCEc1k;HUerGu_8{3ha|7;pz;?tFq!t+uu$s1{}wk+o-&Ocrp@q{xY+7kHy=8w zn6WettWyUF3#$PLOokIb8}?zUf3NEO{u9Y;X0kX>1w$@3O@cQX`C;#R44e_NLz9r9 z?fseqw&Q6XhkQk_*~bwDPY1A4;bO1gmUkhwgLlVqZoRpbJAB?o>!EVve~P$<){ZL` zS<+Gj@Viw%ty(qga{PJldWy=nY^+XT?!mVUDFC?M>73{VASQd;JTx7cDijteC}mMw z*{yB1{~-;&BAT?6U{Y8qX;yP$I{Me)fN4QI%Ta25H)Ll4o}Dc#Beq&;b-J7#p&)+` zKnV3DjrNO<$W)uy%=|jSBCoxU?Xqss+jf0R-}T*+CWui= z;jz-eDzc85%dq)0ET!^lBLab!Q+VLxfx`i@0`u?N__Y@{G#)nx=XBJiPRU6oiE{p> zcArW0#t;rdvzxv;K%X1`|ASQyK5jmckGCF(95)s~iVHZePBIFhP{Ks|?gsTK)U!Cz zp{4Q#NsB8xCpE^tkjG8HQn^jfAxY3;BLDOyinAg?01LEm_#3!55y^1%O1b8nU)^cI zk^oD{Uc;7JF36X-f;{EJaiD4___V?Z^^OJZv7;i5W_HyTmlb7|*oQB54vR>(n0?gt z6K3hsx6xwczdErbu}{OqppcSqITa_wKi25roy`r`-Rqgp&6PeQNT&qJ>vcu)JrAEV z`83b7o;Y=yCF3oIXMWBYK?$}J3$#s1k1}B+&{u0VIoN-csxM8U1*^q6GE$K zhwGKaeF*`A4h6ZMx!J{@9)$rtjZkTqBh$$h#gD&1$WFHF&`DCHXTv$ddkpl0i+ZSD zvxQ*FNv=q%r*)yIQgH#qr$7n{{EUJtt=@5gq{xi&P)V6`~+! zx3rAOV-z{)niA=s{nvp$oH^-Y{Ne5*mEQsH5c81%A~laAC!8%f$3zX5%jKUPVUY(3 z@(K>pBMj^fz7`lW_{9tqeGCAs(GfvT=&~nQk|)c(5epZei|ff5$*65GtM39_4ZO*w z7J=}cnj%k;VQ~${VTtRb+`@|@ob1E@hpoGCi!$op1WtEL2~v{MNH+s29ZENdba!_N zNC?tW(%s!4-QC@t0}Q#(_`bircCX#}@FxuOoadbLy>HQ)!1|T3QsD<&y;>2`okwLQ z^Mak6M#Sy9cS4_40%FIzTDmxu&N&_oX5q)lzU-N->?jqN_hnVqf_&5DukJ+^h~2^& zUfwr4K`HSUI(*ro@$!1PcP$iMuH^Ru>C;7B#PshcZ%y&az7iB?z8*0{ z;&L?KogB^`zWVtA58_<+@*Cw}WENy+?9Ezsi=?~a!9_`!<||;oEa$_Qx9|K3%e7l< zE=aV-12P_3mlDIoE(hq_<#t@(+#;MtcV%rJ-S1-*nf6n3HYC z#lr-`q^Wu*_pH2)nA2W#UL*kFpr1ue-(Oo93bolO32ns zk{UN(edmxla0GT;)>GA#9Fq>->FTn7aP?bU8*ekUupJgL0X5uConuRgHOKYB&?@e~ zo_Ig;7IE7~I-I6`Gm)3Y`jDbv8Sc+sJAbkWh0pg0(FAhD#HJNd2oJ8nnpKf`9V*Sp zfXP68f%|OZgCk6MLVlivy`|a_t6yy2$HXtjnVU-dWs-0Djo)dRdBW?v?D7sP7K3 zK@KjbcJd|4Y@iMnx0`99GK)(An5>6TFAv_2vd?L8t_rLCuQAd|g?Qbt6a_>M?sAUR z&9Z{Nn7k{apHTdh@u3^CwHKWsigh1(O?to!M_A#9NGYo7Q}n3P-+NdIJNLIL-KrpZfFwjydA;b4?mK?3d zDur?W-~r%@4@-;ci^%c&2mt>7`T{Mpc5)Kpv=jmy=f-qJUCAtDp+4v!f*7dvM^#73 z7$?VPunO*)s9gN6^*(=^VC`SHS2#U+$8ev-DU^@!>yw4*q!#ue{SS%AT*}0vUj;&n zL7(r=e(f=geNKm;nv>J#^v0|R;V!|ohnU#(180sB7sZc$k7$(d#=|BW@24D3l`aFF zzB}Q6AnoIdD8=yJbGbj}BY`1m0`(zd>83t6yziiU^zr4UiKnf@^90JiVi-#NV&Ne( zm#_R4I-Dlt*TMr({LejGCgeOaW)R51p3WR69K^^+?Mk+YX@uPTRxYSCgZOrjbBU%k z>TJp^7y><93pTrG`QkPfJ#^v)8z&R?q=xgMtTwf;#73p6o`%Ztp5-RBt9n&Jroc9> z9@P{wRXLWql7>oBee&Eyfx9f+WVGG-`su&ZwvkZBeGNk&)X)v2&%7zlTmzXvKdskv z{qnevczD{dn##F9Khv>D3EI!1Wzul@(r;5FXi4k)$w@Rd@}a6Asm33@LU5CQBxvhQ z{~;ppm2$@g`y_bpdVTiG{cVhxPtWf0U;@< zi$4H3RAD?1nrpvsm7KRR9+W^z1*`!j{Oh}+y(zB(h|(2R&Ko)YBnKQP8UZ{q)RMGz zm*+*udS$m!XdhMwuH>7cX--*wS$=en$bIfVb0~a#hwBA_*<$u_n{+)Hny>U{GyueS z9?%;Oh@XfJ5psZAS4ke{zH|TbSlT36n{{29kCED$eAFMy!~)4MUAX3{Hi2pRbJIe* z)-EvGoFOojkNJa@M3<{+zx`W7C#KE!2GoG_r2+0Q)2eUaw36Bjc@V~54#L=?6*XA? z&^&4ZV`MK^qo57aWU9&pE(8M4*3f$vDqFq5$L*UA#fM5+E+7trcpUAsXu&B)*ac93 z^~qmay62eyS;FM$xOqXw!0-7*ce6dzDzPNBgcJXGV0M%TKHl}7;(F=nX~R7hOmifi z@Sz)=a>PAnxzhaNc73_)@NloQ+E@JFSx`e0sO|4ecjfjiYtH6!Lca=Rx^ql3`$`j9 znOOrS4T9ksx7yp6XDvy+%4!f9&CG7Uc%wwW*Q3u<2*@EUgm--I>Vf@J);1_B@1J8& zA-HD(VsWj`_ZU<%vWSlp==Sk#Y^vox`Sq0Po-=YKL!Bj7&Yty+Ux($#4>xtpZ*xj} z*e5geQ_S<2=XzBygm&tqavW$P6hjyNI&nJ0_~-wu`b#&q7&6yF>x%I9Sh~nCU~B&b5W+BCyBIP%Z*v^G(V8Ee7bC z7I`}uYY$OZHhf#Wf+-=D4k_azSwD>kL}OHc)~<8?)^|U15@(ls^)2va=}Wfnuj{7y*XCsu= zk}hch=)O7IH41w>up%!wL@2>m5#>zIez8bBH^C#Io=gX?@_b7EY&Y#a(&ez-sh<1@ zMIxo0>RV;Kx{+BEai0`Pg#F7{n(K}i ztFFDS>c*BtYy~kQNC|nFv5t4349 zdiPE23GI?>hk5ms2~fk}{1Nyv+__fXQ~+Q~L0W+|NMUZ!PXg|X}Sqcp&lVmtwrQrOXm^RdUR;C4kSfhU;K zRk9Fs)sI+nX3lJTk^ybA$;~4jjK8!Vr^!kEP3Y=?d+%w)LN!?sz z3DSFeEB*dxDXmYWe)hctTf^OM`ieHE+cV9x*8SJJ2C(w*o?L;O4)Tm>miHL-=20HD z-}ct)$_ygViGoR`$E&QO#;nF&Y<;8^3eo6Z(;3%r)2w*Cr`IK4$t|ZwHBK)pB^2PM)cGl zoSu=4GJJGzzuZR>bAP%MMzoc+Ac&n&N&n1mZ~wy)5C@K@lIRs$p-|6bZhzAGKab(T z+lK=)aYnxyoL+QVtF&6n$`IQmkaVkuLYg*qiLTMxp4{;hhN6E?`OTJ+Vin-SLS}<& zWfeM3t`t35*3iu+{lRO`EpzuDbrrnedo*d!axyYMHi2_xro>JYei!Ck)Qe_3tuRti zq;|!KN7hM@iFNPw41XTVy3?uiIzC$a@@2JqP8D`iEv+1JlFWgK%LW8mYe7#o1xc@n zW$oDA39&eyr#9ofu-AO#&$toV%<`AvDjO&Y}2^nAo#uYwI}C|iFp%yB7mqN(+zYH z@V4ZnweEIt)vlpy11*&4S6=g5sKAmuP5J7SuFb=UVz#p9qg3tpW#4oL_~w_KZRc8b z4HpuOwG*uU=ff)rjF=Mg;uPWm!Q(Pn@#0LKUGC_6EWeTGwDc{#OTn(wFKRyw5R&d) z!_F+=KmP1_kgcJlT6ftj9o8S6<*Bz1@+LK_E9|VicWHli`dEAgcbV5cwIL~twI>!C z4qt>_gegPsq_zK>AaN_>BQqxBBWwIyrl@Il!~u&i;l7uaE1=2GC=R5oM4D+;qC{rH zFXSG}Mfkqjp_jnJxB2R?sy9V~^BcV~Ih5f7Mg_5?)?2Ms zy~bF9`k(`}zHL`68P^Zle1AjEFXYewO$z?^>-(5SelL3MyN{j5C1CC}=|?jIP$lLx z9sbSXNNK0KUeuuq{3}NEv5~jmnbN@#EwR<<>zKS{H+3C7{k^hYF=3e9{mD(|z&^Jx ze!Xtix^ziveQ~de0-{$h(XeV^zfMVg2M>V8glV2;m&<3J^Gsn1wD*^ynbi z#*FD93645U%g~fjSOfw^77@KRp$vSZpM-C2{WMkpy88~bLo=-E8&jQ%$f9IHk%gqn zym&M$}0GgM2o0K{d5hBP}tSSoA@@(`g%pI-lEt)4L%1*Mn%Um+rAl#s=Uk+B)d zPiLEx=&fQYN|P-bYAJ2fJnI#O&lC4`FS+Z?s~1=foxR8vA2oY$&y#UD4E|E>>}vdNB~@7zauo909fN%8j*X{5FNjkq7&^L8`m zr9f}zHpXo8pWl(P=(EHe{}jv_M|&Tu$20GW7rw*kMDaWqHXALlw**7ueq4+|t8N^D zUnYK9yL{e-ow;B9@OKvpzl`~3qAffmUR2Fz8i~tAjr9Nr#RBpvF%>b~{$B`rmp=Uf z8vq-+SR0^t3yrY-_nq*~SEg2osp@F3Y~0qSz42*RT>Kk==D#5( zzfR%-owv1TmTUTzRa(zSk8_anjs8?>+o%@UCGI*)t2w_;{J?ie%3_X@yCfsrIFs*k zl%?qVZIV{lo2jHSp!^)000KhGw{~mY-s{ujg;~noXSPBG$E7ja_J~}CRsEI+3JjMq zXOzBpZ?$afG$@eH5|)qb)X79B1~q<{Jv;j>*c4XRFnyc$;G}Wke5`oR=?aTD=R1&U zmY7hoZY=JD(J%0ss=e~-eOczC(WD1d3or5m7J3_6Zi@rl3SSKd&Nh?$Qdvc=-xX;~ zrorZ26*-J0M!6ZYl;;?RdBvFdEI8LH@ zl-%#Qf{t4=>Wc~jxWww(?@-L7Ekx68Fr?Yurn8vASHVth^*m&WHzQ3RP>6k{A1ZN@bq?k51(%)}h^REzf%91IgKr`klchu1M|2FX&-3ZxwtKx}JJgWPg2lgRpTHbHTA?)GNsz$pv5* z*5l5y5T6NpNqeatgH|>`dpatv-F|`quoxq+Q#iSh$a+q>P6+b@KL#g3d?@KzP>s~Ryr}3HRh2EVt%PC!%Oso zI*M`Btjx!Z-2O{IgAbSCrlUWUek- zXrL9o_P3?1BQs>OYS+w6wsz3_lX6?rV0)LPU8c?@+P#wICUGr?SkN`O5)d@&TI`sq zK%31I219Witq0HWCiMp6G8DgPU?ku%91;d`1=? zH(u-90tt7xNk;H=Pz!|tcYA$IFGsi2P@Qa68{PMI4Y=0K+9tg0DjWa--aQZDw*=j9 z^7$1Uf3Md}m}gIYG5~iT2F`@G`+A*>FJ{$Bpr5)nuI=~`Y4kN^v`$<ZJes`lVW3FYVV)V4rtmivm{yfz{`kYU8IK9B{<=)dyLK#g^aax##2!&H3@^!b-K^v2L3yWOYZ3y%QlY zCj~&lP{0TfuL1#JgHF*`h?>rh92~o$HfqMqGp0mvNGSl=4<2Vnt>wHrZVnp*!SqTF zE!%Db_6ER~$`h%cjEZs5dsPNvb@noYH|#$QF&=iV(a`Y;8JPV@2)UUpG{m}(Fw1#c zuyo%v@j2Q1c^lnCF7P5@5J0W#ReO${at97FF_ss!z?Bi|v$SbpJXPPDg2w`ouPBCn z*(p*cf90)R!4!^CBKZ@}6!4%Nj*<=CRgH~(#h}kJ`|J`}VxCD{Vx9xudUE_UeD`hy zqa*pp<+N2FAC#s(5X?0^^<-~6c)1h_T&-b;)*_#)&UrlFB|hwAU`2#=oc-#q@>rld zqB>kG)-6n2@0ke-drh`J+#*sK5Sx9TYFfsvN6=R3mQ$t_}$kcORb1lM`t#UsFO@9MS>Fuz>T_)*UpID zvXP{$BIrV~+l(K4n?2YHTXHUA2gjeCALdT0`kP3m!LjYN5JpXrA>h;x4&xkXYkm-4V#(q`f&@O8#Rkrh|qaU+&%=;SSm6TG{ zkzAS8TBfz576zQNCM6>p>u-q~QHGckG?}aY5;R$=wM;s9kP=CCY+zQxn+Tkl0)X5t zy+TjKE;K00Pl5naaa|}F-sL+2&rsSMndA6*bS(Ne3Vgp|JgTov4n_h*svm|=IZ3C< z3x&UUBADAi**WwLkkPw}%}$o&YRYEaUagLKiiQ?x7XwM{syciBdBOu%=-k8U2^1OA z5yG5b9nsAO$IjB0%4!8>K|fRl#n8`-=Vn$~4u^~;>~SIYpj|dGMZmeXR3B8w)Ahr_ z?G#msv_@P7-MQ1t)RW?Ex97v{&DlxTmr!8Z%o*nBE7X4G45&DeaNfD%$Y2ii_u_1} z-Th_vDODm!3qZ2(G<#4o^;%%$n5%kfc+$y&ciN_4#c!3dRn?&X6JzVlltbxEq3~@D|u8t-ydrq1F#h0GO>rV7+Ni-0PYX^Q9?5BrH-st9ACf z$MdW1F>J%fnv1DkMfgMxRwJEWv8&J#3LamSC)?Zhl^34zNE@dIB5X>@D7!cKWZV;Y ziBqBp8YC(V>e~WRf9_zO9%|RV?O+X!zF7vdlZ=)}Q2M~)em6`-;H|pNTyhnx zR4faVCTwp43rV>$QZS|kn;eh#!+knf;K+FveY8xl`Se&IW-Jl#B9N9E1;8pmj5Cw*p0a&fHE37-~5bf5$ zZAaefa4a~v0Gl^=UF~L;m5`+Pj)d0@phPJKys3RPErcJfKrTR8=RpZ84L*j;Z zah>u=ypVec*8{+F|GD8|>omHPY8Zeqx+>&(bu8}|1EIVE(IEsQ0mN@lR_a~Xu(-J` zEoP??!b<f z{*l6*p+TQN8L|sP?W{%VjFvbEAHFd4AGilBpi2|%g_)}FCd6*Y6q z+ei*E{gCux5iX4H$#COWTVl*oQTNSyU=gn$We5QOFPA@(0ANT>x*&BxJ<7MGvwJ}= zqx>t6LEfKQIZBTI{f?=E(61~EmZvE!!DE>Fsf_osT;)=_Kbzjx%C#wpqXtU~ZBC4F zGX(yqx=lC!`H}E(jpzs2z8La4cR6IC6fxT2SGH29z%=Zi*^i|WjLlB-iS$8di_q61 z#VQ(-vN@gdM!loer@rh&-o&|o38fUa15UT1rXE12(Yt(~7Gq_M>9aJSrW><(7WQ2q zMO90Rc@s~MrRo8@_aUdN{0&80fie1+%wNpF;CIFfdOFdb-+_JhR~92; zy++HO#Vs-SpiN`_GUyIiZ36bN?vYVF>xlV>LJh|=*&qtM(nQaFja5L>>t&=b6 zMD^Bw)Bw1~XXXGeJ5B^d>N-019NUpYq=~#(IoY`YSf-MS`?bAB-^1s^hkQyNZU6y! zst}L66W6Z6`90+q|MZQTLVVL7PG^UoXEU6SL# zu#$`=!8lp2po}ZNQ|*bx(hMQ-kVfybgLm0UW8K~&c?>B%uY%2+efjCs0$$wX*b{)E zZ`_mo4X;0c#2#~cHL-)q?(lFjYRWFuy7@K$AvMZMpy~jn-#7M!hK&y9<^5S-3&oWW zhR_kXWUgB8EqTDXu(GQ<0D0-Y4s3lCIBB1|CDZp&=b2U72P&pr(AV|Z2D-2cvONLr z@|vEdAu3u!Dcu5PB)5nebt2Z@i{hz&dBUvmK&2Xh1}^=Jl13? z3iw|;MG);tLXGw$*1i6rZ{Z6qE)bW@`x7f35FDAxYqQ$$I)O~n<=X;oQtQ{7@C7D= z^PWy1pBcd&8xXC}ny=U_a&li22uuvB5EgcfJ_w$jwP3f#doY330t%_~qkaqE^;(ft zJO9w&kj4pWzj{$SqB)|mtds16TKbT1d-GioF4&CY(IZhu;VfvaMl?5mk6soLD!_RwlV?`FJeOHzp z;P~=p9XITcBUPlWF9?f9)2;*48&_Myb>*-Fu#It-^kQKD4Xe<5gBUB`zrNT7w*}WZ zD{3(&1Q`Rm$O9**Ovp~bTx0o`Jl6=-WKmk7wPKEL@yvZCYx-kssmP%ZP>H8$iE|9) zpY^CSp$)`iUFpHX=z)jLMQ25A=qhSOB`TVl_m4y`^#kX=NgjrW!G%QYK`*E;ub#&# z>p?bwIt7e)KTr;&TRR-Mj5l833TFM3<0j3V#9iBira9j+cAeWBvhTh=OP?9l^mw zgnAk%bV&wpyRPd0ud@T#Fq&9~f2Wj^x^~U$E}fX4SV(KFrr#$$x!>$9WEk!jWMzj$ z!8N)8?(|2?aWVlgBjyT@G8DFwD^X2kh}48p8hl@={ziQ#!xe-B#z(EMmgs^-ILt06qt24W!nVHKh|g z`+~J+({0Ki*YBq|D^%mYt}iQ3J=lHEJWWwhG30RoYG9vbNAv$*=*0}&21}e!949ZS zBL6gQZAX^r<}!d~4(m04RSD9gR9Q5ODfx_mCrM|AkfbBBn5(6J^f6vpNilb-FkXAz z-`g$&q)gd7xi|cWnH{3?S+83DUcSR+PI0U^j$^wYr|9Y+|=36HR^WB^_79<4c%leMc2}WrQbd2ihYmjgUKg=c!D|LozAIUWA_f0LC9+p%`VAPGl$^5myUuJS{lx)r{d7ZgLq0vkg>~(V&lV3L z0T}?AP=qjmYK<9yF)EnuDs+Yto`Em=&L_Z;du&EX{Z+3%sju}T_qp(a=j1I7SJTwK zLbhNGG;T#3j>IgJiuI`d$CgdsN>sC<@@3w({J_88J9qLIp=WoYABw~o>>{AQTNz*f z_gTf=zcZ@G&tr=smJcN?KU~SrgjQcC+%aEhZo+%l-7GAcGL1**J^B-{zd00~b}hiB zuU0h47!DC}ewaQ{CSrz%${Z^ZslFe^dji%M*+jQ`%bxbw|B1uG=F9PsdW!K#M(xm7 zny;xaH&o_(zug~0En$$kCpPAvuAd)~OvK4%LasCKYxolUH14c1XQ&(Mub%`U=4)|r zJ2@wjD^6A8EoB8FmV??yqnQ*3?+0V3jM7h|nUoXOZu;YIu~7eKqVRu?GKl}c`E?0F zHXH_(Adw);$OTIKoFHQ$0hSKp{QIFvB3H&)u=!B@^GR%9Nx-#wMjhcw@U8WkHP{!E z!&&0Xn;ojfuhGd6`0te%jmPtdmx?BOK9#6WMALl70u&{W_xnO-NJYAgZqw~geZWQk zFK4o?Z2oJ)?Vdij2H*OAl1aKCq{2Xb7a{fLxo2k3W&(7N{7QYttDG6Eh@L?xa_-bQ zg{Vj*7O$?Vr|OOvZjHahPzSjef}cSy!d40O*Z_AO1$PI`psTTIU~((W@TqR$FN52U zeCJ;3=hpu815lzA`9wDE5R$x%FXIQ)pIzJP6qiJ-AMl$PoIdaWfk{whDb`X8_UJn# zD|W@8VS^Go2${spC8cbHf*BD?b?3PXvLGvRE|`TCN^obs(m=ZB1e+d?;m?)-&mKA{ zJ};Av+A_xF6_k(XbC+lEE!tTpTrQ#v`Y(h1z%1+#5e80Jpa%d5wM`VpT3%~zclE`# zxScsIUI`6ZA$75xC;M~%2^@Kdo{$1nVKr4HKg81+uz?=JMvQf)-|7o~jt7ca?&yDN z4wQUTv%B0Jh@fYCzImh-fvS)(kDp|uwc~(-Sz4;Aw|*JAm7*n;v#0`AP0=vR{G zk!N^#9gZ;%yl7q3?pn&*S+6g4)WYU~<&KhPU+Lww%lRf_9k5O5=d}lk)dP4ZGJI>r z;{gQwYt_DE&%&T?6#)LfhB})I>=IiQLlr~(yMhNmqPuVPVBKIHTq&oPY0cl&xUTl! zz`GF6;Lk0U%{RI|=)8D?Q zTLSDPR=y*mur(ejgU+Z(NaMPf^vwQz)3ya`O6}!0_fLZ--1T$Idhj!g>X!PY!a^sI zn!TxXV?2y+l<@7NMEJx26c7R~0*(UThB2O(X52OXfFYEAV%#!!qou>#E+AqiwJhl+b7b&XT>$^YZ5qT^rND&%;|5VX8brB58F+>?+82NFm^f^&XJ(qeu_90__EkA{(0ACa;*fAbY9;X3DQj4^04F;KR9M{cPVNK7+6Jq^6{k1Ts2iZ5tRH)LT*p&TJnq5`f&Sf5Ts#4u5e7+DsW2Q<{1U4XBS2Dh*Mj zng*Odx4?4bMf;#EprJvKniu;(wW$i~U8jLN{vzIIqA;bbc}GooaPSSx?&r7l-O#F< z(49;^W&6B84e$!k1Jh^K*heTzU#aJqM=IbMLiy-v&Tk0_ERFAc# zZdIGoX2MnA{#oj*o_Rte?H|GGuRA#dOgrw%wwj~P(5}vg+c{sb3YfoW=2W4`S*FZ# zm7*3RD+94w%);egQvf8_+Vi%C?ZG%rf=;dl^WCyI&p+krlC3xwkZ8-Ga3)1_e5HKt z`1b9_Hz@g#p~Qg8Jaf3^vRA^%SKmhC_+$QUdhjJZnq6iOM6a(TyPB@1!H(fWX*}CQ z`gSN}uNsk|KfT5QFX^h4+X2El(O8q{(Q}DlO_Vv=lmHU5jmC0C1FWn1 zimtP3{Qj7zcJJRGVXt3-Y=5uvwc4MX59Yti$n(n!1iU?mpbW*<{W@U%GaY1i5(ZgN+(^%P)=${xJZt+5!%p(@xt4R5-RC!KmE!=v#j{Ov`yym}$O>bjeXoK=sC z&i_T7oJH7QmP6WS1O@~KQcfmDp?G?CT7`jeCJ;j?#*W4CNfiokeyE~7NswNcHkW#c zi2E+|+H*lnIFfZ}7T7Y7#*CEYf>%4&bCCTj5B=^1wP)0Z0+7uxCrNU+QKV>oe!wLH zzLTv2i>@_iX4qKBZ?O2@_@yq{*fckmI`#atR;^QAP(!3E``zHMDRb_01qkea_fz?8 zzF0=J_vj73wwB&xY7VcotwH1nNh&vwo!L#+La0fgRwjV5f!SVj^HSd?p5cZK5Sq;t zRmRr@C&cS+bGdx&ERc%DLuMr)HkkWM_FW?7ln8*_>FqOxrz^RV(T%9xqj&R(r?iSz zf5P|U2Zh<7Z?aN1x(p8~+;PmP%2GEt3suJYuLeo9lqO^r>mMGBOxhsck*{|xO}HHO zUUY7S_V|bPetSBfHIBL0*qX25Z81JJ8v|!S8Q$ch>B62j#|!P{WQ9A*L-`5F=~ut^ zC{L&Wusx$tIN~T&9jO1>__CHZZ6cakLJn<<+e@DbCnKQ@K1Y0HA*}UFHZ@64c1J2} zX}H`RM7Lvb05Fy+=KogUPjiR?4zTnmMt}GLV9v>o1ZQ2DAI}d%)kRwZYMp0iHMNmr zZD?0ynhH{g-bF$qbSgE+4G zfT3p!tKG%n+GKRkuQz{>!~7e-_D5T@iO=U_iroNs!ZaNxhY9IBj+4F`imA+VnT_}^ zrg3}CK4Y@PPRckCU7%bh_+U0)J0oN&A)iH2yQ?3an(apm416RK2_{#7yXfeH{c5?D zcPlW-a4N2QUc9N^Q;yb>2-Sfsat_N-sQaZkhllT85YWw+ZU1TOM~7iaw?=3syP%!R z`mq?wXH?FH3dNu^{HTrwPZyknMjv8`ASwG%gq5DJ2f2O~E zsfo`@1gNL3h|*)@NiaQoROmce-wi;9qX`y*`ve-R5#L#Ysit z*3m{Dy+Uf_zKj}-`v{F}2A;7GccT|+TY|_*KoSHnwjFRjCtWQE2&%9!oarh84LP!m z)e7A{)v$H405E}bpZ3Y??W_TbxX>;u6Hd}Q!*y(UUt$!JIBq&iIo8%c(W%SZbUCHc z+HrMAXpDp^%mpYX1WQbH>Oq?&0Oktl1I7cd;%B54bE#yV7y!n_WqHBgR8om8Ur(X+ z4TcEE-QV?B9Xh+06H%;(xKNpuI!2AX4*hOP7wlV9X4#Q-Nj6>WzYZ;CW{Qfl?3_0C zyC!hP1v6d6&er%qAsDYIK>^!K_4JbbfX3@5(0`IH-MU_-|Jr3rR}#u{gZpjiST@qg z6N8@yz;thf%EZkAd`?;0&K;1AOZ1}xqN*-#TKUg+J5u4213;&7%k+(k z&u-1}|8nj!rxFyj1+4wCNky#gDq!D78NV6gyZJ0ixT53ah>H-643NM9A$Y4q-4LH} zEE+wxZs~8QC8p{TiM05VdKHU)BU8BeUH}_>!A};O4Vu*@)D%M7CK)6xzCqlv*IHC1 z4Cv*1!om#oyH|VC!_gC@KJmn^Ne(rw}tKbQU9?2 zN|Zh#SZqVT$ISX93f?17hIFJm-fAOZ;JMVB4Rk#uTb&pmm}qpWwz^6nCwHjoU;$k7 zFXB18L~u$>6|!`HLo-8*o#XkhM&m!<&^<;J7C>-#3IR8hMOPORIw`XiuzHZ?Q2=1g zE|%5mTB1Wc%9tS6-cSaU!O}!@x_%LntUOKgD3|HG!XutqDT;+nUz~FfDGTWUNao0s zSUF@cs)$hiYGVMl5q+*=GqnlkJJ1<&E5$XXwZvjb4Q2&{_viaRPBGAi&eO&GlXp7$ z4Q+P9|AxSm{1zQ*cPxf_gkC+b)XbWfPH7<%ZZY?2{s{Clhn<4Ud-6V!*2t(3F%*7t zEwJxAjxNKHwVl0H#5XxCejX)I|2Bswnxb`XcCb_@r* ze}yU?80RL&>xDzWt&(RyEwfrxu_VFuh1MgBW%%3@;Lia$QJRV!dM>BE%cs5?rj8`(- zT9m=b7^&{@SR?oQ#VI)yxGDN5_P-kzy7EBG$&;j*H%EU*k!~wofymfazVrSoRe{Q6 zU2y_~o~SQDC3gYkW8l6X*TVcIL(Q=e^HZ1Scb;}H??yP9GWUJAOo0#O?Z;JrI9k3$ zU@RNMpUC#6iX3BJGW? zb2CgGr{0wIQz=o6`D&hf>aM|RXG@?v^8AvWZFegs z2$4X=Z!a{R!ovq;)LCQ>qGL*2Dx&3GBOv5^^SE$jYR9^xGuVhvyyE3_H8BDdk+z}u z;6A=39<33bdY1Df&gTE*_=AoM!;*NG9m6GRiQi_AufomhVt%tS2`JSF^!5Y545Ow3 zEwC6R^_sTTt8~B)U6;P3XY-BA$;tRcX|}dxeS5pGV!i0fvyZPZy{4zhnn z89&6BsLI+6mY?I*T|;=KrC`RNMQhdG^mKSryxzRx+I;3>hv5PUiFaW-QCswXhIxX$ zhc`G*uLJ@&-&yO+to&+v=S+1ZF&kQfo#`mMmoUm=fIv-5wVO=og&jVMpY~2b`Cey6 z6Yx4@jnv2XPB^L3pf}((ddGFef#Fp2jli$ccmQJjB4e$pMPNuXKyaVt0N^igo4U%= z`r_8?RSZ$PDn-hlP1ACafJR&be_ zk55u24S#DhG2&V(s#CQt7LUdh=DD7S84QCu{{xPgWz!1$L>kFMV%@Pe1~|(QOs8Ez zcdSkXb;Wsx&XL2nMBIMs@lTdsDGa>v?^4K(*TQ}|qA7p5wc_)`v*-pd1&+!G27=Ql zr>JvRx8}|lFGs@&=rSo(B>L=G%*YC-^HzSHv0;5qhQb5G5+Uonv8z-s&-gNC`>uG4 zQGJ$vS26Khln%ZHA>dPj+YteRl4xmtpE&xV?I9V75=>flzFOFAkVmw2xfMV1u4+oI zqB8=&@PSgU(ObO$%8r)7E5@;6j8UVsK7r_9;2bY}jUceR4Isiisg+YJg^ou=!z!h7 z-ah9_f3C2eCdNL)0|>EfEdw{rq7Sw*-2PdP7|U?2?i@klB^ApRj=xb^gul4j+Upw& z4YEX?!xOpBz4y+4P`FK7KTIvQEo*}?cpp8L-qv@`WEO+e1D_uoK9PqB;khG}S+!JM z3`%O-2G^t860IsQ!=Sw=v@dhgjJ*R$1KkM-hWRH7OKEjVIGO;|cyd(_l>^-0%PtN} z1J>+AmI-srY~waXD=N+A2kvQBf(tHH!LKgpHx>U%gvkRi3-r`lOFUkCX+x)FVae28 zbQbJsoy=!*H4{z7?G^ik$walrr%DPmKUcrS;HIz;nSPU_4&~Fn(q}@E>TShS<6yos zIf#h=p9=Y2Wt=v2kNqrUPl2<8mVD0r*N1?v?2ple!deKVr&bNs8(2xRoTa8(XH!#5 z_bS(g`|!Ynta;WYScLmw`dInZr970u)%OuDIU#;*z>ST1f`IFQ|-YjpMhvl2b3O&!(8;=v; z^N#4BJHM;~{*L!wN6v}0HF15pLHoh0S#%A-#V_$OhCP?%%3_1fZ2Z&G*yF`+uG!oq zK_R92ay+{l7voJ;YVm8-^{#P-ae{FI?M1RK+KOUfmYd@qJgclL>%{J%k!BQY+RfwE%q>`<{*~EPK^n=VpDe2p3{Y-sXO1 z@?C?plIJUn)78Q5)Ce+9)I8tUEfkqEk(X+CGHfTIaoiAH3K#ErXUZ5cI76$oO7Fg# z;1k#-_&Q{-wlvS3o57R#K>slKSIBf-Bdxr*tIOz}OM&&swxI_j`MepcZxT#cMzm58 zCxF<%35yO8mIDs?H{pJd7(_)^p_L0KT_Q9`sAdnEE@}Av$@OEah|_xC4j^2 zn>(O$D>ph(L=P=$j4VS?B$Pf%@l6JwkH^Q%+kY#*4#$Q)yRnVT*c<=?6IlJQd^&V> zqBYx7QX-o^GM&rg%l*d<bVg;UhNv)^K_YvB~||b;bd98k``a3)1kh5)UWM27w84o9R4~h2_aBq0A61%%>O8`~*3JgBDLq=M~Hshg8E zbFOvS?qp-UU2;u5EfCZgaER^E+PN-1gPz7F5i33&G_s&%tTB{ayy(B|0#Pk8+;qMx zjfXJ7rOH{nFC54*W9lY`a*3cKfbN}??^TbRW{rt!=o@>-gH_aDTza^dT+yBx;GQ zUQoO}`N`xa74m@YC%p=FE_a%|%*k&iTKhD8qugm2(2Wp^YG>Y)OOU80cC&umxe#br zcCQ0&JRHCES$pt4lMqYvV};cGKY2dzb*|ifq?rZnb z`GhQ?BMm^^m!C-e!;6UJNq!@A3^8xEvRIm~E5zg#w+UKC@&9fKZBFf6ar)3J-Ene2_n9I)Nv{U&=2wABt7c_jvoh-^% zRV>!?(A$Hg?YX97^VZmyt&9hFKAnP_=dfc5X;`GRD0gemy)S_`^FCKB4LU$Lq8P9f z)x%}Y4KU0+5S#J`vc#bMmDf9a>iarbj=`hs78eCQ*@WH8*^bep!f58UQ>%HcPmk_T z)c2Ct(J&GJ`j@@jy@MB_Dz4`zjbx)iE7H#NGieB=CUq$b7J0LeTIcKB+jKFIY!PeE|$ z@F6Th{x{U{Mm|$K(L(~*p7lx{VfDOi!@K)obIkIRdhF%5kW@=huy6IINRM~;gIFGy za>_fzG%cz=nr=?bw`%{)5nOpliloE+0UTLPE;LW5&!E57n`;66#8^idjFl zCgS@bbllQ(bvRng9%cUnp05ixO3df#aThX@H8rBZE4cFbbnkukuzK5yZ5}yElFQ8~ zPod17XCMdrdrqX>TaOf4I(H($eQvoq_qQAsf1VJ~)mr0eeml9Y8E>kZt_rz-UW4Xl zmgZKJcGe*+k#u~Zdw-F=g_TOiBWM+}f$?Hq{g6@n**bgi6G2gWu}9ej=M#B<7$6Sd;{}+gvq` zSQcotSwM<*K!=phF#=-G1&^O0*?#^Q-!hZd{@waj+AVEr#x~)EDaWtjExc^{lAmoa=*3c_!&!0 zD=G8g5pg3!t}@{(OOMvN=2&&QsW9=fwOxJZN{3C@3SKj91&gotbDbBvg~{}A!NOeP zi^WqhxO}#I+)-i-Qy5vx2dbiKb79ZOC+#oQiR6(nJGWlc@6BfHru$t-Nk!96`5edm-Eqvb+5O z1z{s&Pw7G81iKcw-oa*Pc@T7wvnkyZ_wO+_y}=W@^u`8$C7>D66+PtlGyjw0ec&$P z2uq5#!T4{GBhrZfZQQD&hbZd%IpeY+9pu54U(N4a1C0>-(4q*$T>gCywq}>KZEahh zP8Qw&o_t@zg@9Mm!@e8*7LN;x+R{o!iKNmx4-j&pX4+D z&80A`on=*xT>YjW(lYfn zJ&1>Kk_FF2nGq#Fc8y1NyWmhO^n>Fy9wL6Pn*>8_zf zxr$DI&>blmdU{= zd!_i|2NrAcw~1?Lw7j|nApaGZ>lYwbtGr5V*xa8+1vXLn`?n~mA(&n~c7tWe&&P>C zixt_MpVFDVr6eo>4F6ob(ux6NXNv8s>xSm|LPSmgF2r`?GyWB#EPi#Tv*VUaqg`1M z(zq2fW|e)W`06ZLmHOSzqaKnFWIv+zmTxCA!(`}~BkUq=haxpIbEdXYrB4>P8qA#r z*d2Qko8IqyEy8rc*myR`^ulj&D`t!)M|yiTd+BvYRIvCEKg0tB^#vkchu59En9Fx1 z7v+m>!M&pZ6@l^NHv@F6lHR%iQrZ(b0F5>t3wpdnMzLS_j31O%$a`J-UHb8j4uL8wU+x_om?hypkMGdu zPE^}s<9^QvQ@9-aTbPWeMv5~u#GQPW&fE9l?K~C_TEHak$xt4MHO_z*i&3R7vg_z} z=7c0_Q`s)?TgYnc=j&?1p-T8K@o;c~aBp8ps#DjXn;b3WhabLS~4PxCD z=a6@>C&>zdh5|W`UqOPT@u%A3@89CaG|3s{mC+7A@LhqAo7X);2!faqw0vE0`&BLg z9u1r)zX!>TDv{mZc6|@0$EIm?a0ah6n*8krf4o)`LHyX$iEz2_uMjVKs%a$w#R>393SX`fFZ9h;gZuS#Aq-%`zrRr#6e zO=teZt#6%y}#T@x~Np2IO$lw7)CCD+j>HS5H>8mZMKB5L9+a zXLymJCRX5!vr(He?O-UprITeDmQK*FY6^^e2uw=`=vueNFT)s5|1SGtkTOzVu3Zhl z&p%Bc_mOja5m0s8cg!e0q;+8dPH}UTY6IW7U(93MyUqjCo5Lz6dX%VyDc0Qk_P@^~ z+kcrOt>q2=u-TJF%}YbrpGBvD6-nh=~(el?9*~JP>I<*&8<2nl38?%TiIx zP?7sNol1_974xe*T3KHg>f~5_Nk(3uPB1d^7p++w^OoJE-~e@AVyrR;^5~PYVMn~P z$QhJLlDwKc1}@;TOcOpCymSu=`HlJomsEP_18>h1IjZQ_3xUN|0lU(++WSA(ncwpP zC^>we=QwN-e|WG%4exWpI=)HE;!p@aI}L$>9l4bM*OBXQuBqapG5~i^c-Er|inPqy zr59cw`k45=Na3pB23VFmid$jsb}Bn-{IjVmJzSakUa-eFe_5d`Dg1C@rLo%u(+SfJ zfXs^z^VA49MQgNbof%H~RNPJ;s|_f3-2J#=x=^DQW(8m$$irk>X^oX-18Ig=`y2Hs zUEf3Gao7aW@|R&~zdpEofnn|AA6#h;qm!q-3^%>fXb$z48YMnN_lYaS(+FjE#6BYi8BL7i=Kl?DKcpNTc5Y9D>dF zL+ri>c5Y{>{DoJcUaekzYeiG>7_J0g-Cg96M%a>*lYi9QW#WsE5q#&z(bDDGtDdvjld0_0xxUOH!>Lq;>P8P6&V~dLAx~$ zW8g5y@L#76jtIfQ$2KFE)o*$8l2sM03TIWSvlw`CX|UnlD}bDS*kj4(dk&1sSLYx2 z%a;dhZ;zjQ;T z`Q?fc8_Fr-_V>@ydHlvt->Ny4E$eqRqU9)@}Dox_|T~{8RhxL z9k;VsaAE?@WDl1*5Cve>WVx)ab`zC~J$w>qG@-XLT?s%&%VFkF2p{fTdt^1^aJm4d zecb%ZdX4a=$3`E7!|VcWOScH4EMtHmi}I9Khx4YI#0w&bk#7JnsNPW6Yu_vpPmtbO z&TRW=z&qiLZC7%+w$nCjI@9m$d}>}RqQF5JoZci^p#;E5%-|Nc^vCh;IsLj#fjCYC zV{Sz|nZvFnDXHNV?eKIhV<73W%6Hr3%6-3w!<+WU(p1d?}9S}~sB`+I5TW@vDC`Nkr7HC3{hLrvrG0>r9^WL8th<;Bko zajp~6UU3z!$t!c_6ZrrmPrDTuGaPZh)*9{2f;6UqgI&;|(Nmzob2t#YiE7>Lo%&Wn zcd-kpMDG*_)09f+UQ|Kl5nv4=h6)Ps)U5^ufAMfK}i4sW5 zkmp)#0Gti6I1&Pf0`D?aL}5(>O`wh zt^tH55^=Jd0wGGa$HR7$lVwL5vIk-N@UXBjbPp?fQOx^hi}hJ8i;d1Co`|@o|KTlM z_c6?&$d%XeiP@IUOm&4NO>Rmapo8>*_ulq)8MT>e&%c7X`9!id!a;n1DMmw_j@4b> zy6IN_isIPOq{jvVit;sidEuZ+P>!tl`cj%9V?*9DxlNqcM zPH^$^Kw|+@vqpc+`L+BDXC?rA7$b)Av-CjMFaR>p(O4l+H2Okh@q{>iM|>)vn0wx6GWF zj*QdZ_|sFOZT4{^q?fT83?&NTt6kB! z;$q}q%fy@Rx4emWh5oqxWcIZ626Wn#PZ^f9`nR>phD4)KVTWghOJhfn2G@rHknLJ3 zkAD9qi%Eci8*)xr=Nmvyv4gSFECPniQGw?*UrPfJukU9rdO+e*Z)#&~Xr`#oOm?~^ zBcAvOoICBASV3C-F3!QrABUf&(Fv(oo9D&FX$C(InL^B(U9pHtVGS=O)MKsF{Q>sM zY|&~ix)OMS&cyj+R;tX!LRShpzFIpyY2yGQHB}bbBKwZTBvjdGX)7ULViC8a!92vT z=pdEXfi1}}Fo=cwaLUjdbf3u9qHcn?MCnGEDIR3$a<$`$QlqXsKuQM=ihJ~gd_sVSp+!D#{9{z5zxs)}jALm$P#%#@vNt}4l}69_q3&+}+%ZI~oYE@nUfj2Hx zwB^F-1*tYqg;nOiKY5B0PS|EEFh2+8zkc%0O`EPG+c=DS$}<@i5z`)q5HCMeHjP&G zGgS?%it3JG6k+iIP;fnfU)`-_M{^3B^nUa(B3wLh9{j$B6@2a2 zNg>r;G;&@s8{f$W*MYJ2TjXhsG6F11(hO_gd`{n)3lm6$-)L7^gaNMb zjE6NO&yUKUDT$xk(PD9MB*BvWJphrkfLF;j6(xg0OxV}GojO61EksOrrW&gb5Po}h zwRn zVC4DxG-8-CAk=)XDAeHUw7rBtngs)FvYKG1>~LX6eKfh&AKpx7cpsiUKBdZG;nS~B z@(Kl*C;#?L3vCNoaB5t3?&`428l z!)UD>+%rX`6`+52F)wU2muVf6A5@nLAh(LJ1_UV;0$$AQ*UsD46WKVHx3-?UYpeRb zJ)@uMu*@BKlkqJhzQ7m;^$zQX(N!TyzbUWW*80Zx3SJYu;1at?5ft1vKU3b=^Qo*RFyARx7l=(7WGUScpGtcFRq#^i1`wd$ zQkPJda4x%viuZ2n=fW^SD6(T^^5dec>J!2spj;Nlg%9cQe$zzJ#bN6{o5T zqR-F#=98n^meXpwKiLRdki254WmDOu(a!q2B_2}M-l9@SOtL*Yq1 zqns>+mjkQgCX^$5U;VDoIIWUuRKqll;amr5yE;a1(URx$pP_maBVM|Qp`b`wZm5DJ zSI*0r?1op~)%K|cE$E-Z1+)Ny8#tImAZf!+iC7y*+OTL<{GZ-X%RV$kG=1?@DBg(^ zsG}stTqUujzMhFSd%w{arI9k_0j$ z5PSIGGz7tB16S?4RbxO8Yk)%Jbzl@C8&1Q_?xyJ3Ff1WJXM+gA&A=6aGJ_g{0;sBk zMGtk`@YvO1o4JAKjW810DHVJs_OC>MkmlOi+H|aua(E|Om;pxfM^9gNyv^ssZ&MMy z;9+2C*^e*wq1HIlA3(V4_*m}np#iif?>2GDQ!Vy}<;RM>Ynq5^#X zLn1J-pWX6z6sD5;&oW|DBC}xkR{ru@q z#{b$Gt!=LZf-c;kyNlB%$1#1<47l*ntHqPn%@95QdItwEsHJ5kwwp0ZSxVO(z_&HO z3kVPk04{i%vfLzXOaJJW)*iyL{Gj~7@m~PgJQTC8W)t5Ye8lDF5^C73af>QpGO0=c zm<9XF^BHxh*kmwPd{wi(7IfV$Py9{sEdh)!F+gKAQwrD^`ZxU!0c-u&fxnu511Mb_ zSf;hhD#=&QzyIe0`S%YH?Ki3byr17n0e;@M+O6-;zD|Yk_Dqu-x!g{y>O@r&TC#?} z{tm!v;9eR2C`|WC|ADP>-FaX=J|u38IY&$80lGfqmbl9?4S!BP=VCQNH50k64pH+{ zJ_{q$bFgbX!iN6ha~x}VeL^U?w9?)k6xH2;`Jm)Xi}de*=T_sl=jS8hRe0Pll%%3O zc>k==FPNq9qyZ+1eaYW>>D!!J4BxcIcvo!vO2)TU^7Xz=2x|{xF{61RM(g{kK^yJ% zq;V-2^VM0@4$%~@W%_Y^9`b;Y7B#Pe zV)wr;CMu6;j|{>`WDMS98PQ3v28Ihqof1*i-aMvwL^^qQY$F#CwTGXO z4Lb_p`rMsg@1NmMl>LnA{|xw1s28pcINj0sbEEcOA95oukeP2^;v3G_z|@+@2jLCD zHW>gjDG27@Pp^4*Jc=Hvoi$wK_$%w|(}~rTddEeOm+*53g@we1qls05sF#EqEC@*o z46%L2u*5Z4exeU|)vw>sdbV=FZ3yWja zOxN~8sJb>k8`wdmQ1TD}FY~m_T$-*%eR=YS?x(CaZp9GHrEhB^2@D6jso}2{w8J47 zxHyfMCF+_TMNI`e>BX4*oerVkODhwHu1;)BewNUl5lV5NTda5u15Gy_jl+hu4}Ghu z+fqx%3Ci0XLghqKPtxYFhshnv5SX)hpt-~fT z3J5aFJ**B!7qS}j_OT|>0D%%v7>|NExy?`jKaZj1sqtqDZ;gF zX))*;@f39n_p&DKF$MPiw`P?W_ld>E(CzAn?`|-`gBM9qZb-;A*O~dwfP_QQf@))4 z=4Lqn*YQ2%H|PtiC^UhYoTTrKOmIy$fOf5R*fwoicd)Ih|DVm$;RL~8l;41Nz}kQ} zs;R+yK(ye)sa^Y3Mr`#jb%Ic3(dtEeQf(YAEJ|TfQr4BDR*r<(6u~X&##lGm$&~+T z8TYtc`jW+Wz*H1;rxW_a@XE{L26MuHChM z+hGFia?Z}+(K?gFH#_2M2Q7jJ>WfX*+aI<`TPB6RE5)| zE$%T?7JQihJmi;0IDq-oE?vVNalOpnv}h#@UICE84M^P0X9&capS<{z@LPVBrt~=g zso1r2yjD&t#^!Bpu5N|HXp`hFO58@dU-kLW9KT6+=c}+tj6f_Bu67EKg|?PfAnsX| z|Lk)Hi8!}EiVPlIqwz+02D6Ps{4FGb)4(#|G^0o%Ra|)erb9?|vz#~8V5pUTz zY?|Z?xu_EjDN>* z(qwR#v%IGzyo}UsMAMp949f` zHS;j{$ZFM$UHA=dx3NA8nL-WDEqNuKeROv!JjE@|ER9(+#oKJ_j(Wv84juIfabm|s`0)98=&s(DlDUK$L$t*25;kR=f>WTUA6s8xSMX8Wz_%L zNbVwt!q%{K{O)@54}2>^3qp(ke)=*ZNpCv@nt&^I>3jNyChVDi@@c5mbO-*Y>ktN1?!2$uWP{1eH?i6$Y zu3F6C3W#pUY6FOW5o#RQw=Vnf()1HXXU1lPp71e3y7aWh`;BOg@C3%#e#WU#t1BHW;g3U zhrvw04PZ+hi2ncjgaE!0Dz28)Mmst<0pRY_wk#f2Pb<2u`m?BjRZ8tEd{m3C%m|90 zGDn4yABU@Xftr1@Gk-_*&&9GMDc^sava0d=n zshwq9TdmmNh(45!-@IMo6<*pIKH%_Juv&d5ZtthEEx+>QyD!b{3YI4+RlOJ+ve1aV zd!3G8gC1z^A zV*>JlVB6vU+7PGxdPWfH2$eGly#*JIsutWc?~17m_0{z998CBU+?|E*=tvl?Dt)5Z z>}cHPXJ53#T_&R%^t{AV>ee4?ZYt*qxZIC`$K_p`Tx51EA7odAYI2;P3eX9(-EG|H zEgqU#+s!v5aZuaIbPT?WDf-6CV*4}{*H`;!+0;p}n%dOF<>S`n<@kGM?flBKj%b>t z>EoqtX`_nu#OZGtbtprim}YkVy@}P``hAD08+gxV)M0dT6i4OUboqf$cs~pBG#&Kx zHt(bp=ZZwefA`Y=9KAa)UMq!4JPC!wCjFRV8=#o1D@H0BMr}lP_I57L7;!Y#zkhqB z7{kd_y%bRG<_@ z0S%sBPExbL3hjeZu%cXq&9I_(;k!$O7y|o0UZh%)IJuXfINXxeqG#8$zTMls(|IS;apHvP0OV>dGEiJ^+0Njet^?(Ookaik?qGHc-cBIN^ zHcD@1r((*aq+<~F3(IxN4PY{`$68%K$TpAk6*WxSxJK-}s2|iPPp$g&i8yJa< zQdlPojF39)IX)}0&mscHOT8g6YRmbBX$3$BI!ijY%-X~D;TaHhh*Y+mC|iu$0JC1q7bri9h#$bF1RM66g3G zfWSRrnLna{`rrifapEf=ytVfH&zv46<-UK-&6oq4&67#$LipTag|s9ge6Aj|uYKf* z_F z(IKiqD<8Vw$M7A1<32OTG*Xx7YYL(X`wwiNm*2uyR!@)Byf3FN4ADxVT9hyXD!a1=%p%7rdBt=?Bji=JS&!IZD>^28Pd$=UI zCb^dEPyZpxW36&N-7&QP2d0D}jaF zb|-6fvqi*jncn?EXzhc7sWNpcf6=ypi zW?7yW+Q~NP-5)jdpk}tJ#p-_1>g?(t5eB^oFOF-Tb!A8Wlt$pYM&_pr2}B3CG&zI= z4G}958QS~(#mi1p;HRV%{v;N5m7e3;iT@;;veQ|`((b-E7@NjYwdX!Cdv<^C z#r!#+FC`l}NSxBg{g%e+nrC)&5Y;#K!e~L9;H8(=WJ!hOpzU%7leuI6;*VBZoPc$lpK)hZFM%flLb*03Fb#~ zL|cSIZ-q_23Fg;ztXZ@Q+AthyO{Vf2p=o>L|LvBRBh^6w>?HJR<29foWt2b5QEGXw zznXcZEVKsGnrNI&B92tUM!g7rS*xo&+z5Hr3TEbUDmCGAsN!M@kcc{d2cweZ8+J_4 z-<2a2y(3agx3PNxi%I{lsPrP$Rn=mU_B71zemyoymN-JsCnm!>>GO+g%*X%LWZA)cK=cqj8)!43QcB+xU`~BT1B@?~35|7zl__Zu=?M(>ErXo4?*&Jzc^cd2ksecwvkxD8YUJ zKZ;2=E-ju;B5kh^>)mlg zS;JXE5j*79Cv&SnHAZ@u=%N08wcs2A!iY;^`**w;WpaDC({MwW%+0N%nh1O*w_bxlr-)(e~ga$7nw~lN=Qfm0E9r0Wv5AeYBq|Te0!y{?zhLFw&e&G1r zLyYTbbvnA0_3b~$FJDYd54MFyCq}b_{)MoZ8f+O0?&ncc`Oy#{a@WWdTkmJQ|CN1c!lL(J;A%g;iYnJ7ZsLyU2sr z{mJdJ>Ox+F{0CEzKQx;w2SC3QPXp8H9d~yB11LrLf81)|Yg~dtfoqFr?&upz$->qL z&)*ouf*Aa;Piqfs!PBHW2j$Ie2mYRo8|%plwo5&kR3ZLYcpredv#}q;pJLOI20ykg zpD;FWjXTr)gUI?QlTmaVT`jTm*^=ZgYAIDsct`@Sp_nI0zH*|N$#v-Xu_2DP;MO{| zfZ>Vf&?>J<=N5Vn<7X7wkJKx>^9gtv0Nk&w6h6Y|;Xo_0iOzgU#p&cSUV{OpRLG)_ z4|HENQNV7cBC^Xk9E{aW;y!9GQ4bGB@%8fpUBfH^O;A3cB#AGU`GNrk!1IdDOP{OP z5T+eXhGWI?b3%AD0QXc8nzG1G5sVlD$YQm(zG1%ds9ywod)PWXM*jr=b}Eo=0fjxV zkMjeFxDdGkuSG&fTG%ZJ9hm*Gm797p2DcEdY%%D&OZ(M{%*4gz`uC`N$oTbro=#ItrBJ^4ar9L z;vK&U>C{*kT&I<+%u1fDU@sJ(gqnjfc%-r4>xm&*Glyw^8K6ynU9wLHizr*P6R}rH;HYyDz6T%9ul2aeb)HDYujZt{=|!aqNla z*lH-O6i1UYdEt}K?dR&PjnLe10jQ5RP%}4WzCI0>ZX-MC1chLY<`Q?J zbPZq4Bc0b){H@?q_laf6R#e~`G305p%pPeaX1EfGkond#fzyN0W@%{qQR}KtPc)%% zRP`qS^#|9LtzTm*ZH4~t?r1rU+C2Vmfrz#ziOUIqf_IG5Jb=eQ1SDrqY}T*pM&k4{ zFdveeT9Ow)z;_i~_UN^vA7qS678g6F~U*7PoB9*!Qm^`|QtiJ+>t~+LZ0~ zdbfY`q`qbRSZ;vXY4|vGU2wpNseHRSel;gBLX+3%3xZ8?f}Jt29Cn*XqaAi;{PQ3IZa_hoN26){KCQ_4T~B z6S>x^Z_>UA(Yc}U96PQ-da0~lGlN#5Y!vNbIJkHL$UU)N87Y)IN5iGbaO^MdewEAa zQkMGEgjYJe2mfyBAnWG;Yrh-|}KGoaxN z54O2>m?{X8i84CTBvG+YQ7{J*`hE1^uV#SPToFYj?L+^Z<#S7qy5{EEFaY@7IRbi7ARYI=6&>C zd1}E*DFFLm4yK)!uySj{xBtrzGA4cm4|Wgs3#f7QSh66mi_J#+*tTo|RT@qN4~{kY zc%a6Ysq1aEO;e~+Ce?_s zF40!(uW?s=xH>_Vaew)&EZhjYjTDEuxU{e?o77eLAbmkpXOMsSM)m)B*9u|;`l#og zUv7POv9B3JVX0hu@Cyj|VylmT1kBK-E2^B-w!zXVs8+H}3(xL&9TJ5Sa?5sQaNI05 z^f2&;qRnpqNbs=^y$x~J9zZn%{JW*1wda$;$xs_m$rTO{ElgksRz4OhQu={U6LcFC zW0?Rz_5TKwxuDdTc1)!m_@YcxzqCPjn;n%N{X5dz&gq0|oB3kJF-nZ=FKO7U1?Ox^M(I_kY#kPh_U&0+2o%61Q@A z(-=;7mSZ|&01#50I)UQ!OMAJ#Pg%zoS+9|IC*oWrZ0oPZ*5A{z1;qWN7o0;rcCr{h zyJc(+nx$m`f;kT|RrvIlm|e?5U|E#;V`}kLgR~wm;}v73-3f$e7e$~6;X|?vk7*!* zk{8L%l`;`=C&ZNl^_brhl^!=N*JVUa*-YXf&SS)I4;3+CZb~Ld`^jwx)-8;&M=SWu zeL938B8A++nV3|T_p1##axKtx<*>suDK~=tTia zlPc(eapi9Z!#;FS+Dj_I#}fl*DYcjEHwM?R{zU7=>TthPHbX!@ZsjHS!?|` zNI#LZipw!hXLi45F@mu?H5tirp*$Mo590BYP5?(-Tn3MQbe|I53QSVo%rl7ZSkz*~ zSH6P#LeeXdcwzy*J}&}WY$-gJV> z-b#Lyus0=+pV^U+yvRAX|JzHBhV7;Nkd$Dv5ExGb%)fy|siZ$T?nS?zsOHOx&o!fe zdRD#BVui+VRI5gwAGji>;nf-u(BhdFWpjSr+P-GDPjW{pZL+yH_F*ihWhdz#1fsnS zm^3TzGmLFDR=JW~M~!>9yM-0rzZ_fo^Q$ycFE+s6_lMGpG!-HEYH~}#VdvYE-I9fD z=cs;G>PzXvh$UQ8e~bWJ*uMVBJ*y$#TW6)?Ahm+BUq*sxrQ!UZlg4xleLia!Z=QCF zI1IH?pyCpI)kVGto&f1J-3&)evE9ejX#{#OCEX__7^}oOXc|~{;ie$?BL0y8lBp+X z)t(wH2fVkj0HD0T{=Chdk9+F2wL3Zu|GWEh6bTXH#oU6eOF1^yC{?Do1hC7q!JZ<; zID0u=qF;Vou%5KXH)~%-HYf_yZkaYBl;NgQ1v2yNyJmNszE0cqD1)FY?(&vS*HCUT zZ3E(3FZre&%E0<$@h{ll8D?Wq)Va<4F(DhAOf3>r!>_7_{c2mmEwpqVSxXP{s{`=a zVJtptVIr=&tG)xiM&#df4*6`gckJpR2j)u!q&kXqu^Wuij2dFQeba{R_G zmJeziT6c4zrc4c*qnl7erS>=$Znom)#4(b03*W(-BH`e)GoATWG0qGxE-R=DAYDjq z;&rpV9*K@#LGQ-@6wo1+*@+!ShQV*_-+e`}sf9E^8V`iaKLR1VSLQMRxk*-O7J&EJ z%UvLrj~Ud&27|cAP6EqvP{@US=I;_8?4%EWzZjLb^U6EZDI^)Ijw{6o1-Hfd-sCXH z`r=)P&D1|BcCGmaCB?el)t#^745gp;o#IcC{`UIY&o=~of>7H^_NT=SbK35HSl5_( z_0~B!^4C|Ci@w+xOx}j48z=j-YMxRzYr6Ei|{Li79np`rMA$iDit!JOQcV((bjI*XF>lTXLK3LzX-~tLsgirrbyU=p1|T9#AOf;SWu^BCx4I#+x{^)!$oag?kezpx)!?u87zpeA*b} z^B5_U7-9j2zwU2_(nH%vk#n?IJdqqHA$7`84?K{v;~T1UI~4kIIUYHE>Iq?!-n`#S zo^DGGi_9kA(ZC-!AvE@PnAKe?5$}PG9yTmj4VJA2qz0dVvo{F&KW_x{8v*K1)zqKd z`)e6M9hz^R3?~~HIp+``Xoy+)WRmMoG+O$W6UQ5YbuW4rM|tWTSWE6JUJkaM*IiV| z`c37>owZ+n3zF$hdNO%&y3YM2>qXtA>k+=7cK8Ra}H5{t-(knFc>7eS|dcv zQ#g1uP^b?UCu(q(>~&Yx1=v5#-BY{80p*PH6KBA2Yr+;4-VI>~X)=g=$8vI{Q2I~t zPYsw@fne*ss?1g72dH@9Jne;CAK z%LD-5|42f3j;3yo@G%f}Y4Kc{9*_l10XbfS;5`swDqdK=8_p(Ih)rSOSHzLAh#sU& z!Y1YBv9KC^A=(DyC6x{$9=actoP3L{IGy|>#J<>GbY-d0Wz-(Zki&jTg&L%B0T!x6 z#7^4G}ff=^^eM+~{3D@ZP|RXrW0aCoXeAUr_6abmr1^ zYi5u_v-PIKs)T$EI^gUxi~ot42&qLL>BlHh4l|G?c?P@vPbvUdey@aG3k#`TttY2i zm|7t(L-7UabvI`@t?X>3KwlBmUltv{53?Tiu#>I*ioxM`xzow(N5jZ;VRfKNEL1h^ zH6OC%G^!zFKrjcfAI~~6t2lvHWZl`S_!cGRHarU--l@&ef;$b^ZF2ifXw}vnX}fA@ zY&i*%xp@K{FmE$2S5Ef5Yxpc&2GvG~$L&fh(F2WIsUOlIu~Eu49UtQ4C)nJ=&zLuR zM-~yrQO0G)qYq{^kHQuiTJN2VHj99}d$_76m_eN$#k#(=mo`UEF(Kqgi(VjLQ9*xf zG<%9-{jKJWE>e5~E4!W{$`=FDW*+yrJAb`vm=OAS?fd=@Q7>RB+7Y52vM6D}D%bvw z-|1vN4twZPii9aFd_m0-P)d*w@giy)u0Iu4t%6O`Bv6;;JhfKb7bJ3QejHtCqg*Un zbdIzE)WEH*)BEO87XTRB(HT>L?AXVc={*Z9KdtEJOTYA(cmPV-szFcl*9cxO+v7iL zLA&bHMP+Ixx#w^8zs*s5C6M7Uhpck5oq9gS_Ii)0J{mTSozb5`jBH56zpwx%ojHEk zC36BVjxnHr_g!pJ(75w$E}kqD8Mk_|(CA)%vwEYFsCA}^F zi?eq{C2GjW=Y9>R#n?+o*Q%Ulz`}K<3$gVpKNFoz&Z&+b78CL@sM}|lSFB3`5=3F? zwm)qFyPTUD56yxPF3$0Z#leRc?@lz(HPOS@KJ(1i*sDC(T4-qC%q2Bv8PG;iO*+kX z#eB0zFW%eTncaCL+5-;fWVq84a1B+fVkRh43W?MZ31{~R=VK>Zv=JaBy#5lvQ%Cff z2?A-%;7%>}>^?3QQ~24;;(ldim9}`O8=~+cqE6vP>s)O5P<20R4)OTq`gFe|1{VBc z0;Ra$-dxCTsj=Sh^4^k^cFJa%*u*m)vO)uO;bUtr9W|G)o8*Yt%(bP(V0ixXPh&*= z82gXcfzp_|qbsY32MrH9%6wtvst4wc3G0SZvO?oesbdwmUjLgEybPgcqXQrkR>gz` zK901{>oEtqXffRI++FzYdnGAyr&!$o7KPx&;R8AdwDf~`Ic-h<{0i*BxM*#4lu(cE z@%uh33EdHWhLRlQ&@jwWDn%^j^Rd*P>$0-?OUbi?Ox&VZ=O{QT@tuWeX5$icaxGRCX4pV=0-~^*KrU}@vP|ipf6ZKo% zu836p7zS(pqUfL}#^K+95zhv0aUEzLKb4uc&8XQ!v2WrbpI<^^X{$Ngi7s2JJBZLQ zoL^p?WQ@unmguom?BNkaNiLZ@)hLl@KJUW)Vg+vj{d`UQ6~2}_F5#uK@ueK(T^rwv z`@)#w7@U3k)HK<(5NqTvGimffogy9$60~snGq*EHpej+f!zYTs0&c97p^!g9LzOQ}meJrrdr1w(HjI&66YtjvCI%pxAo;k=a}<)af)z9~qLLr@+JuTV$MdyWjJ^a6lPj zws(>x!PXq( zL!hL)U`TXV1i|cd6()`KoBqom+oLA3)y|5H2u^h@4_vQcCzpGSZy=Z$l0?Ae{=i7m z4D}L zZoXu&sB89$pe2aC^%f>Nk4rV8yhB`TV@=feMJyaJW?6Oz_sBvlf^KoFvlsNXKIC_o z^&4cS8TehvRZ`b$AFHbB49dcu1()$JUuk_Tdx9-Z+!+JS6h)uF%coZeVe-pAO&Ul2 z#FS4=ootoeVN)=rRFcNw(Yrw75+RZC!SDAO398pF*YGQ^Kz-kXw4R7|){g3xM$caH z*noN`($r?JUCNG)4G4O3D)6*`rnJDj`J!_tYzmW^GQ2wiP~WKcf8;4YjQhg}XPR?cvC61NjOv{72N9wrEVrW;Wt`Vwl|?DbbCW#p+Z7q22s&Het04kEF^JA|KA~8I4fD< zxjX2(h3%RllP}gimC~vIV<$Tak z$j3Ql+2<);*VvM=2linmf?!44)^O6|-$z1mpp1q-r9O;A)Xg*6c#t7Z52=wKK0f03 zGi&U|CtKq=k&)ZH98gPjNguWr(6(j(C-7 zTmm(i{z4azdLs{(NOOyCo@)U>81Wi*sc5ouo893WOcVFGU$`0tx{5paH3hj=0zPbe z^%K-;rnCjmIohyriogYFra%U5unPtZtSAJ@;#&=jxj@WVmJ8{x+f>Vce{f^fkawcs zvi_Pw+Wyd1>#3Yq2G!d~=(Y>HIxb&2Pij`ADn%`y9;nW@4?REA6D{^bTJfX85?{jX zH%Xqe*k2COC)F{fQ2`dalD@bzoJ6l=00YVHuPuO+0ag(;TASs;`dB30w4^T3o|weY zD8N526NCO5KPxA9JBGoUHvRF{;SElOY4=V!J+=yq7bk4p+I%F|sI(H<^ZP`)4#z^;uf_d@U^j=h<^Wc?fk1rNy)xB?T`Rh`45zWaFK<59P)$5QqQcl*K$1A!VM9c&o1eeyZM z8g60~Qh-z_48rOBgrjnsC^ZEw1?^p&JW*cRv7fqlp!4T=NexjBpZ|}syN=4LYa2x` zAR*lyBGM(@Eh3WA(uhcRcXvuimvnb`cXxMpH=MP6p6@$*fBW6z9DZZ?-#xrzt{K;y z^SWf$9LMtN)>L`{#IrR9`>LZo~Y+XfYO=sJVD8KaHPT6iGwx*;%UYns0KVH_65 z$9=s!YfB0qTY3)q?>Oh*BMm#7bWDoSYZvQ?jN#3T$eJrnMUD+GEgxRdzg*1Xt{yKD zXkF1lUKz$geLO4OglJt#O>r+;OvjrTk4c9VRAyrQJH~|va9&IDbiux3PQl7SeAS`S zl(TQGZQ-;tnb4OuFW4u-e>QQki98@q9-6k%?8hRt_*R2Uo?ZaHm{aur{h=qSKW&vR zaeHfXa{Z(Q)ZWKB`N$Y|*BAcDUQ4=O|1=GpWBv<|KYpUR62fUM_B%RkD8fY#KVxi1 z+@4*cHn6S@Yui>Uhtt`5+QQFCm5CKRbw`TtyVPG-AYtjUj;FQ1UZTn0dUoKEkOQR~ zsEt+sF-_wXvGiLH5wKh=SaP#iBF=u{T<+Q9 zEm792!{)6ot}DNzXVq94PhV2#74I|ry+rhwbFY3!ihX^uury@;KJ#0Sg3|)-zCK%! zc{3~p*RV>BHWq8bQlXNuV(kKDWB=7vX}hCN@8CpfNe|P&0I3BB=i){^;{C{BS#wN# z&Jp{*>RwSEhJ(+Oh4uiO0rA{?b4S$p#++DrG=F9*+>)nqJ)>7rB%yyHHR|p8=<~&Z zpml%}nd!^;2>@8G;-mBdrN0NOnap|X+UU^MSQZ#4i~NffdS%A77=Kl%XC+cvUc@bG4ymFOw!DH` zQZZK(+o%d5&Lihtu^^Zch3$*g*y`kZWwu2LSJ@;d3Ky`u&T=H_p4^!M(ol+wV_SwQLjRsKwYrpWW;W8R&-j$ChMggB|} zv*X>)R;qhLZEr#hDCM z8>by)a!~+VI&g14cm1w(-a|$fge<2vxn$Ya%(WL1+ zB7SbhYWqo?gmcMr8+xa3tcN%jU18G}&{8NCIXSM8(T}dP*<6v~IhJ zOX&?(CZf#Jm}REqO?hiINt}&(*&Ie6+(whpw0*1e$jY>IwnxWR4Kv5W zEH6KwhaQE>H9?)Cx3>S=+_Pt(j{1gbPwHJ`UP3b|&+GZwZzRQ=h-~u-{Ns2~g8MVk zIb1?Tjen59NS<02^{p>S;?pUyk4A{WgA>Ulp;=@W3nFGP)?t==ThOYS_vH~+;_7uh zWz2HF5SMuoUjJ2^|0jfGnkk|BuMDE+=LZ9(u4S4}U&2uFF5a;pk@7mAMnxqcTcSj9 z>vs!cA>L`Z&8{$(X?<_`Xs|KYT@gxoyLeu=WsT5%g*&6GjuK~MJ3b|@obK}JF612V z#W{%vAKyt~9@;8ut6!qKt)TFhiwXX%fcRQ~Fm!LiEh()vlY)xg>2fYA=ktW)8@v*e zjrP^OYq8{v?KXVl)O{W`&o(qN+KuX0A=(0$?I~@7ks@9z64Zo4e79yb+?Qz_PqA)} zJX_O!z5`7VToQ&Xb)>S3Wwo^iB&dCk_rJG!xHfxppVZP!d%{TS>0!GLg1Un;Rvp%L zrkY67`pN{VQB`z2Jv)0LA6c&RQ&nJX(jWxC-n6s$J_!kll@$vk_1697qZ&Y&MV<}1 zp#?|wrfY}Kg+lt}ee2T4vWtZu-WtPqu7t79yW-FkbD$gM95(g(dDXfUsP!yngSJQ6!u{px>E?lIFM)So5L zM|Ge7J%96;MHbPI)IL!RV(1to?;$Y;Tt zsU^|!+YRYAv6hp|o@K6~>_l2N#9#qaNp~ZDgnAO-vFBidH@i;l0Sea-*OYn81D4>O zniyu=1`SHZaSPN^L4C(1x7V`Yx$~v63v0MpW^LmV5;=whezXFVH%Ulgh`xJs3JkQ= zc1?6U)I#sE{qHBYhT~k}c{f_NlZ4u3^rXTxJR?JfPG@m}4xe3TR zmM|+7!K&HH{E{XI6H1PS>Xq9c7WChm3H?|v_KZNkFtTyr$c>Zl z(K%G4%8|eOG;+COAjY}ovK?P~uyn)F=9A;#ALAM-W|7vZR>>&3dBIF#8so^TKpL56dXK!1zr}4ldz1 zoyYE6iBXdpyrua{)`flOv*ETG25$?r+Kcwx{1eaHmXBlQeg5ksJyvjyJ)vs@gB#}(~r@rd(V6% zxc|J#kw#I1opUyk5`PhQes}#?&StQB)-kMep<_EkzI>bqR-ciTb~KQktUH z(kON{-{PZA5#JyuuloYLlx&PLg8WNN>P&Oqb@hGXzU)eg=Y1MCa1V1x&g^Ff1xWbIz6cMv}{ecqsKk1^_uRNil?eG z-~O^U&6R5L=_5qBan^8{I{dd#d5Wb8f0PM~>qYlBjXTKB;8Vt$YhWIeCs=e ziB8qj9WV7v`>9c^ncUZ^Bp$7Pu~Nbhrr+A!Nh=f84erKfYIauu(?)h>tVJ(VH_1Ci zG;_F`-zz`1P%*{CeUVImoH|z?(1PBx1e}$a`*(`0lb*eLBE1*6^!HFztEx6KeQ{{% z--;ffj)hXKsxmU3dcMDi`R5AtalxW;U5Qt^naGk~{geJvJTZzB@9x@*dcxu{1m z|F!DUckyB0s35vB+|(`ZFjKWCx)nTPTftx*4h27* z-|y#7YR%WaJ9KMYnzU#6tlU5MxuUhY^yE?K9QCJdf-b$p1%V#k3X3TMIl&j{Xhc5= zG@a}m5@E{v@q=%M&4hPJ&2zeMb03Z|U)DzNB_9mgThxD@ygVGq^~&XDltTGc5->S= zK_G@7mMmnks4MW*Tx=5T%1(3_Q(9zC1>UuhrMW3&d>cX(np5A(?dMzFd0=Bt7-wyq z!sKi^yG`$q763AMd0M)&?;9rjqpems`kJ^2#6|GZ>nvk`h~cnic0a$nV4co)&U0?- z48`!G#Po(t>SgpkA^DX8iZ}$lU6ui#vWaVux1TXVv1u)~3;Vsv<($gBchbuoUtog` ze(qP|11(~A?(>Of#NUHITE*@LULJQl-d(#ovO%jMRAZa?VrA1xD#v*}VLVV?%Xnn7 zX651Tl#n-&0ifPkV{B6l`p#gN&}*s*st4C%Thc%KTfq*%48U|0V3(*i8bOF^1l42O z-i2apGso7E@Pz~0Yufx5@hE#y6O=hJevjYoD$ezvJg*BQiz17>HWB)8A{hm*{SJ}l zv`$47uM+mzri>xQ-~W1ztL*_Hn>at|C!xNMKHA?B<4)_2>+oasm#0{(d-4nSi*1^X z^j^|97t=^THow7>(^Y%j-ZHNQ2_ablyV+D`+l?8&GJ6IC2#RzzurrCV%Jjsen+ zwl^ANwe!qh`1fGq%nK$z8WZ;-#V$tgKA&8? zXma%EPh}xSnlL(^k@4}H+}uGRe0sM20K+~0D@^OFrsih=99W3mp-xcT3BXNNqhwaJ zh)?(kYUP^*VA9pA+Ener+hoydyf_HZEVFT{?oma{mS=O~!b@+;j3B(uAnvY`qJ0%{ zZRz%O2=9$}sMEe}_|1ByV}0J5UHbUVPQPyoFFvvp)&NJ*D|J`PvZXmA+2Y6)`@8MY z2CJ5I5oD2&tzbi~3h(?g&WTSkQLSg06*rvbvtuR&!}ScZPv3@;=0!-k*ZWpsQMAWm zV$KPQg@Hm5IWvz2S&ZpNwBZiqae2IXN2f;-vs<=3vHlPktJfzpsr6)CYe(-aT)C!= zKl4t8lpiYJP$$v8Nw%(JOj0B1AM4Ui9J_f+gV*M6IgiZGkQh-uD4y16;z4XMBH4LF zzm(ZMRP3X3J`oT~^kRoL^>X#UCYrbZ+(-hX`x(h8{iqYzj;>N^Gd3C|TWM`qt?dE>zfOMV-@U>Iw_T`-`8Gw}K1K{+*yhJqGP$cw_n#0`sjQy~DZKmb#ja|0_n&E3tUn(S1JX}Wwuhwz| zbAFj!S=nn3Zi}NV&a}lLMA_;$d0G_!I|Qjvc2^HqBA)8q28Cz-fF-X9TX zo;;2PrL}2xOm_#Tt#}J*B)RJakMTc0>H0BvGe|k}?j1m$Bs5oiUV*xf^bG2-PV2wV zzcG65$FBKg8^R*F)XHRZduo|4`$yKQ*nvf%QFpXtT+RZHSl!&`erH3GP`R?C7lPaN z#BO95F@!N;IM0M^IUHO((o5=PGt1+t|GjjFX@w4I-G?n8DRg`FHvv?0MvrMEV3a@8 z_45O2Zf*7Ye30>4=DGM*RDas9TmV(K;%)W0^j2~|4>T=)kwN93F|iVozh#?dCtc-l zq?Sm*oGrUTABLO~u2_vn6Vlm4BwWHz!!Q18A4v)>5tTh%s^jl*DAE5TdDq{BH_>|f z^(lK!*|a|RegLog&sV?p$%MfCUK51T-k{a29*OW+;#K?ZlaLiW+<&qcw{^|~R6Eab z63xw(ZAuL#CU9U)9)ec079?t7%~tv5ohxq)X4!2QCQapq+9W`o zWtnQ?HIZhe?@UBzgFZOK-Y}eK5XUt#JL;XoTdiz054ggf+kql%IQJ~Zb!LIr)<`Y2 z8t2SNmn6TPdklh@9!1lv_T@!fhWw57pC+rRt$ch>Fy2)`ng6gZ}coH!#?&yRwUcDK!lOD z>OY51^5mPVe#JY)`@#JP^(y(<+F~o`j+UQ2Nqz7AEK9%|NzMtIiK5|}%-(Q|T=+Hw zz4=kihX9=r&lmL^m7a74q~^30jfJ`S`4xzH*Gd)xmfarhPmkS>SlGCqL+G^MP2`NE zGA{dr?XQR?snDXlUHr5QN2}Go>BMQ}YQ@sqea`y~=d2ThvHR<7x7)I(p~Mz_OpOA} zoCWAqVAZ|#|rLWf5S-7B7l?mMcb7T8S} zUYEmt^Pv;&yokSJxzOLXAiWVj_En!P*Bu#$Gv58u6Mb*J@LL#`n-%j0#?08JAs z-<;*TzWea=`(S5WASKKk?NQ8if4Ex}wGXOr?k_ELPc?nBu(^Bze!~yYwyF2=nJX}r_mXVG$$K#-z@dmkW9ED8gNXQ zWj~EI5nMuD0d#yO;BwHUQAgR*j0KgcrZZH+zw^Bh6K>jz85IwI~XY`eePjig1+ee^55@I z*afLy10-v1+EjHl?bYoX`hb5!&hT>o@*;T&fHT~g8ZJmyHu9P;v5uDmzw?tE4XuV| zJJ$~+-}yUTdPz#OGTnaNi|%_{1%NsyZ0h%HCic?QKGlSn#gCNaVC4efw$VW7Jz8;u z%v?XL;8;nJW7?l^w6Ze=5 zW=mFRY(50mTv!9MSwD>_|m^J)PDZ^r#3R$XI*mt_f9tbw`$F<9rO&d;EEn;c&l#oMyaY#%A;~5fZdH{0! zAA#_HUvPF*UvW0v%|k$HDl?x_$+th_E=nm1fh`j3e|;s9E#zyP6%>Eh(Gh4_)<>k% z5lQyR{%GyFf7Y+BVb!(`sIkn`5kp~bED-D`mDjVpT;&msZ9bsK=S2}~^f6`@BBon) zhJ_>Q>Yq2~gX&(BoBvfYBOl$Z?aJ!_TKP)mwNd`VE4k_{6N$N>3gi_EKl29VR0hzo z{(S=U%0|DZcM1QK7Xc<=JUV-y1NjL!=EY z`=x}cX}vDXuoA$ktoNm>MO@B16Gj}#V=Uz7<<5-6LDV#55z1!Kuv&7B>N*wh(twhl zoKzWsR5w{&u}~s5yWAHC8?VYpR`v81?i~4Q4jrGbJw1vGf<%g4*P^Myd+)l62?Tty zi>t6z>)U5zHJqn+k)@obaKG|?MECfo%?6l{-1%I**}>3;Ku|fQeXu^*obD@vy1)ck z+>LID#G<%Xx@#UcE;^DNkyttQ>ao?wgm%z_eSrdTGXuytyx_ql|9|!bUsQYL6&@2G zM^F6S1TMXaF`Ok2nf`4j@cF+$cU#LgP7`}llq%VElGgDQm!``#jw{;{Qw#+fN<4uX zf1B9LmyK+{-|O!cRDV>p(LvoiaL440!cq($p1BU8jpj9q$i+L$8p)3(e4>1>`Oohd zKIT$*r*gp*N0A5+`JUX#9ccTWVSRw}z$4puyZa{q-G8RRP-{sEY z`FXLf&f0os`ve6apOw`h_dY~x`m#uESM#ir$7@|Z@<5s=X`NvTu`Yvpa1wYu`gM`A{B~0 z^UDwvWYKdKc0e>!@WKZ<#gAsd-f0pr7eTB6`S-@Ndp(jr>I-&v8|VYOv&r?WiL7Yu zI|Ve0k;{}lKKTHaP=GH``UNQdQtEF0vi!_fS#4lCHZQzR`L|`BfKT;_iBBaTSfk-H z??27`mXJHc_R~QKjG8j2R(4l79l`fKgRsjR@VWe?Nd5>smoo^(k#LU-;6wSld?@!z z_9p<$bos}wma(*Y?|tNa1aguX7lrdN^Vq~cae!BR@}R&()b*8t%5X!H3I4Q&k2Qg2 z7{HgP71bpi?2C;cG@}Mu|Jxg+n}z8Ujc}(x6N`gVj|J8o+RuE*91gE%kIH}ps*D0< zkDXShyIW%e8L^!g6-PQK&S@G&c2)CJUK1HI;rx86?eBbUU5!%V=$igq2949Es3m+# z1XwF(D?AeXy`|zp{W2)12V_~(yLo)Fy*${R8S`i7Hno}G1uD~lKl#ns-p<5G4$0)~ zSMV7|NBR5aY%fwjt{R8W5SIWy>xN~|M(Qs(^)X}75-i|*>BtYy~=$~oTFg9`bIk2k(PeAEVqV%fYk`8v)b94Io0!uAtQ{Bo0T5Z;U(;kqgxJQqnW{`do z(O#H0HLfZ0EfeWR#^ZU7e8^pT(8|!u9dFBgkl867@Zy^i2zWyZKFw7ZU!-Jj&J6ma z;m<~NT$XPe&m3_VtV{;V=+{~+p(cP!TGp$ z>fb@K8goN?qP#4-B^|?~c-+d{;plgPm@cx0?8}1*ljS)s2&A#JT#l<{vebRII9``! zEdk4pZ_c`GKzAVHLPVY#?IL#`*eT?qHb@3)O)C`o=3S1Ca<&rsPrvYC-neqcfH5~( zmO5R|t5fu`-u{M#|A(!YhEg+8F%InXIiC=EY)SSj3y6>c{9K#*=SC@s_H%p3y63I? z*aMxv_0dAT{<-jq1QtZyq=el0TZs>83Nar~fu(olCHRT$W@b?fDqZv@d-$cZ7TQR6mx~K-jtvk(cdT?7cHuxa&z0*>>4;^I|00oT^nwWmTxA` z6Usmq76Wbcur-1`&_p;k>TfME*INQ5g8}kBC<}ZhH>FFVJx`Tv+7(b#iZdf4>aljse%PDR&50yQtjoDnC z-p+5?Ec2LV#e3n|-qK$>Wa+Xh>7~Lts19tWMyrYz)CY3YS#;5#X?`iZ0}>U2e?OYR zKF&P~?BN_;PeF40mWHLAx=PTp-01@RV zAApFpiVdLC<&A}+#Os01^W~7L49do6e{wSK4ZU9#?(}zDLKmykzBWqjHFE%puM&=` zh5?&R_!02}A9lYCqd?Q6jYEM8zz6RQ3;NIZ9%sB8j@=inPPtdI2kEOPpoo?5_HcLs znnyTvHqdzujwQ{yLgyPz$EI_qEog#94hQp7zXWol-?`r)3!|03(2}dbr~-sl`br~p z$S)372;Hq8ht6aReikKE843Sl679ysgHqnnNVd04DP#e<)b){46$4-{N4o2D0Fd~v z;fM^`mS^%vmcxL#w6B!{sx3QQiX@RA=u11z<$;cK4=YPKTJvmioOK?MP{5oe?83K} z*v{c=?)3$SMkCh-;?ih;YO)}{#Xa5#K=UF;-VEgedyDL`aAiKzq9M?t9*jXRUC9Z7 zQbeLEQ0ccIVvf?phxxY{VieTcL;H>IiMMDlHoa;k+#*UHiIu=>JSLEP;|g~I2z8ty zQ|dE;!NJ^gMG@ugK4>#G(GXnC5M8EBIV+ufI~{#We-N<`kO&3D+~EilWtm9LC6r?= zYq`2#HMtm}uA8p`;|wrc%q`344#VepPyEk>&jmRqwO?in_6C2+*ro)HGun!3OZLn9 zyUy5Ei-EP@9rQ8&CBcA!`Hg*k;wqWf-I5{kxGZO#Z=JK6#x=zefClvF=($5iC%q{o zc-V0uhD0jCn4_~jxx!FVr(xZ&ZTGn#S&65Jee-JMAa&-YX24$wxrk5|PZ~M}o);^w zrD3=!1oTk#ua0p$Oz5Td`;VUC}#WFRbAOKZz-86q=q~%suN~@#XUMzUhy6o8m)mG_Y z2r&UUnk5y>t});1geFX+H#1zEtU+K!WDOwy_X~j|o!oqxtDm#2IB1`smdW8wh9nRD zf*->gSqpbYz8eoh9S4_fJS7TMEd@<-*$QyK?$*5R{m^_&!};9zFc%q4Z9bkF^#&^_ zjvQPsTqR(et2P>1)?(Yx@2aBm8r|paWrQE&Gk;?OKv*k)MkgEc>?cNQU&lQ&J?Y-} ztm=ivvT06x5H<&2m!jy9v_HG3+r;iM zOlZK!eBr0Q+su&gqd1mWjIEWK4!y1KPo;G)e&qVlWa2b;S>7q)LV#FIpw3Z?Tuls0 z`zoE0^oc7l`1a%zus*ct6oz9Y2Bx94og#mib^#;52qFQH<_*n9_tmtR1d1jKWnY0O zrP-&q62K~2yWs@RDG6qjfI$>kRTc7rl4?#!THGI>&ptl7{7ge#M}cFEUB#V89G}r= zbXJ2hvZt;pEF;}zm{iSdA*4c{e5};Gi#R9U5@{730ifXDv7ei6BJj;$8!Tkd#|^V4 z8*;O}1UIjWwS{|;wEdHz4&g!hM61dD3OJCNH=FrADrnVh_$lduE6zY*%36{MP7@XTiSu1QD4O6p`1f@Pp<(!%M+C%=7@C zHTUXvYu_DsnNjNrK$Jtk%klqZMqHN;ir}WPRoJKcV8;yj+tfY+-8a5s%qiG?Gjy^w z=Yf;C=x>E;pt4)c1Ay^sD*6$S9ijCB#a9I@&1|43 zQd3-<&J9OTvQ{Kol--=O5Mme<>QQeQ&$QNCUI*=9kI<(eK&7BK@UL%i5lz={b=Y5DhtQa)pKg4Et4S;j2Q~+N-1z>@c^DnXbog`VrfTd)!y^S{Q{at zhE+LJDKNu{$s9c9!z=u@4kdV2FibzHuYcAEC3bG!$5^(XSS>y zW@d|vxXkQ7O$KM-fUz`7u5W|HwXtdAI&6U_t?dk^hk(#OwvN$SYx|4(^Gr`*21Qr` zRm(@L)IM-gAMBfUPz0d=JEirx-46n4>{(E+%XMbK8O)~Z`1=tFTBmIb4tb`@Qenb( z)m-E*76K4sf|O`f^QaCFYb!Q}!(Q4?&hB#og2x_<5Oq*L@DND(Q|D9B47g`)O7~W4 z?5~D*lgD^OJX1N;1}5Rp9^M5vGdIM(?qk@S(d72Lnh>s`m4>NEf$0{Nep07ey;!JF zH>ex6B1P22?RZr`9#c%h_e>HcEyK;XcWGXud@sC*kj2>ggM#I*g>WW2SL$oxpAT6F z(Z%xo+?r(gW4jqvRpFETau2;ZSH^$yq?9ZoeAUhFmJE#_jUTr8z4Z>5y*!e;+}~YZ z?7?1UkUaf67-%irIJc}kJZZ*#GyPc=Kc%e`=u3|M+v8O*0AW&UVKBqkj-9bfP((Wmbqa^sTA{pXCA`KG6aeRP1+L<4z zO5Lf6t>H4aUjA7-e`~Z+5UOLdyE{6TermDW;fVs<9!Vr&M>j`$LF03Kz`1BuyJaD< zwZ;QxpS!HuJL0HZ8Fll%2j+YuaFWXjjckqw{3QMkp!|ODLGYdLl5Y3_Wwo%4)O+R4 z?O!C4umlw<9E2PCDlq(?)if~dbqaJ{G>n|^w;)u+4r_8dQy(TY8DAIAjb2!e^ze~* zUyB|E&*Qu}G>~9|O5L{a;}7(O<8SnqV;xZ@NMzi3pO`k;M@&-|2vu3ZLKoHs3CpZa zhTLFnaBsX~7Ipnv;W<@3Q>Gf*VTzUu>0 zJgR&ik$FF_^(Y~(Tk!*J-W;0yN#Z#A+H0n2Q#FV|8{lL8e|x*CTPmRK-um1shV>`l z-Ts8#_DG_D`7`f7#Vt@n#7Uy~DTt(#cr1@Qxn{R3hV&CrFcIl-RX<>U`58ON5Rdd^ z{jhUA;>&WTzdq6_0);;+nEegz3@70KI;kIlt2{3SSS%AK-~f8Kz==jPEdtregQDLe zhNB}_ootIx{F*eq(1eKNH;I2DShrq}gHHg`ukpE(3Sa)2=QaP{6&nlxLn4%w5fi4w zeZRBNHyc?-u8VL*na7GkOP-s&J2}Ly^Rb0LkF2qJk!4l4kYJUE z5yVuQLBXG*plYv_C1<;nIU;8EA11xGyW=i;{V~+a*@V2;E{Aus^4rRAXET97zs;6U zvxcM2SgUDas;;)wlX3mx)KlLDzzUWrlLlm`(0xrGj+7c-(%KW8^zDSYL4d^I`dHFZ z`~<6!VvaoEOW71PL8jYgl`+@L?58cCQTk~MS<6g)Kq@wofy37^2(;op7?;WLawByOtwJL9YpcO@y+`#?IvQiI9*hEX7` zE+`~YO0Q-p$*HZg+XL2wP4rQPKX(!I+f1YSlm`ed& zoNa3O-^DDiqFx}p?T_7@4S8)g1x>leC=I>O4Y6Fzj=SzP4S zR+z4ezEqN|SEZ?cLdy^Hj1(P2UI=+Q-$;yRrq+{#AN)E55F+qrM>ZvFeGYg6>X6(m zxI3z+Ht^fKGQ56(1$4|q8Y2|azY>!1PJ!X`m7kJP6j0=Ac{fI4Y+psLF>o>NybN5Zi3EqGc=6dSj z^aKC{NNnM7xXMJet1?=ameDEt&o08*NRB~103H^ZkkexIl>&RNiR|1fyTKJE_a2vC zv&g0Q4SnV_0_PjQNWvfCu7>s#r#QSlA|c43zO=Vpz>bjcqEP0i=DPji_JIa@$Y`ZD z^SkOy_I@Q33-uunj-}Hq$7K(n=QX;Znzb`qs52!Yzf@w-sI%FUgR@5gjI5Zsw%&VO zK>)>fIeRRT2@C8B=gW+kQ-uQ<#r}^Vu)}6Cp|Fd{K2nXe$sD>3tnttc?d$YsCh_N| z)uGJ6@;>+Vx%*5@s|+|OIMz^Hmv#u78N;EO0*EkUnoZXk^yqIv&iD__O`LRz6(+;w zT%`_|>tn%U-ps;y=^-AaG&+jy`K^6Mlm*F;yHG1vAqRLfEUTLjQ71CZFOM5^G^M$% ze;_SdHlHGIB!t6PiQD+RNQ=IyR#g5hC-%Li90%%8i5$RSF36ckl+mX(<^rfJ(bkQv zX6ym}pI-p(|MggGXWGbdQ})@L2Un33HQFxvH#a4tHxZR>IsOmOeD-*opLp z^#MF-$x1>RnbC)PlEZmOUbtX^}S0|h0AEb}G(SDQBiKz(B`j#T-;H*AH-^ee^ zKmdV7dR77Mix1XT+na-ZsLP~{BDZ2S`FQ2_z00J|I&-;t8NWmOSBk$%Wc>a_Z*hFM zI9JD#)aY=&E`BE<4R;1Dnhm0ECz|0YwpTBCzJZ%mxPee{1aX?L2)Mo6K5Yb+g7HrB zChE!PJ}TD=FLv!z1HIWo!q-S40o-?lVN!Lr);$cnb$xD42S}qqQ?rE^TAFUm`9DY8 z31BEYPFNS^k*pGS*l*Jij+7PDM$O?dxqm3k|LLnC3D0uw5d@Olr3}09dA_%#9029@bD1@}o};h@A{9+N z$Cxd}_r&W9tNOoLb;AUr9BAY6tmS$Z>nPSS`7dm9_}%|v6TP( zAvL;}HzA)xIk_)aAW_6%akV<-t&Aevc@(gKR--d|Bvb>t3!Cr#4>id0!WR{4|2tTn%e!sSo=lIn(=nh+o%m!SD|7=Z;#s!yuwFX@iI=m zAH;+b>l6o_ay+bxzX$pmx;4cnPhbnM!H{f}=1K{kFJ2axI0+9~lPBXFdYTF|Ma?(7 z63J&iSqK+NPw{oa-oJa;T9$DWxO@7c%~tRYf0>Wt)hmxz;va<+p6-qkv03R2R8d*C z4F<>Jpx}*ttJ?8*CA9AQxiMcDME9}t)G`^=7ac8kx{|eZz&}#UugV1_CC$zE+?f%= zK|cSkty@Hm)+XEYeABf)wgkAL$zDp1v51a%e4-6i!Y_LEg36Umhs30keRMm^NlYikyD~c(KyU3gAF9f4A%*mI-+O!vvh&Kvyam7LBVy-%niS9LT_iO5n$` zG6?;;eRn3h?%4&@&}Lm4wGo*>D&_y~+IbIGYv#OkT~8lUMySixjlM`xCuNn`>5lH% zB4!@H%-p0GQT|R(v>*PP7O-bgHwkPQ{pZDvZ8n=|MhvhmubMkux8|?2GrmebAhq_h zhO(ig(0?~Ma13QUeZLWXVjtU&Ba4MT@@cKDVJC>h6Ghjm;pdGHe$E#LQ(QQDJ{qk3 zZOAV`Y2Pi}tI*&uUjX!bJPLqtf*XMUbu>H`mMr`ixyZ;Cxf6lF_Fo}CG7A~OSlt6>DYM=5cc%$4#+Yf;MS(P+pr6iHNB{z`udtm40MZAv^2=>1| zIG^4s8W^ZY7k1(L?~NPjo8)62S5v14Byh2l*8F$>21qMY)TefH_jnn5!KiiC5|u^s z{N@s)iZ6s)sHExoq$ji8_aNMf(Ax9ie#CFl!Yl0#Z8~S%ok1OkHJpmp(8)vIZ0mdp z$1_0|pj!kYNs;-=-rKeQp9su!{k!LE#Wmv?<65>&i=qv_sg>*JHgIz>jH1cc?Op243@Ti;xmw zN1A)B0c%S%X^U!LKY5(Hgof|OhG>in!!vTkTXV6vbMs);F@VI#;R9SCWX{*?<%K70 zCgj5a34@zCMmsS^{V?or0|iMtV}FdRY1O2?ahYl89&1BMAImLgS@NEhZ9^bXgLDsx@1A6{CB@2HqBbziae^`H74I1r0`w2bPk`mO0~9@nnPC$tr_= z-GiksdURT6&e84G{nq_cf77sLwd>rqo?04^bvd!+N%S&!7iX+|8Qdvg?9*5QkT?;a zI5zHBmKy$?7|f3UXTGy;+?SUk&zHk(qE+L%bZtBO@cD$SDcs}Q=xA!bK!CoaHWvVQ zq*diU^^8RqSD`)Z#f7Se27s9V3Hs1@f4bjMurotzrns12i}m^hX_=t$Mw$DZfFqb; zzT9Bo5B-|s<;CvKSVXa;ayIQ6nEfAl$|GJvpCP#pQxnHk`XXZh*7?*)Lov^zMOUO~UHBww8 z3_^J?!!#i|bu<@0)E^(*L-VLdB}~3f{lp*Yf-jWvPUcsR2R4h^pb3-wwmjU~)_;e} zioqopdvOsMDr14vUHT3P>dGmU_!e6p=G2KUQAno%4v+t|U{ar3-#>YPz0{JD8cp-2 zWv93mJ-669jWn?v;w5Jr|1a|~m9&N(*@sediRCZ#&zdK;O*2=TrK_7>Q3%rQ(nHA! zJZwU`3EixrhOJ7~<%-*iz`7@V{sm?Ri2Z0!w3(mmZ-L(7Tw(nEU)Kq+|MpHOR|(U` zswH{<##e#M6lKTZw(dTn4LC=dl?3%M2TH|iT%6T6v&*{9FF;PHXOM+U``McwO2HbI zrC}Q$sEX_1c^+d%*4pV+Rc~W-cc7*3{Qq7|fT4ilbR2(8LMI3wyTjQsU9snl8|#+T zhbQ&VRFSgp(}4iex@uQdkTLYLCMaT@Bg_2Dip&6crG!E3D6(|3z}@qv^XCCcz~IWu z`e|?nBXY+WmY6c32E>Z}L-WGT(&||SShMlr>A9yd%#r8mp{Rk^ zD{PRHd7~H6LH<^1jl)-O89WLBcH;tn0x9sFd~e`Su#gv0Oq|P)-bn7Ur!V-JtrG44 zU~?W)kKOA9e#8zi#A+N*{A3nk0jv!Uv+#@@`V7U@WtO>Yu$_GiDW z*wr5!5pHMjxpn#vA(rEQuyF*8*zMk>@A#*P%s*E?89rRpi17!wpx(XOR!UKGG}HT> z7{N|zD*XH$z=Ed#Zn#yr!PgGqNZY)7Iiv|MwYr7IQgJ>1qY6+1p$gRgsvDKv+96EcBvZ6pm7AlHBBW;G zi;=bD5t{r>E>4c0LnQ$|D7yS1cOPI7`H%n?2Yk}5N%x#s{g+e|8ImY*f3H(vQ_au- zgoDTFztm761$Kj4iHy`9z4bB}paO#8AxV_HxNt*ZrCSs`)XEf`f! zFL_%o^uxOsM)v8`6~UCb$C2~GP4x-svE*WWEeq-|EC4hOWlU86n?cN<@GQ{0rjM0p z6Wy{!->QC2B?!&VmX_)0rtfd%#};V9V2FU=0^k8&H0g@M1wO)EwpcTY>ac(3hpN-* zvNCSlpi%WXz#H~iZ8Dv|U8Afsj4u*A0j3{Xi1zCb-@60w$bJeQ-_)(_xS?2ztowB$ zL2U7ZN=qvNua79_Whzr$CS?Wg$7I&2eQqEAyC3}j+A;v1u;xaQJ^yi(tQ{@<$&J4! zTD2eBZ+*`Ys7mu&^C1iPi}blBt#Y?y9qiYa0_`c^ZaMZ}MDL(#P}WP^5Pi@xYMykV z{-?kLX-kT!*g{w}(HgL5NyXtWrT!t`n~ho<<4Aj*CX3nft%;W2Wuhyvf=N^T^61D0aIwSv$;)Qoio@w zi$gT<8ql|)2NB|MUeU5B|12K#=|%fR_D-zzG2)K&LS%KSTzNo~KI#fQwgmJ$vLyIT ztWJiPkj8)isQ8*LFUKc_WAZarA`;bR)ut!I0$EKd!9$W?*uKVTqnBBpn&+#Ex9 zP~WnQ=~2|)&&$inPJtwo0gT8D7#EGIrdL6JSz&N(xHSD|Z;f+tygAENg1{tvfw+bb z89xC)qEXYjUfpFHyijVhG;S*~BlK%eG#e)#8TNtr(7BWmkwn1d@$?K;kuTp#;cDc> zeWqMvz>XfF?ZSHpcEtbI>}+Yh;mPifMIh;JZ4l){)^xP0yiptO448B~y~0-iV0hYs z;aq*QuJ`eL>*n{0nj2#X4I7YgP)@e1AV?Xj}<7A^qZNnl7bY|8%k7!%M9tYH7A zL&Cs0gs2)YzEwZ6lhcwU_S|@vOs5F;8UuG*<7v=?@NrllyuaoHXdD{X&w_er(^#{p zpS1RsQ`8-;w)!kGrbAUZjD$?KX8X%x+WLk7NI>jemn25O<^FW9BPU&4Au5;R>*$_Q zPC;_9FeUfc8_sBJsvTKm!g*JNoAm#OvbTVWs$c)U0VSkw6e$6*P+E{~xD^#Cr8}f! z=ply&1pyV2E|HF*yF(<0ZWy|Up$8a-dH3MC&vX9IbDnp-@BXaiTCN4UXZHU6e%JNA zYK||V`_1Jir}#8}X>$naG_wsjcaCimT5SYsW?9kyUHjQ(apV95Z}2@lq1Z5*XfF0+ zF6mfGAvfV8&BVHzIMS*tho^p)zp_--eNyyIGyQQM zK!V>|y7N~6n#Z4Ry&%1>1HL-quKqAe{GT~Ug67wlCIQ3$Oa0~z+&7{kP-o*hw!z^F z8ftuwy6tFu8E{6tCO;iqal`s1q~369qio1LbMd5Yh780TOvS^hFI6LR<)28Qt!${LP)E&rnX(^rOvYz z)SIjTynkY9z>3;8nw;$2Sfc@9xaeQu;jpz304L?|xwxM`7R^gmSLPU; z`&s`Ad{bY{dvhnNyxE}XC~8>zH4+bT7sq4I%$3Nm!Yz+G0aiy2iWhNwUW*4Hfj6G* z_Jcv`bG`1oGmXJ-umFA=P3h>sU9w9LY-P1AXkwlc@JW82ydpHv0+>+gr7#rom|8G* zl-%)}Jm#|!uR$zS1qJN7=7_r%4wb&>Vk~0`J_Y?pXyIGh!#^3!z|#Nq?``+W{LJ^e zr;6aEVBEe9pvfg=Zh(_3M(=;uHZ{+{{gP1|O0WsK?@O^j9VT)#L`=%71hCg}Q^ntV zcG>XqaxEW#!-@rHlHmE~@Kyscxsd{>wg1GKbxMnll zugf`V^-gav-_-r0_I zH8gx75s>xm#7+?gjrur#n7{t>6Q}NkUBm$LjoCO8m*YR)1ScSz2jrCWv`=&bm?ND2 zyb1qo1zl|O0IB(9fB&!l>uOW(NXxMN0U-BM9E!`leS34%SHf9~68J80x1+UOj0B+i za*5+V_MY#-eE@H&j+)_E2o^nd5BN18UGu*LmahB(pZfpufG?K+ZVrdgVn{WK3dyw@ z@a%t8Gr%n*Cc~xx0t~(Z+^R^Mf>@vv7M^yRg?HWhfS{gXs|xB+0kS*mExlnUsYO5h z!4HwI=VQe@a?u{6mqtkKY8+GT$p#fr6bsdmF1oa>2g3pa4@n0&MUa`KnGF3K7YcfIrG!n(;!W3R*4YPClBvpl zEzo@@0T%3uDOTZiMWa#HlG+!~8oErsD_XUDw zmO$6I|0DYG=W+je=5hP3ha>$e!t&7yk}Ss_*ROR^-Vt}?^3XWUUjqi6(DV^0K{q@Ih{~uQN*3#9mf>u7{K|}8$k9>Ks#f^R%W8jOA=;FSt@Cj7|`0J=|H(^eI zsp1`ubT_;0$??+9qvCnsx5$eFwsJX(kiI}0p)tGssg{+SF)DAQ@KuPU87|)Qy-5?fi9sQo{iw3uyiWMWACPMAU4@vsO)19eDz!nLdH!ijAV0b<2QU_D|Gm7D!O38K z;^?%h;{3egE79zqJQ9wDhnBFBPr>E!R!+XS>>gq9U@_4LJz5H#zRFU7-MqXx4RA5Fn zz$5i1YFU5B(x?I*BoCj%*`p`1*`b|j%73OHj^$f5Fj%7FMPD=&-_fQ%HyJzMQ)G-#hYqgu^&r&`*< zZS3W3r{NCr;!Q@4FJIZmaOe;+WirGv0toRYq}a9 z`#CuYfQ;CUYKvtw1R{j7JD>8;C;-@&wrcr`cMG1ra0=k@G3aVfhG&g!XlCj~h6=!7 z64Hlv##s&gTrp~si%spmVdCVu{OLXZ9DHYOD`YXeGK(=GDz~L>&=v9O8dZc^zJPs7 zWoZ3Qr6*-~h9^X3e|wrM*!>sc(+q>W_w)!$2@l9%zD{Dc#(B9Xl9TUrVIe4pstuT; z9=#(?0R3SIh>Pb4h?MW~(b$LTakWqnd)h-)eMp!FB@UP<}ZK)!f^T=ni; z9eA7aDG!eHr;29Ze{oqshYT|tiE^v*0A%k^yeN*;4^V&mDkLK*)Pb??hS>O~c8(eI zVEG6@(14Pd@aF9G$zMWMzSFP@F>ANst(MXGcit)p3QnU!pV9-uTV;>mtNkXhkW4(b z8ov*S{vUX|`i23%jvXezd`y-6E*W3hTLYQ$I14pxz~$TeW0>0K_bKmFj)vT>9Xwi7 z>PsRis4L#8%jXoiD&xFGo%_T+0KViqzE^j~&yb<-EY9J;V2+#2rC}>6(vg{r8 zjBu(zWgN<9Eo%E^O4R{w8NO)xi8=Hc2Ef-CFvsyS#^ejaG{ZWVZ%6A% zP%{1$dTIZel$|Bn!*L1!Z)C(gT_eye3wAECeOO3UqYlp7earkSjls|2f+5$b>{mA< z`=WVB9Ic_Ie{!2>hzn`&IZpJyk$}T4fki;l!s|^X5;D!0UWMDc(*F@!G0UiExO6JX zRjTb1GGO@zQ&USl_E9~(yElXaz#^RKs0<}M_U|U=I{s+;NIDqAa`buc#{Tou1T=uW zi#foJ%2ABSdP~c+0 z|8jh9`}ug#NVy73N>bdf`4>l|24g{b|C!MGOmbL)hJ|K=4k3r7=0IWc4MZ%CGs_(c z$kWr#1v*u(pV}`=sYS{+`?3X+(lwSfBlQGB3paIuQBm#mIZu_%lGA4}4ygLMu={xm zZ=-*z zKg$r@CQQ|s;gL_oH#R7t=k-TcqA8=N6=(N>;gsRB#= zTp?xcJM3jOTjezK%f9ufYSVsXW^l`}z`JNSz=3m4v#!WkV@|_OYb-<<0=k$#eRXkl zAO=%oC2}NGJW~&XgP4egiQqSz*nL#|8_NL%HoUh16p$^N?;87u8qo>TXPD*oGA|A4HEhXxEz{LR+^e3WN*i@unAX zlx+3q3m43{OwO5sqvt3*eGkvqjtRRgqhNQ^WXBFQU|Nz`>t%|cGsA%wDZgkwJVd6WFMCCCwYaNy*hc; znSYY5l-JZ~<#2n#P|w>QQE5B_t4gz%){ZOZu{2P=ymYUgp0fdy*$5VmRg4?$1wkcn z@M_26vOWmD%^P@uH|G{gMoseC-{KCeI5P`tfbK1vxw@9r!2=Xs<4TsUHsTk!{wGEC zG{kI8&;PF(fR`-RRhvv@ajk3DnjY>-1{~!ld}RG0lSQDOu8J3)Muv4(se+#;rETxK zG)gZ(i&5S{o|jkJj92{B`mg*bU#_&i`Y4!S6}C;IJyT z1Gvs)SMsF&~>b!s@au)W+q7BOs-aMcPm; zddm5@MqY2D&{TIxgx7RyOqR-bDIL;fa3ow1LT3e^XoC7;4`=#wlhlMWc?!!6%^Q0d zg~s6Oe(!umstzXy%)~R)F>j9*$Kp=O04etWzI|Q)_0RSNgs^zbnU5j07L|WKJI)jV zibFkRMPKYwe)3{+o!G=bXLhBibbZKUr1<^0``^m=e9969+QDgdQ`E zgGel?(j{^-V^F~EdbV&Aq@^Mg9{fb>(Y|FP=;7Ut1_ z>cV>eIDh5UJ1OgQ#2MX#BnPk8V+S}Xvfe6YeSaNNciWA7OSk3GoKjtqYlaE#^YtY7$$e=4 zV(=LUhX-KQW$L_`51CWKF$Tsd0RQ9;r=?vwvz#5Sfy>nYQ*v0^R!zp??=_mqq6&K1 z$aWlh5l5;w)uV~4FrE4L66~{-D$U-lSuXzdk6Z_Qo+d`HFOCmK(ExECfA$67oveEB z{t-Y~78Y}fLc^iAGya4r|53vJDU_JR^-cIXe+!Vq{Qgznldkmw`pJ;tV^_0*5B#)g z-xv7iE|~ww6&bZ<05oo1yz{Sr9QceK z(naeMyk|3y&R2W2G|0+0t};}|@3T^@;4kzfA&i!D^^a)Ny{kyJK{GXq@0|Eg(Wl%X&%!15CYa;e)L$Ah7PT(JMppjj-xOY z#1DVp^lHZ$aM5R#uI>_eeq%M;^#Jle{Q#PAvEECL1UdoWpbnlYu!N>IG7^Cp8?5&p z(Knf%hA(p7UDNHWjJGX$XPGmg%p9MY2+-rFZ&^0*KIO*jSYv^=7F`w9w3HS4?j&QC zpq(wfEB%Pco3N*QTRR2i(1F(txc7J;7W+8?p!+-SJwy66>jne)<3OFHCJT;%8Tmsw zfR$pEe=Wz}|6Ta%z;flND5oV$@p-rH+$d+4F`jBE3ovKfFdghnj~LwcfgUyv1dYEb zO3)SXYnJTP$6-Z3&(~lgk__=W-`vbG7HxM9T_xXe`q}k`Md5Cxr_JH$8W5t!*fY?gR!88{c+EXhVRHJghY`t6P{-tD{M!H3SQb?Oq|L) zr7Vecx*30;=W5jP!d{$aZ)HDVuMdJHuyP=@@tmNhu9?Zw+;QKM=rQ>u$%CS%HO&Nn zW2n{EOBE43|%NNM*Q+$4Is-VO%y1Wlmk^48Gm&x4|n_dAD1qL+=|23Jl> z&>K}|+eOmT#b3V0G~kUF>6MyZUZ}95XlLf!6}dgoYxWpnoo0UIll)s_ZX&1-P)liG z392QEnaNAuaHgTEwma>+wBFV?)R5Pn)o+Xfc>4k}TNYu-3MbXVFu(DV*xdiw<+yg7FkxvJw5vE5SI67hSmtex>Y=U?bhKU2tVwm1 z-1HiTv%=QdVDhP9Evx7_q)`)gqB3dc(93@5TD+%^DRPn4&D*fOx5Q(&TVT=PVHGs_ppy{CSI5uf5{8W=Udv?9k|xmUuL3@e+*7Q zT`;%bEv^J-W}c|PK*Gm{uQ#@V$MRdq2o`-apeKzcVE(Xb;P~S4r6=rwXO)}Kh=~=` z7;S8wQ@b1E_mJ(v!hUOVC2&EfwSBTt%x(-dU1Lfu(cp_co)c{@Q+}OtAoZ4tg|ZR5 zG2r*=`WpSZelG(RC;9EdgzwkaNly*L)(BwY5)o0XlxFNy^& zTA(l~L~eg|JFgzf%$Xbig_&vCMDS>lqTR#$~=i)9B&vyseD@ync z`)gV>!CEK!Q!eIIR=0<_e(I1UMmtE=ooz%n@xD@0R#FaO<-s^^hAFD*HdjEbeP06m zho~>Vx{jLp^H^HPbYwsi@p;?m+&i$!90jHt1)uj5*GG6)BFCCaE_cq$w_$;Zf>^5PFCwfA5#yUfzxw|>A#yH&cLyhxS?BTyIGu-;WfAA zRxlCkL9*Y>2c)VFR-^+*viL21JxBZ2O1_%e1?Qk%*#Krj_gj>n;wf;vxOaS|a@|a2`Ah0yIQl)+6LexE=a~ zgABP8u(#14lv@wGYyE!dQ=Hk|hs(#&uZ~+ee)J_dubmz9%pEqbT2dfL-T8NQ$Vo@h zqWGtN>h}8Oc^Jl$To zD!s{US@Uf5@av)6!FT{X(kzp5O;&NeYBOb8e!YgJpXh!2nSIV|RB;ehoD_LKK8wpVihMQW8N}4t{J))28-_JD^Mm10Z5ByZ<;r~zD z^5U8s_gJO+#9Mz-+I*cDK2q)eZABFO(Pr%|~yQ%$e8s zy2o`Wj!!g87#e;2-@^pIXBvQD%4=a9fg89Bv_Nz}F-cI^)`941{FP>Vs;0Lt!kQ2~ z5BKz2mpq~$F?