From e6129babcba2b2a8a6b2e0db54b84d968931c52a Mon Sep 17 00:00:00 2001
From: simonpcouch <simonpatrickcouch@gmail.com>
Date: Mon, 6 Nov 2023 19:10:37 +0000
Subject: [PATCH] =?UTF-8?q?Deploying=20to=20gh-pages=20from=20@=20tidymode?=
 =?UTF-8?q?ls/stacks@31199f864a65d9945670cd1cfa35de57fc0f8b44=20?=
 =?UTF-8?q?=F0=9F=9A=80?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 dev/CODE_OF_CONDUCT.html                      |   2 +-
 dev/LICENSE-text.html                         |   2 +-
 dev/LICENSE.html                              |   2 +-
 dev/articles/basics.html                      |   2 +-
 dev/articles/classification.html              |   2 +-
 .../figure-html/unnamed-chunk-4-1.png         | Bin 228234 -> 230364 bytes
 dev/articles/index.html                       |   2 +-
 dev/articles/workflowsets.html                |   2 +-
 dev/authors.html                              |   6 +--
 dev/index.html                                |   2 +-
 dev/news/index.html                           |   5 ++-
 dev/pkgdown.yml                               |   2 +-
 dev/reference/add_candidates.html             |   2 +-
 dev/reference/augment.model_stack.html        |   2 +-
 dev/reference/autoplot.linear_stack.html      |   2 +-
 dev/reference/axe_model_stack.html            |   2 +-
 dev/reference/blend_predictions.html          |   8 +++-
 dev/reference/build_linear_predictor.html     |   2 +-
 dev/reference/collect_parameters.html         |   2 +-
 dev/reference/control_stack.html              |   2 +-
 dev/reference/example_data.html               |   2 +-
 dev/reference/fit_members.html                |  15 ++++---
 dev/reference/get_expressions.html            |   2 +-
 dev/reference/index.html                      |   2 +-
 dev/reference/predict.data_stack.html         |   2 +-
 dev/reference/predict.model_stack.html        |  40 +++++++++---------
 dev/reference/prediction_eqn.html             |   2 +-
 dev/reference/reexports.html                  |   2 +-
 dev/reference/stack_predict.html              |   6 +--
 dev/reference/stacks.html                     |   2 +-
 dev/reference/stacks_description.html         |   2 +-
 dev/reference/tree_frogs.html                 |   2 +-
 dev/search.json                               |   2 +-
 33 files changed, 68 insertions(+), 64 deletions(-)

diff --git a/dev/CODE_OF_CONDUCT.html b/dev/CODE_OF_CONDUCT.html
index 6711434b..b6339997 100644
--- a/dev/CODE_OF_CONDUCT.html
+++ b/dev/CODE_OF_CONDUCT.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/dev/LICENSE-text.html b/dev/LICENSE-text.html
index 1acae699..244fbec2 100644
--- a/dev/LICENSE-text.html
+++ b/dev/LICENSE-text.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/dev/LICENSE.html b/dev/LICENSE.html
index 47af4985..7d5d7b58 100644
--- a/dev/LICENSE.html
+++ b/dev/LICENSE.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/dev/articles/basics.html b/dev/articles/basics.html
index 83f89582..59bac5d1 100644
--- a/dev/articles/basics.html
+++ b/dev/articles/basics.html
@@ -36,7 +36,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/dev/articles/classification.html b/dev/articles/classification.html
index 49c130a5..4e09f518 100644
--- a/dev/articles/classification.html
+++ b/dev/articles/classification.html
@@ -36,7 +36,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/dev/articles/classification_files/figure-html/unnamed-chunk-4-1.png b/dev/articles/classification_files/figure-html/unnamed-chunk-4-1.png
index 89e2456771fe28d80ecfc633d77cd9d2f167caea..bcca2a37fdba951af9d95366dc0406b1c7ebda30 100644
GIT binary patch
literal 230364
zcmdSBXHZjN7cLqS2+{?lNfi`SdM}}=Ac!<kdKD0mUPFgO1Stw4BE2XIB1jP-bV8Hf
zM0!c+odggPTDaTq%sJoO-}lZr_wE@cOqe%&=UwYvYdz0;7IB99T6EM`sX-tR-2-h6
zBM=A*{0gbQ^zY^Ch}geBK<yuDX#lUlb$qWO5_qBV);9M8fw*7&`yoqO*?R^8@qr#_
z+<6j|y*(Eg#ub>Oigz|`diWthq_2r!lcJ!{)kU@OX0Uax^&E>uEKzHTYw=kjm#_-E
zUlq0aqWy*7+mk*x<Fa?p9@e;*STvP>o;yC@a~W~1e&&?WyBXYK1J=5_lUUO~7&blO
z#2!P&_}?q;))8|4|MrIR4UoovFZ;JlApZYelG;?D|1M5u|4fdE{a=?htr1xLze}$}
zZ;|o+cPUfUd(h?ou8q5+1NvVpV`Dq`{vX^no_C%5?WD1VMSO~o0sUXk*x0|EMWa+{
zeH5VBq@-5^PMla3CPv2OkH1!YTQ~f{!^baUY1KI8JvDrN_tkMR+PaUx>YrALdrLw4
zJwlWr^8fQoI9{q<`Cth=cbvox(DhAVpOGCM9msh)#^kSVo}Mx73m}lAlN0~7Yfljb
zZD3VQ=7$fgKT#+MNJ>h|U-Sllds`wm3_n&L5s)8v@46tcBahPFzyF+K!pOm@1cN<g
zhk|%c|HsCZ)YsROWQ+_ajQFLj*3E>NR4mBN{^yfuX#DTn&i_MJ{Qv!J4bRkZ|2YnZ
z&*>@-lnsq2eg=XCej_O;(A-tX_kiEP7?bfFpu|)#r4GU^0MQ5f<pdEHga{Xb;WO&E
zkHC4*b_ZM$F#7fQY_a^Bj0-h>ZtZ=aqZKIx8$Fq&K7kaWCvEto6&D#q#|$azL)M${
z$ug#^Le11D)0OHYNvw>?-ZC)6#XNm>!~E%~@f&oHFnU}XjhUrb{!G_Ts<SSqtjUEA
z&Ve##kbfQRj?C=ZmHE^{?Ff1rFQjR)P7ebmH8bOHYH0j(%F#^=#3!BnF*3(5VCe(I
z&(FwtVaDXv;*WGehwR=J<StvwgvD+2i#@c!L*$F>Tn}62S$QI^xR{$b1veS2uMUu?
