From 6d7324214f83227fe94636fce3f1634ff6367974 Mon Sep 17 00:00:00 2001
From: juaristi22 <juaristi@uni.minerva.edu>
Date: Fri, 27 Feb 2026 12:30:03 +0530
Subject: [PATCH 1/5] paper review

---
 paper/main.pdf                           | Bin 547662 -> 543392 bytes
 paper/sections/abstract.tex              |   2 +-
 paper/sections/appendix_benchmarking.tex |   2 +-
 paper/sections/appendix_robustness.tex   |   2 +-
 paper/sections/background.tex            |  16 ++++----
 paper/sections/conclusion.tex            |   6 +--
 paper/sections/data.tex                  |   6 +--
 paper/sections/discussion.tex            |  48 ++++++++---------------
 paper/sections/introduction.tex          |  16 ++++----
 paper/sections/methodology.tex           |  26 ++++++------
 paper/sections/results.tex               |  18 ++++-----
 11 files changed, 64 insertions(+), 78 deletions(-)

diff --git a/paper/main.pdf b/paper/main.pdf
index e4042bc0d1444cfe32af2f7d9ce52f44b52d48ee..daeda5ed18b7a8b59e63d6e6d095536603be5529 100644
GIT binary patch
delta 124108
zcmZtNLwF@jv@YP-wr%d%>e#kzn>$X&wv&$4vE8w4+qUk1&K=%6sOPC+jjG=I)>_jo
zqy?#@3F4qE$%QmDi4O3<41y#O64=08ISad!Ik^M&NY<+%?DZma`KL-6o_NC@WCsMB
zF-YP>-!oDQ`NU|)mknwh%0HuGWMQZ>bcyh(G<vY9wt8?iHA)pylnNSuXdQTsS?Qb5
zs8-Nx!WJrKGo7@Vno!&OWfREE_e<y~Nd@|OhIpRlQ>924#+*EW3HE=a>P@JQ{>gGJ
z!|+#D58!6x3^ue!6pK^UGZ>j^{W5>WhcDAw7?ftqA)NNwrU9AvF=0p1{>+87PsJLB
zxw=gPRXK4nE|F4W3WHRuaaM5lkl0P72nt?~q)1la7M@b!a@L`sYF5IutPYVxj~*hS
ztLJfx=ptyDAZ3mKqP24o^qnY)bf_IF{CQvQg2Nt`OIiF&yLhJ|qk6*J_1+3K>1?dx
zD1tN4r&d7TJZ~smJM_MdRq=ojlp^Qz0Lp=K5qW>C()UMLz7!n@3T_Tp8urK#!if1g
z#t@)MLen^G*vHNK#NKZ+{|-Nk*dUAnqp6{Qk;bB==p`W!JTDBUHf04S9t`|IoKd|9
zf^Rv2_ZfLyl+Y!Zr#@jLg}>Ad26M8!i3Krd4@ki&*&TO-$7q|@^U^eG`1FFvW3^&R
zHFCZ<z>c+hlEMrj{-Dz&Z!?fgOM}}6@|3Yd1oM{d7Qt=uAmY;}KqEgRF^e$7{1MKd
z0Fe<vx4aDp4zNz!{2);H#xKq{B4PW!tMO+>QP(XKuk4Sx-WAZ*c|1Iw*Za`q31suP
z1p;^TyAxD;o+DC)E<cSgkZpk(Z-d*!wF7kzw-@KeZm3%Zhl|8fBp1xUSF!XBXaY8g
z$8T1tla$v}<ovI~#|ZdL8DJjNE~I*45zrt^5-S2&m%qSOH$e6f__VP1a8oXkF0#v5
zZXU1W^H;}`13svnX9Trw+4lXo=2X7T=JPLH7CH=df0R0jBQ6BuZK~*0a?9~aN@eBX
z`3$YBcyxbKk|qR8^QZ2qI`Skt8h`WlM@H|GTT|i{SOf1RnGy^h6H)-8i3cNkD*NX!
z_Zte3J?7(^Xx~@@5^FEQF+yn@A_Edp1#JchExTeee*3vqjLn)1I+%US>QF_x4q$rA
ztL-d8VAdhMC*6Gew_6Pr1_PPuF*LRkXsMWP%hw+Tw%YvLOE&jGHn1J>dJyk}zxu%=
zroPda)kEZAg%{#vOO?5TjPi+9qvS>8eZm2l8Y~#YX_=D7Xl^pHk`PSWeI>?!!Y2FW
zrXO!JoDMoemR0T3#BR@6CUQ<ifHY2A?AFOCZ(Ukxk9D`W(j-xHCgZ7y$><}U`Wj-A
z35|R{wQs5BgX<WT5jxMHmy4y?<|b6a*#R4cj=wd~b8SU}N7;n{CcVi%?^(=J?lTL#
zI4Cxezqm9(5p_m?0#P6fTxmMQW;d{<;p)QX^l5Np2he}Nt?gY&P!AqOC50odTqUK}
zS{qQbn>%@;zkF$JMYRZB80EW{=WqGdZ_To%!4^lM_Ha(lszC6YN0dOBb#INnBSZ3F
z&SXSM5R?>T>CNM<XW)&CbeBa8ggyWr{hr%P<q1~`)dUt<qq|{9HAdzGlzz39im0M5
zUb^@QQ=%#vo^dyqT+iu((Bt9!q%9E9SpR9M4$+~{f@(=gHlMgySro0+sYV(z7bwpZ
zJfkoRtY^D&aT7el`|NCve!oOs!Wj~r^=Pa$><o}KpVQCI$K!7m0uL+LyNv-}tzb9N
z6E;;78HYEhbM`9b8!8#L``Qd+O!0;?pOK-$8@ji^Hs6Lo+5LW(8+c1@$;X`H#cY=F
zRoP9kfA$bf7`mTH^ssxW8RYJ=Gp`CS!aw(eY}8&Ms_Qt8#?941^6Q~)z)cVEbbMPO
z4zPY2SU}<Wb4b90jaJeI!6FB&;pO4bnLko8F9m%?Y_!yB4v$J=G4q}D6A5C@^e&%B
zRy=>^H{g_W#8@J~`_1@=?~J@<sx_smfsZEsvSGv8dZd`Q;|QSF)?K8aG=KfK^wSpz
z|7G8I+fryCSlqI4=c={T?#fj*-E%>lhM*}c=oY>$F!OgXYf_+b=5Glw)kdK+R<Ju$
z<|#b#{93{0cks&M0l98NEKG&yNvUh#r9C!+5OVyg6E4dK(=fghOaP%Z>qYmfWK#jx
z`3;AA_2~6juthtjx--h`L65-{Z;t_Hv0%fK9A%5`Ljh6)WDQI1QOT;^_>$fC)19=K
zhv`i_T($GgC0_DPE~)}BQ!z0=gc9yuz33%-3E4EqX^#a&ZHZj{=Y>z7dTY#gR|&hE
zFSg0X)K<+DO_!~`<3jcRcM{rKbq-WrRq~*=lI41k`wttRtI#U<5kbWrVWYxvyj35^
z^_6Z9&~3sMFBj0Vcv56Y?f5QH<TcP{DW9}`2kgJRc*MH>&36U&&GqxNPmD1PA(`w0
zO#{x#&G|nsBQRD=*7blBt>;7YmKD>+_TMd{izE6Cw=>-s|KTE015Y2CUEyQ|6vc=}
zz*oQ);`!c^+Ot@}oN%!raR)ARya>Uk7v|5@OGIbYjpIG9)ImAJ0$H0cj2aQ9HCBo^
zVMP<dHD^DuvF%E2>y-oE)c#CoesAws<pQI$iiErJ9#HJ~`@|%3#QZhnY&p6Kp|!?C
zvBg)9p^+Q0PG1ONV(GAFK5GsBDQ$YW_==BLNXtSzMNRv!+(UWX7i9p)L5m?ny)D~l
zZN&E2?u!-X*lAe+-eij-C+pYCucgAdskWz=+7OEPhAz3cZD-seoTRocg>IZHENQ)n
zpU4HfM4&n>qDkz53g$PzwPnQIxUo3aqOyKcB2lo6xHTq*GOytaaXJOl1zn}_jd_~5
zbt25kRFRCR>y6W^!Qr2CM#OPI?7}>}st6dWbm9X~WMxr<LU2h`G5^d_f#=KM>7x(7
zjKXeuFy#oqB9b1<K=z&{reYwLs8C;LX|GQ~8R#@l68de-P8jY^z=_vWJwTo5ua6-P
zqk8%=?6p0e5xX|Z`Vsb<FKw2~qZ>Q;+O3iJ@-_`e{d6*Yj5iKZr+I{C_buc?Hx62l
zNMh0z=xw>;HdAwF!s$PPM*1!W&1|ap!dcAV*=>pM-byMNf&a1McUAwgLTKh-uxkrU
z2EKi|?_57}Im8?-kCgA@P{JMNRO}X;0%En}1?elDHu3PvZ}YhDlVshxH`)!NTLyMb
ziDd7+n~EH%Cl06jV@AXIYf!&`8v+}yemmxL3)cA4omd(tk&f=`OjUgLMgoWRE6df+
zCgoVNDQ2}CmZM+fTM6FBBDo;~QY9;^f&U%W8M!eLc7b6tvO|%;BwpgT$q#$})q*qo
z^ysg==?ABrj)Tz4Y27FGXGXoIW%)+z`N+G>;;R#5*`hLwrp{qms-yIDelIs)k7Ex`
z{{Vmh<6txco+#gVx@k??f#eJojYRPk>X6m@B#{Ni0nbSoLF1Dzq!K-pTGk6P@NsI;
zHq>heQ8ikxz2IN$m0RmE0TUaEK5AX9<E?b)tnB0?E4nd`p5ip=E{OxifTYQVrz+dV
zpB%?OhZ5hn7xDDG{AlC8L5UXg+t`z-hYj+_b!g-H57zTr=4D#X*>nRlFd!{S<FH1?
zkP+1;O65_vAsfQ3tKlf=t3Z}4pey0v5;jPECXY0Rtq=bizlnEXQQlG6_RVKHPauUw
z@3}JC3N$6RVMsq{jN$N@avl|2gd2Df#K~pYEq)*Th!e`=wKd9%nKW!fTY%3dTy~mb
zy&p;X1NiT=tbAP@Jlwfe#E*!Frh8>dJF{m8w){%T!hm-_a&XH)Z0<BLptzaw#Y!|0
zKq0B6e31F16uhM0+|2C;^G`k9&S)Pl97&{Uq0{MHx@cXw1wF#Sd~q~5dlQ=AE*~uR
z9ySh2kPL)Ip%0_t03v<aJrm7JX1cRC*qn7<7g{DYK>-?ChLaQx66~acXz=E%#-Ha)
zLuD{MKUlvj?or{Ljp4XPAQ&}44YnO#rl)_aot<S<ay!lM&z<!xCu!2U36?$tAqq?z
zhK1*Ng_{;HTM?x_BoU((ayw?%76xacIaqIfGG2!f5e*#hCjidt#HzxxJ%d1xEjhy<
zSmh(oKjz*OmI{{45^G-+!pS3nvq3%H8OM&HvyYPQISD8WxlXwqKoHa~!aaBG%G+E9
z<d%|KJ`&tH7}+Bf^3bKG;+G&lW{?W}UZdq?J_=@K2R=1~W0m<jibyBK1}u~7;;sY0
zpBHx29WJc7D`wG9&kJR!)zktSDZj^AQgZ@9*?W_%cT;Zzm$8R`nPK-`!q%TLzEqF%
zukPNFoDFBEs$kJBz-Q#u^$8@>ocix4e*Tz}O?kAe-IaaP;}IzLH5+$&{^{zgX&+gn
zUwmft!_pPoU8ybx4ec5aos!wKKjwB{W8c;dzrVYcs1`|}M~W<BnJP$fBbf_EeTS@@
z3l-jRbmn8KS&1&AsU0}6Bq+ZqDd7IOR@qC|N#|h!T9M^(fMQM~Z<r@19k>$rmCu?5
zQR-%H@l)bTyx;|L>J&-8!uxd+?}_#|`zTSe5dZlFovYsx@+s6W-jFTP8~!$V?=m+6
zx#20jEjn(6x0U=PjVOGlB<ONFBB!$m>}^aji^3LE2~~kpl>!5TP!x@k*g1_Bo69c9
z?fy#WANl?w1t$N!d_=VfBTnSv4A8czv<tA!2r4`sc=Fdh=!DdG3XHIR-AhkJ$T41-
z2*ra&Zh3w^o{i#SbmZo;PxW3fiRA-be*(*PBD$OECxfEKDVE4mDujf8qMVRzV-t<^
z!Q>`+?eH=n20@B73|8h|ijng-8U5G9$L`$ILp2bN2dql&fuJZGsqs?r29%<{xgq9{
zS>q39fkD-)Y#YWr=T%F(7el=2<-hmDW*rX;MtU<pa!Kb%sHcAAf5pla{UXdLc?u3R
zj(>^qg=+`#lf^0ng8O6p7aK+(iAR2y#hp*ST+kdOL9_!m7#qZN*!+}jri6Zdmqr>Y
zBeOc$0z-0n`rb06+fdYVahrcl%}%_lQok;YsUhD(9;5D$9vb6OfB`JRGMBwzRitH3
zm{UyQC-+%fph7pv8NS|HLe7|=O!;~-Da1PP=xiY>PM8odev-lT=4Fvs6ID`i!Wk{$
zP)U1pXZX8ocU*;H+46y1HJ002DgBww-?aDlK)24~ZtP!)=fTxdutx3s{RuWgCf!l}
zIS_H*4RgB>@tP36JN{1Z%p34^xJOmlqmGF{mrZHl1;TdXu=ATwG&A+vP`c}!xLm{l
zsX#`U`1?q79gAh19ISaW->0)tZqg?E<|>x7O#isq{M#ea%}s5O9_4o}9pX8ER5F<s
zuvHWIi3^RNVBAMjRkfR$xlU4{mU7!0PNk*qpOl{{j+&xe(+~x$#8fso7EzaImzfEu
zDp?xq<$4YMrZH2_B;c)#+m0$SGd1&m^*#B3^R%i3$u;{Mv#}!ct<rsSTadxleSCAF
zCzs41b5GHCK~TYym%Kvt#_WzccmijhjEqVR#=`a=g=m4=vdJ55sQ*c*+yz$Me9TXE
zJ*TV1GI^2r!mE@n!W{_*ElKHk0))r(ZKcdGfd<i%PI}2eLaZ2+9dFA}h)VAgdv8Bo
zv+(oV@8x_-zCmAX&L`XaZr`DmW?vlDVc9r~@w#&8CWuYm(3Xoq(>W2kbDFbm&ovo(
zP=u|fbqnZUgHK=<WS#De%D<amTS|PDA%Ca1)SPR+ErU32z1DtdbyOf{yDqSBeMfy7
z4vxGfqOCvXP~#U<^TuhjeQTUq64YSP{Q0uAvlm=NqhDq{jjc>V5sCIbL^^>1rI9{C
zGaG)?ESDpNEttx*1EsG^?5&}Osz8CDoPiKl=n9m!RKE3^)iA7C^Q=5(Cf`TQjmZum
zc-++YayEd7byk2FS6q4FnmX&wsPFV#X9d@D2k2jTff2<7m`@O-vMfQ!`NI}BrLFnS
zN>l3QY2r;ZmVJP?|6@e*%2GcnsBsRV-P*M83Dy4X2AFU{Wn@^V3h?_r!kp~+d2rPD
z0ssq=Db}D~NnWeYn;DE{4L*q=>+-ik)BYI~=-2cY7eBeAVo5uB3V)FwOjk+>F&)Uy
zuGPt{Uufv6XA|pqK~>R(*!&w29XO**G81hXH1iP+i{bB)tRY}-B8h|Yxp}U6>(5RI
zqv^6kg)}UpNd3~i+RXRrQDan0j*DI3w*htrovaBaHZ|%GHnrH+cnEx`U!?PA?HTDD
zE6EKs+DuV!>9j-~Y+h`|dl?`wY4iC}7^OUzfWm${g%C~{H~nEU63OhR*id<B+hEyA
zMO^49xytd=v=X=Hxg}F(bJ&S(S&XwYLB+ga*f5asj*bt+8l8r^fdnfLid6%#qz4kU
zL58Wpz+CPa+4&m|j{CZRnTbv?b8bhp29d`j${culmh^Yl?>Gh<Xmk5iryenKH9%$|
zZcnq*jEuH5XC?POb7F{#gp>s(p5Z}Cc0#)x_e1jdn4AiqlA~j+p3Nl&NwU!R&Niek
z?MRq5%ca~p{Jk|$LXu8}nys3vNCbAz%eS^x3S}h0mO}o{9NHL`&9UzmIkZa2l148s
zVp%6?3?ycli^1n5@BO7hH5UF?A*h=erNZf!HR%&UqJem-hqB%!61?rYMKiTwi-VUR
zo<>tMb`StW(%yY?Ov9h$iOMJ~0C3R&5dE6m&<Q%`&O|je`F)q8On(*Q0Dx9Xl|*_l
zm>E=mQM=NPVS8X)0#)PXnX{mO)j{=bYj}tn<_6{IBodP>2epOz=`1}G;X|<{a3ZQ^
zMD^*yU>kuQ{DSu+t&$uld9->dP?dw06PJbC?98~9KXz@vAW`ZoQbn}O4-{{unI?gz
zk&#$-eSSu&RUje1LjhcliU2<fpxG%YoM~M7jRJ?58$Q6U3_fVD4j-h|ARQ}hJ64Kg
zvZvu5)^wfR_QIQHJNNA5sQIA*3la=?aU@2f3}~y>{Xuz>?R*dve&rIReSeIx*nPMG
zyz2T}SGXXFVHi2!G>9g~qp>hS$(0PCRnWhL*B!xlNT%}Ip@9S$TR<MXeT4lc4kL%x
zzcayj!2wy64anu*N*?j3M3Ji9Z05|?(i4*@c7Xs)WZzMKIUYAG5w0x|6Ye}O6KN?3
zsaC=)fwlsyVF~||ueYYN1~Qbwl<Xni8`ro*E$MxkM3!J%XB$v94Yq{NkTX=mB%0UI
z4SG2qVN=RRW2Yofa9|1F_7MVPUq7`Kk<$Q<n+p_^#8U`8Kf{dPrOqu2<SH&ZH$71s
zj&p1M$O$7zAe9JF9KTqI=jRwei~)o<m)^YuaE!;o5-0+O)`ugK$T&Xtyl|oTVx59K
zJD_K4zOhD~uxa~oXdWq15UH(m(!1--_O+GZY&<?}jbNTu3+%Lj@&nm?o6cc|&Wy)H
z`7Qui+W6tkdHsB}#(0~Uklk)c<ybPP4YbLgo2;|IvH9wS0ms3JP=3{ul5N=4NDp4S
z<!2d8TSAr)>QLQ3foLRAmtbD1LLwYznS)toG3p7mN=ZI~4Wz|^BB#TqV(I8Ye{qu&
zp$Mwk-u#0t4z$M@xd1TilI3I>&AZpu=B4!Q`Jj7mZ;wb!iag`A__M%|`11xQw=fvm
zU(Y?H2;66!A4~j8&cILwxmSoZmGrKl2yn)zO(yG&f=5_`^^zBby^uiDD9V}J`hrjd
z$NQo``T@Kc<QMVA2r4Z194}sT>1L)KiiaCEyENQm0k7}uA<CQw-e#1?1V;4-$X=^o
z09Sob1~Y4$k`MZ<nB|n)7t7u-Bm+Y1v3`-_Wp-%Jz8ONqk}#qFxL*+#^oGR*i@{Pt
z^o&wPSQHYG*L>JHf6QnFahWs!kWHt{s}O)bS04u^Bw^Bl9<V2=@Rm^Ef<zm#3zWeD
z*!&Sofe6qWu^-^B9mV+URD6_*aN(^n&T0v(b)WoqwB*BKHUzv{lGcz!<d&yCoLFb?
zmpS#!tI6DO{AU|SwqPY`VDw0B?2BphtWv6DJIT9-6S0RRdf-NtkYIOswD<cLdIzUv
z;?Qkc&5hl-On=4B6Ugzt4GyC-@^i+m8zxcufhF#I_BCx{SWSy$x3+v2=`!^&hARA_
zJ^YaKS50_9EUCq}A945Aec8B$jtEq}Dd?rTas98d9L88sHxj-IJPEgvlikGrv=Qu&
zO%HPgri-3!Pk$pmXkSvjNx%E>K`M<*lIGhpZ^%K5A9c@lT78)@48Va~f8KMMm#D%3
z|0TpHaJSBI&sD0pH0EuJH^))ec5vHmyO%QD9rr`7`2~R?{tn$S6s4xdj8|OIfg~f}
zZm4<Nn}^x7;Sqr^s3de!=nf@Q7=~d@wwn<%WRv{2z4q;8OOdio^bV|mC$O!r5VDsz
zAhWYz?<x4+D;%9m`8Nx7mEVj!3%-#&@F3mE789*1m+nbrU0;!Ztm@`5MRl>izAyb^
z{o+~1nRZF`TtBi_t3Ufkw;y-)9b(IgTSWY~2BIjL;*-R7QE4cOwBF!e`Z1@7z<@@a
zv?|*Wj!c=dc?cwKoCNjLWDFg%G3yjuurwW0+d;C$Dp*kf9WsEh(1O4>xg~A|s9c65
zgp>WI8%3n#a}p|6VT+vkvRgKN(+_bkt8H(g9BFhNAJBm;?^yjfUl2NXI<`0-ot*G7
zAWbovBBt_~(?uGYCx)xt#>eP3;B`b5i9C2qS_sMp5tS^QB@wYkXOBEe)4saQrDf_h
zHSq~`;@1+?$OSW*V-E-#hdnd_;-32)Neu3JlP7{u<^Q6dHV{6!c#{Z3PJb%_ML|o_
zbQ^w=uNhX^@m|6D{VG^Zcja7luY8K^X~mTGr;}kyl0+-^W^$)zNn6opAHxnEHuu%p
z_k-)r2yq}<n)D?SALR2~uo&V!Vx1i&ZpOEV5&HbENVtCnm<d?|i>8PIkQ8rz4t|R~
zs5{F&-W0)<UBCrP*cPKj!P8Vec3b|w=_63AX&~GgF2d~f_5w3(ha`4@Z_?OL_pQ9*
z4|-fX8P-#AOnX}eRsRaLGER4Up=7`Dp{%#5PErU58WKqBYcH&`pf$IV=D1Qmj-hh2
zapeLK)W-U?!2U3uG~+`Rc&9O$R-{jt7jNc}LQ<khMD8`bz}E?l_|vDFSRL90-mjPm
zz6>GGORJK|W+Bm0v{=DS)Frk;J8#2O*e;VYuplHviO*rL9lvBfPhw=$EXJ+LRzJVU
z#j!w<wlMqI5HG>kbUC8*;NsKw0|;VF<>T;S9(U{Och~v@UtYdUaaWoE^#nN+(W&_Z
znis;30cY-D=IZ8RZfyTQJ4X{6I9B##XB%_~R&ExS^t)LIS|Hw9+-BR|mf@V^9EgK9
zzZ=hL1Oovx3dQdOLZg*Pbm8J-xd{!~#+3E#v6V8W+~Ul7Y^3{ybS!O#q@fP&ODoUD
zXb&O&=);e%qekpH`5sle$pft$raI4x?Gj}y3C?YOB99-Kchfn;%bKWQS68Q=TD_62
zqS|(6y7&8EWx)P=8wZH}cT?Yb=+#1BB$DFbBA#0<pxdfa|A*9?A~r8P0kT@zhO6PD
z`o`#cYikdq*O9SeU&YoK2h3$|wH62VPuJ%{^ooaboK>qPXA|e1X-?&QYg3r%<xxd4
zyPOJSG7L}G>PS0mfE#79={!H+lc%lraS4+Usp6NGFYuXF>k+b*H!!t!Lud5<WVsUw
zISI0xK{Th3RPbpedKGF25wx-3cHmH8zIOwDC|dOS@66gDMDEF?vmH!5;*1dHH#sa^
zghjB;#q;qs*@Bl>RWk0BGRD7{>n%poRw8SO-+WsdJ#^7N!l%<u0FSE;^wv|e)z=t`
zwxa8JH{kEom}5MnA{>r_@MctDF>}|VeUiPI!t-(0l|(>6G#yzo#IKk>3H@9;Z9_;v
zdGHB$f+9}zlj?VheWRfgeYo=S3Aju~6Rm<L9b$7-&1v4);HK8_;@{MH=UahMOu4;-
zSjYafmPHh*bBe`GJ7@3eQv|A0ssV&kl68eaYG6wd*YD7X3m+_Ku(Y&D6hH<2Uy|$5
z`YHdCyt<B*C~cgVQVgysW$S&t$BpVa3A)!aGSyRhb|)37?x61SpLVd=BCIMr_F7#a
zS##*Km(3;23*9WJAO_f)*6~AhEnmI^HNta%=csf;F9TevOqwKhdbxZ<3Iv!ITBZ~j
zDNq=lu*eM>pzW5-OI6Fo&tx-2SPF-?y|f#r9=Psb?O@)NDqmTPzhY@9W;CLUEx(j&
zRE(%^ArDeTz1+ug{Wm+xq%UC9>3;F|;XiuUy6$6UdW62A`ZDnMUqNmSN(Gs3tFx<S
zf5A)#A^6M$3(>$k%#dL{I%eHKJ|W-WTY!{VNC{kI575;JSn?_c!1|2lIT<Y&T0e0l
zsEvGSF^D5%UYGl6z@T@y`?NoXUkXxy>}`0Ui#kUfsJbXG9V#-&CBPcVB&jiF?53F8
z<}0l|R|gr6ke=V2RieP_w=lvnNnDL-kZk-u=<n4e+cw=6cf@I^5PBxnp!J=GTi~26
z5yf*-Fo*|zBNC?67B@3$Y><A?o{m%>I>}dpuNYtRLq?#{!4>k%E=FCzi>Bdcr%kKE
zsn2qD5t2Emf`QTLTL^s;)j5sKvTB0nsq$oj!e19{c6;)1h^^_T6W!&bO5C-j<;mvj
zBa<=gx@5b((|G_?Yr(?oVUf{IGcZcSH4P4SiXAWafS4*<>ve`f9x6q2qbbAnFs-@<
ziy7{i|D_1k%-JI2-ev+VB`Bi=I_0KCBPe#^9}>-qZq+~Iyi)LC5=$+~nM6CB5ptGZ
zRtdB#l?;4ZGA=m~SLEu^Sh1uUcR4H8=4|6Hq8!mAC{KLZN|;jn+IIfFZXm~40cS=~
z1kRd!V?B8TS%@Ki^d%(@RaAQlAc+-M#XK;k-3d$f3X%~~y%hW3$QlCTBN_hNd+vG)
z>uj+Lw|R<`1Q-uurn52!+U&AjobiU0Uh7(lW}Z3xwnG&T2}9<Ff%`PD<2UIrl6Dpc
z+CV*I+`U);<a$00^s~Bi7%<K(L?QLaI48E{;Q<RagEw5cC2C1@^a36@$V7#hLAQ;B
z*`7^1;G(NV7lI(!;{Q@EyGk#0ZX$*Ff(HZgtEWQqw<N}mjHdqp;B=JdR5C{tPM#0%
zY6Ch)6pvbw7dp2lXHX__l&oj`TF5Rt50ov#*kAO)_rv`7#dz=&0hp2RKmRHB{dA0N
z{!aTKi*4uo5%?y8;y3>JXcYkWix<+<%2@(Bc8qMvD$xaT$w&uQ&n>OxDWUnCAmcl<
z2CUx#Q<zpsHyM(P@yKd+AR0A*wsSzxkg9wtqG~3?=buyR+e#@Uc(goc<;zwnXGx+L
zcjs&%<(PLyxgFPF2XJ^bCR8qx^;sik$Aca<m)FOPPgwfk&LxzKWPh-a%o!}B0tWLG
zg2cUwFwF#Q)T0X=jHWqkF8IjG2pRzmp2x$px=a#!jipF3L=rST3%P`pyxm@=-F^h6
zkXolr4L6w|TPMf2<twpMb%~<c`4*je{b5KW0vu1m2SMXc2XG3)#lk^Na-P136C#ni
zP+s!m<3seCh1lP5;koOvx^@sG3B|AEwj|;ZF<VX`b#yYcPaUJ8w8B>f=fpW%A$>|Q
zVnqp!t}<C+UOn;jG#4LzD-L|eDKdMX(OQmc^R&g4x8z<)a?ETNWjEaKl5ll<b<M0m
zdcZC#Ch_^r37Bj*!!;t`3ZvjV9=s!vGOu7l8_qj1j%=**5$Xj>N$9V>7iJ!dzl7XL
zyl&$`I<Nu2=P4ri<(vqmXV%49v!Z2Fj0QP5S&oKzp$<D@ae)>$#h}mF9uOvQ54?2;
zz&a7mn5csRXhsTuISyW|?FJ5o)V?e&y^&Hw;KErw14oazHUt&rA+wle-gfLE-C>~H
zwY@ui&w4cCB@kxs%*KPlLlMEGO#-*p8fN=zEtTyFuT`So87(7bxjln5o>=RSZene7
zsKXa%!ikNpa8y7&1Q_@nn%xZ5c+)bBtf?4gou*cyfZ;h9GlisG!&c*kOK66i2Wti>
zq*Jr+KwT*XAzGe~KZiuvIbjoJ+9&Bg=AmO|WxWo|aJ2q8vOq_YYC-}pvyrQ!y&E&~
zRq}Mn#N?>zSsfgB#$Ut1VJ0b8Mr>LFQshlWQfXaw{rMt&xbR+Bk5?Fvq|K9q{Mtuy
zNrA>3McunqFuwZ%e80>Spyak6s?aVd!EY#+f%{=X@swTcH1KNj4tykwZ|0~F^M~xA
zI8Zt`5s_)+74F>@jt?|m2ZVRcDSVow^}qA6^deI!bnjQ0Pl7{q#g)-SwleAMS=)6n
ziDg-fGI2mXTp}v13`|(dTVkR8<hak)(A#%SC*?q=CNrgNVl_cZJlGceW<wiGV1Min
zfa9Pn=|lmg@<oWj{cV${n1&8mNRwvVcP$owB<jY=eq<#b1D_9RfyVBiJAf$C+DWx-
z6RaRmi`oYt)f2<FVEPW%*cTKt8oXF<Y(4EZdSjc~)sTmz${}N{f5Z!Rxt4CfcZ$}l
zqWX23cO<ENc6+K!Kg8F+<-wPKBB8)WfDJhN!%7O{mY!y&ulie_0-ne5r!afzp&=K!
zLh6X-SuxxK^9U?zBG-`**KOo85RK-JwA?I##NvO&zgyF!KV4=u4yr%~x{}vH)5b@-
zw9uMF)sbk`(xldNhY~zGL-_bl(uLcJQJ_l91829~y9Y;7R1E5jV)8DbIukKRfZ2ws
zyy5)cfV*eEA>=H>_e&7^N`6^lFAu2{Qfd256~2f!XqQZb9+tffiU?skspoqesoUVA
zGF4y20PdjLqP&H~IFZi;nAkHYy`?_UG&ZEBY;}H9^xxF)ciGrDkB>-72$mJRr6}d?
z9z-ERLDHsLxd#lB#Y<mS*E^@zz@%Nc8VcB+lios*1UO{NMbFDv9T`z^4uj7-&XyTF
zCdYGbe!q>2JUGNiejn3f23~7*Rx!n8{h^H|RaVa%o;X8JeMIaQ(&S)2>8{tO_Szf1
z+r2&MZ+&IWlCEAXfE(ARcHushfUNZT!4~t$_=LQx3YWbaPRTBbO4|=Tc{@i2<TNpJ
z{ueZ_F+$J(2h5z@$?+bTjRiCj@X4g87~ou-JpWsEFVS90+E_;QU(x(~Z_;1`V{G_o
zk&eY@&O>WnnWL5xA-jtN3lD+AKqTKIScV-45>{Vw2<v&)`FytEkAHpK=FH`2*!KSQ
z{XI0nCDL-oPaRDLbp81&%|;#7AirN4hcA)+Ve&9@760TwY<AspA#dp6pP4O#H(fg8
z8PFgAOnp+;nLJV{>}AgG=-jvx<=Vz0y0V{mFmH`;T6ng)U$AGs;t`t|Xer$hDOa{$
zNXxx4n;%JDB7CH#4=bw3_IRoL5ORDr7~tFXEkqBmzg%5}JXG=ASl>+Rog=cBW!!!a
zxd_9YiMj6L(X|e+%5H3O7o`W!Q#-A{R3uXarr;2+E-QRUnp#eCvl)Vf!f&LGbhvta
z(8#trdQ`-NM;mD$fL&~E&!G-nkxv=RhmD;PT;D@P>%XEn68UsM3c<G^x}O$$xzm%C
z-FvOCwCp@Aww)xmaEV$f%B!EM`I8%3{jrx)J5CygudTk+LRRxH*m@{Xtr*l@&{=~4
z(qh2thkvc3(uMaEMkXQ2Zd28sTG?847Dah(Wj`{;#|Avgh%jNH%*8F6lo48!t)VfM
zu^FWui<sdssc|>O(lS7`s%TOeK&=%wuJ#mt=OGik@BO;cs{Yfh`l0@yHUWEDNGviF
z6W6l0K?}u@t$W5w79}en8|fiKm|J59jH?mWusl-)z@!zL=2;(sAEae%e|>KG(x&o?
zY(3iY1_mEhx76eX@-c5^74^mBG6v28Zd-*A?FNZuVL)m^74k9v12mJ_F1cnPi=Y8@
zYz#>ODpvKv0`*#64Irdq;+jDKg72E{<KD_A6<bfa;<BUvytD_X!zZtOBF_hGfMJ)^
zyrHhhLc3wu%T+GcE6As46Fy}g_>zg>;2F#RvrNDXKPj$+B(jH3*bT-UQ|;``dns~^
z^?S~xR6s}H7^+Tv0_O@$DGU)wpf1KIXK(vmb#QP^2VK)iT)*_y{nUpxTD$1B-U2?L
zT3oA8Myj;{FSlqB7vt}#fDAOS$rDgcG!7?&JA@pZkw<M&azL~k*Kd~-XE!;-ag5n%
zTANcWbM_R1-_Fk0mDG91$=FUs*9H6lv+cs$=5(V-Z0@dA;o%h{z=gQU%MFoyE%@3X
z)mk$xiZi}U^D*p!#oh)u?dc7mfWP*gusY%|XUH)7TkiAqV8;$%lY0P%W_j1=+V!ZI
zS0jJmk|w1c2s8BTG2nh6mHPsVp^Qh>s#)PqG-y|;_r4c72T?cIqm_6(Cif%=(^bwu
z{~C>8EF5XrEyV9H0&HMq@J3o<n1iTivPCY<VkK}ZL{bv#!phz$Y#9_#nKvy^9Dt~n
z^;ECiWHtf&88x9a37SA396G8_8vHQ@vrvu-ZF0D#+<x8<_gbE)W@w|8fpd<fu$EiM
zeZEB*l0{vq?7z|+tIE$J^+z*v`^A?d{V?gkYSM9d0--SpRI51+WKULFlzR$FDi}Jq
z!NP-8nH;r}uNp=u7hyHQ6bUYIP%PREmeUVXr)y?MVNo(619xEHq9D9K4E}6Xx$-&G
zFd6o5TscQKIkqMeD#<u}N_+;~P!?irI+$P7rEvCro&#Ik4}aH)PyWgmUGagM?V^%s
zUQd;f^BCleiuER#7URQg*)y(6P>#ZQc<a*bNcU<Hz0)WWguqoY%RvWoNsywY3S3aK
z;FQ-JP+|!={lWnD#%S&Ti$r(Cpu&#i^Ac#U&)%Y`G0Qu6>z~K5-1Z3#0`sddzUmY<
z(7OCnU@I1hRTtREYXn(rQeP3BbS*{l3*A$|!E4u7w?3FeAvcosr^DlQ@~+mBICtoZ
zJ@Dp<b1B8?bckJq6L+JoO4k_-sB?N#b$rU}j4}1Re<lvBUl{CmvZ8kjZdE)KTZ@&K
z_r#jHW#<Zq3<oa<L{gbL@pD8FXPgBFY5s07f*@X^JLL&Km%e&OuKr~K`+I9q)wza$
zFwRm5`L_sD>Yi1wjaa5j)Nc3%7*fES?QmlBOL`MK<AwxxPxY*iXc0X+8-C60BQ!Hi
z=e?dr=@&Y1PY9<NRS4F-USS<5H2Yf1wBEq2gFA{>A|jOPZ@5+m!)e7HjQEmiL@oyM
zDfc%BJN2N%a2z)B!fX#>1@b9okD^6rmR#qR1j*75fXvn>>SX$eaHAUnC`h1c(DSL4
zxe_zlJBDRY(^Y|J{=HH#&;GXax7gFPQaEdS;sg&!9)?D)d_W8ly7O7W@zXDBO}Sif
z3qVM&LqK-MpUt=tD6S737$gi{u^1`ZRd5eMR&2;2z|e;dxLd1DKTjbll#^mj&Q9i-
z5w1sl+?2NOdb{SW3OyL(Ah7H|qii(6Zq1Cpkfq*?E6ueq?aEVGq-InF<Zt!F@>&aI
z#Ki(rAZ&1}nL&~6rw<lc|1u)>>aIv_MCYzP^ai`|lwh3OY6lU%!|s=xdB5MG7`PpQ
zRH7UJ#Jzthzq;YAA02{;3T9$Y%_?&ITWC382pd5*^Ep|jTPs@3IdUgdmJ)lU#ie4#
z3!!m*Or#-M7EFUCP1UWr>)~-dhL(wU+dT%dzN=Nnk_urZ;b^&_obo6=Ni$o^lqd<b
z`O|=`IrC2_&7n92PD}p9%PLIv5TwVPKBCr*X-EoFAq4O8w1xJuR48SIN#gMhoz1S`
zk!f|zrq63W^Esh_A-A5;lN=+*VQp`J6s;l~|Dfzu%l~5^!&~vG-4lm51O}1DO05fA
zP97+5U5;~R233%QV@2DLFN=s*iLVV}X5Z53@@-mE^OG5@vf^hT<?Hblk}40SCKdRX
z?ie|fhGTkfhS0tb``?${K^5t*RI~(J{LM&S=s9`3>Y)63kU@~&*lA_<yNM3@`K;h`
z+$o#t^IzKLGrU`4#C`@@k1eO37;j^sOOhG&aq*#($@EfgzT#=il!z;Fh2VBM#TYhS
zQ%D*XC=CY#n397r#ea7Y@e@k{VUntJ9kY6+Um=%?b`{?z?~f?C-%9Mt*9;sEFe5XL
zKW?KF7DEvZBbi>14kvOrbB8I->y1hQTP@l*9+k|5l51&qycgEYAw&%W-LnXIq6s#H
zLR2^*pFIdbjT5Z7^U30l%#Q)@g~WyYj*0o)m`zMBk%eBCKhqc+1}n)W93iR$OEN`V
z`no4S3h!Wt_ATTyg*m@=V7+gXIF2E?ODV6A>lI1ys5p4f5Hl~Ia;H_o3{Ng~8Xr;R
z5vU%fAdW4srv3BJk?|dE>cS6Llj?^Om4ht9J|eH$_rUPnA~W(}9u(`hPf9i3)w_q4
zd?IS=GK9gWa830-EBQ0fB^AZR+2bf!Tz~CJ+NQ}53)A?67)rVQ*h0KuAsTOAO$FNk
zp?Hf6CTjOC8QB_1`Vy4QWJu~jRxJhj5By(*i}0eh?llf<Vc~%uC7d6S_)*`5Pe<a|
z<K%_hW@F+Q>(altlm1*OaTD9R`tJ`Bh_&=LMI(R9=ie+_&#A3s5dKX7Z{5=-CKw5I
z==m}7^+_ro1k-kX|4+%%hdx#k?A%RIU<|@P6UFe?qgIsw$bjcxGK5?Y7I1OFULWtO
zJ31^WW)=F2xVxqWM7Q}s@Vx@HV5_FhPoAV<T+mtO_%~(_n+p85d($(rm_eCk!dkIF
zpSKzsG3=#(cjW>8dMIY0ry_)^tPYclN&_=k`Q`)|46VOUod;#%w{{sX{wU$WtVc@r
zqA;2n?DW1e$4(+l|9?2l!$ATd`7cRM_Od}}3=n#R1LaBv1Y<Nt$oRv7vNl4h`@yF(
zz(K-;vvYC%ual>A<sHen(7G=)Z`*tewHT7b!9)TnLe}(2;2|=3>{tO<4O@5Y5@`hf
z6v?0O{!1nsX@s%{?Xh{$Yv;8VLP_dRl~GNe+uR5GfJv*O!i-s}!CD+evtRdC-H{5q
zEQyWA=#-+B8DdszxRLC43UbC*4EQdYF)wzbw!MNl`#?0OLgL>()VOF}yWKeBumf8!
zuywv}(#juC>#cHK&=^Q4NT+0{uEV)BsJ%{_j|zR++K)V-d_4#vgA}JcM(lD!Fj-m%
zP(ym-|G=RmYN=xcVVJ<K#Rye`?a&@$*LJH_gX4l-w=CrgLhC?Hj0PfREn#3;+wcM8
z=<#<D(m<5q!?1nQ9Dt=l|E|O1D};CxuZ(kB-zJEWnRrMZ8GTj^b>V5QkqO!rHnVvy
zqKGG@X#ikfjQ1=f=FnZ8H5)>&Em<xqHztE3rlS8Kg)MNumV4P-tcoxib>8`)aehhw
zWURK+io!@iU^`i>Lz|Oq%gvI!{j}JmVWB<b8)!EE-ochBaNI_b3_c5{6vwMA?t}_}
zKwyyc3nwI^M98^w8TT2?W-wsJ!RC#H#LjeznAM+a9<+ov-txPvbZcK&DH<Q2Elj$N
zsGk)F;Do=Y?PG-6l)|^EKayRrb(2}4m}WXr@*7tp^5|2PBi7=em&405Rq`59!OB(;
z183=tMmpQ()1AXSVJl<l%el|cH3-W{`oeEeC~v3?{mn~nCLf6>US!$)O$Ey#c^`^1
zd+fO^k1y&-0hv8N2c;|v#1@)UX75`r_10L~9Cm{hhi31ljRp>y*i#*i|E`yNvnM>B
zukIV)3Rqze()MSlPEEo*jdhxmKP>9CfJIfvxkDG0OZ&|N*_Zwu$=Px{_A^{tH#ZwY
zWn+&UN9PUN?)grZ3qXY*jxQvk*{96lj@1n0S<38AZ<iKfzVhEke;!cn_8{6FCqr;w
zH^e)wFOOU}J{t9XPUd~sHkvNveOa=9^r_8IjJCZ&qFeiR?W%c`K?Ao8|NVSE0@*HM
zp2;{xDX}-+J};$t7|_&iGUOh&i+YP)D2AY{zVkM9X%x9?TW!3Rj&dt2pbHFc7N3-=
zg%-tC&w+i&132f%-1>A+5xfRF6~zkw9wV0S_w_@p!{>XvZA)*mXR;&co|LjPiL4ui
z^s`HB9*}C#W@{0fYu2w-?W_1-fG*xr^a)l=#(xg0y{hHR=x3uS`l-CbE4_U%@e8!7
z!NE88$NQnYUh178B{kZqkL|$u1*K+XKx+Ce`h^w2`6Bl9>^G{0TneAPw+yP&E}nh(
zphu5PxC4J-9R{sZYelpbQ{d+eVCu>c5EGUC_WtFD#sY-t1@qM{cq1W-1{R!I2d>C>
zY5{ZXSLC5@SFY^ptxu)`=_k(32MHVgY2g0A3#T<O{)YB3V+q7YX(SNm)YYq9pQw@$
zT~>e4-S`x2%L@Y$xM+j(05!z!qoRD^&KRYvGCp9Wp~`OP!bm~G={)%3EotP--aKL<
zeXw>*#tL_biVG>28sjyR1k5VB)X~(ftRYUaORy-%54NwsmU>}t*ZN~{5T@77DJgb9
zLe%_Sz{y2K=F)FtQl2o1T@`tFTF~dTEFOzyq$Sk_rbQ)O0dtZ088qtFEZ~T())1)7
zc{NqA&B9mJt{G3uYef4Nma%v|$OeA&mw1y(DXa8h<S@Z_10(+B1c+)9YWa8)X(!lE
zc_lvm@`?Vh$zx6^i2AlLzM^wJ_yS1M8MpWnj#LCp43T~NZ=I<@W7}ddE)@=)pPIz>
zn7Rj6@b;)3CDk$ZhJ(vdz%UKefXf+GW_+C@0eD(;@M+u06~d%VO9vqMQ6rV&qZ8n;
z%}k5I!H+!NBL8cTG`J8kz;?kd_I^H~q`6EMSMiI6H()6R2-XT`0I&uJ#_Leu(E4+o
z6k7g)!)|qL*&Y8^&;?>@&5;5Z&~(FwW*he<BQR4YO=chE>4W^!q-)DNaS%10&~SHi
zZd@mdFhjA0Rp?~z`75`QeT%5v4TojJ4H>CGc~qu~3CVJS14elaa#zYng2mgGBLdr-
zmWzD*0;z-#2RbP^deJh?$0;^b|K)Ph=l&HhuE1W)h4CU*D@M3AnCiMsEKQtqrV&~C
z^o^<`=3Wv{c+III?WyCd{HuK%;dGf<r_&Ec21@%ENRSJ<<ql>Flf5<;!)cGFuOjhE
zXBQ`lZL|>Jrtz<96(eO+lk;LG1)aPeSw8I~OzdIup>stZvkm)w?o-ukBQin4`d2uW
zOU?p1CJJ^?mao$`FcW;*j`*lkWCt)2(hDk^V4M0FY=mbKvG^(|RPbGusPSv>%{FRi
zN65NVFv3)Ln}WRWpDyS4`{#^1+Hbzp`|*P)O`dk(eyXobm0B8o>QNTT@m50HjvIr6
zQWSc87icXtY5iEonxcjPdYAbjc;0F2(Y~p#>%wFqL&JC5<}@?(;~EUJv?9S{fzfH=
z{Al_4G=>JTgyA`o(i$Y>UZA?a4rjFD6x?Bk=G0Q9b{+Iqe}NuK)&e7wR)jlCTpYcM
zh}%4{<ofcOJ5DP1If&!4+nVDSZVzCE8s6^dbh5TODv+|1A1tFMM`!L4a|wG_SyLxZ
zo0MT@Ri>A@m>LA`jEP~8>j8BShXsZORXNT<mKe*zlEd&fOmj$ju@Y`d8mb9VN--Kr
z#Caki46!)M=y^Nd4f!yQXsOPNO+2zpN>K#R=ibm^)Xn(8@Yu%3;6(i25Cw8%G_7EC
z+bfH6s!k?D51JZf<(6{xHN{*lq1yyhmsMD^NhU+pN9@?SawmlNz9?p!<_?y5sAf!9
zw|E*V;W<mgT%t0SJN<3{8kLto9d#@_K+OR0sE^4H9<5shj-Ym9Ci}`pSZ=$~-KP&&
zHxkMzyfgCe4}Ds6o-Q~Kgjf+?Kp)YQoUmZuA1<I>T2*xX&(`F0>%*GOC<M7D!8&-9
zQ?C!j4P@Z9)fb6CG}ZHwk;y2us$9fTm+aQC@1sd_3i-e*D=9b3ZaT!zbLh>MYY6pH
z*Y(G?Z@^<vbbvPp0{z42U<2RWpvw?&!2MD$KDYj6Lu*4#x5B>G9(VphqAcR@k)5F{
zl8EE{B?`ssSk$~s^TqMYLOJQ0AwE0<j7_7!?=)fVg3ka=EYmH5!As8g9WB9ly07T3
zgqC%ncMxh<DG=Z+bgXkOp>Z~iqiH4`Z}Q3vlIfuBURIVEcC2$9>C@1dkSc0z2{NN8
z?<=6H)|AD_l)h4@bauw_RFa3v;!QD}i3{t}L@WNK338Dz6w+W{1YVZM**JvRfWHR)
zD3k<lANT(Xkc|`Ye+3A~&kx5eXYOF>W<|mVVCDW_?WJnV`)>R<jC7*;`WO1$n;|Tp
z%-gP8I@>=>p3~!{VGEpL0F4WyBG^pt=fnR5Jf(cr+J_8)gAmw1$;W2o)!YeJ7vO_=
z{xb5RS+`^Hrk+xp?1pZ*C1h}Pr%r@*0hp@2uB!ThxZ~|K5LeEDY%O+e=~gH*YBQkp
z#`1Hs0P-#B(sy^H-hB96bo2=7Q9A=TEy_Ee-+NzvyuLVT>x!)OEpY454Qlqi2<=_8
zj~zfPk*!OEaH>53gU`H7!Twv<p(g7xc&WHCO++Jk|0e!6=B;WR2j-eQ?+&?-T&U+j
zFW-gxkS_B+?Y1Hy^_nC6=`S3tCM&mVEMl821r}DaZ;h7LH{%0Jjc?^t8P6!kLW~_t
zoc|8?9Jp5*&^$ZY@$RR>rYd<8zY;0Ji90&}K7l6;p%ve~?AMbtN{<xlbBMeT8{(K@
zb+v0<+{yMhN%y+UVW;clNWEH-J$VE5onn>xQrV!k_Kd$D8mMAG{FXE+7s?RQKrWhJ
z27Y5noE7*b(?uuKc?)A2X9wtnaoO|tkCqYZNoq3j%iHUyNW~0wm-lB@QjN83mgw>j
zj2L(PlQ)W!T*00DedQSH<@VIymc<c%-Ba9o@KwL$pSE3sdb|z8)gNJat$|$Hy6*0A
zVyC!FlNEdKo|f9d;$D*Poha|oGdis61H8sw`{OsmflnJglx3yi8CWQ`3g%o@{xCRZ
z#z_^^g`3OeQ%uoAAQfn~I{8Le<J_r@)Ago)>Bq6rbN+<Rc)#}*(YuqN#__<-Y_+Va
zuU*1~3zebpqPP(4s(ET&sw{CgP-yTJvR1M9z*i!O^k3&?^U`UG<pv>E_hX@hHn7n;
z#U!uB<~1XOXs!Tr898>^F_=d2=`VII+^JaUDRZvGnvIuJhZ3sS^|Kf*Z>G}3QoqKe
zQCQQ}`jGz1!5+=Y5hTw2ycIfs$2Cb)erJ8_M{UKQj{~i%OA<NHV)RX~>#zcN&jB4|
znc=CeexmI}flrsJ5Jols_H##LEZ{s7D$hXG>A3Q1!Ejl=0jgL0f77K-_M-w=NaP0K
zIexk~yY6(-CMe;K1{jE=bRpa^NfWfIN)W~J53wDXudi4x9`p?Epgjdrp&;b9XRBJI
zNeZVE6vP}UJ2yUDcx48bXF0sCj^WnV+^cV6ODDa0;B9hjI?kPcLUZ|2m4RT{ErX+^
za97ubt=AJ6*ut4qFZ1c0iZaM>B<lJTqXzSY;?|~xDe_srXemx3OCsewMKv=aB*OX&
zT1|I8nq#=_%UGb%Q;21sJkC^@C;8)GI!3nQ*kuKta=D?X7Gq~IaM_!dn%v16ODmiK
zXawyG4Zd?u=0M^-mRKaaDWJ1atL2`r{*AMp8|Ld`D;Pw3LlyC`>hA&?$x|kaYWsuL
zd@DVsfu-Xhj8Dv-s!dM{M!I0b6;E56REY3jXe8HJJ|7IBVJRWxs$N*qrLSP$uqW_v
zCS>}TnW64IkO=&MtTJh|%>#h%2W!EOA+~cLAH&4%!P6Sl1X)|N4e(ZB7_$~W+?EXI
z4MEJ5?}Fb59k<iJ92%7}0D{b<R0*Qz0k+3li8VzfjFge&qA&>k`Twx>j_rYVZ4+f|
z+qOG)(y?vZwy$(-Tivlbwr$%sIyUE?dB4r^{EL0As#<%8VT&rW>%zL6Ec+&x+<Nab
z1iBsiUWscp83zaTb^VqSCAVZ%m0f(MS5&Qeaxp<>8Wjy{)x+D|^;7~h*#O2arIIp6
zNmkqe-8v3H_X$3`Q2kqAyk*TSw<B=`#4)-=T!-r<A(#C(Us<O?xH768CBQq_jS|OD
zqS>)yNAC^WZj#yycdor3*hH^+H64e-)QLPoTny^8=x}RbOO_)Syy|%8BDNM4i=0V<
z+xz2}*adESxk3AD-rGD;BK=i^u1rno?Mi;_u1W-ep<KBnwc>-HRnFH53at1^rlKl(
z5FYG4anzrNUT25C*)P9#zmY6pA!uj%YI!$_FTMVDE7b@0{zN1VZ_&0-KcyQO6XQZ9
zcUQdtFHbI2;we$oAddYc$0TAs&L|cG4F(<s>v)mol5a5FZ69XI`ExS&SO-pU_=LM0
z3X~n#wyPQQSY^m?vVhrfoM!)lbdb&(j!1SZ5YBmwfyrcPQa{kzSw;b4CJF{~Dn5*_
zW-%<CEz73z(Jkns5$Qq+md%$TQK0P&Cz+!WG{d{XL#H;+fc`4k)*m`0z#AgLfnYdE
zQf*>yn91)2F1q34cmNY#miAh<3?ICXK|Bj2@cDT^&DA=jSknD^!kWv!0NoG!n8;?m
zdOX5A6$r23E?7^kUVAg4%7Yxb-$w;klcLBar-?0W7}s(_Maa^;9M<4I_t@YWQlfip
z`%}cClOIOrk3YQBKZ6aFWDab*A2@uwEDm<_N#fDKwcpJA6+`kR5x&YJr&=FH*0lzv
zNMS<YVZe^E1u>M(uac>e`W2T9$MQ*udkq-8!A9!Xy>Os=tifIkc(I2lbCH)d29&Xq
zKNwMd*%zh!<)sR!|5t@)$|JD#QWQuqwAi&|ElC}Zn_~atJQyl@OK0V+K9vmB=$SDS
z*An5<HXsFWxICOUm2NHa{ja#+4{aU*i+~t}ev0g&JaO)kY+t~sPieUK0G7E;QPPAl
z4f{E!F4~vrKWNY$<BvsrVH*N>8JC5aBsG<ru|n%vwa^3wRlp1yOk?S)4egv}%U+X<
zW__MhI&^0j7BzrpG8>s?iI$Kcx?Y~(U~uR2+nR2cDkL%72TodSmmv+Wd1Vu@(WU36
z%8_iADrydj&7xEZzJ-0h?<BLj8#sw<9|zEelvQOy#d^KZiVqk_$v3Gu!nbv1=Lu>|
z2Sj#iESel)x*MxJ@uwGMoT1l!9Xr)&6oYH$Dq{WA+k)oflhNl{qpzN8lw}unWHu;X
zAP)AN;(e4pA+J{9f4ZPL<=O!txy`ozDC8yDC<mp}`(!6Fe%r}VI(#b%??65r`CMq)
zFKMFr3G<=SKg|a0{O@in?0>Y9CJ^0YjJ&yjPN#68-$sBD*|-zD-QDn4oUeaDdK#B<
zlWLPs3nKH2o$;s_Vc~Ob7|S@9+}CiOF1_jdS(bo)-@w1q^ut{Bo6`V2S)~n{_Y@P$
zuus&G+>yyG;l<cb{MzB@*K?;E5fb#xpX^M_(yT%ZsTO+Kns!T(7(Uiz#l;Ck(%D2B
z<}ZmvSX+Z|p|(DRE@=KwX1-6&V9Mc{%tdU5QsXs5j@aX+-RN?6wQ0$)T5(`Q7?3U;
zUj5Yrbsiyef`;{9xuF2@5jW!cO7g%IshXKUq7{L@xb4#%eFfda+ewIyuUU<4TiGi!
zb`G`l3jX;haeC~RbKAk<_a7mBEq9(W4?bQBUFk_{{qCPph_M%rI04&Zs@)uJ-~Rek
z+&cQN(9W|C@{T<R>62X_!&1iy@(|V8w}cigaQ&BOgr7AAjKV<nJBl(*e_<={qtCQ;
zOF-CHhg6UZQImle3t<k&2Yc4cZ(Wt~cPG@<BUi)cn>v1mL$^U+71q+WSQYiLH$2EF
zP1Yt_-Hr}l>QEZy)Aro^$`yGwmnd3DG)nbHGBeq#HTH)$gA`DTc}PsO-6Qdw_W0&{
z1&5y9mGc0|xg<dK$kLXL`2r4++uNP|7K~Pb{vWJ~Ymgz_e<}!=vapHx54JEEf`_yw
z1Ut$_6Rr!1mZo`A`b7SicENoAPY=}`eC?3Hjxn})SwNR{&`#eh@g%~rp9Lcm?PI}j
zoypuc_D}!Hg_%_;(5`UFlN9RSyhYQ*MV;=mRUVGF-wMRQ6k%3$#C{Z!_m@4jN2vyS
zqRq_NK9w?UQ_yf7g=1$aYo{!}HK4sK7%tcZi1!b+J)#|jH}p#ej0Rj+qEG>aF(o??
zZ_}gr0&5WR1UTrSUwo}a9!C9nUj_`k^&j(6BwkNwr-m?Ay*=ad-!x^#FT}5!LhKGg
z+q9acC;+2=fVRTRwd?{(Y}cHp+I=Qaa}^@{fr_}X*x~c!HTvzR^eIwqaf?#gax`wE
zv!S6{H{m#46$1HvFB~zQzT7HQK{34_t2wBaIvfA%@xx%BLMPliDln;LLgiWIKdH@T
z&zq11WLw?;I1ISBNRod9p@VV$ANZ}(l26XzM(cXlTr?&Ri#zQnfq<0U;LCfw8v3#2
zy1o|NuB1sT2O8ZgAkY(FU=c-E5pEDq{uGjON}f9I_o^~>Tr15f)Fs{j{P~&PLfl}J
zT;iO1V&nSVN?XZ2Q#r*%dmwg_{qA#reX8o)pl*&^E_SBNUdi9y@lL_&_@AzY%LaQ*
z=dI*AJPx9#Eg9Ewew1_O&Jx?!I!AAzmFX|fYS%@sz4iF1_|qYaBmeft$=7d3A*>^n
zE(UeytfSdHj`!@CHbD`hbgK|lZ%pj^^8puK?C$s?V|(ZYsUEJP(fhRJExQJWcvpS9
z=JGg>?#_Md@=K;SjxlsR#6o~7+l?*3y!`fF(UIk)dZ0|d(X+mfslDOd+KK#{EXnh6
ztZ6zv7*|?ATjmNHezUusQ?Z2Jn+FoF_;yNz9<!^(m(80Vf9Yz4V(Jp*0ef#DmX6#O
zFGCl>HAwu;8pGUcbSzGiR|0n_?_ZK0eh5<zlD4AnWlH(EO=*$`*&cx25qX(1XX0k@
z%rYnOr7;{M`OeP_I)jrv%hpN6zA+O`xRtjIZ@b5&qCz-T{)3!{JeJpk`=ieG4;S)9
ztf2sC^AJdGKfXT~;`=Oq6>+ioI&VZA_mU~ss%tu46qUl%>3Z<!kHpO#pEHm6yQ8v1
ze=Gy$0)MXD&vP5ucU@rSxXJ!sv*vpE4MAz&yuU5s7BQjxOwy@Y_svvT^`%@|N`Pb7
z+&|+=W6nBTXNJ>DiX8D)u0wuyP@&*5KI8hFl#*PgX^{hz@79uhHWz{Aood|8n@|j8
z`%fZ4l9L134QH!)A+d2X5{-}FP{vc~SH$;WXK~J#HY>jEj9q}}BJ$QXi(m`0rb5?n
zF2NQ^ll%dn_0(_*(MA&VXgC8IHAruQzSd~k!PKbB;(zHdZ$_!}`-!(_y^mHk$5afz
z?YZxX^OUy=3yEcW(&>(cH`<bZaK|Cbc=Q@$Wf}fvUsH$xJ5ANNjLre-NFF%2Kldrz
zYDGbWy2__}tvb*moz$s6WRH1beY>JLZbdmag|NqhK(f{Ql1R<WiddYP>ffTBHK@~7
zd)V^4(sJ~DM$w_0O3~utNNuY#Zof|vxrxq^rjP-#nZG>eWgUX@t8y-*AnM~`SCTy9
z*Wo<>L@T<esutQ!`-vA)b+m8EE!u7{ch+*93iIsjEfs(_*4MWhSYw^ieJQDXMrL!_
z_i6!(!!A;=eS-^|;-QY(k0eCt#PS%|aH$Cqoi6m-8<zWAYFse-MiOqA(+FPPw*}cp
zX}T-AIJZKsCVyY|W4iv7_;!Y!MKy8UU8Ee<Gv5_T<n4PGrUnMF%i2<TVHAV<N&m1#
zC#+QisuQ??UWtII%Lp&tO9z9=rT!@X73^$K3o7dT4*@CR_ifnVR#p$?&PBU5SzTcR
znGrG`(%3;|CACo&3{BVH%d~TLriq8_D}SQ=We-uN3;dvfRvRh<3UI1%_4%)x*9C2k
zcimd~TSVT}lBLdN=Bh$|W~)p~XM=NF*{xVoP)or3A>S4m2IiaaGOA=Qw|lr6jzZmp
zfX{*;3HgWkA8bSl2^R(`hErC^-EXE-idn^o+<zj?;f>9Ogb^s?L2prKyFKNG^?|O7
z9$*jFE>pg<*oQe;I_=N{JXnkyBtpGAGa)G<=uBn_B{^XlZsCP63km!%*!@t0mS23o
zYF`2D4M%+X(}lm*ofILwKm9W25+;PBLg<M=>+p|k55a{qGO8Q?UMI?0Mg$7$p~Qa9
zdiOl4mMP>7=vX!f(KC3E>=*>_PvL}dqLDx5^{05qKRYHTv+9ylyKSlCUh13&rstZD
zN77zWB|wWCn;C1MOtBM$q4gzs`eshUmIwg+3OOiNv%kx^Lu7sEva9E0Zq)>Q&yn+?
zmv)z0l9C6)?kxGUZuyV<;mV1RoAq&=f+YTutxE{HeJ&ORYxCe+VFYphtqDDHm0z2{
zz(B6NRDn69#bDG6f$HbAq%kozal@P^P{d^jPauP_706j{HqmX$#&tr4w8KV|1fv2R
zCZ};2?+NP^Q;&IfJvD~ZOTg>;a8*(zfzpNQK>9HH+G<uqB@OjnT6^&H1G)~cy=Mv>
zzSrfnbFNc`BeM*d>0!RJ%G{K3)7(_Hgx>@rVb@dDXHByE^`q0l%-yRTTYK7^Np54H
zBMh1%Gti7s3vZHqPG`MkRTX5@Ne{sKg^p~fb$*&%eEXkEY8&OJBKXv&ErJ~sf#?Xb
z2r7`fPdBp3f^A6@*fyE-AHd5^;vaswq|?)5LOU%}Jgg;Fh)LD<JXAxu8YW);npZOW
z$G!Vpv)BgTQ|^T!zn#oD`14xdL2D-z15R1o_IiCOk`6y-rt{pH$0+)8Zbrak;=EKU
z9j>s(%RI?YW7#q~v79G9c;t2Wr3!lNule+V9Lx{X9Ih8HRRaTcshb7?A@s;yf6Fjl
z2H90%R&)(qCF~lD)-+<}Mv1l=YMLX6({K?G2+4XVvC+a|j=v?_`cb<1N#^VT5NS!|
zJ(ejn%6!m8_<E^_ji-=2Jq@6(pyBrDiz-MT(@L6p;>an|=0N@;LIzRd<NkBA4ejw3
zc@dBW9@p$oFgYf(la1br(R8M?O&Fmoi%rL_YCj-`K3qswW~_e(>(5#FiXED|*6)UV
z90l6@jAIo!7+$50IxU9y${x2PG0RI|7R3YZWWF=${z<Zg8=VdrAqOBZgGPInX~&z4
znJ(ifN{%&V*cLh;{s}a4^x4Z?OjMJ-dJk>!&oZQ!_IRWU{koHvq4>^@@8+Ue&oV?$
z-8zjLue{z(*CI$YzJ=>yNZ7uzT>7Oxzss*PMmd{!++k3Ml*DQNDIls%xfYE#L9Pm9
zR+J-SSvpq;^GLI<<N<q>sAzp!lSh(D>`ZedO6Eb%R(C49;hnxB8p(XhL8VB1=u_}g
zy5fd9dwP{-(L`E!i!d1eYVU<!_#!;wgwlwMCjMo8zvlm|Mz(pcGJrceMrJ@G)1v)c
z?$l`|5oieX0r2d#TL0;+9>oBhcGw=LrO;*v>XRnP)+}b-B7hCb6lkq5Br_Oxx6zYY
z0gG8NyqWv}TOm|Kx5V;8&|j53c%JK$fgcRCzm<w(;to=lwGDWYFEtIZ-tYOgCqH;P
z=Xh7s4zlVbZ)0b4lnfp08d<pscw+p?&WAw5r+;r3S$EbZ=<pgkx)5esnCwoeASRDb
znR4RbVP#*w;sRy@D6J0z+AYY#6ZEFuR5LVZ)U<jLAP_GVI9t#@u<2ZUq^j%?496&0
zzWLgFiNblciL~wD8QSOub)o+Wksi6_eDB)4^=NtxZEO|0@XbJn7B^<pJLW@KFe2WK
zGnG@3+^G@sQy6c_K40&~&P^uzU^xztLdRYTF9xux1puZB$i&fI`oj#^clwkQ_tB4Z
zd);!YBl#xlZfN!&j8zpbgzSdC&t32{4Zbd%T_-E`D$a~0OvS{M1W20FX8D{-kwHf(
zZo=^jEIZenha!FEGEw;wrzJ|{i~4-&@O|$Gwz=wAgWlQ>&FVzd;(|h3dWdhgUsJ?A
zc!odW%xVAs(>FGr{{gO9Ik-9gKXCn@MJxHB6|sA!=7#aSqFJ+ms5yB-VP(=nl}E1F
zS8<KNDst=izm|f4#ro^x>&+hqUv_mK*bpfch+oJc2TL1t2A_AN%hx06$U^?pEtM1-
zMcT$1%r%u&+V%%qRMU4$vQ2N6DGrkZywY++0<rohS>f!FVd3uWQ-#c8Ehj)-GxF|h
zGwRd*s(-Q2)!j`^de7$r&sI&6S;LHL8gb3?uw9Y~{k~xwQe{1b6qjyV9<CG@Ut)DZ
zpXpo%*c>0a02*8Tcysx%i&N_4YvCgzBjI^83sU1gSqDOWZe+j)zoVP!%7ZEMX1wXi
zC5_)kSubhXi?iNB<S7h`RvLim0-<f<A->o>&lUgWfGj3vPknF#YIzE6_@uOU*)aNC
z-#h-TN)IiRG(LD;b8-kvhI2o2Qe8}Z{bdHNd2vk^5ObhaGTIlx+iZFrXRp8K(I4PO
zMfRt+U;3rUq9b;i@X(XOza#q1<5K3cFKSo!9=63rWYXKdl=jn23m51K<4Jzl!JBCj
zXU`mdoaL5%+D91lf4gI_bppdQ4RtQ4B(z=c>uxI+?G=Icy=fwT*<OL(Xt2#-IRAI1
zH-P8%+v#-t@e6T)>VsJ4b&khrCikKSNg7x~7Ebx1AKZ|Tkl*<09}JF6&Qnnp>W}pf
z@%<SsH7q`(zKqk*7!BxoN!C3auQoM|t|d+T7AZ5FM`w>TY#i-%jd#^z6Hv)NCwfng
z89E%E!tIP|Dnz$61T(Y?<sv<&Uz;FW2=~q)+Wx5j(X9Qj>B66qr=U4O5Xp@48>RQ(
zSURY{DGB$<nT=m95P$!2a-yjCLe=kl9l~2}JoJO8|JZzkmJ%>-JYXrf<q{%Dk5g*v
z^J%tA!=dp|&l~Qh-AE?qjiuw$+LTYT7_1%{=W=A7SgY>)v(jUF;G}K$HW8`EL=aXa
zoTr63JFuM}y(*{6n(Xj-i6w6km+Rvh*w-(XEcSDPUvucOC8!`c8ezO*cLWnxX~czG
zYZYns<+QyGCm(=>r!jCvmjsUP{Y^(OMLmBuUd+I_V6re2Oy%3!QyMEb<AG8H?dt>`
zsdp!XQF$RoWO^?($KQz#uw$d_abq{2CC%1ahu7Nbceuc0gOn+du1J;LrGuMwVKn3U
zziY`GUF=D5;euPeZzFag!GJ6X5bVG+d1HV1@%8_!djM!m82FeXhTF!=FEzCVcg*_P
zoo14(ZQ|+and=iF;q-kJSJaHACY*7vP~Ez`qSx>#^7{b)(S`^RZLOfy_moG}v_Cvr
z=#$AccRqZh+T0Xg-t?;B)104`7a_g^VKI&fiVKDvPi|Yl&+C)Y88;5`BI}0KL<4@y
zc*GjnW+0n?XNPf_oA&eCIqBy+FNAVhE%P<Ez~ys_Z^Z1N_+aU=RkjL4kLq{77&<<V
zP%`KqqkUi!g<K9C%B*3;uzhN};#9Vy8cRODbxc2l0cB-e1smDd?F4k$#}-~DEhWL1
z(3jPj&lt!oXGi*<vaA7jpI5T!$UlQByftmo0RSpge+Ml1ey)k&D%*h(#W3jJHn_xz
zT7dNvq`aUXS2~CRO;NCs<Ylv7rQ5$3iC^H94uM1m(~;@kawLnU=|#j~#3W9)U$awi
znm4P7Rx0jyh<+Un7^vqcA3!Zfuz8Z$S5fBU1;wW8a%{`00qZT94FXx}pY_MC*6Km-
zSD@D4kQif>kmr<yNuD-DZvGwpgo>7~%udF&3+(T5@9I1uzHwp<)boBZ6D=9+Et1k;
zsuD9ig!hdv|G}aPGNxGq<U|8_K0Ap~mF-4i1O~nD4>6h1_ezxJH`B{-Z7S&Nr8H4|
zth%wjBD@@RskS_5k3L(RX_m+Z&hPmG3ZUm%W?sDWPVv(v66xiau)v`>N!f-}u*(LB
zv2~U`9G=v7a@ZHN_C}Oq(f*L$5X|EC68n*;O#!(@A@i9Chn&!1$<JP3_@~nK_rIjt
zi+gmYe+{$s+Iaiav5B+JTjg35kp?3r^>)6kOa}xFZM_3#G!VEb;bIkB=~B|09{_GZ
zs`-l5QLMY>R7c0@d?j9c(s?n3PLIc`Yu#wE%A&#<(h_1lZ!A)G;-@mM_G|DV*j9&3
z=4tKneSN@)Rg{nlDtwiJ2xJb6&FW_W-VsLS<~S%A#WHVsY`V+gU-odE3VB<Fc=3X<
zAht}9PTJ!~_D7#G|HK1p$2aO%3?Srzi=?p!)Li^n9~j{070P_7ZfM(ZY^yE4pb(gZ
zkl9B94L<zc=uW!ln5-QwdKropPGx;Y%7}!4AuQ9lyPMiaPWcq!>Ouo$CSW)ezdqYY
z6UdAhE}NfAk4k#yiS|0wseKjd9Q{dWqe3B!ngMh2bH%Aj8)K!bNiIJi6!?rXC!u~B
zD8oJQEzeKU;$9g0($xuFb&5p!D;Ykze^_!*%Cm={77;#8nnp(i{=976CuU3E2>$a)
zx}l`fJcdGIPh7s3CBaqgzoCRl)tS`nsuaJB``CLf7JdSVzT3__8UKE-kE+lt+`*GW
z#LqI(Rx(`bgW){E({&7c762pxzHcSSJe9AB`~G#*-URCdl&$2Lq+`rp;kj^pTt+At
zTyyjN0F4v0G*=^#fVdDOIVc#z<1d^I2cydcTuZfWbq2SryA~^lk+zDQrX*s9R1%w!
zfz@D$fNm3_t4E)HQI~Ejns#w0L4{Q-EviKxMz6)gPH970`f-YD1`ynq0CHzS=NDV>
zEr`gd%1pi)kpZtB=+t_xtirBX+DyIsB=~8pY>B)=oz8$%1)}^H4)wo!aXyCD^*5aI
z`h8$jLHgWBTh!6GD`w(SaUHe747*8jq;Sl>TRYEQ2=k{4O-LH567u)N$ydu%;~GT1
zBR_lJ3j5DhgY@$YaX_ME8-|ib2IE>tAlHmby1d!aSY>T+ZhJ>c?$v)WISB#{o;m@4
z;*0~0i{b}+c}Dm?dk?L)S%4SwDFK%=B{45D30(gLpX3IjMn#R@#@p*(>7UiguS5AR
zdpw2vi*<+e{z7@v>zr7$`K~-qF?#F|;{jnEHlqw4n*3vSlYqbMU^zk)M^8usO*4yS
z%JkE`^6Y$esg6ppYBj|-nb`*h>mVr81&ZgOJp<1f8rfL1qsx1}(P^x3o|EX0x<bp)
z<nB@vmT=+B=iG)__M*asuakmZzl(f*S?_vT2Zx+tjh^`i^e9F5K@}qGBDP-av&R_N
zTqJD3$(HuN9c(J(fU+uf4^*l4LnYHO4NJsSyJM)Tw;JeOw_{mtc`n5AW7G0!W$(kR
z<K$y;QOG`aqIXtSg}zk<*1~g3;l&H17_sJ8Wjukj7rC@U1-Li*PhQhg^UfktL%OQE
z>z`?Ym7C9g59spnF`MCifW!_D;a1Z~#i$F1dsThYyHJ}Td{*}0I=$@g)#ipb4_AD`
z;Y&{G<{?4p&L0x`oYD>!{ywiaLXzBkcmD%@^8BAw;RpBsq0dw;-Q<mCq<|g6d`WMr
z<*0V$aouco`x)tVstz`PtM}g2R?d!g?vzIS=M`)cVcm6Jd)5QoWD=O5RzDC-=nfL3
z?w+XE?~RSh3$8YVkQlT?%f5g9B+0_V<`|aVP8$$Cp=%cOVoBlG4~@i}8)_j_vq&8W
z$MGBAbK49u4FFkRmvgOrg;WsCIv(u)cm<`Vju@K6k00(O|LNwC?u(cs8KP;JU1$81
z9--L1Rxzy0Jn$A1y<xI-x6X((sW>w8g?`f&(9PujsghDjp!+#Qffm~7#fw9QBxW?n
z%l&dAuwIH_rZ%SaCqqOc6a$ndLxV!(Mxy@oG26}E4idODuhqZFc5avw!>S<qUU1^J
zYnH}659mN8{Tt~N*k@-e@7F`ncOwvcU3LDj0JRGJ-dof+Z#38upIojk8^p)4lr0wK
zyKo(0zD|!KFn4GKt>@1kOLPm@ry;|l1erEoz(6mh3(3ipTuaoBlqxmB&MrX5V(nX_
zBuT+TnE=dW&gX_VjL0KBH%e?ZIfaqa);@F3S35Vy)5s88Z<k5O`j||6^N*b=&9&ts
zaA(nC#MZXql9&2V`naxswR6RBfexHXZq_hv650o|mr@T!Wmke{h$gnNzk-gvp=+gh
zKE#6RPF2{)<InCwCf(1TE7R*9a6GNwak8-KECabwDeQLAPr>9IepHoQ`KoUcYZe{t
zu)e(?q4-Jc$kd&62;o?zkhp|ryE))mxK&51PwQGH*24a_R~DURHrB7ZA%V|O8|6d-
z5o+uZS4!+#xbxkmw7y|f6^<F6c~CH{F|AKt(vQk5?Z-|a)_-u2qHf~UNEXIZZ^>a4
zP=HTU#hiuhm_c?0*MYE}gsn*7O<Ey%x{IO|vhm`F$uEd`n~W#7MzV@{cIj5hvru@B
z3<PG#X;Jj#YmjfTl;hWRB*AI8`mJ#+I0Dz4&Q&nEZnn>Ug#@C6UZ>IfSjFj|6_0P#
zqIeMgCPvP1WP<w+-Smfkv9Cri5f(=o%K+(ssvFc{!@u64@`#X0MI2-rMaumpBitlg
zcO}mfj$~3h@2fzs*!~tTYB#=pzOwsWC)1M%61(QKrf?QlzXQKNEJO_3Fw#9EWWn&S
zMWXZbWHeTL<F`i^y1k@A2<W%n%YOCQUAn^56h2T#$Ceuz25SeDwceGO!!AT4d%)C)
zwl+iuKh~b^O;LU0&ClyVL0)sGf6v_UK!9(9Gx=>$#I>b3w)<|=b{K=UM<isB&8gLg
zg{sJBAmM@+E$Qv-Gx7FKB(2UGBn85DrsBfo2XyjEJMmP=2tnjcvIiTstCEjIK|!l$
z4Itu-s+>dZ>~s@&A`mGy@Xq_x4{ZLY`@jHcUDZ7VVUjmxDaNTyFl&!)ZH;>>cH|LK
z?}4)!&Vo{vc4Za7UFA?@2zt(bCl}S&Ymir=<W97;Sn~YO>XbzVTRQfw-xbWzwgM;r
zMTzTjV4J}v)3kbFG0s!n@3Lr9(T<=H7Zp12d7U|hxyQI<DE-(nz5)a03W%WayF(J!
zy7bmdVpz&-UUL30F&s?{zUw!r#-m!f99m8L`K+9}sVY8`Rng>ZESvhf_Y0~9w2Rn)
zJeb1A1-)Uzaek14Pnd0yP=u8cZ?I`dJWs5AvQrHt0>R~Mq<h$75-km>c598la(N4W
z5;p!;%aMje!Iy8cL*QCz7ARLi*t5G~uqjyI<IoCii<jAo(_>)>KQzPKg(q*B0sp($
zXxH?ta<;zj$_KXjJQFo$W=V7_j?A{%E<l-~*Mw$!&Fw!&yaDkLw-!3K(&}{ED$Gse
zFU!`x5FKR<hjm4eeznDCv_tzI2-USaO%Yr8S0?=`#L4UEkoP|=7_bRjAyMJ){<EOh
z0fI2#97>W(^(Zd?;<h+XQ7ipTTvE)u&G^hqqh!Rgb-qvg#2}5n<Cf!dFbuQp5CtE%
zMzEUh*8!MubqCg7dMnK*9`>gbf5I6f4vGQk>^FJT&#<XaOFfc$?x>@6NZg`}3J+6H
z{`hqV$>H!UgdzDmBH-Fz&yBTop>0KjA=qxAnO!viQ%B7seZ?YI-{uZi?D<8rL!D@{
zr{AI4fcz;iy;+M1@9&~D42yoRYgz6$WP*z4@tU8wLFi6Q;V>~?gWvLg3J&_eawqb>
zo))K%cqd7+Ro8uYVaQ=pUa!y{>M|`b(9o?*;%X>j*3C=i0svifGyIl2ATsp6rtXGo
z_aKyzNvmFDu97>&+$*|jeiJ<>{X#gHW<cve_&}pv&=mT{6DRgWvg|+_m(nORPjFJU
zd1Cy2Alx(wQ}>x8`gyH<561`jZJ=Odh5FL)Zw|}Axeh*^a@voaJZI+CgvO6UmP<)-
zRX@)bhPQyL3!tG8b(u4UaYkryH~n`}V7OwIxHzd_$xF#ehkL8)x2Pc-NZPIKFu?V*
z2J2eS3OBtj{@xuv0;5uBde}Q=rK;R}|3BvoZMqWs*N89fimg{**}BjhdS>WnukUxx
z$eD?w42b1)WSYrLNLOijd5{BT$YKpnaMm(MS`fHE7Z4|QC%{!i_210lOc#56XeFZ3
z=0r~y%SUgI?*JdFR<tSj9Rnh%Y<Tx{E(QcI-Qji<kI5>hN#WlvuVPO}*P9pR%SNUg
zSYiq)Ol3#q&R0Y*)TmR~#a$kODt3g5L_vS47yhPRS1EE*BtleKlJ%J&$T0~Sq_VRb
zY}?Br6yRsBB(9!_F$(P?ik>MEYSil^U#>l~)duZSQ${g8_*$6Hb{H}m>P$hwUw|h-
zyAYSRbkwMuA^Du*Z2Ae?!4t#C<*tt#d&6?t*QZf3#B7*%)xnRePJ7%L?K;2oYlfY0
zIDKVhGq4FtWZEBl@EUXu=@W9{m5FuWasPez5pNyx_d8N%AU%XmnN%Ugk*WCIrau6F
zM8BDz{u2O0$N9nfKPsGTZ2wo`9MaK_-yB8_c&@=Y9MTgmT7zb$_O3;okTM}rW*87I
zfY^0uWoxt;{ee7oVfeeIy`6`mMYAIa@x=P=?^X4E<^EvD0MoYry|f+B<M%mc$*f+7
zPDYm`c4e^h?~Y6+$s~Qq8{@+4*{A2-^X5%!Cq$-Vcb@B{zUM^Y^NLCu2)biWHC<@p
z_oyD}E^pmJ_;_DXYvy8wA3io4%W3bM?utxq<wqoZ-fkvj^mB|%^_VOE@$NXWaMn)D
zC+>!BQyiTB-=nF0+hPBA-6is)*ZuDmU@Mf2O%X0Mu2JDtx>DA7$&$HAum0ynSpsB+
znMO}!YI9}MQ3ct!`aYI4U{Gg4okgcu<CetzL`c0aHNsXb^hQIOxVT=0^6-J}Rgt9o
zMYs17t-*sZmG7RrjJW;V@@GEK23yzVFtH%zcIEBHpGMyD)TCS2lG46hRYyHf>v0~J
znHNn%&nb6u#EN*;-d#QZ>lA&fJN!ZHrY(Jl4o&A$Br`}f=*({dfM4)R7Z;Z(oYk$R
z9&+YZw-0TeyOF|PbEvzSF^C+Mnh4!Zc_7G=s2l(7ZiN=fuvLXcWgtI`%5houEEso4
zQc#2K3fDBpM_7!btjs`J^HFo%s3c)s?YU>NML)wAEoHI4c0C!8?OBqjl&O1L7#|k3
zD7SC9V?R$Lzu5l*#H=MSouKdr)JxV-!<S|Is34v5xXjL=i{3H(7~Rp`%A<z5Eh0Ds
z=kY=kWs@BWD)0YpC!0tYd>l&`xavJ%@TVD|*wBNUT6^^Q6sC%OX+(=l&6f!l(sp?@
zVI+_RMNd>_c#xsgz#u&o!6?GHXDvnT?<8Ka7%m_lRt8N6n!V4?)yj;coSOAcR-1V>
z#p_oq{OL49+^D$DYYRxoQ|+o;TvHjQ(#Chu>w?aswu18Q+Gri@tC@;6V4JB~$K9Ia
zP<LA1+mV_5yBz6;saX|idH1;3@)O<`hQgW!2_}vQgD(PQ$RW(FHaJu`dlr533TdQW
z{ds&6L%=?O2+c&XcX_dBTt|}n2@&Y)cz46}Cp+-i?X6LMt)6J@KjIi8K7<a4AmD=Q
z+gZch&9>c5)&eH+&MD?YqVV1*45o>|B06jbCC{JYjw8I(QzM-Ba=g-QgBm*K#k%Nc
zL=Xsp9tZ~#wX&pLtf>+L1Ui{MK9v53_|84OxO{OS*y*225x$wqvzYeC;C-6z)GKSO
zyziepn7FAuDa$QaPcPZQi)RTUUm{cfO|GMV<;{x<h&2t3^7t$lQt68ty&jaH8V(8)
z?ia=ot3-sw!`o3T7+vJe$h9b4r5B0Qti~D^#lJ?a-S9477%Ra%pQe=DRSk(=K#f#Z
zHfUJ@;tI+j@E%QcJi9ja0e*!&BXwAC`XRk7of!itT**%Kj=Xs!n~(c!MW?eKcMY~o
zX@P6njnDH?4KYKoiRA;Ua_27-xuYL<`Tr3G6xwE8z4Jh}p7ZJSdM@Tj@tba3Y5m%N
zdg=DuyTu;U$Jh=f6#4_`pq|pyRHmXC>{4lf8NMYw<ly2s%qw?;&rO#*Cu9--bIKNq
z{4Ggp6g>3foJlgvok;vWaTZ$L&=yWryT=Yr15cZLf}1Pf8UhmepnpSo&5U{QN>iHD
z&Q=nhw*#L@j7}i-f|pGUsWs_ulus^}28$OM<9Sdtg=19X5^(~*+qk7New<1U&H8p=
zY+an(4!N|h4F%rXY;3XbqR$*n0FiQbag75z#Aj@G8XneFY;O(O=hGVPNOSY8O-OKD
z%#LIX+KN9{`76qIR37T1&o_i@N*GRv_S@`g#8)!J%^Nbk2eJX~_l*WSl0o^9vcaUX
zU&b_Njuqm$N;Ttt49!jS?XO6<IK4$+7bzS>byPSWk?n_jGC7}UeK&-G$DU+5{0dgg
zuUJUaqQ5>Ka8_KykSJYg1P`k?(7%;79g*D+qbYX&i#)RvAxWnDqhz8~>PP>s9U@U0
zhxns5;px@yZ>X+F<T{%Lt)0(5px{;R+U^oK`l>3he(@l7|E+Yvv*{j&N^e641c)M8
zLep;aw?<<@^BR%b;@HEjsi0OW<O3_ct-bl&F6YX_hq7$+nu?{|;Z%{vP*Ek|kaS#}
zO4hDzqB+WwR|y}pwJ0)nul(ZLk_dy@!#b*w{T+EW?y{izX_&FSRt5e_{g)=9&L9Za
zm%iz>U_n?hlA0#X_+GV|<a)yZZv+=swuZl*20I{sko*qV<C)R$_9_IQOcj8K`*FjG
z4qaPd*9n@_hD-?JML4l>yci>fbt@0?H?mt~3%jUS+xXlESG@LqKUXh$kVR59Dk(JQ
zRRmqayQ~i@WEdGk5mxqLKnq_jb)=TJ(+6jwxxNxBR9Y^(-)uTmx_zP+D2jZ|GO|Jr
z6V!3!XMT~k6{1`iUG+Ot)67r`i<Y{FwL*aUoZn@Q>oY<)cHhtp3#rTJ;Sx@c>#&&A
z-d$>hJ(~Xkb(EzCH6eFXDxsL(3C>L!=9M#ex7n9sjIJ3bz-EC8)!<YWkFT#A{ti`u
zpYz_o8<o(55`3sLNqS5S80`E_7@oP>*dKuoNuVfU%heR*)jQUaOtR0Xl6Z?#`c*_G
zV`qN^ZI~AC$52Pf))h5>L7s>}-=_QbO^HWTzfE|*v2XDmqtip*GjGB`j?a)WF7eTn
zg5#00;lS{VOPP>DEm10IDBHsr4Kz21hmLNLSA}|mUHJ$@=A?ZQSb!`}&m)6?2sP7`
zgxs!U+&Zjmmvb=>s&PU;9)+a*jOctx2>W^N1uxQtMXc<RMMN7`d}sV=I#q$d?8p1h
z0J9{3v9O*Rw^$>_=h2r%MqsE6Gn4r9B_U;@dQ~w18ZEm9Gdih2X$34fO+2N3>SVq<
zPu27HT~fB6itmymK$UW0<tR};CPEe8CWT{=_tArJy>wr_-4PfEb7b;H#T@p#BYS-u
zvJ@gCP{`P19~w@+@+Q9TlkDkaLM0jTuGuL?j2C-j)ALX%4bm}Wo~qg7b3^Dx$?%(<
zLr>(-M&?Y;E5h?Me@yFWEMiJ}X0ROLvrMnH)fRk-3)7AZ;2gEQ#%v7^xk8_&OW0nH
zKXrOH%X!e)G8%tgs564>1UdB&A4Gd7BXBhfl`B5%fW0dX6m`!p@Yn{rkcavYY0&$G
zUf6JJ#hG|t3^GYVh-|tx9s*XRh8wf~?IB$zwZhS>h^h+m+k^^rK+qHa?yh4)%~t*g
zTi_9O2Mo_2;1ee<61PfuBmjKNh>AgwopTE{4+F+VTV2&q14)3Cin&s%u^*%0)|d9Z
zuAHiupYA0iGmYH_^w4_Nh#B@#XqgJz%hPdu6Y2_+sm&Qu7I_8xNY<#y_Vv!W^$R;W
zBxuMNi}5)IZT`BB$d2rN!OgT=3ST=m+IN9B>|P=Oh#Ygh>vgD{!4~PseOHnfvsAKG
z^j<+gFCAexH*P%^x3~m*O`Fu1Ka~9WJ^`g85+r<nm;_hOnjC^7Bu1DFp>z;cq~5%X
z`Ha2a%)wE9_lG}4UbblslQf`){**Ur$NzCY>0F0=8FEQdK8!nKSd+2-Q*_{QJuMp4
z)&0NTjvuckV*)Q>$3lR1i<4(oCE75S{8avV34MI3_@`AbomTzx=C!TX>+T&F^%Ite
z&z|q#I$C*&#ys*|{zHhD_7cBIToESc(D3J5`EmYr;WZhN>2bYLSx=*=iw%wl7V?bQ
zIoKJ;n8Ei!@iu`vQ}nLj1(!Q|s%w9!Eq<E{0F=|hg#j08_^z!4d~FUS_)QS&QEsPp
z9lqSXs`e^7gB5(3CGuxaM)BOs7~HL@$;T%p47n5Xd$8bxP7AqoJ&Gh>^q6`IN9^p!
zDmuTAQd}-|bn7_joo2QvhQZTKQ*}ByiZpZnSyT)6_7P{pGlzgc!5r&&{h89M6OKLv
z7IO3F^DdTzI%-+cus|k%ySW#+*%*n-{hmZX;>u<L%_HrFHeB!!Z}J{jb37Gd8ct2P
zt{FMoAD|%#bAvBmecVw&s9wg+gj&q(Q|E~#<U?S-lP?gP$yhSR{i<DhoOX@vVnpFR
z`MKbkOv}Uzs2Ok>;7w$znxce{zZ9YY0kQ@^B)RUKQ4UWYVwsHHep<>^sYkLf7swsk
zB63ylZKsCslE~6v1De4Lxr9OG#_AfmE5(Wz8dMrM_$*$R*MVmC*8cFyoE`{8L!!iA
zt2aX%$IDk$xdXi7#P^i2nu&EZe#FF`4z^JI7>9ooMqO7GQ?t2D&G?_T+_-&!hE9t3
zBsYl#N8;B;UQ6aZLSK6&j?~>*pDGI>dQr4i!%D8mXu9JA8=f@gzb1`oXlL~K@Xd~K
z8ME`;YkpFM=jaSGAI>vFn3=723oKy@0x*BIUMdK|ma+cY;LCVBjtKra$fGRZW2<P-
z?75tU(Z}bs`NnwZQYULDt;5zK&FqEGYdSTd)TT{LpA~{@j0kBghRfH8?hg1I`v!q1
z3QVqxLC0lh<@z5t6#I|=7XZ@%8!rF-$mRe1pO3RT24%lrx8jL3+bn5oeawPMy+eJC
zf0m4*E{qH_32)=|O&}JEHc6CE;gTPg3+z<l`E0=+y7H4w%OAje+35jV-|m_eG*zP5
z5thye955@@2`I=|_RdAS-@NUG^usN3Q%`K$rg?q4=6;E^91eejSqtg|m0tg-4%@l&
zruCOG-G}ZaHwe!!5U~#Jzr}Gn{h!}3Yrda*wtMz(TTcEK#I_eagm*S@yX>qs4|zDK
zcYftxdEM!D6fY5bXLD(vw1*e0_q2Z<4G58LN#GgF4XT)_$EDKg^t?VvZVIbV8(g1<
zNz!jPAEbGdXJ7ShMRGp?KUImnOZc=LI=_!OX51g`j<bJn!+T=mMzhG3V!xHsZth7B
z67N4tMae9;F@Aa_W2T~8nUWmX*I3Azm#TVSXJdY2DfoV1$zQAxmiYoM$VS=t&${VD
zX_gxWcNeG1tZv8#T+(u-nv583lQ(b%;E~BmV(ScJhZLH1ChjVLX_3OYhjzntsgcGW
zuWi(2Nh28AxYuX?_qKj$c86=zzn4$-f|XB7@b;nfnpnK-Y8ew!)5sRjm9_&8)%!TH
zkTAo(`3i3B@7REScWr^j7EUs_0?C#*@hk*gV2=xSgx-uK21&DA%A}?qow4KXYGA2G
z)9u=bQd5GG3^oZ+$G}gdX<8?$3HYMaI?4#qhDWEa)NOfaokyDwu}Z6a)^uFUE?ptw
zjE;#2_Hub=$)_Lr)p;rokB(<hlqS9iOXS~p2~VR?-bA7P<mkrB!6ERW?+jtmXve~f
zBxaJvFdX8eZ@$$wCUceYNE9|l7`N$ea6W7-$RhDQHs%e;Zei(zRw};j5mh@?U+fSO
z{84JClYF8PI_RdGkavI`rpun0exDLSS|NMEDU2E!9=~GqY#-^lJu_Q~vTJwhwOsy)
z_O+DPw3|U0Zey7uSQFi&Ap-ZN96-mqk#CsqLwoaHHb+81kFdlCJcg4cvlC-FT+ckM
zSx<h-)@lLQFuH-+9DRIDQtPRwX%H*y3#)q${4ArBc%I}AHeh+|3e{>b^e{aj;cZ=K
zA0<x?w)WG=zpm?>oAfUjBT_sj1r4ZPld&qs8mJ7DOlq;gI8)>5qJKpXN7-2ZA|kR^
zJDofJwY2f(RWK^5akzYAGU~U|7qbdizJu`&epUb+IHY49iAF82UyCA2xPBL%o``R7
z3{9<>=5m5Y`l+T2IVxZ_S_DqoTzKx)hC#_i_+Ks5zMURa!&CEWKP&f2`>YvnEZSde
z9PMw7{MaOI(gC_I80=I-$7sa5O$NiZ@A|d%Cg@pNm25W62pYJ8ho$3r<V#SHE=dnK
zQ~CioiR^tH);QZnRbtFmu4q{FQ4VF2qA2DF;!#SY1rd54Sa_sHIfyd0vBisb;hj}`
zfv7fq?l9tWHdh=i&{xSiX1X9=yy*)x;og59*@@1AgE%<S5wNQrUyRX(=Dkj{k>#7s
zL*4tF22lCkr!W>}$XtBHRXc@zW?Q}VP)>mJ?^nzbu)*lAgbd>E{XLfg;YF&{p4Yc6
zab4M#?PYa?Syi6MCX-OBHQyke7HAx)X?sUL_XVaQqs$_(@W+{lo#Jq@h@*tlv;=DF
zt|uBX`Lyh5JSLEtHs`$MM5?P#4$rUWk|z0UHJhlxQ|u*kw&kB$WU~^gl!xv|e+U4v
zLIEy2=I70swT7IwKe{{tvIl7g+3rMfpru?N3W4CC1-ABW7c1N?L>kV%UO{IP3_X13
zTG!c^Zcd#j;P+P8j}E^`U=L|dX;UM+s?bA&JJOKo5W6ja;_?CtkCA9;)rSk|vciy4
z6&msf>YIp&om(m*0@mK&;4lb}_?<wDD0hd+@ZzO_yU+9I(z#`e6zmbUQ$g?rx*Mrc
zJ^rn@|7fu8AMd+B5y>XmJ7iKAuE&)=N7>yKvI4#;qh0QCaW9CnZABizE6LZra_|TG
zbF{FtJr<epe~L^#j{h_ZQEYQ9+6S!IAa0i5YP-D9#4jMBf;I2GwY7~_;vE3rW6G#s
zNO99K#oVtCiuFsTeM1?#c}D#Oo41!Tr#Y3IoajXPxUny5v{f6NSUPSB5^J0BhCLf8
zLnF!tkqxNM6+@n9`EtIzxOyz1P?8z@$IaSDoiWG4NjQYoFA#7thah_1pT93cBB?C!
zq9ezDD?L+V26WVOWBEr(u`vM@3lL;)Pn%5JtF(>CpjA}_;gm1ViW0$gKLtk~^NVhz
zX$bx(juAb8>R`j(w;}W6=vSi$5*~Sl84j3-$<f_5IG#6%k~MZ5U-m-NXXcwsvU1VR
z2%yiN!Gu_W1}SgBM|2kw;(x0r#tYe^HuE|nMgnR!N+q90IeQN+$Ws96_vI+*=(T%e
zz(0tG{AE`3TkWri-Of8_OFY$nv7)|S(t1}-zGYKv^hu=*_*9(aYL8(ECq5O}hrxO}
zleVYD01p4ASP?n+T1pQz-bbDsjE0u240f%avGw={x<Lv92Go??6Ux-@K&2u<Pi#b&
ziBS2!6CYoHh!*gBhcN*nh%u^@_-Jf53y*37Zw-OUv<6<tP4a>4)siD~H*p;fMyk0<
z`@Masb2#PvE_&`wBgP=2p-O@J@w}*#7%TP*WDUKpe2J^sbP?2*cq3iw3_1J1{Z%s8
ze-5G+z#X_JYLmqVl^a>X%=}VJknytc(a=1qgnu$$7~ak~?sNeG>ybK<O6RwcZ!@{4
zm6ku$XYxGQ`6a?{UruSU1l&Sn<zf<wydd2(y+{s!F5geKNq^pCZtap1ZosW<WiI#=
zzJK7GiiJ;aOe-rOctQ(re9d}ZqJflAph)LDDd>}{sI4X4<DbD%blt!Ejr`i&?%s!6
zXS`+b`utTRn$84pbxWX08UJav_OOkF;Bi#9O}E|lwUGi#p5ry#dtJ7OjVJtq4-*@_
z+$@C~y4s#$5lg5cus(*emNq;WgJGr<<FvcJpP7KR$$f1MzpB+Pz)VPEJXx_;$0nyN
z&9+JLNeWFpIw5Wpfa2)x=dEm$5b(_bErAtdC=M@N)cpd;2tI?2K2%ijOer5FrH>fK
zV8n2;zjEm}w%Y4QlEjNd#+-DVimEP14U`?rf_cWuK>vW3T9A|k*!mxAB0>I@#VFjS
zh_aVM^KrX;KAA0#m1@k<SuJB4eiP&>d^sboj5nC_qCEpDG<&I%nx!ec7JYXBa2`Xp
zSLj?ZODP~;t>^`0WxvIF3G1}#=;!)6dP|OE2G?^wW!7>gwm+YQT5MRL0Loct|6c`h
zjEmIy#DOJkW30s^l?+T}I!pJpa&4Sg)z}SLSwq<_%G>s#J246s9%fxpZl<a~&Sbo~
zz8J7gO82WAH-(g#Rv_QM3%CY|mDXdT5JhP&1(X35UZ%*3;upJ+V=m;iftHZ%LO-=%
z=IuA4vDQ#gzitnj1AdsI$0DCmhvn+;dgcTa@QoR52HL-^muyTS8z#Y&K*I?WPHoa0
z6yT&@)#jW#hs+Ny=_f(LNGy2AB^tPDML8{!0ga-2E2HI8?WhGrj6iQ6j_FHCZT7bn
z#1;TFJv8e;i-?1RI9u>FT;L2>_NPLKCYF2W>5oW2u=X45f|bWZ@$Jf513T5&V_e=R
zPX1B!vf{rXjM@C$@F7AmGz7uB?K)b27YIF3#9v{<!or-w_SrwS>6jzzH4ZaxPEgFK
zR6>A8ra1eC4u?1X%cr>?W+JF!A1XZ*j{`I!aOh-8C9!40@%U3ym|d-C#8Jx5^3kpG
zgtT$`6~20t+x#BBtSYZurF$Vr`=VhAOfyQ-T>t%Y{?rCCq?FYH2{t@r6%a#wGDldP
zY0OxXNtp#v3-w@<^Wnrn9Q`5-Mbjb004e`FPbXn&viXMdzDOIE#{%@=*8!vfHQ;pD
zo7AwsLgWtllCo|^E-hTOfCfvsOej8oBi!WFGX+Brj^H)}&!yFXMt1UhiS!;51oI^c
zHp6>t&9T!VaedV<vM#=cZO#mV(dNwNAl@o*KTPhRoQRa2h_gfv&6*uX3PKH1VnRys
zOQR}Doo-%AJZF;3cXL<0C|*P;23VjnG`t~YODsf24=rkHA$@Du3{o%Bp}0{k2TLy!
zw82uLR&}N-9m6OHbts+n>hI*!ga;d&)m>d}rS=skYo+*2jNJ?qibsJS5b#&G2s)6y
z3<!SMijjc=A|0@|5HTPgRG@C&)dXbgA|c=vwls{-5h0bNzmr)${H-ep0X-&`p3+^?
zVFK)6l6{niXRv2YB&wvdPr|8YcmjpKm_})X!(9ySWU+?ea;lC$Y@tpV6(?s1&mP|n
zPhg%ayNIlFifUO<dmJ)7qkZKV#uTN{>6{ykXACIiKJDbh4IRaL7~XzK-GfWCBv3XM
zR(!JG38v|rgHnk&kB1jKj*IRq9#SW1%~9egLUB@1wj#4LsAQQvNQu)Uka9t(A-f<q
zwPHf(7pUKx63T{{Tb3uyjP>WF0vFUBSTR-QZwP)-hzW_pkFQ`f5QdGzVC>q4)Lw(b
zLkl)N3DAz6uzkrX)E4uTxI^Ed!LY?s|6i(w#qq!T-8|`gnBa8DlJ%H?3FY`0D%Vl)
zB{)`wz5*nK6@OzI;Qw~v5C{^S`gai9rQ~EzfbH|8eeYkM$M%(DO=tG^^_5{)0QNE8
zj^{l$lE-s=2ZOUH^H#CJv5#_RM~w$@?&#LH-*>$C($xX!xWI|M@Amt2@#*9jigzYk
zPNVTmx-(0vo8Z5LQQC80?t!<mE7RZaT>-DOxq96>BYmj;G7jeFagO|4e6D?-*++Vv
z)zi;&gyv=c^JYz~ArNQJ0I~3K59qh}0x<=XD6~q2cTpS)g?E~723aY$v`!qmjnUpD
z*`Ds;fkuDf4c|;V;{S)Ia|+I^>)LK?+jhscZQGr6+_7=Tw$ZV3$F^;BY}@wV&-+(>
zRqJ5aK3XSxuQg|lc@2BRzgbG&y#A+=)aK;tzv%dB$oRFf%-!vl0Zj-(^NaG0C2tkB
zIkG8H>SHD9Qeu<7r3AqtTOy*^0)I|o!qvw{8weGF-m=pA4t~RNr_h)PT}twqikTby
zN%YxD%i9iq#?|d6XZl6!0?nJNF6-2Py_L9>UiuIWM5G!$rs$O=|DK$lP}<OVor9z2
zcV&Nb--#(lrC`QX0B|C=5atuO@`qQ>;eh0`ZXZlDP7SY0Im8?ip9SMvXGJu2XY4Ml
zg<3;Qb8ZWKO-{#`Q^2mz-BEocv#av7S<I5(>81TgUuGtPepl%wya<<lJIW_Um9j?b
z!!QvTYV5*kvvox(t4O9I<+BXy|IUJSF}H`ukTBpA4xtV|An=N(iuR~#S{&_FjVtyZ
ze<Ftf@izDO55(+|YHs|l96%6z7gmfxJC^3ln5Khb9tLZn4YA~sX#CsO!{_Q~$~d<W
zXpCz~s2`%{Uqg2Te%HC=;Z0&wvh7?KpVzIew460}X}UE+Y%kD&M_^j2p!H5~=;9zR
zWkjN?7=3gz-~?{H_8bDU!kc=^`4%X!zI{ZxiAgoEr&Iuu<T^mR-tIwSvlag>xj$c=
z-`8*gp0$Q)x9u_P4GZ6Q#yAVOpxIFV@rO)`St8ar|N3NVf7X2!n5%X6JG<>~Gd;U3
zWJT}tkjHw%3lbtvk?iG4S0%s(5~)7=gUV%}g3u`l@SucRqN+f4)f7J(6P?y0Jdrdc
zXHsNm#sUu+$W0c1a^WtT1?@fWg4^j*Qh(^rW+F^q*x<n#yg(AS3X{r&vu#bHCmyIR
z`&}n!6?vn)T9EJ*<37@;nKspz*OU_fPjG~kmdt~B(TFir*axjgGq8)N8(E!b0FQgN
z*eDARz^6-*=AvV$lb!2EJgl@8a(OZt+|~Xz?1@tI*kG2$d7b@Wc0zZ7cn~-9^F!eS
z*gfZ8^K&(BVOXHTLY?w4(9T(;B-p+=+k3Bi#+0K}fh{ZjUHtM$e2@G{QJ#gKl}_T-
zdzwET!L3!UQ9gUzJH;}uF^;TEA0~>$6S_hLWRC{9ej0Nl&?#}3SjoL`S>--lwUK|d
zZMmo``_sfMNy>g|&_L$16~AcrL3Sw*jh4;*bgXFx0c$%L10XG(xbR<CuPVDJ!T6%K
zEfhN?{T>mto%TtM(C8>WP>5<`k^14J#6dNZp@x-aV#X{dHbc!Dd1_V^*dJ2D=3>(Y
zi0(ld+q>XQoYTK646@~rki9wbZllArbaGLtn9{?>EDTRn74O`(C<+iBd5F|lyk#q3
zQ>KBojy1(p7}ZoSQQI9>Jq`I_$bk2qx4c@kozFcaq!zqEvYhHm<FA-DDjObFkq)ro
z8;uR43645nyFCq7R6Hq93*Gg$U>9))XitcWK#-z{sJZ=OW^KjF4fXaO{`xs3x+92o
z#(@d;i(2gfbzNjN_2=@GBa!!bMsCWFebQe&j(Iwc>wa+V*&+zOJb&vPS@Yl^xh20a
z0A-E@OlfQKtwERI>SCgXOtCcA_@a{SN7PJDu_Mn9$M69~UT4}JJ|f`oSi4gxfH`TI
z_|)VdV`5>=sF{lzLlgPebLtSUq$Yc$N;Yrc<pG151^OY<zp_zTp<zXg@=)HB8b@iT
z6eG2oWC5(0hOBXaphXMS^m0On$C~<(ZNr*N>Ib7Sl=^axID(I@!9)U8!)+f$aX%E4
z6IAS!K=S_NkzMH~A=w{a^O_M^0(^wub_s??WhWG~OKOOblDQT$L_hrk?%G%Nu`;5a
zpUJSdsBQvp++Z>MO)Ms_Tl1@f4|6Yqy4H;2#koBN?rie3c9^IbN`8uQcS6E|Q(LB!
zVuxcH88hu4RPjJit+GNuC0xIWC^|8J3)zK?txTB$GEp}?bErESW(Jo&0Kp*UZ;^Sl
z@ok64YQK|S$FyD{7or`4h=&-rahakswciIOos&+eM7#x8%~kSQ@%l1l(Kl$F`I^j*
z%ABONe4JI3bnqQsgL9+SEC1Bd&!Bynu}2(fhf1z4^{z2>WIfXlEl!yKnL^o}$D|A~
zcBbAWeo9h%>Gmv$zMh$k0PJN-n42D|oA+TZO$;*W4PQQO7hDZS{}otMbvN-PvzxUZ
zZ3d#%{z;lm)ki2)n(63<5O|i6zQaj>WnQH(a{!MU(ni*vpu?<EAD{V|)_GW@J{T18
z&tR~~RO2HcA5ZyP^wb@t_&3JY_&qRA$h#wQcvPO2bjEFl|GpD`6F@Oi;pOi0s6@L|
zwF=pG43*sX6G-w9VznAO#awJ@vjpL4AZz{wvt)A7W=K?}<K7Ur!P+u8HRjf7QyqT(
zIRz5G<1}<}r$1pRZc+>#!f>3&iBnRI4SDV!qTy(r6haU{UqL=vFxr6+nVVZ2Re4>0
z>Ca<B^?HUFP{^{22cRhen1s`_NdT0|W|L%*1?MFG$T!6Xwx#hb&cV0ENY}cJFTLG@
z*8Lv7CtwZVVuZ79?4iUNqbyWsC8P*|nYYx33!EtK!bsd8Kno1J#vyX6H&uC_tar^0
zc(Y)n%vRsvlX1^&v!q2QiouwuPfG*>k&WF^i<ik9aLg4403j8Y24#`+5-D0ot2Vol
zl3LY)&W_{@k$w|TH21Duz=`d~F6OCCfBD%cY0qyZ@KG?X=UT0kUuG~y4FZ9&)yKsE
zm}?q1OWEe5PxJk#P3DXk<Z4057;wmscACY;0uskAiZ)tVM-H+a7a=Mp8b;-^+U<dU
zJX1y5k$Z{10Qh@M>cKCLfB7a&<GQ8vrleH;pY<Z1Z1S3@E=o`ewW&Nrc4x)fFn!D4
zVEiO-orR&loDuJ0V1%K>58$$^Lto8o1%tB~0MD(uC<-TX3(9C~TNghbmB<Mr4h7$l
zVo25~Cj@Mlz_B%?a$Tc3J}Y8jUHD-pDMSF-ePJnP04(&#(tge^ET*^-E?ILB*;k7?
zdgC*<aqv2R2~r4f3%zAsZH@9Oun0dsgH7>YB0D8AH}3()e>x)kh|^#9)nWoGcwU8%
z$z;xwIcO#KdIO<)gN*mf5w3>j)~Ep%+(UMUkZRM##Z99~$kGUGjL9!$_<x`BlSG-R
zGf6$S0kpTn<B*oBvZR1a_Wng1Jmcg`Iu3A=L>YBhG5dK62oT*D!*%+F(t(McFm@yh
z8Bq_-yGulrI9+hm9_UDWs$xa3V?+SHEX6WU`}|W7bK>r@=u5}DNLV2Eoix$LO*YLK
z8Xh&C>RJLQvuj8H-E*S&f`YBrBL#_#VO_pHKp$yqzWNWyG*_>mBAcQ8B?i>*4<QN^
z{aou7&ggo3^r>gVR0eYU9KZl=7tp~d^te0MnJ1X)3z$%ZbGe_FLTKG%mqpFU`RR8M
zYi{Oq3m;*f%Bl~GQ*C^OprxJ$zWltG){`QBK9dT<T-Q_POKBRoY8IF$9RN-|F$@X|
z05A!-zRi8`s8)zc*J>`3`i%VniY-{ITE0D@%Dh@2gSd6mN=;_KY?2OW4z@8bA&ZaR
zU@c1Cx-OgkY^LF?z8fe<fi>zU&u&8rS{16XP;zk6D9;oMz|NF3G@^rHU5--)+ZN=p
z|E#>I$xZ5=27n|yd074*EsyYj5<)h1_5>7aWG>eK%MWB_WBH#|C8@<22|#8MN`S@<
zA?@U#mIsN{=52`&0lnJd2=T62?U$GK-Mw9@DJa+6#OrMWT`O{RMgP3-z<ocZEWZdj
zGRY|G6j(x`PbGyh_e+YWP-A4SI%?l+2Dl%o31(Qw$vekxg2X7HUwX!Ftc{=}&zKY?
z!i@6uvl2()uLKrC*f-6h82}iC%gqgmtwEYp5S6kQjLriWN?x~1<}#RI;20ZQ#431A
zvuoEWMhk98=T9`NGARP?-dnDo1u6-r&r+=b+*!wU=gD;7h~{ME<pF8mtHV<s`jD$b
z;(;}7xzzz_JXQ{Z;T?)l&)!i>%ws<hNmP(<B$sl5yDBTM-^nKP0N}o*TJNVKJVf(x
zv_p&{aid8!;=q(qz3}s*?9(yAy7WTEwV!HsR*DcPQ$b?*Fw>TV!WN76VMc4P3Tug(
z+kh(+(V6sPl3^AEcJRczlJTMppYT9+^qs6E16>h#OX6|2gQ)beNH8kBhuJnnq@1&*
zM;U*9I!KuyS?^<%0{)Km3UjY>Bb22ViVys$OqvUr6Nb;6>b+5K*d^zuji3_cWa`C2
zFeK9p(<g00>M&If9!!C-L7}TQvs5%bm2jlAWt&*)JBQqX&}D+e!_r1s;$C;^ccOJa
zGO%t|beqj`h2ssyAg%_LBQ>*Z)AQCE`wPCYm;>)7sX!^e1?bF>7y58qXgz(l>(Evz
zx#l`Pzj`CSf3jx%G^zXeOAmluiQZj4Ra)p#0{olI8!~%ycCK^laX`$mZAm_GRxEMI
ztz$E-5|j$vc)ps#=kmQS0qFT?x1Q=fr$xQC-=6Du%aEQqE&1vCd;#CDI$*yLX7!@s
z2qB&uvcAr@0pBe9b4A_89nD$i=SIXS<R2mXtm3ObDUGGiOs2s56FW0Ak~~7t;^4z}
z<Vf^5Xv|P(j}ZiJYxODY-B>B4#=CEB+#_(%g<h^!3k=+WE>PP)`SQK@5mH5&*DfLd
zfa$MYfho$lnXV)z_g}&4Lve0D^>-55=3}!F<G5p|19Vz_N4f_(s6E13z(F=2Hk}-A
zy94suPCHsGe!u0v*1NUzwlu&0c3YJ%5i&(n=rGI(Yh0cZ)ry+BE6{QEYw38LIa=^}
z8rX|Ex8)tMvZ2H|?N~*-<}GPWId2|-Giow6(WcjSVK5zPUhHP@?N~&7&fM{<*qGVw
z?{D6m23#7$DQ)_xu0Y>cP<jmAQNea+E>|(@c=<e6ah`8y9dbMoJ&|x_9q!%lEo>^h
z@N47w`eRby@C+D8T=2r1g(m(>W7Rx3xhK@Vp%a-04itn-h{=zRsYz$A(^^UZ<4(ij
z?YQhZz{qrGx}(FXAc^k?xidQtw1UuWhRNzA02;v{@acODtYi>xX!rXLHvq8!9JxjP
zxG-qPwPVKCXAjl5>y#Vj1mqp$9R@)hx#>Rqq=h|j;#-i0`=Rg2Kj#t))to?7P!o5^
z*)%<24lfqB59{!PeluampVETbWd0}&;Z7`AM)xdCZJDc)Rhsvxivz|aM?P-GtS?|f
z0L&<f7j>n5S#PH}hhlJCxIHyGC1ZH<FqcJdQiBCGF|V`#mgm29CD!(o&;MNYL~8IJ
zU+pWsELe12Mk)|`2Ad(tIGmTD^6SysN(#gozZnI0QCFr;BffYuu$;NNfC-Yy4ckae
z1Y%4il=`NTxmLN?vG)*_>8B6351@wg_(vz7Q6rI{@gPlJpSxAHFTR`0L<d@++0IX|
z41oS7dqdo;B3`ZoDR-YNYKgD;#4x0cq^BeMT%$3qFLEh}WFB^~voTkyn9f|GHiYV|
z3J#moXK*q@=|@56fbFYR)5Dle{Tk9hcvYT+;Q;EJgJ95lhat(-<oyg+7T|HRHYObY
zLNVgEBW%Cg2l@`!tNe<Q3PvS7DLz)pKhnNX=@6z8SYmgKtu^)NTG`~YrC)Q*wLFzx
zE*@w&u4N1Y!G{l?K4dagI`kmG6K_gdsf(AqPxMIc=*yJ69kB(C`mGobu-316R(rfV
zQ17x;(Y<t9KT=V<)c0jB1%%cn4MygLPoxS2`pd<{)x1pYdHLN+Ef`L(AMK}iHoLzC
z>Tmm#s?@8N?22i=lX*OS(wiv3u#16WEYk^X$@XLXSaBn~t2gco%gXyBZ94%w_#`}B
z&vmf6Eo+-=dvuzUq2*VfaehDfM#AlMi>uH3vyH+Pn$b{S5-m-g7@*&n>!(M*^rF4L
zzZ<E+V>k44K%`R#o}oXMrVe$8Ha~r*4EiLj=ya>17!OVu$AUagDtF`3ub+{C1WlMs
zKFKAZP9m5LzC6){G!amjmBvj6ooaMBiwF^snA6Dj#M;#rcx{M58bal+1Y?Sk(@<EM
z=TKXCA%2uhNU4JC21tH@v^kwy)0u0Y@&)?py*0<;X$ymzgu^1x`{VjC>?YQDMa$4<
zK2_)hTBeyhx7qgQNSPaU;8pw2Y6pOswXN_F^%IY11!0ookVao&rIr5}JKtwDHQE87
zK(Ng}!T^+o>--nbLgMz@|0O~nYjQFC%RgRw^ifZ(I0OH)3_!xTqhoq#PY>SJC&R#)
z1RcopAB(*$#AKa7Rc-*jzD}M6MP)sfWXFWjKr-{pPV0$l%kqf;kjOkV`C<eq@vg>(
z{_rC3{9sPr?{{-hu4K1w8icm|76g-199-{Pq;32jyzlPPb~y9g-el`PrbZQgjtp{p
z`W)1r9id#k15{rj`~xVq$I~x>`}xcq>kz6J>V^JkiJyi;Bf{I#nYl9CYW*#^GDR-Y
z{nmgvzd^8$L+#MogX0t#ByHMs=@Y*JwU$>vQ=!L5hDoove`;(w`uG--mKP4h49dkb
z7xnHAs-rORW(~ki_(C0F@{*mz=_KQ%@vgu7kE6*TU<C#GFy7N!VZap3E>(o(XMfRv
zQB#oIyW?b@{-nXChjcZtpd{edEjpub2|SldPIG{+bS!iDhtR^$>(@hp{x2Mc+?;?j
z*z<-Dr*^FkR9dCSiTQW))6e<`KOL_gdV#QOKn)8T*KygFU%?Qm+kw7tSw;Omjpd_x
z9Ma_oD$(nYRE)DFXdk}RZ*W%*w1fY(H$gf7>F3ZPSU9==zke=7OD<`h6XCz4<@s{{
z22Bc`#NvIb(<c@MHW{~)utw-=iS<7x8j0gqPs`X$U2gnSCd(8s{va?u{smK>=yra7
zm}A#j|6-P}>j$V52YfTBv)&hu;%$*sP$jLJ+iEzC0@){<JGd3=cbY4rxeS13m%jm{
zfs*fX&=cHc?SOW^tbLC>|HD<Qme2REmSy*MC-f1M4D%wvwTncG=zDOR9>4}w7)TR_
z#7<~;jccJ^k*9fXf__1}9YL)m89~RdzLEN`g4bG=67xjAvz87BzbaBgV>0-FQ!|ai
zSL=7o<gj&v9Ev_WCNw+qoLMgz^~MM;tiq5eC83@*9{}m0{XzPE_-_$}1$K)TFJN{$
zG$*wu2F+JM&^Upwg{KW(CrZ^#g@Cn5@eM92!5^!c)?gG%W{vzM`xm9_cC0Tbp-7E=
z!cE*O_t3V3k|*t;it0m#5h%q=BtlQC9;;vH&$uhgcQw%9G_$o|FQ>8C3Q`qx2HPN{
zBZJDnYXN4Quyx`@+6F&_bV8&?Vr%M-Om2R47W1+6XiD*SF}tGj&ad5GlM7aAHudPc
z7ZG>jX;}0L1y#gIRn5rfmF=3xb2vJj*^;{*7jEjzF2jjF8!si=nsatQPt(IeD9Y<u
z!Ln7J_@2rhhn5(P7(^l#vtSo;w{By>VoRcVTmtxOh?taJ@Whk{e@;_78%GJ#?4K8i
z{n3%8OcGJ1!(RE}uy(tv)3PS&^0g@HVN5uqPK80$zA)-oQ^_>~JZ1YEX=p`EQ)USQ
z^$?Xvo`}ci(=sgl;%4wuzR`;gm)iYiX3lN;lI}2NAx8#TbxJCFvRLH_$bUryC%j(h
zm<e$IDi9Y3tpBpgI}>?D?VV2pX?AI1abPmf=!SBCC7ncOfkC~Gvvf>^21q#arhtC1
z-@x4`!6cD|d0djH53V93qs+YNXyd>$m&#WWS*vJ~8mi(Dper`pXcw?LJR7d#@+GX0
zTDHd%_PRO1Byr2l1!Q~FuBVx_DXqrUkpq0@f0uP-9UgH8avOz0$D2hkV`UhnVZu)J
zGrE2NYI<Fi_EimaiQx1(Qcea2REX^e2H8cT^9Xxbt6@4E>MGF_@|*L{Gu`)p)OH%y
zqa4k4v;GdOz0no<u;R~O(Qd%)<=A<XZm4ioFDF+X(&uhBH5_i2;YQM(zDb`$mIavn
zExPdJ-|MV}y5IW+xtf<9uxKo{D_(hoU%g@^hm(+$S4PkB3WwT!R+$0wGM-^>o&WdG
z;30vm?oS^hxK?}w>KAZLL_gat92_6B$_@6&K;#}kx5=szFUw&z6`=d>s_RbwhQ!Hv
zDl6^>`GOB}#3W&oG93f_Z+NR5sB6IXCn_s41oAm{MD~L*dmO(AS+5Yg9kS*9B$(P+
zFKH<fO7TtfqED@SwhdvvO3D!(M^02m5(rG-%=C4t&Lj-<h+I5%e}pbIed?)PW*TLN
z@{i=K$#9`HZ{^80#FKdQL*Am6AZV0Z-S)wdw*yNo7dV_2{uYMOqkMDnfLlP83eXS>
zn%IjQd(TUr98CcsL>^(^Rh_tUFai(Q#&%h)BL@>OlPMle2b;e5xW&HGj*e0)D?+ys
zX6Z{*2?3*|KzUD5U=^uJ;nHy#dM|O1io*elkuyu786zqK(Uu(!2CRdsHmls*f{pw4
zEqhp|3nLQx{rqx-gZ^1*CJEr2RN7k9%`OAQE@WI@PFAUf7~mi=H;7Or2=>Tecx+b@
zK}M$%98c<_F8u|emAr~<Q>dtGkd5?VY&AIgvpNHAO=ZaYtp7$X$w(xL_+nJO^Mgc<
z5sl~F^WlOYBjGFrKjsiZOo3s~2r@4FQqIR#7KJ-KQB~eknFCQ)3jvS=t#2H|`ugRO
zSt^nzfR`EL$rhu!u!Ek$-NUL22NoaFyTRaqu`o6x6^9+FqRk5ey!C>lG^H+S9vJ&~
z*|14(uNH<;3PU-KI^Bfh{fIu(weFcm^bsigGL4l_AO4*|`2wm*Q;$QSIps|pc1xNC
z<!=Q+@)0s4X@$lNp$tfR`GUB?0;^fMFB)x)*fHf%+V%5e{s^aY0)ydC@1b#6-lajT
zNlNy<4Ly}wtWQ58k{uu+?y{7<n+xTd9Li<LTubd<3kUb@G~9^oAwKF0>OmW>i5xo~
zl9K5@6X7@%L{nmF^<b^iK7@O2@Y(cU38szG($aC}ywga6&ZXGiGPnnWzIG6(rItY*
zJ_X%!FbdWoi^DnoQJs5b^a))S7cuz1gX_OMP}C9h2EiFwNKme{DhE)Dq?^A8VE?8t
zzW!EarTz2;Wo_AX0961>Q-y{`gCt?$;^aybaR#LYXv@W|bE0%#*UVD-BMidH2+&}j
z<jb_%MN$k2ddWKF*{x!c$^C0Pn+ANn_zmRSt^yxiS$PPMj&G;Lu==P_`k1u(zq`A8
zyj%&f1Cmt|O6XdRpmia24-%y3cl<DY;D~{Kd|A|ONFDW@zK_gmy!08Wmrn+A-c(oP
zp?=-~LH*{A&wM7WPBwfsjwRwpJ!%qU;)tG0TcO-B2caw=KYfSm`enthfy$2h?Pl)T
zeC!9EqP>CjxR(U_0qyVpi-}MBT5HB68X4B5?jOtMTAUNC1q+3ngIFUf)jxlA<Q~$o
zN8woyWgs0QH+a&yzM!u!X2b}+xG~$Ba?~yXs+u~<hN7!78SF3T|5Dn>=rvr#{wb->
zjl!SK#AKR(ZLj)V+sqc4oQH7VEw>I4Zk$T)%bV6nXo%8K33j)y7AU58UKmgHg~%<Z
z?-YQi%{B<ds39zr-J_fEDLt8=))u5Bj@PWIM(2Rips$x6#2e?O+-6hCgyyAdj7M?@
zs3a8EQqG4U!xdbZ7ys0u3oYZuVe!K<z0;siN&SAXbb2Zq(WCe08Uf@G(#0i;<_N{2
zDDmfrH{RHPb3e91fS7;PC}?6+0ck&{6;eezjd#9%0FRfECuIIl<>96q6KJT1vd|No
zi=|9t+Pf$ok&)c+M1<#z;iSx(EK9Bc2X|LC4Q8`UiMq)^d2<z!R35QmD_}6%D5aST
zLVpanuiya+&9%WV>FlwaYG|RqSi2n&Yl-M^r%<~WLp!h6IHlr6c}pRVRykix`ye|M
z_msCsShOkFbcm!(=!Pb!uo+WFxTY9q^r~vBuYC`4tQO~ISy-}rx>N_!;)3CT-X>bq
zo}QkpDfse8Wl7IeLi=%34}wq&3ohuUFl%SM-{Beu`H@A^v1ICGT`;sI#MD01w{)B0
z=HW?uE;?4)$t6}V$In5DRLZ(z-+`??RvMlG1VPH>!gP{#Dpr#r$ygWn8}A#0uQ<sg
z{m=yh^tM7I=$LO2V@(4)<j<f0DqlWuW;RrRt3`VSw#p9^t*hC$*to*pUS~$0I|p+k
zUM}#D>0~ww=VOhVlPQL3{rmPGT4BX$sRTxf1CZq}gEok6BsY-Poz(LdIXFhL1UrTZ
z_Akxx4es-`_FdsE#N@wvw%-b4R@uyV?+A|w2{jd)X*`Bizthl8D?>>E{;;lAnj7!M
zIM03)BPoy6TfWuBr~z<+N|JltwWtdQZYud65x$lEBUSKtgro_Etk2vm1i5Cpa~U<G
zX>E;_$j#KkgC_;UTE(z1Lr9icm%V2rkd;1MLdl$WT_SOH3gD5^MNikosBnBv3I!*f
z_`LR&YTU?csU=ipvwj+Y)ovgbG#4>9oU8SMMzh_8`be)BoH@zD$1MI)3CduGDaMG7
zyfY{fWVL4JG03}nAGhB!>nZmgY7chNheYe3=SGIe@_NZ-j@(`cKC%Q~2o-n^Km2l7
zyPM)LMQcs!XMW*W$n#~U=O4u3i6Orke^>C*zUa{nqG2VZl;3y&@E^dqp&^KLVzDzS
zWlK*yT!!3h*`LVSY{@RKhO!{PMdunNM==xcqoj|&^2qDM6o*+z-e6TbYhuB<)5FwD
z!L=wNT7&P*KL@k2115X`rl@B{+EESFYwCzlvk^WHYQHKUrS<eJ@r%^-Va$f1zD{s8
zGdE71pgt5ab)$m-CK(SW1y`Zw7nw^&R`_W;C~6$?Z7`97P>>eNA=o9&3^f(M-qZr1
zkhJW40u;nKxk=!q8l;b*JjmAm`4|@r-$$c=f1++mbn+r=x?wGaul>gBW=uOW{+zm3
z39xrLtPi_X17S7F_OXUi%_S8X<JWh;j8$r72smJyX3)U|KuiWjhi`D0U8$JGExXJA
zUfsV8(Js?|Lyko1(_#_C^<o<R;cR(>rfT{RP#9J?0^b+gt&b}7(Oi+-DOL!}sVDEG
ze^HY~Zj#H5oCv}25P&>gRtXcLB|1(^0+$hubihieY$`vKIfVU{ev$esj#d3GG^MY=
zD$CP=qxjSVpwyLm7khTpyvF9(m)l?ZBwh74`$9!VdFy6s()-JB&@`9{HxZN$i3#+Q
z@dnv<CF9Cd&>NpID)xy~hM9edzVXxAv&WVXjH&?<QFB%*zlt*^?4M*$hY>mZ^pk}-
zGp9o1YGc?3c$TkAC_nXJ7xialPcI$+(EM<hvzpZxAh^v-Ok?0du)i`@-RR)x{3Fcp
zt(%BU0)GhzoGhpS`r*V&=H6T#EjbI#Wbcx1!Tz4p)TXs~2`-J3c=MsbQNj{?S9xQP
zJtB?6czA}rj~<Dxzun^f+|R!`r9)TP><(}ImLFIiPKRZCH}b2j7yZ9x`Tu(5X-WVv
zC{X5>120g2Fi_V2j!x#Z6kkvYQ1+H}Ur?z}Q1<_O$DQH;j@Xj!4b}+`%JZM79E|xN
z`qBYG3+VbU*{~V4`=5X@o=!~z4bdM$hP$PyD@*n@VKYM=$&i@jUo!sg&&iR=SU}f1
zsBBTH%}8}J6WXjLJKjGUMg;W_7;5Kgc-FVpbJmQ}2vs~Qt1fWoi*!^y9u_x39_o2y
zC+x;A3ph(e*Rna10lSF3-W7X)Bebp`+I(#QSVpe4bznh`GZ$JZap3o6YIJDEpXRSx
zuNv>C{&3v1_;iB4ku4!;_CCOgzs^X{PP?)VxCe5Z@wZziYAnU=@5J(33HQf6e+TPb
znWIt4jJJ?RFq|e^ijBA*YKDe6(yWI@EPOk-@xR;|x%1cdt`gX5>WtmF3<G$D<~%(B
z&xbYu@%6^d4|YMiwX1vWknP50Z?Edos-^rY$>kb@Jlj}v6KM82*?|IU?+xriN@ZS}
z6XaUVq|0*WYAE~)egZO#w%0em`+LbiGt^Px02h;gnZKS83Ce%7i2X47P?vTtMhC5K
z<o~#m+;fi?2RLGG*lkrq#;KB_SrHWh@YEI2HDA}GN_T{&Orr^#q6xLXtRu6PY|_nC
zQhkdGrXO=vS)j`4EHJmb!R>JqPG+U+)Q0KhYd!@odO@{0OD_e7Y~%T?{wh*AO=<kY
z98HpNlftZ;*{D_zaa*M->$(Z;#CGqevzdE@QPEys&&8)w(qtveRco`Gef#(U6Y#?c
z^Vmqmoj6@6N#c%PyPizJy6A%g5pr-xbDr@MP$c~gW@L!M`WFneNgOKP85J~KxCnPN
z6Q#W`XMD`c^e*Y&0LFi?ZyLw_M70!9Q2}E>;agbr6N>JbYJLrZH^2irNk51)^hQ*F
zl9Zk>$Yw?zjD|p+fst+a0ek@<s3_9lZ$EG|rRi^*L9J{`=2-W4dnvlS@Z$_KS0((0
zPnf%WpulbmKEs8B67Af&%k)tA%mSNz#1`~L*`L}eBhHyyq7U~kL?>}93=R-G$K|T-
z9)}&vl8A!UZ{iOK-d+9vD%;xlY9t**HCaJF7B14Li^R43fj+(-z;ptDdm)HWST%06
z;UeD_x!)wfamY+WxFRmMyEvAp3_GfW=Lieh3GpfNPH>KOGJ9Wf*t13`w|?`oJ!de;
zH+{L%;jBmK2^vL;cyx%ovIfm`j*18&M?8*a4ZmVHbj^kR;yXgl{!7&VOwzExYCvZ`
z+o!$Ly8(1btLid#GMpDcTiDNq(bR3=kG95?HfAX85hLgzwQSNikm8P1@D{N#8r*>$
ze!E6IjSD4pC0LFa0m(JkqQ@NkjWkDVI@vYTDH){*Q&_zFla&9NpvyKT<gtXm5Jtpc
zB*y{==l~i^Sx{`U99zl;#l}jZGW>V!Ui!}~e%k1^bkEDWoai}#v<t^ZANL1ea#DG`
zpb~BEvFrxe$WL?lD;i-5Su0$@+`6dt2MenFqQ39$W{shN>DPN!USt}v^Y59Q8)n0F
zJLfy`d9Udq?O~b7drCAC$YFlVwO>7M$U6YKa;d=IQ6S1<L<Dcmc;#(iXFKifVZrmV
z8-M-ueJz2JSuI-tdm_y_R!)`7!EZkOacMlkd`2J~BLr5gTExrz07utq#(o<G{0<^i
zjl`w4(t5J*Wgpv|is=IyKW&_<*vpbmJG23!RRdD8=lC=xI8>hO{Y}1&O^Gv7$?M;J
zvdaXA)Vby=lvt9BGIY^le1c)bRIRsL|DiBOXuk|3pshQ=X|rg!NvzZLxHM7&?G*Tr
zczu}#hxZZkzhy}p`|5TW;-RAc%YfLtLj!VP)?uN&U@gCqvrUZm8xD*_3_e<BW*<lp
zO=LWmnnK<3u`Q&^YoV4$PFyPHrT|T%%Im@CREwcPmY5PYtv+H}Cy$D23MyBB!+?NY
zMP0A!wbu*qe`6zPMLxb$M&f@hb4?#CwD-z+>c%8hmK0qF6@8bO{ceFHg9^Wm-8R{r
zD3wBnrlB_+^gcRua>zP84LZP6H6GWSS@C?a^6(SB@ZM#i(B5qTHax_NopmC%&Ed>b
z<g@P<M91OyO@e@|SKl-7Wx8Jlo1nZ_vv!mNSttO$M_HzXV>AJNVsr}j1-R7)KFAq8
zF^7<#CVsuI@NuXGsM7<@xm%YojwtzlJ}weEZB1q%2Fa0TnRSdc8}OJYa4=<F8$~c6
zkD#4%$UDiAupD@xE(~5d&J&_dVTOZXeT!lPhd)4;kcxfESkd_=RK(tms04Be|E%8z
zZ-@Xg4zZ6-knrB54eZ(oJX)(%EUXdOqHTUxk`&62#J5TY)Jz|eyXwvt(la(gTQiiY
zs>f>K(s!WIYBM+qR{<v1ncCFR1|O8Gf%|Xg-6wx8PhI13xij5*w;D5&B`ctPQ@o`L
z6{w|ve+Xl+L?>0L_)Sdykt<Hj?PcZ5RTY5Ge>2h;;ZSNO)Fj}0E}}fZjhvRtS>5x)
z6mrc8CEuC36`BN;NZE9>iydQ?O~?H7mLKRoRwn)LY_7Pv(|c3{ts-beU&85Z3k9`5
zg$R$^NLqaqOu+&k#$!H4sJ+Y%A4X-xWoxgQxY?A!l-6>o^%;IB1yDz9OKid>{@w;q
z*1GxnUzGl#RBTt+&uC2~`7F2n@kS7w6m&UZdbu!!@dg8bYZpx(fj^XhKkC7Xtuze9
znJn0E{N>*ym^$lrJ5UcA@6`M&6FIGd-Y|U~jG#yOxPBiAqrUClN%#adu_>@$l&15S
zOC&7FKkzLGk{cJtmXsQf7)`*LP~eQO)k^PNi19UEX&`s)ZeDmdl(Ls@n-o3KyAxtf
zBFe^0a-t<b)sq7KqtgW8LHEl!Sex0`x91hpfUqFV5eyt1oSldJzXBednS=R%QOQ*r
zQ;CKD%JO&GRta?m1T<;k-$~<&_2otN9Zz=k&<J>Xn8JpzzxMcj+ZT75ERNHiQciz2
zy*;06Vi6Y5znLd>y*a(ikP)Sjjl*Y*S46Lw+`A^xshE)w7$)qNCF>Zgr&jfMAd3-w
z+H&RYwCKwBnmL)@mu?<ik=Bq;F9399Q7>}>7vzi0bEG;Lz08o0nE42)B~ltbM!0FC
z5;sR+EoSP<(%q&^c=*>=&kHa4Oh?k$1W-TGq7rqu2qQN4r`YDQZ%euRer{j94<+O3
z@zJ}4x#>0#A#z%}w^I{gD8;aC#aexsKjWKa+>kNANz8q>Lmlo@U`KMK%K~DdCL2!K
zSFXaj$cMRgRQGXxAD7!s-`UDN?rW~3{T$o`8K|CPjE3AOrcTRZ$miT&6B?ER&qc!0
znSA3sGGukc!s;6jFwJZEtZU{(hyFY{3wfG~Qb{+uJq#;lqLHjd_34sB=A)QA8+IIP
zV^wA#=<8Ho&~kUaM*~n)RRK4dzHcAGc+@TGKEEHmA_Rj7FU_PDLeWfalQ;s-v4rat
zxwfLATQzs>?p#5U;lM>*7Q$*(%3YVd3$)n7;of7FZ?64;`B_n+p?E8Oje{cNOW90r
zq>s2F5f@{p(~amyG!Ut7sLyU*@B#|g;3U>1-+hdMv^4@Uso~#5O@N-!N$G9n2mYCK
z$nk^=d<Ow}6&0AL<7++MyAkE@v>{phl8AU!HtTe@sV#q;rfIF}UT_$mdvT^!6Zh?|
zjOo^$hqzzXb4x7jkd!SXWP?9A<FehNG?ZrTbZ?}(V1J5#!f<zc$D2cZ-HC=MB<@zo
zV6vVVlz=TkVJG@VI|HU|?R8z$WW#}{Gdo0QWDU`FEyU7SQz*xQQ@+CVZLJ2HJEGzL
zq-=fO;T6iv3u`UYD2<Msq<PBB{Of0KP0jhmCrA@+9|+_xILUdIElIDVAWjktu*#Y8
zf>x|Z@D=%9XwOxPLiC^+yX;E9JDepllw}*#6viJgZE?FOQGf?O{&ba|Rb?d!+TW4=
z=HoyDm!Q~Zs-inT=$XMcB!`t7&xlPrHAYN9$T*|=vLG>s#3NVCE{H|aaW#$oCc(9S
zF*KrRtt*ezUvT0fU^+5}dVFze+8*2QnskTex8E<CuDZ0I=lDxS*@VA=TquPkE0Hxr
z`4MGNARJ{}q5y@YU}c!#tEs+?Xkr*9Z0z`$x0+~dbOh0YyUfcy*;=_b`+fwb;~HTo
zQ5hyYBwR`m!ZdSTu?c%!0HK5SZ#2fnBQu9!;0=;GYa^-uI685^q-nj{0=f(k(fMQ&
zBj+v8({l0zn{EB_lk5_Ttv|<sIbmjkB3zb5U)0jkk>uWt5S0vl2*{-vP8YHyB@@$C
zO^e@WpA5qE4O&*Bi}qjF4<oJ80o3wepAR0KgPElzJO<3q>3>iEY8Y=Ya1IvcG>1Eg
ze?<z*f4`Kp>N^Oyv>7mPD1b18@J&Kik#u!Y<=}kf1oS#IBXma)Mi{|;{^QZ|u?;&n
z8-a5Y%^`Y#B-iZ8G4wESM5FWViK_dpyQ{<VQv~}5t%OOm?xsK`Yty(XEvucTd&i>E
z+$OJ<il@=~Bh{E$(Rb&T8C^v4dJk*9dsx$WVI3p%GN-)FNWL|4GvIhtpi1a|g)nyC
zq0jT5wbQ^}9qTe2MJk>zHGA{+ZuZW-!*^if@M;Hd8QGvs=iBL-OW%|nJ%NJu2VX0b
z%>|@^LZ+}6&3Vp<&<K@go$O7DX;%X4SYd~+aVL|vUI*3W+c>BBv2IRG7owhzl3q&_
zFlU=9Zs&_?*8M4;0B}xfd&u<jnUV$IXy@LsRU|8=<h`z=`uZ4V_Url-s&B*hI{f_T
zb91VO3k`THTMb8RmQXn_AH_WfJaf2TV7j#q<Iom9Au|&2oF0&kY9uN5YU+9&yviol
zq9B_=Y*r;xLKAGT9{F{{__aAUKUmNk3mBZ6ry!eJ{J5BY0$i?W)Yi33@nb2Fs_Pmn
z&EQ=i0Bus<W@l?}du*3k5A!cQl$SVay59XJl6~R1uGYPlqqetMW`e~*iZd{He^l3K
zpFFq>iCd6GHM~+D(_!mMLuzHnQ%E-K5sL-um8Hm(7tAy}Nm<}WCYZz!n7YzR3aR>&
z3GQ>pZhgBQ04%NDe&e7DelZRX4a?q7Bx$7fSFV~>2{GWQA4n<^*fYguT4B(CL<(Jk
z+dPM=UYnnjDznnKm<t4HzYA4DPiz0Gq{_wji48Wik7SP_U%}2u0dk8b7ctxBq{FzQ
z-U5rqZ=RC(k`S{{gq75DY;};mo1yEqrRq^`<YLhq0jUY>XD-3AZ25*=w$QNn$EU<*
zVcM|e1Bik#lpV(nz7#dM$piW)#=-QkcqIvglB)&(HhXv-uP!;FLV7^qX(AZn%_x^D
zCJ}aRu#dbf5$LSwjZvq(Km-HP)H)*6z5cfFQ=(vPo%4y2jarA{q2Qu)hFDq?lPCFM
zq@JD*Fy?aZl4$S{6)=bXx3p&T>U);rUKB^IQvOdAJH=}pQnsx)_H6kcr<8(xX(AZ?
zKFI!%0Z;M?T;?0TE6E;@JINlmTfS+uI*tG=XvY1?&e__tQl?93=M;~<$d9Z|RbOc{
zD!7!d>&Jk?oIiCkXxs=TnXd<5MKBysH%<QrY!CB)k%sxKj$?2cPZQ3ZOu%&?$NK%=
z+G6ka40BVc`x@4At^6REI7qxlQ#;`)IeESC>of|6|3pz8&o@UAm0>%P(>%Vdbq_X~
z=ePFdd;aIfws8o99WcHOwr9QqPOlA9!`khD2cEoh<QL2ar9U*<zdKCR>RURYGQ!iX
z1GedGKD!dhg5x3zk=DZ`$Ai}hH)fL#_bAknAi{wItg<lGa}8s)a}5oTK57URcyLeP
zas)o0c~LLr*WEihH(O}*@WKcVeIuY+G<OVlAdV4ya>-aV1}e(++pPD~Pf}dkIm!88
zt4v$n;-&_qXvS&FSPw>}y0}l1<Qv7vfG`6KY|ZB@)R*XyVB0i=8AFSRz?xk-MVPt2
z6d=ir^w)tL?NaON2LuE$oa;IeHFb+9D~fM?n#*qHm|(9y<t9}nWwj>eSa<1Z8XTez
zo+Tc_$?8b)kXGCX;F6d`W30O-wjXu9Pm<L7wgO$^D&9E0zh^remYX{^nyqP%0fZ7>
zYD@C#Dz>ffj^ooH)HecQua}8!25O7@(|o#zp~(55su>NO&pH!>Vyn;m1)HbqEA{Kl
z>_w+(O)UN)g^0H5RlnAMZ>x;Xg^MMK#n6tlS5ozftwTnz73r&f=N2MMY_M_H{a%gA
z%V-!=z9%L;E4vr;%h628YGf<62UJ_cd}rl&mT4t#nmrHP{+XgiHzHs=7uHViyz9rW
zup7mK{CoIE-BYC4$v+v`s#*cYF<ie`dV_C^0@U@Gxdh8Z05j%`{oKMw>rWzuzf*8}
zLgegtMl7PlRx;0_2&{>ed#5kECU&L!fl?WU)&zU?AObP2KG*;xt7fHR1Q?vxfD3;6
z<@I$Yp{$#Op^wxuK@*V%rWe-*>{V^KnYk1=i?5$XcejZ3B-fPP47R|>WmXF?D{k&l
zJ)D%Vbp4y4`!Z3CJ`Vc}yZf|{G5n>0VTenBNkYJ<o#!nWku*|KJ1i0OkW0e0j+mjM
zSM5P0n0|4_00f0=a6l*L2dEE3jqKba5ctyZKppS4x#a~0)t>;UP+axU@E4y^j#1Ql
z(O?X8Vd}5TJ2*;X?F+aTYlDB~A%RrMH$(~HK8TSnqV7>2`~(|(Xhc3erp6vil-Iuc
zdxhyL`u8;F`PN>sVA|SxPeoft=y{rp@IZV4N&a|zhwE-AA~%Fx3Siy#UVZSP7k*I+
zdk-S~-ou#jeU3(Rp^31P&JOcUvXV~nLX2Z8FG<h4H0K?>IGJNi;r%vrTID&Q-R_N$
z_jcdH1%YC|`vZ=`&yrx@fY%+y4sGAvZa+bCB^Z=L+}>zoF{)fU!v-}JUe2wJW|<eG
z!>!Vt|58?pc>-Ap1t4=uFHB*0tJCbCVV(e!{pCJcNrmJZwft~iT-Ass={Ka9>1KKf
z(wutJiI$OQ1bV!m<~qQ~`W`9wmJK0-=X?vek%&udO8Vgr6&BohN+l(~V-sZ2GW)tQ
zcf1{`J7#I?MpRFW>t%Z$gsR3~`1`GG57w23!$@A#DvQWSmMv7*OzB|Ou%BnQ1&vwX
zQ?&Q&Y%%{=4H8QaP?aYq4>vNQ2-@Q1w)k%}y!{dYydyvIPueU2IL1FbhB@%xeS-6U
z_X%xyaAbglSP;pJtfI}x)~0$VzFft4-xj!pD4wR{C=0P><LmuvJN_S1b8D@Nane#y
zyeY}TTw-Qx=N6SL;_1JI+g!^QHQQmUR4^a(kq<|2Zb>H!1z<w3g;TC=uLlOXrE3%R
z*52J5jz<|if9#V;?yIZ)$VAT!4v%?nyEUf_m<Ryh{LAGQS)D|dyCSVyi;LaXclp-%
zyKmh%e~P--(D${EgBU_O8d~gH(j1$F4u(jGHzz3q1(C>Zmd|_k+1Gnth6f8Cw~#uX
zeBhtvg|>gK966(8w)|XP4Gl)9A)!?0DySOq70WDzZ5D93Dy*yRXTI3T<M!>Sa^~Id
zqIUr_6QAd&joB?<N#I;yg>U8&Ayjkrm5>7$cJ9I&rl^0xZ`QwnXMR8W;13mCQD*Hf
z?)gmrkm?dcgx@QhJ<s`7z8otycU7f_pEd7BvHSSWo&C-e&MNem6r!P52zG`Qi%cjc
zFoIkeTef`JroR6#a?-KfJiW|A(BoPixy=F~y1?wjB#>F%Ht>3)o{aeT@dwZ5Ti=+d
zUX?=uy)G2kJCl5hmS(O7itY<a79>#t*|_O-W*U01=GJ<<HNm|FW_d}<t-uIdzbCXc
zNn^LLzkAAlx|9M_(ROXf<;EMvG<!~e#`B;Siot;~k#~!$;j#^=z-ojNk#}*|6sjwL
zT|~H9^h3dIP`Yz^KxYFk|0uZYndfH9AlH`%USAzb_Ejg**0j)f@Tg=Z*Y{1sPL^Jj
zSO|W9N}%D2i%Pj$l(|BotBv=)RZ}{UBEKx$^-Ri<q_nkB!nN!V?-h;NIkM>*NUqoQ
zm#?wS8dNngpZA_F;y!oi%kjt-{^}T@Xyx>h(f;Zpk`V21gev3=B{3xK5=C^)ER6j=
zX_R2fAu%8Q68cPsm37TdLOi|q8|4cDRix~LA{?+xNLO-8r75k6txl}Jv4+><h~{N~
z+!yGeyj2`!kkd(sS`HsONYC>%N%XZ;*EyV1?C(3CAq{7CL+q`_)b;8@QkVc3itw~;
zqlq!Z7@c-f$$ZJ+tiK;&!TzDY#z!D)J{)8~o6i=KKpIpr8U+6n0r4VPoSi)bPT)`t
zMw0Fk%&5xzJ*?+5`mLy7b|O3_<d+{W?9`vxQMnl6I%UNC2z2Y-HY#>vI91Ou!<d-{
zl3NE^G<)?ONK$pu*I_YX>lb=}9C;i)j}>B+aS@nhl|yd|q2KEWW_cAk8i8Wq>jp23
zxGuZQ(y+{uvUbD4Bsh!qxjQk_w%WXP%(V~1u8NI9sarpX&S*JdgLIwTq}y1ha-(5f
ziu6xh!!P0?I4_lgKX>+uT|@B&dor3vp_P6`GAb%Y!-~F#8x<6qIm)>JMf*uP>!)jf
z;p8ToaZW1OPx?h97*;gdKy=<hN5jXzN%94R>ZYOPr3<zzCfzpYswL6!nodSiFcmwm
z<u#CEH%OP6!a+U=3u$}!3DS@iqTjE&DL)O>658^56}!5#k<1}oK$2eQJT`zf?~nYf
zL>x~)&k@D0CoH`Kdf9;huUATx0n0?HVmlmIV@qclt;|!Ng3oE+QIvFMHifH*G?U1i
zGz}q_#i^G|KMit%D}?P8g~HC@H%N!ri>F);6+LuGVYJ<Oe!#{07^VpUc|V@^?KGNk
zkPN^!AlWUCB+qNNwdQ3KCmR7x%J3(2i(;x3pGoAw3kh<^I4lSN5(&qaS`irHm8I3>
zLyE&{Xk5;q?jSvWYSn^<>Xaz)$SbN}&nqe{c@u*m-dD$H?ltOtTn;<TV{t27e=Ul0
zbyu-#t(XM-aq^H9mEGPB{9VzGcVda`f_+;aZ5ief0h!-AWAI@aXUBY}aUpr5)uOq#
zu2*52pGF+OGbID4k{U(gC+z#dl^TT_)A~9`{a#T`$t_YGo%O69)>PMjveZf?5^Ucs
zw(dwPi-C>wQefF}iLR|n@+%6%;yi!49b9GjE<?C}n@mMHm);mIhA}3`sITm~LHwnB
zxZ8<pa`{{*TJ)!(e@tsiqz~FOXCY2+KsDt<-5_HDsfHauKb}kv;kT7%i)_N^Zehpb
zsqBy;(2;Ffg%_yDJAU-+VmO7|%E-X|hY9HeY!+T2G3g}=={M~trTFzlc2cq<4O|W8
zk|%6Yh$rm9Ia0j9bJGwlgf57VJgms2{Qec=-_PI#TDu4bTL2C*F*S3EG)84WTpWw+
zk9&PoShZ?E-q`~-P?{UOcJA$Ui^_JO-+p8If)jnv4<&TMx>0)51J3y+#?*++SGF`b
zvFy}jWC0V+NFEcrE4UDnLVd!-oo`9hu(PNY34)4|NZSok!&}RK8{XV$lc!0OG~!tk
z-NT8NS2R!#0sAk`?O>OQHNu$66hs%vb@P=MziTj{T+5W_c%wvQbG6eaKH^8v<E<lP
zBefS*dvILT6ukEl#r!Tgr-vwtSiA(7LrzhgEcQ)mGW`Y(ou3p;ZUy$}%48?Ocl-gF
zP&_pXY!?cMRs4X_>hI*^#>=X%v>^#$%H2ui12O_w$Tv~#Qfz)xF*8SbS5Ki?Cp_=4
zSl$kRO%XT1k|>SfdGH_Up>K)SxO=6TS-~SfIREmp6v$#;dQLLd8B`R<tHW!4jN1OF
z#=67l=2Mp+#ry!nUmA1}kpEo%395N+=UIo<hL}8IQ=ix87zY_iEV$sIN$*WcaHela
z_On`ww*P16X_&sX)DR}*pnw?EUI9kowkIQ?_CS*R`&%Tp?=fK=Esz>EXh`9`!*I#s
zb@E~)kzh*J=|I+rBx&q4`YHpfBS*^Kb>k^KxX5&Q9!I0*K$wPzAfW$VCfyz9opAG~
zg>W(zl1A)?HU>}PIV?fC()DUyd)iM9P&F)a=$6f`R7~C^W$OLG)^8(pjAnrviu+Ii
zfk((i+USpReytyHAeDy$>~%IObenBeI1N;?XPmzU?hO_1SF;CM9`5AcTLuRIwk7rn
zr1(MU#Y=Wb;_)TO^1*BkYy000Rx2{Jx)sN*z+?2w#F>PFSN$ohjGv5_I}xbZK17;%
zyhAwjmomVN-PdMj7Ay!TO$ZV`)t({-#O<%{J6008;vJ^E#`9U?Gvy>$9ieD3sTKJx
z)?QIqaA8hsGi(YwKTG`Fy%)SDI$^@6>oU3M%BZHaffgp*pP;H`72x9|0eD#sKRnFZ
zKA=Xb9kHy1W)xAoAR{>qn7Tt0@HyFCmO0^<Z)%oCJvKi6b&1jUE9uIOoNfXDOdCJY
z_Zf9F>9P--3mU@hFU5J?lmtE-2BOu#ugtqAym#Zl#3&{_H4wUuD*>R||BtPA3eGfo
zv%Zs#ZQHhO+qP}viEZ2I*tTukb~^0%?fK7FGxb%?*{-?|_rbmPx~}zG&8Q{Kt%A83
z$<aH#@d*2u!@bm<?k{cWBF*O!UQ3Fb^BLYCu6R{>3)PwX<NK8mVXJ<pK=ps-F~Qd|
zE#J*-&H6RY7h_K#Aear_@?&+x{IiI(S5WUo9mK3xAfGK)&@{$>*uEbGXlqmi7@;^X
zE(}U*jc9)^ygmG04CpT;?EnAMcL!r@Xg22bKWLCBpv;W_fusNTMj9t_?`z$?ooPYM
zIh~Dtrj&OA7_;Od{<V<>LX4=+_UaH(>Fx9P`&X3Mrt?tG`cy$6x|Uisb&>(!xuu(?
zSk>Nbk;We8PqA7zb+TICu@S1Bh8y7Zh(4yOx{}s~skxdonphvL19<$7|9Rtb$nzBq
zpnnD~%>h%>2jJc9@sa$;01Q{lqo^9=%baNy!#vx#Y|#25@l*`gntCRWZ4-KV^iD37
zs^9W64oYj`rsk?nX3l_J`i^-hPyT?~%b-*v3=7j*;*5;5de|F#2Cq7vFTb!2ShaSu
ztl_US#F-W$VXLdMsF}b|VJbu+{cNniS$j;fyVtWc0l0lV{&C<8mCFg$Qop~=lTx;3
z8G0j~3#v$EmGp`h-O&1)v*yG+^=*&=Uyx(vY*-$Xf3`+U#)e4qHc*&ea2B}C8#~q~
zc|z$>*hQN0@rbtqw!n~91LTx7&C;H){nJ<O!?rlA2Q^>3Nk3M~ps|cKB_`pW@^EpL
zj+-qRfRclY<5G{VS6XUpTaiquq3b5M<csf9!Yc+45dc<c4=m!gndaJTfd6YkdR03t
zN^}KZM~=Et>hEjkhTY+R%g6*N6&5CzZJJ-$X__ghMvXIH{AhX-THq1~0itJG_QC3$
zVZQoD)1n+EeQ?R&F^4e@IcFLWF<y7_ZLruwfKC17W$LlgrN%rnjl5DPY;rMZgc^&U
z%4bWoDV-0>;`+@rea|{E27~cpB>|*gt=Gr%w%0fIi+1SbZ_jujQJgohNb^yV5(X9n
zOoiMNDGrdG4QGFbytDfQSlLB|FpWrB(j%l8Mu1X3I*vMkCb9W>dbOPr?C}z0gBO7z
z04Y0l2-#NKpYLX4E?pRr>y{3ME?&D<xDbb)*O2jqAt_o#`{^hyarf$%7Q57^wv6(>
z+BcXg@TuGkD;z_@9IJZgFusoR`Y`HlBNM<3`uOTQwQ)H#^0<AKq@9I~$4yuU=Hg5a
z$|Vpwq`atXp!Y_A+c|SWV2FC@pn2%f0eQiWu{_APsZ2yi;XxEIOV#&I3FbyZC^)=E
z*Ms3CebXR(?)IShvz|Sc5y_m>i-&UE`dcDY<4CTF;rKRr-AsXeIk3imz?V6sn&FB;
z)hMI}+bOk1<U4a@vK1;xElZTb1IO4JHkDRft3?%uY7(N8nWw9Mc#e%V1@6|G0sKa+
zwm~ZXsroO*6Rnn78015~gvPhzB9d~8`*&A%xR!+L#>=(m+i~btqu@@uc0<f{wI&BE
zxfWo`eQq#szm0<S4MFA)R$ME*4SRkbA~R8YU^*3`1S^6m)d-lDRk<@3DC7LkMG&1V
z(|wyV8K5v6z<ZPM)uP$GNy=M70KK}(JyX&eLROAFLSvV)*X^~KA;xj*xZ-X1hv{xn
zlE<iih2b{5HibPoo`VO!bEOOml6$HoUHe2!a#B;NguQ_-#H&@*+`*!~N&OsW7Tfl2
z`&&-@MYFj5aN$)Y(@Wt#KRD9unF&cFlRlr3sb2+^&_>v;9e{ba!knDJ0qq%OKY@3@
zN?W>9QoPE!K$5()H(Y?H6+vxHTp#8+p{WI06^ONP6GoV!kl*Rtm70gH<&K3h=`e3q
zy|bl4;zlLM8c2Ns)3s^q@1iO8tRVrk#qF08L7)~si-bA5ScIZ@?-*9EZd=UcUZzr`
zJPVgjx7}($a($E}4g=3!fFf|B8e)iNI=UMt)X8;fRl@XdqxO5j^D#Fwuu$GJq^y!v
zb*1wF5)hR>zg<7m%@!iqa<YjIvIA@|Y<ZQcEDTLn7?&njdCn6)JWV5%pid^5;Xod>
zg+dkv%3qXPi7||2yOqBQEs`AC8FFcqlvvKPcpODRqq4EtqLu(901egRDU|2PjhAMj
zp2`-E?c9*d>o4;1UOEWN6wuv@tTJA9UL<`<Bad;<Xc`D7?433@(30ZOnvzW!gNo<!
z*z~5_BqS!LVJlv2iBk2`u)WiX&Ew`pN4Ha%C28Ngl0SIQOGRA0%f7c03s<LSu6H=<
zw-5D%rMKeCy@ma%089_I7Q6X~MJ$@jp0X<Y^7~^XeA7+Y<*^bgM%pb`$!x91r{PBo
zo<9i0t9GX(**MR`$|_Z3Sm^LIOTssOHiC68hYgZovtgcDVB&-q&d0w)Ze~xcnYS4e
z?_e8yWN;c;;#inL!0l=wCVs;vv^x<;?ey3|>IrlTf`i`N02pdMk#2>_JN)p^<h#LC
zT?r)E7Nnt_tA5GMuYQ|mfLo+ig>s@%jK59x;4wf5=I{aDfNTlQjo}{Mb7F1BY75DW
zWNBDMmPeFx_9r}BXZh&q1Px54PhW;s`0Z6iZaSyP!76*XH`fZA9eOzlbM~={!#cBZ
z<gF8|Cpwao0U^tc$cTFyzp#B~a)zOO*m?KY`jbLi-}5B$1wL=Zdy~2!ews4cn+!$F
z2KEac!$d;jQ5ERzlC3W`QGCCSP#&~qNqIG;7N+a88O150NB#P3x(>9W6MImPGkwy%
zjih+FAPh~CyaVZ!NBWfKP4srUL%)AXeT0?JUYTSw0x;FK({;k?ar_%MjtwL0tunZ{
zd4@(~np$Hszk~?kD3KCKJ<?MWx<Jg%TRmhtlxFA=^sDpc;5W{qLLKOEKFT{5fLC5J
zl3kX@tl+%<=A`@5ALl!t2ObK)<-kOIq`N7XeBQ`kMrDv%<~AbFm5|0vMsZ}AY6uhj
zE4qat0BL4}{cpR8CEYI$)at(nT+AHdx_BUMx_F?3PQV;(|32(PVX*ukcwi>B|BSnm
z>#acn|Fx1gS2sVCRhc?nCP_xow_2fUJGNUfIq(8ZVoYXAr3{n6!}I-oQP(A;WFoKJ
zAt_k=$NK5$%Dj=gHfiP6Bk~*Y<@kEC)u>%M!p`8nae44nIjOFu?2uff%g0(}u-0Yg
z-Y2zGtqXAAxstp3W{j_O8}iGm&A;RDQmX~Xtw|mo+8x1*+?C?kVr=j82Xc{|?h$-{
zpKST|NiW^>(t357woEkIbyk_Sg_nI)YrnGcZPg5Ub~L}(^2~!D<%imGFY!a*lz24W
zfYn(0okuq4BfLB^9%iVI4^2RyZM4g4-}QQRy`ARYl^oe&ZV}l!b*gFGO*!{%CI|we
zjLR$z!IEtK?)@OMM`6j=flGY9Jo^giY(4ss6fBMe=J7AyDT`m4o*^(?-yNhfecd|F
zV@CZ(<DO6_sP(+MA+Y;=0d4Q86oAs92W40Ito%vogJqMxGQurbip|CI$XWLSrRxPq
z7WV43)xNv04MY=<df%0qMLoU#Wo3Y=Rs$_ZZ0@YB*x=TOT;lDTz(2THcPf-P>Hamv
zr;c8jS$x<e73|{Inx7m6{J=Y-W;b*M99{U=(sRJaeva(jQ9Go5y1%ol9p6>va!4tZ
zTb4oDRhQfqw~ZuG6K!#u{mlM5ea<#nH<5N?a@xc~=#b$X?OHO*fsMmAp&NiB*Hz(b
zyb>Z$3dNJy76>jkvu<yJuGfpkPTeNjrp4EmxZ}|b_b=;Kc5&6tmmxTsmB+BuD;88@
z$WP1}?Qe(&a#-#digJhi08A~e>Si4!ocakB#MnIIE<b2uc6!)%AUAph!JoU2g(=S;
zeagU;3@C<cAdFBd+;n`impH(O?nx_ys{O~OxJHD3E3FUq{qGOLW_*nVn<NSAP{<GI
zwl3)mfkZKuw=;uXf1xv|MVU4?c#$Az3$TsWr9j5N`f}+WY~SbO3**MtK=p+)i>b5J
z)qc{#)>{+Ynfo72ya~KkGYaowh?^VG6Rz|*Kfk&v*J3%kfsV&y>MH^3;@tVSQ$fp?
z&7J;7-F&b&k2m&KDHg5v?{T;fgB2(0hqT_ER=o<VwLDsN_!u^)&-nVe%O;D|3Bwi0
z08T8lxF<5yfFVvj<-Nz@^!H=QQCtE`Q0owYRxJ7RZpE@t7`iuS=opf@u(dvE*K<>o
z-D8xs7}%bvSxdnN*#uxUl3q7gHUU@@?u&YBNlF`=f1WVf26vWdPB=wh%UlUzCmW)u
zcj<>Qiu8SIfA+T-h4i->ZrU_~KP5wvWqIKO&E8xQMk1TH6vb#L$NA}%iGa^OIX%iV
zHeXtOl*XE+_eSXz1QD);NEpR};G8jh_(JvoJ>LK+c|-2Q=o}z%CiSXsAT;m2)d);n
zcl2;>P_EN9w>&FGUz-K|b$wH=YVEDSU8zz-EnT{s^zFeKn+G`$t0Yv%2;9EK(~=;T
za?UAYFR7U{$n^p)C;fNvl%Mm=_^*=QzdXqEM)i3i7x0ceBlK#^0UNul?n_Hi9f-^L
z;Ljt=&ABnt_!NNna%u*#l4yU6Bo+xn-pK>3LNuhpRkTKHkJzDh%~=eP%pFeRDw9Tx
z0d+xr>)W|W!(bkFAqy*B^2ld>4$e7U9EaYaoO4!Ic^rqm_?tz(X~yz7ryk?xh!xRi
z;wKWpbNHr$ju0dN_c0#NDC^snIpC74hM*{d{uDX<x;OySb&S>8-;iCD&qBs}RIdZe
zdmQ0X2|^&nvwdMo`8CG$nP11sGJT%xn{&q~;9N3^H6)f)<1t8U*Dvj%KUL$!d|7EE
zscORQn1QYydwkn|-7;uW<}3c$q;Hk$H*Fyu2xnae00nip*u=k)%5`}ekKL1PW?4#A
zdOckqTF!tVorJs<tcH|s)lP+zZHZvcN((|XGBoPXOf^Uj#I2#XU=aSLDP2r#OyP=M
zybxBAYN6qjQm8wR$LifEnCVJi<)f-H!lL(U>Ui*I^0&4>j0r}~i)EgSc?_3#f&nC(
z(%WALTLhuvZBVAuOt8ByMagPlB#60d*uYKAq-OxI=+g;vjgn}qL8Y*w)}F-vrA`E2
zvbZx=-i{)xpaB9>|F9SSq&~#r0Zbg%L_H4l-0n**@F6Wi6gY_BvITZC+2}S$Ta0V)
z;MCJ?QD`TYNupSkE!*b4XkN}LLFZ&N=sYkaiZOu2Nlgi#tY^zB0egIN*{x_?|7(_u
zW)fg$upwWb{+Ex`BY)hi4fXn<ZO>`=uH)?JFK{7x@YX_%s5Z$Qdu3^O$l<}d#|<fQ
zUKiEa0-%ZX>4o0gTK{V!a8SVuAWxWPha^tKtncY?QVgV!B5ex0Zi+f12uf&UMYT2L
zU_;~hBoOcx$wZku_|6^Lp(V5p3}aE191TDZMswJKI98!FI!a@epeoB_d8TE5xbP*1
zJt~L`M<tC6X$cdvlXm}9(vosO*{DLioCpEbuX;7w5$`1QU5*;aDwB%lUNAb$%pj^)
zVw|%%m56^KhJ!LrH2?aBWNt+8P@f5w88*h5i*Hhp6bS8|(+gCvFq9aPAL-hG3Lro&
zl+a_M?tZ;g35#$A!Z@k>NUUiDcj0iu3L5w-R37K+;DhN7s8DW{?tbZFA&xLPEvFA!
z6XnZjFGq;#%4BUT8TkWkZf}!HLH_jzK-D~|JCDTCH5dGC*9IM?C?%GYRORl6ax~!U
z>cki}towB9_eyE(C!TiXKG)f?zcN74%h>$iLNA?5AbK6T*>I|-1^z)OCs?thP-eP?
zSiuWT0`qSML$7dUP;6saR#C2Dbx8-Q;;vv5IC_gkQN>A*lPp|LqOHVk@+#^EPC3W|
z-y-482Vv@=h>eHZl)Lu0<}^KVX~z^?u&d=qEg6}DGn*GT7Je=kxCDyWk|97L#cw5e
zVw);Hn`dy;3CJLNnU^jirSI&;=?M-bln-)TyGd1arpu5zmeDmeWE({@t_hKl#8jS<
zT{Et(UdBW1$;HOrVBgyw<PCpQV*&owEL(9%h2!{QbCX&%>8U^<S=d1IS^4$I5n4gP
zd-;y<v~mdtddQUQnC%g@1x>(|V3yig`Pd4RGNkEA7!7_`hS~f%5}Gi>O$9&Lon<I{
zU}A(rgt))6r5jN@==?qW)P;NqH(K;TX}sK(=3FB5>A)!VhMTsmm+IIKuTbQ3!bC*(
zyh$K(o*y$R;<s(ac&I$Nx<K8zm^eK#M<ZEN1~@RKjW>yvsr>JYj($LRif;tf$>*(T
zf++mYV@Sr*N|pE{=Q4}r;~~|-eQhj_t4b{mXvOM17#?DMM<D0hEt^fBxIF5t=Jto>
zGtMdymFbCfwuar-8SzC0=(>!4{@miiy#{IvTm8V>v@s5+P9#=?XhM%COUeWJSt}&=
z+hcHu#4p7_8hp!`_;Nr*)DBXBC%Usq{P=CE&wEn+hS|k?E?DJ$k(#`Gz<o9h1r%Uf
zMoOw|g_UYE&^8MT{s1nGsei@(kvVfO^23N1e}#r3E8>LA0)E<#x75_{?49XB=fsS!
z2(VDxx<jc9i!7g{;}mp1XcpkgSoe4C3L^-cT*LG@(Py+IlM`@2jJ_*GNGQ^edu9F8
z+ADO`$TK*f;^*w0aH3NBX)Yv;D&bwkxWOAyl5vB%F4#X!Xry!;gZsC5HL_|>Id|rb
z1Onz?oq>C8-e7p-WsTn-Kicv60bNNrTir3})FD-@AyvG-#y>}@Bf1F*d@zttUL7Fy
z7G18f5L~v>c{_mi(^$h9b9_j~mj|^0`9Hr?=cmx3y<_XiL6M8vn1XOli1&^xged$+
zmrhuV(luD6MJu>|t3=vP^sUc585%o>pdbZ)I7Lw9LH*P)vigl@1Afac3Ji(ej{_&`
zv>5e(IsIlohi{0$Kdkq?qP3r-^lIE%wG!e$CUH+0l(GT<E~YF~?VGF#GIib#Ic40f
zoqS@jAN@LZ7MgW}VFrF-9(SoGA7nii>%A{N2QB}dpnuSF#^EaeQSLBJT3W~;Mx>%V
z17;gJBs^-GiySlqy2rN4p9y~(-6?uA+WUJB*GOKYNfJD7iL|mWrWz;Z7fmc(O)p&&
zfssrLekBW70d`&mQCzHw0v%o;;KDemi;$v?Gqq}Y+Q2@(vbc%~_}6d*rsar7k&gSd
zc&WVvp%wA`1eEMc;tm|PHx}Pua8CQ3<<>2y-Cn7_RzkPJPSqThp2{Kn?^pN*!A^aY
z|9`ldtpEEcf%{+f76;RRUL}B3ExF{acBGyUjcGeBwX`$?WUh`I`;FR?q~%d}HFzm_
z#Lz^=-;!7X=Dcsu7i&8O<xn@$^dtofKsd3tYX}*iw~pz1exAu!cXj!DpP#+n?KDx(
zL*^>Ge*WDylU?KvYhFDaAKiT)67IbQsdDCDN9PLW=ht7AyQ`Tv_c{QxKjlFReLV&{
z7#j|KfF%EVrp^OG?cHw{2KLryA#gYK)ZEzk+ODV3DM1f|){O<vC9x~L0;XIh{@(Z9
z$A1fde!ROsSC9FYELBDNYtQO2?q!c;S?A&zZwVF_0=M2i-adVss$T3F$u|b3!{Z2A
zj`fS%@k8e=Rz-!e_ssxep3WgAIh;|Wm97Eom8UfJy+@3ph1_6yrXl}@?!X?b9hAEu
zUfndTj&l6*q1h*5PM<ChJzPJ*K4O&#E``@@ZkAe(W&))lB}30K2(}~<O<a0?bpVu2
zBNIboJiDfd`8)hWaqcQ#{s%@5n%4W{$2&RpC9i?T35Z<)oB*I>HvU?dX0xrKN&1qb
zEB5HClDds?#+x-86;B9!(5d^$$7;&~+ffg7B{_C1QLOS{=Kh3R1KnXiUC#V7`Y8`@
zJHUxg_+ot2VbF%|W1ZE*)LG5Ji@;S^t!LLBW~1L-Z>KJ*XgdMM$+1~egaAkO2U%CL
zr-yaim-}yn#||JGl0Jo%ADf#zN#fH>>jBTNiFR$IHPOLeyg*}w{p0B(-ja2L0^};+
z^@{a@hk!u~X~7fepldgGyPz=Ms;{2JinJZ&iT;aqJTLs3E5e=^`=3dT_^M`tS<EUA
z={aW=QmgfE<dqjz1t*-ToS@LqM8wD_h+o7fW8)FIxDUYEx)z->5zfA`6S#okkAY_l
z>CI!gxXUfWLS|dOp+vaRNYbV45Qw;U3Yl3<qm&s_)r)uOX-46XZQS|I=yHi2k@G!B
z^hA{7B=pU)aBJ3YGhhF3VQXlzvB;bvLlnXL8vzI`uU_V6Zcw1)WITHSp0gd(w1P!=
z(=7(~pSN;~?(PyRNgEj7^9fwOc|WX9eh4&;0nY>^rl#S%-ORxccj@A|O!k&nU0bb@
zE$WFPv`mrmr&#G6ir55kZM&?tjk~w|dQkp6qw5<vptCtJh<0}Z{=ZO(J(SW>bzcMG
z;E<~#ox;0LzaqwZxXN2k7-V{g?P~c+Uq+>_ivW+2x`Errz&NAPcdqQp3E(6F9v_N|
zPm1r;HEWBL5q)Q1Q*?e$Iwdve&hyI{f^7`fr0Xz5LqZ*Nn12ZwgUtoFDQv>(qu3?y
z6S29z_GhHR@}UB1#;Gdc?a#b1ES6r?2RtO%9TSq@<EL)U!3^XS%k7kJlNAX;&Us$=
z-T=$DVjx~akh78!u}x*wNT7SOlWTrDrb#JcO}n=GXz2z<@8o48R88H%Kk(EYe^I%s
z@){m9!7m}Wd@N-n|HU~mwSNUhBJD$;v9)kf>VT4YH~IR)tl6@<F@+qCQ*Oyq!Zc~f
zxX&#CpLCR3t{-DC1J;YBg%`ExRTc+s-2v%>+9RY9!)O_aim%5-Jc#G@-o=3m&Y(x4
zq2&yi@#sqIR<P=1bBumZN_5_6Ef@6=;9spq<@P!un5}>0O5<_imJ=q$pm#j6>YY&+
ze<?|0$|nZdI?2*8$sGo|4#I>oX)05I@K^-JV3w5i_eh%`LsS+Pcb@OGXSlnH4g>Oe
zZ;I$xxJ9uQJyO>)%BPHf*)&s?VlD-jV>yWJZXuS4CE#wSE(*bLMHPCwR<b^#(csP`
zsQv~s3)N@6-&t8?s%N^#M~do;a}AMz(W%e>B=Qlxh{n*=;v+41HYi!6I_AGe_Y)Rm
z$2n?NohNs|Ne%FTpP)~Sg&6Uqf&et9LFs-gLL5;`QZ?WScHgEAe3XP*x2rc<KlWz^
zsI7*x*_9P_n)V@0!zc?MM4{rqleM)W^Rt@w>7y=@s%ef;K9@K-akY&?a&ODTCj-Gf
ztMNXxwM|}I3VQU`qlnux)zwP#O;zS74R=<OVpA7fLuL5#6W0?#rtb{IrvqFP=vx(1
zEg72fGI~w2V8-)%o_M`pu9Nmq(e7`Qep~&C2qBE_?K2wuX{{^JY}_v^(7Zie24@82
z1m5UBSEcDgSnNxXvtzTMVk_O<V4QEh%xpDQJ<lJGx9hxCIzqb&8o#LH!tLgh42d$$
zt58%-8%eIsIqnP8S~y5^R0GHnZMRLFkwu=wq@~e^{u6XD>}T$I0zzZ0h4Wt-#&uN6
z*9+Q*fVbIl-wf24PsR>(7VX}b`4~JslNV;*OJXS5fcw2zxn9DP<6gcUnDf~qi#7T2
zL@++XV(Ri+BJ$M01<VB+-EFovm!dk;wVbuYaaUDKxTbSo;^Q}Sc>tN%y<{_s+>n*{
ztK@;8fj%QP=-T$SO2xc;*aV~#$QH|jgiht1dc?!?EFsHq77p7?bitBUxKDZ!mP+l`
z0HZQe4peEcRa2jzFGD7{T&;^@kt*(&KT?T>p?;E<i@a`HWMly|eod4zkQP?P7;cz@
z34KID1>pqM<<f=FC%_D|@GFPE6%i<d3~Sh;yFS=4GmR9IRmeXz^ppkDx{0!=ViV$m
zW(T?Zrj%>v;lQs}xrmUZJVBnhV=~J>RhKG)ZciSANHMy(0vOE8h=fymeD9c5dE1LO
zpzzP1Asp;4p7wy@IP$(e#)$&as>PGuN&a~*D4MLidCP~bJ3y^W@@4h()cZ4P2og}y
z#^g<aM!c{xFjN)d8}pz+Utv^&;)$rk8F-ZnnJvj&Ig}>-0REUigT4(?0F0=m_@&hm
z8nvwKIld^VZG-7O_yP2)LPIhlcI+0fF=c*FSfpjWJA=oebaMFf;z@g!KIQMS=2Dk#
z{u|lADh3ISZGeQ<VjJxFf||q~wv{u>6u96G8q$F}oLu3TC~NQM-_&y1I?<M1<S(zH
z?K*;Rd=!e}Rz(J52Zg^T3aI1ddk60ocl_`RWvik7*Bl+E_ghK4f=cD{!UnfaT)gLF
z@0n6RM@K4p+uMa98ow{c5n+IfQ@1N!dMM5=kVHHgp8;i%_9wS<GB8kLDtPpGLEz5m
zD@0t3%@IsDJF#?KTAgu3moTSSgaTHlgc^czf5<JD_|~S(tm;#E4av$awB2<}utmaC
z$zv<4>M|7hs8lrt#@D4@CHIhp6x3`H`IGm0Pwp#Y(RH`!F{pZ-?Ye~sb=!Mq#<`DS
z?jA``J^{iYibB#oU*vv%Afm%ZL|8K1h4Ny5xUGoZnQXroaY($?33O%qz6JJ?AU;4m
zw_6c5x{Mkc!D$f3F(&Q)WVPBwg92?C!>3o~<T%PNIp|lF|8N^?{g!RqoZbEHXr%Co
zFh@6RqY4%u8@~uDOWzYedylT5?_F~|nxH>*^$WnJc!)ycy9PPPaP@oQJ1IzUEEM-5
z$}q)KY2F=D6<49jqLT{lE!z*Xlq}OMoe!bcKtgI#LLiUTF0}YHeRu)JpXSd4L5@vQ
z-#yAu!Nu1z@tnrObJq74rN=t-9j7b3HxA{o0Gp=k6I9Om_xe_O4@)yo;2IaYoI{ih
z6uau%Eyb5d(CsJXkBw6?0BuGzV^wTtyQX!fvj**L>LzxvYtG>t#F8aS@c)puZ2v25
z|3}%fFf#sk1U^-3D{f2tKZSfoAdUe@Lp&9JE-{>4(QXOr;EjTf2z;h<{~D)Nvq~tE
z>-JPIq>n?I<nxAWd(|{Fa=BNc+!4AgUw5h`0ex?OFR$9XI--y66V4`XkG^VU-OW@(
zmD29`*}L+CiKhpT%?ZY&@IU{4-}rOxT`zIgw|tNRqMEQ{nGr5=BI17A*}WQuXaD$r
zuB|yN4LNUEwkaer(o1VSA1&-&Co{v1;#8j~)Uwn-cINp9i+_YS{Rzf#NTT*cj>HLz
z<Vo&lYwa4?PjqXDg+DKO4X%+`^C*FX$GthsEmK(<`?%A;@jo4aic7yjY=H704Q!)0
z>J&-?ur?lkEwi1L_(Tm#&37i_%ygsRvCa<C(i|6j@zJW14V>5P>}3X~$X2m(YT^3w
zE*e!8U=3#bA(aj&NYKt6-Q67VB3ZZJ3jKyVb+tD#x_ww(E}2xyYI!bp;s$hMrfTQj
znuO~>!izZ;d8zPlSGQKY>==HXW%CX#6jHzg#(c~JtK)r&VM-7X6gYVrr9!7nl}yBO
zRdFapbolhZbzU<;xj(#vTh5^uE_h>`G7E(9Gu)L&s`jUM7veh$gXb<bRoeErrA&J8
z#61S#8j$|E%Pw9Px4J$~HLNtpj+lmkA-CLWR(vhJr8o(zHS=bK9*;D7Iv&!2t!!!l
z{BNhA!tT^iPx(86ix_|FY_AiSEg!uN)>GcuQ_afSLJ9uhM5R!)PNOH3Hx~B`QP0Lh
z0mw8V(xZi#2eU4tL%Sm800PdAyJS~-;h3z&%Bu_8&UdNPC;j(0UoVecHe!ymrn{A^
zYv(n%sWrYk?=$XAN-9YH<D-QJriDnrxws$_q^Cyo`9=d{4uY5+1W9{WWSB~OgW?@@
z`apRTIE!V<sF5yxJ*^t84+Jh!Jdl6l*kn3zXU^>5y+K&RnjMV>6v!Cxr@CW+*xYQQ
z8OAGiygb<uCz}HC^+;1qxKE_%Wj*<}N1mzKtlCgV@J(=DOPO(>5Zgkr{WcOnPA*Dh
zUv6!`^Y2GGW&CcBe|&xh=(S&LF&x$dV^HgFgVjj>u;M?keVBfu=~ZL9RlrTAR{bSW
zL*3N{5O+WMbR=U+n#XUv$7uyLgAHa?AG`UoSN`cY@IfzsznC{g+CVF8W1oS+oOB6^
zxK=S<efrB;^}-Jq&!w>Rlwz>~9<a6Q*!g;USxU}IB^{JOO0dgSH+u~@gO*MzYi79B
zpBrq+R$N!Zz;DZL`5CWNXMPD_u9{c;NMRf5;julpR=lmQ7S83dRTd60NkD=`h)=8$
zg2o03It5I`u4D?NAv}*(J_3`Lnxk<svz+NDjIWokDs)I;Mwh`KwFIvK^X<8BalLDY
zy*YRZ<nowLGNiw{c5bgCtYrm5D*|R5u2P$XA?c=2N_O1WbSSqJ6WEB^*yps`HUtC&
zd?l;VyQtl_*a3&Z)&Wc7C4ercl$g$^hOP9TByut(T)uqaQ9@;)65`rsqX?NCqcc|M
z2!&7ewk5)Tph%%^emGdbuL@7&EZm^Av!+fOWT_+n4}nUfRgSp|B&ft+<A(WoN6EDg
z$KkngDqJtv31N;*`BW3o5dDyH_-tHj6i74JQ?bT}9c)YH>;yTdwxy_{$kom-Va!0T
zQgdwP6A8FS>fT`&_L#BA2nTp6@C3ZfjlXP4ZGrG)CD9)YkP2x4X!wyWc9b$D%_8z@
zKB6*~qBsLHRM>dqhQ`Y9!>8(<WC3NXo-eDUd2p!z?$DQvcx@%`7CFWSef+KtQT~xg
z=f?KKC;<&W)Or%MNO|?HO%ie(1<dR>KRl&jN}~(R$E$2Au^eZDmh-UqEO3xuzj(D(
zf)a~NrWC2@95E(9Z;$}PhP~&-Y`i)QoH(kj+W$mfDhmdvOt?E@k~*Dyfmu)aLl`FY
zkMiM!76<5pr%^la%1rY5)_Jh5`$=xWPKr(u?_Y?$8<ODW?fSD86|-NgU&z%uqL;t)
zwDu`_C!s@(uup>P2&VQOQ4^9mStVoDq=C<GFTK<$u4Hoo%D(QLi90wia&c*to)o-j
zE|e!p;TN4DD-UdAujb|YBBasJo`Z)h(1P%QQ!1!3_jk4Rz5D>__FqKx3Hj!|o0o#4
zyw%+6>P_D-ob@X>lkvWS%~UK8b&eIDSerN$?G*OhT}yO-Lbyy3u#R?!5V%43HkcFK
z2#bls`sP;wrjxYLyITX;b^T{2t(K@yHu));fs{3pir8pISYs?O+#Q>*%Gh>FznP(c
z)gBG`DV~H7l~#(ohz@&9kr{VHF!3s%ow|BODoViADIAMZfh7Mft{$@hwWV0UdgF=~
z8LA&TyCFR_oLIdm*$`H;QWk`)DbInDtjj3f<cn?tI3MSqDYvcl@GUk4DU1n%)1aW^
zPN;7n0}EZY4aB*2u*<}teDw_2TS2?-`ShN5Srnu=8ypiGqNf<=5NwW!+UiFhqNtIT
z1*K0v2?OiKBp%_xREO5HH7Z&71tT(K4S{U3P!ey2{!ypWJV@2B)V1C%dV=|*y%_gJ
z*%{>kM<d=PCkrhYca$PeKY{j$`jC)A^!Dt!d|u)OFp)(#0rYh3W%`0*iH=lrlZ=s%
zw)AuG2ShWb(aJJS9DW=qnHbXb#IlvP?7Tvjqt#CIg(ZHhn}txbS0v-ALo^(=*cMV5
z2Eper>~xd$lt*HTNlL<Zrehr$S1%$0UW598r0@HXY9toW=-wE}e9jx*`wY!<if7|T
zqjiIc4s?q0*rYQJ7cr*tM2W<b%WBHMf4~n$fF$j7A^lz4-7n5gqY!51Se6S71FC$J
zpe~asO(r$-t=CD)?<5^imJH+ls<!BuugSZWP0bn}dTE7`UQcTzy_J3gR)0>z5eg#!
zyH6Cu71wwYvMuycCi_WEN<W>L6p<y4?@2N|mVb(%vVt<X<F5y#Gr!D>y1FnUs}V$x
zAmVW&0ivo1y7f`yv{z=RX!vDZCQdU7J-L32YczyVdt`Edci#Q+Str9!$+;n8HTlv`
zyh>iu<;ja@rqIx`2I6UYC6ivY%N95QeJ;-$txS^=X}}+3ZAd@$={Do2Nml&<=f!14
zJC@Hyg1<CDtWtVl>7asaPXxE!%Vuyi+XZ@jAa@qedL^6ueTDoZrYoZ~V>*;0rf{(W
zj@uoa8pJj3TTkf1i+K>3)u$enCh+-d_*czTDRRA+G0sT3O-bTwsDq1C8}(=b2GGKt
zOU#~h5907rb3v;)qAv$S0w*T(uf6WI+R(F|$G^Tyc*ZC03l&hNji77Yhx7Rm3s;#W
zrEts*K+*W}(6l$7Yg<cbArBQGnGY&X-kB@p;_cg%1)`Y~n9je2{9I`#NxbIBgsI8I
zxbnMc(d!mC4+tB1I?iH2hQ;0hXiQFuWYU1+9@O)ZU;L*WA0?AN_kX1tElNrI22@31
zdNfw&V5k=)=7PCULdc}<1LL$Nm^#QSSPd;Pb1g-V5X7k+h}_grKl{mraP?poIsc}8
zC!f=k*`RjUj%{UvcEUW5N(vz!kXw2mWsImE@4H+U!G9;4Bsh%P%|?g;T8yjJkVp5$
zvXFmTQ`sKt2gBne?Tabjy<MnG58MAf(v&tQSyJ%+c?)9nL<)MYt$8~MJvukTt74@4
zp0x6!a_2}<MN!O-fR4aTH0ewl6Uqu2duoAZMAik7-HZ($ia&G(f;&7G9<S(--X)#?
z5ju{C;oSF`O2tEo=7lZ>JZuzG6amc(LSw9TEYn(aO7kU+N@0dJ_};^icK*#5LC(?L
zn%tr)Dac)0Gu4hZbjE8XW!G!6awpdRV^HsdD_)7_REN<`*VW6>*M1m!O}61X&jTLo
zUQ19R3_I1h?P)o)b|&xges-vw&qL?_Xsp#D13o1~wfWdFI1&A<l&sq8j4Hfq8rj{^
zZKxzRv!dB6OWn2=(huhnI*d=uo3Bq~iDspbU)&MVMf{@sX9lndgfjF@Vw|LElrtb{
zg)OdC%1);d-23(S5>P}ZgYw^CNjCQXs*&Ji{U5>4%*6Je{1;$Gd&?e&3)yc?<KF2k
zRBkmCgAkKYe&04g&ld5_$Q?qcVBD_G9c3cnIP>`KOjX96cA}O3Ru-Ib`})?pQIl8A
zO)~|uV_$S#Oi6b{siM7T=2*0NDW)|^W})^PIWICwW~KI$Swh2U!ii8>H_|gFsA`~7
zh8~>S(iFo4xf?))!L{!;qV`MQ#B`P%0x8n6AlMQwT3Eaip7fwXSiHGCqu4FD&RKZu
zoJIYKqP?)6-Cc}CM+9#qLdsMY`fgs*Qq?zfZCDo^$%T0KB?M2Z8Ju5Z@SKZB`nPzX
zBfX~@<miB{afs{yXAnaLIIFcIE^!#;gwru=LJ}T|cp5-DZe*D<3LF+Ia^al2I&NiJ
zg(znj`>+7s96H*W%%<QKNTytld6_#RC08^<o?3ql`5u}%gnyZw>Rb^7d{x$RUDi;o
z#f-}MD}#Dunm~?g1O<olLa=zp)ugZv^5oAPm50Z9^a!@+WN5B2`sM+M`cjA!s`&u5
zP(p#qztwM4J!met>4M00eKOX@Kp+J}DpqQ2YrJ13x(flgg68>0YQ{dux|58YszA4l
z>Y#=yXy-9nB~fI`KwJ<wxye(g_aQ9(6)z?2w&Y4NsVMa|{a&*p?}4&`?no?w2Q|R7
zCJ}+qd8#Pl7-WJ@p$jt8?047D*><q^jI7uzsi*+L;)DrngUH~_Kr_r{)p9K0O=T5)
zULSVLUB4b{y+>c>OnUH{jps2jkDvHI$3MZRrrbOZ7JcBH=AL_>JzhKA{c^)I1zuZm
z=(ifRhcGzGO$tDmp945gloI)%#&bJO7PVY1GgTK|6*l^veX*m_s|Iocu@VMb6DOXF
z{_TLgpJV@TuYkTX8qn8AsK7&YB1%ZjFGj!ea)t>0mcO%ZVt$~ONbtU=PxZo&Z-0kA
zEAjjb1UHMlQo2W4aNL|<PHo4k9+CZiIS=%>%`p6Y$BPI~G5P_2Q>^9l_W%^*ra9av
zOOkj8&^#tf&tbML%Wy3m7Ch2%k;4@nPuTz+OBbdKnP<z02u*GVIfGY_g%(~W3j_^D
zUL@qBJ8+$ryVJjmo6xKFvp5(O^d4)!8XfP}FkWWv8}V2f7?pXnrVAyS=j`52m##L-
zb3Hc7eV*oCe0g~@{hfA#C_qPbp=AD=a}KpdjC)b78JWc?(ZOWVr-9zQe{0H)+@1o!
zRUo5wosFoETVR9+G)h8L1D|~b%EL4?ZjOCFNAh{WwywPUmv<ZWq6S!;5fVL>qhF7!
zj>n!od!a5-UqGI-wm}MZ6Bcy4OQUn0f7L(8GoinLQQZMo*ABSAwfpt6I1I2{7~T7_
zW$PQb#aM&XSXE>(Rq<HfplbOo=dc0p%!ejECp&Td+-9PiPR%~HJePa@+zzwW?3udk
zeFoA(%(e?CchAjL1z}xd5Cz`!P4!|J>MUyFzYchmpGJ*b?qy!oeSz7iuLqd#R*aZF
z|5}NJ!V6d_Sneg7C_*QQ*E!bj2rcYCP4)4A&powz)Cu5-&BcqrDGRPe>m>lDI>9=o
zA%t#cQJbzUqS&}fkUZ;^{>=}I256KA`K>^k6YU%4sk(uWe@@}fV^VHxU8<MT>g@&^
z^TA;595=UZtOtI>?t|<7f@3@-K-#1c)~jAI*}8{3w?cXHQb=;vm=#nIPqp^bMU;_C
z>Aq1%FooU9GyV|098tMs_n-tk*SmeU)|-vsO`>3U=EfVtTz1Wn(Z|93i>Yftwkk~y
z**+EVK_HyGofiq1PZM6YI%!rCGi|mWNXo(DyD5AtzV0L7Fy{z1C*k9(t%6MKAU28X
z_&_qX*b3GC&J&i^yR=G;#20~MztpteBEzWW%>?tU*JD|<gMSUrdV2<paq;6TQ0R9*
zrC%Tw`UOIfWau55njz<l73eUFe@Lz$(P5tQ8D+eVKkMKvG9#Zk{}s0i_N^TgtTIUD
zH!e_fU=OfV`+fB9)FMk!wNG2OX+djRp=p)@mi=JuRvzFd(+rF1=LZzH*Q9PZFc$^d
z_RjdBd7-duF+|ho+5&))zOG@U1oO7Md`gz+5oYfmyiED<9O>Nt8t_sHYrugwQkZ<5
zf$&Bcz+OUsD8`Nb)VIQe>|!up!FySWI@6p0Eh&Z0(wGGPBj;e-VMDfde5*A9p?*GF
zgxdvAs&gf{tU$etCR)9}qeoR;+c4Lkk_bILmk_I3qe{0c&jhF*j!I}ZP+AclIQwJC
zjtA)N)NlS0t0-Scwi|J1o|>Ro?2B+8zNGT-ac``S`ijYQMHOZKTIh5tH{`xh=Kv1}
zKDdzLaN$1u%6Y8KNkkyqi7V8>3o3PaCiN=AO}1O>ut+j4Q{IM=GTMbk!hMBCn(V|v
zu|$L6e8OtPY6fIn&$AJ2MCBs?;&rbKo1albpseLrLP$tR0F{B(w)gJ8(Mmg)>w4nc
z0Hs{=2hvOSVfFEyv#y$Cmp``5anwE_t*fol9en0FhKDbt@shYUZQopA4U(%xkjUIC
zTYB7h9swhcX-@uurzrk&+&EUP6f3i;(?hB5kohigMFwE<;R%l0Z`Zm>{}mBZV26au
zAyB<?zvO6%o!fS!7C+hFR-7wWm$CrltscKJ@+tZ9#G&=t7BXY0yg?<gPGA$)b>gWd
zERa$BDmdk~75;2VkcPIWgLi&W*<M@~75@cZFo#Qc%(k2N=tQlDb=ly;UzOlXlazq5
zcNSJ}91PGM2`XUV+L%-DMcb}=7jCRF6}b{Sq)4hTH9NFTWpE%`X0ICFxl_AVd_eNZ
z%$XZKOD|Ann%>+Sy}9$_b_7aK)+uGxWUU3XD&vH<-u6#MX$O`63a=dB-~4(Re-s)U
zIo;^A5LMdjN#09PI)<$An5+f>)l9jd|Hqm{aCJ5_`k%V1f0~2{3GEj%d-|XGe>_et
z85i8P|H5%zQ)SjH!k>L9$`jjBC(-c`Rj8I8;`ZY?LM&WQgJ&SgRZ;ix8F-ky2r=mW
z;Y!CVSg-`zF%SrdMb8#%_I#y&*qiFoz*w{Di8!jCeKd7?tuMc1oVyakQZq_9iS9E0
z6Te@(GNq<1F`%H7^BjA@A^%?K0s?3`?7w7~DuI1s23L~gLn2?P6IznDy!7w*<(46i
zRKH@v!&rd?UdGddKQnKl+^&lzg{E6*1#zYxi=Dbu%os%ock#Db$4np?jz{PF@pi}<
zE!{Ut?Y{y+<5alnzwSs{lSa%KTdhmf(LSF~-{bSjvlym$VNmb<RB`A`s{l?9xHvA%
zCbnMzTNCW7VX94ON*hT<c2&>iHBQA^ADKf#-JR%zBvo>5CNv=CUD@Iz*agX=Wftoj
z2fa%=^=V)wCQy)#iz9*FR#flD1Isu}QEsLv*iOlN^=!dxB(eN-RivIhx+hLiE*yhx
z#WPo<)nfxPW4*(!?jcew27q#non#jIRUHIAs{#9ealGi>r4mF`p$MZAL2}nTh)L|w
zCawdk>b)*kExqB+kqyHF0G|8YQHeS>yN<1IogjI|vv&~Q-wYy$YfxO4;SG4uQ(~yw
z8$nKc$)@G84W1_ROnFKKku|6+;Y|fxU7Ra%OD4J_6O<q>e{kn%en5b3=MAHq|M-wL
zw|5uIYPEjYleQZkYr^q`kWtT54YH}CPJ&QP?uxx#tU>7;iXR`QmFIe_xr};Y+$OM{
zbpQ3*JU*Y_(I+T_$6eO)pT7d&KUXdhhRtMVc$|WjO07(NmPM*a1<gARAXsBE0jyEb
z;j<H@W-b5VZXBC3#sKp9fJZ*vFG(9*+f*_o$Ev_(@I?a@mz(%_Eu0n<^Ev|Y3{J`_
zwboAVXnTGN5N?_i)8+eYAn!bNbajjp8CN8DI^eSG^-M4nLJ*d;U~o78s0T=)$tm<_
zYywBVXiiD>`&F8Rp1*)5U9EXn?FhW0isrmDXZhQwA<tbWEdZZZ{U|TGZ0^?9>+-l2
zm7#e_ETSH501BJVk_c7*!EUmdujp!jr16Ch9M;BRWWs0tQ(X}P(!|BZ6`LiHur!eW
zpcV@rE(;Znvb=<rNgSytjt-A(wmLVQFD_B%7*q2mG@N|{0yr$85^g-4p9dny1cX%F
zYA%s*h`|{H1|aWx9pbl5l+F?-(&iE+5-0q)`}S6N93<Dv8QuXrGktASat^+Q73K<?
zXw!I&)>dF}3i|C5bOGg1QH+K8WK8p-6(UF>%d7eZg;qAKxad+{SH&Nl9nBR?PIy&L
z=n)G*7-gvDIQcw^FvsA5LcJ!L6KdK?#WVH{aQS$!Jb-2^x?zDQPBlhke%wM5**ez}
ziGdc4ce&l`6*_6xT^fo2q9(}M9M`ol{~t@_uYMn85Cxf1(m$JV-k;714gCTfUZq*_
zE}^|)<YQK4o$G0IS|<%wiY`Rpl(ol#>?R+qnFC5D0HvC)%Uzj}R7EPuy6VhQ1eV84
zLSS|m1d#L>Q=va>`b%{D6a<xDIC=lK(CIzP7p&|Gy8LVecn3xoyME4)_U*oQGbe_#
zZr`W5-+X2bR<5Z7XP*9!3cEY%Q3og|;Q|H>bK(=6dnJ%>1JKW|1plAf=&SCM&-`I8
zam_LOMegb)*%!xR0!bsA8~JK(K)w&OIfYn+BY>posQr|hK0a8X2g|BYD-W~{b)8I5
z`PKTIqeO<Zn3|e2ZV|Q2vdW;OoqUvQt%#iz!yCDh{N{;2Vu9CGOOr+<1qj<EO&dJl
zz83)8F=rx@Hfp4**3{K@)b8Ri4A#^RzeLT&wlp$wbY#+=htPo4tGk_D+qFMbR~`m;
z3K)P1pNY?kaQv5!BtInObIUxFmNQCVRXBrF@)MJWtstci#w3>f<w1Y00t$e45Vj)o
zLYJy&$QsG9-L=5FI$)^*QxUs&w}@pBH=%+pr-J4S2aSjwVx|THiioF1`Z2Pcpo|D7
z5P9qEd2NVWloY|p>Q1^sjrE6i;M+qh0t_9O_(f{gky}-PlVL5P+|7sFZj^NdkhkKY
z0a8HYC%2<*_8&Cs%+S@<_^+NB+%qNCG5uGcieCl3{UNSCwxU!ei&lu?jvlnovP@xp
zxNn&!f~(>eJN$lbyBvBGUwokdKw|K9P&41oPpy(TL;S^M*!_sTj~TFe=<@dj1%MKV
z`0j3x#sf@nn-C0G$%u%*2I33QpVb+A`jGe`w9-gO%qOa0<1;h8wVC(P&q*yP)Rg>7
z3G!c)DQh9U$}cQ=l3iHJ7y`TGUq3eD&27mTa%y7BxA${Zs#v}(=+QX*$JZuPIo#-f
z#|7_!%<&9{z4JOPgoiG&!tDC9174%{^G*lsqBmoFgTd|91U$LtaCB|7AH90X0*~%!
zYb6ZLdyUf=G-e|$t?o@|IIQM~Lz&JAQip!6MwH)yOQGB&x2PSXcp-hZC&zFQoRDhC
z_IQJ6mOwFv%V@;u^v@kxi`*gfv_77|xa$8{^*$3*P#q)mt{WxX@u&iS1Dsip0yR&W
zqX!Y1zk3C)l<&hDBYrH>rLT1DE8W~Ll}V46OCNqf(u?}DQKjC|PJ}-<Ur9)F_X024
zZqy%a>6tKu60CV<Bg`D3Lb$NLvJ2F>)72Glc9=8*m!wXgo$Vp0v{puC5;c&t2b!fl
zq|-79qS}WPz<4)&dK*(B0Va@TS<hsc1e5F4(!mh^HbTKf7u)`Vhre3j6<`DbnQ9^a
zMN57gmuUxgFHE80)!)Ve=Jr!U^XH+&Y~o@FqK`c25Tu;~c{4<F(ydoyZ}ZC>A7w}d
z61uptJlM~Swx0zTCm%{RI`sAxk6*hA-R!dkFkAw%(U$@A;xf9*fXsZwi7#^btGf26
zYbVf@?to%an!)mE-h|7Jr$>&)=PY596bN0_{f}0MY9ormJlyy&Kcqq@1&UNZ^oMY-
zv>$(m(T1hv3Y$2!f*JNg_AwCOEu$3iVk}`=dLBQSK}wSlZ2Lyz(=vnNDFeKZLW4s1
zM-B^Jf!R$|Goju!z?Bbyk}oUK)%Glh273a<ZvDhxREroi45f^+3l@`9G%Ol(6SIXi
z;_NEg0QNz<JC16#k>`q*+P``g8V|5u71!Dx0%B5~)6}YT$KK<yck!+gwz@NfUx%~T
zt~{>aHIK(&=zb4mb<910)P0J1{nM!}x}M*MgLb+!s!Wn%0Pq8Tb)Siwcy`B6{K~gL
z=!^R}<}P#9mL7oGAiw?7LN>ZwOjOCsWY6%Y&^q>y`8!?f*yoSdsT0<dwZV$`$Q@LJ
z1|5~p`X3kLs0@)n_L<9lWD5fK6rw~|{;d2j5-@S>l&0`kE!m6s-NmrCiX$%M<RaKa
zY>V~J0*bCNfUcBP?Dd3DqP`Yb(CiqTd}j?~Z!)i}!VZM^PgK^im*#$_OICcwUYrGp
zLP!t?0Nm`?1A+#JYnBIwn>&(NX5Wj8YULG%34*;&z-dJg1B4kAv>MeyL1mDKugQJg
zYkBhaMSb^6V1JzmHd91uNFt8q@R1^9n}YqaPNx?k086jMH90LQ1)K|4Bt}=X{JEet
zY`5O>DL_WRJp3j)<M%M!6Z(VVzn;g?{g6h~TKYZW4(qoyAXtP?`1<GVnzCK|4}|yY
z8ZZqO`%P*8nqZrz*A(axg7h3@MhgX^S`-B$qKyrRbCsKha6B7?kSQoKUdA)Jk0Wnf
z->mu%Cf;OzA2#X3Uo!opR-yeTb~&mPvU|Hd?+hi6jQan_WS`^zi{P`cvi%1?o2t3x
zyv2#^H>YuKg;CjPg@_R}oz3T-<9Ub`U%G66V{Q|jW*CVRwH#oIJR;DyV|gyD^)SqU
zybc8Xbg^aiFksscvF7kM@hAFEc9RD)&ahL86&?7|SRfL}&{6~qAEpAtkNI6#g&}<%
zzsp3pv^9H0w7Dm>*C_qv>`bRIKz#)b>cLt%_%Sqh6MKvbhwjnp_R`X`(sR&0I<hL;
zm<ct^&1@I_$$s*o-bx-o*9|9gnS@VZBLVW^R*p6`1+QOYPU_;W&003}6XT$WG?YAG
zRZU!qyjb8iB7y<GNToA4N@b7RWURan{{Pr|$LK(!rd>O>ZBA_4$;7s8+v#xPOzeqm
zO)|0VWMbR4zj>bbJL{aa&iUK5dav&OvHPxFRoBf}LSacvyS{y{U*RzuVVq4#69^o%
zw1LJ}5?m#_XbAOnk1j8mmF<*QU=M~`2&!R803P1jP5+umf4XE3B+U7AUcq*x;XJw@
zJ1#QN8*rG8jiWbDST>U;uqd)9lBt+Cxq$?`QyF;n$bMHD>3R&_b4dp*{IsxV(CTtZ
zn?4I$vZR*78VAX0XB<@Oee3KydjMSUA&~QI`&CNb2yHSlLPzHINvgYoTRi7SROg%R
z{Tfklq8>_gRK-Spt8Gph;4pLe8ohf#Q6+muU6}8Wte>1nzIw6~I{U;TwhOv$YF-K0
z#J>Zs_Hh7%!>+9LKItq2>`V2rmYHUxTiGqkld%d<{&tM-PJ~{ZlDvj%42aQjS}S4b
z7|tW6xlUc{kO;MK3@?T1@aBJd{_7wy??s)#6-!?JG~s*nu8~w%DgaRyMV>$TiqxCO
z_0U0P^`)HbA>s5H`>n9e2JDil^<Q6M>PXj{U2lqXGknc`CT|_qE*;|2^DbE*HhJVN
zbOcxz_j>P>8r2P!DDjUvBLMVzCz$6T#LP0wr2brIjAb}61leS!@nl_dJuF^oncdN~
zZ(nzIO8O9~-8WmDHZr#qhEN6KU8u1gd-3N*%Q_*4F2WSuPE`Pj*$F4gE>$^K5XeZ>
zUQoQK-02_3O1iP)x*)W3fAO!h4fl_XVYDap9~-OM2D1|pMkN!m(g1XkG>ERWhdP$f
zHquMTbnRjYl;bZ+c7fp5=5hV#PiTsRRgFcFfOip#5X=WL(_R=IHb+)ubi0|R4LA3{
z0r5ZD8th3t)Cd~y4jS}c5W|Axu1<?<$(n7Vu&z1_A}v(-qQ$O>{!pi6T8kyAd8;_h
z4>mMyS>v`sm;L6=z5|K^6Ehr4&IOA9%!PocJ3>T-f!IsAC?b<$?p~P%%)s^-tvKXg
zmkV1<(aGHBE3GWDiC5%&OUokp;a*z?+QL(2A7i}as*9;bYU=k2F1M`A-W<<n71_vE
zCjt%yjz6G6If9eNZ&c4rX-u&npfppf4^d>l5)V@U_%*dEFbn)$rHGaB>gF}r>|Os1
zR-zgnQMc*X;(8z5z;HN!Qa_)AZerPvRC=h4gtVJeBj=>)tX(O6`w2=R7y*q$+SmKR
zQTJ_uEqcZ}c~stRYRzM%Lx2x_Rn;vTM(+p#eT2q`!YfPYir(zz+CmZ92o6&;iV0+`
z!Rdmd;+qcyzcqky9fK>tXd|99WG@fx!~RAl&eq{remkf7NgXR^+yc!oA^zyMuZ_!M
z-Bc{9d@YXl*wp$!9e*Z}{1C&65oWT{RC_akjA=I6Y}>n%Jci@K{r>c0svG(>0*!VO
zZUu*#MQJ371!Emi;3qvL9dqL->!Y!syQC_^D-RMDz6Ox#pk1E3>}B@{#Nm4w7};t7
zAJ1uOFg&iz#dN!RUw-oWzj69ibG7yiVhH%7Ezb$yv0!C)LcIKP?LsMxnSWRGLY1PT
z(@O)PxVnCvrvL}vqs7J+n>IMt{th~DN(4(P)Z*xypW92{t5MZP8i^?FPw*%wf@=T7
zoU(0NQY%2ViY6|Hr9@!{1*bWGnGoU&GYH{wJXb_8gw73FFfItp^Xo1O5YX+@?^wP2
zg0eRhCyI7xHg6b?haMx6?$XSf)|XAeddrDvgr_gmo(TA@4w-`(@H*&gto<E^n_YuY
zuf{H#g3>~-RV?5gNkZb*!$EqckYAgG4)?aPA_&0xMu4Cvtj@DvOG&VAI-Phxr`R92
zbck|SGboB-dyK<E)vq*TeGMQkZN^qqc*p~g`Z!Qsm|;V6-Zmh?zGt}Li{@xC8uGT-
zJ9P>Aec+d`!t_0}1>EDzdzI<nzRlwVtCE?*)8hKi24Bb1(38{c+DlEzlTPS~n|~ow
z27sPlzV8@_rVPy$@+3BC-r9$1sk<Dd_+NJ%JU3lA`Lu8j@3!)0m9M)Z*DAO!q$aw_
zL9xcYR@RS{1z*t})^^cu^q_28j$_=DpRKCLKb48laT|>(*BF;Y`{l)-Q1Yc&txN`x
z$_Ek36B*nSZl&*fT+J%TrC|Mig+*lz!vW2a9WMqPzm>rt7+OJqeih5KG<=)lcE%cn
zp1$LXDC*Ryg^KOiZ}P@L(I|gxTs<ULDS1CjFPs8q=7|PnpzDo$0IHsAYCwT5&5Y%@
zi(2d+{n)cX^q-R}w9_lO+U~o1LMWw17R~c!biH?siF}QB4Y^a+;T(mb?R$xZ3E-}#
z{u4Y{W`$VbFv4~{ZhVI=bZgcTZl*hU_fR;gSTw2FQ;5Cnh#2oZG1p>4i%sI{E(T))
zB0eQXyd&*w25@W}^L0ucE2SR~C<qbUui;W+@rP177cBVNYOAp6c%pd}yw(+B>3R!;
zQoUyT%CP;gk;;T?Xw|jc=|F5K0D{M`hJ-2lrl!%`4eN3T5#RQSZn0rr!r-?qyNp)*
zX%nkp@db|wzkg5aUWS|3JoA-_Vcy;c^VhuzNek92aT|a*#prCOag#1H_0d1}38(V%
z@_D`%YXzLIT`9EOWf2Tb*QyUS7ZCgwxQO(w48~e2*{+7a%RXAWiGq{x12&!SnD;{i
z%JlQt&yyQd?wkxVSTwCm`ivjEZFmVYHQosy!WF@n_EhR3td`xi__y;=qF%xB+{+**
zc4f0E5`D9Y?>w*ZDYV&a0v2($0@e3fll?Whv?LR%c&4;S?&tC4+|-$4-A(FXF>652
z<N6qpiv+r!^WLavIJTj&LX({2p*L}zZ4kAV(}%f|gkNnj{b}RiV}&JPPHP-@t@Z^;
zrqF|56eo_=gR1Q~yYM{u1sc)9+4BF;R@|)rCHr%*a{X6YO=|utn{uK2M_QHVf!Ge}
zU&;s2oG&}?t~!(Ozy_C4A7h%<>oZ`wEMY$%@%8Vlrjk(z&w_c+-m(%*P4KXZ-Sj>L
zA}bhN6K?g7WAAC2utw}zC7O?J-;U==K|wN+=U9Y{zifMYnBi$phZlZ3I%n>4oG95y
zaVXd!PAdiWG|gY%@5RMor-fUDd^Z=W4Hn=OQqW7%Dr-`zIT0tQvi(_7Iq~L~jfXN^
z=-$(ZXcA`-VgF2RmibCz*JNcykkQU{bfY3*NcSpC7LX<ho$74NUt_ZdK)pU0X~AGA
zj`g+Q$QwbClXB~!A%B;~EA|sTwS+=$kQv8gjQVF)QNdfeMoPcA2WJU=SM1PaTm6J_
z9@KBg=_}UWV6D`AHqD0+Ve7uYr1kq&6KCymM`jEy2f>swxMbWo<Da!`qQqUN>9jl~
zGK&S|G;|@(`bqspUDpU#m83e#XY@4x6rBXxA-sQS#exPy%2R1E$I67WQXToleoBq9
zyL1Hr|A+)BU)`H=8lW{qZw37|(^`db{&7#A*9(2qLx#bfo&zE~&#^rJuN##0#MrRu
zV(h|KNb&>AJ2R=r%g@&rpNFc@)|Jg52zTMb%Yl~bI|OarD7_}L^Q+>?IgUMEEvjjf
zXYjCK+|Q1+-U2Q0o9X_7dm(e>PuF_tk1s7C`GXG2KTTM=x3J@#G$@fZa=<J^xK>`@
zl)oW#vdtskPM&3a7lnV>D-1<<MQu!-lK#Zjn&Uexx7&L_RDK6{H}`63VH&)I%kVD_
zaR>^09iOe|hS+|M&Da{N7Ue+e)w!EuqKhIa^VoUW=@ni#kc53Z>GRA=A6vFaX9PhY
z9HrA*#~dG6N5|3`EI&El8*u%{pKiOofT%sMU0dWaA;R>YwcR35x}E0VV|gEVI}uB~
zQWlc-q6gX7=p>_P@QRmFIR;h2%To9O8@{YJKM@idvSh<E@X!;W=d*{!-9AVR`)n+U
zPLtJ#+xQLDE}O+|+{vUo$f@(R)y^FFQ`E&7pBi1fhNq`Ipo}YY>$sSP5<~EvJ)`K{
zDBTQU*Y~QaunC3dr%9k0;zke8RNVVPs0cVEXuV20go%iH7NH&76m2ygP_7g1hHs*c
zD?-nS63ru;>N_INNR%fNN+@h^={tPCcfxB@`aR&{dXiM7R}jS{E~=<rsh|v4E~4Oj
zuudpBw-@i@>Zdnn&CWjU&Lr`I3H_CHoR3r?EKk^M-tE1JNLq2u*7Pes(AT(!(Edfm
z{``0EoW8|5nVE%aKr?ZSlD&}}Y%s8=HR)MWa0jZ1L3%|$!eky+iruE%q(o<p+M9E~
zRb&CDD9-3Z&97{b11+_c{q6y{iSO;TtK0i=zVf>@y=rW<>HLnL(ltIL+f|%6H=aT{
z3}tn_jlUcI3xVI;t6!c$rJPW7_At+dx&k^rkCACCzJ)9}hv@wH+aOF>%pNa=U|lC2
ziMWrN_GBKXJy~S@Zi!(k*9L!5Z1mg3N6FA@dGOpK|9p|vIn2}=ai9$F-l$_h+MfJ|
zY2(M(-0a=Uwbl`_va&;iXfXiJjy!^~?{6G-)qV-bOB9RT<3jq|-5|sfh09A#`{^FT
ziyt3a^|GkYd<BYw@Grp#mRn3M*IMY^ua&joZJx4RKST`gDM6Kkjl!b=s_fB#E4|By
zS_GJdY^>5b-41qnE$%^};|JcWiVl1ymM*1RS7|pJbz!}mknv?&-;*u2OdI76f@pv}
zX0rAC`LWxUiBH+2%5M!#-LEGz?QhQ(H2<9Lo}#E;z~3=_X~XzbzcG<)EYj`hemseq
zq6cCxNz#W7EcNKDkv=!cRwF<i-6BC#@b$u#Zoh=HK<E&0!Fz=PN+H4a?@ULLE{d)(
zWbAe(ddCKNmsq@-e|A<(Z!M&FLeDhRho<^{?R7zrM;?Z@NqM#B6ZNXA67mV&Fed~I
zVcfseXmW+rd#qtSn&-+>4>2(s7d_O!__0qG?Z6lCR~FN%%)pVoe(oFj#(d%SWbji`
zK^P|ScKO{P8z8`7i2Of(5oy<$Fc=W59RGu^NwYkLzyN%2$vqY{2-fz>pX2v?@|n@b
zFv{bThm#C*YiCI)Y`6_9FVIC$?ATv#9+o2hTdnd(eL1?zimudjZbPxAmZ-w!@{Cqv
zoTG*(CuDM1jNS^i98W=EMKWlTU#d31@}JEJXjQQ$;OBC0RU=!zjC!0`foR^O|N8TC
zwLq5wFonY0ai@M`(Wx>+8=HXd(6(T;N~NP|+`eO76j5GrpT+-FR}6@4WR^#lasn>x
zRyaLcgb<=j--?vWZk?T#6Z<+QR|5>j*qQAut;i4;Cu{0{fB8g13kuF8mKfZbb#w4P
zU+*rQ4z?q|Z*`)ps(qvi@tQ|%a&mMN^Omjv|L{CSaOo}-Ze{L$v?<IR6j(#`*rD|@
z!Dv}TV>t$MdE4Qt0l$f=<W&vi(+Q9*U3B}8@jq?4(wmzu{k-_x_TjnC&#{Tlky}z2
zD#ML)XWyG?9q7JHZ8@t?czIENk!KS1oPoE~5JG#}H|`~{rOruTdt9aX-LV1Q9>-aL
z@%y`)XAbYmRHrBO8a~k{(IDzcpjIrF$*fAx-1h4OB>RhdMr2-DdtzL@y^d-l7$f_Q
zr*hb*C6e2_%S#daQ+Q0M;1?us#CRmBAA;vwLCvqo6gd=wZb?yG6|mv6k8k126rH{B
zgGgPgNf{fJALG}h5T#nkCLSe}x+J84k8N9gNcNm+IQM$IoJ=@u=bv+gu<7Zv5XYPm
zVce*10;9r5cclzq6nkv0+!XEoNSEXf+SVQd#mPU_QQi21^sb>BIi~M**AOikE5fHd
zRH<Th4~p8{sA3n$i?v$;CQY4R1Z$%Lm)=WF9oO_p$7_0&_9YeXZmu^+Vb_!ZsxpR!
z482LUXogv|i|5T6@vKSg&xjai&C%UWh7E=$5SVg=J)4`IT><d(zqn*Fee`44EF_}V
zSHq~9N$eWEvIBAb^9QsktjIO^bN58yp|lBH679HC{i8Fhv&DC-<PeG6JWGO5j9)^l
zh2Qs;SxdxtTG-yxW>t)gF>)P&s$08tt+4ud!wn7d{J4F$HN+m)S)F)LB^4j98ykC%
z$r6S~UERc1w^Zm;74dQ|eRcGjcbP*!-@EznN>9fv<wQ0sI@F=oOwpWjgu-v{vs?`S
zTp1{?GVI5j|D5^9fOeNzG|5X`gF8m-vrf5opPZGH9d6$@Gki9rFr4B7^Mhn^DF(;%
zqUMWRmhId~6!Fn!p!;4|_91%C+94oq-=+7ISoIZ)y)Y~FAj#U~T(WgvUGL?mp)*b^
zdWp~yK2Y?3W02Y$scE{b+@0kr5<&_mt4cYXjxCmOJ`pbESpa46=mM>;y~Nwtex0kh
z1lIPbe7xX+W$$P#``B*48rXx?l;)tlGRN<Hh{?Nc#BWPOg^~6Z79k9Iz68gz2}r$<
zVEF11OOxqxeT;d$cn&|HyXMVkC(w4xV9T}iH6xMQi%Yq1|4heV!Cuo(1;k)UgH>><
z7qn)ru*t{w+zy{dlFY&+j+Qu0l-wC7RxpEJ%MqM_0Z5gJO8YYa@-)4C7kbq#b8K=_
zN>NvUyM)9G7Wi%M7#9sAdVXpp0cx|O{R<}Z5;akV)TcC!&%n#iA~11vbJ`d*jykv#
z?Lqu`$KHh?FWb1=uSrG&9%hG+7dXa-dn{yvUi4hwD1}_{4WjIDZ13+0e-`*t#5H;G
zS|Ld7+2X4*J)?dC>=hivq+M4go35$41u&#+Ke8_DD)W_>1J_O8Kz3i<D(mqy6h&<v
z1wN{A2)2hFh4rCg>666fgI`f#OL+=9GXpV&MM_JY(LSPukSq(0jKpnQXY+7_#Vt6g
z82&y#I6ONKUy}O_p%b2JOT!bZu7MuwL1P?sfk%Y{F}XQF#Qm8KQGF+Gfc107WCpYR
zaAEIXdDh4g`3WKkVU&%e?oc>EwDuc7m4VU)aFS-b9F@%7IRS_wWZ%&88oaDxtVW!F
z^h!#Wwjo^@I}&IP?yr)uoy{@b#LP<XE>8ZnRFOiRTSc9*i;&crC7MvSuysT(vL9BR
zj7-P1MEDNKtVosL*T=+WyU!K6Uti3;^cJGe?%#;tFJd(n!*nbM=sue><N8i%1XvIH
zvwvr-F{9;ADlOM$6p;1sGS=p1h!krL)kMvknq8arxPXDRfHM&8a`7{i=(_USM&?s>
z3tt38Qg%iu61HQj37gSOex+v}k+jky+bavk2f+Y8k`XRNFrj1m;QqBBc*^tTC5EZ`
zr1cKG*I75NVPDj<;W83I!SH!dJvDE=T38?B+`QFWiyLaWq;Q|XPfv-&H#0oyEp^w+
zs}arA%6pb|ZBkHfty9QgRPT4l7L6{(#7j6Q$|8fbW=6`(<qTK|hT@j|TnH&C9SND1
z7oZ11aDUP*NdHRc;pgO=)I{8w{76?!#{)6Xd06qWA1@&0`v8UZnMaBv@E5Bwdz#^p
z8D*J#eiW8IeHZ7lcg|Myd>IPh|B(FPb}4#<tS=L~87Yox_Y$ojfOqxLprU?O2*!D^
z`Sbgk74ghz%EB=!N1|~d(QbJR?>j46)^j2t4DR^1=4ffyDP|O%`$x~v^d?7ms^q1a
zS3}f&Kg*pIc$#c!mPYLG?Zm}Ing?6gjYU!LpoMt8&GY4i1S&psl3D`fF_GN(?*)6m
zCAB>Ttu6~?3X%nGn8=goE`E`bw=+%t(K%L$0?}c-pId&rEmBGPTqE2p#siP}V**=1
z+OSHvHRUW$u=w{CPv;FSi~Nb-Nh#tl%jtX}r`*oO#OMeq5{C8whqI>7P(B#l2P0?A
zEV1>8S!B?T%bTpi*n?6(IlHZ3Bb%YdH9rHg7l&c<<6_&NK3&-J+6nA{@>WQOY!EdK
z)mB+>+DF~?E3d~GX|p(KZqwqva-JQ4>1`1t@Oe|DWHpHm;ZL-)*b)u53thwZ#WeAh
zMwigyU9lyc8f4s4O;dy`@o(p=kOWFkkItHRN{n|i#tiF*5$2D>1147#rQH!JEr9qG
zjIg>ssaMQxum&C^sQjEN4R4P&WDW;i+Wk5xBbd1Ex2>h}w!r*g(On{rRr(u1sBoWb
zjk0pVEo-fH{*0D7M8v_;)VQ>PvYIg*#xIb1Sg*Ol+wT3kg+>fg6!!tGe<oT^jxNx<
zyBZk(Tw#zqATb}Hvw`&~h^~7nk_6Xv`N#n@r}~-2CBV0#cyFVIkU`mE$6u19{k1Rp
z1bN<{jm3&>St({yWXvh(_I3kwxS<lb{TyZ6{BR$MxyqxPAFQ6gZ97CPA4>_;m&fgE
zhvokUe{Q2A<J!{`2ZzoY=0GSfs({G&j;C&mby_R!Rei;}GrGwAT`?)I2Bt$yQ{WvM
zef8H_@6}?awR38qM80Yr^ZnUeBQwR+Rr(i{)G+eh{~-<2^aCJa(x{FhFu_^5{^#VN
zTI(M>WC*qMs&*T-N+*tG3+Di1MVXd{uD+=HKSV=WQQ2gXKrrFz6vN!5d2q&F2~Lf3
zc%J#Yh55U-H;&tdep<ti2g>eOz~A$35}!~*CCaE;XTx95`}b`thQZpxhQouYM+>8x
z#wOQ-kLg`}tL9lOGU=Np2C~=!vMHe6!vCi-yxy+G*(PD5Hf6fY=}js2&7m{MXjpap
z_r&Vo%@k7}wL&dNQPRdG{l-0s+@(b`bJq%}%cYB&%~aFXfOV>r%e}4zW`{+gtq;Pd
zRVnfq=c=y+$+w^SK-s?*wC{9Wnv#3Sh5RZhx+=cYosAaD&xcH&Txjh_bke|gmpQnk
z@r6gHyOZy5<sk}%*!Dzn(YGmxeX_E1#K4p$pVYrEy;@DQm)$=$qUUur<Dfz8u=#7r
z!D;t!cceJ{j_ldwfB*x=@99eG<6+3Vva#hLq8)QZFjK^us?pOc?8$Mae#+wO{j=YP
z{Eq+UuIfr6>Ch9Ve6xu8KU&J$uvxewo2(YO-RHSZbKTd24pW&U^E}=nz?}**-^gZu
zDefSajbGIzs1?5JSBIrhkrtQYjj}lY(u#cGquL{^&1e*<Co{K`#kDxCF=yKrIJuXp
zx6gXyeTQC!y@-K5pA7xy8%m4^HB1;~X>;tjCcZfw>OR4sx@s@Jodyu6FGO!sSUmEp
zgPU@Cp@27xnRVjO`Z%vDhVZ2be=*lxRhmm1B51tMfkt%w>7e_p1j-<%J6VNnA8)E4
z@7Jq8Q!xQM*RpI>pS7Ux)-Na|WEl$;Q1|E|77=b=&bV8X6?TO`Fmv2=+a;qoXs3k?
zA;=kB3tcFT0+@MM`QyNObz5`!T17PxbezD(&EI5y0T*sPKRzb@4s#VnXw*-xM8AB<
z_|W!)3@|xI0X?ZiQ~ca=U8yF|0+QeG6MrV9RU%GELojs=tqnHQ<d-zMMI?j6Kjp1*
z^T+k_WQFp_JBrl&c8f{2A;C)B!>-5P1<go#F(W{m($rxW#8v>c=p3xPERA`@XkB9d
zstz+a#9YA;g$mv1CnuT#Ja1iY3L^+@tu!9jvy@nUyx&m%zrfvOL6^t|_$)tj`B@e9
zu5wU=Prm4G;|9w5#CjGweTQ(VTYa3M3hf`8MPnkoe>1Ha{!w>$bHMR3SYf9pI5lh)
zDeoMqNWELkKL-G5nsrqhs63j^;sqhnKu$uQ74pd%AFQSRd5vw*J*|gy8lGasrC-`w
zy|OU=8d#!7pA>Ae_K>V2vfC<9Udle^uH<f{E5eRUFX5tY0&7E-SEHQMIsPhRAoh%Z
z(h@H;sm)Qu`Y?twh@S_LV+Nx^EdwIeI?3m@W02C7k5_=0Chc$NAmvl+dOprSar>EN
zXyo(-<I<s(V3bhjy?VJ>w+Hu^I+n-fw6Kx%2X&eV)02B+g<T|>^_`SKG|&bqthY9D
zf3TaL9~-hrQiyCgWX&siGI&PX&v`wZ>Afcmin|FdwX3ew@z2hHBCKFjwr6=<IQuLJ
zXH1)o*o*>h`_^8HE?`PM{tKGxv|0Iusi0uy8ehq>gaS~(Km9$orIr0{2xJPL|Gr*d
zWKFgR`AKCpnctb0h3mTM=LiVe6N8Fop-wB7RcMXx*jzPEe@E&LEmlNVImJ<K>dOp@
zF^<=`e4Onk&HilBhyOJcwN!UrPxQ3D>&merm;wdp`Y8B-5i#zedDoyuX^{&%?Hggx
zMQU}~%`3FiA`WI!TF?HLovzQqX-ySe32QzbXV_I4P-8C<4-#KMovfHY-*J=kEBHx@
zOx6|cwt|(_5CM<NUeRBFNiN=9G~d>b%3d_+j7;BJ7CiT37Ou&wA{eU%#DtA5j@#ye
zf+hsWa>BOA7WH57<5|+zk8(c1I9h{NK$C(PHr*;V9aP<fR8JBg7fQqvk8#^Qu1%mV
zZdgI~536vBS%S*^^x&Su9FfwCOuU|kE&Z!HD!!yO)Q658uTvQ8O5zCvId^9bO9rnU
z;F@z8O3N5+5e2GtQXo}=AZZ^5ZFQgKfKLJF;?$asJU)CxqDq6aP;Y;gvvL?I4VT)^
zlcbILPfDjz#%wGkTo+i7<zwSxk3Ni9tT%QD(qqq2af0ojvz)S=)FeP`JGak!>`!?q
z^lq^GaC0t@zL>56#w^f3agb*C5E0M_DUlIrBJ!N2sH^cV8ygKK24r1Am|tw9eMN!n
z2W{1qqQRpP6wd9otV{M=C2vr)-1X^KYto*VnLWa3$M1bO3pWm0AvmiEZ_`hrF%9W)
zfdXo2<m$W!?2#Eac_KeWzubO@*&@(H>=Hn3<|+x#oM^kih6YIxOqWZbkDC2Gj8~|3
zB{@I4Bj@0UcoHQ18mQx9TRtNw4iyBJz&G|pA4-))X~N70;zRs13foDJV$1__MwO*2
z{-nT<jyJ!xtvoFBjSkT=`+N*%$283F@(HyW@czvogD`v-LweAmrcWotqstfYC(^As
zgHKjC5jYK^qup~6FFpGSJHYFK3Cf=yiQ^~>Wjd4X%#5brL7;`Rf-Ic`K06FVv`K#>
z-jzbRo)*M)yOov#QSnmbH<u;C_#kkJ>klduzUPfd7yHd(Fei1NI*CRlHKmR*@uNu+
zuHuI(x+*#ETLq5`oj<p=!p%5ol;=9UgHmqyo$*v1)0-9YRjojTxa!5VxY`J)wOKaO
z(b_AGQ(kXI*f;-7*W;S!*sLA^gJw`wFwf8HlT4eOCH4qmRRNakG>E=s*qrf}A_`&e
zZ!#0&>A--2;e1O_I!m`Cg7}YjIiWO`8Qli1X7V9n5qU}+eKAD7!sR5!gP~^{ZYY!2
zIvu1^FGDqC8h+h2(b1D$m%#q7!ElXx#yMQ~QpI6s6LH%yq}1NXBs5u|lb`Yc-ZzEG
z&Ig_T^fJ8St3x{=G#ocjzHeST&S*<>qLFc1uoxXK_dpEKtWQTSlw&#O6r#p;mWS$#
zOctJ6+c4X{@n^Dz2@gR{O^CMI$)*_k>%;)sZhx%>=7mhTsMH=Rm=ef~nx~Ds{^KXK
zHvFm7|EY)waA9C@n0fwh<})ih3-^C(E7jVvj%!0$n{S#<=z&mHKkjl)7(p}1B;eU2
zgW9Fq^C)yoL?V%-2RUYlfo?XoBJm85%r;1s$-Rk;yPUgOi3aJY{nBW%&#ulcuMdSk
zq4n!DgGpy=1J?r>oJ>VBHA|yM$>utZ6HW>G2F|KQm9nYL?d|<a)%AI9tCuI5{~*cD
zQSkbm<xl45sY~|r$@<GDn&zru2Q?7>dPL3J+Ly!9MbvLT-JN)CE?Zm88x9?pp~LyZ
zyTnUEVkvZ#t~Bl_tx$Bm?yu!Ki0)@pl<M8sQ3q+7DrNVtbGJ<$pv2ifSfWBH2(?3@
zh=DdK!M8h32WZ_Fi6oB&LR>H%0LlN)kSsy-s~DRwwp|^2Ztjm5(Sr#n)SqZZ>LriR
zFo#;Nvg!uGvzo7l?hBVUSIiKJ?KL_}mH8KaO`Gy2$#3dru6&mP&(Dkqmja~?L6O6z
z;rZVxdoZF_?oAhr@EEhD#un<#mh{yhp{cz~JWz^GH5NV-*W9SNoC`N=0bYV~u#e*u
zbi{V4ig_mJXJT2gje(Ssi?^P`0No2q&C?8)2=}6@PaTeM_uoh*UpKG2<B-t??z}B>
zR61DKL2?@H(bE9ND4wqpo2Mw%Q;tiTl!a?P!;t*#$!@<&TwU<d(5SfUFIP<c*EPL=
z>0*t5P1H7~=<WUFl03=-5Td!P0GD--UCT@tMTQhP{x%vKus5MA&g^ZgZFuNnKG51V
zRsj?{=DUo)B%ANo!GYF02_?}(^TZu@^dzu>hBhXFkZHyU`F4u$GP89D=mXVD3jv)@
zzujB^O|bPYDC}pqyH09e)D*-laAIb-@S2Xi_K~4Bxl?B-jL2yLkZ612W9d@|bh*l!
zH6zMFCI*9Ox)+yt#B|8*>7wv5){wqY;yh5{r)f0Bgy-e_OvV#>G#6De-a{GVp+Oo@
z&Xkz4DFxakDdIcZ8YQCpytKvcjNDn+@VToIU>bq=dWqWQ{FP*bH!;pmd*pPfYnrl{
zRH>={ispW#L)>$KB9+XYo{+-nRUAtQJ&D0wMXP)jI+j7$z#?^ofw%-WdWI9FO|O>%
z#w=z0BrcC;DP6e=n=ySbm}A8xoNPxny-GD(K8nZ~PJ#^cShPe^-l6~}1O_S|QkWu&
z6B8Q*V<QaQZgJ1Jp^xsxdu25J-%K4a%YUBS82VPnyB)^>f0(R+fOZ@PTnXJdpcaRL
z&VxUGrK>htvN_xkG2BI+MJcFM_K#uAG$rsG6h$55e0kNV#y;UQ_P8Z;=SliSS>+Gg
z)}0^!yiL9W%>kNqv?+rZprdYV5He&%2&FoB*T};8@dvq^t1BY(hJiz$9p)z`RJ(^6
zk1f%=zp+2yqcM9XT{8E!`h@6a5~1lf-}s$D2i@csZtX~iwiVx&k`s(@Ct>X3#SWC<
zYRp+q;4h)i<*1)R-}gtfgg*SvX-v4njT8WcgZZIb+$PwfUQrxq@;M`PO^u6B8=CGP
zHC-aEvI%>0pHB10dt$L(J}~3v!~tTSn0*sI%3{g@g7~Klf~fo5x59h>?;pY<gr@Y)
zl{ebMg!F{I>NIm(j6X4*_EG<Wo!PHUyG4konJh{D>A5o~oW6z2uj_G;+inovjvXf@
z+>j*8zu$5U!4@VM!$gJBuWJ%{S6)+yuFGqAG-Kv<nalICHLflX&yS4x!C6K4HbX(2
zT)6-WqzMtYYO8P{T$nOf-h@#u!x9*Dt3q+`W2?OVEpX61)RFk%=&CRnP3LBP0i|(8
zT4>Z2$r>fMN>C8S!T@SO^r~bv?TQt{Go2Avd-zrMlZtV!mk4G6uEzKT<A)ZKF{Aye
z++^o3j;Cd~k$hxm2S567=T?JoKhW9aU;gPWz}v{57_Jm!-oNcp+2s%*BIvx)Z;|kk
zjrUdjeduyVV3y%Y>D6?}#e_2SJRsg&>iFl76f9O2FhA;ci>N%P#$R!Hu!Lcjk=>>X
zS(I}B6^0OWIH4dTz{E4e14LK!S0qADXZtT}_^cO7i6$g7S{NcKTqN4pn9ssbFC!>c
z0HL15OP18XKkJHk$g*)z;Yn0>AvJ$0eOn@P4<&<l#@oZNF}{Of{Wsw+mQ&KQ*CrV!
z4qQz`?MvO!n9_7eds6<4FGUU?XEgXx_u#c**t8EytljH*VgD=;EqtZT=cAGP8WTY}
z7`!Grhw2_!0UA9oilD@L{>mJHWrUj208pq!4Ki^9ubd3jb8<=-&9<H_7DGk;7+gu{
zSSYk~f$4t~xZ7L{mm3AX8@rNE-U%dv8pOkDLg<-v_~B7+^C$9)RXPUufBWPcE-T_Y
zuLDs9eihHYQOzA&OHdfW6b01azC%Y9@KDb{xa7MS|MP`|gq1h$3R7kX0qG-F2Nt`k
z?3=wDin60`*uO)TOn83-P8Rg?2!~|3+I@Z~t82g{S&Qlk<4hHJeG&3;X(3O(^Otxx
zZ6fsH4mS2_+a&Nz_QK{dtFAK?hegm&<yC7(K_l5sZifYnvoNti?n=mBdP}#2+QDXt
zc_9(TiXJrgv@RaQOz~us9(5+i0GWQ^QDn)1r)Y;Ti4E9462V54ay0Xw$uKDBk<wVf
za%VzHCwfQqB>3aD(JU@eF(0^1lt?s0?@QvH=>Il_fvP6di_*8M{p9%#A*`ODJB(f;
z!I<ou@d}q#;DX9jlG~H3VXksZ&DNL!h2!5a!z@uY>8~j9!D(YUVsz`}1w^#eCw1_7
z`}A}1ifzDhaGUOhK)OvG_nWtz^_O3B8~1xEKHfYTis>^a`68P>W#z>G#2mI2Xc1IF
zTC2$%$RIkV3c_<YE_J_sK(&<l5?0wpCBcZJ4JID+)NYtZD5diD3AG+49<;$^=3>Zi
zbVQcKBg8}I5u6ck5W}xD170oH7R1h(&F#=T_}eQfM%a2`w`u($P(~1tf3HWF*SFTp
z4lI6fcFYDdGx#?NxxEQa?8z$lT_>Vzehtkk=WWSrZrY!zYk?p@@t)e)^o`J|RNl*@
zm%qK%4*t|B_2;jfnQ^GN9!3$BLUvye&U7D;U4yAI!M_40=hn&I11uB|+7nP$W4&F#
zu1*Cme-w9Iy1g{VtSj(hW+OifQa2Y35WFjY9f?;Z*c;{w_Z)*Z!S^={*fj^SzNZY2
zCN;;Ua+!Dj><C7;!l1=MeapLZL;W$w&HfALz6=%d1%&x3^Roy;@#Vprc(Qd8caHF_
z_4f^YA_^N>a^e>_6S)4LUumegFz89nG8hmn%*_9z%EC1MsWNVqt|QH3YVZWwghYD6
zwLF@JQm~u+jSC{2$YvO+BF&=2=Pl14V!X~aMv4Uf1XFMMWIW#_ZZAq;j?y+JOMN#f
z%riXQClfaWR~pQ8o4l15Mjif@9iSg7say4mE)OUxE6Ds;&Fr#dT8!;q8$?uW+MEj;
z#04Br0ba~`yc`$-kY2uRn2~Rgg8q+JY>hOkh{LDBaK;BkZrdN~e<i%v*bgqSo^isj
z5*@H$hZY(<v2Cy^vRQrZt}dCbS?MY}D)4T<-$&Q?OlT+hu!c}!YE?k_>^VMK;6<P<
zR)pmAiqy@Ag>T{`^?)U3z&5EM?GS<g{X`|{0B$b#WW%rFS#iKZvF1l*vI-Y+FVF13
zEbc09ctEMVQn;Rj6#PbCDCuN1N|EoJRXSS;+#PM1{Yit>K%Fh`^Ww$iAo!0(h=S-<
zVUrc<`f{uFk4zEPhZ{i9amL3jWGT-(?kg0kbz=lk{Pp(pd`hXyLGK1lg4;ouLbC*Z
z0hjbv7}wcWQZ~h5mnmGhATntC57vo~Saoe6qZGml4`<&}PsC%Q&>#<N2uYLaNuVCG
z^^?V^ogi6yPVIEPHnOyX-!t9ILzjKp6$L-+qkgY=TxF~(Ee9s%s^}Vu12A{~phn1%
zDvM!0WakpCEOo83RpALn_Upgln(4$#|7}6q5KLaQx#|<&{6r9qVxl<3vGex6X;kL?
zE%H0~1@j;s$X2I*ve9oTn%wng=VWig-{a{cVcBBGCuXF~?-cZe6qj<~bKzhcR8K5H
z%X>1s#jTHRgt{3-*4NqXjO>Km4Gvs?6DtpjL)8zwyM;(0A|u7rO8YV23j~J33Q)Im
zN`-66sZv76!t7N>m*?L5{@QewsP74`8W5Z_**qiM&b{_n;OV>$i!E8mkGha$o%ol6
z`CCcgn5I+$oU7W5ZP}R^%1L{OlpCMaw`N!NQK8TR5{3|;Enf@JpEap}cR)UZDaN#$
zGVaPiSxpZ!#P3>AqpPKo3QVr3B!HFr4pW3=uN|Qp`J{#^-9_<K?a)kdedG4{hNnRV
z#0U%AT|Fm124XpP95%`Ow{qu66%BW9UC-H=+$H+Sinhrt6JK1Gbt(3#f&K5Rl2suC
zgg>K2R1GojBNQ3?KGtC3;tNR6Er*;g%{zMEcVD`L4$Y62+?c<vDuXO|jsSZz?ccA+
znygX7)UU1xAZ}0gq3;-dGNar~uA8^x3DUlT=fi*4t;l7hEMPaaUpH%gP@Dr2pI~Zx
zjkQD;2R}n7aEg$|9xH#d(i>xj3Yn}85Kg)rdLFzOQW{rHj>EC4yYtuf+B(wJGXCkh
zbO-=1jI@3D--%lqaXkc90zN?+9ULSAI5#u%e^a%MfAkRYR@DE}L!LzJ45H19=$YHI
zq_Ly|dnX{IOmN_1*KO$%C?pe)Zl8F<H<~F7*BUGvM#LId%o6CaDv>HjniG?iS#&V3
z(F*OPo-qy@wtuNP;X3k1K8R2_d-180b$8MZQs-7T3}q!?i%>#h+9@yUu)|Z==Xna!
z0cUa$^<tD}2ZU^c<w_k5o38v#x448<ir>M$3%kNPW1zG0G=8J8B+M8YYt>OnhVRoT
zL)*GJZIDn0li!g;gKEyk`f<}t0*gcd!R8W&d%wqT6|ee!Ef~Zi*8>f2B7zxyJw->1
zc*@3>9VR_ugoT9I;e`~0Hku4|jHa8n3Q#c<&GLns_ae)<#L$Ymbva!|%EoJf-SCjg
zE0aQ`ienB0L52LV+a1X_A1S*%X5ZJtD2C$kX$VC4wiBW`%YEPff<M(mVu~1nXl6!!
zy3)>I6eH@{8%MTC=3;K5pw|&BcNknq)Gs`6j3OHcz005x+3RK+DMODnP=MDo0ptxp
z&cU$q#L$<J$8)F7l9N}t$bh|~_9Ix(yq9$(R}~@Y%*c>o3z3FQ(L>9}r{h>8#z5>z
z|EP&#k%4>0Jl1ZATaT1sEYP%7yFCY=Q|zU<2hSv-ccv8)CQ%p>HXc-&w#PA6prw$S
z2#!q#^QtGxm_UeN!;sJWwjyB>1&k5VigY!GKf-a3bisi@nJ%+6<xAZchJ+~JHS)&H
z_P%dc9A14rmli0bSIvg&=uI3+77_=Pxz(mc$MSM3O`b_@OzTH45!pDJeu3J17}b6m
z#a;<nQ647{AJ=*}>L_geV|<ilatU>7Oujwdeb+~P_#F3+jvAHbvkB;{11?M&hfZ}l
zi8&nC9&SxI?3Wf)Y-bJE_Ez?X4=mlSl2637*HM3DRLx*jVDPnRMh~Lt6KdFzG7Y3&
zZcK(R_dOwM*t9F(UGDam1bmq{KR=bdIJfdTXu3<e{(WzKDSm+^-oA~HQ()mEiLGhF
zE2!}xjvfjQaYQTvTdAa-1w1QW=5Ie#j2XJJrZ}r}CaBOY^|&3@?yhL+(ssTFQ!?ta
zCPgc}6#r>;^z!B4HxkANvu9tWx;1tS(t)~JZFh0wDCEsQT<8@c%=hNlm`II2)W2(K
z>ArrsbZg`>{K~f$^nWcX9UmO6(64<My*0fv=FH_<Or1P=Zs^c%0is7s2tC!6das@i
zh4|#%?v+C<v71QW!wQ1Bw{kpns?!-NA#(<b*dG57R(47p!3?j077`c1r&m#JTM~lo
zg;+7%!{P6v=rCy<1vS+`*oW6g|HA0ui219f59MN%pTB+%JuAwVUG!Gxip|)9>H@7^
zDTbO0nMSi7Jih-J2FRZ5us%4R7^^ZGH{L(^d$WG<34xUU%pzDPVp#4HYhG=2LBv&y
z6sWpHK=0-b+4;pWoHS|Hku$YA*ku3d%A##&ap%!tQ+bg8@Z7R+Ysa6<usUkv*Uay5
zbarFGvsty;{4cNA<DtcF931t!%^L3dByD4VnEsA~y4_G}2vnYw{7BB%{sJ0*gK9ty
z!Y-Z(OJZ#~^=`YLZuifZMpvKeF=<k9ux-i-dZJtB^ydFwoxf2Kf&X<#w%V&023@*G
z1n<JSmTlly2JgbDt@m!*>M*S+#Cc@MIlgdzzcjhuKj_!6nrkSIz6v$O2HURM#)=)?
z!ZL08Jn15s1q8h8?LzR*WiS0~4v1k|oL=y<{Z%4qr`qQLF0>?pX!+$iQd?^OJ*ClO
z$Z#CnH!{|@?3~lE%`%|bZpvbQ!NtQPpx!#`XpGaFIfk1&k}=|z^?6GYJ!gpKX=p#(
zn`+JNgD5k59u#yDMMvCp6{#2zbc8ns8p5%y=P-Q61mOEi`C{B8+|C})2?0u0V%$X8
z`-Znz8?JeTN`+JqrfCT^$4i}&!W^?mb<G^{7mGVz`N!;I?<9#1`6@o&i29n#h{t=g
ze`n*lgSX|%`PnauHE&&>djrj@*n?;Kt20(3bWnn~vq1Un_|Oi^o@N6@n~r%<8P|?%
zvZ)n0PN3)9b$_wZkm^i5#A#j#|MlDISU2?Xjj%GxSLQY_>}M#w&MMx%*#n7R@X>|_
z)YO?u^C~a1oL+{GAK$x&`J48SdMy8Flz&?O;R;<15-Ufvc-4tNMc7oYrL`cudR!sK
z8}ZnMc!abHBH6MI`RF;zJ758BE#MAu{n9PG3b^ngF@o=HCpwVXU+%4~4|ldouFeg&
z4|eU}7z@8Apt#76@E-ZZ*MLuF4W67|d{B81Q$;`AyG3eqoA%#!=x0wFUJ|hMElBZi
zb(xKm$5*}$a9NBm{;af`jE@}=gTC-`2S3SBVuEw4IrGV42-UmO7)~?mQ^#?`@|72*
z2B2^-NLt0z=mrIQLOV#%k(Tkjhl98muH|TL4YX16$HD1~?6IuI%=3`D#h~e24wdnQ
z!yi?Miovn0u84@|!BZ(#Bu@w)sWwpsJ{<N|l)epZg_3O&2|z~G91X5jklHrNdOoeL
z7e(@Wj2|es`6%ual{OODTJ@1UZ)pdn0rB3g40lab?{^kIJFZP?eaG*%KYy%{`Joxe
z9u|1NSh~)1RLX-iUvlpY_D?c}6Y>X)cOW6b`pRqc_J=u_=PbWbZ~cJPBzV`S#__;1
z;PExJAvs4yX}?B6KY$9o(eOym*w^=}utg;*hVOYpBc`|$HZ#B%ngx4Q_rZ6h0XiSv
zz+w^mzx@%DgcbD_Tkk)vWlA8<`7)0BQ4u6@oD*@>F!6Nt96IbMa8Bj$kYR)kSrR#e
z4W(UslYo9BXDcW%TV8s{?@(F8HC|n#`8aZ<vpHM0RO5fqy(oD&abTsd)4b-Umwwjz
zg5SYLZ1SXc*xCBoRkX+Tg*{@|2yiT|V(nN~p*<jdVi2XuY}rcPSh?Z6W!Bz1LC|Dc
z+H$G<T@L(1d!3G70drXh^^F7x6QeEoL*M=7I7r)bKfGuWCTxbo*Je*d|Ec!Pwt0B7
zdd0d;ml8BUJy&pM<Ob%gf!dh($zO+Lb-7d=#%H=qkuNGP|663H?)(Lf8$fQk8BMM>
zwAs7mvF&dW8a>gE9p2SQe3RN;X!Uu3J_^Z<MrUc$O+-YGr8V;X5S6JdRJB0m`kbhr
zlD%dKrDDg-I4wpvuT13#3mjj(91EX^u}a><!k2F&)lf>Iop0_UZ0927Y`(&GMm#di
z^#EC>rg_Ya*W+V;%Ii&@4gl|$z7wvNfn+=CmKXhXX@$ypS4ye^Zx*T}DHG%t+Kr`b
z!)d!F%f4cReEle*O23pKB)st0yU=Q?X{iLp)|KG(yqB-|lDn#!bJ5%y+7*759a!jq
ztIVXXX(z#{xY>BnhOQkw){(0OdH18Er=>VGa4#{v<X`VZ$1d&rKkB0SmPm~yS04!#
z6({C1a%8t>{Xo15t(ul<*-kQO3)QEF#0J7|GrrDLS*heKFHedOq=g+VmsoP!LMI|r
zPSsaT#2MR>jvBi40Qsdb_6X(HA`su4#nabD9KORref>ZG{G%3fY05xSJ+G)G^yEXL
zV78E2)E)3Mc#rQ`^ni-pxPr?xj}N<2f$U71DUJ&nm2yffo+KiyQ?Oe*L@;{y*zM`T
zoJ31e72rD=<zN6CCOwD5#_YX7JCd-CAUdQE_bEvkeoKj#VY}+g%1tw~!;z=BI^Sir
zAKjJyaoS;G3ATr(fpRsr;-Q+)-}Qnkt6;=D*yg_&@Lue%Ob{hlTyg)+mBCnfS|HFN
z-9SN^|1&3&7J~t~0m_!-Y=fTWj|n;S|6kR_g508J=lZ{DA1oxSZ0y|s3(m(%!otPO
z$&rSl2ZaI2!p+XkmWCPuMg_z?fxE3=JVGIGHD5&97fSC9eeZ`tU*p81inZr#-Wc-0
zTjSjFo^qW!asAwCS$wO}wza6Kw!QMd^8d1+mc~@a35&Hy3_@$>REVjPqxS<*=0TO$
z-~@q(=R=2w$LFUgu4qldycTltl5x)qDQvAq+!CS}Ck%vAAjwD%LI*~arWT<{F6<zU
z%%SUCZt7j`>gzz3wY0pxOs&p(pi$VISh%1@SU`*H?E~YklM9+XE|CsPEEGs`zWg9(
zaAv`@yuDj-KM#;lnxS~p@o|?y_7Sot0vAcXlY`R*MpY;9%#3~jqBAh*JUlBmENty<
zZLGLje5@<H>2PzbAb{TFuL+PbVadD@_!*pQ%~23<aNF%}S<~QBh_cKKzRzl<dJbn7
z4kr*WVK6#Z^1N8`k-?QAoMRA>dT_NOGf+4W;jHdh^Dj;#koR^T5Dm)3uODwMPliDg
z_MZoKMrLNV_LfF_4i2Vp%q%TnU>M)bGP2jRaX_FrhX7nWEr4e=_#IIXVWJuVK?Lec
zW&?zTv<iY5ME&{dl^~X2;v$<MgJ81yvm)W8mN}0zdJQLfRYw=xKAT|YYfN!s9n=cq
zw4Lv+8+mDNXmMfaPH1FceQ0^VmNn+rMf!r>#Ko^56_c-K(8%qt5xF1*5CtqOEC)sd
zkQpKnH#Sz&Pas-&VFT%ziQ>%AL<w;-$D`Q;1YHSN*wGc8<QRN&ezpe^#>vGA%E$Lp
z^`;vpCk4~U&S4*d86q=tt>-He2@PESb13i5li3s8sxffrCuYy>(pmSGn+UpvovrD?
z=kU9s<V00bRRKM@*w<C!S0x5&y*s!Ut4jko&I-3SC{R^(-T~sEfED=kHKfeW{N+hR
z5BL?^$=QK+0HyG!^MH!S%M-%PyRV29=oh~1CKh|EFvw)C$dd7n84Flv{U<>51hl??
zW!-icymuqLeoadCt&E)-`eb$^eggMUi-#JQU(F!$HI3+2@cVWm?M8QiJVzi<TN6>8
za`Oc&Nb}$ZdWb9!{XOlrXZKA0_ab<7nS8hz9|DZP_`;_SJDn?jUO{!?i=ifbU_wgR
z!Pl|ynz<Qln;?M*+Y1vWN)!K=ERBPyneF>hYQ3`yC;|Zi!upQ3aN)qvDD1TvWOX6S
z%!Q#Uf{{f-y*pSdi145{xSSo~_7|@*qXGcZcx_nFjfx$py^p*L#T@-9!Vh9mFBE~@
zL;q<wFbQUq`W&JKI*|yV;0M0k5f304rM`v;pR)oe5&DnFVD;nwe0M!wgkT10t_=sk
zO*G&C1!evdeEAowd>0`Au~_&|^b<O0nd39sOJrml8SKF189K;?<=;X8OY8(;Om?+=
zmGG~xpXuoPCT6yfya!)W+@Guv`o=)tr~CWF&{trdXdplqIQQmz`-u~1!}bFHC3*;)
zuY3hFe!*O)Z7#0t5Py=O&jZ3r-z|V!aKhM(_FuR&hs0PU881R4Wpt_kmcXn21QqfB
zH@&DWGWY>I2<!426cKn9y3ez*|Edegw*Qhu9NWD7iH9+Pbbs7X50?3E4Fi4FU5xZK
zPlcp(bS}&ldm-cl@8iO@_VsT9@=MZGOT?Y`bM%1;!tF(poLl5D?$F`O2}Jpw6k@|=
zLcs#52e>Cnu1%`E5F!wM;QM!z&e1J0OvUB{F*N+`Cp0cTHPG?)@A24(fG9|bTM*uN
z`3E7cu!KhNG7g2WwsDcA5#YL;Xi7K_e+DMNdUG*CXrzF%>?2<%y(h)l^L-EDpyoRq
zL*Y@PSoj?}s~nlV`uCo;{$EbJN=&C?&D@h0rRIs6(|f5t=`)L7#m$zr*dN>n<r|#O
zgVN`Q=#ayGfV)uSHVm%1(1myBs&Dz5u4q%ByR_=zGT$8u-Jvo6pGVoYg%A2S0X)|3
z$|lW0ijZp+&PkkNtnxiWe%TD$be?cosEYw92-)y}ogb{ezFZ+W0yhd|gp$Zt{%C^+
z$K}6!u|Bwu>UC<y4wqvkgG$3>A)KQHc_j{LT;jjQ0*%|+Dj4L1Xv$`yMO<6omV-=;
zrj^B3e}7ZrWUc4<{{S;U%)f$rsP8XYhO#{s5@V#aMJY0LXr+?fiHn|7$6Y-A%Br6a
z@2J$d|BCGqQ8An05GeG^(&<!@Ei|?7yA^TJEsYR#=RxQ}(gN2aUf!>tygt*^_21up
zKgBJw8Rj}^_jnqAxFMZ(f^$DydQVJ9+Q4RM->7~$6;z%4KsZEhFwAO3n#~XW^)yRk
z@K+`JyQLr)L21+!xQxYU&AN1#2l9KLeAcm1pwP0Oz~190MDUFTd(~a%phe>{ZQxXN
zngyS9j}7FA%yT$7n@V1_cEa(`Oc6tLq0w?VumIbjTJEZU5(9&Rq_8|{{N_;8o8o&#
zBEIqZsnY3h2#3a1BF5jnrliUy&w@Yexqg#Fsr)ekBo`AqXU+V<t`VQ#9mX^|dEF(=
zgw3^3IZGsvqKYRsbVAl3803IUgI6veFo2bngt!$*o-Wv=WzlnImN(QupQ85WTJ?Up
z4gXS45xdiWH@T&=sUgoKT1cZlt#dBFw?~-M(c8?9h72_;EOce1kLuV{Di@;IbYF(h
z8SQSO;$E^j-(|RgHrDr**O|yJ`q!P})?jS4ytBhRt?Wl<v(o6ffayW8@ul3>9>v)R
zy<%_Nre+eT?(SLm9mbMz4>h{6-^A?Y`#yKup9@`oaRNPdmH5EthG5!AkmonYm=dP2
zjNDqwu=8DTlgsdF(gM&0v;@Mr7SlLoe$S@Wfz8l+yr4;T<g$R#cKbSfDUT25Td2Yu
zGRgzP7C|b_<(5*P^M8e!-_6l@Z%J3yMQX_4?T6CRN++`RGQzP93ayFt9}1Atu?~^{
z5aMHhlbUd}dy06MgU6zS_i7@GnC<|%I1icF?yu1Q?gP8LI(8bJUeS#L2Eh*&L^Dji
zyhq}n++$J`i1fl-Xrz%GD3&ttidVt|7C8K*{EJ2hr6j0z93v01e`&fa?S;bb0wbPu
zl8k8Onzt@e8Ss%?>&m+{V<lFX{@4Q;jH?cRqn$bohYxMtz}QK*zTLgSQG`jcl%OuA
z#aNZpQ<*xeu`H8h*HDFcme<^v>!a#?UwAREa9qcV6<1FB=uhZlkWR(TQ4qKf6?wQa
z-SrcO;bSZPNSV3>j?o_bPxk5&ya67|p9p3r`cS9PR&7}O!Zxt6>jr7N4FOGe^4$Y}
zNV8ph7rDjSsT5<`#(b0JMp*Ny7O$eua%?+aM>!oTWn-@S@^N?S3I?AxDmuq}Hr|YK
z&==&s!j)`<VUU{e>I*6!Re8fo6qMdo2tqCHrC=mBroimNa>Tb2@pu5i)nMtRaZjNh
z@hs8s5CsGoKKoSa{TKvO6Rh$<LtpcMp@rndNs`QF#MBR>pO_7WPG);(kF`!S-+LU-
zWc}b!lURCk^HWgk$Jywb&_whY&z7PVnDs#{G`TfUOz~zat>hb5D?wbsBX5z*M>Fwm
zeNCEQ?y({A4=FG-y!gfYIL4gw0(oni9A!{#oEW4r7l@jo>Cn~U#L{=sKRTO#P<0k(
zbv1J_aa)s!>nqb=maOQSXGaP6z7RGr3KH54ao;62%eFq(4=ieB4T$>G4Orbi-j;`>
zw;89h7L4&R`#%9pNf|H);5$~lqjHgz#_2C$c-hrG0iB6TQC9QEl9dgOTS(j=Ncl;?
z!QAcXH}`&AB27!zC;AE>LqWrT?YZ1rFgr}nEoTI35h<e5a#Wkl(<|6%>VI?c0dOX1
z$0%!22$<!FuDPakkX$}<C<<dFxSAPu`F?(6Hy8H~3TV(qDYQQExKC8ZdY~Yhd4bRt
z&*AXI2=2G@P-fP6JVa)MYx6HL(3gVRYlycti!TG3Tv4|-`|WGnr#BUU!!l*gi?TQq
zs^#4iopex@?fzyEo`Q2_B%|l3qgX2xoNg&d`AA*~{3JuC0p2biR^<e#Kg1~AkuZQ@
zNHm@wnt!cj9CJ}>7Mc`W8xhkl=q7(!nzv@R9pea5l-VLF>$jv(XeM}_J!2bnGjp6%
zl8N8_VOg60LYiL38h)IAgispqnG1Kg*AfpSdfwWfmke4S5%5)Px<|X?0&rl)<?1Pn
zVkGc!t<uFieH2CUgJ|J8>@;1tm!t4P@`;JeFZ!DbR#N|psN#4Po5zBm+dkw`zWQi%
zA-Y6a4Lar*q*#yAW9Avd_9H3Q(g2w{fJsNrH+MRfM=)gXyD)QqNp&t)gC;PXO$zJU
z5ElVd4=p7EJ=wW^7WpG(`?NdK-*ntV&aw*a;~x!u$9&FS^jD?fUPS|>gakM*3!xBH
z5j`A>>rcI%WxpoLE&7#pl5uCrbj~0^oZauYUG0Rd^0+GiB0R6IVhg--T)xs0_cDov
z5E8QIF6|}CP$c4iVdfverM3eY*nMl$lBQQ3`!dCF+BRmC-^1bydsUGd{YH&yV1BTj
z#066NzqpI($kn5%@P1yrP>O1w%XBcgku27G(c$C_O(fr5a>3FL{xVTdi?xp}v!ha|
zM_BJ;(~?>thp+j}z(bYKZGiO4nBpyYH4#D2iCF{Pf6;w^qRY3dAGa9xx}2j~Nyd|Q
z>EqakU=6W$W5G6CeBjZtR=V9%BEESLBM{}2;&<71!x*pfJ90ZHGOjL}C}dd@Ma_|Q
z4@2alY^@NZ>3i>Y|5I8@9N9$~69qPu)k^=Z4FHY&Y`t1+6=OZ{y~DvbJMzjIQTa!U
zd!(wR6}Di1EA>Y0=hGR)HQuQJ#su2Ou4m^O_-<!<Osjh1+Z~RE!)A|CBU#7Lc<cvN
zX=rUesB)Ka{FmUpNFj|d`G*KOEV@>tYI!L0^D|kKYZHlC=3*g){12lEwy5Ky`UYd9
zIV(t8B1cZTw;lmH9XRPw<!cGb%&86>1;+=+*m6&QB}v$GAJP{jTzv=6ft=LQx>vsS
zTHT6c#CC&V(F+TWMGjR$gNm^RK|?6xvhmc65ClKrCycPr$|6?o*(JRTRl)G~sneEo
zV{E#~2lWKOW2STFsdE_9n<4$y{8l6$I$h-yRAZ~e9Hmo6ln7eupLzm4Vc?$qJ5<*?
zn{L>DD%J7gD&-y5r=y>*P<%TnMFN-2BCmc^5H2umUJGXLh2me^9?B#ylYP#ds6w7a
z$pJL~U>g@fEm0ferk1XuX*_Qu8c3hIsxV?zlv<75yl1k?QJMv_7?Pm0$SzT=hi5%&
z^Dtx^s;trQTa%-=mT3xV`88m|K(@ln?vvktU&a4?m&BdN%-vJ*am1E!UzyS`d~d@h
ziBMXkk?@nv9@smVFw0@DSvRJ#1aEb=>O7)nEKMF&$_2%GaI7!i$c<SS{{8!HjNH3L
z_dZNXHsT_|31zGVIF5SI`|>x*twEQj0BE7FDtsHqe9}NW<zJ(XLnsfvTfFH%K1Ak!
zjI0WLx!1NCOm3{W8CV04K8D4d+34Jy59B$bc<Lf21Hw*zM4z)HLLj;{Su+XBO^S)k
z9zm!?FAb<iRCFZ}Z{RJyouH6!3tAS)do6kYMrNT2)u!nP@hazlzj!2GRezE>C;LKn
zKzc%Ff=i!H3Z^C;7B?{rj>sUyj^4e0?Z%yS$wf}k^=?3V9scp#Hu)gN^DDwo;*fJ8
zPK04zGjsg9WLMn^IpDM6y-umlSHsQiIf^@nkDat*Lb1$6`1icg0_nCGB5J5d$4lIm
zOgWsT+qfc)+saaa)Ve}m12SB7QjBdXA~!Tv)4k&3a{Cm^#E(fKCy|T*E!Osbkf^Hm
z)z8}Jm1#Q?1`w?(E}y3J%?}%0_ov;-Uyak=!Fxx}Lr^Ilhu3}5bs&I`VGlF9G|}hF
z@O-;h25hYbLQ@zkE1tnQyW5i4^I5Q;YWaTUT9rTjbTruGIaunHY@JpQH)*?(Mr+Sq
z={znRl>!s^+Hw}nD0FtK%g=a!eT)q)>8(ORY~Y1>ziErO&}ivQeeRl$`dXFj?*BpJ
zC?#mkuCJbr7_%k5-&=}qtsX~2bVATwC{00s2n?#7BN_dLZ!59JP<%f}T{noeaa#T2
z(U2jKEL6r;g2mV{)&Rft1n+$`d(DJHQg5`}@;i;^e%~113OsCT!`qsFE+bI7Y0(fr
zam2`fP6(?e`(@N@=-!bYeLT-M{OOl|gJXKuFO^!H%mX3@1E&&qbFiArsZocw7Rvf7
z33j{}Q`Exj!nFvhJUQ>h-z9_W*<zN1W?!RpvK^d_1=lb6@6B6;3G_S9!m!|NCG@9K
zjVy2F(`O?dNNm35J-IA@V9n98r6vAO{k}s%eu~8PIOX|v4*T+Bf2*=lN5bV+?L>nJ
zq07(e44%>rO53!47WFjPg3G56T@+y!YlZClcRfoh(b{u&l-3nW!Z#AyNlKeWjX8-7
z1?CT!a;T~96~055v->{RTJ{~MrOPDdQhCAt6h+yhBdY0iZh>WgKXsqW_LFFDNX)ar
zz??X=(v#Fb%<5Y2hWV?{e280Cc#ik?tGsQeHT_`z)&hS;$p|j3_A5u*q*_&zdzP6A
zKiDV~JHj7U4e}eqy&3m8k6a$d#1SE!#{n<j_GsOv#hbN*%Wb;iO*A=@;P)I@LqE5x
zwr&0UwxSW)zUKaaYy|foA!YG_)i)EIC{xOxm_1F7l|(PSe;5yU!wxRo`7CZb)E$ym
zL2$IaW>3Z5wndsK=G!)%`gO_8(Cbk%q#&Ph=`;V<@HAT-8#*6i4x36l8nxK3Q}LdD
zJCzRgK$vn237>1+3E!Nzz=11StMl<lm-Vz4*7=}3ZAL+V8^PxxuOcnG>P~2(m(`yz
z@SUguEcO>6G3~VKq|c;`7KAX?lOhXLOaQdBBDN2WpbzgBRbSznOV5_xI~i;sR3~jm
zs3zw=!ACO)`RM=ydDO+)QpzY+sI7SyXx0@|7omMW4z4<G|Ic@E*vF)A*A4PFuRRud
zTM8URB^{W5OPz+R@4E#pQD9EF5?Ah5dv&|FmsIDT*RD$1^jhK4-}CI7;ptm8KOvlN
zTiY<KTgcgnn@AyABa|ph5>=@|R@8q%uHxxUdyo&tDan7&ltB&WBhc+H5Fsu5i6bxM
zCyf<;zE8a%%yehj?J1x**qbc0OTlcIw@qoJT<;%$vT4&!3P~5ZeU+BWi;2k-PO>78
z;;gTld6dT5h)6<pMh68fkk_Lsc9!zGeGH<1pJyp;AVc6o0E;!oQGvOr_|eN%spfjC
zKqkKM>`OFxadrGrBBoE^7fBAvlr~wg8w=(KwCz`f)LSi2&w!}Dn>7-qAY8;#o{<6S
zMlHC1`kpa;14>h4B%e0lSi|iD8I9Q=$7(qqAy;5`c@e|%nIU<13{rqe;F&qxKDd=&
z<I8@<rqOnONH-Kcvz18Bln*9^NANx4UJgR1JXD4T7KVmVn?uYb#$j+6PNa};uR4RV
zAD9jWe^I#;p{jB>C8iutkJyQC^Q3IH;}_3=FtK`a4upq5F;9b19Si;@3Y$Vy<Kb-T
zu`wBO#3X9F_+3PU8REMYeNi+VFxGivWaO9O7Xxu|<P{FGt|9^LqoHzdC~iM)%Ei{y
zeLqWPHmOF-FzSk8R|rFaQISOSJFfx$FI#a4m0ecd5}u1?3|My=dxgC47}N}cfmX+V
zg!x+b_MfW1Vu#^kibhnQF(7l{Eqh(5;T_a!vkjDbJ~WT+?s?wm*Aa$;hTAH$UAvp+
zHf2S;d^<QcgQ{MJl!?XZkhZ)q7S-h9c5e2XI}tkCYRV=}&UIH)AOpc`2&wBKxA|ht
z6_<^?ZJ5t%WU`%WEq8(A^J8`4tfj+$rH@q52l41a>zmojG?$97Dno`|?KHQt0RPQF
z)GxuGno>3#2xW1ax~ZQ?*hl!j77BBV#`tK`>+hv>Dw9j@2&?xy>4uA~86$Pplfuoj
z_-)boJFi3i_P8<J!Kn^NsU|r7;qYbZw5V1-ygIv)@_Gu-6!yc_AikO=mD$*TH?G)4
zh0gu!r;S;Sakd)0W!#5YFk}zOs?!&MxqQy__>b3Dcasi%m3Oiw!DlAZuXp$&T$20n
zuY<=TiNr61#+qlb4VbslEt=xqA8ImAgM;^g8t85M2iaCuN~IWFuh8bUB|-k*B?RhX
z@!4lnU$h|JR=4CSvDcKK^<2k)aaRK)`aILA#495C@+?0BP{IxgF4G7_N@|*<yb2TH
zvwh~4)2mEmO3f|Z3IQ=^9zE6#%>aZM_NS#fAL~;~c;$S?Lf@ushqli($Bs$MVaqRC
z#k=KYOpv;7J-Rq~0;G*tX0)S#KCg5W?ax?cRx2Uwzx{``*F!yiA^&8520#~|2Yz`i
zZcAb~FrEt4x$fe!jY>JKnqBNZXu%in<@q2vr;R#|foxJ=?EcA*TT?p+5j6q~=Fk{j
zgPN;NTDas{@>ZJp9D%||lqj0Ce#<)tY1<D{PNoh)xJireE(fo(2PrPj)*ExQQbH+r
zjW0qa5<9+0(M$yfdwg$y*`?(*O)cNN4vgmq?(%s(m#*G@!iCgmUwe6cpfgG^U7E5$
z&-&F?BiT706U4F4c<KWzRmF42QiCd7!UtIbdEPonpO5qg;~fUM%ufp;Z8Y`p@9<p^
z!!z#=yk%z16N7t1TlqE})07A@J_0KvZYf*N?gsMYmr6!BJd-wmj}%GT`b7b~6|HZs
zAz^CTJ#xnTr^ia2sc&m0^<glR4X%8vtK=idRJsRpt^Db9O<ZxeFI#$Yz0>Z`lEy8!
zvV_H4q_vd9GHZ4laSUnOHg-MO#IimbV)R~)!^RqtT$Q!%tT2W@Fkv(~pkyQqYsvf1
zxTp5n=NKKZQBmW6ILK^}#)Xrlub0pXpc6Mqu=tsLz+%K?=D<c)Yy-54Ci(MS4c)HW
zkFf(uUxIoyn5`3ghr50eB#UpfvpSt)V~XOq=)$u{$jx2Q>v7#LPTZSi{p=*ISMinF
zkX3Ci!U=d{a}QuNgT;U;o2_E1gZ!pzSn1-x-?7OVc2l!|)#?!wOVE0#u#?fyVvX89
zdCZ<|PeZRj#D-9E6B(4ca6@~Xu8a8`=7^LO!I=UHhC(ZAXyu<Kr>AX&)$BL+J-E+c
z2fGuJ1UkRpn|bfICJxMy>!Z~+Gnq(Or8g7;O+urJi_>1>)lWTjVe7I3gI&tXbKR;I
zMn+xv;6coP`SZqDBn@6W(>`5m*&M3Z^7_*cg;Rb<U-R{{xz(sMMZ@euCH;NwAI*H(
zS&v3GXD+Cey^zTu_<P>>yrxJli=9dO?J<L3v??fs<^7ANJ@6dFuwsuXZ<gM*)WLOE
zA_@a?NhmaPD3jzOOhR;+({R66(A6nN`vwy_o3#3WV@u(hh5H75IX#c@R?Q+UU~Qz8
zSx{s`8WNPlv_76nad3ld$GzODxqs{l@}#^tan{ERLA9%$gVZKMg>Pub;>`E?%6`l$
zXt#c9a9m;WY;3DKK5&#uwWzz%1uLyR_*T8Dec<<Lpo@xJg2#LtPLDUeQ6N4NEXZ{V
z21uWO%UEpH)FF~<oRRK=aW1$cQVJ8xX17VKO@@Pgjcc)B@T)4Nc6kEmXc_p%!2*p@
zArydeQ}K0S!8R)&eJV~?Zw?riiWY*P5IDJHP&Ht>4=%k8qjW9hsBLv~`^{Y{;2Vj@
z%fy;n|B_<4f)@6+y!aO!=SNdaXyF&=$PK4|WL?6!eI!VtFMP}#uF2PLzH(3@DUjPi
z)^yoI?ijrCX+fdjT@vXOoD*|QoRT)H&#4`&v=8q+7v@{q%lV&M_L2Sg=08dn?>ydK
ze8!0=G)+3}7j}6coXTg~!nRziZy<v?xjx7P?-jo06;U*48@jdPCTF09!b*#;kDz{k
z8n|HHpeq2|4rij-BD!qet%8SB;Z{I<klRbG9+c%Q)N20a%<)9g{I1|-BAlw^Z|^7+
zaIp|)I-=+r^-dg!<gPCNk;)*20^quBlKMa~#BBFt`I^0@l-z9RoKIRFmSoyvWvvYI
zd*<UuZAU@u5fsDkfTeV>oxVEzgKvU=6E&9d!#S}~Hx2T8U3`(}HaggG!g7k~<adHU
zPZg(iFf2CqAKf5Nup(H467>7`xgN`tBF&6l^r?nj$2&$z^HF>*r!3G41%^bx38Y0m
zhjwl4ri^fMVk!n>ijzj<7&uD%)(t0!BPj1N3zfE1$n+<{7p1f+zS|(c^vM)|ac)8y
zoc<UNLlD;$pVYZheU)@k{lzkZxJcuAbNE4gBf1d<QvO@~IF4XK9GMr9phjCxF!9mR
z<4DRR%@G8Uz@64vUU}%aiaGhys{7Kt`--;<S4#M}&X-)%?l<YtnEG569zFZp1O<*1
zUaJpQY%)A?_=g)^kSUe4?`pVzXdmh7`7NjCKXadB20L&S!&hDw>7sL2xChj6mE`b~
zh9NM{Go}^IJF0Z#o{o2~b?Z0$a^~*%mWTF1|CF~Yw#4yN^%2n01eOooQ)^557_iML
z6^CaZ9QL>;`L;W%siMJVx_~~@HrjXQ^veaqC5ITbrv`QqVQvb^`1_fE=I5W=9Ze(j
zI-D=G`P+nPPa39tLN(t_h#04{7+-ZdQ@!|f=@rrnOONbdp^f`tWK^Re?w-v>XtW_F
z8n582)&=>mqO{x|hElYDTEu0>;4QDie#eZVaD3Rp%<<Fdw&Q?o|1r@!2KHvwO3vmw
ziCk-E`0#UGJ7zux?5$6Kj%%XUjXQeCu{_At5)k72a8=t`nmG~0k|50&Y!E8$kN|mK
zVB0GyNL!F4cCxkjl(fdn3A`a(>{nM>jZ$8t0<zzA>;gkHI*@--8BwE06NB@RX{dx0
zfw2a*)6!=k6V4vB*CVGLcs3?I!}5*vRh>3hn-vam`@FR{{@gu(N((#U_NfZ>39h%-
z8HySE$R?>ZB!Bd?1s(a~6JcDCm}1pIK*}nw5CSXzWK+aIXjZf<gQUm9&#`+}yY@<p
zb$uD3DfsmZvo4Xk{PTJj;MufO2qHf^IrxosJg+EDsBc)gL<Yv_&L(YW;$+2e+?6>@
zJ~hIO$8Aj4y-XE<@s%?Al~SLpWa#{~!mrs;`>A&cMfmHx#kF4fpk}$GDQ^qi4KIuD
z;}HJC5iim2{;(?c5?yPl78=4qM2B(E(lexnjR~K@`Ohr!2ZYW>`nZ0*o%_zw;DcFU
zl36BG5Efw+JJN=Tou&sjH@$WZ5aIg+*Qz;eW~gbd)hN4v(p8$>Z>C$mG>#lZkj`IS
zyV$?#e!cH~$hE-jlzLQ{wVG1GFV%A`^-^T3qtT*f!`mMHB8(IO|BS3RC?or9oD@Fq
z`6BCA&=zEdfRgD`-Y;)}(e?<IVf`ikXU;-DVFt<mVO$2Q1q1_s`|M%)e#vYz?WMaK
z%Pw)rXbs_ivL1?FL7CZrpCv$KY<p`WI_NWy{yjl2k>#*&)o6rUjC_<Om-)P<+tZ_c
zSblSB9hZ^OygX<qX@?u_*4gAr+jRAUv#pcH@=g#gi#SpyL-S~*zhQ&_Ew(H_FV}Vl
zF130RGBEZ&R32wen%YN};G1|L1Y(n%i>Ib%M;I7?0~e}Zv_0KKa;u_swX7Q2W%&?c
zDvg@e@St>NTZz%RdUM11N&rER|1M@=>bHIDVmY?E`fZ=BI9rtBd13~2D&RN2C<_J3
zxwg1Ovv>?oyixgGQfp{6Qa(E~go?8Kgp&C+%=$EHB*@rcQ_e>!@h@mMNQYt`)3Fr{
zg_pj6+mkALz5`q`<kzz$E+h9+iN%0&jcDI*@D=GBWP3Z`5y6lK7k4+8^usIh{d`$}
zE@42VTQ)-WPdbM8GIV|YB@SUgpUw45B3PPW>%87$8yK>8xA?{yV<cbqRK|+Ekv$DH
z^8uRU{X;}}DmsF{Q{DmkDr*~wQZo8df_=Y#Xh8;_e#Jqh82T&gEn#uuY-WE3<-1r~
zVtTOGT|3mleKTg*(J}wMX|TjigLJ+l|8Uw7zPla7b6^|Xd6V&w+*|KC{Pn|(Ec*b?
zwz=o=7XN|g_He-5R#4VxTfnbGS0~&h21?GbT=c0DTAU?hyMRhti!-*QB40V%S~hNf
zi$De&y|V;%aUNza#oDZkgTTF>v5zFS9A^~RPWfla^DJ7Gm`1A@LsnOYCuTMy?!T>>
zah9X5c&tA4!UjIm?&0^cR0E;+5Sb^^1XtqB<^vE>J=*V$h)T>F#&^Wnpo_$!*ubx*
z;h9LS&eSTqzFYXvY<+$b0$<C80Gv91ev;=(bk-FqFpS8)RbK8=+c7koG4&NxI^*<x
zO?bBHMOFz!cjLR1*}wGmJLP0cAPr1np2kNwqcSPXZ%rubC6TJdf_(2|%Yj??i8a+u
zF%t7ut#OSk<DF9j5h!-~J5`C5?#UnXx3^t^8xHGN(RsRoQApof*097d^bYucL&P06
zT7)B5=q=Yz#<CxUBqFCFPD9vlDn`&0aU!AtXz@>C8twDcQDOlc{DO%S&A<0O8%W%1
zH<u?%59*5hQH{qapgi`r_ndHhF6%pV(>~I`F6Iw^$%Odg(N_}z1#MM$)JC*ot>h^!
z9)b3pH!D<kDSAkgG*lL>exJU71HVT~(AY+#xRDUbb(5EOj66{xGrPNAb>}-&sJbf2
zq)X!0ZI*+@ny}Km!3y_d*atS+_;b}vDkSQdDVt!Qf7Fws5}Z!`u>j%r3=@Vz19_=T
zbmLI`TOGWN7L~IOZv~>=E;Duu#`3BhT~lBEG8eIYXlRvpEk)veX^1_4K)S32Evi^u
ztgX3CRi2bDOW8(1oGXf-R6tf|VPei3u1qX<E>f(H_}6z&2YzMsHW?13Lr7~zEnz6=
z`%q6XGpa1<%~K~kpTH_UFGI2qmX(jM7sN5uJVF$(IogdcaY+}!MC{$i&~trQ$wkli
z@AYN$D<6_#Cx!ViW;#88+r?r@cQTRRDA&n8&CNXRwVK!s>GpN$K6#-wi?BEd*E#zA
zfRK{8*RxxFdGSRpwN~QUE$3)C!yHC511u#|tG7cDGL#0l`@g9E2;c$y#?zc-d8S8x
zT`F{Fad8bnL$@~qBwikc&p__qm((POpA^A-`?l}IuF;tO)t}>kp?3I(O7DaMt5a*n
zn&nQFh8xbcIGN6(D;C#X8WlFQ<Zs)bvE(mqqdi%lB$W4XDTA7Hukh<)KI=2i%h?zi
zF1uQzHGL7hK~I!clsjel)D@$2HOC~Qzs1(88~c^2V;kG})$gN>^RzW@Ecu=w?_m*w
z{bOWX7FgBFHrf_{#l4h?#)6}b>^Q(Z-!yhkOb7znYAv1OA@t0IJ_CuN%E<z^xS@o<
z?ILAPP}o3>sKP<__xhC8$uZB+@sFil0k0%kO&Fc$6ka7rEHQR12|ZLOGGpG{uOZ(y
z&7B&XGl-2#Zt^2d8f1+DoN5Y*P?Jy|{lPOM#f@?F^33voMy_M}pSO5|s@1DcGL~hx
zMV>7i8JD3Wm#b|x31MZHp|1<pfqbDBBKl>jac_-ZpD3>{$4$$Y7+R{WS;>mI{adG_
zyEd~gj1k^frosV6b~4h@x!-;4mABbhtRRv{dsz@^a9yr6#RFTLk_{@gbfYhtM$a@F
z2LVe`iDrF&jBJE!KKW6HY%63&WevGi=Y9q<errE8J|^ISkxIZ;4R$1T&O#cl!tiSc
zN`@};_O5VBznKfOIw^056gTWuJooj++juzlb)_V3o2+U^#u!-<s5l&Le7lIq3)Fq9
ze|YP>Nh@E}S}h@a*pcWeoA^z&Z_~`BHl%7y8=q}|O7{^r>LUIen(-K7%IG5IMu|Vz
zR2V5@>=VB8w6^Fa=B0qL$K)?ZDTksVaNW^d=27l5SbVs8%JVq?rcUQ0<yj{DDm0Ze
z@upAq-@3kzRm9k)_|N_vUDsag6X|J*8pM%wphF3K@ro>p6NV}PJuQ!VdYI;B@8b(8
zPFOvE3XytJkT)VeoU|zayiIS_DkL29V%&3YYD)Tf6I4Fhl9@G>!&&4*0~Tg!dMb$P
zQc({M%%b{P%v^A^^tie|P`k<__sO2M21nW=E}zR^nl-955X@l`@QG#$Sjy?o)T@fB
z1&0Zni=c)LHnjVY${mG19<`by+FrjOl?oq!SR?bqBo``%?p$JM`nk%}Fs@vYkHHI<
zFD2#9ew#p|#bVXGo^s|(jDfz>;vp~0ytmW<>T<D}+CEestuK$K7hUwX-lo;`RtH{c
z_p{h9D-*VnKs3aIrnTD8Jsn^D`^|Y@=#ZS({*NxmhM!Yh4|mXs%$}1BEXv|WJ`Wpz
zsL(ZV#E6GTQplyza;hW&4!?0P<NLN6u?|B19=>T#1QL2U2)4{q3TBxhd~)w>sJ(1X
zed>g-Xf<9Ta=_$@f0u*OlIpF@Y|*wf$NpY!^szUsA%X40=wT}pAEzjUpvM#MmpO`?
z$F--#Q_XUxU7Qe|`97ytBk;WK>Df1b)Km>$IbI)De_QPn!%VM>?=tSHoriZz>^se-
ztd#$0<GsWF9s{?8obEN+0YkOa*kJu@K_+ta3*;xU+KaVWe<W<qc~#98ib_j371;rS
z$da|y@jK&_SE5d(Ws*=Pc1i2JAAN>3PJIhfUyxGa;4POBV!0~0=cQQ+*(S4pkIj3C
z-E~nMC#~HFy!KZJm*>ZC7BEve&>4$g*`OpiXLe13h&&vtSthma%06rQ-wX<eh9y{a
zG_qkbufH9=vHfc60XJ1w+8!WI2g45|;51fhdKMrrd>RdBGTiz!d~mpLv@0jXvD6bi
z!#*uJ&+VLJuziF5l^oB{nX<BfZ4QG9-N6WR6K5lW+Vx@Z?HPwpe6Ut()1Bzyi1IRd
zq4;vjQl{Hr=s<TLiw-M*p3;h+&QNAAwKTDXr)zcvrIl5SOl-}nfwVV(L&N=MAD^LV
zQs7Fhi|cnnat5L82_z|RukTomNCpm9(^ZDP^cALl&WpOEmP4gJo6S9c$r*Eq=tW7V
z`s*eBKGutC0~Kp2Fc=_&84`E(*2hxb{q5%W8jC5dfwe5u56!s3;R`frHqNDwt>1^w
z{U=6H7M~wTqUl4CF^L3qg-#}6e=FKCIpP7@JnTGH+xotE?fG5r;Jv<BS7040BRoMr
zgF|t=^rZ|Z`lXAR+7tPI+;%HO)55<k2C70XQUL0<^P(JeOxzKa_j{El*^-|4dqqE(
z)_u8IS2vfHez#TQoEDIdBwQvDDkP#$-Nnex<Rz}<U+bo-)vJ8v{3wff!8jjsK@NSF
zim?@v^DR?5<@<u&H3p;L)b;jz{LVLm54oYmzF*+i9kT1DaYyif7b0)&PSV{YnB^+@
z>vD^@MOxm;hW!aX@xR@Yfs3G!5HK66bK#N<op{|t(o%fGw+~@eC!fPZtrYPM>C{B<
zUp%<l)HgdFWjQ@}x-&mv=%209@-QU*^1_m+x2p@5&9sPJUNApY5VYLdx4Ig8HO1ua
z(54=@+?BI-oM0q>B0fFKo3S5;j9ls7D12zOC_e6}n1zci<ui>O4<%OWZqp;JlC2AX
z>Claw`h<#ip)r&GaP3+SsRJD|n4gl}Z-L0G;=rlt<e3-dc72d}giaMs%P1M-Wjy6y
zRA>+ucxaVAJB_$A73cg_r~o_y*D{VV^9526Ot2qfwf`}HA(b}1?Zk246?brt@+ZQn
zHXp@1-h{C7{M&W>{OyquNBHumXVp3MCNBqRd)iS1T@|Q0^$ABr9mE_H>YN|oSNaf4
zn64vRV@%}H6|hYWtOTpO$Jm=iW${omGUEI92Y%eQG7{P$-X)2#8c%@|S`y_ul+uj@
zA;<a5C%8U;Hz>isE^Qz6S=rX^G$rDK?j4>#U6tJ08X{H6?B8RB0K+lIQXutyOhZnI
zlSs95^|m4SRd$wtrG=<WacIxKo`dVF9-><j;w3CiVo!(-J|ie8dh7Pyob^7=T7C==
z;IJF_-F2cHLA!{8X-HI7o94Y@<%v0w8}qO_5~Ij}tjWD!A0!AG9a45*RYd>D(QSdZ
z;HE2q@~Oudfi(ABEF~XUkGkKW){$#(<IYup6bVN+?~gXYQHu4b?I8X2(o4FZSv^(B
z_8ph4i*4hmzLK^{MpGege3Z)1#GEnoqSc4H<Gc<$DH>+#JHB-I#haOAGAF8Vk(^<i
z?f0&KmtO-V1_(c7k<SdOD>LI66`^&Y*o(FGZ1l{2?ozE*0v1qGIvy3cu-P4AD8@Pf
zPj#g#b{pZN00&;rF`Y3dGYz>$GA4yo^|2p6?ngv036Hpk7ZBye;8sks$Sk5!zxb{j
zE=B7FpXNd<JXVkVd~}s{Lj3ZCM88Njx!x0hYizSjr87V%t`b6F*FS>>)>%Kqhc#Qe
zNJJOV?m^*hBINV>3x{+XT*FgY5s|O&Gb(@6;^?H@w&XTZax)`V%jzlWrROW!={qVj
z1Kg`MJ{+U(Lg1R7UdQ{~m^m!BfG(%WmLpM^2>+PO6EpXTSdeR4YvNZa*+pw9Y_dmx
zuL|Q#rIaJy^3%8Hu`)5Gaebr}@bW_l3M|KA%<umB_KHQeD&Q2a%-yX~d6%vkfy(4z
zc&-{-Ym_zy&zU$(_p|M}zngjO9p5<c^{ZLG&tK>fG=Ip%;UC6Z35o7NR$xJzr!T6`
z8H#p_^bo^7)aRQR8M|64u)_>KiX$C=A>5ovFx|LM%!AZ8{b<r2$7mIX$U^kQzRO|R
ziC+SnQ`UQf(LEOBD*iHNgHJYrh?}YU9Yed;HDH}t(~(wk#Z^fr7k-U3K->M5<!8t}
zT*IX$<6OtG)z)l*5hh-TpB{+?eF9yf4#}5~-wTg^1i1}`pz%0^1L2~9VWk6qlzy`p
zU-lHW#>7Y~h^zs$B`lF$RV(G)jK6=6TK{AlL;lpiCRV0q*5m~HO*W<<tPDSan(*Tx
zB{QksdM^~+ZGi=H6!U}v%r-rZW8Bb_m~Y1gIJY3v`h^edE1&d>e4rb?39Y~^m!^Kn
z;Z8Kw1p%e$f{GhCdL%-W5Gyx-cbJ==^a*FD7|-y*Jbc}FxU+Ko*BKmVWJNee$V)_$
z>d<v={95Y|dLd>zIGAQOr>^sUH!mt9H8qTj@^p**>ukW2??>wnE?g7AlFZw>rh1fa
zrDHaP>XooY6*fs!*$_mP?8lUvW~z!4W#kb{INwL-+h~Pm6Ohm%&eXks!zUs-anVqQ
z?1E$NVqe*1Szf6otm>~cSrAHWu-sF}*5|z1+Aj$x8gK@2*jI#fccQK%n`{lj%z*o3
zR{EVaDYv%j1un&!LJoLCGf<xZpA(J~Or08SVPs>`j^nWEuYQuD3J5^U$Ehc$WQfk@
z&QKc>leBF;s?MSf@?SfDMaq1LMAF#E+4Xc8y)dU9Q@=V+tl(9;(RyZ!AZ8c@>s;q7
zmd&`3huSx@z7X=OebG2R5i|(CGb0*o77D#Ft|!fqS)zH6t8}lKp$Hl8I?*JvxmU?e
zd0jqTagVR`wg4k4ziZCWQ*tw_=;mxykol&}Q0;UL?Po9$KG#iuc|HSyl^o)zoK08x
zhQQc`1kC<orEkeB3C2(F1j%^*68RfFc-w!l@~EOmh@=$)p;rBHC!nsxZaj*ZPdX6=
z79CzS<=CkN;k&j;`iL(1B5_wyT|4G|N<wyEOU#>?6P5qclA;Z@8s0vvEXUN|_Z}yV
zvDdSVjbH3MfgW9dHTfFX#N-vtWf^|H!Dxt6zw$*5UwuAL&0TI5dTy9O8>=21Sx&Cs
zc*hTC8AJ1Jrz(yWt&-@%7L2OLLxd7(8s%?iNjKA)6_RO4>;{M&U^j#}tq-BM{YKR?
zEjDU5VIk(bwJ~q4C^bxy{F(8iNO~>><9Hlnt$mL6lULM#yE=A1ODdiuLI~I}!)e0P
zJ^z<m@EQ)^C)Nj8PW0#RVOZORL(cECG;WdQ3%ZCtzZKNM)=<lNz=hwCrh7<yzFF(k
ztE8^ViXwC^&yMG5dl^JaV~CAW{!}paV>?4R_!KF~JX@Gj7FK|LN;S|==eC}vM%gH`
z#GBFgA>A5(hx+;=)K)6e3ms1dr~c7XK7on+hp(ftSLR+4B#yv`7gAq-RQ#y;bESbp
zkz&tYb@PyGUy9w-DJZ2|cIcumop15|8o;R6AvLl0n8uOMrLC~0dg!_v9mqgBRa>VT
z>FCBA`~WCQsmrC3rOUfW^46;vLFFN5yTkjR6#F)RStfKC@=X;&7LyEx1KmduNFjp-
zMl_YUloj+u_iQi8S?H>E?d1}~qQvoqZJ*{o%Q>w+o$5={CiP%F&9a17w`V99z%!#~
zt$4K3xYG(|Wvf^u8-)cWpYn0-Mq_Q)#E#v4b)_(S757`!(S@F{<J=ou3aFtQ)imCz
zuf61dV2^vb!N`i5=DR%Bz`V+(nGOj>7=y?>_|z*t&iI?>Gwt$*PPl5ii|k_$Yid|Y
zR0eY*)o{kD5wd6w-8s$=)0_@2vX}J-{;uL!71#SZqSdmPBas!#Q0KuM|4HF@MsQRV
z?DHIb-nDTu6?0{=iR0y+alGQ%0-sg$FCON9;G|x5nLjcLKR1W7-H*M9o6Xd`-Oex*
zxG9+L<}a5wrSM<#s=`N6U{M7Sjo5cw$fdY+MX7yVV%PfR)@S++X`E^nVWQJtL<<w@
zajKPu-)8i_R|ER%BE*>AmtJGT+ufS{(gfCBZ<@LGVF($;;D_3s>%c{(#f^e9pY=C?
z74K#xYxA!0OZ70OV7T86lnC&E=XA6F{NjfDDPJ!~mW$f8Zl~dRerKq5n(!&UZ7xL$
zV3@f^O%}yLHuh345ItXH^fsI>|64y`DeGf{S4{8JG24LxES!Ivkp5iOg;`7IxkxNW
z9}G<Z67gWtbRJS>y`u}ZXz;H{d#GQ3CRlG@ddm}^2MPQ13*Y$7Bj6)+m95+!SUuhZ
z)SUXbD<>cu_>Q=6KmW?~R$>^ifdLA=@?a2pj(D-pY>hXM&Y<KevZ1)kV<i8e%=79X
z&xdlf4_lJO)q+xG-dsy-TQ$it0w=6)c{nPLB_<>f#Dg(zrpRhyr$^D;H{>sWrIfN>
zm}j*z(aAICbxmC(!FX5{rTjXGGoB**;$P<hUH&D`=`ynz2PubMAiD!!3W}PvHHVy=
z+PXg?{6*-D{Mmw6VLGmqW*DjnEpQ{7Is|2V&A{qtFdq@DPD+PUhuc5Crw=F$kR+?G
zjmr|nsDJ1{7RaoLC6FQ}FsrD4J-<JILTm>3c^xGM4XpvgFp<HIDxt-L%9ukL-mwuz
zuae)rj*uvp${Byx_g35*XfSjhlT1Di(xymDF?O7Z7QE|Cb}@uY#M}K=(d0jqMuv-*
zh_=U3-0fr+R}OAo!|{4!sGUDxg3Oj*@m}-zJe4nn9aS24Zx*geO6iJ!C{Sg#a=<`O
zj{Hs9m81X*u{-#!>2!EgnXno*gFbF1z<q$CkDUPgp1khyIl2h+H={v@G5a!j@IcvY
z!j8$}0p57{C3=Z}GC)+8@=<qdjD3e_uybV!X90s>7kHW{_%w8;!NZVR*9loID8kg=
zN2MO)mO}6uWh$c!xor`Dx|<QYnHUixJ+fL8f*w|zD`$XqgTM{Q>6V$r6JBKbdiW)O
z>btkrcn*ImwWV6v$C@%&qL^P(p+U}+jg(wv-1_<h9YX`{iXt>3oo4kP7Lle`qqDF(
zdI0M`H0xI6DaB;GzL|1h58jQ@dk7z?i5cVhuUomlX8p_*Ryg&4zoKagk>52Fi`x|C
z`OKOsWANE0!dHTk0d>Yy!Jz~RgMI-HVNoMfXRHZvf^~5K9gMl>XOiA12Ewj!Txp{}
z{c2QkD+vo~Y}bOx2gX&OgdPgno;7uu*3}uiBSG`bIp|e^L^!LU%FjeWDxt%=pkGd+
zgC8m}G<FOxM^&+ZS)-&BQI`|lR~~N@VVPb<Jok;k_@_5k^S&f1j&71RmkZc5Bhkc0
z*wHI*webx*48(s9P$6<ih7Z750Kdwvn`!E>mbKHNTcy+065jF}9AoQRiJU9VgKkRR
z=^>sjp)q~z@X*v%68VT-Ix|=kXZ}01SS2?3yG3WVOXv81Qa#G%if)ZBR`IEw0o8gd
z#vzCEHHF;GyA4Afns%nRKLcO+<L4|Kl#5$s6+>*RD3#&mj|nDjK8X{?z(XP0vgEkF
zSPKz6-0j7$!%9y`6Q6_Q^qxA*<!!PqRNi5Vc0YfkB|$I<RMJPAoo1fv@L=%0kteRv
z3{!QtB%Wn|Mf^Zz*YMzN&Ngy7)R%<w^sp!0OUxF-p1@B`VCQ%_lLE0564tW$R_{Ls
zr>kcS!x=Og71k#+6p_czk?ouY#CT;!KPDL73alika7IIEDh{Qq_i*;h*n*?I#j|KE
z5@>ac!mXwcAZ5YDoH#6F;jwJG%EbRxYaySV6)<dnPbl-PCDOpVpOW+mJ6W1wvG<ha
zjgMZwGzi208>!hB6|s!4+sS!B+7p<9=WECtDh!Bswz+IDXMg{?d~;=2eM1;;ffbyK
zL^@g{ZB8^Y;Z%AbN`@6XMloAVLxPEsbmW5bY93|39j}l;*z4Hw1aYq@QRr?5=X2`s
z*72WzziR5O|9s&AcfQcKk4~|PRq8kZM`xHSZ?H0+$A$dnBihkSa3&U5Wr_JI@&`Dl
z*a+;!9^PqC#Lg?rje#SEi$Um8Uj?)jc_<YR(tt{b_k7XZ;9?q_KPRl;&;|$WGI%oA
zrm99y`lhez@nw;MkjL)p+H$Qv=c(}!GH>R8*9^Fbr)z4>F}Gca)gUPrg-u@~qu8x3
zM0RP6QoET=fj5GZk6xqE9B@$2YVQzFjD(n!`s<#i;7ZnC^pjsv!=y-^U0~a|J966O
zn|f}RA6;U;iY{JDnMtMBm@7O=zxflZ;+^<{v3XA_GN3(n+s#t48R^&b$;grnY*Ryj
zHNmu`qp*v%D;Y%ir1=>%S6{e6C;}fp3h&lC&%`p`<HH}Y_W&{TOZ4wrR^n?orFE@S
z3CB6N^yBl*$(2kHtB2NwS<y#+0hwVvT2!$e-&Pba4Jistya$YPvB*7+B;Ui!G~Wdj
z*mOmv%4NNY>$!7Qwwp}7c$Sp{b#=vmtqU}0^#YY~!=BDq;zmDpS0d(W;T9Db%0*W-
z_7KV5#3Lr{Q!k-FxgKbBgYouGM>Iio!YKFhenox1?Y;JD^^$FyBsQ<+pRk-Y*a1vQ
zog6`*(%~wk{a)lvCw#EzLE;t(F>}LqiIpRYZa1~Q_Z=O<l-BJWf_1QsA>bB&d+44M
zs-uZmC@y4F9$<(zd)c^>qSS%Q$Q#a-_WoKQ08i)Wz2Tn6Ykb}J;b?W4!_+K(UV8@j
zE>WFXBP|>Gu*H(R)cmX#c8~`i{zJ|Mi~NvL##8NA*N-Ho1cRo713&g1Pmj^$Jd>XC
zkUUDyUt@yA&y&P(em~RU%=Mdpi^P0MBL3EV@}{LYdJa!j`Py8`#U23`BjMlz-SumD
zOW8gMVc#H?QB#v>5xhE1>1XoCcLzUu1TzJs{A40U7;8C6-KGO$&1Y^Pd&7{ou5@Gb
zLvE{}|FR1^yG>|AYDMqBr}&edVQ>2Fr<>WOV95R42s%DnN2wU|(Iz&3kAatwlL#w=
z)3LzF4%@UzT3?D23!Lf{p$Ie$i5?jIzD45%UTgy=v0*7;{r<A~)%Qj{C1=!RF8%I0
zt(nwiX4dbk7bR)^SoOcN%(BX0Mn%P>x!rIg7xWerOe3diBh;~Z0WCfsv`b0|#T)7P
zGxCsZresr%f!S>JkP4xH0)vf#3X(0Kd|<zh?Go24c9&YNJ+4w|S0pEW|8ZeoqRnF*
zvsFzd)^C1grfW><tdre<&$U68xtqpxu>|<CSTHpkFA~!T8`=01FGg-wdRqJjGFM2M
zlz$Fo7$Y)*Lg@Yrl2lMjWKw4OX6uhuM?2{YImBCqO=0Fm7j(&gEb0xkR%%JH&M6m+
z(tydKuadIfJ9;h^eEq8c=u}vgXluvs;!%?Q4dSjTF49tVCJY-py@aLSc<1ug!X|#=
zhRCH3ni2%m^s4MwI_sdHUn@zvLfzUq*Gvewkqd)Fddg+`r9NG1Z?mdr@!QFzersjU
zz==lmkq^$@0s6y#1Um3e^)8|iC+#Ly+-j$jQj)G~U^m3yGV4d%y_sr%Po!z;`&>Jv
zOm`jtzbb|uydLqJXyEwSqlt(hO`bT2TvGXT9#ILQ`!V+)jv_6xUq)h?Q@3Vce;K1?
zMN)btC3gjT`(C;v8nOB)!K+X8IxqM0QRVZV?%_H<R&SGkg)EXHLeM!$oGXsFNas2i
zw%E(Cd=y6b`P?sy?bVv9k7gM)=%Z%Iq*Y!UHB?<3J^JYc2#)(wIzM_ejJQxToaN67
zYy58_b5T4J^@P|{HJ);eFpB#<^&-s=my{E8(c>Uhq4rTah}^jOLZ+>W>VGj>c#HW+
zMI#2Ya!Gi96mTe(UzS-CraJS8CVaqxBHHtqOobFAX|TZ!|9=8z2ATQMW;-r1i{_?%
zQJ>Y<qa?Qd1OS5W?j)8>^^Lt@!>VZ*899|QA&B|A{B=5%E8OYOR?R87vEvv}Qt3CW
z!m*XuYo=0tGgPQIQNecpHX=0tWMDFT5ja+37_v5hg;j=H-MV)!>Zf8#2frQuV{`zk
zXX(r_&da#X1Udva+9R4NR2c=jhQ%YJ@1d~r=Zxl6Y7}%)y$`@ypN6+<Tvcn}S%wMm
zX%@|zmenb+CVz2?8YFM$Ly3{nL5=AUMO@B$xsSYGu9pxGyD_jFEtX7(i`G!hwd`(t
zK|Z^GG~19sqOVTmi=*l*8FL1mPxeVAAzZ}YJ8LyNMu#qkQ;ky7gj)uR)y7$!+piBt
zG>V=3(onGJ;U_mjptBL|IXvf2urjt}_E+(GoaMhw4?l|F!kh@^<NJ*xgW^|fogys4
ze}~}>5DK^f%}kDg<xM#|PmU_-tFrAQlf9#V@@RY%giGdD%z!hQ!Z%T!TWsYBXYJMT
zVJGe?<Aa+WPa?us>Ov@2g?fpVZFQYofOT+HE|@BhiU35;ys6IfPSNHefS`LT<okqe
zgK!qiL^E=M<sN%6l(NH6I1;83<IGf6B4plkNqRd-BA!E4yOktENk^Qnp;b3JMXwHj
zdJ7Y-3-AB=3g}^vxE1;dJdt#>1(ofBu836S5M}l!eE@jScYKw)biF>x9JM=ryVD;>
z?(lk@0Q`I>k!HWRGtQJ~dIJ7SOY<#qC4hbG1f+3uQ(jeWt%Xq6aXh0I_1(+L%UA-M
zzOYyZ8YavmBo+OudLmJMT$C7;dKd_Q$X>x(ZtlYj)tWCd`E}-+Fitt(IBG>tQ1zmD
zR)U1Vv9B1B0q`W4fR0IKvAlGILO&AsofY^Jhu+GkN=WZQ3yNqrc=anv_+H4K25zN>
zD&&zvrDBs?=K=$JMS;RK!$hk=avR*zODdm45IE|gw;UL(zXqisK+208OgLSCd}IjO
zTAB=BUw<TZkI9J=L;jj*2RV<oDn{+mV-Ciiv}-R-wI2@`=tQ^rbz-4NSzD6=X^_cT
zgH^~Rt2oSR25J{m?hmTVNq>+!YQ8GBbwkS!Ryzh>1cg~fbs!t~bir#ese$(-h0wxt
zkz<HyO81{vQl9PePDL$=EhQ&^x%s}vb$2ZHt^UTzM1U<MPqE#{i>+a^C&X3aL-&>+
zmxpArR#qJBNlY#*^C0;O&_J6aFjJZm<?${n;+}<8$lOy7ifD(u6xGg(@6sg(y(8B6
zh0P}$)zoARC?fQ@YdwJxegHUu?6L;~{W&qtN;xsQX0ted0MOqyk+6(^Oq9Ee%&WxC
zQatl|JFeJyS~t?9n=5~fj#FNBQgjO`YT5a2yjbecL1I~>^8=t^GKD%LioFym&IooV
zGF2F-y<7k%$T^Y)YA7PMSIja^DFACgl)v$5zybVjs9_ROpo~mRtI*;jDE|P1Omf?m
zEc5kB&fqsGo`$`|D+bdblI*hxe~1t)A^)VZCfe<`l5tBoGoo31sgGd#|JLAA42!^W
z+9z}P$Gc%JvZs`I<k;5oPMxZ391WPUbEX7K@`|0<<U2Z%I@*Kb)sX^sNjbtO50AJ7
z4z=JnL70KC`hfA`#@vD>2s^gKeD(=7@E~MCnEO2%bL<Ss)7NOss(HE&f3Y$UGk1tX
z9FF~(NQ)IMBn`48QbZF&i%^mNPPFDwN93%@dk@vqJFGrR1>BjmoEU=DglMxlFd+uo
zI&3o2>C`mYg;Zzcrfwl-=xZ#qzGmyz2bv2Alw7gU+bee<=VlbI^?W;3gywW~-2HX!
zUX;5Y(@RoDQiql=Ay;2Te@_B|QWQjbMVa`=V8qVi_p;<$JU$bZ0@7ac_3j#5)4?51
z4Q8#uF>%ZUFw61b^*HLUb+HL4AeeY+#Y3@P*xZ@cSBD|9y9**-VC1%dvAh3-m*xym
zowlT)|E%LH$r6G6k>!<4+?qg&VjiB#k=*gAEr(EtNJ3dHW<W#&e;)&4G|+SlE^Znb
zM;XQ|bUkKiqP3LA-Y&(CILncwS0B`Xb2rF7VTnYl<PYV)q-Wc>GIy~c`y$TqXiyQS
z23|)ud2O$OOd<#S_ceMys?K;m-*tozVb|fikwmQQZ@T>;@;<b1H3G49^(QiCiP^YP
zlAIw?&J12EvOhT08tmO_#Pj)ZlPMHN_P}h5v?lTl=?`v@c?2q#5g7;*0W_C!QwJ8e
z5@H2Z9G5Y;1{Sw`egzLY0yQ(YjkpDHAD1>41r@gv*aaFam-}G_7PmqK2H_o-Wqt)0
zmpUQ_4g)kfFqaTq1{4J`G&V9gmyrhtD}S8Z(k_fU!QBRe26uONcbCB#U~q>3!QCMc
z+$C6Wf(8ig?(P;e$YslS_CDYLSKV7vHS<30-RtR9?@*Des56RMIGO>a96_#(tV}F?
z0C7b%Ru%vY3p*1F3mXy@m4=P0J@9Wa5|tJZ>|)~x;`^6@I2dT^`X-Ywb$zo_bbkZ^
z<lO85tn2_*E<RRnJ{A@L8w(5X{}?)g`2Z58?lu+xMJ9lpBM9h%L?!O%<OQ~|vUYt-
z^FNOOT5~!8D=#lM!=LT|Q3oK{#@rMHP&9S520FZDG&i*es5_e50A0QQD+H~8wX3TW
zA2YLuhX<3XgA0=**h+|w0pMZdYJUw-1G)gg?m!E`9}NSPOdWuKb;g851<<gzarsNG
z?r7=iVG0HU-URkG=0K3kn}-|70tg1YB?qX>Dgcz7fS|vO75*|{0Q@~004o#gf5QFk
z{YM}h(4Wqx=H`wLPNpC)8;}*i(#9SLP?l0)a`kj&0GNU-{xCGPcX534H-B|EwXruf
zdo%d6a#Mhms4Bqpt--(Bb1?_oIJvqox!BnM(IWF7Vcw=J39=A(bZ`KITwRd<s87NM
z3^acmyBG6cCu;|C^Z@z%4J>Uy7M6cBVd3V)tO>Glb_2>v{B80kLi$H$1#|^)vaqml
z^Rfeg&H$jNxi#}2;WfORfPa4~S^p5fb>QdY<md#jd}{*eXJZL``$6(?F?9z5T)}QY
zKc9aq{u?2&vH~n@%v}LyKr0&%(m&DP#6Zix@Y~^oZ9D<`EN|Rn1+e__`OlN#8@((X
zLH1t%nE!dc%vy?aYI@4_e|7vHotT)TC%}i1g9X6I#>oO;{exuOZ+{Pd{~bof)aLIp
z{^ctRvUCLS{#EVUnEt0?_rD21`}cCt0scFdlH(g|fdJZnNN&Kw$zuNY#rpqD^*>$y
z{}}#9l>Zya|93%BZua(ns%ih!|9{k`4mS2)e;d4!*3I>80TdnI)&TUsO|^l4Ev+KZ
z!p6<ve`{r3P2Uzl6n|u8|DQ41xJcP}0xeW*T+OZjV#{B0%|Eux-Ub9zadffy<5~eQ
zz8%f~(!H&gx!v2<;qpewKUKiD&H3+?k|1+Oi$7M3jgt#t3I>~cA-x^)8{!1`u)gi0
z1<><P5(Ai-K#s0&E`Yc8`~a4YV5C1zl#2(zEb)ivFT@LAmVf*Y;${Od%lw150nCd3
zK|Cw~W|e;sCxBV=KZy6u()1tnmci^F^p?T=KZx^9XYOeK*8KlSI5_^0I5_;{|Ho9A
zE&c^r0nEUEL3RMM)xY4IzxBW1TVXc;f^Yfk{{`QQarg(mZ4EQ%U+^uF<G<kBG@Sma
zlj{w=(bw@`m4C6m<p%!?zU6lL$Kh=#F7~D_|0<64Er#pA;G2cpzu;S~?*Blxx8gki
z1>cJE`V;;a9n9Up;I|v`&xL*C%KzX$uU{b06KIaKIOk~2A7Wb<(tiI{l*of|cT8}a
zYF9g*j?rfk+~M~44nBph=JUsO@U3XdP!Iabwj}M1$bS;qv(Ir$1AJI(l4{#;-zVb)
zwaMK!q=gCGALE6`qV@eGh{TK<B745i&c0efcF--5-{q)so!xlespMnbd-O?r*7uh$
zk9?ZjRo$!Ml1F|jpJI&Bj5Y`=nW4%v%bvxjfOBOeLHK}q;fXeLb32XsW&D*)E`c7&
z?<9u3$bUy~gDv)9)_b*5gUzLjfRaEDp9J~_b7Yd*N9-U~4sYJ4v}onWf@h0B5!n}1
zhO`w%bfyEgtQs4&C{R;xt^3GgC+kjb3Lu1r=m*wLQf*liwlSDO8K<VrI10<Pmfa3W
za?oihw%UHH)op2;M!dv3RsRZ@LDjWB4s7oB`G4KuygKyr2jktRwATBfrk~s0e%d9%
z{ldgn9;_cEt@R^Umie+s(qPM7N1?j3fwI`>3`_1QJWsHup#nZ1FU1}SS+@#VSoC;c
zdm%@N&p2~^FJt|x8ZVSOH)b&A?L9CJgI|NWlwQrgr(IkW5?HfKntc95o0lco9g(*d
zM}MQrUk7ZQ2`*>8*r>u~k|^^cQ%Kz6T(}&dEaH%r?%kcQ9*Q*h=w<uBTflUtRi)V$
z&sm^KzH<`AIC{_+G$<Av+Q)lqaj<n;Q6Xl8bLgn`a8czEE;baqslgqR7tyH}_@&E5
zIYU}B3aa>Sd$t$R?ER^Rm^-;cQyiMMhJWDW_Gnk}lLte&tOONJ(V}9%{GmXY@1o&?
zL{Trn*FX{6h?}NBwz>}Gds+yT&*-bMRG;^jq2C*K3^i$8z7UELscYQlzxUpN=$)S$
z-6kKN+@|!guQDj%=xYY>b4F)C+b2+y&u<52Uat$%Mn}F_MmDc$dWbeG5!YtT{C@zh
zpA3$`h`opJU}wP=8at5z6qeC^UQ;Nd$9|nl%CX(@&K7wji)$9Fb}^`1!@dy+tq?Tm
z(uJg4mk<uuJVKn<d*Vv@GUX|qrwSUVAj*s+*s-ahnoG_hpPxihV1vM`)z8Vq__bXW
zKt~>J8LIOU!TbOt3ISqUV3r1ZuzxNt^->K>g<Ud_JcN<27J~+3<7Gnhog7*X6YOGN
zow(c>gq&+R1a+?uH(WedEz<iSt%=@=uH)G`nOUNw>nP^3pYNpRtvo2CIFsW%F4e;+
zeQREW990Fz_Q%qYkta?3zX@cMva*!<B4oS@v>iH}H#9vgd|n%}zF5l*S%0rO3`A&b
z{Q5OZc!<D6_?kQQ7#G=iBo|gcrSqD{Q$>J-b(Z8>&~^M?24V>fm7go4rP&WG@5Xw_
z{yS4s90JxR6SqOewW@A%+V9p2pLFiU-<*`nVBPsV(4e(-k3*Tmom@Z5_)fMp|Lb{u
z%@fX#J@Y6+mUWVfD@C~n#eXllER`d1?mk@W_{nny_q_lp@E7NyObT`a!8_>WY>g>h
zd8QS4{G%bdZU#K*pYlGY;o^E?(m5SSq$FO;wC^J2X+HI`DYHy#rB!pus!&qe|2}-I
zY-UWt+tBaAKb`@O4n^G$5F!xBkIYlEQCTIM)p&>9$&C<XFt|DTD}Pq9_2GN3xg}=g
zhRtw&$ZY#ijN0FWS~2;``+cl*UAV76J{i}X4u5$3Zqwkx>Wga?wQ$*X?0Da=-yk@R
zV~>ucy>`l63nc^b=B&c8CFR?>agiczHczqx9%pg|i>c~uBFAI@9ab&o=s__-k=kF6
zr{pY&-MV>>-Vc<g4}TMP1A$b1Ro^Fva6C5GHrWgv%X`-W)yB`lvev_BZU=c{4rfwu
zczlhl?+_MJEfC5Ey}0f^b}g*|NTg6pL+}+qoa$E|)Mbpv(_d?Hq`mJ;d5`b~2Wb*_
zXh)vgWuL5-9vkL-c9byz@6uVJ72M5d9dKZ68jHKA2PqU!j(;5bM4Gf`KULuuRynMg
zQL!wh!+v-)$vlSd%S&2MVz#{L&9?8LQjuiK<0I7)rBfGEVBQlb$35f3Di_ZYEm*(H
zw&$<*8SX-+ec!sE0}GOZOQb-Ef=Gu5;qg&ha7@JL?$f}d-7(g_fHD((36S@7R75t$
znww73##nzLPJie?_1aN9c;yPL$gMT%xHEOlAwv*RW<Abh&W4b>BuoCr?nU}OW7`}}
zQE`4kx?k^kRUg`+^0n1_JX$IM4iVG7{Go!PM^3032(zFi7N;#n?6<E)7Z0<dKDcZ{
zz>s$yv{rJihKbvZTh!4e6*B-`-l3AsbkLyyGa2Z44}VjPZ>&G2qZN*UH*oz+HMa&;
zL*@0xJvjFP3HXyC(dlK0Fc=(Qo%Gn8<c5109};asZ|_5ha0gVqn~OIVvBcwICzM%N
zF?FBxCeR$kG5-2x6#Iu>1uN-?@|}Sg%$hhw`T;Yiz)$=>tK}nzK~<{kxCdHV{0n*Q
zqv6x<P=7J2m6?=tHf$K$d3Y-3NP|y$KZRO-5i}6o5TTITc$AmA*T^&bou1IC6%=bv
zYj0KEoxNl>RTZ;J-h5C#n>sseu=)I8%Av}N_jSY{P{F(9tX0%pGExLcbTP6!&3Cbg
z%>upiPUuE}uPWFt<eMW_$nhpHjK#Pp2O~!o<$sq4yDLd_Mkz{j^HRzx*t}3Di7r?X
zZG!B%ydcBm5aXnj8?OZ95qFPOMhlf=MMd#{x_?RvSevrXr`~V8*WI0OpjjM)&|aF@
z4W{nOX9<!Ayb$1IGTR{9EKa#0n&lAoh$Ha~^m&WQ6wQ8NheYyXy3)_}Y}s0MWfgU+
zDu2&LaS|P8(&s2pUtlxyNTxRjxRw>3DFF%bt%)T@>Q<5aDKf}O{dX$(>A9T@3Ozgt
zFm<=2k&IS3C96kQ4w7&%iN6@Hw4_Qm-fdnS2GWv>qS!ZKO`frutnj!oUAxhrZ7)(h
zUJLPGFa&rHHB3?=IkwfT=Xyd9vZT9w#DAC@B*4^WmGT@vn>k2Z2Uxv0B*Zj&YI)r>
zE@4b&Qy9MY_Qy{vadN-vM$IPql#p?|&|k^~7UPYe!~r$yidTuVKK8kmTZp{BE9HTh
zuu}?~iIV>cnLEv2&~-GEWFI8;NDGI&5Uq3_Sq#a)u~}sCQrYwS6f&_2{b~I}dw-<B
zkK(+P?S8Bfta50Vu3}z+M(H?IaY>U<t;2lljgFx{Dje0MF{vU)m5#oxNsQnC%JTRQ
zib;0%_B$2cdEOJf7#H3!NE(yvDpJ%F@RIfepHgzs$)L;h%|LP!nFyR^Fv@2m67AJc
zYBmlJ%rbj#<UPt6Htln)I*--RAAcvxyk@TjyA;aBU$B0TYWIGS!ZciEE;jKhk5;Wk
z3?E81P)+F}lKN1-Vxxh38FOHy1h};@E>JWHF8riUrBSccD$gBPgFvsAF?XU5F3Nfu
z)V=A)I@NUs`nA_C5Tg`D-7V;bsXYEd8k2(5l*~W5eu=)AmJjZlMU#o9sefw}g49zL
z8AFp@+;F8T*m+I;+@)Vi$K6NW`IKut2FgaMB{<ZjnOQdN`KHl(NuvV}m|4$Y1{s4=
zpyd?uaFG#cQ$>*K_H&0PGLfnpl}LQ#r5e|lFO2N6%+`89NB+2{CJHqLh6u~1g_?Pg
zE&V=E;h64Q=L+j9t$X<#Nq@gs(<uY6zz(9XXD%t&JwL9mKzi%YA2Qkm%kF-<+*~+j
zVVjBgBzVmsBan-WlSTR9nIkZ12k3n}IVas?!2=<0m^p=7xu~RDP%=&$+}4cEWKk;!
zIDLn~Y(}%kLh)_!&uZW~cw`#tOZ`S~IX2+!Rq7s0wrN}^=3V_W?0?IrUu?qeGwld)
zP)FD8H}he=I*_IX<U9nkg$_um%VGFd9*gAgNC@}byHbd{B3|l-{+FR&pfai+>>b7C
z4l9!o$x^f3$ypu_nj)o=KccLzKc1CIuAAzIf3&<;@#F}U3C+dvJ%ZPMYT#d%CY{4Z
z*Tte>7yHGzc>T0@zkexXKrvmR4|mU+I+k&2&Z%_ITka&p#x2(-_I%yw!VG;D`OS*<
zSD7b{D#JThA^qwtp3x*}m;UA;J~A~*UPCTy>hMPguWCA*MuJH9p+noATkX;)g+4Vg
zBN8<+Q~U47p7qAL^Sar5b)1SNH8AUYXUo{<k?>ftZB=NEM1KLLznN_dLpS;t^T_?y
zOp?!zC8)KD0|(1@=2UfBDB)z*OZL%5TxAQZ+7SZ=XPFjti7!PKKsXiPRCMECC!m#}
z9t6}*<)>{|Q6*SL1yoRKRGKiX@fxm?ap<xYc4DJ{4(^XQ{umR|uM?-G@?v3Y7hJ%;
zZ`~_oH3`i~LVw8z_TN>V5<U@D<OfXCb$$Al{m@u+#T>&AuTDUNV$^cOx8n*US8FSB
zgzZ^TA1q1BBK?zptQBWj)fp<bY#`ZAgQn#o2c17hAl{h>w6NnA<Z7=V&YGmt2jCnf
zOha-`nD29*Dhd+biJI_c6<4S4A}&y2?e!-G4ne82jejDuSIDI*F8vHiv8TP9ZlpvY
zZ3Pag?lL<#DRIhkLux+*BYo+5mlxJdBV|&nHuPGiYdG9uH2uE!_wD!oDuu(=1Wb|-
zr|PDh96>yg&p^r|%;@fxPJbbgr#dh%lB{VyA^n{tD7H8vYe*ER4##>2cz-?Jiede3
zL|^rjf`4_nF5~V;e|7MSmnkPeQAJm}Y<7a5louw3*bt^{5>FQbH!4(wGca*=eUg?C
z>Hc$q5OQ{JaMO^4jBQvx`nC%W80VwsN-EMGQ;*Qn#9U2H-nd;_^^$!bKXq!&dx8C>
z%&%9#csk3;9VzcDlcs?t;rwk`a|jW28R}GlSAW>{#4{M6I|zaYD=1s)BeB7@R&6<Z
z|FvnAD7lwc`!?U@G(C2XIeCHS1#Id^vvI|jxrj|>OF#DY)U4y-;(=`n>9SOes(l(u
zMVCqV88RJh&1b)-+s3A5-M+>cBjmuDc`Dn{0dkqn^C)GRSA?O_STf^>Z&CaQVRd6y
z0e=Zwa?S2}@9KgKSKcF0F2a|iEz^iw3kPv#2!yEPKa;eb4u9>w6iKSW&qmBNt!xNn
z#Lf^6FpvCnXz!>E*`P>!6$b}~XexQXY%)Iq()F=jHevU*hjJ&WHfg1iwa<MH*r+An
zZFtJNnO5ofdE%3t4<}iZel9rXWO`tP9DlT6eXMK;ZIaz|mbLA4|A<n><L0$O3BF2Q
zy-smMvh*`|?xpAK_VOoxUNxM}ek8i@I(etbPzBlPnoGB5!zXi<+o>Gdw@h1Y3{(y}
zB$3kUd~u$D!xgG@JJ{%Bl;#WHyzKK1&I!$u(KDcm(>!%w=z<_co$%f^!rvGci+}N=
z3RKeAYW~J9h+nO9ydm7~fGwAJkIOjC9&$%wvpp{(vJxL;IUf9M*Q~~LoOSc;P7^rb
zz=TONs=nPL(KfPZ@AX86&z*#ukD_}sCthFG?{L^?^sPqPjnh~<3;w=wk+t|Ynk#ek
z4D!oH%*fUS(gsIFEo&7h-1xVIkbj>O9Q~eUxfQQ9CCOuHKcu0^x#EM#g3(lqwEQhU
zX9ZlL$G~wL%#JMB`CPP=-BR@9x-N!|TV-I+>Q3)f$K?6R6s9g1{g6cB=9VumX32sS
zUabgv6_6zAih**>SRT)SO}Eu*T3$HRYP~hit3OlX_KKjskIyTgNDzY2r+*<|W~xKE
z<6$+F^i4Tfe+$YpipY5&xWEfWmNxaZ-4>lwp`={4%e>Lg*P678GQnG73rlmnC*Gq?
zH-^6Bxl)gB?0mm(ys-Z>>-Vch`+!AY7tsKuiQByh6rnFn<p~+i+)DUCL*@^oB`%^f
z<{Yod-zKIPd6>n6sFC_LV}Ft8CR3dwk~sxE9)#VnC#IxCD0$v{HdhL!A$XX6wm-k_
z4~hIPh}`w-5Aa0WJJ<dO3!kwI3Ui4~X90b9=ku<{URbZp`#LJ+CuWpT^FGfhnG>VJ
zB=umyZ|HM*6Cr>}Ln&2{NoxMt>h~d+a^Y6@;kF+}22T+m5^?69ReyhOmMWQ9hSD#$
zcdmuxI8q?SES6{@wtG)rTOC=An(oBLyERDVPG$)7V(C13tu~iePCz|=XoniKx_#vk
z)?qX3iRUPl$sEnkn!n04>Yi{`06NxvPS-rW_D8uZN=%&K7CNE(QFxsp1B%aqYAvl2
z<)%3Fk@%!0z|7|Qtbf=)@aw^4w>a>(9j~m;uSCm&vd@#yTd(8Kzv;0Si*+PVN>b4W
znT|HeyC!-mKFBSrbHD3a(j~1yAU=RSw_Qdhp`Kb4J&;48uh<L!-qMDExr8Zrld6=r
zOUdP4r%tG$ixKgD8NXT~?UiG!<ZNF!3u2t+_T;YjAwUYxLVrtIVUB@*AJ+w^Qu?KE
z!-%=L%(<Z#X6P2uI=UaN^5H>`w9-37qRdqUh5ZhQnVI$FKyg7t$d*n(+uCGDu{~Vq
zbv}RS)qfMXajGRDH|v|$+Ny)*H(4SMjvU1Mp)n{rcz|q){|uF0keICE#}6AxG2JeI
zJN<qvK_j>6oqyEbqrpDm^PSO55;O@RG%|k%Vwdlc8bHFJ?<1;=+YHyMYX9>@I!(4J
z`=A-W=lkYZGWKgS^%1^uTQiK)WvXUZxdeD-p`UrmuV<f0z*=iTJp`VHUt8eXh3K}<
zzjQYm2HIyoKyx~FqU-PHO-S*hGbmA)(qP$?MH<v9seg9MUdhai{+b}`64u~56<Lrv
z$J<wWMTS?`>0V{26{w_Zv)<62gA3~5uxyc%XtIu5W?HD%fTA-}hn-|34F!4lJu@JO
z>nSx$;!KD`{;Ih{)<up{g9SEv_sra0u(wYcAbneKJ$-r6-Q+Zh!;|ti&hqSz35M3p
z@r|O<Nq=k{2n&%GLDQT==aNM>Aj=5ZU%3@};v3R3{tY!GG+WI3L6*f2ce>Gl!e;>o
z;hbG_$@|CUklT*bXG`dj#lp}VmG#n}bdvGmJZK-EG>bkzdeBv0SYbsD<2cXoZCKd8
zFKVe-!3np9cOF14`rPi~h|1U^l87V^EzKRbO@C44t2z+|wO9VQ3(&jrcJU4U=vUCE
zFx#Mx6<Nd=Xjwk%&|cCb%6=X<ccG5V)+b1iFsEN{(X^2CB=^~1r5=~GiuelG{A39~
z@aX}~7jD?K)0TDeN0#V5_$3mjx@)rfg<wR^y@x@b%oh^zsv&edMQbfS{xeegQ<iE-
zgMa?Ynt*1v4j(bh$E9UqbNXAVjatN0rM49@bxz(-_rcxukTK;s6uonOpX;p-G!ENW
zcZ|%+wEGd@mTBugPy(Vo;41GA=x`#VleR;Ru$9QcRlU02MFK^isPVvQ{Gt}J;zTCQ
z;X!^#jJ5zsE^mWj`zy4O=g80U9E2Z6gMZ{13FcR|i_~{jqDHGl%cW_8`*8bK9`WiY
zrbZ;s*%l^bRJgnGYDiCYceIBX>!gM-B^49q4KoMDBDTIzVnP^RvQY`Bw(M>B??GD|
zZIW&uj+}fAC>VIZG~i=V?Bp>SLIPp)Xmdp7h`<X@8rAdxc@O(>sT*Dksu*vd;(rE3
zG-?3aRW+Pci<H##OCs--9SfhemAT59EPiZO5sHxS&!-)T8WC3G?aL!quZTDCQ@O5;
z%!Rfw!|timH*5be(<iJKvkQ-V@1VWRtfx{Ru&&KuCRj-LJJ2SJLPOoU#{PT`S1JIn
z%@}q+{d3z7nAe~9mm#5FS35^Bn}3MUxwSFMD_cya+QutfhmLHOqok!lBI&0#9Gdsc
zh;g#7GVaW`ooFu1nky8Qy#z+_B28;u2hE}{PvU`<$Inj1qpCA)X)v?0!%6<rD?&oo
zSWBQM&FLESZ_@D5#h~Wcqi8{RJ>ol+%2BXPBK+bY$%2`CY=yp~r+_SX@_%<s6(*P`
zH8>^n3D_%WIZ|8KR6i>(#5ONKO&*3KfmRxApK`#S@R&TEX4yc>J16mZH#n>0{H1pk
z!EgVT8zXO{UxpD`g|#KW(#wBCL-*U){n4J`86k3iGHiQSIfaXb=`CBLm}e@jlgTJp
zwU$*|%qJp!1X$ggdOmuV*?&y^hy-V&hdd6t4+KGjj0#UWJRzdsQnpb!JgAIriFJ`a
zI%h5&h)~esKaLu`TvKmbHJ#k7Zz+R}`V?wTc^$0_1*pGceT3R$3<-qkSAuJL_<5I2
zLI}(Gg1g3xxP-w5<kH64NgwJ}a$UVEE-5TQCg#{2b6&xPw$oy0?|(AmRg07=;675{
z2*Yhc?pERV=inP9N=`A*hi#+MK>XeH<EmYykvIUUP`*2a(Y>$2TAIVEPaKv7>6u5T
zXsRM85i(bw_p%^m+b+c@ddnwT9u%bvYi)Je^%Z;To4!%JH597oQORw0*|(4N5=|?~
z<)>Az4@f_d-wQ1*Wq*hmR>bW4%}&3L(i>W`h_iU52kFkdAkFFC85UH2puaEMYo|n4
z<4RMApt6{C7ZfzKiYuu!RNM#q7f?{ZA6N`rr@5HN@=rZ4q2v1DwR=N5M4oEsVmeoL
z2PT`8u8643+}B8Pu?VhmmzGN-cIVyG)={*mpHS~Q?gK-?0Dnrkob>%)bGzdD2vwqy
zWog~^bm8azkTQAeu1nBS$zZhTjmv!sAXm9t#2!WzJG~FTSU4?9tp623G=!%0;c{qQ
z?t*vz3N<hL_94=d51Y8kY@*$DdDFNQeYDg_kM5(wo>y2s2mTI2+AAc<xSs3H{3aJZ
zGugD@h_j|*DStO^dcC^+?Gv)4@P-kTj@N~^Pr@@CKvc#gn9<@}v$h~M5`RiY@v~3d
zT@^gW;e6QsOq=9L2zTK^_@i>62d!igK6<pKOzGx3;B|AUTZ)fmiw&o|2@09+a$eqN
z0Wn8b%uQE!6J|jly<*z|Ol=l)yCvtQij_^lES~YU)qiO7Z@{0ra&kv%j}LTsQQno+
zB;1&^CU!cW{_Q{0f)@1<C2%;S%UXX&)O23K=N$T11(?$rq&?pyq>1_@Vh-B>Zhghb
zc~}j>eK4>kr$89pG&fdKbEZ(}&T^3QLHYS}`z}u8uux0At>%N_Pc78O=E0I7?~feH
zRyI1r`+pWE6$f&ilPdkiP<mBDbFBfaF{)!$ur7^c7M6*WUu~+Uko-D}LpLcQoE#df
z4j+gShS0uxz7lW=+UtbV2{kuW`e!6*5DKoJEHn{yg%<!;>fq}J88G`VLn?BXJ^Je^
zFJXFJkFF)s9dYShX|EG?*p;#88*0}r@;@6)JAb7n?=H>7jgZvqo8ENAde7BtWVW-6
zH|MY6#p(<76=dtQ;DaBCkDq=~HB%IafrUXc*u(XqlPOi<4ELw<OfmWu(*)V$rC7gT
z+wiP1TccV08<j}n`z|vPGg?Fpl<oK%ay^*vqnlV9wjT{CFCZ#D2$5WL4<(kNP{PWz
z;D47Jne7&+46`9K?JdODB*d3`){5N3zvy>q>HaDh9~=s3?_z5c_!djlw$MX`QUJjE
zg$pV->y~KY?&KP_nA1Z?)}z>UH9j~OH_JFaC#SrmbY;NSG76oy3=?oH%9(kIqCw^4
z6Y3u8_O$pt*h25I;9eg9XP_U|vFyb7Gk@m`wuu>Pp?UO>_|~gCwAKTvjJRW=UD4;7
z2<e=zPP0!#Heyvk`UL3q?ZGE|jMrwJLo1<`OL56qttkSHSGI~T3d8l!Tc0mg&<@K~
zIA*S8OJ2(I!c!%Sk%WoLX(sg!HjFU0F%cdH4!<rKPYl_SPN5*x9}+PVH)Ckj(0>!r
z-QQy+CkWu_@|`e$D$gkPv$QgIbfA8q89rc0r0y_)ukTTE>lP~u6poGk07C_ahr)v{
zP<f?^jJLAZJH7@_v)~afc`;!?Yk&O#&o&g&o4#!Q+(B913h~>j`^#1Ts_BU=zLbR1
zw%RCb!`+!mSRzG8K&Y;X!Y=;>yMK$kzqJ7K@9&&19d7>4>AHY5Y@3sy=Z1$xMB}X_
z*U5Eh#zLS|=XfwnK^e;d4|w5wB@tqR+0&EGTDQr%c*Mf1=ka2I$Ys4Wy*OiU3(5z5
zJ<>Q)Nl=sa>o$HE|EZZ0Kjf#}9d`y4l=fkz2-)&-)(TT`$aRlO9IDZ)t$*|J5BV1j
zmW+}pchCW5uX+6U&0d96W7FfS{m0<Uy34B>hLINX`x=A@)t;4eXF$Msw{%nL$Dl28
z_Obyen2vXTr*yCPblsKolc4f0i*iDJ0r1zUF=$;4h=o+K=dikQgY$0;PpSSNAmCnH
zE<iJ~_|hlrRtK&^{e#k7N`Gz%ntGAiMjA{^qDl!^bb^Q-ghEtwmV$EYPKpa%ab~S}
zJe-*dWsaI`m|Zqj8nF-+HykrO@8sJN9Iy;zglygeL5XsNd<V6ISL#aIdr0!5O%5vN
zrnxq_z_gGKRML1o!MqN038w-0nrI=`oCtZ1#si=kS^t``i**>aC4WBFnNA}TM(^zx
zB?BgTg%aM>D;CBb^|c3Sal3a`%`aXEs<pfJ%w+*N-bvc2j3bS5?8C|}g>xOXDJ5yP
zk>S%!-B>6};w2tYtx<hrv>STevA+x1l&(-Jym%JLr`Z>@vd0$AvS(eje#X=@X2c`o
zL`H?Pt@&6$VOkJmVSoK1)lT5gc@@UI5ch6v%|wOH&d4bKQFaj7Jnag8s(UzgFm7G6
zi^yz4yhnZHsV3AzJP|R-N^`2X@p?)is({<dgF&ld+<dluHeU2E|Aw7noq<CMV_xqQ
zQuW|0R3YBOZgc+OoPas5&Y{C1=`bY2>XIpm$?-fW5YIoR%YT20%gGOenq6EK2BM;v
z^i$(Th2->-Ui34TVlkgYG{S8{ylb7M7(yF<Dqyj{0=;v?I9)bn9Zpz(agoLxKs5Cj
zg6Lp~vzYzZqo7fiPbw)EnbtBqn0K_{lu>I3XFlsoTyF3j52LMDc3F`?K$Obp_zLN;
zshX4h{?rPU5r4AN2xn}ku8Xg{H(R7x{FXhkcL>3mYZg&Pi|l59kgsPlKN}&=5iZoH
z)s1<oWUiehO9-Kj1~)fPlYG4Wx}&(U!;1f41-^iBc2fa29&-UhOg5f!ral!cA=493
zPUHO*J}5)>!_E3EZo{de-)|kPEFQbnj|P67gjLGJi+{_L6VkfLN_LUX0V$8vl(Ui4
z);4q5YpWJI$E-n%YwPJj(cKt{gfKcUnmMUV=oG}#j1Oy<PbtzL520Gw8Ba#fCsu;A
z3nJR;Wh|Nt6!g{eA0X*>spMxWV<Td}Rvc=s-%MKF&j3s1MFli&p8C79KW*uw{$hv`
zKc~nfmw$k;?WM1wM>zIor0f^CUylNyFPcs`=C(ilv><9jNkv*v5L9)?24E4`cJ`hZ
z+<P4W-5XUIsJza}@A%p=Nvv+ys(FNS1I5fx8w8KOeim)?SM)n;P+e#@^h8gJ3aTj@
z%7QYVr07wsDn=_ElHwgHV}_Suk{;eh*Isy~&wp6-f(^PbEeV{DOf>IDzFmm$v8F<R
z{)K`=@|!dSyMp1J(yq<rg~w*?W&$14JB`3c_F4~HY^I34qxh!_$~%bEsdwcfeuo?z
zAw(ucZqrcVs=HPZL$MPVqW+QXYSA#U2iV0;Nv8NG4fcx!6sj=P+U%CCAJ+8l4a_%X
zLx1eh#l9=6rM2Xtk0pw=h~S&g9F^|n6PU)SeOWId4I<1Z>~-Y}mTpPu1`LKCxnz|$
z?7utPgxu{Rz`Y#Xi;OFtKel4EYh22b`$Y>BLnoc<M^+7kI`Mn@@L(j#N`J_Y$dzVG
zM_*9XqvJdozca-_r1>q&{@}Eq?&VdaIe*LG`%AOV>0BASoV9=vZDAo^E~gmR4ezr^
zMYa|#hRsvJoX|4yY_Tk(;GLI<soOlTQKvLAZ4(xX`sInaze+`Z`>TxJ%1a)yxs*Uq
zdAf4cCaTH}`^a#sfde4HQCr`ny`A~@=Psxxo$Qg%qlzCS`Gp-3Fffmz+;(7kJb&B3
zDb74Hlv+y*sr3OG=U(+c-sLP!h;2RDd4~qD7#@fuMPy-7u^K5_QMOO5jdkPs-@8TS
zYlYqs+cFzg3#`{O2#`4iE%o(x@z@JPpB;Hse>)iplKV!&HE!sa>n*l|S_RYr8#)eC
z{OIEH^)de<-NG@AxhUsp#|!5d>VM3-yKv|8s686^!v;;`5{^YWGbx9nRH-nI*MTcV
zJ*MP~9)Y?>nL>s~!b9y5M3rJWvWI;lP#sT~EmVv$A4Dk~s(er>Vs>KP<ESa#M-Y3b
zZ-0MQ>fK2Y$~wg`{j=~HoHyr=-U}vC9Cfa5AnY9Luzqk1j!s>us9{$S;(zoFa4$K%
z;V@6iQ&XuHgA;Ad`W9ltP5tW|2rktZqCwL`{IWrCf1z(twt_Z3^MpzY#|Ob*@u!>t
zwC9deBD&CbB6HUZc(h$V+|jZnF2#|~^u(;$wuPNvabdqLQ_zw{4$0)~z>8U!HHdrG
zNo^*^1JUZ$_j-KHyyoZhCVzW5r{D3jbumX2rdXa3A-u58T!%A`9)9r))F6@nOn|L5
zx2%huksIXgXyZ^c`1xb*Og%zGr~yx|6+1uL&a%2`pgqi?Stz6)iohYeImmZw`CCG$
zH)_#C<noWOrWKw`)*m>2y?Ie}p1ew%lPY(?+svVwV{6%>Vv-u?3x9#h*$*W;6g}Be
z?v+~=V*}+G;dAw6x~E_6M6K|emY7<IJW;?a>XeFwRTwXCrMfGHVMzqWP4j=UxDPuv
z_(7n=AWg>DbMubsnKn9=9^bGwxqu)qg0&T>M>hkdStd^Z@ud!(-utl#rB2%m^ZN{G
z7+gj1R&}B+K17Gl&3_V1D4I~T*x6>(!Fl7S&h=PBw>Eu)C8#f{o#lMHD9{qs$VrN*
ze0rCQVgozsiDewdR0vQKhYDYKPYVUS7;VHEkS>uGKsy6ULU-AOVMRU!dWBfZ!gwYv
z#?D6r@#l#RO+*pJ2RssW1Nk&;4C~D*IyMjx_t*?F<NHwTC4c(wZS}_GW0QlOST_~v
z$qOlE%d|fLzF6V^Ts&RcdrfsgS>YhXsZLVq=3LG{g@CC@n=Oj7vAvD+F6cPRCn(Xp
zjeE^dH|gBy=S{NdN0A&2rP$aGhJ3MxW&@|ooA^4^R(z7!Joe)Rv$)MrZ_y6sufvDC
z+#tC-3t-XIe1A42=c8Fs@x(B!_$}ycOuN2?ClEuZup1pzajG^hdz>d$b<lHV|DIXF
zrIi}EmaX;}Z>A-azad)!E>pj~M69kAZQkTCvjC)~(VnmPJuF>XuZ1chES9Aqv*wRs
zao$^GoK*pRy<7T(<z$I{$psAptG;w8Ymm2EP3El6Eq{T6zg*Sn1{N0BUNty>7S*E1
z=|&OZN0@IEJx6s~4@6U%hV17U7?{>9A~VAx9l0FU#WJ;WExf>F)kACoVI@ZVY6s$x
zq|TGdZYcd?d!OpR?`R6)+$a|A%V{XfWwvH0Pv76O$>&Xnq&!&*I;>L2?qP9FI$-rn
zIm_ZQ)_*!Xb<6YUuEKf$$`}fVPq>k9XX^GCd%)~AHfT}5Dxp7uu6Wb8(f|ut@oNv3
zfeN!ytLI2dF-rXTxfnfEV)xw!XT*VZjJADO<~q5Yzhm1c-LSFs_h4&6%`qxgb)8eb
z+pyb)Uh6Xp1KDKP=!j5X9?D8CM?;L{UIm`Uhkt@kwMneyFi>fOh`b?4AKYEAz5-9?
zE%@m-d;MsXQ;VwnihW}nOGe7n^^IFTV(5aARP6+bhBwBr)T~dw-W0(AdN~x6F1Sr)
zNa`mn2Ud7bzfIL4@@zVhe(s%Cgv`KI5M5bcDr2A@<%%p9EuO2z;nOZ7{XBKs)~X31
z9)I@LLIou>Aa{T_ha5h=CrV9_ZqRdTjDOzueS#ANhpn5d6vCnI$~NR2AtWvv`PY@D
zyTq|E8u*BBQ95Do8)T{g>$?r2tzM42m^Izz(JM8zYF*Nl+Z4yc*svaO=#G9t=MRl0
z3nEsw@M6Up!e}yW0!;h8f`L^id$#D7w||zNa2{24>s0C#!lD|^us#ty^pWO20EO4>
zQEx9f&<`J5U_1P**2K%(G!UqcRm@3V-~VKzZge$!u7U}Ud6yf-QhxP`O9X|E1L0Xs
z<mrHJ==i#wEjdNm;YFyzS4FhlM1O4Y?1ssb6evtOe?_A$O;hR{9BfEQ<(zj+hJX0I
z<j#D`)ZQ^Tj~iz^s-2x28w^lUE{I1-tuITK+9vML(qY2gyQRkn`|^bcId)%Qh1@NM
zf&vVNndGz0V-;V?qmNwgS<(vs%A4aY7x@IPmzr?!N+h_g&NDl?T4&N$@`qXNBe1~{
zZOc_aMWU`|2sAy33~VUT4v9t<&wuqie@zWx%ru-H!h9ew`z8Gj$487;Ndije6APg<
zL+jTECBgb$?dtcqrW5`MQZd505+9@i<6SEy0cW`=$uHBTb<fq_vHXU#9Fx*`?flu6
zl$1mmdAo<k-#$0=h=NkPDmA<|i{ic-ab!xZDjetdo9eDE!4mM)PnZs%lz&J+>0L^v
z@SNuy^Mw9-WpH*d4T|h72m?`+d5-XTq1t@9BeyES5eu@kz4nKi<>z;08HZ1Cg)={!
zqfr3rbEi}b2dmz&w)dLLzjO<%YGfWEj%s|`Qv>CefEUok&tYoeA8iSA!IcYQ@J@OW
zEe#Isup{!64#D8-C35P?_J59_!n{S(3r^nm1G@b@LB_iCs=rZGYbLjzTATTZ-WTTz
zUC4w}C9an8usxUM^@yid*{S5^@P8<WLRz`T`-O-OLu(0!UFFK({56G;93l_~m%Q~j
zKr`#36eV$@mLkW2TfEO@T`O86@NpW=O?240O71hAI0PfYpoaLgvVS`yE-g6$O=Wmq
z{OM)@A#+d`x;EIa_CzhpIzV>Fn^D;d8FKA1o*T^@bE%V44zqxIXvF%SfZAed$+di|
zf12&I!J6g{v|E5}kA}S~<J37g&_ysKfiH4P5Fa|Hkn}zDV*q)dxWv4>fo$DJStU?U
z+3k-=NE=u3n1VjVv43GEtbn|x#%4AlLTvCr*5l%ict7jbVN_=0f??kc1?D^N_V!}V
zw8{3|zE&ud!wtt#-a#{wX3ocN{^Fc7cVLpE=rOh(UVME<&CShQ$^Pc<&yR~h?~0)>
zEAnONQN)d?g30=em%3Y#g#?jQgtjG8XJHTRGKZjo<cVL^_<y}l&r{^x0;IY&CIhTf
z_mgDfbe9Qc4R$t1UWlfvT<{v;BEH_wK>Xt{t^%(YA{6r6bi8+?B08O)HHSYZph>We
zaDMq2gfo3R{sG;^8HN%u-!lFq^%0t#s~Jpvd2=n`hT}Vm&y%pmo~|vw6%NufwQAx?
zHbP+;9BfPFLw{Qi@TKWnU2DDV{8!qTd2<6=v3I|xpJ&cc#Ban21hqdBi{IfIO!3|#
zmkUZBL52FH6V%fl;TsrM5ZeGU{IL#hWNAgEMDd6@=LjnswFZ2&(xqZsP4RyB{WSO~
zbV<+23QR!~hr%<C>El=k92bE#b<s|h6_G_uENl)SZGU@vfgc(g2y-nbi;i~^^z{MZ
z=3Qgk6A>knYTR{Ki1Vup73Of&nJb{@4{~+TX6XWme?f$cEIPHB0(@m+m>4?O2E*2O
zxU0Q;2bHO}mpZ$>EXU>vzcEMBNk%QiM7PMxdK4eX&*Q*W$L{iWs1t7NR+eYLMsDY6
zE!*H$;eUn9yvXtHysR{b=hv{%+z66Cuaqa!=6_Wn8F6xnH#C;Hy`wr!_yQ!|mgp_*
z$5({{H}4Gh^wjzXUwT|~se70Xw_o83({~qZ8=mvKSM^Ku6$FE``jr@1jOVSL^tJVl
zIj=%8BFUr5p#6_U*A&SJrAr567<ooCA@F)>Re!vxXJu)dIcWBwh9WI8>%Hlu7(*Ia
zb26nQ#uEC2KcwM|454Z~8ol$i2-o@|MI3T?(ZWeN8`K0~|F&vNB57;4=TA#D;c4iG
z9faUm+^PhnUX6M3&O)VdUoz9o%nEVB|DEC!_n?#X)=%!ZvkScL>~1F7`hkyQ>pA@g
zynmPduMXLby%MXRb6Yw`5~Dso#<YE6b3ObhPE~i&xm)RU#?6YgIa(M}$$zM#mU&I@
zv3A^a8<;{S+-aV!zp;T^qpJGbSB(naxN;3!8|rsNClXy^v?Wl&;*P4%yZ8pw<tAAH
zOUAkbrRH@Ip<cKmuptI-U*|ON5H0K#Gk@xSaye*r{@MihK87l+f#|SVw`vrwod>bT
zRh;fl+qwW&m^@10h8&|F2bbc#z0B-))`9>9Rh=J<!sfdPzmUO)_7{zPwcQE&>6^qV
zP~j2gtY$g7B&qjLFgZ~+YWY=j!maK+F?TwjA{Y2%>%>3ZB%82BUZB92&HOewG=H|s
z>XCZROeN*<_BS5(xszLN=etc$j1SGSdA>jrDnwjm$`_w({_flMU_K*q8k!c2-=QOi
z0YP}<6aj}l1yEf*liY}?I~C}MP<o`h{*dfD*Jpoc;Lrq}A<M!N$pF(;0;_sYU$a|1
zd|ARJyexdp>T<6qzsdT6J><twN`Ej_;rz4o2Pt#--YxQqkE=S_5L9G51NqOeo~YKK
z^iKHnux(v^d`$_PD?7}v?W%c4T9QzK7nT#PO4CHI9t?jo|6I*8aAE2TcfFz%(+M{G
z4bPzZOE-+aLEmYt$OWrsZ)J@)-lrwK3PS{;G;I4>L^a*VoSGJk4iG_SZGYR#ce?w%
zk_AHIPXrm0kTUq*lm+T=?Fcm)x?PAVVT`3}8~JVwPG#4>z9O=xJZ_jMrNMoQ<-eB7
zq+0#@;W{1S$SaedF!<FLW#2;RF~>6wizgl&gc9_)El;(0XvluI)cx2fg_qgfCAyjc
zRegTP%5v^T#&Kr3l$i>PX@8658$Ysp4@tBJ64TmjZu*t6u7dF>N6XYw;#7)fma?Qm
zRVULA!)gVUcuv+HI4j+{+(!gQD>8Ih8p!uJ)F3_ngKB2Sm%T9j-842%s-eRj#_3Ee
zwS(C>N-m15`0KLJ?VxV@D`@>7Tif5tOY4<IWl9O9eP4@IfKxy}z<=Av@KzgWnQ4dV
zCqdfApcYY~O$JaK088r)*>CMC4g$2=)mkr#ITK6K@}oQ31w+!Ad28{p8@s-r6W=I=
zMdxNyO_T81A@3pNhT(29_MZ8HClvJ$JL9mq<5tk-x9N7PYcO!3@#HBFU**Mca&$e<
zvK2cIyd+4Zsgj01w}0*+&7L-XEPpSW(1G}@dxR$C-FT9|_@d*pVp{}VNd7@6pyTu%
zQ(>Zz4Z$gT?c}V86IqV9K8D-F6Kqx5ZBfNt=dd=o-U7D$H!T;*GoU$ho0I#90vzL>
zFwY(`vHd_?OA1|xF1X%ymX<6k{blX*h;!IgrnJ{M;{(0XPJhBfNOzPBgPM={PyE!L
z9nx`QC|ldjDNn5Awx)^aJNA6ptAU&#&Q@Ts@g8(Uh+R7|pGj!=LEYi|1RvN}0A6l?
zC`_zHa!%RiRaCznWq8mX;74HY%nzvcq1|s7)6G<A_a3{ArtbCToan%gxyJP`Wjl=;
z^u(nlr8K|BQh!@qMNcX;z~pHYLI&Fe_04Y3QRyfYX5yVt1rF4%&!}ViyZBp0mQY_d
zJA<SNy|4mS6jyX~*9V(xzQusvrJe>enjGS#4&ge6uIcpmnNAt%=nB;P*M1OZG@j{f
z2Z`J{^I2$8xTQTAIM9+c<kSfAmlP@$gz`qtMRiaDX@3`4r4C&9(UEMhxHK6#&t=e2
zyVE_E+KX;sT%sd^Y5uNh$u^U5QW(^{epHh}5q6SSs3yu0Ws03MQv}SV`pS%|@%Ygr
zZ3D3;%b|Un$~6qUMT{PazqI9?Ig=2QZhm>%Ll4tM66^n7!I$!f*N<^s-|>2hBlgY7
zh#i$W@_!IwParyW!mWlNZQ|%B5w+}e5628?-+Ip+ziuIGHh<nBFW=qe9k(3<QsNZC
zM+@aUO6H9TdTHmB06e9>lS>7z3`7RM(A%Q6sH%Ejooa<iIaq>MnTRTazU){^i}rOO
z%0~`k1-J5|O}+>@1iOT)oYwr1IwTr)Mpi*dA`vI?g!7`T{&qvi{C@yCdVs^15g7;*
zmy5Us6}SCc1~Xij{nrH+w??=It1*`>1O^tj2kHh&9hV~_1{Swx2M07Km*`st6}PZ1
z2f-wlF}Ma5w}w*(!5)|K=>`?ITyh7BLzhtp2N$>fwFgNomzypJ7PnmD2gx3na8m~p
zw>SX^Bm)63w_OAXJ_46Wat9K(tp*780s=TPm#qv4#kXn-2tfju=CubCx4sMr^#TDn
zm+=({6PK<A2rjqv4+zHsmqy|T6SvP22q6OkI5wA%9|%~tN)`y(0+%HL2o{%b90)|W
zei;bc0+(9^2o$$%90(x;mzo9$6qnEv2u8Q)9|+q5m!t^@6PK$d2)Vb9BM1otm#hp3
z6Su1-2*3lEvJVIqw}2`LDg&3h5(pKSurLU-w{k8BDFT=86$ll#urLVT0s%FV5GD{g
zG&l+`Ol59obZ8(nHa9kvfdeUjTiJ5kNV0wBSK#s9xM9`4MRZJr<*nVeY>&LOm+%8E
z!4kVEQB6`k_I&-E%mRr52-2dgr->Uji-kf}R_4jI0Lp7y$;xYAS)cG7DkV)oCsxV_
zS7~GMk+JedRoYq1HM-Kl8FWmgqrzmG6P>KIN|n*UjJ)xcagyc)bd*MaC!*s#I+){#
zvcjG!VN<IrYc+K=I+*CKK__CBwUu>Jq2ns2OdNUZE9Vl;!AE+`!NwQ`8x9wq2s&Hg
z8^hQyXkx>~Dp==mUsl1Bq*xgeh<R_T=!_kC@2eOk-HKI`K*9%EC8Mx=&{ZNxe6UrD
zfo^#Kr4sdkl(Ynjq8_M!S4LX&62#UvRZ2-0M%pR#h#3url``O8Lk^{E!kDHpZwW((
zUWfz`sHg;Btke~jVsq`NrvbXgdZmq5*bQqz)Q_CUT1|jh;I4E?9#jl^3D9C$rJzbE
zf_n-*3rfmp>P5^m20dp$1?-`h=&nQ0CmI6|BM+L&1P2aiUlWOcUmhb|z%F_T)<M*j
zm2@{k4h!kflhh-`2tNyZLJupfp}U?|pgRH(=s5{`o>n+0OgUOXcAy9Y8|W@!V58)9
z)Z?9KMWS^uLgl0;j_`r`5GIj&(Tx-gAC#geWYHsiFptjC9Yre~Nl4UTxSF&~TzJSv
z<1Rp665K^>L~^@-gt;U(a7!cN#Y^adg!V*+WcJFFO`@ZrI+7SJLWYixH85G4Ocqb#
zdRTfz(!w@FW!WXQ9er|9AJF1-KxR$-;KKpzSS7R}qysFA{vcGr$pdm}tuREcXJf#k
zsMG-ouJwT){xTrqn-;#ytADmFX(n%)(Y6OK1+z(82$ZycKOlcJG?L<hsbXSr5jm36
zv<}w@3N~Pdk1UVW4j0jSOouKN?-PicA({c2NZw;r(hQnoh>Su%f&+|b;!VM7umQBJ
z2u^SUXbtV-WCH23?V<+D0(C?QQUG%X4U(#$B=8y`qlJA{P#jDUChig-KyY{0g~dI%
zySqC<7Fh`H!QDw9xVvj`x5eE(xLy9MyL-8-x|go`x@M{$=3&0->HeBJ)LZrerry3%
z;3`ih&q0NoZrTZ{DMVn=4D3{gNZexMlvsheBV32YC!g5sh}>O_Nq^-Hak!GF3twsd
zw%orarav1Kqe;s45=Voa4!b&5+9v8w&NT#kO!7;S?-WXw@)=2%xjs5^2&Xl6QttrT
zJ|ggM;5wqjkeSX4RQhY=Z~+Fh1W~K<weAh84phhI>tMJbb{qCi2XK3ft_9!WIkY9!
zEP^Y7qy7|-!rmh)-EB4@AQ{0PX~ILBt_h~lR-;L(&fJ2RaM);?zXE3#;TPGTGDpgJ
zG<^|Q*;FYE(sZI?ky)||nSQ5is)|p{y`ek}FxP4HwsjmK-^QijXh2(OFsKu*qDDzl
zvxei56MlyVsjxyfpTXrmUW%&y9L}VINuh~gknF7PBg+9?B1UiebPou&_$yYi<AB6p
zfIi91u=GuPp0oB}n3n)jrWYk27ySGQlMxu0m_Z6cLLKO1C>r$PFHKldKvAHXF>M1S
z3aO;o&PL*0*dNtFHIEx&sZTkgm?;RcU*nKgW(9qk;2fizB*9tCD?YPSWoUJ=U%kef
z21Xc1co8ZYXNwo28YO8RN$lCMrA`U_7i)r&wdzXTi)SOxIO?<vEtoL)e|=&Z!<yAg
zT}f&fN>Jy&S0~arjKqn)bttKk(IHDD_gq>#5T=<3xTIw<P`^ohOI;772lN+}zd;hr
z+<eb*Z2`@}>R{DR04|}ykTvPJ(bkklnXdgT-YFARes^xiUW*MD>Ph+@VsE=yL!=7d
z`>!aD`HA$KlJ@*$ZcTqY1g+n`3P(|fE703<de4VaD^W@pnrWF%>kMs2e+?8nPe(>!
zWjL6i-zfwN|GB^h_F*)@+bp`dtlDTl9>S&Sck+@phk<iJKEz7iCc^OnPnOghe%)nE
z*WRA3#F0x&#;Ja?!B*5pF>FV<@l?4k+(wghn`O#3qT-6F3#xxb4t2xBxyFARe8a4g
z<S*-s-34*9=jP2vpQITxUfT%^_!zQr$ajDqTd<YchOF;Q?^>USMMTXxjhV@^;}4PE
zsjmIj48Zhl&;D+0g2q=Ia$3Vw*GD%Np1@YY-P_Oc(*~5?TS>U+K?ih__*WDkOz)Pu
za{HAkUn0Yks*V4(E6R$3IgMhDkq8EQ_Z@unz%*!sUnZ<!LKo-!tJTBeE}(c?rCC;^
zDD;n4<`jk7knAjN>VvYmn`6rhyD&L@;o?CU58#9`Zl&>L>K2T{@1m}w%1ro6^irDY
zCDlXTjuPuGeFuRv(<MG00m-1wPJMn^CYx?;(VhXqL#Aii_IHpptO@A?3%X~wCbj<5
zBo~_Xm!x#kb~hgY>a+&-{Kj%*&Z@^w{sJU2D_|+oRjZ2I)!&}1sqM<NO{cwP(`CEn
z7~JRT7!!zA(oK}!xnSn-&3-2}H>-t*gcF~aK%y7eEUhYK!@T6t3S66&9y_F&)y+2`
z^RYxLf04V{mfB|Dh^e-5uf*1@H_y#^IcFYQYD!mC(RuK6_fBdm&~X{U{(OklHd%)t
z%)#phDjR`S@`*StO;Mfb<*?<|j%Tb50VgoA#WTjmf&Dzjap+t7?ENjs`PT`Bm&HgZ
z+**%VrsQJ^mJ|Wa!cA?3{&L)Mkw=mV>GUPS6PTSXZq9e|MI@DUWgj<R7g4J9rzRdM
zciltx0|cTkEkoS8ZGU#OdSxnJ&Pzr8Ykb8MSFLFAy=)*tUUZ_>(WllSZohNv1kW`w
z(CKRmJMIOz4~f%=k$u*C-R6?pDGNB%4O3!RY`N=53?Pm4bCQ>>R_0&7{d~T74<XoR
zC=#T)77}jUoXBt(RCJ}h-D7NZ{K4r_#cp4XmYPJKS9j#2`DdTjia=OmPGZ035NFh<
zW__T1Jj%#uWqOe_t)2<OJvNPl4L)AN_ef`jzqp4mQ6??S+{^LI+LY@7ZKS(sk9-C}
zT3yX*-;K%MA-=IsE+Wl8DY#|&@N9_tV~W*`^-0~Y=~fj=9D?P>r7-U153tlogp8W~
zIFTHFy_$&5Q9%P?m;wT=GBc#-vR_M?uIA+MO>#B*uHM2JtDT+`l$hf%!970VR?1E|
zB5@2dM_1**uMYQP2?{H60Xmw0T?=<>Rwiv14_vEG2pwQVBF60HGX=v*gUPo95_ulR
z+I0)p4pT{w=*PI<yWiMsUTpuGcJgH?SEbhbOosDEzTZCHUDsQGJZv3Zb&50k3X<fp
zFzQ>4`%Y>++z3!dMaA75gWm>a{acF>2mUzjmK&drtnjl^DP=XFueX;R9n9@}1NmFQ
zKD?Dok&;~%Y#TGEWeMJ!%zYH(hccrbA5<6764#=FpW_G2uO(j@bEB+xTKTZq2IQ~h
zUe$@Cmv&;k=>qW0ucfjhl15(lo_9_g-f5Mr=2mDR(b2VhxQ!;y;GObEVDCq_{Wb1y
zqwGj`*Y~%Rj;_x**CpvusICmGbZL$0kpvJXR~~6d&Z2QzE5Mpj`_7M302dtDsuLmg
z21sS2-5gM3o;v)-lKPPuX@A}PnD$896%CFCFWs(uSbZW>CU0vtO{sJ)OqOvrdb>X(
zNxnJiOzfAN!IP5(Kat>ede2m9RWPSeqNaE=G?xy9iBn0Z5M~&#p4LBkgMhjQio=_O
zJA_bdboFCa`ApUjnMvbK;?96;C^u4}bt^yVgzM|txsRU5an{}Tk~2B2%cfD#A5R%P
znWp=8Z#LnG_?mjQJGYLeipWRURCibO{#HGLp#{nDDeD?=B$tA&!f>>EEwOPw?J_0s
z6MrM7<Yk_S-^)F+#2zA+8dG3{rWc&b8pgB)-xfYO)OMSAhJOg?`K-4HQMa_EzvIcV
zG0}@!GeThg-KFxcdFiWs(XMT$PO{!4<GM=es?2@-lyPUkLy)D&+QHgrrAOK8l;oN?
z2=20y<G~-?bm?sP`#wtNGA=;i{SjM@#5oE1Fl8vg@|hsEN+MNRpyDr<+A!ANC(jgG
zP;1a^PjTdXy;d*rdi{f&;HJu*(f0ZIB^jCU)1Q2YXeZv1#m~FzCUxuX0Gw0jocZSy
zzdEIAXa3%UEdbtb(y1lO5PpvhN0;DF7e2DQS6wzRzbx-tHZ8P=KQKdSc=h(?*g@8b
zN0;vCc_$gFn(cPUcxjh`#_?Ii*8RbQ-2bE27>4d@@frl`6y^OmZ8K6~l(259Q0`Xn
zSvjGs-YZ|582K<uq_G<dOc~)G?6ClWf)jf#QxJ1s@3$t3sG5G(45li0)>s@1nY7ii
zw6cJk_C^CuEdWNsCaA#gjUD@H4*RBi`o`hW6}efKh^K97n@rIi-tT93r%P}nc#<AE
zG+33F1D$>MKl&GC@0(eX+7r)HsD4lizsvFOc*}#h18^0BJBUVJ<4oiz3a_lID(kbV
z#!lQzTdun)E}2gE6lIo~<^;iyuTL)$Cb!_{{zV9%=0gkjFJsNnEj<sYh4U!Hvfs@+
zx5SV6%!E_?UlvbmmL<{A3{;NzvzE*y6VcWO_pq^!c0g^}28!K)I59tyJ_bW03UVc-
zs*-#lSy4Do^vEFGlHt|;=vT71cv3;%(JjY%=HlbMBQdvP{H|lxnpG$LUHH(&!8|aZ
zN;X#@Qw6<LvZ&z%9sW0&s~YXc*Q=Y6tlj2MuM_u4Va6ox$6}l7+1%rmv)ARc=<(6T
ztLQcQll6}IZ{?A-In48e&eb|E@6Bm!xj(i$eCiu}VEw|3LZ>l@{ep7UNd6wV#pW)8
zK5pXV+B6G~57IQvj#KMQtG9eVj(5TR)-t;DEWgBMSV4j1$H;XrKB~>yj`1dFmhh|o
zRU9^OV6;U^*eSL)DO94-yM2bUa|_<;)zP11RsE?h>%NMx92mpVnHLVimM4KzBBcjQ
z)aL?tnqB6qZGP~O4sS4*uWzLIS@OT(qVDtHf!W<ZBf>Hqc*Ha-oW=VH5~|>fDU}c1
zuXz-2un`))s>^#{&!<!`XC(3{R;|CLAu$}Igq_h!fWTp(c=Ttqr**V%pKh|yGTsOv
zd}9x~&k2TUR>c6_x|jNSfry7MH$+y7$}nLZoUn54FI@8ia1R6!_S34#s64h6=z!Pv
z*S{EQ{sFudkTyps-R6;L7BjH<!<V^3>;6(l#z$8v6OF7Oj-j`e<6NQVGlwtSxI-+0
zWz%WAS+8r9Di5b`q%x-ch3#m#P{%F3Ds58v7Yoz=0e_VOERcUd_73q(+l(Ri1dtk=
zdLHJoiGc(stdf^Ov%DhoEG1KY4Gmae`@bVg9{zDVq(<Etz**7k05FemIezJ$)shB&
zoQu$uOeC$1U{*G3ner1p<M3ZEr>9lkr6+|Lw8fZ17U4}R%x74-LerZ|k>gO8cZ>;d
zhbpicW;v&077kOL+%H@AzFsdb(g3sg>1hS@iDV8p&Uk_S8!rsOW=Q|O@qpGO?`GAL
z)1J%I{6Kf{b1$jqDVTI6@R$x|!fSl<Bz*ZxJ1_A%ef?9@%C{%y<8-mPzyz@qEPtQQ
z>>ZC98|E%MaX13Ok2{0+32f(#MJ76VjSoFuvD1K`YFxEg?wwZ)9q(Bm?A0D9MwMzd
z+N(L`S?te+?@fCm4@CKvdU|f@c-5o_#j;WjZNDk?zC9%-DS<_u3Ct_BVt91cx!x2F
z+9NOOS}3yskrFUNcqgB{DvjXhnVqWb#v8;@{c8FfG8ZJWL-0<>y@K`ydWM&SoCdAT
ztvu;Gx)$O$f0NB&2_He<`<EkZst2D!9fyZJR1zAjHA24eT}#-xR1#*B{933W-5F@K
za}kOLkSI9ddxMvtH!cx2ntJrR-Es6e&%y$mS+2A$i)rNbcrj+|7EE^7E3NK6UHyXJ
zST;PM7EpU){lUb`^@Hi(AP>lUYGspxvOegm=sHZbZ%=i)oz5Aj<!D4hmtDHro1$Lq
zQLeNF9n>iW?0hAL##*D^H}1l4?n1p-`t~HPVGe{RH5Tx^9LC&e`GN}L50)!8)T@aN
z!!1ojZ$~e^V27YIP&l+_QSEgOiFO~ifG75y(E6rpWrb1U1$BJE-m=xPQQ}3_nF}zv
za20J`ewqZBV<9x4lvU=N(I)XL32si5+b+qK4L=<Fqu_OkcE^bH;*3ApYqfzWEMc5U
zz!P_jd=5O8C`8ii<72INK})Mkid62TcCo-!)oz%T<jb&<(vD8k#S8m@?c!yb0skr*
z<Dl-!hT}y|Y?iOs;iUb@rdX^0_@u@pd?8ohn0aA&%vpiH-@<!$&*|K%l1A}+qU{43
zS@LZi#nQ>b-NVh&)bYQSvzZ+VfQyHmo!r?hE#fPZA0mK*larI3BMns+iTM8jR#~K2
zSYCcW!Y3la|1ww7wr=ho<lO9B|0jvX$-%*wR6tJ;PSMqOK5N1Xc&OeoHy|JGP>1^t
z<@vo(YPG>;wuF-Jhu3lAmZ5F=SnJ-BS6Ja*f~4bIxO>NoE8pz|pd+S1k86_K`j^Ql
zCxwuDqU>Y_bGpZOS@b}}5L@EraTjj$Azdx@^7$MqB_AvQ-fY0t5|=ZJ`<6J;_I5Co
z@6}!l*g{-JGKlJnmjRb*5A`s5QeP0F3u|M7SDXM6+YEyDV_z>~F~Vf_mI4De-gkx~
zQ{a{AqJ&C8IwSWNRCHNWB-3;}={`hM0+%ah;yDNux+=RP4TjGz)Gx!{VN?<f=?U0^
z{WUWvYRORq_PrEHW`e^MU$wpv*X6@{c_{&{z&tV{`7uEs#Prsc-!Y@OPDOP~;OgQ1
z`q>sGh)X_4!Dn-zr46ooZLSNws@)y-kOL;)eIBm<G!Tot6*)+(n`;hLJ3fXi^s{Y(
zm93*}X(cyvka1&;15k`XpDa(MTUTPcwx!ho1iU|X_QEdP6X0EwN?FzdGe=?@bOE~x
z;EW7@RE9KhJZ|9<+u{TfEyty)`FB0fdKF9&ZdmWf#R~C>(P%S3a(>@f%#|6Q(UCu!
zhF8Xn;dmQr?kzT5RLpGsw0w?he8M-G>@V{BnGsU0F$RAu(-IE)BK(SvI`h%MjBEww
zAMXdl2Uq7){wwzn`93kdT~BlJbAP|hfx|I(NVN*}if)8S&6zYP^eR+iOUY(L9=5-;
z{)}V=ht8MDFI3woSanzZj^mS&%Lc|!!%O|hH-L-X!>gyP9b^uUFyP86n$!{)U{=ze
zFOZN_FBTn88GtpT7Nb6qppzZ<E?*}0Xd1muq@S9l7&WIyRgqErCx|5-h=K&62P@vy
zBSkPX#PAX9^9Qky3hY#h24Y_!QN!%t5uTXe?f35ro!QjF<UuhfR!hJ$RrxsylWj-c
zhO2l<AjVZp+`!>I5fIH5E$H7F6l%Jkhqu9@hnAR5(ZtC4qiusQmvVlZ-_=&el|a?=
zpFJ`9g;6Shmx(^C6Z%)$$`V}33SKH;dnx-J1fuw@y3on1z1`WRVf<WxZbFWaFfpAQ
zcA^2`0kvs_HyxLaxhP!}L+BHf&%Ij5#{RB_ov7`=&4_PIz~iIP_&7jrni+pVOmhv5
zNk+MTB<V6Sq-xFG6I6z&-`>EC-!aS=uuLsDYx={Ki{q51K`o?|33Q0gfIBFuN|sc#
zNygdUxstyb=fI{m{)tMKby$)$B7F3cqD(vZGTFjv%jjHCRo<`9msijOAF-ybU0j4B
z_gIO1r*9MLB}Js<c+}>a<I)B@sfn2e^C?CCnE0Q3YgwBwOE~MmPS+tD?wuy)x7gN^
zuGjC$-?XwctulW5NL5X=0>OUhcfZSms<~Yw2hpJ<V~PbrG6P~(A#)y8#xK8wIs(qd
ztab8OF3`&K1q8Jl12G;Ab!<yt4Yh}OW;N4)uErd>7H_S*Iv>=!CN0583wti91V|!&
zkyF8z_#lLERS1022rY0FwX={*Ivab2>%Z{doS9L!TqiNg!aHU0fCUd9!pm8rfK>eL
zPL@!V7P**_5yK~3hRCxs{V;GR=}?moAI2E-Ue7u{=Vcu%W3<%%#JIE9CEKau1!^+)
z*bP$A^Wwx+i^Vmk85Y5)HcMZ#eg?&wwUggi?_)F%?kQ^Bl5boDD2h`YX_FiEs%vuv
zx3IPt1bO2sD?-qAaKQo_%G|J{!H^4_p)B1h+@Vz;V7PG^&YEPcdP4woLBrfz3i+RT
zg$e!d+)f_t@}GCQ^_^^t#?;YRq}1%-T%cRNQ|kD2*Ca&fL=6-&MVZ9Y1Im3-fJEoh
zeF3w~W@gbgG*UHE4XD4CDcj)L9I#MN5b<Z4RC9urq<!>ibHIO-CC@&U+yN**^)geZ
z=<sjTW20Kr!_>yk>Kc9d)&!xjat*gX`83aA6%+`xD#(ZX7E{FS0<dMw-|-&*Ey(8k
z;goXtQ<QWUA4`;UjDZbata9t~OoR#{)K_^;-sO_LvC@P6#*)-QuHd$KqTDL#wbm$B
z>$OH++(uNyH6t+i3m{%UNGBi*AMK9PlgP~$?X4_S+oLT8M;W$0M3lO@ID!+o>js{A
zz#B^SJ2h|Z+E1jY$m3s^VTT!XlX-X_FyUw+BU?X)|I*j=)b20_@Ahup|JhX^83koQ
zlMFl7Tb2tCSZ4LQUYDW8$$OWFNsCNcxLm0_)a!8QE4PAEqNGubEX0F;2+Kf+{9XO5
z$V~2)0bP;5hOx{i@kF#D>2}KBTvyAHu9acZV&oAQIt1T@yTvg(_IaCQ$VHedkfX_Z
zBXbipSs9W`8Li;|p9#tx$`Lza@f=?6JV)PoW4=eY4h?NhSex@kusE>fG9WZPY#ih7
zq8>#T2LoNpA3Md?*j6&EvKc~_iKP6ccsJ2+&(1y)xCu>vhHO#=qN-o~Ae$QVt31=O
zG#%ft5mbNMps%Yt@!W~@_q?Zd^%fM2pE&WDi1~V^i7sVU&dBf`Jzj@=`AQ+}Dq_p@
z3_!9={NGK6p-(j!OBeU>nu#q?QQlIly|piZiFeo*u$Ub{-BwB9v}czzQceT8#p^6K
z>yK)D33H1yJ(^d4yI1=;<Ft3Y<LZxfHY2wwRt7!0egn0MpNXyWkoLa~hfPFw*VDBZ
zU9^{xz9$(5cC7{w0ahU<eun$ik{m98)Ys}>1BkN{ZP)K5&(-js6TQv#a0Wm}Lr|^l
z&a8l2+D423jZ;d%?%nWoP=fxO#@{=_<$g>{<J>HqcWl@1caHwgp*xon;Hr}d1rfyo
z^sQfH&C~DB6C8rC<)}#<3hiOf99Dm*^ZMWLo!!oS{$HGhH;qLdi2}?8FhpTfw*^^}
z1GoSvY}(}d9OPW&9RHkZ&dwhHoZSDojT#D@tdo`VKlC5d`435w>kINpN&~n##ks_!
zc-W<-Ik~ua#5n;HJkpXpeEgi!+`Pi%|Nku*{{sE5Z#)h@cK-j#fN0X>9W*sRhfM)H
zn;&LjSbnk{wf|(P(I?$s5NBvemI+cDy-mIuN>Bf?tLR{sjijVvUWn{!HGL!+9Dmd^
z>{=KZc~+yC%|5fGpy;s`5`mNs>LNoR^L?|<^PgHdG`2sNP7q3uFQSiQjw-ABC{&3k
z;;R(p`c%YI#6r#t3+8XYYoIhGO)k1Cz(s<u3gfXch;4*o3U{<=tHfrOB$77Y_r}tR
zX~I_&(S<LY!N|YXxW-lq_QY0<Z~r`RRh=|o)1hjBcrruCRp}8~1I=B?BGLY}1&vuo
zKhiU(2AR94Qd~fJ8Afkru^@-a0ruSHbY>FUGf=5v!p0h@9o&P`z+zK^+%AR*`KPUA
z%BB;^8{!$|f~ynT1nU?!9VFZ!+YnMiX0hrN01c7tYcwIz!-XLC-I+jD5}xyIBh+9y
zB11U(6ivAF&>^aQEGC~SanF&qKP^W!!8m5R5Pgw$w+TmX4#sMLZeZvmF`<KKHAsXM
znV>n6L8!M`!R&FoL7qQc0L#I=p`L**Y|G(c0J#@rx}JarIK={5GRZ8lRhd#UV2CG{
zqWm!hW|5)zszfRN_=cb-Rwk)LmOGjwK7_A;_Z|+ymuVzM7CP+WJ3WMSC#GDm{gP;8
zY{eJ*>#{vbMH(3GiKM7chm$L_8!kr`!pUB}Wi`_PK2I017zuDHNX*GbmUFFRrstO(
zQE4c3(Ns@9^TAt2dWQBfQVUDvIzhO}6uzMis7ooAbC~6~;gls#9vFq_)$|meT;53(
zJN-CyW1dE!7p%g!NdZkG{aL6XZU_cOXWBQSAQKVlNR5ni()_G)%j8A^akGf@l;TuC
zQJWosp^J@mkiWLo5dL5l^Rh|7hz3oa(v9?r&W5t$Ex_!lE+hTP6>SR<6HSB@meR4F
z=>hARFy%}0b)u1RU;Tu|L-M9F$qej=Vfb8QL@wJGAw}h=2d|7^8AYo7RA4>y8l-@H
z^CK6N4mcXSyUQ<xD@)AoLVgb0g#Niylm|5#oX>7KBN-nMi-JuL5Q{7-DQVziH0X&7
zG9zF&&F=`4*I)oDQf5u)dHiJx<=iHawj@Wq$tV6SinYZqv+A0~^4zBRzzP>FhFgjt
zFN&^%GLL=|2nvkyK~+K4?yPP&`n{l64b8j**fUeu1&YBh_bl*ELu;(I2Qq>HDkR$A
z1gAP1r+)<v-#zijL$f!`Ii?-UHr@g=b5#6r%G`FC+YElgFLL_e7OB?s7HS`QnBwH`
z(XK*;@~&V=Kho>P7iONpv_pU6TtO|38-dGTq%p@c%vvl={<8;_KTV6M8|AbQ6r~|d
zVb4I2=}~n-%SOzQ-+1~ePl8C-Bl{x|OyGht{(Wu>t-D}d2z4?j8fwlv^#;K#P+pL1
zgAE3EyxnAmj*c7i#rtfxHhx&R%chm>eY|WVzS|zf>dL#{n(V#(boT1UFgYozx5KtQ
zOWR#tm7Cht*m!Q0%Cg!0YpHa4FCatc)|>R;#=GAfVbh8Cp_0`ESJ{UiT~81IW<yW&
z?%3AW^J~;4w7L?)>t@gYF}ly*s4ZfQx=g$2d@Y}68<qJCiYA+<Cx%(>D>rwJtOMW1
z)PxSx1QZR%Z%kY8T#;xky2unfU)H4TpW>EDm4BF#RaZracYuycQ#3n&``C5*vOJ8s
z{}n4)Qj^btv&%$e_BwB&Ew?-dQ@F+r^OGKWSL|>E4-*AI+?+ou1?Ktraryk+QU1js
zP4m7J^f6y_0|d;F;4?oQ$qLC6s0b7U^S4Duvph^F8t#{#Mu|Gc{F66AIT01gPvVP<
zh#NXQa;_S3D^N$}4$|uxnPov7)x>D2nsRI2oE4l5i3_eNV&JEl^ArVV0+rmI=)S-w
zy8s+o^S}$8<|crt5}~cf)334sbiV5r<gvHKl9897IBJ<7j4$CV4@**p3$1&6TPk}V
z73sBWZ%&2_Bf*j$K>ci3$U3U%AySS%C6nhz0Zu&h<jPOZF?=WKRAcrER3NL3^2(0d
z%nh!+M-Zc4860Xa<rr8S&qJ(4jYQ%b%(PgNvTE7HFTwd_;24>&Y7uBm<mC~Wkom(j
zMLIUM`<qmeWoq}sZ|Skhj(h*u%!=3P7U>DCfL)$gy^7b#V(Ch)_se0zM%{upvRFuE
z%?n~Iu;OaKm++h3b}&b5TZPMuVXT(U_C?Acl!vZymk*<`@LMp!0qP3xpeDghEI||^
zmxj?~&EqYae4Uv_h5f{|H$b9(YI>3+=b@4Qo&jCo^@y%78C`y*CV4V>7O2L{Y}x<6
z&0rh@f)a;IRFL{w#?H>?=wHjc3!5I#RB<?dqdxRMWqxWrCFiE8RjW<jFocIRAiCts
z*+w%oN;4Ru|B2E7za&uqe&5lzAM8C{A&TuJeZ#COemMWXpK^B(Q#TK9H%luN9&QdU
OP5=rGjg+!9%KrfBei@1Y

