From b1f617aff7e351e7dee395bf8d4b880dbd088049 Mon Sep 17 00:00:00 2001 From: scijones Date: Sun, 1 Oct 2023 17:51:33 -0400 Subject: [PATCH] first pass edits to update as of EBBS chunking There are still a few places to check for further updating. --- .../Figures/chunking-ebbs-components.png | Bin 0 -> 337175 bytes ManualSource/chunking.tex | 419 +++++++++--------- 2 files changed, 214 insertions(+), 205 deletions(-) create mode 100644 ManualSource/Figures/chunking-ebbs-components.png diff --git a/ManualSource/Figures/chunking-ebbs-components.png b/ManualSource/Figures/chunking-ebbs-components.png new file mode 100644 index 0000000000000000000000000000000000000000..a58acb9ac6abe82e3161cf54b10381cabdbedd03 GIT binary patch literal 337175 zcmdSAb8uzd*FPBBwrv|7J2$q`v29x&Cmq|i`^N4h9d&Hm&gA>N?>qB*s^+h$sb9_P zs(WvpRp;!z&brI{v!j$0r4Zrp;6Ok?5M`vrRX{+XV?jV5_+X&ED1Fp{!(Tt3E-F%@ zAhpv3$6pr^79t8FARrA1@NdSDU)QjX(%LQ{AV>rM`GAf%l$(7K&8^h5TvhF@92Lx6 zjm?Z*jlXChAUzAgjC*@bv5 zj@EL9W&S)F`9+G8EtIXusQg-N-Iu``3^~L_k2O_kZ_pwp-P>m)U71 zO7fo0=C$1D<_#GV|JSIP$>MgHsnzF?#phQ1&#foMzT5qfIQaNgi~o{JBn>CO?-NU> z>v5sZABNdLr}+fAG=?b6I)e%!(toX9IA8rO2y9j`p51IT|3zG@P)x;={MS5Wvh>YS_f&W+54X3fjQ{!{QI&NH6soyr16lt~ ze$T;Sq59V(vSt6zDB^IG3yt}1r;EU~HNBS7oS_)(BJcZ?O3r`HPWFPyY}gwb78Yi$ zz_0AIdiBQ&dtp| zLEn10Kh5B>!|U$uR?9>A5-}NC)J(h6?$N7HseGnd9@>AiJDc}^y-1ViwTXP5kbVo* zSpsw{Gy+cG?nuIaG>4`}r_~lW909jb2b|CSEYfDFw&EZ1N~1aa7i*7 zM4UF8B_wW#4YeB7uZ20?PUeq@hBk5W@OH+MssAH%*{bU5YNPMp!|Ur=CNtS1;({V0 z=UBkO!6);Ci81Nal>BZKq@`gFrgQKGWl{d&^xXgw9&7OIZy!=bafDsnTaW5KhX$!QuVv^UHIy*KhWd zlB}?q(V8Nkkl($0vG;G21t96v{~z<9c@|$N%j0npZ7J36gu#)z*E0HY9yI5Xa5#Cq zE>zyU1%cN$`xCfdYAGvL2sYeizaDMyLB!J9&;SgEgfA2(|C+i=G7hiZ^SaRbJgY=5 zT~d?mAO2_al&iH^+uPgA7dcG-l4DjN$`nof5|vq$DovHz^x^dBR5;itr6;bLR!|H)}nm>v~LbMNBl`0GoU zMd|49Uj~mawR!u`^EBrA9nO?gP_Jd+!~dW2m;XC`^Z%dMqS60Bx_ro+U7It|Aogn1 zXG86N$sf8~_ItN823kn6+&}KiNY|6d*#rnwI;=T4quV184?Ef zrO=<;WpHd}gII$P`{TJ1t!_`JyT*nFr#rzs8b6qURsJ%kvpvc+XaC@Ihj&yA(hzBC z^2U>Wk%GC@FUKXwVZRE6C}##B6m;nSZ5llbkqswXFUuX}s+{jFg@7xJ*JHD%fcZ-6 z(*g<3S2q~)1c@zUh4A@EGOVVku#ch8@8lniN|g=w@?a~&WrA_sgKJ^2pkAZuee0n| zERL~1o=VTkx@_)Oe3);}!IanAxi-P%y+Wym}F_^{0|@_I^{m z-|ve_JB(lKlM)Yr<_#ty?(hFNp8>ieEfRJ@SR^3N0(WTOw0U}ceIu|vrj0LOAaNE- z808O$`u}Vl5r2k}T?3y{u#j1S83DxP;`g+uCH12b%e;0R3=9lArL(&Y*GLhcq=lO? z{~fGlA2GI>w?7UXaC;fi3=SD)>&kdFa^1zY0B31rcbrTP*&xXuw@xn##V4w=3Sf1$ z!1j#jA*cGD=@UOPvUBmBsi=-4ZcCyE-dDP%W|H^Nk%$B1(R$2w+#f$jsy-wVh1whE z!Tuoi$kr#~qEe!F2(Q`e0rw};cT43oulK+s(y0DdW{M4%ujtfTl1jt2JKjAJa!nFO z^5*F?2S+ukg*xLiqW0Y%c-%keKDu16+v@mt?tpe$yuZPE@X|$Ud5V^3QO?{!{zy&s z%Q$Fgl`uJrdv}$qGZ+8|9{Y-{4WTV9YlL&S7?1vWl}*4!0FoRQQNK-#{SfY59IN@OO!jgs zLep?2&~?TLfio`@vdVpTUWQ-Z(moKOax)GNFQB&tYcZlb1+cCroelniNnco4s#ZWC zdvYk%DqhO14zW4(K-hMz*<3>Y0w2Y8AO;&`)2$LVH=3?}FDEW8Gx*=hXo-u8XJ=o( z<^rDc;x~&wpg1hwUgz7?AewP{Y*e;lZog|?`la&Gk9>vYW;EEPoKvR_?guNX=_P_DZbpgE6E6sQ z8MoUd;15h)qrbIEwx)CvDE1DUQT!F(a|1YWPmS*4-+l{s;9UhEo*1+&#i8tsiH3a7qsG8MiM$tDbj@fpwN64zb(Jxmy>&?RD# zii_j182wiYf2>Di(CITYxlpEBh%kO!sWAtWct%K&_IIc6Tgx`*Eys|nuAD1v>YzPT zlBgw@AMWttl9+8IFT|s}-3X%lS2kDJOn1hGly?`KlckwLh$yKjLH!;YLJybIZz<>k z#fK4FBbnMO7$T+_qGn^2z~ckU!KNP^$DT!+5WZqObV%*_^ukQ!^P2ES;`+OovAz8V zx)sF>_W=g-S5n*k-wA)=9pfDOMO|-#0eJY0TQP>W0*eT53!m%rYYdMQxp<4@*F3G_LLEU}9}PV*61?~YHd*~T>haOe9R0uS zG3UK;1BPqh+kUu<4J!}Q%$G9jV9qMnjF8_(Y)15%675EUc&v4~qCtCHhOT4r+pe}C zOZ{FCJzd>$W%ZUFltSjw=Zzb9;BT0+hd*mQ4}G6)YCpMw3$h+)vY|X9_wC-?IbUrb zTm2l-2%Gx`SnKn|c;IjAw?7oodzAN1ceBcezTo2%4;aDfI$!p?!l)Iz8BT`)D9{S^ zcE{XGoi_|jBk6iv3@=b*cevJj`~q^!4k_l{V2!m z+qvIaC>Ix7$w?_eaew}sl90&s^jn*)vi#N!wV;kq!? zG^~B6v`y0;a9$LTJy{}(!)q1&l;PspyA!EmGk9XOxwBlytJy7ixSihiY3g^Zt6FC3 zLDu0qjSMoi0M2zT;LOC;&${YG#|w4JVhK2`_iV*-NgT9q3E5aEiRjfH=YvXmWP8^`g&h zhXDLIP{ORjhMP@=zgv--GEw>weFHh*Tp0r8f_U7P;SW?|EbcD?3@;Yich6>ayu|Ef z@_|G3php{xq$ix0QL~?C{px_A09|b2K07^c4=yHj5eH>)4xS&?2 z9-7ZP{S+KSptAc}=v5b0FL(*`nve2l|4%LTHE8scJNte;A`O$}^wVf-Fe~)X(7rKk z%};Id^u*{soJSVQRD5ERF$E42^5-ko=#*pTSm%3&8$pbrSLmffZx3bOXRy|D;^DiA ze$J}0I5co-RO`WJm1xXjx`*&{iw$7;v3gCx3cP3Iy`Vts?swd0ENAH^mh+=P` zsdWDHoi~l)XI#{Pcab`~X@68a0khgmjEk zv?Sl+i-(N0%U5^$N1LXY2SM#%!UnT%XF`vLu27&+IeFqKieso-pS%y+6~1$aTr}|n zJIeqxa$04yp?Ro_AFqN#$r(@*OT%8v~lN*FH9tTv}?$X0;T#&m znR!s^2ys8R-XD=vaG->o%@oTWQ3KeB{yRqsB?gSoCrH_1eArk%W&26G7Oq-mdK=fTLKlmS>9&Dm zpSmkvw(9`&mu*<4_b!N{PPg!m9Mx`TeTX-K`RPxo+$S2GO?8r^`y!sDv4OIU$S6iB z|6zUrz;jrRsBU9Nm8^zMg3a))&5EJZ9WNMi_4mNNU&c+OC7D=c8@uA_)X=YZeOCKh zI|}#c(SET2o>s$Nh;5r(>sk_zx zGa6?Qh+O1))>S?!nb8E@l*}zc4CmP!gf!_OsOsg@>MZ@P)Op3^o4FJ~)@^ebaaV85 z+JN&yFV{=B>kY*e<1ZSa&?V^l-XPh17Q^=3)*+r`gADf>g$W{L5yzIi*b99r;L9%sPbf+hxKya!sh+PqqlCK? zQ8&cqIvg|=x@YJQbOUAtV55=~O^FH8M=)NqqI~kxtjZU_8?8j#USZd!oBb}EA$kp) zYyG*nFkt9KR1%a!RVXY@K9pk zXoMl<7mKpiCsVR*H)|%S45r`^T_rG)`X@6OUuTpi=(v9GvR@Q$UGr2J_7CVYBetAO zy+{fx=*peA~Of zlCJgP`_v*9_g#Gm3be^LNo{-YH|A16TSUaLoL0C;*`d*oXC1wVm$7gN)JsS663AgY zpP$~6i?D737qEUttuJH-5mYsk#V}rB@KuuemxF$uAL!a-kNjcUSf5)vx{Pw~(afGV zmk5(WSA2+u$>LV<|C3C|k6qn_QTHxk%0#D#9G7d^%xCL7P8&V`9Lf+hCtzS0Hkpg% zt=DNnr*`O@pIO}oz>I|WB%t6kW0Cqyy_A=7CEZ=3J**M!Ld*YJrE)iOk|#mi(HLwCF_z9Zi5NTDCsfR}bUhP8JK&2!_6RLj@a!A}}hKtQ9c9bq*`JtA;NheD*@v z{?-39K;K2ww?QM4*}c3=6gF6JC{|U#nU5 z*E<#q-g`m&5YNB+EOC~L#$ptKhT_7awg?;PrbMbUcEzDpXqkH>xmmyl@9PK}mBk3Cx z;>c3u{PPWts}*lJ5?-iCr5WEL%iNw16J?R~+Nx0yaPv3(pAN^!oIQ?(nD#c%gOI+w zKcyMFR~q-|g=Q3=q1eoRrgX9FsWefOZCpmS7N*U_9IzVrk6WV^@Lx?mvS|YT*me9G zRyS$tX3JUhH967ph9tvQ2s z2uf!9U9lFI^ZHZyr#&4%p$L#yf!4O%^K|7{VVb$z9vTY3#;QC{+d71ekprXXb?|bJ zAG2f1NHyV8z!gE|%}h3z%y9qk7bRr#(4Mt~jx_3Pk$dPtukk|rT}jLN zBi%|(uJjf79vrW6?#O3tHEX>#1ZQUmVKOpI!ae#F;sy_AUf;Y9AUCYUZsbV9m4IRh zE2L1s+uduf?ik}8?}M)0uztZ-B@OmZSDO$ptMUV_3m1a4TVyzBe98{xUMmj-Tz1?) zWdNXOQ?I3N5lbVC0QLSyy0NL(7I(Lt)Xr@0-wXjSXS6vE2El9DtL)^Sm54C8>D?Y9 z3Z+=vp_9S^5*PVwF7Q88vEqGPoXQkUjv^u_E>NEg(mF)q8y>%|x%*Hm;;F*!UNy~_ zLhM^;0DoNRQ4LZ5q>+n>s5uCuU?p1^4W~5V%(W09gd(|ONPsXA z_A#*JNT4t|{0;9C?a8Kuob|l(&Te;xEJMDG=y?|+;5&{yJ!~~iQZ|MejuX28UcZbu zj3`t`@V#3YKb-xYwVFmiSHmX2kq*9D|33I;;EJ~iYR(B+icT4F_yMWqY#>&zm5m7H z#&44UEH0nLiij@xAeQx00-`RleE04m7W07EH#zh*56rk=Gdv&b*i32+D#zqwQJfte zM{|-q+e7x|G2Q7vSTZ)s9~vt5_;)hY_BuQDe#kuz(nZ@yAHN6Ce^Z1fyCd2M&~wGR zaJNBHcqjGMX-lv78A6CX`vW?KH!t2|&8b`^>84g?2eWHmn$hjauGk3$gWVp`B)w_z z)dw<=clLink|8jT{NQu;y>W~dWX{g*r+%h_mUD^*5IM&4mgQ_cTu^txp-_sSAU$TR zu&}aupASaP+BnVb$^~PRB!U7QSKJxAAM+wq9g^k|$-KrkcBUR9O)!+1FxW?LeGKpR zK-=rr;E)Wg0b))>^MoC+NIvPtw#?*ev_GM>UELBX0m^FjbKs_FE4xRf%H_#YHCVDQ zNF(VC1KNmUrc`w(Ypwc$Irf{E^khfsgpukINXGl28!VM3fdbmD1&Dy%t+ZJ9^QpbVsHceS#4n}_Ki3{9(rO&VZJ~`(oz;i}QzITso|khibvRF6-BO;5 zCzHD=mDbydOYPPYYIaXaJtmjNu4kRV-P?Wn`&_FK0$a|#s||@$ZLa|djn*UD(EC2j z>hEW|jV&h;<5qSQ%4CHYDGk1mDM~TY;z9yp$+;`NW01QC74%nekCD+Z03GIi4C+^i z$Isf`7dOP?6>~ivHynP=sbH(6UM=6h0HfE`--DgxER_omOn|o}5lEWLW&lSFL%cj zB@+PQ?FgRQH+H7l&EQoLuuE?;!FilUEUEYvp!J+I!o?EwyZjWZ-7yk@ux-eP`hAL& zKySIgawuYY(hq1AE88CFgTnGnAAyg*X2=mP>5k8HUfFz&@J%}Vs_&xxuF*KA7xvI+ z&AD(*pnkIfk&xeOR^VTGrxJyOl2uZA_$FHkhT{XM25gOp?{HZVcCl@N1vRG1)DcE| z=*G2=*EDPfy2-boWUFWlnasIw%^C;z=N3rMQqORLo zV}7-=$KA;SJ>UJA2Js7SsJkq*gtc4LbI;M{xf*mB2jmob!ILZMDffE#6|D+ksZGK! zpgk*90zs_5-wSZw7ccFCCrFI`wJ2I?gBHxmWfzJEs$#K)Vq69FU4wXNTOo}#z@~mk zzPfoc!CW69G-NpXx98DLt+9bMnXIDKVFv(gFA@e(#L<+=$B<>VUC+Pyq5Eb*wCu%yx}KM>cC1LPR+b}}BAN}~-yZGSe(|iHN>fL~sql96@e7z1 zdOJh33&x#il8(g(Q%)2)@59lQm#1X*(m^K~S0 zZ(jOMB@3;G#FZ_0L&Y=}g@-XNW-Lhy?Tb++EjoIyOCX znx$QVg%a?1R62#HG=7L4spA(Lf!A?SeY zg=QliT1F~2BP6MB;M3=V;CaNoEN)3X{P8{MCQ_48ltCig^wmVGIHD3i5MCHryzYHn zF``As{hWXw6y3=g(8|D#zg5|Z}gE*FVu1_Y@T`b5k=oZ;G_{|rmPW%Ihl$lflsalo4#y| z3}=!3U}e&39e$TlPy1V;M}=#F&lxktzmS)*Maw{b#&Lxib1gtSjb44+pQ@=5QN8%U zt^;XmV0vP`>`n^&-(SidEQ3OG)E zHTW-*xLOLM+&=KZBVdY(r|9iHd5&NWC@|DNyvae>xY(oy04LbxSo&RAnP4%0(Rz)M zUoY;nF&S!<&?8@yDzyB3(f&`5eDLCNpWc&3U_-02IPxldqGWm3eSh0W`kALp?b}N~ zVeU0uG-XFs{wanok6uSS{0NfCGzbn>6#rJ!{3!(1l<=KYDd>-| zaMBAkC)`BdKbPI@{f>`Utldgogyxl87%uYkHr4W~=_PYUlejtxb~N8<1r!NH$K(01 zqaq@ZfcrkCYEA~t1pHWtE;iTR46NTq#>kbZzaNS^z8;OWA1XcR6A133uI6i|pesnl zg(w>^_8Mj*GrqQl`OBAN6M!b@ZKXzo)y&lyUhT`xsY+X#q0wPPMXCfA3#@pQ7d}iX zlWawE4(l3i)L4gw7OqBZnWcjF>Mo zGV%8^M$d$~QVslAFdn$Pk{n4quJ@+^Q(4rb53Vl(Gf*P`jADEPWndDY%mOSzHMz&W|ymk z6TzlBMs|UoG|LE;7r}5x{p|?jII=_6!{l(f-D<=gAEC0a<2p|iHFS&l=lFKC@M60- zg%7gS2=usM93b=zF=C*wAY34hO45KF9-)HTReoU0S;Xse{sCjQCK@Sc91fzi5B1P4 zSTZ#%!Dvq^z*k;k-?*?(UB~-HRUl~T_jb5W)&0EUMF9S(l)Km`Z`buEz#TLG$&nZu z8&NW$fZl0{T3;M`u6s0spT*AAHlgv->h4LQeW27h1ujfzEwsUudrL9UTShH-Z69H_ z;PPuzl?8N{F}%W*VrbUh%bw0Gqh|E(aKNpTQ)Ra#O8dX@iG{ z;t;9uL@zl$U06md0n2l+{}nzVLy*jf8@b1}Z&0rK?V8^Zl}yj+ zINbYkN5FEn4_emdQB=y}3`b+uZxp0|^0tm1y6 zL2?h|Y_&ov{V?gqF7 zq_^7_PDJH6-wfHsm8sNSRXud;_~QDA-Hi2N^FcIHXN!Dt1hR}vK3wM0i|3S39hx%8 zKNHOw)B=^FGb!UxWQ9sHZSDC7yN`yMyv=WdXN$GocQqr|B>8fcvT!33wVd{E=p3~a zDE%)lFdi2rVXTRwTh52Ovt|{nU@4nVNgZNma@E@mjy?iUex4CrRtKA9xHd)&DU3 zYKCK=2)D_f#rl1Q;>-hC(@(M@b?p<^e()|lMj^dcNU=bia`D?@nf*A|xdEKGL;Dt;YQ)ne2~bF3xX zEe%t)aKf$f@fqJ8b3uoMt95e2wLC#j{>sCxMT7k}*abjjQbMy4qwC2$Qq50piQR|Y zhDCR_A2Cy&VbH{ChrB>oN7k7Np(g=%3a zM7)$ej(k7PJ+ya1V@xvHfySkQj4&}}-l~^(>1gD~2P5E*R?D?|+!fce$kr*u*>ZJZ zd2Js&X0^+R%<=mH*JPF1iB0qIq?KO~H#YEW&uImnqgHMdNO27EFP(c32+1=>W4bBlx?QcH>jwq*J>?hY-cKz7!32t zScE=3ghbN25dLh3+>r*q)du@8$^&`4u%LCFe5YHXjGjwY2+78eU9!#I&?{P#K<9mk zk-&HVgmwQLY}_enhZ6xOg?OJA4+SY^X7CVx!gK zvE_?2E?EcXP52C9Ya+#7#f|e+FT=o+X9fbzsxwlu7q3LmZc|94?r6rN6AX@F+pahq z3q4UsErkXH%nYr%;pI+iW)5dVJwA8LpuRUU(G&r;qkX9$(J&<%I?U0S>Lj%%s45vr z!9C_L)e?pL!R@~fGt2x(WUSScb<-I(txhM#?K7h=ytQGvnv>PU)(1nwrwGoaHx{2b zz(GBnCxWQ4Vkv7(ctCjT(JtgJ6}o%=B$C3sGQUyjF$BU~GM?rmu`fHOh;NF6JvKK?_sX`(!ifu4%F_VQH^i=u4Go{vvk5z*g zax<6D^P9SdK*ok5b%#Pixh`g=iJs0);xOLMSgniz&TLVQl3}O|J}skhL&E53MNRev zv^RL4nO3OBE8OBY8o(e()Gdb1_sYLaM-eu2#`AzyuhW$l_881Pt`f~d+Ay=v`7;d6 zootU;4??F$g!Xq5#^*DNYB#CZ;DGfAxaSlCM8MdhF?TVwe`eiFIHcs*ku9~IqZdmW`|ImpY5AqkRp3>X6Tf z$4ZiG&8OF@PX=~4tjPgv2eLvniC6)*sfOo3!talj;O`J8IX@auP99nY`??N_+>}{; z--Rmt^C`xoyPq4&YJIxxy`iTXQkeOTJ`VG+r-#QUfG!4=;g zj8*+rmNi4ml&v(BwP#0(QPw|Sw}h(@&rr25Q;E# zHY~Eb!}-HJ#~4$I*);SJtmHeadzruKGf7FQL^ci4#^%{M#1;6We#7mS1^__kldr89mEVI-Hh)NfG99r z@YarGCzsPW_qIntoxLncZiDik+K5h$r*5!2me46ba-r(1D){OHTN!`}e|^(2NDJ8R~jX3T#^_gC)4 z<>3;=(H6WoBxo?k`g+C`8G>qSm5$}2M`7$HRGh`dVF>L|2Rn5nzYS&PGfclzif%RA zLr5|RMN{XY!=c8IjQzxZ*|*NOqE?mHZGtVbDejGI(DI6du}|edTgfBajzpmojd9JV zH;#-tasXEGwj>1%4ye95kyNTIUQ=b~^+`$75v_pc%4hK`F2{R74R4-7!d{3vtsx=Td9dUD#4ITvmdw$%BnY<0ass5W|dK|}-UN+#2w z!6686JmbIicKO}=+y$a)@>M?8BisiT9#$dzrnmjP4+ z8su8vqFno-ntEAR7cJ%$1(aX^B-Xm)Z1B6UD!lwOby#V{vb1@c6rOG$X z;#L^4doS;z;EnF5uhNy%$sK&&-W03}wyEB7jBUAED|+ab%2*sgt~Oe=Sfcm>^;=NP zG|vSWQAyUXII)>ZjQ8$%6NeWvuMO4n^mrw&QE2L*vw6 zG}NtlpLE5Yt&5yl!PD0Bm58E4`~G`7$0WE3V`9Ip_Xoy}VVR99Wc7DZ1X?h`qK7eq z4uJcz*A<@uWhc9H=yP9imA3<1mwvslCq~({VLu}^wE`_Xe;?vP~D1907fd66Z z(c2o)Lh|VuXL~o&JPySih52pD9K^AWfIk+_t1~@q(5L!D&AXTiUhEI0mL!)0y&2U@ zs?-&lJs4zo;1|BPP}NaOEs!5&I!mhXus*=?Xj(Vas-1OAtlQCJlPu#QCwPDe_BS-j z#~f&rk6Td1zP+B)EjAI@ao1>qV_z?}#z1GmXOi4iluOl@5}&IIxVnYexOxY-HO%_$ ze?S0FuOIXLL!@n%ih)3C6y1KhlqjZ26jv5_It$rp)?q4lKHf=qhMQ3i2#FZ5> z@T_h3ox^fTO|w6C)98%X^Q)KN=2Z2qyHQ_RPybLNfg>$K7?K{ZrJTD)Z(*B0l2{ja;gx zK&#UlQ{&xLC#pou+HV5dNyAK)col#en~*p6R8L-!K_&{9c`$;9s@#|q;18miuzXWq zYzCPE*cGH};%75vO?bZLzA($nA!#oynl8>U4cAJIP=md{WO*m!u<1b;+dr8_>Nd3&H0)dvMx^Y*|D z#eqw0(<-H4?N`UZt|k~&?OWbi=|k_s?MHGuG|E3)|9F9;;(LLjVZdtVGI!Xzs9ndm zZXc2Jgx!l0En`8La~{AYiw*(ZotIT1&_8e(Ok>pLC|<=BOgW z!klou$^JPhEQ8aSM_z|JT=|hz%l#KDth-SRR(zJ6PvO@=q-zzQN&ZJagH%UP{qdrJq6Ze`ONH-E5NWsHjmMv6ak%pN zb-1e&zV{3pjn!6NVMOMPf=?SMe=c_uS-J??2X1vmb}h4Loo8g+4UHlyRuaR(KX!z? zPYe$fOMvfRXA&QKnaH=YA^z% zP5u{Rf40KiPE2iK>4OOW`i?cXD#yOloG!ncf^oq>_S$wA+djGI9Erx9B=cb)*X3#U10zdF;Q zl+^Z`eGH8`%~uxlHS4u`Au%WWYYyAFl!89sI1>UF0@JiV%L?zIFj_cQ@|kx z0!|ma{m&H$80naijGbiinR|qXv$d!kWveA3J@3rujr;n>QRPvD8+N~VTj}OLfH0vR9zo@bc85&0$ znF7;5>(pDgls>ZUgBzB=oAj!LPWEawI`T%N#*jjry}>>@W?s_sm$2#^GO{#%x)Vgh z0r2}?Cofu--V%f3k+NACmjofOMlX&iQQLK{J8sTuII(5 zw@o#@2`0FT^%Rv1i0vMxvP8FtZ`Eq_SlYtrE3ShXTpelX$ujw4;~RfBP!a%DxH%Zd zbBUE-;H(#4@A0@@@66Hp_%r<7eq}LKO8wjaiaMKJ)@7~)!QVQFJimo5a>?^@dQHS+*f6kx+hqHx!UHTxe59s#X6kptniQ z5t@iVx-rdT$Bhln?>^K%)pkwKx?X%hTcX^#z-l|CVY}HY$Y-}IfQ86z<5LgrbbiI_ zIg)LGt6Y+qxHzC)2Dc$))Nw*3BzfLQflSxNJ@Dq{ zzE6o!yQ;Y>Enga)SZ1rwVLuNOSE#C=qE37xoejN@=ll7$O|4FTjnpKFM9;7X=T&gN z5}v|KxntdN(x3$$3EN@kBP7DH!v{QFk?A7rHp^11XQDz9kTOzYyFT2os#(V{kDNsV@WQFI^O<*Id~f>N2Ee|J zi5W2&+kbS*<>m9k0t!@yGqxZ<(_C17~K5H3TL$y<4Eqjrm3)^hj$LYCZ5Tfu7 zOijAR`ICpa-ATN-xVSloW@FoZ#dmF^r@zN8;H+w+UM_Hn7(LXZ`GKWW;g?v#y}>OB2Hv(PktL_;tRo z=#OWWspl>H{@-~;2DFNtu4oGNcz>@41Ot`xfV{{#UMERkZ&y621fn=aDI}u)IL-bR zk?SZk2dBAhVPEAMF+CW>pBjn{77GsKw9);~Y7^I^1EMIEk)OU8q!#wu!%l3MukBgC z!0F#+dG+q(qo98>eJnMD-n`c< zLo;?Z#1_Cft9WO}fI*7YJbY=Q*q>=meCg`<821Qg*L9LE=WqEb{ph#szoBWQ>zcn= zPdm3tD|*L`?%j+2PKLYj?xE{=6ko2&lwb@$KV7J!P#L>0yH@{WE2`>8&p^+o02&j- zP-N)HLsUlZYlY+G5-|R2v{Ils~?JsU4O-y?@CH z5^s-TLk@PF{IcC8J$oJjTgTH3_rI7Ne@7t?$rY-^L+%eITxD5MVnnHe=?{qKMkVM& zB0b|$U<0=m|dA|RlO(v2iAfa>;pfEGFhSr5tBhu$=z1tDC zHgFuX83Nr|&h`th^>oS9Ppwfu_`-a_;E(Hg(QlQ7--XGHSdTHfUPVuiP+b6@5qs5x zXq;}5jyk#Fe9c=t8g&wAwv$NGvFZq4D9oAsWZdenBODk5+15daECSBn<=qq1q9;*O zfZlJU(_x+D+&eQD0Y|9v-Am(bCe7mMGsA{BISRNsPvi}!4(v7yIEO@i2~@ko?gjtHY>4MI?o zIqZE_fu$~4icz@ainjw|V%mC^h6Fc(^1Z-uiUS3zY-mFGdUt1X2_q^+)B`@G$}R*ewzW6=*jjJZml zEcc^Y9qfFXm>G|NYe)}Gz|c*VC{T#90+}FPe#o1;po)aO()~18)~Z)sT!nTM-z{uo z3FGSFCip;=2j$al9lnf0Gw};^o3wMC8RGbyO@JN8MCWl4{Rx{(=F#x|#98m%i6_XC zjaZ5yvybKnKG z$&n{&9i~5f$rzYxp)-w3JrpC@RKN9PgXlsVoyeOYG_Fw@OqV)d;Q3+eWYl7 zNfZPqY&n&oPVxd8r$JUT51+H9aj0~689~N$P+u`cnO7I@)0Rj9nRHpKUW5iYWU==} zRN>({02MOWs!7E--V6po%5GudH!5RVcs*qZ-!NO+g6e+nB8?DAJZym1g_t_P?TAVq z;g%h)Cb-^k^=CxoDA63GlU?s@C!665olW#)S&N{v!nF`bB;L=5gn^~Ju1-{=M-ROv zfdO@XPtgX6JPv6IH8R~;SDTJTvh|c11?q%H_vh|j-x{8xKC$};d{B@dtDgP!)Tyzk zhvBu>GFsPhnV+hV0|P4zBM6l}f1 zxi;cn%=x2KfMvJ`zg)*3zN`pAw#Q`~7tX6*)t9MwO@w2o!lf9dP=5MdUQMXXfa1D8 z7H@q&X~EUjJ?=e%EQid}V&(r}>@B0>+PZb!1b3IhgS)%CySr9!cZcAv!4e#TySoPq z9w^*BxVzlSTKn7koZIec=hjcCMk{KL+2@eAKMyp_<&(x&qnG60djEb6o zc{&p8+-l%Nw7N}~p2P^Kl<0c}a$WXIeSfW@`tJK1HP%O-77xfbM2#wHF?2WNi1>~^ zW6k^*_#%}uH@ivsXo5m-$tH4A|L|Aa`0Xk#bJaicgbxdsE2She^@Sd@K9R_ywp3~q zb+wstGFJ!cwypULA;KQPF*t(V;@K@`E9Uv6>2OUvJ;FvPVao>DrQi-5u`Z6EVp0BeUi7TG3 zjvy%Q7Ct?$7d=hA$XQGcGrk--kbc&GyA*Ty+EW{dT>Sm!XJ}Xdtx>X>AxpV$ROkl$ zyA+21;jorMY0}nh6855#eqNCCA^}e|Af)Ws*tRoW#vEU6s6b()AiqH1rcVlq!v?O9 z>vXz0j`e%vw+LHdUD*Qla#$(1dE8QPlf=y^Q|uWpS_2p1P*F#<*IhqywI>tZSdF^L zZ&!#rAxr^rmCkFK8=533CJeIQa_t=l>c3)++Me{((>3?RTkQTAA2TM_xmou?sH8-; z)LckaT(w~R&Y*5(LeQ{>$85OF;Pc0)iam+o0wId!&1iCzr9)OOtfAYDGTrl{j6K3` zkz$3V?n*rtud#j}{|v@fK<`XE71+o1kC3U*r9T^o$L$4CAPM3|J);a?N{yA4(_FBf zoSY%tRi|Ou#m*cnZT?D~<>{OvzI#U53f=_*P=}rw^l_gKoVOd# zd1ahnDPtnsOxF-uz^m)QO&#*2sdzHQ!V|b>?oGbN4<6jMV<}xhhn%1`TSfvoSLjNx z2p7iRcd%<^1YX2>VE4`>rOZgMcl}68v|-JvrLk0xB{YxBiuznCRsz81jwD(KH}{52 zq2hi2{kJBR)OJK{->r@rAdBy36x4I^Cl-;xOyPwe*)SifC1CQr>kx&04&3#XlCa~Bf_ep~b?^ogdC*TjaRlyij;hRc; z{&u?0NnfL^{$^zzY_i-BTqP#G;IFJz8g42-TOsjF%{#su39+c`09ddb5bvUn-5FJ3q44IGqC>r+JWnye#s&y`JQ ziCdru;=+0af|uAm$l-}vqQK2$zud)pq-+Fj$OW9eTpTl)7DF;?f_DU3K8xnGHoM}$ zWWI&i+6>@ZGAt0_7~Wudc_#0~oZEVN`Sf4=`rN{NlqKXmy+ggm@U$?yEAX~ovHVofHsKL3@q$Xzt(O9r?b z)M=9> z@}-ed&WhJCGVM$5kD9lS8}oY)yb4zylg$gTvb`T zgb9p@!6@4%lj92;$0dQ_kl>SiR4I*gnpV|M?hi{c7H`wofB)EBPobe<$q;{sc(=8f zgPmc`3ibMY7gF}f!La6z^o;8%O52BH#OB3o*H`CxNirka+8n^hgfJeL{C|llt{@R%EisE5+v z;y2k=uQkV9lNs9~w_b_Ci^_CHqsp%L)0EC1Nu34a5G-uSwk8Y6?O#~|`LM(z0_2PB z>8-IF6|$fPpDTaP>W=M+JPc50(Gh3F2bv=FCqhQgd0UoK!xw- zqS>Z}ri8N~4cZZSf(ajBX^~*HFVs+#L=0RnyEOm!$p-FhyjLl7BU^z&k2Rpun#abQ z#n0ILrxw0+(Zbuq_tgdEG#eIuF2pK?GHvW%=!fmF>b@Ql@81ZVPwwRX-bhs#!5w;- zN{-*dK4w1;T-{Xgq?YLZx((Sw0!~c9uU!FH>_2S(EDtd-v78}-Ja~_(ILqDHE#$bM ztUlL`6cv;fol&5{O-&Hy6!6E8{Z;V3+f75Du97|2KWFMf+a37>EPTdhUsP_fyrtsuU|?VOJtZSdu^1J+ZlUJwxVWgA9( z&-7u_1)^g3CcJHpJk)OrO%-KMTL_iFQoCySE3LH0=jcaBHGvIQ^nKSfaBdoTvDOrl z>#Y4d*S?Tej&S&|`!5+4D&Ms4=VL<$k6gWo0iLiwK05I9nDaf|WyUnD|0>MOg!K;n z^(G49l!DL%bB2Wf$uPm)m$sNlpe_-u!*-^VyOc7 zAi%DHS;;IH$61S}#w7p)Ax0k^pfYl~N{!0QuQHw0C4?RNUZ${Xuo_ zAzxcySwFw}HzOUMijd_)ZR=1SRAlNE6N?zJ{yrspS%d+vVq~hh1$F<#bs3fNVYEiZ z(k-aOi2j3_z*e_Ao_J!x_y+B)tw-8sCq{6~X*Wb)t}r5ZIOc)mHIpK=&oH%cosqag z^S2-zd_E{^`I1Ohi$w*At{b0dwAaDk%Y_NQAtd4YIhwwYrw}c1y03xKDEM7I<1nSt zb%&EF#2DGl6wsMaRSy+Zn?vh~n~RpkU)Ng5VQxVgSk?-pdEWX-2x` z1FJ&_RH-g-22OaY^*aU(EK#T4SmJLV5AJegw^gPvuvl~A0#WUUDoIaGb&4nu4#r~q z!fsUC<-D=t1M-vlpC9G>)(5u>q^$JCT@yd)@5!OrNR+EcY}(d)q$T$~hYAux2cIVb z@-jjAv9<;>d7vLmH<`*q-@}Mhb*%XPr-0}1O}u3a{Wa^DJcgl~vfG_jh4$!@=wxL2~Do~y-ckE;tB*ip9GgNH}JR~)>W zX;PC!^o{}TGn58CZ<$hG?-Jx{2-oWkSaUR>ONYGmg`k-Ls zkHr*lbWj4fM3=Fd+k~mYob(5$UV{Ms7;h+2{hi`A;yNghrQB7W!)_iXyTeoZpm}1t zRn{_taUGW0__n8Le0YD>M6lh0jT1-CAnp5p?ROJ^CC@Hg?#jc$ZzQ7cfc{c@+?x{+ z;u=4Fg8wVkA;)io=d4#?$VPH7OAa5TVa$upMeLFJsxKHa#G%4m7e4UZPGoUo9~7>b z@;FmS?c;F@w-b&$U9iDn*Z!i#sKP9okE71o(GfJ1?y+`-y)&K~Tf5Ne>;=nJsUOp1 zn`*r>*oe-e)r|>a`-8{*Kz2J=rHDD*Bun0SvPdN8>mgk-AE&xe%0E;ifYNeU#>`gO z|Hb)GjbqF8Yb1?OohY}?l9G6Z*`>UznxFSMd<(Vm=KdwI=;gZW&o&#g);Q#F&6%q; z+PG;(O{oV%tw-Yf#m6WSaXD55&#a(tM986#6Jip)D&CQtI8F340`JZ)qk7ol^A3ak z4%mr;qY1F^SMzxBshP$3Gv|@(MWIMV)Bb>p&(1Gzv z0u!tt8^Bxo4=L+*npHOet+*O`FT;^8flECnxBSxe%h_71>GunM3+^$Ma{e?2eUw=} z?^H`aHQB_TNwmt&;sf^3!}CFIdbt1>JTgvbdr3aDn6Fp0*ks{4>VMR5Zd)|R zqNW+w5P?2$P&UBB-n=oL8_nE6hHxxp$5oZ%8H#}p*OIN90Uj>1U3M+eg|Ba_Z3_wIBc5ON7a41+U8kmcYjX;P)4Nt&_jD^YD3^O_}@6o=VlKDHB z9`70*^7X@53W}Nei2cDSO+c?Wt@CyZXMK;cB4|-dja0n5|tm!GSuRhnYOEbhk#ajAK4x1BVWQ z5H6{3bN*CRUo9(lsEW#SS zr49?caBUZLGxO?=UN+LVV`^|$*H$;*hVCN}9!mWx7uCNd%&(~)wag2eOSA>h*E)OX zr$lai(!_`8;by+ry#%(l`Svk$7@?RmMR_zV4Ad9ym(lzT(*v##Q9>X~o5LZh><`aU zdtI$U3W+A!bX699mZ;NWH?duUw;|_z7cppre6j}kuvg8{f=FDcIG*Y}<2c~X1y(P1 zqX>+;Un6TXBI22%n{q?;3d?C+kF__Lqi6(wsr5mP-LDKR&%v3yP`Oa2?uv#r@6Z|@kjKL4|{uI1x1y)?9^zmo) zq1M+f!RFvDoNEn^e#HozT`1sR%#;G5L=j08x(6?FS)a7ZUCLJ1Yn^QoEsa-M?*gs#m zXl@<7zvp7>-09wPy0g8;(`haci*SuZ`lKI)6(!+4U!@Q0>goyt13id{RJ(lKL3;gQ z3T6bDJ1WII^^F}?(DyFZsQ4$;_GTnihNrluGih0jE-qK0r;XJ|+yFfGIO&0{!rrpf zGTzGSYOx68KU`=YFkfS-b`-`{YHjKpfBvSYpWgR1bbgRja1=rr2JQo1m4fRneCsyE z2C}eNKVK|!!4akLzpF%r6WZ6l^s6;SFXMWDK#8{r^IiZR5Ij>Me%H+zK2nN`ogod3 z$SyuLc^#1OSTgC{6|h_OCy8Jp#4Ty*ZQaS;)}*72x-qeHCmdUOU!LagC);|zfST9W z1HSKUlDs#!(mdJmhU2*hRMH}Yt0E3CVN3lbFh$jy$v`^?5#si#pcfM`>w?ep#sV_p z;%MQE(ImnkxR!|jB93Dvcz={2dSEwHE94l`%qsr;S|;$yukf7%-f}W4$|#4=ovv6h z6Q_%CaC#a|IY%JM$f(HHJ6lbkf%AFCIq&Bu;rTwx6r8gd^X`_(uCQk;$aEa4ZuX&D@tucX=@-bT_&!9rr*W#qLRy`3L4tTeAmUUeC z)zEDnl=NEl(}HRjFf=%Xzr2!)|rT_g)y(zZ$b9o^zkx8=)~je^9YbF4iN5iV#bUw*mk-V)3wI$=2R zJQ<#e0jGQ@g8bFcEtx{!@!Tb*-u!2(e@&$Zfu91v^C9F#_bcM{FiI!EXtsQjrujpD zGm0lphGZFLO2%``tpnCUD+r~sp|Nxgj^Q^!a#gM*pQ`Ef1CtgH@;gJHaPWfV-0eC~ zY?c_nI)ue2XZ5ItshP8A9~0Lx**nNB5M)^ZSc(QogFEC|3V=lt2QmpbcL=!6j{Qc` zmYrV55xt2uiRCD+^;tmlcT>^KtwoaAZ8k(brmkRYR%U}yTEZMlZQ#sRSL>_NHPTSB z0rs@dp@zbdHljB+b8WXW>Y~u2$25g|@zzaRz-2f(=uY>6eQLaMps(*EyX9o8QMGCz z43kkOpzCGrcaeOZ*X8!1IS{yW=qse!j=m4M@b0j6HZbgUD+Cv0+@Es)Wqc&31x`-m zf#itO?u*=GpnGd)PrAS^a#pp$K=$yN*Yg`@@W9?kB02XJE?rc4_{2=B9{T6k0ArNv zfidmw2mBftvdI*ZLHRv#IRV3dKZDDSDqpXgZR}2AzoB#|JPN!9QvK?eX=XzdO$`0IygHveSc2&Fc`~mFgMTN&+OtEr?~8Eh__gZuDQmgi_CdnUCH~Dx2l`at zLp&Sc?fheG*-KZ(*Lv@#mS94hq~mqO$^HR zA9t+KNu1sniBI@FV6+W7ldLMmvWO@%Z~8K##Agax!!r+*VPglj{N1VN?N(zDQH3e^ zJWCxWf0r&OkgN|g^@%ri1bX)J6kQ!;V!;D^&wp-u%(vD&DXG4l>3Dh&<5M<8n7w?7 ztnXy|^WcH>`~9ucw3nv~Dz-;Gw0pRazr|^I58f#7A$F?I6zz{in9+LhG_yK7=P|bg z=WC^Pg&EB%$yC-^@1HvUb+5->VHzw`bhb!$cCrd1BXA1{p*fwlLG0AxFxF8cvH0{k1j#ytZy9wu+>unmv`cy z#K`uhC$h+WGUYW(gO`zy3dMv@t6xcEga>xkkyqUUu-0aufEU^$673_&ec{Ffw_Kje@ zf+_&Vk~vz={EB)#lAP$jK)E!xvWg#fZroW(QEJ$snRwOi9{F@F@4D_^A$a-&4*+P z69K2>!?zm8lb70~pO30C6_Awz4o~21z{A$jg80no^!LL9@acS~if5LA0Nh$Q)1Y;e ziE}ta(~M^y6VD4jNte25Ef46I$_LJl1TgfVq17#bWL{rsDDjL@wM9*J7tN-R0X2NM zC=Re-3DDYrf5FL;M{yB0s3VR%+ploi4)S5qfHLeVpfPSBjphGl5;>?bZrqN-pI|g} zTOn=AUrc&c5|Pg|Yx8y}w>=p|6%(rQhPhG*7omaEFe}$SZHoBprkGOnshs09m~_X& zEC{4lY1DyJVV7poH+~72Z@{+s0sh1(vEHiH0s0w8L?LC@Rkd>#*z?QPgGZ@jrksI% zLkTRRz5~06IMQH93t^LfX0Kg8gAP+eAHKOzwQdt(nk$v5Mvq5n`mf+YRkpo~&|*Uv z;7ud64o0f9?Lx0b8RfomL+=+Q#lO9jlF68q$4%lrN$RXqHf>j(^;xa;ByVY%!8NdYuw-jHZMW9upSdA(O7|^U-@}jmp>IovLrjAO`MzUM#vi3 zqgg`3%jwbwy2%VcGmLGLp)V{<4WA@)f*(wmyP&}Qx848?$Hrp|lrhDKO|N%o@N(Vd zGez=-S{XH}_I`;^*Q+NU8$1c(N3u1B5e5zYA8^t-tsbPMf0J1L`DZ8@pcdxhQ|BoA z#zb8b#c<}~KHv&Vmo4^bcm##0E`o+wp%Nr(z&gU>DSyZ@_oV66cW;*Sb-1G2Z9_Xz zrbU4XjDs-82U&`+h*vj-ja#_BjdR9rLYj?92zig}fe8XjC#PYUMV-l?UJY-D^Tn(- zDU+G;bf>cpZ;=gq!8v*-Qs{(VzJ!MNfJD%VKfGsTYyX~zU-ju+{0CO|msjU+q6jGZ zV3<;4gvSfDQ>WJ18Gu(4z-Gpf?7eg3?uq>IfT7f&(1px3*B>`3Qs$H{1zU~wM{o~a zc>ECphAai^*r;OubTxM+g*w$|HAZT>pYX@yQYLi+v!Mr8n6inSHA25YR?$kuv$Ynp zR&x)?A2sTVdmA;pR~zFXM+Q=db($)ben1*VdBD=kddpd75e9kMn{ruk#dpK0-qNLg za;y^7>3)Z^8*UVnz2j2$*scsy0hRJgP3k8Ml%V$dm?-Y5+TIj`5e`Js0)buW&KH`C z_HBju^E$Fwu*fB+z15T!R@z-E(c@oJi+}9B%rHXu-=>+Vp$hooBTx=q_lQ&U^rT<&GkdpYdWXA-|l)lV|C7?$F9gd0qxlEqmg-xN1pFwCX1`>yM(W(snZ5ETeYyBSQ$ zl4Qn2MF+MzorRiP%csWh>496zha6@wY4+gOXzsJ^yS-VSMps4#>Kk!P#p@_4cIRq? z5z;wzgD3QlAD5x0&CmVOKY6X*B+{pQv|Pp@5_Cly~yw_wqwPT~00qy_e#9xH3OLfg_~DA%m@cVsZ2k=icF(G32NGoc9F zQ>Y$|wr3T5r+5vQV}h_mK-jc~s(iW~z$(GUOsM+Sksd5)CONWH3aPgXx=PXV>mN%$ zZ#rti+1d(ffmkU=!xljY`L+Lza%15{I?<6OmVHKaJ6nA$Q6KhsDrc6b%I9|K?Gw*S zZ)QV1R_eVYUoquvN7Y-PjS=JB2BnSt80F5jf^fS*R?%TVE7#27KRa8#B}BaUf@cnL zmCRQMz`zh%em5HxHVnVrZEC&N#lHE?DUu1TU)NXg@d}>PAWuSKdtNwu=7%$oL#vod z>b368Sxg+SbSxe}&W0>~d5;+tUTlPU&1)C#>PqT_G%V?}#IA@|_Q?Z{)@Aa;b0fJhUjVQVswy$!hns2wcDDnWG&4%oM3>G(@U?9Jm zj0XqXpWxrY&b;8A*y@3kiRx(wk9C@SuX=wbt#vqzvtbHWkq9BFkup8SEG&nG>P^Ff z)Ybe}|224HsJGHq6WtVT=u&Jxm{Kj?nqFvZc<%IwhqkvlTJgIpQex=0R&lcI?L)R= z;quR*Rdt1}cl2D%u^xr|$fw*-&S&MNdJ3Dns($@}+Gq3$a%Ny6#FZgb#r&LX^gF61 zs8BU!5_(5i-LW!DsLu*v-U&-0k6kjOqNP)ov?FTs{k3BLoeTpEl*>#Gt5TB+FeE6^;5AJ8Dj~f%t-Dy$rAHH z6`(C{xzIUh@wvVKo^D+v??5!r5~Dr%}3-{#SFpd!Ji4OD*mh>gJzp5t_#swS+nj`8V&(H3Iw z4fR>P6^N@((E|0+I7eok1&fS_|Un+)F{yYwfhC#eB{IJ0L=S+RNBWz(qA)3A&PWm$oY(enULzH zbw_+Jd+;1+$84i_f2}lM+F8oqE==8h#ddXloW6zfJ-Rvwg3yUFwO)g5@uB* z^yXQ3(&4%%9~WUy1b3gaZJc?ZkrQ7NI{TCRzy&AwW=X!9Zb}2Gl(Jofn5;l%Vfc^` zi!aPwEBw7{+M_{#GHr7!^$pPZBZsZJuo4=%YnOwS_SLY=kL?EXw28%V*zduGa3RB& zj~T5`eCx1&$4U2s(72ouv)iKrRMieJaQ=ldTae7m6Mfc_(}A@Oo^n2p){5@8YT8+X zhr!Bh`Nq?ddy|Hj!a+*e}C_ zeK49JK-jM0WI(@DaT%&@lukYVWCW{Y$EzDiHh#b`xa__$IS|)Ty{(9`g7;btzUbre zCv^k~BM1B}qkbr9Rl4`5_{*(`+F#o*XWUc+dO46POXr;w(4lbo5#305P~VOGqo$6` zBTKJOpNJU<=O7E4dB|06ol_L{o`CJv*h18QnYe#nG=|`%%^Jcm<(R`^>HJQW_^_ys ztFX6i-(VP@$^450O?BPAW^_VA0`T-k)XEloR`yMK28zW zb6Qzdw=8y&HRJ%MQo461JS|8cv$^DK9``Y5hpF7H%Px7I&H$LFJ7Uyl%gc&iJ)Oy$P4j&_e6=eVf>iW61tWFP-YKN)KeIMBP2yTkodRR5Fz9N(uoKSs%Ed1$Z^Byhf&M#inKSBj?@h(zq+tf)%uK5L zl{S6I?xs#lJZMtLanr?y{EgN#Q1%Cgny*OHv>-@wzGFnFXe>DYdc;Csl%JKz5zc?A z>2tcALXNXz7JDoZ6nvC=jrw+|L@y`FX9^|QENO}if2X&)kaLv+IIn_^x7>9)JWiD< zS8gLNKJmN_QVtSmb2mMj7Rsm;xL|*NSI^c6gxz~}+Qh<4(X_oGMRggVT`2=hx04=m zSVgCucH~aw#}sy@sa58wc(a*2EvvawPCobZFeAZ^0{=KVb5CN>8o3I6?-SYSN4xUFwn}xLJ)E&Y>r(TA@Bh5!BT^o z!+H^^Ds=_a<*LtNI|r3Ljswj7`t(>VP=UZTtZ_V3`fbm#o992X+{l2RU!2=1hjGD8 zbSlS}`;SM+5Mp0Dk~Lxl6fI0{v(jj;fV<30kJ?a=-2<5mj+>_hku^~uJo&4_KXagr zBI`w4Mp`lBog*e&1-yCGwGI#Q8qq@5;X42^v=vA zM9j#C%~)V24)ACsh~>{Z@|vOotHq=q;lud^s$gqDPsI{IhhrjNPg!Ui0$HV}a~g*Z zZMPF$e5F~cCclu?`H5jXO9*}cmA|)j7GT%){wE^mGb>io13uM@y6mVLJO@BU1pfrF z2HdrC`$jIdqhJ9(H5(e2NJz%hICdU=F?4b}C5$Du3wAJUANc9UfMX{RMx7g)pdYT* z#X43%TwgvPg?0m{Y-HU!85m3k#M59Pc>jzVbc{L0BT5^24IWajJ#Qbg3eUVM4Hlt` z8ipKM$p#ia1$O%0B7rxILn0P@{oe4;-$*Bee!ZrdG{Bb#T;>Qz1G**z2gV*+c zoJrYGhgmc_?&+5R{2X@SQ|`S4u@V}-8{4?x(e}$aiop((sBF$Bjrr%Tr)2-jL48l| z*0!D?$4+#`6ud7d*uG|$IJP%5gWdnFCQEq(!;O1_nB+WkgR91#^UHIbnCE_AnAd#C z(bBK$|53S|6(*Y>^Ggx`ld@XP2XOF#%NQwU8NWXsTm$$$p`{i^U*&PNi`lwrowOn- zP#dbSWod5!zj2^B{aVQ8U=!XuLK}t?`nMi*4V>?OmSn^qV912hJz4U%^tZ3;$XEBAh~OU#w}`{TJ2A1{)6Y^<>KSC zzxL%(TAc7jPRzFFdu%D#bQOn1CYFh~yXDlPz`kdPG!H`Wi+=jgOh69y|K&d;l9ra% zs4+4G10Fn1m$&vOGH1qI_s2!2as&(0X<37KcnaY6*Hmtklja615i@7>2mJbk5e6>n z;#gh_^gfAak*wKInHZ685oeE*{(h!w)?x0uldnO)Vkc^l_?Ev>D@>(C5{?6-7aBdl zZ02+A%=5AJOZL#>(XuS?v;f6*efJZHfKdh=vN-PF=L3CrSm|^ z?ygAA1{|{k!S_}{PE1PCH$F&K-x(_a_B@uGQn7$y2^cdI3f|F{U{=oN!=5cxj4_Hg zv$Be$R?d-N2CL!-tMre-utDA zc?VU;luI@gJK1va1`C{TEl)iEviA-_%?-KKEDu3iaxF|0#nUKX`haW;Mr-o4nBcu+ z$D7>GWIDamgjE}nrc|xrukmR_Y*z6pKKoutQy})!mt?lOzr!2(bcHxhYwx8UeFEMI zfOoDjuqzg%L-dLg*~5GUO56v+kq4^uTME*v{t>RhfW=@8I<=pk*vZMs!TpC+|0vRg zVPRp%vv}Ek+Wu9SncZA@%BM38ZV$#3rpLv_6`nu->wu~M|FLha^J{(-W3W*F2dh-I zOK((m3`?iIBGzcZ8ZFH)jSs2%C_`CQ!<)4Mu^xTf5|QAH!I&TBn;|QbVvMEordi@A z(x<)q=I`fVA#Dy3{*YuzZy$TJ4c>){!-I{=S& zU{DUDO<`6*z@ho*hbWy6;)DPF2o&+<^}#KvrGPb@X+3U0e~2250c`V;6{g{ewTy__)Dnv$E^ROUc?F$1hae*?uT8|`JH@jFdRGmHgN&LwK8~aZeUDF zya0{H+y_c}E$OPGJr$L)O&3D=8y1JlDb7l}cfkw7Xiuedr?VmQ&a1~S34Lrb_4{I5 z+t?R9FmW@rHu=U!=J@+N(eDmx1BJLXo3C5!-J$e7!a%OiW2q!hMPvCkRZpdcQkf!F z2!CT+mximZ{s>c^x#$48Fb6Hc!#%n$7^&q2QmOSL@9f^WW@MK$uR?DfI}qFnEf?@` zVmhIOe0Xf)%#Nm5jGDxtBMyAP;aR-9fAn^LJykOs;Q#QSS%g~d+}uFMN8nS3 zH4zI{L2e&-<6~wc>LZY>pyxkS^)l3mAOE{AvOqW<{Eq+iRp1u=fBcDD6Pb`t|K5{# zmJ3g4*ZbMe>NQ{I!v|giS09vIv~Nc(|IVc&BM&WE|4W`z7D?F0OjPGO>UCqg^1(gR zO+*-+TxWA-!g&S8QEJ7ocs@bJ5>8_eg(J_DPw6nM2cVw?A%J>Hu&~b{57bvo8%z*} zVgU&1v!*P$EQs>=d}RS!2T+E3@%bSzU|_+wN#0oQvSL9I9=|MQ(~b#01w3{LaRo)z zaX$zPaMq6=+Zxki+i>Vip_h z7}`BgVxXmSNhNN{$kd`(Rh%w6+U(4V*YtX zx<6Zsk)2G~AmzYr{HWsBj>;dSbhsFZkhVnBOFK-^>VJUZV8|O$*paYVO+rBTue4Eh zdH4(K>zdnpL_mzTDX84b`c@68EL-%iVo5^crUaa(4lVOnF!kO=7n0K?p$g*IB)yv= zhMF~B*kr~n6M9vLaCV>Y($UfJli#s(7L1dw-*dbJ9xbM}zT8zDeeKDWxr2EwP8j$# z=7?VW6np4M`PDfx@b3jL5T<2gQ){}hjKq>FHx~s9r9#@;+Q1^T`1trzqpo%vM$^;f zrh$n}?!xqg@oynHg1%D!TIsBRQ)K;>J$fI_m4OYKdN(du;^N{0-d7Yp-j7#%R-kgt ze?@oy*V@X=I+419iA-FmZ|WGDxMOw5 ze2@L`7f@KD^i7gO)no2kD|@ zc{tKvD3a)lZZM&weBtN56STlrq#nCdC!aAD&{Q&$$%Ye$u@~mX)Q2?}i`os#peqaz zY@o|9;OYF1JUTp%D?%%;01L-uZf0jFD1gM+d*nLUoni>?HUZ!s`Fz2KvtO zcXS?>{JnE-JfEsQ0=+*>P8`O|Qooo5Cl#Dv&)7C&Q#CS%F_ce*hHVa(HR+f%!A|DR z`3##^h0s280Yxr_u|}R(Acm5vYyYgHdN6BxGcZMJA#$6VmLx|JlAIQGe{rTTux~G{ zS(68O(@OPeA=Y}ETBgWUXHz}JVITrLkLGCG>!bguB_b8kp z8ow_(H1kQunf~T_3?3}|O4!}~_ORZr_d^|P$bT7JtpgdsVpGt+zSUQFJa!8W##g)( zQ?PwYIPPzCdsFJEiKs)WX}%FZNUg3Dl&4P^>RTgMp_9!HA-{C9C&{q$!oL>e3-ble zM(Ki)jkkU0H!Zg;D8dn=iqxBK7cH8yl8A(>!0_Lri_hG*;NJ|ljViKi-K0BKBu4}M zolvDH4wNuorJ`|VCO#?tF<_7bzF$PQI1zJA=1EC0fJKwsQQTleJ>oCzqTPb!d@FeY@LqW*Uf zxr5b(834OFv%$^nZH%rgG!$9If38U2Hx@q>4H6!`(8{HH+IQhUKOb!iEo7Sx3r++( zyLuu>ny~f}Tiz%8OwX0Y04WFHMR~ZGa-Z1z-d*Gw}w~gx8>ta2$2fxQoZ7=>Rk+@SM4-zR9p%{*7_o+9x7Lm5 zi^8+{;%?dbfcCxkjQ!tX>1m1zevUm^{~b2qyT@NJ7Sh@3v-)^f@lg~ax=WLBP{TXm z{pU+-l9$IX6kI{`js0wFbBRA`A=7d(q@f{eQ(GI19ZnD1kPo5op>`7Al6qX!ZvbZD zL5!jMTJfQ?;WIh`lq%R4Nn5iOC0+!TJ|K*}F9CyQBY3k0+OY-`R6tOzjqI6hRgwT4 z^W^CemoF?&d*>i-UtSvE8={uztP&N==>`tA88;0|(g`xkcutgfPWhJ(GJxZn(biWR zP4MaGue4tg=Y{?;ZU}jx8*_lppMwra+qD6mC%9(+2;jo8BX5;{AfM8*n;6Po&rL@M zuI<8mjQah$*da@F!%?Bda}a>N9Jc%*YC~NooEz2>xKFAK;-W!&xuZ4mS2O=XgNTsE z80N<+o^_ljQt?NJs}v*jo)5r0^%^tLfwHp=Nz!*Jy8lx8yAd}0JF?id+BZ!S1o+{> zrBl{pKc|~EMy6?KF4*GoaDCK)iWm=_CI=EG!RH^#9)=uQzUTx#yYWUbIeu<78xCq{J;npLcK@Ul^$4-C zTCS#sxw8$q@{zR)PR@G7B*!>EBmQ}J8sU5<%>58yTOK?kdG_0VrheE2=~NXL`l~mO z|KGU~_$BARIPkC698OmyPsE`q9Rpg(cqzo6nM~stym)r8pJ0)DEWp04Hy5(8D(Ex& zi@li7%(tC_uNB;4AOuT1;X-PTtrz3DMbE8aw{Xgz1i}eL9dozyE9eahXTO3A9+o6p zCvhMc4zYW8r+>lBE8%0@ADy2)159R&Ut7Rph$oX?e1_%ea>n=ZB!)V6b^Mevi!i9| z=LK#7NiYi;RK?0azf`a{CKhrF@r(@bly^Yj?H>@k0Q)cxu?VTc42a&#I#1zw5Mjg%X;L!xqGy}Iax!0hlsB*XVYxaWDZ_|6Gu8cJbkh8@vpletCmO5I z@oekiiS}^KoMztOJtknQ(KrC)VwNUAP z97ZGm+_n;<|4u*U5@m3Zau1YRJ zMDZ%fJTkDv<$K&vyX`uyZ(ve)-zXSZ;NVLnwERXYNh6@d(0ehnazT^CGxNV1fTEy~ zc9I|YO%GCn01fo{BIbIZ29N#KSv+nTAFRK9{128#F$h!P$Q_{p_8aijouq2YSh5sW zi2(o;W1r9@j1$emQ$-Lr%1=y2Gh?MFZiT4xwSIMU?DP^m~cj0^*5!I>f z`U`JZ!Ug`d&m#QeN1s8Js2AyUT7|Bt^Pg^4B?ovCeM!Q!FgX*GXQ*rL@8@M_@y^0~ za<>ci3AZ<4^C4}{3V?e5Awe(Wi+*3*=f?uZ2xDa{9Pn*2@)OS(lSQZ+yD<}anpiAF z5a@81v-2V4L^VZJ??{TcNZ$oJd)`8g3*Lb_3)GWAY>eS1`}n?D?JlsBsAoI-gXFah z_%UZq2xNsv3kAWvB#t%xb4A*hyi^dBQaSbRGdkWqc39a*1$InL+KkS`DYh4{mK$na zY;6HH)fVpm?d*9F`r(O1beQTGXWFpP`!g+L;O>bGd;K3B$QWlxTXA`J-6mdK@ zY%;UdD##*PP)TFdWqNXluO$}HG`U=GLa8EcZ;a1b9;fCLcnJFa49B3o_H(%OOVIwmtuLAs1`Io32}?9{q}J-E)6+1yM1Wmx#PR@ z!#9cYW$OwP*QApkzD7`qPmR=yEh&u z%?UTmL#$ts@^fWxV^q7lf4+IA@%*8G!BwLR6mwWK`vdRT6i$e-Td|G}#BuCjSC4OX zUVvuutOiRh9Tq1^qT)5Lf8wnR%(?Sx5#7yi`~ABW_T(OTSd{~w{i&76JQ!>QM|G_f ztYlCw`d^ZE24h4+(4(OrLNCZa*SgfhFS|AO9$nX<6+h8ZVh&P!@oWrq`EDf7&U(Tr zsyR{?XfOadk&Ia|I6)nF(yF0-&()Az6hy522M-lQ$| zbtWbHCtY;Xnn(ujaIjkO5mBnxfZb6D!|d1*)jXsl4?wv03Q?`E0|8b7yqN6eQzR$X z>ic?w3a%!M&Va|OPi2B!v;1^d<#J~w-`w!o3-jwes|XS_+6pW#JV ze3H`+L(7Y&hJsL>pyEXkt5w7wThQIG2dPi0yw}5d3fZk+D77&b+Ce#FLtI4d0STg; ztpVOs8#}$R$KXg_NkaKhNUg~*mjb5-;&1xki`*SWsA0dn5VBK?pTq;V6|6ss_2yK; z%tg!cy{Iby(}|`MV=r#-N;6{Z?CG;EWv~$_NXPvUO-tzzxsVQK$OGt*X%;!E_s5p7ZvD%1{$+KT7uIj>#g?lN-HNP4Adx&|Dc#ebOPSnGTZ=*` z%k3Y51eOVJhI^wZE71Jt9lE?K=9_ecLai8>KJ-)QucfXM(%+P!Rf-*8kurObQ92F^ z6i-I0S1ZYN+P4Ay#!vZ|~NpwjPrB21M6uOTDK?)Orr;+}P=HKKaBw1*|c6 zs}{7gJ}It5zOD=v8l5d3p|tgo&pUcP)-8BO?^5i-_FXEp`9$LoGR4(-(t1o)>yzQX)lA-R)`czRk% zd@^eKzm)_D&lF06&xg|Mtb^g7CWPa-doU&*E1SEp7Wm9!X35(MY2c zud^7y0^q90$d#*(6znH-g&;;Wn+%??0;cm>kGR26H1$f;11EM_!9+JawwF;R?mzp?)`{NuyTWX3Pk9bdVbp^`-5= z-_vPAtqxRKssgbsrP^}FYNrBdrQAz{$aJ^QyO4;V7q(PRM&r%a-h)nmTF7`c`1IPm zDm0GZy4LT2k{Eg3lw!ULeoK(t-C0}oq5Wh5ut|y+;$zl}k^ObdnGb`zRDl!@n!|~- zCm;oPag=Nl9ytCmc#HgezI91Q;qhSSYznxwi|V&ImDyH5+?6b>!<@-iv73|>@*Mp7 znNRxP(=LpHOQLvijrpBaYgtPo4ETC4Fl6~nZ$ZBpD0nG3H|le;JK!NA%*(U+xA zetJPS%bsp={>?M>I=b%}2|$hlBU=`Oa2obq{@2|HPhthM>;?+^Kj(@|yZ`GkS)7%d z4m6GgnfBPGn`tv*^`&J`PkC_4tUGMHL&=t5f=Xz_#tO`cLBwuJz>k8xq}g8w=YrFx zPSUT=HR(U&;owmC4L^ndY_wVkxH6NLT-r}d8T&@YN`APiebaT6OCNm$q7}<1P_1$- z;0U!0P5Tb$$rc~u5ZUVm>l%+V1`e<&cm=-+@sX4xH6YVgKTsBNfM#6C<#A{`Qia(s z33dn@6h-i@3~o5yk5b4 zE59+KKN08hv3F}1!kElW_kaR(H8tF?6)Joq57+pSb_Tu^bmdAFGOzK6(kZk#%_YX} zk)I70iI+#sfXmUbb9OeNf-k4K`&3#*>fWg~G(GU&Q{zU{X}=ve8TfQ4=!7fD$J{fA zIvUUN`fBo2qH4gsN8~P$`KMqWZM^E>n{^waF#DeL_3U}#-|ah`-j_y8b*~*QZ#DA9 zCnQsU2L|t@F?oj`@Ok+_v<3bUB-$2TvyS-_K6qw1KX3blh{(RnbX;9X8Kia-4L((TO@~g>u7sm(Y5N z#K3tVskBt4LBC7pH@`5FlC0GmR;92Y)Q8DSVOG!Hw8PKitxPGFBD2m@@0Rr^K3tgu zEnXhH=6fM~49p&xW#aP&!kiqQ)$UHHw46Y5bSVfIp@^Q4_p6gtTe}9_yxl6C5PZSg zr*0;beiVO$5t?l&)&nau*H8mT?|8O&2Q9q5s7#*e-M<^qQ#0igi7ZICcl{4N-$-v! zMxTXN2TAi=_WobRT>thGlrirBuPz3Og<5e0xr?F5(H5h*aP>LJfqagW<^#a-CQa`J z{Hs6vX9Lc3@H7qjCHbgT^lcesJlE84Sb)MAI@HFs#n3K0}B$vn@b$`3-Y`gqV z+>{{uw>Wg^3#AexqPMBdFH*4$I$b0&=2+5ShoLAY@^qv(N6Yrc^1f8|YnKsE;Jn9a zq_!M4aQQT^%imP=(w$5FTOFh2R|*!B3Jr=X_2u9-XcQgqx&$lZIW0vQuW@7pK>QQl{c&4S1 z`=&rH=CfNgnx{BCFXl}E&R;1Od)6Uk%s`&ndz6K-$*7(06VL~dMVE*YqC-;=gj8qS z#tK$N$i`h4u>E#rKj3)f)R5i}gzy&rN8*_?C7~*h`7rutWo?5L)7f$sM|GC+&on*Y zl}u&Mx6mjYnEk<=88h(=?$O<7H((Z4oW&{q(>}A`zS#}udV{vGe?=;C>^X!w`V%=o z@e#J7-zgE@{*3(bsJbZ2*v)Nq_4prxmX;QTut>`#^I<1{3*zv1O-tpuwe^>m_}_R@ zn9Re^F_}XjyhJl6NVHtdSHU~!$=6_nSR=O(WF>%15i;QLD&Sq_*K=!3)ilL*&>VIn zkCd?}@lKV?fUytb#4GK5Mp)J$L}|&yKyD?|d}~s5RT-E|rEZN!qxkiyp7eR*rZggu z5?wDiQ?HCh`?E$JgFbpV@M_408~_fd7%t?fl@*aOmnKGe{e=vnsvzup?z^{hF8g;& zo+g};=l$m+w(6P4OBO@YE4@LRSXXuc^yGYQlcfXwviq(v#jAi(7jX8yW(50H)Yfn+ zEsYcjd;h`1o#>KqU#L5S6s~BDZN1@7WS2f=3OP>4c@4;v?&Mo5civ9PB9i*uNMipr zRVw%98(E3%KN2~+pwjJ?wtl*k+vxKe-Q?Rtsf0igZBWYmB`o@ydhn3LoB94kwA1+o zs8mj1kfiO3BKl@>FPBC;LUdRAlXQ^`(W54-b=qWVAHlG4v4c5!WM=pNa$)1;;byFY|D)YoUGSYd4h1G>kD( zBZ=ejOUGl<6Qxb^T;r9|GpaE*gC35GsDq!Mzy#R*#oL5e7z%(}X(~%J0(rW4w zTIuSB#~p7&Zk=%V;wB5XEgRSfj|+KL+v?;gFI83YLE%hIi&Nj$LI$ruMp`(QG73OR z>24*Szi|y)Ahqf(6rSyg^F^#1ISN0ojxSC2df$6Cr3-oC2eOCU{-H@C3H(H#%zmRG^t5^w~ChxmEkn#^TwJ%zZY)9DRkI zUk<7=)rB&D>AlB4{k(GO#z=dYwuolsI92Mu7aT1JJ5S?`n9~}ef_m=oSGuqa6R{b= zSPZ*i)EmR4x`p#ZH%JL)ooc_j)N6;c1=Uv32BrI(g#KBW{f|5JnKG@RoPaRxDl_Uu z*OQ&6QU~GdhZQT=2e3CRpu2^)(qrD3kX5gzTAggpVVC@|NfK}fy14y&k94@Q))}pq zz|?MZ7qF^7L`;<@wkE!@Kb4q^X_pg9(_l6IlAM=KX+MSIR=JlaJ->d<(dRk7@6~#G zI|N>wgtY=F2q4CSS!ae`@6W)<${`7ksXhEz&Dl;04slGoyo1lKhrt~Yeu^Rtgc zUG4G9rcg^<>nKEFw*w)JA?@=8}etPIONgHhFgt&VYERb;T_{?9c$f`2 z+$~tVS&U%ROM^Qh6nj(t^pdkh%m_*0vg^!2T!t#y1IiDHVeN{51U#{i%$Jv{E)K?h zWqKWBXU3LZyw_M~t@WPY3UE0fDh2d+hU z^{TrWn|61Ce#?BF3pLZWys=cG2fnQNdk1L1HIbIEIaLLuHv7QGNKnE$#(YTPc&VwI zxoIFqcp-N6^+z8qI|1;0jx>ybc5qld|5v%uIoEF4q_X*xaRRHpr4}Jak)uktoj*wV zuC4oSS!m@YtItf6b@P~*)KpWi_m_Nu1l zyhsb5eCk`1ZWhd(zI}7@w5dSB^1v#K6M~GpZ@4`w9a|mUOgE+8z-G1%O0yw!76V0F z$E|(!-{`D-?#6-{3{xtop;|1jR$wD_v z4Rt<)w2tjNrWrx5EEoA9f#B*bV`(* zqYC6CsO$T^oe4p`t|>Eegfj;cJsl(uA5DI{1)!#piF#jXCeJVf_@j% z&~PZ}tbVrz+?%&QLzfghweAGL$Q8OW8N8J8>(jxBl-n&^g@(089)BR+*Gx-V$lNn; z-^v?pU($xB&a@FZU8y^_`;zh{Ze7?dr*0*@nDz%zR{u7%Gs|##EM52d@)5n}6L^8!YXkqPgIhf?JdcMGhsEI7_U1bQ>|sOAWF$$D z<|(WF&%-n{ejbr0r|>D3Gz|LmDEq9F3$Z%AKf%YqAz%B_>G=1ZW(LMnX_3ODBuU~_ zhroE&fz-C`!YDGNoWVj+x#o)CF2QEI%lvKohyOl|-X%o&o>qK07>h53jyA~DGbkix z%MVa$L#ws+invY4kIui3>ma+;#|&Dj|eNrFZrvteE?)skJKRde^soXwzb ztLGUn8>wOiQbG!Z_je4<>YR?%6)3C~T08yO^hTgKOTJ)oTbzT2$a=sLPauhl()+H+<@IEPE!DHZ+j zqUFRjj-d@@d?G7L>c8h+4QA$20PoITFvQca@id1Uk1Z^di?}6QX^uHL&RP~+mdm#P zD#$$hNmfc!txKBruu(RU#uF>;t1{{j3~GGT#s%_ZWSpR#1K-h}9^w(tTnfxEvVa}w zum{gk4ULYs{VmYCv6!GV2XUF#N5focjku#==Lb8-!;6|I^A{3&j+=vw&k%|~;_27~ zVt&OrP}%8kobf@4b)y9pSz`iCTx|{7@5M($N{3R?<>*LM&Iba{`+O^CR~)9Ah&rzK zB`SQkpr_Ul^84#mBjo3!0gZC18hY*z>Cz)}kC7-3B+Fv<>;K;RKCcUS+&0Yf!2zj3 zpd-&$rEvlSBxiBt90!L}v^}B zQpMI$jQS40ifTn2DuWfl;Vud?tvYKCOd=N3Y0=a z=^xE3`t1oUo=}Iq=t5~|JaNJ0_GvHm;bL>bRv2Nm)?;$&HTh$~tzSy%r_!w=ttNh3BgH=)vH`c;Y9+twy+QSyrjI0nDQcYHXrh`B4!YV1>uXKrJzcF zX=Op!>QM7w2Yl`$)n3e;v5k#vOf9&|ZZ3Avs=hMIQ&|qDV|z~_jqZf!wTI-lQGY8w z_d?Z}A`4B_N@M72k$qKOJ3+j~-Hu#NY}Y+Vch#sNv%a9ON{HMF?BFeiq6Zo8(@2+& zF_suKjhUZrm^r^*(aiJ_xo+tTuqqI5){SLI-FwA2bFYyjHE>(vg7_N=Y~``m`Lk*& z7;!kPQg7Y4@kMo~)bKci{u6HigZ}3XBXPq(OUf$mWh3ES+}sP{<^&2HdDa*8FF&hK zIrG~YRcYORR;J0W%7%^|y%LIS=djlPLzTD}QHe96YRa%P{&puSY=A!OQup;6dUi9{ zCFtQ=QK1nW55+2HXJR*SVCrD8ZS81tcUFc+l&!=Y`!!`+bH3?FG%QnyAj6#Bb%N#v zIUmlgQ-JV<&)~(a`%sAB))0x$&X?h)S(cTPo0r2%a~B0-NB*;(;33@qP9!hwr%i>{ zLP>J`o$kR_msJ=gGM;8kKiF$;GKx-qj+Wq*fN+Gb@M>*;Ci%i@U5X({uy#+tHyT0D zX@lf0;D6{H2c%o*NIYgOues4xxOK4z;_fP5#$OH2UeYl2UuOLj2Ql(m^q-(fa{5fF z6)2{T+)Q-ut>zNc4JBmhEsuq3HD*4JMD~{S0%62Dyf!lM&U_wtWCvIAJFL^wCE*29 z!$QLpH`5H+@t+zN8P)BE`j#_gBxc3p(5=Up?R1PjFxVMvbA`KH3-3;+e0Rl;A4sq8 zTCQR6eLwpJ?00RELBeFb?xfwqcRd3gytmhgzXL;YpAFJBvJ)IHmPzm@Cp-+y54az8 zG#C{x6PEY5Oy0gPls7LyaRpr-iNQ;*)GY4lYpz#us9xZS{W@D2rGSH+)k9*V56M;t zc7?~B_FX&cdIHHK=xS_7X_}{*rAQYmh4R-?qE2(f_}Vk7(b-)u%pSEf)UxSNm@Kwq zf#I1#ze4-iinrId1FRv01+Cr{tI*+(%kFW-mQNAw#x%+)cJ2f;yxL@L*{*?6-qUU8 zo!xxZ9k+N>OL`)i%XA|C1gDsjPW4{`oXgoX&AUgl zEuLXgu3}E^A4^le@ntRXUg1!?&d^=h#lvXc9Sg zM%DB;3EVUUW4W6QY$$=cfA0)5fneZ%m{) zCjY+|ey26EhHGM)C=wlnvO`&T6{*++>3@EryJT*YQ`<4>1U*7S#${db%>e z2n!TyhSvJi*~k6f%`XD(39ob&O1lDZlVI}LI>QZ-xMu@(!zK_)L(oCm!RQb~6pf7yE zjQT7{S4O|8$E7-pOcY61l_iBLrVG4F--8NtVn@|UlH=>a1J2+94an}gxs6+_qL zsrAbodtD~tDC}-fB40I_-fupp@W@_}kewt!&pXqGOi^;SiT0wW;HUG;+G`zJrxCW0}G+xs~e=q*$FRPUZ+DpG*&QHbk+x zL<;FqQ8JyJIQpUrw_Ti&$(XCRfuTqi8z!5_cUeKsU-lK?2G;t zD39x1pRY6i*-h%Z#;W0HXFF!oE(!SepVjzuVA>qW`Y;#P5Xr*;0?yqt*T-3|L`5yD zI6+4LL9#YVy{O?NzHK@FQ3hA=bCWfS4-Dmz-CaGks{W+9Ovq1zEs*GJ7s;9bC^zQ* zEoB(&=}IV$YRBj7spX_4AdAC+|8qE;4^{8;yJrIr_}m%l4BbFbnAcuJ8B}h3wrhFP zr2zH7Lz9Q**KGfT17Au%!GT#dG7>9kgQHZ|%bvek_7slLMUpDLL^Tby$5r`t?j4qS z{4|dGDlZc>@#F0axZPLgIxBgXlMe9IxAdL8^S#g7x)(%K;lL|gJ$XNz1f`aAbJNEe zXI&I&CS!dfB8E_w;vj`f0{cq6-T05U*qlBS?!PIf5cKku4!R#PAAx{<@1cH zf^y>*=wyRt3pCGeMT6D+N69q%D^tIc2;s6K<+h`mgk8?$R8iDaPO6OQWF8G~8OC48 zBeo%E>3lfsp+n(*y8K@?RudX3Yt2W~j zwRzBdg}^-UC6$XeiHD`&5t*~epA>}f-?!?F&ngbVr&=>g@tIbMM z{(-Q(p@|ruQKw&WEP0n}Qx~q>yZjkE{Zwnk*xr}3yXS*d>OhNG%g-%hoIK-C-NLbI!=l?mjv2yZ*j=pD?oaQJ;|H&34W!EkbH8O3aJq^d$Hw^tRN zWIp3&<&81x|Mt1e?@y4;3#&J3rQZTyQE~FPuZoNeg|vaTpFp_?K`xWCPrOu3edoa~ z%Q|1e?$%G8shX2LGdWALGOmum+gBt*1^vp-@^y2X3-m`$SK<4EztGr5G;DRm3QyM# zM*!hVm-pw4?}5Hzh@qKmk)#`vRr)Nv&G`<*9*1$zR(I7)6(`%SZ&w}RW|ZS#N8|NR z3SPwu`(pRw_*(sMmR+PzyAmq^evZVv0amI)swO-+q#4b6o5fZ9!QvUuZ*$a(e+O&& zKA#u2h1{R$ZHIz%i|0b5V>#X@bG~FIkxt~^XEk`%wGqZdU$u>FAlK9Gj+NB}8@2xU zETvWUJpYg6&?i$(l}A20k)~SP*+#rxo;ktA#{oN|Yi9{&0*!2glHSKfKJ|C2de)A0 z>0`BH0n+&Udp%Ghq6Q8-6Py$c1_=&tRGCgGB^KFqp+@{!9IJ-!r>o0d?w({8R+#wZ zQdn#iVXoPO^jBUmbZ;YR!XrWw8Os@Vk-^^|vjY z4wb;&kGIpudU~QiZCwr-^xGL-NV&5V>dxX=G=o?XGwXOIg*7NoHR6GujDU7W*hFTm zP4;p4i7Mk%+Fz>7Orxzi)+uJqH$j#zps7XYg%jrSh|lS{J$Fgx<1&DHCpL%X}Q3id^>G+RpW{Lk+1FTcHj~G2?2Io$zq5nzcr+Ey-c)kTEtaeiGL4T9JI3G|}Za-#Vz^DVh>6hL2BsZNQsW{;4Vdxa}QnJfpM`#q>aas1uXo6;K-R_( zJ~kie7$>A);K;%ILYX_Zs28YH`UWL4nCW;{NHvd_q0HX+J&MtOH-v9rvwq8q3;8^Q zu+xw~kSgY;>t-ecot}Gy96SguwK}BUDH@z!wLxy~UguNHTfH?|AL5vjq5?zb=LB@v zZBXnbrf)n8iohpVNsqRAJALIX!mih^op4OV3--H{gz5=F)Nix2qYKP1OTO_od4yNe z0hdFVQ|97YU;jA{PjRczYw|oGbffC{!z8CT=vzCqY(b7m&wEpHK8ac~j?L{=Ls4o( z$nhH422P(ic-~X-{Wjd0(brt^`V&{qn_LXAW(8X1N8AQo%n4X<9N7h`!|;!rvwThh z4L!6AIw%!ObH}}6w zsg*TfdlU?lyK4GXgsLAN#rAS#mEu;oP&o#?x)0Nw+@lz5OPL~@GZAD;l@?*L%^d4fvD0^!Jsk z^jsEj(z_$ADso@lfE}!Vb73s~>0nGiWqS{lP~0PgV^-32;H_Y$g=^xTo~=b~p?bl+ z?&;FKzruCCx00f6Ww44`#S_JVB)3!B~-{Lh4fQWfjs4>I@?GF}upZ}t_ z>o6Ku#Dy|zf7&lL&+-mxxx4E`r};Y7>e4oPv^b%g0VYEU`}Rd7vVDrYuxzD7R6|70 z)6n7NV0rS;#4PJ+^xFN&pr=5zYInplRsfi)m*BjErF2nm6 zM$^_PtFSQ)931!yOxd493-@(W!MkDd$Tw{_p2LKM*ftTsmAE;rq>%EB{C{oP;f8hT z5W3o@azQ6cjH=Q3sE~sh&bk_E0le=7OLcV83X+0~k3lX9mxF{7Vd1_4GeNLSIjBzW zU_9&VQ$$hwcy}Vt9ut;PZ{Y1)3#DIEMriRy3>#57Unr0x+OVPZ>4BR|NfOlHd`!*y zx0&*EG4p=%Z|UVRQNvZ3@JH=1)d%YdyIml|6ZPb*)4x!h?l2O~K15kMBkibEJnZ~% zZJ*<#mi%BG`!th14Z_eFpfN+{Aj zFHC!;B&_-Eqv-*w_t;l=+1}j`!$b}a4qaOt>Xe>Aim6aawF z8{5ma%;8>L;n@1)KXEU?oCP;659T9_tgJ*$_iINDkoJfIVo_1Zb5zA6;%0_cj!Yy< z^hxTJ%Uj?8pQ$}qifyMZsJ7_#VFjGtu18)rAfR{4vVG^93HS>e^!ej2<#dQc1e~_V z+dSR66HKuwg{6KzfB9B7y40Xe@X3YrR*wqc#UDY+^G1)zF!)-221+$Fb<}W*N^A6G z{Mc0L2DI;>Gh^E1R@8(SkMWQFqPUu`|EEopImbIPN&q2tzTT?(Ua+i@s(?NSr6`L9 zJXq?>b%dLLv;XH;f?MESs{JZP+2Fh>kCn&fn`vs!q^oIkj~UXzXtzMBq{@lH zUL$Xvq?0fffaYqu_k7KhZ5rwO42_uI{>Pe!uS&j*1EMdY^O8ZeK&i=btJCGwd%Rso zK4=KRPo{_eIYt?BMfp8GB4praV1Maj)338N{n1++x)?Z;5e0vEAr z0yf13^F;w?SX5St@!R>1sDiMTVY8 z^hZ(%xa_N~A#Y9*dz1N?PlkTr#qPvUQ73!*A5WpbKMxFJGVerw>$f~{f2~6vUBiedq^Qf@e&;S<>|bzdf97*Ji%r>R>X#^imH-`2xtRu286or4`5 zl62a2Bv=T|+}c0F5C-oJdP2$pj|(p;q0CKZ%6TTQ9ck|tnuc&14(J7Zt{5Argtzi{ zJn!pf=6axtF;m)um%6wz2?G+f_IYmzac`oj=ij9&|lrBlI*nAfC@sP*9)^d!6Qo=F=^Csk&izV>ms`{U&XZ_yLI!kBtYZG+Bi)HkC=BF4}Gt4XC!p{(Vp zDDa5#qcGt`ptDZX&sEVKtX>%HkT8j=@cgB@3;H2ev94OaRhM5m8pByPJ|+jm2Q}bD zNol;bimBc3C^O|kGSryIq!&S$E;Lc8RdRw7*h!sit}5}TM)<$6;d0p(^J=D+j3fqD zl5m==J(&=H9z-`i;ZYKy_)mB^-I@VpJnL&9%Xyy1vF3YGBCP&$Q-mLi`*X@Moa)=O zT~n0{Ifc$!qs8tu3vqTV*ws-Vt==zXY0iFE?htl%vd#p*@)VUhxjgvJA0O=0JqqjO zrwPBnM=oIBXUz_ndDNLGEx7&R@K3tBhvgI-s6Md@c`R>R+Trua60rRe1IGZHiuwE3 z`^ofOWkF?{fZ&)sY!M5?CB34VMA=8mjr{x|IFiqDX3UmXi~LJ+ExGBZ#`hPHqp^de zaF*Anu}XA962H0W%qx$1SaMeC->9`S5C*p*?b$=&v4h2J~?x9 zGx^k>MrE|Ui4xu|AK#}>#?%syUpyJd>K(}Gdd6+`(hEwW`XuyZTh>u4Y;Sk>mPOgj+4QA7eRVuvh9jV+Kw^a-hX-M(dSmCIIJwx z_&BWpYHDe@3K~H=C=rpfN85O5xog+B=2xA18sPBdNRzAKO~St9{gm2uEZ% z7f;q)Wk{gkUdwp|9kl-adlOgKhU4DonW{#t1x#c=`_b`V{h}*3?3$}tpPud*LX3Bz zPWwSO8R<;PpOK82SYx)Vqrqm_Ia2kor&ocW{Qj5PKcF_-N+T%@kjCtsxW}T?}W#vtWJjH$@tzE?=PwVy#~MiRklTLh-q+3ClxR_h5x2Vc4fq z&--vyEHY{yE9#vy8Z(1F5wLT&pY-$uV1HFDUx|qs!NAEE3PSm{~eMutfczWVGRbeQ4mT#mub8miJ zu#o84a7;FnyT^1Ci`=6)PCYOvlj(mvqYXLxrAo)}eHh@47Ie#_k2+lTebi>Z)Q~%e zI6@`2ZSmFi2BWDnjHo%E{Y=oyR3CDdNku2rYDAk#y*$P1;Yw|PQ?8&B z^x7x^OJX>YCV4O<7hD;8*`F)W)ku;e8N~0=zj?1=ny-{i6Bq%@{siosB2Iwy! z1aeoXXyt#9PdRB$P5$?<_f#A_yi9{o{Xsqgh<$!YY&;(&>9FM^+lkCE4@-$-Gl+U3 z;?H!u@Zu_#Lo;$8&4!Qb!mSK^djiR<(zgjD=`PL98V90Z6}(C+_+%iwv6t9bKl2er zlPO~qVRN5#9cFG7wqLWV6l|gL*k7NXh=nC#{aLR!WAi6YoS?9xH88Wc`?-%NhmQ|R zX~n6liO7HU(cGHbB);*V(f$hHbwN}3{=A>gR+uWgA>|oLBMNWJ3SP4F8MRU!REJ1W z{ct1Z8>lHsHwoDmcn}Kxi{DcnL8WHaU(?&C^Ty`XDU{KN*5{*!>mGwv@$2OvdEj2X zIQgR+#Ta$7QiTcKp<49I18(8frZ^iYmVDb38~AczNj@Nj4{GOh=Q%Ok&FBb~SXHr(%iOin3^2 zRJb=HhtG+b>9xfxT>bi9E7ge$y_xnOv#rxhp{{WgaX3UXHd}$xt#g91+C%1%A*moH zm7PHe6~uCwC?8uY8SmqJ@79rMRmzVMa%xQ$3k@9eBUrKT0>yU1DhxNbrV>r%6Ium= zpPsLEt^7`DWE14TrN?a!1reO4sJ#s{dPNNe!+?K(bGv8;c_wE#EuVi-qPci48FzJGv+3+8Y_*G&q?@v{9s=Gb!66%YlxKX1?s-R3{Q4T>|F*Y16Dfp zW&C`hVs*1FYd71ko2G9X*1?7;c#HVpO1nA#iSvePHw3<5f)p;NnuL~5qJYPl{CLE# z6O8-zahEN`ouW13;)xGDU%@bRa#O8B;GIY&U9i3_&jc)w^oknyc1_}H5zIU*|NWM% zz}e>psqK1DT&*H)MB9`2Lp9h;9@9wBw=2jhjHGUjGgA;4y?VJNw`h#~0TM?E)(ujM z<5*YuFnwKr(WR`+axCd;KLxJ@y@n7K@KqOD&sYBp6X!DjP5K*0hfOo1{q?+)gx}$& zjFonl?)mm}RyP(VcsqWoSU!UZ=Z*jNgu)C(q{4=~qw`UulDKy-i&AKg1GN=5-G30s z_5f&NBD*OZi1?e#d)vZp>P*3@t9SiFyuL4Y&rra`qY&D_TaCPF$ZS!2b(()`Z*~CC zL(6=?r}=C^5Tq7RtO6*{7XG_~j66!hCZ}r`r(&xy88V~9prT*AcTVO^5lk?ARXI2l zJn_Ip=8<^d-P+qX-7ngtAoc1L7)v|?8+#-z4Z1AQtT0$l0K{IOE#+WmJV~sIzg0BJ z5?f}@nY7EQ>gb(_(ySffGZJVRem388PPml9Il(ESBEIKyu#ouf+Hcy709+i|liK<}*ze zMUU?8dgQTfy#capz5hA1_inbr9W^PQX|{iIz6AH_92WUOb$Kh%0p*Z6>ORnffsePP z-QUF%FDE$k9>a=z>X zIYmTSjKUPIBU|Wk?AN~BY^^@*V6hA`=gBCJCd5nG@XlJCf=?*4KJRVN&=?7Gv8nVF zS}ZDulW=XvCQcm5kghO$w9fz-4iDL{If_Ish<*`waNTUgh2R;P|8`18@=#JHcvFpH zK4AGKd%g|ed=`QyjZ4$RXRhiy*1Z_Lc#$mj;D-iWoU~n1IM(5=^e2RsT9M5C7A2G0 z4#P=$te458j0?59`5IoCuXY}|8EG>&B^tM6$mQFP$%?10H?``50Zezr6DElL!-z#- zmB4;gsnMtS?KvD5w-r`x^>{vD*I1XKr?Zs!tn_66x+jjbBUoqVb9soV&6#hO)6Q0D zk~_P+ulOR}CPzVp4;y8<`44C^IiHZB82ba-`e5<}d1XQnT??*|IdFlWj@$Y_O1R*Y zfU7M)J>@w!w~JcTKsMH>Q!?2ClVwR+q#DYQ6OUe@%n&cf{mR!oTKJfW+pek39FXh5 zoo~7K37-K3^#{H@lqD#%e=SMsO5%4uBikt9;91(km19a_@9dnVRwWi!HW{~?KI71}ac72AQE_52Ryj6RQMB4%obP|7`q)m($Xjst zciy_~kaC+6U~FKXz-ig@>ice@a?K-8K{Rq(!%QxDWOFb!ncIoY%^$9?Ga4TOe-5E+ zl+zz1fkt^c2JC1jL|}V#;O3g0pVUZY3i^aQUwi*4=26{S{f$i^i(g<-?0d~1sr}&b za|3(IVkRATOtmdK!o~Nh-IQowuuk~X%iCUW{;ZQFE+-B@RBZ_kj(aino^U89KRI2e4#?if=G z6ge^!RW&oZ#L3R&C?;-6a7Pyx50gbhDdkJ|&t;Wm7BCKT0Uh{EaLzbO<>IwQ5clj>ngZflQePF88V)3u?1M}k6Yup~K!jB=}z^iyRE*q;%7pXInC->G#dDSK#V@wj^NTQTX0UPg-`+@O79V<2I(I%~W|~H* zDtQnxXlNKF@NX!U5I^NGWNWk0&*-~f5)9#;a>ZDTDk>PohsO<$P?;SK52rzhdkMJT zs^|2~B_Dk+?G$uP>okymqmDjV$ot-O%9}=EJ9Mg8btg7arx#B0-wURAGM_%aX>&$+ zqeEz&EpgpT71i|zwNhH~ZJVu;wg#l+);NC3wXn;4?|pW8 z5!~@@3zZN1J)JMsd@9r^3OGsJF|V{`A>FYpAV&GCmNG46B@&&Yzh?UeO?WyRt&SAt zNv_Ki=vp3FeE_omnfJk`_^OrxA5UL>Zksj}WM`F*fzfyQ<7VMpnD4>z8KBwepwsbt zU8!PAG5sq(+7{IM(Ozuuw%Gr~4L|=s;s&n;FXb5lU5-k5Tkm}g_&ON$aha^vY;!1g z$=gcjL+NR&TiP_FF59RfK<|QEx`F?RzpA7Fi38reiOGH#A+j z6PtQl>^KRU|M*mGOWH`THy6l^J2JH4?S$5yqEV>4np`I8MKIe?(AMo-)RyUVextFC z1(F^`K4`hjbN%tuDZi7CBopK&7xct2yp+XOw3hJ425LL3F)Fs+i}#@>icdWpSx#Ba?)7k30{MtC2k2dl9SI* zz`a^aSA(*y7aPH%d53L;B_11QnvW{SVWL!WTnH( zJDn+up1(&p%sAK*V%gxaXcCHI^$xQ435S~_<%O+SJP*kL@27g*2Pzyu zHB~BGSS35>x&G`p+rz!|6zq~I8^pIET=cSz6vwoi1x~n+S`unZqYZR2>s^ofc<^@s%Q=zV?!eLAMWYs>&M_!VJTAw z4WYvsxasnOOLml8&Le^Ydv|~$LrOE_Qe-VviyxY1=Gq?BS%Xd;k>YR4AGd}_7IWZ_ z%BIOh7KWK?1Nm#AZ(fEB&vdTC*{dw)_K6Jbam+E}(Fn zSxwZ}+yrs0#+^@4c^HnW&-FadZ=Sl83)2PChgHDoMlOeTt(W_GaM)cUSAn-kO2k}W zn^Y-`M_80I7Q1dA?)?YPHKVB)uC!(JMG(??u%$nCq}$b~fMo%2wA)30fx6~z!vpm?-F8sPZbr+;}i7e-=~U>@9SD?RpPkC5^6`o zSN}0+`Dv9DdhC13u%yLfhuj3TOa`}qshQnPcf98)gaiPcdoCm3*+$3c{IwRsp_ zqg+1(UjmG+oIkbeBEA=GA*r!WW556HF$?@Z%DyTrt~J|sClCk;lHg8oclQt=xKp@W z;qIE??rsG^3U{}{-66QUyX&p&-QB0py-)XLef6-cewl097(>=)3Vpp*fsG37wbuq` z7}m+jhRbWq_gyLCs>9A1_w%uuYf<>?X1z$_4nJ{sQXYxdvE49{QWBm1cgBai3lhzY zpz$_moY44!uh%@+X4ARMNam_}S_6v8*1th1nw>R9s@gZLrC(6?!bsgA(YOrdJV&FowWj1c@=0~$ zEO$Q^GA2v-@m9!4BC}P2Ruf`^#q2BuNP6(QKzv{WnQf0f=6Y8b7T`E%)YoRz6#YCQ zpzj6ZVKbff|J>XupkMr_(`B8Q?+5Z~5anFMko>#1UT`c?j zbF8yrb8ONb_DcerA=El|A_mdIxsZNA*%sn~!F4qfrD)hh{`L z?TvYtzaGVOsa=1$ki`*n$W&fk?1&5#LcruYP^s;tK#Bwym1@MK2R=`@5dt4yk`#9G zezZIsw3{wCr48t-PI+dL)oHUR=1GX!H*-E#0|13l| z-G4wLfrnh;$g<*tS0}WURXniFO;3n05~qn1m^{%)ki)Uu&Z^82=BIrTPbP=$PULiY zQt9C-wjjf0AQb)`JZ5sa_uGeCRrmu1Dp*`56+MJ$#DpgT9+&~o#NO#E@%~b}+-3ue zp5}{8E^dA!e#p4v@$&0-*boybVBKSv==%t_iRn+w)J__L`;WOSMpPCrK;76gg0$KO zg1Pn;D#j}Fld`>UJ@&@bnjA)zjH^G5Z%H88VSifY4n5-;6BdR;^5hP^hURPqtDbMS z_8thb17VO%E;SWC^n~DAA{_%2w%gmJajRE%w=;VzJwLhio>oviHi)*dc3?S?)+o*DlD_%2No zHTw4m(?kzmI5SLO@xs&RKaW%#q(~E5`xUG{P1k$EZ$ z6}$yxPNhE)i&8n3ZfCt+yQVzquTS0lcu{oMj3+l1X(aS{4sMMduPw80M346eOY!Pm z4=^|~O2n0_-vPck?MDjD;1nL#1WSxuM1MI_RfcqpTr<>LFK?gIp4tkTM1X%Kj6zFDo?L?{@G$FS~s@3~%hmnXNbvFxBE@Em4!kaz%!pFKjTHMk!kijd$W@-6a>gPp;I(w#Hq---6v>Z?uS#Ingna4Ir;q!kw+gDSQpi1O{Uz9FUFU z3f3L)^4gZvH02@Y(rD|cvB`_F9+B5Dp(BRo^&@|9C5@pm{^dJxA_}nnIMiNJ`1+Y0 z5vMNZF>tQDlZM#s@TSY{Qb88P*ZtD^$nC25>Qz7dE9#|@J9AxmyiH(W`14Q$L4)AT zR@pgkE7GoXl==L(%%S5C;}zg_tw}}*C%t5>X@)GW7AH#1YiF^4x-2!-f+f|*I)e4# z5Wgpnvxb+~3_6_UM8%(d#KB|7?1;EEsr_=4^l}`s`scN`pC}2od0^9FD7n3^bZ?G% z+nK#^Ep5$3JGwi#zmHTC49(pPQ2xN3T|{BN8B;MPn@wu_lsw*K1?=p8Zb4k~dRG~1 zlw-Q>UUM`afSiEngrI?fEVI{DgGMN{3r#h+nP2Pg^una`xF!2s8+Rc*ziRFvX1&;S zILYJc5TtZwxg~X?kRQ8*^lZ+4;`;$6(Qz(%7D)O1C}($bCvFZFv9iQfx+DI)T&>KzW`Iw^sri8*qd{!Qt;(2!$dR=*Bgiv(3>FB1z)ij z>eA27BdU;!lreTg3gmJ-(?uqneT6t60XOK-1YkeN#|{-Jn6) z{nw+b?L8)EZV#BzVwlG{k2P#cl+wOJfY7(9d8InWMch_ui5^BquE$`I+R~&TAxB}_F=>Mk?U3ZSRyBJlk2L}6Cr0%BczC8Q)L0YT<&_f zcc>}m=f7pixIVr3aBep2wB;ur7>+h zyQRaAt>w&4(&-V>oeP2}C{qoCI$vgQs@5kpSviX)jAd4M zuJ3gFy<%fdmbp!kk~=3+Lu1_0MQTyyxUY6q{J$9(_~j>Hk30;``fc^ZoudEozQurj#G zr&Dpqhc7qk@_g=6jhUYNe$G@H9OwJi>4C}RSu#*{C^*GH_n$}#g>B3gQ(20rp9j#y zxG1l5IJ&Z5&4thRT1G4?AyUcost5p64>;?SZflia17)y%L0}X+AQGb8Dr&= zTd_<=nc>R-^Aeil+1IA_gsQg|YBNO)I0UnYH~@g{B_V91=$_@reF;jtok+EXZVj0@ zv^KbJgR>-6{p@HlS#lMaCsX+5s4S=C#eGhBfg|`lTVYs5@aB$%Ra}!zUhq7Ql!{)2Pn`-NVS;4eUbFn5}6LRI*Kt)8dM`X zp)TUfl-cS}9s2aEv|`*h2TB%Yp4`>ONtIIpG}>Mi3@Km1H231hhN+{K5~*NR{(PC7 z47*GlBsuF<1cbVjbJE=b@%v1}!_2k)C}ktN5?pIsk_-S34KF5>0GfA8li%^HC~EKI z?(gO5mz|o*iLa70$hy5%5Dh~ReYV6hDAZq_z?9RQnpIkTU)|MDDPH4Qf+9l zfH#^lV?252p!~^r=}6h!f-C7tYH{H{hdniA#*1+ZNJz2eI4CoOW1+^?fI!Yd$+QD7 zUb;dryrsOH8P1XFfDPQt>be}pr?BzHl;AJMlygCu=4do!w;->6dGCO%Mo9s@ei=@k zOAhbA!17k*cDhmWn743!I0|^t^;0Vpe-vee&1cin&z$c>?m{3Mw<~?gC70e)4p>!8No0Ru6)+>DaRQBAIX8y8$vtmzxcJJs z0x{C6M)8iiIE4JvPWgRO&I?x76P$#OAuZGWJVhx!_bj z_E+AJQlW-Zk!so3LIZN!g4KYDFqHE(FW7g9EI5$_X)!}r&!v}>MM|Qq{Ao@t0_+iW z*BY#@JLb-W&F#N6>l>Y}>EJm>43&c(*a&MKl@=%RNe*i#PlJkHUE1D@13uAc3;RW(PzAX1o6JiXY z|K^gJDXcsdh%0@ptLQiTw5-XQWs(p&x45WOsVIY6cQU#r5jJn9hax3b`^)qq3so^G zi_I^Q-oU8OUD>_@?3uT~cFG)Uje+{^MSw0%8e^Zun)uL#Xnp!#N((_Y%tDomt4O+E zSoz0Xq%@z~?_`~%mwodNDx{fQntY!-jjA-CdPe<=Xu=u`#WWfWf~(b@pC1w|CJJO+ z;CSeb4j~@u%=rX)lgSq6=&9E>UUS~uv0f4Won-=-3~;ZdFCKK(`rao#fBS#>nZ0OZ zE*(gN3EkDztd}i>4s0kKl^a>SLZE^#3Mqmqll621Bu!4L6uIK{)QH82OY%|KU60d} zGdfXZYq`J-Ful(;gAgif#5T+xDtfkhsaEf4k+)+Me0EdtgM-y390PU{F}ljW8Iu9_ z@2P;0z!Z53n%+}cANk3Xuz+2G`KZ{F1s9t$w#cs~WLAlmF^FM!X-7zD3hXW%RVE{n zb;C3(9?7TvpU7)zXRTW~-~?C!W?u*T-(FNYC>Cn%L_2JEd5{$CAl|UhFZF78B`s?*igE4M>~xsS;Coy37pvr!^83ahRf#OUS5Nt2 zpzR?~AF%Ffw#@CYp)vk|mA$^6kAZ*1gfAxgZSaLVz}whwEa412By`+rkf{jnG-otS ziFWq${l1aRNuAF`v;Y3IdNw)TWcWg(pPYdY;j&NG>#dUM*#+W$Pf;*X>?yb!fH<|W zzVma2f-%E{QQtAnmzWeD8}_PZxyhAM?yZ~+Ikh?Rc6p$BR|GQ9C?nFGw0D9hXg$J{ zvC{ilG>#9vGfo>w{3VJHMrG{N}w53X* zAowcJ^5XqRD=+W8!FYXu1mTlti)^LA@^!xlPT!#LBE(&ZL_%bBC6Z*sJg)C-8RI3Y zfv8Cjl|=C~NOc2jpxT78*^K_>;1P+PXr*mqep8Ik{Q*^Owcmyg#(HpamhjuFr2a0v zN^Fg!;Df4!