$pJRCV=vES-j7%Mrl{*&ANGt7+|!!7`;@A?}%#>bSv&8H#dr9)IoK zTX>Fly^_?1c9irW5J&oP@c2z#6+BW^e}cg#EwmxRyH&0~B#}&T5tmQeL_k1Fc#HJ$ zbx}R%sueqYvB7|7n!rviIwJ!gK`Vo?x_mdv=#A4c8kf zB`EwDFQ45Z=b`{;)|cSy(ELzDO%^N92~@9#c%PtlkW)d@m#5e0c6cTYN`kyLkB7Ij z+6VFC{7Wy>6N}FWya+#&yt$|_i2TlI_KZ?)wP#Fs)Q~WPyO@WhFBB{ef06ALP$oue z%E=wDgDm%cpR+Qd$dn}ZD2tRH$-OtdvI|1#+18C~Al%c1<)E>bf-jq691`c5wQuxc zFiz|9zcFk%RHs^^F3 z1j!*5QfAIOhR;aJyBN{8Yw)YPE>GMG@$|L*+cC;>KbsRNGArFVeqqo|`GY5MV@W@_|5 zW@hkj%g^QjgwYPtb(jsI3agPJ z2Gf#U3!(>zzi`%xE@!cn4BZ)(A!UVt6tLrwA_x~ zjH^!xM{wK}Z@H00J71ECP`jOs_&b2Z-Yx+ryQRsF(liGcQ;hH<5V<~si+Nw{)4f@) zQkkvj!J^~IjBRp%no$!Z2$&>Ye==W}vgp|F8{r-mCS3dEoe??!6MD|gG}BJY<8=83 za@N)vuB&|%GSds$VMrtx5O3BSzP~c!{tlIG=VNdyl?PHLT}7^~&m{#vh-1y~ym0q} z)IQs&F?iRurYD~%sgxvA6!_6z;XA_n=GVlAi_1o&^hHl~ulB!f8HXEsoZ^W?-qBVv$WFln!}v3|~P z^*84?Ev%b#u9@GyHLd2vGm5k>&)0aM{q2ck&M3>r?TPFsZ2uB+e7IZpaR9x-gU0D@;kq(?#wvbDdw8%_tuOF@as;8ugg<-uw1@}SNT06x=}LvDQrkNOksEaO;SkB=vDF^0-sqBs84%mL)t zI}?2pSM}IlZJJ=VuFlPEo0Qo{oE4#l=iZeX9T~Zl^Y?;HZn3t2Tt zfk{J(`Gp^WZ|=;NcrGWhvQ2SVuNAQIp@)`)YyS#Y^ zpGfKW`Vkmx()Q)|{VxjlWO>yAZar3Z3dgz_8d0z^NeTIoyTl)kLNOAFmVwK4Rmq6vsyCfd z9mfSu*~`b6p7;x3uqtGmc&z*`issQ_IVmyREYTsXlwkw<)g&B6kE=??qZfYWj%K5| zRU1@CRVmL;+KlPm1NXQ@sk7$`vh`EtT3ZKqmCw%k}Z!~DeU&Skq}oFmR+wni?t@D--_c$ z5q1*L(5f`eI|?4>1m7=av=!ZhX~5&i>dnp!EpgM1An9?bw{J{8j|s+m~#F`JjTPqI01dYIn9B zIeL+5Cx_)(qHtC&NXeNxm*KLzNPzTB|LOp0_MnuCF* z;TQ?7qKA3ox*^!Q1lc`Who8pPVJS0w#&LfJQ0@Cao zLo!mmX4cv>uvYJUNxgVhVs#2BvDOx95xS-g3yJQXDdeQn9-`Mwuxn@r6L^%#`CYoj zH@m->EA5r3efw19?7oPOSlX5Zp(f?j>71eMVA*IL)PM>PCe05is?y!AjA+SaW&e~Y ziXW%6nz7&1S#j*3p(xZvOM6!DmXiq-Q`%^SX%InJiAGm+%P>ruO9JT>fQ2L2ur1$Rps zUk|_;tR2-$*6?rUYL80bJ6;OBoVj);wRMZ)C7*b#SnLLgdkP1&3%`58zX+v@^ z$Y1C75yhrZO~4eu;#v$%RFsm9%2Zyom6jD->%j{PAa=@sWH^}5+D+T7DDka}V{G%u zrE4>fnjQ9-Y|5(=XPH2?I8SVpME`>WGV3Z z?4}EKg@l#op!1;O2tbUdnwIC>VcohLv3pre3SZ;iQQHVvkW(?ht`Ru-GlmTkTUO#?na zhaQ*N<2itg^?h6fofy~J!7lNYKH#oNCBz7@TPlYtbtiolekcz|q9JgUcg+;iqB#j; zypr?<^rGqPG2p>G;~Xh`Q_jCIJyMxttMbwysZPj^JKj}r(`J}fm(|X<tg`dzpR_wgF8oFAzn17SPZDNufgdo-uSNN5Z zjKBUu%JDFdf;%sM48atxAZG#aNH4qVO>_dwJ4Ah2RX&o=0fwmI6FI5;>Tg)HF;FBX_8K&dj0VVnR;dA+4P!i5)^;G@b;4N{=3dlAx-zvc@`Tmj*(JzU#(@j&18Ivr-$nE7NEn$H zy*OyYZwre$(2_oXLHAg5YieCMWI=n zo^Ph=Xt$_DXe*E4EBM0gAF)U^Oy%RxSogi2Q7A+X8F{X!e~NB*{=fJPI(fU>&)(br zp8j09x+_r*vB2~t8Xok~Vjt+~Xj#puYa4lVlJY_?x|o+6G$iHQW_+(Zz3J-q#b~s@VxZxYFO96eGM26?cfF z5EL4n#EWkBzL;A^HqD-n|Af9WilJ2TO+A-54rp%*^QA(|8KIj_d=ciCSBihvCijS5 zb~Smp17}^)MwPT$bC4l%BPCzs{E^qlc5CQ0CQ3UIJUU{-J04C8pb7B`#9)>DO!~+V zhIfAlLZ~3HnXpVbD+3Tbz1@cBP<&u!vAS@@jgAFIN?P`wH(p zEK#pn$xtXMkjpaPH&&g|^6|7&2&w^{=d=!O@6CioH~SDwq&;uE;d|p6U1Yv=?tLyD ziC(YGhKmESM~N;+Vse{3gwdLHlx#OmT`%f;e0-j~z7aqTFD@d51?uIaZFl z9gI&gm4jy5f}7;2l)0#M*mQ(+z?V&s`D5&pzvw4>ur>?23Bqib{SM&9xnp;&mnyg` ztdYmt;H0i4yF6*YL5vKhCjr`T=rjU10DV&o9O`6h@2^IE6F*q}wJJI6{S)=lgYJyy zz~F(a4Lr{`=cGj36}+*lngC|IBuRrvVSLVI`sK8dA~?1hOY}f7Qd!-V~HbJz2+@3`% zw_(P~X)}Z7qH4B=7K9K=+}}Vy$n2zEMnP{qKT7Bs$B3(UJNsVdyNwH0>B|gdhHzEp zaM-gXgA=a5?L9OW4Z}d}JpT@wJ;5o?9!z;?i2~nys{xaSAV&0((MGR=w!W{?WW!pU;26V^b%cIb+!uMQAV6t2%jEJFC$%us~n8z zDk^VDG@*>AmcO=0Wb;w&jyLIS?LOXX-UJUfKfW`MmM?1m?kKVowy!WPuWU7tn_Pc} zShHCPUA1XBdzzD-qT-LC0CP_wsaV4;&^{&S6>whDRKljmdW4zracQPy`V4h-;+ zG4s_PX8AgHssb+E~5TP&FtPcTiaqB@TnLl0A&Fb{~l>AtVRc~l<=i%tFORDC^f;2pE$ z*3Rfuoi4mKGAN-1#+-m81=8nJP)jM^Qm z*FqY87Q5>vvxO((jQM?heV>RdlRqJ^=MoE>rCezW7DScJEN(3TIpu&R=yZTT*%G2k~$)1b*f#%aBj5(L?nE z*uF_BPx5Vb(K%VI4GkuhrmO}TcXND{wUr-g@4KdUg%>Tjd$|vERF|05!vK+B!DO~^ zi7IU+uKJE@Pj&}h4Se5yW=})Lpht~PxlwX>V^r^D@x{4}^QuZTXyGTq{Mqe;KC6sG zVo(v>J16)8GNV$@vpLu5_ck&>CkK#70Y z`fe;X5p8@ZIyc-rizf!foz2)E*<{5X)ZzrV_^R8Uyr6b2o2X zGP=&*ndis_QBn)P`N5OXQ~VL34-ew$Ubnvn@9k4)Dv2li!P;8N>65=|8i@*ZnpXOz z+o(mioSJhD`06_FlAfxS_+XPC_C~i~oAHidISF%`AvR)~Ae`?6rl#=7kB3pizfu?o zyUq>L-nQW#HO!^J7#;hKVQIA+$~?^XNy966hnrraFAgdxY?;P|g2f4UZ`HiRGXz?m z@2$#uE|~er*@Wh)+XAy_$Y*vk;6J2iHJKh{a z%4bKQ4fr5yPvc|Mk6AJ+_uQ!ri?D^FUcI<#9>654HIR{=b*F zfLA?NK!Y5fC)A#_l;8=C^WXI>dfYbe};VMpimtEfSgt^liKt^>Bo6CzOjRT>;09y6&4 z{owj`meEIL4)H(8>WSkpBGuKctgx{r>=ynCdI0x-(2PSRnr@2)fU;P{AgxX+_`5GJ>MQ5 zo-H@BxJm+3f?P#MS5@QWjfdxdHoo%SxSG$%BrOT4J$2q+UCLMq|H}Eg*xx3lz^!skrjt)c6yQenqq#&^Tbfx3kHdxj zpc4I30Q_A>9FAVYo3C=%Wee>AxC@nxlA45=3)l4es}p(mf&3!2%U8EN?cRtB)paP+vKqA!lXkWRePv>q>=sdy=c0s}pLNFBEkpaf7s0=$d2pmpAJL zP4>R2kspM3kEBL7iPL&%G8j)UU%wT>WtgwM0(I?(p4kbYVyw+6rwV8aaIaddZ}6sYuwr^UNPvde@jf|mIeZ3q|ibN9j3CX*%ZrhSYevZh+=BnQrbqGVo-5>xyk;) z^Sy(%!_&BjY+^j7dT)wUdU^X359i8!`MA=z-JrZM*7*<1%&yukRu1vjACdY zgD@kjFeMj{d zP>c-wYjA+_v&Ye#_rG$6GmFNQ$2et*kkjP zAT>M%=<_->Mlm!*82Nz0Hu8b0ovKHC5oA?{?Hc@IH~DQpgUWol?wm+IY(auKabhC=^j_9&&jzxpNd%-s-boe0*Yfn z^gyN~cjE6v^wE7@{>Jq6_|~UPS9HSom(8<+?kw5n&jy?I!{Ds;sH<$-FIW$+sZH~o zI9bi^!mm0Hu|8<1LKDlG*9t2XjOv)X307%|%wf3 z)F;%^+1ad7Vx}pi=a58p^t)=mI2p_2%IwN^SK`DOE0M!PUrQF(Qnw6C2bYxe+Sl3x z&hks&1kEOhQaD>)?k~^m@TR>NAQ;e>m`8B!ope-X|%mF14@epOjm zTiC9x)|7ocx;^~_w)hDO>J;EjcT6&!4l(j~r%NG0JmLqcrgl814$Ia9;BL*^kO`BI zNE-OVrV`{rQxKX`PSX)!FAf>DZUq}k&1AE8nqI2tT&SIPsa&^t$@Ep|f&GWRk>=;z zy}I8kxfag~*UM93fv?763fXnqTS@{{qR@3(7O*^Nlxc1FPb3NPU(yXhO5aC8Hzrl30fu$sHyayS~z= z>3LZIvDl4TxJ&ao#G?qKi>^F@-Z3hO(PZb2FSj1Z7ELnV`ki+_C+a#ZD^*c4POkd< zoUXg=lS8h{xqtPW#d%U{F2HbJpccu**c5<($Qs97yCspZW2w{J-A#T=j8a6nP zrHFRdXwl8IOLY@+9oiYG>{YP5J}#wKEz2#ged=Mq)JBuOlNq0}L19LuV*l(BjsC!! zowec9k>%!%qeoH>e|-~rnx;BWdJB3>Fc#i|^8f$jI9+I-A1mJD1KG}xH$?C5y9}@W z?0?1Nh(ylA8>Mr@^8p}`dSWUT36H(Iryl13m-~=8Nl(HEi?{$au}s$&n=-3cUL3x78p5aQkb&MvsL9TUK|sddXV;t}R-@i*X>RA28C#1yd- zm9XUO5#OXk)tp^DgP79Bvy748i*wpg4SG!zde_UQLsUw0oZG|y=6;1Z87$ZRa?4PQ zd7;E87_geqD-dVd7S_)yA1?%N-)KXUNH zfZe;J`lo*iMCZ^=)j4!(h@b54FaD>37nyphPvl~)!;LqzE9%X{C*T@-0AdZ}36e8z>L& zPT;^BZp5)*deXTQF}3HgQJ;Yfm21aPpoC|svTVKqT^@*n)k*TG-{SLMN!%1BDeYom zZlSsXl?1kga&6PWl{akJO0Md8jc4Yo8tuJvrLMGce3!nbVY8@OBrvQt?_LHXn8dPaP!;vYuhZr z6`M7Hz~)*2(f8QT82J+_)dq!qmQHV(f#u{b(dgEJ4uF+jn8JDZ5$z^t`c>vhWtdv| zi+)ByGF_|6L7NQ1a0MLjf%DpIucHMU`~_y}v}S|)5AOc%e~BnlH$C_NyKL%Pv)}na z$juyuwxcdGc4cDvqr)iKGic>3o)Gm>8aM0QP9wOz72}N|NnXR&yQrK!qFwG04ltwQ zbv~k1B`zT@_ECRKaC+wa!=bNO%!#3U{A^^^W-%-% z5i<)Zzd*=>+h>R{u3MgZHDpDAiMETwu53{ph2ayl&i2Lm9m9~Gb5pi-_ zGPPP8?#oIbQ%7nx>K!}CkUsf&0g;xB_GUpzP@V|x*iKIPr}N922lmKBgmZIL<&tQC zG#hn}GlN|nGy!g(N}c%bpy>_MJbagqZ{%v1nRbUiWb?E+7+uL*xPt`1qB^B_4QCr` zyceW~eSe}~(r1glafyF4SyZhDs%E)WMXDEqsK!Esmo7dW-XFjQ9Hc0?vJ5fc>ZSea z;7eaSA4Vwhg+Bd+|M>Zn*Q-c4qJ9Qk4Z1{UxpO0Qrz>Twp%q%w@`}2{UoX0_FCWH^ zUP#C0;K;bfywQS1;kz@<9ANDx8%+dIBKJsXIh!6+6XIuJP#e z%X=TsRymG}2kYfPxiW+tKBKd|A6txY?Odm;t|s`jY*vlf=gg>@T7FD<=>8gr|18mE z1}YFnlZF?t1RpV^=^{!?F(|Q@I*UKNv79?3DVrC^O=u||7QS% zC>J47fMNLG-PPqJY~*Y|kucM&9o~z`PmBzL1g0O@i1gbG0N8xoPj~jFsEgE;5_K|S};QQO? zV-t-pp}YIsl$)j2UNs5;#=DM5K>_(`vdS^Zk3^Lg91`lycfSV^E|Rs;1+o^9vo;1zu?RB@aQ>C@EQId&s?chzo_5EBh z{Sy&>i|Zk2bP&pip>gzvrXnJhE;j3PL zHO;;v`kkg)^xAIH=dY;%l*P3AoHd$w@s6x^|!F1$l52H@a$PhUP0j4oJOGvp(Zd|7T(O zdGc&o6@h|P)|M-*bN8Vr|9oAGr()lxty44E1(A3tr_I?kUcFl=wrB-7)zBfJ_a z3Z#UMg=#QqQbg&oR7*IVp}LyYSKr1eY7oT?J<3#iAupNcC3Wm`;K%3eJ%-Sz*6Kb) zV22@v7;W^^Le8e*)4C<&srIrkbE=OK)LX_gJvcWx`1Eogvy8?~_gTX7BTT0|E6DEH z`&D3uPE=b90<{QvH};y^(UeHplI7EbZXKP%R z)8E^4kABVLhhCkvg^r$?f}OpxH|Lxe5q`m(34aZAxy#wqeyV>ZD9vqzLcd_&q6|zl z*a1y&9WW4Y1myHEwuQ0J)@zmYwsDT|=5f(6Ol<~$>CV+^)M~l#`ZQmnfhG)L zkT5j!E|~t#x!9501{q90tyao~P(uJm?xI#ZG?Z0cQvFzUO#aN0AasB#g;k7oOLx4o z(Dp_qDso=fi?vUJ|L}L;{m=goUw<7H)%OMr-)Dju8l*ccLP`N?7!!k1Qd$~C8flnO z5Q9>rBrTBcu2H0s|MHpXYqm;tyCXW)8FW*>Uab`WQ}ID zmA&}j`Qn)Y)!c~GCx1$VM_67opMHDew94eG#o`uHMqjouFZapru#ZvR0vj4Lt4l3K z0-ecIu~l|{-MgPIiflbtDEf^aMU~Syb3$QK9QIN@y9YDY7KcA~U#*$HcL1or-+2B8 z#TtA*SKTm)w{FHn1JoI0HanOx3EMW#x@Wd2(6^{T)Au?^sG^U`P$%%+SmE(vV z&&l$z$;;u-pA3x@mRTcM49sCf^@)U9nRao{C42$MS%UvBpYy}^L+Ooywh-m5gWp8$ zlMy5RdKKyui9&uSwMlvfVec&wD0{UU`r>sK7d#wx3> zWX0x~mika|bd&rS9dqc$mV4RwNjb-iezKW&fVG$6%368TBwv}jW5Vd=x>4cj2o>cP zpLbhlsT?MTYK#2QFemspvU+c0d0Mo?D6x#~@99b4`$qMwcjtzd*PTB?!Re$_tLy%8 zDZWC!41Tv-{Zsr+|MLHou|6B-62dVNUEPx}Dl3}fl{>;$ZYY0xcn7ICx>nt>^8%~Er4o&{9!;U@93#q#B(0B677$O=K)-slYD{$mBx#%Iaio-UhDZcg5u%V#$Qy^ z!0N|e)X?}M-d<;*DEW3HC*s$bN5!J+02%dN&X~(0?SWz~=`N;|)s5Z+5N@%0(A2AE zccGXWf10I$T&mR0p?eVipsy&?R<~0;)<KG~`KA z_095XP5mW?9~Rf!q5q44Roa`@U{=&!)wv?uAfy3%@^2vfd^#b<$*`6}m^M$g`mghY zi4#5huRV9ZM*kNzZRgdupqr(i1KB3JfLt<;Ho$r04Kz?c^um@i`3ZS+tB#m<1-0Gh zu-;P?G`&HYX%ft^S z)tBcTK!!3&yQ)XP5AvIFN zlSKTh=l&wnU-7_e+okz&(tv!PM?l4wquU2j+s|99qFT;JfvMY=q5I?R9Qq=88;`by zIw$TLx4_xs4sM&=~6|Q2$?NR^>=Rrt>@jz5MMhR&Q*cx5GUJNWFd&Tv$ZeDtY*-OigVU(iTz zPRiNN_;tifrpvXngK9f*jbJl9;C*j=x}>HpyyYnBKwt;0DCT3lSx(NzeNo=R9Z8^= z1x|Lt?moA*G}yR9kiOvk7tM;T!-?Yt_Qf?q8)~f%Bj^^)q>|DW+p}Y2Q*CaIt-kPr zJfWp8vyW|i3!_U)D&V*7be6qjhn)=$Fy8(}>*e%aL?aI}u+KIT5yEb9^{!FjquQHo zxX|z4SkG+AA?%2H@~MHR<`jxJ-6EOrtCp*4(`dQQyasib_6X1Au|6A)y!(l zVDazi?`Z4+C57sdIHG6gM&M#8Rip!xvPdj?b%h}JeG&P4`O)v3c$)W1nNuwzG^=Xn zepXh{nJ>OF8d7AaA|LW_Hd+Zv_65bN!10G&R0Dvm(uhdy1HYp3Kz$i|V7{ZuK;a07 zS2>0+PxY$=TzaH*`?0QVLWIJe=Q?o;L5d7D?tz(#2i6>Jx$^x!10d`u?Ex48M&||5 z$dN$dld|CmVJDsOZ`Cj{u5xF>4_N9`51#Lulx{p%LBuCB_~ZvJh89D7){Q@J<=C6`jX(LM6a63 zt?;~^UyRWL!kS36PbARvlV2n6%V7DzqH`ERT}y4P9iIaisG#-%< zncSEGT*;ls(fz`(nA;@Svw46x)*N|emzWUo>q_)GSAMml z_t?6;Gbz7cyL&2CWoqk0InP?=Dgd2P4d;?`J*)NjJ^;V|_$C17T{~sAo2f@`M+G1l z8b}~ZdJ^od>jB&TNyf=<6j{B+;C4dx08Flr1RwU7vq9#y!ZwGbl#N{*blxo4&wJe( zlX?S~J_u`CD<_q$#SZn_YNjZ0N4Aj64W1)A55O~;v?J1a&}m9>+vda1old;o(VgSH zqm2Agt>K-t&LOcO4IKFpYdUex~)DqH%Y@B<-vbY{GK*7m=8 z-haisy9uBq#_TJ5-P4^^cm>XA+)NCG(sB|@U5D?a-gSr_6f5AaB3}bsFlAYe`eV6R ztrr0539)78W4`voba+jHDmlN3u_NgcBb-?{vrzAH=cq}Nx6@CNk(`t5vV;Snj!2XY z(xlp7Q-a+z?N07H66untS-az=za1EUe9Fn8K%6~r1LLd`UiZgk4X!nMp(w+bo3G$z zaR@$dvwB%X`Gx*&HIV}OBrNG~6{B57z1K@v4R;j}w-@i6ui5u{42q!Asc_v{6vDIi zVqa9L6BjiQWMje>Mbg-&w?b{^)`SeqqhW3qiH1jA%{P=_-e#|Qcrv6qf)2T@K3$jb ze1d9|V0#1mJz@VY=LuBkb%!VY7Bt=Th$r?ZjOMBO$1Q2BS5Vg(p}VQRT6HGc%rI~D zS6w{6a-2%tKkcpX_g+gwE0$R*l%6pfrg3y${Mq_e`P6Qdg!*LwWiy)M64>aUI`K9@ zb)OsNqqF&K?jG!H&268kO_WxtHh}1;mfM(V>-v{rQ@?xXzj&) zl=J-PSGU*>*q%JVH7l#PO@iu=%@sE>qcjjLFtjK^(!XN_5awuwLI0Dcy{B%82O}dG zQ_I}IWD0Yro*&f-do}h>nSrbBrIl=2r8N_DaYp=Ubtx)MG-Xn*gn{1coSncb_wln{ z)$QiFUSHha&>PP7BpAQT$6FukoL}qP#P)n`6T{#hjANfY$4uNsoZmJL{Aap7)c z10wh*@gtYYtTCPW$q$w4v0`PFqYJK!PJZ>@uMS=vgvZ%m?6pWE`(JH0I5BvFs%};` zx~T19yyc}P=s=Og+m9j(wf$eXdb?XoF!#`5%v zVqV6pvgRCV4~IR=Q>^j;al-!mVtQT*s?8i#bds)8u=2B#aS?kjLG=kw2|InoY1bQ5 zl=<_AX&VV02r?t8QWgiY>^x9LpqBRR2nmGiMWmxOPQuafdd(Lp)aVyj-WB`ePr3)! z$3b{?)x9o;R-~iZT?yGh?k+p6Y>uaNS{*U|{R3#M<%np9pZz8HWxron`>~g-q7Ptm zYq$H$^Y~UxGdAn1V(|?j_v}L$a6;m!F;gbg=B2qy-q=$#D*PwBwJv^>yNxif&?c|L z<#ch`S55`_P#4|$8?+|e9AVyL+akMts>o%+x;7xPdKfTF9E~jDe|)VUvu$)WLV@Vh zJ*4k|K$Co46j0LeN!-L+i}Hx_ckWs)j30d0qathq850hl?Q zF|OF`#JRiv1|4h3i6Y&`>Q~}QVwwARkM~^#Q(kKp0OjSlIXYKyDpD;5K$TiQX&sIs z6MT6AI4Pg*QOx%?-@urA#*Udh^@RPmFx+ zw41-`--Twx{51Y)?8UVzyBM(3rH=3ZeryyTibfJC9#HKg$_mTOrvXreaS{mJ$=WPJ)qu1`wl5|z}u4> z$E@C%W!4h2-?h$n#$q0d)*HTnMJOxp@6MfxEIIx7Mn5ISIfOv_o!d?!+T*@4z!mu@ z20*E`CdSB#GTve!fySAuFL>!gaV;cJ%C5><;9s;OO|A59zd7HD#&Oc20)Fn|=cZ5| z1ohBGtENcqK!5My@<9MPlG&LU4U|lk(|60AF(w?Z9@G=*PCC(pub+@$AM+1f%?*Zr zkdfe}^6Wt*AJo$D?#($?1gVyiXg_|^H-e~Ci#|JY4uIKcx2JtjCwpHU4b%mwl+jMe z`z=azdH4=2o0K7z;!_4b;X3SMxfLX%xV0787?BoayRcsY?A9iVS zdTI0cAB9nN*^Ni`+tGZf(HKf-qk1Fj}m)puo{Zb?}2Z0%$6 z$e9nLqEE=QimGNTXqx0cr;8;sN~v>}J5RZ{-GKUNbz3t}UT9BLa#R5bsy^0CaUJw- z56PVONzu2kOaR_U!zN=s3Z;(xcGRv?C;KHHYbI7lJ;+n6Waqm|3``W_ynp$N17stT zT;SDc!FTXvdt|8$yLfuwM(&&AeHebKs>L&k2=|~`VEHWyRr-4DEUat=b91&|Z@;c? zo{rbRCBd;HP#Me-fwhcpx)fmzf6+;u!mj+7cvQLPv9X9+N-pqiTw%v_w|NEWp?kd&a-k z;LfN5 z@|M>w>}M`!VY7DuFZ`VzfT7teu1+HGkC|EWc%ARb%8>%WW(A<#y?7KCpoX^yZlHX_ zY*uID?Er{Kw7Y+pm(u1F76WL@#hgtDl@Q_m9{?uRVV~G5i(c(m2jHbUcw7H*W8BT#Si~DE+25b9p1qY=d?Z= zc#1{|Z{~BIg$-^?nN=^5Yi+;I&~0=v|7%Z-RR41ZcQFs$6z-2u(`9>fFRhs`Mo#udAclS!Y}|a zs#l)lYy@j7PO(?%O^zX&HC(%fsTCNpA|Z%P0tj;dQ~Rek$v`RRwC4%5@Mr%nm6bKS z1`BI_yntUnEwR#w~TKqroGTl@d$QqBHiZgh4elF zDtb2l-J9lC2g^Kcr~$=tN{5$*9!~Y})Ytcf^ctXW-VF(5ehT9Rw=41z={06)N++dB z43P$mBLer>Zf&StR^NhR+7fDfGMKb1&1yhw|k}&hiZLsKvRS@>E~Pyh=tTrPUKw< zZ(v5Rj9&9|IdBhJ42qyqB4e7}mi=N*ece`3^7V>pr_6eV^y|k=1A=g(X0^)NF>*EI zP;By>?55PbeNXT4ag!Wj6zhiQfIWRjg5A|vOZq!Os4BXYYh~P04tFn@ zx9Dv^(TlSJCq8DjIsh_}J-}u+{ttesdFXA?8Jt{?+m~7AL)^~>H`Y!QMYNqX3G~;R z1Usx#lJQ6z*ZM4$Ox(?3AWK=-KXV`XQji2{b&Hu5>4bgC`3%7JP6MC9BLtjp5MJY% z@h)|f0Zm6(f5vU~K^d5YZJcD=NeEG>;(=Syj13LJaUP&h(@V;hf57~WmITT~0=3^& zGZUmg08pbJ31R4i?c?Me7>CTiTI2)fSWK0^gJ44Ue(wF$kLn=YM<<1PIYmgoNMe}?%k0fL5GuZQ^~U+ zxoG4$+@;9p&iAUPC8+rhqznNR)uCSOYnPsto|!rz&nJQ0{X77BMRNWr-xmLr1LYkh z`buGU5E(2qOFu71UrCOjLFxuU%dayJgwZE50Z0)^Q=95G zw&vPMfBu6Xn!9*b%Zn@simxPik?9A&ir&N0xU|EC(6R&az;SuVXVzSx!>6>u!EI>f zQsA0KovHgayLo|u2+QJ>Uatem%WmzY!zglLs+*LXkj_${yKXTSxoq$`NEiR6^;F*H z*I0g*{dA9YSaJz9U2^nt87J>?R`c!yeAq;CcKI5NZNvV@B$$G(uNi-fg+fgU7zcl}SI7KMuV1@F?acb7};puBE0ITkd%A z?V2U_cY6y$8>~}|jg$G`^)W;J{bWXboo=hbpOW!GOZwv@cvo?;y?t*FNc@$K^6;J8 z?po&0AND9&e56k>?W#%jy@FZJ9M&z#8a{Y-j_fXFjC*1S{Q!UNHDlQGKd5?|7iDHd z-|?MlqBo}45a5VQl2dSddft4uaW(7@C)WO#l=X+FiKas(g^#TVpaLNa?Vqx>^=Stg zE9sRf`<)wRzGd5v%VXgKM1@k~V67!SXn~@SF%#!jJL}iE2luc5Cd?Ijd>&~HsOKga z%fR%B#x-tRw9Z*edP6yv_-su&Dib}po%Aj51MpSH>9Jo`%mmI3nM5a9zORA+IT#|D z>DEzxBhA?zm9$t40PI25-3iXS3Y)Dhjqht~kl< z%Mz2XTB_sApsG1Cx0nSAvOdIzx{>84TXJJQkFjxG zEB%5fR;1WFao<>&>1)qAqVo?%lE5r^NU-0{CD=k!=6c@#X?8}==yqf!>QmT_b%7$p z2=XK@c{%q>c!4HhXR=WE1|9!>x#VS?t%d zxg<+}z&|WeitJ*~?(AV7E$HNkR+f4!YJS-&bYxQ)e=Q(Ab(WoQ?a#r0cdNp-szwU8 z>(m)7&#KZ-mm!1u%0`a%=gluVEl9pG#U!bX{g4l@E6yA@@K3i9h-6LRd?$l!?rrl= zl-XD5RBhN=*JYNeJw92Vru;F+18oEv`MC$sNZ9`>W{paQm(d|-CIZF^SeY1JGcduG z#{@lh-Jka}UX>N96b9fPoeRl*N{EGE;N+*!O;!iPeEaX`Q=eX44jQJw5SsCCaa<{~ zjz>S^o+Sf3`78aEMvIQHVH^qeP-stkoeRBil{SQnR*qU^(`>G;i`Q%FuzByTq*g_$ z>y>GE>Y?sl*1CA&XH(16MobtE)AfN=2@U(djfg8k0GAn$5WOHkuA`XaWw^#9s3`&< zuSNv%P$>1dRI)OPoi|p4s!M7BSeBfgTlNf;-AjVew$!%ll~J_zcoYCAYjqj{RI59a z#HN#AO^FNd1p$b?O7f}g66zAGjaZv8x#GId)K)VUVW_nz)>fR4ousRliznI3>+c^5 zzFt9qTte_700Yt1z;JrqQR6n}h!dpoCW%Es6g9d*g8da6Eq+0cmfS|j3V{4y?=x)) zeVRSBr|hRlf~AY$6;Uzh@gl9AohB9A`>BhWifR^ks~LFw^iU*#MTII-Nz~VHTPo}f z|H<{J7n&T(_OhvwUjee8yx}TbfoyU9ZrklcGO#3wgdj}R9nY%9JkhIvK5WOQFf+MJ zO-9%*m5jLR4IAk4zhiU(ThkO_Ji>^yn)zbd(&ku~!lNf#6P=CyV($_);R6@-X_prD z!PhMC9r*S^+C`g`+KdEGhHtZ;@Uv#O<8P%w@g4T^@vOKAMir^RMDpuI-|fRbWenz{ zcWyaHc6(~Ln;1iOEr&8>+WX5=&`{Am*Wzr^+$ZlPqt|kGW#vcR_4je4vP#MAztVuAxyRo_Uc#8B|9Re@cwPAh+ysu@z(+F`=`* zgD8}-Nt@b*)7lu&kM1!O{MM-lkWbJY?6WLOMdZU{ae7&QInj&dxN2sTY$?NiykHrh6uu=2~k+ z&;?*0rI_;Ss&x5WB^OZtccog~Xn6ga8M8OGUN!A`@{{vsEdzkP`vKr^u3VhYO|OEh z@u$&ilMhxpFaF4(!X4=DWUc7KE}bF)Sr`xu=(AApG361nY;AFTnT(Uu_&7dKHBf1P z@g8PO8(@?eu`TU%;7xUZFkjd+9Ufs>wf=N^`4?WP(>HlX5k^Bnf?54SIv2E@l>8Mq zr&W+uNshvxh%cjm{_ zPj<8|7VeP1KmEJ#zr(a;G{(pc>WTtvk7(!kC$7~tv6eU=je2}x17c^r#MFz1qAlZTrKv$ieD{hqKTP{A|B4T^5AyTb-oFo98j9{O|gP2OL~2#vth+)#2l@_`f{q(7L! zb6x9=skn@3ALsb^;ktS7#0TwG6_|w~!?mA|Fi1p}Xgk#2*V4KRQ3Yqz8L`1T4}$xY zJkd8zE`I+u#64>n6z0-K?m-B%*o>0lz{*P%_r1Yw#XG(UTtqvtFCJPg`tt=Yqe~7& z@z<9#M$R`vlA&$-nh|8q3{?`j{yr@+ujrzVu=B5H8hV=$&5bsOM&x_XH+AQ;dh~H- zUIBmhg3(^JhkNU@tvLI}-q3V*gxjat>z%JqGv>52et)*LE7j&Nfx{W+7PLa-F*D!V z?!|w`PkxRdt(vCrHeC%4rlexV1|P7j810@U-r{E|cu9u6VSo%rARdxmV}leMv=yhz zaXg!;1ILzgzq>L%YRethp`^dTx;SgPtBxMUJ{B7=&V`9N<62L4@V${Qq|D{hEca8` zh-?*r9Mt!%F}Aii0Dj*5>Oq8!#B+A2$QT&iUK}u|>3251N^wUE9{Ebms?>7rA&*7059rP9PW;T`ZF8%4;@%L@aASg-x@C6Syb4@bnZar>pxD^K9W$;#8%!Cw!WO#afZj$q#Q1H^g_*GJZYhvZa z%gG4svjeJz=Q>{#4ue`B&DBF*;htXGP0~aN1Ai9@t+F|Z6?Re&;A?|vAbhxgODW#g zz6I`W4vRT1BwWItg`8m2*67){uQu>SzpbvvKw1zV{ezrnK z#T;+2T!7GkG8up0%n15a54(4j+&se`pjdJmoXD?(vDHhyu%)Ks>pl%!oxFCuO8E0U zgpLP(Hz<}|{1ghwO$NZSe+XxPTH~M4!`=YM_!ch!RCu<7$+|QMx)O2w`W2ly!+ zoHRt1%Y( zjqaua;>^Qp2PAg?F3owGK71~YRv?fSeqJqkzfFh2t9Z6!F4D$aZ|iF7gIP!xjLC2E zElt)dJS@Q6x%bTV5~I0Q^mSOKx%sXog(Zmo`sl`@aPpMthN6=HW~XuqCh-!-$&pA6R&Om z3k~vVtF2tnbizy;{WbcFyPpJbWAFVEw0__7di;U(vg7brF(;lp6#tUfvo1_2VPlaF z(bu=Nm7>?{y1P%CA1A{L)wJEKlz)SbP3zQDCeqGvapQmpEu&XMJjHZSLUoHnuc=dhz4M1N>SFy=Lab|~ZQkFcr_*U7ivO_HQ zi~mwYK_OOjEj;g@)R{88i$pr#+z;Jkw>K1ttZs3%`Lf!Ac3>7{e5unWlu|0s= zv}ozq2sg1BZo@3 zGHpwlv$cqRvtiRMBSi^CtIe%H-4Nw*+_CcfSph^8^y|u>q^a=k@Nh$ zh~$w1?vb~m%P5(0kH33+0|BQaZO?VyNsflhJZo3{#&C>p1x{;T@1Y!@_fYk8M47173PLV)7;7TU0($pM{?N@DiB4Uh9Ts>NJ{f-h-f)1mXEk4}(aEpV_S)K3! zFAW7QQ82E@?s}0xX)Yw#cPRjG7}&BR%_$lwt%D1%!g{yT7S#t3wdy3;@(`dYu(qbs zvrJ@JAAc8QQ=06<|Ew=sPs@d_dm(@PWW*sSWa=4?ckcu}8HBHHPQ~MU$=IUhSze6e zzxi9XQX2_u2NTc6ipCmUaO2MJH-s&6L9ZYY+;C9iL{=v5)B6bGI2}T*Rf?s6zL?IcNt(2kolLHNV;n&;xVu1@ZS9Y&j3GV zQ+D2^2N7H1RqR!K{F5`ghn56mtMcw{?cYmclj}Qm30IVS;-~{DsfFipf20;p)DQuO zCLh9OxSiyAB!Y$1>?sc2-hbwro^V=cctwImp5?ppS-rM|w0j1P{}mnYagIC*i>Lw= zQvc)u6UVx7;5LHQ#{&yC$x!x2tb84y!dI*3m$N&p66hPRaQctQX0 z6DTP0a3!dMaR#KUAm7{YvLio!U}#2fDqZ&|pigb`%D=xJAO>S0!G`An;-u9ju>E3t z#-J%yl#5c$?8})-D_X5H_fM*gxxH52U7qQGocc&>ETORSB%o54f9v_LJjA`8`N@?j zsAOj$Fy|#?bA6GzIG&bV@$cekt2H7gaf+u!>76gMR!f5a_U5x#c;&4{e*TFln?m4H z){nnA4MuOuu5={%2gn~~TzMeE@}@z)Vk;sg)KIg>aLdW$N*U!7=3!xEGg&KuxblGn zB2AB~Yvw7`{=Ji9Q%VJ9&_RB20r|?WlFu*w{7P2z-i|B9iW&(O*luU`@C!&kwtMlF_XCx`u zE*mUwm%qW9^7chHK7{JX@`CWc11EOV3hzJ25HYXq^82`V6|0Q{U|r{G@A}1h8J>{u zB{Kyx0|lF*jqG8P94`Kcf9L}FiivPP4sn5$TqX{IzfU)$ahyU8s-pcR&%q`tT zs-UbFKtSEWsijU^ktvv+lrV-Bi#jexb_Fp&jZ0k_%jbuloLoBD*ok`k0igUgn+Y5r zPKuX5a4sU);4>NkvTvFsbAxLc8vvsICLO1egA|{vnsXdM!Z46x@9ig=&j5S+Wn*%yXrCnbx(wq-jZr_|F5^$lVM+xAN44|K)6o=1rcc{xNA8jjGfBI zo&9IyBX;0KvT-WtSn|krzZ2QPnk=qpM$U)iC*Sktl`&(j74tKJuCe`~hbljER(AW` zCf_>-U4wkt@{N%12(nau0w>%aN!|Oge_!1F)6lin7^p9&?L{SN;OqWGsqCs}A2K zF)G)ApcrxfsgZ8Ae!xc8pt{8C@!I9)ilR=$;!EQ^{2yU7=#oylFHu8_6p679y%2m6j3S9c{L*gq4hEf^% zjP#qVK>FIhPGhfoCMUkkW}E{=IOw)yy~75E0W^F`q5zEq8#U_r+J-Tu7ZrfuoR<0J z^d8-xVi@@7NC+=I0wjQa;HeR9COx3BN0P%AN**8I{&z3Q7{H_Gx;&SwZHXP5Q{ zY1RQqGRaoH(;c;~F|OcEv`Wo(dN5H~i+79~J71-aCBy(MgqfiPTc>0)_m5O2P&I)} zoJQckmw)BI)k&yVmITwH2}F<%6o4mT$ll9{U2^IMEow6Q*IGfY$0}4J`F3$&N@e>0 z&>~aFrJN;5DwF>OC-iu;{tR5)J)GQ#=tSGJw7*L@hnDCdfzC{-Hb#V^)j~;z56qfQ zA$-RezflP>T)PxMeP~&6e#_+w6xJX zuX2iQni=@YlfwHRK)U5P084>HO8L_dum5<7^!dc_e;$|rHHE{G?u;w{YZoeN|JQWs zzbR5AZ;s7__QDU=_-;?t32uM1yX8MiVE64!!=sbjT$M$mi4_%fqUywoYP(DrYuG@Y zeXV`99U$N7C9O*G@rbMbMB+mWNjBxb6{T!G+Y{I^;t=gV{;xYbF<8@ZJem~v+_aRH zWfz89v!2g7Jx?eRr!hjcVI0kSFGG72d5B^26oIODD&_;F@<+S=@ z(kEG8G0)2s_IXmZDIy$w^Fc%*kM)@MY*B%xWlEzmros#y0FXHh zSK2S8<(yRS{okLwwGfrVwY`Cir=ptnWGez>@ctN{s z^C^-k)8BQY7R(TT{Z7AzXF$a3$x)#d#fSfHadXP#@laH=Kod_@emepR(Gqzz^dQwcMmzO2?HqnpC0S<_T9qZL)P)La7^ivpF3` zSh)m%v%KKAai4P$=TF?Jnb_g!n*Z2)EnRgs!AamE2`1|%fC8r{9PE|mt2~POC!vrf0_X%4t^ba$FP0E-3Du`k zE)`+l-c!pIF>QMB=vVG!17W* z6hOMWdH}fnRDh4r^KYFM6(`ijWJ4)YW8u@VNr|zJ{KXf21+zL&=>47!8wx3;v*v}r zdQF1W$p%CuEIUHwBxj)NyWX+uVq_feNw9Ai#WkvBPnm-)__#axO}btcwG-%DqK;gK zD}&v?m_D+Vk+HHWC-hJ21^|Y5LCto_oP4!)(5&(J63aK~YykPA^F<1c3e}g0yS%Z= zv|Y?S0Q|*onhehSs;w#X>qj>@ZvvNIR85xy2Kg#M{*}EQbegtieE;9&%tw2G#L4(h z?AGq&2hDrf*{>H|oPOa3Cwk)=M6g5#B6lMtX3xFMYK<+`%Q0B%$!YpTR`!$%X!QY` z-_aAXE{URo+8cAT+?_89!eT#g!cu-44ZqUI`JORVY*B_Y9qFwY6~VORMCz{ofD;bF zoD?-t^q#^;PYV!)wfRdE37(ja{Absq5m5C{Bc^WDzxCw*x|&F*d7HoOKNDWzFE;|H ze2phb!}8Ju@5YVLS@z`=q{HZFd4-E4fizm0jx!QW5k>pr!VagfbaAJHN5r$~9?1d4tQFNftgW1%UGw z$2+OB_T=VzmHgxQc|%)Z;Ow% zRx(jJHJ5xYO8`W<^r{6KD6+2Js}y}3hstj=@RZn|jjxM=^qiU@<-3ksci$oZR(&yw zbtikD?%_tS0(W38sr{tFY~bcPel2MPFGgmReCjbw+uq+cdIj@Ua~aoIga5b$^$uRc ztG4t^aY!3F{TsqJ&;I`#Ueqm|nFJ8BJA~2aZgdbGI|=3>ep_p#Fy>d{a8j8Ijt{z0zhDwLkqRoVEUf!bB8=sgkGx4O2ueyNP6_ z4N^xW%RsMyx;(*6X4a85wLK4@6XepD&2{F&m~GxLBYd2_))|tFzgZM!!`Y4%I7nFv zEaZ)oa_+(3M!>8ENchs_0CxIbMgZZ?=mFr&m;&zY-oPnSfLYj@RFpLEfSJWfblk9- z#&w1C_@hG>k{(GOu37e_z~$P#0i!Z5+ZIyF_?23xJL0K4VV)n{utJ)DRIKSQ9~?GLKr*)dbu$RR&Th;EHiF7QP5 zCUrPS774v;aOIIuho+Op!r!1_7aG;gBqSIjyIIAV|H{4<>C|adcn`00;2rNN5|>yG z|8uoO8<#HXv$?rK(dTXb!=pXzdKU3sWw0{c9dmbpOq}kjNu4fD{P?zJF{4f<2V1 zT$=wzl1VgLgf_5I5j2*P1{DgHO%?^5P zP2@4U%<+2euWFj=6m8k!LeA8OW0GpytoW&!ND*r@D7JtmsxH9yj~Us)+HHtxRBd+W zE@GEz_rCW}Y=XrpUda%6aHI9eqsk-y4gD{=+jnzd&RI@Y_xt5A1ew}k#X1;Oy2h=! z69#%02995zYbu(y#TDTXBCcRk81Fp0Uya!gde)|S4c8=mG0Tkx+6v%~B_lW@A=&;M7 zb^pv4th7QnNy5o$_h{V*Ik1o{@`Q5aDEz=7f9pnkshAQ{{i+4+~QXILwt(tazJ@{yp61 z^tcCp-j3~CMmTb6&r0u2_MW)j7okw5_YjpvoFkEy=4As##Iy`@Ma64KL6uAJUhc~b z?vrJnvho^0No(2w*xksE;M{f^tb3@UCHd9nPW$^j&-)o9&~s8qO5kx}wUhV6daX+5 zVgI)Q08twIM0~^)Wwk^CbsLc27oPQzWl>>t%>hDL)x_sx)uV`g#D$yLYy}p`8VORJ z`lMI?t$0ON33}dJIsLy`*Gcruwf2w?bpJ~jf>_BQ!M1V4OJ917O$zbk8TV$(mn;Gh zXLZj3xb){Kv;p#Lkt)K0I>&Ygn;a9DYoGX0xz~*yJ>40>drU9>8i>Xvbv{)d*`tm$ zy-=g)BbvY1kjq+S3bp2m)|DP|@Kxx#W=-kwNyz}Lk$LX`c-p*LZ+seJX?mo!ES8RK z-({}pBQMUBriQ~u8WJ})E3b8)j!O~xpe>aSiN%v(or(a-f1d@$Cu~47+sgT6;K{wA zq1lkNP6H85(+Ta0bXUy=JBb6Pk0(V`S-gfd8}MY$+gu0Tkc5Ic@K2No zL3;MlPy#Y_k!AkX>3evmYO2z?MnrYE;w+od$^Is9l2^&;0jS(t=&!d8kA$(_hT6_6 zJlEr>?+ZU5ONI~5t9hDrTNr}sctka?H|(M&&{-F#{(j+H-`z~sNH?ijyN-L99vxs) zWnv|D{7TJpA58y3cHe_6Sf_XLhT`neVP`c}h!|I;Gr8>0k=0Jr7!ETDSYppLytB9s zE2Sa9k_uS+>|i}?7TfycyQxS+@^z=n)lI8B^}Z1LkZzZE75##pPLjt)ZSvnNY2CAk z{h{O!8)T~OX8<8$QtaM+f4^SzYI9wef8RmojxLTuj?wk}R8Y;{3HH)S*vbanY2aW|wqsPyaXxfb#opsV;L;4q<;X+_Q%< z5~jD#wm+N3%+F9(7&bYnsQmT(Nqjj0OViWHNi?J&$3So>u}sd%Z5$<)_tNx}Y94-9unf{yEVbFDL66_x9j9FYD%&yrHKq)u0d)@H`5X|u;*x$$uiu;V%*Jqejg~Jbt8v_1D`=y)9+nPw9+ui`IH!Ph~QwD_! z=a79w5KJ8iHI?*fhu~Qx*l|`LvpIj5R<0g^GQCmr+K0JMd2IP}X~eKHCA}sKfbzM$ z^={!DS&%%*;cQc%)fhnYlA4KK@De?}6cd27^0EeSSwHlH^C&$Pp}Ckvfa+G;+qg3q z(OMfi0P?Eoi)w1`p>gg$rA%KPRtfh@0Q~({G=PK2phvl!izWr&l8#oR&E>E;)I0$D zQ5Eb-MMSZT8Ms?^)1;Pu1Yo;bYajn!Ba;Z|+PyM~hFvWR2E`9in_gQ04oL)@hIvSH z?TKsuO{!BTvw*w*TQ3Gv+1XF=*+5%nK3JaV|NA7;%{C=PBakn$0j$tuM;Q&JAej!# z(sve~-0^sLuFL)F{Pk9I^QE`^=kJ`u66}uEtLtmWCV4Cdoo>o-rjEGS`~K15YdKpL zyvGLSql`HPpxRWllZoqRAJnRbSea2gTUSlWl1R5Ep740#LX}_rY$5!G-uLl0N+CsO=8F>n$Jsw}0NmU`_s(Y`BKCr~zG*IU!I(sTTM4zr z_S=$v_2WUv3U4K1{sRdq{&9*7@|G>+t-=zFhq(Q<9wX@dM>0VLB5Twc7sZ zxbq1BLLX(y?q$u_Wzi3y6jk;Q=QciLetv)X{1h6;bp-emrb446H9DZD3d>K2myy5! zKh@obR})$PFz^5(i!9)Zf>Ko!R0KqkD#Ze(9ye5t8_ai>%Gq2!=J!7#fqaDwSvdnfIQ7ka0>oP=Ex8H=^sy}+64`qTsb zw%gO78}<6|{B(1kT$T?^UDMvNvUX<3zM~|=^`uDh^Jl;ockOo17d^0_H+Ic1ot#K_ zcbB=eZ#Dgz({7QT6uov~IhCa*K=sB`P*nl}8juFlm6%x}Dy9Cm{(v!{7(74yAx|U= zHto<$cpD_Zct$#$bZzUIv^Wh(2k=xwoIWD(JA0)oEMrA`&db?pwr(_{#`H=caH81t zbK*h6=dif_l4gpJwrFTOde(OT@+qAJ2uYfs+WpOW%U;G&;MW(2UIIdI0?#rZb_(ZL z+4y2<1Cb`m3Kj;5s4Ww?&kiXB?lMWLLj8_?zmD_3vfZGEgUDNn^9oN5wjdOPLPjV3 z8&XD8E3EGbY?xV&kav)c0EG3 znxgy{VjFd%hg8PH)}JR&WT~w{JzAYW4@VX}712@yY$tMWZS7ck7 zr9IN9*j)GjMru-$5$Y`kHoJmr;gLn_jx~$MzOn6)Xgetnt5+v~^ZeBZ!jDz&)~@%8 z>g?m{0M9j$hWp}HyF+3|x-LEiH-}WZm@3YS= z9CG^TT_Uo09Stwq*4^;wdG0Y-Y?<)c?-w?S)>D%qZHv#D0*UF6hJyslducP#dkTP~ zrdE!CVl@VB$i63CY;YGn#5%TD&k!ImmVU-hM5=QMfY5Rg+i9?VWA$s;#m*P6wDe82 zodBicuT2>;NJm(Ft#pX`g5bMPm;S=+5Yt8>rv|z?>^{4|(2nVoc`>VAf@rgq{=C!|b_R zBMa9mSc6AiM_%ARTRXhT5AQ&~V(Sh5R=iRI2(0g#(G*Tl$8>;%ytqDS7s5W z%5XrY+Gv(rO<_c@Skp}+Hp}@bsXs2?{(1S`quJu@eBiMOg%^2a0;m28yS6%hX#UKq zH;1&n!E^LA=Fxjo-4@dc&5J7xe(zZ_EZh36-cCn>19))TT)U9@Xur>t<1fyz=F_E% z34;&+BYs6L8y3S6qzG3E$#weZ5yp@1b=Cyvw3>ndyUEgR+fS(l+s+?_U3<~^jIsFu z5sQe8_UOtoisH(HY2?=SE(vt2=2)9&{c7s{pOMQIZ%l3c0WI)=aR3&K4Fv=fkaB}M z0pZ7noIU^o!8wcO%#NsUnE0(B5L2hZx|ka^0`xVFl#kWW^8xUL?pZ?evCvsUG4u~lFhSW}eliu5X3Xs1XlD%+lQq;Mm z?N(ZohUx3ez&tyN;-Aa7KuT@cp0ZEn?8WGo^G{Ots|xgPg)~r`2?;@R0trOfu#?*#dZF_Rzl0Q37vEgpIR$xDPTwjEU%Hdd{ zV{xKuo4m{xsM&o%e#?U5gi%Wv>`%IMy?W$J9lw9d@sD`VFD@M$jNWq@KW@h$(J<&s z!JaM~h+pR%q^i0e=u!SWutLM;h{ej1b!@0Wlwxy14@wqfX2GDXc7++R(V8&w+nb;J2_h zAJ#V-9QZQg!rpVY7!9Z=f-`z|2&@7}~KKf9=7M$MppupV#=!Oy%}t@*Ft#q|G2FIZ=N z_UD2p-M{S+o_YN4pciIqVk)irdAC9}a$Rs(CMH%UJ>mjM$VfLxNz;E@%gt~8^QwEs zKE`J=Qi}vc1A7Dn=VA^9MsjjO#HPNepl=3i8O8joLmA4>4Rvt!XQEFoo|WAJ8@SL( z(0%*sR7sV6VzBtU%hjHa)~kDK;iI2w)f(c$%ob8fF`0e%xbs{A`3oK$iUpEOwzfYyTzypVnZk`nk^R*Gz3FbJRZN)@@cN#XWeTzU4W9->Hd2}546i4y*({p zQ2L>u-+qp#0Xt02vMhH2_PJKSEQgr6{k&>AHh*O-r@_L&WslK`Q1_*`LNPy*@~;=z zX7oP`lRK3g_5*_5-^jq<^#FHg@KW(QsI)si-x8$@=HB zwpc&>Z*|X!_`h;1!s2Hy@jjC>&VGtnWjxGlALQ@wCfHr1M$bdY<8hx}h;-`D0AWLe zMNE>Q(sZobQu|bTPm08r|J5tJU2t9Ke-;`f-G47~P4bTP@Azdq(b4K;Oj|j*g@r}+qUhHgQ z4lzef)h}qyy55Apu@sk5NKM$#U87HR)sDQm7E-j}IC<93o*DR&`A~9kdr%q!vNCRi zlsO;;d>rk5!%puzk;jaa^rqroJDeuG7#MZn4$qVgdzfffp%YF6?SCoYX~V>_FeciI z<6}TW)lRp_w^Q95TJXu}eb(5*HbQ#PiKJoT*+EULlWl$Nw?c>F5Xj+bLc3Zz{CLrX zT{+tK5Biq+!|GGrUha3??|dypa(vQ9kJq?em2IQCZE**_hLodMe5v*5`X!$@pUcg! zAD<49u_sVQsw#YuX=?#N)4F}W@96Df-WV~AN2#UL$BL(^d!;iecQ z+gm?wMRZPzmUpP8Fg|L}=JC2@IzA6=&CcxhTuOI{cEXEQW6xv$`G&r6kJ>@we%Q^v&f=bSr z*VloaCvplc&jX*;O0*y@33oX-4E`Q#8XnkYX4d70$A3bE>4erjX3cJ#_w)Y|k?U&s?3UvJt-t=r*pbdHd@$YYHY zci5$UWqoCfv%}klfZ_u;30N(-E3WxjhZpB~KV(7hZrq4Og+l8WZ}wLQOc2^J-POM8 zfb03@r>C6kp&!cgk%8F@8>#DRt8N4eUJTsCqcIUOtXO8`9##Mb0NO{mEvJ&*BE20> zn*+2<6q!oRyH4e|tL~-QLEKjC2^SL}wab-gh`KKY&re%%qB%(q4ivT{(rb?UHkEF) zt!c993#EnZ$Qs_{tlNV^hvC)IMK>m0jp9^Ne=91thJ&8yE<9T$LCp`P4H&6~*>Y!y zwGjz`RhhTF?nd|bvkYu01^-7Wu+ocdE|4v|b8)d?pi^xasL97wsL31#LNxKM5vjZ* zD_={zF_<(=nnT5P`#d|G9d0$#*JdQY+{UZJo89)|;eBxtxqMy~`dVw=GwaV@{!{Gk z!8g8T7woGx$2RWsun80cQULJkX52{$r2K$(d=Yefr z?(Od%38Br`j)&xgqV0pQ>$a@|t>%o}mK+!@)zrL)T62a*phT1UywR(T@yR??+)P0*iux zqD!7^bKA!SyVIqZwXk{_(t&chN zF@Ddm90ku~&lT0vlgR_;@1_i?Tp7HoTkYdV9h>nk#KdO+H_uYSs~_Tl0Za1ve&BYM zku`!#mdtNEmjng)xTu6Or zCMWHEqrbSOf+AagG7=fi|2`65A%oO(ui}vI2A7OCJ(Ll*q~M_m(x_KVs_uDkI!7?W z8--MJOoD+NS~k*~y;RBaCS#o(kq!VR;(eiE#kchpE=Rr>|DE;$jW;9zKA zSX@ClP<-Aw#q|DF`=Ic)Z|WpA6`2Yt^Q}TBzLeM1*D3dotGqaR+`sKpXnjUAWTtZ2 z%4MQoUMq53!ktczk! z&H+xldPek9^#E;eH3@YfViQ+kcmVx7EBU&ODUvz$1rYq;6G-Ub=*X-+Re>~22Bvj6 z7!THMmafk)d~A~xS`CMAczpOoQH#?P=G2cH3)Pc@jOIdV=tLiSY6P6?j|zO=WoInX z@)DYXK+%GPZ^SpAiQ^NrV<6nw^GhjHsyzyX>qQW*M*AUPN(>W_ACQ0CLPlO%UV5}7 zENGIN8U!C94Y1{A@&E7#``0{qnx%sTmEeMeukMbkufnsGs}QP zypgipVY!1bwST|zer1DuzSgxy-?-<((!Qnx!zE$~q5ya*&?=;^yF&~k(^}$~TV(B& zRafV_qBrMHW1LFF9UvYcPKaUpT=PhmI1bI(kInYP8{FHc4v}{_V}8cGQ3g-k7j~9g z8I=AG;xvVSh!US}?J@c>O_=LS%jT0!HBs**t&}nXP+?^&?sJ?Wqu+y*=l90sAUl0JeP)L+GN&|y9gB%-zp4Ym zwmWf$5Lbq17{l<{`I?U$hbIU?zWAzN?%0LXiIceL{K`vL_pUL{s%MNghI- zTVIKfO`VAcVp3JqtZXyK=*fNG0RlhG8h^|7z< zp43U~Y^S%=ALocc1yI;_t#HYHdXR?ste4`il0#H`0uiz`b_9diToXHIA5-@jCU)XJ z;XX;+NUu8{;T9UkgfoDTJrAD8K93btlY{Hv?oLcq^1N3lG+JsaM1V_$X6Il|9asnS z3*F(8Gg@UEny3JgNCETz68S45g#3LrXY(JzG@d}@ zF`pw-_$u-k*HqV3<+0wW-l;~dNZuE26DOWAC72IHiyO$Hbf)}JVk_#SRmnMO1xE^w zY|$C=E~Ye;U}*N$ReruMVempnK;rH^WvnX3HDZ`>Z*GJrPk(WpSRLxWe9QUl!ovyi z&+-agj+sRHIb&O%4Im(QQ0KzkiC~8jMja_4$Cy~-;ON`T81={O2=sl!1YbPU%A&0_VF-^)|mn$r?AWBA!T zvz7P&R3+9MP6CjbqRZT#mvb&6r$g}e)@iBc^mMRfo{-rP)mSr#J_Q)%Fh9%=1a z_G0z*i|FkQHC=*{_y&JL0#GVeK!ZS=&~p*J*}U46cL|56j33} zko$G2@}xI%WNURiJ{})9TJ%voPEJBs&w#E%S8-~f_ELK>j7SA@gU6`dpL>jaLPwW& z`_*$IDPSqwtnkszT}?f?x`2>kHYD2Y9iUQtXCU7W>+X*dw*sQ$a^rGENc<7z5P%BL z`u=!r&+-X*iUUFFp6bjcnA^)%omB5HQ?*^eS>r=drE*B!0iUCn-LC*Wv{lRRmfr&l z^4eZNOoyjRd)#vL@th3`U!R-x9CPE|tzpsE1s--?rJ3>$6;Tx?q>aAgDQ+rl7o(5U zAl%sksvV(BzRS}7oqFGM(PJhMKhz$N6P*DwCc~MU55zABd&_d6NlIj4v;8Hj8LsVlq#Y^zm^fUn`6a{G8tnY!41%P!odWySln@Zye5oTrTjrNiD~V*4i)nbP!I;Q)%7AH{fu#a)mlK zy=T7j%sc}<0}qilQpkE(e_KZMvxmUeM0(1rOg$t2FH;8!faqt*&ysg07DWd?T{$p! z-Z9K0{Vwq9+v<1K@BC(%aUEhWWpt8q7R#2(mV~Z?nG>z{P-smaB%mrNUh!yX@sz;) z=w{7kU?RUiHeK|DlzRY5RR)7V?FimrFFIvyg0Fk);LQ0~Ih0z@#q18#L|;Jjc_!$T z^EX$lRgILOEX4TR4)BOGsqaHuLm4I0m_O-ALmO|WLeYZy<) zSob1xPkdpyN7hJreXqUtZ6iQ6{{wkABKKt%c~5vQrHd>YksG!YlJ{yo755BFn4i3K zF&%5qOZ7Pd*PT_P+Tt#O*Mu&DVvW;u^*y)H(e(gTjgMsS%?r9;7ja^|2G)(wT?7q8 zR0{pz1M0VM8k+o-H6(&3_4}cv5ehsQ^TmE#=Y6cJQn6CXXR^^aSb2x^DG)u0PzXyy zUYuY$RApub1oO$kW(9pw+pLz4U@ayhZ|pU`PvwlUwg#e=ho~2Dof%km)2zE6+7W^Ig&_BV z{2+?-! zU`QgXPqK#UWs^{otVBOFNSb~YHmsRkLeEfq-$f<|po@kT2CF@n86Y!mn)@Zh4$Iko-lNo0qd z(!LM!SVj$}5sh^%H4MWgA_-WK*%}B?qHt(Lb*0kTUDfyj|M3yWJ9d#YaSz`(;hJV^ zyIR1a11dKiXH^+6(wK+K9=qnS_k-3yk^5O8@`1RTp ztBrs`>p!uE>#2$Cl{<`W!W-DQ|Qr5VJZtj1HFW*-I+-zXdZMI>&YWIgvs z&oJi-3~;kYz^(-dZA{h+P^|0yY`-uu#)BJ`yLlnnYLBt@AITl=L4!30vwUmrde_1@ ztD2^?-Obn9gTt#macqLMe~222>rBPE`WCql`=JrM)$AF%vU|sRt(-~&xg;yEtQlG= za%RTlz{^&2y(>NdePO2a8TRE$h-!=D#x-6b8c}5N#|(e&{eFalG$`RhIfoBs1-X)| z4Y+UkhqVwO_bT`Dn1>PU4?~vVLn7YEPq3=FCc?vv<8I4LjOZXwNN~nKy|H5S;w$i` z>We)LmJdiE$s|_}6tpbO2E{a|n#{Dm8b{5LxME+EM_5KEal%b$M0Y&Y=vC)(=;t3=Hp8VL_#W#y|>C4Y8D9a@X3V#oxDDBkD%Hign zk+hBputS&ndabXD@Typl8-pZ2;zn0(3%O(^i_|{OdIY5=R|`>8R90TM&P&>O-bK)# zS=nV)?I^iy0cs9?^PEH=CoQlL9ymOw_QF12plN$+G+f{?BBKFy2)7A z`l3TmzlG$HDyT2oBAu?pviVxXwydDVx5JbxILU3({@$w0lB%L?aIjD!IIilubG%5} zv^yxr^mwy4e<;Hjhl=8}5j-jUSMbJRrP`^r`9<8^7EqZZ&%$vWU#Cd$Yxq?2N=V+u zGpwsg7r~#H^|}t6Xw5jE`T(@D$((dCJ&#kF!CrY7ic2KYf(%>9R8WT)0|`jPnjSb9 z6RwN%pFZ&?sL|G+0~0Lgu7Z8mCNpKzXmIecxaTvGh)_xZdS=C9vc}P3C$WY*H^hFx z5(5>QSDMV{mEJLetCGkiRmw)Ze~q61zeDu@#ESokvHyR&Hd$$M6XtSzZa8TC7!n49 sNukVcp!{yT8`!!91bX@Uz%HLNSJ%?eCF&V+z-rievrFf2XPp1|KaB#EiU0rr literal 0 HcmV?d00001 diff --git a/src/.vuepress/public/img/LargeModel-toto.png b/src/.vuepress/public/img/LargeModel-toto.png new file mode 100644 index 0000000000000000000000000000000000000000..d6a7b9a3573a37b95d8c8e5235fbf77501f17017 GIT binary patch literal 246499 zcmcG#WmFx_5-vQOjk~)g!QEYUg1fuByZa^t2@)W9(BQra39bnQcXtTx?s{|H_pEd8 zpYP|_Yi9LK_v)JJXX>e%>8bgssw{(nN{R{q0EV2bq&fhAHvs@tkB#`Zr5HaU@^3|L zA*-$g0RD6U0Q~>}_kWw9y8z(D0Ra1^03ehO07R}?&1xclI}ptkWF%kyDJZ%(OMhEX zTxIn<{?>E<>`ccuHh&M0Jmr+6koFPqaL_TtN~pg7l?cd5zSZ-~M-dL2z_(axyZkyqMQc5MMpO#>dCMympG3IJs*TN^~5o6b=~iQ7NPO ze@0k!7sCBdzx9r=zC?wR_=XT2I!9~wS>%C$|Mk5G@MRJHcP03{@srvZ753{!PF_l z|5IjN=YvQe%Ktrm+4Zt^*B7T#$u?6b?y>abOA1Kk3OlEM8eJ_>;HKkW;GFltPEzq@ z@af`IY?rsZVcfVnyQ-JTFvNe_*`sLA?vEg$I!QZR&wV0QjONPF z5imo=K3upbMt8DPds&YDwf5cr@Oz9-2+jJxcUSD|5-XjKYuGv5F{7A4_hzDnV!gw| z&JAN-WWVuEQ?DOrky10WU#1`<>(#w6qs@y)FvenYujptOR?X9eXvUIm`)Q!{X}xZ5 zeff%|`?-Mk%2AtRz#Ix)`~?c=i?*ip|79P4sc);y7q~5FROMm%j$nl1B>o!+Pf6}6F>GieaqH28qyoK&P-nJUf ztN#fPmg<8j|8J7mxuGQQJqW}@G@~}eW1k<{_UJp@1G$sF#04LFtP3obujb0JlO$c- zKFd;?2gPG@XE7%i;QoJt6ighE7!uq{Zek%~dAZusA>VtFxiTc7sO->bSrroQcXx+?V6||Dwc)Bx`;XCgcZBjJ^b@5H49LqzN52X##CuO2 z^;xk*erUr#<%6HSOyBL7Rb21DQx-ZKl`W#vx>1Yg4mv^M4mhl{nSloctB%w*6oPSyjt<;dulUzQlk#;61OGACsE}|~ zx(_RrA@wmggT+496hDWu3t#_7>FyxGq=vgMD|VuD^&5(GH^_>|C--Mw?M3NJ86$-O zLM>vpe2bq{{`9t7ptY}!2TgqF$4N^^Hy$jsN9F$^ZocSqcSrPazG!6G8*~eVv0~K) z$CCBrn&kfn$6F8GkhUwj6=QzO+9}51^d60W+0%~pp83E>lyso}!9C2ca3-z#HimLj z#)o_U58~>#sN{dG;0$Ko`(YO$slfZxnS$RIogOUs7lmBhA&Cs;FPx0ozpxaa9?#gn zNhb9d(S-yXo8l%(+y*J+T5$c&?w7>aE@PENgRMEdr!qy{9}i?b-D=I&(e1I+{rzjQ zY3gx9{QCpE-M5G2w$AG>E>E{Fc%M#R4!TFkMssc4=#rYv9v{hbjFEzaSyE{K0yWx& zXj?&BTZJ~D$YgigHB$KMUtmu@meM9+G`5OHT=g>XTSlBOA>|sLIZ-t4YdaqJF}DW@ zGiLrn)qF{{FDN_U%8i}n51N0?^vuIFV!La`k(nW#B$?==Ecahlyoy$r!esQzg4^*j-K*YV#U*!GT(SQ$Fqy;MWV5}`;ZqH7umd^&Rd!H|L?kAK z|CKri{qTQz)ZKw}5Xs-XsWBt`yqxDBU5N9q+PQKE?9%UEq`^(*yy~z2F>1+v8aM6r z%5JI7!8>-QR>GUgU)P_0y!wx%M@2-puD;KAg=l|~SS7P2hE1h7Ew?^Tyi8-p(?}^z zp#MvYQAyaZv`Cie3zd77bo0rm-I{E1 zeuIFlhKw@?U-G}xUB#Liohj|AcQ!Jw)y7Y9qL}iWy3`~}$oZGk@=TW0^#2^*DkYF7 zWu*?K9+?`b7^Y_}rtwJpSH=suHL=c^l|hg)XW48{|0^;!qh`;&Zj?LB3H<*874fe| zXK1fi=^F@|;jnPo8qTMY3b|2J`|{=3SIJ(mTt;MCfS+Gaup3Tv6rQ?W>DoTkgtkVw zB^UhHd!DK23Fr0?OQcsWLHlRR3Nu|S>?LQfkS@eiy5mXsWig{ylbgU%a%ix`bHYBk z5&xW*=*x=&fqAdNuwD#t)>RmPV;g4k)6;Va2h~Q4G)dtw^T5upK>8<~zkmv&7QSDN z(RrZELa&8iSr9##lWaeViNKvD|g#a~FmDAR6eb2?m(wGxYlJD79u(L_U zPt2d<+=OosWhqG_hX_r?O(bL%4T>;K$sIWjOUP6mEHCUZZGg0(I)1&Qm$SW;sx+}= zs;8k)jP%%_9WH_c{KKMes*q@Y_fYi-0G58mHrp78hfG4EJbt&MC%H)(S9Gg*uB-v| zcLpwNO=S+foy?HqLr1lCL2B~1{XM%5QlTBsi=k>&AhRY_ zgi%1vDM}Snm9c6@au0QFqH094A_jsK8R}II11FnnHZzLcpX|r~Kfdzp|*Yl2v$=Z3jz;D>zX=XS$2e!fgcO{1Ko;bPn}wQ9!b z-dtVq>g-W!I-#Z)>r;;NBCr(CdGVLtu8z;v5y-F@2W^Md>_xD-K8L1kZzPI!c_Sw=e(gvoO-fOu6kJd-ZXufOih7_3||9=umf#+Vm7b9O~TNUNaxfm4Y<$J-Obj?AZc^6 z2Esz+zrOrM!yIFi1WjFLM+6SDtV*tp3JIRdj|z_;3%jSUF5DM@G6QR3!Y_A>v-}{9 z&;z+%bCX<9gkdIcvmIKVXQ6~fmkGz_Vck$c4on9u_suPgTOIJDwL|i=Gd1mr4Udxv z_HW<(^vkec2x3>`p8O6>Jt{>nyqsysN@N`zQA{h$M_PegIr0JyefkB1A9hiK1Hr%K0t;N;|2+c=TGe+5`h z&g~6%^j6Z>hO#>C5Yz3iPM?gL!RsJ%@o^iGG829`GVCeT#w(VMXz`@tIFIK3WcSs|Yc859l+)Ysxbu?}a52C9 z$ruLw7uF-8Sz2-^na~;;B1ikp!<9^fig=50naS9fyw||;acpe46V@TJF2{B;xjeU%k~-a4g*vgk8dbOG?zMya=Fh&ekF|8z8bYl$@~BWGKpnUO>EAIK zL!MBaa#JS3JjS-uI-E#3IgPv%FND}zgtgpdvfn%IZO0rcECq&dNnO)!(r=EVM}=@>Zib!hSVOb(6 z*X;C{@Yh4SLdPSx)aE0R1|LI#9thdqZ{w_aX^{hHh6l0U$O3jv1!lSqrG*sfz!0q+1uCw&UAI4C_`^K-~U0! zFTuO%Q#>#AB25+BD%w=Wn7lS0kPxfBOKp62u>u2tu#k?R`-X#*pu9oK@0y+VQ5NZH zM$|mNU=D$Frz_K44CkDx^>$mg>Rt-Qf}51A@fmL)39(We=wkZY(AGQL>?zr^MD-GD(CGp=M3dg zHCEzaxqfnh-5TSfPSrRy4+k@ksqt-E$Mr&F?`~{1SfNIfL|TnnWI#=oZ?n2Jtw^5; z3QMk7`M|u5)Pu~FuIY+7fhyt%htY*qkX823*gyqbE9pKT?JfnB0EiAsyQbomno&iZ zV|Sl-nxTYvcuGm-sVDi5@DV{%a4a@PRS+uqQ;!^y;d!d({aP+mgU`gMEu{8;WZ87~ z+Z!BFwt%1cS7YOA;XVl5;g~sm+aY3POKk-6;%6$eJwL7oulwXDiL-?~uVR(1Cz3EC zj7H04 zmMjsGV=^r$?FciXLj)v(tLM2G#?_AA8g6jr*VScW%|2$SP_Mtq$0r0Ue9I zPqSdy#`0y@3RruQ;)M7LzUPw>UoCH&7|MEDVK+(Ns))Y`wIo7R-y?PsvEdIBUV~-T zep2FKw;YcPVQ6?jMit^7HJqcOz@X3{(6To)D?I79%eZ1!Eot$h8 zZJm4ZUGYD}pFlm$I16-~W1#+St{=YS2+`BJDq7d*^C;{54c6UwOSWb#Hu7;SC5%7K5y3LMNw`MrRWV7bFO@H^J^a;W=G zxpQ49TDMQsgmNZ=2!l^qg-X|`r|;6F#rZv`#KtN2LbAd(q|1k5L`W2fWy>hBg8?}| zTAPOBNq)rvYx9}HJCAALRMCM$ZjOh`c9*9dG0Zl&+r!{LMD573D#~n`q6FCdi40Lh zV8WZj{exAoI~d`QRIj>o{OT&*q^{ZA~RqAr6{Kf&}oG)m--JELtZ5T%VPW0uD@SatCR4wH4tVX$?9#{g_&W;*hO73oCo~zCu|J10I=z4W-JV~_W znR7>Z8SnY%+f}Fe_(&5Ok(GNLj}Yo<}O}G8ZO4Z!JJ8(bc<96>o`%=A|x%vf0 zQbdq=d*=LCar7>XTVLTUnLOk*7mnNw#r9}(#NTtVpR3uecW2bh40!LxOsBgPo0sl& zRl?O%u;gM=#M|RrL5l-_G+=Nu+$2LK$D%u;lRAd5(bwr_MoczWzaeo)vuii3Lox_w zy;pRcG1F;7flMdNI`xivr412$iuEBsEsdN>IU8)$Yoq1H&&4|`Iu~ib^1aw|TE1LF zz~BzFKlTgzhu`&K>`ny-rQP55`k!W;@b9qD58_`Do$K7idy?6=Vi)(BFxxRn$?K2OC5h`ClA($aX zlwn)7U`38ud5a|PxUBBoZ{#Q#WfRX4b`7_c*|LEk#Cvhb-!B^!3#g)v`4zbow3Ex~ zC=CyeBdVi)Ua&&xJnJP}75f2BCSuCfnZwX@h0Qlzp2A8FrgojYa%7kJ~otbrfiB;Domk>+D27rF3Z&@0;7-Ev;5`^CDP| z7Uk4(N6 zS;X~S$eu01r^jm*b~uhpKM2%;Y}F^Yt1ZZOxaY=_KIFN!bz=g5P0i{0QG1XgHqIE7 z@-Q7aj5UJk(av*c#@UoI6K>5?s`hx%>HYTzKJgqa<9J1}<43nx6W?{Ab zz~Da6{mNaKP2RA|^B$8`yF zPIdr+*au+QU#Q~J1(J^~Oa@emAXJjnyOR&bV=D&7cjjw$M+U=B+ko%K`9@TR-U=Rd zmTkXmn}mRlJc84;`>ic81JIGzJ@0fNG~!9969pn!f`x<||042aMBd`_spc5T*hd-uhbL|R1&eNH<{Ve?^Pq{;zf%(O~sin$e;ZDh7 z?FlzNClLcgyv9=vM_N~xS#iBXFEyT-L>(+8pB4I8*Sy$CEd_vx)mqKUL<_d0WE48> z(>5sITxSSl0z6Khd8m0Y93o{i+|<@4r4j?gdFQ_nVNN-4@;fOy%%$M-jdRp@RBPQ~ zO2+Z8Lub9Hj#10uk>ft_zIzSim1e5roci6bd-~d}^)F_+JROFPr=(UorC^jc%%E!p zYbLYZEG^1C|DGT@reu&?g1gPVJ|y9VKRplO6Ri}c*2KXNHP)m#fQMKnoLsDUf951#9bL9 zj!O#HQjo^i76kOI?$G3F8_!>(tnb6AeCrDm!=0Op&8b*^Qlt zZLD-WcO+iCukG0KgJbHTpj@Vy3-FrP&|%Fpbq_YNBL<8Dc{;c|#~rkP5S?}(N@uMK zn!ZcYs&{$-N&5^>*2Op*QyJbg1_nQnxlTRiZzFs<*aZzAwofzDU8;v@3P7)NOY1Gn7Lf`>CT9u2WrmqqV2` z1nmm$m-a&@T{m3z1E!1;Q09P3WgYjhv&ApD6SUkI(77Tr$kvzxu63}keZ21avi`O1 z1va|OpLw4G>PgX?fqBuH;@URJ->8_!=Q@7ly>mlf`?mS#$utc%tI*P}>_=y|3JX#= zjjLDAMD}IzAfqZo(L5RyF?Mkhx7+W3_aXaHA;N8EiGHH~MdalfHXBkVva@elv?Bx1 z^E>IC?ouWslQnP@%pu*Cj!EI^*ZS|==3_o|J<8dzVkBmETOg4Vrlzk^~z z&q=q$7;Vvf%SD|D-6^o0k$pbQK-%*7QU=387@2h@Czt%MohA9uno|qmH$RQ$B)Gt( z*Rwm7FOrQp7ILd=ZS8cJBzboq_1}VylIT41gdSI6IwoE6gI{xS9BeS(VBOYCw*vZ!5o-gv-Wk zUPsSfXgWiOw<3N^wjW&~1v#AZmw~7%sz$BEon;A6EYnLk0Pey+u`Kh+SI_MPtL8`c(Rq6nL#FB!InMg_wap$s}l4-uWGG%ij ziQVay0xltn0>1rLyF-)B;I!MD-&aKiTss{wrah**esVJcTK`wApTkRgpgWPJj4DJY zoPF#y_qZ#e4YP*1{8X-GL;DJ5&(CaI9W`@5{@@|_yUu@QN@7|vP01VCMJeKWjX}5L8m2p7pW4p6)VsGtNN1Fy z=QsM`v``g)kd$<}aoxAFw&7i>8s99fxK9$)`t%z%uWZcYc7p0b2ljM(ieE8>GIuow2M;sVlcmXH=Kj{xTa4pKds<+ z9$9~vhNs=_yu62cxgzo=a*{0Tmb@~zHKU=!TjTZthISy+;qjoAd$7<_KyRV~)_kXx zP0&kpghjsICa{eNEhz;J=06ROdbjADi9v@|e><;z+C6M)Kngv?b|&8W5wtK`XwNt1 zs*6NlLfRQwzj9;F1QrfpBMo6D!*)is_%P8~*V233UJqi*v)-9lO<0M3_R7%>fIGUu z+BG;rI2G%(fNavOtqn{U9N!US@2DLl8n|MQeH~aMnce}uz`R~-r9DA!K1c6l_Td3= zYvC8rh)iuQ0twUZ8U)*Rk_5%i=tQds)$szs6C2pnx2Q@88W5@;170KGfaw#7G*jo$ zH-l>*XOzkz3jW)`Z*JV#`iRf_-kq|e2rBycXdjDpe(z>Sph97aZxtO9le4nn*Dx0d zS7G9#-Ll`w0202=yTRECP1RwoC6v5hJMlYFwUj#@Na*&NbPj?5snEg-lN0pYRvMdq z6^5^gZ#!cpiz~8&>kH$vOFFv!et=*-iz><07N2-cAQHQOh~NlzU(8S)Yr5Kxj`^DE z*J(G9r|(D7$sq_I8Zfcx4-t$-y%tY@5{(5pov@mbY823Pe~-ACMMd*>+8=PZWcd!>vp9Gr0vp?rNjke(~k4 z9|{@Iyv|Fpa#P5Gdo@aGqCi zI_VC(wL0bZO&yXkVlfR3>#e@YzGt?HbMX0!kRbA^-3Sqi#LnJwyAS6zhH54JBqv-m z;o+pI*V*~=GIWNg)4a4QGnBmMM?K7<&tIWnk5wMKgYNJ0C}o3?ebBx#FbTrHu&ANh zH8icK>XZO-_fC`B%|oVD;gXdbs7%u}4k2W8OAgEaCn2Q1UO!n6CE%c1x?va4J)FVf zW;IB*ch%>C93Zg1Lzi>~`84xrp3a>TJ8sOnAE?$8)qE~u8pE?Vv-wo?%?#&Aoe=0B z#iPj(4dwF4sz|3qcx^7uKyMEgl-BdR8;1twk-vjKT^M!hjC9TWdF@@2#RHYU*sreL z!|+`R7J1VsJsy-Is#(r#3^y=bx$yf}p>4+uWLd%<4-g4QYyau;Mwv1rX&X-DWUpDA z8VP$J+2+G1=pFjOEOzXOs;!%3P=q;H98pIgy}dz6Xe*E z!tQbo-WgVy#Xn560wSkoM>=vrmMWvAiM=yjL0}-x+Uro`73SD(nU={GpR0kl|0)?P zH7wvb?mcGb@5Kkt+yDTRldXT-9{N^ALv)ymBmVI^VzO(61&lhgd6Oc>G*I&M#`?s? zoic$3M;wJACm12i8%_k$G4)4Z$beT|vwypiF` z>1vOS-d;fgb)e3J@6LB`cS3j7%qBn2@0fT1rLAqo@8ghb_Kd2V`bpRHpT~V*Gho3S zJ%@8pBhu3Gb_k|zhxOR9JqdIMVHuk@vcn{}C>zzdFF*9n(JWGv61eup26K>^@1ic$ zt7UpC6BIp}jH$o_njPO+@y$l6EwRxYje&Pw5uCpthBKuuE>P=EyE<=QJg`YDfO1*1 zR~aw-M4|g7J{o^sP%-?y73tPFq=}}L)0O}EQH`Al;?>dvZ1q68hw~6W1^|ZA490N9 z8L7Qq3Ls9>KW!S%G^b|=_zjQLRUtVcsFV1Say&0>=Yd6cf9_xrfSk+vJ8pl%BbZj1 zA7kV_wC!U|O2)`d1=dYA`>iiMTn}!Hp_2<_NdcGD`pzY_6r&T%1set4`D}$)gF-ea zWiyKabysuxM!f;fxkKFI*&8oGl(oR0@*-Cl5eUkri?db5b0slp`F!QSRos7tlf4`x`1`5P?2$dPl#$6HzGUav9TqsuIVh#_cn04okFBIQh|bxn!&FV5wPlr z`x<`B-}p0(CvHo$DmL2i*^GKVma1VhVk%nXUh+YcF49#(qleTQyHzB?$tCkAI&^gK znT^x}kffoLPQu|hn?M~NH zBoj6E7iL@**@vGPv2eKybdP4cn%$8rk|K05Q4i7#7q94uoNHi%7gbugNob9K9G@+& zvs9WwMpOM1jVp)IKZbNG6BFq zO$6B4?3(~4mMa?e);knwUmE>*qek{Df*T1&HSw=XIiGi<-uc!$6o#6!E5Yy6r>g72 zG7kw-uXQ<S!U3&jJ)OzD+v1gTsX3U59~n zqkOY|Ex$2NpYD>UUdTrL+C+Uq8|A_FkdVbD5ciFMuT_s9YMH1{G-{;Z5UGn0E<{I58REhF8KulXsr&XB zi84PjxfAh^bl_UE_*1i`(YP!DgO zLakXqnrYC&YV!Pesz;AN!5W(oH-BgX!AC?Pw+lt*5kKZV7SP{BT} zs9gr`1j%N?Syn5s=Q1bE#A}q&V<`D>)-l52v-^>mg2SG%3M@=(Ipq2n-JS9qC z4M(?vjt{-o0Z!A;aJ7dIr%yNinq`L|!If@bd%=>LXGKN)nVMP|2PujN5D1mBTi=ko zlR4j4XnAVfsr0Y@iP(Yz5U%CzlAo?U6){k8Q_FMGys|@tItWDQK*?*j&jlf$+z8JL zKt2_V+s}j$Q7r-7HBu?C+!0#|wz@mEq|75$MjY%-v78 zvAXN1r@C4oygxdOc0czyo<&h#<{0DBLDbGR>r!mBADQ(9 zI;J#+AHBTlTCZoriU6UNW|i!R=g4%SoqYiKG`^)fc)53dkj`)#MOnCvGSoGoFb3u zR@G^2QBvnHw6luAnX*2gs9gETo=1gb2$Bh8e>d`7Xt52c85K^MtAPBBek7SN=U$8D zOZ6}2#^KO6cdl;3O32b0!&gB<9LasIKSQNHlRV6eMT1NyVbnS*fsVlUXj8wSv%*WS z!w}(Kfr2eI_Q4Zs6Wp~@ko%NP1VLUP$u9MCh_y1Tl7kB^C znl;6`s9)l!)k(yTgGqhu-F>e~ur}09;jBUk*g_?IQIb#eoEo>RAX*Ynsp=He)-D3- z&z4p858nw~GJ+x>Co`?l<9x*1MOHgP+;Uau|Fd}I|6Bw}QsV=jBL_O_A?QEI1oBjP zoq#8^$P?ns^5lpky&xo0!$n}M6=fdWNsddt4m4}I`>i)0 zBK90DBJV;L5za%~p$4Ts2sGgtAeX`+hUf3TVG45ftUpx`+e` zy}eiM&~O(HUP^l8BSX&WlK}M^nYkn8E`H1M@9XCOTC}y>=;xtjoP;!Lb-r&!#~y3; zswn8-@l42iEXfsIvdNcpMZ__m6yWZ?6>(`?eE8VAL|yMq#rabKsgag+jDoO<&)B{!JOq#N$wvcE#`zJlY-CmXx)y z8g&-V31Nc-7EPy6Xpv7yA=XG2wy!o}Cl_@hnQT`Ripl^HxPQ$=EF2HwihCC&(`HI_ zYnp=0IQZv{z{i9og+{%#!g0Oo^mt3n_catcQoKczA6!#DMU)*Aem<*#1pI88U>Ofq z3t(&(O@q)fi?v$_euNFw-Cng7Ge}Uu#_7=XUE!T?Q5VDcKNpvT72R+@RrO+ejLs8n zP2A@oD%r^=_UMe?zWoA?m=<> zkKN2Q2W+lgPj+JPPmPmQgHQ1h(-rTYd^Qtl5zloUhVO{JELEUmw(`5a_@HvmcKUWRM2%h&_ z+a+IZ4VKG#3`|G_d9bU=n9DCHni8NUzuK$^eu4ai;mt&rHsx*Ou!lY~H#b$GJI~|d z`Bpt758Glvm5Q(c{p$r0d-YIBBVB4GEwa~52`*o}8y>IJzlUSi>d23f@&PPZlLwh= zHP5axVrLGBxx^D;GM@xm(5Q3{O$gCw7g-V2ppB86akt0`R|0M3&S=z}G~R~(wA2Aw z5l%Vuu0dzt1SwU32hFqq>l;967hM}6K|n;CHwZnd)# z8t9LLrGS96c^|#!=T_2Hb6|3*+C!`Eq2xx3wD7RR6$3Jyucr(ow26CISMx)s#p#N5 z7XLhLlzYQn4{H0bv>{j;Da)`yZ9fRnErQ0w-VxD)v9sIeWq3$czmJy|Wg(#(aKo&> z>qGYJu1C>B)^<|7J=(nNu|SsyrmBYlB5%CSrAKCAY1Qm=><>w9$0Zjoz7Iu!PefMG z^$cO3{g>xgv)I_p1&A{sU8YzZ`{&kkCzG|nUPVkzF=LmAIK;K_heR3yyuN@Ojqebo z)1HeQvw7&v<__Mk8Swh-XN2`}$Zw1~t*QxVaArxYZ7D`!z4C5y6N*4mjmPffcgULC z$!7X2v^Ar(N!5{DhWOPiZZVweJ0-_^A}StZ%kVIW)@sC&BM~^IVXDP;1HO<&E_QGV z{BET!ghU+vRA}Fr`v=onz-Ee!q!_h0ZCEQ+f&e?>9uUZt3q$n*<3{#Uj_AUzu|emh zLJD#!n_O zk|jP_r%E+tyrKz&iY_pE+~m!*^=*qLUqSVR4B{uVdX)R$fwRdgWu9sRSj3dv6~(kj zuT2mF=qTXNS6)4TJR6s6Lo+GG$wp=J<^_yc^oQvS`X6XAC0 z3wJgLF)x543BiUs>_$EPUcD_v<%UdYPF;IAL~WG`n`Q*W>uIw+Rp75@g zBlodLs8Gw4xwow-VhtTa+_Kt7Mzia~8I;!l28%apO0<=tPxk8IT zp3dp*^VK^bF>be(0fa5}(LShEf83`$Z55?Ut=LX~9~$?z$#NqHavJXQu%QfRhw(W^ z=95eT8pVz<27FGg4+iQOcnlHad5Bx5BjQYcIC}Nfi?7bPOxh0fAm*|!2|6+Xe3@!x zPJIinsdimGw7j7Cv--{nECi_}ns3_->Kk7OE1g5iPi-IZn0A1ezog&nh18zcK9xn^ z3~$3x72gy8rXf7)NcHvneP9I%PvK2V5d*dAdxuzwjs`Gxe42LdG@WHhf9}-8*$*k4FnO;qA8UPoPc{W}-l6=?B(8wB z1oW*ku+no2#*=YWoL&~6(XHd>)&|TZ^hYkO(Qid?-_0)EM$#6+yGl^qWJay|>-OZ@ zVB{i?g{$PHbH8q>6oB)>s%R%-CX2crGM<;afLI?i-?+$;MUT9zpu2JMSa--_KL6Q( z{v7QuVx-G-sB=1(o1R4BG@qaG%ws2M9l*Sty6mQCOD%Yu6)hcSWBoOb z0Gf3N(d7LBJD|r0Gnp!12T|z9rM@3yXs=!T5x+2Ck$qKXkH1)f_C=jHS* z9IOgIQHHDu?^4TTk5IB7-)~nc=;u9*cXCR~Mc%)?kW#$<^DGPY73Rol0COrIHV+0m z-ElL|KP_*Ez%Fagi-xa6Vgsh*)^lpKDcv*QQxEf+=e{>1p+7Vj)U>4F3Sx0FGGU(_ zspDA*oL6dOo|C7krvIjvpx`X&Y?GsQ{=AC>)P8ezX4W#e#(fqLyu(yIWOfy9-!6OV ztutDB+4EQx;SBDL5avpK>th-CT7sj7I91?L$DAFfgS+y@AX@Bl$mB$$Usn>L={DCP-BAmnqrQ1oW4Q~x*TqcK zIlNC#=ywb+Z3_DG(mBdp;q@t1dKlzJ(yw#v^*L9m%Lf^@B2>tyG^W!UX03(kgKblv3m6{8?O|HAyrX_}<( zUFRm6_h%ZDK8rlGkvnmCG8LGur#iYSzq$#ib{VF!j2ieOFxD_pD2PD`ilg)i)HI_# zeLLDKYK@I;Ea)n|PPevvtfTXq2ihvs6o+aJN2x&Rg{lp|U-Tne&6upQ{%N;N2QL(> zx0Sgk8L@krR&E^$@;9@AUCRrMo|(C2cLHDSB+z^|%|YG6jn3Hc(&kToLg4jkn9TRv zp>yDxTj#Hyk=Z=PzrWO6TUC4t3~YF#x+M{mp(Nm{8_+f$CcB0~`KE9Vs<>XJ8|L|@ z(=@XzMw=x)l-vYnlSulO^)orJ{2)z2Y6V=`e^E(+5mJb^2gOVmaKE%aU%rnuw*9en z4K?>9=cH+qw#&PQZZ~6Ei81MXt&N1PM6M@-EmZxxxF`?IlG70#ag5h(^pKJH-~gANRt1 zWALA+Gb~r5259Q+qg_H@z51TX?}U6fXb?Pq^*SN^@@Er8=mqmlYd}35HWP}$2l_|B zT!3I?s*fcIf0tN9vPHU#DtF{%XBN0E(1j|inFZI~S_FUGcL&=6x#A-Nkrb{-HE@F+ zk@TljcEZ&mUFA-os|Eck%@GjgG}RG0Cj9zSL7^QLU`Ju+GEYoxr&g&a1Qk{ot@NKn z!rt!F`1?tm>{247CFJEr3p_QM@&hPn-^TEY4yVyJE?&>$z+?y-eY?6#p!#4HgNNw= zP@sX?ugp`)8aUJ9A3hgJTLnRm-=HPtB6wc!W|q$){5GAMAaZ>t#;H3}YVxIr>Z#JdN%aSkXec3)zhbE;&Rk;inygzuYh0;eo||wd zUrnp2JK&3{10UFhzI2iuBHk=QR?SsWu?6a;it|sf= z*CE29^FizT9$;Gj#zOb7JEr6Ly6}K4w@{^>83)RUAKZgK!GJb&c*;bT8 zLWk^_Q9M2pA2xtdl=w|ic{3-ew^31J4YA={he6X?PGfaXCPsl|HUpclrQZ$T)Hxel;y-7A-rQD?s?y7)Bp zUiwnL%@gCF^Te>OwX&Eqtw|WS2s1aqfEQ;#R-bb3-bOm+{y#*WbzD?U`~NrSmXroT zx>I^3q)X`r>26qZ5eZR3=|&LgknTpB1?leYr6l*qeLv6h{k`VTbIzPIb*{N)KG(cI z{z7D7QF6`ic&YLf&-}HBzMm`F8dP^UrxFeex9Fn|oxQ|PtGumBRc8@#P2XcJe_UZl z4vGglK_zy(nT-T@%4hfQu@PL^Ih9}PiFSMoA4^}iEE045t~r@uI|(D=tZG2NM=eiQ z4gh^66{e^^9<;QPeEpuz)XBSt*E@(u7(}&$)*RnA`|uzi#7QlV0A;`8|J0X4*~PrS zE0rh&H?cOSuVPV9E7^S|!c}7?9bRT!(n@GV6duWaO#9(&m$ahotVDgzldjS^Ir?UI zq4A~n#)H5}JBgr6TItBzJDHPY7+01uq49-FI%gGg9 z&ASPZ^##-2cuVgLF9FoR80aciNn~D5z-Y;NxHd(wMC-LNH}uXDz&X}k^a216jTMx( z2l3)}GrU~{4ENGefr)>!UVUW&=c~;x^O`*(qgSuD?%JFWrde#o%npqQrER26f1YeC z(|g?Xui3;ov%X>)0*g0`$6!N&D4D036nDT%cHPcS0UX;3F;}@-eu#JWu+Gf`OyvEY z8b@6f3ZidYl0^@YGeaxSBgHiJqR1_Yh`{}()ii(?U&qU)%zuy!z`%YUKJRVXW>Vcg z$~YAT>5cFbGKc`t`1aqwCAD~^213`V2nbL|5t!KbgKshB6c?MA450d2xO&8|yrUZs z=Az#ajQjX1{Yx=qxVW-10SnUjv6zSx4bWv^9x+z3i!5$){5$E)sIg`~5|V!cBA0ib zyW#I{Ac`XIKi}W^tLzU#R3q%At1Dt@@2Bo>kiNYMW99>GFrQW%^lmmVW?wFNB0A*Q zK2LX21Y4kREa-ce*y^9^hmC9(lttIxzWN*mshd1^GLjbb52ANQkzQzHs2%o@jk0MS z64jjVCK32*ov;gwkR(ST*T>te$PWsDn7_6+1@~x;hfuP?63j57XW;>9%YWng@9bz{ zHRg)Ea~rb_ne)-yQl=T|n_p`9Iw*QS4U(CFt3Le^MT9&_F1)Lv&;VZx>5ldrGhP`Z$WClXXoW0K+&Z2HS4CK*Y=?ol91cYn9oJJm>BzqdBn{}D>J5h z@m9yrw}H>2BzFxxv;1$JOpyvXe||MNPdfp6-IM}f4tA`4sU@Rr0Eo2d@82{*rcT4< zorVBj;}lj$dlHHS$bWQ+EpKD}7zHX`MU8l>au^x;WW9#Q@}v$i@vF1GK@m?Ki`V9K&Lo{yreu1W`YJM z5;-1vQm>u(>E04vlRcnmt% za@6f+0e?H|c~{rI`2{|ZW&esEc@h;7sTa}0I-?uCKC`mtdDXEX3z1a)$>oAqcs-`1 z3B?`uZFyNe;Rdk_t;)1uHLJFrJ}=paqy$2$KVF=e1|jW(vL%|7M_R>m-o7sTvQD29 z;9bi&gbDNF>{G~Nep%mH>*JB?Op2w2QTEkzo?3x6>kro*;h56AVRaQ%hac4%5!AE2 zm;2CvXC;;k{(gw_VD>kPUcKGmfpRUZ22+t-pBi{F0GIAM2D;En8k33wUG)w9W3(~T z%!uuekvz+Ir-2s;&7P-~hPFli_og{=Z({ut)j+7V=O+VIfZAG$EF}`4htvV;?Ex_V zkxiZ06L8U`H+gr5tVnm-E!TXWQ$-vn$&qi2Cv5YW_(|#?vz>4LCwN}+X^wFz$k$X* z*FOj9pekk|{h)T;xW?|ajjm1BPB2W6H6MxlJlq3+#ffb8YJ~mJ*GO!ik0=<8|3Z=W zyk1y%vUI~`lXDUMJCN2Q_4LC)*dk2QjafM5oRx!e+cgv_W?cz>66MqU#_?MatvU)7 zXY@sH6FPQ0hBmfi#$fLIF_HOy1OYy7Qlya^tr^19hms$QnRx>?Q?3r;roAz0a+g?H>Zl`uHdYMFbkuK2eE4}JYuL-2A5XgY*3P!& zrtP!F8>{b{UiJN2s(V8jiq=jbPVBk@XoA@LNdvY~Du-Hgs92|NQl7iz8$lgYuJh+y zcTfdnpR+h$==RP~uvFW;t;;d1(=fA6@9EgF*iAYDAUt`CZ*~buyNc+`x@@lfyy!X_ z{h}z~IF(0j`X9cjivpE$O z0(rkQS~MC9fH%i3gzJH!s_voKe&Bj7wIc)-+G49`dAcqxgp^J>ffB4YLgmzB%*fq? z9x@FLgzN0#c90t*9sQ!MpL8L`qA-Uoel-^&STj3Z;<6an|LsQz0+xFgA-(VH(E6j4 zHm6E$MEO^Wd5dd6y=zN-+=gHJZ;3AWX$go^DN6u`I9JcA1`f2 zYZQ>gKd2&11$?f}L+0QESSR?T1=ve`H(5Occ1Tv4_Z&ct!+i>Z76!dlXU-fF0`Slx z*0-R62Swk8&^jW&pL3>Nqg@Ptp%R#@X;3NU5QPh_1`t(%=z8EMo&myUI7zuv+0n6Z zHZL82f>lJDjxoD%N8S=c#VQINk)3EG08gyqSGy@DfuVxQCv-p;BNX*JD_sd!GqW&P zQ|x73mGKu6Sd|z1T_OU+NJ%{Z$p$ERB_N{ro7Ppg1s7>ccd43drSqav%f}MQ-y2Ys zUx77gRSOi5#;^yYHLH5oGKH3D24x>qtf`NgZsemIBm#?t?GWWE+N-@$&08RKz%To? ziFZQ4&qNWe;;pzaT977w94T?(5C}Ldt4oD84A(cFphTymLOsn9W_*25#_EQKV2hw| ztXjfeRBzIP_n}5+lG9|o1@@Q5Nxl*$)gyZaMdYW!On#oJjhI8OMu$nD(~#TfP2Ra~ zV9RR{6!2UYxh51Jcx3IzIi&=g+7?{4^FhIY?ng9Ogw)2MI%)+q;1mW`56tqp0Fm7; z#qziyMC>N3QrYLx65Dq+SzZEkMO^yak7VI(i7rKYEHprb!R8y{3ozexfyAgj;LzKM z@7UXV9GZANsa&rKC_Z+2ZhMUje7SKTzx^{By@f`Kl3WPbq3PJgvsyP|cO;AjTj$vl z!s?aAD5-fXQDBR2J2JIYUPs|NUTQpG&!Hib*H)nFISwtR+1FGW{M*xg_7oLZ0kQ_q zJ?fVsQ0WR=;CZvQ0D8R;DNHk4;oiMmMOBD8TtU+VCoY59H0~F!@5#BJ$|m-zQT~$4 z5(ywtr??^^A%xk+*}4Ue=1SQi2{Hd@A^Zu&^{k@I-b$MzGm9frB(ZAv{5>4 z+$W(IXxrH}vL&|Jq*M)qDY|6;762m--QlDhpxk}*6IB|#93eGx&GAe8ePg5jYE)a- zRvD54eppE2fP*T$-6V*eb=uhe9QVe!f1)zI`JN5*K_&7TU&t=jcd>6S;BH?05#@_W z<_|oS&v%9M3^%p_1{i}H;2z_bpK`lWh(p^Ryq|nI`(u}&$xiB58P!4ukBPLpS%TVL zrCNlrn6%JvZCW%ciWcvOK!CJZr;Z+;k#B?WMykZBMp15~K26Sg(eL@h?BR)zhTRVo zVErjRiY+|UTpCpMZf0b_1}oSqv6l8SNELGNxrTP^KQjdj!LE{@ zRhaZ`yV$nWpSj;?xUt387xt;#*V+$=zvr$h zlZMZVEq3_CMl6?igzk@QcJdUQ7OFC6%|gX)63WXW<{ol?yLU6UyO>9AYh!8j3pyd*q#f_E=+U$sksBMR<(TzUzb76Z?d@G62 zis|}z{4nbx_fX%pb-vf_er@jf+Q(E=p_GLYx|V;k+P@+iBMv!_8-r7aD8<|38<|Tz z&l_7_zbiNGyRm9|ip5JjOi!a zh{UMurB_Bt4y9YM|8-KahY2mkeQtMfi3z6o3mM5|xNn>I1`#ponNKOb>3P!}j{sR~ zT)Ts@ym^d4ML(t&2g%zhDh6HK0<4IS07Ys|GJVS{4t}qrn?KsA;X)&K?Fkf?Q#E|M ztq_{D6&ZgI&uk_qo8Kn>0o_$TthE%^_B#3*6VhY%JcCvx`<$9@Og}s;f)0#tJsBrIP5An)LsZxg5UkH<@>j5jz;~dMCkR_&bjD*gi}UzdKcl@y6UqA z%vEm{9v69R2`eua`N$P=p|y8iQrbU-Hcy;rluN4d`M>nBKk8n-Cw~8^Vu;B@VbP}9 z%s|l2ZErD;N>`%uJFEo^yPmV*GDn3CmCgH?et^XcQ0A+DZ z;@5aJDqU>COhN^?CzBE0gAOasq-0qT1v&k`qBgWTF1X%nv7S9mYwPGFDH&RkLj5tK zL|Ekqa1XpOUlZ4uUwAAG;T>DHHJjLPg>j}tJ5Wls#=cVE)8qiie6!)UEe2#1wZ1=( zKZcCiubllV00y@lkxpSb4sI;ux>cr)j|dzYf}$xrRovN(6r^=LSB??m*ZYExKJ8nI1ScGq-vtD15}AGU?Q3!iKDoV zj>&)uc=HzO4Y)DdU5La8P1@xqJ60XV;A+n+@2xUjpr#UJlyvp+!{_GK%LbqcSRac(cX97Y-F2Dy58?btQuyW5h@ZOnT1 zZi+|n0e#m_qd{8Y4^CUl(mXlOe_p7WvuiYK4w=Ldc1&EHt*+KV?n4WS>~w+VrAEpM zfCQ#u6`e{H0M%pCS&WJIM)UzS+J$avgS|1xr2ph8juQ2c^-EMsV4Tdd)-*am@4lxC zp?9S4eu_^$L9)Ng3zc zSn5*{9o)zd$>XI~*TrBJUGBC0yKVrKdL7*A`uq~+?9GroDT%|9`L<4%3RV1F`}?)_ z^kjc!Km`xT{9Ysm=!I9qv6BL*ahh*;(iY0R2DJ)euD2^bt*0yVk(flC0K55iZUw~9 z4iBouJG4<`X|W|TGO#!0x1Y`d16%BbG31j1o?@vkSU&)#cJ@1!2+dZ9_n0XT0HQ~@ zgC+F*kv@axABhtK`17S!|O+lrz@uwvW4$mECY@01qL zzqg@{)lHk0FM||hzWM0{(vOUb+Fmrg z_YQjtJ3ls%D60xdZ&ss0BmhJNs~izl5R+D;>kq(=q0DOobP$V^6Ycw1ys@asqtYfE z49=}!$wz*Nb0u$<3j%I{pHwEn+Y02mKvakch7w3{AF_o6&;9XS(o%bJR@le?k zOdPx4{DD!HcL*O&T(HVNDF|G_8M;%Gm29XD)30W>jx6YMs&toZ4e1K=E6fiGA!E<3 zumU@!hjR)8giL3h;)h;W)+PtNASpX|Ww--`y9vJlKswj&o&Vf+eD`n=IK~B|xQNTY z()bzuPHJKb0%##HcEsK3<8;P>x`4F$D;nD`1Uq0A=J&}%E?eOehkR}H?Vl%5G;}A$ zi>m;(@w3f>5$S6LhKUMu9AAyqwShm>{0vA4Q;>` zlZ)=5LZG%eZtX5+ke%1Jhbl1aY5zREO$pd!&S;+uXxC={+Rc(j+(8t3|3P!Iymu5^ ze$Zw#R<6Md3xFJSG0GJMxHw2GQO$!=sH)3d?)YF$McbNB5oWoWy*z**^Jb4(=g zx#G20qWbJ}pVUmBgyeR?)I-VFNEgM_PUdX&kk;yO@~)G!xWny7l{(r{OJPY8PbU1= z;>qeK!?q(cLObAHxV4aIY0dhGcEaEl*>saIIl7?oR}PnCtopw$w8NWh&IGZaR>M0s z+LeQNTXom6$~JUP*SOi4rnvd`Xa+Y~jtBfw)m4>>2M~%Te_P4nTgVwk>)Y~75er#+ zHXc!u4rop~Iu}S`n7mj(zec&qT6MYanaOcJRfd!@arQ)4ca4(F=22W=HKwmEb90qB+^Pz_8ZXq`#D!S z1wmT9`t!WMe1{E=FJmHg$?&++dVcwl|5V~!sqY-D_)vXLI#1%lK@9+Rprf*W$XcU1Ok<-?*@vFF4&(hg89`?*PDip({V^l9r=8L)Zd z8vV6Tz+O}rcctVHyY==yN68S|U(E5U@amwWiIq_v$@9O{e3>MG@4w67(t_`3j8eNU2tIx9C;%O*7gF9OaCe_ys&9oS$K6Q><+RP zo=uvQ!^N1v8CEXDVks%MXOTo5rqMf@bO0#)RydBF2~NZNr1EDM38zMXevoMJj72;-3QueTF%@vAdewvXLu)PB{BPu9+ghn??+)Oap8QK>3joU6GawVrgP8Qb+?ev4+E^?f~ml z1H5O9X5P_G!p-6+-;{gX2aN!QUX!*to@is=4wwp^fUZiO*t{JVLg`H2X+Xw8SzT4` ztHe|J)&Sfv{3y3?kC21kfjI2JfHe4iQ0}@AU8Bl<$rl<(*RR6Iy$xX5sNadUDew*T zh*u3eQ0Wa@ThsypVOGJ|8`7u1xtjW7DMav={wRz)d$1t z0HFNkiCj4Fj7gLGPY>+Ye?)>UG~~}aY7+`BaPh8yuWbqHZaO)9auTe8da21l9DV<- zrG3V%S&ZEEM0t@*fh2~jYx%B27KT>(3eTbtjz&N!hh|)$HN;}} zJCu7R$Qoh?hrOIKimD?*?;Tc;MOh?q#(wOT9!*GdVdvAs)v79No^8@Aq`UEgLd`*^ zDqCvy9V8t-e?E7wg+n?YKaleQw#iDj{Gvl{Dd!hCb@9hW`2I}UK3rZ{qWnbLjtn`- zeV;7R1gdbS*$uJTe}md40UJT!L`;{8?=QV1#Wd)4c!G-B=3()qRyZgmY(BOpDJUb4WeW{DV_%cT22)pStBf9I7 z|JCUNLT*OPSId)s1#`CD8){nzCa}gNl_n(^LmsuI+`@KRreXhwFy;!e<^m5N!9D8Em!pT5DYp-UIW z`yg|V2W=uIz+>WyKqSS@*6nkREvJWFO;f3gIA%ka#Y+cl#`fsoxL=1uH(P;-#y~wu z*R_r;W_SsJ@U?k`9#~^$Vm`0sdSXYVoOb2qL~CIyr{&lLc|K)5Fe7_r-qPxR=HMW# z-oF>#48+GbaXMMR+MNuxjZWi+-6qbDvelpOwEmxLPmkZ;BI2_Q%r;R&)1BH3@xkkc z`X`7~v~K1yURmkdnLZ&gF@czb?9=0Z;xL`_fNm)M>Ps(cf1$VrH$kt=Pqdi)n~DwS z_Nd28d{Ovj*E?CiUO2l-49PT{)LEl-MVR8g=*FqSS_D+|8ICr2B+f^cE29gzn2pIe zrX(%eybTCB{GuiiU4$5SCD~bz{1IK)?nri9Bt*2gg@41tIAPqTzO`4Ri0F1DYBt#+ z9oR_{C$8X=;9E%E*|G|2lj9v-RPc%Vhml>LYg7qtQuzBfb)Lo6aSV zmDDQvr@Ws~|_v8{m&;x?l=!dBb6Bc8@yH21J) z4(cVR_=0+VBcfe@LV^<7@5kvdfuL1m$l*U%0}c2gg{# z0D4P7V5mBeo5Se7vW|rPH6D%Q(tMJzn1vf$YAi14oQ0+U z>+OI8`1Hf9?{~jgrPbyF}5z#nsq8Sy5k9_moM%iCpa6k`q)`BOMIf zS_*p!xWf3niWTX>NF*l#IPBv5@_2Q3$Y*33bU?mZKj2}n1=J93@N(7H&Up`~Pqq&j z?Ko9(m7%M=8CAJ+h=;kB;X4spKbe5}FJrpEpNobo#WE-G-OJiatfzTX$fE8e1;O># zo=?}0mlr`@CLDd#kaD&a#tcxUu3n57Y}UlxKwyJ=5&3eF0HZ&4;LFGFfV!w-a~8U! zsJ_VChfoA7Y1Vgsy263_A$5>1*C0t5sONY+DmM)fZXuMG*dg@cK<3qp8nO!jPB%!C zAENbI;3iK{@6k1Dw~x-~5iS6NL6UsRDeqk(mSZ^5~*U=Q|lV;wrEaP_1D~Mg9CN~Y}FPJ>Hq_0sEygMp# zojwg3+|g35zNe&bc*jXZX|0NuWK1)4=O_a4_%!M{72R=>psJI34@~s8cOk}B8oj&p z*~siINqgo1lMqgJZkr&gjCF@PCxO3UnI??$mHwEDXm_FG{)o#VRMuaYblP}H?eDIR z(58zF23o%&Jg@aByGa-v4^Z6zwxCXL)3*oyy?Q1&we^kbS^#M&C}K^gwc`+eEfQ+2 z&^1U-QFYwFs#cqoqAe4tqv@ILH_cX{Dv{W|m8Cu;%${!c8}}6Vh9ZCJW>OOqe~)Ga zeVyACzQn+u3LMLG9GEPrA`1?p$qgX9CTAOxfxkh-A;h4!_}7+vLf`zu<$1e()&i*- zmWR$Uai)Cl-X;4diE-P!{S@JSb!Cm{05ajbW!Gxy9-rOz7MXG&I|@wfPERJYT8t?y ze_L|4F?rJn=d6@>Z7M#tiX94eo z%M4x27!A^=*?)ceQD^ICMdN19Qorb(`q}J|l-$0JNN1=*sSedzS>CGu&^@bpr%gFS z!