delta 128443
zcmZtMQ*)pVur1(lV%z4#wry)-+twRRoQdsZV%xTD+n!ka+voP|i>~TF&{a>bT6xJN
z)6FDt;;AH*&@}P(aKI{l0thioK(?H@?a7>6B3lH@)ezQt0fGEuIW@O5;3rthTuOoE
zyLw#|7SITN?f<?FSr~$zF7=m5M4G}mAN4nHzH`j4{3Z-qjIzO@NVcI^g|tc7s*tfq
zF`GSCT31Tl6gWFE`&-mfsdUeB8EGX&M|mfCem4qKWLXwn0>BcFM#XXRZ<gozFm7Pg
zIIV8r7<cUyKR!JROmjed;P3HUPVAE1v%SAK)rgvxoUj3nIoJqdn1F@IucF_~BD|fZ
z0gG+AS~RH1vPPilHoB_1*(jbA(!_agq+@^45S1Dh<+Cs&r0K<mFnnv->kNOWL`;$J
zo*vEWA$oI%i~?QzAOc~~^F^c{q)Skbw42zh$||BI$Rjp1^4UFi`$Xtz2g85$gOx%8
z%MJf5$^hW)m>*pXDkt<1g{JcrNTq>Vib<NOV3d#xraX6?N=n6fI!ZGvrz(8{=|e7%
z;AYD&c+`@`+j6rRjbA^GL+cEwv30e%l?_c$zj?^nB7lcjc(n@}Q4Nz1vJY0ooIK)8
zr?_9jR~pA(SU)Q*N)72NJ}+kT<Xt(GYg6UxltClqT>x+s>>B2@$sni?F!Oq3g+9w=
z?<LS8x3*_RrD?Qp=p>gKIqYfaBTCp5mu)cEk1Q1>M%xoUBFKZb-~vN_D7$^iz~A{g
zkPy?JC%|M^Rn-o-g(J--5@v`+`&fg;{+Sw|U)NVBrlLW3Zuz}4r$J3lspS6a!<gd9
zEz(EBiOM*qzxJ!yNotkBvVZ!k*-h%itt(+pHKBO=DMzAe4|?Eou76wc<rC;gG)Y$v
z^ddKj!23~@flN=<gxcE$8wg`5G{)xxi>)?(5AgM;+Y2Xc5`6jM;FG9S3@dOKbg^EN
zj$f;?VI8(l$z@uiQ_JBr6KX!wamBM1HUSNqN;pcLX`2NZclYN=0+rp^j<Tvcfes*C
zrPfHNrE-7~C@237(sdq~E)i6c;zRV%vW{<O+7VER*a3|m6dDz50G6&ND^7vX!>k|x
z0U#W7_x$!uQw|dMy#2%;-vDXI6PDS3t;svfki&V}vFA@shsWLyO}%TtB}F($Afnm9
zXu2)HAkFP(T!EjqRti^AveL}*h4Xw$(%XR1K<}+K3x8hZ?)j0nu`5tnzXK_7GeXVl
zLu`h{22_bUZeM$gL(-I`){*+bFB-T-0qBQDf4K|I*HmOLUvO2JidwS6PkhbGJXRh{
zn_pOVwZ#(EW0Q*MkJyy9LXJv5m_=G_7Lv@%bu3U6{X}gRGo0JE)Eg&d%Li}JC-^*-
z954k77biSb2g#Gn_s$nR`-|RbmgeM<vrbPz-pdOEEt76pkjb4Eq;_J4^Uj59fEywV
z8yyR*U^;Y7Wu#qlFi&||f8m3sQ0D<3qe7~81@MOb=b1l%y(^1uHMP~M;}*l7tM~)y
z7=3U@X6wz!xklrsHq>XO=;viK)&>H-NjeiPcHK%Sj!tAi$>r=~KxGS__BRPJ)jp2u
zTzDb6S^}JoD4wb)cj@y#r8LJcAbCbWv$5=eC#ig-5xyanWbj;T9Z3dSlg-i^oV_E9
zP)QnMk&{QkuFU5jW^d$=;hv*LC>h)``U~Ra$l^fyrm^~?N1Jn!RQd|UF*E+sbb&K+
z(||hGD`!`MGu)4kqNuk^q&e(9fjRevYJ(1cd9yjyOkRYXjSe9aOD;YW;QbNuN(QQK
ziEL4+dHh)O`~mL#XG@2<++3yTOAt$h<x72od}#y`yP@cy1?n_W<7cJia#<<IdLkH=
zZ9@xQ5M2szT{hyr!!sF*Rb{(`>?)dFV~BXMexSbM%Uq~B5kK=R#&IxDL`e7jc66e_
z$cNx0Dl^T0B0`FuN*o#is8f|l$hPCtPO^dY7)`@hEGsFqMB2utGt%5rup+s6-Aixf
zj7)@{#T0mx6FP;s%JTCFUZ?F{WJqWXaY7qdbz7Zmx|46`Q{13L!Q`)~r6sugHsO3<
zJ4E#$VUKb||2TFfwwcjTKe5`(;<>E`Kv;DBUKVi_koZy&ZSbFg$+f-|ja=?4;<rQd
z@|j<hDjgk#Bw+f=BhmN`9QbWdx0%!VwJnfLAJM$~$Yz6l9oRL!Brk4%ohoNEi55t=
zO3O4C>qx0c;dI6|)dq!yZl1-hm+!>Ou^;y;E#_<WpupVX=q<NS>2QY3RjbV(926Bu
z8}UxY)N65{396ZascqW5q}R?D3d@2#16=PO(NCpe7%W0?A$5MWM*#44o6qiqLbUF4
zEs1}!v5m2o)3l+~^zb7jR8}E{id)!qb93f`+n3|5y}of?e;U9o$s%L#mR}zGU`A*i
zc6d*@*?B)aWaDACEt?VV;oANe!+jTW*PB4zMfMrQU2%p)zl-9r{^{_=J0w^T_9~m`
z21O0d%E^?PZx2cfNG5JBqjcZY^bVOt+dzpRb}PfPs2AXvtMf?lbV6=l_ljYY#^EOe
zojr(YwP3R90()K8{$&FzTesPgBOd;(ci;Q(pIcKNHT(TAevI-1gKa*J^{cY|qtc1}
z6a9;A6<G#b%?x4P_DBz=oIN|AFIyGVJ-Vn6Wt!yN?xtq|VBcJq+rG0=y%|#0;V^Yu
zcTLE-F^aTxA}aW2V~fw1GKU}bk^Rc9xc$Zwu)w3xkCU9{i0<lNk?pKo=-#odE5|ik
z5)~ZXwQx8YJ2m*^Tld>hWBD{o#HePUYFzQbQb6`m&I<K)5PbB7KWlA;e#B<A1?zDH
z9tJGYjz9qf-{7nM$xkN^QGSIc+Y~W{Jq#L}v%WD_#>wnKxYUa;ldD0ycpu1rT`{DL
za$xbkO=F@f+eoV))F&Z(^av`P{|3~MCbB>zA*3hYA+)cH#CL5MCMS9iVsf@(7YxVG
ze4Rl$$bIXP%!*r$3u?%tMAFlLrNh*~J@gZ-QR4v7EU@`Nnr)cTUPnwIy~}gqq<}-h
z;BVSVKNIz4{PQvfc``6t98#i}=<c$KLS#9Mj03IM7V@;V;7_&{a33hAiw2_G52Mpx
zPtdwV*3l+iUrMR23tbdY&}1#=9r|JSky-kEGSORPg4U={hLScv{NXiXMK<wJ<453r
z?_>gdU1Fpbu%cM~fpBJ!_L}MW<=@B>hYDTYDlI--TSr_9&FqzF$JCAW4lSCBDgZaO
zvS`wwrgffIMRjh(MX+(Z%#T7lTkdQP1)rb6KA0^BfA7NvW+yVAMB-&pWd&lJIt^mU
zTd=P}YGp`?oWvN)Q!k|Kao?6K)LhE2Q4D}rgESvubUNfa3!ygDP}Pz(X>9Z8U}VMN
zSMkfIL)BV`7J_%2VQiUpt@zfPKeH-S2Lt;x(w7;M7->)D3H*8R&7OMk->5UtCfDaX
zAH?PpL2pcH%l;&D{x8(1e>sOsN$;EmJN>(?yU@gc-o{sy+@fAqB`RHF4bKYj8=V7;
zS>(|TvK8}-jh`&GRwr%g=b|RRoiSL+o%ZMmg`Mis9ENmn;RIO4FXbCYu^bfGF{`2>
z4I?@P!ReT_IKgW9P6Cqcp4#0#f-usbkw|axk#0Ji>}kRyZp*LbG+FZghV(@A_<X_U
z3av!!Jlf14w?MtW^1*9B{J#COnI{NLPN6E~>1~G}yDmO>@bZkKjmrwRnseal&|wT(
zgtsm~06Q~|idk$+G4+DIgMZ<0!T0Vogl5>P^Kr7jzrY2PiGMO|wgjXq%nR>#2x<>F
z!4F@f)rTZA)BC(H*lYjjdp?wt$YI0+y$kjm93$%zaMWfJ&shA_*Bau^>0$;FJ*&B_
z#dr+!Nc2D5*mq0PZ-Px9``yX1pf6%YURsXUPz4Jqn;5AmiYwIw0;}ovU}%#_z{M-q
zBljiqb1Jv~{dHM$Pzwg$5-D)i@Y}QTlU0O-ln=t;nbz8WL9&p0J4-H$Mt25QP*Rn~
z(CACxB>k`yMVfit%Tt1$Yf%TjYh&i0t=p?_n&r1h(B}9?uJ>~fx&=<FVJR#SbVAJ?
zX*Zr?A$*T(B?)yWD1xibozKHc*7otMVJ~OG|127`xPH1M38n1+BP~)QCPL(DuB(#3
zHuFhXU2`U!PJudd3u7yJHKrQF@Lt&HOympxH5!==)oFW<37VH0;rt9}bB_fO?kK0T
z)^UnWd@^w<U6HXv%NupaQq@HK<41`$hH>I{t;@b0oO)|ttRm;0cgmvbo{d;;6gfaO
zXU9}p@l2@nFFOnA&4y7!aRe)d$OtcMz_5p0hxkb{>~&Yh96o0g7}P1rq|$QlZ+zum
zW46(k@|dr&5iQc8SvU)DNj3Cr!7e6<R)WNrqDudHKIHS(?m;oCot6Kog1X9zwUJgn
z^4|>?hFTD%F18<~D!&!<qh!dKNP};pSq_m*Xrh%}nn;Nsw7>O;kVomNs<WT{(IY9m
zM-FQ&zLrQwilizwk{dT!{ld5C$4e?K4N=!)+(FSavh8~zg`oo^o4BHUSI&*3-rEmG
zVF=68f*55tEv9%1NJj)z1g6M@`;lc+BF*kBjrA8j+1L}q!D2zibyQvYB6RkCVr}+e
z+#^53dZ~prfHTjm1wQ$SJ~AfvoRHVKe?qoWGE4_Y&q53NNI1qB&W`9z2IB}{DecVo
zB9YMhPh9pfO-2DwwS2mM)=BgfP%(pk(n1@cJXr(2G_zyHpyerD<WRTqLMw#Kt3!Iw
zyZOR?=$kSHN?bu`*p7bk9$(}6V;<nuxy;?I%WPgCZwnp#Y7M3?;~|j=X)}6|u{2{Y
zsUb7ft*FhK7>4(QG!c16e{W4ZHf3i~FVGv=>aIYiR#kwoASHy8>VV?_z3Vp1=#5i<
zQ^S8iyZTUGXfdRIh}6p@#mhegi+p^aIXUq?PBNlOX{L_U(ir2?=R$0gqCMzYYpA>L
zEvMNh+&3<qweuW)N}UU0GAmGNWQ5;fI|Is$U8((eJi-IWm7MUhcgECx`fk=Qmb_zb
zid6&31ogn7dl9;@)lmff1OkQF1VA@lf-!rB;SH|>e;q`3M|6*V=j3@d-UFq@MBC~5
z${BLb2i*@r1E4`JVyfsW<F$TFy+#S5@ZN7KKX9=Bq{qsCRS6j`RS8(&XP%pN%ZL+S
zj=rzLPKlNyG-q)4k}F5kSr|p!kG;PU@jS%n%mVB?v$WLJ*fx|g=_P#kyULG?>j<9H
zx*64fboas=Z?9y3z1_#7Q@b&UBS1&$`tMrzuT=l>w3fY{Jwn{XK+)DiEp%dx0U^gG
zU)N|x$oa_D7PTywG5*0pe`u|O<^4NylcwMPX;U1<k1a`5LR(ysd|wU$QnQb(&^`pQ
z;|_4r-G%=b9RGvJzqK!(@$1W-30Um9G<5s5;Mp$bFX?>hXM?1V^XC8N8o?ptbd6Ta
z$RI!8w2d$v;10cL{NqY{Ja>VT@y1?Y3&L9qlgH~2Qu;dSYO1E^T>+BC^;BH^+Se$5
z8RmWxE)xD=cx|gQm8eX5|F3?$IQKWQT&EIOOg7!!JaO9_1@<x4H-b@0(ic=t(KfLK
z@vnk<RxnXP58T*+R#%+SsTMH4`5hI%r+QJDCe0h2ZbeZhc+z1BQy<2e2a34d(nEk|
z`U^;GM5X~_<xHGYr3RX`<y|&LQM%qWOPyiVD8@mc6Lyaeo2sUYm$gS!)DfK!$sDB=
za3D@`dJ{e_7HT@eAaF+iB#!qT%gtz-w7)-ID0|nnNiq6+#h>5Icrfa3%cWvzyLV+x
z-A8pi4w3d~{XKHinPj-u!1|VwJK8{Ipn5gV@s(@nU#^&-Qw9QU4p*+9raQ)GrRY%|
zg4w&<C$Yw!-keoT&F(7{+zD!ODEh9Rnh>(XSE_dOlpC@rww5DpZ#Lf7uT6OfBs%>1
z0J|wc4NevopCyhetg&%d%GK-pTn`{0Dcf|&O?mPtxIE?gEZRo5fn*6Y>X5J^p?Ra=
z7p0L+r`}kc70@`xdF|q0)}K0olYO2|aupPg_z?ZCtE!TWeqzu&UM$;lIStcJRd7sA
zw>SJ;xyFi{Up?^JuAMFss{Om_QN#R}zM#%Zo}M1dwRGvre#hfv!Gi%}Svr>H1$$jb
z0CS%G8YkvSK#S3ZXN7|Yz9szEA76>i3`m11SM6GpC4kGHTmJeiL4(vD@3qJ;Yno3N
zv}vd#np54P3oe_CpaR(=tB5O8m%nS2f$xJL2C1i$7-9S4&cmU*UE{c)@`e4ei6Q<p
zFG6$Xy6@*b+st(8>==&Mhp!H$$FGv}%C5C|_ou+F6@aG=Go~@du9CCqeY}NlDq%V^
zd7!EO6(~OIWnqK$NWlCRe`E0W%ylT8GOO2;Iqb(zchOv&xIAL3bNVL+VGoJ4D*9SP
z0;%xWM(4#*M+_)h&gh<M;XPxs6P@cxjk6BII19dLzJ+Uy`Pe<o2rH33?eYw{_T0H3
zDMqZ4(<V|d)I1{K+0oRyaT4ZB->fThop9$b20r`pZ$LGCmt1Yd+N>0qz%;GP6?~kh
z*r*!2YHSUO+D0)~8rJSzFL)z}2mS4i@HKvrs)YKd{$<9Ae2UeAS(|p<AIEirz=_Sa
z(-7u?+Oo99itR0$O&G`7B3DDqMWpT>xrr$eeOvcjg+@EQ&wnwMd&nNZv4^6!jA>O7
z0DdNB<2Tf^U;LO0CWg<jK3G>*M5a6Qk^e;B{fGVZ<Z{!wq`!3k^bG5XTt&BM5aXD#
zUA^Vmu+uIBQJo&GjU2I`XY2~sxIQdnaw>N^@cdVNZnlw^+t3^Fprixm{(&9@Hrruf
zY~~OrbK5SX6pad;&HmZk%=evWU>t}15U}C_ei*<y2W3vMC}KxO1=&aWM)85P<xwZv
zZ9sL{T_i^n){uuAT9C>pm9Y3klD2Aq{-adY#@R*M!$BYWw67r+4eC0BcEj63*yAxS
zU#~R*P;>ouZ7r>_!Tmy(lMBFEQY)6);t_G*Fx{^2<$O?)nGIQXH9Ae>nQ9TA0Qae0
zdE(h5@e-bOe;I*Ip=<>tCOPry5rKVnKc0i=5DSUHBk6?aWp~FSD~=U-OjBLvltIao
z7N^l4KTRyQq2ld+5Rj_P#`z(CCRzsXV%vFJvKDu{?GJ;n1~UEb%%~1kN%&{#a^W#J
z-5us=m80hE6Wxh?Y1^V+X?fu;3cM79XZ*cb=fKm?f`Q8?QB{Z7$~Ft>CoC}p|L^&g
z@UYGl<{<|SSm$--$9~ne1;0Obk;mO+@()hRow1<eh?|RHNt9@hcQCxq(J@6E&#!`-
zJ~s3bK1S%Wj&~(x^SM`r39MR*;C+aA{$8#RP|A#U3X!U3>DQsC;_txa0ty#UWCZ(%
zo9&VQrX7#@*ec3JOdbxI(OmVu8M8-55_jMm1)7h-XgsGzRQ-cqu2U^(p%95+*6|F{
z4mTMm5Pt|BFiEBJ+HCj1B12die)Cn1m296K3z8!*ByLRVr0rlCt*A~!A~x+rH6ujV
zvSa}L_jQPQ?34@X=Ox$A1GE)`ne&0em?&ad8_rG#$B@_$mCmnuZnCzuJz3eB6P^9M
zD5Z^%j#UUv-pCVy&W7&`er;Tu`zE1sB$@Tbw{3V)DDbwRk+nrI3q1}~o&R`E`d8vm
zYbuhhIW^v~M7fWsWI=JYqr0;7Bq_ydms+CzrmK=Qo$%`y&ACZ+7I1KdWkHm*BMvQ*
zh~1oB+m-~)qes1X_)BVZj)Uxw%?A;FSnHqin+(?lJbkeKHab<Zd&*3W1^Vh)K(LkY
zaKp^Gvo#^rjfpX&uN~LP%*pN&*UcxIxEtBk`j*cDiU6_NEf%8BoMOK%C5)!E`Wadx
z@<{IC+6NE)w17Si9blM-Qk5u1u_5d%_hYYQVvJ5UQLxamcNNbk58}b<-d@P>4kgO!
zzQdqRLaq3$M;nbN-RHfHiVl(I%Asp6c6Yb9jOa2nRMLnwPBg#@8L`VHR>Tcf<u%5x
zpV?RHR&%0DtC%_Sk8ky8s<(AniO7B6a4W+Yl5ccgyz{B4GGLm(#2ka5(QgO|Px$~v
zM;SgC-J^9mTGl-K7|?90vzRmzlyF4}LT8#8pm5c})yjwc*~>wC*4%x?`K#$e^_0$2
zmBLRRG)4=i+vJ0aV>!H_zoE_iEWYj6y43d1(Bu#nhVpphJX9b`T~LMs!%v`zDP~az
zir@@3k&m>C8fb+%|Mu60Eu_KHRS0NJSI`uEz;zndRbojRx>LDFJ;oyG<FBjH@;T*&
z{SkVJ<`)(fDbAhN?+jaMb$CQ4yDa6Ve;NFW@DSgse3AYXE_SZi*vf|U3Q`7RVdczA
z?K69lk?3#0>Vf4elZXdpJ;j~s{ej9h@h7wM`ZGR82<TnR-%6AS8%KsatVIo*%U<}L
zp(Ybo`}-;PLVj^oNU`el;f^iAJ*6@C5<*L>^}w0;BCt5s1fPX5^{SgKm`fx*K%SCX
zR~c)|?2EN()koZT|EgVSO^C9!Y^T=(hNoR{j&Jb7ms9j_HCN$&c_Pz`JLKgbk34IT
z?mVSlB)~!R(4gy)gIU@pg=Dpd5vBm^;v{2DY51{Ag%6%?4nlDgc22!;5|P9T?e;El
z*@Sw=JugIlq=u^+mY0w<8%&DCK_oIwZC@!5$ze)Pf^o<uiY;PrM`t(@v^k;nu@Au|
z^o+h(D}<aGD$=vkKjzW?Jey3+<y52$6$2uz0hAOBh3<wo=|=>NCgoOkHB3dq;rP?&
zc(m?88ErjfYUGx}d~0^h%u^V08Z)$m-xmMtL_18$?^1=XO1}{ySO2M_F<#JKQWX>)
zIv4D7u+Zu+${wjtgnw(tYQC#V&3G&udjFFyOM32SN-kq8z_mM`O0)u5#6&GEx-ODb
z2GpSy$k7)c_>>>FbGiP5L#f?3w?2{FS$))}{S$Ehfx^_yO@=aTaYFkM5IFHY-s|v)
zME7zro}_|r7`}t~U^NcInoAp6NCIgXEh9DkP{g21g!v=e6IVL!HaS0E8u2D2>qJ36
zNAiH@A0C)<QMwoIj(Dltcp-5FAM6$0UjThynOV<$zBlfU$Dpz7Li#j2E?Tz%=kt$g
zz~y4t05vctgTg&%*UbKjcY8>3$l99t_EeWyS-EXr$k7Ke_Ul0viI1{vS^eqznyuzm
zj<lwcVml@d>ZJ@)`G;cM-Ghh0sMcrWB2qDAOG{7Wp@w}*HF@}M7hfv!YXl<{IpF)N
zgWxXuAhhO%pFk++o{XizrI()xw>mWH-=QMjY3>g;m>(-+et)e9uhaL@2m9J~nLXX~
zpV_~Mf{s5o2H-+&m$D5B+jTv#v^y*7!a&Al)WgK$h9!57M3yAsLehvV@6Ei3e}7~(
z^7$c(8{~?Xe5@dY3gN^(1lx{*2JlN%?(y~H3eT&y3$EUObygSGzqbGVV4)em=$z$C
z%$@Y|dEPh3N)tYZHM2K$adkE`visl7!Ppv>g)J2Y3>poBg`1o0|5<uTXDfE26$O|v
z*lQUVsthNkP^P)g4`YKur)QAIYAu$Bv!wT)ilV!cc7A&LdYgtSt4UswM#(#!$iPJ$
z+jBhT9N3e{?2vDK-4j3b@9FWnJFlAL6bbQ47|hHZkT@|)ILxG*H~O0}CExaH;poD)
zksbMv$6nIETx7MUDgg+)7&-%u1gh6j?^g`dVEX}tiVo`?1<gI_31h#>h?+?)%gfjL
z<H8s=uUV2I0M_?y4tAH}$yMCl58J6=yj7ow_U===l?^{-yKlFLk|_mK@^6JwrLv?W
zL5xqrRIc?m`;J6q+|8CLJ2M5wJz259s^_IPd>*?UBQiKddB7*gD?%$HHeRVod~TBo
z73}><UQZ>=LV_Tl)HT3)K(dRI&uq@^WP$$*X(;A6=xL-N%&%|LV$s=5ek501-N~dI
zqaf{THEw?5n!;KVLR`<8s6sj%#O%A1!mD&3=X$1;TZMHRK*FW*H1DCj4TzMB8QgBg
zp&$D^ugbug0*vvfIvntH`k9_L2j}VJSV4osaTNH+jy%s<Ccwu8i_V^^y3C@^)0VS-
zt!V#)Y~jhfl5gRN-_Lztc5Ou=_-BtRz<ar`2dq=lQeTIc1{b)=Y}YAsM|dSp(|Bv0
zz`M6aJ!G-}L7=joI#3jUk3fX&poCgdo7y_i7-{vw07}r(3#p~jy4Wve<fB-~hO(~h
zWK*uO+g*;{5$#;|F|Xk9kf?)b%D-C8tySUTl!Gvp`ZlGA^Z%)pl1sgoG;V!uy18QC
zMo{km;D)6E!%ODZ=InvaQ-wHXR~B%dJf3-h;<IX@q<^EzVbWGju<M;LiKn*kVL@t&
zOh*F|1;V*ty2D%`HD6p;ICf=i%%Q;#*rFdnoS`*PL*0W7Z;e2V?0;^={~`{g55tqU
z!S)V!>`ne*hMe!ls-P4JeL1{NrgDh^6|>Bh^$?}U%UAYeVLkv<&TGF@OxI06K8h*Q
z`6kxjqLVfzD0enRnV#j2(9y!?%fHqzppnl;0iYzS7TF+Sc1b$k<zrBAO*<c|#_<bO
z7gsPCFJxj9i0=|?zece*%e?y+zf^^vXKdonMyo@Jpd=)=zObFNdkbB!GH1WJIZd++
zN|4Q*=va>NYj`f<*T74tk3#bWHrZ~lFhYkSHaE67v9mjsF30jX2|}$3SCAba%QyVb
zbYOr>=!6k|xv?5+Y$znq&sc3Co9j6(G(Cc-T&aAqBT>x=thvu08}=lE6TWHpmQ@{m
zkFnI+QKiH>uhV7Ou}{xvp6%e(BC(9>Bs8g@DMW46KW~AQ_S;a5I4uk%sVa{i`(f<v
z;B5DX&=5n*-p($~(57A=izl#mRMsRe5l~t|_`;?@5$cW9%L3uf)7-6IAW5PSNMcTJ
z0JSB6%UUPGRM`|xgfUMGRKv)_<mQoM9TP<Mmkk5A{5iVSs$2w#7gy|-A0MF?IJ+m-
zOOW>w$1H(YXT@wcIQ&ke-^;N5aXr%fGR0x@WbDWTZ6}V=4#Ra6Qk{xr5rU$P3`i=s
z;|H~iRK{;-86Nv7b2f#k#cljMd8eN;yb#utAJVzdpDFfz#OyVT42)whMn{^@I{L<r
zvSc>3kDZ+?7zE!Vl1NRf9n7CrnNi_j9}3ouzfeci7dBG$j3K^C@gL1(mqbQ2Hn`O8
z{*?)dIjwS}jcC(AZj@t|GeE>T37{#a_vR)Frx~P^#c<$6$3R5j75BD*s~?UB{#)l`
zNz$>U)a`+?m1@{^GLu2+t6kzg!SiIHdR>TQaYPGlzt+$YQ@4DnKn<g3kSh*qN@7Wj
z!;S^HF}GmNv^r_6SF&VI<ex+1V1nZ>edMk49wq)WYLuvaz8t(5o`ua|2!uQS)KM^8
zvrKEz=daeiNVi_0hsQo{x3lM6@ZYA4D|z{lkS1Eshh*~FSNhp^!<1}K|5v{FSEA}J
zfhtn<!!tVy6F5SlIM}W;xzKPymr1}FSW0B7DDO3;JCgxs@x1k@ivS-SG%u)DoKR?6
zlsfjr%SMn0OKqdM!=(j$5D;2R;{OAW#j?E5V#lC2luo+FShNyrwU&w8Ih@hOXG~YH
z#wTa-&b<E-p;Pg=;mW|Z$CGkp@{_zZxBc9o<=gu?%GG!3vp#)&+rUu}I#9K%s?C6}
zH8&tYA)i2p0S^-`7Cl(<Z4+1bCnR>}`!a<?jEmA6h5=;Y5udN~7$9}NTb&PqprZ>n
zpQW?Mxx-vK<x?@8=0P40`_*;cn`cx8b>XkeUQ9P{6Uv>+g`}Ea0$ZGk3(Imm`O-A|
z&QmUCNE@tK%>YHG&9a$(Yb@q{OY4liJ7F#)3?I;0#iBJUO|oQ#04?N2{}0nQ8dakg
zo0)g`I)3V3IF@R4GQch5l*~iRB%Qiyp0qk`Z9SVtTF$uWlmJJsz?b<onrm_eDeEEV
zl)|5FTgM$|6eqdQUTLFo^Iq7Z%ucHh-Q<kyg<Edx*aPT_jQ9p@x1V{POS^@`xQg6f
zj9b(xZq<&>tLDr9f!B?4V&}d)L=Sb^EZ#ahyT0WaKa#CY2Dqj$i;7KbCgYCW@t<!6
z!jhOb6NXy1WpT2Ry|5+EPG^vJd^w&#eH~u3)br#-`4NYO-sd3?O<cNps|!dCpVGjA
z)}+}#BPuV}Q(&P?MP1IFy$J|P>WvC!O*oTCf`I+yAqND+$>FF<C?W|3@nL=B2WAw$
zT3bMT#<{jMKz9NOQFK%<2It}4sp*UX-vj-<yV6U%%gz@m4>Q_e%z`R|4GY(M^yLc^
zqV~719*Vx_*9af;vRyaz-ghnL0txP+bD+Co!lxD=9#ik|0yE_OEAC*tT2MQdG?C~`
zrTK+}N&_tuRoE<-$DgSud)<K^k75pbnk;YTj{xg1ARM+aIM-dv8qq*4z@1rmPKxfv
z<wzS62u(tNMiF(DpG_9U=(sdu&$+@UGsSWGpjTk%_I5r$AL9oc<KW=h<QNZrqZJqC
zezog}{JnJYe&dBPs$vTm4c!o*>M4xf#S$hISD;-IvH7|WN{nkTY{QvZeuqy%tcT{u
zb}m{5==%h}0@#Ue$xkboPesks!@Sz1(sjRn7eRaX;l{NZ#46#=2cYox(AMPqyp`6!
zQhMpb3a{1D&a-mf2*@kW))XR^ofZi<(y~S<+!V*Hqg%dIBZR!s5%u2c_^G*P9Pe|n
zVS*!l)pnEW($_bok|d_$am<p5xA_{kM?paaWZ;v{Dik-0f8r(z*In}kFz;DfXgUi&
zMzgRQCN6X^q_b~>gsjTxJ`})Rw84G59Swsrr{=^Cix*%7-W71??O!al2jUY@X}3hn
z4M!D>exsgJjOOSMz+q4s{Fim^qepqq4jy8dqhK2vcRIX@JZ6mbSiw~vD<E|L#un`Z
zETuYjhx>CGWaO{bO;RhglM=DNvW!OBC`Y~=?75@7`wZ9SZNwz7>^qzax<^-gtUBMm
zS$k_lj+pD>aWCbKn59E`77~BfCw*}3xbp*Y2K({-HPA(8@;>$K6d99T0xK`bw8W2t
zPO%@%osb9%+p4z4V1nU>QM)b9!7ocdUfn<x$h=NN$Hxpt(bhY>ObNrqGK}d~F&o8p
zBI#ZY)S$374wGxFp2IDP*-Wg~s$`Apw!>!hz)<XWn23Q-k|)r8<T1_Y=Ny0|Uflqv
zh*%cqs`672KfCWAc$3`BNty#&B2`NW#Fh86=cuBWum1~T3!fmq5a(P<L;N2=`nruF
zT{plElIPH6_>Xy0D3!X5+k^$9{_gAnnb6*|F>ea;U+y$a6T$c~<JBxQg5x<fUhKP=
zY6hC=*_Usq^H)GSK*}wQb)DBAK|@CGH^KF%C(eddaz~ig<x|g8ehH80xjt)M9V5?5
zeRULJJ@3SX`cTn#8+V(gylUSszbPt}I0kNX!K{jNnt*pns%J*r<SoH=DNw!GFF`^&
z&Qs}eGybaABimmdbNEyADWuUX5aE1_>!2Z7cOQ4hLjVY^MWpZlA$^WSWjFK&cd7`u
z)P6TmG;kI!rqsPfC>o$hdn4hX8O3j>M(JoMqa0=GeaON6YAExbhsL5jODzj3a!~44
z<!^tK1(iYW0+tGSYR40$A{Z_LvQVzT7D8Lc!t_>`*WUS!2NGkqKY=Olj7vTBy$4ck
z2b)U)^#+Cgd8<*Q54FelW>3`(qVVx`hrrE$LMN>ea~yl$p9U14EPmyRB^cCT4Et;>
zZhdZ%xE}aTH*|G%>E0kJ*{>6l3T{1<GUi#!E*yUnB;hVi|Jq1b!1Uw6+Y$yi=BYE=
zt!};^Z%?-RWIacq7+w3BQV34mV7eB!T8f#TIM)lUr1!4?r>@%lWM*A9EVBKP=C<%k
zJe~C0N*TpwK_E)SikX+GD*Xw`DAmBxyc>i%wYZ_A-Q^llIpx}5r$_*iF0y$Ta`|^#
z30cTHmCnE+24;y%>jWB|`<`f**X0@J2i+<sTVB|H4W2Uh#=@&o=2x%mLCy@VL+Er?
zX8B^#n}oknd^3r<wrq%|Zu-?K;-cT*t%tu2OJ{c5xWF1T4mwdgVptWK8ho&i3I(3d
zHf8mxskK%|!5@!}%xL=97^i3%Q3g!pxxi(^5&~_~H8h42R>S0DWmjy5MXqLfeOx*F
zq8h1A#Cl1KIzRDGa3ZdM1Haeqm}PyaUZ?ICe?p(;6ZK6;|7qUapn>vjW@1T-qGfm_
zrC+B(a0QB4Ae+JT^oj%rNVLNKc;zz^;k1x@KOeLRN>7}Fxh*)Xf&55m$Y}Qdaoy>f
zn@XW^DF+iB8rbt~WJu)ikPp&`%HziY@l?<p7ooTbI{f094UHo#M8>C;T9RuDF!#fI
z*>yxkBVT%sK0PHpmov5&q+GT4ofdcdZ1coaMFGVG+7Ky)70-!kdJ!Hex6@M8Sf^Al
z49HJWjy)`f_d)+==579DgC3J&hvnTSPj3fcnJ>3;Wj|5dzzv?dc=@or{l@9!;5YdP
zCxjuu_3y*%-vRvdaxPxZ^mayT&`;d)34^fe)aF(S+#8#x-4dN7jPejy@bWo|%Ids;
zp;JKmhhdk=pX)gdrtwmL^ud3fOn3y(wQlg{2d<_j_V&y*F9a?}`06q>J+1Mw6=bw+
zybdcWmIcF#4*I)zjA5Jz1k$eZH3nt)*`$H+k<Yoh!4@xhpO5IiYbFJM_AWyudi)6(
z3XFD1TSG`FZvz>dE%BPutQk!#_IbOpV#a|L-H*t8Hx?7eHU;I{#1{<m^u$W=G+jG%
z*l$SX-T<;dCR18{oZl`r=$DDkfPLm3g#DHH*dK70*16s#T7D2r{50eC%}BpD=^$Kl
zdSU5@)zmxLU7AR9WWlupWeIC!Gjto2GJ#P}%#v^{BUb2VLD||>6<=Hwc0s2}eGveC
z)W92`ma>ByZ%n~8gk8E42htR)cG@V@Mq&R7bhU}iTi&e<wOi-|u?ZDACWRD@qwELk
zxd3hR2e*!I<0lvP-hC{+m{;Hd)mkao&y4dseN0G^oT4EJ;TU@^_pD#iapkU^>_<YN
z;p>u_7sb?lq~NSS`~)GUKaUVBWe$OnOwhQ)hQ+X2(28)JqTB)PL$ErL?Gv6793A9D
zvXPhs0s)*5W+E&~Xc4MnM7v&({*K@iqXvhQTKR%cUQp9r6k^WnsUk9N{j5<DK~=*-
zJm@Vy`c(<4QD}E@o8%i)T|Oj}B<erXkdIi^;6ur(;Sy#H-qP?tP?s|Y_n`oV@M9vf
z6fURzLA+!dv3*2+c_Z5<yrSeu7d}BPo&!3aSSF<aM*2z<%~QCjYz)^J?72MHR!a~K
zCR)tQ_C{ApFch8E3#S3_osCc|2&Z$oq}P?I0gFE#yzkg_%M&2oEgNc@{rOlE5Nn6F
zJ@)oPFdt{4z>cfrN<^V$GZvu87#U1>{af`{KH8!{Jet|h+>C3woIR^+csO_{CL>h&
z>0Qrr#2wZJP;kb0vp$&Z3QZ;M&};3_H;+yc4VdsPHWjBC{oN7PQm8@V71|TqAaAjB
z=SagLH&CjckIYzBtjAQTj1)gI{GUIiG|AiRm>TAOD1Prl8d)$lk8c2iC||+Mcz20P
zg3A#f4-vyHZfooJrlzF7k|;k7-@921;%sMa$?>kK5@L{HZmALw)YSf@ekm(R=~BT_
zwK9}i$WhgO21Z)$1#@NH_3=&JcEKA(kKcAr#s8Qc^GoGV9amT8LenaB0*9kf{xGP8
zfT3`VyKns3sV(z3un7dh{ZWr%=gCq1m({kvjx+(QhwoL1q?qeV*@S-eg#C{0XYODp
z97WE^&O-dc-Gc_HMu^*y&U6>o(UKJug~!j#?W_vzL>02Mw#04EyDj4-{eH|Mkq^Xx
zcrTIz#p;y$iuk|W_U0TaHzB$=Qe9^w5u`b|7|F*@e6J{`6{J8J+8Ogrv_ir?#WJf%
z0GXadzHe0%H&*ShNjwRK6hD~f`DDuUl&LMrcY;s-IHDni=mAECz|ycbTGZ--Nu3;4
zIpI1cQ^Lspbi6oPRGWo47s!+GZWh5GI`<*%whReg5(#+BUgC3UGU9Xeb&|t=^&>JX
zM_nH6)C>Eedl|q5y0)G}@(UbE2Lx8up7wgnN>UX~l3kePCSe&9BwO!MmsuRr)NF(<
zVkA4m6Sx7w#ki>2;2Jrtw?IH9F>>|;t)%&65kKYGudQ=F2R016LQyuOTo!56+8U)X
z%P57ngiDcz(VEX$+%roMV)3WM{|di2mWOtxT=6YJ#Zo})sh<Y9Bmcg++-f~^buLRs
z%)qEhxF+gWu5FRD^GplBC8=*xauHbsuYZ(3w#&I2cl;)4K6SK}7mK+b+I<<#Oy+Ze
zp*68#<+F4ElYnbVg<l|ywX@fUPG^Bc14Lge|JWtfdYIv(jIcOa4~%iC{khHdu_@=T
z(=;WxOJ;zdgF{2S<pX&K*|EWV1tIm~C0dR-xjx~ei|=(%u02(;=ev1IIafxd>an!w
zP|U$4xEb@67UsW^eOW{MujZ+514V#bFYWM~-p&7n{Q$Ey&&D;(;3-3x)psSP<zVRt
zOHXcF`f%7gjL-~hd8ZPcx+8_WY>&3@{>6$@J3wr@$t#7CY;szh5%pTpX_389klu<G
z(SyDkYSn%d*+~x~V_Gxw>oS*8VV~+?(BT@<X?G|5=f9P7%AP91llbcE;uUSLPDmME
z)@lAN^h0S5^u`d~g-T0>T-zE~3wYtJt#eSK73#(W&+(hZVnALZvcI?LNz9`3VWULI
z5|F$z&ALCcE#a~CMf;1BJSJ=9WF^sao7hmjg6kz9l9(GiZ2CJz$XDm`${&vBVFnCu
zNK9+8NW5o1zmqzYgArX7U*AjXEbG*GPk+)Qw!!(7Z2Uq1t}l3BS}5()fw)m`F2mo~
z4A0xv<UyW9R;g|V8QXZlfr6hZR4hw26%dC5BWWOIk7GwB8V8-TuxXQ#CO>_#LPO|A
z;?y1im6AyDIC&;B=N&!9yzHvqLU21BRmIV+WcwxrvGV#UV?MKe|Gm7~o`p-Efcw|P
z8U$3N0K&NLSi}FMNCA3b>f%8Wj~!Y!sjKhXY$>dRjvT81Km&7TL5|`+8D#9y2F4@=
zXiHT_x{mU8W@Sw~fyaD@7%rLmMus=|+dIL4fN#sRl3jA`&pGmb9MD-o@&{m7N`>=&
zVun%E2%n?=+Ho%*t#0<H1^4Kv@NJqe!ypPpdOrMAm|$UJ>!nuoCb;<HG*G0kUP3Gu
zPv83#v8X1o-;&spits?Ban|b*|8~DeZxoB?{~$9rYXU2J1BbvpY$AyOdIO)dKP)Iy
zLx-v_93nd`qnw$&g{vho3mYfP|3>GDKszF^KN*LQ3Jpfit@|qvGU$+SoiJL&!ofEx
zkwQ~2LH6s-zhS+Jo;L;m*Y$Srl-||lb<5%HyLsqgF9R3iTtuGq9~Ftr?2!Gm(V^&x
zK?X>hC5k@w)Wyp3i4cvT1uZK4MfK(|3`yZN<|wk`)63YuyO6ew;qElN;~1O`JhUjl
zjq#>>#lBOk+#G{CeiGlFrH6S3M2D1To}R)MW36)UemE8uf;+(~f`*&|b%QEZbj@H3
z^hO2RLEddh;%*^M&#oqlmh*561?B3leiP#`T2U{ug9~$BCRbTh2#^&SB~mNBG_<-!
zIIBmFHhs!D#WY+~r_wJhjI<9w5)lNfY2b(ZvFco8ME_hiseBQ?Ok{oIUR&9hra56V
za<!%nQBcw?7n;Cj{$XA?c2ph9#KNr;TiTv=i_36Wzpe?RBP5!7f?7$>70$Ldju;$h
z03tRfQ7KmPgvkICY*Q=ap0UVKrfT+pjE&XQ5jpU1tIXbRj7T2RY<y_#T!aTG!P_#6
z=~nz)J`4H@))}PFz`P1-A30`w1;Pp4U2zy>0BcWc>@m=)krDoiy9Pb5MynZ3-dQJs
z=L;__KO|(TiJ`PhroL3YTWstfz))E<KvTVJn4`ScKoS<Q%ku;JxtG(C{K>Yy5Tb<j
z@Yc={)8*Q*n&fz^b?pQGTHX|xHu;iY7BxmxKp|uZ2#QjtH1Inr7HmH=NV!R|e;e*<
zIm6mSvVUvd%4)0W3GFj|yI#2%mhZ{c<48R7m>4S7wp~s`>Z7AtAj)`j?%;7)-eSA-
zf0J~Xo9@>0cfXmU1aHs~$lOKwxfz8c#@pe58AU<(*qwR(q?38QoFflB>2Iyc7Yo`8
z+O|LRv_vT;p0!LYi8q|<qf8eA*<W^V1PZsOtl)0leJhraI0BmGirB1UeI=IJv;rv*
z4Gx=Kh%XCX9abMh&g}11dMu$vUFmk~kH!4CQ{T3qPN=%K9Ey&+pqgg6jX7W|TO=PR
zJzp_2rm90&lg~r$0)l|gBBCfnKc9Ib*DLmcGVl1%hT3+658TRG_k|rBPxAI$w6xVN
z51F#eXGA=~(VkdLP#D@R_R?2KOg*nf<DWGD9-W-;)x6ZbdWZkAvyS73<8INdXR~~$
z?#Y>FvnxNZu;nz@mErCyzigMCrzc&7vv*QQ%31T!g`cgY2`@l^RSNrCrQ$7fvF6>U
z)R5*vB&cSze)QMK@b9s(G%#wB##NE7phZ<>gNs0#SI{r#HakYFiQn=Q6;HlRx-FU;
z+{6r6CBY?!@(-$Dg~AlpPbgo>YQ2_jzMjg#Eu;MstU^j*Us1%lI)_dHn(`f5K$hK#
zJk)K^#+vcMbS9AKp=;_$KiF~8iak{7A*PEy`b&-tBpBi?ocM^ker^9hoal1kQ&(o>
zZXBmB1O}D9+uj|4$kZcYigd##(S(E?l)F|S`|#9KVL7s3dZBELyaZg438r2ZS-h?`
zv@L$|{iDxLSRxboZLQybNzWtCyo(PW8rWYk0%sL!g$XcKo<{2MYtgcJRk0A#!lDFX
zZuOEr*vKD%i+b;h)~m(w-SZt<&Vi$YPy!&!*6;axO7yTlb{bbo4hyw<cSo6ih{5<K
z*?6IlW<D$DV8&dpk-#)+VS_nOP8(tgeDNIyY|j@Kx+B*YoW(xO>-&0#{WT0VWIuN_
zmi~%K#{ghLnU)i_(L+o2KqK@AowbQ-brojJ+7?GD*Bf-s*BBn_Y|gcr9xqdUI^_H&
zYWjlG7cC+<>CzxGbUa-#*-na#)A{EWuM86onXQoTV_Z#o#iOzDxo1v$ncjV=JIH%s
z`TIHIpl49;CHL;?LH91>y|;H|Fe``doZPgtk^y++7g0l1Lq$Vmnza>IVu8$-%qjdi
z?On~l<h&Z#5nMx(XY#+_CIgY!%~=JIpP+wv$tlpw=Y}k6y;G<`uslbltnX>MKoYF9
zrxh30`8y<O1wHJvWK$7MiDR8V0#QB^hz(-HPjl3+f1c~PZ7xjtiFaXapOl+;X>~2$
z>jG|*e!r=nF$j;kuYi@kXJK)lM>T${CRJK<J-0G+(*!+&bx~g-lyD9trR68**9AqX
z&3&^5ON|Dv_g@{Yd?d-Je7v{xbZ8g1kRXNS;jY5h<=ZJ{<PdSG$tGi}xKyB6<a9LW
z^pxau>~p1Q#^#KaL6#VSJ3@s8Yc}Br5rEN5A!<}=W<pxN7#B~07@6y=^+YgqH4`3+
z%Ni_#h|!=9kA5Y)<-dT6l~c4`?}e&HK6&q_OCea|F#NO9Y>dCO3q*<&Pv~Zfo`i|1
zC$uujP?0KUZILS8N7P-;N&4I}HTY&wn<<a(Cr6DJ=Fzo$CJZDFy%i8%>hc6M6aY8V
z9-z-4(7duTg+E<z5j%19u^+Msy-xCrg+R!tfNJpcpg;v~gC&&5XkK$eXqSw%NWteI
zQ2WtpjZT35r=^lMduoMwmw*+2j*R%BEN3eduW#+rZoCskIW$=Nq?Smx^kzhKBU1Dp
zqr6RE@7aXDC>=xI&-00ta5x!iG9UnFJefZuU*{E!PogL}S7OS?J=PS`!!9qcJK^f|
zKVafBE5H!tGpI+it&k0)e%L?2bpC#%Rtx-_jE=OS90UDa?^MDli6VC2EF8kUhxeG2
zKAAng-36Ox=Un#cfyHmjIDUhLg>nbsAK?g7^xjWFu87Kz|0dM>24;xR1qdOCsK2>;
z=(%@DDo|>T(yLVxr=gSw`KvxPtJ4j{A1Vc2ETz6e>?N<Vd<0Yy^E7@c6_OQ-Nc`O0
zou{W6i1(}OoppWO&2WZ8>5?eMsac=NygCm_OB;gPu;nlm!Lu7!4^Oillxcy2-#3IC
z!}MR};IB2{&ilL4W1mv-0K|T;O$ykaG%V4eMbBHEf=~zg2X{ix;>j-LHrrY5Ok-NR
z%ML?bahWQHJEK+1=)__lOrhH*l96#z$2#2*A_!FcbxrlTxjp5N^!KWhJ`}E#Zskif
zvePsi$?NA&e4<CH`e%2!vDnh)ps3?215Hh~b}e5koxD+m4Uu!C1h}1iH{Fq0JThBj
z8Cx$Qmg29^J-iLdMp7r*$BRo~9W#QaudmnGX%5lFG`E1pClmo``+Ub*>paR$kH0~+
zt50daz~a^-ktsGFD|sIqi#RrNw8k#-KW&LL<bBv#)fzMC8B#QBl+JLN5Q=h8n3+)z
zz}kxwe5c2n>)!+D1PK$aSP51QYt!i&;<rxAK&8e1fG!h(h*yZM{fRySXCXNtbF2LZ
zg<_HX_y2p3mHj{Fp+T@Ru`s2g@Icc7vWXiUsNL_Hr=w-{=-uQpO(lbMiEEK7f7sV#
z>!G9urB-kA1OARY+&o`Em4<OwUq%QO&Vk@=8F&QWT{1;k2zEt2z7Gsk+-%cFX(uq&
z`zB8l5?YA!{M>&eolb2CbAIwkVo33LAeT;^%k1;n*ff(~+?i@H#_{`SzSGeKB<I*1
zzHeXoZM3~H$QI_ablGHfeDi%gba<?pr;tZYIgecoHh5()B&d^2*!noS$C=ZH_gOcc
zJ$sz-@U$Njqc0@QQ_DO2edKZ_*AMMN7%XOEcaL22Lte?bKDI*9p20g0Id!!rXomk&
zYX*O)yk)%1Lu<!g!a7@=U(RO$j1|{4tr$)ooTeA+%*$$sea^`gHh<O~xlJfgHw&oK
z-;IZ;ly)NdDq8$5@L(H!NdME9<aTIhca2&*Ax3K0BKA7DUv81k&E|7XB<}((%h#-r
zTeOuq?;ch5;2Yb;kQ(>4wnk;u5_><?|6oS=nL4P{!)b8Lt88WR9iG?;Xm%?nUo_Bq
z31Jvj*z5l0v;*rKExGWM)nwpHW<{VOAKDjIyOBSWVqx$&*EnC8jZJNhoIE@HMTT3q
zWkZtFSrOw=$?O_SzeR8pTXn5?wuNhi677iL_hQDolRj;|vJu=upj~LmL4!GQB*9r+
z{g0KgUPqsy9yUckL|lU#*l92!8>-bpo3ggSmI<OqZi#H?%aSuU=^x$bgDT{d*`El6
zkP4mB-JSz$>4+|D>50gmscEk9{uGPy_59u&Ss_=S;-g<?yBJH>C{qmuEq?f1`1e@!
z)0Jg`y3o_;=2*`Gk@oRkp)6Lk!Q$GE+w*9x<?(OoweCR1R3iQZV0*S}V(cvztf3{y
z=Iw_j=H>jU+$xiYmTuhYifdM}*LNo(zFV|`R^(t7lvoj5m9`CQwZoCvaUJ?besTnf
zeSdF?!T;8bxju6>yJqq$dH&Og8mSqp_-YrDGl!Stw7;shU4N{#js9{k=W;KVlV??E
zsioozh0$c4f>e?ZkTuWghhi7_{1@5l6Oe)$U?YCb?1cT4;hHFRvk9tSl0g1r%3rZy
zg>n2?+bYYH)5TH}rjkRk;WQ8d#n<r%`-=>8w0@=*LnO4XGNc}Qd<|pd@K1}TBgZ_6
z3?|o;A{bL%b^9Cd`Rj&iTtq5F(6;|!>l_=yiW)5)+i7g8v29z8ZQD4pZQE(g22Ik~
zww*M#lj(bB=FWV$f8c!Fd#!cWdUkl9=Yxt(-F>0d3=v`_z%@Qtzna&GpoyHqx+vsR
zAdF<iHgm~yCK#R|ENZBy-&mzpFxXijvC0v$#FqbF-{{!o8){?S(H9*We5RIZ6avBK
zURyM;9HLEy%BK4rJ_A=dj}lScXal{T(#jO3br11n-2ViEXqG()_JX^F8IV}NfZ5r5
zuxlUqJ`l7EI9B|;AHb<x#M2nfA?Ob%)%+DFz#He>tXq7xBr)}80uh`EG5xU)lmX^B
zpKEq2=T*2dGW8FmDIbzL1$p~9Hjp8-va@N=HH^=~msQe4GUWNFU_H+6W8t2f8-dvo
z?Sq#3vv?bG9kS(Q!61bE$JQT|{7|OEUdNnh#=SXgpr%()B9=W=ERo>_9o9w^K0@2Y
zmKe`FzzP3ie0O)(K6ZT=N!1jG=xPd2u1WoVsEXI~)F4<8M&hW!fdA$I^^-Nwhh<W~
zW6uc(dY2Xa)HDVf`yeG8S24>X;Q@v91%Kc{STRL4A^iDY>=TZ{%0;7P_V>ura5neX
zH5MfcaPX6wL+y}jFb*}#8_IsR`g>kw#h_CveJv_Z2TQpxG)ZO$0eRC!h@;uR<(WC0
z(}F3FN1Wd(3u5Pa$&+2m*YXT8)vv%^2VzcX2<LJ};wLo22Dg^#m~gy5?jy$Ocy2fo
z-405~_uvHiG`?(v%Bt|(gw%UimR+F&2~-!Qz*8a{g*lf~gX_jycpH(GX#EZ8FDaD@
zBA3rmm}q9C!DsjMR5-;B8MMfv7i5E4?0h>h;!dUpPxuJsNL+K~iDBZT0506^<t39!
zbL%gMoxJm#N$mJr^`Cu2(NHx<*$Aftr)sN87*5|ntIIfNUYEIq7hKjBM<Ebdce5&-
zfCl`gUQo)Xs0!8Ow_~Rmwt_<4s^qP<NrsF-4D;V%T)!GXaZwKoomxTv>Iuzmx9MeN
z44ukq4Z|_t2!mlc<)w0FzUwX$F~#~yqkz~c`Gn>aHw?k&!t<y+&E|s91rt1@Oxma_
z4~iTxA88V_<_XVvd=a0Gn?@#Y5lkkO0KuBDf2lZLPf#&1D(K|xQB)eps&S?mKnT{{
z7g+9`1W1|XS*e+8UBOct(DK|U#jg}3hBi(uQSsK0p6y23LOe@j)6c4&@ayM7{_gS@
zic=@dRnK^8B^U=pCI5VgDIOJXnOu=V_&XS0h>Em)u2aV)GYx|<`b|!D!ORse3}|p8
zIEvPppZGjk-*xomb~oOxp@Pt<8^!-LHB-4J13Mrbb*hNvwJmXj6cuZyArpQnGoVnr
zfUdCSK2Q5XZWk5k_2WMzhTwK0yr_stkXMB`D{)b+pL;QML;7Sr<^9)>?pd;J_a0T*
zqZ;3SQt@^eIxcXdo(eP{a{E!93v}@%C`}E%pjk3)c0b6E_|Wu^&Yutn!&xMd{|QsJ
z#;*H3Q?ELp{#5kC)(k&L!NhkL%069CiH)hn3Utu0LS4}_rcN^EL9=4k*$T-XCuJ{z
z6vH<_u;gGXW7v@!G)`zr!u+XCKgslGeMM=)xkIcZA*_xDF`p}V2Zrj8EfAwg4Z$~c
zWqLpwDuElhawkrtsXJpcjYX!_WI89+*dRV}rmql9M9zX{yTtZ|p9d%1c?xeZu)@tq
zx<-$1Jg9;s6zmOBm=_cN8(Lb<^iqEh8A@EffVpCqMel(Ry2z+|EDfEZme@_<HM1!h
z;SCmZaMaxi{zFV9TP7+27HHL%R9kDhM-a}5lX?}qE(Q%^(c?6~xso4IPpLt>JEBU8
z9RFQ83ga2pbcd;O8Tfiu09m(hjYyWi6PxmVrotN&Ut?7VY#YQ{f^x^(dO7wa)R%yP
z$Cx%)hNCvzL_k(H2IbDkB^j~`n>3S)+jl+0lg{6z(Dk9cnw$w15@3Hx&6~;3x0m7Q
zGc+sEOADe^yhZb75UEzt>hsPkZ$ofR$lwi2jLi_`y%#~E<|?}z03By<XJ!O-^~(gm
z)){4aopq;L*1=g?%jT99rr;(uMixvF+3oT7F2|o&<jpO)Hrz68TM{Y!qWy`cI6TS}
z+ZUW$z%^24`_LU720WuCL!ebXY-MS`G5$6j^Yj!h$EP31`To2u>i>4NJIAh7!8aiH
zQAs^F=$&hn^i4&=AVDO8(v`jTvSd)v_F2Y_ZnOt$Zp!oOIrBVBBaQ)?rP`5ZAw?!f
z+cCZLV~^Z$X5;~#Ogeob=0?)PS9Ne~J(}(zliRFEheAz233wRC0{u&tsRZL7vsVxS
zm4FW&kQ&Rvu)Qa4S?p|$K8H%bBUJs9G`4<=Z$goufgL7E`yEZp{gD=}yqeu65z5Zb
zS&ddYECv*BYvj2mm0D4f>V#?&o&rv|SZFzVz~0eMQ!F1O(mRQ&oZrTdQOr%naSUu|
zx_lnT5LePA5^!u2-0sagjK**~*j|_U>b4ucjrH$os6~x#br<wp@Yfn+)P<3G8d0|S
zAQ@KDax5vo?712lr`3S+0KGtz>F&94xfIrKV7bxEoEDH`^ZGkx{~Z*f@uAv)c~yg%
zOLlQ_{=78b(LfkRlg3XJ)Xwa^8q3w~EUp@-mIZc16Ttsn>Uo6oL#WeVshBBPmCr*#
zsYwYk04b%HpsH8$MPGoC2ALp>R8;e>t}n+RGJ=>e6sqS^UUQ6@@3I=yIfYCXj}@b+
zqYAhEy56i_RSj43#S#DF2kv_HtHD;(QZZ=gakViUF>dLrQD=AhF)FDe?Is}z3jD`(
zeRuvlIIv7pHhw&tw$!BV7{MgZd9`Z_`eKAeuHt0{8A}<koMdc?CJ6ynXgBig%J7HY
zY<3Mi9=sRvT{aOWOGtTP*O11};dt3hHj5&<ePq{n(70(<)V9`hOQ7qOs0Ti|^?z^x
zoNUBQ#Q!Nq^YOtkrW1g{K>zDWaQ_$L2&mEe=LgzR|KkS|st(m+EzuwhB-gbZa&&c<
z&S|DVMUBW)UX)<w9?$C(d!!Og6uOef#368FHHwK0+KM)<urB<4Fh-v)HmjGflAW=J
zjWWxuJ)Y{v7a8KCnC4C|FWcWNb~+pC&L{QF%#=EZ5DC}3zWK(@og5i>0Kf+zK=<Qr
zpDCK{r$W=$pYLb%bM*mXjuq?FQJyd1wZ|=bKCT9vH<U;cs&DHxc6${F$>)%7<alpt
zo1TI4)@QpM%=2MQI2jkV_G*N_gG9do2fTf4pM$NHA6)Jn7gnTR%RxVPs5||1tC+xz
zE=>rmb8krY8mS!~rM0H^g3BWS_nX?`r{1CRZGF(~!r+If0_QUG$KaZwEyVV1GSw}7
zyk#E2dvixUI@z1*KVWUswcX3g-zuc~x|7SBl;CzJ1!U)UE)hH=mbQYRKv-!&(V)(t
z1w<Ua_<67`UVq_r^LsobLiZ$Y$nob$6V7U*l!-@{(lTx7q<CaONU~xBlw){G<F`AV
z@z7=X%53$G7T)Wy7K@85HDNlw5qy>tSj%~y2aeCZU9Ij^9>^X}58q|hR*+g961Ezd
zLcMo$*NXoECRRFNoT+HxbzN|v)q)|olzS39Tdyts9WgwxNDDugv(=!>2y2k7PPvfh
zn^PVJ-vxH8qQDH~_BY1>wM%d|2cM-WvZu+zIS;AlOI#E~1T6o9M81U$OI^J?#KQUY
zF})n!Db|bZ{>WqcWLaa3A?ep+c?SCa7>#ZN`?G_$04lyULO1+3tj4xs{&mHq5+gVj
zed$guLXIz1iIiwmlDD;pgj+#8RD~fbuc4r3l|CIEoO+n8lM?|z(#iNuIynO7RY~qu
zC`gz9f;}q*MH81dXB7+oBkgrJc%LGkEX$&3Y*7L7GGw6TsL$Bb!b5ORV#gp7TJFM0
z8;s4QqSV5YpG%9K-rfED*^}Ng8ap0ovVp@G{BV)L)4VgF)L}E=JKrobFYYbN3OXY=
zy)%XwPk&9X$NDcI8Zi`o*qhizl1EPLK@vJ}^D?l<l)Y<O<SY}y$#AK9G41i~9DKw7
z?%u@Vrn?0y{?ySMt?;feJ(b}qCh6dF@O}qP=*I`S@$#RRoe6t54ym6Aw8#pch5%Se
zZQ!V{YR-~wD&$M==d7|vSpKl{!F%6>_oIO18Ql!%?|d$RfocokyO|LR-fO&E{gLVe
zuodsbH`@32?GhP9pruYRt{Ya*5buxwaE)}v2ALKbIG{x83+~87jYXPtV9A<-&%t0S
zN&w}f!LQI+ykPcgX(9Poi@+e{h~W9e57+D<0bQ^0;VPLeC_=4YCkdx?wIq!K0yKhy
z9@5b;X(S&Y?G3O|3<LF3nnARvZv3vM{33$?;63k9%@o^&MMlAweR{}pAk>lfE{V$@
z$%bvt@wiFl^v(J~EJ<+NU>ozRJo})HHp>2G6W2x?`f6GR?S|LLq%s7(qi<=|(>@|H
zpo#>!sOx46IgB3)BK<Y-1=m@=k-U95>Rd|=YwHDwqV)N}RwMcVK^$}@;+|_lEYR=N
z>9iwo6|QrNfO0na;=Rk1AUZ?6iJGgTZ*J_)G!h1`iX3Czr^y%ZTS6hE49#~=h7O3E
z<jkNb_^6I@GaV6%*bmdr>M$I!BTz|Sw_T&NP|_+@#xb$|(K$)MCyCqUz1KKF?6#s(
z(Yph_4d3bww2?Knr{QerdTX%4+A^n>3Nx;fxaI}WFg*@=JtFqPO)E@+o;39cJGYfO
z%*v{B#;v%!CPiKahMBTvj4^@skN=P<-mj<L%ddKxC~S4_&nc5wMy|@L$7Jr0Xoy<B
z67bK6H}8qIZOrej<YFjq@pmBj_u(`vv@#%6-{-BL7?evo*h9|;EclvHNP`P;`<~l1
zRai-UAQ63U#}h_%<v|1-jEoi&6!QcPqQHsD!gx8Jkav_~4LtP@1SF)ZK%ayFn{)VB
z2_yrS`;|+3pYWvV7@vkd=yFskCKS_xELsgBl8jGkTK+d;?TLZ~<q#U*4C4sY4JiPx
zW7kth@00I+HXtj+b*SH#)32{eRnIO{>uUH6j$~xoj%0!YPMkIco1E5+i>tov)$?RB
z4ee?A7o%ne{ZHSvC*d=t0@}LdItZyapM6{0WM%Ltw}5aNqOS;!8mSgVy;}RjFPGGN
zWmPp$I0!WRG%4L_DSThlyYGc<`JMm*?R3AU%U&rUwlAeithVZkj+NYq_M0LH4Y~&g
zl-!DXUfy*KcfVS}urRODBBYKkAOan7TV*aL1WG}I_?s|eG8g>tQfXTZ)h|p?N<I@?
z!QPpn!mHsx&=Mm)tv-_=@3D(tS)d0J<Q6u4jhPUP{;=nQa@@_toMCp&Y$5>rACm*3
zdHkEQC{=nwUYdRuX^1f)d$ek{XF7fB{#>w5upIU8!nA3~U4z<S8g92#Sx6y;`;3-!
zS$ie*eD1AX$@&*1YPf$(GITxmd!^dV>)mSk;G(~0X4X|LFK(VDa_)ghg9JFzi-Izi
zpU9sP*(z!a`)j9aWr*Fpc54Gw-KN%&SM~73-F=u@aOS5Q0x>EMAai8NBZh`xTwa14
zC#CgIgaxAGKS23K5U#dlBN3?@cwFa^ek@_XiTfi!h)`~Z`tn8V>rG#`mjdmJ|89~v
z^d|2UrrFi|tw^M%@0brTCdw$<9Fe116u&}Z`1Cg5eeSyRlqoQVw0;9hr4SpZ=!GPa
znBiB(PABrw#d^meERaLZgocNST9@TgiK%FWj$87DM?7$70;|86LPAJA*E2sO#xYSl
z17cj1bxVgE)9rMcClLc*tJ^_u8s>xBz`6P>gjE;VYAIczy17%$b~2#mpPmwwO*$eH
zdYA&O141;N5i2H?lw5%rBWGQ-!ZrkWcW-Zn7`*Z=T_!_OVj3*Nc4z~uUkJxEx8!}$
zw8WrTL5&}x4!Glf4!9tSEau(3EgNBE3;VNU-^nJx386sG6JcZ0xdqZPcr!m(*|U9`
z5Z<d6mUCt42WMGIV`aX<9+BJBrgW8;Q1Txe4m8TKCj2{3qK*M33)DxPGcu=0O-D!;
zx#zLHe&QFI<{{72Ura@6hic`aGk<$#Ft?97Oe>cR!C%M!qEEkGjoj4)-v9i>uRQoS
z%aH0WP|z@TM=?_@8(0>}#jHb)*(rGrm}2d_mV}sP(YBQP3Oz;eT*BG_qn9Nwq~>jQ
zBm+&8`3DI}3zq`!g?0cf$B>w4xKcR~uEJSGM`d*Gc59?$ExQ!8p@iE~3cKee7J_Js
zmXA2h_@hCwd0F=mg4;J8jh(x18pTS3%?8XD8cYHy9O6*V45sMN)Hg5}(;O_tHp2K#
z)ycmU+}mTGP&}XZ#g3u>nDmy;TEKMrg65kP8FBIt9Pk3_4D7hJ0apjOY+1~3Y||D`
zqZw?&5nCSnrxUCQ0Qn0xezr+ojR6i$gP8G=Mw#BJ0itEDtt`jcuV@FJugMgPTZMI)
zXKxoo6!|&c+;<qZM6NHIz^=O<L`RLP)g;N-VC5te!_0bZ22FYN*OGI@%a$I_JS<uB
zYj~}fkuLy5zSjl41)MLH)4U+We0l%Nd_@kH63p1i*Zuab0|DeO$lt2W&_(=X&|}TQ
zJbB-ea4zHQ=(9T20F$W>j867<{a10{jcXfxO?8fJc+nrV4km+;gDbop(2bx$Wfo5s
z9*J{RHq+56f~uaGW}`%JL9L6^Nmqbj1f|)YwI!g}L!KLE<)nqVetM6Lq2tvO{Ot$b
z!goF8T=;~c3t!YqH4D&HDh@H%U++B??cmtZzGZL0(Gxg2%p_inKzeM@L%beupa1Ng
zqn|m3q>=WwPJVnojX^b6E#?pJr@w-sb?pCBD9^?9za=@!e`)gnTe{$2W%{qZV(dRr
z{;$0JkE}LBpHr1M!AWs!>Ttm$eORB0lfQ>g496NY87`7=MvxBp{`L<jYTl^IfnW=e
z@F;#ev!~tavfrCjyfXgyxOv(1Ad0q5KIr_;I6rweUM9n--nBjG)!-#y_eo&Zd2f_W
zZiU*NAJ*rNtc4bfy#AB6NRjOlU{p0B6ms#SIw-e#mZN*T*cF`qdcHJqUm+j8RQfg{
zxMe%<nx2wc4+!Ywj3xK-*JSwQ1^IMfU5??hna9!_b}JT3l42@vdX!fZUPwQ_uXV=7
z#Z6|T#JTRt+v6N|u)TUcHP1)e9aAwb5S903NuQLdhMAz;3|q<KCc-fRntAC(B3wxl
zAJOlh^Y}!q?af4L=cqi-vx|A+3}od`I7S<fMbF$3=XKs&OdRRsGs(^dx;7Gct?4%=
z_`_{crL({CyZE0bN1^#A+9`Y5yff3ESl@N3wy<@fHo~X~w@feJtSa1MDf$b|D_Kq$
zKe`4*eR43tC#pF;zRgYpxye`%i4$i0&D4*XJTUHEU-#tWGh=n9pe9FJxGg2)3kiLb
z)fl)x-Pt6U+Y|kaeAx0BiYXk{WQ*4omO7Ueel%L1hw`k98-AQ;iy1dqDRoZoSo37M
z`L($b+~ZT-YRzv%RD!yp@<I-O(xu4UMeyVIgz0PKnFKt?i?oyhlyjNpnNhw;qwiyl
zceLM0d+DYz>Oi^uzOPuN$^Ae)<9fYCfqChM^Q`pkK`zXZ6m+$%hJC$zuRQ)}frmh>
z;d>2OxzN!#`+W-5y#@()Fw5wK`gMK3%kPg~xWZU`F;I?0y7*zJk?Vsa&D3{g&o_Qz
zD!IQ9;}g-Gfa9_ZEc)5FmE#}^4g`)&x;k%G|I}o4&G6LA<%v!=QthE%=>0Y^U<rcP
zy1JiL&lxs$!=Tuj2j!+M-o>{q$w~($u|~aj>nwgH&Z)!or=kS?{a(ym73m$!wLTm&
zoB|m~HbNEQhtOva3AotLr7K(6hiP>J6j7`YmWDu#QAig}Ancy#JX7pb&exV0ovIK4
zt)>92>>!$RWL|4Wv;;FuSUVsJH?x^`0nZZM1d@O^QoJVJYB_#)udLyr9W%ccoTnwr
z^_D#LfKy8XH*xS=8P0Tk5Y~_nuXip=?<_+A_<mfI(cx1Qe4iZNgkk81L!W%)>kH`L
zj6T$JLOe_&U}0)4dGQQZL}LO-r47*iVMbzzr_lr32a>bZWs*=+b%L{;!%d?eP_*P#
zx5_7KkDDvSDibsUW@+&~K1W$-8v7jQlsn<13rAcJHY>`>P5j!4DHk9yn3!`e44ZHu
z425A7-JYvZu@dt3{!)g@lK0W0+WZ-_i^cn!uhLx{aB7bhK&)Q-K%?URW_gq$OPdt2
z$-#?H4L)mF=V5YP$eUejZevJU35Pe&HF)t$4kqLj>>HiYzHXM``zZ<EZ20A(+gQeR
zJh{XurbE)CkQ*N2=fga2#QP%7D$O9l5>bzvYo|GgRQ9%u(gr;m4g58{X@PKznyOB@
z60s}<kP-pvT*Y+&o`5yfCl?dkhcN<AJ3*agCKu<^maY;IlRzIt;Iv*vl^gEwO)(sZ
z67=QM2TVZI)gIOds+4$|{R9yO=j-A{icZ30d!$%%wbp}vBAnpb&-fW%w);Im$eq4i
zXJ>69$B8Mrnw#HYIl<yvZOHkO&%>QQnK7FI&?-6TCyT@rC5sN1B;7Ut$3kILt5DG~
zHa2RD3j^)u<eDKVRXBtBcx6V%le_m>Y<wP4TejnS4{HCp&^iKEen6c$F(wX%t#L$5
zRC~{OZ;pTNd%ILpL3f7tc|QpUcZ6DQ6AX^VYwveY#C~iJD~{nLF;XZGed)#sjnBb4
z0PCt3pGl2k8^R+jx0#`>1~tz7rM4u}RYL0|CSfz?S&UL&&a7F)HY3+mNw7eP&xDgF
zZC4u|3seODHs9)7#T)QOwefmS82_j|<gA6cM-`p;b+eJFNgk<Rc4?|jkA`!cG2uqY
z;Un|(4MeKt)j4u$Xj}Cq!Js&maY?5RXo=Tgz9+{}mu~jMaXil2cJcGBaZKDuc(QVL
z&E3d1oj{-)VNCGjqCxFmcGudt!jVt5WjM3?sh=d(11SRDx$Q9j8Bs|E50jvYS7NAt
zlcUrzLs=gfhNZAzm1)_yb58-1pLSC;;!nQCtq$hOS<~<~7hn(rwQ+bo0yCWkcrv`=
z7A1a~u62GMUH*1RvGP<%BI~}uF-J&ZuNpkjfcTjtJ}i1r$4j^+mg#E0MN$BQ)_B6I
zf#b^JSu#MVtQ`#}GzC<vr&2S@?#K@B1{HGgLXq4)p}iF7Mgxo?I_DZcnutN8gv7Yw
zy~B7n8oRb?T3{EFZbQ6?soM8}5}bXT+QJmg*-G}FhU8=Br%L2LZ{4)vQF&-lpCBZ)
zrI5652TJ!nNQJJkH1}bd9@Sd&94wM(#n)hXn*lk+b24A3vWcTHWeR)BZ{{6!dLP(m
z1<_l)jNp8IeW!fUTXZ86k*jpO-zzDjA-RlCPfaJ|SayaD*+@PA7M>ph6BlH;XbaJG
z-FQRMCFURbYs$faZ3cwP^)T(WEqnW(84wZQ5K0nENR+az^wx8($Gpv3zIgF5KeSY~
z?k)aEiw*ytDTnoPs)F-*?Bg$a)3IAVUC6(da2|Ax{4k>>HZX_VwNF?r!eRLJMjbS3
z+I{KJBkQY3p<mJMB$5OIIEa}*PPK$6V(vzkG?jTWdlR(DI43P=E0f;K7YZ?yAePUP
zuneCaV`tg0Duo1f?I2w%OGGt3RkKt~nu&Ssnhg=L#0TjG<HK$@)O&lM$<du=neH6-
zA`P5dlMCJa2nz2VSiEe6nJ5kIPAK7wrLl81Ch2jD<pWL3B}}CRIFAN8lM@U6`lgCG
zI42k4x}cISD&fB>Ez+UoXb9%cc&JN~+8{BoA?)@E5A4yBCYFYfAUiM`5>;Vy|9Sia
z(gN+voJd<-c0l~72sZH7>m`c|w{p^vH+<6>3s!3JYOJl5`<nuX?{{wLGLq0;e^no?
z-jchAgSwfe(W~SdKop~O&Htv%HGS(z&NCHl^d*zH2WYtu4K?P<-HwC9A8if3*^ik?
zO8zjuf)mQE!Tz%2l{_zu;bE0Y^_>ZEqrMVQQfgT^wYSr$eiJJ1e-aQKsGe4u@FF2~
zn51_S3Ig}QU2oV{38~>9X=&X&a^?I;!6m7{gsKv^GO~XLM43g_C7|TAYk!H8XO&Qs
zI-xQQL3D?1vhO#DKfMJ&_&V(Std`Fx950QLhgSuCNsp-hrN^bC%oc<S9@~z*dAqxW
z-D*!Boy5W-Z8qSvdZ&*A8N1~q{@q*rlF)?~BzE$ZtGnX*Mjso~^h154bu^$^rRwj8
zt~$Ea6w@gQ5dJlZvq%>SlGqjY$j~!7lXI`|w081tl<K+E*GRsnEj^Ovf@&7tlJshJ
z>DKac-n{+ZaaZacuRCyM-_I!j)Ja~)u`b$BT@}ovkd!S?l(+qXECO%G_CI99+^qj!
zHvE5dC<F@|6X$=*A~b-ebozhVB2RTYiQd!3=o_cmR0>qACR7cZE;7ymV$C7U;$ylS
zOYg6hh~Fw_ydK5D;2{f?#x>Mox`VO21vf))zv72G-@@6R)Ck3B<-3Nd7~{M6z6UW$
zcMp%fZ7Oo;lol-jHen{6t3zXTYaRT$9Du-Q^fIMzjwxy_BG7H82E}dLCksBPE6sfz
z)kGYwJR(Qcd8O^c6CbR#-Z&D}EJbjqb;U{CH!P3&_G&1U!}nn1WT#Npmnn?gYFArc
zRUoIjW?5Z??LH|yK&_<p+na})@;+s`(B1L+lWIY6_fR-99IkEjmUuiB8f=Kdq0p4X
z=KhmJA72MT3xKtldbUG09!{*od)M&y*|II(;GArA7`Autd~AcS(BaDC@nf0BGqmoE
z7qR9&I=dPAFnsDb%1kDF9Pw_R4AcrM-U={(-rs29dRzTbb_|Jo_|)o!IiLk<EfYjO
zs=3h60EEyoV>n8LI%FfhBbE8gkyHBC<8y`c9FRG^0rFYT9ZlXE^US>k!n1zU^B*gI
z5;PN`$F%$>TL^!+gPNJteusiyX+BdO5S{e})r9^*f<350=2$0Z_!uVBLj6dnP1Ns(
zS`o0HQ|qDx_o}L5^;6v2rTQY+k@7oguzLo$#?YM8w(^wVO(ye(qO+}s`dIo>6Vl!s
zqh<-#fskL17m;=IE1*vuT^qKl>d(`XI*g3O?k9-V)nsJhBsS}?Alt79?uR;rrxHS}
z2EG3_rCgn3O1-|42xHJ~0S;T4G%b{A@nTh`HhExzHnf}SkAS`LPIN0Nnh%(>{%c=O
ztIgBQK{>l7f8@8ST?Bz3%ppakrVCHaz=PaPAl6g3i)D^nhKK981n*w3qzLTVVBg??
z=-i?!!0I_UhIq=93o1eGs7QT(CXcf42Zlj5f2SK-6xEL5@Sl5dU}_P$HUp#ZYkG*j
z&vRH=FlNzARqQjG-7~5POsV0WmX<-8B0k6!Y;a48?ZIVtZI}8Ftk&K3cUzQUSS}b4
zpx0ga=U6G&V}YSz?oUuk1s4{-oQ{!zKBQWM%s4?F*5#Lphk1-;3l?alu|HGpQ$`ZY
zSkMkMS&xQz+^Kp&*XNntnlI;y_R$#0EakPz{7xzKqOiXwv5u`UUEuV8AL9s3@Sg-)
z&`GeoRB1)*WcAgy2HlxpRPbdck(P*00hNNyxd-*#0%y537lwUNbWq6YR%eO$%v0EF
zOvb1V$YLRrL$&C$G)n<>MAJnzNOi&FV(v;A(E$0RTOi&f2__S5u&BR1^pT(4f$6I^
z?>J8YH|)W&0qA$!&6=S_ykcf|mEKt7e2LA%v$6mJhu+qX{m7@I01){<I~zns0Pg4O
zQGb}kUr6-r*eu4V5w9i$tt8V}$c-BQn`a|ckf`3MLId<~$bL_s!U}SGy1n2~4j1W0
z=LkR-hkQI??px8dL2btV26~dm^hW#H`f|@sEvPCX4yEH4P8v+9#8c2cG5%eMh+&)<
zg7iD|y<Vt(zybHG2(eRKtt3A-P$u-68CS|TZ4mpx1~=U3NlfWHP;swFA4o#secfZF
zih|kbJ}s9*W}2C5#JJm9u42-Y)>F#Ydr!I(8=Tz*m*O6d`ScZ?KpzqDH?*(1sxe2-
zI?zxv0e&px#-Jajtyl!{bGPwv2Aeyr9{H2`7T!Qp5sg9#!?lr92T`sEutus(VbKsG
zA%DOM7XtTA1~0J~SAx{}fj1G$Ku3Ct*?ojc_q@bJwT+>OKWq1eSv5zeCJv_lRi=%Y
z@ORWw_iK_G!{SsMG;Qlgk&spz_bMVSnwdHLv+-e{p(9lTZ@?vqVmtdq@`Mp+hu%US
zjvMm6_dtTSffO^lxTzW?&}^bMbQd9f`)er#gn!z)Y>S5rwG*=sM$(e#y_{~QtO-IK
zffarK-8-o0j+s0o2v@{>{ilBH3{>v!&rNeiEIWi2LlO>IFO+m+MU0JR9@(ov3W|Ff
zl~i9Leo>!5Z-oMg!ys(QX|dLDGIm>jbow?2kgbKEa`GE<eQL@OfOJkfD=`&?_~{oy
zMz0@K$km_UR_Aax3pAM1Zqfu1{S7w5{mM($mK4lvmBh~B^y<q|qHO$ufLhv$A)=fc
zm!UPypND~YtM&G|OnEJ5I|31sr2;ZC@e;@)5PcY~<8$o<MSb{3CHw|gAt-S#&ant<
zOQljhH~ZS7v`2Czpm~P_Y^vUH>cl5Tz=yo|KC2z~A0QZMCyG9I;v!;}ZDEc`vEz-p
zs;9;KTe)z~R7_rJ>DSmU9YwD(JsoBe20Vnuc&QQ%sT~t<a;Di+s5foQ;z~O$9EALB
z4~JGHovsrI9zzFg(GRGOLv*a4k=52l)i`o@6Br2GtmH`bfKi+(9YLc3d)as!_DqsP
z&tT)tljdV#n7^}Rw?`V+%f983*bIc{g*#ten7uxfIR2z4*2Jeta8$AeSg78Mg$&xv
zPNDfXS2%tPgw$zc41u)zzwI=wj7amXXd}+|B33)#uOUZ9-GMv(DJiQ}mR={SS^aI)
ztS^LSD$-Go05rv^J9LZuU@7f^pzq3`taBTXj3Hs=s!U9mit10%X>4i2w|vY&1{G^>
zH)-bx@$;d--CepciKGxKYyzPplua@HugRiVuffTGEB7c$<&pVTW!{x)Poca(ic%0t
zDycI)-hpa+FEQC3$;~=qrgV=Eb(=_?VV9a9cK@>q;M{$Y;U4zcH2mw5<UGJh1J-Vf
z7o!kp-3w(86>N0OE0|rm&f5iplUgSo7g?2U%hOFiPmMw>h5G|KC?R|?tma1U#nIhl
zWKlvYEJD^fAuDMdQM%lRcuG*x!)@*jxmx?XW0KIeY^3NO_P}plZU5h+06|ki)w9SO
zy4$Tsz^{}>oKkrNt(8<Fa2IOp?2i^W2?%|jk@$Fz{L7TCzekB0n8Iaf4nW7-$Mj5f
z0K%<ksjBDnfNPKn8S*zZfkF)*qRXutRj(8uEu|wTVZ}}>5WiZDD&mR7b9(F*xH^Qf
z?_&vwYG7_^2*;7kfaOgeE~pNTRIs}2&{}*sU=vr~p(<|X@3)a>zQMw3sq+*UFDH%;
zmchnhP+uer-)}Lt2qQl-ZvZ#e+(M~q5!WYdt~7KK@{KGRJqV%y7vKd<BP}ht1G~_i
zzNxTCD^)xDCk*u$ALoc?{#LPd{Av4S8KicD!n!bxRVPbHU(livxh4Ot)8PJ_X~veG
z_zw0BoRyU$c}A56=+@SXKWIhu-!_;XkzMF*{eCmkpg?HbZ$r+Uz(r^h=!q`0S#z2u
zN#Ejj?j7|0>^oRAwTmBy<&QRP+)y?0S~XzkvIddupbz=@b~sr*z4g}<oz|dA%GTbV
zvteaGcn?~}!9FgVi2pG9f&n4)@%LwaY)E>l-x7vf%jT#i;L|4WJ}PKmdj0pv{_QHe
z$N%w1SvJ$XY7OEq!zT1U$i*3ze(DCbTB>U07Ybs98oIT97Ng1^Ngl_z|H9eUjkyEC
zC-LMq6F?Q;wv^X`Aic45QfJ-!yLQa#lNcMF-Pi66=jyJ1Hp~@!qK3Zu`Y*u;JINUI
za+F{dR2Oe{z;;2rWf0mVHD#-1G+xKik9hCX9n0phd!4c{-39O$zH0etHO9~BruNy6
z5vk$Q4~j;8Cp+cZG!%orHSIS<4>`IIj^T3I-~+WSQ;q`2{~?nhioTpEMHXK-sf*M5
zSxO>qTFUM?^J=b-{igj7f5~~AV?OgcI(j#S_9eUqSaM+%!n+$eEIPbuE}l!GIa_+{
zMs`g|wVQ0|9FsaPrbt^A)=2vhpc}UmFb@;B(DrhW&UG`Px`77cKm0kpX)L>xg(ZjM
zqHb(oEAvt*MIz4n>PMVPP<mXfBADonO|EMmUnEx)8HWV6>}<4CZCG;uD>~^hRHZ*6
zp%qycNPv%KzD?aFHn2hxbE*oHVQ1SikPt7N!p!Su;IXM+7%PvG8TQ3?YEJoUjcuY)
zqHeoP56=(#!KJr~iX48pxg_sDqFtounhmuE-_5CjDl4c)!Vuq!6=}$`ecMaDGT@I3
zO%f|hoCh;*q~7vlO(h|+PG!S&zE_Z!oz#~CFpo<9(_nc3&Ehp%w1&KI*@Rt@n^xdU
zyJB{whbE1XnAp%)@WZ|Q9fp1mIRs-3(5=1*A)>Hu%8q@TYM5TYa(dqL8_GA&Pb3=i
z#J)+64!b{;+QY))OeC`6=Jh>VNuZ}v;t{whK9Bqcts+nj{es3{kzOg>BZav4#L<=x
zRFe4qK;8=^#E&k-mfV%ADD?5L%9*%Yu<&?%h9n%A6(vAJ(`E|Y^L*gB3e?1Y=u<>4
z+87;qOQuMl7}_(Ib`lFA3h_2VgO5^j#w|0&-mmW?6=?$MCxgWOu%4}OYdktMG0d_1
zawmrojdgy~`6bp)5_&o+NDUgBA;H4|f6~4EXf=KgDD!bbrAJZ7ac)dNj8SodEr&fQ
zA+VCT)4yvDE=fueeu=tGWD|h0OIUGTN{em5Zm+b9R6S{&t!3$@V^DB15YV=uq!d~u
zmw*3+hp>D_coZ{8c2giH0Tm1irBc2m`-kgyK;e*2eY(lv^fF)Om^lFl9Xb3BXhP85
z2Mf_})_@vXT?g1?mW0{-W!242lbkBPPrly;^*W*3H)s(iZi7i2m|VS<dY@p1eFoFM
zK-SyVJVIJcrF9qmS27Am9WC-PFunL<x#Tl6HWI#ONMVZg4E?abn^1{EOJXKTr#{3U
z5F=mkXHYo4&GT8~1@A}n<^&T1@OtZRQTj|l7~bX_!<HJ`n<6%6ILvyhScO}-%tF@^
zA(Xr{9pu;@cmvMt3)IvYu%}^YRrJ9hBXZ-aZM)Pulx()w9{H>D=psOo*!S|$uRtxk
z?QO9JVjY8?=wM&fH{uHD=sap#@LUmpSjJ5RnQ`mWM9>Dyw_8y+m;4a~db}W)z9eq&
z4>UF#5IbAb)ra`NuVka-JEm0jvEA-WB7$$Q{aEQk3^XTiYzqoh?Sd7WFjT@|vn|&s
zrmOU#YvTMUv#`Az`CVi9!8VGN9(|3$hH6SyzcH`iq`_C_czy1T{BpNG!et}yF?}Vr
z{Yk_;#SfVTwO)=~@5g2gyshMc6|*+G^qO>jXNX5J{V{CJM`J$s4f+-(jk_c3JqaTM
z1Al|#U(4?sX4Q+%lb+UrIdTPm&6|VcNR)d`W_b#15-ALkI_1<B))HJ(GJn(TUerMv
ze9OI=hHmV8(c$41&y`K=n|GY>#t0~Cb{+#i=$?#En*vM*h-nNNkc(CEm<(I~+v!cT
zupa}$D`@Br^}u0A^e5>S)yqn#GJ?vp*n4Dne6l3!i!oSozKCX}tPeac6;^C+REY5Y
zxT#V<#}E9;QmU3w&WvF9K*g}kHb^{T&izlz)kQ<<KxAt^m$YEKRmqGAsNB6K0$$kw
zr|I`v@h=l9-F5WefDXRN*>ji)vgEm0_|OKq8Uv3`ePuxqSDAORdU?*RPT}qEnqdUY
zTm}lTbAr?x>^1Ijx@he&E0SyP9eE|UX~M!Y1Yx$HjcRe(Fj%Gy4lLOovH?&%6UbWI
zb1K&w)q`aG-{XE49bzWm6pO$czr-cUjvh>T@ldSJrIgBv0(uMqYhK!`$L_I1Wio%u
z$TUPmH{#ykHi}eWVJctd53WETXZzHDFvEo@u0zE(x4F5DmaVQnL`6S8b4M~YKAKX2
zZ7I-5bVexS;uwf{P*_~Ys2938Mh?hwml6@YgMVjlI&*Ig!!~ku8Hu&!U6B~)AM#TA
z+rZ`ASEhak1(cVMKZ<Lu;H!~#4WlO+G@qNZesZZ)E=+jNaal5tLJ|z=R`ile>=M*S
zW)xd>c2%tOWPfv5H(TtB%QB;zwQEd;wV;bkcUMT>0Tu3t0`p^o^~x-y!F`hqdTn4E
z4){UPh%q;U*zo0=ue6UJz1!T&P<|Rc!MV1!rvl=f1h}HZDO64d^^>s1k_sDX4+$Be
z?-*u)!l+qBjOq|;qK&X}A1xi2KO|&#PdFh`f&5a^c)8TnJ=w}Zn}-*<JVzkCbi5b+
za_rE9ZR6d#X@<oj34xxKI_AY9nDV3FoaFIACoT&Y3`{{eSP6K@rF)c2DYGM$XYhS-
zmJb<;0DLTp^JJVyq!CIK&zy68`rOl3PndCrnv(|&#pY?WcM?s;hk7=Aiq1zAUJI9+
z`0d=+(n(;&U#+ki2s-pb<5MIci*TRbeDh9g$T&gONuYlCSIt4=<eDbd&L)4;eaX?=
zgFivdXYv~xC~&uhbQBHkA;nn~Od^dQ=>IvQ0;n)^Jo~c!meTn6!96%VCOvg{zj;MM
z^^%i=Y3lDBEY&E_ySLKGn9%$FC*u%Dl-n#fMd~AB+Sk@&aJPHh`83UU$+c8j`W9C0
zPgypfRVVYItvRjNdDuXutiGlEC>5{UX!6m0z=Qeqv7=J3tU>5h`=R~YU#4eDM5*2$
zc3{@Xf0Ln05I3H6Ie!9OX=%7g18Yq~x?N^&fxxG6#fM17OxI>}<0CNpJpOziAB6(F
ze&4)ZoZ;u5TeIeXnhQO6R_J24Ic@rRT(A+YMW3eZZ@1UvF3tP%`~CEDH4Hs~Jf?br
zh~GCDx<2SN4wAUqR7%J=RyxZ9YWqaL4{$<haK!g}3COkM3n=SP^QTwXMQCv<o5z`A
z;$xadmg8ZC+H0A)6Y{MD_h1n5Pf|;o_)g&*(8Gnh(1Xa&YmM<bi2u4fFLF>yu<(fW
z7R<NQw8Oa656nUr*(*Q>H!s|nlH>|%!En>(Ot{Fi{UI;E={A?w!1pfT8-vZv1IEJO
zf^RElg&0Ba1ds81HnSJngndp3!3(=5<9}-9m2!!*$vO$dEq#fQE~4Bds*kgqgxvEV
zL&e#6)6}yAl|+0%fz$jQbfqd@_-*sMGsQJ;S1Xg4q@0aL10z{R{9q^ax_hWUE2fQo
zzR0<5wwF$?j1;PNJacG5L!|Hoz)v3bj)y@>OuwJ`vt2-y^O;H&ljWun-E6l5gJi2t
z`c<Xfl*a#SQAf6_Hb&b*jfkIxypKBL=g5*Ghm+Alu`M?EsWD=dY=vhnDtGSL?TGr;
zbZ;_&Nj_o$#CXk{^w3<8|2Ick&88<^$%U9|H<)dx77e92o8#!5L1{G+U`ARk=tRj?
zq_M;%TYDmjEr`s^8+-1aoAYA_xN^aRPsb2_c;POQMU;<Nc4Af57LsRf)&3ncsv&E3
zC&)47+V%pT;MfJ(%KOPnhM7j{bo~@;Rg65biZ3c_v4B>AWGafK!@mdi-39&S?~U@e
zOK`NwQ3j`}Q71<Vs&Pd*;FbrqdIKF-YNu?CXv0IqJYPIjYa26+=FSjSVs$Z9-Z~p~
zr-|9gNhhW*eW?4=5(|qr6CI^}XL``hlL=BX46mc{i$6^<UrO)+lJY{P^^2t)(%)q0
z*ym#hF}dX4Bp2ao^Wb_bwoH~QC=6_ozT1ac19tPnYuUiydgI~<;M@6Q##aX7CgGU+
zpf`yXhW>1AP>tG%q#2eR;KsV4tH*i(V*!BLzs;3>9bcVS9BXTiri2RP&+Wk9wM0(N
zcFqz&i~lY>w#b@=teMqVB6pYmJ4~u2v!#ppBr5RlN`Ttm%ZF-r9ft2KsYhDMY0>5X
zDo)Eo9O6N+f8!-q*GgUO4BaN0yX!cL%g+0xxEa)IXmh#S{X_d<a2E~sGlE!7uZLQn
z5ib;^gk7*gRZ#k;U#K(l$yv6>7xcbZUeAA9tpDe*SpGAV1pBv_oK~d;TC~@l_lHpZ
zmNl*uX&Hwz`gV@0__*Ac6ll*p87oW19>K+sq*H_fpn4ZC(*y;|%khYNLrqsz^0+wf
zC+=PUG2~S0SLlNVJv0K}KOZ(1(;5413tPsDM!s*^VMS_@3s-^R(Qnyd3es=BosBTu
z-MZGKdmXQxsbfo=q;GHm?e+!+b#jdxBWi69f9LIikMT~x_hl}t!Iq$uHp;~ILVWu+
z>-4jsQLD$WrDhyaR*%zdz19(Pt#jo-Pg#R+{>$BGI96V75t01@?d%q7K`zaUBI#CD
zgXsDL*W+1=O7o{HekF0#Z2j0;2B(AG=@Nn7m?o-y(P3;)wM{3mn5Z+^YL6t|vhU`0
zco3WgMEBn;jJ9!AN}H7lGN!TFUqn4E>tn~yH0Uf>o}~^_d~Qa0zLq6`6Tq+be5ZX1
zvbd7!y3>Fw^Z4Z{_BwZsmnFssAT=L;cm++xgyvFbh7C=>ub*);8eFG;y&kV)b?rRR
z9Is=PN_p+O$Jzio5d{>LTK)zee0McE^K&|v-D92XcE5+|4JHWk45kLgHMT#m*Ovt8
z(==-LXL(=S?K}8Ay8rmx2?6U1(Dspkk`LD%^v9(9nC~Xz9a05@6x6<%IV0VK5M}>y
z$QuiP=1)+VI~+=7**XwEylQ9+JSyhS8;h6Bz~iL*q^AY4($gKxHuE}&eD;l9s>h8d
zo0uTAo?K8MxjTH=*bph2eHN3rkM6yHg5uT=jtzUDZN6sD(|Nk7n3O2|b@ysID_?ob
zBJH6eSS?dXyL!F)w{iT%uYOhwJUXrtv1EusBvZ;zb$C1@-$kGllh8f4Qc9--h54ph
zKb}UBYNCMPXW`k%u{q&;EK>3;p&tdd*z_!^l*Uk!+U^f+9vVIRv%i8v^)-{L*rc!;
zV%0GJ4!$}*%ro{G2TN-<HA;81-^n6JEA4ZjW*zQbEZRWroHjrk8I`CIC@K(@x*m|v
zSagSsNsutI+-V@F^1WKFVkzY?%gR@sZ=ZPDT6+V}$b)^sU7Q(Auxi%F-=P}Q)VFf8
zx(*byt_-m^KYR7AO8xPVz)eyV<F<7Eww}OT&H1-OQ^7GT^sM`!6q@fNoh!5K_t1_n
z4xw7A+q*PuQ-%9HV-^k}WT@Gvdx*!o8!>I=+O8$Y`y?QOY|%4vA0h4y!lvIa;CpIX
zOxpq-8$M3MbU1nS0g74aglR!8<Zvpw)9j5NQ_0Zj^8{_>DkW^co(g`!FX;UF$4!>U
zjs;J0XYZhZG|aNvaO5DwzMQ~ChTyYf0#(GN(x6RI4Ke5)jO{!Xt9Yw-x9>*|yt>7_
zsl<RMEhV^Jzat0cMA);tGX28(^kG)yIEDZQSRQP;sv2-_m^I>F=CzF`Z#|;8Z~w&m
zJFb2VFK}WgVH>*O09TI=r@nD<Ta<&(%tMzI5qupbHlv&O?L&UnUEawB?I*mQkj8JC
zu|0I?b0%(4^A>k>!lVS)hUspiA4KE>ua0PC<IIED($~6DYi8W#TTbndQg&$u4!l4S
z7Nv|NiAoUXeJf4ZU`rP%6?o~8{H4j=h4l>Asgr0j4nujm_>>)WC`OF?Ze}laZDNT4
zt>YCbr5y3{P~aRTH<^FIaR9wq;cRmKj_~+U08h4lKpKR%QjPaF7cWNy!G){L{$4n1
zO*v)Mh1u&e22h9%dtuT|1d{|i9BCi|wjKFH2NQZY;2(%c6>k}{Giu!vC*-w=ETd3h
z<J{Gew0Qk^vzHFa@g{#XN_6&LA~<mOjD_4r0w-pt@={&5ozueZ@z+hy4$hkvFv)Hd
zxgLjWSZ&<w7vkWfZdK*H_`W|4h7kf@_;s}*N;=>ybua6`Z<BOVM9~)I{RmueRy%Zs
zyX8@g-n9Bwy_zFL#I|ElQfk_E`yoBuXfuJvC9GEBzOsR-74x&(G86sksic74u@@g$
zisy}UoO<&l&Fa5`;F)3MF~wT<kM%e=!e|SZ%z!hNf387#s2z>>nMPoC#JMWkVn(41
zB!J{EYlzh3P2{4kyf=hZKLE}f*wj-7t~{BoqWnX{yj3D<52)_>C@I5+MIZ<h^8Id6
zgcYg5wmL`(^q4KI-9_@Yd_7v8M;TehM!=>NoP5C;L*(4a1uD9j@xdpHg`(QhxXJN@
zP!wo~`y6DimtXM{pplBnLnH&vK2lb)hz`Tck*@11WLK&{jB^n8;sF8wx=oE@;<GOE
zjt?np5zjMn$LUK~MUHTdyMcl4<6tTd-Ja%@VTDMwQ~%zfdUcJ5>RL4ETg}kPGaNI1
ze>_zg$f=68p_C&v<=eABM%`YH4aV$FHu}4WY|x{XI(X6h^8|*#Sb_bGfj=Wug!<Gl
zh?b2@1VgZGitfsc4I0QAc$VsyOi(I<cCt<X`lXO*ZCvJ*QDGY1v0aJcWadQuN}g!$
zt5gO%aQ{%5%io}3Nd%#-;f3h~MT*b}$sL!0CfBtT>k%M8)42?7srU2zSisT2_m_nB
zg{ty+-L8vhS?cpD`@<+1>YAR*?M`s&gfjC{Nsja1UqTtNLJ<HaG5+EQ<ZM3OpAids
zFb%4ugt20kVCRczve*E0s$M_GIC3MZ@axie&zCs^rJx!U22!ujh`p|akz~E4tVTma
zRJDgh?Fro|92y@=bd$A3eMiuoI5jYSY2`4_THaOzc>_B$ltr+J5OT<mdczuqEqAjf
z!xvfy2dd72c~5{@Qn~ey9PLYhQi*_@1)}q~q{`26ae<NHdx(0cU})Hb0Q1LytqFzH
z^69eU#n&Pg(5I0b%33Qg9D>9j^6(*SE*j5VL6||R2Lj*l?dvgx<M$Skyv!6OSS8r9
zQezR99V!}bPGk!s;3ey8zIPP(V<@Oc1;@$CKeg|Q$K?WLw2}Lv>Ldt9Ii|?C9c<*)
zNizW<dH;T&)fn`6y|fdE7LyXhc_4gk#Hnz|R@gqa*X%MIrNp%58PEDjYMw68%13c3
zrOQLf5FYT)cd=Do2{z#s*$$W%nR)Dm2+EOz4`a_3&Uf6Eur`RsYVp`3fym)ju;El#
z#7Ze2x=aBMEsTlxQ;azA(7$pSYsz3#$$clCr-utg)9EUcAhfkAE=a}DVQ;sr(*)Mv
zsTUF$<Rs4gLXUOQInj6z;KRfQ&SS~p2Oo<kn8Xt5@GXy^BBc$^#b6j|#W*6;5$@ju
zwT_aWXK;Qrxh>@ejMh<AT0OHM#HQppW+~rFWBvfr@`0I0Y;w0X(D|^c>iS{$-sV7=
z=Da`3T=0?r6}BK&>X4@NhqAU;qB#VujP2J{>e;35uE`WBi<v7MJ$KTHcZ9|Q9<9-*
z`2xsHpeiehYLD9-SBog@8UA1eXQVM*wnR`|w=Ys|p}JK?c^!LIOv7)*NoxNwlZV-A
z97cfRL-7G_MQm-TYMOAj+o7kA-+<)>I#<j>Dqhak6MTEG?dAdVq~=f<jefwcpKK!g
zZ3az|ZnEijE^*b6XfI*Jy_^-tgj!ArJUTfoM}Dd~4I6T))<M8)5eqLnJf5+}nVFV}
z=M`^<KBrDWgnAhx{Bk=!mc|ShTJ9E1lo(LRlHOpmBk4j51Nsg`ZQ1&WEFP)ZV~ZGF
zMP{Kn%d64%TadB!{F)~GO>s|nSPb=jKAgx}j#UXEhNvxdFh-tDpd=sr(LHe6={53_
zJopB@_Ov=%KMNs08Kw*xN{GOqU3toknRd&FTm1m0G_v}Q93CZx5(}-_5=j;{P8&Fg
zx2OKl=5Nsw6j>-fMCQUaLC8mOzsryPUu?Z&kS0;Lty{LyW!tuG+qTWFw`|*1mu(we
zwr$&0x4(VPx%=GMHzFf4BO`xitTpFcV>~1OqxN{49S+oNAK0I>vAga9Y{Z6mc2hP)
z=#{loz6!?m((Fv|Te=RsC;9s2$~)CJ$pc}utc_XYSqhoG2y&Q^(zNd0YTkd-TRpcx
z;H4OGA}R%<`5d}m@2OqeEpRe~D4moL10eK8WxYg5SYNNNjw>apa}(vf(3VjxCwWKp
zJGB~~I5eX1Y<YOI^p`?N-3M)!xHfn|hCO${@gzkQ@#;8%Cv#q$A8%X5$0xaA;BM}W
za{^<4Rp`62kD)GLnSi{Nl3Fm<rjx4-LYQyHWR*0K!C=!!rHewG3zeKi<Odj;4=}l$
zFaxqLP|W|u!--T%jlfVrAW_l3*eDNojK~u#qV?4jUu&d>{02$Xz@FxT1+r)Q6D+-(
z5jlTa(9QOB#v@bONwO)7Czbwwh-$<G@yn_R#D*Mjz;;HgqGf9=51imq?TA5#^!CNM
z7GI?cB3B_nPzv!!zt-#BWjUI@@t-fFtqtK8<n>(0?3^)2-=ZOF+Oc^|2HSG@S#?TR
zQsSQ8Bq@C?Qq56*&E7&(bOMe;g;`*kPLeGa_^CGjab2nK3ov(8l~5`jlq^K$RG3%1
zq8h|)PC8C*Ozw~IF7th>k@3ggv8nGn*Bero|1SiBAxG28;{uKX_x$)7AAre?rE1hr
z&2r&nl^wKQ?%=FNbufD_4cJ%W*y?@Y&)3f)-`640K$ZO9C>5>j_}aj~?-r;dpG-IA
zk2bh|gHyA_dF+UM8__$65K1>%aQ@>h^vS~8UYzD>Y-YP?Rx8`!|9;KtxYj;g8Rb%n
zfJyu(>@k)DD`XcY`XGcEwp9=+STKSJGbBPTG=q6B&I`ZMCev43BydT~TW=Yw{CLFW
zU;krhEw^-1p(5Ld2Q8WB2OQ)dwfqVqw{5YAsM+SNn}YVzk5gNv#2(+e{G2HEi8x4l
zI{E+8_p$zmg9ysW$@X8s!?YT705RFbI5Njk&?V@vbh0uK5LO`T?(e6!cVMLadYj69
zSf+)C^Ru(j6*&i&RnIIy*PVUO&Q`D22h(4Uo||JQT-$2`j-IqOt!z7G2A6Ki?M;m=
zhUqx{%aqr<R)MV_8}075hA8Kmx4NgC!cB-jE}PfFzVharEgN{+=*TE1fWLSPd@Wml
zFcSH47(|#>TN2oY2_O5CYTw?J7oQs8o=*x{J1|b+bJ$kbA7$L#15QTj;~Wmf_35i5
zuCZrsHc;F6F>=<fi5Xl%i`df{xTClYTAdEjp1>i;tDeoDzrQ4`a5p-v83vnZQ_VaG
z!LY!TAJeAUA+hodgecuM02j#dpsQQ2$oQlim$xw^5{xWF6!IUG2FD@6P06Phf!M8i
zu+;kKi=DMPx_z$0LU9RpEM~cs1H5=z$y?2OE6krzbJ0!5F9dTcHOza|T1VCnj%_!|
zLAV9XlJ;y2Zn>9^x?@zpov!z&U+MbIHTl~b<S`s__PJ82@~syR0C_(o^qZ8T_K__6
zuTbh2*Qe0Dsaf>4!jhH0Abe0U_!<maB7a)D7@@ANK1X2%z_hEX7Rw_GPNxX*yii%`
zyKCyWh1cbG-t_F*f94*FO3MM24H27H>J(yDDrNms(UFk`g;ygU)?&w3Q<dF6elum4
zyA{d|0Z=L&e^t1S0pFJ6zMq>+a?LpFIR!FGKej0!>+Ir^in4P+dVw}lMGbiMV$xz;
z4Sf9qtGS+KJw3OMOZY`3jS{nbqin-XOVFfy0wHk|X~xV&MMMe}^U@+G^)&M0!w&|;
zF|{G}E-Q)Xe$D7y6oqL2d^!a5RUxA1J_|1tB*ZCBejw3qfKu`=?5gtaYi^(OLaLa~
ztzp6)0yp39?QI){E1q4L&LqJZQNyqEl9D)VPSZha<D5iMsW~Lr!C!htIPN*3WJ}#i
z$A4ajb1!Qb;;*I22QTBQz>3_GR9f8bfNXLfYPX-n59Zqo5-jE**_&zBd+wuOP<{o5
z(3bzK5^d(A17Z_MFi9jB7d+184vKFYGz6*_x42zu?6fSCVN?TZqt5IU_IX$j1yj~#
zUK+mX&^gDLU*+awid=Ug?!<!ss3tjj(~-6%Ouji3AB|s@SQguvGO>t(If;r6Y~2L2
z!#~fuV)1&EHXI3h6av>rPtP|hal+rH$Ly-Xn?^~30aS|X`K_e)3L3;nu5fPT7qs=*
z!Go67g&R=rxlxplRZYuEDDfc(YzrU}R%Vr5Q!cf~6l+V>v>OVPBjU@OhN#6p)N9n&
z{7x?iL+D4}U!F~}6G!2Tf|YQN!xFMczoEQ|)AI<1rgt{%PI(mIm$UXsX%mFr<zdOZ
z)$xrM05jcJW+!C$#g^!8euXv9VMSv!_ZYGX{7CSbi)#&O@_-B=Snm+!+Q2Y0vs%OG
zD%Kc2Tx<bbD(?Ouw=Y9Zxbuv|^bxx8Ua61rhLhOW6YnLZ1Q1cHCRXJ~jDlpS9ID|_
z82Zx?YVlBUg(#n9aN<6)xNny_@v?<M`c!x^fJgo;Mt{z93Qh;!hpEf47P<Nr_KTd1
zjY+J!qe5_e5vR+<KCdq3U*prW@XB;~acbqZSyu?<3$xD<$zK+&32-iZ5|C@w(aEmt
zlSdV*qDhre%GJ_(s&L3;VxnC;=Cgvrqn^VR=dW0QSO}9*tz(VN<i^xhOI42POXgwU
z0DloN-pl7UE7ywij_^iZ#&{HMV0v!LxdElvG#UkehY%VL=wZf(-L8%ti{%j>7i9+P
zHnHRk5p1+5!wE@6XOnZVqT%S*Mjag#ntn>m$tCfjz2(G!8<VK+!>{s3Q`2g)_Ia<U
zJpATWun_O=XbK#ZH(Lk>^(-hudvMQJ0i0NhAt1OUfM@~BONJ8{JMqo_XCYNk^4dVq
zBNb57VjIfg%^Lxy#Bnf3w^PkdUm?&&k^!6qqhhe)5N9&4Q$w5rEj_g^+;@CA@G%8k
za{MDJx!xdbNn7xv2~p;==y)zksCQzRSOIBRc`Ik-t_rpHe$~fBuR~s-m{>b60Cx0X
za{_@TqMe+X6DC?*7v81B19p?OKcNuSsU|vQ^Ke2T@PRdzvD+n3xNyUA@!-a5RxC53
zwQuR;;F?S_4IPi0Y*|>*)4~Swot~R;Qy66WrD^H_rw4=mOcW>sltbF=WM^?n!2`a;
zmHLsQvl38;1mSF44&Q6x3c}q4pzjs(!*GqA#V83^LdyV*q>-30fo$FhjVO&q3rD+1
z^>>0bgI$|rY?KsxT$E65r&DHt$nhnEO+AY{3#+TR<B3=9<yfq0n@(4yfN<8v;ODGM
zH|O`(;HQ)LSXlkCM8Zv|k#o^=bggq=`m-~B0=n+|Tq<c-5uCmW+)H#Qpfyy9m@@j#
zIkNi}7z#IDhYam*ON=j?2e|bgT!m6Fl4ci?bfkl~a*0;E1NS0$JhSdKl1CZwIU?bE
z3D!LO%HQ!i5`}1eQ|jg8>XenA71A^qa6}5~G^2cEyyt4xkN!M4v_f~{{^3v`>0eE(
z-3Y_=4z)FFJ>Eg1;uUBGfW;U(U3{ov2w35QvGRq3A-0|sEz!NIZ(9VdHUp1JP3h-#
z-;<L-AH@258TP>;(0lj9ER@;?J^w%mrN;opesngKme(a0B817p{A~L$q@s{DE9b1O
z?(ppJO_wru<TZ=TAs9&j1TYt(EE6`!GWxrhY5@w{Ip+4A+&G8{fJl>=Y?NrNk4!rF
zJ5@A}7r#IO{9_H(5bwFP?@SjRP-vYO^%mZ{5n^=ityJf$RYBhnVNNtkDf5In(HK92
z6H+!aU($6u>wZotbic%aGG7X<kw)^x5@eY{g)J%;JT9ahhfC^ZF0dpGJH@9k9uITK
z5gsQkik0XO85xQS*gwe!@#djvHv+)skBgP4Iul8wuGOnZYXT{zJ?l!oH_9Lh@iY5h
znhsYnXPQk)gS}JGv*V>~@koW|wN+7};>RQ}G^EYfR$>^2A(yN$J7Jp150+F}YE4AW
zYvE}gui03^7QtHduXQ9?iS?TQ9*fbNQP6MhVVv6P3k&-T*dmSk5da+n;dn8vrHlfs
zyaLhd&B**DHx^O$CEI#EUyNXM8Zl!LaDAj#5UTFiMk&kDz@V(iB<h%oxueB(5DX$l
zqM}i+-PoJMGg0KEEod!99m87YBf|&%BV{d7+vG6Eh1QFTuv0pK%uoDtsSmR>q*4)A
zFIPR=%t{~w)-`EFMIO3CnG-D)DMi;r%POOoPQwJ3Q-849wUNjlORVBcTR1+XLlz|^
zB5C<Q5MqedLyl^K+O_l&1HxTD#7dHGSuBu5S$bumhNL@WP$NZ*vh*jjKE8NLrf<)h
zhyHAwVWCZ_P^ZlHjW0K+>ZjOLDjIKfSCg;}MX^EyzBv`6jOgDH2VD8TGK#<-sQC7J
z=SXAl9;m-)g!@Q;+<u9I6pjkFV@Sx9_OfJywcZC(S!~Q_9>wuU<FU|}ysG}|k3}d-
zjOE3+zA(WfQ&?7ulZE*+)e|dMoa^RHFvf9^Nw<o*jZb4TMS+1}1+22X-aCA-z6Jk@
z!=xkux#&IJM1PBxf%YK-AE(L}rFOdx0C9<m%nBOoen{fXpXyLQt^yZ|ND}z3M9@1C
ziBQ_9ZAq4rRt#VPe>Tulf0U$HiAv-wA|S^yO{*v|Frqb*5D2pmaa>|sZFoWga6WeG
zl1R~%`>j{lqZ=Gdre5@q>4G?BdiPSdQ`%RwfIDtZbQ$}HepnOPReS_2uOP2p>|A!8
z7YRnMo8U$;dgJHJsRmR0_h6=Sb|;MevY~EaQ91Hu8sbxZi1)t{OF(4Y{%4`h!us!o
zrGpZI@xLkKER6qEbOVO9H|%k^Q2uk;;`9<?j2<P10tJx{+tt`d&zOE_<Q_yApR*lp
zD0Y7yIQ99JYu#)r(?VbGLrVy}c5QuMSE1<^nJkK-8$Bl`MpA>6c^H1BglFcHg_qPk
z?9|Z0U1QmvAIBaclMJC$xIpbWT_fqNG_(}yG)s{fjR0||444~_C1vIwR+=E!VZzu$
zC(kcrkK{%`gNQ%!?<E)3VbqQEH+f~|#$~MOMghqj5=>`|EKCBeqcV*&RKiTO=}(Sa
z3%5dI2{p`;f-j0GVpnH2iY{5=#;k<uVdhiS(Iuz2GWORdE|pw{S&wW1i=LDZCNgM9
z5`BeOh|Cn12Dm2LZ4%2{H}$Hff&2o=j)1Zgr(v%wH0h|gL@DD9Cu=tilIW9s9Rrq@
z_6H%QGCGZ}Pc9PSpn`}-XK=1VfS5GKgUgF@i<AjD0E!ZZ96({zwhO5SRgo%}57G|H
zVK(^<!8jlW=A3CL2)(ar$410OH#ef>DYO^R1Z6})1aQV$cAS-nbtEMcA2s)PZkP%K
z_*<}`yBcB|bt!p<Ww2S$itggJi({o*%Z#Nbj-$KC8jWPy@%~jGO8)EEKeSAB+J^&n
z8)ZyXhsAA-=UZcrT~%ZXVcf^C7HpKET!%EMD&0VJTcHYPDUVmU#tq6Y3tZG`?IV7>
zvK|=50Z^rc!=t6M!SC{7!TdhlmhgwV*5LTOGWZFVzGY}Tvu1u{e7)|ZFiT&`da)kk
zH<B9-A$zm^9^7ZqJ#BM@>S;PRMARclVaro_VtlpYGOjJjPCmfplXcgQE#lz09@CpD
zm3cXi!O<s+k~RApP6;8C-U5aD+u#*cJ-SM90f2<}EYIRtJ)IqRC(`mQiotB|ukkU2
z#m{oE&294va@c<dJ`L^tYr)}=RpN6K+WYBx6nc|cE7J=8eI3^c!62TO_0R23TbMzQ
z@yqG*q5pQpoJnWOAyW)ze_i0K@r@P#i}B-p8waKNx5^u6Db~~-U6bC5^NvCp3*6JT
z766W&_E*rxuj>wbl@;q1`|Y%v>Mf58vKt}$Q3AM~nwqq$2Z51=jsDxPwNDimhk_8q
z!Mrd4!FM<$<!un%F@$!X{$u}u3N-&<wp3tivj6LTBi-T0a;(GFtwW;+3nb#}<do;y
z>#+5k4@+KjX<j^tabxC8&tLwpCDu*YDu66Mp(Bx`w<*vesVy!Tu*kR;Jw%l8ZJ=l8
za7v=Y-!4wt5O3NbvN+LAaRjnv(POIN!Mf7QGgn~IF5KT&i13h4{V3BA@}vymoEU*3
zA-<x);>`DB<f;9>b}Zox#bI;1KX{O_b*e09hzQITwbU<;b$ii6x-bf+&;6c-Fn}tE
zo}&IBgdttua0!sUQ}1Vihhy&rS&T}tIRCm<;mId%LG){pPh!#3)<oVh9YYo>k)Kb_
z-m5V@Lm(gK;<146ot<3J8`LEi7nM|rcB+Se_cWcDlGbYbb$Sb{w92MQwf0H14y(@W
z620kpdW)mk=0%I^LL{N0=q2?b0FNr-Z90qwM8UIRvc$lhu_=#LJ?0DM6Eh0MV-vVa
z0`a4;A*Jd}{$b>-PG6x`V{3(YjQ_DSO!Q+F*0cCqt0A>Je#0?MK*94r`M7A3Et$GF
zWBq$Gp-W*l_1dVUg0WUr4aTcGZ<hDnya~DgXcKz*)K|PWIB8W80?T(Bpv^{{pdfpZ
z=Lki!vM^q#^qKP59|3$?$^PM_uK=vY-`F#%EZm`TP-C$1gtL4NEjRFdk!4i?`pr}V
zexMK4V9AA(=Ga|zkI|h~WIZ(nm-ZWvNOOgRG<+QHQ*qSbT75D>=psp4kSzP!DFQ^H
zo3|#?($W@>)7XDJ(UbfX5C=njK>~c*k|96J(K=-++?>i#IqvMsVdWhSbn7&m>0@|%
zr`v{TIHK&#8+Ug(R*&~PfH5!C#Iz$3p&=Kq@F5>+=Rkx{F<%*9DZg;0M6jfY_unPv
zP1oA|ZYbe9#v@gThtYyY9uOi^GWe@EOKyY%z`8JVVD(V1(9sQm{wRNe$JC6%yTj@s
zJX5>999ffuw(XD99P+{-c|^GrMHds~LpNJ{_$H@(>yoP@@$i4X8P|j!-I4#uyEta!
zcYctbaesJZZrzrNt@jigS^;xu9!C~;tw=4sPZIxl6KGsRNn*==rc;08jr#Z7)UD=5
ztCL=+%1@A?9k2m%b5+J7J-DebidvFil5sKq%9M?o*uwIOOt02t*CPaODf1U#1b;+g
z>f2GwYsvSzAauI8cP8QXU6a}(I4kusg{9e|VVVJoG$nZ~il9omV~Drb8eK2yd{W$K
z^s+kCJK0|{8EElEHpUSIpM+wNDrXAtji4IMy*K{$=T8E<-gq7QB0q{%vmUgB{|FnZ
zBd-n8w*n=x>BIdJq9eQDqvr{vb@kGf{79@?6!A5+O413H$nZL-H+i1v<yffCiAepa
z5nniJQ+OU9+P24N$&ZdIh7Cf|jNiM3`3{&$2VvZ=kH@RQ<@)HvGl;vqbZ~Fgm6Q0|
zlLD3o?%xFLMj)$E^@1S|RBUbJChfZ)<;pwpFdyG6Zi)ZmW+Ei<LhrO`U31~*>Qw9d
z`&hR1nB&v_aw~j$v=nq(G|AT~?PuO{ibiNNa8x{L>W<uF->$H}gre9vIW|{jZGo!J
zQ8Oc%-jhJv3kO3Xs)@!r8^JIZ;C?F6+`TQ~I9>**JDrS`cwVJUm#)jRcf73t`DHr3
z;Kv(8SVeJana6FEUSMmK4<tS>Yq1~Bx9`utzB6v?PuK_fQSw(y&{V+l&I^O8)7c$h
z0(nq2Y}-ab`A))xqHxrNJI~*s#YboHVyN)BZDbG&b9)zY$6X<7OVQ<<Yir3dq17M`
z!;%MZyLIM#z7F2NEA<PHL#(|F6I0Shvp;fbOR)&sXw#4W>a*B#H`3^+qtRPMzfMY)
zh`gGs#T#sq7zlSeM$rvY&^CCfkZ^uxyZt2#bL)Xdf&q+L`D*+frZ{?~08}W!%v}3`
z%iz>S*1?~#O5EU5<18eozK@}lRst`xiU>t+^1Q%CfLJ^}*oyrFDLNoc|39lIRyMAr
zf<aUeF0TIs98v=?C!G&l?s^QE(vrE<cph)%@K-LX4piA=H7X|h$AM*pS;~)SfXY+u
zpUyBSmGMXUB}F+p0{lEoro$%pd|&sbwEi=*e0vsI#!V!(k0j>X6u_-`RLr?}&Bh4(
zxW1J!4fVR4$g<+iy>E(!xT;;aOIBBU`Nb%P@|zDN-wUvkv6sF)H&4Tm-q}gu#YESW
z1963^JtD88?^n7&4lt|Jt6Cz-J;R!#0qbGTABn~IrpD0xa+EC`N9f@B{`juVU!fX5
z;APPA12nJemG!tV&02`L#zkW>x*r>ykS;iC$)rn@&f%*WRtqAn-vx{}(h4%(#V)bv
z`h%<_cmQh4n;*=9R9xW#%LyqK{Pol>Owiso5XIbpG8(%`G5jI4@YA63A-r^P6{He>
z&2`qWoI)=RVyPkfCXgv{JS_}28HoUoH-hRcocTZwo>^g2m<^ZhBo&^{RYh(x&vfI9
z;+W+j>q)@C;mVVOaiS;&B=d$`7IV-;n2GD3l>sthtmX2>h@39^D})}do?(l)ue5E3
z;PDQUhage-6IJ$e^2p?Orkv+iGbl*?2?YoegN7wXhsy}Ias7ZJiQna0xFDgl#QUSr
zy^zPz0ER^@)p?f*r+p5zaZiyT6na;7!A?%*U5swE=$VN}{Bk6pbo{rj6760ceMi&0
z{{p^JZAKQVE$w|ceTg?xTWv2qDHOv+CQeTj2!>eFY3i!_m-$E?wGR`{f;V`12iakN
z_T{5>#!fKq3@dXuMf<d(48Sw-ht4X7w`fEfLqvk$9ZnxT!LX9qc2B|JAdq!~4l!5U
zh^)%Js}P_ZCXwgAER?znf`6NXWu#C_$O71vA%v!-2-AJQ**|T=tBqR%MRyVcLbFVZ
z9_+n~7i>1Ah^z#mPEw7;?J2p<a>WRmiBSRFnffMMvdrrB3CgFuu-j;JZi(JuS+t0>
zgls{L6d%OiB!!@2l|-aWH%OSp;uqUniQY+anjtf)03jINl{vAv^+3I!N-S}yI{;5W
zP<%8|Byc#&*I`Ip6btS$Rc-b&0-Icx<X8Qs=(I<z3+ha>_f)%MQn(?-iM;a$9bZXu
zG@Z0JUh>xfFff^lfYk<e8P>hV5%|pAYETVqvi+Zex0+CLIb;egnASPEAYdU%=BMgN
z9d#$I>Y*LZ9>`3EZ^~-aJvkF}7=YOM-Fm>DB@;uQ&%*PJ@PMCafZWWTm-HlS&CFX5
zZ^aDX(<JXJ#G*^IXPx4d*JQY$l&KzJUR`<;#RE1d4h{M`5=RNN1p{(XOGF3{W{zZs
zQXDeTKMl^G#P61m+&@y$^I|2P>T_Ekh!jXqUjo;d`q@oeztA>am#O8GhyjQ>%@J9&
zCkG-1U&UMVJ})@koOb&a(!nJ1!6&13M1justrnj@u`>4>F`^8fh?znT9Zz-Ue9`{$
zhzqA(B?;7(*0jevWBB<LFqWD>6ZyfzgvJI70*5a`{(4TsS<#(Z2g}TeOI3l-KMI3`
z+UOua9jsvBdv#Rha|uYA^Z--Gs(PqLagPCx?tOx03uY{+^7%2Z34PC2?=2cjkg4*{
zEhHVMWt^!T*g*;@(T_viCuNH{h-nsiyYs>hD56Kqv$(GbX|D`hXS9ru+=M=VRF&QS
z!|_C1irwnCHWc6;1wg0IWHcZ{H1>vV{QD?Xn;->acm(!d==D<;1+XkMun$Kx#WKta
z1Kpa|6~XWK^O9~Qj0?>C7=-;q&Sppe!~2&Nzv6^NrFvo@a|%k8nHHPo_67p^<TRgq
zT>Zd6ZYuF<I?vCJ9>^mF?2N#XeQ=mQcfnbJcK0t|D2y>~Z9bAWCufh=VPzX-=5MSo
zk`(?m^>@}VOs!ASS^yAOttwuSb>jfjs)@?5TyHclp#Oc2QfQtqwQNqI(ZKST1*7b2
zv+S%Ob|y2Hj=pGFNBF{J3dd|9SvztiIep3LE(;Y^yEKt^CRGe1_6G2=63Q<}^yj)s
z<Kdz3LoJj8;1@}=j$!%9BWmA&Pzv4WWon^MlIXUrt1?}0c|iSD&`;GeuTC;AbTaEo
zCbB`^EF{E;(nWy706|%3MkZlZ&=Xc5)t8ZvE^<3oC#JzK{P+)%sH=iXMi#h-@^4hW
zk9?UO#UiTjU{o_|7K%0{8t2O46LezAJ44FiL~zC!^4lflIV5S#kyd8MSYd2@na|>p
z8m0#bXzB&VYyj)N;Ow5OxBTqNBDYb{;GH|jf^T|?W5juVzbo8TE?W2irD!E@ck~~@
z=afu{nq?zkOc8gG4z&k&Zy+PZcIti%%rw&oT4fN-77<$UEJPjXnKcdqMHuBV=}0!B
z#!SoX5!7@Enn^|huRcT#79R2LPi{p!KfFG?%UO)NN5E23wB}Cqz=T7$?~W5Pf<UBg
zY0B1SJeia7(ZyxGJb~)1lkHI;;sa!J@eTq1c92E#sR71jgJXS;IP!K60Z>08^QLEn
zQ1I!J|I8S<4ggZvw^H1Efq9#jwR?A2wqSq_57eJv&aIsaMeu3f<wz+Ut^8mkD+eF6
z5enY_!%CyxcmpE`IeA*`I#XZ4OWD!yro3(4!}qXe5bYHwapRc{t$)|@_wU~+Y#4(I
zQvQaH-EYak<GkCwfcd7_q%M$-dt3~wsbH!0qW6cdtRG;W0qg%Z{{6QqmHj`e)W)0%
z8E{~>G=eY?XHZsl&PM&&>MUUPH2g3S%e37HF!<JpcMwPrP*x6>|0)vx58XEhb24^=
z{kF%TKn4g)lZNz(epY0G)h?1$tYJeX5mckpuAt#CUilaB`(4gnwCT9gJGU1nOE$&{
zS=264`0fc#fA7q-LvBy_q~JuktP1+(O^_~lJ0)>KasL;+CjtNRms6u$ByUsc-0v^1
zqWf8-QvMvfFbm;a*{EJV<VwJ`E53JQQ4;@YhnwtqoW3A3KQX{UuXk(gwrj0|nALq}
zO=Sx|{K@p)1TvV{RER*7T3-MFDv)SuIf~6ra039ZZDEc`Bqh11Uc<3jpAePI`4n&A
zjE&?aZ~2fb5dNC=*K$e?yYHN0Pw{#MeKUIsfX;)EqhBVmS>TdD!UmM;ZQ-AtQg2~4
zpu}}FpqANZ)(*!sxl0IkY<a*BeqTR7kjw4Ti&kv+i@0884*G9ZG~^>yHo)EZ`SxQ5
z?MaMf@CT;<q4jO`R4&#Npt8);M;~5_eWkCBqPPl=WD=~*?U7vM{bQUUMm*iG>2OH1
zFBL@?H=!$;=b@BQvI~$&PWz<1ouH-3nN(&9j|{>(fFx3H1}S%bQ;<#yQJ!wq6zuW(
zHghW{mX>Z;MNsG}<c&Zw+}f$7N{jbzcwM2dq)A)K{4JAKf}}us2k7X1a}DxD_J8^l
zZrqkeMog`WuUeM%LLgQLN_?oII-+%Zs-++V2V#N_x(VjN;Q&~{4wY@BOx242LrXo!
zax>Oto1zF}4Nc7rF{G3CIIJn+?{!2Ya-0Eo<Gl;nDE8`H(u$TI;tBe2Iga_}z*#0*
z+*d_&GEe+vFf9Z^fz>+HdXrxqsa!%dHVKA9OH9-_#U8`~^FXU?v-GC@g5Kfs_(g{y
zJ<u#WmJx1@1`JqP7X5raw{N5?p%%eH*J9sFJ}Sq>o%zKf(Ku@VFpV3phE@<+EJ=<`
zoj@xb>2{Q!r!yPLQNS`J<<(y0d`l^}a6985Af4$yC;gok{gI`lmaJ$(s7fuE(%7nG
zB^!eV9O>21^g`+f%EHDk*s<6DWMVEH_APuma<WS@1q=}Wtxd|vWCTL8NZ`gJEZcMp
zR^*_J01rWY5>*7{alIUdmAKQ{0L3n+503o%1EHpia5ZyFvA)T4K06fAavvkm#V`!c
zSC}GJ^j(egI1MGwQ9&G!_ZLj0JoYnQl;<}&@$&~s62iM|I;6qP?lu}@BX=PoGr~!4
zvf{M0E(ic>4zs&xjbYgLnTL2KmI{|&U>+#|J|GJDjlVPewx^>`X>))}eeYmS06Q^1
z-af)dA2U}$`EUFP-djg%g}EaixhR~9s*8rURKY66`W|P)eFqSd!b5*sxP=J+jGeBc
zDXGhHn)vK$!7?dpPTIj?%mZ2=<TCb_Va2YOMkio<88K)Fkr#rZxuy9+yOo_uFd>jV
z{;2pA3p>Z(s~Kk%oU<Oz7nApWQ`|o{qj}J)W(#=&n}hYs)y<(3RjAGiF20Ne+64kX
zozzHeqFz4YcPd<~jh0wcQZz=3v_V+8A~}B;mI#m*V(kIJ&h3}pihF8*22X1^HXk{2
zO%njjYTog@svXm4YOcwin#2@y{VMZ{$Arm=@<A;(WGmN?Nu2M>``fn^<$<OlwEFqi
zB;lZQpAeUxfKBiD(Wf##BZ+nUAaQ6Gl;AtFqt%(%;kYG-;D%NusDmc=IWW-jRCxXe
za<i<|kmrjs!Czxp0$u~_!0D$ZlYzbY$_7CFq=<9SRjCneF?JW5BOhx~@OaJia#%;m
zl^5CPKTWMImFf1;ct~&qKdvLZdve7XlwA)+IBh_JRFo4+i=IMw&6B&Ay$Fzp<S1^G
z15WV@odM(Xq2@%cqDHOy{jAm<ge^s<+X44;errxGq(1=>RR(`dzc=Z0o-{|WNH0L4
z<As3I@e!G#&f-p05o$G{K+L2KrOUkOg%?XCrFiegPL2U_tP_99L?SHZDB}<7GyG`u
zWD3lhM>Ct+O;8-l6T$W`iBYl0{*3oISvO{OAT+;}4#6SzM)TeoRk#G&vQ!mguU*Dd
za$aD=yuDibC@F{wr!CY!=2#-IY#xBnmZ_Mk-RuWmJ{CPTpFY_qe~y}SBUYnoV5ga=
zhZ0#V<HbGL$22GSK<^x^-0;d05VCAm4QhPBRsSWI+ev;xU&dVz%SxwihEMCG2#I6s
zPWO$)>N(lIm?1>ibehrkuH{Je6~rC<Gc(}Tz2G-O)94@QXo6<{F!k=r${s*t>~s?o
zdgEIt4Nd_)w5v=C7mlN3t=6AX0t>kBjJ-zWSahg!+G7e2d%NDX)Ea$^(ET8Sbg;qd
z3T#?ch&+oTkdvbT)aY9Ek!RK#fA>oWi`d5(ka^!PL*MF}{1^Z2j{7I|<6ANboS2<#
z``$aTeGwvtJohHaHvoh$r3^s37zk-ke+9naQyIE9h5sdpP4w8Difh(VwVL3(^m_E$
zVB%~CjRAZo^hS5dhDzcqkj8cDTjk{aBFDCURZ08(=}H+Q)e_Ia{ef<#iDz_?&xjn_
z1-AA(Nx352G;xq0qLTKTMR4bKI1rY<@*5;4iaPH9YTf_y_#f^*D+g!mpc}A1NE)0i
zFjJbVC$JbWb8Ckuu*4TI^M5vV#{azI{9ik{hdW3YC?+S@|Ku_;GqEuJ7az1+^MCO{
zFEnoBsa4cSuz{cxS2vfR?WxTuwsV}0cAtO?8BtLAO%pTM#%mP>1S1!nFG*7-#6YlP
z#Q%Z&IVb~72m@bmoF0CjjUJvlSHYCU#cj}ec_A`jn2AUd#&`FSK5_YNUC`HmAGLIH
zvJ`kXZvCUj|BEa(v;5cwxP9$TKwv!FDn19<!HhO42*X-*?hP$;Y*o<u4iK#TUSjY7
zsx_VP>lBFve#?zm^E#XU#n~GvG7?lNu4VSS2kyEr34-+V{??-HNS<6NvDg<UaWzps
zg5A|u!``f|7tU;h(dm;wJ{EH4({lzCHhXH%JSBxTj^eZI0W~KK@UqkOAkieZF{N-&
zRCq-nr}p~Wr@QRLd)*pu{e1tu88&&UfcDa#T-{1j)h_PCxCu-e@?S1ui*W;=`}C5b
zNJ^MUFfqptpRM1PWIQbh7G*qrG~7IE93e`i_rn7NF=rokHeLoNd~QS|f+@a4J1b*N
z;Kx*0^RAM#DDhMfKs9K>YJmEO(F^8#fp%n7X>v8-$YFHVOK*@gs^JnA-L2+_2HhmS
z!dd-pP~tWfIRYUD&0``x<gB?j4386N^0;lQBpv=<9@~uqwf(rF*ALMs2`eeg%H2ZN
zIn81hzkcL>&lm?LFYY@04LmY_tagXYx~5B&Svn=~s_6<I@IcFeDIu+ZEA7SPO++5C
z@4Ej+8mx^vG?*X*b3Er2F9tz4&}d2wFQ|J-N0r2;#FJ4$$$<rTPdPc-_xj9Wn6G)t
z__L$S{teJu{>#vbU(L;?1jq4(h;q)M$i<PcOg;-e#FKuA-NY1(v2ToEbcU(xQ;H(E
zmv8Nq!Bk2BNPjki$prVs33X@zF`eWMKtYWdR|aUZl*ZmyYWRv14eS69Na6jmdypC{
zuB1xP)VM&Wfm9aAn#FAd-xUGQ9W&#f-a`YoP(;1;<0Lk$()XNicR67?RnGT{AA6;9
z0kIsLjJ24tq_Lo>Kx&H&htV<RBwRr6W=U|dv;B$y&{@=d@ao#CeJ{j}CRcLmZ<wec
zqy>RySyFXzYoSmBbIFj<6Zxm*3s~lvY=Pi;>o<RDbc$Hv;Qyd)Npu}7>U(=wSAsGa
zG8OlCeFE!zoSNm#3D!NfR~5AQZ>fL!xdiNZ@CuudgO%GKqRHV*CtjZ}936bX*Y()h
zc^HZV7NxMzd&<h6u0WD!{n}-pg$H;Z19;jHVDW2wU6Ga)4#mkgdj+Fcz8$9-Iq$F4
z>?CR1h<dgHi9RL5<&jOUe;=`Fy9#5{!4tslwI7_CRnk^*9Ak+=oqL~oh1SDh>}nEg
zW4m7+c@7wghmuaDb9>`QumBEF-f+&6db>#hP5ARsGBGr169`9^)WtI@FqiEpt$j{4
zw*7eE<&so%gD0r&N-6+z6lv_w8yg)p-pgzOhN1_5=BDqTL+H4)h&6Tz1=8vh-DtDy
zVon2MU}eL<L8W%?HZMTmWr^OTr!7!GY=`|l4wbNM8iIJb8ZYsR+7Vgnag}j)YEBP;
z<&UXBYV8qn>7>`4Fc~bVL5f8)wf>O40Vir?1Mjv{4jvcYG*jTN{@vzsybG(WhFU&P
za&GBp{XmK%C2Oc1Zkd<-N+P@E0A_`?cH=%ERV#dcce^Rq>3&uMAKl^{XbUX64VXQs
zfmm)L)2@D_O+&Ae-pl&e*`q#&1EUncRcsB;Qk#2t>$PvSCRsvH=JQJoB41tK3n(T5
zOUC65tVXtpv+&luA#5y&pnQu@C|mW&(AC>l!ComL5OY&s`Xg`8ME@!BFFP$S=gFV1
zjCn2`P`o!zyapR?poIB;mxQ2_aFm29Ue^)sUr}a}V@R-?i6I1rZI*lqOY^7zNQlL}
zy}sZ4DId<e;~3S0l=#0F9-3aNp#qpu))A1_83Mn1)sUD3qdh+Hmb-m@A07P%_DQ!{
z;?jaPq7?{8SK>iTw!KHD2Neb%_>6a8<B0lU^7L{_JaCmy?8ah7;>n>)pfSvRhi*+E
zfL7ge0BdtyE)sA-V;(7Ma(XcUqBE8qv<pdb=tmr7D{s^(N4tAI({J`vRFDDRxCab;
z5>b-uFgt;s4di`7;jh=-$OdPdai9)L^l?@bz3mPRYC`y>V^`V^E2wFFK@LY4biLCq
z+*S#(85eu4@@LN8wQD|BSp#6k%HaEg3E8emw>UeEM$!umYM(PAsLDh@4_p^x%J0ee
z7ew+ai^=Q85PZ}@^DC9U@&4bpciHJ{twBSy-h$a5w>MvKASKmz=c<WTTa|jU#BGZK
zlM3M^7QIUcY|-m{Y*eZ)V`Ae_oqFxswej##N)-pU3XRqdf92SMFa7NebwOJ_rk`Ui
z*ev;*riwc=RGCM@;JqyYa5&qDMlKxBb1bhqm46log_})dDy%&W+rr!j0si?jBz=20
z>pfl11d{aY_D#5`%e@eL6PUA{=0p2IwOc&&Xb<vXAMWK6h_$Ro9D8UKT&+yI*^O6(
zn?kaq`eie6CK~o{&VS!l-7kzpnZ&KTfUleb3SwMo*^naXq<HE8jt10*OMu71QN8K7
z&LPG2yHcvMRO_Rx;aZZ8v=oQCT_e#c!)qVXH!-zXGybSQTfw+>Z?=%9k=QebDZz6p
zYwDj$Fh}b>wCRAq==yEs;O0Y>1}B`JPmfec6y#m3b={8%&{4ML?|tGk8xZ;PYnfL3
z@s$mpM)t5`d*JUz6xOB8f4`<D_q1VszPFtp_**oNTdKj#0}V9$&C{kDQY$cH!r!BY
zg5Oksd*=@Rj)j#JDcs44X?0*cOY8^SG^*{1WJHZWc@?VSj|cbEol27SWK!1ezF!+X
zFqF9cg-#>%20;U6|IaJ|l$n|Jzlx4(bs2|4cDSAsjaPM?Q_2=nfznpG4y#tnxviY`
z5+6ldGjYw0%;v*ls`mq#$a~&IAq6r2Xh|i~;TzO5-Ze6>`v<x2^H%>KFJ)3Tg<x8(
zwM8kOP2Wx<^fWHSEQ`rHwKOlJe*vAztppGWpQaA{{X6?k;Pe_6{D95v&22!0n=+cM
zUMlYze`vnfur*PlUD{@uhdGA=x6ZhD@qU=C+9dbywCc!5ssp(Sisq;+A0D3W*54BQ
z>F9r+oBXCdbJoXiq<s657@Dr5c}!_G-Omf+YW}XS6V>oME@Yb7)b)2P@QoXKDkeF(
zo35o@)jCe95oIK!oBd2CCQky?12z5(R6@BL=E|1!dq^~3F-bU#SB;%%0J79}dgK2s
z&=;|fo#WGCr}skbVs;V@k3M$f%H8b*;5Kx@_^xAb;XcZXRkSPqiqq$X1vs`Z>w8Q@
z3sRl!FR@@@n16SV{J(K%8SDfdwO1j_%F3#e|J;wn-er5|yWVyvyea?yaX4!4wnzl5
zZ;x-WzaYcQ|1g?wb~_7bVi2eL_n$^TSgsZG_tk0FI3t8&C*YtcUoTz4^~3KM`%E}G
zAQTJX-oDWBkP4_FqNKUuqe{4lK?I52YdJ_Pp!l{ykCIpGwn1rA{Y|AtlJpJX3ux7T
zDq!sM31LfyuS0~!3ONO6{1i#@Q{Fs*K}VD3sX?Agc4*o-{xRA>C?)Iuj;g8D0Sv({
z7C5#E3PM9q97|89i}k0WWNmv!e+aMtTJgd2>q(?`{N)F&p_#tNsNxC{g2Kf(Ylt5L
zRJNRyaGq15R#vb%tuPT5nC)Y|6k$3ss`v3aF)?7>;hwah;^7Z)g)WVA7ao%-Pr^nB
zKqPGI7Ng9A)=~spX*&g-gcbuL#V@5`!gABc(~`La?%$2uVc+4>9)LZS#mNunLW(fD
z@Ho<m?hI?&(Gymh2j3<{{4&Jn;mhe7sacd?9(G4*N9P{A($k{YRT`syJjTu9s8kgE
zAv>4_gV8S~U;zdM-^fLAir8DLe!&if`sp2eb*%XkcC(0(dvY)M6d|Qu5ePd@M^?mc
z3D^J?@c&|F4{){uF<YO5KoP7AaJ?HIu^pnqIcJGO#S~VE6WOzIE!N|~I+rWR2_)MO
z@WI)lKii(g&2o3j{~cQ6MG&2Toi_Ez4<B<C4yf~9Jj4Z17N0VGay-OCX4jr6?JA*(
zXWQUXd1Tu3pK^*QzK0(|sd)-q|1Fo2iF<yQ*H_JvnO?TZJ|XIHA>Tau`X&Cs+*xqa
zQ21w8bI_B_$#(}Zo`k@@2O?XQb;y;D5Mj849rrX25emS(PrJ*ft@?AZOH&)`nj=B5
zez~z~)4vRec?B5P+NdwxGqb~C^C(oiuW)LySB(uTXWs(VCrg@_bHrnv9-914I5z@{
z5U<0o@dU2Ap1N+JO7*&|sSCfvdCK!5;^P(a<aO(q!cM1^(W-esM2OP3gq42hhK{MH
zrQ9R$>ggLmkub19Fwe*q3Sk6e%H=LSsn+15vywOI>*^C)!HP4mTruyL&zKvYZd~gs
zszqFaH)nNZ^cD8RRhvb~P1Qi5ysR(nCShvQbM1GbH~8Jq$6^=~%sh!bOG`=5Kp*Wd
z{R2WPxUTtswQ2vSOXEyS5QaicBd`TlOH!PL`*+mZnj8k=_iy6*@6h#MmuB(*j#{BW
zFd6@Mg2VqOz=4VBf3KDQ8nn&-3UEjlQ9V|@QUnu>7qw3$Zcl7Lk!hqSuq9~2y+ly)
zjeeW;ez<sG!#;!?-CUJ_6V0EqWbN?_oI7?xwd?7LsOfRIvDXJEnv;lHHl=2DhRIQL
zSdk8>o=t5TJ%~8|lWN58Xt;MqsUVm4+I#$o4*L4Okv*LRc+)f*R^1oa;*>TYC^j`;
zdATcG%<$4*EO6>NF1!CTn&ij^fg%?IeaL%RAa?OKeQw?C+&cS}PnKSW(eK#(qv=7?
z*P}XNjzZJ$BpbBEgX_7Wn7qVm@Wlb5T}er>4rm6h<ij~(uVZuX_;2jg*!`9<TE;sY
zj@DLn81{7lC|l}gcb`38U@scEeszd%n(+F`ozZ;9L!TwKof}v53Dln4>(hx7wE{!1
zU1d0)Z67_PHCZnn90EH~RpL1HOvGs9+oXSBiZtya-;cxc3a5Z=1PN8iH+Z?Gy^98h
zW^1e)d*6nR_3$nB87Emq#~Gtc_!})p%^k_Vm&Szw5(NQ60tc1_2aMx~_m|VpSF7lK
zt458FVs^qAm}9P0_!jv!ru(`gzz098HD(hc8+U}Y)cQUzSGiQ9zsz}CmI@F7zej1|
z!Eq_2#LAUjev1_)zD3qc53)+9rRi4Ks#225mMkzXN233n@@)!Nt3qM<wupIA+llZq
z+aikqs_5TNC7cysEzrdJKMdqA<)#X7#oOXKVV#&XrQE$5WKsJv<t@L=&}SJJjQ2f#
zoEC*i;%;fan-HtEaA3OA=7Nff6;EOK;0-<mW<f1TmDQ7Gp4cS>9X!aT3V~jU8<?@c
zZxg}#%<B;x(-^&*|Akjg6Y(3ZdO~^^KAJuNtT~zS!;R6QPzUI-<Q1jCgeTviGKvd)
zpO_IqFfGROf<;J03O?caW4N+T8h?{kS``n{s^SN;K4X&y_~*N#2gm&**t<(t7xo1k
zuX_RK4e_mb9E%8r5loKY)yPJ>m}8N{{1bG?5^9)zUSxIi@Fg~>aoLhsF!SB06=-h(
zKF%H>*J<P+$hPXcFfvR#+?cQQnxi02neG<|!modnNLf&=?bx{|28JNtUdX7|!m0C7
z#?4|Q)1|?<G$#;7qQZQ~l+e+ZI4T5(TxtY|oX&kmFsYchQE-_Ll=Re7D)U&+j@za<
z?S&?@2DSad#jvq6KEFTv^c6-{hne625aqwv+w;&12CjRTkE;Hj&<@4R==?bZ`CEU0
zsi)olqB}P;ydiWD*~+kTc;TR{J~|PX5YE8G-~-cqN#)JoNpPdF5#o1Cq*iFE7?xN(
zo2Gbj=Sce}OsXN@?P${sMzyM-stg~(#sv`MWZ=5YPA@6x?*dgw(7K5JhCNgaC^Omk
za548N7!T>JDz^ntRk^)H#&{6mf#VH*2u}o8R81C892TxxAaF`*y_ppw=@wax5uUa;
zEh$q%h=GHD1b8gc1J8<WE~**aR%tp^(>=ssu7jUIW*3aLgoD14UDL*EXS<Im*6lDq
z^gow5*YS@3hG;Or?GZKBs~>I!=*jEM0&<;8(lWr?WNr{UW9h5#CnX*&cuS|<;XPZp
zZy0^y)114a$hpNuwBC<`B#L~R<b+20=O`O<LLRJj;BxJjdrjT!(lE9YX40@*obO>S
zG=9;?dFi8(wD2>az2K3RfQkz3BM$m^!mNxl?-|){)}LNAM{=!kd&d?7o*6q<^Svx;
z>pRhOn$RA|yu6b%WO?Epl3xzca(Z(5UM`PYZ5-3&lwu$)EHkY5gCU^IV%PtQ(V<>+
z-FS)4pKYvmZ8+ITE->g+&;5#BZd7W~4&Uk}Nh?C`B<{pkkyq1~#kpvC;F+d7%6$L9
z&!O4R-40iqR0osGBNCYg0B92=WGoI&tkCPF5~+MynkhdGp8kq76W@?;UMz5BT7)yi
zYN-;^%Bp+q!7s4~hXi!sCNZ-lQ$t=-!%ip96OL+%|90-}f438Zy-0!Fk$Z<X<cay3
z1CE$NjSsa&&9ehFda3bmGTo<|T0n51afz6Te7{zrPyw~5b~z*mq%UwGQ>vgS)F!sc
zQnkBQcOtxQegr^FN(<Yj0n6eJG7)%!AsVNSct?tKUqQJRHcD)qMfbFd^!}<HyB+el
zjmxa5ZdcekAb&h9W$IqV)O}&37<mw}zIOa@ayEIPzORLQhC!0YLpsepxiYOEsCAkr
z{3=V$a1t?9O=Zai*bumSKc%W!!l(dH@jBGu8SrC9S>8{%jx4joPPFBqA+6@VWC_pt
zDcC1&wrA?)3e)^A<^y*=aw~LLG<@;8s6w0VA<pv)(51ZZD1y#Y1^rswHrg7*R!~)A
z%AnAW_oezC#5%xPa3TM6oQV0k2X@F-YK|AYxZFsCq+%%p;<MZhBNb2xmoOMawS|&X
zD*L^5j#PZDuMV(HQtkoii(g@PXb6~axLx+UKhm6@5xag?j9r_9k|}z9GWCP3!mC&5
z`gt6idLD5-Q{SIY@XgWx+%u2|9&%^e*dEt?AxR@Q>myJ2`cSQ_tKJ`pi!@-Sq)$)!
zJ5rR7+zJ?g_6<*&78AE!tF*N4r@?lELJ9B}nXY|d`>TNJQ;e|pFZiae;f437OiD5l
zy6-*}y#&%_gq#&^7lJV5u^lu{m+(`wt=n?mAJ;cKrTGZq>c|ALX#lLjJ+eo{>qOu-
z@#XzuL)bvEOy|exd3H*}5D#0*=bHsdf`lm6dUdR+g|4s8Nm6>DwA**}P%!n2S_W7<
z5ZU=Ax?`fVezxAC86X+fSG(@5T~CryOE8Gapo;W0sse@HDV}-IB!%8Sk#Ra2Qj@;z
zw&RO|$2(>Je-<ljOlim<P^e(c>};I>Y0uI$z(ElK)u9lA{&7=QSF4>_+7$K06}Xpt
zin!8)FkvjmgOND`d9|JWSjwgDB^9{N8NmM;2JgUEENw|w`*EuvejDiJb^GF&(Nquq
z<!bEc(IYcYp~NFk(9v^HCkVWFSOj~M)61>3%LacAdvGgk+&w?X04eTQ^RrLfby?bU
z3%mowxg<fDC9Uw$`65?w{IhywcGW(|7=CF!<|=DkK{=H_^lU=8>ug_d<V06)5ixkL
zxv~^Mz0Sh+@6`N6Zd5hHSI2~0=&d=toY>?$3_Q4nn(&{G_293!*oI5~fT+&1Qa(S`
z*@Qo^_Sc6kJ77mDBfVD>pp|uBaF_9HPYMHc+(ouLnA`1Mc!(g{trBongZh@FYqb-(
z;pqX~I=>B@#u%db``+?qR_K4ZGibpvn7=$|bKT=%Ng`>08@#Z}1bxk7eCVC6gqo|g
zi0&P2aORwAb<#%5f(^*tDg?5$f@74*$6;#gQI_b_&qdBNO6SUtGm`Bg&Tk;nNo4`d
zP04;kB8w&%BY(E7sTLn}`4P41^l(a9+M(Gex3?43eHW^NXPp<#D?#J!Gvf78>;8gV
z>B{bFy059K93cUy%Cq@k?KJWYX1DrV2e^U5byf3@ELw8|{~ks*yI7HMDZR6t_9b@@
za7_U9I{lpU(qq7Sq@?Ka!nK{)z*PrGpkf~C`x_jIMvGqXX0KKQS-Y=(#KYfAJ*}0o
zTvaR6pqooIL{>u^WmOt`JnwNduNxwN25X_!USqknHfV>yjFhWCou6NRDHF~~39BP#
zvAWi$>gcA5xs~(sd{dKNco1vHRQG2A2sAUzK$j5hyau#c4>%#-$~LxXxW)mcoM%t$
z4%ZXlT%^>|b}-9Aq*3@2L?F#d2v)zyr}2j!3yn}vt)ED-v2t1jiKX)aNB|IMyg!U{
zcQ+Lx!>szlFw^HT=Q*t#4I_YT(f*4+^?kUij%Rjh6t$3==WrzTF3~qygFDozE1Hd_
zb#YXv^1z7SCU%+a>;xR9qzwShdPK4l<ri9!=R_AP6!VdZ@tZK+Vn?m*J!R@f;6rJI
z9I9wqXw`Ng>S(5>c$IxiOBQr+YEo2JYIxW3)UK~BgYVGSVqG-T;0e)2M!X2~+;UiD
z9By)JLDoA_I>xl{<KIHG4hfa@d<#<kIw)S1uK6T&!jE=9L9DY{sT}}+QZ8rs2*YC5
zHubjM0cgO_Q{*apER+O^QBUifc6ob3Lw#-o@L(v_|HIZfMTZ@AX+LS4#<uM=wr!h@
zZ72T*jcq55ZQHif*tY$3=3Q&Pi<#SVZudH8<9YVa9Q2nL^;T!MkBq!E+~aSiKGNl2
zyNfW~_*iouvv`t})Ps14IEPLf+9^Y9?N7p0SSO`C-n(k~)<Kl~JSj7S;E67lfsz`=
zpVM8pR%*tyF*OR!&5r;V^P79PU=Z6;U6NcLlS~vNcq@wZeJOXoqrPimf{n3H=>YUk
z`C_f^L7(<&jvoiGmc38^2z@8T?Oky=0SYUO$PL%SRK9k;q-fNFY9j(U<&&|?XeJCJ
zHS@V^F)1{K$AWSwG#^YF%Lhl!=3vt2!wa}#y~Tgd3&4xC7DoXAbY@+-ITICjT^K{n
z!`&)Ymih2Up6)mq!?H9>CxsT>BIwVe)7Pdp>@L1)EAalU9o$yzpp+5b{7ll;5A(2v
zzv2i<qO6$HglBH5VrauPHIT|3w55iDlE7xZ^51XY0#CvBx^{Vfx}Zp&2Na!ua7&qO
z!Nw<?fvwxqDu4h;RJn{gO4{@@YaL~#Aq%lL70KFx5C&47{Ns(9LF*<MozdhBQuq`^
zq{rfrecOumCSD^v)upwOcVoN$9$UYO)(_PBXEP}@G`3GL?Ri{XH8xaT7v~qd6d6&J
zv6#Dd$<b1G8UFc+za{tfRIWrV7+fg-vmPj;MFhNSrp^PID_)ALpv8r^V=;xydB~Ik
zf9uLmwIfj#Z(A(o9~#-!TiP1=rI5BWd?k4{#b9<PL3zS`+x-fXlTn-zG0jhir(VCu
z4c}&g&~#Udg8H)lq9$uw;NuHWx-<`;QHgRskWXbCuE+8p;?HN_Ri(KP;~bB)0l{p;
zm>)(d4-E#;8fS@G>yAI>I+~?-MrH#yPZyMHm#kb4hYAx~Om*2W<~`X2a`n=UJoNCM
zA#zbuIVpEA=@*n^rJ*;I-LsX4il|x&bLB;d=w(HSnt4jjxvD4ahzlWTB4&xHvas(j
zzimXZ{b;2JjXpu#+uH*lG6_?*HoTh^w=_Ma6*dCy(M7W^jcZY+6j!kBTx?Zr`v>gb
zM=RLN_Y@^&d2uFnIS3~4yhztDWe7K@i7BnPM{X8`jfb%u<78H}L4+I5o0~=Gw^wDt
zrv(LfxBX2C73CZrUMmH~7qe8X&|@v^XZ<CTNJz;nj5%|MYAC1Re;P|h+fa-^c(y-Y
z+@u1?L%LxHnZ6m2t~!Sqj6ZISYWVr9KprKhM3bRsS9n>>qse@I2fitk48)F+2svgK
zAPQri(!*#SQ_=WIgc3wxUaw5>cYhL(iTM%9^`LNr<`(!P0%omoIrFv62c2|Na5P~Q
zUPS0S($F)Q0o+~N$T)Nnwq=v;*~#;qoDD$YsxQgl2k9Hr{xc=UnZY4R1`&L-d%Wt!
z$w}<ca5F~aD333sR1r3rZ|vyYEKf|K4hJ__TkT#1JX_zk2vj!I4VcW8N;_D=Qc^|)
z`M7P5u6_bIeU&$y;`?>lP=etq%(mjuEef_Xue1?CzufW>3J*;*706|a2@#zW^DB_^
z`c1_S{O+s!xZZ-0>mTwYUUHiI;)+`<=b0s+P^2y+$mB?*E_t-XrRowJ<1JId-Zk?P
z8obzeaSlt#zCT#e=LY1*w8VHv|BRBLV<zUYoI*J5gs%}-%?0-uJ)ZJ)LJY#ry(TFa
z%Tm!VHK(V#jie0NQK{T~pUfrI8Ao7vd*VE5{^|&&PJ~=FI`PhLl&R2D<ZsGENA=3N
zB1Vl6HOu%OYtB6>!JmXLv#l7c{Yg9qi}Y~=gE5#WGS_dmW7*cW+|d!cFt5xXOL&jq
zoq|ZZeosedqu(tdBo5Pq^fTLRFRi)0PLCs^+h};j8^P-N^_j96Aqg1q3W$Lvw_7*a
zi;rhFS=xC~5$lYO>COkEhx>3@>+91;13EgoUUjGeNP{TD2ZD9m;c3FZlZ!&Qyad0q
ze>yx1$-0Jl=By))a;rBSgqRb^W~W<v`B*hchdX|YztW3U$c>4unDL8C`o_C;p)+T9
zAv70xjY%eUakQclbM%@igL(@1I{cg1SP{G-{oEIcysEbl_g#YEa~@o8VChTcoqRTu
zfaL#j`>}QCcb%l${opP0SdoW#F!v>DSxc!>*f@{*EiECLF<HM>yt%2E^}xpf0V}<+
z4!v|XQCYU0-xog+WOWa9s}Iyv1O4JNNQhczF1^0K2qmOO17)37pW+tKixFQp<tl9>
z%KsxwrBl{n;toU_n68G3NO2E=xUC~*dCV5~0<j|*QG}6<J%9boYBz4jk-QC6{>8dN
z(XC4{VJOOrAnW0#!{^0BD(eC}X#f4FP$9q4WOq$xEB_lKVXEcl|B;OTVES*hLPAkL
z0vHDi+ke<m8lz6@>`0www1*h-Ns<ZMml+*Y2bnf8PIb<|Bq!iV?1&{-$Ci)Sa5Cxl
z5>?9PK>nqp&C45b8trN=uMZR*&r#O`x?<(J70PjtMw%66kB8M6L^KI?4Iz{=-emDi
z`#OJge!ZWHWK?WEaO7M!kK7Y42*kY}?i?=q>Q?0eR8;e-6$DN9Hp(VKi4^4_UJrxq
z%k7@3)&kM6jGmp337Q@mlTZ|qWpjb*8`}GuCaI5Xvi;(=l<(rA)O@RpKCBN3HmOBj
zh}MM7sv2h}*B~4C4|etk^xt6ZFRj>aJFN&Lyb$|RXC#e3-EJS9a;nib^YNmR9)|nv
zuLB4Ha-nz=n++Dnve^km{em!#3v_y8_BbkB+Al44o_<QgizIn=RC@i?Gv9ojjZvv5
zeCG`{)SO7GwHW$psZQ-^g#vy{i3YKD9UB%ZZxYq!eqMYN3jOwIUbYMGbJ%Kxs?|He
z*I|-7E6d05bp*$#OFXsywWQ3%or?wsdv<qVg2p=a&m`;as(CrZe9DNwI}Bv943iTs
z9Fi_mg;hw9pKH!xITd2iWGQ_qUWyR*aD`LTW4>g;CXa2|DpnmvGvho;iYP`XiQf4F
zS~Twy<OxdHdm84ZejC_BNAR=>OaErNltDk$PUzj|MDI%TRC}uiZt|m?g7<~|sE#^N
zzb{+Q84eaVI5T6lbZT{8Y3Mo`%?me1VvyRLZ?niTS3;OW1qQ2LSMH0eC%_OdSa8HF
zPSdvrgH)4(TFpNh&-@2>FmW10Z|nBePY;?mH}W?XG~Z^ZEmEyz(m9b64@XDJNL{+`
z@Sh}}gG=HB%eeLdP}94|%^a_<i(z=+s*)adWyN58N(G=~QbQ0Cd4U$Lc^s#D+U&R<
z1+iUxwvj9B^o_+(tlv>vR5|HUJAr09sPxN>xqYMmB4`YVQK=B{?Gl)uHdUi-WYKb8
z+IOgBz0a<Q#N5QH>yw@7H0on|l);hLZf(Ek{mAAV{P!@8ig8OIl`9cc688a~47d4v
z9cygzIi=l)JJJZMa$xd&eR?PZQcZ$X@&S?T816#*vCV0}b~hS|-ES9gOKiM8aK5@`
z1=#)y2P6!$g@27BgG;O`r9n87`JsS+_FaBQ7S-~r(9^gkk0+W;$oZWD-M8DrI})z@
z)gZ5esP+CGJ2gqaGIG<kgWwTpJCPGO*V`b+d7`Deme1GL*wNiY*^u6NYEzac>6_Ne
zD?*W*5^h1moIyrz)C``~(#{+{OI16@jyFuv??G@$A=e~6x=53A#zp78#ANrvZeep)
ziz67i-aI76diC5N)DkuZ!ez>#>#heaoMo1NPZ7E(a7?nTms};(@h$)=UYA&F*kQ*V
z{<ddMTCEJq4;MrdwqD;^PkGTDpUhSXWCA^Xm*j2!{^oL`MBJ@2gwM04KT5@DQSN(S
zp!d6CMb|p(M$ya7ryFjmNkX^OVfbCe>jIv+?BlfXru_=T+Efvm9HMV9endhAU{Goi
zn7jz3%WVbeE)%oQ*WLgWrS4BRV1#{9`yzG($lMfiKE>FSwjUFDcgB$)p1s7d!I{Is
z9iq5*0hBt|8XbVk9`4USgR1?5I1&r2Z=T$|f$?Iy3Woaz<0N}D7N+?swMH@8IHoiB
z@yab#3TAmA?x_V<9~`;<5;lhcRosEM=FgYPdbU#ajcMY{tE&K2VCzy7;x7Lw;ZN)Z
zCI9rYF};n_@3vWl<~*uK*Zi)!TlYN6Xk+m=K6B#q6fX#e<$P?0+9H$DjJo9<1H`kc
zv6)T2fl)(}DVMAZF+3a_03>JEwa;HLUdF3-d}>d~Y{t!tm@0iND&aW&%|f9EJj+J}
zqj8BWd0{At8vxk$;85@#7tv1teH1G*KNq;@pc@$aZpa?#)WAj@U2=EF`b4m8myH8E
zAaIM&N#xkM4a+`GH4-041T*-Fnt2oF&f~u5cr942URT)a?Qe_+rIniHy@Gyzo%i|I
z|4(jMEK02bzrCOVKY|7Qs&zUIAJaui74Iqx^T#q$906d()kt|Z`};08qQ5b%iu{8*
z+U6D39}(-PHF)xp&-*7^duH<<l@gB5iX}vnxE^wzGcu1Ts&6?N05xZJQrx7#wFCAi
zHyu~XH{e{(AqSVj%m=N^8up5HoM>>aFDmZDNS<730X$4p1v^PoI{_E>L9bnbH*&gJ
z)qmS^IM=MSjhMqW#F{a2ur+N*!sKtx{@y`CAc8!}E+=A|Z8KU9t($|MGkp^w#Rfyp
z&%h5HU1{8f(gg_N{5t4x;hAAkq@wR_2`oiVh#X%aDFs$N|HsG2%$%}?3FZdM7NU*?
z)~t>NPUQGcrvLBAviYAZ-yJ3e^c@NnoP~+&$AA5aG=NOPf8i9T|C{TL+|N<Ej#8bl
zuw<67QuQ}50W4%dl1ioETl`}r@ZtCIOH_`t{ZUE~%z}NUx>{$}tUHWdU_*4;5Ablm
zyNEflLml2>m`<4dHx*I~vQ)JjD}l|Y+v>39xC&(L__*5;m@e-6U}j8_*oEtdmFal7
zIOhj8iyRHJ?V6<sAaXp6pF-AHVJQBrW``!MwtLlyD6Hp<Bb%#fVuT&NW}S~#Jti8V
zoB9pUvfI@R&-13S`H5j;n4WV0EW)uZuwFi2d$!6D;)i=a+(xR9ff+%!OzT7bcjYZO
z7%x_iVdb+yRl!@9XuB>B{ctEID>eya-xGj_b(<S;E%lM%;jpR?2{iMf&W$2YmZf18
z^S90k*%*ZY9lVKqTNYSP?uCTN=j2-_UI7F;XKU^aFQ-7={+3HA$*Ig9m<lOX5_C}8
z(tsUS_veWKJ?mlm2tvJaJ}z&8`6lL$4Gd584jxS(rik6#QiX@ki*r%?7qW)C^b7!3
z$DP%m4Ugzs!VUDtvtj0*_k8*VA*8#iki~fCx)Dj*LFIu%!^}Vn#R(-3zC%-}BW8!Q
z6`o#)@g3OB?Rv*xq}x`E`+Y53Ca&?saDFa<>9h;e50AIgVf#e(8GdF0_en>OScAAU
zg651B@$4W!tixHn;41uT{0K&&e}d@Wt0qe1DWhFOZ+<z}H}q?1wsH)^pBg+PIq0#6
zd$w`lY$m6NDZY#f{XaJKeC#u>ZnvS9tc8}A)AF_BX{w<=^AK4IS%go59TSx*yWhjL
zL?10<?&&V24r6JiZ=~y2HkQ6S6sp?)X(!>tSuu8Q#zEa>qU~a(N`w)px3dE(StbV%
z0__)UzlUJ2zPZuxzz^`AR<V3+GRF6vxWD({V0JbO2XfHkOI)TD8mUn%KcU)7k84&x
z!B6K_+?-IMX{k+L8mU^Thz?2|nx#ukyFmNfx*;iR+-wKdCw$o4qq@!;`p#_}{rXmT
zZ;db*M=4#+K$=+Eg#UsN56J~^068O7cJvn{!;}H*hKW+<2Y*Z+#ru|Xx1I+V#@^^7
z$8LCxZ~wOTj$X%5#caOcV)S(lF?FG6sw`bIDOE@&ka+wG`kJ41b0Yh~5%6Zq{K*v3
zL<0{;A~&l!&!KPzQweux5jWE$7B6(Nk0OPZX~kZFcn<9?GB%A86h9A`#3<^}wpaN~
zOGwWc!9QiqvV?wa#@uIUwAmOgqLeAuu9Hm$R=IhTl09NeQ9Y=R{~)J5RKVrOUm5na
zw_P>S#9PZ*<yMIS8z(!;Fp8_Q0JAftiZW3m=*XiBF|18@LBjqa3c02{p<FL=degm|
zjKw>lM@brFfgSstpEd$;eh;ebIyqF2r^5h%{B|;?Hy6(`J$My1`PFF`%7g1LW4^&3
z+PzUt*1=F8vNm;UQcNMhQI}~jR#DtL0lfgu?i^K+&{MR^MsuSI0u+uVi9hZ7Y#_PE
z1ELe+E2IfJ66q&+yyty$?Q&7wZ3m)a;_vjX0ip0}$5;aea0<XMb<HtVPeffmIsX1;
zh>=?=J4CAXK8_Qni?Tg>2cz@d%ckBC{<J`;Ku7MoW83#}lcxjEKa$gi$VQ3IACyUZ
z{)X6ExN<htUCAhYK8IHM(tk^eNuEU7e_ZEA`5USjxW&rMU7;M@jM^zMzdIB-WDpbg
zPsaqBZIhS<74rh1EmbpEg+AYsF46~`&XK%@8*C08aiV{)$3W+i9?++!^z_MNOg*9O
zm?mwTUR<Li=7z4|&PO|i<~89cHiU6RSA;^^8`@t5ZrN#hHhFq&T-n-JUEKIOJnbr(
zWs6FEzC#}bD=x?+r~jrV2v}&BbE~Pt;~wt|kTl+aA#w*Y_j?tJLKIvd>xH#bnTuz2
z6D7b5PnOtGyP=9m`A{=s*(rO=OeIv|f*h4zO%$fI&#3|)e+aMt?Rveal)X-MmLP|a
z!|I&<)ptz(>6{Aspud=ztkgqu<iI>aqQ)ALC=?o@MPkbb)6y#|^F0zgusRP&#!~RT
z3fA7mh9CzPWE>#ZJf0bUktxZFZt?aobU+MfDG+NqDn_0sUS;PKNn@iKa*<oWQ7WID
z`Rl`-Yx2MZHmvdDEU(Oti2rE3Qc&K)@<Ary@3Ika`FvyBLtdJabf_$CBe}fQZU~Zf
zR7^n)K<v;RxsRG<?dT}>$};5BMXiG`n;!n>b|nPXwbii6bMy+x;j@^6(8>hOzaxV)
zDZ>i+Y$!l|^dWPjJQacPLeDOXHKXPE%4w>n-&z*$>S}O|FfpnH3A!W>>=4H*P)kR+
zs2kX#m1!zudg4bX%zQcPAot-b3CjiZfr<*Jy?glxkIhXKJDXY%Aci_7U`(d94dv54
zf{6j`^m#a91?at5C!BTBP~OJ(u<vfA@8GWA6)}2+o_iN_jCLdCm=AE|k=qPWYgPDZ
zjYwMYFl3cq<73Zyt@k&|(E|Qxh&K7$!Lgkr75n#a%eluZ(O7|v;NeV7=9?mkzoq{%
zFmytyl13#~f(4C<X=3|y_G-y5Mw6z^m6ZmpeyRQNuY>k0rSudEmZjBbc@jQihUqsu
zI*pUVo2obFYSj?T22T%a-8Q=ey|_9|7?!!~O%^Ie?17c`Zz3?^{41g5OWeT3WAAS5
zifqlj<w@<X@u(*Xn$)U1ZMe$`1H^`Puxus5vT@1z(;`GDn-Pw-T7VoG|69??S)mL_
z*{t!(`Ob#&DIO4!s3<#*{Av85SBBldHuK7Bv!jbZY4qJ{<vRDe&`&Ef^B3O{MFl~W
zc;sgE@D=nQvhTn9OZ6sOrMB2G=a2A$!|Xw@4%$lwd*47Mj}9`?h}q#f(L&NT#dQ69
zh(ggbB`|D(SbZeAn6!D&Zw!~l&O}+jOLU0#RxtL(6uMswiL_?~{0uc(`j57h3Fnf*
zln#qbl>GNT0uTB72B+r2B~~b2t@*|y#mf(WN{685U1;3|%S=pznT6b{#c|nmN7lN8
zX>@ZsCE{@Tx-qd5dy2#q&N_;w1!n3_8c=z}c)?j|JNB-2DLqluI|0O$&+J#AKdZ@5
zoQwKOjhuka-vv~L$e;&BUE5idH6No~i5;k4qxtyaldn3Rk7WK%a>})hQi;DXYMDOH
zgX}A?EZ=T81bs4R{%2JUoT88XeUDlC`E>edDfjvjR?AVFGqiQV7yI-x`gUgGVZBKO
zQ4kNzk&&mW4)IV!KV^^r%2*W8C@`q?kKAqK@g{TX;bMOd9OA~-)<r(~nM~HdC`is}
z{Lu@e>w$-?q>y%IY&V=&vNM-vZS3!r=7ore=Wv{SjcSyp6ng4~i8sObQ@a&hEWLJl
zZd;nNV-ATTa7i(*uT&N3Z#PNTjXFKOix6qc)hfrH(`+Dyz(Cug`r9o)2$3hlsJ&#x
zKgfuO3i})K{wiRvFV|G99zihnt67jw>ZNb$eGA7bmyy)rvODg6;cy2l1-$$7dsXsz
zc)2VyY*=gx`)Q@UXl&XHPP2CjOc_%)pTH`zvvg!456YIV>0?*bU-^aD^q|w(zitrn
zZ_>De@^8GERL*6xc-A`ra;>#-SZGCRY_oC`+YW!n{OM)aW$@;)4ew&Wmc4#=`^y={
zQarJoBFkl&T&6fNY>}LEhv%1hgJmuB9<&Ob?{4L=9L^#0!&}?rTE^D$r?pH1jjXX5
zy0u<QHFaw_G*tV)cj=0J)-^6Cf-P}p9bsWSbL6GfQ!LB)<%t?#4E<UZp+ALyW#($F
z3tFUGb3{zAxx~}32>dipxai10oqfvSiZ_C3HRsPplbVBk87J0&2KQc5M>f6@_@sZ^
zaU;e1ijP<((cJ(Hy;mXh_~d_-wK5er9$c6`8JmWZBW_)i9lS2pYaU;sjNA)~{Yt?w
zTxz)4eC)D?A!!C=-0zWGsHgvwW4d*tkPYR7@O~|$YTSYI7vd{)X6sA#xi7#+gb%w%
zn<OBQjaSON&J|MrA(Hi$8OY<v<H~zD5?G04SPxmG@<rs*rruqxF3^e^`;-q2MNg?h
z<T^K~H#LODX-;l*42rGPGqxig-Dr7@Y&Gh&C9~?Z81CUcLboJvV7sfBtf3N@rWdcG
zc>(9gFh77CLrhSHKhvs)M))+aqw|XVKS~YO|EVQd|EHE1LhkzK0iv)=l0;?XPfWPb
ztlE!tYVfXi?g=2h)FcwzBfe<6-MjPM-~*8~QW-VnnrAsqJ2_hJ%owz+)p~z7``InP
z+wpQK;-o1$`x<L{v%52I3}~9K8r80yFK4jo@Zw>1GZ1F$o?WVIq0j?sqL!}ri=~U6
zdp-|9sU57&Wwr?tWB0o0>JLWm{N<d>&xh-$EB*ov^T7<4tP4j}u7mT4qQ@hdEv^k8
z-`Wei&J8i6I~^W@tM=DS_sf#``m^lQcC`*2f(HUI->!8doxQa-lC?paA=IZ?WAwr^
z)(x6=1vx)AgR5)34Ie+C`LF40-M4&~9C{s~qsaxYc=^56qhh+N@1?Tp$U}JTR9Qb+
z+D30#$@t2l5X+n9O+9E3{-N;hs6*ANS7$rc;^ez|NK2Bgue0w-gzkWv1YT-Ei&tUy
zaL2e~k-YX-aeuEA#+he<5>=u1BYun*v8@dJ=---?xt4s!wl*zSPiNS39F4=Q>hecm
z)F865BgEPuhX{W;(#Py_y(j)wd@FQ#aJ9}aNVD$Xs_OVXrQM@@KYx#Y>dAGFN4Hx|
zSgHTogZ)8^hMqCJYcv8gYcZ;hcdu_I{DSMX$@=~0v_evH7&f6EHd#$fiZvCVR0^TJ
zs4W4L{$I94c7}ESMc>gB^EK0xE{9ow?Pc_&EF4)TMYrpX?$g1^UWZ%-!YuK})k4|n
zT#K`O8;^m{-uIQc2$bk5=gG7srB#@Ir5GzjTMo+ScPGnJUTA)QU4A89ykGt)#1(@!
zW!H&{f6f{x@!49bIg0}B#+EjUI~Xl@hpLwhDl&ZI>OGM$DfAnX+r@ccxHT{VKcp+k
z$9&>bG-P41MHf%Cm2A9pUrXQe`Yra=!?*Emg3q)NqGMZ(**v5Dz3shedyMX}M?xE|
zpCr*kPjOCGvbCO;70w^j(pW(VIlK!5blO|AxYKM0y+#hd7uQ+-321_g-;ufLvq}KO
zy)j((h1yazSvBEAXJgi0joJg?a~P3Z(O_UMrhC0Q=^iv&5j{J3`h-X>INv0rS1@A=
z(>IO7$~d;>7+xuLr}IL5ip9+J-r#+9v{WgQ(P$aPE-$i3XNe?^g#m7uqL_;K&4O#S
z*nQ~RI%#Zp0BFr}1eHCThRwW&1o3dg&?DTUU}86m1$cWuMVNI54r&AtRHbzV?Fb&X
zjepRo<@bP(M?QosL_ibCxptb=%F1uRajVr-SV3F7@4SJf<SWi!@iwf?OM$>3drR2K
zEq5f7pBS&u0<+VY9w8US6Oqd1(U-4IE&8mX_%Gl+Z?AVfGUVapDk2M-5O03f>z1t6
z8Tl0cbj)KMJcX$)ICBOfCFxQUOHLDu!h|}12aDqDT5I5t!Z_qAA=a->B!aC*VA&Zv
z<6IykA8+dpL%^E5r}~KN-7VX!+x=vP$H_-!lYV-v1j*Q@BJ(**y|HWe8M}b(#eW*t
z)?lD%j_&ok^!4xvW0T7EnPK<0X1A3{P(UzS9DP_5s+ow~s>T4qERxqbYb~^jJ)5{i
zFGUgVO}7<KFx?;{RW5!b-ji2<99K{K46Y5k;1Z&sKM{L;NYniiim*P-$Zy0)Nn-V5
z%U;~oOKtMa_wus*3i#IyF$Rasks3Ofno}-qgGcCE!gh=7wY_`{zhbq#LjIXmFZ8wI
z)_X-aAK>G1os$pLZ?vE};juMG<sWpC;`pD=$psLBz#aRCl7WtGR@u{>8Ur8=71PpF
zG2sSeXyc&u(M&gMWNWfpWOq#hS#&h^Sc)cS%*>;g9hQ8`G1YH1TdVrDr0-?;Ne5%G
zX5}X)&h+vqcEXTSV*I#)aGD+*B%GrA>C3V|;rj}xb|C=bt2R8(lUlet;!p9&?%3Z!
zqzpnc9h%NX6eIg_5*_PI;4NH2Vi((YWPMEIUFc9dW}q<jB$)fL=im!*ZyIXk8@do3
zGLxQk%B&ZHwwbflXuk_dx_jVsxAzihENQ-5N)yCu9`D|Y#wK<`hq{2Vz}Ic3)tK`=
zMFK<z+e83W#3<jU4j;x<UMW@TI;*msVZc0H?VI=kS)-(Qp_6~2r}<%r+jhkJ&lkRC
z+fdmir9{Uj#VDD{K{TnaWhPaW2|unLH<8R>TD2Omb+kji-^P(u-^dU9GrpByKySYI
zZcqbY+S2gUNy3cqnLu)>@zR>M8^YSh7P_s|FBsr41$3K6EN+@OgTzQ49p(xGhEx|8
zOCAAkGZTkEy^om*t5bAcbtST>x71r5Q|qGMHFH8ebA&;tFmgrtv}|j|NBh9Cx1ly3
za;MYRk|!kHx5R2B*H+K^t!oaeQg+oRP{*X%1AQqP{yJr)m%<x<$8=c(5h2LZc-4S~
z)D~bAKYc3#-}MsYy-a)Io<ac=4+G5z5<5Bq;Rs?zbZR<+wo7y<m)KAK_9mYaLMd#!
zd*g)o?0AEtk4{&jrfcWcO==7gNSPHWrW1C*0I_~!)hL#!mQ(P{9AwE&$hwi!whnKj
zBks7pzRX%V2MQax@7;1sJ5+oz<xRSqR2i6ac_lbOv;R06MzoLmZsm$0U%e!=b;?X$
zR)MB;VnmcnqP>HbD8$`N5P4~Zr<x(uR>|;JW3&Kc?sveI7&W9ORK5fe-T8MZi}^bW
zW^;nK2huK6WOb40BVbEvlcLZhMl58wtqd2}ic%tgtG@3<Dwj>YtWEomUo1`Hoh4vU
z)c9eQ*WQV}2Qq!MzQ_H<va@bdPdmbSNDiCbjjBlLsVcaUor@4ky@sDm*bM_?0$cBO
zQ+B!bE8XOTdlm0<_;Q;;U*d$pCRq{ev$278AlXpojCE~u-}h{yb`k?#%Kzt6MJ>8`
zUQosq;f$reFG2T3ffkmPvLn(&Lol%aI+L&`TIgV{u359R?U016(q_p90kvmn!&WF#
zPTgh6Hpo#);l&(&5lp`^7fz3C-=wp>LiUdn=NLw&XV0}m-gLHahc4y_**Mprdh!!h
z++3nxjIPHqlM(Lh-eE2H`6?R!=ci|y3iVXvpT!HiIYg4&uI{f4_z8+IVflaNb^v*2
zn?!^SY*zu!Urq9h$VZ{l38X@n<zjr4gjs2^uEu3(L6AsXgAJ3YqZxky$ysreO&P8j
z_X5xM5hH~wAOx902@}G(PLy&c4xf?R%On7+i?7FyXt>N3xVI}&cj;f1olgS86?ytA
zI;5)@lxcm6=?m>Xzvjij6bh{DeYjN|(?$C4U!-J_KnIMGJ=fDXOvkHMim>;?Qa`3v
zI$9?rZ2ktr3i!sb<Fo5%Yv@i~hm{MZrI)VgJf3{AXNLA8J)1Q-w0n!?*HtoP;WKf2
zdGZ9gV&MdqV*X0nRO3IYz<qxus)gsVYT->|oFTt$XSkF<d!-n&?K-fMo@K;yhted7
z*2L^48W#}W82~%Q#wNyGaU9xF#Y_#$Okr%q<`%`89(%fPA{V7p5QeZ%@nrTVbzKC_
zMOe$hc4Pbt;LOiAQaalgDWK0tTSNR3H&%fT;dhVL6H;P53?48LmHg`;zIe4~Yf@R@
zan9)M<mS0pwOwcBvj#}^+W9S+f>2oY`6Cc=y;$}Ko=I?g*uzpA@2`Arv3MW<LG5vC
z9vP?k9=siI$N<7-upw@DvfQC*`-4?f6MkB-sni^$HfBk=C8=6{z7-uk5a!zy(x#ZQ
zxb-hy6JoJYSLeSRQAV9=Y1a8Nt3kpfob1CjoR7fYWoF#@TsHtv)L)@@VZU*F>}8J=
zW}=5P3A}ojB4sYdp+h(UUkHW>wummkoT5)+3I8ytu4mCPUm5T?+|d9kTzs>m2(`m6
zgi-o(OT3r#cZ)$W@FA;_?wx;KcvVih{h1Bv)r2~EpFN7_Mvj#rO#DsV)aMRytFvdV
zBL;i-w_|zJ)d>hS_&Un?z30Layr=qUN~YAlSemGTm#CcJ7?ivADkZ|UXr=4dNSpDa
z76e^N(bEscdhQD0OMSm+PFwU~a@JbTQG-5`3*vrBMGC6ZZ%GR|@8q%RD_68MsvvL2
z=GMfIms7;bwLr|PNjh$gtd+}2vk?>%;fq*i5m1#{ETCbMXL(Ub)X7=8YVbB|e0c6I
z;@tdh2O+FI+%CC=WnazyaesX%KwG>7lcM?#PLx2;4-Ll7#Py%yyrxd|8YhC^RrSak
zr(nVj+vADO)0{CT515VjFD*keLNpyp<kJ;`(aV!GgKk2xF~@Js^S)Aa;qj8;?~*i|
znd~=*ALW{#U$@Vv!|tp76jG0fVsUl^Iqarw)O%KhM}vp8tM53A7YnJE4I71~$nb6%
zqKA3s-(xD>ie~jH0nZX32M%k<R$H90dr*u-*NWF#X3LW`TWxrumQZaxCyT)IF-`r^
zjkkNKg@$C2ZNhQANR2GylrjG@K(--2vx<AMi*v8(%lDnKVU7HSe;To3aK+wT>k*0V
zbuOm@XR637AR{{V+2qaALDql0mir`FB4<Ni9j7?jmPkhr$ml(e9>Bpbm%ewHiW3dL
z=ZxPJEs?6PuPcKbegE_O<6MB|{Apor9I?D^Z+|v@k2SrEG{W?Rc&>X;s^|=PZZ^DN
z9doe%qON&OPv5lV2cD&scB+OY>o$W&fM_AGZnWk1czADu62{MO|9xcvN9^dw=twtr
zRt0~rofVoHa3S6@SdzuG(Byt=%h1nyOF?#eDX#8V7eO%H$)P1uJ{K}F(w%W`{v*q8
z=Y~Z~H)B36lyp8flS>-_XDFYTKMwT3o2(o5<OiU9<gYY;#-e}Gn;7;faa{<Mm!#lB
zuFVq|Oqe#isRZ*8b~p|`MvY7U>l)3X3k#}~PsG3i6gGb>P&yBtvqu?k0JG$_D~IXd
zE9gPL={>L7vK_jecI&ZCcY3&+f{VwFEl*EtD*7s`!UfVPpTkMDcEMhLt0&tQ(L)T2
zz^S;u@3?$_+OvN&_E}Gjg{GT@j}Z+{XjH6at~U>KPzw`Bi{#Xckp23+o*MoeejQ3k
zNBt8PnA~J<SFf+t`3C`k&g?)B-S7R)Pm~cGKxs+9kQFmTkZpG4T-SC4{M(JogX4GF
zQO+CEZ^rl>eiCSHV(N)%5G`f5n}G!R$M5|a@w26ulw$E?LGOSru<hX7gLlqgMGIPa
zc=WV~v_31tGHch{cwXJ)FS=X>!HDy-g_~y$;Fi8kB!dFvHz06PZIa>2goX0+cL0qq
zY+-(@>QZFHW7#H@ecal3Rd9q?j;4|K&J6vl)!l8i28lzKe%(RqFu8H&j`tqy(e6)1
z6Rj3U;$&c4zJl-NFY$jO;Ub^4HT{wyGz<)Iq?xLLZz@W#Dpz2IoIz+S?3eHZUYwGE
zphpTFbQi^X)-YUDjPurJn-_6_^<pr7%1b}<cP7VuDL~83jbEVH(+WGi9CO1f#KD}!
zqeeb%*tI`4&5;)+P6Zb=t93>SwYng8>I6@LH$8f}2y|pF){=A5&^XO(PIe&|X5gY!
zCqWqrn%WsYB#a?t!O`bW)60O=!>~Ib*4nGC<#+uXnSJW^!ymHY#wceGCB7l_)081t
zXjMWRnWPfrG&qz7`d2J(9k?gDz-s{Xs4}cPuwV-xoImH1gD1^@YX290h({1lIyhA(
zZK`}P0!2}^pG(L1ssLuJe&EQ3w!ET?N_y{pqi$PuuJ&AN;w6cO$;=;7@wYkPTSxY-
zpU$_&jy8G&-2FTxf@m(IT*$s*BHmXSqh9DKEcL!Rq0uBas2A=xB?<iDYXjeQm4fk{
zo6)mC^Dd+Lyw%~j{d1iSAc<&rr2G#OezPiEZL3Z3_8bd-n0`R!+^qVKt>idQw7QIP
zTcm~%DRsN;cw#HIKTK?^E|~$Ktf_wS7l2Q-Op?W0VR}b(p3f>7Ri%}r51kjZh4196
zdFJSGb{8e)L#Q`C6z5+moxeZ}99qOZ^6<yr-u$|Qcq&10knK7o{VUR1)0!lQv>;O6
zef>*y280>ef9wO>5ZO+D6-BJY>JY67Wo%eX{TLKnU-<b=7_@uQA1@Us#;Jt*3{XZQ
zIX<A((PGPs_NPpM6qDDgw@ux%teb28HMJ7~aVHi)l{SykL+v8;>7p7omba@vc_+v(
zlO0dq8l)U=U@Ha&D%n?}2v}#gg6Wd@vZRUI>lCKOywbjETzr0wrM)yen+0T2_rd$X
z<aEqrGJOX{wO8DQk(2=@WxCQ)bLgBvGyT(LQ};<~$){2kx$?NQ8q%qov%m)7;sdUP
zKDN_!Mj+DIby9iT(Rbx!jkma#y>p%jxKkD_1iF<6N|y{6q|aj$D!G+pCa$$0?yE>Z
znlI(|R&Qo7!Jh3S?a#lJFvYp7JE^I9vV~!3({+av`Ir)O?QQ^|^&sPAhTTPz3ZVyu
z!=zqRSthIK!X?`vG|r5MUULc5nV>s_V_d3!DdnoUQlZE>9^mt??nOlLN%1Yc1zae9
zrK&<KFyo<am*o*Mk(oYm?$IN&p1*wK(JY@I{SlbRsWB;aZ^HZ07jR#bjy50XPbTt9
z!MySq;9`)7;yi&9gK}&P<~E#Ds9`iEEC%UdHiF{gRHbC+MjL}ARmXgS$1UlC7q+;f
zc*$9&i*>;r%s(QzE#+M9ONqME*BTywKC|y%MyBbgD5&D;tf>7uNAnmgMn#~ioflxn
z1L2*LoD*_HQ`v|Mq%bz1MTYAMUGX<zmXO0zRZ2A0h2jB&$-<~Tdi1N%u4bj)56x{C
zC60=;RS9Vbyjtz6eR!fft#n-mGAl0(?g8vF*e#xbFDZeL-$;SqDaxSGh|~6{y&n*9
zmi|WTM$Oi(mav}g<5lg<K<X-$FP0k<`r^k_ptb1&W&}5%w>1ebP&sS*$OYn2imH&9
z<keVx71hAO4|IY%_$9jX1(se}VokAYa*J0mF!2Ca?GPpKl!X_YfsRm3#{Is0OY>!6
zqp7ijPdTLJYEOrG5F=gE-`QT)ta{oAHGqU1#i>Rg>m=@PV^MWz(U+`{rsleOs|6-^
zL~9GM93ziPUt4~9L#1Eu=@nSfNrn&Wmcyo^QLw<)({D;l;=|tcm>JAG?t4aZM{Gwu
z`7CO^&Dr$eljblR$5ZK7Js0D3Tu}aJ(gr2+12k|n|6dRGB}5ygVx3DPEY9&e1c6~r
z8g^e<azYfEkEjkoeBYkM2xvDjW)6P%*kfKd*}l*i87?$BPf^bO%P08Tmq!%9a$6j&
z^izSAklPsAhD;Q&#naExWyZOVLmH#h>&MYM9)(F#eu((cGFpa16^L-zi2J>WLNEcX
zulvWqjuyhQH4V1iYtu=gHi=gUb5p+e0PmiZ(&5c;&^s5)o(AbhF|FUpM*`O3^1B~@
zsJ=x29Wnp9Y0t?w0iW-P{FnZ-1Ybb(V=O=~M$=9!KpmO@Cb0aR4fU|bKUv4aLm!#_
zdc-BfhJ(k~Vny&zJG?GBKLXv|k#WZH{;RE}^o$)+{nI90s*(RJ7pIV;J&m|^iDO{n
zm4k_-`4Ckg7bb&I_E>g{GgT<6{J788yUionMRt3|eBaC>lBG(Sw+~)l=RW`6dJT=$
zkpp=tZ<JJ>o43+rc>#F-ma|vK;vv0jj`Z8lM`s7lJVBHF>dQ+p2Dx(I_<ZciU0)L9
ztYi*e0PuMbxWf@4{NF8VR<{2Fd{82CGXH<V2NTDC9MS&>AJMq2|IJfp_B5&wSL*;7
z`qtGG?GMh(>$hT3t2vTRP|f~RYKFY_i@#W#<xvtx&YCrf%`ktpl0>V5+f|b;8oGU3
zG@i6Uv@YLxdN&m|d$D%6wAZnbvK1?${M5K9Cm$6xFcc+5G+##PhJnG{DL$ZFNcUgb
z!U;PIEE+3JpkTq1u*T<tb*)UtpIg^5&QSdrl8p`H+M#qS0Ier~Pc#+->r`$McS*!P
z8;PaHQu{Gf+UuppYU5d8f#&m10L2+aW={9Rc&;LvOABl~3jGJfS&~KP^&mI|P7MkU
zM3Gb^buPI%oeXRnX1z8%10(_|H8MP1_6p7h@M~49@0W4W&qm1pMDz_>lZAznSgBDz
z8Yad>%Apjv9O!++YhLA4hIA=5!9Wll>z)Dd!Nm#8KdF$K1G>4ryb)Ds#1+ijGCy?k
zh0M-OemH1MxfayKs2abc$;w!-#D}*oiHA%{KN5na6SJmAPMY&2-sJs)Nt4%}v^-8C
z0iw{vXTZy8b*r#9l?=sMg3!Y){c+K07sFj?F;hkkcVX#}y?R(jf0ZcBP%SB@B)S$b
zS<D5&{B)x>q6QZ=)_Nq*1)VO5mBCh5J5G@94T>EFr6tm2hZ0oycX3w^n&N~_{hm&0
zrXHD8h)da;nFo`v0V#+k6klG6JC<Y!SP5SBT_X_7!{Y9%%`#2UnvjeiuhuKw?VRPg
zPTsBBbP=#&X2hIbeq7D30^*YqFSm7ipD0L4JIjamlWT|M`p}H|hrT>o?FOQ&I{PL!
zvpqTW97jpJq~?-*?$g<=hkb1JhlGb!o(G>l(yDa2wdDXE`oA~!Al;vfANTVB^p|Xg
z1{O*g<kKdbz}MLp)35P#U!jgk_T}Ynf~243%*(X@2IrwQ{H=*2DNX^p6PtOp8qs-(
zrT1gR6MRRVja}K(OIg?kU~F(_ZhZRImllq10kzoBt6f-1bB_=rtXIi*qhjC9TMswN
zm#bc8{IuHb^X)&r?A}Z9*%+At6&{^xq?E<Bj1-N#NvF9gZtZjakNnQeXoHFF76eQf
zf0*sw#j7Eid$IR3^{T)H*m@k)3<hzq$MuX?mt&<Tr|*2mH@(7<Wr)Fd&~SCyxvv8^
z`qxpI5pHf=AG-$Y5F7LNv9Z4FRo~23o~u2WrNg>L`lWfdB)XF4W4lQKC>P?_NZi-`
zo?a}*$1$Uv+i&dW33~-ynV?Ic*J<}4In?|JIrhnSlX+`vPuuAh-<Jp6cm7g4&6l=X
z@Y4;~FZKmz`nXPxM;xlXZJiYOejWa{-fVktnn!@FvTLWjSsBjlXtqtl=JggBs4mK3
zA77ex0HLHHtU*8A7J8Bdz{43OEn%tX=R3J8h6$^mRGFR&P{Tf|@RTtz#X3nWPoUC!
zPmTb;_x}5)txdls6A{8-oPu9fDZa7eHy}=F1Q4w}G7~=O)NcE9a}f$<olYZhotIY}
z62hpZ?Bli;Bz<vOo1b~f_5|YBtwI^&e<0>9G3XXh8?>M{bOQVV+%O-hO=X2lC?vn+
zPK_AwN6?u&{$pqU*?CXQd(PeG$o{dw0P()ej`IXS;5(r>tYShxQJgYDLP$K}(a}{j
zz0QxCh9@b6%}E_ctXrF6TwYze>#BogBAs|AUfWJ9#<-fH#r;$2^R{v+x>J=UT(zK{
z#sghBID&%#yu3u<^W0N`g5uGw&y%(M9A#85e05X@DmOJ9Ay>ncD7;iBx{>Afm+DVS
zD%^Ce1?A>e%t(m<RLPPj62r!q7U8WEpTog%2i#62so!AGo_T7TSlF(Y;Y2oDgod3h
zF@0slf3B<6XDdGq1yi9$p;PYn*PT<okt)Up=L*uoU+l~PWB5l(xk(zbl;u@6IgXMv
zkt=%@+=HC$8hDFn=6`LFbzrtmB1PWk??)H#k2uBb(Tl<_;Fnl0Vn;t<{+<s7k?b|K
zja71oN0AnXJgg1{oeu^*@RL)x9bB3p`lexlMY<VxJ@I{mJTf5X>fyUtZ6RbWPPi0}
z#>0V133vSo2<O!FlM~5~|6Gx?pIkl~#c%cI^&Aj!&vdFH<ivaZ16Plzz0;UaJSH+1
zYwM%_STND0N&XZZOa~#+gOB(=`0mC1@^^|~&b)H)^XSBcE5}dYX6eGAV8J#_y=!@w
zw2sYq{BN7I0K)#w=wDZhQp&ut%gZ!aJ%oTSLqS>wK;7O_Ge?)~2JiI%i@;c_)2U$h
zC%#8{Q=0&MotgwcbP|?4EP9lw0J{e+s+)6-?~+lo)m`L@UJ9byNzi(2NfKwx@09z%
zpQ1>(y_c4DjbO9k8_FL<7j7p4Wv0mtBxNK-7%I0)=XMcDD<_V>wcBa9sa4(gqiqhn
zf57~d0&?}a8q1TZ{)KI3?Tp_>kWp%^pD3Hg0TcMaIZ~wSNCmNYdn@t;+#T8=D0yFI
zX}f8UeN2lfQ^t5$OPt;olROXc&E9-F=nrnbl7{jjXJ2RM4bv3#wIewiM-{x8jBtNF
zV+5^pp-3I={8hV1!}xE18)#&D!;1MGik`Q80Ke9CT|NKAf~*hzY`LO(g??m;pP!K?
zyXw`Fv+m7D0-B67zbISAPWiB33}kBz$Q{U)Yl(&*a13IGi{?$&7$8-lRBg@u?N3*F