*-kI1sqC=TWgl~vg+(eyft+9d@ts!#>!F+=vf!3TyGO!n2H_28QTG^H z@O(S?#GLH@S!C}w<7k!j;9orKHea6uN6cn-8FxmTLk-Uxz1q@uwpDthgk$>#Y z{BWwh^q!qrDp*ZDm8nL|&M-`oxcipgc@jUqU?v_Jf}%HlCn24-K5V(0BAv#P8RO?Z z)o6w!xhwJcJmOUqEzQUczhLpk1uoM0!;;*t`cyKTH5-ga%`fM9EwZp3nD|Ksw4NeP zumvP09K=Te1OoHr2*!fFwarv&sc?=vSMM~*>BaqYa#DDPQc{%qSPG;FK73azzN}T# zPpW(mNOjI08?Thk6p(t=Heog#gvs*Q&*Ic(D9z!l4y^m)?a6FV?CY^KrPIA6yOcHV-rsnLlM3a0k2 z!t}dJKumJx%iI;4ROLvknhE8o0G1RvcH_`cWd?lwhuJeVA1%3*Lp0R0X1}0@&*=jO z&)+w1$R$1j2#lU4YX}7ViT&VNUC+KOB*~iGi4H575>}nA$!xl14bB9YIZcYDv<(_X zI-1?2Rod0;tP9;}N8MR=*1<;L8b>u`s>lHoQp3$Qad}=%&ub3Ri325_|A3<@Idn0m=|b^Lzy z526K{sT28>Fi6IfzquKN(yL>_Ec=L=6y_4%nxQ;TP#8N^^%n@xOQZZke^-l3sN2+j z8O#$`RHaxU|Eu88tfVxe1AAFa@F}9>ujjJ1w!s9k4x%g)kNv9LTB=0Cv{jmTk_gfA zXG!lYM@m^C$t{!*=dfF1ZlFIyt zn~Me&0_`!68`x8Z*z?V*P1r1E;(NDQaJ->aVjJxS*S>}zBbSm;V{jY>PO&n{o#UjL zR^#3UU*})tPBXsflITtZrA~B)kx)5zB{|dtfy0z7b_j~+cuIPbuEx=~>2eTnXH4%( zrl3u7k_H4^_(5^{mjXos8G}mvnr!dEu4;p6x~HndMQ5!Uu4@}7SmM6q0UsX$)NHBS zknt+5sp_Sm!Kck|AUIZ0I4GM}#7%0&_emO<5HpzI&Dy52=^hgH8Ee6aUakMACLvor zhsC5i%JVMPxbkk_^Ikme&YKENv0;dBW_VwsNBUgG=JeMHjgYlDlFo1uf>UtTrdf0Y zWMajQ+2{9|=#%L#2ZltLv%$mD&-jpmtUc8Q@Z3ButMLGt28MXiYw@1KoJoFf#XGM0 zF{}Z3fd}E#nh!92H6bf}Tp%iyL^x>;mH_8~oL8^6lk{#AH_*0MF~8wNEr!3lqW(15 ze2yGMZg%Z(`irMU2g1Qhqk%ZVx8XgW{L5Vub$1Yp<<&$veqbPtH=%Mv;qvayEotaX zVQdUZ&#q`6TNGiy@NBWT_>5&Cec4sbXiSo2jLY0Z6Lg`V>?JFJ)yn(c)-z;ld8t69z70d7aL%CeOBBybR^GE0hNm&l(v9tiSF>3_iMQ$QX9_1zNU?H;Q%ADgJPrcRgOTq`zTDq6oIqK%+M zAyl&+_4XXvJOpyrSilVJ%IJ%L>Za{2+bfpk=xn}R%(L+|y>=&aqK7!d^j&;QLw9pP ziya|!5Zw5dW<;&1rWN`AoSMw%*Kvg;=3mLg+=KHMvor01bC1oJ|d1L}GH%ORdfdLzW*`p;8=ql3ebG z6&=PP<9Nj|kJY;ex|LsJyvGI6q;}9m^Oxn=9p~Jx27sl;9A40Lq7L1EEgirIMe>h@ z;S>B`{axaLh2`M6l~Hh01; z*V@?nprzNTJe(#gzMkHSK0dJ9M7-I|%?-AdRcF20whGPdz#-21IlB1XFItlJH_&M` zl?E*gptR>UOd1j|Sv<%2vp_ZGKCos-)5 z+*N76edLVhhzLw43J7b9csBHGqJ-Z;;~;DCVkjk!YdUN3E&gWz3xB@v^tAlV8X!LYPGeZ;1Gg$45u231q=^B7Kr_{-J>fN7hS3HxS4}q*@K3Ezva}Wbvkcg9@V(ZNRjt*IPnT-Bm|0FaLOP;tNkZ_lw61QA+)M8lUoP+-;+L?V`4K+ ziw6D163$Z{w+|UY&X-N|<*!$b|(J8`#)m(&GuqgCs`IR{hi^2LxNH#?2<3DsQ z{&zS9EV1Om{b*G?15RblLB)zua6V;>C$dl=mS1AN;r;B8{Bh~;tQp+m=gO+ONEOjv zwY6q?-y?25&a7M1YQ%Ay8S10);wBt?AlG%I+1q>gImw?alW2~pRTY;(3+;-@efy4u zoI)j8IP2PJ^Jaj6j&Leq!HH+nbCTYZiHhfK;E4`*bYQ~axS2mItxM)wYH<>C7Pz8{ zSmB`LsoW8edW}2gpU6$OclXJm4a|4H-idi5fEbzTO#-1kn!C~6`%yT<2&Hl?nyZ8~ z>-nkVFb@D#u*LlAg4t=}HnWt96@0MHc`$IQjlQ@(*kFW2rJ1=#!hf;C37(@N(OYU@#gQexkKR z&7+WdnN&7g2|GvAzzwQ@xu`Lnez;t&$YZQw#$o0&e_&a;a=e}0WhHJ)I4woQTK(bV zTl1t+gd4&8)X~14gTrZtTb-Vf!fuEg*BhDTQ!dL^Y&>RTw%I2hSSgW!669{R&#i}! zmOt9=R>wL8mem`sfWl5RGEnE~_8W1l|72X_=v2AQYY#epD4->CbfwIXRwoOX(Fs2~ z=9?vu7)hV;^Jk0nq;^9Qou%qn4~PU^xn`^|Q&zn-65)iCX|gg!ifj(ktwsN_c>2_M z-n7yDTk4CixM10~m#eQoXTlzDb#Y!-Xxu2E8M0|dTB{0nxJJj8%bVVd=s08>n>g!S z8coZ#HAH}l-Sw$WNzKLjb93|;mH3(!R|so~56YjfQw<=ci zedi89?mC-~#Zbaf+rA>Bs9UGQr8W@X{>XXh>jm8%S^d!0pT9(@%_kJ8a{0cF_%zqs z^*9PB8(SKri1Z5FwY-*OZ6{^bRI+pE$nULeZd=_G(m1ZPd?f7E99x*dJT%4cWgOV? zc_;n$*NYafTkNcU5Mq5YX89wEtMO4tM*46H8a79#VpgazfrO1YIRY9P(~C|*HNnUF zW4*q}J%+_(HAW-X=UXh3sk|ZBx>b%Vo;q3Cl|lM@!)ayAl7-GTv|TO-OCn9xQNxR@ zNaYbsxz~D7D&a(4cQ1!?Ke%-k1exIa?D)VUK$cc`bxJ*#(H-`G`G0b1^O4@SxcS@wS*ad8=xyp3?+wI_d<+Nz|#cL zQGu0c=0AGN2&QW@A>OKVal9DwS2)huZ3H@zLx99 zd59g&GK~=;j?m~WCM~#RXN7W0`}!Dzi@wnuyG=zj^ukZQ)rnpS-Z4uk5A})B!W`BA zc;hL-i_uN-&@TE|%s1$L9f|NVGo4P$T&MA8v^0M<`H6-T<3qE!(`;$!Ns*5IlHTKU zG%w_u@P}4jMv&lcqqo~Op`T%VP_|^sBv*Zf?4_?53cX;OBN5U0TH(1RNyKZHhowl*=?AbDACeGF#Xw>j-RC|c zfi%E^mXREG3%E^-+p8nDPsOOF#jde``P2+{oQP_f9tiC>PuSzILdYo!gpGT$&Rjn_ zTx+e3Plbm^w&mqwF$0bfk)}n{w(0Y}^(K05Ge4a@C6zOWQUJnJ`isLa&|iF9B7XEi+b-K=So{${G<65yz4hDPH%iDVg(^(BZ+hXeROt9Z?H(>o z$NA)Vsp8x<0frmEi#ukdVw-&Sw2yZul8$o`jI991UIeU zC)q;Lbk5I7iG+LoUptj^hY}*9EuBN+fbTK9OA`I6SE+C2&Kld1ko!>}eRuA|~ka0%&7MeeJNsf%?7U35A&3*0}_#_8)apmb!+O|*6`(jM4grUv$v5w{iq z_U&VNg|+}oA4XUiv}(wZ498ufDKS#vkCwbw=7hBqa!?a;xGRgDb_SIP(VjSYR`3^l zA&h>k0KN|xgQB)+u-8*CmV#y-nUC6Aca3X0V>DGyj6Q1%So%uQlL4D}k$%Wg!iRJF zxPb4!r>t$RD)tBb*j-FdcKwp-;Af#*>C$U3`VNTt z%&pO;^DdS~uyrjsp&m8>74JhuRMK4SF9XRYAAm>b1tu z(#aTB7hfu04cttfhG6m_?)+{B<W8r=sLRQdDyWe>R zlZ-ALULOLq)tj6c#cVBApzsbaH5Up3_p(L7IaB(pp2)A5N~*^Y{w{YWz#3I~cq%F&(){9joXX_qXM?yH3gs;< zy;JXSCl;8!1q}X+1@>={3rSwJq;N*IUttVI?(S+c9_`+82|pT#ZV!K_=}$(vNXCt^ zos5*j4P(T>ZpO)g4w-8Ia;k96P~>AN11v4|*9R_H0uPqcEeBSNb@L@gy3H}C3*gQVn8&uc1^MR`Qe!V)Ka@{?S!aTPgxA1D>CHQOW%P zk?85#S&4WCPvgt#mjz&R3gjaHem(P9*WN@zmnU)|s?tj22a3R*XF&3B{Yad*)?8JX zM@I+&_%8r$-mAaf4-7K`TBUZbiyES&dYj$Ib+3d%i-|I4xJ|x3@A{;P$J_aIXI9ns z_qH~R;xEZ|n5XTh1&kh{?AbzPX9||ame2P_{jW{4v)ua|*5^@?|@KH zAtv+oYdTu>i_wz6ZB~0N3H1#UEuP<#irJ)I`i%Bt{PLx${nQh}*buAK-l84bqcG^K zh0~*xHiUpO@%+it>zR$g8k9P;q1falCJBqR&@_t=8js1ml)dP*)!_H8E{jL-WcZ2DZMX2M2J3b>Y&M_(L#3akx?fL(TDF{*lN^y%wMv-o!;(_ z))xLQV%$u}sx}{?NgTE_&}-J1w&V?gk@FRPD;Ya~Yiv|p(hGTxeH(aS&9{XzNQO{$ z^3?8^kz=NlL%b{|ZP+pBWU1C1`)=y{c&m_%hRj@4u{V@wA})rE!g-tqXkhWDb9HTna7dsi_}%#+$FH6M;~jo3#I4;(yTo(}=pLP;IrAm$`d+L?pP z6&6fxUgicVFLO4RjWa1>9;!H{LhT&WTUR9PPOfGz^-1{GjL|sODGi1C#Z$*s-c_j-DMgR2No;1s>}QF5l>IwnJ%?BMt}q8E=_x zwQm6H_o-qDBwXZ`PA|bOspHV`QX787w0Zhsw*B|q8P=68e*6KKSifvzNYHJnllRhP z-}Gg#R|qLfn(?K%IvHSQn9u%lT0Q4t)agkDNF{HwMLXA8j5|-h`6dg_<+>t}DBc53 zmkX!(4w3>{8VpjS?-PDoYpHRE=L2Y`3J&!shBf*x*k3EWmuWEJ)5D<1;XD(p{9_+e z{4)GBOE#QckwbiD*8N5nFPOj=^~g2Z`GZ}tk=sG&H}9NJrxPktNlK`B=pL4(m-?=HeP zejks;Q;!_lpF=Cl<&tkws&YkeI=lg}P^&dUmW5eVX#UPiLEvhDm_-}6Fy&%YEJuYB z5DX<6{-;&-@7=HgkIw5sO2G&ct*LqsT2Kc{eO|ghb8G?57RWd&zW}wJv-Bh@*DECGYjKy=)&e`2C zuk@~^v76B%w$B%SPJtM_YoJjwOgFjurF;>M0)RA*#6*N8Gp{p#k4$L~7scbY-<6SP zR0U`h-CBtKOr8@1Amh_+BTD4_rNQn^v3G0YixWK85!9oj$C9ikdYzF}$v^22f#3Ecoai2T5uXevM3APT5{wrlP3B2oE zVX)$u-uu!sCoxL5;`*bQQE+bzjy&XplgWiHyYSmMqLT`%2en>CUW@mNUfin`P>w7v-=)&>v29SkEkO@HVH=roSHZQr zot~eocGGH<`HXi#Fta*19tgA>yLq)!BB?-D!X9~LZv4V{*6I0N!HlQ7wZmdT@O1Cj zhP;x=X@lXV&&p%y7nwXDO0Bu-P&cJMR_7=5oszkZd9fAZQvzBf47g#FZ#l!dc@smF zVHp|U@3!Lm#sJ$bd;0aXF$~)h z#0TdmN;AocJyz1T{q1eR!^L4?+mpuSEnDv^hOKsLZwsKA`?Y(PX(0eZBGl3NV$MC6 z)k;J!BZ|Z8AclC$f46JO;Ydp$YX6P#@SFq}<4XU}V#d2HA5!CA-yBo<0#k`8`ReQC zW{Ebp)4eQG9hA3hOGNf8^LtiL{o00aw*7b(8pGLYdzoBZ*;(dcK?ppapoR+IZ@&D< z*(7eIoeG0Kb;1Dj1vPMK`;U29wue+KDPoc@Zw&vHYWh>o@s=+_&)eq!vf`Os=j%*3 zc5{}RatiBXL^+BcPFnN|K0wUm*1Iv4Q5KAH+NNk!WGsWcOq#`gDJ3nZ60+_Kwo5W* z()oxpP$8*z_lv9^$E#23w2TtNeZQ`h$@*1dPDZ8fiVYOenH7bLNPt1q(g!LDZ&P9o zvBZoKGrg*ah?fn4;Qg!L!>77QZ>cv$pmZCR5(-(!t-B@q-m|`Bchh>MET3=uc3JZr zxF_*oSl0_tck?Bq)JwF-{U*ta71)E!*p3hH58M7-FPjy}mYA+3Q5R7RUZzLmrEJjbaG+#{Cp7%Ywa#z&rILPW{ANOO9hTodYLAy# zF3|WwI&i$9cRfEp9B6q>QRSY1c6`KYh$Ybm=#>#0*vV0Av&J2P^zVXQ_}7PU1h6~Qq6i6qyFgjYu|St?bq6WwqNVTQOFoT@A&gDGBWa~ zaoe5UedFr-Mx)e{`xEe{YNY^kI6BF^!tma12b9OmH|I7 z6uMia&>bo_sf;tAlf^}T0VeexhGta;;Oi~Yd7LvXZZppwGH?Mc-DDCmnS4D%@_Zh# zS%gIza$2W+T~axsAbB!bP=@lV@$&O6=;T8LB?<#*_hO`(&+{R*{?uziy@%Mk(J$b>AJGJRH>x(G=>pX-)3rs8qFm^!}40rmG{Af?59Ng_uC!Z=9$xt zrMO!)N@+QrM_ecaB}p#$LiNmHd*J{L%MW;tuOO(febjjf_GWW&C#iKBfzJ1T04B>|M4AhR)TIiKk{BK`?%) ztslj(<-h+d-@C-?60Vf=(gpBY5-{D=OX5N(WOI+4?JtGhxG_Nls5FZOBLB?y&)__PK$9 z5Ck*|*cZC57h3~J7Er^~=28x&dngeaP3Mj<_GJ2ZTH*X!qJ7T)&-EL9zj%5SNW^1X+*(K?8%szw#XC zUeUybGUDxj&e!_m90Y@ul;0S_!^4}5q+P@)3 zSeGr6Rw1RyGr}_>)^+$Re|`QZ^_3~`q}4aE?)ag)zXz>!ov~Xr%WHHL!jsHRc#K`O z-nK3(m>9LYlR8pu#XGVQzm22vpO^b*kUd4AS`D)MD@(f@_m}A?_32&Pneez>ha}*h z&1tCqq( zrUQe^c?Dh9hyVJ+e}=ML+Q46snK21+zd44n6i)S0AT2VXGb(Kpho#vS_5ThWGL*w6 z*VBd`Y;)SEbv23OjT%VuMzm?n!2Qv)oBip(^NW8!>D(d)iY@DZ?f4FNojfQJ_1=VB zWlo`Dk|ZPDKW`PR8wDHg!;%X0z?nBhd~K7*`||{Wd)^TvlU-j_y8rSj z|Gsnsi}KATD0t?XN|ryU>N*CpPN`3|5m*20nBI&#x@<2@xS>GDK0ni<*?dqt!5}3g}w>-vjVj{9BM949J4+a4P6s zQtJecH+pDi;u1XtbF+&?EY(QcMx0qIk;DF(PFM!x*@LXM#XaynG3-D<|9g{k-+s@YVl7Ey2ae)5xtvJ5nm zedXt!oKKH52IS=R!ooa`HbI6v?fkXTL}O#t#sIxF?zPqtAKCzbmg1{QAdOS%M2U%z zVmHeZ2_3{?_{EE-ueGx+>g^(lcFJ!sVqGA}AG-Ty6cE)OiS&VfDtX0S{pS$_2F7#cB!%kRbF9e1^mVR_z zKlR@B9gPq#;4iCk_-#L9BRHI}+qN^7p$64-YJ!tlsL8hOFnS;*9o}E8XY{x^LeL^& z()$Yr9vM*!MZQRx+Ip$sbAy2KAUcg|8UDZAl}bY~*$f_x1_AFU*CxmP&Kk2RY<_pg zwK05r{A6|um9-|QB{*H3DbuYEE2ORMJ z+?asdz8l+bB$^l|i9!2=R@CrvvkSFiZm|9k%bVK=KR6^7%M?ofa37-s$|KyBxXKvV zuxd7YgOeZjiBNVnvonv^NSD)2x)q|kg3I-t)D<>sZvTMC@>C>D_;D}C-G`ONE3+9~ zFsb$DBVE%}RdC8`F=6dul*#@J?Ps_!n7y~|tvX9xzdn~1s=uOOriZMt8UTr{(*G!E zun*e@I5eZ!?dPmJt+#AGW%Fwx-nytZy6Oo6`AH<@kfNeacFyw7d|Zwz>G!ME@N{nX zYcf^UIdc(eQ&IF^PNe=k0V?n`(qWuqstLv+tzYTsJG(4!)zy{(1xD=gdtj-KybFW* zWusx!#h5)nIyDUdX&&(OPhUaj497>8SH2IZ61+s{sR|5qOAkDD4NK0q~w@!LrCu8@j^nGC6 zP4i&Dkau7NlfcTFuN3Tra49cvY=y%?8oitL;48BOFDlYYSu^&05rL0Yt7wDlRLr>R zie6U*hhy4d=OWe1=v>3n%u0 z7h^tpngb({g?tn9fw8TZw^~e2;-CmkjVBwoP>@{h$mNF+#*5zLB9l#7MF+`C6kGJXGx?HUO}pV*itvqJgN8M1-uht67{5sT6;fv(zRg-=?Q%Murb%Wo+Ed+(lK*m$aSJc=?U|ulsBliW^>Hvr*4>j*Y=Pk042tn0V!f=MLkO^yh~?wLMs{%GGXVi>{p862`iQ8I2p zM5nfqi?MtIx{Tq33QpdPf7Z%OJ&h?6mk-Uh*cqQ=`vvwIXd3-}e>e@_h4QVNnmJzE zhm+i`fVjDqPj@25ToxFd^~zk7uB}#I2hZIpZTYx9E>H8?7GO)N(&wOeRHi;Hi(i9t z{G;-Qd-27xcl>!La787C3urc&^bjFEu=(rmED&5jYQJ|T3grwuSxhjPTVOIn-{S}O zvgKYb^z8(hsmw*!W$|x^?vQBzQY+vGIqWVGBd7ZeKlbyiq#|J#H?o(%8_6iw=g=y7 zEny3wRm@gHo}5!Z@90W61f$q*U3|PO=VT&cnjspq@|pD9^VUh3mji61$1~`ruCkC* zW^TZggW$F*yRy^?kgmyZ-94{`R?-^8TY z8DEW!7TYYnEls|f6c5*ESH=iHQbyN{w~yO|rVpE^q`%1h)c)xl4x00`FYFwCe4I$| zn|fe%+y<~h^`+fP3oQwCV_PzX3F)fmhcw_K+quK+^*Fe z2dN5k;E~qXDdl{9VV_aC0@3D#vh}xRb;e>eVnv9_ zW$SKVJ)|2Cf+$~Qe~pc^FEOnyFMGDnJ$@3lxHSlCd~cmTJWU`FAyixCKhP}jRKLyW z9%6*x%LePF% zue?A(^&KN))epN$gPsP5UD-8vXx#I61R)q5fyaqE^UrU;g15c37PC8h6FGtO|2T== zv-ygJwhj)g|C**T6zs!7mMPB@Nw|df7)TKsMsExcwH&ouQ%@Q@VeMed+-kGoYrhkc z9|aHV2?v}9r6{nH#e>His#*-a(nkY9C~669Z&+1Oz*?Q}AMBvL-b^Ljt@Q#(Sm{0a zTz>V#VD;2!$ZtM%p+^lU&<)R2nIdPUL_3@Th&uA*nmPgCJs*=|?HJa3BP`8(Z0KHl z7f7^MXtTc5zySgeA*&DhRDH^;0*jDpezSW7O z9lzU@;|>c3y`$P@M!@57$5e$?ZbeS_8erHeRRV;%bLRk5;MI_BHZ^H8cX(rnB&X`~ zYqWvYU#j)3Plcxfy#4ohR{<92zWReGpveBYMF^c?EjjLq(1@nXeK}2pI{OJI}L;MjUhjv=|NhP{cTY9 z#Ct1fcvZ(R9c2}!8OPwutyODnv9D}~2=0p5F8TF!^|-_|ZnOd^6+E;r??AHK31^3K zUP;s%k>V65iW@y8per`Z<72ncx(@%nlA6VoS1J|EVF?ebO;lYVW0kLi*leg`WZLLY z5Ay*P|E=n3s!SY(L!&V&WjF>U#|zPubRd;VY}Lf-5w1=rb_sM6zc;7GHGH}uHNZ0L zj~s7}$w6b95zzT|xS;gD`EbyvHl@wMvO&L1ctQYZ1x^j;CUE9D;flwuii4!(zn9+p zvGzP!$DXo<%AoasSbN9l$h+kYbS9Z#Vms-HCbn(cPA0bPcVgSNZQHgcwr%z8dCxuP z|M{+UzxC=~zh2#Y*REalRMqoHS9o>T`xuke2{YPO3pz$aRGu<@I6QYF0_Y$2j<3t^ zSMDjbYqltrER$@`#eRqQwS4u8V(d^wWbK3oWFp|wBd}kFKMZ?IJ`dziMc_x+L-V*f zn`vfMpGa5M7WYAi?XPs^f1F z|KX|r@0Oq=(IZ6lO7?)Z%tVCj7hX&#vqj@7hc`G_9)*Rh!N$ti&`wM~ntsnp*nMSE zO{1^49ZN=Z@k_Y}*-{|xr zv7|2~e`i=e>gagj#C#E6>nk(2`b654q_3(gvFER3RT?@q*Md`!(GApE zldsnOtX_KW*NNwAnwm{rJoAVfzjh_S^^`Xrn6>L2ToHM#_tvLUO#h}$13J`Ou0pP5 z5!@JE!B#J|=O799g_^k?kf-dxUnI*yWCJ+z<6O)aEBpdFW^de6Y(;zTv;~!u&)9N} zYZiy+u~h~QhifWXt`%PMB=moa0QVceOItQwYOJYNj%zy+bb+J5(T_g)(S?Vjn7Gy4 z7Ce5Z4?d(B4ewOW7`L~B8eRKls>l__UF6-A;j&2P!0WZkU=L)L=1KvUfi7@NI^B#v z&9+cgf7aF@AK5|K@{J*MmOn3lkC4d!ofXPoc=TxlZO&(4%cTD=Ix?`Vk3)}8BINse zZ@%qe1kz9suVwy)N98Re%Ee*VX3N*a4bO#E_nQ$rai~h)VKPDC+XoyG99B6wryBq^ zJrJ%EFX|PC3AgFjr|(W*ozYcYk_?XnPSc`x`s_a#u$Q99J&_SeK*jI2ToT~+otm>r)dz-KORk_oH}X0Mc5HO3!jX}tveZPLqB(H zXO)|_FxQPOBpIiFl613}T~6dmbH85+nMD~ghBkn*xyLo+6X|smoqjp{) z7y?8yQcvwAO7-%q-}BGm2((4AcCr z*01j0eI@oG(*VB|8`llO`z)ul}r|n`U?DC#O+_8r=R&(1_;ip z((S})AxNb|r*9(nA@7n~Iw@na*fg`)FBi}L+9-~yP?)?F=^g)RFdq*MHin_}%MzC@ zOd1uooE*IKD|1VyM9E-BYx_+0fv-O-Z4hlMfX;DkkYof;IuZ8G!JKzb`0nU{ z>dCHm_sc0;G`Od&Xlu%MP&U}%G3MA$$6U=bp(ahK6th<6|C`a6GEVbWgyqWp%JLF- zaA=CQ?xxXSBbLP6;GA4}OsEzq-x6c;JyKckjvsHk+#s&X;v0UsJIdFd%H$e9 zi67>8zoKQW))~tP?LS^D!+10fm+8q>{ z2Pg>my$6o!;|B|h+#(NO&;NMs=5i&iA({d%nNPI`MoX7DBot}~O>>NT-@IJ6{{Dpa z@|ten&Q1GxzX1({pxK9DF4yJfhMGn-Mrmq}6;intFfKl!hUe$T!l z<{kPsIURT!jvfZe77A5zHUK!HWrFGKuL3cBaQb;d9*H-rFZUd-0VW7$vXw*r8ou*f z$F6+rjm=cF?T1WtGBhBqgWm4wgDjE%L2R+%rN`ONVHJv;j0PatAy9uPO}Vsi>n3<9 zu0ER~ewJwlcH{Y@>)bql=-Aqr)zz=a?ic<#jGXu^G9~@n5oLKjS)PlyyXfZ zEVTyI;(#F8U{y@C{iVTB5i3V1p`+;P)y*E0XYU$nTaCf+AJCOU8lqV=>RMXqlZM%H!9#wzM1Q-IZ{W?g9DHOk)1EXYPL6zn0HUJd%NN!W7F(IE$y>NS zkKXd?q=f081Px~2q+$gjMv|$RN>S6zg-h=aZf+p%ZPwH5-?BBGu1Nk1KZfYVgHTca zC3h-CQl)1ORnoE4zJ5hSz>C?W^4`8s;-?gaj#EUcYkhlfG2X*L(X`9M0p?F@YVy8*sJ6(r`r*rQ?AI-|fD^ zqiS}~N0;ZGMoLuUi~6*Suhy+AB+^*R(wr0_dJQ!B!x8!u#kBc<6#8gOplc*E0s>4V z!2>TLrn5Ab#MJf7AADs7TY&Kb1gu zd%J?k+`z6e^M%w3_NL5VTi8P6v!S^Xs^&OyGjz~1K<5J&p<*l0S=PP3+=!o|Ig{tq z>YI5o*QI^uF{2!YV@mWwbgG4&HGed~pkwW()@br$Icu)HIyR5G(wn^hT}ht3NV_1E zHT4Er{^n{lkY$R&Jn=thWuRIdbRf}5zmGqf@LxC~Ls58F*Bb3>Yl98xeS&&R(g_<^ zL86q=SN6+!Rb@WwolhfH+cT8T9{rkFOYXIF@tj^mYFm2~H~BkbV8EKvnN>$6-XrQ;%nS)j~Rd@2F{5%(JaPxRd!b2{L>ikP?@CGlKJ}*REnY`)}g3 zv(h;ulFHPZ0_U`&^$v55YQ5-o#EEH4yr3ycmQ}`KCe&}sH?R9XvTiu6#31aT?-KP8 z7bc)^EGBEggfaCpA3l{TO#(GMSrZLWmiuqW<;CVjH)T9BQIBOyhGsviU`kT{WMz7j zfP88Rk~%lW!3IIkDCfIp|Fl zGu7Z-3N&nr#v9qy+8S0D`X3zg+y%m=O!KA1S9$$Q@FnQSuO1ylr`QLZFSv6_Zfv__ zd!@Nmo(`FV>C|UnUq7&;GQ3TH+LaLX1AUk(f{kT(gB1&401SsiO@Vl@i|MM}(LgF` zmp$H7V3RuGxu6KAcT0^I$I&J>Yh;DE&RUjsHoj5=kdT8`@W0`>rIOjX%LFszXk{qc*tv6T)E%G@Y6zQN_Ad?zKzfwC8ZT0zZP43<7RSC zE9byMDIySmNrf9n(oS|_ps zY56o>a7!_6W)_RfYTUtv{_5_N8mF@J#{>kvT)UpU_AnR+Y`FcRKk&1vs@wuY^(ZS8 zm4*oE7-peZmVoCJ#y^is9l3gnHWZ#dRkpNblfw?kwVJF(6sx!|l0iys*8J65bAyg4 z7pb2&w~~kNX9Ool&)S-UYfMQ~FU2GhskdlRmMd_B*<9nF1=Ap!H$>0(E#UBDtqyj> zP3%FZkHH7xTij?LrjsTcg4IfR`_hElV$s5r`erN9f%JTp6~~yX`WR2C)!z%ai|R^_ z#&oeTW|YPX)3>}!G%~1a$wx~1eeBW1t`iO!QKr(T?~noYWm&rCp}clVYJ(F9iyi}? z*Z;Oo{VK;1mcH7?$uf;7`>RJAb0V?=pUju%^e~D<8Mpll0%<=SjdheWEe!S{9fC<> z4Y`Oz%2_CG8j-uiIS|t)>qO2#DdUxeF_jbIHj-EI9em+Gz9q6VJSyeY`PmiBDH%~5xtr;Af{_7xub?9oh=RZ8<}aLixG zJf1>|8L`#z3F^|o%1CA^sngubgSY_HMKAM~Z^LXl_JrOndu4dGc0jG$5JF0|RO`M) zQN`npY1}U|{<7n9L=DwcX?QRvLap84w}|UvzjE+3D+yj7wXyn^$IH#z(wnT-EL(4TrkO;i zJCH|8rQYriAgksX#P(09pKGyPfUL z=4SXP$2jA($p=Sa)AUvrJDp3Y{NYVS5WDYlhNq)mUj@Y>1cAEajV$d_KSI($%WQU%s?~XlJNxXP~F&H`>#(1F`efP{V|APG3!ZWD>V~ zPEDkyGBE+&x01`5aAu4?X_!>29O4!~k~4{=#GWIJPNAsHRMX9qBh`8mjI5jeCV3pr zrn~cQRzba;+S=sM!0 z4oa!%|342EgUt8#6(X92*euxOuj;7Ha)ADtMn zo@`WKVLxY9Fr20pCDX|@S_U!hvsg&l@c5einXR0lzdZ^0^a>ACNvgar!B@cLx#3{n zft5pdKy&win~aG}-Fno5?R+c-I+I~O(3FC?nr*iHk}qSn9X494fbFb!sek8=XFZv2 zyITgT<#|EQ0j{iJTKc;h+geklRIZW2(-vyxi@Z*zh|-4ohrso;x`M&BY|&D~{KpQu z>GFiZ)UA#iFAIJ?WMsGX97d5xxQ&xdOg4#w6oe;>ZD&)x(=25ZUgwZyX$Yl#Tuh-j z#A&02l>csN2&!9DD2rD8`|R>fE0GE@?FBB1msJv?>q@r);W%rnW0?a_mhm48Y2*@zu93bN1%_)63Wf)eh2!ohU-+9#CZtj-1V#tO{ub4T}LW(8_wtD{-IJw>IuI)huPJ35Z zL|z1aP<12%myPzh1!M@l-d%q!;Tza*{vRP42moHZvfd)O?5uE1I!HkJ9w-S!qfiv~ zHNh4z{P+zHbvj*8>$OSzd=RsmSTx+z7H99*owu|2TOwlfUfjX{F!+~N+@=3ntc1gb zSm5;FI-TN*S(iHa{o#WG$CvF6%?Jt5fC7$)e@ruUnNmOT(-f;^b#OrQmm{t16XAm< zF~t>AGYm4L7>j;&RD!`x#(m*^hcR5+>&*ZaQdqheOw1+*Uu{0lM!S674Ud+_%c+{w zGt$A`97f8mPO2zDYK*@t_Cpof?t&RpXd{MvRC!6l01uPFUhyzb{bm;wjt=^SspBz` zi(jZmWdyChIE6YU7bFZUvz5xX7{Lz_0YcQ!Mil znt{LEU-DdS`N~}J`UqTfCM!(h-gJX-Yrt=HynwMMS~0cAT;mIqh}FA)W{?(Rkx*Y| z%DS&l=EXq8?lfK~i`nNaQ-R{whV+_PrZ@w8B%F7GZ2gOrCFYCvR8I^e!J@~*>jswv zbJhS6C?ceA|L5B@uHj{dTyGwn=(WjwaxJ5BcV`EL%v0u^qjP-i7_@NznO)#q_(r|w z?vmLT+zE_?pf5nV+=KgGy+F~u@j!@7vl$yO+-Y&p`kV3c9Gj1bTgh_E&S*?rbt}?M z6Tg1M`DrD))SoKuwteX!p`eO6z`Ro`1twQhTR@Lr^ez)m{_3AZ>Z<>oP=)r<)kkdEuK-tSvt6h7 z?~fOt=AR6wVuOxsAd-{LdPmve{*>RRxQiug>|K#K8?VHbD-aP_Bz8ftmPlS5zHU-Z zU3|5-9VTCM)#^ic%&R3#d!s(%N^oZ5J*fmK4W4-hvVgD6n0{=Gi0lCp5g)*p9xp~C z;(1MkC56D;J6!R=4?R;S{iQ#oU4fM+t=DkX)R{%*n~sJX{G3r7zc*(#!IZf-SrU?T zOdJ<({ravp5kwo$hT_VTFcQLoE#GDhha|%9P@j$H_8=#Q$Iq?JuX(uqqb`Q)-m zohhIf$LA;B^gM^~8Se!PNTepy(3n%kUNX0ae-|gKJPUi(Z11btVFDN>XP6M0&Y z|L-1xZlEu@p6`zA8G zbdocg6l~jIg7X@(S+jo|Pqr$9#_*h}9mSS@#5Z!lo?ok*Rt;YxFG^su;Hw2^42W=E zUky1+pJ>MCom`M$xT2GYD0h7aaLMDk#NZb~VIzYZ1j{|L}sulPU$<*~)2jzclXB0qR_xisjDFe^_vm72`H z-X4s~Y`%+7Xcf{{dPW=0W|en*vIbLhL#n2ZqS(;p91hMGqktVjO-&sSURxvf4@kG- zYfv@mUr#Kab9x2ACQsS-c{7C%J>}@`pbaegQmd15@nn9+M_o^WSx)r8o@uzh%rB#A zfH5Z9;(JAIh)p@Xj2TY>wuWC@^N+XV;)R2nt2}-WKX*-VmjkOv_$UFISylw&_wm*K zfs)2?>&j>BalXsvha;;}+O6lKp*rO$YsIYYC^%8Pe2&(Of-YiMNu>W+!oD#${Mob1 ztLr1AYix@&o&D)5tKBlOCoLl zk(2Ruiv@DpldZhyF<16dcxUJ&ySmz7)~2+EQBQ$yW9+&MwiWMGhhHrd&BdsEJxae` z6@u?9XRF)p5a~~{oH75rVgXCWkayR38?xc*3*$LVSOV<=tuZERF?7pFmp6vT{#?h9 z;g*J=WicZaEz7N>jo`mt|6dQ}w<4P0jM`frJW)lm5xhVq6rfpw(G7ct#jqvi`^i*8 zB=eF*z1opcTS*m1V;{hV8>E8bplUJRlBlJ=GEXt4TKiatM-j)bsjF;4E1WxkEFOIu zfJT)jvt}V;BBfAluyP`o55TQkz#AQtfyG*>Avr-2uA0##9Tg2^-P5RdM&6JjYs{I| ziY+~4SFOSl8&^CDvAu*H6W%YI&32axQe-GS0IUDrq*1bNPaW<`aB%!7L!HeRZ?06t z=(yv+dVafoon;v}fBY~*)FJ+K)=Q8ErY0G&hyh1h^U-s)nob$Z<@7zv8@uNn zgY7$*!WnOg!pqI!iOkeFJj(rgiE1+)Ff&{|<+ZVvP{cuqP&WMPY}somJMjAC-b$3F zUBFBpp}nIar7QMf-edPLA=pX6szz~Okk(x_OvQFMq)IKp)G#{?ukdW<;%d<F|_fl&Zc=UxPcjS~V`PVsoGa zAy5MCJ!Y|lW;UEJ)8w(lcpaL8JQd$vQ`&Ss$o$@7UpT8lqNY~yI7`H#4x?PQJyvR^ z&4GOiWc#N(CO;>6#0p5AfiTOw7O`ZMlEdz}8Z5a~&f|xyZky(#TBW4YP7hl)Hbn^URgjDtk;VKWQ35%5WPg z4~5OPv-sX#Z$=>|JS!$io?vAdBw+~H&02FZ7lyj1rJnn?jl6Bet!lpL61l7*#DClx zDEQeXBb+~-0Z83mbpwMSwVTio5fNd~$aDCZqRY#v$Ye91c?nppu;6i+p?NXhcfNJ) zOtmUyv^r&aMrZHcZD#XF`mry$uop;59cH{xsCN9oz5rf-Z|P>!EtcJNKVU`7>e{{- z{?ZWEva<4%$bF64Xk04=MWoIk^piFcXY1{V-Qylk?UH-{h09l2J@$u!4_5!pvHJCPzTuGkBCE~ODoZHJ{2VG?8W zt!)M?EDkSosjf}xn?cw2PdCrZ-GpMyW&-Z{g9~E41(@9-IsUx;laV+YIuS`UQYnL6 zl=OVLY~J(Lda~yhU*IPk7X6QB_K?52*r23pdv7$JObJx~+BOuB&igr|7dXLL6?n02 ziWG9Ro;AVyz#zT;T>v9k5Z`3B((Qh2*NcNII189{7;m4AT8wmI>YN`y{}C&^*BM_c zwB)#~?iN#Av=*bVFi!G1yp;;-m$3(R!*B21?16av4aJINGsj94^8J95pl3Ha$lAibJHRYj{@`KE+Fyj+knG{@0L0 zP=ZXv3%m;@f$gPC_|9bjmibx&PTvwoBz?p`c{dtT!D0Cv=In0qWQDuO=+dRBoN<;t z!H!C18~O~(LdrI&_~^q{^WJP7M;R$oV%rB43+7EJZ;&5(pm{)(Dm=kdJJo@qlGgSd z?1v(~rA`Qcd63e#U=t#B9;@;%FId29yA@BBHvcd5o2eUMK zRZfaxGj{k@AQd^zl=kR0BjM7_T*(z}jz;H+_R_F{YLRuHpEWpz0W%-;eVXKlNw{F3 zb}A7~V-dBf5rpGrg5S)wg(=_^er4FwTlIhfxAxZ&QdnRF6d2u&T;?G)y!*^p_i>E} z{sste!)A7?wh@c0Eq*Y@$)4+PzOnt-k6pR~4F3O~w}sL;V*dWYtr|M2xE?r!@4;j= zK~glCGI*{rpUP>@?C;rBtdN)|NR80PN$Q?)I#Z_QD=M4P2O6hLCDf@D4A6D zq})!AaFMI@<+Qat(_-4nNQF$+SN;%?$F#A;eJ9CL8^L1Q8#ngsd=ztF} zXDj~xKb}W7_xH>)rgV06knepKTbq0D8uMqFJibRyE<$gaX5byI_EQb(+|zmosW(x8 z-`l2n4+k5bAGHV?+x z^;2R3MxP=fn8Q8EH3+Gk@=v zTPKz0^vcSlvHEIts2}BGv=~tOS|zC9r>4-jN9J3OR%ZkqLdq&UeB6M#0Ctyi=~D4Q zZHSZc1R^Cf$b}A{Lu-^_^+V5u3n<+njka+VrCO{9qh-7x{toZkK?48#^x0*PL+(;dw~6 zSC*?G%nsJl>cx20ryATRt{hPp*fstx`s2Cq`PK{&ueF?`+O9yo;*wF0sJB=A+BQ?J z0(JY6$=1xxv(oWw>(R=gls^xLz#Z1(TfE);^ZM2o?szG&N~gcrhUds}$O!rK?KyTC z1zK?Sbq9d-Q?1m-*7RZ(vDcwwf|biTsQTCR<$*L{G5a!;bEftTh{df>fcs=^XtV54 zz}TRD-#v9(8-D$|aaPWVU@wFZgxqZe{)iR~lh5*#bRN@>6)hDi7Y}#LR9p#s&p72A zj`B>$L^_Ztohi{|KhwEWkMFhzzr`KM{_r}19@fRxJs{gBRxmFUv6y{DZ}kTceHa$_ z=JOiT{qx=J`7Y2f!yWK*ZA?iyPd1PwNp8*hc2AY|TBR>d9oq1|MTd zA3jK`5XWs!TwE^a&uL9EMY=fsk$EtT%GDCFGauj<1o$icB1|8Cuccg2aI91$G<(Bn3*c#n2CICUaE21N*Aa8=}X%1?E8sF%U9=i6M}jg_4@sY4hK5_#vzVR&Mka@ zq-Ij@pKif%`N50uWElezT%vj^X>4#yT`Bn@hdgZrf2fB8oMbpKY+!_!y78pgryGj1 z6bNG@4YLC*2Grjov~BuRv0cM|kYR_yB$N0i)aCqeUY#>yvNWnVJb8iThQEm4i!J9T z$aBEf@B>48rm~_Syk07VHk5?QpOEYVgS6=8*jFWwP^DPNb@Zz?>!7nA%XUb7TR$r$8=#S}XgxvUQ4{-IouL#R2@7S5;DNxj^fCjH1Z4qH&J zQkWYUkNY1Uf;3cF{V}_{!%wcp;1WM zuj+O?wt&gP+WLpSLMBVRlH!}+*4N1CuS1WiL!owE36dcFIJTlvT5%}b1aIyXmCl4hf5SXwY505?EK zA%;-QC6CA-GFLo4-b!LZ=1yh3^G)2f5vgmNqrKJ0T{z{`)i6+B;Dv8DD;0`XDkfvU zDF~xuOKiLI4a*}&)o4QYk$kCmq_$W)6Q%%~LUIz8+a+6gOutk5BgFesH8u|nHJ@y)f#;tSYN zQTIdbERH2P{ZnU;q>4G7INAbwvLl9)$>cV%p@(4b;)-WJfK+MG)z>dZF=~rk1X~@W zljMb_JKQ}bdY#*?zjI6@oIx^6p>;mwWX6<2PZKzs9Ukbfyl$j4P~}*k~kV6rofkWx1o76!Y@p#a@6yzao%I41Bpi_y0T-DABv=TL)iGj8B}% z`PCgi&C84Bb~(g3*t`H*YYI)M+_1|p)&E~u+qL<{ zsjuBQwQ3(kEp#h>)Azn;2&>zq0(CS&74-vZ!QUj1bJ<(qm)xH+O-tXj57u~u80cut z84F=O>Arb=Z;XhbZSVnHhtH%2Mz}5AbmXYp z>f#vNB<^~xM^awjI_iy#x$J9$uA)5|6b%2L19(0_y2JX-f32iILM8iW^lY`(0)+=x zjYLmVtAa=!H=gQ3JhL9onJd!pWy*TuL?IQfZ%pQT7zqtvOLz4CP)t6eUJbB@0YwU8 z0X&16Y`G!5IPI?|Dzx3N`c&JX%|OZ@*+*a}VZ`Gmw-ZlBNGCTlU^Ld8E259dV77I? zJA67O`=G_!o51z&lQ$uZ{>^8m;rAQ>+mA~}6G*9=BN%8ahmL<2vk>9;#F4>hL7}--Pb2Tz1xSGDT4eu5`04=Gpdrou9%7pSDb z|E&>{Fs2|q2*s)W0uk!SF>f-%ta`~inri5|(qix(_A_&IBDFrX{g`jbYGb;#o@}Mv z27$PJA%~xl>cbVnk1sGfgUgd=!b36@{%Gx`)@V9_T7uT0@Viz=g8A0=-B`rgh~Br< zS5&#qs9MiG%1^C^5a5!P{!{K9_bZZ;JM8`gz8u2kde0aTqtVr zkpwp$o>l)^@H$B42$%%g2J^M828#E7n6T&Xa5Q0!US4k)LZ+Luq1+nA7D%VVV#^Tx6b~R!D&xb|}K@WlIC`;qrMEsIsyL zurD>zS+*6_`_U&NGTfiT?1pIHoz47ky`&1@7vLRzCvN0 zvh$hv*3MoEvomG<5L~Y9@?lFtb%fTm`FYEZ7l!|3`u4;vj=V#$fhhdn{dD1_gqzAuEp7BJO9|ooWSza(Wt+iv| z5PES)YOU_$^5OFa)=LD5}T`AeTHT^(#8qm=Qa zJ)8Vq?n|)#4)o^2i0&NcHBgXVaaTN3MgpC3xoliR=3(jNYxjs(mS_+L>iycRs zoyj}5aYW8#Ge@6Hx5%C9u|IXxef+q?*W2w;h0-lRMVR(Ue8-yrXSI6zmEkFL$TtU>Rj@` zzg4ly;>>BiAFK>QRjb^WZT*u~@4EBA_(h_%7YL#IhRK*?j{0idF@8k8ij9+Vb8ws1 z%Ure;(0k|VIQxgs;OptQrH#k?aEWT89d2zRS8t!KaLM=Q``BnexnoD?%qu^RHcfAqMf_OWjL%_nksHwae~UjM~O4cQ?{=M_t|I z*3Sig23W7*`nRW6+k{IVx4u6yov>Sum8OihMlp`_+i!#?bF(t%t!Syzpi1dZ<~gtJ zt9SJuRilwic_2scRdn2aYn>rrZGoku!`nakkeSJlcVquBo_p_ z`}Fquloc1DVku8P@u!`CI(gZ(#qE>(sJ{?a+x?bs*Ior}=9#ze%y0HeuO-w~IyH0H zcxxnvh zEcm<#Hg-=7VPAdr!IA;O(}%vGtIYX-=ElEHKUTvg2V$6T zE6ql_Wj44%<0FoYr^r(@G5_gHLZ{5_DO6a@js`$=t-*}^Qo6=*ROg6<(qLd8>#GuY z!>K8rb2rMNpAO65pgUcWeZ(vv`x_M}3q)tvo+Kh$*f7i#4`BUO>ZH65-?*JFmb!(ROZwitNQUNa0ef+_MK4xVz2=+ z>eYcVTxn0k?W4y}k%U$-w*qon_h4|!y19^zTH?nY&*UsDq;2$b>qvHggp~Dx9*EE$ z1CnHWm-q2>v;DW)%)ICW5niP`Tp&lpJAp?0M-e47`?&g2K9O zk%h!3576u zbYg5aM?urLn;o--t}&#phvJ(GI9XRIy?*sc4^wMTWGU_2QEQ$n+neq&B6NMXgmMYQr?C04 zke>G=Qp#F#<&+QAEzsN3)?PmwittU*3P!b9Q?yn=0;$tXH_K%2Z+o9v3`XUV6Ek!^ z!^~T2R2SR>ldc#IrDps6fvrwv>zUYpxZU4(p4Jvfo-=2zZ|@;7u$v5ZG#fsh z5|%G~NdN1iY=chwnW_kMrz1AuOaC~su)W{=1Kmm&lN% zNsN{XDE6h}WjnFx&AxcIHROnbh`WGA3Q>?(_=eb>J6T5S;>Nq@>`RtXPhI|43)PtT z>Fi4MiH!l6oNX#3rg0${c`P05 z4e3fAk^DA#2_`~8j84bP%p;Xw;*il3IZBww&m9FyO&_}NeH$XKx{|n&qpA+XscU7c z-5tI~mm^aRT}1=-Gk%MhLDvl!+fz`jD*)YB6qhP#^&yMfTxt#BJ}~sQGh0QU558)N z!5OHCBPrrD{xM*#$W-Z5r2og?@Jz9vzREQ+X@VP5>apARX6H64_1;Q8nD%b!7t-LD zia90M0eK7I{j|O!G#Q+I5@lbN4MP>ZDc4qN#?H1&*{(5e*z_YIsykS0sA@r_6C&85>Zl8UC)$9jt~Se)ty z_HALm&4vygL8EFln>-I$KKqP=buLAFwv)|NU64e*hP>iq=h=o32y>+E^h*>WW`&y? z`OE&^F3vw(odu@Y%tZOmMHB`COOc8z=OA;cz$Nhdqo_d=nb#|YmZVf49~ z-hr5)c@0z$;4`P;=R`UWTlKEGQFwbY}>p-mo0VegoxhV%QZ0E^*%3A zMm6JxsK)_EPjBpUG*b{_;Q)Z zPUK)bwBPilgb!9oR#8k%a`jXW!WT`|J?*Z0&Afjt?-Aa#sIkBD&8aENcZ{sxrSfGU zKfWAmwvI!{m5*3Dtl_6zI8Z^0ajHEinhxL~2bkt!=W!c2i;>}A5ZDGspvf9vZ=EyW zJja$MV2jwq$Jx=YB8r5XFa%oXV_Adbl{rih{4?<;?tzkZEN4>j-F@q~7JAu6n&}=5 zxZP=nPhRk2-HDkB91HxDcw@$s?c1#qDll zteyze z3%b1zMshj0Irl&gdY#I(M6fw(uol+)$n3?=9XS)%)#S8EO0-s*#mmP&7N5yud&wPH zjW=*#CdfOqse#6NREn6l@KlwsvTLpMGJE&FzPeE{sReIwoD4af>M!`P;l(~&QRS^+ z|Ik^L91V>PQ|>mPly(-_=+)Nc_w_>+iP0DZ`WqiOJe>=j{-6A7;@j#G1H%7|#K_hZ|>tv~eQd(1{*~o?n(8PLXZ_?L^ zNK2fPgp*w4<~*NrE)~f{EM|8*^&jn1s3}ahtOZz|{0!k)Q39RrwS9?SzaFlp_k>`v z);lPT$CU>PWx;V4$^{DN3hk}gqTrLauo~nbB*vdit4Q%eIk*v3lipQ*``oJbJt+8bO{(}9LHSn0R3$qXfw_%8GGeegnDT-xGVQe&T z4IbxotQvzhij-Wf9(8Y7w}f>pt_kgoHTF^j6-Qfdyx0>e z<YbGByQa~f_y``e`=P1e1q`V2I5 zOM}=$YNP@B+kjU4?_pn8Cw{P03eCoRs{VLRoas2&E)byZx|I;5PSQ^=1+xyC*6;tV zl+nH&dNk;Wx7C#p3JT%4LKl8&(R7&tjn~(plBOrWbrakM?Ax28bOEfE5;zAH1FR|0 z--US{FYk@x?KBzx5IJca_O8|!b2y)pc#5)1R@^{;^8@rwr-J_nJ_%7xaG@AUU?9e3 zDk8H6v3lcFWtHjfG=!s( z3hr?y$-o<*7N7s|1~C!Ua#JK~wa0E`_!%kvb7&({vy>gQGyL)q>}ISVt94K*8re-0 z=e*CJKqX}}$_I6O_<;LyQyqlvUZz4CK47TG!#UB5TO_x2bx$}YK2Wb(P%YdDRx5gn z`VO~?SukE<-auz%bx};(^TZKtqqDab>FL&u2-p2g9xiu?L#$3;OMp$RqH4$mW5twM zbdwWkrNI|^_yQj&b~iA&lu&vXyc;*$!lra{vPf*B*Iq`|&Bc9t!*rTlIeN^Qz+q?^ zGlZf;q>l|LdnTjNKgqE6Ws|Rc_@TVba@}KOTyDskkJg;((R8Ze&?wvp;w1QhxTt1S zpgi>sOP&Hmm6bXKgqOQq-#%T$UH5#9`r`i-&ZxoR5os=I0-n5N5}F=aq;$lvz>+`u zCiQTF9&M|K*fo_o%=?VZgODy=1((T`gu_>#vrkdVXh49v;b3U0Ji7QT8T)u|$bhRbdnc*RnuPee;P4?ZbUrU~IP9%=htzTi}yf zO0;VU>m$IMnKilfCrfeyymD08l&E}JuGsX0(MS$ikM5hMes(Q=75;Fo!3Uh?Q6tj| ze~H)|@-#5Xp7LjBKeP{RTdU_ul6U`;|Hs>GtE6c7aDO*tol#Z9pEiGduDq0{52?GF z7G8!h=?niSSS6;Q77>g2f*k)Ubl7ix*^^hBYA}Csx@)mtSKMJTf12yP-(+0?>=UC( z!K;kE-zvvViF}TwIsPdQE}k#^1J+SM?P5r{_79bIn{k`;7fA;d=bv>Cbd^NliQ`92 zW^A>u8`!d2G4!U9$-zVrq{MePKgir$8TaE-cAoV(h_Nk~(I6-s4|+3@L)Ll0Pw)$Z zZL|44h!9o4%Fk)pk75)f01hL2L9>FffMUeV?Z{Ei=s4;`oe>I?RA25}lED*JUs5Iz zn|ip8QKgk&8yD>}o@Z!&Ly295HdCEO~!{h{J4F{!#Qb?J0ID@rQZ@e|uWP!-= z*j?O5xfP@8=`G7;`a&VnI?{-W_X_d2P89s2O2dnRHls+vm%-^(zf?<1rWzYU#Q`QV zyD^EFF6PPMs$UNYKlj5bxR&nnZ;=fLrbYHfEj2i*APVQn==to`*6!2_X{GgeBZpG+ zSH73ZD)I`PE5E9rTW`{y&VpWl&sQ)cyHH zfZ*;L+zIXwg1fr~_r~3VySux)TLZz}T^eiL-DNuOJN2LWHuJUnR@berbMD!9@3nsG z(U!?_PhMaEge8j>341SrKF<9vwi9UdR);CUX#R9macnyZG^1p-0+*L_yB#u0WlX; zDl^v^mY)TH14C;W>NSE}#Ig%yl*dk6C7eL7Cp^KVpp5vgFKN%H_y@Mzb&~GGl3E5R zA4sm>1JWjpMvYCLHY4LyQtbTl+|MVY=M*zI%ig@*-+J*c)NPG^eBgCGcI^yQaBg}d zI{$^)c;uVo|5yS(#4rf7EU=JsfiJC>%6g1AlE8X3~4y zg4k&=W|2F+ZnBw}r^)?&HV=xBFX%lCp!q#@iJVQYzo+ZfGG__0KPJBq*V>KTB?R3me1x(qg5@3TcYKzxo3G@wS0V{~MJuMBy0^LYr9To+FxC1Xgj$#)fa z^LO~e`-Sc`=IkO;`lsZK(2K@07lgj}=NOLlnA7Cl2K>D`zMH4oKzZr(`R2!GK4`_d zA&w0^eMSJ(AaRD=-8YtEpW4{{regb%T_sKPZo@Ai$Be~n+s@>3OFqq+>{WER3=KEB zkK@0i&Y)pT2kwso1lOa{LG_+b?R~YQV;>xhB{q}>qUeN;5d3Nal{_?Y46$H#G$AjG z^RbFp6dvZQt(=_PcqTXNBbOTZAyRrM#$ojY3pwYA^6Z_~c$8~-nTpK$*?N~o%YVmR z3GU=?=r^~4ri3Aao`d9j0j*<@`3aR^a)9&dq% zVYe0*&Kq35eh%5{0ogFqmW8hkDcK{r}X1 zy+0IeFO)Z{;JFT(+sny?L*Fd3SBmtXOKrMs=LlYQvE~irg3^RdJQC(Ob8=* zl2U3Dlcn7aTmLnA(WBT5*ch5&#dh*OL)SZ|bv0K@{a#jc`WqTD_5IkJS0Wo;G#^dkehi~;xJ6+y$wyLC;xVn}D%0q=Psz$^e2}@CRhO0s`1=O%AKYJQ z3%g-|g+&QlvY0aclq|8FuTadK-=!5;F$EdHxw(Yj zMWi(BZD|5JFXwdo(!+g;4bmeNCl((aPPSZJi1_@>rrpl!kT1Fe#~(1-UERIVC?$o9 z37-6gQz6MXFfe8QOTgXfS6}W!8$03QTBg~^^97a={tAsnUva333f0CcwrEzdGvVy| zPd6?|1^uo_<^`RT(0V3gr=WG%cKtJ$>hq0p&4K~k4C|y)^1IXDOQ3$~6NJx3aAQtK3xF0A7z zvv=s4^Q*t;7~kxOR<5~9J;*&iLpE8EO?u@ngGw|WK-O;3q|CMAa9E!tB^l9dgu8DV z%0Wr}T$FgJaaa%p-g<7bUIF(U;R_j!)Z{2>P*E6d?5LE3GAS9kCCuIW(p@uWl1AdT zK1Bv|W^1qb3~;W$Aa<;))TjoMKetu#xznU`(%|y6ykY(SF&3i!(|DkK>%xT?yJ#Q5 z>=hJM2YQH@5P#|ki`9$7sMtd5en(jP@7KL~bDhQVa0{4 zHnwWc2rDskTBDw2TW!}-gzIULffIVpXaW-RfTnNWHz*@If*}GUH3nAJZwm~j9qhj+ zox$U;$Nu)y{jOg?{ds~Zzp%3<)ZVoLY%j;g`NB9R&yBO)8BMB z>$_3AqkhweVv11Wz<=%jnx0~Pss$lC-_8b9krlK|<+@;Cd-!5?NB^$e1_hZQ5VOI% z_T|nEQWH`y0!YCQ7ln#T)ZYgErnp#a3BVt`{ovt9gsQonWx!XmRg69%OE1*g93Q#P zWs_1J+J@snN$IjJ_Dq2ff~~qXzamvB0?(9u`;`Sgt$BYuWhT6&6x>Aaqjn!m z!UkGJP_{7mxuGw{eUHYR%NAQ$n!4+#h4)Q2J>YF0UEr65v42MJQbq@u_IF8gX|Ypj*RNZP^$r8H#ywueGk=>5*aBv<6|{H1SLhKNQ>FpiAsu!)v;zqvqX3r z2=1-lMo8ZyV!ZEQ{^!Wb;L7IX1s&qq_}i1H7tiNi?YuyL7DsVVTKq_<>XSZMlioGg zBl`Uuc*ghb;hRbS#9V5eO@vhVRSp1Pa-3Rb-MWSDt{t})-=@S1jchG6^9@d=yi3c( zW-7c0vy~6X`9a`j=<;rnG#j?TN;%6GqEh|E9e6+tEvGRiN<60Jl9^a!&hCYwD@C;w z!Lzw^A(GmCmnB(ii_j@>^|HKIpH1=cVm${7rD?yy%M$l0w?a9T+@s=90hvt%O;C z^?XFBT^wz)8hj(72ijXUsg_ym`C_g)FQmT|6`G6Wq|ufwclE$rJVnj1@rVP7@u~jd zM>bc1drL^%+I2-NCJPE*>KM)evS=EtKO+@&IGT>UQ+-`%AR8y-js8FqWJXLg4RqEL zUyZ^MS~FC7f{Z@o(2pm&EESkc$ji9{_z6F!u6tsf)JRuN9ED~09UJYlEOKMo%1z~c z{DI2!F`>*E-Az1V4hpNfHYoX^qRD6SRXT}~s&CZq@x=tY`+MP~}cW0m} z=?3uW&|518lwp(kW`C!4bTI&x%X!6^JuH*sd_eeU|0^cl^ka9=Xex1-tzGKcU$95g zea#n)82I`%y`K<43@hdWxESCJKpuHvwC^gT3fQ2R^WgF0%E4LqL`}F}2QV8>BH}K53*SptoeQTP+ z-M>E_8l;*zxYsE5>6M%kFYe9bqoJU~sMCP_Xm%wje~FCQvr`{iR5!(dA!1FmOrh%7 z_sTjE$hS9s9iAoAc^3K{8%Y;b6@3Tx#W5)EQ!=~LTqJ~AC z20=ge1sG!Pq@v8aY*NU4!(s#8;fM|i;Yaox8acUN;>eyvD=!;CE(}bTe{f;Va}-=m zXKTLLemL$fNH{EO*o@v^Kz#=+_~%*Cqe5)Nm+$99#ZKl@7g4TZqoXPNRWJoSh=u`i z8rJ^RGSTm#U^OxhEpo`;=5aJ~{myU%+={O*NgAB8qS4K8F{_3$v_yH2L1r#+lny?1 zw{*w9=I>dS`=V(kM~_6h^@&$3%5bFF!s(o22?)8RW2m9xJJTJ3oNpI<;=A%-$3NR3 zDiKmF7}Vh#;g4yv(XekhLph92^+93Qv;3#LV!!pl2X$Ru7zWF4tIrdhpL$X%4CN(=Yy)!&_oFFC~4 zhV!KpX!&Kq*lR;9hJE}A2Mi68vVP&tc2*nlShB}TO(LeJNrMeAh_*3PCfnPW4o@ZE zhO|a-nIb5`D5DIcDWj>tajCk|Ru6y3I*gf-^h3J>hySKA>_1{wXRUxKzbgct-Jqh0 z*oT8luA4>RAqrIEC(c$M`Fx@N2_ZkeO|-pvmM$HAnF}AfR#W@IzHeUjG0dPL$IB>_ z8|8&J26}?kIg9W6V2hFT#@PkOSjtre&6a=*saykd zS4P(*J3Yw*1hv5LH;^Rp=7B}OW%-|#6|n!>awsRTyQjSZuQd>9&f)Bbhio5+s$BSj zt57;2{>K*?5XsDeGlJ&47;$*NWbY9A4B4E(L0V8T(Xo0>ZDH#UWh0eepdoz&gVaf! z5f1CV_8j$U;*Wz^ag<-LO6+!02G4(0Tadv%@)IdmFx;jru_}oPiCNi>7dTmwv*Y?T5dCc|5Eh7L85cZck51G5GZ0|alnVk+Xh)CmwhUz!fd)l zP41N5=dQvXXwQcZ(ht_pOwvrK`vo(|u-f|$rsDQ7$F?%PEddaLA3c5C7_x^Qz<m4boe!0WGU^5iO{d+^3T_)#e2gO=fzH!~3wJ#MriNmeH5oFYqyd^2dS zs}r{xbcNsC;I&E({cs|@=|}ct)6kL~ zip4o?PP<&TYzCO%hyy(&TyqU#=}8 z@z!GmykS@zWIST4GykJXRkLpl}Hr!FTSs_H0A)nT?Py%i&YI=RUo&_5UkGC zu7XLe4wk({fhjYNY3`IO$HchSL6zu`#IY6O$O;^mK_&Z>ZpHDA%N`N>=1GNYBqQv) zl{YS`7-Hk0Ynt22xIQOql*iZ!qY}q?|G_d6O~rUz{{7sO)AZ)h8|Y`VY0BIv zBPxfjB1IVMtT;Hr!FsnbpXeOE$voYPw3%*3_eku~SH=eO9S4ABXQg}GYv`-sPT)1}Gse&yI!kT!Se33AMt_SATGk_6NCB-7>jcfQCCCP5b8^cuvp zMw2yWmu<0oLV+IU2L^hgl}P_tM9rneO=v;as_Lv}EiWy7Z29`aw4-IwxN6u4`QZ7b zhD^N~#(j?dx37iee}4W>Cec672yQAi8B0bhW9Lut&;(Z8w1|QyCTDCY#>_5*J)e=^ zpkHa8FuwWl(0Wx9ZvW}F_7DP}ep8Say_Ni3uY~t>u<`>L0 zYQgqh)(bHDHGq1qB*Dp|^371TIJ~J;^+a14F}x&#=E1WY*B)N6v%cRj{}?NTAK(a{ zR+K64;-lHq8B2|dBl}mn`4tV%%dfE1vw8>1{~n!}j67Y4;dGG9GeVO-XEh3MpgwkX1IshSX_pbw!j3SV<>`o{DEKf@V{B4Dh^OeapZ;Zf zTPk{GT_$0e6T7P!u2`~(5UeU~$}ZQ* z>G7h620%a`Wqw`N&EQ)4$Eqw)$d~=+L_r^s4yWep75Fx<;Sul6<&~6CMH06mN~~By zZVz=#gRPJ`x@~qVm+nDjWpvtn87&!sS{M!7o_}R};o%D6`+kZHan^I0)iaYYLS=j) zr#KB(w_%dITK_aLQkCHRczm4v3#u6m35k(mu&bIip-8CvkG7@cXGOCLTX`o z76wS@#vD0ipI=*p(r4xCaU&1I1Dl+%JA_7uQDl#*s?@ykG6pYmJTKj68eBmwxR(bDl*~F%dRmxIfC$#3~W;^D4+1v8# zh{P^eY!Iu5LW3KJ5=9-E!8JWb)mjQ}U%8Q%?iY8r1Aw_Nc?`O`J~sI1IA4S;X+&`l zzN}`kKdON34ByAVrv&)T2gR?YrNMFymye&HViuQA&1Nr?g6Ol4k@4z(z?z%{>YN)L z&*PGh`I@PmczM6Ag937kdh$Fbo2qA;C4k1h5>+G zaxR0(2}-HIp8hfyw2QH4&w#4D4-F}hNFHw<{ih}pPFGs}2O{2$-nyiL-w2_H?Ms&v z0DGr*2mXgUU|3{1yb59#-cgtH2x&Z;21I;^CL9?$c%6A4uI4oFsot2(Jqyd34o7=Q zd5&d_T<4+Dz@<|*AWv%+xa(P&m3mgS#5szMvN|XZg1KyRD27i;$L33^t&HZE2tia7 zD2lI z0ZKZmE4XR9zds>^DvV_4#vvNjWx{@lllre@47ixI5E4Uv>-@dRm~~t-qNgfMPJSl+ zFVj~kRm_Uk*}U+s3{HYo_rN8wC6(eT;SWHO1ih-QP5qqpZ~SXqTWh4&@oLb6%syQ) zwS;yjJ3GZQhh{#DN?iKyWC;Ct9^en8~7DsPajm2SsOLlW5tg{_nrpRN(>Q-T=&fHUgNtIwOO z(^aSQ{hFJn9hG<_o|yuue6EI$fqHuKwGt0lm>hLd1BaFL*IJq~d= zK_nkM*VI6GK6u^+>2j=RysIi_EE0lv4{M-pN)2`D*ln-caY&VXww9B<11}c$3ab`E zCI`B-*zt{&YuO2DsSrOnC+PT^i3d3AyOgqnkFzoo`iyd&E4S)QjcgAz!u56 zK)6IJZJ)ViS*mv4Jp0Zh1UY<#7GD>%mlEAadQr)7!Er=5DK$He*O;r zS$(=}wD_w3wjMx86{tZCIqYUaFg5E|LWpznR96 z>G17ako%8FDmBtW*^md4@VijZDYC@PEMN3-%jrd0qvs*Kcm<7M`Z?lORIW(a#(!t< z@yW?oy$uCeOR1H3jb3tAs|xPMzC_uvtziXB#WHp|=4E&--;MOE^5?VwHo*Ea z341TI@-TnDLKYIRCx|YmH{z|mLr$ZdM-FW<;5#*I+fL?^7hNQr107vuVX>~?r>`QJ zowh?CXeCT|L+fg>%}UM^rJD@iQc8_~8CUt8;%k((;!^WSnu=r_D#BVxRq`6~%t)>ZA`)l2Z^hM5y~`FMFG z6zbVQqY&x~KB;T21b;#EV%-3aAA>!AN-nugyC`tvC}N3b61PJrwj|d91UHK> z<6p}t5v7J$s@+m8R4zGk`;3OzI$YNvw%Bq_j=`Y)C)CkWd?b8uab8Qe#V0-j4e-`y zkdW}}sZh8=gO{A$`J~G8eZ|ov@sw#-QaM&6->_5_`eKkcOR&QWV+6CubA*;oQW`MW2l za(%t|H0AENVq5D*CasfM50!6gNbE>Ga;gJ0QkT6C)gks^vf=(E#TNRjz{bFG-dG*= zy3UV9j2i8z>84bLM&R9Jh@Uj$HDf+@?uTx3)bYSLc*^~4AJO*LoVZeo3psd-Z7=%J zyYkxbA~*u5Dh`I{#j7q-EgV`Y`g3lf@rQ+Vvf7<2wH5BGzlLz^T6L?M1{A~R6Qy+& zLC4QR^dfW(6h=M%uSvAb&5KvH&C*iOe9>&joGA-8yowV2KSMu3%>G+xBW`kO?8xpr zC-UTR3G`+zN~uRm#>Zysz#Dc(T^u|w2W+N!=ld*{t`SF^N+`r#d zX+a-Faa6SfYK8-<1@-;Jlk(&Mp>p-f-;JTAJ2Tq>>)!~iVC=58n7w*2V>1h*s}qry z0H531O;{4Bw3Cagd51i2f~x86>32uWHo%cK8$e31SJgA7e2!TCHNj;%FE&!WV;>9F zcf$(=(iwqR7i<7;RaC8K**rVJ2GMDD+ui|m7~ zNkjUaDGYb}U{u`iK%sY1cpT;2m45 z#FgbZK6Kf-#i8OZt~ewtc!9|kt22wY2j?LX0TZZ9Iq1=nWTO?(IohtWFQg~y7{gZQ z#n#lz#^1=qqY+Y`7oRN&@_bwpI>fGkGr&G?9sZ+togBP($4q%j(6ddoAoHf$?$uds z)S(NwpV>W`rHvmJU35Bk&%Q2seGh#wW-^CGlU9i~vJ1zb@6&xeBDVr3PuLc(FV4)J z0QT#5FGx6h0Oc16qedWaS3*FI;cxPP{cm6DYY4{7RSKQBKY#?owB|cO2n}+e?RT!k-^Kh^O#NtYF>0MdC zJP6;j!?hRrXt5>psHUJ2HBk>wOZ}5-nAFLl+8|xs1q<19fqatQ`lspq{J6XlEUhGQ}wRaTs7YQ=@mKNLzP;T}1gP{$IzRv~A#=%W?r z;%wHR+jh+fkRF-^RnZoO^D0(oaz4jAW*pt5G}xE$QgxnhgKeME)n9x0Q~63+EXdGJ z+1`O&p26M~W-LHQep+?F=ycM-C39&nvuepg&1QOg!606UF)Ufy*s$brEcNO>Z#ag9 zBu$;8%G=JT2L|{Fi$BCuwx0G+*L!v&C(qz%*>CBbHSR`pIUR22x)bBuS!L=)aIzs) zyFi1Ya8v)CFtbtmGt)&3!E<7r(bRJQR|;*k(Zs&fiJ*TN>wbe5SZ-b3!+vu}(5o9Q z;C1$`C5hjn{(w}SaZEY-)Pl@=%7+57xiDrdH!fE-jOY2Y$VU>yEOBcj;`|E-TPoa# ztirlj`QYNsMkBogZ!p-I^inHR+>MT*u}?>#JJ`N$mtW+#J*o)zkLz(W#1xPdn) z$%lVekxQ1Hw+4_s^Wc4uv@~_%O~(HxQ(6g$g{ongLxu!JBa$DJ*EzSf$HnSS8hNnU zeDJR*OzYLE1kT%ax;9J-bGfYdstoTWUBFc$iw%QE^C-ICivS|AzQAtwHpy8Jncj;i z+wb>J!wCv)+AvZQHtfixTeu}lGR3|j#q$-)DPmO~KOD{(t3?HrIYrO^IBxo=$bM(v zLEIe(vKc2{ijiYjFI8_L0#tRdTJL+v@h794J3r`ld8!lmI;0900eTa@_l`thErg`- z`8>uBO>J(KymuWC=q|2 z=r0SOIIRlG%ZT^Kj+yI~P5p-t358VRp4V~>6Z-+~m>Gm$&M5)s`k;Pj55k?+I6ps9 zZ&4li_dm6yt^Bd!gKL{eEuYhhvVa}krN)TclP|Uk zp5Y`(8JG}~qZwQIW0l!~GH~hoUXnV~F3H6?E>vcbvk>TphFp)to}FxNB2`IRmkO_? zq)rP?om8K<{7kBgu#lw5=7etF!0~%D*jeU8E1=f`U4%9t3jH!ZB4co_c;&z5Ahq&C zQMv}QI11XrO~l2!a_LUMN9|x=#84ON+5^oQiy~J^8Vc0RfbFWUui%WbSwWzhXzs|? zPps1_>eM`-AVp60y{IEA|5^|IHz3m}08$cqA7r+zH3mt#w14ndx>D z=?Y@;zaGAhI9)K##CK=mT$n+1Bd4psthlcaZ+ZnjI=U?BkzS#YJTAnQ6b>${R%*9Q zQdjL0w~Rj{L!N(!zvAk?lZB{}RrSay+`L1tmAuD1?@237wqf1jbL0Y|$`cv(M6v+B z4d_X9A-_)QZ3(GMAIqofKPv?uX$ z{{!7q_V=a{7twFoN>EJZIc1L*$!j!MGh)BTI8Q<#Q~&o-zaq7H@7B_?op6h{T1JjoqkKq=*9% zWvQMCb(-$E20Trqz_Pp z{R>fs2FmCT#a@l!B%3)X$%l-R+4BhriFjkYl$NG{{X+hiX+Tui3yW;6dg!pRe$zQC zu?#-dB1ha(E?1n&2A93;NA=TfHgX&QgE>Ift5UYz8Gm*+?zn|aav=ONMVQ`-rrYz~ zhgIYVj$S0-VFQ#I){FDw@*3a*ibw-l{u6ea(_9i|1b)j2*=jr*oTc-W+QP<#_w*|AIV+5g$3mAzdyHGdm8VyNoEHc>SEVheTZSn>zqSkMySgdLz|)(lrIdHN7zvALq;i}tK_KgVfjrGAUNy+cBzjA_05qn~^;Ja;)wabuT&P$P~@nz@|)gP5#@!8BnA z7736&mI^?m&RCH%1QhS}9c2;xSo0esIDR=v?YN7Ji}2-^40kq}seZH8Vhc0q$|fQ+(5x(Ow@)K`C_Wr!x;PyHOnvK}aY5T`ZbG5<0 zBn`dh;0e#@?Ymv!Lf**nf}YKTPkJSwl`aEzx@1Pf?fkvUjapGoX0G9p8&0?9?E{3u z{4AymH>W%L5Eon1vr!Xz-PTyqF|~S^IYy+c?y!d!V(QzWVo%g>KCqP%+U`HbIL8ZY zvq%fWRrr#OvdU4|KR6?zPAIpN$Nc}+)?v}HED#KhVuH;zX|tZP*8U%y?hLtGl*zJj z*~qYJhUSmj?-a8uVgly%QosJ+U|jlHmOZ4Rqq|KGHMv8nfN;^_WDeYW4toVWn36E) z{4*_kNF=cKSM*Py-Tc6agit@)e`AjxTR%*02UA3>Dufvi;=Z?hk>wbLS1!-%&pG3W zFuSlLNo&hmihULM>S6T#`llih^n<~IaUad=J+Mo&uze9P<=U-GdINbO+Xv@Y*Sz3N zA!-qBl4j-Z6%w$f(>c^UqC4tV&s)q8$$a>7fOq zUJSS!)i@*MPS92RgtlbNHfnKZJDNCym| zCJ=ZWEUT*kWDu$HF4}gD`|H33;;R}?+Thxxt|z~liTt|a5XbFL zBjfa4S#Z#3=GM;H^7vMi071PBV$f(}YvMsLnk=pww~ifgx(VKgl4n4REa=|6;`fct zD2|Fs?psq-DX%-q-d$_i>^OgE2pf5S;S^a~+M7(;DBLKo$=+JVcf4E785^3qUTm#I zHzr$h<&@(_YIBHPeO$nDlYKz)Ahds*2kt$l`)UH6XrzVbSX}gO;3sqV4CEY2IG+_K z=&JRT4*PtBd^%o_ju$U3bx40fm^KmLfSn1TCGvX=Xols*BhjW_r9M~8hx_y0c+oo5 z&s=Qza1`O#C2p-M(E_~PgVZyRdL}3H*{!OKF;)$Xw|IQP-6e8Hqnk!JtB^$a$!Hbx z<&2e=3#g09y2L+bW*UM`#0(WlMTN@j>#MiFw?gY=9?fXyisg(?L%-jitVFp?oG3Oq z!H#mCPWJDKw-OIb(+ryHof)d!47Rql;TBOk>Dn>J6J`pTUv6+jno@4M7TAy3pL!at zJ3n0=9f zw_0h>G$${M#U6LZVr5vmDWFAPsqn*YY`-meGP!cJMx3TV5E1W_?KMh!3qXY7Ae@W) z__>#Mp=>UfI}19}?0mJ8VD*OBXpJX+lC;mzs$BkIcQlF4;HVykY*b`lc~ogf>F5I8 z4foy0oM%jBY1np_L)-azd~LF;Dk5#Vh71DiP_6@#MeormRskH_5z>0YK`$D-)woar z>r&e4zYm-FPgjf*n~{pVj8ht>aZ7m@Npv_ zEQRzw%SvmW4=~7gX;r}b2R||o-QUXVLlQA3pV(z=3>4If*=X` z1BX0skCpnp(q$ulLHAO_QQKs}i4fPv_d(ee zMp|-wk@Do{G{(p3nVD-zTI|}a4l|5=#J~u)|2ovphQ=z0=yR6S+MC*I4}lk)$Z8m2 zH?cxK=uAvZ1x{f`8TWR~EP1*&;bd#B_C32Vdh3eDBr$$dlu;K@$@it=AUWTL{b^@C zsn&ADkmE1%jOPY3mss=WeDQ=2k!0b4D4`c~Fp!j?O=|wFBkiM${`F2s2Z|DI%?YD? zHuSzT_=o8%o9Sfw!0iYam%Wl$H(jm7sX9@RBX|~B*w1?~U&U%8=~DqEv&|sA=|t*i zVBFi`JmgSP#)GTj;3||8PG=LNX%fbuyjd(92aBAkfsOW+HYI?c332Ah;qC7Z49wje4Vd|l=ZnF^ zZPv@K;nJJ{2k@w<+`0BxWgbJ>iWB|$6vaNOeZ%%*TW}n_C7r`kmAC;oEbo|7`Wf+e zP|8(%4)KFa)VA+O=2oYovtnU|7}*eUZJBCXvQE1cW*6*>y~6<7q@-#&iEFER_0_t> z=Au^7cj5L+Z!MX7UW4J<=HzASGuIIOhe9mj1->=bjD?h<_u?cyviFhKR0o8Z{k!j! zO;(a)$%?X7xXn*-@9Kuw_Kr(MT^8Wo{n;)AjqL&6g6P?c4aPyIj}UaRDCJ@a@jJ2T zWpQznFy%iU&NpdpUEhFD#Bb0wd%^Q=CVxKzBa<8UXhxN*oI+uK(p}G&iGLzP_Al?C z!(p?-#n!`rFtgmy(d*liLtp-QZ0eF=r*t?)~L z0Ed@FRYo^{QoM|7d8$QKVj=qe4|?8v{&*BI6LMO`FfcC0DpG`&Y_zVD}>h>oTk@OPLz~~C|MGhJVxOPbt{Ovzh-V_5nYDmTU zLaa`3*Q*6Hh$Ru43#6>uvHu+{{b&Mq24Ed5M*X4vs;=hazSQ9MWd6UUc5hAJv3kuS zwYWHFemsLy8_vrzUi1w0BFg|INcBSc*z)4y(${3EJb2oQ?K6o-g)hsaLbig85Rxa8 z7#m1=lBl(j^${av11i<-Ktj}xGCgQ0or>u`X#ryudhGp3=_($U&=`dh43%~hs0ZI7 z61lx>fxUCaLKa!_gdKxVnK4}g&V{R#g5ElunRVT9wbh^;94%Z|5#ErIYp8(m)81Nl zuorYMsY@-a%5r!aIs+vNPfE|=)QY+5#03j&DkF`wz{|bwo?>l~D#P;`37-(i-KP9S z{rFyYYNC|sTtmVu+$l%oVpfv;Bz=EQIJzolAXLxBLaB!F4v)P^w`OJ>bUNZ#X2PTS z`ML#PE(2JM)<`Ps?ZGx?!P2d8GV1t?)y%T_nK(fF_?o|w^X!C2xn|P?@QAxW?e0+S zVk@0T%pzG=tZ_F=nVj;8)A_FZOyT|Ja%ZIDKV3qV?E~*1@0*udi&2_{Ce4grdo<{~ zabtf{PhERs6Ueo(bU|}D!MW`tzTVH*7twP%zFQM?#J1=90{yEHhAf9M=M)8&dy^%m z_OZ8d}P{zyhz3H&$si55mfmPKvwf%9&Fm^c+B16Z1r_DaaufngHFA|=1 zGuc&mZmc2Z(_3*HDpihOQ7Sc|8^y7nRy4N$86&0{3FJHUnR~NMSFMr?KB(%%4jkI>-dY;Z$rOR5;QMCj&Hb@Svf92-%bCl7Bhg*IK zhk~)Fb4&L3LARmWaSKUWPX%M5?f=_KS@xHWlS6PxsHdKvpw!>o=|7i+v=PH}#cQsV zy6M11eLt4@>$M46Lk8$SluP459C^ zGG>AK{x@~9u84~l8#)Ko3aR_hE=JefOskc{y4`w=4r6k}GOqWjE1pKkYcyy5*%n3i zlk|EhMAzqe-_ISk-5!7mWu-<|5f-sA#pm>q^2z2{u_~S{I;niF^~5|m$Af4)ec~V{ z=Mn9hbrA!wnZjcc2`BTCeUd>n^KWtzfkE!uW2j;|AA$RFo?%@ZlhT2!Ygss-<4s|D z+=wX3i!{g(VNAC>F@b)%(U|O<;bMQgD&55)%ngf96`2OL&!uU>XM*%LRsz9Q=7)|4 zTSpDDHtA-61LRHaqO0YP<=b7u8(<8l&5~3p8Nt|z3&D+b^0q6|n8kk^(3!XqKIJwk z#~JfNt}~y}6p+}JOv|SSy@PtT`#M${UDzv@0LCFhiuVj>5z10WbWiR?yQPxA-(#^< z4ux86{l$b7mf1vCTb6<0S1{4U_ldRZ$=eExK`gL$C4PF=8>E;^LXzd9eHtwv|7EB# zn)U#p=rubLV}#dI7m5&QWa`WDoje$g25@PDR(~1JL8{V!XPQWIY`yI9%@Q%kZ|wj1 zWa@A9PNykLy%_XJ6BPQ6apB}v^KOf|&TrXhj=9ven$H&bwae7Xi22#oE5vAHF&yX z+qSJrI_QpV+eybZI#wm=*zVZ2ZQHi(snh41nSU;3-&Wo1{l1Uav({T>ZJ--%5GxLa zr?in%IXh)UuB!VAbDfY6`NO@GVZus`sN%u`Qu635$yZ)lDjUcyjTq=IhQ~DoZ$S3O!pwWL!03!0G|jKzjY^-k;++m9a~mAnEHo zT|E<+nk^o4ln6+kn7K6>$cWin?32{#8qjK(jyLVXPs+~+TE8F0P03=J;DtYGu@7ST z{a@#^V=I)0c0NVcS>le(HH!^F_qUH0$vU{9@K=8%14dLNZCrN~u%5{sGefmyy8qw% z`jxJRGqa~HhtZhb;inZL#c+hP=|Q?KdbHob&y02owgbdJiZ4ZX*!B`y?#Q3AWnz%U z)e+8!E7BQKo3KAS1~?m#MRLE^WwIwkJ^MU~Q&_zDKPw$r$bPCeKgXrsM0+K``21k6 ze5l=h3eoT8k<^?B^NKavj80#!kH*?8D2611!zG?r>M3|OI$kZt zkip~1S6S?q)@qOPwJ40kjI6(G^16;1j+&G+L$W)o7mflg?Fu{)@x;rRv70xLYHVWy zk5yuXQ8NOh_s3@oH!UtKq-fNrBsO1D2gU3N{(2}lFj7th zV$Moxe*abZK4@h#iq>P2i9mDVm_3wG%O9kg9#kz8ntf9aFwSO)hGLlqN*aw8U#&;X z$a1)IG;Cx>1f7rATp8dz6Q6S`iV&)qW&-+4&88deDPTVnQpp>xDgZ>*pictfSEVtId7Wu$SN7m^xM1YGZd}tx zebG`CNF!wejbu%F#^S9xWA`o1rfMw^1(OEuiU>1z&uXtuh3(lE4R%shq%HJ7d<&$F zNs+t1<7f$}YAyPqOa4TO9|Kc(tP>^@SVohWa}`jW`e`ER`H$0ESS_dJNBXN2OH=^z zf98!jZ9bz>%4uW1X|##2wyVTi0XXnRV#`lcjuHo1Sr#e3-BVSVoX>$6o?V!k%&E_Z zv>Uj|63fkgB<)OFwwu$~vnd^3a+JNb@+CNshpoYWivnvdaWcLs!2p)UYI-oJh=FR1 z;J(Stu;$OmLX9x?zUZcEh@{&glhfqsT3T+P3f{aB?)_aPHo12?__V+LEDel@jI^6a7r-4R){in0$Obb0HG zd^En72Lh}`9->Os*EvG=w#057z}78*TUCKR7CHvJAq8rS2+ufUSb20!3&x%XRMk79xK*jsr3b&0QiNqd1};W6n2?q?k@ zgtM8ok+H@xm-1GOf$~L7nA$+6p3^_ASeyvX8fXb#19$DvK8W5gcLtN0eBFcZua9Q` zzr-t`1yBGxllpQ4?CxapDh?x+LC1o0@iVX?j;dYp9!C>q+?w@cLwjA{_EV+Y+^zmXj zJ*bSEyL#t+r8xFZ4~@5qXH?;oa;-*Ix2vtKU4IC;or?=XFZ`ZA5^>mEHriMAARRFO z*S9R$EjpV_o{3_qM)$K@Kc>E3ehtkU40@3i(v)g7@Oew$jdr(?&g~6jcl@(h0J7ds z_AqwRt9hA;bneOe)@ZlugSd;wUF1zh78+JlGP&)YJvlr*E5(T4-+8I#vhTwOx7&wl zYNveqAzKbP8^Q5=i@cgWF92pv_XpDvTu(op?gvOXmcs#x4e~rz4CkMq(Tmq`+$|U3 z)lCKwZ9alaSZo}>82^olob~gicuwo!Ytp3s^8p_rpLgWtpd9$fj(!F)iG)%1nhadOeDp6=$}-L#T++k5gNQO9>}&)9-d(Hun+>Qo7$)#*h$WAo=7#3ldJtw<@c>{}nXR6K9BI|OR=}FCVkZn z6b3W#ef48#MjQ8vNoV8LuJL$6PSUloj7Bti!Opd# zmY+)`QnGQwCT5#g7%SZBB1;qVbIIAi^f+JedJV72weSOEzn-BL5(_n~p)+P?)uCK8 zJYhVo-7`0ny|3hZ6+h=rp+kwZQD^k1jv3d$SqnZgpf5)_a`; zj`C4>rdbGWjaXdiO}?m`ZFYWyN!5n@mnXIt0A(=s@bpiH(pKXL{lie4?xhlB??#3K zkril2=v1|`pL$^|h0>Qw8rfW{New+$rCFKU`vpJ$`C<$+6Z=~E2jyZ&Nhd?%o$|Jc zfdrfDU+y6i+VUuBn1Plb>+>lH=B$^AVWH(?Yl-`KpJJJIK?j3=eiL_VA1g&=(h#Qe z_aKZ#+NewzhnvU^Y8?*qFI5ayBg?f%ANRr9!^=+GpUgC(0ZC zUuN45BE2Z-^I-ukM*C0aU(>BJVb;O^$|dvZ4Knx=Sob)ITX-n`*O#*3Y19eL#|;J$ zply_a{yuC9Ss_*AIV7&#g4!04ZpJ6f0mnM;^4t8hWV! zEoN|~NfRaqW6NV2Eve;Da&}&WhNATthSlGZ%fUf&B%9gs@S=)Zqt*FJBj&$T1^;WK z%`45hpL6l5<*QQ#Dq!h&sUUH%YMk4arYouo?*HK5#7_huofYD*Aoa&;F->C7_`bl9 zB@fW4r|^e5#7OsNA<{|6>#}|KTn!hi@$l^LfifdE9GEjSaaGe4QM?VlSR4@By6#3x zSJ?ZidaDy++#q7U+#%*%v3hWLWJt|Ji8B4+(T|iI+6oAY9k${N$E&Szy9oDx;opL! zkhLlkB~8X0sVU2ec|YpwP}H5Fat`7iF%WGu`}RoI%~i|c`$y-q-5(XkbPxTK8ngQM zv1qD~bwR29cVxIQsUCKXA*CkvfAu4j07xluqWIrriIKg&`4n>14wBZ-Qo61CJL5 zs*@lN0nGYh`4;m1Q}#UX$sfQEGfpX0P|+6SxaRROJ@-PII!&vn)Spicb9ZP^sk!``u6Fd&;Qn zj64ebFDC`>q%|a``n}hCo#7T(K^i26X6Gl8p*d5NuWZr#n?Nq6L1GFt(vpp8U(C`B zRm=fCgqK{m_nmY{+Lf;cRo`;$*k5o?3J8xsl`^&oRVliUkIC9`?%A~V78<8`)xwsA0`$ST$khET1jArAO6(Kj&tcW67?9mDJNj_J4(#oO zdvlj0NSUrstGEKre7KfIxUWWjM0w&RWdgw|v}tT@sBG7q0M02BiQZZo%q4U-D&eu* z5^K6x_60mo0^istbcK3{8?2{7F z(O0F_rBi}oQ^{^7LG80R3M!M4_->jMp*QR`a5|9um)y?pYIT!v6Dhw8R^tOkp`U5~ zRuq=A&LbA8z!J$4>I;~^wePK5W!Xb`J^FBv@$+mQj5J>O0&W)K_SXK&6;WJFv!o*3G9v zyogA6SY2&WPETT@jW2;3wt7Ll*MLKicA^_3L|9 z>#)*azX(ou!;voxP8JOZL~w`|ixj+C+*7})Po?R2X+;9@B~4?>U4Obn&OYJ&Em zO|N0+vmHEUt(!}w|gis0HGowq(~BVLX zdfkgQoNr9V)N)-te5#n(auBuI+P!qCE|kO5){49}dL`H})@B)D+1#!%nch+;`ED(b z&*Cq&L?wJOfPgjmW$N{Q<*(89+YfO3?lJB8*PSINRmXbT5pkI-z3W5aIG}kR@1! zu|(~8)M+QbW!NvDj{C05ZC+NkqU*X4iy9(0-?!f)J-&PYf?ZSz9})TfG+bF^c$N+z_2@TFG*Je!Ay!YF&w?fn3*o%%y@hlZPp@16;(foj;AqvR`C5uQR>#q& z>Gy0dGqvbKsdiY9bhlXACxYqRQ#Nbb$e*xXZ0;s1g0%SpLmDv)XZ+rg*$weGNo+mY zt7t8R617Ug^piVn0#zt-=sW(=6MLN^t?n3F=hdsu?mTRPpCACpM{l4#8TMK&*|>#T zcAXIePn5>+Pup6x|J&w4$Jou}j|_y^hHk44;8XmX8RaJ8S-aik^KjWnf1aWc%^oYL zN4)rWwzJS!vE&7H;%fU$U0suxkWcJv;rPXsy{*kDWst>ZQnIE=m7+0Br9Axfn~THh z(F?1HCT$ZsX~hRGFYJcSx2^>SttG=`&PtO-2uHzE{E(S>nl#B4&rl-@Z;spl!_vW? z?b5BG+JpvmZua(zGdmd8AA2R3Hm+=KYO zzS$6IP8@L}9M_!DZtjRnl;SxaN__^Ms#VxgP`hWCtq!~kQ}Bcz-2Z5>y6=wXrM?;`h$_dCYi$`%abqsi{mbO>zkswT3<^kK zUz_WUh+HW|rk~JV%O~_`6k}QQO=x6}PXHpsJtY@sCWUHFFHN<`r2z5u80ph-qqwau z1M|%+!|9V-7EAYXv6u~%*IDOKYNf6I+fFHXAQM20wFAF$@8pA2ea*g$hHqT9QS zzMANUlw| zRANePKX@HRP}CYM`CQ6c&jm2pZsF)f359TVHaiM_XjoxSu)6g2T;8#_m;!g1`kerjsZW zx-C8Xj)TkX7PQ_cdz3NT>JPWhstjuZa^Oa%kAnqlF7Ms}+<~YeYZXq|JQ*gdYuxr1 ztVv9n+f9Rv3ryHzizm(E=1)o`Ay&_}u7RDlh$50$7D+W^cWGa+mK!f)Oic3JBL_Qb z110B2%9`ews}2_4_n*!r&fF}ja#T~|{hd9SZ9X(zJn#?C`Xkc0+mg!Z%-Cv_WNs)~ z=m7$4i4*#@{a+hU{zuS*EA1LD0ViMh3OJ}-4H~?z)oyc3B=c;nad;5tnI95*8uhGp z8lcSd-rfIym4d%t-Gb=X%gcJ}(0De{&@%Z8MrUFH>~yhQua9>dsOD@1QEz_3~DdDsepX6ls- zb51the|js)Z%WCH#tW3I7PCZS{*U|&_2l*2vdcv+$XUgXDC3yTVpF#mm@CpMGfF1s zS@=%aZg#&H!TOcFYBxWil=GtW^Nbtqj=(sTYbCDzL^Uc;!yDRNiQ}}T(2RPCrtp?R z;XIp<9F}W!Bo$uf6u$Ler+dT=ja}HY;!cVz{V34tg+(8W$8T+N*ptLNi_0ey3&VB7 z?*!(*O9i1tzPC@@%0xvuxfjcqk`4<()%v-euZ|MH(Pz|G^4bYtab8!CXhq^`ARLf` z3Yqh3QmZi#Fk!x$PMxQOfJ5?t)SK0_EJ4yL_7A0Fhr`@CE2!2tn_<@~;s$u#7&pGw zh91(JUJ10rMIRb$#ut{2EMZ9D%y&RiI2Mqlv~x)2bA&0GQR+3fIN>z&q=>rSWVB+nO&^yT55% zEkrL5#sI6?vf}>{jg;i=y$2a9aDYE+Rb434=9*NwUZk}eX>uC&2IfkF$QwU)W)1NWXj%al)sijCN&4JtWVU06pd204w4PiIC^)aGUhN$nc; zGx6~|gFK7#!NsN-xZLEFnS zAE6I3x>bg`yo5!b2{vp9tJ9@UTzn1@?B2t^<>pqm#CU|~qfq%|z{z*mK}h9r;<_m{ zbU^&E?Bxkf?*+>hrtTtX=&GO<#ku_0l+F;6Mkdio?%E@5(&jtIP;`ngxO$V3GeY$0+0R)n)c>_J zKsj(U@fN<8>n549iAx~0bReo_)^B&B_mRqc7<0nZ?V3gB|c6eY6 zP^~`vE}j9$ML_eOuB}}rNNj!;&H`40|KAhMg;1{QMaF zyeqbTW><*tEfaHnwgZQJ3h5-0D5IImRy(TyeC%|Im%EMJ|BKxI-zUwA0wq*DbG#e1 zcybbtQx37;RP&Fx(i~`qb@O>s@-d7yvR3;2umqNzQ`Rzb*D-6}vN!>zH5{wZ#`gF_ z8pRA~s^w-LX^QBh8M=n{GHO1B(e|5PU7ZxRKhM^olxHo7Ag>C+FqG}KHatBAniu*0 z!9Y>4_q4ZFkgo)wP|+ePui-s{c0#JOo3KviPdD&H(1j&m6Erba>d+w|ynPP$(*MMR zn$dORagUz`mFi#e!D%4|P*PMvymfkC<1y3JL^efMt0ro{S&gJx-m6hb!_FGkPua|| zTHQZG)jN<+9+PskW0I1R=IX7#8hzU)s#Lan`Q-Rz{lmLlY+Qb|!KICL+JTpEFCq`+ zL?R&ccYY$@NebRgh79oRV0muW-q`I?_BAA|E|kpld?`q$btiaziM0oo@}%>aEF7Wt z)R@kK{KqyWV$qvPcXgBR5E9lMds(FSUbGMvKh?V7UePLgBJO&L`1mcUYul^hKGUPo zw7-&D!ct`YK)%6ZIh4oU6T6e8@~HV`cgOA1oqe}vUlPnrj|6jf%iS55qXJ!* zXu|cWBJG(ZS>eA=ae#QvA3`j4-w!Xqw{N9|@My@$za$fAiwo~+8yg3&cL!B=^TeV> z(>Se_cZCH4c~Rvm(g#weeC-df@dCw0^y8=DO*I^I#GAvbisVtbWqf1-G``J*?N*dK zfvRy3C}U|8eT0)9*Dpy@us9pMO@WSwAN}L2uGd%3c*4S>B%%g^@RT?PiDXQ$n6%-F z7@vaqu?N?Tg!zGUcbZ39@PxeBK0ZDr*9N`822WSpCD#BB%Z0xAQZ@76w6p|`K*Ts?i<93tT##(Q8^KUc5(@{4SRAM@ zKL{%0t9}oCZ(2qAbqRNmqY;neOm)^H4W+|HKd?0eOgySOYk0x82`ssQs_vVv!o%-v zybGRRGCE0Y0nx6lhU}l;78P`eX{&qrj8ns>q5zp5izG{X-nSG12UE~Sc}Rvu^>hy% z?{)cqOfsr}>JY`xwP>;9zd~|#%Xe1EttzKOS!11d7o7+;{utQ#sA05Aw%4t{P454*YG40DKg>N+h7``Z({5w8I zUV4??QD{_oT31ba4wt*4#lh(*oU5zrd{a;m#M}G3%5IjsyLD#-vq-#$wc#k34m}J&;kZ(AMBHNp8A)v9Yl^p)nAkf+3S+=%XA`>PD4W z;m-No+Q#PaAP)w$Px;qZvJK{E1Q_U}T8~^DUkSC*OKhnoAVx(wI$7wW$V3vqViuiF z5w=6N^cuxtwV93guPtvRbGm;OML3C=D96oRL4q+?nzbZ(i7{@fb_8?D{dKRipT zIdN)jRRE3T?ivxky^l7B8#Cx7tgktSGPz5yfGwhzeB=5*>SXN2E#m`PvgcX%5ZwZrGr8hTw9ZO)eKA*6Uc zaxUB8{Jr}nZs)~zX>sP1gySjuwYw|cKfTZ86eIAX!Cc6Mt4jYNmE>n>svALg*Vf*O z)jtaHi;(*vm!Of$PK*iK(da&Y%Az2WHG0@|F<+`OAPoQ$C6I5-R*B#(earM1?AlW? z1}7r}d%^lZthT+}UF#lK2bh4rG}wMUwU}2Jtia&mZthuk`_AAAnEL}Ms@V~{SX z`}LkwgXPc$|ImB=crxr}?25ECl`xt0AX=oqiERSjdZO@ApG?+<740W9T=_2MOBvw? zEFFB%KYx?JUa7^fve4?yS=E=;Y$Kq$5GL>C!Ocm|(=$&0E7YeC8k>|YhKqw+sy`SAEa(-MHpmmtWK^! zX3vDBlu$%(w1vzGh7pc;@dC7>L0ueo9#y!Z1g?j~mCuThqLL4eT8bf~qyJ$6Cw&@W zBe%>G&Lxt0Aac2EkN4u@AF42XoGA6Mad~0k=jR1d^zvl?(D8J_1t2=AUG{B9xL%Pe zZYAKD{0LiN(AS|)X?6RZ7$h_st~bI?*Au1RxnR}+?}`Y%wQ`tbqK)$q348qh#YVA( zg3@Nm55uT6FGhslc<$O*_Reotkf^L#X{gKl1-m~n9x+a z3YaoO5>bJo-zE4B^ctdepzq3k-=XT0q0|gYZ3qJwTZfR^ zQz~2MVeNqUHUg#sL()H!3h*GlzbX1J)>LQTntFHI^SmHV$uq*9kBclJHXtj>9 zCM*t9J8h3PWCbyU#qJFL$?C3m+6b!3WyCUXXJXX32aD4TUdo z!wby#xVDjvO%#pOC75c{U}KjWnQGzEROVEe6qY~<-pcsmcHb-14UMmR5};(;8Qi~Y zi8L5Kue4?SDX=%3?ei4E8e|gjGW6Iyef2uKvN#lo;nU4C>Aww7Z+@4B^|nPOQ_2jlve^%`hifp{JfJzU zis9RUfZ*1}*A6>(jjdJr_SY+*yi>yi!+bb*zZU5$pfAtDX}iUx=GrA3U^9F2c|k7t zyQ6ARAS=+4`TDP(Jp&Ct$60_w?J`}prWJrXT*$;$;2wA z5BFBl?KQsuH~hw*h7~)%J_($;Eoj2ot-=L%>*r(<4QEl!*^L<~8)>mG#k}Dw^xDz` zFL5>{;b&1JnTll%ls-vEfk{CPHMEw^gyVJ8J#G}=y?_UisFM2m_iRP59L#06&Z<9XD-us}l7NWNb zZ#~1e>v1q`ZUXSmbG&Msc?Wu(lYn_kF2mCe{q)uka9fj2M-o2B?Z=8Orzw7`%_hv! zB&L4Fta_?5=-2jtPPW{`9@9C}YunNOd_~*N&X#lh)D6NHo@>Zn8%8gJ0s3@{K1L2W6 zmpHMlk7s@StMy+h{8B2XwdL_DE}k~)o2x9gvhn;5GqdZ(%x@buif z+}!b2RDIlqk+v}{)^u6ZU$_4Puh#})@tGi^llUzP9t8z-WB=Og3@PM*aIwe>{g7JQ z*N-w21j|v)r&7Ub^mEt9P#<<2P+kyHdU653ijjUpL)v-^V&>so>DXVuUv zpvK~>w&tS9IFF0@>O4e%SY*iI+;KAZES?f&jxamoI*3d%lVY+k-Bv1LAe)$ZZf+EvPVOWU%j`j z^o9?F?GED~U*0fozLoCjX_oD&VYnYe>|gJWh=KLxQ5N$h6!<6b1DQijR6{5}{a9>J zJh}fyTss`_M;N7TK1av`S^ThW*{G`8l8}aggY7Nr_3C=+2Y%4tqP!lYj{2=vY<-OP z#fKOZ(I&z}Rc5-r^BbjXMvnI2&t!HN>X)~5WL1O}qrRCuw{(4hxb5LE6FViFG!H}_ z7P~!RH+N5?fjJm#t~Mebj(S0Y{Z9lP>|d=CGV>;t$S2?Mgm{*+3!K@Uj=A8CsQSxW zxeDZPqGKxBx0b39%e3wX>&6eWU9sfDM5olXloyvE=SwsUuL>g(+(8%_52Qe&fR?Ju zBY3q(nB)^arF#B6$Zy@#6R`Iyh&^oqjR~GXl-#<{GT#T8F5S)Ch`gb4v?%#Abwewz zA0r>UtFF?!zDUQ9^sOxg@g?n?m*UXn#mHpZer!{*mLs&TXzT|0vpl&@GWH7f?_IHo zx>LS-{Gm1ECtk^ZwIj6~FoTHlI(`H+OZ@Isa8#Zw+u&hIve*Cg1*u&Ag9Ku>?8}A0 z#hS*liO{#P2)wFd4x|ZDV`AKWIv#|Cx~V40jlUdH+X1G;u+Z2x0Pv(EdsIfF@N~6d zx@lC{ehT&N^oBfb%99VpAHPc3{+HKIM^hu-dks?Q@!IDk;f!FWXOf!S*X7CtCnwoa5}1~&t22tW5T}3#A*n_mU|;p zFGzvxpy$)?yQQgjUjg;v3_)D*YF<&rOv;U2x*i8E*})6haV=9s;zbPnu^)Y>|d+wq`ifQBQiySW(gsOF^Or3t;e zZbSI&!R33y<l6W~i+_D`OW*gr-PDJ5p&~$p#L~o`N5Mni`u-EdK4{J{iDYD1 zsPxXkvaRLOYjMWnd_^N-(&(+e!5?iQNnwZuIp-Ykd6=dtysY5c136|lg%(IBP{~+^ z9qv~fwJq{?n{_7#zfp^?bMy1f?CuAAy-UBYv%v67&BRh&1(GDC=I|@u*Y!00Uzdga zdHEtG+mlC1fM<1XiV_Liemg(J!aZV((6ht_w;>ucNomWQT?