$G6}{dd)n{O#E;ca4amh5=vRA#b zxlya4<$t7__uT$M6b7?P{O@pl7R>uMPjc(*Uyb9#2Q!BuaZy3d8`hvf2JJ+WOIYy+ zB}cUr-fnst&e$Iymi^XDLKDb5=#i2z26$59-x8^X{*kQfZ71gL9)F$D5*j30Of`97 z;hPEOhrbUP8i6GwOW;{v_tzWgoKaVXcA_}^gXn{4x{bQUR5`ol7b@RbupQb5mGmz?laofOW|@5nh-A-@d(1|}2l9?hNr%??Yaq^c8ioB{ zLXB21H@J}$Uz^Igdo#)T3_xnOoow%4`$~?bxZC(Hk&GXYl>TI{}Yd*=#cDR| zI0L)hiq#y#FQ>nxO~$?3T~XW zXL3CWdjBOlB%nOeYg{d~q~%yiLOeFEFI=;MC?tQke{Tmoh!SakCTa43y%`cU_P_!b zOV}r9{U#@vA}fPDZ52NuHSpKI^I%_flrb|OKt;i^@5|HMufbJ}s8=T-t^W+7Hi~q*}~i`UKEL}@y~jE@;qNq$pyFzoo$*-HMCt^ zYm!u7;+uv0{%2B|Eydi6$bMW=mj-5EmTR8a_3#b_t~?rk{QVlt zFeDsjm$2=3HBfl?YTCT2f+0S!0o|GH8rDAfaR3~kI>{W-xZ_yVUV{D<2f0IEV(4R8 z#cnFr0yOlh`Pqe|4^d0o8*KpJh_WV)Qofib<`{LYbYi~ja9^wJ>c(m@DREUdN2TGD zjG)vCZLqN?h}B=`|2hD0jyv7@p8If-%CY)=EdQX9{(h`on=9qJie+bA!YGe^K6boQ zS&C!lgsvT_I(Ihg*{IBim9;FmaKmfFp1HM|vcbEzgHrx#J422Ri&RX_IM#0?q)t>H zGbo`qE!=BFkc-DXWVL%ik4)&y=8YsmGCtFpmvwv3?~VpP3~zl`Mj5Nl`l=S{IU``> zq7^o+&Ngk{hwdg5_};lA0de4b_K7v(uZPi$)fVeXb6>!F_`7+7=d3M>o3BtiDWb=# z4rOiWDI$v7UQ65OpqmIF z_K}`t=GTh8nq!3z2Y>e&7Zz@}uzz!7zswvqgqkiWE$K+OJit{n?c2?uwRx(|iEZQ? zXm=ktC~q!NKFS<5yh-G5aE1XQPfQgTiXM>Ixsm;P^|Gwh2A&6>&<|6ma$Q@%^_Bf% zV3eKKLY-X12X!U$FW0Qu>~aAe6xh317~(@w!l-D@i?cpn>S8b7mtK7RFTpC85%Oee z@cR=imo`NpKb49FW(Xr|@m4b3(O3#Iv)MgRdsG>*a(iKlm0PY$&ai2gr2~6j*oWA4 z)w3W#?HEqX5BU#Is+w7B^5l^8P+9qW)Y=9Ohsfr}-IX? zcqe&Yj{}hG(Ol~c7+*JFgm{TMd3j`ob|Z1A2&R8*3rIG7^#njrf6NouME@v5H=-# zSNSS0#_8^Z0cyp>V8RmDK!PEe(IicHxG((|f`=8?Q$N9EJe9Up_<2%<3Vu`nkA#Tb zB&RjBhq88SC^-8xr$@qJq#>k!Du@f(e?4F8+qjuEedClGFd*}>88OQl3;hTqD@Xxy zmB3>a>v7@NLgMZI^St_r>SkmMZl`or6+F z-Bq4lJstAr6|F+xL?S#=5h$k9rX{K`!#S(j##&MR=uG?aYfOW@Nl&QAho-0HlE;>T zo7Z-2Tl_x?Hmh2u{Yd{LFP;Xj?j!|cn)-F6*1wPc`!U&B{0}T8`CRv1s`${tQ1fNO zTEJnfV)8?})Fu%(T+mnSsAC~ezHKw~MJxbF^Xmo7Mas(tLAufgUl^Gb<#=c}X_gBp z4`#!(F=Y6Q@@Y$aUHU!ES$|$PT#cFhzpy$b(Enz3e`8M{a5lp#L;9^+7+#$LD1Ei? z8H|zj=4jDeiyaMS!+r7_H)-{6kH0!yp~KBlkJoI-N2HX6_*l2I0>AyqUWwVr8a7`f z;)D5f_@BoO!yA}SDW#aIcig*IFQtBFQzDX1FI-v8v!;2r(G)0Ylol_>`rziBGfhI< zSrz%`DFVB9XF-CTOk2!fL8ik*(-^uaL<7=cypi1!W6;y)7$ty}@8vYWa_HP| z_j1_bOSg1d&&ehE4|+g+w!yCh+S35&Z`_o10SYnlcj>4F|Jl=cHG3Z0uSTi$mZXVu|w~yoR}OC2)%7k{9piE z1~Wcxusv0L-(M(iydWs%(oR1;PlP^@*3sjdq;OL* zl#+6>9;%hmRhSUuZ+$m=IZLKB<0`!SfRtj1mi+L61BXNwJFNB3!Fpd*6uCV1&&U<_ z;Q<^X6aiA!p$ro5U?VumQz+jJC|@EM=ccU1{5Bq_Z}v#E^o!5IJWk6ruP9iVUeQ3t z3OT_z*Se_x^Nyw+@}yU9a7f;hKupdGN;o_)&7m=0WT5VQ!D{ zABfx2gh^Hr_CKr_@>x(qh*6k-8uj+~+`!OTz~&3xKR3Zd)_O(#xbSxkK&7qedgIJN zTywj{%D1~AnS^jlkZ@(hdB}-$w=${l^TEFq)o)#8R*!=Z8XG%fk?kM-jg*Rc2B zX?;MdLAi&0Mbe3qvu?3J0ASqoof3n+6zf7%xxm?=y?bHB|;uoWZRpixg_;l#HG zPxp8)C|=$6bOY`8{Zhcu=~8=?q*S@QQgg~=JkMTNb_E+gue0zM#rKlta!Bl~@6g2# z${mL01SxfO<+GLmMU%i*Mbu0y#Nxsm7r@PN2G54Irbw0MP2FwQnlb_RB5CWTQ<(5B z`*d)#dR)H&tJn{%O+VCZ9=>%col)pr?`K}6^H0ycpg~hyLzYk)(2CLJ#=U0SWqyD; z-`StOh1UV=Jz^eA*to}G&p7z779O4jbP)+EAT{-ZCzHrvfVSWFANES6ngq&z1ks54kP07);8=j;&0`g=k*W>oFYw*i5Vv(Dn}LlHrc!@a zv%9BXq0f|9&iBcOQ4hnI>*-Y!Kl%Y^1|a6o;-dmSMDSgUb7#ObC9WyM+6;FzJ z!-+gD@1bG$d2(GYfxo0Ls@I6*yfp?M#I}q)KPI~zuW@2Nz6yT)OT#iqRv!H19%aZI z#d)t2iDxKw_axtCkKD+n@dseSGy`(~DEds}b3#vE|F6LRx&Z%Duk?|AcQk!Xmm$HDoQQwo*4pi1U)5Z_hr`&X(1j_jM`I1Fo7wVzP ztFwF>yU725Q2##rKPKv5!^)7M!%#` zO&p)g+@EanZXB7OoSbc{VI~yrRDAc}=57D~&3n@SGgZYv(hBW->3s??@NZqazq~cZ%(;c( zc@pm&=T}8)F(x^Ah3o%3`I2Z)rl?*^anhu@rLJz8XX=ZH&CYNp>FlpenYz%{rAWsA zp(Y8Qw4N229cYYX7>*9~c3=KWw5%rh4*~VFtQ0%yFpuQAUVLo!}{fj zOPOe$%Cdvjl+vnLS`!th>c3q`lkclb(o;e^=5CzXUZd%_j4BxkCCCY$ctm84L*|Cv z3FXV~zaAcb&niXwZT6lZ?n`OeF|-K(+cZq-X;neiOU=RJ(hmQl(MrK&3R1+aIxU;q zMOGUjVFPUWH+cX1`~+7nlnicvtCy8`9mlI+?2yMcaMQH`PQsGuamL0-S9O9nU!!)k z`j-sFG=?CX5Ax``cmp8uEI-fSc;rk)>_y}vsS~O=U(nx2#m=X*PR)!=d>R)n=^-kz z!>M|6EA8-3Idtt95zt&>HM@=J&DmZIW;AAWH5!*olcQIw9ahIy??5JLIO(o$n#rIM;1EeZ!ZVFwXFnpCw5i8hyZ(J4;MAPdR(X& z(oe4DlE-^;P=zimjC0gNo3QOiZtixT(N0)9A5?QL1JSQ(cfCvb91516&p zS&qn=Hh+K;q^_75HVyue`44pTTeNAQR6c;(Z0!M+hMfkybs)mY?ttz2=U6<)*ckHT z;V)h_7W@(Z~0>Z73bck`ne zD^@8U4Aw!_YJtaI+6!Th0t%YKeg2i-T(cXWLC2_iRN^M&Q(SgZL&6} zxE^n&2Hz`${&xtadoFk`TqJh{#`>Srqyb3nlM^hSc!A&T`fc=vz%{-=;Ig2%FoLpv(oF`oq5Tkl!wp=|U;U%bkA< zLwQxCk{&O_ymVDjU^prNl+awC9{VNzu0vw{VgLMclIE#L2jKF>1y|q|2CBFlNv!@2 z-N+o(1GF&&7%yMnxgmF&MyJ->$Xv(wyL#E%Afufbp^c&m`2SXKg4PRso>cvq0dkXn z5X(=)KhiXQm^0O|${ws-oy1Zu`6Opx{P^-G82f-DVe~C1!6{*gHUDeYBWKHb%TfTa zG}36T<`}PX+FrF+bs}VFio4M?9&}uO{Ee(vJjsf@3P&F;E?5#75YlUkY&XPq*=vvl z#u$nGx%hEqnymgU%jjv(`T+SZW%QDL=Ei9bqbDY2kHjBM5z9&|s=wR(YgSLNXW3)MkqqR9mOt-tgqDa7tK=r0*!1AGo3OVSTQ4#Ce~?8}Xn ztU>-3Rcb!_1q&!%G;F}lFm$VX?1greOzQHBen}r$X&9ef(R+0N+tlAwaXROI_J9=B21XmT}3u41nHHz6L!tFft#Z*4}Bk4J91 zLz>Bb#_4#AdVw-0YGVa`cUu=uyobFrL*p}p&HcKY?Cz)t(rwlLYi|3UHraVZ&D9#tL`37wvGB9wVZ_aR`7LbJvcWKGpd-K%(rgGnVJWH=Fj|c2=jCS4VucG!>+GEEh z^8C@5LrAF&7hQ^PPV$4@kiDJk4nX`a8c$yVnlZ^z2}K9N)wToj_wK>JnD#hxnYk04CE%QM26FXs#Tv< z;n$!4F*!;?h}M5n$2l{+sofFKLlyHY78^i%ZCuLhiJ$YlqCxi;?%v+YoA#J_*_ z(fz!TdHw^O;vy7AawfEwxTpxct;>-U?7~_dhrGIni34ukmP6Mi^5=U(?i*&R9XX+G ziH{69KO24@*o%c6KF-;=mjjc8k}&g^f%xi@U!Vvq9zH=y^BSW5r-jRuYJAtlt<$bH z@>B0YL+q1Um=5^p%m4~8?(>_kT@S%Xj1>OL02}M56k0XJkiz#(iD+Dy8WU~82OHzy0G zoC}Sjl3WQG)h2N}d{NlLhqD~Dw7V%}NGjf?O*wF(H5+xaZ^wAB{_)EFJFk{U_8TFM z-;KjO3!mifvNxsL9uBE_Y>#qU+Kzyp7IzN}4{+2bkl)dzqbD{4R)95YM7zOq?ZF<$ zOdD2DP3#=i23i8Su64$e&1f_K-jR4(Lok7O0s14XO6%cy zA>|FARsu71%-?2>b^?waoMz?iDU&eEE@#71H?j@Yro)7opJG5$ z|LlZoYlzm!F5b8U?AYHQ#YYmXAgS=;B^hw$)y&-WM2K=zGGDndSoGp3Wpesizjzu{ zphPfdTsN=c_Ob$*ZP4NuGH(ljL_7t9+9YS#xu2m{a=Rm!okoC*y>u+VEbe8FD>n2= z1AT|rUdi0SN*DVR{>s8Ks|phg4C?#9*MT20q$z+7mJY|NFMye;muw73*24uzLeoZ_y<-95MziWGM#?(SZ!Xs}=b zLeU~2xCO{f``z!ZyXFT!vPjN3^UR(-bLN>Dr8(@e$1k}40FALDQ!9+P7&-sgKUu@a zxyO#?elmHM(fYDq?^-gCU96?YEkwV^j;`W_VDnwDlk9=d#p#qxotso@Guvx+ZA_0w z&s~wZAKOF|l9K`FsJb_HJ-^SjQS=)hL-PvQGSGjVZ9V%Dj}i>Lmkm7fjVl$9&EZeLviW5C7DCp?% zzdnIr4WEd9Hz*29Cd%f!2hWQ4%)xf)RrEDUxG3e}wX#c(^XFWAX_OQ8XK_0DbT_j% zvq4KJe|{`lYRX}wV6sRB%$`~Jj#}l~VdVT0rSz|VUEm>g@I$>}8&i1ha84_@GTzQN zl}9RtdHm(@lV@NM*LtJ=0@q*n9{0k6LC3LbRr?BK=2CiIJ?^KZ_=(-I-(j!iY^!o< zTr!BM|LsI+w`YTZR`rQqg{~8(x&O#DN3V9wn0#{bNMLXdyReR%W3be4{}MeN$~0## zxriR%KB!@_VbOIvI3w=eV#78{u7n78b27@2(veaXu-o)wAN9J0F78Z$B6?~kJ4);> zS&iZyYT_BWMjnT18a5ODCyGR$>v@k|$aJ5P8rf+wT>PpJlOm2+ZQ1+Se?sP8>0Cp!voR)~d_m_DG^Z#&^1ieumYlE6;{Jrs>OjL20&;fGf zJ@TU;U!H#_|4uGI+@9&ay_`EG1zhZ@^#Y;s3kdkzT{1jAIIL~2*L5%B{$d`x`L&Ne zm6Y9LlS5dnTM}Mog(mU3TVq2QwZ&vX!r0=u%hvi=73(@w&Z@S6rRJ9fhjk)`S0cf^ zf8d-5ZvfnPa3U}`W9H9{?onUf#<{SHz=RblK^y;NZQJjO{Sqg6$6PDilts4ns%8R6 zCcfBWQ1$*3Zn>pVmBSTk|6|C_Pr7SNDq05zbp%VjlKlZ3gf=5r0$Zhv3n3fl)PA3d z)qV)^4Gem|=??TBhSN$}CbNmMPoo)I_I4f#{0W5q;(Qy-l!CQD zz`t7Fhq6F!zf<9lY8SR(Tq#Uiav^;ZvA~`lvf^%Kt+bB4CJ8{?4jM@nZN&L4(W7=> z2F@ZpRSUGC`h)_wqgCP|pNc*el{W>elb1{>$f2&gckX-KqXTL3jr{bed5leSzO)lY z+f!20z-}zXDTxFK`E!s%%nhyAX1n^xZ6tnRW;LT5cLN$8K@6rpPKMGfVj{4-1v*}O z1#!;sx9S~y)Ig0bBHP$|M0tM@Rxf8I)yB1+;K_w2^hI=m?dA0WV!_C0scW&O?{nfh z$|T$BrOx>c{sRhywm`Mr9}E^h7C*KwRn;H`+Q+v?6LtIDu2?w}Lh#f+LY5uY9o7^!5#-xoMjY!Jd{e0Ny2(n(X{Imb zx+?;ikQsy*Z$_kQVKuOtbo6S~_4+R*(6Z57TDF|Z&uY!<=2nVrrJeodu0_*C>|Ov5 zeGmO_gfq1Y;L`PGc~#$XuYI$Q z7ko5W+tb@Y#}j0}F>y@la<)<8&Wt+rcI54dH7Y_7YBa0C=iJ+5-C*du+vRV7OOlQj@k&;KRh`ASxV@`O#?>2#N_x!^3P<-Dgrhna zVukGYa@AT|Pm8%vyIZ|&^7pbDNuh4?{V9dBa=J57hAw5XC{&p(m5&`58$gR*6JU^Q zyNAJOB;6UF7sAaJw9yX7K*4ti`5ySOG{Arc*MXfP58j$89L%GMG|=Os1vFxX!my^ICsPF>zLBGxU4Dr@|AnMxvdM61d?%1-Ef_18*UU`? zkZ%QbsO@wg%&FgYz=2>Z<8Wgmf6$=_y}~3LrH%v9%xAYtiS?|Ys)<~bcKyr4u%ki7 zgNtHbd5~EM;vsQZGIX7%Yrt#8tQc4>4j~Jcqoz9vGDLObpQQU<^L#n4Dc7M&jA-!| zx&$saD602UM5u!xcFVN~+{&r~f3W@AV6I?iOf6qeVUS_;1eO20+UH`9a1T}8&%ZXm zn^o$_ql};lX?jMx{=rM=^EJF0dyTUcZvVP&u!t|gAKN6Xv@E;TPM5I!!a=;j!W=$) zAL{K7`}3{o_rl=3FOVzrqXxB6wKvxFH5h!b`GQ5bWe3?F%DATNKF_Yujv0dpj}UN< z;y|?eaium;q6shOOu~KJ(7Ifw{^@EJSgU)2L;PN>R{G)HueEva_=NWkMm?U(D&FOM z71J46qW!I39Sy*kr_Ba74fUBQhC%BR9I!SU0k)~lN;iq`wvzH#~K1^ zwZWt0bbM>>gFOv5Cy;!i1YB`kUop}~?TJO6i6rYXtV?W1az|%Jr}p+U^YPSYGrG20 zpaQHv*?*1>JP~Wxfvww_`1|BC4%f@a4Gk{$lLy!P=L5+(s;f~RDDJe1cgh= z6gB1{re2Tae8vkb2;B-s=r|^6&--V$d5kE{0oMyjm|yWCxleEnVx6HaukqT`Txhpm z;0RDHiRspGIlvkm=W7{+ysno+AVp@b2eNW$+>yqYrPPT(ikV89{dzaPCknlFRaY-< zzA}hH%3imA-3Km6z)kMUWs|l(Cld@M3uRK}8-`(t;T&E%z?dlXc7j28MN7R5424Z3jEnYK zuKJAxh_)W%;N5ba8hzJ8c78+J{*b*`m-^a|Ag|rqfuVf}jgjAV@9nWQY>@p^Y547p zSai=W*|l&rUQYJaP$3coIv!)5Qha$S>T@dsiJ+>6)FzU<&;H}algx9W+qxMUVnX1) z<9h{WvO?<{T!H!c1n*BOVdq2!iak1Ym>oFL5{kxiN&20T3pLwc%&^NcUVTtVJV4W> zHdxX#K(9ZM#EXlJjZhInKbCAXyDT|Ogxd59Qd$&hxdH3XtNMFml9PX51l4ZLPUK(B zvq3UR)vZWx z@2u;#XO-fS1!2?Gc}=hOv@jva;e@xE!9SQ|p`0|Bm1vdmm4`|BH&cU$-74>K(RsJ?JLy&6_d((eCeA@V-Eh~d(f9pOyfzu9mckqZVK{FxW4m!Zn-AHS z)r<F~t0=w@vvb^|i_?61GgOf5U^k#lW!{%py<@U0g$_a3QTy^!wd z=U#JG-m^bt`90ffN`2aJQE~h@r!kQFef-mIP>Q?5&N8xXS_YIKET)D1P3`=gK+{k2 z0Di8=xD)Uze8zV(VE(Q&neO-zItZQaTcvN1-98_B#PXs%c3L(LE^3Nz6`G|Q574rb z9YeYr1iiE0+D$QeK-U%++0ARfF$(OK-jispO_yK@EgVUq@~rfTVlK<4!O47<$!@K3 z?RzaV5#tfX2L<}%ncQB3;Zv#a;TtDJvV7#Cddqmfemys)Vo(*&@mto#d7<(0^^4dL z4bHNU@fr|A#0gRF$$}+%cmh#A%?8fQObDGCn{x&BFf4TB$E>|bx9^?z)XmC-^?rUG z_9NXjykp-_>g0qmoEVY3r(tX3HoK!TreW%reo)37NZ>~;!}<8N`EHn*3~a=Njo&V5 zYE>tf_&fU>m=8ZB3*C;adX?}*qTBt#$y*5M%cT{;JTLQBEYQpGlIYvR#Z~26!}%4i zu-Qz*fC1bwLb$i742Zzo!6fWLz7QhG(0*Di_iL24Ah6<0+ zSS*$e+de90Gfzzwa98yeBkOKq6k5plZPI&nk6a~TVV?ajRyx9wMt>%#4fDI!KfQyG z0D&`Bigw7oWF2tHL^(lFf3vE#4p;UkEr~7PQ%w3!3Z^yA*y+`|?H-0QXS&=+ulAT7 zb|3Xc4L)CT%r~jRT{K^Ovq&ZN4!3(My>g_nLT|Ebb2W=kj}FzHKAx-FpX8E-hpX|L zW~SZ#hbA+k_kM&v?%iAN(9G=uo&E|^{aGHcXvmOZ^RoujJjUAFC>$V z)UHnZOt)-D(QyK`!6u`}?&2TEpBN+@8;mTvMudN=x(CH%WZ@|U8SgD6^%QJC{p2Lv z(Hv17RcgfSAHsgGt#m}qZ#Cag@t0#uNreTg0P0vxICRe9Wd(m8{3r{Y7RQ1qZe;Rm zV}pHL-5sq1eW3F~cHIj%{my-m2!t4;yXxu#A`8In*y$DN71>JZv#mVRrkk)T4=u(; zM{Ga6f5?;My4bZ=0b{R3E%$x;bElK;O}WXt=8JN=MgMSm9svBpZIH03ldb_!1E^t1 zS@9;@qSVJ|)m1^rYfW;FZ>g)f8to=U*gZs-dd|s%e&qOni(2I*4W8xZ-cLRZ^TzJa z|A}B%)#}p8G-`NF^Q%Gf>XeGby~x3BY^z@fp@cBW_N_U&a#H)EJv$cRS#U9v3o4b! z@1VVnxI+Qlp5;Re;Q5e^wq+;b_H8N;+nHUf(Zy4Qb2i{~aR5j}6dE%U6EW_%N8}8+ zHiu8~FbonIBZ92%B%Ki8J$UESYDBhRqP=%5p)i5z(pqhpV+svD(tjqDOQc4&Qg1x{ zSB0-;t7h4CE>q}HHy5Sg;Cm2WS%ct(_(V@ruk*ZpmX>PVX{rUVIoaT-f23^pJX;(V zm3(@18I%Y4-M8jN!iq5{@yl_i>A0S zN!r^nOS=kvvm{2wp}se}y6RR%;>B2{IeeJ|nv)BY4{vnI^^f#(C3>$n_|`L?@8Qb@ zL5$D!b?Qe9ZglOt^`~329;JKuz8EJ9gu$_9Nau;>7XsYZurr-^@u*Qpvri^n0%ss5 zFs(MLJ@I=I=^i@&5nYJkiEIxPPqc(ezq`ZGP1?M0p|?I@dSN1pNcwzzh{D2og?^~a z5m4b=;|K*3$@4e#jE15yVSabx|SYJPkts|U~+SY zs+TEN#9}{PJU+?Y#S4@3vA5FaXywv+NTBndJ1n#~wI6dML8Mh8ll_-bsfem46}pGd z z@M!ZF&5(AaL%wEHeA&kJmr&4AHBT$Qhq#^?JwF z*#>cKAQgSeZ@MJ8`nV8vGOBA~LXbbw@|ecC^zEYg&LeATA|)Riaz+tCc`~B=?1JPd z?+YAOFBdLYB{RGRdsvrXQF&pfMyd|x==}jpYhaNV3wjG0DZdj7zt)vasGrj21~$sE z*w6^5YK?Am_nsOtMz*6w_fZADw&lz)k5ip2F`^u8jDy)X}{UNf){d&D=2K2 zU+K_e=d0+|IISbva}k>U8=#6|I2>}1%dyJ=s;TSvDQb}c!ypg3IN zp^rYM7GZYcp47cxkESJO@KRie9$LCY=uH)V_V$(*woC4sHK~IQKC~%D03#=kR3atn z(YX?jJa^fgZGB$l^Jz|le{mz!)F5g%YJuoF015FUGofkMU4{~7tyVBjFiE0d^`r00+>FkL>+F6SOos9zCO8&~8n0FR zf|ct%ajIN&PZdv<7X9m&$T-#@OH~~8bEl1whnS&ppzjr^u%7|z9ZuGb6QK1>sC=>E zK8)?9scZwO?2-poz^U1ba7Q9qN(oYF8tD;Bp$a|haf4`~@BF4?^ z;B!Dzp?01AEXdezBo4fP(zgCF*fURHx$*g=|CB#+ew3powHB`w7xE#DZaaNYIR6RE z`>@e#pZKB~v4L#Y1VnfNK*;Rza_{E12h=2H#wYpOU9rAqdfeJJG05Rp6@!o5wO)P# z!~;4*Jt7ccP+)@*8%ej8mz`9;cOCHSQJ4364vN3@;&i7Pf%)+(!pjtUJM6~hoiKZ+cOBkPeK?-?rqW)RBJ4ic zw#l_b4|9Pj_;Ven$>A35GJPN5U^hKa?@%9NANBZ!VyIarZo>lkD9*#^X<}4R4iPt; z+9oVUReLKrME%0)V8X8Pe!EQxpYI)2d}`;$EeprPX5>{uL6-t&Vf1O=xUOy-^ZU}5 z-A}E^>s8wLo2G+2dtu-E9Z#SI*L$tXF;6&^IhB^isXRfm6B=StBKg6`U;1*PDLZn} zMBlk|$!oIDZ&Kl=agX-s+pt7n3POPADRjVQ=FdJr=-@M_MvYD!Hfg2mzTf~7yu6Vu z0G+H=sUmF)bP^&0MpG ztD^*$!S-4jEhyTA!^D?~H|;A2Rk<&^7l~IG9=|8O~Sc=Dyk+cZyUA&xnqT6hBlu?(7Y%H%i7NfIDE*pzc3-qoXu+UcA&EG@u5lY?gALTJy!m z2`0(Ei6-K$?KG!a24{h;mRh~8=9{~R^SJSgEQ5{}QLrfZCZHvtzmg;HkqAvKKRqwA z_=`3#d+VwDe#KSLw{Y1$Zno9{p72^iPaHm}1`AnL(i1F#qxV#5?@6qu+$S71R~f?Vi==JLZF&7}_5;X<*%0%ePKfxGWdPfmK#DP!lfezO$8c z=kd{4p=&>Q&ML4nKvZZ+F4 z|IFR7lbZ%TJ-#P;#7KW{NEdL0vcoKSKH>jAz6i-h3>-%R!nF&W2sr&KX&A!6K1F4$W^eMqjFi?`7-u%?LKSbAUTJPGN)@9NVQ|BZpVYRKHp7Mwm?Ry_ZqKCXqnR1#}C z?}YuBSE`!PO@7X_UaTI!`J1n2Pix0<)PZtAERP51&s?eKo~NTD{CFkL-}e4#^`f?M za;&Fc5%0YaaC@nl&(?X%#{wIr@*<0U2nO(;ITGTFk4aGaHJ50t+Sb2AmTdLJOBG|ql2)VUUhdnrm}}=&81~m-!f*d$3DRVE z^F{zc6}P+iH>M{Av0mGCtJoN~iV=DmB5ZT==j z`m8bhdQ{3l@tSk?czL-&R%5)4C($C?xDe#@7g_@mm5{Gd%mkGFAN_Qri3a}68cA!{ zMMg)LnGE;rksS@UYKvrWQa5+5uvZ4$Nyfi-Up%D}o$jBaCU*$s?iP{U%r>O#PWl~t zB15DJCh7!XNlw)zF7Wi~+WBYrS*eY2S};@}B%4?S1$o}R=+*LEA^H$2VN%D|Mf@#W zT`m(GykcHy`!(d5s2|K&lIWGAwCb?4ujic=U)MEf69#S#i}kPbIlk4@<4Tor`I&t| z!$q;cAgHg7tLG2#@Nn3{__HCGi5y~q+96pl;V@=s7U&tXJg3NQP%6(eYeT!*WFUFL zxl?tw*^S6qy#29Vw3z=BYcdsX%|z=7_~8{YrF>_#@fMu?_v@d!N652|Yn<`$?10YA zOk_X6visk#a>#{Eywfgm&OUz%%xV1P@!|yF)dHVa+r+AS^mW|zC9pUj4xZBQI_d=3 z&O_gem@T91zFn-y6eU~~&$1Uf%R=4F-XQM|<16i1TAU9oGe*<*C`a9G+^Q1uFTJW-iN|D;}OKYv6~V`Tg5t z1?!F9dSjK}$7y$0QX5T>>vG%Uo$h>Du1c_?DC=iEZaZ#khb@^4XFXc@gd<7fCDP5K zG;uevo{t++SYwuA7P2G%k8cO)YbK$ci4;vFGyu}+Kbp|-ZFOO)Rp{?p?$Y+Z4)b@u z`Cwb8{7so!u$`!nz?eix4imzJzq)8G#?q0x7E`wK`R~^R8E-6i->-d0EK-{;3<)${ zD{M$*-aw!D_*cdfi0Q+A-K)xM$As86y7mfRPVMrhOm{C^o@DR)$4&5qJK)*-a-{h7 z)R33z`K281bH87A7dX~t7aBk9-M#aAuJuxUb1A;);-obyaWg9LhcO?IQ~q)Lm>Th_ z7<5~20A)4blNt2>$xlKnL9<9J713^ZIG>HYI;(b{TA-u1>0W27lH^~{@j&q7>)tz(3GL&H zSLlp3T5523gA4n~;+`lrlelyM{#Ns>Tv8LlRrpO`B{un!T!@7NzvM&0UDItIlo%r1 zX?{*VmnV5{I#yHi8`Er$e9#_N{_G~#d4tG*ifK5)2@@NW^1DMzZC0HBG4&|XnehGiCB~~Ugll$ie<1@4z_nuUw}8Nw zh$8!Z@AizB?lgBh>3q-4yUeiboPUshw5#c8yrc|*&POq3Pxq`;elhQ}bCgn9BjBJk z^--f(^+RPr<_-41-QYf!Q2<6nN}Xj-6$g@B9~JQ zC$$Dt0KQ;xl-~~a2gbX24KN<9|E@~|&NuoXCK21SRLtbC8ZUcvCX69KZU6#wm&P-J zbA%!EFidulig4Wpm@Yl6@N!`Vat#V9*B|G<3hWR((4U|#GTVuGZaGp>u#=pW8LJgG zVv)U8EBP{|F10Ro4;k$IJ143CdrJR_d@u1xdxzRo&}fv3N=niWey9zcc~S4shLrRN zobPnD$R@HN+FyXmehygT`*iT#R3+m+xpa=Trox<}j*^@up1>$9YA~?%$?YoW!St=% zbSP~P(c_(89*EA#^+rg{@xevzGP8N92U64H>1uy(57dA}iTSa zgY=WaQx+W8b!pygo2~t=ctdidIZ6`fRbf^V;;ocq9&j^Iub49+u<`yWs~lwOsveLu zPPyNIEK*c@tdG&hy<*`%k3x=@+wSX(%K<*WxRTj$>0SQL{$YMTjs3ZLiRyKSsm~XG zUHER*Rcy;gN7ZB8rmAoHjrP1b*4aPN)*J1OWM z9vpm5oWJyu94HIRh{%hG3o={|@Z$64ESnt2uv${HM7hyVBB1G}+C{ql3bP z^RPE_ysNzL904v1J+a?AHp^N9=^e%JxY&SB*CmV(QOu(jj|^r`4!b*cbt3{8D<`vF zp)oy2SCgVN%b4E+T5{4Q+%5MthzI76U4>0@gZ5+DCBsvq#-`DioO;s)#arah9p;Hy z0;(Ufs(boNnxj@vuHnu=`W5=dcMpw{Vkdw$zi`6fLI;5RUAef~&34lV60^H4e>A_e zo+-a^l5zoFYsz2I^juOi@0VcVClTdcX2aY5J^CBR_H+)jKO+W4nb#*!Fyl`Jn836`Z)jQG%wdf~1TG?&ERbye1R z^s&m9J@~?4vh&FW#cR3m5<`*0k;9KTMeae8<^gX`g$^lq8;B!I2}Olwh%Q$yt1gL{ zRR3ZbGVv5|_~CEtkHa`ey5BSHpX)4LJKM3iYlwn<2-Bgb;`V+1O8laRS{Jmk>i92 z3xlVAUt6Ip#Yy%{D&a;Wi#Dj*moKMUe9&g|) zTr21py1YCZPh^3cc*T%js#zH@U^Zc}i7yR3LU7z^1uuLwiPYPVPYN6X4&z z!*5N#1RMT4jAHhfkF$wr>keJ{_pkVtm6~%MlX00}tDcBl6d@yZj|w75HF|}q>ISZT z(?4IR0+5rjZ{x-r2Hk%f1k`(~6KG$B^i$ov?j9Y~L6_sI5q^g2#MUEbxaVtrYfew6%NSC)0^R4#B~8LX)3pl$%XCw`p(t6aUo zRSn>OTe%Y*FCY>u^Uj)xntO#n_Wl(76%vsCc98ByJnhlb^dXlX9Dj(Sr}HI_2q*Oc zx&K>d)lmsrYLVx3Dgnz|-*-WGVU% zQTlssTgcX)J;r78c5-Y7EttS*$tJPBNZ53)BpX8Jx2uT~e&lA zDHWtpth@@pWoDv@pGa57t`^|^BJtPiix}+s$gZ6%s%-Le(oV_ zNP~fm`LEK##7MU0a~vA13iVm7(LhVQwLiSd_9)0&@?G#%B0pCz;ECZ^bOl0D+ zWUuMCqE57R1`J*ji|zC6+-ci~(8Di~I}U-pocI2dMDFeK@^hu#IN?cu#TJP(tIHL*D1YtS|NH?hdx8P`xDd#fCXA`+!pE=}q`>jqdB3VE4HA>HQ6nBih zdGjwZ^Qq5DCByc{-uIMpyGbzldCWa20h`B@1vLfv-)tE>)Ih}hYo$x|R0DILNrZaJ zp~;r@jODE5tlsaBkw=P7SRsgpU`@oTGaW_H@28~%tH_1aZd^tQy8lUw;xJ0!k^-uA zd)M52w%|ho#jg#4fc3M|gq?p!bs%>+|EK1ih1ulHyOWn6Kd_F@tt$jz^_*PhL7-SR zPPTbAg#QTFgL9=NQsiCCPtWqo0Kv#>1!Z!3iZt3vT#k$?+vH5&9u5kH>mFrv;C2;3)=vNK1pCFJ7%o zFqZUhbH%{jQ4XY-Uj5K2S%P5zekN4U)OiM#>|>r%{&r{7;2uK;z|r0h@FXx&sChXi zc&4jFNc1#z^b3^|sWMh2TX=O>4YKre86k%cq~}6t^(KvP!*5dp&e3;tvez^G>s~E$ zHHlkp=L5jsd6`?}4h4S63kE=bbhiYzBzs~RQhObphCP{fO9bzRQ0ux8 zd7mU*Zdd9*$N8KH=h4}ZcJ?;vjQz)8maww~u;arRg|i68dmw}sF@xv8DQNE?QVMw* z9k;8(ZDau`_izSv8?J5MI=__bhjMRS<90LzS68Xv81A-&XcyzB0Zy`S6KjQcC-e_6=B28g;j>k))N8(V>^ltw|LM>TXOViTA`V9HV3FUi0 zIFXX$#YI|X$V|{p@U-|%m|Y)IVs;d9&JON&uz!J;7H&Win-XBoKE>8caQvuIXq$M= zvlPFsbJh}#5lZhcXh8NxB#Vp5K^o)@?alV~^8twDBcaW$)5M#*xB05J&rn}K-rqK( zmsHQmW@3#nY0zniMo&_A@ViO)q@CVwT(MG|P`#ymHJ_GXOsB#97PVQm+3E`4#Q>&1 zR|4##-#<%bHj?f_6}G(;gvR&fxJVvgyWqMwZ2F+&Sc&YlB`EYg(sx6@3N%xyQR|+K z?hb~%?)~=kqVe{Y*-8Sv(xP&<^4H}OjC+lfX6N*D^HTs^ve~Q0F((6cR>r&+ zb+vc>`o%oES3oGGSXKLtmon<(^FV+R$fNyt{y@Jxjw7`f*TWo`o6J|Z(!zn{3$G<| zm@D?U(wn{{I&fd;ER&4QSv~Ca4&jR}1U1lSM5czXMXbs@=9IYjwAE_po(+{I&MCpwy-2BC7`jnVyx@o$CwQ&iq8G05wU0dB}BDp3Wwph(V^KMef@@lIUlf5s5QG~e1ULrx+diMm`gns%!?daT^wlY6_kdF{3 zA%RInMV(u|MApbZmGp-mV_03~FzIQwwg~M!vp1e)P zywMYr#fa44&Hdef`yOjg+#5BI+<)Yb6#vM47%L`GYYd*X8FxO!_>TFis7Nq0xolf) z6x&qD^mcADt<~_dGo$p~TOKXGK*u~>CvwAFBdRd<-;YD@sN(M8U@n-|o-f>M*ro11 zzOzt5bTu?;W)5Q%2wHAf3WL}{HdNrng?om_h&6K)^^+1@asjCz8c$Es$}!l? z>1QP?%=+{Bf5g#uu+5jDAf4_-XTgMja6CB%-1D% z{XLRn8=~Pw@}AE^FXFW%E8&l`7pUi#YxVsMpq^Ayf?k`>cH(MJ&Oi<$lxXXu%Pu51 zo|%1(PC&_^b$uhoF1C>jk_2<}xI%{9=VN5fGKcv{DW4JKPz7j};j!$e$}32YWg&60 z@1usXhOq@v>VyuLN#Xg$*94UH3xsoV&lh=|Ggb|G+;Eg4kmmB`@C9lc!R8m(^5}0W zig;?jho%1Y#SFoBc08$qp8SZ@fX!M|7Sfbg3fDK}%$6?W4J7XmMpbJdCPNKXmfpj) z^IdHS8RsSN$aCrF_gTOGd;61IxvzV!fs?`uys)4eB~E>fFPbR+C^)ftXgpF7V;a}w z;VZ%O?>ds`5q5L@2WD|1bsjJ0NXPu)sq$`re+&~!)z-IgNXD{~daUsAFY5ngEC?)0 z2J(FVqH*2HTacno^Vah4Z+wbrs(b40`EglT+yRXRUTJ9Q-kuo{`{Xc8U6!?W2d!uv zSxj;DINVTWamSFcoDkzP~T0dG02#TO;-N{m))%gpxo zn?P;t5GLihh#jYcb9}3iitm?I$W?r&vs~YtU+a2%DYDk(c(b>uHiT@8r9Kw>e^>bS z?+Tk!dRgHb#{;Yb4RM2JJca@R_Bfxi=>wTc4XXJv8Z!?w;5{C)_5>x(Nkrd7G+5`C zxQ%A-^R}bCFDh=}$eA>4aY#?@T;-ZO0zXQ?Mkqe9_D=o2wr+EPWiTO28<0HyhD9TV zb)00z$4(4%^M?N4@knm^b;z2U^1~KZ#6653bA_3lToGRJsMzLLTc)WV6FfVhIlBNWF%ZWVUUbnVij<2y zi+q4_tNfe|Px@MA35Pe$+4uED7wP?=taZ9Nfj-TEuUcJ)$_APj8WZb`AN|d_ti|hc zQXZncH9CS)X-HR*GnCjI+0d4vjOe}GfBX-T`=6CYCq$o@Qc{h@`Q)}Qq>Pe@59KLQ z+IHZ7Pk+gmfzHsVZBEz3IduJr(gr!G+`+D~g@$RET7wf3;@e?~9TlRWWHF6cew8p2 zzEyy$(J_Z}%>Bnv)2`lq;Xo^@-z>(d`;gdYp$Vb9PD;adz*lh|3PuHY9S)jORIU_x z+{0I?1Np3c92hbdO1n!-k*Z(Stf)*uDT=MAv}O2j&(}+~YTI(ujZm?l*gC_U<@#yc z=9oWhAueP(fYSh0Hjf0PjM{9@MN8Pm6S*F4>GQl z`Mv}{yu5KSa$r4@ZGTYou$%Mop&~Y!(Ihh`V?N@{9@w$|+62H%V@;FgM;* zZ-=0tnc=u!v!!5le0EvEU{W3O^!tKlO*^)dfgQ6-Gb;S=zORaY2!?LUkNzZdA9y>> zGp{DdsdD+4(a_-ApW-lXUikIt`Uj zR*-qC{(y6+{o6qQUy_U9jlFlv9JBH2NpGCu-wMlv{!d344?7+2={X1h95+bj19A@gg=unfL|qzT-Wabl8i$jIKe; zi+OQYEru)dWce-%HUbQVo447=>e4?`CQ;-V*R zlLtJQ4(bFK1gS`h2B6}5*UFReO4+znVbkmGj$sA-qcVFYnKu%&JhTTEm*cPRy<&L)cW@SSPh+<}&kl(>OsM7~s@bc^l*<{^E` z{ln$)S{k7g+Ftt6a}mC3-GGFB=e8}lja{PbRH5A2F(d3!}B3q0wWTEgRr8_LDrC<+%^TSc%C@%&0yHW`2B{vtL z?ISKLqCTEUbXm3R9?K;!gamn0G@$8X05+syzW)9Yv7(oWMgwTutyH?S|8b6qaPCnM z$wRC8ockjO^GcwEU=Wq%#sp6jiv}g*ry0W;tfKEwu|y(QDq>ZE*_4-*R4SqL2QRr{ z{Rhl1!vvhB3iZcKgx_)}{p>%t_dc2JMN2ja9Hmob9XU#k6;Eojg^Z1A0{PKi;r{ng z))8ftAU>0JBs6ZkenE}hvE$F53=?z3kb{u10nM#D(i#G(gMx)1de?6fBYq>#8|k~* zOQ7(Bx;{}Ag{72jR>cs`=6gQhTeZKKzfi7rg;ZR=`RW%vmG?f?L-7KRp#}m!2Q?Ib z4xY?X#+gW~ds#GkzFNkzO<*r#lqnzF&r;gF8L~|OP3Du}eO!Ei+dArzsud_zRleFl zD-@Rom%>0t!SS~(`G_JSK~F)xMX$r^TtYuC9tmON%IO(GiwEf#d9|a^=6P|l__&88Wpib-!6EI>+c5Va{_LB=&Bs9b@@VmCt4-Up)Qc74g#@9ByFf_Dp}kItQrLKdGW=Zmjf(( zZ^;P@o=QKK&LZ7Q=iDsFjAeP5Vv&NXI5k!!=1!pXRYq;5W@$IKQOMc{+hWbYAXjz$ zVh+Wx>5tnQM>piI$8M{RO;!KTScV|%{6PzC`~nWX3=C0}_T92*mjZxgIyJ zI5k$WJl@qJIVJ)j1Y_&@8<=5O1xkd7^2-JJ0q>-(z^Y*+>pf}(NXb~dwb6KC0ZnE7OMiE8mod;mI6Tb ze!;`5NrYfOBrw=080w}gXr?!!rNhL%wN0YKQeoamCtWx6B3%^PL+pGdY?tI4?yIqS zeX(_=@m!q6^*{5tW-sRO(nJxX)_*xTWL1H(s?Lb=yVNL+JNg_|F@n3FJ>)A6QK7p5 zZ&#bq@71@1jLNqm{f-H>>?v2k&XK3*Yj?r{_6i(^3D2$+T)#g6j(5n8=%bnFw>9l0 z+re+oDoFxnL|lA zQBW)q%y7%l&jJJhWRI|1a*l1he282_`*3YZtx!zI%4VEP#nS{lqOhN4Ka2o358LZY zv$zs-s}X_Ody<<%SC5xV5Zb6w8q=e${aI*e)kW0WU;4eF9G+W}2(?;n-hXIVevw$t zE74ku0o?G~1`!nMZcKvlhZ(Y~;5h5P^>sz;g_YQqqY6QU@=rsf%3X5)>R|eS!H3mb z!3bzooANRg@2am3Qtu^MPi@tP?p^purtEln5|AF5Wj+~_I`F`Ry@W}xkl{*d6O z)DgpVv_WhLeYJ8F7?%*1@o%$L6LuPZW z7~UsYja@OXR5WSOziZGn(E7SSO;$1&_}b-Z(JK*Js`_4)oLK_jF9)SZk@L}WDzfhAiYcLgH8k@}E;ys3`@$H~Y==%BfCD!x&aJPkO^_+#pub~BmXKmEaZA@J@AjHm%9 z_fC2KHa{LTq1TC0NuF?I1EGmL=>s%ag(si=iv!N;mZ9zz9>)wkIiS_5EU95o1Vgi@bxf)AXeVoay&+-u3g^h8*4t4`QEz(}o z^}1^wGqDG}EvmB3H?><%Eh{#NJp|oi8u*zasYi8ng01blBo65d+g0~GBRdiuGvLPS zto1|-1X{GHT5^^c+(C>N8N5RB07$Bl6&U1IGB5&6w8N3lrh^hUy=#4zv6xMlL6c9k z2Er~$pVO|uE)^>kER_e=*{tKl7tvbKu7`+_9Q&rcWpCGD%#}x;QzXF^401L zTK-GZr6{c?&W>vkgwRGhbwfsX?|t8_v_$ZteZB;Wm<80y-Y(25tCOw|MJ?gD|5HKV zj&gwhVM`VH*$O32;3dC6^Sc}>x5LAgD7Eh~Dapuii=4B<+hbYYV43$u`bd!vVyFVG z*ipYii*gkN*D;e>}R1mP|QkDxv<5RHZ$88D|SvDpXSKFL7}k zl|Om}!P3Dyiz=~>4+=W;#_^|VlK`Mkb}pC&OmM;IoEwh*4^?j&)YkWg?c!3jxJ!!^ zcXw!U3Pp=Uad&qp(xRn6acC*QgKKbi_W;4&A$Z`VzyJH5GiUZpJ|#1I_FgNGtoy#6 zsh=G-kxOGYYzlGse2hcsMug=#{%qa+?mF%NGxoHwU`SsS6PZN3&xTTWx0f95`j&C$8}7&ce&kyGpuh2Q~Dr$~)T z1i&MdzO7)A_Mmp(9W1_dkIJJqtaQ|*5eLGvKHo&Cl~|{|*|$rGe-0Tv6^&4pR^!^@ zb`XnY{Y1w9&ni@JNu_5u9d`6xNO;7p?*c{xvr79x16nTHezl~mQ7A1U zfmjZ^d3*}c&@!)FPbCxN`#)_wg9!Hu3eMHfc(I&6i7d!*ZAztTm>|(Q0$k3ChIm<( zCKZ&GVIaRZ>f~IqyWsJ+0u}k@U)FSP%JR)+*TityexL8Ep5Vy4vI9;3ULvN(Cg+eO zFLcCCtIU0}OSzHke`~Ga=)hgUW6Wn~($d}VYg)z*1@HvG1?2sUIK>t*au!uX&UdSm zFj$`w#z))SyMTwY#ILa>dgVGB1}~FCB^KsALu?~kpq}N&UsRh$8!LC~#oKA?Bu;uC zoW33s*A^|J%A~%M*SZGvyTL-zkI!@N==8;^Xb=--w4a=jD3 z$DNg}xm&_B*}1i_nPXi!CZ7S1%*}j>!tr}fA2a+-i`}|aJJ6Q#pqpTk+>LK~(IfQN z2XwkXtvf(WYBKG`0$Bw6>evXKS3b+25fT*>8GjU4zIPYW)}?*^eAQFYB{t$C8`=jc z0@=zwimg_&(Y~>~lS%1t(LNEVz%90aAdkHtziZ*`79>~oW5CNeG|S?-tCbLFgELN< z{93lZ*E1pKeAoT=!{?4B3cJd*mbB7&M7y0g-J33qMygpafCuJiGIh$UqKezTqu7I& zkTx6OD(JCFusp=mdM2{jwbx%Oh?(6Y3P?K;qShUR`8A+K)7lZKorc9(7II?=~+yl1OTDSO67+fO^iCxX9D1GTh2Z@~E2=Z zmxKXs(SIrDCZQMq_37qU%=P4jo6Hs_qS34a!z||z31>6J+nk5ir0U8UBPUiW92^6J z7sAizW6({$2jcV0ebleP8 z@Pe?+eWw2KG7#T;iNAAeZ9Ld*eT18zwyi-s$#H5uuBKsUfxHg!nv@L(KUzHjMo&FK`wxXrg>?J2S zQyI(tXHB}y`CpYmy?Kp%Skc_SW=QLpsUa=GnQtr%ug%znk~tjXjnwAW&x3R;Lem~X zwH0*$0|x7B{hobnnIAt+;<4y$#2O`AWX>BBVcPad3(F)arZJsr2`{owT*Pl!Q`ji6 z13sQfHieL{1p0bhon)_@Otj@1dJHAr1>B8$JkATeYQJ?3Y&`YvjG;s~w<%GPGOa*{ zeI6I3>yqjd^TK-XrP`E5hMUH4w2Em zI&oGWi;^vtd#BpJUay+18g+e{^Jhp?E6Hs^5``}q5~)ydP%c_j1)|gU{`_$SYSy%R zwq|@$+TtXQ7Gxu%XYR{tv&SJ=PbNHg zzb#@fx0In_2^_=mY{Y2H0zS7ZBd8V@f>O)PB(e|h(fgwSeIxi=flE__Q8^_YEoq3K z6NqtCPe1|p&(WP``Mw7CO!WZc%t!povGJ8RGYIJ)#hw>VTNohKLCjYXQ)5(?XqOLn zD-*s*yH9J?T4h;gTYt*_9K8~N-t#w?b;gYcbci=$8PKw=7W+C(Qt1X&K2UYf57Z|( zAVll|{cB2SjKm1VZ-&IV3(isZd72EWzoV|1(!|c*(4J+H zciZ>De|dKbeJ%q^`Cr0jXw2@HTGsr@vKy9F1*oFDe_I)SH$?p zkETqQG}1vZXhY6ykYR2J=-HLVO(o&qGO4A`j8V(!M(MqyMVE`@H9ZqttlIJn>Yy`6y`mW`gtrtje+qY$6R@0!>HN zsMA|IJKir!PX3v{Ub~_R*Ln?370k$2=&>&$p6G1K6I@u!z+~;;%}fqr$DV+g&D+Rr z3$y-L_DJF++VllFJt5|<#J7Ek3oFg>!1g*vcCvx~QZDSEG4(6;ogw%xb*Z(a_BB41 z^Cr$2$44wOTO?`n^bq#u8#*Zn(TMb&;$h*YIke&frH#t7U~zODCyN#&9_U2^yHdm)Chma>n%S7zDHqL$x-gFiR~X`r#1vT$^1ugv!G9~FjyI1@;zSwq!$ zdIWm%>x6mqEfV^Iq0gK9x0L=zFsarjc_FnL1B&kY`rk*I%st+Xls^d?x@s=>(ECcw zY|bc)2)Z}yHkzh(my_m(#+1Ow6t_cS{gN0IOS`b)XMQ4;^ZPXooa|xp?`0GKoVXVq zc+>hNB-ZFL7hzCp^T)31}FA++Cb>GjBbqi#`J8Gudty9ba+N(~XVwog)KReO{8B;!@-C z?K5D49H(NABZ7|I#&POy9$%ALp7@d6NE~*awQg>=vyyybe8L;C_i+Gv<6+i6)8-vb z)AK-OBWh}k5@oThaq`)=rZyeuT3>g|wcGLTg(-R*QXB}zqEcBh;^+TLrO$VI(SrR6 ziy)ietcM@6N{-7PIue$chLVR?u=doOjrzug`6+3U^<<8B_Oc#2@M|2MlJSATSTQS) zD8pc`Nzm!}T)G4J{_FAfgqFpoi)2<*bViF7n>g;O?~3*2WZ6o4sMW#IL$G| zKDC!cC+P3Um%X*hhPQfFE)#Kuol^Z01>iqY<1hF3c)`A&{3k(Bu%E|v3zA@>r;ex4 z41m;bKScP-HGN~|**TW3SX1J5LPRv8=Chl2IX9o{*==-m_qR+RWi*MX(YaN5kg#+_ zdn6y}JMMy52mz-Z1ps zy6R;s1w3v}_L31P9sShbeLdS&kLMMt1@w7)hNleggTScjc-eWrFg-OnIh3t7(y?o; z9NdADp4zIvyTwyU{ZCk1v{QGBOi&sAJMX5n2Jdp=MsVr#%QJNQ-Jzc}go5e~PDgP_ zB;O5Q4}liL3oT`H_*1jClt=IpM@&=d;>0xBpD2a~e$vQ+0t>x#XIaIcO#w`_inD+R zeWp;qQm9H)7+B6QUKICoVIiht17fn9*O%wC2wjxt|KT+%J2xzn!4=V*;oF z>w9{2g28xhG&771;ErEHThboobGToqwSsHJ!-^A>L`Q6W#hIC6;k}o}#4#66UEBh#j;8U zS7g4KWRR{WM)77+@QCHtmUrJy`5l?7QOXUo%<%riEe#Q%KdM8+7h@~&yX11M8iIame31U(fhZOv_u~K9 z6V)F?c)Pi@S4c;%rR+Iuqj6lzE*Pw0wst!Hc3CtculDiNLg~8aBB0Eiw9ZMlMe&0D zochSXNhFQLW!c4Pf!`QQnX*sqTL{hE^)57vR6m>@*X~Rro9b|xH29h6ol{GVteY+o zcYZ~A(3o-7vJ>k}`B&r=+l;xx5E7%rMg0?Z+1)8;2V?)v>E{6v0WPN?`o}Lc{khMN z4Fx`9Z>{Oz*d(7F#OAl|*-<1t%DXI?n1U%17LpxkG;?v{6gsbHUh508N;y*b-7H|S zo|l|1cNOBRD%!c_Z|Dny4svhI=@|Ic?*%{LWpuc;?N6gI&Dvi$^*G|sv2QH&-gCQb z(r;B0YcxedHb#p=3$>rvU^^W>ZMBA-@YfQQHsdw1fRs)4Ucgqc?5o^s2_nxMrvojT z)4!uFLGfzXA z%aYhW{}+sV6*xB2!biDupqwmDqmr+(;_J`c4_qwyq7WIE<%9 z=z%KVQq!94q)U5zcBXEyo^wet*Sx^UOBdXeNq>m_}nMw>lu1C#5t+5l{(E_Ph zXvnj+ttn7xiCExm1=3;rMw_Zi&`>VSv{3vR$RJH0mr>3yN%9i-E1KSDa3y0Dmcnpl z62q0QIc-4m#h-yx=qxgQW9TzUgdMPNZkZ}hV$@Fr&HI;UVoKKfbbrGSe+OZ(6WRlY|C% za%vqL(*oKdE8$JyEJ`F;y~K(v<~(w;8?(u2XLMh@zBDz9r|V{T#UXC7tI!e_tZ5HO7>vy%C-ELOEuywisV4!><>$_wwX$wL6@iZ7$eWdJsPm!sl6qFUjh9ZtLq3*V12XR|D{?$2Tnb>{>92EYmrjuPfQr~>>HH0qW zg2aBJus3_p0ryg67N_n|<&yP<&WO3h)QGt<^;D@_97k2rfWu86v%Dql&!&j&2D;61 z&C8SaYH)r`LU==re_JW!ijGUdQr<^iOMwt}b0m~M1`xr9YtyoZyCQa9GZy3;XJmyZ z>yA&)YL1Aa5FlwGYc)jleD2ZiY5w(0gx)KZu+TLAn5#&g(&@$bzrau5ye=E@H21JR*QRy7mS=ZvcuuA(2o>HQZx%SO;_? zFTQ{ry%a4|FLAAL&~MDaJceQlVCod+eP0QS)I-opu+*fo6vCt9R0%K}Gd>plZB7%rkY#?IZ7Y zeCXl11qBe6b`#$+m;U9-Cqi6QYWy1jmoGKLEORO}U$wS-(G-omq9J#hC>2y-_#u;gEBZ@MDm)_f)Hm{d>3VQJ&j;f3;W_#b?qAI8n1e(C}=(2sNT zR2Zq2q#IWEbS^^b7CkrMn|$D1Q0h2b58?hT#f@d6Jw2odaT=y_<|BZIZbc!n84!A( z?S*X68J?SE2KorrZN`YATvi&!b+!FHQqqFAq`8`P*p9dd5cDpcPNFXh5>cafb)8<& zc0yaV^ipRen&s23p50nX9ot)%YWzUk+cZ83h{5|1MvL`zml9E{Cm1(p?zL`Uy-|lXm#5Qm`~+O0{G=3C z&wCdz8?9eWo!c;)CD`_IzoT>%&@~m_h;(+>`G-bfYNvW8^u? z<@gce2+e?+c#O|D>WA-5u|*eFDyUH_*IwxX4YPk5tXpz1tqc)OjHe)~ zXZwu#@9WMLClI}}Mz7LgV<>E{3NTTku=ic=!)^8$lA&zd-)_Up#wgJsESyojR|KU+ zv#Miw0p7+NqWHNYr+atqNdb9{rcs&F&a7@UoAS(3;I+LOQaT<}<SB zc6NvzWMn;O|FYNaP-?;1rp%ZFTn(o~&{P7G8)%^&x+As2RPSb2Dou7h&duz%_dYr2 zoBVdv5~nwOU0I%cL>GRiP$^dF6~m(~Nagno_(mo4f3xuEG-gDliZ157t2a)R=PNc+ z>V&u{_cjYh4RfdY4#v%~DLu-spq3H$i|M5lk6CH${Oh{Q=8+Mki~Kd;_@0t= zIRToZ5sPPCi(~vFo==5_9*LjeRdG06&eQoAXHnFUW1qFN3O4mheX%)|(FjqYNe!a3 z?Wk9@8km`stQy20pqzBq?FvZ9%6Ic@*TUU6Uy3s_1#?ARzC|8iS_yyhU|r7f)ik5W(GZHC#6sq|g0Q)H-Kq|U67++E53s^U@N7`eP&^n_ z^i_oWD`K{Tu{PtVH z9oq4PEWHa#S&0?9Gyvd4&XV%;6 zwZ`l5ObGpR|t5FuH@Q&W$m|0&cbL!uvQwa{v%0a_wyU|J_Rd^Q2LKY1xv`cN0AEe zy`$4E_?X$rFmiyfQsWhjDDBUDPFfh^6tBriQgIU0n%iESF}Z;PCpp;T<&~YxapdgAw4vbM1u@CW!ALO#!*yv`54lW2|2}j?=cC#4h(SD z__FNYH;+;r#4PtgZoEEBNXAePmfFIwLQY(Y2f^J^cQ&pmql*)7uAXfu=eED-F9SDr zp06Fj{pimNVCqyZ^SPttHk?`nV!w=?a!a ziP&9P*!PRa!(PXosQy$9=sIaFt5pfo=x2*a1)7!GQCJ#azDCW1@IJKleZu(!`(~45 zGiX8kw+{KNJ3)Y-yzBSR_SRuSs!gbFDvcwjy~Dc*nX@9vGD{?{(3*%d$=ql47uYlV zcpmh3Bm5k{^~&u<5Za8eJYY}k(1l+vQc{Nm5@}%_{#FO;tc?y5fU=)|^C>ONxH>G; z5|sO(b5(cwvSi9+Pn9B`f3GVL+9s{ey3RgeNH57$|1j60t%lly`+(MhaOc$9onQDZ zne^h8RlEzbFGi`Yk}@`Yi{Bx|=rgQ*2V2fzzYRB+E}Zb!l|U?fQ@e!99}OauoA|mA zMXTUc)@qn~^W68eMrm(Z@+5VtJiXHKe&$VkmHl;ztPMw1+j-wBz4`ElvU?jNFIK$T|W|g?a{r|qrwTTA_p|;U=2)oj{jlIDSvy_F%s#fy+v{% z+I!3iUAlQhL~^15sXMc8EmjH@Sx)cR(Db!+!m`i7xy8BdCihG*0=AP0SE+7n-A*iJ zjpbDb0;5wc?hPbaFi3reZ51dL_S51NyA^_=S()KD_OiA1A=3X2mfXx5$0z8d3!!v4EP^hi zD1mfOI&j3CO{iR<8p}_lf`$`bpcG18QjeS+0~lnYz@uuF1Uv%YNBTYiKp!rwwX1A( zv9@d5>M@Q9m^%v=L!8P|rsnmS-x;}vlXYLaMy1JblY3oRd+Ql*Ey2(AyOvqC9Oixr zTjc6O=rb}Z=vU&lzSV;OB!_&h>PyK&Jz29TY(jCp9=b=-fHH?p#-URmFmqMY%v2e3 zR_FYey*5s+$hgEDK~c_7NmPe8VD@8WquA1Ty5$pEDf>i_ptrI+{*o@ zNb1c>kDqV34lgT{(`9f8$C=~8{WgSoHBXI~xQ|`Nn$1Q-jLlpFq0Dq8NyepUG~`oR zn+B<{MW{0wqgPavvizcZB+X@l?T2*GaLt#|6a{`XZThrtcfUP7Ep?D1uSq9q<;?`f zY{=*Y6vThI<8bmzHKhw=hrmHuq7yTaZ1x)wOnP|LVcu3!cE%`{7z|G zu`F%T#rW{6 z4mc6RMcCiJ^SyiLkd9&Tioe;=g}F}UomP9V$^~*!0LNu3UdiLwIKVk+k$We`jahiR z?x2_!ZJ4I;yE}gSsd2qWqO7d|fxB{IPZ$V?v$SnP@aVd2ciM8VI7NCH?99v%U zOe`?+u7t%Npvkt&Lwa+4e$Mf-yt}y|62e}T#k)L|-K6nBMVe!0WR7`YHoEN2g@JoEd!V>l> ztkk+ViIhes=85qkw$5D=-B>{^4}Bs&0Y=Rk(eS$^LR7(^L*SNc3N(mOoXmE6PsO$| zo;jIv{xe^s6!=F|CXI(?YgHk9gb5_uK>h3=;wRu`bk>A!u*bQYzj&)rs@iabIFXk4 z@c?N_)E|1Sh{l{crkKT$Zqi{adqw36A!vW=F40@AqO7=<)?j+C?FseDK=$W1wSS^7 zXuCojcwJSt5&Bkwsc>$AZQL%U2Lpk1_>=ZO143SMQz;8M$B7wum93-CGc@4hN1K*j zmUg&w+Hc}^^LpJ{A z)Cq)G8@pQ?joYU5`Qcg0Qpk=ub$i-c&#;F@QM-wf>@JtQvP#a^K%oFp!Dei2I?9`A zm9$3xXiQ6 z(^8$ka0C@Kepr2$14Vk_;YF9n!bq)$uTQ}t{-hKX^Adm1EyoT!CIfIc@9#5BDrfBK z-=5z`(_^aq5al{aI_=K}>s*{& znLfnof>|_B+H~JEm3djv%G7*uv^v0x2NQ`OPGbauK!GLksn1W761)ngJPlzVw$vpP z?l2Fb6Gc9`7=B*lY-CKW9sW57x2~W6pn1dGgEX_zTBzRfb+?H&rPWa9uQ9-G%tZ~L zOZQewdZ$BxV@Al?@5)!y4shv!oT%P<@Sja@E;!8u7d>ddYirLj=Sfb-G zndblcLN;?4j0I>K&^lYzNIPOahWjq09PXY>`g(@>?zw3c=c-d@4Vf*L3a}_hlaC=& zgabqUTQyOhd_X3%uvZl{MEBEHZx)&M1&bZcEizOV3W{WAp}oFWibGe?rv?}JjDAA4 zQ6+n9pYRvEw!+qqU<+)n3DG~hkgK@!zPn1Cm#DbY98_K_qYd^kb9Aw6fO-|VNd&zF z)zBNUfa9Dv(iw$uTJ4Bh9y7u-4ZobB<=*>e=NOLd)ffD&o`v3{q&&-ug=bs|8@c8+ z?cvvZAmuaI^UGFi+4VF~>N{Ksd}atAs5+iviz*T?;$U-zDzxgX+m=J;_xitV7yhvp z_&3Bl03-d4TaetRbd2$mkGNY98qOldO?&REq zbb0oOcW2@V;dk9%3idyg2Nn3(^)yBQ171|2sgDNs61#bh^M|nR5gfVg|n%pX%^iI$SXsR1keT0 zs)6YKNLy;p%Dbw1?;pAnI=0siIWGQe2vkiZdt4MS8@02yd$@oNygu8tIPlS8+e!rF zz>O#WDJ>M9mbRFJ_TSp!Z=RXm&7GHR@Y$w-E-Ilb=JyF!QFH1Wr=6g_(`J?bgJFQ1 z(q48hQ`ZZIOuOf*k}QTCers2Tih0-5fvIorAHIZ00&N5yBQhm5f}F=e30muJ8|lU+ z4jZFW(GCO&ifdB*>d2cd8KQD=tajx~SnfpEMwXv%Hs<&yXDlk$aB{B`558cBeMK*k zGDuT=DHQrg+XQ^*{{R@n2eAV+1COrtLL5vTtRBLr#(WnRF$eFxM-HqVK=y_*MyGdq z1oEMRc{hAxsn_6Ohiqv64iUoRnw1|oq9GMKhi-G>T~<0@Xa+HV)Z$w8$>tv!kSd>| z7BvAA(~cT$G>QkoP&~btg1x)FlHziXFJlRcFp2l*Nfp@}5(Tw*^82%aPd+*VTB690 z5a6mesuL3S1%8bvzt*fT9c|tZTo)aur7ufgf>5;o5H97Z2gceYfqiOetwCg`W4^9u z7w=rT#eBgF@b>xhn~#?jmRgtQ2%|1P8%a;_9v+_C6ksx@DlZ`CmFNYyesxP?bYuYr zClZFl!rB|gGEadr$Q(0@koPL1AXXpkK>&VOKD-w^_<#Ll(-bQ z-eS{uHA$!;|DOlt!gP~R6Pbh0yko4w#0`gT#%Ygg*;)f&*UA4#8*035Ibdq4e!4Ca;Wrtj5iKP8 zxe?(-VR~x8Cxs*5-S8fY$E}6{nIL_M52;VM;1J+;Y{}B}i`((U&C9@aa1TL5+bNO8 zS9TXKhrg9%qlxgTC|!&!XsTnQe4~BSrXU@U6cJ_Ts>Wq<$D{Dl#L`yxEe?`Vd7ezU zmFGLeZv|b)*5tNI(Nam80(zM&N@h|s!M{&lYuvy4J?tq`9pKguSUB=DI8kJ5Ygk{p zw@G84SX*uWxpo-Bb~xWEMrd|{7YpL3#4-gw^J)!Pt$Qr&IRie0NVI$10gXW;pjHefkSm}P+o2;Hp1nlP|ZCuWyq0 zMvoHcX%4YRuX>f_1tB0mhmVU2o~Ta2^&f_CPHIS1JWnn*=4O_uedar7#nD%LtfkYW z8!Nq}WLNF|I=op~VRz;t{|j+#FCpX1A*juq^0yTY68n=uWon5y@38n0hA)evnz4(p z?N@oY&6v9INe0)bsLk@}MsmnJ*9|hM#pkh? zdq`qkik8HMIi@G(akdx2d(XsuQoAr68aC!~2jrjw1-$=vzr?5;DHbf20iIl)UkV<+ z$>mS#=GlS|URFBqu$=XdB^Su@HTqTK9UHq3sPM8FoTI;_1{N*fhwAY}=Vm@nOT9EK z{vKlEV5H~QGQydWE3Ol(YK3XsO&D89aSwZT+aa~97>n9EbhQCr=V($QHiqR)=N&|f zN5HWjLrbUaP6xmtSNDUVLK{}4fBuUnzUA~~YzH39FMNQCDfaa2UB#sB@ryVLc%(u);@ceRBxjteOs{YUGR;?TVTa9qh7w<(Vl^ z_Aguf>kBh zU>PQZ2+jX<1HB+v)$r<*qZMAIl<5D#F!F4!dT~dNdEhKH@}&F^-*Lgx43c#B)Al0E zJ@|Fkys+*+30)KVrDOtsTzGKG) zhf_IEP;mD6>q@c^jsQUFri))41XHJ>#(%eU;H8kFI+CVQ8t5*}6K$)w{ofAB!Yd$tzS~h?*o>oQ44DdFk8ZhZIlWT%j4S#9(ak7TVSMrdRAH z^9_Tw$5KSYS=Rra{$@du=pXongm)vce=$k!EqZEH)reJ|*)Z$`Gk{Myg2di^_=ls| z%=@ihq+djio#Fpbe)TcTSOzqFtP)C@>a5+8dZC`Z1m4NbR6$n`hRMNnhbrN(^#{%V zrR7ML7drNVkmqEl7(VIiUtDXi0?GqOsEg_D;W^%gn;2XQc+?BdA1iFBSd^Zc2=r=T zl*)b}u2?UtuRtG4=$A<|cNL&}|C1rF@x`q%F)Ap{9gV+YR*IE=9W=#g1bPFj@8QZi z5;VqD%y1BO$-t&h{Q{Z%6i#J*S3hUSOq*ka0t5BdvUOrye7-Inf-X` z_WN(VhOOoUFL?l1p}}1DlFeNbx^B*u~If!p!)*U^o{|Fi&9*`{cw-lF9A98Z*)F6yC`kmRch(lbd&SJRxvFjYQITWU)Xa{G&p> z;RfMSj*EU8RGVhiHXj$oYx)Rys6ZcxH?dBng!BKWo0AsQ9G5^lzX1aDEmpIO>0MP5 zUXX?eHw}nphmO-w2AmOOnWHM7)>-c&>&!^{b*tspE~}ke2+9B{8cVZ#rb#3Sn?j; zSVj(>A=5mHe{HgZhNassw85NqZBMT+5%LZN$I^vg`E&2Ps`bgxj1qHifF;H4C-Z zwLedv&4xJU|NX0njv}_fyu)rO&VtAd{}*gVLZS`Ou5S%7GnlMIe^p~rQMw*)oDu&Q zZhDJ)-5RzWqcMsFO49=B-+gWm$+$f{c6|bGd8rTYY18*AFf9k85w+^S@e6^(Iqzo7yEeH#T@?OZ#Wy>%p7la&oc6Yc zj^Q<~iDdbv@{-B;7z7!WAd^y}ODVF#-G4p-lN@lq%j!#h%n$x#UgkLNB z3)@aZ zO`;UvD4Qx!N(t-3FB=;3ELsEo?@xypb zKY>S99~7D;zZ&tieP3+qHa`<<-qNXhnIB7a+v$4X%dpIUf;CI-+8w2cCb^1RHTg-U zq;KfWpH4-axRE|zvmbj+y5a(V`P%1Y6Kd*a(tHj{3xuPD zom<&GMrsGw5YCQlAjqq^bKm-=T3%LhWbw4$|Nr+;B+lO*Pp!1^zuQUWW+W`vsa$H} zsr(P-aFuur|IEDe=uWw-ARMg0=$T1yCvmFTM zUcd1lsN|Uo6MgNcAR_TqSukvsDj_^L8vx|JoTwa-)N%O)jmn&Hckv!Y{m}C8Y(|O0 zT)C>Qc8vX~<$U_R?B7JeY{`d7K63_N{<_N2Po0f*DtN}W4<#Xpd#YW6Dnc*@>s&kJ}K(=2%hoDuZ}G6%oKH4S=@ zRc6~~5e{zgQsF6C44$+nfLl#dsfq8we*9bVt?_d?F@w~bAWc0um?pZ* z3b5y77%(`^yWr#8({)khASd@7lFee=Wtn?|npKpSJlH9oP6Bz<#!iz^P^bYO{{P;2*}VaES87#`hV(li5AZ{H)~d zhuJ-Rr(d4)fdmvJ@3>c&5S}xP)sOlS1D@BQ_D|Y5A~>ncXbqmlXuFbWee(&f?#n?s zUkqJ=R)z`|QQh}};f^;-QgEIAYC^DMIVyd;g<%Fzrv9}E0Tok3je;8;wM37Y7E@8z z@;IE3Ifh-2W#oKisiKKEs3MO&Jr-h+&Lf$4K+mwII%fMVBD5X|f(C_MN+(`JB*Dka z%&h3=iVbH)L)f>K?{$^iUNK$_aQ?i0iI#__lD>)Ir^d10nov3t7qxcQj1)gH*7GqU z>4kBH=_kMaF%FGD4{#xlMRBHSt$OGAnkccPX8!E;o!HLjO;JTnc-r7ceRVT&@(PUQ z6wq?S@|5gjuyTw+R_jvU;~Ju|nIg>i@jup8;42C*ThH=S(xe*RLcd{S_+XojbjJ2` zojbjaynu*)<1HX=nmC+xL^Xt?DmXt~HMFOhx37l3@#FKLC7#+b7FttrXULx;H{*(s zcACQyRzq*OxoMGkv1Pezhf58WkBee=@;!^h>ldI^R5g)7s$EU#fX@#ndF~gt>zGQo z`6Iw*0~RE|Gk?(1(<^CKoEdeo6y)z|4M}@taq?M+uH7dlFch(vwqH+!n&?z89LZ9S zX~-Gt{;%_T7FgJXh`jLKgsWjbC32?^t(F*{fG2B71=51Egi~(oEB59JrzNb?vSIe% zq?2_y`o_ZJhkg#2c9pZa1>^?5G!^0pvH`DrKSZP@4fYFygDrVC3jw;8C z;ieZs*!On5zo(H=YO^QpkZ_r0m%TR+bm4VQ5n~YYWxXQYu5ya;qj&QvGJ!nA{&t&% z?CQOB3@x3XN@Y3UxfS{GqHBdlIjuMub!y)twmsv{@}4&3lU@tW!d%Fhd-PbIe!q;n z#(8Ej0AKHNdJ4Uxqou8vY(i|+0pZ(?&!N;9?rtF_rB=(6TO0p?wpdK`1aOCZT^Rc~ zc-mCYC)7!y60Pq#x>wPHrmWx^0^w+m9GWs90b0M}VV3%8^p-c!SrMs+MXj_3Ix$ec ztwxd5fpb3S`kuo+l;_}CU#;NfyUoM|J!wBK<(16%LJt~SrcUmzwXPdWYZ@%IaiRd%5n0eR+13qN*FPYuc=u_|BMuri$yPd-L-c2dj5pehs zY+4Y&5DF70 z3cf}_2t`nOFQY9+96w%zpZ(=kGaACO>Nvgk90P>5^zFhab_^HR!jo)2QwZvgnAPKQ z`|EADabYkwO6Us0w+=L7>+r?Yy~Y(vw0=PK{ZFIL%^&4z)CeRY7Is%7 z;1N6MXB}q^=-k-!F;acO2@j0A3JYr`jYq6$oZX8qfOU~{On>$9U9seh`b@hgmIX+x zWvlc-Lrm@zLPc%%A+i>$K(ITEyb-Mr8aXi-ANTVm2q=h!skn?@GN9_k@k7JR4#!X7Q$^@ucvrC)!tz`KsndLfg?87g=d3P~I$PEGysn$8(5qarwRV$J@Ck(-RcBe0 zoa709lkTQn#J0NzHl(Xz`ov|-bCj1x%Ga02zjX@j(&b{!?uwL##U_v{hu&PMH@9Q* z_Srm!)%UUpd`0KCxYhb&sgsv$Dcc@B4U45n^$319V=mN}(>KY6c`5-Iv*gg`I%Mg* zIfvvK1Xp_<;gC+!^_u41?LL!uAfd!?50)a zv-WI8K%eHdIlrXD!0jLWm=~Yi&OP7RLNG~AO%|<`2v)?%`H$P#RO6pS+yRIW+owm* zUYv0JO5PhQPz_w@q-HCNoIeuBRe2p3w(X!)`;$u!Vh8iT`G>mV;*pFW&vixV2A%-@ zrikAt&_FxfP)D`4#F#O&+^? z^EOnrON>5v2O6GeC^x#q$9y`AF4uaiy%4Z|Z#c$;S`_9nu?cW=&QJN|qJe|Z;Sbc! z!6OM-5Ox_%#V4SsAUhO!T zH4027b)k@h=3R8fDHkONTLzETUThZ~mw!^bR}4+AuD^=Tgsf3^5d0Zt$Q)bfEKIC0 zfE8T7&|(JT6I#Z7owWuKrx_-BH9R5*(X2l<8m`TX%}Y$ZdR>dB+AfOR-ALUUGKbJl zHKk6t9N{F;K`U_+3jX6h;+~3>CEsq^VuRMxs!Fneolm6=`_t+fs6<1XEpo}iQ18pl zyvhGEux5;|=yZ)k^-;EZSGlJttaBpEsxxs)?CT?Qb@%1(T(Hu^pw+Qp!w+G z5&gXTiYQkz+FSCalPP0p)Pt;ads z@HrEN#TYT&a{&Z)iv><#>8PjSASzH+#CH?EsVpQBWRS*uQg;~pBh|{)i_lE>gu`B2 z(O0n{EpCi|MnwEvVOph5*)Bkec-Vd~tB;Se0bzZ)c{;3;({T9UG;kQSzPl$B1qnvD z$AMqn(!Vu}w|(Acqip0!uPI3?n$Q@aH;6*SX^jqo{buT^HG<{upR53Kxr^d3IqMi< z%y0w_Bg5dQDdNwj!K)`@z@*ndUgusAE*jT@tQVwv749tb)ibm70dmT*Kc{>g>8xkw zM?HegoO5}EggzF&x}=)x*RNdx1eI~&i|*iuoV-^`9_n=+PKYi=iIQwVuc<&O%lP&v zOo~dHB2y+IsVQ7Fm6?V;V}@Q1~a*J8llbu%jR%% zOf&y%6Db=mOYHU~K33Gtw>SHh$KlKgvDCP`5r1VAI@+yJ&H~0jDR#)qs9mw*m$Dy< z)&UI`8(rOBjbPHki~M0DhA8NCVzp4mF;(j*(zXzEj?S>)EQt@LFJ0CRfTV07-|B8E$kWDSAAF+Vp?nt zKC>4L*WxRu8^L~Xp}_%t!EIJI*BYZh`_#X@YYUK-u<=y&1xm2(Ki>7-?!Wp=JI@qH zw-I>l0I(V3)C%9hml-rI5%Lt3vYK7#j@nnsHsD%## zQc6Kqy9`f5v}i{`5QM*)bWl1dt%I6dYc8%<3%-0hsQj>#;BEBj?{yg}swxgH2Qb9= zMPg5|KFM&pbPqUWG3REB_zA*?$Q=c^!DsVYE(>T2g*O+nm0Q8;op^eCMG+rC_@Ld@ zj0`*hlCABS@aI#BPLL9QeKu8Msx+^d4N034uO#n-@m_p!lrre0I$zKi>x&lw2r|p?mIw=lV{0_ zzDQj3x9?ZBS94z(3$}#e4d|+c_Q%0Wf#jI|Is@0$zzCWGD- zmzdfWT;^2VpT9G>cddJ`=nha_boH}q1*|16lYd7+%5j59peSdL4d6r>@&NW3g)0QO zLQ3D0o$geJ*zP}7yO0A9%1%48eFu1sZ2U#ZFTizYjf99PP;+&@qV|(v&6gKFQ{v#% z_D8C$7^o7S2U{^f{;OS4M(5AD2gpR!6jA{#9yZCHx&RR%Q=U~1D4tkqf@(9r{{%&G`ToC>NsugVVZQdXMy!;O* zY}?p1vJ?H+(G#jd-01lV4(|43~E*A?fwJ}VEo+U!ck zlJ9?e1li>)?m7_45B9dx8Lvuw27gXc^f=(y*)N515Rw+b>V?K zm?oIkSmJ+wdGh5Jo$1*Tiq5IaKcc#Y0%h0OeEt@H=ryyZyz z(;IM4cI{?-GjW^=f*}0W{HFN*;9~>fPb2ILT!uT7Zw#C zUJ8QjvyU6aKZJm>T~5|01=``k<=oGT_WYwA@Vt9CLj|!%)2ms<-C2mP>Hstxr?SN(=*cg@W(jhPnO)r<@2~h#7>K`iusU^?ne=rK{ zix!$2&;oo`wp>&u7cKyNb0Qi<~J*7ftaQM2&w7~fgWnPthEC3!& znXTG25~yuMvd$;V?8dcl*=7CRky0{WELX=(Mrt$l6R!o|QN zxZc>JO+emAyGY#$fIsz{l4OzBk(C17*gSY9K+uU^OX|8oQnN!L*CQb||Mp%NI>LS1 z$(y!}0^j=WQ!76NdsBp?fAFh(e2O@3R8|Hq(hDH6XX-NJFmS-8mVB$|AAi;5_}f;9 z0dheQ{&p@-9u0JNpw^6!N911%vgPs!=_jDt^YwT0U%eD}0do6iS2E_lzx8m}_*}QU z2UV6dI-JJ|6nm8WWSw8E?ELn0$}K9ZybKzZ(x z@Y{E1DQE4+P)IwUd(T`Bvf}jfEQcX?tSQV=0}ORcJFIV*3CU)J@~&(KXdkP>WNzT_ zxbDQ_6+w3;&w8(d5Lw`drH8(T2a4&%m(Kv75d~8FLV3duJ?|ams=EW z86KI^IzypJ7kgf90NSgME^XSY>Kmt@?m!uH+eTGp!-MkD0O=Drq`NmLV<IcI z{0_H|uRI;{jqq5a-UvrV-fx^C!97btn*J=5z#pv`Ape=jm8qo-rNR7_Ws0={1e9=I zqk8T4{bb|GukW0JX9;Uco?HvEzN$uw^$@(U?`QQ^fZfyrHok>{snSx(`R}WTti1YD zMFP03s*|_KAkY@blkWKpBIhr;e|RT6t~qzdp3>lLXuh}1QE)9&ZAW+pxUJP)RXBsq zNTt0j5+Yltb~iMEaJ3)UJww zp!U&pRm^-bFh2d~+iCo4ZYghqv6;baSqs*4)+N@)koz>-WGwK@@%g>wd{>3{mObcG z0u+0cak7rDmgj8DT4-DiXD=MIJPrj*rRP(c7n@vhsGlb|9CA*v33#=A{*%QwdNqYh zpB(QKg(`HMuIPL^N5dJE51?(!>P6O{L zUPrz9!ofqQ?p&z_rRs;OB=PPlK@bE%_#jvlt$EGs!TO!PceD*iXukLF@Z0k%J-PIj z5_e$b=PTx{83{iG{LpxLc^G3F6F1B(;w`+TAJ#H{ZSDv$Ts<^>9}(98Fr1=>nd68VxPJ8yGJ=lS+#Ae^-eP zS$INIe>)%Y7wbjOifrwW1dq-494o#45oVNmgg;GS52NHJTGsa4Ar{olWO$mFWI ziXPyeUMIB3K(JBz1UYNqN%zH@52_*Y;@pBe2S9|rda=@D;o=GNkIoa|af{?xhB6Rn zb?&1YCi=tMT%H+KA!j+k7tz8fagj3Jj$~WW1BX|SQZEpedD*dGw4p;dy~MYu1_8B zyO#7NCn};cVO6N>zzDrUtgt4qjq^$Lmnb<`wi(?{(NBEsgqklZ5Wht zsJbHHD7Z#@&9#dIWhJGr?6v!FjHQi*&BY)#DKb6X4$dhKHGh%{=?*EuxpA+yD{iS& z#Qzb54hpa5ngM)4(wlZ)bA9v6`KN)km&M=e0GZ14-TG;;t<16-SAY2V_J4Fa-<7kY zyNB-OyHnP_T7K>GGkv1N;qHX1L*p}m$QLWj*s>6+(wMyCUTrTAkYAG?fpe_;X6;w- zj3;rw=2cpL*&(%^j)0wy!)f)d*S-j!w=Sg#9G&vh{d-Lzwq~SFdd{o$$))3KXU~EL zv?Khr8!7C398%R=z;8+DLnl2vO1e#w7$6q}LHNscZ_(ps7bi&m>gtl=4*(-Q#H$ts zK@bGt1CwN(WNG*Nvysk8=cHZG`EZvF9Rgw9vvrj=)rEWU_trdU40*!yvFu^ilQoH0Z8 z1*lD>E0(GRdBdfgOFFRL+&bZ0elR{v?vU9Td~3HIQL!XQdz543i$#BU8-2Q2XI%=1 z+n!hz)eFLB?_F}d1UOwVmdh;*zHK{YRLuuY1!LPkv{{mG@qFW2}kzVx~c_Zl~irke(-%-abO7vPV7zg zH$4OWGSk=ATcF;em?()?I=r*dk4FVeTm*AInKN)&I@re88g;*?{X|FwLHH#u2_7Cg zSEyU9^{j%Qfla3FkaFXTl^+3D{&RaJ>L@u7OnzJe*RGzDZ9mxg5$qr@kCBII4*Q`wYmOUzDdC@PE+9#w@DH&w-W?G_{%v4I+k!4 z33kK0W~j2jI?@0=L|g@9~XKI*GFA*y>Sz!^qyR8=1`ctY|4@`$HBRS za|fq!B2NESX8X*L+@A2b^!H2moq`PMR)?q%@bda-LzxcX{$uStg{OivS2bOTC2iA*Frs1RtePE3azvVq(^*BMF1|pOZiWLagE5Xd72wus<-w9pkejH!tXTurm}8x8eF>)JS<`b}z<+ipRn?^+ z@2xecJBt4C-!9r8%ll|Nc-d{Md#o}z98-i@t3yD&?k8)126kVDclLYvpvsRQZapzwpoX^g=R0{4cHOx1zfV*b`zAIi0!a90lBf7``yW$qD z%I**T5AUe%OW=6va>;80!Nwf2&tn+`Jnh!J#y6najSRn8QD1LM{uOnxKKb* z->#!gC_L8Azq;=WaO$$wX1@j;8k2_{hu|@h=2tDXU#JU zy?y_YZuh@SnF$XE-%L-u4o_cv{;SIQuA+`Tx|>!Wj79Vo^IUkMxRa7x5n>lTt)9O1 z=iB?Q42^Q`^7G}UPR6#D8jyS>erMLKUtbP~HO{*=MSi|KH7PM7=NMS~TiaVV|9t(+ zvZvtbycc0l>HAZ&3c~W5XV)cHf~7sq zmPjfA>z`Z-PbmxMy`Q$oegu+9Iz$8%8R>h|rWbs2wPac62hgs5sRhn0p^Q=+#uZ$8>B`GvLDNNXSXu{C4R*d582>}4k1F~|qrubFE>dw7s%!K1vQtC*cI7kN z3xH?n#{;gFM-vOoJX(NA7_*ywbnM^2#7tvP#0 z0>?`lM`-{AZ|zg2-p2nL82<-RJaTNPJp(}ldiAeY5nTL>PYAvNd5>O7yJmvelub*| zhJZQIeArq~^b^ZMOFC=7@L=zh3;iMAol9r0Z3gG9#gxH12-5f5TBj7Kihke&<8NWR zBmYvL`>&1*n{jqMHAkT+$Ge{X2vQzjyz%I?=qE45UWr|m`3)3ab3Ff<8sKvL;krm? z@N)a8cIg7(U9ZK=vh_jcpxUPx`_HMjR6|irb{9Ox)_qcR7x;DU*uKgGPUaDd z5Fr)>;ce{8IZw1{0Pbyk*4x*Av2f+9_2;q=HGX1+yW_4Fj642XLKS5w0i~!$z-t50 zLH)u$cS!)%HcfxUKDfF3eEXQ1U|g<`Fh@WY8u6x?-cP#j9rdsxq=%%Wm>=Z{w<9jeW1S#9Z}O$w5HQy?=UU3W+RnRVU>?Ul5Z0q` zjQbkki~r2$&b`Zh;tGR)YVKZ>%dc-Yn`ycss+V6}UYk11ipNz>WQ7o0cpGaa-(kp4C9!;z34+3e-bnrL9R| zQ(SIqIRq{RO(@8tOnGFgcbk)61j4? z0LvLe;@?o=nt&S+%Sv{FdKO9C1Y1+FiO%;Rd7q83U@Q=MY|)LeaI^TdiQ8*JrnUBo zh~Z#2(ovH4`4@JSgLm5``h(=XFZ;%3kag2lQ;&hHXLU>kK8M@+Glr*}0G3}Jlalxk zkHvPdY%EdV`#xA3WM0Uc4t4|14ZiUVuHHA*&)f>uofU=zPX5eypOHul-ljef;U_KjG3Ke)g2#)0|35ZkWw1ygG6o%v6IeV=_7Z+ie2mwvu09xkqJ zu(RM+;E>CCUHN}sZ^^SKIL+@Xuv=w2UHgVtphWmav@PfdJJ0MmyyqO$-CNJGRtQvc zsP?4t%k%K>I3J4vazPN@%Gt(ki?$I+I;jTBe!f1ga&)`&qri1WzP>!xG4f8j1VFym zm&^9a>q`s4Wukj6oj=65KAnVn4U-|KQr0-b^F=%8`f9yrCx3f~E>JzB z>EXOPLH=0roh<0p`Xo*4?yP2T;mEQ6PZt7L`Gi})J|2c5xmS&GfF3lzA)c}#vbNGj zAd4%^eeH61wqcsN7LbRV;^+_rK@bE%5ClPZ|M)-kxAW}|u92=~oqNKDr5m+d^FtAe zrWF$5`Pc9ZsDGl!0?4RUVQBvO;5@~>m~(fqo2ky0RR>ibxih&SUn51)2bArl*4CmR z>mtdl-vf~SBC8UZUdwnbko*AzpdY31eD%>XweZ}~Pa$Jbc!cj`h-v>r_JL?n_T9AN zd|}ASX+F17TPR$-h^JR+Q0FLOq)-02{Uc%Uf$rCyLPF6!?XRVQp|@wGYbnULs{5$w zYeBQzWxEbwKczm1_WlDygELsG6?6&x3NEUb?mw9en)T=J+-?W9b7gLl0eqUYxm0mB zD7xt$svG`ugJC|W-)?k;C~3uK$Cd+?uh)Ba6jEn|+vgb${^nmz_p<9Eq;4a@@+It0@AMBlQa~`={2f!n0&^CJBJ`O(Prd$uBF06bl} zrsRQe$jM1ulC=x;A2(=M{w9=KS+8u^XrL0Sc>ZA5zo);90LbSV1YD}&bj|Ay;Je|} zu^`a)+1~k@7x)b8Ow}b|TQgMV>G;nNN#?oxH)3~#!Y<>XJ^;Y>f4-XF!hs8)U#ot$G=JMHI2d){{E={|cDj0r%GX7l{^{)V`%H7v zg8h9T=m)?f`T-xkVdM6|!~g$O=N}uCykOI*1VQCMQ{Sdpofh!)#6p+-k0330P1KPt zaKb%JSJo9ujjdcKq!+l<*CF5Y52*VHYl!s<#LeD5AS!I}kwe8GYumMwk-)tzg{l@D2>xvwoGV%gG&6P0GzG=^ z*4NfARP{aVnpER%`ES4ycH)!sr$83HdBpkrARn6ZgE11kQa`#-`Zh2|dLR}4=hQ#@ z#G{4r$06G9`-}T}fbn4ZI(-wkGOYZrQbVEm+LE_IJ$GAl^YP z2*O+C^vwFqF!1f`71Nb7<&)lUqEB$V|G(~_G;GNYH}CrAmQ%G~9k1vJWyjP!6J!gO zs?{G8vH{cuHOm$GzS=9aTwopqQAsOT z9lH!Mv$k|THw&Vt`et}f1-G~^ikepNs_PYL*Io3B|K``*b+6_Gn5LHB7G4By&MZBs zSaZnOcI(^dcJMHCmjBK;&?m&rNNf7x*VoW9p?mr|c<^LqRM&&&v z0^qT>z>~b^pHc}_A&0g&Rkmaf9g8HO?o4z_vlyL!S%azN+@#BE#RFK6`cc>YPw;Yw6a~r6*iEb*b-_ z3y||TCnvY}?=PPiAQuGTt&&>Cb=^W#fGp>f3x$x8P2wGpAMCg1w>M&`(x}cLN(&ou{%CJht49N@@qD zFO35&O3M?HE#x5Uv2l&%Y0tuD+q!h2*N*4 zR)w4|EUO_oB5ks%;j7R87M9hPFzd)y^?&*A%jNHDJEcF(`plf?_t(=@-xO@M{Cq!G zQt!6Rt^zsV>1$fT#Bm}BAENrP^}E;319?Njx`eiav(niO=f4MQm^I8Qj%(pxPP-*| zkcTgXlp7e_C@(mE?^VZPGXxDkVm+&ZTj84`&PT$5(zm;%$AH1t9Bf_l{@2U2EPYqb z7x2h#?v~woA>sU?u{Wl`L!Yql&`5AR-a4>SS;%kiA7qyx`oY^sO*lYhD8DrHrd=V3 zIvP?o@Cw9NZ;)Ow9rP)2l~Z@a1NAqTwoidfr<<~<2Or?~fM)jw#$130C#OH$ei!tG zW4=wv1>gMLir1_H_nHJO378BnAvUeeOTfO4ri^Tj=pX+r&wMurO&ic(mDQBj1oQZM z`Z8YNF0Z@3$P7r1IkD_^Ie6UoyXCuYf+;ZdL3X7NsDdM{Vb*JqSm=P;wYl)v=ljUL z13^{asio~?2pvAq+UPJmyWlv-sR}qQvT>CAfRj=CqrAE3A1^sIv#BJn0`@gDW#y?5 zqg7qgR)dgb18+6_0kq!)`+Gz}WY-m&j!uTy?pu1FD*@Jzj7!XeKb(4+S7cYxpM<9$ zEggQSFT|;LjK5d`oT`*vn&%V*MtAqCISu5Qb~QAY-=2?))y3*;g<+M4yMHwe95s$5 z9Y;g8ZPnD3Cj9<#iYe9wL3pbi&n&rN13XB+*)4G`_t)OR6kelLmX8BC4shM5 z4FOq8SsSSZ`0NZyakK^gYudJ(RW1a_t^vylb1loJSIg73C%4Jz48R*F#HggL-y_rp zN>HVxzYA!a>Sn4s!`k3^4KH+noKjf>|B7<{lCsIQ1_vp@YfHdghv(D1S9{a6?*ZlW z+;I+vfbw~8ez~yQ&$qkS=$KPvXSmQXfv24yq-}U6(g8t} z^Hp@!!p({oVq^Bft>YJMVvP_%X}n*{H#ojPW!Ea;?H%ataPZY~%`@#Y%5zf>Mxk)Gk$;a zdeds?uypv5NAg$8m(3l!KIk@hws9?_d=J|xylzMjn+q`g#z6e+e|tSA<@V|2`vi)T zub(yu#%C`q_v|7Ziab{^V+T|_S?ZAebQs^G7d{Qd`QfFwX-Nl+p>Xr;!x1@yz}VQ7 z$7%(87hM^&MR-r$IZv7s-n${bnV~p{cefAz9k>rJo8{= zPT;HMK{tFZs2_s!F^BGoyP{qHX>>((wX~<8&7(GXTAYJX2S)l&xD1Yy9g8}ag-TSW zyaf(|seO5f9%#8GDMn_QEImS6eKv4Bwh3jsC=(r#EA72A;BYrw{VJGCMB3mRH zz;9KD&?*B#@riv!n`!S~^>U{~*G?V*k*AlAIqVPF1spdEhx$-e?)o30J$Itzvzthz@x|ez;>biM)g(j?i$$4VKT@+ zR=7x32>3YfMd#NOOyNH!W_noZ=E}tWa5eJu=h2D45WeQMJJpjH#ox)^4Uq{Cnx;5{ zpIgXgr_!(1s}5+wpmnibtxSiIq=FCJUi;xx^Gy3p`Gz|ctv9Ue zXbul6-TXRXFB-AwPD60vm;QNUikd^ar|#xA2%E7Y7brbXqeky(B7soW)Eu%h^d#L zHNG|wY1(P_u)chfe*E9xFaNupJWIHoR%!ugfchZKJFv^@NZJzc6)K zn7Fe~JKqFQXDb&;f53U=ZO2vKaRJ|`ykGkCbvbO^2QUQm?qJ96Vg6jm=37iN;Jc_K3m&k0wZd_v?35op=$KA+**qvLHXDdRc6F1~H;aZD|CCf~J zfKsJy1r32}4q=7upMkur6puLG-GQqaQmJ*z(EhCb`CDz zBpo8pES+`O3elHVM<3r0Szp|*9=i#A^0&~J(}LobqguQ9@2iJtZTh90VGwhDu6sH&N-gRNookr8+G**!@*&S{E`levHEakJW3@&r4Iu?uv)gAW|_H`k$+R|-@^TUIk z)7xxM0?$)TW|jI9T)(N}QP2jY^~zTA(SKika-1IBi+>D}4OZSg8U{HJo{UfY9Ne4K z=~QGDcy4Jlt)v6UBGmp0SBS}F>3;+s=-tlw3@lHXi1#mYb zp^M%hG9wL+7FVz>Zj-7S4t^Wjt*qP)Y?}M_br}fHMsI9+W;WdKFr~$o@4&lR%faP0 zg1si9kN@BKiic6w0@fRleC9;qJFDT@@{NU0cLx@kuUJb$K#gweYJ`I0rs5vKdf=vP zll1NT(+*;QToC>R5_UYZ%@_}g8%h^hhc|3civDkUg88LfUG_j@SJ+o+W1rhgA*5P? zVXjjk=v>%-mwcehYrjhU+Pfk$wxp@`ci@4~%{d99;Bmt3^rTcU4l#Uf{_G7G^l{d; z){o(^@7BV*98WaQnV$%VmWV=QVG1LJ}q2*SUN<+|0%O~|SD{E5q#idQUCISfYr(rZUxToyuLZRBdiMCSu)w-)%xTcY3)s8K(SVFO}YX0dfia@ z^`F-xZfBB>F#^(6S)Z80-@d=M#$kgc+&T&#cX?LY&>C!-s(+MS2d^b=*~(_WJ&s}D z2lms1f$b{oP}v)%Fn3FK>ZpAAj@;^y^-0b+%LdTP4TUW0LAOe~T2=t$&86$oNO-zC zwzT06*co&~++jGev;+61PBAGawKL%KmbB zU-q9UyLg?U705o8S*1}RD=2M$m;V1%zfSkP=0UXrr7LCah0;M0**Xhg!4}@VJ$|Xb z{C7ExElm;D*N%(R53c5Bc7a<3AFa>50d0(RcevDf_x^Wy%lS{!SHP^eRax-^pw;%; zH=MJek3~r?v1|cTD1DY#vV(;Ze znV-YW_7At^hQrX4bEZY-g9&YW74*3YIZO1Ff6sZva@}fW=+BR-$JJY*10DaEM{xw*D2W7;{qr?T0QCb z2G9-KGx*pYcw~5bFSZgq?OA?mfuT@oQiRfV3^)ze8s#JYyz=GTeYz*{Cy3s-_U_3f z$Xs=MR&+ANulKa^7yviR7GGL86!N#szt3$DRG%5P*FFjqKC)Ka7VZ06`IK^;qwEL4 zS*{aQz&X>?+$ciP_t-#B(X@%0> z;Bc_SzTnv)6_AgRef;N@*Lu@D#-j7V!el zCiV5j>VRy!%_k}|#6+bZFk0c+rSwClyilV`nANTh7=Mf_Xzd~T!7nK=*>j&wQMhG} z9iMv|qKc&+Ft~w#6{kd%9PFwT#19Bv|K+5{3nAM0W1-`nAo}r|F(-b6w2hbhJaz`3 zRxKRLP63fG=hK(#copA0AH1G|UDmUAiHi0RJjZdgW%|rWhc96YtL)0+ofpy4G6Rx&>l) zq)#(mgU}PMr~zrCvaNgpxE-q1r*L=BeB_<$ydI+RuW~we9v;`78?*ZuxU8t^5ncy8 z>efju;tYyw4tnih@TxR@>)2AMli=CdjbW!-L;9LaBOix=HYcQq$4&67(SB3q&Y?G}3&vQ(ZgV>SKQR7Rm-B7Tn~uzRdY)gr;ow);^ZEy4e?uR0XE>F;x9y`%pt!8u zDbs>UX6$Tv2-aZhN7jjN_~FsUSk$}{?svGcG68s4|MrC>Z&24!ul$dAoOltB9p_p<#mMn^*g5v!UKzm!F{h#|>+_ykW}hoL#vcEpiytdWK8XIyj$xV{_IDP##g7 zmr|fh-;YMQwSiVM>#TGB4RiDMa%x@5y$+OgDVVDJ@#pK6Hi}s3`45J%q`-NU~ z0Q*axE}5PL9UC{!<5~%h)wwt^V+~wicYkV*8)&>$_htQ{|CRR69uwfnuxAwwRbWl} z?hVOhz$BlrXTf`~`*qb!(AyZ>S;oMe&sUd=?+ts#o~)7nhWNL&G}z^$fwr>x zbzO2_?Qipt1O6$sVSe@obwUheyXQ`@JONKn*JjGeVE4UtqDlk7WbII#ZromV2rQ}a!i>ej1P7Y@OKQX9r56opEgibdFtf}cE(rliI|bXxos zV=r+Y|F_c()(zHXLt%9xYxqj|G3duG8wx<>>s97e>;i$|f#Lr1M4SWxfZMX|(g|>& zIKhC8G0@E|1Y!zqcyy*aCdMIbetGb5I1!KIM*R&fje#Jt+v~O+jsJ+rg@~=j$bF;neEKWPJ?PBk~Tmy98dJ zHz`x{GT6=!`@%Z`V$N-7e%b&}To%UepA5D`j(PbE1ou&Oe<=DX*sAlzcrODfzkI0d zcRZPGy=5M6*#udG?=0WIu2vwtEjV2&T|LhuP=4(+M|a})m)9^k;Z%AoB!wS4c}oFFj}LFYc?GP;Eb*3> z;4z?K_u@~%&9Ua*!pA`VomQz@^4^5|t@pHx0dheQe)}I%Kl6A?L&)isjm7fXiX9YZ z)OfKl{rxNFU(FhAU1|LYj6Dop|Nr{6&bz0#>@XL%tc09i*)#sm zc*sK)IkJ~i!WR(7lOPC!@L^I_RCbb10~!MV;~SWR6i+!=&}iesgg`jZ?@WAp0mxe_ zV5_D-6j&A9R`V%r$v(6&^%%^Rt(}taJ=97oXLTq5>O-n*sXaKx*tyB;Ln;}VUmGCz zI^bGS8451#9M>wI|IpG98(6ev)4_xuu;$F}Q^^OwwTqKNF$gN0DcaihBRKTMIs1(H zFhjL+caIluX{hG60Gg4clZ%vw27yUu`q7Og70Idg54TzqUQXzOKAhGuGCrUTo7!xYdgy>{k$9-yYJtXaumWMgRGiBxM+8?T~}$6K&1;&9zmXZS|H zVprTj2oDd*YtsVU?lAH5#W^uGAonN3LTf!JF`&R)?Rs#Fceks_+_Wm;7W8uh{!*F)-_3l}@Q1evTy!Kx~yJZLi+ozuilzs_^!5AAfM^c*Y{gIh6aEwJ==IbN@ij%O~!m z;X$LQx%z0>x9HU9^mDJ$)$CIChJDFj?za~v-I|xZ1t^UZ4p6&Rxo_;dfPK1dmwd;o z?aE9k)K<3$3iZxAGC%{i-L%)`yCEtnF3vEN=L6(ze*1alZNkO{cGTFxjh+vBY!b1mQo=&q|kgyGGz&s`KEg<-oCFsmzcVh*$5?UD*M) z^-j&dI2f|$T$S$hg>&D9F7T-b-lI$G3W@;NVBg&iYe5&|*jD@cjas9vt*y7gFg&I;16EOnulJGTYt#G)4`Eo`wPDXo$%h@V)Omc=&nVlc6!7EaCZ$({N~H z4ExFM8wlO#_Kms-NbNZx2FCw*zRew1DZhez_dK3zn!x3`QMGc?q4e#5Wx6F`dsuy1 zwi@iL1@Hy5pAB?qxF3=Y$CuuYfcWBjysor@huO2-c60`tx;~#cdx1lEF?sMzuqCX3 z_b5=-bK0oe2eLfsH09FQeiv+Q+-aHwhF!_SvuZ+SwR@*yydh0~vCP9&kh>?+oOBZ8 zdu7S#&H$?LAGgV7R?#oBw1Jyn#*N7B4uvjw*RWX+ zHp7*RB`@G4+qfp$|CeYADON}eAaK5O1GO{UTb(dYuY>AU!sgqp2k9fO@!AYyvXbi3 zid*2iyn5*fz`k{%F9XU!+`65|FD-|Z%BMW$Fut$|wI?*5p_PLBfB6iW(;^{*&kPShbI{+i(kAtb=dz(r!44HN7om89oiPEG#;D|l{t{t0Y#g51GlQ2oGSH$k(`@1jc#V2_e_ z7#P3rl5kPK3~Uc;+RJ|g&j`Egirs%&xUFwnAqL0=;cwuW;Ih{y1LBEVmm$I-{7X4G zyT)jrUnrSqE#e^vf*^btew@GSZt{1q@64$t>2=`Z)7zhC8-Tp*2GC8g=_G3eMLLB& z)INYfZ|^GV(GZpqFiTuq~?9%eMo!{D7xT`uVkC zerMGo=S3<*WhY6S2w>JxcUV%&jZ$*={3Bk~W;S~U?(J;7%l+6G_EfZx2t{V^r@kMp-b83^ zRL0ix6oWW*@LDYTQ`KE>ag=wAE3vue^NAe9Jx`$UNoIk3C;a1B)YBGB%Lv8g&Gq@) zR#xSyg&kdYZ13M3hx#>j?oX~iu*n63^!1|#bjfNBA2a~t-JM-UvLT5N!n;? zM_3b(l1IpCso3GQ@v_yV!F^n&GNmQ zefguNqv13OEPQ;P*;2o{OJK#F4=Be~juO*{FI_|*Z2dIPLduQj8}8jLs0iuVsL=Cf z)arUkcv!(S&^P0o`CdCdZIB8Zkcq7RSkF z7s!C9!(B}dIz^(>G1$nAM@~02!7jP;XP(ENXP%cXxlfat;5YRt4G)#%~Z=v67|-My~k_%Cq_ zuk}0j79(s+f0glJ_CIA_0WS>3lr6|5c-Ridy^NR8ik+S(hA63A%GcEd2&MVYDbF_1 zf0M(kK0Uygl*++|(<(b=d61tiD!XeKF=rDl9dqT^7l|5V{L*~JwJRpvR<_qvF11Hl z`xVKWYJ)f|;>XKAOL!7T30~w*_0LyOTZH1VRv~iK({fflVw1LUYn#~(jU)4^pU7sy zTJ$)~j@+lV2PtjpAHmnh+PRiTQav#(01Yl!a}LTEjExL38h}{v3;hdB3s|+tH$7V` zaGhh#ami=M70yvz%e^F1Hew5_byUohg!A*S za}Na3=lzPz9KFyH@v#(2WN7Uy?>ZW!caAcJq}2ei`gPS`T2I>Mq&*ovOlsFi{P{>mj+)4$&MG8-N|5~CRj zr*-u5;>;+qj4=0#PfsKi57E-mU-ePxdBp^{{m6@51*S>WI-=2zv}k zavs93g&YyjT=ebAP5%lzFp6mzschKjrquRGS0O*$1mJk)?wH*`#6CC8QdJ4B}nx5$j z7qK@gAlm_7`-AGK*NE+XuzDanKNj(zIiztYo%yegVUb~WYlYY~W^MOy_i2(qoWQpW z+cM+{6LP`*NKp!=iw}Nqm{lr+x05BD!Z27X{r;iFB1?-?%!#@I%^#g`I>XyzRja~L zwd~_xb@&p?i-yuwsf50OP#X^BuXk3j7pLxgR3PT8&l3816o=YZ!{mfL%{_A{tS-yK z$bh4|!4F@c30yZl1^Uns&9ljK0SFXfl7=$^0R%L#8siQw}4qdhfH-9kzEc`?QpIPq0=bupQ zF}pSHC!|Z;>&smR>1&mmm6{*$-%TWBe?K5WND>L|G#GUslO8S%o?*VepZfKs0lh4h z%KmA+!0D581%ct00d&=M7nV#9i3z0xkR=t{}ask)XH zOg~S%y6SqsoJTsXYat|0&Rs^Y<7J$@Va&P)4FmT`3$E9w{M-(fMK8I0ZL zKE@#Y%sX#~5AELO3VK*?s0Q*OjLz zk8fMNhRe{cO0?6c8{7P+x+PTeFua)$qidBBSP%N&YMod-cQ5TPlAddNB$_|m7+%BM zb{sJ6nRq!FAS1p8rL~n=j~o+wv8up+ldu-QCw;RoooGjI^gysH2-v9(7S))Wx4uA$ zh&|)5LxGvs9vw?>k$ro9F7{3ETkeEsM>%lkR)WP$T*n{$)x-MsUtuu@l|Hd$LlYyk zhqYyb4!+vgz256K)<{FAMQ1Rvv0o%vHDt1}qTV;wAJNViI$c0CqE!!Bvut*3m1T9y2gsSyB>% z2IL0#pgN?r>~v5($-2Fasxq)+L1KOB+slW7w+W1u`G%AZ%KYt)IM{MMlz=TSG>%^- zdvv^L{f=RRr;}9jQ$g1=C2eWk6K3zul&~-YtNc=+AO~*D+wLNj6EY9jB}>d8uIL=G zn5}sIQ$!m|k}J~Z8#1T0DzMh#mJAN4jCRdoFPhN(av2dPm zA#+t{n6Z1^1aybLF(&tXs2dYEy~YC!4JUVgz@H!S_l;IEnTGYBhtp9OK7^!%@t?DR zdsWO18EcT~$XJ-i{syF*X>!LzhR0TNW$C>e7Zx(5_dh1b`6G!$}bHANsW1(*n zX7OBv&$E^ONu_<>gFo(cHQ;CMr$xHh%Reo2#aC}NLT%kq-5_v0!wzJP15fQa*unFu zMZcw#8?D&n{9qoq(aIClLKiJd%}uTj8b>e-vV|%o#}6;67G1LeKI`PaMs9$%Jn&3liyU29*XD#ZOTxafL6zM^}A>F%Cn`zq_$O58^ckH4ySA*;wv z9e0TNG}KEQkh3RX?(&fJLLhpxB@`(PjUIIM8{`xIG9)Y{+VDjH!9bq+cG<=X&>+>3 z5JTJ^1R*;7RX$C{#7h3c#iIyW_2O}(&##2e@_KvHHqxOe!0&XbW%O4XLM^ILWl-Mv ziJ=_h^p%Iub`){~T6u%Szfsi~h~7WORyg?X{)I30>-Eb}Mp#o^bysWltGjkc1MiU| zj?f;9C^o-0-w5L_DcaDx*4Q>G(K*UG)|8(?J$%#aietlLUsq#2#DE|31Gx`HI3iPi z(k`8g+cqNHf2sRKc5VCf5!eOuIDxq!fU)<~LFn=it}pfhvS|1IQe@xrHc;{hYS&*q z8XPAgnpTF8xc$t6+K-}o*3MpPhgxIV7)yp~d!dnI4j!Wh$gXdbFAu3A4IhONuk|s@ zZ5qf=HLXFw8R#^Lc&%0RnrAz2SDKNNkqLIPZ#5v50SgLsG7aBOea_QO*9=#q?W$-u z97fRu2D`RV41)R1B`PNl)3kSF(6`zGT4US#G{Gk>z#$|tWIlY!B1z)7zlBj?gKN|02K~WbO zA$fN!AVB|7ZN0@YD$?{mm=f`PC7~0oe)V9&Q4ooA;R5`5ANymB>|he+5a-GC3@JUc zVS6^iE$xsyO!0j8OB%6(yQ^yYn9ZQ*CBi3oVW@YHnL+UI$Y$W1RkA@3TFJRz0L%}hX7H(FLkBh>B)+LqU zUh4DcrByKYrtmY`i-%{kMs|o?TZ@t@YoNN@kmU{blK$IcXJeW2m%Jmf4zb>WO?QK* zK_)1~V}$*@+rrDDcI#T}kw0l$ypFt<6$ur8o^&5K_heAoUfAZ&HA~#0ci22uTTb97 zumq@^--(Sc`%#sg$eED*k;XX()vVAhm8lRt!3xSXIg?CuJNMnrjSUP;9-%nrZ{~Ky zyvs*Fz?J!zkIvU<67Wc$ch;2VZ(f;NoXYqJ@bx~{i@CErE(_o*=OAqI#mMf{=5d>S z1-${DntizE>3=9i>#7`E#T{jGP23{s6}YD2iBZ~j;rKN~WKf7%$7YQd`zQ-@dzPGs zSzZ{9RkuG&Eo*J9DT2SN;N(=xYTF(m(%L_I%)G&ZNL_^Ns;5yRIMemoVR9r-GZ~&? zrYj)=z?&G@J>gq zBI%mY#rEHI1l)<>9zCI zAWN3)GgUW*%ca^Ijwps#$%G3O4FwK(t~z@86E)YZve=-ZN!wChOj9N#m30^6I~Hwc zA1uG~OD_e@e2EnMqOhSmBQc(7cC!qDD@y&UH}4iEl=bnIHA>0V{zYxqh$^JSK_~ZO zn97G#Do1Jc{ARHw_d&}(G0CL2DRWme6nQ9Hua^1XDf?^{>ZRC=Km$KSI2V9(eC0fI zPr%k{ap+<&OI_t-zFd0cSk`xjx-79h6S``;i8#pVAEJ?VcO@3rqOH5 zy-Lc$IeZ|$>rmFB6JX-VeF;Am#4G;8BNNwahQAj9VpSp0+51*jrCR;A|Ej*M**{qg zW4m^zkR5lubsa5Hy7Sk8)nn8PzUGL`=yBcNtBbJzp=kkv*F#M4g6H+#y-$JFrhz#^ zB5P3!C)sjn^Uh>q!t_uHn#D53FC^E4!