z?-drzVvvIPnsjCSA`z5Nq>U?jNdx(`t6ubLjTrrU?n+f5)dv>OsaJEvRkuDozsFII
z@%kQ+@|S3oXS+$&)0zXhpc6$4w3GoROb7a0Q>PaguNmuKpv)ybdH*}BU}OJ})0u^t
zjV<Mm9Vj)RrRBUngzUFmy)B{(S>A$lN$MG|R)ngNID%4si*#fJ4d+e?$&`Wpi$w49
z=Hw^jFiQA7Z!FLJyz}FOD`^|IzL<tUNBH`cyL+|%R?&kRUWjWprG`7AhN+X9xEhw}
z%c7q5!?Yo9*v)I~5^@6>YZ<r2{Y;tCi4eR5fma7Gbti~Le#v8*wh(GGL#p7B+IE>X
znrC^>s!M6QJ>0Mpmy^%i_PM+f<Ir+U)Ahqc*v9912OV#3at3FSU;}1g=C+`_%<;S+
zW61s6Va9ZoNJ2bSG(&OErK5*0RXlGC&ujpbQ)oTYIILf2eO#rLIJX@R#4R^McDh~N
zHG(XF9C(~vS+}ktDR;U(`Qqa%NwXVb?)^$A5vaALml+Xio`2IzfKE3n-Q2aEQ_qx-
zf`4UtU?{{CV<q}&erQqS8-J|=(akiPlS&SYI-Qk@;GNQRs_#1Pq=tM9^-kZ(p4i*Y
zLFgf>$@5^o@ZD#mB{!r1nP+27sb=A&=3hS}NrF_H`SuMIvl9!h<dZjw_7$M0<?Aa>
zz<We^b4D`{#KQSYG8!&slKj>uX_M$5$JasLP4k&>0S|`;BVf+r+Lzn6i`Pau6}GpF
zOHulVSy}VM`~CZ{3&;7oWO3y|tBdh;6z`lQkKN_=L=-Cz^=U2ybJN_Dq0Gi!9k~UN
zY??0mVEuB8ZY5p~#e>6~SEhAndM0MqCE;k<Ibu?{cSGCkMQ~OqIQW|%>t-__=zFsW
zOjk3f6M0Tj$W<QW35{$mIQ}OwFc0yE^8aq)2p3bJsj;D)4D56ms^({AK{6D-3s?Bg
z2%>LVSSNv(gjnLgYvo_wA#1K5nq2`jdtn4bl^R8nd{<*NwiNLU4fp#}b7gKMS%Lbr
zvztaSpashwoBlwt(3Xy6<`naVP;79YR(%Qdrsdw@ub<~1c~_s1%j+!Dn2<S?pfGUL
zj2H0AJ&dv@ACw$+FHdyhGyOX=<;9MDdT}e(+`hCt!-Ym{{c^NCtDNoYKhc3B=i+Pa
z6>PT8p(zCrp@((?v)YG(u|OzNuM;&`^mh3j-`?m8I=}=q8ebP8O9=N$KXsF4PeHmo
zZ@y?~fMc*}3xbe8ED*8gH`A0>e1DJcw)k|(fW!52^D2d+{(8Pf)a-s6nX^IAA{E^!
z7Da8*&inzj_GYNQFQ+!euY4f)PRLmhk*#Nv%kLNZ42a~w^JqP1EM4zMp`Zh8ZM8_L
z+>)gUOZ?Q%LPB<1^6+SvwxuK|l>1baLLDk^;CU8!kkGVd;C2!x!ry@I0s6CWdx&l%
z=z<4UC!U@UVKqmzz_2~#;Fs+_6%PFhXmN;WY|*1*MDeEygh#@;(LLbD#GMx{DzxCU
z#j)<bL%>K+$}iVSXRZRUd$bkTF_Tfg%TeB{y_zv+#v#U0UYI+Yy9vYue`}mgsTMHk
zcYmx(@0lA^e&jbug2uLtccrSC(GXWX+kcxksC%)_*2h?$i0feOY<H>ewT5S9T(a<@
z!KqLp%g|bCR=;BwnPCFa<B`#|X^~nT%yxXjU!2+NElzI^p#vC3kk47?40Xn=A#Y<h
z`<daF_eyMVHiFa(N*S?RoQz6KQkmvcu=Cs%^N}eu)rgL8SeKOqyzFbpaI(_jOr<}n
zDUMI-q4n}u(EC3U9KgaChb-y_-K=`zAMZ&;;!+Gs3I7T356%HpyA{Mg*8TX6h@?>E
zl!5->rB^w`*Mn#Z&mCkF232=-ND{kYTI3FB?9Gog%5duWoUjS+!ciFb#H?}Lc*;c)
zDhkpb!Es?GDdfR8r*_vHdu)I7a2y~POAl9(BZJZYe325=?CRk!{2E@$C8XSFjlwq(
zWE(V$A^Y*R(Q^SjzCo|s+qYJ}@Lm?MjITWricoB1MP^6aS-!y`<%eKecp%ZET!)cg
z!TR2+_Jm-~XHa#HjI}uU6!s-^34VVGelGDYHrH|P36i4-(1fpkK39fv7U(98>5Gp>
zyxuf~q45BV;k>8Fw0_{1CZ!`gC5kbQQy!y%x{h1xmd*$Gwnm~9MqrU22R0;>o9u*L
zZn$_QiP?<><##LiP#{hfbaLSm5p!+#XZ)`oacU~WJ#0J`ncS3a9UFf{pyF(Mv%WGl
zvxBpAeCKlbrqBT5K_`S0$|pEFY#N6T-G$+(9iS0_o~23~Z>4<*P4<Ef4xzUHQ$zPq
zr$z@uXHOcS)f{MfB23-c%x|u{{)4n^3BL05?ZQFvPVx$Yl;@IV<QpW?GymA>J}Lzf
zkvbn9w<4kc4ZDa9RP))f%VWV*a%qJ_{3FGSjJzS|<gbAX9}=0-ya+N{r)<&)^=ZMC
zCb&(e&cco})I%EOd)c?=8^*U2^Oz2mW>|wvZWma9?XMVK%CK=Cl;PiyDC6&l84-pq
z*cjfj8D`AWe^68v%Ej}BrJbdDp`TR-boV=)-QaqT=dDw`O~s}_{TbBgQ~E%2!t-ha
zf0So4hKM|~&WwOyvz=Dx?z;v)?NhYiV;*qWiC~U9E;&v1^MUWXae(CXGJX&7bv*c?
z<>Ms=aPbuNOA-X_SPf8KI!^^v=5q#S(%$!#f6uEl-TX2RtPX_l6@%~VlJroRhA}&1
z`+sb`V|OOdx@{X(Y+DuEHY&DLv2E*(ZQHhO+qP{xx7IqRox4BmKQY_W#vG$}{IDI@
za<za}6$|o?O89M--$h(+WMP4RXuSi~Na46ve>5^8D$MbJwZiP-VBRSlKUphO4Y2wi
z)k=D|eh=t~%mCw&W&dsq%RhbEmZZ!D1h&4uCn;)$Awb~knfB2j0CRhN`P)CfHK^!o
zF7_NGHzs;|$Jj%e6|!nEPc%<Es}vE^<h=IE+ZvD~)RD}6w$7O2shId`S0dwMZW_R0
z=3Ip`b;2@rLIlbdpXf8d+M)BfGPdEd20%*9QjKYE$N61}=}*Erjd3IfL7{j6mYl6t
z1kkRKBkT~T5lmOp8Dr;6ug!4_aYT~<3YZOt2Mp<QP2?N6VE0cIq-i~eUFV1)ZI+CK
zC5{JAw?a<B$y^ljS&19uiE@q#B5j(&=~YssnZ~{WNApI<S{~`vTP%-ilKu`;7rO)v
za<YylC83$s6|$u{=6juJ@K|pEfpmetm*HCykjmALluKahUhv{ThWt5h-`-4GuZCcQ
zeDELn4Q{WM{-CuTCz{628snv<iNH==Ln4S(L<zv@c5bm$^0nv7cD}$;pd5W@h7G0g
z+loxSf?*Th3|HB6q_i^0!#dBkt_3=&y6WPhVMn0wz{;i~;UAJfRXP7~oIV$!`Npw>
zD|w$_W8l<ioH%qd&Y+H2LVFFY(oSns;4UImKx1vYzJ{pdE7Vr8Y6{(@oJT3FC6$rh
zvi@0+kz1r^r2fEr>R><80rw19u-M`*_gDq<;C0x+h61&@cANtEAi|wcc_a=3((b^r
zX1{gaHWkEsn_uaQa}*E&UGDphk>fs5_mmGu<XFn^4gOz8{^=gA_@ie#UclfP%<@#a
zKt)(`N<mLBO(`tVt%U)*;|@K*uy>PF_SKUXWSgLfeFlp}zp{d|`m&hWL|J7Ec2%*Y
z>iEbI;4A1%BMPcW3H7xb0@1YhE*kzpm*_85`VgvWU_m7N`BmBge%GrxPeoLn-O7)9
zeQbX?K`np4ZvJwlh7Y_)?iC-P0eR_%EgjIeOzmS=WgH|wqj#1NMwGd6x7|G{j`!Cp
zgUaY@K2;oM((e`OCtV!VnFh;SxMxcJ`|?oN4L*{i;XBK#17?{rZf~P6O7YqrIA3y^
zv`<DFYlY^M{QTnw5R#>;DMH(qS{{kmEIix<6Qe2j&qDDXS4$pe<>2rYPO%&rtdtT8
zNk32Q5()x&+t}%?x*|69aXiZ8Bn|nVJn~YWl&$Od5(z~877466vFt{5ew{GaprdB(
z<*X6L#AX-T1Q~H&1F^136Kw>_Km)ECI<*(%6@q=YzM&%rIP*B)p7=?`7kyu-J<tCo
ziVG*!XqFvQ2d%f>98O}>y^Gb)oS<iTXCj|$y+5o3v@G@#7(0N$1wuU-G*QH|pZf-4
z03#G&yC-w>^$R1v-hjGyzt;xA(4#%z&P~FuHb3!;llC)3|J1ziv?QzDq-cVEl+l2T
z>yU1AxSnqV>`3`sjU8T9K)sO;;eE7KCe%=JITm-rH)WOE8j!6T<2T_OxhN-V8MF>V
z#6S8^-en<=6Kovkf_e^^d8ftmNj5&D0RzsgpM#+U6KuE=ib3|^r_8iMwVok{6>tkL
zt~K8<pE+RvljQ$@5s|EH|FHb1g#Tgs^YHxtAt%d!5xJBAbt%V#R;2&T`^<$^G*|=>
zy(XugdG^!pWXJDGl8hXDN!3Nth?@w_YgSa=uFfDRq;$+JPcmVkSg>QRzu$pF$1I4+
z2zEkm`Fq?Q-vyIUC=V1bj2s-<#}Xvy!qp2s92zn|x$9+2R~ATOY%IC6int=$w==e@
zGp|~A)QSNo%lXE#D`3k_R~0T?pzBNw89}xkvpVfJLR1&l0S*1wI0qjL+uKn%tM(w5
zS2iKrO;b>@;lxZ+4(~Y`a)eXT#rEkYUKR~n9#S8u=BY7o7c13_tj|u1feAr<@GjR{
zF=Flp;~*PJWtJP2!)WTNW=l3#@Aja!SC`99E@OZ%g-ib-@NtfbZ|F`-&(qN!*x;vK
zi;baC`**<;l(dnsMapz=vQj3xq<yHBXL#opAa0<C<>G^%CVhRE_#p|TODf|vwNh0q
zZ%ZvXu@woHobghK2baw*vkWvUrtE3xIMIhUQn@o&{HH$Fir!WtZB2PPvsywpkfr=&
z>;S;?zPrg>!d^TDQdqId!|BFT3ZY->XxPYs;Q_X>_=}NvQlzLtFT<Gp(^s*l)Oco5
zeWXg<<;COU<J}#jPNY!fddR|GwUrW*Pgr>30{h#>fPokEFbNUqT&T2CC%Lwor1D-u
z$%;Dx!q`8g{kq2PB>GTVeq1H56Zt*ED~x~^<g9QMzR%Xl?LEAKK8RF!n8rBC`cta|
z?I@Y;y}*phSz&QsW8=&SSRkN@2>+CyMO8CYqJoOlAJdk=HFUVwRxpF&h|}Oqh(A3;
z552CyjQXdo-h6JB;5He?9YiUJ61UbITgzQ|Z+NwRX9UIAUIuq8zX-ufG0w7e9#{Z$
zK;?opfCnYoUA<e4-4Pr92_QiYhgM(yc5X{nrMk%p?<z9NKK!99GOqU4{prv9gA>=A
zR&S^g8$wl<(<Ua6+xxj+?Vzu?*4u#rvyp6O_l^sEAYlLI8+)qypIH~7UEa0dVG{g7
zNDCQyzgB!{{ypfiMVf*<vVV-hCL;lhyk~4CLDd<DqKT*)-;p^eja2)>=%GASXB@&S
zQJJ3hbhG$)<&<NNwN$OO1-S?=!sMpWq*=Isn=G$OYAZE@h+xMT^^`2xnj(F`ve`r7
z9Qk?nl567>&7E9sY~cE}&&+h1x&(K6{o2ge1xnnl)!N3l4i<aw<TR`OFl7K<Ypr}x
zhY*Ejr<S7PtbRd*0`Ey@dZ0IMWizxVkhjPUY>XwvH5piG#;C?C6koHcEkH0yK*uIj
zA&a2ydf@9InHmKBzvs<=N)V-ZNk~A;z<|U8cC~Eals0X-nXhAy;=rC>Fnt6rQl<n`
zKx{{u-%7M4lhU;$lg_%O_J{!=bVWZ1TF{`j@ucURWB<%--oO$k5Dea`Rh%;x(QSxe
zGZuyvzmMlQ2`dlgdYCoa9Ivlx>sjrM3Gqy~Yo~D^!2~X+s49hWi}19Jt*qT};9KEb
zlS3)ead~#!CI^wxTU;DbhnrZlAO8t+Pph9c3B_HuTzn5enm0X|9|Hj@0)TAvP@h8K
zwAQ<3;#b#=RR7=Gd>jJIO&W-1EYf?}1wcKg{sA-;guq4Islqx-Y6Qin!om%OD0!m5
zyqEj4t$lMq1ICz6a6dJZ`T6na+`7#^$c^?u+OZFYNoe$^#Eqk$9kaxKF<#UvZ<T{E
z(LkhPc8P@=H1|#f;wyk>wo*O6JcZNHaOl?@R;im&iQ19<y<7DQj~0QXE#|ic4(iXI
z*6tcwm+^Q5wAnuydKvq`pG40NhUQR0jau!7cKD=`dIdN<8E<D4?di>Ff_THlzdlw2
z*n;W>Y@a-^h$TbG%(6IbcAlbz2E;qb65>!HB1xH%EQ@rosObT->hY0D;U_eB6*NFq
zvHLNVDx*y56^EqsJosgTHKPruo3BAtJrN=@3n1m_PQ0U0DctSCY~?0|U#eHLYvolb
z(#7(&dE?csG2N)j=;xKu5LyEdZ}*^JtvdqPHL2s^l(U43AzRT{CMQ6pm=qWD!@E_W
zzcC93a3(L4<y8QzHu^%w9@YbGUNb17$%%q(Ck#EtY@OajXS}#EuRq*NOflFERTiSp
zzlomng^XJ=!t2QfabX*&XbT);ySDU#0}CCGe6I5Bj^aoY8#X)%WX)-whATZd+DSC<
zg)MOP9~$sJ(H%L9SsTfJi6%_|_0bK&Nw%Xg`yBG{Sh)bO+LBXAHAk?tT%RA1?gCU1
ze&GZEs`2?{40b<*u|hF5vcSZPMl7lxqfkckuVd3{4b1X_=CzF-BTwNF4Q>bVBhC))
zVu#?Y*upOjDbf!yX@PElND1|)TJSe;h;%BybC>qP6lnaM=&x>*R+1q>=#g4jU6ER6
z)8%TP{W%PPRxsmEb&LTv3VSX)io%Wo6%8X8d1}$_`8|F-wN6Fa9z49C6r(Fz>4Wx<
zAD_GBpCxJbO3t3-If$00M38WPDo#$l=xl#2&ussQLh_Jpi039T{N@T?!b}ADe$Zoo
z;~u(qwnC_$Dduyr%&_Li;y1d))M9xpdd*7tL?Bmy1w;?_X8E4FZgd&=Joq99de>iH
zUUm~GQLZf<=36eWPy5F|EQ=dFz{XDMi?=l=MoY_^a2kbyh_Wiuys5kD0uL@{?fev7
zBPYa@V4|gjdioR^=+UqRNgRp{Qf3(R8lNcQlkUZOo)w*u+m3MSjh5n^EDme22|rA!
z>rZlk5R~Z1y!Z8|d77BfSZlc5zJg>V(oC9-<Q#$V2h*^*Sdp}WEoiTSka^ZIl}h9%
z>|+3vTd>Ge&}A_2`rj5dnb6x6y}yq@gR9yR-+CS2cFKtw$d+wOYY1wx#>f%*UTD3F
zU-1!74U#&FpnP+<GDBQZ%yBor-^1f=qht{PT!5F9n(fcy+w;<({671^y`Io-H1GuZ
zXppmUSU$Tcj87wW+-XG8O|nPT)SV<UbH5_v7;H_2Q_8ctp!^RMgqWHPU(L4<2;%Pk
zG*R^h{La%+j=8t#4^-^L93T0nA3N4y2IUhuX+Idz7Fwy80Wos>v@*kS6^*tEEzx+u
zJ7ssX{j%^EG0Fu7Z`c7utiMlPTft7+^kD?_7@VIz+<AfB_#NSfhDixJFFNjMnBa8Q
zdE&cQfMC`wHRSxKe(MB{tnT$9J&{ldqJ1SvR*UP`ZwwHKDKg;Ba&cWwp*E$&anquf
zXy2R28eCxMz+%4$;q*cIfRIr6YsN0d9`^@+a72Qi|0P7`VE$ixN=zxc|4}JV{R@<F
z*pYo+YIc>N%j1JNjvq5n$Li6jRmIEW%coQ=#DxwkgF#Ig4AU<7d$!DOF~j5g%e`+S
z!2cz!PfvdPT^cbj)jM|>-53R~UkFqeP)J4!B%TE!I>eq(a-jHsy5<^R=v|(kNT4-*
z9Eyd;_`n~#tZJrmv`UpO$u0q^Gxy;p!3ij7wVdDPQraEA7$+(y*-$&EB7U+VlVjwj
zjmDjrOSDgbWf1hM%^&_*oKq!Ts;IS3YD-E@8v7%wss)Qek7V@AJ4t|>%}Xj005LYV
zR8TRVro^!Q48`1w>&?c7I|1@Ut8We()tLfmg;p9Fq|K?-R&%-A)aig)6eUY56D#0w
z>=rW%^Loa|tkj|U#T~OWYeDAD<#F&*sUKnNYoQ8~-E=bFf0MxizJJ^~DF~35HAFg9
zX*=n^*jVx*%S*4WOyqsq8BkaTp0uxv!1m~WpeHFh<OA1oc=kcQLtmKSdZ@1>DGbRt
zD+jV>I>AWX<cpx=)#Ct#8@Sb3Bv2jM7v;V<IZ&5h;AB>PI6#%f6s)vmPq0>ceR_JJ
zdCHS~nT7sKCVmf$)De;czqZzc#57Zq!4WNAs+zWO(k;3QRiQpA8ZdZ)Ca+K<4cWMW
zN0>@?P`<}5D9tVpgD;tE9jyGwWudiYSo9K#kqi>%zeBeW_&foLZgP(nfsi->mXJlY
zhUi#Dg2nw!8pkpXjW-`(LD}e6#6T-&a(kW3kCL}+t~xo=E!%5t13N&}Lg6?h0?-yU
z3PpJky$PE5a($&8454IAIKmKIEFHq3#XB)(ASR0w1j_k}v24*sObU1eYpz2K5&<0A
zx}3M!;sH)LIcfkKJm^D@gLyr-09FdrDC_9ELAfOB@M7!UwSES};`|zBr|+pFC~SLk
zNE|g3C9Kw2ko_sR%{=PV`qawaNr~h-xq=Grbc@-UI^YOP!-!wOZhx;sw>&D<fP%b-
z_%Y|+&mKiv5^=AAGnCx|{|@o^etF8Ouy|mldK_bK_rC#J@E%@tSu;lC38UIwxP7~&
z-Fl5%b0D=ETilZM*oV)La1~IYI{va2Kcn?cvAtpbbN9duG8vq(yvm=zT=~x}#`ugb
zF6KgxEU%N7y(qKD7LvrND_NWyN@wL0F<2Er>VW-(3V^>H%@F4JL0_#7-4WNU|FlK`
zuIL3ZT(1Ccy5G$;#Pmv7M&{4x_t1!HT268NH8|yscFsxmE`G)=D3=)8D!XgDB{9=5
z6HEfVhAtca4#r9}jEo2426a|AH{%p}rDmOJme{%J;7qEE><EoOAfwGNR|w#Ah^m~2
z^cM&YbtPl>xzLGcDV7Z}Z|!|TB*ZTI4CKN}^X7otF7PJTf^w#&t7^iayt}?E9?^9H
zP|zz(&}Cz6%xbFwjPKe!9U+LK=&)aljp9aFCZF<lE)Z15P0v($ru*7%FW63eo1dLo
zP0e{$!9r$_!>G#0W!xf@MCLM?@n)D+IJefGk~s7a^^;{ZP{$V!r!j^O`@NI5R77T<
z5k!D`6<#jH&C>fe#5!8Znc-SJ78XyG*%#s#Rd(?a1BM4ZrD5~un=GLp<Rb}xk~M||
zYB5{F7m$MTu#3YGNSVxE(3&`W2TVqowq+5DJHM~pqa_P~6oppsV`t|NZtpp9m;8qh
z=-@@<>wenIup_hJ{ZmGTOYj+b>}I1Q&Ith|ipGK@Y&3QU&?4ZBA^Lga6G^l*I+?Z(
zc*%qUJ4FJR!y>M+a3)N<X`3dfwRp!z?nf1MqyCW<bUHl&P-=%}{Ba2IMExs^vvaOy
z0#n0UdtJ&7c9?WAes3!V2IHN+Uy{BH7)e<tRZD33=*5>U@Q#D5Ptg<l1pRUw$F+cl
zF$%v3+t@BfYZ(MEx87DB;F{nKFD3{Fk$@b$9x@l1qF*VWhRac`O4qeE^78eA0l4J?
z;jI(yvjE%mad*4q@*9n0>rwWYn7%u3s^SS>A{?;Cucj)hi%xnY0^S~Hl!A_&GR9y0
z6%mYxq&-XIQ;7BMUDN&U2P5IoR2cwU(4^b=9CDWyJ^l#t%WnLW8pQjEbdYHtgY^L_
z<Qga(mwdcGe-ijFpcm8*wi7Y+d&fxMT}Bb8e*G<W8U#B>z1l);UtOs)(>~=28F6#-
zc)b;F14zLp&@zft*o5sdHg9czmZ$sGw1+^%xFk8uETughkSJdIxJkkh^6UXT-jE*2
zko@G&s6UTw#ImMPNMVGUUfQ+VoeQz!W|zCieF#<hF09E_1er{Aml5unVTgG8#5h)g
z&MIr>lK%A1&AvK55>cygq&TH=Ry#M&LuGiDy`Ui;Q#LW8eQ>#qD&x|nm$O@=fgW^=
zX5qHXhiEb0Rkz+5VZB9(VbTI3{14x_^;SUqT5{RdvBAA=Bs5aL4Js^tXqCneWwNPS
zLAY?76*Mu#u~5X4>fDM1Okz3T$f~xRExV}}DhJd^8IYMMq?L=yDQ)S)4Af<)W9Rlk
zNa{U|5l9`%;<YJ2%WbOlxRSiB2an|nJ8Pnt6a-l24@KvzIdJR6)<Xc>Xwc(Xgwk^B
z9uEr%Aj9r_+Nfw`maa;ZojisEd*ta;)E>4~FB|Y_sQVgB@N)x%6p1NL3?T&T%@yZT
zVP0|{B-;!4qQIZJga(GN##(#v^j;7G9+qA3P}l`I&?vjTK~p25VfjkPSyugUbf6jx
zq&BegU))@hOGrq8JmP?&dK|&O;ekq78P6Yba&mr6?hMSj-nmaDCI%m7eo%Z0=u)Sv
z`!$=(-iv2Xq5dzK2jI56UxCnrF$bc1KT8q$1QQYc1XB?@KUai%f|=cC0<S2boogNN
zk8_En`$&BOtw9dzDS;X!JZ^CSsg-MOeny|t)FWp<3O@wxEd&7KIDLyQ-Oj>tT!B*U
zb@|q3_4PL^kg&W(lr4VwQLDE?=OpG2*2p?N8OVm0QZt`Q?_T*;4}7ljB@fEy49@Kk
z$n`lSZX4gV{1GB9qKr!pQ11LRH|r<e!IY&0MXk+>xH`ju02NKDMr<)9lwxPYxy9o(
zfP^N){J)11f|G->{{ItBvvU4-XAa89z?ecr21yABQjv1p{O5i8P&*Z$GD4Fq#J`Xl
zi6$6ZzA(D9=1Oaa(3J9A6F&@2{KWiu^9N(~apLU~CxP?@>P8@xu|Y}Cyq<IY4yovh
zA-BptbZ4WVAlBF-agJMT9HXF5SmJt%$Sjl*ER-}ARf$(wKNLCry9w1bJY>37qLYoz
ziX{(t9J2nr-UbAsaY>m*+B*QV6^kX|81_lFo6S+5EFOB$f+NwN*+i)f6uWA?zgXPe
z74Fdw7&~XPx<jo6n$bfrQm~h@(Nd=RM{trum@6RDwu_ZyzQj$ci#B9VCumGFPg6I-
zoiWsML#KQ!DVjr?^Jm$wR<a=%L8L&~qR|8Xk~+7toYPF2?McGEIR{15kL98(jS}?c
zmWs>Pm`PB-9)46B$W>M!NC-Y~9exwM#%m}j`9gokLRio4^@r*f%<D~Jy+wz%ErZrE
zs#=t`xlZELBMOEkJisR1%*Uf3(b6ySb5Y?-;#h!|hfuYHuxs<6FYK+Q3aN~t(W(Q+
zcXia|L&T`%ejD1tX$@t>du@TRvBr70HxmTJ3L8m-hVy@ER)cEEcciP<^f~11B2%Nr
z|N7Et-1aDF6cHp3disYw<X%f~)Z;QnLG`1^Nk*EdFX<t`+k1c2-vIv0TLNy}D16q|
z>6(QLZtmJIt??%(jk8b7BQ*kuIfoyh@(B7zQ$7?4*$cu{6-`51T3}}GsAX2`cWvea
zr-t$C4?a~@Ttm(h8+v=dL9mb}4zr7;GTQ6H@H4T8!j}uKP<-o+41?~3n@`mC?1xin
zwb`_>6IUumHc2&Jzxxk=F7Zz#@=CV${8#eycsP+O$-HdEa>zI)sj7j+Fnl*4l!n|i
z1V&4X7{SX;pt#i5=Jv<B^IA_J@CjICI<TJ3jbfBQbIV=HJPX>%%ga1tL>@m-xDGQv
z1IYy5DMBuCICKA#hQhnWvjn>n>`5%uud;kJhr`|5$XV3qH7j*GnK(S?=7bfRUL>@Y
zoeLVhyGB3Q2YEStni;57BH0kYUTRIQAoe!*73q#R^i#ta?G!Xln)taaiMjCS#-;=>
z6`yKhqOW}U$|Amg64RR-`ho%aVfOc1rw5hU8xQo3bzFI{izhGC%+gHMeT;zihs812
z8@7g}7y@7Rz7lvuk=t!VaOOJ{d6PqTu)u_27gGBW1a$P4SsX=7>x&9Nxk_|+Xa#0-
zYhVJ+R{iBRwY5(u??&;Gh`yzs)#d7W$2uHTS{uhUFc~j=(8V$;^ngR$>n2F>RM;Mf
z7L3^irIF7Z*dD%r4ETi>w&7nh0Ff}ue%RFor`2b{dSR4mdp8c%{Z}0*V&_-`^dD7m
z|D&45Xf@9t5Z!P}FKFYG1S1Hf6xt&&G*DJnhX1MtNH@Sx0qpz+YSwFYyxHLh67aat
zzw@Nype2lX%06Geu;CQ*gAbgS`H;Z+{-c`(M48oTh(xSl!~ytv+FCu(N?_uG6$d@+
z9b3eYh@<Km=I{1Rx?Wk`nzf*a4t7O1oVL!5=P<I)QmB)L`y7>0q)7d*s?GT6MfEPG
z?q4!~w7L`lZ$hCAF@_Sz1w)c#d6S*J8OCG*CT;OdFyC6kTohCtbRGje=VLYY=*8wB
zP}ehAv>U_M&msDEPB|>lDtvTwl8UJ<8XNp6OmNa0Sn1*c&JEO7S3TSNelZA*g=H%|
zYqoF*eBEwM9(Uire_U?{QB``5Vr0|uTV$qgLZ;0D&FWBTAt1$CP<R$uwUVULu3<s;
zm%sEY<8ud#L+VRXSSVii5_PK}EEf{hP{>9G(l(bF-h+G6>5FNrxCyr5@!tC3xIcm*
zJ_f9fC(iX(O<8)X!Z*f$R?}rH+oRz{PC|<zZr^)tBuDzRH!j@{Ba57nJ2yN@ce?lJ
zXW+d7%#k@c=hI5$8*MMwdY%IVJNX>Mc!(8qLNuDt>eyO6-M(Z!G|Wh-&T8|F%`%o2
zs=YyE-*n}(Ka+mZx^y{-$KSDWvnM?Q5cZsmfiZhFx{FQBb@5RB7cJrx!jijgT{(w{
z9fj;}d8>VDo=gp%O#rO##9h=2!O7`GBN!5Zh^-8UL#&d=zrRhKc!@KHJdBjAyxX@o
zhubL>EDGVcLUOEAukS|DTq7?N_!1-6)jsb+KI*CJlEnKuD*tHmSeS3U=HoOU%Z|NC
z(WlupoAW~ZicIB{>%uyZLKf<mn<L4J1Viq&5#F2hWQ=TMWbcfL!YSpX92>gtR>W-p
z32}J8beL$R3PpmI!Zn@G7q4dJ9a0?9F_p$<M^Q%*IuKEkz7{Q?j+dTq7MHMyM5^c;
zF}MiD)fdOUMf@<8TiB|-Tf0_~;&Gp|aQgN^tUT#`u^Ib-I|r5~ce|60XP^Kw`N{f7
z@6nH#=VB<2<ggRPiA(SglO{Ax<o*L7N+mlAr-A|fzG9=CW-%_)rUO`=OcRC)NKz7k
zUZ-lRdaZHD$5M(ZW%^z~JDlRFxWb}8Nq02!#>VH}?df~VQ_=z4nzWxoO=*-lNk1`=
z|GhY9QY7z?x50W1-L)%*aTjALvbhY>XBRF`YuNpo-Fpjl2{&`Qc}TapUX&ML*sdN}
zHV=~}S~~guVTCc02j&n*<BNOKWgSaRXNf4VLSc?8S$t83T4Rt@As~_>5$z}ii0(dn
zBUI96T@@y-C&;fzCKkEMJeQ>N^!YkoQ~(m;3_2{Xwz9Tx+_8tVcn*Q)oJ*)6_yqPD
zUa83Hi(5voD_{l+9j8YU_7McM(8|Xgd#P-qt0>_}&hj8^3prTxq^}lRu&si$kN9<?
zo*`P%*D9zuZAFfeT!NGr?3uXja?2F=(;&$v$A&LCC!N;*_3%jW`<OI~1De^t95IU{
z$+b`-P47tKT>yik0Ze8?B1aLXSwW9jaS@P}L8Ac8493yjp43u7iNXWEZYSptx%;RP
z7Q{I<B+f6?)RYr#83}*Nyb|`AkSyB9k@P>7lXZa|qxR#%4s#Oo#WH`ysRWc&iu)-P
zKr3RiRXSO}(6#VFTaGDyb}sjmKo6F4YIcdGZ0~xJc<$OtP9G%}uln5G;!IJ){!S@L
zI29GdNg>X&fsU2Cm%IYR3pACm_dn>4c}?3$!x5PbTi2$Wv9W{TMtPeN2&?5<uq*VR
zNE)A)7k0w?fNiW07L8{|nK)T!OA+Gi(tod4K&b<xI$2<o&-J_>*s(!_|7lFdflDV6
zH!^XY7mCuo2_|s8<jXhZnaj&gYP@A^!t<SW7O=nO&UJeoc|8M!iT6|=BW+Y1qy*1w
zWI`=C@nB1DktBp}4qhsNfz9#0NtN!R)ZzTFFBuTOK%}r7WP3egWq*(OwKmcNQ$>Uf
zhX&)CVtP3^m0!%_KM&ihNLSM`HPC$_&tPVe*lbCAlLes?Rp9|o1-liZUG?D_H$OiM
zth?|mWrn8N59<L66QdF(2_{xYHOP52A$H)i2za=VWboxgY_0zt)4dT_EkU~A33Vq{
zM=7eCVupwNlEwYZ?&;}A^9@EJ7|B5k@f?X#ZfL;gxLsch)p$gG(tR+aYSA|pqc8vZ
zzWS!&*R${RrA$0+B*>k0O#`Yzl_-u6xFjh@d9>pYOo)KK*!Vu)(tTeAb$!!`pK5am
z<$62G8>ZA$0yb-xk9=MivVoxiApMc+!|mt4J2Y^3zkT@GqE2C?s*z2-_trQe=bMMJ
zFW%Xk!^T11o6g{uRyqa`Gde`V@{=-M!&uhRk?JXA9nOFW)tKwT#u8R1Ll_1g>pW&3
zQQdFb%0d9f3<vf_TxHD2sUs#OZ{<XtZ3GWcN}#!u4SauVtTi4DOZayMY}fnZ3T#J{
zF7ey{P;wSy?ch^LWaXR|(?&m>(e13pl>C$uh(#}U6QVzv;<qW=W5|go8S2Z<oP+(^
zkw7fafonn6E!M}IUc0oOYcOlpjEDQHF(CfO07VyoY_=P<{`C2d8O`VpGJ?4Jy4G<#
zdq0{5E2kRg!(oDNzUy%2v`~Jhp52!9Hg7Y`z8$GtyBbNM47ZcjMqzNE7f)3Wv=w09
z*4!R-lWwsIHL*fd4pxW_LBxo+lR^&pX7?3NuAx~hu6(L~aF-$XiEdth0ggpZ7P-o&
z+;#%^xK}@Cm?FUzjvb}=psU82ph;|VsgI{1z;QH{$0NT!F!ax>Kdrc`A-k5G;ZW8q
z+8G*Veg93HMZ?ITWdj(KXv4Q~3pANBbvd4Z&Mr9ut~s7q-qyrWbEEeAN_y+3XKl$1
z`d4bOpmH@A`1g{&;LT(+hD|a;nN})uu#y8ljd!qt5O=sol|3DTfn6DGPPu-G*@&Wd
zdOqA)9fcV3>UtClsO-AkNJenHDB!&CVVXObu3U`U3iZhy$ET4+z%JP8e)B*8lSW4d
zKE348&CX$SWATWnUA6-urjGl#lrdH_QA9DuDMJvXQoaKJ?QlW=wau(8%|##P7u*Ke
z*%OgV1CQH^Zb%Fsil)~$zf?dG=Jwoz%gb`t8eK3n$M;OCYRD-0p}hkZI9~<KHqj+=
zG0maSRKr1r)to;`Xt7+KELrc}FxMU7vy|;#*yui(T8@z)V}=#NAcasF0Eu#D`%CBi
z0mU3gS|1!!aT(*Fl_O^3KI$GT{713`987l6o=qFCx}fS}SYJ@zBye;P@f8^J<Pcs@
zG}g{W3QO$wOPte%Y!*7g!{O?|Pt;oUF7$94$u(tLjnEs$1a~@$=t$xD))8;RYaf+!
zNS8QAzYhX-BM}zrX@En8N_KTm7mH4GC9`+velQpl5v2ZSte%o(1Ob7`$oRjc>Wmz$
z|3?`7|0I~Zw3+2tscs3=l=TJ5igAV#f0F{PKwZEq^uvUs#iLa-baQHe?MzywY%5pE
zL88ECPru)Xh(BV{)^=>d$#>LR*mpXURfah$8qwFTyiG(+V5Tb<*RJS3D0{S%h+SQy
zbgi8fj*d5$hE-L*Zv^nccMu9{02{5X6(Ms?8@-yxuGIPuBOH9M_B&N<gqOq>fA2bF
z4zh5zcEy_4l}TrGou}w*-aF^1T&Knm>~&f?JT8WvK3nC8w}KqM7;IE~Hork{*6oH8
znH0VawXSYDzU5m8RB<RLwQ$1_R1d;k{~~)5n3|5+-A?J++H%O!Gs#La0koJ^Zmm39
z-<$@pQj(-A)DBux+0fFeQ;*V1`M(s6xaA4yDhkK&L1u`QI`x*ZBQq|)rZyh7;cWgj
z_1la%CaC_J3=8;Db$@+Od|Mb+CcojZ?^_#pBUm8L-+t0o-d?kkt)G7C{&@AQ-{w<*
zyd7#P4!%yKG*pXxAAb;#0eHRA&4s9oBTQ%Ji@ujky9m2MN>mlT_fCO|O<nK%YlYPO
zTITQ1hU%$iFr<px+|D$0l%@aj^}Gc9wU(rypvPoUu#;u-deGchldUf<vvXTjs`bZ+
zv-JsRmae!Dmr?hP0PVw6Bh8aXV_DYsY(^?KLdz*wWyjynb!2z>0@PAO9ab`Yo~P#{
zM(ig_-&qWk69vrjNWY4*UGh(m#ane&V&bD1OCjx7yZ)4AV7s~7P!g%Dh;u~vI_<w`
zNMedt+o78l#ob+dyJ8!-A@ZDnl>SwYE>E2^J%XG_RxwP4B81aI(@9V4BjOwGtXH91
znNOZ1Q38HPub;G90yN2Ny)W~_GCn(vC>^|eMDE$mxg=}2O^{?)sz3F_8A*-#l5|#$
z{?co2wj+{_uCk(A&K0**moP}#RmTvaTr(P#WTwm!zrNO5X!>R!6~B$`fx2RLDx-E9
zgK1xs68eQg{|k5mk)w$+4ek7e2xhKK1U9#!QM?pi>i~nd3gB<0C{0hZ(p*1#A5TT%
z^$yELw~YPFKsiX&O&6sn4?^EG7Nz+&Zy-^%sb32nXiKqo3G!~Z^oKlV5@lt>^jF{G
zS9s?@9hDwrYdh<V1-N82I%_~Y7vKQtw~)@2_6KA1X7f!j(S2KS*kIl_GDPC~#AXYl
zEuPj7W09Ry834<-_%Ye#ClYAg>Cgya@ajXnI%Gnj6m)p*W%rP=NNy&}T)kHc(mOpx
z8vU7+)v*kVX;f+#8QKkrX(7m<{!l^Cw)z6W@6=;CYHBCWCOgd=SrXW~n2DZrqXtou
zC|FquH|EJt=E-lq7SqloCA_>vnFW)S+Ah&yI1V0W2mmHP<p`Cc!(*tbPeBBxo>EgV
z4oU%5S~~A0$l2*X^tz&cc_gWx46YQHZz$(>%w^lb2rqvXP2sMP$TsymPI=qawS<T^
zM*IDGVYq#Ztt9Vyodb@Qy!(3sT|T>NQx3=F>BPg;K_;`r!%6Q@p64eIXU{n1(Av@<
zRP@dt1mL+NFh5ghBZnmffe@G#A`f}&xkLV5{oS@3pBZxc;_VvxcneO9hpY0z;_<U}
zG3AQ*Ir3tEt+~Pj|5RJ-p>TwdA`GepB1I+OjY_(eJ-z(4<}J+G16Co2VrBtNC#iXf
zC62ee`spG!EjxJD5hUau4K;INE}%a!B22s>4B)*!pnwzt{UL3F`&-YTK}y%rXuQmi
zxYs@riifD+$~U*MRcn-N$58plBx~@28XpXl)1%;G3tf<UH`1vhmKm=}c5uGnr9bPh
zV$uErjN56J(`nW|sxLH_<kWyb*<SZap-(AN<i)@p%X^WR38MQYXU>;-DEW*^pTtSX
zZ@_p~o~5B;%4k44!$Zo>%k^$RUdOkgTOyF&dD)BqZzZ-+JIwvgmjGAqLxB~nutNjr
zy#<13XZ|-L+Djg%;NThZZg1?uNT$fM_=CEe&8Mb36zRX`K4TUGfp89JO<hjmj6N^$
z+)C_FgVI<t>@g>sOQiW3^V%Q!<4J|Ggn(QagX`5IAi=5eGX6Gl`%<CAGBBTCi*|Vr
z1fDU^Zm@#CwzhW=#Gs*OZ7s&aUQ-V9Sy0N(esQ`nwb`$`g=aM;ni=bJ0K?ugZ6|-~
z1LJ2fcjyBxtB?w?<tGCK%J>HD8^siO!=y#>)K^7+E-yNi9QbME<qL0$MAaxrV*n|p
z3kIL2vnzG)gu!*54x|R+%qj<J5Gj<mtz?0qvlj#R1VKv5AVFIG*74fJIK<nRvKx%D
z!;ZA`G8G|)2MCm}t>29ChMOb^#cSrF3~>?$*nzJS0mAnzJA>{(^Q`1FSP+3Dwkd_=
zljeakH{#(ldCKd#r#nt1jC5_(v4B-fi+)0%U3!Au@PRcG5!kRRTyyUC7G{1JBu$v`
zS-brUv>Y2wnM3t=eMH6F=6^Ma;6xz`j67gT3c}OJI0v{uqIgBNfly50`92O(mdI&^
zo|>6I7W>HY*JrOMMVM4DkIE^S>%26LeaKyfh^xU4Qu-eeQ{iQ=olY4$B>?EgMPgi1
z<@cF}t(B?j7Z`SG+S&C{{$rEJk@eHJrKkBK@GCax6aGWRh>2Q#7jrYHm&Y8tHwdw?
z)znLck--jyVgaZ92MdrOHOF;#5VxXYc>N1vOzN~1LoFz}>tON7C4$m1KZ==UnR6Eb
zGL|{^cO%jSIe(DYe$W1Ye9rsa-ILrA4%lf%R0ff7V<Vh>(kd`pjmSS^ZR7EoFq6xQ
z#XPKewDpP98HTuTWn_>&@<m0y5bPY}+WT1YHelZBABzE@EUXGCOi{S8s>5gZubSMy
z;9(6+h$M5|;I|Uar=|6vpB6|7<Xw1v(&3WE-Y{&Wepru4AnI|8rZC;rQlJ!=s;!#y
z(0O%`JmW5gZ+bAhio;Ukj37{e850zoQ9+s6{_m}#@z-{P1KaywbyErmJ62^|Vrm1H
z0ndpM<`G$sp#33tRdPSbsw6(s%J=QglP^A=R!pKKgK{=~CFZl5N9<_LdbnYS&h7o{
z=zM?6^MEN2rl@MN+J=dodBnS~X{>sAP3>8V2wjS(P`_toNeJp$;Ae5-3m$;*Wl)x>
zzXYh(gk0v)C%u?mcZ%Sz{=U2JqGBcrpHC%BH<hA<(brD=tO-h~3<fe@Yp!k=#^DNn
zaxJS{+g7XKYkpzBYpGU+Ia}4f+-iC9XgGhdzSLP*RMLFiQE^nab!m}MC6H@3f7xg-
z<n#6+z}fgRfC9c~5BsuciOzUl(ca>tkpNK6gB;JV|DoLnE|T4!HM_W;eLQ$IX|c%D
z=6&3^6=nC%1)Mjx%Ovc;FWC_BUa-qxGFJLAzG!1`TW=|U9s8*k!hQKTH0!fN+zX>{
zqd3=|?`+;$ZWgXx&VPf~1JV~H50}F?C6W>#r&XO|uD<eW^pVW<;_$x6eh2%xI08D>
zRAg{<ZZe834y0Sk)4AYV(3VT~zTufFwq_-$%WWDu{ytkbUD!Ndxo}*I4Cf8(R{qXS
zXuOo_I%ofwwVxC)RiH`6-tNqItEeV|=!gU1LD7(M7l@ka(KgZ@@X0AQ3=R~8SlEI1
zd%e*eF(Q%>Lh1104?-}fEA4s^EeSvyFXqOtsa5{N{DWen8Xl`*$dPG;o}3BJ(LNfw
zhreW#G`I=d@UQ6a1{+^q+=J7C_Z~h8@)vV&?Sv>1w_=1e(ULh>rp>+DwNren`W&-?
zF`Ur}`Bt63^~k-~kO90C2L%7OOr%RB@Agv1Se%OiEOM6WuyoDJ+;fkPwiloSajdF{
zwzD>k(p@1Rv#J?G_h)wv+fg1xn9UQGx+M)A-{;<peJv$Iu8u|aB4xMUu4p>a@fQY_
z=E=dMY>uVTR;f7!fo6RUfz5#mzYH_(?;ePtUne}Q65}Crf%^NCe;f54mpi+oF(ls|
zy`$KFQo4uHXWgZl9n_<IQ#k>;W(w*y{pcr`Z@-k!Mqo<&A>YpL)=`C8&+aa!228YX
zmYKyq>g~+mGGgddh>YE<-z<8nC=IJjso0CAe$8illK4N7T9}kMRd@Cwt5J~!(mooG
zmXhuxkea2dJw10r)!jGBP_h&xveuT2M<xd;?psT)AwRPSCi_-~DK-NhgU7VOnbhDl
zIS13&{IP_)HnF4niMmWEOxY-nY5k?8lc39@C=hYycjGrLQAF&6ui9c`$siQ+Z&MP&
z6Bxv(Xgi|MU!^0*Kx~Cf0{s;x*hf-@(PL@KO33bt?(L}PF0=Zk+~<X%lPcy(Ndsxj
zM46zfRk?UsCH^Ypp^pRH2f?<La~x*|J|@zd+L#~-!O{${{h@t0KU$rcU3~G!)1^JN
zF|W2wAaIY7PD!N92@tx<TK1Miku{}ab06E-n6$MuP8EYIq`+gey49mt&WoDPZOi|x
zm*)qvA!Xt3!Il#PXHw-7c;v{ZLWL_@UqgcLmWN0DZWP?{nEnQc+<H>qe(Zn2op&Me
z-+M>{Ju4$5QW}m+|Bf3LI~*`q2y)l<&o+zvMgJJQ3VTKO9i5ADg#)6!C>}Zs3<gh3
zzDksdF*2ftrY;9FfX@!hHXF@{Hy2%lE!apl+-@frN2$6BfoMxYC$Ji=dGWQ^9%jc8
z8S-oY^J+xd<6{m$ELMDv^$uDnZU)~s=@YfIP;!pd6ed*=G#mY=aM&7K<I1&xqT7~z
z1*ip;lk@_9MVr5KA+)4~L(BP&-M<3Uginigd9alNH|!6MjsBS32OT*_FBD@jbPKkQ
zb0*QMuS$~v0mcXj$7ra9;G_1^)R{aI#)8-)Lahvryt*EMJ@34?P^%EQRP#H2x_~<G
zo_|}EUeBIJt&!o}Kkx4#L7xkD;IDutv4FW`6ffk;pgNO<7a3$!Gw9)H2B+$CwTdO+
z5Pm#I4JFpdQ^xeAAR&}1RLJAR4&?Rapj)5ZY9@u1n5%De)f=|B&Oh_e32Y%od!zap
zCD$XV6?FOl1TX`*6-c&VqF}>R%b9efjSr~i5oIHI?M*v81J33O=lv>m786s9j}L8R
zt)hhT)hqUm^h|TqR{<2ip#zGbYUY<po&pr1JW$h-tkWvX#3EWr8cra0c2%57QPMdx
zGO*y*{Zeuhv58=OPIMl~O@s>_Q$uy{>ETtxMgvFyG)*|HapyhExJ?Bc<Dvk5^UD>K
zA_0+lXuaNF8m_lhU!9ee7cn}@Z;%w497l<dNU8SUeG*8F!o+va_HU6Qk`?L_B=Hps
zm@x#9X7_=N64Y|S+3oBDZiLZ|-;9@0gK_VM!-{SR*FpqXav1Dt)?BdqIIkWYX49ZV
z_4zk|{u5mNncoa<%Lv&`C42=#twWI;2J>MDYK6T+=7hf}uw1+``CvjtB!_S*hj!U7
z`MontP|g1oC-adjOu|Zh(AfMTpF~vd6HG0fDqIZmM$@;`<;d++LqfJgUr9sltfk9a
zqvs9!`;VHY{B&|k7sY=9mc&wmATh&v&QC4@c-#vj@z9WgzdfE%%iGnJN1gS>K}a8l
zDN}4JnMsW9xIt2;7w}PDEl&dAn1QP|Ys%H!F$w<40~?tIOw7eYgK*3qo<a$-a;nj_
zh_v+5OUBH9nf6=P;^%%o-X)EBpnM-kPP(5#5lWp4BV+2@7;&bcQiX9GG}+MMlR&fp
zLNF?UoZWHMQbI8*jAM%$GxC~5Chgs!@UjHAo(bEw*9Ip^^KT;S&I-p|-6({hGU0&+
z(QJ*5w^8d^VvQo=BQ&rjS#aUeM0oOojPe5`iX@bA^NKc=VG<fS?&WG#h##UYThSyv
zt_hWo{3|#^)`<#^M@#mC==Gfu3C4B+;{9;&qFcloxb??lCG2g>Oxl{SoBm<l>ZYMM
zFjs)WT%VkJchAPDC2gU&7xEdKPo9&@*B>IJM?IiYEKIGI4Z7=a1Qaq5hx-}+FD{{8
zp&O8(tu2>9k4`RpoI+=FW5CMtf%EXln`rzW$je5_Zf*x@JbW=|9^!!VLJve4%`=(h
zeU_dxPJ9Brbm+f~nqu8D!U9hp1NKo-xxxrB%(QDD{n%6A4YuswLMdE=fp$4p4&fn%
zk9KtVc)ykq^mNb0h((5efYS7`3;%myfiN*A<PD&JFtMcUEkgXO5vWSp9<U*HpQycx
zgVw1Mm;Ix7vWR+6Aric=;$wz2hmOamB-Y>c-0)D0Ok<x85%a@`JiM9=pIP)ss`YFa
z-F-NFz8py`RnSPC^Q2T<*{F=Gq^U}jKP)wCfBnl-jmX&HFQOEK7-X0|7c!TVP+uUI
z_Cqqb_^h`2;OLwHHmkqop>0c?$xQgcwr@V&LOs@(^Hdi`DxMjxyEt8{^>?kaEcH8?
zYr5dZJPIPTcpeqFT3?>{|Ba`DWQtL$BFxHLsw|7=4jhY+7g}4COnI*xIXE{0SZQy7
zO!IuSCKNw=32?}UyJt)U^c=o4sKB~kX{Q!0yG!pqt4OT?Ot8X9Fzme>h?Mtu1UJ%)
zCNcrzT-^xKM-hfCnPz`{);m_*>u*adlQ{ZIPz`-kyW)zzI6=*!<)R+C4JGcSmNjuB
zceump`#leWOCu6|sl|;@*@SV9E@usT?W{YGL^)EsM{q~Athht<J4Ih<2{d15rxKrB
zdLbHRhp0ILKu)nLkxh87M(0|)Xkx`V2Wa&YpnapHA|95vRAPF}{zRZ&$CP~f{#RRw
z#)j$Pm)&5TWeiGQ`?Q!T8z$`@=*U7eJ&$CnzUDlpy6FDNIAo?b>qShvM(oeDHPVna
z{J1nHj5;K?s6hCt%D;fXEjQevfy;fM5wjVT6jVq67Nmc`VGCbSYk7Ll>E+zpKpK+h
zN~8OriV{HBPX|{-s@OHavAp<jJq^yqjWv$-+D}Aj!nJs!2SmL=ZebgNkf}y_D=DfV
z`2!gUqNR5ET96q8$4RIq+4QB>o$2}y#i&r!N7ADjqcI!43g~zb4x)P&G!!6kt=-`y
zmIAN<*kWX$c1wb_iub52ra&^)X`Tmp!Q!*_`M%Z4KfZSLdUNe9Q?&lvp}_SG<-0D-
z_%x?8#<Vnp$(MWz@6>*qmFH4<RLD5;&kOK=;*SYT<uE{bzriqyN6Ck4&;HaEy(8fd
zpCz~H2cM!waaYJwHWsnB7$Lj#0-uwqMA8WWMz$yWy4tsq9%LgGkrkUl&F@}6dqZ)m
zWfR@pz1DfYzxxd!m{~F7IPieX39;(d)?UKirGiMH*gy8gxb`${!`iXV7)H#WMMjlB
zrMx)rz>kq2*l)9?L+HV1R--SPr1=?3r0F|oRb11uJKV>$<Ip!FYTaA2+d2E_ws)fe
z?EO}qbBg_!X}9E=)@{pm9MF_5#eFH^4`9Hs#2yBfSRdPL7)rj=SKNo;_~1wbn227s
zSb0qPi9+%;oc=@3FTX?4B+RPk5UKj-01ii!RbTkf5N+i~jT#lNAVR&Tpm%@ED)um1
zjm_NGnCdmjni6jzDSr-B|Bl@k#t#t#92nv<UsZ9LKUYaF+K}fe<4C=6oz#kDF&n~G
zQmzh?ixgKRl2#(H0NeFW1rJn|^Y>f^9O<n!X+Gy}3}e`k&OGl(GTyH8G>KY)88JGY
z0?{3etTE48i*EjJLpML*n#QbOBN4|x&djq+3gwz2$9>*8v=t*1^DPlTiEm#35Cv|b
z;$A=#jY3kpOAdq&ZsNq_?wlD>T3RlEs&!{d*fROg`1Wreu@L6Z0dyFI68**X_Ss3I
zRgNfa<!^uE23X59aqRRH{(fr83X%oVVj7wKN#(bsN&1bBAE1p&*z^AxkJ}jiXCuhU
z`fo1yf7XHtl<fp5v7cbDpp2Z1|7}r>e1ahW>c!E=de!wz*b``CfdbN?VvY1@)*Wi9
zM9x(Um$knCk<j8->b>^FxBK)6thUR?1`O_M!xN25BBA>zW59l0@Pz+Ss28ugPmGI)
zX3XCzHN#V^P>f|M&FrejO=6vn0YPCx;<Zy;-q+PrO=fxzH@k--5dBRcK1@Zu)s`9u
z5ClXQk%=AAa!KnT=H?`RhT>W<kscxQ&74jm31H(Sirp9qsUG?h@5;kF5?vG!6B`xc
zH4vL4&@^E(kg&9!{sFe56!{|)tMMBpz=2E;z0{dyQRGMq{5TLB7@s&-1QrQ7ov9Yg
z={#zZh1868SOG}+M#Exu5MNSS=S`#+5HRO0lpuD)hXV__^?um^8wxw@P!9tB0FG`X
zPmHyy=*#9i#uUhSC6GS=3ziE0oi~OBIRTa_HdF5akBG|vaYKkp@_T83*szzdb#0bn
zlmts^SL6e!AcQ8|H<wwbk*pobxNbFwB|>oY6pPe}2+*6MQ8&c&>v1SL_aqY);8_=y
ziPnImU%+8dg~{u<o*J-kMT!!8NBkS>6w@UypQsKd1Xd~`h&7$VfZ7DAs(};@>Iyax
zDS!%#EIi&SJ~Gm3HpX>S;DIqVGF#awmCKNf(I9Nq1Faz`J3*2xbf|A1&SZ?O!YI!$
z1C*)O<_|_#q%;QWVNtNM!)d@6paLwzm!bP!T`RacLC)RDg{|C2@dgxqtdza{{iNr+
zDi|-`xnFwy$;yTAeJ@1tT4m+={8Cz_1}KAJ{C0@CG{e_c>2`kK<yUI+PGfGw`BQyH
zr?tnEj!F%{qo~|<szhh&@%#EAUr^Y4DYBe^QE}m!I<?o=Qn|}qb<2YjU|tx`rUI~~
zD|wo@PVlfzTs!4i;0!KHc~zCR{{50~&F%E$lZ(<v>{Ksdo^aos&Hm&{2)7ec-f|J)
zL-Aw1{_U#Vb(x$p<$$@KsRpR>srC^8#><3Rg)%wgFei#L>m07{J#4Inpb^1AA$m5D
zKmM!oRphxsNi*P@V7_w-Xz$XJrLVz`4cA~JD4rU*5%r$&A0scIUG|8eOw9$Yyr(v-
z7uFAf!V$$`z%?y^S`=6_*9X63%DVr`Js-+9r!UEQoG@LC8e0#ru9d~7y1$GqPi$m=
z3^g%vmI?keUZBZFf)QJlZS4a|g7th6XQ&@o+gjgCH4f#{r@p!Z+=6n({}wQn7lT5%
zwEpJR?1hq^&Kd@Li-p>+lLqR>b@3_`bNLjWFtSai>=ltoE+Dhylnh|!I`MO_TiJS>
z8b^T9?c(6Z`BSPR{cgUZYFh#5eAD>u?)d)zQ$Vc0?)$~+PgV8t<3~N|+N))=TIs9q
zLv=Bo%_?hie?8vSSF5J2`(}T<X@W=b{R39_H2y?dj`Hkvz|rRkq!5A}M6f#g`sfGq
zONC~d_RN&oY72IKRL*0%@V%I#7N@2}7G9Y)rLtsnOhuwTfb-xBmdL_eKw2ovMyE0e
z4mN7wK$OS|MT<{VBuj7|3T16%9ZF>bX$x{HlC?NO%&G265{L6N(n^2sfV7f5mbBu=
zN0azqLteXdM`m%`zAL9~d0QXzw!Ev~x;-hMM1t_pHG6oOF6tq|+L2(udG3!O!8tUu
zo`Xr|)n(n+<JDx^w3DX#-@0!5>8u$qn}JI3_4wCiHy_WYZBsAD^X{^lt;T&f?w3t7
zWaJ<m&(PRT+pBT?D*}J5bP*^{+^M7~6rbQ^<Gci!fPkFW$=E;KY^}%Eml!L;bHwU8
zl#N7?sP;#k5pYu$x?5finQ>o#+DxQ60im_pNpS3}&R7YT8Fq+>(0iSaIfSpK{q_1{
zSdn?%uK&}mn#J{pYVo=6mXqsW`ubu$tCz<Ec3>$hu@T%ki4uQz8wSxbyas;~;Md)3
zI{EzP_6kN)Ervof6vpe-sU}a+=_L0bF=`Jop~N`v@;8XrF_7UX0s{@VKwAY9i|z84
z1HKJRZ_!V?YCfGTyXkzf?wjgl<+`Am(r_eZxnuE3hxCm>-TAu~mcW5;6l-A#@EVWv
z7+0%$a27YntIL0G0w=Mrdzi-pxz+TSdbO+&*0__EYy<ps3lalfxtFt1d<y!8ao=d1
zEF$#ZA`I)C&cXej$WK<bfpZt+o-*KsgtBc$m=V7cWbq6WmBkXb<+dJV#PKc{x+SvA
z@OT}U$(l`CA{!WS8O1VLVjP00NY=~lz0Z@rPZqQ!HxPd%vM$2`^ks8%*omf4XAXuW
zE>?*waYCe0HaCe>GMCMbfPqYzJy-;H!I#a=8U$>aiXxR;@yK4Ja%^yt=^!}CrYtzg
z1E$+^%eHK<-IjZ5uiYcd@H_cFySYym4q!}W7*MBusVpK<IjX;K&5m3qHU}}Jyf}Qx
zGTn+H^EQ7r%-fhSZ)3r{i2>y)<QcEPJzVoBSG#a#l6bmcKM(NC2%emM_#*Q|-L59r
z9cF<|=etGI&S%eeyNXi?hLL`Q<>Y#mzCNt|i*RcMhTJ(JE`ZTUU}NO@D-RrTFDwKj
zvqp#K1OX$@@?7cRJHzp?1{+QuE8*g0`v?erMi75;lZI2rrU;h-EJ;brQ<q$d=SLvK
zJD1+_3|PWS`*QMlXP%T{xHFFPlX8LK{?OB7x{NvAa~F13j^Iwoi}nNTvv-~K(Cz%O
zRAlWT73CvO40$arTAMw&2iE3}Opq(7l%JHN?BQ9slkN$R)@3}<r42{Yx;z~Z^u5I-
z566Ee(fFP1h?tb+Ioly74MBbq0ml^liGwg0_sE7!&&LTK1eAv7f1|zk@Jw;K!#j2$
zD)%TmJ|hUq2pHuT8xrusg%&XKV*tv-1JENN@?C1OTxkAWVE+3u^1vZKEqF>!BD0+M
zNjZw)(4l>ra*o{@bm;JqJ(gvt@C2Yc?CuXN{QYBI5K4dZ@PB7>L@}3<WCs%gGm}wu
z6t~?61z`aJIFq5n6qi#81>6EOGnX+61=s;NlcB>DmmUiR-j{;P1r-xAGdDO2FHB`_
zXLM*XATc*KGcuRZ6b2OpIWRIXlc5JEe~kA9P#jv;fQ#a8!JWq4Ex5Z&fZz=@?(QzZ
z-5r9v6I_COaJL`{?hcnTXJ&Hdf3ND^q6+$3zSdrQcR@+2ti~u}VrK-Bw6k?)WMyLE
z1&Aq#YO%5bSXkJZSXkH)C@Iw~oUK9skt0xQfE=AH>}+}eVIk%S0y@9T#DUK5e})Qn
zwg6cdYXB=dfR&4vm7AA^1;EC_!t)<PJ4arCIMCI?1fakKkhQY~IU!Jr+1YzIT9}zT
zzvubSM*xj6Er6AWhnwMF?f?-RkfVh$&=#NobT$Xsyk|59S_9PVj4eRU9{&@9hTq)T
z*`Ak~+0D(3325WQWanrmNXr0lf3t8l2dILaK#r~;6Tn|B0~CQapntc<gg^;UH@9&5
zTcKuW>g)z|1OeU!))vMfTc>vy7h4mMBj7zdKuty-pkxoS{o7dnZvzIvzlQ^0Wn%qr
zxPN>970AN&U(P^dV>=srpsk06tr@`7!Wslnl9Xq1c6Vj~0BueFG6Y&Xf7!kJ16_d@
z)<C0ogMXC{1W1ag0D$iu{<}XXV@C^nXD22n3+umnWd1A6`<NwcO~mYMY(TcoP6&V1
zCvM>gGJc=C2lKxV*UHw;&DQHbWNKk+V)|DbCNB2OpKL80TtG77|F(G-A^ee<ft&%H
zEG#VCTs#1f0|4Z1Y|i{we*$$6d(gj>tbd8$d+_nHx3dSBzPADLu`mU_|3UC_0=j|#
z&W<i1AFqEZ{+AF~Spg;%#?AmEkeP)o!k_5xVvy<I{C@t97VZEYmiP5z1+e_}^WRT;
z?;B=fXKU^8$NXPM%q%4%CodyT|L>Ons}vQra|d`aa<T&$**IANf2^$RJOFM^7J$$H
z#!&`Z{JRd8KfW@yrgi|Hzsr4}(|<~K{r3*g{QH8?0{%CaqTT!Ef&ess7G0Nxlg0S`
z7wiA$b^n*k|GzE&E6V@tqW|BDBwehn|D~t-+u;94545qc_V~BO`{uehzuy7{yZ5_b
z`+rR}L4RLeQ9EmsfB&nKaR$EM1rb{_>;I0?!b#G?9b}?x;cRUF_qzNo|Mb^Iv$n7W
zDcd<&{PnZ|7+G0Z{*Uhcb{Si}KOau-JNYja=>78iZ%PSUV>^?-ZW$XV7Xaw!2=qW;
zd0!<qPELRq>-%*y0lEKcivi3`wsy|%E`ayud;q3)jtGC9e<&9xfLY`((ci=cU>5yD
z+yG{=KlBm6EdGah0L&8qC2ke~v*aIQeP6~u#0FrN{zL2lW|=?qo?Y$_y=RyIL+{xY
z{?L1N#Xt0(UFpB{@jbipA9~NO@`v8DtNx+)>}r3A1Hi2Qhu&-X^oQPS(fC8}wP^mA
zc-{&4hu$kQfBHl3l^Oq`_YB4sj>axFrq=K0@E<YfyNj`%_4}FsN5b)!;roO3C+uI#
zz-;mld~XZ%C&GJqfA8TRgZEBN|KNK-Q&WpS{Fn9*zP|?V==zU<f9dU99RCsT-Nfu4
z_#V*wPrdKMGWW1I2ig9^;$8NSge>n<vib+U_iX(Se|(>l%^zfaZ}v|p&UaVa_x1mW
z{=Iv<KiS`%?fysoUcLPv^?NY;_b)bEYmn)m0dcVYTk7~fbK`iQr2YFlWcQC*zkjbY
zJNyIRTmQ#kSl|0}`cvy)1akeyAUNO6oh;n{;qac~Pr~;%klESX5%iDQzt`*RX7>+=
z_x@e}e}V6#bNvUtcklL(gMIh@M>4ke@E-rb_sMwvY4$ymC&=;dRR8mi7`r&WkH-05
zFa7%?^B?}#S2qac4l+hqUa&Lf3%0Be?zpcNA#h{dn-rL(+|x{>W%OEh>~wiVfla2Z
z$qL+bycJ0v=|fxFm7uv1ULkq$I{MZK8`hSnf71Ts{cI4gI<wb~ur!T5IQ8{Nq+x&v
zo{&*pc;EZQ!CNE13i2Cxw=89zgUd%0<pPX*x9?K!4Flz?<Dm<CD*H8Da){66KNzDw
zMd=2V%u(hWWzXZ1!8kJ#!O@|gyQ9wC+|HusPQ8)H#?vGC97nSkd1-I4#XQV=u2-tF
ze>wHwQQ&Fg5<%XekIztfi5{lN;w*ZV7RSnlMi!eA=`XYhcA^3uIO<5dm-)-brnbj$
ze;EXbm80&?gvJ`H>`;o1{IXWd?ZQRRYv3;V@xp3Qh~T8sb?}1i8Cf-#<rpM1XytsO
z%odj3pI}WKd_yAy({&hjn6k*VgjX;)f57AWgSs){%}+RfiOneAS*O<v=#G5S1;KrP
z;WITQWeb<1bFKdQL{NPijBtb;ILc~Hn#%`cc#@+zJXnqXW+eztP!>G{A!j*Ww<+80
ziSpL3m~&(tB($Q>zyBbN1hKW~puXo4ykt_Y1Dc7+u;i8PwS^j!dx{`uQ!A*|e@QwT
z$QCt17aFfr01vbauIH{P1p>dOgcnfbw}zQrm)t26@lH0*l>PXE_{*e5)Fi=sMy7oF
zG(=k8?TZ3xb;T5jTwMH&HM@dcGqJEIoN0Xes#}%`n`^Opj)*@^9ZzB8n5;=K*b$co
zuTn8^2rDNAX(xz0Q?NzbviH`!e_*7EKJC-1Tg}_WF2ZvoMcmK+=^fo|O+_ZLA{vb!
zx@U^}`-FL&eXZ;tP@v~Tgf6WO(42b96hf6-?#dCnVmwUMJW970x{bCzO!OP_xDeUL
z4Bjg549C?fx;QS-$_u!dm&MEn{umaYT+VOnRi2B~FY&=`X(fT~>6t_Lf5})n>G_Fn
zVwIS^^1%0YSGvdzC&+V8g%@Ie1h$<7Wnp`QDQO1F*uAYByU-OkwH%)&BM_Zmn?HhU
zDT7mPbuOa;e3stx`GZtvJ_|T)kDue`%ESoX#Tv{J;{tGOQRK3GZW)dFfEVb6y*$mg
zZ<(rk$W2*1-=Vd&Gl^`xe~oeMg2U?K0!9L5bZtTv!9sm))02Mhogm%j;j!rAy_m`)
zWjaDFEkGrA1Soxf^MzYkpE!xhtm;7phvWkfrWvJP*(V7|?KQ0nLVjj0GS*BD5>Fd?
z!7Kd;9yI!>I!L37S{mFoiBSOcefdXs#xsTeIYuJsG#Sy_6;DHye=0zLTl><dENd-J
zkN(IL6oRV(tCKzoj}K$h#Mni*x!beFQH)8koTMSH%~+GtTb(|qxgwWh-&BKiTF}~@
z@2l=|S9HFhbkx9#6<<mE5J2b)%%tMx_!@Ko9rbJNNB00MgFqYoSh<EIj`2SG0DJ8i
z-q1&@0Yvj-1LzYNf9rOv0})%exJ_WjUQ=Mptzyp*@?1CXd47pbI>kh;3GcLpG1fx5
z<%`&p0^3i+aZbl-`PeJoLfoGXUxy#Js=6k8w_c6&&=(a9;Y+u|F-T2$3<Q;b*Z9Cm
zel5GL5`<pbPs2!VPJ`Wp<4Eiv`sfLQ_yk8Ui+ck7fM@jqe-DXYkU`qFTEAi#{F7jf
z?+5e^Us@<0oD`{CMofcX`l-1vm{hj=jyRhPi~WbmY*sMGy5zFs>j6RSiqr88p{baO
zk2}iR;5LVGFyuBMapmiojM6V$Z3OX24?M-L0_GAu23j<O9&w?H_i3;+JoqICI3}D6
z`~@3Y9OcmMf1DVkvFAuyVwo_t62!8%F%?}cXu3;tdRqCIxNRxKjn$dY%hq(QbK?ZO
zp9z~71qtm(xNnnN<=dVbhnBQ+hQxduhOBQMZYm?t+fCA0zfSNn2Rs7INEtAP5IWa<
zqVrKyCh5;%dDu0)0A0x{(bfw`Qq@h(JILH%q<kb0f8ZVt^xOLt7sx+k8<YJ+j-X-S
z_g(KSnH{I+SF(b%iImZ3IciN8=#}iX48Az~0ytB2VpX*%_|5agHr&!WNiLo^ltnO-
z+{}%-{iGk*EhKz`1DkYEi)@ZP?~+xq?kR|7pCNT5@;JOOLcZI3sxoUn{6b-bZx1L1
z8puHJe>WxCm?xHlOfRWBTKx}n95P!<;Fz)(#8_MiKNZ{&9d}Zd@2xV3%)q-blF@TC
zP;3+l{`mGaO@O=_BrHd#3DF@DUgHd9Fv2L?nKXoGL^N3#R(Pdt5_?`|9+nbU9~t{y
z&|UGQtYE`_H`WQVIQyHFy#KONk-6Yi?yOz(fA#E9UTHRd57>$<-?=Qkt_{K@385_B
z6Bq7ipA{ZP%z{k-4;hRi62MUWN3TxjIpENo%gsv!)tFykquSLca~xF}Osr@V?gw3j
zx0A?X>anTZApLa}D``MgbV;I`?ZelBU0?EOKZ6f+p?X9)O}ZB6q*xEK6Bb#c4r6IH
zf3g6%27qa2-WLx#wFht%pWAQ?DGe?+U<)|THib=ns4G93r?v`#zWn?Fi{fuphm2d&
zRXXkw7kQ<QNdY6j3E$IagLPT>7qLJYA%4z_B4|W)Bu}T3#uFbGdBYTiCI9j+GVUC?
zu31FL)4PNA%b%g^A3c-+kzSXVabLaje_Rb|iTjwuLkS7l^OyIN<tUQzFbfai(>nkR
z?0)qbDL>Yo`m@Dx+P7v^-@+4%`qYt|{l|^#U@O>;<AW#zo;}2M6&gRN@klS7t3-Fq
zXFHl+OO@z9>vD32C6n(iyJG2te4c8g#X7*2`>9r>PuS>d`z^gn0blEh;UiTce>V_$
z(1hYObv+qT-<eqxJz&XWs@t#oJ8lWwRV7EOikuhivcLpbh^BaliD0`OKIr#~cBcJu
zGQLGHBM4PkIYIu-DAv33mfRkijH_EN8bzK&S!-<5(+H(FS3A`B$DPkxzzHoSj{K6G
zsS+FNdUe3g7Jx=^uJKb`4Pztdf34HeFE{Ga1xdByn@5zol{L0t8}(Mb^vNvJ2G2|&
zV-oE{_mfK<LXQhQrgfvq%}<V|U#*^H#_~>KiP-n*vM@Tl(3P%}_|GBxQ9_#GiuaKU
zSafa3wTjRdXQ%R}SEiD4%q2pIg<#`Jc4(912EYl@yfx$<(cjK`H=cpIf1Nm)(3KlW
zs?6z*9AA$PkFXUUOH;7t!7>*m-Ta2nK%CSudY68U+C9n>#P-ABF^h}M#f~*X!^&~M
z;1SeG`9x|)NP+=`DPyb;<&kT5>{32O>fm?>)EO)Jv9>+r!}@{{u|M(_sPh;zTcP|n
z{MRJ!yWA9%)Z=Qzon+I-e^dzC8XtRuykOy<0y@<<x>~N;sx|Q9s}-F#f5bdpqWX1F
ziUzHiM_sN`5H2!oUkT>!hv8q@{gO*vA(PIYszI4U%>%SnuuY1hm42Gwrk1UvX+CQw
z8p@oxtTJX*mRXP6zGJe^Q<(#|9Fe57%q>-JL|{E__cUS~sjk!Xf8S7`w~=cJ{x&#d
z%0RZp%<fzGy@v1UHif%@nY*`2V9bv3K$X%zVt>mvg-}+snNZkvAN-9=gymPCc@L(l
zBu{Ow`T~+y98CdL+BwB$NSq(<*tK~#{@vS6tiqdR&jD;{F47XgF=d=21dc}V+sYTI
zonhCOKo}uIHQucwe_mOTz3Sk2^9bs_-wsb^1z1$x*gF5`J008M)aI(|p$&+bBRI_2
zt*-5bppU;P9=plOKycFn=<`-Yh(s5r8>YeeDX~$xV~Ex0Wr2-I%5DS_O*|zxQxuBr
z!7Ka)FQso^$Sk#>+qIk^UlbhimwrpsG#+Qq%RiGHk{;8Uf8x?-l7fE{36Gx|g+O8u
zVn^@Ub>~jG;36mJegl$TMF@P^B_GCkdO;jX9&stci8Lx`Wlr3b>TXyg2S_X5>6Yml
z8g1{+Q`|ZVbkR--#W5G--|@uoXWC(ienR_ww9H-2l*d`Ni!0i^t11IXZz$pclHqEQ
zV(d~8xqo0af7>rPs&q)RN)|{7J&s}oXtQ>NM%Q$#OY59fXZ(}|Lbj>73jZjy_|@!o
z@WX@r#U$emqHpXh6phkpbkjFe7ZUUkem|>66LYqLz`J(|WNZ5>G=s6W<`t5+w=0#q
zkOL=N&pViJUHLfR1l<34xZEYx_Cq7WwEbH4Lr4Bvf7embxC}VI;kVNmMxoOiJwC>(
zBWxHcA2kYMpf}RpwjI)9vy}_=nOi2>OHHar0GQ<OwBQZ<{zf)p%x{U`ePrl18gWF$
zrUX5NGL#HPz@a-hQqiAzcarOjBo6X4^nzKNe`uUP7%>Eqg~{1TvX~gf0rA_8@!rOB
zH%vLCfAq&ItrBRw4*Dl}*AU>+n_f5c7(ucvOGfy~W5xmVLRfXV&*SDJcTV)^lLdYe
zkAntHPMJA_YV|nTheQlO=TZ*~@VbkcamUwhl#Q2??0C;+Xhpe28<A863O-A#rNivG
z;#R}vhS9pYjxHvGn-_d{7T-h&47yIku@LMee+_2Rjje7JGv^}jNo)-Z9$go)=4sh7
zl2_9ceo~O1Aagy;c)gy%Jy#s;R5$BNy54*`)+9pg_P0Jopmc}UF>9PdI|;Gm@-0Fa
zL!84}BTGoox3U(iKl4CsTcad=C83?Bv~AX$m&{UP2E$Z9OZTYq8^N4A@V(M@=sYP~
ze<87uDF_LmD9#ldQ_rMx4=Nwfdn!Ljp}i)t$OQ*?=FrYe(Eyv%v)Ky|(3k~_Ur~BW
z4DheM>7X?Ovv~c6a7oDsA^T}CPsg-YU5k5;nF&9{I1D>70PYji7lu1??z4{y1sqeq
z3E@2td3bln8@4T9Z5&;1GL^4m$dLsTfAZjr{M|3xcMa~^i^t^qTfgTbdQ^m#CkEAC
zPjRBos0uTCnI5T#UHDX(jP}3{FW&kt?K(F6BCUbsXn)C_iMwf!GF2|LYdP`nR+y#N
zr)Ee)Ips27Ue)w6Uz!*>8(|Kg$@o2PdC;Kd^W*hIHp~-o#wj#nzWHaw_JSo2e|+gi
zgRf_%yqANBE|}^Ma|+r>UQb0eS^0GjLQDOe?@2=mL`~pvgM`GiKh&pvr{%OEMX(-~
zS)gMBVPuuDeQ5-Jd3LD!i#A+)cl6)L;DVsLXgkBSI1dN}%pnzj1OgS&mg>tWqgkPM
z7F=Oi*UVgn4*WT|8n^?bZ{x9#e@I`iniQ{JdMyiflsJe=J298LjMm@wzJ7}acg~l*
z^tjw_*t@x)I`g`6Q_-Q<j*v|#aA-xK|F$iRc(!X}%dlywU@KuNgJgqPswzcP^9ib|
z@iR)z$G(hv#Som*!nbTWv<O}Tz3*Q|Ny`Ut6ovd{u_De6s24?;ZmoK}fB2P$`%;DW
zD42~3b}5Zj8v{bOZ97Pz=z?}HGxB*bF+WC-tSO?p7^r9e&R}guBB45^g9d$7)Tb(O
zk@3EH2&R8qU@2=NL*zvSk2AqhgFUaR=;NyX<aVP(Cb9VBM>KtYc_bhi+s{8pl7~8@
zLl)xBf(iCv_XRQiM%&9Pe=xfLdV_>17#Hc}<Jb^&vo?HV?}P!6(#!<ex7{z!X!lS~
zbFSj(Q=Vt&CHQSY<fvkHXu&Om3}6~`Y5{)$VJ+DFd{DJ*yjvLB15M9tEt)suiwWr&
za>uxzhuEbEou!F|p=sRi7(0#eD<m8zO31HIgTcffT$h5cxYC(We_geQ5>w%0ulTWF
z>$H5X(`T=6@kVkE#QPv|FJPIjC0`4LZ4sKuXfE}{gq#FY3blRW9ujbt_;$@e>;n!s
z>w*ai%Jb+mP(lJ_jf1Sam|y4jNTm-nw?8-KQd|0gzZEl^OtV!ubybNQq!ItPXfpb(
z_YmLbop{9RZtEUNf3Kx-2CUnx{URO&3~C0!AnPN-LTv{J;aWrNa9m8W$l6l|6fV3K
z@5@hkhxIyaLuFp~t>b(9Ue^W<gc0w{ZJphr(?fHUwkA=z8xogA)hI{G#Nzymwz4P=
z&Gh_ce(s7p879Vh#x_I2ZBL6o3(<Q7x%)o9^?bt(myNuAf0WmIY`TMMBY%-Y`k}UH
z&dTw^S0)%tBBsdZdhQ~_wJN;Eh@np>!@WE(V0#$tbBJ(D+Lj}sJWfjw^&<)U7_VWG
z2)9_QuNJ+*ep;6*xztY)jqlES5#k#r$X$)3@Cz*dJ9Ggqo6xJC*G4~aY6H`1363fp
zKhK;L*DFTUf95t*Ud`Z{!GT>4<A2hkGN1Uu6}P0+b#V2#HK#erR;Ry$dmjgm;we>g
z@(i$0%=<A}@$%wf+G(KnMz$>Y#AIf8i!aJ0b%5|Pd?cDo{5));bsE=%c@y(ZOTq`N
zF6$&DWFMr7-fnQ1Yi+GkhQaj$V_{bs9FQQ%-w=n-e?FJ~tPT0PzN1Kqy`ci5?>33M
z9u(Q{l}RO06~$X%B>+GT|3z?-K`2^U*COLxl#Gz;JHL`yV=7l>VdY)~h&}b}wP|Vv
zAkMNsE;snvoLC{K7BUw3wd^{!OV=GarL2UnJZqQiRhBbB>Am*q;S}(bHe;F7jsyC=
zGfj1*f3eD~*FxD>14eZ=!#oF32ABabC1*jOUrO3j7!FNl!gQ~?x$L6TPHN_sdJezg
zOZ0sNlbY8-`+<RC+F0Tt?9Z*GlZS*B2@d<q1YMJwt6f&4^hxSQmiY{k!dHwahO}|V
zCl7hoA4);40a2tyoA)*kud5e1KHkm;bG%woe<go|H&QJMJF!LCTn!d`a(~73+e^l$
zLW>6Qj}^EpXN_EX`gcj^GUNRnm5D(vs3CM2s{Fm1mpjd5XS__1N4}Hk_i$8IPoc|A
z>hMWm@&t-J4N|@W^uWnZU_SHXVrV-}Bf=X(H{|H-n<G!TdF#~hKG9C0ZRZb4L^)sn
zf3-39v>g`@pd$H&iZRZ|Y1;>i6di-&z`m-sSGUmcPddE{CI=@+DqZQX8>Wrnu+vR$
zyzA@aV<=R5hYD?cnRG2&@i)&q`U-tNJf5UXzTL<ZmT-~QQxeN<*l)!%WbE47_hJ*v
z`)Z2QdpnJqXi9Na*L$$S8dYGzYH>ize@PY9lYc+uo;hHjXLQ6yLyPAivqhd1Ns+x;
zMkjzt-X_7~WAX)$6_=X_A6v5v)GnUpD|9n*zv?)`4kCRH?$czpN$wl%9wbPW*y><)
zKEuWo!*SI^V2@OoKd0B{x?7sMGtU|5B5hRjli8A2Z!N|Nd}Q+oWHg7vfGwY^e_?8X
z`l4r4?dr(axy>1VUANQb85>8?_Dkt!R?{~dw2tW`_FM-VdL<$@#M0}i;Pl07+M`T8
z%%^ZC<djIxG$?RXT6rVufD8qF9c!#s|A~Z<e&A2+E+`V1!tXxJ`>R?wup@2{*1ODP
zqT$s((1<ii&1$aB`^lFB`WhlOf8~e3J<94cy_#=~jCzV8!<Y+a&2h+@JoaY&dN%TT
zR4<i{CtyW0{=W?i^>evDQD=*V+lNU7_})F3`?0hB9^0Nhr&9GsA%o=WecSh*A-O1V
zA^Glr84RmkMIoXXQ1ZhQ&ruvJ?uhbw`Au5`LT@dyC@`OdLMx9lMIq8Oe^i$_12>_H
zu0b`%FNDy=v@HNz2G=~oFZlDxS*(wG4(V6cW?H#LWhUeiK?O{kqnR{Ecc>2Bi=Dc=
zhu&Z>%JX9v1H4c)`}%n(9U?S@rVcF5LSIAnBUVBC%@g2JmF1I(o%-a^?{umqy{&FI
zS)Jk6+I5{nf8n8SDsst>e-^v&`aGG<{E1QE!EQ6KAo_g966=;u(R`DvOjnFE!JndK
zuyJhm+r&C#IM`RXmW#l_nzZ_rDUg#@&?^TE3`UhuAnJ9M;o_oQP9geCyuAKAC_Mc`
zD56r(^fIt!$m{?@b{AIVO2$dY`ugUJhfLrX63^$U4fpTM%9Tpmf7rW<5}$Eg9?URd
zM4n-ywwzP-2<H!wp@=^7GIO}4UcLG$K!>J5?FQS><qCOV@F->khe32pW>RoY%`<UI
z*{(mOcdpakzxAG5?C7i%N`E^*@#kF-kSh84aC0t=6Io=I^6R^Z>)Y^5A=3`F)kb3z
z8SL=|lJxgiM7pcZf2dbPVC5`4M!fo2QONA2Q*z{*?xG2bU5@hLLx!^@TVe(@li6jQ
zGI3Wo{!fP$a-;O;yV0vI+!18AD2PoAggY4QXyn(gG#?n|$=epl_&4;jKMy(#yQ+>1
zu<yh4s8PX_X_nDd(5a94<7c}I>VueSMl?(9)TfEXpNJV0f1pR4cW-)ZNr?H*l?vMm
z)!d+>EvsAN$eB*2)*WDY_^8;Wu6(*e#4V}AWVM^dHE!1QR?DoetQJ=LgKq;YKTtmY
zE`khj#w^k>HLz7{tHI*7u!}LunDGs>hqV=Kv-0Y}*0!oxiy4X-l&BFfQrI&r2?WN<
zkQ{mX>07`Qe;c&&WS@P<%9RLrUn|?%OG4&b7N)9DD8=g$cOtN91#>$}4Jpa7-t0VK
zbgoC!HQ5?Ctu7z<0Lq+d?ED4?%NS?>fV>qEDVgkil*&8KcS7aFBFxsB%d%yc<<Pq^
zI1u$Pd${sDt$(PH)Ix7s;ifu4@>E1u2a!vz*R`>3e`K~QPG8sd#;_twU<v&3iFd~Q
znBlhg{HVZ}hGL%oIMy506?nZJ(ccL3a0#}06|VK+vuf-{@`}pG@@}4)d_eYYaeVob
zRLP~A%LMFM?pJ;%BV`?8uPVzq(!`ZWDr>2;376>&2(F1$B!VSWx2dXRh+=pu!tiOZ
zrbji*e`xk^lqeJTbmv#jUZO%XUE(BfN%V3Z*<Qvn#V$e!;?f~>ByrW-T}dxrcF4?x
z??t989~nvPgv6U=ccv{0mq;JT^{<3noei^OFdKhfvw<A-6TXveMBle-n$n4<F?3))
z&a;GQSRpC^JnrL<GF=UOl<EPJByau7qTn<7e-Fvs<sMSOXOq7ot6C1+k|1mzK%ci5
z^bZGtddjtEW>CELuZrb(1;2eXUT~hR^9@uIEa}+jlLXjA7`7uxlg5)D*o0swsv289
zqBw?)=cJG5AjjIxCM31=b=#Mhxv6jaJc+PjY@no@0^l5q+okCDX45c>QR>{O$dpxT
ze<;6~vU)A;Y!2z7Vj6ULJZknyM<RA>tRZbTxt@x)upWI^zTH-1F+4tM`_5cCJ34<<
zkM|h_;G#V9g1=dq+o*4IVuq@}in!v8vTWQOj6$3@q?3qN<sLeacvHemHzngr%XvXD
z`z~BF%OAu=JEYk#=zg*;=!yFC(P1Vvf24FYBii?<!&;;Cr!>~_fQt;J6I&`}X%dv1
z<t=53IMlD)omp<E?w}*26~V_yN+MI_2}1NQy9WW5Ri7$EkM=N1Xl{LeI)c{e)T8<9
z%eLVJ+7W!@2RxJ0w^R_#thHm(n~Nz!i5O|xR%r3Nzk+v>^IL0PMgU*ufBGgce|na*
zYlV4CJF~*|Vh=t^@2S?*_n9~6rd+k-lQV1-9(PhYc+<peaj<73*PI5P$g|dE>7iI$
zm`dsMgA>6g<v}h%nw)AnYZ+sm#(jhmD?KPL(&-vs#hjqH4y!&r$6F}iDmJk-y7=xz
zjk)_JK>a?A^W!h2&h3w;!4-Xof4|K83V-(*Vqb(=QAw4%@7d@92Fy^vlPsxYXA^1X
zL4Dj|gDfi-!(Gy&Wt`FURWu}px2K{0nDb&N0j_J{$9DMQXYRdbc&N+2!Uf`joF0{?
zpQ{!8etan9dQxW!INSN$Nbm*-uQc7n-qG7s?yh|>xcSaI24l4EL#0>ge?(ohelb80
zV{0^HIMPe%iX+}iZ+-OYiK|`&U3jz!*;Nio^wxfgCLcuI;>@31$4vO#1CsZJ43GOV
zwxu5u$rgq@UU~iFD66`O2!b$h?>QW(3W_j~cOA3t3nM1@u7{(0_JKo>WhS3{IlqLb
z39O;{Iq2oGavnOBTQkA1f7F(5mTGcKyQOfe@v|1*F66$<Pu$@!EfoG3A8MbQLjsX%
zG;$9pOXNrk)J*9^Z+|tTCA??^2*Oftbga(u_DYFtVjMS|QOJuduGw!`ra~c>I2SoQ
z!Z56u)1n$1+P1p(U3c<1N*H^g={A<uPP^yrJo^2#2X=WXqh^V}e=F~?%@NMpadP;d
zk73qC^W1I5;nW4#T3Mbg`8XaOz2_26s#(;42UDy~ahNQ_QtCAvuKYWh*logTPSvE`
zzw@+dkG@{^8j77HI6$0ARxpHeRfQ-$>!GKKY(O>Qc>?Y~gDo(+BP3p%Vk*Q8&B~26
z4tPnwF6jQCq(q7mf2~vQD32=a-Y)mx8$G5L_(0;Mm3NpSL87l9C$b!=d6eYitP4-T
zeX`}Q?LHSVmk;^vYtUd2FYsEgRe;RcEWtlrhWipk$LE5s(wyCg+*GU5KAV2xDO3b8
zO49Fyz4aSHgII-@vnWkJW%_y(yt!)?%kPy*(?Ifzc&?$df8{x@z)N;uN`G0TOr9!Z
zQ4I`l^jO$V<L&hfwue2phK^((?vje%M4O2u$+W~PH!G`YaCFg!(b&AH5a)xhA^x;X
zsa;e}0m_NabQh+CSE^oH<mTDNG2Nkf$G64=OvcuQ9Ga2edPKq9c~DzR0(zW@??%>=
zlkvNxNC)$NfA!y`1abewk%*wbCq)Z#%-&m!q+6tww7}lOZ{{?8e_a7AbUNfx!Bsph
zsA`5r%_LsLVvRhVLzs`1o@8+f&RxZ>#+L?0tvsfg2<zV%wjPG1_DtIZS%90EmFiY-
zGxO#_Ut8Q!jqyY{q=R^WxHw3CISdE4!D)zT{#Nn&e+n~vPaN9Ab{jIxeM-p&DDV)0
zw*zQ?!bXi)^`N7!?>f-7CyAMu-B~Lkqce(ApY@lM$cQ?WoW13RWQ%ti2xBd)D}7B^
zYK+lxTHUG341KMJ^5;jZTGORi2)ux!XcJ49Lev&WV85kxg$#MLJJi#9>*;TkXLziH
zs5c2Ze>p`=17V0;u%;A}In_)fnqOjb2U9)7X>jBIDrGYixzT1c_W9vt*%?oI3ZXel
zy}Zu*6AO4FPmN_TTFPZpD?$F`u{11tHZ2%dyp=jt=)sFz<!QR2mnNNv>vtNjo(Tz)
zo`|9H*$}%;3Vb8CrXi~AJoLD+tEN;Sx`L1Je~ckRp%0njTRV^JPgR4fh#*f>48E;z
z9G$qfUZM|!ReCg;l>MTY@$$C4;Yn>f;(k<{{T^IqntDHx`Vf8*R&QPjEZ7+xd+jwg
z22AT(z(t^xhR~38t{N$AA9AaGp3=r>G4(^+sCOw1&)SNkfJs#+)7*GP`oITq-1Ih2
zf1p5QX8Bp=s0K_lv&cAWn!9JSk5MF`0?Y%Ok7s6dyMqF8>TJf)8NIh;G?1U*RU#lS
z%&UJ<3Y)lMz6_27nKUgq(g9BKt%H<g{HVl8xd?R};j($$<1`G<|BXLENH2RymUZ>}
z=iVqsi0@ss2uL?qRswl%iOB+!2_DOIe_P1zEb4Yjw??AJt?PsMg(LxDHv_Of1*Fd*
zULGq>HhdKg0-cu@Fy8U!{-rg%p17{^92GgKC($~;n*2y|wheRjmee?0s;^>~&IBTG
z*_9!fRAR-?1wJh#TFe+W@)_eN6ECv)_~h}iHFIN?GPiO;Nt{rLr@_bl)St4Ze@D!o
zpuzYaSitmS#46hJAat{?#EmNWeo&H80@n7N3Ng*GL2m?gHiDrfd4Ygm^L3ra9^8X}
zVAygb(@a1E=!LqM)!uq0cNx7bpN_E_k~C2^fA*a>^$8d(bxKg_8a5an;oK{%Db6jT
zM*`&;#irVo(2BEarf-i~kV2|re;laOuM5#XSnTjDkQOd(6ItbdP*tltvOihY?idFl
zEd4np;&;cu?Yc(wLRdhKXcY6xq1cD^Q=*u-JB=Z!^2{seR-Y#cR*y(}6Q*SNY3xL2
z)sN%_JIV=<vqZs`7Mdr5O~Zk1SY>vt;T7Vo%;DuWzetdHXzh#+PIasHe+t6o{lfb7
z`j5xVSTb;ov0H`clLJE!$x?cdgC<JfD|fTW35|E-KRTOQSvI;G8SFlu(ugE<ot;Ih
zE^dV+D|!-fNU*7;wG6#N5)LnvY<Mg_65R4LXO~pX>FZI>oPKD_xC!6Si7e4R#YPH~
zUp9bT2)Iv-ep~2)D4PzefBRxQ9jieDPcufIIRaWHoHHqP=-&Mj;kMjm`yD;%M$vzE
z(siB1L9+-8HtYM<F<KZYxgg;N()zEf8}6b?+11-;pH?rj2Z$&3j5}E;97Top(#eFd
zFvQv#`xA57-NP0etKQ`?M>P}YIu>2^Gxh-{En9fatF}3d7yGa9e~ep^y9jb3-GUdl
zf%DhO^C`}e7|-WP1_d>|xe%(~eqq9_>mE4QvL>GB&^IEVXzMtn8xL>6o02Ec6*Po?
zHnpb~AET#a{T`8wS)7u$Sbr?S^pQRJ0rIngM}hl+W|c6E1U?RooL+F(%k*-csup~i
z>zs$-&rC4jdM{R}f1Y2W<Y_*CmQh#vEx`%U9|`J;koD`M89qFHEeuvP|7=<M=5&{1
zD{hY&^ve`MwP%2G=#9ujfxhzn<k}#T<ZZ<noER#z@=L0zm0ozFKFNr^5}5o9Ch@1a
z{7<^lm}p}eN=ZDM?Qonb8obC><C|U?c?gHr&CyxZ$TyTuf0gU&MK@`c=(mX#sm|bW
z^8IwMo7G)=-n(!$fyOM+uc#Sx{zNQc#HFry?8REidlPWrTXQ|grlEp3)UOV=gM^x{
zS~n5G&@ydBqb$4h+c@?6^os{@DOI9M&MOC>6m~4O?l-@}GvQ;It$TYVdK@;`cz<~c
zc3_?0*GZg(f5p(FEEPKfgiRy&*2HPs5YQu3Oh=kVnpL%VY-iw`i$u+{p2X)1XFmJz
zqRcK=*GSBj-WDCf(>b6TLSM@jPI87x1jf>4F2kpr!|53GFfEc&Gc@Ek!mafo6Oc4R
z?}&}48HQ?@Xtx&NcjW4>Ts}LBpF#~D;Z<p#SP_>Le{mpvA1yIjsUsga=P6eT`q}$R
zi`j3<@KNtTDL&&N_6J}Ih596b#}yuXI*3|sBk#`3CQPJr)196J`c(ojRq(@i<+sPS
z&0$1+{R*n~%lz>r#+tVznp%}I!P{;H!kZUs3=SX<0n`2d$4z^(!GPNS$I7*63wNf_
zdz4pif4=9UlE?cjhI{j(0+ZXq{oEwtYAN-CsNJpNo)g!2WAnlAaeS5m|EIZaYSyV7
zCKZ$@<sv>DjUL3uC~*JGj?J^#26l0|-{&r;DZxWCYm@G6aQbO%x5uCRqz>3}%zg6j
zn%RNV%H|E)NK8kwgQ2F*49L4Fg6TCKSho7ue>*V%NUNp2(+NM4iiCMdMIW+A*ekOX
zWl-YRbS=Mip4yscr|Kd&<<rZ#?}4Vuzki=_Lkm_x+AbE!o97RY=^N^&Hsr^RT!&!W
zjab)@e@!y-?)z}cg^*-t-1*?j>N$3GEU%=p@(o*w`5QCwT!Vw}LSh|b>&b+)vQN}2
ze-;K0X<v76kuEmu`octhFQ;--bGqVf(%5NDQ?S=mRCbAkFi_0x>jP1<6EM}*j+Qt*
zf&d+9^yLewNA0dhVf^`{)h3q1!)dnc%<bhA7u`!aieEymQo?n*Y0D_YJ(7FS6D}#g
z+t;D8gofl3DC@abX~Z#l1cErp(mR)7e_6M|7hypW4tE%P`8C^5rQOsN<+9f#OO~S{
zRb@p_5E?BUh-GZlY>wr2*Qw<}?vm(G)0^0P#dOv)*$?oC%#G}n1dTMNveCS)tEOZn
zbvmr3dS8i<`j%8M)Yh8|Ak5cpp_t&IZ@(~8A9Kni{(cIQzmav~5?kAL5}c3We^88v
zUVc(PO>Ou*`<N36=R}kx5gSzzKErqw09=h)jhQa_87K#%J%DCHtshVO8s!;oXbNK>
zDj##!#E(5p!K>m4Ezp(F8GQWxK+|&08BsuVB?juFh6WoQKQ~OZhb^>1qEPYG%R><f
z&8cv^^B41cr#8H(vZK%8sWDZrf9YxI<uv{iK@UW>%y`>7np6ey_~CYrYEVowkI{w&
z)|Swyq25J8hRJt~^X7djdD3j@h9th<07*`!KGSVh!C23!YqT~)U0HP7UCvv|o2duB
z+JoGRrtkj}Mt_)^NmGf3au)#~^I&z;a5hPn0qMQSrZM>vzfh^GvuP&$f9kvibA%2h
z0dlaFAA3nLQlvNvGu|j>x9Knk)(>L>qyiMo_C?`xhT<^;I-J^Hzhn{__Rflqh>@K>
zoreiA=9QRLy7e+GSu{bMjbQG>f3n)nWmr+CT;ukHKksCOE=lO27H&)ND?mYhE1`&U
z(YP?T&3j&NM-Jh4hD~!Re~MH{d~5m=YB$ofluOlybbD(dF!>~bQS<x=(z5(*;WTHu
zQQxJJ(*sXyL5q*D#|d-aF#ic-aAFsVgbkZ_;&ezbq)`Ju-~^eB>7=gMAbMS|99p*)
zTXZX1u0=lG@wc%#wiyPr_@!756<XH#tsb0fH<@{2vQf1H=Nz_4e@9G$(~Br<984U&
zvowde8}JpT?{KMfiaA$!%m&AbPz<-T<eAWdoVYPj_(l2rdcGUKSxNc;YN`fBs&QkF
zl$zj1rMlLi%3h5$Nj#8Zv_jj+#+|W0t&gl{p1em0pFcV$p=Ydr>$>wr)M_|B;b-6Z
z?>uLUt1<^VQ9gfVf2CjH_^TQpY=}iSq<yrKD+?w;X>`pLP3%#X7r`H%gc0+(_J5zD
z<FgogkH$Wda+yRWOJ7bzb}Ei>d+bA0ix}n7Wk;BqqZ}%QxTZz}1+G!^@l1a|MKn?s
zaeZ4p+@L_`kdjpzgV{p=y3QHGP3so<)ndmN{eua&;PluBf2Y9yIn5PS;uppzMt2o9
zAy1g)B*i;w{Orm><Asl`3S%f3<Zm1;jXJ41w3>lJca(VX!5J11gVQAH?BC?Q%>_TK
z)1)oJ8V9Uj&fxhHnqJVMe`dsmln}bW8?9TcDZFXrtt)9mgYkUzc0eGrfF^{KI22KK
zwLyF$B7a7We;Z~V*`*1U3;R_phXq_oi&fvM-&2#;?`pE>*<nw5>eJ_0cxbZb!g(z?
zXkDeoTgDW8Av%gQQN#NbR;VW0r=g~@s!-N%IzK=aUHSbrgND&*_)fT7SxYFA4uo@i
z8NvE#mDY%FXUGhQeKqY1i6Ut^{H21@C031hthr#ee{ue6Jf=e=={FPMH+5qYNbzv1
zHyMaGX2gWcmpKGp7L1z6Gj2olr=1x?SX&Sm?BEwmc$UP~amZ%`h}33rxqXbXoq|u+
zh0or>KYxgQVMfP_B|`t%Hz2)qw9!?^)|?p^UWV%md?F1P_82aA&9MI9%UzH0HaZHt
z^Hnjze@0FnY0qtE&@LpzU>6K3K;dg+oY3LVlHv2xiys;5Cq*sQ!+?VkXe1Bcy4IcY
z`aN?0@I7aEGHE2*w%9Ba3}(%MYV~a|h~2~Kz-Kd=C^l#E!O_X(R(kP=Q`LeNnV+O|
zhB+qH0(s-X@3<`ATXqVD){AUW=w;dL!T?$se-xo-;|!8`D>uKztKz7_*jy}!1TuR5
z?~qvGs9!KBv%r^?diS(28p~8<`?em^Ic@tzOmFOWN~o^TgOnkTy*o@YS+wRIq&kbS
z7WA_`@KyK*UhW2xy@%><77v)(&!A!BUbwi4$G=eXYI2%<*$uh`Qz{r0@%R8|bBby+
zf82eXaR}wxP9gY9%ud#8?*f*T*!L`%A?HAq@D?TB_kuk4L+9rUO*(glP-N&Nq{Z~p
z&sXL^cZ`JlUuOh7Yr375##dtkpY979-)4aqPTGPWH)mQdmJsS5fQ<YqB>fXOqQise
zq@VT>hWFK0dibhk-S3BbSUGOzC}2Gle+Ta;sFbPL4lSQT_t(wy>0>)lGEt_PP{OcL
zA@zw-NnEv7WEeuylg&-_%$q#j(QSM-6!<CySY-}g2Ys2u&Z7)-q)t;_g>Uk&c(^e=
zK8;?X689{$tYCPH+=IOGJ!0LZSaC4h7OuR+p+S=Lr}E=!rWfDi1oM617~V**f8@4(
zLL3S#4ApGIxZ`#jTHtW0Q~jX8<7vb<Wo`%9=8o>2a&U)iBcTJw0&&9Cqp#R?JYjOl
z<=UE*VshsI!36$^@>56tzGY)eOW;|G>oWFU&NTq!d`Gl^5{0r#3bQID@Gn|$7TlLF
zhabX!O6O344TI~#9tVlXkElT}e_c4cp1K*CtbSCVN^tNK<k0tMO=WM~pnD-Unl3wp
zLw#j{VEUYEEGH0VDw#TT+*|E=o$?81csC>-As0WgHjaRYl7;Ch##m6#sy=Qiw~N?y
z<bc^$fli-6PqeoIe`kC}pZzDpkc7BUVqHHW_|<|6`bPpvV5-rh!O9epf1zVF$Jf^9
z7$+q)GZ7)-A~C{yRDT)p+7!w&0&fYI-vHwr$z<YC-0jEWt=c-Zh9Wf6v&$sTZ!bB7
zC#l0u<M>X4a3AJG0B@n6)O_}2ZPPh)>h?N)ooi(UK3v=-(DBV07IJ-@$oh_-S}_v6
z&I`sVrq+a%nJNgKS{x|Cf4U;^1{N>i-~{*DTy)wS@4Hy6hBf#BY3MnQRtEY2=J2DI
z#(oCatA3K_y~jqnLJ)6<SkgrkqC6?I9g)@(yx46L2Q}sy%Y|f7S8cOID93*4=Ys2O
z`wzp)aQVvQU;^Trn5_<*SC7RxASB{-o9Yf319O&=UwN$!N<{-1f4!%B06f#$`W<;{
zXgtMSfqua%ZVGlyb-6^sz_55PqF;UxB%gp9fxN_@VSwaAqnV=NGk{fiS|lM}P2esM
zCN4BZ$y_DP(InTa2F{WSNm6e%+QV?;7qk>-mk9FAD}xK$4kj_3#od6qwLE1Fg%G;N
zI=U07q~S+QbO;@Ze+3gSik_Bzc?s;6;Z(z6oLyVYE6jF+60&+P;@N9=`#P+)@kfL-
zc2JyNXzEVN_wAJ59rcLn5@D9a1?1M$jz7-t7k|p!WLXE1vSM|`cn>ckPDk0hrZJA5
z>n&v!3|hMEeZ*j4gAl8|f|g!vryW}_@m1`nvWXpt>BiHkfA)*xPkV|$pV(tkOSHPD
zZdF+6aj<u5sNwaVfjos~suexT2?aYI7%#n$%Dw>lkx{lJtSbi!8x^=p!vP$X@3%{e
z%;BmS=*)A>(cQ1f_LCsOAi!Px452%N8Xy{zYctGH^;XBtjZIj{&qbORX3qqcem?Il
zVcuU9!WNC(e<!M$liauU=Txjv*J5#w*0`<lUwq51R3Ya7eUvlac&pQwnroRcemw6P
zsG3tDA&7glSd*q5e-3?`@<{^{PDyxdkBav!_G*<~@zoS}#d<-{hLkMjn(L?BvxieK
z_eJT&B-tc&{8XM8_lThPd8zNvK((i20<;L8U;pvrf4P>CsrN7}4xuJz3H6wOnQbM}
zKwjgh5%p$0<o0?ld6p(qRKt^{Bi!jHi<T3X=SMX1+w)WWF+m87m@m{UKOXZC0XQ|l
zs@1V={@VrTdrgI^;5Qr&OMfsvX^e5%k}kN~Ofl+-MP$M>SH&yU5bj4A{8OOpmN!Tu
z%V#{*e@Uw>?pLb(=QvbN0Dm;h{#~F4k`Ih89To4kz0?JmRb|*+h#c?&TxznEtL0ln
zj_)GTQFV7c6jvk>BT4o2`O(~Vvj5~1{@i9$p5`<wCZkh0WmeMJp`PX7pvG2{2X8EW
z#M|^wJ|s{ReJl7D7{(P~V)N}-W^1xFcs50^f3-v={(g0|gskwcdkfmV-5M9>lk~?3
zPsm)JU`M@u1WAZ7Y=<Fw|4Z|>seSVg3jG}dAgId5Usu^MWofXo5uf?73JJd8&5kw{
z=W1QIm7ozrcdvJ`iNS+AVLrk?XuquyrCL|4#y<Ou=&sV?-<r#*Jc94brS*4cMDr!O
ze@7yrdZXqNFTbi5ryFt+bxs<f+3bJDcHtWyN|)iNAm!}Pq1PG|2iWiFsoc}WqkK4&
zaj*>jVio#eI!b^$<Gn=gyg}SCP+?OY{jFm(t>dCc{wUYR;6`%{r-sS1&3hH5=xv31
z4}j1OzZ8*IUJt`;<$`k#11R)Hc^byKf8EN@4$r7gbLQJdsrc3u;bcJ4{mYS4)nHJ5
zVXPE#4$mJOYoWh)$MEqetq5rb)6p(s(;6m@JXW&vIMdHFBjjjh`?SB(r%W2X^k;}h
zU4xS{&lUDGEDo~wtSRYa6XrZ%Q+-(oYjYS99C=%z7E-l5?rP_%C(Fnc{r3l(e+S~w
z8N}MqH_Z)qOD6F<48NR{6eFcJu_lXSW_x>`x;UaeNcqIXH&#Kz;_m#TMoh;Kn-Dj~
zXg{tH?g~5f#}`gKd{_+tKS030#>5u6yzj6psWvw>$FbO&i|QBE>WQ`bw^0i{omwXK
znD%)aHN+w~+fbTQ13NOxGE~DLKg=|={C_xVa(AHJ$PtMeA*TPjMitMNAb5I6fZD85
z`5Qv6_^0kG*TI|VI6RJ~emch5#Ww<@e$`qF%#=RSIw76YdbFQK)Y-|~aZDJ~m)8;w
zRYm-PYV*uueHnb!AogLbNI!CT*aX+Q3-Qm1L~8e<Yut)w2nTI6AN`bE{7gPGk$)R=
zl}X~dXjiB{z6q*(1RN(@S{R}vt!0!pFwBVXaM^C&Ii>EYnkuS6j<^?7ssH@I(f-9o
z&{TZm3ROHF$JRsh8H+3^&h1wS`&X~i90Thm475|@H5sOP`gI=cWozv{8=^N-X9yE5
z3Mi#kSxq87#cUL@8~+JP<3#7uD1WgJb*qY+_Vf@snKw)%{5mB~xg-7;cm57743@i!
zdwAW78U`Ne86Ab5-?D~Ezi#$)9jBE{8h^3Hb@MFF6*pw-8TaTR1LCf()hWZrP*=|N
z$V!+{5N%xYMn}mV5SD0-Y<AOeL@*Lj#g?&4REd^IIwn8(4%s@&N^Jawj(?lD3Y9S<
z82L(l^HkO+Zys4i5|Xrf3~lpVxV8R*5x6*XM5|(SG{EN5O)Et_2tN~px$y)}CWP2r
z8tp0Yed$}MP<X}U^p_G(>!7!&@?{L)vCfaVuO!0*FoqGD+~<6g){^%xju?X6qar?x
z>b-oCs2!O<fHNCQ?kY`!oqxy75W}&jd|cGRPm>GFyt`kp^oR998wb%~0eDHVvWY15
zJO$|zXrUIepS2oOe7J=q6D^p_@{80l47<zy_k7Uspmk_IZ|o%+5gUZAVwmUkZ5iOv
zrEZDk64adAAw4Dj{P4U8Uc6_zEnRU;e*9?yC0GxR5sWiA7%U8|J%1kSEcwV?jO|#c
zv<^FBm1tO^aK-;Rb<%;7icL7aN@pXIsete^hE5Nh|Jh{gGQR7&(CoaiV#s=|l4CUL
znF;(vw$T7XYpO^P&`IqZZ1RI?ohp46R)I#!CLAX5^oQF`)11m)!J(oX(K#A|&llsm
z+xFo^_aEagU#jeT^?%np7^WtIg~0AX;+;41%3+^Gd1yd*zNpVX-M!VFkjg(Q1a!ON
za-8y562Yl)x-1T8MTeM>R!IQ!_DvGMbAa~o5k}z$DrCD<3}W&LmbV3!NH5Srerft#
zw@UytKjvxg6{W%)cESP@1Ut4Nyny`&-LVAdhtY;vT{({QdVh3swfib#q%1-Nq<1>S
zLlxfcGh6B%B$fxeGz~#A$Wx})I}N;P=E(k5#4v~*^EXg481`0?m)NKaYR&K+#BH7X
zXm9kt7%V|&FFwHKv6A9MPuBIu7%Hl{fliSB+G^$C!5xt#hdB++cibOju=G3@>N~e9
z3*nFYy616H7JnY2EiJMoKb2PlcA>fwsNlkL`32*qUbsT+3kHfDSfuM?NkC?L#YtG>
zChTvWzOKQFF8bcB9^>0R9@7sW!Nii4a6g5mKEkiWqFTT5$E)}?-c+J=-HQ{^Iphg@
z06m9f*B$S1X;Aj6d3J$NN1BS)32uU-WMF%Aeu&}nJ%7zqLJhC@VTlU9XG(_1<&*>2
zJ<q+{#WYhO^vNDngsK8;=wtsB$$Z(amfeS1VV~R+{M(-eI^#E8wW@}JVUF0>O-n>U
z-|XvOl6}+=u=&__rF0uYUKtAHL0j!J*TR@JRr0T#1$C9fV~dXW1y@P}1Zl$RlE^uI
zOd)0WsDF(%yQ!WQ$emxc>4x#(4)u*pY!4guLUEpM4hL0vA>!P1zXqi~ISS~_inj_|
zfEF2^Mi^lV4pL1hl@Pc6x2md6PzB1vLMX1vbo-`iI>#~)%qbLd2rO$AcKuOoF~oIa
zO~|E7uG;z5ix^z@@njW*1-Mil&DvlF&SZsG#ecF(PG0Mh40N!#nw0{pAz0Atf_yX|
zEgi_GLbA?=4izGvN50Y(fBnfRLI0IjVf&1a917%IfM-=0B>9w;=T$59qxkei&q_e9
zCc+lOZCXHHQv_Z^6*a4rfx=q|W<KBs7uj3k<))V*4yiF%;usAH;ic?RN;gZYpGfpA
z#D8AYxC3DaA+z|nKTwabGx*zfJGoOf#OE04hk46{aUxwtw9l_3$CUMW1mq|i%5~@;
zBc77<oFCX_%i#{m`hQYXH@spS9^d7bCrs_z>mN=*Ifhp|K6wpm5tA(#&ZSTT4_BR3
z9D<&*`Wz7y=m~CfDX}y2kTfJe2)%qY=zpXoMAGM<`jneidnKipsv-Ku4dw)vCVi;K
z%++$?Q}+J^qzGI08cu#0Y8k1hsYrJp9LqudE`Qyq(_aAzLxIX-l2L&=#QC2GcK5}P
zV0k2(C^&t@cA*+7ctBG|7>8xUn>RAcdAt(v4`+(1nI+OIX3KLx@My7P0J^P8(SHi|
zu#c^JGOtnAdi(li7j-N@6I{OQt^AxYMnp)!aGVzqfAespM_&h*22t7BcoYT??*|4p
zhG6&$r<Sk9alYX{bZGET@A2O?3GarW_?z3Oq<#mROY>_H*igT<&NN>)=_KtNiUwd(
z)W<X;Q^q`eIh4H{G+sQ2mJ3rvx_@Pb<iZWQ073<vTEJAFAw{CPH1R;vlj^Ixwk1(i
z_A!`kRT8SHP&X@V&e>lm-xh_0sZ$@wJZwLvr_09ykMDgay3S9dTs`~O<5&#TNDuN{
zqm3s5C?<kYjZFRqiU89IW|mEbL*&!%<0BtI__on-#FYu}VVMAt)S9OVj(=Jh-Zw;X
zdk@$glUuUn(*H2riwRFFr&ZgYsiV~n`KwKMJH5ZUPxL$dHl{z{3kx~|G=rt2%p4&q
zACX{m>!%HRhzC_4jq^LCx%>f<uV%)4rA*_d`zYMp4ni=O-Jm7J_ATr}<}wEV`!a!k
zIf%SO;05N|64+F~D74NizkgV=JvA}5=iYmoWa^P35a;)0xu~7Vi{X?~Pf=0V@<vqm
zLg?t;$ssoU(}&YWOhWVqxyADXyW^0rr6j?hn+8_W(}HTym(6Qb6YB)l5TT84oYyM1
zEip&1$x8MhsKT`wj<Y%lT3(g6^@pW>{Vkdn|ND$y2-7CSdb3PMnSVUqO6UHzxF`Ts
zDifvLG)4^Y24m~4a?OI|ZGq@4fd8w!HQAtmUsz-eZD~)Pxq}kh_M*qJNYOYJR`@NE
z&YY22rI^g#>K-g&Z2V|Q-Y*d6TBtR7Fe;g*1xP)>Tofuh@L1dGuTf7v6E9pdh<rRs
zH7>~2W(~9Ua!25{Zhw{&x)zGCGBkB|;oGIO!oe`Z+`XgFRD7pA>xHeU^jb<B^RS;#
zQT!U1LRbe6on!g2ILPWSKesezlw@}pFCF=%H4an3o2<`d0@R`Hed|-!`h!t04{F%v
z&#=;F%_6^`kfM>M#jrZL^tX;s$D<zVZS&mh`Qk@813El&x_|D|Be?-#Wr3%G=||m_
z);<Su>FRtsr<cXs)<=%Rw8-M2t;ZX|gF_5@&tNu`+KHBtnFhj#{=bk++ArOqbi?5J
zEttQWRt!UwNS55AtT6v8c?J~ahOd|&UM&`Zy@YyrWa6SKv_1Z+T~sj6?>C5_<gNKO
zwxh$%_sLGgMt_@al@XMjCAh&jN9k)3F%eC4I($wI8an3+|8{R0PzY#`(!ZODzKwTG
zMoFu)9bQp0op73w$xMn$5Xlag>>D18r*XNH?^+rA>SR=Xy2F;9Sg8mmx0W?xo%x`x
za<6X&|I<zX;IX?v0%{UU0O~@Bxk-3sXMojPXvV^wwSU|yYs4>IZ@875iyjrsR5xWY
z9hcj;Bmj}O23ZK>bqPY(+?O<zm*ua_F&wg8l?u$zrg3T2&{9UDFYb;u)(JdyC*E{1
zJ&0zbs|A;K{c`}ur-@tX0cIJkQnTEULXPF&4Zl0+vc1fg1<~vwf**wh*OVKJ4&<rd
zGn6(5J%8_{F0RnTS6Q%w+gs<^GiD0-JXAhV!wzC`Kfw+?80)1V^F`5%$hV0Hc11$>
zvSD?3oOp2^DkW$e7E`M=nSsTmn_DBL|4fc<>aA*b-8%b{`fX1%5WKyPZB~>{kS+R}
zhHKRt0E_Loywsh<fW=G$EP^qjNpJ-&VqywpP=C<ugnqooe1&MMv*}&cx~r`bf8M2&
zjDT-mRU4GwoSADx9NlEByKfd>))SYde~-(m8RD$E3X(c`SJpMX)D9%>SP>B?LtArJ
zic@3W-MrGBIu_o9<%kcHA+1dV-s<R<_0x~l^o5$RFF9-g6RYgS*DF$Z8c`u|ykW-M
zxPL%ylk6q|NY2w2fqvY<a$un1wYh60TB4(0_zbl{hgfzTvI3HZz$-}?es91|MdBR9
zNHSP485`|i?SaCwI7krcF3;75QFxDWW8|r<p2h`rwW*Ag+s8n>Fnih~XfC5My^#0w
z5Up_$=%8DTnzSlS9tt;oM==VUU|sCRNPlKhtPY%ye$5Oa^7IUwd%^zX9>9Kj<$M^8
z1B20~-|fBLpq!ATaA%1?i;Serl@-egLt^Kt;u*L_+UXk9bk=h#iUPI@HnRRPA#vaS
zZE}%}_4}&ioyt)lU`%G}ni@$NND=W5N%jTUe(%1V%S}s%>{jdvPX3P2%BYuUhJQ{!
z);fZ=9rXry7qXSCBa-?j5Hc$9Dr-_n??#<h6eG+DO+0QmgAgWtl78cgPYo_E4>7!^
z&<XF4KT0M)NG3cd1s?lD8~mNMlH$$05S1q+qGAeZI$Tvi4a-vud3;BR4-e968Pb5O
z)7*cWF5Mh?YpZ(FkCC)2P;n<Vr+;&x{a<DFhtORzWX(o?`9ivr$|lKmqx4_2Gvo%K
z;p%_%hK?9!!E2M}?BOik;lo<`nXIK<VBfxT+q!zUXr5rQ*&tXkNI%AOuQKAE-W<-&
zTf>7IsUYfP3Q;4lC}VAIC6A6Ks4r|F5UMc&w+=6<JGD@2BYrJ@iL+m^(0{?)tk&qH
za6SY-K!f@3JAV&&ZQpU+bK=s66a{3flT78omXma1W8XNH>6{}S!qtBeurFN7r1$=i
zY-DQ8COQPDkd9aHr!+aO3&t+3lI+buth510Tl8_%T!^+lxF6&=Ui%+gVb980T5fgh
z1hOtD2&-zz*18Grpv!B+5np{q7`ahfh_>+`iI7QM&?IaL^J<EK_)#!)?a{_sGQ|!n
z#$S6*y=Ooa5T0v@TqO#v#E@vvTe5+g`qQ?7=on2#k=|e8KBStOr4kQ%4aqwC{(<^U
z!=z_V3l)DhTbGe!2NRdE_5~KVmQ)2*9G9sS1{SvgaRm=Lmqktn6}Kg;1zjJPjj9F~
zx0uQW87!Ah+y)l6$Myx`9ha{81{RmT6b26hH8Ypb6b2XqHaL@^2Pl8H1yEeuwl0jj
zyVJM@w;;jYo#5J7pmBE#1b2tv?ykWt!7aGEyUSzCJ$s+~|5fi*SFK*(@G<5$X4Rr1
zQ&MFVHnlYdO4x#&7+INEcmbmFDy%F378Z6U78W)HDk?QgCu`u}VgxD;po62OEr|DD
z2BHo?Bd2$nn32=FoxFc72q5EZ4Pa#luyXOTa`UpV0N7Yqc>c%G)`1rwX5?aN3Xo?4
z$k>8_jtEquws!6gmgW{t?`i($6+ml32Vmvl;r{TaJ3!b5=wNAL1Omt#IavU0-ZPpQ
zSp!sUO)P;<?*A2nmfyn3$&Qzq+11sR$;if$$=1PKfbIjp)zW{-0-yqP1Uk3?O#y#2
z3{Wt#0shq)69N@L&BD_0FS)9%nUkxL0}${ou(mV-f*jvHoI$2Q2f%xBfU2|{K+z5e
z`pa1EFM|(&zh?trWn%qLxWB#s2xJNR)7i+x#MZ{n2;^=FG6$GhS_1)!5^_vVZcZNn
zMj+Ea42`TEZQp<Wja-Z@t&NP|4gRd$2p}P>3^00c@bC5<O&l!koE(`PEv^4(k@=4>
z?^6~BnTp!l*Z@IJjtGC$CuZpYG<hGpJM&*B`vqj{3iA9Lm|22M&HiY@)Y*<%9b{?m
z43rl8+vHt@@Q=(K=mg+oVPWCsVFv*10YEns3+6w<tGR#M0smC8{vm$vz}wT#)(&9y
z-UQIw(hT_ihv4aG<N^dZIXDBoJ^!uvZ-l_g3NW=aaRL|v%`HI)|3rTm1I_-z?}zVT
z=?2hYdFLK0faQ<hf8O-o>1ApQvUdN+{Lk}c7MD|ySJh(ptK<LZL_}=e0G^B-EC5C}
zP8I;`A0&U{et+@)?=VV6mVcM=FJEbpnJs|luWH}N^gk85{7nGbzn6m!@ZYf%Y~NW6
z1knCNa$Ocq7L)e}>;E&=|8)8PWB4CY{%<7z-vvoHTU-CBru|d@|4|#+SX#UPZSYQ7
zXQ%fCkhgtb1JM6A)dc>vwDLexOJ|$^t(A5%dS8D8VUW4?f5vF(C}HUaG*z;6GO_rJ
zEq}??|JX8XOAt`W*3t5hYX!jgel-6}_r6*tU*4|{$9GEpsRF)l&VQ#A2btKK{;^_g
zoLm4S2L~f}g!e;!N1Omp*7sdB1-ku7VgNG}$kyrI1@PXUH^9u+0pX7m<>Ce~i~b?{
z3-N!vv*<sFn+?D$^$&Wtl>ZO<$O2$i`Ui0WnAQJ-c-}3I{y`rB%*Owq_Y5ZgL7eY8
z6I<)|=Kn{+!SRR0#^xXYKc>QL`Y*@|U<UpRvICgS{{`RuE&c`H3$y$ee9ve7FZf=J
z%|GycYnVa*g71lJ{{`QtVfRm+T<_?ezP5k=s*LqLx5K~Sdv3>n9NvfGXl>;9ui{wW
zV>tZ_zFRo|3%=Lt@(*NtFV5{>@Vz+qKjDAT!Nl3Y;r&MZb79}P@;~^`>lX-g1DYT#
z%-Ndo1zFVwwcl3@6S^|)jD4D>+R;p*WAt2b=x}~SflZ{V$_Q9<xD`$u?m=7L5~qK?
z5n3dD@jPm&hYe|sS8jXqdNznvncQhZn4iEI7|%ZvuInd;Ct_3++Vy&|_tNnH0@(uI
zDMOWQ@B9%(DHrqJwNKKmuD^6?<m=py@@^HEEaG$N6l0`%gsy+l3{{SC<}4mLj1wa<
z96iRl8|uu>?KH;s@i$VLSOx^|<4Aw@0#EIAwy1|$kCk#YHpeb}N_=fRV#phekx6P#
zk^LkY+~1zX1<M2TZY}%;q~DQ0Brh|fG3~RZS6QlrgPMA4Tt*gtux@830)l7=2Qate
zYf7513>?T6v8&n)!ZDp{*uMaY_kWm)thC>1bemZv6D{&g)x7~`khLt1{F;BeJ)i3B
zSB8HMFy4JlZoMCF`nlEZtyv`4FGytW%1SS8p%b>e#G6i>3|;Cp3elwrl*U5)u;`Nb
z@fq4EnBOztLgbNvbu*uZMf)RkFZc-2DQC9VMU;0%!@0tb^%?Zv)~*<Ofp39a3U9`p
z$>-<!_!jKqh8bUJbJE4TzvX|dM$;(s)dCx40!x|C*DG+C#7f*r<>EFu=Pw2+3pk`D
zdv|_U4u|OmxLZB&<T0IURH(PbaONqKZ6Aj-j_x=34~YZ@_wn4C?r+|fm5JzMAJ}R<
zoL9JpiVR0>sBs77eEXr|_r1$eF;!AH9HQ`UYql5OnBqiD#D&bJDH?xOQ|;5^)@WDZ
zv+IXaX)!9Af(7}0*#rI%uLZq%v4URwYCj>IZ#PYTY_%PV_q1S08E7j}R2jQVkQ4?T
z!%Z3&uLL55s%rPS6dvnfy}zeMx5$1?Zc%z#SLhaT^fd$cI3rRatz)UlesB4uU9Ekh
zjR<=+3u|6gcNMN*B&vT&pBVtI9S@B_iBQ0Hu(MzZj2%k>@=ItkR^<v9u-@k4v#d5f
zGKC&VqnkffI_lQ0V%_iumwnRh(gLSk6B7(oKZKvyedbF1KIJBvqYN4>BTNg!-?prx
znoGzc`#p&u#|DO5qmz|}e!W%TOGg%A7W_E?&SW1w91d)Yf0lm+Yp6Cl=|TlliCsL0
zEQpb}2Au|d{dGbZMFzEs33{QgR#avTOvb4cjJns88zzRU27$s~W1@GW>u7dPYL+nm
zDxA6GCyK;xb60W+&V*>!3)N6cuc|kHTV?*Sy|H9O#7RS+M*d6^R+bVkxYRfPwga29
z`lg5ZjMZU_^VNUsptXtvKe&d5>gsgCVSGcuEAFHt97KbWY-pXtA6FmUl=wMVXNj*q
zIgQ^-fi0pU^KqrNG<!S9I<p?IKc%UQf<arR;nYhxRn$&Sd*8a_k<7jNn2=EEuGydY
z>9)4+awu}Rkm;lw+)20QR-e^XJ!22-nuHUutPz)8%FBN|$bZ*jDIbY;@#I>=OPJHW
z?*%|Oe77G?BWK6|bO)J`sWzo0%d{+ucQ{Pf{Q+0<r>v(@sHnDxWL5_P39<VUElQXy
z&DUNwMV4ue<Vr4SB}z)`r-R4xX2y8jb)7D}qZ#1naQOWo0UW;U$Zu*kD)R*6Dvyvm
znGyWd56*wKKJw*keRv+L&T*;PAv0X`X>Ig{$o)OYWs`3_onysof_-_i2{<Nnc)!N)
zHgwO;zdKb>3zlp}jrUzYfnd~*Tsz`-+bOe66?8?L)APp`6>sOpg$gv;+(`F7+LOtd
zPE~FZ+8+6AvuZF$42gUas=0nVA!CW_*2=N<c%Xkgd6>8x^rPyl=$stJcHLOrVAHcL
z?Og*@8oUTfTl_+G-p>)SIhBCH<!xX^ftyb<g)13y=ei5%T3iJXOCT2q;mLtGRWDts
zOBj!)tE;jkJ?@Kn4)Hz>(Zp@jj=Z!>KU*j~*3WrvD`EgpQdlA7Tuf$du%Rs*3cIL>
z$mM^J4{iE{nlxv>R^a7V*en}Uu`HxO(?1%f9l`eH#IMCOo89zgT6a(>iL>SKl4uCi
zsfx%k@A8-8oN{88ie?Gtt=(l>^HqBO>O!QYXr2EI4U&L~BZmtIO92b|=&3Sq8;9QA
zr-n<rZJ>D$VJ!UWE9+$|k7$57H=V4BzV?4g6x)I9zAeB1#^qO*U8CP|XXKPc3MZt<
zdX&SQ2_|ttn$XDZPC}8oWr8X%|9e8RU;AZ62hyhet<_^ZLc$jY9>coyp^UsoMxYW1
zHLoEOttmp}y{ADJ1GTI=v}B3@A?M70wdhO*1E(3MprcJ9auBk#Ln)JKze5gc($9bG
z9;y(}Kxgc;MkqS&;MKKqb``Rk(%X$oVD<w7@aKm(yVpg6KnGuo_{ZjWXPk?epa??-
zYfl2WJD~F2T#SK`87>z)fz+Ckk;|k9zWOM(LG|}htO4yZRucNs?ZHTls%UwJL1R0=
zuY8^>r6chE70T>5`x+X2^EvIKq0@gT$QYH1Op00SmLJ+bek_|K2|Vuo8f^aklkTT=
zA##cJM_Gw`wH)K8l-M5{{!!Mn*5;MnnTy6#6_G1s&HJUZNwdFnH(nl$IFwm&t4DkQ
zWjveq8U@WoBL#psNBu7+xsIk$=^zx3*lsx3ioCsi-Z>(<EDwCWDD?AE2gHA{g51(T
z7X`5&;qsE)Je1OMmao+D!t>^Y8z5^gckmDy_-F~mhD&}~_?;uA(R{@yVPU+VE??t)
zSEsCVsrMT0wRV2j(=3dEX)aFe1X6e9viQpaUh%Qhm@VNg7p9!yjk5@PL=ip?_IU_P
z70iBT2S;#cy41;bYuQ|JVikXOt|-k!vJ)O>(&5Ncoo6$4O<*tqIF;m|DgX)aEQrKL
zYF7~Z$y3Qle74K@7`W|p^IhHWF|;-%5%gC$#Vbdb_v5iKh`t*vw<Jk6+-;m6_|cLG
zBUv|LPM)$FE`M}px^iYX-C7`jyb|C$|KRH}Tt7*LVB1!;mhA>P#FBsF7=S)EgpZ-g
zD&aPMI<ud=1~7lMiH&S<({R6OSVW)9B-f+x@WD$ivU9oYM$RPu8k>4M-(Sq+Aj0#F
z5*yU4C0ZfM`q<}GYAQr=SNsud;)_DaOt@?{c=j}3Uf1DFytTi?BP|T#e1yVPSRpvy
z`bL52YkAMp33yxu+Vg)JeR~-HKw(beRzGGCW+|j&S0N97gJd+asJLOU#zC&dddF}d
z6}EExm_&iCQb*tBBzmAPWob+Y`6N4g`<)WcZ=PfANJpL!a2ms{3KHaFhegc?UWJ5$
z;~~fCo56%8QXv?#K%@+PV$GFcYBmm6j1p@P#9hi6HqA54TGxM-;DKXB9^<#X9dgCO
z@0dSFHGAnLF!WZK3k}^%Bb004Lx&S|l@ogiCFo0+E!A)?BKP$b0Jo+FdGdyV`CnD3
z)an#kWx1oP;22a==Z<w83eum4v~Kz_PqgfT-t9H>L`VhUck@~yN{`nFV-n!%;<?9H
zuMy|dvVmQ*s8W9sG_?%^;M&SUW2n*#>rPa8+iyu3T{^{d+<nwPp0h2+K$%E2_y<}v
zGfPH2jcUCYG@l)OXVy}gK?a~iNEx{t97H(UBq4;_z3g8TX$Tb!3dEkW5)Erh=lWmL
zjaPd>hdwwbhH_PTdT>id`RX~~E&ZO)p%^ZkXL4)Ht-F7@9P!tzDU`mL4mQGXr;dqO
zJp)&lAni3s`qZ{hC3im^Z_aJgv5bX0W8LQv;mAZqNy9yHP2iX`eYG2p&q#Jza6yRc
z#&*Hxj!G$}l#G+Qw^d^^>C|%kcAZcd&8XIxNM0>I>GdD?ADM>xk{TJz#s)pyi(LXq
zHw<b;JSu;GhJ62e%_itF(+&p%ad_2!^E;&XGq_P6*+=}@d>e$Mr4T%G*99_I1h{+d
z9SL|XA$L_hpNrt{5UCXp*0v&Z2j%hbq)C}BWGoN+O<@uV0Z1!rkEbQ#YeqVu0cQ6~
zZXAA6!P(eehp?K@^?XZ`By(73TA1YQBG;S?SI>XD_Zw2W<kMw3F!!uUW2q-5oC^0m
zrFH^r+%jz<FINqY%#f#Hjpnr1C2rWtA5feGbSgJLj>bzm_BZ?UlB!Vh=y73Dhd$c4
zSJGKF;D@;kA6WI=Y8Hpf^{I&H6RU_AS$7_})fr^}*2?6q<&-a~f?C@>UBWsGgT;($
zt3ZEkAoMMMVz$Z;UhiMXA@g1}OgKFfqt+zy8!FwNQ~umS2_v;uw1+z4B%NQ;4(~fO
z%e0_HbRjek!Y*@2LNmBN1}*#dz#;!oeBN>rR)A)dLk1;<CksLwtm5b!1TUFm#Wnb3
z;S5CcMH-S+Pn;CXiiE75a{>DryO&9;Vw-;v1QPbGJC*HXzY>(?`cBhzeQnHqXehX3
zj^u+?#iv2iZ@J;!c7l?ru@XAOax1F~6sKj8{K+@gioK+44-r)|nD9l7rsX^fjW3Ho
z#-0!~zwPYrWGyGk8vnTuz&T2gj9?%8yU%{Az+Z4Xe8PiORF$ENC{KyC*M|f+455F@
zHj2nzCX*z;_%kTcn)YJ4ffA0S6*#QC!)#-xz$wcOuK5BC^P=lrnqM^vlS-;s*KV1v
z;&6^s_wMZP+v~k9hQU$?Op^Ud(n>r&1bZNx0hflG(b_4V{*Et8wQo`&Ue$a|@|6Bb
zWMM>FkI+vQhV>3WaW&nFZh<nQqx^qW&Z1O{aVNk>)#26Mh!Y^Mq$OE0JHbc70~JZ6
z2URkOs|AJ=9xTM^7q_xDNlSompAjp7nAscHG%O}%6_Sg#<%sQo9pJW{gs{uhBd|Cz
zS5=iW{w29`(YlY1I;o0+e{V6Z`VtsJXEwPl;gN3GG}t7VyCrP`CZsAwoy33t2HhTa
z3I%ilfqlgE&y)xt(%sUiDP`}!GO7?Jb9Zmw;=Pz=z{)Zq%X2%2P6{v{mw%o6w!v)X
z&AyhDe)OwwaEn~BBniD@kH$>iaT0ch^s}b=i}&+wLsPR>UqhrmqTkGKDyz{!GN~VD
z;fhjkaKocfqy`U-;e7icwPSymzOkD!%`Q18wf=g`6bO_Hu%)O=G@=%Q{+y}&L8^E!
z#BC?Ps=F_Q;w$hn;nR%D>-`w9QiXj@!oD6@+iHT>%hO&)!#IF76;Ui1{+<A7d0H(Q
zvU^!WxDZzww$eyjXTSKa*O2YhKj+*`EA{+5_Dsly5wA))`!r@}w6A}U=s$09r1%BW
zFth12ear6t5vk&%v->io!)4OSRiZP3nYW33F9T<{yARpRir#GIBjJ75F^c+!3h*CJ
z*>t;>yi%9hKNN%emS`&tfQtSH#1a}mUhOAfa0JSo_t*OvC3!<PF8VwIvx3v5v~{VX
z)lXdJyTC}0Cp@<F@z#HTiA1_n`6;MvHaD_=!mIpzv@Y0ggC!GpkHa|49&|@+x%FF0
zXgS8;Y&`JgOS1~oQTok`3yt5P4HE{<sOnaaSlh^gwfi$E9(O!WE|S*GoM>G|zs*5|
zeq)uSGpB)MI_!PJ0&C$DsuOd>4C3p0<jCeZ!aB#d8rBL>sKI}en1Ht(48yKjsX31Y
zCGlf%Ke(RInf!y{y#7>+r0gvpXBkYs>)=rv)V8#P{h4qvyP0snnijgHbGcv7%69K%
z$K=`Z6o!@q+5xfH%`I<q<f17FtV#jo3LswC2_5N(u{7ob7Tsp6QEC2gtHtIxkIqby
z^BbJ%9v+Wu9DaWgdY_tXiP2}uZCCT*_(sJ*olOY0a6<dRz&v*-Y1*W>b}KYaxuR0(
zE|UgtFAI_`%2*GvO-%Lio*36Qtw@HBmvU|1v9rCt@%;V_*3QcYYu^Q67vUhdq4T{E
z1c4V+`7tT>+;ZrCecFKjA{XH)bC&z$lcCXh4o2Y+a+rTk)mRvs;na^2@vOWaSAuTn
zV<Qqmq#Tc3%S$<<AY2S@tDn_-!$MDaVLRUazHX?yXPS-Bu&Fzs5XYz#77#s(CrXvI
zpmvGJRe0i0jBtVGy^klPc8qe9)I)hskY}=n0szDMVyYg)q}<b$&SA$=!B&@FZ3Ftc
z&)?|du;+hXlz(m%D;S#vGc2|LSPjauC5Mk(C{l-S_n5pgKQtdT+K!5Gu9wK3Oy%#z
z{QT&?(p*|T0r5!R4l!hY`^F*onN6=JhND<2Z8SIi_hp)X_k_J1(6%-sMg8Q;2kEXL
zE^dNb;FxY8|0-1q6q5zfT3jK_O@81h_Em+SnazLcMZSOV`oVFh(C_IBkM!s3IJ3Nx
zj7iANxAB)J2F!)R&*H~LNoYe%hZ|&F6TRg0GE1u5C|!$MBvo)k`_N}rOUT63QwzfT
zGDr+%yP=&eZEzTi7@uyE6moVbxm;>h3DmUEzfmmVRmvs5af}t6?g^%Yjnmv7-}OHD
zO5lH*YDmh>ePGzbal|f{e9d3iXKpUBukVE#z6H04=tnJoc+e&(_XrXzaS}pezXM{V
zrGMX-pBECaqT|=JFx-}J4;6U(ojd&IvjJQ`(GZiF^-6AS{fz28StRNZHiSE%HY7Z>
zk7$PX0+Eszm!Ra$2OUN}-7b4OO)(a$mfe4ZB60VqyGQVHr#}-9NlXBV$d?M=<#nh6
z5Yz1oK$dcz;d)c<f0;<3$y8<^GUjumXpSOfzamu~;Vrc?Mn74iYIc%|g=H4_nWOl2
znnCQKvFhK0@1|GX0@E%)w|VxxyFt&-I`aXN)Ak3N&R))h1RvT51?plNOv{om-5P%d
z<!<RqshQF13DPb>HQp1Ud5JUJJ%u+!Sk=$nD=anq<#cTp>zZ>g{v8}<EfQi)7ST&g
z^L1(vbo#2$lZ+(6AXo2~4~U`K3eDo!6QbbPRd<M5h><GLzy^<=ncH*r_9<P2#(Afc
z*H^6#PQz$i2_J)Wx9-S5NcAkQaGHP5aSekZL6SnK>T_sZ(ulgGsX=?ow*t?+!x{!p
z5W@nqg*^1qEZ#WN4Z7r>^Vo1_?COgi0~f>2+Y%XOkRuEE!8b~4#Xsr9V?sZo20W`5
zWIVdkRi2w;hW)~}pW$6MwW26!sanPkwScuBL@dZ?ceF)jY!Qk>kcE`wj^2MFukccy
z2!YrwecS<PUwSxt1qXQN^~ufFt73)~@cNmR&f2sW^$4?{MbDk9BC_>;!jGNPsWWYw
zkAIfQ&|R*>A*mp`#4$Nu#PfT8K=p$8<@Cdfb#fqGc+cTA47;*xvho#wM8>7(gDj~R
zIQ(UO@OYxeYD`Q9LdtWxa!`N0&hje1diQ5uBB+4HB|#I0TdMUM_!EV;Wf4_Qp0D?T
z-F4uRrCH>?bA1_g=KE>~?JL{*#wD8la4<`>we*yL2v?Z$`+Yj>u!#7rV0|nFGKY#@
zEsp~Jg0IxL4#|ANrqZH>hRmV<-Uy6V0B|l3-Cx$1s3R|78L}J%^rL_NvW)n@S2PP$
zca*|MD}_rXX#)Fj`j#JY>n5f~#Lw8~C#00PyK$>XPPDc)hZ$=nhB3tDV}I+V4T*f)
z{7#7hrgy<c#jo75yXmtFX`#PCvUM<WT<x1TNU=ERX<BIKIvGR^X8CA&NNNkm;}Acp
z?g{ev<;|t4clAlhVC#Pb$NyV{8lYWS%}%*MK}Dx1>`u`(|3y=gtCY!fV55RSh-~k7
z^1iS>K_%{<EMn!dXcHfm)AGn%a2qr9t|~*b=76ydL7m8#&}a%9%_U}Sr8?g=%@4+(
z@(G^&EYr!=R4uBk&*pF>d~w?hp!ZTT+6JKBe&SsO1y`^97{z~RB0A&NL@zCGF`Q}}
zFLN3`v{DS0lmrQ-oLF+G-!sEUOTS6EFyH<_b!1jwCNJ;B*N+iuTJ73z7Jhve^(#Mm
zu`3)^o@q;lnw9<)?=!tDAaI4b2zpkZu0m^+gpDW!HAfvre1g>`x>G72b&!gKT^J&s
zH+G3C)3J5qm*#&?=)_QBf_hegQ81Z+zJ!z^v2sfCHg|__bN5#N_@RKml}6LE6tF8e
zCQGMYGMI>BC;HnN#ylZ+5rsVP{cE`)>^9={7d)$=rg$}jY$Ga~_ny{3d#c+vq5I=s
zR(Iu7IG7k7(naz)Mv|Y?7(Z34rq>km3P~OURyL<z4qtzyHj)AmU@Wx}$3gdgAV`pY
z{&9yJSU61L7BYt`mHsV}7D9l1+Ty+tIUU~PsQ&8}^_F?l@y*(%B1peauIhxx)*_#u
zx)U=1VwW+<52{}Qrs?75T_!OBH0LYMDi8c3IvbEn6LULdxL3hx<*u+OzX*|tV`I#I
z83*!<#)p6QE@K{*Fo`_wL)G;VoF>F>B|aYx-ciDYL|q-|HYzpvr>=p^cBKX)Uxa+w
z?jS~&zA_6*4)Z=yXcmN*kDm*s%KYQNvvqhb@)EbcB<e?OdPc~C!WE${%rCmCu{IlZ
z^kXa_kcAJ6Zo5kw1M0+@mJ>=(D&8It1`sI(78ie0h4jiI_q=DP-$ohq%veNO+*ACu
zW?m8IwC?os%IO*IOLp5S(NwsS<-Spw&boa1q-P#oRIVq#=irk^P7b?oI(U`rXcEOY
z^|FYD<Bi+y0cjI<qNasmU*7GPU|75?q%w0~CBen=X@$GER1&^B=bpBfyhZhxddGGT
z80>!lpp?l<*}I<G5!HdK5DqIz?zW~2J@WyV%2{(-gbYt`K#f?x*dqsWl{$ahMGt3Z
z@Z=K-rDcioxx|kEQ@1!+3a-tb_sCtQ=3(DDKsfYd6IGgx`*Ky<G%mpqAu-aU^(eRN
z9#Y4FxBVgc4V-vf+v(=_1{WSP>GY=&dv$;LVs4z2I#ui2XGAl>b$y7>?&lt!u`e(H
zVJX8vM$^V-%}-bee2J-rFP_nN6|m?BzeDzB+QdhKxbx>j9~JXmX~hfh&?3~OiZ@Vz
zcl#gHNS6dz-qsA53(zc54lr!dMl!|ps{-8UaZr%ZhS_YB;8Y<#j{$=2Q-0G8T{3^E
zVv^<yM}oAW<}=H{4)Xhci^x|W^ZMS5c2Ye^^kUS8VM*G^q_5ETL=tb7Hs2}CIwvEJ
z<g4RRf0(-F1)I`!H2~kwSfAXZo-te%7I-bRY>FM;nKFoEAWUr%d>yIsa%K%;NNmdf
zVtwL^@%D|N+-?-nyu2R!Nvm@H>*#+vWp9-ka_NAuMtdEU|4}47VE0Aun9}IFiO`N#
zSa)n%zE+1Qq^vE5XzO;)<IT@ps}(^oSHsYWuINh0tCqfywVcdkUB)tZ?KQYIrJMr$
z1KEoeR+*gR5Pc-I)pm@t)zOLsFMjD_7R*r18y39cQP2A)#CJqkak7d)_(Fd(=aznG
zejy4vOh_NHPH1>#saD7Hyrfr42+5oY8ATkV(|s@RrNxhWW%_G82;0@YwI#zE=*7w&
z!#i1IFiwg469TKmVsu8gutMN-o%-fMSQ)`DWSmp|zKyIc7f_iKd1DdV&ST|FX;b7(
zI;Y<!LufHzq#nM^R56tn3NnA?`o+QYi*?r%To|6fH2I*+GifS7P7+Z^#xPFaMJ&iP
z!yJHQ_mhegu@QCc_yrObB8Ah|(A=P#@!Xzj?gWG@Ka{gk6}q5X6EV`XtjHecHK$%%
zm_~lPEG$!Nxof?+j=Nn?mh_DvBlMX2{w)bELN{fvic;*^wmXJS0~3EwvbSGn#PSlC
zf7oVz`>7F3K-g#wXCj9_x93nziXI2AKfb(}2)2!mNr~5dVty8T^tso?P=KQ)B}peP
z;epoz24<zbN8tXe(fFRX-^uy)=o45X80T|24?DGt{6YGGhrq=W2~a~_i`l7#^!FW1
zj4{RA^AWWc7MARdL?wT+TLm)hz5AfAj4ljl3Ja*sBjjVhUre&BvQ^|dofjJVi0sM?
zMIoc26w&2me2<$|+lPIK%;TqnEh+<B-6XlwpIAdt;8!?lfqa|dYS{ioieo{iMnjz1
z$^v$g`j(`A|MbGLDY)!YLA7W^1IdxFrSO_+!;IMeuS+U-66=5I_1aU0(Vcoc)gk#Z
zOxw~Z()dLz4I~`VQI8O6tcHY1^ZB0_>zBtBQu>W1v5iAhJXyvG@(|t=oxbOHd=P;{
zK>}$k;B5(b<JldyT-c3l8c`^P4xx8d7}e^~IQ`8(O@mXL2}umF!*kP?(!cJ$@tS;o
zxuvn_Gla@L1#y3i{mzp$7~v)tb<hZ&<%EAMSX809O#SW!#eAEbXB1@)sF(N(Mz97q
zQYXUn*5-_m(iZJHX>^+_A_#eHWO1|mVO<{IJo8IiE0$(aOIcPPL7kl;;qtw>@tys=
z+#YsLGc#1_B0POnqoluSkSjUl)tF=aSRbc6c@)DEh4Fv!UazModJWT*wm#9S{HaN5
z0^?_|Tgy^4at9J_iNmFjU#@ZDR{MYTG<t`0e&KGt&c%{bMXjTeq|1~tS3vJYLP?TE
z#v;#JMizy55Q2AJl~4LXyYVmvt;wk0C@b-(ZhKn+H3OtOXivj)p~SM7KOz6+kEwZ^
z@$;wR!Crr$13e2hVzLSenJ`w_=FOWXVFxT9_T(j18`8(PtG5hXp9o6v7Ha2N)XP;u
z@0bt{QddBY5ru`ihlpr;799hh=Ehx#o$4I%bvS2)grR$HMPYXCfWlgda^sqR*rDM|
zifxDQZXjal@7q}0F@Cj#ZGBOqa@@rNqQH?6O#pvosr!Q0<fWT%73)ErfP2d$47+=G
z!N>&#wb~S7Pd%1-?b$A;UwkqHR-AMvN@e&GG$X@jf~`embKQ--!|a}=T}3k+&D7!+
zO{a+{jg}A?h6V@jwmN7lt^*&QG>1B3Myjioq<S?qkHW1iJ1LPG9}>UR;4X<Hv@<DN
zy@-EBN?)>Rq`X|v5AW-y^9!w6=6P|0T|_-dU-d6riR6`o*2^tU%Z%LG#{3Z@O3c<N
z+v&2Dtx$5Fdn_!G`(@axAoM=?1YcS1PmL#CTChro<;8;W65{X`U=3JqZI2bpUbiPy
z$zePd->p;^0jLc&eSvLxrv;o~>AYcsI<J4Lp4=gz=_ZLr^O{N<$mPW>Y+}a7b~lqU
zUqAeX!7*7{OVu3k4UZI^PE^dVe6wm_)9<hjI_Ft~)ffxOf)OdBY$hjWi?vV-ZMVJ|
zrb(7inhv#(q1-Ts>XTiMUMPm?>?)&2DpGPOqnH%%Bvl=JMpFHFjNzFXi!z(3m1}<j
zk+xL0ENqyI3@3R014(tqo+zOBXX)2q|4~b1doh)TLbMdarZ~AGx*J}l+eu!d2;V!Y
zxL`}{-<$EnqrSWXSLws$yAMMCgN!VZHSJS9YH)s#pJp!3drpw^M}7pIb><l&$)DhF
zb}ap($P?(-eI>y1KQeE!T;xua;AnsKb$v|g`qlWm(Z!q6VFP^aI6mvc2NfHkbP}0a
z1w#9J16ovylM1W3gM+##YH;u)_T4&9a+kdq7EJwBA1t#}g$L-*O*1(2O4<lUQZdx#
z;=SY79gL~Q^VM?2)0Zq>z!B0NdhlK4Q=Nj#lr2{I)$Uq#&bHLy*6$zXE5LsSm+mo3
zY}h-`U48^TK|D^d5DVzS_4MOH*xYcIh)GTuihW3R9*gvz$q%sXM0`3tu?s_dIkEIJ
z<5!AFPE8wz{EQuMbn9kisnlxs@lwv}Zh_4(=0hnsQI`c|9SJ5D$u{&*%+?M0I7vNc
zchEw}{$K{gPhTCvP(jE1U1NW%0yh&+R4bP0FSWtUQQV6+MtnAY;=u9Q!h4m46R$z*
zQ1W^|6*hQ-8ygpU=}V>T$K1UE2?LMfbWg*rg%q?5(UQp>*~v;Vz_*d+D|yShLHRD|
zeCc#LLiD9NUZWtHt_{AEJ`&{y#2Med&dW3$WJp$_*5?h~Gp-?-CC7g(Esb4=OT1VK
z&y-V(H=3swji*?{{RGjx1QI7O9|OXX*Qj&jEqjxZ8QW_cU#r?uVudz5-!l&XYS|b?
ztAiYcU`Oxqyka&a8aEQRD6KNU^zn!mel_-fRT;>oM{r;=R_O<i=mMU8O+6>!srowb
zS|)p78r7DI^yj14eQkdZ&5%p8Zo6pIstNaLj_&J7EE<hIWUT6NQiok9{<eX9`(Bca
z9m)^Dv59ig7e>RqqKeCI^RwljiK@?J7Z)cvvB|SwY2A*&dD7sps5Z6?=ATJCTRj!G
zDo3diuN+b53z7ugqr_f01kJHKSH`RSQnd4TbkWj=NaSTGKBRxC%Vl@(7_Mr&eVc3`
zeGo-Kl5^;u2l<;KHV2=$DWw!VejIv1ctN5F2@t5*kd}Jf2xj(Nk0eP!S~bk8F3+L`
zuFGM|_9-dbMN9?@+@C8&rnPUN`ql~vpRYL2MK#nL#RoR;Ota26rtRyD+(@hp-?ZT;
zo3AxUP4XxVKS+P6BPn}MQ1pQ6FiGj#)Okn7#CBxRKNs?(k8r_0U-jj^?Y0wl!XmJV
zUjd#tI%qBQ;X$!j*KwsnB*n8oV=8U#d&VYqTTjpIY|<ABnc0<&k-Mt>Io&QL@i%dY
z&s?>_agx#pf7Z5ecJF`V(JyZ71UE#HL6@LWYcnfnh%|pW5_I})aF?gSU1{-a=;u!|
z<2G6iBWH|zbySo&&2hCj+!*1LsqJDe`|9*3JR*nLNk79;s61JZhh62%6oQo{CIPq6
zFL}kyZSrZVzP+$1>&xf?&U<M(7vQXXSSOEj7%Wrc@ZUZk`)?7)?BD$0SeS?6c+Lch
z);1(kQ>=gLxMtE}9N@v`uh{lc8KO>|>&-50n?WbSYCduXt<0bZjVq~+i>jS>9%|%n
z&6cE|2}fU;DWLW22^bOrY?GXK@v`AN-#(QO*Jc+kJt&BF2jS5bl#acal0urMICa1g
z9ntb00Kg@C3SnibKRx}#_^MfTEEvSVpm0RO;<bOPa7sZw#fY@j^yw7Bi|?QtW;U}R
ze79Y`Gu|9RRBR|_oz_9_yxy1FhjrMF_BbM&uQPQ7vtBDLsNVJB#amLlMu~Nd$FdGS
z$i`3~N(aXz&o8*{cTI?ms=d+y`yBlGxk1gw>MNRUY?rHhuC-#)cD-3BuV1=kwBdw}
z3-EtRzjI!<SmO+J4O&hm+9|o<Tyb1U!Wzo1mWxm$c11Tw`e?6p$R;2b_xe>*IHm78
z{YR=C3TOfm?<Z2(vh>&KYj-|<hv80*p3sR_wg%>0PnuRGxX>567@?SJc_xJ1i1El$
z&+O<~8+!t2W!A^j3YWNRr|Z%&#)NCxX=#6Tu0~P)hOv)@S|C0sxsYzV0Nf}8wq-*p
zxy#YTpUa|d<G~_MqrVWsB%cWdA`n!1G(mha-;I}8yFG_QxUT&!iY>73h@gi1d7;Ac
z8)Zt?2RZ{cTxe2$9+WRqdRD`je|es&;?Dwo8%_%fE4}C_Wxg4}6)n0cHC$a>6uy7#
zME>wxGf$tS`ZFpCoAyj!YZmD{#N#h4{z^vH&fsP9Up*qH88=DS+h4j{#&P7)AE*`C
z?1pWV1LDjA7B}Y0b-o<6?c3IEcT}!+1ub_RZWV2^w9V-*H6!;;Vhmr&WKt9AlNeuD
z*%6B=SkPTjl73NxG_q<S_X)9Lqj!HB-E>vr_-;pBe2`t|24+Z`@#<`Ou|P+9V8*K$
zF?{?h5BQcU%N3#;BME(Klvlrg;m>tYTuk@t;i&MOdi@6+AE_e2&yde!(h5_FMbkCD
zXo+LX-Tdy%7(NQUYsBZ8Z0=&WI2{o?S2U^q@!80cbh4;voH<<A(GNbukSc#l&UF%}
zs<<U3bUa?MsfB<m8a2;@Bq@j|RR=5>eSBW0xB>y2%JUz|zbhrApjvvn-^dim#n8ll
zJ((DcjXi5`hcKN8X!&QKVp_Ow5oxIJ*!1PmT+0i~TIST_Sa=vpn3WHe#NO@Xdgpk)
zD*6$l>wl>|_cG_{<><`7P``gjEz5CX6dQ-pYanU`cnE|~Qpv<%QGb%;9Qd@6wPyD=
zw5d|hqz?xN<|atw(wnR4WFCMwG94$Ces2$IR>kmuV0KegM_wAz+6!kpy&nq{WK|cV
zA!FiS$K=e(D7U!uzl6=Cy(xrU@rW?8HwCAp(6O}so}KRx8;kP9-35Qk(jlADH3&C&
zKde4ge3TcsNio(vbE$;8^@c&W8SIxNE}puvpoXK?hboRFiy?Vq9~B<Bplh2#U+kat
zMt=Y;-<YD`_NqRK(V6h*gA6IaPCmXFv569t#om$>>V~=u*$y50NaaA~iQl0{v0zxg
z!@6wFyS*6BT%SmOqkw-h^iB5|@^+O=!ZU<?+jUD0q0)Jq)S_WN)mF}v!Vg>#=l7;e
zXyq-WhpM4U%Z=LJBv%!`ybdz@m=t5Cz;zHrO&{}jJt(iL6}SKvLx28ZsMJZ8B=G(O
z&p?J;&tgg9_3DY__U=az3&svAD=o6~nZgmT12Dt@$~#n@nkauBCwpNL$Sxr2A>3o_
zTHi#?!iz4*#l$U@dUCkmot9rz4ZN#PiM=2keDTh5FTB9KQC*T3*jHhOl}dEJidzBY
zhR`9Ktgn4-Xz7Cj50Fynwg$5jE~v<oQxHT#SuG<N=-s4uGn`A`Hh)rHKgB<HDaPYH
zZv6dZu#lT(^M!x$@-V-PotM7kLwL>S!G-IOM;^06WLd1cB27(F*_b;6hcUa^hj%A~
z&bv~ncQauEYv(Cftv*mv%?znn31=CR1VrJ)D4vziaoq9I_ofWaai@51jKyxgD?8-m
zM*)?u2DCzXy;5N}i+VU%eNT2{r>!X!LV*dm(mGzmoPmGbU?{C?_0)llrGv0%I0?f-
z_1}u&(R__~za|=0$V5gL3<0R>lqbIuiFH>-=L|PYLw`5+a-TOa>4?{BDe%W7^D}mx
zxPB6YOQ#5ymA!t0(+E6$UclPo9gJp|c??Uyeckg+V$S|m+%n_}`N{8z`{o4Wz)!$O
z|7`B1A3J{@kgLiWQ0?E{QmxkS4p^h&$*Sq$SE%LmXHaSLnq%ghz_7oZEIzDZUe;gK
zn}oV<`)LTOJ%FBZ)MCGp?Cl+Lf-|W6?3#=VP7o1Y6;AereN|8$UC=Gg;o$BT+}+*X
z-7Q#fmk<u_?gV#-00(z>cMVQ(cStT@-M{{)TXkP%R<G&V{o1v5%MUM<aNJbSkZ;?<
zkpPEIcW`+!rP~)HH2G%Fz({J@WKYRoH?1~W0f|YJHypnFAfE0-C7#no9)*@m69du9
zMhQjt2Jh8Eq>w;+D=)QWR1vEiqn=^e^_|IPCF`&28p$Rw1mrn8`NjtL*o41QY3tzd
zuHiDW8;NbFe8%fQ|MpmVTlD(-L1hbw$`Or#9XA_uxQ`hBxdgfSBlPN7l!k9ZH{Sd+
zetm$W2r3&95K4<Jqnt(;fALt#D*<5o%Y~ekMk`~9Ttmo02L6>&)m%xwkK?IB3yTmZ
zEj?&B=4V-!(-wMDc(|=o6XAIaRhq?2K{@dXrMGMm?LGFcZd$bXglV*y9)((EAbQ)n
zGbH+DAUUZT*3vITmm2B!?a}#5O*mGuUxtFMR1xJ>pb(|!HmDF&v(IFj=d~#dAyKQw
zr`HH|Drb#>f>*DZnhQPlrX2})a{U`?9sk&6w1{TrH^5TKs@XO&*-zBvqaq*~eIciy
zi;#+BW-O&%SJPkmAiuNk1R1<|npr2!D>T{sZsxmvL*PADMw^QnAx(Q<GwM3HP+7xk
z)t9J2kjQzQrrrveDgaixdZ_4JLLb`%9LpE#8{uxJQtr-000&FcEZ;T1keBOF39tr0
zT5zwpB*INp<#-F26~b5&mQ7NQ!eJjc7jr3N?1d_83tE)Z{b{0nDrj`SaebhIyd<iz
z=sB#&A7Q$O`+2xSBlG3*z2vuasVyJl3}vD`2zBny6q4>s0WLtL;Bc4Z{8vTRAnj(N
zM)J8hNb=Ui&MaS9I&jE-LZ&zJyj5~t<5+MxPrj`guR+}J*#0%hoLnkse86fX^pu4!
zc#+yeEcXpZ?JiA$=0c*&b8f3hL28-zZ7?K9Vsld_`;AedQk@{w5Xv-%>-K$ELYAco
z#D#R`_F$_bz$I^a`O4GuE!!YdUnFo=N$Y18?KdM6Zy~MQ2PtXd>^<Be_N96(6U~&d
z4>o@xHVfprQ_GTK8Nf~H;E=AmTr4#X5?M3{nrC;>r5RU)asyn}8o&9dW}Ea#nue`z
z|LVR%?BXj)2}5P80S1)%a6f7+3RcD=2%;o-WdIWJ!UFThuM-v$5&wYncGm007S12`
z`E!^ty&-(|wH3dT8TsFD7kn|X40M^h7^H4=T$&<?eBKx0>VKuv0#n0MQZ9b&H;l2@
z5zcC;RFr&wrgtMn#iiNUnT$-#IHG%4`WuB3YY`d>(Xrjs=946@2z#)*QSkNr0yJ0u
zT@Lp0)nz5h5>KZk(_6*<#D}tuD(qDum0`2+%I{GvAIZoB8|q&^Um^t!HkHKswY5u3
zQlavK?ri3AWS}mLsP;EZaMhvi+`~Aldwed>KkQAUV#-aTH`zd(Wc|Zf8@?)zE`$1`
zfYvD>$XqPhH`;zMiHi#TVbjEY0d%h9O{a^lx%=85_Kyq<uEn01=<#LymPQNS)GdTh
zg58k3>fboN{NUWoT4DJ58c<HUtWe^drh+Ea1d;YSYNh{?m~}kd{Ud!b@@G-Jaz)^T
z;k<ow4yejHk}47hM?Q#tUoQUrgwTlcry((k_UnEQ0KG|%jn-jnuA0B)5fpNEl$=E(
zk661JTxXHQh@D4Z*#B5~1~>ikIOI49bLho-<~(VB5oM=%1f2u<D+k_Nq?F2q=OL<v
zOb7nASyFC5Ts%VLChT~5W*DyX8=3a^!#*AEQEF_2;h>|yveR9q9uEw|Z0CsZB|dFF
zoLKVNctmjCl#5v~UN2aP354L)vv`M=rnOfWBaYW)QN!{J=f2A@KJf1gt*22e+@N2^
zJ-c3RPDP%$BE!dYxQqBm_uXQa2`I}2ao_QnuHainwb&22ZwVRoN3;K~F<_Za+_>@%
zdPWT@h&pj%WheWxP3oSSL?C{C$-_u}#?5igm2jWNy9-J!C2ceo2?dp}{jG{l`{^yz
z@Tz2%>&aWlq916+s&>2%gGnH1qeRzLfp+^$ugAcr$+@lP1ayCE)t1oH_tja>79i&+
zz_5k(RCg*N|DfFqTgiask%}m;&>!d5wiQK{dkGEfhEI(Pcn|YF<7us-anL(`@U*e~
z9F0Kor35wT(AlC{TM877<vv*?K8!32S6RaU6)i32bntQ}bZ2N9_0KC<Xu87jNF!Ru
z&l-Z_77e4&6WkdWv~7w#hArb?v3Wzg*@EcVd$l{!Q0o#6)4#W{tj@Yy!zO<HqTP7H
zc96LODbz`8?=c%s4>FrZ)DR#4qsm@&LvPxkGeN}Tvd!6kMhoH(-jsb~Q;@G|Aez8{
zh<Rx$KPv3DSaImj1=h-(TAYp8cvpu-BrFYT^Q~s*Ezwa4ec=o<m&<6vl3lA9SYNp}
zXBj=_LSx_`0H;B5yv#yswuf$tgnZj~>aAv?^r*?@9z~=y0A-udR8<aWnbsHHQbxOq
zKuHwmjZVy*Jb|jug4&KIFPI4wDe!9eI#W{|$@l$Jdac2{3NAib&am!`v_%bj`&^^$
zk-nSBxBMA?Fm~0dU9?gYjTLgt`}W!DSp03unt4D4+Wz;z7F@aF1_>7F51$dZ{%L=0
zW<Ro)sck%qW}iC#Qqu{Zo#qP(Z3;<2LzDAoKdlB$K2SeerJ=${KybjZzOI32EDj?n
zsg6RPfT(RM_fBUD94c<@)}0oOl5g!TK2<mFO1R4GMlGg~Kz;jfgT6z3ZAqoJ>kNi>
zFjAT0rT}>OgqFv~E50VD0I!ZkP><w6wh#2dHn{GczbqeD=iwb7-1?g6M5@N=YSpqf
z7a6sd0-D#&$i>hNd%W0ikcw#MH+X%_9DJT**e}%Ha{t^;nX%%V?)aX)(yb)d`lcC8
zW9{DG(w5;i^!tN9#4Pm(6+ga|;Nhx=bCrn~Lg{oOMbFe-RbCXp(~o^Z)$QLV=Y?By
zXhD+>niG$FhclQ7k_o9FM{n6?ppsVjefl><K8Q`qGxIs<#ZG4sUC0eh`MvX5XXrFb
zsRU(q&tA$&xD=X9(4~p18K}`&&#5CpIE1xH52v<X^CKFm)mnelqoFj|w3M&K@%$cR
zsB+OIVsn>E>+xCMUi(_>J~u>!yT-+~O-OxXn6PJ9BQ#4AgCbym<vjLIF``lUu9ImD
zLs0%tCP`z972R+b*^%EdON<e5IQ2|yo{O$rlUS^PU+-;ih-qRm^3;ncQtIbCz0{47
z6iu|5e~>p{i~Ig>N9Z#dJa&t|G8~6t`jiwrPD@cGToqe-@z+n*ViQh*I9;L~9=B!a
z1&)Izy=d{%0fijCmoh#04j!XrM+$zBB>m|b9<!z8D}0PZI^f?Y)GmitVob+}1z@%`
zFTzifqz57bK6^^<;K`xWTTnjP$wzobsPuF6Pl)&e{{WRviVeU@C;A89*fR1CF9wx<
zj|aeN`Lyq$(#MECKYe;Jx2Qt_MxfFGWS{mwbbxev835qWvI7fnfJ%S)@;_wDC%F!l
zOeTz#K8^yAZrMNtY(PQraksRh063u2*ZBa*$wnUVEuROFLqqbmkYfWDp&|LwKi?BA
zefR))XlNc*)|Lq(fZ%_KE&c5aV5nt=3_u43#lxBo`3{fOf<gh<f`a5uCU2)k<Kton
zvM5?QTD#kjaj|o;rQbxt(18r~oCrtK{8#G!jNl_xuk(<#20Zbm5yz(wM()Dv7cx<&
zlqYD`{s4V0mDJ?wXkROqT6;2s3l14N)mP5>QBiu{WcPL7!$xt?%iccXK8JwDVL#SB
z`|-YppahD)U5%XJG(*}h8hztWK9Pr^g*&Ir_VX}h4S|wfwHrh-L>dUkU~M%8!vl+W
z2O(~0QLcL~D$X6K(d^P!xL>{xAmksGs)+AIUWEJ#tIK&+KMjA%+rKTtKT$CZUpOG3
zM7+M6V;i(fw-I79cca9KVP}pREHAS?K}~J{&@T5g;dYxl@>f}$lX?Q=P8LkNUqWmW
zy%dq8P${GCUc!CG!y;%Q5RF2a{!p+!#JfrpCY<78HtHGiYicuDz#~Jfb&XAU;v#|E
zS<Dwo@K8o##hhdjd#~IAvt?&WEcG%SX0^aPK_i=O&>n$h3BqC#M|D(XJy=^$qG3_8
zq+uCT&^)feIFT|E$Fk237wG{b$5HNAR0<ge76HYaE`RVNzXV8~lalU`@gQ<Rc6>29
zBQz$5+?!JJlxkA?6c&Lt5{ov$ijD<)7rwfa@`Y|ztUQE53X68edN_i_s*k~`SI>Fg
zpmzjQF3XAx_Uaj5z)xLQB5wzJ$HEZWq*`{Km@b?cp>fACVbvNk(x5-hlwCx~Hvyla
zTMjH?0iP0IECUj<CKo!t#<(3uNjgM7fp~yY9`TMKXb&Qm!*s|O6qk;SEW~k_QH5wI
zqRxU`nxj=$V&vu=BKoyX+J#t*6c(_0>vC^9qWIzo@2B_xNu<+<v`#9aw9kf*HjL!E
zO$gpZohA)$h#Qow(FhyC_H-4w^)|*_HM|1`johIKt$}jWHZ-rsbVuX3IeLkIoZXgP
z5UvTeoLwVXM6GJfjek!S_MB9csGXaGFX|ShBJ=^>Pp4*r7T{@~C3WqaI3<AJ;&=JP
zd8yqzE>{<9D7O=;WG{8O+OA^wxhX!fzy~9}A)ObqQ#?>c-JUax)=$>H25%Nv$Z<-x
zPxkZ0qo8(TAp893CbfADLyi4TCuPD8ciku;(ID{1)gU=RB@XBIb}UJXmp|A%R(ms@
zlFhZDHxcSvQrfKZ93&zWIp3IbtYBEyjVMLZs%<|#6lOJ~H-B1Z*jI~R_Ple@l*FMC
zN15E#^?g3gy5VkP>wfX;7tKd`$Uh+*TAEB5*FZJQOlKsvg<c3?YD$z&1}NL~-GS-O
zFZQG3Xf5&RyqOG=d@o21kE^w@1Vmc87m+fYRrN{g7_)ua1e%mbU*q709`h-dDPj4(
zFztM3b6ju-Z~@i8zg1a_8eJ9eN5t}cofMlz|M;e;W3$b7S{tdIuOT_d16{+dshmi6
zB|w;SO7CAfvB4p%%8(AuvzR5cTyiKtxtjo^T1Y(70xL*^AM>2ba(L2|+P<FH^3*oG
z-lAi`XY~B6G?~-p{b=MJ!b3m8OY6Ezs)<#zWuIBG%w89IjHcohSqeS<`B{03dTBZ6
z#H2==*ie@Qeu!U75Tf?LYHd~85uj)h35Zi25F*wld8$=qc#+v~#xdPoxvH_a(S)zO
zh2urPmfbc*L0%upr>edQg_w{8U}de3i3uFPmB(bethUHIkY|<3MbJLxf^C>a33)>|
z&TlO`54vsJO!CumBHM78in;KQ*FtL5^;zOs$C(4$0UPglzc=;g0@R~cs+W(sK-zlx
z7Fv60FcG}CTxTGt<z_3=bE*&+ShJ3NK}ch@Z_2c)s**Nb&oS@OKc@bqlaqCWWwqzM
zZ0s@+w2<OGg(?|L#|dC)rqb}t0Cas?LxTrNZn7x1r9lAnnf|)(X>5rn`Gz=gIqLF5
z$1tKC0r<4!ieCv~#5fjQUoElDK==Yu$eX4^?hsU2Q67<7@d(vjkz1}Sk8MX-n>a*L
z^cj5ceQI5Bh7>gB_5*=kXtO2R8Oi8`rghFoX+{}NuBxK4l0u<BG|jPApql1}LKD>M
z8A4DS_sKi`V#}|*nx2g0QY_@7Lal1s1Gpf!$zVovD#eu%(B^lN<sLGvK$T{Evd`vA
zl$Io;W=6h`lg`j{zwRUZ2!>envH4D2-1pPIj+f2CQLfrQ>OYFRI==ZyXvt6!T!2yL
zA*?7J&h)6y|D=8;JtiX!jW8FqX>9j+{i;EKW^T?i;3$eez20zU{S2w^fZ1xL`xZ#W
z$v~r0EfX0GZ;L7bUB>qw402j_-jQ{t)s`4ZkWJWkW<`*A{=#n>yMHZ95UHn+IpJG8
ztO);LN~e8y@^b-Zm?@o#PyHytikMvj7mCJWR`PsCZOA}YU`|<R{xm^^yZ^Gg%V8I`
z`9o0qiI@gGusE`swZK`Nv0#+|e8@TWD}Ig&Cy|3E$^n0~P&ec|GRV}apo_dgr=E&S
zO2k?J8y|+kBS3FKkdg3`XdC(012uvueVI%AAbPExov_{zTVADS6`5QWwgLjZ1F=H9
zMbX3LkF;Op^fg}<()o&A4lFCaBs*u@!z^<lB8OA!maYZsCu3qaYZP-*-VFwZE?6kc
zHy8g1Y%Dg^J`_&x2?lK|>9bE&`gh+n6G9wX`GJM%2rF>BW`t?dHshEN&v1;#+H@Q(
z<W%Qe%|aR)!9K9b<B3IJ?4%cP0&gL&@(tHlT&Rg$CX8BSOXxkN1E@)tEH@n+|2<B~
zQ1r4`H3j9e9e^Lym(eP84(R6kHz<I(svCFlj-mQwI6iYmE68z&v6R}gxF#cQ;_AZF
zExuoIyk?s=e&ZKOmtb7TY^TWUZaBrw>%<ZmJsv0hjN<0>b>Jr{n@!x{-5BhVpWq(s
z4E5RJx!avGi2_ElrHkgrlh0DucSQY#)*}>lzbG|irOxbHoX_5YlggLfaOqB6W@Qtl
z6|j7#_jNhC3W%9Huk`8X)0D_xF;=UIrPseNU*)F%$!05#2lp@3H8wuHM7go9bqZpj
zvZ&x%H%_Sdbxcs`B3?6e9}l7R`N?ck9kjoK{(&;n^u3vI;|7D*M&cI^?=fQhtvfSz
zgasq^Cx%|_U(p4zFu!dF)2+t)IEh;kTO~+3wx)RuiGp<CUfq-9YqLM=98aUVGw6SJ
zTSo7w&0Gq*{u~bEbDkUaipk~DbJ-cAuBF4)qEc}h@pcMy<TY&^puINgHYldFaqF~;
zEv9f^t<vr7jk?kLk*ox2EM#SkKShDdGC3RLuDSybpbtBv=P9-2PQj!oLS4khuC_T)
zpy)fqw1V#6q!ou8iUJ>M99B=q!)|7O{%SIyK5tf9bxKLprQTOs-2TQ^->%`(sakIZ
z4S(8CZ?RMDX7?;VqN`CmwK>gg@~eI_=uiQR1o@&f$*_~wQVHM5%&itfi;t~t`ol5x
z(AuD;oXp<$jjdD8fJ{4s4fY|&&c+tmkPU}l$Q5+!2ODq4F@7}4TK}_Arc*k)yyE$l
zikJCctNSG4m*0ypGcD?y@nMJp7&1EoEq|Em6ecC=6j}`JhLyE>%Z{X#(CMQFMfvV`
zvw|I&Oky63a})j@CNg4gx9DXZ*X+0VyY?6ixn)YUnds9wIp$bx@OxmJKX^w_749(n
zMP3HAVViDw26$)%lr$LpUgz&fNN%FnA#5}G#@LzJ<QF(bTC0X({qWrmUz2XRb*i1f
zpCr<-rKFfgOSvbv-`1$kKurhVBX-d3#rV2o@hcOzmei3MeH&9Ydz#a$e$}9d8>1GE
zYe~X_lD=hYnCJYpvJrEI#hy`VXu229s{Rqky*V^i%Ugt@<hNjcMlHhSQ#(z-*xn!C
zCOo$FNSsis1M6S8UiR9DZQ1~Tjdu?>sUNL*8~W(-a;t1V$5|;7PwgfwnEfrXb{8E~
zBkM<Ng?prS>|zGF0viqSVYD^6@n%~wL=mc=p{7z&rSFV*Ogutwj{+7k>5yh{%R4(j
z_Z0LvBe2dsPl+=J+&3PILP2{v54`nTbu#Q<8D1JHPc@`Q92*TZ9*<n6Pk~NQCWD7v
zT#ClQc4&t6cW)?{D~*Sjd2gG0;xc++H~%jF1$yF1lPc<zO{^{X)=%EB!q1JU<mL+c
zT$?=QZGEqzO%eL3B3wgYATud2&FS?FlA#^HMuntx+&0QlW?5+^D_!Modv)P|bMGF$
zGN*M`y4xgv|0^k|r*hP|(2<^Zv>>fcW}?F4-<K6nano#H_=RjlMsoWfn%9t!i-{l_
z>JY_^?mKO6MA@FCAm+@_%wFiT(~Wv3M%@tMP4u&t;vl>Wy|VSh!dZ0u-HEjes6B&)
zxuUBzBZ7ks+b`&uR@grq;qSkzif36!6F-dOz4wWwmwWZe!8~(j=E0KIY$N3%amMm4
z3>{2h_ua{e$0ZZ~{E|DLL3U)%pnCP-w^tqqlPTfoCfz)iGL8Db$Kr<3JwqZx);Ch`
z3h#F9hVRx^heh~oy#g!4=UEX)pxrav&#U54$-C>Cr2WFVG|A-aWbEkQ@~~TD!{hp#
z4Hz-$uHmNk<p<`POzCTfzOS}?nR!#%I(K8Z_3hNub=J`n?6YMYMB>Nsb$MOB{&{~s
z=eef?31G1g_7*@=uv5a=okjKvRbB$NT_fx8UBSa2^xILt2d|3lNK4PgK++vnd#Ll-
z(RS+DRoZN|R^~q4WI4~03fgyLr-s!uE5(@A=s$2uBy{f|5Bjor;xncEQzR-a6Piy2
z*k~`UW}(>!PzNEu)XJpjgrhwioP2~oaj3b(``%4BQ1xJHkzb;}1;@^sB#nw-;m?#+
zX)zr#&KHPw@r6A8Rx$2;1s(E~erz-EYuBb$-2UG4VmKWOy}IvoYV-$PYItOf@caBd
z$1eA6i>78UkVrc*plo;16TV0}{=vRaO6T60@2n<1L?%pzG#T||;tpkXm0Oek@sd`K
z>gwd=|MAhc|2vbeqcdv?l6&a+hj+t<Ys<iyN*Bc2o*y@$VeLg}5Oi?wD}<elxHtFB
z)d$Y2H9Pw+pOISv;|CFS<lI8znxF6*an=luUB;3<W^oy=wdwOcJLtnhDAR4{a^&(w
z_WAGrF9?()hxdnfU#YqGy0a5tK@5NN!~OX{c1shg4~@v7Bg$XgMn$!BI553NQsCPA
zIUHS}2rY`>-;=**;h-7a^nMg0_O;%mp^aX53nNzn_pv3Efzj@s&KZLH81Nx-nMF-#
zU$bx=jUF_|1l%%DowF*IkN@rcQk9PB7)soqSvd5Xo@k%YE#~kh%-0q<z*`D276q;3
z`(KLh3&gu~&&!!`FF$DSzKaVRN?Oijn^<#f6)zoA?wNxumO(nlV+LVpC{dg=sLIOQ
z+Z5ZKuMXCZ**^uc3HE8WA$y=5+a#*njcx<7eQ7|-5h5u#N@?dvWrQudop8v8y$^o+
z$}|UZ>}SkY=?8(#wZ~^`OYaJ7NIPPP)7RcoC@!A9LNw(Q2oE+h-#pohj)m_~qGWrz
z{5=1iU|MZ6Pk{<qbBJn_DU9DrulhQ++PU=lnri<R2km9!uigU53#}HHTs_$?F#QwW
z-Y<?v>P))H`%SKDIsFXI_u8-CGzpHhcb}yY29oLjJQAPB`8K<OLD-}l(c(gXhl2es
zHuC)a|Kdg6y*^CC#IIUki^GVJ-eEtboVuXK6JFZZpn_Dc66Tu%OG!=Ed*@Xw_r<24
zE|X=Ra)l>~FIBNq3|ohnT4g+cE1zCgUy##D<n%pWau28O#Ls2h9Q-ZJ1OF;CG|5fu
z&S6elV0OcAKWiN@LN#7S!+)Z#qO)qg`sq7)EONV-PPLVVqt#=p*6zQgZrsyj+}Tzv
zd5(k^IsjrLFs?^pi;{>juV$5RkdQ5N2-({OAYX~Y0eoJtaVk!aNgN4_+<B#gtHQ~}
zeIpWT`HG9H(AW(CR6e}o!FIvq;+`VnLMUqvA&a{JhoxJIl<v}%R08~_*re`I<0r%$
za&d!iu~`8wBzj%334BBaV(5aRwDuJqCRR)F8K6yo@FPxbxrr?D&D#98Gf%eBSepkX
zKZ~9i{JqAKMQ*L<)^WDe*4Q`^q8$GVHc4xRf<!#8bG3)~Y3TzGL>ZyUr*J$6TX)ie
zD6wHD3inTDv;h8DvJID1k3-gR-wlyRNYNo2;9v`0JHdI--sv^*AtB*~8{0!GV#-kl
znWgJ2nh_Qm@Sd6}^9+uXnm%|E7NLuLnBRWxqf~TsiS4|O$W1qt9NXi}ou1a7>UbAO
zO^=;!pZ)VbQ6+smI`_x$kdr?==&6NopxiHPED~r8IKJ|u5j#b@ed5VBvkY_a2y<|F
znfAd-4LIcU8^CW{bg*Y+ZEPGeAwO^eS&n<PKj1TeuH<?Y2H_g>nV23!M1j_K?5taJ
ztPd2SMKni$%cos2y4CdnH(Fz%PLL+czd#%UlYx+>WZ9Wq5U+O+zRFttl{WqwnuJj^
zp!qyd-=IwC)b!fVN(iHB)lCOYmqMnttXP(`g`O7^#PZl+uPO|Hcgt2shj$(5cwxrw
zFAqT4G!rF_D14}6gpDqe&9lYWl!Dv9VtxXNe=tLb$vjlYYw{Xxg}CsIw%PQK+$E-W
zwVr?QN5{sW?6Fu+80|lJr02ZV4d0QHXY2V#e@xyA4Ueo^7)4vDadf16bSKCEW*}Xl
z5dMRH|APij1TFZCZ9X3TN`S2u)aul7X0|snxWmuJD}YjUPBJj)VeSka>(rLOR{vwT
zgJl21e|RJAX8B5Ja2e3S_8->13Z=5s<&Nt+9W>qhFKo<BsYPe~+oIfK`RxzsW*aN%
zW(<ZH@BoE?ZDt_&9xt%nqLYNxKfs)nPUCy=l!ll*GU^FsAv{JRNC>`q1WVP6UDf5c
z<5F@2Npi#~W#PH8uFwG9kb+0GwkwhLcFnMO;5i<P7i)L+mUZVToAE~1I+?(*LYEBW
z+xF`7b|e$q;h=GTcTqTvV+a>uv1^A7hp2;A2r&PVBcwPY?v7gk57`c1hFd){dm{e8
zs=i|CE5;h@UBrwhjB4&uZZxO+eg<_zhVnE`6y0I8ioBmaDe&DF+m41KB8hSTAEfv*
z?LE-a@vED=tEHL4f5^%F8<2;SGrfQp5D3rC#>@JtZl=@m0f_$BP~-!|e#)X;pV}uO
zJ{LETMcUTY&7F*soA-ZJP<9U1^xZ{xdXS!8{QfYyf1XL7j0(in;}3|iKs`dSAWF}K
zeL7Ygq6!&_7)g4;$Ll&3twuTQu~KeSXxqxO=hLcFs|s7GKHxrZ+Q&?@FO~&~A>^}Y
zGW5E3wor8vYaDxgRJ=l5;W@x^2GJI1F&K;$3R~bDY5}xIw3Z~--R;BYfwjgE2Em4y
zcK6D{Z3gxR;d>upATc$<>THYX!VyUCW8f8tOz8D>HD8C~25}&p^maMJb&G0V@cIR|
zPUkl;Q>O2rK`FpEw!vm=3+rA2Pl#S`GB9YNrKu=05RGY1$jSrk*~A#g<buh`h!1BA
zW<zhM2XHlPYE{>u@F#_JWknfKLAjwUU`wPwqCsf%-aJIJdSqj=WJsCp>jiO8dwu=#
zvlamCbUw*uj3083rfk1b%Ar9(Mn+X93USz0Wc7GeRzBTumU)r)>w$qckk$uZ0KVoE
zDe6@oS?}}3l_lXR$zQFLf82ik>LVb31QTCc9sO5}CC~iK<-hu#t!bO@phcve!|xy0
z`|*@Sivpn>R_KmEw4%(pfAi`yX?t;06ubqtEsoCz(jC=VbL`L%uXYWZUKy@~vDhhP
z2nGjqW?42#Rl06k_YL%MlIofFi<5Qi9bw|=G5TLSCk?BpR)VZ5b%$L`b=rR@$hB$X
z*B(wbX?kq}PN{>Xr#a=DL08hVNmMq7Jm%C|8UKZlD9(t)GVdV*TWhFkOeoHpPf!+m
zdZuM$9y{`2l?_HC-S|=GR{dZ}@k-pynd!W<e~T-V4LjL5akdU;YF|-pr#4NjU0lU9
zvT7Qz3VD&8dovIIP&XFWc|Y#R%Nlt4cRtY!f|<66+3OJL;#2~e@ci%|AuYgs(We*v
z)GbrxQ%;*S3l?MMMV@wpn1?bZof7mtDqgW;<Stsc7S*On5Ko9bM%DHzU?71CQ~+hh
zYP1YpUE7fH<<X<^#$^=N;G=AQFfF>~o@gOXub@0FYv8*qy=nR0UF+n&5<|R#De{P$
zm%5ZMrxbkv2@?oM3|k&66+%d#nG%kQ9}%f&T3?`-@#n%a8VXekNiP*HJrGt?qA1|j
zS~bBMkwRJCM-j(B4$}z*7ArtT3`*6wu0UAZWLrq6Uu{6AAC=5q?oPE&n>H_NMy9V-
zB&3~?-<eamn}-+OYS|pxchR>hS7Gs6?T2c*s$n<aJ2S{ktK0&hv!f%>u1&ge>J2oV
z+F`BixBeq6SCcpKh&CM?;8W9X7~XweRSNefPOYT!yF=~fcaMthdkI2OF-EMP1wzP6
z6^`HT-78QW7PQiLvo|9h2oE*hdNvPjeH|q{=;i`i61?1e1_Lt#1i2kvzGG`#u+G%h
zp&a7(A|!$qF#IaGE@IW}c)3_u0t|D+iPvupKa7`XW?dIo6ar3xYsAL+Ho3{E3wGsI
zDCb+C@P^AOx88=IjB=TSo;e2<fjtaRZJZ;{!`;Q;&d~N+6sKLkFem}Cwz^t?3_tL>
z2I0ZLrmdwGxs2O=Yn4V3v&3y_U;!8qMdgv^+YSn;yfmIDK&)?Vc^`7ekZkfTx0_VD
zKg5kF`jM8UWre>WNU?HcNZcLvs<{-P^JmrE1sX;Wj}EQ!mCeyk<)p*^BJ1(`wVoxf
zK*V4ZpL4_0yY|~<P8T`@=S6WXgGIbMc7}e3o|^5D<y}wK$gol_qVCJ8k;Dr+u_>*J
z7!v5i&SqxTOSBwGVgq7H7u9$2-ttF~<@XJ*lpqz$En?#=oU<t{=#gVs`4S<^T@8&M
zZRdvE0xCu#?g9K@*2-HxO&b-0l*6Y9`n*n$+^88PH+%TaRq@c*rfO6d9lM-UjC_L|
zK$J$@jUD+k^8%Yo^4>c;vN8{yaB*4SIW=gA&nnxJ$M(Y10MDu6C$7u+wH+Ep=-DC`
zaHYb`gtZxCugLo+i8Z;f`N%QmRKQ+-`&+=x*yPqYpKwmu*k|>7UovgnuVH3M*^B#k
z`7Z|^Kk1Wcl|{{QX$O039ZiitXZkXEkAtvGp>>t0bp4JiBzF)B%PDbw?9!mrWnF@D
zgQn#LemT!-g}n}}dGO!MAWYX<6$Cd{8SV71EvcWcN~#~2os8(wp)yRj<pxC*oh^cj
zV7}Ni10stU%2MKr7)tDjj4W|*rE{7*19r2kC_N&Is!GoX=av{vAq3Xv-JN;pD!-67
zd(y{G2yTOKTPa60lJ=5B<ITX*7(t+)CwTZ&K&m+T01}FFY)6;<hSq~klkXQLF5?CF
zgo3Q)Z{9GfkIpO#--6z}_gWC)H=RLvEy+`b{YPJX2~F<^9%ncmD9K+_6%cIJN?<w^
zS)@eg*D-r{>=&##dDXU{BmG?3fN&`{Cs=1;$N02Cd2wN(MCM8z=H#E1y)&TE`h)MI
z=yUqtqjPKb5t1dM1#F|WyQ6cm6D>q0Vu(6?&rAwPJ6@>jrJK3i(xb!y%(y@@@7zS_
zL}Tj%v{fgK#rl9rE1X|J{o8WVvB*u}w)B|Dz5|@l$i9?GM$?|J^0urhxfO>RF=ifC
zkQTZint})?2RD|2GQf<33kVcJ%l`0!ruG$hZ?=gak|otMGB9$<v`3NkTaIyWa^14g
zfPR-aVmmtKiM%6YKUCWheLm7)=!(gF!J6s8X3U3s3;DYGfyZ^(!f*P41fm-2>3h{`
zhB^GZx3PS#0hdK;+w%9S8^wnwOqD;4%fDqeW3q2f27^svjyZdL=_5h=XX5aA$_w{G
z^3HI3hzP<5ncb+u<QYud0VA-%wMZV0YUUM#yD9Yn7cS$ghIP{<)-`;8Oe!gCR<XvY
z|2_A+zY}$Bq?B$oQ+8=~gYweEOT@<Fc>dlnojq49SO+!ah<iF#HS~0qn$pD&CMn@s
zamUtjZ}G2PSoqvn*m8g<sGn@}8^4&Tc;Rs8m~(H18SFkCcXnAT0Ld};z<KuVXU!+L
z8A>7^HH*Im@1LUbYW?inlffsIE2R=M^qzHp2<t?`QBv%E>v+1vUwn~=-txw~1$k)v
zyj+8fTNt|@UOfKHXc1E_Or*5ppI=z+(Y@F3g+2Yd>@P;LT2Tdw3?U&tiwTFmk-E#}
z692z4EiN)vvj1`}K|vskrjN5F8H=vEr4^7x&he|IH<<w|84oA75s*d0*4Oes2xQSA
zGhid*AY=O^)t#K&KS_?yBh>%<!^-K?;`|>=lFUF*l~0<Ri%*P4ijzx{lZ}l{N{X9@
zM_f{hi<6Uwi<6H_N|;RW|LgJ@=zlJ^*?5!1Z0SK%=?eBZT6kf{S=0aimB0eKROJoy
zeT%>;5(*PtDc|Y5XpvG}>2~@RkiIJF?r;qoAhlC&a~pFvBoBs!Wbls3DGDMQ*>F-b
z58`>0xD-e$Ip?M`=jL7s-Q{t(9es4WQAeF$y}4iwB}%Co-LL>td86)zY~akn+F+>=
zEi#apokSG{V<7uF*}8Z?uw0r@+(48Ubur&TJWE9v(I#}t2I#?76e48v{$*KmQ!GP#
zLA8C%7o2P}iNg~xyyr3xP->ZjNz3OdbSZP;(2ZISDV4E-oS1!`^&W4<1{c&wF6_4i
zAKKW>>K3#@yBG3tGHn58L%E}dPlvbi;Xw}7vkem=UHPaELhDrhsrjT1(CbqDTBZ;V
zr0YbxWV*4Q5EoFpXta(~f>-PUDt~(Q!LT;Rfbc%Rx;VHo<gAxgpLm^d7f(078E}EL
z3-=2hne;11NLH^Em>IkfV$iD%*0L!BjP!AXn{4QIVOHas(Jv@>QGQ`M@T?>BgN#gh
z2)0K{5uO6NY)0ndFqMOT*p$u2p(_V|*x=5^VUR$;&oLJelMBM26d6eiD3-_V;nol6
zWHQg?E1nz`5K}H9=ZJ2^m+n5W(*4|?^A~3tv;|Apv~w}-;e!Wk-T)6!(}ip{r6cnb
z0`c6@&FQKv{wNnjKhq>jck8orK@L3RVTe)<GbQq#egr9dyaKd*V|}U)YKLLWY7Oal
zrii`cYy6`6q3ktFqT6_o2AQt-UGnzxe6%@od-;^q9?)$BM|$8^3CyLt4Sj!wHWvlR
z{WfV|E<!!RG$>cD1owiu!e(5qfKtPL+Hw3Oj^qK^F?*q|Vn^NSeBjtDNK`(IMD$CT
zu;^>3zvz5uQ-C{yooJcVN1e?$<SznKpE3n$2p=qI$Tdt(LOx8{h#f|PizSHGt;E5=
zYQP3W8(K3oJDKTA00J~M(ki2Ulg%DFa<2jMcOOwDVL3YrrgO||{e&}Gpczf062$V5
zEy_HZ>7nQz3=)y-ji?X+MCro&b?$G~&sZWPN_$oPSPme=C72nHL)8(=6A7z40AYX%
znBp0XA5FU~rJ$T$G9ZdnDGDW@k5q0}3mao_KPF(WWs97FiW3l;kHQEQEheNI=mbgs
zFLZ^Zs}PbPpmSLJ;JVSG$n~}6F+e+i{5J}J0Pr1{->{9WIjL0v%3^Bwlo)t|T0^4Y
z@A`T>P`}12Q+N7IrutBdRJM5Ro+>qiYKygn*2G()TRMQVb$zM}PuqP$Ejp&u0>REg
zSj2@=UPcGYxdls`bwVOquUO%G5+cQuH=mzu3Ot9PH`!=~ZrHHptB5>7cg7-u9_iy#
zZ9Vu9;acS6g;Wj#VP!*=^&#a?T7F8FJB+OmN4-*H3x=gg9OZS?f=+W64dROXY_TJp
zBn#e(`x9xKpXulhhKb#r4h!8u&iCy)7QgE?aa-?EH^XZsh@NNP=<Yu{ETuP7Rys~}
z_FUgzlk|D}TQY=hcKF}-*dj*I{oB?aNEZW5-hm_@{d+VZff6(9crm}cp*p6dFnquG
zp&4a{gYw;<w!6q8FkyZ9LoEMF6u?~qRZ%3>!*BsiosXCG<<A`|hz8z<dotDZ{$jXx
z45TfU&EJ7dm`=@=>sF!b)lHcD&~-xY{gpq3b?TZeHV>M1mGy65Jv=4D&pNqRphXSn
zg|9WzI@zi~ze%9hb#4+yB>v(>7p?t@8X@}|LPhFVX|h8Eg4@A}R}eJ5JMM!$s)0f<
zjdyJrilHK{S%g?e_X>bxSjxCeVVbC@Tnt#n5UGs&#!Bj3>e7_v_fo^pOTsrTp&U+=
zie%|Uv-829`j4b{ZSDbD8C<)SbhW)i3!M+`9&DTs;+Yy7t>AR3KNz|4sDD(QXIa>C
zgK*9ECMwxhJ0uPl^@2HXI7k=%h#9CC0Jt+}!fAe!ne&igdf9g|8fg$97O@J@W^X$9
zE%$JWR45UyjGrB}nVp(cw^TKt@pb9U@el1+GPusD7@fqE*8HMC<$Ol_QV#XlR075Z
z^?j}nP}<NmxRgI|S+C;6%!3h*AB{6Wnt^Gp0L9^ID&S6RxI2Lfx<>mO$pV#wcKdU@
z{)M*h5qVRZ!O3*hHxJ#DtbdEAU|06?Wgvr>N5zum;ac9(Y4dFkymc1|!@t)K{Vwea
zBCV<!db@3PyArL^y7DEbRQo#ps_NStm2b$ppu=sFCb*h)-zY7GX)M|vJ=5*)ut&e?
zRlPbyKenKhda!Lym}foPMMvyl+i#(M@E{@yYMVxwRR1stHveg#uBEcIw1R?|)@a%$
z?0#=K5Om0>JN-Y8*U|+ULxJ8IzRO8AkGKjq#T<)eBI{#b?v=8{?$B%RI4`$s)||7A
z_^vN2xOu{{pDU}>C`jS5k6a(u4M`>aAP@a+4+yJ$f4dD?D4|OJgLYhcyh-<x2aq6g
Paq;j1si>q>q=Ekj5sT5G