z^v)-*4duZ%vmo?RtN5j~1L_dGO#>Q*YF}j0EgsP3BfhuU!;z<}&-<QZkMGg))mOtM
z9#ocG3c~r?EuF2;wLArxdQLww$)PPG)4l4^zU_msZV%Ob;p<Q4l>>&`fUa5`t8gMV
zg3(b4NX27FeJDRa<USClsJ@l41i{t3fBqy8mbTIL=IEuqaMPdQMl&TJim$F<rfu{~
zE4vX5zGn!rO%08c_EZZjr6uB5%{(Ts9<lUHeZd&ZyoFdvL6*oP2clcOM)3#XS#&|C
zGH=-#l<Q??RoEEg61msFloxZJe;{7D<UH1_=i&P`TVVX`oeMR2T5R3DQb8No=-H0k
zPjFXxMHL##r$%X}OxCTt(KMIghOVMSa2_^AzRjUluH-UP88{1%r<JGsj&AXS_HKnl
zVn7YIE_Ot7h2~mZ)Zqpij_Us+Dw(l78791-=tWz~C}>`h#;TOOImi~L^POG@e+ogD
z9fU`jJb!>bH3g$|lQzo|VrgSpDNC%HqAqglt<Xaliu=ys(D=3nwRVslL@tMXTo<zc
zqpd-OG1+A@c)qQM4e=`N2p0kpO+>dw8Vqrhf6%9%r`nQ2Oe8_F^vSFT=haTbxrE>0
zth+y>bNR6o4dFteTb9T?9Qp05NUaa)Gg`?D9TJGV#!-19CV?(!<EF%IW;4Yv7`UE>
zqoE<--)B;^*diQ}p5zB1*_w;u`*N};LQbxtaK2<YgS$RgY)qv@>9J`I3Xf$oMgIWy
zeA@RMu1_(#vbdf%o_J}txb@yn^CGf1lO=gLppN}a#juUq$6q|Y!S{9r9D12oemIm}
zRm+)xaWZ-g)5?C+8#}KW3^SURE)z|!(f5oDMTmhQwCgz0qw@oOA^ajE)=AjAO9Z>o
zgjiT=oEm@%G8ZN;-c)4V@FD()qc7Mcj32xLJVD0fp+*cFkf}P#nN8Ve<(FJCa5zqq
zlL;t@Bq4#yCvdp~S!Nf1<6tJY3HmzDu9LKq***6mR>Rgq8lPvWNZtR~=SB-W8={IL
zW1mVw;$6J`%`bA`DXm29TU9}4BHyX9K?+Y6)rbs#JC0F#6lHVEv9;>6Gc6%jfJ=M5
z63@calu^{0QS^5w%#&hcY2|u@<6~`w%I_17=pd8!_Pf=w-G&SPWT|gfQSz6kI{(0O
zT2a86{n(V^79B)+p<-RKq`Ot4?N3dY|JSkXvHb@+bnMTsJBbF|tSMJ=?$ku~wJCa5
z##J-UikVXKND4`DnQusdYsX__yS+aF>7Hmj!*zbD5Gr~?xW@n6)_u;_M3<Mj?i4rV
z!@<v#D;mA}!Z7s>I;-eYi`s}IitK=<-%CwJ{726t>HE(tW>wL9HB+sf*a2UPViEyf
z)j55{UpcrLa;+x9%2q?eQQ|-Tv2XQ9-s0+XK_o9f|9Ph;<^B&*cQ1%sKg4%b!0#eK
z=>95P<9a<Us6kW&)Lf*1H|zX}ZS8E0^55lE`>C+mFxs_9#foC^ZuE%vjgEK+J&ynT
z&TC;-`2CKiMvT&b40x;HQzS3Pxz*eaekPBSqD&0DHHc4$h1hSc?rI-IfIa_w2>K99
z`?^rh&3KmXyZZTM-czXva(uESIPd`nh{qDzKZq-D+LXM;W~S6LTk}xJCp7ZYc$3|d
zM^feFcRr>q6LST^)f_5a4(_IKMn;a;{~~IOi2;OZ=~F1LbQfSTO^uV8n7YPVT0u#k
zpfJS?6-hSY^gXmPuYlA$rWY6+PbK`uq3d%tj@x(5%LOL;k3V9MGSC=L@?R%kn+NnL
zRRe^)F>NG_I7+SMyHEu#Tp{3s@Y2eYP1`9R*D}I8YHbJgo}q<j#c3yLG9&Ajz+Ouc
z3@wpg4wLWj^WV$Kq}1#v-+qFPV?j%w3PjGxRrI>G&vs*KV!PFKbU><+^7wX9^?($d
z$2(^B^?mm{yeU*+PX!D|Y2QEbK@KxMNB1l9&C@rUUvgKa_!%MsbwyPzVVi^zub0<J
z*8EJ0=EdrFvyU~@8@ao8*ED?!p4pJ$b9~gzMcEYL9Oy|AlVy5;Lor!K;_q{`uojYg
z4MCr@CUlM7OrA5yvAB*=nR69<ftr{-SFOJ?89-KR$QabnD1t?5u6jOpRH!5)E_0(J
z8wdPuMskJ+u~cZ!z2ldn{ZEjbQ>0w?1&ed7QdscrYBy^_5d;xX;kR=m4SPcyR{X##
zB#Igt-l<ZcYQ(Ilw20J1Ypc;{C+e=Hp-$PrgnmjfQ%Ha)7eq4n39%qb<MT!5j3Rj3
zrMPZqjYXqHxJqrxoy$`0`La6hf$f7m^r?x|u&oh=Oto~YQh*E$)o%e~R`8)v2pA&_
zdB~M*RfI`ZxL&RB2IJLWh&Er$j6pvI$qDlIlS6Z?0T8X(a1Q&sa|bF3r9vHh>?i0?
zt-<|_j2jF<k*uvQ2jX0-`tIAMnEw^*;!^>$+f{&mgxgfWv-W(@?Fq<G%8E`)jqB4!