+^GX3jtkRt7348E~uKbMk7+Ly=p++{o&L0 z^e{V2RbZ$J|Et77XwDVjkl1}D&_nG@{>5`BpvlN*e8P5BRc{#O-T1c5$ibq3x`Qk7 zr(0ZfdZ|HO8pPdGKMWWO#=ggmr*@k@g4H(0n18pJ`y-TbxCH&ZQf^TV495_G7Wt7M zX*nq(Te{b-txtIGJE=#r-R&XH*M+Y)?Kd?mAabj4YWs*(h%jR&gDU2w2dsECsKn#Y zeR;mfRUi*an%|bmO}Yt#g>rr7=w^Lcr=cv;9r+7c)kFr*WE;%A!BzLCvf!J^3($0U!Pt%l%!*x=rgh|(;uz3M^AKNyOPZ; zt<$2Nww$stN-)gm)4X?P_Q_rS>{ieK(V&JUj&@UVB3v8aP<;w^9K=|;blhT#;EYp! zdPf5Q5T-y>$6%|un!14|+&;HZB6E%v_jcA4kqgzj!X!T`nfFEG*3O_>9s_V9BPg8E zKagSEk*Keo@jhvE?6VRy%FW`t(%SKl0mVVoZvKt{1I6sq+sIqYd{M_s`AqoQ8K3^! zExgwGR~7qec(RO@^Ax49Adyf@PNuxUvz24TmOkD;d-mjz8^EjF)mzb>1Bf-r0>fo{ zr*$>EULx;%56n{&dH&s8JFOl@uyNK)t1H}0A_q8K;d1;8f7h5!{Pg52;jXl25?Pm+ zAwCN3f{xF1DmZ>!T)*~(f9dLF6XH+!>hMt`>2@w)NWJR-DG8mqa>lIJZ?Yu(R>Oci zm4%1=eyj;=ZM-G+9WS(bW+KHO^;;3c#4}=+)`q*@>&(^Tj7L>iJ0ekUDZ6&nZl6nh zq@q{+;RQn%^8^Iq?kvMHS}n275!x_t(a#7@1}_581WwL)J4#p@3pJdsTteS_11HO3 z=|4$}i0_kniJt)n?yDBN<<^l>!BVUf=}?20g9k!HREr&H2)Y7&qo?5-9CV^vrm4** z)mW{wgG%#=N2Bo58UAEGq|u7a2%Xidi|PbntCYi4KrYerjzD}@O2@8qzm&KXAN)1C zl@zr`1TZBR|FZ1#Mo;oZiz2N?=GOC{G$%eOYFavhnNHNKA?w#YH<6cTZ4c$7=r~8_ zo0MaAysT|6g;~Py&krde!i$kP5IRDE(_W+iuhg%c{~0>JG$0MtsZwP~E)ZFwbI39D zM?e&T+-?P0&O%Hn02hgY+D7D7LO^bL2aHR)R;=IJ;%; zl4UIDFt4KINxnEZys9AlSgv=i%%*xsxA#7)xy2XOb~IyIm=o5Oo4c%DhbE^#Tt;|d zxI~vH?wbIA3M=F-^}+LE_q-&cw{6CPReuS299HMl)*5~jrCZCD0B za7O9xi62~8iK{Uh?~M#strio1{o7d_IpyV%exXMaeCFszy!_Hr;~QSRrki+-T2d8D zX#p?VF;Cl-&9x;kSl4ykyE>aK=vDpJWTRZIH%g9;{eE<>;t)(vTt~n#UmpjyBCOMC zv2_5@V@zzrfFgNxp)P_^`^@od9XIjD$eW_+1%V9hUrT)jU5n08-XV}R`m_FiC?2R~ z+btCSLdj`Bp6`@}0$<~!F(VH_oO5tJpq*v;UMiR#_Z^SqB8L9r@yk2=ZqY0mq~QVZ zd*}uB!;x&4k(TUg!y|LdW0eP?F;1}?xBzM zsnqYa2VqI*P5qozPZa|;9(Epff7+c7?^XOw)ASklPZD=KUWBX50b(W@Hhl2ltaG&S z@2Cspm46XQvtvfDG88-d!sFwn`+f{ctr%^N?MLFiNhYpuziz-G+V(*UIhU#2850xafP*aTR)pezU8dYty-1A@Y3O+fE$r`YQN;$R9O%Zi{ot zd;^QR0(4#!Vf$fMq`|ViCxf^=aKi%;a=WYWe)rZ-xAm|o&-%fRCCW_-zklXZy=3b! z>a0KW!p+bXw#d#7?9Rq^`_Ab7k$g{|+6j`ZG^yK$)N=(1H&wy5bPUap-i2r8*KN0j zL)x9cd#Gh8)Utx_TZj!|4cX!vl#ARf?K?8BqD*t}-!DZJBI*m^;8s}TEf$sC? z&&)k42;2?HW$$)gUDro>WB-CR^*^Mo|C?0ZX8n*!H3LB49wvN-cq01wM7L7B-G|+fuw7UO=pW9pyUXUFK=;5I%F3DzlqrpZPg@7u$Wj zU|Bh=A*dViX*cKos6L)$u7iMlpN=apwFz;~9AF-1fQ}vhbirjC7}r^yJoh9@Ks|4o zXSpa6VdzvsKkpG&I+eryuOw^g_NCa|v;hu|3cU}Vnr-(N>s$3$WRzQc1C79Cy*eQGiQa|b#f4&K1V5`;S5PjSayC(dnN37+aPvdEIUYjK z=ixVCvNFy3MR{|SWonPy(udfaj0&ob&&Oq$EiJ>b^DW+xu~$z~=VpM&UJMsOb7Md) z;<*B6q@>T=^a~UJh9)^82+xKZWbhM^(Ei8qcAzTus&x4$^B*?_CGJ@#y5g~P(!&3PnS{4@659j&%T*E8qSqur$ zO{^A??tnkoN#RI(c?FN2SXi3y!LEW0Q%hv1*7GUOClQaxivwTM{-(bE7J6f8*3{^< zg?h|!&C1V$Z80C>-_Jq>0c7i+UYQ@N5S%s-Nko0bn=}qE0u!^()F*xmK`e80PZc7@ z@>;2wn6rcF))OO%qbTkDem7a+qTlp6Z?uo`yVK<~Ku5G3P*I?lLWSagTng5gccT5b)t&`s1x8tqbzg8tCRG+mW z5`oF+lK+6Q05Th$=dyR7OWmn2&^llpdtCZI`w1`Ycr<;rwi<~ip47H^dG>7xBj(qc z=o@NJ9WKcUPC{QkKUkisHT4u4CfQ<_&{FA^`UCO9w!dWj;ev`IXU@VkO7-k`LbaRl zP?L2qlj4#TouX(ef&@;sEX8hAV^g^_QQ(K@11MLL0UPdN}93mZ_C8{D$nGJ^J1iv#5x zf2m=wzGb1cU`zLh5uof%q5M?SwBHXfe&M(ht;rFhI>c;-xEo>BwF9SJphph^NfF}h zkg<1SLm-|T;-@qMvnF;HHW+nG0t#3}MqE`3Il=3*78D~ZSsbMD`{Qcklv{Y<`ejwj zQLrS$1F9@sUE!MG?j0%czunxfnRNn|+^v@XaJm|NO*IVSC{nDLAP=3f81q%FPxw!a zC&~7hB1G#x0CbyjjLOaV>|c>$7P<`SZwM+KT&+B0KdGl@(Df~#0OA!3G<3zvhln#e zfkWPLGq!;v%?!{(7_}{%c89n8aXJhZwNE5-z|5NBbxou%tqA2qziN9NuGjB*^Om8& z&zs9xb?DXlKlu`FXdic$#GfdVZ>}0m+uo4&Y`JY~3&o(L4~a8XMg19FBpP>f2O-63 zzfbtHs=<$nn$&Ct+*$+BNs@c!Gs0QkU+rXVew@Ovgu9wQu}lFa<<*y1*G z-!7fhwH)d9m6xk8!?gevgF5LzERHV=!VJXMi(_)V0)-8D^^LW9LE&ALv0evI_fMy! z%YfdySdF1%1@~}Yvzsjl#_?b5g98E|PyX|vh480dz5$$pBg~q&6xhEecz+1zy~<6f zF`t;=y9S`zws8mCcKtj%KhKaY5}zuq(w=Yw0`4|Vh<}hn|4JWym^lzddPggjXy?y7 z;hp=8tioh(O~KIDcnDZLQlK<=RlvV&vZih;85nENW}TqR6<)GcgW3xspD9Y6w9LGN{WBM-B+CNVY_0#B`(>b_r!8}N zy&h#|(n4?|2d&$4_x?0>me!}xRlWw10e0qveKGJ`m5ll!AmFU0|FMn_j(yeiZp;c} z!&B4cHNe639To4N=!9U#`x8Sxb$8_X{&Lj}6F!Nhm4LO~ z$XosRV%A(s-|RN*-NQxa)w)H?pC3|hidpVe&o!p(HOb@gLwnT^`xv$IdMpP-)QnIM zku;Ebzg*|Lm|Oj9&&n|JL=ltJ%hI3}eX0tlA0Y75q-X9wLdX|s)ztFvs=J;8t4;X7 zLbX*^3F^HuFA+BTSPN{yR{fD2kslL`a|5p@9!(MjMZ=4_NZ)*&B|Jr4)vtf2vR)AR z@WI7s0S!v0Qx0c9+vMjLXxGg&1a0ep*~JJa;1p)0{6)%V7sdPE zWFEDcjhQ7Nl>3&`1iLfgsry{$NsK>he~;nu5WB(qAB&xb=nhP_zNk_Zx56nlm`+?{ z-q=1OV{ZKb-=3a&iE3|U8Y(7De<~2wyd}$9yTMcpTewvo3IT)Jcu&T&V`*i z(#J6iwOHLZfJxO*OxSf&e$aGK9tVqpA%|F-j0TydusEB{GAkKTCEC+l_kWr)Zq(vX zap2wZxUskdJLxu)IvK2dj}=oWOdnd{QZ3l&6w zIFDY6I>Ta>+9>y*SXjicrP0JuVsZvJAW8|$OF3?f723E9dDaSV=Li$$Fw9I6-x&$n zGLI3|>Q}_1M2GAozYJSBh)yGRH*?GN<+l{a>%ra@V@rPZ-1Lwx)xmeFwGy}! zSDEZek^@L~UQ>n+a)-f*z?-#Ct+edv1gxO!0_6m)yK&7Fc2xy_s!^6?Vb zM88U9j~YL5-Dv~u-GkM|o~5!Ng-Tb?3d2__=3hXRQ_pTih8|?BO)PciFlFg;r@>@I z{b1EiBDxe!~X**yTBYb9yeB?M?1f9GwKnlh5UU}49BbO zSpnoqNPP8uSJsVW{Z4)fcck*AP~rM)RJ)br)?;6^6)oodb3yVI%Fpx5Ua*aAs>TU7 z_gbrbcyoHG0zV)9EPcDkR1{%Kl$x3&gdM#+)CW$N{r%g{OK94LkXM`zi|QF5*r ztF2g9KsCRk5+35q0~bZZC4jD;^-O$Am(b%QxJ}e|LEBBlHhvib6k1Jw5Fo?9fa8Ch z4{UAOXtCU?PA}KK(D3Of4aGkhx};7przhpbdpcZDA}olU)(Ckqxw$~EKnE2pgxb0j zoUy#$qr1$h#huC$D8V`|)W*z9P-n$bgOouX;bUwbG zk5JVW?*{=4PBhq6$&;t+uq{DRAGr43_7m+eT3Et z`#yFo2n7gK)ieD7d!FPU0DP?1H7LUz4$ zgmVkBj%OEpKb#4h&X~EWYz|U&>@-(zljY@e>MXzJ8ajZ-?o-d0L<1S%Rd2=vL;VO? z^vi?Akwb+(bmj%dcA46HM3NVJ9_X>Him6oPcHBQzsBDfA!bkiAa{IUNdmdB2+pTSK9&5xdu0vl0 zHEE;VCxe=bxz0-Gm0*OG->%9NuUBWuS?Q15`@a+!VrfyZLP8jQ{c}~_Z_tni8T(k4ISbAk^b$wGkO=iMc`}8Q@D?D{NdpbLZ zxBm6t`cSefPne4fVK#Ecg3bxsGk5eh_tK?A*imP>t%k4Cvx*?2R$&oycSVyC(mneE zD-`$Kve>ad7ij*Z$uh$7<>&^v{C_oAf=cMrP>6(KR<_=RQ+VnvfZ_v=wcm6-bXJ4? zNr%g2kv;DLLt1EYkijUEzr?M<-PLWOw8ZbJg8 zPqmyNU_f$36QBDtlj?;v%(9~6H{2gU>eZAU&R*__{Ut5^3ea9USae!VGgdM$r;O+W zU$AwmZ}#5Doli*qbo&TJDp&Y(Z=H&FCwx?_^s$Ko-OXZOtfI?t0^pMWNnfyeeAw6! z^k8AfNwMgc^zh;aq(~~?t?QRerkbNlqGEakJ8(1ej4bLpsxQS{3&qIcSk<+`RD>pH z2Q&Y#e~JawTGPzIU@jgv_Q8V~r{to^QOOn}Yhq7lgvMqkTVO;JY#fx{ZG|Cx7!{*kIRhQ8>w@Fh%^8H3fZieD7HGu%ApNCUj?n%XJqhb<> zQJ+4esjQxGVGcVYkQP3wE7851|63*GVPkBU)-SBPejmHgH#^(b3geeo12|J5jv_wO zP8cfP0u3Vt+9<=qUV?$`;QO-il8Cg!KUwH0Fva#&3f5dI*ha#+wC9%TPjq$5{I z^Nni_l-;HIK;-*OzoHEa&}}yk+DGFxNSSQ<8lW$wO%8I7yZaI6mTAeeeG#8+l4VJb zLOkl7{=U8~kivuob=1Z=B9kd-9e0^_9t{Z^)ds-V^EG%-E|XI6tka2WS0GL(+3(Z* zyH_nkLKg+dmIq+q1)B;Ky5+T1yr?sxD306N?hp3?V!R{dG3@2+Y(_GJUY+jLy}W8K zO87pb!by5)Q^x#q#|;q)@@jm)&`(dBO~y97r;4>R0|?LqUVEM1Nv!^9n=qL>#~I72 zxx~7vTkL!ma_W#FTF)9RiGp2Yb&iOouBGmh;Cr{!Kf4g%oE?O?avz$^eTeW$s`r>U zcF|N)e4jeM>1~8=^lnxGu}WgwHoHW>-a=fNlCbE+=5q9+RDA*7lQNtO;uqB8K3xt7 zC}_H1ZAuA|c8RE{W8ZrRQObTO$%;z_c5YPb*49XEH!tlYaua_C{?pNgujOep&RMp*3YH zCB6oJmFa}^+Hz`6<2LRfaGc!?wTy(G4v4R+B3%_{owxuM5O$#|ETJ@}XvO#Eo_#%W zGel~W1T^x}*E(*UL=f!WS*tq)i!cK`$B92KPv*CCr_T@L(Llks@OxBDpZj zF5lR42_`B@zw;DT9kv}ML;51Y;VU1A)kI&_V#Qi$J9che4bx_r{;Wd~c-&~lXc$!8 zaKf?dOwT!Z7MFx3y6>@^JIk^Bn1xi@O{K)Ke$&H~4|Uw-?DPE~u&k4I^aPsECVH&u z12wa-miYz@@*Vqdnnmwy(wxCiLtGO>eNL00r9A_ls>#&7CU;_cz_5MCXawJu;Y?kv z^M_W7Me?|cUZ*OBa$5{(jL#qC8#|~vBlm(w5!EonBBX$KbJ<(XD_t$C3Q%ibUH|h; zk62ub%|sh>XjWHN*)zlafBQKWC)VL%?GW6L*2Vuw{pUYkCipNkDFSK*&++G+9Q$~j zeebpD6ZXJm+*o9KoN}hI{t&8AowLWB1Ms(ya*^Bd$pGB*eZ|xxwoMcO2;-c)9AGQmVoO zG8H4|I{IvqUhO1lmfA|s%xJ}glne}4P)=x2g4NMX2MY;YJ?m!AMRy4ylSLFzw1Fc$ zc#aTVl((Jpm6FLEa-etm1Y{lLM?X3HOOt#FK*A&VzHmxQW(^UN7<~35g9w6>M+elP zOs$(HdMe8=0L@FuU(h1*2Uud2~CA0LOBA=t46`M7~M{Da;en%^_WBy>~L3NT**VT#rr-@r5 zkurQWK8zKwXxmq>TE>gaHk!SH3t!M;(isxeO1ggr zX?cFH7TkLkzG#h0?nPpZIC}TY|KfaAa!qng^Z->lxrfhnXvJDN?f`ecN2fn0%%D`L z0?QF|Y`S;Tylr?+_JILIBj`G0$t{%fAje&L1wCAAGjik+g(Ue`>1i!_VFK9`b5BVT z#kjUzH}v6Y&c*})#S5bqUs(&P$49XsY0`q7f|g|al^XQX?Z?MQGrk2?&5&>o+(jat zfcaS9?d#d=R_yg-xT(MZaqTwDfv#52X(=9OY%Om)JX zBIwV(w#q?y9;?yGk8CKGfu^hLZ*g>s^J-0qFfsh&ssxB7F5LOoCV5RN+^P0Kwgq!J ztWchBHE)9${I8H|!MDHOBULz2fU6%@9VJX{vnEA%f`#T8hxGl_?=b8bYCNrk@t#AD z6t3v%b@gbM=SG_n+o<{8sTx4PEumRJ4-M)Z2#>Z1X9w<1DRtL8%YqzvGGVrd-K*D6 zVMbhn`R7@-(@++)wg_>~cZ#539TzEgXkQ>-AtO*+G+f@?KqrW<>1*bQaC`OF zB;k}|KXy+Xr^jr9(*+vj&L`wi%_uRTB9EWLV?zvr&T}hcjdP9D&p)LpDu13w2(q() z&DMl=Cz7ALhnwsSD}_Fb{jBj(Vhzm>Jrq5JYq3)6xi=EbP@s;z<Aso*7Ict16B} z-EHp9I_`^>BGGoS{*bBLKgh?5I4BgbwohJ9%o^CQl*RV+Ge$)L^Nh57;7h}>l<|Fl zdkD)J45zibYJ8oJMSMp&f%#RtfjdZ?h8{gmv3; z35Y>=s?J`E?R{)1B67OD4~~dI8%x z5?kx^Q)NX&6U5~7etbtGBB;=wsAr$=OCbyxL4H5i-5cP3^2iaFgE9I`XxA1yVR9F? z{ZthLA=~trcgmh8WF#pwxRu%Rwn@y?TJ`Iu15$AZIbx|#E~u>X>?_5Q zv0*t4fXbeFG|Ms?(7q+bEXoRD1G+oTX_kHcl5K{B(0qPqu{J2;`Tau=YKwt#5!Q(Z zpNOpR2-*#1-s(y8$+|pIu!C_DI8i*CSn31XsRbX9#u8PrX!>zWtd09jQ)G)C37q#@ zZZ)$5>zny%GTvbLzUd@$@}i`?MzPXo2JCRQKp^3Y+HuA>E%fjQ(_!=&L0)mxuA*~8$qYdwa;q(D_BD8Gv((iN>Tw;H!v zN+6?K`CQp|!u!tUE^o+)Dn3ILc9fh?ik=9vEE07H6i{R=+4LUCx98U*^%D)5$c?F= zykV`4dGoT72=7JHxl!7$x3owm1I4IJWZ|j-UsV)|Y{@4b?Piz`>4qx=)Q^Xg`sXDW zVnc*t)8bCwz7YD#hGQt?pe-A|@{sSX?6>Q?>UKgw1whGs6;mJo3%l8c&7%qoJ~Rs# z9GVo{{u>SPPq?=@t+=475gu;q@)p^y04U@m8XG=doIu{qArj|mc=MHbE$tc*tMn4L+Yhv^l4irbLyS}ovPFSgx+#d$}e)hqz=0s2! zUE^#&!+&omJ}K~v?aha@XT)FnYPCwq%3ZFLYm0ZG#E^=t=1d`cW+bUnEAQLQSQj7I zh+AY|W&c2F6nBmw0%cMd9laP+s}-*6>XXx4~!)OuS(zC96Fy4$LtREu36v29t!T?KP78Ru{5qfFurXpyu- zTuG?Z!%h|2jI@GH82V|p2>6#jIby*NW4Kd%D;ye#$oULiO>NYZH@3l1LePhm@rPQQ zV)kc03g^!dc7f$^gT{P)8%{qrtjG`P&w~juh+uLo<<9gVVP5@>e)6ot3^STXT3e z3%Guh)!?QgR-#;oP z^x@Q}KCUeGG}ahd*wC|3Mg&%@xnu~3>)AK&qPqlYkS69PVTxf#Mo=8eQlI2fX?+=hYU22_VkO5I8*mXlH;dr5y&bdnJ8ZLmkOg@?TDP~U?-$tBr*}|p zUhjqo1!B$qO&@Q~o|vROHb6#BQa8I|53Dq6Ft@%96=D~mWf>DZnSF`KfV=fneW_W7 zN!9Fp?R=(Ft8IQ_=7^0FoY2u$x2QD8Jg{sw3k_;x1}tfv)u-G2jA$Vi3BbnNCFyNgX@P~NYVj_VH6Pr z*8|vBgMO1FWiOI#1i&I`J_X`&mN?-L-e%Z7QmNEvG@LP{fz1uu@Gk zPczrClC`;--u}uV23dV*#9sN?c(&o^0SP~#Jn^f6skBDZIQvZK{1BW zN_OYM)OFb?x1UI_?i&}P9!Q&26JGRhsj0`A-z5{Fmu44MP7(%}LN|2k1K3Py4Xp)d zBTHH}@P>dX`-E!m%A?hC4}D^|vR@9QfV4)C z8dn(hxX#ej#9+iT@fT>skaTn)IR6#aOUL|%9f+toAH_ot3aYO*Mi>u}9=+ZyOpE3* z>m|WWGRuQwnI8H;_3`7a&KeZd^Nh4Jj$s(Col;u10g{25ONn>27sR>UzSRlKIF>Hl z>VsI9xGwV3Ko*vEg4TEJ+#O+McinlGc{+*gUv8-LZv?+9ia%?V27H*CmX#Mbl_AKC zlBE~#v&`I#@g5NBe(7ePvXfO%o;V782Z(;2zv9KyY_=x4{N?*C4^&-Q7cQcV}>S7+iMV z@9fz<_4k?6HQim^Rkv=aNcOk#DIv+8BSHwpKsj-SvhVC9bw9Px zG;KQStuJG#Uo52%i|gAt?#b4?XPYNge6d6q1FD@ZsDbQRDyGI^g()lK$UE#pi2GI$d=4dm^TIHif;F<;3ty)((MVXT0M- zu#!e&5CG|2tmp}3?6vebW5nImIEKwlepl9qNvZ`|P!|bH0jk;G@b1a7+1>*ONk!?0 zLP{rcFEU=a7%yQ4%A~F<6AN78qx=??o_EmH@gLN9#f$)!F(;zkuq%u)H=;icMOEZv zfX(EUzS0bx+Qi!#C-BTv;&*XEn13i^vXa{RirSJ(tkHm;tIQn4(6NHJF{)90E;UB5 z1H;ThgTc`4bfaIf9?8L_H&Fa`={qITP(7vdW97hb!D2N!4B>%LVL>Z5t*jNP0j(^| zy`k{=r+D}J8Dm%HaF>}RRk;4S)+E7D{VNya8>9u}RxWn?+7ZUmhs>f`s~4j@kA$`I z_P-h7%OGXafHl{QgHSti1&c6j37bBrLUGKXF2hbev(o90lU}?wXBnhx=9@_;_ctFt znj#tk{m$EtTb1sm=tau@7}6R{Dz%WS6J%-*^W3azKAGS>>Yuo3(Twr*m62Ar7FRcc zo(0;{N-c?@<(V0o3E+M)r8(Ezt@b43oQ9dSoP=e4 zC%hIqb6b&I-Ve=-30=WpLE!x)f$>W}gJ0l@eO#NPV`2|ASG)2gj~B^hq0GfvGt=R^ zZ$C?f<7bo7K1z0TwveZ6DD_m%omV^jh1I2=ry+{mYx;$4JZv?BzLr}7DL0NBFHi zr}xU2M$=vP5AUGvNlgPEX&-45l+oGI*+@f0ZQ80Nb-;QE*5wapuFm{lP|rayZ;IXl z7p%!Y;vHHFf<}y68*hu8y%EE85?-}Ca2dL}+|=HKn2WFLkomH05nKxSSNa;t`^9!& ztk|L~vha(7bYH2_i0FfWP_deZP~7*myZ=#3jEuV~0o6UdgxO1zC3&o!K%oZ`fm@?O&d-SUg#(uSe4Eam5D$i(+py152QF;rNqEAkp9k>i7!R)-|5V z4)Z#u*lvVnJKZ7yaR2ZY?^P?uUZ1N^Wxpw-1j@*E&rr4XzGv^#o9N{7y*+-?{B9?Y zbPWo({iH=R3WLHyn_18+S2|s4`~-MLjQO))<^D^6j(8`%9t{S@SkWDRyi6R1C0^Sl zkvL>{2lT4w8Z!>gOxL_dMNb-s@QL(PpmbfhIT>?M z>NC{F@Z@Xse#K8!jxtMBE(Qn9O>5+bXP~=JE!^`ZnPSreX07dF{iOzdZ`oW%88m`o zmF2=>vIMlvueXMq!}B!vh9x;iflJR#O47e%A(t_hh~hGMI_lhT<@g#_`gJIl2^{yU zd+tl1MV|`KLzvQgXRu20L~l`>jI)JRGbEf@pr4+oOPBhh3*4Kdw@2eo`#j4|3nT55 zbNgpSs1+3ECNdd;Ngmd+Fh%#R*3S5Om>gA*R*H9cz+%9*dYg6t<0{jJ%nS***zC@6 zhU6qv$S|KAv=vZCIS*-#DHrT>LV^?-(N<-Tz`CDEj6RV&ITp8WM3Oy=)%q|sFWHfm zgA+qu-L)vLgoRCIV?%=?`uHnGod5xvd0_}$70Z#|wONGWU*nRX##vaQUORN;tzm~= zo16{;AeIYF?J&SPBu3hbF5s3*4MUEyc0D(hn_#0jiOG{#8H2wKnzx@}k9x>IZtF+rcqF1I4&Fr}{{Eb~D= z&`Gv{L?*^jv$79ZXjE6HLMy*XLHeip3`p0BO-zvoSR2x7N_8ZGvE>a!cL;t%^GpaA zSYYkb0&i%0M%#Rs9uZc5U$V_^rOa6ccLX#G{Cj0t#Aw#RU#p2&J3a2!i)F?Z4*muL z1_mE?hEKDA^np|N6x7H_#sVhfR%vqx9`{rIx2}eOFcC~%;trurXuzq6jq2F<6ig|v zz4pXzlg=DvA=hh?x5uJ`=cjGv?`1ty_u|T4atc6i+ERsC>1lynvNfTc8(I{ngwXgF z=Vrs*cf&}IP=^UFh?B{7B~LVx#`

8-y%L(mT6WA35N_^3`Q;HII5*qy%v@4$w=| zzt#luZ2~5fsGZS;_70nWW0IwTRzI(d)m#tM6$dQ1pTat!n_v!5^~x=Ag~8g3afMwR z^0h&+JWP$&c-qz)G_l#4Lg|7#crQE=EiV_0v`;U4dpCRa)sRG)LMxHI5zJG!Q8MFr zg!X3bxqs{(oqX=-T?AGCITv+=T#2_nRxOZi+B9}x`ddMb2(Z_9!{&|lK!=DX%K%$6 zJR&qBLV8109py<6*FNxA9wv6a`oyc{CY!;-cDheZvbS^V7U27I1O|K8br(lEJYqUqRW|TwB>jC9CEl_?rveBwLfZKON|?IUViP`a0~qk zC59dMx7H-nppNo|*i)BzTOyaJVsLwGIbA0~67T316F-_XsRa<_OVe?$97HQf8N$G0 zp8T?Cs1miN@OEXVTwx`c&z zx<$h)={>Dg&n-2EGX!b|E6!e0=KjoS)ggvchvX+@Kw!P!$(FdM^2fgLxn+gdbyRnY zm|R7!=!^1RBg^U$2h5|*i(mEOU8ETr)+P~Lo;;$L4pjt0xH{^fqnri>M}Ov`v;Anu zPHEDxn$mT=8t&7R)bBm0&WLqz57G$oR(P=sYQ??|Kt9}xy*-EBv{&TM4R)BBi=XB( z0*K0#nreD@J`F4Ct6sKY_VXk{&gFEoD}kPD03+79f{gDM>qT&Edb9XUwJA(g|ebAwC%lnV*?cfm~4qK8WTw? z8zNfGb0${~0!_ZNraZWX9sjOz94iptn617XV;c@xhv^sp;emSZ( z<1#x-?YwpG$RN4)8rgUx{ln>KfR>ZlrwwPtNwYM2ja7-y0bgK%-@|lXyJSD3OIYvb z0Ywt?M^JlBTk^H6<->cb35ek*x-zKQePM+$r6eJkFxJ;M`J16AuzRe6v={Od72X?g%fFFtf?7s7 z=ZKY!1{D{-VE#k#_N7wsj1zH6A4{FA)uIm*7@?&+~e>ci7x8C@LJp1wzmkUJ*BYNJ>X*#<3HyJ29<% zFrJPQsR0O&AXC=kb->Rc&$!Kq_w7_es0xkRkbh-`;*(gAEakwx_VhI6G`3~GoFNgR zv=zMY!SA%MQbuoOvxWRHhG46J0#jp+F)|1W(5h6DH#@A%$bf_$HQH4U6PH7ah#krZ zd@+NX~ z7+fDvL#@vhqs13cyt64dZlk9bpe+hQIkREFn&w#eBhWylNJ;D5Oa1~cpexJD@EI%y zxgn7@Ldux5d#=A-T_g$7pERJN`IK^?jP?Jj7 zS8&p!MxRB!d0yS0EwTJGk)Dr>q6bN9r2WHYmydOIwA1xbmNiv&BiAgz33K zFfnGdIrk)+Ln#S^u(Pmw(8N7;O}a|78bm_!zmVu~`xr}-*S!0k->+`ehJ0M&D;&NL z3k+86cC0lRSoqCeoD8MRjs2z1u@LEnd8_!Oa1&f?7$@NNEWY|Wd)&PPN3{i)ZkEjUuQUmcLI9M6vnx+>DJ+<~ZyP<0P!-m*P&; z*7n{^pAYR`XX0(}*bhYkMI4RuKe~`veyy7wSZA29?d9V(blEX#)d9;D{rgfHj+k?^ z_vu_@g97(2I^G97@?r0Lrw4}`_TFFLNF_Pp58^!r8Cp4s|7;PmQj=ySW7kV(eXin@ zUQF67*BsXzPf~zS5a%TrNaiEi*3;s9*8IM`DywnAH+gO?ExIF1cd-^i5R;QTVJZ#_qm}h{kN;`ro#`ZV*AHnV+25^pN3uWan-KHg z$K9L@j9qS853aBtt?X(rn(X1H96Y9oMjJTEl6_027ECwVh?uFvqfg=CvRnM zeR+7K5ZnIyaWK|LKoSS7iE<$Mi^Iyjg$|l4-MoOd3Ps!hhBW>2E`Jk0_*}=>0PY{L zawxJ_6351dRDd}OKKkcCK}kbN)AUcMB_|b;71P*4X7fMYbz@1?DbbP@|AH3&*;1{8 zJ#K$o z;=<@`0cLhtzsV@f<1Fs~{9obvsHVJ^ zw0owwqu0QxU4+lJC$z!XW4B^!{v3W{!xl$hUd{C!Gc(ySIa#Gnq0Sw6QLJRP%%!~2 zz$Z((WOn}}X&pFu%e?i-6{=7cW)ESD7zn>Cb)hH-5wk#Q7<|Rvy*#*e86TmVV&f@KANji{Ogj_L=p)$G(wtd^6k}f2D5FNr=KS zU?UT{l;I*VoEo8oJ&X=YsUg$tn4`EorS_smpl0y5VUA;5>)4De4R!aFk}a;4k;z-qAxHQr|GJM)nl{GbqHMOv z1Pis0^_cs{)@b~1A%6@l>zNhdLrvJ=#?)>iN%){)&=CHO-Q%0eH}BireF**efwU&G zNkF=S6#I4u^oZsG^1I?&dJ~**?|sZzqg+L{C4Jo7Je_O^a1;st~=dDZ~To;dQeFTY6i=!%X&2TQ5_T`{jDb(w8eM0lA1vfv33jww5 zYCY|opkV}XxYucLNh~qGZ@RP)M8nL1;EK~Wf@O-D)fsFks&e- z#CnWg8PDAH^)!NxzzPQk(ax=@3uKHBWk-o~?X^Tr8nw2B>dhy#>3E5~+~R(6z)NJU zYh$ur)R&?GdbDMgEEuP#^RcZ;HRx;2USYo-GhA7A3w)cEt?VIh7glxVUN=3@HSqay z-L||qGi1g1#ZpT`sg0gH?mjbX>7}RQBSKwARQ+Fn2EtSjX-e4EZ(nHUYT=sGxhKZ! z{!{A8=&UbqSQwf$NG~1=gDznj@`OumhE0VmO=i z3<6|g){Aa@{$9IIWLYQ#IOKL#_x=9!4W!5XeYt0^0<@+3dXY_pXCceXS6A&ZSG3j4 z^wAm-lYw8MezIVu*uGwc+UAqFc@KJPse~!J`qkZNZ(kOC`$JDPNBiob{ zGN9LsHVC)8E$uy0sr7pcK$fHumi>9>pPH|Dt^}@2Zc?<=&+CHs9n%zC=A_~OM;yDy zjqUDD{hgYMOr8l5{VySV4IeyUux;Nl3)2bN3Hf?%doLofXo@%l&&GSP^A&?Gt_+aB zBTl1IK7N$o;K%&bYm?yH19~*NW<`FwQxweHT%qnggq(oiZ3`Nm}`m8fYF!+qIne(F|T`O}1I+1VdM~NcA z6l!i($&8ha#msXKRN}gMjqJ0TFGp0sfr(s-@Z{X&w}yQLJE9G5JzU$5$S3|h^i&A< z)r^1poAJx!YM(I^z60kC2S3^>Q~OiGPThsr@{YrfG!Ef>wPl_&m1x z=W62glSMIKppI$xNl4(w+cTa_)oi;w&C#GW^@(h*3s`%MCAm5(% z99}dox17g)VHwK;d|kVBw5uB?&*nQCo=MSHAM+B|_>96g9S<8XA4!ixR)}Hzvr14} zAuDn zu8&Q}HADwILAQd3Ve)=Am)E8hrWJciEW92r%@J+6a?PB+j9(0it&OeEkiMP9Sn}C_ zt>kU@+)wayEZynzk+km}5ibGcp!-*kRwwc`1>KHK4e0VCZ!i2M`gw<_{A)s1@C4}E zS^Oijj^FvXlYpDio%ehh>Bc+9B6R5!Qr8zxS_T4Lx>i+3UdnSRh|A-$s3Mn)0H6EY zCNE8Wjl!e4J^2EOCXRlvmK3c3ef|A|%OI%dmB0HL1JDm2dQt9X7``1>UUT)^Ug9DY zRh6%Ac!R};D>ODo^MfqfjgSedfGnEj6=kibuA}>X@ItHmUC3AY4OR|7aQxBcy=&9? zo`%NNX0JBHIx>?!vWZ_)xeix7?Y?2SaH_@v;juo{MmDL6b^+t z0S_%dH8(Nj(RdjGWb8c1*qU8Xcz!sYEFG7Q9cmI!V;9w4Eys(|;Gle;lFm|0WLV=D z$s7bwcO2q6J<1?sEY8{z#*GloMuo9Qt*7?m;BUV_Lk0r+ z$1n<=vOrwqsqqWjo_nJ27;!|{I{e>KUYX3Y)uj=}7Mu6aKKg`ezYh*##<7=#!>5 zIm67WyKQ-q{{**hRIpub9pEIgnFO3287P36ojSMOmTr;Q_2zCTYX{>sseBD=S-CS! z&i@+Jh(cgm=y!|l^*22mE<*nlccsE0-SD0;NR;E7*e&)tEb=yQVugul=d32AOD$3) zb&bx9cbJG4(cZfBf6K#&2`RGf66kn z&uLRvkqj_u+=lL#b4-(|)}Ri2)AMrZ$eTX%yycC_fIZ?%Cn*5%yHvuB?u;Hp8c#bK z74U*%1CEL24yhL79a*=5f3Y0x4|K^TovujsZMH!KnNEYp3Zq_YyNe?ou6@?UYDV;8 zG_}>aW`eN2L^wEo#;pU`QCwGBS3&BS8ol?!=ge+Bp(A?KLIgH&Z8g9&V5rz{QSrqu zvi?=)&CP3Vu$of{C*IsOe7l#e>w54t(%s$%YuMS%9_?pZIIXW&fDs$&+<`t^^z_mI z6jdli>WV~;g+B5O5zc)^#eX!$UswUYNDuZ7Mge2W_d$ASb0*Uaq? zQ4i~gVdP#-XA&!`L8lJYReF(9OgnSy_q_eTXB}IH+q{QjMFJSj8Z<^QYkYlMpD$>B zF_38p6zdP(Fh^g6Yq7Nif7t5*06XGSFw8^>e1CJ;nJ6$rn%(gxZSmHu7icr zf6096R)7}{F|yL}k@kB5zZvz?#^PNS`18R)bDVF5*!}U2PLD1@Hs{ZeQpc2Xh|5)d zImA!L?`5%MaluG&*xp$TUm`L|;=zRj$27UdB=V$h>v&e)vuxWUC`2(Vf{FsU<2qLO_}Df^ITd zXb~mS6)U_KtRnuZ^w$>W6H3Yt(Q%ebOr_l1R2#AbemUL zzh$7QYTq(;^^(v*y8};>!MdXiJzlHw)mB-ShVc1s?QgEwH089#wD$fNmG;rqcS3%? zgdjE_Q(R1*hC>cUAn{cZN$;531M_BNU=$U!$U7tXvudi@l>4<*oyGx|C-DZ;I zuFr07%{C9+TJgJqyV^tcEXM1@t77ybrpLo?N=toJNo&)Yzc1~ZgXJO+#F=4Qte_kX zJp9J{-5jeIwJK2Qh8MrBMcG!7168Nb;`MvU?cb*mT>K!<^n&93r{Ti2ahoX1$%!Z3 zEnlvy32vPc4`XkD2=^~%oNMPn)8ctc{ahl|{iJgn;19u6Sv8gu>-mLYG=^KPgQm`u zr&bhZn%CCp@53781{3@*TQ_?$m^CD!JCl_BQQ0srbN4FeJ8XYI)tlX>Eo`a+I$+x( z$*sV#oOdW;2mmlV3OS?x2Ly!s&FpWcFjn&fE~;iot|{MO`PJZoFKO+4Q8}tx^j7WI zJ5SyX7qSSifD17fcs_lSryJXYp1;sy*yE}pH|=8xuA{ZS)Qg-hBCzFK64eU{Hq zqi+z2U!re0eA+&H))QOg3Nu#KdUWqBE0W0Igl&K9jm<_?kk>)F&M>wXM-PU<8`~`V zfr6q><6d$?GfbMwHgNmWm~3z1`=Y?Lxe2U74Q57t-k$)3V_>2pX<_EUzF~IXz$KDR z$T0wfl`=-5n=Fc|8i_D`8q5dsi?He_mW8$uk~UOoyl>MLo8>l^i>r)!KsU8HqP-9E zyIWqpL1r4Alei(lw5)ph2ExHMjsyNgTfEE)#(kvI?F+qUCY60&_fOr_-V~5ZOi}SK zST9rnm)3XEYo*r_nbxVsO()MPp2BsR@vPjC%%Pj2FWLw(6>jdNF4*>@yx;>xg4V@X z8n5?0$)aaecnoADAb14we3$yV3{M(|GC_gj2nVR|kO#mqWBygKHN5 z5<|D#Xvsn#^;B=Z@aWDINeH!`dmUtaW822w&YA0X8m;o-x7Ysc&a_S$F@_S>F@z=jYh#-&Q`pLRoW|YrbF}f&o)Mvp3ki1H|;#c-K)94%Tu# z7rWGXOTaPf1ckT7V|chY3vhHx>)XNKIX;D~95C!5GSg|6s^i|s??k0j`ZiS6l&{@;A40Wl#lF|O=)5$78l7MM_6Z6AvWo3|J@#Hd*q z`>6T?%*-$HUBo4QbIA(M{C+6#G{S=9 z&C0e#%PdI8gtL^Yh^pi&LMOjoIqT$hKDXX)t=4c+3sWvdt^yb-oIGap3(>LOzCMeO zS*Q^EEIksMGdk*miowH;lEPm_TPl}S~9S*?~^)5@o4*0)F8flcopRVy7(E{9< zp?L#FkMN$9wZE@Vk=()qV==r<_aXJ)PTRQRCQ@ruPA8AeMQ87VFLyAy{rJ>Unk$#f zKJ76o?*KhGRAueht760lUuzvea8Hc>|H8S)mQDP+*4qy8-kMwpSm+#X=I6hgc*fRU z%mE$(jYrhRtwC_8d_+WRYh*odGA0-;T_Q}7yLQamY_(inyFY}P*;4bk8}GxRuhY74g6ddv%W6ogeQkuG+4Q5m zl`$AnRv4Xt^eJ0s-GpR_%X6mJ4J$pOJ5$#}gOnoIFAJgbcBqq2{5?}02XFfd%nH3) zIcy4trn1akRq9QH?!8w2k{$eAdFv10MUD@=_XNF(DziCx)ZA*x5^$Gt-${CwO7kQZ zV7ljFjP;VK=3e@V_K6JAV`U_tg9AOLZ@IrQSSGv@wYeu5kIOOuDiLUII<35MFgYId9?Kst`ImP>1nbMX%9Vci?^=q( z=T=59qc+gZzD=x1r>p2laJJ*l=JiI20NdzxeV-NYq^+L9*|IT65 zAzeMR3!_VOJD_&>KIK$K$NcP}#W@PK`U*ZI?XPTJ07fD&-Rfd+o4oTMhGnRX8wcH% zDlP`w4nt6a)7IX}Jgp?vdpg+$Wji1m=Zv{YX#YwEiR{xRuQwC}x|X)j!jI)fzqw&M zD#P{#mkKF2nnHT^I$_e`i?j(g9hFk=u1N6^EGq*D*4k|g_Z{`XU?ULC)n%GqC9daQ)QA<$$_lBrg1sAOa7P~0%7A)4Lw z(tZ`-(CGbn@Is>U5^Mn4fdZ*kRYz%7?gHra?$qzM%!o3_4aX5Wr_)LTAUd#@kU36( z4=(PH?v#5TW+r26!^1PFVxz*oQIjuxNI`x*fXcFI^^zTav)sY(WITy0o;co;pa(Yd ziu(+n+w6e_3}(*s$cYz8X^a(zsqkmm16;JW0oogugbySUOIxd z{`ArI2QlZG?33A*kic!;W>5Xoxh~kL6w%rB8#5ub+PD7<+j=~N(|}%2-vPZ*XS7># z2+rGeEq`vtv*G}lwr|ugofcZN!ZfeTj%%isLfgbv{{XnG^yidoA}j{ z)$HTE3cj~l^ZvK|LZ0F&ir|9#U)tuC*Nh$+Wmooft99gh2N%5@8R$x5)}g-F;sx4_ zQAKO9s!{YDk?J*gJ{%AtKCJjYFV#OG2iHbl&CWJo;8s(2L8mpx8_A)P1!X~{(N3F` z${>F*U(iM46%%TN-sg=3^}ahdwgUxTB)6<{&edlN1|boF>;2N4t;t+Ix=XHj!VGPW z1ET{n^O&?ZNJ81s`QOA5-x9^zk?q6zkm%{Rs+e@WptX^fY_AWb zyLp4A4&4*wtBg)Zqw^5jlp$fa3|ny$bAP2&2pg9vwJ>|$H=1^x*ZFifODo=+`^iz7 z$C6-mwEY-O@u}=pRWeV%#pZWjI7@E~T^QEw zRf|cgC6V-dav-PsYAeSFhO$fvt_GJsQRvd)W;r}%adM?rBNToJLClX)c+60K{F~BG zNy92T8Tox!Dl}B?wn&@E3xK$2Um%EUj^sx+pkx`!zK;A$k*D5$F@dd zGoVw_p|g7tYBdenCaU=~^&3V^9vEq)X*K9Uo~B+&R!6^c$;dr@G~43guKNly)1%Oeg8FVL|s(A*8af*R@2uv z%%g^Pd~RYFg^#0OPj$S8`DYND_C<|AzsjfZzBwom+3Zm58WGhd%>Vn6F^6i@P{*9(zP3@mn9baH-}eFmoNnkEfnis;Tg zx1B9BaQb!B{jfN&!Q&$2X?=U7mB>7WUr6w9oV4aKi1j?bxG(v0*y829|Dv*&=q(cO z9^s~qwS(j42@k^*MXJ0;7!}C0G9_R7B%2+n(Tkm#BP#&9b?&+9^0;DWaS$lHGs0b> zslbwDt+a9@eh^o~r1%h1O{`kEu-r8v{EH{&ktUB7_^!>MaSYDoHERV;6Zt!4oB4J~ zA!1t(!uwB&T^@72e_vWBiBxAG0bksm#Q?g;IY`1bDuTaE|9Jr zenB`)P2FPN3!HEB7b)QWP=~vIIc!y2qq6$DT6`Ea-zvc2M-GqR!?%I$A+NmIie45SBWtwX% znoXXhC<|Et(YXVjGjUeJ?H8$NID^&f6P&?ppTTe2u_5Azvo5SkI16ZHO zSG>AOs5eT2nXrrQZEFAfUgsy#aG``&-aY{*JocH1jJ5jZwqlh|;qqt`acCE$Eu7teaJv~ykMm14M1714txh}3OD z+I1&*v;B6EdcH}~0@BIpVVhzQ@{ zF_-lNPrxo@WfJTrU$FCR+9T17(ht&3_D2Y202mcQa6*RSh^7%@D-B+5Yq+*FX9w40 zDn#y1YQIr5ZO!D_xHGwgG3p(*M++f)9lSRK@2$D<~#9p1J9DYP0EQVqofyY4z!sC>o2YOYkBgkKO^NV8U*TCF8KK!+-EDNb^8N(W=T zrD+k#R&~+GB~f;9nhofj(P>)dcPuPrR?Sm?f0(=*oi_RQq>EI#7m7Wivv9I8+HDG% zyq^bJ+>n+bTX3BVHXGd!ktq}f1dQflX#Qy-F%cq6XE^h4uvDUPDPMmtcNXl*3O>7= zdbftE^kD}dZc}aweGxSkfPl$-R?K7$Dfe?}ziaMD4k^>6M0ad1ASuRupV*D-G?I#XNFed3~j z`b2VqhT^K94j)oj{^K)Y|Fn%uC4162SW}(2$BGJ*`X3pzT^DT{%>b>kKanI24U^g ztm@ggXTO2Ol*_)anL-px+tcK*C?6smw$s`(Dns4^FuT{ie%=%k^5M_2`W&^YZt>%w z1XC?6#x};F%daOnCWFz&0P2hU=;zU(t;gOjb$hN-<;u4De*MzMYN>8FSGGSj^+j5j z7{!Wv`jnU2IOIdjZ3rG}LBwOudbcP+>I;aCL;~jwGw{W)ThV~1;3nrlnu$4S!=FO<7o?? zKX{wU4>uKVjqkk2pQHjPd9XU|aibWz1Wmdt(w6RF7WzC3!~}-$4Dl`3hq*CT`8G~6 zV*PR^Q*dHHvU-IrVC%fd{QN`6QDlI(54kF$v9&!}YNX+lOaYVhV z417cG7sg0cf9W)QZC#TFnGUPJ*XE$hH)YSkr2Zz%dOIASP(%s#~Y4ftKqf?JjfGT z6AQCRlE+neE4-*ZKa(jT~Y) z1DTp&In?q(m+>JdZ1nD^rG0cQ)DR<%7$dD@6JnLLD&un8UjWTWG5(xe<9X_!#|j z%0KpAY`!+y&D|SC;VPP1{LTtN=@Oa(9nN=aB33LiVa0zV7pxuW*q;fu6m7)a;_7z@ z?SbL3vMH!(dY6CTx5EOEeJayTe;1orY~S;NlzGWY74Y zmxq0zLF}b2W_5@>CLcy0?(^-`eE)p^9sXYAgxY2oBfcruaCP*{-R+1&r~#%Q{7<3q zO+2>axBaJz!x*fO>}k5~;8IDFg*6#rgELJ0ESnB8J?dpTP@hJz$)|M;H`=zAt@`H6F(M$W4r%JSgjF<|n28H|s>gYr`w%$&Q{(U!-8AV+c zJ~xyD94x)W7g8h7?Q94M0A?>J*|@ESS~~JXz?b~HV9`{O8dd1 zk)G<`(J?U<8k#V4cq5?hcL=O1d#(!*Itja-0K2E>$h|dQHt=QMK9r(aTM?+=;?X`!?WhxXQh5v$E*MXPGv6;L+<5pYmsOHl@5+wb$)N6#Z8pCB;i&oYv$ef~~)L~R$ za$$l>T_n+cnSJEm9(vuoxfKTe_}-o-q4hCH-YOD++EaqtB}ZI*a!tG)fQD-dy*)6cZA@d?da} zdP-n=~%W7c?!xNw;Ry!-Yq zI}RM_3e{g6x`KY#{lP8C7)3~+hltzp_I!Rr9lt9aD;*0@{v>;|VhaDASf1O_1qltv zR;!;wXju4T)CTzcAL11D(Apt+aS0m}aNrfw|KePUuWM!aWPIgvI!7=y^p`Mf;hgIo z&7SlERd=c{L#`$RUNuOWNBV)56*isAy1RiD%Yb>4d;4yvrJ+Km`b(D*X69-8GLP2Z zv1jI%m(5_@Q|(F{Y!oxT6fm>vaZL`FXn4(1zY`{(;3oV*%0$q_-@N4&ESAW?R z-Up8HG2yiHq9N1}^4A#PIc}soM~&q-iURxl+to7EC&+r6&(}~cE>Oi@FfxAR;zA90+-!Ho;_%_RT7R9_ z*74GcF;isK8BpGr4pJb>0Wb${4?Hj@)!bcTz2;l}`J?(Y*NJv7tSUNadv3VND13F+Ir#wV-1WHhI}~#MltxC^PqvaJo-w z-;9lh6Z?WB5ulTuC@62wSzkM@tUieu{fzqUCdM2TGHKlC;Q&;AkCC#UwUz36zPNKgG?hzPRN@vqVmt8p ze)qxqNm!nv*Kk!~2X(IPVSUmw(IgSYzopaAHLzOCnux68`0KJ%uC!Va;jf|AN7+l*3pijmFpcI9;wHb8vOJurgCg$MlJW{z*%Ly5O?xM=RM_&PjCg;5SRgrT}_SkUM zm+$uSw(&>`CAqD|G?iBY;W66+UrR(6Y}H29ypc3iY8bj$=}ey11RN?l z-GxV1DY@`upbB6m5Tkj0fghN5q?=y4NGQ=Xsi_|@G=#JeF+-axDPex_A{`UpZ9o0t zmr17U&9XZ~4|L`xr>m8gc+a4AC~x0DH8vQkj?+VdUW2vwk>$H@O|Y@9=1-fdTdxO5 zj_SgPnXO=O%Z8aS#%}Vr;V~t3W4X5pl;T*wA9YUJ@{I}7y|9J#UpA+RhN19$X@vnL z9XGaY98Y~f00{4Q7k&15*FeJgOwFd$Jqp2w%Q;Zmn$~;};ldj6g*LlMs-PZZFFd9n z@%6q*91XRr&0|uN&10UXBnJP-E1>8`g8<>fvwx?;YND2rZ_OvVI;SzDs~GH5@^QH5 z9b1hZU_v|8c2zx~HG9k4ms-jkR6w58Skm}CoqRn?bAw(dpiqNdTshrE6n)~mYi}#) z2oCHD>hpkDo}M~uFA~{$2*b!1?F8XOKmD?|yEd!Nj*Et;e12=EV{^m4>3%^{)k%tI zuv9Y-ApbFJLZ&VSC z=#D^}LGwbN@%l2PvLU=7!?wSBE0da8OT~(#8d^SA?k}{`C%eILcX?z#5<8wUmYX7C zVD@Q?=;C9>djK}1Wrt^4bSA5K3rw!iynaX_p2o{u#2$=}T7lh-%@m@a>0!pno<8&F z>9ZC$TGV2Wg3rLVT*;3)UoiBMND(^WCvnefu< z`$tc3EkhVH@LifuMOSH0{X#YaAFR%kP6H8BG*H=)g&PEn6*MU0r2SM@wH0BQB&dvlo9-&(dV@22pxwp&+(3i4IYJseJA{%C(9sz0eR0PdNMwrc4Bua@29%2 z1g75aqZ_50tjBEx*fy1FDV6e;Siec1itzm8`T6u#a`_N)8I@dvD;m*8$r!~EOxR{! zUABs(eP*du7y+N;&Lq@1K1DFrvC`T!?13Zt5`2JegKW>sg}3As-(&Nh>>&xM?QqD& z^XzY+)AQ$D|MEgXKtpk~O?t=MPJ=yZK4M-bv zT4M4R9J=)Zzs|q%6Y~=TCW59zC@4kOME{-yUn4uby;7eT&G#K$Ur}98)Id*?8R+{} z`JJ2|=VljW7j^$;-Kgnce&-Y3P|8wiz&xW`$!gL;AKisWvZB4^3Vf0K-?X3-oh$>4 zfF@WFoO>_~Jqqy{7v=5CL+0<7$S?$N6d}+J;9tk>T*?2T>Mg*U{KEcmLO?)3rKLnA z1f@eHCLrD2CEe1^XcQ0>P?~`VNH?QHlRs2A08kG z+6cfL z>d@MA9O{X}!UL~Cg1n=hrbO#{zdlyu4?*?IUKX(ZOAz;AEC(C2kXlcEI_k*ls3JTI za0&^FefI1x7k^>nToskYPBEzA7665$`8{OpZ^X3r+^ktm5KXbThSj#zHY*F`A}Wq| z+@*WsZ?9m42VU744FiCDY-2Z3q20a*O^{%HSgy zjORDHT6r8ooBJ}@cK|Aiq>TeZ&4=FouQ*toN&J0&{%OuR=HOtb{riVe=a4%*82)|e8vX4>5xF=h5aJ>e0RGO_K3%}0 z>qj|JjTDP7U7N9nZ-PJbG41vJ(30vC^klE*bmx5Lujutt_oUDufxBE}l~?7roNJHX zpLB}nW|TrXwxGJxB|kSqIGf9CK5q0_nIi%YLLCX3-FI@2nh7IF31ee_cEf9mb$GnG z);0%SYajv>-wDer>ldA@9p5{^K95wAFSq#=%pc8-pT6&7Sz0F2%PXzy?n$^Pm8j5e zekVtFAkw(GC9PR3ztWZIk#|@wQE~eLs1ANHz4`4J|Kf=n&&6@VH8>s; zxE(@_+&%)f74zbdH6!-=N*2aOE=3l}JI}A#{0C$d>s1pJgAbI55X!4Vmut*aEqYh_ z;W3VsOCkQPr&7eu56dkZd@trdRY+x9;}+ogz|A68#QiQ7Iv| z>Y^Lt8f9KKlasPa<*>_8mNvvFuYOfd1KnXJlW1%L;EhfgO#X`vy1LyHAG8pf@Jj$> zSAkxKw1?j2IWOLtCa3<@AHVIDuuyv(MMXw-c5-s&YgSi$N;c)wD&>|OZ}H~)%M8$p zRt)72H!XK9BH0NqIY|;2UVIEq$gp1zH#qI;m=I#u(TehmC<_=B;}P$!DczhuyjUMi zUwQ;v(mYct@%>XI<7DZSw|MvASzj!>3r;aMyFBBKz$L@&t1Iw*rWedNWr5dsb{9$w zg25h^(-4Qm36mtvP?5tuNf!5cgg;~>5^lNEVx+_2hvSp?S~G z9sYvzYouKnLBJr?5h~22tIov{X-~fd4?6NY^%3(IJCN%ylUTDklz6JZx0REH$ecSx zKvOp|HZlZoa|t9?_UrEa3XEzeuY#jdataCxeJ(*j6hH_9RRH<&!XvV<*8w@-q5G@8 zYR)V4miRn?cHSGDVI#>k^>VrVMW; zl(DqGa&+!L{&XqK-D zX06f@Sd$*Pp=GuuIwD`e66bu?T0?MWcUFV@Ea5hetL-Oa9RiZv=UchJ zIqY`S-XPaCMCy@@Izek}ZZI$a?UXBX8^r$3#l^SV+Ma=hN7k^7q>_UbB?>d4>m6uTzh>T4INj(%llUpsb# z*X8GLebMXa+ZX@M)VSLxvZH)6cRXG19>R=F^t|B3+=S>Cr&x3S6{DvXEruo-RPEUS zWb6wzW)0m{CR-JAv+!%VTFi@vF?J%~8i~12IBQTYY7!t5=cwF0jr*Q;;y!Hq9AuzIry-JXdpeNUgov|ToK%io`9nYauqB01hN$-}u-f40_uud8n(uxgF#P-p~Y*IS6 zp%eQwgYhggNoE>BQY(m_bEYk47#fB!!bUCswF0i(+dCsoSEYG=v%F98;!W{-E~MeG zzDzei?L4)8G-P!#s`r2){#PKzr$NLZ#vy6fm_KHmt>)lEVmzB@J~&`~=O+{G+Hg~y z8*SqoFvCigx4=yxy!y2MwEp-GkG<*F(VoH(tnC>5rQoT8f)B14!Qj59>_x@Qr&5=% z{{^#k+*@@Vi(2xM2e=%R!(NH1d~^9{Z`Kk9U>z@jZzlAyMB}8>>PnWk_}giP3lW>D z^xR&DG`!mqUp%tea&q|hs6Qi)RC1=`9M}*N0#OM6-i{?nx?Dk1SHsPNk%Ha0@7yGQ z;oqqWtA!|7e;cgL`hipZM7oB;d6khg_$lG+BWel%rC0q>t`01264AznT|Pbfs9?gd z8|yW-sN9WaSD#K7CXE+H9~ue2cDj+KH{P9pTGnHl$l60wIvEP{b9sSfo* z1uK7VZuv2-UYX@ds7&*sxxMM(@IeMbsPzotY9XfLfqNbNS8CQii+mE-7I&O1t6z-8 zK9t3i``RyhV1v8jqolpcGJpT`J}4T|aVG!I+v;zt3@Z-rpRQfXa8}oMeQBdaqmS&X zEIo|bl3f1IX)RUH+sK?a(zWVQiQc@#ty5d0TX{fs)(;l+dx2}+vl1#lrGJ5*8(wTo zTmRaZxg~oy+FSlH=N^J#Su|*#vOLCww{h_X0V`f>KlDoIz^7F?kcx@!=Igc0+GyM2 z{hjuvUDQ_JlVV(NqQs|eT=o&Yr_?!YAl_)K zMG}PtP$eA|F0=ajxERGfJ=?aP{^O(?vKt=;~V%KQ96x8S=R)rL+Jf4*-3z<_O~KmIKd zWbDX3ndpB0`%lVLCgovdStKH;Px5;bU97*CC-UO9o?z1@JtE_aUHAUN*Fl2?u83zE zU6(3Nz?H55)bwrqGx~qFH|E#26fEm$lHotryuabTX1{O00VFe3ZvN)G$ooGruq!*u z;}i{5&;AYB2cXUmeSm4`L5Pat|GFsw!;ull2xEe{#VxI^7FEv`TF zSVnnESlG`~RR6IVS}A-8yWUzD{=xi6@lJQ8fZr{D6?W8pVE0tC2CWo^z>T`SuyqP= z@t$jIsT{W@y@t|1Pyx%>B1k}g(QWpBD!nu#bblR*v=Vejjh`hxWBWI!e*PX7faO_z zjO3x?7)8aG0dmjiCr3cF>yP!dmSy3EKRC-vEgECj#{bjl zMXVIVyeV5PV-f!g+vGkM$0>SGKJ~`eA|AxYb_^ka1veUlfHh1$}(?#Pn z{fVUXjPZ=IE$@G!y!{Bn|5lmU|66dhrl$TAm&ycuPOkHlCm*-49={T<#8`rBfUr>G zf86mQi39eZxV*nX`D4ClzUW>ABLq5+E#sgYtX*T>{CGg_?w5dxsqe*qreopmqllJz zyyPqnVI?%R8)MVLg|At4leLx0DnqIESLKIk|I_TWN{Xb2joA1!-bEwN zf1>GcLumE?^Hu|`DfM6xvV{mUh1Rpi2ac30=*Wok>FD?5^=v}+>Fj-SAIg6&L>I+p zKI^Vf0L!$KN^@6z6VDVTjAylrf2k67DS9EYucR(x!xm>7=h76p(I`#!O^NxxF$q@x zKgVc|y!l?AXkH<}#FHN%;hjyJ<}G}qvZ`M!@h=oVh*8pg+b4OXbMe|tRt(FKN)0u5 z4&XZdXjp6b3gz~_8`(9$mG9IYv8w3)ObF#FXWRyEFQOuo!5iPUkJ^kasiqpN9bapWqFSDh|kf1OP|f4xN?3I|yR~ z$l-^bH;BusjFQoOuhh-Q*k)|x_G^^V>MNcR{ZCg^{!{T-n_A(O=&1{obTx0GaHUX^ z?LQImmv#RpZc!wBNmcx@b?d2!2mfgn)=eN+(d=vH9dz-nM*9Hi^9mi~;={po&^Uj- zwG36;z)kNnOU>TPNjdu@<+u)#a0UD%LX>pb-4&J=z~PIi)+om(!;l=1dF45>ScjpM zOjuapGRt4YBP3a+_xAks{Ce-A-z9f%n|P05Qq+5^bz8Y=?B+}r0)xR-$6%mA!S*(c zckbR$C+`?!2z%p2kVWp^wT;Nw`7I>P9(nrrD|7^VvpX+rNU*&fH5E_<0$_A=I|1&K za8=&9o3q<-A5K)@MNL3a+5sW}jc&UR$>leD=k$9Kh{k?>ow^_a1TrB*a#>m`IkK>X z=nu}_6KmrgJ7@g+AAy}?aKv0b1e#$!sxbnW1*OQ`rcExDt#kD_ae4jpE=_j6#=%N} z|0d_}w3okq-H)+LV=*`PcO@!ktT{u!=2a|DG340O?!A@mp)g_@B4p?h5t%pMSh<)Z zs3@y*e9XAyJH55_HDS-=)8>l}n4Jq^|H667-FF}zCOQW0z&thdaBZH+Cr7LCR1 zBwYKA9GkaVFzGv_8JM-=KPi$^jdw}s$Lvm7Z)^1S?Dx;S6jAjDn~HNC`*-x=`%7s;<*U3(mNaj>!0g1RsP+u06l zjKQjQGCDARXF@AAAsSSN+^w**$Rxoc4l<2Lxv3{;iAX5H7->Lu4dncGHOqeSY;O{r z)`2=l6{_%R1YWPE*$1K~3+DhKo&Ph{Q%MTQQ>C=f+toDatcywP@w82StoPm#TYJae z(H;h_>g@h+_KU_e+{dt=P{E36p&dU!fKKdNh)XW-Vw(ee-$x$1R)8(%2eN_9lT|FR ztqkAx>-O^l9pnX9T>#;Wkg3<>fC5tuLWxmK{h3L?YKtf=e;*Z)8xm-D8r?(zZEoE^ zp1Fav0mdlH)b(F`ZB-nMFN`Z$O$maaSu|BI@xFBe#;%k+gf}3_bBVS+%_D0cscd_s zCv0Tu>g?KiOmL?-?PW<%=Pfqe8il2x$7}-I^2&~UOE$hU!@|~~$~J?arYG}= z$L~0)cGrJ)q)dU+!zo#w9cz^05snh76MnR`an@Ztk?*~YN+jNXXns)4u!1vFRcfI( zA}y+iCF%|FUV9aPvYo79*%4D|etj&_2;pYpR6Kvi&UUPEh)0u4(>|PL(_uQ$lWg*= zb7tK}V$2`xfucy`Hd~HwDm-oPp8j3`anG3Vy~ulg_RuENj1e7d%WC2mE1tF@8k^0% zt!igBsRzzj_S;5k=EArQ?BJ?s)@HfMx+E;44^_~W(G@UeUBh=&M&O+-Oue4rxzO2kz@q3?hBK+`i=$wjO#=z^a617Cw?exGgj8g>D_nz*SlEt{o%zQN5}36 zc6Li)fC9O!?7*hV4Kw$@d8mz1_5EDzmoP-D`&T-A7+ZRK4m8aE(O)Wj}33Z8#uHrFp2t~!ZnJNc=s2d$2J1U7capQajX2H=V~YO zy~XF;T=6PA|LNs_x7f&Wef%4}guNLNk72F!4wi5j7S5S|+WoY+@3X1SlwD=Y-$Nm+ z0s&0{Sf85^p$zxx?TO|YR0RGt1B@+p9P@8WD+fN&dbdeY!TP^S_x7V`7M>l!Dq+$l z^HZ;A+5Et0HInk=|CS!MF~0;+jT^`9)MTVp_bZg_VKvxwSLVL2NX3GdKB9Gv7T$2)Guqh#=(w%fgzNkm2Coo7w&?&?QyeIj~r< z|6j*Ibr{SlJDNQce-S_PMzhlC`it#IIF)5rn!(uTl=VAIF{u+Jo5}7Grz!|U%U zA9AX!r(1_O0-nt&B5{A!D(I6p4x}HynvmWOCcMY+lbtU23OaaGxneM7dO zH2huWq}(-WLdjz5Uj4f@&tVUcNpU&Q^oj3VKVlu(U)gWLR~amw9SiSu0PruLn_fH@ zuo_BBOIy^h!zLt&IEz8HHDB&64V(&@xTV$R#R=_}mfmbz%<*h@90mCTTp*$It|%`> zhm3j=DXJIcszF9Hu^+RD+%kixUXZ5K)|7I9@G-tY!XGBVvNy9LI!wBjLL=`RJO-*b ziu-LR3V*FwA(K@4J~`&{w;S|a^2;W(QXDs&sdy#BYK(qG!OmsN<~&{gS+Qs}Hs`%q zh{yy*T3g*)cJY3n{~7ErY_}MM$Q~yq{+&T@Wdb*WhKZywu!CSftpZr9ZoP(=((xO~ ztebthtcRx#ZFLIt25cGbeufy4)zO89o|cffUa0BN4OO`>Mwb zOad8qN!629I$w*@iD+6~QG649C-2?&Ugxey!ueNCiv#&|4XLk1%S4kM_(g=eV6~kY=%`Iv3M$IQ1yP5Nfp~_}bDIR7KMCcjR#{9A zPC^9>J7Y_~#FCSV_$QtFwg3u^SH26z6VfnqDCqx{0vnxLV-^6@jguR(zn;$j7$4m= zbdpAD{xWJUSGqx_C~WV=lu?nz2@1g0PO`!yO7zZp6Y^Uk!rmmLojDD+GdH~093?$= zxXlCJ92FiglmZnLo8&iJzFzu0TQ1w{Ie~gNCdV#on9?80`ia#_=ey2rGa~J&t>%87 zjfG-*&Ms9a@ib3n#eRGq)A#ltFDcfi3Xsvc5pC+tgS4i&Ax_?=09AsJ&;vSwa&9Wc z{*WL26Mo>VKrKnp20eMPC%Fn^_jte5tUy$FvL?A4$;*K)hy?+nd406zM$7Y)_gP~( zL5m4z?e3^^&*C(M>`f3$o#9l*j+Le4Z-LI2^gXF5C8Ro!?6;~Mzb_O#KRx!`JD{E4 z5c_e0fQ_ku)u2^bAg!w321>S+UN)&uJLDo>R|d$3gjyt}HX}E`<-%ZdK+yd3RT@99 z07NQuEHC+q`T}MpFst+vMf$bfR*IAm$l7=AY_fHTd-X)|%h@`z?{#l+Z<{}u5u+=f zB*uUcXo2)W)1`8GLR$TtB`bk&v1hXyi(mL&SflxRjypEN-Jb=`API@Me?m-%&{eclqZF*bxy-EO<+WR@s(aPm5kK`s#&pPu<-FiQYrK^?R8O`j zJDU3l#{3F$3Il-9`xpDNN3f7dFbJFTD4Cz(sa^0bbgu6kpbiS@Gq3tHrv~BA<#+>V z-21hnvMx0AeSxUHXS zcs;_q6{<*ws>hD zAF3&sA8gtlM87>x9rOaS16uEmyP>G!%)35Z{Bp$p%i0Unj410GiJ@ROFCmPK_cnVa zPoO{r6*)X0te>s2XnTW}EF=wzKTAQ69imv<`AsNHL?y_)?bzpr!`em18PB%>? z*_x45anLI9I76vY+k>QQEK#M3o6eN;@F6WLET;nj11H@dth4L$7qY>zjn9?X~&0bMNigEytxo;T49 zTJ1MdOg9U)PS356M+v6*RYgyYR0cBVkbfLd3V68c#4>`MyL1Axa4DpcRGo78?Jt$M zEQ$?B?iIFaV>K zpWy4(*6&a<;5S~Rh~MchD$UlK>^37FaY=Nf=gRg-(bCAqB-@KFG&!v_{$LRiCt=t8>p)^%aGM{+8+y#G@X|k%fn}H|alQ3QO+g3M6 zJ@DC-iQK8}!97ntKQ;%Q@7-PQxML7co?$P7C!+EOv%=hYfkH#G^hX$lh+ z$-{eM$0`@CN$qDX;1b6tZZ~CU*M<4lCF6P04aXx-yMK1Bzt}u<1W^)M*v6?L?BAE9fP!+=O zX7m;?E2RZ!L;T!JD9R_B(+&R^{H1E=AAr%n|9{vIpgL%&QtF2GO=6?iFGe-#2@`%3 z{+pqSD(Qg8EJDhB)3-J0gYxeku8n+IV&jj&zeW$kX{PgK?VDP62M)8WA3JK-I&M%{ zT3P9LPm;%d8ius!;v5RSwrh9fVPp*KwVlK5Umcf{P|$gQz2H%9a(9^-&Ue1(Yq$@# zz0jD+OE*Hp3kM_HGt{VKx$m34aB;KrJ$b%0w>H2(&wts!O@`|BbA7Ww+Zs;v##K@K z8HHjtsT3uuSE@NjD@oV>;>WcAhv&>oQl-y^A#KLu+xCi0U#9udtf`({y5-BR{#sL= za5(dOtL_oE;&2+Bo)yj+b)yr zXDI;RHKzbxfJg(H2wQER&=>)`?U9|Wpu-#&a6!*QYRk5fEr>gb0pdrPiw zpgck>yVpAp5aVGN{>1-@h%!MXqvPVI^R|PnMzr6eA~)YX&94G3#NS+pT`|y{iR*Cx z539^Uf=;4WRA*TyTYIemg+b08&!btrgWn#nk$!HG^3jT8y2EmBjPezE#NDSj{#Q8~ z=I-v}rjxoeZ)FYNOm~k%?`LxaTdZl#JH1zIup_^>G!fG6zc&6s5bnntDU2-`(r80r zXkxTl?-41()~-yOLymQl7HUSmygl;{IVm4qdk6IwEoSrQG>Z@T(~*_*F712qFNPQV zeft%AMGXo3?bpX=G}gmzhUmNBIy2I0jWhWrt)$->2J>QPN%ZmvsX&$Bc~9^Tq#bv} zLKi1(KQVS?|JSh$M>#kO6LJ)cN(z#H;f23VxGlL&^P%iMjLI74qtbvErB`?DNd*(| zn2_3znxe#*+0E5|*SCVcoOrc8bbzs%$KJ{^hpAMO8{y0ZYzjvEZhO&MIzxIwoq0{u z+NbcrS2C(7PByG7JBu35q0QSoaBcmIuCwm}&&%#O2-mIVCBC$Lz>pq7ZVn+HT%RM7 zNqO=;GOO$B1HT;*CJOl~yGCtL>JXR+5km0f6zNr z?NTjzjDH6*y`GNBIY?r6Q;pVnjyz&u5iquTM;Oh%{oKNZH2b0>7kV`ocFF>!4&p*P z4gc2YT(!IOl24HpR7#=!dGEznnB2ygDtv;X(`Sw%C(A^mKG(rVC)DVliFw+btWf?x z=Q>2UbvJ2*kT%@8?c(kpt^F+}sN;h3AeUk(>5%}{;YdYi1F$T(!BhXl zNfW=wfX*1WPIdBnXO`W9xG8wu_n_`6tAo$HC~kXK;U{YY@X+{8K+n>$h3YqdiD(u9zmQQsvKgirW7uX!&gRqyF{;}^T3EbCd{pYA^GEiP(c&{v zmDk2nFiHxxM5Va8@=?B?acQjUGfB!!yykb)@s|7GK*_`qN?}gWX0+kO{#4=M-8Z*J`#jV5@9JNn zGO*Aio!IoQCgG4AyUaIVKWk?>=R}KKA7~&i*IRF}C|rXnccAdkjEUSHdj-_PeWJ#+ zilKSYaHT&)9hvr-{Hmbb4pU?xOP0O9>t79;^z|rnW%N)DlEb_#c&lvLsDZ_S6JgWy ztPqq7@l#*KNp>4S-1dOQcU|81z|G}JFS%YcLR6fo8eXB$4(8Fu_!o2Y9jA2=**DuI zbh=-s98?}ko~iWuA6ZfNiJIbV8ed4hg3|{Zv|ps-&Oksz7VWEcNH6xqC&P=Btmx0^ zQ4g>vvgXneAcn01fi0lfo+Zjb{w3;8n24^pJc9K15{ZF@TrL%P!hh~BPry)HaeD`#O@H9X}rm~q}KdjF2 zasd=_$zh zMQWo8ScaNnQ<1Iz&R1EFefU_)UIscJX*s*{EbAKa`Idd###8b!p!o!iIDkAaH*y#w z(_B+JCW!7UR$C@>7rCQ>Ep&rufG@MU0iX?0E<;D;ugG!czKOkV+Jl9GtrdDbGrwj& z*%#3~3+%qJW%(7JS!@Bjq7K#|Nbu*s+dFf6-TgV)a&FS@rz{;+P73VJK;n3a+<_O$ zE&OAvQrMogB`&i&3{u+|AX~=##`O_iTlNJqS!2CEW8txaU+mrDN=uOy>~N1KTzDq< zro*Ukzr3aFV7V2);{(5leFukAyg)q&AhfvMgkRS=0xhMZhv%TMhxQLJFJ|w2C|q8e zBKxFwo!~3IRkWK0(gvmVR*?nQ!vGz2FTK_XRb9l^ZkpYXL|dE>D;2#yB%mRTlFr+Q zd)s#o-@!2nkxXy4*imQUVzRBBMr0Fp@6?_)&>y6oaG$7TXosQm_x;6sa0^nv(tTR~s2fP)1x(WB}OYCiD+a$VH|buwoRa&Ar`=g3xb z#b3O8OAsVP=250>XV zii)xG%ZjEu4`ct(YB1wwHNNl$0C}uK%Gw*h6y@a0-r#oVm4ztf=?TMsX-y=+Cz*fGt)E{=E**noilf_$iRNuotBi^QTAzqN9z%UE z$o~F}gB6VrJDMA{A!=5C-TM!;s_jK|oWoovwa_0c@0jHV9&z}KZ0;38$+|W^{b7jT zu+TtJ-%z8dlJaj0H==){-!s?nu}$I2!Qu-K)OKzvL`NpGh!B;tO%3{12=78g3$*X0g$74LQaWS=F3vn_1PYD+t=J}WgY5@na-%;XS6>si>M zI~YRuuBdN20m)3IMzxSCzZN=U`?a`(;oTX~Lkb(s+2M#Gyi&smlBi_>i*gvIb)|Gi zl$@ajQf+PapltlJ>{#N8n>mmU&M|YK#O;p9i5b({)jJ&1ocH!O*MHtmUYt$zeEW>B zsZ=3|kJK`M%2DtVTViF#GTv@>Dg1Ky2k$dp6{A!avM3ZR;0iNUHjdM^k!!>=6q_+7 zSXSSut4&mNOLj3({?n14@Jr?U_u~s5tv51(ewDIXOB)M%pASmnv!GgI&)Z5pqB?+lpy-R$)Hd=0L*v1dm-ebJ)tZ^`%0+5+@NYjFB19PlYN|3tJPZ`dCUz z-7;H|;{MX`u))!N5|25lgZj!ir>3L5ee=Q4+8A%_PbsUi4TJ8&uL>#vFgj%Ppt9nBK${|#?WJCw zL#oE>J7mOuJk+XwCv^2= z@@T+=ZwRzUVMC_YW7+WRy1JC?oQ$R; zqXX_4E^6ZfY~IZ?!Z_*AKh+Xd??$cptY<}<);4l?6jZy?tBmW;#yG8Z02v|tydlhm z(w+}ya<>M77S|R+Aq9cRJNHKJm7U9BD~P%t-Nz&7Yrfz*yR#9?{xYE|y&)tx{P()Y z_4=PJ(Qly(gx^1DaBF0kj${J;d&ka3spdtYsfbItwSwRs4;T}b23K3&c}jV3`^uwz zs{451aG4VTKDusIEkKM$gmxa_?kfK3X0zeMn!0x5GIwq^;3Ni7tY$oWIKL=~Kok#$ zZp?HQsGPcB=-TVnPs5#^S^0ryDR;zQ(Rh@spbh7J(I-g^|zFn`vh!moGZ+|48{^ zKKb> zi{^PTR;e5kIjH^Nf;yvXSNBN=BXGQ>=@KVMX|Q+>I=1MCS)JMuFSkU-Q!7E1dd{3o zC`Rjs2dzI=M1LBT7pMCrz1j6}=|gH{pVLgiw7GY3D2TqM!>X@MdBL3RE^HN85IVPq86 z;$EMdhpJTCyC&vmdI#|j$#>}8`6Uq3m``aINhciDgAU0{#I?wg1Dy_rp5R)fo*f`a z2-!P@(^nqqZ@CilNi-B^#QH*HyrErLKr^?u7UW?Afgr0b4Ni+&d z@x@&(SDwI52EENgb6mJ;Vd3h!Kd$6JZ7)U<(c_>_{?2tC_>H5t3sC<*n!~s z;r<=VcW&N;HiC2z;Y`Of{HW}Nt4lM2SM4`WUke&)H>QKMwzViNG}8yNxY)bk58UIf z+#;F$l4enRciL*n_fKyXIBUw@>t?)h2&~^(6xr<7{4>7%I8u)qoMbwWKdZlzQSDFq z5BandMB#jwry~&z*Na}dW_GMHh{(t?aHr!Y0=*k^fjxgi z!?E-j9uv$9FXuYxNz*~|7J`<1YLE9dA4IsnWei%PZe*0G_h=_jrK|T}drx#%l{8}m zLPN3$)!;EDEd9zW+BsU8-G|^g)~dL&b(qa-{BHJKb6EROyR;7Ju_NRa0DsyL1YW^4 zKaiZIs(OKh^aaicolFZ7FQ-W`1q3~No33C$9E2ZW%GlE68|nVKQIJXd&Li`3sl4bV zi>fhD11vY5Bhg(9pMAC0K<#Dz@y=k$C*a6yx2h-+4Z_NZHi2S;Yx|zfd|q?OE1rX< z40y&t(28rT<;i(DDoEM5lWq-vCtgj|Bs+R7V;yiu$?BqzDUH^Qa7Y)Lg8>bLA7wOJdoJ*c89^Cgl^>6Pblm48 z;a(b|m01@kG&PNrSzr)G`o9Wa(S&lwyJ(DmbbK@+v`6id)$#j$a_sXYKflmXTKUBy zr&?{tlOIwVAE)mPW!mvz+9B>~Vkp6qF+F8$;c@YY+gZ0SoJ#}pSBn|4Qc%F4HVCXZ zsJN^Um_aRi>ahGsZ~VsN{5kSEhRn#rTqGoz>zq1@iiDGDifUu&&*z75^r-+sCqXke zH76>!N*-3eaxajKX{zq{@m0~tofM>5*K+q+g2zo~6-sbvT?c;2nt`%h8e-cFB@1&< zx*GiLa^TwO*J4@2QD#AX%C$0oA+6szbQmkDYkw>WBr>>70tIIxrf!Bn7d@SeZem9Y z*uf4^mrC2sncdyIfkG>xY5TDKv5X6~SG3}adQZ94CmXK9rYkHzh-Byaj&7pRw_oTH z)|*)Vth&LLvZuWxeINUZ!Ze=PuS1_cSwHPiP=*|hAKFc!KegaXH}9RHVWPEGV?mVi z(k@e1q!SCrKh7(Ss+Bw_*6wvSQ*4QYr3i`+aX}yKd12;ZTmE`!TY8#p^7$8qX%X8# za6J9kwOI+!f`MPyscL~oues&a$bR0+9S2(jNW&D(Pb=e43o}49lU4grH-o0u^BdVc z7iN>jh6;S0K(gl|m_C;wupv#&3ckAY5ieA$%X{nr127$r7er7V?0esZ)Htd30B`#? zj;vCGICl&SIdO;QwdZr=pGgn!$RWRFPz-;6`ey2^fpO{Co7J|_Yq($7c<2{y*~Z9g zd>t=R7EUXjEq{82&)MImzqG|j#V%&tqsN)CAc9O25tL1D&A+`i=yQWDXZ)nm?&(jlTg)hD=X#EomGJQjxe z{^<=HOK~bKJt-e~N3R(pe+WER`dBEi$iYjMwP94G>t4NUKf8K1-Usb#cU+Z^w(F+q z`RoDb4Q`TR`2rfk*a0wg*xW=%NAUUXLrO@8u`hivcl(QdFUsjcVW$iK>Sjx^6%pb% z_0tzRlVs(GFM`W?Wd{`G^wTaPpR2_lW?`1F2P^Q$1_O=*iUZEOI{8PF-N%~kmJ7xW zwkJ;R+)f)Sda6LPyehw)>$QxEXKug%(ULD9DWd(F_E!V!`hJ~+%K7;X$=@|9B z9AEU2=^o_sS_W^XRpLO*d3`82WPh#x_)1*d<(Yp?H6*Qh=#hU~XdgM^HAL2^Q+LT* z@^Z4JoCzBXSv!-wz6M5IBywEWUUJ_Q&95pyba42&v(m=@Ls$={(vEYS+(t^$fU5*h zf^~X!xC(cmE3<>g?plD)!#$u}dtMN%);5q-1dX;{6{hM?{*)@px&!54t_~C;I>8sQ zt#Fqpg=Yx7IohvSq=Dg0X4UE^i{59U^5U`OE^p`8^Kg=kG6;aOh~jLRE9faEMlnQ= z%>q~N?N7{tpS}0nEsUwseP1w@;WTDlyJ4^}bz=O0T6cA7&3N5q+xS4-b)qC4F(h8v z|2=vE;!=>PnpJfXdC*;)Ba1rf)c?* zja3#RXiwl8pLQHRJ%tI2p2q{=HW?tsu;=-H*CvY_k|5?+gLX0bHB+?1QZ@d^Nd#QK zz_YGc4kmM2@-a4kv(0c(~T@@2OA0V zy$}`>5(?)wNIZhqMMSy*j#@KbuKqcZ6=g|^agtE z61lAcFU2CYqBDHRgF^9RNnKX`#mT-q#U3kahD`{k1jf~B7TI2$9rxvZvl!a@;~w@) zZNXFox(T4g=u37|0sVlXb2mPo-lDJEecoLBydRZL_|v*5yfSyzs0wErEI+s9bmCx8 zpWt;@=+%^!M#+YyPj_X?UJ7D}HWelpyyf-ksKtJ;N9w-RO^(;H$mg82vcjpn-8fI` zGbS9@2>G(I%$h0gQ%SD77L$%UDVTn zpNlPkf&M?n+Cx$9Oo*dMuuA0D6P?&kMBuGx5Bbdk1?1*Js_u7!H&Ey{gaks1Hd zC*?vAS76&;BW#O3V3zu&lgqttD%m;IBwvzvFJ6m;gx&6Toi`6A?oB;$pMo(aU5nYz z)x`jBGXT6Sl9%K;q|PE5&pju_oSV0CvK))7_yS-pxJ>Hd?@G@S4=F+iy)H&Ra)fGF zVd5@Q+-(>JUNm*=-h-y_y$EnCDGFNm{<_kN4M5WQI#rzU2y(dGtV-|T{WK)gar)A-8+cq-k8+8)wjf4~gU~+39!d>?!W8w2C^n=j2Q! z(a`V(oV|vUGp#}}-{zX%@v3)qd^{M0G?6Y6W%`?~mbamRfZ()6(-ameolnR8yt18y zl%#V8GV85{JT2uIII>(IYuD>LejD?Nx3OVL`L-+hrEt|Wu@hnay!r4U3U%(|eM9EP zXRQWElEi1V3OP(jb-{XDfsDqoqNMYa*>2BNhXa7ub)mZ*f0mX8_F(R!t^4YohN*ZZ zG9qS+Oiy@YBtEgg*ZtKculR%+cRY4SJo)ot&zc++2(`GCFpU|?{!4DZnv$`Au^H2tZI`&k6{!Mam@l_FR z6cEyY+?z|Afdq+s_5a=S-;Uui4zxI%w5^GPO8l!w9KOYBSJQt#oQx^6&{ci;EM>sA zX)ExwYTMiL7D6M)(FWXLxtI5PZWvMLGaO6QE7)xIvfSe;(l0TVnPiZ1=@`U|*&GCu z6Z+nv462(PEM3zam$MTlBE$%6x^32JcDe__v=<2H1&BtOd0T{^@ex&ZeYhm)1nu6* z-h2$s*Lq(pZL1%Cpzg=AvI-~RZY=St2mKWh+?{_JE7b-%|6ZQ+?Ecvx>2V|X4H{Y9gHgpGix*!ao&5XW zvhb0K@)ie^tT+SxTJ<0N>FO6p_Bhz&x|K>z7H~^d?dAIKl_;1EVuwh%iU)~$4I1x zDcfkZxeKRt4r;$yxCVLVf^Kf`H{E(jFuSi92<6l$Ty+Uvvr`aPUeZh-O`0|Y{6_r5 z${exa-!0GJX<39{1pc!#c{AQV@jP)f|K{&_i)7x)Ne9;-Xs$7y=x@pWwL|Z$cZM=$ zN%5Jt%OjURCV<^FM(OAv0oN)chNTvRk|Ty)1hs&=rQy-D52wmF-{1hkfYE+QCTICeVA=IwI7h3mF9Z$ zG$xSHuR73*)){%fbjmwjv>G^#oE5EIxpofj`K>WQSL*Jte~7xl-7>9|0c-cFQu;UP)9V_y--pkV*f?;t|x7^gA?`o;6U46k8`oT>X z5rMpI`?qgdj1__lNM+nU_uzU)($W35|LCViV_MYir9w506^*fosNs3_uGItbm){x` zowVQn|CffF5l={^U+;N7dq_)zL$961Ax^U`YZi-YBOYJ6i(jogM@&do9ukhS9+gW_ z%*^E%1UNUV1@$%V;Nm>f+68d~C#$cRN)q4O4xAcj6w5e#%Hyc%&10VxjQp@8( zJ(xOPAnFHvbM`}wG%hf+y8jS0=g>$J|D@9lrF?kmdNO`R(#2fcxtP@|%u5z5NxY2v=4vx!H3OX+O7x%0IMZg8LWi7R~Y=oA5P_d`C%slub*ZbPh_!-PR z0O|7k!|%N=?NYDTK^V{eLyP7(=kKkpNA=|a0eaE#LU=AO8TB(+Vw-Zcp4l%+p|#Ev zwr1XpU}n`~3r+M6_RVq1OXpqrZqTTUQ_=4?92gu53X7#{_BF`C*d)hfL~_iU!7lp3 z@A;benR-50;W1j@dn6ULFqlTO#JkPP7l)UMV%cV(Pd4w+qU!Hh9_?!``m=)ogKOS5 zIebHJh8b@#ZAF~OU9b5Rnr_Me<4@_UOv$5lZukK);^2T&3t)W7-vXu8U*pfy6TZtO+G<@JsuRTWodw;FRnQ8P=$ z@ewob=_m8wm^imoK=%sVDAmpW za*-dh)hhu>?69gSOf$E>@z@Oe>tmfGEv|Ak{!>=NX3JH(g+;T@>f&x}Imf|fxsKqY zW_E26UNeiHTpF$uzW5~13A&G_wKkpXc5Q&!hv!Z-(o^@~yX}Xt?`xl^>pWQuyg_Fk zxqIRWoj)jE6(<@b@~#u=uIG$F;%k5=8B@R>oRf)31Wg?A1@zbysb(*wL-)lT7&8K+ z`P%ZuH=n{Oo;=2@psS3V{DWJj@>V`}jbGSUlqn*Iu1^n_Tt@!>GBWFQpc?@4Hry$d zuJ z*N=8xs^?Q4Qr=L>PzgCFl>RT1cM2d564)@6!FLYW`L+(l7VxNCy0%cHl7=h6?Ql0#))K-qB!3 zX#Gq9x?0_qC!pWM~j4MOdGaw(=>mC=8)4Ml^uMmOM8{(6<_j<&1g zXyG8R-R=8m_;~#2&$O|bqT5lwGxogl3uW3T32^^-4JBGt`K5`*2h{W|{yTM>oAYyp z92V4fiz>cB9e2B2X;1)8nIlW;0XGKVwA$VX2L;EMGvIZu|41W%xGu-&9m~7hd1w^{R>`d zd@nb(YoA`0lIqrIR^8Dqu>TS0%m4H!DR9?2BbF|Ij$yi_V0x{1H2%1)$%NzJsv1Ct zpyWRmn!J$X11SLl*NtCQE(_FMx-jY0Ic*MZUDdixMepizN*id|Vk`GnLKLejkB)Un zqvzS3Qn?uH+8k&~=t=C>B##TaLDy*Mz2Agq6tcuHJN&9RefbHKxndPYo(968%BdLw z>hHRN_u~+3h{;&!dBK&FuREej(~8Hr^!?qEIAG6>-kg_;=T`XUD*nkouqw+Q$U7!* zngSE((-6ndx%?*fWIk58|4_@;1UMP-F-m%{&YDuzUY2Ux^upmXL~d=XV@ zYJD?$IYxO$Ii>o|w9jQ)BM`)GRSu#pZo#EXcbgTFt{0z<{2Igm&hM+3JmSDBa583 z8lh&kiWn;X4J?)ilDTVWl3wXr!x*uE%t}qXaRcgGzE>=P6mQlr?7p6o5q9*A$w89O zTZ@t5CA7d{)5mkRZ6(2P=*n1yGnym{4Wv?jQ|D2qS{87iO-I&;BwFton7^^c)pTCc zwXO;ymg}z%wOl`eHJaQIdoLWFRj5f$p3D-@T%%l*)!?d(Y%`h=#WTl?;gp(XTd zu8ULO*Gh1POOoI#e#HP>j{a*2J?Y!q#W5c*3BSW(XRDDuhMEW|?-nnH&3?-!!__V7 zXP&Osc6?ImZFC@r1`!i-cAu9WxC%ajISYxN2Hu{HTux4z*-E|7$7y%Rt9W zBk1}agrIf9=V5RmVJ8_@Rm_u|tv^$^F2GD|c=;|wgBD+FISGOWaT%k}W zuA)^90BJtk+uax*Buzz4uB2?!<~4g@hVnd&DKgjH-vFI3g3(BE@G7QxgYoBkoYx`c zN#$)i-55JxcKF90*VQTNmOhgHibI3wz(uq6ON&55Ia>4*T!w_XoOT{F&Zx#kZWeR> zpS-_C!)1Y>fy2_=o|L=VTI`n`md+Vmf==(|8<%Wa=En0@idTx`(7_yFj=sW5C_@Be z1mgpWP0I{k(SC1pBt$6Fm{vJ~QI5-B=qbZxt8&w%-xxWmicwX0F$_#KB20q+*!kf4 zP4I8zCXoUU+!A=pZOUde!g_b${cH{J!OFw&?4q1z7HnT>{)KnT@aQdYpmkdbwp<$< zDZ6YkI zGZ#E-j00b58L3Y+sIer(rI9AA`bbuvdC!DIi%tS>wrMDn!v+o=adAhHzb?rNFyNrK zWV~e*(UYw+1bb~`hYSYLkv{JZ(UUdbFZ^*b(9r5sz>|T4jHg67If$K81zg)Qrj*kx z8@4MCg{Uwo#u%#IEG6!6n!$VALz3rl{U460Fw0Y$vLoNt>5k{zKLu6ZMXGSFIE*aI=gOI*;zT*j-7@X5Vxj^S>L9`pXg>l)*N?& zb6J^WiGdLw(d@zN?;b7kHR(nlAh6ah)B4}{A+Ct6zuNT|C4!MMW7u7nF)RhyiQheK zRgI>!@R6wrNeXegW`AfxF}XNxBdEuSRZGk|JLx28Y`ocENJe8vc zufRB~>U9osZ!vJyk!HF9bgg`F^vNpjIM-X3AoL!<_>Q&5qv3s{vCXN3x=xChaC&^R z2Km}Z*X8J}&0_84++5y%f{uaV&BFwKc*uzGlYhoMZ;j;z>;dvr{b*F7peS1>kZ@f$ zAS4=u(M6L^M2P@BR4~&=X1-}I4L^V5OqBzgWkrmhAvj-nbC44u$@%ZHdbcXc>sC==C}g5Ju3vcR@rTSng+(Vo^krsJf` zd)@AtQR6Ji)xD3p5z`L6Z@`_Np0$m$H0ZbJzX3^O$cJrnJpA2g5zaeKiU!KYi%9u9(#PXlH&`r_m`I`Ng$>=i#gGti&l zvmlbFlPH{j=bp0v@ngSheXdw{>*3I<@9e(ct$RS&8F-Rol*i4!)|P+x1LJcmjSD-J z>Gxi%3zUSi_2IDU(u*zFH^VmrM!XH&BFU{`#PNkEBuip5Lri=%qLT5`yrzn1+Q(eN z<}`Ist1@vVS^}CS=X;|>nY^e=kt6v3i1SV+FAr&VWWeRsFdp@%mJlc1t{B7_4ojY5 z$oJKZ_7Xc#WH%R#10wt^7N3HK&k@D$b+9DE;PUKsY}rQ za%tY>`Z{5nw@m>m!R*y+WSa4T`)su^#BAGCYqT_E`su;-Uc+7W*JBewCEhAcKFT>S z$-AAnyz$gUNs#?{ps)nL)8DAkzkHvJoFxUTFV;v@Ehq3{|LYn4D5&PCY%A+wr}8XZ z(Ybf?eg&Xux(dDRv*LSzo^PE5ZM}n%Ok?Izeqk1YN)W`il+^4Sze7hJckb2q7JuDdK*vFty1u=$+61@egK=sDek_P>GUQL z?(Djz*i~7`{C)izIlCSLP<$4*N7j519{I}CJ zrZ-ck1s`xmv9cBdAC1#3MEvfOK?Dm-^#<<5(?Jdvsl-ffZSp}g_l$N@Gja`Y8?gH4 zsTVl%AIUB$2?GjrHHMx}HU$@M(KXe;t+y<*_aDcx@Bkbg-=IZ*=QOmnjJGr#$C( z=fDjCAuj@}P%7f9sasdEUQH!NC9+Hcr-I*g0RnwjkFXa(wY(%W#(y`LN8v#2++uOv zu=}d}>}ApGWW%<#U-GP|%8J^-CVT~d4>_sJ6LGw1nOTibYV%j0%y)3J?B)15mhI*B6XMWPvBriS{Sz*MwPoX$``>tL-(Qbp zbNKW+R4pyue<289fR&^J);o9Z%N1KBjaDPMs8<8FaUTvYDVOX-a2zBqHiH@U4?*DD zx^qVIQqpqL@?4n;%qki|Lo^`HqTz=_6tXV2#@o|(`b$r)A5=4t*Me@moj_g2*PNJ) zWsD~;G_iG9bhGuR3Ae}KGOoY0BJO{E`e1+Zcsq6e9;jSZKdfG7TcizE{rTlsv`g>- zXYG>4*xQ?B@R}W{zdA+0!3(0@W?7#N;gY@WqD@lHQF66P_UTe)b{#1 zd9AwWPNjgD_DNmez(A$3^sn zd8AB=bYL&G0vHQbXG8)We0buUqxs=XILQ`hHWLfkNmX)X5LrY7Z%Q{w8WckxSbiqX ztU{(g0&jUj*zM>O#H=?NxNxTlSDFpr+ebi|T!YEjE}sNiA9M5XJId1T5?J)`%=)Jp zOqF=QAiomX@jw;BknW&Pv7jh;H62*b*|UL>Y!Td8RK^f?6B4ct+?Xpa)>?lA1Ze~p z$Y$y<3q1BG6!qN#&Nn5%-csVA(~jBkA>;NI{0Mcn_g$2WVrC!CD=$^wiP>DZE6m@| zh;*k;z63@5o6Bwj!HF#VDfWZjH^+-$b@{VcGv9 zZM}7=E=!gPLzvoCj7SYv4OcGUYtvec&P0{T4m#}@DQmTg$6yg2(=>a*eowB({i$NB z^Y@e7sy}@_NW)f;I;NWqO{=yRT`3`tGOL|8V;z6&6#Vtb`Qhrr)icP(=pOGl>HYUm zK%Xf13-WwVA{X$U@ogoeqIN=!(E=i42)rX;on8mEyA}U)oXu2M>A|^ihmPNy_B^G_ zT_4j}1$&G>&d)7tz{y!x`k@*^Opl(zqNWUcl5|M3C>xLg3Oy-(vV+Zzumy;?d-g{^ z2l!bQKGvTl5;+gpb<(d(^kTmc#>@v{$P`c2_r(}W5cbybT?_zkmE8#H&j4`{Wv&VE^@zogBZ13}J|5NNnGm+tz&cecOrCSmr;% zOY&v2ztTs0b?!sEt64q*DCfB78UqD}xwB(=WPQiJZyDupbp)_=9#h4eLoAu&9Z|aTx2UC2zTg)Z+-xxfyz~fR7yk1$0-FsMLqz4=b@0nt7 zF#m0oDL6C+Uw^)@RD1=n+Y!T|&R%I)kuUsK^(~tMkC;(QOpTMs;8*j7y$00}%^x&x zQNr4ji!+MRg^u0Cwz#1uqSY>;Jnv$@Ta4&XW^y2%?7T&XGtH!zxuo71D4bKAjN{0U zk1*11jhmDbhxz;PmEE*Q7}92Wf7`&&y53pAo$5VOfYtRSIR%Q{!@RHK^3yGp@lR;| z1LU`os@cmcV-OzusT@-AqP&sF9@zoeA^(lZKgDxh^00^XwA%CsCh-QQo|-DdF8uAj z{sHM9oW^gT;<<)x*eR<)beaaVxsc)g#LoRpLP*!TZ?5K0_kiEmz#A34-Qf@T&5`Ti z7Os{~fnXb^nYv;+weO1I4;Jwu5Z5KXOjRMr+|acP$SP5BTDFeov4>FWmDJ|cv(v^% zr(COZT6PvCatX?h9(DAa1~TQk23p^+XN#|X1JBWi9z3tF4;~O0O>!nJeFItr4i*P9 zGrz0GuS()@;?UvCct%plemr@zRH&|IU~q7k82{ZfM`Ej-=I)|Xcs;o5#vkj?ePFtI zIuzhhdqgaW$ujT?6e8MFH&qGbaHig){g3%dFsz2VCH*OeGrRK^Vw5~WTQ*V!k?^#) zoO+{et-rTJ`i35`Oh5mEb5XJZh`6fRRSD1cKYd{Wids~iX4kkpx4$5rbA80UNmuFe*f1i}J@5zz4%`*oj9!95y zpSEp0`+6K?vbttxTzNUXBP-3TAE56>e)ZG{yxSN~IDNIDzDZCkVe>F<`@FGo4fh29@*WL-Eq6rvNbnN4?*{}=Xa+RF`;aLT_F@C7*Iwu#G)|aU zTJXsJVI~T{&Eh)cq{CFo3`uZynJC4>YqZUytz|;r;I8@Z!WBkqLm=cwD*gw#vjHFInnz6 zIF5g^-}Y65zGnQFe-YR#Y;4Xhdmnv6jF~OQY9zY*mRI1A z<2A!Svd#RwmxetW(e*!wSD04vIhxyJR>e08*yi9QnkB9#n*IOS2v82sy1E^?XhT%r zpGl8p!o{s%eEsU*|M+8-$%Xv2wxOqh;y-;y6~i9fuHVz_Bf;4%gY-{PaCYzXwy!hF z=lLCPXX)+zqay#Fg>S|x@#gTav?$e2Y4j#H-Q)DoS8zL3UE%_Bw{eBJ_a9z}4$vf% zwEa^K0*SYH=z|P|Us?a%vSlfwmIss3ymaZh0IAD&9va%%M)<81ZL!bb+O%=rKyJ=E z=u1j<^BBBSirh&Fofq&Hs1yfU4)_10&7c?OBDqX1X7xi?41Rw_>>FTRar8iEcrDZ~G_avv#YYmdXmvGbZW0y&o z11`AYW_~XYT!~+yGn}WR6OXc8uo;qyafl80`g8pq(e2Oqxnm$}z@y%3Eb-NTQlKDQ zM?hXgZG?XzLhy)nhWGzPk%YuTFN)7s31|p|BtDSJjVp{R*n%H@8AMC3$)lxzrx*4v zhU8UL=MM8tt%fGB5o@&nM#jx`uRkJ3=rc*y1H8mAF^ZqIvw!0Y55O8+1;*(g@Tjsz zS%+9dXO$A;-@W7lmc-9S?t`c2_T3X={Ak;4#hp~qx}K*gU;9ZMp2uAS%F#}bF)4U$uJ|1^RG@~Ji-AGL5J6?{HsRh z|CCGm2K&{2uET3&vNxRnhV0%ul!zLX7TNdNC)`&mH7Zvy4c|2dfsRR+{QrZdzXTdD zd^iNXEB@ULrL>ue@$67VaZUl?>u2A>`IdfL_hbtP;9UzZLn67S)cZQpq=f6BE|683C0(gzv|Fk#>X$N%_+a2Z3&QllC9OM6=vEYolf{UQ( z!gXqe7{_pp(AgiaK~U$L))doUWyzDTc}F zRH5-*2A;L~*7LuC#0?rZ+6kSTR_1;PYUkl-93*4yO;uPea9X_mzf(O+_>(9(uO`2` z(gWouud-$~g*$cmiVYV6*#sWKfhN)`529-eM_M*}qLSZ8iTcrB>d$ERX#upF(e9Rn z|9?v2N6;|pz`0#u+$UXJz`?^>$gSG>kDamqo7+c+lDQJ596O|?t!D5_0A8xQvBAex z=|NfQtd;*?*dstlR2UtmT&B&}A{`(B_5Q&ysrQyt*&h{lSk}QOF7>icfe<6{I9j9L zleUa4lVJhF$eb`S+!5+Dp(yrn&IBCS65F+rXz zf$@;h;W2ltd|Q_77ed>Ai;|nzDAcClD$Z=sZoU5!h=ec9R{!5&43|kfSU#B{J6Ms4 zOB|Sp;^8~n1#o-0ZIiQ0%ESE+@(FJxIp_4vkjr+%#@|&TUly0PM1|1e^9NS$p)255 zIM1}ucI*Y3YGYN1JtYwgGk(7hKA-E$>=4z$ zi)Kl#cQb8dc5+(6&a>Yluwp@tDmhBHh_xO0Ji64qbh zN3IA=1&=zoshwfGM1ZGRbWrls5Dqg^Fmwd^Tn^i0Ar0R*K*i9C8a}r8OXD3p3Bl74 z>3;@(UPO_XWV+S>j|D(-$eGp}I%Tu{T}^^k=#29pDZSf3L90tDc|x2t9P#4;m*dH6 z!Y3w7G-_g+IK%39LBy{TYp(FIo_9rczx?z_LHKMQ%fP#_aQI-aXq{ z2Q~Y;-wv)AqlLFlOQON%QTEx5ADSN|TyWkm>IoZ~yeN!I8O#ksq8vq7Sx5#QtR(1& zlm^aX$x%LWa`xjX9d~33B_q8BEDez$hd41L$=#mt%ffek+=wSz$=H3WC9X59?TW8n z;ch|e9AnmLF|3mrem62|OSkC`TuB<*-3hx@4Rt?fjeZA{ zU*cz7U6~Bh*Z%Yj(az2S0l+owmEGy;Pb_UUun5wKvIvz^M~~AHZt~xF%`9eL!yX8+ z8Rp~=`3yZMFO++D7V3Zj9Y%Uj^%3n6ZKsa%a&o_wYx}Xzsg6je=!ai=q7?2Z9+n^D zL&Krr^%SPNc6Lrnc-u~3m-@rJpz`=>MY=@+Mt)C8GGiWFhkW#r7LpGn3(%BlLGbr- zkyp(sthtTXYw^z|*jNF}Xth6}Ww)(Vvp{)sgCmU3*k2^GZ*VGo^&RT>9>8;KB`%>Q zpV&VOVI3PTI}gd<>BcI>K(H{i2{{rGnNrMbB7ZM~jx$_BBg5H0mu8>CBEMjrP-`QM z;-`ZI+>k10IN~^Lce{P9d+t?xN>9QAPvby}CAx9iPLauzyT^3>rh3IWrB#TQ zDd)I#0Q|%gB3bh_#L4S5=J?7G8|fm+^p6IU5?GCPNmnJS;mI-+z7a;~w+`^5q=ugD zi~)-GQU?}%Z_a3J<7*zD9ZoORyG)Y=vmFk`A^+;g7}iQa9SOCNmDmjPRywKS!XEA& zmg{RbT88UU8u|$OFz+FNY4=k16N?i|Gg9qKK^=gGo`zmN>VlAdDf=t458aY$stq!f z`Nl}^kW=`HC`{##?E{AbhCGMfB;Y6GCcZ`BaBT*_3!+8YitX*}b6l|2p4T-sJK#d^ zp+XBBlaio;17ZSWRz^;^#>5Gh=WB(4*<&Bs1Q0UlksDy#dV}7sU7zNV7jR_S(zT2u(qG<7+wqdVL<(5`tF|F57 zHsrC4kDs5d@u6pRW_5L(^Lk#C7f*xg=B<$__WrE$Y{8sho|NLXx}#($NTkLvJQV?V zkF%cMzifKGBw4;L{FkrG^Ma+Y$MSI%FB0?DKEa(ddOJ;qTc;RAO(|Y+SD*bKRy5`r zo1-K#Y!7EQ%UPecPWlh(t$mQWeH;Rp^7qnas7bAkINE(sZ%~Zz#6U<2LAs)<^pd4C zi?@zzR5%Gt%5j=~%bx(1Y5bY`B6dDWh+USrD|npVN~~hn=W|@{QazQrOB>rW9Ai3G zY=E_LUcE=sMbJUjivNIjLTrj{;oHY`-?I-gHLc*hGsThIMV5*Oj`M};x*!}F+kWoM zCodk^mx|^PD{9lG0Z+_9CT0LLDZETA&EiuR(ChWERr}axmbRG-H9@z1+KKV8n}d*8 zv$xSo3%*ip7N6;;#~@%>`D;zfVEo_B7Qi(&o?cs6ks!70m1pN=!SvZf zeX5Q`e;xtbsIb^IvvXL&+CGN9S`eQE&NCe-*W-@j>ZWSF3j@D|D~5#rU6Fs+KXKq$ zLx+(YYxvhnv4b~@r{v6>m?s-bU4}kPnp%OXk$r>RbX&4fTOJ<3J|;&wAY8l8i=@+h zdssqg?n~Qk_*k|I{VADeU@%v=;>G3Q(CPGEU{PjH0s$rEvo7od>f`j1>;0^Owj0+P z{}DBB=uDhu+T*}OG}JO$dH9{_+HbJ*DovhAt_mAWL~3;`{GIpvz-}=V@*GlV6%uA3 z&)iD}g6+rUEqM!CKC3%BRXenDE}6U(z-)p2UEyPTr6Lus{X-`tON8|{@S`XO`#O`~ zML;vL_LIS#r^JnM=01K8?nymEgF{V{ZH#rx&+5XrDbC0g(7R9*cjI7+5_-PE`!qI{Db_+0_WER4j9veEG&}ork($$gNB7+_ z&yIGG3L#khzdX=Wnx4v?>ACg6{!_)GgXQ0@v?3mG2Yh}ga3Xvm`uWVckhAY8q7}mM z)*EN|3QrHI{?JmA>`vD8y=23>qG@M&kLp!QkGJLOxMy2y#;>PjROuVrp%pMIT({5< zG;wywk=-JDtb?aD$x%{c*fj%rxdrJM^ybhAbr}|zszyC`53Rl=)eF4+1|?mkdn%fT z??@8bj!xI3LbT) zyWpG>r=C|+_vvpPoX@XWP4Yti+DE!2&J zi>tL-PSyCM(YbA6yn#Eg?c`5jmlmlNrlzEFNaXox8F*Z5yuucLVmT1uJ=AHh-W=PA zwaICl#G~1p{On2Tpe;FO=qH_zd8eQK2eqd{-*2usqm~8~l{#VjsMvNRQ3*01<8E$G zVRHrk*NeS5sfzEEa}h=89poY1lC9%|%t4kC8so22;)61^AMs6c?;U}oHapNYI||U< zX|1u^ow8p;4r-pyh!e?|n=sK>-uO&?yt&!JubBH!3Aed5nj_06U7n?mwLB=B}xUopSZ1A5@(#og6dmdSCDHEbO71oX4!GMiUd zQD!n@momqGXib{Wj{Yw0GIN3~*|3+(f7P z>qPp;U6F;_V6)0jW|nHHY5TpPbVRc}Kfi3eM$^}xu?Fu^{fb(D*&Q6$UGg1uucZhw zC4iFc=h~~qVl6NLR?WcrR##`1#E_>?bnRA;4F9Z4r3`bir?9FF7`jT&ujGb{)7yRs zlUjf2aQo?r$trb;T{YwYk67B^9SO?={=N81cpp5t(JTro(WR`MM&J2}7pDH+f*Sn1 z*2t->fh?oRcsMqhS=&m^DBV13i|$dwz%&lo=ULrtea2QiCVwykh<_nzFEN*Z~pj7HG4-UH+dhN)*m|NIgop< z6d7Evx_i_4sVZPMi*gBTGbp;6rjEqXNfhGZ%{p7Q6k?+8Y|wBywkS_vNpkV)3DNso z%Bor&!)!k8%DYfWRf@IvP{0@-`w3C9^00(k4WF>rGo%yA zh9EKAZ?XGLo_qS&e*e z@ANZ|R!R3e-DhG2%g?dfq=79B$JOk}5#+ED{Mn4z4c{6$ zZU^hB>0!K&IiJ9|`v|0wkUP&9Ry&{)r-xWK$27u$0Hij*MfkUp8`fw8rAeiq(4!}= z_sqm~WaN7(9qBsheVi|m=W$*^s0WEQ5DoO61zFMgsx=J6t{g^KGSZOcK5rRfv_l1Z zg%43{u(0@=s>)t^a8u3r5BUa;5nU_0$S-cd6r}(p5`KqTx9~7G5Idlil#pv+@lAF_ zz2+gGO+CeMY;{+j%Uz|b5vOnM{TGtcVem%5Qis%va8tUiK)Q0{w~sReh<{gvp#mz=mf5r|xNI8%T zb38K(;9W*Pdic6#J(klHYz?;7*WG^qhj#f?GsqR8wisgTDCN4%rgn{KLIAG6HfKSy4al+Z!+vvQ_?l!LKJb1BSUaJe=!Mu zqBj=@9$+Bg@;(}-$*c)FDO!8#}IzLM^ z>DTXD3|W1kP zOiDzj{Oe+t98cOIBQ$UCkhw3#r3PlV8CxLPBoP8jHyagBjx=LjjirWok*szcQK zV(YZXh(taqnX%pINg+ERb}MOHg<;hp3)9`=@OMc{bgqhnXS8mEesh|~mq@c=8Uw`! zKdTH%BB&}+51PxwwV!{-l_diq;7J%i@1n&(q7?7A93;XrVOc1?CsIq+>DCd57|^ac z7C*uK)?t%4+ctT-x7R~1HJiEKx3IKz?bG?qxpnh6z`5eGj1CodtbsL|E=jqa=4r4bs-2LsHxK>MYFAsct_xWkP^3s`uW8(9wB!G z?^%}B@AgvjS0F!Xj++te)-31s-YW)E_4)q(YUl1-oES^-M1Fx~#m53Gt(`^kJxF7aOiqMCTc9 zvMi~78s&>(>F&B)tTwdI_ZYyU+cUE8-z8qtMxo-pu@xFqr;n_}))b-wr9%0WVq>)+PGvv&g$!Ld=M=#m9Dn zNgYLlTu;gi6B|+fRMVS2?w_@!I2Dan9&e?dm>nuS(XA(nl2s}38$pt@eT!B)UsGgQa%D- z($LPEhq|y2FP@Wkl2B(F*#&L9XmqF0^SH3^t_XUzW5`)R5V-D9;h$#f)jLx?n+2Ux z^S9#fXgBMr&Y#f$?f0N4)SXToy48%_&YD>uqUME7C)rInX^|ToL!KaBE~l+qO(~>- zp#9V;bX9>h_m93N<%FU=?iv`<9vj*}Bi|p3g`_8b7?4p3VK)*mSw$V*sd&ERj$pKVGu4&r zUUX}dGuKJ!9w70CDxd=UHyW3fcw?J5s~MO)p(+ue)`)Cnc>N)Vvxf~9(@=iugh@ks z{H%`5vQ-5j$whVWW?