diff --git a/paper/sections/abstract.tex b/paper/sections/abstract.tex
index dd447ee..4dcb08e 100644
--- a/paper/sections/abstract.tex
+++ b/paper/sections/abstract.tex
@@ -1,3 +1,3 @@
 \section*{Abstract}
 
-Microdata surveys often lack variables critical for policy analysis, requiring imputation from richer donor surveys, yet researchers have lacked standardized tools for systematically comparing imputation methods and selecting the best approach for a given dataset. We introduce $\texttt{microimpute}$, an open-source Python package that provides a unified framework for benchmarking, tuning, and automatically selecting among multiple imputation methods. We apply it to impute U.S. household wealth from the Survey of Consumer Finances (SCF) onto the Current Population Survey (CPS). Evaluating five methods---Quantile Random Forests (QRF), Ordinary Least Square (OLS), Quantile Regression, Hot Deck Matching, and Mixture Density Networks, we find that QRF achieves the lowest quantile loss and Wasserstein distance, reflecting its capacity to model the non-linear relationships between demographic predictors and wealth. A simulation of the Supplemental Security Income Savings Penalty Elimination Act demonstrates the practical stakes of method choice for downstream analysis. Without imputed wealth, the microsimulation overestimates SSI recipients by 167\%, and the SSI policy reform would be entirely un-simulable. Cross-dataset benchmarking across six additional domains reveals, however, that no single method dominates universally. Although QRF proves best at capturing complex, non-linear relationships, Hot Deck Matching better preserves marginal distributions by drawing directly from the donor pool, while OLS remains competitive for approximately linear relationships. These findings underscore that imputation method selection should be empirically driven rather than prescribed, motivating standardized benchmarking tools, like $\texttt{microimpute}$, that enable researchers to identify the best approach for their specific data characteristics and research objectives.
\ No newline at end of file
+Microdata surveys often lack variables needed for policy analysis, requiring imputation from richer donor surveys, yet no standardized tools exist for comparing imputation methods and selecting the best approach for a given dataset. We introduce $\texttt{microimpute}$, an open-source Python package for benchmarking, tuning, and automatically selecting among multiple imputation methods. We apply it to impute U.S. household wealth from the Survey of Consumer Finances (SCF) onto the Current Population Survey (CPS). Evaluating five methods (Quantile Random Forests, Ordinary Least Squares, Quantile Regression, Hot Deck Matching, and Mixture Density Networks) through a 5-fold cross-validation, we find that QRF reduces average quantile loss by 76\% relative to the worst-performing method and by 7\% relative to the next-best method (Hot Deck Matching), owing to its ability to model non-linear relationships between demographic predictors and wealth. Its Wasserstein distance to the donor distribution is 24\% lower than Matching's. A simulation of the Supplemental Security Income Savings Penalty Elimination Act shows what is at stake: without imputed wealth, the microsimulation overestimates SSI recipients by 167\%, and the reform is entirely un-simulable. Cross-dataset benchmarking across six additional domains, however, revealsthat no single method dominates universally. QRF performs best when relationships are non-linear, Hot Deck Matching better preserves marginal distributions by drawing directly from the donor pool, and OLS remains competitive when relationships are approximately linear. Imputation method selection should therefore be empirically driven rather than prescribed, which motivates tools like $\texttt{microimpute}$ for systematic comparison.
\ No newline at end of file
diff --git a/paper/sections/appendix_benchmarking.tex b/paper/sections/appendix_benchmarking.tex
index 3f23738..d546fae 100644
--- a/paper/sections/appendix_benchmarking.tex
+++ b/paper/sections/appendix_benchmarking.tex
@@ -73,4 +73,4 @@ \subsection{Results}
 
 Matching achieves the best Wasserstein distance on four of six datasets and the lowest mean rank overall (1.67), followed closely by QRF (1.83). QRF performs best on space\_ga and brazilian\_houses, the latter of which exhibits complex nonlinear relationships between predictors and house prices. OLS and Quantile Regression occupy middle ranks, while MDN consistently ranks last, likely reflecting a combination of the relatively small sample sizes in these datasets and the hyperparameter sensitivity of neural network-based approaches.
 
-With only six datasets, the rank differences between methods---particularly the narrow gap between Matching (1.67) and QRF (1.83)---should be interpreted cautiously, as they are not statistically robust to the inclusion or exclusion of individual datasets. Notably, the relative rankings here differ from the main SCF-CPS analysis, where QRF achieves the lowest quantile loss. This is consistent with the expectation that method performance depends on dataset characteristics, and reinforces the value of \texttt{microimpute}'s method comparison framework, which allows researchers to empirically determine the best approach for their specific data rather than relying on a single method.
+With only six datasets, the rank differences between methods, particularly the narrow gap between Matching (1.67) and QRF (1.83), should be interpreted cautiously; they are not statistically robust to the inclusion or exclusion of individual datasets. The relative rankings here also differ from the main SCF-CPS analysis, where QRF achieves the lowest quantile loss. This is consistent with the expectation that method performance depends on dataset characteristics, and it supports the case for empirically comparing methods on each new dataset rather than relying on a single approach.
diff --git a/paper/sections/appendix_robustness.tex b/paper/sections/appendix_robustness.tex
index f143f0e..fb840e0 100644
--- a/paper/sections/appendix_robustness.tex
+++ b/paper/sections/appendix_robustness.tex
@@ -54,4 +54,4 @@ \subsection{Progressive inclusion analysis}
 
 \subsection{Implications for the Conditional Independence Assumption}
 
-The strong predictive power of the shared financial variables, particularly interest and dividend income, employment income, and pension income, is consistent with, though not sufficient evidence for, the plausibility of the CIA in the SCF-CPS matching context. These variables capture key dimensions of household financial position that are available in both surveys. The fact that they collectively explain a large share of the variation in net worth suggests that the common variables $X$ may be sufficient to render the target variable $Y$ (wealth) approximately conditionally independent of variables $Z$ unique to the CPS, though this assumption remains fundamentally untestable as discussed in Section~2.2.
+The strong predictive power of the shared financial variables, particularly interest and dividend income, employment income, and pension income, is consistent with, though not sufficient evidence for, the plausibility of the CIA in the SCF-CPS matching context. These variables capture the main dimensions of household financial position available in both surveys. The fact that they collectively explain a large share of the variation in net worth suggests that the common variables $X$ may be sufficient to render the target variable $Y$ (wealth) approximately conditionally independent of variables $Z$ unique to the CPS, though this assumption remains fundamentally untestable as discussed in Section~2.2.
diff --git a/paper/sections/background.tex b/paper/sections/background.tex
index 73b77fd..ef7cbe0 100644
--- a/paper/sections/background.tex
+++ b/paper/sections/background.tex
@@ -1,6 +1,6 @@
 \section{Background}
 
-This section establishes the theoretical foundations for statistical matching and reviews the imputation methods implemented in \texttt{microimpute}. We begin with the formal problem definition and the key assumption underlying all statistical matching procedures, then discuss the relationship to missing data mechanisms, and finally describe each of the five imputation methods.
+This section defines the statistical matching problem and describes the imputation methods in \texttt{microimpute}. We begin with the formal problem definition and the assumption underlying all statistical matching procedures, then discuss the relationship to missing data mechanisms, and finally describe each of the five imputation methods.
 
 \subsection{The Statistical Matching Problem}
 
@@ -56,7 +56,7 @@ \subsection{Imputation Methods}
 
 \subsubsection{Hot Deck Matching}
 
-Hot deck imputation replaces missing values in a receiver record with observed values from ``similar'' donor records \citep{andridge2010review}. The \texttt{microimpute} implementation uses the unconstrained distance hot deck approach from the StatMatch R package \citep{dorazio2021statistical}. To identify the best match, the method computes distances between each receiver observation and all donor observations based on the common variables $X$, selects donor records within a specified distance threshold, and randomly samples from the eligible donors, with optional weighting by survey weights. For continuous and mixed covariates, Mahalanobis or Gower distance metrics can capture similarity across different variable types. The donated value will thus be an actual observed value from the donor file, ensuring plausibility.
+Hot deck imputation replaces missing values in a receiver record with observed values from ``similar'' donor records \citep{andridge2010review}. The \texttt{microimpute} implementation uses the unconstrained distance hot deck approach from the StatMatch R package \citep{dorazio2021statistical}. To identify the best match, the method computes distances between each receiver observation and all donor observations based on the common variables $X$, selects donor records within a specified distance threshold, and randomly samples from the eligible donors, with optional weighting by survey weights. For continuous and mixed covariates, Mahalanobis or Gower distance metrics can capture similarity across different variable types. The donated value is therefore always an actual observed value from the donor file.
 
 Hot deck methods are nonparametric and avoid distributional assumptions, making them robust when the true conditional distribution is unknown or complex \citep{dorazio2006statistical}. However, they face limitations:
 \begin{itemize}
@@ -186,18 +186,18 @@ \subsection{Current Practice in Microsimulation}
 
 Statistical matching and data fusion are fundamental operations in microsimulation modeling, yet the methods employed in practice have remained relatively unchanged for decades. A review of major tax-benefit microsimulation models reveals a strong reliance on traditional imputation approaches, primarily hot deck matching and OLS-based regression.
 
-Hot deck matching remains the dominant approach in European microsimulation. EUROMOD, the EU-wide tax-benefit model, employs a multi-stage imputation procedure combining predictive mean matching with distance-based hot deck methods to integrate consumption data from Household Budget Surveys into its EU-SILC input data \citep{sutherland2013euromod}. The appeal of hot deck methods lies in their simplicity and the guarantee that imputed values are observed values from the donor file, ensuring plausibility. However, as discussed above, these methods struggle with tail behavior and may not adequately capture the full conditional distribution of the target variable.
+Hot deck matching remains the dominant approach in European microsimulation. EUROMOD, the EU-wide tax-benefit model, employs a multi-stage imputation procedure combining predictive mean matching with distance-based hot deck methods to integrate consumption data from Household Budget Surveys into its EU-SILC input data \citep{sutherland2013euromod}. Hot deck methods are simple to implement and guarantee that imputed values are actual observed values from the donor file. However, as discussed above, these methods struggle with tail behavior and may not adequately capture the full conditional distribution of the target variable.
 
-In U.S. tax policy microsimulation, regression-based approaches are more common. The Tax Policy Center employs a two-stage probit and OLS procedure for wealth imputation, first predicting the probability of holding each asset type, then predicting amounts conditional on positive holdings \citep{nunns2012tax}. Similarly, the Institute on Taxation and Economic Policy (ITEP) model relies on statistical matching between tax return data and the American Community Survey, supplemented with regression-based imputations from the Survey of Consumer Finances \citep{itep2023model}. These regression approaches assume linear relationships and normally distributed errors; assumptions that are frequently violated by economic variables with heavy tails and heteroscedastic relationships.
+In U.S. tax policy microsimulation, regression-based approaches are more common. The Tax Policy Center employs a two-stage probit and OLS procedure for wealth imputation, first predicting the probability of holding each asset type, then predicting amounts conditional on positive holdings \citep{nunns2012tax}. Similarly, the Institute on Taxation and Economic Policy (ITEP) model relies on statistical matching between tax return data and the American Community Survey, supplemented with regression-based imputations from the Survey of Consumer Finances \citep{itep2023model}. These regression approaches assume linear relationships and normally distributed errors, assumptions frequently violated by economic variables with heavy tails and heteroscedasticity.
 
-More recent microsimulation efforts have begun incorporating administrative data through record linkage rather than statistical matching \citep{abowd2019census}, but this approach requires access to restricted data and raises privacy concerns. For researchers working with publicly available survey data, statistical matching remains essential, creating a need for methods that can better capture complex distributional features.
+More recent microsimulation efforts have begun incorporating administrative data through record linkage rather than statistical matching \citep{abowd2019census}, but this approach requires access to restricted data and raises privacy concerns. For researchers working with publicly available survey data, statistical matching remains the primary option, and there is room for methods that capture distributional features more effectively.
 
-The limitations of traditional approaches can be particularly acute for heavily-tailed variables where relationships with predictors vary across the distribution. Machine learning methods such as Quantile Random Forests offer a promising alternative, as they can capture nonlinear relationships, model the entire conditional distribution, and handle heavy-tailed data without restrictive parametric assumptions. Despite these advantages, QRF and similar methods have seen limited adoption in mainstream microsimulation practice. The \texttt{microimpute} package aims to lower barriers to adopting these more flexible methods by providing a unified framework for comparing traditional and machine learning approaches, enabling researchers to empirically evaluate which method best suits their specific data characteristics.
+These limitations are especially pronounced for heavy-tailed variables where the relationship with predictors varies across the distribution. Machine learning methods such as Quantile Random Forests can capture nonlinear relationships, model the entire conditional distribution, and handle heavy-tailed data without restrictive parametric assumptions, yet they have seen limited adoption in microsimulation practice. The \texttt{microimpute} package provides a common framework for comparing traditional and machine learning approaches, so that researchers can empirically evaluate which method works best for their data.
 
 \subsection{Supplemental Security Income as a Validation Case}
 
-The limitations of traditional imputation approaches in microsimulation have concrete consequences on downstream policy analysis. The impact of wealth imputation quality for U.S. households can be illustrated through the Supplemental Security Income (SSI) program. SSI is a federal means-tested program for aged, blind, and disabled individuals with strict resource limits (\$2,000 for individuals, \$3,000 for couples) that have remained unchanged since 1989 \citep{ssa2023ssi}. The resource test requires household wealth and asset data that the CPS does not collect, making SSI eligibility determination dependent on imputed or heuristic wealth values.
+The limitations of traditional imputation approaches have concrete consequences for downstream policy analysis. The Supplemental Security Income (SSI) program illustrates this well. SSI is a federal means-tested program for aged, blind, and disabled individuals with strict resource limits (\$2,000 for individuals, \$3,000 for couples) that have remained unchanged since 1989 \citep{ssa2023ssi}. The resource test requires household wealth and asset data that the CPS does not collect, making SSI eligibility determination dependent on imputed or heuristic wealth values.
 
 In 2024, SSI served approximately 7.4 million recipients with \$59.6 billion in federal payments \citep{ssa2024ssi}. Because different wealth imputation methods produce different wealth distributions, they directly affect which simulated households pass the resource test and therefore the resulting estimates of SSI recipient counts and expenditure. This makes SSI a particularly informative case for comparing imputation approaches, as described in Section~\ref{sec:policy_validation}.
 
-The frozen resource limits also make SSI a compelling case for policy reform analysis. In 2024 dollars, the original 1989 thresholds of \$2,000/\$3,000 are worth approximately \$4,800/\$7,200, meaning the resource test has become substantially more restrictive over time without any legislative change. The SSI Savings Penalty Elimination Act (S.~1234 / H.R.~2540), a bipartisan bill introduced in April 2025, proposes raising these limits to \$10,000 for individuals and \$20,000 for couples, with future CPI indexing \citep{ssi_spea_2025}. Simulating the impact of this reform, and determining which additional households would gain eligibility under the higher thresholds requires household-level wealth data, making it a natural application for cross-survey imputation. As we demonstrate in Section~\ref{sec:policy_validation}, the choice of imputation method produces meaningfully different estimates of this reform's impact, underscoring that imputation is not merely a data preprocessing step but a consequential modeling choice for policy analysis.
+The frozen resource limits also make SSI a compelling case for policy reform analysis. In 2024 dollars, the original 1989 thresholds of \$2,000/\$3,000 are worth approximately \$4,800/\$7,200, meaning the resource test has become substantially more restrictive over time without any legislative change. The SSI Savings Penalty Elimination Act (S.~1234 / H.R.~2540), a bipartisan bill introduced in April 2025, proposes raising these limits to \$10,000 for individuals and \$20,000 for couples, with future CPI indexing \citep{ssi_spea_2025}. Simulating the impact of this reform, and determining which additional households would gain eligibility under the higher thresholds requires household-level wealth data, making it a natural application for cross-survey imputation. As we show in Section~\ref{sec:policy_validation}, the choice of imputation method produces meaningfully different estimates of this reform's impact, meaning that imputation is not merely a data preprocessing step but a modeling choice with real consequences for policy analysis.
diff --git a/paper/sections/conclusion.tex b/paper/sections/conclusion.tex
index 3d9a412..362be7b 100644
--- a/paper/sections/conclusion.tex
+++ b/paper/sections/conclusion.tex
@@ -1,7 +1,7 @@
 \section{Conclusion}
 
-This paper has introduced $\texttt{microimpute}$, an open-source Python package for cross-survey imputation, and applied it to a substantive policy problem: imputing household wealth from the SCF onto the CPS. Through systematic comparison of five imputation methods using an inverse hyperbolic sine transformation to accommodate negative and zero net worth values, we find that Quantile Random Forests achieve the strongest conditional accuracy among the five methods evaluated for wealth imputation, driven by their capacity to model the complex non-linear relationships between demographic predictors and wealth. A downstream SSI simulation validates the practical importance of imputation method choice: without wealth data, the microsimulation overestimates SSI recipients by 167\%, while imputation-based approaches substantially narrow this gap. Simulating the SSI Savings Penalty Elimination Act, which would raise resource limits from \$2,000/\$3,000 to \$10,000/\$20,000, further demonstrates that method choice directly affects policy estimates. Reform impact estimates range from 0.1 to 4.2 million additional recipients and \$1.0 to \$24.1 billion in additional expenditure depending on the imputation method. Without imputed wealth, this reform is entirely un-simulable.
+This paper has introduced $\texttt{microimpute}$, an open-source Python package for cross-survey imputation, and applied it to a substantive policy problem: imputing household wealth from the SCF onto the CPS. Comparing five imputation methods with an inverse hyperbolic sine transformation to accommodate negative and zero net worth values, we find that Quantile Random Forests achieve the strongest conditional accuracy for wealth imputation, owing to their ability to model the non-linear relationships between demographic predictors and wealth. A downstream SSI simulation confirms the practical importance of method choice. Without wealth data, the microsimulation overestimates SSI recipients by 167\%, while imputation-based approaches substantially narrow this gap. Simulating the SSI Savings Penalty Elimination Act, which would raise resource limits from \$2,000/\$3,000 to \$10,000/\$20,000, further shows that method choice directly affects policy estimates. Reform impact estimates range from 0.1 to 4.2 million additional recipients and \$1.0 to \$24.1 billion in additional expenditure depending on the imputation method. Without imputed wealth, this reform cannot be simulated at all.
 
-Our findings also reveal important nuances about method selection. Cross-dataset benchmarking shows that QRF's advantage reflects the specific characteristics of the wealth imputation problem rather than a universal superiority, and predictor importance analysis confirms that financial variables drive the majority of imputation accuracy.
+The findings also reveal nuances about method selection. Cross-dataset benchmarking shows that QRF's advantage reflects the specific characteristics of the wealth imputation problem rather than a universal superiority, and predictor importance analysis confirms that financial variables drive the majority of imputation accuracy.
 
-These results carry two broader implications. First, imputation method selection should be empirically driven rather than prescribed; the optimal approach depends on dataset characteristics, and standardized comparison tools are essential for enabling this data-driven selection. Second, the gap between statistical and policy-relevant evaluation metrics---QRF achieves the best quantile loss yet not the closest SSI estimates to administrative totals---highlights the importance of validating imputation quality through downstream applications, not only through statistical benchmarks. Future work should expand $\texttt{microimpute}$'s method library to include gradient boosting and deep learning approaches \citep{alaa2024deep}, explore ensemble strategies that combine methods across different parts of the distribution, and address the terminal node sparsity challenge that limits tree-based methods at extreme quantiles.
\ No newline at end of file
+Two broader implications follow. First, imputation method selection should be empirically driven rather than prescribed; the optimal approach depends on dataset characteristics, and standardized comparison tools are needed to support this. Second, the gap between statistical and policy-relevant evaluation metrics (QRF achieves the best quantile loss yet not the closest SSI estimates to administrative totals) points to the importance of validating imputation quality through downstream applications, not only through statistical benchmarks. Future work should expand $\texttt{microimpute}$'s method library to include gradient boosting and deep learning approaches \citep{alaa2024deep}, explore ensemble strategies that combine methods across different parts of the distribution, and address the terminal node sparsity problem that limits tree-based methods at extreme quantiles.
\ No newline at end of file
diff --git a/paper/sections/data.tex b/paper/sections/data.tex
index de0f564..d51902a 100644
--- a/paper/sections/data.tex
+++ b/paper/sections/data.tex
@@ -4,16 +4,16 @@ \section{Data}\label{sec:data}
 
 \subsection{Survey of Consumer Finances}
 
-The Survey of Consumer Finances, sponsored by the Federal Reserve Board, is a triennial survey providing detailed information on U.S. households' assets, liabilities, income, and demographic characteristics. Its dual-frame sample design includes a standard national area-probability sample and a list sample deliberately oversampling wealthy households to better capture the skewed wealth distribution \citep{barcelo2006imputation}. The SCF is a benchmark for wealth imputation research due to its detailed financial data and the known complexities arising from its design and the nature of wealth. Item nonresponse in public-use SCF datasets is addressed by the Federal Reserve through a multiple imputation approach that generates five complete datasets with different imputed values, using sequential regression-based procedures that incorporate range constraints, logical data structures, and empirical residuals to preserve the complex multivariate relationships inherent in wealth data \citep{kennickell1998multiple}. 
+The Survey of Consumer Finances, sponsored by the Federal Reserve Board, is a triennial survey providing detailed information on U.S. households' assets, liabilities, income, and demographic characteristics. Its dual-frame sample design includes a standard national area-probability sample and a list sample that deliberately oversamples wealthy households to better capture the skewed wealth distribution \citep{barcelo2006imputation}. The SCF is a standard reference for wealth imputation research because of its detailed financial data and the known complexities arising from its design and the nature of wealth. Item nonresponse in public-use SCF datasets is addressed by the Federal Reserve through a multiple imputation approach that generates five complete datasets with different imputed values, using sequential regression-based procedures that incorporate range constraints, logical data structures, and empirical residuals to preserve the complex multivariate relationships inherent in wealth data \citep{kennickell1998multiple}. 
 
 Specifically, we use the 2022 summarized SCF as our donor dataset. 
 
 \subsection{Current Population Survey}
 
-The Current Population Survey, conducted jointly by the U.S. Census Bureau and the Bureau of Labor Statistics, is a monthly survey of approximately 60,000 U.S. households that serves as the primary source of labor force statistics for the United States. The CPS uses a multistage probability-based sample designed to represent the civilian non-institutional population; on average, each sampled household represents approximately 2,500 households in the population. The Annual Social and Economic Supplement (ASEC) extends the core survey with detailed annual income data, including earnings, unemployment compensation, Social Security, pension income, interest, dividends, and other income sources. However, despite its comprehensive coverage of income and employment, the CPS does not collect information on household wealth, assets, or liabilities, which are key variables needed for wealth-based policy analysis. This omission motivates the need for statistical matching to transfer wealth information from the SCF onto the CPS.
+The Current Population Survey, conducted jointly by the U.S. Census Bureau and the Bureau of Labor Statistics, is a monthly survey of approximately 60,000 U.S. households and is the primary source of labor force statistics for the United States. The CPS uses a multistage probability-based sample designed to represent the civilian non-institutional population; on average, each sampled household represents approximately 2,500 households in the population. The Annual Social and Economic Supplement (ASEC) extends the core survey with detailed annual income data, including earnings, unemployment compensation, Social Security, pension income, interest, dividends, and other income sources. However, despite its broad coverage of income and employment, the CPS does not collect information on household wealth, assets, or liabilities, which are needed for wealth-based policy analysis. This omission motivates the need for statistical matching to transfer wealth information from the SCF onto the CPS.
 
 Specifically, we use the Enhanced CPS 2024 produced by PolicyEngine \citep{policyengine_us} as our receiver dataset. The Enhanced CPS extends the base CPS ASEC microdata through survey reweighting and calibration to administrative totals, improving its representativeness for tax-benefit microsimulation. The SCF 2022 financial variables are uprated to 2024 dollars using CPI adjustment factors to ensure consistency with the receiver dataset's reference year.
 
 \subsection{Comparative analysis and characteristics for imputation}
 
-The SCF and CPS differ fundamentally in sampling design: the SCF employs a dual-frame approach combining a standard area-probability sample with a list sample that deliberately oversamples wealthy households, while the CPS uses a multistage area-probability household design representative of the civilian non-institutional population. This difference means that the donor and receiver surveys weight different parts of the wealth distribution differently, making proper survey weight integration during model training essential for unbiased imputation. The two surveys share a core set of demographic and income variables---age, sex, race, number of children, employment income, interest and dividend income, and pension income---but differ in their detailed variable coverage: the SCF collects granular asset and liability data absent from the CPS, while the CPS captures labor force dynamics and program participation variables not available in the SCF. This partial overlap constrains the conditioning set available for imputation and underscores the importance of the conditional independence assumption discussed in Section~\ref{sec:cia}. Together, the contrasting designs, the heavy-tailed nature of wealth, and the variable overlap constraints make the SCF-to-CPS imputation a particularly informative test case for comparing methods, while also being directly relevant to U.S. policy microsimulation, where researchers require comprehensive microdata combining income, demographics, and wealth to analyze the distributional effects of tax and benefit reforms.
\ No newline at end of file
+The SCF and CPS differ fundamentally in sampling design: the SCF employs a dual-frame approach combining a standard area-probability sample with a list sample that deliberately oversamples wealthy households, while the CPS uses a multistage area-probability household design representative of the civilian non-institutional population. This difference means that the donor and receiver surveys weight different parts of the wealth distribution differently, making proper survey weight integration during model training essential for unbiased imputation. The two surveys share a core set of demographic and income variables (age, sex, race, number of children, employment income, interest and dividend income, and pension income) but differ in detailed coverage: the SCF collects granular asset and liability data absent from the CPS, while the CPS captures labor force dynamics and program participation variables not available in the SCF. This partial overlap constrains the conditioning set available for imputation and reinforces the importance of the conditional independence assumption discussed in Section~\ref{sec:cia}. The contrasting survey designs, the heavy-tailed nature of wealth, and the limited variable overlap make the SCF-to-CPS imputation an informative test case for comparing methods. It is also directly relevant to U.S. policy microsimulation, where analyzing the distributional effects of tax and benefit reforms requires microdata that combine income, demographics, and wealth.
\ No newline at end of file
diff --git a/paper/sections/discussion.tex b/paper/sections/discussion.tex
index 6e9627a..3600d45 100644
--- a/paper/sections/discussion.tex
+++ b/paper/sections/discussion.tex
@@ -1,56 +1,42 @@
 \section{Discussion}
 
-This paper has evaluated five imputation methods within the \texttt{microimpute} framework, finding that Quantile Random Forests offer meaningful advantages for the SCF-to-CPS wealth imputation application. By preserving the full conditional distribution of wealth, QRF maintains the key statistical properties of wealth data that traditional methods struggle to capture.
+This paper has evaluated five imputation methods within the \texttt{microimpute} framework, finding that Quantile Random Forests offer meaningful advantages for the SCF-to-CPS wealth imputation application. By preserving the full conditional distribution of wealth, QRF captures statistical properties of wealth data that traditional methods miss.
 
-\subsection{Strengths powered by $\texttt{microimpute}$}
+\subsection{The role of \texttt{microimpute} in the analysis}
 
-The $\texttt{microimpute}$ package's design philosophy and implementation choices provide several key advantages that contributed to the success of our wealth imputation analysis, and will extend to future model comparisons across applications:
+Several design choices in \texttt{microimpute} shaped the analysis and its results. Wrapping all five methods behind a consistent API meant that performance differences in the benchmarking reflect the methods themselves, not differences in how they were implemented or tuned \citep{policyengine2025microimpute}. The \texttt{autoimpute} function automated hyperparameter tuning and method selection based on quantile loss, removing a subjective step that could otherwise bias the comparison. Using quantile loss rather than RMSE as the evaluation metric was itself consequential: its asymmetric penalty structure prioritizes accuracy at distribution tails, where symmetric metrics are insensitive to the kinds of errors that matter most for skewed variables like wealth.
 
-\begin{enumerate}
-    \item \textbf{Unified interface for method comparison}: $\texttt{microimpute}$'s consistent API across all imputation methods enabled systematic benchmarking without implementation-specific biases. This standardization ensures that performance differences reflect genuine methodological advantages rather than implementation artifacts \citep{policyengine2025microimpute}.
-
-    \item \textbf{Automated method selection}: The package's \texttt{autoimpute} function streamlines the imputation workflow by automatically comparing methods and selecting the best performer based on quantile loss metrics. This feature proved particularly valuable given the wealth data's complexity, as it removed subjective method selection and ensured optimal performance.
-
-    \item \textbf{Survey weight integration}: $\texttt{microimpute}$'s native support for survey weights through stratified sampling ensures that imputation models properly represent population distributions. This capability is crucial when transferring information between surveys with different sampling designs, such as the SCF's oversampling of wealthy households.
-
-    \item \textbf{Quantile-aware evaluation}: By implementing quantile loss as the primary evaluation metric, $\texttt{microimpute}$ directly addresses the challenges of skewed distributions. This metric's asymmetric penalty structure naturally prioritizes accurate imputation at distribution tails, where traditional metrics like Root Mean Squared Error (RMSE) often fail.
-
-    \item \textbf{Computational efficiency}: The package's optimized implementation enables processing of large microdata files while maintaining reasonable computation times. Cross-validation on the full SCF dataset, including QRF hyperparameter tuning, was completed in under 30 minutes on standard hardware, making iterative experimentation feasible.
-
-    \item \textbf{Open-source transparency}: As an open-source tool, $\texttt{microimpute}$ allows full inspection and modification of imputation algorithms, promoting reproducibility and enabling custom extensions for specific research needs \citep{policyengine2025microimpute}.
-\end{enumerate}
+Two practical aspects also mattered. First, native survey weight support through stratified sampling allowed the models to train on data representative of the U.S. population despite the SCF's deliberate oversampling of wealthy households. Second, the full cross-validation pipeline, including hyperparameter tuning on the SCF for the QRF, MDN, and Matching models ran in under 45 minutes on standard hardware, making iterative experimentation practical. The package is open-source, so all algorithms and evaluation procedures can be inspected and modified \citep{policyengine2025microimpute}.
 
 \subsection{Generalizability and predictor sensitivity}
 
-The main SCF-CPS results are complemented by two supplementary analyses that inform the broader applicability of these findings.
+Two supplementary analyses help contextualize the main SCF-CPS results.
 
-The cross-dataset benchmarking exercise (Appendix~\ref{app:benchmarking}) reveals that method performance is context-dependent, with each method's strengths mapping to different data characteristics. QRF achieves the best Wasserstein distance on datasets exhibiting complex non-linear relationships, such as the Brazilian houses dataset and the SCF wealth imputation, where its ability to model flexible conditional distributions provides a clear advantage. Matching, by contrast, achieves the lowest Wasserstein distance on four of six benchmarking datasets and the best mean rank overall (1.67 vs.\ QRF's 1.83). This strong marginal distributional performance reflects Matching's core mechanism: by drawing imputed values directly from the donor pool, it inherently preserves the shape, skewness, and tail behavior of the original distribution, yielding low Wasserstein distances even when it may not optimally condition on predictor values. However, as the main SCF-CPS results demonstrate, Matching's distributional fidelity does not guarantee accurate conditional imputation. Households may receive plausible-looking wealth values that do not correctly reflect their individual characteristics. Thus, understanding the trade-offs between marginal and conditional accuracy, and evaluating how differences in survey design affect record matching becomes very important.
+The cross-dataset benchmarking exercise (Appendix~\ref{app:benchmarking}) shows that method performance is context-dependent. QRF achieves the best Wasserstein distance on datasets with complex non-linear relationships, such as the Brazilian houses dataset and the SCF wealth imputation. Matching, by contrast, achieves the lowest Wasserstein distance on four of six benchmarking datasets and the best mean rank overall (1.67 vs.\ QRF's 1.83). This strong marginal performance follows from Matching's mechanism: drawing imputed values directly from the donor pool inherently preserves the shape and tail behavior of the original distribution, yielding low Wasserstein distances even when the conditioning on predictor values is suboptimal. But as the main SCF-CPS results show, good marginal distributions do not guarantee accurate conditional imputation. Households may receive plausible-looking wealth values that do not correctly reflect their individual characteristics. The trade-off between marginal and conditional accuracy, and the effect of differing survey designs on record matching, deserve careful attention in any application.
 
-Meanwhile, OLS performs competitively on datasets where the predictor-target relationship is approximately linear (e.g., abalone, space\_ga), highlighting that simpler parametric methods remain effective when their structural assumptions are met. MDN consistently ranks last across the benchmarking datasets, likely reflecting a combination of the relatively small sample sizes in these datasets and the hyperparameter sensitivity of neural network-based approaches, which require larger training sets to realize their flexibility advantage.
+OLS performs competitively on datasets where the predictor-target relationship is approximately linear (e.g., abalone, space\_ga), confirming that simpler parametric methods remain effective when their structural assumptions hold. MDN consistently ranks last across the benchmarking datasets, likely reflecting a combination of the relatively small sample sizes in these datasets and the hyperparameter sensitivity of neural network-based approaches, which require larger training sets to realize their flexibility advantage.
 
-These findings reinforce the value of \texttt{microimpute}'s automated method comparison framework: no single method dominates across all settings, and the optimal choice depends on dataset size, the complexity of the predictor-target relationship, and whether the research objective prioritizes conditional accuracy or marginal distributional fidelity. Researchers should leverage the package's comparison tools to empirically determine the best approach for their specific data rather than defaulting to any single method.
+No single method dominates across all settings. The best choice depends on dataset size, the complexity of the predictor-target relationship, and whether the research question calls for conditional accuracy or marginal distributional fidelity. This is precisely why automated comparison matters: researchers should test methods on their own data rather than defaulting to any one approach.
 
-The predictor importance analysis (Appendix~\ref{app:robustness}) reveals a clear hierarchy among the variables shared between the SCF and CPS. Financial predictors, particularly interest and dividend income, employment income, and age, account for the vast majority of imputation accuracy, with interest and dividend income alone more than doubling the quantile loss when removed. Demographic variables such as race and gender contribute relatively little to imputation quality. This concentration of predictive power in financial variables has two practical implications. First, the quality of wealth imputation rests primarily on the availability and accuracy of income-related variables in both surveys; researchers lacking financial predictors would see substantially degraded results. Second, diminishing returns from additional demographic predictors suggest that a parsimonious set of well-chosen financial variables can achieve most of the attainable accuracy, as confirmed by the progressive inclusion analysis showing that the first three predictors capture the majority of achievable performance. Nonetheless, the inclusion of demographic variables still provides incremental improvements, and their contribution to capturing nuanced wealth variation should not be overlooked.
+The predictor importance analysis (Appendix~\ref{app:robustness}) reveals a clear hierarchy among the variables shared between the SCF and CPS. Financial predictors, particularly interest and dividend income, employment income, and age, account for the vast majority of imputation accuracy, with interest and dividend income alone more than doubling the quantile loss when removed. Demographic variables such as race and gender contribute relatively little to imputation quality. This concentration of predictive power in financial variables has two practical implications. First, the quality of wealth imputation rests primarily on the availability and accuracy of income-related variables in both surveys; researchers lacking financial predictors would see substantially degraded results. Second, diminishing returns from additional demographic predictors suggest that a parsimonious set of well-chosen financial variables can achieve most of the attainable accuracy, as confirmed by the progressive inclusion analysis showing that the first three predictors capture the majority of achievable performance. That said, demographic variables still provide incremental improvements and should not be excluded without reason.
 
-\texttt{microimpute}'s support for predictor importance analysis and progressive inclusion experiments allows researchers to systematically assess which variables are driving imputation performance in their specific context, and to make informed decisions about variable selection when faced with data limitations.
+The predictor importance and progressive inclusion tools in \texttt{microimpute} let researchers assess which variables drive imputation performance in their specific context and make informed decisions about variable selection when data are limited.
 
-Together, these supplementary analyses contextualize the main results: QRF's strong performance in the SCF-CPS setting reflects the particular characteristics of the wealth imputation problem---a large donor dataset, highly non-linear relationships, and informative financial predictors available in both surveys. Researchers applying \texttt{microimpute} to other settings should expect method rankings to vary with dataset characteristics and should leverage the package's comparison tools accordingly.
+Together, these supplementary analyses put the main results in perspective. QRF's strong performance in the SCF-CPS setting reflects the particular characteristics of the wealth imputation problem, namely a large donor dataset, highly non-linear relationships, and informative financial predictors available in both surveys. Researchers applying \texttt{microimpute} to other settings should expect method rankings to vary with dataset characteristics.
 
 \subsection{Reform analysis and policy uncertainty}
 
-The SSI Savings Penalty Elimination Act simulation demonstrates that imputation is not merely a data preprocessing step but a consequential modeling choice for policy analysis. The variation in estimated reform impact across the five methods represents practical policy uncertainty: policymakers relying on different imputation approaches would reach different conclusions about the budgetary and coverage effects of raising SSI resource limits. This finding extends beyond SSI to any means-tested program where eligibility depends on wealth or asset data not collected in the primary survey.
+The SSI Savings Penalty Elimination Act simulation shows that imputation is not merely a data preprocessing step but a modeling choice with direct consequences for policy analysis. The variation in estimated reform impact across the five methods represents practical policy uncertainty: policymakers relying on different imputation approaches would reach different conclusions about the budgetary and coverage effects of raising SSI resource limits. This finding extends beyond SSI to any means-tested program where eligibility depends on wealth or asset data not collected in the primary survey.
 
-The difference-in-differences approach, by comparing current-law and reform outcomes within each method, partially mitigates the limitation that SCF net worth serves as a proxy for SSI countable resources. Since the systematic bias affects both scenarios similarly, the reform impact estimates are less sensitive to this proxy than the level estimates. Nonetheless, future work could improve these estimates by imputing asset-specific components (e.g., excluding primary residence equity) rather than total net worth, and by extending the analysis to demographic subgroups (aged vs.\ disabled recipients) to better characterize the reform's distributional effects.
+By comparing current-law and reform outcomes within each method, the microsimulation approach (under ceteris paribus) partially mitigates the limitation that SCF net worth is only a proxy for SSI countable resources. Since the systematic bias affects both scenarios similarly, the reform impact estimates are less sensitive to this proxy than the level estimates. Nonetheless, future work could improve these estimates by imputing asset-specific components (e.g., excluding primary residence equity) rather than total net worth, and by extending the analysis to demographic subgroups (aged vs.\ disabled recipients) to better characterize the reform's distributional effects.
 
-A practical path to resolving the variation in aggregate estimates across imputation methods is survey weight calibration. Rather than choosing a method based on proximity to administrative totals, which conflates imputation accuracy with policy function artifacts, researchers can select the method with the best conditional accuracy (here, QRF) for record-level imputation and then calibrate survey weights to match known administrative aggregates such as SSI recipient counts and total expenditure to ensure national-level representativeness \citep{woodruff2024enhancing}. This two-stage approach achieves the best of both worlds: accurate household-level wealth distributions that preserve the conditional relationships needed for reform simulation, combined with aggregate estimates that align with administrative benchmarks.
+A practical path to resolving the variation in aggregate estimates across imputation methods is survey weight calibration. Rather than choosing a method based on proximity to administrative totals, which conflates imputation accuracy with policy function artifacts, researchers can select the method with the best conditional accuracy (here, QRF) for record-level imputation and then calibrate survey weights to match known administrative aggregates such as SSI recipient counts and total expenditure to ensure national-level representativeness \citep{woodruff2024enhancing}. This two-stage approach combines accurate household-level wealth distributions that preserve the conditional relationships needed for reform simulation with aggregate estimates that align with administrative benchmarks.
 
 \subsection{Limitations and future improvements}\label{sec:limitations}
 
-Despite the demonstrated advantages, there remain limitations worth considering for future development:
-
 \subsubsection{Current package limitations}
 
-While $\texttt{microimpute}$ currently supports five imputation methods, expanding to include additional modern machine learning approaches such as more complex neural networks, gradient boosting machines, or deep learning architectures could further improve performance, particularly for complex multivariate relationships \citep{alaa2024deep}. The package would benefit from implementing ensemble methods that combine multiple imputation approaches, potentially leveraging the strengths of different methods across different parts of the distribution. Moreover, model selection and assessment could be enhanced with evaluation metrics additional to quantile loss, ensuring a thorough understanding of each model's behavior at every step.
+The current method library of five approaches could be expanded with gradient boosting machines, deeper neural architectures, or other modern machine learning methods, which may improve performance for complex multivariate relationships \citep{alaa2024deep}. Ensemble methods that combine multiple imputation approaches across different parts of the distribution are another natural extension. Model selection and assessment could also benefit from evaluation metrics beyond quantile loss, giving a more complete picture of each model's behavior.
 
 \subsubsection{QRF-specific challenges}
 
@@ -58,6 +44,6 @@ \subsubsection{QRF-specific challenges}
 
 \subsubsection{Missing data assumptions}
 
-All imputation methods in this study operate under the assumption that wealth is missing at random (MAR) conditional on the shared predictors $X$, as formalized in Section~\ref{sec:cia}. This assumption may be violated in practice: wealthy individuals may be less likely to participate in surveys or may underreport assets, introducing missing-not-at-random (MNAR) patterns that the conditioning variables cannot fully account for \citep{andridge2010review}. The predictor exclusion analysis in Appendix~\ref{app:robustness} provides a partial sensitivity check on the plausibility of the MAR assumption. The substantial degradation in imputation quality when financial predictors are removed, particularly the 116\% increase in quantile loss when interest and dividend income is excluded, demonstrates that the conditioning set is doing meaningful work in explaining wealth variation, which is consistent with MAR plausibility. However, this analysis cannot rule out residual dependence on unobserved factors. More formal MNAR sensitivity analyses, such as pattern-mixture models or selection models applied to the imputation framework, represent an important direction for future work. Researchers applying \texttt{microimpute} to settings where nonresponse or selection effects are suspected should consider supplementary diagnostics to assess the sensitivity of their results to departures from MAR.
+All imputation methods in this study operate under the assumption that wealth is missing at random (MAR) conditional on the shared predictors $X$, as formalized in Section~\ref{sec:cia}. This assumption may be violated in practice: wealthy individuals may be less likely to participate in surveys or may underreport assets, introducing missing-not-at-random (MNAR) patterns that the conditioning variables cannot fully account for \citep{andridge2010review}. The predictor exclusion analysis in Appendix~\ref{app:robustness} provides a partial sensitivity check on the plausibility of the MAR assumption. The substantial degradation in imputation quality when financial predictors are removed, particularly the 116\% increase in quantile loss when interest and dividend income is excluded, indicates that the conditioning set is doing meaningful work in explaining wealth variation, consistent with MAR plausibility. However, this analysis cannot rule out residual dependence on unobserved factors. More formal MNAR sensitivity analyses, such as pattern-mixture models or selection models applied to the imputation framework, are a natural next step. Researchers applying \texttt{microimpute} to settings where nonresponse or selection effects are suspected should consider supplementary diagnostics to assess the sensitivity of their results to departures from MAR.
 
-These enhancements would position $\texttt{microimpute}$ further as a comprehensive solution for survey data imputation challenges while maintaining its current strengths in ease of use and methodological rigor.
\ No newline at end of file
+These extensions would broaden the range of problems \texttt{microimpute} can address while preserving its current simplicity.
\ No newline at end of file
diff --git a/paper/sections/introduction.tex b/paper/sections/introduction.tex
index 8228cbc..864fe41 100644
--- a/paper/sections/introduction.tex
+++ b/paper/sections/introduction.tex
@@ -1,19 +1,19 @@
 \section{Introduction}
 
-Statistical matching, also known as data fusion or full variable imputation, addresses the challenge of combining information from multiple data sources that share common variables but contain different samples of units \citep{dorazio2006statistical}, which is commonly found across fields in empirical research. In its simplest form, the problem involves a donor dataset containing both shared and unique variables and a receiver dataset containing only the shared variables, where the goal is to impute the missing variables for observations in the receiver file based on their shared characteristics \citep{dorazio2021statistical}. This framework generalizes traditional missing data imputation by recognizing that ``missingness'' can arise not only from nonresponse within a single survey but also from the structural absence of variables across distinct data sources.
+Statistical matching, also known as data fusion or full variable imputation, combines information from multiple data sources that share common variables but contain different samples of units \citep{dorazio2006statistical}, a situation common across empirical research fields. In its simplest form, the problem involves a donor dataset containing both shared and unique variables and a receiver dataset containing only the shared variables, where the goal is to impute the missing variables for observations in the receiver file based on their shared characteristics \citep{dorazio2021statistical}. This framework generalizes traditional missing data imputation by recognizing that ``missingness'' can arise not only from nonresponse within a single survey but also from the structural absence of variables across distinct data sources.
 
-The relevance of statistical matching extends across the social sciences, particularly in microsimulation modeling where policy analysis requires combining detailed demographic information from one survey with economic variables from another \citep{bourguignon2006microsimulation}. For instance, household surveys may capture income and consumption patterns but lack wealth data, while financial surveys provide wealth information for different samples. Constructing synthetic files that combine these variables enables richer policy analysis than either source alone permits.
+Statistical matching is widely applicable in the social sciences, particularly in microsimulation modeling, where policy analysis often requires combining detailed demographic information from one survey with economic variables from another \citep{bourguignon2006microsimulation}. For instance, household surveys may capture income and consumption patterns but lack wealth data, while financial surveys provide wealth information for different samples. Constructing synthetic files that combine these variables allows more detailed policy analysis than either source alone.
 
-The central methodological challenge is accurately modeling the conditional distribution of the target variable given shared predictors, as different approaches make different assumptions about this relationship. Parametric methods may be too restrictive for economic variables exhibiting heavy tails and heteroscedasticity, while nonparametric and machine learning approaches offer greater flexibility at the cost of additional complexity \citep{meinshausen2006quantile, andridge2010review}. Section~2 reviews these trade-offs in detail.
+The main methodological difficulty is accurately modeling the conditional distribution of the target variable given shared predictors, as different approaches make different assumptions about this relationship. Parametric methods may be too restrictive for economic variables with heavy tails and heteroscedasticity, while nonparametric and machine learning approaches are more flexible but also more complex \citep{meinshausen2006quantile, andridge2010review}. Section~2 reviews these trade-offs in detail.
 
-This paper presents the \texttt{microimpute} package, a Python library implementing five statistical matching methods: Hot Deck Matching, Ordinary Least Squares, Quantile Regression, Mixture Density Networks, and Quantile Random Forests. The package provides a comprehensive framework for comparing these methods through systematic benchmarking, with automated model selection based on quantile loss performance. We demonstrate the methodology through an application to wealth imputation, transferring net worth data from the Survey of Consumer Finances to the Current Population Survey. This task exemplifies the challenges of statistical matching for heavy-tailed distributions.
+This paper presents the \texttt{microimpute} package, a Python library implementing five statistical matching methods: Hot Deck Matching, Ordinary Least Squares, Quantile Regression, Mixture Density Networks, and Quantile Random Forests. The package provides a framework for comparing these methods through systematic benchmarking, with automated model selection based on quantile loss performance. We demonstrate the methodology through an application to wealth imputation, transferring net worth data from the Survey of Consumer Finances to the Current Population Survey. This task exemplifies the challenges of statistical matching for heavy-tailed distributions.
 
 This work provides the following contributions:
 \begin{itemize}
-    \item An open-source implementation of a unified framework for five imputation methods, facilitating systematic comparison across methods and support statistical matching through accessible, user-friendly software
-    \item An automated benchmarking pipeline that evaluates distributional accuracy using quantile loss, enabling researchers to select the most appropriate method for their specific data characteristics
-    \item Empirical evidence on the relative performance of these methods for wealth imputation, suggesting that Quantile Random Forests are well-suited to heavy-tailed distributions while highlighting the context-dependence of method choice
-    \item A policy reform simulation of the SSI Savings Penalty Elimination Act demonstrating that imputation method choice produces meaningfully different estimates of reform impact, and that without imputed wealth data, such reforms are entirely un-simulable
+    \item An open-source implementation of five imputation methods under a common interface, allowing systematic comparison across methods
+    \item An automated benchmarking pipeline that evaluates distributional accuracy using quantile loss, so that researchers can select the most appropriate method for their data
+    \item Empirical evidence on the relative performance of these methods for wealth imputation, suggesting that Quantile Random Forests are well-suited to heavy-tailed distributions, though method choice remains context-dependent
+    \item A policy reform simulation of the SSI Savings Penalty Elimination Act, showing that imputation method choice produces meaningfully different estimates of reform impact, and that without imputed wealth data, such reforms cannot be simulated at all
 \end{itemize}
 
 The remainder of this paper is organized as follows. Section 2 reviews the statistical matching problem, including its formal definition, the conditional independence assumption that underlies all matching methods, and the five imputation approaches implemented in \texttt{microimpute}. Section 3 describes our data sources. Section 4 presents the package architecture and benchmarking methodology. Section 5 reports empirical results from the wealth imputation application. Section 6 discusses implications and limitations, and Section 7 concludes.
\ No newline at end of file
diff --git a/paper/sections/methodology.tex b/paper/sections/methodology.tex
index d41a552..6623906 100644
--- a/paper/sections/methodology.tex
+++ b/paper/sections/methodology.tex
@@ -2,22 +2,22 @@ \section{Methodology}\label{sec:methodology}
 
 \subsection{$\texttt{microimpute}$ package implementation}
 
-$\texttt{microimpute}$\footnote{Complete documentation, implementation details, and usage examples are available at https://policyengine.github.io/microimpute/.} is an open-source Python framework for statistical matching that provides a unified interface for imputing variables across survey datasets using multiple methods. The package addresses a gap in the microsimulation and microdata research community: while several imputation approaches exist, researchers have lacked a standardized tool for systematically comparing their distributional performance on a given dataset and automatically selecting the best approach. In this study, \texttt{microimpute} serves both as a methodological contribution to the broader research community and as the analytical engine powering the wealth imputation experiments described below.
+$\texttt{microimpute}$\footnote{Documentation, implementation details, and usage examples are available at https://policyengine.github.io/microimpute/.} is an open-source Python framework for statistical matching that provides a common interface for imputing variables across survey datasets using multiple methods. While several imputation approaches exist individually, researchers have lacked a standardized tool for comparing their distributional performance on a given dataset and automatically selecting the best approach. In this study, \texttt{microimpute} is both a methodological contribution and the analytical engine behind the wealth imputation experiments described below.
 
 \subsubsection{Core capabilities}
 
-The package currently supports five imputation methods: Hot Deck Matching, Ordinary Least Squares Linear Regression (OLS), Quantile Regression, Quantile Random Forests (QRF), and Mixture Density Networks (MDN). This approach allows researchers to systematically evaluate which technique provides the most accurate results for their specific dataset and research objectives. Additionally, the package is designed to be modular, allowing for easy extension with additional imputation methods in the future.
+The package currently supports five imputation methods: Hot Deck Matching, Ordinary Least Squares Linear Regression (OLS), Quantile Regression, Quantile Random Forests (QRF), and Mixture Density Networks (MDN). The package is modular, so additional imputation methods can be added without modifying the existing codebase.
 
 \subsubsection{Key features}
 
-The package provides several capabilities designed for the challenges of cross-survey imputation, each of which plays a specific role in the wealth imputation experiments presented in this paper:
+Several features are relevant to the cross-survey imputation experiments presented in this paper:
 
 \begin{enumerate}
-    \item \textbf{Survey data weights integration}: Handles survey data weights through sampling to ensure that models are trained on a donor data set representative of the true distribution.
-    \item \textbf{Method comparison and benchmarking}: Allows researchers to easily compare different approaches and automatically determine the method providing the most accurate results.
-    \item \textbf{Flexible methodological set-up}: Enables advanced usage through specified hyperparameter setting and tuning.
-    \item \textbf{Quantile-based evaluation}: Uses quantile loss calculations to assess imputation quality across different parts of the distribution.
-    \item \textbf{Autoimputation}: Provides an integrated imputation pipeline that tunes method hyperparameters to the specific datasets, compares methods, and selects the best-performing to conduct the requested imputation in a single function call.
+    \item \textbf{Survey weight integration}: Incorporates survey weights through sampling so that models are trained on data representative of the true population distribution.
+    \item \textbf{Method comparison and benchmarking}: Compares different approaches and automatically identifies the method with the lowest quantile loss.
+    \item \textbf{Hyperparameter tuning}: Supports user-specified hyperparameters as well as automated tuning.
+    \item \textbf{Quantile-based evaluation}: Assesses imputation quality across different parts of the distribution using quantile loss.
+    \item \textbf{Autoimputation}: Tunes hyperparameters, compares methods, and selects the best-performing one to impute onto the receiver dataset, all in a single function call.
 \end{enumerate}
 
 \subsection{Evaluation framework}
@@ -29,7 +29,7 @@ \subsection{Evaluation framework}
     \mathcal{L}_\tau = \rho_\tau(y - \hat{Q}_\tau(y|x)) = (y - \hat{Q}_\tau(y|x)) \cdot (\tau - \mathbf{1}_{y < \hat{Q}_\tau(y|x)})
 \end{equation}
 
-Under-prediction of an upper-tail quantile is penalized at rate $\tau$, whereas over-prediction is penalized at rate $1 - \tau$. This asymmetric structure captures directional bias---a crucial feature when imputing highly skewed variables like wealth, where large under-estimates in the right tail must be discouraged more strongly than equal-sized over-estimates \citep{koenker1978regression}. Compared with mean-squared or mean-absolute error, which penalize errors symmetrically, quantile loss directly targets the conditional distribution and remains robust to outliers \citep{ghenis2018quantile}. Averaging over multiple quantiles $\tau \in \{0.1, 0.2, \ldots, 0.9\}$ provides a comprehensive measure of how well the method estimates the conditional distribution across its entire range. Lower average quantile loss indicates better distributional calibration. Because quantile loss is not scale-invariant, the absolute values reported in Section~5 reflect the scale transformation applied; we supplement raw values with relative improvement percentages to aid interpretability.
+Under-prediction of an upper-tail quantile is penalized at rate $\tau$, whereas over-prediction is penalized at rate $1 - \tau$. This asymmetric structure captures directional bias, which matters when imputing highly skewed variables like wealth, where large under-estimates in the right tail must be discouraged more strongly than equal-sized over-estimates \citep{koenker1978regression}. Compared with mean-squared or mean-absolute error, which penalize errors symmetrically, quantile loss directly targets the conditional distribution and remains robust to outliers \citep{ghenis2018quantile}. Averaging over multiple quantiles $\tau \in \{0.1, 0.2, \ldots, 0.9\}$ summarizes how well the method estimates the conditional distribution across its entire range. Lower average quantile loss indicates better distributional calibration. Because quantile loss is not scale-invariant, the absolute values reported in Section~5 reflect the scale transformation applied; we supplement raw values with relative improvement percentages to aid interpretability.
 
 As a complementary metric, we use the Wasserstein distance (also known as the earth mover's distance) to measure overall distributional similarity between imputed and reference distributions. Formally, for one-dimensional distributions $P$ and $Q$:
 \begin{equation}
@@ -42,11 +42,11 @@ \subsection{Evaluation framework}
 
 \subsection{Experimental setup}\label{sec:experimental_setup}
 
-The \texttt{autoimpute} function orchestrates the full imputation pipeline in a single call: it tunes hyperparameters for applicable methods using the Optuna framework \citep{akiba2019optuna}, evaluates all methods through $k$-fold cross-validation on the donor data, and automatically selects the best-performing one to impute onto the receiver dataset. Using this pipeline, we evaluate all five imputation methods---QRF, OLS, Quantile Regression, Hot Deck Matching, and MDN---for net worth imputation from the SCF to the CPS.
+The \texttt{autoimpute} function orchestrates the full imputation pipeline in a single call: it tunes hyperparameters for applicable methods using the Optuna framework \citep{akiba2019optuna}, evaluates all methods through $k$-fold cross-validation on the donor data, and automatically selects the best-performing one to impute onto the receiver dataset. Using this pipeline, we evaluate all five imputation methods (QRF, OLS, Quantile Regression, Hot Deck Matching, and MDN) for net worth imputation from the SCF to the CPS.
 
 Prior to model training, we apply an inverse hyperbolic sine (asinh) transformation to the net worth target variable. The asinh function, $\operatorname{asinh}(x) = \ln(x + \sqrt{x^2 + 1})$, approximates the natural logarithm for large positive values while remaining well-defined for zero and negative values. This is particularly important for wealth data, where a meaningful fraction of households report negative net worth (i.e., debts exceeding assets), making a standard logarithmic transformation infeasible. The transformation compresses the extreme right tail of the wealth distribution, stabilizing variance and improving model performance across all methods. After imputation, predictions are back-transformed via the inverse function $\sinh(\cdot)$ to recover values on the original dollar scale.
 
-This experimental design implicitly assumes that wealth is missing at random (MAR) conditional on the shared predictors $X$---that is, the probability of observing wealth in the donor survey does not depend on wealth itself once we condition on the demographic and financial variables available in both datasets, as formalized in Section~\ref{sec:cia}. This assumption is standard in statistical matching but may be violated in practice: wealthy households may be less likely to participate in surveys or may underreport assets, creating MNAR-like patterns that the shared predictors cannot fully capture. We return to this limitation and discuss partial sensitivity checks in Section~\ref{sec:limitations}.
+This experimental design implicitly assumes that wealth is missing at random (MAR) conditional on the shared predictors $X$, i.e., that the probability of observing wealth in the donor survey does not depend on wealth itself once we condition on the demographic and financial variables available in both datasets, as formalized in Section~\ref{sec:cia}. This assumption is standard in statistical matching but may be violated in practice: wealthy households may be less likely to participate in surveys or may underreport assets, creating MNAR-like patterns that the shared predictors cannot fully capture. We return to this limitation and discuss partial sensitivity checks in Section~\ref{sec:limitations}.
 
 To create a ground truth for evaluation, we employ cross-validation, splitting the SCF into 5 folds. For each method, \texttt{autoimpute} performs the following steps:
 
@@ -82,7 +82,7 @@ \subsection{Downstream policy analysis}\label{sec:policy_validation}
 
 While distributional metrics such as quantile loss and Wasserstein distance assess how well an imputation method recovers the statistical properties of the target variable, they do not directly capture whether differences between methods matter for downstream policy analysis. To address this, we introduce a policy analysis layer that evaluates imputation quality through its effect on a concrete policy simulation, under both current law and a proposed reform.
 
-We use the Supplemental Security Income (SSI) program as the policy case. SSI is a federal means-tested program that provides cash assistance to aged, blind, and disabled individuals with limited income and resources \citep{ssa2023ssi}. Critically, SSI eligibility includes a resource test with limits of \$2,000 for individuals and \$3,000 for couples, which directly depends on household wealth data that the CPS does not collect.
+We use the Supplemental Security Income (SSI) program as the policy case. SSI is a federal means-tested program that provides cash assistance to aged, blind, and disabled individuals with limited income and resources \citep{ssa2023ssi}. SSI eligibility includes a resource test with limits of \$2,000 for individuals and \$3,000 for couples, which depends on household wealth data that the CPS does not collect.
 
 In current practice, PolicyEngine's microsimulation model \citep{policyengine_us} handles the missing wealth data by defaulting \texttt{ssi\_countable\_resources} to zero for all CPS households, meaning every categorically eligible individual passes the resource test. This produces a substantial overestimate of SSI participation. For each imputation method, we feed the model's imputed net worth values into PolicyEngine as \texttt{ssi\_countable\_resources}, replacing the default zero-resource assumption with household-level wealth data. We then simulate SSI eligibility and benefits under each imputation scenario and compare the resulting recipient counts and total expenditure against SSA administrative totals (7.4 million recipients and \$59.6 billion in federal payments for 2024).
 
@@ -90,4 +90,4 @@ \subsection{Downstream policy analysis}\label{sec:policy_validation}
 
 \subsection{Additional methodological analyses}
 
-Two supplementary analyses complement the main experimental framework described above. Appendix~\ref{app:robustness} presents a predictor importance analysis using leave-one-out and progressive inclusion experiments on the QRF model, identifying the relative contribution of each predictor variable to wealth imputation quality and informing the plausibility of the conditional independence assumption. Meanwhile, to understand how the superiority of certain models over others generalize to other contexts and datasets, appendix~\ref{app:benchmarking} reports a cross-dataset benchmarking exercise that evaluates \texttt{microimpute}'s five methods across six publicly available datasets from diverse domains, assessing the generalizability of method performance beyond the SCF-CPS use case.
+Two supplementary analyses complement the main experimental framework. Appendix~\ref{app:robustness} presents a predictor importance analysis using leave-one-out and progressive inclusion experiments on the QRF model, measuring the relative contribution of each predictor to wealth imputation quality and informing the plausibility of the conditional independence assumption. To test whether the relative performance of methods generalizes beyond this particular application, Appendix~\ref{app:benchmarking} reports a cross-dataset benchmarking exercise that evaluates \texttt{microimpute}'s five methods across six publicly available datasets from different domains.
diff --git a/paper/sections/results.tex b/paper/sections/results.tex
index 4396ce5..6c1f130 100644
--- a/paper/sections/results.tex
+++ b/paper/sections/results.tex
@@ -2,9 +2,9 @@ \section{Results}
 
 \subsection{Imputation results}
 
-The cross-validation results demonstrate QRF's strong performance across the wealth distribution. With the asinh transformation applied to the net worth target variable, QRF achieved the lowest average quantile loss across all quantiles, outperforming OLS, Hot Deck Matching, Quantile Regression, and MDN. QRF's overall consistency across the entire distribution made it the best-performing method for wealth imputation in this setting.
+The cross-validation results show that QRF performs well across the wealth distribution. With the asinh transformation applied to the net worth target variable, QRF achieved the lowest average quantile loss across all quantiles, outperforming OLS, Hot Deck Matching, Quantile Regression, and MDN. QRF's overall consistency across the entire distribution made it the best-performing method for wealth imputation in this setting.
 
-$\texttt{microimpute}$'s automated hyperparameter tuning via the Optuna framework contributed to these results. The optimal QRF configuration identified through cross-validation included 202 trees, a minimum of 1 sample per leaf, a split threshold of 5 samples, approximately 95\% of features considered at each split, and bootstrap sampling enabled. These parameters balanced model complexity with generalization capability, preventing overfitting while capturing the non-linear relationships between demographic predictors and wealth outcomes.
+Automated hyperparameter tuning via the Optuna framework contributed to these results. The optimal QRF configuration identified through cross-validation included 202 trees, a minimum of 1 sample per leaf, a split threshold of 5 samples, approximately 95\% of features considered at each split, and bootstrap sampling enabled. These parameters balance model complexity with generalization, preventing overfitting while still capturing non-linear relationships between demographic predictors and wealth.
 
 \subsubsection{Quantile loss}
 
@@ -42,7 +42,7 @@ \subsubsection{Quantile loss}
 
 \subsubsection{Imputed wealth distribution comparisons}
 
-By visually comparing the wealth distributions resulting from imputing with each method, and comparing them to the weighted donor distribution, we gain a more comprehensive understanding of imputation performance, moving past the test quantile loss average measured on the SCF. The distribution of wealth values imputed by QRF closely resembles the original SCF distribution, suggesting that QRF is not only the best method in terms of quantile loss but also strong when achieving distributional estimates close to the true wealth distribution in the United States. Because Matching directly takes values from the donor distribution and imputes them onto the receiver, it is unsurprising that its distribution closely resembles the donor distribution. However, Matching may struggle to uncover the non-linear relationship between the different predictors and wealth values, resulting in a seemingly correct marginal distribution but inaccurate imputation given certain household characteristics. Meanwhile, OLS, QuantReg, and MDN fail to capture the variability across the distribution and impute most values at the lower or higher ends of the distribution.
+Comparing the imputed wealth distributions against the weighted donor distribution gives a fuller picture of imputation performance beyond the quantile loss averages measured on the SCF. The distribution of wealth values imputed by QRF closely matches the original SCF distribution, suggesting that QRF performs well not only in quantile loss but also in reproducing the overall shape of the U.S. wealth distribution. Because Matching directly takes values from the donor distribution and imputes them onto the receiver, it is unsurprising that its distribution closely resembles the donor distribution. However, Matching may fail to capture the non-linear relationship between predictors and wealth, producing a marginal distribution that looks correct while assigning inaccurate values to individual households. OLS, QuantReg, and MDN fail to capture variability across the distribution and concentrate imputed values at the lower or higher ends.
 
 \begin{figure}[H]
     \centering
@@ -76,7 +76,7 @@ \subsubsection{Wasserstein distance}
 
 \subsubsection{Distribution of wealth by disposable income decile}
 
-Beyond the statistical metrics reported above, we perform a plausibility check by examining whether imputed wealth exhibits economically expected relationships with observable variables. Household wealth should increase monotonically with disposable income, reflecting the well-documented positive correlation between income and wealth accumulation. This check is particularly important for wealth imputation: a method that achieves competitive quantile loss but produces wealth-income patterns that violate basic economic relationships would be unreliable for downstream policy analysis, where the joint distribution of income and wealth drives eligibility and benefit calculations.
+Beyond the statistical metrics reported above, we check whether imputed wealth exhibits economically expected relationships with observable variables. Household wealth should increase monotonically with disposable income, given the well-documented positive correlation between income and wealth accumulation. This check matters because a method that achieves competitive quantile loss but produces wealth-income patterns that violate basic economic relationships would be unreliable for downstream policy analysis, where eligibility and benefit calculations depend on the joint distribution of income and wealth.
 
 \begin{figure}[H]
     \centering
@@ -85,11 +85,11 @@ \subsubsection{Distribution of wealth by disposable income decile}
     \label{fig:imputation_comparison_by_income_decile}
 \end{figure}
 
-These results support the observations above, with QRF presenting the most consistent and plausible relationship to disposable income, with a monotonically increasing average wealth as deciles increase---a pattern consistent with economic expectations and serving as a plausibility check on the imputation. This plot also demonstrates the caveats of the other models, for example showing the lower variability in OLS's predictions, and Matching's consistent underprediction across deciles.
+These results are consistent with the observations above. QRF produces a monotonically increasing average wealth across income deciles, as economic theory would predict. The plot also exposes weaknesses in the other models: OLS shows little variability in its predictions, and Matching consistently underpredicts across deciles.
 
 \subsection{Policy analysis: SSI simulation}
 
-Following the downstream policy analysis framework described in Section~\ref{sec:policy_validation}, we feed each model's imputed net worth into PolicyEngine as \texttt{ssi\_countable\_resources} and simulate SSI eligibility and benefits. Table~\ref{tab:ssi_results} compares simulated recipient counts and total expenditure against SSA administrative totals for 2024.
+As described in Section~\ref{sec:policy_validation}, we feed each model's imputed net worth into PolicyEngine as \texttt{ssi\_countable\_resources} and simulate SSI eligibility and benefits. Table~\ref{tab:ssi_results} compares simulated recipient counts and total expenditure against SSA administrative totals for 2024.
 
 \begin{table}[H]
     \centering
@@ -115,7 +115,7 @@ \subsection{Policy analysis: SSI simulation}
 
 \subsubsection{Policy reform: SSI Savings Penalty Elimination Act}
 
-To demonstrate why wealth imputation matters for policy analysis and not just model calibration, we simulate the SSI Savings Penalty Elimination Act \citep{ssi_spea_2025}, which would raise resource limits from \$2,000/\$3,000 to \$10,000/\$20,000. Table~\ref{tab:ssi_reform} reports SSI outcomes under both current law and the reformed thresholds for each imputation method.
+To show why wealth imputation matters for policy analysis, not just model calibration, we simulate the SSI Savings Penalty Elimination Act \citep{ssi_spea_2025}, which would raise resource limits from \$2,000/\$3,000 to \$10,000/\$20,000. Table~\ref{tab:ssi_reform} reports SSI outcomes under both current law and the reformed thresholds for each imputation method.
 
 \begin{table}[H]
     \centering
@@ -140,9 +140,9 @@ \subsubsection{Policy reform: SSI Savings Penalty Elimination Act}
 
 The reform raises the resource threshold from \$2,000 to \$10,000 (individual), so the relevant question is how dense each method's imputed wealth distribution is in the \$2,000--\$10,000 range. Households with imputed net worth in this interval fail the current-law resource test but would pass under the reform, gaining SSI eligibility. The estimated number of additional recipients ranges from 0.1 million (QuantReg) to 4.2 million (OLS), with corresponding additional expenditure from \$1.0 billion to \$24.1 billion. This range reflects how differently each imputation approach populates this critical region of the wealth distribution.
 
-QRF estimates a modest reform impact (0.6 million additional recipients, \$3.4 billion additional expenditure), consistent with its faithful reproduction of the SCF wealth distribution. Most households with near-zero wealth are already below the \$2,000 threshold, so raising it to \$10,000 captures relatively few additional households. OLS, by contrast, produces the largest reform impact (4.2 million additional recipients), because its compressed, normally distributed imputations place a substantial mass of households in the \$2,000--\$10,000 range. QuantReg's near-zero reform impact (0.1 million) reflects its tendency to push imputed values away from this intermediate range. These differences underscore that the choice of imputation method is not merely a technical detail but a modeling decision that directly shapes policy impact estimates.
+QRF estimates a modest reform impact (0.6 million additional recipients, \$3.4 billion additional expenditure), consistent with its faithful reproduction of the SCF wealth distribution. Most households with near-zero wealth are already below the \$2,000 threshold, so raising it to \$10,000 captures relatively few additional households. OLS, by contrast, produces the largest reform impact (4.2 million additional recipients), because its compressed, normally distributed imputations place a substantial mass of households in the \$2,000--\$10,000 range. QuantReg's near-zero reform impact (0.1 million) reflects its tendency to push imputed values away from this intermediate range. These differences show that the choice of imputation method is not merely a technical detail but a modeling decision that directly shapes policy impact estimates.
 
-The baseline (no wealth data) scenario starkly illustrates why imputation is essential: because all resources default to zero, every categorically eligible individual already passes the current-law resource test, so raising the thresholds has zero effect. Without wealth data, this reform is literally un-simulable. The microsimulation would not be able to distinguish between the status quo and a fivefold increase in resource limits.
+The baseline (no wealth data) scenario shows plainly why imputation is necessary. Because all resources default to zero, every categorically eligible individual already passes the current-law resource test, so raising the thresholds has zero effect. Without wealth data, the microsimulation cannot distinguish between the status quo and a fivefold increase in resource limits.
 
 \begin{figure}[H]
     \centering

From a1f5b19c9c51341ad785a652c55b18232ed7a4ed Mon Sep 17 00:00:00 2001
From: juaristi22 <juaristi@uni.minerva.edu>
Date: Fri, 27 Feb 2026 12:46:51 +0530
Subject: [PATCH 2/5] updating readme with paper

---
 README.md                  |  50 +++++++++++++++++--------------------
 paper/main.pdf             | Bin 543392 -> 543552 bytes
 paper/sections/results.tex |   2 +-
 3 files changed, 24 insertions(+), 28 deletions(-)

diff --git a/README.md b/README.md
index 8b055cf..a5a817f 100644
--- a/README.md
+++ b/README.md
@@ -1,31 +1,31 @@
 # Microimpute
 
-Microimpute enables variable imputation through a variety of statistical methods. By providing a consistent interface across different imputation techniques, it allows researchers and data scientists to easily compare and benchmark different approaches using quantile loss and log loss calculations to determine the method providing most accurate results.
+Microimpute is a Python package for imputing variables from one survey dataset onto another. It wraps five imputation methods behind a common interface so you can benchmark them on your data and pick the one that works best, rather than defaulting to a single approach.
 
-## Features
+## Methods
 
-### Multiple imputation methods
-- **Statistical Matching**: Distance-based matching for finding similar observations
-- **Ordinary Least Squares (OLS)**: Linear regression-based imputation
-- **Quantile Regression**: Distribution-aware regression imputation
-- **Quantile Random Forests (QRF)**: Non-parametric forest-based approach
-- **Mixture Density Networks (MDN)**: Neural network with Gaussian mixture approximation head
+- **Statistical Matching**: distance-based matching to find similar donor observations
+- **Ordinary Least Squares (OLS)**: linear regression imputation
+- **Quantile Regression**: models conditional quantiles instead of the conditional mean
+- **Quantile Random Forests (QRF)**: non-parametric, tree-based quantile estimation
+- **Mixture Density Networks (MDN)**: neural network with a Gaussian mixture output
 
-### Automated method selection
-- **AutoImpute**: Automatically compares and selects the best imputation method for your data
-- **Cross-validation**: Built-in evaluation using quantile loss (numerical) and log loss (categorical)
-- **Variable type support**: Handles numerical, categorical, and boolean variables
+## Autoimpute
 
-### Developer-friendly design
-- **Consistent API**: Standardized `fit()` and `predict()` interface across all models
-- **Extensible architecture**: Easy to implement custom imputation methods
-- **Weighted data handling**: Preserve data distributions with sample weights
-- **Input validation**: Automatic parameter and data validation
+The `autoimpute` function tunes hyperparameters, runs cross-validation across all five methods, and selects the best performer based on quantile loss (for numerical targets) or log loss (for categorical targets). It handles numerical, categorical, and boolean variables.
 
-### Interactive dashboard
-- **Visual exploration**: Analyze imputation results through interactive charts at https://microimpute-dashboard.vercel.app/
-- **GitHub integration**: Load artifacts directly from CI/CD workflows
-- **Multiple data sources**: File upload, URL loading and sample data
+## API
+
+All models follow a `fit()` / `predict()` interface. The package supports sample weights to account for survey design, and validates inputs automatically. Adding a custom imputation method is straightforward since new models just need to implement the same interface.
+
+## Documentation and paper
+
+- [Documentation](https://policyengine.github.io/microimpute/) with examples and interactive notebooks
+- [Paper](https://github.com/PolicyEngine/microimpute/blob/main/paper/main.pdf) presenting microimpute and demonstrating it for SCF-to-CPS net worth imputation
+
+## Dashboard
+
+An interactive dashboard for exploring imputation results is available at https://microimpute-dashboard.vercel.app/. It supports file upload, URL loading, direct GitHub artifact integration, and sample data.
 
 ## Installation
 
@@ -33,16 +33,12 @@ Microimpute enables variable imputation through a variety of statistical methods
 pip install microimpute
 ```
 
-For image export functionality (PNG/JPG), install with:
+For image export (PNG/JPG):
 
 ```bash
 pip install microimpute[images]
 ```
 
-## Examples and documentation
-
-For detailed examples and interactive notebooks, see the [documentation](https://policyengine.github.io/microimpute/).
-
 ## Contributing
 
-Contributions are welcome to the project. Please feel free to submit a Pull Request with your improvements.
+Pull requests are welcome. If you find a bug or have a feature idea, open an issue or submit a PR.
diff --git a/paper/main.pdf b/paper/main.pdf
index daeda5ed18b7a8b59e63d6e6d095536603be5529..6eb6047d2d8c920d321ab4a2d3eaed270b9a82f7 100644
GIT binary patch
delta 4558
zcmV;<5i#zdl_9{EA%L_2)pZ0nF*hKS(NidYbyQ1l+b|5i=U3=98JJzmZ|}HvDK_k|
ztpR%Ic2I1^7T`xFIqkY%KT=ehH0Tf@A}R9YCq5jrj>YVD6<?eJ{$nOsnXsg;qOz`8
zGp>HGqO7Rg)r;WFB93FH)&2zQ)#ou`H}mQ{{El!T;Du*^A2+M3by_naqB_qLw%NIV
zjCmc^S<W^sd*Ex`9gGUXB&&GxAqdmD;9uu1T#TITZZ@N8@v`QVTFC1uNJ>7Nr6;x3
zb}@Q;SoAzuGC>%{$YCJM&sb!S80RzTaer7OhLAq0rJq3rIQt&XBIC_G27>RU0Rq1^
zU*LVA$f7JOi1z~SX;ty1S16Q3OX;qE$8$l}XuZ2$lV3Ye$jXUljH4hdM9nuxl;=?X
z(U$xyPy(zcOIJQq#rHrEQkh0gE|={5#k?cWd25w<4DyUis>&PucTgKWbqE!6J<+LT
zgNQw8TRmYds)RrF`uqZ3GRi^!RhD$BHwca*3ay5Ur^<MSWVCYJ56(ewdy3wF(sLH~
zR#!wdUBJ02q9CW1gzIm2Z@tRhr%8dGKNS!3RvXnUD3?%I=}Sm=&}nG1wyqmln)4fJ
zd*y+kNXk;2QG25-KB?j!3Z_%bi9D&`!$H{yslDvl)jBWFHvlN7X_Uc+ILdh>qBIDF
zNOFE3)Bs_jdv;OSl(fv<xX7u0Q3!guO@Ngoktj;S+62%nxgC@zfIDeAznnjASL`87
zMZuF-{G8Ff&J_hV3*QTFG;ko?JJ@qt;O}X{jwmu*^s%BoV*F{WP#gWCVdU`6Qm|V{
zS99>7XgMtXDU|_tzcc@Bd?r#NjRJw3?;~2>B%$E<(TT@>#1u2mJPyH7-napJ)U8{S
zOh5_9--2QAS!^Q*>2d5~9Hn_C1$PlouylHo!&{`!KLp@GgWEI*Bb#R|`Rh4iNm4~3
zD_K~<_cQe^7xgQrPn*?0j`EN8hc9>mhc9>nhc9>ow=Z}E@;(AKGLz9$D3?pO1UL$a
z)9mhQvRFKIs+Wzo1e|{v4T5pVW&)XFqvK1e0gW0Q2OFy-$C05kK43zxQpvHHQHr7g
z<3(aLx(=&jMTb%$oWvU9pbiFwMY)<9Fq8z$TWyk|LTCFRG9Fi0Btl;exFOuBO;Ye)
z18k(2rP?ZF%u+iI#QHHU3k+$-sw8+~6G$>plUt63E+Ul2e@=fnpl8fYS*(E%7?x1X
zLPZ0f;Ddw#6=uOIVNk-;YIhV}3Rwqc&_ryn5*knpHgbuhp`oIo)*&#4Iiq_tY)(pe
zmDrvHmu$JA0&s`1A^@AJKxcHCYQca2at<~+CzlD<QqBOUA()HHW*fy=IRWE}znqKp
z1DOINM9Af0XficOV~BwYo>UWtawT8}3cQw>p%1<)46LdFUep@Zl>}oeEK;)CSK>+m
zE+SxDDrOJ?+EP;*mt467J_s-w{J|UaU|E-<xdbbJe@|Bk?&lMT;#MIj2w43veW<7}
zfNL#27vmrGv4Aro`ON8Ih350|hxt`I{6JXn9?hfVjn{6}$=RE;9k0$y^a;oX_trJP
z{QEx+9+hGG0(zZXfBN+PK{+zt*eR1r@>F@ZT#CMcu6D|C_Qt-u<PbP~0^hF5olYl<
z&OE0OhIp8*`{Go=&!C>=jj=ztmtDF9H-FKnd-Tutd^|jvesSG`Cs0_fge814oZ(wz
z<fUAPl{rSWC}_G^5VY(2V=)!Gn;T4l6urg5RLCog0e12gK_FFNJ>L(i;nh_;IUj%V
zCr`olx*jmKw`i7P@QhebB%ajxo_v-3xE{|pL7l>Wpk9-E0Mr{Ynv1t^u~SX}R)5|p
z#=^cJ$UT77HF&N8w-0#cEqEq944$uM)9b4)quM}gTB>H5)g82UimX0r&YE|#;beYg
zU5>uG=EYg_vi)s5YTv#(am{PDm|eHMU`xW1Hz<m?bTcX-hIEX;pL{Q}F2<MFv)0G{
z@Es|7%VBfymIg;Ez96G4?2%>OtAE2{`0>+*3LSpZp3uG{XSo-WMz3(N-K1!6`sH6A
z|7UvJyVsvBMbD2z4N^eYi(Qb*x&^t+2SM)66tnDSiZwZe5qPkr<h$gMmU=}<*e8r2
zZKRthbP8`k@_+LN>5P=^XmhO7mauKXvY!PD?{M1OYiPly4B%$lba~sbfq&U8osVz<
z;`#{YX%q8#&m85hoME0UQ>yiGL9BGFU;<>?2PO>Lj$+$H>cvs&cIr^-Vaxn&UCSzU
zC{JA9^>h?FP>+q6R9qYvtS+-{`h1k~a58$)o+AzA`RJm*#F&w%+rD9`3bx6%YdBI9
zgGb#neW-h+4;wO^kY^=5oPSi-dE_p`tYxh|Fib65(5tB@hN(+EtuWCe(@=<5>IkMO
zSa~uF_y7+ws>$L0@H)T#ILF81>BNhd{sKJ++83BGBC$)HtV-guSVbUNLi-5q@fHm8
zo-oXNg!-)xjrJ;X{Brm{X(RVs=Rme8_rZZ0w?ds2nU3Q?E8Lh)R)5o4UbRmyWT{wX
zm<4<Csz!s@wnU3S(boqM_&rMnMt)W!xYdW*$4Y;qJ{w+<r7K-fVsCXpDbPbxr=L9P
z6|&(KR?^$YE7s-iC+AHW#%4H#Ju|*<KF~V_YfM|nk8(8fd(G;Qulpia++km>dX1G#
zKw7E^q*D&yH@&gtu78zWle=dnA6TI~i*oEbkZi33Q9q$jR}is!cXy?*y2d_<DI*k^
zpCKA4_B>I<+aZQkSV4*V)l`iNK95JU>G<>2^`h;MrJ_bpwiCb6E+_01^e>_vv3kng
zmYpoBVmo7JDFsZXKUy~q6`4-D)s2zEAmQmrH*-4T{hafzTz_}QL%PkOTOA@x=juuF
zPB|vqt#i*wik&lYFG&jX7DaH}iz1YDlM`sg1swXMY(`p4cIKK8Y{d=<0UkS419fxN
z18S>J)B{OKNP;CrT5fv4?d>$s(3^JYak2?eru_iACiiIc_@2d9m2I5J(Kz9>!bx6T
zSFwcacZeMvLVs8Bj@Zd?wUcR(O|B2K%<t3b`Rw}Q0t_SV<iEqoXgs;>F=|xHCF`wT
z6sxyL{iF4gm_5VIB68fU_Co<MA=0*p95+9#)$r4BXd0cj?NyJ;z_(-3TV*sLJ*&s+
zVtyoh`kSo#Tf$~EZPwQAkt6Jn?Y9Oy<#<^^KR8G4A%88`Z6DU7eOOlc-s(J0SH&G&
zRebCx0DZ;)A+PcFn2%HG%21CT^I3)DTakpk_&zs;6&vJgzoCe}Wc$%c^~;}0s4pbD
zN;b4!Bz|Htm8?u`znih6y=p)8@x=Dy6|ytQN8Dv{J{zB3B9)3^*8VnbfA7<Q19CaJ
zRZ4@gaWeh0mD(t@b>~%9)Do}?9I7IMtNri?)OXNnv3a)c$tX&@ck&xDiWs5wPMP%&
z=b>`Uhqb>1hqb>2hqb>3x3#|qaCrnYHZVDtDLDuq1T;1<Ik!wX2%-s5LNr1#GDb!?
zH8?gwI5{~oLp4G$IYTfxI5<N^LpC=!J|H|oG(s^lMn*U_I5t5zIXN;zH9{~sLohiw
zI73B4Ha9pvT?#KuWo~D5Xdp2-Fqa`F2q}LhmrHCFRT#(To8H@d`+%0xLQCnpl*geK
z${R{cc@)Z{P|9PVyon3ZXzl`|#D$3q4AG4yCO$%lD+5LoT<Fma#%NAF3*v&6Mpx((
zV@yyY!TJ5^;@9*2XXebzng92F=O&3gdXgqdnpl!F<o}XbN|V&rq8Uih7NgY=R)K#4
zS_fL0un4NrR-%;%1&+pN&^VfE0%nd|&~#ZHs7D(?W9D=Yn2WX>O}A_SjcEJOc&c<B
zXhJ)VrpKKRn)BDmQZ(KuZ2=3=T63*(k#HesMQcO5K-dNrp)Er@PuLC?qb)}}N7w;6
z(N>|ICF}y-XsglA5cYsxwDoAG3HyIQKiUSgQ-n*v0NO^hS;D1Y8QLbalZ4B`3bZk_
z6NC<|L_2_%5e|ZU*)7zd9VZ+DtI!(IjuEZ~!)T3YM+w(}wP^Fuju5T`>(QFfW(YTc
zjcA={hY2@<&1n5-(}Y{V2%1AXM7R}<q79=RBpd_V&^Dt@5sriHXnW8O5bl2fJJBZ5
z_7m;`6KDs~_7Uy|d(iMkWiR0**o&5-Db&gP(N3W4A=Hhg&@P~9Dm~;Ow2NpHgnGgJ
z%~O}4=`pl8N6>bl?IhH~9z)xSrdQBM;W69<+IB*H(35ERzM|Kc8BU?;`)(tYk<Oq^
zp=s%4!gFZcs4_|@1M>=Q7EOQ4j`?Gr>}Tth?Bq;Tij!qg$w^AavL&UfIz`D|WP-I|
z5Rm5;WoR{{B|u)&M_$J*J2|SvCY_^BCW)hxSXtW)YAk8|K{q0Gqg+>Df-XSeaZ7<_
zSpZrAxf&ONZqQ=M;%3c4!p4QnqU7Y-ic(iOO&UZ!nx-FS)C>9ni5Pze0O=Q(0g^7R
z0HjOQ6`G~TlBMT#OL8!hbg_>~H34}P*8;L9k}h#0AW0&hBCfJz_^~F~45$%KFT)3o
zeCdb<BHoCYCgQM23Pr3JNs>5f$!M2m!IM#$qQ_N!=$SEpBvYb1r02tlj&kGpcur46
z5=RUiN&ARxBLyl_%_4s#C?2$AazJ;Zf<($qRO0CADJ_l)7qMfcSVl}3DX;M&Rf<lN
z$!4>b%zUL#COplk3KT#QRD%*I103dPJ8r!tUw7)-dN3C>fJQJ6G=cd*-Jk`i2dJ2L
z0F`jn`#MXWTcP<Df;rqnF2*#j7u13^9PI)<K()OOsG=_cs^fnu((RU9KB!5Tf^H@t
zg`D1FF3@GF-h+Ueol$%<s&eKmdA&{(3~};uM&n?OCD$(Nz&b$5&9(pvZMGGR@=-J&
zMK#+2b^+>Swi{3#vq?a$%eGtc?zASQL}vQ{g*MyIRCFuX1M1MIgxNGW3}(O)a1P`y
zx$&u{Itq?~;~;+nC%{QC3r>O4+(WB-2Al;K0Hz5{6%3>Rmjph1m;-7#@4O|S9n*Q(
zCt#m|eFF9gWh;DnHLrSif7dCPA>fUWfB5gNOT$9}g9I!PutjLL<i|I4l$}CoGkc*{
zVTai(ziNaTLc7_EPtU_#0dEA%5W3AS-_ZfA5-1Oqjh27D{E;+V67WV?X6aQg4KsuR
zORwBjNab>vI;fuy0hTEYTKdM6PEk`|ZRw-W6>b99BB*0;Nb<iYjIc$(C1I_lj|X+K
z8u*x1wZ5uQ{ddHwieD&Hi&p=|K0#d<4+ZtmajR+^)_H2ZyRC3-9(~a1wbW7%05#2h
zR(SU>oveSBrH8~h0p|pq6R=9a3;}P1X)D}#M^~tGvMX@<<`jy6O9Czl>Ur|UtQBru
z(G};d@VG|LQU%mb=wOcS;!-Qzxum0PbzK>#jj<bX^Fa$(0O%x+Euy3GIr^JxvO@e`
z6SRY7z|PVw0?Ji3xNfBt?v3d@_Jod&u%jtCCDVUaK$<#sb8Z_L2b6F(Vuhct>e}sK
z7tg0PVEgMxNjDAH@wiDV+`p@n_X2H6W`H&rhph0RsH57K90hEOoVFDwtng3JQSCfP
zM5n#QS)kQD544Za-Y2p0z+2k!7+{}Wsj_1EQewrL-*iCR1+6JdTq#?z_9q>nidO1C
zJ(z!E#k!k1pk1l9q1pjxqsgbJG+VL$x=v}a;?Z7>7Fe<J^a&PP@z!mPTCEuSHEOfs
z!x@bhS#`@^joPhx>JN<;TlK9WjXJFQmp?V?wCYEPH0rWa&j%WHTWR7Yje4x~?1V<W
zR{G?=MtxSgw@agbEB)P{Sn1)(9RI$b<G+88b3D13FH?FuNAa5+<p(+H`*Jj7Ihx+i
z(eiDMw)b*$JfEYREBZdvq}Zo2V&%=R5qcKPvtXXi<Y^tA1@kPJXTdxR=2<Y09X$s2
z7}#TAk2^gs_E^+oTd&`A`Vq$t-m@R_^e9i2_w+nZz4GkkJ(~nS0N7A@+M=g6`4t*~
stnm8nN0NB`;v2H^rt$oZb_^%~1G$6t*_U@p2p<YGIWP((B}Gq03N@vVHvj+t

delta 4400
zcmV-05zp?xmLZ^(A%L_2)pY|kHZYg5`3EY0QcZ8%AP~L#S8ywqVq)+oJ+|pat$Jy@
zQVwYk#7k^THr@ieRsMZvz`NZ|BTHr&=Dm3gFDE_8$$gT)S~>i4!bu~DkS%MZB;B0+
zPFPvD;piNMG|%%q?Bu#a_V)8E$la2BgYQI_0bh1O_+gveZVE{_XR@jU**=Gkm1MGi
ztjKmp9_eQ64_2p{C|fGtr&%Fu`gQ3uZ<TS+i=Fij4<$9)$ErW6Xy{`75VyP0c{@HG
z{J0nwQviz?HBB+|Gw0<Iacie>-XFX|NZilb4-14q=WbxFOWG}S%IRKN5b$gJ1@4#e
zl9gp0b$ci;L{0smF=$X$OZC_f*v6xOYCQh&aTAYsi$PVGyHw25tSl;er`&LYOry~*
zf~?+K?X*E^A5$=oQJGkIOsf*^cmSQ+F;sbe;M+~rT#g;o7X@2qUNO#!G~--U^dXhN
z;heV4>k)~YDkxN@(@6RH0f9*f#_M$s+q4x_JxzLTk2bLV>WA<533<#4UQ=;@r7k4H
z^<dWFG4aCuG))K|0_~+Ns9$B_bz|Zwz`+>j`D+vLT6LHl;%;8w7Ln6C7s0^;Vdzj1
z(Q5L;zg8B?*Jpam0m=Zj<V6&X!T%P(Y_I?Zs|7-MvH-SN%uRH`-Y$qdj=RzHSO_DC
z^$rr#+QoHhQY();%(JpDakD%z#}zXDXw>wU=u7WGMX-_Tx`S2S3y%ESMMQ{}@v;F9
z-m+H6NW$e+)u(Op7qnm`Ux&<i0f)?Z0*B0a1Gmh01oA!tH8zvcQz)0{w*)u}Fx>3!
zYO+{7RdtstxCESkcvT3-!E9oLOu;yUFR2DJYH%DPhA7E#<TQ|y9TP^DN{+?UD2fIQ
z8Hv&8I;@V>%u^wp#2VtD4rU6AawTh^_W(<%O)}Ivh?Jc~m`*BQ#}e=gGK4<VCMoDp
z0|wSiXxL0(RkdX)V*QvZ#mpZ~ImHtjIO#y6-E?E2uL$LT@y{uDcj$}78hFZuonjW+
z;R9CFgmDyR!74$Jpi{Mf6kQ5A4a}g4*j^<xvVj0kU@0^z8fqN^V_eaI96)BD<5gmN
z5?snKI8*@cFjfR$Qx)ipS)__3QUEyz8=aHO6!3&Iz^V46xNIg-jFl6xt-yKC#rlCv
zfp!ryf#Iz+0W4#PfeM~f6Gj&$Kyii9OUyR!><XK04Dc{C8r78)Afu4UYF~+inSo^(
zv{a}W0oqbi9+&yK1U?7=82rH-^e9J{Ke_}ff4`?I1o!g^f#OynC<s{nFny?~FNi+1
z_|A1jeFSI`4t(bHutM|s_{03F9e$ud+M{`tyv4N}b#nIRY{#qf5`6-4!M%0OFaQ3J
zLqN+geF6R_*PlMUe^8FhH+IUTl03l`cgm&c3*2a@9A|IQcb6Ozhfi?kHM!I2WYL-D
z5!4V5vvpsb!s`g3XL)1MAF-GDx&$|W*{FN;&-Q#gJehuR-GV0s##{+Y_+~i6w}{Aj
zDc50Tj!`WOnl6?E?fU*$OvUcz22%)%-g03o<h2Vy!B?b&RDtz;Kd6RRSMB6{{KcO<
z1>5U-z|`KdU5ddoau1PsQj7QGtK`S^c)khh6!ru4n%o1R-jLB;yrqksasse_@=h@p
z_5~^L0j#dUa}BtCz%y^bGwEUQd^MY1Uv-RX1FdPPnlY<8XzdhPebk&a?`Ffv{K~o<
zeRa)?v*u;{+j!KzeRblR*KRSpZhOI&gk{{IDBkF1R6q=Lj1WKhUSwU2FRy2<kNx30
zQuM}QbMQ;{R(t_9D(t~B@6};{G5q*xLxm1MX-{b1kz-H_kVda?u-!q?;PlJCKK{@2
zws)^T8%58LLk&_u*Na_{%en=*%m+d4&J?rkXNomBq!D<qQSx1KqQNUt!aivPw2^MI
z&?&qD<p1Uk=!}%@Y;&yBmauKXvY!PD?{M1OYihx!4B!S`0u3<c9}(JrD?(kY)nnW6
z2<T}O^m$K`a#zkkPezq$eOwSrO_aRtRu5zh+n!?kMC!#;>Na;M^{{pRwy<Tz9m*3I
zc%6@83+lVgB`%H&R@Zh}2)0qm!))}TJqHiv`RJlQX3QY!wsBahA~D*w;Ydvk9(7Ot
zQ1{>u8#0_gw2~ekRn`N4$=wjMMq7IbF|}+#ucn?tOkL_}m5Clh4L0gnYTG3jtUQ?o
ze6U?J(rpg+hno5I$2mS8PbXfy^q1)op?!e~BbZ&{WMvYcMW9<UM*B$E<1HBGJz<#l
z$n{$t(lQd~a`;ARgM6;XK&xMsK4PF%Z#IBvRi@)&pjEX@C#&FpEU(&UEM%!zWtat<
z@@jRcfPG7}3>1BR0D<4LR$#%;iUha%F#A~PPt<3_E3$O83rg&*Fen9jXzKKnN4-Ke
zyu#{~`*_8=-2K6MQ--k{4q;Eq_ss`-55XGK7V@JUjr?9|9SFNGV$~h?)r#j>$poaO
z3Yu5S0r5?5Y`JTHCD-KcS;+@h=+2@XyAC8<>p;{`D%2H3tl-^UDXgxs&%l(C3e3-t
zja02vE#6Ksti+0txL;A#Lc!<pXf_>xzPett{Y9y$(UWb(Z?wzFC(^%&cEsu_cYAiS
ztcv}NopcmLGX2rIaj0NA=~g!ehXLW~NjGyk;{BZSuAGs74<+4>p<5k-rE~Qpd8Ztc
zZP&SHl47S)+zUxz-U<;M_X-ipx)~E_6$?1@8L}C)nC#FsA=!$Z5&}GSs)o?bRSzhx
zK2Z;lkdOpR3R-S@!0lO<=uNxyIN1a!(|!P5lY2CKe9vmD$~I2qXq<3b;UurFt6akM
zJLC=yp(}ZRN9<&{+R-$EO|B0|nct_=^V#*q1tN^Jlm8AUqw(aj$E;C}OV(SxC{}No
z`bX;}F?&0Mj^Mai?Uw?GgrIE^95+9#)$r4BXd0cj?NyJ;z_+8FO|O*E0D4xB)y4cs
z_VhPd_qT-2Xj0iJN7x^0j=SY}SwcUUqxV3|b=!x3^=KcK72jJO=;^AsqpOOK{RW`V
z93bR1-k$SuDqR`sv2#AFkbEnWkQd+QhOlCTT*ZQ*FFD!IPka^JtNKE+t7Jp#MdBxt
zsbpp1Y?Uu>A=Eaj#NF>z>`?L%cbS~e#^;xyQZdZh-^T6leL8Remy=tiG*~pQf3{K^
zg|_Z1z{-kRf~W$As>tBl{r1T7LC3}B*}7*$QQEze-;hz{2(5R@tbYKl%X1c&p#=jI
zhikwDhikwEhikwFw`;%$U3mmDGB7ch%sB`j1Tr!(F}LqI2%-s5I7CJ_G%`XnK}9t%
zMn*C-Gc`goH$yc>L^U`$GDb8uJ|H|eL`F9>GD0&!MKv%+Mlv%qH9|5sLp4T3H8?pk
zMl?1)T?#KuWo~D5Xdp2-Fq6<zDSsuGOKeqD6o$__y|?$$0xhL4=)0vYuv;GG)j}z>
zlt+P5%JY=+#+aBG<4tU0oCr=(W1<OhX3*$BgOBLI$8n;GiR`dVoG{VgOigs)gop^3
zdcGe9AA9|0pMB5S=U;2>Mv*<3kE1A#EQ+eK|DsTcqd2WVGZ3RSqb(pT1AjTR7PKN^
z9+ac?p%n-Pu7)*eT#dB=D~D}px~&o{L>oq9<#-WTjJ6j|KUoE;(GH;TRPho}gLV>4
zkGmAqX0MaPXuMHe2kOxpGp%%qumLopEknCVxC|^uTZwjoun9DytwK9b*aBM7)}fsv
zYy<6R>(R~<c7RT_&1h!`yMMq6v@K|-3A;fL+9292VJ}#THiUMHa1~gMwgYX3(1AX*
zNwkEpA7tBZegWD^!nI%>S{2#}!u4PPts3n(;Rdh~Z3)^j!cAZ^S`FG!!YyDBtrhJE
z;Sku0wgPRMa2ps#b7+SNw}TP10kkQ?9bhNgR<uKeqhJheAKE10E`KnNHi34Ka5vb4
zHidS8a4*=0hBr$42`9jQv;<9|CLctbLEA^DA3B6~5lu_!ArGTnLfb>A7tG!~b`hE$
zLz{C9Z5P@&p*Hpe+BlkCK_7+3aC^|k2=zfvq2c?IUSDQ7ji&FrlTb!Fi*^W2TPG8q
zN8^V|BZM+Aui$3Uw14fGKV->%ZlRK$oC!*CvMeY$Ny$*OsBl$Nl<avH*a-Rod0tY6
zmNV)G@|r&KCVsM&t4eIrIcPFT9F)Y$+FG!{qU!JUL!@qy>vAm61}HqP7ig7w&<M!Y
zupG35I*XcXwF(Iv8dyci$-O0|u5y|*2zoRvKftIHtN<ip=zjsEUswr9y099ME<tx_
zl@5z~&+8}2!9dc5E*8}S<WblN$euvDgh4=(1U^MrXVJiaT3{=nMmW6;A2jf#0~QE)
zBVd|<!vZN3uwEcZ!iYs9ZCV9S24#vKSNWl5#{7Xy3G$Gh4<|axjpO4vJrzkDFmNF4
z1GWtms6aIflz*TwWzj^BevJweC^tcgqo=2|I4WGgj)7tsFkzs)hD%f_I!z**&02Ky
zD}^%Q8AfFw2lAjC6hIN+Fh|>Qn=SgfRrf9gi$N8r21`H<SPIk)>VSHHig^oA30J+Z
zwCJhTTCV{t;y2`COyfF11=zsVHqZf7+q-}&dN)uVSAUUivS@Bfi}r$c79fS3-eWP)
zZK~e=fSR39d=sj2;w-vcsRh<Dc@?8ku)(5N=X7BcpyVdo0EIT$4o3JWT92Zd>;k(1
zbu!rtsE)}5pw=a07G0axqLj$w0HDw&2U&`4<vKtm8kI1a21meAa15LWnM>aJNJ||D
zC%{RNfPWcq3e1Aj;0(W^-8~D=fr|jsc$V@8Vt`9LAKotl70kO}(I+P~5BqrR<FSv&
zKE7zVFaFG`-jBa)3TE(l!)G7<yX(^MkjEe%3wUhdYc0C<vaWJc@XO4eu29%w_WY<u
zn87!hJ@a@T=JI&MV+P-DHutSAU=>e!C=FUXH-9e;mw3G4S6Y0vSsG^WJr=((ppeSt
zFtt}dUklhK-*53tpJ|Gk`g)5W)F>PR*utx0Z;7&h^Ng^C$0dHF#gE?5WHs;|R@QJw
zq5AKzmF4Rdszs~+Vjr)ri-){==%|%dy{UO>y?ZV9%5(HVr`J+Todjx{2P}8(xF)M*
z>3<<{j>kD3=Xk8*F@wh&e%f+x|E)XJIXM+LeRB$h$0Z(@c=bGaW7cvve$gElEcfV1
zJxdu-JE4O)x{He~_svaR<*4h5Ky8fEh+7KkKs}(7IF5*p#^>m7uEuhqQwubKTENND
zEeFa~4!Ew*a<|{oJkErUgRrA1IVIC}K!2J#PIGQ27zLDYH*C3|f7QKXU^maFJ>dB3
zNJ%#hIPtg%%l+1;$@_thBu9Y`7>6zQ*F{~`x#T$DP~>#1n6cdbZe7*MgG6*XTbu*h
z-3vhH2%UW*D|Y{(6ORGT*`+ee7j8zDUofl-Ixc8W+2T^s@)c`zfht<61Pj3;%YRoE
zbU~+59Yb{j&_R<=QL4?RJk*pr%O87Dqk79f`}heOEdODZMva!gbw#6PmcRF@M$4_d
z_I-_-to+cpM$J}!<28+1to+WTMy*!<@5dUoS)u(OjoPiS`v;9WtnkEp8g*LXgE5V|
ztZ@5XjaFFU?v=<2_rA<<|6YcN^M4uU3)wcsRT=V)8HxupR9?=o`11@)vSL)MYt7Jj
zH$$_}(9Rv5_q8bYDGgh3=qJLI4NKXul+KjWI#M<)Wy4Z7EM>z|HY~-CDF#k4aEgIb
z+?nFy6pN<VHr3yB`Xi1Le9Cz!rAMVy`IMfQQm<0Z@+pUev<GmYN@<HJwH+y44akaA
qfBTUnss7?yYsJCq*&B_$9{mrlr1$NY7fT2q3Ntb{3MC~)Peuy8hE8$-

diff --git a/paper/sections/results.tex b/paper/sections/results.tex
index 6c1f130..d6f81d5 100644
--- a/paper/sections/results.tex
+++ b/paper/sections/results.tex
@@ -47,7 +47,7 @@ \subsubsection{Imputed wealth distribution comparisons}
 \begin{figure}[H]
     \centering
     \includegraphics[width=\textwidth]{figures/models_dist_comparison.png}
-    \caption{Log-transformed net worth distributions for all five imputation methods, compared to the weighted SCF donor distribution. Dashed lines represent median values.}
+    \caption{Log-transformed net worth distributions for all five imputation methods, compared to the weighted SCF donor distribution. The y-axis shows the percentage of observations falling in each bin, which normalizes for the different sample sizes between the SCF and CPS and allows direct shape comparison. Dashed lines represent median values.}
     \label{fig:imputation_distributions_method_comparison}
 \end{figure}
 

From 2a015c95687cfa7c7083950e910c3e2fc22ae5cf Mon Sep 17 00:00:00 2001
From: juaristi22 <juaristi@uni.minerva.edu>
Date: Fri, 27 Feb 2026 14:39:57 +0530
Subject: [PATCH 3/5] documentation updating

---
 docs/_toc.yml                                 |    2 -
 docs/autoimpute/autoimpute.ipynb              | 1592 ++++++++---------
 docs/autoimpute/index.md                      |    4 +-
 .../benchmarking-methods.ipynb                |  170 +-
 .../cross-validation.md                       |    9 +-
 docs/imputation-benchmarking/index.md         |    4 +-
 docs/imputation-benchmarking/metrics.md       |   82 +-
 docs/imputation-benchmarking/preprocessing.md |    4 +-
 .../imputation-benchmarking/visualizations.md |    4 +-
 docs/index.md                                 |   18 +-
 docs/models/imputer/implement-new-model.md    |   54 +-
 docs/models/imputer/index.md                  |    8 +-
 docs/models/matching/index.md                 |   20 +-
 .../models/matching/matching-imputation.ipynb |  366 ++--
 docs/models/mdn/index.md                      |    4 +-
 docs/models/mdn/mdn-imputation.ipynb          |  781 ++++----
 docs/models/ols/index.md                      |   10 +-
 docs/models/ols/ols-imputation.ipynb          |  188 +-
 docs/models/qrf/index.md                      |   14 +-
 docs/models/qrf/qrf-imputation.ipynb          |  154 +-
 docs/models/quantreg/index.md                 |   16 +-
 .../models/quantreg/quantreg-imputation.ipynb |   78 +-
 docs/myst.yml                                 |    2 -
 docs/use_cases/index.md                       |   96 +-
 .../autoimpute_best_model_imputations.png     |  Bin 56963 -> 0 bytes
 .../autoimpute_model_comparison.png           |  Bin 53000 -> 0 bytes
 .../by_income_decile_comparisons.png          |  Bin 55602 -> 0 bytes
 .../imputations_model_comparison.png          |  Bin 90532 -> 0 bytes
 .../scf_to_cps/imputing-from-scf-to-cps.md    |  924 ----------
 29 files changed, 1997 insertions(+), 2607 deletions(-)
 delete mode 100644 docs/use_cases/scf_to_cps/autoimpute_best_model_imputations.png
 delete mode 100644 docs/use_cases/scf_to_cps/autoimpute_model_comparison.png
 delete mode 100644 docs/use_cases/scf_to_cps/by_income_decile_comparisons.png
 delete mode 100644 docs/use_cases/scf_to_cps/imputations_model_comparison.png
 delete mode 100644 docs/use_cases/scf_to_cps/imputing-from-scf-to-cps.md

diff --git a/docs/_toc.yml b/docs/_toc.yml
index dc6d677..cffb6d3 100644
--- a/docs/_toc.yml
+++ b/docs/_toc.yml
@@ -38,5 +38,3 @@ parts:
   - caption: Use cases
     chapters:
       - file: use_cases/index
-        sections:
-          - file: use_cases/scf_to_cps/imputing-from-scf-to-cps
diff --git a/docs/autoimpute/autoimpute.ipynb b/docs/autoimpute/autoimpute.ipynb
index 988a9aa..59fde5a 100644
--- a/docs/autoimpute/autoimpute.ipynb
+++ b/docs/autoimpute/autoimpute.ipynb
@@ -68,7 +68,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 5,
+   "execution_count": 1,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -293,13 +293,13 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 4,
    "metadata": {},
    "outputs": [
     {
      "data": {
       "application/vnd.jupyter.widget-view+json": {
-       "model_id": "0fb0cd8da8ad4d65bf84cb943ebaf1c8",
+       "model_id": "08c1a0e8ab514a84a14be53225051e6b",
        "version_major": 2,
        "version_minor": 0
       },
@@ -315,18 +315,11 @@
      "output_type": "stream",
      "text": [
       "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 8 concurrent workers.\n",
+      "[Parallel(n_jobs=-1)]: Done   3 out of   3 | elapsed:    7.3s finished\n",
       "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 8 concurrent workers.\n",
+      "[Parallel(n_jobs=-1)]: Done   3 out of   3 | elapsed:    1.0s finished\n",
       "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 8 concurrent workers.\n",
-      "R callback write-console: Warning:  \n",
-      "R callback write-console:  failed to download mirrors file (cannot open URL 'https://cran.r-project.org/CRAN_mirrors.csv'); using local file '/opt/homebrew/Cellar/r/4.5.0/lib/R/doc/CRAN_mirrors.csv'\n",
-      "  \n",
-      "[Parallel(n_jobs=1)]: Done   1 tasks      | elapsed:  1.6min\n",
-      "R callback write-console: Warning:  \n",
-      "R callback write-console:  failed to download mirrors file (cannot open URL 'https://cran.r-project.org/CRAN_mirrors.csv'); using local file '/opt/homebrew/Cellar/r/4.5.0/lib/R/doc/CRAN_mirrors.csv'\n",
-      "  \n",
-      "R callback write-console: Warning:  \n",
-      "R callback write-console:  failed to download mirrors file (cannot open URL 'https://cran.r-project.org/CRAN_mirrors.csv'); using local file '/opt/homebrew/Cellar/r/4.5.0/lib/R/doc/CRAN_mirrors.csv'\n",
-      "  \n"
+      "[Parallel(n_jobs=-1)]: Done   3 out of   3 | elapsed:    0.8s finished\n"
      ]
     },
     {
@@ -341,7 +334,8 @@
      "name": "stderr",
      "output_type": "stream",
      "text": [
-      "[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:  5.2min finished\n"
+      "[Parallel(n_jobs=1)]: Done   1 tasks      | elapsed:    1.1s\n",
+      "[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:    1.2s finished\n"
      ]
     }
    ],
@@ -375,7 +369,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 5,
    "metadata": {},
    "outputs": [
     {
@@ -384,17 +378,17 @@
      "text": [
       "Cross-validation results for different imputation methods:\n",
       "\n",
-      "Model: QRF\n",
-      "quantile loss results: 0.0155\n",
-      "\n",
       "Model: OLS\n",
-      "quantile loss results: 0.0124\n",
+      "quantile loss results: 0.0126\n",
       "\n",
       "Model: QuantReg\n",
-      "quantile loss results: 0.0125\n",
+      "quantile loss results: 0.0126\n",
+      "\n",
+      "Model: QRF\n",
+      "quantile loss results: 0.0160\n",
       "\n",
       "Model: Matching\n",
-      "quantile loss results: 0.0231\n"
+      "quantile loss results: 0.0233\n"
      ]
     }
    ],
@@ -415,14 +409,14 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 6,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "Best performing method: <microimpute.models.ols.OLSResults object at 0x17b05c1a0>\n"
+      "Best performing method: <microimpute.models.ols.OLSResults object at 0x177c5e900>\n"
      ]
     }
    ],
@@ -442,7 +436,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 7,
    "metadata": {},
    "outputs": [
     {
@@ -454,307 +448,285 @@
        "data": [
         {
          "alignmentgroup": "True",
-         "hovertemplate": "Method=QRF<br>Quantiles=%{x}<br>Test Quantile loss=%{y}<extra></extra>",
-         "legendgroup": "QRF",
+         "error_y": {
+          "array": [
+           0.0000367764891179606,
+           0.00005051876144414505,
+           0.00018829257114425592,
+           0.00037848750942190044,
+           0.0005477958061141973,
+           0.0006245017815431527,
+           0.0006789805698947316,
+           0.0007155653018236752,
+           0.0007295643674175897,
+           0.0007569494312623643,
+           0.0008639491258964105,
+           0.0009801026362712217,
+           0.0010518961452707697,
+           0.0010297648072906266,
+           0.0009590446159744011,
+           0.0008660289152915666,
+           0.0007846310723713395,
+           0.00065956869485099,
+           0.0003346140844985172
+          ]
+         },
+         "hovertemplate": "Method=OLS<br>Quantiles=%{x}<br>Quantile loss=%{y}<extra></extra>",
+         "legendgroup": "OLS",
          "marker": {
-          "color": "#636EFA",
+          "color": "#88CCEE",
           "pattern": {
            "shape": ""
           }
          },
-         "name": "QRF",
-         "offsetgroup": "QRF",
+         "name": "OLS",
+         "offsetgroup": "OLS",
          "orientation": "v",
          "showlegend": true,
          "textposition": "auto",
          "type": "bar",
          "x": [
           "0.05",
-          "0.05",
-          "0.1",
           "0.1",
           "0.15",
-          "0.15",
-          "0.2",
           "0.2",
           "0.25",
-          "0.25",
           "0.3",
-          "0.3",
-          "0.35",
           "0.35",
           "0.4",
-          "0.4",
-          "0.45",
           "0.45",
           "0.5",
-          "0.5",
           "0.55",
-          "0.55",
-          "0.6",
           "0.6",
           "0.65",
-          "0.65",
           "0.7",
-          "0.7",
-          "0.75",
           "0.75",
           "0.8",
-          "0.8",
           "0.85",
-          "0.85",
-          "0.9",
           "0.9",
-          "0.95",
           "0.95"
          ],
          "xaxis": "x",
          "y": [
-          0.004631754564085741,
-          0.004631754564085741,
-          0.007341856096256823,
-          0.007341856096256823,
-          0.010956305360759139,
-          0.010956305360759139,
-          0.013853899253214152,
-          0.013853899253214152,
-          0.016765778862585467,
-          0.016765778862585467,
-          0.018170938524875734,
-          0.018170938524875734,
-          0.020431459700043803,
-          0.020431459700043803,
-          0.021268421552075877,
-          0.021268421552075877,
-          0.02152068721079699,
-          0.02152068721079699,
-          0.02150832954854852,
-          0.02150832954854852,
-          0.019861473819837334,
-          0.019861473819837334,
-          0.01995636925773867,
-          0.01995636925773867,
-          0.01894943399516724,
-          0.01894943399516724,
-          0.017987324791786673,
-          0.017987324791786673,
-          0.016858787546663905,
-          0.016858787546663905,
-          0.014928394018716465,
-          0.014928394018716465,
-          0.012899519430338449,
-          0.012899519430338449,
-          0.009831864043753134,
-          0.009831864043753134,
-          0.006811614968932836,
-          0.006811614968932836
+          0.0037857585632115076,
+          0.0064035381731015984,
+          0.008682116948116709,
+          0.010637028223496603,
+          0.01228354449019281,
+          0.013704743173963509,
+          0.014829116089757849,
+          0.01572984140710315,
+          0.01635768192573994,
+          0.016757835444730255,
+          0.01698876716674359,
+          0.016908544804384324,
+          0.016466981551585432,
+          0.01573902508984663,
+          0.014655697638186978,
+          0.013182752515157806,
+          0.011274496133844888,
+          0.008723043411546087,
+          0.005424527941128667
          ],
          "yaxis": "y"
         },
         {
          "alignmentgroup": "True",
-         "hovertemplate": "Method=OLS<br>Quantiles=%{x}<br>Test Quantile loss=%{y}<extra></extra>",
-         "legendgroup": "OLS",
+         "error_y": {
+          "array": [
+           0.00011053862485353776,
+           0.00013152308789946208,
+           0.00007328304807248766,
+           0.00013761220727722502,
+           0.000225390831825253,
+           0.0002387781342265084,
+           0.0003690300025792442,
+           0.0004971448458544792,
+           0.0006841372848715391,
+           0.0007855762966044914,
+           0.0008657286666497124,
+           0.0009609612674453248,
+           0.0010760799108166322,
+           0.0011252152270869854,
+           0.0012113725779785698,
+           0.0011552041991586672,
+           0.0008205546613073608,
+           0.0006875490202660445,
+           0.000559862252788657
+          ]
+         },
+         "hovertemplate": "Method=QuantReg<br>Quantiles=%{x}<br>Quantile loss=%{y}<extra></extra>",
+         "legendgroup": "QuantReg",
          "marker": {
-          "color": "#EF553B",
+          "color": "#CC6677",
           "pattern": {
            "shape": ""
           }
          },
-         "name": "OLS",
-         "offsetgroup": "OLS",
+         "name": "QuantReg",
+         "offsetgroup": "QuantReg",
          "orientation": "v",
          "showlegend": true,
          "textposition": "auto",
          "type": "bar",
          "x": [
-          "0.05",
           "0.05",
           "0.1",
-          "0.1",
-          "0.15",
           "0.15",
           "0.2",
-          "0.2",
-          "0.25",
           "0.25",
           "0.3",
-          "0.3",
           "0.35",
-          "0.35",
-          "0.4",
           "0.4",
           "0.45",
-          "0.45",
           "0.5",
-          "0.5",
-          "0.55",
           "0.55",
           "0.6",
-          "0.6",
           "0.65",
-          "0.65",
-          "0.7",
           "0.7",
           "0.75",
-          "0.75",
-          "0.8",
           "0.8",
           "0.85",
-          "0.85",
           "0.9",
-          "0.9",
-          "0.95",
           "0.95"
          ],
          "xaxis": "x",
          "y": [
-          0.003808647070946412,
-          0.003808647070946412,
-          0.006490379484883273,
-          0.006490379484883273,
-          0.008807739971118509,
-          0.008807739971118509,
-          0.010699017371221815,
-          0.010699017371221815,
-          0.012339891153334075,
-          0.012339891153334075,
-          0.013697475098254898,
-          0.013697475098254898,
-          0.014736873094883992,
-          0.014736873094883992,
-          0.015550928634621153,
-          0.015550928634621153,
-          0.016134574039820177,
-          0.016134574039820177,
-          0.016485493155635602,
-          0.016485493155635602,
-          0.016654564673178742,
-          0.016654564673178742,
-          0.016593511197972737,
-          0.016593511197972737,
-          0.016203732622264988,
-          0.016203732622264988,
-          0.015451587766895878,
-          0.015451587766895878,
-          0.014415220561443473,
-          0.014415220561443473,
-          0.013044914508801573,
-          0.013044914508801573,
-          0.011156959998495038,
-          0.011156959998495038,
-          0.008585027733878748,
-          0.008585027733878748,
-          0.005229999830606041,
-          0.005229999830606041
+          0.003648297363886403,
+          0.006220739206137793,
+          0.008560805562643258,
+          0.010594467066356217,
+          0.012137407181277772,
+          0.013580247763948514,
+          0.014709634995905748,
+          0.015584271813924223,
+          0.01615515457661314,
+          0.01660627579116468,
+          0.016924948899226498,
+          0.017013821272759678,
+          0.016792947494266755,
+          0.01604088566988471,
+          0.01484899400756596,
+          0.01347829835941078,
+          0.011713354867882659,
+          0.008831687333185652,
+          0.005731739757859993
          ],
          "yaxis": "y"
         },
         {
          "alignmentgroup": "True",
-         "hovertemplate": "Method=QuantReg<br>Quantiles=%{x}<br>Test Quantile loss=%{y}<extra></extra>",
-         "legendgroup": "QuantReg",
+         "error_y": {
+          "array": [
+           0.0002099742280199984,
+           0.0005577204009778338,
+           0.0005913787202045731,
+           0.0006465761742329032,
+           0.0006427657425320314,
+           0.0009665112262689732,
+           0.0004039564380447954,
+           0.00046810196513831623,
+           0.000920067680617316,
+           0.0004569072602015561,
+           0.0007209820742233905,
+           0.0009019129537077271,
+           0.000654822071277212,
+           0.0002964891583124209,
+           0.0008697756394143753,
+           0.000869354334057907,
+           0.0007434922481074839,
+           0.0008949128843092795,
+           0.0005736016727487648
+          ]
+         },
+         "hovertemplate": "Method=QRF<br>Quantiles=%{x}<br>Quantile loss=%{y}<extra></extra>",
+         "legendgroup": "QRF",
          "marker": {
-          "color": "#00CC96",
+          "color": "#DDCC77",
           "pattern": {
            "shape": ""
           }
          },
-         "name": "QuantReg",
-         "offsetgroup": "QuantReg",
+         "name": "QRF",
+         "offsetgroup": "QRF",
          "orientation": "v",
          "showlegend": true,
          "textposition": "auto",
          "type": "bar",
          "x": [
           "0.05",
-          "0.05",
-          "0.1",
           "0.1",
           "0.15",
-          "0.15",
-          "0.2",
           "0.2",
           "0.25",
-          "0.25",
           "0.3",
-          "0.3",
-          "0.35",
           "0.35",
           "0.4",
-          "0.4",
           "0.45",
-          "0.45",
-          "0.5",
           "0.5",
           "0.55",
-          "0.55",
-          "0.6",
           "0.6",
           "0.65",
-          "0.65",
           "0.7",
-          "0.7",
-          "0.75",
           "0.75",
           "0.8",
-          "0.8",
           "0.85",
-          "0.85",
-          "0.9",
           "0.9",
-          "0.95",
           "0.95"
          ],
          "xaxis": "x",
          "y": [
-          0.0037797191400323325,
-          0.0037797191400323325,
-          0.0065355258283926075,
-          0.0065355258283926075,
-          0.00877891950537431,
-          0.00877891950537431,
-          0.010730211151855228,
-          0.010730211151855228,
-          0.012356819692361727,
-          0.012356819692361727,
-          0.013725116901874718,
-          0.013725116901874718,
-          0.014793067048809186,
-          0.014793067048809186,
-          0.015515637986475386,
-          0.015515637986475386,
-          0.015911871896009084,
-          0.015911871896009084,
-          0.016247057308316024,
-          0.016247057308316024,
-          0.016486402970914003,
-          0.016486402970914003,
-          0.016660653134295333,
-          0.016660653134295333,
-          0.016356605390546602,
-          0.016356605390546602,
-          0.015780547245353785,
-          0.015780547245353785,
-          0.014807938631402215,
-          0.014807938631402215,
-          0.01340137722886329,
-          0.01340137722886329,
-          0.01153643264671414,
-          0.01153643264671414,
-          0.009016559672302424,
-          0.009016559672302424,
-          0.005650809365013223,
-          0.005650809365013223
+          0.004958981353614306,
+          0.007709876078177458,
+          0.010610160150823023,
+          0.013147576784275966,
+          0.016230808947783,
+          0.01765523354548008,
+          0.020390817639546265,
+          0.021631607060529953,
+          0.022142205928372478,
+          0.022393833653832424,
+          0.021357150137205804,
+          0.020907723382855792,
+          0.019930544841328238,
+          0.019095495355673266,
+          0.01816255322567264,
+          0.016320905888505866,
+          0.01317617842711581,
+          0.010650497941205097,
+          0.0072062531829631795
          ],
          "yaxis": "y"
         },
         {
          "alignmentgroup": "True",
-         "hovertemplate": "Method=Matching<br>Quantiles=%{x}<br>Test Quantile loss=%{y}<extra></extra>",
+         "error_y": {
+          "array": [
+           0.0019368529298846857,
+           0.001756890524910662,
+           0.0015769649797891588,
+           0.0013970905356792654,
+           0.0012172898458174732,
+           0.0010376012524176477,
+           0.0008580951787241413,
+           0.0006789164152700594,
+           0.0005004166749212909,
+           0.0003237211525341078,
+           0.00015512287742088833,
+           0.00009066430412933424,
+           0.0002393033878359685,
+           0.0004137649009508685,
+           0.0005916298348888364,
+           0.0007705449475824869,
+           0.0009499170265880006,
+           0.0011295283905246845,
+           0.0013092805654417116
+          ]
+         },
+         "hovertemplate": "Method=Matching<br>Quantiles=%{x}<br>Quantile loss=%{y}<extra></extra>",
          "legendgroup": "Matching",
          "marker": {
-          "color": "#AB63FA",
+          "color": "#117733",
           "pattern": {
            "shape": ""
           }
@@ -766,85 +738,47 @@
          "textposition": "auto",
          "type": "bar",
          "x": [
-          "0.05",
           "0.05",
           "0.1",
-          "0.1",
-          "0.15",
           "0.15",
           "0.2",
-          "0.2",
           "0.25",
-          "0.25",
-          "0.3",
           "0.3",
           "0.35",
-          "0.35",
           "0.4",
-          "0.4",
-          "0.45",
           "0.45",
           "0.5",
-          "0.5",
           "0.55",
-          "0.55",
-          "0.6",
           "0.6",
           "0.65",
-          "0.65",
-          "0.7",
           "0.7",
           "0.75",
-          "0.75",
           "0.8",
-          "0.8",
-          "0.85",
           "0.85",
           "0.9",
-          "0.9",
-          "0.95",
           "0.95"
          ],
          "xaxis": "x",
          "y": [
-          0.023393616982767946,
-          0.023393616982767946,
-          0.02335740713728864,
-          0.02335740713728864,
-          0.023321197291809342,
-          0.023321197291809342,
-          0.02328498744633005,
-          0.02328498744633005,
-          0.023248777600850746,
-          0.023248777600850746,
-          0.02321256775537145,
-          0.02321256775537145,
-          0.023176357909892153,
-          0.023176357909892153,
-          0.023140148064412853,
-          0.023140148064412853,
-          0.023103938218933556,
-          0.023103938218933556,
-          0.02306772837345426,
-          0.02306772837345426,
-          0.02303151852797496,
-          0.02303151852797496,
-          0.022995308682495663,
-          0.022995308682495663,
-          0.022959098837016367,
-          0.022959098837016367,
-          0.022922888991537067,
-          0.022922888991537067,
-          0.02288667914605777,
-          0.02288667914605777,
-          0.022850469300578474,
-          0.022850469300578474,
-          0.022814259455099174,
-          0.022814259455099174,
-          0.022778049609619877,
-          0.022778049609619877,
-          0.02274183976414058,
-          0.02274183976414058
+          0.02420659950459554,
+          0.02410222421800216,
+          0.023997848931408782,
+          0.023893473644815394,
+          0.023789098358222013,
+          0.023684723071628624,
+          0.023580347785035247,
+          0.023475972498441865,
+          0.02337159721184848,
+          0.02326722192525509,
+          0.02316284663866171,
+          0.023058471352068326,
+          0.022954096065474938,
+          0.02284972077888156,
+          0.022745345492288172,
+          0.022640970205694794,
+          0.02253659491910141,
+          0.022432219632508025,
+          0.022327844345914644
          ],
          "yaxis": "y"
         }
@@ -858,51 +792,51 @@
          },
          "tracegroupgap": 0
         },
-        "paper_bgcolor": "#F0F0F0",
-        "plot_bgcolor": "#F0F0F0",
+        "paper_bgcolor": "#FAFAFA",
+        "plot_bgcolor": "#FAFAFA",
         "shapes": [
          {
           "line": {
-           "color": "#636EFA",
+           "color": "#88CCEE",
            "dash": "dot",
            "width": 2
           },
-          "name": "QRF Mean",
+          "name": "OLS Mean",
           "type": "line",
           "x0": -0.5,
           "x1": 18.5,
-          "y0": 0.015501800660325096,
-          "y1": 0.015501800660325096
+          "y0": 0.012554475825886228,
+          "y1": 0.012554475825886228
          },
          {
           "line": {
-           "color": "#EF553B",
+           "color": "#CC6677",
            "dash": "dot",
            "width": 2
           },
-          "name": "OLS Mean",
+          "name": "QuantReg Mean",
           "type": "line",
           "x0": -0.5,
           "x1": 18.5,
-          "y0": 0.012425607261487217,
-          "y1": 0.012425607261487217
+          "y0": 0.01258810415704739,
+          "y1": 0.01258810415704739
          },
          {
           "line": {
-           "color": "#00CC96",
+           "color": "#DDCC77",
            "dash": "dot",
            "width": 2
           },
-          "name": "QuantReg Mean",
+          "name": "QRF Mean",
           "type": "line",
           "x0": -0.5,
           "x1": 18.5,
-          "y0": 0.012530066986573978,
-          "y1": 0.012530066986573978
+          "y0": 0.01598307386973477,
+          "y1": 0.01598307386973477
          },
          {
           "line": {
-           "color": "#AB63FA",
+           "color": "#117733",
            "dash": "dot",
            "width": 2
           },
@@ -910,8 +844,8 @@
           "type": "line",
           "x0": -0.5,
           "x1": 18.5,
-          "y0": 0.02306772837345426,
-          "y1": 0.02306772837345426
+          "y0": 0.023267221925255092,
+          "y1": 0.023267221925255092
          }
         ],
         "template": {
@@ -1743,7 +1677,11 @@
           0,
           1
          ],
+         "gridcolor": "#E5E5E5",
+         "gridwidth": 1,
+         "linecolor": "#CCCCCC",
          "showgrid": false,
+         "showline": true,
          "title": {
           "font": {
            "size": 12
@@ -1758,12 +1696,16 @@
           0,
           1
          ],
-         "showgrid": false,
+         "gridcolor": "#E5E5E5",
+         "gridwidth": 1,
+         "linecolor": "#CCCCCC",
+         "showgrid": true,
+         "showline": true,
          "title": {
           "font": {
            "size": 12
           },
-          "text": "Test Quantile loss"
+          "text": "Quantile loss"
          },
          "zeroline": false
         }
@@ -1806,7 +1748,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 8,
    "metadata": {},
    "outputs": [
     {
@@ -1844,28 +1786,28 @@
        "  <tbody>\n",
        "    <tr>\n",
        "      <th>0</th>\n",
-       "      <td>0.020909</td>\n",
-       "      <td>0.036174</td>\n",
+       "      <td>0.015232</td>\n",
+       "      <td>0.013791</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>1</th>\n",
-       "      <td>0.004626</td>\n",
-       "      <td>-0.021702</td>\n",
+       "      <td>-0.015421</td>\n",
+       "      <td>-0.006742</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>2</th>\n",
-       "      <td>-0.018766</td>\n",
-       "      <td>-0.004827</td>\n",
+       "      <td>-0.005829</td>\n",
+       "      <td>0.008123</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>3</th>\n",
-       "      <td>-0.025080</td>\n",
-       "      <td>-0.045854</td>\n",
+       "      <td>-0.027782</td>\n",
+       "      <td>-0.007257</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>4</th>\n",
-       "      <td>0.000522</td>\n",
-       "      <td>0.021657</td>\n",
+       "      <td>-0.029530</td>\n",
+       "      <td>-0.002555</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
@@ -1873,14 +1815,14 @@
       ],
       "text/plain": [
        "         s1        s4\n",
-       "0  0.020909  0.036174\n",
-       "1  0.004626 -0.021702\n",
-       "2 -0.018766 -0.004827\n",
-       "3 -0.025080 -0.045854\n",
-       "4  0.000522  0.021657"
+       "0  0.015232  0.013791\n",
+       "1 -0.015421 -0.006742\n",
+       "2 -0.005829  0.008123\n",
+       "3 -0.027782 -0.007257\n",
+       "4 -0.029530 -0.002555"
       ]
      },
-     "execution_count": 23,
+     "execution_count": 8,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -1896,7 +1838,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 9,
    "metadata": {},
    "outputs": [
     {
@@ -1942,68 +1884,68 @@
        "  <tbody>\n",
        "    <tr>\n",
        "      <th>0</th>\n",
-       "      <td>0.085299</td>\n",
+       "      <td>0.063504</td>\n",
        "      <td>0.050680</td>\n",
-       "      <td>0.044451</td>\n",
-       "      <td>-0.005670</td>\n",
-       "      <td>-0.034194</td>\n",
-       "      <td>-0.032356</td>\n",
-       "      <td>0.002861</td>\n",
-       "      <td>-0.025930</td>\n",
-       "      <td>0.020909</td>\n",
-       "      <td>0.036174</td>\n",
+       "      <td>-0.001895</td>\n",
+       "      <td>0.066629</td>\n",
+       "      <td>0.108914</td>\n",
+       "      <td>0.022869</td>\n",
+       "      <td>-0.035816</td>\n",
+       "      <td>0.003064</td>\n",
+       "      <td>0.015232</td>\n",
+       "      <td>0.013791</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>1</th>\n",
-       "      <td>0.005383</td>\n",
+       "      <td>-0.070900</td>\n",
        "      <td>-0.044642</td>\n",
-       "      <td>-0.036385</td>\n",
-       "      <td>0.021872</td>\n",
-       "      <td>0.015596</td>\n",
-       "      <td>0.008142</td>\n",
-       "      <td>-0.031988</td>\n",
-       "      <td>-0.046641</td>\n",
-       "      <td>0.004626</td>\n",
-       "      <td>-0.021702</td>\n",
+       "      <td>0.039062</td>\n",
+       "      <td>-0.033213</td>\n",
+       "      <td>-0.034508</td>\n",
+       "      <td>-0.024993</td>\n",
+       "      <td>0.067737</td>\n",
+       "      <td>-0.013504</td>\n",
+       "      <td>-0.015421</td>\n",
+       "      <td>-0.006742</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>2</th>\n",
-       "      <td>-0.045472</td>\n",
+       "      <td>0.005383</td>\n",
        "      <td>0.050680</td>\n",
-       "      <td>-0.047163</td>\n",
-       "      <td>-0.015999</td>\n",
-       "      <td>-0.024800</td>\n",
-       "      <td>0.000779</td>\n",
-       "      <td>-0.062917</td>\n",
-       "      <td>-0.038357</td>\n",
-       "      <td>-0.018766</td>\n",
-       "      <td>-0.004827</td>\n",
+       "      <td>-0.001895</td>\n",
+       "      <td>0.008101</td>\n",
+       "      <td>-0.015719</td>\n",
+       "      <td>-0.002903</td>\n",
+       "      <td>0.038394</td>\n",
+       "      <td>-0.013504</td>\n",
+       "      <td>-0.005829</td>\n",
+       "      <td>0.008123</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>3</th>\n",
-       "      <td>-0.096328</td>\n",
-       "      <td>-0.044642</td>\n",
-       "      <td>-0.083808</td>\n",
-       "      <td>0.008101</td>\n",
-       "      <td>-0.090561</td>\n",
-       "      <td>-0.013948</td>\n",
-       "      <td>-0.062917</td>\n",
-       "      <td>-0.034215</td>\n",
-       "      <td>-0.025080</td>\n",
-       "      <td>-0.045854</td>\n",
+       "      <td>-0.085430</td>\n",
+       "      <td>0.050680</td>\n",
+       "      <td>-0.022373</td>\n",
+       "      <td>0.001215</td>\n",
+       "      <td>-0.026366</td>\n",
+       "      <td>0.015505</td>\n",
+       "      <td>-0.072133</td>\n",
+       "      <td>-0.017646</td>\n",
+       "      <td>-0.027782</td>\n",
+       "      <td>-0.007257</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>4</th>\n",
-       "      <td>0.027178</td>\n",
+       "      <td>-0.067268</td>\n",
        "      <td>0.050680</td>\n",
-       "      <td>0.017506</td>\n",
-       "      <td>-0.033213</td>\n",
-       "      <td>0.045972</td>\n",
-       "      <td>-0.065491</td>\n",
-       "      <td>-0.096435</td>\n",
-       "      <td>-0.059067</td>\n",
-       "      <td>0.000522</td>\n",
-       "      <td>0.021657</td>\n",
+       "      <td>-0.012673</td>\n",
+       "      <td>-0.040099</td>\n",
+       "      <td>0.004636</td>\n",
+       "      <td>-0.058127</td>\n",
+       "      <td>0.019196</td>\n",
+       "      <td>-0.034215</td>\n",
+       "      <td>-0.029530</td>\n",
+       "      <td>-0.002555</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
@@ -2011,14 +1953,14 @@
       ],
       "text/plain": [
        "        age       sex       bmi        bp        s2        s3        s5        s6        s1        s4\n",
-       "0  0.085299  0.050680  0.044451 -0.005670 -0.034194 -0.032356  0.002861 -0.025930  0.020909  0.036174\n",
-       "1  0.005383 -0.044642 -0.036385  0.021872  0.015596  0.008142 -0.031988 -0.046641  0.004626 -0.021702\n",
-       "2 -0.045472  0.050680 -0.047163 -0.015999 -0.024800  0.000779 -0.062917 -0.038357 -0.018766 -0.004827\n",
-       "3 -0.096328 -0.044642 -0.083808  0.008101 -0.090561 -0.013948 -0.062917 -0.034215 -0.025080 -0.045854\n",
-       "4  0.027178  0.050680  0.017506 -0.033213  0.045972 -0.065491 -0.096435 -0.059067  0.000522  0.021657"
+       "0  0.063504  0.050680 -0.001895  0.066629  0.108914  0.022869 -0.035816  0.003064  0.015232  0.013791\n",
+       "1 -0.070900 -0.044642  0.039062 -0.033213 -0.034508 -0.024993  0.067737 -0.013504 -0.015421 -0.006742\n",
+       "2  0.005383  0.050680 -0.001895  0.008101 -0.015719 -0.002903  0.038394 -0.013504 -0.005829  0.008123\n",
+       "3 -0.085430  0.050680 -0.022373  0.001215 -0.026366  0.015505 -0.072133 -0.017646 -0.027782 -0.007257\n",
+       "4 -0.067268  0.050680 -0.012673 -0.040099  0.004636 -0.058127  0.019196 -0.034215 -0.029530 -0.002555"
       ]
      },
-     "execution_count": 24,
+     "execution_count": 9,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -2040,7 +1982,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 10,
    "metadata": {},
    "outputs": [
     {
@@ -2194,138 +2136,138 @@
           132
          ],
          "y": [
-          -0.04559945128264711,
-          0.003934851612593237,
-          -0.04009563984984263,
-          -0.10338947132709418,
-          -0.007072771253015731,
+          0.09061988167926385,
+          -0.012576582685820214,
           -0.004320865536613489,
-          0.017694380194604446,
-          0.024574144485610048,
-          0.0342058144930179,
-          -0.0029449126784123676,
-          -0.0029449126784123676,
-          -0.04284754556624487,
+          -0.037343734133440394,
+          -0.015328488402222454,
+          -0.04009563984984263,
           -0.038719686991641515,
-          -0.04422349842444599,
-          -0.0249601584096303,
-          -0.04972730985725048,
-          0.020446285911006685,
+          -0.0579830270064572,
+          0.0025588987543921156,
+          -0.046975404140848234,
           0.0025588987543921156,
-          -0.04147159270804375,
           0.01219056876179996,
-          -0.0318399227006359,
-          0.06447677737344255,
+          0.020446285911006685,
+          -0.007072771253015731,
+          -0.08962994274508297,
+          0.01219056876179996,
           -0.04422349842444599,
-          -0.011200629827619093,
-          -0.02083229983502694,
+          0.005310804470794357,
           -0.005696818394814609,
-          -0.033215875558837024,
           0.03833367306762126,
+          -0.011200629827619093,
           -0.12678066991651324,
-          -0.06761469701386505,
-          0.0025588987543921156,
           0.010814615903598841,
           -0.05935897986465832,
+          -0.051103262715451604,
+          0.06310082451524143,
           -0.0318399227006359,
           -0.030463969842434782,
-          -0.00019300696201012598,
-          0.06034891879883919,
-          0.001182945896190995,
-          -0.019456346976825818,
-          0.0080627101871966,
+          0.005310804470794357,
+          0.02182223876920781,
+          0.041085578784023497,
+          -0.04559945128264711,
+          0.014942474478202204,
+          0.06447677737344255,
+          0.02319819162740893,
           0.04383748450042574,
-          -0.033215875558837024,
-          -0.0029449126784123676,
-          -0.019456346976825818,
-          0.01219056876179996,
+          0.03833367306762126,
+          -0.005696818394814609,
+          0.00943866304539772,
           -0.015328488402222454,
-          -0.0318399227006359,
-          -0.030463969842434782,
+          -0.034591828417038145,
+          0.0342058144930179,
           0.020446285911006685,
           0.014942474478202204,
-          -0.02083229983502694,
-          0.027326050202012293,
-          0.02182223876920781,
-          -0.09650970703608859,
-          0.02319819162740893,
-          -0.07587041416307178,
-          -0.07999827273767514,
-          -0.051103262715451604,
+          -0.008448724111216851,
+          0.039709625925822375,
+          0.001182945896190995,
+          -0.037343734133440394,
+          0.03833367306762126,
+          -0.0318399227006359,
+          -0.011200629827619093,
+          -0.019456346976825818,
+          0.039709625925822375,
+          0.0741084473808504,
+          0.06998058880624704,
           0.03282986163481677,
           0.05209320164963247,
-          0.01219056876179996,
+          0.126394655992493,
+          -0.027712064126032544,
+          0.024574144485610048,
           0.017694380194604446,
+          -0.051103262715451604,
+          0.04934129593323023,
+          -0.001568959820211247,
+          0.04658939021682799,
+          -0.005696818394814609,
+          -0.04835135699904936,
+          -0.015328488402222454,
+          -0.04009563984984263,
           -0.0249601584096303,
-          0.053469154507833586,
-          -0.05385516843185383,
           -0.00019300696201012598,
           0.15391371315651542,
-          0.039709625925822375,
-          -0.02358420555142918,
-          0.07961225881365488,
+          0.006686757328995478,
+          -0.05935897986465832,
           -0.035967781275239266,
-          -0.04972730985725048,
-          0.0025588987543921156,
-          0.04934129593323023,
-          -0.007072771253015731,
-          -0.013952535544021335,
-          0.04246153164222462,
-          -0.04284754556624487,
-          -0.02220825269322806,
-          0.12501870313429186,
-          -0.05385516843185383,
-          0.053469154507833586,
-          0.06172487165704031,
-          -0.04972730985725048,
-          -0.016704441260423575,
+          0.1277706088506941,
+          0.006686757328995478,
+          0.03558176735121902,
+          0.016318427336403322,
+          0.020446285911006685,
+          0.0741084473808504,
           -0.026336111267831423,
+          0.08786797596286161,
+          -0.04422349842444599,
+          -0.06623874415566393,
+          0.04383748450042574,
+          -0.04972730985725048,
+          0.020446285911006685,
+          -0.005696818394814609,
           0.07686035309725264,
-          -0.07311850844666953,
-          -0.009824676969417972,
-          0.030077955918414535,
+          -0.008448724111216851,
+          0.03282986163481677,
+          0.03145390877661565,
           0.05484510736603471,
-          0.07823630595545376,
-          -0.007072771253015731,
-          -0.027712064126032544,
+          0.024574144485610048,
+          0.06034891879883919,
+          0.07273249452264928,
+          0.024574144485610048,
+          -0.029088016984233665,
+          -0.0029449126784123676,
+          -0.011200629827619093,
           0.001182945896190995,
-          0.09199583453746497,
-          0.07548440023905152,
           -0.007072771253015731,
-          0.1277706088506941,
-          -0.012576582685820214,
+          0.04246153164222462,
+          -0.004320865536613489,
           0.10988322169407955,
-          0.0080627101871966,
-          -0.013952535544021335,
-          -0.04284754556624487,
-          -0.046975404140848234,
+          -0.037343734133440394,
+          -0.016704441260423575,
+          0.006686757328995478,
           0.00943866304539772,
           0.041085578784023497,
-          0.030077955918414535,
-          0.02182223876920781,
+          -0.001568959820211247,
           -0.06623874415566393,
-          0.11951489170148738,
+          0.03558176735121902,
+          0.017694380194604446,
+          0.02182223876920781,
+          0.041085578784023497,
+          -0.034591828417038145,
+          -0.037343734133440394,
+          0.03145390877661565,
+          0.05622106022423583,
           0.014942474478202204,
-          -0.005696818394814609,
-          -0.060734932722859444,
-          -0.02358420555142918,
-          -0.06348683843926169,
-          -0.04972730985725048,
-          -0.011200629827619093,
-          -0.046975404140848234,
-          0.016318427336403322,
-          -0.018080394118624697,
           0.010814615903598841,
-          -0.0029449126784123676,
-          -0.06623874415566393,
+          0.10988322169407955,
+          0.09887559882847057,
+          0.001182945896190995,
           0.07823630595545376,
-          0.03833367306762126,
           -0.10063756561069194,
-          0.03558176735121902,
           0.04658939021682799,
+          0.02182223876920781,
           0.05759701308243695,
-          -0.005696818394814609,
-          -0.037343734133440394,
+          0.04658939021682799,
           0.08374011738825825
          ]
         },
@@ -2473,139 +2415,139 @@
           132
          ],
          "y": [
-          0.02090924803691378,
-          0.004626123494339312,
-          -0.018766371041145515,
-          -0.025079773212822724,
-          0.0005221962276462163,
-          0.0023782190339952007,
-          0.008649160755021803,
-          0.014061356781762173,
-          0.0254923032629858,
-          -0.011912383665013354,
-          -0.018407832152535906,
-          0.024214427396466377,
-          -0.02256780214714429,
-          0.004919903247914598,
-          -0.008170622618974065,
-          -0.018548330310735238,
-          -0.007105185961827832,
-          0.012008661347633834,
-          0.018618084713260215,
-          0.011580347815175038,
-          -0.007449829402368672,
-          0.01264639252518127,
-          -0.019867269093045897,
-          -0.015250402274648698,
-          -0.010913040915993408,
-          0.003691389849940784,
-          -0.0024928165805605546,
-          -0.0007776304487895633,
-          0.006325872630235647,
-          -0.005115114618450097,
-          -0.028534716572419568,
-          -0.005768925548395457,
-          -0.013047736065762188,
-          -0.028820647738861192,
-          -0.0035482860212044933,
-          -0.018530818319814676,
-          0.004165370131410912,
-          0.006749368070504823,
-          0.0019226569832983451,
-          0.02264996119206698,
-          0.020792167630816705,
-          -0.0038184998660977126,
-          0.011053015302265596,
-          -0.037005112578249275,
-          -0.01585513668744488,
-          0.02560374181865984,
-          0.02982157455613024,
-          0.016710154185276348,
-          0.006970480052674291,
-          -0.004867785078419127,
-          -0.021188768112257218,
-          0.015228956794648838,
-          0.031766761148240404,
-          -0.004424832962744514,
-          0.012040504341003278,
-          -0.021868959740455263,
-          -0.014030859836463884,
-          -0.012065482400146238,
-          -0.012421658895909645,
-          0.008993450695372609,
-          0.014045154218359003,
-          0.026886444708028364,
-          0.03098052571069148,
-          0.013277525924978313,
-          -0.0016929388165091135,
-          0.02107173313979645,
-          -0.013609680442188583,
-          -0.0005291128611427387,
-          0.012633195111147563,
-          0.01585177092183963,
-          0.01592201625116141,
-          -0.013505588816878359,
-          -0.029989934542940452,
-          0.024035107762849686,
-          -0.0036208937433724725,
-          0.00986821795135186,
-          -0.0025563066174634427,
-          0.015401158704997121,
-          -0.0017121976478600984,
-          0.0191111920844166,
-          -0.010060858295794301,
-          -0.020294974572644696,
-          0.03240917262316084,
-          -0.002681574554502042,
-          0.00011636986553689114,
-          0.005574322230719957,
-          0.013739407286284734,
-          -0.0066654621851517035,
-          -0.001037285439235573,
-          0.028032372378139492,
-          0.04316051018347186,
-          0.012379703775175534,
-          0.002297693317367038,
-          -0.0097459261449615,
-          -0.002440354209744388,
-          -0.00032389027157234823,
-          -0.00040304651144529,
-          0.023079074995717545,
-          0.0069802916130410056,
-          -0.008488594622839591,
-          0.0011337923289514538,
-          -0.00845554673950999,
-          0.012251890585047658,
-          0.016676498249086746,
-          -0.01474459352050577,
-          0.0050141833128581155,
-          -0.0021218581167275433,
-          0.014514906599180545,
-          -0.0023141889635125527,
-          -0.01592533089818831,
-          0.007626497719330275,
-          0.007189881329180314,
-          0.0014891685535230936,
-          0.019136067420616056,
-          0.015459647789467304,
-          0.010780314394988014,
-          -0.019002304466773874,
-          -0.017332340171545398,
-          0.009044765347958866,
-          0.022799569535975638,
-          0.0174367303520582,
-          -0.013712887832598355,
-          0.021126195472609306,
-          -0.015975502568585168,
-          -0.0028846246835958628,
-          0.0020442019813950094,
-          -0.03266377372798207,
-          0.03286550024138621,
-          -0.013566458567642754,
-          0.008925403172101965,
-          0.02019472487087253,
-          0.009589831559328486,
-          -0.028183039372319
+          0.015231685126095648,
+          -0.015421076522561732,
+          -0.005828944561737012,
+          -0.027782168227463364,
+          -0.0295299254606757,
+          -0.005044688697438994,
+          -0.027131385829086006,
+          -0.014967919098312174,
+          -0.018483150909256105,
+          -0.01901799646632892,
+          0.0019843111862389486,
+          0.006493388405783137,
+          -0.006212982280698071,
+          -0.02150559103893539,
+          -0.02260473952565378,
+          0.0049281533846481825,
+          -0.02442570645715391,
+          -0.01064156355108878,
+          -0.004259935401689829,
+          -0.008386295263997364,
+          -0.018015115574451734,
+          0.0007205377946915463,
+          -0.013088402487870042,
+          -0.014816805743643826,
+          -0.03829371114792571,
+          0.00926588506800341,
+          -0.03464258186869676,
+          -0.008082039196945013,
+          0.011364869658653243,
+          0.0009835337264478081,
+          0.011194546620714103,
+          -0.03700414327288421,
+          0.008013545770845095,
+          -0.014911745323678609,
+          -0.016557630340603,
+          0.015794064646605606,
+          -0.003318999321802166,
+          -0.023694428109623075,
+          0.0037830236465225455,
+          0.022283588751957895,
+          0.020259778298767666,
+          -0.007782211098720479,
+          0.001633485541469888,
+          -0.00336730837946193,
+          -0.00196576220854981,
+          -0.0037485742600287713,
+          -0.023103696888923084,
+          -0.02538158359937403,
+          0.010877528444466239,
+          0.003396870126202189,
+          -0.00931576420516173,
+          -0.02393598933865847,
+          -0.022060080451219353,
+          0.007751428520500037,
+          -0.017131675982750533,
+          -0.016893899230157834,
+          0.0016655963268504305,
+          0.014312087552817005,
+          0.01972723784114364,
+          -0.005721194774450792,
+          0.022818103401221923,
+          0.008053488228865244,
+          0.018756167519343916,
+          -0.013320754164637287,
+          0.034324781399503214,
+          -0.024136787863848923,
+          -0.009247336090314272,
+          -0.01918941549182599,
+          0.010341943198896466,
+          -0.038897981926843614,
+          0.010881436382772332,
+          -0.01566089827165408,
+          0.020615764532841765,
+          0.013009030249356294,
+          0.0073269433667188175,
+          0.006450426551069444,
+          -0.002280843953683123,
+          0.020357206907443796,
+          0.020805411773395996,
+          0.02742199422986559,
+          0.011743537467864502,
+          -0.0027340598135190738,
+          0.011738907246611663,
+          -0.016267946667548862,
+          0.002143508330794054,
+          0.007558658619120678,
+          -0.007106225052230407,
+          -0.0011209748205877786,
+          -0.019965748788626093,
+          0.003937343011525476,
+          -0.0213359342001851,
+          0.01935601032943401,
+          -0.010330455512844257,
+          0.03782723369570142,
+          0.0146552688808932,
+          0.013230530856790384,
+          -0.004987096013140689,
+          0.0333764888489525,
+          -0.00968681496343167,
+          0.004587070452381084,
+          0.005656357169409813,
+          -0.023245933365898277,
+          0.015162951494691587,
+          0.0026953573522491715,
+          0.0017848887739108383,
+          -0.00334948187707321,
+          -0.021721428589336345,
+          -0.010634357965284093,
+          0.0055700317473139855,
+          -0.0009479228487899657,
+          -0.0020297213245332613,
+          -0.013546639944096802,
+          -0.018097376394800455,
+          0.01808774523221887,
+          0.012340078071508173,
+          -0.020561983875713168,
+          -0.0017113426742070572,
+          -0.01261277549337654,
+          -0.03618135352270031,
+          0.02271639240732281,
+          -0.010989417897220211,
+          0.03516725603493444,
+          -0.015797770140132998,
+          0.00854976213952673,
+          -0.0064595754436547435,
+          -0.0008266335032858316,
+          -0.005144033539497858,
+          -0.03487386294915826,
+          -0.017215007399405,
+          -0.01813056307774026,
+          0.007110999760004685,
+          -0.01887035231907848,
+          -0.032472689994759454
          ]
         }
        ],
@@ -3602,138 +3544,138 @@
           132
          ],
          "y": [
+          0.01770335448356722,
           -0.002592261998183278,
           -0.002592261998183278,
           -0.03949338287409329,
+          0.03430885887772673,
+          -0.03949338287409329,
           -0.0763945037500033,
-          0.07120997975363674,
           -0.03949338287409329,
+          -0.002592261998183278,
           -0.03949338287409329,
           -0.03949338287409329,
-          0.03430885887772673,
+          -0.002592261998183278,
           -0.03949338287409329,
           -0.03949338287409329,
           -0.0763945037500033,
           -0.0763945037500033,
-          -0.0763945037500033,
-          -0.03949338287409329,
-          0.01585829843977173,
-          -0.0763945037500033,
-          -0.03949338287409329,
-          -0.002592261998183278,
-          -0.002592261998183278,
-          0.0029429061332032365,
-          -0.002592261998183278,
           0.07120997975363674,
-          -0.03949338287409329,
           0.07120997975363674,
           -0.0018542395806650938,
-          -0.002592261998183278,
           0.03430885887772673,
-          -0.047980640675552584,
-          -0.002592261998183278,
           -0.002592261998183278,
+          -0.047980640675552584,
           0.03430885887772673,
           -0.03949338287409329,
+          -0.002592261998183278,
+          -0.03949338287409329,
           -0.03949338287409329,
           -0.0763945037500033,
           -0.03949338287409329,
-          -0.002592261998183278,
-          0.05017634085436802,
+          0.03430885887772673,
+          0.03430885887772673,
           -0.03949338287409329,
           0.03430885887772673,
+          0.10811110062954676,
+          0.03430885887772673,
           0.07120997975363674,
-          -0.002592261998183278,
-          -0.002592261998183278,
           -0.03949338287409329,
           -0.03949338287409329,
+          0.03430885887772673,
           -0.002592261998183278,
           -0.002592261998183278,
-          -0.002592261998183278,
+          0.13025177315509276,
           0.07120997975363674,
           0.02545258986750832,
-          0.07120997975363674,
           -0.03949338287409329,
-          0.03910600459159503,
-          -0.03949338287409329,
-          0.08006624876385515,
-          -0.0763945037500033,
+          0.10811110062954676,
+          -0.007020396503292483,
           -0.03949338287409329,
+          0.03430885887772673,
           -0.03949338287409329,
+          0.03430885887772673,
+          0.03430885887772673,
+          0.10811110062954676,
+          -0.002592261998183278,
+          0.07120997975363674,
           -0.002592261998183278,
           -0.021411833644897377,
+          0.03430885887772673,
+          0.03430885887772673,
+          0.03430885887772673,
+          0.03430885887772673,
+          0.03430885887772673,
           -0.03949338287409329,
+          -0.03949338287409329,
+          -0.002592261998183278,
           0.03430885887772673,
           -0.03949338287409329,
-          0.14501222150545676,
+          -0.002592261998183278,
+          0.03430885887772673,
           -0.03949338287409329,
           -0.05056371913686628,
           0.07194800217115493,
-          0.056080520194513636,
-          -0.03949338287409329,
+          -0.012555564634678981,
+          0.012906208769698923,
+          0.07120997975363674,
+          0.10811110062954676,
+          0.08080427118137334,
+          -0.0763945037500033,
           0.03430885887772673,
           0.07120997975363674,
-          -0.03949338287409329,
-          -0.03949338287409329,
           0.07120997975363674,
           -0.03949338287409329,
-          0.005156973385757823,
-          -0.0763945037500033,
-          -0.002592261998183278,
-          -0.002592261998183278,
+          0.03430885887772673,
+          -0.03949338287409329,
           -0.002592261998183278,
-          -0.0763945037500033,
-          0.07120997975363674,
           -0.002592261998183278,
           -0.0708593356186168,
-          -0.002592261998183278,
+          0.0003598276718895252,
           -0.03949338287409329,
           0.07120997975363674,
-          -0.06938329078358041,
-          -0.002592261998183278,
+          0.07120997975363674,
+          0.03430885887772673,
           -0.03949338287409329,
           0.14132210941786577,
+          0.15534453535071155,
           0.10811110062954676,
-          0.03430885887772673,
-          -0.03949338287409329,
-          -0.03949338287409329,
-          0.14501222150545676,
+          0.08486339447772344,
           -0.002592261998183278,
           -0.002592261998183278,
           0.07120997975363674,
+          0.03430885887772673,
+          -0.002592261998183278,
           -0.002592261998183278,
           0.03430885887772673,
+          -0.03949338287409329,
+          0.03430885887772673,
           -0.002592261998183278,
           -0.002592261998183278,
-          -0.0763945037500033,
-          0.028404679537581124,
+          0.03430885887772673,
           -0.002592261998183278,
           0.07120997975363674,
-          0.03430885887772673,
-          0.03430885887772673,
           -0.03949338287409329,
-          0.08670845052151895,
-          0.07120997975363674,
+          -0.03949338287409329,
+          -0.002592261998183278,
           0.03430885887772673,
-          -0.03395821474270679,
           -0.002592261998183278,
-          -0.03949338287409329,
           -0.002592261998183278,
+          -0.0763945037500033,
           -0.03949338287409329,
-          -0.02583996815000658,
+          0.02360753382371283,
+          0.07120997975363674,
           -0.002592261998183278,
           -0.03949338287409329,
-          -0.03949338287409329,
-          0.003311917341962329,
-          -0.03949338287409329,
+          0.09187460744414634,
           -0.002592261998183278,
-          0.10811110062954676,
-          -0.0763945037500033,
           0.03430885887772673,
+          -0.002592261998183278,
+          -0.0763945037500033,
           -0.024732934523729287,
+          0.03430885887772673,
           0.023238522614953735,
-          -0.002592261998183278,
-          -0.011079519799642579,
+          -0.03949338287409329,
           -0.03949338287409329
          ]
         },
@@ -3881,139 +3823,139 @@
           132
          ],
          "y": [
-          0.036173881319049764,
-          -0.021702339029765656,
-          -0.004827344613356406,
-          -0.04585371324963408,
-          0.021657333531173127,
-          -0.019557492599216835,
-          -0.016391484441786622,
-          0.004826645265102516,
-          0.026677561541063308,
-          -0.020494778395885635,
-          -0.019045046146368566,
-          0.020546868283924196,
-          -0.03767286475161845,
-          -0.025381010739657484,
-          -0.010055873741307193,
-          0.006350417632745559,
-          -0.019896188609062136,
-          0.006649615916558177,
-          0.02621139123628575,
-          -0.014090746522772122,
-          -0.001262154321621097,
-          0.011773637398311868,
-          0.011258135449664754,
-          0.008944049563999953,
-          0.006999038596481625,
-          0.012591280717550547,
-          0.006376848293006614,
-          0.009152159296335255,
-          0.024846242702600232,
-          -0.030649754557607656,
-          -0.03226696508115764,
-          0.007357869247591211,
-          -0.012233157667270823,
-          -0.005235643701531674,
-          -0.019963033597623834,
-          -0.02294552944805546,
-          -0.004610187191359073,
-          0.014874207825065582,
-          -0.02541069014530403,
-          0.0007426496564170359,
-          0.04671645093525994,
-          0.011725793349133122,
-          0.046377763208378976,
-          -0.04412613967921996,
-          -0.003307935367806759,
-          0.010059493995214983,
-          0.057837475697029336,
-          0.02464075426683707,
-          0.03232886503700636,
-          0.015181239790241672,
-          0.007730755650970042,
-          -0.01288826786920192,
-          0.019209494823279768,
-          0.017901355466069357,
-          0.0010382296741955493,
-          -0.03525362630963111,
-          -0.0010484738172200827,
-          -0.034411594412093106,
-          -0.029311390546604275,
-          -0.026183464604745584,
-          0.006211365730593833,
-          0.007955487839617569,
-          0.009444435976199097,
-          0.029891333441255042,
-          0.006322089058982078,
-          0.01240042724866708,
-          0.03372764125476694,
-          0.01563560428520054,
-          0.004682485638364525,
-          -0.005096275037787555,
-          0.016834842476349577,
-          0.02777791931447148,
-          -0.04590318168669977,
-          0.039194821530438365,
-          -0.03277665344423763,
-          -0.02034178179038435,
-          -0.021864092307613023,
-          0.022562752824849954,
-          0.011182533788756978,
-          0.0013496505057414734,
-          -0.03312248173854206,
-          0.026222471799173384,
-          0.0035463695703147743,
-          -0.018847397811071878,
-          0.006249373998074355,
-          -0.027683935756017426,
-          0.00739858101231083,
-          -0.02736892056924086,
-          0.0003022669656364365,
-          -0.0007816308585437103,
-          0.01700147809620549,
-          -0.0061187072597464025,
-          0.014402384577943744,
-          -0.0046736485052005,
-          -0.02700606133328556,
-          0.014764184029720314,
-          -0.013134114176875222,
-          0.028808489012539,
-          0.02254331768624029,
-          0.005034238597544158,
-          0.02928319889004928,
-          0.0029064231086462057,
-          -0.001784491148696758,
-          -0.00893612903310391,
-          0.0030842824733004984,
-          -0.021545185329186856,
-          0.056749951384627524,
-          0.07189639275286117,
-          0.015165622525579,
-          -0.02023890492078882,
-          0.011598392971903482,
-          0.020595548628953693,
-          0.01685751810776878,
-          0.0014485791272968842,
-          0.012291320055589793,
-          0.012358662118871731,
-          -0.012061731114809418,
-          0.011744440395706587,
-          -0.009350081063623469,
-          0.02917756393068547,
-          0.006130601782959003,
-          -0.01613858615607142,
-          0.017689170808935114,
-          -0.032720269822326155,
-          0.03713214633579796,
-          0.003601191679832815,
-          -0.03355297405509526,
-          0.051788117379947894,
-          -0.02597487243898064,
-          0.008109666350814736,
-          0.0271125648455035,
-          0.013573789266825956,
-          -0.04185854704123922
+          0.013791087535585032,
+          -0.0067415479818104754,
+          0.008122595600109256,
+          -0.007257066606169001,
+          -0.0025554928421562782,
+          -0.01730642961557491,
+          -0.03973989461680715,
+          -0.035215263664104976,
+          0.012493718436498158,
+          -0.023285725079843646,
+          -0.000379727084753736,
+          -0.019743775483154393,
+          -0.009392833506987743,
+          -0.035302615781131336,
+          -0.03905225690190696,
+          -0.03256533884487734,
+          0.005676995172456719,
+          -0.003969962854634164,
+          0.006605608754211003,
+          0.0022337245636956634,
+          -0.003705667926297255,
+          0.0160131771773543,
+          0.0012497484695890783,
+          -0.014183640472615826,
+          -0.024897592873238887,
+          -0.01193835158341511,
+          -0.011352501675463188,
+          -0.02211270494407038,
+          -0.01327963980677072,
+          0.014679377827269066,
+          0.028201762472852287,
+          -0.0497156513407594,
+          0.015564039073432956,
+          0.0059047509621594245,
+          -0.0002891958481750804,
+          0.0413585013037903,
+          0.010549489171225434,
+          -0.04074701058422851,
+          -0.00726694018568288,
+          0.0049352105579747434,
+          0.010367920997501279,
+          0.00007617849551178838,
+          0.0272796380000965,
+          0.013649300291033686,
+          -0.015212767758098487,
+          0.00543914128208987,
+          -0.03867304908811106,
+          -0.04146073514129962,
+          0.037180730183873835,
+          -0.021689236815534446,
+          0.006731851807441467,
+          -0.004709125236394509,
+          0.015675703756764758,
+          -0.014059228838315884,
+          0.006718223688295505,
+          -0.0325135193563347,
+          -0.029345864135223824,
+          0.016299251273056356,
+          0.020736592779718496,
+          0.01752768688371003,
+          0.0036211377320435026,
+          0.00582584141816508,
+          -0.006378574065545362,
+          -0.03888936237339751,
+          0.0046740506114529325,
+          -0.04102493363171138,
+          -0.0036541432594371158,
+          -0.031746617901640996,
+          0.02287841546841343,
+          -0.017455500382433535,
+          0.004170520880805757,
+          0.02950946182329947,
+          0.004286409933083225,
+          0.00949468220189915,
+          0.00848214300974643,
+          -0.0003506567655354102,
+          0.00622718589199493,
+          -0.023382695759762345,
+          0.052460227354301635,
+          0.04570683846266361,
+          0.008142368340912764,
+          0.003042472748240509,
+          -0.009428181419382162,
+          -0.03414234589319587,
+          0.004322787221553486,
+          0.008760128728215626,
+          -0.021232834359493004,
+          0.012613683896552527,
+          0.016963752411511428,
+          -0.0011747844791662968,
+          -0.006854960778212722,
+          0.03271464639099751,
+          -0.020651434806323598,
+          0.013655494578349706,
+          0.0325822587161095,
+          0.01240140962718438,
+          0.004465601030109772,
+          0.033314444821553636,
+          0.005823579974981913,
+          0.02011640909906851,
+          0.006129851307818842,
+          -0.03152688781818483,
+          0.02140947382021927,
+          0.014218926009913056,
+          -0.018311539156434627,
+          0.02234257426494738,
+          -0.001616018921944109,
+          0.0205886244053183,
+          0.013896438117549436,
+          -0.024496748572054607,
+          0.05235423105172521,
+          -0.032992757155550216,
+          -0.024303522197554835,
+          0.01212643354855765,
+          0.03392301443916518,
+          -0.006049284904496647,
+          -0.007631149465924967,
+          -0.028761809138011133,
+          -0.04800248232671432,
+          -0.02509379567625947,
+          -0.00009153975376653429,
+          0.023345773261553425,
+          -0.01910895717320799,
+          0.017869477687563542,
+          -0.02194514854495601,
+          -0.024607711275376646,
+          0.029117388128151366,
+          -0.035875224977058126,
+          -0.029496018124353707,
+          -0.03515240882947335,
+          0.006314919004553546,
+          -0.03774485237314226,
+          -0.044343789309302856
          ]
         }
        ],
@@ -4919,13 +4861,13 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 11,
    "metadata": {},
    "outputs": [
     {
      "data": {
       "application/vnd.jupyter.widget-view+json": {
-       "model_id": "5dfb811d4a2b457c80f89229f35c4d88",
+       "model_id": "4798c8e161a24d56ac7449decf7a734f",
        "version_major": 2,
        "version_minor": 0
       },
@@ -4941,11 +4883,12 @@
      "output_type": "stream",
      "text": [
       "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 8 concurrent workers.\n",
-      "[Parallel(n_jobs=-1)]: Done   3 out of   3 | elapsed:    5.7s finished\n",
+      "[Parallel(n_jobs=-1)]: Done   3 out of   3 | elapsed:    1.6s finished\n",
       "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 8 concurrent workers.\n",
-      "[Parallel(n_jobs=-1)]: Done   3 out of   3 | elapsed:    2.9s finished\n",
-      "[Parallel(n_jobs=1)]: Done   1 tasks      | elapsed:    4.0s\n",
-      "[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:   10.3s finished\n"
+      "[Parallel(n_jobs=-1)]: Batch computation too fast (0.1265852451324463s.) Setting batch_size=2.\n",
+      "[Parallel(n_jobs=-1)]: Done   3 out of   3 | elapsed:    0.1s finished\n",
+      "[Parallel(n_jobs=1)]: Done   1 tasks      | elapsed:    0.1s\n",
+      "[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:    0.1s finished\n"
      ]
     },
     {
@@ -4976,28 +4919,28 @@
        "  <tbody>\n",
        "    <tr>\n",
        "      <th>0</th>\n",
-       "      <td>0.020909</td>\n",
-       "      <td>0.036174</td>\n",
+       "      <td>0.015232</td>\n",
+       "      <td>0.013791</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>1</th>\n",
-       "      <td>0.004626</td>\n",
-       "      <td>-0.021702</td>\n",
+       "      <td>-0.015421</td>\n",
+       "      <td>-0.006742</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>2</th>\n",
-       "      <td>-0.018766</td>\n",
-       "      <td>-0.004827</td>\n",
+       "      <td>-0.005829</td>\n",
+       "      <td>0.008123</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>3</th>\n",
-       "      <td>-0.025080</td>\n",
-       "      <td>-0.045854</td>\n",
+       "      <td>-0.027782</td>\n",
+       "      <td>-0.007257</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>4</th>\n",
-       "      <td>0.000522</td>\n",
-       "      <td>0.021657</td>\n",
+       "      <td>-0.029530</td>\n",
+       "      <td>-0.002555</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>...</th>\n",
@@ -5006,28 +4949,28 @@
        "    </tr>\n",
        "    <tr>\n",
        "      <th>128</th>\n",
-       "      <td>-0.013566</td>\n",
-       "      <td>-0.025975</td>\n",
+       "      <td>-0.017215</td>\n",
+       "      <td>-0.029496</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>129</th>\n",
-       "      <td>0.008925</td>\n",
-       "      <td>0.008110</td>\n",
+       "      <td>-0.018131</td>\n",
+       "      <td>-0.035152</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>130</th>\n",
-       "      <td>0.020195</td>\n",
-       "      <td>0.027113</td>\n",
+       "      <td>0.007111</td>\n",
+       "      <td>0.006315</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>131</th>\n",
-       "      <td>0.009590</td>\n",
-       "      <td>0.013574</td>\n",
+       "      <td>-0.018870</td>\n",
+       "      <td>-0.037745</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>132</th>\n",
-       "      <td>-0.028183</td>\n",
-       "      <td>-0.041859</td>\n",
+       "      <td>-0.032473</td>\n",
+       "      <td>-0.044344</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
@@ -5036,22 +4979,22 @@
       ],
       "text/plain": [
        "           s1        s4\n",
-       "0    0.020909  0.036174\n",
-       "1    0.004626 -0.021702\n",
-       "2   -0.018766 -0.004827\n",
-       "3   -0.025080 -0.045854\n",
-       "4    0.000522  0.021657\n",
+       "0    0.015232  0.013791\n",
+       "1   -0.015421 -0.006742\n",
+       "2   -0.005829  0.008123\n",
+       "3   -0.027782 -0.007257\n",
+       "4   -0.029530 -0.002555\n",
        "..        ...       ...\n",
-       "128 -0.013566 -0.025975\n",
-       "129  0.008925  0.008110\n",
-       "130  0.020195  0.027113\n",
-       "131  0.009590  0.013574\n",
-       "132 -0.028183 -0.041859\n",
+       "128 -0.017215 -0.029496\n",
+       "129 -0.018131 -0.035152\n",
+       "130  0.007111  0.006315\n",
+       "131 -0.018870 -0.037745\n",
+       "132 -0.032473 -0.044344\n",
        "\n",
        "[133 rows x 2 columns]"
       ]
      },
-     "execution_count": 26,
+     "execution_count": 11,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -5093,13 +5036,13 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 12,
    "metadata": {},
    "outputs": [
     {
      "data": {
       "application/vnd.jupyter.widget-view+json": {
-       "model_id": "99f2a175dec14d80be25b0042cf5fe39",
+       "model_id": "4e0b2ffc1515464dbc2e2f23be9f28ff",
        "version_major": 2,
        "version_minor": 0
       },
@@ -5115,20 +5058,21 @@
      "output_type": "stream",
      "text": [
       "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 8 concurrent workers.\n",
-      "[Parallel(n_jobs=-1)]: Done   3 out of   3 | elapsed:    1.2s finished\n",
+      "[Parallel(n_jobs=-1)]: Batch computation too fast (0.1340169906616211s.) Setting batch_size=2.\n",
+      "[Parallel(n_jobs=-1)]: Done   3 out of   3 | elapsed:    0.1s finished\n",
       "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 8 concurrent workers.\n",
-      "[Parallel(n_jobs=-1)]: Done   3 out of   3 | elapsed:    0.2s finished\n",
+      "[Parallel(n_jobs=-1)]: Done   3 out of   3 | elapsed:    0.4s finished\n",
       "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 8 concurrent workers.\n",
-      "[Parallel(n_jobs=-1)]: Done   3 out of   3 | elapsed:    1.0s finished\n",
-      "[Parallel(n_jobs=1)]: Done   1 tasks      | elapsed:    3.4s\n",
-      "[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:    9.3s finished\n"
+      "[Parallel(n_jobs=-1)]: Done   3 out of   3 | elapsed:    0.8s finished\n",
+      "[Parallel(n_jobs=1)]: Done   1 tasks      | elapsed:    0.0s\n",
+      "[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:    0.1s finished\n"
      ]
     },
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "Imputation results available for models: dict_keys(['best_method', 'QRF', 'QuantReg', 'Matching'])\n",
+      "Imputation results available for models: dict_keys(['best_method', 'QuantReg', 'QRF', 'Matching'])\n",
       "The best performing model is: OLSResults\n"
      ]
     }
@@ -5157,7 +5101,7 @@
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": "pe3.13",
+   "display_name": "pe3.13 (3.13.0)",
    "language": "python",
    "name": "python3"
   },
diff --git a/docs/autoimpute/index.md b/docs/autoimpute/index.md
index 1600464..d85a15c 100644
--- a/docs/autoimpute/index.md
+++ b/docs/autoimpute/index.md
@@ -1,5 +1,5 @@
 # Autoimpute
 
-This chapter describes how the `autoimpute` function works to automize the entire method comparison, selection, and imputation pipeline in a single function.
+The `autoimpute` function automates the entire method comparison, selection, and imputation pipeline in a single call.
 
-The pipeline begins with input validation to ensure all necessary columns exist and quantiles are properly specified. It then preprocesses the donor and receiver datasets to prepare them for model training and evaluation. The function supports imputing numerical, categorical and boolean variable types, internally selecting the method corresponding to each variable type. At its core, `autoimpute` employs cross-validation on the donor data to evaluate multiple imputation methods. Each model is assessed on its ability to accurately predict known values using two different metrics: quantile loss for numerical imputation and log loss for categorical imputation. The method with the lowest average loss (with different metrics combined with a weighted-rank approach) across target variables is automatically selected as the optimal approach for the specific dataset and imputation task. The chosen model is then trained on the complete donor dataset and applied to generate imputations for the missing values in the receiver data. Finally, the pipeline reintegrates these imputed values back into the original receiver dataset, producing a complete dataset ready for downstream analysis. 
+The pipeline begins with input validation, then preprocesses the donor and receiver datasets for model training and evaluation. It supports numerical, categorical, and boolean variable types, selecting the appropriate method for each. At its core, `autoimpute` runs cross-validation on the donor data to evaluate multiple imputation methods. Each model is scored using quantile loss for numerical variables and log loss for categorical variables. The method with the lowest average loss (combining different metrics via a weighted-rank approach) across target variables is selected automatically. That model is then trained on the full donor dataset and applied to generate imputations for the receiver. The result is an `AutoImputeResult` object containing the imputations, the augmented receiver dataset, fitted models, and cross-validation results.
diff --git a/docs/imputation-benchmarking/benchmarking-methods.ipynb b/docs/imputation-benchmarking/benchmarking-methods.ipynb
index 9f9c3aa..6881559 100644
--- a/docs/imputation-benchmarking/benchmarking-methods.ipynb
+++ b/docs/imputation-benchmarking/benchmarking-methods.ipynb
@@ -16,7 +16,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 2,
+   "execution_count": 1,
    "id": "45bc35df",
    "metadata": {},
    "outputs": [
@@ -29,10 +29,10 @@
        "data": [
         {
          "alignmentgroup": "True",
-         "hovertemplate": "Method=QRF<br>Quantiles=%{x}<br>Test Quantile loss=%{y}<extra></extra>",
+         "hovertemplate": "Method=QRF<br>Quantiles=%{x}<br>Quantile loss=%{y}<extra></extra>",
          "legendgroup": "QRF",
          "marker": {
-          "color": "#636EFA",
+          "color": "#88CCEE",
           "pattern": {
            "shape": ""
           }
@@ -166,10 +166,10 @@
         },
         {
          "alignmentgroup": "True",
-         "hovertemplate": "Method=OLS<br>Quantiles=%{x}<br>Test Quantile loss=%{y}<extra></extra>",
+         "hovertemplate": "Method=OLS<br>Quantiles=%{x}<br>Quantile loss=%{y}<extra></extra>",
          "legendgroup": "OLS",
          "marker": {
-          "color": "#EF553B",
+          "color": "#CC6677",
           "pattern": {
            "shape": ""
           }
@@ -303,10 +303,10 @@
         },
         {
          "alignmentgroup": "True",
-         "hovertemplate": "Method=QuantReg<br>Quantiles=%{x}<br>Test Quantile loss=%{y}<extra></extra>",
+         "hovertemplate": "Method=QuantReg<br>Quantiles=%{x}<br>Quantile loss=%{y}<extra></extra>",
          "legendgroup": "QuantReg",
          "marker": {
-          "color": "#00CC96",
+          "color": "#DDCC77",
           "pattern": {
            "shape": ""
           }
@@ -440,10 +440,10 @@
         },
         {
          "alignmentgroup": "True",
-         "hovertemplate": "Method=Matching<br>Quantiles=%{x}<br>Test Quantile loss=%{y}<extra></extra>",
+         "hovertemplate": "Method=Matching<br>Quantiles=%{x}<br>Quantile loss=%{y}<extra></extra>",
          "legendgroup": "Matching",
          "marker": {
-          "color": "#AB63FA",
+          "color": "#117733",
           "pattern": {
            "shape": ""
           }
@@ -585,12 +585,12 @@
          },
          "tracegroupgap": 0
         },
-        "paper_bgcolor": "#F0F0F0",
-        "plot_bgcolor": "#F0F0F0",
+        "paper_bgcolor": "#FAFAFA",
+        "plot_bgcolor": "#FAFAFA",
         "shapes": [
          {
           "line": {
-           "color": "#636EFA",
+           "color": "#88CCEE",
            "dash": "dot",
            "width": 2
           },
@@ -603,7 +603,7 @@
          },
          {
           "line": {
-           "color": "#EF553B",
+           "color": "#CC6677",
            "dash": "dot",
            "width": 2
           },
@@ -616,7 +616,7 @@
          },
          {
           "line": {
-           "color": "#00CC96",
+           "color": "#DDCC77",
            "dash": "dot",
            "width": 2
           },
@@ -629,7 +629,7 @@
          },
          {
           "line": {
-           "color": "#AB63FA",
+           "color": "#117733",
            "dash": "dot",
            "width": 2
           },
@@ -1470,7 +1470,11 @@
           0,
           1
          ],
+         "gridcolor": "#E5E5E5",
+         "gridwidth": 1,
+         "linecolor": "#CCCCCC",
          "showgrid": false,
+         "showline": true,
          "title": {
           "font": {
            "size": 12
@@ -1485,12 +1489,16 @@
           0,
           1
          ],
-         "showgrid": false,
+         "gridcolor": "#E5E5E5",
+         "gridwidth": 1,
+         "linecolor": "#CCCCCC",
+         "showgrid": true,
+         "showline": true,
          "title": {
           "font": {
            "size": 12
           },
-          "text": "Test Quantile loss"
+          "text": "Quantile loss"
          },
          "zeroline": false
         }
@@ -1516,7 +1524,7 @@
     {
      "data": {
       "application/vnd.jupyter.widget-view+json": {
-       "model_id": "757ba887bc594faba40db29be449fc8a",
+       "model_id": "e01f756472f74ff7b937e4cb30273e97",
        "version_major": 2,
        "version_minor": 0
       },
@@ -1530,7 +1538,7 @@
     {
      "data": {
       "application/vnd.jupyter.widget-view+json": {
-       "model_id": "c4ac70f725d34a4ab138535bfefbf164",
+       "model_id": "c2b8d34026e0466893d9871472ee7088",
        "version_major": 2,
        "version_minor": 0
       },
@@ -1685,10 +1693,10 @@
        "data": [
         {
          "alignmentgroup": "True",
-         "hovertemplate": "Method=QRF<br>Quantiles=%{x}<br>Test Quantile loss=%{y}<extra></extra>",
+         "hovertemplate": "Method=QRF<br>Quantiles=%{x}<br>Quantile loss=%{y}<extra></extra>",
          "legendgroup": "QRF",
          "marker": {
-          "color": "#636EFA",
+          "color": "#88CCEE",
           "pattern": {
            "shape": ""
           }
@@ -1784,10 +1792,10 @@
         },
         {
          "alignmentgroup": "True",
-         "hovertemplate": "Method=OLS<br>Quantiles=%{x}<br>Test Quantile loss=%{y}<extra></extra>",
+         "hovertemplate": "Method=OLS<br>Quantiles=%{x}<br>Quantile loss=%{y}<extra></extra>",
          "legendgroup": "OLS",
          "marker": {
-          "color": "#EF553B",
+          "color": "#CC6677",
           "pattern": {
            "shape": ""
           }
@@ -1883,10 +1891,10 @@
         },
         {
          "alignmentgroup": "True",
-         "hovertemplate": "Method=QuantReg<br>Quantiles=%{x}<br>Test Quantile loss=%{y}<extra></extra>",
+         "hovertemplate": "Method=QuantReg<br>Quantiles=%{x}<br>Quantile loss=%{y}<extra></extra>",
          "legendgroup": "QuantReg",
          "marker": {
-          "color": "#00CC96",
+          "color": "#DDCC77",
           "pattern": {
            "shape": ""
           }
@@ -1982,10 +1990,10 @@
         },
         {
          "alignmentgroup": "True",
-         "hovertemplate": "Method=Matching<br>Quantiles=%{x}<br>Test Quantile loss=%{y}<extra></extra>",
+         "hovertemplate": "Method=Matching<br>Quantiles=%{x}<br>Quantile loss=%{y}<extra></extra>",
          "legendgroup": "Matching",
          "marker": {
-          "color": "#AB63FA",
+          "color": "#117733",
           "pattern": {
            "shape": ""
           }
@@ -2038,44 +2046,44 @@
          ],
          "xaxis": "x",
          "y": [
-          12257049.472826088,
-          12257049.472826088,
-          11844892.22826087,
-          11844892.22826087,
-          11432734.983695652,
-          11432734.983695652,
-          11020577.739130436,
-          11020577.739130436,
-          10608420.494565217,
-          10608420.494565217,
-          10196263.25,
-          10196263.25,
-          9784106.005434783,
-          9784106.005434783,
-          9371948.760869564,
-          9371948.760869564,
-          8959791.516304348,
-          8959791.516304348,
-          8547634.27173913,
-          8547634.27173913,
-          8135477.027173913,
-          8135477.027173913,
-          7723319.782608695,
-          7723319.782608695,
-          7311162.5380434785,
-          7311162.5380434785,
-          6899005.293478261,
-          6899005.293478261,
-          6486848.048913044,
-          6486848.048913044,
-          6074690.804347826,
-          6074690.804347826,
-          5662533.559782608,
-          5662533.559782608,
-          5250376.315217392,
-          5250376.315217392,
-          4838219.070652175,
-          4838219.070652175
+          12257052.081521738,
+          12257052.081521738,
+          11844897.445652174,
+          11844897.445652174,
+          11432742.80978261,
+          11432742.80978261,
+          11020588.173913043,
+          11020588.173913043,
+          10608433.538043479,
+          10608433.538043479,
+          10196278.902173912,
+          10196278.902173912,
+          9784124.266304348,
+          9784124.266304348,
+          9371969.630434783,
+          9371969.630434783,
+          8959814.994565217,
+          8959814.994565217,
+          8547660.358695652,
+          8547660.358695652,
+          8135505.722826087,
+          8135505.722826087,
+          7723351.0869565215,
+          7723351.0869565215,
+          7311196.451086956,
+          7311196.451086956,
+          6899041.815217392,
+          6899041.815217392,
+          6486887.179347826,
+          6486887.179347826,
+          6074732.543478261,
+          6074732.543478261,
+          5662577.907608695,
+          5662577.907608695,
+          5250423.271739131,
+          5250423.271739131,
+          4838268.635869565,
+          4838268.635869565
          ],
          "yaxis": "y"
         }
@@ -2089,12 +2097,12 @@
          },
          "tracegroupgap": 0
         },
-        "paper_bgcolor": "#F0F0F0",
-        "plot_bgcolor": "#F0F0F0",
+        "paper_bgcolor": "#FAFAFA",
+        "plot_bgcolor": "#FAFAFA",
         "shapes": [
          {
           "line": {
-           "color": "#636EFA",
+           "color": "#88CCEE",
            "dash": "dot",
            "width": 2
           },
@@ -2107,7 +2115,7 @@
          },
          {
           "line": {
-           "color": "#EF553B",
+           "color": "#CC6677",
            "dash": "dot",
            "width": 2
           },
@@ -2120,7 +2128,7 @@
          },
          {
           "line": {
-           "color": "#00CC96",
+           "color": "#DDCC77",
            "dash": "dot",
            "width": 2
           },
@@ -2133,7 +2141,7 @@
          },
          {
           "line": {
-           "color": "#AB63FA",
+           "color": "#117733",
            "dash": "dot",
            "width": 2
           },
@@ -2141,8 +2149,8 @@
           "type": "line",
           "x0": -0.5,
           "x1": 18.5,
-          "y0": 8547634.27173913,
-          "y1": 8547634.27173913
+          "y0": 8547660.35869565,
+          "y1": 8547660.35869565
          }
         ],
         "template": {
@@ -2974,7 +2982,11 @@
           0,
           1
          ],
+         "gridcolor": "#E5E5E5",
+         "gridwidth": 1,
+         "linecolor": "#CCCCCC",
          "showgrid": false,
+         "showline": true,
          "title": {
           "font": {
            "size": 12
@@ -2989,12 +3001,16 @@
           0,
           1
          ],
-         "showgrid": false,
+         "gridcolor": "#E5E5E5",
+         "gridwidth": 1,
+         "linecolor": "#CCCCCC",
+         "showgrid": true,
+         "showline": true,
          "title": {
           "font": {
            "size": 12
           },
-          "text": "Test Quantile loss"
+          "text": "Quantile loss"
          },
          "zeroline": false
         }
@@ -3019,7 +3035,7 @@
       "0  networth  wasserstein_distance  2.818652e+07\n",
       "Model: Matching, distribution similarity: \n",
       "   Variable                Metric      Distance\n",
-      "0  networth  wasserstein_distance  1.798087e+07\n"
+      "0  networth  wasserstein_distance  1.798092e+07\n"
      ]
     }
    ],
@@ -3232,7 +3248,7 @@
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": "pe3.13",
+   "display_name": "pe3.13 (3.13.0)",
    "language": "python",
    "name": "python3"
   },
diff --git a/docs/imputation-benchmarking/cross-validation.md b/docs/imputation-benchmarking/cross-validation.md
index 5e440e8..8da745f 100644
--- a/docs/imputation-benchmarking/cross-validation.md
+++ b/docs/imputation-benchmarking/cross-validation.md
@@ -1,14 +1,11 @@
 # Cross-validation and model imputation comparison
 
-This page documents the cross-validation utilities for evaluating imputation model performance. Cross-validation provides robust 
-estimates of how well a model will generalize to unseen data by training and testing on multiple data splits. Functions like `get_imputations`, will then build upon it, to standardize evaluation for all models, ensuring possible through a consistent experimental setup.
+This page documents the cross-validation utilities for evaluating imputation model performance. Cross-validation estimates how well a model generalizes to unseen data by training and testing on multiple splits. The `get_imputations` function builds on this to standardize evaluation across all models.
 
-Microimpute's cross-validation automatically selects the appropriate metric based on variable type. Numerical variables are evaluated using quantile loss, which measures prediction accuracy across the conditional distribution. Categorical variables are evaluated using log loss (cross-entropy), which penalizes confident but incorrect predictions, see the [Metrics page](./metrics.md) for more details.
+Microimpute's cross-validation selects the metric based on variable type: quantile loss for numerical variables, log loss for categorical variables. See the [Metrics page](./metrics.md) for details.
 
 ## Cross-validation
 
-Cross-validation provides robust estimates of how well a model will generalize to unseen data by training and testing on multiple data splits. Microimpute's cross-validation automatically selects the appropriate metric based on variable type: quantile loss for numerical variables and log loss for categorical variables.
-
 ### cross_validate_model
 