zOy*<9-9|vZR<Ne!kCrYJny3GFHp8zZBM-3l(Ab<Z^vsk8`rgA6=e*Ba_F+M8jFTQ*
z14sQ4CWKSp2`<2N!Oc0Dm%<_hy!WXqPQ%#BCT1qNlRetwi{Yi2pGMU9TRUc^!mFN&
z!ESNqC=-5%YH7s4{+*}yk{Q`)>)-o&d{~^zTD|1eUHK(EWZq_WY<^*RjL2qx?!d-K
z^(*G4*ri+i_r{5g**7jqtRw10T3;?`d956$(5g`Rpu0SuN$d|lGr2wkT3!PQ3lgk_
z(9lGojNC1COHR}9!Tx>?`%k?j%~BnGY+7vWO&}XSjZa!C-I6yTv!o6kDom_`pmf*x
z(>ACh;(%)E?+>S|S`p_9e1;aIr_Rj2=3$a<+vx2qjlo(jY;rboo(2FU9A&`4Os?ph
zQhdBOPm!%>jhU9tQqnYL`-fd32D%~5q|1h5m#oUytcaCURuBJEl3%y_XPFKcx$1vK
zo6CLY{r|jw{oT(i@&AQ2|Hlx~|E^m4irO%Yud;8&mkcyLGjs9Ri!s@3)nDo2FD@}L
z(dI%F*g^aIrlu2lq`8rik8ZdA2kIdM#XP0O=f-yM0I14!7FiB}CVzi_TKqg<vhtdm
z8st31Kf^2npGg_Q2GnNJr$I&J2<_#6_$lS>TM(#wU_jH}ei4Lw6RZ7KR$l&MqeE6$
zc*~&c&OeA*Wnp1KdZPU{Bcu2~_XNNg2n25^Oij?poHzb2CguPCCZ7G@WxfCR(0e}d
zKP*C=bqtx+H)yXHSRc~e2X=Z*wzvch8S0?cVyjr<@+L|dK4O^MQNKf95pvbL&!10!
zh5Vl}AiX10*$rfW<RCyKU4En<K_3a<r^fFF=l_hESt*1h2!aBKAX1N^@D@#t>nHz?
z>z$qviYGE4eKrXth_(UGx(ZW?ei~$m2Yax7OPV6cox^*J7A+^yAjoP}P8so6&T#I}
z=_%W7V)X$SKh=OJePZ)uSAo*16l}q*u76d+i>o3+?CqC?hk)0SU$sDHI`K*JuxH6u
zXi{10+nF+YFH0Ftb-z3zGNOO)`K0HT99);|S^>{1P&Ed`PZ;JFz6QTzMpR9NMqY%d
z0QDV^nJFd8^DmA}0q4yFC={u$z_*K326Nq7(PTs?<YhQ;$mA-8@DcR;ryrA_g~;0x
ziRKQRk?;bZAY6O|Mz%7)l8#6~A?i&oW~|Df5kxr&XReClC%8P>pOUvKc0}-TGssTn
zf`%4g14&mtK*V`h*4|kffnifTHsH@fhd2_jKNoe?ULQ5|n7tcCIHd1)E!^w7nmpCg
zaO2+VK8mwtZfpk&My=%uH<2)-gy8jpG@pSaB<1%{h4}fSqB4&(9fxb5(1jhaVAUTO
z;CbyXAD_`-L%l!^2`65WI&NmIq(6e#o(ZJwHhE&IPw5+C;CK3<5OF*5o0e5mh=J%Z
z<!9bfpvDkoZ#-s4M>Oh!l)HbVk8P=`asV*y=a!}rF5<$6`PN88fniM^Me`xG;n-y4
z>F6Z%Y|Annz$@|JnDBqb8H3Ipxi&3MyfD{&X^CPnai8UB&M7YjHvPdv+e!UE72J*e
zOb=q@xRN<zPBFHKZ+X)P%AIdl+#l^CZy6Y;PZW(qK?<ai<SC-_)?(PH^X!z3;?LBR
zfsE;LP~sL0x&6csB|qz~)H9-otWCLD&I9@I^>&z{1TB1iTY1PayMX3Hxdp^y*LUyy
zG^Y#H;B*zJyhEXgtO42>=j)I<juDPvzjdY|d2q4EYVf)|??SM{&!#Zi!WSnVIdbkw
zEU!eJ5rlH`-E8jj@D#kiq9TtErZZQz;+G@1PsndR!qt9xx1^n_F_<ZA>1D?ATC<*)
zPi#!CAL(jnKskZ*>S#D=)lGx8vw~`opO1!<P`AD>$T^o2$TOLjf)~%%jRtQJx<L3o
zYji6U;4h`*q!m4^8w2<jqjFpl1$5HhYkVxLRAPu@wdXx6rJ)rSP^esz3a5QLo>Xv&
z%Njid_eR#bd~?BK6s5)M2?G3!No#v@!qHup;U5U%?bzAVfg%OMyr_H**>9M$mZwi}
z8d;<F<ZR8vmSUUtTxs_#N{w6J?n2W|az;!D-w=M51A9--{uAI2cs-sue=g!YM7wn%
zt=SMmo*=<DulvtgOe^}NjCx)F6~%GJKr-#l_x}9ntVWY1TXfg$(MUKO(sswGy0R!y
z%+5>Fd;c~*`VGmT(ZGg^>nyOu1!SWL4YNx<KIfrVQi25f5PhqHiGrYMCHC}5<3Fde
zZYT&!k22ScaY{$hme#w`_-E9GXIe=r5IRT`i=bS!4nMLqtv|AZ^s}loL17$hz3<n3
z>(jkeEv5<1)dOf81O8yWiqou-vF!KN`7(1YF)cMR@214HFCU5RckD(w7DDG`(gPHb
z7eQH_qNG14-z#JCy!2LY9mfYQCB<y|LoxROI#3X>U5ug@n{)0}C;)=BsTLLq7TjGc
zfQ6=&f*(mmDvRg$lZ#u6<xelpj`hYgQz0IX&9z>DR-O#(pp)jIq+>};LQ|99_?hbV
zZE&s`eMc6T=#_6yl)RYHl}2yvhxVO|_2yCKIzrcSm#o1ml=HJ>Y-tIph@BjCtUrb-
z*X;K;C^saE<<&eb+e}x&ZKU`d<!2!+FE#zNQ~C~#hhOh7Fmg02)f11gQiw+!<IDwb
zJgl_7{_d)a8SUi2cd`JbX+p&BfZEfgWNe<Ib2+4ChT*K!V*z8E&vBYOsd{B0j2^Bu
z^#zaq)V-6GJ~~a-7)q}Y{su|-qmf9_<iIK+m(`W5-WrK2gk!%Hpl!?OBJ=ojesami
znq+%!Et~Z|Ha!k1x(<4*CH)i@(V>pZiR~UPAV+jd7*Abl^e0!a`ISKcl_auV0VRn-
zeyxzV&_S|`PfqscMf;84SZ(|HUBef{5L3r@JiC8sKK(9IwpAi~*4Z8|d=dWK1yXZm
zANSW0@qo69D@Q%pdU-l1{FVqM$(d4IN!hv#T9ZS@V-1~c;8T+JyZ02ew^#eb!}IjL
z2J|FPt%D+0rGZlL4LOe$tw^TbZVgusy4W|rl+|0QZnwRS3LSgtiJ4}5IsBNO1X02L
zbMm1;aAbpI$Z2J{LI`T7ulqHrNsW=C-ZeA0%=hQdBB?P?$ZF!!8m40wi+YOadSnYw
zET&%ImZ^5x2>S;KQFBQvbI6{`sWYpASbO`aylE%wj<kgAsgnU?7a2>s0x#Y5!9}Wf
zO}wiYuTB+RR;-NsOWYkR>I$i}<0%D`z^g3Ybd^;T$UG0{y_llkS6LZU8rc>6r_RE;