|EYAm1>%qp+TKGrRej96Uqk zG1@#n`729iuZA7ZB)jysEP=$~_3}b3wTq|P!b&DOOuF#yqXk}LLan)N#@L7ex{{WUZ4qt{dZgK(Qnk+ z4Hh!A(JaX#5K9C23YxlU-2`kanQ@_WJI~`vLbi3^@+@%ognSg*rlm`BHfMfynF)64jo}YnJoF;z%B0hMgsF-Vb9Bo0X!Q$PnD>c{Rd34hn#-z5KaDH{h zmN(5kr7|4dPj@ull6$eth_;F=Y+=u>oA+XIR1y*r<~5$kHs2TO%w|V6K3}v&LoQ|7 zYn(+IL&=^uMbJ)URF>pt%gIQY*7GGdZzy^_RFjv*lyuIPp3YXSco+&+*jk$QH9hN> zqe}^{PgCN+LWPi70CqMS(fOT$ z{{+SBTZcKujt=m+xbarx56u#1@ST%&+zQZfD*r6$Wtz-+Hb-OJ0Mg`*B)}%h`g&Qx zKo2Ihg9#&ry>8b{5EJ8MoJTDk`6;<*(j;o5DubJckoqg^Yu7$&LzxuS%9n5D7vNz~KejSS$SrvVp1{83FTW6aHNwavRZC}MwZU;eGCiu= zHj5S-_F9)lWU8Zc0arr|x*NkTgqPQ7su-^5$2;8WG|VkWZTI1wd}Nn=5B*0(r^r_e zpv>I5`UV%2JQ^#>jg#EOGKeM@s7W>;mg(o5(F&Ug6eU5G~SFKi2Me41SAI$%!*H+xzYr!_dik7>Q*U<Fj8oMr4e;1`u>;mfR#%a#m5l5#K)x{g|sSq3~&`u0)l_7gtaq zN3Yfi+Z6ZLgWJHxc;gH!ioUDp z3Lhu3<3)b~Sk}h!lBm)1gK66nAW?*t$dyX=CQ?mbIEFnsk-tdph#Q-x3*}-xO+STq zPx%)tAc4?K8oKm}*lYC4IijVWpY+ch)FAwL`{k3Ql~J!y&mOMc@AMjOaRw+IMo>5` z1?FPFt|qlDLm;tqIW?O8sjwoi{ifVs&x<=VQL^^~w_MH$oF$_8&+6r5zpH=LRub>9gte?=lk< zCDj@Ud=XdJx5-r{oN!Ik^IudbB10C;lN#0GE>kC9Sv^5-$Z9Ae1lk6nJg4uWo^3>Y zU>KutFiTj-^r2KmzBQRycsCJ@N7y0xXcN@#O-&TQi-ME|y2&X+C5-#n#7<>ys9ZwP zCX?9L5YQV&$4^)2XD@n6CXgSPWYxnuBpe00J4=j~ILE{c6CAGyPz51nl-Y#3QjyKM zdp543f9+BsJ#WJ=_YXOL886e@wf45T6kja$I3erq960q0;Z~Qks)92ifWS~|P6mkg zL*!K|NGFmz{@3*45W%^kLi(}9t8T3)Yi1M{-q|T%OgT?^LUJ6a#Shl_K;S> zr-n@IJZ82#eJDB*BaR}cURC!+qeQz|giRj6BEb#O{rN@LR9JKBV|p*1b8n&^cZ2$( z!t{A2t}{5yyOtH8~L_@|DBm)9P9J^!OwzX%AjX8}N`C3ClTfLF-S`IW|6%SI zyvy1y*zDmJ#CCS)z=a0z^ov^F$J&%A8fsM=&u1%v=vI{IpwBBSZzx(vCH!rurFm7>0ihnHYZqgg$= zzVsJDAyY;v>GklEHvTFtbu0PBNv~$2LKTIzMDnAdxO$Tz|BK)cWDgIQJbL>_9zdi_ z-kn2oDA`9k2aDnF$rtAVTnfQuQEzQ3kjX`H?xR#;n8Qw>C%`(Zx+8(7hrSfunM9cz86*?h?Lns(DitwAvG^dMm(MYMy?`jq$Q zmZG!i{sHJ5)^lA6PJ2?vUoh&E2p3$Z!%TUDNDfJaIY7I9vGhPML=QEOnT}P+kY>M2 ztnG;Xgk8cqVJO^{le;rap~cz0{)_lQwb{V#d`=8cEr7H6v?InEDc_GLzGQ!0GwX-k zp2NUflDPvJsofHEf4uv14t5ADe`0FbM@jAGN()2~E0s?>6H5Nbr?H?Ggz|a2otf__ zzuvqhs0SzEouv&M+o_5M#Cxur53a4!F9BV$DBE6~k)d2kwr7pk5bUaN01_j`wnHBJ zu8k$ca5!r^#U8%Vj8eQ_@*PI1#cpY6X?kY!H*aX;@cUuvGG~?WD_mdS2jjQgeh^q# zHm2M->%s7&si_f7fxr|PRXOFRX^2Z|jmM$7n6S_|)+i?d$Jfcjq>3swN9ecSF_Qjg z8;ODo8rxmuo_Gzdr$cb+Xd+iKc?c4l5-9lPf`-*1USmt*ak+rZqUy%TW)PB%asd|9 z#`I#ky9w;VUAPY@-wFLt1gr978$8tS&-{gW6?}d-fT=8AdwFHR%?`Q(WrTDL)t1Fw zn9i52`|S3e+CuRyzFdUgpyQ8sFCei1Z@gm3u!<|NHr*c|yz4p~tNk|6V9nnU9E?v} z1=dY8`SosyE{6_r9=p=+vOh=Mq~af<+K0aCi^2V_WBP@ZQ+P0jjP_~R_=fScQyM!8 zl^@8cg}x9!&A^Vfr2}1@W;K0b<@!$K1y)+4U}_Ee(gEZ5*TU&`v7GXNkRgmdpCK1+ zTjuMuL1_Vf*SFJ{O8!JK43SUkGyn;db?mYveqEd}sU^8w+jJ}&34^W=yrsT29ps); z`RcaGs}vu9Xo1~}Xg~UCkgo`PYYG*Y!0^P*Kc_o*qec+D+d{FFwb-dgR11BXuO3dN zxbXF3(Km!BhooZ$k5K_`U1dj3KXl)e)nzr>rJk&k*LrQZ?5YMNb$VCeUp(2Cwoc{a z7}MOlmlOVPF4GKZvihUwRVc49CBMkqXI{J#Z1J zkG}z2Et?V0Xcw52AEi-4D&XO^_V#$3RyZ?o*;drRz{4Qz{Oaqq_WyW#tDrX6wrdo3 zEAH;@E&*EH-JK!@g1dWLDDD=DJ1uSj3KVyDcbDL@>GS^Io*d>NlbLm2y4Jd~H8BuM znjVVC8}*03cT^bV&)Ff-1natB)gS0z_Vb(7P7q8~bX8S)#U&PmxJ89}2}{rX+)xCP zNM4nO!@2+^gG_Xdy)ee+7PJGyV7@tUh}3?=%_eWzWzr)YV_WgC*>;utp(;qyghGsV zU{D;S?2OLQy!J)pB&hXhY^6P(IL%1BjFIwb9ikRVT*YUCsZw-352=lChVd~*sSXUX zg5>g0s;Wnlcp&8ErqPUzJyX0`q)s{|$#O&tVl zS^d5MT#cSHngy@|8Eu2#4l8%*PAZ7!nV`RM%M8bISU(}EMQ#91&b=pLPm1@D#|<$` zN7GM^fqpsqe)nX_LWYLZ_gIWqOtOT#8$ax&6{UZ7Q&_!Tar)35ELdwz)!WK3W7kBx zL_Pa-B=<>EMIuR!y_>$x&U2JoSd7KmF8>V zT`U;4)Tu126ZAuQYaG~27*qGyuo_Kjt$Uqd1YIWhc63*luVO4KP)q!8S0IMA zZg+e9DE+f8PJP3J^P%rD*+Wz=h}%=k>Rpz)!;}D8kH1wjgyscq5ztaVomTXkY6Jz( z{M=;N7JfhEoKmCD!inKwVjDoAM<;aKCQ?fC6;6iGmuRb7P1S4qf{pO@!1M4n*gVsm zxgq#aI46*q)0jSK3~mr@P$>{zoFBo=0o!U}jK*T_a z8+zw-#EIB1SeV(wY-uAPzVc-AJ~hh^%EEK=1bRb z4xXmiF?RjfXyA)kD$j!wTCB2?n7j-cq<9nF6`9CU&H&}20tD;<3a~*A9VV8)nA&VI9T5CsYsRzxBi(A--C}>3Nxx z=$jW&aSTj-f!{*D#)XbRyCa+Zx_ASkA{Sr>r|?_8&b^(B5@_&d?nWZVb5d#-tfS>O z;eh8`B>`onB7>^jjnm&e@m%SDxlElF{#z5i_eP3NY^bmu20-hJ5yEs4GzFVo)ZI8U zpYGpMWo^~q=e3Ie~bCblj?;DV@$q=W;ua6=4l-@>U-VKX~ja66>ZtVWt zww0X}7Z~G9JRddGlzD3Fg=HnY$nKM+3T$ep6V3CPhaHnskQ)P7^eiMza!|QzZv353 z>8jHooVb#Z%0593(IIoG%wJCpd(A>B-DUNgOkKp zOhRdb-NEyduzl>h^rFMG>+N|P^^PETGZLJChQX*p^og)0;wa3(^RS(i5T5tt+EqYs zhOV@e;Dk?q&E>N$(mK}6;+xHa+vmeJqNncVqCXo)9FknRLKC*!wtr@hr}#gg=IuB8 zv5h}&_&0TtjbmWxh=WEe)|}g(4^^U;LUX*x>)IPPwzhNi7T~2W?tIDf8ahAelYD72 zh!4t61Bau9V&_GPaz%8cq&nt75R^m=b?RR6u9|Kif4zTKE}A4h*S&bzcI0#sF3zQy z$Cd#jLrWx-Q;6vPJ8mBChOwkRrEJg2hA2w?p^Kd< zuH_5v!s_t!00emNk9ZRFxMYI^Wx06s>Wg^g$%qBT?rp@nqLz&r!J&DRez3;1-c39_ ztI6*znPuTP-)M_4`j?+&@_>1Qtd#1V+Wj9X4pbrDd%;7ia=Uszhn8H$sz5A(*pe4f z;BxCyC-)o;w@NkZ(*SSl+b1CJgJNHoZYaMA;!-$BVf*Mb=3ynf>`EDyekwkUm8`s-oI z{|(Y;vX>H)exv*2(6DU&K;|t7ea&*YY^jp^iOO?pG(3&{qH$;(o*fS=;_v5Xedu_- zVPg4tsIy>+1A;H~M^-{W6D~lmV$s~x9FIdpZNOLU`i@Qc&4acp=#7Yd*}gL0Ve6Jn zriH3{WAo%!jnLh<(C2@dH>fif} z??ttBDbf^7S=;0>x2GH&np*@jWEy3$%bZsMS!Ss2waP1-`yJuM0Kc%jhdh!7rC#!9 zWsS%trNBAtcjl0a^ABSy5{bS5f0uo3*!*Ne^twl7%EC*M_Y3W28CHhj)?C;S2fywwi^Te5zEGDj8&~~Lek-R8+ zfG&y#*`n>(tomx?Pt2)ex-A^(_N@H|CF@Hfxuun-D75%|?Ao}(i*dqy0bGR4N2mu* zA%8wH@fc*yR{>EeHpW2CIjN0c+e5CIlVuPUxctmd0o8tSZd(0O2^H(Xxehd~WGeriPSLJTB@*gQPM!%Vk# zvhbk^>pVB^&s2bnfbH`;oD&JYxG;4Podo{P377Mx{gSi6hkXeij?c$cfxmnnUSBhO z_PNzmvNEEdk;P)5@Mf6pKG7o{}GK~5=b|=3!6I*Kl~zB zD?hDg034SNmi>Q_X87nBvZP6pGR`FNG<3)9Qhj-0{yDa0A#BgVRm@J5r zCvN3RU?hY**ZWXyi-B&x+UE)qOC6I}yV)=jI=Uw7i@~+RMErh5zwUk;f7Yj-{?DJn z_wNdCPf%5!gx_uX(=vGDDAOiqJCmcve@X8}+ZP};tHtm_;b_mOTu@Jmd?%xUC~qHhll@&(qd4pbaSz;3pEs`|qoCyTrA&U$j0kL)(3kFHj|l zek_Ry508}j2uPsL&3(9fdAQ}`a#^RR4YoJ3FPDK$Rb<7_%GzJM#Uvx-XQYNfIcIHV zV}Jxrvv6|K!ky?YE-x>JnAkh>;zp!fOsaDs0_@aGbjDzY8B%B2JgrXCTy=KK3}C)~ z!oj3Xffg|}^jmJE1TuOh^t~_MfC3#AxJ$%C6f6M)%TL?K|D%tUeuBf;0QZfK_R;M_ znv5-VXBuVB{teE5`)Cl@4aU_-bZ*I6omF{#&Tgi08EV#j|I_p4k-h)fFBk7G5f}cS z`||moy7a#f+x_Kc+-#sQBA$J+sroHma;J88VdY&=XM}35|Es7E8ZAS`>|@!_e$=<> z7dl;`kt{^#ZnRe5=10nSy>S7K7t#S2UK15 zn0pA>3Kw|wCD6rItMoGQQAKU3_2vwFywBsL#CV(-&D1Yw{Q1pLPtVF~rO6BkZ2RMz zA&B{|EBio3#zW1IkAgdJmsD3b5JmZnGZUz8gVQfHMAg~f&C>Pk+kL6R5FC~WkYB7k zy+NJYv%3MYW3|!w2p5`Q8e=b$pz*f?Z^+FOPnIV5Nz>`{8M$D55Aui~!#w4NlIoo& z?v&b!Cqp;CntwCD>oirBcloACTF5j+8G$qKIg3wV2r*Iocg_?9Y9Z;A06H_!B_|OW zd?WH*(2}FJTu=O`Sw`Dt81Sp#3^1L%9(OBrQE$R{sao-{%0(w$b?}`1S8a_B z%0l5XE3J`1%J2=$i%K>POp#Wi;7f6 z5lNTNAJZfVqyJcVY00OzQ@o8n!@fr)AcT2kF?Ey}>s<2OW?lx|NHYz1q-PJF_T<24 zkO=Ow_?1<^7tcI-#+i-t@N^R7%syL8NG`R74uKiWiCZ_V;cO;UkWwl;XntP9^M%`(feCf+M*e2saByfw+j(}+vVF>ya(&D;Isf79G-)s*FT%E?$s;_;5^ zj-MOweTs`@y6<5a&+NRsjB@d%N-?g6A8z^7bJ?Nhd+NLOa+o@qWye>R^b`I(yTfEY zj`t#SM+KS>#vZQt&-f_+`r<2$(Q)%@V#$1`80Ra*sQ;hH`TexZlwU<;#hv68PN!kF zM3~+3GskJu#cS6+2}J7jlbULDPHJVi*4u>QFU1frw#}dir@X6UJQt_9P9l=8o3g2~ zm>Tf=o$@QbyS3(Vc~ekt*F8dybY^I`RLOjQ*k<;e^18q>ucZ(PS{vFKh087^oIw;i zr&Czcjhy%O(prN&j9pG!y@No|)_K3J@)VDx>MmUyI%&#hdFiFw#V92wN{bBf@BVf8 zQe{}JhV0N<6>NR{&i=YemL&ZHL9esikqP=Dy}_Ep&LzE~t41vlT-0#>}I)W2+k*6UB*C)HM{t?G+l-T#+y|4q>rqYB$xIUkX~VCWV+3=AliL zX=*$`h?|;=J4g8f``>nTg z2Yj+%b|UN+m21wtHK++ro2J&Spu6zZ7;zqDj~>c&|I6aYmfYN9@^ZOrIePp-=Dk~@&7(+#qd{1EglH!#^bLKaed z#NI~#KIA`AAQsK*Qlw!p=1_)%%xTPyd{NpSC#*f&E)9<{G^HGvXA{v z9XU8XUF7On5vM|j(ynCmvxxrMXC0hK;$XKi3E^qr+J!C)I|ygLYyc3GhEMc3>(B!?y`_OfCRl`(q9ty6z#PmLC)vW*m(lD9`PI% zU1GFtNLyVeud;@=u(k#TZP`>uB2;`K^U|OcfPcNY;EGC4l{inkH@NrHoRk51Z9i&_ zgy5^y3*n2{M54b49_Pr^%#;Vg4 zV@|JQ4n9}+TYF_Hz*}Pq#*v28HBlmoRw|bn|HD$uwlcb>2SNpMbG&ilv?Y z#n?*hqsbdn;uYynK|y0~Sx?k#&VD#XU4EeYb|97@w`XJVhxU7RSI`ooKYIkOk{=YN zX_S&&06M!W+u9&(esomW-vMZ&G`)^^0w_(@U8zC-c-;Mc3E@j5BJe(zdi<1$+bSoD z)C&&2Elba}M*{jR!N%$Ld!ft%%CF``#Lh816qKqFSJz~ZU%tb3NSrwjQSM^wn$^bC zI(%_?*L!@6D=)Wap*3KrRv1bO>@QbliEm118%~l0nmY?<1*cmOd4Um_ni!vg-U&TH z2Cbbp|3nTYzCZwfTEoB245LJR69)9jH+#k49nZ=8v4TE<0b9H-nfDzNQu99pl5ou2 zCrjhh?cO~;?gjK#>y}S0uq$hOr~=dt|L01%!=LphsJzaRd8N~%v!}yEe&REnNKf#`|n){aH_lsOf(NqmTh-bFI(u!B@(Ra=}`Y*mbdudp6&VM7xrXxAQN zkxpiuQ+HD}aLSNgqVhm9&L3`yw+M?S+Z`x4@t945PD^Bks>v@_!%!TL>hA{+4xCDZ zX8mC!Q7(T88^+%?=s8fsHeO2P(UhdyiS)K}FKF7xxdcOZbzuU5{OuF%qMy1Qys558 z5kV%mkYZvlo_<%e$FYv+?YZ?6gPj`Q&i5ru7UAVdmL|ioFBkvieD@-x0bZBXP@8M`&_|s>VpJX?O>dS z9qE}xyXzuWajXSu*FUc*>5wSx4d}`(L4CJPc`?Q?{a&BAgc`|$#V2YV%#LD`^MklF z11ZGXQua#>fONs$%cQoSJVXLk&YtW*KISeOpue-;-J; z8g!3z?iVOb2bTR8|1EmWI3N-o6nk^;x*TaclYm&rSWLlTnGfa&gq{16cumn>y3fg%AR6cbe zbi`Jd3U%GN(}9C`&Z{`_s=ai#1B(GPGt{8@UNBrn5-+sX0K4J~6eATA^y6Hn^pb0Q zqnL%45$QlR(YdA*Y(x`;eOi=~h`K2u0&ME=LY5ONj=<}N?XRIE`Or~Z1~E5-=VZKV z{0A_($D_z!tZ*q|#jWNz;Z-WEMP!J;!03#UMVRhBY+?Iq$tzb6HfbOyMWH{_D(ojG z;S*g0$!@bY7!(2L+h4Wu5(tL~@)Bdw?Rz>HNu&HjnIW*B2QnoWQ>%4j6R8}ewy=z) zy#v{Q7d@iD)qkt{Nrb9ew3#)7+0_~&4{zr+U^v38p^hjjvw*-d6k_~MpFy8B{UdX( z{NKUm&c&wiaw z#Qi5Vb(|*N$=Nk=Wg|R6H;J^z(q#R|b!0qaOk-LO)ef}=em{O^7MSEq6iT4pd={q@ z{}}+jcn_67(3qxbYTvnzwvQEA6>)+7FK&MSm<5-1aSH^lOf@&MfHG@)SqKuNal3n5 z;$1KdHQ#@08>BZyGp&v5rVcY1$YIAGf7n=uFomNjG_i{9j!6yp|KxC(FSGSxOZf3K zH+J1#bloGO>!eYa09m_qD~XG2Wh2L=*EmxrR*e=ac zpp0Ml%EdS3A7^S&3vW@Zg6nf!8`>EU#*3QdHMxCmVK<%N&D$&+-{|feDdf*-?`Ndv+0PRz)hn6NAa%lWXbC8#_ zoGpmq7qs0cc|7n#j45psy1)Y5(OD}M0u$|9DY+FPbUq)puKg($E$wR~J<=a3Y`Rt? zIsa@=g&-0ousT}ueW)eKhdhH=LBLbU@sVfei_NO?B;v?8^SONB^)e&@RHdTfc@r45NF+HNmL8kqBtoNUl2N)#_t^Xo&XTqksyQ7=2GdAi0A>NbMHD!HtNy$qUufQ{^zJO~t z(Ewm?70&ufz)5mGx0=r@ccgeO#}IleGo?R^Ju#}}HZ2$EmYp^Or5qQ|!jvLu$YCz8 z@Fjj$K$-6fb`WmS)oxFEXy52e)ysq zN7U;EN+$XWHnQjSV+cEz;LQ1_8&T<+@ga{1b7&VNj=YT5fZ-^V-eg3T{l!O$?QF7q7QL5}qto@vt zoJCe^nn8HRY}<7?Xk2lJI2gHCla6&Y5~&h8lUd)MWg!;`Z>QA7U>_vJy#ja z%`AXlgLqhnGldCX!7sQXE=9ux?uxAbhXyHpS~0%jPC)l_>xaZS0NTU{!C^7sSHIIl z!fZiWCi0Wq1%u^!SRbGdQwtheb)-gQ>$_Td`^Rk0H~ArAvxr^5cLx2ZlHT{UJQnmS zC{K&2h5m5=$nt}Qp6N?a%fE!{gbiUj1@0%gl(@1x{@XjY+0i8}w=}&^@q_u2@z;4M znw_4;U&1bb^pLGTy-O}Q*92b~;y)@-(`9}9{y)={cCSx*^J?=GA8tjL#N30FqlK|& zd0V-a82K(Y?!>ycqCX|}(ft4Nbs9caZ;Km?2`{;?^4D-UaWGz_bsZ$jYSIN$Ina=X z^ga0p03CWY!6_oWs&U}29F)W{*RLZ;_44x5npKhlG*#O`guhS6ibN`arJo0qUuOXl z>})bTXget*!z8lMIJB^|rdvsVKNL_d$p*Om%vbPkVyE!!G^ilnw)+cSX7#h?QAIF? zVzCO$;8y&D=2h0d-a!O%Smt{g#+#hUGhODOKRcJpBo|H<; z1-LBQp~g@xYNC(2W;mP}!RNh)pYHWOOE6|BvWGBuj^YLH`-|WPJJpD%_h8k@L8{JV zqgm73RlG=~OzBnT=%uveT=k7Tq8U@Ps2y=_VcfT^>?W{;{YNgx#eBmW5p0jAMY-s| zv=o;(ecL2_KaQcyMH||fJC^a6@|@kTmre~g?KUA9nn$8Y%785UyMf;DvGidUJG-mc z0v>k&{V}E|`DZ$7Ou(e_WAG?ZHv`msXUy~R!#|Y3_?Y?#wP2$;MgESA{HJV{{H1tV z^_x|X$nZz)z->=hGOjk}thStg?t%yt?RLV%skkX=M;5pe+}%IFKD6N43;e&;51U|5 zwdy_0Nlx28xHM7>6$8K_U!1@~>Y-cNW~L?R!o5FJE3nlvllt#lKKZd)9+cYE2Dfxupta z?|EChc9adzYdayU<#aUQ&x)d(jFV>8L+%9Y4@!4}ZrCqe+)gE(_R zTIHU^o}o7p2^FOAMoIIjW8w9EALwub;dZ%}ANM{5!-&4UiLB><@Auv|FgZFOp|6Nj zuJpe2pg7kV(R6FhHEB)}L^`bgH;nGBe=Pt(Pk9%eFi68GXDtdka?MZHHv!`( z^`{~l7zG@=Q1?US%#>nEvJY5=cWZv;3*v-%}lz{FjpY5MdCQG)E*ZY5QG;`(C^|EiyOZp&J9T(=n~AsRx@IQbA}klX}qoOb-DhmGtLV~ zc~uV-5Q5e2bvxz~(mC5M0+K0;odwZ^&GFMsHsxBK17}2N{n)wVwL|`!&pj|w}&CGtdLVP_l)=^zyY0CLN%0b`cNFGUh!>Y zk=!VSoL7Me-lF)Skrc7mgNopVIYeh=b}pdVbGU2pC{rQDZ^C|3+IBGuSUVrz%u_#aEan-B z^T1a&OpW|lG7FR4E)H_)d<=KLtN{nXbvUt?+XP#u8GS1KOv-^aPViQNFTuqCSH<5n z=k{_7si+?KKsTu7rSaN$%cWY3XCySHCoA>!ydcidUTW=EK8rFe)nf`&Mg8lbXP%d@ z z!y@b*9nT##Q9(s)rLbPlPUmf3!Vm(8E6YgGT!Tp|SGl(=-_aBKmKauVMO?~|ZPWg# z`(a9Z*PLx&!RvVVK$siOz!$GuVY_}t0Z`>V!#4wum4 z40OP-&atTCYsv~<{(u_)7Cz6j7p$`uU~mXM=c9fOuMOW(I5k874VD^X%KZe&3lQbZ zEl-_y3Kh9E*IJ^(`l}$)8*M{K`1U~YY3xaA#vVlO0p1itJ0kh9nR^E?-dx=u#_$cY z{%L?xFA1JrG6dyZ-CnXmLBe^uleP!7zMQe=k08Ei7sdFJxXX7GSO@bj=&FjK9QEh! z4Y8{AkWG>!&lFUPnBd(+l^cpc1SNEHsdAelxI$ryXQ2-pAplepyd$x<(T}{hxbH73 zO-%LmnQ{MXvF_+37A}b({YII{WzK7LV6|d=3sarWTz(1p;49cj8UvxD)WkRlqME2O z0OWXcQ?Fi+9|J<9X8bPBIfA-A24$^>?>DVTlvv(>xzO@lztf6fwyq#8XwrA4R%%i#gYPAjy!`EA%J^Q zz4g2GlvzRl7X#MwPp?0+9CsyLAN4gq$NL=aJY`~GUg%*( z4$BsufvRqBFB_n1go<2KB^U*!+eExVb^0MGch(!U9qE|GBq3=h5#6t@V=|jAW^a^F za}$(UPfP?e7V$#+BR`hU0iUnks@vKCv4cwPO|Te~nTIkj&)SSOnk>R#1mHH}s>XI# zrid@#LQ^$n?wA>8#&;tlTZ?_0$9Y`2GQ|YhLD`KL?S=G)d&w?v?VfL*%6$C~p3Amc02+ba0AIfjvHnY>-$ zX~RDVYcMW_kXojv>vfgf%MZtY#>Z{vV~vN+INKSOXp|tz&{$fO&KiKDr1Ko}>Nq0- zDmVGheU)UYV%V*DMur%O9fc-}z{{DO8fj&S^6i_k@QTgiIn&$Q`qw80w)wi^Yd^7b z$g7+1jM*&EKW(G{18^>TC|a}onk9R}K+<1ldf3##IX9+1AJ8MvW!v$?1SedHO4IwP z#;~V_BNCU>E?#G~L#YDo(WE0baKQHAf59RnLikj%Fb7L~j6=EnroYJVY}L)2z$8MS z;Iq!l7NxtZd-+78c=Buc&ELfFRYxuk6N$O^3}oWnbJN`lvqpz(%`FaB;_9*5<6)MG zU%MoX=y3h>2u`HyN!}h>*29_2`{MHD6{O=Ha8!jHv&oAoDagQv?FTCXjR7K4)<^Mm znXUGMda>@vXvFD-n|^Y|;d~+Vp7;zieC4bOaEpIJu)iJe{wSyZrU7c6>KEFR+AHM| zbGk#kW=TMVvFyIJKWjx-7L_4GGu<bPMBF>Wyux7^2L>1I07|ML=NkMp5;Sa)Lt?RQZ zEHFI1LsHQ3qH~&`aiTrlSdNWB&1$O6h`9S$5>SqWcbLn*?#2+p;}9RTpCSi-$=(ep zr75Bw$saFF-l;j+2wOVS4n7QWrVC9Oov^DFrc+6hZV#Xe4KcBzQuB(KE0)%KB*`Tcu9+(_e@wVS>GOcIQd-EB6zXQW--Qp1U5QZQ-j1$8L z9LH6;#2d$!<@VUGcZbdUEh5dG@F;J8aR8kXUhgS~%%p%}Zhajici#6c3wkTO)1)Ox zMOOdd6BmX}=N!<_oSn1SMs+;*l1-~h-`aaOBCgZub-rq|eZB|G13?hS5ii1D$;?QU ze!x$zTJr2D{)I@0HeUCXw_fgn6K#Nf9r#0q@|>@UmyL7Ff>U%{s^Fx*C_qh=-oA2I zP%iTV-}LS2P*_AbOGx$4+M{1F1tVtPQ`&(pFPuHX9CpQ9_Ujj!|L}+QmQQmVxb9fR zFPLWBMk0XeU&Dau?vI-Kb+Zj-ESQ1nm+fD#cMXsoqjf~t@&TgsP?Sg8u8-$ns(T3_t@Bu9_4k9~-^WlWeK z%*f;j99BnWyXNSY!{P(X*zsCN-$)0ZVCI*bLQAwQG#-`Y#)S<|AJs2JvleyLj3#RLLt=mbJpY@FHlRw4Gq4ydW8%VP@?zQ7c?*7^8kAbxAgvJ zDe-FV>`x%a>zk`8c#rUsg5R(=q6+AyDi7v&Iqq{ZD2Q&eUN z;H2CxgwyP`4>zoiZ3eJ6A09C}=sBh4i7sbuWsqUoBnjza@IAw74Gri`;LNN2RfLgT zW9+_bVI1VM%}N6tcg_xt;lA-X=D1da~x0#oBV5ix4fyqq@ENvYK|_UX(|5?pwc?P7~AvO-nv0tS@{=$m^VD zT=_+-(69Tnx)1ljdmy8#%!mn zQMekt2KH3|sqo~D?YG0pN^+SBF2Y=3kGLW4Nt7xmV}U2%@KZD^UITM;^9O>k%@$qK ziP!YP8Q7yZd+L!#|G7Sn&afTL!Cj5BKZ52f4GCG27zxsayfi!iV(b0*#8NW}(<+tQ zNz1aAaERa!&p)Get; zIO;pl4Lr|f#YqHVfWp|{f-6jo;%55H@(7mcpJDVkWcSb$0geY}<$7H%=l*ML1YM*V zrw#GjX5)D!fmY%hna!V?oo24D|Khwl%NeLlER`?_`-M3I*V0QNaW;+cki?4MDYO+n ziaAvg_$9P+=yjdiei=-ZMGSir8|iZuH3B~j7Q^6*7&9*K651zj9?7hHG+kH##>Ly% z($&6J-c?d=o(U+`4XEK1n!4(@VuQdJR|mG?EJYMsO?jlWxyP*cl@zsqReFWfDg%pF0fcgO79aQrU|}*1Il|UdnhyE_>s7lGVgHm*btZhiu;oLL>__GFH$nk@~5A9;#I!A^x|)H#B+1FK?(Y;wq@0$6yCio8CXpi zt&xg%&EB7-7cOla1c`{ca4=F7L52)4ACas(49E^>lqBIcv(-|BC|Nh@IiMvDomMgI z8FXoYq~2rng~js-^FRGq5_N+s?2PheD7yU^id6O(Kq3uxneeQ5G#rT42xUdPwa@TY zYptTYfk;*<;M@-q6yc*=Pg}wlU(y)F%Q;I?HMi=hsdz4x#+S1I?E^`w^{6 z4=Dn0*>u$=-=CmP28F(a*B~fd;W$dZ8I0B_!=3lbr*RX9^e4^G)1W)_)iH#F;UAUO zo!7P1L|P2IsY7t9UC@(J4hwQn?sWvOeVzf}!!!qU7=@|SL{;kzh8M62_6&znM+X6b zZL_R;2e!z;w|yG^vBF-H4P}HP(*IZ&rI-GSWy)tl9O9t92Fo|u(ECJWTf@{dvjxJO zvN;IqYU_ZUWNND-uzf5mwg=I-?lT<&o@S8cb_I0W4^4w`#w8~pa+qG2FuJmeAFfD(`0gV-a=};yaJdT z>qc>s`h0^*oySGpq`%)^KB2dxrjUcFK~=LVhmY5WfemTpWu2R2D8F*mRMfiDKs(nN z0FjCeh;D8Vf>A`Vuir$gy<7bjfQoXT=j1+54tgr^BId%xgB)tKG6O<0Zh8rEr@r|E z`drr5P}*;b@i2d!(U>kg#6opiaO1_(3g|s_rmGYhDA?I^C7-A`tU1UbM;|GwhM#6B zCM|x#(qeFyHycQDH?lTzGcp+<%sI^C+bwXD<(lLXz+x8XUKIe$vbq@>(_U)|x-)tNeg_FwGl!bwrw%s?AodsYgU|b|7KHX3)`FfkOhWz_mwuE;R z@`CUuCsYL(sy<$WBWQ;y8e7CQUCq-Ug3h-sszY|Q1& zgQN>5M|Yw)3IUXx=U9S~9!0?Z_LdX=L%QW~kIpYRFX(21Hm@X`Q{xA)kUr3M8wkeN z&?9LMzjv*elSw4HgRNjE(yZF4e5(mBx%*H%=R9=pA|T)C-L~?L?pS$3QRv*KCj@56 z$q@nT&R?LQ^BKx#Ab~RI4NlILPE|<-A)g>AVsKpbwZYC8{sk3D4BwxJ5@P`R=J6fV zvROUK{I#79quj<(g`$HvNnyF`SF-fsKqPhy+<-<1>k}*6sAp~JZ-0ANN!Jjo&Sc?n z)2CVM+hYi&tI<6{_xF%9rTm#r6OIf2t`D7L@fOP8$5xOB;S+&{_5QCzHGGp26C8k> z$$^J_7U=X#R@fb(EelVX+mL0FbM4p~%&*<@-!-Iz`FdgRHT)s|X@Ig#!%bKR@?)Dh z@eQNx7*&5}vU#XG?mJ8~VydP@ROzjSLi;j;*QZ0P8j=p`=Y5=;w$6@cmnia7h3Cir zUJ%1)BX@;`f~$Ss;wy?+ecVMvr}egZ8nAd_wS4Tj-u3frn_E`J*C&J>fnE8LHHcSo z(BvxhC8Yb;h|nzF+jH`h4Uq@Wyp?14pKTP}QSHseVW2no1YG}R1&pUw5HJd@P102d zasA)Sh|q*|3C1y`kpS%YcKTcdOJ(b-w%;`lS^>G@46*vR$8!Ok{tb>K(#JD={=NuX zA><|i$DuBjT?!AN<|m(0cStGmc}!E*Vb51~nDsw^=zRAX(*Uh@FDw2i^BvudL8K94Qr=(U&GGq9M& z)_Y^Kr?1Wp#u) zMjA#MUf!S}pBQ)(OfSxu%M|DT;TE~hAS6`eWzW1bmg8pPeiO9uJpJdx!Pkfz{eDd# zD>RIMPJPdYSlSBbH4Kt$uZ9%rMY8Hp?!yPURn|#Ygac%-%8}}O1!c0R4SO>S_TnW$ z8gz}WCDbcerzSXHI2e)z1#BYZmuHR`@9MAYZVmzt%+ibqjPYuDK!KgdXWt!RplyJV;J= zxeGa`o#J3Q$>nRPUl=LfbH4c1jzn`Kp2Kw}v|P!MLW${!6&l0rABa@p_=T_mUwdug zSzjy_qOuW)vHE!n#9o`2QZ%LNghP6Kod#FiH#qiKaEu8TsO_c1CK`PepAdpI!Pg2J z))&cgty$`5&oaBh1PKU78vCkin;UIsz6Tko$+Kg}#$DQ%uOdu?l&SoP!(@^t7W&Nd ze)c_C5V#K3Qq$Hs#)J)$!}dcvOz<|5jj>}_y~KdSIC483;~(D%@#`O#RR{qI~}&cWv=o+NvQI-6uuIl*?* zJbWZ7rL5fc%C-sj{3C779etj$o@N*f8cON&RwEFK^kol8Jt^B4K{78mp5d6o3(ZV! z%k{d?>zF+WdQgA+y>#1|cOByuk1FA%h=7v7z zvl9LN9mDEK<1g@i%`7g6+d50EbZKd=x+|DN{+jg7Ga>-L`*TtMq5sw^VVgn6wilBT z4DbG>9s;b3qG)R*SAoB(Nhq(j2YRE){|zDTfzD*^Y_3esZga80hFrC6vuD?S*c)1_ zN$}&ip}lL!=-Bj%<4v4YQj1nY_AJ_1utehG!un&QK;~&p==eBZ+NEeFb0?=qeetEw8mym_RJFG zcvr6`3L?||Il~O#B7RmzNagvaNZG4+xkVShq<@t4Q*z1D%aFTUi^@AB0CV8{;iTItj5u#+)UPqA<{M+w`zaMQvOpS6q6WE=TTXL~XVR-dhid0~i zqkiG_j6rKg&`7xyygc_W?_=glf*s!gvBtAQwg;WwoKwCxG$!w<}Wk#)A*-Xt~&5nTWV3XA(nuk0=}O+d?i2 z{0F$zYMN77lDY(zgqEj?{ONudx30rXXKehq{S8Q5%j1wdK4Sv4BRtyu9DGtmsaKLH znTHzPH$KloS6HhT8}rz3v+?xGfomvvD#eC3VB#3EGoyRbhr5^wbR!D?u%(<&6HPa& zqlqoo?@%{_*u~3Q`h69}6&y^`=tvpEHJbDrI$w)&y~gHvA3C5ac|TJFrPubwv4um9 z<;Bxr2wDQ=lRolLW3rFFNC^G7SDWwqkZ)Up937$O`}Uu1RG`Tcg1X_)pVZP6nDHd$ z?;{nCn78vLu_gjSfe42saFujLv;%BiY}^loCp15{EH2J}x%<5u=Ga)P_WMGf9NF=5 zm%J>2Xi)E6pmW`(79a(r@%2SG*5&spt3ILb*(8l z4^et@Gd6#n^hrQ%`ZHh5!7}29l%o6pG4&Q+QN3T-H!uvHB3*+@r*x;%B}li@HH37{ zFajzFNH-|mNarw=AT6CkcMaX}@cli{yVkqTT4$X<;Ou+fdtaYx@3!Z<4#3vMGM8jw zSdQEI3y+mqOB(x|v5Cnd%C3;BzJQ5B2VR8&kDIZ~$mLtfgT_-eZJvCS%yilG?c(xJ zfb${~17oPU0{%a9bRReQ{4%_pO?L}}f=~lBql`(Mi4R7(i(yb>dCUipzR>-v5Q4aQ zi{FRnr_uufn@G^Nq^=WD5=Z=&upc>km+*C6*~LRYC2EY_x4{)Ygc{xLV$lq#r5{Z6LjBk>wl6}h8KH1^+XZKz#uiih;CD~_If`e zC0F#V@Rrkf+R&s;KizApyBo@)XWZzzp8@iuxl;El=Hu|`0EH>7t$ySV_Ed(I5jwH# z{>M~~CA50&NcS9H-6K&CnA7Rvfl4HTpZ|px2!yFt*bUNFERb;<1_V(q=@KRU*r}mA zg55MJ=^+8*VTdE=?`Zbe2(P3X;6#6gs14!OUB+9fa)b}xTI?PS1)hD{wwh<*URiWvId!%TO`Yt zI38L1QK|#wdvARLAJUVZ_pSb?Yy5Xtog_Xx^!K?>zgs(<9(eRasvrm)so(tw?&%M@{xKkJ6^4td_-?yqr5D@SuBLS8sW*0UqobESf>>=h&Ni8#|qO zsh^cK`f(_Lyk+R~oPr~`^qQi(+vbt)55Yu^&``E2Uy|yl$8?Rehx7xXi5o}!<=ua! z#nVKc_3Pzg_3hZ{!{F9dW_*9=JqG07iwO=>&N4-|PZt4&wuQT~J~~-seurocVd}25 z=eAccg3;2wWTw3}HQ=EPrbTu_pwYGL-$e87CiKAX^ejFOqas(TCHmU)64Czpp*rkA zO#;-CYPyPcZCLQ4!T1WWEEmt?t9)paWeNp~Ic2}&=;|xTDY%p?5u!bedcECp*5)5# z`5CUji!M**9%YCpOU1m8^)u9-a!i)7Ok_(z`?R^mzY@mJ?v;^Cwt6gMJtn|yPHjSo$5U%S5T)3@DM$oVcl+o814|Q zcKB?$|5#R;8&BV0;<6*Z!xVIu7sAi>LD76`AlCgd^b$EuQuSN;ai4wp(r7;owX^*G zXjA!+IC^3rm*F{fy9+R&rxicrnH_Gx_sw&##2<-3^l0yC9fmA4s{~|NoFxa%E-D77 z)1=99#b5o$-xj(za2pmUgZ4W;KiUB^g$UEDx;ONLijPD=DAK~D(go}Mr&IBzt~OjN zo?Z%(hi(3Q1goI>v$HHmZzIv1j#dhosfcHbNy`(p@pop&5wGB0PUOIKw55n;2s6#( zuQs`_V8$K(7WF@gei4j?^0n}MO$;0FeneR%B)C&8l+apiIj`5>;!Z%WJswV z+RvY+a&tJfg|B7Ms&3umgxGGbD)V~B>k#t@icw8pk6Vh+N^3>%mI8`PV>kOXy(Il= z{5TTp=y?5t+RE~pt)Ewvf~&6U5-5zV`86o4?r}ZP$f!7{1Yyvuz=+fN>1RsS$#K1f ze{g#0|LQyQK(AsoGq+8z)jPbR$A+4kyQe1!5|R{>^k|~k_+JVM$@e?m99jw--%Ol=uf zA9EKe)@@xk|JNaBJwv}i3!=+#@e#hbyNvxOnX9fSu`N&u-Vb0x+ZX(0&3E^^THJ1O z&=}*iwvv;VjbJI;=zck3lEj!rkXsO4dQPZzFUKEsGVC9@n}9LzVG`=XNpgAQVzB=z zn8Q0%{`pNUNc)TYb4+7>`^mNMFRJS2(PcQN5h?1|-c0K55g72J)Ewk_|W?%?Hs6!#?|;y|%XF-c0<`E{d; zR&>>dC=s(gL<@7VzM>dvO4VpR1()u5g*I@IuuC84j5eS}$`0+3b!L-#-iuaJWi*2` zVI%w+G%U<{qDS&(mMaLo6W0nCV*a#Yln~n`arNQ=(u+okDkAEkjov$|(P9Y1TRkhq?f#}pSO_y?t z-|>$x|Hv-yzwq*$K)Wq$iVi2noZbxh@s~Pw_EvrY3~Ek_1>MfwAh~U*JL1rP&gfn@Eh7<|%e|*KxBQ*;P=slj<88Z67iJ zfK90PJBM5(T2Rc$uSCE|#pksh2|3ptnFV3=6gD2x8yTd4j$CD>v-U0INUc{`i&C4Y zkns~1%poM!(gz0DmMKh^_2i%~xuoQW%eNOy18o9!xq5@$yc!=JzcFqv-8h#JcqBsJ z)vH)$(%;*ntYY=AE9(w#(w}qOnQWz2C*7ZL@f}6k)8VAZ5WKWR^Svz4=r<WF0PwHp`Ojh?Pp69@hZ<@IW_*w$`(yoW=1@H z3M$3p{qYUIq0IMhL=40ywgEu{Dtt!-S~ykE12^@|*mO(!ALSq282Bu3O4qZo)we=- zXZ&H35`2qOMCuk{7nh6#QyA|wLR+Bsi*y0%Fj%C{>&9mF(&(H&cf$Qc%$9W%?@{;bhGOu!6+SGG3)wFkE+f~}gDKZSB0CH`|K zUA}a$sC-d@HvcAm6z5a2zDgx-R&xBEKnTWVjQgps1%}Dcev`ZsWNO4f<(~9?jZ>=~ zB`OiFF@l}NgbOY_0Q}C~s5iy`{I#0pBBa{vK34pV2{xQ28kSWl`(r#pJQ-ptn3Kd! zb_PC+Yz7Jz#$BiZ*8GdowD5t`hb?vSl#q|(<$0e^WEq&uMRB|>4)I8Np%%G;7K-h7 zV;qmvQ;&b#15146>@jWawu>O#%N9SE%Tp39%Nfb3B>@vA!vRSD*bN_Utpz_6gVv4k zDP5IL`_JiH)*YTPi{LxWv#pLwxcovRY~ktuM!ku&C*{`Wtwj#brCSBGxD5Dw0|8ZA z@?21wk3M)lGatMEKGoq5bfK(9uEvL{5etX0oIVu`V!58wKCzF+#O=lRJm% zqo*~hA`DITe3HqkIUb@TM86Kl8ge}^LFFm#WW|M*TE-;6E;JA zm%o2)60MMb{Rk-@(59t*FR#vDQ3(YM4wG{KbQ^h5hfmM9JC7PEj;Aw|Y-18w{S-F8 z(a|878#4fe*Orx6B;}ZUCZ1)Xj~u^Jn^Iafr!!#7$Jp)-JaV)Moh5tfEjv}I6I3!( zVXup$J2|s<)V@UJmGj_uQEL%u4+PlVszQ!JPm{1o3ciSS_aInTy;qn?yd!PG%yXtE zG`F)c^y)}&=5s116df64DfvJ1gU$sne^_9gKE(31EAx5b_zwirB`Bt28O*&mD`5tj$9PR$oBGg8ECKin_fmfiv zNjx;zw7fDOYGdH5h@h7xi>@LyeaVoa9E5w7!9)j$fA(p<5(7nldcGg~ZzXUvbDV^Y zA~WNZu!`s_OCF7H{~s;&J?vCsEs@3`EDX&=y3A5S90854d>ciHJE1wpky2?gYke77 z|E*^;TRop6?$G3e#*uSmHLUJ_xu0+{siNF!m@pnGOJ zt=%KV*dsQ--nKLNA^ItKH(+jrzUhprnam!>1d}nUB+vlTV{vleGgXV-;DN5hq6b{sw%#{4m$)U zMSerwL{1-OxkF5^t|>k(&gear|EC-#!_Gs$-r5UX3>}%&1LoQGwx-5Oq+dva_M=D{ zpHRPFB%ik=9tqQxuoWr4;?4iis9>IsgGo^9-S`9=oX@FxHvTFs7n{P3b8R|8@?!0R z?g=NYyL&Btvl0zQ1C_f0jgtnA9%X0?FOWFjI#3xkG0D7`pCA|?M(6mqK$cCS0)cpg zDI#5u0;K!52a)@cIZ8{H-k-c6_T=TiCbxTO2p);bYc85En92wYSIjeUJ)1{vj!!{- zEu0pSk^sx7%Xq7orSij3GdyCs53L(~#VmT#Ln@CTNMxOSxNsEWpEex}Q>HvZIa;Vd5uM0YT*zo~Z0>JM-$Zq1H^!mpZ z&pR$FP6>~=-Fl@n(rvLvu0J!5l>#bbI|8eQzcWy05N$Gc&13dk9+HYrnxpwDF#e$! zy=ZCRRFX})52-cB*!w$5d@IQZJ1zaNLws8?Po8Czf+uSx!~4ai27b?X+tFB zJa<$CU<|G8M)p+GH~!_&gn%^_!%MkE?G`k&kXSmnPM)XpmK^k=bBDzJlTk5rKC(xJ z+|$AkUMbE73G%>(>WfAb7`s;NY}SKst^|b4U+{Z}$gI&QExPKzsI=ib5S>6w+*GJ& zZ>>f%hOyKLorUhh3jNxykX;%p2HkIwXKM|xU=u$j7s4s-e*&THXnPq&Gt0QL=D&Yc z9a3YE%Hu$^^{>=@vDpShn=vYJZ?_31C@@$4ne_SmCZqZpR5^{Q$@jSlK^XMqL1>%R zuxH5b2~QJti!P_T!UrnaIbl?AG!U}}6CIA@uIz}(CAQgG&{k!C3*!ACjMKADz1qU7 zc^mLYsci2LqVcJgu@dZCvh-f$&H3{y?ZYHYQSu49$G<#X6+3EZ`z@zV2J3!WKLzXzB7`|uP)8wu5lO@tQT?>(j$E-e+Wch+nt z!~4BTT1|>+1{b9E=651NTX8a%h^tD!Vxz?UM9+@F)f!lV2=oi~kNB?X4g>?^4Pkxf zl(ETO10fn7nTA!a)ZNBG4xd8D?^RXb1zfkIrgY9G(W^Lmax+du14~U&DQlw>l33$k zhrqT~kD#m}vn}p-M1f8AGh;&MThDrYCqO~8T%B`^+|g!$f&{upHiYTI+fpX0(9&9% z4~+N7kHRKV#3%1*&M9}_Qw3|wk1e@v0yh?;YKB>IKIdDw2V(a_o_>UlCmQs~Kp(bD z+WyQ5D)!1cJcPPRqyN~^kh6+#XonF_`6`=xkd6@fb!Q+G!#*f*Xwum7$ICNBu4NNF z4;CgA*=4doi7`eqz*)eEcyxY%&Ay7gylJ;H!#lJkHYvP@?ya@h#hE~3m+neh32T42 z4w(QX7gsH`Vt_v*7u0Vs)~zn&H#e63&pV0|_;Y6XlO%w#1xB7}^U%P*l|Q^#tfv}9 z&&*%2i3vDYyx|L>b;V5~EZ9R|nx#-stkIn-Z^_y0b3A+)w$EzK)0nV;m4SKH;SG4} zqy;q|)O_5yJF5-%e?RQLod$qGjac1;YF@CZeqE(&2CnZfH?gA8ErgYDFa1aR*#{=^ zb9(&juJxgp@p+l43UXQ0SK%5M;y7Zn=nK|D60+R8(pzRkaE??LGaI?~3Nu?CWwmlm zPu@{m|BG7=$b%1H4ZfC2xWEX_9k071zI^H>h`cue)<66PNAyC4EQEdf{4CRa-`d}3 z1z%2*{=wsxE#n=c41kM-mXj!QiGLNo=+F^?gpipIE|AAbHuk;Pyk6&}32z8DOSZxAk1X7V!K2&UkN z>*3*GG28hQ+v&O`w2QrhOK*Dy>P}J?wL7-{cl*2B_!9gpd;8Mb@qbo?^39=ik&si3 zy|rxFEzbn1_a0uq4U&=kFHQ$^j-N+Wg1i5RMul(v1)NV^cVAQqpF}9oj;(Y(+;d(W z!>VyNsuLM(pHkSB58HG?xhj94x4ZWgrO8UV1pWngGp<)+GGLGH^XmZ_cF8d!s0hpv zv%>FTAl)-ye>8*H%JJ?jBA!e8RaIY9r!bXn$b{C~Vb>IK{lv@#y>6qxGlqe7b+RcX2AgyMLCplFI$C|8Zt)N`^c zZ<{3d?`S);k+bh9%Z94CiH;#bYQuR|uMuVyI7gvzQUZPFFO|VF^XV}VFKuD=My#CO zr5!TJNZ4(s8;Ud2$9RI%vY#vvd8CAL)^tpJjh|r$ZqGPF#|aw*_Xo&2?UM(dBLqZO z7VX>CmoYNd(zeN4B5c=5MahAIvL z13!zda;UW^#+}VWEU0{?Bah(-as9 zvVI=ZjwQn;5YoMOcsT5<2Av#eW6A*Tzqh$ZPgwu(;07GqvDGeajC%bedfYk4-l;8N(MfjTG^rm&Ak{BKgQTLIHU15K&kc9QszUNv<*ukI_>7pKci7!QQ}`d z=4XAdihZRP`j=pp$jNIwkffwPP7nv>7ABFytBYSeBzODwv;s_6q-kdd)*4+vidt}N z!-6bI91x^wd^De%KA0VQk~kSw^ts4AbB-Krk=u17N+5<+Dcipkg;<6@I^3&MR^kpJ z;F9&K-Rf0&N{ylFIl{vnJ8*2H#LBca&CvS3aRsr*h1zC|Ao(Bv7tHt{Z%XFtbdu|k z_}0awu5D4VrFQzuU*2(rYsL1D|6d4w{p)-jFCB3gk3TZtN!|a6OQ6lV;6Ji;>8Ip- zA`aK9K?#YQ$h|7A-mzz#)e25N>A$#NT3&tmkuMF$O+rsP;yO2PJyFt(S)j&^S`yxw zo1xRb>sq~yya;V8)o2M*-A=zwx;N#=hVNs8>SnOBWp9h*b@KH>7(`D9=1@xIG?V05O8`HOZOlm+Y}$X4%>?oyFe-w@h&dqa473gSMeyCkfvpmQ;(C%Yj#{1j1ERi_bK~Ip8M}OL( zE4EV@#V9NKN}5CwVmN;H+w{KpQ_L-wBRR@tz<<_^I`C)o$kR3ZOm0~$fO6E^x8ml5 z+@oS;dB~}?)YA3f#o^cGBKI2ARt&?lwi|DhM4Zp32`!Kap(*$_Kjae*3vma?T9xd- zWoAh>ECOE+7zw}ZS-jIZ!g~_6J%z>Kq;w__mG369Tx zkMnu1lk4ZHOmXE{=O=Lpw^(n+=!T_BQk#-J(deIMC3d`5jUpj#yH2A^28u;yk9W!c zf?dL$Gx(pJa#9DJLj@zSvZ);b$xMWpxIm8t@=N4O&$5LCTc-|ipBTeG4Gg)e*nZYb z<^H5CJ&xYO@9h1@e5rE&Z{DKRL@g9yKJKA(e$-d$(YSVZJ!L9;h1$zn71ZqdtVQrY z+N4Yif48@EwbF}<&Z!t}J#hK}bFAM5joPgbtGzG#s3$m3Xi@EA4iAlSe4-zh&zNeg z27kVT{m}ppzcc4kXp1}lxR-=E>=q%QDB-_SJ6*A>3;Kvu=`C+trlzn{na5Dou?NO1X!fvDFSxjDP{C@2v66d+de7#>iq;Id$K&?N57)+i@M_h|CmG@iouU z>LzhcUf6w?sl^*nNKB)u9V%MuZ}v5zn?BAa`b=c6dE_V}dDC#$9gX^o`8gMBrvcU= z;05lKC#enjL9GK)l1G|H5(A1uTX5+n?EN%*#WJmV|JpukgS5j zy(T%ag!-XtKoMheT&pc#9ntRHmDw6SI$OTW_&gMxpIb8xRJ}cO<`L+-1h~%Dhw97k zRQSs&S>#On$GI$R3|=k>Al{wZKKNpTc1_N*5l6~M(s#6S!(`eW6c;f)Pl&s~XHAf8 z3WwUplfmQY2j=K}u|>&o4m|l?*DZ^EK;ZeLmMn*L(8vVgI@_b02`n5|R~yqpT}j`DSnp7Y@w166=aRXi+EP>3`w1P*F0pKI`%+@z z!4GML|7Ep1VzI=s$^>t>viS;UA8*8>hrw;(1eyi!#>WGjm(t;sFv+nqC>35)8f7cZ z6az>}LA z%Q7`Wg|1_+mb62Ua!Ru^S?7;c0>Z(x*qc5I!q*}|QqM1UG={5P^;rApJFWNlYCx3B zxw^JDlwESWh#T{+Wvh+G2rsYgA$Y2&*UyiUO_mk;m`-=NQY_(-->a*8Fnas|K&WbV zqTNGvjOo#l`zr3Kua`ZL^r?c_-6LVmo$6H+-<0ua&^8Z#asnz~A1}V=68A!MPq9{u^ffBsw^*f2bx#*j}EG>Zaw9*n*Yu7o}0?oOmb zZWS{u9M!74Nm+4qWs~k~XRMdej&_d}Y)1wfj*?+$#jzP}^eJPZ12bB%kW)$KTsZbt zTtK}w<+=)#Zbq7z>*upg$m~bnRmvY-|pih|6BzAa95YU&8 zpZ|fPo+(AvF@p1it8^=iopNF(BKE|$*&PL3q;ay-Tva-4-XoL5sgB1ltMbc#gO&+(+kbRZF{|f|~Aoi~|mStR5cP zR$q7)KRGv%f`|!HEWIS#BS!4#(nv7Z)Sr|j_dxDM<#M|Hz^a0i#56b0m9RsVB^d@P zt=Fp`eLLoE*Sv$IF7D#G^{Ps@=F+BjAs3kx`Mm!b&hL)1vV-SyePZlrPq%ervSV

-^l`(V_b@|?il*P5R#%0LOLTffnr*=+gAOk>c?mY&?|v!Mi8a4`(>(<4 z2;f>=PeDU^e?1p4$7m38b(rHOQB#TUVYkeVZ83Z!!?%d*=n_IbNcgmV8X(tf75Q2V zm`)=gJc0E-UcLt(#2lf1?1Tjs8i*MZ!MgiK?i+hh-cC#P_;MI+Pkl)xpyR}^-ln4g z#Maok9LU0#JoMw)98QK4115_WVCkStqGFsxy95=w&)Q)uGM^M1x$Mn?)MUh1ZAltu zVVr626Jsl*z!76EUB`Y}>0K>B%JpyM9H0s{9RSBLW1eNg*Bn)2Q%;0Z_X->nTFu9PcrvT!&g>MYe^ECw{T`&RLv z6I==QT2EWYs@DUPE)kzZw$*NdeZKc`d&m!{3(bNjX}Y~(k= zn2eNb+i&tfn{jEMK7?yO|L({V!(`2{h?HhuW6&xr-A{JmXV`8c;zB<<&dNxVn#!Uy zVhs1W&;jd~t=mQ8o*Doy}B^)3U7AeuKL z3#)+QLBW{}LcJ=^MC=w>w~r+>)ix#pOwik*&0ru2U8ng#NV40Pn8!6*a69M7nze4S zw|;}5#fT=H21{~b5KLD)SiqK6S{aqfZ^NL+y+QmJ2k*{OYgXktf_Tf8WPB!?Go2S+sE>#H+2Z&uBLn)11OHSsm;Lt{NBBhkFL9I zs=ol;()?N6%JlQk$X|@}9k;f|M7>6%=DLSBwxNc-C-Do<@+3E3d)R?Q=!>qkIGW>8 zX|WG?pZG#t{s<7o?lrrYJfTIR)3lkq+3$(syn9^wIjp+7P}9$Rb}lY=$W}C$ID2%A zsgIu|GT#xCcJ8SQ`!t>Fe`kHhxkUdt{-3n7|IvIkQcX{LVCU%yCey1Q#yhkN6lUY| z)w)nkFV%?7AvPs8E&1i7h3%B)!hyCk4k=T?{)H+EO!iKwA35H9wWCTWC~~*9&AOdw zEh=|^*O(OUgxp)-1dh_H~ZBdZ}<5DL8 zzB&R|$4Z<6{0{EBaA>igCGPE|r7Mm2Vmn6ce433o0KKdKo8MSRmc-J}qFYe7(YFEO z$4FoXQyJw+sEwX5SXme6rr-`0m@Ie1%GfYSy+#ghd<@D zxtD|Oi+%fajkvcDo1>uvfG1T%@oubbl8w>#Z$H9b(7;UDa2NMbodE^y*6SsEK7>wbKBcK^q5=*iZ)+qIKGRK)y~ z_5|>8nby&*p?2rH>$!vhr+M?KWx5Hl02?8`f@U%f2Xl2!$vkZ?l4?Ew7xR9K6t`iI z!~9dq<@mmTM^Q@{ekYsssas@JTWKUR1wcbvqYGnkTK5}+7mSK;mFmNf)mT40TJw&+ z+wR&!z3b>$3UY>h=El`{X2HLDwxB$|WJPLyAtrNBK{Wi(;L3UdI4Y`5GJ4SnZ?jWL z1<*}OR8#FASzD0K(GR9~V%=0?c_No9@9|+fruJHohoB&m(Wx6LLz(^%8Gm+6xLJYS z;V#I?Ie8HOQL=~PF#Wb44126zCH8=0CNX*v;gu2@#k+>2ieluc)=zCNkWADVwr9vd zFC))}hIK&%l-0OB9R=3J_$FAZV!B#>WLFoWoc=QCYIR@w5Cb3~!MeBt!2LPVjSVV2 z-A;L(Fcw+ElHLp1n`#qMVU-h$4IZY(iR;quX?-ra7oTh=jFz1s)NMv;ogVVqiAq}; zEXVKb`bt@}q1)lHIbpJUoo2na%?LtvtPDJTso%8b5>QvWfgQsrJaAg^mQ|~^$ zf)mU^&mQntxKw<9in!7kf3E^`|K*VSoIv&XtVL6#>YdhPM5v!ppYS6O$h)>I(E_+* zGVik)_;k2D4I3hom;U=g@s4u0I6k(6frK+ft)CZqhv1$304NT>{?P4zJluy2qr8eN0sPGS3QwK^W2Qwgka!Qeo zAjHK;r@5WA?@5KS^gW$gv!Gr3{{Jh7B!XN$d$AQt2use62~i^}j7JUlEJlTbw2wTP ziMVMFrkYm=C+#{QAGizO$?dzKa3NTC^jDF%i;gM*1qH=c^l0+vKhbK684qXCg>S?? z9L8Oj>cU79Aft_uOMMYd zkd&gDG}MFgQLvtfEt8+O@dVxPI*G5!CFXaSIY8%F2cuTm;M*m?;ieKjQ8QjA1CtIP zG}bJGXCK&3?l<<57Gn9MUyT{zzUH0Y{c}0e=hLyF9E}ULB_1#))|baqxQJ+)9{eu0 z<~F8k5Uae&=esz@!ppVWyh`-*k6U(!9iTzVC9*B3i*r3B*4{0vd6DTirfApUbB*dy zVdQBhFNG*;J-IOS+_GLvKCS~PCLYiy=$6^~PCy!eY+O?u3@1W*;F6Cuz@$y(%y2TS z&_cRPxqEkmKJnULn--h%ik2C)NYfFdnrWG~h!J&%-ng~By3!a75Ze|0@@_NP=DheE z-+ui35|Pen1OND{(R;BywHkjys^TkB3!%?*+ci3s6I8Ko5CJ{X*0;LU7!Aq6&+SP6 zQ7w6jX77;2n6L=)y4-VB3E;R7obza7zH+r4y3D}LH}`k9dot5ta#gtk%X6RQJmGmy zn7suW*I3L>SjXhUp^3^gB3Z^Wb&srS=H)zgBDPv)-wQe7EL9mwp+2Ba*|keC4bIn9 z`C|bTgEIXF3VK!0=mXa+6T3dR*e0!O9iJ?K8)od~&iW?Q!JylQ2#4!w#O zYA+U7&UCp}_mM4PVk9ess^D@sHK2X`JS6X={cOBhTo}q{E(sFEcCXDZ4J7_7-TBXs znr+&NGjS4QIqq;+r2;p4x?&+Z9Rp5IO>7n)5^#~|{MV$~<1T*Zpmv(hAS9r<+FkK! z#6!pfx?Cyu#iQv>miMq1{5U7DATamh_3Hf~_Ds)7z=!8oJwUkCHI;nj+pOAY?D!t% zL~CV}#DzsYbbgkC33ngB2$O-{lJw{npXSmnV_%GGGiw_nrMPVqO*9FE@3;fp*|&=IJ+AtMk02Ltj6$Jr+6sD1HhzGji#Ync6@pDnq~JIpTm0 z)>|0|v0yHBI+5VXglN9j{ysVdGZ16!(O((r zBVUuNg3vm>^`@K!%YnooJ$q8V4MG~HM+$YaP7mxsti4t`6NS=Hiza!8*<{%0+obNv z5wN7XAFScph{?@xc6$K;64*+eJ{kn>RiMbB;~+Qn-xp%@@;Co#Tn?Y3k0Wx94wMj^ zgReogq>UdAqczMi=WM~Xl`@7Ob9sRwSUiUf>g?F71OqJ|R5LD32c!xe$7|;<$Fv>) zwrqEx=>2D&_Xb14JK)}iQ`<%CcDGWf6%K(c{sxL_JP zTyfZzM47aH@2SSQ&>X((EP?Iq1~J(HikcRM+Hw&O`oohe<+6t>7CmX$-HB8D1=)xg zQ-c?1UKyU>S;-~O!3`QuiCcI&xw{hBB&$r;?qP|1cHxb9A3F`avp>M@R4C3r!{d9O z!eK2%-skFHTaRBmg%p{DFPo&A|Jwl`DXDa%I?1}k&yG#vs{;Wn^Wl(3pQcl#3jp1! zqt%V>PM&A_r_tV9ni5EZ*0-)7w5A%23!23^#!(&<21NTuTO^&%IlZ?@4oi?qy-i#L zz0txwv2z32--UKQNWjwMpyT1Nr9ZEI)`3J0L8d##$1{Zqz*1p!aTcR2b&7BwDZz=u zBV*Y)TBYrz4BCCsMV2|eXs*vkqDEe6(I3rZtzEhtHX0bG5lm0rq7HHC0{87)=zT8Q z`Y-V&*%F5bRagAl-k$lK1ywcgRy+@X*;*HR<*o8t^#i#Qn^G_bai^tU>UscOrd9vKW-poEVj%woBdpkM4I$RoON;x|FSxz4;ov_FR;PTA)j^nIp zU0Zg?!J+5m!2wfS*WX)_mZvk5_GcwpJZEFWl3Cx*4qx4-4Ni4JYDVE``|wfIcSfG; z&0LFYFg3}snf6Pz<`4WXl~Mt8QRp796(GtsJ&UzJ`FbS3ozs7@xs;{R|C|f=@oTi1 z#18yQ+EAuWFG&K@V|)9K2m65)U}ETu;=4I)ih8eo;#Xd>l-({Z?;1v2T_|glZ!Vif zK{75p4aY`_qt=XyZ$q@-W{EDI?9J65;}yg=}|E#W{OX9ruX6dg{2C?wBEC$9=me z@IFSL=k(v%q*Y8o(5YmW6`lK2gYz-q>OM34ue9iqB%Ox>?3r$9!T_(r`rrL~g*T1U zk7?{y42?XZfBc{0w9&K)p_Fi1XzGNvm0kviQ+{eM^68baL9^lrz>nIO{4hd=dl>AwbzvqE6`46|6@_=*OX2%6`!(M~gwv-Y zmGkf#(1W6UJ0@!&^~rTzC8u1y>1@GUFw8a5fAFxGsN4I4)-m}@^AHM0nJtCV2nP+bvOV>Z)3l++tj z-~cXp>*E|=h{_Ge{mUx55qH)yhDLC=mc}V;@T~!hWD4M-?e}@tCG7e>cJtvKC@S89 zEGBef`XP;(!g+>1$uq~|otFJxM}w$Sb&G`W3ox&yBXNZqauc4P)9_OS3VNg5){}v@ zZ%w&t4cR4-w)ZxCMU-ae zt!;8mnQ4cjSQS;+TRMfIqZ%c=1Xuyi(T#k-Gbi|+qmA~i?eM1|U|X`ZVLJvq(}?{L znNy|`G(<}9l(jvbo1j7R=vItwNym3Mj^TSYU^X^|acr?A%oU}Y7a;`u6qKz$G_M)TT*LrhBauXp~C|R>9#P`xT-t*AKod3 zz9znArHczj*7R5@AotEWS9#>2#a_lM>{WZ=}^mAm$h-)+&M3|Q<=7gk`5 zY5!?|NW8}ux&)O?zFv8nT~TGNY~kcpR3OM=;WeuDJLb^sYrN#0$O)M@uj(-;;Jb0{ zS9VAA0pj}0jk_4T!`rZzK~7c}w;-jxRS_9b1 zh!#m!OrXBEWEuw{qeeV2m6x6GkNiN|cFvE28C*?U6C0JF!>MxNzl&LeBa6Ap`t_RH z{?_LWc%LvIUFDSa=gnJ1?S<-KH##UDwY!6p}5cCgT$z4Ue^$ZxxIWq$pdBo+TQwi z9SKC@X5lH)fpCkJq1gW;kuR3mEob6W6@@C^z)3hX4^scR5PKWn`R`wVoHI9bWZ{F4 z`+bdez26e52RhbkGCEDV{%|5OX3XbPy(wgrEtM&70{uH(z1#t%?Rs`9=XQw$hTw8! z+vkwCUmo#3+SOd&JUiER5skCk+dSQ3vQR(1o!yd`u^sf?t9EY{qPjQiI#qa@_W7O> zF{qGjY@5C^%8K1CB}pz#g|cdrPRudB+y(0gO@0{jv;*u? zF+eJXxN}h^ZI84J=9Lt)WXEL7k61q>Od5JukrT!noYHNZ6kd&i?!R;11nm1F_U>r& zeN}iEG434|#uf!KHVBm0?*4OOG0@V8A(%%S)5*Uc#zfn((QLrmgr35I_3h%^QsrwS zxi7Y44Vh+sC8nCZ;XDL1O5qm-Cq|fY9LhU_L<1WZzmhIDh-kr!z;gB_m)fk57R|uD zGfqn>8Z%K}ioS8c^qe%7MK*z{5JFeDBRv?MtLLQ&wgzTfhU4U`D*ZtUs%pFhIfpKK ztJuU#yzGo5m@v>}eL2SYR-HsgMgCWEl5^o6B z`w+Tc@plJ=M&%EiugIBgptShXywgiBbcwlt$@J&H4Pf4w$Vdp}c_BPo==R@sl&bnE%h zQx`Gx!W0E%jIr5XBkh6a7EU5 zrH=eXr`TGCJfbQ1=4A&sB&+kC2sG#e?Qaq-tUr~hf0Q*rp{sOgOUcxWW~r9VwC^W^ ziSkV9^$Yl7ljiS+d>}#cv_B7J9Aag9i;I#J+VSVSwSY^MIX(tmLgab!NwQ)X>yZyb zC3Pkv@C)I1V>DHKFYjS~IKrTXkrp2tkyS5lCqM1J2#gZj;{uatj|FK{3F+ zA9R%w6&QaW6e7bz_$%(CjzH7@O+j1H8vG~tv9MvgQY(lyv1@KS1RQiuIh9PCzN^;N zaRVQ6b-c6YSbBXBXC4qv6lt>0_8i@>$?vxy-XPhdx8>bq@;sLi86_k^Nt)9f=uk7E zZ;IMKzVmqNspe7QLCw=$#rX~BVxOct;jxwi3rsZrTmOFNO7abd5pOT*e#r9bG{F4EWj9=D99MU`!mP!6N9PEXdq}|h zYp-oMtI;k(g%J5@|L?`s!yfO{(JmkY3!EBUK~LulXYZXQtvWm3S*+&@)0z&f3N)gT zT<|T_^w^A~j43_WWJWrm3-kxgcsCJ{(Vp2?+jq${5gE3k5}XwfkMqm>-jwUSCMTxR zX^*!aSN13}y607oW~Pr^6qG3oZsWj|saBt9=D$@9J2#EGguwygtRF84TB@<9 zLTYk$;SX*(2Y<}%?j9O99-~+T*0vq_ZhnVv8EdHOlIc5}!0jh74bM0GtCz49Xpe_z z$~ih|7fpskg33t?9x2@O8ujUL86QM*R2)>sA+~r4j(_9x1!l7)LDeh9oaoC=I$AJSq(JmI-)~&T%#LaTRRlVUGjk$$S@<1eCjynGz0Kh;$zZ(czQ6*Ac z2JU+gW73v@*24R~FA9|X8U39pV7eaI)xHv}L!)Ll>++ivE|G+Xd^9kAjys@5d2jn-z*i^Zkjo0{IgeX0fjq~rmgQ~Gyxw8HF;B9qwRE-cYT%f8 zt75`E5VnbLh+lwxYM!H!)j`;5DlH8Jf*uVUi^t?LQdgh=y~qx}&oa8YlEHK*sF1xP zNTcmUHJvN+6&1w^=KJtximj zkb+>o6_Gm-kZu0it)U?1G*6U!f>21lB)-C1kwd`=c<@75X>cgc^N3L8^5bjp}X3K6L z&{Hpl}Mk6Hrd&QR**DsyslfJSZK ztWmT&G+))puVkMm(+2OgHW#@7#dlR`9y1S2OUw(TUl>>|o}=RWn5$5>VfE~>(|aOe0xE^par>nAUSu+h=+{`G!)KB7CuE)R@_$ak}s3(I%B}r z+CRn84dU|MekZv#tX}rnYdeR4enp?5M}X^ttYlA5xO(DD>AU)qsZz#!!&mXic5ltS z7#4T<$^4XV@n2+&d@|qXQ?k@c@Syk2k%=|oV1-T7&%F=FlpU8YmWJe=50|7L1$t4N z$xr>@eY~BY)BSln>U7^?^%b~${p8BqWuK&%Q0;!_RNIsFf6o8q_9tuJnE`j+x>D!i zLrC88@TYXwlXUN0x;{1`>B)N49yTpfst~kn{Qj^yqoC5A`nQXB{_XYQwm2_)cfy5( zN0;67KFR0R6Q|Se?gZt&cc9Pr?CI@&WnIRpHgIs`rZH#jAU2V12nV4?=qsB`e+8Yo z44u*JJt$HCrFHo>gId5>QcZv@b(a)4)E=%^KbLWT-oIb3zc;S&U)&v-vhm6MjEbp4 z9E~C6etbm6MmREP>$&qK;c}A`Nw>HEV|@PIeE*!Et0zz2xHB8#(jT?^2WxLSlEyt* z{^#vHgWr1Bsqmj{H@|m3`~UR^D;3lao^0pm_3s5=Uy+av=tUiUvT>X>7!1#^T>(2& zc9(}a*JfpWehzA7)(Ne747x9R@9lS28u9sGv23)&Sw4YLUySKBA_TIVvX9KZ8iwZo zC}`M9xKZ!=mp4utar$q&5q6{X&3Q05Y2@%x1>w-)jF)Z?1WVuC6T{QNvNU3B&?vBO zi76X08iMCmE>qxbF!c>d@LOWUPa5y9S;HY7i}>D}quj?V*`_?5JVf)?e)mmTaIa_$Ka z0QtQC7VCZRlukI7J{LUo6ANT~58?vzR_Qos9lal@yTOw)>D$aZpeLzlGRW`xm#`KA zeUf%sUjnX8aj&Jm0h;2Ot#kt=DC3w@0n+(^Z^3VMu9gu^K(v@0asW6x-s+Ih795Lj zHcn^=Zspgm( zu9~1(KnrybR=R?+H0^EYdhm=)7?}>Z+CQkAvJD(XZ&0=C#Hf_6>mthNOGB=a~4dKcxgIuINsZgz>U2)wV| z&^`paoNHOcJ&-bOe&*;W?L(X_e=p_(VUBo5Xa>HYG9S8%gXvCCL3@R#2F9=1&X~sn zWrY3w(kqRA(#tX#)EusE$}#Y@b0&J;2YFUNCEI2IpuOQfp(vmpcMnkB0)FB{w*89~%YeAuy58gj zsgl_%?*Zu>zc5QbFqI7{=J)hCNTQTe%KPM_-b*j4{X+U$l=qbH!c$+&6fX$n|H=Ma z$3DtZ2BIf59ak-n9ZT-{m4`KI9ye|4v9{*#7DF zD=&=$Z(mPK^$>&>kF4tVEkxR5Zu)2a`u3xD#uf>D^sb|0yaJgYruB8Zfx-lE{6C-1 zXC-zLw+c6aAO7}x$KH%Lz0HAooZ#~(-=~jFwmUZhQ-9?I5Gvmgwm$xF;o0+%Ykh$= zp(TI;fBSvb`SkXli*S9yg(695z;}GeOHbxsS$y`bggQ{}NX>U*JA+@vfJx^0Po=xD z(818n5clJqa_PoyVK5kew~&t>Uw$e5XnNqr;~8;AYz!}w^VyGlaib=rT}<+2J@sz3 zc0&ff-)J4=DN*!LOfldBxfu%PmX@<}04!ZTz2+9+Lcq}J&A4A&1#O;{_ zVCK95VJoIQNpDo_V%xh9AUr%9LDONu-LZ|{QtL=F%C5tAjR8`$EqRwsroMETPc1&a{PXa)aYx`uDB@ zuzBJV$DtMf7_U#~eGs;63gpXIJR$pB=vR0KZ(W8>lfIjE2sqMcYnAg8pyH%@2Z|Mg z4mSq78b3olq>;2KYc|Z?HtNvYig0t;g}o1EfIQ!%OJ%@kR(@2ELD^X~cNgjc9lsc| zvPn6Ry34Og|KtMw{-BMoo&Ngqhtg2!c$vz%hCu5vedK!Of!kE3CKMlD`P+Q;VQjB= zYAvV%NzEP%OY0AN)~+pcvME%3*tk>4GQYoG^XDtRF~`^cW?ZMhORF2!DBb|hjX7N7 z#t6vFN=gsPbg} zLnk*+Ja+&No!lID?f~>oncS(>04PM+b-8}S{--t7F6oQFWpRx66#IqwmDCjNFc1p( z|8ea`tS3DX_x%3-Ouw$r(W77`vu;zn+ zv2nCD7!1#o3x&^jxR4IxW{&ku%mxKo7VJ^rA`Iz0^!q_^U>akpY^rR;>3=P>cxd~O z!!V)h#CJbA1@A5Ek=1(;49+&F|HoAU1%Fm&NZ zzTw}2`$K(YDXq)V1!I|sQ!v_t(uU77W$S_dvIm>%nfzl>pud60lN`(&y{2bIh?)=Kj;Ile> zo=mVz58oLO0j|*xil&ADi?t@&6!4Stbc^Z@S-Y=ZjT;Cq+x=0=qe0wiNt87(jSaT= zl?QFUYl|`uJWG=DX4V8tazwiT8Dyv50?Q;&Pdk&mt-!tFfq&}9VCfOnD)=PmihfV; z3Hl-Bw68Rnh6S1JUNF5EG}(R@ybV*UXB`6PxLaou{XyDci!w!mJRqQ(?J7ua+lS^e zAQmxIlynfxQks+i!gz6(;1Bv-{l2~!^h)YI?M={YE2Gt9&_*a-)vMr}n0d$*0?P8t zk6hn@_|WW=Yk}=p%>B@>Ks;&jlNW$K!M8*W2BkvA2*(xBaw)6Ti@;sJ;wSJu%xdf2 z1Kw{ELo-4_J`>p8{u$sFa|usf!`CPIzEXog8I!rxy$O`N8LvApgIqN*r@aGcYn9vT zY0#(WRrGm4No}&;4jdADl1DrSl=yA!Rme#rp$d>wXh;>%59odLDZjnlOKt34%N>v$w(rcjK=>Wo13piR37k*PYlZV81zrncG`CkGNt02{0~9gY^`KU1CXUpv zK>f%)(_0OI_n##`#vSAagW+GQw7%{)xobm2n{3zo`#t$yo$0%%TjAQv=Uz^n47ZzH zX__(}v|UmigKA*Ycx+iWxR!z{Ov~~I|$ZDM4-!l|^!+hJcUwC(Vge~U7fC2yby*^Yg zKn$dBPwCwR0=z-)SG}Fa5%HAAg@57JtOy_=R-mM^v!; z8T*65VEA{HEcIx5AmGuX)puMpVq$m^J^1$4)ub{{KKnNd>S&t)dDa&_6!XzPE?;<7 zxxTq_|3W&S-fWtb`UhNKTmGP5ngS;t?4EvkE!eXBJ6LScks*tkG>5{=%DtVd8Mt1_ zI_&8SJ7%pmAHVeL>3M3nUh%4Me*U40H_k%UqfPghvVzo6t}iZrlFqROJJK(nhp?BU z4+Qx^iK$iM^Zml3zZ;ByJEn36sAR3*qIi`*t{zropLKd$-^T--yP^1&%7^kbg|07; zY}338L|@F+GHe8#=(W4%N;0(Q-eYp@+mKQvVN{00 zIL;ajhUchX6#U@+gP}0^^MTVpt`63>E$>)NFrmk!5613=pfN!Yf?hM?^kcsVf5xa(69Tf}C6z4)$j1i|lnzx)czqdV%nl$3jn0UDN+THE< zw!(;kqbd)Jg1lD>l+G6de#Z;E8odgF^H-`?zz5=3>jcvRBYyvic}2vsz_-EjL9Sxq z^}x07en#?W@XwXMSac0w37a?uy4BZ0I}Nr0(XWL}1HF^}qaF%cK_$O$4OlluT@Bg{ z`b{-hO9x?!G+Y$GZ+G5CQJ;gbL98xL0LRf=v576fS3K*K`$G`RSl5~W)#0>z-vduX z;=#-y&^4{ARtTh)cHO)J#E+~iOl84+ByLj5T2S*jR(L9dy4d@nZwM%B(i0q2LEG$U ztn>#<$Iy=hMuOk@+*-ufAeZz1*!nK`+GlihZ3E}kJA0Ep0Y}EIt%(i)kG;DNv#Qz~ zIR4#ra=K?2y1To(Boq)qK|vanGAKb>L{dRS6h%TxML@3p{qIDNNkCcA*;vRQf-to3s=4Vehm@EjLHii7-<>3!u}@J-J==*k1011SYF=YU@R z-oW@W!%kQ5e4O-Z#!hgrNSK@!1nQE2Oj{k0oXYqAGBAF?w@uFg?@#HQU2gDbDcv&@ zK@IV1XI%uwZSOk$A{c#*MMhO%nQ_^u44DV6Er~x5UMa1#qZF9aBXR|rz_dF&HX!@J z_$m2}B!fIb?V{WP&&X7#!vks+{|?p=P{!EaH#Gp|g7viN9ezGAp2`WXfga+n?;8Z3 zyXh^R?Lb=_T;ARS%vHlb3Frn=G36;a-$NQ;009JmG*~Jn-2}OrI$HSvlmnK=rqiGX z_`PSz0qQJ&pS38c=loY%zXPqMe+%n8P^Z`?na_f3H|@~cgY>EVgIp7&;!=01$$x10 zFH6r$PSAD*cd;J;Q|{2a{s+NVHKV0#H|QbWUcLhVpTmbKM>&wUTGneWkXNchl>I+n zDJ*~8am8l<9+q4$&apo z4cg_PbN10-eKFU&VLu-l|HawE!~nS<2*1c;v88L9A>X8;SHm7MCC1?`JCtk9U`zW| zb$J><}c^tTRso` zX%Z~XoY`T2Px#izdi~$Z=c77YW9Rhi3t;P%)yL0i@KxHppd)4B^q#{Zx3hIn{qqKe zTed>+4izp(KlStdTrG6&tHci=BRzGntIcn(H|u%Z3hf~m5?#9EeQ(&WZ~sf{S+3Z2 zX&;I6LJ)*M>cua!(x^ZY3*jGiAtu(cOxyaOmYW~C+}qe4=&kz4j~_lPhbuS{1~-_I zyW>H4?t=l{8odUAfgzpzo`s-lVc+_%e6YS_X!=&?BmCyN-Hv`E>%9-po$Wuk;bq8m zq~NruUJsV1)J+(Yu^7T$j)@NngL2Q*3M@S9=lf|H(DjR&5m3EP^IE0;gbRB{JN=rm z4o+0wJNII?sU>q3C{QNi5~K&F%yS%vkUNn(0!$Bvzk4Uan?a?rH1} z_%rJD&x&ilSBgECqdK&Hx<}#KQP5d>vs|+S5Y#TLfnT;~>zzqDH23PCaL@An{MeiO zZ?@PTykFQ{?wMSZALRRf`F<-H+qsfs?+UdISyy9c+KN8VqVact`(usMM`;)r!Tzn;35Ry$?ZZ zm|Ea~%xJ1=0_G#VuRGH7oJiw001XQ8Xy_Ab2{a=FeOV(ll+f z5&_mPV$KFP1n-*}gI#BUmQqJP2J^be*8?|$e$JKP83#(BrL#5&Y-93ti?{;3tnb#l zK}P2iDJ`I|bJQQ&h+vzcQdb9W=zu~FVi{FQv%$tC)dsJdyo%fo?$ksg5mV}>sP^b_wI(&3*cRqmg&d~ zT4F%3bs%1(w?lV;cV^mqP7An4B)y%{8`R16QRdme0Ar)!1H+{s(q{oD zj6uu;XR%v)QYg4H6XvBY0%6~B->Q#(mPTK>06Ksn$D^!c1TXT;Lf4`Z;%)*9q3*(UHM>K$ZN0 zEkQqTXHqPUrK%v8P}1c{P`|eKvJ3;QS5R5|QP4IAy>A}|YIVPZmP(+wEbo}oKyGI~ zq%H?J)pSz54)S{QZfy`K!)%ky6+rU_w6=Z^rendg?2ABs!*8TzFUWhfuhp?2Rh6bn zgMPXFD}8MhOnbo6DtblGP|(jgFME4~zSC9CJA$kq0$hLSrsXIH>{hzTpZt7bL$@U+!{tKw_VV zTo_znHk4ZsIz953edCjt&%_pSuVl=H^N!=j-OJ$W?3BFG50=|o2I%HB(5QL4B}Lx` zYjJxn?Tu^~$nvIN@sxzUr#El8y#DvcL*J#hHJ1JU?WLEcfjs1g1Nv0GiV-W$3qcV6 zDCuriy2K`8Aw0?ax&?Hz9(wRP`MtZN)0#s1^pxt3&Ocv2uEUL|lYBpaKR7Ky^&j=< z$2A5U?RiLlDD{QJei=()>4cf(Hl@I41tvVdc0bJ9K4#~tD{y4{kMqwezrCD6?}R@e zP$t_K%rvFFH2r@c|MdDP`5Y2>qdj?iSF&B#IKFS!rPZ*z(mMTEX>kAOJncz^0%J?1 z#J&kJWAa$T9^nx?_U-}QE!m#Ic`f$braMa>vf^NxG7awDykSXx^vLE!pc)o+M3!;146h11>*!Kl7!ub{QZ4?=d+>MFf>4%X~m{qfrJ@X_uMGe4dO z`T_k1{j7-7Z*@Oz^Xzkne|q^rTsxI`antkd;ptUvO0=$=ZG+XJ(e_OsP18aYG4sOj zSB^c*20y=|^>e%hzLu_y-nwAi^tSN51^x{RD>=%7%Nf5pwGF7c{hZ|zSPn)O4mt;( zh*Zt736!zsC0c$kcIl##48RhE?8Kp1vxlzxH!17p88&btlNxdBtHg@D;oXDRFZRw}7r9P%yKFM#*` zw5d)TxQ8W;$mjt|Zrdj&JJ3YhMibC?`HJXefK*O$9o*j~gr;@^XUp3YlAWOLwokL{ z2HU0F7sB%a<)qnk18JyqP-+KqZd16L1H7-Mo^-qiipTncHUrdVevy`|pj@(EG))1@ zG5cQz#`hRa84bQ(8MR#TAcc~TK(OSB-WRk2q-|P=l6?l)dv}P-vC0lv12x!hwIv^D zZw3|eD-B9B>!+p(AeB(o%I*Jzc060!CY1oWhq7&oOW7Q;Aq5_6miJEINlnT_k)mTKJeLHf2)RfpOeqfQIVh6lc4Fb4y%eyfDl*27Mth6a@X^Y1K#Pd<>KmY zXSYF0zQnYLT{wSKj&r+D7*MX+rz0jv>S!j~}{JT{bLZ#6SD;Il&{$HsA)yEgiHWSePH-Bh& zYC2r1e`?9i77xmCX!y3^GYN3+?IZ78ee1#crMFer3;giM>-EoyYtorJyVJUZZ=?6T zJ^(V~(hE2ZxY_B_#`psd){9w}uV2`uKQZ6$u~zE;;H77-E1!1@IJ;%&u3F&v*qzV2 z^|!YJ(?v^v^;^h&yYRTEdOuICv;EpwKL7c4{=I@dw#L7|-Tcw>EPPFlUcuS=@%Q6@ zKc22v9KGFvaP#KH?ssd7&sjnAV}-xJhE*GVQU5p$AMsY6VWXkWn!5XI4;68F+;xrY zHsU2GJaxH+y;%oXRA^!AC5>Rk#pSENS_3qvCe1~he$~A=t^9mN@2`8qf~%kJUnoC_ zYma$7w{;sK!t|H_<#8)*{Hyn|moeU^uxdv*0-%}PH^f*^3ZxJvjgj)lOfq7a~a!?oW zuE?10N`uU;m%GGu0i%<*i>`v=XZytTBKUGR&bU7ZSMK3u+tbccFvz_U`iv;x%H zel;w9AaAwo*22Nr=yunnbKvs3o0@6^BgP%zs{lqFJ>2*fd}rLro^Qb`rGM$z4({4X zUDHQ`yLgf_y)3v}B~8lsG+SCcscw2Za2rW^(`$iebXuXzZQ$$ay5$J~ql#Wr{~S2U zYpi&1JeLx8CG{e>cE?>yi3O#Et*~hkC?DJEnlpfMbfX;qKj3PN_P*ka2k-l7O`O}o zJv3=Z#>=4OvduABLHbI5UD^$L8*h-lIotPzOUbn}+JM7*y>>!FP&!z?Gj#&{_mE>i)@W@vKOY$1V3cMIc=Kd* za8?Crv3b2(8q6aiM+9ZHnra9;69!BMf%s^I(c_~>dFVH=eeYU8LPned`3yvcVMjY z=GD6bpBm@**Y5>cQ=w;=a;wEa>l9eVJ`|M0_Lb)OAblgZm(+)z2T5`USSv=$!5;&u z7^Hwc$Z7Fh=eG}(-y=`9^ww5@77=vJ-r+y2I9o0(z4&1LKfJ$3&EKP~W*Fk5QV{-m z+~b{Py(Rxc2jVqpw{-aD@7t37LbVTmc{|@;Z>HY=$<3#HLaro*LB4!NhlM$z>5>jh zitmHq24M$m5-^{c?1kj$_%V*$@ZIRe#ZFk^#=MJlk{{toyviMQw7dZUOSu|Nz;w<0x;hp@&PL@4%n$OnN)I{PvmQUK zHO{-bAj~*C_}J2Yz%{9^l)XQ3Z>T|9EjO0RW&6ACxb90@4ci8P9ee079Jsor_UVfc zmfzZcb#_7uY_0fJ^ueZo>_aL0i{NPyuLtA<=~LNXs`cRg$s7AUw{SU3tu0`T1up2H%yNl zIPuGF56UzD^>IU1F96-94>n$foIMKqMMgoDqfI-MT=vK7_0Njy`HzoRxt<&5jURJt z)h+m7-y0W~r-G}nqpbTbm|I(4(Dp*9E>*uMF!ImKcSy;oF##_m{z%K#R?)GPYQX8yhWY@ubq7T)j=OYjjlCY zRIdxuVyBFoQ3JNz+5E#+8_lE$0abHV+f&8+Ag=f0Ec}<+ z^+D@E(?GDlRYcAa44Kl^xcFzmb|P1uuunkQV;P}M2k(~-x2Fi0E#W%@3WMBG?WC** z-zi5KPYp1y4O{8&1gW+)SXT1J;wwYbM5$T1TL6Di_O7PvI zAcZS$DqX<-UEW3!(?Qyi6%S&mOf1M1Ogq&X;Je}Y-n|K=LDCuN0H~Aw*IP@1(%;(0 zR2}Fhd8EDB^7*1M)QAD^yJ^LpyFl%0kG8A?b8=X{fXi9y^I^7j(?NYz83x7En`6M|lvl4(cZ5VUGygD?cyI1oKmorvf_wI~c{B}uSBouS(1|!K>&O`3d&e1bZXkd`8$JhRUI!cfy+70TuZ0}#x$cQ1+&F*uq{kY zfl@`Q4$Dyxv|v+yCFwz0vRCD8Ic7ds@Bi!P-hS@yR}ci@_wlUBNXjExZ)`bEsd9BYCQvWe`@n=`JVut(kO7|!lC9@Uj@r}>kyL#npb2u zo=5m}3NuB&9kdtv4SlCs+gG9FfoeGloCWJs_G{+456WXd5SU`C2Q@FZs8A*jde0g6 z<1=}|a>;ti6!XXH_4mg0-!Bpr8dG{-&XJIoJF!@17{ot!DY%`1f{lza9SG z^?z@>`J?Aq#JjQMLiR(4HwM&gH1WkBdtw3J&!~`@|4%t%%#6QUE+s%96~*;9Zk8 z&v6^{N{)Pnr%Ppx|6QCU3$@YU$HW@dJ%HYf8{L!-ryp7W0ocY0XGHH88 zHBer*_p=NGX|+6BIt2RDUc0{G!REdhnZsRoA;Wrc?wuKc${MbLeP!NK5u-sqZ2io1 z6bK*(0U%${Qk8t58}6OnFfc5>WqMW6$_8z>nT14O_+p8eGuXR>F0`~F(?}HCV zjN1T5;?3t0FMy->tsj$)g4|Qhp&SP5p_sUkPrz==GbADuEPbNS2Au#?;gAr&R-oMq ze%gNML2CLk^aKCaU@e{_XGnXnzmRWX_lpaiJPC1e*Sn>q{q@JI zQ>SIiLW`kb!ICAzXZ?Kn#~+U$zjNm=Z{Pg+7uJ@Dff!rvU4ePCeLjRcT6pWgkqf(4 zT<-{KL3NIN07{ms{7jBTVA^QOsbrgK)dS{0%StcS zDfz|EkDu&J_}H-;;s#uAp0*hxrpJ!)uMgj}{q*wD1n{(S*Yp+q>+|_*>TPUoE?ful zFDw39nCFkz+o6^}*1z^L9Bi?@-qo&9d`|^?^hK!lN#g=YpRa%S*saj?cd}i0FBzTSH{|E*UCMLnQbHrxH2kZ@r$_yH z?J^=FcAmdKG;P}PY%%c+MM3xzy^YyyGHKfAzs(etSyQa!Zf?^L+1EzxxjMHQiY4#0jYMeZyVFi$kYtZ**(A7E%i(?a65U=<)fZ`TqBI_gmxk zSNpx~=8v9d|9;%R`g|1rJBk(!o>93nL`7LA<;f%B@ptdX{$5%AS#d|a!{Hp_a>4D} zckd<-fUvNTnL*PkhfY%2NZa$;{ghhuA3Ecx9)cNqK83iHad>CM7|hJUKs! ziW2SaFH=!}#pnwuXZKIPP#DbJgM;nm!1&agLw^oZ&YXPe+P7et7tt@MGkA`r7SAXL z4*LyDVmO#i1zol`1>X;@6`n+JZoAzuWd`L2YtGCqV9aK8@-*Qi8~F_XJ2^Z_GK_``F83TzmI_|pXfk_Gd$$vR$IqV+OQDDFPnVfw-!!E`sBNuo^rG+|Yf%cjI z2&v2DoG2ZVW14Kjy4|!rHkzcQ`XQDB(7QK}2_GIgM`8>BIsNf`%{r0kG?d`Mht>ptTL z&_{ZP`@EpP?8)@L4>Z<)GNypB$tYs<1*3}ROWy!+?Md96HXVFZGj_Xjf@N*w_d&fu znP?ww4ge~#fPTP1dhs#%W@Hp}b^+gajxRhVz_`ICBijv%hT^m68aRKrwIRt10q!z2 z^G=6==gKz9^U2@+WZu~ztAANz^S>`q#V4I02)|}mniI;s3vF`%f&;>%Z5<(w8#U7k{`L9%HGZLir%efeT|KC4 zwD06501tb@BxrU`SKbC@0|Q}a=(+|MMu7kMpr0%aA*4V=DOHqhyvJbRn8*o!Iv^1soz2(LK z`gr`&a@==+Ak}i$*4OmSBdxdp{|8?G!~sX!4Zk6lby?}X{oeRr#N+qwNA}3?OS2&8 zL@0hddVKz9z8})=Zv36j+uz%6{^)u3(Ble%APBVaRtRPT(v?|lFRll!-P zZ9cp)aU)t8qCVQdxbR=LKdWEtU%b>Cx!!`5$o)+&3v!{|&IHp7 zLBnnJL0=pi>pvIV?UFjDR|9Q*-~@YlFb@rN`1^sYdAu_<7?hT_SaSy0Pvtuk837q- z7r(!A8l3I#klYMdL0>w6Wl_wA;H4ntP^QZ3L4DC?F>eF+f#ms_r@(i_<@Nf5_KANN zOAnC8TVu5dkS@v{<*$Jp)Tbiz!Kc|!d32mK1wkf7}1KDNnX()XJc}5?sQ57A!NP#s(DtDO3rQ z_JX|M^uD$jlmXT_(`Asinl35bfv==y(#!u6(FU}8NSxz=cA(sUJ=d_&3-no@o8DJI zk9H37lm*`_&c^PuU?KA3pbXHLJ4brjf@9LHdP$+6RI~X^uY=`<@Z5pl0iVfpr0$?B zuphJ>1U`|@$|2w@;^^yY4aR9>mhlZpKE)|}z;Z6?a*zr1+U}a(Ct!40OtCHq{{Zf8 z&Qxz6$efwp!hQSa>90we8lNZg4OnYmbpCifIM@GJ*SlfhY@6A~lLDT-?!~_J2dTr6 zDS0-*>Eeew-3f;8^DXIgVk~4dNxSJj^Yi`t`udLPy|H~u^x_uoI!b$`OGp%1omtjPGrJy4uSf*|~MS$B@-zSn=h+%qekUrR(r5TIOG z?S_R*!q6IXSG-sldMKmYwD!P@(QlS$wi{HZmZ_8!aT5L-d0#BLF}gaGn_hc!;dKvQ z3xXgBf*|}+QEF7>$bK*_afbUn6NGFGjSB4xZ*&}7euzIDSh)Yn!R8q@X(48Xl| z9lm|I#}U(ma^H``{kYtZlkoQ@pRm4b8Vk0a1&LY!j$JqQCawWn$y`Zc=^%e?dP`jj zzQWEv-XdU%32E;C9w-N`ai%!%mdse`91ZFS|2vl4py@%0wyxl-=`g#)Kt8P%S6>3V zk?-xu(V(95TW&c5&bD`ECD#LI^37IBFM@H}XVy!DQA4k2>;mJZ zcs8eBaOMV2p44iNTi_m@G&B88kP52)%1uzK1yr-11O1xoCvRTRxBEKl^B(NhcQZ56 zRTk1`UwHlYB{1IceB=uP>yli{!&K0Q1;69x|6qM7r}B;bKJbi`DJ=wV=Zqq*55ZJ9 zxVh~Zs2|(cnYM#eOWi0p%a;HBMq4I<*GNxvCWCe^pt!9yXzv8u{$*hNdpSjJ3)+r= z=dCqCe#`u{x*H6$FHH9b=a$=(lSYEOfAX1(i(tEwGc{}nnA?S%@?Q(;2>WYhCzy(d zEb$)!+JK-y+h-tmH5XKO0CS`dCH)W7Lk~%34#=fV9o6EXCIuX_UIJ6=&{Y4=L5;B0 zHwgvVK1*p;2mhCX zU$kUl!`wsp7{Ri#}aa#9+d9S6bItMcQX7qC(hICU(HP?~9Pd@+K zlsQT{`6IBDv3az8;P^D7pSw5cbM>A^gt_B?DF=bOLLAmv|GRjIDb@wyuQFQ7 znmXw1|9MWb2FPE(cNgH8ky%r#TYq)(B{9YNeoRPcwe`jA3*qghBV)&OhL|^Vm~s?>X{Bb=c&{Nu6^VL2a@y}dj+kQoe=9xj@t3Z> z4boga!SI55+pmJ97bI^!?7FlY%uj`14=})2)Y-(d4l;9J3yv=UN;~sDwKvEn(+u@A zxRT-yr(OnMF6TDSNsxoh<+Mc5YkRhOM?;4DLf<At;Q&}fH6wfjV)jp`dfM}Fh24|>LWq#=4tG`270nP(3cy0 zt23iq-GP2m1L-uV2mPj64uR2BZ(t;V(ZZAFJqyM`4ssAk;|w>!8SD)H&}Vt^^#RY^5%@wmGqz1^JTTW^9A8hY?FSD6zGWaHD zo^(wG<0c0<1Gabbe-Rl8)=BvmL`MGI&)S_0?@h73>ED+;;!W_PLnjErL%#4;avt@{ z;5vD0Pbv{T`uRlEkst_yAP9o+_wUCF zg79a!KOj7#{QK`rb;GccLlWNF0V8{ju#FxBeLD8@^vVJ4o_e-Io4KHFQ4gq#9-s1g z>v}7ApNHjNEc<$8511AHe!CC%LD3b())dJHV{eVGKf1yn9~c+@9_$SZ_l=2$JNT6Hn{87t-7l1WibnD=&kom^7n0N&=Rn4QA!BjEq zs(&YN=DyQ2#Ra;})7vY9JVxuPb_aXG{4*oxK&IzPhr2d#O^M%@x(W2Po)X>$U^|t& zb3}8{r?7ybAoY}rOLIXj@2^^KfzinKqrM%y*V9)yRsu%GW$7D`E!uf?9>`y7tCjg6 zw1~ob$!8+qX+@(YEk5UH6W0ZDs3V0$@$gB?9<-iwg?LqD1cgXV7Lk7mrFp6)% zyCAKZ<7trlSvQ&vf;J`imVIQl{maVJ@-onZ19I3-fHZw|SkwRa_ShI8-Q7s1bcZO7 zbhn7KAl+;rAreD61eEUX8mM$hH%NE)h-dyj-{-pa=l*)VPu}-A_em^>PA-JF7```| zF>*a>t`}<-O%0Cfqgn+YzJVK>Xle4^(o`=3Y_b9?#mHPCH6BlhIs;#reUdSjn$%1G znBqDgkIIztsJy334F6QO<(9yNMBo@+!D70k&pZ~a zfu;QQe0gWE@6q;oP{(mfZb`GTPm(0w0EHg+UQ*xas|OkOyk_Iq`K&i{Do!ItRMZ+4 zDTD)%W;wg`rx{Bzf@W4biKf=)Ny>B}2glC8$t9Mz9A_$HgeQW^w`{d2N-l#s%)wQ< zI8QI=De}no_$Z?VmI;>i=(8uxH4CAEKFd1sypU#a--p7m8gnvH-5t>&O?Yw%Lm-k*=_Ni}bG6DO{@X}_J->w4ww!sZ4aFbzY zgx=$sZGsf=^0ZAY#3tXoJnMlvLHCPPdk9@)JDY{8vETi&pd}Qc$P+Kga{_hhfmh-> z(DtyNg!zIrms#?WUSq^V2G;P2ibsxzu03*>lWKC_rZgO&Hb)lMP^KRGpi>MHp5XLp znm(hyZrBn*E*2b;Ti_IE;Ofe6fv2G^7jz`@bq&VvDFL6d)6InaWEOs98%_*yZBq8pOLy- z=2Fto#Paot5TNC=9Y%KoTPG)Hxt9>9P!f_8`p*nejkNl?@L%?fO$zx|NTWM~lb#>c*>VT+rnH&U{e{>p(63Hu=WktaGrjdeQJhS*SpQ z29<4gsCRIxpo0o_X5iJ(a&7&09pmlFtYC0U@1^O9+l#S~IuW5$col0hDPi4+Ij&l{ zTQv4AC>Gpk{-y-uitfCPsG92Q0`0@28*rzXHsP4j4hs}qC4?e1-9DBJwRPO1?^PSi zgqk_czylc1^ewSmOB$K9g#peNh7T-CK_z-wIMcLZE}xX?~s(wqWE#bIfBIbVwk zjB&f-HbGEPP$i@`wfkCDI=A3dm^}BwxJQW(yS;ds@*?c0&76~b>GFdD`zk<_!vr!e zI>hFU8Y zpe-AiBPq4}%+0wzgCA4STUOqL@+5!J=ni*rhG=-#zpM7Sjoo|BD(;af=`zZq^MfU> zby)#)KW7TroHld6L<)uiN`V;qvACvdpY<-HbpBB-tN^mirN!c$%Il7ubEx2q?rSsf zIHhNDW?;^P^Re_2;DPnJc6pyZMZZ%Ike}=sX-{g`3~Fy6hmkoXP2xee^Du;;BJL9l zSvTcwe}&J1w25vD7QrV8%Qp10Ej@lOv!O%2GZ`!wsQc-uj0~qak~SHR>7YH7r|U90 zsN|dSTzvow$_(wwxgjy@t{730mq&JsC!Wo7CuW$R8Q%^WHv39Kqs3LEb!=7{=b*v`q|9C4(;^AzI>@DPA06?z}AKhnhH%&-R37bV445d<(DojOI;k>C*l=z`!@ELzXP(SRQkB z^XbVuk^~TqPA`-56yV9!Soz%?W09_hwLl+bJ8yxGH3g-Ie3Tq5)n6eMR2kxFs&)&j zgvQyZ?}2XqqSN29TlCR|+((TY2*W4%Y>c;@G{4-5)TGP$|9qnQf;@VdEMZIzASnVY zzLnU9H;TIFA+VK3*;qGh;h7?CFD+l>00UuiJf)`bHu8I;;RwH#~0vCD_wadhe0jY>PzvV;@&i@VmgK{A-WiX`K zsl8wvHv01@YJa8uLUQr;V*YfpQ8xRa)p3aABViWy*}h8CB#5zfi;IAdwB*R;-eH*> zCZp4!itM)Jd_vkg&X)^ibM;E~%Z+QE9~MP^{;zKRSGw{cE$^;Gj(}~VLha?QhE;*1 zLNneKJOl;HjkgSsYS5{u-JVU-yJ@}tYV-T|R;^aFlrEVOnTQokn6VOfUQ6fGO^<_h zeKmNTar{GGV^~Z9&Y;L9lvjgiq`8i3(Kx_=;V>nqz}Vd}25szVA9>=n8At0p_k&_A z$>nw2WwnQH3F_m=d`7oU)Z)SGK=%&b8}$(x&JuxHsx1WDI2y$x8RwEOEmwwhTUSgM zKWmBo7!LAGYZfa5b-?RyOOm~^b2h&62(%%IZr&Tkf0uMG$4j7#@!5%D^b`YJyL?47 z4-aMj@JfsCr!W4jmbnQqy`XEdP?}H3!a`n*soi>yn#vPCedYod(ikqMWz1i7?D!4cZ%#xI=R{H`hK4jl)~n|V?5k7VdJ(ZApBVNQRorh1u; zCM3A(@5sDrEm(urlY0ufmAZ5nlToVOH1uWlQ7F_M4#1lP1gBiGjxC=&O}|CEZDKb6 z0-(i?4(_!&arQ2A=akhb#?|RMGnVn`8^n6T`ODnGR%4jsNw%heyiO%p*w4s%n93nMNX`B7% z8LN)-iH7XTuokSnHznmSlj`+k%I|rIMUqkt%P8%%jE#l?#c7keCAXGOJ>EUnJ#nke zhhhiutDg`zV+pO6aW$>W%OzDdYsp~?;qo=i_Kt4uZ2UQPi0;PkPcZ0D^EnhDmMJd7 z2YNu`Yv*Uz>;y;<*Vgh2&#Y7n_#J5{Az>%_LDnMEPhrTm(_TiJ48&h`?e_&8{31{@ zj~j)H_q8IJ6_R@w!~EG>PMW7O@rSI~2PsQq5b_9o_W2WmDQFF-tq0?}L-pnFqfxgg zCfxZlr-D^=H5@2$PihB#$#j1WaX%h@_IVVFK-{R?8+09#1+yLE7kNF2loQFRW9>Mc zgPdQ&)ko@>qv=a=V(tZrqcsCWc9Xph`S=Tz8qRAn>31`kd z*Ezu(%ln?tHbaE=1Af`Jx6@%xDCZp~eUK?Qfz9^-NfY=FI~60?L>*qzzN~XY6{TcW zVtX1Lq=B37mp_d6S542;0(eh@t$vomt+t?~bm32u z7zd=r-7a{dqbmoQ9>n~Ks#{H7mL2#FqLwUnA|M+Ram|EHY{c!+#^){6dUWI#1IvO8 z>JN6n>=(Ef2UKAKTx7fU_xciXbiMGnPeyKnob4}>N8iscGw|klpRZFICH{Y<^k2Hl_V`fm z)wLe?VuMln+4#5nZ(X4Cubh9}eLlIbL{uYT|#sr_WkgJ2RKI{UeXWx~Nzy>y>NZ1ju08pi(V++H zE3jh7AY}@xjbPItGxNJlU}y8Kb?)@bn)98@LH1S0RqC~B$oRPMKpT+DAa?}S6S)07 zBmXN0L|kr!(-~xwT3@p3&$mx70i+B?#N0&fy368uUZ0{t?!%jF8-UT3Z3%&ja#Cy8 zQ1D5sZk5T9;K{Vc=y4JFh;FU7;1D>vGii=Qkh!xPW?zs~XPe1bZ2+1}{7l5KwDp8u z`J(MGr!kXY$5l4CP?EEZPIM4j-oIPqKZDfi3ik)m8}P@CXQA&m&(cC4;KVdFqsKz9 zt&;BpMgY>U&X%U>&zIu$LjwK?Y8~R-$zG!`Kdy2fk?)w}8ta}=BQkO-YJ(l6QO*<# z;vJ*_KLAa`;;|4H#ezM%GuMMif9XwnLyM6G z`T$D9aclHXK^TIoZ0C;=IIdY9rX@g9H+*fdn!^tL6Pr@2+d)F937je|XsH>k?7`X? z%>wvdX+_01xP!{~tJv#99OHbrQNt*N)@KgyP9sZvIc&8ph*MG-jpzlcOwV>{>RYq_ z;Dj3@p|aGZhL((^jv!`IU!({Ea$$_m9|B>q33tI73E-{R^c;TBKom}5|&-F)+0 zib#u`t+>eUNq%G;lBcYUCmXcg{H>Yfpo>zXU+RUF7ir^a{&1$@>M+pnSLh9&le9g~ zqyiYInu&STGrcRYQA(RraCNd+z?+MbtLSb%7Gy^cDR|D>_jO^OCdsHipG; z9bBM8P}F1e3&~s-alFHU-H~y4i3#I=?LEQ~ePS$%ddS&iyphqlX-#BHA@*6zs?WIx&6;M$xfZnoiJ1<_YfOACr1x_=BLAVh*xYg3L3}| zUeYDmAU~86LI}kzlx<@mPpOADZA0RJ@J@FE1_%F4gigqgCj#bhwcDsU;(*+YRo@S( zm1O#6HbPd@@x2*-<3$s%C_G3Sx@?NlXyeyO%jq2Y$_33FXctQ5NuyIb$2hJ|e&}Vj z0A(KyU+6&q&D|pbESen+w1P>UjYn-r+NHpz#ES~3ni9-oVcSRU+cDxS@Q%2(ZRg0W zO1WA8lP7w==KKI&vjgjK?ZuQi&@*v)=#0^j5!Q#JQlF#RP_+2v!ZMQi=DW z(c2m<=A(E)*&CxoLb&&Jz_1@?u!ik};tYQ{DT#!T;4|zvtu8xk1(Ow_2Y1(RP5z7| zNR?G{8yE0d;TGl|4j(uy@IZvT?N5Ipbtt@+r@Ygeg-yu$*_ID1uJkY|*OKc*eX2L) z#d4pO8rrVjp-Dc9T_dm>C0)w3s(hwhfc_c8FOg~AXl@fQ3zX#yAD3BhEP6G4d7-iM zt$a0TgOxDv^O?z4(8glA9yeYdB@uS;Nv3i-4%;bF;&l??8^XysN2C8!K883n6e@62_*feA7zU4hu&;C@{Ja`PiCM^edTL&A-~ zwVlYDn;9cobcOH%-*Er;GNu&z{Dj=FQ>)wuE4Si)nMxV!uiiPwz;_KzL9BmSiq>Cd zUDwLy6t5QVc*?@s)b7?^IKUJ)dGo=Rh4_CUwJn|r#RBgF{E7`03oRwm5n(6R^b+;E z1Krs!710fWw#o3>ea@mv;H|2vMfuRWyZGZc@P#Z7Mz_q!-S2&g4_@*<3n$W6(na^y zn{=F!RsX0}WsV)A-Gc;1(%_i;XO9M|?VyUdcE7E6N@?50v@*0eW1<5KjM5<|qV8ac zgvBn)yOB1vRupZ$;=<=y?awjdgghbl#k+*WYM5LmUcgC*^MkxuGKRgv$Yrg@9E*yM zZ|K2jKO3Wq5O+>&ufG2Q|DT#f2)!gho%{Fj6&Ajvs2cRwq1H3yB1B}+q@8?*^)9O9 z)3RDVQPLF^T#-kM%KJUT?ge*sc-OV zk^VMgtGIgA=w_yaoYOle z5p58zrc;xp5Tf;Xt|0F8tAr!fvD>i=F1xG`TIKF^ClWWZQ+f-<8(Kx=-#_g$5e`5v z-8#nl*e&O1sc6nnho4)`rPa#q0AXb+;ca;QQCa@GSAr)J8QylU3}F0qp9c!@c4uNQ zEA5vsv_DLs9?Gwh-h;B_c#w_ihO3jenkO-O%2t$z9?>Z>Vcp~+<3q*asNdzAOPY+}>`Om%tQHJbIsG0#?&QF8x$v(Y zgs5BI!39?)Wq)b>OPg7dkC6D_pm%0UZ{}nW-Q*`*DjHq|Q1#KN=VdH6E$>O)z)*x= z_sK=2>;i2~-zYfouQ0BqfHOI6Md;}t`wZE0FlkK|@LB%1MZ!7?6c?lLP-+{Fl1$KX zLtW^I`08%zG`t#Z<1{^coFb7k`F#NgoNL#Ju&e2NsR9b$@^oO2l4YXw_!zX@=^+I; z8Lyy2mghX~r3`Ffx86LYFYH&%d(P)IjwNx^;X3J72@L#Jn&~af#y-6?Z^YM0I{!-s ze?$XXnUhEi=d9|r}a=)2}ccMw+Fdc2H+`)2&wlc9Qsh)|Driz zDcttR0`v|q1b!_EwTRKiai5dpB)G_bcoy_4WRAcZuL3n(g<@LX?~nW4p~9+x@UaF# zcBqFTXMS6%b7)k6iVN&RYB(n;VA(%&koTXo!2eMoGSGYleZL|P{g$)G3Bm2pv-huK z|Cp4Z1QoT1Iz+l65o>s{E%s?N2%&g305`$sGx>r|5C z@a^NO2;w#~uWSs$ioVREUQS!xJ&wH&57nA#T=Ahw zte&nfP`6aUyZsY{wgEJjxunpnFfMWY13y3Gv&V#Ystw`NT;hhKWO46ie6x`qwxO~- z!YhdDWaP;%WU|M@uy+pC?JPwAPX@^>ePR@;!Mf#JR$h1-0i$7iU zP-|jKl6Pv&KFmj+Ctv#9Mb@0@-#km>WUT-EV;;2~~!l<2dNle`;?