 ```python
@@ -98,7 +95,7 @@ For model selection, focus on the test loss (`mean_test`). When comparing multip
 
 ## Imputation generation for model comparison
 
-The `get_imputations` function generates imputations using cross-validation for multiple model classes in a single call, organizing results in a consistent format for downstream comparison and evaluation.
+The `get_imputations` function generates imputations for multiple model classes in a single call, organizing results in a consistent format for comparison.
 
 ### get_imputations
 
diff --git a/docs/imputation-benchmarking/index.md b/docs/imputation-benchmarking/index.md
index 337e1c5..d8b9075 100644
--- a/docs/imputation-benchmarking/index.md
+++ b/docs/imputation-benchmarking/index.md
@@ -1,5 +1,5 @@
 # Benchmarking different imputation methods
 
-This chapter describes how the Microimpute package allows you to compare different imputation methods using preprocessing, cross-validation, metric comparison, and evaluation tools.
+This chapter describes how microimpute lets you compare imputation methods using preprocessing, cross-validation, metric comparison, and evaluation tools.
 
-The benchmarking functionality enables systematically comparing multiple imputation models using a common dataset, allowing for robust evaluation of their performance. It supports cross-validation to diagnose overfitting and measure performance on training data leveraging the availability of ground truth. By assessing accuracy of numeric imputation across various quantiles, it is possible to gain a more comprehensive understanding of how each method performs across different levels of the distribution. Categorical imputation is assessed with log loss. This process is further supported by visualizations that highlight differences between approaches, making it easy to identify which imputation methods perform best under specific conditions.  Predictor evaluation tools are also available to inform decision-making when setting up the imputation task.
+The benchmarking functionality supports systematic comparison of multiple models on a common dataset. Cross-validation diagnoses overfitting and measures performance on held-out data where ground truth is available. Numerical imputation accuracy is assessed across quantiles via quantile loss; categorical imputation uses log loss. Visualizations show differences between methods, and predictor evaluation tools help inform variable selection when setting up the imputation task.
diff --git a/docs/imputation-benchmarking/metrics.md b/docs/imputation-benchmarking/metrics.md
index cf2f6a3..32f429f 100644
--- a/docs/imputation-benchmarking/metrics.md
+++ b/docs/imputation-benchmarking/metrics.md
@@ -1,14 +1,14 @@
 # Metrics and evaluation
 
-This page documents the evaluation metrics and predictor analysis tools available for assessing imputation quality. These utilities help understand model performance, compare methods, and analyze the contribution of individual predictors.
+This page documents the evaluation metrics and predictor analysis tools for assessing imputation quality.
 
 ## Loss metrics
 
-Microimpute employs evaluation metrics tailored to the type of variable being imputed. The framework automatically selects the appropriate metric based on whether the imputed variable is numerical or categorical, ensuring meaningful performance assessment across different data types.
+Microimpute selects the evaluation metric based on whether the imputed variable is numerical or categorical.
 
 ### Quantile loss
 
-Quantile loss assesses imputation quality for numerical variables. This approach provides a more nuanced evaluation than traditional metrics like mean squared error, particularly for capturing performance across different parts of the distribution.
+Quantile loss assesses imputation quality for numerical variables. It captures performance across different parts of the distribution, unlike mean squared error which only measures average accuracy.
 
 The quantile loss implements the standard pinball loss formulation:
 
@@ -62,7 +62,7 @@ When predictions are class labels rather than probabilities, the function conver
 
 ### compute_loss
 
-A unified function that selects the appropriate loss metric based on the specified type, providing a consistent interface for both numerical and categorical evaluation.
+A unified function that selects the loss metric based on the specified type.
 
 ```python
 def compute_loss(
@@ -86,7 +86,7 @@ Returns a tuple of `(element_wise_losses, mean_loss)`.
 
 ### compare_metrics
 
-Compares metrics across multiple imputation methods, automatically detecting variable types and applying the appropriate metric. For models that handle both numerical and categorical variables, the evaluation produces separate results for each metric type.
+Compares metrics across multiple imputation methods, detecting variable types and applying the appropriate metric. For models that handle both numerical and categorical variables, results are produced separately for each metric type.
 
 ```python
 def compare_metrics(
@@ -106,7 +106,7 @@ Returns a DataFrame with columns `Method`, `Imputed Variable`, `Percentile`, `Lo
 
 ## Distribution comparison
 
-Beyond point-wise loss metrics, evaluating how well imputed values preserve distributional characteristics provides insight into whether the imputation maintains the statistical properties of the original data.
+Evaluating how well imputed values preserve distributional characteristics tells you whether the imputation maintains the statistical properties of the original data.
 
 ### Wasserstein distance
 
@@ -126,7 +126,7 @@ $$D_{KL}(P||Q) = \sum_{x \in \mathcal{X}} P(x) \log\left(\frac{P(x)}{Q(x)}\right
 
 where $P$ is the reference distribution (original data), $Q$ is the approximation (imputed data), and $\mathcal{X}$ is the set of all possible categorical values. KL divergence measures how much information is lost when using the imputed distribution to approximate the true distribution. Lower values indicate better preservation of the original categorical distribution.
 
-When sample weights are provided, the probability distributions are computed as weighted proportions rather than simple counts, ensuring proper comparison of weighted survey data.
+When sample weights are provided, the probability distributions are computed as weighted proportions rather than simple counts, so that weighted survey data can be compared correctly.
 
 ### kl_divergence
 
@@ -178,11 +178,11 @@ Note that data must not contain null or infinite values. If your data contains s
 
 ## Predictor analysis
 
-Understanding which predictors contribute most to imputation quality helps with feature selection and model interpretation. These tools analyze predictor-target relationships and evaluate sensitivity to predictor selection.
+These tools analyze which predictors contribute most to imputation quality, helping with feature selection and model interpretation.
 
 ### Mutual information
 
-Mutual information measures the reduction in uncertainty about one variable given knowledge of another. Unlike correlation coefficients that capture only linear relationships, mutual information detects any statistical dependency, making it valuable for mixed data types.
+Mutual information measures the reduction in uncertainty about one variable given knowledge of another. Unlike correlation coefficients, which capture only linear relationships, mutual information detects any statistical dependency.
 
 For discrete random variables $X$ and $Y$:
 
@@ -200,21 +200,23 @@ where $H(X)$ and $H(Y)$ are the entropies of $X$ and $Y$ respectively. Normalize
 def compute_predictor_correlations(
     data: pd.DataFrame,
     predictors: List[str],
-    imputed_variables: List[str],
+    imputed_variables: Optional[List[str]] = None,
+    method: str = "all",
 ) -> Dict[str, pd.DataFrame]
 ```
 
-| Parameter | Type | Description |
-|-----------|------|-------------|
-| data | pd.DataFrame | Dataset containing predictors and target variables |
-| predictors | List[str] | Column names of predictor variables |
-| imputed_variables | List[str] | Column names of target variables |
+| Parameter | Type | Default | Description |
+|-----------|------|---------|-------------|
+| data | pd.DataFrame | - | Dataset containing predictors and target variables |
+| predictors | List[str] | - | Column names of predictor variables |
+| imputed_variables | List[str] | None | Column names of target variables |
+| method | str | "all" | Which correlation method to use: "all", "mi" (mutual information), "pearson", or "spearman" |
 
-Returns a dictionary containing `predictor_target_mi` DataFrame with mutual information scores.
+Returns a dictionary containing DataFrames with correlation scores (e.g. `predictor_target_mi` for mutual information).
 
 ### Leave-one-out analysis
 
-Leave-one-out predictor analysis evaluates model performance when each predictor is excluded. By comparing loss with and without each predictor, you can assess its contribution to imputation quality. Predictors whose removal causes large increases in loss are most important, while those with minimal impact might be candidates for removal to simplify the model.
+Leave-one-out predictor analysis evaluates model performance when each predictor is excluded. Predictors whose removal causes large increases in loss are the most important, while those with minimal impact might be candidates for removal.
 
 ### leave_one_out_analysis
 
@@ -223,9 +225,13 @@ def leave_one_out_analysis(
     data: pd.DataFrame,
     predictors: List[str],
     imputed_variables: List[str],
-    model_class: Type,
-    quantiles: Optional[List[float]] = QUANTILES,
-) -> Dict[str, Any]
+    model_class: Type[Imputer],
+    weight_col: Optional[Union[str, np.ndarray, pd.Series]] = None,
+    quantiles: List[float] = QUANTILES,
+    train_size: float = TRAIN_SIZE,
+    n_jobs: int = 1,
+    random_state: int = RANDOM_STATE,
+) -> pd.DataFrame
 ```
 
 | Parameter | Type | Default | Description |
@@ -233,14 +239,18 @@ def leave_one_out_analysis(
 | data | pd.DataFrame | - | Complete dataset |
 | predictors | List[str] | - | Column names of predictor variables |
 | imputed_variables | List[str] | - | Column names of variables to impute |
-| model_class | Type | - | Imputer class to evaluate |
+| model_class | Type[Imputer] | - | Imputer class to evaluate |
+| weight_col | str, np.ndarray, or pd.Series | None | Sample weights column name or array |
 | quantiles | List[float] | [0.05 to 0.95 in steps of 0.05] | Quantiles to evaluate |
+| train_size | float | 0.8 | Proportion of data for training |
+| n_jobs | int | 1 | Number of parallel jobs |
+| random_state | int | 42 | Random seed for reproducibility |
 
-Returns a dictionary containing loss increase and relative impact for each predictor.
+Returns a DataFrame containing loss increase and relative impact for each predictor.
 
 ### Progressive predictor inclusion
 
-Progressive inclusion analysis adds predictors one at a time in order of their mutual information with the target. This greedy forward selection reveals the optimal inclusion order, marginal contribution of each predictor, and the minimal set of predictors achieving near-optimal performance. Diminishing returns in loss reduction indicate when additional predictors provide negligible improvement.
+Progressive inclusion adds predictors one at a time, ordered by their mutual information with the target. This greedy forward selection reveals the optimal inclusion order and the marginal contribution of each predictor. Diminishing returns in loss reduction indicate when additional predictors add little.
 
 ### progressive_predictor_inclusion
 
@@ -249,8 +259,12 @@ def progressive_predictor_inclusion(
     data: pd.DataFrame,
     predictors: List[str],
     imputed_variables: List[str],
-    model_class: Type,
+    model_class: Type[Imputer],
+    weight_col: Optional[Union[str, np.ndarray, pd.Series]] = None,
     quantiles: Optional[List[float]] = QUANTILES,
+    train_size: Optional[float] = TRAIN_SIZE,
+    max_predictors: Optional[int] = None,
+    random_state: Optional[int] = RANDOM_STATE,
 ) -> Dict[str, Any]
 ```
 
@@ -259,8 +273,12 @@ def progressive_predictor_inclusion(
 | data | pd.DataFrame | - | Complete dataset |
 | predictors | List[str] | - | Column names of predictor variables |
 | imputed_variables | List[str] | - | Column names of variables to impute |
-| model_class | Type | - | Imputer class to evaluate |
+| model_class | Type[Imputer] | - | Imputer class to evaluate |
+| weight_col | str, np.ndarray, or pd.Series | None | Sample weights column name or array |
 | quantiles | List[float] | [0.05 to 0.95 in steps of 0.05] | Quantiles to evaluate |
+| train_size | float | 0.8 | Proportion of data for training |
+| max_predictors | int | None | Maximum number of predictors to include (None for all) |
+| random_state | int | 42 | Random seed for reproducibility |
 
 Returns a dictionary containing `inclusion_order` (list of predictors in optimal order) and `predictor_impacts` (list of dicts with predictor name and loss reduction).
 
@@ -282,7 +300,7 @@ metrics_df = compare_metrics(
         "QRF": qrf_imputations,
         "OLS": ols_imputations,
     },
-    imputed_variables=imputed_variables
+    imputed_variables=imputed_variables,
 )
 
 # Evaluate distributional match with survey weights
@@ -295,7 +313,13 @@ dist_df_weighted = compare_distributions(
 )
 
 # Analyze predictor importance
-mi_scores = compute_predictor_correlations(data, predictors, imputed_variables)
-loo_results = leave_one_out_analysis(data, predictors, imputed_variables, QRF)
-inclusion_results = progressive_predictor_inclusion(data, predictors, imputed_variables, QRF)
+mi_scores = compute_predictor_correlations(
+    data, predictors, imputed_variables, method="mi"
+)
+loo_results = leave_one_out_analysis(
+    data, predictors, imputed_variables, QRF, weight_col="wgt"
+)
+inclusion_results = progressive_predictor_inclusion(
+    data, predictors, imputed_variables, QRF, weight_col="wgt"
+)
 ```
diff --git a/docs/imputation-benchmarking/preprocessing.md b/docs/imputation-benchmarking/preprocessing.md
index caa352c..736d73e 100644
--- a/docs/imputation-benchmarking/preprocessing.md
+++ b/docs/imputation-benchmarking/preprocessing.md
@@ -1,12 +1,12 @@
 # Data preprocessing
 
-Preprocessing transformations can improve model performance by normalizing scale differences or handling skewed distributions. These are supported by `preprocess_data` and transformation-specific functions.
+Preprocessing transformations can improve model performance by normalizing scale differences or handling skewed distributions. The main entry point is `preprocess_data`, with transformation-specific functions also available.
 
 ## Transformation options
 
 Microimpute supports three transformation types that can be applied to numeric columns before training. Each transformation automatically excludes categorical and boolean columns to prevent encoding issues.
 
-**Normalization (z-score)** standardizes data to have mean 0 and standard deviation 1. This transformation is useful when predictors have different scales, ensuring that all features contribute equally to distance-based or gradient-based models.
+**Normalization (z-score)** standardizes data to have mean 0 and standard deviation 1. This is useful when predictors have different scales, so that all features contribute equally to distance-based or gradient-based models.
 
 **Log transformation** applies the natural logarithm to values. This is effective for right-skewed distributions common in financial data like income or wealth. The transformation requires all values to be strictly positive.
 
diff --git a/docs/imputation-benchmarking/visualizations.md b/docs/imputation-benchmarking/visualizations.md
index d6c9450..21dde02 100644
--- a/docs/imputation-benchmarking/visualizations.md
+++ b/docs/imputation-benchmarking/visualizations.md
@@ -159,8 +159,10 @@ For quantile loss, the plot shows train and test loss across quantiles as groupe
 from microimpute.visualizations import model_performance_results
 
 # Visualize cross-validation results for a single model
+# Pass the full cv_results dict (not just the inner DataFrame)
+# to preserve error bar data from cross-validation folds
 perf_viz = model_performance_results(
-    results=cv_results["quantile_loss"]["results"],
+    results=cv_results,
     model_name="QRF",
     method_name="Cross-validation",
     metric="quantile_loss"
diff --git a/docs/index.md b/docs/index.md
index 354005d..979a4c6 100644
--- a/docs/index.md
+++ b/docs/index.md
@@ -1,21 +1,21 @@
 # Microimpute
 
-Microimpute is a powerful framework that enables variable imputation through a variety of statistical methods. By providing a consistent interface across different imputation techniques, it allows researchers and data scientists to easily compare and benchmark different approaches using quantile loss and log loss calculations to determine the method provding most accurate results. Thus, Microimpute provides two main uses: imputing one or multiple variables with one of the methods available, and comparing and benchmarking different methods to inform a method's choice.
+Microimpute is a Python package for imputing variables from one survey dataset onto another. It wraps five imputation methods behind a common interface so you can benchmark them on your data and pick the one that works best: impute one or multiple variables with any of the available methods, and compare their performance using quantile loss and log loss.
 
-The framework currently supports the following imputation methods:
-- Statistical Matching
-- Ordinary Least Squares Linear Regression
-- Quantile Random Forests
+The package currently supports:
+- Hot Deck Matching
+- Ordinary Least Squares (OLS) Linear Regression
+- Quantile Random Forests (QRF)
 - Quantile Regression
-- Mixture Density Networks
+- Mixture Density Networks (MDN)
 
-This is a work in progress that may evolve over time, including new statistical imputation methods and features. 
+This is a work in progress and may evolve over time with new methods and features.
 
 ## Microimputation dashboard
 
 Users can visualize imputation and benchmarking results at https://microimpute-dashboard.vercel.app/.
 
-To use the dashboard for visualization, CSV files must contain the following columns in this exact order:
+To use the dashboard, CSV files must contain the following columns in this exact order:
 - `type`: Type of metric (e.g., "benchmark_loss", "distribution_distance", "predictor_correlation")
 - `method`: Imputation method name (e.g., "QRF", "OLS", "QuantReg", "Matching", "MDN")
 - `variable`: Variable being imputed or analyzed
@@ -25,4 +25,4 @@ To use the dashboard for visualization, CSV files must contain the following col
 - `split`: Data split indicator (e.g., "train", "test", "full")
 - `additional_info`: JSON-formatted string with additional metadata
 
-Users can use the `format_csv()` function from `microimpute.utils` to automatically format imputation and benchmarking results into the correct structure for dashboard visualization. This function accepts outputs from various analysis functions (autoimpute results, comparison metrics, distribution comparisons, etc.) and returns a properly formatted DataFrame.
\ No newline at end of file
+The `format_csv()` function from `microimpute.utils` formats imputation and benchmarking results into the correct structure for the dashboard. It accepts outputs from various analysis functions (autoimpute results, comparison metrics, distribution comparisons) and returns a properly formatted DataFrame.
diff --git a/docs/models/imputer/implement-new-model.md b/docs/models/imputer/implement-new-model.md
index 839da1d..edf2125 100644
--- a/docs/models/imputer/implement-new-model.md
+++ b/docs/models/imputer/implement-new-model.md
@@ -1,15 +1,15 @@
 # Creating a new imputer model
 
-This document demonstrates how to create a new imputation model by extending the `Imputer` and `ImputerResults` abstract base classes in Microimpute.
+This document shows how to create a new imputation model by extending the `Imputer` and `ImputerResults` abstract base classes.
 
-## Understanding the Microimpute architecture
+## Architecture
 
-Microimpute uses a two-class architecture for imputation models:
+Microimpute uses a two-class architecture:
 
-1. **Imputer**: The base model class that handles model initialization and fitting
+1. **Imputer**: Handles model initialization and fitting
 2. **ImputerResults**: Represents a fitted model and handles prediction
 
-This separation provides a clean distinction between the model definition and the fitted model instance, similar to statsmodels' approach. Remember to check how currently supported models have been implemented if you would like to ensure full compatibility. 
+This separation is similar to statsmodels' approach. Look at the existing model implementations for reference.
 
 ```python
 from typing import Dict, List, Optional, Any
@@ -178,11 +178,9 @@ The new `NewModel` model is then ready to be integrated into the Microimpute ben
 
 ```python
 from microimpute.models import OLS, QRF
-from microimpute.comparisons import (
-    get_imputations,
-    compare_quantile_loss,
-)
-from microimpute.visualizations.plotting import method_comparison_results
+from microimpute.comparisons import get_imputations
+from microimpute.comparisons.metrics import compare_metrics
+from microimpute.visualizations import method_comparison_results
 
 # Define models to compare
 model_classes = [NewModel, OLS, QRF]
@@ -195,45 +193,33 @@ method_imputations = get_imputations(
     model_classes, X_train, X_test, predictors, imputed_variables
 )
 
-# Compare quantile loss
-loss_comparison_df = compare_quantile_loss(Y_test, method_imputations, imputed_variables)
+# Compare metrics across methods
+loss_comparison_df = compare_metrics(Y_test, method_imputations, imputed_variables)
 
 # Plot the comparison
 comparison_viz = method_comparison_results(
-        data=loss_comparison_df,
-        metric_name="Test Quantile Loss",
-        data_format="long",
-    )
-fig = comparison_viz.plot(
-    show_mean=True,
+    data=loss_comparison_df,
+    metric="quantile_loss",
+    data_format="long",
 )
+fig = comparison_viz.plot(show_mean=True)
 fig.show()
 ```
 
-## Best practices for implementing new models
-
-When implementing a new imputation model for Microimpute, adhering to certain best practices will ensure your model integrates seamlessly with the framework and provides a consistent experience for users.
+## Best practices
 
 ### Architecture
 
-The two-class architecture forms the foundation of a well-designed imputation model. You should create an `Imputer` subclass that handles model definition and fitting operations, establishing the core functionality of your approach. This class should be complemented by an `ImputerResults` subclass that represents the fitted model state and handles all prediction-related tasks. This separation of concerns creates a clean distinction between the fitting and prediction phases of your model's lifecycle.
-
-Within these classes, you must implement the required abstract methods to fulfill the contract with the base classes. Your `Imputer` subclass should provide a thorough implementation of the `_fit()` method that handles the training process for your specific algorithm. Similarly, your `ImputerResults` subclass needs to implement the `_predict()` method that applies the fitted model to new data and generates predictions at requested quantiles. Check how currently supported models have been implemented if you would like to ensure iterative imputation for multiple target variables fully compatible with how other models do it. For example, to be able to compute quantile loss accross imputed quantiles and variables and compare it with different methods you must ensure iterative imputation within the fitting and predicting methods.
+Create an `Imputer` subclass for fitting and an `ImputerResults` subclass for prediction. Implement `_fit()` in the former and `_predict()` in the latter. Look at existing models to see how they handle iterative imputation across multiple target variables, which is needed for cross-method comparison via quantile loss.
 
 ### Error handling
 
-Robust error handling is crucial for creating reliable imputation models. Your implementation should wrap model fitting and prediction operations in appropriate try/except blocks to capture and handle potential errors gracefully. When exceptions occur, provide informative error messages that help users understand what went wrong and how to address the issue. Use appropriate error types such as ValueError for input validation failures and RuntimeError for operational failures during model execution.
-
-Effective logging complements good error handling by providing visibility into the model's operation. Use the self.logger instance consistently throughout your code to record important information about the model's state and progress. Log significant events like the start and completion of fitting operations, parameter values, and any potential issues or warnings that arise during execution.
+Wrap fitting and prediction in try/except blocks. Use `ValueError` for bad inputs, `RuntimeError` for operational failures, and include informative messages. Use `self.logger` for logging significant events (fitting start/end, parameter values, warnings).
 
 ### Parameters and validation
 
-Type safety and parameter validation enhance the usability and reliability of your model. Add comprehensive type hints to all methods and parameters to enable better IDE support and make your code more self-documenting. Apply the `validate_call` decorator with the standard VALIDATE_CONFIG configuration to method signatures to enforce parameter validation consistently.
-
-Your implementation should thoughtfully support model-specific parameters that may be needed to control the behavior of your algorithm. Design your `_fit()` and `_predict()` methods to accept and properly utilize these parameters, ensuring they affect the model's operation as intended. Document all parameters clearly in your docstrings, explaining their purpose, expected values, and default behavior to guide users in effectively configuring your model.
-
-### Documentation
+Add type hints to all methods. Use the `@validate_call(config=VALIDATE_CONFIG)` decorator for parameter validation. Document model-specific parameters in docstrings with their purpose, expected values, and defaults.
 
-Comprehensive documentation makes your model accessible to others. Include detailed class-level docstrings that explain your model's theoretical approach, strengths, limitations, and appropriate use cases. Document all methods with properly structured docstring sections covering arguments, return values, and potential exceptions. Where appropriate, provide usage examples that demonstrate how to initialize, train, and use your model for prediction tasks.
+### Testing
 
-The documentation should be complemented by thorough unit tests that verify your implementation works correctly. Create tests that check both basic interface compliance (ensuring your model adheres to the expected API) and model-specific functionality (validating that your algorithm produces correct results). Comprehensive testing helps catch issues early and provides confidence that your implementation will work reliably in production environments.
+Write tests for both interface compliance (does your model follow the expected API?) and model-specific correctness (does it produce sensible results?).
diff --git a/docs/models/imputer/index.md b/docs/models/imputer/index.md
index 3cf0dc0..26e20d8 100644
--- a/docs/models/imputer/index.md
+++ b/docs/models/imputer/index.md
@@ -1,11 +1,11 @@
 # The Imputer class
 
-The `Imputer` class serves as an abstract base class that defines the common interface for all imputation models within the Microimpute framework. It establishes a structure with essential methods for data validation, model fitting, and prediction. Every specialized imputation model in the system inherits from this class and implements the required abstract methods to provide its unique functionality.
+The `Imputer` class is the abstract base class that defines the common interface for all imputation models in microimpute. Every model inherits from it and implements the required abstract methods for fitting and prediction.
 
 ## Key features
 
-The Imputer architecture provides numerous benefits to the overall system design. It defines a consistent API with standardized `fit()` and `predict()` methods, ensuring that all models can be used interchangeably regardless of their underlying implementation details. This uniformity makes it straightforward to swap imputation techniques within your workflow. Thus, all imputers will share basic functionality like the handling of weighted data, using a "weights" column for sampling training data, to preserve data distributions better.
+All models share standardized `fit()` and `predict()` methods, so they can be used interchangeably regardless of underlying implementation. All imputers also share functionality like weighted data handling through a `weight_col` parameter.
 
-The design carefully enforces proper usage by ensuring no model can call `predict()` without first fitting the model to the data. This logical constraint helps prevent common errors and makes the API more intuitive to use. Additionally, the base implementation handles validation of parameters and input data, reducing code duplication across different model implementations and ensuring that all models perform appropriate validation checks.
+The design enforces that `predict()` cannot be called before `fit()`. The base implementation also handles parameter and input data validation, so individual models don't need to duplicate those checks.
 
-When using the different imputers in isolation, and not as part of wider pipeline functions like `autoimpute` preprocessing is supported by `preprocess_data` which can help normalize the data and split it into train and test splits. For an example of how to integrate them see [matching-imputation.ipynb](../matching/matching-imputation.ipynb).
+When using imputers in isolation (not through `autoimpute`), preprocessing is available via `preprocess_data`, which can normalize the data and split it into train/test sets. See [matching-imputation.ipynb](../matching/matching-imputation.ipynb) for an example.
diff --git a/docs/models/matching/index.md b/docs/models/matching/index.md
index 849d851..ca58c4f 100644
--- a/docs/models/matching/index.md
+++ b/docs/models/matching/index.md
@@ -1,27 +1,23 @@
 # Hot-Deck Matching
 
-The `Matching` model implements imputation through an elegant nearest neighbor distance hot deck matching approach. This technique draws from the principles of statistical matching, using existing complete records (donors) to provide values for records with missing data (recipients) by establishing meaningful connections based on similarities in predictor variables.
+The `Matching` model imputes missing values using nearest neighbor distance hot deck matching. It finds donor records that are similar to each recipient based on predictor variables and transfers the donor's observed values.
 
 ## Variable type support
 
-The matching model can handle any variable type—numerical, categorical, boolean, or mixed. Since it transfers actual observed values from similar records rather than generating predictions, it naturally preserves the original data type and distribution of each variable.
+Matching handles any variable type: numerical, categorical, boolean, or mixed. Because it transfers actual observed values rather than generating predictions, it preserves the original data type and distribution of each variable.
 
 ## How it works
 
-Statistical or hot-deck matching in Microimpute builds upon the foundation of R's StatMatch package, accessed through the rpy2 interface to provide a seamless integration of R's statistical power with Python's flexibility. The implementation leverages the well-established nearest neighbor distance hot deck matching algorithm, which has a strong theoretical foundation in statistical literature.
+The implementation builds on R's StatMatch package, accessed through the rpy2 interface.
 
-During the fitting phase, the model carefully preserves both the complete donor dataset and the relevant variable names that will guide the matching process. This stored information becomes the knowledge base from which the model will draw when making imputations.
+During fitting, the model stores the complete donor dataset and the relevant variable names. During prediction, each record in the test dataset (the recipients) is compared against the stored donors using distance calculations on the predictor variables. The algorithm finds the closest donor for each recipient and transfers the target variable values.
 
-The prediction stage initiates a deliberate matching process where each record in the test dataset (the recipients) is systematically compared with the stored donor records. The comparison calculates similarity distances based on the predictor variables, identifying the donor records that most closely resemble each recipient. The matching algorithm efficiently navigates the multidimensional space defined by the predictor variables to find optimal donor-recipient pairs.
-
-Once the matching is complete, the model transfers the values from the matched donors to the recipients for the specified imputed variable. This transfer preserves the natural relationships and patterns present in the original data, as the values being imputed were actually observed rather than synthetically generated.
+Because the imputed values are drawn from actually observed records, the natural relationships in the original data are preserved.
 
 ## Key features
 
-The statistical matching imputer offers a truly non-parametric approach that operates without imposing restrictive assumptions about the underlying data distribution. This distribution-free nature makes it particularly valuable in scenarios where the data doesn't conform to common statistical assumptions or when the relationships are too complex to model parametrically.
-
-One of the most compelling advantages of this method is its ability to preserve the empirical distribution of the imputed variables. Since the imputed values come directly from observed data points, the resulting dataset maintains the natural structure, variability, and relationships present in the original data. This preservation extends to features like multimodality, skewness, and natural bounds that might be lost in model-based approaches.
+Matching is non-parametric: it makes no assumptions about the data distribution. This makes it useful when the data doesn't fit standard parametric models, or when the relationships between predictors and targets are hard to specify in closed form.
 
-The technique demonstrates versatility in handling complex relationships between variables, particularly when there exists a good match across datasets. Without requiring explicit specification of interaction terms or functional forms, it naturally captures the intricate dependencies that exist in the data through the matching process. This makes it especially valuable for datasets where the relationships are not well understood or are difficult to express mathematically.
+The method preserves the empirical distribution of the imputed variables. Since values come directly from observed data points, features like multimodality, skewness, and natural bounds are maintained. A model-based approach might smooth these away.
 
-Perhaps most distinctively, the statistical matching approach returns actual observed values rather than modeled estimates. This characteristic ensures that the imputed values are realistic and plausible, as they represent real observations from similar data points. The method essentially says, "We have seen this pattern before, and here's what the missing values looked like in that situation," providing a grounded approach to filling in missing information.
+One limitation is that Matching does not incorporate quantile information. It matches donor and receiver units identically regardless of the quantile being predicted, which means it cannot distinguish between different parts of the conditional distribution. It may also fail to capture non-linear predictor-target relationships despite producing a plausible marginal distribution.
diff --git a/docs/models/matching/matching-imputation.ipynb b/docs/models/matching/matching-imputation.ipynb
index 1cdace2..df9d7d0 100644
--- a/docs/models/matching/matching-imputation.ipynb
+++ b/docs/models/matching/matching-imputation.ipynb
@@ -77,7 +77,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 1,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -112,7 +112,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 7,
+   "execution_count": 2,
    "metadata": {},
    "outputs": [
     {
@@ -161,7 +161,7 @@
        "      <td>-0.002592</td>\n",
        "      <td>0.019907</td>\n",
        "      <td>-0.017646</td>\n",
-       "      <td>False</td>\n",
+       "      <td>True</td>\n",
        "      <td>1</td>\n",
        "    </tr>\n",
        "    <tr>\n",
@@ -175,7 +175,7 @@
        "      <td>-0.039493</td>\n",
        "      <td>-0.068332</td>\n",
        "      <td>-0.092204</td>\n",
-       "      <td>True</td>\n",
+       "      <td>False</td>\n",
        "      <td>2</td>\n",
        "    </tr>\n",
        "    <tr>\n",
@@ -189,7 +189,7 @@
        "      <td>-0.002592</td>\n",
        "      <td>0.002861</td>\n",
        "      <td>-0.025930</td>\n",
-       "      <td>False</td>\n",
+       "      <td>True</td>\n",
        "      <td>3</td>\n",
        "    </tr>\n",
        "    <tr>\n",
@@ -245,7 +245,7 @@
        "      <td>-0.002592</td>\n",
        "      <td>0.031193</td>\n",
        "      <td>0.007207</td>\n",
-       "      <td>True</td>\n",
+       "      <td>False</td>\n",
        "      <td>438</td>\n",
        "    </tr>\n",
        "    <tr>\n",
@@ -273,7 +273,7 @@
        "      <td>-0.011080</td>\n",
        "      <td>-0.046883</td>\n",
        "      <td>0.015491</td>\n",
-       "      <td>False</td>\n",
+       "      <td>True</td>\n",
        "      <td>440</td>\n",
        "    </tr>\n",
        "    <tr>\n",
@@ -287,7 +287,7 @@
        "      <td>0.026560</td>\n",
        "      <td>0.044529</td>\n",
        "      <td>-0.025930</td>\n",
-       "      <td>True</td>\n",
+       "      <td>False</td>\n",
        "      <td>441</td>\n",
        "    </tr>\n",
        "    <tr>\n",
@@ -301,7 +301,7 @@
        "      <td>-0.039493</td>\n",
        "      <td>-0.004222</td>\n",
        "      <td>0.003064</td>\n",
-       "      <td>True</td>\n",
+       "      <td>False</td>\n",
        "      <td>442</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
@@ -311,22 +311,22 @@
       ],
       "text/plain": [
        "          age       sex       bmi        bp        s1  ...        s4        s5        s6   bool  wgt\n",
-       "0    0.038076  0.050680  0.061696  0.021872 -0.044223  ... -0.002592  0.019907 -0.017646  False    1\n",
-       "1   -0.001882 -0.044642 -0.051474 -0.026328 -0.008449  ... -0.039493 -0.068332 -0.092204   True    2\n",
-       "2    0.085299  0.050680  0.044451 -0.005670 -0.045599  ... -0.002592  0.002861 -0.025930  False    3\n",
+       "0    0.038076  0.050680  0.061696  0.021872 -0.044223  ... -0.002592  0.019907 -0.017646   True    1\n",
+       "1   -0.001882 -0.044642 -0.051474 -0.026328 -0.008449  ... -0.039493 -0.068332 -0.092204  False    2\n",
+       "2    0.085299  0.050680  0.044451 -0.005670 -0.045599  ... -0.002592  0.002861 -0.025930   True    3\n",
        "3   -0.089063 -0.044642 -0.011595 -0.036656  0.012191  ...  0.034309  0.022688 -0.009362   True    4\n",
        "4    0.005383 -0.044642 -0.036385  0.021872  0.003935  ... -0.002592 -0.031988 -0.046641  False    5\n",
        "..        ...       ...       ...       ...       ...  ...       ...       ...       ...    ...  ...\n",
-       "437  0.041708  0.050680  0.019662  0.059744 -0.005697  ... -0.002592  0.031193  0.007207   True  438\n",
+       "437  0.041708  0.050680  0.019662  0.059744 -0.005697  ... -0.002592  0.031193  0.007207  False  438\n",
        "438 -0.005515  0.050680 -0.015906 -0.067642  0.049341  ...  0.034309 -0.018114  0.044485  False  439\n",
-       "439  0.041708  0.050680 -0.015906  0.017293 -0.037344  ... -0.011080 -0.046883  0.015491  False  440\n",
-       "440 -0.045472 -0.044642  0.039062  0.001215  0.016318  ...  0.026560  0.044529 -0.025930   True  441\n",
-       "441 -0.045472 -0.044642 -0.073030 -0.081413  0.083740  ... -0.039493 -0.004222  0.003064   True  442\n",
+       "439  0.041708  0.050680 -0.015906  0.017293 -0.037344  ... -0.011080 -0.046883  0.015491   True  440\n",
+       "440 -0.045472 -0.044642  0.039062  0.001215  0.016318  ...  0.026560  0.044529 -0.025930  False  441\n",
+       "441 -0.045472 -0.044642 -0.073030 -0.081413  0.083740  ... -0.039493 -0.004222  0.003064  False  442\n",
        "\n",
        "[442 rows x 12 columns]"
       ]
      },
-     "execution_count": 7,
+     "execution_count": 2,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -347,7 +347,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 8,
+   "execution_count": 3,
    "metadata": {},
    "outputs": [
     {
@@ -477,7 +477,7 @@
        "max    1.107267e-01  5.068012e-02  1.705552e-01  1.320436e-01  1.539137e-01  1.852344e-01  442.000000"
       ]
      },
-     "execution_count": 8,
+     "execution_count": 3,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -501,7 +501,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 9,
+   "execution_count": 4,
    "metadata": {},
    "outputs": [
     {
@@ -540,7 +540,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 10,
+   "execution_count": 5,
    "metadata": {},
    "outputs": [
     {
@@ -643,7 +643,7 @@
        "73   0.012648  0.050680 -0.020218 -0.002228 NaN NaN   NaN   74"
       ]
      },
-     "execution_count": 10,
+     "execution_count": 5,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -672,7 +672,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 11,
+   "execution_count": 6,
    "metadata": {},
    "outputs": [
     {
@@ -691,7 +691,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 12,
+   "execution_count": 7,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -710,7 +710,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 13,
+   "execution_count": 8,
    "metadata": {},
    "outputs": [
     {
@@ -744,31 +744,31 @@
        "      <th>287</th>\n",
        "      <td>0.024574</td>\n",
        "      <td>-0.039493</td>\n",
-       "      <td>True</td>\n",
+       "      <td>False</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>211</th>\n",
        "      <td>0.030078</td>\n",
        "      <td>-0.039493</td>\n",
-       "      <td>True</td>\n",
+       "      <td>False</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>72</th>\n",
        "      <td>0.038334</td>\n",
        "      <td>-0.039493</td>\n",
-       "      <td>False</td>\n",
+       "      <td>True</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>321</th>\n",
        "      <td>-0.013953</td>\n",
        "      <td>-0.002592</td>\n",
-       "      <td>True</td>\n",
+       "      <td>False</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>73</th>\n",
        "      <td>-0.031840</td>\n",
        "      <td>-0.039493</td>\n",
-       "      <td>False</td>\n",
+       "      <td>True</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
@@ -776,14 +776,14 @@
       ],
       "text/plain": [
        "           s1        s4   bool\n",
-       "287  0.024574 -0.039493   True\n",
-       "211  0.030078 -0.039493   True\n",
-       "72   0.038334 -0.039493  False\n",
-       "321 -0.013953 -0.002592   True\n",
-       "73  -0.031840 -0.039493  False"
+       "287  0.024574 -0.039493  False\n",
+       "211  0.030078 -0.039493  False\n",
+       "72   0.038334 -0.039493   True\n",
+       "321 -0.013953 -0.002592  False\n",
+       "73  -0.031840 -0.039493   True"
       ]
      },
-     "execution_count": 13,
+     "execution_count": 8,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -808,7 +808,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 14,
+   "execution_count": 9,
    "metadata": {},
    "outputs": [
     {
@@ -826,22 +826,22 @@
          "x": [
           0.12501870313429186,
           0.03430885887772673,
-          false,
+          true,
           -0.0249601584096303,
           -0.03949338287409329,
-          false,
+          true,
           0.10300345740307394,
           -0.002592261998183278,
           true,
           0.05484510736603471,
           0.14132210941786577,
-          true,
+          false,
           0.03833367306762126,
           0.03430885887772673,
-          false,
+          true,
           0.09887559882847057,
           -0.002592261998183278,
-          true,
+          false,
           0.030077955918414535,
           0.03430885887772673,
           true,
@@ -859,7 +859,7 @@
           false,
           -0.06761469701386505,
           -0.002592261998183278,
-          true,
+          false,
           -0.05523112129005496,
           -0.0763945037500033,
           true,
@@ -871,40 +871,40 @@
           true,
           -0.07311850844666953,
           -0.03949338287409329,
-          false,
+          true,
           0.03833367306762126,
           0.03430885887772673,
           false,
           0.020446285911006685,
           0.07120997975363674,
-          false,
+          true,
           -0.12678066991651324,
           -0.047980640675552584,
           true,
           0.013566521620001083,
           0.03430885887772673,
-          false,
+          true,
           -0.012576582685820214,
           -0.002592261998183278,
-          false,
+          true,
           0.045213437358626866,
           -0.002592261998183278,
           false,
           -0.007072771253015731,
           -0.03949338287409329,
-          true,
+          false,
           0.016318427336403322,
           -0.002592261998183278,
           false,
           -0.009824676969417972,
           -0.03949338287409329,
-          true,
+          false,
           -0.030463969842434782,
           -0.002592261998183278,
           false,
           -0.02220825269322806,
           -0.002592261998183278,
-          false,
+          true,
           -0.051103262715451604,
           0.03430885887772673,
           false,
@@ -913,52 +913,52 @@
           false,
           0.0342058144930179,
           -0.002592261998183278,
-          true,
+          false,
           0.0080627101871966,
           0.021024455362399115,
-          true,
+          false,
           0.0025588987543921156,
           -0.002592261998183278,
           false,
           0.0025588987543921156,
           -0.002592261998183278,
-          false,
+          true,
           -0.016704441260423575,
           0.03430885887772673,
           true,
           0.045213437358626866,
           0.03615391492152222,
-          false,
+          true,
           0.07823630595545376,
           -0.002592261998183278,
-          false,
+          true,
           -0.011200629827619093,
           -0.002592261998183278,
           false,
           0.03145390877661565,
           0.019917421736121838,
-          false,
+          true,
           0.024574144485610048,
           0.03430885887772673,
           false,
           -0.001568959820211247,
           -0.03949338287409329,
-          false,
+          true,
           -0.00019300696201012598,
           -0.03949338287409329,
-          false,
+          true,
           -0.00019300696201012598,
           -0.05056371913686628,
-          true,
+          false,
           -0.06623874415566393,
           -0.002592261998183278,
-          false,
+          true,
           -0.004320865536613489,
           0.07120997975363674,
           false,
           0.04383748450042574,
           -0.014400620678474476,
-          true,
+          false,
           0.03282986163481677,
           -0.03949338287409329,
           false,
@@ -967,61 +967,61 @@
           false,
           -0.04422349842444599,
           -0.0763945037500033,
-          true,
+          false,
           -0.035967781275239266,
           -0.05167075276314359,
-          false,
+          true,
           -0.007072771253015731,
           -0.002592261998183278,
-          false,
+          true,
           -0.07311850844666953,
           -0.0763945037500033,
-          false,
+          true,
           -0.019456346976825818,
           -0.03949338287409329,
           true,
           -0.007072771253015731,
           0.07120997975363674,
-          true,
+          false,
           -0.008448724111216851,
           -0.03949338287409329,
-          false,
+          true,
           0.08924392882106273,
           0.10811110062954676,
-          true,
+          false,
           -0.0249601584096303,
           -0.03949338287409329,
           true,
           0.03282986163481677,
           -0.002592261998183278,
-          true,
+          false,
           -0.04422349842444599,
           -0.002592261998183278,
-          false,
+          true,
           -0.0029449126784123676,
           -0.03949338287409329,
           false,
           -0.033215875558837024,
           -0.0763945037500033,
-          true,
+          false,
           0.08236416453005713,
           0.07120997975363674,
-          false,
+          true,
           -0.0318399227006359,
           0.0029429061332032365,
           false,
           -0.04972730985725048,
           -0.03949338287409329,
-          false,
+          true,
           0.010814615903598841,
           -0.03949338287409329,
-          false,
+          true,
           -0.005696818394814609,
           0.03430885887772673,
           false,
           0.06172487165704031,
           -0.002592261998183278,
-          false,
+          true,
           0.05622106022423583,
           0.07120997975363674,
           true,
@@ -1039,13 +1039,13 @@
           true,
           0.039709625925822375,
           0.07120997975363674,
-          true,
+          false,
           0.045213437358626866,
           0.07120997975363674,
-          true,
+          false,
           -0.04972730985725048,
           0.01585829843977173,
-          true,
+          false,
           -0.026336111267831423,
           -0.03949338287409329,
           false,
@@ -1054,10 +1054,10 @@
           false,
           0.08511607024645937,
           0.03430885887772673,
-          false,
+          true,
           0.016318427336403322,
           0.02655962349378563,
-          true,
+          false,
           0.020446285911006685,
           -0.002592261998183278,
           true,
@@ -1069,16 +1069,16 @@
           true,
           -0.046975404140848234,
           -0.03949338287409329,
-          true,
+          false,
           -0.0029449126784123676,
           -0.047242618258034386,
-          false,
+          true,
           0.04658939021682799,
           -0.03949338287409329,
-          true,
+          false,
           -0.007072771253015731,
           -0.03949338287409329,
-          true,
+          false,
           -0.030463969842434782,
           -0.0763945037500033,
           true,
@@ -1095,19 +1095,19 @@
          "y": [
           0.024574144485610048,
           -0.03949338287409329,
-          true,
+          false,
           0.030077955918414535,
           -0.03949338287409329,
-          true,
+          false,
           0.03833367306762126,
           -0.03949338287409329,
-          false,
+          true,
           -0.013952535544021335,
           -0.002592261998183278,
-          true,
+          false,
           -0.0318399227006359,
           -0.03949338287409329,
-          false,
+          true,
           0.04246153164222462,
           -0.0763945037500033,
           true,
@@ -1116,25 +1116,25 @@
           true,
           -0.062110885581060565,
           0.026928634702544724,
-          false,
+          true,
           -0.04284754556624487,
           -0.002592261998183278,
           false,
           -0.005696818394814609,
           -0.03949338287409329,
-          false,
+          true,
           -0.015328488402222454,
           -0.002592261998183278,
           false,
           -0.001568959820211247,
           -0.03949338287409329,
-          true,
+          false,
           -0.07587041416307178,
           -0.0763945037500033,
-          true,
+          false,
           -0.005696818394814609,
           -0.002592261998183278,
-          true,
+          false,
           -0.007072771253015731,
           -0.03949338287409329,
           false,
@@ -1149,13 +1149,13 @@
           false,
           0.039709625925822375,
           0.10811110062954676,
-          true,
+          false,
           -0.009824676969417972,
           0.03430885887772673,
           false,
           0.01219056876179996,
           -0.03949338287409329,
-          false,
+          true,
           -0.034591828417038145,
           -0.0763945037500033,
           true,
@@ -1173,22 +1173,22 @@
           true,
           0.024574144485610048,
           0.15534453535071155,
-          true,
+          false,
           -0.035967781275239266,
           0.07120997975363674,
-          true,
+          false,
           -0.037343734133440394,
           -0.03949338287409329,
           false,
           -0.0318399227006359,
           -0.03949338287409329,
-          false,
+          true,
           -0.004320865536613489,
           -0.0011162171631468765,
-          true,
+          false,
           -0.07587041416307178,
           -0.0763945037500033,
-          true,
+          false,
           0.06998058880624704,
           0.07120997975363674,
           true,
@@ -1197,7 +1197,7 @@
           false,
           -0.02358420555142918,
           -0.03949338287409329,
-          false,
+          true,
           0.06034891879883919,
           0.10811110062954676,
           false,
@@ -1209,22 +1209,22 @@
           false,
           0.0025588987543921156,
           -0.03949338287409329,
-          true,
+          false,
           -0.0579830270064572,
           -0.03949338287409329,
-          true,
+          false,
           -0.001568959820211247,
           -0.03949338287409329,
-          true,
+          false,
           -0.05935897986465832,
           0.012906208769698923,
-          false,
+          true,
           -0.037343734133440394,
           -0.011079519799642579,
-          false,
+          true,
           -0.07036660273026729,
           -0.002592261998183278,
-          false,
+          true,
           0.027326050202012293,
           -0.03949338287409329,
           false,
@@ -1233,37 +1233,37 @@
           false,
           -0.0579830270064572,
           -0.03949338287409329,
-          false,
+          true,
           0.001182945896190995,
           0.03430885887772673,
-          false,
+          true,
           -0.001568959820211247,
           -0.03949338287409329,
-          true,
+          false,
           0.010814615903598841,
           -0.03949338287409329,
-          false,
+          true,
           0.024574144485610048,
           0.05091436327188625,
           false,
           -0.060734932722859444,
           -0.0763945037500033,
-          true,
+          false,
           -0.007072771253015731,
           0.03430885887772673,
-          false,
+          true,
           0.00943866304539772,
           -0.002592261998183278,
           false,
           0.039709625925822375,
           0.10811110062954676,
-          true,
+          false,
           0.11951489170148738,
           0.08670845052151895,
-          false,
+          true,
           -0.04422349842444599,
           -0.03949338287409329,
-          true,
+          false,
           0.001182945896190995,
           0.03430885887772673,
           false,
@@ -1272,7 +1272,7 @@
           false,
           0.08374011738825825,
           -0.03949338287409329,
-          true,
+          false,
           -0.009824676969417972,
           -0.002592261998183278,
           false,
@@ -1284,31 +1284,31 @@
           true,
           -0.004320865536613489,
           -0.002592261998183278,
-          true,
+          false,
           -0.015328488402222454,
           -0.002592261998183278,
           false,
           0.01219056876179996,
           0.07120997975363674,
-          false,
+          true,
           0.039709625925822375,
           0.10811110062954676,
-          true,
+          false,
           0.05071724879143135,
           0.03430885887772673,
           true,
           -0.0579830270064572,
           -0.03949338287409329,
-          false,
+          true,
           -0.02083229983502694,
           0.07120997975363674,
           false,
           -0.037343734133440394,
           -0.002592261998183278,
-          true,
+          false,
           -0.0029449126784123676,
           0.07120997975363674,
-          true,
+          false,
           0.03558176735121902,
           -0.0763945037500033,
           true,
@@ -1317,16 +1317,16 @@
           false,
           -0.0318399227006359,
           -0.03949338287409329,
-          false,
+          true,
           -0.033215875558837024,
           -0.002592261998183278,
           false,
           0.017694380194604446,
           0.03430885887772673,
-          false,
+          true,
           -0.016704441260423575,
           -0.002592261998183278,
-          true,
+          false,
           -0.04284754556624487,
           -0.002592261998183278,
           true,
@@ -1338,28 +1338,28 @@
           false,
           0.010814615903598841,
           -0.03949338287409329,
-          false,
+          true,
           0.08374011738825825,
           -0.03949338287409329,
-          true,
+          false,
           -0.04422349842444599,
           -0.03949338287409329,
-          true,
+          false,
           -0.0029449126784123676,
           -0.002592261998183278,
           false,
           0.03145390877661565,
           -0.03949338287409329,
-          false,
+          true,
           -0.06623874415566393,
           -0.03949338287409329,
-          true,
+          false,
           0.001182945896190995,
           -0.007020396503292483,
-          true,
+          false,
           0.01219056876179996,
           -0.03949338287409329,
-          false
+          true
          ]
         },
         {
@@ -2363,7 +2363,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 15,
+   "execution_count": 10,
    "metadata": {},
    "outputs": [
     {
@@ -2487,7 +2487,7 @@
        "[5 rows x 20 columns]"
       ]
      },
-     "execution_count": 15,
+     "execution_count": 10,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -2518,7 +2518,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 16,
+   "execution_count": 11,
    "metadata": {},
    "outputs": [
     {
@@ -3976,7 +3976,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 17,
+   "execution_count": 12,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -3997,15 +3997,15 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 18,
+   "execution_count": 13,
    "metadata": {},
    "outputs": [
     {
      "name": "stderr",
      "output_type": "stream",
      "text": [
-      "[Parallel(n_jobs=1)]: Done   1 tasks      | elapsed:    3.3s\n",
-      "[Parallel(n_jobs=1)]: Done   4 tasks      | elapsed:   12.8s\n"
+      "[Parallel(n_jobs=1)]: Done   1 tasks      | elapsed:    0.1s\n",
+      "[Parallel(n_jobs=1)]: Done   4 tasks      | elapsed:    0.2s\n"
      ]
     },
     {
@@ -4028,7 +4028,7 @@
      "name": "stderr",
      "output_type": "stream",
      "text": [
-      "[Parallel(n_jobs=1)]: Done   5 out of   5 | elapsed:   16.0s finished\n"
+      "[Parallel(n_jobs=1)]: Done   5 out of   5 | elapsed:    0.3s finished\n"
      ]
     }
    ],
@@ -4051,7 +4051,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 19,
+   "execution_count": 14,
    "metadata": {},
    "outputs": [
     {
@@ -4062,8 +4062,32 @@
        },
        "data": [
         {
+         "error_y": {
+          "array": [
+           0,
+           0,
+           0,
+           0,
+           0,
+           0,
+           0,
+           0,
+           0,
+           0,
+           0,
+           0,
+           0,
+           0,
+           0,
+           0,
+           0,
+           0,
+           0
+          ],
+          "type": "data"
+         },
          "marker": {
-          "color": "#00CC96"
+          "color": "#DDCC77"
          },
          "name": "Train",
          "type": "bar",
@@ -4111,8 +4135,32 @@
          ]
         },
         {
+         "error_y": {
+          "array": [
+           0.0023955551235670214,
+           0.002189517306342527,
+           0.001998045402184313,
+           0.0018257279461347798,
+           0.0016784747085066194,
+           0.0015633841084603756,
+           0.0014879382056280817,
+           0.001458303139487877,
+           0.0014772385383559053,
+           0.0015429572455057856,
+           0.0016498782177496946,
+           0.0017906358968372744,
+           0.0019579462215216135,
+           0.0021456065826219173,
+           0.0023487442428032587,
+           0.002563682729688605,
+           0.0027876937544495414,
+           0.003018758272598448,
+           0.003255374672423644
+          ],
+          "type": "data"
+         },
          "marker": {
-          "color": "#AB63FA"
+          "color": "#117733"
          },
          "name": "Test",
          "type": "bar",
@@ -4175,8 +4223,8 @@
          "r": 50,
          "t": 80
         },
-        "paper_bgcolor": "#F0F0F0",
-        "plot_bgcolor": "#F0F0F0",
+        "paper_bgcolor": "#FAFAFA",
+        "plot_bgcolor": "#FAFAFA",
         "template": {
          "data": {
           "bar": [
@@ -4998,14 +5046,22 @@
         },
         "width": 750,
         "xaxis": {
+         "gridcolor": "#E5E5E5",
+         "gridwidth": 1,
+         "linecolor": "#CCCCCC",
          "showgrid": false,
+         "showline": true,
          "title": {
           "text": "Quantile"
          },
          "zeroline": false
         },
         "yaxis": {
-         "showgrid": false,
+         "gridcolor": "#E5E5E5",
+         "gridwidth": 1,
+         "linecolor": "#CCCCCC",
+         "showgrid": true,
+         "showline": true,
          "title": {
           "text": "Average Quantile Loss"
          },
@@ -5022,7 +5078,7 @@
     "# Plot the results for numerical variables\n",
     "if \"quantile_loss\" in matching_results:\n",
     "    perf_results_viz = model_performance_results(\n",
-    "        results=matching_results[\"quantile_loss\"][\"results\"],\n",
+    "        results=matching_results,\n",
     "        model_name=\"Matching\",\n",
     "        method_name=\"Cross-validation quantile loss average\",\n",
     "    )\n",
@@ -5050,7 +5106,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 20,
+   "execution_count": 15,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -5071,7 +5127,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 21,
+   "execution_count": 16,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -5086,7 +5142,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 22,
+   "execution_count": 17,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -5107,14 +5163,14 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 23,
+   "execution_count": 18,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "{'dist_fun': 'Gower', 'constrained': True, 'constr_alg': 'hungarian', 'k': 5}\n"
+      "{'dist_fun': 'Manhattan', 'k': 6}\n"
      ]
     }
    ],
@@ -5133,7 +5189,7 @@
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": "pe3.13",
+   "display_name": "pe3.13 (3.13.0)",
    "language": "python",
    "name": "python3"
   },
diff --git a/docs/models/mdn/index.md b/docs/models/mdn/index.md
index c48797c..4f83612 100644
--- a/docs/models/mdn/index.md
+++ b/docs/models/mdn/index.md
@@ -1,6 +1,6 @@
 # Mixture Density Network
 
-The `MDN` model uses deep neural networks with mixture density outputs to predict missing values by learning complex, potentially multi-modal conditional distributions. Built on PyTorch Tabular, this approach combines the flexibility of neural networks with the probabilistic richness of Gaussian mixture models.
+The `MDN` model uses neural networks with mixture density outputs to predict missing values. Built on PyTorch Tabular, it learns conditional distributions as mixtures of Gaussians, which lets it capture multi-modal relationships.
 
 ## Variable type support
 
@@ -16,7 +16,7 @@ The model supports automatic caching based on data hashes, avoiding redundant re
 
 ## Key features
 
-MDN offers several advantages for complex imputation tasks. The mixture density approach can model multi-modal distributions that simpler methods cannot capture, making it suitable for variables with complex conditional distributions. The neural network backbone can learn non-linear relationships without requiring explicit feature engineering.
+MDN can model multi-modal distributions that simpler methods cannot capture, making it suited for variables with complex conditional distributions. The neural network backbone learns non-linear relationships without explicit feature engineering.
 
 Training leverages GPU acceleration when available and includes early stopping to prevent overfitting. The automatic model caching system speeds up repeated analyses on the same dataset.
 
diff --git a/docs/models/mdn/mdn-imputation.ipynb b/docs/models/mdn/mdn-imputation.ipynb
index 674f716..824d78e 100644
--- a/docs/models/mdn/mdn-imputation.ipynb
+++ b/docs/models/mdn/mdn-imputation.ipynb
@@ -115,16 +115,7 @@
    "execution_count": 1,
    "id": "cell-2",
    "metadata": {},
-   "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Error importing in API mode: ImportError(\"dlopen(/Users/movil1/envs/pe3.13/lib/python3.13/site-packages/_rinterface_cffi_api.abi3.so, 0x0002): Library not loaded: /Library/Frameworks/R.framework/Versions/4.5-arm64/Resources/lib/libRblas.dylib\\n  Referenced from: <21BE8260-F4D5-3597-9DD0-6953BC4DDF3D> /Users/movil1/envs/pe3.13/lib/python3.13/site-packages/_rinterface_cffi_api.abi3.so\\n  Reason: tried: '/Library/Frameworks/R.framework/Versions/4.5-arm64/Resources/lib/libRblas.dylib' (no such file), '/System/Volumes/Preboot/Cryptexes/OS/Library/Frameworks/R.framework/Versions/4.5-arm64/Resources/lib/libRblas.dylib' (no such file), '/Library/Frameworks/R.framework/Versions/4.5-arm64/Resources/lib/libRblas.dylib' (no such file)\")\n",
-      "Trying to import in ABI mode.\n"
-     ]
-    }
-   ],
+   "outputs": [],
    "source": [
     "import warnings\n",
     "warnings.filterwarnings(\"ignore\")\n",
@@ -502,28 +493,28 @@
        "  <tbody>\n",
        "    <tr>\n",
        "      <th>287</th>\n",
-       "      <td>-0.239280</td>\n",
-       "      <td>-0.307822</td>\n",
+       "      <td>0.577170</td>\n",
+       "      <td>0.438805</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>211</th>\n",
-       "      <td>0.339311</td>\n",
-       "      <td>0.358265</td>\n",
+       "      <td>0.481174</td>\n",
+       "      <td>0.488350</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>72</th>\n",
-       "      <td>-0.276636</td>\n",
-       "      <td>-0.277714</td>\n",
+       "      <td>-0.721495</td>\n",
+       "      <td>-0.631198</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>321</th>\n",
-       "      <td>0.540901</td>\n",
-       "      <td>0.551054</td>\n",
+       "      <td>-1.211108</td>\n",
+       "      <td>-1.460021</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>73</th>\n",
-       "      <td>0.460601</td>\n",
-       "      <td>0.446176</td>\n",
+       "      <td>-1.162974</td>\n",
+       "      <td>-1.194782</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
@@ -531,11 +522,11 @@
       ],
       "text/plain": [
        "           s1        s4\n",
-       "287 -0.239280 -0.307822\n",
-       "211  0.339311  0.358265\n",
-       "72  -0.276636 -0.277714\n",
-       "321  0.540901  0.551054\n",
-       "73   0.460601  0.446176"
+       "287  0.577170  0.438805\n",
+       "211  0.481174  0.488350\n",
+       "72  -0.721495 -0.631198\n",
+       "321 -1.211108 -1.460021\n",
+       "73  -1.162974 -1.194782"
       ]
      },
      "execution_count": 6,
@@ -769,184 +760,184 @@
          ],
          "xaxis": "x",
          "y": [
-          0.15996141731739044,
-          0.032140739262104034,
-          0.023270580917596817,
-          -0.07243411242961884,
-          0.06383831799030304,
-          0.0471138134598732,
-          0.08792641758918762,
-          0.183455690741539,
-          0.034883130341768265,
-          -0.020984657108783722,
-          0.1328243613243103,
-          0.002175786066800356,
-          0.8124674558639526,
-          0.9326153993606567,
-          0.1008240208029747,
-          0.06365294754505157,
-          0.3842753469944,
-          0.46691128611564636,
-          0.017988024279475212,
-          -0.016335798427462578,
-          0.06642946600914001,
-          -0.008764325641095638,
-          -0.0118333101272583,
-          0.010635918006300926,
-          -0.012584746815264225,
-          -0.0853499099612236,
-          -0.002956878859549761,
-          -0.024943694472312927,
-          -0.033975280821323395,
-          -0.07795815169811249,
-          -0.07504265010356903,
-          -0.03126005083322525,
-          0.03940160199999809,
-          0.053814880549907684,
-          -0.04383951798081398,
-          0.019542427733540535,
-          -0.09670525044202805,
-          -0.0822581872344017,
-          0.06511250138282776,
-          0.09403425455093384,
-          0.00344745977781713,
-          -0.00915258377790451,
-          0.0950278714299202,
-          -0.048670198768377304,
-          0.039458371698856354,
-          -0.07556705176830292,
-          -0.006443846970796585,
-          0.020815376192331314,
-          -0.036735083907842636,
-          -0.044321704655885696,
-          -0.08580169081687927,
-          -0.059006646275520325,
-          -0.019963562488555908,
-          0.009087007492780685,
-          -0.130485400557518,
-          -0.0006790757179260254,
-          -0.037002887576818466,
-          -0.09501223266124725,
-          0.012499166652560234,
-          0.02030245214700699,
-          0.014375532045960426,
-          -0.02604287676513195,
-          0.057647883892059326,
-          0.05878327041864395,
-          -0.04572222754359245,
-          -0.03820172697305679,
-          -0.06333865225315094,
-          0.04185344651341438,
-          0.05389529839158058,
-          0.01455413457006216,
-          0.09297429770231247,
-          -0.0037352601066231728,
-          -0.014067748561501503,
-          0.026063010096549988,
-          0.007550985552370548,
-          -0.025209808722138405,
-          -0.04939274489879608,
-          -0.030361073091626167,
-          0.03330788016319275,
-          -0.014227045699954033,
-          0.0016633199993520975,
-          -0.08225220441818237,
-          0.010647851973772049,
-          -0.08228402584791183,
-          -0.07383066415786743,
-          0.061680346727371216,
-          -0.5339921116828918,
-          -0.6321333646774292,
-          0.025442680343985558,
-          -0.022219836711883545,
-          -0.026420146226882935,
-          -0.10970157384872437,
-          -0.041401173919439316,
-          -0.0358879491686821,
-          -0.003495529294013977,
-          -0.04557208716869354,
-          0.013635829091072083,
-          -0.043398745357990265,
-          0.026523813605308533,
-          0.00031686751754023135,
-          -0.09354585409164429,
-          -0.05077875778079033,
-          -0.03905845806002617,
-          -0.02609078772366047,
-          -0.006901499815285206,
-          0.05072978138923645,
-          -0.0048773703165352345,
-          -0.01315099187195301,
-          0.083067886531353,
-          0.11018361896276474,
-          -0.042783450335264206,
-          -0.042475901544094086,
-          0.007468030788004398,
-          0.03531891107559204,
-          -0.09278491139411926,
-          -0.008651846088469028,
-          0.03183288499712944,
-          -0.08680566400289536,
-          -0.01793188974261284,
-          -0.08094043284654617,
-          0.08602085709571838,
-          0.06315568834543228,
-          -0.020230960100889206,
-          0.003904549404978752,
-          -0.03103073686361313,
-          -0.06506631523370743,
-          0.036732763051986694,
-          -0.07117997109889984,
-          -0.0006138663738965988,
-          0.03311414271593094,
-          0.06245873123407364,
-          0.002597265876829624,
-          0.01770462840795517,
-          0.058917708694934845,
-          -0.06402486562728882,
-          0.048038020730018616,
-          -0.03813735395669937,
-          -0.028543436899781227,
-          -0.0075743068009614944,
-          0.03675296530127525,
-          -0.014315500855445862,
-          0.040746308863162994,
-          0.07579956203699112,
-          0.1126079186797142,
-          0.08204692602157593,
-          0.07413098216056824,
-          -0.0642552450299263,
-          0.002614614088088274,
-          -0.033612947911024094,
-          -0.015575357712805271,
-          0.08901567757129669,
-          0.12694990634918213,
-          0.12089148908853531,
-          0.0254050400108099,
-          -0.005140587687492371,
-          0.06322228908538818,
-          0.017512474209070206,
-          0.013693263754248619,
-          0.059920668601989746,
-          0.04606965184211731,
-          -0.03190714493393898,
-          -0.01763644628226757,
-          -0.03793055936694145,
-          -0.0012233610032126307,
-          0.012116450816392899,
-          -0.026453981176018715,
-          0.0048493873327970505,
-          -0.008992618881165981,
-          0.05157541483640671,
-          -0.028200700879096985,
-          -0.05151493847370148,
-          -0.05033167079091072,
-          -0.11441226303577423,
-          -0.11985128372907639,
-          -0.07205710560083389,
-          0.0047546266578137875,
-          -0.018449068069458008,
-          -0.006497040390968323
+          0.07409832626581192,
+          -0.0053354548290371895,
+          -0.12105628103017807,
+          -0.014382392168045044,
+          0.04452350735664368,
+          -0.07941135764122009,
+          0.04921489953994751,
+          0.07562752068042755,
+          0.08775658905506134,
+          -0.003818345023319125,
+          0.12863661348819733,
+          0.05178040266036987,
+          0.04000586271286011,
+          -0.16531065106391907,
+          0.06332286447286606,
+          0.08498921245336533,
+          -0.010948514565825462,
+          -0.14337725937366486,
+          0.17172245681285858,
+          -0.07638034224510193,
+          0.033117081969976425,
+          -0.0034198835492134094,
+          -0.007393532898277044,
+          -0.026365965604782104,
+          -0.12046261131763458,
+          -0.027624964714050293,
+          0.029128486290574074,
+          -0.06916160881519318,
+          -0.05742168426513672,
+          -0.10975533723831177,
+          0.020898770540952682,
+          -0.1172960177063942,
+          -0.04515509307384491,
+          0.026763781905174255,
+          0.14253804087638855,
+          0.2255028486251831,
+          -0.24605007469654083,
+          -0.03075171262025833,
+          -0.09652869403362274,
+          0.20666047930717468,
+          -0.014879127964377403,
+          0.024144992232322693,
+          -0.09012438356876373,
+          0.022548293694853783,
+          -0.05704033374786377,
+          -0.11739443987607956,
+          -0.06362107396125793,
+          -0.02206820249557495,
+          -0.03970325365662575,
+          -0.08100323379039764,
+          -0.09121052920818329,
+          -0.01153072714805603,
+          0.09131903946399689,
+          -0.05644859001040459,
+          -0.1356145739555359,
+          0.13108958303928375,
+          -0.08045151084661484,
+          -0.0677538514137268,
+          -0.013112485408782959,
+          0.03593648970127106,
+          -0.09408782422542572,
+          0.10447344928979874,
+          0.026929795742034912,
+          0.004074967000633478,
+          -0.06840845197439194,
+          -0.020132899284362793,
+          -0.03740965574979782,
+          0.01231334824115038,
+          0.13786044716835022,
+          0.061320580542087555,
+          0.03548247739672661,
+          0.15061387419700623,
+          0.0036609352100640535,
+          -0.02198277786374092,
+          0.13655224442481995,
+          -0.08470708131790161,
+          0.0033969569485634565,
+          0.11218808591365814,
+          0.15245385468006134,
+          -0.05473918840289116,
+          -0.01783975213766098,
+          -0.10014862567186356,
+          0.02230358123779297,
+          -0.1551317274570465,
+          -0.021243643015623093,
+          0.024414116516709328,
+          -0.020500551909208298,
+          0.14334836602210999,
+          0.09190063178539276,
+          -0.06325235962867737,
+          0.034143995493650436,
+          -0.010623842477798462,
+          0.017537854611873627,
+          -0.021074065938591957,
+          -0.11523549258708954,
+          -0.025434572249650955,
+          0.018908560276031494,
+          0.012480437755584717,
+          0.02391844429075718,
+          0.04059349000453949,
+          -0.04446566849946976,
+          -0.07962190359830856,
+          -0.01866019144654274,
+          -0.09491998702287674,
+          -0.10590796172618866,
+          0.11002372950315475,
+          -0.045452192425727844,
+          0.048709649592638016,
+          0.16254602372646332,
+          -0.23081794381141663,
+          0.033685214817523956,
+          -0.0006430193898268044,
+          -0.015713322907686234,
+          0.026455357670783997,
+          0.06460753083229065,
+          0.09441250562667847,
+          -0.008670782670378685,
+          -0.08653200417757034,
+          -0.07236170768737793,
+          -0.09426410496234894,
+          0.14369378983974457,
+          -0.03581035137176514,
+          0.008754465728998184,
+          0.046415455639362335,
+          0.03268638253211975,
+          0.06044789403676987,
+          0.08978655189275742,
+          -0.04920205473899841,
+          -0.1123436689376831,
+          0.0034600854851305485,
+          0.02713024616241455,
+          0.003619933035224676,
+          -0.02590913698077202,
+          -0.09626477211713791,
+          -0.026048922911286354,
+          0.01121927797794342,
+          -0.0919136255979538,
+          -0.04032476991415024,
+          -0.04339555650949478,
+          0.041690241545438766,
+          0.09366445243358612,
+          0.16090711951255798,
+          0.12886887788772583,
+          0.024945732206106186,
+          0.10310709476470947,
+          0.09357581287622452,
+          -0.024438243359327316,
+          -0.02837105467915535,
+          -0.10025747120380402,
+          -0.15249663591384888,
+          0.045413028448820114,
+          0.16909754276275635,
+          0.15786541998386383,
+          0.13495682179927826,
+          -0.02144046686589718,
+          0.08984969556331635,
+          -0.04134964570403099,
+          -0.051257528364658356,
+          0.045112937688827515,
+          0.047064878046512604,
+          0.034162651747465134,
+          0.021895015612244606,
+          -0.05455716326832771,
+          -0.0889664888381958,
+          0.011050431989133358,
+          -0.014817613177001476,
+          0.163625568151474,
+          0.03653216361999512,
+          -0.06846585869789124,
+          -0.08761408925056458,
+          -0.08431434631347656,
+          -0.06531454622745514,
+          -0.01888906955718994,
+          -0.09313887357711792,
+          -0.10920677334070206,
+          -0.05248451232910156,
+          0.001476499019190669,
+          -0.011065206490457058
          ],
          "yaxis": "y"
         },
@@ -959,12 +950,12 @@
          "name": "Perfect prediction",
          "type": "scatter",
          "x": [
-          -1.969916820526123,
-          1.6161797046661377
+          -2.2862510681152344,
+          2.5850915908813477
          ],
          "y": [
-          -1.969916820526123,
-          1.6161797046661377
+          -2.2862510681152344,
+          2.5850915908813477
          ]
         }
        ],
@@ -1942,47 +1933,47 @@
        "    <tr>\n",
        "      <th>0</th>\n",
        "      <td>0.125019</td>\n",
-       "      <td>-1.208731</td>\n",
-       "      <td>-0.793742</td>\n",
-       "      <td>-0.239280</td>\n",
-       "      <td>0.260678</td>\n",
-       "      <td>0.750109</td>\n",
+       "      <td>-1.071396</td>\n",
+       "      <td>-0.291549</td>\n",
+       "      <td>0.577170</td>\n",
+       "      <td>1.379413</td>\n",
+       "      <td>2.154069</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>1</th>\n",
        "      <td>-0.024960</td>\n",
-       "      <td>-0.430616</td>\n",
-       "      <td>-0.080356</td>\n",
-       "      <td>0.339311</td>\n",
-       "      <td>0.740931</td>\n",
-       "      <td>1.117335</td>\n",
+       "      <td>-1.594807</td>\n",
+       "      <td>-0.647442</td>\n",
+       "      <td>0.481174</td>\n",
+       "      <td>1.623415</td>\n",
+       "      <td>2.562861</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>2</th>\n",
        "      <td>0.103003</td>\n",
-       "      <td>-1.449445</td>\n",
-       "      <td>-0.907212</td>\n",
-       "      <td>-0.276636</td>\n",
-       "      <td>0.273373</td>\n",
-       "      <td>0.844860</td>\n",
+       "      <td>-2.563489</td>\n",
+       "      <td>-1.645817</td>\n",
+       "      <td>-0.721495</td>\n",
+       "      <td>0.225897</td>\n",
+       "      <td>1.118832</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>3</th>\n",
        "      <td>0.054845</td>\n",
-       "      <td>-0.039490</td>\n",
-       "      <td>0.249726</td>\n",
-       "      <td>0.540901</td>\n",
-       "      <td>0.896468</td>\n",
-       "      <td>1.195148</td>\n",
+       "      <td>-3.448399</td>\n",
+       "      <td>-2.342187</td>\n",
+       "      <td>-1.211108</td>\n",
+       "      <td>0.049215</td>\n",
+       "      <td>1.042724</td>\n",
        "    </tr>\n",
        "    <tr>\n",
        "      <th>4</th>\n",
        "      <td>0.038334</td>\n",
-       "      <td>-0.671013</td>\n",
-       "      <td>-0.092434</td>\n",
-       "      <td>0.460601</td>\n",
-       "      <td>1.070943</td>\n",
-       "      <td>1.577428</td>\n",
+       "      <td>-3.044891</td>\n",
+       "      <td>-2.123165</td>\n",
+       "      <td>-1.162974</td>\n",
+       "      <td>-0.149736</td>\n",
+       "      <td>0.683350</td>\n",
        "    </tr>\n",
        "  </tbody>\n",
        "</table>\n",
@@ -1990,11 +1981,11 @@
       ],
       "text/plain": [
        "     Actual       Q10       Q25       Q50       Q75       Q90\n",
-       "0  0.125019 -1.208731 -0.793742 -0.239280  0.260678  0.750109\n",
-       "1 -0.024960 -0.430616 -0.080356  0.339311  0.740931  1.117335\n",
-       "2  0.103003 -1.449445 -0.907212 -0.276636  0.273373  0.844860\n",
-       "3  0.054845 -0.039490  0.249726  0.540901  0.896468  1.195148\n",
-       "4  0.038334 -0.671013 -0.092434  0.460601  1.070943  1.577428"
+       "0  0.125019 -1.071396 -0.291549  0.577170  1.379413  2.154069\n",
+       "1 -0.024960 -1.594807 -0.647442  0.481174  1.623415  2.562861\n",
+       "2  0.103003 -2.563489 -1.645817 -0.721495  0.225897  1.118832\n",
+       "3  0.054845 -3.448399 -2.342187 -1.211108  0.049215  1.042724\n",
+       "4  0.038334 -3.044891 -2.123165 -1.162974 -0.149736  0.683350"
       ]
      },
      "execution_count": 8,
@@ -2054,8 +2045,8 @@
           0
          ],
          "y": [
-          -1.208730936050415,
-          0.7501086592674255
+          -1.0713963508605957,
+          2.154068946838379
          ]
         },
         {
@@ -2072,8 +2063,8 @@
           1
          ],
          "y": [
-          -0.4306159019470215,
-          1.1173346042633057
+          -1.5948072671890259,
+          2.5628609657287598
          ]
         },
         {
@@ -2090,8 +2081,8 @@
           2
          ],
          "y": [
-          -1.4494445323944092,
-          0.8448596596717834
+          -2.5634894371032715,
+          1.118831753730774
          ]
         },
         {
@@ -2108,8 +2099,8 @@
           3
          ],
          "y": [
-          -0.03949018567800522,
-          1.1951476335525513
+          -3.448399305343628,
+          1.0427237749099731
          ]
         },
         {
@@ -2126,8 +2117,8 @@
           4
          ],
          "y": [
-          -0.6710134148597717,
-          1.577427625656128
+          -3.044891357421875,
+          0.6833502650260925
          ]
         },
         {
@@ -2144,8 +2135,8 @@
           5
          ],
          "y": [
-          -1.388054609298706,
-          0.7144660353660583
+          -1.1571630239486694,
+          1.9580988883972168
          ]
         },
         {
@@ -2162,8 +2153,8 @@
           6
          ],
          "y": [
-          0.9335368275642395,
-          1.994225263595581
+          -0.448160856962204,
+          5.394647121429443
          ]
         },
         {
@@ -2180,8 +2171,8 @@
           7
          ],
          "y": [
-          -1.2012914419174194,
-          0.6220993399620056
+          -3.3663272857666016,
+          1.5296334028244019
          ]
         },
         {
@@ -2198,8 +2189,8 @@
           8
          ],
          "y": [
-          0.4870220720767975,
-          1.1506199836730957
+          -0.9083917140960693,
+          4.910729885101318
          ]
         },
         {
@@ -2216,8 +2207,8 @@
           9
          ],
          "y": [
-          -1.1105694770812988,
-          0.017988024279475212
+          0.17172245681285858,
+          3.2345242500305176
          ]
         },
         {
@@ -2234,8 +2225,8 @@
           0
          ],
          "y": [
-          -0.6731794476509094,
-          0.15996141731739044
+          -0.11874838173389435,
+          1.195131540298462
          ]
         },
         {
@@ -2252,8 +2243,8 @@
           1
          ],
          "y": [
-          0.023270580917596817,
-          0.658467710018158
+          -0.40085601806640625,
+          1.4143391847610474
          ]
         },
         {
@@ -2270,8 +2261,8 @@
           2
          ],
          "y": [
-          -0.769422709941864,
-          0.17357832193374634
+          -1.4394127130508423,
+          0.04452350735664368
          ]
         },
         {
@@ -2288,8 +2279,8 @@
           3
          ],
          "y": [
-          0.3068358898162842,
-          0.8011546730995178
+          -2.0836896896362305,
+          -0.2426186203956604
          ]
         },
         {
@@ -2306,8 +2297,8 @@
           4
          ],
          "y": [
-          0.034883130341768265,
-          0.9577965140342712
+          -1.900681734085083,
+          -0.4198998510837555
          ]
         },
         {
@@ -2324,8 +2315,8 @@
           5
          ],
          "y": [
-          -0.7857669591903687,
-          0.1328243613243103
+          -0.14701643586158752,
+          1.0303322076797485
          ]
         },
         {
@@ -2342,8 +2333,8 @@
           6
          ],
          "y": [
-          1.253185749053955,
-          1.6908645629882812
+          1.3525761365890503,
+          3.828951597213745
          ]
         },
         {
@@ -2360,8 +2351,8 @@
           7
          ],
          "y": [
-          -0.629646360874176,
-          0.1008240208029747
+          -1.7197819948196411,
+          0.2868454158306122
          ]
         },
         {
@@ -2378,8 +2369,8 @@
           8
          ],
          "y": [
-          0.6771109104156494,
-          0.9602589011192322
+          0.6424643397331238,
+          3.1236233711242676
          ]
         },
         {
@@ -2396,8 +2387,8 @@
           9
          ],
          "y": [
-          -0.7771245241165161,
-          -0.29959893226623535
+          1.113405704498291,
+          2.3700172901153564
          ]
         },
         {
@@ -2454,16 +2445,16 @@
           9
          ],
          "y": [
-          -0.23928046226501465,
-          0.339310884475708,
-          -0.276636004447937,
-          0.540900707244873,
-          0.4606007933616638,
-          -0.34471726417541504,
-          1.4770541191101074,
-          -0.30302998423576355,
-          0.8187563419342041,
-          -0.5293239951133728
+          0.5771697759628296,
+          0.4811735153198242,
+          -0.7214951515197754,
+          -1.211107611656189,
+          -1.1629741191864014,
+          0.41260623931884766,
+          2.5850915908813477,
+          -0.6760748624801636,
+          1.8518712520599365,
+          1.7761045694351196
          ]
         },
         {
@@ -3472,9 +3463,9 @@
      "output_type": "stream",
      "text": [
       "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 8 concurrent workers.\n",
-      "[Parallel(n_jobs=-1)]: Done   2 out of   5 | elapsed:    7.6s remaining:   11.5s\n",
-      "[Parallel(n_jobs=-1)]: Done   3 out of   5 | elapsed:    7.6s remaining:    5.1s\n",
-      "[Parallel(n_jobs=-1)]: Done   5 out of   5 | elapsed:    7.7s finished\n"
+      "[Parallel(n_jobs=-1)]: Done   2 out of   5 | elapsed:    9.5s remaining:   14.3s\n",
+      "[Parallel(n_jobs=-1)]: Done   3 out of   5 | elapsed:    9.5s remaining:    6.4s\n",
+      "[Parallel(n_jobs=-1)]: Done   5 out of   5 | elapsed:    9.5s finished\n"
      ]
     },
     {
@@ -3483,8 +3474,8 @@
      "text": [
       "Quantile loss results:\n",
       "           0.05      0.10      0.15      0.20      0.25  ...      0.75      0.80      0.85      0.90      0.95\n",
-      "train  0.052342  0.080491  0.100151  0.114462  0.124877  ...  0.116854  0.105562  0.090379  0.070140  0.043119\n",
-      "test   0.060333  0.088338  0.109102  0.123207  0.134303  ...  0.127896  0.114887  0.099006  0.077486  0.047284\n",
+      "train  0.046876  0.074835  0.094329  0.108058  0.117289  ...  0.125522  0.116335  0.102234  0.081915  0.052980\n",
+      "test   0.048934  0.078406  0.099249  0.117245  0.130767  ...  0.133579  0.123578  0.109620  0.088337  0.058313\n",
       "\n",
       "[2 rows x 19 columns]\n"
      ]
@@ -3516,8 +3507,32 @@
        },
        "data": [
         {
+         "error_y": {
+          "array": [
+           0.03837709044008565,
+           0.05971215726699471,
+           0.07364494245759541,
+           0.08337091425123092,
+           0.08970334204623862,
+           0.09257452417327834,
+           0.09425214250614794,
+           0.09727578509414461,
+           0.1015340536222146,
+           0.10484112107099822,
+           0.10665196062518409,
+           0.10767398977417002,
+           0.10657423431327813,
+           0.10291839541792067,
+           0.09739380134017506,
+           0.08998945316371061,
+           0.079000703908655,
+           0.0623493126962366,
+           0.03897342645724234
+          ],
+          "type": "data"
+         },
          "marker": {
-          "color": "#00CC96"
+          "color": "#DDCC77"
          },
          "name": "Train",
          "type": "bar",
@@ -3543,30 +3558,54 @@
           "0.95"
          ],
          "y": [
-          0.05234219330354317,
-          0.08049099191203442,
-          0.10015070706030556,
-          0.11446162331927275,
-          0.12487722321825685,
-          0.1324132843109902,
-          0.13768463404454853,
-          0.14056006658781792,
-          0.14185292803744393,
-          0.1415992977069213,
-          0.14005457428331253,
-          0.13682163676619796,
-          0.13220755848574667,
-          0.12584912155335878,
-          0.11685420327991669,
-          0.10556158423365573,
-          0.09037880944328178,
-          0.07014017488138428,
-          0.04311920349617806
+          0.04687617706207272,
+          0.07483489712462413,
+          0.0943290184639551,
+          0.10805756595645155,
+          0.11728923842580923,
+          0.12286641601968225,
+          0.12616856929600748,
+          0.1292466478910964,
+          0.13271784812663345,
+          0.1354354106464196,
+          0.13713837435319742,
+          0.13782672818961056,
+          0.13631602177577212,
+          0.13210069705430905,
+          0.12552150971602297,
+          0.11633498833082316,
+          0.10223421559735249,
+          0.08191549168418728,
+          0.0529798258038449
          ]
         },
         {
+         "error_y": {
+          "array": [
+           0.03618523330977212,
+           0.05800221625311354,
+           0.07291441840284316,
+           0.08871777392907335,
+           0.10179376511491461,
+           0.11098038464484912,
+           0.11340640962306367,
+           0.11611839981806513,
+           0.11748612497231077,
+           0.1174333555879695,
+           0.11549620009870612,
+           0.1127200411691131,
+           0.10807914038546579,
+           0.10210520685018526,
+           0.0948305718757984,
+           0.08725433626495321,
+           0.0765462308033389,
+           0.060192339856620924,
+           0.03773031440262491
+          ],
+          "type": "data"
+         },
          "marker": {
-          "color": "#AB63FA"
+          "color": "#117733"
          },
          "name": "Test",
          "type": "bar",
@@ -3592,25 +3631,25 @@
           "0.95"
          ],
          "y": [
-          0.060333009487974085,
-          0.08833805422856528,
-          0.10910236343419596,
-          0.12320670350109023,
-          0.13430269442274195,
-          0.14292452496806163,
-          0.14938276914149362,
-          0.15449110540720765,
-          0.1571806630200133,
-          0.1578150543631282,
-          0.15689612315245363,
-          0.15328965501428699,
-          0.14702280821997657,
-          0.13845547628705684,
-          0.12789622672114154,
-          0.11488700831441748,
-          0.099006403363506,
-          0.07748605801852056,
-          0.047284249954029064
+          0.04893408004321757,
+          0.07840558199482109,
+          0.09924928077798252,
+          0.11724509694752339,
+          0.13076715505019737,
+          0.14024932427368472,
+          0.14459225917030372,
+          0.1486228918064114,
+          0.15166062433499797,
+          0.1528524649168001,
+          0.15231682792337978,
+          0.15052964914121986,
+          0.14697318308426555,
+          0.14127787715452564,
+          0.1335791298846442,
+          0.1235777656562175,
+          0.10962020934072403,
+          0.08833690511355873,
+          0.05831330288784895
          ]
         }
        ],
@@ -3629,8 +3668,8 @@
          "r": 50,
          "t": 80
         },
-        "paper_bgcolor": "#F0F0F0",
-        "plot_bgcolor": "#F0F0F0",
+        "paper_bgcolor": "#FAFAFA",
+        "plot_bgcolor": "#FAFAFA",
         "template": {
          "data": {
           "bar": [
@@ -4452,14 +4491,22 @@
         },
         "width": 750,
         "xaxis": {
+         "gridcolor": "#E5E5E5",
+         "gridwidth": 1,
+         "linecolor": "#CCCCCC",
          "showgrid": false,
+         "showline": true,
          "title": {
           "text": "Quantile"
          },
          "zeroline": false
         },
         "yaxis": {
-         "showgrid": false,
+         "gridcolor": "#E5E5E5",
+         "gridwidth": 1,
+         "linecolor": "#CCCCCC",
+         "showgrid": true,
+         "showline": true,
          "title": {
           "text": "Average Quantile Loss"
          },
@@ -4476,7 +4523,7 @@
     "# Plot the results\n",
     "if \"quantile_loss\" in mdn_results:\n",
     "    perf_results_viz = model_performance_results(\n",
-    "        results=mdn_results[\"quantile_loss\"][\"results\"],\n",
+    "        results=mdn_results,\n",
     "        model_name=\"MDN\",\n",
     "        method_name=\"Cross-validation quantile loss average\",\n",
     "    )\n",
@@ -4531,9 +4578,9 @@
      "text": [
       "Categorical variable distribution:\n",
       "risk_level\n",
+      "medium    150\n",
       "low       148\n",
-      "high      148\n",
-      "medium    146\n",
+      "high      144\n",
       "Name: count, dtype: int64\n",
       "\n",
       "Training set size: 353 records\n",
@@ -4607,13 +4654,13 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "Categorical imputation accuracy: 31.46%\n",
+      "Categorical imputation accuracy: 30.34%\n",
       "\n",
       "Confusion matrix:\n",
       "                Predicted: low  Predicted: medium  Predicted: high\n",
-      "Actual: low                 10                  8               12\n",
-      "Actual: medium              14                 11                4\n",
-      "Actual: high                 7                 16                7\n"
+      "Actual: low                  8                  5               17\n",
+      "Actual: medium              11                  8               10\n",
+      "Actual: high                 7                 12               11\n"
      ]
     }
    ],
@@ -4647,9 +4694,9 @@
      "output_type": "stream",
      "text": [
       "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 8 concurrent workers.\n",
-      "[Parallel(n_jobs=-1)]: Done   2 out of   5 | elapsed:    0.8s remaining:    1.2s\n",
-      "[Parallel(n_jobs=-1)]: Done   3 out of   5 | elapsed:    0.9s remaining:    0.6s\n",
-      "[Parallel(n_jobs=-1)]: Done   5 out of   5 | elapsed:    0.9s finished\n"
+      "[Parallel(n_jobs=-1)]: Done   2 out of   5 | elapsed:    1.2s remaining:    1.9s\n",
+      "[Parallel(n_jobs=-1)]: Done   3 out of   5 | elapsed:    1.3s remaining:    0.9s\n",
+      "[Parallel(n_jobs=-1)]: Done   5 out of   5 | elapsed:    1.4s finished\n"
      ]
     },
     {
@@ -4657,8 +4704,8 @@
      "output_type": "stream",
      "text": [
       "Categorical imputation cross-validation results (log loss):\n",
-      "Mean train log loss: 0.9622\n",
-      "Mean test log loss: 1.0936\n"
+      "Mean train log loss: 0.9076\n",
+      "Mean test log loss: 1.0931\n"
      ]
     }
    ],
@@ -4688,8 +4735,14 @@
        },
        "data": [
         {
+         "error_y": {
+          "array": [
+           0.027550481167352696
+          ],
+          "type": "data"
+         },
          "marker": {
-          "color": "#00CC96"
+          "color": "#DDCC77"
          },
          "name": "Train",
          "showlegend": true,
@@ -4699,13 +4752,19 @@
          ],
          "xaxis": "x",
          "y": [
-          0.9622434565816542
+          0.9075664863898126
          ],
          "yaxis": "y"
         },
         {
+         "error_y": {
+          "array": [
+           0.0913599760690238
+          ],
+          "type": "data"
+         },
          "marker": {
-          "color": "#AB63FA"
+          "color": "#117733"
          },
          "name": "Test",
          "showlegend": true,
@@ -4715,7 +4774,7 @@
          ],
          "xaxis": "x",
          "y": [
-          1.093641015261423
+          1.093074681095708
          ],
          "yaxis": "y"
         }
@@ -4737,8 +4796,8 @@
          }
         ],
         "height": 420,
-        "paper_bgcolor": "#F0F0F0",
-        "plot_bgcolor": "#F0F0F0",
+        "paper_bgcolor": "#FAFAFA",
+        "plot_bgcolor": "#FAFAFA",
         "showlegend": true,
         "template": {
          "data": {
@@ -5565,14 +5624,24 @@
          "domain": [
           0,
           1
-         ]
+         ],
+         "gridcolor": "#E5E5E5",
+         "gridwidth": 1,
+         "linecolor": "#CCCCCC",
+         "showgrid": false,
+         "showline": true
         },
         "yaxis": {
          "anchor": "x",
          "domain": [
           0,
           1
-         ]
+         ],
+         "gridcolor": "#E5E5E5",
+         "gridwidth": 1,
+         "linecolor": "#CCCCCC",
+         "showgrid": true,
+         "showline": true
         }
        }
       }
@@ -5608,7 +5677,7 @@
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": "pe3.13",
+   "display_name": "pe3.13 (3.13.0)",
    "language": "python",
    "name": "python3"
   },
diff --git a/docs/models/ols/index.md b/docs/models/ols/index.md
index 5312939..13ef8bd 100644
--- a/docs/models/ols/index.md
+++ b/docs/models/ols/index.md
@@ -1,17 +1,17 @@
 # Ordinary Least Squares
 
-The `OLS` model employs linear regression techniques to predict missing values by leveraging the relationships between predictor and target variables. This classic statistical approach provides a computationally efficient method for imputation while offering theoretical guarantees under certain assumptions.
+The `OLS` model uses linear regression to predict missing values from the relationships between predictor and target variables. It is computationally fast and provides a useful baseline for comparison with more complex methods.
 
 ## Variable type support
 
-OLS automatically adapts to your target variable types. For numerical variables, it uses standard linear regression. For categorical variables (including strings, booleans, or numerically-encoded categorical variables), it automatically switches to logistic regression classification. This automatic detection means you don't need to specify variable types—simply pass your predictors and targets, and the model handles the rest internally.
+OLS adapts to target variable types automatically. For numerical variables, it uses standard linear regression. For categorical variables (including strings, booleans, or numerically-encoded categorical variables), it switches to logistic regression. You don't need to specify variable types manually.
 
 ## How it works
 
-The OLS imputer works by fitting a linear regression model using the statsmodels implementation of Ordinary Least Squares. During the training phase, it identifies the coefficients that minimize the sum of squared residuals between the predicted and actual values in the training data. This creates a model that captures the linear relationship between the predictors and target variables.
+The OLS imputer fits a linear regression model using the statsmodels implementation. During training, it finds the coefficients that minimize the sum of squared residuals between predicted and actual values.
 
-For prediction at different quantiles, the model makes an important assumption that the residuals (the differences between predicted and actual values) follow a normal distribution. This assumption allows the model to generate predictions at various quantiles by starting with the mean prediction and adding a quantile-specific offset derived from the normal distribution. Specifically, it computes the standard error of the predictions and applies the inverse normal cumulative distribution function to generate predictions at the requested quantiles.
+For prediction at different quantiles, the model assumes normally distributed residuals. It starts with the mean prediction and adds a quantile-specific offset computed from the normal distribution's inverse CDF and the standard error of the predictions.
 
 ## Key features
 
-The OLS imputer offers a simple yet powerful parametric approach with fast training and prediction times compared to more complex models. It relies on the assumption of linear relationships between variables, making it particularly suitable for datasets where such relationships hold or as a baseline comparison for more complex approaches.
+OLS is fast to train and predict. It works well when the relationship between predictors and targets is approximately linear. Because it assumes constant variance and normally distributed errors, it tends to compress imputed values toward the mean, producing a narrower distribution than the true one. This makes it a good baseline but a poor choice when the data has heavy tails or heteroscedastic errors.
diff --git a/docs/models/ols/ols-imputation.ipynb b/docs/models/ols/ols-imputation.ipynb
index 0502f68..f09b1d9 100644
--- a/docs/models/ols/ols-imputation.ipynb
+++ b/docs/models/ols/ols-imputation.ipynb
@@ -71,7 +71,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 1,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -102,7 +102,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 56,
+   "execution_count": 2,
    "metadata": {},
    "outputs": [
     {
@@ -217,7 +217,7 @@
        "4  0.005383 -0.044642 -0.036385  0.021872  0.003935  0.015596  0.008142 -0.002592 -0.031988 -0.046641"
       ]
      },
-     "execution_count": 56,
+     "execution_count": 2,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -233,7 +233,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 57,
+   "execution_count": 3,
    "metadata": {},
    "outputs": [
     {
@@ -354,7 +354,7 @@
        "max    1.107267e-01  5.068012e-02  1.705552e-01  1.320436e-01  1.539137e-01  1.852344e-01"
       ]
      },
-     "execution_count": 57,
+     "execution_count": 3,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -376,7 +376,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 58,
+   "execution_count": 4,
    "metadata": {},
    "outputs": [
     {
@@ -408,7 +408,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 59,
+   "execution_count": 5,
    "metadata": {},
    "outputs": [
     {
@@ -499,7 +499,7 @@
        "73   0.012648  0.050680 -0.020218 -0.002228 NaN NaN"
       ]
      },
-     "execution_count": 59,
+     "execution_count": 5,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -528,7 +528,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 60,
+   "execution_count": 6,
    "metadata": {},
    "outputs": [
     {
@@ -547,7 +547,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 61,
+   "execution_count": 7,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -561,7 +561,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 62,
+   "execution_count": 8,
    "metadata": {},
    "outputs": [
     {
@@ -628,7 +628,7 @@
        "73  -0.004692  0.007055"
       ]
      },
-     "execution_count": 62,
+     "execution_count": 8,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -653,7 +653,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 63,
+   "execution_count": 9,
    "metadata": {},
    "outputs": [
     {
@@ -2004,7 +2004,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 64,
+   "execution_count": 10,
    "metadata": {},
    "outputs": [
     {
@@ -2128,7 +2128,7 @@
        "[5 rows x 20 columns]"
       ]
      },
-     "execution_count": 64,
+     "execution_count": 10,
      "metadata": {},
      "output_type": "execute_result"
     }
@@ -2159,7 +2159,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 65,
+   "execution_count": 11,
    "metadata": {},
    "outputs": [
     {
@@ -3617,14 +3617,17 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 66,
+   "execution_count": 12,
    "metadata": {},
    "outputs": [
     {
      "name": "stderr",
      "output_type": "stream",
      "text": [
-      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 8 concurrent workers.\n"
+      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 8 concurrent workers.\n",
+      "[Parallel(n_jobs=-1)]: Done   2 out of   5 | elapsed:   13.5s remaining:   20.2s\n",
+      "[Parallel(n_jobs=-1)]: Done   3 out of   5 | elapsed:   13.5s remaining:    9.0s\n",
+      "[Parallel(n_jobs=-1)]: Done   5 out of   5 | elapsed:   13.6s finished\n"
      ]
     },
     {
@@ -3657,7 +3660,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 67,
+   "execution_count": 13,
    "metadata": {},
    "outputs": [
     {
@@ -3668,8 +3671,32 @@
        },
        "data": [
         {
+         "error_y": {
+          "array": [
+           0.00007176011467255402,
+           0.0000913355164601245,
+           0.00011083575969884443,
+           0.00011210242100357653,
+           0.00011563162775412741,
+           0.00013000531898907708,
+           0.000143271836739722,
+           0.0001506544338309766,
+           0.00015261535043567369,
+           0.00014081751874840215,
+           0.00013834799510673368,
+           0.0001357194581903946,
+           0.00012375064613863142,
+           0.00011764060922499956,
+           0.00010758816302847906,
+           0.00009853561930363294,
+           0.00010941603098900106,
+           0.00011884481777470878,
+           0.00009323804980730522
+          ],
+          "type": "data"
+         },
          "marker": {
-          "color": "#00CC96"
+          "color": "#DDCC77"
          },
          "name": "Train",
          "type": "bar",
@@ -3717,8 +3744,32 @@
          ]
         },
         {
+         "error_y": {
+          "array": [
+           0.0002511960288130229,
+           0.0003728781059406671,
+           0.0005215834564121717,
+           0.000522694690812438,
+           0.0005201751472757763,
+           0.0005498446940786355,
+           0.000608514199932397,
+           0.0006423333112739606,
+           0.0006362169672265104,
+           0.0005805283237282669,
+           0.0005503271496842148,
+           0.0005306193310896558,
+           0.0005077865393829969,
+           0.000485382018208908,
+           0.00043665631509002455,
+           0.00036712860842352926,
+           0.0004221172731648885,
+           0.0005054827478209771,
+           0.00043716321985420154
+          ],
+          "type": "data"
+         },
          "marker": {
-          "color": "#AB63FA"
+          "color": "#117733"
          },
          "name": "Test",
          "type": "bar",
@@ -3781,8 +3832,8 @@
          "r": 50,
          "t": 80
         },
-        "paper_bgcolor": "#F0F0F0",
-        "plot_bgcolor": "#F0F0F0",
+        "paper_bgcolor": "#FAFAFA",
+        "plot_bgcolor": "#FAFAFA",
         "template": {
          "data": {
           "bar": [
@@ -4604,14 +4655,22 @@
         },
         "width": 750,
         "xaxis": {
+         "gridcolor": "#E5E5E5",
+         "gridwidth": 1,
+         "linecolor": "#CCCCCC",
          "showgrid": false,
+         "showline": true,
          "title": {
           "text": "Quantile"
          },
          "zeroline": false
         },
         "yaxis": {
-         "showgrid": false,
+         "gridcolor": "#E5E5E5",
+         "gridwidth": 1,
+         "linecolor": "#CCCCCC",
+         "showgrid": true,
+         "showline": true,
          "title": {
           "text": "Average Quantile Loss"
          },
@@ -4628,7 +4687,7 @@
     "# Plot the results for numerical variables\n",
     "if \"quantile_loss\" in ols_results:\n",
     "    perf_results_viz = model_performance_results(\n",
-    "        results=ols_results[\"quantile_loss\"][\"results\"],\n",
+    "        results=ols_results,\n",
     "        model_name=\"OLS\",\n",
     "        method_name=\"Cross-validation quantile loss average\",\n",
     "    )\n",
@@ -4649,7 +4708,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 68,
+   "execution_count": 14,
    "metadata": {},
    "outputs": [
     {
@@ -4658,19 +4717,19 @@
      "text": [
       "Categorical variable distribution:\n",
       "risk_level\n",
+      "medium    150\n",
       "low       148\n",
-      "high      148\n",
-      "medium    146\n",
+      "high      144\n",
       "Name: count, dtype: int64\n",
       "\n",
       "Percentage distribution:\n",
       "risk_level\n",
+      "medium    0.339367\n",
       "low       0.334842\n",
-      "high      0.334842\n",
-      "medium    0.330317\n",
+      "high      0.325792\n",
       "Name: proportion, dtype: float64\n",
       "\n",
-      "Data types: {'age': dtype('float64'), 'sex': dtype('float64'), 'bmi': dtype('float64'), 'bp': dtype('float64'), 'risk_level': dtype('O')}\n",
+      "Data types: {'age': dtype('float64'), 'sex': dtype('float64'), 'bmi': dtype('float64'), 'bp': dtype('float64'), 'risk_level': <StringDtype(na_value=nan)>}\n",
       "\n",
       "Training set size: 353 records\n",
       "Testing set size: 89 records\n"
@@ -4707,7 +4766,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 69,
+   "execution_count": 15,
    "metadata": {},
    "outputs": [
     {
@@ -4750,20 +4809,20 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 70,
+   "execution_count": 16,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "Categorical imputation accuracy: 40.45%\n",
+      "Categorical imputation accuracy: 37.08%\n",
       "\n",
       "Confusion matrix:\n",
       "                Predicted: low  Predicted: medium  Predicted: high\n",
-      "Actual: low                 15                 11                4\n",
-      "Actual: medium               9                 20                0\n",
-      "Actual: high                10                 19                1\n"
+      "Actual: low                  9                 10               11\n",
+      "Actual: medium               4                 20                5\n",
+      "Actual: high                 6                 20                4\n"
      ]
     }
    ],
@@ -4791,14 +4850,17 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 71,
+   "execution_count": 17,
    "metadata": {},
    "outputs": [
     {
      "name": "stderr",
      "output_type": "stream",
      "text": [
-      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 8 concurrent workers.\n"
+      "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 8 concurrent workers.\n",
+      "[Parallel(n_jobs=-1)]: Done   2 out of   5 | elapsed:    1.1s remaining:    1.6s\n",
+      "[Parallel(n_jobs=-1)]: Done   3 out of   5 | elapsed:    1.2s remaining:    0.8s\n",
+      "[Parallel(n_jobs=-1)]: Done   5 out of   5 | elapsed:    1.6s finished\n"
      ]
     },
     {
@@ -4806,8 +4868,8 @@
      "output_type": "stream",
      "text": [
       "Categorical imputation cross-validation results (log loss):\n",
-      "Mean train log loss: 1.0673\n",
-      "Mean test log loss: 1.0776\n"
+      "Mean train log loss: 1.0681\n",
+      "Mean test log loss: 1.0794\n"
      ]
     }
    ],
@@ -4828,7 +4890,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 72,
+   "execution_count": 18,
    "metadata": {},
    "outputs": [
     {
@@ -4839,8 +4901,14 @@
        },
        "data": [
         {
+         "error_y": {
+          "array": [
+           0.0019313286420746783
+          ],
+          "type": "data"
+         },
          "marker": {
-          "color": "#00CC96"
+          "color": "#DDCC77"
          },
          "name": "Train",
          "showlegend": true,
@@ -4850,13 +4918,19 @@
          ],
          "xaxis": "x",
          "y": [
-          1.0673353634599498
+          1.0681169294698512
          ],
          "yaxis": "y"
         },
         {
+         "error_y": {
+          "array": [
+           0.011486633730960285
+          ],
+          "type": "data"
+         },
          "marker": {
-          "color": "#AB63FA"
+          "color": "#117733"
          },
          "name": "Test",
          "showlegend": true,
@@ -4866,7 +4940,7 @@
          ],
          "xaxis": "x",
          "y": [
-          1.077618288538511
+          1.0794183318669401
          ],
          "yaxis": "y"
         }
@@ -4878,7 +4952,7 @@
            "size": 16
           },
           "showarrow": false,
-          "text": "Log Loss Performance",
+          "text": "Log loss performance",
           "x": 0.5,
           "xanchor": "center",
           "xref": "paper",
@@ -4888,8 +4962,8 @@
          }
         ],
         "height": 420,
-        "paper_bgcolor": "#F0F0F0",
-        "plot_bgcolor": "#F0F0F0",
+        "paper_bgcolor": "#FAFAFA",
+        "plot_bgcolor": "#FAFAFA",
         "showlegend": true,
         "template": {
          "data": {
@@ -5716,14 +5790,24 @@
          "domain": [
           0,
           1
-         ]
+         ],
+         "gridcolor": "#E5E5E5",
+         "gridwidth": 1,
+         "linecolor": "#CCCCCC",
+         "showgrid": false,
+         "showline": true
         },
         "yaxis": {
          "anchor": "x",
          "domain": [
           0,
           1
-         ]
+         ],
+         "gridcolor": "#E5E5E5",
+         "gridwidth": 1,
+         "linecolor": "#CCCCCC",
+         "showgrid": true,
+         "showline": true
         }
        }
       }
@@ -5749,7 +5833,7 @@
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": "pe3.13",
+   "display_name": "pe3.13 (3.13.0)",
    "language": "python",
    "name": "python3"
   },
diff --git a/docs/models/qrf/index.md b/docs/models/qrf/index.md
index 7e15838..170ec9a 100644
--- a/docs/models/qrf/index.md
+++ b/docs/models/qrf/index.md
@@ -1,21 +1,19 @@
 # Quantile Random Forests
 
-The `QRF` model harnesses the power of ensemble learning by utilizing multiple decision trees to predict different quantiles of the target variable distribution. This sophisticated approach allows for flexible modeling of complex relationships while providing robust estimates of uncertainty.
+The `QRF` model uses an ensemble of decision trees to predict different quantiles of the target variable distribution. This allows it to model non-linear relationships while estimating uncertainty across the conditional distribution.
 
 ## Variable type support
 
-QRF handles both numerical and categorical variables. For numerical targets, it uses quantile regression forests. For categorical targets (strings, booleans, numerically-encoded categorical variables), it automatically employs a Random Forest Classifier. The model detects variable types automatically and applies the appropriate method internally, requiring no manual specification from users.
+QRF handles both numerical and categorical variables. For numerical targets, it uses quantile regression forests. For categorical targets (strings, booleans, numerically-encoded categorical variables), it automatically uses a Random Forest Classifier. The model detects variable types internally and requires no manual specification.
 
 ## How it works
 
-Quantile Random Forests build upon the foundation of random forests by implementing a specialized algorithm from the quantile_forest package. The method begins by constructing an ensemble of decision trees, each trained on different bootstrapped samples of the original data. This process, known as bagging, introduces diversity among the individual trees and helps reduce overfitting.
+Quantile Random Forests build on standard random forests using the `quantile_forest` package. The method constructs an ensemble of decision trees, each trained on a bootstrapped sample of the data (bagging). At each split, only a random subset of features is considered, which introduces diversity among trees and reduces overfitting.
 
-During training, each tree in the forest predicts the target variable using only a random subset of the available features at each split point. This feature randomization further enhances diversity within the ensemble and improves its ability to capture various aspects of the underlying data relationships.
+Unlike standard random forests that aggregate predictions into averages, QRF retains the full predictive distribution from each tree and estimates quantiles directly from this empirical distribution.
 
 ## Key features
 
-The Quantile Random Forest (QRF) imputer provides a robust non-parametric method particularly effective for datasets exhibiting complex, non-linear relationships and heteroscedasticity. Unlike linear models, which rely on strong distributional assumptions, QRF makes minimal assumptions about the underlying data structure, adapting its uncertainty measures to reflect varying levels of variability within different regions of the input data.
+QRF is non-parametric and makes minimal assumptions about the data structure. It adapts its uncertainty estimates to different regions of the input space, producing wider prediction intervals where the data is more variable and tighter intervals where it is less so.
 
-QRF's primary strength lies in its predictive approach. While traditional random forests aggregate predictions into averages, QRF maintains the entire predictive distribution from each tree, directly estimating quantiles based on this empirical distribution. It also quantifies uncertainty through robust prediction intervals derived directly from its quantile estimates. These intervals dynamically adjust across the feature space, effectively signaling areas with varying levels of predictive certainty.
-
-Although QRF typically involves higher computational demands compared to simpler linear models, its enhanced accuracy on datasets with complex, non-linear relationships frequently justifies this trade-off. For applications where accurate predictive performance and meaningful uncertainty quantification are critical, QRF emerges as an especially valuable approach.
+The method is computationally heavier than linear models, but often more accurate on datasets with non-linear relationships or heteroscedasticity (where variance depends on predictor values). Hyperparameter tuning via Optuna is available for optimizing the number of trees, minimum samples per leaf, split thresholds, and feature sampling.
diff --git a/docs/models/qrf/qrf-imputation.ipynb b/docs/models/qrf/qrf-imputation.ipynb
index c59a999..ffba33f 100644
--- a/docs/models/qrf/qrf-imputation.ipynb
+++ b/docs/models/qrf/qrf-imputation.ipynb
@@ -97,7 +97,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 1,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -3658,9 +3658,9 @@
      "output_type": "stream",
      "text": [
       "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 8 concurrent workers.\n",
-      "[Parallel(n_jobs=-1)]: Done   2 out of   5 | elapsed:    6.8s remaining:   10.2s\n",
-      "[Parallel(n_jobs=-1)]: Done   3 out of   5 | elapsed:    6.8s remaining:    4.5s\n",
-      "[Parallel(n_jobs=-1)]: Done   5 out of   5 | elapsed:    6.8s finished\n"
+      "[Parallel(n_jobs=-1)]: Done   2 out of   5 | elapsed:   19.2s remaining:   28.7s\n",
+      "[Parallel(n_jobs=-1)]: Done   3 out of   5 | elapsed:   19.2s remaining:   12.8s\n",
+      "[Parallel(n_jobs=-1)]: Done   5 out of   5 | elapsed:   19.3s finished\n"
      ]
     },
     {
@@ -3704,8 +3704,32 @@
        },
        "data": [
         {
+         "error_y": {
+          "array": [
+           0.00008551920671991658,
+           0.0004107245170541804,
+           0.00018326432680445731,
+           0.00044296698746322717,
+           0.00024075509310405528,
+           0.00029083292370858524,
+           0.00043090603620434025,
+           0.0004606508356991907,
+           0.00034842456444518586,
+           0.0004626832277981355,
+           0.000816089425956785,
+           0.00038467021226627177,
+           0.0003398728669511024,
+           0.000309964218752641,
+           0.0002952487775964039,
+           0.0002930002413745392,
+           0.00022755896739762955,
+           0.00017022925074651195,
+           0.00005512096011768218
+          ],
+          "type": "data"
+         },
          "marker": {
-          "color": "#00CC96"
+          "color": "#DDCC77"
          },
          "name": "Train",
          "type": "bar",
@@ -3753,8 +3777,32 @@
          ]
         },
         {
+         "error_y": {
+          "array": [
+           0.000623604695409784,
+           0.0009828712930208695,
+           0.0012293155627994283,
+           0.001167454357338009,
+           0.000969638311140881,
+           0.0007127881853378359,
+           0.002636442666073569,
+           0.001259531904761815,
+           0.0010923805491178616,
+           0.001634028502692155,
+           0.0006800631282736149,
+           0.0009467176016052857,
+           0.0010995284899625922,
+           0.0008688007594503854,
+           0.0010158843160004165,
+           0.0011140556744687275,
+           0.0006082527278363497,
+           0.00045185788868329796,
+           0.0004799569375129921
+          ],
+          "type": "data"
+         },
          "marker": {
-          "color": "#AB63FA"
+          "color": "#117733"
          },
          "name": "Test",
          "type": "bar",
@@ -3817,8 +3865,8 @@
          "r": 50,
          "t": 80
         },
-        "paper_bgcolor": "#F0F0F0",
-        "plot_bgcolor": "#F0F0F0",
+        "paper_bgcolor": "#FAFAFA",
+        "plot_bgcolor": "#FAFAFA",
         "template": {
          "data": {
           "bar": [
@@ -4640,14 +4688,22 @@
         },
         "width": 750,
         "xaxis": {
+         "gridcolor": "#E5E5E5",
+         "gridwidth": 1,
+         "linecolor": "#CCCCCC",
          "showgrid": false,
+         "showline": true,
          "title": {
           "text": "Quantile"
          },
          "zeroline": false
         },
         "yaxis": {
-         "showgrid": false,
+         "gridcolor": "#E5E5E5",
+         "gridwidth": 1,
+         "linecolor": "#CCCCCC",
+         "showgrid": true,
+         "showline": true,
          "title": {
           "text": "Average Quantile Loss"
          },
@@ -4664,7 +4720,7 @@
     "# Plot the results for numerical variables\n",
     "if \"quantile_loss\" in qrf_results:\n",
     "    perf_results_viz = model_performance_results(\n",
-    "        results=qrf_results[\"quantile_loss\"][\"results\"],\n",
+    "        results=qrf_results,\n",
     "        model_name=\"QRF\",\n",
     "        method_name=\"Cross-validation quantile loss average\",\n",
     "    )\n",
@@ -4739,7 +4795,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 17,
+   "execution_count": 16,
    "metadata": {},
    "outputs": [
     {
@@ -4748,19 +4804,19 @@
      "text": [
       "Categorical variable distribution:\n",
       "risk_level\n",
+      "medium    150\n",
       "low       148\n",
-      "high      148\n",
-      "medium    146\n",
+      "high      144\n",
       "Name: count, dtype: int64\n",
       "\n",
       "Percentage distribution:\n",
       "risk_level\n",
+      "medium    0.339367\n",
       "low       0.334842\n",
-      "high      0.334842\n",
-      "medium    0.330317\n",
+      "high      0.325792\n",
       "Name: proportion, dtype: float64\n",
       "\n",
-      "Data types: {'age': dtype('float64'), 'sex': dtype('float64'), 'bmi': dtype('float64'), 'bp': dtype('float64'), 'risk_level': dtype('O')}\n",
+      "Data types: {'age': dtype('float64'), 'sex': dtype('float64'), 'bmi': dtype('float64'), 'bp': dtype('float64'), 'risk_level': <StringDtype(na_value=nan)>}\n",
       "\n",
       "Training set size: 353 records\n",
       "Testing set size: 89 records\n"
@@ -4797,7 +4853,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 18,
+   "execution_count": 17,
    "metadata": {},
    "outputs": [
     {
@@ -4840,20 +4896,20 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 19,
+   "execution_count": 18,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "Categorical imputation accuracy: 42.70%\n",
+      "Categorical imputation accuracy: 44.94%\n",
       "\n",
       "Confusion matrix:\n",
       "                Predicted: low  Predicted: medium  Predicted: high\n",
-      "Actual: low                 13                  5               12\n",
-      "Actual: medium               7                 11               11\n",
-      "Actual: high                 3                 13               14\n"
+      "Actual: low                 14                  4               12\n",
+      "Actual: medium               8                 13                8\n",
+      "Actual: high                 4                 13               13\n"
      ]
     }
    ],
@@ -4881,7 +4937,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 20,
+   "execution_count": 19,
    "metadata": {},
    "outputs": [
     {
@@ -4889,9 +4945,9 @@
      "output_type": "stream",
      "text": [
       "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 8 concurrent workers.\n",
-      "[Parallel(n_jobs=-1)]: Done   2 out of   5 | elapsed:    0.7s remaining:    1.1s\n",
-      "[Parallel(n_jobs=-1)]: Done   3 out of   5 | elapsed:    0.7s remaining:    0.5s\n",
-      "[Parallel(n_jobs=-1)]: Done   5 out of   5 | elapsed:    0.7s finished\n"
+      "[Parallel(n_jobs=-1)]: Done   2 out of   5 | elapsed:    3.9s remaining:    5.8s\n",
+      "[Parallel(n_jobs=-1)]: Done   3 out of   5 | elapsed:    4.1s remaining:    2.7s\n",
+      "[Parallel(n_jobs=-1)]: Done   5 out of   5 | elapsed:    4.4s finished\n"
      ]
     },
     {
@@ -4899,8 +4955,8 @@
      "output_type": "stream",
      "text": [
       "Categorical imputation cross-validation results (log loss):\n",
-      "Mean train log loss: 0.2656\n",
-      "Mean test log loss: 1.2133\n"
+      "Mean train log loss: 0.2666\n",
+      "Mean test log loss: 1.2308\n"
      ]
     }
    ],
@@ -4921,7 +4977,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 21,
+   "execution_count": 20,
    "metadata": {},
    "outputs": [
     {
@@ -4932,8 +4988,14 @@
        },
        "data": [
         {
+         "error_y": {
+          "array": [
+           0.004587677764544018
+          ],
+          "type": "data"
+         },
          "marker": {
-          "color": "#00CC96"
+          "color": "#DDCC77"
          },
          "name": "Train",
          "showlegend": true,
@@ -4943,13 +5005,19 @@
          ],
          "xaxis": "x",
          "y": [
-          0.2656462787483515
+          0.26664213025700334
          ],
          "yaxis": "y"
         },
         {
+         "error_y": {
+          "array": [
+           0.21683973320460714
+          ],
+          "type": "data"
+         },
          "marker": {
-          "color": "#AB63FA"
+          "color": "#117733"
          },
          "name": "Test",
          "showlegend": true,
@@ -4959,7 +5027,7 @@
          ],
          "xaxis": "x",
          "y": [
-          1.2133400644133157
+          1.2308133756846842
          ],
          "yaxis": "y"
         }
@@ -4981,8 +5049,8 @@
          }
         ],
         "height": 420,
-        "paper_bgcolor": "#F0F0F0",
-        "plot_bgcolor": "#F0F0F0",
+        "paper_bgcolor": "#FAFAFA",
+        "plot_bgcolor": "#FAFAFA",
         "showlegend": true,
         "template": {
          "data": {
@@ -5809,14 +5877,24 @@
          "domain": [
           0,
           1
-         ]
+         ],
+         "gridcolor": "#E5E5E5",
+         "gridwidth": 1,
+         "linecolor": "#CCCCCC",
+         "showgrid": false,
+         "showline": true
         },
         "yaxis": {
          "anchor": "x",
          "domain": [
           0,
           1
-         ]
+         ],
+         "gridcolor": "#E5E5E5",
+         "gridwidth": 1,
+         "linecolor": "#CCCCCC",
+         "showgrid": true,
+         "showline": true
         }
        }
       }
@@ -5842,7 +5920,7 @@
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": "pe3.13",
+   "display_name": "pe3.13 (3.13.0)",
    "language": "python",
    "name": "python3"
   },
diff --git a/docs/models/quantreg/index.md b/docs/models/quantreg/index.md
index 2e31d4d..38637a2 100644
--- a/docs/models/quantreg/index.md
+++ b/docs/models/quantreg/index.md
@@ -1,23 +1,21 @@
 # Quantile Regression
 
-The `QuantReg` model takes a direct approach to modeling specific quantiles of the target variable distribution. Unlike methods that model the mean and then derive quantiles from distributional assumptions, quantile regression addresses each conditional quantile explicitly, providing greater flexibility and robustness in heterogeneous data settings.
+The `QuantReg` model directly models specific quantiles of the target variable distribution. Unlike methods that model the conditional mean and derive quantiles from distributional assumptions, quantile regression addresses each conditional quantile separately.
 
 ## Variable type support
 
-QuantReg is designed specifically for numerical variables and does not support categorical variable imputation. If your imputation targets include categorical variables (string, or numerically-encoded categorical variables), consider using OLS or QRF models instead, which automatically handle both numerical and categorical targets through internal classification methods.
+QuantReg is designed for numerical variables and does not support categorical imputation. If your targets include categorical variables, use OLS or QRF instead, which handle both numerical and categorical targets through internal classification methods.
 
 ## How it works
 
-Quantile regression in Microimpute leverages the statsmodels' QuantReg implementation to create precise models of conditional quantiles. During the training phase, the approach fits separate regression models for each requested quantile level, creating a focused model for each part of the conditional distribution you wish to estimate.
+The implementation uses statsmodels' `QuantReg`. During training, a separate regression model is fitted for each requested quantile level.
 
-The mathematical foundation of the method lies in its objective function, which minimizes asymmetrically weighted absolute residuals rather than squared residuals as in ordinary least squares. This asymmetric weighting system penalizes under-predictions more heavily when estimating higher quantiles and over-predictions more heavily when estimating lower quantiles. This clever formulation allows the model to converge toward solutions that represent true conditional quantiles.
+The objective function minimizes asymmetrically weighted absolute residuals rather than squared residuals (as in OLS). For higher quantiles, under-predictions are penalized more heavily; for lower quantiles, over-predictions are penalized more heavily. This asymmetry causes each model to converge toward the true conditional quantile.
 
-When making predictions, the system applies the appropriate quantile-specific model for each requested quantile level. This direct approach means predictions at different quantiles come from distinct models optimized for those specific portions of the distribution, rather than from a single model with assumptions about the error distribution.
+When predicting, the system applies the quantile-specific model for each requested level. Predictions at different quantiles come from distinct models, each optimized for that part of the distribution.
 
 ## Key features
 
-Quantile regression offers several compelling advantages for imputation tasks. It allows direct modeling of conditional quantiles without making restrictive assumptions about the underlying distribution of the data. This distribution-free approach makes the method robust to outliers and applicable in a wide range of scenarios where normal distribution assumptions might be violated.
+Quantile regression models conditional quantiles without assuming a particular error distribution, making it robust to outliers. It naturally captures heteroscedasticity, adapting to changing variance patterns across the feature space (unlike OLS, which assumes constant variance).
 
-The method excels at capturing heteroscedasticity—situations where the variability of the target depends on the predictor values. While methods like OLS assume constant variance throughout the feature space, quantile regression naturally adapts to changing variance patterns, providing more accurate predictions in regions with different error characteristics.
-
-By fitting multiple quantile levels, the approach provides a comprehensive picture of the conditional distribution of the target variable. This detailed view enables more nuanced imputation where understanding the full range of possible values is important. For instance, it can reveal asymmetries in the conditional distribution that other methods might miss, offering valuable insights into the uncertainty structure of the imputed values.
+By fitting multiple quantile levels, the method gives a picture of the full conditional distribution. This can reveal asymmetries that other methods would miss. However, QuantReg is limited to linear relationships between predictors and the target.
diff --git a/docs/models/quantreg/quantreg-imputation.ipynb b/docs/models/quantreg/quantreg-imputation.ipynb
index 79308a9..2a93247 100644
--- a/docs/models/quantreg/quantreg-imputation.ipynb
+++ b/docs/models/quantreg/quantreg-imputation.ipynb
@@ -74,7 +74,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 1,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -3633,9 +3633,9 @@
      "output_type": "stream",
      "text": [
       "[Parallel(n_jobs=-1)]: Using backend LokyBackend with 8 concurrent workers.\n",
-      "[Parallel(n_jobs=-1)]: Done   2 out of   5 | elapsed:    5.7s remaining:    8.5s\n",
-      "[Parallel(n_jobs=-1)]: Done   3 out of   5 | elapsed:    5.7s remaining:    3.8s\n",
-      "[Parallel(n_jobs=-1)]: Done   5 out of   5 | elapsed:    5.8s finished\n"
+      "[Parallel(n_jobs=-1)]: Done   2 out of   5 | elapsed:    9.7s remaining:   14.5s\n",
+      "[Parallel(n_jobs=-1)]: Done   3 out of   5 | elapsed:    9.7s remaining:    6.5s\n",
+      "[Parallel(n_jobs=-1)]: Done   5 out of   5 | elapsed:   10.0s finished\n"
      ]
     },
     {
@@ -3679,8 +3679,32 @@
        },
        "data": [
         {
+         "error_y": {
+          "array": [
+           0.00007068607962860608,
+           0.00008566173778664199,
+           0.00010354861421659208,
+           0.00010446681832917282,
+           0.00009958982144334462,
+           0.00010899487172081684,
+           0.00012524333466419057,
+           0.0001401559350155429,
+           0.0001419856727246125,
+           0.00014148804255191002,
+           0.00013964325829394816,
+           0.0001352059984264296,
+           0.0001266211747154425,
+           0.00010806937892337481,
+           0.0000955126673952158,
+           0.00009572945677819657,
+           0.00010825711304818197,
+           0.00009924455051088341,
+           0.00008058254845100971
+          ],
+          "type": "data"
+         },
          "marker": {
-          "color": "#00CC96"
+          "color": "#DDCC77"
          },
          "name": "Train",
          "type": "bar",
@@ -3728,8 +3752,32 @@
          ]
         },
         {
+         "error_y": {
+          "array": [
+           0.0002923819760799368,
+           0.00036360537094371777,
+           0.0004548700007812027,
+           0.0004754851566973023,
+           0.0004310373487827412,
+           0.0004963764336370348,
+           0.0004984041180187477,
+           0.0005592567272666623,
+           0.0006264645396437899,
+           0.0006000923054654875,
+           0.0006170472013545412,
+           0.0005280976014239853,
+           0.0004788123955147881,
+           0.00041499916196403646,
+           0.0003537849752500568,
+           0.0003903169140390032,
+           0.0004234486790403488,
+           0.0003675153867336991,
+           0.00032353724735584744
+          ],
+          "type": "data"
+         },
          "marker": {
-          "color": "#AB63FA"
+          "color": "#117733"
          },
          "name": "Test",
          "type": "bar",
@@ -3792,8 +3840,8 @@
          "r": 50,
          "t": 80
         },
-        "paper_bgcolor": "#F0F0F0",
-        "plot_bgcolor": "#F0F0F0",
+        "paper_bgcolor": "#FAFAFA",
+        "plot_bgcolor": "#FAFAFA",
         "template": {
          "data": {
           "bar": [
@@ -4615,14 +4663,22 @@
         },
         "width": 750,
         "xaxis": {
+         "gridcolor": "#E5E5E5",
+         "gridwidth": 1,
+         "linecolor": "#CCCCCC",
          "showgrid": false,
+         "showline": true,
          "title": {
           "text": "Quantile"
          },
          "zeroline": false
         },
         "yaxis": {
-         "showgrid": false,
+         "gridcolor": "#E5E5E5",
+         "gridwidth": 1,
+         "linecolor": "#CCCCCC",
+         "showgrid": true,
+         "showline": true,
          "title": {
           "text": "Average Quantile Loss"
          },
@@ -4639,7 +4695,7 @@
     "# Plot the results\n",
     "if \"quantile_loss\" in quantreg_results:\n",
     "    perf_results_viz = model_performance_results(\n",
-    "        results=quantreg_results[\"quantile_loss\"][\"results\"],\n",
+    "        results=quantreg_results,\n",
     "        model_name=\"QuantReg\",\n",
     "        method_name=\"Cross-validation quantile loss average\",\n",
     "    )\n",
@@ -4652,7 +4708,7 @@
  ],
  "metadata": {
   "kernelspec": {
-   "display_name": "pe3.13",
+   "display_name": "pe3.13 (3.13.0)",
    "language": "python",
    "name": "python3"
   },
diff --git a/docs/myst.yml b/docs/myst.yml
index e7f484c..a078097 100644
--- a/docs/myst.yml
+++ b/docs/myst.yml
@@ -48,8 +48,6 @@ project:
     - title: Use cases
       children:
         - file: use_cases/index
-          children:
-            - file: use_cases/scf_to_cps/imputing-from-scf-to-cps
 site:
   options:
     logo: logo.png
diff --git a/docs/use_cases/index.md b/docs/use_cases/index.md
index b7ea6d4..32c7af9 100644
--- a/docs/use_cases/index.md
+++ b/docs/use_cases/index.md
@@ -1,14 +1,14 @@
-# Imputing full variables across surveys
+# Imputing variables across surveys
 
-This document explains what the workflow for imputing variables from one survey to another using Microimpute may look like. We'll use the example of imputing wealth data from the Survey of Consumer Finances (SCF) into the Current Population Survey (CPS).
+This document walks through a typical workflow for imputing variables from one survey to another using microimpute, using wealth imputation from the Survey of Consumer Finances (SCF) into the Current Population Survey (CPS) as the running example.
 
 ## Identifying receiver and donor datasets
 
-The first step is to identify your donor and receiver datasets. The Donor dataset is that containing the variable you want to impute (e.g., SCF contains wealth data). The Receiver dataset will receive the imputed variable (e.g., CPS which originally did not contain wealth data but will after our imputation is completed). It is important for these two datasets to have predictor variables in common for the imputation to be succcesful. For example, both the SCF and CPS surveys contain demographic and financial data that may help us understand how wealth values may be distributed. 
+Start by identifying your donor and receiver datasets. The donor dataset contains the variable you want to impute (here, the SCF contains wealth data). The receiver dataset will receive the imputed variable (here, the CPS, which lacks wealth data). Both datasets need predictor variables in common for the imputation to work. For example, both the SCF and CPS contain demographic and financial data that can help predict wealth.
 
 ```python
 import pandas as pd
-from microimpute.models import OLSImputer, MatchingImputer, QRFImputer
+from microimpute.models import OLS, Matching, QRF
 
 # Load donor dataset (SCF with wealth data)
 scf_data = pd.read_csv("scf_data.csv")
@@ -19,97 +19,107 @@ cps_data = pd.read_csv("cps_data.csv")
 
 ## Cleaning and aligning variables
 
-Before imputation, you need to ensure both datasets have compatible variables. Identify common variables present in both datasets
-and standardize their variable formats, units, and categories so that Python can recognize they indeed represent the same the same data. Remember to also handle missing values in common variables. Lastly, identify the target variables in the donor dataset which will directly inform the values of the imputed variables in the receiver dataset. For details on data preprocessing options refer to the [Data preprocessing page](./preprocessing.md).
+Before imputation, make sure both datasets have compatible variables. Identify common variables present in both datasets and standardize their formats, units, and categories so that Python can match them correctly. Handle missing values in common variables, and identify the target variables in the donor dataset that will inform the imputed values in the receiver. For details on preprocessing options, see the [Data preprocessing page](../imputation-benchmarking/preprocessing.md).
 
 ```python
-# Identify common variables 
+# Identify common variables
 common_variables = ['age', 'income', 'education', 'marital_status', 'region']
 
 # Ensure variable formats match (example: education coding)
 education_mapping = {
-    1: "less_than_hs", 
-    2: "high_school", 
-    3: "some_college", 
-    4: "bachelor", 
+    1: "less_than_hs",
+    2: "high_school",
+    3: "some_college",
+    4: "bachelor",
     5: "graduate"
 }
 
 # Apply standardization to both datasets
 for dataset in [scf_data, cps_data]:
     dataset['education'] = dataset['education'].map(education_mapping)
-    
+
     # Convert income to same units (thousands)
     if 'income' in dataset.columns:
         dataset['income'] = dataset['income'] / 1000
-    
+
 # Identify target variable in donor dataset
 target_variable = ['networth']
 ```
 
 ## Performing imputation
 
-Microimpute offers several methods for imputation across surveys, which are described in the [Models Chapter](../models). The approach under the hood will differ based on the method chosen, although the workflow will remain constant. Let us see this for two different example methods.
+Microimpute offers several methods for imputation across surveys, described in the [Models chapter](../models). The underlying approach differs by method, but the workflow stays the same. Here are two examples.
 
 ### Matching imputation
 
-Matching finds similar observations in the donor dataset for each observation in the receiver dataset and imputes the values for those receiver observations based on the values of the target value in the donor dataset. To do so it should be fitted on the donor dataset and predict using the receiver dataset. This will ensure the correct mapping of variables from one survey to the other. 
+Matching finds similar observations in the donor dataset for each receiver observation and transfers the donor's target values. Fit on the donor dataset, then predict using the receiver dataset.
 
 ```python
 # Set up matching imputer
-matching_imputer = MatchingImputer(
-    predictors=common_variables,
-    imputed_variables=target_variable
-)
+matching = Matching()
 
 # Train on donor dataset
-matching_imputer.fit(scf_data)
+fitted_matching = matching.fit(
+    X_train=scf_data,
+    predictors=common_variables,
+    imputed_variables=target_variable,
+)
 
 # Impute target variable into receiver dataset
-cps_data_with_wealth_matching = matching_imputer.predict(cps_data)
+cps_with_wealth_matching = fitted_matching.predict(X_test=cps_data)
 ```
 
 ### Regression imputation (OLS)
 
-OLS imputation builds a linear regression model using the donor dataset and applies it to the receiver dataset, predicting what wealth values may be for a specific combination of predictor variable values. To do so, again we need to make sure that we first fit the model on the donor dataset, while calling predict on the receiver dataset.
+OLS builds a linear regression model on the donor dataset and predicts wealth values for each combination of predictor values in the receiver. Again, fit on the donor and predict on the receiver.
 
- ```python
+```python
 # Set up OLS imputer
-ols_imputer = OLSImputer(
-    explanatory_variables=common_variables,
-    target_variable=target_variable
-)
+ols = OLS()
 
 # Train on donor dataset
-ols_imputer.fit(scf_data)
+fitted_ols = ols.fit(
+    X_train=scf_data,
+    predictors=common_variables,
+    imputed_variables=target_variable,
+)
 
 # Impute target variable into receiver dataset
-cps_data_with_wealth_ols = ols_imputer.impute(cps_data)
+cps_with_wealth_ols = fitted_ols.predict(X_test=cps_data)
 ```
 
 ## Evaluating imputation quality
 
-Evaluating imputation quality across surveys can be challenging since the true values aren't known in the receiver dataset. Comparing the distribution of the target variables in the donor dataset to the distribution of the variables we imputed in the receiver dataset may give us an understanding of the imputation quality for different sections of the distribution. We may want to pay particular attention to obtaining accurate prediction not only for mean or median values but also look at the performance at the distribution tails. This can be achieved computing the quantile loss supported by Microimpute. Additionally, if we have performed imputation accross multiple methods we may want to compare across them. Microimpute supports this through multiple easy-to-use metrics described in the [Metrics page](./metrics.md) file.
+Evaluating imputation quality across surveys is challenging since the true values are unknown in the receiver dataset. Comparing the target variable's distribution in the donor to the imputed distribution in the receiver can reveal how well the imputation captures different parts of the distribution. Beyond mean or median accuracy, check performance at the tails using quantile loss. When comparing multiple methods, microimpute provides several metrics described in the [Metrics page](../imputation-benchmarking/metrics.md).
 
 ```python
-# Ensure all imputations are in a dictionary mapping quantiles to dataframes containing imputed values
-method_imputations = {
-    [0.1]: pd.DataFrame
-    [0.5]: pd.DataFrame
-    ...
-}
+from microimpute.comparisons import get_imputations
+from microimpute.comparisons.metrics import compare_metrics
+
+# Generate imputations from multiple models using cross-validation
+method_imputations = get_imputations(
+    model_classes=[QRF, OLS, Matching],
+    X_train=train_data,
+    X_test=test_data,
+    predictors=common_variables,
+    imputed_variables=target_variable,
+)
 
-# Compare original wealth distribution with imputed wealth across methods
-loss_comparison_df = compare_quantile_loss(Y_test, method_imputations)
+# Compare quantile loss across methods
+loss_comparison_df = compare_metrics(
+    test_y=test_data[target_variable],
+    method_imputations=method_imputations,
+    imputed_variables=target_variable,
+)
 ```
 