z>&e3$^@!GDkMBnlPIIN+A)glr1col${054sZ^8vl{;;s?QpBpzD2|WGnTa6)?*;wy
z-j9A0NM5fknDp-zu)Y=9pcO<ln6g5ld$F{fcTM?WB%ANRj(ha^?{j3-GMtl{-szhY
z$R_*E!0vg|FFEWPonH$B=#}`XHmJ6qO|=>gli+;-oN<9o@!|!k^uONqh*LKD+!QR#
zDa!y<nhp*lyxEP1m7j}$Uw9El5Zykt3Z98%B%a@CUR74{1DUC175JYxJW>OB*SouH
zR9utg*2X{1Ft%tXWO78Wpm%A|9?}A+q}*QDC@<z)K5)DF+TqBSO1epSP0>{mMI>9v
zmJ^w@jLFUGre0D(kDj!m@0(Cd2=PuzFtBH+4u=Z0Pj%`<KtqyJ{Wz4xa+~PG&fs<{
z@3APq)y%q6W#0^3oY_|~Sx{qm<`Olx&01)j4>~;IBo`)A9mwlY4yqPa_YE`X-04>)
zDXS*-oEqvYCZ=Wx|M7<TDxQ(6BRki8@2V*i1b!o^>cZDLvc_0jGjMHNA2B;i?2o(@
zpQBX{5NwIw+)!KEs01a;kqz>)ak9GS+uHQiSqk#p>9%$r=c183=2z>g<=cD?1@Td$
zNZ2#EoaM@`kuT>c+kdbgs5jbm0m3OI5G>9<w&6H-EYGvc=>mjJJ3J-^oqC$1kKIr~
zPw>)*Rn|IP#m`=ojwGuHx{F~ww@VG5|HjFoKu{j3p5Zuim8fK|wEXE*QBMnyW2<8C
zZB=^{V^|(fGV#P^d|EJ5p;ItqQpUc^=vVj535_Vun;0of;R^8Y^T((~p2|2co-F#3
zSq|Y3PwMDKzHh?<SSJ1XDQ=8ycyoj`8E_0x|9X}Z+Cut1I1?EnKJ&x*&6{PO$7#`a
zuf_L#)Zx^^0a4~^l-;yGKOa9;T)U47zCH(}_~O9)mk$sO<-2*C17>!hpe1sA@5sDY
z(<(eMV8(}sxGqY<1oDsk0m_KA8nj!lg0)r4b#Y;DA;gqp$xKg4Wilvemrsp1JSQwb
zDR|d#hBoZ4iKp-CRXk=Ulub@1s_Q5!EvR{Uo7C8?w^sRdDa@@vKv*<!LNSysDChxx
z=c+}<WmlJ_^bU?-bo%Zi!DGeo7ds_I<}+P}Dvd8S=|9GH)H!QH*xEgPKvTu&?Mn4=
z2`kh50vckPym|V<Lu(ou-4~TV@~dV7ZI^J99$^E`&SWt9(vx9V>`XcIrjAsl-<Ff)
zi2N`uzFyK0k2x;_Y7_E`jIsbmMxa07I2pX8fdgd%Dc7Wm=oXE6>=-c^YA$MMF<@d5
zxtk8zzU73y^>&#YySEbF;)Sw^cRH|^R4Yf&kY`Ftad#i2@KKpQqUdT#ADeCs0MlE8
zNlxu(u^+A|vOJq1bbqDP#f)st=|u*LjR8_pc=TPjK4jEt@oPnP9OBKv0H(-1rd2;`
z$C)g@=w%<s(5Ms9`>hz=HE~2<G(g^`Qb!AIX}KVOdu1WG+(C*%gQ4@aGWJ~9obKe0
zAoS^3Mh?R)EN^VJGc|0$=?h*3<!*%iR`qZUorHqwn>2>zTJ!K)aZh8UW`Q$CrXE?J
zjTR5wtFvax=PYff`l-}bVs~ZlXtlVR+qzl7I{Vf|G|1t%C%IxdepURk!uu;90x2)Z
z@{no-6x<rC@Z}N5b`9Qif%Y%?1(_hd4Q$Xav1P6sBj3JRXYqgqa_V>X`kng@_EYPr
z*pcY{3|mmxGHY&0xQi_QTt14;xSyL+B*W2z_g=LNI6=-v?df^oJu0S#F&W4ODsAf`
z=*JYovIv5+eWqGGx<F2EyXg04@)x6&^g+-d^M+rlws-wx=ez7-Y1l=Q7XLvm>gpNc
zj61DWZ1q6iOlH41bzI3E;NvgwJXoQbGm#(<5NZ%K4MP+>Yr+T%{7!#e|Lhl*UC1J0
zG>onCDW3fZX*@C~1fcAX{}QA<J386uV1~D)E#*{$_fF2c+iryLTiRegNqXYh92dv3
zIMv$^<YB92x!=SL9IC#Hs%GWwoZkJbLW4NJF^6JauGzJ{LhAX^NAdni6;q!I^>ceM
z?@IokGj2O}no}qqo=Nkq9x8MZ&i-(|IYIfOmrFJ4U&5c^U-^}R6lWjq9=K^W8h2L;
zen`>0(-_>#eMZ2kQweX+H@vvu?(LuWB`4T3K2%9+K#F9zL@3`6VZv9=u)FZ4wcj>G
zc`0|O_b2x?VU!+Fp??MIeh6Pl3K#x!#0n{Q^4Xpw=QVDGvX@GX76VwP!YRamVkMR$
z;Cpt7XQBOG;3>QN{u!49dhK{;6#eBk2{rJmvj>;q6){K8Xuz#6)?qp0uFA-OCyp&*
z6<>(!v8n}biU_(;7RVb|%#3?7aGeo~J|WjUFhSk3Ynfv@))2H80c@Z;j%q0BY>mz-
zpCE{?S0Ep}^=hFGJC5MVB5lG2Wnnq^V0>0`*U&ZTqDwKYiyzzq&Ipt4xl0x*EkTaA
zU{`FEw}rg7+Pi7(4xhLYmH;qEoEOCO$hKoGbkn{m)=N*?Pj;qmvo=rtK0T$>F;oA$
zSG{oKHvfjNEVF1)_?ZIN|I`AMA2tqlIaMV*>#9S}$s5RVs+s!z4w`rD?6AW2bRkvP
z5X?4P?4>zVel~x$?c{Ysg;E@rv}kh~KY@mBiFkNPnFFnNpsJ=7*4shOY&0q6pa-04
zCPJ%}gw=+-uc2oB|G|Tqp}j>+&ZX`LkPrHfpuL<bG%r1fY7hO{%H*eR_eyl{Cn}`y
zpbF5D%imN<baa=)JG<3h&NrB2d*fCZt(I~=d;?3t%NCw-OZ_3c8_mxEP^3aff;EAn
zjx7I6PTZT<%Ou4q6mrbpudA|Pp;+%@wRVHC%wIiFI-M0Hthszl6aMp_1Ls8BrC(5+
zKi{)Db-v#!xZAL<>?gWUg_v>r`}zU(>fQ>eq_yfX(qOKq>+|P$TxGV5-84H^g9FBX
zR_p?5ZlNM>Mum4JrjD6qr8E|GTY#}TqRhfyWZe~cQ(F%WO!PUps=Nyp>>dog_}2Od
z1&%gFen}x>#Mtm9bK%Z6IH8y|O%^N4<@ORvQ^Pw@{YMlN&+9@`^6ohNm^`d<<hK<N
zi;DY$(vOjwnUz3n3SLkl$V+l=-8&1I*&vS>5n<xoRg#e344#J@1is4Y>vW7XQijDU
zi)M5rhnoyu6XrzHux#@~6y80XpqMX$)h$_H!jgto;qZm<rKgDxvK8Jr+|##O7;=|`
z>Q4+VJAce@h<K`)GB=->AuwuQx1(y3S?x;=Fv#8uQ06Q1WB~>}RNMV@UuJq&7C(J!