%LQ?vQI*^ZTPgqhm(p5f|6#XJMq?9l z>!8lC7S(;dvP+@NRM#I9#hX>%yU;Ep)GcQ{e)cilA-gJMgn152D4$W2fs{L0`KzJy zN%|fByy2iXctN6-0Pl2+|19PO@pha?jA-R3`bPOm8C1dN)*{bQZ_NA*ET0K4>OvMY2!!tzR+kbQ}BZK}O zSJejWfN7YQNKZ`FkCI=IN+u`v4Z+ci{e#_7qK#Htvy2v}$VF9Nv&$4A4u`oNyG+>+ z)Sn)5UZY;sheM5jyTokTYi3o`2;!i}_W#Udo&(0qQBzoAUtXQ&hZO2MgjF$4NL9x{jT@gMLtBe{MW1a2>xzF z&MMzAw0(RWW(J9LiHru9S1e;VfNf%Go~5fh1+|^NvSgE|;P~I?6%{auqomU%XsgV> z7$^geNHq?yq;-w+zxQwzI?>>y{$`Dzha=&6=7+^65@|=rN0{i_kh}FLOSLsdnfLGl zD-o~j^8WB-L)>!Xa^y@3G%zu}u*D-QwM%6VzHNAR$=3I`;H`rw`2rVby$@fqzr;-z z&^o#@T#GQ8EeGNdv(7|{GMxW)90Sw&R+!#fu z{N&V?D=I3{==Lp(nj<*VH?zeIg>~FAiquxs#IWR%CyM&lNXsXzpG@PQ{ zKF3gEM`&e{(udheB_fXmw~dx;p5bxg`K7v-6w-*kS=>p#!8N?@Il?Ok>*n*hNDBYH z{T-B#9I??!u0acMzVyCqjdOXCNODHyCOVdMC3#6DTqbAj2R)XMGOxTRPkBmjldS*| z7mE5miEXBEXk`-8S0=6`{tWWJjd54tug+1Vyu7T|2)d=}m}9eOuc*o$8<|{s!#Bg1 zBvKhM@4s(aG6xDM_VW#VwAOv(G1_izJ6pbeGU1hyJi2^R9F!Tk=@Ui2EA270?zJ|2 zq`OtQ!`Rqz`(Pi()_Kwye|hlH-g3IMsoQ9m%;fdE|DZyo1S43*o9!cRESLEBT_MZK z>k(eB;5mSbq@J7WNAwLwvTFI)ng9O#SJ~1+>%`cD`1$clCn5x|t(yH$T9BnN)meTZ z`;FE~?@la?Hqh$nu(FF$u_a_Llz%39(3GHPSV9dr*g^eG{v~|rc~H>2=}EMD=^+O2 z(gfjqjt1M~4BQBTe2daW?vrY3q2Gm*dG$~^RJxp= zQWzKC;_=b>%)))B5lyH)v@6?prYA8&PTROB&DHkN^AtnwNV&Q>bZweVGcoMQzLt(k zYyq));~bWo3@#a9anDCe+4n1sI_nytHRD$mw_;TzpR*`TDeK5WOy+fDqjb*t=Fod) z$h$m%+eI1Mgq#p*BX*ure9kMpw#r(toL=RL*S z&mZU>*+3iF)zS5?&0ACTjawnxsg*TZ$EV!EviANEA+lFyQO_^5&>}5tiECHi{v6D3 zQGgz&YR!{BktaWeo4*RC3n@UmI#~b`WPduu4*FzqM4l4u;r26(DaBw#36%-WhkDS0 zuk@awF~lF;DWYnPVF85xbav+^(44l2?bo(s?MQ~cm>PJv_^scye)&|7ELJ9H^LjX8 zEiVAyVffVkMjFHGlLrph-TuPv=+5X4*{WOY`QI78(7J`@fkVIhf#V0!(t845_stm7 zXV4TN@2!a6r^Aukkbp0LwZGRoRt7zDkQ(K1sqL+L1lhFKkuc(J-Z3I$RvXD!f6&!T zqq~6@-e~~p%sMY^i#d9p%VKF;oYezjgu`+zX@2i6(hZbrnq_R-rW24-3E;Ot_HvQ&xWhL zUYc`qhe9N5K5Vh$vBzA$krcr{5E#Z05=gbc4Q9^~CojJU^5a!naqpNow70C^h{St_ zrdLl_>w$KB5N$oeY5DkqU&>X~XJTB&jTGp8^h2pRW^6PE0yaZx~D68jiXST~RJNq*cultg8&3Y<`B zttFUR!NzZ7d%~XOIIFep29T z$-Vsr;JL#@hb4QDJ+%aq;5O$A2YfvaH~-nsy^Z%X&PJPYjBabWk}C6JQ~AxM(5yZH zBfU!OEbW!(^izAkV^f}cge$^NiI~f@hFdSVp4Q4X zdYUqJy%ppPL-8Lc0JnibKM(Zoh58!hzNjJyyRUrLcbH?83sCxAg%*D%W^KPEpAO`4 zaefdq-SuPT?*au`+@7LJV{X5Q-&fI3ZGUKC?EwZ zQ6{gQCx*kiYME|}qlH&*McZNL%rNcE!82~ns0@LrMp-}kZ@CwsbY(IM@Wyuvk|+6a z4gT~^cu+1KT}bul!r(RY4Q6w?fMW6RsU558bEx#Q}<-RYAayw8e4#um`c=8 za-~QNt^)8~xJBtV7}>@q3a{aQ$V%D|Rrec|%5$mgFIQJ4$}UB`gQxhKxzXI$a;`za zOj63AiS?l6hjkgB9~*YJ^F&&)j%Ndk78axV+vBWIVU7LoE!hwvk4+cO&iD^`RKI1F zb-yNl)Oh)z%9+2S|KxnBoO7gG$f*J69ocj|6WVQ@dLm__NxZzsKPJ+AOj3AZeiySH zY%;c`T==}RxAT5LG?%B;@cVfvQy_Nn2H>=E;K)}YGvJd#wdcj*7x^ooG!fBjg5Sc+JRy%4te z3m>n}J^Vv59YzP{GQ)bts|=}C$0NrhqVwy2w$n`{$4PaM6xT4U+~g%c6@JIlX)Gv> z%f@_|AP!~rp3KQHwhU&mNQ(Tu+VoLEaa{VutIwxj54COAL`-hJq>VuX z?eJwEO6T*cKw1grIvDnI%|=%%`Hjsqda3b999=9=QeQ^4m~;1d{LhDh=++ zcZj>Z*ZA1=_-GMi(WC|9Zwf$s(9v}NB=HajLE41H)(JYtM{Oj``SA=_wV+sX_Wpb@ zzMbyrOmdB8WBef^mFmZ{|8q6^`)?1yorj%yi1vs&>P7Ciy#IONOy7gr(dTC1_bIIM z3ZYTRpRi>fsZ*(wz@3Ck`LV1kNde;v>wnEMvZmQ3^=GZ)zwzi*7)6eQb<)PKg7!~l zF~3`c*m(SkfF8=Va0l`RV(zY(e?9g91=cyn7t*AHx@B-gL4N+;?8nq61@GT@F^3>` zV=k_8-roh<6r($E%W%n{e7oEWN53ePvR}5gq&^S&E&_N7v7h5FsKvljim6i>2!aXJ zz1`zlWwK|-Ix0^Hd~hAmA>QbAu*nx3(0XfE~B6 z#uMn&>|T!D3Yw3FdmmOp#b<-FY-lx~1#?KvwrkDS&WND%%U3Q%Ti_P-&HjSnb_m||w>G*zNkic9m3dta}oyv)OhjKClH zM1L#@Q2$oyy|aoB3)_-cbnLYrHU4Y3RsD--wA=oC_pPnT%Xk>06xI~_-Ne(Wf8wmg zN*oEYUi{Si`9ZeJ;77Xbe3o@#1rN~nFvI@yIs7armW08aPa}Gk@C2wNKUqHTwy*7D zeeaTM-$7uxNjFc`n`m16dy3Q#sLOSgXco^3M|E%lT$+uW5g_m-sx={b;l}GSbP9oa z`&D~)0%;$+WZ@N6R&CUs>q@Nr^l|E986^gl0#p3N=c?utDfFFoS%uBeyK(dY?2c00 zej2n9`$pC0y&-<^BbO|!D=J^(SK&}$>8g}r+^8u(c?rCypyz>|{$<`6?7~^FTgqM>!O|Syf(LveAOSFj2PX) z%2B{U0+r3F2K{TgB%ilG-Jx{Bg>+K#(zb=Y~7aXfmQ zcnQ^hdW}ExY%g+9jYm@@LP<$9r{4pe%`_eTL9iL)D2GA#$(0{@qhprT-RR3Mxeg4^GS%bVTd#_rI?IsvW6# zzme3GNiicT@>P((ZWn{nuJ)r138PV34GWh4pDtIX_zUzJ-#f3~EcCi(n_^J0RLE2{ zQ?$7jO(bulL4w7eD1#>c*Md}gDJ*jYVxfeoP7N#pwFTmmJMRgbdX~&!#(<&U-K^+N z;7iA+xnvVd-eiCG8|g7j1iBHZD{Xz`*E`T6h?j1#2@^?fzJ9n4HJ+`kWTxEEW68MV zS^X|fAzKhX6Nzq}a~wqgXw$H=COr9kUrdZTPBz}?hJO-8{EGtqECAH@HgOY&kHatg zP8!`Ck$1Pt32$XA&pfvuJ3}R_E`;Q3=pN@7$eK9@FQ17Op#({BC;KB7(%TmU;0h-j zY77GSneyvYg~s$bM!sSzMw-iKye`E&=@BQ0NR+du7we=;V!$5mT#1O_z`$GgU14r0 zMcs4VUm7?8iBpa-m6(hS(?DNLTt`l zg=!X1LA4`ZBS=kV(ivd1I?)Vnwa)w#*fJ%xU_jA^!;dSm0$v#yIyR$`tvrgoL<`mD z+95&ITJnDNc7H{sXN+RBkR&u4K<80^? z+?QnSoDO{3#{3eYEQtO${(f!I=`G}5WP?*VC7G4t{q* zv&A{KUzZeJZ#$;Sr8^i&`*|CCM@0M`F!Xr&vz6s0D+%}Ioo=@3A+W*tjAfTe^Sv_N zDMOKZ>U%EG2hB*wA1$9c(O1O7l}AG~gu>6cMAIpcl^<8rojkOr0d(4to}C@pzqI`P z9XUnoM27b{uT}%|@hYqEoJ=-mV*ad>J?HGf!^dZwnNfTx8_BO<3I1D}#N$WJMypVT zHpw}TX^pAGR6Jm$rJ>PMW*e7r^pNLrGf;LhEXm0KpF;;Dg|dA=d|2@0yG$rKBI@%9pFV49+1xJsF|tk&%J=@m_l$O)ATQ&_P+lJQ9|)q8WF@+z159(K`bZ`m(~qa> zI&mQMi_hVpRGvk|JZG4P#@o zK&_sb4ROeuaY(-{5Ll9hy4v$I%7*Kr;xeu4^mok0uT6RcqTR*_U9MbZs;?76l=b-Q zbktt3lTiTTG6Akw!{H4!PaLS(#ydEsq^qImQ+C;VS~Kq9<@Q15ROP#R8mZF({+RvM z*E0R8uouf_29>xXTlpPpm+%-DogV)bxwX3?$BqG*$JC#pMuGsBeT{SWKC^sL@z6|- z5e<7-gzEc#00J%H_1&|{(;60M#HG`7q;}YE!)FyUQZ3b_7pgtx)+jfocTSfiMb~RC zoXJT0A^#4?%Vef0Qe-GwlO%b}p%eWAYPqfcKeMn?4()RkVKB7*e4F!LN zMgA%X?-?`y^o}0fb8=?9>DifvHm(=+tHHb~U*AkV;kct^R+)2RdT4%Hi82-)^_$}# zohqz3luNWWkozC;9dggY5056b9&_P|7Bzg}?jKX09k20Ql;-kFVLvsocBkmQt#p`< zU8}NFjwlP|_L<9tSpc=4ndNjFPqt4E-&IXgEDmFfP_2??84cLFuK)1P{n0&a+g?M6 z_SqotxNnf5zSSjJkc`7_n!?ze4YQnXSC)grhQ9opOG}%`0>T>whbMT$h;109y`Wm3 z1LyATmMh>w*Nq5pz!J8O|J9_pxG_AkWx%MN`d3puX-5EzDSLj*fav8kC&{Jmn7*n+ zOrtF2OQ{!-#}&qh-x}q^_0pU|NOnD#lHo0qSNqq&U2Cotew}`s4Zpb}3Xe6#0kz^p zl8@75S=gTT6NW@i?3IbF8P{* ztt^Tv*waiBFKTX_aA^V_Mg4Lp6 zME*vSRT}Y1EyXaLd`z{fsd$Mf$My4`7lGlvRBh zRu#7&PMi&z=5^P==0*KAp5TYqAaZs4^5SSUi|AT;kJ2^IB7jF0xad_*GH2QTq0lp7q1uy*J6L67FJ+9W?Dj!EVnv5Cd{sxkM$b93wTxS_Ch(_Ie6US2H zMgRC8C-fuix`1D54XC-q;2nT{hUx94yKDZ&um&LzCGwGbbLs9;NWh=|8j&XTjf(!d zRSr})j4wX<82=S>p1zrGAKEWSq|<^WqXVTdMb}VW!FP}gVLn&2pl$MD{21krZCd$N zyT^oJjfsB%p(gvG?Gw~Kd|XW&_j~AJYofvT+1tK2wV(7CRS$~H)WBtdI2cyQ3iur0 zOLK$AZvi6tls%N8+XsBHT@$-rd5wwIqKrkI!%FK z#Fz4=<|!{%hbqVNNz}Hq-GyNshWk8m#p{<9Zu1{4sQL`v?C(22tA1T4AdORpIsQ7A zPARl#Q0$N7r{HrM+_F_O1#5e@=N8xXuukZg-X*=HZ?63-?%L&`nQeas_Hwx=XoVzb z?_Gx#<)mWNRh$!Lm;M7c+aqE8>g!hkXZi+%ZL}8usTbThyGMIb+!Sh+sMB#^1WBrrtLzxBM z+wAH88VXijn~`dKCuMLXkT^S7abL49)r;ltd)g}OFNSoYcy)2Wl7(r&x92h)%_5L? z^h7NOB7q#sUTnSzawSE+lHKx)0yx4i^3=0oz5RJGLlO$RZv2Kf7<^tskhY@lD0}w+DJE zU;+0haVeOPn%{mS9{3leSymm@u$=|7H79;zae=l~fqgh-#+NFmR^{c^&Yxi(P4@$t zdQ>$ZMnAVT?r?vF^OhN&&L&$jgs-l|4+q~9?=PR-fy-R&4Svkj67d|U-@pX`6cWL>Bf6yUeb%7TTH)y*emr<^H4>7;rU>!Ui%oA>J z}K0nUN5O$2$%U}SxNwRbIpg9ve<996D- z`O^NFM~tGv>x5LR1Iqbrku_1y^=*}7%RCzHipPiXhLHMF!D&AF%2B~TRCr+vslax2 zm}9;EawDZ$bG~~|23!nNGT@B-@8^5Ok8uS*Fk#BwPz}}yt!KFoRrjs<5sd@O;_V+Q zdKa)~SY6Ube0^oH0*g@0Y!j;12>B6(J%$yl$f$9~0Dt&ZY~6B`ugmsp6hl1M+nNp5 z*N?ZijZ(~kF3%x2BrBZP%h3@)qSV8OYg^D|$PI3Xy_;22-~yuF4N#$Wy{PTgoz#<+ z?%*%l-fN~<{_ITu%~!PMC1bJ`3bx6<`$Sd}%n>1>D}@lT?8H!Q`h7|31Cqm?a0A*n(8P0liXI2BZGk(kHN^+Kj!cs zpZkI{{tJjTgp>dY9b%1dXV#hW;vNksQwiWg>fcQ!$_n(IbNQ+rqa+#ian5<2ln@Uv z!Zb+-0mH;Tf4$%R>hdM*!kAj1$nR=@9;BBJJkl@L;r9=z*U8HGOVXvy+hEiuAscM} zX;N)+Wu518bA7~7m7#g{@0*mf&_J6wea9n5}B(QK)+oWe39dJ z_$E*b7H6<;QyFwP2={8h$B>!KG3qCUIY|3$Md6rA9d{(PT*Pgk)5*woe3;PFv&S-@ zsz&d^H0-Xte%>#4mKj#Q1S=b8WkN)-75o-k$4czut-^V0x z7BjJi&K=G0RiT*)p8G8v zxDw{fuO0yr+p2csh83!O^DznlNFD`H^qjY$v(rgPPppKU3lOn1MRd;%dw04we%g4o z&FRO($7^7o5&#q6*vS#cDIxAlS~(N^6^k)hKW*8&k%U~Z^S!p!;pDwTBwJAQFTa8t*kfNno{+uW5`*RdO%?O!ZT_lC=}ekOz-&N3s| z5q@t~@XCdQn9h-B$Cs4x-6dxrTG7wBchMw=haYWcZ;VS&)66cDwFNUjw#aCn!~4rC z+V*-~-SfS(Yo1Nc+{EkQ2P&3tbE4o#+4T$fJjbQ-#Ti!7-p9j`979=+oROb4f%&)48K z2qumQpqz!mB=eRikoNNm`uP4w@IAmcy8qbPmoc%7 z#AHGm<*hrIbc`i5-K3(|)zYPb;+RaP`QyogWean@tQI35%4>uIv7k^WcA>{L7@h; zuqz*ZV|mm1f)CA|+SIds0f)*oHU6f|Ry6Z-ljFyfPZ?K59MP z1gJPk4En)69QK^@VnVUF7ttq^ET3B!ihjZ|FdvhU|Hj z+mxJMGAT;$Vw3^HO{x6${|!=UakxW&TQY_K;B|Zf%}bTL@Un7;KM$)sntJ5L&Y*vP5$ z=Py9SS6JSuDld0<{5dy;+4A#nB5jrKM-Q_C+aWVpB8=nmI%E*X`bsDPCe)LQd&5hU znctg4)r5_Yt1fPr5A~|CjXoF`)n-ANlb{z-S2GUIWg_mrNWK;)?*7zBl(#Y-F<|r8 zl>6GF9-F##!#Jk`T5gqAJ``$Y0v)Q^qSmacEe>fEO`Ly!J5-C|_N#~}P(-eki{^y_ z5T|+#|Kp>}p9}#lhiB~HJ4EuTT?4ZBVedl{_RXMshDyE~om?(pg#WmYH8?SuYfuV} zc+d*>L>Ee~&b_`bcjfOm%vdrf{*tQL>0nsc^@H7Do9q%(L%=v;Q&Zo1VGuyE)62LH zLo{_K)5VYA%FK1-g-~c4?V-H|fLN^&Pe&3)mi)Qsaf%~)K;?_~pvuy9YXb`b)qgNv zulyvWh?gllqU2FeL)VY2U?|J)@54v@1u?~=RQItTvV)@;f&Y)Eua1kdd*0rlq>*kE zX%LX^RZzO6q(K^K=~$FfDWw~vySr8?$)y_>>8^#P;oawXzQ1?x{cQYk-@ErYGuK=* zXU+^^+n;Z3?5iyQAgVgL0z=6eSuzfshA-%e5z+ReUXCM!EO1-xgtYr1(6lm=Y|Cav zB86nN1mK%YY05bo9aDsJ>hfShA(1TmQ--JwUFWCsQz8dbr(;Ln)%`{UvY6%fy?PEk z4sJ~$FtH5ihS)x26_~hqu8#Ni6E!-Gj<_WrBX zAgUdf+gVv~c{-+~L&PVih!S$ZCkg$9KkhkN;prc^Ay>oI=R5){*IfN+l8E0s)n0nd z>c9FwdD%Ce{vH!*

#yrE(5kO~u5G!LPKx#yK^{q0S`$58BN4X%WN@)JC2@?RenP zODd?hk}JCScUA#NaD+7VGaLev0iP8TlGJz zmRx_7eJ1pIR&T;}R-W>&DDs~w9>JQY4)-`+ey^WS-2NS)hZhbI#_DQ+HH>vH6z&qi zvsbG{d@Q>V^R=bJd6BdHI&aU1{aX03nUugcH&zc5B}NwRLS_mxQY z)^28HyW?l64_-3$M+7(ogtVmS{*@GO<2aQeHsTXkuK$#4`G+AY1bsmlekJF!#jsRw zN?JC(3_8fLG9#ON9I6giSu_rYxOn~Z&|frn^`g+HqcB`tT9Wc+`o)3~vG9y~Cs4-8 z>a+AG!+;>Q`PMxp-xjB-54V}0yM6fyOeFws8rF`Y%k4(q+C2yI((VU(`d|~ctk^cv zIsjW`kmx4$L?JekcV4hZ0Ej_6w&Tp^mB4`uL`32&a7d7;Xe_mtO%LcPeD!$T3WH8F z-umfXjcrp=GH-B8x(AL_u;2!r8HkC1&Mo;pa5Z}8Q+y39!!M`xV-M6&+oVD<5*#hm zc~pFWkuH5#lw0>m!4RftRYUa0?RS07juGrb<;Rz*HW1671D}NeVYzg4vm?&@G~vWR zs&Mu<99#e~}rBrUr^qINF+BRU@iyR2A@Zn6~gz7|hGX1f~1Wc5OU9;yj zv<|DO)kH+klECOfozC%;$fW>^E;=E-QkNh+J4vPxH!bH}2XgzfX?(^5B;#u>q6Aan zJ5CWXd%GLk%6fD9j>w1p80* zU|Dqs!OEzzu|fk`AqYU$3vSq*>C5>%uhang>RENBED0v#aM&s2!e2z#&G+jwNaCSH z%;m6i`X*|4sKU(A&p?8Kw>$zZArws?e?2T7oWms3ntAT$@bi`ve4D+TV6FLH%Dlp) z$mTQ6dyO~XSv);53wORD8-5HL4{ScNZ)4HJzxak?!asJXL`f-w)CyQQtx_;U4kW~O zbYHGm$K8wQUILF{k0SvpYIBWsLyx>MW=!5@ixOvSz56WB6a42o$ro_}z%R8B?zREZ zxl_I_P`~i^Kcrj7m5&z43emCG=H20om8fSM0<6z(i6QnykIOtf_+>?yvR^A$8Pgpz z+02UG$*rJzk{-5804N=v#1e5p#~giWaw66XC%1m{8tke@fzCVzq9V4EPkiJM&AiYn zXR|-9qLB(PGal@0YGR7)468GY!}+j^866;l$D=5XEI>`g$gyUj(7DzPmrZl#J$v%b zGm-p9u?LwcX$%nIWcn#M;8icRe6a+u5tr2-y_*{XCc{~YSZE4Uq?`wc?Z#l z>H7Y0$;W(QG;mr8(87QdeLWY0tzeUDmTwSnWG&;tZ5LVF(f0y@m(aQUql7X0fJR27 zL*jUK_FGafr6Xn764i69OJU;KxE=2NvKy?iDKg;^GElXlP!8f`%woGp@QgwKZ< zlwqdpQdFj00nBQ8s#8A(!OY+02vb`CWV4F6=WU?n+7%mi6oBn-n6}tW@x`OeGiI=R zkKork^g!g;Aux^^F^T;4q8D^Q|9g6m6-dQPc_ODJm0pH`U+0C2kV z+?>0NmC9350-hIgn`3A=Hd$FNnXLR0UNYTPfkip?p6sRrpk(!67JNeunpmEAKP`Gd zEU5Z`O;ANqGDISKv4?t6?aEwt@GVQEPw;r6ZY!xTV6G%NK|~t6jH{$~K*(g_Baij% zC-Sc{Z32uh|49!KvRkdy;HEF)@)!=Bg85#SH4hW&0{HdMM5?@+ z!@%#lB`eeU_=6Qh$Ueyd287SPy5#MyeFWvPJ=;CpCDC;sIs_jw?`CJ~Cmip3-k>lG zu9vX@DwqI|yS4>OY%!!|8=uTnUBWIPWo3Mt48qz36C{aHOWS5G1TmQ4Kgf3yHnsBZ z6D{GdxE0!5k$QqCUq}!C7{%nvDi$w01iWXAuO244g4&!AKZ5$}iv@2${gs-dKSES` zyCGt9-qV~dg)fODKaeJ9cY<*kXqS4cz!4vc7s}%?1W!r`sxZ(!G3V^B0nKc_^Ql>w z{sm(3eAHNS3HYTHHQ)Efjz0McPIC%C&vI4G0onG&sIfxztCN=se=OgTHWhxMxA(zQ zn3+9GbQNC7FcDa`1<>u5Sgub(TOtJM2Z85i9$L{rDxjBKVCz=AkUkqZ*1FQbvq{jZ zGUo(~`^t`S7jZq7x4#i(_6VWv<~zn%hL^YQ2!e>I?ma-J~tlWa?-D_zZ|tyt-oC{I<+{0De4QN=b7K(hV%I(dh$n8EU)l zh%z->evikC6;|T;zIm{gFgZYl0Z_7!)B!=8Y>hu@J&xE(&{s$dMt>pk5hXs%{U~S; z9qpLUe8WR5#2eUrixV-D>a7_KoFrwD@W39bO`X+xlcm9JyR&N3 zkr-tqX{f{jU=iU;0!E`IoH#Pc$TZMWnv+$3N$&*{Bk|$;p^6$XgL))#`aKqYZi1sN zFj+)LEElwbdPVQ-NGQs!gGllM9Wo1rVGz4V$%JiHjvY0R_uiPWjZamX+~y>{{7bt0gc#Q?M6NPWV>s|CvUjnuK9t zQDyV&^V#W*R1rY08f}`r7BTT1-}QSBIt@!r)yfhIN49_MDAndx2>tqy`j;pR>)aQB zlqKWM6l<`fCzUw$E)bv_IOH>Ogg^bq5YQvyt*~zf*oi-hz+Qi%q{{;2i{bAtIK@&? zD%MpdT9%e}4X&=G!`8QA&eD`E>|R(>|(>a~K^mUSVt1XhftlDj!GT}UDsu6LpGByoPdBhUrOk5Eal_lvnafi{_ja{+wh z199<`s@-_-nQADZ-EiVR9EB?Es{K!(sC-VDI&|Qi)3m&b9<4&&Efagew?P!!$&F+y z{J$(K{3IGk|EEK;rzQwRT>5pb@kid9TAN8f$Lj@es`2;?ESme~=@jq7J*(t%{um)Sw|<7TvIXPQICYHIzAL{`oq%42v$?Gu9AbU_LE}06_%+;OsL9 z76&b{or5E{DyGH2A2FQ?=p~_&X+Ld5hdKAtFO!?EM{y(6co?;!u--FAw@sUPmZ-7g zX7j%cvcfQP)zdLm#MDRkN9MwU)s#;vpJ4CjPR6X$KvPO5l{nsDQ-b;E%`u1>92JwPQnw8U&CDR|8qZdD`qh(}8r^fq4KI=)CcB zEGCjt0D#(MD@s;tjR}xFMY|W5cr@y20aYLe*Z~1pWY{aXm5e)A)SGPnN7@}KdkT(B zXAzyFNMQs@Y`xzo@gJ1z;*+6BxY2!7u0bP8@A*zViFS6z4RI$c3)S6vNADnxGfFld zKdv8ebfN#BdWKy+6wRXT;yxb!&zCYWX?WW51eUv_i$87a(^Iu$lVv#@^!l%nc$`hOpN zuziqt1b;5eD;=$^?d>frxSu7WQe6%H8_ipNeS2!!$t=_JsGh29KF)z*kHUC zyqjBAU{$!3R^gdrZ@{aI*(xjQ|7OfNI7}7x-{}Czf9%16f+G>+j*}shE?jvj?IqMy zI%-kzDudEhk^kRYxTPY48?~u^cZ6yOu*vjwLg&J*Kj^zqj0#QZjHiHEG(fKresyii z+h>0*pzpI84defJAD`O1b{RZl1<(w_iT(B#{U)<}%TgEi(l%|nE~l%pgY+b4GRV*3 zr&{5R-mPJo@%j@%KNes3hasz8z9C*&yZj=pk#6-lboEy2&n(b6=#LZv^oYl(B^DT1 zTa^%8;$Q5fzhk*;-9G)@IyhoCziAX(`v%2%s@AIJDdBiQ#r3T{H|bGEgWNWm*1EOw zi~nZCH1@(5thLRQlt+{iVcx+7C;qAdxZQ~`*BZ@_Tb6(Nc*pr0;`pi=LAG)9NWcZO zXIlOv$eKqTphk|jav#DZ%U6nb|&!QSa6P@ zT3FPt#!r12o__mSGYjxKNA&myQqVk1tt$lR#YF=r(6-C=t-}OP;opSbr848PV2972 z>$8Z`jP|WgPG2vIkUu{1a@BR@|4-yL{bFBXH=hoQ_eK%myGI5>_Y3p1=kG&^j#VDL z<}+(DuM)cbOePUA&E(nAb=^lk`8|F-i~Bg)3$&k}%ci)aCK0ZZKXUr8}z|{1r?fds_L~ zYl9VtPg#j=l>zr;!yD(p(-I5MovE}8G zCk58AJq!0CcCI*gx<;0y|4BX{A`T(l(_&~i<4@QD>CQeT_1OEL`minJ@!W!F<(Yd0H_a#*oa+?a_nck5WZ2jJsJ#n;A3gsV zLcNjzUz)jkGmwQ?tML=JAs{~ctuao8D_15c9sh9kC<`IG8<7sZFo0V*h>J(slpvfnRZE41Mj zkGL24$Wy(37Ct(MdiCH!VV@>@*2lIsXMu*6#XDmznd-Xr?9t}}at{qFbCsyJRlkzo z+B*NKy}zDX0wqH68>6rLx35wE{gEz!YQdwZI(w-6M6`C3SyB+i!+i+k`GhM6aure)IP00l$k(v(;4 zCSBu~ zuLq)k^ewfbn+;8sp6~=tZ&5^_p0_t)p(nGH`MQ;|wMVOZ={G-v>Rmu)x<&I>-TlI< zuKyXRE3)OAfOzdP0aJ|pOf7XgY^nfsmJloSg>gy+_gy5inPXZhS{-p8bXXfUP7PB` z?^b*+vN_`<-$`J}heZmsYI*VHpocH()pW(Hde>K#D9beT&sxSNYDM))Kc;cw6y)Ar z+U=v}>macJeBM*Yxe|0FK?f=D>A6EFZscvR&f`ytME?J+V=U7GM$j%#ZG<6xqmdzr zzP|sjhI}}bU%59g)gM>%ME^mTx?{2qwE zjYFFU5^nyG-XesH<3K#k)V2aQ=fQim6lWV0@o z0TVng7CxIxGc2&ZVpF3a5qF)MV?|44a=bH#p!S%T!#JvuB-$u}a%V%m9qcOU$(q3% zI*w0Dg@>R?M&^(a?(;7w<-Z@%f=^F;l(V<{E&uCP z$YCpc-MN1Q@&9I*#T@Nl8|+6eiP@?%LZ0U~4LAe?xv=t%e2D*$#||>z?5N8erTn!C z%i_A*SwMuFe(0RnqX0h=Bei&}7Z`M7S7=Au!v+W!pi`5Se(WuMc-2H{6L|0I>OMDo z--rU*Y-W6L4+J&F9Zf|CQvOVp7m6vtFf#S9%n9QA=YBJR5}_?;@6m#fv?Q%l$GKF1P*AKblpyg`+BuL`srINi3HG+P@}R z99YQb46b%xzmcv(1bdktmEYv=aaZ1Sto8q0jCk7I)iZw;|I>R)JLhh9?5@%O;^UIZ z*ZgX-oCjV4*Nneh4K_jn0S|9VyR~}SUImnVIgeYwrKdV+oW*3#BPY$Q# zo;4dV{eM|`jAmI^&m!TG@b`vHW+w-jUKD^7TAPMHvlM(%w!Whtf@Y0#PPG{%e3w-I z_`^U!IoZ0v4UD>51j;$`1;IpoC-4sM4lJt+aju9v7Xog4IhFeo_~0655Y~ym1u#}N z4p`b;-?%so8gUr`0JdBRw*a4GehrRd@j3BVh%AG+7xeEfNkBvCD6gyjiC{RxQ9fCmv4+i#DkL+k#?&h&wv(Zc`P0xQBIbnzlc41QmJo4|!kfaqlB5YA z?1(!aj-LWynY@Qh8Uf}xPu4=;#NWU8S8U|ue$Zz0XniJ}%u+45KQKLENE83W)AK*T zCY02pLMoPVHF!SsF2m@N5)+5_Kn!vM8st7OsZw`6JXFq2< z_=FZmBJe=|az})T<*b zi00r=G2Dko`-`=?`7vIq|fo-X$>=h>Tro-Rj z@`s&cI~{gtjs(3pUO$cVczFZZyPuXQf57xDJi5!?!-bt!7`{uIB{DrSlxciWez0m3 z0XU45Pdj@9*!AMs2BZMfMFh^-crAxJd|1bk$0}!3sH=l*hbb9ZiuRZ%=n$q^sxp$U zXgps@;WdE&5x<|M4d^o+wi*?~BNdh8I(O%h%=pMkMR#H#4CeV~7 zeFAWQzk&Jv5#EYIRPOHD1E^*3OS^{%bP{h|McseL_^_GVllnktlXE+3G?V<8+1|$VT=K-R1156;=)Po-D{E?3FVQ;eTMv5 za$;`Dl9}w%a)<43VKjjDS(|zNg+%$oX?I%gUR9>y(JMN?p>Ka3a<%!Ob@mmR{xe-m zojuf<;VWN|0JwEaGKYxs3$m^Pb?u|_heY~NJ%{IO#2sNgx^eCqKV;XJ-iMn7i5hDrqkM$0~SX+q({b@c)F=-Ob8_v*Q0uru8W zMwRrGHdBS`*x{UL=ptNS|1had+D^>wrjhl{*J6qrDk{Uk^uWuPX7LBMolm@=1Z&d^ zPGVTy544@4Pb6Z!IZu*DsFRzm-uzw!K^mi%WW_Ooq?;<~!X>Xy-NsRQcX2oi6w*v> z^ezhG9=o%$+3Il}7p*tIjotXxwpGmOCAQ&57Jq)lFip|u_f)0qBC(Pu#i$xw&=3cz*hf=5Wz=wkJEts1^g41;qC2# z^8E)kUPRcbydS=IRFm`VWzC%zG+-`H_^o>oV55%2_>Nt^5t(io%6N^RAg@QHP;lv$ znB?aY^ESX<6Iw5)OHj7;!8^TDLhQs;?rP()1LVTkOhr(KSO8w(FvcZ8hv16+$okrC z+6}x@ik7dAPmabHaBpzCfUwVKyEB@u<>g((9)Er@7u(CIu$?t3^AeBlFd%QTS_k%r zl=*@v5HQ4j(dblzN!%*MD%pUu(m3EemEh|<6W)G)VI-4{`Qk2(Lr~V$%%VqeGuKox zp~K>UAMn4gU6-UTbeeu*=92$vJ`43a-aI*7oM$+ zE#vYyhQ622(!OeO+(ppr+F!<)S|FtkWjSvas8=SB5`ym;tP| z?gHeo~GF|@Heug|>SKM3v2wS4#x=f$wM$8Qns@44JT1fwjs6Gg(^84k@ zC%27D=g05r)>VNwR0pby%X4&;C*p- zIP}cJEPSTXp--`tsE(a=b_|8$8jb*s+*pMT?Zyo@O2@QCHYU5LF( zV7KB<(03%yzE(a@+U140L$^F*XrQuCS^FD#$k5#OE`L505CBVtfuUVyeawV6(6^~? z^V<)fWwD`^#&7)IS!=$ePUl1yl8R~Pl&6F-l$jzWa1)VYV%p|ky^BuNmn|nyGvowk zIC-Rw3mc!Jct2+KAj#o6-3MKsS4J-)2Q=orXjTk*-HXEgPOsB1+#3D_5=U^-_=wss zgu{{LSrR+mwZqobD&c@%hMyQai|~v%-8e4DNtv9?jXqyWL(E=fpB$#b8z<0WOD$pZ z?t38H@w2-g#?Aq^A+ke2U&^Nl@Yf(P^M^~dovMi)~79S_K znAxAZeu4Z{UoG$c4J&VHEWvmV9I8x%3ot%<=FRHOD(ruApI#s_<;6`9Z>)=Q;!uZG zr>rgJb*BMbd_qhUN~ZOPXcD$_^dUzu&lQySsblPwFE6k5lDgWTFW<4?mF*5kKNO8jD{x;H40=k615P zf#$vsth+BSV7g_aNfhzUel+LApq!q@G{+f>6K4%yojC6ncHL%UIF=*rk8( z$IRUAAi_rJ;1qM8@w#HK$fk(5L}pbSaozkNu}mXEj463={FZi-$skux&N&K&^-i>* zX}bFed2m+0m!&A+ntp_bz_;512HSzaN7ev__(0LqG39K>sTv(+6eh!hOvmkeJ{R~- z<`UQjHZ##^J}=PU$402KpSF#yQY>J#IV1sl6&LEBvniB^TOV|HXlHBL}$>PT)>puiu9#!2WvEN(#GpqPb?a8pj<8Reyh_j z1~iQqLO|(?uUU=6^K88HL%)Oud<5)=eLI_h=zA0!wc^B(aB3scVA6 zc%#x2PNTcfQcXQs=eoSC?w1{IIkg%UWGxP zW`*SsCXE=S5bW13+XQX1n2C>jr0+h0 zcUuJyJYlk^DmMvm*>Z%Fjx%=UeI_}-7CrOnm~H-p%=hpkh|17sa<7=X5JPua-? z!#hcdd?x`*wCQN2fmx}B_28oY5b5`yos!26_d`X)zFIygIeZF+CIShL^1!-HWy#f^ zQdNzYnQ_<@_mc1=W}A_Yxm9Rf4qIbpwW4P-n=`zc*^th~fAOTGkO3f;m)=#Ap}OqU zIw(BQOOJyhbUO3rCmwwP+fXxH(MRO+s=o&_q)L}uZPU5HFHBazhD;6mkJ8w@Gm*x;eb7?LbkJ9uUeVYTwRuGe&hX*XM zoHy79(a*6)5R#pjKP5EoovwP(9RQD<3v0Qo4QZP+PAwQajbKX+=%sf|U@QDky0sEJ z78;WKr3(rp59T>~4$9mQkUX_hG&>#W|1Pi^yyspy)8>QSMDHoV&PPO?@kCsHNC0M; z)@RnZ`7eQzESLl(#gH*lR5n$qqAdu&W+t9$)9tmJ(hkHuyfBmV>_t9}jF*C~h?``b zNC^)B@N!I1?&aH=QdYVtSJcD9HDg<5{%B-U$n`|*igLU~JHhK{qExiB&LEGx7_(vy zx8T*4j3OSNMq7qMGHZx^9ScUwdHV^cLJt+VGB0q%PJS^&SYLIWB;@(GQ7=;08L2f8?B{6eSM!fj@{lv~4MXiNV< zUgWmHm{4weuoZ98xslNKgI~8yMtvWRu+`==Qe9A3d<}ca^l{&S4yf|ywNJhdBTFxG zd-;1wdW&J3LTTF02z*=e^2ai1gW*?AM0iIF+2#v;S+YI(06nP%jsI z*4NI0_8unHRT^$)THx{0@{jFk3qPc9@|k6H*%waj5LK%0pe|0zv-~uMvT_2_K-fwL zG6C!YF5C;6M{y0%q-HN`I3=Xnbt7-QblaCyx|2TO;CjoXFCf*>J$vV*V2zd)aJ7o0 z9AaMeulUIlTWN6b<3k514uAW)1f%I$%bDkLy1##e(A@NUYAnJ1DR$GPzcBv0F$%Uu z@J_-RYKI2YC4k@guq93(JrF z_WEubjM|`cn%xT~T&7pA=KwDncx>$~nFHnnjAV3qa^6)c9R#~0;Ztqiv>FbRJd>2c zIpDA!d|S+>NSEf-bOE1-=k={7`=fNTjJ3|aPn1{A_5}{$B7j}|B2w*K@>ti8Vdp#U z;EQhv0ZDe?+YYUsmRBw7H=Wxa{OhL#fc3@0ng)jrLi*sh^x8bA3Q*oFPH*Oo8-by| ziOzbBFZ65ZKy($3{0`X7TYvqoZp~B!!`x2k{GdUPJMDsWfw8!A@_fgiEX6pzV(YcD z2aOv(c5zQ&7};{=OlIC1q6v{u$>?>GIg1E?TDgQ~dt02{8}LCFerkU#P2h=ST88^7T9QIVZJHBoc1bG5d%;zJ9bGbORVPN^ItC zalLwx6@HdO|CscnczfV8o+ypk_vmGqp-M<&mDAF7FQ_Y=fL7R+X+?Zyc-V??A@quz ziX|l_B{elU(%<3RrHgmQQ#w6iN=7m^p7&}0<20S1)quupit5R1JHtzFnAZ zz-Bs#PNa1H9u=8>jr8AhjPrlVyD<&vU) zN7B@ zgP%sfiHxMf6dm$vI&h(n1?597fhCe9AkN11wFGe`g8YC}*Im%=WBb;a&X{L@y+dD8 zC=3Zl^QWZ7^RdrzKMI}4K&vPER5K;1k`Dy^P=b3gyjima)eNo|z%Ba`?Q~yIiC+oR zRIoZP{9(V~`%u0dq*J!@We@>Q5JiLdQ!E#rRpFr?T8#sZ1I=NThQ99s;a>Da7*6lo zzxTH-YAT7u2H5 zHqyRbmxsWCQjf9gFzzL#FXcLkVM5*Yamr+7tG27f&HckS?HbKS75>e)^%`c{_o{x4 zJ|JeT82!sFF-WbvyPOn3Y^vgL-atQ!$nSB3;6A|Yr2F@b1&7&z^}&a8Y^II)(`Tvx z3_w_DSaaAQw57Zyw*ZIuBL_ODhXbog+fc)!H}m~<_Gt$Pf=Xy#$AxTHjNY3I#K9dV zrV1+0p8$e!LJ)SqmMA z310C^7edc@yzQ=YIR4yn$5Ec!+g+}DgwW5ea@2kO4dW*E`dgoqjJnGlkw$c|EZuY) z9TiA6#f!5@)uIuA;3iHhegh7@Ci-U)h2`?7UDfwaHiu~H7-Qf|)o7w{?f+%4T(J-n z9%p>)zNvIb*S|R;m{$?I(yht~q;Ogjb^=SNG%(J$KN2;*^O(KjhR(QFE-|D8GYw`B z!Tq5O!cwXa+t}XS-rI|&DSy^qe#dr=u?A=hsB`dJGD881EIr_{Fk4v_iWPY0gX3JI zgc+i%BW8(v#s3fq=QYM?C`Y7v&3uG0xAu8Aro%q}BtlnU;S&v1P?3>=?R?jc&1I=- zoLTD%}m{y#I;qHX4@cf z3#y4zh;xz)605kG7V4_ILzGx+ABasov&@f`>iFH}9AXWVSVUt%o9k&=yL5rSi3EZZ z7Lc;wITH=({uM#0`4R>x3{LVZ$0}2w)7Rg+yF68?Nqb)CR39jQoW1(lj))S(A3XNF zS_oLHw&`j;4U2feJU~$LsC)98Yg7U>!_RAav6gc5)wj!EFpRs9{s{(fepph8kw2Zr zu4nN=&p`fSATL{II^7Q}{bXlzcs^0Et8%Ymmnp#BroMHanhIgQc*;=#v9w&PKGy}^ zgPo^vFG(Zl=$n~2WvW~*KjA9dgLg+#`%@cWF8Epev4&TEU$|-QNXWruM~!Oj#mvVR z@WHMp?nQJw!HIE(k+$iNs+#R2UX7{kGlW(QKoDZ`8 zvRr_Z##Pg2Dqw%pYim^rLfdeZEhcQ+&O^W&P@oVeKO%^&Vwk-7;spD5>uZlFf5P?M zAA3$HFyyl}LF!a7{=Q^O?esFQ#>0>6F}q-GHn=^ZoLakvh^k-1Fm{<~r+HUrhX zK6&OnnrwT9E?7AQ-QCLvCXXX<_ z#xq-%f;Q!cvD1cuBgOEH=2;^B)lTo@uktWr{MQe9(J{}-wK;k*sf^!ybUfh#CdAo) zwB|jOf&Rb!EhYppN1mlqy9GWW z*1ETv8?}bEy=+Ld0o`yOss@ZwC+qH1;;+>LN~%l#mi&TU^Xy$as^QU9i(9z0(QA+j zX`UzCi1VLZo@tjuf3buO?d##Y?C(+S6@fC3d`<3q&=Ef5iF<(oK$TTEW7bR@Dqi%t z1U6S_BVw$N;jxUAKSI4@%FC-#S31hY`KnLpkH5dy;dSxL#cV#g^AqbO5G_6*hBpCQ zhnY8976fLPx<&@ItQ#B15hQhRhLqR-JT1mzv2WvE`w+A-<3h0#&JKrxfdyWy0OpEl z$CE8VNWDVB8*-ep`3dAPEC7}<<~6e9j`@Oe(1_cTA$z)a-l|j&P+{8X`LL;)U252f zq|a*id{CFY2$B<KMK5 zFdP7M>2&qs)8Vd^%sRguxw@io#!zy6-rPD%mF;#tFgQqrwN7|S+6+D}9L#QIP6hac zuWP%>Py)MmfO6|N2pVcas#fX-noY2K>PWK?j})}$A;W8FwF0tZoWGZ}jS(-d`Oes% zF#AK^`J@T(TI$fy{SROE?GN0>K{@bWo!_Nzn)2GD-p|MO3xCGtb8t#T@8zS~l(}8W z7KbjrJYAbvi)w0^(7d6RpxtgCKcH(5uDl;ottL4j;Or%m6%U2gkV`fts{gI|f))y;nAhiy0L zp0$ZPgN5$kp%0U2nci3;+<1 z3b+g|>1k2g5cV(f(>3as{1{Y~v<&ILab znsx1*lQ)I{_iD;J{Q` zXgu#e~vTm=bVjdfjzv zXc$fKo2H_|lE~9UQtZT-;pu#Q!v|G^RJj)gKRVMFf>Y zpHsHz?#)VGbp+MjhhG)6iO7ooUj|s9W_P#@*<*BubfFGJ>YseLf1b#jvGP<6yVz0d zHzysA*6@Za{_mNJM0o%RNA6o)`;Kl_mk~4qu4g<4$t|^RDRj^jM*aQ{yfhzWVJHH}bVN+lArbIRYZsK#wo( z$jZ>}1U1|XtW?2u{&V(v{Z1Ms{R5?UphvHp;S+9r--{SXN0au8Oy|ZLSurqsuu}Hv zDRNM|UcX(x{*sw<6q_~bZqiM~HgS}l9|U>x{+BM(Jm))7cU%Sb_*j|m%w3%hGQnQc zTt0FBR2oxByNX9edWnWSh&hJ1UHMHn1J?Jccb<3K0j==G!!1Ceh1zGjL4SWgZ>@{! zt_dO7<=JdR34dmj}y|Jp76%^yGi-<3%P=CJB{G)Jx~4QH$XMlHtp$l z+_C)pxzk=mF0KQZQCYV;2Y-K6g@-d*VYvwo_y`&DXKO^gT0_Qww2HOpPfBpRdwLG0 zP^yQG&%K({Pv-GfZei{|&#ytRWxqCX#A9uXH>XV&6COKm)IR@JiwKN%)_Lgoc%!ze zbi(yzCKUv~{bYj~4oEoCb<*$PKfn14eO!jYb%Hp$Pi9Ob8+rMhWy4nQ-)I6_u4G<} z)4K2J>yWj0b&Lk+&As_o*8F+wC-oc3>~${3Eymg24A_1?JL}LIvsF>?@O)F9xAr!N zRo3FPWT1cK+EqC4H0NPj-kzq1hL~NfxRF(kQ?a6eX*@Tt)ob5&$@ju(aqY30)`;0x z%rsBE?e)Ijnt|>WE_u0l3)?rVHQgK0?tB{O3R9Zo6Cg`8yM9sqd_}nXLu)CTum2jv zed?2* z#hc#eSRUUz3eUU%X}-b@)9T9&Xl~bak`h>zQqIX@$D(>xdGAv={GJ{lj2%_Z_n#s$pg8CDrI=0Y~l_XjbQ#nejy7jZ~{^{ zP*WOg<}3(?Vc9!CDC~8!?j3#VE_DDj4c5fi>w^3TZb03{q_r z6Z&x`B@AUO%VVmtOw~J$^h$u5`9Ia)ATX44o`3A{URd9YS`yeCSXYB#-r$Au@_=x& zVWwB_v)2=ToPV_o-f_6+6yE}xB}}!%O}#OLbVHrm!vGCcp#&lq7eRls-BdZZL}aCD zog60XmtUe|3Z2%&gxI(BB3_Gu_;RUo2~qC|)@8CfO_qN{o#lKHb=CcScAtdIQ`?U^ zo#Fyj-G%6YD*X~xy+e-BAXHkbvorSZg-~8RUNG5!-nL@kcYBGh!n*LLl=;_6?(~m5-#gx1 zO?#g=-ke7Hn)!V(soF@?aKKt4D*EvZD^7~vajElh@0{9GZqKg@ECnYiV?Lnw9rqbm z^)ln>Uec$s@vT$l9)775?qC0AS^3wbwzefDtgJ<^4nq$PGA#qUI9BD&HsYL_H`E{~ zbb)NZ`Roa~n!wE~t>XNO$l^xHR2o{(MA~f8DyYrhs&%ROxdt}Pa^n{;5;?Ezph?HFVxD5){;_ zb~Xo^0G$4O-F0?Bk?V_PhZ`(*8eCNS=leOKGqx=&8kG*yYO>H{#@8cW2sq?ZRr3#> z$DKi2P&qn>Q1sHX%uR;VB-9JJx@LDx69z`@`Jgn*Eaq+N`G=h=0P1eLC_16E`vW+} zmLL>XRAGd`5iJijOH;)bUVSy@^!k*JmS2X4i!JB*T9TE2^HA;uqg1Sn42%p_QqPWk zYK`KEmIb;e7X2Lh62~c{)uQJJM3uzXp@Q!lciwLY2D)9@6aHekKZ&skehI(42?^A= zLV9~j-(8yZ59PC@O7r!HPrid`naT4vWFW(2_L1M{QGU-3u+*7WiV$zat9CI3}=p@T$z;%eC7;agF^jmn;RWi4vu22$g`@q_DFY@btO7}^68FpZY zuL|4-spF{bFENOF7eR{#=%D~ds^DAJ)8H}#NNQ1E#J1-sw;Z7^W!6FY2%0UT1j&Sk zUA{of?OaN|LbZ=c!THgeA95lVY1vOeYN+$QaMMk-E4EWN&du2lE)VB67=2V{6v`39#8(Fd6R&k)=FD!Inb z;m{&Jl6SiyRi7cG{crV8tW;>xflFjqpCh5jmzv|i8jFXB**vs=h0&N$am7= z(zXz6!*8IU9*7j4*w#4D@Wo1olP)M+L?Qw=C2Wfcn3tE#ED}RgHHu-M`cm9a%%*IC z9(I+6J{+4bEx}Cpt0%N>lV2Ftg^WC?T_`RCU(}9tlOux@XHSWsG>et2f;mhq9lb*! zma$h+zAfNx2f4jVLX8?tdp*M;5Hfi@2}f2V!H9NYxc5-*e1ftmcmI6f8igkYeSR5S zp4(%gocx0$Ykjy+-M-CFE|!a7^Q5PcPy?z&MfZ>DJO*fy?L-*HS78btdm_uV9_8vuLXOK#q7m zZ4Ea_6BeSJ7g7D292yOMArIG)=z;yn@^QNww{<<@l0Iz*n+?8saMz|1#-BY+M@Z1P zC+(@S&S=wQ<^W;{;@Vi)ts5@)WmkAO3`~=%hZPOJE(V4`)F-pJWa!shJo36M5P1@P z1yy5~%;hVe2lU6-apg^SDzG9s@dPU|6v@A`MHTa~3zXwv_qe##Kfp3tP?w`*s5CG{ z3iWeaniQm}wZgTo9SO(ZO`wo(3N(qn z2!>GJ#PO99IkXr)V3~`)AcZXoygMtY$a=dit-+~8w65~CITQ=>S@4b+#iZ;*MpRxI*)m| zv;o}Bk3|TuL{J`68uqH$r%%dFoV*uN$r$Hy-|2;4S@R;CwQ}DDL8rNk){s)xrN{a+ zfLbmfL=AkTRYM{d53{^t1%fM4Rc3%fkPo8S!@?4`^n1ny8IGJCIS4%~yAxvV=}tJFJ2cJ{E?y$wV7(V4 zo!b!anB_40M*``ohNI(h>NO_v?23Z58H3_M*W%XUye_i)>DORf=JKh;l)Xv@l4 z1fYAutvd*+|Xl$kgZl1&DCaH9@@ujIDr-B6u7fxLy-x_!i`-Ss|g zD+Br@z~910(1L#y;CJmo)~ZlD_hd|vXBJ2O;xHvU#NR2l zm+7^hcQSl595?ycWqyvcAwHE1Cxzf1RK52f8RMjh*Ra zq*&Qe*5G3igQxoELQXGd4Rq=9zOmQEw$|OZI5qxHa!mAY>4_<2;C%5t&q`(nXXVu-fFYP%cMya>3!o-@S2n) z@T3k`6uRaC9n0mCo5br`GQBd8>}o zk@e}iEmmWLlf~wX1JzU^ez)I1%BEk%I2LTLs4VDB%mpWASwtA_EwDz|fCHLvlb@>8 zI&H=Qeu_zwpHW$Nk0~KHkbOaTMn!RY>yb&{-1hm7LTyseomqO&!npg5y-ilLC^y1a zBq{8Uq3$@?!Mv9Q)nF0d!*kke5wy3qJsx~BoFCmq5m)CM{(L@-T1+$T{AJi+l&uu+ zUg`$*k&f}>m*0wXkk?}p6ZkjqmZ#6c-H~;@>MUP8>bCMjEFdjW`WEY1RW|cPZd4t% zGjdeFM_z;N=F4|5qFa*$TbYjUR$HBrx+M2Bz}UPquaLH+ur2nM_neG6GP1UZI zz^fp5iV2USBek@)Yx?qX3Lq5MdDEdF4{~@XomOjRXjK>wbP)y#nbv@s7X@`@ztqP* ztv3uU8haRM(a$z!rRx=Fk!`M8O83dPd)`p2$T7=c@}%6@;C$Y~Zd><=;70T+a7(!N zw^xZbVa256w)DRGUV4Ajaj-ip3%;`7V`_%iA(wSECwI>CtWcP~@6zFalLg7&U0~1F zxtP+QSla(#N(Y_}va5>rq%yrS)1;#RbD3v3$7e>mk_3=Jwq-z27_N(O)Q9lJEf&FE zSs!!w$easNqE+mbU~K(}=2L+$S##5PWND{a@#$1a(O~lGL9R1S72-dZn7@EIb$kDq zT0LaADWXyKDP((m52MnmIu&1_vjhb~@l9{%f3qo@<2EI#$&H&(fV9br9s$iIYaR#6 zCjT^Y7}`Q$_)f+NqCTxGnSBl_SltZb#p`#dE^TTb!m!%LXW&rO433p=FW&A9N+2x8 zHhauZ70L0`hd3zUsqw5kX&;ibHe%?@9B~IBP6$GdpC`*ro11eD0VnuO-BstdaH3{{ zFZ#yzBq07g^3dS0-S(oKkwD3aJUFnG=xMO?)!PP2H44C~j|=F+cP&bC>W_gFB&GxPlw`4Nfh93cfo%8df~k}px?jK-Ii4B`%S7=?-ib2 z2B>{8(T_ss&w!O8@N~5FB`KPIGMwsIlzoe2DzvB7!lsF9HcdK9+>t^a|;m7!TC-1GL-G27MG;@k7_*@iHI;gMHu?NcR5 zs!ElA8TT`9V$?n(VlF+qdDODiv%0)n$iogndJ~s+TX7Sltc{w4!5>xon63u7&};)AI_CETKyNkZ6b?MHxLd)rXp^A7f6aT$$MbfWbN@<< z9`oZXU+|$sineYFEsK5IrTG2^c31sJtFWQGT*KU>JVyH8vfep`1|zv8zvxVOi0wA5 zT~Bl@^elmzK89vUJ$Zi4q%$jf|%Ns_RONbMCc7DOl>jhWHO$R=K6>u9xj1MvJ=_89Jzoce7Q6q zJPrahq0#r6{tei6&UGwIqJ*ijATEZgAM_Z&Z|y8Q|#F z!ZB-2f(;lF8-3bU7If|>>+_o#t&%)0z87&z#Z*qC*)4^^Bp8Z_%EU_WL1lLzqX&(Z9{Y|1qzun`); zZV$;A*cX-L&knsS#b#V5_z3j{bz;PH8m?nGbTx-f!1Y&Sv#}`2*%)i;0#7B}`EW zg6)3$zX^iZF4hKG1dInX)SB3B@v`>|QI&i$s~O|mvPHdo0kQbV@jz?74mC;gC28(+ z>FleMAHB39*G~omBYnK{XNGw0=HH5>LLkz&t#kQ68{;w+KuGrVxE^T5r4F)bqF6uD z(6H0StNb2r=RqMy>PcPMh3D@}F_;)-z_LYXN^*s?zn^PJe*twc#pM=*O*dBS?k3pO z9UCO3jIkK^WZ?E&B=)WhS+=uS-}!YZ_|3a@NUk-*d#*dl+QUE^80h>ogOy!=Ytod0{!hipMOccPN zqlTQSG8Q!*WIr0qqlrMI&@{q6wC-9oX=Y7TdYap#?6%n$kA|c#Sqy(KByk369nU~o zTosP{|3u*acu_A;^!=|K+nJd>$pB%`4?15(QCyoSv!Zz!&yeglo$lN*7&nDsqxin@ zdrdS}4CYouX0z|CP$I9Gte=CrCuz>mUJd$bst8!tx+$@uh%gR;dUZ@*epM0WXHw8@ zi_L5{+Y*M&PG&RpxVo+c<*OF+j7Q@Il@3gG zrBljF!I_AE;J(VvNhY75yih!U8KmqC zk?2w1IyShD04-V^G)B-=G@Zhd;Z)}#_oFrIELV3p4EQEA^CosXHD!gd3AIPs3`=`! z)RMA$i8U;=Kg%bS(7LB@X8R@8C$S>qs&>ZnaADJ*YTFK>9)qqfAu<}S zAgxIb(+_Uk$ARhz_i+K)n#wf7OHV;93uqSc$GD0f- zyw^ZzfK-KuD7iViMFHnr8|z<3U&79^I>zj$x>4oHS|JaFG!h?SM4p_LmE_Tz2DbKn z)!O>Z1LP;J;N!_$m>2>ws2kLFN~g?PP~8uOF~yC1e4_UmYNN_ zK)rrCtxuKEOn0R1<73GqGtor(Ez$+g{A%~t!rvWF;WXllfw`?U;jResaVkQBcmTJ~ zMF!1>so*Egt1$)wMbWFIg37)oF__9zaG9M3{P5k9(I#-J4R$BwYd||~^}^_CKiOOS zK1^QHU^XkXRqSs_#L4G+HYoWSsanaX&SGPDJ$&E^Mel9Xhp8tUFN;%mzUU$M zSdiAfu6|fe*zriI=8^g4S%!5j#An+c*&Gli5myVm7QNR1iTy1m+dhOK5Gln8m zjJw!9`$qma6V{P|8rvf_*riug%xKfK#XYVXuedQkY`D?|KAV20a7Y`i_e&#z`V309 zjk0qh($<_rWzyfnTqo$&H25^$X8-1qbBAmf#|XPlEmSA|l5cg;`NB2SC)5F-#%W7W zB2&~`s@^@@V3goxht&cw{} z4l>?jqIEZ4zzLU)fAl9-!Ap+1C{i?C#hvx(T=u117N{d;JCm6RQ&n`w{4_T9ckD&% z-MGL#o?>^?%7luIq8S4#ztGoP{A~|Q_@YFNkDD!_Tje_1X-NyXotd<(t za_)XD$(1@O98FGq6i@8TxS=m2TA;4~k);{M9xuwCsoX&+02WfI!y*jj$i{ToKLk#( z4}_EHJh!~ec%bsb2|B2T4|a55I;LKI$$4RN`-caS*xmE3uTKXlY<%`^#98;(Ps>{w zmyofJ`QLG_GBrIexur-!{4NKs{_C85%j6M94E~VXIrE#oKgz8Ud4ufl4*229tOS96 zI`;3!o^nE$nL=p7BZq6AUXI;-!}oehX1Hi#?*#&y=1F116H2e(J-avkB=K19W|ldJ z`6TKq8kn`5E1QwCuBn^n=D?WUN*T{CeP2fOc{n&?*TXS>gs3arwU;^Xpf|75L@T_5 z3&*4#2jTDbf)gFjq)ZQXoKKsJsK}G;lIibbg_1AczMaJb5^OcT0>SgLX~wy^!-Y%Y z2=`ZX%=!sURERqfrCpwQ(gCfe4n1&FfBM`g$imb-Cpwd+yvYw_b5XxFdTjJ=o%!4* zQMmW>dGt{~xjp3Ge%}`j)HuVAzF%f>ZzW$XKLRO(l|q@>kged(I}23)Q})!*d6;YI zgSynsgYmoXQ`2G=#YQh+x<35kWYxXSZI;{_BI0gSghHyl0TIBO=9o)1~En$N?VfB$(dLbPT%WMfdI5_w$pHjqa;({&*I*_O9ozM&y^ z#0%=ZJ~$>rJQFRdb}GUVBfxQewGW+2%FRIr1M1WIT8l?w1iQk48aD1xxG~NM;MVyV z8tkt3MaS&`WH|6zQ-3 z8hu@xOF$$clE4Q~nt*?PxhHOluHoNbe|P1>!>)VP>z^G9G2#&^22uAg5UX>N)lZg}s+Ox-YAvCQcaew#E~sbi z&P#@bEU`rmDc;xU6Iz?x)roh4SgxYa5$3fVa{7FXpSKL>LK^}-x{BjF`gYFP3+SxQ zxOPz=iDeA0Q}dX$OsA)&{}%XyZkI+Ae}<|Q10tf7GS* zZUo)f(jB#}{a!*l^$N(xNhjuQhWZixP(nE(U_3d5 zirpikZ|%v#Ig!32FKl#$WuZMGPYZtg;ACC}`iDAo->j&r+W-3bB1alBP{Hai172hq zWkFKOr@F%*w1XDwDIN)#i{9UK)L#n*eTi*vs^3C?6>2p4VI*dI!tESmf?|nmz3^6D zFK+q8)ws_rzRf=6;>1WG!iUe7u0GN{Q&#Ej8!}bw#0%D+IBSPUA$Cfkcwbk=-NZLgBzp~9BdeMnOpwOB$Elnt4Zab zgWHl*7)prU5#iXIV7N`S3MdLmzy)nOpa{d0wih?=Qop`0YHujI?NTax;UZGBN2%kg zC&0s>d!>;x|vD35@NuA#s##XF`b9d(*&b7A2>aPkooFII*H#GoMk*%lv zPQ<~ObG-zGpjdtUc3f{UyIr<;il!vb=zusWu5}k*^UTQVV@Iq$EE~kH7kt}m$4%0Y z;`-}cB!nWVf7u=KO6Y{n5Cz*;mO5tolBrP`O>j`f6r5aljvbQ}$gaXNOfl6SMu$*Ud7@xETEg1ErPMrFyxfXGdB#<(31S?Spo#5R_ z*lixYVl<;KI^sD3QH6fjV^$6#L1}qodeZbb$kJYJb3{UpFrBsujeH;}nC}Q8!8YnJ zd}v-uJfW{T=+nYmm>#*6M-4{^R6%$twX# zSGfg-M4*6?e8USUbXK+p(D`K%mKgg6#AA~rR6_Js|E0XQ9skcyJZ;D;fXP!URCpv0 z1%1;C+)i3h%^sv6MVH=+Gk5=U&aX!~qVGw#ZFp0;Wlvi8ITf#x2isi&8yvaQbE?62 zs@jY?pSf$fPVMd;Yh;;}{hpn6==WFzCSRUoaZCm1R6N!yqZ{*KZp;I(z07>BqiCj2 zMC1$BcJv*&Bg@!EAEd5Hk}0}MosOXD87^8W)W;cjEv;n-8mD}$+`4c?8qj#+JXFtpmFAr zN#;-;Gv@9!eyBSgTQ_FURCS7qnB=BQq>`P+*Y=%RI))Eg>YbX%RO2~jyx<3TR3(DI z*ODhm-Kg7om!{Gingt{HD{aqX3&x0U`r0o|ABb+fW$ZIlbSZaCA9V*+IG4G-Ui{Yk ziBYQk%BzOs7wO}$924wD=MR}q&&|0q3`T|ULW)2ybRqLxa`rqc2B0b$Q1ZgL1J-v9 zP#2xd)xu{0jZe_MAF2PI?0O~)(uNBhBODF+h{i;x+7H=*9dFFDcCH@-mzjD-=}0f` zDzn~6;$1G}#^ADklPI8NM>NXesJKFIq(!1p+7@uUm$8?0nTN`CyaJWxy2*14~tyo+0BQ`I$#||*&z)9d6_#EfwjH6 zPPx8b|K#Pmw-(tP$K)?o+eT&-XSGLNS5*`uzLo44c)28PMEM?ylO)JD)X09BHmQ-`2Z^8Fg`R@i)1fUy9|I-D>>{I zVhDeX9_|{uU~&jiMH7rlS>NV$y(jBDNAsiGi4Xuo^Z&YU0%+h;(W zsIPWa2*;b7X)wofZ&$_tyTIDm_%R{nE3ZudCJdTq^DRAqzJe5+>4Q#p)ja2$Qsq0- zA+q8rYlf0%3wDH$Nl``7-%Bxn{8fnil<)9cRTFjs$QhqYLhOwtq2c*gkYVKn38+x^ z`21oCPI%Txpua!iX&W0lf@YL>&rCX-Wsqc7WhV?!II5D_MA}R8H)r-(EoS-sn9-5| zYt2E@SC+$?uzJ1N(NfV>9nsAcrkG4#qXX@G)f3f3Y}`q+7_(waNxx=$hQo47_t9PO zq`4#S3NLcrfoDCWNbbf#b5rfK7kmHa^W587>5l>u(-7&TblhcM?U(@Zm}m8V%mPwg zj#!?>*_W{)0;UAfhWIj^gHu#*YyoZC!F3aS;H}CbvYUqaCGZ@Css;W0Gtrsrby=do z;-+nzcvTkH4#CXu#1ML?Xv%Ij)tqxS)#B&Dn|2{`Y*%??zl+0i0>wmt*>?9>-Lvg+ z1$*ik*l7?4I_}acs_~%BkjVz?V+pn97n^=Y?0n zEG1|2X;xV0t@xM74rwQHaC2x2V0T^m;~T5ajQ09LiN`6EvK z+rcI&De?-+fWkVe@Qx0HJnnnt+p*+TfG0JBHUIKykhH^)yY{72FBOskq^0Vq(4@Ia0=BM6+Y!-AJXc|6Cd&jQpghag)A@kQCj{h1R?#a&5FwG z==2$rPy5FL^?$T=|6liOx$CC^&z`Uqo^~U!u&~J9DH^ SHOW MODELS +IoTDB> show models +---------------------+--------------+--------------+-------------+ | ModelId| ModelType| Category| State| +---------------------+--------------+--------------+-------------+ @@ -432,7 +432,12 @@ IoTDB> SHOW MODELS | custom| | user_defined| active| | timer_xl| timer| builtin| activating| | sundial| sundial| builtin| active| +| sundialx_1| sundial| fine_tuned| active| +| sundialx_4| sundial| fine_tuned| training| +| sundialx_5| sundial| fine_tuned| failed| | chronos2| t5| builtin| inactive| +| moirai2| moirai| builtin| inactive| +| toto| toto| builtin| inactive| +---------------------+--------------+--------------+-------------+ ``` @@ -440,22 +445,24 @@ IoTDB> SHOW MODELS | Model Name | Core Concept | Applicable Scenarios | Key Features | |-----------------------------------------|------------------------------------------------------------------------------|-----------------------------------------------------------|------------------------------------------------------------------------------| -| **ARIMA** (Autoregressive Integrated Moving Average) | Combines AR, differencing (I), and MA for stationary or differenced series | Univariate forecasting (stock prices, sales, economics) | 1. For linear trends with weak seasonality2. Requires (p,d,q) tuning3. Sensitive to missing values | -| **Holt-Winters** (Triple Exponential Smoothing) | Exponential smoothing with level, trend, and seasonal components | Data with clear trend & seasonality (monthly sales, power demand) | 1. Handles additive/multiplicative seasonality2. Weights recent data higher3. Simple implementation | -| **Exponential Smoothing** | Weighted average of history with exponentially decaying weights | Trending but non-seasonal data (short-term demand) | 1. Few parameters, simple computation2. Suitable for stable/slow-changing series3. Extensible to double/triple smoothing | -| **Naive Forecaster** | Uses last observation as next prediction (simplest baseline) | Benchmarking or data with no clear pattern | 1. No training needed2. Sensitive to sudden changes3. Seasonal variant uses prior season value | -| **STL Forecaster** | Decomposes series into trend, seasonal, residual; forecasts components | Complex seasonality/trends (climate, traffic) | 1. Handles non-fixed seasonality2. Robust to outliers3. Components can use other models | -| **Gaussian HMM** | Hidden states generate observations; each state follows Gaussian distribution | State sequence prediction/classification (speech, finance) | 1. Models temporal state transitions2. Observations independent per state3. Requires state count | -| **GMM HMM** | Extends Gaussian HMM; each state uses Gaussian Mixture Model | Multi-modal observation scenarios (motion recognition, biosignals) | 1. More flexible than single Gaussian2. Higher complexity3. Requires GMM component count | -| **STRAY** (Search for Outliers using Random Projection and Adaptive Thresholding) | Uses SVD to detect anomalies in high-dimensional time series | High-dimensional anomaly detection (sensor networks, IT monitoring) | 1. No distribution assumption2. Handles high dimensions3. Sensitive to global anomalies | +| **ARIMA** (Autoregressive Integrated Moving Average) | Combines AR, differencing (I), and MA for stationary or differenced series | Univariate forecasting (stock prices, sales, economics) | 1. For linear trends with weak seasonality
2. Requires (p,d,q) tuning
3. Sensitive to missing values | +| **Holt-Winters** (Triple Exponential Smoothing) | Exponential smoothing with level, trend, and seasonal components | Data with clear trend & seasonality (monthly sales, power demand) | 1. Handles additive/multiplicative seasonality
2. Weights recent data higher
3. Simple implementation | +| **Exponential Smoothing** | Weighted average of history with exponentially decaying weights | Trending but non-seasonal data (short-term demand) | 1. Few parameters, simple computation
2. Suitable for stable/slow-changing series
3. Extensible to double/triple smoothing | +| **Naive Forecaster** | Uses last observation as next prediction (simplest baseline) | Benchmarking or data with no clear pattern | 1. No training needed
2. Sensitive to sudden changes
3. Seasonal variant uses prior season value | +| **STL Forecaster** | Decomposes series into trend, seasonal, residual; forecasts components | Complex seasonality/trends (climate, traffic) | 1. Handles non-fixed seasonality
2. Robust to outliers
3. Components can use other models | +| **Gaussian HMM** | Hidden states generate observations; each state follows Gaussian distribution | State sequence prediction/classification (speech, finance) | 1. Models temporal state transitions
2. Observations independent per state
3. Requires state count | +| **GMM HMM** | Extends Gaussian HMM; each state uses Gaussian Mixture Model | Multi-modal observation scenarios (motion recognition, biosignals) | 1. More flexible than single Gaussian
2. Higher complexity
3. Requires GMM component count | +| **STRAY** (Search for Outliers using Random Projection and Adaptive Thresholding) | Uses SVD to detect anomalies in high-dimensional time series | High-dimensional anomaly detection (sensor networks, IT monitoring) | 1. No distribution assumption
2. Handles high dimensions
3. Sensitive to global anomalies | **Built-in Time Series Large Models:** | Model Name | Core Concept | Applicable Scenarios | Key Features | |-----------------|------------------------------------------------------------------------------|-----------------------------------------------------------|------------------------------------------------------------------------------| -| **Timer-XL** | Long-context time series large model pretrained on massive industrial data | Complex industrial forecasting requiring ultra-long history (energy, aerospace, transport) | 1. Supports input of tens of thousands of time points2. Covers non-stationary, multivariate, and covariate scenarios3. Pretrained on trillion-scale high-quality industrial IoT data | -| **Timer-Sundial** | Generative foundation model with "Transformer + TimeFlow" architecture | Zero-shot forecasting requiring uncertainty quantification (finance, supply chain, renewable energy) | 1. Strong zero-shot generalization; supports point & probabilistic forecasting2. Flexible analysis of any prediction distribution statistic3. Innovative flow-matching architecture for efficient non-deterministic sample generation | -| **Chronos-2** | Universal time series foundation model based on discrete tokenization | Rapid zero-shot univariate forecasting; scenarios enhanced by covariates (promotions, weather) | 1. Powerful zero-shot probabilistic forecasting2. Unified multi-variable & covariate modeling (strict input requirements):  a. Future covariate names ⊆ historical covariate names  b. Each historical covariate length = target length  c. Each future covariate length = prediction length3. Efficient encoder-only structure balancing performance and speed | +| **Timer-XL** | Long-context time series large model pretrained on massive industrial data | Complex industrial forecasting requiring ultra-long history (energy, aerospace, transport) | 1. Supports input of tens of thousands of time points
2. Covers non-stationary, multivariate, and covariate scenarios
3. Pretrained on trillion-scale high-quality industrial IoT data | +| **Timer-Sundial** | Generative foundation model with "Transformer + TimeFlow" architecture | Zero-shot forecasting requiring uncertainty quantification (finance, supply chain, renewable energy) | 1. Strong zero-shot generalization; supports point & probabilistic forecasting
2. Flexible analysis of any prediction distribution statistic
3. Innovative flow-matching architecture for efficient non-deterministic sample generation | +| **Chronos-2** | Universal time series foundation model based on discrete tokenization | Rapid zero-shot univariate forecasting; scenarios enhanced by covariates (promotions, weather) | 1. Powerful zero-shot probabilistic forecasting
2. Unified multi-variable & covariate modeling (strict input requirements):
  a. Future covariate names ⊆ historical covariate names
  b. Each historical covariate length = target length
  c. Each future covariate length = prediction length
3. Efficient encoder-only structure balancing performance and speed | +| **Moirai 2.0** | Lightweight decoder-only Patch Transformer with a single patch size, multi-token prediction, and multi-quantile outputs | Zero-shot univariate forecasting where model size and inference efficiency are important, such as industrial monitoring, energy load, and equipment metrics | 1. Approximately 11.4M parameters
2. Predicts multiple patches per decoding step to reduce autoregressive overhead for long horizons
3. Outputs nine quantiles (0.1–0.9) and uses the p50 median as the point forecast
4. Uses instance normalization to mitigate distribution shift across series
5. Does not support multivariate targets or covariates | +| **Toto 2.0** | Decoder-only Patch Transformer alternating causal temporal attention and variable attention to jointly model temporal and variable dimensions | Zero-shot multivariate forecasting for observability metrics, including joint forecasting of CPU, memory, and network traffic | 1. Supports univariate and multivariate target forecasting
2. Outputs fixed quantiles from 0.1 to 0.9 and uses the p50 median as the point forecast
3. Supports cached block decoding for efficient scaling to longer forecast horizons
4. Covariates are not currently supported | ### 4.4 Deleting Models @@ -656,4 +663,4 @@ To add a new built-in custom model to AINode (using Chronos2 as example): ``` * **Complete Example** - Reference implementation: https://github.com/apache/iotdb/pull/16903 \ No newline at end of file + Reference implementation: https://github.com/apache/iotdb/pull/16903 diff --git a/src/UserGuide/Master/Table/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md b/src/UserGuide/Master/Table/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md index bc20138a4..284198ab6 100644 --- a/src/UserGuide/Master/Table/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md +++ b/src/UserGuide/Master/Table/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md @@ -46,7 +46,7 @@ The Timer model (non-built-in model) not only demonstrates excellent few-shot ge ## 4. Timer-XL Model -Timer-XL further extends and upgrades the network structure based on Timer, achieving comprehensive breakthroughs in multiple dimensions: +Timer-XL[2] further extends and upgrades the network structure based on Timer, achieving comprehensive breakthroughs in multiple dimensions: * **Ultra-Long Context Support**: This model breaks through the limitations of traditional time series forecasting models, supporting the processing of inputs with thousands of Tokens (equivalent to tens of thousands of time points), effectively solving the context length bottleneck problem. * **Coverage of Multi-Variable Forecasting Scenarios**: Supports various forecasting scenarios, including the prediction of non-stationary time series, multi-variable prediction tasks, and predictions involving covariates, meeting diversified business needs. @@ -56,7 +56,7 @@ Timer-XL further extends and upgrades the network structure based on Timer, achi ## 5. Timer-Sundial Model -Timer-Sundial is a series of generative foundational models focused on time series forecasting. The base version has 128 million parameters and has been pre-trained on 1 trillion time points, with the following core characteristics: +Timer-Sundial[3] is a series of generative foundational models focused on time series forecasting. The base version has 128 million parameters and has been pre-trained on 1 trillion time points, with the following core characteristics: * **Strong Generalization Performance**: Possesses zero-shot forecasting capabilities and can support both point forecasting and probabilistic forecasting simultaneously. * **Flexible Prediction Distribution Analysis**: Not only can it predict means or quantiles, but it can also evaluate any statistical properties of the prediction distribution through the raw samples generated by the model. @@ -66,7 +66,7 @@ Timer-Sundial is a series of generative foundational models focused on time seri ## 6. Chronos-2 Model -Chronos-2 is a universal time series foundational model developed by the Amazon Web Services (AWS) research team, evolved from the Chronos discrete token modeling paradigm. This model is suitable for both zero-shot univariate forecasting and covariate forecasting. Its main characteristics include: +Chronos-2[4] is a universal time series foundational model developed by the Amazon Web Services (AWS) research team, evolved from the Chronos discrete token modeling paradigm. This model is suitable for both zero-shot univariate forecasting and covariate forecasting. Its main characteristics include: * **Probabilistic Forecasting Capability**: The model outputs multi-step prediction results in a generative manner, supporting quantile or distribution-level forecasting to characterize future uncertainty. * **Zero-Shot General Forecasting**: Leveraging the contextual learning ability acquired through pre-training, it can directly execute forecasting on unseen datasets without retraining or parameter updates. @@ -78,7 +78,34 @@ Chronos-2 is a universal time series foundational model developed by the Amazon ![](/img/timeseries-large-model-chronos2.png) -## 7. Performance Showcase +## 7. Moirai2 Model + +Moirai2[5] (Moirai 2.0) is a general-purpose time series foundation model developed by Salesforce AI Research (supported in V2.0.10 and later). AINode currently integrates the Moirai 2.0 R-small variant, which has approximately 11.4M parameters. Unlike Moirai 1.0, which uses a masked encoder architecture, Moirai 2.0 uses a causal decoder-only Patch Transformer. With a single patch size, multi-token prediction, and multi-quantile outputs, it provides efficient univariate forecasting with a compact model. Its core features include: + +- **Lightweight Model Architecture**: Uses a decoder-only Patch Transformer with RMSNorm, rotary positional embeddings, and SiLU-GLU feed-forward networks to balance forecasting capability and inference efficiency at a small parameter scale. +- **Multi-Token Prediction**: Predicts multiple patches at each decoding step, reducing the number of autoregressive decoding steps required for long forecast horizons. +- **Probabilistic Forecasting**: Outputs nine quantiles from 0.1 to 0.9. AINode uses the p50 median as the point forecast. +- **Patch Decoding**: Groups the time series into fixed-size patches before the attention module to improve temporal feature extraction and decoding efficiency. +- **Instance Normalization**: Standardizes each time series before model input and applies denormalization after output to mitigate distribution shifts across series. +- **Input Scope**: Focuses on univariate forecasting and does not support multivariate targets or covariates. + +![](/img/LargeModel-moirai2.png) + +> Note: The Moirai 2.0 R-small model weights are licensed under CC BY-NC 4.0 and are restricted to research use. + +## 8. Toto Model + +Toto[6] (Toto 2.0) is a next-generation time series foundation model developed by Datadog (supported in V2.0.10 and later), primarily for forecasting in observability scenarios. AINode currently integrates the 2.5B-parameter variant. It is based on a decoder-only Patch Transformer architecture that alternates causal temporal attention and variable attention to jointly model the temporal and variable dimensions. Its core features include: + +- **Univariate and Multivariate Forecasting**: Supports both individual target variables and joint forecasting of multiple related target variables, making it suitable for observability metrics such as CPU, memory, and network traffic. +- **Probabilistic Forecasting**: Outputs fixed quantiles from 0.1 to 0.9 to represent forecasting uncertainty. AINode uses the p50 median as the point forecast. +- **Efficient Block Decoding**: Uses cached block decoding to generate forecasts in blocks, reducing repeated computation for longer forecast horizons. +- **Large-Scale Model Capability**: Improves forecasting performance by scaling model parameters and pre-training data, providing strong zero-shot generalization. +- **Covariate Limitation**: The currently integrated version does not support historical covariates or known future covariates. + +![](/img/LargeModel-toto.png) + +## 9. Performance Showcase Time Series Large Models can adapt to real time series data from various different domains and scenarios, demonstrating excellent processing capabilities across various tasks. The following shows the actual performance on different datasets: @@ -100,7 +127,7 @@ Using Time Series Large Models to accurately identify outliers that deviate sign ![](/img/LargeModel05.png) -## 8. Deployment and Usage +## 10. Deployment and Usage 1. Open the IoTDB CLI console and check that the ConfigNode, DataNode, and AINode nodes are all Running. @@ -143,6 +170,8 @@ IoTDB> show models | timer_xl| timer| builtin| active| | sundial| sundial| builtin| active| | chronos2| t5| builtin| active| +| moirai2| moirai| builtin| active| +| toto| toto| builtin| active| +---------------------+---------+--------+--------+ ``` @@ -150,8 +179,12 @@ IoTDB> show models **[1]** Timer: Generative Pre-trained Transformers Are Large Time Series Models, Yong Liu, Haoran Zhang, Chenyu Li, Xiangdong Huang, Jianmin Wang, Mingsheng Long. [↩ Back]() -**[2]** TIMER-XL: LONG-CONTEXT TRANSFORMERS FOR UNIFIED TIME SERIES FORECASTING, Yong Liu, Guo Qin, Xiangdong Huang, Jianmin Wang, Mingsheng Long. [↩ Back]() +**[2]** TIMER-XL: LONG-CONTEXT TRANSFORMERS FOR UNIFIED TIME SERIES FORECASTING, Yong Liu, Guo Qin, Xiangdong Huang, Jianmin Wang, Mingsheng Long. [↩ Back](#ref2) + +**[3]** Sundial: A Family of Highly Capable Time Series Foundation Models, Yong Liu, Guo Qin, Zhiyuan Shi, Zhi Chen, Caiyin Yang, Xiangdong Huang, Jianmin Wang, Mingsheng Long, **ICML 2025 spotlight**. [↩ Back](#ref3) + +**[4]** Chronos-2: From Univariate to Universal Forecasting, Abdul Fatir Ansari, Oleksandr Shchur, Jaris Küken, Andreas Auer, Boran Han, Pedro Mercado, Syama Sundar Rangapuram, Huibin Shen, Lorenzo Stella, Xiyuan Zhang, Mononito Goswami, Shubham Kapoor, Danielle C. Maddix, Pablo Guerron, Tony Hu, Junming Yin, Nick Erickson, Prateek Mutalik Desai, Hao Wang, Huzefa Rangwala, George Karypis, Yuyang Wang, Michael Bohlke-Schneider, **arXiv:2510.15821**. [↩ Back](#ref4) -**[3]** Sundial: A Family of Highly Capable Time Series Foundation Models, Yong Liu, Guo Qin, Zhiyuan Shi, Zhi Chen, Caiyin Yang, Xiangdong Huang, Jianmin Wang, Mingsheng Long, **ICML 2025 spotlight**. [↩ Back]() +[5] Moirai 2.0: When Less Is More for Time Series Forecasting, Salesforce AI Research, arXiv:2511.11698. [↩ Back](#ref5) -**[4]** Chronos-2: From Univariate to Universal Forecasting, Abdul Fatir Ansari, Oleksandr Shchur, Jaris Küken, Andreas Auer, Boran Han, Pedro Mercado, Syama Sundar Rangapuram, Huibin Shen, Lorenzo Stella, Xiyuan Zhang, Mononito Goswami, Shubham Kapoor, Danielle C. Maddix, Pablo Guerron, Tony Hu, Junming Yin, Nick Erickson, Prateek Mutalik Desai, Hao Wang, Huzefa Rangwala, George Karypis, Yuyang Wang, Michael Bohlke-Schneider, **arXiv:2510.15821**. [↩ Back]() \ No newline at end of file +[6] Toto 2.0: Time Series Forecasting Enters the Scaling Era, Datadog, arXiv:2605.20119. [↩ Back](#ref6) diff --git a/src/UserGuide/Master/Tree/AI-capability/AINode_Upgrade_apache.md b/src/UserGuide/Master/Tree/AI-capability/AINode_Upgrade_apache.md index 259417fbc..0f614b413 100644 --- a/src/UserGuide/Master/Tree/AI-capability/AINode_Upgrade_apache.md +++ b/src/UserGuide/Master/Tree/AI-capability/AINode_Upgrade_apache.md @@ -416,7 +416,12 @@ IoTDB> show models | custom| | user_defined| active| | timer_xl| timer| builtin| activating| | sundial| sundial| builtin| active| +| sundialx_1| sundial| fine_tuned| active| +| sundialx_4| sundial| fine_tuned| training| +| sundialx_5| sundial| fine_tuned| failed| | chronos2| t5| builtin| inactive| +| moirai2| moirai| builtin| inactive| +| toto| toto| builtin| inactive| +---------------------+--------------+--------------+-------------+ ``` @@ -424,22 +429,24 @@ Built-in traditional time series model introduction: | Model Name | Core Concept | Applicable Scenario | Main Features | |----------------------------------| ----------------------------------------------------------------------------------------- | ------------------------------------------------------------------------ | ------------------------------------------------------------------------------------------------------------------------------------------- | -| **ARIMA** (AutoRegressive Integrated Moving Average) | Combines autoregression (AR), differencing (I), and moving average (MA), used for predicting stationary time series or data that can be made stationary through differencing. | Univariate time series prediction, such as stock prices, sales, economic indicators. | 1. Suitable for linear trends and weak seasonality data. 2. Requires selecting parameters (p,d,q). 3. Sensitive to missing values. | -| **Holt-Winters** (Three-parameter exponential smoothing) | Based on exponential smoothing, introduces level, trend, and seasonal components, suitable for data with trend and seasonality. | Time series with clear seasonality and trend, such as monthly sales, power demand. | 1. Can handle additive or multiplicative seasonality. 2. Gives higher weight to recent data. 3. Simple to implement. | -| **Exponential Smoothing** (Exponential smoothing) | Uses weighted average of historical data, with weights decreasing exponentially over time, emphasizing the importance of recent observations. | Data without significant seasonality but with trend, such as short-term demand prediction. | 1. Few parameters, simple calculation. 2. Suitable for stationary or slowly changing sequences. 3. Can be extended to double or triple exponential smoothing. | -| **Naive Forecaster** (Naive predictor) | Uses the observation of the most recent period as the prediction for the next period, the simplest baseline model. | As a benchmark for other models or simple prediction when data has no obvious pattern. | 1. No training required. 2. Sensitive to sudden changes. 3. Seasonal naive variant can use the same period of the previous season to predict. | -| **STL Forecaster** (Seasonal-Trend Decomposition Forecast) | Based on STL decomposition of time series, predicts trend, seasonal, and residual components separately, then combines them. | Data with complex seasonality, trend, and non-linear patterns, such as climate data, traffic flow. | 1. Can handle non-fixed seasonality. 2. Robust to outliers. 3. After decomposition, other models can be combined to predict each component. | -| **Gaussian HMM** (Gaussian Hidden Markov Model) | Assumes observed data is generated by hidden states, with each state's observation probability following a Gaussian distribution. | State sequence prediction or classification, such as speech recognition, financial state identification. | 1. Suitable for time series state modeling. 2. Assumes observations are independent given the state. 3. Requires specifying the number of hidden states. | -| **GMM HMM** (Gaussian Mixture Hidden Markov Model) | An extension of Gaussian HMM, where each state's observation probability is described by a Gaussian Mixture Model, capturing more complex observation distributions. | Scenarios requiring multi-modal observation distributions, such as complex action recognition, biosignal analysis. | 1. More flexible than single Gaussian. 2. More parameters, higher computational complexity. 3. Requires training the number of GMM components. | -| **STRAY** (Anomaly Detection based on Singular Value Decomposition) | Detects anomalies in high-dimensional data through Singular Value Decomposition (SVD), commonly used for time series anomaly detection. | High-dimensional time series anomaly detection, such as sensor networks, IT system monitoring. | 1. No distribution assumption required. 2. Can handle high-dimensional data. 3. Sensitive to global anomalies, may miss local anomalies. | +| **ARIMA** (AutoRegressive Integrated Moving Average) | Combines autoregression (AR), differencing (I), and moving average (MA), used for predicting stationary time series or data that can be made stationary through differencing. | Univariate time series prediction, such as stock prices, sales, economic indicators. | 1. Suitable for linear trends and weak seasonality data.