 ## Incorporating the imputed variable
 
-Once you've chosen the best imputation method, you may want to incorporate the imputed variable into your receiver dataset for future analysis.
+Once you've chosen the best imputation method, incorporate the imputed variable into your receiver dataset for downstream analysis.
 
 ```python
 # Choose the best imputation method (e.g., QRF)
-final_imputed_dataset = cps_data_with_wealth_qrf
+final_imputed_dataset = cps_with_wealth_qrf
 
 # Save the augmented dataset
 final_imputed_dataset.to_csv("cps_with_imputed_wealth.csv", index=False)
@@ -117,4 +127,8 @@ final_imputed_dataset.to_csv("cps_with_imputed_wealth.csv", index=False)
 
 ## Key considerations
 
-Model selection plays a critical role in this workflow because different imputation methods have unique strengths. For example, a Quantile Regression Forest (QRF) often performs better when capturing complex relationships between variables, while a Matching approach may be more effective at preserving the original distributional properties of the data. Additionally, not all models can impute categorical data. For example, atching is able to match any value regardless of its data type, but QuantReg does not support categorical imputation. OLS and QRF will use logistic regression and random forest classification methods under the hood, respectively. Variable selection is equally important, since the common predictors across datasets should have strong power for explaining the target variable to ensure a reliable imputation. Because the ground truth is typically unknown in the receiver dataset, validation can involve simulation studies or comparing imputed values against known aggregate statistics. Finally, it is crucial to maintain documentation of the imputation process, from the choice of model to any adjustments made, so that the analysis remains transparent and reproducible for others. For the full pipeline details on SCF-to-CPS net worth imputation refer to the following [notebook](./scf_to_cps/imputing-from-scf-to-cps.md).
+Model selection matters because different imputation methods have different strengths. QRF often performs better at capturing non-linear relationships, while Matching tends to preserve the original distributional properties of the data. Not all models handle categorical data: Matching can match any value regardless of type, but QuantReg does not support categorical imputation. OLS and QRF use logistic regression and random forest classification internally for categorical targets.
+
+Variable selection is equally important. The common predictors should have strong explanatory power for the target variable. Because the ground truth is unknown in the receiver dataset, validation can involve simulation studies or comparison against known aggregate statistics.
+
+For a complete worked example of the SCF-to-CPS net worth imputation pipeline, see the [autoimpute notebook](../autoimpute/autoimpute.ipynb). The [microimpute paper](https://github.com/PolicyEngine/microimpute/blob/main/paper/main.pdf) presents the full methodology and reports results from this imputation.
diff --git a/docs/use_cases/scf_to_cps/autoimpute_best_model_imputations.png b/docs/use_cases/scf_to_cps/autoimpute_best_model_imputations.png
deleted file mode 100644
index 477f98d28bcf3e3610341819efac55f8fe2f4748..0000000000000000000000000000000000000000
GIT binary patch
literal 0
HcmV?d00001

literal 56963
zcmeFZWmFyAwk?_vlAyue7w!@ycyM?3;4Z-yP7>T5f+x7UEIhcodvIR_cX-9O_h~2l
z-uLdG_uBh)XvLyc6^%J-)R<$AK6)R7D9TG9Bj6*vc<};RN>Wt$#S0iH@U02=2Dri>
zX`BW8c<H1pA^f6jgmCZ03(^--qC%=}`iJSS<5hcS0=S@1R8ma`Fc?g<WIEQq41Bpb
zz~NIsL?j)A9l%#lf<a3@{Bl>aQS8YXE44E^%tNI*S5x4kCU82B8rNL4`Dp9609eT{
ze;uTX7;xfi#03u`8~<_rBfKKSi8T5J{C^)`Fi?LufMKY|MbSxL{`c|2VF&l`bJCa8
zq#xIyu<}_U6dz&!I@U&5|Kl7+{^OU9W%x8v45`0o^ra)+_&?5J)4se2Xu|jqLKXY>
zbY2AT_Ws8?{D+q>5x~?A=yDW)O^+0~7Wp6N=ngP2m{9rO6e*H_PY<|8`0B6e!chE%
zg{4_a|0Ts9^OwQC0N1$R{D-Z6!NVy+c)!M3fc~=9i)RBO{kQdiss4`v|7EuS?+u6o
zGX8xqTdrAu@nR(Hj=+AUMPl2zU#Zy0Fdw>{KbpyEvaY;28bd60YUa4#dJnfTn${Fr
ztMl%MOJ`^>7P<5nKF8UwlXKv629LVZ2t_iH-$zX-5RZp<%G3(}sAS*X_PsCRa6H=P
z_eUgn^VwKza`2#@q^rBQeyr9MIf#GK;(9!8$L(^RBld}Kb#2sf>+O9(<>o;6JQ$ty
z>kkQ1MZ_y3j;5cAP^8@>r5!=ar|lz8|93>_WFr383P*pSYfg#w-D##e(Y)4HIZ`pW
zpN+&s8jKoXN?R^*FCRN4Fw0i(<g5~f<Fs>+8Yn;RfADKFX!Y=uMC<h2JDyqYWpTJS
zVo0CB5PZU<{;1jaJ+`vebK1fmo(Q3s^C9wcwYu@?Z_d70fdIYHX!yAxluT3sUEHSq
zK?-*J+)JLDG`5`5mG{mQY<(0)JEv%_wTAYT)`U$csS&FvncepE;$*sWW0#!W9<3C=
zlPWg{B`SEHJsp-C=Bqvgr!RrndvW$bp<TmT0Cicrfy^^6{;7~?g}>P_Na_0e1S3%v
zie_PZUTXN1;DB=Y*n0Jcwut@cDZSzzt>a-kJ!r1t0d-%EKCllP<E_r)^2wc4$)gKN
z9?DgPCtS%w4Nj%8xJbJ$9wF6|?Pos6eHbSFmXO=&&_!Ej`rC{z20gLInH{{(u0?8r
z=D@wiG~Qfp@b!1Q-~3j86H3MptLm5ES$?Sfz=pchdHgU6;oF^A7%ubS7Rh}QhVc5e
zA%C^QLwiK)SZ3Z3&)VX$6a16lQ12B<4uaNwxXq1f4*K`%_kn$xj{}XJsjzvlBNT8a
z_SL(8>t2;Bw#J92RA(X2d_SVs7^x-W+dYr?^m}*_EQu?aNPCg%x&>ODPD-Ub<#f%X
zreMS5HB*gX2}HqBf8F=h*w+wxTKfG%0SDEvPEh)tyk*nfOn9e;S9cwMZ{@=k2E_4*
zbH(@C-{|Ijw)CFYHRYZHRTB?NN{G=7+s0WEv*vDtNG67$Jd!F+41rNxOv>5qFmK7C
z>x&?TwBC^FBs4~8(Hw<z*obB6weDiBKTGGlK7x+wwqYdHJm{3c5$$>edro%C-f*MC
zcMiK<+<ZQ_=vCj8S~Z#EL|8(>-e_eLqmSwP#vN(3AKdPlYn2>FmJbqoV^nr3-G=Qy
zT<l;yH0cBg7wDIL!MU8ZU!Dyvo}&@VDnK6&M`NriTnR!UVfCEHmJ5P7JykBuX4z>$
zz?r{oi^Aswe5YNJ_Lp9$;MjGwZ|u|zpBMUwr;206FjUU+Z5ih?8*PkqY9Hsju}))$
z5F=V2-`lSI)F@xg601MqG3!zgM<WSc<*h3SEHSgBAi7l$S<)Pri=*@O0`1W<>$j>W
zX6KcEn@eYw;Mrm`pI0ZcmO}5D<~uvYQ^~U3kY@4Ug*-A=n$9X63SQ!L_`fT>s<G5W
z`)n+6b;-Lm8us*Qp(A2n2JhH3baU+XOv8>VcL44fFO17jwA?KC%HETDRKI=Xf!f7}
z&+S~+sS|N~B-leO)M&FN;o4<Z$+ou-uUw<{Lnuxk4BnI2nW^wpUkr=;3Xi0|XA@O|
zQnfHy3)KZ+$w^N!2+4;_>c9(&Z@JdswQG_jG;(c<c!qen-q;hZhSO0ZB*J<)V!2|s
zllGlagRz`5+>2zEgkvNf!hnZ8!SYWNzd5P&FTF<jyzU9p-^fcu)zwz){Km6caK}~1
zY?~PM+DOS&BsKT2zZn!&6OdUy5FB|b%{()@ss15SU6glcQY#`L(ZORokjcSOrrr(G
zSM~@$PS{~Kuomp5=6;_qT)ki#O`*gmo6)AP)e3~f<OwQ~ATzaSQaf4i!1186(xr->
zF+3uP(}x5ecVQyq?Nx)Go!c&wq-k)^)wgpFoA36t;sW0FnvuDTO;P*&K??d~cPbQ8
zKOWguH`WQ>0Jl62Q1zep?5`gse-#Ea-Q!;2b8DO_iJ3_1*L&?5+}>gHG-+u3WDxUL
zM!ssl?F0$JfSMCaSF_h8d}<s`qyxIt9lwe7*)BWDc58;fXvMGy*)0E{7H&tC0)?7P
zroo*BqjeO1L~knMB>7=$CY3}OmD&2MyeYDf(c0|leJ-x5)r7*N&Sr0m_WV5@1`oT>
zvWw``a*K-`)*odPzwPbSNQ)7Vs;(1<r!|#B9ye{Vq3WG-2k0sCuzeD;pL<ol*i;{y
zVy&zXpXUjJ9B<TggzHwA?;v>9`%led^eE9JUSTZtRGa%O&PIwXf~chF=ryAJqq|sl
zx-5);%N!QBvRlcbhhlO{xqM$&*5}@Fx}I){4vVdx7CIgZ1?|(k+;c(!jeu{#sqAeO
zfwDD7Kaceyn+!99)2=@ogEY2PKY(v|-lcyiUw68h(x*ZrOG|Mj&c;X^$UR+O+c4*8
zh?Q#wNjIO>Z<)`QZhaC(B6;W3<IJWPi%GC&9&!45ywws|$TrR;V+G79GfbrzjGo>g
z;fu}rLAN_-3)^4<Wo;;v7SlrG4g`T0FX*Rmfap4GrP^;ihRud``UVNkSOi48yTQEk
zsoR->0D7)*$fN~U7?#ZTII9hu0uk5KHbh7KLX2ki3ioGRFzANuU`K@NMZ4eY$^Mnt
z+NO%_S?}9EI@8mGr?9B5C$>jZVRG-7(U@TNleW!~!CG0%SW5zB|DZW?F{&L9#S_Fq
zPRwp=;%2P7rBOf5oa>n9U8g7e^ob>Ae>JmWF8RJ1V_@RN;DNZE4k2Uq-KEcH@UbZp
zoXDDP)hA9q_e*8LCS`1e!^oVF4XrKa%L|ffMd~uVNeYo3cB#(?Q1S;|JOW1zo4M>b
z^JGjj=_JA(=2gX)1{C&YNQX<mYC>raMdM{?l~{_QjM4p(a~Dn>5FV<nyP@AnWG>2+
zR>7tu2-^d}sO^4+ospfl^NFk`UK#`bHitvgJb1CgV|P)Hg{|y8jV-H;JsC)7M}9rb
z<5GdL;--g#uuX35B|OP-z`D5fs@HaT^Yl-S`t|!46-4mO8ZMt9JP&7iUWdj&60-rv
zJ|)QJqQnOuOSjroP+Y+vfYp1MLzvQZCDa!G>y>5P@`)hRPrXdq)!c#ah&?wOjXvPe
zAlY}K4tU+y?J*Vw_bgH|B${aB=(VGXYd8<7aLdXr9u`D}^kSTQW`o~_yMvV|wy5Wx
z!pO5Fi^A_AQe6`?9pbpdB(vR^v>0oe`NJ$&*azE4V?`?$-4H4x!;W7wbnhs5^qYVi
zh3HuJiY@+iPB#m->=uF(VMOG4C1V&u*z&t(p%Ca27~NZfd8`F77r}fsByn>$>^qBp
zN{#uFb(90nT@SK_WoWbXrrY>_I-R8GciiT!-X$k9K^1@Hta<DDL#_xcoyWR+q_8jO
znNUqm7_^R{08tX_T4F17+c%5LZ%2R31~+`?EJ;UT^Yk`?Q$s5hY2;v&bLe_!+*wen
ziWd?(f<NC*F@cO7M&){E_qt<O<LYO%OID4E@|8UX?2Gzie&Gvjzw<U*cre;8bTuv#
zW93yZuYsZv%t1o1@x2TMBt}PyIT*NY97~z7RlVZ5$(Q~IFVn>{pb1RK7|_(R`aW<s
zA!Mxx&)b<nGG@9={+0QzzYx{Cna~q%jCUQKRk-bML-<e*m&Pb2OcTs+v*d0YVn(A_
zN47joPk-0-#Xc?$q{tFeydV|hybg^}4u?NW6ca(CbcB779o!TtW6qcxu90@SEibm1
zcFK~U@T@D%y&5o=x+lC{?Z4_($W&^yQ=pK-)acE`^la5ZT4QRBm%N*y6lo!Z{S;2w
z#`0&cX}akBwZt=&gX@A#(}slM;wZJ6x%y64e(um#HcN3K3L;y^w_X_P;<m2z>Lg~q
zpipMWmq6N``j~WNi<Rl`nU6iq-gu<KDq5NrZJa|I{2?azEYz~4Wof0VS2W^{=g48P
zZ<6oe1L!x>xcv>Eb%&nMmt5Q<+$+g*xx_)P5t5uex}K=w+E(1zH(=X>82)TYXT_{(
z^v{WsZJ21vlvdzb$?~ZrP|RLm4CuR0-<Cb-9#QSs2K~Axr)L6Zr+n#SoV;<fTtcO3
zW{!Q$9>3HPYEwiXxD%gptXW=5MUbLsIq%8@>3Ci>yX6uA%nhnsjGoVfPq)b<<g$(s
zmk{g2kCEd3DkeWVS8Y{;K=|w}SFbmktSA<muPK&>#nP6p1hFG*h>K}AB(hEujdKs8
zweY2CDVHc+ifxbkdGbsrENnZAz#V#BnX--y8Zb8SSJUtBK>BI8RD^+T;guV3Yj($1
z;bb0^!KCf5*?pBkV#|%Y2tP_-wdps^%f}N6T8ex3ivmpNU4CPFM5R3~UnNw=lG2BW
zOGnfmV@a9Dz&GXSL>x{JO*ONguu-{)lMKCXyfN2NKV#?(ZYTw{nsV5kb&%bf|MQmZ
z(!wbZO{RNe{e1CoS97;lO9i~`&;FIKbHd0oE{RR?`&@Qla&!J}b-Q_!yMH-01*=uZ
z(fWsyG0&snTgn?2n^U$3@ux-$xGeF=;0WG#8;1ebR8M`%wG1!sD3;7455G9m`aX`7
zgLo(f&u@?3LcFhH@{0SEoyuKn<!;VTU!~sfz~V0FHzeI{NeZIHi%+S`Lv}xWYNR!&
z^QkqWiKh2UDu~Z@NNT=buWfV>iN-^z#7zIU8<<7y7tk_@qkX0mylpGTJLrCg_7)za
zYhwDPNJSq%2w^x34p!?SS`F3MsMgVgWSEz0;0&i+sZn2{$^Y(fJUyDKyB^*jlN@HH
zT!eqvPrzpzec>WKD@GKJL_B88O-e2nle%A{UIRivPu^qoC3qIdl0I2n_0Z_FC56(U
zsjz*H5s(t}-oOxn=$v9Td80!UeQHb$xiyhTK&^~X`W#-yGLSF_LP6^Y`5(EGDba^c
zET^km_ABsUsY+J$#G#`q$|Xj3+|&@{CvIWL;vK(nu)b7Nl~>TW@Q{1%No$0(T2T1r
zosfE7306pvjI*9-^_IDmazc0umt%SDal&0xMwVmTqs4Ls%)Tm<l}Fb~r|*y63hq2C
zjBJz{^|#?tOOLf19P7**?t@U{4-d0g%U%qb@FC}e(tf!PSJ0+zRA?3cocAuFijVJd
zQMa>TWToSt!J7%o!DIE}ikrXIyUyTy?3t>V63yW<I#RwA)tNNc`3lOLq$*mqM@Az#
zNB@Xn`?&2gzx!akT#MGB#+suzhx){Nt9#smYH7@&JOc6VJl6MoZ5YsP_vW7J(g#Ri
zRL*AGptw+_-h?29BNcarGDAQUnjz<L_IPU!T7RGvIm#UB&djaRK@K_^bnz<V8mepS
zw8qr%utjjGq(g~duO+$^QouQ(#vqnVc7G(5P|4S{q+BA+6Oi*-(n>z-y0WcWOc|j|
z9*DE-q~GMTcF8?Brq8L2jdM#bfxHIWdtp|r;Z2L2?^T3CGp|@T^%@idad0GT4!zPm
zBh_Xqxy2cAyX<<E8(P=ncPORVXqU8YA;UWH!wVg{r%c6KIx<JTgi`jD=xU7F>3Tyr
zs98_BB!($7kQ}qYbhWr{(Z8~Cj3;uQa;(65crFDQe4iy2Vsf@gG*7{tI#im@$h>hh
zoC?m%{=`VlZ{$a|WGYsmT-Z$IQ|I55`zwhcP(iCq;_JZ9`9V^7V-c;T-Dz^H8kbf}
z+;lEM{jo3g($Rxm%#To5Y;1R^Kdhh$Iw=OP)YB>RtG!nNo}}vJ>=Ce%znGpEG9|Ny
zd{TuMZJ148@vNPHs4LoWPgISKWCD%HX7E}xlNx%eCH3~E7yS|?Nwr%kbfK$}`$29T
zB*hgSoM76utiNVAM^X6NQQ2QKQ=6mUOQTw=*eLM2PS6SFYiI~&hr%|U!!$Fv>fS3;
z$yjKQ)Lr$eEmTQBWcbk-+^h57Z*^-EB#38}m9X>$DrYaXJmIRPqSVl@L*q0PFg!&G
zlM$t@>#oHiqq<dE)XK4W<Mu-^DjwnEBpb(X%vY~BB@n>m%N<^E`RzTZ)^#9gA`&H~
zh(J<@XZa4U+}9JPb*Ro-t8G6EGjm2PyvQVzF>A8~hF$>M#3sXo*HMq&lqK(yLdP=j
zHlE<XQv7i$Uu;@F4M<YIU-kP`G0V?_r)<%f6#oSknz$!0VRxV(N6K6htM-z-yT>ak
zY$|5AO?Hlv9NKzg$9hZo7TFn(Z?a~!w+)GA>1YDBL==VswGaM7twU1wWFFJp7glc_
zQt%8r)s)J%{jb>Cwq#B!d`sl*RL!G{Ffv$%Y>T?Ash;_b#D<(IJQYsUTMbCqaus~Z
zr|C$1?)UO(6^R>6N-o@wj(eNLR6-1(QnA^Sl#Xh{l^1$OVAU=?8xBtX0hESHt<l&P
zo6vZV#7C|ABT{jBJv7n|J&^lNvGZ6#qYWm<>YTM*6r-;W*3DDis$|W{bZUKG3qp^X
zP6+xyKPiZFYfhZ@^0te2&F)aCBCTXDH1IB-&KRzY_bT)EY=B>j4QI&|q`d6U9Q{%*
z<IQ2exx-P~V=%`16g%uB6>H8BlEbYBco2&>{gNp<U7S)jZl;T|!q>_H<K%2xmiJMR
z|MGvp;+=^~Eve-8mo=-%T*j0_RQqg(^_cZNp&}pQ!O0&7uL3JWrg258gsJw-#RVE~
z_EdW|`?QO~4g8zV#d=0AbWqsE`u?%@!?ZC!fc<{Sl8zyQ`%w}}3<s49Rf(53Ez&&3
z`%2?8%pCR)vGf+WPV=_}&?Zd}d>3EP<yZT^!;<C-7?7fHv<ix`Ga6PBUx)~1>DNne
z(%szU;RY!u>Xkl__QTVHRyN-<S;Yrm+dk1wD!KZgL5Uyo_CwH4@$bMS0!GI_-#6OW
zwZf05IB;=-Lo1w%HQ~@Y0=ny3dR<@)H(_NmiD;JsM2a)w&WDbuYNZdT*3Cx^Hv0Au
z8<mv3NaNt8Dr0C$Wkompn%l}(yEktL*pInLl}&Rp9n*dUvz@r-#=1LmY_9lG&pBNq
z0rkaF0(#q<#gs9w(KRjwsaUy2AoQuX(|fazOyLk5UW5RtMC3>I9{<hww}`~DDp>dA
zOQzYCqTX9hD%HY<4>Ton5RE8;cq19epe5ycu??O|QPfsjKJ!Y>6|N`y<~bF2w^UcI
zO&K;Tjim|jw2tM6YBA*SrWiSC-^nv0e25N`sf>FpIjZ2bsH||xmd1wXl!pSE7$Xx8
zO_3JP_%}$_m@HMR97%gW47e%8uz$5vOd|A|FDCEqV@*16B;2eexMH#IR_8Zg_Htn+
z1wl{w;05}?<zAx$bg_gFmTAN$Y4IvFU~&F&8FlWh51Lw^Pl_FqkQ*IKa>7k-d=icI
zOgrmu!nW=^zjW}k$e=BV!%%!;{`!%5X-i^Kd%nRdCqscj&;~srq1iUOFV^?q%@z>0
z`jdg8Tf^h8{^VI@6P4o4-cZsc67u8FW<JT&$p{d42CuJRVTX7BCb36gtvgyJ-Obdg
z2$L9RTz_8MqcP5Y2-SA}Kp(|N*jD$-p-#*gDo`g4NFRWTp*(7#DV%|!4oSlapuXWS
z8%Ls(CnamNZskpVG)2dbfRjs2kw01oVV{TZ@69YFlx<B;jT&m0VRS_kY$n)44;jB#
zx>X}EW(PaO?H1<jOQNYe2A(B5X~?;{+N!-mRJWTwYZoigq^~=CuWZbqpcy0A=Li!X
zjG7&cV$$}7@PMW}r7Z&IUUT8ob<h2rlwVX1MSrS|nA`OXHIP+kea2%oZP@D;pMRe_
zHEHWL*s(+<wQ;g^&uhl;s75WLEDG5giPiSDrCo|R`Q%k+5^IEsrk}2(&Xvwdk)?Y1
zy>CZ26pg5kiY0<`Gv&ry#LesG^fhhVgUP}UnK8G)WzFtPV_p&tYKYq*zfnf0r122g
zPDw86Z31#uPb;<#VQFRkj%Uvt&0@A3<D@4>_@m^2=GQd5X7B3Cl9BL~)`>yD5jrbm
z>GGKu3gTR|0jynR-TYDN(XL<q(d_VbG!A`#&NNNWo_5ckSkI;iRhA=cj+bk0pXbYz
z89HMg0t2eWss*w=dhwX!Wz~N4qKL#oHURzNdcGU#&6V}qN{3_3mKYbj3;Mb=N(dgb
zXL(Eo_P}{KSgSQC!FP1Eq)FCI#1+&W$F|S0LCr%2VV=k;1K)JxNIuU0F*z{7J=>A7
zYhdV<abD3Bxa=1|dk+exfH9@#=*;H~KRETH%x411`aNa?Az<4UuLtZ*0q@Jnmdk}*
z75<Q=end-AQi;%OTa{8%wDIfPVbA4VGpI!7!sU@m|FriRlpiD!{#)!E@94UFwY9E2
zQy`|2(I*0pBX$^v{$_>>uRJQFc?`%5V@?W@wG5+2d-4@Br|f$gRuZrJuOI%bO@cF(
zG8gJy^{8G|rI=a^Q(pB^Ics5s5Oc&PjkcJ2wys$mc=j-RSGYvU(hv)}m(dGvAAe)n
zmsv>qlQ<b;Q+Zj;ALzgZ+nR$nAg;>2v8+8l+FzYy6?Rgm+9sd_jl1lsCld)gpRtcL
z`Au@?&iq`t|B&)A8Vy?a5pQs9dncd3Wvlz+X>4Z5D;;v39((4GMr6-~nFi%2bSQTI
zD1;8>5-|7|SQvt@q|8eZ!&X)EMTQsXw*&Se75&LavGugv_1U-VSDcPGb7|t9&z1OJ
zXaG${B$QzLOxb<>g$8%pp{FRda(*tBDt9w^%GJ-ytoMPpD2dtP!Zzw<(09g|XMW5H
zwL&)G^Sfe8%I8T!-n~phg$D9c;#@7?UBMAhg^Qf!>zQ9y<O3|7P{^0V&(SR}4Lbvc
z3ZlurDbLQ~`~zgsVRfY;Z{edVG5+TAn4|!zk#c!NKJG8d@B;zB4Vs_f9#Q_q4?0Kz
zL?F|it82<%q#xjH0K=EsYw?fufVn*%Uu4h$K2)l{dh+j)M#O=U=J~`u|0XqcV*v(J
zrZGL8>hF;_fssbGSr=sfq9K3B0*2F@I67qiXN-T3RE-9V#2GcN|KGGL>B}SvfM$)g
zGtd8fq_hZNq-yu@%fGo&2nFz<NzS~ya(|Cx1B{eDY+RG{_k&Ubv}IGy;=JYGBbk!}
zBiRsVxBUH}|6|hse@u#??<4YGT7duY*#FlaD`dCY=u+&bZB|WzS_2<Xr$%Q1j&33U
z_qr4Wmh=_*hfKT0YIBr<1o|Bn4t?sE=q%*F_x9is!0+Jyy(}u^fmJR<?gxB)nzxCj
z`nX1%vphKR9K~M1U&jG!Q1InKWBUK;wlahljx^QfKf}#0c_1dZemk}05MT0^3BQnf
z0mv(8{!3o5_#9N(d4Lbt#NjuVz^@7|v>X!tD;fy`EB6lLN9iK%|KT=iK$G8#Dy{z|
zbQ}Va#EzO~RsBq%>mf>K^vFlI54Am-O5ddEpc(`ZYX*$;b)L5dfSy2X|J?>WE|e7I
zS`8xTdxw^;{$N<IrqmhKv1^5QSY}_kK3i^Zvc5$i8jOnlOmLS0wQ9!)4v-j96hNil
z9cVo^+}xJlN#3gv+NXZ-*#@<GJ%kb_E8pxWsUN4;8zB}Y(fb30&AiRs%yJMa0bRlX
z&hJrh-5Wvg$M75ax-6esbT}fOc)e{E$YMZ9!t<QI!0N^WM)xB;W&jFtyWwJ~@g1r4
z$BMN!FV7F=K0bM0JsKmW*T8zBqzhd;7+!cR_9Tp#$a7<s&*NQ@2wVsspGVHa5%YcD
zvp*K}60J3R7%SG(+Z;&gG01sl+MBT!*Xt9EQR&51Q*Gta+lWa7d_?cerb}kH-7nFr
za-O;s>P%=BU>6f3I<l!p$8G;0m1OXvu5|=nPRy)f?6BA+Qme}zALN_EZpuCG6&oC}
z_--$PNusN5WC%3{rKr6xF%iI?G?nlE^)ic}x0dLgbRmdAi_hBx0)X{$_lp5&xZn>y
ze|BkOAJSeh>$eDnr06N;%KAWN%fIxYbryf}zJv0k_e7O;c$*w!OJ?#EhWs$C@qwN)
z(%s_7?h=#6XY%_76exi_78Ai84>yCeJTY^XUNxDAtp1+ObK~RYo4Q?57_S%!RFUF4
zmfO595J?Bkl_7TlpyPj10>JzK?SpAT5E5>8;^n}hU-tC9K1nd|t^27C1sw2Y^3H`T
zK0SqH5kN#CQAe`_VGi}8TCG9~=N3B8ba{YheWitFcD(qCL7RKf{_tJ4ao9dhq9DD1
z*LART#NDSAh+QxtEli+-eyFWyB$;KFo5dugYK3{+Gugy#WWRjeSBqA&@scazwl@Sz
zj)WhggOT>Z3Ep=V5(eP1_1!&0$s{)TTd-Y$3%&<ZkgU2-C=TMDGUa$kZ!H0s)#Fw=
z?Ro>9Emtm3fM}mwXzBaHN>PL-i&H;Uv)2W7Ege>SnrWL=m!5x9PNy^R2jnLHUiUR)
zyw9t?=L}|y^Wk9{_8{|Z)!l{0#6oW!@4jU2TKKI(L1yFSnOC>=hOul~!t5nV=mBSD
zGi_Sc2FmUw7t<Eb3mwRCXnY@1W*PH+Pj}zL3sQ#%y&G;TRdyy}!{|md6B+l?ia$#+
zxZ2Ap)BI|)yD%X)sozyr+7tz{`OczQX~>1&n9x<X+}r$zXsjVdd})!rC>&sntrla2
zH^*y=rdcUx@jipcaL`l#mCeARKL?9VadT8??5%R19T;o%C`DB$ltM7sqUM+jFL{DO
zLZn_<9sY)5$$r(-f5N5F@Hpip7@HCuFn!>kwT@jUSCS8}qgvyX%G~Ubj#e<uNA%Rt
z7-If>HO5)*Kf~FbsG=%+bb71zcJx<XfCw6y*9Vn_?)1ATwd+A?q4R<mrTQb(bWF*V
zDA6SV8hnBGTL2v~3mV^cAVWUi37ADd!#B*Ceh&VyoU8amBOC}92?%epq$gWxRM>H>
z;+g#u2waU<W2!ZIYPlX8uTiXO@%<mu@j>4E6yMo=iCm8aP0^vTZRxq9P9^EQaQ*1L
zT&!$3P&TQ8NsnNJX<>J^J<*T1tyb#3N2#R(-(MwNrhtk3781gK0+NgmfR4nBX{De<
zz~|m%+9%mY>r#!1SA4$V7<QX!G3&wW^pFxk%{FRRiF}VjbYVz+fhZLRxwG(@abM!t
z-BQAl!szb((=UT6Xvc7VS%BEdgtpH1G8!ux@DKMj0tQWo3$6Agyzv$}TfJe+->kaS
z7FzE$6*RIO{_kJX#DynH=BLx8T0jj$im=n~uTOk|3TQ>9>^N{a=&^(X(A{kiz#*c~
zSR~u5hrOSr(OyPHFQl*TP=GjyOK!aFNWcxh8rxpE^24V?7|{^SmxHs%J&zX{l@8Oc
zOO0o!-aJzxJw|e@?!j-BDHU?AIQH0;#k?+M3uQYVy5?nY*$gXbx5PkQZ4@%WJsdK{
z4sFgyR6Qsb*;koO=P3AK!d%z1HZLN8#P7_Q-nd8+Z5QLS*J2YEyC0mjXD@2=eV|~_
zs>Yxm(d{~9)vGC6I@lH%qgdLqSJr^F9AS70i%vxZYoRH|@YNO9o!p$^3{M*H6;j(R
zds<YZ%=q8^j8H)(^V00|Co|}F)#T1*roie>e5gMbIG@ls<IEKB4f&?FU=6)F4ti2t
zh0AAs&|dHCRmhWymt6s5TH)9i1F3Eyf3On+9R#b~Q)ow5W6NXcsvX4z;{+)i4M;V_
z(@nJKF`@5ng^jBJmB<Qxd)_mab{RUBcORGEDzG@O8}WHRN(Y#Zu2sjd=$9NjLLzR_
zKOawJ1?JgF!V^%juXl$SX3^4uHq6~SoMv7;ECS?1Oq(O`p$-=srL{|UQS!rB*%ih4
zktWG|Y}+YJgn4%Dq0+d#s_&Wl<OD`%J3|(L({ifhcu3UBfx0)$JdB19Uj~zK<spFk
zJC8T@MKV3lMqa7f3;Zf_SZ%$+<v~e%T$bWc39Re=rnZR63M{hQ;G8o_iUrApm^y+w
zlCMDd;-}l!k~|;4yve%3bDXO*^cY!rwSyp$lpM~xPqa%CvRZjD-r17=UOj#C5TAn#
zxjP{ETB|^0c-i1%5?)C8(SGNB#iZL%$Uw;`t<FwS#8k}6zc?+_O1O71UasFJmVsdF
zyzaEOFBhKC+OcWSU?&4P*5O4{0!fT<-bSVb0kQ<~;ZQLk;+!vcn*9rqd}9Jo$J?8k
zdH*Id$HAF{uNXS!7LSlkRpzv*&V95Pf!qPfF1-oeA+B$s7L3une%r!fs0rs7A;P2j
zjamv}A=n(Td6|7(WFjdh`h3O><4ReVD8&3CiRZM602RNiOW*D2^h~@dJMY&Lg>LBN
z`GR%V0IKrORRnKdUrP<S4N%rnKm+$vwq5;(f8%VoE9M-vH!H0oW^>W54AIF@Hio+b
z9}CXatSXD@ygP%h))0w&nV^Ou71nTn#TeWIu*yPP6R(}wKp1gWrid2OahOuvbQ*2c
z@;0f3#b}AH^XO2Zs02X6UBsZ_*e@6ufB0NbJedVKFMb##8=$_=@|Ni1`CM<w1AiR?
zxLRh#jaS@zI53(KprDu-PIOei%bH%Sv>WunnvT6Y+!lBd;Ak!!w_u9OrS9D9en$-C
zB4cSTnf`a_o>7WL@?78Zm1Q&PvBNZuMg8z8$E$o^FaM_5V?XsG&su8SffWr!eV4A&
zNM?3U6P@rP1UAgOXZ+h*M}X2-K^INWT6*=Nf)tO9I0r5pwo_=XAbJD1T!YvSE6idb
z$mTL$>Pt>^Uzag-j!81*L#v%^Lc&WChzyh4j51?|-O8tBQESYz;5(9?vMjsBpL38;
z1>zT6M<4v#F|dtii&%M-DyH88q)Zyrc+h9CRXcLdulGr~_fdr-!1VFjE?Vm2&myAp
zK81dlC_ovO(cWr$r8F#sLd25;Rf#4S&PTc!Q^k{6v%8VCezxe{{c<}##c!);G(5a<
zSndk|@VlT4t}7nD=VbxeeotUtlP@2G`>KhZEw@d#!BvNgdLU4e7PK#Z<suRCZQdIb
zs%!w7I7(5UUUH|EkWZJW9?7n>u_DQlyQ!u%SS%GbRFMwNnd+0`0x^hH$iz4kaMMys
zBqx4=7*^$VutPs6EAXH#n#RV^TU$j$D(H~zKzt|IQE5+wbF7AMQyq^MFEfGxp&AI#
zJAP+}q}vG}@f$_Tqvl8=3=YMlv*R<;(p8<3irso`FJUpHOsP;Kf^_e~Q^Jl0#of8d
z#N_w!c%2gV2(L^JT!pe0l}!Y{*z_cLJp@H%&L8^SRK#9x`pgr2!c1nD;u&R@@~yIL
zo2or5c5hlA)ZNm~mXYC_DpnU}4v9{Z>+ZojkBwz_UPmuku2SF*>m!-TF3c0vx}3CF
zyid-)g`|3!HP;la?%t%(kqgwTf8T)qV5Z=df%*FV1;8q~zLP_A{wG~P2hxQJm-#~)
zdpTqr&LJ=}2HO~}%O)kyPhp=YGc_GJu^K49GybH6^EroU$cr$C_(^GosO)&;$cE}%
zjx8yIm}QRYw;=aeSOEc<5rU+&dlobMGr+Yrr&BXqLC7J<nS#EL&|{FJ>FIcmIvz<T
zb@A(}lW9$?XoCC2w}U|oRjb+t-t2Q(-!`uXaVt!Wi;152;yRY=R}8(;9ekcECdVR|
zzk8nxBB<7F2v@Vmf`-s`-@`gOqg_Rdc&^Ij3ttm4G|y6x@F33?gp@F1#oXQsExO&*
zDY{~Nt7WHMc5wxrkfT9~y;UMT3TZ*B1%x5|X?S2oo=<K(9td3b?~*uR1l*x~8Ip+p
zND7obi2DeIonSbgqgg|5Ie_W5VGGSDOHJn+R~_!p#Y7{}UnWoZEdH=ksNt{p0*B!3
z04sw4mQXZJr9g!MsO+uX(^%s#v1i2i_&jY*PJ*20%a&JFP86RK03rL^H}G=`z<nR}
z)*`7jD7|ipEQ~ZDgn#uFRyPU=90xaP{Fl7I1lp?4F6DuFd^N2WN&#M7v@0{O1Wfp|
z()`Q+GXVTYiT>~Vr2h*7)JuDzVyf%)qT;R<1i(6a-Aj#200p)SG9URU{z3gX<iDz~
z+9eD#@{D-r`6ch9eq8~s;yoG;hWsnQuatK@JftdI+qSzml_C(3VTMjAn1a2J=>`_r
zzkY%vVG3V9>eo6dH)!)~PLmo@t!HpN{Mr0+`rcH+mFtm2^?qYXt=aA3&5dKu)v2@l
ztAmsGI7)@8txvy-eqBgPRQ+&3(QVD+p!h=$#DFKb0WWbyvS$FQz!T>1bbCS1$92>_
zM49@m&C5MeY1aHM*vcE3!4s9yX`Tv~K=Ykw_#m+pSU=BzkkzdNWV7&Da<zaq%lA`~
z_$s^(*VR`G<zn{D6%sWlzt^{NgVty5Myt;zu-!UrT)sNDG#DLh0ra$;5UmVN)|T?}
zMWkH<`HU8ShCN^ST{GJ{YWvEEyIZ21n=SLw4M}WFsNhhk7b-j~{xfXu!D8?Fr$3-i
zY6P^jQ#uFeJbgR=WDLj9+B$l5Yos+@qp;;j<z;ydLbJc~<q$nKxTA#49(E1$qOdrb
zP`lZ!gsBl+8C~v<;M?L<5+b#uQmeBwi)OZYMFg*Pl8FZYeg}H+uEsDc7oX@lQEYW1
zHV}MYfe2pHD%BD^)D&2l^*%d522=&zmjI(>B)$gK2yKo6O0?L3KSd&=S7ZVX_E6%b
zHi_roB--zkU>#A|$wO)o>AZR|iV9T;^w{N)@}$)&MszL3f1b*30KDU_6gSlC0=dbX
zdIYg(eidnk^QFky^40ZUCvBv(w6C4aMT54dS)DtK?%hvFFb6Zmo`}e`H|<?kn>#Bf
zaoEKuof<KB7b^u~DI-zPV%wlR-+nHqrex?fd?t}47=P*cE~95GM8+BGqq1}0g-u|%
zWh|hSW^Ak??JhMkWG~QHe!6SVHs2LABfcbgz$jFq5U<MIAlIE^5fESM4%Cw2+*qGZ
z1+W=bV_jJQ>}O-E)E*N_b(Zjuq2G3L(eItC;#wx!>g`vppCU6_zW{33EU}hT$z+C~
zGYzgAOn{ay1jAQXIIpIEEMLT1G#vFEGJoL4{snI%T`WMXhaWC?Oy&Ix$znR1%4HuG
zaUDlRG8S1ck_dQ-=Jt5A*z0nMMq<7s&u8j{zRGu=*i!$=vailqb`-1!6h(gh#PBr)
zn8)4ON&#^>okDuL)1gMSFaYgd<MMmIp2(LEWo(KBsfW{8O#eWa#_3-x9u2M?2#4Ge
zkP(o+%zrM(Ix}UrKao90rFQU}HJ`^*7fHIo!lUXE@h--IyY>xVp{}(`#>vlH3${~M
z9B>iViR$Ax?>#V!DG1z<B&vB#In2HG?=QZ$TD&rg*e`);&^wm8K3nHN+FKFrc3D+A
z18^UcD!tB_D)<x?h2oL+9o4X2n6KT+bCO%RakcSh>Ap$)F(US-1;%}&+Xo}juy69O
z@_u}R%F^gj<ZmCy)VM8tyP#Vq_eTs9CldeUP@l)6Wx41H0u=4ROpMD8YXivV_yTXI
z2XRi-0ZmQWqz4>din`cOzGP=~a**3}$s^=;ln86+U<X<J-*^&hQf(0XPe<=3GFomx
zciWkk+qMqq;Bm(q$N(hgPbYVmB|ENLpwx=0ZZ{X<Eiq!$AFmG-na~@NQUZ|?13@Oi
zzj=gM2mbUA*0ceDSjm&`R$jbey}{+1)Ao{DAB8mN!drZ#d-*QPC6*%BW%OiszAQ5D
zJ)Y}@g^t7>Hn))*gU(gzZBxF4-f}TOOq;SF^Sl8HCZe@fub+j`Nci=rz4fvZyJJ7^
zFG8{$u5biaWMSl=`3Kr%@$4sZASk?pIY#Qx^LHCHtgbM>a4W_Xg(!Tl#y1Vy?f?LP
zh4(S5v%LyIEj`ZkE$9s;qZnu%d0G0W?FJnj7Mp)k9Yes~Di`#V^J=2D`Q^m$+{qoh
z%F1<>OVgj;n_}FV>|}&K0EB+(_)({Q)pK-510<|<%p|$v9y%R*SsldQR6jKVxS*OG
z`H;Ua`qQI9cwr{In*fbg_(P4ABHFxklQkxfb)Y@>#?hUe`pL7>I0yS^2@K=Z>pJZz
z;QROusD>Gx<YYixj>!ULHEpC#_XJF+e54W;sXuIs-Gcvg8F)m+J0(2|6%%hv<Dhrr
zoV1zt!}%lNp)(t_eJ!CmG!TRbaGmHNbfe_f=Bj=2d#xKrqWn?BOJf{T=^QxCRv)g*
z^333X;<}BQ^q}44R$?|R`=UDY!SBkM07b`?jv5X<C|gp<Y+5%}u5mt+*M&Gv$`r}o
z`KUNJbeb$jGZ4g|soCU@%B9d`v<WC-LLWPNQ>A{<0^2Tc?xMxHujun`{2QpR$rQ=l
zyy5?Ba>fpjfVaHpprw2!&8Yn?M3DM<fn@5bIYlLCXe8r_04Q~dCTDSv=P8-m2i!HG
zkc^$IN*=FI^Ev<m*9bP5n4KdY2L$80dgYh2`AxtJlqv70S_tn>De6n&^IU}k?`;ff
z*t0Ws(Ez0(@lt6qjtdFcO3}nzSslpBs0_X<U<t59p$af8a21<bmAe+VTWB{i09!*S
zXqwlGU0d0uXZXA;W&RcF(<(du#+$v>VlMmfS~a<cxbs4fvzH&5Ekos_fskBAkGyR<
zFm~zouw5#*_ITfIy;;8W?rug-v|n4V(#lLwZ#o(?8!wD=zsGQX>KUDe+DWy<fGh=#
z^UAvx8vNp$a=M-E_RKtLIM$d7*)po%eBZ|`S|ENe?#h;XdwpRn2J9h@z#fv81Z*8=
zw~>iYZqHtA1O>brl&dApK3m|r#+O`z0G>~?+JtxpS#3o6X*QX*(IVJR)rLcSudl}`
z+{7SH-VbF6GS(dr_WftT`<z{Az54=mK^fKWd`RQ-#B)6U2ETlfVly{nP?(-Kpcw|K
z-)DWt&6Yc(_VUbV^SV5390kzTPsn&n42u*flC@w#LC|A<H=`7XK5vk2@40nujAw|e
z4Akr2kg+(92(53M0YHT%CR5hDQ&#&SIb8FhkT4>)uQbuTuj-Bj@`Swx?p+!|q{Ren
zrG(ixNAogG*6gc&{RZ_eZSpB*QF~b(%S%j_lPM1Me1<`Q6B$t0B*N-Gw<schNY+DQ
z0W9${aKc(-gmkIfXg-|A$aWo2^QLh&gw8Bp5f>Iz4}Io3{+Z)S5l*6$J?E(V=@`uk
z*wUScvACEY(V(#?ZI)|_-XqOeIxbFy!A|5{zKc~#!36a9+vQ6|F}5N=piGRj;c1vv
zRKYG;dV`h^r1#`N!j12p)a=?nm_+D1e?Q-<<+9Xpg(}ArAv}Js1sEo_NTvQr(MFd=
zHG93o@{fP_T&-33w%eeI#eT&l@Yi{ay2FcT=@Ywo*O(-1rKeU}^xYw93|YlwRa$E5
zywD|2f93zG5N_S!*vtJV3{AT|W7BpeiE^vVE{;wDVI+&fuV*<8;=4UPlH<D}iQJ&O
zJ9n0GthKuD7!pwr=3S0Y;ac|!0UXF~m_LA=5uaO2Us6|xy#(k}{$Gkb`3^Y6=1eR8
zz%ICJNvXWKiH+vI0)M46ss1Tu$8{19>gXud7Kw~XS)~zU#k6GVw+p}~XO(=+kQ9R^
z_fd*iMitY@r_V~GD85&Z*?H;Vd$Y<4J~0X3UK(S6dZ>GK>otGv0pPM>wWQ#tv@Q^V
z;+ANVkRCp{XPzpmr&x0kGKTbGoO2rA{KP)R54}t&kHD51wdRlNSgVQLPzm{r-nr!B
z=FJTmpSF0Q4R7y_qWef^@8}q0Lo%IE(o-R$MYOD>kk#iS0rRV_*nsA#0`2NY0DNtg
zhdu3@NXnnafM`GTnCZ?ddLEvO@fVs;{C3q};CI~TOK~AC1?ecKhTl7JbLWYi*%$Ix
zvDWrh%s1j`s1wm8jg+UVm~Ir#QO4n)v6D&>6|LHRpD9(z#bwKp&+{ap$X}-EwssB7
zIqk!HdZd+2I6>iGk%1X`-hj9t#EYK@o|o!6lmQKV29=R_|MRI`wqaD@C)!tP7!E)s
z<>CM7BR2WzVwsfMa+)Qh6=RH#)T<aTr$Nv+HJkTQ(lX09<@yJ7Uvpc0xa}VYlgd};
zsR)dn%w@t-u)+$Y#6w4c5XG@@ya@7+e*b4!Au5KmYo)?a`0gMAXw@;<RU~ss=*|au
zB4xC=nI@zwT-|Q-1T@)OR5HQCA_G$s>Wn(<j`}CTAMwG0lC1jnkTluN+p6udkVUr!
zKV_x43Ld^yuSu*rc|~%(BXNPtDZ|ysS*WHyDSb)E6EQxizHj-V=Z&Ty7Zip|8PfD6
zZe9$%0%%DdArpeR#Q5P*pP5i@b*l!s+b0u(ROO6IYDa9bhJI}sj9mcjWRM7YqbX&J
ze$1;`<t(_RTp*si+lRXY3R}-&hL69WnLfya$N>z_bK-ULzY&06)PQFo#HIKo2;9HZ
zx+$UxXexK8WP_&x&fmeJJMaVd%XXYD=Z*ms^)NEB$os$Xbu8=8ozjmKzhgV}<Q3on
z05u(_xS4}&ogWWpeq5$?`?X^^lRU1z>Az_Gqjgx^&)}1nzo2)aAEPm$K>u6&?+_VP
z3kAH18Ca8@*0}Es<te@o@}oIrYLzU2*z-s4QGY7STRF{dK&C)uwFzf2o5O~%DO-N!
z0UpM-MF1N_h^SM&qy((?^&?D~1mHXuUL(YJ=^vvppc;=2ej|Qnq*5J_+pKiU&c~~7
zDMW3VqTe8_(c4K~K&u~X5eG@|UMGJv&<}G?%#HHdNL*y_&DP0m_rWf^!y7*HOCg|P
zs-*=0BUkOcyi&gaeFSeWEIfo5u+r=qx91@kMo1na2=uf8yiTzw5c;5E5FI0{2}`7X
zRoZVb4QxL?CPdD$*pnKGM`cYfTxNvtZy}H>T-I=w_j?NoK?Vn1OcFrC{uSzzL7QSC
z%=$i_pJPho?`OB7oHWSx<WGB4-O<KpnGG%|6(i}uQ2vf1&!`Kz+*Vrql~y2mBP(S~
zNRkLZ!t3b8(U5wga2Y%eP@2(yG_gyypJ?=&q?VY?bpW#))+xYz1rWOD?=yHtgl5=p
zX%1Y9B-}3TI%`bjpUZ)ED)r%Azp<afKaDn8KUCQ9eS3|VCkHT*NVKtC5<i!cZS;%v
z%YIBEUb)lPed@`QTR@Q_DblWcaqD99<Ehp2i8WAO`&&!U$Zjg{mdD~91{&XgX#xI4
zVD`~Nfg<64?o&Q^XeQ|$cO!#lEi%x<`Z4qJg2gA1dZpDfirDtx!)B#M*>d2{_UdN<
zC#W)|qX4aa;>{me&$UinXP$F`CwZH6LyuQ`lRM1gmcQ5u1)Y={I0}@|84omAmM6he
ze1{9Z2!@^h-*mb`dDh!RbB%X2JwuuyN#!eaS8t|evb?LUu0zY8Gdb*%1-l}qM>;)<
zpnmff@uBtxnOQV`D+*^zHH?5W6Gu1#33ll<C|r?E94rjiON6UmM<m)f5I9JV!_)l@
zq6bo@-}PssAKsI6&n?*6&Z%G02PM8YlQ;Y^6b=%OX$FjDt=<$4bVTiew|cxQd{#aI
zeJ&`>1}*J=5(+VgJFxiUOiI;=Me+S8SGQQ_rI~E=XfbIlJ~miTa?kpYw^LGbNDSE>
zk@)-DK*BO>&<68N)Xr|0Ot*ip!9z~}_h#nBJt3s-xJfAuz)N8uy-!fMYmMW;XTnGY
zYR%-%yNVvx08eo7bXQYep4DV{w+0C563b8-4B8NNT#34;4382_0VbKHD2lhrTzae9
zr|?Xr`TM{Xz-rT|)bnD??nVN*))?5&;_1>}0X^TH074&g|Lsh;$fnn8z4LVvmDv@^
zjlDQjaO5FEoxjI;Fm??f=wjQh?+a3CmwYiT1%hB<P@XNgfvAms4G&n5gv*+P0>?*J
z5R9b*v^EIO-?miv9=WM>x%GABUUM4~#?!`5ZrAc8gz?lm2s)nfH6kZ}T~&{KV0vvf
z6Dv38{hK{ecJCtnM}mBq1Ic1}0?wKpyojhtAbY>6h4d!Gth=Kxm7ho)pcLZ1(3v9q
z(2JwXTij+^g&CFeyjZovLbQ(88t{`C9rGRHI~2U7R?e3UHxcZWfK(W=ueXjl@d|j~
zme^~ycpswy;K4u6yqr|I%fUiFpAe9P<=6%@&a>rp#d7ef`6n@F5ZK~thyxoei(MWy
zP#D3%c=KRSUcL0*cI{325avJ}{M)F9LNxnq$r|s2vTP4Vg#b={f6t#qN32y*ypH$u
zJxQgD0fWA)!mE)N<IM-SSdWwlW|MgU^Ov(Hpr_sFerK!HH2sO)nLY&lb+u<9HHIHh
zTQ%MZWep}~M6Zgshy+X{RD!Eit3N59t(Y&(iaC<X{l}N!vs01Eb+CBs$B)2{f?|<n
z!%Tw#^qvct->4t*I%3DYsTazDTbrs!w;k3g<X0a(w3haBI2L%RHn=HKd7o^nPp?lJ
zRvvJ-X06anq+lovE3qxQDw+hQBwauor;gDL*$@39AG1k4NxpcM8V)o-X8+WoEj9ys
z69Dku3uGEBrUh>WyzXId>T~QRtt2)j*TxGvCDU;!WC%%~%PE$KGsc>uX2GaT4EQd%
zU)z^ctZJHUk3-y^*0Mtv5jE_7%>Ln~>l1@0{7cQCs>JBNETND-yL-_vIIC7e{NRG$
z5eQ#8+UoR#=HWRdk_B5?8v(GVS-K}n&yaJ;_Zrrv82f5swv&3~r*|7IO)T?!BeAH^
zcf(KAOA(=g4xCYtK|1llH&}A(x>Q2cj}ZSoavr+lHU88+fAN7lqM!<(h_EV$KZJ}H
zC}!>4u`Sl=T-(_!S>yZ1U+h_zAny>`;L?$tUM&|_>XW@gZd#S8*yPBX8L0XLG7{|h
z!{Jo~=~oh<Q2G$GzTKIY_C*H%N-g!$s5`LbJLix{lvg}DgB#28HJ9UpG=M%~Z1L4P
z0A22Q*(j!s9-r*%Jn*c|&FaGEXcYZc3Ful=3lUo-c|iM5Xh2IThT>75$a6uk-)YvU
z^U^N`ad|MC*r+!gA>_|Yi7G6X^kyaP5@je;j4dtDI@-D0%Gff6RjHe2ArDczo*@b0
z%$LuYqxiymnOU?)9veyi#29nFztt+%XU?!vfKjZDOb61j1P>9+B5S?B&zdz+1{%NP
zkNulY{zQ=ouANll6zmBozOJ03s_WaH;EbTNuI(-c?@|C&6snb~m4As5kfc&$XxwMj
zYvZGoO5m^NWw?xo%kLD+wc(eicugO9(!a<O%?M?0$WUEPIaAPrn#~!P!@J`=v};8s
z8gtq}b4_(cd$}WV=FDpV`Lw(^B@W0C06fbGpd%!JRA!wf04R#Qu^Kzge%jU|Krh%K
z#~up{188kLzpR@W3;knqJ}mxLCv)H*Zd{aHnKSDf{+bhumMg9LS>b8Tg*o`dm<3(t
z(Anrraqzv}Cf8=<-aCp8#U^9Vo4FJyCHt|!%C!mgqsHa=T$zU)5**KW<bPshxjCmD
z6AZK$azq1qiWUYXTjR+*C4;W2oC?gLbfu@wadUK_xjs+HjustRK7kL-N{#N@DYA!^
zc4(TeF{adUfm$Jr^vwRZhXZQMj228O!x&Y9!Qi|3GhY$WYT0I>Sb>Vpnh8{jt8x^6
zcC`3hpwl_z+AMum6sz7c>C`re%4q#T%Ln>GNn{S4E6cN*@oy^mZL{L8d2~(d!@%c#
zi4LlI>uQ<fiSVp6iG%Nq1dvKWLA^R4o(PrkUF|quBa-DlS5}ZOiNSi!ebM(lyzf1Q
zT(4JLg^j;lEDaNWLxO8)IU52v-vJ|n;_C|_)gu1!VO|rJ8r3AA$>4!%5BpSIZ7?Eq
zA_F|jkhv4i{Hf7>=jFi6!B6TO2E99ix+CtufH)6L{;IJR)nqf50V~yIR466l&Y1VB
z#EurVsLy>dv=`t)?m!lO?(ED^FIteeNC3i3x&uz?v$~>(WAOLjY<VAmt?ghq{F@pW
zFX6x;u7f8@pe4#)e(+2UDp`vwl+-?;=|&S3jhR*`L3&FpSNVqhC|~s_A&m~hY`s+9
zw+pxP^MK3lsy<lLQJ|2+(O7Lc0djd3^SF9bk@=*3-<|xKVOcUz&4z6qS}G$>>H{cl
z(quKQ)ESI;Tzbkg!$@360`8(f!#mL*e*<|xv748W49s)-{Pv~~&;KrB@d}Mjm8L6<
zh{%!Y`oH+nJaPO$!l3yaSG=wWEWPaJXQmWG8b%w!tl5BkPzRmGP~SYq4r!IqZ%QjZ
z-WEW8{<j5yVC(bFiU|82JA&Am`2}c+AE-7$Pi@b02$E@R=6pWBZguZkPIP#3WxCgq
zGiPPS*ePk>Yb)Cop95a$6mlKFiJ!y@m@ux})sW35maj$t6{@w_?oJm<3Hh*?;YOfB
z(gmox6tl&A<z~@R`MHk8vS2?us~k+!@>aTC$kz%kw3$!IXeY%(Ma=e}Q1%yK$)5N%
z1l|7p=IO^}_le2#a&ZKA0{ZQ-k~53LXNLVz4%unU0j0u>cIg!oB?0JV{xdfEzqtF#
zu&BH4T|rbrKpF#(7Aa|Iq*J6(8dSQ4p+V^e=@uzP1d$pT8U_?31SE$p5g3}GIeQd*
zp7%ZHT>o$9!};Q+Uh|8+_gZVO`(F26Ki=ZE*)P$YgMa$e)n@O*iCK@)Fh^Nf`Y_$n
zE!l(cbQ3X4<P8;?*L0FEUKl|{M_Nmo>olj+^M97O7e+`|$*JgN&lq$IBvnmBtxn<s
zUDgkt0mf#w#kTN|aJ;8K%85eNm0F+h*Mn0LB+o&?EZYUmL`^ll$_707mvp3FVMMh<
zl1jEGLB^nF3M~owa_pN0mV14ZK!Y_sV5GAxva1K7eq&S+R8u`z`{gN<PPMjSx8tPu
z6e&a15vU-5^UV~Xv@#Um&!M)FO_F3h8znUv750KPkf5v3azMXNKWNt?-@E(K_94Pk
zkW8*G&Z^?TsCDv_jMH#liBphQbat7;`kqlowr{~@j~BURKtE@`LAuKD(hps-b6^`$
z?`1}piD-oV9g>PWOM^BCHAC-?3E_>uqu*a7?s9obgmF6aqpGU>NhTg=JCp2Hn(ncx
zD#dt1b%k%v^WL9zkg@cP^UAK)GEUz(4eBgAx=`;$m;A9UzJ3+xcCD(F%VZ(JQ`nEC
zV{X>87{)sse0cA;t6<R$GA>|gaN_<@Vp3_Ex>tIH8V3tqw|qG%SAK6D&2z+XrprKu
z-yPZvCSftyDllQgH_dnEGah^p3iRyVWcHN8+yy_Y3MO8TX%qlg65PN+bH<&|{dqyE
zcYD>)H43W;2m}`!Z(|$+MsuvELx+zA;T;&Bpm5f=%WoN7AQjCjYxhA<0>pJ<0@M|6
zf3o41N9QYj$Y+c4enjZTSnJHd(|j?@U4(%#HGUqstpW%bd00}0vl?bgr4c;Ce61XX
zR(HN}nw)rLODc;ncwzi$%m-a8WVR8Tl%7WN&=20mP(tHg5{7Hrph$N`H0!h6DHYm<
zOAG$1;K>2D>v0*X9y{RxV7^kH(Bu2}(7I*7h+yw#M{I-9bA_$FNQ&HKQoT*t9E@is
zK;|ykzdW?I9Vpl2f2HhD>{R<-)x+T0<Rx&bGUNv1Ti>(__lE7Qp~<K>lA$7#Klb;Q
z%IX`x4&oKfJyUAtx0(w%2paHX+kDHY#~HPWd?GP=Eyh+2dWAuyQU>#-KN`@jU;K!D
z<9AV9&O8+Adwf8a2~vxN*DBD{_4j>yEQV~~XeOyZ#v}(ofJy*34;!`M(m9=!D?Z&d
zh60@$i1>jz0_f-p?@<ls5djtJ;_bJFX$GIHZrUMy{KQt5O^(=9YrRp_qN9V3U&?4i
z-J^g8$*iNc!a}VClRIA_jJM(Fs=L&$PhXWvom%Xk#eiluSAQbDFyx}!R{@`yI+|bZ
zc+vcZU;gzSYLKj+WMlxS(iXPlt>T*$`U<jxobZh>q+ggq<SmC#^AXpS7lt+iS={l2
zfiw1-3pVl1q8Cg)vk+r5fN@?7rMHlcLAP$ZEdVO3SwM)JSbU7%;>j-u%3vF0MvllC
z@h>q<DPfyH0%p1y<GQ!FAj)P~Mk25#(g*|(Q`PRA+O~-BksM*`^5cmXX6zv25E9l@
z|Bz8~<l2O9<50<`>l>9rw^LemWm_2wjj|u8g3TNUXmSTbL#01^x(~od*$?kQ_Q}6d
zjgbKoMiTh=I)F?TlSMbq8A!L1g*%Cyc#^>c&$wr2+f@yfZ-|(bhS^8wYt<c%!#pys
zJd~;(=9>Cmbob;wO5n*3Wi-1|5J;`$*O94pY$qQ)NRg*x?qW$;x6rJHqs_?Y-@X#l
zgDwWdi|AAjCJPn$%Z9Uu(;|Q>Wsm%;_c|7kb~P1inUz}2QtLOk7zMtE$VlfZW0Jeu
zwF6RLt<2P9t}By(KA+8B`tyf)7C%>&!sOIzv}~*w)Dd^dW&M9K?ljPdcm$o8Ets5s
z7pMKc^lwN6%P-_DetZKoooJ5133u3BA(IZHB<Na>m)hA3EXOZp)N6~^{rsGnTFfVE
z6YiX&GTRYC%z}xo2Z#%-nSw%(VU07O5wB)Ib#tb7M)9|Z$<~CesDW7RW#SI&Pf?1J
z3F$OLl22y)7Rn&qN70DF{krk%=C`)@l}L;FsPz&fBmjG3JIGhU97iA8c3SCl)f665
zh~i!4#dRqpl{(SCZez%rE>Pw#>)FT$;u?sr#5h>w=(q+aDW|{lJFZdK;-G|+vfK|_
zuhII`KYaZ~><QVmRe9Xtv_pyXQ87GPNi38alt$!)0s)JJ|5KZkhdY6T>O|<;B)`Bq
zWfY)tAGDr=KX0EV{&Ygqyr4l2@2`8rXmK%U!?UqmOrKn7EQ+oSV^24~RSPDfM0|Yz
z@Gjtiu)mV6_4VO{bBz)u)6SYP;SLnBN7J)%kuQ{Z_0_!q5D#=tD*5n6^kA7c0H5SZ
zmiOWbJrHmPOXHYYqib$5pcn|Mn<T)^LwDh|A;ugA1+cIq4PV2c#UhOS&Yb*@jo=KI
zGK&!C0VUb8OYh*(Zo|9AePe_>ueINtb7OBwY3^$xxmL&AY1H?XyfnV?XEvwW69<Ue
zOl8w^K*Nf4EZZa3@l-rp_r!+3<VUgow|2)JBb_%6rzN6|i*XW-=s4|`Z}Z&+U9<Xv
z0s_gjMVe2a7HRFXD8&>l253^p#)L3_L}71awmdiK`L&^)#Y}|4an{bU`K}LjK@nB1
zi3TBUtv(DV&<P@dJfIEeD!%uABpu~1eg%}aZk9@|PN1MQt)2bpq|SYaOFlz{ONh55
z{}egwTWEKWOpW#5kR?Q(s5x89KkN%o!fv&>NDW?{G(>!_pM2UHLOu`98;j`oDN~Dk
zzDz}4<O|E>QsE{;f&+`vX|f3wqo7Fkg7_W~YTLUT9WU}f=R71!gHR9t@PEi&NDqVn
zqig(iOm83T5Qki+zq`Ob`AQE$Q~0L1_fwMWj;H59;M@n*lm-G7>m)wMm@7Ve>qLxJ
zE=>S%xiB_rmli?(g;&P8T^H`6``!VVU*`VNrA3k8#wx}TTcBrS=~E?!Yr5#&iCnce
z>J7U0<O9QUCd6djNr)}a>E>?eG+mHn=yS{8*iZg%S@6UD^#4{99M`ved{z_Gd}$YA
zniCcBEVP5-zmW;c88UfW38dw+&I^%rOH@E${)03Fz!KWchM|0R#GCj0iA0c6nYaDi
z@AnuVOWqQ5d?+ybTp>#`(DV2g<0D=pa$6rzJ{$J@8G&n_6&A8s*maOCyDu3d4{KT(
zUcws<n)}48KuQ6McLSl103X&{Yv&%%@mea)5U9q1ggX$R?o?_<5{wZuV`_NS^WbN7
zz;fR{`muXsL;w^b%BkFa;IH=qsWzuQT3~mwIK+ZDb*AXOK#c7G{2L=G9>3-udM_88
zyzYdEj8kHuvrU&bjcVa;9~ai7do^S?QgV~8D?DWWUm!#xg-#Ikg#aonsVL?5d>;0O
z5&B07JO}T;B+*IoX}*h@#6ivQ34egv>6cwa9{@Q@7&QT<D5hyUyKvFwV^CPQ8zlo(
zih^x(LiPs@H#_n5o^jKrL4sZF3*U>rHl%0Xpo|Onk%lq>>caeDh1A!|x?Vz|?VmGk
z-gfC0u~c$gf4O(M0LNPdb%{5c(<}TX*W<9S;O}3sOl%gFxc8Ze7+Vv)?Ij3XM8pAS
zgT!1!?*K#3%nIg0RBs{3Q~H^7xY%xrJj=-ihb0hszA=D7jEw4h&rv7$?LoTkcfKp*
zCC%PV9k#kvF5<E~j&hp}IJ^fO?yIlGIcNAZMxTOxO9<>+Axs&v*%l{7h5@kjj@K}+
zn?ODy$&3rMyN_PlQ#*qG5)B!6r@Jtf8lckoamS-ZQ{j`(7=|dH+-Tx0@0>4KKYzZ*
zOFiXG(|xv8^eL|214!qK-g6sVR*|cTS5Us`7|EM9o;h1v<od-@G?cjY0hsY}cfFm<
zkw&?e3K@v;+toW=LeXvY6(s2PL2i()2Q8wFsd)~B_u_3rurTpKRw*%ydxIDYc!%(p
zP@q%FJLI9TZ%#CjbJ#hc%A>t0Gt^Br_t7Eg1(6_2+x?b0-IJob&t>i(Fd&r3x-AuH
zxl|KkoxkT95Ul}<xEY3`ap}*uS&5G<Up-LnD_fXPL4%5IQiX(A{EweEv;ttt(*F%A
zQseBdA5%EiE%sWfBwPn#^xa7(xC9eH!D;QQ<tn7fU|*L9X?`rn_TCcSfQ{fI)y`P_
zs_`d7+<91`x=C?X+V&>~J<uew$mzN#$<j6H4qKot*JK{K0`9k7>aOx0b(<e?DrHVb
zxhonm^}A~W7xHj@l{z<&ss2Hpe&{XD_|C|`L|Y8lzqd4!3<c;-?744FL^VY80cGJ~
zAx}cXI&UgT1+CYx{3yO`qe+ZL5!R6ldX=p93Xzmz=|40aA{JUX6l}SA5>Q{MtL38N
z%Xi4tiDu#!i+{vff9nL2aU$-g*lDOxDxl)3SdMTPWS8ZyK<WuU&=Ff<aSDs7T^by{
z`M3`DTy^YJh6)#T`bsEE0W%741W4U_oAHX@>onO7s!{>3x<(bIfXV~9jbi`uEARt3
zE<|TX<d7;JY#Udet$nr^2RcERe~1P+ZpM~|-v><Izua>~kOF<Pw^8O*RFWVP((0le
z{NJj=#kb&>1_K)_l~U-ZVX=and;;(Q$N)t{gp<de&kwGRJ^()i2hxcChwJH)1kD}K
z{qLh84AR)O=kH0=E5Egzy0ZUyY%M(;eFRFay=;sC4S25=MSuaYrs2<^2LSaS5&e6H
zjq10S!#IOw=45ACrxr9dBBBeZIH1epG@3i8qM!@JCPF-D_3<i0@7Ldcs>R-UV8cmB
zLA`0p=w2G{8&;dHX!!iu>b!jAXo7iXc-)FNPNMG$UIa=VdPTORmx*|t229TB#_TwQ
z@$X}QvrUf!4nckXp4oS%y!C}uVNG_u3&%qIy}py635pamd((uJIq0yeTQ^bZ_>A|1
z1_3C@wjBqn*7XY4neo56Lb!1IEymai`dpaj((~4El8gJ!`Er>yOWPQ;&)rS8x?8|$
z=lI>~Xa2Z%j6|ZBrqTHuC<i>sx8Nu*GHNPRGw0Nos%qj{pBquC6ML)U4D#sjnh-z(
z`~+zq6TV&WN{V{wNxjF`8YQ1?;K?Cqqsaxz5U&}+C_YgU&V2+>$LY#<5#sfWU})kJ
zv+NAU0QA88gMxUlQ6GlZ-Z;&Q>5AKb!j%raChoRP`qk&5A9#s3(p?<y)S1K)%Q_OE
z$IQmhug+)ygEQN%JuUVbf(KfS4oKg78gh-`?NZfqMh{!|QBhyGcG0Bkg;;tSi6bc2
z`o=6X?>jYIfK!?Q1H9dpN4y~hs@@_8%55OQCIWqFZN|MxTbMcFsvBmYy~qpgK$u9G
zLj^%c5l||j`y7G-d87(nFdb02-~C2NeMkbz9Iao(kbv$=a7B(r!rl$P6G&89^)F}f
zZR@LOc+U>f#!ao(3Hw~k9s`iDwo{Y`U}1&<kS7uQRG8rcFmDUe`~LsuYDl@t^qppJ
zn+9zVy%`llmCfbju7-<L&CY+QdiS{W6Yd0&M&c5~8^sh%P(lq`s;us;YIN>5Jg(A%
zH(NZa%RQ+Y7Btzl0y4<d%AtBot*4#I#04j{5iWX!78&RRZqGgkfb(z+xQks!AMsHy
z!Ga3_ogflk9<(a=#;{@r?bx}h30OcLTP1}1bn|tIsosh{*JaQZ7Y>LbkTpN`eh$<x
zA8}D2sd{5mM}7aZ`;F#6{Ew@yE8=8u+@~^NFD?b5G5}Ew;a-OeKa`uj(Sn>hcbbLh
zwS1J<$<ZZHeuR$-UUboF)WQvfeQ--Xe!&Pf|Kd5@-5=l44lX^M=pgR38Y;%695VGK
z{Z8T1Qq&2o!cOs*QcBIZmRI=ia)n04?lBX+M(Ui!PwkGow4MQ;7o(@$_V%vlU)^cE
z5^CjnRSpRGc*&TZa@Q>eOmZJy1p?4$BM4Pf`K4v9B6A*TYzp#7p@>GH(u@UgS0AK{
zhIQ`2bK=a6pIsWJ!Rfwa!OtWN<9Q%qeP>WjX{E#Q3TP~i7iJ>S6@yU`P#Jovkc25%
zmb=?`#XU-d^Wis(cSdy#xN{v-^EJ6GA>(+WRrF4+*p%Z#<DEO9&V$3G$k)^w@2ouI
zhLdilx!wpA8R^e<?CQ&RwW_+YA_H9mh0odNADi1QNmX<M4RP!TeO4gnCjqR#bMGB|
zuPzWL#Ikznh}SvMae{d6m|<NUO>5l12T8ks(ps$No}IVpioA)#J5Q5id%4qb7{`R?
z9v0XS;6bZb)b6V?ZEwCS`2LJ2@y|)G)6O-1dpV%S{Z21)wZNC$QLtw1v=lP2p{D71
z)3=XbbcB~!-LIDf^Kv|$yhIEgxwK626q)-fnde%a(KQ?0^}DX6Q(CcaLyRdbF(e6b
zSOQ4<JSTcm7<tlSKBr}jWQ=<+&=@>l_$D|xner9VlewQ^t*5UTa!?*h&$X<tAsqOW
zd1p0Luej#X0h{bwxJ6ZR6#YqqYKg9;N9eGj(JtSsi^q;z3HfGlCc{xgow(8g=BH;}
zX?^mwcecRkUs+>6nqN0>IE-$mt$$r6d@HB_0@g*IGm_<1M_})}{qabt(Z#u-#dtlo
zyLzsTIX1&t`ruT~E^Eltr)IAzs<o~<V`m4SSeQ%kbiM;$J#Dee$_Np-p5?oLQU0dO
z>uu{-_ZLgpuaiO|k*egNdx9~TZpdNjLV|wQ1-L;DWfdb5p5OvS9;@_`wP51AJ*%`-
z3?4kSQy_7UCY}D8l$gDzE11Y$LxB=IW+JYUn;xz19}_VMIkeR=Vo*^fO3Nyl4q%~@
z`~<(_ycM3ZMWY*_U|BG=qEMi?D6)MdOgUV2g+t69vpdDJgty)BX|d%*YkZPkNh(Vn
z+g)j4=F@k)4Tjhs-H3LheO(p%?dshQx%O{yjUFGARejz0^={QvdZOi`?&-32ZiW!n
z&=G9DGv9R1dqIn^bL|3e(vK8qxNC^u_LZ_NE*l4N%Y&W>9#!j&*VfBK5VhUxmi@iw
zt82t=<h>E+T7ck<9F4r#1N#Y2MXNHH%`Hbrl6+YPQrd3@`SAO5y$U7?$Dxo<Jl`Dc
z$F~>uM|?}X3_JwYzuokOd>6IN4=I09_$|S?`}JZBo@18N0k&l^#M-d|nR#j~ZPkaP
z;heg>J%|S}cKao<GU|y<VQ?J6hNQ#$)?OlNkpK-Rg2WE0hrp6&>unc(LO401C%D!E
zCT?SSE(p%voBdY0S>N3{UVAN8+o@wkI8SIdLdWNlg%YBxhot`{(rbmUNUvTZhq|!G
zlm&SZfoCm)S$XXZbWAdO@wEgX6(Nxx0X!iv%ITerjRz>Z>1)-alX~^IMm|*!ZZwf=
z2s}5~!fwk-4eK;ll~UG(@~+k7SCX4#H>qo<bQ?Y-qemu&9r}vz8>Ujzso&mJiHMWU
zu(n#xU16or-?%pDTYs&-{;~+FLvC*+EsKp9vO)LmOAP%&Z<_tgye8MsiNk~%3;fxG
zX)d<oo(%Otir5mvbo%`9qY6Q()Ud7Z9|&((`<A%E4ZMil_A@y1G*cb9VmBRUE+Gmi
zd;!Pk!R9s>MDFra9)Y6ifa9Y$;tdNkM9tpHifU%T)e$?_DF(x^Y-lAMs{*Ne7%{?N
z1_yp%v|pGxxEpvEEmVIEY20wJ%RKpvYBXW<ICNghQ!T~$O2%pRkXoR&b4F}uJNCYF
zGxmFtt@qrLJ~D;Uv}<(19xf_WhF*I@5w}tUgsDC8t$j}P3KwM-quQBS2d9(B!Z#Kd
z4Ad)A2o0-^qTcUGZfwR0<oK?+Ey2f}3nIi!aFw6Cin*?Iw+?*}FqO!n?z`h_;c$ER
z)|WOH6kE(fUAW4($p)w0$Izd9w~iOW*$fXZ)z;R|6>$d>&@_!CP59#*vA^zn9wroO
z#QND8@1A!DCBkQw?Y-BQ%C{Wg#FyANq2ZjQSIV67Dvh{2Q#Cp3@`eWv7P6t%D#`P4
ze#bbE5UV}uIxh?`Z#fR#sqDUue7j`&^l?$@ueC+i0#2LA+<i)}Q$)MonEtBdlv1b4
zhw@Y<Yj+v4+Pb>j>g`0k_Kv%`N!nv^MqG3>lcI%ii<&!n5xe0>Yl+2%cRPq4TcyhV
zu(n>#71^6^yY_)Rm)I!6%6i$>u=Izzn1@KzOmX|WPY)T4Z-ND}J#yp6NhWy%wH}~J
zmqGAZ)$5V5_NfO2VM!jDHg$m>E|9Uk4IAQKdhlcFoMs_GR`3_u*c_-eRwBHH7yLsh
zI4ROu1N=|`UnK1#4)`A?UG#5^9~UWs8^d0>jocStlyHRfH<pd00^_%Sl`vCGq)omI
zl3gSRwK0DWJD8Gj?iHJIS@1tMUe1U7U$1shax5u+pTHeoo?x5%_Vbet>Au0i@AZT2
z8VJ6n61D+2xRZfMqfMqBu?0KQQC2*6G5#xC-$9Vs@dVxemQ?=%J@=P(_Q$2)8G5f{
zB#}a5C)PKTCW5c6CS^g@p5N}fAy)lUs_4}ArWgLiecw@C{gaaSeJ8c_GHy20hhK)1
z@Xh3J+tqko?vo`zzA*CiV?k!Ph$5L^l7I(~u*x^?V1pOn%(sc1#fGLs*F=;3$7npM
zLUFkIB>79kT;h}=`&Hs*I|nLm(@h^}Wyd))POnnKX1TB4L=d9HVWi(<O`xtwLHV99
zt+vtiQ9X$w7tG`z1s~o)MlILYqRCt+%usoRj`$6<j4;N}155uQ0=~Ip`qizpZtH6y
z`Ha|Uw$458O$%kpxVvi<;oJJYBh~W4r`koj2k**P>W1xVeYu`zW(0k2xl#e+v>k|X
zrJs(VN19^XZoG(bQcsS$x4|kaBU4ds{JNkZ7{0hP->5R?O3#f5p>`a9_)gs%@BT>N
z<t7!j$^r7_sNGP`bi>Td^^J|n`OO;9q?zKe-cP0sz0WX7Q5-q5p$ra~7)d0k1@e=V
zo??>L<K_?4aQZ5Pum=}C6oNk<ZKxI<y5QNtJecxNC;f-_<}oed%B7(!e8<FQTqgqq
zuBa{%&-DIqPYn$cvBODK-OZEJ7^6pQ6>>#P2i>n=8|~u4j%)q)9k%HrcK3^G=4`gt
zsG>UxZ-~^diyqnUlwhx;{9{h!){=v_@-r8Bt5S6DFLe>tKS2G&j-V2x%u!W{)mAdt
zpQUk|-cfy|ZOB+*jr_&}-S5Dgo}Xu_U^g=&xAWcBQck;%J7nl#7rtA3u*4~{hfrtp
zIAXf0O67nVelgf-<69?<sq=x@-u}TG`;z<L8a5@bQYVK*?b_~{2^+etnvv4*bm584
z=e~sco-lG$b=-xITiVfR52)bGCH<J0ZBX$yGrc*%Lt4OeLpb*VaFy=n=hl*qA6t(j
ze8&R?GV88~iq<II3%q)aD*FLFJa`N}C=S%!`p&~W2Ve8LWaJ;sD>OM#&uXsf?BQs1
zumRbd3a@ZGnD`JQ1X`Z=22fqU!o^S>%`iWCymy6M+>wOf`Mk-<qIFDyGnP_j9e$Sg
z7KzyL0dwvC(Q2QCr{M!4k*SwI-`!IYveLmLGg3}%h)7A==x%4-_<_1fM0_h`I2n83
z#}6~4j0WS&6J)UXwMkK8|2PkN;i?P$9Mv~Z4cr8R^0(u1PnO^`x5X-Nwp~@Vq(%1B
z9dP;VJHr@;dc+*?WpfEcvPUlkGVYllQY+uLr4$hr**N%IQp)JfY3RA2{i58yGs;{2
zF*74*!0SGKG-V1VJfxG;u7S`sA%(st2P*tptV9JfgXF%Ze&nKdTBe+7TI#L9GWCZ?
z@V;7xFwaGOcPt!TMKF)v-T120nKwCQJ2r?Qw0YJ(6AY31nJ!D-!|QNN_nLMRPH#U$
zobSo}CQo693*VA|4$N@K#>r)PiyFHBB2VY1n-8eu4L|Hu?M~XAr)Fu64_|8}%&HGj
zZ7>Y}w&Eq?J*h>t%g4pdUXN9a8o6tUb!gPDniQZDAw<J26tq>4H=W<vlLY$B=d9%#
zU1;UEK3M9rs2cW}EGNO6)Q?)dekFI)u6_nH-Kp=wz`&qM_fGWetFpG~9hA1b5naCO
zb${09S<mjyrph^X=V~~9y%hIkVC!>QR#O9KdcgQ$8RKJd+khpRxys6sA<?ML=PEDv
z;3oZ*q9Ioao03=s6#KQl$su%(x0&0n(nPS@`%()SxtOHQOw=6V``Wa`an?oKCy9DD
z=LwHiO-1(PCY7wE`G)49YBX-iIYPqjmA<x92H93ylX<_(1GHIoGfy<Oa!GMLJ7OU|
zsuDXQ2xNQ)olQ-lbe`zat)F{fIT|orT<uBAk<3qS9zivVtXHsX8O1mCbVfMu#>+3d
zxqH$HaY>6AE&JMkcq@#0s%U!HQz$TJd3&L~f*fUZqS(#{sS6Qr@Wl5dywwID5)C>%
z97=gjD$sDV_dv*bQJg)J0dK#AEN65pOBTvSNjNS<LED6p4CyR$JDor5Eih-RKc0rv
z9!!?357g<<^-dCQ*|}eio^%P^&NH2CYPnNVl$f9BwKxCdsMI})chgAAAUxLGAY6@^
zS;^=^ZtT{+suCSr{!#4#68EaoQgCd`>nEBj`f~cJwJ}~x-&icxyf3MtLMg_BKwgTA
zQ`+&aPs!ff=M_}INFu~XZieHR*F9F9gBkk#IeorGn{OiKwJ$v64#N~XVlteu^dQLg
znLMMJRf;tUC*S^ZPMwiE$YNrJ$`&mgT{>jRi5IpDz6$XxWOn_q-lLc4#2W-;##h%d
z`XAI-tr^2s(2HR3G=Bf*2b$8USnIq9o}}?jHe{Cgm4eLl2y>%jEU^hgsL6!;wSHnd
zkEv%x`n&NANmQ%WobtAzp7W(oYCYFV`dBcsFY6!7z3)4<K3-e>p%$`Nr&{;YrIVuE
zJ>(iH(4M{hF^L?lFzQZg$mY1fx3-np243ukrldyzo<MB|H4%Hhi%M6!1ZcnaGexx)
zIjbg}>=z0fNvKa|Lgx$d@_K+}gz=8Dt2>9CuJyB$^Yrj56eY_N-}=ef#c4m$dS?rH
z{P=ac{o}2CQaM_rY2nwO8kZG<yL^p_9@6lXx9soY<45kzExQ1}v7a}>EKU=qy6pd2
zSofCmmQ*;SNyAo+Z-qq?`%FD2&lQO1Nlohui7YH2uc#HcT^Y6OUA7=#`LO=JY)C!j
zs{I;XoG3N*6Mg@g;!9eI3Yfo$K1n0&PfeMve%;6mXOg}ACci#*rq~c$_jMYH1FIv&
zR)&$=F;9^`DrviKdgVHa!!g1+Pq3FUZ{BSVj=F7VMaHQY=Dubg!58@sQeIMb(w_*`
z(HWRNWL)ZFIe8$K?3Jg$TXBNg{dGsyt0%>i)8}+SYwz&-!~4o1hpeZ^xI8`AgWoyi
zY^&RE9bzn8u^+3Zfryq$6K~&K=(13Yx@N<5R~$JI_w;>cyen>Z=F@||&EBmwW8+C`
z=w2<3x<N+=Yu*w^SK7680M2k|KT4K6Z$>N9vDdTN7#`>IYRQJc_BLr{Hvx<#P(e<T
zoTiFp_mAmCd<jRz4xQGJSr!m9@ttZL@{>7_z%^5cGVAjKd&4Dd!PyB#?`o%ut>sA-
zS+#D>CDf~{Bn@{(Dt=zn{4p@~-m`e6JhJ=61^9T%TX;<k{ma1_A6az<ZakHoCDWVN
zsQiAwqSO)NrOuzaTxq91&rA@B(t2i6Q`_bbkkHrA$r$Nz^17Cn?Chk|Qsv4E-?!P+
z(%l=5yzgDkB9p&A=uxOFr`%w8j90&BcX7gJqnkV!$Faew`|;*2C$1Y}^V>vyu#2WV
z6sTq7KA~(blR^$^A<d0;x5#(Wzse9Q<w5H_G$953`AhCe$YEOv`PXLqP-yt~7qN_B
zLXu}W>b*h)kLY-!hFcp-4xjN%xShBLBzwnN6em5IF*H;Q{JuEhSCS~^)134c<Tv|k
zDDf?ASVQ9Q{8Dz449HI<p2ae}p#y^!b2KI$MBa@lxmR02r01;2g1pOZt_En-Mt1l(
zwL#<cRs#n&RGF6Bd`8K7->iV?&o}jbNhvMvazqp)EanN8K7v2e$ABy%FV>z*h~slQ
zc4$NtZadXb_pi^;Q&ij<t!zq`=hTgDG_IS3C$Yum&A(!$Za(OtXdW8!TQuC4@31lx
zo0-UjhA)YFd%zW(e|8CMI2BT6Q`nL<)Q23MDxEqrzdqP$y<$|<^m*PoH$K@U`$3;2
zbtd)tu6jPkV|wI_Qgi2;S14Y$Vq7s9`3*J|M<NzJc*Vrgl%CDrFGd!sr=Ev>JOpoj
z*RYn7h&I$q+_EdOvmYGv%`4u7k0-qSS_!2C={~D$f~^x18qmI-XPe<jZFo9zx;{?W
zpyXTtXJSBwSJ5m#Ldg7*ie<?c3_iSWrkZGYA0HoQMH5kwDMoDmkdl_lTx^B|ODCRn
z&S367#nzC~e__0uiGN`{H9ehRqW<E96Qem}dkzid1fs29GI!enz_a7JBzucKA7^zq
zBeX=vzSU1mG*|7U*H~jbSh@Bnz>t=EIQca7liL!-`sSt(Z8Yru#_O9|tw)`lF<Vd?
zAulX*BM(A;uL^AO(@L^vb~jb4EVmKOZ??WaD?4-5e7UIJ)p?taM#+z{)Z2t?wME@}
zC1=+V;f&y`@qNq&x`BcYV#t;23{$%3x!4nqg9^~!VK0f#zs85>U~bWB0i`4p2)bNB
z8Hrq}?cT+E$;n<4BfHflR}CK^vbX2l+t0jD(@@7Ptaq#l9Cd}~$#Zj}p=;M|$VHs|
zHC4WgHFZr+>O5g>^jfwsGVu0)k-7Kkdo87=Ia`LUxEvyb-L$9#h6V|Gts{nY;4jvs
zMCRa49znfJO#`(zlme22T0WYAm)pHY+$#W9u70*<v|0>}*tqil3)NwOpHX1brA-Z>
zaamwAfNi?u%b>`D<e)O-@4ke_I=4e?UTL5KB&pbcfj)}c0Q9}Q{Tb(>ir-AJ+8!F(
zJuBq^|3LOlKbr-I3gP3IUVUM@&xd~a^nXLj|ECWpndwG%o)^OW@#9CbkV~Np^M(az
z2Zx@?F?>EwX-r!*;4(Ywjd}q`7m!bOL`>Aun5pPF4obXq)_T>?8a?^iA|p2z>A~>p
zlS|vq+KNtH1G5=yyeOj}0SGH|^YgZc#GgXIM9?XVgkbT4ONOqp<DK_W9ijub_~`bT
zuUigGnirQ@<XJP+Mq0221Wn!Hul+z?*4_R0%ag`LU@m2n-LBzS$bw0~{)AlltSM@-
z47kOUji=V>(cp3WrfoOQ-&hKJSPH?+EMxMc8?)MbD0(^Y<@c8amYkf5m>%dC(v#Mu
zIcp5r%K~oy@M4QqH9q*uSU@WNd2`t@U<<ah?PUs$+F+P^IGbK)mpG7MdXS=mYGiHn
z(pcqg7kOUr$Ru50A}Y1#S+=_I$WWi&i1+sJg-HqfL+$3S5?M#c7D3~}`Mc#|ZQnt<
z@I?-<H~CcKe6RD572z1|W7@dSicR`l%Xs<5ypsB<np%hwZpB+HYAZB3@e#{i?5w*i
z6wOAYVbx?jS4U9M_=dq|%^Gi<t)ZU=pdy52l6gNczzulH$tjc_t0>6`&p)n-0&4_W
z+tc`ecio#Sj^gUt{?RAhaTgt(Y_P2a$UCx92-IO>ijG53AM+SB5bTWb^Y558cH#Vd
zfGATc-1t`L?Zm)@ij?k6<PL*T9WcSi31Wn+vEbtyT*ml++DFPavVir;pi@rOdX2V<
zaJ+n|UPRcHvo^YZo@k;5a>Rm+iWdLx>+VY;%+pEzSF7i_vuGWgbWc=^k5V+Atqi%`
zLGGH3Pb2_pA{xW`+yoy!Xe65PS$Jc&$8j9;&0Nt*z^g%OA55;%P&RZdnh)H*=<&sK
z<FEq8Qb3T^l~4d3i|Adzygy^dBT0UKFXRJYv(%1uB?+XE_V%n9<R0ax%V+NN1`iXS
zuyf~5N$m1#<8A+jCZ!W)$03>ITiQ5bu$JCVihQWB*ySROGgJ8as&RBYaL=jiGxU4p
ztY@}wl|&ef<L&I8;v=xc7yr{VIKX7N5M3VEegz)c`}+rjzuRzv*?a-ssW|-J&q`wO
zPQgL{{Z6D1gMbHB3Ba_(c+oHI$9L}M2Eg8y>8{p%M6XPC_J6${VD>3YdTklOWD-W|
zg?VbZu-A%+8sFPXzxFb$!wT#2Ud8p4s9iy^_`gFM1|DvAQ(Q+p4ZH+4ECISmMmjr9
zF#(_O7kpP#C_DPD33{3@h>g1~HxWxGQn~K^bfq(hi{YApf!IY$`YTFAzbvw5rM~?2
z2q6$uY@ONeYeubFC-9~m%gNE&NxT#4&3H!&=Er|td*-CWSLZf!12*C0u01a=?@WiV
zAxQh%XNyAV9Z$=Yh^VdySubY#Z$~I2c7HDK*?hftF9^#8+Ot9X^xP69sKAnkdCG}~
z&J=fhKUZS0V?^Rjg06LZr<>RNeC^%$w3I1%NI^yrRla2p56g?D<3?=bAG94(m@SN0
zQi^&#f*tnB;+F=B0!sOh9T^Y-vJZ5h|0G(VUpyNFHD)_y);SS5OV7j!*DPBH)71wC
z_!LhQDX;sb9>~tcYbs?$Vu*$C9MFmuD2p+Fw7);N#{ODd^<h|ZVTNLgIl-@>wZ7;t
zhXo9f7L;&&Pu!s_okh61;fpn6hUjqZ1XekO%6%4Oq3Z_@vIXs#%nUuhPCIw%e&(#O
zzD>oa^Q2So;Nq9bFtOuP>ee<+ipG`}Da`%3w}jt0xyWZ#Rcs-X-^$uuaD-Ip&PE}M
zrYtq$ay~f)(|y;7gPG)|@B#^u4(ZU_%6vnD2fE4<!P9S`8i*aIh8j?rNK(&oTe^5u
z=1mwDoiJM{Xl{c1rZDV#heBtv;+nOXhip{m45a>5Nz)aQw%khCTE|HLHnH5G|5-z9
z8>T6anR;7WjwtHF828AmF$(I;)!7_yf7|+M&AVunUp*uSFYUt>aoD;=N=n^gpj!LV
z8I14i?8BzVfHcK7V)H8~uxvHJ_9+RlP)@t|dEDFixw0n9RkonBhul*2o0)LEdIt%7
zVKO8C1~dVRC+LfO`5_ij5<IY%Abt|MHh3I*iO?vVYjHecUWr;7P;-`(d=SpoOZ{y#
z5v5XN;{q(U$%ZpuJkBF`7qC-ORFn|WB6Q3vd_><*FZ8eiRU4oYwB3RTXUeIs$B5?i
zq9HQLhR%0vZLECAYHT0F6Zj@fsOP%UqC!|3k<ghbDeoW=d}tGO0q@*FGk}9?AXtz>
z)z`<~2Z4KLUJTwCe)UXUQN9$PdKeGYljR@s=~{ez;%qF%WXsy=ilkoy4YJ;cyqnVc
zhT57%R((t7=$|IZF@E#tmxP4D=27AFQ7>2>?Jw&;*1E{ask$`ZKbZFA9?8@it9cUF
z<m7iTIjrcR^L>X8@W?FSLcX;-JDq*{h)&kR+ysT{v%6lv`I+X&6Lz>%NAKX4@LErd
zm-a%^iKj&Hn-r~l>*E+_1guS^tBjWdQu<`cKF!tiApU!@+vp?EW5Dijck=_VsV{~I
z%OQ3kMyW&c&W_?S3BhC^eBeVLe5IagRg|kQ0oV`fVHeyy-Ivk`tmV#92s@DUJajC8
zE*}OB=_At8jC3ab8Y-Vo9lH3M{5WVJBEP{9*{pgUg6q-t0mfTeA6w>n9;{W-PA?2M
z$U;4L{<y+~?4l^Lw`e?=teg%U5>L?PBpdsZebI3o;S9Xv;bgh4@c9N|a(3aQ0Cc?7
zKskjdXVs|WM2mR$x_M@hb2la9N6iDIxpl7lIt^R>m2tYr1E#jWYXC$xzgP62zM78$
z)li9STkfoX;L-K~u*94nWjzjl#yxqOYH4NFOjgwpl44$9@@uJOM<t@-ykRzw5rOdn
zkMnl6=0Mu}))EK8pY{DbXspLs?c3|QnPoOSRZl5k6;dBsdOp@EAT9d6^7Zz1U2tze
z!u56J(4|G}CxQ^+(qIQ?TvQzppdE}la}h+q1$YLS*CA9_aWzyNHCnK~yV=UFW7zvj
z+R~jdzq|U1QT-0~V#P9Oo^$pPJIV8vcn(&A#^BRcobjA{_Ji@{4AQPC=(u?pzV^UW
zGNPX4pZB`__q|?p!H@e-j{~%p1*YfQ?fsgj===Aaoi8tlZ{zwLtX$gLb6ZX${%74K
z&Yva=p0*WDKKlzs-bhItfMcdkbb~m=|MC8w8(ipU8L<If+Pr?i)cSV?($0M9pnimp
z0`mO`0SYPh`!|m2T&m(tp076=QhF<LbBM6wQ&Us_`?_%Az3nXhqDc*gD!aZ$`JxFI
z%@YwmL0nb{X|gvmC_y;tOBORhMDwUY$H}y4<Z!A+SZa;zQgZ7TK%WZG734lSu(`Va
z0JKn95`5-#-O{_I)Tcx}_^X1NZiNQR*0%@a>jpJWT`XV6*XtfrJzS%D+48|@ep8YJ
z(_yLM$_vU;TWe>>?MH!L-}b7#r>93wSS|&3lzQ7{dP$sq2mMN}{d!yX_V_v3?;JGH
z<#!6|Bu)@@wbr0@JQjbh;_p`rMnNJe;Ft~7a}9okV4uyMS92wSlsXkA)Af4@+e+hl
z)z+2U(087v{KOXiXmhaZihST$6_m)|ZpDF=3y+!xUx}q<Ex3Sj;=^1%6JA9v>+q(5
z|D)Oazh1ri+txo;i{3#F_Vz94f+g-;^KB|HIt?&K!Z0_+bIL6*DV^__#Gg_@#zEy+
zQexTKL$i{!FNY$K-Go|vn9W9!pcA^7g&B@bXuFKu)Mj|k3oq-xDcfy~ls_-H3d*`?
zFSzFQ8CU$Y9^bysA&sBi_@9?D005={s@L!M&r5yzPW%nFBN%ZjF(B>=N~&isQJe~Y
zCGCPeXMK91ireG6X9-(5*f}0LRxO;ofnFxejy~4{=t(K1nE_|kIe}qziW!aBt?JaD
zl1je9EVrckXLJ@xejma--Bo=xhXmwTqwscGisE2Z#&2AyA>EFPD$hAFQ^P$C3`BoL
zO^Vp^-EEv4GPIp{Im?y(u)bkK3!(Oec9L2j6%cT`M^z8ybD<x4w*+U3F{N*Y_k!TJ
ztn*ka@;G?Y_$tm**XuX^+xpK|Tml~xRCF~wWTr|Q5)w3mGe|>!jKH0?#-0d*vZw6Y
z{nD#pFn;}yELzTv_BJ-nAJ%*#Q4RN0l2Qb$af{PM@qO*O8(#^Cw*wC@&94_pJPQ-@
z5{&2@$JJbq-4CZjma>?987U6&+?Jr<gd}{C29_(~sPo>x{BF=pF}{wq;tzcU_?~pY
zC3}wcI)UysBRz<eawO4=A)vWK%aDG$ek?-V-%^FPbk8E~Z45Gy^{QNJ?^(qo?Ad)u
ze(P@gZiV|M##tJ8l;Vug`WeC2Ob6!fu&l)Al^v0+zU%j=wZ|KZAQx1eF`Vr6Lx&zu
z&hbDoQ8V^JHp27NhWHTxoxrM9>}Nt?eC6eW+Rlzfn<oNazOuHqR2hs)A_PuSqJ*i9
zUV#QB)b*EXY@f8N|Kmcxqp+T{=Oe_&?7Xg0@cgLp6zEJ81(#<V7AbjBb4$y`W8&FQ
zM`>AHpX>=md={ULmPh3M{Yc0Q`9HQq4&3mM{Xo=gH|T!xlz2s?A4yh|X|!d%f)%G-
z5;yYaXaAQ|kwWtQHCH*_KtBY1?z}W2gGjU1d14s)_BD4aBzn&3y#c{>)$^U(1>2!D
zgC04!%@Fo`h=av5#VRDa=k^>Up@R(&n<^e*Teqm9P`e2|%w)%Kp28#Fe?{?Wg<$aC
zWlQos<dIhVvRg?(U3=pthR@(YO8eiTypw38o<XK)_<uI5P`5oks=I~|r$ooN)pobN
zx5m+%)$P~0be%uRPpCTT(DvRXd*G;plZy|x;6TzrU~Ug8+&_HyAgBEhL;K_D%|#U|
zf4$<`3!qlO&-=3A??YY$`iZuVi`QZ9JabXs5jhr_=A(<4nMg2Rn!1HX0&l9JRYR=}
zl`CyhoJSPs-O0-fob!M67wZUs6hq)rI^Mm<gBvW!M903(JHeqa-4M%_{i-RT8Ivj5
zvtS?l^7)@723!#c6wYR6OH4q5itv7%3|ZdNE!|d=4r1-RY~8P%sCfQ`mtGLQ&{5Lh
zpD1a{aaiKJW})U8ORPTUO7nW)W{Ie%QCecD=DCHxB>0_N4I=1BA=Va_&P59|CoOGV
zuXEGv#l#GMj^_w7W%83YJ;3-6<9URrz7S^x!GIg^#T#&dbZO10%Z=_VO2zo)W9hrJ
z9_!kXJ(^@F<q5dyu*2v7!~s#sPa#LB_2slJNrQd)IVq&`zNLHo!i>-TwL7@D(T@i%
zm%hEb2-6D)yK<f|c46sjA(WgyeAk?`AV%H0SX`etW9Mv+0o>aEt?e_gfHXck^yGEN
zc`_BX%Lk8>#oWtf+Ba{Xj}8M(!KfSo1pQe2EmW^fp)&`a2M)jfLhZboXRgGV(*~Em
zy(Z5!apIQ8TjK+&8S}gD?(R+f{ZgH~pKZ?X^K$lgo?vZv`Hvj~SM^<PSp%P4Hxcpf
zA!C2?2yo<I%huXbqcdgv8cP1BT_`k3OhVFzl%vi;J)HcO;FDKd9KP6VbuTCmwj+Mz
zyRapk2RJ}v2_g#-mq;O6_eM2Cz~YQY`x<Zbm4Ln}g7VjsL*85vzXb@|(_<UHUF!Q0
zkP~O-g7#rj$;@+(%YeQSocaX&jq~6ArmnJ*fU~~z7OVF0-HO|dlY|LcKl*&Ls@HJS
zMW9*y?8~_WsiQ)gT0^&a?_Z;)?x;ODL)-4Ia)-D2unptSUeD%_h4y>@bL;9+pkXT&
z=eYuJ?;s*Gk^c2Y4BN4BMrn+PU}hgcC*<)+NMuw@y-3J=A$Pg3_2FhzY$@l>wipS-
zi~Xy)k+;~i8gjfx0|L(@mESHFLIx3xx+){1_7Q8~+Ek$5ql;TmNVcaM0(L=VUIS##
zrt#jV*&+&IvoN{$h^w7pom~Ja>4=5}+rvm_<?vsETrSJ)|BA)4bG#D`QsmBU?gSee
z4kZl7lXcIgL(<0j0MW(^xiC+PH}Ac|7G@@fhXZsbk8iB`mTDIs)IQd5m|w9(=ZWOC
z`+^fjUeO&YRt-=A@jU`LTqTFKNQ$ca@9MKLEEgxk$dqUjDEw+RPT<Mzt~X=}X$Fw>
z<!VwZ+7QE#vk4tQCq_iis3Dym<nVFshfk>Us-d78I}YB2ew3$udL+o#6V=6W9ZL#}
z;jND$Qy}X}l&x7U#ko!+1m#qWO2~Pady77ppj9wVR)H1v@U-Dd=|<<JO)nHPg?ETg
zXep(yl@;vpd8TML^^XO<v)x^S1dS+Qd{p5xcKGbJQDC3U^4=kN5q@vxb2yVyI0qkk
zFPjkQu<dyfQ&#MK3uOGz<6n5+dxt{=#P+lmHQd#m?Df5eB#ZWcUQgNlH^f#g_rG}-
z814fqQW!p-TCemsOH!IRy)5V+*jMZ1>^Qk+E`)UR!tpuUc2|g5E$ynFW&xWisj}it
zCIjTw8voSdWYu(&-brK`M;+nKdfnRv`K0@co{Z9FqXDt2$TS|ax1xnz*5{Cz>5tC}
zftkGVWc1dDYUmoMf3PCmh|0{BB}^RO^OEHmnchnD-Qu~tyZW+ytyVcLu|n-v27vVM
zXH}a5xTNyx=-Q9Mv>K)pnC)AWx9r+B842Hdz6dB^ONCL)ys|WNY#YvAMOvR=8dqIj
zmpn&yp!kCxCsO+FINy%W$PAbp?o$fx2RgClM~Gj77zp6`)L*uJGAR3=)%S=b>%?Bv
z&|}5Jvhdr*o;v^BdQpe2hMu70zS-H;z?%yh!R*4j)A?5n+d7`9(lXqC)8g&=K?IQS
zf1Y_4kd+#Uk4x%jQQQD1YJcU;C)~NqBISjW2cXrm|1E(c+14*3AM;D~hR12;*zsTb
zPkKKlHBa{Vz3y`o*_`p*JI*rbn|TB=62}7-T$T4r7FpvroAAyVQN=$KT0{JILK~oS
zJli4RZGq;IKk_db+|epp80{Tq!0}vGDT=zwJ(*mRKkB6{cUbcXKU4DTRSEuNF8vrR
z=SjJN@&h8ytnLP{p#LFJHmFVa==U#?-|C1v-096H`zKB|qDfO=!5Kk+Em+{F)ZmHW
zj)XQfF^Op}d*AR|MpXo#YN`ghs_}=pY^?t?B9B{tM|66CBCr{Ac{4s5eko=p<+1F|
zh`&!kR|(JwTS?wo%0z_9&u5B{*9N$6ctI(1l=pAc0@V<y-}zO3!+C-h0+T<y@6s*$
zXY&qL<fEWpvlck=KV0njp3nIAA@YrP&z$jWC>3eYk|(DEVyCVb4Lv>T`PohW!4le_
zu!3Ipdw-Ul4Q_C=`0t%w{6P!(aq53kL{f+?^z-?~-z)mZMjtAk+2~zlL8Sc4rgphP
zcRL?iZbF)-cP0w#Pn<vqQ({eUI!FJ+pExn(LSXg~U($K4ulTD(G$F4<pLHzqn0{I?
zjy-Z$7og|%6{PcxQo)hx-ZkBKK1;^4q)0FyrRqiIAIL6EUNyOV6&+wrJ}x=?QApAq
zJD_1@_a9g(cKAwaez=}rRqV~_&pWO7_v%`hD-Tj-kB*vr@*e)FcGNeOxBR$+boE1g
zW#lD775;j{_T(T><|E8;!hIdZqx<4N(riEa*6~RT2^qr*M991PraQw*jyjx$qY2V~
zY6#ZgL&S#Z?SJW{`%$16)S3x&#5)hdV9(OaMZ#u^84VK8Kcy!Gf;1YyiJloS8oK<2
zXg@Rm*a-8Pjj$j`9iE*p*c(rD-DgitZTNL|ZF=gcFzI7o)~tP}!r2-!__*UrT8XkF
z#Otc=-%<cJhG>|l&1WJ(;dnN#4XNgsI<2pCQjQRLomdV<N60Ntka>=jozwqQkAUTn
z_|&IcJr>CQlkMF9&o`%waW1+%&$r_gS$T^Cejxf>u>Xpz=}vLZ=E?i7YP!`qjVsOq
zMTqP<&8++PaIGOMhruXbQ|pH;R=<=5fl%S%RF>h(gU>=sWd&KxLP}S@5<#e3lC`za
z%su~qQv0xLZniD7xr=uM`}+?9H4qrvH;F=X9(TM_Rxrz%zWI~AJ$DCFIkbiD(Q7@H
zwdfV10MQ0kBjdLqu{1a+6E+N(K}M;Wr~S!>5w9=6>sAvqX5%Dd>$Z{+%(4c(Bqy}_
z36t=_CkW5{AX?`#?lm!!7^25Z9kpIzFue&wS4&>_I7hNO(ODAbAoAPd6Aiz-=dSG4
zSkw=L3Eg^(dtM9qvu;XY-NqzUGn`qFzZ{;joCkEz-!93X5kF&xC+gCj2mR0zn|Hp?
zAb)4jl`qu8VC>rif<W&{AeVR^?P{3O#xmBYz=Hh${GqPQpAO{%y}GXX+P7jIy6RpD
zgZZ?29}>FX`NEbo%jaW9<L2fuTu1oW8F&>M?<S(r-Cy{UN_?lpvQgA%^>qmsH`mpq
ztI>H-W~?{7RX0?2B8l;&k}+ClUa1-CS|8k+3XG8#za^UMwG^s~E3?~e#tl=W*beXw
zi9V{HdL&p`(iJC;QpT5BSOD>-`8NIMKBc*yKXuZY7Tre(zB*C;N{295XHX*kpSb1#
zw7C$C7dcmM#5O3!E<6$-N6pkKR(IY!swJ}EedmGa&}yKSZg3X>K1Lgn(6jyHmJ&L|
z3#`4wcZt+(uxZfBIT_qF4ue&+4v1&Y6pLvj%Af6dslT^p7G&z=^BNc+nIN<#foFTr
z8)vj_3B&8Q4qJajfs*4B&_Cvn$yWVNuQd?kGurmw<-=fy@O1V+oR|Yz2+#=^t?<0p
z5pI=Baus93;_9}fW94hD=e@o11wD6L1d$vL8hX80VX#XvZrJBm9tlF5lcT+rg`|OH
z#*0_MCCr#D-fEe8jKME{1-OCVQl)biJh`ASD+QDZ7oTsOJHRCj&(K5E7~b43RfVPh
zGu0@hL9hd;#=QcaH}&5bx-xw-*9C4oiq&CHBKjdfdC9YLuBY^??A*$uKH)Y(9I0Jx
zMsp7vMM|ei)+}~~E0&>98gAIaPdraS18RU*tR0h<N|eK3C#KCe&LXo!0oDpF(p)-d
zs@Gm_zJ}pybnEDhI#5#o^r8JSnNz-8>wX;DEBeT;fQPM7+Hq|=(J}NQkYy0b^1u4M
zVvXZyG)Sal)(Up<k@1%I+`mzH7%cI{WB#)lviQyC3#>f>=Ma9%_CGEqg=Aa`_;LO%
zJSA`kbSXj!NjfCdW(%(?EydZ{*||sQ^+dY-?PoMKVJlmm<6ZoNefY+oRjfN(dyfXC
zA0=61VTWiU>(`F&Y#Jmh=G_7nOS2i?q+DCQ4W$&OWB)u)48M~@!QlWQj?x*6^0NC+
z{R7p#MxrC!%nm;;`3T$CYeuQ*=x~M~9`|s_ym(4koZ(5($HVP7mQylY#z=^!qN8@J
zp9CBK{yMK8R{C!a^`Rpq%2TWBNlgYcDM80{;$x2z{jZ+dqgV#y4nF>wCjJgxixgwM
zz(fNjPtU`apVR)M<4OEhA*rRtkI(GEm<t_C!7Og9&TwA=HxmaZ1Q>QhntsN`IgZzH
zD$E!@a=Y4py7s{4$O%J4L?m}AZ1#Y$&GhK~dyuHR@7_;Z!nXNim44nhMU#2~AA=JD
zgBPMAq&{UqU<qH?3Y)jAcRLSNX|9FR@WkI*rYC7}__{t!aC2icNJ_p(tvF?n^VTwo
z^y{MX4a(3Tuz2p2_VIuqcZABVWP-l>o{-97P=c)A=y_4QwRE@kU_j-5x~C>N$~#5y
zuTVVak5HVDDMA619uvw61jSyRm6H85&}_R1VB7Bu9Zy>xUn6=7>et5e1B31e=8Q2s
z!+Crswj(L8Pm7Refy0^aO}-$CiY8L;eF-Ug@vLI#=R}h{HO06|1L8!A8kf6s-mI->
z^&+8a|CEUiKVY|T`4)xSZ2?DOd5Hu1OAU|+^FOcCL;@L{<Dr?-`gA7>le%%+ozS#}
z0&mLwDBpc1kFb}O_C`NbL@OvW4lUFR@-PV*wLUicV`S(;dRp_f-;S1GgkkD(8)??w
z{`S7qdySF-hg*oEBkl)|0`TI~cXwJDQ}QXD%Lp;5h*07J=Ef&dCFKe`<AIQwzWpB4
z{nuB%iwa2!E&L1Mpasmf$h3I=Z(?dQT7Tk}K1h->X1a+WNwAN6_Az52p4P8ZM)o%~
zua~Zk=!n0JuA;C@hEP|o3<sxCyIp5Z5)F)*@*_Z5QqvlqCEwsU_-|iIs;%I&jUNm1
z1PmO|N54!b_?%mQE4p@%+UL}q0`utD0ku=+(5BlT0<^-skpy_)^Q@pg>F^jd%|ac^
zeRyPGpZdXeeYZ8$Kn~<me?uDT<DfG_P%D3REQ+&0=)`fZa%O_sqg2QW5H#rqZAZ|5
zO9%XTh&}}aGqw_UQc`|(gZ!ue@FQ_8xVL13{L`YIxYG7r>}KtOxB10t2UX`$uIS;x
z%F-f^Zw>~T`;Se*^jSN}rX-DwC8UeK0g#QD^|1bt>_EA|XRT3UL|;zCbfQcxByU$<
z=6i%l(Q1oav$0JBRjx!%r9Wf3{gNN&1rxGjC-P2-mO3%H_P|UY4PnQ#$FJKBwuk`0
zAtlWBpHs2p1mWK^cHx|zZ;X&IpE*%=Wtli(R1z~&bYiHV%w~(q)p6`8EPNp49jCdr
zj&0;d^?p_}h2KML3%cdlO9oD(6hE33JojjN3>`95U%DbMUvlgCwdWFM-Xv%TpB-=R
zPXh}zLUE3nFtDD4AATh9$f^IzP;1Is_zwPAoS-u+;~9vm1P=3>R}hcvoD|R+qtvcH
z5zK+@Wyb#eo%cmatg!~_kvpEa^F48Kbfs1W+++*i9;oec)cMVbF<c$d3x7uNhs5`a
zqPCDeDJFLJwUE4^h$aB-J8bA>kKvpbT=Me2@lz{b8JWNNa_KfK;$Z~|kQ`(`eQ-9C
z|JB}GMn&C4eZwjeA|MFTp`cRIAYGEu4NA8H0t4dE2uQb-fRfS;B120IAP7i<4Ba&}
zL-U-`>$)%3y5F_lFYl*!J$z;UC->QV|MvdvbAF2Xjy|Tgi4W!Sn<IF;dRN&iaP{u-
zei=w9PTp1!`8Nx1cOMYWXrX)TZHbz%>N61gufdS9QhWXwSb%aof=^>AD2Hf#xOkga
zg8bpbzu3PG$qLDXWC=B>yMG2KSfO<u@RZ^|`6EX{^cM*m8d{@CPEDu~Txm-~2KXnC
zAXwk|IA229#ZCrn61Pz5zdk_%unW!wc)4AA-c8$`U(svK^Y0oH{Kln#xn$2v)&#gQ
zaEu>If2`ojZ#9m50V!oFk@5P(telg}NJzbEJ>1ir&#f~%umJpzhkDPL{xN4?z$|lN
z!p{+xCqg$Z<!nFOke@YhEg3fy)haV1|M&0J5b!y<hy&$34jwk|h5r$5Zt5ine@2-c
zEfB#yalmfsrIY&l=NL-_U2pvL=K6#9ho<<42kkMtls#l7W;zsaS#y)^1OL&3bG?A)
zvn^?A2|m@iHb48X(B1Iu&_&94=f8aZL@)i8nbbKGk{Xg&7s)_s0>ZUU%WJ3Pf5@1B
zCNBV}cEaECxi_Ikp>?rSV+19S;gii*XBPjU1cO*K58&Ol$;RaK``DkSuP{`61rHGs
z2g;efg5;mMUDH*G*Sn9$P0!jOy)-bw8$Z(UaSoY`Y8WZb{8Nbd`!Q$OgR9mgs2yM5
zPA=kuBNB#+YSt$T5Pp}%-&)@8FaQvo5>PM(N@WQ2x{FclAKzl}Mjjh7);-XCwB0jF
z-^_`8Br%~oP3~b}^GKp5GxFib-}*R9$!da?j~bhwwMkka42Gh<9hmb_lc&C3H2Ol5
zc2H92I5ubwxp39u;LnuypGvquqWI~ZCy8px`@hSc8R$QS(1D}DzNmZEDxS;b2A5&(
zwj<`6t<p{Q0O{k5{CvP6>gx!ASAhlm!)6oU@om|_jUyZFQVv$-R$x#=$V;$ZD`7$c
zKv`xTMgP44`yB(+IU}5t_a0%kVEhXkPW-*y#i2uwMA0P>UZ|?AqtRtY<5M?Fpm>FE
zkJZfL>>e_441an9q4zwi-_XM_>Y79|Ci%&jH7(zJ1ekS~D`<`K$>*I)V`f&A5Y;@;
z9$5aWsR2q3X)ST(tu-y@Gy#}3Oq!sHwUC}t+3U9tturZ3M6Xl{YAEBMdOn~Ifxk7*
zwxKNH451?oJqtH?s%&w`G!65?fQ#O{l3g?WqJUqSh$!n~4n$Tp2F>(n8&h%cP%>^M
zt2ZO=|2lsX?QjU6(ne2X(`(e$M>MqVFx(9jOoIkZJ-drzuY(yXoZIz!Xi-qRy@wWu
zg%9Aj+w!DR*Y2k~04?j)(Q(IC4t<8|U}PZuH%rr1Dnzt%b?h{ES4HX%&C&%vp<Deb
z6Sn@Cw}|{k^xI@`dTj<Qo+bQF^nWr$+}aWeoNT(&*|QQ)8R&0yybkg|A~%LaL0Xbs
zH9;11g{~F-gw-BUC<D6EL>L}nLDdEUo%LE@H;BHCB?8^mSntM%#v8e?y{GAq0g3aM
z^fOw}8#~Y(E)=ARB_wEl9RZ6<8oue`lA|PGb`y_J&&7v#h|lc2;(S-o=Fzt{pd$jl
zv15bP@Jk{CPOT%lS~ZJd#(NaqZ7~rFw(ygb!!J^xDM))4IEzu%c<@QlxgR?3-%bF1
zRp&)kCVb_O>xY`?UF495K#vJ4eZhJ51OCAIx2fNX3rd1&*-)q%54^&0J?Qk;>4@{{
zx;ny8sNN2|TrR-_4)Lj9!*A=4e2Bz9h|s%C<?a(J&?VGRi7H1GFAt1n{i%@!`viXb
zHVa3<AC_i?*Mby%z@7Q|9pQq#u2@GddCcUTGkHZdS$9XreTtZBO+mo@`6IzE-bZr1
zqxh*e{qEmr#aRTF)`E7orSt*5Z1WWNvGeb+%8xZpwB3ofZO?gkucL^dA0wi3+R`ZK
zO5Bo8R{qp}W^SH;o~mJfKEc0&WgqX^^BZtzviQ&ni;==96h4_@QIuN`s=N7YWe_oO
z>VIJR9f(k%X$#BvKA$PP;a;q(;u0VTaKmOA2$O75?BRkH6S@mBGc?zmdkSXo=ET}t
zZ(&-sxcl*HB9x5<+)vEc9pn~h5nBuLM@vJUX+VRJfG_Es{`ws7kR=HbMYvjeA7lUJ
z0+2w#qtLmq6#J=O9hGdD?PWCJkoSGi1rF2Ct6;AQ{AR4QSCkMPZ*IazL}j?+iE>}J
zu89WhF7sf^%+|fOtT}$VBWqEjDeIMm{f=II{)4BdUg=%C9R~>)L;rmTugq)e&qjk0
zw+G{*%OvR90s&W0F6mke5tO*GH>V5j(Yd>Msv{0=-$1kKB+!Dbyi>9L7zg+@#o2!{
zv1&wap>;QG>Iu;268%Bv|M~#HbLnr|Mk=B%c2E*4CXPbN+Zg5NXr_V|MThv>|L|)~
zut5jH5=#%B6V3$J?e#*Ul)30TL{fTZI4vABxhJ+4wKR3N#9445PcxjFfaTt#^ZS2T
zLIA-%<P1|+?@n&`rfY!c4~wq_7ucf{5bGO-dK#3!`3=%P-Q>(ax66VRD#tAXle7ZD
zQih49kX428N64o!dJW3Y_N_gm)Tp1AS|pk3P-vq1&8|LFiRUU8t=}VmbD6U**b9N4
zF@*gOL-q}T_|+H|Mwysb_2FI`S~%Q3+HIoj4@iQlSy$=m=#iod$!ca_&0cu+nf!4-
zphTz)>-LPV1xCo~cJfP7uMcx2O+l1!Lmw{F2;e{Pf)+=wu!L<yzomsY=z24->?*lU
z!byiYKt!#5j+SyW$j^V#$w&SaN(U+q(UNuI(OHuXOH9zE|29oFWn{!C4a&&>K-q#U
zualc>wjj*_E<&2?y)Z%R#9qi&Uf=5MqE={KLmKW^<33PJ+S8X(cJ!$f<Uu343TzBS
z&R-kwqYPZ#+~~LDiD>nY#JCSFTgh(+r?98_eu?pa&_c6WDSG{pFDNje1v=v}{f8~O
z<Q^>!o`Tw&xtV}#xOR|2NJi|ta3LBKYaD#D)SU!muj!=TyU$-(C(;XX<Bj%;Vy{vp
z!8ZFxcgEji1K!D+W7@B)rfy*8egWlc+ZvzEa$N`cNJ7eg)!8-u)`%6gjQ7ikeZfaR
zvCA{O<@=apRafj5wb+)>uGq}%ddN3N<G7TbgD51K-<fEapCXdG0}vQ`QP-UTU6CfW
z+ri_ib8p&PkhhC|b5Z;!mY+g;8}@frPYgoK4>Z>@*a>K)fS@y*!>K2u%{o=~!~J#A
zcLQrep&lQu{&unoV_~i*ZY_U;)6{FfgVTkc86nD#>8<5L>czd~fqG*Jqk{wOWL>wL
z^vqHzk8+K&+<V+RNguFoNoY$Xydv^d1Aklo8<uj)#t-o?rT|i%Chm9zKY!8B^-*|B
z((~j5gY)Rf<m6H`IKr6yZBWo^cKH8F2p-e-QpUbb0i3D2mouAVcA>(A$%MhbApvjF
zLiF5}{jCLRsPNQB_SekSpnM@v7O}Ul53;%6dUQWV{e`QYUce1pjs5?F)ir9mWe%FV
zhY6;AbPD!k3tz~k+px!~Z`UX5qbozSVKf7J=nz=Vd`LfX0n$%%C>2^KWv`jrl&25w
z@|e<#Y`}KjP6OQG<{AGv^F@L6fzz_tFwE_J-UyiZ&H(mGJYmF3V%e{C4!z+aD$`e~
zhu^kRbd;Tv$sn**HzGS}eIpA9O4reBKYKz8Z2p+vc9h@j+@tIDr#vIH#C5!%2szcK
zD{kv+a%j!$aFB3Jf0=QWYuWaP8^8HKsQ9GX7HJfGGbaL0VIBlJ)A+?%H9z}!Q5`Y)
zaH#er6WowL!*M+H53CKX8@VM=eH*2Ei6Ry=+95XbvT2OlRks6mW*TCszsga^D$MEr
zsYeB*Ees(<uFG44DiFjjl+|h41a1ZY;}Q*?-$oZ;#S=FW)(~?)IU{x@P+koKS2Pg)
zQ+>I8Dqcr8vlUcVnijy*o-8{g{OkX{CHosn<w6B|+Es#e;m3mDP=2<f%>qZKl7B1(
zI2}`MlUwfYgs4xn2YJOj4`Nk=aV2WVqV+W%c#+mDBz1gpLnILl)TAksK$j!a21ioo
zkY6&Ke}!ryK0SV2TYQxB$g(+g`wHEolhyRrJy@beh*o$`{bb{8qrk6;s5x4z=mE`<
z_++Q=&Q~8UK}I*F;C`X%2)4IvQbQncuA9cJxbXov7NXM1*DQzeVNf`GO=nDq=@H{P
zIUMwzE!O-xI!E9CndOk@+J~-8SYe;m-fh$cgv6yPhv?3wiNn4Sv%#Kif!7~@$;cW=
zz$Z|PVT+CgeldF=-Z_z9<kW@vLCFA3P_<2E9HIW5omPpQJnEMm(iv8)x)He#XI{R)
z^4Tgu*MH7lkd-!{05k`Z2caS8AP6$WJlTo<>pMwVi)@g=K(|wz3aT@TJ8ksX;f7MM
zN53qAvt#)93&Be;>ba>Tm>lYc?EDT2-_x_fB%G_5X0RTzs1<~$Rq;^|ts@SMo<qKq
zG>@kNDi^s;2-Ctux9JRtq970D6SwO25&;zF5F@d?QZ8j%{LQ5xQ)@Wz8P}L+KD2|v
zjMmAVLJh>LmX6rTIIxAv7^Ui!=$K`w@J}BtlDYR!7VC;KQr!@I?hBpZMw+UJMUuxP
zQ2g`?`9!ehz@M9_@+E<yNKs=X3?Mq|{HqQhr#S}ifoL8%#Nknal=<Y2^lhR>nCsM)
zNZ7;yyGu`)I^q*&2?>-DkFszB+#m{dd7@@K5q$1L))f#Sd-{}u7x4;Bq8BDc4Io%}
zhxwnTd$!`i)_zHnK%pLLFMa|zhox4<4(f3DOcS<3!6vOqSA8wGN+m9ImyC2m{5nSa
z>&pD?GYrel3B!kp{ZQ%q*IMx&$LAk@Pm<Ag_3rVm8M}c&za!PIe$2B6OHqR<3dt+V
z0%P)HzU0tZK28oY>Q~Tise$0b`Y~sSHrt=dZ0;eGCPK?EIV?$hR+1o$7BgzOhlU&h
zbM|IrX}1G6HsRdz(4Kh^;di+TOx_7kcA>QfqE4aLcUc`lgOir{=K1fb$^kCxLU3IO
z2FMy_6bk|byBu66q&k_9t%j&dFMuS0Qp2m2uHg1Uh%RFhdLQHvS0iE0ZHAv13l6}K
zimjKndG_xjn`oINH+KRaT*EgK$PRi>&~;nFKyVnuEM%3@9lco7M)<fx=_sj`IIac4
zPZ(SjJ__NB8R6!3eUNZ02nPfBG*_~Q(H`0YQ1-Wh?hJ5nJ(_wYo@qC%O42Sx)jd4;
zihBz_$zmX)hY4D|bN3pE8VWd}KLu?F{BC0pR=D?Z6J6bM7_QyqM+-k+8@#PHZDLb^
zolHnH1d+uFty`P)!+Zni)+H)r8DLulwF4QDJ*vcjtlH~5oUHMi4fMU-odin9af@2u
zIWrM1r|WFe9@ZQAu(^`h-UZp=qm5WECLulc>Z45C_a=D#x?PgiIM`tPXnX?-%uruI
zsZyj@VkzP|;`9@d|H-_wu&9VAA??iIVg;LM`GHo_$lS%Zq;oY*s3#7sfa&$rIeX@p
zw7I0dmUJ7kJI%N?Fu_s$C$N|~dp6_HI{Z0WlW$1s#qN0=e{tMHrovZ;0t{HT1f<{Z
zC6V0m`(Ye!umfSv5U4-@P%rK-!L-X0*Js+`r}kbuBxin$?l|x?GoyqEQ2z!5X)la$
zaQM>cOBL7|3XJ8$gxVfb#Z*c?BjOc<2isNNc_Ikn(7m+Izb#KP?qHl&xy&KXJBM^=
zU3F3L!HH$pa||=nFD<Qi!1*nnqP)Ergk}-Y6vv2Boi*GeRb%C(`YBItA+1w;`_d?X
zRsi;FHMT*c)zPu?JwY5GQ;JEx0_i1tVX3;rpU)Aam_rPeKX?1>`B)H1Fh=OD7UQHY
zx;Sg=`_0oGnvJ}0!w9A_JhZ+m&KSNZ2SL9ukJ7Vtg6wCu5Y+yjUsKE+@*lrvnFZuu
zPXq2wZzyWnaz)Wm2X4FyE&C;f&{y1xep((JeCwwcVm>ya^YQ7NEJSy?Cf|w~RAs#U
zmH#NT?vunGH#?}OEKva!KqOF=`0_pw6Ml&l=f;(($&nN)hvVbPxPGMHwFw6sJnW`o
z^9&bms8~xG|Eds{s5Lw*|E$=cj?;&ZL^sJ>;f{W_i81;4#HFALbt@4x!mrzm5AY{-
zvDtph2s9kVm=ievcKl2HMOGFezr<6x5*QFgE(!>rdz2d0L8Uf=e}lE9!)LkcfvRFi
zd-cTLVjmg7dL+cFRro$gp|?cNFKOq%3J0$;Arrzpl#T8vbSeAOOJ1D{E;*dhq7d(!
zA9N{LB%t^=dBymnk`8Ev;-x{Ymgq{uV`|Z#!DohD^&3Cm{eu>Ht0zF<R8xw0O(Cl9
z0QyraDQ5Xl!Y$%uZ*B7g@r6gBT6BK>9$_MNXsd7>b55}w&`)b3tsfh}VI!9POX-)W
z;DC##PCrun2D%UWDWw{v;)Lv0-H^~FHQAkV!o$Clleo{8WDKZ*IfOL@Ff!DE+?(>3
zMqXu2{8re3K=;?OPVQNNM%x>BAWfdHZ;|LQX_~4i^L~Bt0MtWst`t?E>xCXO#Nj%U
zLdeDiSfY|-_0IhoAZq~)c7Qa^KWna&G;hqV5#Gtw{nh;YH6;>gioIm2=SN^rZqz-o
zo)hDIh?7FVEM6_@<9*w+AwHblkO(D;*>#gk5=-^b^R#g@kSbZd3xeD?gcHJalYm^j
z-Moh^&y}Ii`W089Vkb2(6-~G_ZHnQIR%ubWaQv`?#DNucgno|y7mJ~^)w>g<&U_UB
zQ4YId)p!6wW8AnJNr<8nf13F86A78$xeTzej2ps81503x`>TUsR2ot{{izI20UQ(*
zK{2BG7fxffd?e8HN)aHC9lem!9hCGq*I?-^mN<vfCfWCV?lqAFL;-sB1NEckkotqc
ztpPbT9^P0E`<paBM35?D)dn?PK=o!S0v-R6^X_eN3#KrJ2qnB$@n;=s9%lia-jjac
zz&4OcQLU#e$fS!*9xZC^ghikE6*Uv*ewLnlk+Z8|e>sGM){W*}*@}QEVd{Fogfs=x
zplL_>bbI?yVjBjMyTpfVx+UDeF0BZK!RW*^LXs=yDMYR{I36M{9$zi9i1aw5lUrOB
znpMAeA)?WPj+ALM?cJe%9^#?)j;`y+ZiQ+RxhkZ3mE9#OmKmu+ZVlSRiqL)|fc91H
zfqlNT)Xkvpxk!awfY)TNnS{W|N_()CC~#)!R>OW3xc_!Az1_o=QM9V|Pl4z0PPD}k
zF6w2NN?jbyBz_-n&9%3XI50tbBpR4c8W^CObbLi4e&3((B;uB|6x;^k?fKjf38=wI
zJZ{yvf}Ybg=p7VYA~HVf&Nu_1KjRx{%Sw4yB&dMU4H5w-xi?<3?yD0^ZdcwOo)h})
zs*Hh*En3EeOZ*GLk!HM--C?scJVDWw2tvsM&b1MO1J!rTFO377l58yE%UueT)GcWa
zld_xdFuP2TxyqN!zw;(KvW<Oz{4hU}2<}7bRDdT3{6w?o4OO_x9b{MMfo<-iO8^0j
z+zpJ9yGW1fI}G!>i&0Zgt=s4up^ia2AdaB<Mt4#x8OSUXZkX+MbD{7;OAE0-=klEQ
z<2$J@vtgqn{HB}s53&?NF%8@{@=NEueKG7Z^_gE9inM!^{5Dm1Y4u`vTA$Kt;MP~x
zzM?^op!sbU%ac@jxSnSo;vpo8j>4DCLt()-=V2iI`WF4Mvw;E=frIYIN-!;mbK))X
z3`!=eC4H`;RPNw83tkm6TZ{bI4<9A=oo!_5<FkB1tpB=+aAAu%UcBS$bCahLBEapa
zUgc_zjUoo)Cw_xj1^2Qk#j(3kKE~iPljeOGVROv=)Y=VE+{037GihpZmU%`*EyhQi
z6(|F|w=!`d#dx5NqzhQ*j5Mc%17!ydjLR{}>z`}Miue%NX<X-f?JCeHkRI%|E`t2m
z6%SVrk!{;5`uY^1yCo8OhbFjIv`zF)G?CzHD6HewV|tI2-K05tT(>`Ie&dSF-~5VM
zX^8lR<!2}-?kzwSXy0%D;>ur?pN~ebVJQ@64McxJ6uS*1X?D_w1Ob+RGMI>zKSr>n
z@plq4plg|b2#w%Q7yHy3{GwYDHHY1fG!WX0FGgF2NKxM>TOGHpfffx9Zb>&Ka!Pq}
z=D9s{8JE^R@%(K*P>WkXyz}yHqq`9`H8>vLS+I$!dDt-=c%HA{mVmox)_KZ#RAc;Q
z$nb5{lS-Sq+nRky1Xp54@NJz)+<@EUs1CF@_K*9ue@JSvOSQ5^#yRrNz~FugO6^sc
z4X*3GX7<$O3n>Z~%dkDqWvL!Kcq%v3G1+ydc@HVG+M%XvetTy|&G<;0_w|PkFV9Py
zA|6(AvlpBd7dl<Qp?nat%=PDDOjSW$e29<J6FM!LzH)Yj3omQaFm=&;NO23+IkD5B
z1lHVN=51b1QJi+?l@j(x(VCfMEQ%B_hs@Bj9~I^X2xL}Sjeq!KrhOn(#|TtHAFnlQ
zz#P(!2m@5vmtIuiC|BjAO?y7^oLODW@PLSagy0>Y#&O)oPz{rn2#z8?yrNUX3ao}<
z#+x}oo8VJ?h<fLE@efA!u>B`F<8jd=dm)<b@R@xs1yW{^3HxlH(Rn6Agra3&D9=&L
z=Ydax-9qE7mtx8Rt9LOIO8^5-KmhNULV=D7@*GM%nP1-;{ehAo5;&mL&P!-zK;CfK
z(13lwmviSTMvGTD-vCBZKzvRg76)GVjg}m#uAYmt<EMlaM9J0or=i<B(>1%xBbt`K
zC?S`&vAfts4tAU@;EokN1om|X{rWr?Pb$tFD(_!~XC%mABGh>bZcL2~MHPiu1GZla
zKF~aRizNX0LIaZbf<1obIPvn)rgc-zM;Qhw=s)`Q8u^(vkBO@o`RMTueA3faw4<R4
z#`y7ei1-h0O!5&%yc(+b<xEPBT&K24k{CUH8pzg&_;tA-jXzT#ixE7qq6f7n%M?o+
zz~BGG=@*tZohk2}cdH8<>kSi<3@@TN=h33Rr-t;E5qs^6na>a2-5cDH>HpexzFi%<
zg6)EmlQC5enxh?<r0b|qGJg0OduM^XXu=@{=2O9B$lgaz7yO;4NKq2pOxDs{Vun7-
z-?Yrz&Mz}tVyqLIvr6hRT8U8Jxi0>4ak~$Md&%F9`iL>{%h!K={w3$i3+{$1>vW4@
zzZ5F970*Q>&lT>6R>VJ|&+S{cjfZulkp$XRa$Im>NJUpGx|L{vBJY|jnmpL+gOVD+
zI?L9$1M7?}2P>+j^@w8eTi==wfnEwc8nv-YbNrZu;GR&ho=YWJh?)$F6K85wMF!Gw
z$Kl}<Kk<)w^|(pIR9>LsYjQTVt$Ab}ZJSqr>}97xTVfEA=DhlZ38g790!tR#g}mo0
zD86aa7whQW(7k~Ty=i2GfyPAX1^19ylz2=+>1Xt#qRb>UR+`tItHy`W5rxo&###TO
z08!aWO}Ct9^hilrW346KZo-lV;rKa%y9LHLxIYH(ex==1;iC7bdzO<D$LV<W;VV*;
z3LONoYNUkr%TI`?*1A*k4)+KB*Ns)Qmy>d<fm9Mi%DWp08tJDokD7I*GkK<!ga@ip
z6Nc=m*c`{&gP9tv`3~4Ec1s}My*R6oBiid%zb8=t)QxNC8QjbGGh?K767BOH(+ilb
zhC^XngG+K@+zqbawej*ZVOZVv?qeW=6Ic2+hJX@!kK$2t`t0^)0t3Qz^$i)c$%Iae
zmMgoNcp?HtkPY1EMX~fC5#&gy$9j|L4gc^?Zbe1~AZGs#G#IF3t8^kaD?$5tjTXnT
zN7yBW!5(^V`3deM$qt=)ia;)h62t>a^<yW^7MNVL@Nj<^6=+By6759pk2cS2hG6&E
zJ;TO^@>e&C0JXLal|f9qJImdCNgDQe+@lRRmzRgj)ONoz@sdC9^U-%dsulq_@<dfc
z4)K7XnC!*c<GbJlEM-9#(XnEoJ|<m&u8}i1AjNUD>^_afdOkOBg1zcR9A}QcF!=bB
zrHVS3gsre&u1SQ!m1T3`nXjYXTgn36w-bcnM04g@zl{h_aB*~mI<I#F5OJePs?E#w
z{^KQ94n$f4iEq_uoG}qw63ZaScY7(~vP4<8-?q58cu*?Ce*7-BC8kQI1`Bu{sSQ0M
z53+$y)C^`q=3F#mei(A=eWL~7h<W=RcvapbZElKp9a!JG<-Tt}0X17DGca=M18p50
z>I6c+tHn!2g>1<D-1hX&jl|%4#B~hiP5n8vpy=vGS`MQ=E7A<Pa<5W+C`Qzukcck*
zCC#S{N4a2+_4>(4Y^oj2cy5|9);U_yDq-)bLL~}nGQzir;w}UGP>;$yXYlrkV(Sd&
zm_8yXl`^8mu#0Tuf>G{XpB)R}2?(?jYn}^%6KDM^v(3xr&ex-P@ws|h96Zem25>)~
zq<7cu%%q=Z`9<qFTSeE<n?0G?9cFqlEY<t2Zq*;=?Ez+vZ7H{o0-6zeskV^B1}T!!
zNuo8+ENi!u-aH#34-QEC!CwUz6?;HEJ`qn;70cLRw$IEq%~<voU1s>es%gIX4(fe9
z%h}$i-AvalP5PhCjvvlh>~dLYn2K5&v+nE;2cmFf@_bI$sbHZnIsC`;&S|?XwrOpI
z3&~H3z|z}`Wg>APQHnP@Oe>qzrQ|x&k#C(!H4vNo(L>{wfOCl5rURjy7QemzhIO1+
zgZPPBRlnU%mKxxWx~3nxdYj#HxOKhbXyQJy+`l#2{PDop38O@)$0^|bm+atMO6`!K
zzRr3zRbIKAcq2XKw%#@+AHJO^PRCNxlG0@HJx#8Is(hJh@n*z_QZeAo(a5%s2qK?{
z!{iR|C7X?85c<|%(Gd-1q{KQ0Hrn6#X?M5uAAbWSp4&CHE!WioTyIjp+d<k|It-lO
zO6SGZnygGKpY`rjxw}ZI)b45~K43tKUu2yTiMmK!_ZCVI#|2kbbcBbteQoGE`<4OH
zYYsZzIU~Rm?bO>t(g`}%tA^f`eLlTSFb^E~xxk9O3;@$Z>d0e!Ae-LD3n5-DF$Z3$
zXEkh)%L9oU1#6U_&7})QXl$Kdxq|{)`xctS+o7!!@lA~V7*Ar-GCV<OR<VMm|Fwv6
zi35xSBn;Gzt-2Z)SC=t=bCndtJopABz5Q8^j@(QYK5-;dDrwT;n_}^f@Fxv7@43>%
z4SiJi@>eg@b(`)?#{6)NmFlvoLD;u4-+kuFMT*z7b$B_)cm>JrRaFuVEfV!$4aytw
zqKf08%v=f;GVd*(#^Df+GYiMO!V4^qq%hanyZpHKW8z2M-cnCPDay)wYj1V$d%Z_r
z!<GlKVHwpc3=&1huzA^D8bpmSt~S*C1*KEP>p{#gFT&GUH|Ux)8Hessfcw1%%`-E>
zm0e{Wlf4n|V(lWECqqMPKe*zXn>V$V$1W36uLOCEPGo^gHQh=CbEBIZECbZqiurj%
z9~5AkD}FV7lv{}f-K$&d3CfSlrweUm3~!tn3X>?5!;2Ci8^few87zH<iI5)_HPxw)
z+-$USa<=i7vMl**VYXk|aj-6~^8k-#Ey2REdhR2nO@<s%brZ?x=er?Xo36wBgKTGX
zF#1DLiaVFcH)6|PYi{uPRkAHsicyvg_B=wewM<6t7P8Xtb*sRX<3Y8)<1M8|jA~(t
z_f0pEu$s<Ivf&w{Y$JgZO40p7r%5%0m?gOoLykY9riU}i<As~}(R8DK(SDsON#wSy
ziVDH|(v@CiRXXT(8Ta{JidLRGT?o2xXF0qMY7YcWDWADxb*74J+YcEFR~&b!4^~hV
z-1OTA3Oym```{k`oz<<}168&Ilv9V-15aglMB47HZx6C!)3p=3m}nMrrwfkBQD#(#
z9{V$9T@I3+K;!itlcXyf=gpkc4df-}Vuj(lj|qLCIusnL{93GZZ$%-292m+XUdn4v
zB4#dndY$D}1`i$uo_gGUIqG6zFgh8&F}D<29VB7`Cu*13xTp4W#(QRk&+fUsz!J+4
zKAcB2zPh8trWPMkKh--xQ#n_~FZL9Y7tSY&6h`3j&*(<%2Te4lbT($64s0x~V{!OE
zUHm<IOg@TOk}ql0W{Q{4DeHJT6&x3FQ?q0h`tW%mC=mD^pD6)|Lcvm0Ck?WV$m?(8
zEvJ$rtL$3^&ueJ(Ms8(08^pLk4?n+kba1Vy;e^?`2MjklVxc_4D=RAd_m5F!PDt{1
zL2K=h1DYL7p|7t(b0XFdREYEXZHM|<#->_HZ+YHY*7{eYr)uQ}SN!Y7BQlowFXBq5
zLL<Hh8eM^86&w%h_xO8khfD<i+(3qYb83V+jpQ+ARIK)a>x)n^P1GF`T9bW0v;7)~
z+_{yxV00?-%`sttUN-7%)MgiMe8yB!#_)LkTCZJ4mzw3PMmF}{NnK8XGwgxZpA@-8
z?&sN+@n36*kBYK9WuLRR=BLMyaUAcza)svLskEe)$R17a!FlS<u{T|ucAFt<vq$H2
ztzNuK<_H>Y*fia{cMR=P(r1pM>{VmfD8$T}dt`}o=aiIX+aU%LcK0NQ9|^ni)VC!0
zRy?zcC~~Q}LFp6WeM)>d+cQxd>swU4uZR$_#2PWNF3>TJGRCz;d<#ctP9STPm!)H0
z?llf$<re4Ch?bGGoX!*p_<B}flU#gg_RsQ7bJTTI0l!nU(-fzDiFDQ7`;-xqc0-LK
z8W9ie9Q&@1*l{6Nw3`$-=c*|v`)_OeN4p|^N*pI1HWzoVKCA?GuqXeRAf9U-QvDES
ztl0g;s<uS8PrUOc?MtYVzegPZ`#Jp9aQZ^*`S<<%QDE4f^EgF{yn5o#vaUZy@}F&t
z<=K!cymMcT`oL+ddbW+igwN3ad_?UrhbJy9X2eEK6-luMb?jDPqB1*rIorPZXhYnA
zcTfh#(RH15TG!MYM|Ip4i;-j9TWrx_7mLOk2U$1&p3^5cqoc?ptHL$UAbn2TrX|kn
z1SQ4b`Lj@w55Bl}6Zxqy;hT0A_V;%>1mAXj@KBSyH5_<5R7xE%DsiRnj4d}Q^2dK2
z!gyfaLXYX@k>t%Z!Huh}GwdDX(<4Dv99$#Qt{)q6S<hPQEz5enlB#2m4`3<cO6u29
zVX`y<ZkVWK^$YnLcW7RCHI=>Vm8Z+$R`E7kx0l0dmGp|e1$sAl5Z2b+J^Ar;xERRZ
z!;Lvy*Hit}s;=)mRfccUnnmfBVj*PPOVN5o#=PoH3uJpb-?|4e9O}<_bNYn*CZN`@
zAqTIe)XIX)962ss?~i=q^JK7p#(f>4k2m|XetG=NT1nrp?zo6^N!61TCBf87Df;4m
z6JyIbcVYOkSX>E(-=3@}21=>wKmz9tBc*rE3pF-8vI|au?_RKk$d7gAD3NQTGjmIs
z<cO^#(m^}AAO0R;xETVrubX`H!n~LR!|bv(F68TDBID{rk+YIhIIuX=fq7az9zqBt
z@UBrI<o@F`TNpE{Mkr>?{g<N%ZwX;&JT`tt-m3!~S7Y8Y63S=4$cAlMrtj*H&7{VC
z2+Z=KEP|5+t82%%W*L^~1R#QWuA^u=N@jCk1#;}N(S;AG9<YdcM}$h|8Gh&V_diP6
zX`mA$nsO~NbhAJ-2=)*5+|j02T23{FG1FqJP&6t%W}*ld$YNyHtraKVImPd=7#F%1
zEN*fH9XfB8=cK%_z97-k_VZvHtZD7RP-3nupSgF3cLl<KidY(D+I#k*Eh1wS&f!@*
z+qQiJ^#n+s#q<EaV#QMX&pb-d0Iss7ZP4AR7lEbRR_0SZ!VFqBkW2F$J*ja8`wK&t
zUphp;X)mT|N;x*K)Weo_O)D9r3wN&MbQbu+UL$@Qk$fqBd`X|E>qusf;E8?R)LH9V
z+rrV(eZNv#=Vp7hdEk2MIX=DP_m$D&fhs@9)Jn6aHf2DTq*N<!mzhqIb|pS4Fd%+U
zN=4%#_aNT<#(FiIW@n+>j?t%P(X^AM)HuJ#rFb0)i<&cQ1?vzxahH26aS_sqhpHr7
zSwU&3*GST@U@}Qu#Z~)ee+>_LtQO{w8}-!#yW1-AiPmSNuSCC;`A*BA^3RHwqIKFN
z{L<i;Z~uIxR*wmu`^>xiuo^Zyiwvwbeid{sr;<6{`BJZP#{4?g_Rg}-HiYPZelKzL
z&UNsCL_LHDGfVuGR2X%EVNB<Mw>fx&ExD|E5?6@*$$zb<zrW(|arG`dR@2L?e!sGt
z#6+82;lF=N_&&M82_p-QV%7X-IKSs02Ig=GS<?AEqkqTAbQKHrWCvpT*GO>3<lZ(<
ziuEFJ|4jbRPjMf7n=xV%D^dEZpR1=hoUn<GB>jwk)(QPni3~84oA8c!W^}Lr^kYJI
z6$^SQ;LQ3zOA7wM6bGh}aRlT3yPpjP@URk{wSB;UtN->|z-<#p4iY-uoWJ^c>jnDR
zoc~e$Zy)IH($&Bu1Lo$t{_dwN3fMZYk2QDxXE@-eBnoev8`L}vlK$@JkjTWbI*O<I
zueJXFXfX-u;)^!yiYE8TQoYY+pKO}HIi#_$uxjp@&@89UUK|+UnL#440C4r_g2g9e
zi5qV{*_`)jsCYi2prWvKKAJW?HDO$L%)x7`g`Jv8dpPxF@KxW4`QXUj`086EvW!%v
zo%E}_(41Ro`?2+EZRx^i7yZuxE-l%G^78eO?p0f?Kl3wW^-XN+Y+t1mefh|Cpno@=
zzLwM9gQ!@~R%~Q=oquP!MC!xABuo9Td_Ei9(ziD8-AgqEFUpoKwo}d=1_was-*J=!
zbsid_S7s+Vz*F%v&&Va*Gnch~7e_H6f!NXqxFBQo^LOh-CU?%jv8Z&meRctFvn8iG
zrHT9`5;P+rzP~<Bt4uZWb!lF3h}%YwRgk%1S&i_`SQ5Bx(Cv}(sTJlg(PJhXE#{yT
zmx7fh!V8ZR5wu%a*g-o$XZ>c3Eok!HK>5pVH)E|R_*w5PzCP;vy55g%ex++45=Q0Z
zzVvRLRW4X~`fW@>@zQ$r{0c{7)yoZ@hOeNwyAng<!kFBIvJUcL*NgP(XgXBd=aUj5
z)SBMhiF-uEB3_uiryB9~+sGT;eEXteDgE_PQ;y6xQ(Rb3VSnMSrS}e#vddHR+b{f|
zkJX+M{u1dY%&#8g;;fw)tz$}{IqA0>pKq`ny?0i=py|c2w4=GaJ#w06?!9614hwsx
z_355^l@mq%$WyX<XJdtKw$lUEqKPy9hY{|?a}$@W=Zt0>>9l#eb)k5AF|+qpCeJCT
z#Yls$np>Hy{)oXL*oTsId7`A)Y#%UYqjo2o>A!w^%Vqpsb-ypf)y*JH$U~i0b#65M
zz~>_3du2j~)^|tV0^O`JY-l!GNE(@F!;U|-mOCeyk52hG&RgNO<E`CmO>}+f<@H+H
zHhZt{eeBKy`7cs5)71Ww4aU-gfi7j+x%gujqeMzn)4KhG6velwMFUW>ebF>Vx+Rm+
z*}Lj_IyHEyw!`EK%W0IaeAzZja)W~4wdG$4UKLN%k0{+)n@4M*J~!Qq)0rDe$>rCb
z#i@dgq;gEA79>f<!;h>Qz4xY%rdz9it}SLfZ1%bsaDh3o-Pk%<yE{4!J-RQJ)JQ(<
z#r5N}NcQ51&vtgbn&`!fztbnTuu*#EW73m-S5oCPpIfZUk&>=2qCHEuf;oPcztZG=
zrPc9rd${83!cdd72UEi`uNuq73t8ER$u88v#*Gv~;88<*tD|-BtjKU;{beWfyMXI6
zi3h@qrLr|m8+m0fb00lf9s6=n&k@>T=DoZfSnSbXW{<y5C5Z2E;HpE1sHztO2Q6*-
z*!uhTqXi<<H)lfuN2Vt@$JYHxdq-jZ+w0Zy<xZP#!Zss#p7%Zbg7uv_|HVMr<2M&x
zBn~b<I;t73FY^0I#|DsC*vI-sRP(*O*+L>7V*NQME?qZ^WQBWXN{EWG$XMM2h4v}K
zMtW%d1imkI5PXLwYC7a*$SmJ`*nB=%nvkw#L2hJ!2{RDCw?k><YR>z$IGB%F^<@n~
zL%YfTTV1uv=a@CU$r0d%mBVxLZ7Vqu)GsEx^Ov{=1_lcrbGa)lcN8<sCpvZ~jpd$6
zu-PiA5KsBgC~OIa`ZpXhxW9OG=GoxL%`Qnh>MPqU3Z-j2IV8h^N*od{Lur^++>A-0
zWNbgSl@B|AyB|AKvUt$1-sfT4EAGrVZ|e`9X!M>Ad771Z;A20!haq_V>2;_gdOgDK
zF|<DPjL3<wVR$5<_C_bEX;<iS=*3ODm(bQH*VsmshZQXM>q@kOmT$99Xf++$h<`K{
z#zXe1DN=6EuU6jp&}1j-eQF+s*W+dEMc`tiRROE2XTQr}wltf8U645$8?orluc^^r
z+s%1)(aQLczkr1er4$wO;m#fbOG4Ot%ivjo=|_e^8D2uOJ@?A>nqFjjr!{`4eiE7H
zGryS2r_rCI$5#YT%&Grq1+QmqQhVmEnxeV@%MJzC8abEGbWE$$i|Em^nstWngf&m5
z$4r+$`?)tYaQrAz)7pD+IZoGcboXQ>Z@GbQIZeaSxY7OQA+1;l7e?LsEix|8imk%U
z1wBWw6f4=I1#ANY&+#x$tiB74`#$t`K~C?vNxxh5HYF{s?-l3!bxqNdsj6=Fo-gTo
zPP6~;{%Mg`5It-{qv=@LbN0t*f*(@Y-$N1}I<`BP+hc#uICq?Tfz5THb{*-EP!w=9
zu3jIwzj{BYT(9<RuuEI@c1fT3y+lEu9bBPC!I96224{vZYf8KrP4uP{)z<B!PS+yp
z7W5W~&g(LJSX;|2k4a#)Ywid*yynTc{Ogyc6RA86yA?=&8%hE?VFvN`YR#Z@r(+G$
z;=1~CNUBPvwZf{XHMfSdu}t4|Wif5DHBaLAdpvWgAqX=6y8PEh9ai>INxF_@i=K~>
zlG>H2o6i#7$By0FlPCbP)8FR-)GsqEj0(J~Jnr|M?r=o6_yzVu0w0VfhjH8ThNP1N
z)#NIf=lYi}+r~vluc$;s5sRfGW+FaF*3#dKTZ2Mj(|$bfHtG7q>T}Z5Vy&Rj5=J0H
zHBGe>d!EptS6nb*o1Pc-#!%!!C+8zddaiau1E%MVywZ`3sKed3d0cpR->BTQbLuGO
zLQuTV%<D55;DJ43Z(9BaQ93dsO6aAw+`SQ(KqcZ$n(F$6J=N|o+Q7OORR}qd<;IeM
zwq<QjHCH5M5#FJAI87YUd{dQQvODL1mlRyM|G<{(+)_v9E8+vW_y^%TRz@QUAF)v0
zL<(E^sHY#ZJ{xRJwdQnv6MKPpB!XICa4o1Y?b5`88XB_Dq|~6|N%qm|D3edc4HL>%
z8Wrh2rSA;xGlW7kYd*x8d=HOt3bsJ*pyoGjmA^do{;he!@>wSl2KW5x$mcJ%{f8HO
z_S#XwFNe{3CxPvbX}P_E?7I~zoNS@`m^d>KsFhncPip|dqNvK&S2AZJC^+JdQ$f8k
z|HgYM+_pzlHCIJSX?oYMIkC^;y4RnE)PL9J-DeO_3Q|~9=t{9W+z@B?xYb6YtI!=n
zVM$s%7R;RG!N4X;EIJj@9DHH8Eq{JW#N)*!f|%SsEZi)+?0o9BH=2w|2SSpLtr@Me
z$(zxCoE^d1sN_>d`~9`E5m&bj94MqU?nA0dxC`oFXTTw}*-*~32YuwwCT&Du@eU%s
z#A9ysd)nxxL@E0#$Jqhxuu*kup{-)k%OC!AIAxWj7op3kc6eOWS+-3#T+g#}7k?7%
ztZ(JeU@u(9RY-kaE%TgubWZ5?cG}?f&S6Uk&vJ|tEB@Ajd-FmKi`yrkue8LhBS3QV
zaNPF304sNNr&nWAPd_Lr>DlAq(&Wu|m5J~f<Obi1`HL`4=k1l3ZN~C(^WV+onzGXv
zM>S&1D04|aVNdf!)xk_(wEG!5iZWy3uj?e56T_8g=gpd8O~q>U^`CvtUUYJD;(Xbz
zQFLq`wP8raUwgQEQMB=jh(93WVncreaeB+st9}uNB31sFTW#oRUsV+UF>b1IhY->=
z{H}F&9<mEdEjPJy(lA1xOgR!`d=YNU)1;5d&tGjOUKq~TVI|pjn(6xJ`TYIJQhqE|
z*)99N4?pJZS8ExCloE<hh@BtemmQz1yciit^;uG(f89hbO4`Ks@$D6HI>9M@gq{1`
z4@@_n0vMOJQheS}BvV7KEZWB$=+PhZ*(W)q6}q}Zm~+odh~a+c>w<Zp!SpAtaV@+U
zs++VBegz!Qe0yAtpsypPnvWGIH+cQ)u=0AfK7VksS@P3<IH)#WaXGi#|7HI*MTxWV
zC-<n>J!`7zSE-{;HQizD{ljqh!mLWoRtRHJ2(6|y?Q%{+)%(~1kFP_8Jf%Dyvo^2Y
zE<!xhmezlysTX@y9{~I9sbkx;X&9NTlGbAo>XgmUpP@{A-cdF@NYr=g_tD3U^!Co@
z;Tx?+{P`OF{thCaY>_DnDrjlY3*%s7R7sOtU&6<IEl>8D4d%&rgX7>Br?82}VW9oj
zFJ$tlvgEp&r<(p+wS<JsYf~%~{qyq6gO1zyh`IQb{XyxA(bLvVH2QM9f0fHF!)+dW
z?_<eO%;6yIDSeI%9_~KR>Vmp$on6@L_hygDb}%!>8E)mt*lI*3xa55&5Qb#nAyo|A
z5H9~A7T-AljEp?jlGEb;LoDAy0L=UZ+l8(Fg?HcpzBL>T@rg74hpXNa-{gb|Llz;U
zf1$EafQvkK@@>5}{_Ex{BdLiabgpZ@<S*<<0?-?3PIL6^^IzSlJo+|+Y>V$O%>56P
zF`+RgJh$Zd>0jM^)B|ubgf*&8<v%>A<_W-Cn5yAktj0eG5kNA5a)8r_b!nlW{+nU?
z5rCLvX|+p?+5dIpXa>+EV~Ge-=)W0C>;UY>A?DJf3Xt%a`dz^Xcb<N)Ce)_Kv$rpK
zsNV>-(8!1uXCGBHPn+MAf2=rqk5!ne)WxRQd-hSx?+%fmg=I-eFiHsw-v>>0?F-Mk
zw~a~(R(|G$d?&>UKgI>_#uHRzanY|`pgp$rvV?Tl3WuLmYYaLrxKdM!=qWdagbaF1
zSm}_;?r&udTRmg{JwS2*$M+o3!Gv^Q342E}8IF{LtC<S#)~Yq4XHs1kv3?z#UC~ca
z^(`z6!}72(jG_99gojp55DA~KVL{hBVI=5ghQVHIGg|th_uvW<qIyH@-O$A9HtL|=
znohbEE<;w|<9`Zp9G7`B3wWR!$Ml}BYr%1=Yrdf~wYUqxmg*5D8M~p(xCW8a<cC>@
zbk8Jt(Q@(<Cvd96DaH&hHoL0sOW%r+Szb<F_c-CVn+)3WKBECjK>}daANprh7YG-e
z<=Z9cJ5s3mVKV&;vl^zFiv+`<ZQx$+ZW(*;RU!_5ihbfbSWk;~7gk0FftvayW!mfg
z^QoS!cba`K#$d1LL_`q%M;XbksDtIj%RQOjBC&?c2`gOxK7fu)$X_Kw$XxL*9lg0S
z;y@pW@}PK?`odM!vG2!^C(|*57I{9|Qi=!_#g5<)HROkfimn?mLV9(p9G;Jz^YJFx
z9VG)QV!++YL#z%{kSkK+ss8w3;xRtP2d66GX|;rWNLI#CqEblt^gUpj=CM%Uh6KCI
zohAuhzq}Nng*9<|8#G{g+&T|CE|*$J46I7IeDWhPY<6ooVx{C25fN?pK#b{!$`|-5
z@CPPpTV$4fKEPeIa5OG|L-m~B;jOj4m|x{GF_-Bt`hD+3iq{?)m{He%A1<|LZ<`G*
z)kx*Qgzxgh!cvi|bg0-2)M?sdv;M>_1`*!Q&mCTuF`6r)k4VLp=rXpnFLzVo1x~h}
zd5R%lq2cVg_z}44EGY$miAhWHoQSQk+A3u0%ueOZ`sM?SqrScNF3zc?_*rhHYMbYD
zJwgO+=_`>hE{85HWU{lZ)A#h}`x~SbkNL}|f?}8EyQ5BKo0H=Y;st^7q!xK|lA6H{
z9_Rk1>MXnNF$S-FNkQ>}bHjxe=l79MuK-80EB#R<I4E!39Q!&urK>N9(a4(OMMO|n
zFzuFV({n%I)UVIZ^%=Kkv+|dD_qTca*XiXizkTYuKbyRAX2<hoMdUh<`BeOD-Lhh_
zdf{`K*K48#idNz8V9j4T>e4!=4o+F(1{_rPw@*)xJMVGTG*5aeHDu3tA3wh>Ml-Hx
zoIqY%>AWSr8+P6J&#+3suxe8uMWBa8LMUbF<bi-U-J%ipH}YpRKZ_nMbDw7)Dtd0z
z%TnohjOEGYfGOKBI4s>eIf>7fJXFa7>pS-5JET6+jSNc@ctxbsOB1#_eGwiOG75ql
z_&@aPNfoUlTYR%}7miSOjH4u_^`gB!=oi99gD{L@nnm9nM(OeS)PKLw!$mGtc7aoV
z{r4&V1cvqI>mv|bVrsv()NH7xlyuJE{j7;tn<YUu%YE3=ak}zS;3OhB5K}WKQ^Rc%
zKeN3Qfx6V&Wfl<?-6fgySh;?JcWDrrf!f{w_H8Bd2q5Ts(K%;9k%ye$bXCBE3w~ZX
z9r8*SW&D&Q#q8a+&WzX`2NY2W%4_oniD?z&2<Im1;0*Da()t{K3rCK2Y*BHbEYoYi
zz?Rw&KohA^WA5T2&GHly8|kE!c8;t@A9KTU5w&bVrE!tCiTt^tDGG2UcK%pR_0mHx
zr=>))td6c{=E;2C5*oF2HA&Gq))tv=9lpB`H5mqo{S&R~S*M5OTq2B6rrzKWZbztn
z!SOvhBFcB!g^#<rj>4KKeYw?%Nc~Sssos1~F@Q%?aoHOoaJpxVsH5z>t#H+5nn+9>
zFF6isUV$)PC+<Qoi@Df{$EmR#r^ydIX{ar$f353-`$n-Ddix)j8&6PnQ7V0prBqZR
zKR%GXZoPARe$-aG&@n0n#;m<xaVg~UvC-NH&zDHbc;wNM-eGN6;?WaA;D5H&Z_H`a
zi;(+g8n<QP+jRMqf*?%J52H7IYDrZ5n%lkDgNr2bb5*ZoiXO>@PxE_UA@)FoZQSd*
zB?~OnTd;l@TGybyXn`Xn6kX2LHF{^aXXMh)!?9#?3wK`=Q_)WGjBKU!=Ke^d^jp-v
zV^C>SbycCG;2$;c6sUm*pj!JsQz&tj4Jh`iA<OZ<j4O%S|NY@_n+#0C|6QJcBliD4
c6N;J3na2BNHgCTuUjctlWmKd~o|yRkA5+HDX#fBK

diff --git a/docs/use_cases/scf_to_cps/autoimpute_model_comparison.png b/docs/use_cases/scf_to_cps/autoimpute_model_comparison.png
deleted file mode 100644
index 93c6bd6dc248171cbb593d2d4a4dc55c2447c355..0000000000000000000000000000000000000000
GIT binary patch
literal 0
HcmV?d00001

literal 53000
zcmeFZWmJ^i`!=kIAT1$Xg3=9wNUDU22q@jsl0$<uBhsZ3N(u@HNDD(VQc5EY0}S0m
zcRhO$?%(~a|FhopuJ_CPeE5BExg2M%YhUNyXC23JUct)G<Oy)8aW7rEMDX~L%=1f^
zFi_x+<~1zviA02P8u;t7<8%23m-0Gj7cO05y!2S+zN(x4@+h|3i@tiSH4#Nug;pL$
z52E|h*G<Si`dfbyFkR(lBx6g!C~6&J^x(ZutmW?;^mK@3sMu+B_ON3jMd)*5#;akL
zvQ@9-)%{!5Q;FlEzfY@K=aRF=YP@nTi{L8zVvwa?y2>Vf8CSV3?W!t1ihNX(i}vpa
zfB)~xC8(Gf#=k$wmG*^?;!+38emVc;*%zT=hyVWP)&G9$!n*&p#Q&d`sN?WI?4htT
z%u#37A>UQKS?#_{oR)5^Tx3zBTAuvFz@)u&{5l;el;`_m!Vivw9a>pK<!i4?tStIB
zwg3E1OEZQ{PhWLY=OB>B_E6xUZhZh;1%2Rs18_7wAr$ZotvLQ)(cFUSZM5%9YUGs+
zn8cZI12*Rb4hGkY2bgXN(e}tWiyXv%80`K~lqn;+)yb}#H|FUZVj8-jfASJ;CL=3f
zZ*C}7tna;3MCZ1`5cIhI-pB2#Sf?WF$mDdz@!v&uacqeL??Rf2F5K1dB?dVIBJy;b
zD5GN>^VHWcH%W{(V1tUGE?_&XQ72l$O3ukbYrx_)@iXH))Fu#nfq&oMeXqfNSrUTl
zJ$1+jOTanY=k#$iX*bCzUKhFu25f3|lh8xHaZ@&RJ`s8ej{vTZh_3t>FZ6WUv?~a-
z1S@IYIPlv{^(9=wM|G1p#;(@Irg#Syzh^EoY4L5DQJAbmabG7T4eitMZd|YBnd<J*
z$a?<y);*sWL{~RvKg{XZOi;+k$$nO9X4x3O(##_R<>i>ng!)V#&5W?AaY-Gx+J=-}
zhHIAZYgBn2lRIa}$uF@OFlNe=S9#L;dP{EdWsMAImY5MGi~BIkA7MMa%1H3uY>lo3
zn<WcXIj-d$YdGqwEPTDQ%fM9`|E}ou=4H`prdwRJYMkC&iixySO2vCq`K@<fFEVY+
z?9^q9%A4p}YH$XkR;Z13idOD=p{O^!r~NXvCFPGl-Ehv+uthWv(_ydgHtMjcvxy9s
zWrcT2aR(pEcNjLkv$;lqC!ef4gq)sdY~_$cKp#Aa?rv9$ZSCkFSL68b({)t)*D1M<
z(d^{vk}Lhs<SUtL!hABN!Qa1AM@M%v(ArbH_@r(p>RZc$h#I)QJwN{S<I(Ds=8nnC
zE18g$orfQIk{P&bvfnPI#Kijalq!usQ!F&xwyoMBdNA}nfG^F!%Kg|Y@Y%!frT33{
zInor1KRMur7KnRpU_2q~fNqNsG2gvK8`DbBoNCK$!az1MWu9MPecTCyY)2d&Ql!M)
z&$L^->Dpmk;oV!R#eK9j!hYnAx;HGGAKc_D8Ruo%0UIk?)rC1=kL=Jjr_AO3HkK*(
z#43#`ot$<4T~WPZXS5j4w>VToteZ&ZeB9sHheMQ7)sQcL`c}R_tjOT?osBnE)Z=eB
zwy%8HPWziAzGV2yV((JqkELl16b`}+Z-I>r@9yzw=(J*ikBvQ}c30Bpsqd)EYj-yD
z=5(1{2px#Omqvq<bPEAzDSf#)pd0qgluF-Wt}UdT)5WOM&L>T8Q+w}=$<5`x8KT5m
zPduxmx2*}#-AwNwfuJJ%g&~qgycziL3fRN1IuK|3hBAYJ`=xED_)3R3{29`>E)B&L
zc3o+QZtnGbMLt$0ou3|K&DFE=7@k(JLt2qx-#vVq^Nv{yy-%*`4MiTi<BYvfM^4+;
z62*E=M0%L8tqj$Cru~{!ZbB16-V$Co?nRS79(3H)t_yWwppYBrCs>=cVQgyKQ;E&d
zZhUT`SxNFNxjLlXBf@A@*N&_$j#Me7wyEq&=4bW3A`8__l8$3X8ij$?`bfvOzwH>+
zt%rGAcf3~Z42{ztbo-Zo9Y9@SHQlp&zUz2&xOX*!XYkD!9kl{O;cYXXVQJ5LK1*pG
zlg@X<2|fpOtfw{p+G?%0DG`lAY0t*nrDry7n;24v2nh=dXQ}4hEs0QE^on?$oHBV7
zBc@rnPSKXqOEjZE(fvb-#Z?1;wFj3<H)%r56<YO@ElLDi21bZW7~@RYpLyexmzL}&
zJO%!wukwd)1qPK2O0IrH2pX>U`c=^Q+Z;xGe*N{&1{IF|6x%@@Op7dGUY@B7Pe&z>
zc1W;Z-G=$FCALv=11nXfxw}a3jhChyCXZev>)|UCRiT6Zwm8R=#xi*nm99x;(41sO
z_~OmhbQ^5`@P(!P>{c2DmB2h3>QpPMO-Y}qPYlni$S+EiRZOmXuuxymt!Y`&Q;jxf
zyDsl9v|dM(n@VB0D5FgMF>eQx=VEFs0s^=OZUl|H)m2?p_wTqTuLV6SK~055a4)qC
zkl><jNU^g#{5qI7?j6MQ(Cin@TXn^nsQmAUDh4Tr1H6&gi0-TP6J9!>UsZE#tQzIH
zOg&4V7%jUallqhk#;UrLm1WUuAu1F5@Qqku(p|@^<4W2g={n}~wiYExIyAy%JXGP4
z4kj_Yp>}*Sl;gGDMYKN%@FHba)W=&PTn~!&njT>8OXRS$@UfLX8&{NX=SWFS*O1t^
zX^;{@H70O4E^pml*5=n2`N{LZ*3INgm7hRUmD=iK5S${b)`=73((jK7j|s%H3N`DA
z-Q%^rTatb-{CaoyBay3!26W0Z<9n<R`P`o|&2ai&LIO)~76S@rauKf_GlH0z>pJ5q
zo63wzg!n5nU-3|ratc)}!s|9ozrJx6zJx!T?p*m?U4=RgP5AhAY3JoQwZZ<UYbbXI
zv-+yX$4-Y9txWzNu}@J&+$CNzr*D9Z-(MwmB*+L2rpem5lCh}YgSA_)ld(kabYtZ@
zHGTAKr&UA!1l~pFxusQS1og1a#TwuG6cp)T7L#6YM9SPNT4esSjlX)jmXyYx;;@PV
z(YS?6EYmn&adIU`vPb<lu{-f6uxU1*`g;SMYzF@jx!2N(1&X$owMJzeA8sr_3Dh%)
zCrZp>isrX%nBV&!t{f_Ly+jf@U`jxjD))CU`+dIt+4fl6Zd8@QhPMlAm$Y(L`s?%G
zak8UFSN!>$O+@s}Ukw>!+$%eHK8Ak{o1ZSrDHx-oHk4g(a#sG$BGJ{|I6DkJV<)0f
z{z)w4*8Y+va)crHBhk_*Zr>-OcRdN(!nt{3&0WC?rk51iV4;kt<!RsO;}J@y*EY!2
zo<e@@qt&LT%LT`CQyU2+_KwlYhp)azti<28vuF^u=_YsYxk+EqcXGTJF;qcA%|Daz
zDvrNM%KPLMP7qDb_-WOfRuR<HP^*aB&QF7WJx^#F0oXXDt=lUZz*ZkVIg1hA_sn6E
zTWs`J<BWYbK-yzYo<ACTg(joL?p}#bQh)u7;(VNeYLtz~X}2TSCjmyd%cG6VD@6b4
z4lucdVWBE7`FD4F{>VBRPDHC*-^_RAqT$S>x<!UC<s-m<$6Sqbbyr~k)xuZnF@p)6
zx!w5d%U~11LX;t7b!>-olx_F$r2=i7AtZ$#^LA*%cQU3t&RH%WHkeiYXW{v3pPrkG
zV_t*Sqe4{2$*NzC*kR@OXq<TLC2MEIoS17SBXPc8N@qX>KGo!!^9<cTI7o1-(|9Vl
zuMF|=QKLFWCdZd|rn;<I^HUmQ{Sb+_lKO2|txL@I8OhVaUL$+`rGXFW3$upu2F17W
z<KIUc5YGHb3&;!xC%9gigVkILt4WM9<kWbMRu9tet;6r3YbPoz|McNKkA;lH0iI+!
znEl*m{E-8AwN3&Sh|n9@(8@h%d^%4}H1SJ4#Fuv39#WlM`g)XsQj49_mIyK8(DfBZ
z?a<>ao9z*JnxwlX<wYI*j?1tR?)5X#*`dVoPeDh;lVFuS^3mqc?#I~<!*5TV@K>HO
zEt@&H@?S$4b>zE5<h!(l2wMkFHMoy#ZOe`>rljeNV42~ejK<0-ac=0yn2=GsticN;
z$A9g&8&Nk_#Z)@KP;ZblqA!hqn8$$FZp|YIsekPQHQUZ%86M2QbX9p9UKu$#u>Hb`
zx%Gg8#D`+@a9iG)6fdkPKJ?YC$@@C|uX9HmtQaUb!PZf1R(9%E1)KKNgw8S63|%Lh
zSj<1=p*N{<-z`jxOzsMg0}pffhkv6Sw@|P0n74xPBfH%#V<Y2<K1Ze|qJ^M}<W*bB
z+3$z|aMN3&Fe}bm7*$!V2plhBT56WenIK}<6==pVu}|U8$a&V_9}m;$-irNG>ULw0
zQz5QEAE?wGUwgW{a5EEk-p=Sov|ep;U4-e>G{>#b{-awDP2is8koi4P(Q0_H!z(=q
zMSAJ(iPu}vDUa6jE~CYOcr*K@G3GtORVD_m8$@Iwb-8qYiWb>n?93>m58A!=#jV-Z
zQ;2h7aueplX@>;vi{WznXH!y<_ZT)LS{A;7di*psxIH%*ihV1`Q~SB3jQep1vnclQ
zar*}(k6f@AQS!~qT%p6F>?8FDAGp&ia}r06Hz#TtTW-&`hhn=2Wb*s*^E>hjaLdI9
z?pOyqe8p?MVI7-0vI`G<@xyy^1drdJTka^Oc1!*v$<@NmI4M4_b@iZx;^(rP6U$O%
zd`2y}qe0#d{+07OMn6&q2Q{C@S3i?vxS8=Z;H_+`ZNkL$<1cFOsqgyi6E=nH&%B(F
z!*(rn2-D=BEiP)D;kfT(P!W8SH1-!Zv{B^B<ZO}k?+5dvg@si^WSm!m-BC0Y5&lWv
zhqGJ?=9QSqBhE1azf-b;U&tD(Q=B$~YV6wF1g<N=CoJj``xVWYphRGh#hv!=-0Hi#
zSvzy75tr47BP8A+H;Kb>yT6u_@3kMkvJrhzd?c4GzDym3)w@td%;i-Nuk^~tT`5dA
zPx7Q<WR4Dp6nskNg$Z@YBV%)VgI191jslMsI_)VoKQ=y$G-{oRmd$Hm%1{kbY->!e
zwJf8JzLB6gA+ww>^@HhiQD`cgx`F8H#do>y2OBtUAB;b>sNTw4wJ`9gwH-ciex=7?
z(D!;bb@p5Fz_dUtQ$eU=?gQr8e7un|KRBh3OLSr-x3K;D=q}Oml(gvha!s+*z<`9W
z>4P#l-p`9t6H7NgGZ>Prl&v+VOou!)B6S>YgnluqXp7YGc|UGw(K_Cp!!vv*pF6X>
z;fDc2_2ZjdnZ3xMpx$<Sik^suG7Jx*6ZTHd*sDn(_a>I!7G61WvhfI+k!iHNGfY#s
zdl>!FVB<a(g}D>my%!n`>#pkr%5!!#A3E8BcCL#kYJ1qfXNoN#YxN3gk9X<XxpLxj
z5j}F1>X@J&&#4+MGOt;!cCPm)%jXjFU}Wx9p5WdxItbaTC8CS#b+FhSh?oyC<!Y?|
zar5Q`#MSZfno}BrsW+rs{Wy@2l(Ul3{sz&<kULe4S*}m4;WuA896J~4P9B<)ChxeE
z3iPs?Dzxc7#U@Fgd;E%Yf;GAN#!T<H%+M8R!1qwp1mepFH6MdxJ6;(oV<RIHT^c>{
zRBAGk9r%vW?7>g<l6=_59!b?)pG4Si;{4D+$8wc(s42=DlZ+U_{k?n(TInLx`={mC
z#w|t0&#&Yki(T2x_y<DzO=<{HN;dG97ldVAvZPLV2bTz^@r}zo|3Y1nkMD2o5&5TI
z_m+M-s~&x^c$fK3(<&)O-<F^EWUN+ceLwSdjJ{fb<ORTmcLNzz86b=wDtq3*pIyo8
zGM4KO_Jw89IGCp2C1~XhA(GHX|Nb8M*1!Mx-)~)*^}m+*uO-kv^nbs1M0H^5A7V3c
ziDL{Eo4l#6)~gz>CBT=1JfbWzaR_{$Uj6u<Rq65D`FKjFS8C&5rlg#@G9=`QDaril
zSExU2SM@^#uw1PyZpTLdv*Yn9bsjAaw1=AvVawxK;ftyfuCt(9Yi_TM<OAiSDu;%6
zH)r6Rw^NJl*jbvT@wVd(<koVkPC@`ZQP|r1vdrf(L>NR92@t>AGkmhkt;oC}Gn3nA
zaQqp@xQ;Q5)X&!R=1SN}g8P9KhU5`8h8|9k&J`ZkvG0Yi)+22_1#h`4VeQ^}zkcbT
zIA%DC!2}Z<9+|2uEY_6jqyeB?ui7uW#%FuA1n+!$h*1A;@hsoI;@Ut;dm>yMSve66
zT2c4QE-s}2$|I{;>VI%jHfUsIRQS3z|L)0vsrc&yT!41T%-Jn2v@oC4c+cHj7KPy2
zU3oY?Ze!CI6c#ocRO#$`KQ{Rv`zhB^UM5C_lKlz?fSn8|eYz?i8oO({$=srLrVVEw
z)8m6IKhT5t930)Lp}R})BRf_${ay?Nl`^7P;mzQkD2Pv`n^<7f<BjlIBOBj6!4ekc
zAmfhL!P*_gewOCl9%AtyeB@UC_R4_s^!9AtV6a5(aj;9f)x`UE0R>ZWQjn7k?%Z{n
z2mo?U*P3&Ug{WK^ZrR`a7k0uAqv`2F1paH`6<15(66QN?@jslzytPEATuLp2*m)f_
zP(Mq*8Y72{3V2|l*oq>nq9LG*fAGCDI)<X>m8JSVCSCR7%l)?oYn|a9>lrPyVqQ3v
z>R|)n1(6BmSQF9FXq1bIIejf)K-qJ=6|bE}wP*D1Ke&YUcJdpefDOZ+T@f3r)Ztjp
zBW~#q`y}|N{?V(am0OdaaVX52O*B-*>ArVMyV%-Zbk)jr)L^~U&hL4Z%i1ANY5M5;
zBo8$sihQ0IG2Zjpyqzsf^92tD8tDl%0IQegn^$4f5*A;<^k6&C|FBi$ay)-UAd1rd
z3gqC(nQLu@m%;nSgh5yyMMsK4StKkrCzSN<wrzlAJN8RRWUR$dK?Q2{AuCTnI#dE$
zdB>%yq2*|~k>=@^P9%*h{mt|QJS-<0!p<jONyV4@tDCPUebo=gogcn;@<&#sn}7b~
zm<tVvfhJCA`OlHq1+YU_@aVD)gu>l?qbmFwJ|Wzfr@gzV8PnQ8m|V2TRNxxwAq!WN
z)hT08YJRP0=O<zSVeK29w}##e!+KjAJN2j-(kSlGN>quFs=l=(bojT5Pjq>M6U~b7
zZ##*f%xRPe?jrNs#mY=U9wzMoQlpt_3CE@sc!^WWO~=jfp=+FwUMg^?wUk*SB6njc
z?u~zCeC>~^xv4DG8ce8%z(_#gi;FtP1>*pI{4+*uUeJN(Q-6*>E3ba*0*}#?kdvUp
zknmc=DOCEN^D)_|D%J6y(Arb41#)sls#Fir&ohkM&`Qleh4z`<SVVPAO_p65`Mitj
zr(~nU!qUZg#wtH1n;q_jyO8F4>E>R7dJH?Kd?+u9?`J~w?&N;*&HXq&`g<E7J^6CU
zYq`2j$TWb#$i_aUD%eH75+zzXMtUFbJbap1{H&v_;`P#Hs7H$sRp*oQ04~jOZ64c|
zIJS1-4zW^lZETPJqAD)~FKisPc|8fcGwOvp0R7Y=qQ+}ZF{wuNQ~*qL<hjQi6utIc
zL9R9MNuz3PgMVdKMrTPM%z2gAGtd<cvJRH6#(Q=-w;C}_cYGuW9Y8gfdNuQxm@9nX
z)YHr}EH)XHluZ1D>vgM-cYVY8(kB-x=s5th6q*kWA^UYTU+xYGSL8h7Dz@wT8|Q(W
zf-{k12#-DdIi7Jdh|+fd^i_5=i3yC)7v~Gw<{65W{7)g#Yj;KnjaDN`P3OIR8uWB`
zdxxw5Z4imSY~<+T@L_YHHI@T&T_FeB#Pe>mztP2^j3I-2n2ABU@ANXuF|li*fuV91
zLP?yAQ<#5y_E3L3)w1#V#`q*vBLlhZ+jy#*!YXF`*alBWCW!bx=7?ziaz}}~`6t&O
zLm)jw6(sEj)ii`VF7p8UNQWN!_jyeJku%y)tHwSxfs{q<nDSpl`bS=V$SA(@lh;@p
z<{u*gc7R-2u{O<8Y9S}4y1MRTo7aIuPfkwH?h~;6&_}7|N3Oep?CGWVKGJ;DE|dXM
zE*;&@T;t_5QBslEpe3<Uh19xj6Usf)3%)Id?+>k(@D_6+kINg6afMFiF#abh0Bb!W
zxGan9x-uX>5*Km*StQq=7vj}N+Zuf;1I1M;a6NJBo3q;0wWbcPjWu^_zIDr0VO_Z6
z!c#_aqgx{oA)V`kjKWEe3n~sULI_c5X+~ooG##rA=FG0A&s~w@tvhjkfm~kp)jL$0
zY~|(nWpMPz7N^~Q0DDT6s>bgh3p~Eueo7=HS!os+C~r-`GZl4}?GGbvhix~zyL<A8
z|5Y-8W9jCs7x%X`o-}%LuGV~??)fu}twHgOo{=HD%XV_kqtDXWzx95I*ACsi@ZwX%
zWc@dC#5XOWzgf}u4qg!QWY1qB@B1BmREMaag9H;5;(NLMdo7nnY}fCRZX82Hzweh-
zgt_Up`BA*khh20m(w7y84IyP8VhgGqz9b-R-M6@iFN{DI#9N2#&|klau7L>tlYj4H
z04lwgl3bk(%N5V+I{8KF@Y~`xd<@r1FM_9=3wK#e=0n~qbAGe&)g4MF(IjH^D*IM$
z*@?8}d7S+B7cEP7NiMViXj8@CuY2e30fA9V*%md2QZ!xNoop1^%XoaZmqR6V+U7p|
zGoUsE+$8zk%}Y+w!SmG;XkEEj7TGPsS8%U&4i9lJOI-->aUs_D>yPw^M!U6K#1SXq
zsXhHJz2tjN4E1lyB+!Z<(z9VyCY(yiWhy7{ss|g!IpcW!;1V^lK0X^qceHon4ZCPK
zV!5$u*S*NPA5tQ2bcf|)UfAUc@uN<-*Uaqkpd*qDVD}LZ_8o3A3E;L0AB8DstQytW
z`P>y9Zbj;Y_~CYEr{OrTbE3=>8!@WO|4Nfp4<c1+9GfSWmFlnEHlb42gv|4mGr4hb
zR+J;8UlX7}rpQ{~Hsngacir>}S??VSr>FSZyE`%ID0FV4K#U0Gu<wdOmbqf{MgnaV
zn&<rwe~)^%EL?wm=OskLq`2t$MRwH8L0=ExMIkQ7nUlLK#nbs?OLjfvHhzG=nx9we
zWQ57f$^G_`my<Ova-{%O;K~-4$i<_Q70CH*Q6rR+s^32LUMjE)x2rPKDVb%(#0;pb
zw6mh>?L2nGqaPADORXw*c|EHu8Rbi#Rv?oBO*<+&Ec87XPzwc3dDrvXXzev*1dB*6
z7RDyaSdF!t;L7m1$F;}aX$gODWQ0m66U0_D5mQl`#|y0-Zo+bwkZOz^A^e3LTN;sb
z@L8$<Mne(SOSg)e^$)Yn3z?_KM}=6(srz4)FJG^H89>sLP)AhQyluT<*zkmJm|KVo
zLtlc`&@H7O<cwj%?6xFqrL`4bh649!qK)a6rf&jI+(`@QFIEN$3d}!N5DJ@Or(?<G
zlkHm>k*Hv!tBzAKK9>Nl6*#g>Z`_x=I<v?(tY%t5l-Z$Kb%R&GHiom?LLRXX$-cAY
z^shQvvt0T$A<F3L$2%k5^*dh86Aah6bp8}Se}%uW5MZr^n5i8nLfMUBeO8fYXrmv{
ziKqy5Mu5Y8q*yn3k#D6hTU`fdPOsJr1LGl5%<jt1pG<sSPDFR_4eC7H{AxiZ+AJaK
z?(T5m3~W(CmlbePFQbKFkRftn;-)@$<Bcrkd$=4y+P8f00He!mCpG;mS&L`aB1c%4
zMPOf}AWx7Frla9eSaM+oHeKM3m8t#GF}Lg;16)>Q8@E2j$7|lLD*N;F9q!ktBr3#>
zKT)~=#7uT(M*k8l;q_-c7}@0y7b{co#cK>Bt*mSa@)i<F*#?+N&;jp-N^;-iRZxo5
z1ZzrPtoP4z#<dt~*PRjOd%qylqT_Mhzx4V3EXx{0M8v)sJ^O{si_b8Ty>O97IDWnq
zr7&18A@Rvuc=gkK3pcWt5G5SPWWkr{P=$YSS++F>5UQ5xW0~rZw(Q(+hkt<AXmki0
z;#w$D#z!p_4Qai)_(|9ew4Kw7?r+umTvp7r*R>QIoNf8o;A{~lc4rrZ5;25#v5eF@
zpKD*{ko0uagzb~;%2GAPstm#GIyZz@4yC_8Ik|`W8?gMJ#FWgi)&FEBCM9K#R2v5V
zUP5l_b#WCw;GaaOcu|Z+A3JJO>SHbLK_?nZ2grbh5LzKXdom@ClJLYbXO&m|SLC8%
z=^P2daCb~q?D>Rc`g&{Yc2lQ*Ag2hbbwIGZ(r5=Z#-`0zrea$5Hi;7P{d0__+{FX_
zdFl(7*!DkcX+JESt7Au(%K;lR4*<*_YqLicmp9hvIer3hDDj~{Y)(rjn7xI2A=>Uh
z?Ck;5uOn?^3?11`VN}YqhBy@{71tP}@C0EAIv?iJ&#^0M6=2zKq=@u1W96wOxTd%>
zW6KA&tfhyOaW8Z(0@xlDUL9A)K;4@il(C)7mVLUQ@fNOEsMs2Ek(LYzFJtZY;`}>q
z8Jgz^TU5LXEv;h)3XrYrT{yDFd6l0_tBmMMh0w!l>J$#kFM6SHq-}(&Q0M*glF!!&
zoV6FMu#|fq1n-af0i!qW=Jl%kQ_A)0WEddNN&({d{$XSyd?UP;5wy=<_k{1)566)8
z@q*Lcmzaxx5ea9>0cMwk^xi}<I~~~Gi?B_)?BO8>HQV}FN_}<1E!8Rk>y!#yi=V&n
zuID61Q^%2tWTN&rU3*8@U?DLEsK=-7&7T*pr3nS4?IPxGR2QVzMvxYWRZ@+Hj4JYo
zsmU+rCzJB!DslE^>9TYq^CY@lQCQPkzruioN9Uo4V0IWjN{-*2^`Gs`IlGc;cWrFm
zc;IMla>G`)gi5ZVZPf-*6sY=0BA4bOeZOmSLp@N5S_HAsk{@bQe|~tdX$hqyGMrBK
z>1<JnYIL~n=IYAv>&3zK_!WPLT)0UT33@Xr5RJG3X!Ky^Frr?HQB!!1BE`r5ri2cY
z0*lDL9Wi2;{-`JQ+uXC|P@q<5xy{%(ylQlKkYhRDzmwxeRFNUW@c6r<04y_BDF5Nb
zwy5Fq!7}wB1FMXFAg$XwywSEXcB&=rv0+M2n~CIQw>O4p808c_yRcZiBBKQM?%>)r
znG`VsHO(NEWa_f8f*j4O92Q~2DPp%3!Kr?wk4QLA1Aa~{QOVFs3rhmk$CmbQcW&9=
zeq|1D1bch)e=L`PcD#k})!gMZ_^4RKJuaGyy<W#O65_~GeVbkf+E7cimZzPD5OdZN
zG7wx8Ug5+-FK!}p)kvsrlsUr9qi4-ecJ~24n=UJfF#Gi}Dg>zwIV$Q2+I*Ppt^e55
z^D_Uiw@v1i<YYd4lqhFos!Z#(TQ4t@q~igsL+FO<QxLy)$*^V@{;Bs<!1mrs98GJ-
zZV}Xq0qCvT9SS^!zgB!Sj=nw(;tmImE+L+DZuDq}x+(M>A<tvoz@}gC9Yy$H#iSSa
zHoWuq%g4q>T?-5Ly5Sol*^#}jze7cbIO1pSPoyH1VHI~9J+2RB`Lj8(Vz~wgbLE`F
z$;xTyLx#PZ5lJ-peA+6#;)UaYzb0zE{5M=Lm$1ZPGym>Uvpc$D>39}_)<R)9Tluw7
z2~w(Iy>P7n)w(3R)+1Lp(B7?s1Gm2_US9MW%~lU*Twb>B%g|>UYUYQrCPcAb5VDLv
zFewo+bO;UGu}BH*XD^4dP;4J~DVUhj3#{uqVNTmTvxhqX*76E*`ca2ecB0(7;_{u+
z?mB{%Xw@Ht#mW9^c}j6_oC9Y|f4~c7JfaE{WrtSsKlu;L5qjMaqG1b(42ifOo{wz|
zWF2fT(o*wkWt$sPo$U-n7xh<)Bl!VDvvpF+A41W=ygTGNFGMf?Izo~z0l5$=5?|<s
z$D>~jigbCWjqFUkVSq4vMl=T$&SZ|S4@+I!uFdkLU!iGn+}hcB&wNxaUX*Im?HSKk
z!rTy4c?~r=ZgdC)a#_}@l>LV^Zvy1cq*)+xY<PXpOU`^=KIu>k)J-p!#*7_GBt81W
zHtwYOttfwf+HC?rE5v)nU;k`vls4GfmjA8Dn^ctlSZaYMUSuXCpa0sT&x1s6$1YUq
zNe$j^!Y*wisDJp2d}!VP^ltfU0-Hg_HCu$&n-j@Rd$(36P*kLY=$u^b3H>ZlJ>46J
zQ#=6AiDs=iO2g<dsDT2F(#swuhAbOZhmZ(w$UwxeGCLJ)ZW{Htn`2UZ*RE8S){TPe
zyUbEs>kHwcE1Rj*S7;fOK)Jc}?#tq?Ja-$RZ@cAMV@&ucvzOD4FH#|B*RWh?uy$XV
z=f8Y6es5W4iYA|vTYHlf!~X$tv~1f4%XP(1MWx$~^!ghfs5vrFWpv%X@JIj_$E!yv
zF<eb!O5t}3AQuOrquEjUJzE45)G8F2+}(Zy;Ce8JDao4bZK}7sudeFo+yS7_n*~V;
z<q|_%_dK-bE>*#?!P1D|@wf2khXi}UfmDcJWLJ%>O>vdPA4Ld-t07NE1WS5jljU_<
zYAg{A^Y@+bD;Hk+C>c1W1GMs5XHrc<1W>}%jg57rrtdw~2}kE>tEGc=|3iSc&SaHQ
z+SdZPUcdDr!KrR7>y^!qcp?QBEz+X*c>O&R%<ib9J5af!3?hyKNX_#M0e#MZk1~3r
z_cVZt-}yF+_c%_9k2VIc0d^`-NH}PRYp+t&zAm+6W3aVByejbwy7Ru0ll)|lwEuwM
zA?SC1FYTR*X|KbNhrn^2FQXShi9e7Np}?OXrcI7{!WzW7T9NVnK{fRaljiFQ@fzDx
zu~!CRhgkp%D^mk4vKF;3>*otj%I$1#T={=#1CR}Xvrx3g^Pa4QCT7*^XLW0m-zT)n
zOTc0)LS0h!b6lDE*i^Ax`Jaqcd8-B0(dBff-am|v7tP4zfw^vFhXQ@!D<{Z6hyA!0
z@|N(jc$l+e^^CY{Cpomxde{}HBz>wn#GK(QRj*0e^4SH<YY_BJu4^(osz8-`AHi(J
zvV|I?)UXw=i^5STuvxI!?+ZL}S=$rckGh{ntEE!zi34bI=9S5u55s*ezp?QNQ|=Pk
z54f5v;oq%t6|pE)p9ucv)G7PI^~&LTLVJI<4+{<l-p5N;u!h`d3)eWgtXX8sU_0&+
zrdC`N$BY<zAIB`e(>6a|{xMN<bH(>tmiD@F*!%!BKq`%oq6h^NLy4UMQh_grkRn5*
zNcx25dpxb{wDNMMQt^qXr2MsV7eY<l_dQHl(aLDP;VZ=1Mdp;JJ?i#*n&j%OIOgRv
zPp00`gsag~#b)**OV$Jcyzt{%`-^*x&S3kxJ-RO*gxs&qe_WD+`)}YY?y=b9$w{W1
zh34JG68eZyPF6;pSl78}x$|@@y?Bie#wi91U1rIsupW1%J-G}odJ(L$bW@_*seEZ8
z{E!k=iBBBe4Jjy-jlumo9Vrw{?8<Lrh6+jAnE3O&aRV;*^}(FFL5rGY1Hkl<HK)I*
zj_hZP5;>SlU)HpuQELbKD$R(EV>Q!fDYZOv@YQ2`66-$8u%vzF1}gyW+k}rpf}H0M
zBR&NJH0ZAAJ*17{H4e|`m@nj=1Y2ojy5V|a*b5DLq7??CpTzB$FK{tT-71$=AFN%e
zlZr4Zg;MI=%~!Fp$p6F)Rj_+QpfkFHiTSq9bqwoGfUM!cJiUWy{O&1_x0aeua`|4R
z=Ai^o#8EwU<59__XEsx%6w3ihW|C-??q6%Q25vXTv4Yse4Sdw}$kn)Lv?}{{9YPf0
zo}#2H<H0Mb^GNSGIO!HGDiE6B-6b02uAIhSQFhD$k{@E<nO>Z3-1$^{*YKfH-GAWR
zSv12tilir_fsWpfBqpPIrOIrkCieNOj%`*dzvXFz3yC0I!=v(eYQdq(P}+skX`7?v
z9|8?Nht7I~gIUu-CqU;?=-7C7|Fwe%I0nQeG=H(MXg5ICCvD8FM_F02lz3S*=!sL_
z%otMTFwiXvQ+8N)7r!WC9{*hekG4G2m8xa1UH7w|hXlVD3kysCyb8p)^&`n)a5)O7
zUhDULY?;$5@t#ol?D}Sl%8kgn-=IS@iZd5>%<g<Oc4}P<suW(|ZVXfpcnri{`8sby
zv2PZcHUUY_e8%V(8R%x>MV02MiEb{yv_$&MHMV;&JhOEL9Y7vzEg|s1afSPK;}!Uf
zlHC;(GaNX1zcEm>Ra@aE1q-bdFLbLJlq0af9y$F!@XFYj)b3f#j1Z^7xgK3P1I%ea
zD+N4C?SDM?g!s@<hM}>tOz%e0rZFH>e73z&(YOh~GT)-FxC<o>0y36Aw9B&FwnvD%
zO@D>CeKLxua(nXeyuM?UzzcdshNJkk;J$T80Ur#I@Q<2{-o59LgjtQ;=UQA{o;Kub
zmu>SK;F&PKz0zp({4eUw^#s|d=OKRA&`Oy_OoIToH~dYdeArNH3$g=gyYtLcmi4I!
zd^BYEjqKOIK~polZc}FmR??s@4$m5)6|h=+T*e0TkV&Wppan>MtE%cw53pQ0ExE{J
z{+5)6uRtbm;(j0h!G&~JI+AyAYFNA8P9V_{W9Ic2TgX6`_1aD_!F4M{ivY(RWHK{H
z@KV<;Syr#rR%CL`NXn0M1;H(aGpCSXdKw-oHO##A(H<`0$WHL&2Y#ToX|gloeV-_)
z=pJB0g8D4hUM(=K?}UYfH06j~u292yQAGP#R0K#cN?^PfM+pmEUl5_pDTbJ*3&FLl
zv<O;CL?mD$OZU`w^66Te7QCVIM~6Rr7^z*UxhW_=@|8^a)ZjJdf0T6FqHV;T+3!4#
zwcEQIt)epDQ~0>*_*qhQh-j;;;U*J9=<8-dn2xGr0pIx{S|S0(HfRhAXRo1n^{X2X
zb53l=9La!CI!(RLAzmoA$EW78%)BHhmB5Lya)3mJ_c7iT!Us4q`Sx3YJ!j1tzB~`^
z{kH*tiL^x_^*=vV85|lmavN8#cu}?_a`kq|w5*%V2Yl3|tBbd00Ts{@V~C0A^kzRN
z_Y%9)7?ZHe?{-A;=4}9ToIH7n<?}20;g-Kl;Tkq4H{$X}Wq}~R!1}<@dT3J&CG|SO
zx6Ux}c#a$>(nE=lM^Ygtx(>AjVnoJ^zic;Zcj?8sTbXRHHH?o4=iqd<m`+<wA{;fA
z8>$04r!520#3!32YmX};hQ`~#p<tcCIWH>~<bB~u@KFTZ3jkM?5?*(_wIzt3MeM-b
zccYJ76U=_B?KUP<5brXx@}@lv#s~@udyo?CI79Y{M0VrlR9-#o5Xt^_Ys0@fi95N*
z4O;5OYP{?^p7xVfc71P%^8>reb))Py00WrbrTFYMlX2_@P=r~<#Xvy{TSIuHM*)~G
z*r>&h&||Z^R7I`UJRWEJay!<b(5lfbD0=_3O*KEPc_2&gJFT}SjH-~okW%E!!w^r9
zXELvGDxEVfkI?uFK5CW_waKzm-+MJ-4WRGSy_Ec)lI|3X>FUzj&pyzF4CrgL`J_}^
zV?QDL4F%Mt@zU71{U*wIRP`fW?UA43DI!9oQ4Vu=goxgFG01j971S{@K*nnicnBP<
z-ErZQOBCkvqQeHYqmOcvCCF^^c(TU5F<rQ5LV+0Dc-Qa!-pm>mz*6~SB`*Oq;7ttc
zxrcrs0#yyx?gzlBybhsMNtqnIN>D==u)SHmK38s40pd4ph<lF96hf-ZB+F-KgFss-
z>n$PoQWGAkIM((dAT)vust3kpAVheP(jb*jw;J85nMz&2!E#+sT|C%lb_x_(=EHSt
z+Pxo*!SH!0>Z8N<$Lt#lMx#3qEqZcUsQbPpGBJQT6ZW4i@6W@`pFni5^LT=}8)nP1
zS2TJJD3t@IGEd>l_SG{Mr7jI#D?3n_+jbtUtgBN&;O>VRHmzF*{&2nJ!?oWEfU9Q#
zZMA{gY^{H#%W&8ZCLrnt(tV7NJ07B`?QB)H;K<M$e450~i7A+1YkH^4&z)rfAVW!T
ziQE2_ZoW<c^SKM)<-ZEY83O-GfXn{+C?R;@V(AYMpw#+0hAsfe3IMThwf!z}Zlb6v
z);yjD>X+!+(wuv`^jP8=(Y4r|#-Kp8c4D-uHSA9#-cNvf0+bhwW8|{$u-74zjY`D%
z{{|W9{j-z8egn~16alzSI9#y@2?FtTL}MhYm((jN+GU%G>3QGImK*XB)*nqE6pmva
zgelCKaL*nX*1gw#?|aN+a}Tvp48&C7qnypgJt*d&un@Ub!f?|Y==lr6E1AE)2~6G`
zl(q8y8S?QV{)vu`M&Z~|ShvKw3gA28TK0}nw_jj^!4v<zQp3iun?O~@e~1k#`=Q6~
z5B{PC(g%PnF^7nlLzG&?SXo%efwBzDPvFP2tYN1KQ4^pWmBmR+ogdoC>Q&^FWE>0*
zeKoNkmTs5iYZ{!SEbY#O2pc&W=o-&A+|E*8Hx{x}<}S~SbQS*}H_>nbt!V-Ly+Uaa
zj!2%;NI6D%tL*)&;r7^?`}$JHJFGiLF1W_ERzzj^JTUvin@#Q7(BETv>yLl?o8{B}
zpDq{2+{Z`T{w&uuGp>t=7R-LFfpULus?Bsi%6eh)sxpA8%Jn^;ScAj(xyA9~Fb0F?
z-++4DTWC>NQN(i1v0x#8cb=<0p6yudkYJ1!2bQa8DW}yrT=3||nOCVUSX2d(0ei`x
zxAkm~7}ZIAB#Sg5%wK2e1VQ5EoL@0sS9=L5+hZ}hd>;0iaNb<awm;PHkg2+ZzbeWp
z&pgl^?Pk{~KK1O=Fa5K`x$sCl7>zUj2MF_>!Fz)>j68ad{YDZnV2va4Q|U#-1&{rv
zJJ1PWe-(e`Llj(3)aLi6GeAW89)%$!(s#J&D2Wjnz|F(+<Kdgy2?CXm`rT`6KasZH
z_Q~rH!5x-q&a+Yg2y9@zUpR}Y831x%K5+3&GH9pKyQ%lF4}8jS@u|+)r$Z8nKc<5I
znH`nBd1f)(g_c|_dBhO*x!<(_WYHX=;X-7-a(A*r@|@{a4mzuO9#jbAz={8nAo&&n
zIFkwyGg_}D76B_JI!AZN0KnaW%!|rvvxGDxd7ZQLsn+Oya5de9^hYqGmaqKJN1?U<
zzii1-6&QuoYDak4s_5d3><u~~$=)MZvgF60wJ(d$FuZ?eH0j6SYe@Jj`KoejV`Xl6
zB7zVR+(7NG{2tC!z&!0^-`pOAf;!?qwSY7`5DkrsPWv9#W)Wmh>#IUHM(bZ%0od({
zPP2t7thDE{k`m?E^_g5UC<!+eLXm%lzn>Y<1`;-n@6MdT_XZ9hOddYUJiN@^grvtY
z7ul@*RQ=iQCmDzdO2q2PAI|*ZPvzOZeOtY}uGcn&w*xEf??;lIQL>Z>MQ632_V}nL
z7P9?*XW!;}c9!FNpZ!MKW+!e$0@Tx+rugmIPDP>1E__smix`&ci<u8Tw7?Tb;_sXd
zCi@AY4<SV4#|S?xfg|&1iV|@dpgHx8Ln`~<+o{hMES(J&4AA-EA#mfE7bo@$_h3>U
z>sX}6{Zmaz^1tS1{V(e!)E7Trz|=Vgu*urTQrvAzRq28L?U->v&4*_p?@{dE$Ds>?
zkE)8!FvR5nm}8P&OYpO6a}&T+F28-M>2IST&zzo+bH9~A+?%jC&PM*c2M1_QE$`n8
z{XRS7B0osa*Zao?5*ZQfXC4Y4EVjfbGINRJ9vwHAtfF@RS?oXsjRhP(4|?9R8e6Hp
zqazlsx4(+l9|deC{nrB&S7d-`TG&rvPBaLE8bawF9%G2n0Bx2k$Izemb2c*380E;m
zEIEJfTDKK_^g>wwF2w9_1pmSOpM}(JB5__5?$*N<rRKHWPZmc(7yNLy)PNtxS-bjo
zGoFJs6Vzh8giU|<NudlrD#b`iAQtwML2RzBX?8TIauq@FE2a6$%BKM{u#4m71n7YF
zCF|UNVdwU<g;s+3r9Lxgf!6{h0QQ-5DgfA4bCP{;@pUW+1OHDyv=~OyA06u9-bo1)
zlYL@9iGFz}KLNws9D8}#hs3~J<ucHWIaT{QB4SS88lEj69`pj}mDe<!uY3T0Y-7mR
z)TtfXGnZdK^qcs+gX)rmRX{c_ec)DmG=K-l1neK$?P{hUPHX;N=hj(B8|{7epX=z2
zQ~_2eDT&r8r!{Dxk#_GX>g`fFXRTPl-veyOa~6MI9FjmQzd5u>MSxwjX*)lh)Mo)T
zVR+^5;l#TaqIjVN03Om}XMDEXm#$v25b~8XCGC6=pjoA((^f@;?gYO{g#0yUGcxZ;
z*zD{<m64}DbcfruE^d$p!cX&e328hCHu{1d%vQbT{q#8W%m{hsi~qNqT|)MGfl|~m
zbTA0E?yO^;`gE7g<Ts6}3o|-+z7V_dKNk<SRFZ?{GEK}r>s5(b(?|sc>&wx?r)8oh
z{mA#W;9S<MpS#?@p9CGa->>srb9)&Ys<3i>!p?2#MTUluc!&(kR2vRM3)0r*d1E5T
zQvXmebv`(dY<ziyqE{dvaP#J=Hf|7+Y*mu~3ctCY|J~XMbnu{!6f6U|x?I*>^2~98
zp6u`XS<vej??8ZHGD|+Qe-=OMn9kCO;z~T)iPrBr;(_b`!{UXalQxJg#sdGnDe#>C
z59Vd_(*&9lrF6gWC&s<&k4J;idF-2)T&2{-W{}@Nq#GN7!Tp5F2yuq98dx)4I!rFW
zM`@T@K-pATpzNsdmkJXuN4!)Ox>QOy5H^H}s6{k?)M0G*Km1m#VL#OYsK^X@APML)
zpVmdwFUR-p$d1hZXwy@=I3kZ2?=@lh@H$9$Nz^0t3m)U+x`p_@I{7JZmKk!;$8^em
zmi$pMvc|KX_g5nO6?BX9b!m43!2I)wMTHX3EC-f74NQktmoK$!o)NUZ=xQjizQ5^c
zoZ~-bSJfu_m?Skrj43!ehhV{H=QYX<?$y|25XJ0a)|;nJ7=ACJX@d$HEmekHL6iU{
ziGRqW5{KIKG%nVAR!)uf=CNgH>5VM-X0qa^ME}xa$GWX8C!JK6PMV>w;6V%ZaJcGT
zRgU?G3wG+Y&SJ(p65&}cah--H^c!>K{q*isAf?DnS_%ggRD4c^IMKOz|3+dxFl~WW
zB4li=-{qtL56<)8AXrcX%%ay?40VxyMCh$ADqa#^5#^XGcl=9h&j7L#ok%Ou#f9RE
z*ps_F?fxEsGPhRc7#tdU(ma=cw7%$kDda{+eT?acu42uzsf*K_XXwKs4Z4g05jg$L
z6ne!D)4aOn>z5vSxKDstj_$luRX{dX$XtF(*R6<CntSDJwVvQTB<LTBV4uZ1FL|8i
zuUZg;y(xOKUkO*hP%B!G6Wzm%2U;K5V;5#LRtVmO@Z6tw>2Cpv1Gb1&3TAcPr@f7@
zJUdK6BZhyoGJ+`c6V-=IK`rzdY8sU{G`9++j%h{H3489}c?BxUThvUO@GT?7%VF7J
zzrVMfaU7a3-#`9`oAZyVB7y~!%K`|5nn%}Hc~>*uSc>eXQGZxXu0qo1Tz0{dYD%A1
zhC0TDZuq_q54?f3`^LvX;WkQ9=xwW+LqifXiu`0fHmy;i80md@LJBmi4*?y~o2^Zb
zrrk|U$Z8gmP(98jcnmX&eRxe=6qv%VgiKi*a?0Gs;FZiEpG$A~>X3`l`728@N+}qb
z_p=~wu0VY4(|y4XNbXybgpk{uu+T)>oafDWZIjMAG{@96m0N#7IPwE2m!&8zs8WS2
ze6$|cr<9f=1)cq)#mMnogCYD0L0|KXVH|7ZeL8LlSm;RDVdOF~`O7`R;$pSW%KB5@
zx=$0Q)~|Z#>R1iBp5bu-=$J9r&^8cj2a-^jrtIukZKrac)60EnOG9Ng9^L9IKm*(}
z958SO%K4F@A47`zfSNn`$nrXZ^R7vK;u67kprkgWt#lW)w$&Ayzuc|12+w<FJeY+q
zU-zB_%mF7S5G`Nd)KBgtwx>aH%Kc>NlH!4h-DP_1%ex#(uA~l$?8{13X*59Sk>>c(
zcKg`9$eA2Zyk(jlmNmsIzihvkHH>U)46*F`s(80o?r4o1>VfOMr@8;FM+gQ61e)Me
za{y^!26DA2F!-uc{ij*^_SUl^2Mh9)v-)%r^EqhoqbJ_@4=){B>S~{5$r7_A{0|HC
z0bK3`ndP^Be19$sYgUS53M2Gz7nYSn$U+9VwgL`*;?yI@3g~(jN5G!Q%M2N=M-sa2
zbrn22uC0hb=sahEjk{Y!rVwQje|0U6;%DySSbceye#ln2c~E!%WlsicFQE%oegX-C
zh%T4`QCVsVQCjo}J*2_?q6P8U6uwq1zjsq9ZyB(*m;4|8gdhjC(BR#b3So1xzd_3X
zDp<QzWW>)8B5~_?SsJ9fSicD9e{VB%hD$8&C}!ExJPQtPU-paMUoI4%ueeXw&acMu
z6?&M~w+=n@ealXPup2F_Z~5U8)LSXi`sv2JfhCcyaJw*VJFg~2Hv9!p1s)9<uc(33
ze9~j*GYO!yHPnc<IpuXqf2{F{hI0gCrIbtR+@%?1p&qKYw8cf=aH`n%@W}8JzN%Vt
ziAhkq>dJ4*bZ|s9$U|#R$=9mtSEk{`MN|Vvo=Ev6#uzE5nPT>?wuiig%r!^7Lx(+H
zy;>r`wOfp*mHL_IjJ-!sWK3`TO1ov=AAl~BD{chc$jPC67Ap5N7k7?)h<QGF;zO{%
zqLFf44-9)x67wJtKAIKyLN?*(5J7nJ=35)&H0JaeKVd1w3o~@dG43IFt1BAF%0|6)
zS+v)M@K8QS&QyZ2){RGxkE6uqD!(#BCRYS4gq9k1W(O<R`m`omsKNfmF3nHAnx&Aw
zk;6eU&*oGM#YEiH$&Baf6o?suM;M(ikLeFMWa*zN{w8)^zdjROyHe;FIib6}F=M>m
zp3aYtdM(jU>xc-Y=t*@<cr;jb3dxouT`dRIjKblq;=~QG{gJ`Jja6!u1%PRv;mwef
zk*kM~k(1W0hHZSs(}}o#P8+ZQO|0Z*yYt>bir`&&rGe)bHQ$T+uI|o@oOC5h;i$z=
z94s_*Xxwitt4;7yqYA}Eg^Sf?*eAH%+-&3)OOJz+5oKl!5bu6U37ly4y`fy~X|$#v
z^i7j=Cno2<&utGZ*Y{EmWIl$UuDX8D)zq3rLCqmfCGvjaqSK24h`)U*6-Edis?$LF
z&H%PBFFlP=7x8$21(n-_Zus$iJBDS>c!HbM^*?A6YT1GhTWQ0!n<aqCWmv6VWp(*_
z)~eo}c`=n0te<&u5>Q?Tpom!Ck0bf<XcL)C5;G}!hL<qD!Y~@foIa(p<$$t%Dsc>{
z21RCBMCI*p*!a8{^>@LVhU0lgXaq!{f}5V#)eLf)65>L)yBh~JfhZl!VMhk>MrFf~
zC*H@+OTG3^TPg4!y9T=}#!|uHdWL2AP2&Z*kMOPzPOaF~&)l^JH0gXX+ugejr}wFi
z^IMR~U`CilfpSc|8V{BEG7clB<(6{<sL<({jtE9F&!l<5hd(sdYe4a`)8lvekqO*R
z0&s7;43bJEiL$t!4S_F6)g}qiEpOZH)StEH%l5|aCmx36mEU0-=OiI~#?N8^n$JDt
z$z(VkkG?wd6T52rAJkckTiCj>k#{Vhl!9E?zmdrmOEAZV(dkHH`Ycv(kED&WCfYC<
zqaVn!6z^|US4zVsplSI6W^XilrPzxiJe+Vl?0wc$kCdah{h?EG&U8t>69%*o-^Sph
zhFv4}{IaaX=+?I0Oo+(D)}Wv1w)c*f!AGE{%Wu~-LKoR)<OTi6ya&s(Gs+CZFOloH
zdQv4^m#|zVVkTG8*HyF%CA|@fm*o+nleV)&B*{ECFoBeS2!}w#a_q=wg@TNh!$&B@
zsx9+D=k#rm=~O}8-T6+#&b!6umt#W`QRbp6QHQ-!V<fx&6WlNE5Ta}N@bpYxm9$5a
zgM$qQ-HhccTmQ$&lc}Vg+8r^UePJL93<WRi*u%z5cg*4Ku3pjj%}pJNPDoLU82QF&
z>s<E`$O>U)_{jqn#Hf+gTE*8+`^Y{>gO~dn0aQ3)LX+UO@C`v!+>7^nSDt#_jR78U
zfKARzsRZf=bODsnHWCD0=F{?kZcKpg^Go7+ersW2-VOedtkTXBwY6WLaH_?Gn__5)
z_)LD1SnJ%QrW(~H;5)IZMZfn0j5~D~y>u4$TE{YQW#Da%XC0_C@SH)D)^*yL&jPC(
zcX>}`3dXb%qJ2E^>sdVXHg}Hb2(WhdzPAv&M!=SKaU)<lt*j{yEg#L`dYC%~k;!lv
zoUY4#4Ij+9qhD%OvM7P$O?totG|9i!BmQ^5VQy(vkU>9O90(V}P)ETl19{_QTh`;(
zl4BH@xx`7r#lo9HK&i@GCSTP0S8Vf0(eK5}6N~LFMhxSrjJMN2+I`zQKMBHUy04>=
z8URREslXIMa6pJMsPIm~@8PCaeb`D*Y68*7>9gqaz9rno;J2(L$Wr;|T*Gm<(27v?
z)PoT{_QUp^k)n#auPaHa@YwrwlZ2&1D>Co<(!(Ry^`Gv1SvsL6agk>Vsi<KLc5tTa
zBfvcL2}}RYl;c@y&*O2g%IwkC6|oOCZDNR>Iw7ZL(f)<F%tzSMC(mMtXS~;&=nMJZ
zS`CT?ibI!*qS-fEzv-%(-Ll3wb|%QThsbPQG8m-bw(X*VWj_U)Gkl+F;dNKMUF8tz
zPgi%Re2$?vXm-(**1U)%y?Ps^|FL{081!Kt{9#n3<+=K=*^tfi*$^zpUeb}(h0OuV
zd-m*OSXiy1D8c5fhPjkjAq6UT60{3_V~+y$_+$s&psgN8LsP)$VC07JGw7!^f+`C&
zxZa!k_DWj&sn`iRnVdx;73<Y&SgzIm)4bpX74gqE1-YsXI6?UppIijlrneq;R2{%F
zYK(P%e~>0KQDydug1G3MHt&|$TNHD>0(f)j%Gn>z1AF<$+9KbAOOIJv?XYTI>0a?M
zs0t>2p<#q-sA1D;5T6lJF#WqgxoclcAdyw}bS}kjFBjTN4U*{;+%j`sxHg`u08*L{
zV%nE;0Dk~?^=RhdhrN$LiNKHU62kZ9CwR<AKzotLU8|`k7v^RTah5v~7X=mfUaQyd
zZ<+r)3Z&H6Izj{#2z1U8txM024Gx1}@f(d*d+$;7*r4A71xP2BOgHgZbaKxI-FoFs
zVJ2I$(S9nn9tqr4UTk(ba8o79txi<~wWi-xN^%D`Zxv7UYKdhLR}3Uc)qPk@p|484
zeeH`STJi=5TyJo{j(B`_();B3j^JfJ^jla2yYd|cb*o-u|Ifh|MrbIhWJyX=61Qd9
z0X^8npI{TCdQ2NQ1dSrVYqoseAtu+|&Nh_e42oWX#GYJvT4+mE*z7@qg+(RFpwi4v
zgq6J25c>Y@6|bY2Ue#l^iN-&V7}cPMfCyz&w|HzVR`8NTS1*xdq&=%PG*!a8UprCK
zNAJY#o$I)I){AQK5~43WiQ<<`6>hxq6aReNAbUQr*18n@>W)#%Q}FJv_Jb8)m6(Y$
z-RL_&F#+&j5$-j&hBHs@Hk>*z;QBj|nAHlX4L6H@19dW`v$1wdZjsamEuu;Pcf1eW
zSX}G35v0RPIw&A4gssKB=qaLdx4~6r{7ZM>qon(DrB9!GAd6m=!o@f;3~{osl>CO_
z9mwr7QThj9>$XS+dRr_%RU*_Rb{Iks#waXI4#eO%>NkOW$Si@w7>hX>J+R<wgqeE>
zR!gYL{J%(h^LQxR@NYQ#zGV*~W#7xbD@*pBvXdy=AST&GWG{qlm3=2+7;8fIh-`yl
zk}ZS8knKH3-QD-^dEWQ)KA-pf%Y0ndT-SM>=W%?G?{OU8^BN;hk0uxDx(d>aJ&`9H
z3v7fNJHLa8)zcWu+XmkunyAPIkga1)i+(5U1*O3AwT_m<>M88Iw%&7H5`%UP;r9;8
zm~yk#^Ofm<o&L|1ACyy^V6M_O8M(d+tI#UH?D30%h34tXA@p^Oq}BGKPC~g08|0E&
za46pu^VG%%vllI)7l!T$8!@@i%dqchFeB;|oSryJq)eKSzt*Qns}VfQ?&;T2JuMy-
zJ_s4p?k~ve?J_(Dl|&O>Qpd?7zY=wJk-{&+-kD~4E-|fp;(o#I!K%V_SbBaUGK4hx
z;EESsiSGI|BGhK6r23N+zLG?3_kwEdbne>Z(pdT}7dGGCweZ78*F6jPi0B785z+?+
zgN~31S9?MXaWRGY)`yBiKa0jMl+V1B7zRqjS9h<Q<Lv6~A}b|kfHqScCfy0%kAl|?
z1WDk%ahVT&hUx7F`LQh$QEKXrmJ!r-Y5J1^<nHz@{#PO0ZvfNeCr3L{9O`B8Z|#*P
zE_Fze+c<;D*K$w!TLM6`JtzH{&>q6x({X;Z0xj=W<(OTMpR9hdZXi6lO2+!2t9(*Q
z6*C`r%tpYr(vInZD{i$uPAB(~Srj6adYUOI?O}eMuDU-otA40>Qp)3H|2)}uoaRhL
z$GBx%AF%xf^E!t3dgLdSrs0ClD|^UNZhWx|6x|XtJhsEGnK(O=k2n{X;IFkd+(;)7
z7bYj~+c`QSe7j6CV#!dghV&ETC+Cu9r{$Z(qu2Ky2t|II;pf`&K#Wm%&7J}ORa|Gr
z3&EB59|v?*!%j_!e_i(YaO}A()1>s(BiK@^$NJ&@hBe)6UmwJ2+po>FQo=q=qGgZu
zq5qW`?piGjp3e08w~wp|YaO!>+OkSp&@;G^Bc&tiE+JRVtfU^tvyn#LRY_@^`$<P(
zw_YCRI%E;JH+O0ip;5Z^y^%Czx@W_d5BFo{w$Rw{MW%PC3^xHU5zD*s@g(DQdBwvo
z72FqojFO1a@IU6W(3_HgK!VI=-A`$ZohWsEKa@>uRzkmp#5Sgiu*Y5sQ3&O~Y|vV{
zq2$1Z$%|4{hkcu+a=Uk!xQ|n)(VtZ}Yiu*YSxcDvAol&ByVtRbmDHs^!dNpoj-El7
zY|J{P%R{!thPV%nD=?OnAN8nL+5CC)2_g0Ha!XV7nS?j;jr=`8h7IkoJ`I72OU8}I
z0~OH75as?`KSXEAwBMMtEb24KlrL4${b5?R0#%c0Qkqj&?G0aR?|TNmE>=Zj=+kc%
z6qG9p$()s$y$TE&;pkV3nHQTB`6|nLPiRb-uXY94YT(as{A{$Bo1VPB$Hr3EZFdTT
z5}azU;@$&U{_`Y<dmketLAgN8>4tgPBkA1+&Y`B#v212cyO3>2k1CPuR?v1ORp7^t
zGK}}E^T#i0vky+Q+j1lcaWPaMf=+WyhGoMfe3#`{{BB_8r`KbUias`F57~P!!HAS%
z$Sj^(o8x2V)4Gpg*|17USX%a8Mu%@*hXP0nt=!xpO4r7=d`wvo>ro%*cY+&X$V%I)
zL>A8p0%*|iKnw1oLKgp!S+E@tYMeGdvfu<!Z@0>_zotbPwr<;Cy1CSZtPGi@B6v#W
zkA^;#Q$1~ZT9+ry&o@Eeag0oY6{OXHPYQ?Iu#>`;N@6tC#Fv<f5D>y+kd+Ph9v|9%
z{y^bubvi+aF`7V4TU(j4AgH*|{*kv`D#_5lc5jRQY`Bs3c+9is=r|9&&1!yUr)n`A
z9xV6eCF?72!m`)L{1F1^ans8(Hl+s_9e*W`cJi+?@+5J0SO>$tAG0mc7#5q>l~-PK
zSc|hH?38Mgxb`J;F7H9tK=ViQ|LMM*QQ_CMRu!etL(OjvH*1Hyw8w)k{L1dly_)ex
zWF=v@A^vb++y=~Lp3)KXX454YBV1;31x)uvdFKj`5HD+Y$fm4E9@-cCN#xNL^wFEv
zY2-<L;=*#NBrNd}^Kx~|n%gWZ^!Sn5yV621`Wy^l;%6|=`h(H;nr!vPPn)NgWQ?&B
zyq<GVa5lkPNk(8hU6$X^L52zRuL<39HUfxIr9e15Wb;{*-w1X%gl&{!E*jfZG>BS}
zgWgwcT}n3~!wZZPHgDL#D&Ivw9-rh<pig_|{6o!6ojgxl6ci%xRaG~2V`y#^FcwG^
zbcXVB@Qah?g%T%l6t8eh+82kDS{R)06XQx65YphRX5-`H;PNGT&tS5x^rlGCzk2)N
zN`V(I``^*qn_AC;R@4q}qTzef*G~<Og{M+G6qfEg6$Cx}L_&b5!|Z+-6DEbZ9zQGu
zYWO(#mqbRM0i$;7vZ6YD%L3N?FNq62Cl#Ua`rNj@VYBuYj$udp<A-_MuQ^cd8|_-n
z>hrt?Zb<ZlCZsp~7`!eIW8P2i$@GZQ@MI8mk*^rf7wNO`Gz=lDiTvrcCKlJ&&?N4m
zbL7wXi`lgkCdU*dW+h{z5Fpk!x1<V*xT>$@&E<YNG|Yh4<`FKo^ycLuE=TJ)-af*g
z@|FjWjr;1<@hx0T{CqyAZ!S;8f<<UgB|chLKq#X3d&a@=aH5Q<)7(BIvUN{8tm9QE
ze8-Z0<q8}q(5KpNG-~u#lB1tUEW-F7AL8AMP2DR(U#Jw;8~>`4#r7iw7yUfv$MOk-
zhyc<?==0~q5!{G)mN%hrd7GQ9x4d@^2@p>3__Cr){9bTeH0&<sA-vMX8yW$=Hf!M{
zDr-~xbe1HMe}ZGGP7PPgi!%KY1;ZR8-zxWag!_jH$?;8VafUqq)a2@oBbBYZGB*OZ
za&x?pJGKT#yuS6@6WYABQlxv=#uG=k;jThATx59i{XM1<M-_4&ozk+xUev!;9CI|s
z#o)}aKiNi<xQ4A~unpf#o!S)=hRgH4_CC@?dIUHZJm=>+FeJoO&zqcT`8h$V;Q9*K
z^`1?CUhc9;nrP5&z22iFw+sI6<vlP`Uw;i2eqK)F0);vFkD~zx#Q9LOF5yr@sp(SR
zIS~<X9-3!8hUQdmuNW8oc6NnZID(v=cb`TSxkW<z%{uUgjPO*&?UAsxr|b63fhsb>
zU5T|SSJO*PhIaxBawHbsd*uC)*Wwwk7gTSQ<`I-k`((3`bov8I5JQg$IdW$V{^d?J
z#8X5p+jjE4IBW3v{%c&5P}2R!xJIq-4v+j{Z^tD>M}r#O!d5aG1%?;LM095DZ<{ep
z@!gjV8zH~Ow4^+>=HTDMvf5cHUXH4a2OFsFm@<bjY5P+t*MuDowIv`6q|)pmn|{%J
zEPVTmIq<c&cP{58lN#W(8MZGGvMPCm@=d6#4Bch>RBQObod|u;JU&|gT4zBaa(Tbh
z=-b7J)*V6Q4?&xO18<{>i+7d8$*+$3k(~^P^vv=Wwmvp>0?7mAE4_}?9k8vOkfmSc
z*PuZk%d=oHc;La0!Hpja1#zdx`<D1bgmwm=zOQ4`thcZ6D-#sA3sr>TqlY$I!w+^2
z-uuc<@2V%h826Ba=xa&ZZ)-4IwmnaFSAV6RmS%?-{Wdvsvd?ZPtWwj8=SwbC)si~m
zqfUC2>H4#oT%DA)_{?WHmBFqg#Q5Cr4oF_d5~J~<p)8`aqt{!68O`00CtcIF)dBNj
zM*&9$)*dP9#pmc>;dT{-Duo}0QwhP+1;Y=&+TcSs5~=Z_a_{rl-aGOv!+LsRl}dPJ
z&?RYOwyy8B_+-%2deW$&Y~xs0!8ZEu3@or*9qA~IQwbJ~L`jtTZYk%D+kD;|pDVkv
zgKJ!N%^;#xVWb9O>jE4?1O3O_tlP(lN%+vhJ{QOMt7u7T#CUWy)d`Pa1v~W`T08I4
zs#3*do6b?)ldI^!JbD(`>DIb%Abd4sVz*rz6~!s4Tyd-=Wrzc&F0dq&ON<6>J!1!P
z*;AIfg}nbzs(i^Y=pKKONVr@_HWF<ZdKHbbmu=N&awJA8vU?vlbvocd+ssm>Gzf~7
zU(-RFx>Cr7dp*+0TXT<JJQ`B;!~Z5T+?n)xYLJJQE}=%G)2K!Lb!7`jJPDXdaLtyW
z9Ub}V$4Hwbv1E6Ta^;%v!`F@^`%Z_pu$P=KG+AVd$=uAyj@7}{NetRoCEIsL<e5;3
z#L#uOIg)QLHj>~?(IS@{X<#$CIV7&8b%jm$*T1zYQzQOdf0i*%>X26wG9R#oe{!&R
zxH4MF*E;)-ZC3VL=fu3{^~?YkhsAUpP)D20$L-1w#UB;@nvEmXWGzb9LH;nIMT)!Q
zq2?^FMw4&PV=lY3F-s{3G>!i;JC9yZZa2BgfoL9mkShNr#Mh~;kC{|piu52nFj_<l
zvx2{#2SE%4g0bp|@-?T*%ICT(Pfm6u1YaJD<lFWt=Nf+A-W^>ft`_UbUeF@=X{J0a
z|Eb1??Oo+RT7VrWQ|qy&=tfK>F)GwhH0sIz(Q^-ah{hfA7m1mnMMe!|n90ctPlx*(
zH1cm+T{dIl#2V+>GTEdXCcAs)WBoKL*In7bPMH$cCXeCzxjuKHuGF+0DOQE7>*X@w
zEF^CMGaEWY3e-WKN}bIsrZfnFFQJDS%=l1|W5Y2dy6+(MM)`A;j(*g_V6SMl>}9Ln
zDr0BI&>$%z>$Ud?t#1b9@-r+XxiPfdob-rjmTU^kS52><k7bzjWG`{(lQntdY+wA9
z?}5{=^5{-Ys$bp+@r%b}gd`oa{WvK_B&2jKu+6y;K^}>32StR7nej6mVS%63>o~`u
z^Y7MthhSF?O4{q*M|gbIwG>n4zx&X~e`10*9W>DfVcNE&lBXA<A4lDQw87_`uAO+`
zc39J{u*?Q|`Fk;0-6}Qv#_?)FQ|^+$J$`bWq+pWUP4Azy9Cj^xKJ0locI<aP?j}C;
z)^tUJigXKyW^uk2q`?7oSF9xgIW*BELI;dy!lt=@sXMW4qx;nFqH<BauFqQ2&UJ#~
zb4cj`h9?sJp=14}tb}-33hYnno@|!oQX@PyAX0XNP@H}LGGg@XOyR38kzpEOi!!8R
z%!|qg-M2*V2WLEMJ@Gu4jTv{n(H<A;^V;^kNjlV>=E=fieZ7uQqh+u6&$wAJJP_y@
z3=i#9_5Kcn#7$js`{9tmk4={qohdZrsI9MBe!4#DyKy&MZhk>6uU!ZSt>)wUJ(U!3
zLOMB|73PH|6Xnh|_Uv|vIL^3L{iS4mhvhOb%Q}@V6Xqe9E9&*-P<V@)v(6fVbw)fn
zVtaQU#`uLQj|8n4vR>W!o*(cI;lb*vmJUK4kpz}LigeQToAnrN0%G(p)6_<5#v-kQ
zjV!NI)rTj;louZ>nufLPX}|X>?1OY$Q$`o_B1?X`Qa3N^a;jg3u#djx{Nj9>xF;=B
zj?2wouwzza{$0a_E<hr71%E;#Ht2!Rs<*7<$y-{nk!Smu{g7(buN(9lSAN|uxV2`r
zjvG`+1pG>vDEXhTv^e|&AJbre9EKF`>1DwYGJvYNTj|mn%ck=?s?^_Q#_O1MuHy5{
zuf^>mL{Gqnzq`pjO5B$+3lb&hP{EPaR`D@GZqg?);K(MhMRFR~&J?~2y~ZK!+%WjY
zLQo6RCeeNz`va4gBId~kb;MbpW^^)T27YBN=M8A4U0+?!=@tnYgAf;Nhi#Lc6k&uE
zzB#=<!ni%uum4(ACnbb;0KUaSU?uv#?vAjI<rfmnSC@H~?h`J%^E|LX?|FR8A`u3y
zlDa**S`WD#dyU#;DE>$MTe6>#@yH=r^fV4<^9vR)uB-NvF{!WhxV1v>sXuCxO~=Q_
zrJX*C_@pL6>KIeI5R^XIK+BNNPM1In>`T%ApgP&oEX)B96Y*h3nx;w^=F?;~S#xMG
zBwV_#__Eg|!&35&7CMp!p?#MP6WlSrz2E%>R&9)rA$7MZo)xN9mV=)^1I<L-TYg+~
zwhq?=o25UC)vx22SnSHjm8>=4n_KXe=ZI2p88wiI<bb#=cF$GrvO9eULF~P6H)a*_
zG0zWQ2s1-%Kv<lmYIEF`tr;_MYt)n<Yb02ks5BBETehAvBgI1Vbm$eBB}8nZiaI%9
z<Hm<(XPmSq`{c6Q3Io9rX9!xGaOl3<*XATrKiPi#!Zd<WoxtPOxifgh@U~|gH{vAk
z-P$UWMVIqm$A^v617}U+@B6OsaM5$O=C;tvbn8?_C1+LDBpvL;`_mhX%lxyK{(!U3
zvdQUO=u@PPo5+I;!XR1`qTv|!PFUKvC9Au6k4$||4%ZWoi!Jlc^P9&F9h~Aw3giSd
z+v0de4r;x`o$5*Ks18Yk%y?LG40sLN#YJy5vi^ZxEcCnf;GnbCcFcw31~jm%VS9pM
zd=Xra!`}QuKfQ9?LrX`i`?E-E6#b^|lh?(hj#n=qsfT@U^Xn9cK8Gk(zFk^4S$>e-
z|ETro;Ky}vJw|*)!NyR*`S|HXllA$QV2Od;Va5cx-<)R4_1mq5ea1qxsE|PHce*<E
zFJ27}Q#^N7ie&3TUlRhLjd@=xZk4sxkGXi*w2OtbQ@g1NVhWNE$hi#y(|qrw&0*He
zS{hm<;1P5ju!&v&P|eWQ{SD`I#uoly&Qc3JwDSW(OlxAt!;oS5OiTa!@Mx!>qyjhg
zYPdv){VzTt*VR4w$Rx-vg`;2K!Pq3=W}4AiRBi#E!*9^uDAZwQd}L|PiPR2<m-jhx
z?v5PR1|<1e9X&U$h@v&MyeStLV{`1I)L~sl+^_PsRL8zYp&mnp4Vx3E@9=igLLN61
zDP`cT`bE(yxy?p0p4QT_n71ahpDV*U8Lv3NqZ10WMs7}mr^9ux@d^(pugW2w3cZ)J
z@``SksbTnG7|Kl0>i!dLJoJ*IHIud16<iW@%sep~P}-pATa-U=?4PT?d?|fgM1Pp}
zv`w6#>l`$pPu(CLRoJbZ6L^RP!kmoM_=!17k`lo?Hy#oT7vwJ&OE0Z<tRVK#Gv<ai
zjUK?Hk693yxXgsGux;rL1LrOwLN3MFM)r!w9%x3H@I#OMrZJHlsROSq{OhI8E7J4m
zsPe;Tq-0s@mFt@GeumbTl9%<l5Ir(W?PZt?Fls_f51P3l1ixfB%V^GSwO3I9tn4q%
zX-f}JCRw(tNjI!Bl8K1KKd3`C%gpmB12YL=(FH~bXO;UW<dxXo$#0XHB^0!_@0>ec
z?6$PZ>gy_*_KJ!%Zefe`&uJi~lY<2WEXXIGIpmxo!I=?T-{x~@e=%QS(W0KkobXR@
zdYXCpVxTuEVag>ewP6EqZV~~;a;U1dhhy$1$(knIQEdH;XZ1WazE8>4CBtao!WhZw
zKXl~yPb{!!vS)*OYMAB}rQza3TT=FX<PyGWw};q9Zd-*ziiTdJ>^<nxio~NuzH(r1
z6VG+VWw1^aAQ%3ldC}w*zygeQ6g9P77h7E!F{W)RxsGYFOeh=|FUFX0S~wq92B`=Y
z*KyV(b(U3cBlbbe;$u|T_@o+gW$;i4MmrtFqCLB7?@tB<E!L?KEh+SW9T($=zvSOl
zf@MrSWm25-f=t#kQncu{;o^JgJEN##YU4s_-`5+e{K^n^9OKHS&yguxO@kx=ez=GT
zrR*Hopc@<=0F6-SSSa}+?3j+10TN>W#er2Z_=02^LCwqf7YRMNYG(CpxkH(xvA3)Q
zeFU@*`?_ZPmULs>ijUxz+0xpN>jro1bl_P&r?d^<dk?Cvp-$B4cPD6V*SL!Mnx~t?
z-W>QuC8uT^2UWDaOImL{J;W&#TKE3m>t}7cZs$-IMGwIgSexIzKt3bsIErcG$L!ya
zeFB;DQPMFq+$zX<j&l#MFazH9#A{h3oZXt<6*>OwG<0N*k8wpCZnR6;EJjk>R~$K9
zuy5F>GObYHP>|T=V$`QH7dM15J~^F^b!8Fd+oq+9A$9*%!XkRp;ast&P&_mk9Zx2w
zFs^1MG4*U`zM9{oIapu8DUajek+79y2duorsOLCQU5=VzvD<nfK+I|3{CMIZi*ii1
zKEE1AN68d%--f!9qdOCn2=&?)Z8RP72t%3D(jSQ=E_!u35E}s5ie@f49}2U+fA}-U
z(r|T*kjYq>e>iV!DtY*nELU!rvwtJv$T;1E^2bR^spjl5GxYEny{f@a)adoeFMZ(R
z{Kw$5J{8JScdE*}4_hnx{CF_aPyGX4sLlmx{m46uI(L}v=A(_!YHtN5ylj~cqP(r%
z3V62vy!}?XvJX7zqrO}fW+<Or(fyvVsIqtJxuxwkhgb9sY!BUTC|TLC5}LnI+0Glm
z9cYHXHZoP;%lp}!W;t(C!p}A1Zlyr)8Q4VKo&Q;;rOz+1VLfK}E$a1+nwj`N#SLMB
zm!P=Ol_+#R`l*kdcZ#gvO+VvN-O0w8Xu0|_@Dj`k%4=2L=j*eTh=a2Z>=^$oZ!kB`
zo6tDzsdwZQ{xXzpU46DeEVLZ)RSa9*P^@tlp=3RO%qyg)V4q2WYBZLa5K3GgT^!SU
zO#=Jg_X9Cp$2L^po#c7aXgkHI`y;R79ZS6W<aW-j-KslM{T?3<wcYkh-#xO5VQ)-0
zEGML=tH_z5O&q5!w7;<?*^!v7T530HMg{91&U-Mbq#JM8km$T@va$b(XG2)<_Fa{F
zwxnYwI^yzo*=F7dCnZ^d@pkldU9}?W_VjYy{;GTV(}mo3>R+xTnw?cYnyqSEy$3JC
z30W2GRX>>Prl6kBmeR8o=fExwhqT-T)ev7v7eeFIHUIAMv084cNq3VRjn({dcjzmt
zELw(4ZI>!VtE03w0Rur~&O`JihN|0}%v+nquU%MKSwo-rmM#r^nAgiS(oi%L#(2l8
z^tEY^dXXUVo=@;&f>-Mjy-$R~0y4I^)ZQeLqiJXf1&(d4jvnKp)n1=Q+6wa{g(oW%
zHE}R+ss+~hW{ba6R-tqsX{fQ9qHfg-gav%%*qsowod)&4w-`3nqp#*Y$9Es-iO+%R
zp4o6i=gV(XGNw(%gKY|LLIdV|7<nf$tb&qo(bMb9je(%hMHW@M;rX?PB_5jFs-$|5
z`~@BVGvw#>x*+@STFBtOXN8l`H?)v*jfd~$hpX_>b<flIdYs_rVdiAU9k-io2@s!6
zsMGXPI$Df*>O!!!xB>|;-q|Xk<~V$G+(3h^LBf<cRSBT+FpwV6LnhYz$7FSIpcU-S
zo*_5DU+a5W&pNDK&41l5nt}=>=8T5D`RRz3?h^^V3`vw;WG*X6{H{TQ$Zd8s%-OcW
z#RRA(ERmR#AU3a-p`P3^C&cvKV<+9_dVJV6QyQH1y$^Zu;31>0e8?#aJwn1vj=<au
zls%V4LdZciY`tgWn>}_&>5#YOxaiwTiqsY=U}&=v0vAo&TX>Hi<}J%ADb2YYA-zsb
zURktTKNpwS;1_@UmYR+JHSan^)?RW}@#@Bp$oYYRToGsC*quk9d;<LQr1`1y5X`6B
z`Xc=7A49N!9Kxpp(kOH;GN=}63W(`?LXh;sYdBPxfOC8$ps%*$siKw2yaUlEUz-K{
zl4F<G%g^~BGEivuk3d^4pMSI9+?Yv*m0VFU^9%`wu+(<n2~OA`H+;!ul@Xuu1I2^<
zy#!R0ea2N1K!swo&XV&a=k69;h_8HWO`SMOc&6-AcjWY}cq%jWVE;EAF}sT>K>fo<
z+OkAGes!5(QMhf^S(V`&iLy;^>U>u&_||wVl6Jt$IRP%InT{G_g?jc_<v;@j03&Hd
zqUw0rXkv)u8rqK%TUAsS3CK6+X&(19hwGNvXoTt_9v$wzm^!V>Ig7-7WO!&CLfT;Q
zpN|)!Eqwdv&ZW41y<q^m4vQ0D<}SB!?dpnM9N+bPOXb9<&jGVdR?|wo^_!{mSo4{3
zdpwYI3)53k<}<z95yfDo!_Y=|vqYR`Cx^i5{^604(`Z6$S<{jUvtwn~oOfy5p6c3I
z_;da|g<!v}^gv^$uX_lYR2I>Dj#0jVJoNhK*kH4MZ725rRH67iA@q#~d*f<kyjQUi
z=LN3Z09p~ns~?t4ohq#cZ$J!Au5j<TrUhTY^JriJWt2w;9|h!OL{n<XZysx|N5@SJ
zw*TylkoNiLom9?`a$=$<MBq`}b!(kw-`2(@aK$n7xOUM=!zG^>y~@lTSp6a9!;S4*
zO!Gb>eL`1Y{g=)oHnbAp<xlC-w9O||Siu5qC2SWTZd|_7W3AZg-yN`$+@!{2|F=Zw
z%&;i&PjXgD=3GF?Ln~zQZXoXG0|^^E+%OXNVAYj3jFF0G)+Z8WmE)I*T`%EruaKR#
z+&u3{7nqfB^RJ_5Rt~h}cLKox*GpPyq`ObwIVV3C3qoti+><GIA1<~xGUqa`W8Evl
z(pxnv#nQ0QQsA1Dx}kjM2CrG`X}Er66|d|w238r#ATm__pem+vZP)(0z>-6(2$`)F
zo-;xQm@g|0B7;9Ko^1$lsilXI$kaZlSE;0*!%LY|zyncH`kJ=U@}^E-^-bv7Wzlm9
z)ACi)XK55iQ2$=bWXM;#Uk}_PqqnH7H1#@SB$U2>Z9H7#O3g`OXjfD$m)pB2|8)EF
zw}sjRV*`G{8=~9Ljo>49slwp{;>~7xLQrK26~lPKTAM7`ce%pgJ7tOXzaBM8U;hk~
zA=lU8h6R+=rU-1@7B>em5{f#@k3q@FiQu0Uj~TOUuhWtr0do16<Ngn}uHg=$avDxt
zygEm|BQx7+%d^2)E9y78NoGuYemKu#uuqZ5p+rl!jHt(caPeL=KU<FDj}pc?m3fbf
zPiq(J<{kxasOx9N+OTb2FRH&UG-o+Aj*nRtO^e@a1ZCZ;c9isOTfbEA{2`rb{VL}X
z3N;+Z!ok1v!{H)xVkciTYQo^}Mn(lZL&A5J{O_M=kr`Fpy<<P-<M}QBb$UJ<oyQNK
z69>{qisJN^tyy@Tb@7+(wjPt*H;Y#TgGik_t+cL67#wLDeyHRBNVI-a7C(6;I>G(r
ztDBv-rTl%>pt9iU(e4j!np&{79h2|8F^?3sN_Fj!L7S$Bdsl2g;P&EHRo=cGRiS^%
zj3j6(V$&L8G(kvI%DWdJlRaM@g6Z@-3P2c;vttNfv?+)cJ=ZzyvF%3O(|==?(Y!QA
zx}KHCGmNYu@!39(nMg67zxBg(D4~Ljs3cSV$Xc#sO&1eY^vZtrKGP4o!>s42+Dfyb
zn7fLp$&y?_bJi}#ze0YDxXSg>?|5OlE;9*pFFgHPl<FGLFnWoj-e~B#sjY2M3hJaI
z^PcFP0JLqrfP4G3)8Q8&sy5?%4k%dWJx-Ktt8{Ds8qmhHmA>kG#Z<k1U6}r956=rT
zhN54im2z?{n$ih8$vGxy!?){|1{6<TapWEv;$oo7ciuxJZGJj_u228|=t7#+luQqm
zALe`cG#5-9*G$Ufg-QUexi~BYVh{69G&=goA}aB#Dc(EKXL&QFt7K-ZhqB`Ocvs2o
z<0uMm-dJM^7MDfama(o9ts{$aVwTK`l*ixSR=Ot4iWe%&vtB4nd7QE_t<@-YGDQx0
zb#9pZg)}#pDYRrXdn*Um)ZN<s<7CEce>#?E9?@Tnrl%@5OVIXPXSrlA?X2=yVC_=Z
zDZ#l8F%Jt|YyYEQ&Z+1WlgC?&B5Y{ohkas9Z^a|LkmGU_cDaY7SsqfY8$WdutnYcP
zDVbr^zW5JE+^lu&OgAz8hM|^<n_w{PJ4r3=J6dKF`0h8nYn{m$1`Z{suI0ym)o3go
z7}6Ku$M8<2O4<E7iF=gz+*+pR_*f0};>=S=e${xm*OkWku|BNcp(!Yh`>xTuzn`Ge
zN2v3bor2r$r>3u0DG2OY;o;<cizZVZjiVV%A@&h(rA_Mt8t+pa6?O7}|Mz+nkA$z8
zb9oir^?G+_g7vGqW*nKLcZ-O4EroP|UJRQl?adjj0^!aFd->8i=2j6XCoQfKf79)}
z2Z@efs8qKOl%;FQW7rE&0sM{jWZQwgFOGS2e>98poMRM@2~Lr;YkGHhncUnc(o1<d
z^xl^ja!Gs}O=P@&1r+J677Y2{v-l4T=R)rtOsvfbKDYMAD>N?p5RlSwj91%441I^1
z2O$0>u`=q!NYZXJ0k7pk8EC0^A;b2B0gmtGb>ZsVYgL~Ly~#Y>$+h*lItcstv$9V{
zU3_{=-P$kDP$c-GGQ^nI0`@LYUAC26d|R01I`qwNLHI^KTUWBP`csth)DO{=n2$SF
z&IK%9yICte^PNewa_Y&JE>#achU~Xst4Ky9ntVZOxG2TE!lh7n?;X`;&S4Y^%Kh5S
zq$HE!)JRLVi84FR)@;~t6mqwDSFF?|t{aqLQcnObr1=ec|JwIn(C(ACd9dYs-=q)c
zAS<oI#FycR(&tM4<2llvc}=%6tKl|>v98kARy`5~&j4T587)-Y7NBnqyk`#H5f(J)
zO=$=y)X$KLX;)GhnV)S_X=s4q<amGLX#bL}Na=R@e$72g9i8;jee<~>$b*Lplk=_m
zGU)gaCrF5}?lCSm=xMfmeiNp3%1&kZCiGBtkg&f6*Ay_u)6x9$Kv1SQo_shU47%I!
zB+8N{(9L$m?50)nBP;7%uc;QJgy-AikBVmp4P6_O3kRA3k$uoQ@+u?O8gND1{NR!#
zz++FQ>V)PBTCoJSO;TfLaP_4<3Bozhzep6k>^69B79P3(tmF9eWqRag&|p?7dwRXs
zBY26UF8IhCK8z)=GArzSfI#jB21k$r{`jS8z0cTGy@>>nSg79kw6*+$FkE!kg2w7J
zDYmm4qDwDu?PicB@}!4-jZstvaN5&N#x+F&Z-f@Ib<5Egk#xlS`0yQc{CwVXb|*+;
zBBOvpFY?4{i%~0UplQ?+2h41Ca3IJAZYxo?T?eXnb`2$m;aX`^@gGL6^oY9PxDdB&
za1^dk;X#0iUG#4r@v%<wLKcQSWkX0}1J&tKn}WrDZ&TKNaYyd9_(;#Vo+kEsRC1N%
zvhrII#ZlbJZ!J09%R5DvxMjtQwE1k0UkeL)g_KEmeVDIX|MX)hOR@F#H43Lo3jRON
zA88i|va@~STRvr1s2t)osVe;<B3y~mNXT56{-F@p=Apd%T`BcY-fELw^d~EBUIzt0
zWqfqcH^HdwH_lwlQ=qfY!-Y1z`xR?8n7l9DeF|nmzBV{e+x5*SG^W8>Aa8M-V4gTH
zad9<26fUwMD{K?7T^m{K<(2H6sNwDf*$X27rj-gP3kiZMsFty?E@z{kUP;nDHe_xb
zFZ1fqS5|$&<pYqbBo@vVw3kVpq;?$dlkBRU^~_U|{^<PSrS!`Lh-8V_502XK6z`<l
z$9erKc<B45sDP7#l*cBEJc8%j6hB+zVA5-AWF%(@KRA31dqN89_v^pr_j-`Sx7)ft
zfPW<1bAm+z>7ksV$@nN`Hm)hNuq^nU<ddDD7wHI5hWWG;ZIC)<do(k4Q_v9a>O6de
zo^?>JL+ho))R6@%c@M-1%~VXbm^UfaIgACxSH0@U@fYRpB15&G#_x+1$|Nqv^q1^|
z@FNqz{{DbnP^MlBE`O|vyf`w26FYd-Sl|#C7!||qXV1Aivx{2r(Y>*XxRrw+wSrSR
zf<_j|`dT!xX;E)7KL=JepcJgiYHkWKKei}(VOfHzWFrbctUuEX+zIr5-~*HpN?HNu
z-mztXXq0li;?1(PBUwv`Q1FwJ(p8<A&oOMbu0LFkb{6G-(1O)ROuvA{$@q#O+DbqK
zpeR{UAM}_iHr2JZ<DF7~j>LINr6;}CL%zQ+VP}A{CboiGunw=2cN|GVQJHZMyAnrt
zb+ysaeVP21cHXFLJ0^(8H5nAD`{_*=W3#=s+s>WCmQEE4!71%6y){=v*=AAOhjo}}
zu&T~XkT0u^%KofL?LotD?5-SLR_a3rFO*(Ler^pMVzt@>xUzih+w(l8czV3Z%3a62
zH3onhDCOfqN)U+rSRu@D7a*8@F?m*|iG2a5ZfoXYmTR3U=mI)O+J>IwY`hi=7d`lB
zwz+U4`9;Q>PY?qC=EUj7HOHRZ>40m<O1ZZqUq1yD`~T3cYq>`GU<J=OkGt4VcPWai
zM;8Y95T3lQ8_r(!LhRLDsGOUMo;|C92I)!v{29q*h-huNqpL^Hd_YUb6xv`LeBRg9
z<VOa&cFvX{b}B-wGr)eLHIGSd$-3O~gfu+dZQpzLOIG0SC!{i>*xJH{I=^}UTqePA
zTp)4LuuXgCaB~<l?UTHzY~Mh??AGtmGUx#>jAV{S3JUEq&EWXx-G{!Pd8!lVVZ+h)
z1C1rVXJ8XLcY_+%%l#^S+wEsX`VBa!?SE!N7=NyM?;yz~+P<@sVPdM@Czd98t<<Q-
zit^6k4<8u`HjzP_C+>1o%D)^$w|h@kK8aKQ2p_}TO;AJq8u!Zc8?DnUx%PR^>WKo8
z-dC&Jrxz<CS^aHosziRBRc}FA^$;J-_Lu5^Q&je<KPjp$%U2Xpf@KY)<d$YCAN%_I
zl^6k<{iVPTi3Govk-RHderg!H#Bqz#D3}2zJwDTrewte(2W#6yYkiO?=F2Z#;jDef
zb>HBXZiz4#UZaVU&0(XB+`;DPyqhH10QY=w<JRsmwqZ5bO$RPk8PE1lez~T<KRRDv
zl0m?J8zOeA{NzeO-;IF)21v=ql1T>hJ$PhwiZM3RY#hq+T8@LW@}B)^*c1;@tojfN
zSCSLMv|u}}bKO{>0w_T9auh&>tnK@2rG0<*d!Ii?39r}vN$PsN(h{U<QaiHnte?U3
z5I1=Kq;;)k6&mQrSJA((?pgKR$EGF}KD(fNe0vQy1>p@{hfoPJc)fyWTQ6#9KmOT<
zV83o5t;6|plu5DaCP*G;XAH3OjFlhU^Sa*_Q6joIeTn=EsJr15451>f2j<?4HP#m~
z=vUF=>5Vn+%<E!P+$rPlq)3;4e$}-5Y(~um!D|NTuT%v6qXqcLXA=Fi>2*_b52cwm
z^0<RwN+a<-Y@dSZd(MSP87e1D>8-8LqUJwmRjqb2n{bAz6rZfFFNFsVbr0G<bP(4e
z^i6H6`Uyy+9r{lCP!?TNN@-s@3eB(vYS4vNKn<E(=H?H)vg-TC<CL$Uy(XOs7}WC9
z67ejriEg|O>!5~!S2Lk?nH7_;TRLU#^YXM*-<haBZwI|!?^WE-t$z27P^UH6`Or0_
z3q;;$D4Q(G0mVS#Z&Ez(?aZZ|y`$9HcZbvLD^0qV)4G|VpfJN?q9%h94zg}ecQBLO
zcvdyyMxZquz|egi9?}1zUnR${xN88U@_>5Rdnu>f?g8(gQ|L?q9od2)Y>@>O2RBpt
z2D@si&O(*tm1x?h(${p2bwl}qZaBPQlQqZuA?OW7PLzLgrUvP-&C07rzwqM$W~SxV
zbB;e`>d5kF_MGo_+%BqYn4Q~deuED!xwfC!cB^Ka$cv$L)mr$$j6@Eya=Y(ObIomV
zC|3RIC%Y6pr?S{2j&T;-9qV(!bNs?TzRyV#2nG9778*z4hn&j#N)k>&xyGH$6SRF1
z<~bYj5&GkU-uX)Z^j7-^vuWzHkm#K>N;$!IxJ^I>N~Et``qiiv5pV278pPPJPlD~_
zG;pqYB3tjpQn*gUB7^m>GG0VTB>=M?HVi!zMg`MZm_9LeY7T<_+VO9C3|ETv&kNu>
z>zf3G2gs-Mqn=(If?3vlr>${O$-Zsir#G(X=uDJo2tHh#UqP1vPR%YwghQb*4H^nd
zsB->JP!^b6h=VtHisdZ3g4t^|`zd>T=2B0lDk#zVPv*n!8OHx`Q^g1Mukbs3ra(}H
z2)!B1V1bo-)S(t9D~Sy3=>ed<tS*gf@YFq3N<>yZPQ<!`XXke_A`CQqda>2f6gfb>
zDf*_6#n9Ay*?|_(=u>)u=d2Smn|H3P@8RJS!Hi`Zgzwgw;>bOHmKdS0_lR~?Y{a3#
z<5j5yafkXQbWiB};+TKEU-U7IR6O|u+qHEHC`Wy8z3BQ7GbR<=K)ONz!%3#t-d=QA
zdoGU(KFU~?zFiVt?Vr|br<{6SOuE&|O7tQf9zuqnMaAABDA03WOn!6EI1~(d7~8r6
zLvBKPh&!TwLPcd96F<4D`4HR0HnEIX;LN>D7PmtF>5~5R%9BOoV+Q2!ss4seX5ln`
z`vVI`b@R(G*T*q`utXz_ZMZE#1474Woe^`nU=YW`w}$7u;`tt*aQH-(RGZ8*aASc#
zbES@K^_aFI*O}#iU^&a0kXo~0(0O5{LMB{p9Y$8u>767<R6;@g#Y^O+cT$`F%g8Y4
zXtE2C#A`&z^{<Nf(5m^nCjpl_tcm8aW;geQxg#4?dM52nnaxs161aw*VC)^{^(>Nt
z4VY}^h+V>(1n#5mze$fVa*oDZVsY+pF-X+T+;u<GsC@X@1e0d`bp<S`+2kA0Ziis@
zC~k=lbHUb;ShdmYj40OZNYEDvP7wOoW*yxMofUI(ef{-Ovu!dBovS=+d%fW|t-K$1
z@(6-rS-6|;u-FYX9Q|?I$4Gil(#~2SL?K`OM7@6Xd;|*2`@Dnv@C4{e234&1{yH;h
zimQLBGP$3=@k9teGGtPhMgfT5Yk9xp;R#Gtc0ci~(mC)~0@bR^q$-0NDixn=H9lUD
z4~J9QU<EKOe=mhqX)SpHXKD`0{59;44`mCydDck1`?o20Cgci={WX7fRO|obs2T~H
zQ8?2J9c~fa;eyYfI~CVna5YjJpksxXeNnM#q@?f|x|<^XU6^>|gx*<zlS%EYyuT|7
zemFG!S-oX6pGW%#a2&_*IaJR{M$>aX>B<gv_`zqRp{RX>k8vcqD|qlRe1N>J(&r)5
z)-YU><DpE4_1>*E<l4hPt`2LK&YQj5g8QGsV|y3Vn8Obl>}P&)*Ndn7yb?G+krBzs
zaTe^bww4H5vH$xm2%#nQUh2XS_K6h0+!#9)K%?51(0cR<QjrAovEDnjoJCPBNSfMj
zNf%by7_RqMHv&_kbH=)L!)EVuzl*>Q6VF0E)(&OR-rph1CLZniY#IPJKt(vHYO8t1
z5(hlu$$u9c@C>;kZ#?dUTXt;FFJ73v>@})0RKV&u(C+D2V8X6cSs$=WdCKYp09?kW
zGFZeBA;GEw?ggoH)adreKcjQF_PV*t&zMFXtl;m+49w}c(54{RR~!rudLG-p*G~wf
z<Z^iHHIT!9<Floc<ttanHS)y^<*I6|XMEz%S23zBvd2wg;A@!QUyFOlaNK9zVIL4f
zt;?<s-ky}?-4uT4T(e==bs?RnlmfCW!<`DGvl#-V!q3Dk6g_Zrv}T1}x`3*dKaV1M
zfemGdu>@{;|LwRKaS<cFirfHrAbJ+5%!r!&GjXJ&yiN~>qz7;=BI+G3T41SJe=IdJ
z!>et=jB?XzjoAFTh2_lczW5mEl>N2nEj>u?x$A@-)-i*B6?(TL4Jb;G>IfIC)8@5z
z@D$l8m+VL{Xwoi5AajIP7M|d^N&ycE+T3n#GY2quJpwD$nR{}}zP&|P*<a^y2Pdmv
zqMvr^Q>6O!TDway0gu0x^l)m#Ms)cd#*4;9Gob|nZj*4^qe}zKAdpEKTuAWaXyOJy
z0c9N)XvW3IymvKQxe7cLqe?VbM@tN-avHWba0~pVeL{5tfVhA|n+5-d24I_PM;iZe
zI}jh%VYLbHBf&qX>Fq`V4X^COXaCdt&m!bSB8#5cnTC?nEDgv+N7`9Xwf`R`0FD-)
zW@C>Q{l6X!P<cPFp`P<{^M9ZN*$)OGW_K+9WME^n$UuHmTa^B7`-a%MFL~KKG&L^b
zalOl9!a8@$GbI`Ej2lR2-?HYjy?#cilKhjLR?)&i&EX9!NT^>0tzN1v6av~90NH~e
zi;Gq(X`=MstSuv!pZX#kpZVTo)wZA%Z);#VMgNc3d=Qyoa`n;4=s8vy?p8W=Uzvmp
zWjr*9DiS6aQOBL}lW5yZXi669C3Tb`KGZ$S*E2Jmp-rCXbIEnp{o^yXLV|wFd!`yW
zvy0IGfbsI(#>T7fU(h0U0@AiO)vC^Qw5&)F0&mE%1@70=@z5oW<8zMJXED*F67jdZ
z<tVIrjVJ=Uk_v4Y(o=LUKylwKfq%XMbYDyd@5k5*ot&O!zd6Ch>YHd03Z!R?xs1`<
zU2dZV=L`RPJ{F+HA4wb5nKHyVd|{HZyA81Mt20(3HoTd=HXOnOqL4KhxI){XD1?tm
ze#!va&p<NwtKtd=wH-F*J?kOzXiu4+89P1I1FPKw)-WLWUt9RyYLF+btTwVxu)m<5
zkByoKEo$0B^CVztU;bX&zoA7BlGYQZ>2t1C5B)eI5jSFj$QjH~=C9MUv(u54Qm{61
zp9c?YhPC8aXFv21a9L@8d;iWU`&0i@n&CILjRMkc`ZNh4OtX};*ZY>HOjaMIVKqGe
zLKC;m=#{S~01*?oVe3TbA%5R#-+@|~%L9@e3d!^lPU#Ekk1@c<J)fqoY7kg2Pfc^~
zq4+=Qh?DmGp+Q>(T}y<L?ltxskCH#_<=hpPhhZb+KR01}^<7d*1x!U4KJ>VK)=-M_
zDxp$@%)LYWVLy`6I43b7ta~@HYKuCKh?9wnk3~Wj?Au|5EY!_`g|Bk+?pemIg0III
zy^bfVgFAgI3HC<9=*iMvE{q031^>+(4JWW^qm#tZOMf%zA#9J1iwK{wNE~>&w5X{%
zkqf*_H-!n(<}E?nq2f?Exf_^lD;M>dgb(%Jfx(#q>{bDyLJ~B)8dP~XX+^$7{Gtwn
z^GODOCpLxf>CXPmtqz6s1{>O!4isRyx?0^o8`&3-Blsz#WZjb*0ePmig#P&=nsV^Y
zr_Y9m)a&}h(trI-!i6UyZNxi{HC#jR*~!i`&ARyI=~(lq=4GiK{nd_R(?}7~+o_XJ
zf~n9V1GnQ=iizd__ud!ek;=4?RSl(W;NjJmUE^0^%a87`_NIk4Cfm#U80io&9bX#o
zx<ZQx@upB74X1`l#CeeT<c(+&U2@pK$GDP;_GFicPe+BZh}I;)S*d^&y;S_li=oqR
z-NzRY1@B&%Q6+S~?D|Y?Dl4z?q{!wWQ4qXNwJhKQ|698J*)Xo{bVX;GM^KNjpeTK;
z;l|)fl>(Nwyl5f=`o@Z9n=10U{}taME~TDX0-hkz%{y(+aDyq(%OZuZmbr{;milEx
zndC*HGC+sxrK#gcgQcrl^3k-@T_@DY_G!nCKE;Ra{n~w071tRGvH`XZ@hbJ>O-H!$
zT>7R!1{=%Guo_>p-vh@h^cL!X9z27j*3fGGy$N$+pQ!1R_MBYo{WSD*Ivi=gEgsTW
z@yRyVeifK3XNY4RQ<HgN3x}qahmE+jqprhI>zg55Fk7C|bbjz&4>@dG6Eyilk#qD<
z*9YDg--I-9D%u<sIJr&W6<5b#os`P*0NO@~(dJuSk$oT@d`p3csJUP`u0c^fHD-hf
z0dHA6ZgYidZ16uk8COfkOn<JvRWZ5snew|LgnT#w9rZD~8dr^)v1l$>2JNA_eW}&L
zO0>P@j$W7)aECMZ#K;g4qk$m!#cVVtjT%w+@<{O>1K}cvFibqPQzw;=+9vg*I(S0R
ztsV?eh#y|oSJQcLC*iRM=sZR>z58%nu}BwnY7I9o8dkZT^74?D=3=;~VwT~nSDdhU
z8>Dbo=SiN~ccsfi&r{<GUvA_O^nVH8$ih<3-{=svm2(-E7_|z64HI{?DD!ryPta|-
zLekJ)Uk=N>?^h-=&ym6R*t#b9HXUYvm{Zf;=}cR%f8&_RO-w=AY!9#g#KCEi&hK;V
z+>>kMlrMYSwpnA}<?K*#LI^$gxUyLtG7)j6f2d@mNk_F3Oe<u9<~O$8n3nr&Z}94N
zJ7u^W$<a5utpG?{*2{#nyCR+S*+kT!i1t)}Pb2DuR}y1>tCWx6g^8*Jmb!kSP0+r!
z@5$-3%6h5I6fWn3sN*#BgyGffo-HP}ISptWCKJyzM~EMZK?=`T{w*J|rFqX3q%dwQ
zKicq+bOD2g7XW>^!ovd{&hKoxGY&e5FDx>GO{9TQV8H-zns3aTul4zzLBG5DY^MRB
zrDs<!;LWKref1$Uz^e)C(`l$g7SX9{zn1iH5Kcgr=ofew%Q4eXzz4nX3K#D*SdL2q
ziZ}cl3czkANq0_|YZR*nd~RjrIk*H6A<Lo#iTISr0E@%Rnl0o8BtROQpkr4`!`$j=
zyU21ZpC&#OomKjj7(^&g?7~Z#{dRU<(x3B4(D2BSy_%$7E5}m`*r1eviwFhr-nU(>
zO%w~y@&92P{it(Ly(L1O+dlyzW>X$W)B(X2!VNC`FRp{FNdQAACzmF$_kpGT=MZ)X
zETXIDtb;wm0Y`rPn?3*M+p@}OxtSawp4y#__TP(wIP__0`E20nNSh<%oJ6tsi|%(3
zy~v>@FDdfIL|CH_KUB~3@tfZjMG4{W*GQlJ4h>$*CKbGYHQvTD-%HJ?YT22eFyH(0
zWR-Xi4jM_oEhxERj7l&idIT}oA_`Op*e)8)#TT#L61`v&2&F}+fQi*6jWB_5ovBgy
zqqTdIpX*v|y(VBx?h?6*j}e00dy@X2BquEDj0Lh#V;3m;;2HKFLzl7l&?of>H!Hiq
zd7L8MW>Z^*mzwExy>o*D=E(tb?IysQWL?$MGhRM(Q<OWKVeVq4b90np3D_(`XL(QR
zLVPus{50pvhcgC`=XX$ty)^m?_}-FeZA9aUOFkY=jUVTaEDyYIA{6b<92Xg54XqRM
zy4S$va*by)2P~xAg$WBi=zqD;8eIH*5_oP|GBX18jKbRG+BE_Fm!E3zuQCYLa&TR*
zW4=fj=c)ynAk{oGECo{5GaasMB5lGo`?#q!rijmE;pcR@F9DfEd}H?_aM2DEX0})T
z!@5}O5ItL85DJ9#oD$&jlmLQD#C1|fa9WSX@PUE%{4PuVS#WUP5N!QF79vH2o}%P9
zE``Cgy)8dIRk8~Rc#c%jn3W!2I->rTzd@UD^(=OU{>}rk&t31P--Z0UL8+$m5AK(U
zpVhtO8E&P0c+mYMk1e_~Lnw~fsN!v1CQzy1*`#V>^BRO6q)i5*<TDp1(kjWNV9h2P
zuWn~jd&(I&cO(2zf*>8`0kk|YYGH-W2&xwr&(e8Y-GC20TQi!tWg;HsurxKNxj)7)
zuV2pZ2vpAwORQ(ly~oK39?S3=C$BPOf|ElF{2-gs&V{%4$7xFIHJ_)6r`w;OkCOWN
zVfFG=2H{$cD{A<dUh4+VK*bJg2CPd@DBk$9m8avuTGE!=|H7NW2BRfw(bfk9An}$5
zGHYfiThdeRABNZzDmvv3<kWW>p1vp@F4%CJsOjCvlT|}67dmP)j@{q#VXapkaf9ZE
zT){VAQOlhPmejyn2Y}xSc}V`tgkW!4gv9IZ$Sh+TV^l|KfgUb30;|q0zd_Hqtoo<C
zyn;=V7*`4g)zwM?N!MK3ePHvv<9Io3PW3d7?IWa&k9?$1Iq>{vN-vxR0;AkBSit*R
z&pdc1+<@cSO4dGX+4X?q%mMJ+vN<VJJhNPDWtgg$KpJPp6A(MZ^k*y2CeJr;JKsTd
zS?r9eU^O$=lvly#13)*u{5H?%rYCD$lpl#c>MTDHy;#JyhHHeqGyziu0l7po#X6NO
zaZM*PW*YX=)tabdHr!2D(!_*r%j8&s*k-$sS`OI5t7vWyEu8Q}-CpzR<lle=wh5f|
zOi;fw!50^UbFY!%c~LW3W%MolOUm&-5y#-5SySU*iF$E@QcOg}kqWlNa8?BR3w<D8
z0s8>{gFt^N_-|1CuP^)uiLqz>pP}V{K=gM6_!sQ{Kiw+824YJ;gA?y&KVdkZz03px
z!cd8~0R)d5TTtOx10Mu6m3ZSaJn`Qt&(eeHtU%uaEy#yJ$>~8m+sjX9lW>U^ys5D2
z9!U?+tyz0nPnt0)f+oXOFG(Hs5x%p>WD|WEMwo|9coK2{6btd8y2>A}`MmMSEtlV<
z1~d8UI&Bl!dN84KKOFp~+pZAJ{$>eU`nbT<K&iZxMYrcp2XPuC>rVhWMR<0-)gO32
z*?v|M-~hqt9EC0&9Gl@=VRN5;`@4TRvFcZ_Z(+dpLMT@c`TOlN<h=g}p+XE<uS%m5
z@+-JiXu%kiuhVu_o&z|>_&ZPw>n&lSF106Y>$b;-_<@Ki-pW;(4BjP&vu(U#RR3<{
zeH)w|F5o!&#Se^NKa$%+;jD`Id-KshOLJCQ`cLfpe_YKy9!Og7q^_Ps=Mr6Ht&`P_
zPt7+)vt{^>7S3Fd{C^3kf8n`W2?)xff2>nTexDT)e)D})kOW$u4JP%&XZ=bX&D+$t
z4d1jW8H=U?SAo6y{(tXR<|!zUd3z^KbNsIFcH=P?t)07e&G0)m`-X(Ig&mekNFKFb
z7pzxAp7_9a0zgqFix@p6>e4#jh-fbq0)>G%>3??=*<b$kH%TY`h^09t6z9Kqf#|@t
z1!hVGJ5$n;$f3X3u9IqCN^b0w4uJgsr+s6c<Uh0gFY(y_f`egbr%kpOkB0BwJkOMb
zncU5PRtnOOHCdIHv|KrAZQkaP$(&29XTdHN?@eGi*<Y8Pz-c-{VB;OpJU=qH;IneM
z%RDY$rQkv(%gm(_N#677?j%VpE2@wXiWe35P#O7e|76tX%gD!B(8_K>U<uhEre#r6
ziR7Je*ZwT%|F~|>v=7rGshD>5Tg#E~95!E(CsF3~T9{c)25`1D3kyC*hp9`50`ezO
z_Ab%#>IS4R$m6|hBaM_e?cKb6?d9GZU}ND`5c()eLzV4RGjpk%R@Ce*?#gI_&0ILP
zE{hB%fX5!DmSqrAl75^Zkh;I@-S{k60?(dP<zVj_9OtaUssnJ~i`PT~RXHk{3iM3p
z150-yZ|;|s%iBhuAcjGeW|9*?F&PDye1`~@$}452#I5=$z3<2);iYGYt4Z8bFS|?e
zLwZwvgH?~`Mdj^8{nlS9GS2U5kAWryhD-E&D;}6@Yzyv;Q`nd4gpE9<@O%|I|M-xs
zmVrx_A2*R|JpsJN-r?w-|GwJ^9Z1P(E;ky&tCQ@?t#<IH8(_3fd0O_LUt~YD|8m~#
z<R`^s`3qbEhW}{<V2`-+<OlVTyinW6@N1THYaLz|wGS`6(TZCBxHU$$`O34w%*9v#
zy4s8I`eZ%W%V$q1@wky?YM%|bM?X6U=UTEMmT$~Py`Jdx!W@MPeX|qq0afVfht3Id
zARE~h{ZAd<Dr@>Iw{#7cK}&w<iRCkLgkMySG#&0te{?fG`=9cV2kp=GvlfTADYw!d
za{kyD2V|W)jK-gKo|R1gSYl<WG}XyC*yAa_CBXoabu;<fpZrEi+Kx6wn#DGa0~@yr
za5X-Fmf9)fC%n|58}BP_=@8)=l)l9)=qnMiTYYjCpx_!|XInGlfk$9?tAF>c38wAe
zJ^jDh`|7ADyKjF`Qc6;!8>Er$ZUpHZTBJj|Lqb}l1q7sPL~!Wt?(Ps!8kDa447|qs
zUHA9rZ>_uTx~%!<nPHyi%sKn)y+5BFrv!DK2=G-+sBViikal&R{kKN24dO1!)|3XE
z27<Fr$NIW7)Az~J<dV<Nua5<|jF=k?4&JXoTob(FV^^NOdj4olOLP__C)>I2@M$$U
z1|t(<)*;mRwrafDn~6kXeFTXIT*yvF!*Ck{Ei^-^gt{D1Mfp3@VEp(w)V{ABG;7R6
zN>OejMdQK7E(ww3u3iM5&^HSThpj9^9-MMk$!E+ve>1%-^FWW7w5;fN+cHrrz>jRM
zpJm_!2#xWG;#7_lGY4l@OZ!PNNz7Y;ZYsU2ki*c32b7$RKq(OwkXP=u6ycf4S|q`F
zY-R6SJ8!<QW#J{{OK<#<(vgVYV?|G3EGOqVE={A$^t6jYQZ>w4Ph5wi(ofY9+Xn=9
zhAu4}8tM*L1p-05yHGWzAc`Ee^NInOC+7L>;uFngV=kk<ef=rO9$kl``=hOd#()em
z!UnxcT;f>{QFHsH#fpm)U!lHXo5QJQ))WC@G@yni34F!hSo$jq;8s@vfv95oSW>ow
zy|X6(%=P(`#z98M3WJ@#9gWVkO2#7NZMUw)0X87!wsc*R5g>VdBSk`-Lv+UdQ;_jj
zF2M2?|9YMsl$!jl-|kidzJ-SCoDlw!WNytf(M1C7!53tpr<tv%M3vUy6%<1U{6%5j
zeT7?IO`!1>7eh(TyZ8RNt9~lSI%x%z|Adn3|N7{U$HXjW@7fCU-FGJX%VfbtQM4hD
z<83bh*<k%?i{md`Hh|IzP{D+g#~VGv?K=oQizcmMg9-tPXuB{7AfOtsgzl~~VAcPY
z{f|O~)~~Lb9T1N+xI^*gFGRWo1ie~mOQtT%d3fTOwxseH<P9%e9yedTzZV;jk}+N=
zB_mAj(_Jt<>EF|jwA-U#h|dcCjsVA!{-iDHmaH7cVi=8n(D~2C0!nlV;AG0BCV^Nf
z;6}KqQ{18@PoQ3OAxnc%#}aS%qYmyP5+(>-1q6V7_yL}H3JCqWxJ|#2+0|2QMf*<G
zg4v*rg5R(g#}&EcHk0OdqAxT$$x`)20Okb#yXHP51(?ev^-vBduHS)kT?c%g4j|LQ
zvCyS!d7uDjDo2O`%m#lRkPeil>g-oOdxy~O;#~iDHtH%PmcNVz1OBa^oP_PMb%cT^
zqC1HAODxo1RPm@IemlJTOyS=v2w<AuCh?cI^IPjW`!NgOlEcmaAz)CC{9nNGFR1|i
zew3b;{a`(Ip;PDwxmjcQd*6|^s+4d9zU32!Jbq2ulKREF#bLtXC9$iGNW@<CqqPJ@
zKl0Tu51yw8y`9ehW+U?X?>-W$ZXe$L#`rVPDiUCJ{9DkHv^qhEb@VZ}97jFt((bGK
zp}yM7L)KxGppPYSgu|RbP7U(z>_egSCFT(O6u-ly<Bj_~iB>5)UBS#81h<h~oKxpb
zYt+&`slr~cj7m3ASo}**%4O9eR-h=v1{%J4^|a1JVAF>m8u1L!xE0iH*k-_whAG0b
zhEga$fU+9JN?2cJRFOj;XKwlLSPZUCA3Lt5S=VQ--m%|q&xf~>7neXOzCc>O->6kc
zDb)e+dHs#i?-4|Op)7lME-7`M!hy#E0)#{8IbwQhz-_V+<uf;c*|Mz2Km>f3Qu^jw
z4kmpGNFC$ttAcVa<LJD%vtoJ2T+#57R5=!++q<N^hY~qFcdDno;I{Vnkmn)LVg{3x
z<@y2FZB;)MW^Tc}sJx?JpwaPPva@unuKxddb4mYijY9MHFra$p4+i^$xC<T`p1PS+
z$cBeU(tK!HN@3+r#ceo}|BA0WBB+;9Mpn5u<Z)!vS1@!p3EVE3#cHPKEY6qHHQppe
zVyq5)VCMtHntTa>hs2=qi4B<z>ko`3+CWhF4D{GY_-xHH6H{JPeGFt}{j=P=;u{9t
z9Lo{`$mk9~q|_?*#HTI>n+*q5imQ7x+uyO792jphFF!yC9BcFAJXyg4_%J8(Dy0Cn
zvs40F8up22DdGMFx<R*o7c^I5;7{8O_B(WsJ@-PHs%h>J>MT+n<E61>i;EWz$~-iF
z`A-zi`;EeO53~h`{sY(gQRbZY`->PQJc(33l<8BB+!T-2;ZK5;qM99__#%O>D1jW<