zRdOpEBWszXJZlL6aA7f+bSQf9pa3{~F(kOa5E<p~nyRqoMq2^-&5WQp8R1Ut`Poxq
z?;}B>-{)MTXAfROcC*{XBY&ry1^%Ih?y#1NH%-U`9NGnd%sB4f&3F@DFdqoE)5xs;
zFlCjJ@Idg_0snsPvHMuSRrPS184FafvAJ#OGGTmZf&E=q8pOFQ3&v6EMn>A=#^&vA
zaelUdrQ9Jp6A0X6@2)T?S6r?Qyb~n(Vj}6(R-$yI?J7BmOPE5{ymwjKgjyUwhsw$&
z7s+d*k8%0_6gx1t$m5FVUAHuMMxQtsW3En-qbsFMzC#@lZ17iNej}QJmpgRs(pS`|
zMw<g%IA*pp;rIk35Vo7?Cbp59@jH}k&xvxa4mvvCgY7cq**>5DF`XGXwR|MT)}`g4
zGJ1My>WVD*QW}5WZ>>y{{pqX5HAnR#PCx+j;CG&wWof$b_YEr7rl1bO-MCzMw!__Z
zq~^7q+>Y2pG3-QWy7&GQi*}qICnU_58_!8lcqcLc7&)Kh;D<P1t>3&)yQjFWx3dac
z;$Kmxf;5$LJGTv=*C`05_;pSUJ>?e^;0SV|!Z-HF=FRH3rGqAed4i4*$K-<Oj?q0S
zo_2s7i1guX!L2W=9>e5!eSg%~ma%9*T+UQ|C0;2oFwt%~+JDg7l~$GYhq~cSBt}i`
zm)}r$XE@FQI35nDA5pSbx_VzEO_2xbhgH0Ecp4Tkt7LsvN<S{SC{;W6*x=>(Av;x9
zA2I02*~sf8J@+8mP_f=j!Ds7r_s)c2vYQWg6{BFexTb~%6)5@X4beEE1thD<KKx5f
zpu~{J$7ole3+Te4?gP)Sdz0E4Kn701Wv_H%0TLg*`zmX|7<0u`im$;}2a_k7i9W7m
zyNxKR=2DYVVosMY^U};Rb}kE`ylB7mO1{T*#lI?wto~jaKo4L$b!2vKIoS(ZQP?0u
zIT3%ZyWR0G8GF*2UYCAU9j*g*X1c&Nj7O;bKE=t&WeITED7Vo>4*9?eq&Y5`u`G5Y
zn>trZLd1;;*5VyH`IHioVWF!a45eHa&Ti^3Br%OWukVpb?hhm6D>jImGu?pwQ|SjK
zHZy^0g=2a=ew|T3Jy`+&g|~WQM9e*zN<`D(ZGa9tk-vfN_~FMn)5aWY3X(1l6}gN0
zrbF9h-fJINR~2<`_HH4(nx+GPMLZLTI;VVSMwgq>RvlE6OCM&l_PxDlPsRYJsZBZ8
z2BiFBlG7iUHJVchiJ}ROgimdQUSq{a*%+$&h8i43UcV>Dicf%(#3&N2ee(-@;+BKh
zR`s;o-Z#FiQ1#MJrNltjkCjD4D1Q;ZURSgBEOWSQ_o#K`H_6g$Dy0u^)`;R|wwK;V
zB$blos`YnC{Bg`qJ?J|z!9F2O=DiH!IdEEI7!dn4bu6MQ347<hKwRHViq#hRS5~rD
zB3Up%<2bU*RujKj>|RhN<zpE*=rm?}(?;RS8UKir1C01*q}09ocSvY{M<5u4jXTIb
z>qhl9Z}M=A4zYyHTX0#4W3w2taKjQMeLY?9(Odbym!s|Zg3TNA_vEC9toVDOHrq)A
zNe5Lf6Y*_jl&vhw*RPZ|RTUsR9e&;!Z&u5IdiEy<I_un8Jvz^@3-?OQB{$ZWOBW=~
z7bxlyM9Z9gK7In?(dW*I!i>(F3AaP;OHhQ+XHB0%FUa8J%1vo=P~N71^&4dO&BJu(
zw&J*&9ZAILjb;0&z`+{0EfXib4K+X^`B~`RKmOs*)NpWSF2ks;G;E#BxqRq`I#v_H
zcaHQ_r^er|@5=q+yL){00?CO?eUGk`bJ99>EQ0Qmq}o^Rq9IQK;X=x8*~7r<&<cHi
zgXeK`sk7?ZYx>Avn?alYzWxJsTk8h|cERzFKB0%~5?iVVGogJCASQYncP%+*NK(?w
z?WF9w=RO`Zku#PY;uQF%kvu98B+T~1m;~QE#zw8x?##AyMd{K6kXowiDsXLm$n@jV
z9e>~V>qma!DP987>9Of*+T%|pghi<)CD?+tp5Q(FM(ctqV`J0*L!d|pzpo}&|B<Kj
zgQ$QV-9knNOM6osy;iQd_v%wMM8W$O5xxtWFAubAVJKmlU2uh0Z2DBV_<+i}9Iz9-
zkg#kqQ$NvHv;Kfg=pyH`sn-v>d0pR|H#zO35e#Mf+`4scWp(5TvcVfkN#3$gayw1V