`~ID_Z|kB$0C`5U~1(>2|6<1Bmn z+gf8?9l1y4dZ5of>bm(I+S3bdw8sCZ3YW>?QE~40EVEj!Ibi{{ph!t2X+Z2NeXEVO zD!b3OhhpaDMHbpwX77}g%-p9*}{|xE)!Y9gD zM)_+ldgj^p7jk(_@CHmS2FFRK!f_fU_LjPLS7Ozwbg(dgxPhI%Ja+BnUB5I)1F6wI zoGyaT=*)JzTpgGkjYb|=KJq$dLE-u(1ksds{OvYT+UV?cxcj$d{CeJjE ztPLoUvyTvgr|{T*@OYG4-)A^EEuyuf-K_;4uZO#qIcL`C&f%6CLcZr5ge&*gKWI? zwD%xhoH?X{=O@&~sI&fo`O+=VgJPE;#PTYw^E5D{ebPypAGZ~yIqq`F)SgCTOMEz- zdf)soF&%LnKQ2N=7CSRP`wQ}?tBlm%*8d=WBnde7+~(8cH_D#(4)jM<9ZR>J5Fd~9 zn)GS{5`d4-hFu>0ffL=E8JN2sV~1>&*rxgCQ#qDe?A32z&ukQLZLvX{J;mN>Rc_Ow z>0xR-q&n_cR_eRPQ=EB3Ox0aYa#{Q!w;13N{x?tXsRQD$m3-an`=hL1b}4H1?=`}_ zM5$RN(yOQ)Osa>pDjC#~z$9zU=b#Vdp;v1Yyt*K6Q;<6$Zw;Hzymb!jM>QT(LG@KG zj!K=U%M+Z&XU2TPB<@xEEXHpA(dsL?JkD{YO^lBjL5>qsx<^!`*-lxaR`DyBJ2I%# zcqnswYOY${mkK5)OEw0;~lN)pUwah(~GH5!f&kbsTYrS6F~3Y)a= z71G{+ZLwEX?AZYj*}12%LDI`fs^g3KUByLI59uv#P!5h}OV|-UTN~KpKc}&&Z}Fh6 zUQn6z4%>MQE8|PQ3=CA~^8+5*FF6!&uOF^t!FgXjMA6f5CWQ`B*^MRbEla`~B*2!XOo10YzDWO))jhi$hEzv)uv-zbS6)s{IJ3;I) zCyE+}sW-8+J-GcgI7g{hvq-TQ76G?WIE|BGdai=sLZuWTL>H}!(MKKLDE^ogJ!2=N zACk3X3P*?_VzBgsMC_@aFPWG>~Rv;|7&2m@_Q zoe&p8_4MH3eSrfv%D?`?Agu0E#@@<6T3C1!$N)q@?|=e358lmIouL)ve8pMf>pxI} zV|Zq>va75w&c z3wl+PC1!pi&4lIGW1wLG?#381`jM#I1Q6*UO#OGUzbcz8t$v7mGXM~Id{J*+NE9)KzFce$> zV=Ez`T{r?~GGxz5J~aU@%tn{Vo|fB76{Ty5(K?XCh)L(gY%V;tVk=9@o`eT>9AWI18m3~3Vy!y!973{x?_q8{Dn33?tx~smKqEF4 z6q^)M`@5Eq9om+5fCEO+=GsN7s3u}Ym+i^gpN>27f3rtQt=UD*hK?nUf)MX9%p5@eZW-E{%J`duLEy#S@D z!NNK55H3+<6r}S?*NX7IZEc_U>9{(vFUmOR)B8&-BVMJlK_@n&+W0M*g)ldq>RWYI z5?J>^!04#i23hu3{tz_ut0vDv5S>T~U7*;lREEE>B>*z+zFlh*_ePS9Xv4H0jtl8U zxcPM`&h6*hU2wn^f~Nu_YPbg1z;>s4^b%Fcz7NePM zium;O*Mlyrgf8IFX7Ma@QO6%_5@vIgb`_=Y>-0v;n>pg0P@x0OJcd=U-dpCsQ{Fpy zg)dPxxlUf;&i*^qAw11N!5m}}{1L)$t}CcUY7ppiCz3hpKUk?hK#|m}F*lqyK`Hr0LAMj9|Pd^_|%i~ZaopGVm zBC7yVg0XPUug*d;Igyd-yzQ=Xtwc(>dcZ&5spti~agvyPaRr$!WI8~9RDfu@XnQjG zuPLhi7q;7K}+KQJXieH4T`?Csu6Y{qh3tVA)v5&(=+1E>4kLmYZ9Fc;A#^@ ziE!FGG6H*#$NuAVVgh@+9=`Scy~V$idGeyL6o-K8<=f=jat`cHh!}zTgD1}FQv>*tEd9V#x9J4_v1F{4-^PpJ>GZU zRR?U(u~fjjfX%7c?6%zF^>hT0wKT)`KP_8sVbu}QQBhB}!$jVw>u#O`od2-Eq*M@H>hKdZ0*cpmJehT{_(OF*i zYLW%o6Qh7oF1=2g4WK$bjZ#pD{h`KRdi5ec+A~nm%a# zbM5C1D4f630pfG6U%AqF8~DM{^KkDl`4zOpm?4ph7q!(R=Y^Z^$SXdV1ohv~YiiZn zG;2yPNd+GN1d0p)j999JE?H?5%OWvi`MDxJmhlaX%}lx{nZ$c)q5>8+%_1o(DOoFf z{vV{VWHzi?GwHxIB{uav@dDJ_n?pbo!#|b$zVs#)!ITKjD&QKP@aT8fq7%af8!)*z7>>X(# z1t&9F_PttlR1pgkr@2X>hDYnZk_$E?iP&1n<;MlI=+BLrED<^py7mf;)956&QL8sG z!%|_xmb`?#3#JnFMgy&3q}dTX4G(C20y)IL677@@2KChMk;7IRoE)Eb3nqrkRUY|*|7W^J|mAxNN6 z4ekyKdSlTw>gMcF+k7{X&fB5LHdDe84_te`6}q(QOVLl1csf-r>zP6z6%=Jt+OE=N z$WR7&W2KphVDd4^$8TUC!^2FNaM_KiCZlvzo)lpbeOi&Rg_F=oIWhgcHTg`g~5ix<58D7<@65@1^5yEGq$zk#Pc*wXC?5$(qNRNzY{jdP> z*@OF&t>!<`A5m5HHDZk{rFJ`Ec6g?7C16|2jWto2#VdP6y4w%#zE+xBejiZ{RSlQH z)1|WC2hZZZ2S3Kd_Fy)1cE~;^mI&S#r!Z{yt5yer zV{QZ@fli5bstOZ}*^NpM@RLEQPIQ|Qi;tDNFZX>RP}_oog7leu9%Y~pU~K3QG<=7V z<{9SV(xM`LfpiGW(hB{VcH#3ZEveT*@qt;={a>6Fjt>XT5kd7N3JQoh7I_jpje{21 z*@x!E<=CI`tI**ES@mH|l$-nqzMp5*PPvDo(tv=6!S;&Akya%&D^)c_`P(~K`U($5 zyQa&p!cR26@0?^GNA<{-X7?iPxjhWr&T(Wn-85EdD{$pJZlqhokIvMkbo?qq@adJ} zxJd~rI=V!B3G()`RhO_R7H*2RAp+>s)Z4u{}bCt7`3q5tE*yMAYM-GT?gI!Hpp^;~Q*LpO#f6YpF= zq;orp4%%U=Wt|j#D}URpBJ)J1Us}11yMrj7c3nWkztZuyXN;;USA<`lJ+;hA{rz^1 zQHIBl3?~@$vLqAjzu?VcIr!sDd*=(5n3gh)DPpe~lh{*&Gd7PM3AmXpG7T%}rj;4> zWbOuiNvVF5!R`s7rH> z{`LiIdU8gZ?|;@B@521|L|hzrA2$PbZIz0^dK*cWc0U!fA?&;ijTyBG7&6vayi=I` zqwsqYDU`N14sF@X6u%yXQM4%}ReNmM5HS2n=u1&(@_++0%MG$_{)5k52O(9^k~g`D zpwJx1%}UPtij<^gv~gh;Ru;f$xHpoab0q~l+dz0Lq_d1Z>`fSL8cv_l4~N(IEmFi} zs$i;o%!7zVDgAWNou`FxKNcD~L(5Y0J`XiJ$SKPB(eq!en#&$4U-~26WT}h@f-X1ss1P0wYd)s!dTg!+Wjcn=gDJf&~P9 zutdd7(Z&$U1X-cM^TR*jmpXnDj12NLg%i!YZ`HF;4Wa|6dkhNjDM5nAi-Yz@=1m@g zP|tMW=ih?=FV?;)D6Xz+Gq`)>8r&tgyE`Gc6Wra>5Zv88IKc_-(l{izySuwj=Y7BL zueq42nTt6WT}2gD-RGRW)?RBpvghfN_TM8Uro?E|G=B_~ZBcvzRK}YB&H!Fd)ayX} z{n2XN!L;lVT3SE8uGtRRWZo2ACM*?{p8Hx zL@@W)ZH#c%dMMCin`$85qSks$CWoliW>rHM7GtdR*jo^B%)|8*?OfZ(UZ*RoW<=sAL;9CZ!tbCrjs(XxO0&3w;*oF z^z0!M6X*zTh%_z} zUgjeO%>qZCc^>2pAVAqL_#@-VF?xI#bp;|)lgPWO1o>t)h%NQrbTpWIePJ9-efsxO zUFzuuGaqR2C3BX-1JQzY6e8k&W-wL!PQ07UtTKU`!B&dG@Ol4mm-VWSto3miNxmF$ zqACnO&<_}5v~U#ZS&;l~7{k|!b6`v;0REQ3kj~Jd$hx$U`I+&z4MO=9*a+d;kwFfxA_>nD8wS9kjrc?d*Vy?uHS`r(0mI&%dy|oYz@IJV5e> z9r+)m3CsjKkNGZn5uG|5iF9#2n^dtb39*z5JFKhG~)0J5^YA)IpYQnMQW z!rBnt*60PLuVun=4Ed%!;k2z&!kl&zXIq+UQdR9m8k553bx+p*@jpcfn4-k6ISU3z z^1pr@j0|kUxYD5gjtyG*_0jsPx-xaVDn8*1_K}E|?ByQq9dxI9Y9J1%;z$_m0nSpk z3-I;LFI>E|KyzTkQ-RU36Ye)J<#UACyRD3I$bHQ{I21&CWuP$}Rdc!6e() z+b+j4rYAtu@L5)dfy*Im6RULtjop{&->nb))7P=8qrmY>1*Li-{%|T-&Wu(Gp7PA0 zly77*kjcw~8tvJ7K6R@|Z1k;(Uyc|>8XGjQ;5Zl}jh}JR+Stpq=~DfqiJcBn5BQu|zDLN*OQWkRq2RQ^0h7|0-1r3uJX#v?tuaFTy8J7v49*=6>@blLgCZL#Gr7ee1!;*@d}@< z?k7k;O_rdsaISeSvAe0;3DYU>6RjoE2Is=!yVViX;uHprGrdSUW;>NGXF9-;6>Orb z+8Vf&`Onn_6E7w!$ZYxc^62R9uAz)iO-A-1nf{BK@}Y%;Lo8VDRz4sdOA^gxw^TY1 z9!7ghDB#KZ^75j-4GKX4U2KBBY&%)cRfLyoR+kNwDCeT;>gvw&yc+#-T4IQ#yUNBaS&$LB-xzSbVyhL*dn zC_$NT0+31Vfu=U+jcYRc2w>93@O!JKbfyH}wGJC@$z52jL{7aBP>B9Pbi34q$nKab z>kIiKaUCP?1kO5*^$Mi6dgoDU>k9!-9@w{(02;AB8&}b7o;N2bYm$Rv8A4c}%zlSM z_RG`VS09z?k!M~gq9xJ)Um?`L58(RtHn^>=&3dJ`7xL}xt?Y8Vy}fA-D^*EKc0W>rPbCxsOcAU8MnZ4zta-zxDbS@yTe zL_Rao(tWRp2b#jw)xb%M3)`p<0ktw-uqcqf;0x$z9e^_!pMf1DsSj6w&WV0TlDM%e zh~%>WXCTTdkQ!SAlBGVV!wt&gLd#ofJSS2V_6tYXwCUe1%hSljQ05mKcKW(WMJr}e&KwR9h+XW^f^yoqY`E45TO5!R7{OCx#ar#)`0SLF~FEKboeAvVi5Q6w<9~m6O9u3WAdV zU7iv*K1N1h;6%botrbz9t=UZ#{#JvC&>bO@hPNlrom0K3D)!0k_=h!$t}A{WfrWu( z15*gCsh2D1A~E*VPg1gfM-Sl7C&hni^89*P+YlY`_D%alO^^5HE3`1rOMr52=tnhv znzz1s0X~K7XrXs@RpI}M5Q&@T|NI2aRufA>I~1HZ)5wHk%-cX%O#A|o)10(UcbE|E zy`9wm`7ZJ&rf`7Vk(224{&;9zU7hy`D+B~Ysrt_ZZ#iKfA0HxNUp-DC?;Fc{t7$1t zuo~beSV|MPxxZgRK+0u{7lKAwMG*VB&RkKi!}|-T5b()$VQHysAY3it-Eh0?j$@e4t@3kti0IR*@1R& zZ~!|ym6DXJ7Qha~l2#MYkmt`hyuZDs@wt`#e^|{m>ezT~t*xKHYg$fx3-YgWp=_|7IP_u!U=Yb;=_pL{``}>)oQ&C^pvm*=t1Tz_l7R_dazf#FajTNktOl3f2jQwIQqFR zJ(FnDd`J0dvr(qeE?WmoN4o!HA3BU#QNrWmo5U24bvQG912qhCIzO^&e@ZhN7atHP zQl*cr_2%Q1+}1vDBI0#IO?6o5FW7guU_PRd!2i_47|DB;LK`alRvL9QJ3iO-JAg5Eg*`a7uo5<|i;!Vek=?I}p-zQ)QT3J4psDJc_$liK^(DJ#{Tgl9Ke8Bswe-0nZ;fPeEV=;GYCFx?ui)YYI*w?W%RY56m#NvrLPY`t zyW%EUV*f&YOH=^7SYm#hI`iLZf&<{nlh!Vr6c0&1%)}@QJgj&lkwrn?@pQ7q)*u3= zY&_^WaY!Q-lcCT0b%7N*k{kJSwC4L% z{yh)xI8vvWajg?uecCUGObZl=GQBSAHF>~kORX0s&$p6wtiSR3@0GtUGxp3X;P-|Y|0(_{eg~Tn zJ4SOnXTxhj_nSgErR*ETCRo>Oru8+zIWK8y=K=2xlLb_eNJo|=j)U!AlgffY^2nyW zM;8$-C9weZNxvehUZY-ri+_1`MP3UP9E{WyHYZI13#R%z?;BFVEqMpVHCv;qaiOMg zAy}*SMW5Boh&atU`%}K-S@%Z)*)1+vn|*jyTR^X&5NLYeIP9Uc$>^9nLQ?$!33LzF zf~@KQy13fal^7`OgH;Hg+Tk1QsII;^8OCb&Obi(06UL%Tuc71w@4@gXh2Xs079^)O zm`ro~mlrnh7ml-%e@U3Z!NIw>zt;qZLBfpH;C)hIOgmN>N;x;^{DSx8{wDh<|CdYx z$s(d|_tYFj<$I%euM!ZLYW|JBqMA-Ve)W%-)!Y1zb#QB`9?33v_$^l-LLKDsf zj`N_q=n*pcmp!&CZk`5Xd{p5WB-|Z%{_RNpR1j7Lp~7f^Y~9EZ^V>}+kEWulwIS5r zmX-jLPC&3fg06DJ7=3~1`v5KVTZq^G3n{%E^JailEA@y9j$#%;-cCtKHe}MZa03Jb zs!jb4U|;m~PfTgXtIrZNm0AMd9fH#73i3LXC_Jwp;dp_%4{D(~jOtG*Adl-&R53R4 zDM%yc;vKTqM)^ZbuId?$2g%mraB00;Qesksnd>S5T$6&C6=3z^~yLN*7sceCBcMXjFN7%OF|uv^j!crt2lL>W`Jw z1YSHdwanhu-kS3K|S3O{900kp4hW!hX4b$hp}%$c3|uM4l)z9e^l z2{v8-poCVq;E>KYM1tBj9OQ->0I#DiYhvN;iGdAltZ6}N1S zzNc{d{nOEl6u^g6o}WAT9wr*4IJCZkovZRVFhLF9p4l>1|I+-Gfx2 z68n2zWHn7%iHsWhY3j2{bJd?1u#*JvJFyguR2k`mgjxkZn7|r_xIQcW>#W;zxj9RV z_bX(heyq$z6LjIC%Fk_2h37%cmYblMtJIQOlV3C`B<=o9uzT|PQIZoExw3e>S|r6s zgImv@>uA%`3cUkJ)f@pT#8dmCl~ty!oytr*6W2P44!Lh~mrnv3F>Zqb@6XeMc^~>8!g&Jqp7+P9ogF#Z#h{@pFk& zv2QZPWsGA4n2>B=kSidY`E?v zi=nVsEnhVBBEz~k*?+w4P^2wBxEL`Gj!rYMF1pa6u6SmWwa+L9MM)BxeBZaMXQ)<{ zUg_@EayNxoYRCxebxSf2v@q%&tK@knzch2+$8bE>lXP5vG=AoPbv*(7t48_@A*n6sW~DE~ z0ZWe2BDecG1%iX2LQRgvK`d|5AywbyL*KM4MbH5>N22KF_oo5zzOl&25L*EQ;&{`3 z&g8fQ+kG2W6CB7nPI@hKKa=rb2sy>PqQ7fmJi~{g1=G)qgoYKs^Chw}#vq}}EiM&| zc?9>jhOcG!kNHy6elY}hZ8#ydm ze4g&*h3QptWGv7V(&F64!P6-wqg0cMPBsD1PjDk*XkqcC`O@p)ct*@ZD{Jo$#<_TH zz9HO=p{0lx5zP@qfcd@T{N(e+!Uj+`W*&mXk)T$Rf^m*=@cR4Z!5#;cFDzV%6Ts8Y zhm`uBI3n63!Lhpt(aJS@m&MUjPSX{?W%37@!XU6keK&p(jVj)_btet(oE{p4u;?0$ z|7q?1>Gf_#rx zhp|!t+tUb9L(%DT3>OM`hxc@6ts`+>wz|pc!IQ>M881TsW$j2!o7mstDe+ zpCyq=NYcLhCI+AvZ--BOXq#aQ%-rpOg|2y*Lw(>quGQbW@OV! z83c?c%|48OSC?XrjU<~c5XsI-Yj$Nog(e+TeJ9r9s-j^0h0wl*hKup-q=CL!T?R4# z_*Wti@TuMzzUL|?xPDGmtwt~KqeM|w%=~Aab0%zD;dT%+2xOK zbZ?drY}NycTA2n0B(YdsMGSd(Z$T56*EL@A&Nn~5MnJ1ZonYMAB6kzse$7f+Dk&I$ zHrg|Y?kw4samly#@fchG_8~NuRQf___gg53IPd35rsr&v%ahH|mbiFvWYT$Jtg@YY zeE#*$F%R@OXpKs}fcJd_Qe6B3nU1xKn??V?K`}>0%j&CdyvR%pUnduoDKJ`X5jECj zUSLLZ^tyvt{Bz;)$09-b8iE{_Drg#w`d3x;MrfE9RT})drOCSt;-{fU77w>pm?>k)A zg@u5%5cn@h5B{WOLPFDu)xYNeAt78XDRv+A+dEc#hJE~H7GiY=No&}IL|vg30g=pp z++IVv{``4#(?zeK|7*z9!6y%t_GjK7a+dmnD7gjD?w!x z7)NwKi6QIVkfJ)c?mQj3o!|Bmj4crba-$#91X~fDt@T%r2Y`^^oAWFLy7Q0k5Uwl zzr#kN3CNTY4HvQZfAutf5yD&igSmT^4q}uiNh&)HQyonKRW`QFbz(k$6>cQ*#%4*2 z+rCY&n;4mxdJm3y=MaO4_7%ZS(5taEh5AjT_t7tsd-HPd>5Am6Yfiyps=z2be}b`w zk*hqYYW4V|bbwt^cbmKgI)Q^}$(Rsv%=g>2A6|F^w~sd5#@IM5y!lG$?L+3)5hpN7 z*QWvyqp8TA(u zsICAh&qRt7^EFp61+M!oTBoaMisd-cBz0q~>@1}MhVFelxx5D&I>#`krx6sesqffk z^ugw)bH-ios1-UqNbKfgy(|6lY&DI}omcZbE?mZj`%Y8uix^OZ*e_G_7ZwS68kW-{ z{OHMDz(B`9ohn}_C!87Q3~2B5G(#JHDQF% zu5rE~>$Uj!P5|E@o#^y3pkS)C;karuCV}gHpB$;C`e>Qgn_twbkgk*RF)Yi+;hHY2 z!?^m_-2Yi}YAx_V&_v?Rdtbb#`5Z~6;2(o@RML|qE$;1MlS-qzCyC)Mctjzor2NeT zC!5OBH~OWpBJB@W$)tO`h7HM#<}re363oIVVI~VR1G{Z-DCF&Yw5;`sQ@3}gOMX6W zH4n{T-f8f1Hhy6|gL~f2kTN0i`UF4N%dsXiePpSwY3f7G9gkn37gAs)>aUbAjV+#u z(V@bb=>5dr&z}X)=l?KUC)rP!9~+6Jw*|HY;0MvCUXjL^gHNnV3MY zt>5#seuI5Ev%I?{nkQ&u3TcV%4#O~nYciM{x;Nvct5=o|>VEfZmeQgY+vEF#H&y!G zh}S96Ih*a5JPf5b9<}J0^k%_FAfKRo{OUk)UiL(lK0j0lDqcO`+iz5>b0Q2USAJB= zZ$0Sy4k4nza?!fWx(Hb+R~PFK*Fh|JERY_XX%Hxqk_nKGX*KX9l?-HUM%FB1wIM)E zB5r2pSJs+x>>xB^j8BE^Q)`XRk}u->p%E<|F!t|b{Z{^bzw=KxH~D-72=-&&i;H^P zP{HgD5DRCF69OU!X@oB)G(8(_deRwu6La#9MK;l!BH)4t|D|%3;U#;#8d{I|KLTmB z>ev$8?iL$d*TV$IgC!eOF`=^De+r1M>NFc3zsC*@mEXCS{9T_Uqn~EuYE&yBF)YL#McXv&#ftrYIP(%$50^sA!6%t% zNJx=?D$<#yLc*G{G^v~wOtrN35 zih{3Uyp9d!{V=9Os7SCg<#nm0)=i>VZbdcl;K&4c2g;)MvsSMlc4TF302~fhJ5lTD zhv_ASctXtT8m*dS=E#P%(P7xFy1G>7xl9~gbUBVWA*WOAq&I?};WayPYPG!<_^gAn z{B-jU_PeQq?+(*!_MLWty#=hyVbzO;Yx5e(I)${>4xxRhph6i7O7tfNNk5Ll!BvMi zaA`N$?06G3fWl?+wOS1h$J^RHbFM}dB|0U41UD)UDcoURfwMKwP~T;BGire-S@%}S zaRt1`Nj)z;!o1O7qQx+S#_YI1E7F#Fy2Ecj`;d#&QMM$|e)hV_V#6kuw_2%*gPnf_ z^(Qu$PIk4+sKIk7GTodUV*>1KLxnlc=i^qpe@g))v_>_4qY6oCqGl;Nn$%|w`naS@@`t-2gUAnDI z9hhQXabMiq22(3$91txj@x%T;QNZwpdKa2`htbGjXStGb4dU8TU>#+gQ2I;~m=@&p zr7UqRAHf#MMsFc?73|a0Pz@n-;{{DJ3y+?39^U|DN-LxJVp3%kdhe=`Yd^#amHDSb zks~P%r#ehH>J|n${X#m}M5pqr84Jm#6FG$8O}Q=)VLhBK7`dyUGi9|EQRd;Q4JZOS zj0s&aUUv5tQtR*maBTCC7Hw7pIPIbcN%YYX?8kXl~qIr^$^`xYG^rxjjBOw_LV2Dc6?`E`;7 z0QmBG-w-x4IfSsx7(#zNa>QP02{gxh9v&Fpf+G~}A122?FI1u{Zb7rUVN?}_x5ryY z>CC1QMYN}VDcdPT3OB zKsr`RM6jGP1utmUj7#)T6l`}B{DH_Pp4A(3zY^}*x^=mV3Jv4a?g$KfcBRp_OdNx< zl4U>nuX}jE65>?JHH5y1@YhXCO|01pqKQQ_yaA=`XsYrilUP4^vGaC`q_-p{60Db5 zKUgUa+BnNYbFUfs6+SogL&NuFIu=Tlp|i^BeiFQ3 z8VUn^vkgPw;HQ0iJl}4@*Cv!HhlBI`gXCRoBKxwtvC}5jF~G+X^^D|UJ)sar($(Rb z-p&bLpds{Ur>cfS&j_o`a8}?8;EmSP>wVKXZ>6vRb>x`y`YKwPW$$|4%C-49$~Ni^ zee~p(7f4tcGOKEt9=W7Gam(j+dij+bocl%`0L_6ZElCK_e@Q*%HY@x z5vw9r5ATbU!E6K221sPcG?Pb)M$L_N#cx3gNgU>;%RIIwGVsak7t)%qqhoW>M-HqY z2E6}nZl;4x4HUbe>Te<7f#_8}@9`P=2f1c>CDu=xhw?vuM=>p&oCXy8VT%;^K@uce3Ut+=zf1;<8+mLgklu8(YsOM8GXB`N*po6wVo6orN z)u!_w%563drIRCLWx(7~X?pCotC=_2WNp=Sgtb(Z&jQrow=DHqokd)(NAEMgznM)0aFf zT&+s}KJ7@bZ{MLN%z#g4uzmM9y)?=XL~R1UA)0t{0_8Ws@{ZL%S)f_z)+!tDGoP1wrz9wMkBu7Y64;khmmZ{5bsWg0&%XsfhrJ;g3HFAg- z6)d7|3;rSAL>Ac+)m=bPWLMO0yxiRW)=i~Ni~h((KAB=qnmj(aR`7`olu9fS!`<_l z&2b7sV8S+>Zsa>rQCV-58+S7@F;ewoc`ZK=9{<3y_3deet`wlbDPTVW$b(q}C6pMN zFFpEHs`iNnoK&mW?l<>&F?5Ae`IO>;I?$g#pF&jQ-FVrhf7gw}hI~d(aNa$lPep+XKl) zM80ts*>Negk_DJcJ-M{2xd&Wc^uMC76hr_@Wq@GI&1&RU5ZT zmzuHO@VT)@8(}uIy4@MGCeZAT0D(V6JS)O@!Uf-AziiXhDMuQoaI@5-iZ2qs+~>X( zpig=@-aDB%0WoF3j2Q!j)47cZRwA8R)Ok!VUPHQjMd5ew9V@u(!^;Zoj298(WzOP# z4;HJIJ+A@OwQ4?{W%8Mv+xCjBS66NbV^H{f!F@Op46_q5z&kw(QmQ;L(*DVs6pFeN z;lS_SX$Z~CND`JuXip~hI;5!vu>6GT0y`_g)j@SOy~H;dPv+6tt}~eYvMabUdxk|0 z(LdPMTrkVs1kHQ3hf%sDa2-q~RAOa{1TyN7Kb6vK8gHtU~D`DR)|OfUn$DUVWHusk0 zvhL5cF$eNdrd1z#c53IIWD!$E5DrTOWiI!K{us#>nd?SQuCJdU2cZZaM`;ew<}-}QXk_-vG=_FP}DfNx0} zqU_lB9w1V4_hsi0i|5_IZzG9AEy;2c@e9^k`j4y2E9WSEPgIB5>8GJHhrx>Id1#$D z>LhSl++w8F?5r7hqaBR2(TtP1AEu*zS<$v6<@5V7nwC?V@cCk>=Xf4N*!X#1V@`@2 zSP5qc);!iOs2y^jNm1A=yX$yv;@9nm5AQd!5~AuaEK(^~W~wUJMZ;Aa z54D=!`-EBL4xRgbMPLxfSN}VudUZ5On?DKY5f3#v;6{DbO^Tc+r-MS& zd}BwTA_@;XNTvQ0L%=y=rSBO7tbw}bUit1A{atHiwGk06hsOC)PlrIG0m)KNJ!x~~ z4^#7xl3w6RK_#B0ML3_^Q7N4K3~7yaLF^i-GRY{_UQ_+T*7AU@Cx+@1U*x!|tb7Odwl%;1L*YO)?=71~9P zwevewe$cyb>#`-%v(O_~h3shkRqJdUDwhR>$++Z=<1jNWBNLFtnF-Qi?+$`ba6O`+ z48F6Hz0)aGRXV2a3nad--3+-eO{S8Ki!=CS?MuTQ<1-4M~Mhb zeHNfRIunX|w?IVg;vyST0>>DgobId?M6|wyyBQ6a{>*EDWNmDX+Ye&KuJ$#)nhN0$ znO)B?gIjJ3gs0t;lq^^Kn21Wk9Vsrj4GtSJ+O?j}-B7PIaGSvzHOv8rq#YZAMZ|J= zGy6~D+WY5%@AC3Gc}k+u#L^i41?BS_zhc_Nzyr)0mqi1CsReWLI_sLMvL;Jz)re=~ zb+*3^xmu6_BFlNmR^dCdt#%hI12Co6VwTVLay2ke^dpbQDI0mhYn2f%nBh3M`P#Wq zV~wA}tQ&mG|KfOaCEorTV-{|*zO8zoda~TA)IuVp)nW{H2a^qY-7$n9e>&_NQkz62 z1}l$B9Po&2GB-vhZI1XKi-t_u5$j;*MR4AFDTo32H$&Qm$RjCCcW!2|sl{`rX^{InGb$1&SspiVw zN6ie)byEy?E2x!cg@GQpYIGc&J1fsu3^PWi2Ko77vUn~27xi}DsALz9Er|oSm4|c0 zKcQ=bKDX-$Tg&@)0Y}lt=a^nWpql3S?5pD}@_oO8E6d3v<;u=<)2iz3jZGW$GOs~}9pYWKH?(3> zw0X~Hu-n}Wcg5IDqJ%I43k{e^uSx)jmM-A}`N-gMRDFh;izc1~ekUAI%0gy2Prt}H z3a>o_u}&1x@HJR}*O@)eQ+yxR0zRPLtV~25{HU#|G{BxM`y$V(g}Qcq_`b8*yv2|8 z&ch`l_#AVi|1+D@juFqR9bCD$9h>Ev=*nahnc@g33LGwS|2mi3#Ka_m z_<$$Qn$?ek=8`sG;BM}1^OMqbLzt04HPX^R zc@HBOOYjyiXHx|7ye4YT#$%h))swz}*CnR%xt`d>0Nuf=73E6~;>(t5pBVM?Q?t`7 zwPood(ap9S zB$lcY&+FXi+3faCRrUCPRD2-lJdqQltFx-e*Kh&6r(hhA-dN z*laWSajFFQz^0pfmX97ZJSgekVV0FPMb)S7I_lkHRAeH4{Qln6AQ$$RUY?|cGFhYO z(A*ju96D%;e0U81974=$e195r{-ssTa=nes+4tUwUg}P*Klxf!*_XO4COI(F=8;?2A2-G>25g6Q-!dwP04#^+Hz zRku~X#3mW+C-J?@t2Ez!aAhp;vWY^DeWfWV(EsfEI!L@d;A1rNW6jd){tz#A#?xy{ z4`s4?Fn5C~>!$8C?3xLT_JFzfK#YK3=HC+)%RCP!{DEx_CYT+jTwGW9Rj?I3Nu5j+ zD~oXZu5J-j@89pw*S1&NJP<;fNRvjxUT}|L8%3q>i+BJ}>c>cLG>G56dBHs6Huu(! z0*HL`OnEOQf~o>-j=Au799YD1rID)TXV`;C^ddHimrM`lbRRyP@|%Vq+@ovu+CJlP zSG=B2@6y$J=_6mCL|j!EfXQPz3M&1>#4-KBNB$$|?yo#|cFSRk%XeEtv-@Ehg3MAW zw%W+eMwXC~ojC`Z{R+>SRx_XV1l>P$>T=WVcc6uelWukl%$ukd3s0D;d#KBl_-=&8 zkvYuL)25JfYXSG$qet>(K?zHnof!)&nC>aUzYf#ki+No(rgI8ZI*an)*V@2j$|v%R zEz_QLu`MFE%W)Civ_Z7dDedk&Cxqw;l|zvJ!#HWo851FBUw)xJj6*46W~NigmZA7y`x1MhN{@b# z5g;fZ`$Ueqv96@UB`93d(6d$De=fc_tarM6(6QSC2~64!0fvfS$~?nt`w2?r>3-_i zxWPfSteAypyqak7fWHp)kUk#Zs%iWsn&{4q*To0jik9sW=66tNPS^H_&hw4?4XQ&> zAI70CuA?Jz8muYu+NQQ$_d}jHL}%bu2eHHo;^vW+?)nh ze!#7J&buL#z>Eite&2o2eAQhcyTV=IpMptzB|SViS4@OFud3<^Q_XZTROUP!ij3Tx z^1(!1VRy6>S_xi%@V(%p)3XMf=($ZcAF)GzhB7QILkXj7nH;jC%-DO3-=+Jk#<(Vo z*gpx>t+g*g`=(*v?cPO})%MQSU&iq5n;Vt|^65I1>r`B5_25+%a+UV>+ zinj2}^%GW$0`)@e9Pz^ z_3jeL=hOuG7gI!(?2=)0+~^cG(tc~6Rd7W(9Fo-;k#W?gRST2N3GboYAY8_TufdYs zEMtyOxf)hm4(93*tJ1e(RG*>N$qL$Qp4uP~Z93^9_B~&ChN-Z8V*ucMBk*2s->1cc z7#%@b8C}C+@#QAH?gLagzHd3aQ40!ry6?lR(7efF>_*^9#O-#SCs9^5WtBbU# z1;?=$ROw1%d-|BOnV)&uUi3evP0Me%U1(#n>@i&IJzz5OGZVp%8p0S&)jaY%bJgBW z5`4X{HQiRVym6P5=<*+OL1ML@^!irzfljj1oPup;PdepFfiAuo=m-(QxAa3vT$%?3 zA4-|U!P7^vyR?V-;Hx1TmmgZ+nEw0)HLu#2MTVfKKLc^7T%}g4$ofwxpaog5N#Qi% zbV<~p%)Ei4trpx|<&){so{MMv0cWEnbaaEHTl5)n*_XP$4Z&O;j$u%B_q-1o5#dZQ zP}-uoRk>UJJUxw#)1Ekc!_p$y+YZint!H^jpJvF=_&q4b(be^Z!}V@p5)8V6o1QoF-Aqf^0e)%mB{<>n z7ljKf9px}4*|+ufey{Q$&c>%qvW z?i#~h{vc`b5gg~&7t$w2y>MCLa*D3-Di8gLivFV?)6HKNIDE(mQicmLSVu25PUvnQ zEDlvG`qgcOnDM9xS&*D}<)DQT*6 zaEcv%>c6$m9G_a_?FC@-d6`bP4j&ja0e&6$^$cBGAR!wxW5~nJDTS!bAeg-vk}~S? zA`qAFOG{;kD&YXirxCU_?0Qf9QBC=}EjWVlh?%AKOUeMB#?4y4Ya(A~? z6OUvb^Nk3{XW&R?$C%byUz_k~vcJRlj!#yL6{@JkB4r5s31wwx&wdU;fQJ_WpTr9` z1(&$tFVn80t}^H-Bghp3^4Z$iec6^+KRW%$&4J&HR9)ZMA`{=HL>fi1hT(=d(3Z=+ z$NQ&(7_BMY-;><2R#^P^`NIaZYYILa zxXkHElzhCmTpz8X{gKWUYv1Yn{lH~u>gm3rLD;_xu#c|q@*z*39nFh7yyZkBXy>9y zP|pMJkxPlCKS{7?77_a+3RP2L7SUp znwIXX@Brp#%_#}q>~Al}tL9~z5u%JZH))GgmJpi!4KDmFMWPW^sgq7wVXi~9fUn;Y zgnKeb+|$0%A6AM>?EPl*c=yyQ){Gha*Uj-d!Q7ESGCaqo1}O4)ZZKz?i>bBfPXFhd zits})Gz|ZP(cnebO(UzC!18y4J1q44+xLM|u3lXE<=rpJpVh#8^WN1}eNKC0V+cP# zzu5;10|SGLn;Z4$<%NaPKaXJ7?uR2jB@$kLByN(8FF{xBNo&a9<9d5@Ic*B$RNdZV zR2?o(6x9Q+3wg5XoXy6;idHIEmg?N@;p_LUy^ej)R|DR6SFo!B6E|WieQa0RD;kU| z^@pG|_Oz9H{y6~|t$dmPx?YA9y3k4B_##=vX7~0J-d{2@Vmj~oCJOq|sW+4b3L}|w z@tnj-3v~Ch?>oMpA+^Uu`OowLv|X^p4u#*qc+MxFrX@L2SD?HF%jVPRv**nxqVuF{ z+!A)F0YU80QN8&mk~=I(*4QKiy9F_MfYdKZv~#RWt|9MX>Va>x z*Xel$Aq%0y|5+J?j=>Zt(^73Q{!>6{MEi%%?i8!CT%FHT(f^Bh>o4AwM2?5|`6{9> zZ?h4vS@+sRfQ;@StY23uzbt%0GsE$Oa0OcYwYUCFS*M;*!`7|;ee3mTmJlyeL8UPe zcMtWK#2@IcNz==d2`yLRcH(=Os_H=|i)3eB>}@8RyG_0u>$C7g$ipyNQ~dhdw3v^L zFq~%su{|Jh14rT(5=~)FmKiXYKLG)1dxjcw-`rK=-fTYE0+LEqV}`H`VeR+T%XcL8 z@aMg5o-1;k`@=w9i*@Vl0=LBc@8a?R#FelDo!Q4*m;pA=nt3Q@)5h&N?i(wzKUcU< zV>!WVI7h5h5Uc)9*$@h;DsX#2}-*I+!#fmL?vM{RzftCfp}&la(`G%i+$|PJAfS{269gk49uM&Dam^ z3Xal=fL%Q<)5_1`l(5<0`XLFYu(&&|s}VdaqPW zhd$e=39!2%-bcUJ%{0y_Sl&mJn1d!v7d!0YsxP7aas$|Wy)8~h@R42VFW8XOD4-Y0 z;C}GAF&hj3oMVdrgV*7Wn?ki@$h3U%meF62#0WyY%|nDRpwo; zLPMQs)yHY7UpQl+bFnS0m;}E6u+6mos@~{gx|;ue5_{r_wO~K)S^8*WnD(|LWi0{*2pI4YW1nMfrVTBqgecT) zx4x^j;!tu)XD?v~8A>bQ8H4~)kR)njopuho(5>5&tMq(srU@a#fCYMD zfaedo<`sR(mbXw_FP1A!f^x-|)7{UBNE;2;c^fUP%w%J6=gJFi2T=2y6s+Ik*!j3C zuYOY@?kc+*Z#6SDXK(BgbArA38m)uRUPHTo+tI@EHAquLwzFa&3K+un_-vWu<=``* zW8)Vg_DLp;(dvL8va{igAHp{Ac6Q)1e=WMfez+SkL`~{O%(voZcj`PNPp1IvYb~SaMON zM3wlfsxl_VD<&Y`(`{c^Psop;L@B5Huv{7A=k2lt^t4Xp1dnjn^It*Zi0Qw|LJ=td ztxX<%UYcWRPy(-OVKW;GdMgl9U~YW|Lk3TYayZg?*X0<+`F8{$z;U|-&6BjcC5im$ z6Z0F*O++E>`~<2b=3#{NN4h9jx6!&}h)D&^#?p^CY4?&;H~@9FN}-;}jl)3V-S17S^nmBn?A$@a&W*gzd> z9oQbxWFs&TT5xA9MrA%_40&&>D}XL~|5ZHHZRBRMpO}5Yp{ITq7BI0F!QK27SjeZx z`C0l%l%==&QA2tkscL^w7>xI${CFC=$e>c`nwaJZlWeS)W(-IIk zj-}_r0G*qLI_YeZFgq#GbCjlPLt8bk9?;AmNxgVhd{1Vw{#YyQLFcLTpSAq1PeQ0S z`f||8aZjRLivg+m&9nYRhStKkul~+HY zxr)=?@w#Q-)qy90-v$fBb*r|@N^MNS(>qnMz`9lAgx+&ZqvO0at%kGk3%T;{e2k0^ z>h7A1l3$+Of)QNAFsPCN7(Xv8d8)h5j=aCqP$V95ysu1dsoyP#+ZI{#TT^0BxDR-1 z0=`-ogy)PQ-}KVDW36jrx&EyIydVh`Mo|KuIQyx z!c;;2sg4Zvo`?;Hrw~9}glXi_`$Iq$=kLZy)<`qc7bJfZTu-)}G9weQlFaNf_THcP z@%TW51)^`)-TJ{{@>3H;^!0Uv;>5_uoi+9rX-*bA6*l(70$u!}@Ugz5 zsSol{+$t*p&Wr)G__6O>IK7$5M|4k&Zt4vmE9sZiKQ%zn^q)8Q#i2qt_}HK4-pwh| zxm!pHy#%94$J4sg&VD%!#6Md2m1itu&`b#}`5_J2;QwPo(Qy=yTV7F)w!s?XnAIM- zug2FsBHY8bdsIq{$;u5(5TY>d=09TKYklP8d&epgoh6PRH~j0FR~(=1Ly%WSn4i3p zdAfTQ?>9?B=i#X0R#HDy$Ff@f|s{#h;hkH0DWLznxd>y{%8F}b~FB=jm_<0 zf7tu6-`^0`84`8Xgq@gVXKAU?uKk*$R4s#bq-+wl8AB0Ut|>RQm70~_14(yD$BPqrF=xG+PD0a#U)=ZnE`(2dT!id`Ye+`#cEXh&3piZ}xG z;NQa$4}evRoVF%47)Z{7-}P%p{Ha*2de5qox zXFjg>x+LS6>~#I*51S@R&wPYD<@HWue~WC6mIh1~O`D%^d#1DLGhAOXeuH$WR@&_?u$K)y6YSd-3dElh^`y zU+UReI+zP}w{J6HR$7fuA7ZkE!hb?fb*xXtn%%0!j02o=g%dvNx|yNhzN1nR=m!mk z%#lRcdPa<;RT*$_=zTz5TXfiw#1SCmTcsbe(UP^W3qFR0g zIOd<#7qAx_eaQ8eImMDNJdH}7)&w<@TfeJYb@mJ~Pj1PeKfHZ;e(3o&Z|2Iqk+ zvS5v_>>MEwk{L?kD@Yo4nQM?&Ds|L>TSk+MPNA1VIxv8bavwm4Q6L$D{O&6Z+BI1v z?l&wjxWkyTqyXdK>>*D#1Lm4wAu!eJf*#)pzMpr-(B>Qj! z`fL&%1?}m$9l{O};tk)|1V7(?Ci1<$&Zi#MN4lmjP*p5&%0-q+QK)7T&J`BJcbt*S zh`lxXr0*B6hA55rw53jiTm}zr8#S>)EhAcX2t2l6c~qC*H+7^u$0M=n1m}(wPYn_5 zhS!h(2Sv607nu{L!VA0J9)tEC-Fa_jgFF9`m&w4ZxM>1JDo-{XsWT^okgWgY;e@hF z>Q%zc65d{oL#RM7UmG1Zohy}r%o9@0%$thGYpn(J?~*b=cH%O_Egh{ASP=Gq#0%-G zj#$L?F#ePZD;~ly^j+FW8~P=tU5TjVew_afl58S4OpQ0dx5<@WTV2K_G2=!fawM$% zQ3fvnQ95!_Of0}TF0Z=`g#Ft)F@@T*VFd2`O8ttQ@ssNK0Pz`z`*v7nA0Jl34{`Gs#Y!oF)>!4e7Hy)zuQC1-!s zo_TWstpdv8AVtJRD;|y266kcB`}w@IQ)Lf~d98>gjg(_#5HNw-FKI@jLkg8%c-YOc zYIp13d{IC37Dc5pQ7Hr~DVGR%fs)2L=dczcW%otweOti}&pU6yw6NduKepI*I-55}MnQN9c1_gU2~l#jtyikP$-~Vo@As?f;AcwdDfIJ7>zb zWek0KeBYjQfH_bF4I+dAa~K5xqV(s-s`i#&EnI)tSR(aahYMt(BH%n39>PwByo+dK z-5x2y>!jmWVU2Q>9{t}FR!;V_4OP!JS4;n(Yl#y7Wztc4oTYj7O07RVA}=GodOD4D zulBj&DI{WoS3X#NwKahWeHP+*&~ZoN=pRJA{W+;R&PN$um-o4Az1Y6g^lmEF$+`C{ z^3mGs3Z44C4ShJH)t-*#`&~=%Rk176zO92hB!qi5ZaT2=iEguyEtc;E71en(R)<;E-~VqoR#NT%0gg3rS@Q-R*n>^?-4G+0 zlJT{7oAZdfXwb3Y@w$s>*$TXL!2Or2J?5FhKM~fS@Zlemq9m(c%jT=Y$4R){2%@uW zz3%W|Pm3+4aBe5419$um=LO(YHM+&`^621q4!QnnvJ7@xffcOtYwtY)1{ntAW)&*L z<_VAf(585Ckn{d;kJo8%BH;wBh6a$QTQuwJ^gB4 zg@+}YHGwyN(_D)YQ9$`IZ$$L7sKvi-iq+dLzm-q(AxX;}&2TYvD8K8*`rh+pK8Ius z;1^A|zB${T^<~E8P4XFU#O1)?=r7R^I1oJhkf3~{%l$2uUag4B(TH+ zPA01fgzuQw87F6<5~Qx06XIFZQnIvp&g%V`$A;^8oo0`7UgW29q36E?nZO{BijqT=6IX>=Isybf1} zHv|XQR@@!;#`l0sr(Hg=*;=Nrce$P>+jK05xnGNez{4m%gvJnvbsIj(4p!o^*popX zA9e{x80k~y9N|N_#u)KL{CwO<_;I~Ww6ez?ZRap5h}v>q_FEQxwuCt9Oy=4TA)%2; zC0_T`@i^k)BCQS-N_}FR{XyQ%H=77Oxh7pzVy@1NvbHf^kcN29_*mw=2oz(@V0W?q zw5fghOTwc$VWEAu5^z~Pvw?cFCm*QWWw1Zp$+-Nu_FCVYzt{DOgJJD3mg03f3n-h4 z#*XSuzWJoX3annIkG5&^HxyqPDtNT%h0LhDoU{qm_m&npPUHm~l#~x9mjI=cmyT*| zra)pA|D+Z*Ok{(4BsiZG;|nF~uV0Kq?_Dd5)EC}G>$(MA4&iInoxS4natjvc+F9TR zdEqt8p!)d{1f9{$au#n>LzDS6u_T*LV+&&PWaB|^VEuld%aPHBahXc?Fdewwk&URp z&W%zTw}rXnO#Q&$*Q8u=e|j~pl*>2}Ya1va{Gj<&?{;}rrWxSoEIVMHI$sl-uy{A*M{Evbe!G#XKRoT zNzAhF#{zjdy|M$~kI5a>9|gCg($Jx(Nm3l?wK9vCRYpi~c|HF}6_+-otN*W$3sVV* ztuh4Zx3_DVB%^5e3s*T_#l}^BQT3?Q4;t3D&i(e>`Z%Q1NE^av5TkrOmLzP}1uQC; zcH)WB4YiY%*gQJ~#ChO21Rob1>T&S~D3=R%19-x*_drM7C?cwd++uJL8{iC6Qb9XE?kAPZ+sRroo zxI2WN(dmPElhyHA4cRD>mEIYbDM)6HQTf5&^Sp)8d3#h6{^94HszxyK(na^`FBK$Z zMG+_RZDgj6=19qrosP&D0EaJ?*+?2DY($Dke^+UqbGU7wFX>EE9BAj>pgPs#jiL9U z24-sB!I(NUzVtgKQeLPEkh-hVw*DjJTC z=PAgIN;x9Uj=;eGMeL@}cfsAoMO6_2bkg?6`k;hsV#wU2iw&^Bwly^>c)IQml#i#&2!tWccQ@BT4-D9$R8-iv=Z zkFR%Y`x;Euc-B%(j6)ypqMM@}Eq6%!&nMnb{Z~QYpGpRvE^MTf%7Ma@&cz0AR5v^A zH?ix!tY;#L_xNC_f|h})8tZezXw^TVs{)I^-sA5!nmUB;>57Uop7ADD`P7k62-)7$ z(s<^eD6S+s{#*T90?r50=@ROH!P5TVzXPK`qZAH+*S2>5bo8C!`Cs^P#s1&Kre?Ez-3aIEK>&Mt#51m+{oAH0*v-Nnv-=HPyc&2l{I z;pVKY=&NiA^w4(n-?9y(Nezq7`2?$l^*iK>{2E$*Adi$MbLr1w`J{qurj=nG3;>is zi9pcTzGYZ=xN__h(O-2G+CWAq24LT=aMHiHBZf-K;+vbP>9DzNsbIn2w(*HsD{+Z-Iui=Ql46}a@ z;Lpz~hyO4CGUXoGoeE&uUoJpu{phs6+(76BaT^fs9WWYMom<3oMgp<{ygH%j`Nv_B zs57SOSa+AfrwNH^8@I zzizpIIneFjT+Q`+vu5c3O9ymT@R@8qKiHzk z{u@T3%zrL)p#0mM5?UgZ&uTlMu_4IgvHdZIL@%ghypbkLI{%qe{>;((=>4h@cU+=O z_$7}~58rQJA55;w?017E;cFP5#Yzm3Pw!iERI@#K?ynwn;YK$M2JVg5Sj;XNL;h$r z{CSnU`a@UB&6bIh^L!4K@GUf*d(5J#%Hu3N9$+Q%{3VhA-7@lq!H=6BSHx{_Hwo-( zNrHf`y-yy*zP*S9|2{-a7@9YNeu`SpY%bh8j$!IYAVV8qRC8!2y8rJ=??GA@t@`XghdYc?xi>Rcy|drKK=ux z{PgJ+oqB2hA4=JV^8YU5$SDrmRI+J4^+h^CRX1If<3)Dx?&Wii7T|e%t$zcon2&;H zLj38uAaoEbwq;WL>*wxDZURM)+Me4ymnnffXMy|I@2gT{0XC=~t~{?&7N{z=Baq74 zN{Tqdgu7`s8nl|$PzNgo#1msI47Z~0&LvMmzRXt=r)h1A@Lhv?OdKt;klJHA=e~$r zyhSP!$1|~t8q$T^WOrw6d~484A27XQ0%|Xw{fW{Fu}3Wnls>@XCWuuy=_~y0y-Dhf zezV>O-!5qgMJ?s1aT0P^&EYT}VFiTWgT=(x171pDwOWTsQ$Th{6JpRLmD=yuYY#U8 zqP*;V85uBp+Jj(&%ac0Co?Nozkq(Kb{lTJY7ZA=dSU4u~X8nW21Y-C{sv+f80$ai+ zk#wjCh?ps0yXtZCZbC*+*}3u*fwH?;BB(1P#*~ma7BmQ^0D2G1TA9#MUDkrmD61>upQ|^zPjS zc5H)K8cPTtdpNmNT6OStxKSWf&cTLDfE`(GqbBQ?WkS`&aU$~y-Mmm2?nSI8Up}+b zr8;U65fL>|0+oDVLIP%cM+XoH%<&pqUM6ta9aD~;sglL=#dnU#$#!rWG;@2L z^W+%dT@dZQf>!zX`b702t!|#W;8ek)o?I1A`cr7!&}Dtee#WByuZj$vvv8-5VIG09 zn)L~^d*A<8iNxOpo>&_g6x3x_4M4>10U+V3f$mq^Ju#E$HUIEvPun`v)6>0V@O1xR zNDIhS>dL_)Ex|$(`0Fy95a|zvU+rOV&-9x=q8V97_D5nyvi=0Xw0f8S%28T_=7sMb zBzeyEf><-tbNhv59WMZ#8b6@W@b(DVoE-Q$ywi?kpy%etuLmsfk?#?q3Od`$u}<^u zRQCQI0G-#}U_ip@;EJdvar^jKv|#W`gQ>s`rhbTr5d5kQVCD zXUShyLeOLEz!R$-2!VIEST*R)-nJ+bt_M*FaM>RqC%$?u`5l5HXC3p}Y3f)HZ^)4J zC&zLK2kwshQ`_!)p-JRLDSL6p>n`>O`oedPB~{O#UFtYSXDB*i~I>kFgV@wb{DbSrg3L-Z|>EVQxZ zRpkkvsf?q^D8+q4JH#-opeVT4B+TV!V>&8M>mfoUD1S7P4`$a?2Hp_B>mp2f*eR4e8@cmM>_r0nAk4fu>G_riZ=DIIT5kjK$tZqdzPBM0)8b)6)NI;KKx~~iSi57T*A?X2@es(uYj8@4 zkTU%2@x>^+Ji=YG#-fkLwsQhw?#dgje|=zH9qwH3JV-3C{e2#5jywV$WH=EWN1=DB zwc3$zYY~px@v8CH37xXB$Awhh5$w+SegQ<88L2DpaN6jp)cjn#0ao(tqhD>JWJllH zWWxDI11#Zu`L_ubQmD_z@AO9fK(kF$Dadu@GV4gYd>xpK*Zz3+7TQ7`43*BVv7LKo zd&3_w89Ll$@Nd6NZ2T7`ahf%I-x+?BoM@DD2_8yNqHQ%!T~ukdU?EGCo*Wh#vVx$0cNCrO>y=0eivzBKbW8vht(%^4rm*N zwm&?@5Gz9PLaADe#!e-rk%Y~PYRpDH3&@&4%HV}#D1`>Q$bM4Ec$;au_uSORTozj~ zUOCr@aXDd}UP-;c>2Nbs`}lod>u>OHh%MibEBUfkO-5ji96d;K>-;|2&k;{{Z8_)j zTH(Hc(MPE6^y{+{%xac>z&$=eE-7f%!U6eU8qd{wR@`*(6mOFY2a`%M=tj;#E|cf| zsi;11x(-u^vlw7~z5H`ydydCPawDE};abF@-u-K^X5>eUCXg#T!c&Ui$;t5RScpN@&L)svp;o5}bC$yR^8jr(Csl06^R0xbl%nWfe zm@LIP9FKzndRJ#hOM_Qi(<+BadfX4+WMD|?Z@gm!#2O)}(>gC4gx<9Up|?8_a#b3} zM(f^u*xY|zj%OP!5J)@Voq@s=!P2tJH|l6Nr&iuknAcD_gm*{B{h2zY+mVnJ8OFOw zG)UOFOm~a+oQ!Q`^P4Ib$y?Iw88qu)#^3sIPrH9Qa)oSHnn{kh?fXxmsN-rA_;P?I ze7xlwr;o=o)BP6elTO<-Q3Tgkct9({c`v&Ij!?dmy1?nKf*k$_m-1Lh}niE zgXk#F)-C+dVR_pFryiC1NA)DLqGt>e@_;XKHcHE3lX#w<^b;-Z=k5U*)^Xpr5>b;x z@A9m|n<*G*O~4-=5&4>2yT|lX&IR~t&iQbdMR0lwwAsHB?gjZ;o^Hv@naD#@W7>T(6uP4Kp(Kc-~QTnD1{KPZ*&?}ageG${7H@ZQq<(&vJ=M*J8Y?LwJU1Hmds z>@oiy7W}yX9%1;kei0CAL9|x@Ieuz}Q0IG={?tLpJdps!18v^fL_N+C1Nl5z=g$Pb z{rdTdXtj;FY>m+GzU~p_5tpHVxakeoORrqKTKpGp*$0mrJO#fIO6 z#Pn$h!S0B`p$r2(zo4}xIn#Mkrl_`y4e!}kGr|6lQ<%!|%UIAYHCge`KyQbNmHUPH z4!Bdnqo1d(H|z6!N9%bJ^BtB)#evPxw;f83uLH?H-oZs74pY8`jlB2n^735+{1_E< zmp*Md6P%bCXh?O^mtCs0Kz6O8jrGEJcEChx`qB7|vb}f;BcQtYu?-moU$6$IfSrec zl(kl|JC;#$s>}qV{}Wr^kqwz&-xJ|rX1L%W#Y3!4x(y1+0rnZxE#CcQG~ zN4pPFnZ(>jRU*)(EtaDXH(XYlL)-`x42q;;zV89D*ebcV%Hv|7atGaaE#n|HG6MmR z1&mcN{|M(7InaPZ(S83kreq1)@w8I$Y3LrO{c6}UK$Y+XPn|ETZNj))xj@Mfpc=q6 zw)kbS$@G3YN8Kfj%RzqcHRi3$kyl~MF7(2>XWl)LxzYOPyCoEI z*je*Ei4_PaD3@m36Q{}*b;`DETFQ=yg@*V-Uq5uT)fX!7)T-HB^84!&!D))lTQ|_4CH+)R2#F9)seO*^uCG+i@X?-* z6VP^6vAm7(;hhN<6n}7}FN1i{>wTP$%X!u+bZ3^g_6CUp`=kxyJ7@e0G^_y{KlmKn zsvgtfxC}itfNJHqGZJ$YZ6Y%ogXImku3o|;G?dxxTYd9(tM50eIktrs?CgYQTpfm55MjdYfN6yQx2tv${`ZwFLVJLqCt>>w@HxKx|%1-M7Zh+IYOqEUyGtH z0wnJ`*>5W}bO4xI#OrMu;F~8JF%k%`71)M(B%cE{@=vpIqKSyreIUu+=xr6I zV$C(wTyFSUAPpMeu$u~bum8a2b0IR{lD+Gv=1z=1+ zpvOxL<}Fe%+|Kg_YM>T_o+r}m?kH$8U(T93U+V|baPuMoO~d$?F$T<* zlugxGF^z(z(`u*(CqQ^JJVnkoJ;wDOp2xtb9#|9gp6~TXdf&w~#xZ7xM`jA_l=~M@ z1wi9jl3W@V97|tTE6}yeZ8P{p^;@}B*Pk~>LwRm_R$Do&N_cysDP@y?$VH}3vsLAo zF-(E>UZ!~*6iCUcQMSf_s<(tGJ$0jMDGJx@x=Wf6Gl%IIx)NJK*qOM(dHB__MoWaq zqWX_jJ18P&HyEC{8>E&PLc3ZKRNqFY3)qZxmV+d5Vw#Nhrw<)XAjL%#LB%ZaR_sh5=?j=$=wSu(uWg8lyfKCP6LPM|lxV8n2W6p> z6X7Mgz1n*FyDll6ns<)7+mwl?E769{Y&)uAjh&OM4D{fjX_-JK=Zktf&E}Ijji1g; z^d**fC`+a4l-bisr*ZTd{G^@X+CI~JOj6x}g{rS>shg^^j7Y~Fh(?E!7VJtN%j$1h zeB@~vSQ4eD*tWyVDVt5nMzEDnpLL`kzj|Ud7JYz){twzWxXbNTiGdis?rSZrB4i%l z@3JFsBH59H{KKZpAYBd4NyqNe2v|Z^7udAY5 zM0_&q*JC}c$fy&!FVG?86?Z#k5!n`C=08 z#uhpZYH&C#XK3iZV|U#lpql^vlZfER(D8K4wQijQ>Zrx!eTl5}5?Mb|KgN1D-Z&eWS@iQCx3P4`=U&@@8Qhh2r?RNf*DeN)X=u@K+=? zSxBB^$YM{O5`CB~5f`LOcFP%_oYSj!^^raiB@xo_s!WSos>Se}06GmNGj^EVEGXwp zzuWRX;RtoJGrr#9l_L;L6w1_5sxqOa<02k8jFi!Ld07x?_6Egl-DamD^J2A8a_6~G zS&m;tO z(^Ymntz4<|Ig(qm1hhw~NL6t5O3)#9w#uIawZlmL3o1p|beWt?-I-K}zRZZbojgK8 zCuHhkf1J2KeF!>%iz1;Dof}qb);vc{cYbQG@cBZXAWkQ@<~>4L`?ehoYKwA5{Q>># zP_Yi{W-Go(!8|X_A1ry^!F$3*SVtZffy{N5FMr}42xC~a+C;bJ(ecpX;s&Qx2odM!w ze$QyhXJt8>`qyvURdh`CGBV1Wj{HVr@`u0sh#8gl-wYUND3B^N;kcG4)$;5dwOq(f ze17C|!im(dHyi9sfYzr^7ruJx%IP0`NZk_0S4Y&ZbRlpx`F2=_w>*pTw)1;Nh%9o4>7pL*2EOEpepm zQ6ex~5SXgR3iY+ITmH^e`VUMEelVpYF+AX{oEiv4DeC5Q?dvODNY&N-ht3at?Kfvz z`6%H|Lk)t*UG&sM(7jPt{Dh$XPgs>JhdKBf4o+6p2ZN8A(tLpEu}Vxm&5fk{F`9po zrGkBjoyJ|+Dl@{2R#lAT_R!$+zF`_Ry1CD#`ugg_`qlz)( zekmXHbY%U~44}e2b-m5fAO1--_hr%(>yq2U@nS==J?f+Q+!tGUxVd6A8S^4>8`CvS zJJC&ytnnKLdbk&{>-$W6L4QFK>X&GNeJL0UF)-d`MB}oAGPe?dJ z3eHG|-yO2DOhJqB$>&UQh~(mXv^tqNm5FF1a-Y1QMO>nT%$WP__qr6 zh-`yJWpistdfGx>b;E{;F1{c8rMe9cxIsT%wMwc~OKb4{?11FVR5Kq2Kh)85Ldd>< zy)E4ofG1c9T`L&Q1#*AZit_u!2RqWiSkfjI;+^yEj^>w+Eb~h$mSR(w2y>_VVA3y~ zV4^Mjy$U6p>LyLwJn3^n-735zk4uRfJ7#O|mC`APz#wd0xO5~c9G29*x3xSmO%j|R6N zzU9?XRo?~lG}Vi097Y@VFIx#(Y0*)afC&AOk28lKO;Pxwvn=q`^9~f>sAR{7oEEk+piji( zhc$6>q|N5S(UF!ymJp57s_sTBog`OsT)*1p7Cp2Q(6PJKPsr6ciL`aqZm8|}T;N=` z$^j@|^Fza2ID{wulQmB6)Cq0&7(C z#CFt)Iyfv!f4QC}hR5hG`TqK?3l( z4quruA<5~5GoNND0l#aWJX-sGJ}7?BkSJ2cq_^e- zX}eA|Lt2C=0GddcL%hBhTfXvp!85(C7)D~5qab_7r!1A8A}aY1WM~L+CD7%XDcKkt z>|pPF**T4=o-daUdGh!b05@NAP$8#`Y`aOH_FH zOMY_?6@CL=)>7yXa72K(ZO#Z{ins@0%L1GIKrtqp_9+9&AO(1IeAg|a?dr*s%;$db zaq;hXgN%!58&;B)}VKcdwY2^d|N5 zSN6x+V~eN4K$ZFr!ihAJ0OtY}^E(Rjqk7}Xv99K0LiTPeyrkuFq4LY6iBnMWsq>_> zLoBtXT+@iCd7(c&PtIoP z+x_)XS2h91;c!Iy0ElIuOfoq-*QFN`i(V(tIRw;TyV3PqM+I&jVg`NywIc86ji@$C zkwvs}6s{~LWt_iXb7X9KPMw&gRsC_{mo~7~fx6KdrffENt($$uZnMZEfE%5!niqk^ zs1`iEgR;SD9&v7}j7#87I9iIiZk&y9QlUSVP2jgu$YKXJAX$En#}Vm!qmqpUjUYP3 z4?p_6_(}jGAzlVVYoiE+rxvL2zF8{%nQ3Do3yl>8mP0!*$CnZ9@6bK2K2Rs=pp>b` z3tF;)5ai*x#|=M~7}%ziFzM@GnzRU}#ZXTe;O$Xhrko+|ik5&Jo@V@iw4EpB+}j77 z(mq&>^*yE9CAptOyXYw`en`@%I&^cPH1rv*%JaWK)@`Cu@$%jGpZX;K{j5v#I7|KI z5k8B$!p#Jt@k<#(qH66_L1y}V7JVrarf8C#@a8Vj8|K3{)HOKhc4(CLlJTh{b1iD z67)ux1Z<8zf+`(Ni*o!_T0Btz8ofBhsw$P;ip^@pp8$@kcg~(tI3@?~^xw+T{d*fE z^!Z}6`x^5!J-10QfR8*MCqs3M-0GZVNYm6DJ~=d*YJ?YcsOCMT7MB(NoKGBoAlP3 z(eL@n`^R73>fc6E=J=zgCs$vOT{iho-#IUts`f1ybOqQIm0!hSJ0{8J`4xJtDMV@4 zN9Sph0&y$bMUKjX!Y26Acr-go* zo?$0UM26T>My@<-ds5$$Jy^KqxK$)0XEy;vHQULPU50G3Rt1K~g(bnV9rs^T{bw&~ zs%}8t4P|;A_r|TxCtQk#2M8@4(dfSqr|ehhAO?PNKf#<=t>@ljv6rJ>)78efRB1#D zIS&s73*rLFv4!$^F$@BGqiZl_TFs8e&BM_Rk|g-qFB*Rr5Rlzw)7=kWkdP~XG@tXh z4XXy1WwWso^YC9w+UGHE?9AOrbuth2BEU3x!aZ8IrY73;Q zlz3+Zm%DROn&jwQr~<3%yr!7QXSJg^@MToE+M;on471Jda`Qs0ld;LTwJ!G*`dv#f zaPy_I+;2V}d4g9BTY56-Dz2S~UDq=N-8b&Jx{9c*N-iFYe+a?3oJ_1=rMX8@7c%3V zGwVvan{D3nb5#y1A$Wr`QlKY{ng!%`IyRg(qOcdoyr)a8xkLK~TQ+V1JD@MMDFyjk z0XIMVSLdBp?gIv|=OPz(OrH(Z-0*4Cx!Rnh;lWm0s`+EFAJr71T54SdE$t2>dV*yh z;xMVtmL7?}#M6Dfg$4P1G|0{x41>j0A}RTR0?cdq(G+4`W=`?Px~CP>R& z-1AFu(;MCwN$%y?js8rj^+?FDap4W|;ODOW(x3E^ofGB8N8`*tADBT)*uw({4T?N$ z9jKu9cS~;L^ub|s`RlZQBGSCA-3rmnGi-^u>8|bDt7o}oOCY&>7%mCufqgG=`lR^1 zk6{4&SxL6x;ia)r#4FWp-{N{TMndn-^ z+St5B-#VGIe!>3JUmMd8#Kyo8RW4UXl7NC&RNDgl3z@){`GMd8f`|6miUdPr0dB$P zLCZ;pX(1mhZpOFFu})XnE;8WCYmb*Q$+cC|m7G+xotb@bIbtCPfD{(0=s zpY$o=lcMFy?Rqa1**gJ|!pdlxt6nqy+E_SF( zjcBK~I=^u0HrR+2-D1r*doCNhEV~jsh`jA7izCegD>vuMgk_nvk#HH|J0~iY*YJ++ z#tJV><8;#{w)3vJa{1PXY$cxA{cP`9Wc-3cxmnJGH3L7|y*R#BN9KJ`` zI42JFU%{3uS)J8*I=WMldaS_Ko&+{Gc=B}YWjB@hl~2Pt`kZ=ji96r8{;Eku&T@St zOz7SU>E_b#?AvxC*3Z@s`er(YCy)SY^!p-8QfpTL$K{q!6&%+8kp<_^&aLd#R?cB< z7RCg4Fhd`UUD#0< zKkSm!n(NiCA4~i7s4Ij&`7lU_aaX-KIPrKAa52naSMvU)jr!KIS0>5;gIF*s0{BB0T0(9sXIuX)5`XEcQ4%1<$?;;O}RLe<&0sh(WyONZ~ar}fceo^ zWBs!MO5X#wD1+)Oe+bFew+E_KKi2{y41i(bg5{F;xo4jv+dcsV*f5Zgr{tTvHE;N|~l@A!E&bav=R}u0%o`#twUGVhQlk>g6;K z-Qj!=|L`GePp(G?g?F^gs8N<-+S%brYoINCb+*#^g9`u>rQIB^4~Ob9>5>lQIaGY| z36~ookGr%7Y2Q0+Q|Ee}h>mA?KKcokdJGR~RDZ;#$8T)1b-J*0mZ;#&SQytDm9I;g ziL2jKM9ykRJIR9MZMLMBrE<&DfxI=ou{aXCvk*JV-C?((IU+2xs2%qKO@3sQ0zuUZ zJ7yAs-~^Xi%`nzV&HY#Cu-F(z-NN2(0`Wdbs;cVWsBI5y{ zEsidefVx2nkl4~@HQj=@XMtrdpTQt1L*^i4J$B}DS-C^TblL6X_fIlWyu~*5N>2Ua zH9IouY{_zRK#~7t+Je_5Vk4icd`OJ7!h1m0R=G2s*h6DMc~sG30H3*21g!&_U(w!F zqYgGf(}~PkfGD0oHFu`I&6Y|Vp*^)cy)SS-$n#0&t}2Z|d)TZdzA_8V-{EwvrME+J z0=|CQgkJ?Aa>mXPrWTrb65y?iPwj1;Zk6^udzfp@n{zXd6siukm( z=ALhei-wnTwRZjcjt|hRjN2sm@eT=aX@?!2K-qBrZ~^vlGHedKETnZ^C2CgC9ht3c zpv_>U97o8M&MJA8v>XMT0Cdo#@t6owaUeg~0KHBZuP`kavua_qSvvR|B6mN?t59PU zj7b3ReN%bkU&&6=4OV)gS{*;zDF24K`!AT}9t9?h8 z%^Lis6;}oir_z^?f2+^+S9do{e`Jt^u>exia+b%Rka!D260<>e$4l59he_DMEy!h+ zNSNJjSgm!QsmGl*Q+humLk^a1%3>_L`I=6KJfR6=>lzn{Yqe1?6iG6vnIY);hr$h? z_;4x}?RdD#_X;WBdwD_$l*>$!^UR~wFqULzj+zPi0z_>&d#2gT<4vX>Znk;BwTY}c zD!582RBU?l+NU}Uk0@~@)F1V`(URjh?Am+<^bHYa6b&vS842brM3R+{t&Z<1T+%qM z{as^rB5!nssgV?uzGb+E)|u?tb1|ll&JHJ}3b_a&c_oYZ+QSe|q@B8Q4v!$4O$Cy1 z&?bal4QDthWTghG`x*D-YS)hwCsiH_&lmYyAP5bHL|cxWM|2g-3_rBd1tm$5J%~e&iWYj3f zC}MY~Dx*oVy5pt6Jq+FPjoIQIf@<1F*Sw+Pl?NM6J{Hj1l1#ey@;BO z2d-gDE4bo*@h;YBQKQVjOCN1b?y7(^4G0A9K>9kpt>*$62enCT=hBPYJ#z^wfm2O* zA2N7~c7BRI;czsz15NDF3pOj|Geat$1srL2J%8GifBG`DZXwg#qe+9IBh`!@ zsGW+dZmkC*&Cn^7%PlIfvP_=DW*)-4-QdpIT~-9o^|iPTk{kX~bN-lnH&+`jP@|`i z;^v^7#4X^ryUX>aU9L@w$~qNuKjJ=#-22E~-)=JwZ|~qb+X9#ypSzYg8RtU_pZJ!8 zf48n$EA#3HWtQWiPF^yJ{{oZyBQ}|Hq11Y)G+bsi9_$2Kj`;x!jduIjGU1p^`2jivMXI=d zFhmdqyg@Z!{QLr|_uFpL*(NBt)U=$(%S{)A6v#4Zp;S;`#wTRG7`&rRWoDp1sxmhs z@_8-!h^(_Js4{2rCgq-`y=M@T+Q$~Wt>q^hTs4f8Xn?K>04p8#>pQ|=7%sg?mDg2UACBPJ2Yd((fEw{b@w#V@`ScueIUOJQXkSm#H zy70c$if!wuUK@Bl$06wi?}VjQpf+}{J-%p^XI%S`coM&v>Vn;`Eyp3Z*2|4PGI)3g zaQ|}QmI{be@sZ2GQt4w%f4@mu%_Z8i90~Dd zw_o$?4}+JP z=Eb1Wn5XC^pjHq+YhlLr+&7syEj@kM5WZM%I=onyr%cF-kQ|LAt@S{4cK~6?Y zQDfq{axu(wwcw|Yf%zi}Q&#+081vmuhere0kBhN{_5v0s;=-whnTz!CXci`8bwYaQ z#*E%=+U^MkFxF}aNS!1@UviPMfnaw&u(Or!-UF6o8n-_de0-I43XHomjh;X`G%7uG z_Pt8EB*+#nx08YY^-%EVBgyqLcva3yM5Ahw8n^Sj;F!=QfE8A~KJDZ}><`hEP&lyg zU~4u^gOQ%^RKNJd!6+>a!L0umH25fDPqD!qMk2$??ROb>J*r6um|0tUkH=LC|weC&tzur|$KhuiUWYTUkhlkT_GM{($t^%3Gp0!6MiFfRdYs6_Boe}5X&b$|@j5wZ2s zISYI@p50c@y|r8nyNvZj;!%9FW;;s1cBKTrdA0X#c=8)*ShP+DWZtD(-~zTQ{$NDj!wggYUeF)`>DOJx+0ZMcyOHg*bgI&@6u`=*J;I= z%USO!i6m$QIE-`cM%65Vs~rPs23T`f80xGRlGz!_lSQ5;Zip#;->!}iZ&VHDz9(-v zC}WwcH2!K@4;4sC*Hh{}M4{?vBF8vgY74VkB$8fqPVZz>bAm%f_B;W^QmcPW#fnZ| z;=%6>({M(kDs;+`P8i{)N7@vf8Q@83>`zOhHeqm*#qctlPX&c}Ww{n#I~=ow2Eos+ zd;RofQE-zbHZg_QL*!_-UJISShQJs1!xLBD;rozL-kCe-&P@rItR8oC8053t10K$t z@JCk%Ff%M&c9RyW9wfUR{iIA-$7yZ0xFqA)32NP?j9Qx^FzAPoA7!Pth0;n?@pKYN zRcN+_>Ir04O$>(7I2t_5U3Hbp_&DnP-UGWrF=-B*R_Gfi40%!e+59b@>q*Cgt>`%PsYV zT#TJM^SZ!MDkN+Yb5V>E%3U1g{%cCl$v^G?IO(bS$$ zGlkp7`i{$oZSo@2k#E&x_ZCp$h4AvNMr`+Cm|a_YnWYZB3|TTT*v%d?`z|1IVV?m8 zw7Y?+95t}@B^G&&3#@UmFCQNZ4U3`E(zZpNY=~x zx3+rbZ0E6B>flzBn~g(@G^td8bJ9LMGGkH8F=U7>%0Jp89oqFSZ(=^71o!#825gIA z#eEcOjbCE}-O1$bUKRJb1iMT&ey)@DSu6T;zL%{l`J$(?rn_7e7)19 zzNZ6oPBjq*A9Tqj{0;@|xvqlt5}6b7@=4`AY%w$|tbV>p#Jfnal>^cb$ntAyaR#xl zp&gq^Yx+&_dBl@BL!$VoLotme;z?F3gZ=`Y0%iJAM79-E(a6eoh~QC^qlpg1eq9?E z3LXx7lKU-q^*Xst%>rs~nl#ZAyO{vq&u9sTjasH$PNWuj^_fEJ9u%R9pZk`x58Khy z;D0!0L0}~iU-X%$*+G~xdj3H~;Q8<$uA4hZ4jW}CE@l^1ePbbpK_c%)s|DoDj0|*m z<#j6(JyuZT)xvxfKl>@&*C12}Rhfe$0xqI}c7TECt@5w8V7^+u*N(6T^DSh~t zixcbBTHdMb@x5cpg)xy+hhIL^${ep$FEoD@Kj>L@CtxV) zhad;c!&V_Z7WZT=Z=?om5c6c%O z^V@2`SamBkqPbF6t^mh?O39aH5Z`3mtj*B!+Q?Ho2IErgyTVmA-XMu6h;LJ=-vc(= zCeWfVBsd#nu)YK0vfC`sw&-y(gcXFhttHkn@|ala-HgNy;WI^+wb$gRZGL%}YGFCz zy3wEO=2P8mBZy#8_$9|go;mx9rLn1kzq0vp+F<~{HeP0&?NUIKDs;J=9RYM^#)8@; zf+!<+Jz{|xK*iqL2G3Di2049&-Xy0gmYt@#dVYQmT@F#P?U`XPfC{fNOX##^Nw(ag zJ1w(=nCTj`p_yEFyMVpKvJD&kGsZqzt#?M&~zXaI{X8v zCKQUDk12Pw!8Vp_x=oEk*0x{U&e6Q`# zpSvoAxGcuN7^8gC+6>dlT0UaX!$_g6485wU+8x-Kf64^VK<`tl-?MZ#UBOq1X+_v(YJ;DdA{j_YLC90n?-NGC^DwVv zIViS=;W+CgQB)@VWRhNzftozu&5PmM#-tSzE&{W}VNfYT=KFIKm-p(NskZyUueRW0 z6r0SK>NPF*v-dFRY9Y-!z~ePwdP*55Khn8j*jeib;;F)cP5Dn%ACqoT?I=Y)NEkqA z=u_-+yFyZ+)kqoC7-#n~$5eq5PtV&R2aBt=^oGVXqMr@8+NK8^R+9;~GnSIubL7W2 z&&2h2pvuwVFojp*YyzgP4OW~I?S5e&1%$+ZgP0-czaLHju~kZ6TQQ3}Br~l@f|UsT z^+%eq&gD9))bl1rJvZ|6LDjgIzy(qiVwf`9AIsEZ^iLUf_r3Ny_J{T} zj{d4w?eymaHpOht-RS#jQ5PoF&I~@(qMZ}|6G#ZF&1BI#=C$==U>C99iHb<6vw$CtN|%>S)4-U zVJV$X_grFCgkH3fnMtkD)`jhN7Gf(#Dt6xczRo{AsW@B%i}urqHv%2zW7*zaEk2 zPU=@20%p{mchycjj!#*{oVFqgcpz%Scu5~UDWdOEu(pNqOp#FQ%vY4tI`)(j@kfcr zO5;w8UuXc&r?a1)B{zOW;*?;rN9N{dB>O)Rj*eE^*@}RUntCTqMqSWZd3L2n7-Plm zfG339j{D{f$}%SSC4MgUrZaSkv>I{Whm$W9TB%|k&HCq#>S{Uo+vX{D`K|%a*?qwmupF?9=bd-tfF)1L+wXTs6zhdVE54q&Zp znr}%G$2V=Ol<^;4P+oRcBIT<)m1o_=v7EL_q_+fC0rlJN)d&mWl8DWl$%8fJbMNt+ zBD!kLY1OYVhsxW-3{6k2r^?kc+APXHm_~0FC@5!IED{i5UGUx)D9q+qwGJ6CvtH~V zaeZOU{PL*ti|df>3(|o|@0aPqRcBfigHLJ-5P+a4A;~N_ePfj0%1pi;aD`pJQQwY_ z{j87EszrwF1jf2pK7W4$EK@9<4#xNO-mZ@l{Iy)htAulHa8#>8M%b2&s?;=o9YEJS zNa+||)u3wLh`(UIlIt+loXt7`$*joTWY8b2bKm-KZF-jp#CU#y=qx`hhVqE(%mnB25eiOnG$36%urxmqp z?|u6s?rLjNZSiacaF-vlUB6D}2{CBDAa@HhEQM2sqaM>IlFi4OdlGUty9FNMC|4VI zv0m$(vwr_&{()2swG^ZD%E^pE{+4U|pTmzFTQ;MK+N5IU*WU$C+dWg4Q)a8&hRK}p z+XfW;o8REtJr4cy0g!~mf4bnykrr1=$QsDh>m*VH#Pv0t0&(G;^d%EI6Str4Rk)iR zBE4D54Th5uEmxh>VP26*H7mj1>MUl~j9~QM{>)b(r|tcLC0LluP*s_NI5qMxw$A3& zZ_{OO0M*q-!)6NIRb*VVzVm@)gwH{DitcV0`v zG~)X%K1Il!u(NX`HfmgEE3Ld?80(x&KS`E0UKuktbP60Owh&dz`N*_$nPr+)o~?YS zna3zR{J{V%C`2W%o7~y=_;VSnLegUS>%FW9T}t-)}(uweY`91H@zfI7dn?R4jHdf4=)py>W|FxTP=a zgM0K2P4B26My2bKOY)X|-JU?H!N~WAHa9c{P9;(s#)C%|v(zft1;;lMGb8!-gjwI7LACs!5Ki)`% zBr$lA2*My6-zN4Cso%5VH!CyMT99@gzjb{~UaNoJ=ZVu>%<_T?gZTV86F8W=Xv66| zDI;lII0=@U>+AeJuiu7BKoqx4m{dG=Z6a4f?*c;Xk!80%pm*_NliQt+7J#>>>lJsE z2E7xRf-rfKNu7{Z&Mtgvl`>o=*Q zP?DC1hZXMDf$>KX_KOpq^!s+7h7pp{NM8c+*eqLg1JkGdw-Gum!DISzc)dZfVg|ql zbm_noDQP?oy}27_*UPDlPo=@^XEBjPp6GTMG~(;2{eec!+aaij_HhQi(!f2C7Cs;R zT3DvyCKoB8(hlwxxRT;of`ISKNJovf$XpE=`_;P4!<1N`PA&G`FqjNEdtyabdJ=l5 z1(R3V@y7!ua^9@Cspiv3VGRUzxiu2^ZrQC^?A}wU=!ve-|9r~Mz?x})wB={BT$o@F zH6L`uUTxN%F{D>Q+x_O)Go!if!`Je zg*X_E8U&#{>D;L{>=%H{3F_Q|oX6c>_PZl2BVsSzxH^CID)3ikD?>uwm5CjL2WK#p zKe*JSV-yQjaQa8J!d+p{>nFr#8|KEnDyO%2Z6QJ9wS3`r6^LhWHwIkc8Y)jxO;`o+L|OL)B5Os_l5LS(76^l+(K?N?W6 zvFR5Wr>~$}vJ~?51qcvSvvBH+4)Hc_<=;X>WAFZJT5wDba>jp zP!dxja}{!_`E+K0nX08rSV>sZxM-(7VH^8IC#0E5D3n68fgVEGT@Hkm&zH7J{Rn~0 zExNk8V(qT1_%Bo%h?wcUJR$nu-voq+mp@*IKl}3zN_ENO%;jfx{+NGQp~nG+A@=|{ z(+}}Pz`WEA*jHi1Qa$`bSgVmGmG+PucFd7@X@=2h{Xi;iyElj>t*I$|r5o8B#@w^rD4}(Zt4p*JIj?R{tHg{P$%;B1c-NV`o$K z@di8Z%0{e~s8MtmH{e|whXt-!GB1XIF>C(K*0iP8GgqmBrz)STEB<|@P453%%L^>^Z-n{3y&WRvUw*6n%Lm_Hw{UbUjg*?sSLIF@D%JC` zUx99IvWT{@2f?+<9>ebx7WWR3S)WqM<3>W-y)B9>9!s%Wj`PmMbeQn8y(6+}=dbH? zuGSovrm!=v-;TRJ4%=7_)#n=>bd-QUu>DRdL9WvV{BLn>Jomdx+@9Y)l9$Nz^30yz z^cRqW2xpxu*E0-)%dzx0@90JaL!W<9=T5+^JKp7VUT17f_Yky!){#Q%wuYHpXwvvv4~e{6{L{SMzTLui&fj{;9aJ7H4wzyU&O#G@mEZ`6aI8pzwAe#-&6mqrHHrK z2{eG+N2x16wmUvi<&m;6p8F_X`f*le0{<&91RBEzrPc@q@DULWd9h%~{B` z<2>NDM@qWM#toc~4lv%^A3o%Ci|egJW_#=zZ*}WDx)1@bHiat=h^Vb6yj%pVIj#*JA2h-Z$oD)omRLzoRLzB2bpC;US6x6 z>Wgn@IcPrfbTqPF?d*zTN&BI$i@^vfxM3q{O(=oJgU840p9ve5=u^3X6$-YWvjw7w2tq!ds(Zw zX>wYBW~&S==2V5qB%esgkg$(!j_*n1aL^6@yhrs~k zH{9IN;ODvfSM~L0R$ARxO)9XMjJF`Zf!@W8X1iCKfIl2og5^iIGCte2&zsSVe-8T> zg*+(~Q*R+@L_G&QPC^}W>Nkc}9?$b$rz@MTV3X6K4diTjdSSj*&*&ODk`Lf4^)_0+ z{oXfRgpC#<(&@u$(=b)j}x7ZrMrt%wxnZj zr;TEj{Js`z$aX{)4cU&1FHk0lV+UwAWl6F&|@XxbzIoiXw0Auh4I-AXEuVq5=m>Yqf*NxkhHB)p?^N1P( zW_m z7Q=ZYQ)Lo8bQNd*{$Drn;D7A2U$&T5I-KiKn8f<8j7sktJ@{9Ujx;b2M`0v+?_C?w zw!9DzV1{>TWGp@{Y0otx;=qU_5ym}#A{R^q5XnukOhaD$?x2?nhO)#7QyU$W7{p|D zpmjE=U|U^F3EAV(B~5s+k}$aPVY@r1cXH5j;f*fqHXx=&u1C%`+D&p<0rxpS?-07@ zJ7xA}4R#yS?a86I7?HLnJLxuo@!;7ARkO%XXRPyTZN8LGtMKW2@4HlazS9CV+OL8j zi6A<(p z!WU}hy#hX?w_Ssz42=d0=S{d&p$4lp*s8~o?0y#2$rGj{q}8708QDi*ap3Z<2pfustfr6g|<@J*W@p&mQXp7dVKeu>Ha*JK#oi;S(}9#O3SDA<7oBYhb+VIL;2QDOt%cY39z>QK(~3 zBId#4QG!Hl0h3 zzLeN{^*u8`=OUJ&EGF%XBV4ge<>i_-ZJ`_Di8hY;I?h+285~CiFAPE(VlnGtIKrZ_ z!Xb&Jqj`|oNqwoY(!zmdT3z&{@$e4IR>^9__eQ>&P?erR$wF*@F)J5}A|K!{hKd$J z1rWY++be;zPsIKM3Atm=g}^HBt_yGiN1X{>KK>gjUY%{tZ-mq)lkC1?gPsyTKz1is zCWRaiGdN>-EtPkp$EkX&eC`dGijAa|RA#h!Rs@1~bleQ#H1tTVF3irynw!yo-0lr~ ziC7A`Ia$7_D2A7g;}=7{!Bi=s6&3uf5D0%Y%FzJ+U<(N0%ZE)$f5<8k{ zS7_MxN2WYFJav|Qvx^xAe0m?g`ZAMb3*J;zR#=lKAeIo75+a zP8F5Ql7zIjeX$|5WRgSpQ_ci|UM~O$QS<;w_PL7dIDHTa8qlB+V%93sHHZAq9^ejUv zQ62Z!6sMkWt!|T$`hKE-IXs{FdXqa}j#yJ-<(GwCNoV2QKk>t6U(n&_(~pnUn^;wm)$uszbY2_$p2 zs=+AUqgniRcq&;RFhBlml~6fjXl_&${&wlwpRUP4kATQPCv78IefOMPO3qBLU19Xz ze-o2{z5*9>@j|ejmgb(L)gMska9i@=FPwSLa<`@`0QSML#gDz7^}EX+KlZ5l&;BhNYPqm?XAS?UB{$B$UB z;2k`G_^ZJm1$tpt2eORKLIxGKzjNR%zRjJW2u~!KLj6@8mrFZ| zlh7_Hht)OL(*wcN>ipf^>g(fWFt6`pc`tpPG94u2^RnrIHu0QhYMQ|mHC2;fJKNa+ z>q4jqS6kW4XVczJK&;d~gXb-X-1GRGUZ!++;n(d7_PKxyoL)w|ZO4lr%)yFIW_>Yi zad87#*t~XM$-^k{(PTsn5$vIpj60v{+K}CkMJv4YmAA$FKbfn} zfr?xvMA_&3!v~3+|He*4@sa;U;IRcUvU&P)ACkjjih7?nu>%C61WTZNC3}yyS7(?{n`h0&yx2Qh94a zXp1qJ+|NzFU+oEn+@n!yWl{>N=#XGLc{9VWABS=6l;$EbW7$P6$7`ogk3c+g{9G*p z2XGS*z(^&ZD-kfg$TkIQtIS(4zTX}=UjDDm&;Bo$@;-Qq1fUX?wG+hP*4}jpCl%@| zy#f3cG=E!E|&3HnY%qs;QAk@Khm9=(1%e0BTQxx|I!YwAFLyTFEl+sCv__vaLaTv0)F9P5AH zS+S%8q~Ln8X}6NO|y^kU}Vs{wELemMULfk?PnIUwQA5qv_xu z-u!FZu|*mHsIUM&_ahDlUog^YH8s{$4mKiZg#HsWvh)*buAacFZ4OVrM|Tt1N>DVj#AiyGoa$qq^w5%3XH&C=DL@aRuJh04a7=j zy?4DkGPu$~9u|;{vj=L5bK)fM3jD_@161dpKrfOC;Qt6*YzTqsjo4`PImP+G@C1Ne8-^GQ!a++%4<%vLn2mELO zPs%npxkF1i7(ABw;_~hHUtdWiax`HK4;nJ!;#GzvlLV!rLT#>@$hep*?38U+-Z1q% zW;#9P7(e{oUBb}+$yHixV07gwGWrk1Af>747Z3qqF@ia;F!%)4d>k#C3dhpr|8FLh ziJ$T~A3mZOTmIO+D?Q+@BGvVf6B#BoTBw=BN)RYi9H^||rH}laW=FkU|LyF| z3m?h{_z7m#dLw@0m`$#50HZ1O#S1gK{z7_|0i9xEgChDD`qWmBo^nl0U*7-ZDIGr*`-@6 zjs(ve6-lVfwsf-rQYO!~=HWLVX2~IFlWSMubW9k0o<6bMoyL7d!%-~WTa}N!E83M; z#mD2G$7ZE|<%Dbv9z0_W@52zv-Jh~D1T56K7-x^ZTWAlu6D)tsvbi}JJIE&XP>$IG zpY#`uDbzjPp>WHzF#^<3Y-3w{wW(r*zK7nmvyJ?lF1$tkPZfnoaN$+*LJ@-?9x5yq zTK$K8Ve!$B-KPOs6@^^Vw6r+oa!#9(^ky5!p-sAlgBThe2UNbztSkul(#L7x3Tmny@V=(6;Pk1pivrM)(=G5E$dy*;2G$&R{2 zfGmVAuA$k_sR|Ue+u=9uZd1cn6XebR%1y){-olsr5*kQ_x_}#`B^~X4LLssqDO-;cKaH+>QM?V_J5m2}8)NNI#TL_DWXm zgF|K!4mj)sk0iYFC1RY0I7+94?U=`M$UV(_T1V5uVv^i1gItSORh-Wfm%^Sbj#VwS z@Uv>ilvUA?MPN2tf__$Cn97xrY!@l#yD4D|DB^WIQNT=O9`-H#!r0GI*-zk%{QqvQ z7UxooiQ>e=iJo8BmHXHH4*!o)!0GPS@weNfAy?dw%>`0kQZCB{$j4NmdV3r%uaEH1 zrMFIad0P*Um?Fkre>cb+I10drU?vKh`GuqbfCO0%wYMe z0q;-#mIRCl+qfI-f*-l+E3Ahh-SJu@$zdAvjV65gHr(4`XJ(KT%y7l1$Tl1ntzyqEoAJSofN#)Qdw?e~@s)YxJAa|Jiv9U(t3 zLpb>0!7aIz>C8n+s%X-DPR5D}0u-Q-WNv0_x#VuS;Rz)%iuo3FHPo18nM^{9E>=Ye zBRwuh9~Uk5&h#{m`I`RIl$IOj`t1}~$}VfOt` zl31EiQyGfJ*b_|%dgdqk?B4SM)qnM*n~Fi3_BW%D5!DgSYp92_n;Y^^r^CJVLwB0r z@a^x*A&+V5T|%@`!P1%u=QVzJU=)OM!i$DD04c|LHoRd5AW=CnZ8men_B7{&L?Gd* z3#a#|+~hbIU&EZFj;!4~T%Jj%b%s_!UM}If9Vwgky($ilDs=K5WlF4Y%p2k8oalq9br`Iv??NHh_k^?&CFKqoJ>P9b1gJOgEBNt zeKkv_9g$7SDwYyY-1%*9hG<&N>&Moy*5-ySH z7Ctm$&x6#(_Pb%F6D9Kjyq{#%qk~jC0V+mRhAha<%w`e_g`aD>d-SG8{P*sAKakZ? z3hwyLbh4lUqqpNgF}PoCwrjL!?Ic*vERBCD&}}Fa6ryRjzmJO{_(t znYepm>xZJaVVvS@G;^SIhi&&N(@ps=^<|>rB&=_Guf_V0CgK?e*(!gSa7@gd%KS-~ zK`e5ZxQ%YA3bXsQC8zVTrDj!FKmg2JI8^$VgXk7LjzaF6L;}EnSSFb{2m+`I8209J z;9h3&BlD~~9dq%V0a)O~m!62VsXApU+-y-?KqDJJb%+o&F+WmV=g?02S7vL-jqS|n zbQ*W;sxn?8hqp3wI4aD?j5=~3!7nbc0^%p^-1+VIC3$SnP0gxm4d$DUw;qxCxH@*9 z41W>yGSEGp0P&vTbv({eOlQmR0+0!iQV6)5Xdtxsx&yJ4p|~83hm{b34WwcaTH%Uv zap@D$VEe$nTej4BkVNOB?!Ljr6~k_1Y`04%+IAxU{<-*W0kP9lSJ=(c)0%asOwb>A zn4$~?vuXEZmh$?L`)AK$gv;d+K^hrv(1an8%F{+H;EJbk6DtGrZPVB0)P?2@1B`O} z=W+W7Ux#j>xIplf+HITATk&VQqkt-}?XIxJl@``UHE}$;w@FoG-p{;zVc}mzFgRQf zaT9FZ&xY9EmwLWVP5T(dXb)(kOUyW=#BE>V>hR3~K5n6er?n0ZyGOWe+yv{U;Q?_| zMdQhM)a;)lh=k#z=hfslWNKevKsX~e0%$ayFc95}_6|NR#I_=Vsxq6`zb+A0M1njmyfMQ9%E9Ls#_4t?;1lC#3 z)q025ceZF8tk59BdN!tpi}D9g^7}6&PJQ#K{q|rvsvQ`ihq^bA5Jexkj+@O2gqHwk zfJ|R%d@1ztIc*LP6UG%GOaF!z!ee6}uHjP&lXupG(GCk6+T-1O4GC@axc)^N1bRZS zBs3~^kktnMLbpq)B_?0T2hfCvIjl6_)lLSQLSpFFrmEDZ$Aq(!1Rlk_zh?6H!Rhhw zk15DfqF0H37^v~xq0QEhgScwA%%M`wr8!M!@@ImjI+(X$?)cKxf$JmhN?;FxB<5KB z{+)tuBaAkWmx#&P;>3Z@t!mQ7g**zItFqnhh!yMhK)irD>R$-$?}U!v-z$w5+h1TY z?6N<($&%jtbj}-WCc3Lzm*WkVWP70tSaC(8n-5PRdSC1&S~C|be|z5Ulf1u7x;+N# zD_X7io((?hANl2NIm%-_hZX)K{mc=eWxPxduNRa&RVz}dKa4$-C`nDuN?44b^Fm?t z*F6vl{gV#}1U&U7F+3mxS{R3lV-zw^WQ%fL!TG8^vuTaqxLq;~w8}AAL!Xy+uCw&N zH-LI$*`sUsKS9_4qF%$s2^u^q$gm!|gbMv1s0bAIKQXuv%(DHpH1o?JN0Kkj2`BOlv8dv))1 z#C+cT$ouZBwMXnm>U)*5!FseKZUdUzjjv;qicEv&6UDK{Y^C}9Mulgc;kG)f4H#y$ zE5G6VRAGRg^Vey+d$K1NTlGMN7W`6th210)l4e6h#>_41za9r<-`n)4T77`pCRA2Fvk4L^2xyP>*>PgM)!-i~u+3%^I`iHE zlhwBE012F~!7d}hGMMa8%}H&s!qQXb{4qi6;Crfxxu**Hl58NAl5{LyThtz+{FkHf zDwTk{QYUw1KE=SdF;n`WvdY*H3s0flPvr)d+HH)GlI!iSIRCHX;kRb(`3`ZM%x3B_ z-zaG#vL2M*rxTkq*ann~%`O5dIUKk?7IS4?wDu|2zb&~3|AT>Wz~Y`z)Ic(-6%TF2 zH90^E9}WsAg>7lhaFkH4p5CNTX{EK2pg@G3*v?hlhEVuxr+n*|LPA(pa0=h+*a!i6 z=6d)0QrEi}*BvOnl!`&Ng{}WpiUlcM5Pk^Uw7ck40k4fjfV#$b8#p_@$43Txhm9q- zyWA`MC+P+2-mUc&B}&kncI^Df9Cb!47SB$*r}ND%@9NND7uORRVT3i!Cn|uR;wE*9 z!Rq*BIG=a<$`jx2?S~>J(O;Q#-kjoZaePyUs^2V*a1j((iWZ#bt>g^n)B#v)brTdn zZH)Pgyh%RM+#ng0YW!}sH!<*bvOd?n$(MoWX-n{oZ*8E|WQVYQ-9Cmv24u@7RUoFA z$7auo$_(&MTBxHBNhoqa4&(3EFDT@Ahy&F{DtNR{uz8Z zcfWMVHJ{nZel1l@XpjtlhQeyIGPKs=+qDCaP7~PN>^}Fv^^r_c7xFf}3jhG1w|e(S zlLHfLoljP7uq;8LSRD2{l5;IAkSVbFknz1B(nY`Y17uwGv4$K;_20Ot3{kKyQ+BU4 zGJHtos6Y%#D`gCY(^L0%%`3wbyy`ycP)=cCY^9Kw3Pk277WJuo>om?OhNS)Y!Qd{7 z>YQVqxS15RFOyoU-$;3s4HhB`-x7cjXCMT z1+Rqs^jzGH)ye3zv;IKIBjNs&2jJLHx;(E2w>^P;MqT2-gQK;9>*~O9>qaSEvkn^_ zk~Q9bWYSxlta4KDdIXd4gW6lJpfgf-IlNccwf?O;uXkM_Og0W8pVqh85z}hBgC?_p z7H$o?<&v8b@p(xWo6Mmc)&)zq&ihdtCL6hBC^$XL7O&;&@@`z`hOFBPcVpiON2}?C9XHrR za%Czde+fY)4 z)J48_#>^ZQEHV<`{=W!&%cwY*u50s&L4p(9gF6IwNFcboyEg7F!QI{6-QC^YrEzPl zac7$Q{+{=nnKi<1SgRLRU3Kd0z0WmMA2hq#+0UtlkIyke+dn5RIrANj(m)(Vjmd^` z%ACD39J`+)_NY;NRc3Dao>|^oS!JV?60e}uPMBl5R)oMhb$xbaw(QQdBKfWimoj~l zHNYM#@X4bVQ!e?HG%*2Q?{I|+oqzKO9o@0if#}(qn}s_CjJ;=~3Re@}C)SM@?sWy6 zhyTr%X)$N_Z1}Iv9mJ1#^FKm2*_Jc&%tMuh2*yA&-%rG3v?CW60G4k+#a@NIEbSkf zEHc%MnYcyR5at^|2xg%Jpz9s#6};8;kx6FMJQZUnPW+ z^RhJf%D&8w1A!a0bLGV_(1Xrx2z65>OE3<{k7Fy()cF!VH&Gy10C*O4g37O-s`Wl^ zC?#A%6eSA*yaR?z1ZtRiREe+p`J(jfZD!OvaM-4@&_J~lA=vJxFnr8tcVHFAfEx4hG^~oep)K*v6Yb^IV-N#c^BG3i}~t5 z?e~f#rCccQ`FV(DA*Lx8cwPaSu@Eeg%;-N3=|ly+Wbk=Sw7cE)>_eQx0UW7g6q)xjdpu&ZBFm=|}%@8^zf+H+*Mm31@0~Snt;aVgxdSy9-ZSvgiVF~{> z!!--{V@Vn%g8``>w((DZi;)ZNn3Dp8WulSCoeyIGtFC+jc_*#jA-d7=C?eD-iWc)& zaOI8$N{9J!@MP1`zP4GQe_YKk@7n|rn?%hX^}5Y|v2_Dxh7qttzC=na=8!4ox|riB z!9jm)ZZO6vYzMNs!LNI%7nbkZ|60;x$F^NK7w%mn_-?1zHWpVRn=}^?NCQ=h4WAN) zMD6Wy|9_IWRZag#;udJq`*zFwUzOY6p)bF$Dt!zPN3+HoTF`aR!}Z;WlK)}m%TU4y z*3+i(8b@QK*R95O@#O@mXgS~nQ~vc-W#0R(#(0%7tUvPhQCmXY1^a4yQn1| zUa@afBjgDv^xmT7HSX_N z5!rNu-s+DTpR(ThyyJcwdwKV}4p42HU4H1DC{ZnbHm@v(fd_2c66 z*z1p{)pPvhx!WDX1casSUN1Y>Jg(t@f7yz5NPcwK5y=@2 z3Eif-JXAX)+meOqKeHvYU$*q1RdxkEaZ1_N>`O(?vMJ3q#P>r$LG3-cQ(Uh=d1P`y zR@*{x6I-6{G%=i@N#7@KPy8;N@{KAwv=kE6q10iH6JUp@dptGbMLQm}a?m_bTLT-2?LhhFhc_>Q zcZFAaq9Ceg&z?FyefVb13M0b-~ zZ+a)*pK0#PxOjLj7n|NTN3`F<1j~>h`X_45%s3i&Sb?JJ+TL)h7&FPz(U`oddOGIOb@cZ zG`!bVUdU@P$_8F~Vuu1Jrb)JyZVh4uXxWLp8o3{GOMzof>hRIcM^P<8 z0y%H<#uc=)UQL;osX~V!tn@t>0D524&Ah95IJnAm3==iTy06ta&mHW`S$x{*A-Rej z8C{x;S<~;pCW#@Jf;p>+9-IQJ5gBk9U+j2i!t2G_#m0%gO1zZ$1uI==F6G$YQ&^LX zl{{2Eik}VQpq?y1u1T6aSTkT+D^)U9sFqmi9u6`q>cIs}_J}S+CD;qBzq2+|^%X4~ zA0@Mx^?zn5k_r1faNEpTr6-7COL?xb;8Z+(;bCOqFW^M??t~D?rXcyriWnv^NQnVy zVHz$$tdJ1g1!8Jqmr9?|2W8fux=SRi8kdT=E&bT>`sgM01pS+2Kri+3dsIO$ow6(5 zk%969;tc50v|;63d_lPsMITZcMl|Pm2jVBj*UHp=hs~irQGsQ;9nATfte$FGZ!k0i z2If;{hWKJdEeDwYDD`OdYuFywub?E-cWo!Aa$^w(6Lj&|$~O*WIk+mLT_vvQ3#J~F zA#I+C4^~0P>|FVMZtJmBjO4SekWT65IY=5n%|Lbf>D63478sejAzx0Uy9`|7Y^NKZ zQ`+#I2%X0RU9A4Ts-oB0`e3d(m4?Vs8sD|_3jJtllPl8wN5?$VcL6qWrZh|LY18`SnH{(UGMPfM=zA2ewquMsXMq46W4kWX{+pQ^YqmG| zV*>&eZiO7aCS<$tU9a}!WX=Y*y8PhAo;f5A;L+j#*DxWs)fCmmuBW3>p}_&k z3?aW;&4_Mc;lrTcqgeFb)j^tY2S!h1Gx=|FbdT!A(0_!`_{sv4mpw*RDy7|Ta`=8C z@G?cW@X_x(HkpbXhUbo>Tj>4zeZ|F0N&`a<8K&x5XE>&nNdf!{&Y&WE;2PhHjChcA zb%^P@_r-(>dqHF>u@P3HbdW~d|L#alcCaDF_>mmyXkd3>oQ}Pqa4}JuxYwqWbDGw&lW8tX|E?=YTa3*N?Mg zx6j(4E1J&}Jd;N*Um$^82nVBU6DG5z(SSz#X54HDb_35w7wY_mNs;@JyR! zu`M3ce@z#$8Y~mZOxO`DJ9?ihzXe@dya6Mwk7jY+(?ssW-$_Jt(onWL5B{P22z^H{ zJqbdg$Hz9Cr}Ov^q*TrV)_9Fu!JuyQK?aKm_9hvy&@5m^oHj+{uD+(Fvn3q~Ta=xI znB?q$Zw1X|;X>z4aJK_rGRT6Swpx6sB6d~>i6jDEN~aEzA+au@1<#0oIzCz#4hb)u z$sUvZ(zh%I0B#>G!&r9?z5-L@?;NUM!vr$GyTPh_lhbtfukdF{Eb=T<20w%m)akTJ zY$lRJuf+)WK&u*^`s(b(V#yKqj_4yqVIiLcZR~fV`A7+YC_VU%3lr=PaCC5RsODmG~ ztRcgTuQ7)&YpP8rY5Z(+*m7uA<3)n%OKvpROT5F$DA3E*wotyB;!#Qf_Oqx}^Q790 zR%8SVf%W^R5@uNKiGX64bs~3tL5Nv^=n{em%5Ky6hY>{8tHp4W5-U83Y@*>R9AAZ$ zp6G1-d@mb;%P|&#Z0OQDdNMS#68=dz0u%r3C*B5cnjCNm33Q3~qc2hXAKPaSS!Gj| z4klCNqyIV%HoY;dkUf3K>?)rp*2zo^nZJLg=P8fnddNdUmYS7S2Bl`FY~c?FwN)xa z$D`ROeoD|hZ|~}0L#P0iu<*9<_qc)T+VBZZ(o)(q7nExx^dfYSoRG1N#F zPpRC_(wl(W4YFem{Az)fz8C0*Us@afsppG540ZVjl*v?J;Y^>Y3ER?fixkZfv?x+ z1eV8(!6V7Jgg+y?A*j3H@|T$JXfb&l`|JvQ=4;MWK|`aDoSMobuc2%b@~Ct*S`Q z3~x6GY_>B4Z38=>ui9oTA@lE&n0|Xdtl&RNj~m{9+*!