2. Requires selecting parameters (p,d,q).
3. Sensitive to missing values. | +| **Holt-Winters** (Three-parameter exponential smoothing) | Based on exponential smoothing, introduces level, trend, and seasonal components, suitable for data with trend and seasonality. | Time series with clear seasonality and trend, such as monthly sales, power demand. | 1. Can handle additive or multiplicative seasonality.
2. Gives higher weight to recent data.
3. Simple to implement. | +| **Exponential Smoothing** (Exponential smoothing) | Uses weighted average of historical data, with weights decreasing exponentially over time, emphasizing the importance of recent observations. | Data without significant seasonality but with trend, such as short-term demand prediction. | 1. Few parameters, simple calculation.
2. Suitable for stationary or slowly changing sequences.
3. Can be extended to double or triple exponential smoothing. | +| **Naive Forecaster** (Naive predictor) | Uses the observation of the most recent period as the prediction for the next period, the simplest baseline model. | As a benchmark for other models or simple prediction when data has no obvious pattern. | 1. No training required.
2. Sensitive to sudden changes.
3. Seasonal naive variant can use the same period of the previous season to predict. | +| **STL Forecaster** (Seasonal-Trend Decomposition Forecast) | Based on STL decomposition of time series, predicts trend, seasonal, and residual components separately, then combines them. | Data with complex seasonality, trend, and non-linear patterns, such as climate data, traffic flow. | 1. Can handle non-fixed seasonality.
2. Robust to outliers.
3. After decomposition, other models can be combined to predict each component. | +| **Gaussian HMM** (Gaussian Hidden Markov Model) | Assumes observed data is generated by hidden states, with each state's observation probability following a Gaussian distribution. | State sequence prediction or classification, such as speech recognition, financial state identification. | 1. Suitable for time series state modeling.
2. Assumes observations are independent given the state.
3. Requires specifying the number of hidden states. | +| **GMM HMM** (Gaussian Mixture Hidden Markov Model) | An extension of Gaussian HMM, where each state's observation probability is described by a Gaussian Mixture Model, capturing more complex observation distributions. | Scenarios requiring multi-modal observation distributions, such as complex action recognition, biosignal analysis. | 1. More flexible than single Gaussian.
2. More parameters, higher computational complexity.
3. Requires training the number of GMM components. | +| **STRAY** (Anomaly Detection based on Singular Value Decomposition) | Detects anomalies in high-dimensional data through Singular Value Decomposition (SVD), commonly used for time series anomaly detection. | High-dimensional time series anomaly detection, such as sensor networks, IT system monitoring. | 1. No distribution assumption required.
2. Can handle high-dimensional data.
3. Sensitive to global anomalies, may miss local anomalies. | Built-in time series large model introduction: | Model Name | Core Concept | Applicable Scenario | Main Features | |---------------| ---------------------------------------------------------------------- | ------------------------------------------------------------------------ | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | -| **Timer-XL** | Time series large model supporting ultra-long context, enhancing generalization capability through large-scale industrial data pre-training. | Complex industrial prediction requiring extremely long historical data, such as energy, aerospace, and transportation. | 1. Ultra-long context support, can handle tens of thousands of time points as input. 2. Multi-scenario coverage, supports non-stationary, multi-variable, and covariate prediction. 3. Pre-trained on trillions of high-quality industrial time series data. | -| **Timer-Sundial** | A generative foundational model based on "Transformer + TimeFlow" architecture, focusing on probabilistic prediction. | Zero-shot prediction scenarios requiring quantification of uncertainty, such as finance, supply chain, and new energy power generation. | 1. Strong zero-shot generalization capability, supports point prediction and probabilistic prediction. 2. Can flexibly analyze any statistical properties of the prediction distribution. 3. Innovative generative architecture, achieving efficient non-deterministic sample generation. | -| **Chronos-2** | A general time series foundational model based on discrete tokenization paradigm, converting prediction into language modeling tasks. | Rapid zero-shot univariate prediction, and scenarios that can leverage covariates (e.g., promotions, weather) to improve results. | 1. Strong zero-shot probabilistic prediction capability. 2. Supports unified covariate modeling, but has strict input requirements: a. The set of names of future covariates must be a subset of the set of names of historical covariates; b. The length of each historical covariate must equal the length of the target variable; c. The length of each future covariate must equal the prediction length; 3. Uses an efficient encoder-style structure, balancing performance and inference speed. | +| **Timer-XL** | Time series large model supporting ultra-long context, enhancing generalization capability through large-scale industrial data pre-training. | Complex industrial prediction requiring extremely long historical data, such as energy, aerospace, and transportation. | 1. Ultra-long context support, can handle tens of thousands of time points as input.
2. Multi-scenario coverage, supports non-stationary, multi-variable, and covariate prediction.
3. Pre-trained on trillions of high-quality industrial time series data. | +| **Timer-Sundial** | A generative foundational model based on "Transformer + TimeFlow" architecture, focusing on probabilistic prediction. | Zero-shot prediction scenarios requiring quantification of uncertainty, such as finance, supply chain, and new energy power generation. | 1. Strong zero-shot generalization capability, supports point prediction and probabilistic prediction.
2. Can flexibly analyze any statistical properties of the prediction distribution.
3. Innovative generative architecture, achieving efficient non-deterministic sample generation. | +| **Chronos-2** | A general time series foundational model based on discrete tokenization paradigm, converting prediction into language modeling tasks. | Rapid zero-shot univariate prediction, and scenarios that can leverage covariates (e.g., promotions, weather) to improve results. | 1. Strong zero-shot probabilistic prediction capability.
2. Supports unified covariate modeling, but has strict input requirements:
  a. The set of names of future covariates must be a subset of the set of names of historical covariates;
  b. The length of each historical covariate must equal the length of the target variable;
  c. The length of each future covariate must equal the prediction length;
3. Uses an efficient encoder-style structure, balancing performance and inference speed. | +| **Moirai 2.0** | Uses a lightweight decoder-only Patch Transformer with a single patch size, multi-token prediction, and multi-quantile outputs for efficient univariate forecasting. | Zero-shot univariate forecasting where model size and inference efficiency are important, such as industrial monitoring, energy load, and equipment metric forecasting. | 1. Approximately 11.4M parameters.
2. Predicts multiple patches per decoding step to reduce autoregressive overhead for long horizons.
3. Outputs nine quantiles (0.1–0.9) and uses the p50 median as the point forecast.
4. Uses instance normalization to mitigate distribution shift across series.
5. Does not support multivariate targets or covariates. | +| **Toto 2.0** | Uses a decoder-only Patch Transformer that alternates causal temporal attention and variable attention to jointly model the temporal and variable dimensions. | Zero-shot forecasting for multivariate time series such as observability metrics, including joint forecasting of CPU, memory, and network traffic. | 1. Supports univariate and multivariate target forecasting.
2. Outputs fixed quantiles from 0.1 to 0.9 and uses the p50 median as the point forecast.
3. Supports cached block decoding for efficient scaling to longer forecast horizons.
4. Covariates are not currently supported. | ### 4.4 Delete Models @@ -675,4 +682,4 @@ Support adding new built-in models to AINode. The specific steps are as follows * **Full Example** -Full example can be referenced at https://github.com/apache/iotdb/pull/16903 \ No newline at end of file +Full example can be referenced at https://github.com/apache/iotdb/pull/16903 diff --git a/src/UserGuide/Master/Tree/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md b/src/UserGuide/Master/Tree/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md index bc20138a4..284198ab6 100644 --- a/src/UserGuide/Master/Tree/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md +++ b/src/UserGuide/Master/Tree/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md @@ -46,7 +46,7 @@ The Timer model (non-built-in model) not only demonstrates excellent few-shot ge ## 4. Timer-XL Model -Timer-XL further extends and upgrades the network structure based on Timer, achieving comprehensive breakthroughs in multiple dimensions: +Timer-XL[2] further extends and upgrades the network structure based on Timer, achieving comprehensive breakthroughs in multiple dimensions: * **Ultra-Long Context Support**: This model breaks through the limitations of traditional time series forecasting models, supporting the processing of inputs with thousands of Tokens (equivalent to tens of thousands of time points), effectively solving the context length bottleneck problem. * **Coverage of Multi-Variable Forecasting Scenarios**: Supports various forecasting scenarios, including the prediction of non-stationary time series, multi-variable prediction tasks, and predictions involving covariates, meeting diversified business needs. @@ -56,7 +56,7 @@ Timer-XL further extends and upgrades the network structure based on Timer, achi ## 5. Timer-Sundial Model -Timer-Sundial is a series of generative foundational models focused on time series forecasting. The base version has 128 million parameters and has been pre-trained on 1 trillion time points, with the following core characteristics: +Timer-Sundial[3] is a series of generative foundational models focused on time series forecasting. The base version has 128 million parameters and has been pre-trained on 1 trillion time points, with the following core characteristics: * **Strong Generalization Performance**: Possesses zero-shot forecasting capabilities and can support both point forecasting and probabilistic forecasting simultaneously. * **Flexible Prediction Distribution Analysis**: Not only can it predict means or quantiles, but it can also evaluate any statistical properties of the prediction distribution through the raw samples generated by the model. @@ -66,7 +66,7 @@ Timer-Sundial is a series of generative foundational models focused on time seri ## 6. Chronos-2 Model -Chronos-2 is a universal time series foundational model developed by the Amazon Web Services (AWS) research team, evolved from the Chronos discrete token modeling paradigm. This model is suitable for both zero-shot univariate forecasting and covariate forecasting. Its main characteristics include: +Chronos-2[4] is a universal time series foundational model developed by the Amazon Web Services (AWS) research team, evolved from the Chronos discrete token modeling paradigm. This model is suitable for both zero-shot univariate forecasting and covariate forecasting. Its main characteristics include: * **Probabilistic Forecasting Capability**: The model outputs multi-step prediction results in a generative manner, supporting quantile or distribution-level forecasting to characterize future uncertainty. * **Zero-Shot General Forecasting**: Leveraging the contextual learning ability acquired through pre-training, it can directly execute forecasting on unseen datasets without retraining or parameter updates. @@ -78,7 +78,34 @@ Chronos-2 is a universal time series foundational model developed by the Amazon ![](/img/timeseries-large-model-chronos2.png) -## 7. Performance Showcase +## 7. Moirai2 Model + +Moirai2[5] (Moirai 2.0) is a general-purpose time series foundation model developed by Salesforce AI Research (supported in V2.0.10 and later). AINode currently integrates the Moirai 2.0 R-small variant, which has approximately 11.4M parameters. Unlike Moirai 1.0, which uses a masked encoder architecture, Moirai 2.0 uses a causal decoder-only Patch Transformer. With a single patch size, multi-token prediction, and multi-quantile outputs, it provides efficient univariate forecasting with a compact model. Its core features include: + +- **Lightweight Model Architecture**: Uses a decoder-only Patch Transformer with RMSNorm, rotary positional embeddings, and SiLU-GLU feed-forward networks to balance forecasting capability and inference efficiency at a small parameter scale. +- **Multi-Token Prediction**: Predicts multiple patches at each decoding step, reducing the number of autoregressive decoding steps required for long forecast horizons. +- **Probabilistic Forecasting**: Outputs nine quantiles from 0.1 to 0.9. AINode uses the p50 median as the point forecast. +- **Patch Decoding**: Groups the time series into fixed-size patches before the attention module to improve temporal feature extraction and decoding efficiency. +- **Instance Normalization**: Standardizes each time series before model input and applies denormalization after output to mitigate distribution shifts across series. +- **Input Scope**: Focuses on univariate forecasting and does not support multivariate targets or covariates. + +![](/img/LargeModel-moirai2.png) + +> Note: The Moirai 2.0 R-small model weights are licensed under CC BY-NC 4.0 and are restricted to research use. + +## 8. Toto Model + +Toto[6] (Toto 2.0) is a next-generation time series foundation model developed by Datadog (supported in V2.0.10 and later), primarily for forecasting in observability scenarios. AINode currently integrates the 2.5B-parameter variant. It is based on a decoder-only Patch Transformer architecture that alternates causal temporal attention and variable attention to jointly model the temporal and variable dimensions. Its core features include: + +- **Univariate and Multivariate Forecasting**: Supports both individual target variables and joint forecasting of multiple related target variables, making it suitable for observability metrics such as CPU, memory, and network traffic. +- **Probabilistic Forecasting**: Outputs fixed quantiles from 0.1 to 0.9 to represent forecasting uncertainty. AINode uses the p50 median as the point forecast. +- **Efficient Block Decoding**: Uses cached block decoding to generate forecasts in blocks, reducing repeated computation for longer forecast horizons. +- **Large-Scale Model Capability**: Improves forecasting performance by scaling model parameters and pre-training data, providing strong zero-shot generalization. +- **Covariate Limitation**: The currently integrated version does not support historical covariates or known future covariates. + +![](/img/LargeModel-toto.png) + +## 9. Performance Showcase Time Series Large Models can adapt to real time series data from various different domains and scenarios, demonstrating excellent processing capabilities across various tasks. The following shows the actual performance on different datasets: @@ -100,7 +127,7 @@ Using Time Series Large Models to accurately identify outliers that deviate sign ![](/img/LargeModel05.png) -## 8. Deployment and Usage +## 10. Deployment and Usage 1. Open the IoTDB CLI console and check that the ConfigNode, DataNode, and AINode nodes are all Running. @@ -143,6 +170,8 @@ IoTDB> show models | timer_xl| timer| builtin| active| | sundial| sundial| builtin| active| | chronos2| t5| builtin| active| +| moirai2| moirai| builtin| active| +| toto| toto| builtin| active| +---------------------+---------+--------+--------+ ``` @@ -150,8 +179,12 @@ IoTDB> show models **[1]** Timer: Generative Pre-trained Transformers Are Large Time Series Models, Yong Liu, Haoran Zhang, Chenyu Li, Xiangdong Huang, Jianmin Wang, Mingsheng Long. [↩ Back]() -**[2]** TIMER-XL: LONG-CONTEXT TRANSFORMERS FOR UNIFIED TIME SERIES FORECASTING, Yong Liu, Guo Qin, Xiangdong Huang, Jianmin Wang, Mingsheng Long. [↩ Back]() +**[2]** TIMER-XL: LONG-CONTEXT TRANSFORMERS FOR UNIFIED TIME SERIES FORECASTING, Yong Liu, Guo Qin, Xiangdong Huang, Jianmin Wang, Mingsheng Long. [↩ Back](#ref2) + +**[3]** Sundial: A Family of Highly Capable Time Series Foundation Models, Yong Liu, Guo Qin, Zhiyuan Shi, Zhi Chen, Caiyin Yang, Xiangdong Huang, Jianmin Wang, Mingsheng Long, **ICML 2025 spotlight**. [↩ Back](#ref3) + +**[4]** Chronos-2: From Univariate to Universal Forecasting, Abdul Fatir Ansari, Oleksandr Shchur, Jaris Küken, Andreas Auer, Boran Han, Pedro Mercado, Syama Sundar Rangapuram, Huibin Shen, Lorenzo Stella, Xiyuan Zhang, Mononito Goswami, Shubham Kapoor, Danielle C. Maddix, Pablo Guerron, Tony Hu, Junming Yin, Nick Erickson, Prateek Mutalik Desai, Hao Wang, Huzefa Rangwala, George Karypis, Yuyang Wang, Michael Bohlke-Schneider, **arXiv:2510.15821**. [↩ Back](#ref4) -**[3]** Sundial: A Family of Highly Capable Time Series Foundation Models, Yong Liu, Guo Qin, Zhiyuan Shi, Zhi Chen, Caiyin Yang, Xiangdong Huang, Jianmin Wang, Mingsheng Long, **ICML 2025 spotlight**. [↩ Back]() +[5] Moirai 2.0: When Less Is More for Time Series Forecasting, Salesforce AI Research, arXiv:2511.11698. [↩ Back](#ref5) -**[4]** Chronos-2: From Univariate to Universal Forecasting, Abdul Fatir Ansari, Oleksandr Shchur, Jaris Küken, Andreas Auer, Boran Han, Pedro Mercado, Syama Sundar Rangapuram, Huibin Shen, Lorenzo Stella, Xiyuan Zhang, Mononito Goswami, Shubham Kapoor, Danielle C. Maddix, Pablo Guerron, Tony Hu, Junming Yin, Nick Erickson, Prateek Mutalik Desai, Hao Wang, Huzefa Rangwala, George Karypis, Yuyang Wang, Michael Bohlke-Schneider, **arXiv:2510.15821**. [↩ Back]() \ No newline at end of file +[6] Toto 2.0: Time Series Forecasting Enters the Scaling Era, Datadog, arXiv:2605.20119. [↩ Back](#ref6) diff --git a/src/UserGuide/latest-Table/AI-capability/AINode_Upgrade_apache.md b/src/UserGuide/latest-Table/AI-capability/AINode_Upgrade_apache.md index 907aa13b6..fc7d36253 100644 --- a/src/UserGuide/latest-Table/AI-capability/AINode_Upgrade_apache.md +++ b/src/UserGuide/latest-Table/AI-capability/AINode_Upgrade_apache.md @@ -417,7 +417,7 @@ State Flow Explanation: **Viewing Example:** ```SQL -IoTDB> SHOW MODELS +IoTDB> show models +---------------------+--------------+--------------+-------------+ | ModelId| ModelType| Category| State| +---------------------+--------------+--------------+-------------+ @@ -432,7 +432,12 @@ IoTDB> SHOW MODELS | custom| | user_defined| active| | timer_xl| timer| builtin| activating| | sundial| sundial| builtin| active| +| sundialx_1| sundial| fine_tuned| active| +| sundialx_4| sundial| fine_tuned| training| +| sundialx_5| sundial| fine_tuned| failed| | chronos2| t5| builtin| inactive| +| moirai2| moirai| builtin| inactive| +| toto| toto| builtin| inactive| +---------------------+--------------+--------------+-------------+ ``` @@ -440,22 +445,24 @@ IoTDB> SHOW MODELS | Model Name | Core Concept | Applicable Scenarios | Key Features | |-----------------------------------------|------------------------------------------------------------------------------|-----------------------------------------------------------|------------------------------------------------------------------------------| -| **ARIMA** (Autoregressive Integrated Moving Average) | Combines AR, differencing (I), and MA for stationary or differenced series | Univariate forecasting (stock prices, sales, economics) | 1. For linear trends with weak seasonality2. Requires (p,d,q) tuning3. Sensitive to missing values | -| **Holt-Winters** (Triple Exponential Smoothing) | Exponential smoothing with level, trend, and seasonal components | Data with clear trend & seasonality (monthly sales, power demand) | 1. Handles additive/multiplicative seasonality2. Weights recent data higher3. Simple implementation | -| **Exponential Smoothing** | Weighted average of history with exponentially decaying weights | Trending but non-seasonal data (short-term demand) | 1. Few parameters, simple computation2. Suitable for stable/slow-changing series3. Extensible to double/triple smoothing | -| **Naive Forecaster** | Uses last observation as next prediction (simplest baseline) | Benchmarking or data with no clear pattern | 1. No training needed2. Sensitive to sudden changes3. Seasonal variant uses prior season value | -| **STL Forecaster** | Decomposes series into trend, seasonal, residual; forecasts components | Complex seasonality/trends (climate, traffic) | 1. Handles non-fixed seasonality2. Robust to outliers3. Components can use other models | -| **Gaussian HMM** | Hidden states generate observations; each state follows Gaussian distribution | State sequence prediction/classification (speech, finance) | 1. Models temporal state transitions2. Observations independent per state3. Requires state count | -| **GMM HMM** | Extends Gaussian HMM; each state uses Gaussian Mixture Model | Multi-modal observation scenarios (motion recognition, biosignals) | 1. More flexible than single Gaussian2. Higher complexity3. Requires GMM component count | -| **STRAY** (Search for Outliers using Random Projection and Adaptive Thresholding) | Uses SVD to detect anomalies in high-dimensional time series | High-dimensional anomaly detection (sensor networks, IT monitoring) | 1. No distribution assumption2. Handles high dimensions3. Sensitive to global anomalies | +| **ARIMA** (Autoregressive Integrated Moving Average) | Combines AR, differencing (I), and MA for stationary or differenced series | Univariate forecasting (stock prices, sales, economics) | 1. For linear trends with weak seasonality
2. Requires (p,d,q) tuning
3. Sensitive to missing values | +| **Holt-Winters** (Triple Exponential Smoothing) | Exponential smoothing with level, trend, and seasonal components | Data with clear trend & seasonality (monthly sales, power demand) | 1. Handles additive/multiplicative seasonality
2. Weights recent data higher
3. Simple implementation | +| **Exponential Smoothing** | Weighted average of history with exponentially decaying weights | Trending but non-seasonal data (short-term demand) | 1. Few parameters, simple computation
2. Suitable for stable/slow-changing series
3. Extensible to double/triple smoothing | +| **Naive Forecaster** | Uses last observation as next prediction (simplest baseline) | Benchmarking or data with no clear pattern | 1. No training needed
2. Sensitive to sudden changes
3. Seasonal variant uses prior season value | +| **STL Forecaster** | Decomposes series into trend, seasonal, residual; forecasts components | Complex seasonality/trends (climate, traffic) | 1. Handles non-fixed seasonality
2. Robust to outliers
3. Components can use other models | +| **Gaussian HMM** | Hidden states generate observations; each state follows Gaussian distribution | State sequence prediction/classification (speech, finance) | 1. Models temporal state transitions
2. Observations independent per state
3. Requires state count | +| **GMM HMM** | Extends Gaussian HMM; each state uses Gaussian Mixture Model | Multi-modal observation scenarios (motion recognition, biosignals) | 1. More flexible than single Gaussian
2. Higher complexity
3. Requires GMM component count | +| **STRAY** (Search for Outliers using Random Projection and Adaptive Thresholding) | Uses SVD to detect anomalies in high-dimensional time series | High-dimensional anomaly detection (sensor networks, IT monitoring) | 1. No distribution assumption
2. Handles high dimensions
3. Sensitive to global anomalies | **Built-in Time Series Large Models:** | Model Name | Core Concept | Applicable Scenarios | Key Features | |-----------------|------------------------------------------------------------------------------|-----------------------------------------------------------|------------------------------------------------------------------------------| -| **Timer-XL** | Long-context time series large model pretrained on massive industrial data | Complex industrial forecasting requiring ultra-long history (energy, aerospace, transport) | 1. Supports input of tens of thousands of time points2. Covers non-stationary, multivariate, and covariate scenarios3. Pretrained on trillion-scale high-quality industrial IoT data | -| **Timer-Sundial** | Generative foundation model with "Transformer + TimeFlow" architecture | Zero-shot forecasting requiring uncertainty quantification (finance, supply chain, renewable energy) | 1. Strong zero-shot generalization; supports point & probabilistic forecasting2. Flexible analysis of any prediction distribution statistic3. Innovative flow-matching architecture for efficient non-deterministic sample generation | -| **Chronos-2** | Universal time series foundation model based on discrete tokenization | Rapid zero-shot univariate forecasting; scenarios enhanced by covariates (promotions, weather) | 1. Powerful zero-shot probabilistic forecasting2. Unified multi-variable & covariate modeling (strict input requirements):  a. Future covariate names ⊆ historical covariate names  b. Each historical covariate length = target length  c. Each future covariate length = prediction length3. Efficient encoder-only structure balancing performance and speed | +| **Timer-XL** | Long-context time series large model pretrained on massive industrial data | Complex industrial forecasting requiring ultra-long history (energy, aerospace, transport) | 1. Supports input of tens of thousands of time points
2. Covers non-stationary, multivariate, and covariate scenarios
3. Pretrained on trillion-scale high-quality industrial IoT data | +| **Timer-Sundial** | Generative foundation model with "Transformer + TimeFlow" architecture | Zero-shot forecasting requiring uncertainty quantification (finance, supply chain, renewable energy) | 1. Strong zero-shot generalization; supports point & probabilistic forecasting
2. Flexible analysis of any prediction distribution statistic
3. Innovative flow-matching architecture for efficient non-deterministic sample generation | +| **Chronos-2** | Universal time series foundation model based on discrete tokenization | Rapid zero-shot univariate forecasting; scenarios enhanced by covariates (promotions, weather) | 1. Powerful zero-shot probabilistic forecasting
2. Unified multi-variable & covariate modeling (strict input requirements):
  a. Future covariate names ⊆ historical covariate names
  b. Each historical covariate length = target length
  c. Each future covariate length = prediction length
3. Efficient encoder-only structure balancing performance and speed | +| **Moirai 2.0** | Lightweight decoder-only Patch Transformer with a single patch size, multi-token prediction, and multi-quantile outputs | Zero-shot univariate forecasting where model size and inference efficiency are important, such as industrial monitoring, energy load, and equipment metrics | 1. Approximately 11.4M parameters
2. Predicts multiple patches per decoding step to reduce autoregressive overhead for long horizons
3. Outputs nine quantiles (0.1–0.9) and uses the p50 median as the point forecast
4. Uses instance normalization to mitigate distribution shift across series
5. Does not support multivariate targets or covariates | +| **Toto 2.0** | Decoder-only Patch Transformer alternating causal temporal attention and variable attention to jointly model temporal and variable dimensions | Zero-shot multivariate forecasting for observability metrics, including joint forecasting of CPU, memory, and network traffic | 1. Supports univariate and multivariate target forecasting
2. Outputs fixed quantiles from 0.1 to 0.9 and uses the p50 median as the point forecast
3. Supports cached block decoding for efficient scaling to longer forecast horizons
4. Covariates are not currently supported | ### 4.4 Deleting Models @@ -656,4 +663,4 @@ To add a new built-in custom model to AINode (using Chronos2 as example): ``` * **Complete Example** - Reference implementation: https://github.com/apache/iotdb/pull/16903 \ No newline at end of file + Reference implementation: https://github.com/apache/iotdb/pull/16903 diff --git a/src/UserGuide/latest-Table/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md b/src/UserGuide/latest-Table/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md index bc20138a4..284198ab6 100644 --- a/src/UserGuide/latest-Table/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md +++ b/src/UserGuide/latest-Table/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md @@ -46,7 +46,7 @@ The Timer model (non-built-in model) not only demonstrates excellent few-shot ge ## 4. Timer-XL Model -Timer-XL further extends and upgrades the network structure based on Timer, achieving comprehensive breakthroughs in multiple dimensions: +Timer-XL[2] further extends and upgrades the network structure based on Timer, achieving comprehensive breakthroughs in multiple dimensions: * **Ultra-Long Context Support**: This model breaks through the limitations of traditional time series forecasting models, supporting the processing of inputs with thousands of Tokens (equivalent to tens of thousands of time points), effectively solving the context length bottleneck problem. * **Coverage of Multi-Variable Forecasting Scenarios**: Supports various forecasting scenarios, including the prediction of non-stationary time series, multi-variable prediction tasks, and predictions involving covariates, meeting diversified business needs. @@ -56,7 +56,7 @@ Timer-XL further extends and upgrades the network structure based on Timer, achi ## 5. Timer-Sundial Model -Timer-Sundial is a series of generative foundational models focused on time series forecasting. The base version has 128 million parameters and has been pre-trained on 1 trillion time points, with the following core characteristics: +Timer-Sundial[3] is a series of generative foundational models focused on time series forecasting. The base version has 128 million parameters and has been pre-trained on 1 trillion time points, with the following core characteristics: * **Strong Generalization Performance**: Possesses zero-shot forecasting capabilities and can support both point forecasting and probabilistic forecasting simultaneously. * **Flexible Prediction Distribution Analysis**: Not only can it predict means or quantiles, but it can also evaluate any statistical properties of the prediction distribution through the raw samples generated by the model. @@ -66,7 +66,7 @@ Timer-Sundial is a series of generative foundational models focused on time seri ## 6. Chronos-2 Model -Chronos-2 is a universal time series foundational model developed by the Amazon Web Services (AWS) research team, evolved from the Chronos discrete token modeling paradigm. This model is suitable for both zero-shot univariate forecasting and covariate forecasting. Its main characteristics include: +Chronos-2[4] is a universal time series foundational model developed by the Amazon Web Services (AWS) research team, evolved from the Chronos discrete token modeling paradigm. This model is suitable for both zero-shot univariate forecasting and covariate forecasting. Its main characteristics include: * **Probabilistic Forecasting Capability**: The model outputs multi-step prediction results in a generative manner, supporting quantile or distribution-level forecasting to characterize future uncertainty. * **Zero-Shot General Forecasting**: Leveraging the contextual learning ability acquired through pre-training, it can directly execute forecasting on unseen datasets without retraining or parameter updates. @@ -78,7 +78,34 @@ Chronos-2 is a universal time series foundational model developed by the Amazon ![](/img/timeseries-large-model-chronos2.png) -## 7. Performance Showcase +## 7. Moirai2 Model + +Moirai2[5] (Moirai 2.0) is a general-purpose time series foundation model developed by Salesforce AI Research (supported in V2.0.10 and later). AINode currently integrates the Moirai 2.0 R-small variant, which has approximately 11.4M parameters. Unlike Moirai 1.0, which uses a masked encoder architecture, Moirai 2.0 uses a causal decoder-only Patch Transformer. With a single patch size, multi-token prediction, and multi-quantile outputs, it provides efficient univariate forecasting with a compact model. Its core features include: + +- **Lightweight Model Architecture**: Uses a decoder-only Patch Transformer with RMSNorm, rotary positional embeddings, and SiLU-GLU feed-forward networks to balance forecasting capability and inference efficiency at a small parameter scale. +- **Multi-Token Prediction**: Predicts multiple patches at each decoding step, reducing the number of autoregressive decoding steps required for long forecast horizons. +- **Probabilistic Forecasting**: Outputs nine quantiles from 0.1 to 0.9. AINode uses the p50 median as the point forecast. +- **Patch Decoding**: Groups the time series into fixed-size patches before the attention module to improve temporal feature extraction and decoding efficiency. +- **Instance Normalization**: Standardizes each time series before model input and applies denormalization after output to mitigate distribution shifts across series. +- **Input Scope**: Focuses on univariate forecasting and does not support multivariate targets or covariates. + +![](/img/LargeModel-moirai2.png) + +> Note: The Moirai 2.0 R-small model weights are licensed under CC BY-NC 4.0 and are restricted to research use. + +## 8. Toto Model + +Toto[6] (Toto 2.0) is a next-generation time series foundation model developed by Datadog (supported in V2.0.10 and later), primarily for forecasting in observability scenarios. AINode currently integrates the 2.5B-parameter variant. It is based on a decoder-only Patch Transformer architecture that alternates causal temporal attention and variable attention to jointly model the temporal and variable dimensions. Its core features include: + +- **Univariate and Multivariate Forecasting**: Supports both individual target variables and joint forecasting of multiple related target variables, making it suitable for observability metrics such as CPU, memory, and network traffic. +- **Probabilistic Forecasting**: Outputs fixed quantiles from 0.1 to 0.9 to represent forecasting uncertainty. AINode uses the p50 median as the point forecast. +- **Efficient Block Decoding**: Uses cached block decoding to generate forecasts in blocks, reducing repeated computation for longer forecast horizons. +- **Large-Scale Model Capability**: Improves forecasting performance by scaling model parameters and pre-training data, providing strong zero-shot generalization. +- **Covariate Limitation**: The currently integrated version does not support historical covariates or known future covariates. + +![](/img/LargeModel-toto.png) + +## 9. Performance Showcase Time Series Large Models can adapt to real time series data from various different domains and scenarios, demonstrating excellent processing capabilities across various tasks. The following shows the actual performance on different datasets: @@ -100,7 +127,7 @@ Using Time Series Large Models to accurately identify outliers that deviate sign ![](/img/LargeModel05.png) -## 8. Deployment and Usage +## 10. Deployment and Usage 1. Open the IoTDB CLI console and check that the ConfigNode, DataNode, and AINode nodes are all Running. @@ -143,6 +170,8 @@ IoTDB> show models | timer_xl| timer| builtin| active| | sundial| sundial| builtin| active| | chronos2| t5| builtin| active| +| moirai2| moirai| builtin| active| +| toto| toto| builtin| active| +---------------------+---------+--------+--------+ ``` @@ -150,8 +179,12 @@ IoTDB> show models **[1]** Timer: Generative Pre-trained Transformers Are Large Time Series Models, Yong Liu, Haoran Zhang, Chenyu Li, Xiangdong Huang, Jianmin Wang, Mingsheng Long. [↩ Back]() -**[2]** TIMER-XL: LONG-CONTEXT TRANSFORMERS FOR UNIFIED TIME SERIES FORECASTING, Yong Liu, Guo Qin, Xiangdong Huang, Jianmin Wang, Mingsheng Long. [↩ Back]() +**[2]** TIMER-XL: LONG-CONTEXT TRANSFORMERS FOR UNIFIED TIME SERIES FORECASTING, Yong Liu, Guo Qin, Xiangdong Huang, Jianmin Wang, Mingsheng Long. [↩ Back](#ref2) + +**[3]** Sundial: A Family of Highly Capable Time Series Foundation Models, Yong Liu, Guo Qin, Zhiyuan Shi, Zhi Chen, Caiyin Yang, Xiangdong Huang, Jianmin Wang, Mingsheng Long, **ICML 2025 spotlight**. [↩ Back](#ref3) + +**[4]** Chronos-2: From Univariate to Universal Forecasting, Abdul Fatir Ansari, Oleksandr Shchur, Jaris Küken, Andreas Auer, Boran Han, Pedro Mercado, Syama Sundar Rangapuram, Huibin Shen, Lorenzo Stella, Xiyuan Zhang, Mononito Goswami, Shubham Kapoor, Danielle C. Maddix, Pablo Guerron, Tony Hu, Junming Yin, Nick Erickson, Prateek Mutalik Desai, Hao Wang, Huzefa Rangwala, George Karypis, Yuyang Wang, Michael Bohlke-Schneider, **arXiv:2510.15821**. [↩ Back](#ref4) -**[3]** Sundial: A Family of Highly Capable Time Series Foundation Models, Yong Liu, Guo Qin, Zhiyuan Shi, Zhi Chen, Caiyin Yang, Xiangdong Huang, Jianmin Wang, Mingsheng Long, **ICML 2025 spotlight**. [↩ Back]() +[5] Moirai 2.0: When Less Is More for Time Series Forecasting, Salesforce AI Research, arXiv:2511.11698. [↩ Back](#ref5) -**[4]** Chronos-2: From Univariate to Universal Forecasting, Abdul Fatir Ansari, Oleksandr Shchur, Jaris Küken, Andreas Auer, Boran Han, Pedro Mercado, Syama Sundar Rangapuram, Huibin Shen, Lorenzo Stella, Xiyuan Zhang, Mononito Goswami, Shubham Kapoor, Danielle C. Maddix, Pablo Guerron, Tony Hu, Junming Yin, Nick Erickson, Prateek Mutalik Desai, Hao Wang, Huzefa Rangwala, George Karypis, Yuyang Wang, Michael Bohlke-Schneider, **arXiv:2510.15821**. [↩ Back]() \ No newline at end of file +[6] Toto 2.0: Time Series Forecasting Enters the Scaling Era, Datadog, arXiv:2605.20119. [↩ Back](#ref6) diff --git a/src/UserGuide/latest/AI-capability/AINode_Upgrade_apache.md b/src/UserGuide/latest/AI-capability/AINode_Upgrade_apache.md index 259417fbc..0f614b413 100644 --- a/src/UserGuide/latest/AI-capability/AINode_Upgrade_apache.md +++ b/src/UserGuide/latest/AI-capability/AINode_Upgrade_apache.md @@ -416,7 +416,12 @@ IoTDB> show models | custom| | user_defined| active| | timer_xl| timer| builtin| activating| | sundial| sundial| builtin| active| +| sundialx_1| sundial| fine_tuned| active| +| sundialx_4| sundial| fine_tuned| training| +| sundialx_5| sundial| fine_tuned| failed| | chronos2| t5| builtin| inactive| +| moirai2| moirai| builtin| inactive| +| toto| toto| builtin| inactive| +---------------------+--------------+--------------+-------------+ ``` @@ -424,22 +429,24 @@ Built-in traditional time series model introduction: | Model Name | Core Concept | Applicable Scenario | Main Features | |----------------------------------| ----------------------------------------------------------------------------------------- | ------------------------------------------------------------------------ | ------------------------------------------------------------------------------------------------------------------------------------------- | -| **ARIMA** (AutoRegressive Integrated Moving Average) | Combines autoregression (AR), differencing (I), and moving average (MA), used for predicting stationary time series or data that can be made stationary through differencing. | Univariate time series prediction, such as stock prices, sales, economic indicators. | 1. Suitable for linear trends and weak seasonality data. 2. Requires selecting parameters (p,d,q). 3. Sensitive to missing values. | -| **Holt-Winters** (Three-parameter exponential smoothing) | Based on exponential smoothing, introduces level, trend, and seasonal components, suitable for data with trend and seasonality. | Time series with clear seasonality and trend, such as monthly sales, power demand. | 1. Can handle additive or multiplicative seasonality. 2. Gives higher weight to recent data. 3. Simple to implement. | -| **Exponential Smoothing** (Exponential smoothing) | Uses weighted average of historical data, with weights decreasing exponentially over time, emphasizing the importance of recent observations. | Data without significant seasonality but with trend, such as short-term demand prediction. | 1. Few parameters, simple calculation. 2. Suitable for stationary or slowly changing sequences. 3. Can be extended to double or triple exponential smoothing. | -| **Naive Forecaster** (Naive predictor) | Uses the observation of the most recent period as the prediction for the next period, the simplest baseline model. | As a benchmark for other models or simple prediction when data has no obvious pattern. | 1. No training required. 2. Sensitive to sudden changes. 3. Seasonal naive variant can use the same period of the previous season to predict. | -| **STL Forecaster** (Seasonal-Trend Decomposition Forecast) | Based on STL decomposition of time series, predicts trend, seasonal, and residual components separately, then combines them. | Data with complex seasonality, trend, and non-linear patterns, such as climate data, traffic flow. | 1. Can handle non-fixed seasonality. 2. Robust to outliers. 3. After decomposition, other models can be combined to predict each component. | -| **Gaussian HMM** (Gaussian Hidden Markov Model) | Assumes observed data is generated by hidden states, with each state's observation probability following a Gaussian distribution. | State sequence prediction or classification, such as speech recognition, financial state identification. | 1. Suitable for time series state modeling. 2. Assumes observations are independent given the state. 3. Requires specifying the number of hidden states. | -| **GMM HMM** (Gaussian Mixture Hidden Markov Model) | An extension of Gaussian HMM, where each state's observation probability is described by a Gaussian Mixture Model, capturing more complex observation distributions. | Scenarios requiring multi-modal observation distributions, such as complex action recognition, biosignal analysis. | 1. More flexible than single Gaussian. 2. More parameters, higher computational complexity. 3. Requires training the number of GMM components. | -| **STRAY** (Anomaly Detection based on Singular Value Decomposition) | Detects anomalies in high-dimensional data through Singular Value Decomposition (SVD), commonly used for time series anomaly detection. | High-dimensional time series anomaly detection, such as sensor networks, IT system monitoring. | 1. No distribution assumption required. 2. Can handle high-dimensional data. 3. Sensitive to global anomalies, may miss local anomalies. | +| **ARIMA** (AutoRegressive Integrated Moving Average) | Combines autoregression (AR), differencing (I), and moving average (MA), used for predicting stationary time series or data that can be made stationary through differencing. | Univariate time series prediction, such as stock prices, sales, economic indicators. | 1. Suitable for linear trends and weak seasonality data.