zDs7CvF^E13d^x3A&0xI<ex7IM{N1mhuXBW|Az^aF9x+FrAn1F2#3f1YV)<n^JX*$*
z%WpYcIU{EXn>z&$kv2|)4O-3!MZ&G^@^=!iu`b)lRN}N{ZD+j{1Ec4vL)-{{O?gS>
zQbB(w+A?p&p$q!JUMaa9@tYXlOA(<I0{uzv6#`PXo(3g-nt^{<BB%zx!ZF5s#xOQr
zZ;c5?i^DT#j&hHc8C>L(`KWGo<7`ToohQE)Z~u_V0>xGx_+86RzTLC8-CXKo2mWM7
zo^ZcWORLGd7Di-vd1!XU*juI08+KoyXS10&(>%85TtDWAu$whVAXpZC`sf_~inR+Z
zxZ3B66tblxE!ePr1EG(%#+g8^_f-Pwl;0Vv{*fWOsH7m4z)Ek(QbuWgkV5*&sS~_)
z_{;Axx8IBf+7F)p@~)Fp;<INp=A-vdFQ$l$RzC?D=k9)o?=N>@w*?;FEjMYlnnQiM
z@%toHzl`vrC8{;K53S|3cdl&R)Dw)FC)?~$sM%Aitqn}@F?PUze#-*q6fRv0HkKHx
z26?f-P17^kzsN+86J&pX&?nD@b#Ip0xr09F?VdIn%+jA7xUKS5WjbMPG32BLl~wTJ
zKPs~3D4aeER%^H0%Cs9rXx7v*(ZPF!c;IBE1Qn5F-%RyHYG^@go<O69rcv*Dig!8B
zTK%<q#bXh0)hH&m7yi!d*tYat^5n6lXJ2Wq^$xE}A>M7S8`M2~xOj5IsXk&ivk#F@
z$mE?b`vb6HbZ09&NJy+u8Axq@6A1T4G+~4QG56gJJC#uGNqCIFB}*=4)j@8`*(%V@
zZYok>dUeR{2{82szC0Lxts6b8&qkT4?v${mPQ?s*C@A^B-U&%)SoQ8ORM59|D2ge1
z#33SP2C%o*m|mStZ3%Q)luy)6GKm10u06A<`83@Ghq)gD640{FR>E#_pSq^N<%ygQ
zb7{KTaKVc*+udCiIj4NJ2&MX3&|zR6R`K9R50~h(`(a$nhe&s&1vMhh2}C>;v-`yk
zPo-tqgJTakkrRnN+xdjaHpOG~s|j-bijZ&5SNHJuRZ_~9d?yuI<0c9L<qn{*>6WOn
zms(57q>@|xOh)sf--Iitr><;FkM>)2fG-l85!_Vr>Ta>Wo4lgDGS^43LMLON8#CrZ
zM0O#mVJ>TRkB6Afd9f!Twb{kc!jPgAkHIgpR$6H8v$LJ%Zi)*(Bc%9jJ9B`6P&r5~
zbYj+i2xQyN+oFPxjWt~PEXlZE7?)6AU3Eed1!3;l*5|RsJHd?p>Ba)Vm~drp)2OH3
z+b@RbqjpP{0Oi!^PFUDGW<}TMJGxL!{zrG@1A#8zkK-LLxjpvO)ZT<fAH`-i{=c2j
zX4SH><A~BZ{$Tg#?n~1YBQ_5O-nrWa>=Wc9pRb7<!UAOQtz#+|D&wwQ57oGBpLuU=
z$9wLp3bUB0a@2Yuk7WhjLD_-=FHXh4!i7Y&i|0Go#llw4vZ>a=j>yKOu0`h~!p>_K
zm!<O^aWd74cX9FB6;ni>$RiNGEHAN0b`+u35S@UXuR<92nIw*V^bVw$@+h4TcV%Wi
z2_LuLP`x?OL&;<LOJ~iSsI+P-FR4g!MuE1RDhmYOm0L!-(bB$`9aYHeUeFk3^D?2H
z@Hw8MR}~L74;%WPMb|$-X(ONUmh!G8r2p}Ay1XK2lBBU%h4wmt>s-h5qNEqXswfn^
zlh>Xw`zlgDZ;2aBJ7pyo%U$;eTk4i|0mQ+)Dx}dXx^9P(CUC+)hktlv?pbO^n`eyH
zhaU>p*^3h52EIF=G^6rh9rTlwtaP)ra*WBO!hg<^@~#^(IE0~V%o+Hl8NjS-f=lXM
zw~RI&>&_<>``Tb{+^8t9n-ejKq<~>IlU!3-7(Ekqe8D&yEY(GIh>gM*!5B4L`RPXC
zP@o5x*Sr*Y$C$7hy;ABD;v%W_P*bXPWN|$!ALus)eJtZ++sL>T=9#9q$o|dn{h06S
z@nHy5c;7E*VS(E$!k@};rXyp5)Gu!L<)hm16dJ{|$_}k!v6t0D?@8Ibhcp^~b6e1S
zWO5s0`%28`3X&sHEF!12t^5YyK{sbRdcHj`eOE567!_PKp}F&7j|*cLa~cp_uj~!x
z6O#3fyNzglMbZKJ`%5aCj&^9y3B6f%8{Ir~Gr0|`%~d1cZkh~j4zO1)@#MAmeJX7?