{D+fKK2QUd`m;JZfF zFW*rsQjOzzlNLX0z2aXSUpXcgNJ&iK7RP%%-_SQXpDFoOJDL~n33eGe7r}ztP*N@|L5Z z)lf7_G)M?&K$ zk?Ov)H3|E?M4g;?uGoMqe2EI=xsWVjSW&|8lBslX04k{)XYsJ@@5(sJ^UN zSbf{2$A~^Xy}QZwkt-7~odMn?lN zre%yq?AyRMTh)`bCUoMoT5MMO3&C|G!P_^cYDei}k~K9vM=HG852xjXi_Ole#jO>G zm$!DCFMWuqCG#zdx=(`##eBb~HT>t>!8=F#RvBNO6Y<8Nb6L1f6#Ar-rbe9P?Y|OU z^rPVOdHPt@;oWiaYdu_@h#s1CFwl5lb)204hpiD}~oXgfyN4A;Sp>Woo zjses8n!>HA(%ANCIS2Qmy~oakrN#1^0Vi45FfrOFl$9ZCbJdT{#K~*q?j8qNm2bVR zsB_W*n6qc5Q%9)^o*)gZTC6)OvMm@uMI2CcSN4GYuatA!@xLtWA^L=!Xg&3+k^ts| zm}TC^$!YKS($=V1IgX>}W)FJ|E^Y+dlO@rK4fh<}7_@l=7GywgyZ3*=&CQ_A4Rb*ptTp|Fk&KzMr+Onn21MxN_Jf z6Q(dmK$WdDnjfj`*Fpbrgv0%*PM^v|9YPKTgmJ`XrBnJuD5crbt^XEv=KvM1Z##1^ zCOyrlSR~&_bA_Jvr$Xaprdqh_{;M5}pRu#@9tl7H8!-s&Wjj0tubfcI)B_PC|cm#6neGw~aNvF0n>K zb1?UC=SJm5N5Vif>au5+LVpS~KrTM*e3U|En{?Hw6Yg3)iegfBEW1o&X1dWti{l5dEdtG3h_BC6Ad0LZ?yNM zhgP)O#-V7-sd``LXYC=yggw@RnLRdFjKgq#u8ZoXd^G2>!{Np1*4BfE8E}qwz+Yis z8jU7-A;>zuGI2A7G5us`^CT4E7D%sYJ+{T~s|NsfAS!Cu&L4{0;AZ=m?+izy(QsT) z^t{S&?G`9KnJQJ*9=ca?QrSNN_a>ybY7~o=OHq@}k4Yo~ z&+B>2Usinu$S1DO%b&P!pO8>jmj?$KWWUcC4-39-#}_HKetd}Wf9seVcIqPcg8r)8 z3hBFbOzH>(e&qUp@4G*6DM79$ByG~c97BpYnTI@OH$!UBq-P)AXyruJ+tBP+43drM z!BL?Ma{?;-OjK;w)=?Fw>ctZwGaQCYrFiJY>f+&wPmHeeHFPIG+xqkt zF6D6GQ4^ZOb1wE2l=#!mo<1dU_z(Ba?7zC;R}}YTuQjU_KV7I}1qDm3-i9jnI8M?% z>}Nl5e;9?YZ1MDofd+&|m6oVo^W&XcYPocHgEV^TCWmNrGze(l4|~hyTU9yo5hY95 zvKm(zGE_=R4%?bHkMJE)2A%+u$d>mVyW1o9%qDaG!K3TVn>Ua@%ioKibcnD>AN^YZ zf7ozRb3<3>K+paz3E(&yUZ!|xFb4EzU$6Yg*kB(ESpH<9Rf)$G$tz!$m^|~KHK4!a zx=@M8X0bGoTxBg034E}iQEf}=4*&!MWKhM<2mLz}{9bs>)oSAH25{6NU(YWel$V*0 zjY}3!`;r7#*69`dd4X~c73@oPpKJ%M(oFPv0 zMT&;Aw9Z@Y>_ij1mGlh)m;%SKMczqn9zb>eG#chhYkqrQE&i^ZQrr#ZR8PM7f^o>e zvbyW(2LFk~QT$Y*xW&gCoFyY?MBP<<+fR3V;peB5H%7WPMO~$>8(ysF^lIJ*r%p~a zATZC>Vqhm$pouha4{n3DYRIf zF9AJGO9Y1*xP#@=9wn?N6HWf9%TBjsTJOKTviYB9k29O2Lq1)gon+}^?)eDpJ=Pn3 zi`E0e{oFr=!@s^uAX26UybDo{s(=$5*!AUzpr=t#v<%fb|-kQ zwiJk`csT~vGG0`?-a1bP*pBUapYGcgVFM~S`%zLILlHW0xRTa7INBG1T?^|ui#)oE z7iVGX9+^WL?z?Lf ztMVM}H&Ts@(7X-I?OvNktvaj@-$|Ab`7lP@$R{r=fl_1;$6HUsc`nj93e71nM#?J< z9^=e2>*{v1aRD!>ZAbe#M@H!Juev$GYOtr7r8=CB{>PX!tqr=cMD^e!%=)FYxK^C72+c?@ro=UyP;LESrnu9eu=f9eK5{Ppg{cwmlK2 znh~O@;rj`wOd~#GomwPy$pbW2P-%YU5WdZ*^ECEI`cH%_zVK0+W-KKiki4OK*qj-L zz1mM#AmArS4|ALis4}TB0gBui{@qE_?6pDB5QjAf1PS>)lb#F=Ef zPzUsn?VX_8PyfqqOzokEMP=|#;_FljsEHijhVFY_dL z!%nWgX1qW$4rA^j87kC=FqZ9)wbwJFMV$wmPbb*(;nf+F<}l6tmKHr0PJL%S<=>iX zgTJJnuRgoW+rZ4_2s!Sh`T?CY_s;n|bM1U#;%Kxp8(sz(ywaf$&YrxkA9MO--flIT z{kd4z)w2%vwkY9jTmNK*bBR8g;9MzdDHS26&e)nQ8cD6X|An@*8m)xW^6U**j2QQ6 zZm6SwNz@847$W}y8SIy|oM_W3p|*I8U+ZvPS=I``_J@p5RsS1ulX77Di3YiL!KD8V+oRS|m(*BHDZVq=*t~H~cjpjrqHJmj7^#xy7iX7xoTxi2_zth`lOhE!9nASYJ}%^33`0S-QY44Kb>f>`DV~z zF#!9JEL~kMtWf1NB}??`o*jNK{S*#Fbggm$tE?o7MlSP^xOzprG5A{WGRQ#N6;13C z59mKGadPf4l%JM}$eF=F1M%(#d9dWCj}ese+M8plYa{iQ#uxH({-jRRZ0Jz{1)kvc zxDs_M{6%#yr)SIvYAlZVrTj+&1T4*U9<*k&drm~uEoeZEbcV-6-PfSmMDI>1o$i2e ziK6rkb3RaJ`riHgl&@6Br;W$H_Z#k)81U_%Qz4}|c6+9l3oV8+6Y8*DnuE8I7(NBw z;A{b9?Iu!X)a~jT+f6yFz#L>+?(HIj-()7 zWx&jK87r--c%jArGU@A4CDlV>6##mWi}g8v!s78La-$+LRkb#=xIh+vK#Mly6-qux^NBC$}eaDfp;h z$`rtdahp>ldM=nvIIa!5Xg8QJYpJL=f}qxQeda)sLJTkZ?rc;?KE6qvsKSO}JBI@9 zwTRZ5z2k=3^;4RbqYI8H5+(>lVrQB~Qb`D-@w z;O@hl{{f%NZJwlsjeRkjc9YW{NkzBCWC0w98BJV}p^m4lUATz;SxMV7xq4Dq@wC!H zLF}*#_p-n>7LBbS+ew(nmww6F6~Swnl#jX@OtDX|juVAvN}a#DnY^oU6qm5iG!OJx zYaFc8*KlJNW3wqe4Zb^TataFF+Cdid#(lJ-B$rgXEIE;r_Za(bbW=7SsS39TitCf< zAO*~M;oG^P6J;OScv(TKN3>Dv^xho@$L@ddSR=4Gl?}hP2^$r1ZHjGdO8?1Efyl+n-3mdNG+P!YHkvnrCM*D4C58NY>xKlrvxk zjL5K7io(;(rH(e;;C0Q^!eI2-cciIvrX@dVVH126j*s#YgiQj9mWKr9`llv-nuM4p znq%Th)zXg@H7VpBK0oDB7tOyXJ@Vy+Xq{N4jLu|ZE2{Y0{|kp1v?ig1ofh)C;Uz1^ z-A0i=!gHsDpJ=KKE4d#+iYd^ z%XW&cD!47yexyxW-sAAyt%l81VDCcT!+(b7;5ivl)2_5eF8&i)&l!|A+FO~KoQ z2p$`q&De2cfYS|D=@D??eG{VZS;4*kub+kO+par$$JHB~ZY*0o5(^s;&<&VaL`UKD z9;?JgFmQTi+Xp6}2%XN?wmKU5x})3BTI0~LuHD=&Dd!k~Vj^HmUqF&zb3>VjZGa?G z;B>tr%JHJbuh@lkEDkH2#0=`p{K4Tz6oD4h{t=ND1~IEd{{RcKD5Y#qoae4|pjU-h zcxEe;b)R#=c+yytZo7$QDHWgeQYm{yGMT(Zin28;oqz6dj}sQHw(b=4(~jLag43A` z_#~e7@z~`!4wXJZK6~4A^Z!bIc{4tz>&j7C{TcD*q^j`qXH_c~pL%5Y`ub@#Z0o;2 z5$Q|*;lV^OL?>sBn=xGC6goVepcFeto%MAW9^V~@UghM{e z@zhDTVY?2Ib_YOz>$375GGncp6b{EaiOlFYGG&9Cre=>@FU`qSjgD82A#hE|l3F;XK5ln=<^OCR;DVx$V!nw7BhV!^_h$ zwrBbiPsKl8ae;eRf!nxoV1cy&px!qkvM?uBZv6ZCPXo+pi2zQ>6_H|_SCIx$#zPd5 zE3c`7guX5BvBQB_Q-iR^`m0-u_FP+qG=--7fWhcq=kfMXqLpl(LbV7J z4aY&Z&u3jD^y2qbAMn>fT_|#S%50lURPo+_Sb3lEMt%H{_@u%mao;&cZC)>|Nn27K zi3y2_xRO*3H`7cpwC%R$o8o9})T$C^F!Pt!8D5H+jIidhBn;2W$(%wGdQ`YXptg_h zcP%{o;L)r>eDu7@=yc2h4#DGlgW%V(T8Bk*KB$g2EBfTuQAuo0Xu}UbAB8Yt3_P780LEy(;j7MoJmsfBeC@D<;fdi z?pPC<6|_1_ZPZ)#2>E12Whx_VUvT~e>$%Q)`%B_(Cw*gXFIA-Nyo&J=Y%w>0b=+bx z7)pkZ6-U?3yf_?|mgMgAjc9Zt(2=9Dj1OOfS$BP^C6mtehQnc{nVKtYv8}Z1fS83g zx2HADHFz@tI{{OLW;oail)u4Z7RjF5iudaVuXwKu9)FQKIh*%M-;cm~8d0?t`q72S z9Q|U{KW`JU-0}YK$!2f{;F9V&h-JEtJ&FQg#-uw;qPAZTe@h~q9Ps%;oxx%FvuWYC zTXQ1YA`6Ns4}@AyRHENY7z2T18*e5|hUl7?{fw_96f6yUPmDD#QszVtU#0LWnLP&T z2#w3C%zq@1P)?@<)?u?0cCy@#hjxHRec~444;{pSj9DRstT$&%Zl2a))!x?G8Y8Tb z-BG~_?uUErv$S0*nY2p>H9bC4!q&=;y+yjA`7K)ylw5ViY}?q5u_gO79H~s7_O9|f z>;2*R-m=t%)BrT9Y1|p0p+p@J9@~&&hP_n}SVu8AN{fJStHXE_(hr zwDoPS*GS3`9!WDKCfdtVp$W^*VGSxSWq_LopQRrmF`(O^eCNyCaSZ)KQU~w)T)N|X z^@QYmdZ@I}jmxB_gWTh025e_LjaZp7ZTO}aXiZZ>)?tyR{pze(X}6zlH&vfQrqy## zn1Q24BKF5UcK*`a>0CgKT1iC}+KyLWSA~8cFuI$_NX|CCSiJQEifYXs1vz4cuGl$Q z+yv&T1dg+<6-Z6$GuG2>YwEp<{}wCAJUTPGI@f4h|CEe zwA3YsLuP~sE1S13DU!Dz_@7gDT@&ArnJNjju-nB9%T7sg4pn z%uYu(kkA~t@JSV(nhF_p_{(;A8RqX)>3+mXvcXSGk=lA zFi;m?T`9bvUK8D1U>PNiObPoHz#i0(gE3ui^1O*S0e3Dt=ZUqd6rOLl&bT|8l=hi* z;m>pteBlbRXt-p@L_q9Va2-{0NK$vYU~!J8y0HF2m8UCp$nK;RmH+(i%`Oc&u2o-j{GrJ$3Cr8Iw*83?%^4k~-|3v>Rx4r#n)*GpzwG z;El6~=U}b+!BVFYRXdJimf8~*uInT0z(7kQq>GTDCK*umcmf+I@nw5k%&| zS$i1MqjExo>uhV`!p8}XWmO?xIANdZb%xPNxP`ds629lb1r2cPoonX0nh*uSB1h*K zP4Hv;?+8SuCN|*?&>5{yO5`f`1~9`hvlI-bxOrw*Bo-bq6pHC%*$U?U0}Rqqn~dtj zp!K}Cm#Ku6b1}}6qlR+vF0Pt-hKdf{fR)yqB2~}@0I;@Q!dI%?3-l{ZrvsL7!m=$q zt?T^~5=ZUuxt);=T!UguA0{zrbsbMqZgp1z@Ms#|2AO}{+(L#tI=$xE3de70D4}d~ zl8d3vj!3^Kb9X=yFu#R#xF+H27tzEIpde7QSyt8-DxFS9XxL!}7SinpqRjRYjlb0- zMdafrzh1l_F{PQ!>i=F(48;tf?~qKwU2=sbgba#wB}rW((YkqOa)gj#Hq{iB#mn{+XIF>*+?j;FO}=AU?vSo%u-0URVZFhVqJ$~(A^?KzytQiOyd;6db z=l~G(O;PNq*A147qt#A6zEPXXUX&rDTN(2vUQM3u#CPGRvAM|aW_2@MLC6@1k#u8s zKgLb0(15K=Ll730Sz|Afo|Dt(I6`v9axZ|LEy8Wlv*ud>j81zDI3k>>_nu6@96npw zLFkj>F9WoI)#;3G6pSkH)HejT%3P{s_l0OOJ91WSS?4!u&|wterXmuoT)kHA(!O$o zBdL1S{zLR$&CrxLyfUuD+iAR{J!7YB=X4*j+nys_MvI@M8-m^O8Y$%C+^Zqr;>oVv z8-Smt!29>?YRou8DdJZAhek7rT+-5Gy&V8!rL{+YD48qwz6MW803QgIzIV%Ydl6$W zX{Lf@M><7GAc!D;=kyc>m&cWP$lgz%P0L@Uoy;@pjJ8o4kICYVc(D0FOieo|$RN7x z#}~O&e7mD`>6-JVP#ZdC?q&~g3O{KDDdlY5C2UfTdMW~X>CrK%jP1_1(V0gfD_ME!XXJ&-jYkj>Uig&s!VSM-+)RZHWks&V^}--eX~ml)|2XYhs;^cCvOM{GwF^n8%?AP zUl+$>GS$&*sxRj9cZX$HLD^aq?)5wOo(WXyVO*sa8UgCm$q_W~*Z@WM_# zKRKql1enW36)Lb`zpSQz=pA%7>rI|b=C+|BhQ7%Xs9lMpkMh1RSgiE4G0h<#k*Y2U zH(AXme}RHG($RL~Y_t+xzw{r9`>{=*^$mH0`@B3)Ix~I^&R$kUGHmy)x55#2#i_nf zc=5KbNHawvxcsAL+iV9%e_lo`2UY@T6Rdh+i}%tvt>@!+;pMLTK|rXNNctTV?rkpOX=Hfx3?m=8e#?@ z{CT@Uj+pe>_@(#znIxF_IqeS~HG`-Gl*Rnz=7gq?6+R*^0oThT8=fj=0=?{1Efk&= zKG&rzhWpq1n`q+S%Aq4fA4I>uQ5lDr>h;&lVkQWW=B_pha6s9>F&0P{tnr+kCq_eo zKup3J0a2K(JRW1CS(CD>F+aPlWF!fKbiLx(y@wxa&dP}+&V55g>)G+tM|Oizk}-8s zYQR(*jpZeE5oMXJ<9T?BV3;YumL-C@TAv@`ZJn7~Donp^?8w0PcGmkBZNI17JIJlr zvFrM;*n_lvo^PkOTIZ66=$z_W$E<+W7Si6Dz+h;C$k_l84Q0hUVz9@izB@6Jy^A4% zx8~^qfyoLgJ5WYOc7wM2eOX?xYW75bV#MS|-+L(nR;{@0lr8lJGk8wc+ETutJkGSp zNqzt9gfcZ)JBD5D^Tg*w*l#6-#==`%Yd|O-MQOHg--oDAkHH_r7aVZSVkZBi06=kZ z^Y}erqPuhKfMhwHdN)`h}1#8x>Gk_}~) zl0HtS;cx6mxEgDx6ZP^3E(F)@`UWIsD_^y#v(Z@ug#2|~?Aw-F!%n3IP!3wbQ(%A5 zi^`fH=^Q^VC5nBAH)e84%dxLR;jnL-SKov_$@N3dVuMa~ow5N|Gou)&m!%9g4i}=N zm*_K|!i5PV@Jv3pqN|$19O8Bh)!+RT^t_AVj=9?unCeboP zDIou!vE1Or8J^YN26?(y3?)S{#j+J^?y-^b*`pRp=5sk2Q@O{)e>7L!Y@H_QEMNOR zYMkcYge5zm+xbtRgD&$rS@|QA?(hEO(Tu?RuTDjY16@ ztoLU349EBmaZmak2tUe^R}2izgH$_tFX}AE|6u36)Y3V_67~qC=W(7M=gEK7T5sye zqQ)a9*lJ!rW8}3nljL)yYXNXdbyAUi0V$Ac{?!a46ZjUA|21anP?Musk92X|uafFi zzJY#Y{OU)u|46|PQdkw710ska*B{oH1we2*TRx`cHYp)Zq=s*yHSxNL{3cnx*t~tY zC45FS8&dHiGr=lkrWb9l$38? zb!|=d@dtcWrBRTY!Hev_32r2>^2x-?&&Q8)vCX{crKGT1e}-ei;IU;$0^S6vD#?|b zpZB3FTAq7b#%F13KnRSAC+C}5 zba{xqj(~)!c@-f`*45vNY~78H{!DBf+X9Q%iTG(!WA8Q{Vy|^ic>x*yEOKGNj^y-< zxYGSSGIB$? zwYuuySMx*Fe`FUC`|ly2ji(UXV`V2=9?WXGu$$0~{t8Er+h_fDv2dzGZ|~PngNfq? zPgu}rYptW`ygTP1coe++U_g`z4H zYk$Pq>&?m(h>czeeu*5f{b-X{f7>35;9$JwJetXi|YR5T?1nt2yOy;y+Of-LaF+*&0Ob< z7-Bv}bFn7%W_Tgcc|%p``~#>7&mhU{PS_1|+35qq??*;uaw@^S90m{k(z3Fk{VR(9 zwBy`a@hP0Rh&<`hGc9PyKc`D&iyA`O7C?c2#9PQ+*{3Zw+RA42QG-Snh2G85pYF3{ zaT!w#!>{@nv9|@_@YK;kDXabG#_<{#+=jB(7d}_rl{ySSKcy>Z)AfOq|)yG4Hkn=ufN2>e8nS7+$rt;5DiPYWX;CW zb)wdI@}u-5FvWf`C=oUS1F?F1>NXU;TxUmAc!%C6+wS0$7q9_%M6L# zlt_OlM&Q6q9>1cRJx*5@axd0gE^R&4T+U(uSJRi}{8E&JLOak~@@F1o!Sp7Waj~!{w;#luRXQ5nb8Y8K6h)YmBsQu7TC;Y(v#74OZJb!!wm&64LDM-BCqU+B%y2W@_WE-TA_)FN0a~eZ7ZJpZ3!_YghG}k zp8n)ZwjAK=wRCkCqsh5#bWdpNYNL%nwUrneLuf;`JQyORg5?W~gsxyjh$TMBC#j)DCLtG0o=Jzw^;%^BhT`6xRS0?y0iJpMlFkQ zV>^d3BWP5=x_!QUlI(Xodqht-)@Z7>DB1EDm1~hh;k?G|8Un{Ab1pyexXn~}ei4hS z5RO!Z{@TDc^mi@8TJJ%GbDmN#8e+j)%t7IS4%fN97nMlo^taI$`kCh>oU%}pyTpfP za?;M72*g~Fkm1}o#S@K%f;ZV@V$mNF-LoO<%*Ygn?L5FL5IRkE8(uoYe`Cm&av4cP zWK_f-;RTK$L>A&|&qW3pip$okDnntec#g!`grzz4-TP&6J9}Dwl6Pe(w5{%mD_i@k zr1`{PKc_}}iXVRTU< z5F>r5wsdT_+AU#oWPCGXLn&0<%R4o3>$hG3?GR58tPKP>fR<~{QJpp@-HBBV3zmyK zmnZ{Pk^57gPobuMfeYyA)6Sm=gMgWpwW4Q#Wjm=tdeX^>z22A!#AeP=*G;RHYJP`KLNV=2cq8}pmM%OjP(N}c*?RwW z86;5qG*hBvcp{dNB@rKB4=FvF=*RKqoG%_?k6xwU8$&#=T~=8Ch*pGtU*Tc5^F_S9 z7|IhGr8AK7$k_#5D~-9`j=XgFt&-z*hdufAP82Qw)AVxf*dFU^;VELGB~MYr0@ zLtblAH^;vVOi4v?#_MK!GZWa1A(I=8Sk&OPM}7w;HbKirl|4PDjk;kl~nB$Ym2L>H=Xf(R*Tqh#XG5SzOwZgK+D`K7RuFab;jK7Sy*gDw6V7SDYa;0&nh1b z1~V+aWEdw^tZ)14H37sx_I@_w$1{Ktgo^fC8MwSpGaVe`Yf=FYtG1Y-5DfY>w17Ic zmvO|xDHW&42JXWu5sc5*FVY4sPytm|GWu?9%&Q(uEOB-=s9ZkElT1vXGux< z{KbS%A7LF<$Wt(Pqu|HwY@j@gkrLXEs<+WKmU3?nQw^>rOG$SHfQL{ z4QiD&one#EC>D5B2mT)CV!S$(@2JF(m(Cza6)UU|Z_1asLFTn@TW5-r(nkc-*i&P? z6sDG)@CCu|D^(zEC+h2_Mxw{ER(mYVs4W>1|E`Pj`6Rw=NF_S^TYRZO7Z<)JkWQj@ zLLm?gT42l<_)@bB#i!v5X{3g*V2xbDUmdAT-4<&8sh5xqj;2odS?9_7J^5*<t7 zIXF2KN|yU$4@#=LrfQoJSv&ShG@F3jc7IsF08v{Hu^Hn$)?`b4?2!y1S3|c_glSNG z1EypEo$-s$Eo#_Y$oaWJdPa$_QjPwpGpFLXa7a?NNW}aC$*^k1LT_TV#k$F{(F?26 zX(lCAa>}H^(Y+C>gsCCfMVMHQJ3MB!JC^V1*tO3UPDn=m2l<6cBRzF4#xHKd-gl1L zKIr;N)s}iZjFfwO({|IIXN&{Plz$k_>}c}{Beu>9kAC(7C$G=qCUJ>H+}yy|+Z;PP zT?X%!s+%K)M$gUrQDy!fMf*=ov# zk_~c%(Xv44W+THa-$ z;*Y$g>T`3gi=T02s5BMz4>&eMPkt3y6>~V^_k@YEFn1H&CL94Gvm^A>zi1?6%r*2$ z^)-JbZRAl^sk>(dU3}TJPNQEt5Z$!XwH9v68xTiXJjvAMJKdUh)bVA#8XTu!G=p{A zd(#nI>RlIIFkR@tvcor-2U<4x3z?)X|_Wl8V2`mQbQ$+g~kkrf&a$A&Plqo<~nHM8akGu#s(AZf?79 z=r~84EUfW)KPhSWOvz%S+V3f$fep7BdLVFUfAFN z6nZcC$PD`SWYv%L}wc3K1DzO`9u@&ifsG(3QbYyt%(BumwLBa^=56*ETj$kI*(*$MU> z1O1@dB@H^cPV9q7T@hWVt@#~#^s^-HVuDPg?Cu8z(^q_cg>gbR3@}FL>LI%h9`>8e`{YhPeahpQuu?$9l$ESYufTPC6(%ViHlTLrIqs2N3W} z92>K>vpDuhq@IuLc?&XDOJGUS95s-?^6I_4j1$Y-Tv-yCN^#dF_%s#0wyIoq96NnN zzHW|RUZCB>kp5WYZ^Fl#5AXs5cZ2j8#Zds&zEsBWct>GXK$vK3=ws2i6=u}8 zz#pl^a~D~oRSPM>o38cJL5sN{YHYSxcQT#SGFazzkKsk$U;&jt>BAKy7hMl)ugJu= zT!q!QHOD`C9+^`t#|vH1o?tyn%<>)!8E-yfWA$oPS*T57MWE|+ZREU!0!6*Olh40H zSV^hN4;1QKU~01c2*GN8Qf5{*S6RG%D8TCyU*q|XiaET(aq%jG0G;-${R=b(Ma66- zhSNMrDQZc&bbQ^l>qnv`IJraoXi6RBfdWhI6krpMX1n0oc~5 z&AF+IDIjnQ<_4v=7)AM$Lh9vR`7GXIVC?9*e-Y9%eR}&erJHk zW-FZhQe39a7lJiNh0d$B9lAPLPbxUD;A5@Ps6r4cD&InEz(3K(a=*B8*F5cWTm{*P z-^(xXA5!A=%EE>L@yYSY9aWAu{9G@l+`ZR^W`CfDQ---gsLuU!Xx_grDGJx(GW^SKn;dhg9Nt%(`#Vu+Tqi9Q11n#Ra?;uUA{_M&dUX9osVPuX0Me8w~kt zVo?VSt!>gG`5}-m{1{{Y-EF|2@8VWC?6OB7i%EI3VKYT@m~0Qjuoc%A)T5HJVl z`rmRq8vJS4fEebZLnb#5a{#OIKBL?A$KlPdFL&#Y+4&lIlqnnaPrm|lZrRpkgH%ct z5yF)zsi$J>TMGMjPmn5_FioN4`luz+B83^4Q0frtttmL`X@eEw_gywWi0TBgn$#1E zCRi9Tvo&QbBpN&!sp8#@PNp=Dezkp9HD!h(o9HaFh%sXg{T8o@^7W5IlseskiVUuK zOtXd2oxz3}r&DWjK!G8qHeH-S+U+=T{@~9(YTXX*{CXB$HLb1V=fuTgC{c@F zNmGbf8S;LEm!(WYuCvOd`sp@2_&ttW5-+RT3QD$l;KS;$>4leHPa*@HuA{F0@IkN< z|9aWHIEqx#Pm46AJSClGFPOGx%FW+iK>i=G3GnQ#m(BZjApWNDm)BAs=@26V%k`de zGRl%UGS~cWne=|rlMZVGJN(j8W{R09axt^4vS&Lsr{Z~s*(>Etuydw*A?$e^^nfr| z+mQ%coXED!lV8;e-t~bfm`un#D5!-NX;MZK1!; z13u7$JlpZMsJ48rbSc~K1Yoj5=)|Z z0KFK+3C#*QDas$W%l^X=pYb2%$!n_Ox+`@aYnrQ9@jwxA;8l>Xl6{`EeSfFNK9ofBeoB}LxC72GdfTC=PLm79bmte3 zZ|JsM;dk|_hpE>Z(*fyLHuEhvW>D`G^{sJv{KL@YtKLlmNFnwfoJB<0_tt9j*HrSGRfH7A3 zt{!md#Poh9bPQ7GEJ-dJq@t;Z*}Yh-lmizQ@qPO~skyXZJjV&kAP_82aCD|PYK=A4 zS&zkZ8?D%@>TxHhpGMEWBjAD0sZ~*t#C-z*teF z)qXbb&Q@|(7_nq_h7c$z7?8T;=jY$Oo5%mV?~ekUI@%q>xi8k#>Rw%#CG%{I>vqb= zH5o{j)2tZc{p7~pl5wEfz3lk`x;3V4^Xf@s<1chA$^|m< z4t*}~no!t%Xg|5fT2N5M)$ODdE}@)CI7Xq-iSbSH1vqQ!6I*V?!KzniAs&yCnAi38<+Q}-;lvp;zqV9mp?&MLzT1Eufi zm{bBfLKi^DXF+tmYjG12ks=db(`K{2gkg<1zcy}aah7XCHkm~ASDV<2%fe;+`AiMn8H!V#u0{0q*=OqJ^2)MTeJhAj;LHV%SpvUQ2@YBgal^+M${59 zzeuQ*CO6oisAH4eWSOb{j2Dt1$8WbzHz?agx&7?%OENY_`+@;22Fh0N=je>jV&$DI z)H3gPKYpa@7cg%pFy-fm?grYF%Kh1#dA+4^$rMm79Ti)4vEGE?0pxpkC zXIaowVuefMRgaVICh$66dbZ_MGWS*-%m%3=z2qwD8b*~vkzt_HMM^-iM4}P_w4oPX z?4IGUHh_6OuS=iI((XHR1#F!=G9`>wZKIp$uW&dRZdlAN9~)5_Rh-W^Hp(X=k%nfN znoS%wI@~(H17Ys3INCJu2)eY%RnV&R<+>fdCq_8u#bS)V4%&Db=x~|u*_=d@ zhUqYXT>wR&gT&}i4|5DU#v6$%lB zfQBDc`NAPsfs?kgh)qIj1eKYH$JzMZlAmsjmIxofL_3wNZ{*uGyr9D{pX#mM zmuZ{;52yb!x94RGx~#_ex{TbVJwh0 z8>md0Ya4xe#v{3J=YE-v`2Nl?K^i%BE*ZL-GT{{KtgRy|LRFxZCNLvE-=G9uM%KNA z1wXVkU41xJQa9YOlglwlu&t@ul}I&SCGT%QaGx`rLGa80y3UqA!E0)~*o{J|L~c+> zh^xL&e8IN5bK>P41WCOm%gknGPfc3P?Sk!Ex<~@9nw&r!HWQu&AM^`kRK8jSSlt6Y z&0t?)V zRROlM0VN`i%%RfON0rJewmF7U{1Ga&{LgC&@s(;JgPW4_MW5|0TyREG9qDc?yeeQy zt(cEz{lnkC>GTMzo3hr-*pTRSY#j)tbvEa;!%K@w2SS7>Y`D@vziu(l&E?SV>RnNw z)T9TooQg7T3)8FiXkz8sgj}R@ePwNE_Jp%gRd4iazV6sHH+&KFKAHs*q;DAo- z$_y&QXHij+e--id5coiOmYR?!R{UP6MCmJI9~J$%o0*6|0z7OinQX~i!vat` zR=P&PH4`UOAc=S1-KwHZtrXrc>4-0d5^F{U0OW7d&JeAQ6{b+nK~A8YlW*&mlkrzQ z`{#&sV*hc(Rq>4ILEO}6md|E+zOIQW<$Dtry$+E}&IriaEoXZRHYOKlLZGjU?kK;K z&_XCDiiK%W_N>OeJw|eaBH1}_Wun|5)G$p^@}-fg=^c}A6t!vZ7XWpCRYkwk6o9yc;bH0 za-q{~2_iQ)r>`PFV-lp0`EcO>^dpc<1IH?+RciK*v;?^ zE%QFjVQ5$jj_W95$dTHgjKPMG!ni;c=;`(YuKltEQA;r{d(vXm8Hi@rZ#R>y2bQ zzPVHD)Tcmg`3JJ9mdz$+a+D!;{0&IBSs6BVTm8HZ zw#$Q&-oU1V>#j~hom_b>pvNqt*U7u%#ZQFM3yBjf5+|_t^MC2`LH(gbCHT2^0p`p+#jjYO@k}GKtK1pa;$f|X3B)V5WL{6!F59l>xFq4q^g*)o0D5zA?8FXhH0afGyn1T%Km;E1Dn65mRts{twr#Zz6@ z=u8F`#!SJh<~X&WJ*EO4R7_-H+#;3-_?ys=Ra?RvK?T$w+((avfs?tNsL*qD2?8++$L(8u? zv3mC4nas3_`Q{Z;cy%kiyPM-82baU2N~Ri6>U9AP%7(jIjc_-Nq-0+sSmc@a=-=%A z+2(iPR>fWT@F>)gOrEqzTrR9_f$fvEo{iCrMRqdCscx&3Cx&FR(zbi%HjyVDPo+A# zT^$TZt2JK}s7Qwbkcazxw?i*^xdc8oVEj^qNj1@?G^i)|DnVEwl(|G^G#P+@Kkd|} zy_m>IB%7v1u1Z!mw8Y~_fisxSAFRs{q0eRazO5SrF&F}VLRg9lq*A&eU-$M=sWB3V za3`*vUCZXVdUjoeB{jK#ND+0tet|?)J9cuBdH{iL;uEnnt~!rNR6O7WXzE8}_mj&} z<5BJAE7dWQRu~!(&G!KjzT@hmUo6iRc!2#_HbA;>00++& zc0(d*I)m`<+UZZHC`tPKxO`>3_r(pfaG z98PbUcpv4kEpP4*uz;?%M$;3*)LM1n%4t{jx0jYZGepnu%fFaRSuEh0b@UwpDxk9i zzl8phU8r+nc@9j_ukUXRBa)sj9C+GyAw3{3vXjMwQRm6OVM@ZR2dwh zBovYq=|w`R*lLd}Q5c0foZ+&nu77Y7T>h}YaG(8gAv=pdpVVi3Wf*Vu3mh+O;btG8 z!%53!@tj9tM!$Qiz~~))ULJ5fN9nNfExPwF8=YJ<+}_NPgsPclJ3}BcTG{e|@2u{+ z$prK&pp06*kR?N7E{1ao{s53GIF!ZCc%K;`Oqwh92&bS=?f=0St(gALb7Os$2=%v&$91lGMkIZOx;Cfud{_Mo z{c4T{w!)OfV$Nwz1jF%6Snl@()RzU$o_!-dRG70lq%54Y>#<>@*J;1pz0DfhU&-EL5~XFfSY!x%=eD7Luk3c>F`7w$i`j66 zsoP#>R{xR=wc%^`x^0J#TVQT(?!fu=yyv8a*)lq>@9l7*Y}PAj{7#I8nM1VF*4g}l zF9A4C5JiA?%*3Zd_{)*%LjE&*9N&+jm(Hz21wQaai`JyAxbfWvwi>)VynTcGBj&v8 zFJv^(GnbkI{rm)@cNvc14?q2n;RLJwv3x?UVy!XyVnaF=Emsm|q>9TvJZP5V{^4BN z&e~Fg|KxB{t+Mw=+v|DErp(bqvwRpHspH#L?DXAjQ=PSX>&MnU80_gd8acg%w)M5T zKIbu^=fFX8tT2sgm(L7J_S^pV+kGCD#YX+P$hRi1 zcCwe6JMBMsC-D$!d7%I13i)f~Ug*2A8jy}k|7hA`UvH#I66co|$}R?riDab9q<8P9 zNa|;y5*w|%{cG_Ps6wYzdL>iSg{tARHufPWk>}|C*p?0lr!cg&zDzsud)MAwyM@K9 zEpOM;r6oQ@R9IwXjGjF%E-^+&j5WDR{jH`bHj%<Z(z9?Q3aPt}Z%pFXkC zO4|se^|`*YL~5W+Wf|vQwnI$CdV8>}gyoS9WbVc6P0Y?{a9p}_{t3-v*ODroL5;v_BaZqmF#VY+pmlHXZVFoO^iEG|j8&wk}wPVexU@}h& zw0S&$3!xH1Isb-X5+z%ydMQuR$S=Q;DXJ3J^eHB-O;O92a$O|@uri)61qi!r;Yk8S zm^f%dM0Sv9%ykA6DLw}hS65x1Mv}3O+HK4!>p&;bO6uJ!{yx8b6Qf^&2p|kJ-nUNw zv@;27<`)(qa6S(x-JLP;Hwx_o1<9pQ#iCk#d6>H?oA1aX5{o~5W$(+ZkB8MeNvx!4WVAST!F z<|_7d%tS89xLvD7?kstpMTkU)AXei}?-vS7HNI+SidV149{=QnNWINk9Sr(BY$yXe z$Hf%AQT7|C^_?nRm~E$`Ny|Bz^iC;V%5cRKrednAo)7tV1CShjMATo$u(6;T_t<$J>?^?}fhYL$0x z-o2kRXW|1*QzpD03>ZnPjt*bW9C}yKXV#>;<=-;Z`l`pKZ6u=g(oKboXPkW$BuIJO zvkNz*>yF0QUv6&i&KzYLrBJ%ph%MO@+lnZWpp*nf5hfcK6u{Ne$-IiwoBQUU&pBCA z<)B`fG-;Um8uMCh(;-$HxI$i68LAd=n8I5}X^sjoXpqp&Sc~I~XPm`2wZ(?Xih}c?XI(+0%7euo;mi6g;p$2Rn(JrFDF4}eLl6N%h)Tf}C zBo`iAEz)u^)XhP>?4Uu>N}6KXS-pjD5aO=qi>r>wiz|B*p9QLX zDmGBI{k%3!Qcv=<-=0At9hcVZCe!$Q|`6F!PiQVNz+Y*9-Sb$kR^FmHt)?Mp*F2Z(L{3;2`7BDBX=^vfkrV@(b?m9rH4XSN zmm&BtIAIO+jFZ?ZdIMeM$q$7LeZ}B>%--sONu~oG!9*iwJ)5Y(;Z0lKE)LDA~uVirj1SVV=jl#Ufh^qZ8*>|$iR77rf)tUW`ww1KWLZz-G5U%5E zb>iDIECovd{A4z_ITYFv#s@uZ@;iq-zReaZHe8^_D^Qli^UM3X)}MsFp|&>eIn;v8 zELhC$V2+^^?$Y!v;I}r)JA!WI^EMI+tYaVoL!C`!lj{v1f|V|Cf+G%-rHBkSbvjRq zfeQ;xqYwj)>eOSj{VhIfOZ!YTp{R?%ee#|gikX${SSF93)H0NRAHj(N#L3;4R{^Jf zw$i1RMA144(MAK`=uB0B;lglz4dXepZ>*$o>2%su?(g^Fcg4^J)DA`%wLKJ)=}!M4 zkaqCui8vySR>aiLW7X*RSC;5}(>u~pjp0-p zPYxe2EuFUXo~vJ>XUHTr2=rMA(ipUKV(!bhY+tcV{xO`6t-_HnzOVn;-~P@% z_=N0_JhRXH#yITxTe+MHH3(7srR02@5lRMDBAM@PZ*L9rY$!-Tslff0)lHD zcZ(~eb&unWyP3Wg^LRdbcL}2gka9t;22C6d5KU zXBJZeosnil)MCF9AK5sH|C|t$`5aj$j*8%R?#1pJUV8*kcWkq9JpCbDvD!8;@@29d zXlA&8m4(8s{4I3uibC4pc+jAp#$}mHZb^hffBMwxj=doL`b5)N8ka>BuZHqMpu!jU z)jh1~d|s=sInotqO1~dnt8e{3<09}u`(vBnk2^OfE6N-g((&;|XkAX0<(?zcmSf*= z&*rj7)Fn|#ixt>@R$kS5fqL|M)RR79&+QhS$c#Hg^U3Z2EL4iAU*X!R6BkE%w|WgK zH|vbevASRppl&@%OxZu>NUdN1ntyN~QRfCdXpOeRF9c>aHyWZG-dRI5?*D_kJ`aoV z+n@Yfa8~z(0zu!!)_HuP#u-Z?ttndVmO_7q_+4mhE5bCRD8uNR^AuR7JEqNH?7pAZ z`*FD8IIS5;q@7AgrF!lX0H(7WK_mXX+jjFzZTC+mm4%uZY!Ae=T}E4-D9!?dtmI(^ zW2%3kI1vg+HFgWt)%zm5Rz`QwPa zve~5iYXU|S7^YAYz8j})tV)!+Q8O8yKYf3FzP;EHi=%W=+OG)?J05W*EPuKUp!1lV z&`R@dVaPU-h+jb%OG?*ei{X7@{q9oE$!s;95p#BFcjhO4Zk0`-UXNGzn014cC^$sv zYUwe}c$U(JSF1lvJ=%n|?d!bPtAxoID}q@b!S{|tnfkQBdW28^-c;b8rcEo;@ooa< z(t5QwkcCRE+5$?`-J_s}CY?eft#=?|nk&9y`i0SAq}8N2<4VIp0ZT9!I5i6l0tC z6toY>NZ)WSC{n~YtWN4g?qjg*Q+G;cRCCJ6dyHfWCmDpu@4etAPY~uledJ1kiBKSX z0}!}Ab8}9~I1BulQ-LcSmXzx6&VsquNr)GL_91&&XonO9Wzs>}h_F`K*{mDt99X&0DLDoCzLo!fP zk^jA6`GPp{^6$b6q-C(X8MSQViAi=?%emg5D2OLqQbb0L^peU?D%wsMkTQ=*w^bDY zC@m+H@OrE%o-=xFT>5lUVi_$e@e7_e@e4LY3Xa4MSiCny~|Sq}pP^-lxP zO;s6-#Bn+4T6%+YDOk{QK)*Q?CW{pde<*Gr4Ret#NBuiIg4H1G@!;z3^R=dv)rO)V ztsNX*ujHj>$KUE6_Yt8e6x}0fY~Fa!+QJY8-y3eUZX8H0X+;*_Sp8xoJ4frUf_mbX z2zYzs0#eCQZX?;OFFR{qrD?!L&2*m4hfGAb%RQ(p?z$qH@(}a1U=;J(=t4Py$D~1# zqbk_QBO12IjO}XApMnyw$VPO+D;tRbcVms9;wiJ5=pcRtAL*^jGR?`D@YU6zrt6pg z{}N+EFcG(Vo%JW&z{7R%CgX22Mi$&5&Fx|O*tE;rzI+LnNU*6>dVBTRJbY$it0p9x zEhbE(o|C6$`i3R4&h*W@qHz`QRBSFJi&<@aC+t#BjLUO2j55B@R0jiflq!ImAKf#M zb#M4T6y5%I9-&^&|4YHB*3&bGWiK7SaYpb#U+K>xYjlrG?A~8ry?4i(d1&?%3-tge zh+*x*^&?Uce%;+EQw&2gS;MWe6sauEOnIn7^TllL8)9mNl+l#j+Pm>q*W*u$bhp&k zQNOr^E;ayPiMSakU{sN)yiF9u)TkfLICbzb_=9rRFws06L6;@G=Q{3JvpBQ=ae@z> zIL*+Mzo*cVMM}a^yFiJfwC7gu$~$D+guN~8zN9{y7lA0<`a!iX=QH>zM@yEhH>+#E z3kw=}yx27tg>8pUROWwPD4mLCIRF>El z`8%%#ca-JDqdA$>Ym8tRjY+Lv&4>LH_GrL2hKH5YL;i5$uSqlhCJ5)7ZAO!kCP-ctXjO8(@s5 zP!4|;Ff{Jf|DLvuq1;DFcub`9E96X+0_QtaMwr^5v>qD9wST{G@>@3gR8!}|C+m@i7=E?Ro%>P*A*p8pdXaWQLhjPvA2EtLa0 zwC2Ug?~*ym&p&M2 z;U;m`xUJuxAOkP0Oh!&gAhh_w`9rfc&Z{sL+%6$a4AssR+|6F$4~sK3UNJei@W;)N z?oOOch-mC0>c1FbS;+$mWGD>jBrPxmMnY<@IlQjvn;o}VN1r9pu~8`${L|QMZ@Due zx)+;+31{xq{@0Ghb(I|LMmQjJ{EZ_Zy+0!*6@+JBs&U1dB8f8<9JFp~FY@0P;Nj(G4FJjk(Q%$3v1cZT!NL!vDgh@ zKLe^&g~qfeGFQvdK@|Zzx0lFpd~nKkB^}FsJ*{(~kJ>f!u`vX-w6W&+9CyV%c~@oW zgm?EeAzR@PFh!G*sm8Bbe1v-^;%c}H51QEH1K&3gjgl?~gsM^0#xj0h1?|oC?J~6E zx>L6=UUTf7Vk_-Uu*6G;X=TZ*`8~#h>#WY1m&l=FtzEN=W&p$ba29B zsSwDDQLR=<{0-Gt^sqcb|Jh|^+gs}z?k`Hsvd31iWC zBnI2c++QxlXR_E}4TnMhW7|HH+nwEMCwk0K_gSjhR#e>++veZ?N<4cRs7} zU_3+txNqCu?r61Bw!`Zt+%X5_UA9`O%Lhtl{D;S{3E0Mb`BqL=?K5jZ>S8OFl0Hyf z&Ft7zJn@0b9PP1qK2v_kwc1-o6(+1t_ea>GXe3yOf|@lwm4sR`DpVg1g6wS;OVuv0 z6zQ~kL>FU`Olp8hlJ&#uE63 z9qU^CYPR+@@1u-`-l-$!q8W%*E}lDvFKD027LW7SFTL~%ssPC6yu7@YpWFj?7s|Jr zZfUZ~Xvgr^c6Km~JbKTr_*a?ekJG$HhmR#jEXpxD&R-yyKuW_kc=CDP@-Xj;|~yz-F7sM=Y)qio1q@nd7VWB_&$i~w;#<>5e@ zzcC?_$e%akJERI``x=D~W#58|g`Q$Zvbk1!5=pdSaIshhk1M1%+Zyzv`0T#Xmd7#B zHBaN{btniQ7F;NAu4}iIo-6*Q9I>88cxdCUgL1IQ-Hh4oHAL8%IVG^3GB_edYx3|C z>j1mK@HPk7lrqMPx?}GMsU>cRsM2)}qBp8o8#|+D4Pwom?zdjLl$96En{BbP1bs?% zP3|E{ihjztTdqGhn+j0}X8y9dSKo01Z*@C(uYfBj09ft$a%-+t}h(ujjO8S^#mIEbP5A|KQfz%hro1i4I0jGbjXF=yNl z5bF~zmoW<#5se!s$D+8E7YePg=R&|iI+UwzA1HO&*x3VR1YJ%ErFsNB?R~o35Zc~P z@-g0no9rys2xT6oT0j09{Mhflhu?2)p;w1tKnPn(#nu$DtZd5;Am!Z_Ps2&e8i?JD z^o24|O#Exf$j($Q-NW6&WQYBm$D}E&K7_CB>2q8r?AYHVwTS!?3Nvyss=)*j&Lyc$ zDdh64EZQCyg}%0)3oIp{Gl^B+4?zOeaq;klYaxd7%Zm4_tRBp*nN*DllpYfARcyt# zR~~j|OFzUlY8d_fvvR8Q0srcwceU+m^7|*r(E8KC?Y&r{SJ+JhO02ti`60K(ls9? z3D~Dv(@j>Eh?>X)Ly-g_Xf+2D0(c>hhell8=2NMwjo#KzoBPEw0{FBBNP9jE^=fsb ze1bwUbnIDu^je)CYUYVLSd?7U9od`q^ zcY?zyzKK!quV{*_93~yM&>Cp4q$@9J%$BMcXt*woyK6G6QWyLW2C1^1wnjEjL_uEU z#bvcW-ELJ_$M83=bZ9u?_4ZOPq{?!tBHMQP><$OLxg%-?f^}oiZI5-=Ch{e7vS@X; zN_uFHcszqNjaEHTqPMY+biof&jqc#dW7%tIwRY_07Z~dJDapaL<*+mA!g_39_J&-V;(yvGAUo^p!o;Ogr97N0 z#KIdXMtSJyQpWTBL)|x0VH&{jjs&>(y*5BpT&7ru2}7QxxLZ6ga`x>^&Az&kw)Xg` z)hy#Bxzn*#Mk`SqV2UhEp$U@(UO6Z{n;&bEUE#`tZ~u}LGm-ekvL-^s>JV%|pF$s% zaOds=MG{1rf_feO`_RrPN^s5zW(TL;!8I&M%4qV`Ya0upQ(o(43X93lrVp7aoWfT8 zbfP_(;X9=`$~TLLeW)^c#m0?9BP&k0yLA}pZ~E=R0us(P;xQNi{MX+%H$lKq)J@11 z_#Z(90ayynCkt}2Qkiq){Wzz{b<+g`Botc@^|p>%hk||{D|H7P^}G~eWAiCM^Payx zx#K!S#P`sOlF823)VOfFr!Yo0HE z{z)3%#CrK_I6RRV?0#(Gr0q6xeWnYnZ1fJn-?9(Hv`)Q#kfv$5eb{IId5Hul76(C3 z6{|%0zE8;3NF$);TsQNm4V3%UcUCTFhLp#X=B(Qde{+W7%kVeshrn*gJ9RocC{Tu| z9FMt6+}&$b?~`Jb7$_DBh@N^8wiF?X=JUlQ0r4#|E}g?)bLk?OkcPO>Y`aalYfbu6Exak%AhX-!-R z=tO~yol785Ra&)0mbHNhC)OU%z4bs8wm<2o>CAbj1P)bbFp~nMU)$(haWokrLFzh< z*1p%Ld(i62t#6O}I&XFDN=xl(C2{i1*nw)nki0e$^#2@Oa-tC;C;{IUk0}W&>2!m` zhr+YybTF1!TAq%k+qGeoTQP78D8!J zSno8XyVswhl?lq7I(;I;UeZ>$YjKPz=OafHH5aLID8I8t^?T*a_4B*CW9r+~RCzXC zFOVbQeBqFmlFrQrohq_seafk;?w)@<#T{Qj6>i5;TmPnU#;+ic-*4^$m>0pJe;{xQ#8n!itdB$%mcL#oJ^2P2a=@V1Z_^X2-0AJ7eB`_)8V zzgS>&jf$-W_H*MWJD?1hdj}9j8ajKl#-pU^ogEssR1edr|CY7qq|?LGz8fmnyt+Lj zxJqTI!`*uILN&lnA&&d)JSYI8{A1;3!voLDJk#i6&%<=Uwh--@DWY3h+vz~Ed^O_E zr>QxkK`Dp)gr`lUD{p`aLJFlmW2}1ejo?WRVW2czUl=N7#v@m&A*#SV+}?HmZmZiH zDwWr4!|tR^h$Gq6=r5wm0cD-IhB zLHBT;)Wip@F^;%{T`U%~Uk;%)w%9$!x{f{EwCe*Wx3l^C~8i27`t3Ip)%?#qI4P;HE z%)UIyonXB%8iAjM;O(YQbrQF#Q#XzhYT88EWgs6}C+hQ_kQB&^#v zU)fQq&Xle5UH!6|xJxUp9+3Yk!oOHA^77mMJJDdijQ=$$d&b!DVbpgILD*V%E4^x0 zi6fWBZy#FDrq;llrQOMGA6QbAi)0aUH8d$yF!p&mFz|(`U9TYQ9CuVU zSze;o7wffa-tiR!n!-^s@vBvuMj84KZsguaW@zJU^>+(yA9)=kA{_GK^zV}SJ&O& z^704c_P00s1cK=mAVoeIc_hPcotK}iLu8~9rTw=pV*JGr`p3$ zBKy|S&1EV-9WJyAawtgZQ9l|uWInesJv&uQ7vsm~TA)J3s}qkO1G!Nhm5b6*`&u3M zp=`D*$x@?xJG~dM*HdQHG$^aO`^7lKRO3DmCD$CvF-oWw=H})FvFoC{<9lA3{1FnB zON2T)nK?G4x*;05@^H5d)MK+hX`@H|DkVT`#ne?({hMzfNkuszk?g3_;~mwq=L)3@ z71$E&;J9_M@9G7&Il`1v(2#F5|El-=Vg0OWFvqY`V1A@5(KIN^Zt^<6yhcno<#Zlq zcTwYN8a}D9((U1r=%`ad<#=~Ah)sq!YuD0^wfcz4hW6O)VvmNlbGREizXt@;Tw!lt zFCW^rJNmv1>$=Ux&e7i0M}wTi*S)~wl?0qT>}hm)x@L*@b>#bMY#=drRp^5i5U+TJ zEGjxn@9k3F3VW=B`8bo_b7MOqc6E`vcA*Qp=V*ggn_1gD3O&>8Y1p`QoQ-o`Y-jL7 zY;qSKP6_=G!#}$7cZ~@ZgbF`{+aA%qp}^~xkM1St- zR-u^L?A7w&dI0FrT`H43CgdX-gQ2Gxfc!au%|Nx^h8<6HV1u{3NV|jl=5(u2EMy{{Qa;(M*&(LU&b)-B(R3+wlKbK?7{pbi zHCkx9#upH9(NZ5_9@o~G8zPn5 z5p~z0o8_TeVV$~u=9p!EDlrP(G)_=9^k{vTI#wkEUnIwjfAc;mJ5SqO0)@g%&4+8R z4*SQjwW{04?o-cYjiFyh+Pe$&RF`pxzc$VqGTAdUdzvcX;TCOpPtmDDGPEImQrGwa zLu}bYP0m_gxE%vM>s!~HVH1i)fnBw8E*FQ0 z$9&xx#wL$ROqr+l`?Jv=FBt1q1Ehy{p6Kci-(y}H3zO?CoHVMG(ci$vXIs?bNzvaQ z@}2Ehi-n-Ox{fd(8G08EP;=cJZY5UI$<%tZ@k9)vlU_Z&$JXgY>3hGC`*~NR4wF4? zxsflBU5H4FVw`2}=R5H^mfLx`!ie__6@M!4`xl+w5XU%ygi*X&Cl~mXQ z=Qd8xDE$@RxR(VvajvJIe!YM6NHia@R4i4Y=`40xJ}UDuy9Zslb)_T_?Wz?r2fRni z$y@Wx7OK>~|52e3oF(yJpNO49sgxRU8!XgcAk$2)?_#iW1?;a>`5wNE_bd$me|)`V zR9wr}H5?Ko1b2r73-0a&3-0dj?jGDFxD(vnT^eZIU4y$jjl7+ld+u|8eB-Mh-DC8q zu3fcv?NzI4%{8aQ{ImPz`QWs_5oJ8~tEB{J2jX7HS$ib_ghWW~0&D$~H0V;kZ#rK} z5H%=h(zk(FI`v4Y>a{~r`BUMA^aAboEaI9i*t)Yz-m_rh#WxRq@`~ey0-LpN#ehz{ zsEJT5%pru+LY42!4YJKd;8z6*fZ9V!6o0~5svUf#=J*YhA;*s2hD=0EUaYUmQ3nJCSa&Yp9ovW8S%(&*jh zoWY$F>F#x3{j4utiv6-egx$&M8%d=iKa3H!?Fv{(rWWIPCEv+|I=u$qcWk!#F2wd! z;vf2-ls3o!Q|4Q78@Qs4FDvU-eeV1OM5aCHJH5R61K>dFj7Pc_?6||7`x^@$%T@QB zjqYgaY&eCqFwKWnHaI+XtGscA%6K!TcbGD3?n7Vc-`WzLvUMG{7fno%%)TWC1D1BN zThAF1=fezVmgO2$UNMb{oTB^GBIVpNqHA99~V%}x< zT;OPDHak1uId&v>snDxvJ2QHo6nlC`#D^f-4>AyN-oJCLQpG3ONLa2AtJ#VT&|g8A zv<%nR_VwE9-ozj=Q!k>0t8B*gy)^92Tk{qB|M!Sk`ZJdy|GvVLNZ>I~GXa7t7k{0AzB%&YZTzxb534 z+7O+ZnB}F_l^HpL-#efYm^S8X4S=nGixrVV=Dck8*lXVmq<>RIr+ zQ#&_Hkh<6gNEaV&kcz!Pf)w6vx`^B+8b4pj&G7}qjNgGB?zzdIy}Q^V`{}&si}5)A z01#z_*qyGnUzH>Aw!J^ma+6giJq>cHE>4Lu#f8(zwj%-f^=WHBe+ z(g4u~AugmJ<1E^605|Y_jiSec-pF0LeR|Awpw-k8gD`5)NXf;CH4W6E2(P}qJ<_Fz zAQ~w5(OMyM@nxiot|S2#y_oGjme4MewWZ=;} zAG&SPKI~Y9z}k*yowpZHAcck9R)hKkTYo;z)HWd;{j||`=}la{25@%oTsO6jQMu9e zOuUB>l8j=tMD#6@b|GbTp=A%51^#b_c0@kZy3A|adG$>|)4x!v2BZC!T_Mc)soh9IhLc96nn{xaxG1Obapk79Vp1gxXuASFRy!sj*qV+0p zC6;d8ombC&g$bQrCDOo0(O#lKXE@-UwRvgKPGXTuV0nYV%?aHN?F)3_i=YPV2aB5#}OOcN5v9-^-U#Pk1whWpu5WLDI8~hb3 z0-mgkZg#Gx97)>OYeXU>o&^#Vce&4K2kQj-6FSuFPIUC}}3wGF__mDQ6BN4PFhFRte6f*wP}UB=evWx&abYwWh^tKmj1NvFm z+0j^rpB(>||E{t6g$Mzf-e97H8K1v7rUNF%@vh6825%`Ycc{IIi$3doWhjWgOH`@Y z(={^S6T6cJDW!1DXlKQdEXWD(L#xcDky;z!y<6X1Pe;l*avB~>0|(*afN|F;CutVX z8eb$t*|RP7l_7`h5~Bq+9uc&+F2=mLqhg`Shcq@@(L4AT6)7Nz7Y{Q(6 zuk4UiFQ(%Y1VEJn8`r>xvD7T;e`DpL3`cG*PUjl9UB{h*W*eD?HXsd_Gwu)qtgH)0 z0rtpsEXDOT-hm_eijox@y5!Nd8r#Iofr=4SRf3)Sq@`kc-l7Sp`R<$MtZcz^aiDu1 zcA&?mlN*iqo|Zzna8yzmIcsf-PHZ@~3VOM6Dhs)4z)d?p&rcu8%8Q=WIq7>CDJDRr!qDA6yqbOK8TjXu_!K8MI+Q9TW) z=V!$+91jE5Q-jLm(MAwm+p80`&dKLd(wOthxfyrG)BnZHOgQN`%=^g=jO#`3hjx%_G*KYX*$_SJ|?ZEAyiygx{_@+@GOhrr8Hh8G0xt-W+dWt7<~C~g!$S1 z2S0S8v%;(DTGo=K$qbigq=51ie+*`_rzbc&03Ngf#kt3BCFG9F^4nJs(WnE+w!vb7 z4--ImtzJ;*v?5Ql?(svhNY>uz9=GtmMC5K!u+gx|k_NZQzfFHof;H^Zl-J@nHUf_R zL&cD#Mkjddh_CFtL{n`JK0WX*FfE*O$2}%01fE!6(W%eo{ft)2HU1?4n8`L%`X_|+ z?%pxY6gk?Kq5I(VN2>kvz=$8}WV z6Qh6e2ch2!$nAIi;L9?Hy$1JgvyQnfjV!Z9hnl{jFOD3$OOeA+D%9*d6Daio)5 z@bOds+2GEfW~5Xc-S1!T{8r%M5adAHfyb`ulp|9oIJ)K)=!odr>LG{&Boy>D$~w48o}xz;%{xJUQg z0cP5Dd@t<$pAQU*!93L>QA&b3vO-V6r^UG(df=oI`a^+$d4PKDF@NQHC!nItbg9b* zx&xQps5{HFa1JGSK~lSd(m3%9T}}j`VSd2*p{eZJOr@5Bj9MvI0g5|Z6?-^r65WNO z`nzUG<6L~Yiq0?K>E0)<1?z$CmUK>sbwBR=_LWoQW7C^LL2b`@swyu9Wn0MbLW>ki_`<em# ziF9-#f3{|C?i(L8qHdmX?&D}8I7krne28j@HUqtazC<}^gdI`XMCur8wD7WQk(B*$ z4SONYNEy4$fR5J^gVfLf&tRGKv-LWaQ&x??xU)j~v-j}^ks;vj$C}u&o9`y^y}!+} z{PlFcrbrwNY~L*=ETzsuv;YAW_JK3ewXOrXDq|~+_B!FnrL%-D^tQWs8pAQexE2}*Xed=_XJf8~Fx%dbbsM%m){t7i z_&HIbuW=BG&o(L@%%xw2C1^~)A&D+&_3pTd3d!OA+fgJ+5MoIFIl$IrgHi%;HgO(%bSGh=eV zSV_LmHQI{p{6Z3uWD2@-7+M_hCs&kC!-yc*5}7Mr!kf9s#^#H;UEas%v>Uq^yL`#r zqm>LJ?L0pkXd(HE0h9yF@Yy|72W&L=FFB$YN1F-d0#e z<0dY@b@W<#dHd?Days}Ib&ABuJ1SP0cOrM`-e6J|PZhTM7F3}+TV1gqfvnX*w>gdl zH(v-nOO#&e4?4;>JDdH?+?{?iG*>GzS_Z3GOPCINhN0`ATSAo^^0n}!m?gUaEgyV*k*Cy z^A6O#gPJUL+BM%OLCbtEm3f*j-#d!Te^e|VW^lCx5BJisQcDaMr;0bUs_#5xgYOT< z)(HlJX-@t#sVPF^f^oMonOnqVdZ#UO+^4>)+ao&yXGuk>7ieb}(N6&h5XKKRmY2A3 z%LqmmG2a%24}z48O9L@*m|D;tw_m?+i;#GCY{Hkv^^FeIq&4@0m@0I|(d!Q76o5mr z`no5)`= zf@v-sxQ0<+47L7H<2v+92N%nI9h7Ws~eD@Ub$(co)y@5G@vT`MW@mW{71?Bh3L`LMp4|4hWKs1e;==z|Z`2u`mxz@Lx|oXns?{;5%m?tR zxMoZ>-FKxSPy0uSHOa2??%fS~M3FUH#dkzo#!oY)j+B5JkSqr*D)IZEX4!v%iYS6K z@d~9A3iA^u-%AqaH+Tutf}}Am5Vb~2YZ^WlhK#S@qJaO&0fFnCX|F2x6+ZqpY`~Rp;)+_@P{;N7kom>LOfFmjUe}dQTh;}~qPo1yS zf9$1%H`k25VeImTJJKSmwf!dMQQ?j4o}+EYLv;B&?ELH0mBqwoa6-+uMC2lidQA`D z1%2e()u9l6N9TR|dO8{Ba1v?I`F&pz96r}&568A-#LT9U)-^4Cw0f*V4-^|YkG;EB zfFoEA&i+x09N|*gY@eyqfGHksBkFMg63=%MDVd7^YNU()sr`WqOMRd<=0t_^v$b67$KZ?zuS1JX@F zpS93k-So3nz3Fr!;$UfQ?0p^aEoHw%&kn$+I@Uh3r`E%%ejnkqG>A#t?41=1&snFB zc0&nP79F(BE*srm`#?vyX1pjnX1d(`<>qiX7+P6Eu3bU1F#Cc?mm^nrL_H4G1Qdz2 zb&mmCl>48+^Ow;&QuCTSsg2f%UBHUm7gyBWkRz!vf9#=ag$97z{F%nc#S^DLf$wAj z*^q>@=};0zq=kBLq`i>6^E4Bo`iTa!$D9hL3%z*U`Cu9#-MQ5QUb7g7nDLS!^n z)&XH87VOFllrIhR(us`4AYq4+4V`+_IbAyhvr~(Eq{m`viPQ(S6fZk-;VncJ*cATm zH6Q`w@=n1NOW*ORQR`!=puH!A9g)h@*;yUp|0aRNT^?m@Kz z@1aeSg~&DevQW)@o;S!9v_M2v;M*dLD;TjlU-WI$3%x`sCMnaC+38s-$%?1jT`V~i z>CBnP&i>Mx5^js^yV--SnFxu#e{9>w`$9+Ooi1rdoRuQlTWWN2->FZtmP=sSIUuv+ zkSJo7YhOQg95r-D6MdJ12Wum&H;@cWB85tRb55uBHg%z~r;@b%URoGmF7R7my(b4v z%vpeb+qJbvgj*ky6*qFQnYxcG-EI2LKON{H+TdskEl|g~T5IkbUqu?-z*e4W!c6 zs`=eO<1AOsjy%27@?cKloR3hFN_pxFpSSOca;-Q`**&A~#|>tQS;rF-Si=&DNn2lX!D_(tvG%j_F{>7%`eic>@*RN(GBCweocf1E=2V`$pU(FplR;X$b#{ z3DOva9){I(OmF08fK>)>`1)zOT$zKqL?VSuSe$XOTm>I3H#da z#zrU0qDht1N{^y?tvOPNZ07cKq9q5U8B$0d35Ctc9HW(Wxd1`>$=pe*$zl?st@hx< zYo?ic7Z=8fgnX_Wti=D9BqnY+9$Kz~$7q;9_KGVPe?{%%WrMfb;G})WR->J};#Ved z8QMOg5h_h)kA-`peWm+5{%9SRFiM~I+8oMd-kT91Bxv}gR7y?qikhe!FFx0lcsf32 zzIyr?yV<68HH_rk-3ZjPy&`9}-YKX(!oAKai;t8K~W)acLj3`_g2dQ{+L@oyI zr1L08no(poLA6W>TlbF!`CJi%8Q1z{;y?x2bDM0{y8@Y(qi1d-a%$Jit*4n7AHn9~ z-f#}_>7R9GP`EAxfJO37?J%+TnV9sMZ?&3g8o}mEw6q@zt#uBFdDf5ubX(RCoPtls zed{9pm1h}9C#hI5^#_x?LWa38g31;OT~2=~65z6!OE-O)QExNHlt#Ca0;3=e`h(N` zZ0;sK-M4UbaEkcyUsA&RB&d@_nz=u^SKrzA+Z?K;ri6k zG|o)Q_yLA|Th-;+GJ%}P#z0G1wxPgRoMWXD?YZDNOp1A6nleUmPCA z2@H%+Crm`PPRL58a^bv#qgWePSjc1W`agpjS&EfHo$LXdS2F0DOc-UNGtu7IY%WYQ zrsD>CktI^ttjM571d3#O3^vS7XAwFkHB@$OK`J(@keBYCQYlQCsdXky*29p&NC)Fx zx74xxQPTN3eP9?*6wQ!YYxCPf_Y=};%PoyouZH&z2BuM^ny^#NN2^6N6N{GEQ*PH< z=JJ_4*jcIPykrAbBO1-Ln}KPGmm){}E6GT3=+q&lY#>q_ql;trlh_*@+U&mMsMSW2 zJ@?tgRMGdzILM{psr(f=a$2?i`IjON-#bW9Q>D>aAVhshY!)_(JH`?**ZkCTSN0m8 z?g7tJ%PCQn%#j%e}7C$iTMN^&C_5I6= z(r2N_?~(qqDrz98L=V1})+c8}LZ%M=C4M|2JTq_-sIh3oL;uQJH*srD$8)14`=6`$ zeA^E{rl0)aA7qWfbzds#5MOw`HI({xftS0K?(9_VPu@&r8D~5p`}(Y(@o9%7z}k)* ziXa;S0e$NQE_feu(Z5SxgsfCob)Jh4A5Zq45#)}0^ZG@ZWy_RCAc%|haMqIo&7=yv zk2ka;kOy11x!}(d3A&a14qtjjAm0xEGkN~{?80Di&DvPTR$sY`r~Zi84XSF1E(Q?P*%h?kc2N!YJl^Sv41Aa%GQ~6Q~p7oy4s|QNvWDlrh7^ zGbWp?+ue2_KC*9XeOFpm<>0$Kf8wU7;-g(ZvE(;DUc&hVQ(ICG!mp-ix8cd~w zWs~>u>hNNZ5GIxQoqlKT{`i1$Hqa95y*rhM1T#hFTTYbx{+ zX(~L)o8|NM6}ZcVe@Pntx#^cPCAs-ul$kbA4GFUOec4JHl}euJO@m7v7Ox>(34DK9 zS~`0%8Gkou4=wFKoR?mI?aTOX`!lW9xE9;f8|<*(F%)UHUvH7%cRuy+7lB3X5~#9O z-9Nso3IS_i0Btk2Q^;DG>U}Wlv{ik#dlm$=`-15^M3|UNLr)jHD^?nqwLTfU6usDm zjognFDClqya`h1y(y#SV2{z^24&l#m*aibtkpS~p;w(c$8SB1FU0R%D?yB2T-+3zS zpW-5P)oc>^;LLr)dVw3i)(~jojLC(-@0;uXa%x0QySyL}B$>?Ewj{zf0&gT!RK2yf zaY5DnaD23=KA6Pxy}H~JP=Pl=lE2n*mc6%A_xosft@&buHJ(lNuX5)GU(x!-z4v8w z(S)Hdu0sQMZx<9Q_I-fgAz%AFQM$%Q6!QPVJpWE#AFd7&gE%OPpEnMUBxXu_*4Frg zYu(p*aM{j1s+2w*?mtNUo{SNwC2KK;h48mmygFLbwWD*~!NMA{RbpWV{S+$Y9J`?N z-MmY7{sDg_nF(zcr+-A2gt57vSoY@|%1ragldG6o9SO^M)%Muy+==HVNXpUylTu(*H@Er9)yy@Z zxP>xO!!`clmA_uK^Fu0aM!pE)~BmBK@k@gP22HYWIID8>$Xq=2K{|)620H z^mhSO2#D4@$(p}`Gp{+hi^_jX{MdcTVayo|RW7G0@CJ>{`R+Vqgq>W%H72@q+2}oQ zMfA9Amd`!vIlk*Ecn_(OQB`3q0II=vdm67M zx?WGlDZJMm(rRxuzn>>};@RK|wW*WYLf;wB{;!lKLAM%4hyjKqGSCezVt6>*j4$Rr z1I@zA3X-%9`4w0kDjCPJlJb?Vo18L`VO#YP5#`_Q75EK#o|_epu-GKR?(*_e{Bt^8%=25Y5)DcKO)1JdSjD0v=XfPKm$GNPx?`rKLQ;jR$S zj0OodYFuyzz`KfsyP}1bE5<j^?+ zNGk@@*Z|bgxXea!_zazPx7(j-+;PC^`PKBCr}uD7Q-@1u`=Hgn-kK}6gC1i(^N+@4 z??e!?-!s()@&7#e(5{TchK5U&gZ0(jyW{R%pQIKl1-3-n--rDpFGA*OZL~r~v<7pn z&gJTC$(J9NT_}zP_k;=7_W{7ffIHE?(dk~zpLUVR0$u8!?4u&a}jxf5@EpN-%i(Nw(hZF$gL_u8SJ%^=G;YFtg$ zUphw2wOtbSIqpBcr8H@*5S=UJ#o={iESoPk7&@`Pu~tu{q9wOGNVVmvAomH&g11gP zgK4_FK>st@{v0T?|H6CgyIwo${M~HH5k<9-vmHxGqt+4~WvN=LH;l|RU-@mO2Iir0 zsULV^a*rk)fd@P{f#dg3&Cv0IcK`&jx^qPqI{M(n|G>FVUgc@Xwoad~^SLIB=H}UV zjBQly9~3_p1gBf4)zljHlSIqn9ooCL4onfGt2OyBd%m@ea2KRgm)A{|`byS9d!16a z7prh~93n92655-&=xWTBMrseCCX`Ia%8;no?G?1NvGF`SQ$Cb>fBCEM^zUXGoO9z3 z1&jIs6AV7N?s{8?Lgb;To2gaxo_<6yF5iydrevc_=}kc)!DiJbA8E?m*%YIfTZ(N0 zsByi~w_djJ3&i^`8=6ekHTxD9xq6QGcUY#d8^eX?Q@M4Tk>P4)$Tte>oI`%ETtF&OzI;QuD`{Wwg`gw!BNWb{?%>8}WT7xsCfvBg( z4w`kRk0GkMlXyuC<8RAM_QDu(4<pux1!<9siDO3PS8UVno0t6r zuJ`nA!HhdMmAF-04u}C%1(&|k9e5LTr^dIdLy?U2_xVfIi|%Kzl|EJ0;DV&vb z<~}(W)0%+|o{*FImB40mZOScG_w$aWwavvs5h*-w{~Qt^ll2sN^qTiX(-<92qDKSi zJ^k!n8am~(r_M#*?o;#Jf`kt1Q~u%zUyiP9mk0&#o=;XA90smZ6V~2}0icuS&+Cy^!-bK@suTkS7u>X3z${Fj7@FrHMB2~&e0C@LSI_bU6h3PaXc@ourOoUZxX6>#*kXw%`K@sGK zJ=sF*OLVn#?T-ldbqPf?kqtB_N(sFf)UiWG3{w|HO+zAO>c=SybyQ1RsWEjdY=;yx z=H!yW$zQUlQZUM+g~DF1Q)e4FDE%YXaAV1BFo#{!dX}LwDKwgiF_@mL15L~ZXFX8S zxRgG-9zphn7s49 zNQ2&{Sz~yrw$@Oyy_6ZcxR&c5TCrGzt@rB(#b@42lVHiFz~YVb4*T;=9Fwz*^!?X~ z$83(iNm>n<4wG~LwQ!x{Sm@I)(^{>Go==1F&dBr2)_ss8;e+XVpbwNc66}CM))mGk zb-qw0c_R^pa_Ca@&<+6!H||V7L!QHiM32Nt*M6TxI{ZNLffm%Id8IOiqR$pi8G#6( zMgy%Eqr%+zNztWO%W6Y={uE4^1Gkc>_vE|`t_w2HmV|Q@WWiR-j9vg4ngc}`5ht`K z7fOd@s~?@E-Ue@__GapIp5^E9n`>*Dy(}nX!ud)lh1(4_Yh01|oC+zDS(3?gIT{TX zz4$07FJVXX6%k;P0gFN~7jZZyt%g})SSZCc9*-LXIDb7iRi|1HA{YTL7=v0xZ#;uD zZ6BB2CUUM!)$(sOZkH46LfP!f)JmNWu9MSK))|x}8uj;4nEiEm3DY7?`v5&uZ zZ_AorFZkQmRdkE+sd)-BP||3jH!g4Zlsysfc_Nn*IGO zg59-hyYQ%OjnBL1o&lxa`8KvHg^_Bm7y)g&UA}{wnU>Q_cf)Vy7UcGL2Gy z;9%~uGi3P=@Tsua)Pe@`4*kN>Hw@|3VJ-oCnx|i`Q12N?Lle#8IQ zRmj_IYZb!(3D^C8BGAK}x zQHh}fX(seVV>7~)p1)VN@?D?CVBYdM+Tw{1g#Lw({RejCm9(<-ubt3=@6~IC=w%_W z6H7wV3YFzqG!5`HKhjCrL|zsdMq|-eut9uvfftYudvZGrSzxCJOzpDrw0C_`%OH3R z$Ih!g;!rHI>NKj-4SM6|X?8@SU-#Iw)Wgni<@~wQjci<|B=Xq1r0A{n^$7MK|m1B0BLyjSb0xazh@I-Pv9KL)~(m{cfDr@XfJ7cS`8N0`LbT z3JOX=Og{L>1P18Hu+W31g+ggOtu`0w6cXZG-8t|qo2<}m!iF%sZQ%R!*DgfoZZ!gQ z6+>_Xp`p`=;NMDm{xtJ)N2l8v3FeU90+WUIGJ`KdjS?oDmese?WGmKawPXPKr}>5c ze=bP?zL-M5pDuyRQ-RfbJ5KypS&fpH6h^6{HYVsdUUZnf~6lF8tTK!>9F&ov9XZ> zD~K1|jnD&NbqJ%p1p^uD-@4!r<4^F|XecOCF`sU4`xXQ^(Rva{1B! z+|vIoz7&6+!vB_6xxXFzzjjbwa(YkQ5-j zg7$%~cD~2q$rIVXrvRSKp4VI%l{!5I*QWMIeT?{=4(LGHtR4{{p4-J0iI-3i|F~ZG z`B}65Glz|j%EyUk=x5!nxJ(2@t~L^3x1$9Tz`ZTs#oe7uQ3#S(fN!3&%zyi0F7y9m z=cdlT2k+y-=nY0=UA*1lobx}HWQ7sOp^H{Dnov}d1o5yS`|)LO-fh8Vd4|asdm+qP zPeW^axC(BgE(LXcH_Tb6vFO9q0_7rFv~AxW$6Tl;4II>|GPapuZc2Am$AZ*mTM01N zuVMNgCoTG+7H?Reb~Sp#{3rC4<_N1F@qIYS;|@M23;OH~{X|f2AVoIMQdT6M+~#k8 zC9dZ{-WAwn>W>AJJs?u-enj}xtvV5N@kCg|DxuGGO+Fq?+oR$S9%sTYuyjH9^0EfA zwT(?Q*dqV2)!;($!odMnh0;SnK-k>c%1xDtEGwgyl9CE+*^$>hmQ81qPKh8oCqM;H zCS=NC#5Dxqag*KK5;e~{Kelm0ZqR+LQh#?tdbQ6Qm(_CZ2ey!j-$(KL%^j8;67HM^ z=i!$JYDJjKub!W5z@acT4OD5*TY>at_s#QH2%lQvm5>*Q9h+qP%^cr{$X-0v@ihUx zhgEt}X!n$-E(2_q)aX1q+smh+534Pk@I_^+_N-U>F^L2Sxpb@f$_z)bhV=`|NMB7? z+d=Q%I=uZxEIso4Z5}j@_4!LQxuD^+`JnE(eV$_6`&PhvVujed^s+F<2(jeqbsa{}EsV?S z<<&!q_RpkA!2dhn3)0Y`qd>j28Nw;Dp>&ZNGCgO!X!1e?qj0xBW{w~ZST0$3Fs3}h z3!!TGs90GTa7aVMZnb{c3kMJuTceo%f>1;GH*TA zC}et*(Tod@0s(`ri|qSs@Lt(HNx4#zzMB#@41&{R4mD}*dr zDz#DiBUD9>{jTkwv~RSqyDDQnBc?jdP2q`nv8~hZfR#-Sn~A!Tp)X1)NmNnO1pp#r z>VZNA?}JJ6#{?#y_P^HfIgU$i*0^YyvoOPNxr@L)z-CF)r1vL zz74F=Uzeg;wEx5g7Bc=($;UTCgFoqg%a$u)(t?l~yR6e&jj|Jb>>ZPzwm zuo=9GHwvnhQQ(av*$E62=M?VLyu(eV=Lutkc34IcWN#a{b?;5ao)vBSX~;NQOYHDc z$}fil9n{Q|R{tHu7@VjsJ(>HHuXhoULg6irFxi|7QL&WUdJ&Pr*9c z44nJGrN3o9QNJPPdrX6diq2dS5QhSb<1jCfX~ESyS|#!Y&QFTcPySqaktYXY`tuNE zeoI;I!qr;J9BJ}VJ8=ll$vtk|fBHTR^LHgM7bX_B=~nY^Z*dfXiu@9?Py~h!Mo#;w zC9Lk|MiwSe`@LtCD-%}uSxEE~gVjPcq|*NooXnVBuP3-8t1>2mr5u6~;4Vy^T&=pu z?TIi=JJe^JW=veRM1#*=195Fa*FCQ2$1D^}qMi@6URb%R-F#_06VA>x+<5HW6d(ZaASpA%75ssqEsaM!4wm;mijb)tJm? zrU`iSN%N>s)Ockqt|1g1#h#d8QGK*RuJkq+Iv zdrH8Lz@v$o4)JN#J0~fPTzv@6-rf|$e;??;A7?>6IDRt<1T03+-+C@KRPht!1v@55 z=&0sbxIPqQ?mxeJ5PG+;^Y9m`cB?su>nDXtNfx5#=pmpbNQEfpk7^(|mmd8PC;w-T zP;d%MVCZw{9M4w(27i(Swd}NjGc`+<%g9InEbFlE?TrM~;dzdT_{t3ZqAJ-AN(qaZ z*p|>}m}~=s=#Jj`edIhI1GJEk#1Xz-g$!P}$uqO)JiK!VV-(Hq2w!5wNqfBw{C#AE zR=@E~*Pi32!b@Yv8LS_)co##V_>spS!Wf2l3q5wufJA4>X)aP3KbhXDwImw254e_# zASGkmN?_iblZi28o`C2IiY-ggD&u8q3t?xCpVtfx0S4!OmW4C?HFS5z=sFSF>=`TS#7!*-w}KU}(L z%hsoFz9TAtd~zPpTf4(CvxvQsZaJ?ausfMF*YbXVChCIzmH~K2MwObQyido#K2F(A z-3aLqm6L6MzN3TR>_$|@|L~bYk`P;kB|s!5%gj7O=wU_J%l8A&P|WUiKxl5m*Uz>Z z)qsE(ok9hZHLYGuH6nZvjh46U`x%_;Ld*l9u8CCUGlhNv(sHjA2b9-FTQ5MDkk+~B z`|bP|!*G$0=|`{CtgW6?O*@xn=r?FM{j~zQtrEreYxdVClo^iA0m(!in^;L0oVRd! z8RCZZ7Bl#Ba06QY97dgPjz9>k6Z*}DsgG*+O!HpZtH*SAwoB3R>}$wzWyH&oX_1|* zjz#vk$+J+A#+|lyz)j*4`i|hK!N&t-jk=wS);1zd@DU0^6Y@^wB6J@4y~x;V4IgvR zF_*oco9hcpbf$AR@7&$*U3&j4!7td#W*_RjW?zbhL~=z&ZW3fn7dM0hJ+%xBFRiPwk1+%TC>1(Anat!7NdmIeh?|;gyy_ zPYUOk?M5eu{Z=4@*x!W_ht_hw{NmyQc@H&72f|Q=5$t6!{dfk6@w7JM&41ANHP*%c zv!E7A^yQqTlJJfFP_O2MKxWJascyi~j-?vaLrNzcAzj5lbVA~SpA%;7Xh9IzFLR+n zN^iFRW`-qNe-@aE#q+bSoDz4oBCvgZyjqhJgBO=YV51@inWCMY4yqj=^Ov`^qLZcv z%+isj{!_U0Vy7|29)e}?bJfX(?ZyCuIUfRWEmxtO>U~CyjT#n9l_@lHU8vZH`XfgC z_~dM05eFqR;KFX-?Z#&9^VN%Ux^s`yN(HvQ3Qj>dy5LZ$^rZcT5={f=2do(6&GRj^ z(BzP!yi9WfC(SK*`NU{!KJPC{8nQi;^8+878PJj%Mkl>raoXX|FR%ih8)!wFSVSJT z3gkMJ`K8ob(N>T+dn9wH;EI9{7FR!U6M?$7mk<_-R(@P+msg)kn+!rw0r6I*%*-_i z`z4jd%6#H$oc3MtTaTcEdf}oYxN=+OWO(i#L-*NF?5zN9PL0da#UgJt&Y7{@`!I}S z)9B6~uSl*1bCWg)Mx*I(^rEg{&MJW6WT}L4iY!;4Dr~t#Du{B_Mi(mzh%cyDK$y~Q z#LHHZ;ir@MZ13R0KFe+$LngDn3E`Y8C3UxjPuB^LgejDOW|Aig;L-N)7Zx5;ZMM_b z5C8ju`#?K)KbRUI{k6y_Wg#^bFrShbepE7{3;JyQKHUyssPJN_G>&%R1?8%yYfvQd z;+!^q)7IxMM;jo36RcC-uWn1$kDS-JngBA^F1j}z31dua1gB7IG+6sw@a*O9kLlE?OXqJ%)&?g&VJKp(btqLjog7Y|AH%c z69Hzc*haN{sS8y*a_8abFrX^?-0=7TMa}iR?MPc;Hp?5X2#6lDf+}-~Ark@8>sB!R zjz{19nkghtDU-~vyQ;Q`u!gxIa#vVa?YQ!NHgbK4>jEu{*SS9(2ByYsoQ~&0@tcAJ zWVF>_OuwPcueme?YKk|xx%rhg6nokKYP?DsqCUIf1NGK!Q7c)Q6UQJ+|XY=zHLXdh+j5- zDohVJ_~1iz@%ZqO1+N;3$YHj>94&;QPyw~0KWtjV!4n&XucuYVnsIwrRUnS2&~V-m zOsf2yYp5^gu2MFe2JO`r-~M^febwy-m#Ezvr5s&Ib>dzB@VvnG-I|^}rR=C<;<;zD z&rM8`pxDhQ$`pmy<@(%**DJFw$@~DX{-{{pm-*N#H_$AUs|_b$GmRgYu3mIMpNTG? zX+*e`i8Wt^$`iZDWJ$5L5-oWce7!3B-G?t3El);MSMYeimu8!zdYLguxif{VhE($7Lg-VDz*(f6B5nsa;kr3QNW3s6*N{jou_uuAko3ml_6dyGw=MdEn>Bq!_S4_rTF_`;OfNeg$~1 zts!6xWQfW6Ll|;uHQj$30f*({XYfbxPZ@rxy>DtTmaS=5W*+T83-2Xh|9!GIPJy~# z0l~pmObm_x3cXVab1fmElw}ig*wEVs9lm43!JrRbCul=1)06WgBY&A#6ga?d|1iW> z(tn2tdhdRen&)53RJbB7W}Ty&w|e@S99}2bbeCLj=4zZFA>T(^?4x%pVy4E#k9b`l zr&TYEbnZhS0v+VCq19>S;-ikUJ^2B_4@nk18nJF*QL~LNswbOuAM8ST@nSt3bTm#-Rs;N& z%|uPH=fVULGs4O`w&CymsROkxmaAazzK(dj8okFa&3k5{VR-C@-#WZ5A)9NNvXdT#i?Wzy>C2?6aqvF=9e~ApQYyUw#{WYkB zghsy2h|dW2s+%O_MXk*VuF_}R?ORR+u$)koOp#C%RrcHW{Eyc za8)fZUX}2n@=?Z`Md$7v3@Bo=zVDOoQqV827Ub2I-|uBpKYBzUOWZ#@AgG%SxR*XOYPBghTwD?44^l~Cp2l-YYWo3v1m$2fPLe2!e0$g#08 zoPUmvhTkC~BK~At%qCzku?DX&6uhCNlS3sT_)CtTRbTxnbb27ixYYEIhCkS~y2JY} zpllxYSENi;@TI}8hlxSh&xlj6-W$9=fj$PJ@1`}09=h2MNXRu;cOK~`p0n3jPVkP^ zUCxNEF)~kaX6egwuRYkq6ZIp!E+62;?ZnsK9xr>A$n}WkPW47O?fIo@%7ksd*EA;EeVS%^k15By@1jY$?% zVIVuK!AV`8H^IqQHBuHGv6Lr|+H$nw^EYeEx<@8THiwWucyN~{{ztsSH6p~E&I|7i zX6Eaorj2&20}~5G$7k?Wfl>n8&PIZ7o7pWGNQB{iSl({<4rd@Eyy7^VZGL1_WxHb8 zG$aN`T?7ylxAeKd$VJ&KM~%QYGJ3BGnT@Y7vEZKLNL{oc;C`A}fIqp_ol}$|aEe(A zNVgKV9yMs+DfvE!%W|+zZcj#=o-Jm0a1EIneh;1L+#7{`tepnp9OV4g#M5fh9{z(g zR-EJtkN%A;WY25o*HohEn!yTj0G`m#W*4!V6)sexW9?uw31}P^4crNM-JPHPeP+PZ zu&x|g)1{UDsHnrEANs-L_au!?9z&esaRWWA+rw!oy0}7+bu=QMQ(GUdWi`@JuK{4K z^9>wMlEs~gkw>a7@9ef>R_5n&|AD+i$Xx^GqrTlN*X*iG=fh&qtmY}+s*r^tqiCofBb7SvfIrAf%7wcw~37$zgd>z7S$fR$(| zqg$S)*s9*NsVBS4lPhbIw?*p!3M&@9A4RN+2hQ4iZC=D|{u+I3GBR_5@*Eb`5d=>j zc7wV-x=83GtL`7px7h?sXQM6>(12i`XTYq9B4ru^Y@sHqm6bwSu1byv>D^a!PyQ!~ z*NNq7utatQDr0?U%HGHF!qF?HTXoRgJuvlx^_khccm8}~Q zUXl|<|1Q$fl{2+>vBCSus$2Q%gL~OF>D|jYCuATL ziJLqWkFAC~$p&)Td}RByHNVF;<#u?SRGr!4WY@))n4~`APxDa~$W5)T$lS9nPDnrV zEa*3h0(_cq_HW@VE{Fnkf7itc+>XBLB2n6y=f)Cpdu4+C^SjB;;xnLp9XK z8hiq{CN2|)(d8NS^5)@_;|W{yRhpNgC+EF@#SHd`D-bc+NMywOE4q42OOgz;uxU*y z0ECK1!}P(K)XQ`s_^(aRX8E*`UbHuS?{^*Pf;Uv2ex1F=>m{peu0J_lSA3pJuSK|U zJc@7VIYfS%V0qh z5FXl;2O$FY@DRt@csOu;liw?M>+A*A*p2$#RK9Q`BBe1Tlnw|r{_5lJKXiTIlaKE- zeY||Q57De(+dea32a6W2pm&RHKTK0)_fbCSrDBop&X0$j2z;XqD%!3zorNK-EjQl_ zq+b)}CG{AZa&cWnaj)Kh3G))sIn{0~lRJOXvdWrfiXBczRj;nX{hMGP`j8N;sA0T| zJNz>($VKX82zvW-4@Z-f*~F#p-A}yLI{Iu2s5aGg>FBMU85-4N64HbHsiX@5aoUMy zYoi1!@}YF6yI4@b2AJGsMx~A6iDmoXTHA>%0Ci!;?0ZWx0GA={lbKM z3GZit3EQ=E)OvUTu0*V9?dZ4pDNB;00k1!Be2HD+QXyeeeW-#H+SS$8FQEFH0>-PA zTn*JT6U^3QLCwE`7BQX+(on;PNtRS)SwvhZIBb5|HOKST zS>LgsKuB#3{1p^E-s@KCP{LKQ((-lar=(|s+D=WxXX-hMT=H`hHISI zuf#Z!bzFy)`3w~THNM)@Fc^Ez=%zb}Q<*(+0jg%J6Lt+v^Y2BmxJkelsIb8b@W=Za zy%b}*GI{tFy@3t-&v4Wn9F1HVl2kXtZ(xQ}tkSPDLHo23=t_Igm7pCk5QuRy!~%0J30zbawR^VL`ER z3IVi*V`qKFbiPBd1&I<&tLGu>MLX}encoLp@=PY}#IHN~+c*Lu5NWb7XcuDnM4FUK zZObE2($mNDPdxHgAaUy7@i&}4yZ(EjcZMZD-nnR@uVHs=oAHq{v&rVED~VL>1uy)+ zfvTa{NGAhnipiSaBf*qiz+fc6C%8yD+K5dBl`Og zXtEyln_meW#EU}tcM-Y2oywo-s4)r_KDX6@rf0x-tv0k! zea|Dr!isSW_FDDkmW5)y@8Mh&is*_8i1XD(+>j#bMl>;?q9lU$L7?Zjq?e{ck1ljQ zuOYzKncK-6Q0pX1(0<8KzNjEP65e@lCs)l2iDOsvu2v(tH5zQMhF4wBncCp77wZi;`~HHvaIZjdD_!z`m(S@!_9 z2uJ6feq(2!Sc<_MdXiCRpt3A&;a8N7`&4x?EabI;1HU>eRX9#`(?Sr3$`_#VR$sfs z-`-@!vQ19Ow1OD6f;Vdz7yf4za4KnRr*Wa;AeM&TOo@7x>fXfe(?=6|{J}N}s`* z{OMokYtsHW!>$@CDT9RY9I4NdFBV8B;N)U;Zp41y2{^@GE!Nq&Y_D8|&N*Q>ANl=O zX(H?c@ypLfl~waFwgWfzpZ=81jY^Yz7_X{r*O)*`pP9w3koL9Cl81!aZ<@04x1D3b zA9{2+2xtgcuQhwe_3SyN2= z*&rd6nh>QCb6w)pbCEi6M0_-*eXe5Lz?w|8!nz>sXTwD8d@`(bhfqB{85Y)dblt@B zvyhS}NyYnc ztRK^4^+R2~eFeqe9Y0Ysbw}sNm)-i<&g{bEm=jn-veFCaC@$rbSLVZ4Rw@~n@CFRZ zfvRL5zeV`%y{)Q-O8bys``^reWBQ$BIfSIrXH|yxz7kUi{~CgISiR=DB4ADK8gNSF zV}W1jPqb2fV{KMOc%R^}GFq&f;G&e($%~>{P=tAaeSyULzk&DU20@o0uP7Z9CI-%_ zP9+z@1~gU&50f8aSHNV{fe`f=m|zfo0R`Q~PF}s$K#W@h{TcxR{!Z<94&`7|ydQ-3 zROBPLX%s^NzKxdLhbBm&0fb)`>g4Nu>#dk#PCyE|fmmc3xci*p8r_VdxWI1+I8H}n znCUjtB%`0p*q;`^Yy{|RuYzhDeD0uHJwG;(s6!8^CnIVyP&t#0?=|+xhhnY675W(8 za*bV#J`$Ou%MqqPkN)%(dt*>M8q`bYWa*cZ&ZX2Auf7z z*mz-l+vO%cam$^EI%#0a4RT7^q{*;pow;Hf7%wl!uoRaB%@`0tX)%6vn{M&qZnnBV)=T8NQEWV%!g= z|2rE6x!-!7Fgn@05-&{1B0ju-nDAm2JsZ?I9#8ssCBJQcz*Xt!%|8GAc}diaN)`7g z-l=hn%DA5r0oQjO08HoeT#iC^??1_!C}P(cxQ{`->r>sm^F7c!Y3wS}7v)6Kdx<4X zf86en|4b!*o3|x{5i?DGcJ&I4lkdHc$!c1Y9rP|*bjr-;;k&RIm4**oK^v&RO0_x` zRML*i6!=i~L#vEFC|*R#fnq3hr_AL?|5LSZr53mOYph1gh=a22cVsGXMHjg)tWaCH znD~|}&bz8$6nduuFCT@X!5GJAv14{I2jSOfP%F-#_WRA+hWE-p?3|LXvSs0$xFWnbHCiX5i}FzkK8CXr^NKJUMW+t8@VA_`_#$O{0O3_;}Q ze?%XK(|^f3sftvOU)miE;JPw27^~8*`1l+AKRNlybPWV^XPN|Q+th+v6aG^M_!jAU{nwLy)q%N*)rtM@kQwvLpZjBwqdzTtf{4pN%{i}0O zH>p&67UI5$jtc)H*cj_R9Q<*-8sBNazzy0EZjrg-ff_V%Lx7@6^Mz`cn!Ap(p4J&z)~b8EV)db zyNGga)B0GC{As$_5R)LSHdrx@sn8IhZ^QQ#)P7)ra4ouH+Vcu^X1EbqH4!6y#}q47 zlUVMSuzmBiWt00ZR&%O>f7=x|{9C5F!Rhqd?;P$Awb?fzj+2*5Z9hzQ2}ZxQIUamz zrcz9|jkaXN^l~4(@Xn(h!>tdta;v8lo@W>w#3fbl@ki7;Y{NgN3JiPa`E9F(1Y({a zY(XACHC~U%Z1myPCqAsS+Vu{yBhpW0`~ba!m+6Vtpz8w#Z`CT(k*)L4mBHz*WqEP< z9d``BCOA;_$K{tku1gc|uT?Q`uSb+Z(xb~mNYG65x-b>bNG5QosACad(WUG;ivKt9~#qZcNEn08^z-51y~%3;P19A7OsYVQBWAQt`L8 z(2{qTAt~4g0g__9Nzt2D>K!3J8gNR%%cB;%syjkr1T@6 z&B_R#z{s+hz1U9pJ$d*^IfJBDz;Q2+=%3^J#r()^!{s?Q5}6I{Ptxb&;a} zZJk;*3$`NI9y5ys=OS<~X6e*<*O;#%(3Fp*i=Cuq0BW$7iiUno(_DDgox(+GYkXV1bQ{H&HJ}_n@wK3CKZ)2phN<{ zM8dSzKa@Hw&$$>?@Q#PKGjvYnhdv!?F(Y}M7+6MzdqIP}$n&BV;KeA2|98k|%4bPC z0X>cv%yvhUApBRuPIp8wP#YVnjX-LU-%0=Pq5Im|sl%!-fj6h4D8SiAURfZVC= z#P!b%Y`t*z7WW#VkV<8Q z!$%GPjnftBB|Y#N11dzUJjjE2?)#{%<29DLk4-Z1V;~Kn^C!Kq+RNU1)srZ&spO#- z@%8EtY$7zDYfNMa$2S@%X zQG#(w;9oKo#6P;Z%pFv0iCSnBR2e{!4HjG8jNlEFR>KBwUK(oe5yCTpWVUhf9um?5CyO7W=$;xOq$4X4=`HOlz405v%NXuWbi zi1qivzMEc>!MDdMm1yMPRqAzqXd3swxJ&w!YveXyz9Ok?<4EwRceF>mWGeu%(EjS3X|@t z`1QsPO%Y{JjH1Vwcei>-w*+=^tcoGQk*95Jwms2X0(?zWW3c zKE7NOa?%C!yBZ$;+bVt*Ur+HJASg+W8Rt*^dH>sY*2w_w@V$eqOo0|M6!Re>a2Ano z;YR1G_x&$?&WLH;V+ytYP=*a>Y%~=hrl|2ZqDv8BB-XzI)LM8^>;)oQ89$eLu-1h+ zMl-NiR&Y5Src_ZFNdvYtn@?Tn-UTF~Anb)yH-{R=gV+4S3L5@SH$q?m7z*9z8x z=VDJ9a+^tZj*dsV-d+n%21A3=O%ub2Y08L;RARFkEWeXB)jMB~E+f_nteUS3Alf*N zo`J;v7bCnl?mwnG1P`D!usSp%-+n|$K)b(W6Gwg;<}uT$c3(Sxf5V%^&%1cZfq`Y> zk=`K!eAmT-H>m(W>=)vQ?4!&1(q@^aN2e-Xd*uCtEc*L7C8pq?Z3CX+{gJ^kZ`-(L zD{D|HWBXe!FjYhUlJGG#hW}&n|4T68g8g62b1q{mB!ZEG)7XDM{@;&4zyDu(=>MWc z|2_V{i1DfNRaD9&_UoM<1OnemIWmMZ*o-89*nz_Z{wQ~Pd(_pja}uD#*_Z86=EQze!qJGaBiJ7yGiw_OYDNv5T-OEoeQt*1jo?Zy1jG0bXtN zLQjOb?&1o=cNy;|HC%VN9U|2MfH=5CmppGF4|L)?gCw&B5YhRo>!MViom6~dN=))t zZT?^{CjF7!fH~7(BlIt}5^BUmf1kPr(c96v(nsg6A{*ZG&c9w2sq)3hud`tFscvdt z4`(zBk)#WxvYgVp_}>&0_i9v%@xO66th)IB@1>`Dl498ObH*R`IE|aBR^60H?lp#;^r3)EmyD3ArY}?) zDbfv^6M_{Vo98vca0Z(F&wkW7vW9zUoWpghD_s zfebgS&66d9?Wa|xg+`0CD}95ubTu`p%V&kN&{f&h3+d;9{-rWr$c=&KlYrTBpU?$2 zRZdF_g8uJ6A+>WW#mbXIPDwEkN*`^x<4UM4YH{dr(fhLbOoBB6_FqOKS5bF8H(Wla zHFc3(Goj+#i-ANeYLko~?ZC%w;&v+`sJ+1F%UDgN7(+_~i&wgA0b8_9CRFaU1BY^FJl9u8ANoPDpABg-@OsaZD!7M*1_N5b&)T zH_+$cb)t}pX;b)3+)$3x1;07ymbGECttT8b<)XxxKKS?z>@>Fl>rxkdLM#!kE<+^A?j-uaAq9c=EL-IB^lb^s9zt zi0H-53sm3tf*x81jO~5WUZW>C>9Z+Ynesn>wU0TgL#UPNcF4B#gglr^=KvxS>z}ktNkD!%cO;H~74& zp7gokpM@-|&z6!YSEq-W{i#uhdmcD`Rcc=g2>D4BKVv^caN}$RV>GeBwAoun>Kl|5 zuk@~!*j;pGjo9BKQ#<3dCF1`@d-l4`Fa()5z-)9Y~YWZY5uCmj#eBq$Y+>XPm|9=cmFp(r$1*Dz=dT#JZwoX#Q8@2aZ=W7+d#XnCjv|YTx*%L zV&EyT2-f%JFiIa*vz0`Q%3tr!V&m6%`_>f}c`qQG9a;MOUxwgWZM!0~YPJI_4ctd; z#`B*!#FvSunkoW{%s1h;-&dw}RL1*6hJTR?xmUl%rCrf`TO}}33-wX`RtE)@6(4FuckGaOl#eCVXuf$ ze;x9y4?e%XUb6S&?p79iq*gX62flvl3gNWlvky~U7b*!!V)L)B8_W%3*HWSl^-Sy4 z6>ji-ki&ST`#(@J{wv{QKm#qVE(v#YDrYDDoRvbStmVUOT_oR{v?Z=-%ui6jrZc^| z5zuY=gNE;96n454vCC%@{}v|w&C5{W9b<#X4awU` zAoAFDsW2nzphD9hls72ny749Ry*B*kGXOgww;j3Q;2!zY$M=l8hGqc->+(7ZDn&8? zdYyo9)Cmv6*y;yIqF3DpT@`W#kn@RE`zS40DB6ujr1!~UR|W{)q*h8`cjkoCka z`>8U9^^JTo6l(oA_uL{gqTqKSPKv4;Jn&4Idva=4?`zn5izLxB)!x+3E@&_@;`A)% zO^(T;qZ!b;@hsQg;>QN(NfG~ZU&t+~b zVy)A&jTk>i0#BLr^Nca$>YeUg%@&1Gj+U@LCH8$b;w&S7w}!t|Lq~ScYe#~w6CUV- z@T;=1xPs7y-2@|0=bj54D-@Ql#Ys2gAUU^`A=X`i61oWYs?jDkaop(&{9IcuD8Q@l zBG}Tww#QjIh1ZgGFcLl^+;N@^^0ht`HHTryoae)(Z+D_=CFI2OSZiX`rPrS1cONs- z(IubHup9q2VCVZFK=K>#+vDq;=F%$5@MxXL<+-gFOm81E%eV~q?M1HuSySS{BsGD!0#W% z!mR_^MK4ap0aWQJm??8(-q=ADUW#siAi7lZm_^