zD+81OfHrr;!5?4au`H&k>Ro>LD=LUZC&2xiY`~sd{GA1IN>a$)=6@JjZpe`j=*!zS
z9!8AhRTmG#rqp8qynm8e40nPIg*SBPKjiH)eGwdikuw_Eau9qMkthZYb${nF{3jp@
zurokGeAP#DNqg4}Yv+TP?uoO0oQv>F8zsiu02#%;J0lvrWD7Lyn2HR$_n}b9{)GJA
zXNELrb@_8>pc+J|C|r#|W+93+r+7ryJRc0%YWJG)d|$o4)ZoBMf;HN4d|P%8IPD%0
zW*N<N0ktq$K<|TKpgqVkiV@YzLe!UY5Lk!PK=|+k%sKBj{M&{2xX`<M-ENLu<%|V;
zMB}ZO;y;Q6^lS}B@c?inMYQ9QCEaUz%K4+170cUb6MeL-tZRa~i>AutuDb%9D->p%
zx2PpECPl1-0|^P@Lq?iVEIJ>;3{5@)cGAiH&JwMvG1Nr)$l$}z%z(=RWJY0P>i0%g
z#U!&5&bJVLGyTZHZVf~#-yh>VL{c#Sfi3o^p6-h?R~-iSg$qTMm0IXsV+~hQW3<{D
zc<8u?{&8G@9pnonN1hq|&K-e<w#%W>^NWWya^{+FuB3d>^1oj?+MUSUnBg@qVhx2u
zTG9#-KqFjPZEqj@?cf$B>^lX;w`?MhvQjvw?C0+OPU3N>2j<WliDP>+LER7drIY^(
zkRe*<%$swxMD3dc&C3~UG51&irhF0iS5^{OVa-097si10M6^YEg(mWsN19;OiceaG
zT3Z9T{Z+#MVQQ?nbw4|lpN5NEtS1{5Mqp$X1-eKAqPO~6Mg96^LV*k}KnrpE6Qe&(
zpMHTzZzRp_uqAypwtt0Jp`WtQj~F+mI-D4x{_Dp;C+qf0?7Nb<_y1!|a3>!rn2viN
zyOMU?!Jl7K3rNy@h<!VlXZ3&lqZT@v?yzIk^7sGzd9HS{;TvmtFhDe06{%@D+f*U_
z0u!C&nMgsYVTAV=ZXHgmw;Kw#+veBg1Gn&mT?%_vr{cGy4>PTw1i>Fx#&q-2as^tk
zV6Eo$x4!(5m!YQh#-8i9d-2Z-*2CDh)93>X(x5a~$Y|(aLC(A*RhGKf4f8sg<AP|&
zyOTKh_BXt9XM+E}L#?JlPwWGrCyXFz%z%M{=%942ac`odAuIkXok6b)?6G2kaNr)$
zt;oZ+qWzP>6+UF=iI~9UR|3CmcyfFG@zw-ZfXsIyqnKd+;#`B4)W`uN9p!GUa2>Jv
ziilI=R(E?0nRe7zII^)i7Y$YXN`jfS#)^{Wa>^VYVoXA{T|WKo6yDMGHL%USFCW<a
z9ikE%oIw9fswryrPMJslHvCoYuoKgLmePx6J!9NDbF^|aKiU_BcoO;gV2C+0T5vVj
z#L!ScksnOg`KO*WsN#1$t2obx4@NDB^2IU=IL^ek^{fE<9z4TMVN&Wf2PY6;1NxTr
z$1mOcEJpaq(zQ2Gd7I$(va`VWmxXEqB+-wUop0UjZq_Y&UX3f9Icxm#4FUa@&0V6P
zz;c$n<o*5=4ZRvF7rt-1S4gv^_z2P!hlrYLs14tyvE`piJoF$drQox*@cZc&gYc1{
zR(WlK5#O<SvN;ydddeIRwIj|k%WCPS`S5}U)A+q>+Zcfnk7*+#k#GAj!N8vQQlAP2
zVZTWH0JDVp7!e#9?0RszHP0)xx`v9#fvkz$AXa@IhM)D`Q?YL24mjEb9wo-;?Wchk
zZ|k1b?PlO-=<hM&FJqm{)EPF{M4IUhvnDhnU9N&bNcxpp=3B*jC@B12+TKT~wj78~
z@R?tlIGa+1HiX*jik$^(-!rV4Z~E|>>+*{_98j_9kFJ@}T<04y)Lp=Q#mY`kh)mEF
zN6^=888UX{Z*OJWNd%|1_f^7;mlH3$;vLwa1f?li;^4fdGfnKwM&3~7nZ}Lb)8)L7
z^rqs7`(mpd6LbP~H@4ZA)aNu`!U!RJ8}KLCp?JlMrcxCh?aiVhi<`Qbr>*qLlY9<t
zs}217j=66BN`!k7KiBb14e)F<HkTxv`s<diXVYO%P4cEH^g-M@X<m<lPbmZXue^g-
zAx$V%OMPllv#2zev?avMwqG9Dlq`!JF{1}cy)k$LHci?e_h5L(5>|6ud+|`tP6t(7
zBltXIU*^&v{W>YpLCbuOh73=zl3U6W%%UFSTBQtmp>Hqj)f09^`0VY9%(aA6%`p24
z|5QZ@wsWLa=muejv0X(+BaHql&JcIiC+6ZkhW84coIWbxvbpHl-uJm253Wa)0PR`4
z4U@XKL0PSKYhI7aj5bO50SB%cXW<?0Agg;jhSjax_VQ^d>hmI9psxJAT{_Hw!BdkR
zDRvd$P!{psh47(0JXmH(JwwHV1ZN+ZbTgty;4EDc+%MM!Lg4kVvv0NY`EHu}pc7(<
z@o4l)VuKv6e^it@>kYL}=(Eislu_eIXjYqbgOT0qBEo>>iNMfq1ht*765kKq)4S~&
zDwoy3pQ}_?CDhJLFi{f+P`q!ByR`cH=2#;1Axz6|D^uQ8mmxd}aEXYPu7d!F?2|0y
zJP?MKqR=%Im@|%mnt#I|z|Ht*ySk9OiN9pG<-B;b<zP4$A^%XE+BB3r=z&ctou;No
zNWPF`s~@qi_QC8;+g6eGXk>*fiXmTx=C=!(L2@@vMzoh{nn9~}FI_|qUAdN%sB!Wa
zhW$&+R@jqYaIX~YS<Ccr;+-^XA$_c53|bD`WY>I7$AwB1R>_Wz?7e8i5aRE-E3xTw
zP(x{1{k#N(r?6r~6kcg~WyVhQe(%W0s7V~xN2F=c;DI+k0Y~f4+Cn4gWBBD$suMw^
zc^qWV8lD8Too$KaLgQ_jn(ND2v|88jL212uMh?M4p@>l0Kt&fcf>7;Xb{U7Ft3bjM
zw=)$hsvEK>=ZX=L60i$ga9AwPiu%FLSvio>=T+zVgqj!>qS(Pww9({dV`<Z&OPVFu
zZjQP;zmRV7`UGe*2yKlCv><&BOK*<?)%%@~|D7eFxHXKuv2s`TR;Ak&Uxx$j=K1?u
zz8X*)vp3-3i_2(+-GRCn1iFYa-i~wqY<tFlgn~xjG;J;R*Z%`ww$+CP>=Q@bf0wA>
zD?AG{c0wzkhAWrt>+Z7wXY4$%PZo>)xT$H?_T7nmc2~sEFX^@Bl+{s%x9E;Re^-yJ
z<AYLn*R(U6I%-hk{H|ZNyZM{F#y;G(2B4!q|Ezz82;1P7j2XPP;GbKnmNV*sA@GmB
z%-*YHc;j8o{#`%k(&2lrlaWx6JB=PQHxW8wv4*^n!llh@%hbt)xoA4F-NoQuUCUJB
zOm41$q`8b?!>XxmZBS@p+!#aYG^4)o&1CIH%haG!IpGxugV5L21|E{0;KR95Z{Cxp
zrSC_`@{K(vCK)YucxiM)xuz~)rjkF^o+nkme(S-x!dyh;J(+$QpKKsN_;4T&;VE(i
za9GbIzk0xfg3Ugq#vT8D%Z}F|ot0tOv_p=oS{D={{cuT!aY<`{qh0O|V!{0;hhk~4
zsSr?_xOk2<;$6?bT5#`(h5P(#QNPyp*-2NA%hHD`zNEvSZ&TrGtA$6781>IU12;aG
z+rp(<0^y4lH;u;w0u8xV;j6-&gJS(Uh2!+&cubOrI5|41{VK|A?MZ=wNdfp`RF-&{
z;&tzhHR3-8XXtQ_Jy%jrEBH7jAaK4gv(sJ6U0wZhQ4e{;(f1<#gk$3a<ivH)`Pgo+
z{Fu+cH{2(aeVe%wP5Nn3MAX%4o{#9*HZIT4J=IMY&O)K^p{A&=yeHtCDD13NhvgMq
zfde--)SX$99C=(w(fd^Yq^GX9Lh}`x&-8=p4vn4m5$R4OWV=j%n5B@_Zi(Mt=+uQ6
zJhjrtqGZ1izJ7&AI-nie+UfIrwPbs>eR*RP4!1`T{6c(V)mZb>d{ab~^Hi00dRYv+
zw99FQ)f*}2!Y71Sd}gc(zT7s?vG03As_l+#?5JXOCo|~cm_S6WSM?C<suk=Gw4m>~
zZQfd>+)`J05&PW{`^Ya%M&=I+S7TdLo;piPby6Pmwz?Hd7`avEhq!JvejHEa6MDjp
zC5l(e16zdULaww+lz(jOj=DQKuyaso;(J{dnRGRLj<XYrKZ6F*2b<3*5~mB;f2Hz3
zOMbtizreI0p_EQz8>vZrVXU>u`f5ZBVnIIGj4@LePhli(899E`y*L1Bd(*{}db!L2
zUaSh8G$;tn^yVj1+D*0XC$Bk{bdqi_ZYo6MAO8VbuX($$lj<WnU(wGewA>e5D`dep
z<>pU~{Jd&$!&dj88{JxNF1l-Ze*Ek3=S8<bcxSPAxwUP{!q_b;jy%0&EVsM`;aKfl
zg_+v$n&UN59x_G(B>@doZoB#BJ_m1z>>ILOUN`y`8pJF^|GLfQ^PjOfL)DR2_tttK
zg!mjTQ6618bB=JcE3`B_Nl015p6Di8jp(v2Xv%gPn4k-3rOUw4#x&`)_bi1CZy)Ay
zTM9%+F>R$J`)!X@ZVEjR^(i^<INAVO?c!}OxN@}Llg1O?628FdCzueMC{AYczSc=S
z|Bgx%y&YP<!SYdj&)!<=wDL6e9k1hhw)nsReWumTC8;m{FykVQ8`UUfQsuTP5$~}e
zP2@hiO_pB94+pm7uck(a9Snm-$c}rNtN0vT8s4)acx6nrkz%Jy%ygi#H)YA+V-#li
z^ZAZ#2(fCZU>Ss`{WsM=jN|Rr2pLVq97le*jKQJ`wjV@F??Xzr=c1W^{0Zp1Z&J|f
zN2lhq`-J+Z)1%=L?iwNTNyR8okO{4?l(`>w8157E4HOIFPr!n`nXf&$QB$E1nmvu|
zphfA-J~dCLq#Zv%cE6zJZ<4C^+#{*G2+Z<5Nyt}*+?#v+iSM}GHqQ@<QyOA@NicAE
zjkvJIhb8|KT|Jzw{Nq+EJ6>`7_i|rDiJKF0y%)m^xw1AKKIJv<UykyHe4Dv(`|(Jt
z%DCKPcVjS7y55~d`AdclQ8F$l-QKK~IGv-2^aeNw2WR)Bde8KcwhPhLgopCHwu`}a
z-7B;!uasHsS1$8Q`_<-J7dbejZX*SSVDIT8EPso!VccbsN-t~)Rs<KT0-u{B?nUwy
z-|)!j+@L9SWt#8(p%(|1kd2{?D>DA&2Cpmh%k$YY96hI*d@JFzdNZa={4!603~2^c
zc8XwX&$kK^DiWuYHpHfUXL3C8rCLIsJc?d7MLJ_t8^)GI&y59`j$b%nKCaiB=4JqU
z>Gd7!myvlEU6B<h^FIheS(EP1y;7*=DGV<H<~PXul!e3YCw<h1Q)S(%t@FS!W&I1^
zhdh|x-YF)RKdEI~1S!3qQhK#aCpo>W+dR+yY;yIffP5wevtw%b2zffWWYUF9qb=~G
z+tXp@6S*z7aY(vX@nD2*-h%CBUC^s>T2$TOMV1s!1sl|C=eio>M)5rdz)6pqflo|Z
zcI8JC-)?Kdv7;tFYXB?t+*1f+^F@eA3ax?A7jbj!o`!?H-4S|Q2HmWfg|Ik0Dkcc+
ze=UaR^^d&?YLR81^b187N2%kzH(`Tu%C3r2=eQL+LisLKHsR%4R6m!71J1@0bx2vX
z;UHFZ87>vhVPNl7eF}~j4$EQo={yW1)M0pH7wgXBHWBj)gOds_T9PIdka>q_P#vrR
z5l`@gH93X@S|diK1m?_9i7HD2Tl!kghWix56@cs%|0pX+W2!~ZPy5-mEcgy=vrZU>
zEMBwny(~xNs{ZR~qO6X7Z|{K3w>Eh5{SNuASO&xoz7>r6J_nCgXJ7d<9KRU3Jo~wK
z`h9Yd{FonNnXv$qgsUmUOmp&1Rszpje!^CKVXm5$uL|3CzTq`r2|@;Fl+QA+UN8Aw
z{)OcWCgt@oZ*88l<7hR##+x(lZQ)lcx>dC+*jVc{M#`K;8X|+us_z0lrA|$yP6Nx?
zz7M#KAIZ8;bOrlokH!<r9^_O!+cZz>9|BfV1|JhQuZ2ru2=9p+4AX)SZcf-E7puC>
znaU%6WO3^mUNu1jYOjG}K=)UiV3D8eK;$e>bmeq9ix`e8eI1aDNPV&;N-l2>l>`xP
zn3?+C=U;=jVjEL=Llj}LhrfuTIg_kxjBY%{oe__G<a#MN*}zD&yHKC<MV7kz$)#BC
zR8ELuNplULi-9(@v?Br`=FV^|WmK>aPA=wb%0{u-n@N=f^!4Vo#U$_;7Xy=!VO<hZ
zL9_wNF2U3vG)dk(VgWVJb;0UrTRZL4QqF}aCPgReyAkv+6c*T0VvGe7i|tR3Accgu
z5+S>~!;lDg$hD*G3Z9mPxuyw{!wg=PFF8ywfoBpunRFoVgxDuW18pKDtmvT@0{i{0
z_?%6&1@u_QV;ir{3c~EBV^D}5M+U{dR^J?295iUDc~+~OZ|FlLcuiTceNHf6Pnb2%
z|0yMkPjfr`Z9IHHor+JP0`ifR==EfA)u97hwdUlKEd5WJur~Le6r9Id;Tqn#VkD-p
zv=J}vVPJeV$mc5zXE>&YlXbzm;%h*0hir%0PacJW47^{T9r&^s`wBWPC|11;R5A-x
z;vtOPjnD+8r{a`j9gx~dN1rT;CaXw(cA9r_TN}(82ucF<*V{;}X$PNLTh6*O%NNcI
zbO*iST0;D4)TlDh9cwom6;J50Ql5tRW$<D33U)9tvS-}O(GJOF`nYFzD?FH5?+0sC
zPToyO)d6Y|r&f^8`*xmfE!P$G)1Q(yF4C5mpfyk3+j*dm*moGh;&3UNWikY|usI@1
zbOZtA=cO0#JS=y{kC0z-nUcp(@Ji&7x<zu{8e5;($q;hy8EWJqG~E-!ALya(H#Y+%
zNn`bjfy$J7&G;jgf}3Mr&0i{NPMoEydeKG6KY6`<{al6WYmWL!sT!$SRG{Z!u#8^L
z9DBx=XReLYc*Ay_>}gl(o2G2NOEW3cGK2WiQDzN2p*Bt*8O@Mz6ke)LezFHfqCy3V
z`cFeyK($OyE<3b|^&ob0@-)_=C?I~8K8{{-9bsk~{ym{<>{7S$C7e&~XqcaEeUdhn
zUIMbKFhtSzvGan`v%-3F=v$^q-ovphNC6tiAhwTkPqx_S%_qy)2n$T=5vfo2+_&*O
zzwg8%l>hvR+!_q*!XD@@tX5nYfxCccsIzwT9^)$K+F3Yxi-BYBuO1Ay#x=$6M+Xxl
zthUDZJn@>0KPa?XG-R+lqpGO4+c29kiU#cKmge$kV*dTr?pC%)D?Pxl9_QHXct%Tx
z-vpIFd`{2+c8iF7*^!O9IB}xcZQh;USVN$VJLJsudB}Ujs8G>Z1WvWoT7JhjVDHbL
zfo{<dSlY+Aa*6Z-7C$Da>O&omy+)6aH33&I*36OUx7`cqY+LHriI066(zR8Gk_e`H
zrWVXQzxCC_9c^amsBlvxO50(r;nk-vfjw_d<3J0d^)rQ=T@1(j7V;{QMH862SQvDf
zB;q#cMY$6in%nXh!4c_sSA{dBPY%W?QX{$-&0BkTDW4aYCW6MPP{gC(c7I#GiW+4l
znvTJE{HW+?$hEz}=Ec`%k*T!yRE4oWoB%%-t74>1x#Wg4<9cJA0&BsI<zcTGntCSk
z0HEm)$a&-5*L5kDIJk`^6BDP~#|^hvZL~bOuNvr(lcz-BGTKXPt<6gq3x@8YxGhs<
z)6{4-yer)V$)sGXu_I(>I?h-A^Y^q{=7KL8qv_*6lo8TH|Mwyo7Sr<`uG~OMQVu+q
zPu3Q8$I&C;0L;&8W+7tEokd`8WbGKq(49x^)P2Qj2htGqo2}FAFUnl<6v56)I;b+=
zz6aYF6tv!ShcjG%v=XjHb4Tw^8t*pG*}?`-*|*d7f4TR2*1u+7yyMgz-;FZ0`i{dY
z8mjCbM91HVB}SR7r&6nEoY8o8jEIYIb+Xl@x{s1<eQwD0KGSNkN{FfDDv^CdUCQ5G
z`Mw&_b?KWkx(hcp=Mnu)pR-Lg+;UqS=@t-)$XdFW!j5ql!?{pvZUJJ$z)D-P_jLT3
z*CO|(Unyn{7H?fjSPplX;xR!t-O4;S5oo5q1f8Hao6Jc52Q7R-8fj|>X(Bj7cb^LP
znxa-oDrO-opaup8-9uJFR88}6@yZP$2xVQ<eegx1@`|XV>0lv*$HC0N<)J367Z@+3
zF}*SWse0FLg=xK3gAk@eFE{0K)@L1Ka!7ReT2H)ScSnhC<#^_^$F;^6P0O*CY)`HX
zN=1a6?KPjuc4aqO&%3m5uEJ-f<v!!&;k~FzE@_;qDJaT(3O)~<ddE#C6;b<zx87@H
z!e9hgYnv`~M@y5D>BT;h*cu@VJ=zfbwbTeCa&3-wsbg;T%(Bbu_1Fp_^2K=bEpDlx
z)vgmRsvliVY>sdDhk52|oN4=4)hB%xd|TA1Kv{vY-Den@VGKg#7A}pv*OWBqTM4Eu
z00=T^)LIUpA1|qsO?X)!*xQv##<u&4rni{qhwSVD{lQljK?8K562g2&zIQtXZ#8}3
zoNO{b>{+CVJM9qeJj}g7PxAvxs(G8Hd1+Sx{}{k0{6ZEJh?lj_w<@y%`+&meG4kwj
zVO{wvzw$&SY-;1LB7L)V8mbo9s<nzFiia?9#5U8)ahe(6I?Hj$UX*>}B6uD+3-jh&
ztDRWmlb(KiKyPQrSCOc(c`lYZq#h5O*E@$*y<bcDj1n!ga4_rBX1T_&JO{SriK@5b
z>>|_i@o>tv;_Ilu5hGZTr))BwUzeXyh@<DqUEAvtjv!F}@Gl{@XlxUc^UD2lu;+C$
zSQzd4*dgCcoQ3nE1Gen3$;HZcb@<!K?so>Db1~yt#3S3hqZ*y<ykiG8UBW~E)a$H|
z4o5T3)L(Rm?~Il;^z2d~qP}<<dn_9h_qO$Z9ptGzPdSTk`Oza?4J`C8d;7>GAeOVV
zA|dj|7~*!I!M_^+ImHyP#5Pyd^<FI#!KLyx3VatS(^`PCJ)3wPh#@8Q>3L<`n1>01
z%Y+;`-Q?D3TR3jIS^?P_uf42LLmU$d!m)KxS$Y+}zVi;@vZ=*cOGP=x_4IjQM6c8y
zMst!sY5vPn<vA!ap@wr(>07*_8zeza{ytt%V9^zutTYp)xlEUw{q-^`PHwF?KN4d)
z-@elzRS?^Hr?*34;~IOs{F6w-oTU*(INSWY837CakS^HuG3%<W*qWoS&Jz?O%O4c4
zsQHVdWv-8Wg?OmUzCO%L4X*UI+5E;=ts&9ahM9XLpIUP~Z_(fdI$ST&-SjI^ZUZw4
zQMuM?S9@PX$x$XuusHCQrIQ$(#Lb({`q;MzZ;#2GhM5}4_9nsy8NIsbkg#C%Fim|E
zNyJ;mdx?~wE!?$8?5(FjW6~O|jG^sEW8j8_A&Z&JWQ8|sA2PPN($zreJ8U6uIa#%Z
z{e3yM`-)o7i=rCs4()HnzM6+@agSVjvHw6eLG8kq`@nGR1J#d!Q{w6k-3wL*_#29=
ziK@hB*;%|2T)tn=k}oNJchAG4yt}bzDxB9jq_0X=u?L{jup1$IeY{L1vV3iSjAA8u
zk4ogv%LNWd#UYdNgZuMpx(B?+G$dP0f&3Qc&+B=WBkJGRZ~fqro>%E+#A-=mk-`3X
zW!P`v|9wpci}EnTzEXttbJ*Vx28J`h_SZmw&+y1ZmTmnNjd5lF9*$p2n#sSe;TT2y
zTITPiwb8%(Yj|&Ag7`cCb&ag$2NP5a&ljpr_1EyoU>~?f{p%X56b5(?#R;oO=FhwU
z!@=1lx(^+(q1aMK|7_R0GfdO|_iO&8=R0$ffD`z7d%DcHbwV&-?HW0bUjNj8EJ*5Z
z;3PzVe~sArApy7~d38_M#XIuFR>kU2qY}C108zeL1FF|7XN%)tob2R#QsE;M5(Z^B
zUPn_^7e08243l1ueBGj_9;ojlP8N$obXYVMCi(W)@1_oe6u5K9CnYUPEHeKPzov!P
zox`rDPJtm#qL7|JGWmqQsnjY`)s2#lVtZc!4T>w5I`0c<W3{&!a(JF<F;!kie89=7
z{(Z@$nfAU&9SB9Y3xE*@6Lj&KLa7#b2nMG=K%_kf`pi37YcGO7nB*}_HPlrP)vBtt
zF)zRXe%J8F;iAp=s=^<d1Cx7M8S2u{PevfKrLS5Q^{U+9KQ*c#14rygeQwd$Dsv&U
zovRAE3Q;m#T*l4^t-^Z^tuw`Wex#C;=rA%_5yB;i1sF2M4{UIhSt~;yhpqBezCYn#
zW+oSiHP#SWK#gP8hMTV=%<@<}Jl*=#$ZE4d;8n2&1J<t%1z@RE+gor~0@Idwe`-AC
z?TXqf0>VNW2$f}`hxq`8rU-HEM8WNuxpG!r@hP4^<;{)Sjm)3!Juq*+lwG-=-Sg>a
z`s`TB=aPLM`Q}i%{xDOdK>(lIn~gq>p#+>(Zhn8A`I$(mftA{)L&lXA?&aNjPyD$&
zVlb#N^n3@w_&S>mheZd;9XMLTcpmW7$MSWW6|17t8*m=67>Eil;#m$qBzPQ>!ydBq
z6(bMej|db9mh)B)m7vT*VCmg<(n<lu6mk&_p@qegLA)32rq5+%9wweqdqUiP`dB`m
z2k1vHHBL45qBTKq>c<hA1o~cP;6<VVsd!%?jB_Z`byG}vsgP~NxhudyJR}ule4^G7
zWB=()1%AwOJ^D@N{Jzt)%BE|5h@#};Kz|INf&v2vCAFT@)~VzOJblp<haTr9f4hJE
zBWM6+IoxIu4JXrt1t%e(L^w~V{%%L@lj2m6QsrJNz`>~^o%p6-B9pHg$GXhD(Da_&
zK#ZMXFyf4rK2~eF+B_oM-y0FT+L>G&Xzdifmy!syvD~GYNO2*F!jeQOm~zc>I<YNA
z{|vs_mTzM`muT-eIkEwZnQPtYpW<i0)}FcQoQZ`&&c^oF7tdX34^#SRby2y)h>D&b
z{xPV2Xt1(6jAbx||542syKL}u^#&4+R42nSbF%q8X(@4+9TdE-@U<;hL!kIjw&m(L
zjsM&CQL6rydXM-+w|dEp8OWz3xr7w{o9gh$*%RZwI7uyTVN`(i&B10`NVwI5WR!qr
zdkObq2GC=rME)!oMB+_vlJSY>kU0xA+I)8x)N7|yjWvzbw~=tX$55aEv1Ot@i7Pj{
zhmYO#Qcs3n8gsXZYrE}zWGx$uB@8pDA=sxiujqAgM^r8Q@<Jas&Q)z?o7Fh|6WfCJ
zDh#BQNP+5C8A=;wa&d%b3oO^}xMzo7)t)d?E94o4JY-DaM9m4iNy`Zix*cBAdq6>#
zOxqQ6BEJ<ucqntZ!kWvX^*UWCU3G#A{wu5RUbX<xjYMB+{~Nx{n}|1GT{qP`E9Md7
z?S);Jl}<wjnx=mwj@jvQ`8)&Bwt&ip=Nf*Ho;rJJUhVvxM)@{nC#I6$w=%A)J|m0#
zT6^D!^*lPv#f|6aFn$XuQV&MkR0faWi9;;n%F^&2t5b{Dk!Q8*EDc1x3KiwnT=+o%
ze@eIZM#~e>QmAD1R~<}@2I=`D{2dZ-^rF_fV7+>9gXJdO{Kq$2devo=^6P!f<7z5`
z3$?Bo)sAbdSM&XVYbqx4bO;|&nOb3e6GGUPLYvoH@eJ+o-kqQ#iutu%`5Wdb1D6a~
zZ-0gZC%H)W1&ZJ>$1R3EGfcZv#Ec4nG_|C476Q^_$_RT_8nbQPyIXBTlNeC~UlMa`
zE!^k<<*&5kUXk**2~>faz;RG9;~)F?)-Mk$!~SIK!^_&g?GvLU@C=pt_MX3u(f{4_
g|K{|6ok*Rsbb%@bhX*@7VSqnbNks{;m|?(w0plyPRR910

diff --git a/docs/use_cases/scf_to_cps/by_income_decile_comparisons.png b/docs/use_cases/scf_to_cps/by_income_decile_comparisons.png
deleted file mode 100644
index 9adb8d1d6c4565e266f8505d6f026f7b38564b0e..0000000000000000000000000000000000000000
GIT binary patch
literal 0
HcmV?d00001

literal 55602
zcmeFZg;&*U+cm0$bayvOw}8?e(g=(0k_PGS?rsGnq!!)XAl=>FB3<Xle$M+nd%OMq
zfOEz>#u^H1vDUi#n%A83x<lk;#gX9g;a|LXfg~v*qWIzk6#a`AFFRmg0nbnxg1UkK
zUfL^)3%w{ECfs}Rg7k%?h@i5w_F)FJv+~SzU@Po#w-$W&(o#tsl#`Q_uiNPA@v(f9
zz*kd^TxZ+bbtnY%M=yrsl=mry)N#_dfu~11^&?c=cNQ*JeED-LYU@9L?M+1KlD$MH
z{lnum`KnNRM~I4_@E`v`9s-lT4o$)v>fev_$QRHE`5NfR!GCyB!rK@6>IGc%zaQQc
zXkT>KK1+AJjQRJAFW}_CtHgh5C%Vl$pegBAl3b*J><2vdA|#~%#J?Z1;@-Xi6AW-L
zB7b<1QlQ<}?N6PPXxoF82ir7It$F{u=fE>>KsG3U=-dl9E~1z4*s`^hgs6Y&JTPL%
zf0zWINBI;2?HzlG{*dH<di{%TJMW)5N4feLXbRU-tc}3Gdk#DUOdiZ1+yB3l_ov18
z|BK06ZA7DAZgxY#QN^iitoNub;`5wVr<6q{m>R^or^e%PxrA7*oXK|Y=jn^t%STP=
z45>^iw>o%PWd=gC)fcUu-@Zd3VYXOT#Qc$f*8S52E~Z7;o`S%Uu<K^9b~MK!LRaqG
z{us&81i58xf5<Hum5*?}J92F`Go!MR=|)qn0(t0efLk>w=f5l2XXN#%7A!U&lTOa(
z3~h-tR=Dl(quZOwmN9HvP9ksmJ<qPTQlyrCyv6z_*eMM}v9g+K`EuoDpm8-Tj~%OU
zVdc*HOt}*HE1%L@tXIAkh;y=+e0(Kio`e+Ba%8qSWv=6obZ|T=lr|Ek8jQ}J-MWXE
zRfX7yavzZ~C_cwFgxbhCn-S24bXhvq8$}Y+YWLc3C>mWUuo}fDv-N?xWzrrC_h4i$
zZiUZm_;1HSiwwj{fNr_jj$R*0kz(S9<A}|W4Ue><i<F!^T{P4$l?d46qrWVaFt{so
zAAd|evDU7*okD#~L+UEU%0VM@2PwVIvgf)_cWeDYZ&{0<&Y@my#<boQ>eJ!E>Jjnv
zomQ5P6gWCWz2b5DGGT%N`|oFzj!&Bt8K`Q1E5c^2W3{YxB*%7W+*EsKr_B4Ev(^Ak
zM-<6RRn?@`CDWX@&|%in;JUjL17&kQ*k!+Yt`c26SiY4a-7|(_h2Um^76D-$MKG9<
z5ju<s>}?VI-vhga*+-rUwAgU>f<qli^h$-8%k~of@trL(!48s7T3S&y5w8apqfVn=
zf9L!STO-cOzN5!$3Zn|B-RS}0gV5Ep!QoCW=S_O_?x=4%k_W0xdJQLSIws6cH*Y@>
zj$_Adm-G&M^q%cs5FBc}Cnh;VTeu=@k6ejbAD>z{^|&LZf;SeOs(?OK3KhvS;&WSj
zXUwqVtMsHf;YPLXkLHvko1p(jm8VwA0M23s*EhhuM5n=uv#VEQ(Qp8J5R0@lAi17t
z;gkTG<b4p^P8pd_<gYiWOqGRRW>4eQ?{RrN(cZl+3O&SAN@=)1ZAX!5jF@U5-a^<r
ziOOj3&$x?PTa976zFJe-Qs=TZb|_~axX`#NHy-4>f9$1sNI12WZ>H4m3I5Jcz(|&n
zrrMqJns2sdhpEZ=<`v!qrSQ`CZpg0c=xUv27ZPM5-l7YHt^Qzms-FdtEea5oVCTC_
zB4eh7?~F*}`IFXaWrQNv6<g@!2mD@B$cTH>s7g4G*>@tL;bmGIQr9eI5MG1&=KK$~
z_axzI_4u48JfpMt3KZfZU9f8`9){vxmcZBIGq}mhPgn4VSJ+&XHLNsL6(3t$zC%e5
zbT0Q27^&f?-j9jH0UKRqT|b=Ch(zJTLE9OPa!BJcB^H)&57w^Lk4Qf>^Dwbm6~5No
zreDarQ!Y;2Oc_Yw=9uR!EXJ1UuwL{|OfGk3i_B`tb>o(+`E0rXhUf@HGPszcz9POp
zU>Eyk_5JQwUw>FI9tFG<PXE_C!OWnu%4yT^a7p0>vhX{vjq$!^^BrzHNsMlV8M48j
z9MgU7h)aP#CYx8jd%8SSgRzqEL&8`s*Qg~vB;bybQ+N#{n$8rUXN#Ske5!PrA-O(U
zh4Y9s<TsQ}B7}U$$vjnQ!fS*`1#j5RmX&lHK6sw|-W(sxF79<mRR1~W8!wldirgo(
zpG_$zFnCwDM=)m#?X9-k_YO-d{5XRVu>looT1xe(={vLqPCjEX48)1WkQnrvH)`d!
zD3`>1A@8q9as-Bd6)1dV9)F0R+}3oC`S1RaXk&pAYStx#@#P+@Sx>qVw%P7*_=V4>
zfx_MM((P#a0hi4KiyFl;$^!4nEjoe!{{64CY(Yr6r2eQ%9P2M?wdJl71&Y-C#5}Iw
z*M*OyL@&IlySeTM1Z9zx$R$J~F$GbG?V`U=>!Zd}hlW|;DXx#j?aWr(F*REs9lmrW
z<>*R)vNhOWlYdOsuF;DkhTR&Crjy2RUdb$ns__s3M|XzAD+W{)p^^r9!Wqw%DQU$s
zq)M0dgYI01ri2eiy6woU4soOXmILT&m7CCFL<jcHXGXJoox=~8K6Jmx37%Cd_=GcA
zrUFqype(PKBiO6pTEL(}T13perLBsC6Mi`H!P@O{9V_)SUJiuMFPcS_#|oT1rPrOC
zUyo23EpX#>$u73!RoQj0`*75`t0Y*Qu3T>Kp;H|Qg7QanvW5D2T(yuvtWWg@jbt(d
zU60d0cmhSUV(iOzM&U3AG89l4$;t7^LmApTKwfwnzRu;*ug&q(o1KY*A~Ti7^ne%v
zOL82yAR=%6#y6liA{e<sB?O~K!OBjzmYzccMRYS5?Y`RtjT2grCa6qy==(kc3JJe(
zchAFEv8)<<p;}LsCpl&On`ERb5}x)<;nyDtlQ|l~jwNaAM&CQ`lWVu0O*!<(F_aKn
zbVHdgWKwXs9EV2IeZY01l))8Wh;{42k`5&dRC-DVq0Cj_h-mIDe2tvqOuuO+kN+l<
zg8TZZ{=1N5LIt$4dpt)HX@S9At;q7>-UrJEk8RfZUn#+i4Buh!H80IeVps&E>0WQ3
zGK|y(WB&~ATT|oeZ%FG+Q>)%{!un*`{BuVliqm>!Z#kVUuP9Ws=Nxy^bpTf+eDq3S
z?JyXZEFE6B!a9yBGXGBG_HD9>R9Z)tyfGhg;nhqCLIcdETB1phdihm-#f}!K2UK?_
zzBxr{D5(5`h(j*R0nK*eDIW9vCGnI~k1yO3WH7tR>mG!^BSxKYfdXv$`+L6ZPv<_s
zVdNfuv?7I4yM^!k{hPO~uIv3I%oc&`KG`BT7j<5HLPh_f26DE-6gOKu0`X=;l{KO`
zle=xRh)E}H#&Ri&qw*E0XFACT-+=oyelK^Fu1LJ0m5z`XWH5X=eM!70+y~5_y5_u-
zc<UkI9M+dr>Xyr)N1trTf-Wtd%Jsn}^gc(p!eC|`+)W$#JS%I=!7MCcH}>&`&9FQ9
z6RJmKkB*9i)xBuHDi$3#XoR&lJUdY*$0MmC$cT)LS(@5nGHNeF)>}s#^F4hME9_TT
zevwqZcz8c0oRZBg;~Nkx7Dj!kysg;fn|-i~=Ng>=S%Qhz`lGW+4vl5MhTtpA!G--R
zGhS%~1T-Uj2V$aCuxa=Ral23#_{>i|jou)S;T}3{`fyS>i|{dbwlXl#6gyrz_>gQ)
zx|NY2Kmx5(xJ7ox>b)k)mu>q5rgIPRq3$X)qG;}=s5diT%CvPhmnJaTC%8^kX%j{a
z1N2~F$%zpU=0M#PIl;JZvUPPC$;d&I#ULo6G!>R9>v;$I(v`;<V`kYLetQ=A%dGsF
z{LXKOt#4w;zbI#De{x{?V(+1J7Yiq@j}t6l2HH;MN9b{V?-~)#0_7NtEt_~J#g)<c
z8M3vEDF3bx%oyvC#$ILlMG$nPCAYz=%Q%h?5zXf`C|<}+*Nuu-!-yYgsz9-J*&K9g
z$}J*V=%a;SFqjIGWk2AP6&2s8uW)~Y+Np;hZfwJj`T~W3mJSoCwR_TRdvGi~Y03e$
zCA}$_by?jy>*YMh{I?9q+XG6dT33&-B^<Jhlm%J%eUjT=k|+&t;E-ZcKuK88fvN=>
z^}zH=7-5Ryo~rQ1;O89tZ==~(+!$&%`}Es8JxQV)%#+nd4$61jyOUpW%9h1CWwlQa
zL8>M2)Wd2)TLf3SM~6OnRYCe&qvl-}i2L*MGNngE(z=68UTxClJjzwayl7Ebe7&J&
znXX#&vLZYzBtiGun!>49xKVn%X$89mWxBS%CPk~R(u?g6N2@~j2AgkiyQN4!#Y-e_
zP0jr>=Z3{P+YMC53fuMQ`8ssZ%#fhd>Jqlj(RhI4a8to3N{^UZoH@5b0)G=rPkqwk
zGpz__7NPX3wXGy1-a<74Qt*x_SOKw^=BZ)(Sa_kIZf-N9+U@!)S^1;(vldFl2;AVo
zcCBJRkvHmZzeTBU_szK6&^DV);qsij1W3u;7F|Re?JJ<=ZLS(0!fd3dF@J?oq>|f>
zGLiS~o>zlJJ>l<&T{H#pSf+F0mV2Hu?cyO3Jd~u&k(_PxuwR&*GEeuQenb2!M(;K-
zd6Z*?XmDE%`wPjmxHe~|4>z2nUB7d&{SivmFYe5(ykVDDKaziJ(A^h>n77gmh`;3T
z@DAvXg58pXUDo$ErfTa8&4ev5edIM$yN1O5DTEC<Xot1Rkt&m{0>~^t9Vo;0Mk9@+
zZ-~de<b4C`G(!ewYp5vho8#s$Gjgrok!M!?J$-i9g7~Tq#^pBRetXQ78^qe#^l-nG
z_KAu#Z(4tt-ss9_<Mdam^}Ex4hqDV#<ZW?TO@W|CrC&uk*Ea;BA2z!)gf#gmJ-uyl
zQ}?Fu@DyJofGg~Fmm2Qm9y;9bj)ggRv6XDfFHzZC3<%dB*A!a74aVWgO&=Jb5v5Vw
zdeLWVV)<9QH#Lz7wnfKXL$CE=@$8tre*JQJp|?>#LKj~+qp<|nRnn&2E`%J5VO;VN
zO(e9ArZ35<Q>X)#a5J^8GZ_bsrfXiLSQfE8Tc$I#K()HA)%B-Cz3y>1e(ma35$7AV
zMw2f(k_;$UiM|13_$BGfavSluTXW}0Z!}t?w0ye{S+rspKcY`G2X^~`Ac94`o6Swj
zboD^}xcZ^N!YBOKdd_9kpG3#%T{s+AV*S4SkObkoEo%N?d&OgR>>6RI8|)hBt?Pco
z2}2UEtLG9GQ{Jl>&TKuQ?j`0ez2U+JCA^{b)}KteyuZoPpt*GuKcq3K>xh}jl0Zcn
z0ovTR+vft~!?c)P7S(INN)AC%t^*nai4^*;{0cjdov#bVKa<YC{T|%#@$t?Auftof
z&@fzBGd+Cv9jo?+7PW6qovj+STNa($3bf!^^rIg-JtcKST?N$TaLgUU%;v#kerj{v
z|4F&qt|Un(;jy>9u3yzut^4W`E>g7_hN$xbIiEJ~DLxE!=MSwPquuQ4i|DbG!5Ck-
zLp;dx9%~d0@x89Yyj%)9xT8srQKCJ2he4L;^j$_G5Z(PGV~TQ~*n%SKkhp@khrq_S
zuSulpQUt@&Q~Oq#*s%V%FM-1*?}#s$n!xPrzzw@zVA5*UXdRbgK_0YgD;h1xVvWCs
zO%(M<;@%|yphiq_f(r`{JtD#@dn+-m4PBGEZcu5;+O~nq=-bOUUTkdX{9l3<BDOM}
zm6-I0TEaje+LI8WhBM^;Zq4R!h8Bid{sAp4xeu)~M<?HP|JLw+`C8HH*wYZvc{_F}
zxz!HlEfZm$6;(|I-7GqNE((RV_n4y%RG1$@(?dyl7fB2+{fKXI$+y9zv+l)3*8w{w
zyn_d-83}(Zw1^*rlWRomk&^O3rsnC02gyrx;%Vz&l50uQFAxuTbz!l^eh}waO1F2|
z)W__!4IQ<%%SPN9g>8u}f><DUIxqQDFIy0N*U2DZL#vy1Hi_ug!A59u)q%v{T1Hyo
zxy=8|u-TA9d{qHoUtAgBGCRH$bE^?}Oy7GT<6Kt?7WE+{&kc5|haz0$vPPwSOZ?>}
zTLe91Y@y&|9nztBl}qO|6hg?9e1#)ZNrUIGfD&QYgIA`C6EONhrGqc!AuL0?MJM~v
zqRC|JBCPc=pZ6hgr7AeOEjKo${U_Mv=9Xykgjvb}J=j#QE*NB2@4CUbUqgZ?0A}u?
z+};@v`eAgkUyd&>2nEw0t~+SG2y9R1ow7yAheH@A^IepHF2b13i6+}6cGTRhR7uB(
zy|A~14k6{7xjv0oBwy9T+8xzEP!6~?PDp+fAY5`k5KdnU=b}LP4Q?yvHbbIVL}R#M
z)LRaP`p<k`;a&3*#U9~IKXjzQvn_`HG&O;U@=#163CtBvN|*z2lD+gDp+B@V-9Ob3
z2a>z66^12L>zC=CQBC({3RkgMN$ZKzcX#Rk;NdF1A@r-v%mTi#oUv<aV4Ssm+H1!5
zTyyT65jHf4x0oS7{q^k$Ymm_2_{vj~=V?GqLY74O)b)*7$5zsMzM+<Lp0hN3J|S$M
zs$XE#woRGWP}*DM-23x-ssi|vN%f0p?*Oqc5`$WXGT><4EK%MD=u?y=?^txuI4V8j
zgr_+kaxTAH?@w;u?)B<MW{38k7kIdNKURXN4jfbNcO%NaCcfMc=Ry-pjBe{8Kay}e
z*hZXWsvy^uphn&(^3=VWr33Xf_uMjNEG|&;1e;3@5@SqKiTl&oo|S8|;rWGk+aXa#
z<jhQB0f1(<F))x6Gg&l*illQ$WPGy4`C=Y3s}0uiO`$y%V?gkWSOuYCU6EwzOv=n-
zv(6@+mZ}*`PIMsC_=rP8Qv-!yP{JUJZ+->U+ac6Njg}Fvf5<N!9<xGVMXqm45i}d(
zw?HnwomJejdW1#a(uF@~A6!EX!gUi|<mj^Otbj7%<rfDtU&a~nWPN!I!%yEC$aE5v
z=Q-4~>~bv+??s!`bk6vD+Gh(T^i5YUIWo-zk6D~nXO(Ka87ebQIazU&S)<U_%b^{W
z?K<K-aJPN<=L@nF^v$p;!@Ihx6N5u;cL<O9nK4|v8ciX}bN~BtCR@a8@vv~MyQs8r
zK78MRm~N5VA2Hlm9f5E;2Jzvq#;qscWQ~8o58~4RSPq0ab3L*kk~NBNwnMD01-zd7
z=#=10G`SAbl|C-gb_vnZo1Jkaq)#;6sPvmfhY1HxF%<TnWKw9JGFzPErdigFVbk7v
zQ14yjRS$j|9m^llUXN;h8v7j8LCWPDfJUi=e9A$GC727lR_<|vh7}x+3p?JrmD<l2
zz3TTfK<meb5@kf$Mh1^p4)IUn4yvwY4t7l&5sT=_`dB$p#qC$r$tjHjDUBV$3nBqx
znGIzD8>7S#G8&z*{%^`PyS93Ew#<uX!Qz!l$-R2zR3KcTtSYi{Nli^H(jUS?U{pcL
z^?Hf6(T#j2!HN<Ie#nN&t6qUcCQ5J-Pc6YV5;D7AAd8tBKvR{AjT_5b!6%*&P)r9M
zO4~|f6w#wGh{U^W++lk?dJt!%swmJ=xy_v0Lm|w`b-reps(ZI?M|nazmpo&&-22VE
zRz}4v@g!@NNA<D^;cyKN(;O{$;UXiK#<2U-2SuVmw%3U~KRt2iSh4?m+L8K{Ns}tW
zWA~68Cz`PjDKLSmg&{XmO!$~vR$pP?2^>Fc&2*-+Dm=Dvgc6cceQ(K)V@DPy_m!}b
zJ}39`AM_DMUQ+BzSN|{r6RA;<B2MWS2<1=6?ZWN8(jq?YpQUdkaVm|byE#gG&LSlb
z%(g~9$tkP$IeWAs6Z3@WjU59vH2Qh_n^@K=%TD;b_@JL|wkM~A9CA-Op@giB)y1VM
zwCGSu>qxlgvBH%KoAL`K)j`Bo5#;6X4P`cMK4FGzk?G8Ldvso>E;njfD!)nTUBT=Q
zZ<0U$kf)q(r{K#PdFxdKi$)S&-tSBqP?Bp9k3O5)kB^iBhI&1${dQ?4M+S)=R1et=
zZ}@SH9IRhuWc8Bt0`UssT%ZiVtJ5|^i5~bnU2lA#Ti8oO0-|AIxA&B`QqlOom@K45
zBhrbXtywMAKqfb=;$Ru%!k!$Ln3Ia<J*m?GA#S<($y-4w@!+M3)G&SF6l>=^Z=ke~
z(Iu*Vrd?T97K!3hvV<0&;Ywc%Z8RINYfXuf6rIIlrN;;wB7Um`WVw`pipG|CS<Dt2
zE}<;BR76I(l-5l6MeJ4ZAtxFTfk)bwAp`r&_s9(mX;{7FLP5+9k79>>*R?Ha;p9GA
zRBc7o=>H*sexYHaxPihFlDj&&+z&}Gl;+H})RiZkSP}k~v9!%lq0#~iF+y2!x!H9i
ztRRL0nu`Z|s+O>2Q<qG~Z>Tk%Acdi;s@bgw|FQTz<wYhGCL;Y^-%dDHLHU3~`E4{J
zUaapp*Cvc(7F1arWEaX<&NV@!f@;f;fwk3#yDNY|`~E7pN`*3GZzG?N7D6=SzW@8C
zeJBXI$?Hil1t!1m>**Gq4pCW2Jxu;k+y1NSz*u9|*^qz;!n_udn9J@|@fP|gCeh0>
zn@Bv~P~uGC3JKJY<x;F8QhLiRXYRA;Z}^>fhQVSzgqsg!W*<5O9mK2N#6VQn-7H>I
zB_;9XlQW8j;j1YLEhzmb{D6bR<u_ulCj5HkF7TqSZj;=X{qPaW{)^jQz}+VS64vWE
zq(HM~kQT^uZUMYO3Z&a;xn3=C9|Af_Uk0I}B1Fww2JW?c49}J#A0!d|CiA(w0IF0A
zk3pj=GXzM=I8==ITQt2bPoL@mK>G%W%qD#x%ekut?`2u5iWiGOeD(IO1M#}9`ynl$
zn1ZL<#w!DT&FtX+&RtNRYGDA9qR1c<3q~H(yhUgRsW1k~MTMw+M~;vl(qTUkNnhaN
zId%=@e|&&{sh=8+_P2QkKxtd89H0?mlzzQdYCXFUy|!7~l|J%6=p^&-fIz=iV0|aw
zb$4stj`sJACAuw%T%dhjI_4p8A>nffr`1y!?pQXd;}6r7(+@2fUv*j?NJ7?ok=b<%
zSrE9t@OwN8l~{p(*WyU4Fkq*A14N=$>QA^H`P}e;Y)kYoc}(HE{98lWtDK3FqTDl$
zn?Y^Fqp|CvNSSfWx5upbaO>UU%HNkXT#|bzjITcqY|cy;3~N)GOZGl0v3Sax=Rd(S
z`A{IJ9{5^Hi_H=i<tAs;JG1#ozq{Msxc;pPLjD2;So9*ryS#SDj9Sfp00v}d*s0dI
zistgP!0byP6Il%~q=;=$5DC-2sSGr%Y_~F>814i(+7<k+(TaP=Q>bx(TSCX3RQNU%
zJhwYpj1Q<mdX>_)Xr|Magp}3a;!#NM!M44E%5M^8_8=1!!8Vgt`uzWxKmmvqbRT&-
z(BYyJ;)Y2{`%kmXV00QGZw?oLLce@+z<<(g2+0ju&j!+|TL%r~D1sUik?qekrw1qE
zXjFD0@ieg*BktPzCu?!R#2o0Mn?vErEZLuyE*6kU0Z>p?npQ*?iq#his5fR_u{XlU
z^$t5PV*&n{{D$=a6n>XGBGXDNR-8N*O)()y3zCph;&vzDbCqTDuWM{1bL&y=OJqr?
zaoM`nBII*KF7sIVv>lCD7`Jy~z2p&W&Ur$nIm>)~esV66$i?_`=R?%40>uLFmGsQ9
z?tDjt^W!=P{Puvi)VH5x>XQ<Acbq4;!X-+8Vgh~RbUWD(-dTlf^0-0{urm=_oMizY
z)wYr{QjuPl5&VxAK=h~aIJb06L&|ubJobo(Ho3|6WM2wZwf6hZzlvGkZrh%;W8Ayk
zlPJ?C;E+fEBI3Zg_>L{UDS<Zn#kF4`B@kgFQq<g0e=d;MRK6K%y*H$=IH_bM0zi&r
z;f(~vCABPkY9}iwS>tuE^%Z$yJ?#B!>+9=LOXW&QsC=5Z5`m#<PQ-7*+_@+A(gVCo
z=?i11JfxHs(ag<_if1mH#XoBXoEwW(S{cY-V|Q4ukmTE*?K4!{(WEkkHH0Q>fM(|@
zR>l#64XA#Xw#}i0N_1IF@5v+JMypg(Uf3snNnBsHx#-bSW{g8IVOnmB(0J1S3BE)f
z3`eha-PLQ-?{DIA{RttP)fyO#*74P1Oq0ug4c}e6CGhIYTi3dT2D^fI$QnXZClap6
zk8HPOdQBi#fmY|O8hG-B%lpFCgYoQ!y_WcX^e={A4_#a1=LT!^6ID%u6FT!UyuP^B
z`vS9-K&&Fl6K(6b%L)B0Y*Qc>s+?oVY$lQTO(K67vGHAGYQDZu{dzq>7C~R%wsc<m
zlspWrOfQycw)z3cx%28sBc3)M^2RP#iKYF`kXyKT*gO8BS<<H#`;)JsuuM9w?T6nq
z9Ei!YgZo&myB!bJS@6+1f;gs1G``jK=z^LYaRM(*bEheD^2>yL>-YsqC1XC&Xe+7e
z_irwzmR8K1&h0KX(PoQAgagb!%s>^uH@rJGI`C4$!A6S%Bn;&lu^v+L$m)x?Bw4VK
z=p_2bp+~r!U%rLXZyWt5^$d#Su=Sly=5lKODI3Gir<gC{uO?lqZ}F5!>o})g1QV6{
z)p22s)+Sdn1((2yume!almYB!K3XU0{PR9wSoKn{L0NLNMbgO2H=vrgLzU>C2kHwr
zf9&Uk#S|bfKPkT{4sshLe~G-+R_C!N!Am+o)*?B>VyajauKuo1&?!&V<YAx5S1_fa
zc@3bI8e(~_Tmo1xTsLNna(*E+O=i4nf&>H}Oq`qz??TtnYD}rkBKIAi39!o=bDXYT
z70p%1Dv<JS7?)1JTT*u+b*}^Gt%bp2aH=3yeb>?Us(l)nh!*}Obdu@XTGeVa#$5fK
zxyKR@r+s=6VQaVm#x4t|?3cHN8a*^Zc{a0u;430=GSz8zK$41Grsl7;)|{iX+8l`?
zj`$QmZcbiB_|6r1LhLh`1|+fMp72f|U>$uL{FB*uo%UYcx(xO1r#w`n2!B9tUtbB5
zD=w}5p|+jC;=uaZR9*BWqd_;P)bijJ);$^O^bXSsd;Q>t%vQk?LH}0F8?~;OZ3oZ5
zl4|t|E1XNhcZKWm<(CzX6)JL>x4R%GdZI!Vs@h{dfe01ha_b{!09LjGsq{QC90L>e
zu;1Tum%jvTU}LTAaiEwfQLzve3?;c<`Lte)T6DdSHaEyC>p)@2{-V;IO)G>8;}H@d
z4`hq?tX{#M^_<9K1l!1Bxxs7IrNO8y*&cSy$(8N|$ID(&BaseMtOmWy6u|z=)^2n?
zCq}k9eM^*A%RuZ?punfboQ!~tKGI=6FI7ETpjOWO=_Bp;4ymDO;jyohSPmD_<U0Fw
z9D7l`MLnz6DTC3%hX4_qSD08Zg@MrQc0EX|XMQa6DG@GFcdRcb-i-c;_jS(x2($vb
zj`5KWJ%tq3aH$F=IqYPaIz+%v!%1`2GS8eH8cu11u+|aHo?WvndV0j^ZQlTnJRn!J
zeP3R)%xZ$CvqalHIVC>OP;zH`JTCLgOlM(*KnN<#>4fi#Wlrer9dwc7>X{PBXDe+P
zFYR*MsD{wsQA6_*>K|(P*zW`DC{$ka&zuyw9nThq?XuaGn(<3E%4IgU-E#YVe!>;x
zEIC4J8UHPyN|08e!dCkR5K>2FwD`k8YQO}GlCFDitu0TO^Rebh2EtRFr)t7=X|<u!
z|LBIk*=K^0xPf5w?H<_6-1j6GZ5eKCX!!JHWS`;Hp59wP8?Cdw<HL8t+mHj#*1Kbp
zi@}y%r66GUY@?e8<Are+4F=SG>1%J0-O{I45YVTbJYYt@n~`bLc|Dv)&mMjUNSzi#
z&Bop2?dbG~SEze(P+|C32syf?a@qc{*ns>#@zh^Omo%JTN~?%`@9@e#$$G*9ha6K<
zK)Lf)2r2=PJzvp+nmo^^X_Yj51G--QHL~`DF#!2r-h_$(FAPQvBk05zg$g5ZA}-Jf
zEvZheImO{|y(I(`mkCG2A{;dOX9K~pu!-J6?AlA;fL8!7XBXw4h=Z{s_&1K@FMkto
zXfTi450I%Lf72&=whdsB@<9~LTcEyju>67z(bMkL?oW(w(h=;TTc)efst#9GtV%4G
zc~d|qGE!5G%0G-^I@>S3BulxwQI}%Q{x+aKAb_KN9gs8dxyGlaM73V`41>OvBCnT{
z1x=qpYM7-5mrtm6?rhP?1mGvHq9yB+m6A_yXVSWKFj0df1Qbu&Em!x~A^PuQ*!46W
z+N0|*&<}*m>Kn4Jg1J9^)dzv%%|~${8`<S&V#8FeHH>Hlnar#3_3h4CoZwup9-D`_
zz2MqArqv~FMii8i4924!B~Y32SaA?g(dhl<w(=my^7NqNnKB9VIs?b9vC%a;HNoH-
z@oOoH--4yjsL(TKS|NbGA+RnaQ4CXVh}tgXOTImnDZB9!5w}ucX6)+7?X#((8l=g9
z>OG)Kq(O+K($gn~kJ6JXc6D?tQi=Fd96ud|YsZmh>hCkD5q2?b^@YeClTLXudUI%8
z3X8z!pxC#s17?$=_x-<Oh&USix3#=JiN!ukxvC4Pp9)E@LY&J1f5E$~UxjfUmBLzp
zw1>2PMUbWh6()+|_T5=uyLd=zzi;iPyJ&{m3Htnw-T(wZ0Et87D7%0UGzJV>Li%D+
zz7$Wk5opqhdb|V&gp_eoGwp2bgMsl`VOz>L_`=}m7a981<S>pYVO@c+;e}ArSc6cu
zG=NNtbn94fU`=E>G^YrT{=tOKJ~2*ZI6cl_<AaboMN?#e6(a~m+F313PcI@7-S)j^
zT1r1JJfmJ+OD;E_CRK?ozBr9dUdyZN$fzsVsP5^sP<sSmV#F@3^2~Yc_pl>8^eq2x
z3sW7;oT%F6hr^^(gb>u2me%?)2{NA$ok0mSU&#?(c1_Ong7K3i8g+PMtHw8-n+zxe
zL|K=ceWc~qr&Xqs0^8P;1*cCm<hH1M`R>DueiLP|5JR7mL~9Uf$fl743Bp|nVqK`*
zXPIAKOA@Wo`}%hw#^{F?{*atU8}x4r&DpjXQJ1_`L}My0$${c`u(z*tU799_8@>fI
z{vLa|D11(Yz}JeaL(pMu$ZI;^SFB64fI*GVr~eIgKmV$W0vHxA<tr4$>4^Z5-*3*K
zL1<kI3?-|3>fc?UOiRL=FD)qYr#IpNZ*<htr}`Uz^4~Qpjt-<`67z>)zXzu0<Fbtr
z$N>*Kx}^T_u>ug0fGe{jrfLr7_l`a1U*1}P;c+$~B;ogC_phi&C<!ET6lRr7e{54D
zVCe)SV9@=Rk^U#G<BA0`Tk&yYoj<n85iqUl;ZUpm*VFdjuA<ve0wWQbJJj;WHXQ<a
zVh7^~Aea5aC$CBXBS9#c=JCfi&58ou(Mkq|{BaOto~@s;Z}0xtCU6+gokr$z^gj+F
z1>m!_p|rdd{ZpHC|94{l-{tea6Z=04@BdRf_A~&N$Mq*+*~&G!?2lbog0~;MwKMs=
ze1Nl?!Z_R%aI79a@q_}FfmOf{S5=_F*!T94>k&o8@|NgX6R%*Ti}t1hw)7&#?>iP1
z9`+YL3qdOhUjW~cd5(Js&r9RUf^$>NIy;07m>RL;Vzb8`yr`^dZ1PBevU$_@*74pG
z6R_LrpLgg9P8&f3r{W)+++U-;$w0Ys8Hkll@9xsR!eWtwB);mR!Q=49%Z6ZMfBme-
z&G63?&F3V{8y0)5GcW+eL7y^-&>z>u`SLWg68Sk<-~eKHe<cp!<HG=qkuXl2oSTC&
z$P5Pw;sCI}SbpYX43`)!7HwhWvRWq(j*qPXt73D1hfpztG5vBlq*sR64_pO>8dJu}
zQcdNP{DAJOY8YYH8ccej)He(N0w!ew0#ISf{TOR~Aqai(Z@JYu3%=wy13di}!DVN#
zw#>6M6{oufP&xB1veV9(NZ*r003wpvzTfy6J!pGqv<(emnb#ioChrxE)(R6vM3k>5
zH3PO%#L%sg@4nrQPA6k6E0}e4b*2R9!krW5#IChP5de;s2VcYAk+M7N{qhZXV<-C4
z5FV9Xj2O}7G+QJRxA+1z;Pj~RO7K-m`A4~1td9zq5`+2&sV-U?7uqeZ28!$q3$+dm
zAW~4u|E%)AVu%E(G8!{f*tM-^8vr4}4rez0woaYLki9In$F6|EgCyX)BMTfhz?X|1
zEMt_L>dW0Hv-?!k**!4sJK&n#Y5_S<%x!NM?qbPT?WgNA3aRPi0G_$=@4B|@hgem|
zM-f;PJTY8?7A0&B?PBST&X93|H95Ly-yM@TUT;+S1{8RpGaC>1A{A)9;~&qH@;yvZ
zY~SfN|B&8@47g_n0Y4}jn;0gryl@L+Un{$oxVmg?*3(hPCP-}w10K+e9U}tI&UKFW
zz=-qyG?e!9%-AQ>BptIanR0ttgA9~Y&)a~<nI`(^#TcAdJz%;}Q<}n5RN{K|=n+n~
zEN%1nli}GpN|dT1_;nHQ!ck*}u+?&w;vaE@E3>UPs;zM_$;YXs{rEm7?LDwpyfrra
z+FPzP>f2H2KYgV4jjeX;D6+N(NJPOBS^!1(^l(T~cx}m2x9eYQdnC7*VmQ(h!Ktd^
zOt4E&?uQ(g@W@xlJ7N}$!vSo(_=O#`%uFLvO33M2$9KR7p^G_izPCuewS5Ew05=*(
zMiSBXb&WY5*>|$}So96ZnWcS?^`9}eLU(!+2(FlO?_S5{3^hIqY_Z+|pp4*I*QDb-
z?r@VQDA8%|)&-1>h>{pRV_907T9fEPzCWt=*x-OnbTw#Kf!XFz=>C@hIeX!aqz!Xu
ziK}_dVP{z@N$2b%!&JkwNAg{s1)fs-G{^a+<rKz5VO_cY7l%%GF8jxKKtp_{*zU+&
zC;L9B7#*d?9Wx;R69Wmwd#{M&To3*7U5Mdrx0+$H(sw^(hB~(8Q+w&(qYsPyQ{f|D
z6`$7`I#H_f7~6O%VC`XYMD+6Q&gOVuu6HL`-Nej94yv+NlpN91j~Q!K;svnngtAe=
z6?)T|N4{_OxI-h0%j2sO1MM#IW|4LaVBc<t*{`OjEB@t44t4(AnFlRA<}}IUd|_!k
zbs_xoHZrqj{8#i=(IlD%HqwM}i;G&_ttyZE{f8+*8n3MS;wy%f#f4+n><^*cV&(U0
zzcO#=oK{mtq(57zi++?gpZW0~tWx%zQnSu-Su^WDpvqXAop9p;UTZR);a@6cSOanW
zJ_`X_|A0AdGtWgYZ{6(Mg@)f--hP;58_4h?*@Mxx-C?c;&)V-o-G-+ZVPnToh>rG<
z@VEv7CMffVbT!2syNQ>5S!{mlnz*8#Nn{RoSoD<IVt^wE{8yR4?Fwm@nAWSMd)E<Z
z)7?25XoQz^r9?60ft_0iPUWMGud~%dK9YS}w$D;cH3uz9gt~ryw&qS%kbkW3e*>6b
zZ!IOiBm=(NY`}irVCD4W5&mfiW^MJAwby}ZQRbUCzlOKoxp1)+&r~f+J?QgngYmLY
zwX2qf>u%Uu+btYESS++48`zx!F->a4+UIX{xY!SchkgEW&iFpVT+jB7krRqfRrtPu
z4cY7-uG|rfNv{pq%H}PD$fO%crUo>Q@%YUnddfj>tIINT0I!Rtctdq@WHFx7KAr@t
zFTg_RG`W}TDPr_QeG}xI$or7(F@VRQ(Gy<<hP6|z^ksmMr33MJok?xT&}%m$02LnM
z=Kvom8S4@Q!qR#*)JA?n98AJV{VdOYJ8&$lBlGnya4}P)?&(_|THWIy;bwjEANY1|
zhB595P;}+y-KJhIxh$su#|(+Fiv#O7nE9t2GkVZN(KZ)$B@M6?dM!${%W62rD)MV@
zUGPM^a6`(;u!*Ki;c)&gwZ25ZCI;|0tUCCbrrT?2xqHBUZ9ikSN;kMC9T<b$YJff$
z(B$dj(3Z-pq8V+}VxW}4RP{>y@|2TT+6ELA{4z1qS=ZxnT{n<2r!`?R+qmA7&U+b(
z<vJ89I{D4B<B(&v$!7KnEXv)BO1%o4t1NQmRRtj0GO>EB8PpIQvaQ>aJJWef!obhg
z!A)}$(4nG$8=UQI@Scp;T|#!%^<$X%(!v#SN6r?i7dFyQw|+UGW6hToV~rk`hSyO6
z-P0?NlQ2#ePbviqR{m4w+kTZQBb5fc&x`;-!7xKN;k~t`iKU7R1M=D$*AlA)xpTdL
zRN48A%6j{DCu!)RQFo2M*Y~&aaU`F{Jy~w8{dB{~8PeTYg~1`!CBR1OSwB{}H7eC+
z293ZX9xG8DYb@>`3NR{P75AFnXs{<4|1h0eASl^SxWc*bwu|Xn&$g*%1aU86z&yV5
zE8;x)s=A=Ae$>~GfXnUT;{Ae+?(K28dRZ@6?M)^>-Ls3|@`qXtR#fKg7x4ppu)~nM
z=JbAuT=I1sU?0>e5zP$-DIM+jzENN<nlh(CFhVGCi#z{k+TrzF5+@wpb|u;d>^Fc?
zl`k4%u1@Cy6pB})nANiSY7&l-*t|XF%md$z5Rz4-vq$;d?={P$A9Yxq)B%Mn7%v{`
zq{-tX=2m5hU`EDI{gQx}geJC2I4uy1Lmtp`KE6qT0r<}<b-5oYiPqOh7_X|Q3+f`?
zxwTgsXc6U?m;)tBBH7pFF}wz}X%|g-CCdEejuCXicH~X~lfQr+B_A+8U43FcXjk)h
zShqq4Aib4?0sCd(r+iF$5eq<3cxSbIqYBvokSTc3_9o+<O(22bg+QYe)?!|AQN*ZW
z$ozJIYiXu0l(};%HTl^(U$=n@Kqvp#Bif`+)9m=})KZ(l`_@$CVnw%G7kuUyhMEQ@
z#XU8=;->(og|s?XqbJIdQQ;t?$$RTmTIMnT{p!~Z0wR7bZ0rrCzW-@-9X2|lcxZ|F
z%E9(lljMJJ$zT7rFg?qsWf$8PlFtKuZj{penQmW&;h2dzufzyLunjO^_#<tU`WVrN
z)ulF_U3?U5djlBdyQ7aI*uBFQAc{;hG0af_Yrj9-2Jsh<LGxJb0)XtFZ7@x4y{(x?
zZDl$mVyg&9cG5q<uVwp3)b}?AzFX*g-_D8>HQIXu0b9p)Y}xDD8;G~5;~9aUy@2T%
z?P~<&D&)PqIpTl3*IgA@AnvMzL;bNYTHO9!b-B7g3dpGopx2SHP(q<Epu?ic`yGGF
zr(50tQ>BB+MDY6mS}qA2{`}TC98VJd_2W+P^=bR?r6wQwF^VTFPS40N0Vo1WDDe4D
z*7{U_|8!>H)5#!bQA$CzRD`mbj^#z|T(;f~Qyyr%70vKL7sejiO+tXG1|h?&JELo%
z+;FuDr!Nu~Zll-fZl#1=|D-`sCj15TS*U&8T;3$>?~4rhxt&^|oqV<rqFE3sZ0n$N
zZI|T^0GVgr#z>D&?Lw^FfO(b(u%-1*rCqTb7_<=XjTWkXyhr5?jkH~`_R<5$n7lJ%
zsWb?n4oP&M1Y=|PLA+pDAKY5M`6Ct|sQJ3S)5;mn8T;Z22iWB~3<c)K27sE6Xf8p*
z#tL=y=?DCcY6B_RkKK&zLW(xT+_o>>P_q05S{$yv%90z0zmdq~B8J}{4!-A}pa8Io
zEC)Ao;6K8Z1nDo-FS=-~=6j~?fTCnM3oQAj+qb<@k{ANRNtKCK?fsRi%q}1|-or56
za{DLS@Vyy`=g|4F^{~4~2gRUwJf`kywHzKBaH5+)tU}<_)z{xb{=Dy|6A4A-YdaAR
zR9`J&KYEAl*T^Z-;au)}*i&&lgKHh__qhUI5FT4xR=cgQfAHfJZ8RMuYp!e{h+jk)
z*U9ObYE~^>={|<^+L$`v+#PFpMVmMPqpErr|L(Ox``SNLd!Rs<<$0Pbxgs(Da-?)j
zHR#^<MWPnL<upB=zN}V9QKCcOHc3UmR@_-`jZ&z-Aq6-?R1&`HLS0)Mzgn3WncC;_
z)fR$=gURSWPY}&w4Q>KgPS>{xlo`@UTq?eG`_Gx5?e^OP<{O-n{iK=8T&u)iH07Tr
zw1H6N2f;m&V@`)65`y^>C~)a6yNmqlmG7F@a!^CyV+*OFH}9(h-nh!y4@4((=b}%W
z$9X-h*;5tJy1Mv{KqBiM-aLDOcQKx#{_1SN%Y9`0+EMo98SCW^w_)|eyN&pUAIc0G
zFF5G1$M}-07yV|+JZujqMG1Kw(11NDwvqZRhSo)h$8$f&{+;t?@pOVqjTwl=@NFnY
zmA(@A)a3^M+rclPMzI9>uZF}TA%rZWWP{X;Ky-&bocqdXXZY9;m)Aq~+j#m@KOYLl
z97HrA&1PF1hlMD>Gq!=+voy2<1suYF<#cRpG$)?FT}V@6+G;7v41rVcE*1^I@F7~h
zqD9)vrn?I0E|vLy_jhX>Cb$c&RLd=HQPa(@$z3NUcBg8vv&BM!0Y{tO^zQIJ3#mhQ
zsPKM^`^76j#uAGB!2eZvf_BzpmFw*x*XGCrrOvlV5N{=awrGeykbdYtULOP4CJ&yF
zbtS3Hqk0J-xeMD!OX<c-jkM0skT0}-?@%%UDM)Id04ONgov&zUsQsbcfTjP9{1rrf
z<+tvRkFD!|StPerWf3XTZ#VYNlYoLaa_uJ95(u&Rcu67S94<k+x6lX*&6zwNKC>)2
za~#SF0n5!^Q3WdgNR9Y<|6)gB{r@_luz~dl)C>lFT@EXESp&cD!;$tk;bZzJf{GBv
z!lpvuv7cM_`J!f;Xt5bhWn=aJf4l${CLfGK_0V3cd{!c;ro-Y;BgJj4!lZxJ#}eTc
z-ytW5X2rf1S`f=BMtKI$9f9u2Nq3<*eGpMU11{zPHeVMgwV{;){8;E+5pZZsDnq+C
zZB!yqp@TCgrVB9Ba-*a<i4GBG+rz2Y(BD6I#OobR;R}Zb6ZOjJ^Z-^rW$^XJi5K%=
zQKj~KuN&$H#MIIc5$4Sc(?X%SXTrvQTnco84gSG5LlZkpsBB`&6JXKmx$3979`Kv^
z*BFaS_jE?G?EJ_9znv}df1NFWgSuTvpH#}GACL)2i$yPy<;rd(GAW5ur)hJ1yT8k{
z76r7o=a|?4iS!TLImu2W0RAS)>={6yGJMK?!Gg&E9R|R*qQ;PT?M{;-I#HnD=VE6q
zE^#d%Geo{wQxFRakR`KczTp~R<HWDRV?S4;Hdw+;>I<){6Y_h&0kCb<1%yGbNG%=w
zoNP29+rq?<m%fSHv22U&Z~hpMdH`pe6=Qd|EX;n$j{iO>YTTKWyp&$^f$pcVT-fkX
z5t+jMELo*O`j-S!=&-n*(p?CAf)B-vKs{VOdEhlf2!t_S#uVIk_rJCg>tEYQ5ZFez
zrKons@jr;gTrTEuh#6pVOw!bW!OP)Kiso#wRse8S)pu2G=vTsESlYk<z@=IjgzGC>
z4xt9PhAsu5!^Q(Da|XPMSt!@KuGP$y>7VN<i)ZV^4QwQ01X7IlFI5f@o{O8%dDM~H
zfik{m1_Ro-Q?z(pGKVnN&z;ZtCZMCuM}b^|gW1eX-(%4OJJKXnMvoDM%VVTzpJ+K{
zh{t+7s_2VJ--Wq##V?y862c$(<Ojeay$Rp1pVSocv`++NbcGN|e;20PxSuh9ii1tA
zE*V847Xo19>Du*Vh<PVU&g1?Qkqz5gWhTmHqrqGSW$;Uy{v5Mo&!^(#?Lu-NUO`U4
z!HXJWI3FBo<76+YQ|6Lt3TR2V)7a)FD3i_v(laML@izoAzM_Dt#y&05u8B7Vo$DE-
z06?2M^|%k|mB;`>WLE<v&tNfbZzLpR0sbt(JL-hhgFECBH@6&%a!u4391Y)Ypa?jI
z*^{0(O}Zoer2#h7z9$Cj``)vXSxoxyh%JCs0hY&?oJN-&9iW)L19R;oJu~|2;1V>c
zpi9%kYdo;{?_8bl^OjdL7DvF=|Hj0KVE*=GqB1vLsGcp`TpC5GNkqC`2*SE%ms`*P
z1bBEjMWDjgp)0-`ezlly{U$B(%QVJjvrDu&@wQh4cdoIJIwQJVdin=L<I{6(FvmpS
z-nlA^gELGu7*zAf`60Dd`-#`df-^k5g3t{)W3(Y*aI^E)5@u>jSX1c70b}wgq!!Ew
zGc;U6@f*J$rk?JiS`J8)8*~?CSU8-w;p(!ctD~>5KD+DNj(^EXYVel`jFx%-H0qs=
z?p_IKfNvB|CkTaw%K`tx<ATh4lntQVn;oJs`Mm1ry<|e(XMU=w8)^0lpmS!?v73@Y
zI?++1;J$Rof1pACSAVT`EFuO_a?~wQ1cUq5tT1nnaH$NsBM5<jHSRWCrEC1FXEl0X
z0teD^i$@SpRtIA$L(fdvVxT)idE=k8;#JlR6hs5$rO)tCMX8av-8M2o50PF+-2yL#
znNkmfEw7LU+&EPivw4ibn;pibuKKW<a?fL3C8<<9q}y!OM)gk+oqspk|0O`{{DnQq
z=m_P~k^D?Xj?9@XK%`3ZSHXm1BP|5J1ytBj<mg>t=a7L{pMFloaRECIsE80Y^t7Pd
zgoOeIMfJTWe`vMFrFqx5%Ee&XqA=yG__Ht1K>4+onuw5<v&FyE96mr?LbH3nwjut4
zk<pQJ7owhQGqg@UxZAgm$Nh+WPKxrutR_55F|>jU<z*uj!VVsz_-_qAZ3dX-v*uEs
zorh+0z?%~M6b^M}s+<$Q9bNqqxD6s|O5>7DDRmRT5OY@pap%;3ldpj5R<4Df(K#8|
zA`vP*-dm<!NF-RB!gynJ%>EO(5|<n`r4%ZGt-jdIB%9qsPli`HV5zwVtkAcOaDMa2
zcD?WR(z)q>TlCh^z@oRd8RIfUzea>`toN5dUi_GLxz;r|m{&?E69y$6Lr$Yfn=l-N
zj|Kle0A<|dbhYF+%q%OlS_u{AEV(-4LaAxcb)99r1K+yp?|z+iyD9_x?IYYQU@@-+
zcq+sh>!+|@Irr1crT5RES<F!e0JQb7|EJ{WKR>kLP@Mwh4tJlqSxN-8ioF~dW(~rR
zv5_GSjD_=TV*EUqte;0?1rACYe7(}RmhoGIC{JCW_*B8Tn)&e5H|cPfr237bZ3EZc
zF&yrbx+@3aoYx-r!6;Y9FHctHIGUmu8DTMA6KkfE<^*X$8h?`)_VyJmP?M7N|3~x_
z;`|Gdws_hEtW>1f@K*GezIs5XhDp8fcAI?TI92<xe+K}`k(e(T+a(HVk0kLYRm<Ej
zv{h%k_dr>#&H+)LBxpeV8zIi2z`vgEIe?}1xex@BkQkHJ-`d{}Xp!Wgx?&ZTc734K
z1=*sqNU*hw=B#{T^hphT4UpB_rL&PscTpR_%^uOt!GMWPIN$gq#}Ds#%hrbgzz1gV
zp%weF=huEUQ_m03x0wj8P>=VMf`A?FVJfXbmty(4MKlR}>L^u7Lb(Bec%hW&)MN$v
zYFFmroCPfJHXPuwhwEqwG5&EHL<_w5>o`#2QU@nD>JtDpfKX*ELuUr#P{17#$bZW%
zL|{p#zw4Ik^$Hd*xW0%BtP+jyLnDLn!l@&bE#7_94F;W^QB*OACzv64308F30vbjs
zX=7FK9Y+trCTRv(2n>bL4{x_OPXvH*3|(5_dISMnC(hUD<Uzm5v#9ug1*(sqb9K*-
zA7C&`r=cgr9Yp{t>H`KUPLz5{P+>wqI#r1p=$SbIZfj@*ZZNTs^}Olfb$)t;-2cdc
zd~-FJRAzg9%@6RxsWmHkdZOVOE%3leBofn<^2?<Bc(kExc@?l0ae3axMwSB(`c%+5
z;6|$2Wg#vy{>4@gc9ziySyY=4KvxXUC|57x%YJGl2n}a44le+%9iWm;Zw~E?>Bxx@
zrvdDeBmhZ`AP>+8=6PoWU5UjD-?E3plFAe)b+A8MDB=P3lJ<UkRj}C($-Ap1m`AYz
zLpb?-I|ZP02~NMB)n)nB;X4~bChZo%+et~525b9frYg)v=r9c@l5C@qGv}X&0gRzW
zYMdsd*+9CcUS~u6+tQL10~U85eJ}-dSP|!Ow$)wCY@@NJH_D>jbrb>JJABvp3R=*;
z{T_AdXQCV-5rCvq#)K^f!1hNrwNO;DC+wog=NcZDSm#}q&)CCS4L*QwL2i(`iTl<T
zZvb=12k_6P>_=W&_)x-a==Q7+J^P}tKkWJhMP=nE#y!rqTKUpgiIuHYDo|P*V-Ba`
zWop~U@SD7fr;pcSc<=79vbNig02q0(hIQ%#E;HlG-L_1t@4?0Iy}FSnoNEssG1!tU
z#AXtwVt8!b!?UXK&(^y{fE!V+4^GZ^xpZlH25<t!Z+yAzev)sOEk*)0;%iYGPr3v5
zNVxo9jt+m@ULcYJ<kR=)IZD~_&?VpQ^sR--9aH5MV+l<DnmEKPz#jjPL-JbtuQit7
zn5%2MyUUGup&dSi5Co0eR3{0P{Dx>-nyX>aZ&H_wMK|cbRLGZN0nAQ(-q(6Srq$?x
z0#w9L^Wl^}ek{S*jyfsjFI#RxF=5o~^d7$W2(J5l^j7TBEN{kY$-^7JBv6R2ud$&{
z9g7|qNdWc3?+jtKQEe_YUOC_YfRSkkZ!c&WiXn5Tc;PVAy8zTr=gP+5Mwll~vjm@x
zHe87uYVr+Gg6<nEFai3jtX+I?^2qOIqjBx{p%*ZYGXe~%KcH|yy^l8QJIrHRc<wxK
zX-^ELAR+sAh5f<E;F7-=5j<y*0;E91@&8c+072We0S5Mc9(*^I2)Nbn&y;gNMz83^
zypbhPiBmkHT^IRifc-^z?a)-AHk}_g(YHIT)_SZ|fvTiZe}h=F#S;f8xSG!De%J({
z=yR1UE$F(*JpusZ`y95FeLD4Vp#&`1qHv_3U$}1(x_=r8>iPzR%XP&71qY%7n{Xd3
zlHX=i00Nbs68M=?QCHmrJXO7<R%GAye<{Z709Vx{GrKVW*Z)|iTGzkwV9}E&$d$@O
zO>=tqdL}?F3LAwiQ&i9>WlHOafDQ3DUvjs*QmAd3%D?k_kdF9Ima_$9IqSml#Y-3f
zZ$oO5IGG10afuR@=r@)lBZD~{*AH+sR?D{wS@;ur1K>*_z3k-eLZvTVYAP$IfdK9<
zNq=7@PSJdMJEVbGHb<0(jrj2_TPVkKwpzc6CfUR<G6&e%<v=DdAOrJ7D&-!iMuxiQ
z^AwqL^<b{?8YBgbu6cm7r>usi8~hWMGqqJtrz0MkKxUfEZb=gyokb<v7uhWSd<E3A
zEA(S+O7t9Wq-A%WuI<CL$H%l>9=AvIrJ!LEK%#ptK2y5IGpGXugTO`tKdX0+H?su)
zINja(|C(v__kd*!P1-#WDCE%p+?ptnL5@bu=aN@W{{K++)^SyCTidX7m%^fB5lRXu
zE!`>&!lGdT0)nKpfOMCn(jXz7(p>@qA`(hB(jg`F&9%?o=fHWM_xs-OpS|~Q{}yZA
zbIvix7}qtfF>Xe>Tp0dmLA4^9Pdg%6@1(;Q*4GQRL6<}KcU<UOZ4*$}0(D|)Fb)BW
z8*H3@QOb6R{~p}9w|*MdvUN*yXp=y!C5l-kCpMHE%&Y1Osnq&fUsL1NNosGr6u121
z2STV!!jCA!O6XJO=lEl?(bJJZAS+I0b1M+&)A6ltFp&tpzKf4ZIYtfo54tWelzF-e
zVU_?1`acFOsyy9G2Y}HAqbQ0}W(WXmz<w98fQrBao%iGv%r;pj6T-#8=e9of3H_H_
zWF)i(vCTJv{VOFE2TXnnt$f0;(k=b2eKfrfTNhqCW)_QYfIpOrQ_(8E4JZh|=Sbl%
zzpYl=j!hy}%{~$Hc)}M`von+J<cH05Ra15E*s$e-)4i%Bb3nJ=pGwRT3v|2r?MxoO
z-;>gl@p=*?zwzGuM4s>%qcemA1<cHO*0!+oFmX6{NLsP{XN@PdUb8*WT^5a$Kk3E?
zByhx2=sP*~jVDJ0CXH`bo;`@P{^<tWA<>FLO5K1Oq|gH3ffg~#ZEme~hoXFmY@{>?
z9fPEXg6P%ul(#<^g^;f+Z@~%-6zO>0FJ1p20?aAVKlIgo9G6S)_5f|y<*oSLtg?^-
zVi?f~1J)C`wy9_Zw{PE&Rbkq1pgVx%%Y+nLt(cTQ8~Qo1?4>*I+n;cFo$Z#N(?w(~
zrkef3z5%ECUAM)HZck0=ZKZ{$JY_18s>+X%oGO?tdu8#g1sRYs)N69CM0y5I`*?J3
zKcg<a59a_>rVuUp#^e<#9rB|*ov~_Ey5q;GtrZSBeoWhR-@HNkAej;|jq@9g?I(2y
z91QfScLk!@1AZ!QPs?FPWLXbBehxMHa4pE6<BmKovo>^%`>C`*RV{I(@Nw>7f%op9
ze$bXj<eFr6IFU?tODHyzjKXt}8U@HHzvc;uFb@)vf6rWEp!;c_pWUERJ^o=x*3VLP
z5q}&;x#DW~3b*U!cuduc@Q4>%?*$2TUrwDJa|gO|wt<;z8b9th6{&28yEhbDLEDmF
z-XP%KRD#|oq6ch@RAN+DD6adTLGIWtC!NGDI8VXyTl;%*!}AF=zj&%Atu}4&b>D~+
zsH_I>LvBu4mA62zxzO_gOVWB*A47Tk6XT8zqN`!KTSRf3K7r&mR<*;sbId#4c<L**
zo$-_|yuo2awK*Oe6wGmt?j)4VMo7sGy%2|lC{5-LF|z;O&fMcikJJxN_fqh`%PWS3
zFAG!`UwphSE|Sodh;Xp5YUxeh3O4|;L5*!;THd&DX9NjKF{NhfyMyszr(Im2tWXZ1
z(;b(@6{5UZTDpf|&X*vo*Wl4_5_?i2wgABO1U2^t_>!Xm;MMs(Vv=A1)VN*@M<WI)
zR59)UDpx19U%fPmkfn%A4rprZB+d<%<B$0A^WUZKMMMbyT@_D3EOY*bv8)$}(tpMW
zV|TrH)srC80os<J2yg-P2;)`@QjEv%@`7S!!))WL-KjAYWOCF_RoJ~DFHJ7{sNeI$
zkefjrSk%_)5HTo0@>zq?Vq;tIp%hP3PSdKXYaFLPXrR$;(qvY?>rr0uU)7Nj1pXT7
z?`ew_FT(;bQrPnnS=amQDawc7^q5-hq{I_@c$Pv8mHdUtf{}~OMxbU>BM}Q+U?8yp
zjJX+olNP&b+lr`XtGj?<881A!KY6pMkKB%yrXL73g0{)8Yp~>wZ(p8~jlL&nZ^N6p
z4CH?PhqdsKQLIGTQ!!oW-3&vU4rn^?iSc}o&^6!}jc)w>Cgpfe0~15Kf2Ata)y4R1
zAC>tnf7qxMsOkqXyYR{05DyMM0b0e%65JjcWWzx5Xy1V0;#$o{uf@V&dQ_^6WMQ4j
zptB`WR{y2Js<{&$Hz!<m*z7AV`~J4iiM<JYdh9~LYvlpDi<L#eT^q_RO0Did>Y!+b
z17)CsNjG@-tvPA+d;ILvxMx;}j2JuYrAmMnZ2p7*qOh_*T3`0p5e-5*P0vN2TOGPr
z7l$N>G?hD^GOIfgnJ{(-!d}4|?&{raznk+wRGd6uW`i-PmroJbW9rMO{9CU31&M5H
zDqe6)RWOYLjAE13mtr*VkK3r(&bF+vC;Qfvxl$KT-^Mlmaxw}i$dkKaB@%|=>yuOs
zN3`Q{f`PW*wmb;thaG~&!qDq_JqCpe57(<{LuR(D7&^S!*%@oL2R(0(WRZYTL&k^i
z1-Yyag&&Rig*#o<W7#=fEIqxu&Cz?irwfek0>)B<?O9A!5B))YEu^+;0olYtVWxQx
z<L)!e2@V*52_Fs5m<;?FLeppEL~3icMN|X+*YfcAd5vt$t<>8o4x`~->zYtE_pU5s
z(7`g{v4&bQ)0ijdkt44V855gmwTW{;FQEmjfZG}px;hm4O~zr3<L&J6usZF)sufUC
zosF$K*)IE>$3HS2_;UGzgI=cd{R=Tt=@4J-Vgokq&)#DYxgxx?8!4|(sRXEZk7o>u
zKAzhkD$^|!02BX&ey4|qNd5eS(|F1hxP7od@}inNH{s4gD82Rr#(|HW7184FIo!j?
zjpCm5L%5uFwzmV1Z%<cu$nz9lRMxA?-x_ENQbsWSwNA%SFs~^B(1(^P{h&RBLFTQi
z#Yh=_WJ+xYHwSCn7n8~~Q%vpVd42`5?-aJLXERmJIbEHL3Ky%KUUo$Kfi#^D!LL`Q
z1xp%FO@A%1H%k%lAgBJL*!HQ5foHa~*vb%&H*xpcS%7*H<zMKX{;lx*d4rbg#;UoN
zAN(qn)tXV-&Z)Xklr2VF9<kJDSG*C(W!4x5NNz`JZ?BlahPvO&Ff#Umft2=OsN7WJ
z0VhIzhp)(*ZM3c6o<xd+X+K%U{<faSDlqOni!`#iLA|q_pZSy40piDET`AG6$$&0x
zk2U9v2J`SDoDhwW077K_Rdaqh7HbbWq$wtks8l!bu>LAISFe;^D>&anQo@MPMeTh}
z&(2gL6XKiB&TOrZgSBFPvy~_%^)K!cDps4d>SqGGS}4%R%f+Wesa#x<6F%lbuoI6(
zmE<Cc+u|(dDk)!8pxey2z3Ie)Kl+hhM9}l2$EwB%nt-2oS>BvUslka*a|NEXBWf&8
zg%ASgJ`FcQ>F_f4g%|tX<KyYIYL$m7xW8sF#d73L3us^>i&-#_+WJ~L1+wE?Te=1w
zyKCexYN#WN!vaqWZ3`n3OK0d1vI)`7PbC2YuleL}xFjX%zaVKKCgkPS1xkOKVO3!p
zuCWQ+$rpZAX#}Nxx%w}%V*|tPR3MpCPUzJOO%gHbDEK(p<@ghLT^bnKl{?84w$wt9
zzCJ|HPCSr^24;?qMlGi6F5&9b+t^yC@WaG{9~Hm`hc#PPoLa_Q{W`jA+UNx$yq-GE
z{--qldjfQdW{lDNpG)KZUmPC&#eO8~G=v?VfvReL!&Om^uHF77i2}`-=mc%vRu0`e
zD3jF^=sr5ff^>NB4S$IuEuHyR6uU+7Od9CiwqA}sZ!#Y;qz7%L&~FJM5vY7}R>x?X
z<HL*9=1MD0z8tU=FbNDap0zLL4aLwYDMZZ9eMZCDM%7Ng!AQx|z4{BOY^=$*Eli&g
zFzXfL<VI}<D_E*0-KC?Aw5q1b)r25sOZua8XUTDQu{1hdlyvXKo_*NPXJND3{Y+l1
z!<oQ|10o?RKff0HY%NBq;n@O$>u_Js@?y5ptK?bej`SI8Y8=gyl0^G(#_}-`l@Jsq
z*q6*%zLkU(hTT<*u5jJDUo`71ex+ZmC-J>RL*t6W<_wX%9)YrZ(PA_z)?r~h*T7?c
zvR>@EuMHSeQ&V3X&jb`FxNN)d5gJAydDYUo0!^9F7fZqUTE(bONcu{UBfl4`JSYv6
z#{63Z31o>@^w?I`jl#(oLULHyWzHL3h+3_r+k-5bX@?hWw!PW5K|;BEp`&51t<wcu
zL?XfTP(%-@W$^<<AMT*?TU~j~!1$KUQRjZ*ULVL|-!I`0msz|mtPEJxP$)H!u~N_{
zkJ_|J^7Vp}44Ag-MeUEPz>Z-GU-=~~WvXF)lMw{U1opqE0gica0P7JA`IBCZM<wK>
zty_|)A+_aM*_m=IAT;SrGc$gF6L*kko3*<8<L*ZAEPk4JZdLE0-7HA>Nob{S#SkLH
z+n~Nu+t!xlcB7>ybE`o!TQBBMe)X^MyVE2nANVI1hNbBb0EoGeVK90--=KsJNR~Ca
ze0=mN>EI`JKcmch+$qA!z^J*kZ&?(ek{GJ0uGPBZmp=B8vvjsmX7BxUxo0mXXq4V)
zVY+L4c4e}2KO?)Ag6~#K10?|}7NKtgMxKuva~C#oVpf0(6*jR8jtz%IsFYn0I!j4c
zj;k)Sjvlgf$d1c)PH*BZ?dQ(?^*bpC=~sSrE-N$UfiI5+;v*yM_x^BzMjkbTr~Vu@
zJ0jhiW|L(d9=@Bg%(|n~is%AbSIzf!xa$ocox1G(u-1KWl<i5}rkIPX<Zuw$doY#>
zYnyribDNdQ9`3Hv+Qj_164dKZpwu5Yx<rRYq-L$hrDj71buN4TXMqIW*$SHX&GoWT
z$HpJ(yhWU=|MPn+UsPJNe^Y8PZpasVW_))1)LLfJloY}GN1?=@S$3#tXnfq*KABvV
z<J{KyASTQ#j)bl92AMeH78>F|8U1ge;2h-NeQspG-foyQS548NsdS#P2Uo{vc|(W0
zHM$<_BlF5^cKmXm^p_0IiuB}NUQ;9cm2a~!;bKFDufyI`$Zjp52Rmza;?#bdqggME
zb06;Cc}RJzQu@c7>|15d`N1wre>A_ACeP)cX&~gJJf@HEqn6H4&70=3))z+Ct~B+Z
zKyx_GbaeRLlZT8`K9WOLOox6;<K<6Uz-!7ZV|PKv6o{dpzmZ?WFsY-HQw2|RWRE{1
zrj(kPfc4XLt0Wzii^|qKMt#sgklOX*sdxX*-cDi!by=|f!>*F|L3~P4+R19DkmaY4
zw<BiKMyg}QZ>Wl)_}rn3=?zg!CGQsW(o*g}sInWJGfhkU-BJj4v2&0T&O5%A%9EuF
zb?(pq@dA^Nlj7wk;;tf*rLG%$=%WE`ZEY8h;^s!d*<;dvO(a@j8F-FNb4DgOpo@+Q
zUvE1z?CVlP?YPSs&%oZf5bc<7_SupaP8@im=si+z_<5L&z6=xBi%srsu+({AOGng*
zz0t*YR}>r~@uk5uw2oZ%*kn!(RxOP&l4NdF^rrY(kL)PcZJU#}Y$74NBrGjbbBz)e
z+`<|v>5Xo~Os|2^bX@kMqIc}(_ktG>WIfBx_Q-0V_cC9xx8ET$3MRs(FPoSb5naK#
z%1!ODo9|OpAfNe>G0=mI!@aNsR4m*t@2J0Q>^|fn11|YB3L8IdbmrsVP~@DlUCiHG
z-=za@XAK!@va=}fh`t6=bNwFwHtJ>iT*sz<u~^1`v;dxVbH`#!j=z3n`uuPvd`1Ne
zE|%_!B)x;~(douaU;T09=*lwI?GkH`^&d<8+b9>XxyrbPC}Wa*0w$KrG&i98IF7&F
zF|$b9Y^8>?lJ5onwq_dFK5^TFd~kXtzRKrhb}uTuDUck8zqb`Wnt)P!zDj+nud%5h
z_-g;Q-~d#p3=@YyZU`UGOoYit<~Icd{-F~HSj*gxR0P}MZ&s1x9NB&k<B=cH2y%|6
z0dme>@MV_td?W1_u9Jy4LT(X8_v0r-FzKjoCorS(7e|D3%pV0hzsJ4FaFGayCLg1O
zv3XC&;Dm&Jq=I843^aK4?3kB&Jt#X_-8OL@)FCnVQRD8D@@Ju$9ZT?$34&BFQ6%hO
zK!$CXJMM9E+aKu&GD!JTtq2ffKzauCy1qVsYU-=jtO4dWH67vjkH0M@{52kU$}liY
zW8nAiTl?2uwaTH9!ID<H<iE5^R=Ezp*Wdl&aeGA0NoxMJ;O>5Px}M(t*t8t%a)hqV
zbtN}du6%XZt<EYqJ|3ST_kJt`va}IY79OmM)g(;O=S18!NDE)O%0mvjH}rbAsPLYb
zF0ZN*N(eqy777lxVz$(*WxioymY$yDbZJoeL93<UI=!KdmSBZcI2Yjvm2}d)CS$su
zx@0(U@BI5;enG}BjG$w~Ym*;YIy2rlH{Dz885~e4e4*{fw{xsoB>1axhdAMihtT%i
z@2ksi;ku-cnBg;=lnqTf#gK?*TEF=w-CneZT?<)oq+zCq=ozp!J#-YDa;1<7Cng~a
zqLHVa8bpp%q2PbHY=3j|Sn|EQK!}qN{EH6G9nsq%lxZ}xxne<3DsjWtsl4%#L2gC~
zp@sz;87;3|OjK;RoIbU+59Yi?sjn7)zWQo^2>vp(?GB{hFfeTA{h?L;;^&Y356Ws?
zw;<?>Fm{(ALL4D4&0Nfi=Yw=$TW(AEV~hOcj~;hB|3bQTrWC6#*~#s?0Iv%RfO$(;
z7yh}t4~3iKln-_dlH7^DUz|O~WCGer_YE?QiIN09Wa~{iet%`FSMNnRNdB!)|3PO7
z?Nf8y;7XI{@)l$T-j6WP#*|_*pZ>mzOoP+Q!Pv>!CTXbCI@4~`JL3v^ZPyEHmAa$-
z-uQ+~5z(D7aOO$%{X+L#V1-oJ<5bo_&Y!Mb0~NB9NP?5tZIhkkuqbH7i|i+9s!fC)
z@@L<>g})s>H18<jwf&yzE1fT1`B{iS=K<04fP)s@<8R-k?2U*#8Djd*9e215&d}%@
zp5Cn`831Dlm<g`O4`7K{EUiS4m0`urR(j{ci*r9N`$3&Ll3%}7^-UnI;rAJ2d4qQn
zcS>D|$jO<;6YZXC(6w|H3U&sg*(>JE>UK3D`}!{HKs1F?^^DXur=ZPqVG(ybIS()*
z4TCvZ<HTd@4f8Gg<odhAU%?)s2^DKb3Y@}p7b}TKJzJtKk=pF+%Hp@YuTJ8E1k$_P
z3`TP)7Q7b9rIo$8$(LJj5U!jswP(m+zVcguY5`2Nk|WEwByv~ihT9QehZq?dr>(M4
zc_Py;T1rP};p7g(vc2A0e>O5pGt>x+`Ln=(1n-G>EY)(gmiR&j@9780M@Q-1M;Ccq
zxrUDfa-2FNbP2eQ^5T<ts$$P=tk;PimB~|DwJEEYCR41lfv3RE08Qv!bx-+CTl(ur
z;&3A4*t7}J_D!|on?Iln9QoY1-12%z=#Z6zu??hCHv51e^eo>~d~QYQa8nhb^|se8
zn%+6P!SX;jqpajZ80J+^(u;HaKxXM@<cfe~NBeSBC`l!H$8g&qRiXm#xC#|a`WW^>
z42bb!4}u2`>Y>LaB#z;qIFJ4PnqC(87|Y#^XG=`2t@Lgp1fK$tu{bh>$_Vj5wUk`R
z;6&Yqm%$gOe0Qbbhm^6cdF<D%<wKSm=>!f5*v3qPwCPVUq%PpV3ru&0aRgrP-=-1v
z@>vKtm@9meqw-O2IEh1nT{DI{?_W8{-=6=r9=M;rcMi~r<u#E*QNn&atuekhVsLnx
zI73a@+v<z)SHkl5hndBFI&T{p#lXHb5}>^edb7~FFg%hd;CfCE@*jUhK4Rgcmwzc^
zSr`9wo9cdVHorC6jlYk60lepjpK|??s&m<hpPQ&o7xqRgxB497IBvJHF#Gu($HPW%
znlHTUCvR2%lQ1C+{w^ZmSE;&-Ag4{zu}c0p{7oT}@<67o@9q<Aw1liT?@%=`AQElR
zTIWTY`+9XOE?7j_91lJ`^;5)=b4(<BC_-!yc;mQFfNKhHn+avM*c<25ZLBS_)>RKG
z5)b*5X(vx<nTF0d_pCFQkUpAaRdYAng=m92)UY;rgX{S4%V^0-N$j_ug#u*)&VR7S
zF5StrF6$Z23_xEVNEsBGF;5-M{WJ?V&6L+LS~t{htmC6P;#boOG1*IAmUUHD_BD41
zKF4_C!Yx{6F=ZKUeKOW%viREex93(AI!(4p<r8gJ{7s`WM`)(8Lrq`%sUrn@jp>$+
ze?U;Sc~Bs291bRK_xQ^o<1+_k)4rrZQw2vuWlO*NS8kL!q03qt!p=WFyeQ2I>{gez
zH!61<-*E*QEo&SrZVkPg;QMxBr?H$H_Mm~oRi>hNG>rY22sDYDqPHll`+e)Wx{~qe
zIAYk<smB{-OKkIRqd|Y<vfiHVAkszT`I8Lz_E4`&3Hj9}3!2ZUwKI7XE3{<cN1k~L
z<2j3>&xmQ+thO(`D~Rqwz_mznrPaJ?gUm*6K|UON0%yG^W2D+_3We!v8_Sm_aBo?h
zb(j)!B;i0a=ed0XFu6Z%i{j6YBjxK1<LOU&`K^NYhg{VLV0uydTHhbE_^}I`H?MN^
zl<iLgn;al-l2*!H-My`+yoir!ME%me>F(=Y$k{-+i(IUTKgZ%_8qX6R>;2*-6H3L`
zN5Ak7tQPCp$McBp9;TW7ZYOuPQf+$HS>2`d&V^Pu{wW<yS2t6;iS?vJdm!n^d(+)e
zi!BN?G(wwV8S1MGqxdg<V3k{RDc=M#^ef-z>sO}~i8=gqh`VoaE7_SQ-t!_$(jIqh
z2HBE&tNhipjj4P(L~+9l+CNiwg*pNV^{N>CuO|74*0FZXn#?s+@AkPLVt7mv^o5q)
ztG~iOa&|pgLVM16M&hm4#N;VdUBF<jCS6mi5P}SJ*ZODc8f;M|Tg@>Su7`JTRVG+>
z^&WWlRVZ#~Uakd092<*q*}r6dw)*!q{wqRb-ACN0PH@n;zkXn6=X#g~9lO?P`@wKj
z&+D8P*9E*!pDP$aB?A@6Q=(^B@a_)j2EoxB`yXt0i7>X*hH)P9jcF26N>u&g#WMNh
z8ML2+EwTARyTU>=m<8I;x7ijqpK1mxx4xh`?Rq`vPrJ*<h|8;k-#)_|Yiwr5bX9ZW
zJ9E+S2a_O5{5aEKqS1l}MEGDZDlW@={CWPK47tLE8A=v?HkY&bnE=6@$@rZSU3VAj
z@UU@P<IOMecAV;{Ad3Q%$$OHTtSOycm7_=ColTv%f$y>#B#Q0`nzY>wR>@P&yL)0e
z{9MqCNl3WP{tPKlK?h7h>t0x>o+RV-6dW6(#Y)vgArgN5MFEv<CvnuGiu<Cbx?DDy
z_U_`re%IFuW8SIS3za}-O44v?{P8c`Ta**U0l0p`-Ubnn!@{gVEs`>ef%}bn1d^rr
zctm99n&@QhLQs5k2v%mFmD-KF|C|(fYMO5!FxgkauaMMNisdTZ4L`Hc!GHUnsBN!V
z1UV*N(t^05qF_GZxyC+d@Q3oDoJ%4D5unhBA2LaYw-Yj;%%ne|M-J0HrUj+o!F{!b
zey>s@M+A6@LWhV&Z9%2b>~T}8P@hdrt1Lg{|8B{Tt4kUsF~&|lU%A8ROe++Grv3+{
z%CKL&(vh0M!Qs-QxfkGlB3b~>P^qfAll!u#4%5~-)_xYx=qAlsmM5~`6mpe~gb?{W
ze(e*Ggii09=4G;KGTs+L+laf9++@h-+)9s38|5`(Z?X7l;>XX=Rw1pD{&yobG!nMh
zHZ;W}?TCWNy7q>q<;Z^26ypXA>iwbYaIcnZo+?>!V_B>VU1*hjW0m2fly|`#PDauG
zUm3T!Bo+WCQPyf`5cqocd~71O+uCvd*)(R+yGO)%S0`*rAf=W=4AsodhQIJ1uT7&x
zg)BVv{K<pA_AS?dig?fL&r5k`ckfx)WfPg(@O;m&J#@%ISWD7AG%^Id_(<L-#bI48
z59#6j`a*agNvQE*f}h3o3ETHl?Y&r^t~B1*<ti-u*E9K?pV(<o#@Wf2)4LY3>06nT
z_#1Fhb$AN`U&KG@b)?rLL~4IbTS|Z&)?#xJW*nQ?=foxofpwZcXD$<PIYrmB5xupO
zj?wko5|oE9aiYX$l@v|Sq|;l9U}|Zzo+s=|DGiZWevkS-^3v;R*SU$Y#<PWb>A&w)
zCC5T=3|`vh)Z`BZGM{z4u>YAkzM)!ps8X~~!A|=>lf%EjalvB}Sg;dy=Y49FhMLwL
z^AgE$8^UbTEIC-PxKz<IXfBhfLOKM#7&JR%aA5&pRzFfOhfvV54K>E{g}$UH`0HO#
z`V&{NvI&o+Mm>)Oa2=MbLT(2fc$Ga|>B;r)Y)->u&yurz&0co$r{BS1e^!VDuJdIm
ztj1_y%6R`O?{BQw5cw~x=*d&YWePUq-zaTw5X->Iba!s=<x6tdD+|@9+qi$cjEL~=
z%f1Re<ddG1X}Gh%C0Ir%_4vILd63gVV1FjZ|A|X^2B425;fB!WBWI*V^jG4=1!jw=
zkV$9SprK}UfXBq^WF=>Z50(~U;r;=bgkz^v-4*Xp%<Sxb{dxJDN|e%S1y^Q9uf%nW
zlv$21c=o(!X6*Y!D*~(5lJo`@Y}=HTm&@XhE6~m*Qvh78v#Pv}!<GtR$mK{>H$^`_
z(Tu$N!|_CTMQIa{jg4AXLI19tW-t*1j3Ua?)>1zI^dd`dwIlk+nVCtAR^e#Y@UaP3
zqYIm$j2!zC3uH;}fyu2<k}qalWrUOc&s_#k*;4Vr<m%2OayC&RnJy~Zje96KF%!XP
z;xu9&Xu6ba0_Jw1b+)FaahtH26z%eMt1@e@|A7(yHUn~4@YZ7t*ZZA5ha+^(6~Q9y
z3jw9HUQ`D-X-%&g9{oH|W+fbfN`sF4pS0k;Ja0!tnqTfR3038Dv`ld7T_m5UQBXjT
z(#;mjW>(8m#=d5IqPRu$0L0q7!y^b<1yDy@sQo>y@;=ki{qYW?25`)BP)LyJ!?sp>
zFPel5^Fj)=c#26X9A8_=L_CE<Rm@2QumHU<(}5HTgR2V`=&upMh4W?ar8<z_81<DY
zIaCF$Z2-NH*=$$VKY)-s8;Y*1$3DJ8Xr+L=e)+QT`9ghpe<-yK{P;Ag@T%2u>;)=%
zz-PYTF1diZ^VyvU&@X&K6`9^}H_tzqTGE+DfI(5ZjHgQLX>gb9orM!qU@s+vVacEf
z|I}t#E%STBOQkG=M1^0f_0m2`ZbDeY@&>Ut^2H1UY1fEAviUEoy~VkxsI0|K0%~EA
zt(<i7HWCyU{SJFBS&gv95xo|1QXB<?i?g(;2O8lkgvVq8UeM-z75eAcK|e*C3=L3`
z;rF73`iLc9Faav+P0?CYaU-ry8ydKO&sk7XBy=$3>W<v&=lfG0W@668Jorbn3U1ur
z<+ydEsvf1rW`86`-|&7=7F^<aJbMol)19p*bpF3kH(fb`r;yVV5UqO=p(18BKJpW}
zq>I65gDj^=5}UKy<jT@|)*Bm)xA0Tfk(;VLyk+986Y^N<QKkm>!E4$`fSvH^g%?Oc
z!WR5A_*4wuKvK@DFdE7<oPV@Rc%Wr7>&?E}wEH)q`A=dIt%3|PjeCXTC|@mz4OLK7
z+a5&z2Q%p+FcVviD2E}C#NCorI*>&$ahBA=UkG8rBrpK&e5QU>(h1W`ZY#5Wigp7T
zPxF?j#30t~11r~me&TQFBaaEN+Fuct?i8Du#<(0lB6{w4OeBoq-s&-85RAqzV+36}
z(6BjYaKU=qJe0(deJ#%wFCYXkhLwilkpxl~0*8P&&_2q<&}^%X0Wy~mEcUFLCm~HO
zMGFmV%*>w}KWdgQ$&pg1s$BTd(_0V~cK@9GK~`jwp8~_vEMh$?m(j;f4fB+eTK4so
z60Khg&9_}k;if5)y2ga~#L)wUmm|#bDHeH!2(hizY@KPJ!t#!LK8f_084YK;aj?Ux
z?|4L|@t-vpg5(rn3)Yq5U5Uj~wrq<wJ!Y*6BL#W8wN9qGtuIwX`U_vqh{6(si(%OW
z<%8YP1{}+@UCm8^kc`%y(O}m_&Gxb;tt@z9J$TX=Lap8@BGb@0CaAdaOd?Oy$@H1j
zD%@~Vr4G#TGHx8FKx8;JL?e0Xcj|UZ1#nXz4i&Z*Uf*##H-HP|XNH1`v)|OA0ou$*
z!j>EU(*aRFBL9{Xqni?Xq2o6jXtc6sldn;dq&>;k>gdxrGh0Ig{N;A-!H#-3?GjW@
z)|F#Au#)tuoAOHB6@On`u*R4B*k)$*j(VU=g4N4lsqE3u(TgjoL2H7q@ipD?WV?S`
zayzd)A?{0;iX;p1Bx9{vu9+>rnxL&{f%{9JHLIH?_tlg5XF8Ub+U8Gvw(OsO%Yx$m
z%Hqui|91aH2tOx9z;7i>qqP+6c{lFs915d4FrePwZDrcT@^lE?T-yTUI0Q^=t^F#J
z!Jj{wc9$vnD)n60OTEv>RzfcH3aiP2Q`znsb$(w4^1z4Ra_xibdNS{j&Ku-1{|SE|
zltzOG0!}*rdf2R{Vw~Z=2`wA{PEWs<*V=4#-TB7Taa)r==Mk`nE=daxo;~OEF0vmD
z>Z*zso9xr4Flue{?sC1le}uQdHr}5NE#TXV+UggHVKEWX&N=>EXQA4{Dq5-{JmUzj
zfzVuMQOZ!kG7@&<e#l{;8;U<Oib=AAzX;0q;{X@bH{L*Mj+m`1*anf^fQN%q^lEXS
zXE|k;E`1L5b*sq!LuzL9GN&UPEjfCii~Qw#jl@xa#{-`0ch7%_(n_%*XT2wD8&3I^
zrVD2^#_5PkG>9SZem@a);9ys?#Or1BS5`i<vr*k4hUO9nvqzHfKablQf2b6z{c;_u
z&c8B)i$Z3?uKUo(0L0VxuDO|DGlslbZ(Z2-%L6T-KQJOL7AkK)AmbXJrzK+_Plh!e
z(p8u~;dEgnQkR0$0_GNu`uXQe46|A&GFG{fc?_~3w;9Yl*wdz~xh|sRQqMqAuSIFn
zSfJL}Jq1kwFWzd78#^rTBDk?mZ<!a7eUKy{o9tD2fs%rJ<0`yHu-QS%qguKKYT)28
zv1U?I6QyS&6}}d{lVW$%kv;`}!Wc{Wo#%(vD}>p@g=9ds4KfxSfr&{^quQ6hU>}^S
z_#2AJEQiaSna-;$!cr*c$HIa|X9(D5T26~SqLzj3JZkLepX(6xyYrK`7#b$x{^vGF
zeMHp5EQ#|b7%fAy!<M^pTiHyp>Ou<<q=6v4%V)OhJ!E?LD5!TYEo{2vq%UG~5`VUe
z)@O0@xkTaUXI8j*g0P^zN@=}E<LXa7@QS<}HX*+Sys+BYo#$7~gj(YEMtd7Gu@hB`
z%(;<k;Z!B>?o4sBKH+SsJtpJRMJMsQ%PyN;%9aU&H4AY1!C{=qhdteB4E76CPQk-a
zuuX3dhYagNSRBVq4zh;r|5O;C#klt4rPY%0vL?2TkM{Ko0mWqCPVz0>?L7{JAM`uv
z!&{$0j;Fi|<_MGR812&5Br;a4bbkiejCZa>qXjyAZRu#X*GVj&B|2akg2HH@e2_!#
zWeAk*rc_h>&`%lu=roOLs}G{1&S46(3yq-DymH-}Tv)nfk=eYFEv2o|Ui-N6_Dsw}
z-REnE?qGxwF}z%{+;ufOuX%TkpSs_hf|<0Mig%jw)WJ*!6eR2~_Z0;LK1dXY0zhrn
zxZH^k-SEUW8LOqi#KH@bYq3ON6zjZ9N6|B-zG%74=kWZgs#D74s`s=)eBH}pw*4|P
z$3<zlc~$t}zP8a~-du($Vj1MZsC-}<%wQRWkKpHCE365FRMB$-cZ@p@laKf2_8i&u
zns7XuZNpi&nZ#fW+5L0r!;1~*%m6HXdTjmpr|srz|3I0~VqwLKa>rIdQ-XR+=`}l#
z0s;zM7OD;DruvC}R$K8Mj$;t%bMP<#(dhjl&I5Y2<Q|bJiIa7qXu=*&7R;o*Jg<}#
z4#}o>9=Rz?EqC%2FM8<&JS4qJ7_sTBhwv6QQc4)F>G_Hg&fI{hz2blqeBg>{jf{3%
z6=6T;%apG+#i+N<!un#6hoR!Hbh{Q3G=&a0>d8H)nXFv0Uye}NL%MmE7?6G5WYm<E
zfAUQ&9>9?4(uj@z`c+`Vu4<DTX%WLML#3IDds$5L{+|zITbZL4MlW9;wjIuj-IMjK
z+UCV-&{7OnpZUrQg6NX`+qG#)_-yCbI_cN8u533vcl|AtFzI`%cr4K$5V6}((>f7H
ztJ`H$`{8M}Zq!}N9pWir`UkfuYiJ{WY->o6JOgt)+>pvo%&VW|XpEW|U;i={x)PAq
zH4_n+v4bEW<LZjqNuRJsT|S;r#x8l3aNT(k%i{uH8<P3OLxt`X69BTe%CSFg-}tiD
z+aT+LEvTEPb7%vak4en)DE0bRu*id*#de2Tv!Z<gmeP(J+=zOUAl(I5_x4PZONO`u
zf@w`jL!XqqmPH_8ce1s%`f_5Ru|I7!GJ-mP{kjqF1Jixy{Z!q|<vGiLJq;>&ns)W)
zv@E;cNPa$uI+xw%muE5tEKa)WGN~re8y~0Z1rN`7%3y@T9G#DhjuJL`Oxqzw2jlz9
zNr%~y$1XL%YNnd&>Sa3cmdN|^fb)o2B#M(QK|1OPFJsu^cJ*E(`Hi(034*n-uWDE3
z5fdVZj+epey$HO-97$%x2tpZ-dq%x?J3iNeE)7<Y0p~xtu}Bp<3GnEQm~5O+d`#KE
zJtBML$b`K#5b3OF)%i<7#SZgDHI46RWe_NbjhjxY@`YdDWn%ds2nTp{V6wrIp$QC_
zNTj{pN>BNNqeI;B@WvtRYy(;}?TndB8sXEF8u9b2QuxM$izCT^zbED{Cqz?=K#<(v
z&Iui*M*^KphIpFw=Y&ymt*d&Ry<<O**J0gGta%i~&i*$azzclA+u4W4Va(Cdz(#R~
zG0O*j#f$AMto>r!_1wBgI&<Er#~tUymK6PRsrn|LkNw66%@7$Imhh&awY|b(Ligq=
z2?9DqUP;;SWu|yL4jC>xetKhgkg@^ui>?~HX8+=d>Z}TO>6^PLUJPu!8qTil6USl1
zRd!<5<*e7}rK?I{9<kN-Qjz2_Sn2h0gi$#7yvcW15HgF_8+A#w?MjpW>(6MUia6PI
z`^rjUL=-R8y;|tr3UWQ{ru|VNkgZk+=blXY_Fv%XFB(UlS=%5HxmO*W3Dg<$)NN1G
zdeW0h`$0~SIt9j5{V^FJIuXg03-;r*p~T{>8!0;xxDPLAm>x-j;{n%R>fU|$O8Y?4
z-<k_fn*V4nK&2f?r&xAaQ9Jm?v1Gi_0XO(X7c32#MHf5+H=qbV*jLFN@2vOyvfrH~
z{BF&Q3amnwk90MzxNYnvbKk28E^mb=k2M+EZE5h=^VP2?4Nm?Tw;M$X;O!CM%BFtm
z0Mp^f!2RHspy&;|$^8)<8MV2*K~7)-)-2$>!WyWg;|s{1tk)kU9q_`&y?l0cDoEha
zLA(Pwodhb6r#DWJu>j%nmRAOtqn8lo)MOBqvvFhu2KZMDgLubHEv<m^PPho~s8RMG
zEx;XYFbH@l)fnaa!HWBQ2kwLZP*nFQ9GUhGqDZCXwf}|mYRV&g+zaXR_FF6)nS=JR
z?)UZAlT3i}ZfhXwMEm;-#PZI_1t?*r<0yPLH@6Id$RK-3KOSRuV&-GBcj}OU3)2Y_
zU|K@ws}iKmEo=FktNI_ey!O=vpF4hf1n#eI844l3W#23pU%jv!O0X-dQT#-0)^T$w
zaU5VNW@2KW(vV*uax=#!Pdk8zleh~gc__fGF8l9W9SeL)QfE&evAjLPfI#q!X7hkl
z*&48QIlaZpy=H8;(A1j2j>lc${y@RbN1(M$VLJ+tnspSFi-4$=vqHoEUu$VK1<SvC
zXCIkZUKQay!La_h`B>5lDI%g6SgkuLz!8D_OgL?V+3G2gd{?=wHT;UIQE3s^%P-=6
z2%le(MymT1NgWLr9nw-X5eevSw)!A4xA`@ry!@i~l4p_n;?<!{#wI4kxfBSz`7`bF
zXJ-EZxz`SIcchjjB|D7Fa{S_zjwO;oGNwPFQg`z6{yL>l9(HiR^-CPx4z}Y%8AkVt
zTB|ftm%5}vW7PWk|Asr&HkJ18GYY&~xP5m(JIIouKBGw2Cut)cV84eD8{eIcTT27+
zBn)R&5}@d}e=d~X0c%>mCD}o*Qf^HY{-=CePA;HR>puMhwHfuQ{;_WCzftH?HC}4i
zw42m2n8g{eXJBQwcqH|~#X&k<Nl^uexHS1+De>NCy8%PdfKWx=haf^R_YD?u0WuB|
zeejl|?X-u#q?==g<gfN!0?a^fBSMGT=X{yuw3^g&b>f%H{}&*%!g_a7rD|;@L_(xl
z=GpOEM2JYK-@Uh@xEX<{Yp(7J0>Ye(d?c@H;vJhRvp4?h*?s*v!fSvOI3mhU%n(rZ
z>f?)dV#7or>{*q%n304Ipr87^_Vw1d?NraLp#V_nApZ6!$0yVAdBAGdb(@Y>Kt$jX
zjE?83bUvuSq%VEi$b^l#OU|7zfM-M$xL#SR6O7wTbwru0m>jP^F8lPyF4UiU3m*qZ
zt{Q?;?%0JIo>Wf--bWw{_;2n+p;rs^AYGJ6S3USTl5GsiIY0S;{KYy+Of7*VQ}WOj
zW4`9uopR4vmd4l1L6`c;{f_jm??k;B#=cNP@<ITDjjGX`<!;GP-5OGDa<J6y@mLf+
zK|{^_zs0N1*b;aGlbmj>asgp@wQ}9)>aC0j<V{AtJK0v5B$!KysPC0``(c7H5*T|j
z9*e%JpXcBAe(3HjnexmhrJ}aBKfAnB^qT@X4+aEsbxJ7DyRq1%qszf-d%sXki>#<c
zJc@`Jk(D-z`5Vrzrd@-xd_%A@mKleE26y<>n<HiefvBQNaIjQ@PUL`o2m~C0Eumo@
z(zH&-@Nu^N!npQVP?8Ew@>1in<<ltC5Snr1SogN%b2PZ&_<BL6C*$BHWKD_og%db;
zAm>AOb9gS(kjq<U-I12mN>9f{W>`ZV2>9!sA`;`s=b=Q@@TQPaT>|>2xxXJ!{ey7<
zSDhH<?~cDosIH(exi($S{rV_X@>DJ0>&eE@vY`Ei(959txxiockU1*Bfk!hAc*3&A
zdsVbC{E3B)-FubeW$*(p_OvkqUR}J9W4%qj9xWS-(u6g>P8=}CC4xnCd{VBY94h;&
z(?zNF26Oj@36ra=%s=mosM_2WBEFOp=D@5SE|)~@k;ACZl>2#^j!NF)rYc!E>vDes
z4FXR<Nev>%D_ulN<k?6I3qE=|=|}~NPeke~$wv%aPST`Q!d^s!<SC|e^pSGPhFLN-
zd-k$C!7ZeoA*Ff^GwNKR1{-%`>)*BAVQxR?Ql2a0QgvbM0I+#>r<RTBtTW+DKte;R
ze!2kl-tYfEU>BOJ@`^^_&b!o+j=?cmC0)YFBwz#7pj?a+Wowl@%J;ux-!4j6Fq%-;
zLR>N=^6eIm7bU%%sI~qBh@7&1_cVqwca;{K_SJcOa?%4A5g<yW&YK34N<wO;=0d_$
zvmMA^>lZf`Rf4+iNi$aK9L;R|*SZP*;L@p#Z4Q;Xy(sWzY#u%R6f+W{F`hy#G~U8E
za6q6!F|a-uc3tK`SBLO#CCY(YS>o*oh|#%~-trn{NAb*VYp22We3Rq}>Cj(Ex1SEl
zz)vCWfe+7yLJ)-8)Uwp(Tc|TLk(D-JCb}tM{y}mX`spldQ}t;nvQWrZFck|MC@cIy
zs@R@A`-=U_GDGjX15iRO^~aNlJmTp*?dM;956Pkp1azU~*cnxKk-Z4vZ3=u)*yF0O
z)%~F)wF>Mpt|32oS^+mcna?|0ul9EkP~x5rO**I!ixl4LU@1lc+##DNr|h-t#mAHt
z#OZgUWPD-BSd1C0W%3z(ukEx!3bVn#v)Hz{iinPXUeA&}n|kV1hVwYFO)R^|1y5zH
zSo{lbrxFxC|3&whroZc9mv8c^LyNAXBzt3>-Jwz=Q0jTSqjxCADL)H_rAzsfO}7-<
zmZeyXe+ft~U;XSR=q|-R_oGlsO7MNhS3&MklUFS}OCosjcP6VM&n!s>_0Dy-JL$)&
z^R<_!F|nmqZ*SD?ie689<ez|${`ljNhdSLJBEcY_X=$wl9rQq<@Z$+Lz42mXHh{X4
zu|n}NyidDPEic|yHw=HWj1a&vVXccH%OUb$`@SXwFL*3(k&la-6_p@z*XW#%fVeM_
zVP!}31`3u?pAkE5A15gGP;B3Rc4sk;|5ZfB9HHP_M)#WN34ZQZN+CG?HzHvU!csme
zM{ec5L1Z&H<pv!Wazo4csC#Ax3PFTHnuP&Mg<ap?p@dsW+#bI@Sz$Bf#q+#D-H{*6
zKO%Lgk^^A)i>IYQ)b$J*0#r!tSv35Qo+1I`>%M5j?!sA?IIu04`ySB3)#`}kT%3Xl
z%3IAg4~ZXXDTo7h_VND*a!ab*)B?au8ptWQ?9S;;9eaZjVJ0#pRRE&!D-7AKf@w0c
z!ua~WXNdO!Wmrr_I<2lUQZ*t~TQn~<-k);H*ULl&g`5cBNf})5Nl{UofVlT{f*XCM
zqfqs+x!Vn(^aeQ;#2IEd9E$yc0)g#KO^FdQCBJ)@;!UMfs?}M^(gNPgcRyV^Q^f{)
zrN2;_AiD7dbf*ZU<u&BbktHsP4ff-91yLj44WK0MqYOTC5=S_sPUnP$xUZ7QAj*f;
z9B;iJeUn<Oxc<%9_Im6S_c3Jvx8a8{HlIIo2t5cFyFXmil(0fXC8au+ol}9{cDKu%
ziOWVDV`7rKOg=8JJNf#tw<045hB+zoL=GJQLH4+rz&+xf#Gl^{c?x@wEd~-_8Iy{G
zo4eVLScibF!8_z$)z%NKlCPL~*r^NwojnjH%a9!A=&;E!c@r^yKn9loaSmF6L#zTd
za`=k@;@Rdw2O0=TPy=PP8OXZv6SvUf5V|3ouD8VUcz;>r+)Ao<diSpbtCm|-Cl>@K
z%||`>J6K;|2r@D9HWF{5T;ImG*&x?mI?<T8i_Mq<@HQMF6RkgUO-Y#e3ClK`p<0`$
z;M-pvOt33nioc%1iX(*~6cu}=w~*Ksz;+lL-`~7{{hOJizW^Ixti)(vQK9oe)4+fb
z#l*Pot_cEhCB0QbxSqvBA6i0A<!kN=3A^zQfNK0c2&l$dHsQTV5dN-?K2qh~Z8WY5
zWy)HIAi{ok$D62}u>yh5&{wyZqdO7y3dsI!(0~DyxL5_X4bTWbkxS2WDubsr5_zNW
zwz%fUQzSEK*DoHi`~ZN|Y6l}`v&_4hh&^a4NZlpC?FO-=yRHv_D2oYu9_NyzH{riV
zz9rrG(+qK65WeLJH4WmBdl7U8qWmf#G|@{0DUT{Ev(vERCfr!TlS3%-K4k;+dfgzS
zD}X_L*j()yQx`8ra2+*L=G<*oc>lUp^+{*e6JDrc<J;Xp24R#l^07&*|0XIXR$**y
zl_a`@qe?f5h*$uTaJ^oCfG7+F>@SN0#Z2wl&YdaGU1A_Z&v0LU@DV+LI`anBdYg_o
z|6_#-dThoI;29)&Bo(fJfbjuvS#Pv#k)r{*`#oO`;wC`C2u&v{RlIRrcDm}#D8yv~
zNY8u%x_*^)+%Qb>3Zgm8Fw+C(2|{efP(Y~sMOJ`U$rC>EYxI8gFIajzXSfmT#;zE-
zAqbA0ca=Q}9N>=rFSztG-u92%_(z`-ihF0SZ-{u#8ybpW&V_C&inYwqwTO@J={LaP
z#6(3z<h~h3*hbxqs6)&Ed3L=-HupBl=$d8){q^5g`zSs!K>5|?fUm$M#!0umyO@W4
zxyKP5BHMJgNZN-O!G;m)zF+A}L)%Ioe^Q=m)^YtxL{Fi%ml*mS78%R(vNiV0!?f}{
zTP1xAq|32)tw&HDkxST^scYn)r;6~^kBO5%x;N^i4OFmP2Ec9*3iD0T*Z7s)M{laD
z<M&)l6>mC3S-@%TC;Bk3$LW?cqN(}4X=_CgT7XFp*ff`xW`vZ?1OSJg`MOuqZjg*$
z-}+MV8MwlE=yz@^Z9_mANK%j&1!q~4{jtw?y6c>jSl4G9IgCd#X8*<AxZX1g@sS~3
zxIb6=TR-^rM%exL*wd#BPYOHZ7UKv<O^d{R7)e!xnHYO@`_QkuOga*5n2r$<MM=Pr
zfo`K~rWf?;+$~$>pxYI=xZPQ#25Fxt;2`Hbu2;}~7>UcOTIm@T!q8?ux$dfX8mZQs
z66>5jD3fm=kFUuU&%*vN(Mk_k=Ceox@uv0kkH}%+;gT>{R6uj!jy=@4&&Y$TsVM~e
z^@>)n10;h1c=0&c_#rOt5t)t!bg*kTm`)qqUX@ES5UXn7kFh_X`w)}E5dIRhG#Kx_
z*9XoIe=2lIV*Xrk`Cv`-I}XsA-Z`+Q7`>#Da<7i_Hdu8e165yf3m7VE0*9}v*r&5K
z$;`d@JJ8+_f*W?5zY#}}Y1zcO_Prpo@e@EfhQ$9bko!JwWAHWiT&95(iyUTSVx&Ja
z&0O!YNZFZ17~)~)FSFo&U_&h5i_CJr3@8x7G)3-aI0W1lu-~R={x}~wSCdIL(G>6+
z9|5-db_YMFiiDk4$`-~jK!KJmIIUyVC}-RVp+7P)FOkkX7u@yickz+>ZJa-qFzeNG
zY2uBjBoi_){ya_%-SZO$*&rRFwB#wZ403w>L<RD%*-|KaE%vl1!ldwnELHiPIP}25
zoF&rem;Vu$CfqlAVr1|1-8#kb8x&i$fc#)Gw?oT$@N0hhyUafZ<!nG$77`WD0{F}V
z<S>9!;Gj_ej1fXA&}7HrcAvlOFt7Ez;x5p-AIoe9gqH#Dz^u|!Wt3s3&+M*O!pwu7
zVz%E&TZ-Q^w=DIDYTC?1C6=ClZLxAZ9ZxKo?f!jAS^j*Rt84jB<f+V2o%`~)ipb+`
z&s`i$Ng(|)TGQL4?9?hrl#@a<T}>u>(5rX9s(7*!<2an<_3X7lrfcJBr{ml;vy65?
zQ@~<Z3g46L|0+iKAxZ^r?w1~t*auEb4wNwEpJrn}OtP);;>sQWm})}+Y4-(DNpsyj
zK?CeHRPMgNoODKL17&^H#~^?L<|eoA*cn}D2u$2N|CDQcjxyZ1;rjifrh(E0pgI_3
z8U9WZ#v+>EGO%H<^hT~5#`#po<3EI0OY)JLS6Qm*CwT9llOVKNv7!d{JRr5s=I_R{
zhytSX&q5yf=^ZiZiOa$GfC-tUYj!}idt~p(Eq0bxW}I@aKt_}SMTyYP^>~!XTyPE?
z^X)#l=Db6IJXKOqaQs?Y5k>;$#ewdzIl-?eGs#Di>`z#b5nO0rgm-7XTSv^Jpr=_$
zQy0frnl^%On&bxicQagiWSO;jXxka+%bh1B1aj=7eL9;>NL4<FD=CnVku&7EWdjD*
z;{0&6A_A~2`6JF`_%=y2=159X%M`BR?F*J-^tSi4+#(PuXxGXGXuzX=@rmO0PTd6a
zUOeT`LZ*7Wy|-=3zqzlWYhje@^;p261Gj13(<)g<@>GqE7ZD}z4jSNSC5Af;1esAD
zb)x<w6Z^X}Pk|>E#O>GPF0AjzE|5Xm;7KSP90>SvSjofk4i8~XTM^7vTqTR~szB@P
zq?vD7lP3ZU6rHnSR$W>p`a3LSV-37d&Oxim@qDGgZUby2>(S3kF2Nbca@7Sk{tDYi
zuSdL4BW;t>w!9bj{oOzxUcT0}$c*F5<i-txYbr9II2LdQr;a8@nNy)mEYTDQP71yS
z-js6$-}`+z@oBXdmtBabGyuFd5xz5gFj4IQ&vKQP`L#^yP6P|)Wk=9wU@}^#_8GJu
z3h(Sn9Qb3h=o<sPgwVFAYiMSZPVWdp<`!4CQSLUpCa^VAb|K%m41`|c6!HCY8!wj4
zQo%b2I*t`@0f`;tAK8JEEmW?+TAPQW<o;kq=)~R!pv%tJ&QWMweesMVnR)wqGzT~u
zxYzBz>${KaHv{jiS{49r=mZS{C{|W%KsrWu@-lde7g4<kSpaRB=}8S!7@+sE-8N_t
z!PDf!KIP-m1HQj^`jNx)e)hbAe-)btfF@*aSl!U|ck+GG4dO7J$oJb3uj?wZfI1&2
zsB=>!zh!{e$AWBT#sMA_eJC2lZKX%x?;qHaJ<AimIjMOx{Z%mC^Il8`@58=25+0b%
z&+`a}rij1G&DlfwZWER64^qAfwSrox&#L0J?CrKRuq+2nbVy$@>@D&~2@u1fLTs|^
zKnzRHzb)MQ)KH9gF#Qf74P<?Uf7~9BWG`|%ENuWDp-xW>RK$josvmYHe7NWyl(U#<
zc=omnITO5UJ4~<z@PVL2eW&yS8K_$@g)yAP0h<VjE!jW>$mZPW(o*uLdq`^2oyM?g
zrosT(-5kx#)MFgE0kd&J|AE=}jaaY2Y`k>-YlR)I22g8S{($@MoC4&aV16lRFp9;|
z@!G`1#`lScsHTEOlgUY&GD3=HnB?R;MWyKo34cR?{=BTR`ZR+D{sdrbixg@_oui~T
zh;YGd7@4$B%SP&{=6_cwLN@Y&+m-883fbXs{W#;M0WY_bo76aA0?6OE8v;6;gEAdM
z<_A5lWerJ)JL{l)eP_8ne(?#8(z`?!?O&NYE0K;|by<1GZs-Ps35YxJF;{skar%>q
z+6#*>&Uq^~l@mwL*e|t<)mR)2-;ZC@p2?$u)G|&?`afU&%&6S!EG9{`<^>GTQ;1gr
zh#i`39^nH0Rv7=K1Oh*A>Oa(0s=gixfOdimVC2kmJmVX8b6BoC;MW_Qf%#Qh|7Mew
z{$d3LIZi7(*!7r}^VhIttHLWAb^c+<yaFD9kLNY@UKyhIsV`mR-#HnG9Au(q<OQ{f
z;?t#~u@^K#mRReWR|+L}xZ&##R4}n;mlw6^T8LclBsb9G#jY{!HOpj<&Oj{l;K4QT
z4LS+CroGeQu$?hN-4w5JbJl`qtSmJohcmSVJ)bTSl2_of<2NK2$N2I-&O^%xu0?|c
zP`8pitu()}+CBLwuZK&{X$#EH*jw~m4_bG;KTq>MfeeHJEkK*qFFx=u!I^1kH57`o
z>uM7uN09KT-0!{a6Qs>O4bK+$(`!JJRnWJtgN~U&ba*1()f*#0j-^?-p|8dIo;*eR
zEW8spo{L!C4zY0|<o{%E0ETqs2~m;c00!n%gFEd%*$84amjGL4o<9TsxaLq9y8wBM
z5cUbXU9S5z4^d2_MB+PtZ1cZ<KpEm$h=*1hxc<<ChRq?O*{hibheM`DF;IPqU?LJZ
zC4E^S@%uVI`O{%o@p>R4E@MP$qH#U?i4C=l*l<c9B{dea%~y?A{!c=oKRZ%~>?LnV
z7$PDr^zZEV;<Gh6q=5Dzn+lx_qPcH{{ftB?6@p+5D?r1{{bGjqqlLaeSQ5B8RJgnG
z&A#l@rzgW?);9`p-*GJLjw<k+&vT#YKPyE0Gb#}D3MVMPgpCy}YY?JZge*MC)0XmP
zqmS)E>p|rp4W+(SMtExMdVLYX+W(@~`Men%%!hmv@CuEA^be8QxV!Phh1qC=1nO^y
z+g;<&djk;BqwxrM4x=Z3`bhhd9XX8Xa0nRX$&_HbB$PNzl`b9;jzQfa^3H0B@ygmZ
zqJ1n-K`$IgVWGmpl1d~*4D3LQmi`1bfp;AZF@vtVOpPZ9@WHDCih=!z-X_om(n+gz
zC2`nimNCA8Xww&@mZo|(J<Vn2YTln<x^wNx*g-`#;627rt7HVEU#+zQ!7%PT42dBD
zErUTI%?iL-TJlX!xS%ON{`Wm-^#y1>-|YmK{kuguI~qHph`l$GHSjrp8f~EEBm4(^
zQ{+Ejvh^M+Pw@A`StR;DSwe~$*ms^AWv=*em#JiTRS=h1BT6r@S0iKH6rPWG%dNSO
zt}zm=8R8$jZ)JUA4_Z0BGpeR4cpx=ikMoM6X}$ijmXA*W#aSE$s1OwUd;>3$9msyJ
zjh(~}JjThvisx_g3!LXsI{~0o=-q1*gKNqnIwx%DlT)dOR;X13JNes28kuifTUmh8
zqR(DeJwWA_!jCbYnrAxJBYggzW!3wX-K_8KsuO>E9?Dg4pG~#(En<g2Q)venF=Th#
zroJ@r?$D1~;x7I#Obm}lPd}N?UfbPHQYmyu=q!qjRYB4dh(f_j#jpv})(UURxnVv^
zCp`3wYUC%*Vd5*VjSDs^kkQ!n!A>Hge(j&*Y1gKo{H17*bPuoIa9kjjUHyCe=>)M+
zQ8#v{v~Ze;{-5^VGA_%dYadpS76}EUkx(g7y1P?SLQ1+7xae-AQ&PISOIjMFySuyl
zKiBr&aqs8({C;^KKky^G&U4nxnweuAs~8xDHf}>F9T7#8U)vP}AxPS&TAKJ1aU|qP
zGN7A8&Ov0-<sTZe3KNo~roHN5a&eib$yuhYsOe;5_ZAifZU*{cuqwVkhn|TF{ZC88
zUl=5&-TL-_zP{5@DEjyktE8CvuOR`wA>yiP8B!cOIF=_?X3SeeIVa0}ge~7{vGZ4t
zfD|Th5}*b?$NY4Zww4-ij3~d98lQlfx`tFn^It~7f8wuUK=RN)h7=b-Eim&oNp7&)
zS(d0M$6PoX{(mx03@Cm@CFjJ`B%B{_y_xVx<l9BlWuE3VSRxmMDYy&$#DOf%Zs2o<
zIOVCnsUO3pO|&G7_$3ZeABOJsWq^hMR(fjY-Buo3;%2L4gdZ%sfToRnOmpJEMa6Q5
z@t-~YPm(3n6fX4y?_#H<E%sP-)XZ8uufw*WC5$0&HUB6{*InrJrJ*tfv9vd*;%$QX
z1gnl!<yNP$-i-8<T#M)OP|f845g_Ht(J+57J29+|v>(;!m#(8BznyhUT@0Fx+gttx
z0UacOApF;x1+_|cS=16ZbwIL2ddK-Lu-TofGbNGy!C4Tg61`lR*<Ay)I8vTJlB0|q
zu9p%@)|h~DH^n&rTcCGGVbs&U$u}B*$H>Mgf<u8_IvC7R8wMCBYs)_->c3;I-76}v
zU_ANJZ<rr}Br$;u%QhOd@cKWm8xj>1Fwy6B;J7&aq?IBZ03@f4&T!L%s^%kkJ3;c`
zgdc$*IeH<Zh!}aj-yb;xB@F~OMHh7b$!CR&@;V`*g}mo!<751=)1qaD6(vY3b2A->
zr)<54?rs_zq}EDmWK;OI50s6-<AF{qc-!D76=dZ~D3dr5n^#;}%I`ho4TwXPN<7E^
z(5#{_VeUDLn=Kq6e)Oca**ir3$P}$4HciD|t&vTnCQ4im3Y@5QuUHi2Hz7)Pa+R1K
z!yGx_nn)kt3LWhp%dX8-fJzhsmFS;qIZ=Fz_#@gpVwewaJST8UvFFPN^d1-63CBM+
zp(s2I%ig4xe^r#MyHMH{Mz$PKhVYIUF12U1>obVgnmAOHL|KxN5eznu@k7GgzbTGm
zVXf&F1{M({2IZQFQ-D{U{2?E7HuWgACZRlc^9gX;Ajo|DX@StcFjn6IwKefZa{Yi+
z+YC~{ODQIQzx{1I@3X-8B@(=3Zwvv{-m#;fjpzVstz`xlzhx?(px1p|dzB<RH=#_y
zQZvR++tDjL%9Mc#i^AkU$xwXJwkhG{$@x}4G)fu=4q;t1Z3|ps$L7Cp07Wf~_$^=*
ze?GVn+L>@g2B5vfwof-*YFTSC?myQ-3=KMVPFqo(+2NA+6>&-;!`5K3^m76;GAVNr
z@fs)qm-!0tF12hpnAD<3z&|Oocm&D3^g_tK|IsJMjR|Ms%b(Ri1-*Q#{@6QO3R;IC
zpEK*tjlXlpujPr&yw0_yz$gvRMXi2W?C4u|=&i-;1mlqAf^fQLyzhcNqV~9ulBZL@
zKU0Jy<SpwIIPs2JK8yQdX4vWhHLr<!J(=%AHxu&M=WLM>g7r{P!NS8w@n}(@Q+^a8
z@z+i+s4X<YlGE2Jks3#zQjEn|%i@M~Dp6nx;XxlbXIS7CK1&+PrWq5yL<90Kequs{
zoP;*NtQK)h*zUlb5!RDbF_<Y_Ta6TGJTw<N;@6?1wxMX*CJW3{q(Q=uN4wX-e@4)<
z$>M+T*e*Z@#mmPC9X*|4%cNW>Po&+t9hL0Xx+h!{sh+6Xz8@ZqDCge&{YfV8-RqSv
zs4^~u63N(L90)xa*xmKNrHYjkQ8w*Uf{hJ&OcOZY=M66Xs-DRH%OQ%ifhk)~>-CWA
znhYBJzwuA*Hg-vuo8CQhQ<@ob?lxFzsdnUNe$Xf+8r3qCIsEe!koIf%zKk?i$$Hrj
znW0e>MD1Aq?EZnpZFmFGgC<O`iepFCm4c#?$jkL(3m`a8bbrN<J25Ufw=O<bt*J+r
zx29<sM`@yT7Wtw3R-+dVU8;!ZREnj9P!DV1J)e?pbK;apvY|`QWkDcVG)ph<T31;D
z?Ub1SDEaFjF2>Z7S}E{G(*6jGR5MQ#LBt?UBpNeJG-5CYL$%%Ymq|sF-<3g_`64=`
z^FE6ET{{Y`z&&rM#%**RcuGP7RN11_-PoX4ItXVGaL8f=ksg%4ZrfBr1bm_pC^3Vp
z6?35rZ}%qj!^QO&tdo2zq3xYBlnd0dF!20eqiQ0ZUUK%nLMKQSNDrG&X-lv_6!6U;
z$h=uAaGHEwZVGttt~zi<`w!jsfU|KuO2OUxB4+eeb4xBI#H;OHhkhqfxYBIvZBcnt
zJ<(<<9t%HoozjtM@YvP_8g)hwmh!LeNv8Ej(iGM|r(4P1w`oewar}WTHx+nZ!DDZA
zf+hbO0erG$B-x;Sl2rXDss7dBc!8@di3~GM%n=#E&QsTrgUGuFsnf*R*uloM@SoPd
zB0YZxVap8KASm-Q0+#OomI4B3Nu$sOh}#qQ=B@i>$~VRaPrZ9_1202tx}B(&5GG+e
zOuwAcosz*55T0gha$vH347^TxOYt@K$kS8~9y_yuLR&XZuLZ4C3|CpuYa1Oi`<?kI
zY5`I3X5b|)g`{dAZ4RhMwEJqo8?|K~?>u-S6Lu(_W$&1bLK+DPWpHuL9N()!?tatY
zn@R{jq*cqTz=b6S0KtS{Q%{k(C&ILP^2PR#EThpFw&qJHla7G}*f<Tf?ES+5Vzn<}
zgp;d<`ge-PfuUtv4BQPjHXHdnw!p!hY<}ARzMQu(axau-YW`orKKC~27}H(MS1ydw
zH|*5<1)*YAW1!lMsiEw>gtCfC+w@)I-d3@B+7sC=*TU9uCelM8Z&C8LBXVu+Edfu%
zNBH>54R8I3by@a{Tl|N&_IFp(Mb`I8!#L_N23|He-_bE59?IXxlYdb&LExg-Whm@J
z`)*uf=l#S*m}Z0mCV*LrESaHjQaC_J_#xUG)fXofR}zM1Zuu0cq0DIxbccFpKkcba
zr!pH_m*zFspM<@Z?X0G|-*UXV$15T-Ivm%hZ3&7wi>L`C^4|7p70<hyAD(`Xx<*^k
z;_17!{k18z;~YbDTWe<Btwe(qvCdJ>SB-mXnr4mp+kqedm`6L;GLOY}v0crVq+=;O
z!KI$B{`YM6%*vitRt4y~KWpC@9d8_}6d^u8R<Y<_zp0~q5wCT>T{jqN9Q^u~K}>1E
z=4NN5UGc<Rc2nx-ZPAr4A#P)S_c8HDY)Nvk0rcu*q0Nf7>%L&YULs)Y)DO`AtfqX?
zxP`~BvyvJ=nJ1j09P#r$L8iOu(`6&-!9XVyjW{HP@q&#QLz}IUT;ce#tMnN(pR$1_
zV3xOXm)jlWhROowgN;@f@acdQB!%mHy#OjgjH*4+Ivw-u45#D*B=oB+pP7dH=d(U`
zpharVJKOl}U05}I6P1_hwYF(*xNT;t@!i#`(n#(Ea?d)GC>~4_I%!L;KH%Wslr*sz
z@d!NA9xJX}cY95ixnpAH?zC#EW%iRuElk!S;ff5#LL;xEv~nyhy+@vPi+_Z^=Ll^$
z!@E)Hra2-dK;JGJUi&;KFTP#Tlhb@m*jp|nNob)vDa_CB{XHD5)z#-`;^&C%G?Ta)
z743b2M(7!LmBm9Qix;-4UicWLhWRI*O8f{0A5O=X=I1jSTMjqEu52YZua_pB7EO08
zM(R#`2QeC+Y~dh2WwAQ4;$7hx8^Avd+@*hT$y4U@%dd%)Q!a$;RVzAP>fH+t)(u&z
zVF6{6k^6<Lal<bho=;l>FK*CfBZp_L5gjud`j$S-|AJ%e)M%<!N0r-J|1k<d5cJ7C
zADY~j%@$vB{WU!$oRYqWX{y2f{?_81xPf<>dD|B${QTO&?DxYwJl6MkEMJ+-&eAUv
zY^8|(W*S~4qbT0bw`B*al&}>s+?;N?qIW)aRJdSaA-xSk^ouyKYEZaR6FpYG<quuD
zt=4EkHOLph(EC~Y9!tQA<TkN<KqU9R$`Gj9^nh*B@e5(6UmO|3X$lK`hPR{lXD_Iv
z*mvGU6Q3vF&8L#(%N`5yE!1F;eZzA4b$xk>dwSU<{tTojfznc-z@6d3J8`^|{V4(v
z6BTCiGNIA(sg8a=i2mTU@cm~_RXEEx>INeon2xGGkF!SNJJ4-A0Hi<-REnXpva`2O
zV;-K0?i_sTJv=*;5~PiJBo%r^CvMArB6R-P;xiBF`_qE$%q~w&O9h{(t0bnC`oK$%
zXS1T(dV$f19Q<cjyk$BWTX(`4U4Df12%kPT!Hq(MyfX?%WjQ?aWFc*IGl7={@zt>x
zm_&M#nD8Vgqtssg_-MNk*SXV#heGPspZh(JQ~DLhX+I@C%(GM!QBqs3)MDP89oi1+
zy3-~0d~^G-FBK{GnwlyBmz8K%py<VfU)zL_K9wg?Y(TSLFJCY`L#BrHtF}f;onSvv
zF8*)5UTa5UqoZUfQO$w+kP?~&pRL+o>>PIM*}+4>xC#KA{xqn&Dni*ekJ3##Cc1sc
zWcBE~aPEC`V|YEaa9y;RY|?RyDksJA9kuo+^&t39>3Q!&<eYy6oKIF`;}7>#LnLc#
z!O!tW=3QuLWF>9a+ADd`NA2Iy;?3$!n(N$i`6tB&-cGK|Cmr+g@ntsN{D6UzO94(_
zQsCT_7~5?xa-6$e_|TZivE`!p@DnNusGj7C|MQ#Pq4t>0gek7V<GAAIK@Ejo?4wl4
zmdnH?dV_}5wX8k(0q@hGjJ;w-{w4m=JXY)eyr=otXG56LR}EDTk41U{i$1?1<%L|w
zrFZ#KyopL?SYm<8wb0K*UaC`Ex!WzCDt~{9)%Jd|PoO<H{vH>)R#Jr(?|a^r$wDG>
z<-C%37|VotuMdK^UjFYWa>d3r*TOIpS$n*w$zkkmTA_aT*5UVGSO)@E9xvh0gqfzi
z#6NYqjw~4sBudQe?$K}WlnhqDcTcdz4qW{#GVWe?vRmFitd%B9n?o~;g#V=^d=pK7
z=x-<Tr@<foegPCF6P*6(5!%tO-VuD81ubv%uFf5lmc+yz#;-DlV4kt5%&~*u&;kc(
zL14aph8ENz513d}$k%&;&5#JluKWV)ixeH5k9Wt+&=k|(F?IqTryY^Rx`1(XrTP1E
z1*mfm>rW8*;Iy?=-c;*w;UVL;i~CFETf^$kuN7LS5P>upfc3p8{$xsubsq}33z;Ur
z>euWNVyU`2;{cH{slZcdnFt4pTj^v{E7NC}IO7Nm7ta3(K?q6GS=>_JICj=i_(Df1
zfcynYcvUAUwLKDs=2P}jA+gl*&qM%sR~Cpq=`PUS09DFzn*b_e!vQ==RXBx!0LJB6
zLAUrl*7#<!RQ%XR4o&-PWW}G<F<>Z1-Gvx~Cyb9<qFh(9p;;c<{c&CI>Qt36wfIjU
zT~o7$)8+6xyWAgN)4gUD46a5$-p9NG_V-6%g<4Vr5SA}!1Q^*qK`o3L2%NR6x}3&l
zV|bW-Ky#TsTmYVS3Z6gKpRY^iB}E3+i+-9f=dnDrH1>!JK9+rYI7x9)04)1gfG(Q>
zW)VB|q4lXh{T`BGCqJ$8N)UYO)H=Fvoa~9+wmbA9Dd-DIZSFpW$zSdR_Ypd+Xpm*W
zprVfkU8OzfDuMMeXlvrH|7OG_iCeEPDE>)RM<xDGIt4l17#eYPbr?6s1_pooM=*d^
z$}jMG2fDF;@bsdD&^VzEE6Gp|C~f+OTMHHiqLc}bZP4D=w%`|UmlvYoyu7y=kdh`(
zv<H)-8vM7oz$FZ=2|@5R=L8Q7ibol|3y7LsM$c4yu*-wX2Zx;3Ncq5oHdM;ML_uBY
zAAXu-w;b?hNi4~>Q>aY9@%LLMJ!375IX&`xY6X5$Q=0S!o&0||ta%^8f)8CdK$F(^
zFMFB+cOAtB6Huv%-CQ%UzGaS_>?Z|JV|xjsQjo7WF4c{kX=>8lL`s=Nu>X%BAFZ#z
zLX|&}vsVYhB?gq)DpcIC4DODoU(V<`vW;Y(*u3xJ8H>mL?xQS&2O94^PHSs~nOIW~
z&x_xv2b0kR4(U+*xidi#vviP*6Bc>Zbu$N@1QO^ZNTSlDB1b)fmv#r&Gru}@gv@4k
z6A1}TZ#Dd+vOxUHpf3=WziAq<)MBqe93K*>a9jYL4*r*L*uk3z9i6z<CiodQJ|*JV
z6=G;C(~+m%j4l1daVliV&UPY@BwC;bK|hiIqVE7Z`iDbEu311b`%TO<<r4yTjaU!p
zjRiS2=2&UTr&*OBu^5m-qB*I2;<v>ESL<}CsMdw-kY2rbx9IY4AYb!A-?2#UH9=G!
zN%^j;sO-M)mNZ~A|59-fYV`mf+X-n2>n%mG-Ei)!y$^-R&)=EO<EcKY_~OM3XJ$eI
z8$jlmd~KGv#UgH@<C`t*scmMem2wr_mFywxiFp6OFwpdC%u=X_1{#@Tm+?PG4`^hN
ze|~#|dNv2hn^bA)jW`#Cf|7Yr`3bARfTpFk`>8sVw>h~i!OL_sC=#@^FT|-|kdV9}
zp`Z-1wzah-OY|tWW|1565Tz6s74;=iud}{MI^RgVFXo8f+3Nju&K(jG!eP6YaLr~g
z#?|P!<Rgf@%MM#o-$oW(6)?-xCiLLpBXr(BzqH7>HId$!s^4`#!hWOaJ@Wttk;Lt<
zFAJqG;MZ!9Ma|DJ0S7E=PyhblAKww6peql6zMFQC^THq1Uj#_&pIq)?6@fkx3_?_6
zxXI2h{(ctpJ1=qYHU-lkNdFwdA5UtL2Y=y&)G+_o`~UAVzgOe`^LcYGerk7j<IQP*
zCHbl2D~1hDP}2v8jbqVNk&u_NoSdWw>OQ*#pnIg8MkDrNpohvtw<Qzte8<t%sfmm0
z;IkHmg~JGgB@DK3!JsYN%<-UX;Qgg*|B9<lD)jZ4ti`*Ty`J;Ir~bID)H8+s-Wa?K
zyKsm!b91587Z|M0cQTt?pQ4}9UHCKRYG5Wkc(gYMGPmao!}o3PXx^ndih0<T=zSiC
z4}UXS7sH0GT<+l8t68JS$(808vj3Fw15HS|nF2;0$)oR$c9ftsSJ((;xcK~8#i1oh
za$;J49>jh*%7t)C)So|T0n4rF%ItbIp3P)+vkYg=b{sUREwLnd>$s^OTd7sw_(Ay0
zoa)k2zoj>%Nt0z?{C?1CwbwiEXyfoHW$3H26LYm5>vpyq6wv!t>?$~c_MGS<4(Hl7
zpX}1g`B79NFRyFF9b`H9F<v9DbYeZ}mwKJJ##hKIzXAzi5Lr_mxO+cU`bv#sw7n-l
z4=N*yMU$tjbbWVkA`R_!Iyea1Ail*CUwbMegBHu^SO7vqrP|R9H@l@sX1W`evs%Ns
zr7XqbSlwqDu6c7So|EcaF-6ee%K1(k2z#OQfBYe)WIjP#Q)WO?jF`+{G+#>%nsY}4
z9s5lDIyg88?F_S5-Spe>br#VbsD`8e^4a8)1Y(&0b18Iqv`TG%x)rticsKNe$vb7{
zUVP}tg}gN1X*;0dd8pUBh+Ns`k{M@dn^ahOr=P6Xw4PXzwzOI6%Evm?XK0^>%cTf%
zvqAR0mhUo%fxcf9NK5u2UTRL;j)*%Ja?ZuktlFGLvPJnfE?cSAc&v)iGTW`^M+Kub
z4`8(Ec(pLjGR>XaT08n<v&^<D_gG|FaJ4aR){5`PBj_!U<W#r|EDq*Z+6~6`-^y(%
zy`d61(zDX>ZnC&tiS3~`oN>xgrfLK$64pfe;Egj(mJ9?(a)(b1*@yRdZ#jeC|A$=Z
z=U-Y0eUxV~5$i6m@RV##6j7fez$_aS(({;|_ixlZxqTHsf2bF*+T0V;l;k=aPSpqB
zRY^fcQTsvR*tlPNUOeT8N`{>~ft|l7SmG_lE|@6x?kN>m3hf)#_FSobf4L0Fhd@J6
zY!FEnbipYqt@HvTgB6spBP&MK)&rdf+izY>mq`VoHw-&VH#*!_ZoK_Pd`1}M>>J#3
z5=Ph-U%^Ab{6HNiLpEzLlPj_D<{JUKjZYJ)TTyqv$X%#Q+FFc}X~`$MSXmA>p1T3(
z2IUvk^6FMuTC37SS<bX<b{A-eegu<}iI?O(2rsI?I)2Z!TJ2%kc@^PXskf#1Kx(18
zB3`bs^);pDjlBKBvFIDkdP<-`-<}Vq)fl89trS93jW%oRb^EGdDP%qFZEGQn>bv4e
z%X+UC&Mk)GtM(->eVq0Bx4u+Dcp9!8!qO$8By%Gv{h*2+3luv-QB;ufiMvpz@ewFi
zI`pMUA(o{fFZQTuQJFWXVRz0|fw6kvG80Q;Ey9qddktM8>-lD+4>qKyySCsOkp|{c
zZnhdC(=c870rO`FS;u|#>*@2Gmwm?Hm}RC}=bjC0D(h{Zze@?W`O5e>zRWbh=~xk;
zfy;|b%RLI;HPVddjtX?Qbf<jqP+dm9)64s4yJR?(@F=rFT%w+;O<9uuL6OaELl_R7
zMmLU-WtT-`64S7x+9=(t3Mx$HiHOVF_>-NsyrKMQyUAL+>Rm=>!*o{}#+*yId6`G(
zpXndMpuEZp^^+rj30N3kX`>#`kX?UApDn?4$tcpk&^Q>>|HEoO#Tu8fFaN&zz4kro
z;=y5=zTv^h;x2{iNOq~-IGyFn3ftz;4=Ss8&%^q`V9>LSLq&ARJ?d1!_O$omk#tM-
z+T%o&@7%lBoTI4eW}6O=Yi)$4eXyQe68Q~xQueIe!D=#ih$MN~;?t{lW8IJB>ohNF
z^4cOy7T>yktaXqt)t96-X}A{y9SdnQkk>tWq?`8V)5rEph(83C6*Q;usdS%XK0O<-
zME}eWl_x*a?>#w?XEasX!f;OK2V;@C&*efhSy^I+RkOuhO+O<7f;j_@*)w(1;e8A`
zX&M!t^LrEtpU@dI>2VAhE}vgNyZM5)`CTum%3XDy*WRThtt4LLt32yuS?Pf7&>b1k
z(+?u8)P3??Jc~7OXIPRx({Vnb+^B<Zvig`LKo()Ssz{o$U_{rg7jIZ)P9pmC^Zn0B
z@;7%<#<xUj?+S{{6!HrDj#qJ##7bZ$XAyijgqnmNA<}$IPyJXB^9Y-7JSG^^a0-%-
zYon2);4+P$DtM@Mkw@I{p-3Yj7_T$#U64IqYMPw&Ytx`3Z~G%`Yk3`UQFP&2?@y^0
zU&w=PCGD?vY9=4HQpaVTn17&tUb|=e_1CFV(i5qMXX+O+;{?x#3fxzIMbF*7^|0?K
zaBw!Oj31d-A$W)aI*#&jcZExXuBGOg?2m~Wc|nU~GQv0o4_fqxiMdQRSK{`EN8XRO
z2cJ{(zy#mDBz$L|V<&tVI)+LW(@1CS8zpt@`_*>m`Na!2w+JM3WmXk$!N{eJhT~dV
zk~*(PAJf=ZW>hg|?`RU|6L!<eF)}E<x|}z@&INIVgz@}R#~`FQIGIGbOcj+jQYlKX
z6S{VUSO>eoS!*2AiO4YEQ9K|$9h8MpnjWH!t*gp?S@KCZON9-6U~vSt<5xu5E4r8u
zc~R7zKlXW5zwYtN7u_PV3p8Ia=S1z9rmMwog|Rt{KP4!Y%F^K^G|Jvi-KdqW4fZbF
z{mDOaS3qzp<*@G_^7uJ}jJ0p&^TGPw{hQRJceSL?HRYxG7Bq24iZ8~yE_olJBNn*5
z={(z+v^pkFfORi^Ht@c;<Gcj4CKLdW!gn=;Y^TC{Jiu^}Gj<7G+vK9Lu0$jxDQ^Eo
z02P7;!qt7NZ$k)$M#4m8#q+y5RjWRkv3Z?Z6sdI4u#d;^ubE)XNqDsu;iGzKN&g#l
zk}Md51GrF0Clcr1xRV8)r1-O^_!>6GUx*kAM{CG!?;kvPo-QWHrvP?46WARe^4@g^
z`~{4EzrYP25nZ{OkHRY8pP(7~x9T`>0_KfAU;S^ijn3PI3KVuGIlM;bA0Ye>1pHCT
zt*K?yIPmCyj2<40-u!*n4C%kIap6lKurZUc{eO=hj2IS-m?c}+9OK^te|QQ6uE0zC
z@jpiF4n~ZxKwsteZz1!5I##O&$!h$2#525L#HduUYJz`S6MqN-gdBzL%7XUSbpIal
zKbFYt7>V=5p8TBnN_@x`)6Z9Ag$3E}_{`ZKyHEayP8(vG$3ELdUAkDTWH|X(9Q|?Y
zOcbsId-6nh>8Lxx?an$e%m&55or1G_2f65iX^ONfM$}U#kSj~!kNGpqP?pPV+-~Dq
z`8%h*z6+6(2)MGL=e*e6M`G_$iu>X1nOZB&RAI;sm)*%lPD0H=7P!AYraB6tMp(`P
zI8}?5Na<NO_U$JUp{q=}B=q_LRnU=+7%F!2CyLdCPmD2FYP7$!XtL~y?kK+_*MD48
zTum_kTZ@_~f&2-nc&{G;O^A21@#v7&d;z*|Mt5iMI4+-YpeBIMT<OscieC`|RssG>
z*Fq=EP93ciTaL>}cc|fSuFcF#Oi%19J-zPyfiz1l^;OqG(aA9n^lb<`WKin9suAFa
z^aXJ|>D<Vn)?E#BseE93?w&@k)$7!iz?~7eJ5Q=G4mrt04!ROs#>+^T9OTrt>8q*_
z-A{X!>6vuAjOOW_gF~m;?GZJtyLEFqRI0(2qrQ2}0-SS;@Uayuj{X?45=2}sq_kVn
zQS6%7sAs3J?6Mp;QE`t$Gvpm4j?(=JzE_%Wb6;+43za&qf<6Li^DIBirjfRMf)iK`
zx^a2TO|7;rV)IsdFM0HCnv^ivEYp+cer^9&yJ|0>Y!V7jy#*h*`@PhH!*<Z5&ODDu
z-~U-P`$-fg0KaO_-$udpBh#A!>8~~$^*SBT`d<@a1P#r|00!nv$zpi($9>a>22=L;
zapROa{5jv~G`lWhsr6dVJj1tc$uX&YK6s(gRPJ9Yj+xF;FC&ycR;sH57xD=nw}cZh
z7d9c2e2uj`3rv7!@b_ru`d8gv?giu`(bkp=Rq{hA_qwt4`H4uEb8CpwS}EGy_1E0t
zZ(Rn+bmlU5AFXp!u@Jrn4RfROaqk-}P<?ZmiFNtE4<)TX`9;Bs@O#5$fkW@}6;-Mm
zA_+PM9Aojt;ljJYl-;Q>gu_KOf?rS0<z*Y)y={_DjgG&2c98RNT2LyoejCYtuYAX9
zD`bt=G`d~2TCeNMgE6+aP8t5{Ok}KZo<*)imz(m$NNJGr)$X-3nHj!fWS3Rr)5Oxy
z^N2vX20@LkFt)_q`l-o}6<<OUSgll2g@Q;4<D4jkczVp)=gPjh(AA+4l+{=9jV^&{
zN?)r7<tQQn;OmMF_@Fo5kQPp&VVWbG;-}jaDcmlp`JyqY_O*SfL1?ZX%8$XWneyf2
z&6)4q_gO1f7XrsCGqrS$EY^$N5C(psqKxq(rNYp_xYC^<3~cZ@l4B+Fs3f$cXtZH+
z7Q%TME8WLOcRK|)HtZFx?2cXflZV3=OV=0vD?OKC<%ODai{)Mw@#o!D63kttl+Hhg
z8NObAEI5a;r|<DGLl8^~MPP2>=NJfnLU>dpaDPkWBBBg6xv&eTUS2;*jC{7=KqhqL
zN>f2~$EBLG9l9CLu#Es9IH<D6L&W-WSwBDWVd9o}CYpR)k`%z+Bak{_EGd$eY}74#
z)n7H#mEG=|GTrqE4KD+Y=Ce~jHaN%&OR1X+^BHmppMQN>)?tHmvVF6%$)IvIizZiL
z%E!<qvfbn2)dp~wpeIRUr8eOep=iSE@03~dWaK_dT7wlriYm9aGMQ%E^@a<&&U7P2
z&AGeu_J(T5vZ}uGy_dgk)I^+qo(3Da%w=lBL4gNu;&8I6n92`*-;zsKdA!Y`)URgU
zV()jwrsvyt5mrF&ZA|LSZ@6Nk<w{2Ku8l+-WT3xxxbnL+iB@BZ48aso&i)6)@}Nqd
z-pV?QH_`q-Xa5Hbpo&p(sR=+8Qy!Hs*IMmsZ%&*@<P6(XpBSqRpDB!R7ekHx@=b;=
zFI)&UPDF>yckz9UvUHJcJh_8mn$k;OP2Jw|)l5%JF;X#r83R5r-dkN&FifYf;jA&_
z2glkJGhTGYy{G_jTRmPV^K{RR*<0(*2;UuF;w#coCh6Y3w<Z^viWHN2{Z4z0N~KOO
zX(U%!5a3fXYL>g#KonZ54qFEI?gP+Qrf)5&H``+gjwkHaSGu#eEnadx*R(`NYTaUS
zYd7<0W>S`k-9{~jI=l=+V(U)qaMuzO$LdO~#9ADl%LMUM&TVPe1E?Wuq>vW5S>2~n
z*=8MBK>1?4a6X!SS6DR6iC7{wwPtl-kb&N;*&&vpFfU!=)I9mzz+#o9_(K1e>R=L6
zp8AwEJE&yvhB#g9%`N-L;W+ncmuteryA0Q?+B=P5lI0EmW5DA%gb3xb@6DL_R=6nM
z4}0&xqU@TE1DllIIi^Jtw$^c8q+L?&1(;H%-Of1fgVTkvOjFp7J{;7*6Q%-KF7feQ
zJrjn970sn>N22R|!;*ETDpmPN`cH^?sNs%vqX2G=b>Oi*jI%b99SwXi#E!Bzm^(gx
z2ZM4;gPK%MtLAGn)URH<W>Mf5iZ|s-S6i7+g|uwnGX&v@*>4NhOY6lWKih4tb@2|r
zl*~@u{RkV?uxrw@ep%KPT<@E4zflwZV)t5~ee4*APH*c61Zs0sI-uoq85JuxXlrgx
zX8It7-PL~iu*YQw+K;WEqb{qS#E{IYDY9fbNG0N>4qM>a;*#a*pP_vw3|_dhBKm#b
z|AGRs#}&M|01E6rVl19$SZ6s-SK@#KDD8)~&7Uv|^zFh10P|iffM{8-Yh|u-5PqtQ
z?aNiTp@Do-=9ZsZa;6t6GR+Q01N)~G{k$T{gA_s<S^*EcFOmjbcXnqxHP3Q~RB9QE
zNU<bVS?G)a7Vmi>S7iC_m;sgMjEMbBPu6MFYa*hftx=OM&64_Oorx+wr>3TJRvQ~e
zMzfBo`Ik1yDt2=Fe63G*cQ#W3<f5}DXD<w2e{wY(NRt`GahjcF&J_@NdX0f?_01v-
zl<Wh>y{D-Osv$zd=ec}+vrxA^8J<_kP5iBC9#I_i7;jYgdVYZkhY36L)ewQjLv_9(
z;)BCZx#XV!g_fetqCt1Ft~KvgQ+aPCiz%|F21QY{0X|(aJj5&y2cg7*_%k$0#k#DO
z{#Y^Tgc{wcsVEbc4CCYGQbRcz26~cW=Imw<X``*V+S(k8YXOmI0xk&mggpPw{PyeO
z>`{m8Nb0*BTv;7L)&A|OU6#eMX#yp!loxI;K15^Z{31SvZ?%Y(HxQ%S6GeJ%kY*_$
zU(+)aaf&8qnc+nHN1Hw^rhgAVN|Kdf-_oB*Z!0x8Wwk$C(o5dmoXA00<M^aLe>X?O
zWB)Sg!!{`(5Xz*G-p%7DiO8iYq500oSCuT*|GwC7J^-wGAH|sjpt3nFc%>oY`>ff|
z3^z{7A?+5Au$4qUW??9IfU=O?r)uZA+e?c;nRglcRIloHXG~*<#Ls$-<~O|52+RCE
zr(?!Xiut|$J=G*CLx#jl@Mbft%DSP{4a=5TR*GSVCWo(){GIC;3BuD9f#Hg>EcZ6h
zYSC=bOM{nVn5i4562b^2y|D04Sx^YJ3W=`l{hzi+ozLCgfAYQ3tuyH{1}xPaOLy{O
zun*^1;+O5NQl=LEo-{H#dIM2C$0BLe!5ljF-51H%W!=U>!X^93-)lZAg08dXPJk3O
zFLb<bCSyo#;^r*jers`QL!T_qMX5wm*H}rUpp|m^<>o}lcOC=Pj$rqbHnFlbF?*{Q
zf@d9bvARIk)QAZWzmnv($(A@rrRsbU1X*C;Kl2S^x-xS`$j#Zc%Al4c4nArw>YdXC
zn$fgvIJrauZ|*lb8qH^2MwucT7-Q~25rsN^<%ritq<GRs7wfjQVxp)8`)z@;iR3!9
z=eN#u)+-^-jsvpP?8t3+gnHhfPc*<;XWFS)ym!T)0zdw+<Qh0jM=?IbWtt8=7}z9Y
zckZD2!0YaftvWp$YK|(!+Ts3(CE%6^Nd9UER5=%5SBO2lZsND#P0D26WgoL<v^MaU
zBu;-z_7L2|=3|(}!bHj}c`|JfE~G{>8tXoHzU8%PN$YE&MC?UFFYjA(#4~rH=)jx|
z-i+YPvb_BY9_vCEK*^brk>!t{h?Nqt-}1o*tkY?T?%PzDT2$OS-KIEeg#i*_?U<3E
zE9$gwmV7A7asJGDsx~H=NV&o&B>y{GlZ~@FU5aWzb0y^qc8{gAS>HPwp}~}6_t_+k
zx5*gT?2k1$ddIBM9l!lxW4N65VJ;5VdGGFB=$f2{L_mm#d@-xiPE}PQ;U_Hu<zI1;
zbO@Dv9h!d<g>Ti9XA(iqpJ>i^qz0%P&l-`W$rF@w2_K(Cniu54O~RjEp%HP1SE!wJ
zr;bkdbx0A;&FyG`#c6XZIVN{Wn9vVG+#cAZ{dV``P1-fHt0#53PcZgu`Mt&kD$Pl2
zIngOmh5F?MH%20<PC>?^;h@xY(^5#;H_y@UjCRAa<e(oi6JOAqTrsW1@WwMrM(F9v
z`x@95{O;{)z}9lF!aBfL-Z-@$Y=x+u$1`Cc>B?BZplokeq-if+5p6?6<S}=8o`*N_
zbLdaK8yCWbO!?hv@3yA|2(1AY<41a8Enb0-Xi2|iOwH8`RGM;~yhMY<CpwKWpLpAA
z$2<yak^xA4c3SMBA5<JuW0(fQDTgbCedbMoo7h1ED9fnWh!AdtJn=}%c#e8bcI)RC
z*-fO`CL9j402@tgQuiSSZIZE1NRJ9BIEKWSWx5NVe>JXzAI)K7F`s96AsOa0+1nW<
zZ$KB%gvrj7XX=8jfL~4*u&0pH91pKv+u%3W+AktKjh*wj%b6_>Gj(;k{yrP9LzDU2
zP+tRsF}Z7D$W6s)s8jTJ{J19m#<HrtnQw7eUq$00yL#Di#xTHEl;^ir75O|p?6ZrY
z3o>9zA%}<zeInzmgGSnCD64m_RHO0oLj1<fm%q<2)&O9fl#M)E1;$(z=I3Y1h6Y7S
z@hST561%*{T8qoq6Ozq#EJplgB&mgTDRmwjPt6j4r36rKDfz&1j7RX_KYn;#7wk4K
zZTFd7A4SP5Lac1O`#>XTby?C#DoYa0nZ77-4rFo043hXSb%d>M+r|bLGJPUrboXZ3
z<xNV!xu&VJv$KmaLaJ5s+xx2Rsw{9=!7!VyPr1xw^h>qO<SUl4VOAYqhwMulP$EA`
z<&{8|_0OQEJu?DY=QtCo@$923fCnEcw=wF9CreXdSXh~)e&RciP$?~Jhk!kH62gp1
zCAyjFN>`rCA+d7BT5qO~X;L0I<<j^$29Po8`4HR_gYC7)%qVQ|qwrNb?Bub-S&@QB
zOK}x+s%HsWHtp{n^CAj{C9Q)n)zNM4y3O?Y6t6d_f*WpvVNg(efh*$E7~G&YZPm_Q
zTe*Crd?f#I4>{{RO5+PJ7nTU}uLd^rgUs=qD*TskPaR*rN2Dqqp!I<@kujjtV3H4}
zPj}SRXmFyGXt-|9hm;vLf6%WDkrtu6Qh8!DoBYFG-3%7*Ar>CmoagWJi2;x>tOB&}
z;lcUD@zCU3sF=Xik1zX<u59L&RwIdcjL8GbQ$ZYSUAa$CtDX?r6r!MQJzab86gw_B
zg$NXW3k|uOj2^mciHd&`2|p@+Q!mlkS0|TqIKM5ju>sGOpR%t?q@KoClsB#XMze8x
zry<3+g2>Z{g6t(7mh-0~-(bs)p&z*`J+WOkRSlWjmJ>2agZ&oD_32Ks<lopt1mIEO
zqop^=x*}dO@(c&(`w(Bn5!GU?onR2Qaa7*ZazaXaiW)QwZ5jsze7m;eP|nC?&+aVw
zU-qk;cp!6Gt9#2j$Hf&;z}E3NQtt+a#Fw<>uO|Dxd7B&HxHseYb1fAdV~~c%>C*n4
zc5TSp-E9F876J~)+&rUC)g|e~H?^X)fC!MhFruG+wu|?HzS)bY*2McwWe(BSrwY%~
zQ^)v&W)~^HrYWuSBK??6<gt|Pf)i8q;gRWj1n!}RdEKWkHxGy74)twO)Qf~fupikN
zYNeP2E31{Lu5UeIX23RMJk90k-{8Wc3tcwb=ymiwt}nJg!&^GVW$Q3)&D$>8c=inS
zX_>Alt-Mdg(?&B9o>t^%)~G0`G%zMI>{l7vp}3Vcndinq4ud9}(wultwo=r#$MThb
zda+z_S28(ouQkE_!7L>R0>Vj*%HtE%Ma_p~=qoCsZ74FZ{~FmDxp4c#&%<Ec5k3Bs
zTALk+Gc4iif^D4wZOGj+C~BRIo{Dy8RqBI26+7(x!jt{vBbQ7++D5Udb)Ay)k(uiw
z8&`MdN_fuN;%qXUZRaVKq8aPZ6I1zF$%2cbC$u8H&)xCis!ShN@G*^1G8`_oAB{7g
zPO182f0z9^JS>mwGC4?dmLnfBw`R9+{JM8qL$v#NbUa|iJ){UJFMmI6UR}t;vEFSe
z!X>LBaq|gdaD#y=u`w~7wCPjR4fZ#{t5LFRi8<51&iR{UOHc%os+jm#{^3Bc{^1r}
z8biO&0UTdRDl9FtXo)S1i_?nwECh{<VR@4q&*GnJ4EBD?7}MT;{E5eMW4@M-P}_Rt
zr~RO;91&62eL={o)o*SN@xw!yfX}O|OMjeodo;;1zxqF33`^GgV{pZB$O~3G@_bnf
zXWU|(lJpmkAk9&c5w^{P8Um4&aqF6m^T*!`#crzSw7I*tG#fFv(SOq&9Dr_&gZ-22
zLq~q7o+ni0c4~hjxa80g$w(P~sYcHH?At2s?-Q2fAmo8NoYcdAlar4Uq=DF*xn3~;
ziW>my?G9)5FP0Jf_<#TZKNI?Y^Yi<F`2U^$EV!x}vTI1;dv`tn|HOo(1q=DLJpLbj
CUG9hg

diff --git a/docs/use_cases/scf_to_cps/imputations_model_comparison.png b/docs/use_cases/scf_to_cps/imputations_model_comparison.png
deleted file mode 100644
index e4675dec7238b264708163bb165783b4e805935f..0000000000000000000000000000000000000000
GIT binary patch
literal 0
HcmV?d00001

literal 90532
zcmeFZWmwc-_b$#1rKFUU0*aKB0@5uKf^@f1(nvRhl!~-~h`>llhe&s)bV&`}(%t_(
zsNd)JJkNRczs`$uuIs$;a@aGoKYOpe*4p>F*IM&dQC<ocivkM;1qD||T0$8G1$7zx
zqq&9wMheKJg}{Fh2W2U7l)Ua+D<~*5C^8aas;>IKCNUj}hsH6s8194+iZgym;rix_
zCUFCeSm{O@`i&cCXawO1@d!k)bS(YrYs&tL5@Hfe5D6BEckiw}ji8|MN4W#p@!D)2
zz1%S68NH`t>XO9Q>Tqvh<9t%aG%-%V#dCIcR=Msilo*bJ&L|4OQiT0|8H;uF0A4wm
zB#{1}?_e=PFdV6_{r6XHUwZ&wgA-_?{f}w>yhP;>`=68l9);ndk@Y(DUSom&$FsrP
zkc8I%daj}<CKP9x|0UW@-@mK;=Sg+2ZlizJ{?8~J6)g2O7c}_aws!Tlq3F`X{~qo2
zMM15=qwsz5U-H0u0N*A1@6lyZ7(@><r0(v&B=`5}WW8=U|2^tRQwPiQlQbmwcX1=X
zgtIJ&@!zBTHw;Bb#Aplt7Xk;zfcpQA0V%To%Z{Oflp571sT|fd^{X}0hy9L=9bVPG
zyAUzEa$W9|nsG|MWz?XDMN5=hhn<VVPOl_q(%IiBTO-bfJ~@w0PI??Z@E$hHIF@r8
z%v4gcRtf$7OU##1m1dCbw4f+r{H+Y{tJPan%qq~)!sFSoeY78Ej9B^As<;Xou1(KR
zqjA$yQ>RNs6lP{-1Hw6_!XF<-M}*i;l;F3@e|}<II>K4_Nr%N_XV8sB=M)!VrB9kh
zegB0+oX`0>+OX|tu**1)xaNf*cYZ56(_aIEeMb|#54RmV5TftW3mDoBEa!UpR9fa~
zMewar)P$i>azszG<@U=_>nd&b*=P#svgkK?wl#qh<@Uy|I_}!fw>wxht1OtEJuifz
zsAmbvTrMv+Q&Wvi2n<d8QwO9~J*&7R_ifEcvjt?@UB5J=I9jMtCur9uH&}+yd)>G&
zCjW7JYGBc8qQVOk0`cDJG#!5y)5EjbviD@|ftfD8>V1QoOKkdLyDlw;c8BUO_jbN>
zZ}L>5v+_}0o*EQwa-!STZe#J6<Rl^JKXdyu{E7~zR3F!P<~$~xP@Iy4cm7ttap}eI
z&kgbl$akmKFfvXfzL*}0_saMr&g!qqssCC$qv3a$6#5>n&f$xWsz;i9E{s(7)q73R
zJlK!AHVg_jvS0p48XQN&my91AT(}aWwz4**-my(lFWzw;&&%eSe+S#ZTW75$b$k3;
z?G0Drt=qjO21&MvfP%9x8G5;|yMluJ*TgnUB%<FfSZOc)DDMoXblj0pv*F;o57)8(
zkbWquZrdzdd_TpiJ()XtgdF=JJYC&X)TJe&)$6Y(3MT<Bk#cYTdS<S|j<JLNbI+40
z*$TYnY@YR|AI)n7#V_w+1uPVc)eGHx^+ePq=pcsXVk~$_m#~-9eg1Q;BS{y3Sj&&T
z&9Pg(U57InFq^K2DNc?pJ2=ZY%iX-)W%z%E@r!vn8lqPr<-I=%`GxQybF*J+a_A9!
zdozDXyifcXw}P;PH1`Ngp!Skv$B|9yfQI;NGvSF_tJCNYu}p38irNk$?$HuC6Dsd5
zoB8-I{RXhH=XE~=T(`^LwM7Y#ImB2gZZ00I4ie{52!*Mbv1g@kh*0zAdF!Vd3h(3D
zwyN(F5R<`6Xceksy=!odFpFA)I&11`lDt(!*e{nGm$%i8qFf&O@STyCY#yVZdh@@N
z{q`!0r8vn8%lgxBeQS2fi}8)5p&r?r?5T*+X#1C&s~u5olOng&*MyZeo(t+6KP<A*
z#T@mc`tbN`(dymZRbkT5H6EK63oW-TMARCopDgiEz>VKFeEjZ63(gzuIuDW3_S^pD
zalVEo{gT?5HLsuF*fFHS5|sBX71OggxDrEeonr3Bc=`HCT_|rxacAtl|5WYRRl-$S
zX1P%5a#`aZrZ%QgmBzeB`E2du#Cw@CH%fa>qmhfpye`}K6SH^F^lddM!$fv&^jq4_
z73NJ8pvISOYVcUUzVry8xoBZ^q~uW9Klu2dEamwt^Cf(dQU8c@L=$b=yUe8Y5&GDC
zU8&25yAP`rbk4{ea4fkul{bhmsso%1*6VwM!m^FZr0X$xj9T!1!4I;9#9|qSl5#KQ
zZ+q-e9Wpy5${bcZW?o7$muA=u623{J^oT9VPhOjj3uN0E`*ycAsY?9LK$p(jpIiI-
zzb=nm)be`6br~-;Wrl=w5w|Z5XS9dt`@imR>2|Q%%-(IybF@sXHbyOxqHf9^fBPo%
zQ|Ni^StxoZkI{}tf)X!F{>*jv`AXJTGyQ53mv&^NMstJT2CE+%HEw?WAa&sKXal3=
zNsWoU=iB2kb{C`i`I<8Ob$<FDJ?lQD4u*sNw5hUK7^{!jywiKi>!pL6Csii#cQ&Pi
zr9$p2BufRgSXf=N9xI%Kas0gOq+XTUDbl~NG2c9iC%&kq4A}ADekIK?R3dVlKh*O}
zea7M0z{`;{I;YQ_mXMvqlV7id<XWuCTsDQ>apL`g-UVcip#5OTWt|_oR^G?p!(3FW
zr8i~uh8pi8bU|q<YwFptTc>1Cb|Hm=S&u=>7<<b0c~9+3M=@m<gF<7)(BLa$!DjvH
zP;%;A4QrODGd%VD9W=>jIx$(n9_~IoM`%K7Dm$2eVZgFn-5RvQzry8idbZ2p+Hj74
zaBTkbrNf}XfVJezbDT{>g@j8!;Kx(MlXM^0gk7ME+&V*z?u>mh=RP!4v8Xbea4$#m
z@^*5~JN55(DFn$}v$c;>?k_(jyRS;aj(6^@Y^}mpV5!EMkBEy_ysU3<a@}vvVhm?T
z^T4~M)Ja}DlBK8Hf$kY0emvytBjAlSb$M2a#hESasSt2nd%+jvT+NWZrMFaW=QP8*
zf3Uw@-{!r(H5#~i!(pDCg!^822a3LZ1be5Y!f;w5NfL|SR9KQa>6DX&ztIoa!Pfr1
zlGD>{HV%b`QaSdKyc|SxER1F$#<ytx_346k^sD8tTD6ZkTbfK!2#m4Qrhdu7iLly>
zOcr`vFCDk`y;cLW-fx$u7V-U5gpBu4Wb2}sa$S1UJloq{h6*`&wGBm4YECvV_ig7~
zViwL<4(ZvIc+BtB32(PO{#mkS*3>-J#K4Jh*Wl#_*ZBC+QMffu!rsHLVXI$l)Z6a0
z4!$`QQM$Zr?*C@t!69S#VR~nMHFNdr8&>ONp@%VZPsJmOG!TO2TeA)`t4HCK1hY9#
z<hIKB6=O-7(%~A`YT1P+qr(Xz_k(A=#-^i5sV4RcxMaLi_511V8^tl6Mo|iIm~X-@
zTi<PaY;e4ph@HBCjjXCaUJ1g=o~qwGXwINv;$bzuG}t_oiB5Ozd*F2T6h0X0lbPb*
zQSqoTg4xWO#JJcksq6N;pkNX%G{@^b<@Okv)4w)1gTR%QsaR`z@qn25%urO~`UokT
zUgJBy$8uTa#``0hd!rgZ4@U2`eZ<vqK6}}xgs9~4nVKA~<s4J7VTw8sVC3<echOwU
z55c9G@%;6kWc6p}jw;`Td{v4~Z?$S=VMV=}d8BpcroY@!k~cIgn#txU!SbZTf$V2-
zfCZ*`m^p35^k_#IB{z}79vb2{KlVAo+-6%cWRbPTemAW{C2TsD)+f_Q;dRNx*@uH@
zN}NySufpDx9ideTO{4Bt4sNAc;Cq`LHfI=Z&zYlcy^Nb(HRHG`7R+XZQ=9$ePW`Is
zXc9wD)^W=we0;~n4>Gx$&uRAeMuZbV^0AlmOFR_xGJN<Q!e?Z@IzJpcT396;tft$T
zQmF5=#c6r`q_MzN+h<61Cvhd16?2U1XSAT$hxW7Ej;uKCQS+tR!<>9*_Z~c@8N}Fr
zTH}GAUv&FjBpufCz7I7PTI;jQ4ZC}NlQ;_Q3rsudYbDv|GS>zW%?A8l@R{p5Iqc(y
z+io`s=pJ-N{bEg$7nR=lylS>1v9_N#r>aQC3HQ76UbZa$a>bP7<~^>q59)eFZVoN`
z5@|d5iGk)~q8A1jxsli&aYf<j(Q{MS-(L2a<WxX7vJ3>&Fx-?+59;u5B-vq8-V9S;
z!%f_Z!17}@#aCBYFzFwTzEog`%iefUQx<DIV#_IkPIJAFLxvsI?J0wpB(w`PMY9Uw
zl!Hb6*rP;U<oht91((HFBUH%T#oQ;(xFCz#`&cg^EW&F_Re{^YV$}Lp!_I<|!Xs;I
zB35|kNz2BLt@Sjwo4EDLZO3m9sDeemhlh)>75cq;gmcLb$$UJONq<M#v_r*|Tu3k%
zre8WlBY3W=RjgpmA<SJ2sryk`MmYu7<`+_)Z0o1Dgy=0wd2w|-cwMEwLBDTFCdpGu
z2k#6)u?&r=7yzd#+3_{tr4jqFTl&So)~P}IH-S~ueGg%)4#NF;&$V~g>4gN}PyV2M
zI_|vI!VWiUrMD<KvyQTE4~cdPmb}Yr=Dc4jA1cI$Q7j{Q82b(Nz3nY>+o>blz8?=K
zack3U_daIWgi0&CJ7CpU&wlB*+#Z6ql1V#I7RNmJOlGcDiEwOu@8i+5WQR}fk}?KU
z@ScBo{8Nthso7FYIhAUmqN(RPmo^XJI%iIc_<wP8Zv;#O-EjqDqHOLOnZajE>pP??
zgf+%!u5Rx%YS+V6JKLL69KB*q`akt!EfY2^%&M@755&)lYbQ#3mH(=Lxi;3E=1+fJ
zVeH=Is0U7bZo`K&-Ng5Yx45N*&SSA($kKk-!CIbL7kcwtCFX_W<pKR<ds=t#-IGkV
zEH42bmCB(+ah<H-A5jE@Aqn#@w9n?9UNdbvaevUOjN4kt5z7j$ag2>|rxY=d2@uh*
zCZYB!X|{X(#eEKuwf1W)fL4x-wv}DE8owWVd2t~s1b_?d*!(8)l1!m?M8%8bx6cm`
zO|xy<$1L$m&>Tl=_AGiUEQ8I@xgYoC)H@tG_lONO>p&v~Gp&C<=dCEVe8Uqz6p(A3
z93W9%{sN&UPz_@<8GuFvGTqdeo6V(K8&mPxHF_eR=HFQat+3J&A3qYaC3hzx<9Xo@
zsrHFoe0gB7No*|y6h&X>!Dce+I$bqykx?rGf^<~M&a4u)XRhztaWPj;*yfmCii}*a
zDio!)zF|$eF<4S<GjolY?PHH<&as7cdccCMy1aVQ+kwz$Zs+soGn<-coF?nG5fi<U
zlsfiCD*DmkGL<u0x&20gl*0XUqV&5{2J4%V&PJk@1f|~FjKn`KYkT%`$BKx)-^}Mh
zbG$nFbeG{7&%fYUmvveVfa8qA+mG;RPl}$LlV*MV*>mCQ=WgJBn#GhQTcRual&AmE
zZcuixx|>v$)8X~|D6%(VXo?MwW9i425j7F+6bb60u{QH#MA~!F3)Q+S;*<`1?2qa`
z3pRgXi0=;=7lixldHV?JAJjMJ%W01mr5D)y3K=O8Czt$E+FeL6cf(XgbPdtVLi7?N
zI|uB`U&d@%3X%uO9+&&uSvp0ynNYD=*uV05ctG*VnId50HpQ1`1Ejd|F&kNr>Djmi
zWd<<ZqS#JbP<LfCG*>WKs;{S7?l9ZpsqxSR?%NnC%a^~54L;n1R=6D-)9y%|#x}O%
zK35x4su+&fktT~V-wY(rl6svB(L*rD9@;OEMKP532WTcnZVK6<#n+7HCx&)(p|7aX
z?0*=!Yjx#O4im;D&w4aE;`BvJ2SXQcF8X9}nH`@qsvr3n@#u!tW{%P4!n}Vqk?&Co
z8SLK*clkB`C_h+x=swvPP^KEYrv`#r8_%QeYGE(8DA*U16iru{o#X7!k02mIv_57<
zg)dtp)g@OxvF#Swu`0x3Drn|}N*%}dmT2ki*OUm3mU5eq4<|NQcAEQbwP!<i-UZ*<
z-Ki-bek@rzW9Gfv>?-o<P)g-AbA+-;re!)j-e&RiL%c(C@J+>P`gz+)*<8%HJ@LNb
zFkw3~zZM^3G8)B;jkYQ6bz6AJ=e@*;tvn&I(uK{yGvSAD9U)y_zrQ3`!ZlYl(UHVL
zIm$jYTu)O$R^9D13SBuiv<0Z9hdongTNI5O2cv?Bh<k04d#f6o{Q^_*_TG%Y{;J+q
z5%eT#;px{^ZWe_h0V#QnFVAdlwO!2X8M>wK?A4O0pgXP$=~z!KLiA3{tB6=O<7KN&
zIBFg~XM)Gqo?7YhFiqv^ybU~br1J-!BFAWDKu=pV?AJlXY8?C-_|lp-GlV@pz}B!<
zIUUP$I_2Q`O69=L2ff6Sbkbf{uhVhC4^K55$nWJfTNbOFS9HWfwMq4LP%Emi*`6zk
zAN%_GWYmwBc=A*2ynC`4K`s6|y(+e5F<37UW@H)v5;8fVft3hOySd?(C}m2$;2ZP&
z*Qq4ECOU-u^VL4b2~lvid{gNjEH^X0YM9Lm<?t}neN+C~P#4o&d)2RDXoa(b1VXGw
zn)LAfS;9B&{t7qI*qu&ZY(hg3y-K}vp~zYl`|k>r(!Et7MX(y%IX?}XA{m>r_RTPy
zF^0{9PNI!B-6?~vC_AM}0*giQ(J|d>i-<7}!d{cl7`wLll$eey)xFEP1Qt_cb-rUK
zV_mrG(b(lm+6AJFi=mt5$~fee;@)OOFq_UPTsLSBEH8=~|K>mEnhn(_>0M}_)Hz4+
zxFsn?MdVEqb`&#ck+K;)PQTd9u7CHGm^DL^JQh+qpUhj$lmg_vXnj%CbG8yu9Q*<}
zI_7jo_Sp1i92^IE#RIB`?!cDUO03TGOG|L~5q1HE7I5ykTO_^3E<>zB%_AMjPURyQ
z6&yBUG7cJ<1U>5ny#pTaa@I$NVQkM05Qot&y>xFxlq){24zu}T54(B6DbPJf4r{X*
z4nBR)wVwS2%L|dE(VJWN`hh#4))l=AK_0b8<T!T_8L!gs@T8&OU~eiBqUTQ3<{ZL6
z-fYm3iHpSmId96`&8#@?MDgyIC&yhTKNg6JfPuJ~{E63eSpdBe{l`nae(1>vGd)Gi
zc(}=YOTzn=2j(EYauB00SGvo?lxFGHX*0Kd_MJp|U;ZJ&TOC?a#bb~nL;a}RA4kw!
zu)XfHJ*r7NCaPkd<13PeOCjp`EaoblyS(lv>48_GAH_#?cQ<42{!3-o4Iw3;7VMZ(
zDw*XwviR(m$+9xDf(y98{TVuhQY+Lu?xE(l6UmqJj!3eV7Jr+6Z6Ha0`C^4P$|*NM
zd;MH_t}pmf*o(^AUB2*V^W-qYa?66=w+o#gdKK5lDnG|SE4X)x+%ozaU<XZI5fDAf
zrtHQw_PGS^d~Leec+#EG@A)wyoaCDMGOYxwZ*zpe=S8EBPq!8W3WpIfSnSXB#WMvV
zkEmotZQ#;vF@2RGQr#uR&j1S8nJ+e#m{LWpFkk-|$FQtTYP73Vv%smLBCKoeAR5cj
z276N+zt8#p*lzEqNS4&w?KdYY0nC;Py$iA^%u6+9M?!-bYn)WX&x9Y*SF63Z#cF4!
z&z-uRA4<mx*Ky#YzQ+bUAqQn|^W)G;z^SBR<&V8eM|Zql-(rvuUW>YOXt*%(WmUz4
z!+a94kl<Bb|Hx9#Meux(R^8{a5ZFb^PFCNH$VI)GLm-*WR$E9ZTLh=w>XhuyBIBaf
zZ>iN!p9@c?Iqyn%ytY}J!o_gZ>3N|W`q$BETz54Vjm?WX-o3E(H>>wbRaF`aQTMpZ
zloHYT<OKJ)d%q>&{sZ#b@hnz`P1cUe@txeQLVQv+(%c<4E^97FSzLxB(?|s_8F;i(
z{erm4>(|9FwkgPlh$$q|=YVNnCWR=yJ5zzEJ<3*g6AnDof_Z<t%F_0p%EI8%-2CMd
z`&7=mEL;pzCU~c(b@-!|WR@hY)6+e47OiR$mq=r~N~J&-{oDI!dKKgHAJ5|`X0#-|
zVtO9eV;(L2cyhMiQyU1WEg~3RPzz8G(N}cRId5j$m*I{3MbJx^y6<XRg4NZprtPlM
zW|lO6J=M`|xI$jX(pI)L(3w*bT|T9LspiO3guYuVQ$3=7X?{R$%vd%a>FxvwUDH^D
z4y@8J#v>1pOsddi^J&|5R`#jFAJ9$4q-g{?X0z`rK6m=urBxvlLK3b{>f(0)eyz_=
zmh=nd!qrDKNznrc-<xw4xp&n@tjWD}YXoSL)T7eBD(Ag-d=*<E%^U4Iw1KD&3h1m=
ztzB#JRlv61xn5zKFBs%;s4|#NXj1Sb4tTMy?Zx#b)g-O&`<w4`Azm#~N)|Wx+|Hk1
z@P6oFep+g?-boY_`UqhEsic|&T2-=GF@}Z3l%ud;hhN*L9=e%GowA*`&>6|Kz#dF_
ztD{zMe}oUFo6IdmaC+RRLF`AZeuGz2vY!ryTCS+|XB!dC<8^0xx0<kEA|HOGE*g3k
zcC~-Gj-U6ZwIJuU=`Jz%^MxX!H=@@{H=2WzJiom5mm^HXboRIn81JnnY4J+I@(4Kd
zuKhw=Ud0_*!x;0`0vE-U4kGoCh1;?#lG$F_8Xh}VtT-}1Bht0)mGYnUd!6?6H$S`_
z0Pxn&=4@UzCN})urmQ7I3Zas~WfnaXKVA4<_Zt54aKnSboeo@q2OjMhQQ?f8GQ5#q
zPYwI&ud@+0q{i2BYh?%jh;<{(crP-ZT7777R(?-Qy(U)Bbyn)gSTeO@`%aXrXPINv
z33tgsh|O{NHselzsAcU$2zqBNM^VcA8<$pTt*WG>zfsg_jCdxFBAtTf0+Ghp_Fdof
zDS@s*8Yd1dphK8%4rD~iztL=-S_}lpY7q3f6;<3ZLhLrgaQ7v`I?SMSMR5uoWtz2p
z*f>tT!tG|6xj9-NZB-hgJmXJCrsOlUW(kis){cz3xCh<XbORh^Tm8@+-GVgZjjMU$
ztKV=Fuv3W7ROBZ=w651~<k-nOJ(#S`;Ev2&;V4MhPDN*5cX_-Uj>5pxdGBM#Jsx3%
z&H1}8mZfKMp>)@CwU?uk+P~eao!c4ZwRUtkE~jgS9`C@lmioy0Q;KKe5hZ5{5$gb!
zX~(W-JfT%$WXY%K)3eziQ{Z^vlT=EpJo$bwCc=8E)`xoGDJf^w8vAgMnj6%wqE!3<
zjuN$QKH0)rebW$e04`_G<H@LXO?ON^n;UXu6MY3t<ezPJE)Pq>YYoP>X?>V!_1(Ht
zmm;cmXlT1EeXvv47mCqWX4&WX+RPYC3IY65w~=jC%rc<3_*XiyOtdZq>MnVH&$E4i
zaW^8oSiwE6HkftuM(O&a_6Hx8-A|(;VxuBAvanz3RJ6;EEHS_031@JTM??RlFPL7+
zzI(hQA+zdoxS{uG`tyO8lTIn^q??HsldR-Ou@iL(3!p0lvB>X~J7}F1eL(y2Gke=&
z-izFNN-4@!zlI_`G!`UEw5@vHm$*oFej4JA;BV{lhoOkfw<la^VwJ=w*yv7FIZ53%
zE4j5)s8_y1qd)2g%?(|HdsF8%XZz0G(=;`StY3?aj8w~qSVk?<wkm12DoftZJx^$V
z>DBajtf!|3-)$EyTVIGg&UjMrPI-<nd2GBrhv;WFm!0>&XgNDOaFtdj!x2YRI{~m$
zCZ-iO1{3McO-a*^cBg+bmM~7VDt@>QYRsWOyKc=pCv2%tx@et5pIU4~M7Qrg4`&Gf
zjQp(YYuI~Kn$3wIv?bN42TKzF$_r0ZfYb!`#@Jz2TDIN({jhAd4N`Uk6;zq_$|%5h
zh^Adus8xI&Ez>|oQCg(oBX)VZmA)Ycn*lgyW-CM=w5G62|7N5>^4KsN<Z{uOR<J5x
zoVx&~T@v>OeU#(3+#?C|A}dYG9eHjB<U|{6@Q^^Z$<E4}(cUe*>dKMeg}*ofym`M4
zmMGuZep*~pQ={e0FrkNkBPJe7yrm0qvzBT>ZA)MdT2haEl=4rq+U8yzY(p}nIa&E%
z`g6rJ!InS@*~4bM(DbkW`b$7fOM=v|&7daB{r|{=cg}r^{#y%x<(db!vsA)rDfusn
z{od{W&PV>g$Nc{|r~fO2hIdn1z9#s{U#E@q(|End)9Papg~TUcpl%TYyg>tp>pyNI
zs$KAX&YUn7o|5?0)+TYfba^Qul^BMN_RGyC+FGyL8@1eFLE-=o<Fahr&BCULuS`Wj
zXF--m)WJnD&^GST&FywoZqMbnZW;Y}SZo^c`glw3DCMPcK0g(c+K+-q<qt9@t7_Ya
zO-5H0L_%H0s^01Heb08Qy(Jt<)wrn1(p>endV6p#O>S{gUakQUt;r@SOMR6p12`WL
ze8KwVH^X|v4UbMvj9&&RvtOUDzbtR-5V?IqR2|sU{3n%4K1N0Q^~xL%Z?yd=SL^n%
zrz+w)_gaL4^iRm++9a+YF(7u6f2NdevFdQWCwt@7=YkK!TQ`C8u{Adz=In$rF)=Ok
zxC?vp9$~HyX1p<D772Y3usa6=nWL>JcDps_yTXawC}A;7c?W6edl*iMV(Vti(+z3a
zN5@;Qkh#w0VB)%t^3BW}r=2sky+%K+9KCiQ5wjDTDtB)&$kHlo=w^1*yc!o4+-_@R
z+_j5f8R<-<>|L$eo_h#VDfb@7UpP)W>(&qkd2e#@<iy(Ta@YK<p~lsB#Q>Lt=}QmY
zk5`a+k3D?U!LldxS#QNQcIN$$o`+QW<^sgN5bx!wG`@zNpxK@1wKsrn2r<Y@T>V;F
z0!h@^_sJ>n|F_)|=3pwEU1J}~darao9%r?g^qqA30EVwvVyyHd{%7Ysc)Wo`&O{mN
zcM$|E+97E6%L{Z|$a%`)>D+c9U7@ag`Oocbne)A|v7YZ$k5h1O@g3kJ-hK}Kfwp&H
zw3zbp%NadHsJd`en?Wb$yGh($N8|C80}Ou%@kr1quBH+?z~`5eZ=F9xNOqb18rXfe
zG&E>!(aARuW5VHFWM_`$8U#F_OA)7~Ii~L`P7Onl&AL(iGPT*Ic8T>Ax<l;RnNBn~
zpucWj2*P_d7d_0LCja`|AYfE5xes?)79i=O7?fy?&C2}LUMo<N6()okI$c@)aJ5|w
zWc#wxF_8@EP`tAkBA;2S>V|vO5L%DAw_NVw;mR%<w+U0=p?PqahOOFKP~d#S?htdb
z$7x_&gpf3Asr!Jun#=m4UkjhZw)=#1S-|t^@r*=Bl!$JC<p_}~KBU%aRI${Qu4<@|
zs#~w>wHl>M1_INpTTlJe=VJ67GIJK57m||BuCms7fwkC^!Y$&Vi?17|ptwLey7{be
z4Yi+SJF2Jj`ILgx_K3)m;`s^#O*L_<h99@}<TmZWnJe+LRgKoy3kKqBKU4Mhlk#7?
zjMcoWL-8qI&aX{qT3kOJDoLH_r~M*xPVvTdfluJp!wIyjOD2`UFyA4&B&k}2s=4kJ
z_QT&*65Q33l#9Dn9lKdwnAoJLFh6v%-rG1ox;#QfD-lscFuMUMr|+7XQ<wZ9^5^N~
zY*;xUG;9F8FuJTT8O%N{*z7)B!*)W)>8&kDh^pN@E|nwnq$9rcxOj&(LC8zj#O+i3
z)t#Spvqx4FQQ?_f(N2R};PmUVMq5MNCOW-i=GSp2vJfQEV^K!VE&uzc!x@J78^tE<
zCHbGMlwDSo?1V02ur};bMH@CiqI4+4H+=dtqc*Pyvf8Ef@`99_MK>tcQcWDuJMr@y
z55~Cj<$Gj$qTU_k$UkzqJOZD)6|$zLaT<Pp>9Cl5f9g!8T50B+^6Kq<z>#M>m`ZtU
zG~Y$^UPwf^jOn4bqEwaQIa6`ogkpF+CgIUIS8htn_3(|5!imat$}iSU!Eb^UME!WC
z9v?-S^rq_LqM;pS(y*Jv*!+e%ignx21O8GAy6BBxAp)U_ONnnE^WXVZ8(*_B<iDhr
zs4*ZQ$(sc2CNvaviQzh&8;0(DI1IRnI(TZcBQeo`&oAZCuLrjPRsT46#DY8#F8d}1
zGXEmKc~0~P^^w&%9?9pW_nTi5IV?BfoiHA{cQ?7L#@OT8vs78kUKzx&-Mi<{mLx<b
z=8G?xbz5+T_SePCgO(@Gq6ri;Aicz;mndziaI*-;cE?S3UMJAYyc!}LN03>uJ`Upe
zh4sE#!=}@u)bVCL%5bSfRs7{K@%7SU0i!__caTQ2To{0!9Ew~9UKriXFf`aaP9Jjb
z53uR=5Bo63!_1Jmajq@8iTFY=#^u#lFk!hK^hK^c`6J(nUSro=knPLZM&tc`w>Y2|
z+S?Q#;8hHYx+{gmebZHwbhl_Xv|kAsZP>lvlie+JQqF^=Neb%q`otiVXf_j7A?~&~
z>Iyjv?Zs&;78AJAD1(VsIuy-%m73w$!4u_Fo_C*NFUORW+s?@apB6gaxuf@;wzAqr
z)B7Sl{+rc0&(Vt&&BpX;Hr<-QH>FbKiQJYD&5DdpB1h}Uyk#$`uEvj$^&uiR541gx
z!;QJKwCpl64_qEwx);N<`>ur(je&>9*4LMnsWW4b1vbV)f{L-FGSU~cMGbh<^2dU>
zpM9)W7NRqFX!fSUw?BDJ^WKD9)3|NbYmLPG2H$0?BfaUnNC)_fE<D=7QX8Vz!BE{i
z>_v`Ce)xo*uN8L{fpy_M>Z<=}HLhhht2)%wCJ1^!WGGsiL{~aL19j5DZQvy7%yemG
zeKuCPpq()Ub<dXUnJcasvSSs72nBV??7MNX!h6U^h{9ryr_9DssOQ_LUhW3?dsgHE
z7K$<^)@Co8d!1_~CQoQ#q{0vIh>47pBDt3awy<kOuDK=qFyR?m-P&z<aHZUEBD_<;
zS7j-`KJ6zj2^@RA)6c+1;5o;1A}jdXo3pLX%_I9!S0YktMH;trxW!RQAgpG0_mE&~
z>U6QN0TX*<cIk1enrN)dDw8~id%m3V2%EJ*@#@A*5O1~=K(ci!(2_(Q6FYVC)S<C8
zqhTyI1JvXBUE6{e=cTKrR;@AnZu*a~mId?q&%>ASn8E=HcCV1og+cMrj15ac!fAFF
zot6dRJCZoYmLRN%`_gpfuJ{^%9xLrQkc_0{tI|ER$)g0vQ5#9<$s1v7rOSZ(fnQ^6
z+e%g*5<~?$isxF8<X||Gt1@^f8h(Q43A2-e*x7x=V|H3T5OSdHOqq@#7?f-R$;KYN
z3K>1)buE#`d4aXcL+Xa<x52nc<pX0cKrVxxb}Sy#F}I=r^!fR4wPuxlusZ{ytWqS*
z=PKu8_ZgPEq`uDi6;Pns?I!O(!r>@aMQgBbq?U&Y?3A1QOUu!XX};efB7L0r=5Tdz
z(3q7xP8Vvnuwyc`JjbC`@g`j0xbU;*=2_$QYPlM}jSd)_!V#c{tTUo++G+Gqew8$i
zW!Ubg^XfXASsmZxF!#$5+IKa7EIRUDIN{N?G-~&)5NC37Vf(YYkAKD{-4=L-VY}}!
zHGjQg;s&Eua)G;<e*BLqu}k+Q1u~)Y0o9aU`NT?XA@oc4W(l;Vj_0ngZ7fAc4j527
zy|%k9RQ+Z~UEkO}yxL|P9LBn&US6M#zV-vdQFe)JE_v*<g{LPQ-_cFKYnmO6G#v#6
zl&xYA7bs0%oD-99Ekx<|cjZ~6Nkxcoq4gJ^2d|&19j=>jVv=Jde{&|-9&*71^p#uw
z2&Kgqg^-{aDU@6>*Zx?_Zg!CGebRm$1%dZ!0oy9Q_>Vp!SC?g$eXX()%3*aL>iO#8
zBz)?iSE3?L2Y0nylHe74wWB$C+8P+vQR#C##@dLPSlgC-L>UpsYV|hHLRx$R&M(@@
zb(6u#*XZIZA+w19PK_OW&JgYz23!;|(?@C!yNTqU8s*8;wD?-N5<#(8rKJ=f-uGG2
zR4OU;vZD+>W!hfvy2qr?xvB2!zQ2NKR>abq6Zhx*J$b*R&#1(Y0X#~*Q;a{i0#MWt
z>67`*<EXG0i5kKBHwtAT?6v1&xRDluFSFdXf8FkN(_e5KYB7MYN&2cHi-Q>%!y%CJ
z@bInDnYHRw55(BL=gx+uG04i#Ezb43;)x}lDlXQM){az15czYD>mcK=OiCma*i}Dk
zey$d4t3zKNzC4`w1*@JT?o!8_ezF*(JI?Sb4se~<uNPOvmYR<ruG0I-6=%h?_UhY`
zVH8Fs$<P>-I1AWJ%Byy^!2~#ZC%;9R;VN(&OYCb=+bT+O$dtOR2-ga(>I^;8k4L=J
z#vJwZdbRTOGev|Refh&)%Ol-=A^SM>?%VI01Y}laD!JcPp6Pb*njk2gQY7e)KWMMf
zv%|fDV+S*iI7Y#}<HK>cO5IFL!G>F+=ClH9<%%*;tmA{mbm#7r>a}=9IAo+bWT!<|
z4;3b!oU$gZ4|b~+e%?b{>pytpl=5?|>FUzp8!IQpN@WMZdksM>9QFiqtojpFfWR+5
z1~4i=U|)Q{X-@cZ(ihwl9BuKQX1W;9qS$7AB~i1<M;@~!zsT3u=rqQ6-{Rwu9mpMb
zL6%P(?=Eutn;r+c{Tgze$89&^u#4Y5A>3M^EZIqfb9Yo1d$p<T9Z}*R+)-HJ<ZQo~
z4}@7U`@FO{Iz2Tp<<t$j(uQC6XqJZKOKoU<5aBG=O8SkF4iS6d{lX%j(s0MhIC`=L
z^T%8>fGuf~Z2g3>&4^0Y4JT%hW4EdBiPT!%TxwrRvR}Mbcr9jFkje<byfu*TlRWaK
z+%<Wk@#Dj+_@OzeF!2^r!kxX3L`alp3GHW=?oP7z>-LQ7VWg*<Dt4V?%XKqP2pc6N
zwvz(Uz~r|fO;fC8*}6Gsh3BGqcYnz0cuRkErqdkhmd&YTTTB<}4w^!ZUzNw@b2N;j
zJIP|t94+>z9Vj{a_kPAP+x=?N0xXn@Q)Z-1%}UpRhC?A&9$|QWr}M91u>~xXq4bIg
zli2|O$|B-U8OX?WY4#{B>inRygeaTVV{yNsSiFuHmMko~sW|-_8-a~`lTV7DhP`l2
zDhbr*9bQax5_Z;mR&Iu6tOr{f%AvH!Cf;&}gkdjF{klMZg1yi<7bfWMK$@RXpILOM
zG2%sn8z*8dLAp;2*V}9-6DUBvAI%;NU>)4>LebH3>ISmXa`S7R9`7wsi~7_dv6Nag
zr`7|zZD~-_XDgGUUQZae71;^_D~DA*z3XT7DANG-;$oC0NkGu;q>7?fsihZq7%P@|
z^2`dtuumm8VISDO);vlWt}5zowSr)keMbS&<b*qm0Q%mY=~s}+&uTRuqx8vmtAp31
zKz1{EyK@NF&!0u_JMM_>gldY$$h+mBXR<A12)9)}D?IHgb?l*R!RJnip46YN&JWy<
z@o2~>^lP_{9};EM)0*hM12Lfn+D<cOT<MGMaIPgS0qR5UQ~xxc^1{47%o4%YGK`HB
zHv~c9+e3;Y1*&a_2KM^S$1`05Od=Kd*Uv6A_BQ*W0FuJCd7{#oJx3#-Y=Qh*Znr~v
z-VecvKlLdGH^N8S<})kys@!}omVn(*&v)W*B>brXt_k5B_-wZBDQLbte!zIQRO~k`
zWv2rP!(HZJ><X6&9twC2*jrW9A?ruig1_#GwG##m)gIpBeXTI;?N>jkliUf9`&naA
z_~p(5R6?>aM!{Y)e&jd=!)APoh4(iXMw0>!@)mjPokBW25EsFy`+z#|nby1XBzLrD
zt7;nYkr3^*Qu5cf8Vb!vT1i?7BL$6i!Ao;bR8CBbq;(UojE-?y2}(uTo6fEsge*W@
zp{=A4UK#w_DuJ^yN}+IZi00K9iW)Aq0IC*(HO8xhgx1XJMh@$jKFv4G&dj3s!R6Jm
zcab<Y1rI|#@b{F4&!z`QgO|CcQQzPGNd>}1T~=-o1I}Z2g_IEBcL!bSzj*?x&(y+-
zfSWAf&$sT8x!Kb?AnEv|!^NO(X_?)eN1`Xfit*~wgu|q;K;p{)1U|{&IvJgFrvc?i
zj~kza8u*i1`lldv6|$39^~=#UW^D`&M{Z{HrQ#2>-ujYwt!>b0OODG@k;ut1P$%I~
zIrk}*n__|TA2l$79QVtN+~MpiUboD+6enrS0a30vMl;1kb@a)oLeNE1Dqr^C>56&R
zcW2?yUkCiC{P;pwE84O^ni;b69DL(Bq6x9;V9s%Tbdq74*m-Run|q_~ecxobD+BrM
z6Z!fAg!Nz)jA?hcZi<}}89motI75+y%<uvc;FhF5?N*&EKVP^~s=7J!yq3v2frx>j
zx=ez6L9>kfBm?d<j~Oify@Ypk)HsTMIe1AG#A#27m6LdcJcNf=k;diudSg3?m|6Nt
z#muj{>WYGU+a+r?P(%yEq4iMW=~Jw)teU~O0zS%9(y?OJ6{WUl)YQ~X%?uem&)S!Z
zb`mt)*sE>2FcsPcKDDL<|Iv$f=*~Ueio$pWd;RF$wXfDwM)f;g;hSk1xIV;27)4Tc
zdABTHS?>1pQ***ScKVtsu0j@3cPJHU??86JkJ_jYI0n?zz{RI&pDpG}0Gyp6UU~ds
zSWxGn)8r2^p2%o->5WCWIibW;)|FngsC6~lEI|YR?*J$ls!!f)-8^Ie=hU!xHDHj*
za>NAXl?5!zf(9c{fZhLxEw0iA?`Fp*MI%Y->3x6;^nI?~_`@~BsR4;~UW6_yekG}H
z5H-2f@Es!MR!3uZcJ!*LsR^!IZl9s}Mf|IAp8jrg2cyqC?lZBqH>q;ge#}au?9f1%
z_pzYz?A$b^IvG>Fn_O_~CVVZeToVtgOqGV8pMsQ4S9E=>nBeBk#r{IOCE|iuz|K_P
zE5D7@oVXjju3I$D>tg|1TXs&q5pgTCbOv&9ME<Xz9}<kxKIDv_<|pnAf!*v?>HdtC
z*W8fq%0$B@f?8pt{RaID%o{B^Ioa-H&s;)60xt2E@Pvs>R@|2b^p%-G!m^Q!dRrF2
z|LW-FHR;y@o30&PC`@ek87yGVs@!A*3{`d~dLY?9B*;V~en0}kMgriOegYofCYw0c
zfdL4}#ey{^8gTf$;B}a#vN~fhkfIja{3`L*T`<IMVa1+xAGLz?5qI`e&|0Jb4#3B{
zfv{X1NWS=DLI8b22E0np4mK%P125E>*scN@2$%&=bb)>nc|y7IVFBqd6~Lb)baW-~
zA-^EgJAsQlQduOHV`DS~OX{Y54J9u4Gs%)Xcw=Kwn}r!1>?`n^Y9uvB@g|t!`?a<q
zrs*eONPqCBzN*^Q8r8uTdXq%O#i5{w@5ZfL@595x<wzbdjRJz{q(OuRo$ei?XOO#$
z5Pk119Nlq31}!pbnuyq+xE5t#cK=pp>}=sJB)5zH!pJD%<Ht8pbS#<&4|)UWX7ZQ}
zMSvVmwd`g8>ZGw!-HqOXF+5cFo6+U`zZDZKTQ8rr(f&$}QP8VE)7~UDWrEso^3@Lo
zcoU4@{tX4)krs5l#eX#vyE2M8bg&gR(IvLOg^&d-NLl<YjH%KB^Tg{UIsM7PEX#rC
znBg|eUKzkYBK`AJ#_$;Mps?RvCtIdq*|s)b!wJ6)9L@>;1`$v2<12HJ2io9rpU35&
z1WMAsTKY6^0!@iP-eGkC(AF?&DLFZuj%el}ERy@iS@$&Zk6)O45ZMX*3U`uF{#sHW
z%?KxjG!pG~E=3ShV@?$`NMG+Z(iZf_)MM{zwpXl6gZB0HmD$ZeEk^P}3JM;?#smnY
z!>U2Q<2qgYI&vp%{&n?!nBOc%*!+Wnva#R)kRxS3e$nxnh8U+Ztc~hz+$h?}7nluN
z5iHSu@tz&)P2px)$d_aCUZ(+ZmTTyHv^Pjd0tV6*Xz$*I(mnmqt5B~=h-EAWg9u@}
z>H^h(i>f0v=YHG1&@&*=0qqa_Utn`!0k_RMSu5qikki2gC-Ni#drt$0Wg^9R4QyW>
z$V?}#?6UyUnp7;m<p3pVMH-uKRH3*Qv7u<sT^fya7%6f{$;d-Vjl8!y<C~TmmZGQ+
zIMx$OqL{W~@FH)23Bfe|SEsXc!BEN5SA0l~zymFfE-3eTR))c#&7>!bAo^!u4I8R9
z8$V>gko@aK8>FIW0%C0Gl^8??EQyK$tmE$_E=a~KefWhIP@pitjU1NIn*<ax0=WM~
zG*2GF4&FXYZaGNK;y|wPdlvZW1sWkC;SEyKpy_Ete}8}DESOBYFt5Wf=>>#|QEAWn
z@1e57=j88(e@07!)1)4uf4rH$BwbA$y8s9qUB(-fl)+6+62y#(D89bFY%JGZo<UB5
z9L9Shb&~&N0BM4O2TdGy2C~oiF?6NiqWNG??$+cnfA5_D@VEZY<s6PHLm~r*;`ZQ5
zyTKfAO0d~UE_CeQE}RW1DhR)e%#{X%eV8C+L4IYI(_LU5#KcLCz|}m8qGsT4e#;v-
ze+L_XMfy7sWfb)-LO#gzw;AzZgX42p`@ZqV!V|zY)^fx&Z(Uie9q_UhD&^&zf9>}O
z?Vo<!x<aI;;2_^&Bj)u!<9+esg<_sApK6BU9TmEA_byWAl(PsPF7#a5NRDFMJPAnk
zHL^swfpFM|Bi^Gx63M4_R0Ffy0?Y2^Y%zXNHj+1?p)CQ}q>7c!*8Twj<wYzLe1IGS
z{+e9|P+vfi>Cy101mEaYI^Pi#)YUUIHmz`NCskQQ`(pUhOX5KK#}E24Jm6vA=zAYH
z8GJ3jZ87O%UF!nXcJVh~h%v&ceKRr`KMQ+N5D*YBDkb}SW3@yo(qIEwY1PdE)j&Zn
zL}q!_#PT-H!C>NWat=QFEe4>EygD|cPGBhS`%)tEBtcg)CInp$*p-_E*nSB}e2w34
zUijxaNJBvvL|WF|rw=3@$>5^82?5TsN*W9WQ;Y{^b-*^Bgmh(rd+Y=#?TA5qfkYnC
zC*_J9%0vFqX=^Z4s`joE>EY~P;2=;H%VkP|e{xw{G@4<WRs|Z6Ztq|wk2C^8E*GS8
ztBXV)3O3e9WG?Ov*pk(sNy?+Z1Oml3JHbunio7UWA2Bh1cml!LV}aT`VYx<tA<{=B
zq-xQS$((Pth^G?Wx5RGZ`xY&_^*wAvF4dV>T3T9ESNGOyKR*IO!pKZk2D;PfA6f=S
zn5l4yzSFod42@)13fjhGWmxKbWg8W7xCQB3m&oa*SX}SXLh?m($ZZ!w-}!q-Xy|vE
zd#olCX>Y6P+Qfp*(fa=Usd<l{K3UWkdJPB1F(vYQ25|w~bOb~QG^@RjFNP9kEP*=8
zjPtOEt-T)eLG%=BH$z2dULV_mzOVNQ<xphHYT^W6U$%<MV3xK%Ydi>yb^oZe?TsTF
zTO9fy`fou|yJXsqiE6Yh1N}^M4C=Eik;2V*WZ%}R88gpD-Hy}vD~1)eEH!uH#?zuF
zhA@#0L}hwO@&62Z)eLftt}UB?0Uo9CPAfkWi)=kbqGX?Q2RzgkHqmy_O$!k5F8u)q
z*Dsz*AF(~(MlwWY$c)*=h4<=Z&@-*qua6)1<$687CFG2C-pW!p7SI^M;F8C+<hG#H
z_*4bV2#0B0|Djt&g^73J3%Xn0lmCQ()-1LCS~8E%7K!<PYXNYr-Q^l$6j~_-3r#vP
z3=F8x+;&(%+YDu=spMVTM8o&LW7EJ;)-|I^jlOV)ogxA~8il^`{lxd*PzB)1$A!QF
zIT?Q|yb_fm8wgg<L8_1!gt=Ku{g1}wpeM*aw9Np`y^X=aCb5nAKcOm(4+2w9&I2#r
z581}Ke_uf}K_gomol5DC%)>~JY&OMIaeZ)*0YK$#3UcVlcI~BFy+`xsJDE5!x#t+)
zH1bnG;nGmm#}tMKP6&$PoIO*^GkYh{%iL#IQ{z4Nm0x7{dPOU>d=AZ(UrR;C(t6oz
z_?VZ{B)z2h6R*!PFYD$^sKtFf@3@xi7%`qJ5=dv*PP{@pu=xl9xWD-Nin~UMF@fNv
z)syU95Ld@zlKys)b(U!S`$?vRy*~VV43agM*Y)$xx$GCNi3hYk{qnTgzGLqEI4MQC
z-qSV3{PzClZfssMs8ae_`ilNjEhk3GX=x+T0+sr1$;J@@vT-#ku%-*P@kok@8JHLF
z2r+0?_u-c(<JZql#$`Y>;Gg6z6bL#-TN;?WhUy;aVRSVes72vQI>gwhZnn2idRgxB
z*Y?4rNd8&f57ZMtw0mxM9k#{tpw*tcVeKTkgHFmUs+WylS(*nFgkIC{XE7&%JbuQM
z>9ra?Arf2xoYrI}FU3NM=Z|cAd})u2*LgI#(%4ZeMs7XJVZ55!)k~BIGyJf7H(*gR
zXk=|>$a(X5&k6ZzHD6py&j}7bDf=4?r!4J+k*_Pkf?3GB7wkr#){v^Vi<zp}1Yg~D
z>@8^P7DMR7I#`Xr!CZRk)|~Ss_FGn|_IP5{gH$t1!Q&~SJyrocLyw}WC<O2(mZ9|O
zAFgm1S#KBG{CU~HrGA4)H9{1$Y{NsA*@B&J3lbqSvy-1@)R(GNvkAlQ|GsWuBiB+%
zd-2b`2|=SlOKfLLHvHl+4x0>wtveALTTzQS4WRqv2A?;9ehz^R*}9&-SS6tgYQ&_z
zndsrCl(!_n4}&e_<1gFKY`t8Wy>@LY$gy*c&l7L#k<az~8ch<F#a!Q71MWT7?CYtZ
zRM`K!)3F37HjG#lq*DLuJqRA)FYhdTfVw@t5h=g`N*jzoW7{>Xy;czK+8ubZ{0&%9
z3;qPG4In3&Nid;dJ>g59Qklt~=eR8JBzM<pp%oyc1;ZCYS74wn8KzO~$9BiFak%;#
zZhP8;yZUZ;w{?azP#fFRjQ|nFWNgRhCgo*!7?Yv1TTir8w76Y{1Q+2XzH_Z%wDj~C
z3mwsq1F+&wIhuwnYc$Ch!m7S7y}TXqA$rzM4=a_am(QNRq8b!4-QGhI%GHGQOGB7b
z{62CCA`S~b@=Cq~npd;HlQU0GL3C(K>Y=Hq2VnJmKYSpT(<!*A<8tAzQ%W=>YCk#7
zqr7C)$hZ;PLrW)&SMxj|g79-|k?t8Df0!zkB8sS^BNv-)8EmZBl9q+#Jw5T(H6-;j
ztj)m0L|#f1H8|-li|qIXx$@ozpDMRG%)55qr1(aI+vzi#-PW|&ud!n6nVFeYC@tqe
zvuw>r4u)&!jT^LuQZG!o>zx?G%>ip`-q1CEtve1C{FK4hvF`IzCg~~%smU|$r1{Ls
z{4o5p^K+c<{{@jbC&hImpoz@?4w1Je-W8ha(@i^p_*pwP8Yqb)Ac*r$>0dw+K>V>`
zY_|~#wL!jW329B%0T<{}{hxpUL@P~8c}`5=<DI6@zc9jO!5yr}jg1}s&rJ&KMT-D{
z<9Ade`x4+?1uM1Tx6n{9<no@mbX6M+jhr}a0v8{PgkHyxy47@WfiH>u9~YfozkdBq
zaPVV0y9eUp;v4}VKK%0IH1nFF4u4JfWYf%oQui&?ouC)Wzi|)RF0*wVOZ=Sy%zWED
zbWrw{79M~i0nl&XzDdi;N%%t1Z;+8KH$>Wv5Emo?pmEMpPWRiy0p?;f1lOr`5JV`r
zL-%*Qwg+HX-bl>z^1uEE0Q!x9kna1xuSO8tBw(Wc_LMMa1(58>GnE3GD|?-W0{N1{
z9b*4jFV?@}IT#dyJkze+o{Fn^z_HVV=X}R_{_5&Eb(sL+$uBH+|9L90G+<m3S}}k7
z&F^5@Jf*^G^1tzSZyfl$o5H{O%aw~s2XIRuI$q$P@;`-t{pJe`IOg3ppzMs(;{R^B
zBphmy*(09cI6A1^9;@{7-(^x6pMFZ*8w{JdezLGD{Wi?cR`Yo(+R{T-s#?Fi*u@uo
zjMK`#%gZly%j_V<mLrmynjd958xmCN^ugK>BIlivYd033p04-%S!_w#(AYSH9+<YW
zS$oNSR%YBDnR&DP-yyREQyCVBA^#gPC(*x<0WrwGLuQFAd-Ej)*(Lc&)z^TfiGwww
zp5&beg?EjRsxb>;_|#fip<JEo1EfaJi2knvHal0*OV3?4IRoN=TcdtAwx8k-dR&-)
z@XmN6K#q(tQP6LJW7&A1ZS(UP0N9g7Gn!UdW#HF!CefEC(i+ikQT3)_Xyi2;wbmGX
zb364$LLOw*N64yBFlm;C#D*^4Eq$4dD-fj*CfFb2V2RBJ5sw~S&$)8lIe0ah-=#Ub
zhyIzcHzatwXEYdDA%ZNn2DLhaMEhU&E#VDDY5nZ1u~&!o9Eg@`7@i{^gN&o)+He&C
zcp&So{+d~xl`N0Dyv<`XGcHtXxO+g2Y&YGW;8bWB8Nmd3;TB7$?B~6etP!m7`5KQ<
zm}LR<E(LnyQYvDZ1jwNTDfB<pskzF!-3e%5YnzD5pd|d;tudQ;hTumh2q+@4y9{yz
zG*$)u#L>m0hX96pQhMk6gX8K_rE*U#mmMQy=!<0xcqJiOPu)>eutuQ*B`rEEUSzZ(
zRxptxixk<CPX&4Buy;yg43Bp?f0nvCvG=HQ*dnqkRy9*op0)8R-%$z(HV;|Z?}>jS
z8lX-f%0x|MoZf<6?0ykZ$}kirX8qXHw27!b4NqV5g<PyIA!%Op-JoPeO??GN(*&dn
zYW{ny@Zwe6UX5Skzc~<<Pf#1y|Ky|i_iwEXI~mV_J)K@_YH7Lq@F4*@7O|+Ip<!0Y
zj==K4t3~zC%wRD!pd0<@^auKT@?J!$DaR*|pc<8J&}H97W$Zlblm|IPayz0U&=qt?
zJ<`O06j2Ox9EecvK3OkkMn^N#bOJb3#OTI`zGV>Rp3*?f7vs0$FQ;Zrh?w18>vJZv
z3&#r(<C8KkZ-lalARRyu2#W>Y$50M`0q2KD8So?-77uK`(!9~l?b~gtUNfI6$Xl6q
zWFPzWV$T=rNL)1#`LHc%#J*IX-Ih{42?^?vR^dDS4EgNAu^eXfTg*VyH4SVgA42i)
zm%$Gwh|M28`N*3}WTEONj(dY)#&hg-@zGrVfM%tgzmY{tE;8D%dkm3%{uwffOBwm3
zy9flAUsuZRqE~6xDU$)%v--Hfnt5{z#>Nq>VN*Y=rbD6}`b(_LNkGZV#v1%`4H6@l
zHXGF=wdSV|P!UA(L}De_U8PR&w<_XZWQm>dm~UEJ1xy86LA=|rV0m<YWZuLuq54Jl
zVs6?_%U09|{Or~i3($-n3x#3&Ij|jao;eC)c-^HwCbL#ys;?h#f5v4gumNa=z-++}
zq)~rn1zWXxwV_}OKJIQw5EBI5xKqclywNAmVsZkNzrGCO0i#U?saHFL$EtLzC#H{u
zWfW=?%0QXgG#&Vket&PyS2d(M0=81r6j&f<N{iDTQZv0l2|(4TWug0$;8}a&R_CJS
z?2Tg^StJWDMHzFY*^TI-)FPoDKazc5*1#jVuUf$S!o%IKNw0b_I9lnVjq=m$VFDRE
zc1{;jh!hqX^xwVjBwelFMjQQf3DY$9|K#)Hd2F-nzXw;2b5UbkJgJ+3c&&W0<~?hb
zsS>He_Vo1pdE@F9GQ9U6AV;B>U4N(f1D_ZpnKN~S5aEg}r1WMl)V(7vs;*e`h?pJS
z|NkTGEugCEy7h56f`mvTA}!qr0@4SN5LCKTT2i{>fC!?r(s7U$q`Mm=4@kFkcjted
z_x;}c-S6Jt9se;H>Hs-=?X~Bc^O?_l=G-ghAj?-b1zG;1|K@_hED_4PHO6t(YK-@v
z3&X<`t08aC59dKUnP9c$GzHR7uR*lgYz)<uK&s4z#u7c#{e-}`Y8^L=w?>g3rV+Fj
zJuIC3-r&kH2mP1o{uZzhtAqcSyB51SZCv9ykTg925O)9lz696KIv7Evje))A%E*Ic
z(BXbF;CXCI|9apa4D2W98C{+@3L|hPLMZXI?`x1@0+~i07o_H^TX*ExcE!_ntrG(k
z4=}<t5YF{*e>QQi0UxHP@Uyu@5jqE+_k}H042L7~T*R|J976)%c?$6D*}?w&%x&<G
z3LX2GSZ=l8a+OplDO$eSs7GKZRIq>t#P|A;lag5rTa6_~@+wpZ2r51q1>SedVzZJ0
zPNC@@upzRDs$dOZ^qy*tr6ojNfHd)lIjbspgLIXPgtjye(@)4RuIU_1a3y-A%LhRz
z^0~tn)e)%xu1<`CUn3QI`0w@oW+TAN;2`-JE-wCE7qBMVF%vc}G+ku;;`Z1Xhap$L
z*$~p9F-O@{7O4jWjJSXuDTG{r(OrtH->F7>sequkufcAk3Cr;}(pZk@0DJFORrMSd
z9i5hrj!tDxTtV)nIli=E4_8+F)A!S3RgsuL2q`d^|G-`AE*@dMdoI;rg5JFQf5Tnx
zccul@nxp9DAG2JX&iQ6v#IG_JNvAM?Q^PSQ9ve9|C0!Q`M6JB2M_~SqVqqU%eiHC}
z#H?%Y3IO{B$ii2c79@lnB&0DuF}b0R^jK9eKN@(pn)*E_ev6k6BO(37MMNE>h~%kz
z(}~_l1JkgbhRs~!8qc=rGoa`-go}cDs8_wwlM>TIFw&ul7zq$;IlvaJJm^|52PD5p
z&xdJHHc$r6h&6GBmUSDP=$j$6IuXOiSU_e5Z@P1Kc9t{nWBYQQY4+VAi(uVNiA+n$
zU)4=+P)|r9<t<o5(#0vm3HkTYl8}s@u+`45V&wKKUx8sxniuDM<mZcaNwsdo#Hmc7
zlOrtfKk_5wIzp^E3U$$_KvS{T6XWr|Cb5#>r%$ZNPj`-74A*}%Ww{!vvK_H6<r+S3
ziJR?)+}#-AvpQr;4hQ+%Ld08oW<ytCQ?=0!0UQMH<)E%EGfRZFi)KIlP&cX<Js-Pp
z-u;c#0y<aYV(&@52uN5P{?U-&`~PIFL}%Rl3&>7E9m%Z%n6`Okq0rb4Qlm^RxLn_#
zquTH#jpcx4>VFYcjR|%(`r50;pr|D(r|E)K^#<%FAofeWktRKmRGZ&YSk7GMMfzEx
zgh!_jS^q{cCV`vmOUTwm=g)eNF_0?VB+fJ%P6l&C)Shg6D0Ga&C=_$Ow6GIPP?6DH
z=knuok1e8$a4o;``$0(Q`n&!fYk_J?q}Tboly$^JZQMZ)WC{xlgj{BZ=&BqB>cocw
z(dPR*hzU(T9s<x?WISAE>Xza3)X7>)tPiO`Aw?X!p%%R&LNtQ=y_pK8FMAC<>2<3D
z-&^<L%7^lV00QN$K9DR(ny_L0`9NN;!GVHNAyq@3L$9=t4t`2`PV=}?m%`an^;vJD
z&mu#Ot3`(BdBwY5kCq0=zONjU^P#ir=LK%4rGWCgdN~%~*Wk~<0*Q+Bzz(^va;yu(
zc26i@jO=wHMM`y3X^}K=PB8EeXn$lHT@&(J+Q>y<xrd{MfVh&iN-00I2aV;;emiw8
z%K#Sr%j{<LNC1uinR7<Q6|TVfj$T)j{h`>kgh(<aAS)!b=Ys-#GqBeB=12p8{Nm3i
z2i!bwTcF%_@wsKaaD4~sppY>H+CfV^q4*3kga1-tSnX@u1bs~=L-NTw)ISFrfX8x~
zs`cFdo4cAzy%h){>J8#g7*$e>KJ+Z@PcH^q*9_i)$#b_ML?tO==9>ELIrF?Biq)#(
zJty>5$HO;-0DxDjNB_ecX$n3(mvEe_Z;63;%%}$xS;$d+WmD4J{)a#UWD9)_cEYFa
zJO;Avr{?R@$2iHtHkEvZhhz`7^lfb;02J^-S^OOQADjenM7*ERaRR5h^f|4a1(vo9
zox3Qt)fzX-p&x*p%E&%p&<j$-676Dp>vF7$)bEYB1OBSgO;rC`=I-tP2*5viKmaZ=
z_7lqT%R~DzkoMndYRDM;{TrLskm3s;^qG;L?bqhgO{OS%_uhNv9#fAzcr!~<d1}H8
zuLLSjNso$-=vxu-xst^@6*9N}K&8L%F9K}u*n1CEkAsbpLm%(KhHANBA((Not!Hf;
z4(<<vvUvYv06MPGSb-js+7F`cSYIx#Oul!(DN5AtJN#$8PM^AqrADqv7~WI6?hsvX
z(-v^}r;)^H@OzKu2(8U%Cs}^K)s$mY`{6#|VjrN0Ecd>g*JkqyiG@fU^1y3<3|1Rp
zf#F?e<d6TZ>!o`fZ!jss!Qtqi3~B$Bgbb$To0!3k`}1KjeC=Qo6>|Pl-_)t`*yOvt
z)6*W^_1~I3%FuA~MhwnT44L}XRCOzk|0NoMwV)m5x`P&p7p3=jl`74uM%+0^ewds(
zI0k@lzIy)=`VJxEoB6It+J|>*=rI==?@oFFz$a#Qq%3*p*8n;muDn<L|2UT;ry{;|
zRDZkOg8%pN0FxuZnI0NvLq~IBG4#TjskEcwFlewQK-h@5Zj^bjJwyZ1$JQpI3gmxW
zYcnzLRlCZG5T>Z7pe@3%W6jp`!zqq;wxQ-OA0TxX<GAe}TwRx>b?Us{+q7dGhCbJc
zRd}&8PyVid@hP%t5GaEH03}{L@IW$k<^!GI!9XF+gXTNDT^oD3@=8Bm(9}&e;14Cl
zj-G3h(S<31#chDa9f8tPLE(wvyDd&+G~7*}sWF{g9va|Gd|T#psj=StkR8p5w8$1f
z@bQ{M#_5{moO|W#VBpH*M{h(acrF4IMm!u2p|kRZ|Dxvveq2p+fcRUD3`J8FDf*@T
zk-rDG`#v(viO-d7d$MUZkPaCa0BJNB<s{<oxU5L@m$uCedU})|e4Gv)1t6+)rL+7P
zxu60We}HH9j`xel%k^hN2mU6{?9YEEWOsU$yaEqN6q(&laImSZpn>UB;tR<_8}}-a
zq7!)#2ab9x?~Y~o@gp-jNmfr+3Xr7uei8VX?weBb@*K?AO4D6m;A8Iz1l}8~3fUo9
zTS4BljST{sIg=w5(R%~%uTvYLm}x!~;Mzbye|r#=r#8l4U3#N9$dz<)8nt1hzw+k^
zz;d-|8FtUSQFv3IQnZg(SiWaQhWH&2nRi5O+M@&wfTKNMnWp`>T%zv_(uz5Q!qYp8
z1V95cot&6z+a77@z^HG~uJ6r*($3?qSJl-x$k0*=_&3+7EaT>6Ffexq)nv#2-`w03
z9oQ(*#dM}FLoU!A!NaWRaG*DAJ4<E2K?WMcboq#5?jx&bd#J!eeV(ecTU_Z&1STDH
z!CC)n$b2*^%KQCFKIhe4WySAn*GGKK=R}Bdph>Z3#Fu^Jub_yhd<A*cGwqmBu-ViC
zJ}8WkOEjpXvy+yY8Iy*F2AsCa8Wli?bbgD_$pViLgjz75Rl?9O^<pVx|DP#ynmoRb
zLWnsc%C+Cy`~`mK4880l2aT*qNIRzH17)O%tLsw!N>3%Y;hWW%T6|c~RjGMa6%QR5
z2NH2?p4bY1Mnai`iIRr&6=b&whHuH2i@M2TLi3`N9t$F@`^cU*^J1*O_%lCGOp}!b
zi9DLSHJ;_X@`?7K#`h)B{sER-G}Np=4M1q84^_T;z#anEHj~BPZd?tFG`*iZ0^N4|
z`i1<8`R{#BHB25L?ZN~~%|k*`fqK-=3VD#ZHoWrYo?M$0^M6wpiC~$mkaznPbjFy(
z=VLynhg%6XdMyhA0Du?6i7Vl-YLbTxwtI50UR~V`v`jrWV2{gxmh8U)t49{=nG7{h
zJd-3R+Gh%0w}eh;H8G=HzftBk3QK7+xV`!;Bll1eOgY8#U3|DeXnSz&D8$di575z?
zK@%kjNsQ(BOoPk=F))1OnhKpQ?Vp-BR2y<49K@Y-sQEh`Sel1d3;T;Ooe!YrPjyb*
zZX$;d6kG1dgk1-S(yqSkIlu(tb+i7-6eU5Aqy1Z5=sl;PXC(X^Cg2z_1~|GBz2`yk
zOpyd3HGjW8qw4=3*!Q0z>X!_Sd6}UwbjV3v709rZmj7K>-eNTzaAdk$<306Jr!*XS
z9KZyzJ&;^k5}p3P@VG_l<a<_RhA<6nN+BTuxdvjM&FCXC4Q^`>WS)2a>yF``Gk+Ez
z61%uT4XW+YU3SI3xD$HDqaJZKOoYdO9$VZbCPb1^I)=p&$j{oM0hVYFQiLR7cLJ%;
z52egFi5Bco4LzOkr|XjMO5LB+kXiz)iv_c7EAjspgthF#Yrns%Qovf10P)J>XiZ7C
z*5$!+Ppk~66o2fAojA1p=-Nh%RI39Q0%~Xr+bF*ksSPMm(e5>McA^V7qW;y<_9jtd
z@Xv89kD@LIvV7Jgf|ftCFJ`}h8b&VQube5Di&VvE6rwGZJ}||@i}UBk<$;@bL~O3f
zK<}?Ho4<jK8U0bIBf_&hc?r^z>t{eKhQf%cDHy&pRf!wRX&`;PF`@*A!mwjMpop}Z
z=r&{il^F#6-<d&@q7?8mz{hR^AFDTaj}7Jt9BHkLQ1Mry_efw~+FmWDfeCVOgW43o
zEpe2Mb%DQ{$Y+uAm7nYb9_Febu}L=q($iTQ@GXk%UTx$>X=iUy=|evJ+2+?z<43x;
zLSlp@P<%d7JU_g4Dipz<`M$u1=5&}))KjZd#(w`LEFWChTq=CCCK_oZNttB!V5X_`
z(Ywtz97ur}6r~m<`cvPc$qo&Hsp8t|s)s_ffLLSzLaE`*irn}fje>N?C3K6CGqhpf
zn|DB`-Sg8=QxoZmPVf+(V4_rkyx#eT8A|qk8xx>y6Df7j4Tq#437d)Sp2=$kB(rgx
zl?<joJ_J$FV$gdt<P&IkN1ytD+#(r3FL*i~mN>WoO;9B*1HTjsE+D#2;T<MwLqq^=
zBKs~tkCj<D-Z$2-fd&=U)(psNQ-TYml&>rk-sGQy-kE<SZ~v9Pl|yha*IajJS{tZ*
zB}e4KpKvu9jBuUrh|n=I(t<I;W214Ne`y~fC{`%fC)-n1?uQHxOI>L<J-uJO#z<m0
z%sX!a-eO=^W%CV|I&P4qzI#BBeR?PV^uRQ0rYg)7<v%1onTPa`a0og95lYEvxGwyd
ziYh!KgAOQ!n&xiBSGs!JcGJwILpPtxHPN7mi2ixF0p!L&vkeO2cxop$wZ78B8lGIf
z%62j%fm6u3BU}e8YbFJ4wyyreTSvgoe_gIfVg|Ji(h26f9@1r1Y-rQUO_8fg_naz)
z>4Yl~fMuR4%YOs_1Ph2(7~2`UUH9+*OaZDBOZ%O%6aXqthL<Xs<{KcB?!QTqzjsPW
z^O*1-Er7@y;o>C#JjQ?I1I_rnGBAeALqJYlzdo*!y6041f!!m8@((HWI~>@s&!j&+
zC|?M_XaY=MibQcr_X15&u5DhkU%^z?+h(T!5|{`QytjbzQ{uh3>`geQw%;TPmw4a5
zCHUm6(qBR1brb2#1p7C{tlso(c>SSY)c5h0y~Z96CCwqMU?zgL4^PY#q0kpEqU2(j
zFrw&IlN@8%Wbt138{Q$7LcvG73t?f2L($@9W4piA;0T=57FU{imU&j~>aOFgdx@T&
z_pS!7-mylNDT9(VgN}b5h1Vnr3YL?>2e(wz?*Y)~Q_39n;bDTN#GH~(5cEcu*;B5Y
zMvCi^wfIz2*{tk@k#{g7{D?6=l&bvc4S>vGl?U)$Jb}V?KI831L!Iz2%yr~oloZb0
zEczc`z;)W)CtpJ7ppZ0by#6<V<GXCVyuM$87i_b%tav5}LIhidZ)u^h{z)i=jvBRP
zB1Rcr+8_Vv4q8pTTtG#{9n5im=yM@rqP2<~Zl@OTeZjoDm=WNwP3SX1I5d1^D0>%t
zIt|TyI8@AO4G*bkY6-kADZQ6$-fE?OtrF{PXU7(jv8aVd5*ya$#k@h<58ssNe8E6B
zkGJ+RDAH$z$f%I>!%<Roy-v3y;<aYfvrnl0wD4?pB+!17BK&D6Q=S%}lX(({dm@HT
zUR<cCObE$4-?XCcV7}l-5lO}hMZoT&7rE~06S1nla^`~IuUYXEwTsC(P>YH_2)xgr
z0+(?;F0DIFfGXT>QolcVWE4Vl_h2a-TIuy`2!&uT+%KiV>GEV-R!&Z$+F>~&E^ck^
zev+~Z9Q{ie?nl2nXdW1FZg~P~*ct;oTLs?|R?ycV<SJO^J^eFLw=T<XdV8PmcY__+
z&Y5zdqgfSrP|DLeC`6qa+yt*2+)9{f_@^SzmK`lx&2>*4C0O@|crbk<Pz*9NmS0m#
z>csB5nz=)?_O<}SOAiX}?ruv*kZh9+>QPxH*mzbKK~z6T@`xX*tRE^iatY$F9V(}O
z_+@8LcP=yHr>;j?GU^xPmSsZu!}uM!ae+A<enn%aK_Mu@1UnRBh*oo$e~~F)Jhmmi
zie2-XMTMwF97Y)!_$Kyc6G+-@n89-mqrg;zPZg`K6JkbyQMqgDs*3PSjTcGcr{S^R
zR8-(P)khXWn2%(j3c$c@$?aB5C?P~^E6n$`B_MQ$gb)uwU?w-jd}uYkb@@pNkeTjG
znW>je&mkU@%2-Z|Zaj>KcyK7cf);{+fEpzc=x%f2(9^;qX2kWP>GCkqnmUKb<Qqr_
znBE4gkgPLNAU+7#p?kM~4AKriAh&E~11Rbul9Ik`O_b){R(~Gf*|dAN{#3R;4)ba8
zdRNrB+`W)53%1d6#8S2v5DF3%Kl$9?UkSwP_VM+EYtNpCU<|4e1ml1u4e^-wQ}gl?
zBdfyU;kb{n!?o@|rs3fNkGlp1FCGmTIp%Xhu*SU?y|ItR`uSbotcErO`;2{Fg_kxY
z#QRCYFOBKYYQ)OoVo=${spBf1DWWqY?$URzexZZHZ0TOFNFmSN9e?CL))q?WE>Mby
zv11*rF<W`aL1r;ff5+UDi-CwJI(YTVf{m0OGSLVBAN($5lpy$nAU~`3D`e(tyF|Jy
z^%k&-*03bbXPDs8MiZg-QG-W|3o{fXPC`CVZy!vUy9A8#5x=Mu#Dx~`mCj?P`~7fq
zxv&A!mwu2L*@WKDNb6za3fR$=o%C^X<ry|hDCdi2&tFQT2N%<B=uJk=mjM5jg<eW1
zlmXnpGo=+72p!%rD!73U#`lsTA{%)*xhRV8#2PmXa?B5*P%fKyaA|NT|K}Yvksd`1
zqx8drQ1h61fJvGbBbW#<gix@KifC8wzeZjqlR9c3n_A{1-}ZColt}bfM&u1~p&)Ol
zAxwd1A&%Glh3<l_KbEm0VS(;T$&JjOu9)M=(q0vgtS1~CgbUoJ4vKJSrF$UUlMWsY
zmXh1Ap&R{$XAKs$=qODP%mQmTti4|2J{!TlQXioBAc1!9r)!Y#k*G823Dp@H8215P
z9d%avlQzWiv(i)g;l0BrU`NqWd<BVva41@UgjuQT{ae{iYCM?FxdU5Nn7O#Ft{}l-
zritI$4!IMSljkrGkDEOWJ+a#3qh7=A`2x=QUvj5;jqC=Mr1BbeE$zcSZdF@0kIv*&
zOn#Av1JGhPe6&1XVvJj25++&yI6Td{bz_4x2&y3X<%lO6&(oVGW3_jCMlU76_TvVW
zOtwi8X{FwmCx4cHMKkHuJOeske;+Y44+KU$3<dSO9Nf2Tnaujy68E0Vy9Fo-2o5)%
zO;v$R5n{NL{J!@tPi|p5tt*Bj)3C1OO~JxbY^J+r;^3t6AWv#S4>}Hc(`>z`QoiQX
zk!Fk+aDy76&%R+xP%p2?SR}8qpp{zQ+@(UdltP~vC^`*Wqaty5cuc!LjE_IM^4hbj
zo3;~tsIVPaSM6xj)z)5Beimvm0vA+oY$P6HP#`}dFnL(76<hphwiK;9hGQLXZQYjw
z^$n%ut4Jeg$`#R?Sph~_D5?Y$IFlOhOBXtN`ZUnL0n`=PlUP_saD48&cQIakgBl#q
zs$V%SJ*p`yNJL5gvZRqS_F$6)NAjjlLuO9!K^$z=N*!8_zH{VO*wX4srij|>k1E?_
zi+be5!i2bHb>0EaOQTBp_Fzg(&W$s1@H9b#L;n~*Ectn;txc-RX~UH3okjDXs_#sB
zb+#>+6|22So$Us@aqTaxODt2%Nu98c-hX>CCka`h``f*mPYnK;|7&GMUUyx*E(v#-
z+<6EN>E|OaZDsZo?Vv;ZzjJ0R@kD(+v4DV>NJ+~>;~<6%gl}%jT@iy)qxVPW&p`ZU
z|Hr=`0iU^#rLmzmcl)K8P+z7z<aDuSOUwRryL}HDue@<gXx11=hP5QQSo3Y95tm8?
zW<G`8e&VdMuRHbj(PTa84x5_l3;>IaqSr?J9Y>y#5E0$L&eI6LUzls}<e3q#^@E>r
zQbUyJ5w5rI@6`G}58j$6F&f*s9zR&7?4=k}Lo0C4Jyo^VemaU`-|^kqGBxSTj%UqS
zmVKxn{E!B?YEPsaM7nAm5D+?9b#uchg#ukgFTkubhWVh>+a|_T-=Nq8(Vyz@2ahT-
zPUjq1H-xp`#5=A`yQ|s_kx(@XP!5zm3U*)AZ)WEwRz9upE+;80yudi@aL$C{YW&$3
zp)+*&J4{5AJn6e13{}N+(vT?G6*frGwNSc!DLj)eR`E0W`$^v~UWZ9CChd$nma0Bo
zi}ei7d8nw#xBwE3CxZAO1&7Q~2-JHb22#ZeDP3{MJDGM)3O@fivF-b4<K}ps?dCXo
z=apWOdU^5dO+6b{)6Wx!oDVee!24$&2Q+~B5e2scaJ%s<Vm_O1BJ110=jvH_QDMe?
z1qk7pDn!qd|1=}9%wc>wKj-0EiP+A7Z`gO=)ea_uRbmVLnatj)<M<FD8T|Mod}ech
zmXNr6vd#k`0+0g~1BBB!EqLvr7U2PGk81&-ATsf%NA6y?i{eKV^>=pJDRVzu*!fb3
zr{X*q`D`k16-UZiWVu6q^jH8_t$S`W_=H0T5tb_WS$}&IodKbzXRLaPAmIZzg9bnX
z5czTKX}B(f@qp*!pa_}@ArW8c-H$MJPrla=1v2lZ-)fXCRa&h0RI!*8I-@YQz+K_@
z4163Ep<qyvXFW-z@}m4pKh#&Rp$gglkeS9O8PLB^&c!7S-O6jIL;-nKoCjuWHk;%q
z+x{a&CgzQ(7dCrMSZB*i5!E}RutkzIAaL1tIMinGzYK9sd}?apnW0<v;h|n<t~q7(
zE)F<J#FXG9F);pplH6&A>im!=i3I=Vq+=kG@sA){-mFUaceU0H7Ov@Bo>eZVAp)=S
zZ3WA453nhU$HjR(OS3y@FaCbsQGO}?AMzG&C`(sxk_P3cXX*oZ)++K##D$J$I}LuY
zza1*j7umdW8S&ANrA_Siybdxj>>qNqKR6N)cG|UhG@35ppdU-IzP=v!+?<%tr{?}D
zXGRXxfg)i{dD|YUG_5D{2ipT-+Jh}it&DnwJW8}`a$a^!n8-Cp<mSe{r=eF1J*lll
zhBD6zwlIsG{&#gV_f2V@+JRGJg}0ZjlIHI8rv(jUczND`4#W&r;kC4pF_ZzrpX2ce
z6T{fz3(r0*Tj2_#)hK*;HU9z{#v&5yg(Bc+B9ry1c&zB9yu3_z`)A9&spBZE)A6`R
zZJnJ0w(tHE*IXOF0dSBrivfiYe=nO|#Bl{%Sw$rhlwHrx+*<4BDV<s)xg{x3zrTTR
z8n)0OAqE~EGR~#w;{`)c$275=d_O()evbJZ)q{%@*wIJIRC-u20oQkS-fk2XtfO^F
zo|?|xWTkeS1TQAUYZanrAApbDK>f0&s`(Obb=lHde3k5p_<6I5uoFY8@lk*E65>K<
z_(I2$hy=?iU6z~57XvsB$<ObAfyp-AN1;Um#T^s?%Lf#~AtD~#2`N#2LJ;Y)OzB7f
z7eeR`_mRNwFJuKki#0VOS#NEaA^IY_^p4Bw_gXE7e)!=k9u@#iY4C30mra7Gy>tj(
z%PfKofEkjz$>9JtqxdfxgL^%pnV%-z8^ySkTm7l>N0>*ybR(7WPaqP{%!oRVg{Axr
zc~zbfASPcZZ2!o<RN_Q<W4<>yO)oB)=rzZ3N+f2Q0YG{$6-wPQr5Ds{0Hd^hx)b*5
zxj7<0D~qzX^kxi06#peyOk~D@YqXy|>Jirau@eB@WbmGAz;?7vupgW2(uXO&AzH(S
z42%Kz9|vyF!!4!?owg*veN>n?G-izk3M&XA_#NSQ4nR1%rgQ^pRiQOA!W*Ca+8fL1
zHwpd&Iurs7UKH2#<9^zUh8?RD`+GB<FVT>nz&CsxbWSD@V3lS{Zz$>$siyMR!;@+X
zzGJ^*q`PB_cEq!x;i^v)SL%t@vY4?R1GtXI3_)Ju5+H${6EN3wi1Rk2OuiPM%$3z>
zK_30X{K2v&9mRluRh1Lniu^t+ZYig|D@h#(d1uWzTI8n(^#3z?yY@skkGNWFAEv{D
zK>pyZfpn_%S2Uwijng|&_W#iIIZx}ODjc1)3rJK>IZckKbxA><^!XWZ2y}+pkU!(Q
zD@t&K3!HbkI{1Vd_lmVB?*|Wpgyn2i7N@N<JMDe5N#g;CLH9`rgJ6CPh~Kw@7)in(
z<DRtp37=nk@(AVJ>>F7UV4%R>m;D6HxQwzg0kY2xG=}22fmDY(r+1tdU}tq4;Ei(_
z@xs6mOOFok9S`$nfdEQxF$Dn3SsgvB6(D~sbeIt{!xT7*+BfJ@Fx&Gd8FzspH4|XS
zFapo$2J81;Zyxn>6jgxP_B*7h(am!ruSv890{N&sYX(RwC71`L!DS-;^(*zzQ~Hwn
zKKP{vU&gnO@j+j})9*rX8xvr#{_~q}e-6L(AVu?tgC6pNi>=WAwLst(+<j42O1?m5
zbeF}gWT_SZUL8m?ex{oFgh6I{GbN|M5owX2CQrsF#X16M(WcRs?LbxIW$R3z+hqb8
z$PNk?iq)QAIq48HCVw*e=M%Qi2-BgJ#<@d`z*jY5d`Kr0hxL;>HL{ozF=j!-7h1sU
zJf7f%ReXbhlcFw`a4RGUv+Y#QQ~=BK!N)LVY|)Cdk{-#w$i7f;#_{+NuC<6=W1trJ
zsTJx65y6f)$0ESE3sbE4nQ)HWnzuRKZ++3wJjNpr1HmVl(cp#k2MEDZns1LBWfg=%
z<HnPquL}GS8JV5b$3_foZCSMPE37`tAk`(|V8FSH@*2Kqg!P`8_^DRu6A_M5U%Qp3
zl7BsNt_bM!N%g+z|H$X~5ud+k%{cV1)AW%24tku(-{{+M@C9lMj-X$^MwDNncO@MW
zCFIZAi!-9NKlUSwI!p*!-8){_yr|?5kw<)6@OzTs15r>#)Rw`Dd!~kWaLHH@p%0e|
zZuTxopqPo9_*hPk-_U=4B0!#d)-|XgPD0mcOxC5b2-<~j*(19C)C9u%cg%?N%B}V4
za!k~r2JLoZc}yoPmpoP0a2I8`P6NieA2B%WG6lnsb?A!#sF?3fUAp3Q-w&P;&!r9P
zvif+4$fR7-&14{ydOJ_eYYeq!#aPi_WS9r+r+ThB(OL{Av!EgCUn}T`3gXQ|>dz0_
zSdc$%kc&EPl>!e!H{+a=fGY<13B?aTa8&xCc~rGm38Cs9P}YRte@_C&|D>oF^mpI`
zGDr6sBjc_(4Cnnl5eH|!$=O#hN*nq&zS|Pi5ho)o4S)j3%-DL%h*l%g9jUL0G~*>K
zwg{XL$!ImTQS$_+T_Xw&q;RM?JBWS+7#ZH>S=oL&^gBu7m6rH2!?svQ!3(QB>EDpQ
zCpW8y-lFAKdmnYR5Xyg9;$m=w^mYr^iUe>dF1_B$Ye*X4rFK5Dqj{{cB!xPHgao`4
zi}w)xI9zuE#1k_G*iIt8%7-xUackq;<^&k@$lMsqNu#uiQRJ)GUCDziC_gGK4-LvK
zGX8Y-bPcBwt-*SevJGyC-^IGn^6jX%wZ-Cejr7DCM>+}LUsxE5jN_z-3nx*1hvEV<
z0K*?-`HU3O7x5~=^$mDKjw>geGA__-q?<Fm1Czx#`p@(Xdr!jPVT|^|8(2Z1;_I&=
zkeL^O;ws;f-hQcj(g|`s<SHva6B1X&9o%=I2{q)<3;Ay}75Z0nklc;O!NgIY!BTF<
z#OIfuwh=MkTl%$Y0NA0wjolAQhk60X;mEIFKR7IR(>#4Tc!%@IR2)_iPN;$=4W2lc
ztyc)DKOh7ti5lRPqpS$;zK@$Tc{@GhAZtZ`;!E3p=(?2WqID>WR>Rv9svZqOZK9yz
zN0GIL9}MWvyz#Lvz5mYCU*3TiVfeD|joQB_p$mhE7?Yl}rp$aRs`c32tep2U=us(c
zA1nP+XzG=Y5BP~tc$cs&wHN9WnqJ=B=K;1*0TnpeoxyQ7@T28-xzpp-p}O{w-frWY
z+s_o?=4|2bx5Q7??<S$&1vvXq=jm^72K!VWmj$9u6C$P>J|<8BCguY(;(?!q^+Tey
zj)O8)TcS^qtQ;k{xSsVC7+6ko5Iy-c4dGhWx2{H0dV=-4^SeyKjp^?j10f>QZResO
zX#35cjNB$&djiibrf&;atp8(0P(?V1N7>Q`>>j{aHq+Xk(cx)9Jf_AJb>0cWYy(AB
z3jzVbvSGR5dPB74Lt8p<7r7Ss|5^*y(cZ9ZgA|f_6ZsnV5%q~Y)b8@JB`a7z^(&a|
zofpQXfc%Qd&U8qZ{i16>5m-NQjhKRANW(9&7*_LQwCF=>SAu4jS9h*f2wV-yVW)ve
zmK6mv4tWo}!vLs|LE|7ZcY-_2#9_8<1AY#lzjl2Z_PpzFQW~_?GP~uowheuu45z&&
zANu`<mVu8EFPasekn$gJ6ZLCQVNh9<J@62+`#wd$(84}5H)X1}ne>GmRrNNT_TX56
zc?!=ktMHJI;PG@kk9kBwLdb^?r~rEgT{w>|85k~GAM?95tF_v`f^#mNK2>y}@dVak
zVsUIv&KyaYd%D4BJonY*t(}Y5kqb3oV^4l2&x&jU07na}#Ms<%0I<4j^zrSL3x?Az
z<)M;2I$1o~;S&&*;g#OyLQJpWRzBLwIV$kPf&^uFB3$;%LxFy(J{{ZY_Vgg>#xYbT
zc4L3NarA|tlP|T7<)IwVzl`B}sg(xdt<~xX+(^j@08QS@+1_(?<%bszMlAnj2|tff
z(Q3-Odss%1ADscg7ntn@ZbCfmg6r)?-E9nw(0$57os_{A=thA)tbjjR$n3_2oC$F}
zCXD5Dogc@|1>EgLl@fe&kw})M<oH$J+zBt<jk?E2$c$X3DFwg|uNBojq{2A=Z11H1
zB6T7pDk39fLZBG%tfdjGHMF_){qX%kaG&oiQNT)nPhAO)J`f?`#t4P^d<$ocT$>zq
zyf<@ZK!gTj|E#tYtRR-?_qim1c2wFla+_=j*4WGSgi4E0J8H;v^mb+L)$}Kt`nz{&
z9E5(90_+?Fb=AORM%3Rv;}wmPyKFeScv`CywX;cTcUs*h^2lSwmv0iahOzlIjIz2$
z4h9@m<w-S`)3$*8w$v)6Q++!{g%b$YlH2D}FbpSl{?$lFSDrO(^hvEVs{v%J+G9qv
z(gK$fqXNIYB?Evgl{9rP{0`d1$oSKS54Bcee&2^0JyzIxwUptTKVK8AdAxj@EC#&8
zbsF*aGT?s`p!Pee;Hc$LUw6eo^culiJU6Vg>SEM4b(P?kO3Zg<f4F}1<5@##3SE$q
zc!KcW|1yuQ440UWme)Pl$t&`n%Z6q{Pxa&rvf$7$`H=N10G)Smf8u3=)CFIY&|Dte
zmF>bt?$k3Zr<t5M1TJ`BH+`V8n+ZKlFL9Us`>r2?CrID!bORXa(;)TWA!#}IloZ{^
z98HOYpMTlb?7W?J&Ee{RP`89#+ohoyLQ;FrmRq6migk!%TiSa1XYeOZ3dAO-L73Fg
zK3ni-c^RXIuYLy@D>McY$vkV#h|mUe3W&%$*U&(c#MW_OL<*0qh4{GBy<`cVB1>a5
zTj{ABez8(1`?>!ta}+^N7nv3Or&S&{Ov;GqQ_N=IUom_fQ*vwwh0LHX6MTW5nl@C@
z@4$jt4R#m3y-~MXMDfJR%%^Q?-)CiOz^#(eR{c!1djp{{Rhz8hi^`e*iK>|v<ztQZ
z2O`OX!xbd|`2Rv##_yGE5Tqvhd&#{28F{F*@$k34Z0Ic%koeV(KTA>@li#%ztT!5I
z)D3<8b4<Ek>DC*K-if4WTr(WfWAh(Hlt-EjKEhJ49dz)P!Z;8Do7TQ~a^b;t-ne-R
zJ?u_?bU<Hl)bI3I=81>HLHjHVo!9xFH^Npt3k|3>j<b((f!%qS_w+qZ2Ke)KvY%<^
zhNpCh5G0PVRh#URSWfjB!r8xBeFA({<^@n$a`e!xyfEEL`{HT|(xJF@IPTuu{sA#G
z2VKe&%J87-s%eUED{L>OKVg{Dq{nX-0#&ODp+H01H{>DLi&ILxaBdh+oOEMc^g2(|
z06fjuj9;`6SZr6K+LzkH-#Cz&2~LC3b?IRnEGJr~1b*CFfHcI3S^h}ED7|JXN};Dx
zow0tW$6i&tiF%;s(=y2SxH|<hqnLDKCJi>Bc<GQCbPVQEDd;8+V?$=9sQnIuCgw7`
zJZP=a(GU!{mjo5+_P=APcG^l;5^G_M=8eq2>-fuZ+Gm57di7WFk<jjT)wXt13Eiw$
zVPqip&7)6=VRppyl3sDaTZ4^$w;)-&;dndN6A#hjWX>W4b}zWm90|ISy*9YF#P;|d
zKsRLn&&}^Gk|_ZK)rm?j{K^I1sP&&78LS9=CbyHHPCmr^2E}EF(<pPs@xs(GShkJ1
zY`EI1w7vW>FsmLN7x<~sW!&bT2ze<Qs0vla#B*LRvubczUN~4!6r*c#cc$rwb|Pxt
zI-6R&K&@eYa@Rck%d9cwt9KWyh+{0Z`gQ+ek)#3bW-0^u>DFJYQM=q8_e4}dtiC%T
z9{UM4*95ou)Pl`~a3;F&v%jxI?Mqo^cY}SheS3E%sIU1)_1#FJF=yM<69z;mj`p%V
zqKXGGbzrbc-?n3hD8HKSD-&w$nX{N8S|gS3W|_Lz!j5ph(q)frD}ac6O$La)NA1uG
zTmSk--3u`1ul^@V{mv6c`D^Hqc&*0(1<XZ36W_#jWj&LG`INu-2@uNDh;0S$v2QW{
zdL#_#1i)8OGpAn>Z#cJJ1hz(q?T##x{liBG2JQCmm2LXytkTbU+|@eQDU!bL<a&-)
zN;EMO<zr6u6M=D|!{9Llo_OTLHIJS?U}~ke9vaRT1E|WW$9sH^2^Iuv{dv}3^mjVh
z`OFHI$0kQj_wvBIlF3<OIn!-9w9jeYq`sDp9d+)bLFm069TX@pHFO}}Nr@4kGA$Ar
z@+P#~#Z}vaSb2-zl#b#+^sLB2ZAl)mt5D2ACFe+%<H7QC?!4?&NO#E-KfBsy4*v8>
zDc4j>|Cf$!(;+YG0wPUvY^lgalIFOwMuFnQZQ#5iW)|CdaKg;h<^15~p}72Nl1yR$
zSi$qN<yeCP`&!V9X9mSrChK@DwAwXCL9*RY>*?{p(QGp;*JW->WYtw5`*(^nhk>hU
z!Puz9o7k8O_5$5$|M4^KuCL`rP$rJdj)lV3iQ%VKQnpoiij<M<gzE!$?oi)(`9fUP
zgGnZg{9|2><YwNSxo=@lEa&1v`{x_cjw5jbw}Ndh?Cn`>@UoO24wL_;!SUlOtnZxo
zgtGcxg&YyS>~eC%Y+HJ8z5fzc^EJuq*rJv2wIEtT;Btzi4ZrS`Ltf=#%A}0lc2=dt
z5x<z7H?8X_n+6Q~*;(m;Xlvu`&F2te*6(1PgQ#PFt#!@7v&c7ws<jpwR9c>4C#*u-
zwQFONuePW`BueKdKU>7p-w+{ARR7|fkyLC!0SZ1ci5<KV5$~PVlSz<;MO2A-Ku%OS
ztrLJ5N5i*}>Y*XOub-jTY!STj)B-Pdd2(}etKK(j2RwFR`uXF5r}1o+#`~Vv8t7#J
zA^k(0OI27;L<l&V%^P72aT-XxiphuS)=^An;e!179$|!?o3?hsg;|>%ZC4Kw>YG$i
z8!VxGOMrgDqlDZq$Wv@bsisEXh+A)skS0X_o`nuRK|^HNe)AI=R{zOWmWV7DZoPmC
zx*Uk<|Di`@ELt3G4`aAq``+_zT`4Sjm$&uOE?aP7%Oa>!_>2#52n1YFe)(?}iJt=m
zOtm%i`?pd<KdYvQD;^ecSGIEHCJ*tzs2jF|DihgrchSEj31FQ{a`3S(_Z;OLePOM2
zA1TA;g9p!jHrPDM3#l?+tnnfD3Vi8;5@kU;0~nk^ux}raC^n@4BcJaG;wF6xm6xH_
z0NRZdGh|Kq?=N!*L9=&zO<x>$nTRrnzQpH(MJ~8Jp|Hao1?pcVS>$zJ`$6<%_EQBN
zaU>bpn}Ou`1E`98R|vY7bMeT0z!`4*&*xQr1b3DAZ-y9dKhwR2sEeU}#2SfIVt_Ab
z8AwV`)0d+E=7G1MXU$(;BDqc^-AM@hoEwWrMi_8*)&!x8V*A%EjfaA?YesJPv)F1B
zr{SiCX9RF<;}won8)Nx?iFQZcbb%9i6Yb%Wx0Ia)Gh%su0iQK&TH<_u;>R*aPrSb@
z3-UR8K{SfW7GAWl((^sokJv{?-Gb^}`ctvTT7f0UJtBavCpGV<r<JM5&$&6f+hJm>
z$Wb=<C_;VS;!Ad~6iBWqrJU0wZ`Zq`-&4Fvec3VQ>C>E-E|}h%+|Dl3gyYjyf%(Zu
z_vga(57;-Wj;=RMadDl>%i?0aYPxK=Nb;i?7bWkiA|D|-ByH?Na)0g$N}-^+wIboR
zyU01%Fu~*YXX3~IxGYG_x;L86?~^E%zvXmJ@jQL&sFZ#X5LdrM@#^t3PG$&6G4_Mn
zvtm%Y)hsfil|Hoz4ZF@x-%$xST2FZGW-M($yxPP<a?PdjvRJrRjk2sm6g7Wuszl;a
zmMrx0b=f*EsSVj!)2SFGs*zEV_p0b{&H-G|w2O_z=2;e9e$(J1h4}ku9z>&{G-^hB
zekC=nozGQM?u}ETTL`@o?)u6+s{CwgzH`EHb#QMvlF6a<IOq126zKf8=S%gbB{YzL
zUG3pft8RPy0hfk5;_)S3QMX+W);6upvlZ9<j!A{R{R%sKdiDp0j7|aAK3(571rL+B
zUXZ0{=ZEfptp*7CEd`$V4Jh5MP3*Yy8e5I@aQIkH*AXIzU@YX;D}7&k5Vml3K0kN1
zbhx{B`p38`BFev6FERc?R_x+!fkD(vk8w||=ic_L0RVBvU8(%ARkOsri;2m{3XSYM
z0ycpsgWY`H>Os6@6@L(si_g7|J_x$chw0db$r4vN&_5Wcpa`jzoAJl5*M4qY<R5CZ
zmAEO^PIXJJHg~|Xyesw7GCVGX*5-?>{a)rRW0S}lUm;fG?KNk^%;9H(rNa7ZrEM|a
z-NnKpCv7{TI?+OnT~32)|5wGuS;5+NBw|ZjbD&t+o}3x-{kI=!RsGWIGwaS%P0>;@
zFpf>to&0K%3+KxIk%8CY-LO>QfLuufK0WqKsuh+f3l>~0OD~$`Co62;xC-huRDCI)
zExLi?R@PF`0lS{))<2e#KM|sn?H#4RNcJX3YVB9H-0MzO(19NLy{38o@^Aonjrx3M
z7H{!o&|SC1?LN@oL#I7l__2F=S50jYi!UUZjL$caEq?K-;q;T<$1|j-)A$W>vL4Te
zq;<Tz*;ldAYL-KK#R9_Vl2tvEs5m@@PIhi74m(^WL5uR9ZXZ_8+1`vT4;hy&jnBm~
zXE|yEDB)YotTEr}KlUGX*_ZrYaY|Z$XM|-oo5|Ty3mqr}+Q2Xe=<Kur#S*gQ{E;)f
zC8=0X@oMGPY)sOR56%{FOv}!$$KLZf{Im72#|6IPe$W<8{qlbMn3&;Kp4|`FPirk6
zp{;7C^9}fSzHwr1$$0x=JcJHoGVDX`gK~*W3r4ccYs=q`1Sz>E7@+FPk>40NJS_i+
z%LkuBujWg1_P|?{-I%kz(n3(u=GJ#FPHBB;#~1RP_$~Pl`NGhH8w~>k++4C&k7omh
zVZ_$h#A56QtnaaT11$PeAtG3E1}FXX2vfFx`s|zuor&zD$qY)(V%yTkXb(|124~7Q
z!y9h{y~quA_$lPGj-_d+c((sYto;f{mIJ4~UE!dfXZBovEZ?V#-~?0xSyvoRcJ(qt
zj9wJRkaC43H0WZ67ucjcAklA}Bg~<i+d4;6St)Ez^Hs>HKNXo83(rvucM-edIKva0
z&lAoto-8|jROb1IR9Jt9(lnG@{IFj}h|7Gvc4Lk^@yM$$gr?%_gv@;u1?Q5Ud5zb<
zPc+k#D@#KAFKbNmTFLBuzh5XtK7XthQ-PaO^(fWrhTWFkg*L;7BW|3cGPdg1ezcQ$
zIvagz!=4Qi&z{bzK>}(U{5B6)aefVGhKCZkX8Y9QYx39?8B1k@y~5<ejlb9Z^xO+e
z-*;1zP4Dn&$aF-oJ|};Zcc%Qu-%bszNL`OOq^xC;n@q^~kKobnwC46Hvo2R7f$x*y
z*v9KmlM1}P5~tV_bpvbXOj>1oyrrvYlp{wSzoV&SfMh)H^IUW4EqS{Yp{Gfl2v0gy
zDtNyX2nmwMa=7zr`zlNx#?9EU3zK7sP86OZj~!~drS7+rEDwXJ#BMGa0UrG%u9vtF
zdAZXxd~OvXAc~W$GZq?`I=B_Y>J;xharj&OI5gsVCbG)wThNxySMN0*X(uP^MlGPa
zs5{;<e>9WLd$1}Gn!I@if4p^8b<1(0)14Qxk#15+#?IKrB!0|VcBo#V#LAB^dOY=@
zb1#Kmslw@#B60SBmuF~F5KJ#&zAd-3>l=^E$CnRX`BDU2Df?=^NCBk<;%MT>naNb8
zX2ppy@m)sMSbj3_+C}L;nO{cyKfWb6Xf;c_=7&)VWnMT`eq$ntLLu`N==O8IExsQa
zlk(Arf{b+v*~W!N=&8BGG^aT7nsrGY&~(aJ#i0P!(mtK@P2T^Pq@q*D@t$DVRY{9Q
z62(;nuGMA3@4Xj+C0T+j?Q<+^{kXZ3uLVP;YqNNjuJ_SZW4I5*4v&wsrp-A|CMIyl
zxUV-A2|m8k{WAWl___N=zINnLm944oXVHxHOdb{<j>52w470E@LApF|x)A$8vimuW
zpA5x(X#y{QOKLyv)djM=?N86xMfjv^k}Lf~4`<z|N(2X-M|3iJ#?G&ItiO8CPL!I8
zH@*oaf7vs#<8Nb?{lY{(qK$x*``uIZK8mgJEE6;`y&m`dAK^uABMe^~e&$R*yx7R;
z?}}yA82=F<6lTS5-5PS7YBac{k5j>|HIn0EI_)S^Tr|-^f+hnP;paXTp4$GC&vv<S
z`ZhHr%ds?0U+UOD>;K^+3;llHGuK%{>MhAlKvsaY*^jH0yMNwaANM>bxfT26DTroV
zS#zIdtK&jWMQ;e~?c){yq}B5DVtnYa9Xmx`kjbz<KG8Qg-jI-E(cgNCub}3G_j*HJ
zJ<m$pIn__eTRWcR9#VGjEwO%b%(aU3(%X*Jo-koFZ{aL4qqchQ*g`Qcih1)J=_ZB^
z_ea0licVhhcPf=W)g-<Yq9F&k)60)ExjP_p)tYJE&(cFnt}f1FKj!_quVre|(UTn7
zcy$u~_{`U{8BZ6~__=k|eTS`hh;+tYsk5G$6nOzyWxt!;uEf_SpGL1P`}<A$zES#P
zOKu4c5x$Snl|p2CbDUEcE1Yw1b;p;8WoIhe&i-AWemo_yuY9$oN)>)1??fIMUsftj
zF)l}avoG0@Tp<3+qK{e<C3T_APt4)xa5{%S<c6AdG`Hif0-vh?aDAj8Jv%#BN9fp5
zx9~`<e8D0~?RpJN<<b^5c!2wrdD;9|JZp6x)ey}$$GJ-<i=2~&U-Z6K6E+ZS-kh+g
zcC2*oxm+ek8T%20&Fl?@S-3m>!jl;TF&1I3?Whe|+xvr-frTFQPf;>j8Sc_g()VmU
z9?Sj0?W!M|Pxel%Y&pCnRO)`t*=}*P#gIMrXg}4Lq+78|jW?ohS?$n#gXh>ou&2iL
z++&5`MjyRw(wkaLYIK4xSJLkrP;wgF%bT70Zf0EGUlv*6)(uciI;oBB*$R3DEPyW*
z%xjK=cmxTNG4|;N3S$ZkmNZ9)37fNl$@p{Ana+A_n#K55vvVQR&Md}<t8-}=i>E=)
zK6dhZ`ULhK^vLDJ#pPv>os|p{OXu9}t0^6vxc81T(S`E!yH=t3uqchwo1T-l_|Kzn
zFM`8x-ja74R)M2Yg_<`v_c5MUNT7G-WJJGL`zkEqEOJ|rpmDxBn@UCI9I_4+Q0^aX
zVo=nCMNN>IM{A@uzvMbsSMZNRbeRE*qzPf1;qdoi)SgBR{!fM?`$Cx%G{j8SjkZB=
zsP~Mj%nA2}u;+R$04P*~xE{pd9oCEqVg|2gVmpKYaK<+<+7746Tfylr+MpTgk>Ux4
z#CjYbrx+-1Wl31xdEvu3l6caGmm^~~EVaauFs+jLRKAFG>>vVTDsOhq|Ec#anP$YD
z&+p#7Tf26+TQ}`n18g0?@#)H2QOM3Dr3Pzu;|ZV7?PYMnpQt;=B`=)YFZxK;9|jJU
zLZ><vCzldj;d~#B%@k2R=pe1<t#czCF5X$+b*9WIru>(>YC9;Tb1O~<B*Cn!4y<w#
zy%@6|sKqgZH&5BfO`VRN_LyBYKbhUl?I=C_?t8QZ=W-?gO%aE=ZI-ItY17@@ic*={
zRSFJHgGw!mXx9CdC#LD-&9M=ig7i(xQ5AJkWA>J{8WXj0^MM?TTs}}TUZp1i<eaC8
zVE&#}kh7nn9mS^;z!nec-%Yxj^PT6ko^2r8*3Vf#%PSmwuHgj&*|3T(><oVg-$cY`
zoH=w0y<o(FYc^@bz!~Cmxrs4(l=Q{^ddl#&xlvVBON)U)etSu~#DfrYS4O>Uo5L2a
zJO4w@)zv#Ck+DcEjuCjBjtd5N@7lG8H(q^0(Mx!aXElY-fbI2bxY=xJtt<Eb=no#J