2. Requires selecting parameters (p,d,q).
3. Sensitive to missing values. | +| **Holt-Winters** (Three-parameter exponential smoothing) | Based on exponential smoothing, introduces level, trend, and seasonal components, suitable for data with trend and seasonality. | Time series with clear seasonality and trend, such as monthly sales, power demand. | 1. Can handle additive or multiplicative seasonality.
2. Gives higher weight to recent data.
3. Simple to implement. | +| **Exponential Smoothing** (Exponential smoothing) | Uses weighted average of historical data, with weights decreasing exponentially over time, emphasizing the importance of recent observations. | Data without significant seasonality but with trend, such as short-term demand prediction. | 1. Few parameters, simple calculation.
2. Suitable for stationary or slowly changing sequences.
3. Can be extended to double or triple exponential smoothing. | +| **Naive Forecaster** (Naive predictor) | Uses the observation of the most recent period as the prediction for the next period, the simplest baseline model. | As a benchmark for other models or simple prediction when data has no obvious pattern. | 1. No training required.
2. Sensitive to sudden changes.
3. Seasonal naive variant can use the same period of the previous season to predict. | +| **STL Forecaster** (Seasonal-Trend Decomposition Forecast) | Based on STL decomposition of time series, predicts trend, seasonal, and residual components separately, then combines them. | Data with complex seasonality, trend, and non-linear patterns, such as climate data, traffic flow. | 1. Can handle non-fixed seasonality.
2. Robust to outliers.
3. After decomposition, other models can be combined to predict each component. | +| **Gaussian HMM** (Gaussian Hidden Markov Model) | Assumes observed data is generated by hidden states, with each state's observation probability following a Gaussian distribution. | State sequence prediction or classification, such as speech recognition, financial state identification. | 1. Suitable for time series state modeling.
2. Assumes observations are independent given the state.
3. Requires specifying the number of hidden states. | +| **GMM HMM** (Gaussian Mixture Hidden Markov Model) | An extension of Gaussian HMM, where each state's observation probability is described by a Gaussian Mixture Model, capturing more complex observation distributions. | Scenarios requiring multi-modal observation distributions, such as complex action recognition, biosignal analysis. | 1. More flexible than single Gaussian.
2. More parameters, higher computational complexity.
3. Requires training the number of GMM components. | +| **STRAY** (Anomaly Detection based on Singular Value Decomposition) | Detects anomalies in high-dimensional data through Singular Value Decomposition (SVD), commonly used for time series anomaly detection. | High-dimensional time series anomaly detection, such as sensor networks, IT system monitoring. | 1. No distribution assumption required.
2. Can handle high-dimensional data.
3. Sensitive to global anomalies, may miss local anomalies. | Built-in time series large model introduction: | Model Name | Core Concept | Applicable Scenario | Main Features | |---------------| ---------------------------------------------------------------------- | ------------------------------------------------------------------------ | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | -| **Timer-XL** | Time series large model supporting ultra-long context, enhancing generalization capability through large-scale industrial data pre-training. | Complex industrial prediction requiring extremely long historical data, such as energy, aerospace, and transportation. | 1. Ultra-long context support, can handle tens of thousands of time points as input. 2. Multi-scenario coverage, supports non-stationary, multi-variable, and covariate prediction. 3. Pre-trained on trillions of high-quality industrial time series data. | -| **Timer-Sundial** | A generative foundational model based on "Transformer + TimeFlow" architecture, focusing on probabilistic prediction. | Zero-shot prediction scenarios requiring quantification of uncertainty, such as finance, supply chain, and new energy power generation. | 1. Strong zero-shot generalization capability, supports point prediction and probabilistic prediction. 2. Can flexibly analyze any statistical properties of the prediction distribution. 3. Innovative generative architecture, achieving efficient non-deterministic sample generation. | -| **Chronos-2** | A general time series foundational model based on discrete tokenization paradigm, converting prediction into language modeling tasks. | Rapid zero-shot univariate prediction, and scenarios that can leverage covariates (e.g., promotions, weather) to improve results. | 1. Strong zero-shot probabilistic prediction capability. 2. Supports unified covariate modeling, but has strict input requirements: a. The set of names of future covariates must be a subset of the set of names of historical covariates; b. The length of each historical covariate must equal the length of the target variable; c. The length of each future covariate must equal the prediction length; 3. Uses an efficient encoder-style structure, balancing performance and inference speed. | +| **Timer-XL** | Time series large model supporting ultra-long context, enhancing generalization capability through large-scale industrial data pre-training. | Complex industrial prediction requiring extremely long historical data, such as energy, aerospace, and transportation. | 1. Ultra-long context support, can handle tens of thousands of time points as input.
2. Multi-scenario coverage, supports non-stationary, multi-variable, and covariate prediction.
3. Pre-trained on trillions of high-quality industrial time series data. | +| **Timer-Sundial** | A generative foundational model based on "Transformer + TimeFlow" architecture, focusing on probabilistic prediction. | Zero-shot prediction scenarios requiring quantification of uncertainty, such as finance, supply chain, and new energy power generation. | 1. Strong zero-shot generalization capability, supports point prediction and probabilistic prediction.
2. Can flexibly analyze any statistical properties of the prediction distribution.
3. Innovative generative architecture, achieving efficient non-deterministic sample generation. | +| **Chronos-2** | A general time series foundational model based on discrete tokenization paradigm, converting prediction into language modeling tasks. | Rapid zero-shot univariate prediction, and scenarios that can leverage covariates (e.g., promotions, weather) to improve results. | 1. Strong zero-shot probabilistic prediction capability.
2. Supports unified covariate modeling, but has strict input requirements:
  a. The set of names of future covariates must be a subset of the set of names of historical covariates;
  b. The length of each historical covariate must equal the length of the target variable;
  c. The length of each future covariate must equal the prediction length;
3. Uses an efficient encoder-style structure, balancing performance and inference speed. | +| **Moirai 2.0** | Uses a lightweight decoder-only Patch Transformer with a single patch size, multi-token prediction, and multi-quantile outputs for efficient univariate forecasting. | Zero-shot univariate forecasting where model size and inference efficiency are important, such as industrial monitoring, energy load, and equipment metric forecasting. | 1. Approximately 11.4M parameters.
2. Predicts multiple patches per decoding step to reduce autoregressive overhead for long horizons.
3. Outputs nine quantiles (0.1–0.9) and uses the p50 median as the point forecast.
4. Uses instance normalization to mitigate distribution shift across series.
5. Does not support multivariate targets or covariates. | +| **Toto 2.0** | Uses a decoder-only Patch Transformer that alternates causal temporal attention and variable attention to jointly model the temporal and variable dimensions. | Zero-shot forecasting for multivariate time series such as observability metrics, including joint forecasting of CPU, memory, and network traffic. | 1. Supports univariate and multivariate target forecasting.
2. Outputs fixed quantiles from 0.1 to 0.9 and uses the p50 median as the point forecast.
3. Supports cached block decoding for efficient scaling to longer forecast horizons.
4. Covariates are not currently supported. | ### 4.4 Delete Models @@ -675,4 +682,4 @@ Support adding new built-in models to AINode. The specific steps are as follows * **Full Example** -Full example can be referenced at https://github.com/apache/iotdb/pull/16903 \ No newline at end of file +Full example can be referenced at https://github.com/apache/iotdb/pull/16903 diff --git a/src/UserGuide/latest/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md b/src/UserGuide/latest/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md index bc20138a4..284198ab6 100644 --- a/src/UserGuide/latest/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md +++ b/src/UserGuide/latest/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md @@ -46,7 +46,7 @@ The Timer model (non-built-in model) not only demonstrates excellent few-shot ge ## 4. Timer-XL Model -Timer-XL further extends and upgrades the network structure based on Timer, achieving comprehensive breakthroughs in multiple dimensions: +Timer-XL[2] further extends and upgrades the network structure based on Timer, achieving comprehensive breakthroughs in multiple dimensions: * **Ultra-Long Context Support**: This model breaks through the limitations of traditional time series forecasting models, supporting the processing of inputs with thousands of Tokens (equivalent to tens of thousands of time points), effectively solving the context length bottleneck problem. * **Coverage of Multi-Variable Forecasting Scenarios**: Supports various forecasting scenarios, including the prediction of non-stationary time series, multi-variable prediction tasks, and predictions involving covariates, meeting diversified business needs. @@ -56,7 +56,7 @@ Timer-XL further extends and upgrades the network structure based on Timer, achi ## 5. Timer-Sundial Model -Timer-Sundial is a series of generative foundational models focused on time series forecasting. The base version has 128 million parameters and has been pre-trained on 1 trillion time points, with the following core characteristics: +Timer-Sundial[3] is a series of generative foundational models focused on time series forecasting. The base version has 128 million parameters and has been pre-trained on 1 trillion time points, with the following core characteristics: * **Strong Generalization Performance**: Possesses zero-shot forecasting capabilities and can support both point forecasting and probabilistic forecasting simultaneously. * **Flexible Prediction Distribution Analysis**: Not only can it predict means or quantiles, but it can also evaluate any statistical properties of the prediction distribution through the raw samples generated by the model. @@ -66,7 +66,7 @@ Timer-Sundial is a series of generative foundational models focused on time seri ## 6. Chronos-2 Model -Chronos-2 is a universal time series foundational model developed by the Amazon Web Services (AWS) research team, evolved from the Chronos discrete token modeling paradigm. This model is suitable for both zero-shot univariate forecasting and covariate forecasting. Its main characteristics include: +Chronos-2[4] is a universal time series foundational model developed by the Amazon Web Services (AWS) research team, evolved from the Chronos discrete token modeling paradigm. This model is suitable for both zero-shot univariate forecasting and covariate forecasting. Its main characteristics include: * **Probabilistic Forecasting Capability**: The model outputs multi-step prediction results in a generative manner, supporting quantile or distribution-level forecasting to characterize future uncertainty. * **Zero-Shot General Forecasting**: Leveraging the contextual learning ability acquired through pre-training, it can directly execute forecasting on unseen datasets without retraining or parameter updates. @@ -78,7 +78,34 @@ Chronos-2 is a universal time series foundational model developed by the Amazon ![](/img/timeseries-large-model-chronos2.png) -## 7. Performance Showcase +## 7. Moirai2 Model + +Moirai2[5] (Moirai 2.0) is a general-purpose time series foundation model developed by Salesforce AI Research (supported in V2.0.10 and later). AINode currently integrates the Moirai 2.0 R-small variant, which has approximately 11.4M parameters. Unlike Moirai 1.0, which uses a masked encoder architecture, Moirai 2.0 uses a causal decoder-only Patch Transformer. With a single patch size, multi-token prediction, and multi-quantile outputs, it provides efficient univariate forecasting with a compact model. Its core features include: + +- **Lightweight Model Architecture**: Uses a decoder-only Patch Transformer with RMSNorm, rotary positional embeddings, and SiLU-GLU feed-forward networks to balance forecasting capability and inference efficiency at a small parameter scale. +- **Multi-Token Prediction**: Predicts multiple patches at each decoding step, reducing the number of autoregressive decoding steps required for long forecast horizons. +- **Probabilistic Forecasting**: Outputs nine quantiles from 0.1 to 0.9. AINode uses the p50 median as the point forecast. +- **Patch Decoding**: Groups the time series into fixed-size patches before the attention module to improve temporal feature extraction and decoding efficiency. +- **Instance Normalization**: Standardizes each time series before model input and applies denormalization after output to mitigate distribution shifts across series. +- **Input Scope**: Focuses on univariate forecasting and does not support multivariate targets or covariates. + +![](/img/LargeModel-moirai2.png) + +> Note: The Moirai 2.0 R-small model weights are licensed under CC BY-NC 4.0 and are restricted to research use. + +## 8. Toto Model + +Toto[6] (Toto 2.0) is a next-generation time series foundation model developed by Datadog (supported in V2.0.10 and later), primarily for forecasting in observability scenarios. AINode currently integrates the 2.5B-parameter variant. It is based on a decoder-only Patch Transformer architecture that alternates causal temporal attention and variable attention to jointly model the temporal and variable dimensions. Its core features include: + +- **Univariate and Multivariate Forecasting**: Supports both individual target variables and joint forecasting of multiple related target variables, making it suitable for observability metrics such as CPU, memory, and network traffic. +- **Probabilistic Forecasting**: Outputs fixed quantiles from 0.1 to 0.9 to represent forecasting uncertainty. AINode uses the p50 median as the point forecast. +- **Efficient Block Decoding**: Uses cached block decoding to generate forecasts in blocks, reducing repeated computation for longer forecast horizons. +- **Large-Scale Model Capability**: Improves forecasting performance by scaling model parameters and pre-training data, providing strong zero-shot generalization. +- **Covariate Limitation**: The currently integrated version does not support historical covariates or known future covariates. + +![](/img/LargeModel-toto.png) + +## 9. Performance Showcase Time Series Large Models can adapt to real time series data from various different domains and scenarios, demonstrating excellent processing capabilities across various tasks. The following shows the actual performance on different datasets: @@ -100,7 +127,7 @@ Using Time Series Large Models to accurately identify outliers that deviate sign ![](/img/LargeModel05.png) -## 8. Deployment and Usage +## 10. Deployment and Usage 1. Open the IoTDB CLI console and check that the ConfigNode, DataNode, and AINode nodes are all Running. @@ -143,6 +170,8 @@ IoTDB> show models | timer_xl| timer| builtin| active| | sundial| sundial| builtin| active| | chronos2| t5| builtin| active| +| moirai2| moirai| builtin| active| +| toto| toto| builtin| active| +---------------------+---------+--------+--------+ ``` @@ -150,8 +179,12 @@ IoTDB> show models **[1]** Timer: Generative Pre-trained Transformers Are Large Time Series Models, Yong Liu, Haoran Zhang, Chenyu Li, Xiangdong Huang, Jianmin Wang, Mingsheng Long. [↩ Back]() -**[2]** TIMER-XL: LONG-CONTEXT TRANSFORMERS FOR UNIFIED TIME SERIES FORECASTING, Yong Liu, Guo Qin, Xiangdong Huang, Jianmin Wang, Mingsheng Long. [↩ Back]() +**[2]** TIMER-XL: LONG-CONTEXT TRANSFORMERS FOR UNIFIED TIME SERIES FORECASTING, Yong Liu, Guo Qin, Xiangdong Huang, Jianmin Wang, Mingsheng Long. [↩ Back](#ref2) + +**[3]** Sundial: A Family of Highly Capable Time Series Foundation Models, Yong Liu, Guo Qin, Zhiyuan Shi, Zhi Chen, Caiyin Yang, Xiangdong Huang, Jianmin Wang, Mingsheng Long, **ICML 2025 spotlight**. [↩ Back](#ref3) + +**[4]** Chronos-2: From Univariate to Universal Forecasting, Abdul Fatir Ansari, Oleksandr Shchur, Jaris Küken, Andreas Auer, Boran Han, Pedro Mercado, Syama Sundar Rangapuram, Huibin Shen, Lorenzo Stella, Xiyuan Zhang, Mononito Goswami, Shubham Kapoor, Danielle C. Maddix, Pablo Guerron, Tony Hu, Junming Yin, Nick Erickson, Prateek Mutalik Desai, Hao Wang, Huzefa Rangwala, George Karypis, Yuyang Wang, Michael Bohlke-Schneider, **arXiv:2510.15821**. [↩ Back](#ref4) -**[3]** Sundial: A Family of Highly Capable Time Series Foundation Models, Yong Liu, Guo Qin, Zhiyuan Shi, Zhi Chen, Caiyin Yang, Xiangdong Huang, Jianmin Wang, Mingsheng Long, **ICML 2025 spotlight**. [↩ Back]() +[5] Moirai 2.0: When Less Is More for Time Series Forecasting, Salesforce AI Research, arXiv:2511.11698. [↩ Back](#ref5) -**[4]** Chronos-2: From Univariate to Universal Forecasting, Abdul Fatir Ansari, Oleksandr Shchur, Jaris Küken, Andreas Auer, Boran Han, Pedro Mercado, Syama Sundar Rangapuram, Huibin Shen, Lorenzo Stella, Xiyuan Zhang, Mononito Goswami, Shubham Kapoor, Danielle C. Maddix, Pablo Guerron, Tony Hu, Junming Yin, Nick Erickson, Prateek Mutalik Desai, Hao Wang, Huzefa Rangwala, George Karypis, Yuyang Wang, Michael Bohlke-Schneider, **arXiv:2510.15821**. [↩ Back]() \ No newline at end of file +[6] Toto 2.0: Time Series Forecasting Enters the Scaling Era, Datadog, arXiv:2605.20119. [↩ Back](#ref6) diff --git a/src/zh/UserGuide/Master/Table/AI-capability/AINode_Upgrade_apache.md b/src/zh/UserGuide/Master/Table/AI-capability/AINode_Upgrade_apache.md index 6356a0489..915ef8691 100644 --- a/src/zh/UserGuide/Master/Table/AI-capability/AINode_Upgrade_apache.md +++ b/src/zh/UserGuide/Master/Table/AI-capability/AINode_Upgrade_apache.md @@ -443,7 +443,12 @@ IoTDB> show models | custom| | user_defined| active| | timer_xl| timer| builtin| activating| | sundial| sundial| builtin| active| +| sundialx_1| sundial| fine_tuned| active| +| sundialx_4| sundial| fine_tuned| training| +| sundialx_5| sundial| fine_tuned| failed| | chronos2| t5| builtin| inactive| +| moirai2| moirai| builtin| inactive| +| toto| toto| builtin| inactive| +---------------------+--------------+--------------+-------------+ ``` @@ -451,22 +456,24 @@ IoTDB> show models | 模型名称 | 核心概念 | 适用场景 | 主要特点 | |----------------------------------| ----------------------------------------------------------------------------------------- | ---------------------------------------------------------------- | ------------------------------------------------------------------------------------------------- | -| **ARIMA**(自回归整合移动平均模型) | 结合自回归(AR)、差分(I)和移动平均(MA)，用于预测平稳时间序列或可通过差分变为平稳的数据。 | 单变量时间序列预测，如股票价格、销量、经济指标等。| 1. 适用于线性趋势和季节性较弱的数据。2. 需要选择参数 (p,d,q)。3. 对缺失值敏感。 | -| **Holt-Winters**(三参数指数平滑) | 基于指数平滑，引入水平、趋势和季节性三个分量，适用于具有趋势和季节性的数据。 | 有明显季节性和趋势的时间序列，如月度销售额、电力需求等。 | 1. 可处理加性或乘性季节性。2. 对近期数据赋予更高权重。3. 简单易实现。 | -| **Exponential Smoothing**(指数平滑) | 通过加权平均历史数据，权重随时间指数递减，强调近期观测值的重要性。 | 无显著季节性但存在趋势的数据，如短期需求预测。 | 1. 参数少，计算简单。2. 适合平稳或缓慢变化序列。3. 可扩展为双指数或三指数平滑。 | -| **Naive Forecaster**(朴素预测器) | 使用最近一期的观测值作为下一期的预测值，是最简单的基准模型。 | 作为其他模型的比较基准，或数据无明显模式时的简单预测。 | 1. 无需训练。2. 对突发变化敏感。3. 季节性朴素变体可用前一季节同期值预测。 | -| **STL Forecaster**(季节趋势分解预测) | 基于STL分解时间序列，分别预测趋势、季节性和残差分量后组合。 | 具有复杂季节性、趋势和非线性模式的数据，如气候数据、交通流量。 | 1. 能处理非固定季节性。2. 对异常值稳健。3. 分解后可结合其他模型预测各分量。 | -| **Gaussian HMM**(高斯隐马尔可夫模型) | 假设观测数据由隐藏状态生成，每个状态的观测概率服从高斯分布。 | 状态序列预测或分类，如语音识别、金融状态识别。 | 1. 适用于时序数据的状态建模。2. 假设观测值在给定状态下独立。3. 需指定隐藏状态数量。 | -| **GMM HMM** (高斯混合隐马尔可夫模型) | 扩展Gaussian HMM，每个状态的观测概率由高斯混合模型描述，可捕捉更复杂的观测分布。 | 需要多模态观测分布的场景，如复杂动作识别、生物信号分析。 | 1. 比单一高斯更灵活。2. 参数更多，计算复杂度高。3. 需训练GMM成分数。 | -| **STRAY**(基于奇异值的异常检测) | 通过奇异值分解(SVD)检测高维数据中的异常点，常用于时间序列异常检测。 | 高维时间序列的异常检测，如传感器网络、IT系统监控。 | 1. 无需分布假设。2. 可处理高维数据。3. 对全局异常敏感，局部异常可能漏检。 | +| **ARIMA**(自回归整合移动平均模型) | 结合自回归(AR)、差分(I)和移动平均(MA)，用于预测平稳时间序列或可通过差分变为平稳的数据。 | 单变量时间序列预测，如股票价格、销量、经济指标等。| 1. 适用于线性趋势和季节性较弱的数据。
2. 需要选择参数 (p,d,q)。
3. 对缺失值敏感。 | +| **Holt-Winters**(三参数指数平滑) | 基于指数平滑，引入水平、趋势和季节性三个分量，适用于具有趋势和季节性的数据。 | 有明显季节性和趋势的时间序列，如月度销售额、电力需求等。 | 1. 可处理加性或乘性季节性。
2. 对近期数据赋予更高权重。
3. 简单易实现。 | +| **Exponential Smoothing**(指数平滑) | 通过加权平均历史数据，权重随时间指数递减，强调近期观测值的重要性。 | 无显著季节性但存在趋势的数据，如短期需求预测。 | 1. 参数少，计算简单。
2. 适合平稳或缓慢变化序列。
3. 可扩展为双指数或三指数平滑。 | +| **Naive Forecaster**(朴素预测器) | 使用最近一期的观测值作为下一期的预测值，是最简单的基准模型。 | 作为其他模型的比较基准，或数据无明显模式时的简单预测。 | 1. 无需训练。
2. 对突发变化敏感。
3. 季节性朴素变体可用前一季节同期值预测。 | +| **STL Forecaster**(季节趋势分解预测) | 基于STL分解时间序列，分别预测趋势、季节性和残差分量后组合。 | 具有复杂季节性、趋势和非线性模式的数据，如气候数据、交通流量。 | 1. 能处理非固定季节性。
2. 对异常值稳健。
3. 分解后可结合其他模型预测各分量。 | +| **Gaussian HMM**(高斯隐马尔可夫模型) | 假设观测数据由隐藏状态生成，每个状态的观测概率服从高斯分布。 | 状态序列预测或分类，如语音识别、金融状态识别。 | 1. 适用于时序数据的状态建模。
2. 假设观测值在给定状态下独立。
3. 需指定隐藏状态数量。 | +| **GMM HMM** (高斯混合隐马尔可夫模型) | 扩展Gaussian HMM，每个状态的观测概率由高斯混合模型描述，可捕捉更复杂的观测分布。 | 需要多模态观测分布的场景，如复杂动作识别、生物信号分析。 | 1. 比单一高斯更灵活。
2. 参数更多，计算复杂度高。
3. 需训练GMM成分数。 | +| **STRAY**(基于奇异值的异常检测) | 通过奇异值分解(SVD)检测高维数据中的异常点，常用于时间序列异常检测。 | 高维时间序列的异常检测，如传感器网络、IT系统监控。 | 1. 无需分布假设。
2. 可处理高维数据。
3. 对全局异常敏感，局部异常可能漏检。 | 内置时序大模型介绍如下： | 模型名称 | 核心概念 | 适用场景 | 主要特点 | |---------------| ---------------------------------------------------------------------- | ------------------------------------------------------------------------ | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | -| **Timer-XL** | 支持超长上下文的时序大模型，通过大规模工业数据预训练增强泛化能力。 | 需利用极长历史数据的复杂工业预测，如能源、航空航天、交通等领域。 | 1. 超长上下文支持，可处理数万时间点输入。2. 多场景覆盖，支持非平稳、多变量及协变量预测。3. 基于万亿级高质量工业时序数据预训练。 | -| **Timer-Sundial** | 采用“Transformer + TimeFlow”架构的生成式基础模型，专注于概率预测。 | 需要量化不确定性的零样本预测场景，如金融、供应链、新能源发电预测。 | 1. 强大的零样本泛化能力，支持点预测与概率预测 2. 可灵活分析预测分布的任意统计特性。3. 创新生成架构，实现高效的非确定性样本生成。 | -| **Chronos-2** | 基于离散词元化范式的通用时序基础模型，将预测转化为语言建模任务。 | 快速零样本单变量预测，以及可借助协变量（如促销、天气）提升效果的场景。 | 1. 强大的零样本概率预测能力。2. 支持协变量统一建模，但对输入有严格要求：a. 未来协变量的名称组成的集合必须是历史协变量的名称组成的集合的子集；b. 每个历史协变量的长度必须等于目标变量的长度； c. 每个未来协变量的长度必须等于预测长度；3. 采用高效的编码器式结构，兼顾性能与推理速度。 | +| **Timer-XL** | 支持超长上下文的时序大模型，通过大规模工业数据预训练增强泛化能力。 | 需利用极长历史数据的复杂工业预测，如能源、航空航天、交通等领域。 | 1. 超长上下文支持，可处理数万时间点输入。
2. 多场景覆盖，支持非平稳、多变量及协变量预测。
3. 基于万亿级高质量工业时序数据预训练。 | +| **Timer-Sundial** | 采用“Transformer + TimeFlow”架构的生成式基础模型，专注于概率预测。 | 需要量化不确定性的零样本预测场景，如金融、供应链、新能源发电预测。 | 1. 强大的零样本泛化能力，支持点预测与概率预测
2. 可灵活分析预测分布的任意统计特性。
3. 创新生成架构，实现高效的非确定性样本生成。 | +| **Chronos-2** | 基于离散词元化范式的通用时序基础模型，将预测转化为语言建模任务。 | 快速零样本单变量预测，以及可借助协变量（如促销、天气）提升效果的场景。 | 1. 强大的零样本概率预测能力。
2. 支持协变量统一建模，但对输入有严格要求：
  a. 未来协变量的名称组成的集合必须是历史协变量的名称组成的集合的子集；
  b. 每个历史协变量的长度必须等于目标变量的长度；
  c. 每个未来协变量的长度必须等于预测长度；
3. 采用高效的编码器式结构，兼顾性能与推理速度。 | +| **Moirai 2.0** | 采用轻量级 Decoder-only Patch Transformer，通过单一 Patch 尺寸、多 Token 预测和多分位数输出实现高效的单变量预测。 | 适用于对模型体量和推理效率要求较高的零样本单变量预测，如工业监测、能源负荷及设备指标预测。 | 1. 模型参数量约 11.4M。
2. 每个解码步可预测多个 Patch，降低长预测范围的自回归开销。
3. 输出 9 个分位数（0.1～0.9），使用 p50 中位数作为点预测。
4. 使用实例归一化缓解序列分布漂移。
5. 不支持多变量和协变量。 | +| **Toto 2.0** | 采用 Decoder-only Patch Transformer，交替使用因果时间注意力与变量注意力，对时间维和变量维进行联合建模。 | 面向可观测性指标等多变量时间序列的零样本预测，如 CPU、内存、网络流量等指标的联合预测。 | 1. 支持单变量和多变量目标预测。
2. 输出 0.1～0.9 的固定分位数，使用 p50 中位数作为点预测。
3. 支持缓存式块解码，可高效扩展较长预测范围。
4. 当前不支持协变量。 | ### 4.4 删除模型 diff --git a/src/zh/UserGuide/Master/Table/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md b/src/zh/UserGuide/Master/Table/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md index e1af1e1d0..5fe5c3771 100644 --- a/src/zh/UserGuide/Master/Table/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md +++ b/src/zh/UserGuide/Master/Table/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md @@ -78,7 +78,34 @@ Chronos-2 [4] ![](/img/timeseries-large-model-chronos2.png) -## 7. 效果展示 +## 7. Moirai2 模型 + +Moirai2[5]（Moirai 2.0）是由 Salesforce AI Research 推出的通用时间序列基础模型（V2.0.10 及以后版本支持）。当前集成的是 Moirai 2.0 R-small 版本，模型参数量约为 11.4M。与采用掩码编码器架构的 Moirai 1.0 不同，Moirai 2.0 使用因果 Decoder-only Patch Transformer，并通过单一 Patch 尺寸、多 Token 预测和多分位数输出，在较小模型规模下实现高效的单变量预测。其核心特性包括： + +- **轻量化模型架构**：采用 Decoder-only Patch Transformer，并结合 RMSNorm、旋转位置编码和 SiLU-GLU 前馈网络，在较小参数规模下兼顾预测能力和推理效率。 +- **多 Token 预测**：每个解码步骤可同时预测多个 Patch，减少长预测范围下所需的自回归解码次数。 +- **概率性预测能力**：模型输出 0.1～0.9 共 9 个分位数，AINode 使用 p50 中位数作为点预测结果。 +- **Patch 解码**：在进入注意力模块前，将时间序列按照固定大小的 Patch 进行分组，以提高时序特征提取和解码效率。 +- **实例归一化**：在模型输入前对每条时间序列进行标准化，并在输出后执行反归一化，以缓解不同序列间的分布漂移。 +- **输入范围**：模型聚焦于单变量预测，不支持多变量目标及协变量输入。 + +![](/img/LargeModel-moirai2.png) + +> 注意：Moirai 2.0 R-small 模型权重采用 CC BY-NC 4.0 许可证，仅限研究用途。 + +## 8. Toto 模型 + +Toto[6]（Toto 2.0）是由 Datadog 推出的新一代时间序列基础模型（V2.0.10 及以后版本支持），主要面向可观测性场景中的时间序列预测。当前集成的模型采用 2.5B 参数版本，基于 Decoder-only Patch Transformer 架构，交替使用因果时间注意力与变量注意力，对时间维度和变量维度进行联合建模。其核心特性包括： + +- **单变量与多变量预测**：支持单个目标变量以及多个相关目标变量的联合预测，适用于 CPU、内存、网络流量等可观测性指标的分析场景。 +- **概率性预测能力**：模型输出 0.1～0.9 的固定分位数，可刻画预测结果的不确定性；AINode 使用 p50 中位数作为点预测结果。 +- **高效块解码**：采用缓存式块解码机制，可分块生成预测结果，降低较长预测范围下的重复计算开销。 +- **大规模模型能力**：通过扩大模型参数量和预训练数据规模提升预测性能，具备良好的零样本泛化能力。 +- **协变量限制**：当前集成版本不支持历史协变量或已知未来协变量。 + +![](/img/LargeModel-toto.png) + +## 9. 效果展示 时序大模型能够适应多种不同领域和场景的真实时序数据，在各种任务上拥有优异的处理效果，以下是在不同数据上的真实表现： @@ -100,7 +127,7 @@ Chronos-2 [4] ![](/img/LargeModel05.png) -## 8. 部署使用 +## 10. 部署使用 1. 打开 IoTDB cli 控制台，检查 ConfigNode、DataNode、AINode 节点确保均为 Running。 @@ -143,6 +170,8 @@ IoTDB> show models | timer_xl| timer| builtin| active| | sundial| sundial| builtin| active| | chronos2| t5| builtin| active| +| moirai2| moirai| builtin| active| +| toto| toto| builtin| active| +---------------------+---------+--------+--------+ ``` @@ -150,8 +179,12 @@ IoTDB> show models **[1]** Timer- Generative Pre-trained Transformers Are Large Time Series Models, Yong Liu, Haoran Zhang, Chenyu Li, Xiangdong Huang, Jianmin Wang, Mingsheng Long. [↩ 返回](#ref1) -**[2]** TIMER-XL- LONG-CONTEXT TRANSFORMERS FOR UNIFIED TIME SERIES FORECASTING ,Yong Liu, Guo Qin, Xiangdong Huang, Jianmin Wang, Mingsheng Long. [↩ 返回](#ref2) +**[2]** TIMER-XL: LONG-CONTEXT TRANSFORMERS FOR UNIFIED TIME SERIES FORECASTING, Yong Liu, Guo Qin, Xiangdong Huang, Jianmin Wang, Mingsheng Long. [↩ 返回](#ref2) + +**[3]** Sundial: A Family of Highly Capable Time Series Foundation Models, Yong Liu, Guo Qin, Zhiyuan Shi, Zhi Chen, Caiyin Yang, Xiangdong Huang, Jianmin Wang, Mingsheng Long, **ICML 2025 spotlight**. [↩ 返回](#ref3) + +**[4]** Chronos-2: From Univariate to Universal Forecasting, Abdul Fatir Ansari, Oleksandr Shchur, Jaris Küken, Andreas Auer, Boran Han, Pedro Mercado, Syama Sundar Rangapuram, Huibin Shen, Lorenzo Stella, Xiyuan Zhang, Mononito Goswami, Shubham Kapoor, Danielle C. Maddix, Pablo Guerron, Tony Hu, Junming Yin, Nick Erickson, Prateek Mutalik Desai, Hao Wang, Huzefa Rangwala, George Karypis, Yuyang Wang, Michael Bohlke-Schneider, **arXiv:2510.15821.** [↩ 返回](#ref4) -**[3]** Sundial- A Family of Highly Capable Time Series Foundation Models, Yong Liu, Guo Qin, Zhiyuan Shi, Zhi Chen, Caiyin Yang, Xiangdong Huang, Jianmin Wang, Mingsheng Long, **ICML 2025 spotlight**. [↩ 返回](#ref3) +[5] Moirai 2.0: When Less Is More for Time Series Forecasting, Salesforce AI Research, arXiv:2511.11698. [↩ 返回](#ref5) -**[4] **Chronos-2: From Univariate to Universal Forecasting, Abdul Fatir Ansari, Oleksandr Shchur, Jaris Küken, Andreas Auer, Boran Han, Pedro Mercado, Syama Sundar Rangapuram, Huibin Shen, Lorenzo Stella, Xiyuan Zhang, Mononito Goswami, Shubham Kapoor, Danielle C. Maddix, Pablo Guerron, Tony Hu, Junming Yin, Nick Erickson, Prateek Mutalik Desai, Hao Wang, Huzefa Rangwala, George Karypis, Yuyang Wang, Michael Bohlke-Schneider, **arXiv:2510.15821.**[↩ 返回](#ref4) +[6] Toto 2.0: Time Series Forecasting Enters the Scaling Era, Datadog, arXiv:2605.20119. [↩ 返回](#ref6) diff --git a/src/zh/UserGuide/Master/Tree/AI-capability/AINode_Upgrade_apache.md b/src/zh/UserGuide/Master/Tree/AI-capability/AINode_Upgrade_apache.md index a86a0ea7c..255eaff80 100644 --- a/src/zh/UserGuide/Master/Tree/AI-capability/AINode_Upgrade_apache.md +++ b/src/zh/UserGuide/Master/Tree/AI-capability/AINode_Upgrade_apache.md @@ -419,7 +419,12 @@ IoTDB> show models | custom| | user_defined| active| | timer_xl| timer| builtin| activating| | sundial| sundial| builtin| active| +| sundialx_1| sundial| fine_tuned| active| +| sundialx_4| sundial| fine_tuned| training| +| sundialx_5| sundial| fine_tuned| failed| | chronos2| t5| builtin| inactive| +| moirai2| moirai| builtin| inactive| +| toto| toto| builtin| inactive| +---------------------+--------------+--------------+-------------+ ``` @@ -427,22 +432,24 @@ IoTDB> show models | 模型名称 | 核心概念 | 适用场景 | 主要特点 | |----------------------------------| ----------------------------------------------------------------------------------------- | ---------------------------------------------------------------- | ------------------------------------------------------------------------------------------------- | -| **ARIMA**(自回归整合移动平均模型) | 结合自回归(AR)、差分(I)和移动平均(MA)，用于预测平稳时间序列或可通过差分变为平稳的数据。 | 单变量时间序列预测，如股票价格、销量、经济指标等。| 1. 适用于线性趋势和季节性较弱的数据。2. 需要选择参数 (p,d,q)。3. 对缺失值敏感。 | -| **Holt-Winters**(三参数指数平滑) | 基于指数平滑，引入水平、趋势和季节性三个分量，适用于具有趋势和季节性的数据。 | 有明显季节性和趋势的时间序列，如月度销售额、电力需求等。 | 1. 可处理加性或乘性季节性。2. 对近期数据赋予更高权重。3. 简单易实现。 | -| **Exponential Smoothing**(指数平滑) | 通过加权平均历史数据，权重随时间指数递减，强调近期观测值的重要性。 | 无显著季节性但存在趋势的数据，如短期需求预测。 | 1. 参数少，计算简单。2. 适合平稳或缓慢变化序列。3. 可扩展为双指数或三指数平滑。 | -| **Naive Forecaster**(朴素预测器) | 使用最近一期的观测值作为下一期的预测值，是最简单的基准模型。 | 作为其他模型的比较基准，或数据无明显模式时的简单预测。 | 1. 无需训练。2. 对突发变化敏感。3. 季节性朴素变体可用前一季节同期值预测。 | -| **STL Forecaster**(季节趋势分解预测) | 基于STL分解时间序列，分别预测趋势、季节性和残差分量后组合。 | 具有复杂季节性、趋势和非线性模式的数据，如气候数据、交通流量。 | 1. 能处理非固定季节性。2. 对异常值稳健。3. 分解后可结合其他模型预测各分量。 | -| **Gaussian HMM**(高斯隐马尔可夫模型) | 假设观测数据由隐藏状态生成，每个状态的观测概率服从高斯分布。 | 状态序列预测或分类，如语音识别、金融状态识别。 | 1. 适用于时序数据的状态建模。2. 假设观测值在给定状态下独立。3. 需指定隐藏状态数量。 | -| **GMM HMM** (高斯混合隐马尔可夫模型) | 扩展Gaussian HMM，每个状态的观测概率由高斯混合模型描述，可捕捉更复杂的观测分布。 | 需要多模态观测分布的场景，如复杂动作识别、生物信号分析。 | 1. 比单一高斯更灵活。2. 参数更多，计算复杂度高。3. 需训练GMM成分数。 | -| **STRAY**(基于奇异值的异常检测) | 通过奇异值分解(SVD)检测高维数据中的异常点，常用于时间序列异常检测。 | 高维时间序列的异常检测，如传感器网络、IT系统监控。 | 1. 无需分布假设。2. 可处理高维数据。3. 对全局异常敏感，局部异常可能漏检。 | +| **ARIMA**(自回归整合移动平均模型) | 结合自回归(AR)、差分(I)和移动平均(MA)，用于预测平稳时间序列或可通过差分变为平稳的数据。 | 单变量时间序列预测，如股票价格、销量、经济指标等。| 1. 适用于线性趋势和季节性较弱的数据。
2. 需要选择参数 (p,d,q)。
3. 对缺失值敏感。 | +| **Holt-Winters**(三参数指数平滑) | 基于指数平滑，引入水平、趋势和季节性三个分量，适用于具有趋势和季节性的数据。 | 有明显季节性和趋势的时间序列，如月度销售额、电力需求等。 | 1. 可处理加性或乘性季节性。
2. 对近期数据赋予更高权重。
3. 简单易实现。 | +| **Exponential Smoothing**(指数平滑) | 通过加权平均历史数据，权重随时间指数递减，强调近期观测值的重要性。 | 无显著季节性但存在趋势的数据，如短期需求预测。 | 1. 参数少，计算简单。
2. 适合平稳或缓慢变化序列。
3. 可扩展为双指数或三指数平滑。 | +| **Naive Forecaster**(朴素预测器) | 使用最近一期的观测值作为下一期的预测值，是最简单的基准模型。 | 作为其他模型的比较基准，或数据无明显模式时的简单预测。 | 1. 无需训练。
2. 对突发变化敏感。
3. 季节性朴素变体可用前一季节同期值预测。 | +| **STL Forecaster**(季节趋势分解预测) | 基于STL分解时间序列，分别预测趋势、季节性和残差分量后组合。 | 具有复杂季节性、趋势和非线性模式的数据，如气候数据、交通流量。 | 1. 能处理非固定季节性。
2. 对异常值稳健。
3. 分解后可结合其他模型预测各分量。 | +| **Gaussian HMM**(高斯隐马尔可夫模型) | 假设观测数据由隐藏状态生成，每个状态的观测概率服从高斯分布。 | 状态序列预测或分类，如语音识别、金融状态识别。 | 1. 适用于时序数据的状态建模。
2. 假设观测值在给定状态下独立。
3. 需指定隐藏状态数量。 | +| **GMM HMM** (高斯混合隐马尔可夫模型) | 扩展Gaussian HMM，每个状态的观测概率由高斯混合模型描述，可捕捉更复杂的观测分布。 | 需要多模态观测分布的场景，如复杂动作识别、生物信号分析。 | 1. 比单一高斯更灵活。
2. 参数更多，计算复杂度高。
3. 需训练GMM成分数。 | +| **STRAY**(基于奇异值的异常检测) | 通过奇异值分解(SVD)检测高维数据中的异常点，常用于时间序列异常检测。 | 高维时间序列的异常检测，如传感器网络、IT系统监控。 | 1. 无需分布假设。
2. 可处理高维数据。
3. 对全局异常敏感，局部异常可能漏检。 | 内置时序大模型介绍如下： | 模型名称 | 核心概念 | 适用场景 | 主要特点 | |---------------| ---------------------------------------------------------------------- | ------------------------------------------------------------------------ | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | -| **Timer-XL** | 支持超长上下文的时序大模型，通过大规模工业数据预训练增强泛化能力。 | 需利用极长历史数据的复杂工业预测，如能源、航空航天、交通等领域。 | 1. 超长上下文支持，可处理数万时间点输入。2. 多场景覆盖，支持非平稳、多变量及协变量预测。3. 基于万亿级高质量工业时序数据预训练。 | -| **Timer-Sundial** | 采用“Transformer + TimeFlow”架构的生成式基础模型，专注于概率预测。 | 需要量化不确定性的零样本预测场景，如金融、供应链、新能源发电预测。 | 1. 强大的零样本泛化能力，支持点预测与概率预测 2. 可灵活分析预测分布的任意统计特性。3. 创新生成架构，实现高效的非确定性样本生成。 | -| **Chronos-2** | 基于离散词元化范式的通用时序基础模型，将预测转化为语言建模任务。 | 快速零样本单变量预测，以及可借助协变量（如促销、天气）提升效果的场景。 | 1. 强大的零样本概率预测能力。2. 支持协变量统一建模，但对输入有严格要求：a. 未来协变量的名称组成的集合必须是历史协变量的名称组成的集合的子集；b. 每个历史协变量的长度必须等于目标变量的长度； c. 每个未来协变量的长度必须等于预测长度；3. 采用高效的编码器式结构，兼顾性能与推理速度。 | +| **Timer-XL** | 支持超长上下文的时序大模型，通过大规模工业数据预训练增强泛化能力。 | 需利用极长历史数据的复杂工业预测，如能源、航空航天、交通等领域。 | 1. 超长上下文支持，可处理数万时间点输入。
2. 多场景覆盖，支持非平稳、多变量及协变量预测。
3. 基于万亿级高质量工业时序数据预训练。 | +| **Timer-Sundial** | 采用“Transformer + TimeFlow”架构的生成式基础模型，专注于概率预测。 | 需要量化不确定性的零样本预测场景，如金融、供应链、新能源发电预测。 | 1. 强大的零样本泛化能力，支持点预测与概率预测
2. 可灵活分析预测分布的任意统计特性。
3. 创新生成架构，实现高效的非确定性样本生成。 | +| **Chronos-2** | 基于离散词元化范式的通用时序基础模型，将预测转化为语言建模任务。 | 快速零样本单变量预测，以及可借助协变量（如促销、天气）提升效果的场景。 | 1. 强大的零样本概率预测能力。
2. 支持协变量统一建模，但对输入有严格要求：
  a. 未来协变量的名称组成的集合必须是历史协变量的名称组成的集合的子集；
  b. 每个历史协变量的长度必须等于目标变量的长度；
  c. 每个未来协变量的长度必须等于预测长度；
3. 采用高效的编码器式结构，兼顾性能与推理速度。 | +| **Moirai 2.0** | 采用轻量级 Decoder-only Patch Transformer，通过单一 Patch 尺寸、多 Token 预测和多分位数输出实现高效的单变量预测。 | 适用于对模型体量和推理效率要求较高的零样本单变量预测，如工业监测、能源负荷及设备指标预测。 | 1. 模型参数量约 11.4M。
2. 每个解码步可预测多个 Patch，降低长预测范围的自回归开销。
3. 输出 9 个分位数（0.1～0.9），使用 p50 中位数作为点预测。
4. 使用实例归一化缓解序列分布漂移。
5. 不支持多变量和协变量。 | +| **Toto 2.0** | 采用 Decoder-only Patch Transformer，交替使用因果时间注意力与变量注意力，对时间维和变量维进行联合建模。 | 面向可观测性指标等多变量时间序列的零样本预测，如 CPU、内存、网络流量等指标的联合预测。 | 1. 支持单变量和多变量目标预测。
2. 输出 0.1～0.9 的固定分位数，使用 p50 中位数作为点预测。
3. 支持缓存式块解码，可高效扩展较长预测范围。
4. 当前不支持协变量。 | ### 4.4 删除模型 diff --git a/src/zh/UserGuide/Master/Tree/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md b/src/zh/UserGuide/Master/Tree/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md index e1af1e1d0..5fe5c3771 100644 --- a/src/zh/UserGuide/Master/Tree/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md +++ b/src/zh/UserGuide/Master/Tree/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md @@ -78,7 +78,34 @@ Chronos-2 [4] ![](/img/timeseries-large-model-chronos2.png) -## 7. 效果展示 +## 7. Moirai2 模型 + +Moirai2[5]（Moirai 2.0）是由 Salesforce AI Research 推出的通用时间序列基础模型（V2.0.10 及以后版本支持）。当前集成的是 Moirai 2.0 R-small 版本，模型参数量约为 11.4M。与采用掩码编码器架构的 Moirai 1.0 不同，Moirai 2.0 使用因果 Decoder-only Patch Transformer，并通过单一 Patch 尺寸、多 Token 预测和多分位数输出，在较小模型规模下实现高效的单变量预测。其核心特性包括： + +- **轻量化模型架构**：采用 Decoder-only Patch Transformer，并结合 RMSNorm、旋转位置编码和 SiLU-GLU 前馈网络，在较小参数规模下兼顾预测能力和推理效率。 +- **多 Token 预测**：每个解码步骤可同时预测多个 Patch，减少长预测范围下所需的自回归解码次数。 +- **概率性预测能力**：模型输出 0.1～0.9 共 9 个分位数，AINode 使用 p50 中位数作为点预测结果。 +- **Patch 解码**：在进入注意力模块前，将时间序列按照固定大小的 Patch 进行分组，以提高时序特征提取和解码效率。 +- **实例归一化**：在模型输入前对每条时间序列进行标准化，并在输出后执行反归一化，以缓解不同序列间的分布漂移。 +- **输入范围**：模型聚焦于单变量预测，不支持多变量目标及协变量输入。 + +![](/img/LargeModel-moirai2.png) + +> 注意：Moirai 2.0 R-small 模型权重采用 CC BY-NC 4.0 许可证，仅限研究用途。 + +## 8. Toto 模型 + +Toto[6]（Toto 2.0）是由 Datadog 推出的新一代时间序列基础模型（V2.0.10 及以后版本支持），主要面向可观测性场景中的时间序列预测。当前集成的模型采用 2.5B 参数版本，基于 Decoder-only Patch Transformer 架构，交替使用因果时间注意力与变量注意力，对时间维度和变量维度进行联合建模。其核心特性包括： + +- **单变量与多变量预测**：支持单个目标变量以及多个相关目标变量的联合预测，适用于 CPU、内存、网络流量等可观测性指标的分析场景。 +- **概率性预测能力**：模型输出 0.1～0.9 的固定分位数，可刻画预测结果的不确定性；AINode 使用 p50 中位数作为点预测结果。 +- **高效块解码**：采用缓存式块解码机制，可分块生成预测结果，降低较长预测范围下的重复计算开销。 +- **大规模模型能力**：通过扩大模型参数量和预训练数据规模提升预测性能，具备良好的零样本泛化能力。 +- **协变量限制**：当前集成版本不支持历史协变量或已知未来协变量。 + +![](/img/LargeModel-toto.png) + +## 9. 效果展示 时序大模型能够适应多种不同领域和场景的真实时序数据，在各种任务上拥有优异的处理效果，以下是在不同数据上的真实表现： @@ -100,7 +127,7 @@ Chronos-2 [4] ![](/img/LargeModel05.png) -## 8. 部署使用 +## 10. 部署使用 1. 打开 IoTDB cli 控制台，检查 ConfigNode、DataNode、AINode 节点确保均为 Running。 @@ -143,6 +170,8 @@ IoTDB> show models | timer_xl| timer| builtin| active| | sundial| sundial| builtin| active| | chronos2| t5| builtin| active| +| moirai2| moirai| builtin| active| +| toto| toto| builtin| active| +---------------------+---------+--------+--------+ ``` @@ -150,8 +179,12 @@ IoTDB> show models **[1]** Timer- Generative Pre-trained Transformers Are Large Time Series Models, Yong Liu, Haoran Zhang, Chenyu Li, Xiangdong Huang, Jianmin Wang, Mingsheng Long. [↩ 返回](#ref1) -**[2]** TIMER-XL- LONG-CONTEXT TRANSFORMERS FOR UNIFIED TIME SERIES FORECASTING ,Yong Liu, Guo Qin, Xiangdong Huang, Jianmin Wang, Mingsheng Long. [↩ 返回](#ref2) +**[2]** TIMER-XL: LONG-CONTEXT TRANSFORMERS FOR UNIFIED TIME SERIES FORECASTING, Yong Liu, Guo Qin, Xiangdong Huang, Jianmin Wang, Mingsheng Long. [↩ 返回](#ref2) + +**[3]** Sundial: A Family of Highly Capable Time Series Foundation Models, Yong Liu, Guo Qin, Zhiyuan Shi, Zhi Chen, Caiyin Yang, Xiangdong Huang, Jianmin Wang, Mingsheng Long, **ICML 2025 spotlight**. [↩ 返回](#ref3) + +**[4]** Chronos-2: From Univariate to Universal Forecasting, Abdul Fatir Ansari, Oleksandr Shchur, Jaris Küken, Andreas Auer, Boran Han, Pedro Mercado, Syama Sundar Rangapuram, Huibin Shen, Lorenzo Stella, Xiyuan Zhang, Mononito Goswami, Shubham Kapoor, Danielle C. Maddix, Pablo Guerron, Tony Hu, Junming Yin, Nick Erickson, Prateek Mutalik Desai, Hao Wang, Huzefa Rangwala, George Karypis, Yuyang Wang, Michael Bohlke-Schneider, **arXiv:2510.15821.** [↩ 返回](#ref4) -**[3]** Sundial- A Family of Highly Capable Time Series Foundation Models, Yong Liu, Guo Qin, Zhiyuan Shi, Zhi Chen, Caiyin Yang, Xiangdong Huang, Jianmin Wang, Mingsheng Long, **ICML 2025 spotlight**. [↩ 返回](#ref3) +[5] Moirai 2.0: When Less Is More for Time Series Forecasting, Salesforce AI Research, arXiv:2511.11698. [↩ 返回](#ref5) -**[4] **Chronos-2: From Univariate to Universal Forecasting, Abdul Fatir Ansari, Oleksandr Shchur, Jaris Küken, Andreas Auer, Boran Han, Pedro Mercado, Syama Sundar Rangapuram, Huibin Shen, Lorenzo Stella, Xiyuan Zhang, Mononito Goswami, Shubham Kapoor, Danielle C. Maddix, Pablo Guerron, Tony Hu, Junming Yin, Nick Erickson, Prateek Mutalik Desai, Hao Wang, Huzefa Rangwala, George Karypis, Yuyang Wang, Michael Bohlke-Schneider, **arXiv:2510.15821.**[↩ 返回](#ref4) +[6] Toto 2.0: Time Series Forecasting Enters the Scaling Era, Datadog, arXiv:2605.20119. [↩ 返回](#ref6) diff --git a/src/zh/UserGuide/latest-Table/AI-capability/AINode_Upgrade_apache.md b/src/zh/UserGuide/latest-Table/AI-capability/AINode_Upgrade_apache.md index 6356a0489..915ef8691 100644 --- a/src/zh/UserGuide/latest-Table/AI-capability/AINode_Upgrade_apache.md +++ b/src/zh/UserGuide/latest-Table/AI-capability/AINode_Upgrade_apache.md @@ -443,7 +443,12 @@ IoTDB> show models | custom| | user_defined| active| | timer_xl| timer| builtin| activating| | sundial| sundial| builtin| active| +| sundialx_1| sundial| fine_tuned| active| +| sundialx_4| sundial| fine_tuned| training| +| sundialx_5| sundial| fine_tuned| failed| | chronos2| t5| builtin| inactive| +| moirai2| moirai| builtin| inactive| +| toto| toto| builtin| inactive| +---------------------+--------------+--------------+-------------+ ``` @@ -451,22 +456,24 @@ IoTDB> show models | 模型名称 | 核心概念 | 适用场景 | 主要特点 | |----------------------------------| ----------------------------------------------------------------------------------------- | ---------------------------------------------------------------- | ------------------------------------------------------------------------------------------------- | -| **ARIMA**(自回归整合移动平均模型) | 结合自回归(AR)、差分(I)和移动平均(MA)，用于预测平稳时间序列或可通过差分变为平稳的数据。 | 单变量时间序列预测，如股票价格、销量、经济指标等。| 1. 适用于线性趋势和季节性较弱的数据。2. 需要选择参数 (p,d,q)。3. 对缺失值敏感。 | -| **Holt-Winters**(三参数指数平滑) | 基于指数平滑，引入水平、趋势和季节性三个分量，适用于具有趋势和季节性的数据。 | 有明显季节性和趋势的时间序列，如月度销售额、电力需求等。 | 1. 可处理加性或乘性季节性。2. 对近期数据赋予更高权重。3. 简单易实现。 | -| **Exponential Smoothing**(指数平滑) | 通过加权平均历史数据，权重随时间指数递减，强调近期观测值的重要性。 | 无显著季节性但存在趋势的数据，如短期需求预测。 | 1. 参数少，计算简单。2. 适合平稳或缓慢变化序列。3. 可扩展为双指数或三指数平滑。 | -| **Naive Forecaster**(朴素预测器) | 使用最近一期的观测值作为下一期的预测值，是最简单的基准模型。 | 作为其他模型的比较基准，或数据无明显模式时的简单预测。 | 1. 无需训练。2. 对突发变化敏感。3. 季节性朴素变体可用前一季节同期值预测。 | -| **STL Forecaster**(季节趋势分解预测) | 基于STL分解时间序列，分别预测趋势、季节性和残差分量后组合。 | 具有复杂季节性、趋势和非线性模式的数据，如气候数据、交通流量。 | 1. 能处理非固定季节性。2. 对异常值稳健。3. 分解后可结合其他模型预测各分量。 | -| **Gaussian HMM**(高斯隐马尔可夫模型) | 假设观测数据由隐藏状态生成，每个状态的观测概率服从高斯分布。 | 状态序列预测或分类，如语音识别、金融状态识别。 | 1. 适用于时序数据的状态建模。2. 假设观测值在给定状态下独立。3. 需指定隐藏状态数量。 | -| **GMM HMM** (高斯混合隐马尔可夫模型) | 扩展Gaussian HMM，每个状态的观测概率由高斯混合模型描述，可捕捉更复杂的观测分布。 | 需要多模态观测分布的场景，如复杂动作识别、生物信号分析。 | 1. 比单一高斯更灵活。2. 参数更多，计算复杂度高。3. 需训练GMM成分数。 | -| **STRAY**(基于奇异值的异常检测) | 通过奇异值分解(SVD)检测高维数据中的异常点，常用于时间序列异常检测。 | 高维时间序列的异常检测，如传感器网络、IT系统监控。 | 1. 无需分布假设。2. 可处理高维数据。3. 对全局异常敏感，局部异常可能漏检。 | +| **ARIMA**(自回归整合移动平均模型) | 结合自回归(AR)、差分(I)和移动平均(MA)，用于预测平稳时间序列或可通过差分变为平稳的数据。 | 单变量时间序列预测，如股票价格、销量、经济指标等。| 1. 适用于线性趋势和季节性较弱的数据。
2. 需要选择参数 (p,d,q)。
3. 对缺失值敏感。 | +| **Holt-Winters**(三参数指数平滑) | 基于指数平滑，引入水平、趋势和季节性三个分量，适用于具有趋势和季节性的数据。 | 有明显季节性和趋势的时间序列，如月度销售额、电力需求等。 | 1. 可处理加性或乘性季节性。
2. 对近期数据赋予更高权重。
3. 简单易实现。 | +| **Exponential Smoothing**(指数平滑) | 通过加权平均历史数据，权重随时间指数递减，强调近期观测值的重要性。 | 无显著季节性但存在趋势的数据，如短期需求预测。 | 1. 参数少，计算简单。
2. 适合平稳或缓慢变化序列。
3. 可扩展为双指数或三指数平滑。 | +| **Naive Forecaster**(朴素预测器) | 使用最近一期的观测值作为下一期的预测值，是最简单的基准模型。 | 作为其他模型的比较基准，或数据无明显模式时的简单预测。 | 1. 无需训练。
2. 对突发变化敏感。
3. 季节性朴素变体可用前一季节同期值预测。 | +| **STL Forecaster**(季节趋势分解预测) | 基于STL分解时间序列，分别预测趋势、季节性和残差分量后组合。 | 具有复杂季节性、趋势和非线性模式的数据，如气候数据、交通流量。 | 1. 能处理非固定季节性。
2. 对异常值稳健。
3. 分解后可结合其他模型预测各分量。 | +| **Gaussian HMM**(高斯隐马尔可夫模型) | 假设观测数据由隐藏状态生成，每个状态的观测概率服从高斯分布。 | 状态序列预测或分类，如语音识别、金融状态识别。 | 1. 适用于时序数据的状态建模。
2. 假设观测值在给定状态下独立。
3. 需指定隐藏状态数量。 | +| **GMM HMM** (高斯混合隐马尔可夫模型) | 扩展Gaussian HMM，每个状态的观测概率由高斯混合模型描述，可捕捉更复杂的观测分布。 | 需要多模态观测分布的场景，如复杂动作识别、生物信号分析。 | 1. 比单一高斯更灵活。
2. 参数更多，计算复杂度高。
3. 需训练GMM成分数。 | +| **STRAY**(基于奇异值的异常检测) | 通过奇异值分解(SVD)检测高维数据中的异常点，常用于时间序列异常检测。 | 高维时间序列的异常检测，如传感器网络、IT系统监控。 | 1. 无需分布假设。
2. 可处理高维数据。
3. 对全局异常敏感，局部异常可能漏检。 | 内置时序大模型介绍如下： | 模型名称 | 核心概念 | 适用场景 | 主要特点 | |---------------| ---------------------------------------------------------------------- | ------------------------------------------------------------------------ | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | -| **Timer-XL** | 支持超长上下文的时序大模型，通过大规模工业数据预训练增强泛化能力。 | 需利用极长历史数据的复杂工业预测，如能源、航空航天、交通等领域。 | 1. 超长上下文支持，可处理数万时间点输入。2. 多场景覆盖，支持非平稳、多变量及协变量预测。3. 基于万亿级高质量工业时序数据预训练。 | -| **Timer-Sundial** | 采用“Transformer + TimeFlow”架构的生成式基础模型，专注于概率预测。 | 需要量化不确定性的零样本预测场景，如金融、供应链、新能源发电预测。 | 1. 强大的零样本泛化能力，支持点预测与概率预测 2. 可灵活分析预测分布的任意统计特性。3. 创新生成架构，实现高效的非确定性样本生成。 | -| **Chronos-2** | 基于离散词元化范式的通用时序基础模型，将预测转化为语言建模任务。 | 快速零样本单变量预测，以及可借助协变量（如促销、天气）提升效果的场景。 | 1. 强大的零样本概率预测能力。2. 支持协变量统一建模，但对输入有严格要求：a. 未来协变量的名称组成的集合必须是历史协变量的名称组成的集合的子集；b. 每个历史协变量的长度必须等于目标变量的长度； c. 每个未来协变量的长度必须等于预测长度；3. 采用高效的编码器式结构，兼顾性能与推理速度。 | +| **Timer-XL** | 支持超长上下文的时序大模型，通过大规模工业数据预训练增强泛化能力。 | 需利用极长历史数据的复杂工业预测，如能源、航空航天、交通等领域。 | 1. 超长上下文支持，可处理数万时间点输入。
2. 多场景覆盖，支持非平稳、多变量及协变量预测。
3. 基于万亿级高质量工业时序数据预训练。 | +| **Timer-Sundial** | 采用“Transformer + TimeFlow”架构的生成式基础模型，专注于概率预测。 | 需要量化不确定性的零样本预测场景，如金融、供应链、新能源发电预测。 | 1. 强大的零样本泛化能力，支持点预测与概率预测
2. 可灵活分析预测分布的任意统计特性。
3. 创新生成架构，实现高效的非确定性样本生成。 | +| **Chronos-2** | 基于离散词元化范式的通用时序基础模型，将预测转化为语言建模任务。 | 快速零样本单变量预测，以及可借助协变量（如促销、天气）提升效果的场景。 | 1. 强大的零样本概率预测能力。
2. 支持协变量统一建模，但对输入有严格要求：
  a. 未来协变量的名称组成的集合必须是历史协变量的名称组成的集合的子集；
  b. 每个历史协变量的长度必须等于目标变量的长度；
  c. 每个未来协变量的长度必须等于预测长度；
3. 采用高效的编码器式结构，兼顾性能与推理速度。 | +| **Moirai 2.0** | 采用轻量级 Decoder-only Patch Transformer，通过单一 Patch 尺寸、多 Token 预测和多分位数输出实现高效的单变量预测。 | 适用于对模型体量和推理效率要求较高的零样本单变量预测，如工业监测、能源负荷及设备指标预测。 | 1. 模型参数量约 11.4M。
2. 每个解码步可预测多个 Patch，降低长预测范围的自回归开销。
3. 输出 9 个分位数（0.1～0.9），使用 p50 中位数作为点预测。
4. 使用实例归一化缓解序列分布漂移。
5. 不支持多变量和协变量。 | +| **Toto 2.0** | 采用 Decoder-only Patch Transformer，交替使用因果时间注意力与变量注意力，对时间维和变量维进行联合建模。 | 面向可观测性指标等多变量时间序列的零样本预测，如 CPU、内存、网络流量等指标的联合预测。 | 1. 支持单变量和多变量目标预测。
2. 输出 0.1～0.9 的固定分位数，使用 p50 中位数作为点预测。
3. 支持缓存式块解码，可高效扩展较长预测范围。
4. 当前不支持协变量。 | ### 4.4 删除模型 diff --git a/src/zh/UserGuide/latest-Table/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md b/src/zh/UserGuide/latest-Table/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md index 99712b134..0a75fce5b 100644 --- a/src/zh/UserGuide/latest-Table/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md +++ b/src/zh/UserGuide/latest-Table/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md @@ -78,7 +78,34 @@ Chronos-2 [4] ![](/img/timeseries-large-model-chronos2.png) -## 7. 效果展示 +## 7. Moirai2 模型 + +Moirai2[5]（Moirai 2.0）是由 Salesforce AI Research 推出的通用时间序列基础模型（V2.0.10 及以后版本支持）。当前集成的是 Moirai 2.0 R-small 版本，模型参数量约为 11.4M。与采用掩码编码器架构的 Moirai 1.0 不同，Moirai 2.0 使用因果 Decoder-only Patch Transformer，并通过单一 Patch 尺寸、多 Token 预测和多分位数输出，在较小模型规模下实现高效的单变量预测。其核心特性包括： + +- **轻量化模型架构**：采用 Decoder-only Patch Transformer，并结合 RMSNorm、旋转位置编码和 SiLU-GLU 前馈网络，在较小参数规模下兼顾预测能力和推理效率。 +- **多 Token 预测**：每个解码步骤可同时预测多个 Patch，减少长预测范围下所需的自回归解码次数。 +- **概率性预测能力**：模型输出 0.1～0.9 共 9 个分位数，AINode 使用 p50 中位数作为点预测结果。 +- **Patch 解码**：在进入注意力模块前，将时间序列按照固定大小的 Patch 进行分组，以提高时序特征提取和解码效率。 +- **实例归一化**：在模型输入前对每条时间序列进行标准化，并在输出后执行反归一化，以缓解不同序列间的分布漂移。 +- **输入范围**：模型聚焦于单变量预测，不支持多变量目标及协变量输入。 + +![](/img/LargeModel-moirai2.png) + +> 注意：Moirai 2.0 R-small 模型权重采用 CC BY-NC 4.0 许可证，仅限研究用途。 + +## 8. Toto 模型 + +Toto[6]（Toto 2.0）是由 Datadog 推出的新一代时间序列基础模型（V2.0.10 及以后版本支持），主要面向可观测性场景中的时间序列预测。当前集成的模型采用 2.5B 参数版本，基于 Decoder-only Patch Transformer 架构，交替使用因果时间注意力与变量注意力，对时间维度和变量维度进行联合建模。其核心特性包括： + +- **单变量与多变量预测**：支持单个目标变量以及多个相关目标变量的联合预测，适用于 CPU、内存、网络流量等可观测性指标的分析场景。 +- **概率性预测能力**：模型输出 0.1～0.9 的固定分位数，可刻画预测结果的不确定性；AINode 使用 p50 中位数作为点预测结果。 +- **高效块解码**：采用缓存式块解码机制，可分块生成预测结果，降低较长预测范围下的重复计算开销。 +- **大规模模型能力**：通过扩大模型参数量和预训练数据规模提升预测性能，具备良好的零样本泛化能力。 +- **协变量限制**：当前集成版本不支持历史协变量或已知未来协变量。 + +![](/img/LargeModel-toto.png) + +## 9. 效果展示 时序大模型能够适应多种不同领域和场景的真实时序数据，在各种任务上拥有优异的处理效果，以下是在不同数据上的真实表现： @@ -100,7 +127,7 @@ Chronos-2 [4] ![](/img/LargeModel05.png) -## 8. 部署使用 +## 10. 部署使用 1. 打开 IoTDB cli 控制台，检查 ConfigNode、DataNode、AINode 节点确保均为 Running。 @@ -143,6 +170,8 @@ IoTDB> show models | timer_xl| timer| builtin| active| | sundial| sundial| builtin| active| | chronos2| t5| builtin| active| +| moirai2| moirai| builtin| active| +| toto| toto| builtin| active| +---------------------+---------+--------+--------+ ``` @@ -150,8 +179,12 @@ IoTDB> show models **[1]** Timer- Generative Pre-trained Transformers Are Large Time Series Models, Yong Liu, Haoran Zhang, Chenyu Li, Xiangdong Huang, Jianmin Wang, Mingsheng Long. [↩ 返回](#ref1) -**[2]** TIMER-XL- LONG-CONTEXT TRANSFORMERS FOR UNIFIED TIME SERIES FORECASTING ,Yong Liu, Guo Qin, Xiangdong Huang, Jianmin Wang, Mingsheng Long. [↩ 返回](#ref2) +**[2]** TIMER-XL: LONG-CONTEXT TRANSFORMERS FOR UNIFIED TIME SERIES FORECASTING, Yong Liu, Guo Qin, Xiangdong Huang, Jianmin Wang, Mingsheng Long. [↩ 返回](#ref2) + +**[3]** Sundial: A Family of Highly Capable Time Series Foundation Models, Yong Liu, Guo Qin, Zhiyuan Shi, Zhi Chen, Caiyin Yang, Xiangdong Huang, Jianmin Wang, Mingsheng Long, **ICML 2025 spotlight**. [↩ 返回](#ref3) + +**[4]** Chronos-2: From Univariate to Universal Forecasting, Abdul Fatir Ansari, Oleksandr Shchur, Jaris Küken, Andreas Auer, Boran Han, Pedro Mercado, Syama Sundar Rangapuram, Huibin Shen, Lorenzo Stella, Xiyuan Zhang, Mononito Goswami, Shubham Kapoor, Danielle C. Maddix, Pablo Guerron, Tony Hu, Junming Yin, Nick Erickson, Prateek Mutalik Desai, Hao Wang, Huzefa Rangwala, George Karypis, Yuyang Wang, Michael Bohlke-Schneider, **arXiv:2510.15821.** [↩ 返回](#ref4) -**[3]** Sundial- A Family of Highly Capable Time Series Foundation Models, Yong Liu, Guo Qin, Zhiyuan Shi, Zhi Chen, Caiyin Yang, Xiangdong Huang, Jianmin Wang, Mingsheng Long, **ICML 2025 spotlight**. [↩ 返回](#ref3) +[5] Moirai 2.0: When Less Is More for Time Series Forecasting, Salesforce AI Research, arXiv:2511.11698. [↩ 返回](#ref5) -**[4] **Chronos-2: From Univariate to Universal Forecasting, Abdul Fatir Ansari, Oleksandr Shchur, Jaris Küken, Andreas Auer, Boran Han, Pedro Mercado, Syama Sundar Rangapuram, Huibin Shen, Lorenzo Stella, Xiyuan Zhang, Mononito Goswami, Shubham Kapoor, Danielle C. Maddix, Pablo Guerron, Tony Hu, Junming Yin, Nick Erickson, Prateek Mutalik Desai, Hao Wang, Huzefa Rangwala, George Karypis, Yuyang Wang, Michael Bohlke-Schneider, **arXiv:2510.15821.**[↩ 返回](#ref4) +[6] Toto 2.0: Time Series Forecasting Enters the Scaling Era, Datadog, arXiv:2605.20119. [↩ 返回](#ref6) diff --git a/src/zh/UserGuide/latest/AI-capability/AINode_Upgrade_apache.md b/src/zh/UserGuide/latest/AI-capability/AINode_Upgrade_apache.md index a86a0ea7c..255eaff80 100644 --- a/src/zh/UserGuide/latest/AI-capability/AINode_Upgrade_apache.md +++ b/src/zh/UserGuide/latest/AI-capability/AINode_Upgrade_apache.md @@ -419,7 +419,12 @@ IoTDB> show models | custom| | user_defined| active| | timer_xl| timer| builtin| activating| | sundial| sundial| builtin| active| +| sundialx_1| sundial| fine_tuned| active| +| sundialx_4| sundial| fine_tuned| training| +| sundialx_5| sundial| fine_tuned| failed| | chronos2| t5| builtin| inactive| +| moirai2| moirai| builtin| inactive| +| toto| toto| builtin| inactive| +---------------------+--------------+--------------+-------------+ ``` @@ -427,22 +432,24 @@ IoTDB> show models | 模型名称 | 核心概念 | 适用场景 | 主要特点 | |----------------------------------| ----------------------------------------------------------------------------------------- | ---------------------------------------------------------------- | ------------------------------------------------------------------------------------------------- | -| **ARIMA**(自回归整合移动平均模型) | 结合自回归(AR)、差分(I)和移动平均(MA)，用于预测平稳时间序列或可通过差分变为平稳的数据。 | 单变量时间序列预测，如股票价格、销量、经济指标等。| 1. 适用于线性趋势和季节性较弱的数据。2. 需要选择参数 (p,d,q)。3. 对缺失值敏感。 | -| **Holt-Winters**(三参数指数平滑) | 基于指数平滑，引入水平、趋势和季节性三个分量，适用于具有趋势和季节性的数据。 | 有明显季节性和趋势的时间序列，如月度销售额、电力需求等。 | 1. 可处理加性或乘性季节性。2. 对近期数据赋予更高权重。3. 简单易实现。 | -| **Exponential Smoothing**(指数平滑) | 通过加权平均历史数据，权重随时间指数递减，强调近期观测值的重要性。 | 无显著季节性但存在趋势的数据，如短期需求预测。 | 1. 参数少，计算简单。2. 适合平稳或缓慢变化序列。3. 可扩展为双指数或三指数平滑。 | -| **Naive Forecaster**(朴素预测器) | 使用最近一期的观测值作为下一期的预测值，是最简单的基准模型。 | 作为其他模型的比较基准，或数据无明显模式时的简单预测。 | 1. 无需训练。2. 对突发变化敏感。3. 季节性朴素变体可用前一季节同期值预测。 | -| **STL Forecaster**(季节趋势分解预测) | 基于STL分解时间序列，分别预测趋势、季节性和残差分量后组合。 | 具有复杂季节性、趋势和非线性模式的数据，如气候数据、交通流量。 | 1. 能处理非固定季节性。2. 对异常值稳健。3. 分解后可结合其他模型预测各分量。 | -| **Gaussian HMM**(高斯隐马尔可夫模型) | 假设观测数据由隐藏状态生成，每个状态的观测概率服从高斯分布。 | 状态序列预测或分类，如语音识别、金融状态识别。 | 1. 适用于时序数据的状态建模。2. 假设观测值在给定状态下独立。3. 需指定隐藏状态数量。 | -| **GMM HMM** (高斯混合隐马尔可夫模型) | 扩展Gaussian HMM，每个状态的观测概率由高斯混合模型描述，可捕捉更复杂的观测分布。 | 需要多模态观测分布的场景，如复杂动作识别、生物信号分析。 | 1. 比单一高斯更灵活。2. 参数更多，计算复杂度高。3. 需训练GMM成分数。 | -| **STRAY**(基于奇异值的异常检测) | 通过奇异值分解(SVD)检测高维数据中的异常点，常用于时间序列异常检测。 | 高维时间序列的异常检测，如传感器网络、IT系统监控。 | 1. 无需分布假设。2. 可处理高维数据。3. 对全局异常敏感，局部异常可能漏检。 | +| **ARIMA**(自回归整合移动平均模型) | 结合自回归(AR)、差分(I)和移动平均(MA)，用于预测平稳时间序列或可通过差分变为平稳的数据。 | 单变量时间序列预测，如股票价格、销量、经济指标等。| 1. 适用于线性趋势和季节性较弱的数据。
2. 需要选择参数 (p,d,q)。
3. 对缺失值敏感。 | +| **Holt-Winters**(三参数指数平滑) | 基于指数平滑，引入水平、趋势和季节性三个分量，适用于具有趋势和季节性的数据。 | 有明显季节性和趋势的时间序列，如月度销售额、电力需求等。 | 1. 可处理加性或乘性季节性。
2. 对近期数据赋予更高权重。
3. 简单易实现。 | +| **Exponential Smoothing**(指数平滑) | 通过加权平均历史数据，权重随时间指数递减，强调近期观测值的重要性。 | 无显著季节性但存在趋势的数据，如短期需求预测。 | 1. 参数少，计算简单。
2. 适合平稳或缓慢变化序列。
3. 可扩展为双指数或三指数平滑。 | +| **Naive Forecaster**(朴素预测器) | 使用最近一期的观测值作为下一期的预测值，是最简单的基准模型。 | 作为其他模型的比较基准，或数据无明显模式时的简单预测。 | 1. 无需训练。
2. 对突发变化敏感。
3. 季节性朴素变体可用前一季节同期值预测。 | +| **STL Forecaster**(季节趋势分解预测) | 基于STL分解时间序列，分别预测趋势、季节性和残差分量后组合。 | 具有复杂季节性、趋势和非线性模式的数据，如气候数据、交通流量。 | 1. 能处理非固定季节性。
2. 对异常值稳健。
3. 分解后可结合其他模型预测各分量。 | +| **Gaussian HMM**(高斯隐马尔可夫模型) | 假设观测数据由隐藏状态生成，每个状态的观测概率服从高斯分布。 | 状态序列预测或分类，如语音识别、金融状态识别。 | 1. 适用于时序数据的状态建模。
2. 假设观测值在给定状态下独立。
3. 需指定隐藏状态数量。 | +| **GMM HMM** (高斯混合隐马尔可夫模型) | 扩展Gaussian HMM，每个状态的观测概率由高斯混合模型描述，可捕捉更复杂的观测分布。 | 需要多模态观测分布的场景，如复杂动作识别、生物信号分析。 | 1. 比单一高斯更灵活。
2. 参数更多，计算复杂度高。
3. 需训练GMM成分数。 | +| **STRAY**(基于奇异值的异常检测) | 通过奇异值分解(SVD)检测高维数据中的异常点，常用于时间序列异常检测。 | 高维时间序列的异常检测，如传感器网络、IT系统监控。 | 1. 无需分布假设。
2. 可处理高维数据。
3. 对全局异常敏感，局部异常可能漏检。 | 内置时序大模型介绍如下： | 模型名称 | 核心概念 | 适用场景 | 主要特点 | |---------------| ---------------------------------------------------------------------- | ------------------------------------------------------------------------ | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | -| **Timer-XL** | 支持超长上下文的时序大模型，通过大规模工业数据预训练增强泛化能力。 | 需利用极长历史数据的复杂工业预测，如能源、航空航天、交通等领域。 | 1. 超长上下文支持，可处理数万时间点输入。2. 多场景覆盖，支持非平稳、多变量及协变量预测。3. 基于万亿级高质量工业时序数据预训练。 | -| **Timer-Sundial** | 采用“Transformer + TimeFlow”架构的生成式基础模型，专注于概率预测。 | 需要量化不确定性的零样本预测场景，如金融、供应链、新能源发电预测。 | 1. 强大的零样本泛化能力，支持点预测与概率预测 2. 可灵活分析预测分布的任意统计特性。3. 创新生成架构，实现高效的非确定性样本生成。 | -| **Chronos-2** | 基于离散词元化范式的通用时序基础模型，将预测转化为语言建模任务。 | 快速零样本单变量预测，以及可借助协变量（如促销、天气）提升效果的场景。 | 1. 强大的零样本概率预测能力。2. 支持协变量统一建模，但对输入有严格要求：a. 未来协变量的名称组成的集合必须是历史协变量的名称组成的集合的子集；b. 每个历史协变量的长度必须等于目标变量的长度； c. 每个未来协变量的长度必须等于预测长度；3. 采用高效的编码器式结构，兼顾性能与推理速度。 | +| **Timer-XL** | 支持超长上下文的时序大模型，通过大规模工业数据预训练增强泛化能力。 | 需利用极长历史数据的复杂工业预测，如能源、航空航天、交通等领域。 | 1. 超长上下文支持，可处理数万时间点输入。
2. 多场景覆盖，支持非平稳、多变量及协变量预测。
3. 基于万亿级高质量工业时序数据预训练。 | +| **Timer-Sundial** | 采用“Transformer + TimeFlow”架构的生成式基础模型，专注于概率预测。 | 需要量化不确定性的零样本预测场景，如金融、供应链、新能源发电预测。 | 1. 强大的零样本泛化能力，支持点预测与概率预测
2. 可灵活分析预测分布的任意统计特性。
3. 创新生成架构，实现高效的非确定性样本生成。 | +| **Chronos-2** | 基于离散词元化范式的通用时序基础模型，将预测转化为语言建模任务。 | 快速零样本单变量预测，以及可借助协变量（如促销、天气）提升效果的场景。 | 1. 强大的零样本概率预测能力。
2. 支持协变量统一建模，但对输入有严格要求：
  a. 未来协变量的名称组成的集合必须是历史协变量的名称组成的集合的子集；
  b. 每个历史协变量的长度必须等于目标变量的长度；
  c. 每个未来协变量的长度必须等于预测长度；
3. 采用高效的编码器式结构，兼顾性能与推理速度。 | +| **Moirai 2.0** | 采用轻量级 Decoder-only Patch Transformer，通过单一 Patch 尺寸、多 Token 预测和多分位数输出实现高效的单变量预测。 | 适用于对模型体量和推理效率要求较高的零样本单变量预测，如工业监测、能源负荷及设备指标预测。 | 1. 模型参数量约 11.4M。
2. 每个解码步可预测多个 Patch，降低长预测范围的自回归开销。
3. 输出 9 个分位数（0.1～0.9），使用 p50 中位数作为点预测。
4. 使用实例归一化缓解序列分布漂移。
5. 不支持多变量和协变量。 | +| **Toto 2.0** | 采用 Decoder-only Patch Transformer，交替使用因果时间注意力与变量注意力，对时间维和变量维进行联合建模。 | 面向可观测性指标等多变量时间序列的零样本预测，如 CPU、内存、网络流量等指标的联合预测。 | 1. 支持单变量和多变量目标预测。
2. 输出 0.1～0.9 的固定分位数，使用 p50 中位数作为点预测。
3. 支持缓存式块解码，可高效扩展较长预测范围。
4. 当前不支持协变量。 | ### 4.4 删除模型 diff --git a/src/zh/UserGuide/latest/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md b/src/zh/UserGuide/latest/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md index 99712b134..0a75fce5b 100644 --- a/src/zh/UserGuide/latest/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md +++ b/src/zh/UserGuide/latest/AI-capability/TimeSeries-Large-Model_Upgrade_apache.md @@ -78,7 +78,34 @@ Chronos-2 [4] ![](/img/timeseries-large-model-chronos2.png) -## 7. 效果展示 +## 7. Moirai2 模型 + +Moirai2[5]（Moirai 2.0）是由 Salesforce AI Research 推出的通用时间序列基础模型（V2.0.10 及以后版本支持）。当前集成的是 Moirai 2.0 R-small 版本，模型参数量约为 11.4M。与采用掩码编码器架构的 Moirai 1.0 不同，Moirai 2.0 使用因果 Decoder-only Patch Transformer，并通过单一 Patch 尺寸、多 Token 预测和多分位数输出，在较小模型规模下实现高效的单变量预测。其核心特性包括： + +- **轻量化模型架构**：采用 Decoder-only Patch Transformer，并结合 RMSNorm、旋转位置编码和 SiLU-GLU 前馈网络，在较小参数规模下兼顾预测能力和推理效率。 +- **多 Token 预测**：每个解码步骤可同时预测多个 Patch，减少长预测范围下所需的自回归解码次数。 +- **概率性预测能力**：模型输出 0.1～0.9 共 9 个分位数，AINode 使用 p50 中位数作为点预测结果。 +- **Patch 解码**：在进入注意力模块前，将时间序列按照固定大小的 Patch 进行分组，以提高时序特征提取和解码效率。 +- **实例归一化**：在模型输入前对每条时间序列进行标准化，并在输出后执行反归一化，以缓解不同序列间的分布漂移。 +- **输入范围**：模型聚焦于单变量预测，不支持多变量目标及协变量输入。 + +![](/img/LargeModel-moirai2.png) + +> 注意：Moirai 2.0 R-small 模型权重采用 CC BY-NC 4.0 许可证，仅限研究用途。 + +## 8. Toto 模型 + +Toto[6]（Toto 2.0）是由 Datadog 推出的新一代时间序列基础模型（V2.0.10 及以后版本支持），主要面向可观测性场景中的时间序列预测。当前集成的模型采用 2.5B 参数版本，基于 Decoder-only Patch Transformer 架构，交替使用因果时间注意力与变量注意力，对时间维度和变量维度进行联合建模。其核心特性包括： + +- **单变量与多变量预测**：支持单个目标变量以及多个相关目标变量的联合预测，适用于 CPU、内存、网络流量等可观测性指标的分析场景。 +- **概率性预测能力**：模型输出 0.1～0.9 的固定分位数，可刻画预测结果的不确定性；AINode 使用 p50 中位数作为点预测结果。 +- **高效块解码**：采用缓存式块解码机制，可分块生成预测结果，降低较长预测范围下的重复计算开销。 +- **大规模模型能力**：通过扩大模型参数量和预训练数据规模提升预测性能，具备良好的零样本泛化能力。 +- **协变量限制**：当前集成版本不支持历史协变量或已知未来协变量。 + +![](/img/LargeModel-toto.png) + +## 9. 效果展示 时序大模型能够适应多种不同领域和场景的真实时序数据，在各种任务上拥有优异的处理效果，以下是在不同数据上的真实表现： @@ -100,7 +127,7 @@ Chronos-2 [4] ![](/img/LargeModel05.png) -## 8. 部署使用 +## 10. 部署使用 1. 打开 IoTDB cli 控制台，检查 ConfigNode、DataNode、AINode 节点确保均为 Running。 @@ -143,6 +170,8 @@ IoTDB> show models | timer_xl| timer| builtin| active| | sundial| sundial| builtin| active| | chronos2| t5| builtin| active| +| moirai2| moirai| builtin| active| +| toto| toto| builtin| active| +---------------------+---------+--------+--------+ ``` @@ -150,8 +179,12 @@ IoTDB> show models **[1]** Timer- Generative Pre-trained Transformers Are Large Time Series Models, Yong Liu, Haoran Zhang, Chenyu Li, Xiangdong Huang, Jianmin Wang, Mingsheng Long. [↩ 返回](#ref1) -**[2]** TIMER-XL- LONG-CONTEXT TRANSFORMERS FOR UNIFIED TIME SERIES FORECASTING ,Yong Liu, Guo Qin, Xiangdong Huang, Jianmin Wang, Mingsheng Long. [↩ 返回](#ref2) +**[2]** TIMER-XL: LONG-CONTEXT TRANSFORMERS FOR UNIFIED TIME SERIES FORECASTING, Yong Liu, Guo Qin, Xiangdong Huang, Jianmin Wang, Mingsheng Long. [↩ 返回](#ref2) + +**[3]** Sundial: A Family of Highly Capable Time Series Foundation Models, Yong Liu, Guo Qin, Zhiyuan Shi, Zhi Chen, Caiyin Yang, Xiangdong Huang, Jianmin Wang, Mingsheng Long, **ICML 2025 spotlight**. [↩ 返回](#ref3) + +**[4]** Chronos-2: From Univariate to Universal Forecasting, Abdul Fatir Ansari, Oleksandr Shchur, Jaris Küken, Andreas Auer, Boran Han, Pedro Mercado, Syama Sundar Rangapuram, Huibin Shen, Lorenzo Stella, Xiyuan Zhang, Mononito Goswami, Shubham Kapoor, Danielle C. Maddix, Pablo Guerron, Tony Hu, Junming Yin, Nick Erickson, Prateek Mutalik Desai, Hao Wang, Huzefa Rangwala, George Karypis, Yuyang Wang, Michael Bohlke-Schneider, **arXiv:2510.15821.** [↩ 返回](#ref4) -**[3]** Sundial- A Family of Highly Capable Time Series Foundation Models, Yong Liu, Guo Qin, Zhiyuan Shi, Zhi Chen, Caiyin Yang, Xiangdong Huang, Jianmin Wang, Mingsheng Long, **ICML 2025 spotlight**. [↩ 返回](#ref3) +[5] Moirai 2.0: When Less Is More for Time Series Forecasting, Salesforce AI Research, arXiv:2511.11698. [↩ 返回](#ref5) -**[4] **Chronos-2: From Univariate to Universal Forecasting, Abdul Fatir Ansari, Oleksandr Shchur, Jaris Küken, Andreas Auer, Boran Han, Pedro Mercado, Syama Sundar Rangapuram, Huibin Shen, Lorenzo Stella, Xiyuan Zhang, Mononito Goswami, Shubham Kapoor, Danielle C. Maddix, Pablo Guerron, Tony Hu, Junming Yin, Nick Erickson, Prateek Mutalik Desai, Hao Wang, Huzefa Rangwala, George Karypis, Yuyang Wang, Michael Bohlke-Schneider, **arXiv:2510.15821.**[↩ 返回](#ref4) +[6] Toto 2.0: Time Series Forecasting Enters the Scaling Era, Datadog, arXiv:2605.20119. [↩ 返回](#ref6)