v@LXt!`=;f7j}1e!3Dq-5(&t z4wqMc)j}CAcjWxK4D0>1_$Hb19We7c9F?+#{qe#ITFDb!h+i zs6RjjkE9B5GaSb6@ST_^y-OM2m|;+vp^!FOG=7vnE99P-*s3>T*h?;P%%4kUp@pAcBUGhu-r z19m^NDr?O!UA}?|SqOwNkV$)zJ&c&%T~Oxa*!%YqkKoP`pR_Zd6DGsV(hf7M|^4F!nNHK?+pcL$cxdp0OBhU>yo)MOM)i>0W1{y|)PBL}KAH%s#qp|j-PBm6dKe50tcgBH!u_J{8UVily zG!>uKSnxNRk-Zr;Kv&$M$Pt#8&=qgU5}d7^^yd|R;9p(ZbiK*3>8oTZo0#c}q*cBR zyIAJq#^&`-9y}fM5k^rn`Utqk&F$--$CZ{=Nb2Yts{UXmt3ASKT*uOc>~<)X;3fCg zzss7qN$$!u!zEsu|Eb}z9oosbD?BzMSLXYTN$(gV?3p8_k+saJ&3k;3St1WAG*?7oL;>5o*3sQm*Wl<|yr-p+&Wr~NX(EUSt=E}{SKMQ>y zK&HoI>6cO4YP0$kJNqNM#Dbg|rP(gw^XJ?9@{Qi4e9PJk!4NX`|bFP<9TVnc8W?wcT#sgGsth z|Co){GWpz>xNzGi^k4VTGz?by@U+zFtk$UWW=BX=#WcUf0*CfTNt*Q33vK${dnkby zqLFFUb(^v3n4c3TR`-Ma8uuV0Q#2*nuew0XG5w`GFb+#y9|wWQh&{Av7d;(lN{Rl; zEo5akx5M~ZXk)iV@$1ziMMk#+qj--cE#E70VascbDL$*jtyTkqUF7i5ChX098 zn1z1UIgU`sYrxkOn_xc{iPLV62>=&hWl<5&!-5Lf4=XUcj|uQ5HSxcYW9W8u+Oso^ z)=y&C*bi!0ec|iONej-Geqh~pV6d7+#9ml6TW>g;Mh?>X;k-7KT_(4gG>S&!XCAAP zuu+J!C!Uq?so{D}hGe;w#H~CIH2M~AR>@Rwv&uKyT(jqG-i3(Ps~GhyRc3@x$aV4! zp_Mf!jL|}_ILM-eod2$D-IG1Y3uD|%INaCUnrHnAy20w+F3wgH_GzZB=f_os-1tm& zgS{SKs=7%1cY055n&Ot!D+?G%_5PavQqMN{Vu6YKl~s%M0iWfn?B%lrBd2uzWska~ z_AewvrD_uSEVYAm5W#hyoWr4_bPxrsDCq_OyC6J=FXo=V`xU4AlZD+FImM&_iLB2j zYi)u5mx497ia9DOndbf)bRXs2$S-vBZTQqE%X8P;&zp}e0JB#;6&aO|Wv~?P3Gq)j z?jCP{l;wM2TF3=j8F}66$zTc%6OleHoatG+;AoCGFGE>AqOX2>>`6bfn;(FW6KHMb z?SDXVhD3@dlWcwVBl*bpH9m@8dR9LoO%2ks|FHgG#lH{qL5uNUnil+><79__GR?71PI@ z!KHv-2BSArl4noY4G!keyJ`o?a=FNK)ahFV3z*0Hi0{X|4N6~SnEJf8VkO~m;p#zm z0_z$gK_`x)uXph`toE?I-T@(!6;u&5>j?tu)zm@RBxkecT^U=;Td(b`U&kaC#t4L+ zQI2T5Cs<|QWU)D#4WcVqV5C#euvVxcUF!pMs^l(N(iZ}HlD^MLh2J$a4v;FiL0%i9mkB&yr(7R_In$=P+L6@@ zr~xY5aM%{u_{YJNr!kg5H}#0LE+L<%2+HZm@5~LG1l?`L?FFhTR#UIPNLk2iK9zB6OJ{!DrU>{BgG2`PUR(L==xDRq;kdx=xEt z7(`?!#(n2>HQKj6_?m%kMd807-MxaM&%Aqyg@k_Rd}2^W_FS7;PD~>XU>H ziM#amlO)J2Z4d^+T@tbf_A$viWp-P&S?zn^7?NvDJ9(M-YGVfu3B1*+;+VW@KZ7-> zy=L(d_4TNfNX;n49*^Yz0=imbIUMUy$%bq=G5Yb_&V{j|e6VnU#>k+QxiZaM_#Y}` z2p??(L`Aw2y{c4{Y9Au}?oDK^E|ZCJN-VE*XlP_vXQ<3&faxRadu+m3hGdRf=Szrt zSP0VpT)ZDGYmK1u1k(~7V3P14@&Co&FE}!(qouhpM$hqpc1N9s(KJTCELl@tIu%bB z72>(_wDaro>BL7k!J(K*wg||RtVgLBg|lZ57AN!3C2HZCG?c!&=NE7kZtW^>lTkWX>oBq2BwV=pQ2gwU z=$smexGl`^<(E3b6^%-b3FaQMjMl)e_xqE_KkFMvKv2qN&g|tDOSG?@id?3MuXq=4 zUdx56A&7nZG+y~F<1e2$Q>=)`(36e8D&F(msnRYaD7sorLBES^Gz8_mu|ERDj0+4x zx3KSJEpcNp?(pi@sPj{IuCs!s8yyW=RIX;w20gpgUjY{x=f*qnEdFvSePY4Uq}$Iq z?TL5Ptvf!ui;$VY^D)x--_nYQK`Jpo2d4Byq5{LSH@Y+DcnVg;6n(@BX0^5eYSJ-5 zUfW;Ul~o0;lm#tIXM0Z0m&Od5Nu?iRGNRg?y$olm7;H0n^fmFR6xD3~ZXhhRVw1!H z_{Dm;mT%fAh2Dp=zv@isX<}{3e;*k0U z`$K9~^_A0cgw$W95OjFIh_=-H6)EEO5NVv(ak_VeBChRE3^My`xEIau_ju|%mV%HO zXkQ?cHJ;6SIji7()4)T=vFmn{T4?RBy$!=He{*WOHs=rt-Z}&TR=>%{4yf7M8}HKl zETR6Y5~w`-g>S9TAA@*5QFfyWzL+Tw5h?t6Vo+CXpI!I%EY#C+wZs52!-ym3lj_=Y zIL4OI{#;Mt34u) zj!IVBawXwqqdNf8kq^xx9e~GCFcF_{$kj~uU8goaHAC*(*CrY`*pm1l;7|Crwkn5t zqEPnw{HF^fR!h$%mNJ83@F(a0wx80xGc)N`Hc(_bj^1B+CX~3L;?r|HI z)n-BqyT3T;dbg`iPW?J^K@)~F0QBC#_)cHs$161FKivuoJGJ{>;ED*%!25`dz6Pv; z-YqHN^ThMZyB>5Zq1WmR83H#w($-Py?=I?mY=|f*S8}~3hZRpX;W}RziaA(vFGv&J zA^A#4jl~S3T{u-$PZqY4zQcWXfY!sIq6XUTtkPL_8kL-C(H1mjS=VnjD|$%m6i}bQ z=69puS!rqEp97G}0}Z8J&sGi=ioSiNke~@dU7cYy)9>Y1)OH30rev7!-$}V#>4dY3*xUu* zNG?$a3H`0v(;Wq)D?;=~iyysVh1u(e=6|Jd3T$8A)A0wrhC@>y7k{gwV)$Ill;(g?1Y)D3`L_Zx7hk@uS~c?R0hpvT5idj=~0mQ>FlzU z8@VI+Qd_CeVWRv%c-NU_-4zNyF(1S#CVqAp#{-=`V19+!jfF1(0&1&kJ!0W|v5yQT zpqh%_5v=Xh+hbvZs3NQLNJV+iV`a0==7$&-I@o_-BaIlAv9q1GN_}mGG~$Yt#qG4P zBJ==+;Et2m6nfxOD91DPK)&g5=GnnQZO-2rw>Q`a7=0Y5az-qi-NhD0+V3{R~R3-k_eJ$T1Z72Bwu)Q-@0qCON45Kn$N1KCB@ z027fltFYrX3#FO5e-(X~Dvx?XiB8vvIa9K+mDymhRHwnyz2e9nCsf$p2N5 ziX*<0blu?ai{%DL@@Gts#%-iJ|-wy&==ca!c!xG?ei}zz`&-e%Pppm=~m#>JnciF;3)CQ1MX#2EUy>CuOWAita zQpYIHE3r#7CMdZ|->yYRjcoZejAoGIU88N?05q3)iv_tk)1(bi-LdCOG6gSraBziog z)X!dWU9IT{v(6}goq9sseksIWyrPoavy8;eyv{UR80fY*g|+7Snq9PaEV*oB`WLT9 zDJO}mbqA8MR!g^xxbYRV@9;6dTwd;7>z)UMc+Cu>rv|z8zX5`Z)GAsVPPjf`XONG4 z0#h1q?Sl7UXQaISEj&rOE=Z9sL(-rmRa`A*Y+OmifD;t!0`k5m?T4OW`>80mbb=?45=907%No_f3@%Xy z@JMj%!^0!}yPq`B{`!1o(oMg!b6RnPA(|$K&7G<`AOztdKa^cFZ|CLH3-n%)A1;iOqc5$VqBI^>zDpCl5` z;=jvXxAnL8`7_Xwx<(d{QT#rzBG{*rHJ?U$Wl<8ZxdHv4)oc{d-~33q*kHeJ^uP=n zHK`}GJ3yL}D{)+didU>=p8eSFb^;UVA7#$z?uIX(w5>huX&#g*b7vJ*FyKKSfvy{8 zZ_-q9IB(ei#F-kX+69VWj=uvxvL?o6E<3ZJ$)bF~Q9u&eSvas|OO1cQZB=cOxCD4M zIB6PcKg}E#O)@ZZYk9rv1cK7s7Lx+0J1}R?2ky7omWc19Gq7g5R1S*If}JWU)*`jW zj9gttU?1pwi4gk%h}!DiafsQTWTAZ-XFn02OZeT!(WyG<>uL#ut?*J@Z+;NSFlou8 z6)Sdkw-b-kdrCuJyYi=->cjjW`lswD(I&wxuVQD4HR0>?afD^*&6g;21rju0Vf7A) zvP6pfJo_azU26%w)q^@+yy{ZZM;7==bIS)$Q{M|@caL-+=EtKmMH;t4`OMM2PQa2c zg82ze3^#Nd$zPNM%cD9~I*E!_y;eaECBGAg*V=Fs6Km@96U2K{wN*R=>{PiF#H&-E zAKU*U%#5#{beJVSUg4c&c+NXn8w2yTbqH3y-l?O~*eit3_3HOCxf91>&@~a6ER7A8 zrPQOHB{lJ91!SBkl4DKAcbK-eUwvab5uIBTkx_2BFpt^=X~hXmj5Lh87)erjW1?Fg zO%O(xrhg#xJr0cJ+`lKksy^19qIDt)6_c>o0QFN@Y70UPPL#KHG(`t*3=v^JLs_CZ z?DtPpYMol>u(6=)d~Y%k(q4>R6a!;0RMO$l-ox%4FYN8SdUwDxC)ZDwU-eHlFgm4N z(W|78!m3nGZzNeSOk~zvMuWn93+Ipb?UB+|Lcf{X<&DQ*fd!1$Gp=LKaL6mb6(!G- z;2}{MteoF_%2(#wktFq6HdgNvsD)E9JRh@w{RL`plG*joqww1zNWR$t>Aokk?@bhn zQl{f?ZBg{`Fr~!Sz!NY?yLV%#PUN0@%FZTD#$p_<5rqKki`#lg7Q?ewVGZCnv@IF7XsrZCvrlIwzolNy}3yI>P zOm0c&=C@Zo2+UX-F6=%Wuo2~(iKdvwp4IzthicC>m)TB_{7-u|(3v(8%H(`K(w=Ti z=aMdLHLbkaJq<1{tURuNFoE)>Nz-kpGcwqxMV9oRIry%p zYK%0+$`qby|2UAoahMy4d`;9z`pg7;I}&) zW4O3nl5S~k2$so}GeB8-f>HGNVCVt3J_~FKZfxgRVdW2XgU2s&yNq&9khzt^+8UE5 zBndC%u1sN&(JQLWyVy%#^I&f^dMK`%x6|eZ=*{<(dq>}@XsqhK`W$^*3WtVB%S*Kg z+7xIe!eGg9Pz~Gz<&2AFWz(iNq15J3BNAqcDN3c~NiC z<#h-@_g+UXICD71h(QiUv}_BA3(YICx0#ewPhnB9Kcou&{$I?!Ra{)#7Ur7-cXx;2 z?he7-t%5rQcXxLQn&9pl+@Y}G4#5j|cfGaG-skk`?%Thu_Y7~lVc zPFx(kiXW;ot@Ew_gsp=4;A6+xv>tbGO(WSwMKt}{paG&n(RHrkhZw8}?d9&3P^MQ} z%yn!Z!v5(5MSW+A9oAe`mL(*@isX?& z#VUIyBvkZ)6jZMm1c;=-u>qTIp)78-uNw7flS8pnW`a`{_pIh9OcDt{S0p&h^6~zU zDa;{8aqy9=HEDP)!w7R)-F<)i=Nekhm2*#>%xLYzQq^WIrx&nU@fx`OOML;4L5`Ax zYc^g08fLvAbR77(@1kC7sri;ch+6Lktn&nDd&5tF=l=+Mfm% zbopF{4PEAe{Bx26np@gTu%S7cd-`-l$G?_IOBQkP_9sN3KRLqtbU8txa0+(z51Xi? z|3vPNOvKL_BKVH~#Q@aEEp5fKychga1JrJPKgb%G^h>GK`ohb@?q>8tX7GfPJsHhFe znEY)O8p)bFbUxG365n24Bc#LgIwZ$`hh*~(SoZm5$}|~Eem>k}7bo+NaqvL_Y5xt8 z)}6j6Tt#mw_B8!hiOX9-fv(ZZ(?y8;Ga^-XL@HeQ zOl0108U8riGy9P68X%;S-PzhVyQ|qi43(qT2gfAvGUtc#btKQ}&%r#BhQ$8vsAU<8 zzx+Y4V0!y0A3DjhEackBAB=9&#D8hyi&Vhi0tISYCWHSsMepz79g}h=zg;FsD~aJm z{(LBBPBE@grw~ye+gh`vM*7#}Nc)#{6-SMP59jUqD*l;HMD@(k+ z4K(-?$}Dv79(!mN@4kW1AJA4_pB0n=&0`vf$lu zdnLzsZ;j}?lU9!9d$m9DFb2&Z#I~HFDIv+-LAkpt_T>6cZmWr}iZ|mwq1c8AY~6H= zPMUMULRd*KRD8;r81J9zzTcMdIc&q)xR^3$lVoUv&*9MKsB7XP|4vNno89x6eUGvz z=3nGAX_;TSJQdaAq3qK)gd2=oagd&C@2{3;QG>pbnPzyj z>}$qQ>5@EHWh;mAFE_0~s1PxaApG-t2re2reh}Is-lGW#ujBlzAF{&Vk;zITU%NYx z1t%=yN#);>bf&(OwM*`D_;N?gfTSm69awjAq0sS$-Q|Www(j666fhv>Z8%n40;@ic zd3Ct|O6ku$T@&$){};!b83v!DIygEDfz;rgXpKIc;FHn10;QAGddPds zEI-P&?{}E8h!_kRj7ajh?@q&NbJ|q@#pa;uvmc_(lb?UxT(#t^`w0pL_to~}y|9dB^iuK8b8HrKj(XX`?`xSpdvu~%47C(pAP^xzG$viZ}9^Nex=%(k_zJvh%hn5$wS4Mi57rXw1MlCV#uB#JEcFHhqzxAo|PZNyyz0FXYgc@tD9swf2IQ|~3jvTDe z&gMbzl^Y<`B<(CsYh1ahE6o9GVK8F2OZIZLGMI3AouRnE#Xk{#0vml+3<~~ciH%U4 z3&qCDErG{GZX|3K)fP*26V^Y@C-qh#KQ*j#izDIWpY6r!>u=%x{<=oViS!P{GoP{a zJuds$H3e0JjI%5ZMS&6et!y4Vl_44{T;=r{@$)+OhB;O_wQFtA3rSA}w(X!LpK~1# zJ00oq$B1}h>Ft~puuu_MK3`FwJuYb} zY|7=Phhk4k%4iiittThhl@l?8bI%D**9m9iKH2gWB+{xcYynFPs6XZBMb0;C&EWaw*hDhmRvc!JhjeVsD!3tZaGFAm z9$>b!wz1{E86Z&EQW>r=GIMc&cR!OWrk!b^dY(eL+o7TEQ?kf~j}z9AEtm1`)?+%m zHxo9)M!?!o4QL?Evhio7#I(n%m(624+zrNIBpt0yr020kU%rRBER5Nv_o`q!3_&V$ z&kNLOt=+xMs?&8mACm=bOtIR6vC$%W8M0eZNP`DA!_$UAX$8q`RecTfrh_H@TeAa` z`M6HRyR9H(c$z^TZ-rcRkMYbGb-Ccsw4B%eHv3TSY&un=~{+?{mQ>$?{{$n z5>JYXXEuTEJwo>vyKk&6`nC8N1d>ozDS;&FX3aPRS`0n9^Fl_Q#t=4>tG&+=&}-iX zN_lBC6|ykwR}BecQ=+y~r38BTpEKBSGYkbiBGlTf*=2U@5;GS3z>ea` z3gV1z%NNJ58V2Xx*;Kw-@T@rQwmcopAKagBsv`yTVOQIrx01cxQ$_MvZ>pA)d9A|o7GvDQ8I z&z>Q;+aW;p#ncSY_WcUoSGd1iF)HbX&dyTge_o~~p_NI0hoqFng9}8UeCd9o+Ez2W zc*g$R!Au{ON^`feK1Z5AM)okMNQmzKg(S(6dme+F33gPds$kF?!8@e0-H!BBidQ~U z`Ycpug*soMY7p0-7RSrBX=s3meY#u3@qo@`jZgt;{`{1u10iOvgyxU)Ql&87_TKS5 zJ-=gRBHl}W3-a!(y=^PpOU@wZ=w_%Y=ycM*@Xew6mxTw$GDsx8{ zIGU%z5%a-TRP?rWiigMm({`9>7%tJs%t+nF=wEq&EdBIOXJO5o@&KE@Sv*>)c@cnh zGa3d1J-K}lV@2T<{Td?0qRr5%`6KF&2MW$YGowZuILU_cu&pQFM2(+M*EfG<(j_YG z;b-erAoL*D>x0LGJ;G*(Rp4GVPlxUfvy4=gsq4`Ub+xBzmaw;(!<7p~KX*-i5uaC+mMXx@m1a_A@i6Gbig2By6R?FgH7qUzQ%^1aGF zP0c2XlFx@`2a=5}o(6!No41z&Ll4yX5W0M?F{&darWRuP4H3|ah!RyL(&HA}T#s46 zydW^9nn|xM#A?1A&KZpSxS>z*Z;zlI_cA`NxLC*_@00|%Tz-Hq;b2TVq19~q7Wjv| zTdbnjSU`Xc7s^<}N;IoDSF>HgK}}la^t_^lr_6?&``eC3(;Okv-bT3mZS^PL69@Yg zPj$I%Q&{hg4}0U-+aj>J)^EbA&Br)O@gEFM&a7fsaYQvzn(vmLeE9hWlZv!ozbfD} z51RadG<)ZR3IEEur65s)=7g0bzicwf4^`M`4q7kJnU;P$+M5{Y5e$)OOMWGp6TQWg=LZbfJvnp21&QP+G3vi6TE~c4WgVD6U z9E9A;`O9(mT_-Tr+}T0ORi2Ue>T~A)4=M+FBw&^Qao+qahW+|R} zbF@@DpOnyQ7YBn#tKN6eB%Qu!pE1bf;(TZ< zO=GNzU6mUx_kI&%p4{Rt)~xq#Yn_9x^`O7q36wk0guxfz8d+s>cl1{D;hgLHh2-eE znxk2hDY2Shj>VuY8kDZG6C|$U*q`YSjfcPYp3dUu!HmRDkWAbJHr83KlGm^_vntt22qBFIC?kIf;47T zlZo4Kq$JZ2{38#zLH~F8m;g0cwUik4S&MyanFGnyIkeeU05piYFwcAv6LSDZ^<;lo%r2w_jZ14m^os85p*7)HjHp5loajnHVxvZzaFV zYO9kzC;x?!9CrSV&mOtc;yVlL7c%6tIewZFXSU!dO#*|@^{q( z&JJ4-U}=K)QXPoxMLj8M@!9Zr9nm*~3QX)w+9?m2V7J-yc!f1@q%He~S#-e@Sp~CE zRDIIls!}$KxWuDBl=#(FS@pU%$HD1Xq(kn8P(=KOdRkJ*M6KayhJ<2JX+W6W!ek=z z>x0FSM-1*16r*(s8<=^ihMI!!{-qmjoAq?#Vls#I8z!{tx45Gp43=#80vq z>&yJZ&gwe?Ks=qsH^(FHnG@-ijh)UrhmFa(V>X}s5&V^%%4jZ^$x>g?Bxu0uu+iSM zx_-j+(>dtQgzP>`8irC93n7Op-=+7O9Ndd1F{1yLY}nJ+}6see;t;5NQ! zd2Akx{pNWf%Qr6;cW|h^@Pq0wXzin3AQ|nnGkoo_1GpLB5>PXI==ylXO?%<^uFlXx zh=rTFLm%<-^K9Jd!NY^pYZ&nPK78Z86a|Mi7!93HJVq=C4U=Dyktu-b*SmzAFKe=& zmf^2+SMtC>7J;>2INTZ}It89@AxS#}lp=@3ozpaIiufP)_1Z63dV^zBrGMpc6tQi=;!PS1 z6C2Ghc{*HZc2es=AmXdCYm`wYu}eCI=rT&>+DqJ42|%qAeT|1W>uqV?N5v%aqR5k9 zho7uiay>vZ&iVKY(<-3bExYHZVyOE5l%YU}BekNey_u=Dt=Xz`SRaO#J5e97g(L>r z|Fl1nGfP?8)d|_p1H%2g)%Cx~R>@#WNYfv9u)$ky=+U(KmXXIGu@pZtl)waGNN1YE zKt?0Hymj&CMMAWqzn!8mV{uy*s14sA*V0Y(5&4w+?(g2GI`y4n^!-0FcMTv}(7kMI zmY<3uTCXEw=;9lmUN#l*;!47i?vrE6N=wC+ieBZLBa-uM@aFx8l>78>o&!fHt6dXM-R0)6@0KA2o0bi62}%Ur1+HkQ&Oi8(C`dLGW$61-O5WR9BA& zJvz&03l6_RXw(l`vZfWqb6PXq;uo?Y;`SYr=0n?NW|EFaD@}*Pa|e6J)~QHw0Jxq7 zPE))oTq$uoUamKLbfrbSsmpdF0J^VP5v}LG5QAEG&~a?yiH|N)6yOk%k5_ z&^-qUTsFWe!@|%es(x%0Fc;Ul+xn~+Uh+MFC!Q?kW~<_e4iy!XCM085ISYELBv~v^ zk)o7bl~MKkkmNUs8xmt=g_zYzGy<)fz6JKssZQ_15CTO{Jz&$J7^2!y`;Zm6iz zH=fTNko8OZGZAxG>1^o+8CwD4HN z$T=t^7b3SF-k!`hdg!Gju%`AXy$@iX6ez&zbylbudcC`itG zFt57n=}I)l$Ck=u;G%69&#{rrJBU-Syg)Z{#T#vGG+8_p-?nbYj{H@3 z;K`D2pGL8UByRBW9k@*Fs?t4bIi}kHf4smn)FA^gRW3uav_}~4tMi=a(r`)TN}xEj z$6IAsSP%!%0X!^YR37a4FvwU8B3w4yjf6=5(}XBrS)8X;R;hfiDrBlb96E9zu6Eiz z)h*wEf`KEJ19d&!D}5^@YyoJl4&|#_$t{yn8@oFV57qRlCDKLE7)cW^s$L;Q+=hc( zA0)qaqWM}!RcRvxuM{?0!3pg>S&^D2!P4Yj6OAplZ`8R4mnrCvJ!HFRbVuquckYfl z!*%U&&}WBOlP&fjClVp}R;@)$Il?iNzY^G&1YTiQ;jmMy6==-D9I=C^ue;{12;dIh@Eo2^7Qc-kF2EGP7c8(cnZ>?)(pu zg_FK$4e|ABL< z#fR5m{{`n#U;$h@k`!wj#4RLo!v>OpRJ0|eNrS$};6|&|0v%lFtA#+nFE$n^TU4uX zmdo0&sP<;zD*wE)tzTh%GYd?W{zS1oGZlHMmQub!tfB~h^Ng$M7jW=dJM$RTx$Pa! zhmuB=E5PKp{0k|h0zGX#hsQ_}HV^L^Rz(40Jji6xF>C4`EeVR(+SfAb&wXrRxM9NH zW$DcY?G0mWTo{WdoBzjgCz{ZKxAlK|(63N~_8W9g^Y(bjArR<#ppd4LTp;vR=_zHf z`uQ{0^bFB%x9lNoA4h4c54;|}Zw9mI`C`^sM(trMC^IQFY}zr9WyfzGI2dSR(g6PdxwSK-VRZ&c(-A+cgD^hSoZT-)(ak@0JppP@djuK`n{kz4LJ zq&%KN$ttpiqLQlup%n=7weupTlOiK#!cRsi2zA(r-YZ+-$1Nc}kBmkJ8y_lu@Ewa^ zMeGQ*)YT&JF?8ZE!d66I3E~*vj@gw~GsieoUJPUn=uO#8+m>duNsl5AzL_9Q0fe&= z;SI7_3T&k+{qu?Wz7S{(5)VxK%O}#4QxO~l$hh(wjq1446Tw9R#()dE5Gs267-t4N zU@5&~PsJVN*tgpc!HlV($Rh#G+*vAJeH|Pni`)KsA8SWrQQio!Y1HmF%N18=j~-5I zPI(?TNdLtdx%|dP!kWcKvDd>z)X%h$M%2~4ZHj#skbjawEf*}&h%G8&;hNmijZ^8D z_x*<&zr{|kH+?WS;E6&TRM{oFZ;SX{)ozqg@wE4Y$x2Z$Fj%cZj5%ME**hk59N*}2 z`-J6(VYIK>SpJCg?sMZe%NZ~l{TK7*672S8Ka5wI|DDtTUZ7;@{q7e`ZVql+jtQ_n zV}IM1L@OYtN#Mq5F!Q_cxJI|{2+FU2IY<^w3q8J6nhol3E86%T94uyvxG*(^%EZL< z^RU0At*x)WUp!ro__pz_;oEFT>%)JX5MhNl(NmsR`sfuW)#2)~EMpFbPFMuAFW8zk zual;2wq!gWJbI}~yUwc0;Y&Zm(8X^=`@fbL->koHf(XzHwCbeksS%^T8omv<9;72A zbpMI+(@Wi6_D1oEo~{6dFnB>Jw>`gaHcaVW$DC(1;<9=ODR#%gszsAdSOr+6YT|Mm z!7{`+e~41z5$qzWbnBXf{3SuoTCaf{R3-H$k^0^^*aEZ%fi4;Ao9n~CTooJ!#H2*7 zCH4-RhbgGbOTkEx`Kv?W=0=N?-KP<%nN(Q}ma~TsuTAcQzn};%@?^)NA|%hMNE5KbgEBDzu$gKylWM5TL$+OXX@^;F8IJ zy8&`LC9m#!fQ}TQ$;MQMfmIg^z_1YU{f)a^^0ot`pf1!6W+iNnDTr@%8IqU|eN z>1-Q2gEj7F!kc1Y?7t_U7$lGbYPwbs@Oi{05%LG#Z?b$TkIwW&QZ6+0ROhZhNZOJy zdn9|h{m(Ak@L%!dnNzJaC5}t2d&{4H2&XMQt`6F3t-pdocsR9Nm zn7=JO^9@k`15^}GZfE(Z`TtowdF)nvLQHsZ#2SB zvA_TIrsDr#O#1g!{+`eOFPu{UFBAIzLa4!J(v6Ua#)fka z$j;7Y0dtYT<8WYcQ7iouJQiIDwGx=43g_(V8X6fH31-}Ya~$bhwm(^Lg#SVDfDh@L zAp{I^e@JpT@G9QAZcg2Y+YSgXrVCM21l}338vf9l`Fh<&%7I z89D4)-z?m&&Pgs0u0#ks?pU*$)u6i8d1DIh3!qC&VsM^b34A&_L0Of5InD0&8K@qa zW^daqBLMGfgf4A_X}uqOZly{uA`j2+bx82s8NdzUe~%}2R>MrY9gknn1~K@Y2;OUc z;kn`;2-EJQ;^VL_jkVz<2Vzv(Pn8+kHxnc6*26Z}^Y1K|J$VKZ?V`ixky@)C* zz4COz7c6z#$)j*;q81?bw{O}DMMwR42DCD>HW2idxujl?%nVJ{KO*rV4}g2G(Wp8o z1tWH9jIJkXVzSL(+y_bhzq71m#&Dwu4Xt#%VYCF1!0!p(TuaW>up3#n zhSwn?uQ_4}=Cg6zN4GjH?SckcHf_72Z;=rjy^7Jw__ z;!EThL5{%ed%(>~Wqgjye_c#!m-?Q3m)#8*&u?lQpkz?w8nW)Z3-xP?j7qK(JK!!S z4vRt|QNn|ZT}3vkZ?3Yi)`t9Go<)q4Xj~pT5p?)dCEe+S=*soiVEq|oE@KC}1pF%> zMcwJaCtQ-_aJOr_kb90eh01vN*p=MLIgRwG&*2xJDHRAp>E^>d`RT`gtiuhk?ZEEG z%kkfD3%h0l?l7O-cm}!^X(7tuv=p9!^|}oabq(-*H#N~O=tkKy{gc9H=9spa)@wLb z(bU(uw1Ap@|Swl7c#@Hx=UsBA!MIo)J@12bvyY0EB3-vnhj zuc;@2r_PtqeimG%L#sYPxe5Sx`iAPk_+>Z}Nwob*;0(PVV~_f8x%24KIR8x|M5Z7X zYGgf`R9s;?hkHAp!6 zH~#mpHQuJfH+!$oVQ~+&P*BUnyd*yniqOXy#Z4F%YtPTo7>+yBV_3*PD2-5gL>O$? z9YI_Y`8yW7b0yCbz1O`iqItwWLOS_v<=cVY5+QcM(9w?rb&jo-AP194yOtNK_U9To z#Ihp*mPp~xg{{D7RWXK8_3u})(zLXp<_dCsi^uE)-(MnPc>Ze{0C}J=#nQeq8SYC^ z+?;45-#)oZg+?R;>-U$sU*WqDe%(AL9+`$$o5#Zsjw3}o4zeFk1-PAdrSq*44iw27 zz0hBEwqV?a2~Jnp_}-kVd{hA|3vIglI$td?BJ;Dh@|7@NQCa)9NO(2nR6HLvD^JzH zDD!`K@8Iy0{Nn=I>!<4GY@J|>{R{#2qr?>527>Qcy;RrTWxrzP!`PcdH~qIqc$y+2 zA)>nlY>IaI!QS^zz#M7m)Tjf#dPdngjLRA+P~f)rS)Tbh*D!TzzVYOGX@%C>TZbDHRYG4KB%%z?=np&z$7K7W6QY40SN%D1Pm?CS^f zujiWat|Vh@kHEn6`%Yuok&WaQWowu647GH%zI%`+;l4^Z2?VTbWIdQOTphPI6CP9;_>@I&>kY zB@Cu_dwq!E%pnMyyjL>b=#6fr3F9Ez2y zx$IrR>6K-5xkXg2v_ayq+d#rjx%#~Po%H6(CB0EET+4~JT)s6ZS`Dan3GLMt_MC>= zzm_XL#Y*$BK!DZ1q3@kSSNK7c_!Xu9k4mkJ$__gAm$z}Ka4wsn$dwup{RGT59}$aX zE#(L#zfp4mv6pK(CdGgSSXd|E+8V1ff!wmU@gcghJ=ah`n5rlzL$}wxa(V{b`34E6 z1lxe{Bk#CY1$VxgGUoO>ZUYZA9x=03*dPMQ7be!fec+4B3B`oB*ME2NI9v+p`(KDC zrwXuCT0%z?3r^4Q9O~PYtwEDs4`lEow;1}v)x=8nw58SUy*kyjFLj&oE3UhUNmoZ% zvRkn|EsPP}|9Ual9=R0balR#6_wbR(rGY_|RS)`FF>zpQ7ZXwI-G_PM>xnmnoD(>^ z-TbgaFV!Rvs$4HA)?l{*Z9d>a=3j?64mJ-K{DC_VMhL8uxFvP@paenI!U`n3Mrr~r z9jGRDGhd8A&K))&fiDrrL6eEiBO8p*jqO8d2XV(GN}H8HBCw%jShq9^bZCrnT_9y! zLs=-8M|&@kw(Oe!G?8JU1j;<(Hz@C#WoDpvI>M3UO&YYEFXk=Bk*ZN8-S|aow5Zji8r6yx{cxF3VV-bJhJRly;ODQd)jl6W1Nybx zg!@Tqh~l=V4l6t!1|bX}B!=_&CN#LLLlyNhfIL!LWzqfSJxI9idA6Tk zQ90`LW8Xm3*z-oaLW-yS5z+Vhpx3l7roT->Mv(aOmEt8?WOw?%?>KT3OEdRNR-Iity zyUJ_q7()C6*r(o^`{Jq-=jF2(jdFv4RYtQ|IQDL$ROW@}XR(RkGC_TD*<+K(__H*4 zA6MRp*cHXiI|74r5ch;{~Db zL-;sp1wgg6f>s%B1NHoWl>Mh%Dwi6NbhO3?63m}G%(yXL@iTHCleZm`;QW(p*g6D- zNGQA-B_Y_^({ZD09DbB5CKZT->&h-3)4YWDg`S z=+!NEetMeJkDYghDR3JWLbT(Cjwc6G^#QeXC8zFixR|ypF_TK_3IjN0i z^P6b!kGq_Pr7TX~nyHU8E8^FG**ZR7$NNES{@?716AZ~u)a@Xy7MpO_W-xc7(?oAI zUs=o!6>WrfAeloy=_?{k*R^7~3Ec$_{UWFpZu+?04lr8DI3OrQJS#CT5L+9>_2xiL z@sc0>ew$J2B&no+B^>%^?WjO;Xu#o0v4y0I=0TV+djY%nmT$YpEc}uNK!RnpD7uTz zdwKihh4@XKz?SNK;8$)xgj-qOO?!kMPIR8dB04K_e%tL+f%m|0Cs8$w8)?~^VDT3L zVIHK^V)KK?ZvF&K+xXt3edD+8g6Y;HZo-!e%5rtPIGJ!=H1ZdTqeU4~ed6pmw_?{R z+emhC4L;+Cn^1t2Ae5L60mWvYJYa!M?EcZ>f+=rYvr#WA%O;SiVoka3-KJ7s5GU9g z259HNAM;g!AnA=arljtTOYk0eL*=YxTXA}vf0<4JTRFG=ezFGTgDFJ-eX&m-0x9hf zzIkR>^ko{-JzVs1Yd|QMr^qr7;24C1)Efke0-3gTh7GHN?sT8DUCt>GHQV1` z&F${)PSf0i`FVZJaHo^U7~I$2x?V3s%YbY57@xe`a}p)e-=^1w@;DC%7U9<$sxrBq3l3B%Kui^a#ZcOH zgrjNlV&{rUj-vh_f9TO-?gCONbY4xY6UgUa*fLkpE<^G+D%l&r?Pz3$2k-l1y^ ze(vB(1fCk72+EJuhe`O=G;x>v0_L|Y-?E} zh-?M<*f9_uZ|~qaz;a{Y1GlBp=DI6$vCI$aNdbMa-LU=uKky6qi}j_s+wnjqxp(w?-lK5&n;Mi}bmZOqDv$waURt^kvWWH!ggtUDB)`p8eD@44?pj`ONvtD8TZS@ZX@ZdF%o7`ngbTwJH7NvrviyV;Y3 zsH$WnLg^0X+Yw#14_O0DE28Tvu#J>chRyUI2N8}fl=}K8CNDvk><^Y>Ykh#&G;99+ z-Ow|h2X`P%Ri1{&oszRUp06xUgVs_(uMe}$$#>2>gZXG>m*(#NbFZEV`h{h2k#;MrrKbH!4~;b|xl0 zbUw(40p}|^OM|6w+Ab`Lc?$9IJrW?0~b*+s&&`kIe zs(v6iUqUHo^LLi*X16ZXvA&>hO@ZJdUtepE^}=wD+?~LsEKSJ9M(w?TX~~5#i%W6S3THWiv1qy*gXZ`p0q-${Nl|G?t~ zFaUN%M~$*v^xpisLObmi^P7>`rSNJ;qw=w=ddl`W5UY+!Q@-7XTLXA@S;}7;Md8)q z4hiD=`kAm)3XNeJYN`-QeA?Jxe5FqMF%-VNur=xJ7}gmkLLs$tOX$lX`6x~H9qitt zmQkiJ!@&5zLtzN&i`lAu3SOw}O`!P*-AgBD=904;by`4As-f%k&Xp+Yij%?BCH`71 zjCa7dg5J99BG!3x1!Mi|X2G|L#n%|aG|uaL*RZRlU)1Jk`rL8C#}RH^a02fvbP(D6ThE)t*Lz&4a_ol|APZ_gag)@LXaxrnCrAmAJ} z)ProySP(DD@#{^_b2X%&jFT=H4nRC~$tj+x-j*rlz_WmVUdEkjkGjG8c@IQmb)5vwj4X*KHde}6qmVQf&HI3LmcuVu*cKtj_pl`Q( zF;Uob7Kn8-nU1VlpJ&~EWjNqzYKg*bUG}9_Cf>ajo{IPa`AnBO-*}=WxDuXa;ksxT zb<+w;p8!;Bw5R||q`I6^3JtC`DgkgQ74H#}w*CV)X)>0$;bh;jn%tymPlK?oxKAU) zSY3pT6WbZz77LP_!e-AFXBi(%{vg?A;eIt8@?q;%j~7B(5Z2&?>LS(B3V zfV*4!)zmJCc-(1p_b-Xks@qMJxLbzt?5MEBYziiN{-%Qw@(= zXGY|oEKOx8jc0=J(JsR}G+}nOSX{M@i8(IN(4?eRQrp<6;Rt(OaD02E#$OC%@R=OsTEZw z`U7(PqJc?ahSRa>FmajyQ^~yiuQ90~L(S+4>*iBF7Hvp9ND2CzGeKx-J`G(p(dcyk z3|wdesoyu7Y{U6S7?=6fA#`&(hy8NCnJXNkTb(+~3w7haIDu#Cp11!3J7KpRl`+Ha zN040ej4>1L8v9K7*ynrdXKJMmQ?`!RwP4VFCl)w3F=cRuKQDFy@J8*D+~$^jX?EmP ze29FN33UQ9^btSHGR)+wSA4{sYn2VD4g~)2Q&ey1{)UCziE~^l8cC;-NHqxj7C5Fs zUyuOke8SRAvmIc6Et|Gen>blkQWYFr?mA_ixYggmzUIm*}KrAc5QLuMy3?42F-o?c8w#dNs&RW zuVhpqj1YJc|IdZWzW6GCdq+k*daqO=|20v;AD6sNd|BR!UhkG)pygf!p2Z(7^#@%+ zzDr&DVBZXUDtrN!IHo4eT+7EOnF^GFalI8sN?CXGy>67BbzLmZdf6ApGoKEf%*F6S z;p6q5@?Qa3;zER6rF_PMRP#!*jejqs;LpUzfwzz}<6^%+qGKhI-A)WKK6mjF#Vo`U zJ2engcx@{4&&B6W^1+Oh>%kvhQ4G2M>y;{H`HehLDw`&my6kAei|&Il&Q28sv3y1- zi$wT&_WN6^^fb4>Yyl7Fzbuc{N{P2OVkeasl^46ptP13wgO=|rF4wKm+Fk)z<5CR3 zx)P_IKXe%IJJD02uDA#QI%AIq!t>joffD$bRvfShYm=aZD1%F}ej2GwChqd7QB2Y%RJo zYz7qV@p2}(-tCWauGKy*wZUNENmA0GFS}AuI~FGzTHwC;L|K}1-t&}1=W$LLG!a(q z%RfuUa;dq)b9+V0j&LIK2`P@Gy_}u`8!vGv zgw1|89ffyS0in%CBr?tj!t1ha16ej9PaA@EMpR5S~pn@GW00<6ZkR0D7)zVV#Z!!>VV{^3j+ zER>q;RZ=p#U79G>+sCQ&`3@ua_9wJXLn304uu}PGf4%fLcg-M{B#v1pR8ECFolI-N zq@ozA4fkZ38_Oz2FBDIK&{a>v-z#jJUNTRjD9Td62diTp?+l;QEwWLc)5o4La40sO z71HBoCKpgayX5G==!LzA@~Lca#RSHYtbdLrjtq1@Z#lR8*}~W`jxASZsaMEVULz77 zg+w5+eO0q(aDrL?tVj*s17T@p`C@w|} z>fOHRl+iFD)SR&VB()|HYC)}OsD3S5y5zy5ndNP<`BS>`nB|g&M6gGVH?D%QKY`$y zdMT|HdK{R{H}I0bG0DK5EEW3FMq+B|vkn|GbW{P26WH^z{D7g9J8D~aW1GnzFvwu7 ztn_T_iW{U8WRuo&v%Xh&#b@4Wn|5*RcpC#4q_;D1kGlRShaRlT45W%{yS_dA16PSV zPnbf-HF%-1p8#3$Pg4mv@%Zc6X;LYv$S~9O5A!GRKO!#G`IOZ?40rJ39Cka!K~ zo%MlGr=5g%xmilcWHS0Rs$l3R&>>?U&x~!Pic?H~>!uuJsoKuCmwm}f#$){@zfvo< zq&n$xw36b2`!cXmLy=nEfDP4bx4z)cFXTClg2dcnoy`zRuGQ?9krgD2-}e&E2SRl8 z{p|`Be^J9)AsbUDdgs?_GGxqN3sC*qC&qU5IjFD3rh%2?ZGF7aZ&oM+x8qC+;VQA> z1;_ygxqcg%Q|Xkbc-OotmO-eWwFL9cykyrtl(Wvz*cc^=HfM@r#N_5iSWF{RdR6us zHGe(Or3${A4Mpt`=!6FnEQ+qQT5Mu)9;E>moAGFa?6)`X*gcBYP!*I(gLZtUe2^uj zKXIRp>IeUh(V%sCYYn>zM%+Dh=CQW)fZ5f{?9OlBU(!5?Kpq?7`{Hw>8jDLs%o9_f}4rC6Jvs zWDb^tV!=HKx|x?Un^6BL1ts@`I~P!eU~2ciFg1MB8t#WU3~1BS35Am8EFW5GH5c}A z`NK>K>+T*C=Im+_c*`* zktViJH8`}B0x%UHFnAjTT@wR@R`Ei2?S#Y(JPfrb6CHeUCc|4I>-9N5F=h9M4g*C7 zX(TQcE!9Mjg{?%<@X~rHzcD=ZqAFg3t)D4PTws z&kcehcvtfE$Y!44z{Ticve<2`LLX!6w{FGP6kLHlhl5IAF zSAY!?9`F;u`l$r_SgE7f5Tckxb(CTq7=z{CyCF3JG#PNZ z0tGQ>)qTHCoonC1(4$;9S-dn2FjIm9_6nyh3r+d3h)I@1#cG1c46ZH<;JyT`>eb^O zk#@o{`XRAhKCxT)8lAPR`Yaal5Mo`9Zb&(-oA8G52BNq1D$c2q5V*guDEguim0!xY_8|IN885tiBdZS8Z^=u=Wuq5 z((w{yG&N%OP^6Ck*w}BPl1uUQTLiQzXI;Bf8WC1@;V>#GP_=?3%qtx|Fv4al(?Hxz zOg(YZ-#_?BCV8yhy&Rmfk{=&Sh_k2v|1kAcVQqC=v;~S3E$&d@r?^{iD^T3sNpY9p zuEn7^#oYrGclY8B!QI{E=0EqI^WFWBFAqBpJ8Q1J<{V>=F~NJgq9=TqU-spU;y-yf z6;FY_+9r?@&U~%#8c+4fWd9P#?{T%8teaVO`L=}pFmT)53ZlRnVj?HL)fMo2G}SQ) zQ||0Ki#AuPHC6q=(>z}%+Rsd?%zWZ>Kk!zsi+Ss`ptTzcWQKHDsg@lEIDBw4v4)h7ieEEl*w16(}*9Ph{LW(vEN<>_cH>*|)=o-+WuQs|yWjIQ?a#jnnO)W~J0(g1!xGW-kfKdRT@GHTPId)g z6?^-4iK6alEv=gV=>fzdaMRp$J`%9s0@~`!|CBFb`_TWB!~0tjJM4=BALq$BX@CD8 zvs;c6y(^Y11jB*QQ>(fIy>5$zj-y7J&Dz%%fiv4CoF!|M3mqOtnvIZlCA!=1a3 z=Y!DvcB%K(=ZRgE!3H^bwFOO+qej4CE$Kp@p;`&AJJTv&!>_=T?uGHKKoS(in`5QC zKovrYuKf0xEfup_uwIStI&n#*>i643)1o72hjrfwf0Gl}nfc+devon+_jvo40r8aL z9Y3_l(ls>mBf_7LzD4%Zu)2C?XMp?R#X&Ei_;hX$;Z^tlLa zdB3JXvv#+Cvvw%pPorP`>N%Tl9DTxt2&`P8teaLh0+fTh{#!Br{6&EfwLN@)fz|Yq zf`~g>oUord0;{d-@~u6}2~TEqxd}@OHM+Z`5OuCGSy`ME*h@HdC4FzCm9Nj*!(q!r zeb#h3h<4~v0?!WE3kb^n(#ZEY`(}zTY5G7M`p-P(=OCkeB`{adBYv87j?Vn+G^RiJT(~j7Tsvx=AcR9 zEP)5Ox1iRi!MfGWFyO+J#UO4giTm5kpV8lwEll-I$UNMS)T$KjGXcSySv31U=-a=a zwr^fK9fmZ$P%A47&w{v;L#8@(NxyGO-4bU2+|k)mOmGb{mxJ4RZe)ER)~r~j?yor! zj{{hDUbrC*RXlHmYzwWPd(GLO=vNI_(9jUFCEs91H+;ALL2t1g9JnpNm{|8S=~1#x zE04q8BTX_S3HtI;f`t3a1C`LnRE-i&Uz*GP0>4W++W=0>87pwt!&SfX0+lE-mC05a_r!1%v}kPK%F@Z!Wn#s43ExfZd6 z!yg0otBsyI-E)zH4Mn09+tE}O-#_k?yL5Ai*8R{uaKfJ=m$3RwQen-MKlPD{wa5{U zE0>Zi9~+CLFeFb)k<}$ud+x&5f&6KW@mB48+(bIoHPu<$t^tP#X)*G+H=D z@~4px@-N-5mUZNXhW4n8wYd_VcSBTJaSZ1sa~0dP3#DlRVXMk*uDv~=(3r?O0-a+2 z`W`F~WZYoe;NHARGMd+>@GSs=z>iBU@GxWTc35ayjoX7rMF;LR(7MFO!TkyK!~+qa zMieNmPl~1&N8~-74C5HZl|Heo&)%VIvL3X%XNP?1@hPINByH5+)hD2vV)-3qOSFa4 zI75QDE}IisxL_{X6N0lHMEPAA<0iNUAqn@b|)Eu_51g&8CY$% zfkFF9N*~v|%EHSpvu_BAfv^^sCm z;u{4dmV8qe+jC}*lx`j1%qn9`=Q5G+7g)OK!c?3LSnT%uD6NU*(v`^5m^d#48_3>; zqKx;2*84*LnqNxmiv8h#k8zaekfbXIS9@LZt+IJ5Q-rtu&{x}MIS#Z z8c1k8o0GdsG2I~0UAoVRplUseG}_YH}6dRel%(g?cbiwC6 z=Tr@wA(bG&xz@~l|Nb@=ll9xPB7KbAOipeT(y7c z1#TVswaK@4XWvf@AAp0hr)kPxR9~kavQ=*kYyKJdPYvsFd=+ev!UtZ(lPXUEF&br3 z!s8^z&}vPt3X4L`!2jAmF)_rUkL|$PM5(iG7YLgs8sV`}hjDmboWTzgd>W*L{2wO`8O>eqeWTqMyyN}rau z_%QyLELb}W^Iig}>rwyA{Z|=%3y8k>_^WDO&x$JI=Bs`-ZkscWzl|6d(j`} zIZ_S{+y1?du65hx4L{xYUMcag!FTS4XseK{^0VFs7ji~zL4rcmn~BQH*1#Z|UB`ds zz>aqt9^R}s7lyhr24t%!ic^#-3^!BKBrViB%*Sb9RHQN<8pn%_*eXx%>f zHfD)2YvS8k%9}wA;Z>etsVx87B3DqR}%#7`_gsH zzWE%HMsHfo^F~_<$7InwVy7^)N3Hxamb^obIqx8FjrXC z;Bp$jX2OyQ zxpKF^Zo)Ik()Hx9fwK<>+(^Q+R^enp14bj+*EK&1QMm0u{cGaI4rfo|ZQ&Cv0#T1I z2*vN|L76N)joj%sFB+~%W0nKNxAnoP5LreRueK32hg&c)fgt|XHcV}1 zR-6Qt<|AG7takr#;M0M3MXy54w&?725#M*(Tp<51_^9lZKiMa+@db~?APNQ!VTgCe z-Q7LMW~tVRxfy9-?X7UqXNDgFcaDUUTpCw+2WYZ8ixfkErQ_=c0Lqwsp(9o>&V|0P zKK+pkH%*@=Xs|?ul|<_wcm{70mrA6-&flfQuXqrlvsi}I5#f?m9)vbrD{L>7$Np?9 zK_jWf6A60^yOa`~5*o+H$Ga5=TRvZEtwTzCVCnuOO3WCB8oS$@KJaP)b=KAc7oyQPP$`_eC2=@!pP~?-_Vcq&6@Ph zA_E2QXkl@&nC?yB70(`O>>z}W*YG`?iQg)AEgCoT&!BDl6}?=3Pw%bxrIZK@)uD@ zOpSEl3fgaZmfiHDt|tt8S*8Jfc-8XRXiE{G-r7?tC81E-?DuFW7xKbr19V9u;OP&!eG8vsV;$f4X?&`I2nezXKkc@@Mx^1g;wL zJKP2_U1_+9~96=SoGbe|Gq!RG+`^`zAtBh+75Hjn-&<>k?*> z$yYC5rj2KU7g}45Ep&bl@!RgKw(|t;{96AFSTIIXP_&8);5Y1Ee@1JIR24L;`6w^L{%$){P<41cs1xgw(3Bl$b_7yv5ioo z#LA$;AiQx2QgmNy47*MxBskk4FwM9Fd!U7O0p;{jS6wi`-c$>l$ktvzhEJm`!|?M> zb!=a+EvQ@S(;{ak3^##^UP{Pe6|{@KLJIOMmOAY#dw6g6Ga?-86R-|e-f(sL?i_n; zRrMgF(6&P5N>SY#(nZ|4TB1&2!lIs6Q~fF9DbrJ*XC@(Ub}avS-x!}Fx$oxY)G3-= ziP=J15PrjtQeKgyL@ce5Aln`WpIxXfx8{vxRZS& zLKdYK5W_5eJQ|&koyoee^Cj31_)>4jM|8}<3Ux30VKrZ&ohszh!l@575FGo>qnQ35 zVOn0F^UK3!H9%mk2 zrA~5B3|FaAMYJIH2~}`|O#mai=QK8XNaspwmN}ag)9Y6>OZ2+yO-Tbh7lz%y#1ar4 zu5>-CuVowEVx^xXC2vMq!sTZPhjlitA!g0uUl?{?E~tCHtyw|&+2*r9CG z>kZS3kZu>Vtp)&(DXdIAf^tUFqWXqwR0O~!v?dfN7{osbVNdp!pA#?tsg8 z^K3$1WPZv_)2O*M&Eh-1KxZlRt-d5g_uW$rQ)FnwsFEc4TK(kf=SX+)-jXBzX`T|Z zd2tDC0KX!A5mW7-lXTojESKv=4-;kD+H3Jw?lT4|dAh5I z^$656b@DAA8RHust)Y!WqSu$`=%>6tW!FuL%VxuzGfk$zrAWU9xE`Wa zJA0Svud5%{sQ~xc7Vs$|osqy4_-V*U}15yqyHgsC$LR;gdJ^+U! zKC*Une@c+#4;C$$dg6J>$CFRTL`#edhTg|0a8T@Y@U%s3`*autIGaSbl_?CqI>kMK zbRZVnw~Kxs8KYI}|74q<&3-k+x z9f(qdn(j=cB+tSPl~exs@u0*W5Me#2H$2$9p5y zq{1A%PwCK1yYxij4Gotc-0&wbV3R?Y{{t=rNRDO7gceAc@Q?5z`8puo> zY_AU$Z)l^{DS|DVt3W9>JDsdtw@(GwGp#b~1q`kh4Vl7wx3qlvJsuWIqVVsUts_&D zojLYv+xva|`0@MjdpN%<`F)b1A}T#?&}}w;@C2Ki1&}BSi{h>26ay7G-rRR$%sFZ< zH~TX!`NCp?@XM>u<=bD?)n*seRt-7v&OP}_=QHIt&OzR!ujb|R-jjw|! z4}@}BJ_(CBi45EQVJ*uNW z&bS{2XEwOY^aLr2gcqwbT0d;`1@e|+e*O~+;mmKd^kW3xP)Jn6X0GV&inA718&47W z?OW9dgT$aen-5F}purnorT)-A@5CHx(!0fQ!(6rgDL_4!z9f0656p^ps9fRBtksQ` zJTHA|jp3&t4P9cK1V!k0kuIP z0||c}RP6Sd`_4I*Z$JAvVvyg=sD{SR=!u$ErAVqFgEx*!3Vlb;9x?RraOM6Jv_5aP3{XZd;J)U4!jN^W z9nexYHVi8Tjd=4kTx|~PcQjQ0&S6Il>bCW^uz?n&ktFFrV1gy*5L=$v_B*-F zw189u*P?K}+T)vt!3Qe;9D)hssAMxn)NxIg zm~pE)Jz*{t6l=?@LiR(gPtJaA4b%HE)PWDbynCy{b@KCE#C3{%$pa{Fw}%l!n;g&7 zzH8=VTA7u9EY?{91>o=MpMY8s*2-utYsKC)PA|oEn!_6h$Q$hic499(fT#qjc^69@ zM5kd@{UHmvNCOdZH#eAcogYSJASU~|2n>48Nc?B~zN2xS}3 z2b3{qD*^J%*7bwt{CAm87~m~sR4!5Q<$CGsV$CqTbzk#J3$fJrdiP2Wi+kcMeMz0E zNGa3DcwsIa@-wf^jd6L0rK9i^govyC`{mmG&Cp8Xd6RnAXP&p?h@j%Cts47jRIU@mv%j;ehzw_g=<{;{J#`ys7LO^+T?XT|n4_1! zJP9dNb9bG@HMpd+8a^;EMb9NC-`uJZTx{r;-~$skp7vf^PZ&+_(O79qi=-*S;<$z= zeSb5chVqL9pW|}a&FuC{T085X=N1YkErw?hm<-wo;1E#boQp0mE<~1UO?OFr{M6j* zs_Dw2?)&!DLZ(coxL@X?Aj;U7@y)cC1J+bf1FtZ+MzQ$zZ_+rU38aTA#go1$s6Ant z7!KE0>H>|-F_I7j6Qli+p}hCA-#E1PTkFF54tGmF8EfYOt-e>9Y$1ZkvlU@h)W;hs zpGcK|ZNEW@r73)N>v=}GY$5&hDorri-rwIe@_zEjb8s&+&0KhT81Rg>d2?HWscPb= zb@&nG&N?}xM}tb@0XNZR-L|gcfW8M8S-cyhFiobeUrbVtgEU9tV%9%suc&nH$$pZ7 zW{eFx`BGhUTB9hjb(ow63(`aV5Xf&1ST$bTKk5pceOabg{UH*cUODAuq&O?ziT`>m z*F$%DM$2B`cblKE8?r@(V?S=+{fnB|;bPkPfe!7o4R;#1D#Z0QEcafqIsf&i;jIxN zUp>@zzHTKW@6ILTZG{kKf8J#8WDejmL!)y{S&m9QS#Yj}YOjf93H0nrKpF&m&0dwt z8NYbFz$ZLjPCa|%Nne&>NfQa7VQ9@;YAc_9)7d?SNjdFR8qvA8`XTDR(E*F1dDa!K zyM`3*G{mIBpC`EzK(N+GaWyzyUv_m0Z(~qPaD10F7@2dhuV9=4XL?jGkqx7-MCG*u z%u^EVZ!z>T*|nv{N%ojLJKsG}2429l}-WluRT7O(!3*eo>Gg>gUKgC-18J)*_KY1lpFO*mAy-$W31nY$87n5xO zXH^-smZh9)K?HAO6*c&_G+sSZb>G|Jc)<6J3R@rc zExv=IsERz=s)0>^aryp)Fi639fD3Em44ju(fgERcUFS0qc@t5jFAG89ED$~&o+(bo0-_KH}`X|%nZ>k z$Jo-Sdm2P0l;IRzuj$Oymqz>}Sd0g<+@DAJ0x7GlAJ~EVohdaiN0l(7+Z zB|ev+_(>_T_p4-~f#(TS9xZ^t^79)#!zbqJ<=zc8B4M<(pxEVJBYkUn1tuG^dBM6v zEDau7wb+_GYv-|6`nxVH53h>e7~{e^DH)~B!>53Q=Bs4acgFbQ*LAJ6;*vWgLhJ|z zKeyfsQgGdE2CmGlTlWk1(#}GiSd7t_qWDJ9im@DT{b#1A6NjE7JBrQwa+ckP%WY#e zEh#A}y=Etg^qAq-=X>QE6A4~(qDy_hu~asvaaQNK=x>-749N!XM(fFbj+57)@mT-O zyUk2P7?WB}YDXzD__re-d--#H3xCn3@~|K$miT*;h620ESoChRVoM>vw5qjabH7Kn z6*A=MK<)_G%n6fO469^E_l(r`Lrb(Ngm2nm{)PU0z==GN>ht8Y;cE4m?8&BIXYqgB zwG7&f1$R=iG^{Q*ckD=}K}%RvbLCX2Ywc8_%z{5QVA?)M+&5xuXK!Xs0nd0JEKR(A z<@StH%}ljJ$EyljKySEs6tp9ML+s^$kMc0^Rg&^pBNP^Wu>bp^?wAX1Wn7R0ixDsJ zqS={1Htao0xoQIQ8gqK%Z}+EzH8g1TEC7MIdI~fbZWEF6yU<)DI%|gD9|0QPBmFLD3{@GeqvF+>2*ndWBrU` zW+Q0Ms>Ns#Xs}UxEO2B0%CYC^DaGrSi*)4?=63$gGq!7T5qZx1!Ech~9aB{otQ$ql!sf^JROv8E*Q^%x2yfEaW8{to!A8Wr_|SR6+5{%yhSshVwj zE!P9H8~>mj^sa5}a@JaBj4^>r=BEZCpgu`G1MfzDBTw6Udyi!!+-Jvbf)InjfYHIEOaA$)4+5-zAlCW&O88+rqR~>lyUQn|I4mAWT{ewv26gR%qa^ifEJ>Ydfic9p zr7K-hSHq1wu~6119XafO3@`OLT=jC{&<`mEq~+Tlfm7zrN{wZCXh z-;O7P1$u(2ZDqr~;YoIlYGT-jXii8x5s!{-w#8n1Q{D+kR;EK!^Ut_vUXS?u0A$qY zzcXkJJIUzx-aI?UzEI|!`zOJX$f}7}N=8q$nab8#Oe2BHl7sWbS9NK;p>3BZC5uhC zI*oR=to2qJgw4OS;(kN3yn|8X?hwfH?&S}r_*Ofr#PEyT@iGUhSti;>`%f1=Yj8Cm zzXdS^hc~7c-DeDaCoQ8HO^NddFTzvV%$d#PiNi!Qk8`yUPE%N;h3%T;dSs^b7!gEM z@PG%o(sm}vMDCv~wIGI2J$Wx27mSDKYA(fw_SoqCVit1@ibFB$?g8*kWRk%YG_V251?uu)7@QQ&1 zky?d5nq{6)ibiMe=I|?<-JlQMVp(;a)SJ#J>xQ2yoAaHV9O(?Xt@U15yUS0`i)VXM zNesgfU5e>es!u-EpQsR6dI%xIGk80wv>IRzgqY0GCFgyky@L3I8I|Wo|4QbPg1gp; zIjTY9;~5t0A4)y~5o!4;+htYp>B-bKCmd8>4L9FQA^+U!^2lY^l41xSqW`i08FY?{ zM5O^p|Cs&l+F=WtY**I)Q6Fy*`t{ z{+!M|0|Pn^%(5&`IWr2A%%%MD?Bd4_c_a3{M1tvQM**D@$|=GNhN}7az*s48veQ<6 zXCV_{8~=iFEQvN)wt}FLA*w;UTtO%^@?6#i`iyu+j7BIeWav1^#!CxK>%>e?PFjDd zPpz~;?6H2cc-cLyL`l~!5UY}rp8^g6?zDk->-s~BLP@)~@(^`d+$!#9-ngs&daXNk! z^eR-&WF8&dyCVcCcNg>mZlqzps&|{x)dHhm8Af>TyCy%)usQE>Mv=}E-bR>C!td5g zF>Z`=zkt^{Np52fi34~?WVR9?!Q$c1D+|Lwhsc&XEGy0D^h=WKSxC4~r_A{O^Gbu# zws{3q7o4u-co0G>?Omz%P-@D0SIldoc_v2Tf6%*4d3aVio=Ajncr;!E)OmJ`{WY|e zy~g-X^myQUeeI$L(2nwz|J)r2dXhDq=mf1>nTI4Tllwk}9ft<5b(Im8J5t#sfmjDj zY3BbDgpySf=Zn5dTFiIg4~(s#E^u|`YG^Q&yuP{s^}Yr#6C=*IZq?L*lZWOVkQ-X7 zdTxvbr#jNGi8x&(BkzQPlRLPJZPh5qg>m`x=PVmht2!&6cPpKn4$LFPY|Uhkf3uem zFLT=YKPfOxfrjKNbcG3T*+*kP2Z1*)rp#LoS^wQT>a+$Pl>rqMQHWmZ{Ey#T$xE?eMx`!mgZuoyB)=@j(o zcsvL%WKf_?ZfN*a{AnP1?z;Lu2{}T|L%DH6tzy|3)*y=vuyMsOqEmi(7ssG5A__lB zBF}ccj+V+_hQU{F7a*x8RG4a+N}$9m+)^T0GTv|_uEiw#Qd-;>kjiR?g8$bU9evOe37JChfO2o2U!V=+2cJob=tKgew`l!!6$=Z=d!*|LY3I2`3^K(CgJf(#9Z(UIe)J813RH_9Gk!OD6lCfKCjW5^2H88sN!`bFZ7^LqH zaG5d$I{EO^$%ip%f4p*1d5N$?4SM@Woen0;aQM)zqVy2>=WG1M`_C-YCDHlHUN`o! z+4``PU@hs<`c7)l;#ay{b@uI|L`h|aU_5cj)|(mRL!^U&B;?4qJ|~%iXfs8GGOBDI zTjvCVbr^hnP@$IK{>CH+1h3e|rKc(FRq5-|*O0ZY(A}^vx9)cFr1CW5-Tw5-K_h|G z?COVM&8oU1g77{)N%F@Id}6A#8Owxhkc(aL)|5?<>H!7LU9Pj^u(Lz{R^u1qUQ zf)SdM2fzgzSwhg{)ZcRXY; z=eTeMIpl~-Y*sgEfE|Nqo`LRtLpr?m*{yf6l3p89j>U+hIS(G7W9z6%rW9w2G7x}~ zM~hZ3XVRqRdkGKerKgD2Gw*gG%wan%=467zw9P(3p+SU>ITV1?DWhrFkBI@jfo!og}YN6XELKY1;&?+5yt zYH=(6*JG=aom-l7&a+Vm93 zhfbTlwAv!xk?=)ZE4~yU4&zxy<{!qvFSdZb>g>e6H3{$hVz?8W6m(KhZh?E69r@hL zTPLziNOE~$P2l_g%iS#P*)lT?L)a7mK5=%bgy7&O6KM{}}^t2CX5 zrt3wFL<(!;jgM=E!$V;tuF0UqU^@^TzI0LS$*>|b18?7+D+`+!#4hyPE2DY_>*!j zQ8=pZNS1zYQ{C2eOWAzO;_+qAQWQ_ulJL|r>JWXagL3LJC6#pYDq znzA_bT7J#7)8~qf1JAq7oi-Gn1#?D~>f-#qo1}6fyAnijKkjmT_9RBcAh7?NIa&M5 z!gNh8aY6wY+5McuZMzm|%YV1=uSL-oLrfg?G@g9~1T{Sgwf733Y#wEH3S-o zv-chUQ$;0yE6FF@58>XO%n`Kt5&{m%b)ueB`R|qaRko&@N8?PqtHnA-)PbnT-x=BSJ#X_`z61{-E-*u;$Vi!X;q=<> zMA3a0b6l9PN22~!0&pYeOGft$t@GvO1qldaYX7JVyHi+-7GP`;%Zf0Km+)17G!pbZ zy&_IN>x^+TJ2io}z-18U{5<=z)mtjgIyvN`6tO!bNHV=^DbbXC)4L@m0STh7Tc{hdw`xL_ZJX43 z9^lqcV^uQw^quH-5-og*geUmjtKz<1+5O%5^(pGQihnfH-ucuX+^@fIx7u`1Ng29{ z|J8K!D}7eK`eBtR`yZkh`jPlEvFWq{|9u$aro!n*$XqlP!2Kd8z=dF;vX<)iOMwQI zG4n$2jGRs~qRkTWVVd~!Ev6@A(?}*~pok!1kLN0f{e~{9ZJG;LUCT^rswcD+H)f3Wzgvn-qak;TpVMzEwX%YcMF$ElWf;orcqOFQ*MC=}XJL2rCe zbh_99abIq*M0?AcMFd?hi^8A<8X9&Dyx;bc4+=}t@n>Id1lW0N(gEr~Na=jM-E;%9 zAARQ6rkE?eCL*O`LEXKRgYGYSjbGW;OalVZr7r?1U`u5+MVpZV7D{aB?*@M2dF-Ca za3MAMeL#_N5rL2&<;tI9rnXp+-np*@xWNL^=ixJ^M9%CDt~WW+v%~oU|F9^KUrt9i zTP7jiX39t%p>yAC`UoWa#-q=+>5nT49arC(Ex6ZNjo0-rHxTN6|lUGuR zU@E=5d+>FaKyyI!*xdM}5BU+F&zwFNHz@dKHKx@NtyBVgd3Q!lvfhZsEW2-fbns?z zK8qyG(rmHy&{UH$IK&t)B#YMzfu%v%wQV|mRYXtMrSRgmUImyO2o=hQfxG#jcN!&3$=&Wg#&82rXV2{`z_Vcx@9`N6s8c59*V;Yy ze|uD)-h%Z8&R#bm-|5L)OpY#$3Ao3;r?+8a+@er@Fuyy&IKI=&y9)t#zFFqL96k%oy4={SDiYkSs4mHjKbJPQwp-1v7ZjN^axS06BwAT(yecIzA{`6{xqB{G61Q) zm-4ot6NHCz;G6932?88n;bG_ItJ!YDn?Ml{!UBrWWXnE6ZAdML6&as9DJZsxwb(mMt++PLjlP`iymF6^WF51LTBUre!L_CQu!P@U{?$y#Sfqig1of9%Dl z`M-0t>|a#(z$DRZLs&~B*L`~E&^OOnFqf84EnS3p9OCB%75KIuris_ZwjIT&GYB%s zwVA`k1sk7ebk$5*#NzQH@VF zk*|>NCewg$`=GPTT@7X@j3 zw8)x;-?=XfWK{D}ru>w<&&P|MR~S(bg=j|_;21upR+Fxw@YH4pYVLF1fMXWeTm@sG zi9S%wBdUj8!aYsSh9boc*tuHR=sIyZ38D`rZjq9<1bjX9CU7+r@6v7}b;o=}q#wRD zz-6&|czy^q>?LGR6Yvg@C>r}t!@@9ta)-fx5T90Fk=i)+J657hj@D@74L>U)=SRr0 z1(yT!{Fp61H%Z7<8iR(7A4#)~FPm9aYv<%EJ@5Z#e%;aXMNyN)6Z2gvi6Z5`>onj} zBsNK+uS-RavF@F&;Dgc_qp{@`eZ95XKW*6)6BrQITpw%~5;hFtyX^i-O&BHm3ZoPX zMC9a>V4Y~4+);nW)IYw618N+g?Xkg^L4c!HL3yaJpq~hl2b`E)RzQkxwPLvIGi3$~ zaoE^UI+yk8FL4=tfB8pkyEHm2%8=WI<>S;s1#2+mSWD}NNORRp&F&g}GSo4D=2Z1dfw{LP2 zS{#2=M&vrqm^jJKfQO?@b+IDC?64<09BHCb8)2R&)_Tf6P%~DX#8F7WB$lY5MC*Y1 z{T-dFeFq8DIM_7IRXkqk12+ym0avjJTmO7afvh^Le`i(wj(%SjtW-{A;E#Md|GPmD z870I;;2$D+N(3S7;xRuDydkAZVb8ZXNg=D;CP%1=IoIDdOXq;i zk#Q~p!ER1mwik?ofF-A8Cq^~cA1Dh}6jmIK3pMG-Iv9WblP3HbY)01?oEDf$iv35l^G&`bg1&um(1ew4f&8iAdCr zM=Q4 zd}k@*8D$ZE(A++a2Q3or_niBaXIc+ey83K*>_k|vZV_7C>9KwZW@?R1 zIho1IuTytQeLjt$1v@f5JXEP;qL8>LZfU3;ed(p7S1d2ntBGjqAn#HBewSV7xxBJU zqoG%>FS@?RTuGKmXE}aPSgV2pbW6*|Gw|F9-ZNxFx=9nb^+R|X?M8+9u=l_6I2{86 z%o1Nl+Tf@0>7uH2H=lXjf9L2rv$m~2|Go*OQiG=E=C4h zLk37RkuF}tu9Y}SRw|~OfBuYE++p{Cuq}E5dzLkz$@Q^e)pHp+B$F>kCzOEfn9gux zcZKCMt!xzVrBG{MVlCqdSR$&VBoZ99sxhS6^3lc705>a(p8*lcByV7_+)k^}lS>0^ z5DC!~I+!d2zBgypqx{oRJItsGH8r+%2~i_KH|-dDs3#oDV!>)>X2w7 z0t@P4x|YjmKttdUvkD&d%{KhUtX`pe&6X=XXuN+yC+nby#kCE$DIB3FerfNj_+^~d zIQ|8&J^>)*IY?-*TJz7pdTa)`!y5O^W|Ap8sA`_tG91 zUUvpTRoLPTI3?%7CXNUFY!a8DB`JpMjJt3O&7_~aSnV{MnSF=82!Fce51(UT=_9F7 z73!JyTfJDdfWzijtI{PsEL69uE7uE2!Z)#dEAXyEdv6kqy$x`Snhm%>LGt^VHm$H; zSYZs4;%2lCVY!~1c^Q0=l@6&pwCUwLLr0!;^Qr)%9P6a;IIg<3$)R%>;dxdWTa#_>*7`h(jU64 zSxHc zWxxI8NgZ#1%{bmojLF^0r$P7d+^z^K<|M18v_<25Vr$<-T;bkT-VZ}W`(;$|uM)(d zQBTFFs3wS6*z~dJA*1$fJ5CkQ_CV z@#w=}rfG^v%wj&I!T#UdRkV79GTXfw)uG}G3h(IoVyMWw^%LUub0i$c9K>+fH!13% zivPdwi9GPewYug>CyVfH19K>X0z78%K1d^Esyb_xv?BRko9D!jLO54m6`NESKcosX z6$^v=6(gT#js4||!6UNNJ4pb%!P`$dy~H9jnmRgq9tKsRzM}W&fl%F0^3!qy(*A6! zy}5GiRgUK6{=JFfAl)bfq*(3z{ue39#%T<^i6Laubv$BIVaCZ2Irfc7&U zbtH;)qSaP#;8rD(VaD1YBrk=>?=Xu|!hE?!L3X7FCra%=U7VRrK-(HWVK07WKzDB? zuDi8h5p!z}{~Dh5MS?ThDbZFhOug=Gh)cRS82CI30uX-*)b!X`tq@yW%#aRD8)=ZR zLR<_}mu+--*~q_O+o4i<-YOfVjas^)eIZf}rJU`nu}lb6(gQ)lyspr-@U1`4lscqC zxm>teh5|pu592MTy(m7-a?m)E)zTjAZk*(Cza1C~!OvGKS*jYnLg%xrt7rps!6lbqBnz6>iB(_=Z(E@n44Uh;ENVe%+CUOycc2 zw78?}>D)r^1Zu^z{aC&Z*Rn+jq^~U1k$$MN3GYzdkqLZTdY9u?YD>uJN2-gN`F66P zwBCCJwfoGbqr+it#P$M1D5=~6Vf4FobjaZj@kFCCB^Mn-tpR*%-i(7ZiD4vD@ro`r zG^hM!Tsh%fbN!y@h<9g~-zCKK$|!aF6H@oELD%Vaiv3v+#z(NbrA^}?)M73?Q0#Pm zLxuS^SFeeoY%x)4rxaVsb}j+#ak50lqVDWYCFnF z7!tpoyFLm!KK9C;TuqvCnZhBcu-tymidkQt8N9Un9x^jv;$BugIw99sfcNy=9=d{E zKJQcTK5iVrqvLdulV8~^kRzJtg}`?QqaliN%l#DA z?62%~HL3jd=-TM|p~RQZ+%C5Wxf)HEou$f>nEJm^;O(4z&$>h>6AB}FRLi1@D!PGL zd~}}&Z7<5``0o3BQ@Q!R9)Uxj4=R4lW>B)vZM&(c4lc_UVIBF6gke_W)%I*fJZi9{ zTo#Wx*rl3~&G0Q)`V$(8)R)Z6xcJ_d)xUVg9iE@>FawUUiFG?l$Qq|RL3-Bbv*Z0&8w(EDpe3CX*ZYF;!v*@8$C`Y>{ zF+1FBy#Ce2_pmXnLC?$NyG?_fdLC0;uCd*<#t_cRT?5g4WA6tl#C@$;{a;MEo|R~s z+RSZ!(*$Klr*A>WVOYs?jg4QT-}2WHg*4$O)})q?rpu2Qmj?WzOwN0K9&Dj4V3u%N z{M|UYmHoI|Hn@nfDR=V=<0!joeo5r!)!mDP>PF2|byiH{>72tUPeC^5gg3T?CGPe$ zEq-jVsiC)?sII;Ui5LjLZc{SsL(nSv!dp&+Xo$!cjn*!Ic2aiF{p|xMtG68-oR-m< zp_ZXdCh2dA&%{%@!+c6dn4505P`Z9?DvUVgJq&o2dqn8Qw2wPhbB0y6mv~(}IfTLZdcgrwNxY(HAMpV6(sa>tfW8tVT)>W(x??2ZfM)O6r_<0$vJA z<$5!9N#uGWpIgAQ`)-qpKaL-}hoDNZ8wv^$OQ}6=GI<^y(^%e%eJr|lezH4Vbk+t9YNVCELjCEI`edoyD*cJa%mliV`bfz= zGbUgUi^vk<<(_R;QA_d(*It%Ze8GjA)^?B8`+D0lGh2kMZ;+=Y=q3wQMfC!+-w%iE zX_$t4i&k`X*gCS!cZa zg&Epm_$rzOXVBhOM3eFK(gV(kXIyaoQ7;d+XZbf?4Ly}@V7Gh4ZjcshPgx*8TjUxBDgd5-K z19|)Kn4(oI#D9Ar=-4)^aYpF>t{t(Gwe|duf-bJSGS@!GM=$I@DXWGM?K>K)L&!_i zpHaM!V?710`aR)hY7_6)d6wY~yu5xBhMJ(p3m(eBB3zzV%H_MSuQ{909lNH5peX#r z@&WygB^z4wrE&PLP^Mnfht?*f7`eXJt&9<-!r^Q>BD_8qIw^_6G5WeIVU$$Mwsm@k zx}AdYEKkx!MRHPTJCf~=UC}zZN26SR%GLZ@$z&{9&@qQK!}RR(La)hyvvKm--Z)gY zh+{h4dHuyjOQ!Hk87xV9&XFieK$r+Zjz5CW5Neg=u&4v z>Ubg5DDf+qhylEBo3VJQa{>e(agmwajI34iw_jLmaXlEK=IGH|Co)T$#^P?2hZ3a( ziFntEpM^v#vw!5T!aEvwdSQX%tCLeT5Chr;4CkVI8tP|}kHg4x&eqK^p?z&sb)uEKOJc%;$LPv>*j@Y%RV+Fp+USo2mZgr+r^9SVU zT|??r15L!5G^$SNgGArRs0*Vh2J5stMj~99H(!S2bIUm2$FzN7mK$NzuAI7#*GZey zG81dynDz>0LAo1zf3Tk!7H5R|+LjMlj<)7`iL+$K4n0Sn_qu4Mx1!Z*OpQErK*;5M zvfh9i^Y3l5fwi2FUZpcfwIY7Mx;hYW!r&!q=?qnRn<<9)dPoiXWoe4U9v8*KoPhn~ zP$)Miies;j! z$-w+^6k>SGbmX-n*5rdZv^4%}+x_B*QlvzULZMOKo=G41d*8a$bme-3<;A>6Q0AWI z0na!_TIBV;B1F@7Kioevd}ascLp1I`>~I=IP(J4sQ|FtVQ>BQ06~(Vo?<}FtkOp5O z?>j+)vM>0?;cV1K$n)2I|LptQk|sf=dXVc6QW5~Syg*YxgseqoyA0Xek_fzD|pDqmDx5{I*SzHk>! z6Vzeu23;lbKCI~vpy1B*IArUVb za-LnX+M7f!H!&SQXSLr z%E5Td1{}vRcDcFClYLE@l3@0KdjuA$_L;$ZRYYcD(MeHu{*v zJ}tx+kMqc15Zqsa``y1UrHJ4})tZ^^h*0_w9=1W;t@HS|i1Fv25x0ncmF(Z|g6#iw zu>3z!8?3pc|M&FCzp04YKSz50{YgGW|6^9cM_nrZbAIaIpLLS?pJK=F584pydJ7)O zf3E+(4(R_LYp%Z?$`tJX`^ZhL7;@2=H^W&nV=*x?^^R*1T%-wC=lglaT@j8$T+2r{M#Py#8I!{PfVsC$+AlFhpS|Hc2W`Lwa_-8jdpYCx> zZr7cu`$|em9vHvp+lL;kWD={hoqNacWabg$uoIvDuknK4{(d$-IMy~cMy96J05)&v zP->_7T%O2Wru=zne9S(Lp|ynta`20%zh6PIWsk*1LjkdySs%{E!o{Wf7jDyEu)sk} zR0{%XYHBim?^6n?cwVAT@#M*~p+6>|qWW$J9twF`6&M%@UISUpaoL_Y!Uc=AT`~n+ zUPR(E=Ub@3Lze}GypaA_IR~qWSTOa~+p6GTgq_{pHIl`ZmB7VCLw>OMQYI}r8hx}# ziFMui3~U|@jVA8r2PaK$iMP4mT%L|r)8984;e3~wN&0V`F{6Ql`uqDUm#80+WbU&V zGC|v314rGMc>)Jj*|vsFe+#n%Iw9CxnMgxJ!!hR1TnxwvImKSi3Rl5F@sC>n<3Qlo z#0TEJ`g2nqpjLn##SHrhA7l%o*=8@d`uck5u|F@y6$LEh>tcIg!REVnkG#QDr%#xm zn*6>~+4xR7<+RlQbzx_oo;?Gq`O^Min!b*VsATgIedZnrz@@LXuKj}(9U}#APXWA~O82Gjyu5`!o*k}j#PE6pJYLvwXd~!g) zuC<-h)|QGT6V6elTWzSqX&Fn0pJYtb_mV#0!gt8sLy;gG|Z1_C$KM#qDjzVdV3T5lRmRTiHZ9{{RHk zKK>l5$?_xh*)ZGB~4-ZYGy12SB87ow9 zWXUimB_S`996&Z(d^GRClvknonbBLU=3~j(&i?z zot+(jky@RtUcP*q>QI_wwJiP$iWFY9Z3*Y=p0~3NPVdL(s#Qx8|4;9K3RiAxYwM^; z`|^W&@ypU=z|zh2g#2#3y{|7+vhv3AWpGLg9CXLa^@uC*4t->mXO!}m)6IVlr1 zrWR`0P?op07?gr>{O9Y<=Nh?rY-humb!vN@oIjS8jnvxc@V`mV_Ps=Ia@#j!`=?+B zdv`oFVidg$zw^c;5uZz{b?0;*8x&YDGfw8U)6bVn!AM~>fKJy~tu-tyk22R>$g_ty zj%y76@z~!#dlA6H$bW!HH z^TW$h$TNR9M;0>#RDAJT>mT+f8!CTC<-g;g$`LrJo0>nQvA}lXW10{iC%f+<%up?` zxLh3^;tDCZMhharpJ{1n!B$J+R)>EMcK$t_L|tHPGDBIyUO5`2YZ);c$pSq5j;qi1 zW*Z8UyF7tSq!e|8zpbLz(~W$4*y3}ws@w5OTO0Bs7=@7I^+Otk^q?w}Zdrf(kzm&s z1t841Y>l@cEPX9-iG9q8DC~764mMs{qn4*_Jl*?VN+95}MKv5Q|M9a}D8|Lj`2xGc zviQl~9C0uT9!JBB{l7*F0`-{hMzTx(G ziCUG_B+D1SJ2I<@kIzHU9i~yv@YO+slsnKD1VPmb;}m0X5kDiERYMj?!0pX2P;6enIiFPeiF92gF_hW-S)G_1$eho z2RxTgfxq_f>oHmo6BF~lN!>{InyIw`As-=2)Gr2v|DFnQ0uqv_wf+>jF(5fT{7HnG zcy)Pce7g1BQBlg7msuNIP*70rzOk5i$}|{iiqpHsR($4!|3)*qO~5ged? z8GQ|LWcf5vrbE0Fqdlx#s%ZjyBz+cRh0#1Qj=Da#*RQr*wX5l3dVx`_(MvyH{AAkS z$NyD4($Hs7_klAyLNBF&$5Fmefff5~i;ovSL8f^UN!(PH*A|uB+L4|#9LL@w6P=Zb}{(yu@e6Oqt~sFWlKTz=f$ z?E8UrkMEqI;Tlk^E>kA6|3-;RPgs=L2Iex-_#9tPA~G3@6;&u%a6Eswq0~R^!=|?(bf*eb1f_;_1Yd2q+DC*~%XTs1u;Zw`5 zGD+L`=|8L#EP{#!abgq_W5N`M-{lF8MpS)v(&i6;M6v>6LGbbVa6f-03LZm_R+X_M z#>Hk)cK)IJLW>XMu#C4#V0ChG@{#&$ig!!_&7Qu#zN&5$hcZ7g2)oaH{+p6cTmV>; znFj8)w7(C)uXvkQuCnwPq^nE|I$oY&Vr9zIm^a_#R*@m6eYDyOj~aO(r{Vlr;lH1e z_y}eOG|=)Cy#SLB8k&2@%X>)tg~5{-OlVH;NWai`dk%ZPU|4;0eYX z$Y_fn2tiVz0{@*j3JzGBT@A-?1ui?Z$z?Tx20~3xQ4y`TkIz~|A^QY#O^Zua)uB&o z?PP^PznTD~%i(P2#~R6hjay|FHZxhVv6z5+v8+2Ys29KWhev{z2wDnb;o>t#Q(De7 zf(mOs+Agf({eJFsmImm(sc+{kj0PJJa z$D!;EARFYnpYv#{qMkyKLIH4q#kV7pf9xBw8hPL3l7?_Axe=_enril{%ix>?sjQ$X zic&UHfE$!3BEN=W_B$=JSh9T`UzF6HcZ1;r+3wlh}KU?<}TSWBJk}#sXke0x*U`kg}xv6*C3!z_W|n<38V;6Eie?A|xcl=6+zf z>GGd3Mg1NV3Sp#1H%LB?EVdKndW=fJU=R4$8BKLDbe5Iemh-9gU+m@@U5*kJ*ZPvy z=*PlbA}2hdaE&mNOpGbHdgS6$*b*DoVgs;Ry!NKX2X%TV8IkzG%jzjS6DX@z<+Che zL6wwh^N|ajXU+QRJc`)5`EouDdTl5YIOa0`T>T-|h9NG7Hup2$*sB*$Ap3ScK(*TL?YQ=bn$gTh$q2(ngCzmbJ@ZLkT{Ut1S;JxMb%h> zzwL2-QH8r2)AQrUt7&{J>%GZ}Y$44eAoYxTJZjHr;tj^?+lv@slU0=KW0U7DMH6p zCEol0-h6e43*wUB+}xZ24yc-~~g2n2m#tZ#PjI`@fIQPA<_$D|{eT<2Dtf!u=(=}k1ZRSnRfB ziVecJTOaja{-kWoj}=?o?Ff+a(Q0;x6fPqpqeUZ#0`@R^^aoJU`BI2vqah==gO2um z3t_BntZ18MY z;CHgznyj!psJWm5fJ|Wp8l52F@%huIQT1Zw=VO&dVhSK4aswM5rNc!D?B`%l6Z|nX z)l=hE4|o!xZZf?Gt;#-NEAlJ|0`2YX>FOn_k9FsjvA4hZ9*#I~e0#1h#>dLa8n_=$ zAO49CxNcyDgo4xhP?D%j-iYs0&U!Q)HAe*yz~bdvC2@#|EDr>|I;oOJwx?@K+g&)= z|JRk41=|%!)Mi9jGx}2K=k>2+#03Jibg)V07s-_=YHlAj7*++sLSI7=PV2yFO;U<^l=L8Az%5E;mc7toVs(m?!OW>T!U1fezSS6U0uyc)kLd{GQh?hKI(n zKsHIa-rk6xK#*PJDpvHCa3War-#jY;{VnXKB@xaDGmPN1mUN~b7)Lo{dFwa1mI?7@ z5Ms2cfB*0@q49gO`yoUq^dB%5M)8+OJg@~eJ2?PP10ZrCD>*qiegaM`w)rZPP%dVu zTJ*F|z(YbEt<*C-$*4LOsIo@s_y6Qriioh^vNlKQw{OZjGj&Qso3Hy4k%7nn04g4% z1{z8*9KdObM@rcuT>v8$STq9zUI``ledomP1S0=@v)6x7gT!Alb(-F|LsT$;&b?~T z&8EDV9_6}mz($gaqWaS~X(Mpy0{{c=b;6c>@TrTi&jcFP4v>3+L!1Qa2G|xk$FM!% zNcE5ss_p9P3I#GyrOgb_=0&)RQYzc?X(Y99n}KXm(aW>lNQL)2xd8ctGXAI6y%aU0 z`c=lk)tJlffH*#`vHZ@_u*UPxOjDD4Gf0C#Xhz%U4+Ae8rX(w2b1K{6-iC%bAWn}J zq66<=%P@inZY9tvbqo(H0TK;j)>l9-fqX-cpo2=l5_NgHou^i$s8w(ObZ@>n@@1*U zQ-yS{BSCpexG&9~ryXFYnS!>mhi4J^-!E5q9ap7i8(riscR)hw-;c0`kvfhFgVyAU zn%BM9)piSHq=K%ADfa*tifz5S)v>H;Hf)CjDvpjUfaaax$rcRbU1{b#yZ?*ypCEoB z{8v3>K=FZvu#gZ44G%v>ypDR9JUuhBaOYt}L+Zz$pQO4ELPkcZ@nj(-4o z6XhUcaUU3U!6}GFm;iUa0F!HfxGXVVs;T7b_S%@3Ow^aeCSxO>T2A&Z!tfXbtanev z5P{l&^qA8uF{ZAk&U^$826mE5bsIj7Q|JuP{)x-KlPS2jL^>FMP-8Ul**1=rq#Q2j^jOrO(}8iBpB zqSy>h#Ro|ek94aO0acF1eTaW2N&PdttzNbuAUZl4_Z`o;#A&fgUjD=oh(;Ze1gxtK z&l;UK^AEEm4ty^TmQI&=u}{fM4CO&X|F$L^&KT$X;G)zR4KS z2sGOGbxGRBSnx$(G6K2q9jtxtZq6$YgC)*KLIS!xhqFby0piZL7=iNw*m$)1DL~lo z8j~h)m?2&6Fc#JgN+fFyBd}yUjVX*}{ePHMakbov+#^-|4=vfUMf*5S_a4*FfQ@B5l>iHC_=+A zc@T-;?g^3Gn$m6de3=DD#A_P`6J{GWC~5=n%i<;M1&rCIQbG9ZQ+pPUO`U%ZwtW5T z`e(}00dJ3d@-Y{`9@u$eqa?n+kqLPIb2V0)9t<2#Q3fB?85B@{qFl1SMHoS^4h}x7 z&}zLodp~Yk`$U^1@-Vb3b%4D9#0J@Z)Mdgmu&{SXg5Z$w63Dzj1ObYn@vMdL;p5uc z+JDOqVey>53jlN@ghbs_AS4-I8?zVM$%~(l>g($}Di#49SJi4NjD(c5z`;G1RFGyk zoE*y?@41N@jJvA|K(*H+o>gruYqM&W>#AzbR22Z_uw!zPmuYPvO-*0S84Dh2JUy;4 zp>ob`{O;c>Q9SG4-y2rtKrZC|WZfBLO7fCZOCV_~7b`!_5b{vXGRer*fC5k(SL7`2 zHZB!I^4#9OQi%3Y0=+JdjTe8>Z|xe- z7U&`>h4PPqO50P>X$wB;NWH#XuSxa}L@OLnLFj}@R^F;6Ha6fB3AY>no!$+(KLgva z6tFB^1SJXxxCe6}2#*u-QBjHeQo9;^{8rz`KAFnpD`sAnI!lofWQ+KKDuDR;>_4|C zZh-UG=r*E&EjGGXB{bzt;Fu7CbDUI$HuGhZWQkB4RPrQc1Pqz9MWxigxI6r(PO!24 zS6s%P13Jzl7>snDuCt3}c=@4iZcZ=DL_lm#6{ZUVaUKd&q4*t&K#8WLp8eDZCgg&l zv`x*HMd5gn|C8kDJpR5lJ&>euxnk%59OZLdeHaiBu*zQuI%7F(XX~G?J4fI%GazJ3 z5KJB7v6~a6+_M}vjcz^%VY1%msx||jiJ4iY*7_Cy8#3cePZ%x!b0vNO0yy}OK_{5% z)e?J$-5719yb(fc(~NI>#mr0{twa(TF$&|RCZ78ciAbOrZGA88S1arV!4TAAo~}b8 z_eI3fA9b%a|M^JqlilTnXP_;hSi6P+hP6Qqp#Xg5NS7j_7x+v5vx<4Z;&E^n=PHfh z$1Q6z#t`B#N%)?uFLwhuGi$VFJbFr-mj)E70KXge7KaJVF!hau&3^IO@6t(^{6{xH z(d}Jblh?%w4nKEQmoNuY?lnCc6M6J+q-h7iS%u{0VkMT(yGvc1+%3 zQ%l(`v?P@i2#-kAWM^k*L^gryh2+*x?I=A^8D(ByrUgm>;EIGeKX6qR2*xmF#B;s? zmvcYg%S;>wAO;#s&Z^KY{=oO{_R7w}bEzYEv^7Nh+8QXH`AXUPLtRX2h4SDAU+LuRVSuD6!~brnZ!wF=vWZC_(j=J$#DU#Tx)6I^(+WtFzsNqGiPgCK8k1U z<9grkF7goR>9S9OT#6*16RZ3JUqjGAItYo)Xk zHO`x=F#VZXyZRk8C`Q&=4q)T_Q%S0c_Y5x?fWGOT4I&=i=UWGf+YXr~Cqq z#zD653^$k7)8Wy6q}BP|a_BP^Q9(-%*9GjWm-*6IQNV&hSUu8(q_80_gj(b`(i;LT8VjRZGKf50qgS}WQxIwrcS?y05UiAF(l@;gat*Rf}d`b`M=T&j-&OO)O>k<9|Wp?H)wY5qb?4BhulY-%Tm zCqNwB^@=owcI>aTw0fv#oI!P#tDeQg0&H72dIfT=^?{3t#cEI7=#}v&k7F~7AD}K} z36#ddR{;I~gjHBwN@@4}RKUEQ6pVF-<8T8}WMpTmO3Y~mioSW3dvN{BDX`yFt(ol-Mav>hq#n&C^iC2 zc9anWo0{=H1Ls|m{>gN6SmoaHbgQ=^HN1R-(lr+ZBC!N0UyIW^2CN{xj&`*aO8yPN zq0LlWCN1SGRG3yml;Q!RwCec`^h{|da4KcWyj59NFK|2UvIT?yrD=TGc_0!5f%>rA zm}~*|UwnHxW$AK41!R)I$jC@8g_ANIW(QILFZ5qaMiJfzMaf@DJuodB6p5tIYr12} zzXH1l>570w_Zh%3hx@H}lG_CoG?4BvP{4zL77x|PuCVi`RYSikUL|Uyl}LWKzlz4S z)x6!Jr{iCpJXu;np067m*ElY+rc_}%825%RGs~8jSbtYC zOc$(>$a8RWo7B?QCel?rQN%jkx2a__la(z2!lPsF_%96|!pM2s6XAs@BJRFLj!P=- zJ1HtpQnqRm95s1*P|!^Ww6CW?I+QsI^pwOKV-NFOp6|0xstdpd@mk2I3?EAJH8qTm z45ZKy01k`*R0k{~P$(r#{%LtyeRW4fG~w{%%#6;E&sArY$IT@-sKgMia{%2i-X+pfIzwu0kEIO z_|-ri6M*un6>Jux`R%ZxC)n}$7v5J;QN`VC#1-~>)RnY>E_t#eA)YuMzF!K|$0R55 zM0FT4tismoOksE_`&ToOkEnRc7z_ZjWNgUticla%1%o^3Q*+3>J4^yiEM&r-R0_mS zpd{}RL1z(Ra}krT=!{jdPo(86#_Vg2od9S`9+*tk^KpDgf)!A^1ht3) zKY+36Z-{Zu!o+0e>SlGKnwy0M-1eB(-{R0J_NcL=qN8^K(JB9MRs3EWoye3xCg^Y2 zUuaD!3ZtWkNh8M8zQ2A~`%ju3U0jliI`{=aJ6WOC2qbvCk>W#GJB=EMBr=w-e*XMP zGt54u<@Q}%rmId@5=<5IX&*IJU9KwK=E@_UC76Khf~q7gqxoX*+~a694tzxw3^=i4 z`&~q^kNk#fEt-)2b@0|+eo`#Zo=F`EbUDHb_o+%F*u*Al=PH9nlk1Z)CDlq8Ce1P- zfk#Ok_vts~GCn^6nQ-M)(g3$#ryd9qxgaC}+7b`Bn0N)@#RM0?2;sp%sfieQdmkP~ ztzj)&pfl13kh2t_0-!y>3T#eQI4ke*S#-6wUR;9G{E==Z90E!XXgq?o$mQ1OCJzI# z3XfS^J2F10CQp0hxyX&|f>KtiQc zFMgLn30XA*djx?7HEJ5fD4etlv0Ko=6i4Dltfrp8}WYQ!8&JRIRIiXR6Q8hp8%XmrxptmO*Bk-vBOl=?| z=9^T4b`tqaftQ}<*e+Kll-f-Z&tNiN!3|C;#rW9%6%4ElKgzE0I=2OPPHsM$kNolD z$5rYTT7Xtn&j%PaKE4I!f40963oDcImH}STCznP8LIR)`k3nlw2xw{oy#t;6KnZFB zx)TZDj64$>Eif(wRFZs~v23@hXi)F`9T)LY3vaRI5^+Y3jKH5NtE&^&Y2=48_(z)D z?3nbwXbIR}I`V^#5H8DcI{kIYo!F^~rLUoK%(;2kdVwICD+yo#))jyImMsO8ePE}C zU7SuD6s%-&n9BabTM_@2#$g(vP7r2g1l(uTVFYKg2G%hGb2do+DtJVyZlAFLkXE|f zvyIRw(>4M?bF}(`co`H@N2`I{xGR4r)dyDF|B+ z0001#)e>ic?jJWdh~q<_uEnHXEq!8x{l9f#sjJ|b#m3eob>P;6K&q$R?FWgG8=#%< z&b!ku#|q>K1_62Jg4Hnfalm-ARbQ^z$esJ$?ww_{B z096a_X;#N{py|oZoSg0I3 zy#cf`|F}rCB%kU`g~`gC6e54V^FwVq1y*C}mZ&i8Pd4iRUSVJW@+C1Sq1D+fOu!iZ zsDK26c6IS@0LKCN?)X4Dx7--$dsJ8-%H#%e`WURv;HmJaACFaEXj*`xkWwn9M^1`n zXbIK`OAQKgf!iZRK!~vf{i{ziQ-EY^0gSZpF~Ilg7(lurc4Y<~W4a9FpbUvPnWqFs z@*Nyjq6!c-(Wd7&98APYKJ6XddE++B^#8&=ahkDI|w^?TV}`Xq*jZ*J#@vB!bON2E(*-1nZh z;zY>c(eNrd^ViNqk}+7@lq2f{BqS*>GN;>XkHGy%`tT`9YeF`~ARyB=RU(;#`X4Xh zJsbn`5KM9wW=a@V{BSRirAUc_?qbrw&q7yEk5V`TNBo}rM>`O|RC5l&U%!L0H2+TX zvQZnjWiopEM$la(z0K=pIa6EatdF1);`npG!k!;XYZ56&;_aLko7#iNKpQj&yuSiM zzTEv^p!O%z-O=9ubkwV>3f9P(-}VcP#~A%!GvT2!@L?^_1&|)O@bU3Sl1qTQYdyw) zJ^+G-gbod0rs+TclJ&cmq=qXpevGw&hu22#cb(e{($>iVHw4yo5;sBb`{YSLRKUZb z$4$1NSMmu$HG+eFUka<8J%3yT9%EsNdP#g@XS9~mgRIkIN}8%Kxu;>{b#?{{fcTvF zV&UR-KiT4d?A(H!m(HkJF5M4fLBL0{%=i9^yS!ge&PJsqy8N9tC{W1daY+2XE^QEO!k@DGfhR~eOtWWeYfCu$DCr+k=ky^ zMp%L;PsmBQ1Rw2U1FXiC#R{?fxI5}`!lGUM-sjxhf`%wYkaDRjsz%txPU{&vsH6H5 z=~w1lCiOmLgoUKl+wOvvL`@(#Kj@brsy{0O;fARw8u(#3F?y;BevI>+`>?rrWWEpR zSH%&|qE#ZR0R2~=n1YYUTT(z-zV}nC4|i7=_y}ZP`J%)<$giMTO^E_nQ}i`yfg8 z`p`6NrwjxgY8v>g=|VBa8FM~HtiVTyt4uCx55<6%lOD-~Vx%f^G03CnnZsY8lNaZc zZW;U)T;Hkrc$6oBPZWgRtodIq8pZS+5xd-*y_ipg4q<%J(jTHu%lBv{OyW#NU*r>xn4@?UXR`4u_5UNnQ}$g zaKQQ7V@Ck;v6q($;<$tH{Vl zR&}>ncYHQum*NtA_Gt;CS|EE=hgT82g4$Z+j_Hopjkon>XT6g_xjS;ZKvUmCA}dei z`asOXw;LSmbN0ffNDHZ44EQJQ_RVhhfyhC#5piywGU#f=Z*Gb=q}ZMo=Zh$5vsNc|>AKyYtEm}n~$yFbmIeOa5O#GhOMUB-5j!z{lCYt z0c1OpPu*O;ou|HV8);F)(JPCf*sN5!1G3Aws(yehjqbBM&?c}I$0=MEh=4*s6#u#0 zbVsY%v&J)omYA_V85zI6ZAAH=J+lcS3a-HOXcV}nN3W62-aD*7d)K&i(Uz962uR#@ zS!Y1+VvYE-d;)z}LxZt!PiFNFLISaimR%^Cz(I=(b~>^NVqm*rz6P7ofMHN9h+iH6 zwy1ge-t)u=C8mA@Oot=Q^cH=)!%2az0+&UXuieoVG(5@2(5!hfX+4l5TxrB#BY74) z3seCdu4;)q6DmfC49&6+gaTe!>R{*daG&FSBDn35WVS|?Vn?*BKt~Ru`E>N0DvV~d zEsS?LBkR(UsG4N+=8~##pXDYCUr|M8U@kcWPMZS;>RYUD}_9V z$9{M*bf9M`&9=3sd>D&Z%M+y-!a+ksl@qla0cbvzy8`FCdIw=7cHe9Fb*F(a9;2x9 zNW|5PTL){3+_yJqGo+O7?)UvLZzK5-MJAJ@*m3P$+&QMK$VEQpAGfaD0_`y}XS0sg z3Z7RhpKM%>mwRo3L6tZt>q=7R`Iw4Z1*0}MDqIrf=jFgN3iS#c%IYO|F?jfKwWXV- zRI%ppY$&~Xs{fJDkcBi~-Wc3OAe7mCpHB+nz_*30I~>bnzhhJ#4DfExu)kBZ$l1Hj zn%L&|wvD|FWSo^ET=VU2W?6DAS9?ttruMhcs*q!v#O>I_=Jv0%R^<6wbh~pDFIK#lJV~%Vh+wy_l&?Pt z@EP9F4X_C_>6i_1$Mk&c%-7!YfM@(XtyM>whf*ct<{MekBbO*o{_8@>ErvZtWyPz5 ztH{eG`$^`9RDp4wogy*kVJk$v-vFHKTX4(BekUZ+T(bR32H9GTWZ2y%gZ%J^gRRt} z@RIPoGd$}B!&4g>%1<%} z}K7vxUH>TG0MyRDi zSmYZVU&0;yqjosvW`)K#6^zRF*jPt_{II2~h=gKchCRiU*aXKM%dNXZxw-3!ppXVY-}@VgCgXGklVB}Ta8d^i3KB|+?B66FB;DW z1<-j0t~9Zwe64LqpvpNd$wpnV_t)PlG(xB331B*FTjbgb~$vUEsO{ZA@=>=i6DiMFh))Z>kXivyogJTx`E zCSAL)K0e*7+CjaNhnz?p*M^uu5O`YsB9_DH0qtyIca3hfKL;i@yjg(U*cB>6Rux(w zhx7r2HXHd0j-gfMaTFC2Lykf&5gBDwAaqx{c5`1t%qMu8-%LD#O%hImHdsZ*Hl?>x z!M6Ey;sO!He2P?I1xb*I$C~b(m-Od(M1=h@JLfm$d#0u*?h^bzaG8vv9P88GO(M_)cjq*kE~yu3g7GA(Rh3v($l|9$CmZrk^TQuoF59@ySDW6gsXp0xjHN(E$kT*k=h&U$~rQAinxUO@BFZPK==nc_A z(R`78Ot+hL3;x#jbxh&gOU@IP@AkIh!id;yQz=-*oShFH`PT0b#hHIJj(FKYi0(3A zC9J?Z_57>b_=1tnVrrYm*Dt3;0+LhPowex#E+4!}u4v9ky&jJrZy?^2PH!Z_%ME-h zQvQvvE$BGaV|Mdfuja{CH}fozKd}rjOP;n{wI*X^%s9M%&ec<8lv%#MSnq5jCY=6I zgm}p9BD(8gK=D3TtlQ@*bYHou-CZZ798KQj=lhyFb)XR|aR%!O?KZ}#z()@KWa|4i z2L4oQW%LfB-IS(EXbor@@xTao=^!MvI57|2pj&SA?eKbLr;FYy=<8K@g}x(aCR`D6 z>6Ia+f*B$pxl=(XgNXiSz*AlKs$mPo0i#3U-J(mTjzh-cG*acwNLz=#B+@**uiF9H zP7JSk)Z0rlp%^Q8L8GW+a|K)h-L2N^ZVtM+x0tWsFRa6!0GDKR;G{LID(5M!> z>k&h_sdG+1Ptb0d#%Q@|m5qwKLSLn_*&^I5%q-y$a^4wYbekffWY%lF{x1-|g%c~E zesNm-Y^~)7>C#Ns&9`_v5c-;lLgrkV z^S<7;++e(VEMur+=*-O78w2f|nlBc7w$J3`@%at0r%8J364AoP(nCR2JciuCb!w>& zLSzLdBs#YsOI+){yA5i>NW@MvaZW03&<$E(*^UhrsBESDm2cE4*#Wme9{=8qG(GLv zS9a@U|0=&H4H$fTFHXrMf4#h$$jiO&nb(l~)d7Lx*%b0!?ANI}A{9&-W3}Ryd)<{;tMMy99hkqn44hx%Je;SrT$CSun{oWup!vN}iU6K$> zG=Cf04mbSLAp80@ga&%84nn*jB9?maRN6$A8VO!{KBS3y3^LFS5)KqCr1jYvb-9|t zM);aGzQ5J8lSO;skfS`|6S@a-13AVR(P$34Yo0<{Q^+PIaswJ zeXkPlRDRZ`h>dD5g!I-34@T&S~}{H`gl=4m4awYJKdwsnys7s@5oUnTsR~{2 z5?*Pl#d%dFPq^LIAS;F`N`(^lu-h)Rfn@J}q!Z`VLJip|I$MDMuT)YkR_r0|;w6WK zcs9vlg@lD5b^#lhow{tOtleAJ&4`VQs;PU5UD)W48xq4qK4vu8`XoHWhjX^<@>^4}GOX?`j zZf0qA+&YCWpyrCt_A1>_SJy`ha^t;Y)MVktn+95CQ6XH#<|I%7#}gkp5D;22@a#38 z>vQ9|`#FeeTMlIdk0>d*=j2l;EmF}BMC_76IMC)%Q8;8iTVF+|@i8?W2Ljp~jX|l+ zh!XT`Q0xUf6_HJ5Dw%kfdc(iWLN%V~2edyNAXFkm_W#XcN;dy6j4cosi>+dY6+3pl!)h>M0<343X z4P(0=1<2Ib4=upo_nT5Qy3GIr(?50z^oRO-gdf3a;Jw`?J_CxwW&XYt0rTrQE^(dU zk%F55Y&Xf>5H3{Zdn6w!m2zgSV)Bk&L&{41LaLn-poKqC7SitP`g+#^)Vk}v(LgX$ zi=EK(BYXsZXL}YK2@<0q4=$(v_*$U%P99hIm%SEa#zPf#jKER%L4VV$d3}`~~ z7Fgs}01M;eHYhGyIOQ~P)$n~*S{<}ibAR}le0Ji1zX?i>Bt5F~JK8l6^!fc*lA!|_ zke>;EEPG@yCzf$0p7T_q)M=g**NjQJWg7tDV{MQ-7_ln}2P@3ms@@G_AZQnD`3_zz zQl=?7*6!Zf>{TRZfBU?_9prOvRhczA-!+y?0w!Y!n7nngLE&e(PM&R6yHFPjE5V^O zDSlDhmChm)LiI)^4w7onHa_fPvWCJBX z3-naA*q;pFPj3IC$S^oWR#Nx-ZEEB*;Lh@|Fq`-u+!v)vJ4oq~T%Qmr2iQPjzH?xozU6wd^5XaUr`dI! zhuSk^RN6*|wVXK#ZNbjmUt*fO#3y-(e{i=839a9^-LmQ94; z{tnnYCJZ>EN7a-ZArt2GiM?7{gz&7 zP-N&zakaMh?_YTq$A<^YCG7b^lP>lP4nbqDl176Ntkq+^ zYOz1<+@udSJ(Z}-TiyWW{8aPnt50g39ZxK1cKPIwcAlqdoTb_aP)~h>R_(@e$efxB z!dAIRQIN}Q_@~|=Ayo-<3seYNfiM8hp+C=OK)0~j#z_0!OfJnPS{5gbhDFq`Z?osw z_UMZvdM2E6&i9?q?oR_-?ll3dZ2!ZY^9ker*I>D1?!!9ziTYC3BX^Z<=nJe`A0JZ= zd-=z6VILD@>QhF~)#s^A9U-s}KjAM2nq{|dK9p@qwOv7i*e)I1>yXxy2F7lQsTw4^ z8GRhI#u?sQ>ci*W03A?At2nfnF!$T|V^6(A9>}HDyV!