z%->hd*|l6gCpww}o_9pgtYb|#zL9S5*<hp~-HX5&mCv~f8kg?ArI-v}$F8q%uz>uW
zW3O9a&4ULS-q@}}RKcN1q({2GwK(yC7XzxQFs%W41I0OCt!;YGC{t<|r&RSYHP0tI
ztizih*{v7?ovB-*LrYrv&qGw<S`$_rC~~V|n;`%+G^@P6d^@;ASyim~om)XYZp70*
z*iwT&Og_W^kKFJHk9jmwEoB+l)>fUjHzp?Y)L9d~+#y?apNuE}4*t%Q+466p&Aw>y
z0ZPc?+LL#`z3M_cMD&)^u=Rti+PLtBXRknH^?V^6ad*G9d!XyfdCNDd`7}>E74bR`
zSBpEjk^|)wS^d!Z%agoj`4`jR)i}~pF0o~1wbSP`k@m@)rMCAYq3?k}tf@&OK`=}=
zvN_kQ+VT6F<AZKAfoS!!@Xg!rSRYMR40)6_qz5!%t+$jKcg3b$c3Owiwm}~%e$NPN
z>4@Vbp)d-*q^HltYfRc|=~sI9;vaZ{%rC6h)bkVSuI1FydQo3ZwTD@tOA1c_Y^>A3
zSL_32>nePq^>n_JqITafehRU<d|?T7{5@*N<u2<fvy!@frccqwYo|NbWbmjdI>p*a
z|0iB9w6sMQ^<vzI^SdI<3sIqN10;)oa+6jYc2~u|Lc>kCbZzcry%;1LCUDV06S9Sp
z>B?3IKh$G1oUxLOr1exct<1N!9)9~(TZHl2P=5uKNNo!axRQBir5BBi3wJ1@Se4Tg
zJYenz#;+RrEZ56Pr6Z0njFz3a9dE!3DHg(_`18UxcBpWS4(diMnaf>lfc7@4-{AAN
zQgms$ntaNJ0-xt%>E=4J&tr@{4kTl^1XI^pSA%>TtT?5QZ9@ST|3T^~z>y_+s2zjB
zM_Y=@9nB##GwnO@)AD6Q<8e3kl+MwxP*yPxO8loS9!pq4agu0VKuC}x$d#>i`6l--
ze^=b(r|>_@?oZ-b&QH{ubnnN$r*TTp7i4Y20<!a`k7fZ)XrY-GWqjvlL*`G<SE@Qa
zWz@~h^&^ehGZSk`0fo@gG^_@u(#0s(QMUMR0@`F^K2V)$`^`SL+vq~BcRQAB@uOr0
zUe^I*hE-`LgYnU9R&&SKK5ZRs3k|jMEVZV7;tf*Vr3HE=Gt!WjZj(=k#5iQvd8C8!
z2@ny%9=*!36zRpJ@W|ldoGe?$JDnvQV@mB5b5|A#_$Izr#q!!(&c!#aoLm{6Jj+At
zJijca?TKm{j{vXPI7VPAw>V=wPCo4YqpoCPtOBo;z5OeFIpIoO;CAiFA6<tK?GZ$-
z8UW|}^`-%RXDWqN_3Z>=JUyfWZ%=e&8zTJmx5;wS!^$RDY(`)ExQfX1il=QJrg@i|
zzY^Fq%R<^;$1N(ITh3NmL#C|ww`yzlON*YdII#vYZu2nyn?Roxrsh7nwPfqJQQmX2
zkiPEOz4!MFzKZK<QF18lXW~+qZaN%g$6GsfTu&yFjj-_8l&4}F$=^|n$EYqw_%ag2
z(FA9vj<i+$Vn*2LlywNK*gosWPEWY;064QWl=^ZrdgQl!cKPeVmzKZEpHb3g*QbSb
zI?UqQm{FO=Sa-G2%ZmDnOBuyv8msGtF!Q*9iF08SB6$}y$5M9nQ6rBdF(OP>1Yb{m
zp_#8ymb%}ix;)73R;Nqat9K#_cyeU2+S;Mun&2?nXT9xJ1n&21zPCf9VZMWlYv;rA
z{a>q~mmH(C(+T#<<i$mlno=QlN+8G)KusWXgnvc_u7!yZ+&p$|q4ZKxCntX?n09Z%
zB=Zxg&2aaAkDYHz8Jd;adpUs+yYVIZHN<y5pu**nRFV-4HqF>fj-LgRsH~x)VD<y?
zT>&o+>63SvZC|xyu#e}TU1F3_Qst*ehcU#LFHu}Mo;?CHq44@r^UUID#Z1Cud*gbd
zpf-iNMe<K+y_4eAQ~i%=hmMCd>W7|;Tu#m!#T*1bQ)m!#gOo_fRvc|j*Bh|2#ylqj
zQb`_DGRu_-`PC?d5YHp8^$AXN{J1AWrhc*&NzS%v=<Rl4b<atL2qIoXVt?hA>CQsR
z%=!{0%=e@wjG~zN_%fze{L7yX^KP}W8yK%|ytbs)f79z;H!VV{|L!AoOq!Wt<`w9%
zQPpNBI$m0Lv=}qT`XTI57=r4))kgX|)bb+WS1)bX%~^BV3Ju<x@~cmVb}@^+AI#uB
zaRTFC4)+~)y{|IIw%NIRxV-;Nu@KZb5uv+!ql~Q%*rbF=%Ug>#eJQn7kE<WXGT7~f
zv$M#56$0bJ@+#*)4=H?m&gvM`khfNzW5AwaUWaa!jvD->&GZ{}Or)XlM^{k6P9tJ7
z2|#U%5Ls2DVzuhZG0O@6x*s?SR@7j<>U~JC=gOS*V))y<^}F?VYh%Sv!zjf9IRea#
zn%gK@auhN+e6P9wKkD8(s>*JA8<s5~sWbx8DJdWzQk(8jx<R@HDQQsY6bb24kZxp?
zD&5^FAl=>X+M?$?zxSMXjBkwZkMDcO*kcbK_p|T4?t85@*PPe9<~3gdhh<@@*2eDj
z*Nsi{5tIChqS1c@XHII5*x=AEk=b-;N*#t1-?o)$U)$aby(ro_;xtviq}S;64D9+d
zZpWOVTZ6=>`?NV9kJK>p@=W&AVHXu8eTT0(<8i_%O4ZO_5al{>xYq;nKN}_b7~x7_
z`o{<pK<P{{XNSzhzSz3hqE{w2`xUCS{r)*8Qlh;}ZBo@H3qZz{CG`*2huY@dswZ@d
zZ8kqdvPP9_#C)zs#zrZ}<XSm|m7)$wRPMueIG*_+)*|WWt8QY$3lx4SuJ~v05GdJ2
zQtO#0$@N^_9bM{{Kpj~2c*hi&ALFOtpO@5iM26;A785AGx$DyO=)Giz1|EBN<r4%&
zS@u-d{QTa9uC7Jc4DBaVbnXS>^3K5{0^S#&J!dUkLC}Gpu3ktjshFmhm_7zGEe(!n
zG#c+xq24qMX|OS;F`d4sh}x_bSo{33Zim3be&~SbrFKwaU%4KB06LJ<Q7MTg9?Hez
zuBshpxczOy>11}nHNEpmD0!cRP2NDt`dBIV7NhErl6wI`(}2=4%X=aum^fO8rD?*c
z6j5GDRvqBB+CKbT1b%a+u88)N8VdJ<g?ndKbzUSrCA}86>wvEp;PXcw=xU83MVN3J
zg_Wnun<hEE`qbq-<8{~*OjeCv?+O671jh$&D)#SCFFTFLXgl<rak3!4xcqJiutm!J
zLRcC5==J6R#wxo9s|<d_vu2k8R^ASI({u6SPW+CvPw;6S`5w~O2?%tEGh0QNzodw2
z%F?}Q=HwbD?M)r%s_|L*kP<CM%@_gBI)W{5att*87Vc&idzcDcyk21dOF-mZxo;SJ
zWX&DUGsyVepIJ91E;-fe8jy_6>Q{DNg2EXoSy|3*_QHnC*Z0*228K(V0UQgEeca-$
ziuARwYSrZk#a#Qyc`@2iC{$K06tUa|rK4ZPeaTp8NfCr@6NTmNyj+2}h}}Zamxc(U
zy7KkN^{8>3n^YNaau_|k{usP=pPA)-oktw;aACfwz6WWi>Cs+>=Ln_2s3uS4{>Ya%
zY85I9RPita7Sph;{BY>y+@S7#p>vk}B!OG6UoD4z)u!^8JV<zJiNE!uv)jrLMC&T<
z)Yk>NrSfuqbK)F<`9yCtBx-lQ8-q+ZoaL#GdO{pF*z=x_4JVKQk`m6@8Al&Si#5hF
z)>YI%E)p^SN@;iAeP=v;+O@dLCie{uX|23=!?#7sP12>k@$hZR+svSRqC!lwUuqrq
zr66NpZket1a)JK6?k!&c9tapA8S7V-fBNIC*=K)quf9DWnc#+M<F5EkIYMY|4BhS*
zMX4`xAjVtF)Kfs=cxol8f$B;U#-x3=Mn<AXgJpv_MQUb>kL--WP3?{^j?#|p;xelz
zhySk1Xxx~f=~SjV9*x&b0F+}f@;mZ?mNRBCKeLkQ=;qv1((uiu$9fXK`i*xz4kDrA
zTm?O<zF!_sS>|pRTh$QBVU53h(D}>Gr%ej6q+Ih5>31#!a^*@KR_3$%I0{h;ddC-9
zSp{-7n=4!nWIag&5v`qZ-*S&D4FGiLFIhBfSU*WK6K>L(#5=RwHJ?e}SHG?Dp*r&k
znqDelVwHEmxRD;a>hdEcWDiyca3A>Q&6-th*o%qMjZ=saR4S4}qj8;fO>kUS%T87~
z0X`HFyEqEUS0LFxhu`%cByPkIJ#6eg9-i@?Ty9g}pKJ)5;QR_tOO$wV(%rQ*Lu2D7
z?W9BSDwpl?@~*yi<UeanHox-VIy8ID3L-~yovPA@7iridv<8d$lZ5)$K0G5{elK_k
zMKYhp3|nGnvQ_R&;F@^#OEhX@jhDK_piGun7fuymH4Usey@O1z(Mp~FrprSR^U17R
z21~4aR^o~#Ls|Ha3|a!i8P^JRWYz4xFLkOvXsx^i<xQ5W=PcrAT4_V(8RMvk9P5=k
z-S>tFLQ^)^7gDKPi_hu`;suWI7rUamH?{@@L4h1(#;>g-m{Y&C^c_wtwlK5!%J%6h
z1sY)&`BCP!NPA0Q?ZX>N-|s}02Trke@kgPnE5dE>J$yZDo!ji}jjwD13^gKtg?p~!
z<m6anv#(6S3eP4%|Fof$wzo;Vk@0K@)NWNOc}T8NK2eCZ-LUqVK0WGXqw!EdhREx+
z{A$O8L;wZlR7p&H2i*h^A@p+(1M`#iw>C8@b$E+XF3GlJ!~JTux3q^x9_tG;E2he~
zL<WUmCq5*gS4<YK%O+cr;mn2rzK6@^(m6ccMiGpPNMhZ>yh0N>C0m!B_PpA?Fa0$O
zsIsw(=eOe`kl$K;WDj7WM^0B-VC!NTnHDOwu3&h4dRS`Go&S{1WQ#fc>2POaGwp;S
zE@Z}ZT5p9P86+YgAR#e3Z?nFIJ4`Vk;G%;W*aiwJ8~G8ZkfII=C`w%dkg?gln$HwV
zjE5!T*-N9fD1UWJoX*sa8BwL=UQi?q737CUTMc_cBn~a!{p7i>vSFqQk6#3Z3Um)q
z5<poV6w)u%m}<aj`jgXQi}3@4*8PfHfsC&AD3j|#7Z6bxkMecwk~e5=CUP8?BKleG
z;|EuLj)DjyqFs=`dFNQGHPir82ifKO;*Ep(iCyNHsihcpJ;ZZke!KhM8v32um}~0{
z7=A7+gmU<vf~@7+NS~@(Q9J}4<llK}dbQM|<rAuiZ$IUEmiJ?E&omR!SHR%_5tNhl
zN2TX4TIgh#5`X=LBw=qxdy{4je1K5*&6LV4y`HBNDw?PlJ%mx6RKo{~v%UgU4_+TM
zagY!yQNfh%aOzI_@$kMNp(+{6n`|;~tUK3F@}F{+PSxp!NmdVOkUM<d6KR~zPh|Ft
z!@Z;DTAAziywc!+jEfv(1BKW9hufnE_{EI>asj|or$Pi<PYt9EKkqh1hXKfjFqdaD
z1aAJL6&Bxb4}<e77vG$t5i6V`_A|oLH>{qt#JacU<nYw%SdveQ0jC-i-eS;j<h{#I
z@4QjYvnAoj-i5?>uU_8=F++21^_vIh-Ly(NaHO~=WBz0}PH32HuBTv)f?eM>)YA#6
z<<RbpcF43jQk+HRI}~QE)AR|w{xmgOlZyP4NZ<5NgJ*M3kps(=qz6=RZcQNtJ|-b_
z;V$&@&a*r=-qETVu^#j9b=gwQu6z`@q{;{!qP*_%fSY_VH)8$FKKh95$pdP)3MD}x
z4<{b<iubtAxxxWBXUu0#mZ1O^{j$|5D9BqVpSV<xiSiW39s*C%=@DHpPH?~0LVBrn
z+Cv~mGK1jqq4wd=%KlsM+I5OzgJy}v&g4OC?<J`LX?!^+o$I{LX`-h0AaL_SJpeWj
zzJXpKdMR_R?C(r8k+mhzl5r9gC?hQW3R@3WCv3PK)H0bZDjMKf7FQbr^5r>Dy;`Y8
zb4Y2jNVrBwSeW=GJ^43~BMtLH+NXwR<0BAfPR;KK&Q=g+x>dEb51RFRZ-`8(WuSmW
z6d7<kx+!9>>05~$uL9w}4gdMKh+<fEXJ>BsZ)JPjN&E9?;nlJQv2RLTE!n1&C{o?r
z0SRizS`e#D!bPH88!e!Z%2Xb{d%$a+p2FUYuIn}Wq|4@y>P0+$-6cv#OH~q33EM*v
ztI!O}K(1izJfHvvECewk1XMrj<$40STDPNv!fo%CQZ90%y~v}xIGv+5EwQBTAE3*B
z=>pEQFW0(wE8f$OI;r<sn_+c^(#reVjwUEJ+>BdJcCN<rPVtLr(^hs)oF|2+ObaS_
z+Q@;NHYNR!ezOvMv9JuVWYG?=RG3Wvk{`}clNw%PCnI;kD_F%|%pl6SZ`xFJDMDIz
zK`}CXpZS_oHB!U^h)rd7mFYKgz(*Rj4M&J~O(tVZ&9j^t)Yck4m$e}H!zy?id+UwH
zRcoB=l&JP-Jl~&xg>w}CzE%M0QHYNFTSVgTzy0|TY7VLq0%F-fFR>dU#XsM50S4t{
z_=4)6|Mur^6c$j~-M445bu$b9<2@||&>UESX8NBg`sauLE<FD;`8P}8_q!t~M;Lt<
z6!f2;NKur7p@??IfetYLzO6sE;KK?MZm>t}*#GefFnF=R^D;l>aQQE{&<9E-0IiGe
z&tw1R{rmR83ldoLe<}QzTY$m2;@_LZ{&`z}{>IV*LrGuioRSxr8~D`u<kc4c-cq{>
zPDqzU%lIPt$?^_6*a(5Vn2-B%e@H3<KG5SS4Tc8aY7VB~1N4=`y3g)J1b}}YMt+rv
z1;8||_Dl7;ME?{#9e8r`H8P?7=y8_oM_$yi1n`UtLU>@wHK0P%uC4}&&t)4VGLD$;
z^We1w+}W<fbV^|k0QTzsGbq(v#bjQW?lj?mtHE7BA|S0?lKy)4c<R{$HZD(Mo>>iw
zEN(Uc<)g>iHU7z6i=MH1b2dc2M_bid0`9}C<nWx_uahUn-S5YDt!o68?T@BjGbi?s
zZeN`Io)Ixl-*<LiKCkU+Blu+Tn``ZLr!&TiPnj^li8SC}SLK?e{}rD52T25M5|YgG
zYKMxNv}ts%{*7vSe!;T!^VI&=b8?}p=TwaR8=7tFUSHzgn%XK(R|L6vI;(SAD;^FQ
zJ;mt*nJry%`p>2JeGd*C9Z8;&=zE>x0(#sV$*mgE?hj-q3y<_XFG-UcBK=0`jsjQY
zb^$A-sm;P}u3PgqN1x5C3o&b!j@Y`yFq}x>QMG-G+iteNS`VPcw#*;ymkyjPY<riy
zI_`C?Y{?7^>Yrst`zWkH#HLw?3Nno8+LOx{Ub>@z7|?}#FB5*q^>(Dm%UCj!xj!mT
ze%uhY1g{!{*EJ`c;Bm(FZY(?RXpmSRsi$=YR|WSvwZcaIDcSY?Y;oKW&#D*EUCjpX
z4!K?kgA8su-!>E{p<2Lclx}UeUwOP(U|2N5LJBGb@*9AHJ|KcvxiR)eFmpLm>)km(
z(#kMmuYsnqYe_53huTk%K!EbVp~$$nF?mcEWLNXAchIXK0zhja{!wwkk4k`l<FXya
z2YMwFr<)E3TvlqDrf@d5AhW;(PBU853{qk)YVIY50X@ZYmOeJ!dcWXa=O2N3EF36~
z`rPXkqw28oa>oW!Uyt$TbNC%%SwN`zcs%y-6_TYTGACmCVExKOav;}j&p+yX-<HiD
zH=HJU#|qhV`)anoa@4p8GCl309s%X?OAJ(@j%_`+<f|Pg?ke-G40;ifQfgnpcVAm~
z|Ao7y7^W$i8lcpHtZqh_7j9&od8*g0z(^NKC@5Lh-@nLj-iDeQ^Dq;AlIVw*wlC=E
z2JPOS%r)7WIGm3y+CA&Iw|h3E7ufj$b(u}42~Kk{<EwA&y>qQ5d%ULRjn_MxKX*TD
z7H|eDH>U!C+h^D_a9nJ6z0If#^6P0fgV|BEuWGd<SV<pL+dh1*bm`q1>$KD6StV-P
zr;7w*4)D&`Q|H$&qk>g_$sPz{2~qP(v*z{MWA2#tz5|cHqf)cR%MVGIz1=yzE)EqB
z<)TOmd+N<P6Ocmaj%)ox#;#f>Q|AsJ)_sV--t7Hx|B)>aV~js<Ik(L7S$tNb#XD4_
zioG9O7jaK;@?B=D!hHPzZU6cba=43>o_77z8#gYp=FoTRcx;AD`D;ST{p=c#%)hS#
zp4J^Uo}!p<DI?SL8DNT)Kwg{OaqmT2n(UFaCZN^Y#@K>1n<dfJd8)^!r@??L5O;Qt
z8xbRZX~sAgn}{}F)evs5l0_R{+5wa!@E9W%F}>tcs)#|&bAOHK?N;||Q2X6E;t|SL
ze~PgDUU>+o4Y^lE?90#V&33zfC;o}Yh3S<FwKc1<6{W(lubnM5N#KFw5-sW1rFHIw
zkRoptC)tkcnHrfifQ7GJ!hL_K#tu-8!nzsmu{_#E^4^fb0mv}y9^FY@2x{H>0S5X}
zIL-aswA00FBai7NDB>QDm8#J`EGf;kiI!FSP~c{V?(DH2eRkBr<Ko&&VEI`W(Jka`
zm|`+r@WPsq^|PMD>SS<+QRf7M*FNUj>jHU_sZFGc?(WlBz-j95@0s#ely<b)XGUA>
z%}j~r)`&f{yKtm8-VvV2@OWKld)pnLV~hpsOA{%Uo^UtzR#APuUWMX-bGhXXjV+UX
zv}JVV_L;EBzH7DnPDjY0iawOJIxe2AYT|PR@C!`djIr5t+%}Q<@_2yW9Po8O2qh{|
zq{|UKYu9D9GW!V?1RxtF@(05WyK3B96N9=n-fB+J4n~*42MpJZy5{3)#95@yEa-L{
z*8<KsOrzfZ09DV4mlhIP(*;T<s+XXTrZZf4241828^)Uy0J)T^$7037x02X1sDY~x
z#8!WW0H;!uaGs~<3<_?mZrd~iyl7`UH^{O{=dF6SFQ@8)Qn{4yg~yzo)Q{(R6Vxk%
z>FI89C@1jj8gi}lbVn}x_}MU-$KI)n%v<gPb#c@x5ZQ*l@g{SAyx|hV80<H??8Ma9
zFusre^lrQ_E8v?87sy0DTNK76_~KtQqr5b{Co*ow<~~6vw^HaadNiTxeSt8PzZZ!X
z%4yX9lr~G>_L*$A^;f-w&C~_~aa8lypH+`h8OL8K3i%II;Va*Uz?nNTcy$P+g!0Ik
zth)92PuT`E69hhYay>BiU;6wVsrYi;^8^4ixrR0%7a~LNrygHZ=or}(+tu;d9mE&Q
z@gMDe9^r^9owqxzRZEiB$~vjFbm^ocKUkawo!8Xf5bSM(q)x84?DDe?C1hloS?3U#
z`jszVzttkYRFWTHt*M;`v9TSd_{n{I(O<5ox-{sXSl<L#bj^OK(+^~yG_zKVy%bVD
zG|cJw!I%1_XKoVRe-8Fwd}u~)o^pxh_9JO}1*29TA%i|9Hn?`c#XGoH>hlRT5bC|0
z$@VBiU6yO|5|BI?UYXf1MR&yuUJtws#FDReT_yJ2+DyD$-@cxkmgl3ev4e+$B0Ucn
zbQl^efz?+=E*2l^So_-baRHbz@|8mePuih!yRvemVW#P#7ZFm4bhE!)^vMe+&<o_F
zon;1itLz8DXVuTtHqK974&}X9W+oqaqS#0`Y}m(Dx-Dead`J$n6bgqa#<<iRc2x*C
zZIMU`L6JYJ=U)%8yRz5>SU2M&&YYjfmfLDfmv9JA<k{$NHXok8$~4?;jf&*tQ>&{{
z#jNiopT%^Uaw`k|vS%D!pUXy7ZytvcF3ZPfBt~B`+^d|plIKl*>dHtmdgOR~?@m7)
z&Z)UpP|0&8zkBVj!(Nwe;3Y>gB8-W*Vlw-A)VQKf9@%1VSz3uo_iEEtscArnkysI;
z0DMDDQfo=5*yuv<u@a_Pwwlw_8U2fmj|8`e@94cv)brh9En|J7uNu{>-c9;JINsU(
z89HFB*gI-ia+CYuRN*<JeBUjb{a!eIKFw+$*ZAai&12!ccX%E($G1kjzmf)A3HZ_~
zyMH0l7}sWVZ}WGy3#VuNpt_Bfu2FS}Re`9}bkHTfPT9G_!%7mSGpegAPT{MgkRs^b
z&BxWik9t5HL*C@WDzCz!$*Nx`EyU$MUn{93R_vH7=xGH3Stu6_6)F1lapDNxzV^nz
zz?ra4i7yH<Y`(0?c-6+$Ss0-w6L0EW0mx9ds2gCzQKO%IIA<06wZ@sDTopfW`L08k
zgY2E6{!VK9ZC%>oeFuwVXt&~NVi!t#ZkT3f&0Gp!@WWm2izXuG__1l{2DYcaUGoH}
zKA(3W&ceMZ<&F|!X<V~X9$ht}>YlWkfmfHNogHkNnoF=dx>gMY3Tvy*Y0CB8R0eSh
z>zT+h$qn7;^oL4J>HMgUnswavN_+tpdjfIWXR%z-4>mto;||uZwl6F1h-J>?790-I
zy~bI&<l|`pawN~U2b#+A30{Uv4yIkAEcPb)0H((m#{_8HuuZ9eBCADeF3)P-&y6-{
z5Sg!Dr6X#^&u-nsoNVufFV;Gf%Aj*ks=S(Dds>yw#NJ8DW{Tu=Hks3hcRWYmYdw<H
zhnoD?8Tqczq#Jv{n%LBVJzGnBV6$huAjtSWFVwlFMj`1bI(C;<le#yO8xZXQMA;`j
zDK%*lu!b{RA?||vd3v9?I$pIpsZwuFaDft+iYeU?;w|~KPC4IwT2AEBE*<T=4UXNd
zmO)c(GX;RSUrW8(xLcGtK<&cBzA2p+Qmn^cQKikAf(YA(`tUP^ZCCFTayT*wPV8YK
zD|(P_GZPn@?o`wo(@mZ-aANlkP`bN+QyTurA(AF5xdgPTED0`vxQ^1~o$_HUHze3P
zGz3`+nasFFcq2yksc#`EnoYyfN`1kz-T9hbI1kdT#eu7H_N#(PYuED+S2U}DJ@3<v
zVh+)*4=GLZ+$ufZ*-cy+?pgtKERf5)&RN+SxLb|8GE+akC6<PP0hd;?N#rkxEvczl
zoMdL#$|@LQ-JHx9w>hTiOTII-q^TJW7dvkaq8m`)MR%0Pa8E?pq3Ij#BF`}GT^e{>
zzkkGaTJz}Q>^zm!CLhIMRp_h4kI}cqa#;6i`|1wu4nRR+qz(Y>#3P)0%@`PoTqBh&
zvO>#CcbdIc7HKV)lJm(v@K+SLai2@~lj5`i9f4&0@3Dxv+#6wcIzN2R5i9^(8GFd&
zu@?XxExi-zeXmApQVsWNxx#6;;PmXic3IN%4wLGUbn4b2oTos1gZ~GkL}1_f*}1+W
zSe{g4d7hqGpRzNr$3}-rH%@V3yhqy7t-?>SN<(K3-41;Dh8^r}c+-+lnPj`N*zpXv
z87q@h;kO^hCgWyEPl~Z@vWjD&XYGW)qLnT0-kYtzp>{rRe6MyR9bF<VTSu-LsP<5l
zK;FK?K$KdB#x|ZZ^?m9we?!tEH=}VLG(lW!(fowVDwCyd)bxn9z|q{YiWm_rsaa!^
zM!zMJ7?x7fv{!ksuh%KsHmB~H%D}(w7jM)|lidhE+w{*<m57*fq0fE1^2KzJRxKzG
zI=bKOOb@w!`T-H!W#Cuap0gt4vbjYuUG+^~Ug4Byq5!I6kOkFbL%D4IERMH0pMy%`
z5}~k{*_TAsL~8d^^<I)t5pY~vMbHCH!;VhTQvI2aaj{VwtFZ#36ZtR{N8VxaNs%Lc
z7+0iM$B~nI1m<By)AS?;EafVU%$L=DhQ#SU(sB$Wp$=!`u&Wx~az}y*U4P?~eB9U*
zSvBu`wXOEXs`Bif+t>Yv>zP&kaA^o)8D2C`JcNf)MzaZ=D7*Hf$k_eF3i33T5Z%}o
z{F>^>H}PH1LvmsSRz40nY}i|Vsoqa9#-7>sF5M^gE!<of;n<f<<GT|yXdNA8P9zX(
zbwxrlG;<6pL25Alvg?%JenOEOzr*IfY2vI7nJ|~_(aG}%x3@VuIB-tgr_1+fU2rPA
zgmbp(Tvj62fSsyuY2}FxYIUusHWH<KUmc5mT;0`+GK{7`yHY#7bqz3s&384QuRIXY
zx9*=wkIAl)`nH7#<7c|fg2N!GhS{bkTu?7QulSx7;1OeiK4nnCsx0|PlcmyQ?bvK9
z?K~YMSc+XGA2!2xIpk+1lL>*wE^#S#W%RyV6D^?hHt3vcSHt*`h~E!Yw?#4)Mwf_5
z4b}u3U^QONLUC6L&s(O!%Dr1uusAToZv68_g_f||FAIn+^6|O~+cxyzw8wLW_eJ5v
zn=N2+yxq=t+z?@v2Q#{wT&qvCn(^+BHgt1MV)A2#n|sH7BMn;OeAWMy>tU&3yOe0g
z>g0-<w7RtzC@+?HC#`-(O(EQT7Os~DiIRqLEX$Gv#q~Q7RmaGp6VuE7FlngJ88E1u
zvLiOG)I~gd7xKljA&$B1y@8Fxr_XvKVW#T6>@LZ|$FzOfUHX|v%bq7K@fiJf)8h8q
z>>h=|dgs6LuPjf$A|?Y0njmuB03Hrv&-5pjh2v${5+xKzb&~VYga`ql%U>F~K>IWL
zBiiPZr#2CFWzaZ(U8*uZ^w+4!4E;_OHn+bE@5RxsCO}7@n4ix)Rx5M|ZN&RO7tD$r
zrCBX6$edPga3k;*b#BF1_YI;tvP2A;z26-|znrtY?OpaeBo~JiLUjEC+PPUZS!fW1
zE4Isj@1&85ZQ4OMzKkMt>W{(=g8H^{q$A%p`(xbdyM56>3($JW)PD4fr-{?%zO2GO
zjUU!H4}>xUX-xEcF?G6!NacfiNk%OwJT}@N>G}Xzk&tr)e0de_!p$!m_H54?um}BJ
zUY4&<$e^>FBwJ~o;H*q|7BK6q2jF{4mVZR#2Z)#1bTKUg_UM3t^=ySAOMrXMh7IFc
zcHpK9_okBYb7S&KN4muUiISyR2eF_Bv$U<{Qhl7&<qTT`D{br+mruF;Lw|MX84+2;
zEe&6xITnagBbbu0x+iM2u>PD(hK|ODT3{A+(K!-7QZB%8U1~OacJWKZQt{+i=<4$k
zU+0WVkbz#_5(WNlrQ??jOC;zFKksgvr;Bl*%C-)I2xZ?p`R!}+hRUVR?O`u*o9%{L
z?a2m)b2KJ?$ErPxi~N|icv3UpLYDQoob^(dI_L{Bxr7=m@fgy<s>AC$tVg+*n-08^
zi{3CzZ#=oKyjb0iU^MJAaE@uts5eZbI2p?i!c>RaE+$EY)Mbg59^VRDVTwypNORP)
zd-(iTvu(XF=Fhe(57e<`;<(0GA;G5roi%WW&n2$w&yy1bn_+1p2e6~CckAoNwcP2?
z+x0;*0cYVI`mcb>Ga0H6oh6?Dfq6)+5H{SI$F0#@#jE67h^ma*MweaCM1d?Gf~UkZ
zkTvb>@=hd8Um&&sHep-gF->pzv2LHBT;aB94z(m-RqmVT&2~*Gsb1=n>x%=3u#2i1
z?q`bEEz09xE3-?@bdJJhvFa@Gw*xSqcdpVg;#{m#tB}h^K;ypIYs&AOP#@3v`<<#Z
zNKCl$qMGe%<EnUv5~7Yp0SJ#&Hb(FvglgcE?}t}UOh$BSXd2afk1!tyYn3;263vRz
zv8a7lVzoe7ijPMvn{_~Olpb{Hva%%qpz}U1Y5%EV4`TxWB^F}GXY7uza`y&~tejSa
zbdx<9WgC;&Y^v+%_Oe82%&IyCxw^@SG~+xE9TtN(4$0ULF@Zp3c8{XGtgLe9KJ8|c
zFx$iQEw^fC2B4HF+f7lU{ZkJh{TN$ke$?nTcGc@-j+!<Uoy1G;6+R^0V>7F=EL*~H
z@;>=sVwomd^RzOj!1Nn=ZPf=Op?P1SyXpIlAH(;eP`=@?)E~FKJ~=!VU3-Q~Mr~Fi
zOE13Fx%<k&qH}MtVM)#bwVZlmX}U*pSNg+v|6p96=Ym5!6dE^j$dUArDY@~eJLMUv
z<mLdfXgO4Df<s}m{)FWX4yY*{n&<KU{Ci`oLQ0!0fzF$;h69Ux<tKrOV#g0gRv*i?
zaimnyZJke6P*Z2E$B7?rR#d5;v&7}&L=(NJHty5myQ6cSQWVd13($#_fhcd)c!m7_
zP)qysHG~BM#+oI(543qaU!bl7EgjFV&>s?#UZP1;Zw40~4OMu?984@iXT!sWu&pL&
zOY9cS6_y27IE+}&r)goeIxjjy!csblgB8L#qXmOe3=1%$_lIb=LY(vSv;Qs|E5%P!
z&G!ntsU(*USI7FW-|5|RH)LTVqJ^e72hBP_BDhhEhi9=i=@9JT(Tz}spU-1IA@U1!
z5Xm=VBJI5ZwRz#0tvU#8du&7>N5kn})v;j?fo2LrD3vIy25h`F7kf)gmUw!&1y_p?
zZ(~bFq&`0~=azl+aZUc*H2zC0zj$Pa=Iqvy*7?4IfJH@5+Xz>5RJdc(lh<l>>ZYPo
zx;QAmPQH`{zp43(uoqe!cR8b3X4<19$nlg7;c&|7eYuM?-kcGrx}Y3&VjL*0l0{Y=
zD9qxGg{|+7a)lKO+V5eMns)yjE`;v=;`}{PCJ?v!aRADxNCqwSCU#;|sP1c~G`x@f
z{4?IwQaSWzX0va9c>mDea5(qoQ0ZObVfzNQ83qz#|FiG!LM*jY4$R*7>ML!2$(C)Z
z#h_a3PVS^8cd^NsxeBWS0HU;ldX%uG{$}AQ_vzGY{6so{q1e7zAa4G68tJ2nGNzW+
zDXEJXl2o?g);?g2ZLqb~F4-Nuvgh2(j3mQXiyP^qn|!BRrEEdjznn+RATg@CGGOxc
zm-MlRR2mt19dc#+UG@wfF9}vs{083rtm}#fd8@~Wi|3S^+01-Sb;j3LK(5x7Z+>R$
zwN-@Z0@5p5pg`2tRc+loQEaT@FRX2&SdrArvg{bJ=RmNM?s0FQ4ph(HL6`&weFtMy
zUN0!NGU3A6R2EF3!ZI#bIudG6A=Nz6E^09fN9lL)-`n}pS$y3%Yx!0I64Jh_;_$l0
zEc`_l{uNea-7`4%;e3k~mDhbJJg$qCf71DWVP}*oQe8ByA;)@y;l8@1X3GB6xzeek
zFBxNLR~+JZS!m^@oV!iI(+~~Q?=-=O0PH1|RfV$+K7nQ5)`6pFG{+=Q>MklUy+1-0
zFv?dSh6tXLjX97%K|We4$I^6Wk!@FRXaI|GS*=7oR6~Bb8x75I<T*|tE2I*41t;{&
z%Woc*_qNi#BNnPFe>^xh1OK-wlGXdmDh^AwezAaK>-6x{_jMBB0>bCNdJSf;g!(+k
z9^@zRDY2yPIkUlX(d#O35XRTx6)BXG`BG`6{ME01j=XpmV-RQ5E#fI{ruTHYNPYu*
zk2Wy<=-p{FkSC91;Vey63AYW7=*@z7lV0pUqA|$aaPq94b!1MNkVW!(Z<K1w;*}Xp
zI=`7xb@U`bbI|1#PNa`1Jb&VH?xniyO8-l~vCaj*H9CfD=PPNClp;-GkagmF>LFLS
z<Y=nTCL+u#s$xqJTYgPdVrlO<1Js?tfm<m&g=*vWfz`5hz>*|j1XS@u^Z?$Y1F6U_
z|3k5MvcbPk^iM+SB^8CX`C=9l@18Ujqim!YZ<q)x@7?EH?fn5@C0RyPP&1<@{mzp5
zAZEn)6d=I}G*1h7Zx3UWawH0tyK`2^Nf(ACy5vfYPV2joRcK@S?N<+%;?*y*@T3h;
zoAA$+t?#XvU$XbX<tf2y7SWjGV^{9+`jAwz1Esb(@n`{vz+(7%D~W<P*FU?}Eq@Up
z8{8J3MSk>K4jUxd#^QZ!kYI?PR{2bQ*PgUt<2tC?jFIfO)MfNQW349D-FMG&Bl@yM
zZ`TgF=PtxESJRVX@#;S_ofMe18vKfU^eJUeOJJ)jX|iQFpzx-AS`610&ce49T{dVR
z(qmAeAYU{uINz?u2T7~!N9N0h-WR`ILz9mDTit<(><Zli7x`Z<K&3zmPodT0`K%(i
zRROP;m<W?-2krZ6d1X=b_iVF!WMcEATl-fgzhG~jU1PuyyWXcc+=e}F9_&6dnw)qT
zUROF$@;}sM)8^-UYt{1b-)#`R4Q-&kNyV*D?X=EbXE;Sj13hD2xR~iu?M7HkCUynG
z$n^tW|7CsRLba@R&5@KXuFc0SHtziPN;NBfGIoE9z~A|>Af5T)lazqE?6umqB;Hb9
zomhr=pz9h!4^X+QN`q5mr!9wG=iW>JUBC<ZxTq=AL2hl=I;<54pSlaoxMzrHqB&k=
z-DE_q6`=~87NahV6I=yBYk<wQqpo0evRU*Su05NCr?WvMI&|e|cg$bWis5H+o-20e
zW;7mH#mP|b_37T7UuV6KtQ-lQbP{yfOW4=CUIoq71E^Y#XoBn^AzM1A{{TFWpaj(W
zJXW+n7dX<5I-ibR{y=afdxA?67cK!h*wxJ}U_AAP>uiFTGu_LdLz?BAT$)G<1;;~~
zchwhua?AtrQGdp|D|6D2D3T`W6EE!GX%!_>JH}-(ci3`)KqaT_nZ%#WYCX!BLRwtj
z^b&FEc3mE#Y52xlklK?=Dnla?n*P)y6_>^h#92jvMzlgQZ$Nr<EHq=VgSV)Zt!~Z8
zL{)EQ3#eM;zNl6FeK`XY2nQ~!?6@wQh59E?-RpV6__71tw%4oorrk;W56#1qMorY)
ze_J%|kN7r!a^AiMs9+512vla$EiFx;f52A4B0#Sq1+)ohCv@3dj)4*smKpKQ{^7#~
zti;~}tv-{7;PlXTw>Hx|2>Y`|`LP3mOun%bQ2lyO2Ui!rA<q1H_g#U@t{aboCIhIE
z5&xOWW`I|Ysv)8sP$2Bt|Ie?&uLLw`)_@A8kc{A4CaG53aXTno_T@VNooSSgt~ZDq
z!W5vRQ)euT=!+Lr36Cva+{<41yMyBc8$m}KcB2-Irsn2!qt<XimUCM}Q<}{%(T}lT
zq-xhodk^G5y9hd<;!c73XLwLNxIAFS_s@AZQ;iKYm&mX;Jihn>kL1l;X<srLsGtJm
z8HgkWg3%HS|4=O6f{eRZ@%|@dcnTed4#_R@*69;eF>JsWU|_U{$^4!uENpPA*x!2M
zh^4@{rQ+cgvx^HfSmNN)nK{L)jK6dFP%)5dmqg^#w*)tTKw<d&=DGd<Z(x>?X=QFE
zG*yf}Opm|T`rj-D3Sw{@cQiZx+~keb`p0`Pcw&+GzTi3d{FiToI6V-w2CzjG|Mi<O
zb~71&-r@7L4*xp*8UNu!{F@<R))$s@8#EbEG6qxiXTSaHk0jxHr9j*JzpuoX$p7zt
zGy!_eaEet(L8@rnfbj9fiwBxT3@ZJBqmS-^bxeW<nr?hAHR%LRHv;M86UG{;VSk7B
zi|Xa+o=I0c`)zxB`<1bhOvQzI?RX+-_XVZTCX6`SunkTD+(>8+Dv+y``u>0UvibjD
zJtshqKXeC+;J@*>P5z&hpFm`|1&ZbPKP)I8xWJ^bIDB7I{FiS7qn80z6p3)u>;G7a
z@cC+kcewa%Gyb&*|Fw*PX#>$PHr_|D9{;_7Z~g|F%lTj#|F#f+*U7)GkR<%?A8ouI
z<J&&T39iJPeORs}g8k6X$7f6_fgJ5QqOe#N_%D^H8FxY~UnLr>tRTmS&k>(tD^ZE#
zDSbN%4$e8CLzF_Hq@1-2--4mmxc6+Wnr-j*o_Bma(&;re+v4Qx{f~jE_o*HMs~mId
zWAY#S|L;xj!?dDxfDi|rO>srDkdmy+$p}P2e??dalVV}XH6#M#zf1A*>f&|2PBjh^
zI)PXsx6K0sZ27<Eg%!4xyu7%M&I2lH>i#@+w(%BHtt9+q??AD4;SU*3lya@z=qM2C
zP+ubH(i;95f=D2+1Sq_TpKmO2ya+HIcpLdDqF_=Pf$_{kY}=s05&<Swv~B!x%<myL
z;(-r88<>3}2Ig8FH5#Hz3pJ(p#e&bYNYCdlH_!PqyoU=CAJZ^s0tAx*)|wH4gAesB
z7O=`#4&3szzt6cxK)Vw5a7^=yKllaJ?^)n-k>k=7Z-Ma(h%B?XG3h=J!H4qSdSrm&
zv2P(b9$+#)T?!CN_#yH(NnZKJU6BJvjV+q9S}rC0@jiOkrwTE4#zCP3W2IcQPPjQk
z4A8)b%Ee{RF#(0m%<p6ThtQ;F<8@R=N8u3NA2V$kQjPQA$4Ivf5|(42S`#c;Q*z?r
z8=7|RsUFlwQBl}|p@9@Yryl{Q51d&ry8D`eL5L3?_7&;Ke_t<C6G2(y#dWXQC41tC
z=-FBygK&I?WTcd(gpkC0%tO~y*aSq@B6$=Q4pyfyxbkA-2F~1<f!3qPS1`l4)1A40
z+(0M^Bz05ZEstrDNLL#xy!o-p%Hbm3kPYt>Q6!`DhR|#)^{6YGVbGea#Osc2R@ToF
znPWr8hRU{BF7)Sv16ifco~Ht(CShXx{#<?4s<&;3JD&epW{oX~59s^6&lm8J<>Mos
zH=|RZw#auS54CQlY}V9&+q#@Ja-8Mf^6q?3`~2)W^94a>AL(gd-Q}y3-L};GPrL+2
zSnh|_zIHhEl=QV-nCbm|<B~8?Vbz#ZA9<g1vg%j+@4n$nb2u%1USCGYnl4||e&Tv6
z=E94Ojg4~e-Ve5hRA>ZtSqZZ<%HDE!*yv}rA9bHX(v>@rHLm+@1KD4n#TR1bt!$nt
ze0%$iRB3hA0iP@QFTt<T7e4%6eYwVM@8npW3ggX642b7+K#l!g7t+Dz6!ROV=D3T5
z?mGtiKX<=UdzT*+9_KtMH&?+OCiCoDt^vp~22i)XQthMp?JW6XAPP>Nv*fDPl6y+@
z3P;=xn}%GD8qw3v*KydM*fhl!OopR-(;GDAu}N*;y>!J>UM`Fo2%ff<UsD%NUc;YS
zaQZ&nG}&SX^Zm1(MZ$a*5MiCZ`kv?eHW1JL+=$(!Dxn46Qd*^A=wq@X4<VU21~kgG
zL`UK`b=$pS8XEB`Z^zu!D^;EidDZ&$y08!r16ARQhWit;@bi~aE^WVYl?d_d+qT!Y
zF$>ve4Qlv{!5!fN+YNUj)EMZo6S%hp@K{uC6(kNvU7a6PH|%=E*`K^nX4Z3c;%~vU
z4!F}9^I^tj`t_~d^SefSf$Adhtnz>DI7JrX*?emF&mF}EYDP)t2mBt0HT7p?=r7J>
zeXU@3!<xf~m8U$ZY}@Hplafwg+v~b<BbX2SPStOSnphl&9OwtH2w4~2Y@RDoct7?!
zTT%>swWd2M$g~M1!7oEpC95WsPc5X|t#)8R#hk!s&9?gdv-)mk$)nHYGSj+ZZdGR2
z_cUO(d;~2A_lRO2fB5qlYq3!t`lb;1JaN$jjT0SRcP;oQZE23IlO84~;`TOaZfAfV
z-sT$FxxFqn3B#YD%d_}5{$Cqb8*?8gN&WnSfCh-ctbT?2J*i@C8LRAGmuYXpE(3MG
zQZ>U9M=7E6b^6~$uM5+>YhMXDZxEOb)V#xxCB88fP+&Gbq6F^SdnfY;M0sjux(#~0
z6n7+Shhc^^?_*lceIds8tj|*~!$Ep_U3jy%MEre{px0a84Mrzf;wLZ1#iVn{`Mi*S
zxK+MC3EbR?UQjcd_G7=#VNbcHR<sg+#2TKaTkC-@3TwE}!EC|s=MZ~a3kA~k=Q4zn
zBl(&Xl}m?Yqyf%sLqE!#@~BKflYpR5Xi>O4W##t}&=FW&P9$aOnb$14)Pw8dKcjdD
zHT7k+Rd^ULMBM+{;ip80yj*-W4$uzymQ!!@^5+4P<BAY|h#AIvNO5=y4DonM_rAac
zPb}5fqM34B&Q_;5Tb<QDA&HKC*QGlmNVJ=vpTO@ErU+5QGNcA;e%o$3#G}8O=6-kT
zYe=#dS0v;|P@Yjsx5iy;9fhQ4<RfLYDY?(2WtO)|!Lup-*Qnws?J~|gd#ksFuK}q(
zi`wvo&gF99{`sd}PW7G#IPbeBj|{JZS-yUFiK)so6R`Vcg4(5!AuBuE=<LF2jMy7X
zdNleI)Ing9$N7#WCVomefYW|+fH`nq=0lH{);@el{F;(P<+*u$L_<dlmNR!zG@{+M
z4#d;_=ff;Y(!-2ks-Lq`I~f<l;+7?&a8$uZZ0_Q8U!&TA7L9;8klSV=5)Ji<pU>Ca
z1FfG<i#<f|0t0s@tzk|~r0qPmQsfPsLNA5|_?a;@OU0i0A%%hG@+cK*X~_uG`LKwH
zg7oU#+PPQye&#>97NucyloS3I@Z4mO69IKao?;OZHUfRvS3%dYC)HRD4w>Y2ZLOP#
z)&LW%wvnCN+SvTybetQP=KX(gdgXy=st|{wEE<XeVy=6XW`im>^P~(8&qTlE7)uZ)
z>C(#m-+R!1mQ6cnG<Ay7Vze@FzPY!5y!$Ev%4gv~=O4%7e|NNefOnv4A_;d;fJgBU
z#`w*eVd$$^0sf(uk}=$^`rlqU!s_F9fIk9vApSa6|N5ga&8i=KSdrGoadWAEzJK*A
zEWM|6<2C)4%Z_G%53PiwO#b)L-hw-2CglI8(Spm0lfOhg*bK^!^miNzVGIr|e&o`h
z7cw-EK^zN0yt!MD<=)g*pcVP7*g&}4Vz32)pdazin2QN~hlYS4Y%?VX2Tq;a;d`5@
z^kz-_s{kpLm3TkDd!IT!11ZG5#9P2O#rga%7px8S07U)Lk7Z&Q@kLb3Z+x*^sOFN;
zFHo1TYSi5&@DzUZzJE-SRs-7Z?kelZeD%+^QWYM5?qbnrY=sZh$636Kxo7~w0{Rdf
z-JQHy?Q@9WKNWUfm`{T5&AQJ?)>b%jhLZ}tao>LEPj^TzIZe+<skp@;8JP`rc}rq}
zh<aZHc%@if`JkN)m`IAZ@bZJ=ebN`OEy#59YTT@%LF<;?&V07MG}GFlm0Am@)78Bg
za@LTxK$OX!S<tG2^|6;YkCK4HHs7X6d2>tz!Z!_j0AHWZ4@6jquB8@nQ?6T=wiQi!
z3fogcEw0x%aP!8SSvq-_9SOFn(}EoBV*&R=MSFv9mXin^wFrDN@y(Q&uJ7F=p_eO!
z7D$(6b+Z|+{eJHmA_4;&gE9Fc>DBRmDXdxXMS<=+a}9SBBp}3~ALi(NhLGQtFQ9fq
z4qSGec0&Q`;_J1xAr4dvXt_Q2nuWgov{e80N>RW+_+E_gSi@)(t*)X(O+EX{;FpJ%
zf=9&#@z2vHJ9Mp?QR8L615&xjzrhhG5aI5#Pc*4*#uh{uG4TANb31F;fwE`+SNjtF
zYEFE!YyQBc`N5D;r3jF~W^yShp)p{DJI1?Z4!_roFBaHaWDo`CuR$>M19I<ET#)b*
zZU6a>GYv0qwN)|s$B*M#>dW((Jn6OsF8J9}f$tCpJsw^GmUX;?j5`)y^9HHCGtGaL
z|M6zl%28f+VX}BnwrfozxlO&7|LVGm)^g;{dz-N2@baB{tI>Lg^uXz>8bQ1j=fek+
z@{keg#&*9$kM9IG@b8o8Tdg`~XTzv!tE_#ka@xVApSHet_6o)Ls=K*81%+J!_+|d{
zDLCTRo$n#jm4!w>q#GR&pJ)~*$&OLn?1TWEx%hRZ)~Aik(@1B0R+T@Gg%lSKDM>l!
z&i0!$YlrqRt=(NX9G1EawLgG<nogu73XspZZT!4*#@`&{iV<$i#9f*>8;_;<b*Lm?
z6AidfRAy_xedFWW6CIYJFM0^ouf*q1+V|rRdnBe*7w<NOuNzvyujWinq6qm6A|@Vu
zA*LJm3k8m7QT{X!i2qUpB}_W<ycCj4y(Vj_el@DRU$S?9I6JIBz4sjp-pbum*)$>=
z0|=NT+_W;8n@Ng<OLID~R#klSM{y*i#Ae>=sGz}P)540v#;!zOw&X{GQBO%xOZ_We
zyBR+vy!k`Z1KU@xFmYEbEUI3fc2rEh1nD&rLqo&LW+`7=?a`8X8iS|c#|Cv9k#B6(
z+gtF!?Pcj3xgp&&dff)!QzpTt@M(R~DbJMCc`mEFYx0lZfz3$)ZV6%j#fv|-2L%pP
zDUHRq(?GyVO4fhSQvTYiKhx7_O8=MbTgkU(j}EsoT&Vb`u9#LPaTO^sS91te@vzDG
zAZmnPOx{~*vVcQ+7$r%_n~X`lQ0$u@QvYXp5~VclZc*!dzjw0FG-}poc(2F1QBhGV
z!+CMW;X=MCSo(MW5%^Abo1K$F!T!2x*-&juq|XLSt&^{nUzOAN`fJLA5X<4Zu`G(Q
zHwFp{`3Njbnc`To&lbdUBqQvJ(-sxoJ@4kN{MyP?$aIwLr)M;2=-~0_7MCjhxzU$k
zQ=6F;^C5s0kxJi_v7bh?axBBC;#&MT82_QUiz(3r=tIKbGH@TK4>UG`6}oEvMc)8@
zExB<E(UEY)io2ECxnCkRDG7%rTzhKsD!cJ0+UE+`heNY20DHoBl1QRY4B}b3_&AN3
z%8Cbdd9uj*Z7N_%37J;Rdi3|kb2{$DRU9U-=m_7-X<%nZeft<pT%@y5*v*=|>0279
z`K}QDgRsgO$UD%kr4iK{0y?LHamhbuHod_D#y}ku%tp7>|5~;YG`9rtmB-1d`V)tP
zC&2hTN*F3aE-Cdo=vB^9uFEJ$2FBj%xMY{cfB`%gQTsBEn>C6BZ&T{0;1b^srhg&3
zCv(Bag28??`Rm7oU}DwIOxbn+C-!IjdK*Vio%2)tavqG^gv9*azAwRA@~1X~_W=U0
zKkIf2`pSuLMRs=5Uq2$b=|q^cc(G0~GS*oZX9cuq%b6)BSL-J7#c4(xCnx|81cw1|
zASx@uFv0Ej#Wj`~wPvA#6OH!L?$`B5m2>RhYOQqP#>o@Qrua?@`}F>X^2+BfzJm56
zH4M)ggq~Jd`?!o)JE0#8nz8*n1*Yo9_YwvVGp~Y8j(pJV;fGQg70$~C;J#a{KOo;&
zE1>3u1Vc*xPIpQJpBuAY!lF||^z6HhEd}m0zB5<v+~TV|&;?H~C<Jx5`$x9Yp;Z;<
zW3D)l;A`W{i6-x#Qwo46_ppJXkps)BqG|m;QQ7@FcS(ioS9IFm)8dTcL`OLa5Dm6E
zHZR1G7jwyg=1nybV5XJ90!|E0jkyWV-=e_xu6?;^7udV#3Ove#Sl7*l4^A(iu<FP`
zs|T8Q&EGjND3|6uydQ0sMZv-Gef@GhNV>g^Xi304xyl8gJzpbZ-uqC4bA|7F*keef
zbL*!+z{tY<c5=gRyaD`7gu+`4X+;_HN?<(_OCvE1G$U0nq?<p~WTHeoObV)-E&|}h
z4%mWYhfBu*tH9|d8G$aVVP+kh^<0LbmN3~buAnY<N8@@v<q_@9T{4+{mPWFvJ0c0`
zf%k5BpzrTT3xE-tuyl6nnzZMq9JbNwHWNYk>n?K=GK^$-j^q-@Rw34!Wm4Oob^API
z0T9^<DZu?<22*47x{TWizU0sAugJal_X2v=Dn092OU-I7L)?o^FOOOuVkA6*ml8|t
zg(b%GeTk^aIV9g~Inte#(q3;?KX>-bHuyew7oaiB63xsK>3n<X?&r%p3sFMsFE2du
zwI30`-e2512Mta%6xrb=#7IxieRNTpSSJE6XRks*UsuBm@O7Q~{GkJVFCkwtYEvW4
z4J<%sCMlBF_ji{ijE4LGc#s``B|2^;h;14L!M@*IV5-FH(@nFRI6`VtcNu-P&?1k5
z&C-)5wCjD(aH{c0uc`@YJ!zj1cjsI``_qQ(h;1iD-+8zv;HuvM6x+CkbQxv=Eji;J
zx47BJBKY<sWw+fTf<;3lxSnnH-vLqV>DUjj%tn;EBq8adjiaippdfMZ&W7!(nR|oy
zaPV<v?+Wiqe|m{_8k=?%DdCkuqmR0Nq#Xaj!ZW$#8Um&T{~xn_`V026J>47O!JA;G
zJ@WOj;>TC5DP%pgMn;06Dj3EH<AU38xBB??w&zkyZ}F>b%31-NTi?Hb_v3C<Zz-z3
zJmvPi`kuGIZfn5DuThxw!DhYZcHpeb{o6H_&}apt;^WJYTa|$}9wxH*;+w5ZCmy32
z&sfkLy9HVy(v7ddvr1kbRI13i0ZYnXZ>O85L(Hf1!{09c0~5KZ+vg%vB-@qA$@nx4
zj0z)I(s@3u`LWan!B9dx+pMXkXclBc5T_5!6&^0grwI*PG5}~o^h-^M)bQr`_3hZg
z$#6pf?Tqz4jSUTf=IPL{w24`qjxQ>vM{8IMNMJ~3M@`C7qEFYv@x|Y^-U$F6k%bL$
zCvf(Pi4aDcDAfyf{t<5i$d~&SO=3tF#8agUN9FaR-eXT#g4i-R$se0@K1eP67#vhK
zb9lN1Ad|ZAR&=AYXb!k(K$L4#P9tQ~djL`%fuW&dMS68vr56Qh^i6R;PRXgBr5mB4
z34S1cIeo?PK<k6Nn{N6(rhKP~VGo@_0#wLlo+ANBDpu@EWNJk5J<i^cPpewxLjB?1
z@HXH|SNPg>*=u9$5>w*PCd6zqtSKdo6PxU+*Rnod;{&rE3g)`X;c20K84NUOaSum&
z@3W6$Ot8}5-=CvfQ#q+L>dddtwAn)~60#vf0a&hPgXoK$Z{Y9?T{8bG1QwV=q;poc
z+p=fsHb7xEocL1&KPT$;7sIE64rhn1kD7|AFXve%M#0|H)m8`SCYQ0GQK1g}2S;p*
zt>cCs<D84<-&LjW>(@FWku$%ZW(lr1h~n`TH2x_vRG^2gI9f=hT46ri`qlN65!uDc
zcTHX}Lw5Bq7vS9R(vO{CZ`;zV%CN+>-Vl>tH8Z2v<<L?s13Efj!5o74K&DXuRZ0pp
zIOYf7R+p(epU|?><Hwz*aIA|falwA$ga~7OT&8q!&ZWV?Van3NC8@!szGZ=OTECnK
zn>!BdT{J{~BsVQBVj#T_v4c!&gLXd!bHbJ1qJc}Hix@P%vzvl3)I73V`Y9Ns%g-2e
zMN(_}>QEiAzj+>hPLAg-tihYF6TQA4x;BX%q3(5y{naRytTCdSlXdV9A<`JgytOVn
z>T3tdQ_p2IoK}{tiyt6yyaTF}lRbYtgpU(9n=MWB`ilm?Q6!$R)4>q!*X6D6EB*EC
zU<dGkw!HY8aPR{3M1$GqYerweE#q0@Tph9<7+!Xwpf_n{3;BnPwvaS#0Jy~d3*w1l
z11RCD&Y3b7qx8FTcYp!M257ROj=}qIdU`a+p3(Q2v?Z=w?Y$z}5}vKKA^%l0<|m6f
zs#NjQ^S)4yfWva|d$hg7o>^Bpa=2p&O#>VEHf_s+8XY(%8vxK!HDLGf1r|10p{&?(
zS4_w)r+Gs|aMtZrI@A6EbvAe4XnmB`$3R38P}1xsRU<?bN@(P{7e+F0XR9ls!R4xD
zA@Y_aH8_(VgM*_GRf6CJHr)TLinVPC#!O}8P){<n(z&)5w<vfY2wIt^81P9o!oWxh
z)=4Tuf5$70aIb4~8#MDFhr4vAA*chI+wef_-JQ-f5a2!WJfe!Nt->76<qynGIWZ%t
zc5SNhL&~sOA_s?Xbx0U3!e1cf#+~|e2s;MCoqhc}UNUVLvc~s3cg&!SVrTnB;jM+3
zoLEHIhT2nqLl1utqViyHG2KuCP=w*Z!iwXa#QYj4xzH=5t?HV<^Kccy&a8Ew9hE#v
zAY*wSAgWE3fx3|CI0OI=lor@n11Hmm8<C&>u8hngmz4|yJLYfC;uh+KC`F(BNOWR~
z(H3wfvvL_7>xI>$d|tr!nf<-B%HD>S#0402Z^JWP@%5SM6CyR+3%qrLP<T27pJ&Vx
zbdM_dg<Tndm@i4u7xc!ebgm<NdGVWD`Pkpuf0R&Y+Cn~?Z{gurSzoev@{eca@JR@X
zD@?n#dbK-Ap8axLpjj!ujr8OwU^Z)A1HM^Jd-)Bm^g*-L2~TT&5cz{^-$sH_+8WZ{
zc!76;c-j+L-z}_Gz9EchB-k9?)`QuSbM1?xR_P1RQ^qakNJ7O6{ZowNoQW7VB$%#1
zt24aUDkS*{NHZ>zjNAa%2Hp_g@G`uWqvPsm7T~P#5m6w)qU&TQr9kq<lJ0H>($vqw
zFmesF-%rbj6E%m_j8nPaa%zdd#q5Zuu7E=<EfrWZa<4P|@CO3nAg>TgcbGTdi)&{8
z$0wfQUef}t^L{(}08>=My;DGk5$+>H{Rc3b(}!q<mCg9%61hL@8y9~=2{wR`*BU>l
zhV;jAnGQ<)@+Ie41wX9ZHV`+vccIy~$f$MgVPHp;KzyY()jcE_?Np-#ycE!!%U+37
zy7(~#clht6>w~C*)n}RX393GiXgVAU4`No`CwJCMqQMTl9b0>hIxt`>4G|C!AO)(l
zKXQ4gl`=FayD>Om6rc@nP?;c3F*!m3-Cgi>_64-%exqY)LI?#dfLC<s*-X5My29E=
z-{VFoj<zJ39gDfslTcvm>vgsv&VRk;YOV2v;N5l?+wAj@&uP%A9HTL2oMNn->k&9G
zCW5B2z@0ZBg5NyoOlfZ+#Utk8%i5E|rniqupg`&qrhP00sk=l!d}Mb@0Ln*&J-%9J
z6p+s0{4+o_G_*~GM~GH%r<nrb5K#v=;HLK+_0y-s6;`8nzZ#$&_(ZYa{yl+E6s#RQ
z5bRovm71Va{XSo)HP;nAH87-GGoEG5+b`BGci5~Bht~1GVGT(|t;GV)ioR@Z`2S+>
zt)r^?-gR+7P(tYx>D+V(2wOs>5e1R%knY%Yt8^$WDJ4pGNvD8xcXvo`nmf1nIp=%g
zcYo)OasRwyI2?mmo3-Ye^PTVeywCH@k6gjFbA7l=!9DL;)o5(lxYVj$EkT3I`N#h4
zau_%WAaA<8Ez&z5h1;ti9KS@!kNLN>2Tpa$J&?ZVx%iL!G{HR8h4Jv--vjmU$Jy}3
zrdE_F7OjV%edEtf9}HMj0AbRSx-BKj57v0~m$!{KbDi!Mh64<l5f9u8<<+{=QcwJC
zcofrum}jb}Q-+2q0lLXE0CM7ou-*V;z|s!D3&!m>G>C{5XDXY?G9p-i7=nNL6XUR3
z;^0L!Z4FW=4U#ziT`OB-NpM#2A8>OUq(J8FpXk;65i@1E?NyQSb$A?45r0$jHD<CA
zIm(7QO@%Kw0WOTgKGHi0AeGZSGX|44AXx58fIMwb=%4!{rf)sy@gUl0^S-Bjg+15Q
z)(-jiw%zAu@Ic!XEo(cb18nB#{Dru5x`<<tV;nHaZj4jraKM*XwsgNb{uJpW1HV*@
z<`q-MD)JNGJ*i5w*d)Smn5+&1HQY2PY&OH<2>*f-oQT2(R_7--s+BtXBc55@7JekQ
z++k$?;~&JO8Vt!OA;&+j#TF4nXOJG6)v+TBol0P=6rA8BY4D2XCkoIxtg>si#UFSr
z&Y$nv63E1gs+bY@wLrk>Z*(AB%rIc7t0{ed1IY{07x?>rbY+oU!%E_PP<VDI)DL@~
z8NTn~xA++w%I6OZMzj*4)_q2g|H-*yEGLh6`P@tI9x%Xp+g~S{Wr~iXvj)>6-Nos%
zU%=@WYebSVMgAL{E78{kdS0iNq{kx+N9bMtg7^=`8h90K0*!?j!Q#Z)c#J??5Or6K
z=n9vzVV^x%^RbTT_-hF4n;x9k{fA}n5PUWLv?l{{bJH;hJ6niTAs4X5P)hCn+=qX9
zRjX{d>qA*&0Ou!x&Mn_tiDFeI!Nr{y{G4sw4^A=Mf4fen%ZRLe{mGg!#^T0b<`z0A
zR_#KLB(&Fict@-3d%2zOBzXwnAXj5J(BHT11bdAIp(zR7O9&*37U>+9({4Z2Ax&2#
z$b#$dGfuq}-Q+qxjH$_#33_z4S>CC+`(&1d+zA9jw_&_K@zk$hzb<aQ?+TiL8T<SI
zonytQk)XgzUn}u~LOD8JETlX}?5kXPK?P2K6!RN80Tq%Zx`|NR>s0ei%9t#2BrTL`
z*sl^#Y0zDP<7!ADOW^jo?t{mVT2=iE9u438zQ*+8mq}qTV0B+_@b8AFD5c3yK)IW3
z-$!t>PLxneDGbVR){-a3wN|@pEp6eC5?A|`9Q`6XUVui#FLs|gp*@0uh%i?Laa`OL
zw=Usc&<N<+qSMtwwu}6a(MkC&dPQYUQ+xN#JK>|H;%}?&2FJKJ$FOPO;Ntp7gg%hI
z-jWNcxndLABKj02o-WnA*x}L}eQ&+htFkBCjSP&F<w&5#fF`)>u(W@>>m>$kFf{<d
z;%0esMtM(>fCCS*p{ns=Afr)g0^0(jBLW@}{Hi*J12JEI1~mJ2|KF!fL0M{3twpUi
zZq_v5Q0W`jJ3_QsJ~96#U0~5r@w<TV1IQVFx5h7jE<mZy(d^wEW9qx_=+U7h0mftb
zZ)4_hfcMmr4`K#p{hfgy@FK_zjsPpZO?i`?`jSzl@BTJ2wIXn>lu=Q<>7C3`fg!sK
z8ufsga<kM)WJIk(cr{X4-ZwJs$Vk7p68w*E>;LSz{||=ffB&nF=nk2!x$MAN={1<<
zh!QU|@-Ex21zvOY0YkN$x2sH_Y7`<=yKME7wO49u=IYnmn$08*q8Z;7$Z{Lc+JEyf
zLoX<NTLEy!DzI|XBq+p3;x)(>@c}YP`Z-A7_7G3vs4^G;;eUUd|NS+rsMSeNew70v
z&lYHR{h{E17xMY%^Fk6p$fR#~zqv&i4Y<X!G*LLrzdGrnv3vtznnSMv{sMxh^iWLn
z*AlhrJ#h5nCj;S|)d0nn7i?n|vSzOcou(uMfCM<!;2WHD5^-0+5b#g$%m(MrXvk%W
zj(%e3x=CXTnUOLLz+EW`b-*A9*flCD9C*Y3b7uoT{bCptQ_ZC#-~I*CqP^?6MkdWF
zHCtf$sT#wB7R0F4ox<QkMDNqv4*^Ma#QzP14X?ozXIsYNlaK^)K7Nh?R1C&%U<Dnu
zwdtaxMNY*gh&P1ckRTB9i}&5X@_TYPz-9n=Q-};)N<$SLWX&F6tC<{ivT)Hwc_x!e
z`Ipk;)Av82^!y8yt*0kkTUafX*YKoe2n=>PH5FqYJ{}$pC1v>9n#oYUmIi)JWh-JA
zr;|b+jZ~DcSf0CA)qd<niYR0+C%?-<i}WwCrlh1>4Wf6N*E}shvY8WQWM$5Wj^`9^
zcI|>M=|1!4f)g7bpp0V}feC|#iY$YA#+I?5JE18iCYGv@E`j&JYjq&9*GO{Wh_%Dh
zn42ceiuu`3D%&7VE`<oh^I6Xx-G}X7Q>34l+b(wg56C|hhP?)Scs2-%@m;}ahEsZ*
z<0g8<xmuJQDQhno8CvJ(^-h-I_-kDHd@-$9Pp^Jd?M)oMt*mh1axfdewJ9=jpNruY
z<shLxVXf9IB#yK4HghzG>;hs!&P96@fugyQU~S7rnHJ~ue`NSjw*3Dyd~zI3L21$r
zIwV#UZX*_0J^(C^dI#^UoUh>39k1T5{OyE%=f|1v;Z9#TFx1~shvbd@8p-hF1udMX
zt`&ycd~}*}KP=t(+L=cHBScpm%qQhTjLD3<J`s3p!O9*X9QBu^*imdgpyu0iUOrc-
z7E?&9kMm|oYSZ04=fOSRJ$5KM%D4EImVN1rnzF0@=sP&1SF*q`o=#U0d>EhO!vyk2
zK7O;8zD0vEX`Xq^9fZ;3*t1^P6nXM5nXM;&z~xKjkxH$UCC_#TSFhR*wA$u_<%*o=
zhxbVM3&|;1RbIPRIIbeK>_S6+(rTkuvvmWkbIt;McF!?K-g&sF6#5~+3R>q=JSE?R
zZ9R5AL4$c52@$m%#v+e8kUt0}&-q&UvxRM(E-4aU2kue3IZ#av7#p8KN-+i2(KzX|
z(1t3^*uBPM52gr_2O^jIjKY&{X#bVlqxV1I_VB(d1;B_XH4+Q@VRkmt`@A|KHIOc4
zcs(g)N;uElob_doA|)3Ggd|s^-az|R)cxhB{=_M5CEF7Ey>0OWRRVi25|7wrh&bMT
zaQuGeKANSbh2bE1DbPoPCokFX!e{4u!qdf}(}=2J+_B$#6uILM!+Gr`sX;%*G{kc9
zNB*1a$ND2_n<%Cg>XS2(TtDEl&TF%Thv7~;?P1dkFI@Q@@9SED;MR^;m?rv4M#k<y
znRx=zM?NXe%fp(1TK$+3KKj$1Wc({HXg1ze5p#N4F<GekWVY}Qpu6<1*0}PrD5RQ?
z5USw-!4H5#E<L0$hj35+TRsp5;p9AHdtO^LssZN>Y}HS`t@ZF}4S=jZICv>&-HrnX
zD>f9NAQP_52Z%)n>=ZxoMaNskTO?R3c8;e)L#C$hYi$8OPpuz$&4~gmU1}xUzf5Bu
z{kE9?9uQwcq((#Amt&w9koOzp<BcPPkr9keQL|3>@QB3<42bY-5M3I7!5E`cy5FD<
zm*@0)ly<Px5*n2d$7xSrY~1Cm;4wMZvW^Lkq+urNC80yrB^Po;2ap4&YpFG>xQ$M&
zDWg%?X4n!LfBzS+kf_zYPV9(3$1pmBpxzUP%Zvyrn7CJ0BefmGNwWLa2gb>wUjf$N
zU=v%Q*<4B5wKrTLB5OS4e`FE()$Glufm47cQ80Tk16~ghjo~jRR~`Pio1`%{{`?o!
zpXR<~Q%mJpuhuGDcF`E*>lrH+Gv(L^Di)2I@65B+_k=?zhbTbzl^WEwgR2G2BPV`R
z)yb@=pEk$qilFj2q~(P#!Ry^bY+Thl0>ng>J5N_Js*f&y5<w6#MvF>Xx+-w<CuCsM
z)B2VEYwk2V!vjI--^V@lebMuGK((3q2r>wEr7>@BuuUmsk+|FuI@r}&ndPs7OK|B4
zo#ElAL~DbhaY`T^<ZzxEZ!4@a0lo76`l8c{WNx*;|9fiL7%303p97$HvXB;CC)tGb
z{o;-??)ycv`@{A5%T)92XF!h~So{R9kQlLdrst-K@&`TIyh-Nz61)evz**SVuL|91
z!S0lkL7puNS0+~(Ruc39DV-Ry3?sG#??>cmrh(iWn@~|}WSMqQ3DgQT?uR1iv!D3>
zW^#OtsSf+PrSTGIj#X@1cDtcQO=maA#FjX?Gjmj*9hw}6oSLtU&e+wpXPGm3AlHzN
z18D&u`ywUa%L}@2P5^e%AHd*GA(@BHL~ej!IQ8y>ycYZ^fPVa8>Jm{?5u;9zx*;Cu
z*Q4jDUE>{#Pwd{EBqOSq#OI<`nvQB497%|0n-;kI5OWLU18zwFc5&((Z)2i5#G@rh
z`&d1)1^u8un}jr#{QS1%)R{i)b4>?L=+!=Y9DqN2r4eba>t9wfQZS-}HiheRb!Cex
zWYgCDM-Kr*Bpy@tPyF%|m6$RB>v(-xlreke`(lPtCVpHJ$|I@>7y79q`q|O?n(s>(
zuD^euP7*(XuKc0G<>3G36Fz&O@t#fdX+^c<xNth6hd{UepYRE!=uT-f(aN|u6bEYE
zccQOu0OhpwbcGt2rPh`W;rDhTCpKXz*9X65g|%vK-BhTG=yCLY*3^V|2Ulnwe$2+}
z*NS<%KfWbZlZ-mP`$9r&NN)Vl4xOY`D-3703*dvXv~&l_f8R+k4yQhSBC=3=38as0
z;B@>RB2(-0@g`dQzB+D~Y1{|tbF^UyeiQ|QlMFx^|Kwb%@4Sid(FGKlz)zRY&}I=G
zLvKYi6R!N%snAURzA#bnba$V1$%0bhPcq2*Of)#~T_S{FjbP2r<J~R-2WWtxl#-}G
z`jQ67OeCqm_}~5lHoH$mAVG6PH{tpiQ4)rngshbx{#yu9{5^h;xF>!ecDOZcX2H`$
zfv@76bbK1(h3+#6yjLHW+!Fwdr$70fIJgtw4fX!NUY<zelq3s(A?dRdML|#v1Fzsv
z=-;PItJt$10t^QXhbrMF_@CKAfUF*p1BgfK_e*IJaDz|Z{wJ`O{><iJoIZUTYGVvk
zLDXtZ%6T`$z}a9Rv9&2-ID9?4Gl7d}po*F(fECvY>NzcB^%rzYxqx~%de&oaZQ=au
zvu60sRfQZkSDl}wuwf?yaQgRFMxy!)eF15ubri5}2vJJoM6<KCD{?bsqF*bO6V+SV
z-Ji2gc}QLN2RL#A73z!o5~&Gttjx^Ja}1TpO<Yo%4QmhMeUP;>HOk^z>^^-TKnL`9
zW;7F_c=>(;xe`$B21UNVNtc<ZQjg+feDZGzMm|P>`!cbAIsCeKpkBshh@US}hoU)H
zhpyh0%{ih2ITBccOXw(qS$FK~t16V8gBRFm)jbpe9Zje1i#!+)8IHd-nJ%7>C?h@t
zFVC780GQtzr}}CI&z9>x0P^5_yeoeqls&M8SJa}JqE8djAYn!pf<UCXHC~q52iH_x
zj+&K$4>}AWid20m^(S6|U5sGu>Wk^#B+)yKhS_G_Y_305d7NwmD>!jyc*Yz~0QNiO
z!-%6vF;G!TBF2&W&p7%}fmsPOXZ@COHxNhzCemo!Miuvt3{f{mb%os)vNns)PVs12
zwH5LC_{$?iOyK@2FKD1ceo4F*U({)7JGxNu0#jw*FO!Hs?G9pUlE{jNiozL|wpsgX
zKVlmNLs6@XJjP_oz^gy}KoqI|ueVMLZe4@u9hp^E)9iCLrs6;AjT6C|@R}i5Yde;#
z8ATMre(EpG{x`Rcn50*FCVvVp+5e>dz|4Mwap-LWSR5wL8t(ZdM?`p^oH&a~N~XiM
zCq9pTDaEm}Rgno=zlqz(^t$vuo|xYr$MF&WO&YT3QhtS%ZmLCG_85wZO-IR5zCg}T
zz5ATa;ox5v3<HQHy=dCO)1vB?pD0=X;Wjp?B3=+Uu#Rk!a$B1W%mwU!W)mk8&%3HF
zvHa97D6~Yh_hQ}~Sv#3p5mBpq<Dy(><*lgIYZ5rAsP{UU`L@`iPNVIXU<v)%vt<94
zZckCV304T!CIrAS^w1dAK$Jv@xxO@uPtrePWORDZ$X@wX8HFNW)(dZLG$&ISdW}?w
zXcU3Liq`(}7s5brjWW3-$~$9@MW?|uWa7#ijl6~WCE4LNN5U0aq8g_F2t{8RDq6Kn
z0|_=|q{pG;;32NT%S6apbb%HHe4bzRgB)}q6L~jc$9Jh0A_6X1eSwHV@;y(%>pTA&
zC`kZ}mV`W=n%`!I5Y2okOO*2>S;WpX?$yehCv^}t2L;&gwnN7fGjMy-+L%7cRp?Vq
z-v%IHt!n?7tq0hCD5VUdPqTRy9GCkkH%=hG@+4BvF|UX^pDsidJ6YNS@JXgSc^%Bv
zv(Uex$q(&awJ2rsC5NBXrpA2{-HonM=}a=}k&x_WY2p4!qO<4LP1jKcv7)2jk6<AS
z`9m-ahs=zvB+wztygo!LZ9&hL`N|1Jcc#`w1&8h!oXwmGJU>#mtl*hoxDtLKH&u58
zAvBj6bfNviP~h=Xy<+*$pR2B@T9p5T@0q$UsXwG|*uLU9Iv9moun7MI<)>gZ@c`uL
zSeF6qPF1%#FEAz_s8E@{YWb`@`5c0lH_Vv0fK9t4uXq==`uI|NT@?UiLT)TEQeZq;
z2-fy*X^AXl%DjDYn<XU@A|(n1?-5LdBZe5myU6O84%P<o6@o)(lu(;8JF=vv88qn7
z5O@`rIv^)FezC1d1770p`)W}Yn!4x=h4l{+zh0K_=VL4BCA-MhaJ-gbMBUE_m2M<-
zFnA38LgckDK3JELLANfFa}@U59OOSoZmq;6g9N7o8!_3qy)u@VrC3!(LjiNvGL|I<
zf|&+fDt+5)(+!h<X2CsyXX2^nc)nLC`YV+7`2a|<&@NU??c#|}c+(N9+e-l2B!^@g
zdE`j3w*vDIW_Lcm>)olzL-s<Rp4|0cRc#}5*uKaV2N+~0<U8)#RZC?BkqKdvR+Zzi
z?6>!#@X`tTrEN)HHX=lbI=(H0;=011rq|jLI|it@<34YUzKX`)mGWWlmCBG86p&Z`
z?D;MCw_A3KBdeM$@(fC~tg7>|6NUrZw*(C$)M~pc@ORWm<6q|##UW2i6`1M(xI`w4
zV6EMXc1>L?0$)N?L1p~(NxmSp>@4f@*%wz1)e@9UmbLs5b6!Ya0w8P5yh~E_Ohuly
zkAAiOcm?kwd2;bKp>PkiRUJL}Y!4h0k)wyQRlfO&_NkDv2f+}~zO~Lf1oPwasSoH9
z6k~qt>>+e;aR=`?L<}R30kzW%Q<xC36Fy@z2A>-lE^E_u=?OS|IVuXNEO7si$XYHr
zaj!Rv;+{<>djx@rJImBJjOh^6JqP_lJ{CVMQH2iseVCUt_>E;P{+nW<Mh)fXaX~il
z?7z{WPgKUVU@rJ;FL48n&e+GQ|I8OqHYIyuHlF#PE`gZS#e?x4ye8(_Es&!z>P&t5
zhYRrUghM?&X4zkH(-`>jRXl6SY|7QVoc4lNgIGT%<Jdf-OCLw@G0WvVN-{}_Msa=i
z3>h4cfJ;suI#FiM(nt)P2vMtZX!02EJ|#=#OegfDB;<?!=|DHfp_*VE@z~3_Pbgtk
z{V8Y<^_vP3SiUP9{xV)cNSaNa3{KUoirU;2dy{>5jI2VT;!3-vqS1Z>+`Oa3Q=+Nv
zB8qkAgAU{yKRt#<;^I0g^3^J!my(26`C3I)l;-(^DDjFGF=ak;QQsF!Wf2E^e^k`2
zgAbb0Q5mYrGb+od;0wyN^Evfg9d;p9*FxVXC1vU_&<<&x-F_NfP&_cD+<Yz3kOKBR
z)*q;W2f07X)OTD)Vp!HX-Sv#Vi@wi`w>(mDN`-Yy^UZzlnq;o5*hC#p<gY+h#5lbE
zkt&hu=Gmv97)E&ZZh8VF)dL0uYZcbEDnU`mS`f%8=sPV;kD65!XpzRNVC0sn@BA_+
zSlc-NK`5+hW{oQQMi$_VY)_4su-}ZS51V<xoHf{DC)ej6ut551-4a>WX24PC32;o1
z72Y~;$uL|B=}e~fA4I8s-AHhYFGU&zC&hb)j_MwkjnQ)jp=PMOsApcn{bs1sN@-fk
z1695hQ?r>$cJ#ASPYn%GXlQBP@cv9l?ceOqt$-#)Yfol%N*Tq_<6t09Cy>C2l%{9H
zuwO;YDEA$mO@4Gtl|WA{_RT+Xez?Z`7`0l0oI60DsgF<;YQJOsB0wt;oajVlot(Hf
zSVd~4%*L?{(u-T14RkUpvO>X1%x|KzFbjWX(O38h5v*N1Vza6Al@Tq)LjjxI)N6!V
z!4^=?XX0uxYU(a1#)~D=T$&%OO?`jbSDwJuSIeupUS?^xr-FQ$q~b~;U@}J3Bjt0I
zVh<<@UU#%WOZ|;Ai=6R%3d$rO*eHB9|Gyb&+~TRmR>edtQm%<yD&8aJco+`LOUOx;
z5DvAPC|3-JZFSQZ(lETWWU*Y<k+PFZp0@Mc*2splaG1e)zBQ-E3zYz+!++<#<?74N
z^!)5hVpGwqOnC=JYPs{s$r7Xn4~3R@K`0<z>s+5;%p3$m_%Qr3@4I4Z!ssfb@9kJ{
z!3A#ucEd{RqOIhW3B++%!PcZdn*qUNr$?zK^3~c_HejnM)Fg23{TZU!SJ=H>?3a4~
zT_*X~7t!n9bQx6FmEpay-{)4FM6Ly?Zkzt>dy5#=Z^h?dHd1tO1h|;~3+F9Ef>o{N
zKDr8ie}z5IMHwTW7v-sOS55uT^rZ56SKyUr`BOcwwC2Z8LWhRQa?1l(w>e(7^S(K@
z&(#&WAMz)nK#DXYxVdUmEb7z`Kjz|&;~Jl|&i%^iG)%m9?7RubGp~q1)@l=3<zPsa
z{>n9We)ZzrrJKj+u^5Dm52y{hV93)lEX%GJrX5M<6^I83Cfu@hO~GiPv@`W1T(et#
zaFmQF(|Zb55#0Bw2AFf9ru5ITJSaAm=d9ifz)-v6vr|p;>mVerf~7C_Wp_%%C-WMX
zf{O6L<m;8~8K<gYm#XUR<C$5EYC`4PT8o(#R~GP&b>$7P!FJYy?J_a?BLs}R3FYMh
z!9Pkq$YPAev+`WEzD&pGr9$;8XYlZXi}$664o@Oit$yu39Zh>q$G~vLW~3b^S(@Ee
z{C3K1xAyg<=jn3nJxhm6H>&$i(YP;WL^l*Zji_i0m2&lQ^YADQ?8wX6uMf~$Oji@0
z?k_+0yK|jXu?@Q{%=@_19j|Cszv<L=G;2h1^fdkhl2XE_PrjR@1$U$T1BdaQzYoi^
zhOGO)b|U_WF<EkbmsP8R9I)ILN6EjPoX1L~L`S(_<sR;YAH5&qZ%M@~H7Uutr&Dfx
z1wDL}clFx~cI(c8^HrOM-qe9{bGbD7fVuT?^~|YjK~!k{o>D%Rigs^Zj`*zn#Id|I
zBg&Zy#CAYvO{wJg6EAe#K{<m#oaVa?2914IXI8~?Wb697S8XdB9#`kTrlFg#JITAz
zV;n?uCn&E#l=?6YU?%7CX{oL8$M^u-e!T&RdE!75(FcG?G0D3kYpA<?HwXb<pD!o^
zhzM?Tq=N$-NoLOJpB(ODoV7#Y2JKe147Fr-OC3JBY9)8Hne1=6TdIf)1o^>M^;`DU
zE6Jm*%R<ja5(3B@QJHB=Rn}x{)!Tre_F6Kf>Za{3MTLTXFTN7TF-N6kuq;PvghTy^
zi(R`e2vE@OJ$>-#Mswkfas>eJ{8avyZmnR}kGN~YMG&lO2ZDcr1bwGo+z>JlsJpbN
z*na}{?g0v*UptG4s63*E2LEsdK$=2WdA%QPPQq^>P<M804EpT|Lgd1?2bw}E-eJ-=
z#fa1#AF{sF#Np{?A@3E{r+mhA#}dIc1QJKW^6{ot1(pn;Igj0N0I0@*))I;za4U`r
z^0zyDLSsX_%+S=_6hR-nn`)DBgZTiz5kxn;ai$J14P5Nh+{irlLkZUX`~!llm)o&e
zQAMDo2WANAUVL-sPRxWv%tTcO7He5a4E=j%q?!<8yQ?qccJD$wUU`vpY2@?Obn889
z#S(ySpMlkwpM217Wvdh#@J^%JP{CsMh<7HQ3k`@VVv6@S%IJQStl$SkM*@{$_K7^{
z6OD@RnG;M84DQD0C#-G!+B-QKlb%w_zV$M>Hh}ALLOpD`sQ%^RG~9{CYi8>4#o#W<
zOb($8LCFZAwBa|ciXu$eZB`@cFES16i}8tGCjJkp4Q2-^+xBB}eqq^@2#~e7eI5JK
zqFnP8P{V8P3Qc$|9J&DIrW~G$b-@|h*Pg3T>1M(CevD<-x67MFEVvXDVPFxpZ8^<I
z0$nm!@hkJC4yz<{$E~LQ^xk8G?Yk-I2eS7$9Tp}5)0>Im{%%MO#;hJ`3TDyBR`e*C
z#N-bJT#?g$2|*vl<U>jap2tkMH2)nsZX{@kvCPGDXp}rR`~6G3hvrqEx48Y$3870{
z4xMfQByq-T;Gew?inq3I039kWgGSfG-GPVcU)dWMNc}SAV9GA8VTqwJq2{tIm%UF;
znLw)v$!XS+QJpv6GMZuH!!kmSGf}o#S))p=SsCY@_8t=v2U|k6^hslGx?;K(&gZ=M
z9`a)sV6S_R8qBsfSh4*U9j{><d|5oyLkQFTmQ=<Hi9#RKd!o2NIq~J)CtK4s>_yeC
zVZ&H~7ujRf#$Od@fa0Sy@|IOr)|bh?_>+Srf$sP*%w-_9HB<k*Twb|5b1ejvtUk=x
zv8_wE1#=^L4)5MKzHWqrQADta?NT625gZlg&tDj&1Yd`tVe>~9j6sS_Cxsm>Cvkvs
zdw+rh)!sJgi&{8d%z~`zM4m?0l^jIhg=csuRRgAHs;pyB+e5AEzxM)tTfVp%!yqm3
zRn$o{1UB`^<;SOk_s-V-w)|0B^7=W`gP{F5?I(Jt=wPqs5&_FIU~Ru+%n{OuY-%=f
zv27D2dpSPDSo-Bx8Cc6!RyM6THQ|$K^BYK`^;e1gi<W*pWm3*aom>SP{>9bVu*ISA
zV31%SAs9jPmyN9lOYxW{{f@|>F%@{FmF!euFgtqPJh|&h=kW$lh0XF0zm$w@tZ#6&
zPL*A!{x*WIShQ&T`}ZYWcV?+B#=vFef*weVk$sOqmk1Y{??2zJvDGwSvs(159k;OA
zISa*=X<Y9;mqHJt%*8wkG?!ER=jpQ_u?fw6So8$(3?!iJYm_fhO>Q+8{wOAw{*~f+
zY$H!!?X*3KkRzm=HOYHx{71?rQR8@z){hVb#eJ@pK0;}gX@h3&r}y3g*V`7BlZCnP
zlg%Xo>BqvV`E=T5`x&p&7O>8ER@tR#W&%?68w&|2AF^x8L`*=+H&%mtGTpFAfdG4=
zRvsH@`wTi0d9%;OP~DHVabJ*Key`aPd6gKAymv&xMb>3}U;HEVU~`7O7o+ZSo*x`q
zjtvf}Ys;uqa{e)JJFdh5e}-gaytYheUWO)rR;qEJY!aaUVNa*kNhTx|;_+e-qK1F~
zw8ICyWQd+ws~qRPLeuLDx!~wcgrtsz-#sX^==nXJ^@`iFw3ltwiPBVKdoMJHT~{t<
zrCz@L*d|<(qPxxMsu{DGF@#mpE;g|@>)wm87@a7x#l7<(h};e)xD7DR+fSHVnj7{G
z=w*S80d{Cg{)Z8v*#<@yMrvT~K){NV)F^{{D{r$eo=N=#>%w^nyW6vt7#x6;0}WK#
z-E9FP#SB*>IWi`3J7R@eXt}#O-lzD2clb>QHP@F|T%?uf`Ogq~h#kv&QJE(4T{{J|
zuXYmgWQ%0jlyi<eu{`0JZ7(%~GzIKajr|#L@T2ZiZa1c_-{yHEWHB(wq_NMgsNDPh
zRjU;iF7KNu+b8y#PKTsIcc&+L+ZLmDohlB1fK!SR_r_P^>b#O#*{4-9<Ghn#RLO+-
z?deR6)`<lW3mi#U!)=jlpDdoIc*XkePk$6yoQ;q15QQN4uYJHdEQD-~0=NyRbTKe7
z%ifn&@Tm+1QWj*O=umz`iT}}C4F#FJw<RoBJf~cSl7>UOI_N%T`-3`-4ze84ZB7GC
z<pl#Bn1fr@$G2@_wc9xL*ZQMV9&dx<DV(GNMchMGl#80#7xOjiUz__YU4URRhzKSv
z>5MORL`T0^r`a@K48Zh8)n;eX7b0jS`3~2I9yK2u>y@J|(Rp@<-eY1Y#|bXKa@@}y
zQntNwS)jPeX|#doV0ys%KuC$#j|873x3^x9qC{sm@VtuzU!Ii&Ftt+9Ix1i@CZ<Z0
z0g^9HTBw*jpA<08MuhPKU0G?3S*)t7^_9us9rL0ZQTK+*oPZOOLq;=^E6xwc7R$os
z>^%;cw%T7^QBBWlHr2|lpuo+-XCEFN1ycvZNlDNVE6Ja0$J#2HMy#Dg6?gyyxC~{B
zC*Ka7Ezh`tIB#)aImhIx8y}&fu(FIYfG}&9ZWvsgwM)OCxC*k`j$nBVZW66GN+4&i
z`J)I<67=h#!rsvra>q+fTI0L3k=_q|MB91x+(ymcJVca8s4``p_}Sz}Azd=6E+x(O
zK20D1YQwh9&lm$bcJ<xyC~J9o??w&fI6e^YawaRtHS*VL(-N3PHTNGz40+_AAy3fP
zce^|bo$N^rI^fZOxxdFQ-)?UDb{mI+f<cW{IVi!2)E`J2rWvM=#EMOhlpRbIbkxOv
zySIGzoLZ2}J`fWhDo`80-1zl=lbT_d*0ny0x;B}`(x;dp1bRR2!-(8wYGY)0HK4Po
zv;$LudC=ONDu4d$k!a~Rk6DsZuOiPo$*yN38slC%bwRywa)}b-<lpK?KrGXGK}f0J
zx;1ar&)(x&58UQjYG*J388%#QvaZ4BdA(Hs?2+q55rJcmY9SWCb(Wq<Z_I0au@v{W
zeGX3)SFbZR7vB{??<aoX|KRv+9Z%fRH0z;kQSGJ6y0t5i5$<{Z0`}&~F~Z&A5@M+?
z&5Ribhw38WzK`m;+z9jOwPUSPaoW{d{IU0nzM1-MsPl&}fAls&%zLri2$wFAh9I?O
zZWP-g`zMO}7OAi}F7}+FW*LJt3Ru7xlh&R3z_^bEKvz|KvK_9t#>YwoReq)LI?2k^
z{fI^3OT}Ji5X^*YCkXojsPEX>UJU0b2sA%Y;6I7Sz-q%{IcsLpR?hZEp2pDTOufI*
z<q?M#eYH!=(BL%lkr+5k@TDHleSdtm3PErWhOI+wtSQTzWHgLoo{3sr(JkJ^DXAB<
zx^Tu8o^ZK!zzDbZibC(fJX^-AbN{v~c5l1Cvvyg_{ju4kF@}-c+Pd&pc9=`ITCucX
zkX{=tL+`YIZ`0>0+UcB=1;OY2%0;(Zjsgxw7idFQ+3(O!LYa*M>AHEW*B#2{6L!5K
zdnSXYAE9C{wMJDa%hnicLXksEa`<CXyiUnm-g%wo>UAXX_z#5uu}5Y&6}fh<wOEPj
zUG%k=RDdO<euPu+#FmMjvryFJOq4yZOdh+uRoEQIWbEghGLB{2{T=8wKzeBQL5p<w
z^I<FC^|RJsLV%qH0hIgv_d+I%$*2=Xa=wzdx;GxC!u`z)$_Jxfnt?U<Wfuv&NbYod
z2LT$Lw?!E<*7aeHgOSJLq`^mMEIeRb`D9_#X-gwpeZKDA<cGZS0obxS-sgJmmx4N%
zef3Yhy012zm(4zvLYowq)bp@|jsv{iX;R~jb!$*rMBbUy8y8k8CBUBLUW9E8ss~Bx
z9>$v3zQ~wJ$FBAx4LasEfocjqXI4JZ!rxvM3~2+}gmcZ>k-@t(aiIZQHAXv35O$ve
zX<xK4mCOFSnWH)U?awu>5B3jUt&f-9duwb2(T0)kGiaN=hCZLG1*iB(`CzBG>|av}
z!)j~xc%~G=X^v*<cY(_`Pi}2)07@whRoM2!t97QV@bwk;0kg$;JdJHmcH{D9c7wEs
zUe++MEB6jy6)4$DRExA-A8zW9aJk^}vAW*}JnjL)uPHmHMB&Pn#Fh!e!$M4jiR3eF
zECNu?8j<BrzWv&%N|y3yY!VUd<sHb1_Pz62YcIXVAnd&r^XX8F@3Ymud+pKYfggHX
zS#NizegM({lnLhH0LqXazbx5Rk5*z8>u|!llkKw&w{G=PTp)3yovg=Q$*UY-m~cKd
zqiH)dXZcdvq%ub<87@1Tf2uBBCZj_T83Y6dHj=Bv@KljOnAnHS1u>h*m4Y7(d50KW
z7GSvdkR)<<uy)5gg}$i)v!jdM(R$AB)@1yh_bGQ9XKsra#<|yMZ+Hwn9XjYYKL1G1
zjR(rU{f(J1ak9+EsK80au093mMgcP+(&{!&_{^k5Er=#qo*qbG`&`ssx6Lc|Dl|k^
zww?&Y7P-XmtEp$wqO7=To{r<sOmWMcX7-2&x*snTv4*aZT88;;)7$dJs??BggzY>6
z920Lfdn`b}6%)9c*}|bTY7M*BdzyGsakn!jLV4)D=sj%ANaWTaTN((2wBWu4qe^uC
zZf&T_)z)vK$<}PD;w#aP&>8?|X;)VnxopY9FP^#wIKwY{GtA{fNt~h|bIW@@(W(iW
zx96gVJ!ZR)T0QuL*&sY_E68MsCdhTgevT#w3^}^ONY*GFaL9Urk+tZ$q&yJ<Tc%{}
zS%`kh^V@35X1&ted8rOrW6!+2f69H>>4tPB&|Fmo#62;}<Hn)8rtuv83z}L!KdPoz
z(_4kI!9murl<R}i-rRI7V1e{pbDfUCa>4Vojl5kSC@_I{lBikonAK$VrOCR{JAHAC
z&qS%)o4a1EZQ=0hMa+OP^o7Sh0sA`hJ_97&dzQH&WFMY6U6h!ym#<Y=YOBPS94Mbc
zb)kupx)<Gq9SP8{kFwMppG2lpY*ku_O<`7*Lr-f@Y9CJJy)SK+dg6wC5l}Q3i%_AB
zRXTDUQ(P#VB@x?7wp`>(99#h?sr#NJMP5UD3OR7^F)~T(e`*-Lk2`vTUUV+A`Qh~o
z@*;y5nqI6Zd7HTWn@`T91Rmuo4wu@h1Paw%6v1|wLK3Q`_G^%%#!QYLf#xYZqHP-=
zsjmgQtg^auGJPpHWk!d+wdx0XO?{*N5rsX!f>ov*O3c1{Nad9bS|@Eix#X{Q3Ydyz
zd%^evr=rwg{K>oopcVX<uHAYY9usk=niR8TeCZ4DXOAq_N%PBp+z>)(agFDOedW#n
z0oIoJE{8<mZP6ioXg7L=REB#q+pqGoO+)4Bz(;RTyddLNsZIs&RAs(=NQUFJslQ;f
z!lz-2Iy*N?ypXi+hwx7YjI?*VS5Ef;3X~2MYy%s0t4mox)%mHzWeI89ZxVIveq167
z0X~ioM=KZUs8X8p!``J3bR$h7XlQqs8?|xl^lZ4By}%BOmZmxz@t%4M8&djtS)(63
zMszXEfExvwh|+Si_R-KM<3W#cg)Re6diS>yIBN7BVp~rLM~O%CkI`JqtwVA~^JOLH
zg>^)l$#22G>Rh$3_!&#7Q1xp6_&MOC;BiWOPcJ4u7_t|fgs9ZU-5!(*MR0iPLyOgP
z*e~-+9nIqRW^B_7>h~S|6zh!4`^tQopJ`$bUYAu32$&V=^MLAwNu|1SmZLvS=fP<}
zxX9p2y*~%mRe*Wmb?$5U&Jf*mPKuH+!3>~DoeRY52aui9i+9gOD5A@QYLF+;ZTsqU
zLL9WU#Rhnxq6uTev@jXNpLmdI#_O|)u-B!N=sYSi#v58t4^5B7OYuc+u?D^Ycy2IC
zDV1*Q9yrzB3-;)nj@-l@bJ!H70NK`+_Cs<!uJ`$xa#8tL1PhuiJE~_heCzP(1J%ee
zA;<z%z^rPHP~FyVi?4)!9n}+zVO!A<msse@B*4WGg`}|d-UqtA)|cfl4}*+2nT1Z(
zl6EbD>+^m8dMzZ0-qt0H!gBx>)VRrupWnV*U@U)0>Twy}^~8(CQeS`9wY&1Fa}f3A
z>F2yOjc+~7<?~4b*Ze_KYN8mfix{0No3f<^qW<5>XcWECREa4lcC-m=emnV|mm{%|
zJK`A6X*G^xtc17aI?coiGX<cRP8H#%mwyeapL~}SxO2;9!E*`#Zk4yk0u%5*)a{iY
z7G>GzgitQE^kC+oJ{_ppEpm6mFYIz=kQ`4Tw`MyR<=c6$dFoz}=#5iwy1xQz-@+Zu
zJu<mYJbxU4qseg;7l7^uO#>42PE5YXnP15lo>z=zGN9x+FZp&n;MInR(}<nV#uj{e
zd4z#upkBq@%2jPm_CwGiYqlX<xwv_d(x<02*U<lUiW&s8q@#c6*=Qi2E~w<~-_FwU
zTDHt&(*e?%q3*h?PKZv){hc7soC7$0byfKJaH!R_&vQwtG_v5-51_V*q4}->u?atk
zquOC4UR_RcP5bodY2>7l`O5JSy?KQ+{7bN9!UC=(F}x!9JY6V#s7BvNl@0Av|7G~m
zc(^xt8xQ@P+Ht@k&X}zvek(hjnPHNp@Kn@2YJ0dS@K8bV$%2^{?qYLL&T3sZLE5n&
zxz4jhH!MtSIGNb7gv|HlFh#T(N{?#F*=zR~)t*Y41q?gcaAK0-Zd;Z78dPrLlBTkJ
zyJOW$P^S(H8QPq?d+ceP>_~bR)N^mhs;xxx6=6~Vfb4YxKK-Z;spH#7uvW;XF5Q(^
zb96B36D|W-EbqLE;}Carnzpv^PsdW8NoGdoJSICQ(guexzDyJjFQeSrcePB#*@vO~
z8p1^0UXIWrCU}^p5p-!myCb?CzJcyu4%Y~nZGV~fK-c^NDN)cKw}kQV*^0OMhWWr(
z6x-RFN(={}>UvG3==IdF9ox)0VUa+Y!YSJM^-s-iM|^_t3MT3eDA=+DJ;S=Tw}W-e
zCfH;>s|ZBKlJ2~S+=(Sv^<K`e$AeoXfG5DOGfEgAuRW4w39xSz3-}rE+XlF@^o2_O
zQn*{|$65Wr^l1CrLVC)H8(}-z=O>qykE9RWTRtLJQAtT?+8VFSqHe;lkf5Q)c0-q~
zwUzK>eY^2fZCYf6a{(}RBo89!<j^wx%)?wGLBFZf4RAM%eicn<)&=TEt@Fm!;&#Yk
zox~dmZcccO9gHbHSZ!WiMF6_>8$g^`+T2y4$`^K!@~z{CTx73?v)^HHjKnJUU21vJ
z<9v2`8#!HXlrV*Asp3&6NgjgLWb$?Un|}1m#HSY1rC1n50WtS|4Mzx_$~SW@3fED#
z_>zAj9T3Bxv2%c8j{|wew<8F9QMp!PD3%=dz$yArqDFJ`dkYauCIB%%@v}}4mFch_
zamu2l8h9oBw3+@?zgf^yyK$9ofhTfN1HbkT!j6&1+<$vy!o79Dt3gafEpI|4N#3JX
zX()pu!8@rKiqXeVHxX=qxv=p7cU&j;`w6CR6o)mt1L;N9xs|K!&Z~^sB3Bu-L&47`
z{Kzn{F_5?gyHUsb?@Eh?iLOu;+qFZwe@G0au$i@KDda1(^DbZLeKprwwkVnau_?!;
zC2bphr7N@FN7r4;>j;=TTxgG%T3ljToS5(<vR5da;+Y>~PO5miI?)kI!dqkHx-h}R
zjlPycsu1>knToFuNK>~XPyAk|CJ85fmg%2vT?}3ZCMA7<B<rx%N@(~pqhKVX;QUHw
zPvawOGv?e=9SVn(PsgJ?P8Jv}D3#GbR<4(L^Q}y#T}vT3zMWTSyMIw##kSU}+^gL=
zb1U(481nSD<l)AA(RxiID_;ej3vLiFc004&cB?QWfyJQOmi#(1cG0K))L%C4MG)n?
zzO39s=qG|3g)4r}^-oDcm$Dq(zK<hKxEEcVuGt2N^5v|AMCGsJhodd=>+jQ8sGdB)
zx;!CkvBU--Gn<wdPvv<q!G;8kbA%8FDa;s{`z+wlrmj3tVEOV|a3`qb)QDm<%77?W
z#7x;kRj|BVzxnIyCn?lI>S@I<Yt)u43K0Ntj&sw8?ZPf%5gir2S|P_<>PODEBR@Dc
zr4x2aQ{ZTrn)6Z4X`7aO(HJuf#_w<+dYX#{GH8^T9|9fZXmal2R#%<o0i@%o*fA`o
zmEv}j;9#QEUz0_+UMf>ctU7mUC4V)>U0G%ZoEKSJi;l2<1Wnoa`4C|Ch4S2cQH^sb
zuKg#9Z<WLw%|+8PNvkQs?PzGe>HFWz?R@`m0UEM>&J2KQsIw4^-YRppDL^~Vc4y)3
zAvw?8;4JKCK+>4I!q7Yc6pSQht4fBk9G-Px+V($k+Pb=9T{kklo!DD{@$$^Kf-|fB
z;Gsk?t8{}7P8+a;)kymV!{}St460c!dfx9L##&uw`R>juVP;!GNv9Ro>5ws7A=but
zU?wPl3??kvo)G%2t9+i#p<*9PoIs`JPVCuW|IX+ZVAFXfln?1y#QegNHXB@IxzP7=
z;Br&h`QnsTCs}hK+EsSuA@Znyq+aAqo2=Z1VF1rtRA`6=kWYUsC0C6l(z$3cOAIh`
zfor{&&+EKhohY&gd#gBGWc@{Habl+?r<LdmR6mG5O}0)+@$Fji$XqGJGtfA#@Ynu?
zVIrOZYRkU~>J;6xI4i5HNhc4#S8#VfAV1(r;vo}7mO;09q6qkC&wyxWv%A<ZuoRVu
zS{)n^5(Y48h40RKC(wLIn7Nt&=XVaTAG}5U>JACs5ZW3kO6nHc1?wMUmdF6)4zSA%
zo@`Qaee?-g3;R2@X4cqesL6%<BeyKUjW-2M=dIS)Uua379;DjN<#%2t*4?5aU%XQ;
zDJE9AL*NFw3;=<Z5^-NptMj}IPTGB-zgfeA6TK6vi%R!<SH|jkbVa8BXX|W4z#FS*
zAL-3omQC}ncYOC?ty+E);b}_Yfg~d(gv`QkBsT1Uq0ri6-3KjZGQZk8XGgHDv$CY&
zlQlg>CF!5Mqrt<L6g!Ra@t`@^^swqUu=9Mzvq|zwejH4SIco4Gg>RrNZDG&x$J_Dl
zFKfJJ<zVY~4X=^8ceoO6?R7#36fn1Yr+wXyuf40fgd2b~MLT%d{wlQvD}4aeqvC=n
z8@Mc%z#As4C1casLrM~mA9mouI?arftz8}loQ`jZO$S)(J_E2CvsRGL)=PR2f3nF*
zr0H7;F*S{eS2-L9p>73oQ}Osu;zzkLZzH``kmLP0)kzf#ktSZI)C7zN&$j1t)W=?}
zAoNB6r%@qR6e_5MdT}Ai35Nc}xv&KuEO8b3n4r4!18#awWbbT<Rrq>zip0+<z{atf
zhgIzy@ZL<?doe!dky?x*nqgf}cMt4TaK`RP{S$a#iLe?wcb@a-TNnX7k+SL;_~FhR
zR(iOjG{Fg~Pf{X{*-7}kB1)PA;0ViktdD{%7?p9A!E6ip=miBFo-9W9o6UJ*(lyuh
z*9NM;Bt)o+6L`htE3pgJN1;EZW8l?T#7-3}dNx4#GE-JQ(O=3bGoBe8<w0tmLT!6r
z{w@K?aPM2DcwwY$pAxa~dooX_rotnEV^SzItQDX|nUfEI@!-KOtO2+zU!@Q_j+X2+
zPdi(8t;Rjd<=J7v2?d9zrz+;$$yPf$R)W<|{@Rxog@7O^Bh5vQJ{Id;vB-9Xoc>v0
zhxrSuzOYu+=KzXrvga2c<a7$L+FvC<-#SSqg%t!Q*&6+#j2inQ`2fQ~D7*LtKa!5l
z#r{&K57y;Zy+P<=Ml@_MKq)`gway1%(hFw#ucbU72C>vgj5xmuQ!PQM7$>+Z1RAX$
zaBFq!5}0=wW=m|ds3$vGgH9L9Qc}%Y&#A;t`jXi)!%>sN(Q*e3S}YLmIx<8obhUV<
z?;D6Iu}*I~%@N+wXsRw#(TfgjdxEZ^L6NZhAaP;bNJgVK#Ou$iS9t(Z_U+)V0QAuL
zl~4X%rI>gh?+Omjt`STBlV$>)Uz)w3!{{GyZx&cWQ^PYuv_7CDiZ-Ztm#4kQ5Pd;2
zqjUJ#p-20PKBnsd`(mM6bHay~L>$MU=z|omb4;L@qTc@e^CBQR;JkAx6nMf^>$HO2
zG{Wux^ag29C4vb0xB+OtFr!?PXzcK8HtqgR<GBLA@T`($m8U50a?VK-hk_^!*qn(U
zTaXqsoP);e_^IDXWu4nKcJ_&`=^ML_%XL6b7xLeAq(+z%3=ngVVvHbkQ1JO8RjBo&
z6}DMCJDpAkpYCVU4Z6rP$fZIDOV`q?i~HiM674asYBUHc3?oZ3^9ll;B6`(|Noimo
zO#C*hYeKEV#dkCnUPHx~-~c20Q2A8}B^1A2fsleJw8?3uy@vx!Z)VApVEqo9OZlc{
z4g5oDLTl{m@Qc3Od8O(d*eoaw9ba0(4pz2q{SXdY%*=8`m?1rKE)JqyXafv@BZ9io
z_YEq*GfFL(Pez3$paFYvRx>P_{Bx?sw;IFxsPzFVBA`i2B}MAJWtrAz)vV7oH@up8
zuGUH$R#&=ndIuPw%{$U-lzcdWNLgs<6zTap+^trX6LCHv&OJ9FX28DQ?JNWC;!SxP
z6`ZA#vWKAE{OZXW)kH-$M>O+-%2M}7t)QyL8wH6sN;jRZzpYAGFR1g@_$(MUgcTHm
zUJ()ztvFr%9{pw?_pm}e=q)YLWlMX|9m@~KXlN`dUzKKotK2XTbzi<^O<h~i5fp&)
z>~gLtxyJ!pvBfwI&+w%iY*EC1w-cx2yaxmWDsoF)<DQmbqlF+Io@CHtcQ*dTXfm9^
zx4&|_I*~*fcHsjsFfOOpl#q_hO6`}42|7?~0^pyPO5h^AbELhqeX^k5HZ=4Io_|Z|
zw)*GB0LQ;BcCp^WZCdnnaF0uTyvR>`Z$f2umg=^gze$su`cG!)CH)58XRlq0X!DH@
zkjBdNK0sB=wAzh;h+XJ|={;KA)@Z)w5Myz=IeFyzOIr*-hHlBcc8bPT9g4PVrAqAQ
zu0P`DU3p66h(~kef;2*bL!6Ax0{r0#ZqZ_Ox@P>sAH5zhzM-hQruY_5V+o{Z{a@{Q
z)@3>}cP;$_Bs6}QJiWKxj~7jDcc;f9rO*?hJ)rWLLmShZQqqs-fVM~;BCwi-0!Js}
zgVWeqX@k9{xT}F3&GdXnC4I?M(;NSsYP`ZlVi7H$LK{3AG$h;Pa{yq8gQR8Z*{E4o
zAkUsY{pHmAuI&Ae+JRF$T*^hz`~WTD#+{WN(DK%woe2P1IX#3;P_!o|t!r6ZRd8&-
zQ};vG&%#p0kvC%bTj+7(4Ool$cziQkYtUo1QZr4BiFif|E>9<lM_PigW7W)aOG=B+
z$Xe39x(}Ms9xn%{=ryR~0!ny3O5N#|3PP`i01k&EZRm0H+-iM%v~(h_t^s*Mb$<nC
zspIO^(l|(7K8(4-J8oa+mLJ^6o)9XQ7}4I3C0yAWm{H8ys?4enp_6r+=089p+19IR
zdagSN=)=dvF<D4RU)}cxZ=nW*YdGT@fNRehx8WV&3ddsUk@9#`Ykh(j_i?`IYSHUd
z*`i-sPk=xR7^~oO+L}lLLyoDXm=f1N;}7H?3`n@yN@DK2!(Y%5s>cHr_W+m3a~a9L
zT${Wc^$mhGGjwi|KV%i}bo7ghN!xC%E2Njnb>1Hwud%mvvN@76b*}p<0KZ_;@;v!c
zM0ckcyO7(eGTo)`@!!9}8R<PXz%F(+a&mh>j@3huR~+<w$nn7lPIu3=R^L#4ETnaT
zhdgc)BQZ#KJ`7y}f!bQ1$enHNY=7fT<+4mUi${McGh^}IioS9{hh0TnynPckvv6=e
z>Vv}x#KMn5Gv^)+66ZGWMXusN>lmAwn_rmXv;6f<@R4Vji$aXzm#JV==%3FMl}FKp
zSH(iq2tbFk;a9@90McP`!LL34L<Ok^XXKaYd~!Zk?z4J8T85wpVNp?)mr-8qpaI^F
z!s&c*NAh-x)Q<aKMi7W3)N}6;e`V#69HFYcp%neg|D=E?U-JC#Mg^n<|Mt%t75wDJ
zrSjju`Oi<mHQ4|4yAl8I8nE#GFOA!UmzyyX`Ec3$`OxRybLvd;sgE;NZPjz-r>3XL
zeQWSDDm83liQo(D?-%4VD)H$$G_JFpW(>KdjhZ-r-n&-Y%v0fZ+7W_7opuDS%b)+-
ztM}<K_D0FYaXy}sOn%}<9k|f`(PqBsOkiQgj}Z?Xwz3zpviEa~Ld1i;u5Oo3JCP*I
z-s9x=Q>&Rejv@mI;&7&qa$azkQ*{bqPX=kc*2|^M&naM`{n|gChMLNkRcfu)?(G6Y
z_`Q1ep979HhK2<UFb@$r>>Xi;tc%x7UzfV#x%#Y5ez?=WO0J6#Ltl7j=;UJgNB-r4
zTvunk+;b_32hIBGEEc%5>X|GI|2SIUaoPR8v35S(epyla6H{`E;#BRmDvcN#N0Gi5
z&*kNG4Sauh`Ys+n>|WqQh`tTS`mWpEuuHbC>oK-ww@2lhi7T6!8mo{R_TQw~ZR$<%
zZK$a#+ge%6b`3VY>m%&IjlFB$9^<VCujG)~fhuTv`9laf><$uIPyb+ytUp6^#_OLK
zp6ugA*M>h6cgVhJRK>|@jaOY~5Ohb6ZxkJ<SVg0<RQ#+KxDCiToUYxcUHlo^=eW#q
zWW!8)XEW9;<(*ARp5bNRZX4aL{p~^zOk*P-stv?l30gGi?|GanyO!sf9;}a8i!jY7
z#&iX(dl+8#mvF<*Fit&os1oWa2kyq0k4ZjI;-Jn9TEZx0Q5Ab*GWDtr+k#DvClh1c
z>fj8qu;+?VLaOpa^Ive6YS4<neX4A(KbCM?>9Qp3?6%dzmxj%Cq@$k{)i1tEsbQ9@
z4vA@vbd?!sp%fPk>1#EIjV>}`TbiGR*={wVPt&V?LgCYzoOsIEtF7#J>)xhG0*t)K
zP5TFIG3<@CfGWd%Q`*_gSCH}^#m1ddzxsf(7IPh)_33=&2xAJUL8j&pJH@tRYlvR6
zr{;MfARk@;eYRlA9lXM44LEBjCa0yspO`!?1{IFc98|ZseqL4Cg$zrtZMAx3=t)|u
zaek?xySBDP*ipqla>~n)`YgW^zH3x%H7Sfwuf0VALpSZxm@Hq7%bh8G)pl&Ich;fp
zgB5Pq%5a&fw*A`d2U^cFx(rG|HzMKS(XEL0ITc+4KWm51n{Moebw!Y#r!FS=Rr5up
z?7JHa2G$~bEG=c?iCDe(ruZ$pKCB^Fqfr0uo|&3lCqY+U%5zcvg9ka+wXow`f%sjD
z*JRo{KNdxX8#KR3d#w-W$GFE{kqKGbthBC7dq$HaQ!p&P7dnW#OrJlo9;5A>XU!h4
z@Gi~PkdPs|YTQejCzWn!1saK(lUv2-DgL)m&QkneVm@0=a}kSC@c116WGV*vvN8QD
z_j1MT+ukpOg6pv-kgsnz;E}FPxM?s@A4U>V&ReZ0yEcEH{qCs-f!kBmlCsjw<t5b`
zMzY3F$v&UZsr96N|J@;`j(k(92zeYu&!r^RB5L~XE6iuV4h&y$9OV#166i1v*k8E(
zF5Mb)dUh>r+oQgHn~0^pN0T0H@`DPs%#swhn}cy<#@W{U@Rq|6hDkVu3hrk{w>ze@
z*}h6otBnSQ<Exzgqzl#A#e6r&h+g3v6^vXpdi#hgpV-VlYU!IQ@i}+CzVho<|MB#g
zLYS7*D0Nmff)rnn#BZ>1arLmSi>Qk&k1T4Cz{m99I%3h7{pSI3!BYEFe5Q7^epZBJ
z*ZW?b)%?yMgc?h~12I1_yv6&b^rdjACq}zBu&--@w<geO*XR>eqq_H?VK4X38dyQO
z=Mx6s3_5Z%Y%LXDAiXsB9+xTFKK0mbk0B{i8-yJfd&4N@qfVKq4&1lhhlIt3+b`vi
zg7I0ZmpVUdPw?l@`6rJ?tmAP#Ls_5wmVf1O=6hs!)Mi<=-)s3z?SHlR)j@G~&zc0+
zAOV8ATOhc*lMoyNgM=iwyEC{GT!TX(f(Cc@;O_1WFu(vo_T{tRcXzjL?cZCsikdnF
z%sbOZx=%k(OKy3QKyLN1py5t%Df_qm))9j%dBU~~zjHgqOD%8a#21m1Y(}%rwnPI~
ziL7*1{E!J#mZX6fUw#?JnSOa;x$FJ2SdHg~U1gSYU$V@ox{mE_GYHVN)Os%q5`PV|
zDhJe6+(w^Ik-8Pg4YvD6T_i&_-}nUL!b3&xDXrgm%H6rgVhvnJu1j0aSFX5OA(D~Y
zOjp@pfl!QpWeAu>lZiPP{{Ae~tpgl;Wsl#NpH;fw?wlB%t&VZ`C-T)-_TMkndOih8
z1CGx3)m`^Pr~7j^cHW1$$@BZ2d3ezQ=K|gNciJhb5!@40XVku@uUM3h_&N*ZGe{WJ
z%sEEmG-euR5D6eSF)xgole0R_6l=Ao3XPA4DlK+!iFsYpUQ3wJMbW9Iw;5%T#>kV-
z7Pg3Zo*5`j?@^=H9&Cu}NSCh_ih1tm{oEIPtvzyBCg|DvOPAS-50a9;g<-+87;)h@
zP%-iR{qbz6C7r)Hg_uLQOeBe3`V<TO>3UCM8D3W$aU@^gmeqsO&bHQM{=*T!qO&@v
z*`f^pfjKmhIbHw0-a{wkwJzQ8L1OYeLF5(b(WMWgV$`nMl&=3r=Fz}?QW``60b~R0
zLOfBH+6ShNb;H)NH~i2=iA=AV`-E}7A#dXKDT(!-%R~$vI}7}F7_kPL6xl&FC3Agc
zk%mbO5@@#EWGfUhfqr(CeAl9U_V7vZIf3pdIH4kp+vw`nYX>}OD{O$iHB3_&ahvn~
z%dZE}cGO-T#xtN%dC^zuTxBSWQQO3b{vWCBz0WuiZ+fn!3w%T49>P0fOSF>V$>DLG
zlqcTOYmIuU>M^VosnGb$&rAz`p9!0br#EF`2GONNyc?WnOMOPI&CgEs)d}>urnA7Y
zB*5`*8pLafeQ8MrtWR8rmhfO3$3?-LWyl3Y;l<!5hBQr9X#TI>{34DNp`?gZk&9Ge
z<MVFsC#D{Y0$iK@6Fcw_{;1Vd4K5urF%QEV&yghOl9p*rUZSEeL;7E`qmLIBDGdtB
z+h?0TjB!qwZ9i^K?x?G+h#f5iC^+BjL)T^@iLIf)3d}d^NT5b@$*`fQJu}p0u)j*j
zS@g6A_5)aW!}1!i8fgGIvK?9Ejes$_WK6y1SUNfU2A?__8<MB7RsOiFcu_L-QmNXG
zTBW0$(Kk7aUU9GpEn3}EV0+FGvtCYyBT|jl9toD#8{<@C=-8ylK5P7zGGmaFE#Rb=
z?nZjRID>@|S@ryW%s@yt)WiXssf)b!r{#trvA9*W@mQtH2v@a1S^Ai28HTbz$Cw$L
zLD`*T-|r0mmNV0>*W>kiADE43venyz^nY|ymFSlg-(z8rvOQ&ZYMUQPmERs|zrRS}
zM}6G*G;`95gnIGn_#407?H1a)h%QeeNBV*$YL7wA!Z?2z2!EqJ4;3rex%_Ir@y7Hj
z*CX(Y??z1LOB)#!KiTW@oIw7ETBGIwOwrIM@X3C(QDfd~ra=OwX}oQkI5?3!K}IAl
z$V0u8;jz>&{wUjLD-Bpr;S`U-{?@W$7|A;VIV~sfQwN=Lb$RiSQ_;}-_+JQL24V_n
z>ECssl30YGa#>Ly!?|u{`xSmX$h9<&rr?d3#?#}gu3c>Ttq{GW|2+0n>o21uzGjA0
zspTEEB|2!b9dEej=WPDNxYMeczU^S?DV*4PT8lh)U&k%1TyYBpeR1mJRQchF8BU}2
zt7fw)A`O`=t~$=d&d3<v(6OPNlAIgY%XJrv$$tHvPP=&e@!hI-yiF=4=G)kDV@6PX
zxD6*-#+8@VBlFVkTQ@c(bCzL%&C<xn84}%<{1I(0hQD!F|B+wgSM!3%ZWAt^J<tZh
z8AW1~c`??%ULth&`;u#F!Ie-hx;7v>tEl7JoAVy?8N1~KVHo8%`b9#pEq;h(ogjJ-
z^(J=rUR1ba=PmVk#RGnSCSP#KSK>%A_rYO|&9-OQs_+aT=^T9$WO%$?7=_rWMlg9E
zftEQVrsa{HJv7mV^-gFyDG#Y~-)tLxDY+$AVCWd(%Vw>EAxVIM`yzb36C`eGU4m&G
z83CnZk0r5>mB6?E1=@xWCibDAca1ju@p3~$tJphm=(@MD`|o9GP-3oK2yUDA8iFmV
z>QEv_@SX^&OYiy!YOFzn?UxV)k}P<b*Sd%Nqa8c(Kug7NCJLn1#(gL1XTYyl?GRg{
zu-N^rc5)+Hp=5U~&-_c4$VJ>Uz-Hb$7^V*pBua$V)@m+g?IXqDbLEl1o!s+9^({AN
z?BDv=qdGB@bgakO2AlKcgX$bL_G?Hv2ESZRFNtuvX|hA}HDCNr*h_6gY`7+~4x9cw
zD6bfFvUb~c{L(yK2yQ6<=O2qoDCkHR-NK!B7xl?Qx?DeoP2`eOgtihB7SAt_%=@Oy
z;Ni=VUX32s<uu8*(s!)YF?ma}X@M#^mv|oLgd5z)wT*Hz6*bX|U4@DTiwL9DAwM2x
zBF4+{e#9XpAr70c4=QxU##w6%@ApB$ITYev)vaF(*E&NJXs)Z~=|&xVRlZlV5=@EI
zFCSScJ_-8xpyNVlG8gCzqSAds^<tjM)ikJ))ZF-oy!5Vz1d=7NtjSu7fq1*y&%Rz)
zTkkEjx=~tAJ~<n~WeEM`O++y6moc=@ByDZC;wI&*-y&OXPdeWmcLvVg{1)Uem1;8Q
z8JkHf|2k`?dv3;v{*zNmDaP5-cQofr6lAy{o^<{onvcVYZy$naq;&0pl4#eVeS^S7
zw+`!xGC0OyXP+H;&j9VTac9+ncap}s!69X==)~0t&cpHjsQK3%M4=XL`$KP!`}IO&
zu^39mqMsJJhN{8yobIg;KBo=uU#Cc6B;`{3usU?qk$nEhBfU%Gi50B>gTNL;AxHn%
z7B5jko@@Rxnl-)D#+P0HCM1k|=(LVMXpJ_p>BtXeeMH^hdyqxcV@4!h-XkGHDsnfy
zEhO^{GyA8E)8RycQeF?H0QPsZ^PFipIX|Y+rR|+zQU{>6r!xw|*4|>EKX7LPQO9~L
zB_pJ#WU^YlCD>)k8!v((YUrkIEs<uEns_`J12aBJwopk0Ht~Iq`CacV!w!ex^PqSe
zLje^SrtDhM_=RVh6oLa;$K2OiGCEnP*@SnkKcbM-7It>iDr3~Iw*;^@L?J?~Y8HGh
z8*#Hg#PF_ZIFc)-ruk4Q!%}HuaT10}+^XQTsiyMOV?FeBW{O4U!tZOLUq<zNaN>sd
z4%|8x^sZL$<SK-v%ejl}uqOW3n5v!igZ#Gz$g|^S6o2}?CvlJRtl;cT+Vo*wbHUt5
z&M?8YsD|zdaT+#1By&s~N82tJ<vj7v%Eyphew5{&n;+C1KA)6o_m0GWX}T~rHlf{@
zhB0$H#5{YAIQAAbTplcB1;5nyAh%L=XEs^Qhl0W|%7rcNaVh1r5&iKGNq7hDLREi;
zfd)3i>D9YX`9wWe-&9{qTvz)kd%*6SNmxEL#)gSNk>#b@7gy4nZX2<);!6XI^a=jZ
zg3D$v3<pMdo4Q1J+lrn-sqn7g0X)X<N#KzF&uo8QaEy(lWer7MmJxm_SR&Mg;(EY{
zKH^)wJrinR#GY#QwK=_O5ZHtjH%RH>Y9;w)xsZMhdt)qmQwu6T_lV>|(?Rk1>E(kP
z8xe;=eQvG^Q+h~%d7mD?V>@ta{H%qG4TB3cJWF;+>Q^t(o!v7xx*|-_*=w7$-}o%J
zZAQIa66X6+5E9mhORJEKS)wxKJv>L{&G9kHJl*i*s^5v7={c8_JE)6j+0z;p^3F31
zUZU80;C*<Z`9<k?UC#^kY86RA*PD34Md=GqVEL9^@Y$reS{%i9rf9rRCJ$T^7$%RY
zhA2EjKM#qpUW#ko{!uE#yB-lZf202Lq_TR{w$VPLah{0SZV7Mc=~ZjaS(YDKoI8K=
zl5TTpzNefhtDmN7H~2UgW%K&fP1<;)HBKlv)A=WRWQf&Q{B+!3%MHGQ%zB$BqEB;o
z1gi>ohuxmo_#mIYrSs!!k70u>Hz$BlUSBVuh;jmsB{bee_%0|p!)P$M#e_WopOZ+q
zBN&5G5wACiXNB&fwwraDU4gyLr+wEyz%?}?t}L6%kS*HMU%pxIXgt030akcmxhSED
zfZla!VWb>QI!k~IE$7Add}Bp>@|0=)c<?Ux<d@>=1CZ%=&GX5h0A7y3ZIH}b+CU1Q
zq}u|gGP)aHWK~XDm-uXQ^+MKND2Qu*a_xxTG=1~F)Z9G_*D@6SyPVh6RoeC#G4hiQ
z2qzrXB<8daPtZ>nTWwir-ze2#uZq#Jo!v~YYPNEtSi)oi7`Z(HO7)`5Y$0~ac?=$!
zQ?}G(^hv(O<lY#|q3Us=r6bz>xuEn80Ron5^4V|LXP5}zKSk-j?iqgj;pQ33CQa|J
z;?kJk{55f%k_1Yd8n`x_Gi!FJ`6xP3fz+cFW;S<U#WPr-5Mf7chG9^C2N{<BWYpo5
z4920!3JkBhQNQb%d*R=)07sO5B7dTjHMpE;&XHu+hYf1YTgE@Kp2qT#G>sQSDos}u
zc}w$h9UI~(W7vD&tYT%G{8qLOwD{4^9aY~`=_nMD(H%p)WOVe52whI)ZJrYzNsX!1
zLU91P7SCXPXB`82T*qv5$kFP0cle&&Q$EC!YBFc&Tb|UN?;zIlmUgFQt1dF$mvmvI
z8iqh7tC{4s^UBpi3%=9a&ezK~^<y|5I;{gEZ=0TKopBkxv(n%>kp_#RdX%+ut86@x
zMwVJP;grv=NN%gkbrgAPY(?uWW^zd)=(!JHt5<cj)0yAG31nXx&H6)K#VMbtgq>98
z_hl5OkWIfo#E!A)CY!T=b&D<Namf?*&OMJ>=E&yCwvYj#!0W)Wf%lcf@Y|Jgsy2AD
zA~#0G43-jK0=cLsraO_bBdo(OYOMvZWyCv<jE`!kPdrjN-zj0WarJ@ZQORq<+xdiR
zcuX@_=lEL!N8LAbUVL6;?Xx_sF$nDa6TqF7gWJ=O%@Ub7=TLqjxUD%PpZX9*u3#-}
zH7dW>q$Zko#kx&6%g?1_@-&=)wqe*Q=K30^CbpwYQr1%MU}4n(V`~rk+_H^xno{mw
z1bvhwZ{^lg0Q;&V49moJ&fBmhnZKoFpO;^Zzo(PiQ!`^uSDzI8EQ;}!RZ{%RZbuYR
zDv<!h|4fZy8!qCiY_^Eq54MbXMu{=&Bcbv>{=?lPvjbv$bjua9`0XLRZ+*vM_sB4^
z6Q#tGm5gQRD*PtwPsH1r#(r;}y=3WQYDhY%+B-fUH5utsQ5yMinhH5?S0}xf^(n1B
z>o%g09igW#pmLZ{9zMd?yR-VzVQ9QRD^R%1Em-Xxr6N@#u0O%u`cy^(tR$}0%<B2?
z=&UbZ8MG@DSkxYUT=GbIUbIiFn?Kse2A;xsL~Sq3A;Ao8DgJ`b5#``=kHZM1qlvuI
zO>X5HNF&Pf?y=6gOwQ$^lSik^RecUqfo;u#X#?xO|0H1#U_mG;WUT;siDM0q(^CBU
z;l2}v-7<kqF~B%G(&(>lpx_su;Vk=&m1lRv%IJ#BG}%)lCgJ!1;B*GaI6n3??V<Ry
z_0Yd@`MSr>%=~1U%Xt4n_WuO&{SSZ&L~hMorm=tJH~-P<>d#qRd~Bse`jo{k{EuSS
z|N4tU+XY~T8kbWmdjCAF|9-Uv+FveVp(NG*S@8c~ECb!DC!pFuc0;qKe-SZ{l-plF
z{6OWp<B-4s{Xfgq{{8DeAYoD9$dvMbJp8+Y^?!u(Cj~N@-8v<=*#Amx{+pWs{-uuM
z@n%WNeVO`ye*ecd)QxL7XUYFiB>xNyVga!aqKUo!EWrO?zW-yl>?Qj*9?Jg~_|I|a
zBV>Hh6D7*<ADR9V&i=8fx2-^N6a7EZ{xkTWW&7Xv<o_(&KU?vy!T;lF+eWlqX|Z|!
zDG2S^yo(4n!xJ6aa>yHZk9(S^;iX?mcRv6!vVD64bkv5eKckSE^1>S8n){N5Ld`~e
zIXK4P^fukV8#=j7p%a#@qAq*>^V*%lTpT=@K|6a}$Y~;`5J6*SBbJ5oNbgISKem@6
zkq`Evd4hsh=e!pWGE+`_SqSRYr?_3~MDe)1Ep&I6i9cEExE7xI4ZB*D&y$U&S?dVC
zE3onzNGV%sU{rk8i*b(M``RcGMk%IHqf#QirLNuDXfwj4m_?WCoxm2CJtDrtOYZxO
zkg>^S2b}XjkMThD$PL|hGR8AWW{&viHIB5N!BEdu<BmoQujB3LUVm48zWa!J0UxXs
zgt~Tg4Bffy2%R&P$r7!(#xvQkejZNIxxIlP#91e@+~UP3CRTPqj>q@?D1=a7NMwD@
zLhyFKg*|ki3-rz`-a1YkC^gK1ks-X_fkKf%(Z9VMipCgUcyE;vi`|ca;58Gv4Kq`C
zxlygSstokIxi5_3fj}eaS5?4bep}jU)$R$Emmb$j0DHi0$o_4w)o#UyDZhw4MI1Ye
z5TMX5pheY6N=pOX?@#K2D{TjT#Z7rU%=I7HC2>DeA(ObSXaZ@HDW+4!4K#%I5#GlB
zbiF%z+e820o^zywt?05cfKKSyzxZX;LJiDj3R4{I{8BamFAj_#RVk0b>)m<t7hBEy
zP&4|jjh;HaH2Fo6CfZPSDC{WLqhdrI$#{Y3irs*j&H#YnpDMf05Py?Rx9nPjTWm%g
z^eSDyWQus{W#6e-Bq(?8)6)Uh%=u=l1#HC!uGX}Mh@BRn4RipESHIJ*5AoSN_GyK`
z8A&oFV|e|Uebi>HVr#_oWu?wD@y;FSS#?xqrKSfl<P(LBWwB(R+VC8_C=+2!!;3+v
zZcF3PO1p6etCi5+XhB|A4Dp?lOZzrdNF!&XHl*g=TO;w-GL+aXr9IE)0)_kU<E<J?
z62I^FFG%YF5^4`dI@%6yJ|I86|4Gb=4cJIj7eGZzFMTSPrd!p8|An`!Q@WCz{v)LC
zFFI?VJ`{zwdnSByE;2>E#l(nNYZHZlVJu$`m(9KW(>05U0l}`3AA!-=q`NU^X!5dK
zSjI;~(+_t|?RcHgiB79*5g*)5j{v+uEMZUOhhHui6o)p)9HY3*r;9s<?sCxb=h1st
zqz}spo=5TnXC=*nWUO8s>ndR1t1b_V39kUmqiF?4-)y;#P;5qhSMnr>D((Kw+#{X{
zgn$X0f`?@vCgqvqMBWV8U=fnE7M%o-1b)E-pPqj!w%75>nXVbSvzJFI%;ln}xzh2D
zf|x(2P$qVp+U;08o)b@Qy*K)rxfZh1_zUR5QEopQt(Vz%wiN7UA5nl-=(q>(cyU8c
zEbJ59brQvZ&^Y<q091b6^owwpGvpQ7qFm+sD?gsObzb53hfzuDOc0plOK@Gt8d6Vw
z+goF3B<*rGAHp^muy(mXflJR~Qasl%L^9<zJ$UlSVX<YW_ru?x+)K?o!iSBQ+3|a{
z!V0OJ_tNUo-6j`~;z;gL0C3hQ)2o^<Ebv)TRL?h{?vJC7ZUUH4mj$d8?Tf*I*!tE4
zemn0%xo^}BZ5=ee=Rda^bd>-`#X|jC!R?99TaUXQooF<tUe__F4rr_M632x)N#tFY
zjn%;EM9SuVwD8&Ho|B;Cen5fpMYXdP2&7K0%I4F&Gf)JQr_+uy$D?<o(-Wu&ZSwj3
zYyZMtgOf8Vvq_V~u$dmow~sD1yX8|v>t^k~WUC7dZ0NKSS9J!uNnD?*{QB}})-Pg+
zVySzx!l>e2Rn#bDTT5l}(>qSzY0J5Wom3IOQ>B*$iVU=#Jjp1r+5LU}%9~wjta$8}
z?@M&*QxjRUbXsSeuGsoh`GOzGo>q-q8N-X5c7)Qb{Ji?4Mh9!ae}2E$@c9ppCWSW8
z*;-pKnq!;7c@YwjwD8^>Cg%w`s39ALn6kB%$<=2UNR$rU)FSlWCmWf4$0zb=dq%x#
ztkL)m*J92z(^J8ZIdQ^79McT_yvB#YGD7CFlI-0%<w%p6Gq#0R*>_5COdyVpdai@G
zo?w0%;OZzN-K>^k-J&{&w&UOe?^-?KJa*~nxCod`1h_+iv&*~8!8G_KLP8v`3^&T8
z<s-|qn{c<8;+Em_Ie+^uFSyoj`1+TM3F-9oCQZ8F`|a<|jgQV3{E|D$#uDEJKw-{J
zE_*ILV$*ZV{QXIOw#vCpdIcI2N_brF5W$ibtFzbX^8s$m%j_$@2{Iut^TTp3#~O)|
ztZV&n9I+;68@(xPsB=kcvQC+gRu1j(yV;=bO4y0fO$$pRW8}~8I@<R_ep*838F{$x
z!k=};ef@y0w55)8LC$Y=Do6gDXsalCmr8m15Yyh*Ve<EcNnqPE;KnHM*!5Gx0s4+^
z)%#Id=zh|)6_Nl}vNZRAIf0&U={rCI#tR<ttLJ-De}l8e-Nj)hC(+SV<M*6LCn&**
z26j4r88vsom8zgKuGh_?7D48>kLMFellfmp3T!&j9k1tG$9ug>*|H|3pT`G(b4`wh
zhV~tK#cHm0LMd70oGlhVBm!iiZYOkYEfwHVBUHkmX_0)WO$|V3wR)$-*gl9ilhk7x
zrx*;vUW+S?+>QtKXab;z4?KRme7Dld*{H%5%}}VMJN>Y|9z74!%NMc`KhR<>G9>#6
z=pb{W;m*wpoJ>xy>Zs!dgXb~NOBXf4VfPe*9)^`4mDqPURz6LO_72w2)6zK>gps8A
zo1;LMUu{K0^evW;bi((DwI-fX(dau<GET_-oIw(aY4(BSQbam+?D`|wWQ1(pp*lNN
z#5{x$wM$hymfSjCZ6c*Lzt4}s6_M?>bhi+&S@T{QEDw*|x@&|n;MUk*xC*+RHb~2+
zaU*%gS%Q+L-1hE@tgM$^G^efSzm3G#qmUyLqLwN;&Fn$ZI-`Wq9u;0I=Y5Mih{TzU
z89)h>zkcvd6sRCG${hY<PZn!ccut5BFO&tN$!1mBXbDKeb=VW}Cx3aQtECy(63<H#
zO1PNHS2C;JQm%!G<X+#BLIT<iv3Lv`O}?m&H-cXIv|>}#N~nKP$b%R)S&8P$C%Iw`
z4np*?`+@|mK0rHiwXrJNKi_4tk2cm6Gc?N$C0AFW?1}_B!`(J|)=R-l|HRcQ_kp1G
zXL1CSc<VbqXyqG>7rtfySg8p&KFkzc+8`ZT!0ihxRQ)|4f7fWYwSDwEy@g=*cD)#8
zW}gQ20@NxsP?Jye{WWx-?qm*4Ksjea4#!c*=AQDbw<d6({&@ChWmn`gX`I)374Pri
znsan9JF3tjkz>iPk3X;!T^uiY^&90j50NiuHJ1ULRJ@=vUq2=!CZNeVNUcn_Ri1B%
zk}Xmj@+sqCIfaA$tHQG5k>+kke&F=i9RZ7poTdm(=^UWFonq4$HjRofWQh7#e4`T7
zHu(MZ>I{;sDA(h07eKi*S?lqrJVXbdjWaj%bdh(k<Q%kj##q@I38~Y_Mj7K3g7h`3
zSw@G;&;lirJ|a~6)j<M#AG9I(gu_$%Pp4bjR*iYs4jZsOGb0GiBpB#T+~_mqUyXJ$
zjuc3S#=ZP5Z#o<pvdf{(nW4!Pr>cs;b!QVnH(t5iyZHN9+Dz%#dT*-D-xOnKFz+2H
zR-~hQ2o(|A^CUU325LgS7i0`-qBP%NCGNw=K+po?X%wMW#LPYeYtcnq(5Wd#N9v57
zFkP{ve~r_FqF$Y4<K9ZlRXZ@#q*CVnn_2O6aVr5qCs7?lV25A{O2H3la>9=hEOr%s
zJDx@BK{&f%)||eSL{(tWzA$`nR;M0?vzNQ>prpK!iGKzVpyFw5a$@!NTcQjqq?PD{
z{TlG(g#j8Su<hc`#Ew{{?}DU{xI4`DBi`<`uai6B@Crx(Ww~8m74!-5^c}N45gm#L
z?;E>YfYKlQ<-=Iu_Iut_&;S8fp7Zszg|LdH78^}B`uhn~B6nCV2L<l;2C^s7X;nx!
zbx(dc8-lqAdfv#ua<mmEi{rLNPaiOx^IOAS28;=S_6saIWu3=>c%)a<$DDOQKY=kL
zb8o<+^MWZdf!*ha@`zbT^gY%D)4OjHLi2Wo2QCmj!)>h5pU&k~vj%^z9aY0k>!7Wh
z-`7`v4EmkS5kGj&<v1Q~@19@05OOW8871zsouBvgp>(u3dx&^KuK_c&h&mi65;a`T
z+<gw2vRP$&YJF{K<THDqVos+rB82ZI<YYhGX?mck0yN1QgETp)Q&#8_=s5qvT?`<m
zOtf-VyiVxa4w}7SlAXI1IlLrIlQ><|5nZA__fkSUCFg1)cZ|2EprDlp^~TWe8QgCo
z_wSsbrBwJUm6ANF`X`|>(Rg~I3#bR;PWS0Upc=i`cRsRMIJ$7JNyfEB8*!F66@VZ#
zHPtO%_l6=2DwcNLrp`XKlJ~x%gw^mTf+S7KCz}wEDBl}!jQ#cXMq^sT+sfD2#1mz~
z1+Pz&NbK8$?PAW)7S4W<eAM4>HG6`wVDjv_k9vimvKU}U>NtW%ufRVhncgpJ+sx9E
zx>3X_O3xQdgsw`49WK-Q^@nq(d>9yYEuzuF*+mUTVdWw6FdnkkbV`=Mj_3^`51qm>
zrBK5e6<zl@7IlmvNyeEzD&A%TQ{ZE4%%Ljf(9VDCU4B-y5KF+_;JE-K*B|nmy~+v2
zAk~f_>|=0BwowP$7F4f-?t5#b@ffEcFnXxxQB8S0mWAg0&c8nf036QxNks>5uU*9u
zrF;Uu16OFMcbE<Wvsz>55GAA#W`XKFq`2|%hju=4mvCOhiJ1I(K%oSRc!54|>1iY-
z8F=_bGQ4OVL{<J|M-XsSGl5!tOV>jlvUv6^VBmpul3{v5O%Q|a9~7wo=P;1CcJ$r0
znNTyVaZQrHL&(85tf5JeyKj1_SMy5p)4D*|8%?uVV%l{4;!#N~MAShZyV>!+lN{4E
zM7w@~Lecu(0ga_`Gf$TFhJUtuf{x-W8uS@P#5bsUj<FLWVKY;lrv>(1AEL+dk(1Vh
zM?aKI%+GTiS*xP=+MXmqjy8u2Qh96DVz<&?^f~myXc-2NW|aw}9kY>0+2q$ZGPN<R
zYTp`}&q?wcZMX9F>E#0@UL-=iE+TSGyoZo(0DnsH`l0Cv^;#xNyoZVtyJ77W|BWTb
zDF3m*cjPFMIc|ll!#x0r$7gWZ9{_G`td6^C)&q_XF~x}PtP6Z>5cyAe_V20*0}J$i
zs52mh3kTPs^dKM?yGFCj(v7ye-dzU1T`=y*nSSc^1#o>9&a5TqXhs&pn806}+=p#R
zDgu7uZwk{&AwZCu%-kqNKH%+w@3sLNgcjv)yjowqahQ`gg4&3r^%bg``!+i%jd!{$
zIc+!k#}o*6QGBM}xcPh(t-b=PbHSDIm)O_CEQ0`x&GP|uOf*CSC~Of)GLCZ`^IMiA
zCC<>8dDMZ;lF``?w1l19ww>a?TzJ_cPA>RV<T()63I%{a;aJ`>8F*@`$E!a^^~qYN
zl>*&a#PhNx>H|KA&Kk0+=iO6UI-{5<^+t7EcH1EKgz{iwM@E2yg+qxO{{~{$n1xJU
zw+_ZvH%>0hm%nvjZwc(S45;IJmGxFrDRvuIPjt0UC)kYA(D8f(XICvxM%^S@#N?dS
z7KlR~*pNhXJ})tyP>L5fV<2DfFAcSyR(J7IDs$zXlCFN}(`zSu2j0ptmCWM%pq^G`
zT_w>*(w+WB;Fs5c`o&>c*3?RNI|6K%m_5R?GOAmIaQ;F#D$$qxpcpl@9Bj`I%UPfu
zrKULr9quieUa?(J<-P&gBYF1TC(vum$M)djXDP++K4_h)nLjo3lqW=XDncReA;#zJ
zlA?$s<=mv|cw$l|hby2Q%hgn*(AZXIQfB7{EY_*6*pFDc2ugjVo1V|1clab2q<r5E
znmjdV-x+Ima)+(IKV9$Aqz%na=SFH^9|^(Dt$Zhhj@*-o^sS~pXwK@~$9{EJZ&+2{
z2Oa;j^@3MGs@EvOv)yC}$DkXhrShWBiEb~P1riX@slArX4y`>qzVoVwtyaV!0jJlI
zMC(D8Hi>GN-EH-1vH}ea9P^tqVviVN0R&G-GiPZPJjs2-D0PSH0EaEKJX8FcYoalM
zge3OCMbT+;Cl)#mdc{degNHe0xV_+>PU?Szy9;g>7f%vxRRz%MmyaHIh;W$_{A=1(
zHgF7?BR=NKOU961f`uU>lX|9f1hr^*6LTgOl(7IZaeaXLz3m7=%cov24^_=d0fJ6;
zX_#!j78EsobPtbSv<T|RuM{Edz}YtMGnJ=s&Sm*Ll^-1}CZacD)HG>%sJr{|funl+
zbH3$I!dKrk-nM!Hp^zmLu64{^<n?2-92gi=f~F0{C@b?9#$=o;2P9cv0o$*5O6hb1
z&DlsVu8Y}kD`tsx%72p(Cyx!*sqwWFPnsI0WC#;xqc{~kf1dlrxW;_=DM2t1<)%>J
z2Y|k^s-Ur=IPbWwOuzZKCkZI2uViAU4rCJrDm2oKyqaT(b~jgZ-*^)#AEFIRCeb(_
z?DETSP-(q*r*3`ljXP)vsecyBVL_&%;>caBc+Y<<j(Dgnzn8zY&tx7QGdeWpH2GzD
zk(X<7ifx*3Ew;NYIHO=?KJF4ll3%{-#t62JM(HRK<}LefGG(6pyJN$SvqKc-QMmkm
zgFKK2BmYw_Z5BpmCvX)`HbZEI<!(H;0AZW@olWeN?s5<m^DLK+$kaPMR4j|Bzxic&
zD<AA^+$uPETJG&;g24r4ImV_G*T|<4=~ajWt68jP@}Auc>-)IT(jY>x!&GEmtoa1C
z<Y~jp_)M#mknz&tsl$`I(3Cs%B+GVo(<;%4Q#fz#xlstbqLIZcgUbDaOnh}9fhQN-
z3uE!etybC#P_K35!t%;KTcGq5Au7$9Q*|3gHUIB?usUwG(JWEF5^BF<U#lO)y=~m)
zst%0q@vGqmJSAVWRcgOT2ANU^LG)~)A@;V`Ga)w3qfsZY72k_^AkGN=?4`i{-ESDn
zO*G-?1=XrillKSc^6ISk@|mH7=;lyoXH0NSb;yY;5m}ym<a%#Luw`fQC?*&AB654w
z(MUN}cl{7VSYRh;Rp<hy^}`?2?pc?`mnRxL^YA5T#Z=g;Rn4Xpx<-H}5Zyv$`F;-4
zsQLa~27Ori=xAnIpsjD~>XR~7$9)?c_mETTPWKISRrZ@zYq0G>71#}ZnR43Kj#I|d
zU8XW4wAxMQ&Bn>*9)?dX?KqMwcjMD64rs3n_XWn<;tt0Ws$<znPO0gAJ)jRn?>gL9
z^}1d^Pllo|o2q6tkD?N72zIY<YBTZe6r1sv!@k6YydM!g)dez`MkN6^!Z`}TOA<ai
z+CMfn7PNDDg4*I<bs*I)Ktyr1t0t=Ab%DHrf8R20cP>KUSItVO1-{zJop@=xZ5bNb
zp?btB0|WA@ob65{XKb(MT=z6aY(`*M;4C1-$E$^KwE~D}O#vV(#q7&!=SGGlb(={8
z;=`#o+BS={*G?aKY*=?bVeE@-m~2Jc5oM^RFx`8F<E#Y^_oBE^)od)dokyvVE=#%z
zqMo{LM1+-<9W5!OOVKqQ<;`C{3%!j?@y_l~fZ16eD-58H(aY)fZW)9&n9<_k70)l%
zaZ9aa0G$}{`hci?US7Z)5}lGwW!;;~C*5;Wc5&}{$%BL@vl)RI*_w{Yz5h5X2~aK1
z03t{X6L7JNc%E!pcYgCYoRpQ^t%<|1iG(ett>REbqrwL@n2y+Q`h?0zUx|Dr5M~J?
z2tlz<L+M1NdGWIFRW5b)6M|NZa4*+$-Ze#O8I&gTZDCA-8zGVG16=!RGlX0SYOA~!
zG|EClq3CCSp)UB)yY%z<u1`2bl`(hgg94Qdz&M36`5I>aA7qq2q8M=X-bwg3)AaW^
z*B^k|n1~Qv!T*=H0bv4||DnSEVZJ1ofX<;5=*VEuze%*e$G}V={K;7TmscEgz_5n~
zNE$!?-`R&p-Vgrs$Ul$&*B}TUAhvRiab}VJ;j{iT07T;cTa15y)PE<E(?XCk930%d
lqKu>_U{(Hq6A$qoB~Mp_n$qNI9gq@EQTDY=sg!ZR{{rk{#xei^

diff --git a/docs/use_cases/scf_to_cps/imputing-from-scf-to-cps.md b/docs/use_cases/scf_to_cps/imputing-from-scf-to-cps.md
deleted file mode 100644
index 18e09c7..0000000
--- a/docs/use_cases/scf_to_cps/imputing-from-scf-to-cps.md
+++ /dev/null
@@ -1,924 +0,0 @@
-# Example: imputing wealth from the SCF to the CPS
-
-This notebook demonstrates a full pipeline powered by the `microimpute` package and specifically the `autoimpute` function used to impute wealth variables from the Survey of Consumer Finances to the Current Population Survey.
-
-The Survey of Consumer Finances (SCF) is a triennial survey conducted by the Federal Reserve that collects detailed information on U.S. families' balance sheets, income, and demographic characteristics, with a special focus on wealth measures. The Current Population Survey (CPS) is a monthly survey conducted by the Census Bureau that provides comprehensive data on the labor force, employment, unemployment, and demographic characteristics, but lacks detailed wealth information.
-
-By using `microimpute`, wealth information can be transfered from the SCF to the CPS, enabling economic analyses that require both detailed labor market and wealth data.
-
-```python
-import io
-import logging
-import zipfile
-from typing import List, Optional, Union
-
-import numpy as np
-import pandas as pd
-import requests
-import plotly.graph_objects as go
-from plotly.subplots import make_subplots
-from pydantic import validate_call
-from tqdm import tqdm
-import warnings
-
-from microimpute.config import (
-    VALIDATE_CONFIG, VALID_YEARS, PLOT_CONFIG
-)
-from microimpute.comparisons import *
-from microimpute.visualizations import *
-from microimpute.utils.data import preprocess_data
-
-logger = logging.getLogger(__name__)
-```
-
-## Loading and preparing the SCF and CPS datasets
-
-The first step in the imputation process involves acquiring and harmonizing the two datasets. Extracting data from the SCF and the CPS, and then processing it to ensure the variables are compatible for imputation are crucial pre-processing steps for successful imputation. This involves identifying predictor variables that exist in both data sets and can meaningfully predict wealth, as well as ensuring they are named and encoded identically. 
-
-```python 
-@validate_call(config=VALIDATE_CONFIG)
-def scf_url(year: int, VALID_YEARS: List[int] = VALID_YEARS) -> str:
-    """Return the URL of the SCF summary microdata zip file for a year.
-
-    Args:
-        year: Year of SCF summary microdata to retrieve.
-
-    Returns:
-        URL of summary microdata zip file for the given year.
-
-    Raises:
-        ValueError: If the year is not in VALID_YEARS.
-    """
-    logger.debug(f"Generating SCF URL for year {year}")
-
-    if year not in VALID_YEARS:
-        logger.error(
-            f"Invalid SCF year: {year}. Valid years are {VALID_YEARS}"
-        )
-        raise ValueError(
-            f"The SCF is not available for {year}. Valid years are {VALID_YEARS}"
-        )
-
-    url = f"https://www.federalreserve.gov/econres/files/scfp{year}s.zip"
-    logger.debug(f"Generated URL: {url}")
-    return url
-
-
-@validate_call(config=VALIDATE_CONFIG)
-def load_scf(
-    years: Optional[Union[int, List[int]]] = VALID_YEARS,
-    columns: Optional[List[str]] = None,
-) -> pd.DataFrame:
-    """Load Survey of Consumer Finances data for specified years and columns.
-
-    Args:
-        years: Year or list of years to load data for.
-        columns: List of column names to load.
-
-    Returns:
-        DataFrame containing the requested data.
-
-    Raises:
-        ValueError: If no Stata files are found in the downloaded zip
-            or invalid parameters
-        RuntimeError: If there's a network error or a problem processing
-            the downloaded data
-    """
-
-    logger.info(f"Loading SCF data with years={years}")
-
-    try:
-        # Identify years for download
-        if years is None:
-            years = VALID_YEARS
-            logger.warning(f"Using default years: {years}")
-
-        if isinstance(years, int):
-            years = [years]
-
-        # Validate all years are valid
-        invalid_years = [year for year in years if year not in VALID_YEARS]
-        if invalid_years:
-            logger.error(f"Invalid years specified: {invalid_years}")
-            raise ValueError(
-                f"Invalid years: {invalid_years}. Valid years are {VALID_YEARS}"
-            )
-
-        all_data: List[pd.DataFrame] = []
-
-        for year in tqdm(years):
-            logger.info(f"Processing data for year {year}")
-            try:
-                # Download zip file
-                logger.debug(f"Downloading SCF data for year {year}")
-                url = scf_url(year)
-                try:
-                    response = requests.get(url, timeout=60)
-                    response.raise_for_status()  # Raise an error for bad responses
-                except requests.exceptions.RequestException as e:
-                    logger.error(
-                        f"Network error downloading SCF data for year {year}: {str(e)}"
-                    )
-                    raise RuntimeError(
-                        f"Failed to download SCF data for year {year}"
-                    ) from e
-
-                # Process zip file
-                try:
-                    logger.debug("Creating zipfile from downloaded content")
-                    z = zipfile.ZipFile(io.BytesIO(response.content))
-
-                    # Find the .dta file in the zip
-                    dta_files: List[str] = [
-                        f for f in z.namelist() if f.endswith(".dta")
-                    ]
-                    if not dta_files:
-                        logger.error(
-                            f"No Stata files found in zip for year {year}"
-                        )
-                        raise ValueError(
-                            f"No Stata files found in zip for year {year}"
-                        )
-
-                    logger.debug(f"Found Stata files: {dta_files}")
-
-                    # Read the Stata file
-                    try:
-                        logger.debug(f"Reading Stata file: {dta_files[0]}")
-                        with z.open(dta_files[0]) as f:
-                            df = pd.read_stata(
-                                io.BytesIO(f.read()), columns=columns
-                            )
-                            logger.debug(
-                                f"Read DataFrame with shape {df.shape}"
-                            )
-
-                        # Ensure 'wgt' is included
-                        if (
-                            columns is not None
-                            and "wgt" not in df.columns
-                            and "wgt" not in columns
-                        ):
-                            logger.debug("Re-reading with 'wgt' column added")
-                            # Re-read to include weights
-                            with z.open(dta_files[0]) as f:
-                                cols_with_weight: List[str] = list(
-                                    set(columns) | {"wgt"}
-                                )
-                                df = pd.read_stata(
-                                    io.BytesIO(f.read()),
-                                    columns=cols_with_weight,
-                                )
-                                logger.debug(
-                                    f"Re-read DataFrame with shape {df.shape}"
-                                )
-                    except Exception as e:
-                        logger.error(
-                            f"Error reading Stata file for year {year}: {str(e)}"
-                        )
-                        raise RuntimeError(
-                            f"Failed to process Stata file for year {year}"
-                        ) from e
-
-                except zipfile.BadZipFile as e:
-                    logger.error(f"Bad zip file for year {year}: {str(e)}")
-                    raise RuntimeError(
-                        f"Downloaded zip file is corrupt for year {year}"
-                    ) from e
-
-                # Add year column
-                df["year"] = year
-                logger.info(
-                    f"Successfully processed data for year {year}, shape: {df.shape}"
-                )
-                all_data.append(df)
-
-            except Exception as e:
-                logger.error(f"Error processing year {year}: {str(e)}")
-                raise
-
-        # Combine all years
-        logger.debug(f"Combining data from {len(all_data)} years")
-        if len(all_data) > 1:
-            result = pd.concat(all_data)
-            logger.info(
-                f"Combined data from {len(years)} years, final shape: {result.shape}"
-            )
-            return result
-        else:
-            logger.info(
-                f"Returning data for single year, shape: {all_data[0].shape}"
-            )
-            return all_data[0]
-
-    except Exception as e:
-        logger.error(f"Error in _load: {str(e)}")
-        raise
-
-scf = load_scf(2022)
-
-# Create mapping from desired variable names to SCF column names
-scf_variable_mapping = {
-    "hhsex": "is_female",  # sex (is female yes/no) (hhsex)
-    "age": "age",  # age of respondent (age)
-    "race": "race", # race of respondent (race)
-    "kids": "own_children_in_household",  # number of children in household (kids)
-    "wageinc": "employment_income", # income from wages and salaries (wageinc)
-    "bussefarminc": "farm_self_employment_income",  # income from business, self-employment or farm (bussefarminc)
-    "intdivinc": "interest_dividend_income",  # income from interest and dividends (intdivinc)
-    "ssretinc": "pension_income",  # income from social security and retirement accounts (ssretinc)
-}
-
-original_columns = list(scf_variable_mapping.keys()) + ["networth", "wgt"]
-scf_df = pd.DataFrame({col: scf[col] for col in original_columns})
-scf_data = scf_df.rename(columns=scf_variable_mapping)
-
-# Convert hhsex to is_female (hhsex: 1=male, 2=female -> is_female: 0=male, 1=female)
-scf_data["is_female"] = (scf_data["is_female"] == 2).astype(int)
-
-predictors = [
-    "is_female", # sex of head of household
-    "age", # age of head of household
-    "own_children_in_household", # number of children in household
-    "race",  # race of the head of household
-    "employment_income", # income from wages and salaries
-    "interest_dividend_income", # income from interest and dividends
-    "pension_income", # income from social security and retirement accounts
-]
-
-imputed_variables = ["networth"]
-
-weights = ["wgt"]
-
-scf_data = scf_data[predictors + imputed_variables + weights]
-
-weights_col = scf_data["wgt"].values
-weights_normalized = weights_col / weights_col.sum()
-scf_data_weighted = scf_data.sample(
-    n=len(scf_data),
-    replace=True,
-    weights=weights_normalized,
-).reset_index(drop=True)
-```
-
-```python
-import ssl
-import requests
-
-# Disable SSL verification warnings (only use in development environments)
-import urllib3
-urllib3.disable_warnings(urllib3.exceptions.InsecureRequestWarning)
-
-# Create unverified context for SSL connections
-ssl._create_default_https_context = ssl._create_unverified_context
-
-# Monkey patch the requests library to use the unverified context
-old_get = requests.get
-requests.get = lambda *args, **kwargs: old_get(*args, **{**kwargs, 'verify': False})
-
-from policyengine_us_data import CPS_2024
-cps_data = CPS_2024()
-#cps_data.generate()
-cps = cps_data.load_dataset()
-
-cps_race_mapping = {
-    1: 1,  # White only -> WHITE
-    2: 2,  # Black only -> BLACK/AFRICAN-AMERICAN
-    3: 5,  # American Indian, Alaskan Native only -> AMERICAN INDIAN/ALASKA NATIVE
-    4: 4,  # Asian only -> ASIAN
-    5: 6,  # Hawaiian/Pacific Islander only -> NATIVE HAWAIIAN/PACIFIC ISLANDER
-    6: 7,  # White-Black -> OTHER
-    7: 7,  # White-AI -> OTHER
-    8: 7,  # White-Asian -> OTHER
-    9: 7,  # White-HP -> OTHER
-    10: 7,  # Black-AI -> OTHER
-    11: 7,  # Black-Asian -> OTHER
-    12: 7,  # Black-HP -> OTHER
-    13: 7,  # AI-Asian -> OTHER
-    14: 7,  # AI-HP -> OTHER
-    15: 7,  # Asian-HP -> OTHER
-    16: 7,  # White-Black-AI -> OTHER
-    17: 7,  # White-Black-Asian -> OTHER
-    18: 7,  # White-Black-HP -> OTHER
-    19: 7,  # White-AI-Asian -> OTHER
-    20: 7,  # White-AI-HP -> OTHER
-    21: 7,  # White-Asian-HP -> OTHER
-    22: 7,  # Black-AI-Asian -> OTHER
-    23: 7,  # White-Black-AI-Asian -> OTHER
-    24: 7,  # White-AI-Asian-HP -> OTHER
-    25: 7,  # Other 3 race comb. -> OTHER
-    26: 7,  # Other 4 or 5 race comb. -> OTHER
-}
-
-# Apply the mapping to recode the race values
-cps["race"] = np.vectorize(cps_race_mapping.get)(cps["cps_race"])
-cps["farm_self_employment_income"] = cps["self_employment_income"] + cps["farm_income"]
-cps["interest_dividend_income"] = cps["taxable_interest_income"] + cps["tax_exempt_interest_income"] + cps["qualified_dividend_income"] + cps["non_qualified_dividend_income"]
-cps["pension_income"] = cps["tax_exempt_private_pension_income"] + cps["taxable_private_pension_income"] + cps["social_security_retirement"]
-
-mask_head = cps["is_household_head"]
-income_df = pd.DataFrame({
-    "household_id": cps["person_household_id"],
-    "employment_income": cps["employment_income"],
-    "farm_self_employment_income": cps["farm_self_employment_income"],
-    "interest_dividend_income": cps["interest_dividend_income"],
-    "pension_income": cps["pension_income"],
-})
-household_sums = (
-    income_df
-      .groupby("household_id")
-      .sum()
-      .reset_index()
-)
-heads = pd.DataFrame({
-    "household_id": cps["person_household_id"][mask_head],
-    "is_female":    cps["is_female"][mask_head],
-    "age":          cps["age"][mask_head],
-    "race":         cps["race"][mask_head],
-    "own_children_in_household": cps["own_children_in_household"][mask_head],
-})
-hh_level = heads.merge(household_sums, on="household_id", how="left")
-
-for name, series in cps.items():
-    if isinstance(series, pd.Series) and len(series) == len(hh_level):
-        if name not in hh_level.columns:
-            hh_level[name] = series.values
-
-
-cols = (
-    ["household_id"]
-    + [
-        "farm_self_employment_income",
-        "interest_dividend_income",
-        "pension_income",
-        "employment_income",
-    ]
-    + ["own_children_in_household", "is_female", "age", "race"]
-)
-cps_data = hh_level[cols]
-cps_data["household_weight"] = cps["household_weight"]
-
-household_weights = ["household_weight"]
-
-from policyengine_us import Microsimulation
-sim = Microsimulation(dataset=CPS_2022)
-net_disposable_income = sim.calculate("household_net_income")
-
-cps_data["household_net_income"] = net_disposable_income
-```
-
-## Running wealth imputation with autoimpute
-
-After harmonizing the two datasets, the `autoimpute` function from `microimpute` can be used to transfer wealth information from the SCF to the CPS. This powerful function streamlines the imputation process by automating hyperparameter tuning, method selection, validation, and application.
-
-Behind the scenes, `autoimpute` evaluates multiple statistical approaches, including Quantile Random Forest, Ordinary Least Squares, Quantile Regression, and Statistical Matching. It performs cross-validation to determine which method most accurately captures the relationship between the predictor variables and wealth measures in the SCF data. The function then applies the best-performing method to generate synthetic wealth values for CPS households.
-
-By enabling hyperparameter tuning, the function can optimize each method's parameters, further improving imputation accuracy. This automated approach saves considerable time and effort compared to manually testing different imputation strategies, while ensuring the selection of the most appropriate method for this specific imputation task.
-
-```python
-warnings.filterwarnings("ignore")
-
-# Run the autoimpute process
-imputations, imputed_data, fitted_model, method_results_df = autoimpute(
-    donor_data=scf_data,
-    receiver_data=cps_data,
-    predictors=predictors,
-    imputed_variables=imputed_variables,
-    weight_col=weights[0],
-    tune_hyperparameters=True,  # enable automated hyperparameter tuning
-    normalize_data=True,  # normalization
-)
-```
-
-## Comparing method performance
-
-The method comparison plot below shows how different imputation methods performed across various quantiles. Lower quantile loss values indicate better performance. 
-
-```python
-# Extract the quantiles used in the evaluation
-quantiles = [q for q in method_results_df.columns if isinstance(q, float)]
-
-comparison_viz = method_comparison_results(
-    data=method_results_df,
-    metric="quantile_loss",
-    data_format="wide",
-)
-fig = comparison_viz.plot(
-    title="Autoimpute method comparison",
-    show_mean=True,
-)
-```
-
-![png](./autoimpute_model_comparison.png)
-
-## Evaluating wealth imputations
-
-To assess the imputation results, a comparison the distribution of wealth in the original SCF data with the imputed values in the CPS allow examining how well the imputation preserves important characteristics of the wealth distribution, such as its shape, central tendency, and dispersion.
-
-Wealth distributions are typically highly skewed, with a long right tail representing a small number of households with very high net worth. A successful imputation should preserve this characteristic skewness while maintaining realistic values across the entire distribution. Examining both the raw distributions and log-transformed versions of wealth values can better capture important information for evaluation.
-
-```python
-def plot_log_transformed_net_worth_distributions(
-    scf_data: pd.DataFrame,
-    imputed_data: pd.DataFrame,
-    title: Optional[str] = None,
-) -> go.Figure:
-    """Plot the log-transformed distribution of net worth in SCF and imputed CPS data."""
-
-    palette = px.colors.qualitative.Plotly
-    scf_color = '#1f77b4'   # palette[0]
-    cps_color = '#9467bd'   # palette[1]
-    scf_median_color = scf_color
-    cps_median_color = cps_color
-    scf_mean_color = scf_color
-    cps_mean_color = cps_color
-
-    def safe_log(x):
-        sign = np.sign(x)
-        log_x = np.log10(np.maximum(np.abs(x), 1e-10))
-        return sign * log_x
-
-    scf_log = safe_log(scf_data["networth"])
-    cps_log = safe_log(imputed_data["networth"])
-
-    scf_log_median, cps_log_median = np.median(scf_log), np.median(cps_log)
-    scf_log_mean, cps_log_mean     = np.mean(scf_log), np.mean(cps_log)
-
-    fig = go.Figure()
-
-    # histograms
-    fig.add_trace(go.Histogram(
-        x=scf_log,
-        nbinsx=150,
-        opacity=0.7,
-        name="SCF (weighted) log net worth",
-        marker_color=scf_color,
-    ))
-    fig.add_trace(go.Histogram(
-        x=cps_log,
-        nbinsx=150,
-        opacity=0.7,
-        name="CPS imputed log net worth",
-        marker_color=cps_color,
-    ))
-
-    # medians (dashed)
-    fig.add_trace(go.Scatter(
-        x=[scf_log_median, scf_log_median],
-        y=[0, 20],
-        mode="lines",
-        line=dict(color=scf_median_color, width=2, dash="dash"),
-        name=f"SCF median: ${10**scf_log_median:,.0f}",
-    ))
-    fig.add_trace(go.Scatter(
-        x=[cps_log_median, cps_log_median],
-        y=[0, 20],
-        mode="lines",
-        line=dict(color=cps_median_color, width=2, dash="dash"),
-        name=f"CPS median: ${10**cps_log_median:,.0f}",
-    ))
-
-    # means (dotted)
-    fig.add_trace(go.Scatter(
-        x=[scf_log_mean, scf_log_mean],
-        y=[0, 20],
-        mode="lines",
-        line=dict(color=scf_mean_color, width=2, dash="dot"),
-        name=f"SCF mean: ${10**scf_log_mean:,.0f}",
-    ))
-    fig.add_trace(go.Scatter(
-        x=[cps_log_mean, cps_log_mean],
-        y=[0, 20],
-        mode="lines",
-        line=dict(color=cps_mean_color, width=2, dash="dot"),
-        name=f"CPS mean: ${10**cps_log_mean:,.0f}",
-    ))
-
-    # layout
-    fig.update_layout(
-        title=title if title else "Log-transformed net worth distribution comparison",
-        xaxis_title="Net worth",
-        yaxis_title="Frequency",
-        height=PLOT_CONFIG["height"],
-        width=PLOT_CONFIG["width"],
-        barmode="overlay",
-        bargap=0.1,
-        legend=dict(
-            x=0.01, y=0.99,
-            bgcolor="rgba(255,255,255,0.8)",
-            bordercolor="rgba(0,0,0,0.3)",
-            borderwidth=1,
-            orientation="v",
-            xanchor="left",
-            yanchor="top",
-        ),
-    )
-
-    # custom ticks
-    tick_values = [-6, -4, -2, 0, 2, 4, 6, 8]
-    tick_labels = [
-        "$" + format(10**x if x >= 0 else -(10**abs(x)), ",.0f")
-        for x in tick_values
-    ]
-    fig.update_xaxes(tickvals=tick_values, ticktext=tick_labels)
-
-    return fig
-
-weights_col = cps_data["household_weight"].values
-weights_normalized = weights_col / weights_col.sum()
-imputed_data_weighted = imputed_data.sample(
-    n=len(imputed_data),
-    replace=True,
-    weights=weights_normalized,
-).reset_index(drop=True)
-
-
-plot_log_transformed_net_worth_distributions(scf_data_weighted, imputed_data_weighted).show()
-```
-
-![png](./autoimpute_best_model_imputations.png)
-
-The logarithmic transformation provides a clearer view of the wealth distribution across its entire range. By logarithmically scaling the data, the extreme values are compressed while expanding the visibility of differences in the lower and middle portions of the distribution.
-
-This transformation is particularly valuable for wealth data, where values can span many orders of magnitude. The plot above, shows how closely the imputed CPS wealth distribution matches the original SCF distribution in terms of shape and central tendency after the imputation performed by the QRF model. The vertical lines marking the mean and median values help gauge how these statistical properties have been preserved through the imputation process.
-
-## Comparing with other methods
-
-```python
-donor_data = scf_data[predictors + imputed_variables + weights]
-receiver_data = cps_data[predictors + household_weights]
-
-donor_data, normalizing_params = preprocess_data(donor_data[predictors + imputed_variables], normalize=True, full_data=True)
-donor_data[weights[0]] = scf_data[weights[0]]
-receiver_data, _ = preprocess_data(receiver_data[predictors], normalize=True, full_data=True)
-receiver_data["household_weight"] = cps_data["household_weight"]
-receiver_data["household_net_income"] = cps_data["household_net_income"]
-
-mean = pd.Series(
-    {col: p["mean"] for col, p in normalizing_params.items()}
-)
-std = pd.Series(
-    {col: p["std"] for col, p in normalizing_params.items()}
-)
-
-from microimpute.models import *
-
-def impute_scf_to_cps(model: Type[Imputer], 
-                      donor_data: pd.DataFrame,
-                      receiver_data: pd.DataFrame,
-                      cps_data: pd.DataFrame,
-                      predictors: List[str],
-                      imputed_variables: List[str],
-                      weights: List[str]) -> pd.DataFrame:
-    """Impute SCF data into CPS data using the specified model."""
-    model = model()
-    fitted_model = model.fit(
-        X_train=donor_data,
-        predictors=predictors,
-        imputed_variables=imputed_variables,
-        weight_col=weights[0],
-    )
-    imputations = fitted_model.predict(X_test=receiver_data)
-
-    cps_imputed = cps_data.copy()
-    cps_imputed["networth"] = imputations[0.5]["networth"]
-
-    return cps_imputed
-
-quantreg_cps_imputed = impute_scf_to_cps(
-    model=QuantReg,
-    donor_data=donor_data,
-    receiver_data=receiver_data,
-    cps_data=cps_data,
-    predictors=predictors,
-    imputed_variables=imputed_variables,
-    weights=weights,
-)
-
-quantreg_cps_imputed["networth"] = quantreg_cps_imputed["networth"].mul(std["networth"])
-quantreg_cps_imputed["networth"] = quantreg_cps_imputed["networth"].add(mean["networth"])
-
-weights_col = receiver_data["household_weight"].values
-weights_normalized = weights_col / weights_col.sum()
-quantreg_cps_imputed_weighted = quantreg_cps_imputed.sample(
-    n=len(quantreg_cps_imputed),
-    replace=True,
-    weights=weights_normalized,
-).reset_index(drop=True)
-
-
-ols_cps_imputed = impute_scf_to_cps(
-    model=OLS,
-    donor_data=donor_data,
-    receiver_data=receiver_data,
-    cps_data=cps_data,
-    predictors=predictors,
-    imputed_variables=imputed_variables,
-    weights=weights,
-)
-
-ols_cps_imputed["networth"] = ols_cps_imputed["networth"].mul(std["networth"])
-ols_cps_imputed["networth"] = ols_cps_imputed["networth"].add(mean["networth"])
-
-ols_cps_imputed_weighted = ols_cps_imputed.sample(
-    n=len(ols_cps_imputed),
-    replace=True,
-    weights=weights_normalized,
-).reset_index(drop=True)
-
-matching_cps_imputed = impute_scf_to_cps(
-    model=Matching,
-    donor_data=donor_data,
-    receiver_data=receiver_data,
-    cps_data=cps_data,
-    predictors=predictors,
-    imputed_variables=imputed_variables,
-    weights=weights,
-)
-
-matching_cps_imputed["networth"] = matching_cps_imputed["networth"].mul(std["networth"])
-matching_cps_imputed["networth"] = matching_cps_imputed["networth"].add(mean["networth"])
-
-matching_cps_imputed_weighted = matching_cps_imputed.sample(
-    n=len(matching_cps_imputed),
-    replace=True,
-    weights=weights_normalized,
-).reset_index(drop=True)
-
-
-qrf_model = QRF()
-fitted_model, best_params = qrf_model.fit(
-    X_train=donor_data,
-    predictors=predictors,
-    imputed_variables=imputed_variables,
-    weight_col=weights[0],
-    tune_hyperparameters=True,
-)
-imputations = fitted_model.predict(X_test=receiver_data)
-
-qrf_cps_imputed = cps_data.copy()
-qrf_cps_imputed["networth"] = imputations[0.5]["networth"]
-
-qrf_cps_imputed["networth"] = qrf_cps_imputed["networth"].mul(std["networth"])
-qrf_cps_imputed["networth"] = qrf_cps_imputed["networth"].add(mean["networth"])
-
-qrf_cps_imputed_weighted = qrf_cps_imputed.sample(
-    n=len(qrf_cps_imputed),
-    replace=True,
-    weights=weights_normalized,
-).reset_index(drop=True)
-
-
-def plot_all_models_net_worth_log_distributions(
-    scf_data: pd.DataFrame,
-    model_results: dict,
-    title: Optional[str] = None,
-) -> go.Figure:
-    """Plot log-transformed net worth distributions for all models in a 2x2 grid.
-    
-    Args:
-        scf_data: Original SCF data with networth column
-        model_results: Dictionary mapping model names to their imputed dataframes
-        title: Optional title for the entire figure
-        
-    Returns:
-        Plotly figure with 4 subplots
-    """
-    from plotly.subplots import make_subplots
-    import plotly.graph_objects as go
-    
-    # Create subplots
-    fig = make_subplots(
-        rows=2, cols=2,
-        subplot_titles=list(model_results.keys()),
-        vertical_spacing=0.15,
-        horizontal_spacing=0.12,
-    )
-    
-    # Define safe log transformation
-    def safe_log(x):
-        sign = np.sign(x)
-        log_x = np.log10(np.maximum(np.abs(x), 1e-10))
-        return sign * log_x
-    
-    # Calculate SCF log values once
-    scf_log = safe_log(scf_data["networth"])
-    scf_log_median = np.median(scf_log)
-    scf_log_mean = np.mean(scf_log)
-    
-    # Define colors
-    scf_color = '#1f77b4'
-    palette = px.colors.qualitative.Plotly
-    model_colors = palette[:4]
-    
-    # Plot positions
-    positions = [(1, 1), (1, 2), (2, 1), (2, 2)]
-    
-    for idx, (model_name, imputed_data) in enumerate(model_results.items()):
-        row, col = positions[idx]
-        model_color = model_colors[idx]
-        
-        # Calculate model log values
-        model_log = safe_log(imputed_data["networth"])
-        model_log_median = np.median(model_log)
-        model_log_mean = np.mean(model_log)
-        
-        # Add SCF histogram (grey/transparent)
-        fig.add_trace(
-            go.Histogram(
-                x=scf_log,
-                nbinsx=150,
-                opacity=0.3,
-                name=f"SCF (weighted by sampling)",
-                marker_color='grey',
-                showlegend=(idx == 0),  # Only show in legend once
-            ),
-            row=row, col=col
-        )
-        
-        # Add model histogram
-        fig.add_trace(
-            go.Histogram(
-                x=model_log,
-                nbinsx=150,
-                opacity=0.7,
-                name=f"{model_name.replace(' imputations', '')}",
-                marker_color=model_color,
-                showlegend=True,
-            ),
-            row=row, col=col
-        )
-        
-        # Get y-axis range for vertical lines
-        fig.update_yaxes(range=[0, 2000], row=row, col=col)
-        
-        # Add median lines
-        fig.add_trace(
-            go.Scatter(
-                x=[scf_log_median, scf_log_median],
-                y=[0, 2000],
-                mode="lines",
-                line=dict(color='grey', width=2, dash="dash"),
-                name=f"SCF Median",
-                showlegend=False,
-            ),
-            row=row, col=col
-        )
-        
-        fig.add_trace(
-            go.Scatter(
-                x=[model_log_median, model_log_median],
-                y=[0, 2000],
-                mode="lines",
-                line=dict(color=model_color, width=2, dash="dash"),
-                name=f"{model_name} Median",
-                showlegend=False,
-            ),
-            row=row, col=col
-        )
-        
-        # Determine correct axis references for annotations
-        if idx == 0:
-            xref, yref = "x", "y"
-        elif idx == 1:
-            xref, yref = "x2", "y2"
-        elif idx == 2:
-            xref, yref = "x3", "y3"
-        else:
-            xref, yref = "x4", "y4"
-        
-        # Add text annotations for statistics
-        fig.add_annotation(
-            x=0,  # Position on the x-axis (log scale)
-            y=5000,  # Position on the y-axis
-            xref=xref,
-            yref=yref,
-            text=f"Median: ${10**model_log_median:,.0f}<br>Mean: ${10**model_log_mean:,.0f}",
-            showarrow=False,
-            bgcolor="rgba(255,255,255,0.8)",
-            bordercolor="rgba(0,0,0,0.3)",
-            borderwidth=1,
-            font=dict(size=10),
-            xanchor="right",
-            yanchor="top",
-        )
-    
-    # Update layout
-    fig.update_layout(
-        title=title if title else "Log-transformed net worth distributions by model",
-        height=800,
-        width=1000,
-        showlegend=True,
-        legend=dict(
-            x=0.5,
-            y=-0.2,
-            xanchor="center",
-            yanchor="top",
-            orientation="h",
-        ),
-        barmode="overlay",
-    )
-    
-    # Update x-axes with custom tick labels
-    tick_values = [-6, -4, -2, 0, 2, 4, 6, 8]
-    tick_labels = [
-        "$" + format(10**x if x >= 0 else -(10**abs(x)), ",.0f")
-        for x in tick_values
-    ]
-    
-    for i in range(1, 5):
-        fig.update_xaxes(
-            tickvals=tick_values, 
-            ticktext=tick_labels,
-            title_text="Net Worth (log scale)" if i > 2 else "",
-            row=(i-1)//2 + 1, 
-            col=(i-1)%2 + 1
-        )
-        fig.update_yaxes(
-            title_text="Frequency" if i % 2 == 1 else "",
-            row=(i-1)//2 + 1, 
-            col=(i-1)%2 + 1
-        )
-    
-    return fig
-
-
-# Create dictionary of model results
-model_results = {
-    "QRF imputations": qrf_cps_imputed_weighted,
-    "OLS imputations": ols_cps_imputed_weighted,
-    "QuantReg imputations": quantreg_cps_imputed_weighted,
-    "Matching imputations": matching_cps_imputed_weighted,
-}
-
-# Create and show the combined plot
-combined_fig = plot_all_models_net_worth_log_distributions(scf_data_weighted, model_results)
-combined_fig.show()
-```
-![png](./imputations_model_comparison.png)
-
-Comparing the wealth distributions that result from imputing from the SCF on to the CPS with four different models, we can visually recognize the different strengths and limitations of each of them. The implications of using one model instead of another for imputation will be further explored by evaluating the impact wealth imputed data has on microsimulation results.
-
-## Wealth distributions by disposable income deciles
-
-Lastly, to confidently say that our wealth imputations are coherent with other household characteristics, we can compare the average net worth values per disposable income decile for each of the four methods used. 
-
-```python
-income_col  = "household_net_income"
-wealth_col  = "networth"          # e.g. "net_wealth"
-
-decile_means = []
-
-for model, df in model_results.items():
-    tmp = df.copy()
-
-    # Create 1–10 decile indicator (unweighted)
-    tmp["income_decile"] = (
-        pd.qcut(tmp[income_col], 10, labels=False) + 1
-    )
-
-    # Mean wealth in each decile
-    out = (
-        tmp.groupby("income_decile")[wealth_col]
-           .mean()
-           .reset_index(name="mean_wealth")
-    )
-    out["Model"] = model
-    decile_means.append(out)
-
-avg_df = pd.concat(decile_means, ignore_index=True)
-avg_df["income_decile"] = avg_df["income_decile"].astype(int)
-
-fig = px.bar(
-    avg_df,
-    x="income_decile",
-    y="mean_wealth",
-    color="Model",
-    barmode="group",
-    labels={
-        "income_decile": "Net-income decile (1 = lowest, 10 = highest)",
-        "mean_wealth":   "Average household net worth ($)",
-    },
-    title=(
-        "Average household net worth by net-income decile<br>"
-        "<sup>Comparison of imputation models</sup>"
-    ),
-)
-
-fig.update_layout(
-    xaxis=dict(dtick=1, tick0=1),
-    paper_bgcolor="#F0F0F0",
-    plot_bgcolor="#F0F0F0",
-    yaxis_tickformat="$,.0f",
-    hovermode="x unified",
-)
-
-fig.update_xaxes(showgrid=False)
-fig.update_yaxes(showgrid=False)
-
-fig.show()
-```
-
-![png](./by_income_decile_comparisons.png)
-
-QRF clearly presents the most consistent and plausable realtionship to disposable income, with a gradually increasing average as the deciles increase. This plot also supports the behavior observed above showing the extreme negative and positive predictions that OLS and QuantReg produce at the left and right tails, respectively, and Matching's underprediction at every decile. 

From baf9181a2ffe2e4a93ea20a7c636ddfc641abe88 Mon Sep 17 00:00:00 2001
From: juaristi22 <juaristi@uni.minerva.edu>
Date: Fri, 27 Feb 2026 19:19:49 +0530
Subject: [PATCH 4/5] fix minor test bugs due to pandas new data types

---
 microimpute/comparisons/metrics.py            | 35 ++++---
 microimpute/config.py                         |  2 +-
 microimpute/evaluations/cross_validation.py   | 12 ++-
 microimpute/models/mdn.py                     | 36 +++++++-
 microimpute/models/ols.py                     |  3 +-
 microimpute/utils/dashboard_formatter.py      |  6 +-
 .../visualizations/comparison_plots.py        | 91 +++++++++++--------
 .../visualizations/performance_plots.py       |  8 ++
 tests/test_metrics.py                         |  4 +-
 tests/test_models/test_imputers.py            |  8 +-
 tests/test_models/test_matching.py            |  2 +-
 11 files changed, 139 insertions(+), 68 deletions(-)

diff --git a/microimpute/comparisons/metrics.py b/microimpute/comparisons/metrics.py
index c963962..29336ec 100644
--- a/microimpute/comparisons/metrics.py
+++ b/microimpute/comparisons/metrics.py
@@ -175,19 +175,25 @@ def order_probabilities_alphabetically(
     return reordered_probabilities, alphabetical_classes
 
 
+# Accept numpy arrays, pandas Series, and pandas ExtensionArrays
+# (e.g. ArrowExtensionArray) which newer pandas versions produce for
+# string columns.
+ArrayLike = Union[np.ndarray, pd.Series, pd.api.extensions.ExtensionArray]
+
+
 @validate_call(config=VALIDATE_CONFIG)
 def compute_loss(
-    test_y: np.ndarray,
-    imputations: np.ndarray,
+    test_y: ArrayLike,
+    imputations: ArrayLike,
     metric: MetricType,
     q: float = 0.5,
-    labels: Optional[np.ndarray] = None,
+    labels: Optional[ArrayLike] = None,
 ) -> Tuple[np.ndarray, float]:
     """Compute loss for given true values and imputations using specified metric.
 
     Args:
-        test_y: Array of true values.
-        imputations: Array of predicted/imputed values.
+        test_y: Array-like of true values.
+        imputations: Array-like of predicted/imputed values.
         metric: Type of metric to use ('quantile_loss' or 'log_loss').
         q: Quantile value (only used for quantile_loss).
         labels: Possible label values (only used for log_loss).
@@ -198,6 +204,13 @@ def compute_loss(
     Raises:
         ValueError: If inputs have different shapes or invalid metric type.
     """
+    # Convert array-like inputs (e.g. ArrowStringArray, pd.Series)
+    # to numpy before use.
+    test_y = np.asarray(test_y)
+    imputations = np.asarray(imputations)
+    if labels is not None:
+        labels = np.asarray(labels)
+
     try:
         # Validate input dimensions
         if len(test_y) != len(imputations):
@@ -291,9 +304,9 @@ def _compute_method_losses(
                 log.error(error_msg)
                 raise ValueError(error_msg)
 
-            # Get values
-            test_values = test_y[variable].values
-            pred_values = imputation[quantile][variable].values
+            # Get values as numpy arrays (handles Arrow-backed dtypes)
+            test_values = np.asarray(test_y[variable])
+            pred_values = np.asarray(imputation[quantile][variable])
 
             # Compute loss
             _, mean_loss = compute_loss(
@@ -327,9 +340,9 @@ def _compute_method_losses(
                 log.error(error_msg)
                 raise ValueError(error_msg)
 
-            # Get values
-            test_values = test_y[variable].values
-            pred_values = imputation[quantile][variable].values
+            # Get values as numpy arrays (handles Arrow-backed dtypes)
+            test_values = np.asarray(test_y[variable])
+            pred_values = np.asarray(imputation[quantile][variable])
 
             # Get unique labels from test data
             labels = np.unique(test_values)
diff --git a/microimpute/config.py b/microimpute/config.py
index c2b0eb7..ba5896a 100644
--- a/microimpute/config.py
+++ b/microimpute/config.py
@@ -55,7 +55,7 @@
     "ols": {
         # statsmodels OLS uses default parameters
         # LogisticRegression params for categorical targets:
-        "penalty": "l2",
+        "l1_ratio": 0,
         "C": 1.0,
         "max_iter": 1000,
     },
diff --git a/microimpute/evaluations/cross_validation.py b/microimpute/evaluations/cross_validation.py
index 664adb8..05f7a19 100644
--- a/microimpute/evaluations/cross_validation.py
+++ b/microimpute/evaluations/cross_validation.py
@@ -265,11 +265,13 @@ def _compute_fold_loss_by_metric(
     for var in imputed_variables:
         metric_type = variable_metrics[var]
 
-        # Get data for this variable
-        test_y_var = test_y_values[var][fold_idx]
-        train_y_var = train_y_values[var][fold_idx]
-        test_pred_var = test_results[quantile][fold_idx][var].values
-        train_pred_var = train_results[quantile][fold_idx][var].values
+        # Get data for this variable, converting to numpy to handle
+        # Arrow-backed dtypes (e.g. ArrowStringArray) that pydantic
+        # won't accept as np.ndarray.
+        test_y_var = np.asarray(test_y_values[var][fold_idx])
+        train_y_var = np.asarray(train_y_values[var][fold_idx])
+        test_pred_var = np.asarray(test_results[quantile][fold_idx][var])
+        train_pred_var = np.asarray(train_results[quantile][fold_idx][var])
 
         # Compute loss based on metric type
         if metric_type == "quantile_loss":
diff --git a/microimpute/models/mdn.py b/microimpute/models/mdn.py
index fe78e68..1649ecb 100644
--- a/microimpute/models/mdn.py
+++ b/microimpute/models/mdn.py
@@ -64,6 +64,19 @@
         message=".*have no logger configured.*",
         module="pytorch_lightning.core.module",
     )
+    warnings.filterwarnings(
+        "ignore",
+        message=".*isinstance.*LeafSpec.*is deprecated.*",
+    )
+    warnings.filterwarnings(
+        "ignore",
+        message=".*given NumPy array is not writable.*",
+    )
+    warnings.filterwarnings(
+        "ignore",
+        message=".*Trying to unpickle estimator.*",
+        module="sklearn.*",
+    )
 
     # After import, also update the rank_zero_module logger
     from lightning_fabric.utilities.rank_zero import rank_zero_module
@@ -148,6 +161,25 @@ def _generate_data_hash(X: pd.DataFrame, y: pd.Series) -> str:
     return hashlib.md5(combined.encode()).hexdigest()[:12]
 
 
+def _get_package_versions_hash() -> str:
+    """Get a short hash of sklearn and pytorch_tabular versions.
+
+    Including package versions in the cache key ensures that caches
+    saved with older versions are automatically invalidated when
+    packages are upgraded, avoiding deserialization warnings.
+    """
+    import sklearn
+
+    versions = [sklearn.__version__]
+    try:
+        import pytorch_tabular
+
+        versions.append(pytorch_tabular.__version__)
+    except (ImportError, AttributeError):
+        versions.append("unknown")
+    return hashlib.md5("_".join(versions).encode()).hexdigest()[:6]
+
+
 def _generate_cache_key(
     predictors: List[str], target: str, data_hash: str
 ) -> str:
@@ -168,7 +200,9 @@ def _generate_cache_key(
     predictors_hash = hashlib.md5(predictors_str.encode()).hexdigest()[:8]
     # Sanitize target name for filesystem
     safe_target = target.replace("/", "_").replace("\\", "_")
-    return f"{predictors_hash}_{safe_target}_{data_hash}"
+    # Include package versions to invalidate cache on upgrades
+    ver_hash = _get_package_versions_hash()
+    return f"{predictors_hash}_{safe_target}_{data_hash}_{ver_hash}"
 
 
 class _MDNModel:
diff --git a/microimpute/models/ols.py b/microimpute/models/ols.py
index e0ed4ee..f2fb6e8 100644
--- a/microimpute/models/ols.py
+++ b/microimpute/models/ols.py
@@ -59,8 +59,9 @@ def fit(
                 y_encoded = y_encoded.fillna(0)  # Default to first category
 
         # Extract relevant LR parameters from kwargs
+        # Use l1_ratio instead of penalty (deprecated in sklearn 1.8)
         classifier_params = {
-            "penalty": lr_kwargs.get("penalty", "l2"),
+            "l1_ratio": lr_kwargs.get("l1_ratio", 0),
             "C": lr_kwargs.get("C", 1.0),
             "max_iter": lr_kwargs.get("max_iter", 1000),
             "solver": lr_kwargs.get(
diff --git a/microimpute/utils/dashboard_formatter.py b/microimpute/utils/dashboard_formatter.py
index fe8dbb6..5d56309 100644
--- a/microimpute/utils/dashboard_formatter.py
+++ b/microimpute/utils/dashboard_formatter.py
@@ -700,10 +700,8 @@ def format_csv(
         # Generate histogram data for each imputed variable
         for var in imputed_variables:
             # Check if variable is categorical or numerical
-            if donor_data[
-                var
-            ].dtype == "object" or pd.api.types.is_categorical_dtype(
-                donor_data[var]
+            if pd.api.types.is_string_dtype(donor_data[var]) or isinstance(
+                donor_data[var].dtype, pd.CategoricalDtype
             ):
                 # Categorical variable
                 hist_data = _compute_categorical_distribution(
diff --git a/microimpute/visualizations/comparison_plots.py b/microimpute/visualizations/comparison_plots.py
index 4e1b85a..53248cf 100644
--- a/microimpute/visualizations/comparison_plots.py
+++ b/microimpute/visualizations/comparison_plots.py
@@ -200,11 +200,19 @@ def _process_dual_metrics_input(
                         ):
                             std_results = ql_data["results_std"].loc["test"]
 
+                        # Loss values in results are already averaged
+                        # across variables, so create ONE row per
+                        # (method, quantile) to avoid duplicate bars.
+                        variables = ql_data.get("variables", ["y"])
+                        var_label = (
+                            variables[0] if len(variables) == 1 else "average"
+                        )
+
                         for quantile in test_results.index:
-                            for var in ql_data.get("variables", ["y"]):
-                                row = {
+                            long_format_data.append(
+                                {
                                     "Method": method_name,
-                                    "Imputed Variable": var,
+                                    "Imputed Variable": var_label,
                                     "Percentile": quantile,
                                     "Loss": test_results[quantile],
                                     "Metric": "quantile_loss",
@@ -214,21 +222,20 @@ def _process_dual_metrics_input(
                                         else np.nan
                                     ),
                                 }
-                                long_format_data.append(row)
+                            )
 
                     # Add mean loss
                     if "mean_test" in ql_data:
-                        for var in ql_data.get("variables", ["y"]):
-                            long_format_data.append(
-                                {
-                                    "Method": method_name,
-                                    "Imputed Variable": var,
-                                    "Percentile": "mean_quantile_loss",
-                                    "Loss": ql_data["mean_test"],
-                                    "Metric": "quantile_loss",
-                                    "Std": ql_data.get("std_test", np.nan),
-                                }
-                            )
+                        long_format_data.append(
+                            {
+                                "Method": method_name,
+                                "Imputed Variable": var_label,
+                                "Percentile": "mean_quantile_loss",
+                                "Loss": ql_data["mean_test"],
+                                "Metric": "quantile_loss",
+                                "Std": ql_data.get("std_test", np.nan),
+                            }
+                        )
 
             # Process log loss if available
             if (
@@ -240,35 +247,41 @@ def _process_dual_metrics_input(
                     ll_data.get("results") is not None
                     and not ll_data["results"].empty
                 ):
-                    # Log loss is constant across quantiles
+                    # Log loss is constant across quantiles.
+                    # Same as quantile_loss: one row per method,
+                    # not per variable, since values are aggregated.
                     if "test" in ll_data["results"].index:
                         test_loss = ll_data["results"].loc["test"].mean()
                         test_std = ll_data.get("std_test", np.nan)
-                        for var in ll_data.get("variables", []):
-                            long_format_data.append(
-                                {
-                                    "Method": method_name,
-                                    "Imputed Variable": var,
-                                    "Percentile": "log_loss",
-                                    "Loss": test_loss,
-                                    "Metric": "log_loss",
-                                    "Std": test_std,
-                                }
-                            )
+                        ll_variables = ll_data.get("variables", [])
+                        ll_var_label = (
+                            ll_variables[0]
+                            if len(ll_variables) == 1
+                            else "average"
+                        )
+                        long_format_data.append(
+                            {
+                                "Method": method_name,
+                                "Imputed Variable": ll_var_label,
+                                "Percentile": "log_loss",
+                                "Loss": test_loss,
+                                "Metric": "log_loss",
+                                "Std": test_std,
+                            }
+                        )
 
                     # Add mean loss
                     if "mean_test" in ll_data:
-                        for var in ll_data.get("variables", []):
-                            long_format_data.append(
-                                {
-                                    "Method": method_name,
-                                    "Imputed Variable": var,
-                                    "Percentile": "mean_log_loss",
-                                    "Loss": ll_data["mean_test"],
-                                    "Metric": "log_loss",
-                                    "Std": ll_data.get("std_test", np.nan),
-                                }
-                            )
+                        long_format_data.append(
+                            {
+                                "Method": method_name,
+                                "Imputed Variable": ll_var_label,
+                                "Percentile": "mean_log_loss",
+                                "Loss": ll_data["mean_test"],
+                                "Metric": "log_loss",
+                                "Std": ll_data.get("std_test", np.nan),
+                            }
+                        )
 
         self.comparison_data = pd.DataFrame(long_format_data)
 
@@ -805,7 +818,7 @@ def _plot_stacked_contribution(
                     )
 
             if title is None:
-                title = "Rank-based mmodel performance by variable (lower is better)"
+                title = "Rank-based model performance by variable (lower is better)"
 
             fig.update_layout(
                 title=title,
diff --git a/microimpute/visualizations/performance_plots.py b/microimpute/visualizations/performance_plots.py
index 8e9de4a..6c88f70 100644
--- a/microimpute/visualizations/performance_plots.py
+++ b/microimpute/visualizations/performance_plots.py
@@ -67,6 +67,14 @@ def __init__(
         # Handle different input formats
         if isinstance(results, pd.DataFrame):
             # Backward compatibility: single metric DataFrame
+            # Note: passing a bare DataFrame loses results_std, so error
+            # bars will not be shown. Pass the full cross_validate_model
+            # output dict to preserve error bar data.
+            logger.warning(
+                "Received a bare DataFrame instead of the full results dict. "
+                "Error bars will not be available. Pass the full output of "
+                "cross_validate_model() to enable error bars."
+            )
             self.results = {"quantile_loss": {"results": results.copy()}}
             self.has_quantile_loss = True
             self.has_log_loss = False
diff --git a/tests/test_metrics.py b/tests/test_metrics.py
index 302df50..ca416c8 100644
--- a/tests/test_metrics.py
+++ b/tests/test_metrics.py
@@ -1289,8 +1289,8 @@ def test_compare_distributions_return_format() -> None:
     assert len(results) == 2
 
     # Check data types
-    assert results["Variable"].dtype == "object"
-    assert results["Metric"].dtype == "object"
+    assert pd.api.types.is_string_dtype(results["Variable"])
+    assert pd.api.types.is_string_dtype(results["Metric"])
     assert results["Distance"].dtype in ["float64", "float32"]
 
 
diff --git a/tests/test_models/test_imputers.py b/tests/test_models/test_imputers.py
index 762d45d..7c50ef0 100644
--- a/tests/test_models/test_imputers.py
+++ b/tests/test_models/test_imputers.py
@@ -264,7 +264,7 @@ def test_imputation_categorical_targets(
 
     # Default behavior returns DataFrame directly
     assert isinstance(predictions, pd.DataFrame)
-    assert predictions["categorical"].dtype == "object"
+    assert pd.api.types.is_string_dtype(predictions["categorical"])
 
     # Test probability predictions for models that support it
     if model_class.__name__ in ["OLS", "QRF", "Matching"]:
@@ -279,7 +279,9 @@ def test_imputation_categorical_targets(
 
         # Check that we still get the categorical predictions
         assert isinstance(predictions_with_probs[0.5], pd.DataFrame)
-        assert predictions_with_probs[0.5]["categorical"].dtype == "object"
+        assert pd.api.types.is_string_dtype(
+            predictions_with_probs[0.5]["categorical"]
+        )
 
         # Check probability format
         prob_info = predictions_with_probs["probabilities"]["categorical"]
@@ -333,7 +335,7 @@ def test_categorical_return_probs_false(
     predictions = fitted_model.predict(X_test)
     assert isinstance(predictions, pd.DataFrame)
     assert "categorical" in predictions.columns
-    assert predictions["categorical"].dtype == "object"
+    assert pd.api.types.is_string_dtype(predictions["categorical"])
     assert set(predictions["categorical"].unique()).issubset({"A", "B", "C"})
 
     # Test 2: Explicit return_probs=False with quantiles should return dict of DataFrames
diff --git a/tests/test_models/test_matching.py b/tests/test_models/test_matching.py
index e6c8e40..7c01bff 100644
--- a/tests/test_models/test_matching.py
+++ b/tests/test_models/test_matching.py
@@ -208,7 +208,7 @@ def test_matching_mixed_types() -> None:
     predictions = fitted_model.predict(X_test, quantiles=[0.5])
 
     assert predictions[0.5]["target_numeric"].dtype == np.float64
-    assert predictions[0.5]["target_category"].dtype == object
+    assert pd.api.types.is_string_dtype(predictions[0.5]["target_category"])
 
 
 # === Edge Cases ===

From 55f0a130034843057019bd1f249e8856430b2df8 Mon Sep 17 00:00:00 2001
From: juaristi22 <juaristi@uni.minerva.edu>
Date: Fri, 27 Feb 2026 20:36:47 +0530
Subject: [PATCH 5/5] changelog

---
 changelog.d/maria-paper_review.fixed.md | 1 +
 1 file changed, 1 insertion(+)
 create mode 100644 changelog.d/maria-paper_review.fixed.md

diff --git a/changelog.d/maria-paper_review.fixed.md b/changelog.d/maria-paper_review.fixed.md
new file mode 100644
index 0000000..2354004
--- /dev/null
+++ b/changelog.d/maria-paper_review.fixed.md
@@ -0,0 +1 @@
+Updated paper and package documentation with latest changes. Fix pandas 2.x compatibility for Arrow string types and dtype checks.