zbsd{pAb4BYqjcDc>GO+6_K)qsY6J@ye`o`00Hj*cYlZEye8oKJ*BWNpcH~E528AXJ
z@T!SD@!p>P*oW-m=YU2X7M1<j5@p)j@92`OZ3+TUEET&hV^sN59r`>E({ns<`s}*g
z1L*SdO3`MXhM(8nx{QV}ys2bi=hxsP!I2S#9ny6Xn7gd(>D+x+{9lC==Tu2^V~p8s
zc^Tkk`XBLI{a3K(C5R#d`TQVMu<)w#eaqA*dv2$Stl_$x4rM8xR{Hq6pVV)2A|hwF
zg0|nLRu)1Q`eg+{^dHh;S~k*jL3WZ+6BAp(s;c~S-2>~#+2X}NOuD@bxzLw&KvztE
zwD+DW0ECnssiy?ZCU>-3w(rD_zG&W3E8g(m&u(aDA$E)2^AZIZB||_>Bh~*$hO*p5
zKOUl5Bi9&RK>yxbuFM%$R*)jp;#^jFxL-a_p)a8e;LcXPosT#zK4#x3L9f(;6t<Ez
zFMOMPRYSCz(3?Acxbs6c4F)a179s+V@K?Mjr5bO)foIEK)Ox9QGKwbK%;dNCfQ&4R
zh6uS+yvjrdJpWVu2$$AKEjIj2Q}`M)udqYkVf_^C-8y014t&roM~!FFzb%@RPUXh2
zgOp<8>XCW4GhJxa%)q%vuGp*{&4F9j6!mTyqVK`8t~&}tJ(W^}Z%JE+3)bv6ev#$X
zLx?EUNsSK#=^!H~!RI8$xWOv2P-7QH6u6;$*hOoR)~ZP^pJP{)T2x$6H;ExnNU$op
zK^7Uo3GwjCSyZ5R4y-OJr`f8U3KKmoxe2Y$Gl0~FMl7?jrG0FL@-Q%1u*<vrsyA<q
zx?rKB9Znm^R|PskAz1U`<|H)|xf0x__9+Ptg67L*nGG{xzh;&<b~9@1E_;5pg5rNa
znnyVY_}>%u4wg^a<88_5xr&J91q2V{Lr+Z~dr0%*I)V}x6vDp?JxYJqeNfcNjEuy%
zN8~j&y!c@hBt~o<v%(ZsN98?#tVK}W9TLZOD?QZB<*TC=RPp6kpm*HqZ#6d#JJp4L
zxbQ|3j_+ry;I-So-N#a9DTZvl@*-GYv^_RbPSI%{3=gZb8o-tkf<``mEla(#W#{(x
z^L6jifws8q7qh`gpoP~cVp39`>D2VKJM>gvh*|j-l`?|_+VP<SXSq-OH^gt(T!tiF
zXRY=92g(L-YBB|}T3(d;e5SHL?j3V7oL=ZoT5{DE^n?cm5tN6h*z_(1rvZ9Q+J9{(
zMlCjjbWd{g?vVv@pgC9Z<t=h<<#8zmXvk(Q=$-n(r=$;HOc)1oAJPPA%-x1#Nt&yy
zyH&34awX}dx&5pkstz%ypPapXmX8KrYYVafDT^&)sAhM=VY1k#yJNA>GqEg}0>|Iu
zv=we|J7O16X@wYy8)FVhOEA_VvGPcxZw~_fs&d?R9U6r+wxrp&cr-eUT&0KY`;JOV
zjZSx!j&<&fnfOLFVq62hyB{XeNP~QSUb)$L)rNu>?b;eypNHedNopbOWp?udEH2HG
zHRi*NdIyo&{$jL^aOm6I{=D*BXk$Qq;8>`_zaIDw(ScoA@U$Uav>4ocpoH)6HF;oJ
z{B}M#x65n@yMCxN*cX0IZo8JEb$GTNRYE*b3ho?gelaJXBr4m&z!bV`*-igV`^}1(
zS`lGuIO{1$oYw&vwHD55>bv~wW{N|{?6eZm;z;@BxE!P*9xSfA(bpq9&hsj9*RSHW
zqhCViL?9)*L13U|;r0Eh#wKa}Rgh4bgyg9KL9!di>b!;vh06}~INMbeX5>%^jMn8m
zl^F4;p$?7C`+Psd2s@PVh1p!s(@EDRgz4@3?m^`${Zt5pxMWK>9}xRPSv6K`J}nGR
zSxttoQRFsME2R6!B^^Xs5&WwtCawDqMI$OlBhx+>&;%Xd#Iw1DzSAZB4BVgo+(&W6
z7%a}caUom(!iP=xdWShoXk!~p@|GaNJ#x^zNfOW3+?p!lp*EH*#-IeDjFw1{_D=KQ
z?Qht%lR8cn0UxOpq<-3kA2A_SWHUtIl|w`S;2gskZ&LaePl-EuO|BE_rFa(20A_9V
z?muZo5TH1AWS<IA3c_Q6<h>S|TTcx%xL04Sh(+!Eanhglk^$L<au7Em?p_{SSc;wv
zUue1!GjgK<8N6f`%u1DUi}r1k^EF}H(wib{T+^-aM@&>XJO#FQeX11q1+oegcbh7{
zWpsj7%;$JOHF@%Evx*kIVsDLAm9&4SSZN#hUK9SZ7HwqI$uTssMdabg>g<@_KAR^i
z)J>pSwW;GEt~oi|Q2uDRc}5OG^Btpj7ARbND29KxiZM$ccz*{5Arg0K*4+-33mYS{
z1;nX+&31T-u@~E6CL9o2diMJ+EJ?`5a6zK`u9y(yc!<}21<%|U)gly_OPM=Q)50-7
z?IgB&o=O?Ydh<<^!{OR4P4*qDk>RTIS}wsi6S}6bb8pkP?XKe&Gn^!k?|JRVL}Vj?