%fjU?p3Nve6I9EON@e_OA? zPd3my&JMUHa8q_yKk&C9q9Dj%75=uj#K17Tg9Yr zHCjaI;hI&{#o#xfo+-HUXLT_v(HC~cpJSKQk5()5M?fW%O8ov_`^-nJIFoO`&3(xZ zxo~Tx%LLpmz28Zxqku|Dc=tV4Li;gWhmhv*Edh+0&U>PVxS9$caJ zP=>Mpnw8k;4e3Thk=3mT*_>q4xCAfFJ~@7Rdb|ZS6L1n#x=wz;t1Z&X-UA!n!+Gyp zb$YAr4~{6&bT~5jC(GX_Igu;z&QPz<8d z2^0w#LBrWJ%U|?E+H2{4^?V}Jo@=k|`KGsQ1l55`6~TiylKI}fITepXsLxfE;5gpSR!{#sv(gcJq75D-cVTeRSaJTXny3QgJZLdeb`72bfuZ z;(KZ2I=#xup+4Dv<*E-Uxz*g*(?d>UsunW}7fw-?E-}wqznBq>HZOqVuISCF;=13q zoAA~~0)X_ko=v`9JqAup4z;h@q4czWLWog|v$Te)FiobPmLl>~jg`StPFl+M-X<(4 zu99o7b%mTL?!X$BR#5My+bC6!KQAPCX4%mXveU)7`=M2t%)PAqAC*5pTRrW{Hu(@( zaf&foLRl%%Brj@+twuejeoT9_*+%0nfyNt=7l>%UQbB=)a`3h~@n6zc%34&s zaZzlQxGZ6UJpMDobIOpzMx7m-+cumS7>wK-d@-RCONU56U;FIogjYUj>CGScIB0U! zD`Gg8t}4qs+ed?m#MEXlu<5g`kSA!&PIlKVU!k3`OE=C^(47P-JAXc4Tr zHS3#5^T8d2oy&c5_Z>@YFoHB|Pi_`tIwuANE^wP%O;vMT#nQLCr>3=Wp~+ri<5Vb? zC%iC;x%TE3LrrQUw#THN~zMWbPCJokv3 z%C-znQoR;q=VDI|cX-UUF*y4Wkqvrw681X`yXCd>I!@GV4N!R;0e<||>w%>jHX$|v zLtc?9{CvAq=1lBY)xe?P5dF&I<}i^;{2ZesvxIUjpC_4+)9eZA%fUqb;_0NUYN7J_(7hcky&kJLpDWJWeYJAU(pIfQ@LcC4@8i!8C+liyL~L|&EQ!+x zHW+H*e1G`a!=X{f9?D?suhd}QavUh7&rqo|a^->w?xz>3D zm&10k_iZOVH<^6C-{EFxDzQR|6Qxn!8)bxMsZl+*emh{mPfa3+)6TXO8)UQIejh_% zxHx<#OM!UnNl_drZQaRKFQdIfiEgeZOz_NJtEfik}1dXnZY(O?;uYq|ZfkS_n6 zPA_H8o6y(AzzNb1&?u6=QKpgVv+A+Fu%FS3BYM`VEAG_Jp8Rz}N1XvJa`BeST)|wk z&svmP-H*A6N0sBWD#{8sf3J;J6xN@)tQ&QjTxHaiw1AdaVy{<$z|BdmRlo3#*5j1{ z-sS8ey$D%0Cv*W$O#^3wPBiImU)u4P>H|SyVSQRas}9Zjc#S-oCtFb+(f5A5_-JLL z{?E}(F`{dD=f-BmMPI+RJmD4*BlkPMM!s?-E%dWx)G1x4gm!w9k855N9RSAQtrs50mJUO>`G$Snx7vsJhdXxq%*xQ0;iz;$e?u#YQ z7>D_5HUew%Vh+Yr6faTeemBo|r?=kM$^f%{s>A6EgykQ=0r~7Es64$S=PsPe#K4B~ zoa%`h&$;uuTMcd{>E~3)$#xznI^JSSsBjp4!$?S8t1wVvTC#8oVskm@SZ04aEqqIP z_$hFB{UN1RQ{N7WI@4G*(S41v$xRTNCSlPn`IeV%yCnV;`dxSNUBbvCL`XuK3yl1x z5^I&Q+t1rVeUzO(c;VC};?FF}EBqVJeU@;{iRPAlze1e*%R6o${aKs6?SS&@Gobgg z=IU3>S$)_j&{vr(Kv7j1Ji0n9WxFxe0FY1i8Zg;Mk0swK-C#hJQnAnO_Y%`JxxI8^ z@?txBo*R8Rmw{Z#S{h#Pvz&*}+}`aX9PU_#TorxS<9*roRK)`&7YniEXY+&%ywy4a zy%4`}VbjOfNzg1+z&tQDg15XA40luZD3w zbc#_J8$G-D8tS`nOdU=;LtHl?Gvn$Nq6=J?pEs*1)Ew=Neh?L0faA|o&}~^COt%ih zpJ@`Ex!5M$Cw-PvVCm4KWWyXQr_ZE_>`2pp47t$@F68{G6R4nEU7-*v&B7+#AXzpm z+p#i{k6ATATdc(gn?Bj%7aj?_etH&vp{6+x$?o0|^sdgC9Nz)yrMV@ff6Zap=o~)up~ME-MaM=PtvHVx1s$7X zh>q2+8XkVJZmn$=xqw=|aY@0h|8CiXf~qCTsqBz@5qr|7*-n2m>7l8U9`o_d9UlCa z9e=d`a?jVyx5W2FXD-z=)Y*j#6x%P$waE7D@Vy+yo}aCyZ5r>n)Xp=W8%%vqNgeys zz_cA#HJkC!nplLRAvOu=zCC82iO>Rj7{g8Az}V`T|tNcrHijEzxIi{`V=*m3tx zPY9Q+1sNB|+h(Ok8my3tVuspuhQFBpdbYiJ(oGhsHMb9wi_>C~4@#KyLs(KMHQJ9Z zFlyrXq)CA^RQ#x?Ks9y)V_bdCCi!oswQtv z?!j7|_$>>ZGTJxDu)f4*Fu5NaaOP5_s(i;Anq8eH-YtthuaBVL+A4JlPZT?=^me+Lbi7zK1-|P0=Ar4iidQiP@E<+`tPSNHb!4r_%S~ z2K9%N*i^yLN-})j?!(W=zgH+xX*mS6A$Odw#rM#Q&4qJj=;%#P*PXK=rGDG;qUuPH zp*<#Je7PLtM&a>!9Qs2;SS4XB{;FbeaA1|g2CZ+_BHsGwYjZBK`i-hsl0hBGi?)@Q zyqKbQoAIqtU(RhbdFkV;0%A^0k*0AB5_3D>=i z;<)#nw7#8;zL}xT=N>&XdkgZI)QD#`bYm*G;iE6-D%Cb@;`Saz&3l!$ymfwemy!7P zf{3}Hal$oG`I0p(PAhy=RhfIlNBH|9`@FPrt~c(J>4RZuBOAjQmB{SNS4Li`Ki{_- zRHI9*&SLc+R%fn?uC=MCsztqD%yusHjByo_P{r$g6=qexarFvr0_yo@aC7yero@Gk zupd52-yFh-UZ}2X$GEP=ksjK|ys}rMcJ+xX>O_h4Z#zwH>aa>&W)ANw+EKfjpQpWy zkz4-x7a`)DbHi7ei&VN*VJDz|aP0KxM)V{f+mh-tS}GT%uRyqm*_G6a)zxiob@G#c z&Q#s@irG#~$=xV3W*K+yvK+(USaOy&xgwdp!U$2`)$7v{jg;{*HgWgKCW#-A-yeI~ zn<&6{r^qmn39p{oNP_U?JKsZ&6RZJ5_Tw&uP45r1E?O*6$UYV{$~WiIs*gRVFwIJT z81{@&)~JgR+&4L=saeEO<+C%Y)KZAb8WV7dOpd@P#c~`Z&A0uuzKtF4%duWUe=YWI z^yzXsY1S082O!^7`NHv>p|L@o-p^oyfOh*qj|__Ib4aoZInj2`xl!VFU7vbBd&Mzv zDf-n+wZJ7MF+S}zV(yo<u;(FpA)VCW0hGp`3lWX^-K}w%`V`xd@ zLR8&@y1LB4y=s~GmpeX4%4Pcl8$Le4%6|Mi!rcre<5$$oDTBHxX@AUGl-m684u$}#n%3c$KrsDnEI-JkYy*lsz z%!sKQ-HLG-``%@5f8m!iY5Q(QJ%-&|D&A!-f6QMYJB)q<}A)7PeI?suZFn&e^{Ylhdlq$qW7rL!d@r4+`>@9g)f2vPS8>d@3L zW_Wp1|2~xP-hZ!&0$=gH?%BH52ZcB3CWJ$%`)V`6kTs);3JHv~V|!6*jFQ#doDDj4 zHDsw@U2Y5yZ}i*hwbj6mR6gU+U+>&8>;YfZ^}RHenF}RBR=)s?Dm+d-HxLxzU8X8} z`D84p;Y2}gG`tu4n5K@5l;nNM^mW==SHvqo*}AXI=(Q?X3eNhE<>$o5A3x_=HJA6m z)MC@H3av*$yJC$oLh%QN@>8y6e##UR>+>Pz2Tl!v1O6V2Xk(%Yc}zD82AcH=W%9X|txqeCY1Wc5WVv6JyVoG3(RD6MGN zHVze=ZQnZ+HRoxwW3wlu(xhwrIEKa$9iLk@FK6~P7M`Wb|@A063 z)C*IH42qwBh^y8hs>h5!kn8C=wl9(Xj=3=J*?-=~W@$~lAqd038lLUj5Yl!S_4Si^ z)0u;MNfEnp7TLXxyUyMZ_dYdQ7OPD4F{;L;xjWtCFC#mFW!RODzr5Tp5T9|yP>AbF z{AWydU1v=)_p(Y;x`xPVE{$NvPxVu)M_1ybPt*Ktnp>>#_Zctt+jWiBSQaZs-g9h> zQ%{pKWwC3dTrqP&+Y`1n(JZ+i#jN7;lo^=3*%+q23}TVTGlHs?N=NrW9^Q}gSV|B9 zFQ{m8FBD8LY4!UgrOHbMRx;{stQ;zj<_oS(&IR0TyU9bx6LyuPxgnVT9t`_=_G@Fx zPt?FT*>u}ONzl5R-Vw5D&bJ_Z!PPT!-%@^5t(P-Wn|Nen0RStw88-xl?S1r&wg8g6ud}#f-;S4e!y*TJRX9o*2E5M=6zuu<6n>YkH ztu>WQZv5+kA`Pz3DV(VARO{UdPHxTb(XH` z_0KD5D<2S7&6cFmSWBrvt_mpzs_>l|-xIZiCSF~35WO{6q4 zfeO;}rQRpm+g=O9ZE9}}+k*unc9ePjy4bPgK0Ub83^dXCOot7=R6YC={#&)SDtXLzdk@_i9&#G)YDs)`8=1AQw$3zM-x)O>A2m=p{g zbMD0&Z6F7`9P?BMw*$^##KVGirjF@$Q=)W!N_ilMT^r3aVHc#&^l*f4^gi51?<_G_ z=mD`-3%9nz=LqSZaCmhu5qYXZaL;xVxxB|>nyYr;pca3?b|~p&h27mVtgmx{j{?Y$E0Jlp&r^J#Cdx3_vuMh_z0Rbf=@1ELD059GQBetp--Tiz&8Tphzp*G3 z*9>-n{t=nGAp)H-ea*1b7YtlPT4{;ove|;Xj9Wct5u>@QBm);Y20WI$Yo4G?dCX_s z_WY*YFNfg|Ez)o*rNyuUMv*r_>;K`koustWI|o}Z)3{dISU@6(?Yol*)8twO@fm1{ zN85rqb-bM>wB4_NSGW8D4nS{cD|T{!Du}$l@N||8NIX%9`Scxg)|0i7?)h_I6x?3u zPG2FV5l~3N7fEn7?tEDa&eV^Or^F#$)xLZFF{F&%<6Udtx&wPaT4zcCHdJU4_V_h# znba@XH!HFycFz>gqe{IGTy}YSq}`Sil11(AEV@X6qw!S|nFmmM>77p_VB<$4u#eq_ z)zMS94CG9>+o~ctaVWh^bK?HA)4bybete?)lTp+^Z zC?CB$B{f(~tTa9kfv`V?WgWpCM95fy=N5(Jx?;vJyfx8O1x5MK?xATe#LT}0)G!Ke3VFY}w;a(<9 zVc_U2HqzT)DXW-3Yw(4-rmi>#dv=9dQtc)c&!5YGZc_H(*&!pOg1EzT0h%(y)stbr zX2G!+`rKcr?kKpj!q<4HMe}}sahu-DeUq1>cIX#G&$?WF;`0qUqx2qPDoCU;up@~k zoV5x=7+~V_hc?Wm5)LCGcTbSr=Bi(QTUdjg^fTrA1NnGp_^NCKiyXPNme4(`*zgKqv0NHBx;tvkC!KUj3{p!H@k0<@-S) zOp^2p2csq%Q4aG*(ewj7Ch)ZhXGkcX6TxxvD>`%8aO(_YY9ax$zu3o>T z8kC&1X;vPfQb0?~@HuIdPOT_)p7o1Vld)G$PH)lB=>+E{V@p!hO$kD2hS!2_-y znWLQ_6F_1ClD#aMe9a7Skac}K5(U_MLhWH4?h$wdMss!_Q#*J368oCZR+ zE$giH_>=$RiwjZ*AJ%Qw|N1ogh}cL)`BKNB)=7~Mqy};#S1WW?Dz5(MQV^n~8U_UD zZ1v&^0~tR9)OQLj{SI8*qyU z1GeiP*^{%XM4e524-EVA`8V3>+&$?W-X?>tIi!anmkTVfQ%n?&+$2d#TKV2F+oUZR zjS^84%(QHUQ=I81rr_%P%v90`C)%5J>a)*!bEJhjoHl;@-` z?j4G3@%o@p-05h&vO20>0gb`98SxAAR3aM+L3C-`aQJ3sS+-G|dNbE=6Z2InWH{MRY`=AmC24e(&TeJpgGLuxU)R;&7KeAHDqEbxEeZS-8P%G0QSYUQy&36EIcnVJ z@KK#BUodGYRx0B{b?!&|%GAbebUZQ}U+@Ss+RR>I%^LT4ZqNt^67YpM^-&kQu;_zW zel+0vS&D&DE~n7|CP@@M=iu)jwo_hB)?@}w!YSC1Z$B{HP|B(j$yVwLIy){S2*= zHOS5_+ZRMWO3*GO^0);Dg=mJ)SvJLQ8CS*hFIyz=ipYs(Qm+exC_j_AV3ALMjJ3d0 z@ih;(N1xh-XR2LiUaN@>w^Q5eDOl>c9>tyRxT|(KhuZZ$hCXqTrix9UoYf#}hC#MZ zB&z5s1$1nAjWXSVP@jsA&?9Qt1^_222#3dfQ3%AVe2 z?3JL-HDb;VVq-7HgMOZUQkX(6J6_5cZ_rZh&Jd}f=6;a}=D~ew?%7T89UYQ0ZL)Q& zwLO>nVth~-RxxkB)lQpp#ingHFZY?>*~(qqWpE#3?jev+**wZpWs667@EG>EqzkmV z2dn(vA5-J|)!|Lg3V=aEt{byD!`o0aTY&4Uajui!EXmuAccgywTw{oqWyzW-Ro&@4p-{;u1na z==bf1#A8n=vC^g)IPsVet0QCy=9{%6&jK`u>fL7frmdaAor%|4YO`@O7lR72Lapt&dgG3@M4+JeuMQoHGJ z|JEml{CneOgty_&*G0t9ZOCoP$QlvXQ5F%C8phfa{ZvpSsB>IRrX2{!T&l+|y^g|0 zxL#5r{n>L#$STa}H_72AOHY{z5E~sZ?kvUgNnp%VAO@_ZWROI}ULZztCqZ>q^|3;u zCsTc5=?KB4sB(&eM1k&LznP07zs93mQVk$M4OK%B*{|ohBO)kH}zcD6k9erNBsuCexVE zVbdUWB=>N#Mx1%wJ#_F0%r;1*-MnO53&!g%1hxd?xIo4R*;ENr6WM*AK~6TeR|jR@ zyr(lHOK=wS4OE;K;Qn1sDGk6r!^U_8S%Y%u zc&N#X?8)jhWBH)5d%Dctd(-Kmmylg)bp5MCpxwHY({Yoh&Nfb_rft6UXu<&^*DpZb zk%WzGbq?;q2P0`F;vJuN7qTCEzwdQw)^#eCG@u7^<}XS@KYw!z9v`hIC|w>4w$$ImxG-DYy*Q3j_|Wk=yhN*SGw^Kqiww;q`xQlje?SF zjM~X`7dDvodi)@ty*^c@L|lAb$uguz&0~mt^PI>8WAMfds=(FdBLC&U0lwSi5Hds) z77gqPP%o$4tzb%kOJ)@NSmz(0cF!TLHWxUnWvFMo6c{K?DBA+n{~V8dE~aluH|#Zg zHA)iTtjnF`nncr?OuaXi1HBG@>*i@Q9E?8QhJh^e{iaW4tY>6GdZf`iHqLWj&+n>T zmzbOqrKWO)y55=Z`_tz}g zf(}4NC5h)Qi=V5r%+!%tieue@ivByJ+ZmmQFJOj7?<65g7y=MX7w}Ikp6Opf2&OF* zb)=nRPPrBEcVH}n9z3Q#qu*)LIW3PRz%n0Pz_#7b2rX~8zYyTK;q^k`O*xM&XOwQr zVaL^y{3-=}cA_sz?I)>tHEAf7OCX>21BJ{T!t?sQEL8y=6+2{!Rg0N^bux}wjd-JH z6Gsx{vW1ui+47r@;eQ9-`F`=rEjGD=#9Ogx4W(g|dBALSbUY>^Z?ggG`~6#V^HTJ= z4e`ysms5h3$m4H9I>F2)Bl~L4zRAYXWx`85%uiMD)ym(r1rtW!7Lia$KNXJ}GF=Te zV~g!o+9JhaKD({3m{3Qbw0D)c$vJ6EM9s(ZJkEwIPSJjl`$*0Y|1@=RC7 z|Gc5*rFXXi|H+7pXF~oK)aV?s;asd=&M8+?ldh)+U#ztWh`c}Mo&p+$MC^-F z%Bc2}O_nnm*O6AWCOMt3>CLq7J&3?e;+2PO&f2q{)u*_%H<%{=I~CgIUNRV9B{PH0 z?$#K7@%b%aPyDDVU@f}m=>$b36JyAL)WT$b027PNW2`?iYT#Gjk3>s zKVPbi5@@Dh5B&-0@RQh}QzQp_--LU<`Y??$t$6)$1`?I5pKBz&h3ewa=rC54iz=r91 zSH&uhUE?C=DYeN4&{iZJjDDB=j3GGyv=^N;WC*bx@Gst-ljn?^88?j`;zV4C&v40>Ey8{Z1 zeYatAQ_dHl3IF9C9StB(0XrV#o35a`qamQgi^+Q<0Q&Gho{*GYh6UIM4GQdp(IZ4$ zSQqcv8GBt(8pKq*}0g<_jR0V?sAY0o44d@~& zluX7E4z>w6$7%qW>?|~{{xn#0R~Q2#Oe`6_#1x3f8@&J-sD;2_iOil2CI+d;v6IA8 z-j*hX)u%db_PdQ&tOUT}2S(oi`C~r@0e?m=_P{6+!<^qQAlFVRlbRcXl}ZL^=qX#H z9#bPt=!nKCL}0n-0^W@IK*HHMLbh25-YfZ(zxdI9_x^4Ud=`hiYM#3W9NtqX9*^$xEg77I0kS^{XEA>L|98TG(Ld%+Dxc7Q|^5%`FQz|Qx+W}z>#N>enI6Ns_WJAiZBpM3X{Eq~e zt8C>SslX$xaqW>EtOkj84Wz~as#QtR3I)?U@rU2pQg$kCmza!0eryaD0Z-)*H-a=r zW+eAo*yFnwq&ANsJnjlpldmD9a-EJ@`Qj)}d+iI1^qBW@=M8NZtqX_>Elzsr^DG;P~Z>E``arlIc3BlS&u-$h0jE_cxdOHnh_um_tPv#UG`4z}xWC^f`gt z4{+kMa^ET9@&(Md59VMj)&@a1^Oew1Ii!4Q6SD=+L|XJk6M~hW=Jmy3Bk^FOc|$*m zAHqjYkhux)1tna4GAB6LL^Q`at03nZfiQ&O`Pfv`?#Aux~$F8FVb!3m>^ zN4%lYPZ~A`e&gy7w>syKSH8nss{>gWb%ts-K37FPXGNY@jzNx4M9wKn z1t~>%qRL8-pHu;l^E2TbH1P&{4ZNsI9W6p<5Hoa~b|``yX0TW;KhN~XsbW_KZ))d>jjf1N zRQA08weZsMBJ#Fq4zV1RtWDONj)Ajdi?Lg6Vpe(pM*T1&xObl4{xrenY>a%zFEd!7 zg>QmusCS^Lf3?uPgjCsn;EUh#O9tw~7)kl)`sdjs8{|q)Q`Eu|cq*#9PJ;H7-@wk;MaZu^S_w-3sZ84?eFH6JMec8nzb$ty2vueOC)w}TNVb0I-V+vqWC(^0z~*sVM$#?j7}40UV_c1cU8vDJr?q~ z^e3>43k&=~7AaYgcxfr+(W44T3JHfc?ycx)?}PbdaM!p`c&bLx2wuMcv0q2?QLvgK zA)B-CKf^2{PE)o*k@U+DmQ)M$f%nNx$K1u(hX_Y)FjwduX2h!~S|PFLJj}5GF_slO zx?2UrHPW&|6)Qk#fYSysEu<3dXsgVOe{^VUL7X~9rzU)`;4!=OpqMG?`)T>Rso3iW>HCA#=s{;oSVl?V2 zF_i~RQ)|4?|^hN)M$^CsSzW=%k738pl=rvUY zFLXJ6ooo_c8zKA|0)^fd4E=`s?Wev1Y<)j3JpJIdatE1CBucFV9Hl6g7aWpQ(9xH3 z85E~^K&ScB6t6eyKh-Kr)8FS#D+&azF67ln);dngZ9=#5^;dt=2zF1I{D<1UPd$`B zK7rXUuHJsj2AFWx!mZHpmm?VK(&X)J8P`koC3r&*{<>y}6&&~HKKbCDI71`c?&V>( zWS$lmO7Tw)dIWe+;ldvFvcs%mga{SoYzBycg^WS!)xx2AarR4?+L80-6pK1vs&=Lv zxCmBv&*;4~5*4L7x4wJCVH}rW`pG^el^7y+Kc(Ja=;MgWJ&67`7_lRbyhv~xiYQtU zBq9IS>xfj>rr5R03URZuSg4#jAO8E%v22ABT7pDcWNF}@qgHHrS?Hl#doq7!DiJa% z2Q~od@JkLBBX#7ih-XFL30}bqkS_B&mH2G(Sk&AR84!C~_$WB{Y&$YRCUgf%*&AK# z>FN8s)$fpxbN{d|TI|6$8YD0Dtt2z8h|Oip{nbHO2AqtDh!n>Yav#&19h6G^2=+)@ zZfC{Q|9mzg-VY-RIgiiCKI_#$>Rb!xo-G=K375~`umeWys@K+ZHF)Pl3`!k?Z5tg_6wOP@aYDRUiUYj?tUxQm>C)R3J{@>bm5GaUA0x(97S| z{wJp&bQ3f`?h;V#A*RYA0Tui`0E1X4Nf0utaw-rKZ4&NW7Oa66*?D8?n&btZx_)kgHEUY6Ce+1=wRR2?cc5O6wP+?rITwl941o-tLr3W|5?M z2E{ttQsM|eI9Z@-q~d0KUNyuArkFc^q+G_I%Trw<=Z?@Boa>ddWS);svHA5)ag80E zB^&_9X$S6d+Xv_PbR{fK`T=<-3i8lJ=kI_U?ldV02;^SBIc$3W>p?C)Tfh+s`LkCt z==<)Frgy~7EOJ4i2csj~4%39#{(!5h=y2_yC6II0!<>Xp>|2vXTr(`-8tCFxE`Te> zpoAtCG|r2kFGV@Uf%_E$)z&st;1~zEduzHUp>1@L3ynkffgz;?Fu@5>&h=M3R)IY& zy~q4jEz(8(9_ap87Zp{`?-`Isshknk*1xE_Dze**&;LY3L3b&Klrey%3Ui<;Mmc^i~yK2ufi9DN_8< zZFRd<=J??DY-?N7_ zTZj#@PHsXKipd(-E<098kWlFwWKIdp2k*(8I$HYTW%$7=938Ug13p6&fH~4T;lZ@s z1H`ZbV9ly@6cAe;nEhfzX@r4+R?7^hc~F(TcD-2vhkUDc>@|xpJoD??)1$^9ujzTf^@Pjkh`wtXcw=^;~XCnF0kbi~k20ITwlF(#wRbOWNo3Jg&4giEJ0nbRc zh?VqtOysX1d<`QePJQ?3Z|6G*H2EQ8a(rAf z#<1V)+zejetRXot{r|J|isdJUO{(=S10;IucspI%l!qmTGQFn(RK_CGl(t)vFm26c|wYGJmC;r zZf9f;qC7#6T+IkePd3`122Un+%X)~LN?U?^@B7Tb8&+Ty&!-iDiEBZM)8GZ*Y%Me= z9#F|gdi%k}CU#a2P(D0_f?;d7@!IUkZO?{1D#Rx-w?!_y6Z>1kLrv^u2^C-LiDTrP zDwFkO#vw@ITF77~fVg9i)bvAdY_XU}IoyNzeoQr{?)fu3sBYJd;fyzdVqo4k;MERT z66YpSD`xxTf0F_J)42-V5=WpV!avQt^!)&YE5kbvFfT-GpNEAHDtW{MJ=c?I+J(Q~ zM_s|(Pp|t}z?(hfO+VHf1*X@U7|z#PG<7@eoSFKr*OBp@NHTYqu2B+r5S12JVQabt zmX2Fpy5$=<5Ag4XLxOu4*r&t5oYCpE&j*2<$ZNoxcrK*)5kQLc4?N&8yW~$nv}d7e zZcug#z_@P~u18io#VJaw<^bq71Kc$QEP|vaPPk7-h=0@q^7Uq*na~z^EDG6Qy3i1) zk}*8`?)PiyKR{&&yQ>rKy;|LX- zi*y@NowYY+dTKkhZ5-5svvyp5G6X#>0wdGwe~P9jzBKGrm~C7V(BFX+JkCFDysSAiFYqLzW2G?Q$G(wfG#eV<5-s7}KAFMJ7p-U2u=jU@SHJ zR^TWb%kal=kBLDpY6)_PrmaJ$9*yvWsI-6A;5kbqj53BhtHXNQFeVx`pqPqG>I0ai zaN+RRWPSb8cy&dpD{z^!muS%s0wCTR68c=0okvN7TKZ$O)*IN8V|*5auQmC1^e!1LBXE`TJnEc z{r~;TzcO(|g@3PtSc3>8neIRRAM&FQ+JXgjrIKX5o(MvjfPAd~sw|P8{|7$P3nZ!6 zW4R!&89c8okg3!RBZ;o606M`}Id$q_A1No!z+lq6qpwiBEJ5b&KI~)7uvxc&L>ieO z(G2wmpNY@zZG^V^Xtf19)Jvo}eIfENejgy2dE5{3mmJEf39;F4*UPGd=GFTgwVa{Lwcf#8MmdhM&G`V;|YO2YR=!UF5=*x63 z@SC7`&(nYU1?@`U#zEC9gTm6)xdqzy#X&sd=VT@$ZT4n0vK!Jhg7Z3fb&Q4-2D&+A^@#1 zFHEga)o`(~^lsEt>}7DR6b?b6A!6>H+jji?ivw{tRdqva*3LEy!I=mL_E%RdY>9!+ zGx8=0YaiTO_YSP{+Q-2NU`HZHA6098OP4nAh8iUY+WF3ufP4iWNU=67!K;Z_5unr$FQnT<`s)DYw7AlY)X#j@|ApCMsh}!sZ zTm=j-tTOTb1P8Q4vhXa3(6+&2!;&csXi$k#I7Rn-Dg;3!pQNXQt?e;iXz%DL8d2j4b|D`z5< zI)Nbix-wjrH|C3&`Nh5Wopz)AVU!97nvB%})|O;{#gkUx|I#~!5Dd>j1u%pDOJm#G z88W|W9*E<0+5k-Zr1<&&g8G7{<?w)ulOd|uj3*raPBh=mxnypKUJhK^%4JfFtknE8U2FTz~Wd!KZ>F1+4 zg9hZ?g;v%X@nK0aXwdM{%O(nOpY{aHW zTu3qvdY;9+AiKgvqOz@&5UUoA&lU;CHb-6@30Y$QQ+VVNkaG)-&M~(Bk7Na&920u0 z5ppoNS?R|HX)pw%cvy;PcD7+U==Ad9R{ubO1F+qKSes8;p?*QmLbn17JyveODRUmI zgm0@hV5({Un}44a$eaJZ?vK(%#%=>GUjfcT7)Eb~fk>}>;4pT6tgNEkBpB6xq* zIIs(|#j-W}IIsp-F|Gf63lCI-9^}eUH|_Q+l83|j1AT9(t}i^47C60r znl`?KGiP=4?X<}k7-wT}A-BIjEf#xC<@(F-bx>fjNm=JEO3|`{ ziS)PsnMcrgC;q6EZG{V&dn^D)MIz~H7BpQJF!jadpHWviLFetDTey-J_vhN;b;>PA zNWr=7dW>+!h{CrEKOoqd3l22J>yrNuGK|$h8EA$YSQrJ4yi>66OuzIyq9L?r=-vW2?X6I|J1*LdAE( z+yB}qAB;U2b%3S|Y6mI!SCDa>Ge9Hocu8l`B^z!+hE5L|!o<5us*FU!J;--livnIs zI{7~E7b+uB0LtVGS1uwHoM+dgGf92mL|R30{#Wr0TV#_TQrG?b_m7?YM_Bw%(`M*DD#3d4Tgf|R_vZI5}1v6HqCrjaP{uarvSU9yh9UF)0 zLVzfZyL(<&bh02%CKUlH3;aPToL|r%hstyXC5$ciWiZ&-M*Q~%@|~L_>dJz^g-ytj zW`erDA)Pg*-u~Z_AejVj0NgLSeO^ zs25{5!36^B4?vN!lBh`p(+ko~V)e*QHXT!lK6 zi5$hXdq$#*h)FfsG!tyhkSqo9_&R!}<{VK+ zW8yT}7spWb|4fT*pV8o}QtDDhrhm`CiK_!G9=r^Q$fHX+l+gOWUL_ce9xTXCxUM$} z0(epCF%St9y`kWcFbDJt>49wdVs*L@et;OMO=h6>qDCNiKlQJWjrQXcFuh@aH%w+E z)Jhl1U{PwuiESw1I3uX}i;pSZBc{IEC_uJ(Fx&Lui1*ZMy-F35cbLD3){l}ZK7HB9 ze&m2ukTw%FRjJam0LI+bI|)eFZM|bFB(VuU_;=lZKXa~MFP>$yLHDJYW76~qAgzu+ zaXf!#KbMCDFXEBQWGw`1Rw90DTRxE;1qMh9gBs}SOla*V=7M)AKg394j-a||_(~-rO z1KLeIg1abt2XUf^a$3bq>(t))vp-AT2Yt`}obbTPvbtT9mj|RR&AA$f_ai%mDx;@T z7on%AwY<8s{SPoW+}v3kH6UYOv)D61WNyl!H0wv5pqO0Bj>4(hshJ>MO_9;Z8{IJf z$q?vD8T*(C1jgqG20Z*oNznr@e%5{5E!hw>n`!*et?k1KQAUOyc?(p=2!)g#1MJdo zdaBne?Z2N}JUJ*ttkfu2N)bMKG4?C69cd;p0)YYPK%yuPQ(S`&l&9#IRxI=_edc|^CG884BH+QL|^Ye7D5Bs{J~u${yiVuFlqqd zr~#QwCY&^TZ!&(BHhdU?Lwv=MRJ|86o*(i}>wR-F#z8c24&(uD-i?=JTR64TxB%2< zU38%Nga*f^kh23vy#q`$p@5*N^T0zUZA0t!9Q#=%aHOtHs7{rHGx`i3u=Laj&lGbS z(dTbwKK)0RE{3s@)Uu z>;U&WB@5lq|7q_nqpIxI_fZuAY3T+fl$MlG1nCwO1(6PEBt&Epigbqvf=C*mNEv{1 z3rI?XbP7n9)c>Bi-@W&5oN>lEXN(hH&iY{R`mT7^Gv~bT>%PJen%YvRt3814zY4~0 z%=bX@We8{T%lW(sKsA5@=uEv6EL@mH;7{ybOnR zQSPV@a>a=t?JYnw5v3O$kiO=ML!wVol+)kB|7ZdZtpCGG;7s8-=HER~!t%AD>Z>Nt z2jHm45`W})PkVhd>?9^SQSz!&?(LtpF`ayb@{%@PncHdMU;rZ1zxxSbm4dBBppw*q zfOXe0r|$IO^YrHI(ZElAQ&^E+4be=i(z4rSn?+gg%ey0cylw?|4Hl+@fX zRCqb-{Vfg93}xc9@U7tCcG$OL%QBD+kP~QHK5IM9F20vKA7SMA=%JL;dFJA>6?YP* z%b9VTB$9q0*kY@=HgrfcXahHz^Qm8O2A6SNgah>aw2d7Ade@{`g?nB4P1MCLQ zw~@?Jd=X5NTyI5&$UdQyxHSbdlz+#{q5wh#aWn!OkwpbCrGLOU6$2nQPfNmo`|pP! z2m06l{rG?ICz)!73?hrd&~tV`V=FS#2D1=3G{a1if`vc@fQF8X%;Z0P1(ah5ktLoH zF9DZDRQhy&h4K`6SvJqXf(k}7oH=8(BNp6k;d)zge~twhz^rfo`8*sFuqvq#E%gVi zYHGkspZyqs&O{z619(ZmLa1R7bIZj6n3<$nG1#i;w?DQ2#samPR6qDRB{IhRfjP@K z6zWIo0@viw3f|fusyFJK|I{%rvp7;L`=y??AB%`@}3}`S0-;`u9)o2D}P@ znl1=7maM-$mI{$~z^hg${=T8gksL8<{!`(lYe8_6$g`_UCfQR!BQpoGl*S>$j+ctnGi@2`a~xXdol& zmWJXFNX+}!!*mA7(BtqaZTKGl3+`087=11bJW~H&TLf$!x$mg|12A6UHD&Jyh8^`&z(|NMue-?ea{3j6t^v!{d$%zFL2;^YH^_J62cAT}-%LGpBaVUMY zUB(Qcb{hZnK9>pt-cF;Ck4HqgFku=8anMZ{ha(ccq{o*;au%4ij3XOT&{gFZLH$yP z_(=flbEA1vTo4Mq3j!)wFuYcy{b~hxPuUH8QU+~*L% z;a1wYh6ch8L%M(xUoVBJxJ_^$CzNKF;HWkVN@w{+#;PMFJRJRHFjpHa&M_8%FuR>ScI44l(BM2~mwN4kMQFxYiNtom76s2-F*{no98v1&K>Nr;yFRy+hGMg#?8kY+(o za1`+9)~O%Yk;CCIeDH~u6sZwZN?~S#_hGDM2&A?$s5Qm7!($eY4R{gXs4i7 z_blF8t|9S&3ELJB8qOS6fD`Sc6A40jLnt`O4u^oM);Lk6Z=|9Cqu9yvkB}+2bD);u z{U>WpoRbt98B|iG4J%7Sb%l^#%NwJK3YXkh&C}=aU&;&XLCi=Aa3KG0sN=v4Hm-MH z>egUX9hZ3p$F(YF#vgxDrrLzWOhu>NEC-IEfqEyNI`Bp7~W#Ehb`)Q7zcgVKL6K9C?8hq7m(ZcvR*T|Acvds%?_DFl-u|EK+Cvcr=zM%RoQPm3| z=%=o1sFRp#1nw`Js=+PH6RqF;*O`AkV#f}szCPo$A!7s=S~=!1CJP~NlQAk@PlO<4dsM$Ctcd(WA!<#+9wY8IR55ORDTXXL z$KGF2m*(-HdC>t(?G60n3nodPXc}}6GK`%sWhj=~(q3{*PC&J3M<~-)RH9^93nX9$ zdxFLJCVx3NNp+WqO#tu9nXlP6Nxh1x9^IU$s9-5>f<3|F2ZNGZ5$Tk~P0?rqgy`1I zTB(K}iy|`^T5D0@3uTtK9P8>mn%CGVkSxsfjVx+~sAayWn!PXKs1H+01*RK7XI_n9 zmdc4!n}t@&5W$YQFX()n6RiOQX`O1_EP^H(!Sq@^i~Xz{X@)}^3`a|3Pr+t`zm=iz zjMi7*-U4}c?*XCvuO_%r<*k(DY`e0u&K8O?Y5SHoY?TWk0sYx`Jb2- z2sLcO3}bJt*X`T?C@DJeA~A$euPh@z7Vm>Ait{Rte60h6r0X{txnp(%ZggG->xFdi z>2Ur72+6ab{&KCVItYFXkbz5e#_2s2sUvqS8 zJU(st!2FQ+XSfPA0nkmU(D;l{4eG&{M2HD&4P<`jZAXo0h=Q*XU5SAu3tC*~%q_?3zslIp6ACbW0CGw&a*8ZIMiBxa=&R{Y zW?@q$yiDN%gqCJZJ~5fz;H5`oXzxFB=fxLs>HKHUs&~c!vKWNe@x@*KY1>ZU;T5LR z;kgQAEa9g7rFC?Z+oY#m|dc(z9QE$LalN3T7 zO;_MB475Lnj@?+&eVEhO{6ZGd-0$XZsCWzmx#^_j`3|KGfG}^mP<)3|* zU@Q}3YtBa%-b_n(AoGWiC5I1l}`46XhC_M*$|#)IsX z7@8x}_uFTF0;9Ji!SkL9O&Ou)51SD*SY8KE)*zoNuXPN6O z+%rfb`|ffBE_TMR{R5CwvKLQzWIK%@zORauenuXf#jEow9{hX!zm+ASq{s}o-I{s7 zZS+-VA%!Zp>~&kjTd5H!QK0>n*|(h#6x51?F5|RH>$qW=clU=oI>Ya&B*PY&UIWW1 z$STNoeHpR482%(&H@beI$CrqfN&T8ex^7Kbx;on65o;oGZ3#&Y;Y7aaa#Ud*6w}SU znTq}tbX_Go!WVc2A;7$To5fT=Szq5KCwKoMs5d%AbeuN>=pDeJ&qU0|D|Z0uD9wX= z)$OX@C?9qcgYbcFBM!-C$i_{Sz>-aMImM?!;7DS60|*nfh{F#|wl1T++n=nW7p#x6 zt|>U7r5-Zv9;8tI20oeK`f>lLsrq~U2S71f3oR_Tyj#Q@OwXIr)o8Jr>;bd0X$P9% z5STkQS$}|m58Cbk!{xaI3DAaY$?T37mCM2stxO zxv#;xzS)`=T8Z){q5&{9T7ik=J8_jII#GvgaO-b2O#gJuzXK^7pwKYMLUz7OoZa9z zxS?~@`1P4R@%zS`XvPX@8|koH+59G4AF!-(MK=V5i{1D-^EEO*#Uu;D7kakGGCx}d zzUI5}CJb3GFpyoOYJa~2Mm}V~k){NqdaOSdp_?R#6ju7oB7;GX^`2c;cC{3uo}E1| z!`_0z2O5*tzbA=*EZtf|4)t(Wb0X~v9A$hS_cCxFz;-4NhMwJ>);ID^>*Szd0hG7g z7XiTV0Ti%{+%^D-tAn*!2Pf=1CcI3&&+ll!bh;1HyoaceKnR`1m@cA`I{@g^L2q0K zl_+A_n4cvjVrb7efcpCxo;?pW0l5Stht-u{m8(eH-b3=}aQ zK;77)95P}UWK5wiv(A^yGsuhdj{$H)E8kSU2ClJF+zLWfWX^LXa@6%GntHd%CpZFPkG{j$ zH^Te${1VBoP5jhQ zp<&Gxzoi}IDYg3(Ksb4@{qD{Po@#umn1jW8cwRk`tU;V~v68U%P zq+*C9)C@Z)H-)47h$>6iGR_hLNbf^9ti!ng>Uw9`zXpm1gPA|HghD_cnS{+yV^QRI zUnKNx6Z(b?FbA9la*{~ZNL~)aYTo=X6$Z{${E$UC1i<|;+`5aNJi12=&&A_^&_u3m z82YR-=o{SOGp(XPjwLv;mXBK4^`njlYE9Ko8jNsaZh0eVrdt1WTgno|0^TcK3D6r@ zLuy6+dP=};)<9G^hiGTRfTWZaorrXUox=kF^(`1pPw&Gn;`|Z{flcIEHX~t5`S8LQ zJo0m5@gi8el~4+I3AH|i{EG4TM!2W9s&JVf!kS(|4EeZI4?(mu-n#}!^g{^lU5C2x z0ulq-;I&xn}!OqVrm|RXtpxMCI)xD9Ame-bGVMIsf>V0l1fiC^! zyc70tXuuLp$^?u75K-pbRv}dH?u9=eyy%6mW(gKy%x=OLSZR3;KE$Q>?BU|JTIf49 zFuFYiJYWrVDqFiyaXf_WZx!13DIx~3cEs_~pFzFgGoIU>#SuUj{bpgeO}sAHCmILH z(P8<6MZJ5_8_q+^rzl?Tpy?r# zEZT)qY@u6TWw1`BcFd_O9@KZstaRG}x>l3@ldWAST^`08JV@AMS~hgz#SvqMielan zWKNTANFjD^AmVU$>f1vEq<_|V=v*O~+hJhB#Qh&Ok>$P7f$B|Yn$i4S@VXl4 z{1{0skyTr9HgRlCi!(Q?7oaoTU80MP1+`qmHZK5;Z zQ-D4a%(1iu0o@oX*5_eK^b;WC0t&@V)f9+T-98z7ku7H!GtavOxrjqNr3>&M$j zHJqQIt8<(khdyRpR`3r+zLVW9ZJpvkQT z&=m~qLb^vOwErGRrsv{rN5Yo9;ZeQiQ9~pO%ZRnpPvFZ;5Gm{ADv+(VAW8^4i11v> zh@!i=A(}qj0*z>Qn4-llJ~dTJ1VDAr7G#h5xcJMux1>l0N#;e#E0unyRbeqzfE;1c z4C%4w0hU$WbJ5ZYc=ZwS$X6oIyNzBprqYmwUMBQ%gUOwu9!jpf5tziNTS0y#~lX5{goKuh){zy1p_$mEeUvBs^&v6TiuzBXUAu7BBy>MLRfsa=` z1pIlz2GC@_rpWB|Dvq$yqjsWC!^*fF*g6+1Ta)mTLzj-vkp4iN3S(NGQFoW=r(lBR~Uq?9%q zj>LUHqms;0Uz|wf+HU7q1QeY9ywgO)Ef)7jCDX6+HboRg$YW{jCYD0kSYX2cg@Eqc zxKL{8V`p!e%#rcXbLAO&HWVFr0{}Sr1$zHaF1JrMxmqoUDtzM*HQxOR77loERT6Lw zxkdp96CBp$ zmy^4K&Y6VOe}Ig|mD_pO($?3jlj-FwUP)R?)U+)v>H8rga>j9iMiwCkRBVwdg)z}|zX@KalX zz12I%QivK-3?h&)yS|@UjpjX6q@?Pp?PhQ`Z2{yEm0fpy977mSeJ!_nXH3F#~hk_NtBTc@;TQdfKemQLgKfeGceEn_}g z2b92e?&0?#<|m6U9vG)|wD(nh4x_zWcO{jj`dZ`4ktdYQgCyBSEKVnfzoD(*emJNlQ=x6a|_(wmV(&n|LF7HD_+$3|B%sywvM2cTZ zlq*)hnO$epLk>IE-TM?ak#hS)pPFYNK9A)sOsaF;=OkJ&2Th-2VsF#OzrkO|AT)A$ zSly~}Opi}DRpx|d{%Icf9FM2AkS2L|RPu#pT98*m_bDY@u3XK8*J&nEw&FtHC+}z4 zvNJGe@^@(ZykV}n;)~Cov=yV5P%Y+xQg30I$RYoq`VDeA*NlTxy`wNjFI5{h{zS>v z!jBfJW938Zti;Q{xM<<)Mw99x*fVs!A=(6JetTzgtJn-!sNtz1-Sv@2%Zh5 z%Vr3w&!ll8AhIy92sIAxc?2?;lNssOcWdPARJ9$_)q?HAaK1mN(TTmWESWe@OsO=S zF%)PY_?3X|QS=#wggwa03B5NOZtMKAv|h=DBR64rNHe2Io(={&S|on^7i0Vc%5B4z z??z&9lwO(AdY%hQ5A;HO1bWgG=Drn+;EcYYi3|OtAmMD=gbo;vie!EC#wvbQg*Ddb56rmO3qe<8H#hV zSmvG=dnJHv=IitT6oe4gdRLB|BW!WF?9QV2@gix$IKrx^iZrfo6@Pwle~uG+^0`+2 z>PbA=Ub8e<85zrg1V+6Dch5Aw-ElXdh;VYss*^AY^Lt;e8{io1QpzkOK;I%>e+^6m zws|=G&N!yKUcE~8p)vaT`it+t%6NtOa-P7 z6gyOU)$;u36Rd76m@#j?;Hh=HOI#9KZIQV}qg>r*w=ibIcrS3jEP!r+!C!-%%K+qn zfD{r6(FTsNd8W*MVE$c}P-9|th##k~(JFo6j=rWb^A=WPBeA@ZOy%Enq59FZei`;JJx%tSPjg6be)Vu1Zv)3`c`Bj^ExPnYL_Nj zTGm1*bAmv2@5M9WS>91S4O}iawP@{10YO8Q;107%B^D8~dVhS%m?_T6!<;)>V|pES zmqUh2HU;kj&@6E92S2>P*I=Gy=M8=Qcp{PVgDX-FNYeL2z#(rFDbe;Nzv^2B?h4s}F zG~rEDJSIaR*j@tpO)Q=oIi2=|lfId;FlQ-A$&Ba~ddv-*%x1i2rCmOEaL%4R1qLtP zn}SN;5{NjW_2m*KfOh{h*JtsyN$iE)oet!9ncQs^7AEa}a_Xc$I2WaMQnM)2On5Yu zhu0HsUI|d>@2hCnJ;ust@<=!()ayhB+gatTu>oR9n-<8&mLS>X{Ux-}3#qB~-a(6z zOSmb6-VC{c1>8Mz=Ru#8NiUu1_ApOESeUJHMlqdwLz+gu=^iZ=QAlb>f*LZOyAF`bDxV9yQy`@BE~U5FNs;t;W^5*x;;TRm0Rn z&aa(W#=tcg9ZTSvALP-{_hdu2zADsFl8S4=FJDsoMqcG|%^Sxp8-?f8Ir$niFYw4H z`V5-;e0Ke8IK)!hm-P5J$5d!C@2ks2Ebq%BYC5&H~d zztNNoZIQm`LasNu%c-9znSWf?yf{uZm;>mE42J!7-$@Dav!LE^&B66?d9yRfFejPV zqC%6zDr2vTf>Q&8g+QXzSe(SIncBeLlhbfNb+vjDuIHWO&-X6l`ry+%71#gDSi~ok zmi`@#yGmF@hj^)IXfMGUM>JHn=uQZBefOhKOlEZjh80eHfzqqk;XWek?XJ9mciNsy z3fCCaWNf7eB-eO1;VG#l7KX^2ASyexawp`l#iHd7TYtv&xx9Jqc5YHhPpu22@3i=H zu+)`^OTgGRdFX3k2Hb<$Si#C`AJa<%{n&HXWJSK!V_XH9_`g226=&@GO(Ur&TQKmM z;e@i;!Ua0A@jn8k=RFnHc5XstZj8)AZ)*)>unkhXIq`CW*c>PmPJ106xId0e}on+1W4aEjh zfb<9?qJKHBx3#L^^-|LOE3NLWM-F?;&B@D0%!86>J-mtUP6T8xDG@_S#I^o!mSPbQ zUu&UQ4yh;mbs@OUU^xU4NQ)mAop|#N3573Z0GXNPw<+)+@4<+}hU*iOfn*))^(X+I z?{)RMTSM|6ilAZyww=@Y3r>Ufh5q5N6J=VS9Da0!oUDd$pA0FgcK0-X0iEZaT1w$M z#DpKmSF~yTmIR|tb!_9qRf53`XdG=cf&goK$N31DI zts7}b>!OYgCK9p`=?Q`fNbm&d7|%E#S$EPPblu(Bm4E?%I3oCW#=~@~)<_kA1VOF- zK0m<$}+8#tCca(W~kk}6M@-A)Z3h(_%7bf>fQp8W%=&il7J&FZ%-tL<^S)8K! zB#L^yKJx=8YDxSoG(U|bgmKkuydew@zgXhOAGdL zT^9#L%9*g2@Rmp|<^WJ`s$i#%^>UeNV#mr&xnOif3k(u=ZX0aej^;09Ox4#0flod} z8f;LDGRPv=|Gr|Ew@D4VbV7I4z7_C%c|B5a9dL@tL+C4`4@+Pa$k%bj9=gnIBunFI zP8$lxWWv(egqcGq3?<|>q#(?(K;!IlKuSZAe?|*-#CuR>y)PbxK@STc9E_VoCKwn` zPR&;V946y%t?L@%$7@WiRY0T|BFJ!v1Kw>J4mYZV*6UsHf608)+KxA&>3H16h-Tb$ z*pQiT86?S^PpvuY%A*LM>9#NU-OUyT{5uI6CAf`&#PH3Yf=`o1ddxtv$WKwv@;tdL z`-~l;e;f z*^s(q)cX8!-~31+is|e0t<-%AFlu~8WcB3clQVez`DThDZVIqE<$*l|>IJ|b2=B}p z3(<>6XK!jHEsa1c4U|_4#?45_W%|eXp#VC*t0}+-ca-Q6fIuUMK(_iKRb^_9! zeUWJVo>l2O571DOvIg7l_HSfMy*!T+G%pIBHR%S9LSAorAJep}2|!-plJH+?~dA1l8{WO3Gxz#hmFg33#RAj95vrfKT1G9RaIC6Jh7 zBHom|uXKC?7nC)zKhmBzkR&tz-UhN%2o-t#{hrFYO@NOeVY2-?MOxbV8Nln7xH$uF zJ-hf*;(Cqz5@1`b7g}$Gy!^iUlz;QbsvsT0eiHF%zy@TYg~l2zo2YGw=AIf+sEJ->@9rL!O;*Bg`)-6-F9#P`+oX z=Z6rd+a_sgI`SNub`O#I2Ksoq2Zup-rGntzNTJRy6n*|I+}l`6xHoV_10=C)3qD4F>)P-mj;OR1oWa( zY;{U;KmI-;X;70-fANn=xfI(b=sFZds1y5X$Tk?Na)O=I@1666jcq1xu+u@z16urd zS)Y!Nq@)aIr#&B(?DVpQM#F{+@|}FyA66}Sr#0I^tw0J>9JO}WbS$O3;e||56yu$! zI*@5M`|}ZajARuHMfEA{CkMQ`|B1ET-|V7i)ryb(gv6r+lL!^uyBQAfk#s?`+sZ|w zGH{pJ4Ou`oTP&Q2YX+2&?DG?Q@GKPr<_|=ynq&tCD06Q5KgtZ7)!h-!AKTNa)yts=oQw!3D;%RlZ2l0`eH@`8Dh z|Ar#n#r0`XkHkOzw`P>$(%Y>I@pH8@6rN1-_wmLrdR3BKx*EJBc`KZ*on^#s$ak&G zK)|Xx&y|MqY>deS>)4+hQ=;%#VNatmZ>;YlJb8idS<$>?Uh@HeV2r#wQ;mq3Rgb2R z?QQ*Cknoy8La2<@GJQ7bJG@?l2e+HAVL9aGlJ5`mHL!E`Cf!YMYx5ux(RSePD3`i4 z$7rW(&NkOlU+hgMH(+s}jra$hFwV7*R}0iM3PX0=5r*|a#(Tf1jl4FVC6c8B!LlGt z?r4PnFno;JWjrMY+wGFjdlO0VKdYI_iGc;R;fH87yzDi?*Aq1 z`@P7uBge2~9!jni3)Q)0!@Bt+M#9=PJ{8A~vF$3$UDd`@<$ZLp7tF;`bhsV(ai4rfi*SXRyJ+U=9OARw2u^cb0lV4sU!2jmAeLQT z9lNzEjmx>xE}TTU{~p>A!6xev34~g(F*NBsOxC02Z#$&TTEGz_6kxs`I{d-9oVUxD z^z&4&llk5q|9)+!`I?!>oli?`-XDyYNTx!S$on@JpLCjH=lGCRQVYy~+ZMW^o(&p3?OUh1r;Qp(8z+C>YV77#7ulx95o&jSLPX;qRXOY(UyKjt>R& z-(G7Mrn9;dyG$H{e~U+9J=FHTbh!^ou0R^Ts7x| z={lEl&e+=@od-{uMz1KO1sH9z>XQWPE@W&T1P_;WP^`*c%{3v~&6+RQF4Cdk+S?<% z9p3}fA#U%7g&?dBMVw@7dJMH2@c%SewZhCL(Ci*_1L!^?lu<)d5^Ry0xyvA`ZJHl} zGOkJC;VcnN%jp(AZ^hRvqMpg zR8@a{rFZ$J^`a>Q5!v@z&VYvz>yoDKE>>0)$tVm%gjzpu_o}~X;n#iOh;?aCY8rn` zNr~G`neSOg;$zk$zJq&<&L#%k{1&IAa{o=%v?6cW=Gc{?k>B?C;P8E^Zfgl8= zoiNtWx*@Y^qn;bCgiFl8fhuSb>N^h@ukvhZ2^-dNc!9CBQdcFT#dok-ZY&5l*!=7v zs5lrZF@#y@{oE#=A9H<15Y8McaFwuRBqR^#2j3C3uI$@4&QY@X@5&E+An5QQ@+bQg z8y4qvm5z(kJY?G4yZ6J6fj835aaEQtsoFTRhdrHVD(;{n!1G%Ivqs{D7LN zw+m6%bpNsbz+_!b)BHmAOoqi0uvKA0o<1PAF)G+%{xlT%T4~FJcVOvHSsd&^ZZ<2tj;}cC_6yCqgH6K|>K+UV*`9;) zYw-fIaKZxoaOlo?SXsr&XJqx}ZxMNvZI+htnzFOES^)DwIr`Ik%5>qT-xgt^l0y@n zOab*IgfDGKT*5;ovM+FMIfh@%Ki#tSeWupgosF=?Bun73gC6ZIF%j~5?+VqtrMGl* zlKGC24hYjueIxTRP#+%`+iplo=>_P`)5k(YDt6mof(+hdrZ4#;Ha(Q*X4tXIp8`&= z(<0x>(7FKp>DZdo1jh)WrKyPF>aTADjiMJaytf$Rfmw0B@4UcCkshD3`QR&Cb+YEx zE@{ga?cx>V3)H{eX(BwKB3d1`&HOe8J{{TG^6$Nb7m(a4J7^AE3l_W>>{XA5Y?}I0 z+!ei2X`gI+H4bu{m!_2jAMZvE^cr+j?}DyUrw(ca3-L!Q<8|Y2UO1?HIcv@N;9zg< znm7aORPU!m_%W|vSg3fWAoHK?*^J1^%Q-0>36-uMuXJ4)>6L$4I_am1IynC{DuiPb z1gGMd9;|?qiZwdI-ppXA|Hrlxk>uW-&>Z>Xy^j&0Httc~wvcML^46GH5ZNE8*TwCB zwTc3XKx|y?H2w>Y@O<*p(=Co&i6|;?xI(P+qz@H*2^M=4*S}lH@J5=CnNjfV0G%)m zJ~vIpE=k1M^`M?SX1P_yF>eJcWd{=XZ28@?K7Wsjw$ROu82sQ4Fc>jv*$G@4Vd#|T zb8LY*k*cOiT&x!b@1Mop>KEVbspd9oJdgtdh)6S3&iI6wa~Xi3KrsIk7C9cmB4Q$B>juj;0RaST#Z#N2TP{4YFr$?~0|h8{RpsR>#u zodjnMtrm85U{jE@Wwo!7P7Ud*{(2zL-y0I0J*eN)wwCTbGQAtRS>}^6;{e36bKB8D zXgtZ#a;kefu@{xL!6^+{FVi$YbaH3el+i$iz1ZbcA+1Ln8El z(#QDe-r|tAc$?JBNZ*Dg?#oIcPqlqsyI7BTMWoWxY#QE{(lROu|2}0aVG02gBb5;2 z!5QF*!-a&(C!FiJ27Gz?d)?+_uZ1356Q|hLOkmYHgPY6#10+Du6Q@j1j4S@>7B;UHRRP$-P{qfD6^- z!D;crL-d17UZyKwbqz}ZajFC9f0w=$nyeK;^Z@)(NhR&f+q2^ShM`r@2U)wmNS~!B zGutSOj@n)Xl5W!G1kgKU@v4lsQVNU)m$t|j*B17rNz&qzH1p~Xp15d(V@kKt{#623 z1IGiH)ti$TP!HMCZ`E}3ufG_ZtHG~C^MX}5b(Q3q+^h&$ zmQ`!VrpIOpNuTU1|33>wIsy^OVXsJAGQl}iPA-3(Yioy>0LS-i+J@_Mn5Rm(a@}^Y z^pzuP%AkvhL_iTFG?(#dsKoTf#o$d?8FIfSs|Wmr3eH!`=J*Tog*{cSQcXQbIExxv zp1{$sw7qQ{ORvm*RXj?rzRK`Sh^p)={G%M(f@64`lj&SCYcX0vCO}SjDaOq~~*uy=jwhSwg+ySzd7(+!Q_Ib90RJ) zDIT*p^+W}a!sa-(9$_;@jw7K`&wNiI(pn)ZeiAY62<`8=klhz7P7L^ZSItIxmF*~M z1xoMngZZGtlig+G5rU1|?I#$N`-5f$?#*}F(T50MTHOnFad@@$42kPPu-CpCsW!RdHIaSQKY4dT%)>83$>SI> z>+TFeCu+lYOxbb`LsWLHhMjUTz{}es{2Y6u?Z@Udlvw*;i~nloO!^r2Z-MaL(R9La z=MG?bOfBAvdR3dLVAaR;F6?QNet2aD<1|e}KX_?pOp}e=6Qu%s2GI2{q@~+WyOIC$ zsGe2Y{AzGcJ@M6>6C3X5*m7Gm@Rey-@GY!dw7G~Z#z9OCIF&p78%J06ixtZ2uacV` zndg-hH=*%d<^63c_6kJmV>W-SE2!O$liy^3vaGCNrVR8Wts6o#-s8w~)gxx6j*=+} zeE0wdx?#Ik3VxoQ#cG*BCC2Mb&$wscr_CPVVpUocYW(@jW7t~UQ3M3F&hpAQei~jf zh*7O`K^tTUXmS}KGJrH^-zjVbW!p;~P`zEqdCVK%aERDbKx3ap5(i%QdeA?VfqC~3 zxU0lKKAHofMHuQ|yHQA#eYYyA+i-*iB$kGUs~OkgR{A;s5-%uxR&zU%O$*BIgqN96+MZ7 zI%*{Abz2bk5C#B}=t273#C^|NsI{3Y?wQdK0e{I5tdq!i!F}=SJ&VZx7rPgFQTJ@)3eglDFx!Dp(ywwM8VBLCK#JH)2=|o58y9hkni_prted(qh;lgk+Cv} zfy(ykme8ef{K{cKzb5M_3#P6B11Zu2%0ySb(B4Df?@Zb*418|3?Js!kQViXHqu2SI zcJ5q8I{+3VM%)(GO!F!?FPccX@mdC6QJg5MyH}6IvUDf%+Tdb%In#saYG-odR>^;Ln>3*_8RP0t>ZsNQNzrK%`QO&)8@+%}%f84PDkBQ=0So7 z>XOzlWC~z2cTAt4Bu)rxgNUx_M2>~Nk8=p8wzvgyrlF0c)Q^1oVD#(I-K9s~C#Gh= z&gH$x$*x?Z5pbeolJA%7D!0RlVExG8yVM5U366Oi6r1KkTMMpH7X1Q4-o9I@X~|9A zrC?59z<`g$|5HjHv^vHqqRj~$s>N}#Kv1l|1h}$Z=Nb}c%sA&10S~VG;|XZ8Cn4(1 zg|TUjE+QI;x|iZVV_8VV?r7o8Vv>+wmgTiG_tMi2Mu4fdpYlJMYPML^JrG3YlPT?U zqSb;W92~#CdHuL{wvkw$;*Y;v?#>VZpjnDC^q{NKALa`S=tN2hhfpv$^^852KeZ@TKv)F@p{%0u$ z(3~VeOs1ObaH2G5F+_$MKO(1EJ5Ty`6HK7UJPn1jrA?Lql8t|M9Jny+Ob50R3+F+hW#3mN= zz{={F$rX}D5vc{0_dtxyRFomOR6!ypBd1oFen`l&lU6HT6y6>aqV#?nt_(wsYTGS>-4JTlFqssLa@-%zyQ zJRO|{ z=6t@=88`LUKz!a^0_^ARnBzbI4mTpUxG0<|lJdSr<4;<2x~;fW2blSzHR4hd_Yo`H zR$YnfYn+g*sCG{~{{Y;M$dU~LJMRe+y5hb8Xhb4jb`--9^}GbiL338;Ugs>*%nQyt zAWBVY#wB3GrRLF1P4>c}6D)}dNJlm%L8{d=TG!?8=Ljf^5SJ#=C*4>8zx-n4@gf1Rq;tHn02( zHnV+Wdbbq81J`^4;8~MuFD5&;1(Mk_r0U|y=z`60Kj*&%ygJr6pXxh2TNwz{`f@UC zdYrH=m@VDQK{%Jv`uS1|C5LfpYPk7%{8LJwCaRgsoT9T&=~x8|>n{m(I(`t(g%9;e(I*}RBPodfp1kSR5c zTn9bgl`y&a&u_@c3!_fz`^;I*Dt@pQl6xlOT?S*+)(t@#H5wsgoTq!H-&djEs`LR+ z2+w8K7JL9T?zw9nXDAwFVnw4Cq*LvDPV&Sw&2l0X>O4(W(DiIyvYC}VdBl#>^N#0E5r_$+O zz3rig#7zRz_s*pb6J|6T?{0H4XOObdRcE}N0&y_7;5(`Z%^R+N``qlXJshQ07bM1E zln}+!Fl!A2=8$Ve=}yAsmTcSt*CV{jb`U<-@pako`&Rf8cE&QYC9!D~YR@}narrXt zd~e&;(a(Wv#u4ulHAB4lXYjxNH<{p3px;~QO_nfKO z%W21=SJsQ|&<}RA!woyJ0pz(X%ZCLfkAMnH&KugELTVx1wbs}E?dFo=v|BcLy%k}^ ze&y#}vrD61#*a!t5I`>JfJ4_Yk1?~$Fn%eUSCPsmyOpU}hX3WG?d?mNaoxKtbTVeO z!oA`+Db%c{XU8HGPHc*k_SUD-T`*(jf4b3^Q>EZvz4E?!C})l3h`Lh#m%8#VdF&Y0 zGl9&mtNshf>&Rm_qZ9eqZUQfKkpcgj?CT`kmtkRy|J)U+EB93hdgxCppC4kvmi=)e z#-BzDxAEG#%mNUC?efC@u(J_GJKq0*>vK{k?+%?w+3ZWZNdADy^B9T&{;#nbDravm zP-Apn%)4P@B>tETf5AcS5gSTYMT`V-6$=H;T}IVHfuFuSEvq`ypPfS9xVCWiVLlVv z8jJrD5s`k)MtV3K;!AwqjP;o;nD>H#eT?nTam-%-9fE-E6;-j2ZRD)D8OV&%?$ZJAZ89EsZjP>!i{dR;SQ406MMc|h4&-v34_C7kRd88d{x#+_*aAd$Eer__qxwah?bb-g@_+&+vgp| zQUgGEGtS2k_7(rOJB|)6S&lyLM7{ZMX9$jrW{?peyZoRMT?)V_-v6<6M0TN2my;u@ z1-J?YA~Mh3E)e6v+KRmviiWAnq{zJbe=A=%j0O-){3qlnB4LU+;5(Knz+b2}QhXH) z*n9d@-$d1H%`};D5K0o0@`RZcU?AU0Zq5`EPvo4i-gdvphdU%NPL z+d;;7L?QukZsh#smZde|z)B(K{SZ(Kgo0^gW<|h*^;s>HiXd(P1>w#kF?hUyZAtq=Q%~9fp1B(8r*B5a&DTJo)C0WGgU)3W3U{(ZaVXKMqXYMW_b= zx2i*?o5H_h&#OH@10s?OGg=XI0B96JGdcCg(GxD4vY{-_u6`A=vo|d10R*0@LQHYY z)#-aK?4zDQ0v$&s$?gMTF#2I}08d-@h^pPXi%AgjN;pBU8h|D99mzrvdPl%58r*xmH{xiDO2tw;j z=~AJwymg4oQ?%%MAb}%+CmEvoY8y?Z->)ATL*t=rKnMteN33@$t{Xg&MQ$6QU5!jh zSU7~Ikp-X?_;N2~1skRTG8v<+&GxjTXb}lRZBG|a&%g!Co<>-9(2;Kmd=_xRT|GqV zumk3B_@>@vRWA3S=a-wqi>aicqNzE6uTo=QM2zr8*pvX73SfEi;Epm8C;`!4S*892 z@NP-}noP&?jV3!A56AOepO&0(d?WxD3nuR7kaw~=+))&`fVewG`?B}UhjgeNy~1~Y z9l0GJT`9S_6Bw;yzgwt20MAGFpyJ-h6+~k~OAbCN69rh3L0xa8Q31I~tA|#o=de)` zTe_FTWbzz_;4JB{ps0gS#leU~LzBzy=!oAk4XuZ#<2ZZ-!}wZ|JWU-LvuI(Wpotb? zuLXb}#y3c7^v3UK%BC{%-A~|Y1H2>!(O-9QKL?U=VsKph=|#GszQfJULB<3J!f4f;q@PSyN~1!G#w^s`CJ0^^ySCOl&Cq2~bF%E_)tI!I^%<6War0 z>!VMPJpGYXwH)ZsYN3f`C9eX!?rF^}=-t~)p|KX!3LD!5H$)Z>c2=Jwk!fr)qV7dJ z7jhbxTLD*~P`J3rLi&<O1@yZbfQ@h!k2~9Tx+{ z#(CSCElvRc8n`;iLs~=~%5GL17pG?sbCkA&Y0~!;DI-D#F-<}m{xer0g+pyP4u|!~ znrSDxv8H8N29bD@N9$d&lh7f`_UpW!1zK?Cv+eF{oeoiHMR=qivRrO==c9~-@m^)%}jIrwSRy1zptZ(89U!g zf6+uAB|;sYl>d3^IeD<*>C$#gC>66Am;+c9#SBuSIGRWQ4(vkv8cX^o9w?Md!Bhp! zd4FASF!7+V1l1Z!#rQu6QO!bfD9g45v zH}9at&yD!e>P`LtR1vma*;PZGOToP0e=ZdI(`#Nh-N}&&D9VxcVh^p-Au;;_J5U`Q zMRmZtr~~uQznjI;CUNu&hX@6at|nqJbiLQPhK`c`?v$Xvo7{h{!~F-#B5Wf6^W|Q8 o{l_uy?^|w4{eST%QGVM}30;+Hp_rD8WAIRx*O1GRF@EyD0CZzUfdBvi literal 0 HcmV?d00001 diff --git a/ManualSource/chunking.tex b/ManualSource/chunking.tex index 50bf321..7c4d7e8 100644 --- a/ManualSource/chunking.tex +++ b/ManualSource/chunking.tex @@ -20,11 +20,11 @@ \section{Chunking} % ---------------------------------------------------------------------------- % ---------------------------------------------------------------------------- -\section{Explanation-based Chunking} +\section{Explanation-based Behavior Summarization} -Explanation-based chunking improves on previous versions of chunking by learning rules that are qualitatively more general and expressive. In fact, any element of a learned rule can now be variablized, and learned rules now have the full expressive power of hand-written rules. +Explanation-based behavior summarization (EBBS) improves on generality and correctness issues in previous versions of chunking. EBBS learns general rules with the full expressive power of hand-written rules. -Figure \ref{chunk-comparison} shows an example of an explanation-based chunk and how it differs from a chunk learned from the original algorithm. It is interesting to note that in Soar 9.4, the arithmetic agent learns 1263 rules like the one on the left-side of the figure. In Soar 9.6, the same agent only learns 8 rules like the one on the right because they are so much more general. +Figure \ref{chunk-comparison} shows an example of an explanation-based chunk and how it differs from a chunk learned from the original algorithm. It is interesting to note that in Soar 9.4, the arithmetic agent learns 1263 rules like the one on the left-side of the figure. In Soar 9.6, the same agent only learns 8 rules like the one on the right because they are so much more general. \begin{figure} {\footnotesize @@ -114,49 +114,51 @@ \section{Explanation-based Chunking} \label{chunk-comparison} \end{figure} -To achieve this generality, chunking needs information about why rules matched in a substate and how those rules interacted. This allows it to determine what is generalizable and what limits there are on those generalizations. Unfortunately, the information necessary to determine this information was not readily available in prior versions of Soar which only recorded a trace of all WMEs that were tested in the substate. This trace, which we call the \textit{working memory trace} possesses limited explanatory information, which limited chunking to learning very specific rules in which only Soar identifiers were variablized and all other elements tested the exact values found in the working memory trace. +To achieve this generality, chunking needs information about how rules matched in a substate and how those rules interacted. This allows it to determine what is generalizable and what limits there are on those generalizations. Unfortunately, the information necessary to determine this information was not readily available in prior versions of Soar which only recorded a trace of all WMEs that were tested in the substate. This trace, which we call the \textit{working memory trace} possesses limited explanatory information, which limited chunking to learning very specific rules in which only Soar identifiers were variablized and all other elements tested the exact values found in the working memory trace. + +To remedy this limitation and produce more general chunks, EBBS creates and analyzes a behavior trace. Figure \ref{fig:chunking-wm-vs-exp} shows an example behavior trace. + -To remedy this limitation and produce more general chunks, EBC instead analyzes two traces simultaneously: the working memory trace and a corresponding trace of the hand-written rules that matched in the substate. This new network of rule matches is called the explanation trace: \vspace{18pt} \begin{figure}[!h] \centering \includegraphics[width=\textwidth]{chunking-wm-vs-exp-trace.png} - \insertcaption{A close-up of a trace showing differences between a working memory trace (left) and an explanation trace (right). The working memory trace only contains the literal values of the WMEs that matched. The explanation trace, on the other hand, contains variables and various constraints on the values those variables can hold.}[A comparison of a working memory trace and an explanation trace] + \insertcaption{A close-up of a trace showing differences between a working memory trace (left) and a behavior trace (right). The working memory trace only contains the literal values of the WMEs that matched. The behavior trace, on the other hand, contains variables and various constraints on the values those variables can hold.}[A comparison of a working memory trace and a behavior trace] \label{fig:chunking-wm-vs-exp} \end{figure} \vspace{6pt} \index{instantiation} -Note that this trace is generated dynamically as rules match. Whenever a rule matches during agent execution, Soar creates an internal record of the rule that fired, which is called a rule \textbf{instantiation}. (Each box in the explanation traces of this chapter represents an instantiation that was created during task execution within a particular substate.) The instantiation contains both instance information about what matched (the working memory elements) and explanatory information about why they matched (the rules and actions in the original rules that contains variables, constraint tests, RHS actions, etc.). +Note that this trace is generated dynamically as rules match. Whenever a rule matches during agent execution, Soar creates an internal record of the rule that fired, which is called a rule \textbf{instantiation}. (Each box in the behavior traces of this chapter represents an instantiation that was created during task execution within a particular substate.) The instantiation contains both instance information about what matched (the working memory elements) and explanatory information about why they matched (the rules and actions in the original rules that contains variables, constraint tests, RHS actions, etc.). Note that WMEs that were automatically created by the architecture have special instantiations that explain why they were created. For example, an architectural instantiation is created for each \soar{\carat item} attribute automatically created in operator tie impasse substates; the explanation causes the \soar{\carat item} augmentation to be dependent on the operator in the superstate that led to it, which means that chunks learned which tested that \soar{\carat item} augmentation will cause the chunk to also be dependent on the operator in the superstate. Similarly, architectural instantiations are created for structures recalled by semantic and episodic memory in the substate. \index{chunking!backtracing} -All of the instantiations that were created in a substate form the \textit{instantiation graph} of that substate. As chunking \textbf{backtraces} through the instantiation graph, it determines the subset of instantiations that contributed to a result. This set of instantiations and the connections between them composes the explanation trace for a learning episode. (So, the explanation trace is a subgraph of the instantiation graph.) +All of the instantiations that were created in a substate form the \textit{instantiation graph} of that substate. As chunking \textbf{backtraces} through the instantiation graph, it determines the subset of instantiations that contributed to a result. This set of instantiations and the connections between them is the basis for the behavior trace used for a learning episode. (At this point, the behavior trace is a subgraph of the instantiation graph.) \begin{figure}[!h] \centering \includegraphics[width=.99\textwidth]{chunking-trace.png} - \insertcaption{A visualization of the explanation trace of a chunk learned by the arithmetic agent. Each box represents a rule that fired in the substate. Arrows show dependencies between rules that create working memory elements and conditions that test those working memory elements.}[A visualization of an explanation trace] + \insertcaption{A visualization of the behavior trace of a chunk learned by the arithmetic agent. Each box represents a rule that fired in the substate. Arrows show dependencies between rules that create working memory elements and conditions that test those working memory elements.}[A visualization of a behavior trace] \label{fig:chunking-trace} \end{figure} -EBC uses the explanation trace to determine (1) how variables were used during a problem-solving episode and (2) what constraints on those variables had to be met in order for the substate rules to match. EBC then uses the results of this analysis to create more expressive and general rules, which can contain the full gamut of tests that hand-written rules can and can have any element variablized. - +%sjj -- not sure how good this paragraph is, but it's my understanding: +EBBS also creates an identity graph as it incrementally builds the instantiation graph. Production memory has knowledge about the actual matched value for each element in each condition of a rule in the instantiation graph. Some elements are variables, some are literal constants, some variables are the same as other variables, constraints are placed on some variables. Distributed Identity Graph Unification (DIGU) annotates each element with a pointer that maps it to either a new or existing identity in the identity graph. This annotation is used to determine whether elements in different instantiations refer to the same underlying object and supports identity-based variabilization. % ---------------------------------------------------------------------------- % ---------------------------------------------------------------------------- -\section{Overview of the EBC Algorithm} -\label{CHUNKING-ebc} -\index{chunking!explanation-based chunking} +\section{Overview of the EBBS Algorithm} +\label{CHUNKING-ebbs} +\index{chunking!explanation-based behavior summarization} \textbf{Basic concepts:} \begin{itemize} - \item Every condition and action in the explanation trace has three \textit{elements}: + \item Every condition and action in the instantiation graph has three \textit{elements}: \begin{itemize} \item For conditions, the three elements refer to the symbol in the positive equality test for the identifier, attribute and value of the condition. For example, the last condition of rule 2 in Figure \ref{fig:chunking-trace2} has \soar{} as the identifier element, number as the attribute element, and \soar{} as the value element. @@ -187,7 +189,7 @@ \subsection{Identity} \end{itemize} \end{itemize} -EBC traverses an explanation trace of the problem-solving that occurred in the substate to determine which variables in different rule instances refer to the same underlying object. There are two ways that an explanation trace can show a shared identity: +EBBS traverses a behavior trace of the problem-solving that occurred in the substate to determine which variables in different rule instances refer to the same underlying object. Identities can be shared in the following situations. \begin{enumerate} \item Variables that have the same name and are in the same rule firing will share an identity @@ -203,16 +205,16 @@ \subsection{Identity} \begin{figure}[!ht] \centering \includegraphics[width=\textwidth]{chunking-trace2.png} - \insertcaption{An explanation trace of two simple rules that matched in a substate} + \insertcaption{A behavior trace of two simple rules that matched in a substate} \label{fig:chunking-trace2} \end{figure} \vspace{6pt} -To get a better picture of what a shared identity is, consider the two simple rules and the explanation trace of how they matched in a substate as shown in Figure \ref{fig:chunking-trace2}. The connection between rule 2 and rule 1 will \textit{unify} the identities of \soar{}, \soar{} and \soar{} in rule 1 with the identities of \soar{}, \soar{} and \soar{} in rule 2. So, the \soar{} in rule 2 shares the same identity as the \soar{} in rule 1. Similarly, the \soar{} in rule 2 shares the same identity as \soar{} in rule 1. In contrast, the \soar{} in rule 2 does NOT share the same identity as the \soar{} in rule 1. +To get a better picture of what a shared identity is, consider the two simple rules and the behavior trace of how they matched in a substate as shown in Figure \ref{fig:chunking-trace2}. The connection between rule 2 and rule 1 will \textit{unify} the identities of \soar{}, \soar{} and \soar{} in rule 1 with the identities of \soar{}, \soar{} and \soar{} in rule 2. So, the \soar{} in rule 2 shares the same identity as the \soar{} in rule 1. Similarly, the \soar{} in rule 2 shares the same identity as \soar{} in rule 1. In contrast, the \soar{} in rule 2 does NOT share the same identity as the \soar{} in rule 1. It doesn't matter that the \soar{} in rule 1 uses the same variable name as the \soar{} in rule 2. It also doesn't matter that both conditions with \soar{} happen to match the same working memory element, \soar{(S1 \carat number 3)}. In terms of sharing an identity, the only thing that matters is how the rules interact, namely whether there's a connection between elements in the condition of one rule and elements in the actions of another rule. -All literal values, for example all of the attribute in Figure \ref{fig:chunking-trace2} (\soar{superstate}, \soar{number}, \soar{intermediate1}, etc.) are considered members of the \textit{NULL identity}. +%All literal values, for example all of the attribute in Figure \ref{fig:chunking-trace2} (\soar{superstate}, \soar{number}, \soar{intermediate1}, etc.) are considered members of the \textit{NULL identity}. \index{chunking!literalization} Variable identities can also be mapped to the NULL identity, which means that any elements in the final rule that share that identity will not be variablized. When this happens, we say that the identity has been \textit{literalized}. There are two ways that a rule interaction can effect an identity literalization: @@ -228,13 +230,13 @@ \subsection{Identity} Because the condition requires that the rule that created the matched WME to have a specific constant, literal value, the action's variable must have that same value. Otherwise, it would not have created something that matched the condition. \end{enumerate} -Identities are the basis of nearly every mechanism in explanation-based chunking. EBC's identity analysis algorithm, which is a fairly complicated process, determines all shared identities in an explanation trace. Figure \ref{fig:chunking-trace-identity} shows an explanation trace after identity analysis has been performed. Elements that share an identity in the figure are colored the same. +Identities are the basis of nearly every mechanism in EBBS. DIGU, which is a fairly complicated process, determines all identities in a behavior trace. Figure \ref{fig:chunking-trace-identity} shows a trace after identity analysis has been performed. Elements that share an identity in the figure are colored the same. \vspace{12pt} \begin{figure}[!h] \centering \includegraphics[width=\textwidth]{chunking-trace-identity.png} - \insertcaption{An explanation trace after identity analysis} + \insertcaption{A behavior trace incorporating identity analysis} \label{fig:chunking-trace-identity} \end{figure} @@ -242,59 +244,118 @@ \subsection{Identity} % ---------------------------------------------------------------------------- -\subsection{The Five Main Components of Explanation-Based Chunking} -\index{chunking!ebc-components} +\subsection{The Nine Components of Explanation-Based Behavior Summarization} +\index{chunking!ebbs-components} \vspace{12pt} \begin{figure} \centering - \includegraphics[width=.8\textwidth]{chunking-ebc-components.png} - \insertcaption{Note that the two rows on the bottom indicate when each component occurs during Soar's processing.}[The five main components of explanation-based chunking] - \label{fig:chunking-ebc-components} + \includegraphics[width=.8\textwidth]{chunking-ebbs-components.png}%fixme todo bugbug need attribution to Mazin's thesis. + \insertcaption{Note that the two rows on the bottom indicate when each component occurs during Soar's processing. Bold components are new or significantly modified in EBBS.}[The nine components of explanation-based behavior summarization] + \label{fig:chunking-ebbs-components} \end{figure} -\begin{enumerate} - \item \textbf{Identity analysis} \newline - This component determines which variables in an explanation trace share the same identity. It also determines which identities are ineligible for variablization because they were tested against literal values in some rules. - - Note that this component has two distinct mechanisms that occur at very different times. The first mechanism, identity propagation, occurs constantly while problem-solving in the substate. The second mechanism, identity graph manipulation, occurs during the learning episode. +\begin{enumerate} %Mostly from page 62 in Mazin's thesis. - \item \textbf{Relevant operator selection knowledge tracking]} \newline + \item \itextbf{Operator Selection Knowledge Analysis} \newline This component also occurs before the learning episode. Whenever an operator is selected, it analyzes what rule firings contributed necessary operator selection preferences and caches them in all rule instances that tests that operator. - - \item \textbf{Constraint tracking} \newline - This component keeps track of every value or relational constraint (e.g. \soar{<> }, \soar{>= 3.14}, \soar{<< disjunction of constants >>}) placed on the various variables that share an identity. It is used by the rule formation component to make sure that the learned rule only fires when all constraints required are met. - - \item \textbf{Operationality analysis} \newline - This component determines which conditions in an explanation trace tested working memory elements in a superstate. The rule formation component will use these conditions as a basis for the left-hand side of the chunk. While it does have a few key new differences, this is the one step that is similar to previous versions of chunking. - - \item \textbf{Rule Formation} \newline - The above four components performed the analysis that EBC needs to form a general but correct rule. This final component uses the results of that analysis to actually build the new rule. This is a complex component that has seven different stages. If a valid rule is created, Soar immediately adds the rule to production memory. + + \item \itextbf{Instantiation Creation} \newline + As a rule is fired, copies of the exact knowledge structures that matched each condition are made. + + \item \itextbf{Identity Assignment and Propagation} \newline + When a rule matches, identities are created and assigned for each element. The identity graph built during these rules firings is manipulated later for further generalization and identity sharing. + + \item \itextbf{Constraint Tracking} \newline + This component stores every value or relational constraint (e.g. \soar{<> }, \soar{>= 3.14}, \soar{<< disjunction of constants >>}) placed on variables. + + \item \itextbf{Collect Full Set of Inferences} \newline + When a rule creates a result with an identifier, children of that identifier also become results. This step determines which working memory elements become supergoal results. + + \item \itextbf{Dependency Analysis} \newline + The first part of a learning episode, this step backtraces through the behavior trace to determine which rule conditions test working memory elements in a supergoal. This works similarly in classical chunking. + + \item \itextbf{Identity Graph Manipulation} \newline + As backtracing occurs, the identity graph is also manipulated to build rules at the right level of generality and to prevent correctness issues in previous versions of chunking. + + \item \itextbf{Constraint Collection} \newline + Constraints recorded during constraint tracking that are encountered during backtracing are copied to the relevant identities in the identity graph. Later, during rule formation, these constraints are enforced on identities included in the learned rule. + + \item \itextbf{Rule Formation} \newline + The above eight components performed the analysis that EBBS needs to form a general but correct rule. This final component uses the results of that analysis to actually build the new rule. This is a complex component that has eight different stages. If a valid rule is created, Soar immediately adds the rule to production memory. \end{enumerate} + The following sections will describe each component in more detail. % ---------------------------------------------------------------------------- % ---------------------------------------------------------------------------- -\section{What EBC Does Prior to the Learning Episode} +\section{What EBBS Does Prior to the Learning Episode} \label{CHUNKING-prior} \index{result} \index{subgoal} -While most of the work that explanation-based chunking performs occurs during the learning episode, i.e. after a rule in a substate fires and Soar detects that a result will be created, some critical aspects of the analysis it performs also occur prior to the learning episode, during problem-solving in the substate. The two points when that happens is when a rule fires in a substate and when an operator is selected in a substate. +While most of the work that EBBS performs occurs during the learning episode, i.e. after a rule in a substate fires and Soar detects that a result will be created, some critical aspects of the analysis it performs also occur prior to the learning episode, during problem-solving in the substate. The two points when that happens is when a rule fires in a substate and when an operator is selected in a substate. % ---------------------------------------------------------------------------- -\subsection{Identity Assignment and Propagation} + +% ---------------------------------------------------------------------------- +\subsection{Operator Selection Knowledge Analysis} +\label{CHUNKING-prior-osk} +\index{Operator Selection Knowledge (OSK)|see{Context-Dependent Preference Set}} + +As described in the beginning of this chapter, chunking summarizes the processing required to produce the results of subgoals. Traditionally, the philosophy behind how an agent should be designed was that the path of operator selections and applications from an initial state in a substate to a result would always have all necessary tests in the operator proposal conditions and any goal test, so only those items would need to be summarized. The idea was that in a properly designed agent, a substate's operator evaluation preferences lead to a more efficient search of the space but do not influence the correctness of the result. As a result, the knowledge used by rules that produce such evaluation preferences should not be included in any chunks produced from that substate. + +In practice, however, it may make sense to design an agent so that search control does affect the correctness of search. Here are just two examples: + +\begin{enumerate} + \item Some of the tests for correctness of a result are included in productions that prefer operators that will produce correct results. The system will work correctly only when those productions are loaded. + + \item An operator is given a worst preference, indicating that it should be used only when all other options have been exhausted. Because of the semantics of worst, this operator will be selected after all other operators; however, if this operator then produces a result that is dependent on the operator occurring after all others, this fact will not be captured in the conditions of the chunk. +\end{enumerate} + +\index{chunking!correctness} +\index{chunking!relevant operator selection knowledge} +In both of these cases, part of the test for producing a result is \emph{implicit} in search control productions. This move allows the explicit state test to be simpler because any state to which the test is applied is guaranteed to satisfy some of the requirements for success. However, chunks created in such a problem space will not be correct because important parts of the superstate that were tested by operator evaluation rules do not appear as conditions. The chunks would not accurately summarize the processing in that problem state. The tracking of \textbf{Relevant Operator Selection Knowledge} (ROSK) is a way to address this issue. + +\index{chunking!backtracing} +Relevant operator selection knowledge is the set of necessary operator evaluation preferences that led to the selection of an operator in a subgoal. As previously described, whenever Soar learns a rule, it recursively backtraces through rule instances to determine which conditions to include in the final chunk or justification. With the ROSK, not only does Soar backtrace through each rule instance that created a matched working memory element, but it also backtraces through every rule instance that created preferences in the ROSK for any operator that gave those matched WMEs o-support. By backtracing through that additional set of preferences at each step of the backtrace, an agent will create more specific chunks that incorporate the goal-attainment knowledge encoded in the operator evaluation rules. + +Specifically, this component does two things: + +\begin{enumerate} + \item When an operator is selected, it analyzes the operator preferences that led to the decision, and caches any operator selection knowledge that played a necessary role in the selection. + + All necessity preferences, i.e. prohibit and require preferences, are always included in the ROSK since they inherently encode the correctness of whether an operator is applicable in a problem space. In contrast, some desirability preferences (rejects, betters, worses, bests, worsts and indifferents) are included in the ROSK depending on the role they play in the selection of the operator. + + How Soar determines which of those preferences to include in the ROSK is determined by the preference semantics it uses to choose an operator. During the decision phase, operator preferences are evaluated in a sequence of seven steps or filters, in an effort to select a single operator, as described in Section \ref{PREFERENCES}. Each step, or filter, handles a specific type of preference. As the preference semantics are applied at each step to incrementally filter the candidate operators to a potential selected operator, EBBS incrementally adds operator preferences to the ROSK based on the preferences that were instrumental in applying each filter. A more detailed explanation of the logic used at each step can be found in Section \ref{CHUNKING-subtleties-relevant-osk}. + + \item When an o-supported rule matches, EBBS caches the operator's ROSK in the instantiation of that rule. + + Since that selection knowledge was necessary to select the operator needed for the rule to match, chunking must backtrace through that knowledge. Cached ROSK is used to do this and incorporate the necessary operator selection reasoning knowledge into the learned rule. For some types of agent designs, including operator selection knowledge is needed to ensure correctness. +\end{enumerate} + +% ---------------------------------------------------------------------------- + +\subsection{Instantiation Creation} \label{CHUNKING-prior-identities} \index{instantiation} -Each instantiation describes the working memory elements that matched each condition and the working memory elements and preferences that are created by each action. With the introduction of EBC, all instantiations now also store the underlying explanation behind each condition and action as defined by the original rule: which elements in conditions are variables and which ones are literal constants, which variables are the same variables, what constraints must be met on the values of each variable and any relationships between variables. +Whenever a rule matches, EBBS creates a record called an instantiation that contains information about what specific working memory elements matched, what elements were created, and extends the instantiation graph by creating links from other instantiations to this instantiation for those instantiations that correspond to rule firings that created the elements tested in the rule. + +% ---------------------------------------------------------------------------- + +\subsection{Identity Assignment and Propagation} +\label{CHUNKING-prior-identities} +\index{identity-assignment} + +Each instantiation describes the working memory elements that matched each condition and the working memory elements and preferences that are created by each action. With the introduction of EBBS, all instantiations now also store the underlying explanation behind each condition and action as defined by the original rule: which elements in conditions are variables and which ones are literal constants, which variables are the same variables, what constraints must be met on the values of each variable and any relationships between variables. \index{chunking!identity} -EBC uses this underlying logic to determine the identities of objects used during the problem-solving. Identities are not simply IDs. Each identity is a declarative object that describes a set of variables across multiple rule firings and the various properties they hold. +EBBS uses this underlying logic to determine the identities of objects used during the problem-solving. Identities are not simply IDs. Each identity is a declarative object that describes a set of variables across multiple rule firings and the various properties they hold. -\emph{When an instantiation is created, EBC assigns all elements of every condition and action to an identity, creating new identities as necessary.} Identities are created and propagated using the following rules: +\emph{When an instantiation is created, EBBS assigns all elements of every condition and action to an identity, creating new identities as necessary.} Identities are created and propagated using the following rules: \begin{enumerate} \item If the same variable appears in multiple places in the same rule, it must be assigned the same identity. @@ -309,14 +370,14 @@ \subsection{Identity Assignment and Propagation} \item All variable elements of conditions that matched superstate WMEs - It is important to note that if two conditions both match the same superstate WME, each condition is considered independent. This means that each condition is assigned new identities for each of its elements and will produce its own condition in the final learned rule. This is a key way that EBC differs from previous versions of chunking. + It is important to note that if two conditions both match the same superstate WME, each condition is considered independent. This means that each condition is assigned new identities for each of its elements and will produce its own condition in the final learned rule. This is a key way that EBBS differs from previous versions of chunking. \end{enumerate} \item An existing identity is propagated for: \begin{enumerate} \item Any condition element that matched a substate WME with existing identities - Each element is assigned the identity found in the corresponding element of the action of the rule that created that WME. This propagates identities forward through the explanation trace, which allows us to represent that the variable in the condition refers to the same object as the variable in the action of the other rule. + Each element is assigned the identity found in the corresponding element of the action of the rule that created that WME. This propagates identities forward through the behavior trace, which allows us to represent that the variable in the condition refers to the same object as the variable in the action of the other rule. \item Any element that matches special working memory elements called \textbf{singletons} are assigned the same identity. @@ -328,56 +389,29 @@ \subsection{Identity Assignment and Propagation} \end{enumerate} \end{enumerate} -Note that rule 1 may conflict with other rules. For example, if a variable appears in two different conditions, then two different identities may propagate into each one of them. In such cases, rule 1 is always enforced and propagation is ignored. During the second phase of identity analysis, which occurs during the actual learning episode, EBC will re-examine all of the condition-action pairs as it performs a backward traversal of the explanation trace and fix the missing propagations. It does this by creating and manipulating an identity graph that can correctly incorporate all identity relationships. +Note that rule 1 may conflict with other rules. For example, if a variable appears in two different conditions, then two different identities may propagate into each one of them. In such cases, rule 1 is always enforced and propagation is ignored. During the second phase of identity analysis, which occurs during the actual learning episode, EBBS will re-examine all of the condition-action pairs as it performs a backward traversal of the behavior trace and fix the missing propagations. It does this by creating and manipulating an identity graph that can correctly incorporate all identity relationships. % ---------------------------------------------------------------------------- -\subsection{Relevant Operator Selection Knowledge Tracking} -\label{CHUNKING-prior-osk} -\index{Operator Selection Knowledge (OSK)|see{Context-Dependent Preference Set}} - -As described in the beginning of this chapter, chunking summarizes the processing required to produce the results of subgoals. Traditionally, the philosophy behind how an agent should be designed was that the path of operator selections and applications from an initial state in a substate to a result would always have all necessary tests in the operator proposal conditions and any goal test, so only those items would need to be summarized. The idea was that in a properly designed agent, a substate's operator evaluation preferences lead to a more efficient search of the space but do not influence the correctness of the result. As a result, the knowledge used by rules that produce such evaluation preferences should not be included in any chunks produced from that substate. - -In practice, however, it may make sense to design an agent so that search control does affect the correctness of search. Here are just two examples: - -\begin{enumerate} - \item Some of the tests for correctness of a result are included in productions that prefer operators that will produce correct results. The system will work correctly only when those productions are loaded. - - \item An operator is given a worst preference, indicating that it should be used only when all other options have been exhausted. Because of the semantics of worst, this operator will be selected after all other operators; however, if this operator then produces a result that is dependent on the operator occurring after all others, this fact will not be captured in the conditions of the chunk. -\end{enumerate} -\index{chunking!correctness} -\index{chunking!relevant operator selection knowledge} -In both of these cases, part of the test for producing a result is \emph{implicit} in search control productions. This move allows the explicit state test to be simpler because any state to which the test is applied is guaranteed to satisfy some of the requirements for success. However, chunks created in such a problem space will not be correct because important parts of the superstate that were tested by operator evaluation rules do not appear as conditions. The chunks would not accurately summarize the processing in that problem state. The tracking of \textbf{Relevant Operator Selection Knowledge} (ROSK) is a way to address this issue. +\subsection{Constraint Tracking} +\label{CHUNKING-constraint-tracking} +\index{constraint-tracking} -\index{chunking!backtracing} -Relevant operator selection knowledge is the set of necessary operator evaluation preferences that led to the selection of an operator in a subgoal. As previously described, whenever Soar learns a rule, it recursively backtraces through rule instances to determine which conditions to include in the final chunk or justification. With the ROSK, not only does Soar backtrace through each rule instance that created a matched working memory element, but it also backtraces through every rule instance that created preferences in the ROSK for any operator that gave those matched WMEs o-support. By backtracing through that additional set of preferences at each step of the backtrace, an agent will create more specific chunks that incorporate the goal-attainment knowledge encoded in the operator evaluation rules. +Rules underlying a behavior trace can include conditions that impose constraints on the values that a variable can hold. These can include equality tests, disjunctions, relative comparisons, literal tests, and conjunctions of constraints. These constraints must be included in later analysis of the behavior trace so that overgeneral rules are not created that fire when those constraints are not met. Thus, as instantiations are created, EBBS stores with each instantiation the constraints required for that rule to match. -Specifically, this component does two things: - -\begin{enumerate} - \item When an operator is selected, it analyzes the operator preferences that led to the decision, and caches any operator selection knowledge that played a necessary role in the selection. - - All necessity preferences, i.e. prohibit and require preferences, are always included in the ROSK since they inherently encode the correctness of whether an operator is applicable in a problem space. In contrast, some desirability preferences (rejects, betters, worses, bests, worsts and indifferents) are included in the ROSK depending on the role they play in the selection of the operator. - - How Soar determines which of those preferences to include in the ROSK is determined by the preference semantics it uses to choose an operator. During the decision phase, operator preferences are evaluated in a sequence of seven steps or filters, in an effort to select a single operator, as described in Section \ref{PREFERENCES}. Each step, or filter, handles a specific type of preference. As the preference semantics are applied at each step to incrementally filter the candidate operators to a potential selected operator, EBC incrementally adds operator preferences to the ROSK based on the preferences that were instrumental in applying each filter. A more detailed explanation of the logic used at each step can be found in Section \ref{CHUNKING-subtleties-relevant-osk}. - - \item When an o-supported rule matches, EBC caches the operator's ROSK in the instantiation of that rule. - - Since that selection knowledge was necessary to select the operator needed for the rule to match, chunking must backtrace through that knowledge. The operationality analysis component uses the cached ROSK to do this and incorporate the necessary operator selection reasoning knowledge into the learned rule. For some types of agent designs, including operator selection knowledge is needed to ensure correctness. -\end{enumerate} +Constraints are stored relative to identities used in an instantiation. If a rule tests that \soar{ < 7}, EBBS would record on identity node 1 associated with \soar{} that identity node 1 must have a value less than 7. Similarly, if another condition imposed the constraint \soar{<> } on \soar{} and \soar{} was mapped to identity node 2, then EBBS would record on identity node 1 that it cannot have the same value as identity node 2. % ---------------------------------------------------------------------------- -% ---------------------------------------------------------------------------- -\section{What EBC Does During the Learning Episode} +\section{What EBBS Does During the Learning Episode} \label{CHUNKING-during} -All of the previously discussed steps occurred during problem-solving in the substate as rules matched and operators were selected. It is worth noting that the analysis performed prior to the learning episode is persistent and can be shared across learning episodes. In other words, EBC can repeatedly re-use that analysis if it learns multiple chunks in the same substate. +All of the previously discussed steps occurred during problem-solving in the substate as rules matched and operators were selected. It is worth noting that the analysis performed prior to the learning episode (such as identity propagation) is persistent and can be shared across learning episodes. In other words, EBBS can repeatedly re-use that analysis if it learns multiple chunks in the same substate. Every time a rule fires in a substate, Soar checks to see if any of the working memory elements created by the rule qualify as results. This is when the actual learning episode begins. % ---------------------------------------------------------------------------- -\subsection{Calculating the Complete Set of Results} +\subsection{Collect Full Set of Inferences} \label{CHUNKING-during-results} \index{result} @@ -386,62 +420,57 @@ \subsection{Calculating the Complete Set of Results} The results produced by a single production firing are the basis for creating the actions of a chunk. A new result can lead to other results by linking a superstate to a WME in the substate. This WME may in turn link other WMEs in the substate to the superstate, making them results. Therefore, the creation of a single WME that is linked to a superstate can lead to the creation of a large number of results. All of the newly created results become the basis of the chunk's actions. % ---------------------------------------------------------------------------- -\subsection{Backtracing and the Three Types of Analysis Performed} +\subsection{Backtracing} \label{CHUNKING-during-backtracing} \index{chunking!backtracing} -When learning a new rule, EBC performs a dependency analysis of the productions that fired in a substate -- a process called backtracing. Backtracing works as follows. For each instantiated production that creates a subgoal result, backtracing examines the explanation trace to determine which working memory elements matched each condition. If the working memory element is local to the substate, then backtracing recursively examines the instantiation that created that condition's matched working memory element. Thus, backtracing traces backwards through all rules that fired and created working memory elements that were used to produce a result. +When learning a new rule, EBBS performs a dependency analysis of the productions that fired in a substate -- a process called backtracing. Backtracing works as follows. For each instantiated production that creates a subgoal result, backtracing examines the behavior trace to determine which working memory elements matched each condition. If the working memory element is local to the substate, then backtracing recursively examines the instantiation that created that condition's matched working memory element. Thus, backtracing traces backwards through all rules that fired and created working memory elements that were used to produce a result. -If an instantiation being backtraced through tested a selected operator, EBC will backtrace through each instantiation that created a preference in that operator's relevant operator selection knowledge set. This behavior is off by default and can be enabled with \soar{chunk add-osk on} (See Section \ref{chunk-add-osk}.) +%If an instantiation being backtraced through tested a selected operator, EBBS will backtrace through each instantiation that created a preference in that operator's relevant operator selection knowledge set. This behavior is off by default and can be enabled with \soar{chunk add-osk on} (See Section \ref{chunk-add-osk}.) -Multiple components of EBC perform their work during backtracing: operationality analysis, identity analysis and constraint tracking. The following sections will discuss what aspects of the agent's problem-solving are analyzed during backtracing. +Multiple components of EBBS perform their work during backtracing: dependency analysis, identity graph manipulation and constraint collection. The following sections will discuss what aspects of the agent's problem-solving are analyzed during backtracing. -\subsubsection{Operationality Analysis} +\subsubsection{Dependency Analysis} \label{CHUNKING-during-backtracing-operationality} The traditional core function of chunking's backtracing is to determine which conditions in the working memory trace tested working memory elements accessible to the superstate. These conditions will form the left-hand side of the rule. -The determination of which conditions to include is analogous to the concept of \textit{operationality} in explanation-based techniques. In classic EBL literature, operationality is typically defined as nodes in the explanation trace that are ``efficiently calculatable''. In terms of Soar's problem-state computational model, operationality can be defined as any condition that tests knowledge linked to a superstate. - -As EBC is backtracing through rules that fired in a substate, it collects all of these operational conditions. Once the entire explanation trace is traversed, the operationality analysis will have determined exactly what superstate knowledge was tested during the process of creating a result, which it then uses as the basis for the left-hand side of the newly learned rule. - -\textbf{Note:} Soar 9.6.0's explanation-based approach has led to one key change to Soar's operationality analysis. In previous versions of chunking, chunking would never add two conditions to a chunk that matched the same superstate working memory element. This made sense because chunking was based on a generalization of the working memory trace. More than one condition that tested the same WME would be redundant. Explanation-based chunking, though, learns based on the reasoning within the original hand-written rules. Since the reasoning behind each of the two conditions may be different even if they matched the same WME, EBC must always add both conditions. (Note that there are some exceptions. See Section \ref{CHUNKING-usage-tuning-conditions} on superstate singletons and user singletons.) +\textbf{Note:} Soar 9.6.0's explanation-based approach has led to one key change to Soar's dependency analysis. In previous versions of chunking, chunking would never add two conditions to a chunk that matched the same superstate working memory element. This made sense because chunking was based on a generalization of the working memory trace. More than one condition that tested the same WME would be redundant. Explanation-based behavior summarization, though, learns based on the reasoning within the original hand-written rules. Since the reasoning behind each of the two conditions may be different even if they matched the same WME, EBBS must always add both conditions. (Note that there are some exceptions. See Section \ref{CHUNKING-usage-tuning-conditions} on superstate singletons and user singletons.) \index{chunking!negated conditions} Negated conditions are included in a trace in the following way: when a production fires, its negated conditions are fully instantiated with its variables' appropriate values. This instantiation is based on the working memory elements that matched the production's positive conditions. If the variable is not used in any positive conditions, such as in a conjunctive negation, a dummy variable is used that will later become a variable in a chunk. If the identifier used to instantiate a negated condition's identifier field is linked to the super-state, then the instantiated negated condition is added to the trace as a negated condition. In all other cases, the negated condition is ignored because the system cannot determine why a working memory element was not produced in the subgoal and thus allowed the production to fire. -\subsubsection{Identity Analysis} +\subsubsection{Identity Graph Manipulation} \index{chunking!identity} -The first phase of identity analysis, forward identity propagation, occurred as rules fired and instantiations were recorded. Unfortunately, forward propagation alone will not produce correct identities. We previously gave one reason why this is the case -- conditions may have conflicting identities propagated forward -- but there are other, more complicated reasons as well that are beyond the scope of this document. What is important to know is that a second phase of identity analysis will be performed during backtracing that will refine and correct the limitations of the initial forward propagation of identity. This second phase achieves these corrections by building an identity graph, which represent the identities involved during problem-solving, and manipulating it as it backtraces through the explanation trace. +The first phase of identity analysis, forward identity propagation, occurred as rules fired and instantiations were recorded. Unfortunately, forward propagation alone will not produce correct identities. One reason why this is the case -- conditions may have conflicting identities propagated forward -- but there are other, more complicated reasons as well that are beyond the scope of this document. What is important to know is that a second phase of identity analysis will be performed during backtracing that will refine and correct the limitations of the initial forward propagation of identity. This second phase achieves these corrections by manipulating the identity graph during backtracing through the behavior trace. + -\subsubsection*{The Identity Graph} +\subsubsection*{The Identity Graph} %add stuff from earlier discussion of identity here and make the previous stuff less complicated. -The identity graph initially contains a node for each identity used in the explanation trace. Each node can have multiple edges that point to children identities and a single directed \textit{join edge} that initially points back to itself. As the agent backtraces through the explanation trace, EBC will manipulate the identity graph based on the condition-action pairs it encounters. +The identity graph initially contains a node for each identity used in the instantiation graph. Each node can have multiple edges that point to children identities and a single directed \textit{join edge} that initially points back to itself. As the agent backtraces through the instantiation graph, EBBS will manipulate the identity graph based on the condition-action pairs it encounters. \begin{enumerate} \item \textbf{Joining identities} \newline - If a condition matches an action with a conflicting identity, EBC performs a join operation between the two identities. This chooses one identity as the \textit{joined identity} and points the join edges of the other identity and any previously joined identities to the new \textit{joined identity}. + If a condition matches an action with a conflicting identity, EBBS performs a join operation between the two identities. This chooses one identity as the \textit{joined identity} and points the join edges of the other identity and any previously joined identities to the new \textit{joined identity}. - Note that any time EBC uses an element's identity, it is actually using the \textit{joined identity}. + Note that any time EBBS uses an element's identity, it is actually using the \textit{joined identity}. \item \textbf{Literalizing identities} \newline - If a condition/action with a variable element matches an action/condition with a literal element, EBC marks the identity as \textit{literalized}. This means that any conditions in the final chunk that have elements with that identity will be considered to have the NULL identity, just like constants, and will not be variablized. Instead, the matched value will be used for that element. + If a condition/action with a variable element matches an action/condition with a literal element, EBBS marks the identity as \textit{literalized}. This means that any conditions in the final chunk that have elements with that identity will be considered to have the NULL identity, just like constants, and will not be variablized. Instead, the matched value will be used for that element. \index{chunking!NULL identity} \end{enumerate} -\subsubsection{Constraint Tracking} +\subsubsection{Constraint Collection} \index{chunking!correctness} \index{chunking!over-general} -Our definition of operationality is very clear and allows us to almost trivially determine which conditions we should include in a learned rule, but it does have one shortcoming: non-operational conditions, which are ones that don't test working memory elements in the superstate, can transitively place constraints on the values of variables in operational conditions that \textit{will} appear in a chunk. If our learning algorithm does not include these constraints, the learned rule can apply to situations where the previous substate reasoning could not have occurred, which means that the learned rule is over-general. - -To handle this limitation, \textbf{\textit{EBC keeps track of all constraints found in \\ non-operational conditions that it encounters while backtracing}} in the following manner: +Some constraints in the behavior trace may not test working memory elements in the superstate, but could transitively place constraints on the values of variables in conditions that \textit{will} appear in a chunk. To handle this, once backtracing is complete, EBBS: \begin{itemize} - \item It stores constraints on the value a single identity, for example \soar{>= 0}, \soar{< 23}. - \item It stores relational constraints between two identities, for example \soar{> }, \soar{< } or \soar{<> }. - \item EBC stores all of these constraints based on the underlying identities, not the variables used. For example, if a variable \soar{} had the constraint \soar{<> }, EBC would record that the variables that share the identity of \soar{} cannot have the same value as variables that share the identity of \soar{}. + \item stores constraints on the value a single identity, for example \soar{>= 0}, \soar{< 23}. + \item stores relational constraints between two identities, for example \soar{> }, \soar{< } or \soar{<> }. + \item stores all of these constraints based on the underlying identities, not the variables used. For example, if a variable \soar{} had the constraint \soar{<> }, EBBS would record that the variables that share the identity of \soar{} cannot have the same value as variables that share the identity of \soar{}. \end{itemize} % ---------------------------------------------------------------------------- @@ -450,39 +479,38 @@ \subsection{Rule Formation} \begin{wrapfigure}{R}{\thirdwidth} \insertfigure{chunking-rule-formation.png}{.9\thirdwidth} - \caption[The seven stages of rule formation]{} + \caption[The eight stages of rule formation]{} \label{fig:chunking-rule-formation} \vspace{-40pt} \end{wrapfigure} -There are seven distinct, sequential stages to rule formation. The following sections will give a brief overview of each one. +There are eight distinct, sequential stages to rule formation. The following sections will give a brief overview of each one. -\subsubsection{Condition and Action Creation} +\subsubsection{Conditions and Action Creation} This stage creates the basis for the left-hand and right-hand side of the rule. -To create the initial conditions of the chunk, it copies all conditions in the explanation trace that were flagged as operational during backtracing. These initial conditions contain literal values for each element. -To form the actions of the chunk, it creates copies of the actions that produced each of the result and all children of those results that came along for the ride. +To create the initial conditions of the chunk, it copies all conditions in the behavior trace that were flagged during backtracing. To form the actions of the chunk, it creates copies of the actions that produced each of the result and all children of those results that came along for the ride. -\subsubsection{Enforcement of Constraints} +\subsubsection{Identity-based Constraint Enforcement} -This stage adds all constraints on non-operational conditions that were collected during backtracing. As previously described, each constraint is indexed in terms of the identity it constrains. So, \textit{if the identity being constrained exists in one of the conditions of the learned rule}, EBC will enforce the constraint by adding a new test to that condition. +This stage adds all constraints on conditions that were collected during backtracing. As previously described, each constraint is indexed in terms of the identity it constrains. So, \textit{if the identity being constrained exists in one of the conditions of the learned rule}, EBBS will enforce the constraint by adding a new test to that condition. \index{chunking!literalization} -One situation in which attaching a constraint can be tricky occurs \textit{when the constrained identity has been literalized but the constraint itself refers to an identity that has not been literalized}, for example \soar{\{ > 3 \}}. While that constraint references a condition element that can only match a value of \soar{3}, the relationship between \soar{3} and the identity of \soar{} must still hold (assuming \soar{} appears in a different element somewhere else in the rule.) Since these constraints still need to be enforced to ensure a correct rule, EBC will invert the constraint and attach it to a variable in another condition. In this example, it would add a \soar{< 3} to some other condition with an element that had \soar{}'s identity. +One situation in which attaching a constraint can be tricky occurs \textit{when the constrained identity has been literalized but the constraint itself refers to an identity that has not been literalized}, for example \soar{\{ > 3 \}}. While that constraint references a condition element that can only match a value of \soar{3}, the relationship between \soar{3} and the identity of \soar{} must still hold (assuming \soar{} appears in a different element somewhere else in the rule.) Since these constraints still need to be enforced to ensure a correct rule, EBBS will invert the constraint and attach it to a variable in another condition. In this example, it would add a \soar{< 3} to some other condition with an element that had \soar{}'s identity. -\subsubsection{Identity-Based Variablization} +\subsubsection{Identity-Based Generalization} To achieve any useful generality in chunks, identifiers of actual objects must be replaced by variables when the chunk is created; otherwise chunks will only ever fire when the exact same objects are matched. At this point in the algorithm, all of the real work needed to determine the most general but correct variablization has already been performed by the identity analysis component. So, this step simply needs to replace all elements with non-NULL identities with variables, making sure that elements with the same joined identity are assigned the same variable. This step also makes sure to skip and elements with identities that have been flagged as \textit{literalized}. -\subsubsection{Merging Redundant Conditions} +\subsubsection{Condition Merging} -Any two conditions in the learned rule that share the same identities in all three elements can be combined. In such cases, it is logically impossible for those two conditions to match two different WMEs and cause the same rules to match in the substate. (If the two conditions were to match two different WMEs, at least one of the other rules in the explanation trace that had unified the two conditions would not have matched.) As a result, EBC can safely merge those two conditions without losing generality. +Any two conditions in the learned rule that share the same identities in all three elements can be combined. In such cases, it is logically impossible for those two conditions to match two different WMEs and cause the same rules to match in the substate. (If the two conditions were to match two different WMEs, at least one of the other rules in the behavior trace that had unified the two conditions would not have matched.) As a result, EBBS can safely merge those two conditions without losing generality. -\subsubsection{Polishing Conditions} +\subsubsection{Condition Simplification} -EBC polishes the conditions of the learned rule by pruning unnecessary constraints on literalized elements and replacing multiple disjunction constraints with a single simplified disjunction. +EBBS polishes the conditions of the learned rule by pruning unnecessary constraints on literalized elements and replacing multiple disjunction constraints with a single simplified disjunction. \begin{enumerate} \item \textbf{Merging disjunctions:} @@ -492,25 +520,30 @@ \subsubsection{Polishing Conditions} If an element in a condition has been literalized but also has a literal constraint on its value, then the constraint is unnecessary and will be thrown out. For example, \soar{ \^{}value \{ < 33 23 \}} becomes \soar{ \^{}value 23}. \end{enumerate} -\subsubsection{Validating Rule and Repairing Unconnected Conditions} +\subsubsection{Rule Validation and Repair} \label{CHUNKING-repair} \index{chunking!repair} -At this point, the rule is essentially formed. Chunking must now make sure that the learned rule is fully operational and can be legally added to production memory. A fully operational rule does not have any conditions or actions that are not linked to a goal state specified in the rule. +At this point, the rule is essentially formed. Chunking must now make sure that the learned rule can be legally added to production memory. Specifically, the rule cannot have any conditions or actions that are not linked to a goal state specified in the rule. -If an unconnected action or condition is found, EBC will attempt to repair the rule by adding new conditions that provide a link from a state that is already tested somewhere else in the rule to the unconnected condition or action. +If an unconnected action or condition is found, EBBS will attempt to repair the rule by adding new conditions that provide a link from a state that is already tested somewhere else in the rule to the unconnected condition or action. -To repair the rule, EBC performs a search through working memory to find the shortest path of working memory elements that lead from a state identifier in the rule to a WME with the identifier in the unconnected condition or action. A new condition is then added for every WME in that found path, which is then variablized. +To repair the rule, EBBS performs a search through working memory to find the shortest path of working memory elements that lead from a state identifier in the rule to a WME with the identifier in the unconnected condition or action. A new condition is then added for every WME in that found path, which is then variablized. -Note that there may be multiple paths from a state to the unconnected identifier. EBC does a breadth-first search, so it will find one with the shortest distance. +Note that there may be multiple paths from a state to the unconnected identifier. EBBS does a breadth-first search, so it will find one with the shortest distance. -\subsubsection{Re-ordering Conditions} +\subsubsection{Conditions Re-ordering} Since the efficiency of the Rete matcher depends heavily upon the order of a production's conditions, the chunking mechanism attempts to sort the chunk's conditions into the most favorable order. At each stage, the condition-ordering algorithm tries to determine which eligible condition, if placed next, will lead to the fewest number of partial instantiations when the chunk is matched. A condition that matches an object with a multi-valued attribute will lead to multiple partial instantiations, so it is generally more efficient to place these conditions later in the ordering. This is the same process that internally reorders the conditions in user-defined productions, as mentioned briefly in Section \ref{ARCH-pm-structure}. +If the constructed rule is not a duplicate of an existing rule, it is added to production memory. + +\subsubsection{Add and Recursively Learn Additional Rules} + +When a rule is added, it may generate another learned rule. Once the learned rule is added to production memory, it is guaranteed to immediately match in the supergoal. When it does, the chunk makes new data structures and EBBS creates an instantiation for the chunk that includes new identities relative to the existing identity graph. In the case when a problem has been decomposed into at least two subgoals and the newly created chunk makes results for an even higher level goal, EBBS then backtraces and creates a second chunk based on the reasoning in the subgoal that the chunk matched it. This process stops when a chunk only generates working memory elements in the same state it matched in. % ---------------------------------------------------------------------------- % ---------------------------------------------------------------------------- -\section{Subtleties of EBC} +\section{Subtleties of EBBS} \label{CHUNKING-subtleties} % ---------------------------------------------------------------------------- @@ -522,7 +555,7 @@ \subsection{Relationship Between Chunks and Justifications} As a result, Soar uses a limited version of the chunking algorithm to do that. It analyzes the substate problem-solving and learns a new, temporary rule, a ``justification'', which is added to production memory. If this temporary rule tests an operator in the superstate, it gives the result o-support. (Note that when learning is on, a justification is not needed since the chunk will provide the correct support.) -Justifications use all the components described in the following sections and are even affected by the current chunk settings.\footnote{ +Justifications use all the components described in the previous sections and are even affected by the current chunk settings.\footnote{ Even though they don't contain variables, justifications can be over-general because they don't incorporate enough knowledge, for example, operator selection knowledge.} You can even print justifications out like other rules. The only differences between chunks and justifications are: @@ -554,7 +587,7 @@ \subsection{Chunks Based on Chunks} \subsection{Mixing Chunks and Justifications} \label{CHUNKING-subtleties-justifications} -If an agent is using the \soar{only} or \soar{except} setting, then justifications will be built in states where learning is disabled and chunks will be built in states where learning is enabled. In these situations, justifications also serve another purpose: they provide an explanation of the results for future learning episodes in states that do have learning on. EBC does this by retaining all of the extra information that chunks have but justifications do not, namely those extra tests and how things would have been variablized. This allows EBC to learn chunks from justifications as readily as it can from hand-written rules and other chunks. +If an agent is using the \soar{only} or \soar{except} setting, then justifications will be built in states where learning is disabled and chunks will be built in states where learning is enabled. In these situations, justifications also serve another purpose: they provide an explanation of the results for future learning episodes in states that do have learning on. EBBS does this by retaining all of the extra information that chunks have but justifications do not, namely those extra tests and how things would have been variablized. This allows EBBS to learn chunks from justifications as readily as it can from hand-written rules and other chunks. When mixing justifications and chunks, users may want to set the explainer to record the learning episodes behind justifications. This allows one to examine the reasoning behind a justification just like you would a chunk, which may be important if that justification later participates in the formation a chunk. See Section \ref{explain} for more information about the explainer's settings. @@ -580,9 +613,9 @@ \subsection{Over-specialization and Over-generalization} \index{chunking!over-general} \index{chunking!explanation-based chunking} -Explanation-based chunking was pursued to address the main limitation of traditional chunking: \textit{over-specialized rules} that were very specific and could not be applied to many other situations. Specifically, EBC's identity-based variablization and constraint tracking/enforcement has eliminated the core source of this issue. +Explanation-based behavior summarization was pursued to address the main limitation of traditional chunking: \textit{over-specialized rules} that were very specific and could not be applied to many other situations. Specifically, EBBS's identity-based variablization and constraint tracking/enforcement has eliminated the core source of this issue. -The nature of EBC's algorithm does add two new situations in which rules may become over-specialized. Section \ref{CHUNKING-subtleties-rhs-functions} discusses how variables used in certain RHS functions need to be literalized to maintain correctness, which can cause overspecialization. Section \ref{CHUNKING-subtleties-repair} discusses how testing or augmenting a previous result creates non-operational rules that require repair, a process which may sometimes over-specialize a rule. Note that this situation can easily be avoided and, even when it does occur, may not add much unnecessary specificity to learned rules. +The nature of EBBS's algorithm does add two new situations in which rules may become over-specialized. Section \ref{CHUNKING-subtleties-rhs-functions} discusses how variables used in certain RHS functions need to be literalized to maintain correctness, which can cause overspecialization. Section \ref{CHUNKING-subtleties-repair} discusses how testing or augmenting a previous result creates rules that require repair, a process which may sometimes over-specialize a rule. Note that this situation can easily be avoided and, even when it does occur, may not add much unnecessary specificity to learned rules. While over-specialization may no longer be a common problem, it is still possible to get over-general rules. Several of the sources of correctness issues listed in the next section can produce over-general rules in certain situations. @@ -590,18 +623,18 @@ \subsection{Over-specialization and Over-generalization} \subsection{Previous Results and Rule Repair} \label{CHUNKING-subtleties-repair} -An agent may learn a slightly over-specialized rule when EBC repairs a rule that has unconnected conditions, which are conditions that have an identifier that is not linked to one of the states referenced in the rule. Such rules are illegal and cannot be added to Soar's production memory. +An agent may learn a slightly over-specialized rule when EBBS repairs a rule that has unconnected conditions, which are conditions that have an identifier that is not linked to one of the states referenced in the rule. Such rules are illegal and cannot be added to Soar's production memory. -\textit{Rules that require repair are caused by substate problem-solving that tests or augments a previous result.} A previous result is a working memory element that was originally created locally in the substate but then later became a result when a rule fired and connected it to the superstate. (At which point a chunk must have been learned.). If another substate rules later matches or augments such a previous result WME \emph{using a path relative to the local substate}, then EBC will have problems. It will know that the WME is in the superstate -- so conditions that test the WME are considered operational and augmentations on that identifier are considered results -- but it won't know where in the superstate that working memory is located is and how it should be referenced in the learned rule, because the problem solving referenced the result relative to the local substate. +\textit{Rules that require repair are caused by substate problem-solving that tests or augments a previous result.} A previous result is a working memory element that was originally created locally in the substate but then later became a result when a rule fired and connected it to the superstate. (At which point a chunk must have been learned.). If another substate rules later matches or augments such a previous result WME \emph{using a path relative to the local substate}, then EBBS will have problems. It won't know where in the superstate that working memory is located is and how it should be referenced in the learned rule, because the problem solving referenced the result relative to the local substate. -As described in Section \ref{CHUNKING-repair}, EBC repairs the rule by adding new \textit{grounding conditions} that provide a link from a state, which is tested somewhere else in the rule, to the unconnected condition or action. It does this by searching through working memory to find the shortest path from a state to the identifier behind the unconnected element. It then variablizes those conditions appropriately. +As described in Section \ref{CHUNKING-repair}, EBBS repairs the rule by adding new \textit{grounding conditions} that provide a link from a state, which is tested somewhere else in the rule, to the unconnected condition or action. It does this by searching through working memory to find the shortest path from a state to the identifier behind the unconnected element. It then variablizes those conditions appropriately. \index{chunking!over-specialization} Since the conditions are based purely on what happened to be in working memory at that point and nothing in the explanation dictated that particular path found during the search, the learned rule may be over-specialized. The chunk will only match future situations where the previous result can be found on that same path. Fortunately, new chunks can be learned to ameliorate this. If a similar situation is encountered in the future, but with a different path to the unconnected element, the chunk won't fire, because the added grounding conditions won't match, which should cause the agent to subgoal and learn a similar chunk with a different set of grounding conditions. Note that if an agent designer expects that the path to the previous result found by the search will always exist, a repaired rule should match just as generally as an unrepaired rule. -But if this is not the case, an agent designer can avoid this situation by modifying the rules that test or augment the substructure of a previous result. \emph{If those rules are modified so that they match the previous results by referencing them relative to the superstate than the the local substate, EBC will be able create a valid rule without any repair.} +But if this is not the case, an agent designer can avoid this situation by modifying the rules that test or augment the substructure of a previous result. \emph{If those rules are modified so that they match the previous results by referencing them relative to the superstate than the the local substate, EBBS will be able create a valid rule without any repair.} To detect when this is happening, use the \soar{chunk stats} command. (See section \ref{debugging-explanation-based-chunking} It will tell you if any of an agent's learned rules that required repair. If you instruct the explainer to record the chunk, you can also see whether a specific chunk was repaired by looking at the chunk's individual stats @@ -611,14 +644,14 @@ \subsection{Missing Operator Selection Knowledge} If an agent uses rules that create operator preferences to choose amongst multiple operators in the substate, it is possible that the reasoning behind those rules needs to be incorporated in any rule learned. This topic is discussed in greater detail in Section \ref{CHUNKING-prior-osk}. -EBC will incorporate relevant operator selection knowledge if you enable the chunk setting \soar{add-osk}, which is off by default. (See Section \ref{chunk-add-osk}.) +EBBS will incorporate relevant operator selection knowledge if you enable the chunk setting \soar{add-osk}, which is off by default. (See Section \ref{chunk-add-osk}.) % ---------------------------------------------------------------------------- \subsection{Generalizing Over Operators Selected Probabilistically} If the problem-solving in a substate involves operators that were selected probabilistically, chunking will not be able to summarize the agent's reasoning into a correct rule. For a rule to be correct, it must always produce the same result that the substate would have produced if the learned rule was not in production memory. Since a different operator could have been selected which could have resulted in different problem-solving, the substate could easily produce different results than any chunk learned in that substate. -Future versions of chunking will provide an option to prevent rules from forming when a probabilistically-selected operator was chosen during problem-solving. Until then, agent engineers can disable learning in states that involve such reasoning. +EBBS detects these cases while it is analyzing operator selection preferences. EBBS adds a flag to inndicate whether an operator was selected in an unreliable way (using random selection, numeric preferences, or reinforcement learning). EBBS checks for this flag during backtracing and if it is encountered, EBBS will not add the rule to production memory. % ---------------------------------------------------------------------------- \subsection{Collapsed Negative Reasoning} @@ -626,7 +659,7 @@ \subsection{Collapsed Negative Reasoning} \index{chunking!negated conditions} \index{chunking!over-general} -\emph{Over-general chunks can be created when conditions in the explanation trace test for the absence of a working memory elements in the substate.} Since there is no clear way for chunking to generate a set of conditions that describe when a given working memory element would not exist in a substate, chunking can't represent that aspect of the problem-solving. +\emph{Over-general chunks can be created when conditions in the behavior trace test for the absence of a working memory elements in the substate.} Since there is no clear way for chunking to generate a set of conditions that describe when a given working memory element would not exist in a substate, chunking can't represent that aspect of the problem-solving. Chunking can include negated tests if they test for \textit{the absence of working memory elements in the superstate, though}. So, the agent engineer can avoid using negated conditions for local substate data by either (1) designing the problem-solving so that the data that is being tested in the negation is already in the superstate or (2) making the data a result by attaching it to the superstate. This increases the number of chunks learned, but a negated condition of knowledge in the superstate can be incorporated correctly into learned rules. @@ -644,10 +677,10 @@ \subsection{Problem-Solving That Doesn't Test The Superstate} Over-general chunks can be created if a result of a subgoal is dependent on the creation of an impasse within the substate. For example, processing in a subgoal may consist of exhaustively applying all the operators in the problem space. If so, then a convenient way to recognize that all operators have applied and processing is complete is to wait for a state no-change impasse to occur. When the impasse occurs, a production can test for the resulting substate and create a result for the original subgoal. This form of state test builds over-general chunks because no pre-existing structure is relevant to the result that terminates the subgoal. The result is dependent only on the existence of the substate within a substate. -In these cases, EBC will learn a chunk with no conditions, which it will reject. But the superstate result is still created by the substate rule that matched. If a new rule is learned that uses that result, it will be over-general since the rule does not summarize the reasoning that led to the result, namely that all operators were exhaustively applied. +In these cases, EBBS will learn a chunk with no conditions, which it will reject. But the superstate result is still created by the substate rule that matched. If a new rule is learned that uses that result, it will be over-general since the rule does not summarize the reasoning that led to the result, namely that all operators were exhaustively applied. \index{quiescence} -The current solution to this problem is a bit of a hack. Soar allows an agent to signal to the architecture that a test for a substate is being made by testing for the \soar{\carat quiescence t} augmentation of the subgoal. If this special test is found in the explanation trace, EBC will not build a chunk. The history of this test is maintained, so that if the result of the substate is then used to produce further results for a superstate, no higher chunks will be built. +The current solution to this problem is a bit of a hack. Soar allows an agent to signal to the architecture that a test for a substate is being made by testing for the \soar{\carat quiescence t} augmentation of the subgoal. If this special test is found in the behavior trace, EBBS will not build a chunk. The history of this test is maintained, so that if the result of the substate is then used to produce further results for a superstate, no higher chunks will be built. % ---------------------------------------------------------------------------- \subsection{Disjunctive Context Conflation} @@ -658,7 +691,7 @@ \subsection{Disjunctive Context Conflation} If this WME is then tested to create a result on the superstate, the chunk that is subsequently created can produce different behavior than the substate would have. In the original subgoal processing, multiple matches produced one substate WME, but that one substate WME only created a single result in the superstate. The chunk on the other hand will match multiple times for each of the items that previously created the substate WME. And then, each one of those matches will create its own distinct result in the superstate. Since this is different behavior than the original substate, this rule would be considered incorrect. -If it were possible, EBC should learn a \textit{disjunctive conjunctive condition}, with each disjunction being the superstate conditions tested by each substate rule that had previously created the substate WME that was repeatedly asserted. This is why this potential source of incorrect rules is called \textit{disjunctive context conflation}. +If it were possible, EBBS should learn a \textit{disjunctive conjunctive condition}, with each disjunction being the superstate conditions tested by each substate rule that had previously created the substate WME that was repeatedly asserted. This is why this potential source of incorrect rules is called \textit{disjunctive context conflation}. If this type of reasoning is needed, agents can move the conflating WME to the superstate. The rule learned would then produce only one result regardless of the number of rules that repeatedly created that WME. @@ -674,12 +707,12 @@ \subsection{\texorpdfstring{Generalizing knowledge retrieved from \\ semantic or \item \textbf{Justification for a memory recall is opaque to agent} \hfill - EBC does not have access to the reasoning behind why a piece of knowledge was recalled from a memory system. For example, consider the case of a semantic memory that is recalled because it has the highest level of activation at a particular time. In a future situation, the same semantic memory may not be the most active, in which case something else would be recalled and different problem-solving could occur. Because of that possibility, the original rule is not guaranteed to produce the same result and hence has the potential to be incorrect. (Note that this can also occur with episodic memory queries.) + EBBS does not have access to the reasoning behind why a piece of knowledge was recalled from a memory system. For example, consider the case of a semantic memory that is recalled because it has the highest level of activation at a particular time. In a future situation, the same semantic memory may not be the most active, in which case something else would be recalled and different problem-solving could occur. Because of that possibility, the original rule is not guaranteed to produce the same result and hence has the potential to be incorrect. (Note that this can also occur with episodic memory queries.) \item \textbf{Knowledge from semantic or episodic memory recalled directly into the substate is considered local} \hfill \index{chunking!over-general} - To understand why this is a problem, remember that a chunk's conditions are based on the conditions in the explanation trace that tested knowledge linked to a superstate. (See section \ref{CHUNKING-during-backtracing-operationality} for more information.) If semantic or episodic memory is recalled directly into the substate, then any conditions that test that recalled knowledge is considered local to the substate and will not be included as a condition in the chunk. So, even though the substate reasoning required some piece of semantic knowledge to exist, the chunk will not require it. And, since the learned rule \textit{is not} incorporating some of the reasoning and constraints that involved the recalled knowledge, the rule may be over-general. + To understand why this is a problem, remember that a chunk's conditions are based on the conditions in the behavior trace that tested knowledge linked to a superstate. (See section \ref{CHUNKING-during-backtracing-operationality} for more information.) If semantic or episodic memory is recalled directly into the substate, then any conditions that test that recalled knowledge is considered local to the substate and will not be included as a condition in the chunk. So, even though the substate reasoning required some piece of semantic knowledge to exist, the chunk will not require it. And, since the learned rule \textit{is not} incorporating some of the reasoning and constraints that involved the recalled knowledge, the rule may be over-general. To avoid this situation, an agent can retrieve the knowledge in a higher-level state rather than the substate in which the rule is learned. \end{enumerate} @@ -713,7 +746,7 @@ \subsection{Determining Which OSK Preferences are Relevant} Add any best preferences for remaining candidates to the ROSK. \item [Worst Filter] If any remaining candidate has a worst preference which leads to that candidate being removed from consideration, that worst preference is added to the ROSK. Again, the logic is that the conditions that led to that candidate not being selected allowed the final operator to be chosen. - \item [Indifferent Filter] + \item [Indifferent/Numeric Filter] This is the final stage, so the operator is now selected based on the agent's exploration policy. How indifferent preferences are added to the ROSK depends on whether any numeric indifferent preferences exist. \begin{enumerate} \item If there exists at least one numeric indifferent preference, then every numeric preference for the winning candidate is added to the ROSK. There can be multiple such preferences. Moreover, all binary indifferent preferences be- tween that winning candidate and candidates without a numeric preference are added. @@ -733,15 +766,15 @@ \subsection{\texorpdfstring{Generalizing Knowledge From Math \\ and Other Right- \label{CHUNKING-subtleties-rhs-functions} \index{chunking!literalization} -Explanation-based chunking introduces the ability to learn more expressive rules whose actions perform arbitrary right-hand side functions with variablized arguments. +EBBS introduces the ability to learn more expressive rules whose actions perform arbitrary right-hand side functions with variablized arguments. -It is important to note that this ability is limited. EBC can only learn rules with generalized RHS functions in its actions when the rule that created the result contained a RHS function. In many cases, RHS functions will be used in the intermediate rule firings in the explanation trace. Not only will these intermediate RHS function not appear in the chunk, but any chunk learned based on their output will become more specific. This is one of the sources of over-specialization referenced in section \ref{CHUNKING-subtleties-overspecialization} on over-specialization. +It is important to note that this ability is limited. EBBS can only learn rules with generalized RHS functions in its actions when the rule that created the result contained a RHS function. In many cases, RHS functions will be used in the intermediate rule firings in the trace. Not only will these intermediate RHS function not appear in the chunk, but any chunk learned based on their output will become more specific. This is one of the sources of over-specialization referenced in section \ref{CHUNKING-subtleties-overspecialization} on over-specialization. -RHS function calls in intermediate rule firings are a challenge for EBC to deal with because the problem-solving may have placed constraints on the intermediate results that cannot be represented in a single Soar rule. +RHS function calls in intermediate rule firings are a challenge for EBBS to deal with because the problem-solving may have placed constraints on the intermediate results that cannot be represented in a single Soar rule. For example, consider the case of one rule that used a RHS function to add two numbers. Now consider another rule that matched the output of the RHS function, but only if it was less than 5. If the second rule matched, it would return the total as a result. How could we encode the reasoning of those two rules into one rule? Since Soar's production syntax does not allow using RHS function as constraints in conditions, there is no way to insure that the two numbers add up to something less than 5 in a single rule. This is why RHS functions in intermediate rule firings can cause over-specialization. -Because the chunk's conditions can't represent constraints on the output of intermediate RHS functions, EBC must literalize both the identities of the variables that appear as arguments to the intermediate RHS function, as well as the identities in any conditions that test the output of the RHS function. That fixes the value of the RHS function and guarantees that any constraints in conditions that test the output of that RHS function will be met. While this will make the learned rule more specific, it will also ensure that the rule is correct. +Because the chunk's conditions can't represent constraints on the output of intermediate RHS functions, EBBS must literalize both the identities of the variables that appear as arguments to the intermediate RHS function, as well as the identities in any conditions that test the output of the RHS function. That fixes the value of the RHS function and guarantees that any constraints in conditions that test the output of that RHS function will be met. While this will make the learned rule more specific, it will also ensure that the rule is correct. % ---------------------------------------------------------------------------- \subsection{Situations in which a Chunk is Not Learned} @@ -769,7 +802,7 @@ \subsection{Situations in which a Chunk is Not Learned} \index{chunking!correctness} \item \textbf{The problem-solving in the substate violated one of the enabled correctness guarantee filters} - During the development of explanation-based chunking, we have developed a list of possible causes of incorrect chunks. EBC's correctness guarantee filters detect when those situations occur and prevents a chunk from being learned. + During the development of explanation-based behavior summarization, we have developed a list of possible causes of incorrect chunks. EBBS's correctness guarantee filters detect when those situations occur and prevents a chunk from being learned. \index{chunking!over-general} For example, the \soar{allow-local-negations} filter will prevent a rule from being formed if the problem-solving that led to the result was dependent on a condition that tested whether a subgoal WME doesn't exist. Since there is no practical way to determine why a piece of knowledge doesn't exist, testing a local negation can result in an over-general and incorrect chunk. See Section \ref{CHUNKING-usage-tuning-correctness} on page \pageref{CHUNKING-usage-tuning-correctness} for more information. @@ -806,18 +839,17 @@ \subsection{Overview of the \soar{chunk} command} =================================================== Chunk Commands and Settings =================================================== -? | help Print this help listing -timers [ on | OFF ] Timing statistics (no args to print stats) -stats Print stats on learning that has occurred +chunk ? | help Print all EBC settings +chunk stats Print statistics on learning that has occurred ------------------- Settings ---------------------- -ALWAYS | never | only | except When Soar will learn new rules -bottom-only [ on | OFF ] Learn only from bottom substate -naming-style [ numbered | RULE] Simple names or rule-based name -max-chunks 50 Max chunks that can be learned (per phase) -max-dupes 3 Max duplicate chunks (per rule, per phase) +always | NEVER | only | except When Soar will learn new rules +bottom-only [ on | OFF ] Learn only from bottom sub-state +naming-style [ numbered | RULE] Simple numeric chunk names or informational rule-based name +max-chunks 50 Maximum chunks that can be learned (per phase) +max-dupes 3 Maximum duplicate chunks (per rule, per phase) ------------------- Debugging --------------------- interrupt [ on | OFF ] Stop Soar after learning from any rule -explain-interrupt [ on | OFF ] Stop Soar after learning explained rule +explain-interrupt [ on | OFF ] Stop Soar after learning rule watched by explainer warning-interrupt [ on | OFF ] Stop Soar after detecting learning issue ------------------- Fine Tune --------------------- singleton Print all WME singletons @@ -825,16 +857,11 @@ \subsection{Overview of the \soar{chunk} command} singleton -r Remove a WME singleton pattern ----------------- EBC Mechanisms ------------------ add-ltm-links [ on | OFF ] Recreate LTM links in original results -add-osk [ on | OFF ] Incorporate operator selection knowledge -merge [ ON | off ] Merge redundant conditions -lhs-repair [ ON | off ] Add grounding conds for unconnected LHS -rhs-repair [ ON | off ] Add grounding conds for unconnected RHS -user-singletons [ ON | off ] Use domain-specific singletons +add-osk [ ON | off ] Incorporate operator selection knowledge ---------- Correctness Guarantee Filters ---------- Allow rules to form that... allow-local-negations [ ON | off ] ...used local negative reasoning -allow-opaque* [ ON | off ] ...used knowledge from a LTM recall -allow-missing-osk* [ ON | off ] ...tested operators selected using OSK -allow-uncertain-operators* [ ON | off ] ...tested probabilistic operators +allow-opaque [ ON | off ] ...used knowledge from a LTM recall +--------------------------------------------------- \end{verbatim} * disabled } @@ -845,7 +872,7 @@ \subsection{Overview of the \soar{chunk} command} \subsection{Enabling Procedural Learning} \label{CHUNKING-usage-enable} -By default, explanation-based chunking is off. +By default, chunking is off. \begin{itemize} \item To turn on chunking: \soar{chunk always} @@ -867,7 +894,7 @@ \subsection{Enabling Procedural Learning} \end{itemize} \end{itemize} -Depending on your agent design, you may want to consider enabling the \soar{add-osk} option. As of Soar 9.6.0, EBC does not incorporate operator selection knowledge into learned rules by default, since there is a performance cost and not all agents designs require its inclusion. You may want to enable this option if your agent has rules that test knowledge in the superstate to create operator preferences in the substate. See section \ref{CHUNKING-prior-osk} on page \pageref{CHUNKING-prior-osk} for more information about learning and operator selection knowledge. +Depending on your agent design, you may want to consider enabling the \soar{add-osk} option. As of Soar 9.6.0, EBBS does not incorporate operator selection knowledge into learned rules by default, since there is a performance cost and not all agents designs require its inclusion. You may want to enable this option if your agent has rules that test knowledge in the superstate to create operator preferences in the substate. See section \ref{CHUNKING-prior-osk} on page \pageref{CHUNKING-prior-osk} for more information about learning and operator selection knowledge. See Section \ref{chunk} on page \pageref{chunk} for more information about using the chunk command to enable and disable procedural learning. @@ -889,9 +916,9 @@ \subsubsection{Using singletons to simplify a rule's conditions} \index{singleton} \index{chunking!singleton} -Unlike previous versions of chunking, EBC adds all conditions that tested superstate knowledge to a chunk, regardless of whether another condition already tested that working memory element. This means that EBC can sometimes produce learned rules with seemingly duplicate conditions. While these conditions are logically correct, they may be redundant because the nature of the domain may make it impossible for the two conditions to match different working memory elements. For example, in the blocks-world domain, the fact that there can be only one gripper in the world means that having multiple conditions testing for a gripper would be redundant. +Unlike previous versions of chunking, EBBS adds all conditions that tested superstate knowledge to a chunk, regardless of whether another condition already tested that working memory element. This means that EBBS can sometimes produce learned rules with seemingly duplicate conditions. While these conditions are logically correct, they may be redundant because the nature of the domain may make it impossible for the two conditions to match different working memory elements. For example, in the blocks-world domain, the fact that there can be only one gripper in the world means that having multiple conditions testing for a gripper would be redundant. -Soar allows agents to specify such known domain characteristics, which EBC will then use to create better rules that don't have such unnecessary conditions. We call any working memory element that is guaranteed to only have a single possible value at any given time, a \textit{singleton}. If EBC encounters two different conditions in the backtrace that both test the same superstate WME that matches a user singleton pattern, it will merge the two conditions. As described in Section \ref{CHUNKING-singletons1}, there are several architectural singleton's that EBC already knows about. To specify patterns for domain-specific singletons, the \soar{chunk singleton} command can be used. +Soar allows agents to specify such known domain characteristics, which EBBS will then use to create better rules that don't have such unnecessary conditions. We call any working memory element that is guaranteed to only have a single possible value at any given time, a \textit{singleton}. If EBBS encounters two different conditions in the backtrace that both test the same superstate WME that matches a user singleton pattern, it will merge the two conditions. As described in Section \ref{CHUNKING-singletons1}, there are several architectural singleton's that EBBS already knows about. To specify patterns for domain-specific singletons, the \soar{chunk singleton} command can be used. See Section \ref{chunk} for more information about the chunk singleton command. @@ -932,27 +959,31 @@ \subsubsection{Chunking Statistics} =========================================================================== Explanation-Based Chunking Statistics =========================================================================== -Substates analyzed 0 Rules learned 0 Justifications learned 0 --------------------------------------------------------------------------- Work Performed --------------------------------------------------------------------------- -Number of rules fired 0 +Sub-states analyzed 0 +Number of rules fired in substates analyzed 0 Number of rule firings analyzed during backtracing 0 -Number of OSK rule firings analyzed during backtracing 0 -Number of rule firings re-visited during backtracing 0 Conditions merged 0 Disjunction tests merged 0 -- Redundant values 0 -- Impossible values eliminated 0 - Operational constraints 0 Non-operational constraints detected 0 Non-operational constraints enforced 0 +--------------------------------------------------------------------------- + Problem-Solving Characteristics +--------------------------------------------------------------------------- +Rules repaired that had unconnected conditions or actions 0 +Chunk used negated reasoning about sub-state 0 +Chunk tested knowledge retrieved from long-term memory 0 +Justification used negated reasoning about sub-state 0 +Justification tested knowledge retrieved from long-term memory 0 + --------------------------------------------------------------------------- Identity Analysis --------------------------------------------------------------------------- @@ -960,37 +991,15 @@ \subsubsection{Chunking Statistics} Distinct identities in learned rules 0 Identity propagations 0 Identity propagations blocked 0 -Identity propagations from local singleton 0 Identities joined 0 -- To unify two identities propagated into same variable 0 -- To unify two conditions that tested a superstate singleton 0 -- To connect an child result (result in rule had children WMEs) 0 Identities literalized 0 -- Condition with variable matched a literal RHS element 0 -- Condition with variable matched a RHS function 0 -- Condition with literal value matched a RHS variable 0 -- Variable used in a RHS function 0 - ---------------------------------------------------------------------------- - Potential Generality Issues Detected ---------------------------------------------------------------------------- -Rules repaired that had unconnected conditions or actions 0 -Extra conditions added during repair 0 - ---------------------------------------------------------------------------- - Potential Correctness Issues Detected ---------------------------------------------------------------------------- -Chunk used negated reasoning about substate 0 -Chunk tested knowledge retrieved from long-term memory 0 -Justification used negated reasoning about substate 0 -Justification tested knowledge retrieved from long-term memory 0 --------------------------------------------------------------------------- Learning Skipped or Unsuccessful --------------------------------------------------------------------------- Ignored duplicate of existing rule 0 Skipped because problem-solving tested ^quiescence true 0 -Skipped because no superstate knowledge tested 0 +Skipped because no super-state knowledge tested 0 Skipped because MAX-CHUNKS exceeded in a decision cycle 0 Skipped because MAX-DUPES exceeded for rule this decision cycle 0 \end{verbatim} @@ -1015,7 +1024,7 @@ \subsubsection{Interrupting Execution To Examine Learning} % ---------------------------------------------------------------------------- \section{Explaining Learned Procedural Knowledge} -While explanation-based chunking makes it easier for people to now incorporate learning into their agents, the complexity of the analysis it performs makes it far more difficult to understand how the learned rules were formed. The explainer is a new module that has been developed to help ameliorate this problem. The explainer allows you to interactively explore how rules were learned. +While explanation-based behavior summarization makes it easier for people to now incorporate learning into their agents, the complexity of the analysis it performs makes it far more difficult to understand how the learned rules were formed. The explainer is a new module that has been developed to help ameliorate this problem. The explainer allows you to interactively explore how rules were learned. When requested, the explainer will make a very detailed record of everything that happened during a learning episode. Once a user specifies a recorded chunk to "discuss", they can browse all of the rule firings that contributed to the learned rule, one at a time. The explainer will present each of these rules with detailed information about the identity of the variables, whether it tested knowledge relevant to the the superstate, and how it is connected to other rule firings in the substate. Rule firings are assigned IDs so that user can quickly choose a new rule to examine. @@ -1146,7 +1155,7 @@ \section{Explaining Learned Procedural Knowledge} \textbf{\soar{explain instantiation }} -This command prints a specific instantiation in the explanation trace. This lets you browse the instantiation graph one rule at a time. This is probably one of the most common things you will do while using the explainer. +This command prints a specific instantiation in the behavior trace. This lets you browse the instantiation graph one rule at a time. This is probably one of the most common things you will do while using the explainer. Tip: Use the alias \soar{i } to quickly view an instantiation, for example: @@ -1320,7 +1329,7 @@ \section{Explaining Learned Procedural Knowledge} %\frame{ \includegraphics[width=\textwidth]{chunking-trace-identity.png} % } - \insertcaption{A colored visualization of an explanation trace} + \insertcaption{A colored visualization of a behavior trace} \label{fig:chunking-viz-identity} \end{figure}