zh_Fj*(WV|gy4r9HT-J2Nhdat8))wk6bN6@siX$Uryl)&=N+tUH2VJ3%f%PX<{i`7I
zwRFi_QkYH>O#pom;X66W&kY|3WA@Fs$q+q9AGWEFEtUrh^>#TyXsG*#v;zJ7eLw#3
z*5jc7S{TE>o&?vMP&WE0F{XV>!LI+fCf^>T$qcZ>_nlM<06u%<%QjTei26Nd8QkEs
zBfff|DkqFjZ*_R4KIeo@ARO<$FL@3n1*NY&`Phg$Nu@JW<S>&=jplc<u6s_ZUnq=w
z6bBQ>A6L!v3$Yd@J@x(5G>lri_65+e<EoXgr&$U$!WIG++}7w3KAN=9Y1=k(bz^~s
z`%2ki?!O#;!`_*9rYgUR$G!5@XG}%{(z<R5CfC%H=7dJ}h%j@!hq&<}{UwUW81R`R
zCw^UO(AE7&iBYt8F&fk8Q%b#G7RE4Mw*7q749kn7`z|^&LUHD5>JGgSxTc6qjo-WW
zx^izV(Y)zmW^)H_UV2p?-MQR99r7y>jD8Fx%`BDD-2Ep{S`%uIDYzoByDu_QygctL
zke*mD*UkKxAu9QpR+9v~MHaOIZ^321+4d0Z1e#W@>L<DOWt<ECz9zWeaaLy1!*?T`
z{Oc5sx<rZ+%29xhprwJ5U9F7(Km0y7XME=qJW)){@nnoL^aT{osYRy8)0zLsL;7tp
zbw4j<KTip^Ww_qFRD?7pwl${*vehbnFy|01CKbfBR&aU2w_2;(`K4-siWVOry*u-a
zbo88v-#X;#ln(fb%gbwNt&DmaU_O~X8|3n6eMyB?4>aHMuIlWZBWs}Jse*vUeigD@
zrQ`EEf=B>odQh1EE&La@dkyC=tOOSzZ>p&rkIb-sob8-c&g^L6F2#4<5+Oe>rRpgc
zwgkjG9G^7_f>_mpSbytsw$mF7b<Im4<PChn^2g|5h_TJkgkYXgXEJB;w31+(rMah%
zj5<GMYL|5}>&3JV9RW^utAIE!a~hg_WO5C*(>NsBRHruI<#<SbfgEN?q4TgRSsvZG
z*NEi1%~_h$U@9YV%coVC>X|H=GyB==-b5_B1TcXCj7}y${Zt^_82zK>iWDEGd!b)N
zyrZp-TM#35pBWoaH%W${l&Z-!k*}l({H=}Ep~G9Iu&pwPQHKj|0gW;?jg<MqM5oKA
zgNoTKTvVo>;1U&5=<7~KdIyo3Tk=1D#CMnWjfL6XI61x^z=4ho3>+ptHlshYqtTMM
z-XV#QkW51^x3<9)u<MR|=i0Z#n-{lfrOGpYi3^kEwl*uk^X&{nq+kUkFI&vW6Is1d
z^|{CWj7n7p_b{&>8h@lP;G8i7qyt@HrmuHzNvScP`^v77HE31H{PcaphCA-}Vv`ys
zq`7hMB`*|zXPykmEl;X9!Q2Uzrj<}ww}GwaRB_G8x0RgyakUpW`x27PU5g5R`Ovzg
zZjXD!$YcvPxh2JZi6OM(U27N7qUpl%BS+xS(%!<D1DE(BeTY+`_t|vQHGk_9soPO;
z1DuB<)ILe&0ByuU5-NKs$M5r7o3E8#`6oMZr+N5jXO<2t-+(kIrBR_e>Ggq^95(=?
za$@nPW0WUw2br3^+U+eMR3_w^j@CDDk-Xn((k@I+Gzms!fC#I1bE6uD7&5tmhBRFc
zwq+)lKM{3$FoW=FT)e2iF+Ht>2&S+)y)|4NM+RT<NHW}u8u;b;CdQ<!gz5Q5uA#=L
z3q&==JF@1vvo{4SjBz*4E8YV`SkMKNXE1B!4e2WDn=f)TW)P-H2mCv^$6QnZ>%it{
ztt?60!APq$@@1DZ>wb)@QxP!au^#tBa+G;ED-5w?HTz#$xh4D+&2Zcy-L1RqK_yZ2
z7keWV4LsnAE7nciXZ4_A4XM_UDxD3ftiS20aR1rgv(rsIY>9Z~yV9RAZ04f$CZ9;I
z^7tmb@oz#=0~AL!{zi^3w4JJz5JO(%A9NjuB~{|e39mkY3{-omd)2p$ek!oPME#E6
z9q88vMY?7o1BZ@D=9QfkpXO%dfBH0$MJ7nkf2GWwz7izzvZIWK1)XLC%7t~ruW%kF
zKg_i)UU^Y9-<p%aIvzs?Ov!nGRN-z0ZNXlTTv&c5FP2}esr_w%dM1&4sY9O^a(3Q?
zHc)NrSZZlzATCe_*<60JbTr^JD5T~eN5jo~i$X4X@(#+=WauJGvFI;A?0C1eU1COw
zu{BuP1O|puZ1!$C?sxg6{Mudm(c+gF4^8xg1?J%TyrVAUK+jmI0kX!ddbj3_#9#dc
zBmjxpqJ0#Tt!0$1G$ep|Ej{Gm*8SXUb@nuYKe{$0CrTf2K&pa{{z#sO+KBQ?8U6vl
zk?!Z~KkUasI}%x(SZz6T7{jvOvNv9yx1?2Guf(Pd{Gy&=qnEbZwpMw^J~UMmIK+V4
zZ4aKcw+5zNc0&9)EHM`(8*KDFG;HnixB|2&0cAPXvAHB>YJgTS>pNc8wcF8kBRB^<
zNtlB=l7D5N_L&b$Z2FdLODVn=enOP8o1nntfG8BL)UYcBtAw%MuOeQ%cG}^>QC>)~
z>4)C{r6onPzgpKF3TXlHAex3Em=lDzDYXmfQG}e9c(DWoc7%Cw_N^Tc1I8$WR)u<n
zt4m(;F*OVegQ2PxQ1`z+XVO*OxkYoDv_FmIeRg8h>$9!YR*Qws0=u=hZ<SC)R4CSy
z9dI7%HmYn^cKerfVh{IwXx}R}-tN?L0`c0+g$|^+P}KKd-6=Zw{%4uUSXCN9;=nqO
z4>;x8vM&GaAJTsSGWsFV5b3GUHla8LFT`zVe!a5;5Mtj)kG%+rD+jM8&6#vG_mBKm
zQ!d4jxx@OtMGEKGbf#K8?w+H-&xCI)(*OnH*-(S}9rhd^Pl4az;!ilcEu+yfb!WN~
zj}09)!>q6)lC6Ur-@O9|=^X7ZgT!Vp;oEmx0<99Ot1<mQLP&Nd7EH{nMdw%8HTr%9
zpNB>Q0(8x^XZ-Ci>D<{=E=N<ys|*Y}zm>BPdetoUYiT<hpvkPTK^($4+j2H<#WKhp
zufK|H$)w?!EKEUn`wvnWyoRX$%-q3G+%r2H9#t)$C^-#S*}6NZGJZ`s`oSmeHM3i3
zS%VW16(Fq;sRr79F&<@sp^Fz!t;INdUGE7yxO%QFCgshcX;%E?5ruNQn9|Cyttjn?
z|EUE?vM#Rv07}mpTAnTAcRMf!df_nV1B^JgtZ8LnU8eNj#1K{UPH!>#p;kYI>VDNR
zVaB7;Ot={ADk@^Z9{)$<g3Mjm?OV3knuBvdvNPY=BPG%=3v=s^Azkp~<8Wb2%a#)b
z#w11+f*uV$5xQA(3T8gt)fvf_acg}X$iLSw+pM;@qkJ^fy#*`#Dyp+4<#b)hTO?cD
zsVgueIR(0$K|9HbZQCj5#)Z2-&;fa74eox`e)HSK$NQ|{RPkzFU#By{Qu`{U^lb>y
zpD;~M=aF3*S^7R<^w(4{X>zebZu8667fFvRq(SYVh{}W{=bK{2(rC*m7uEdv$zO?0
z-eYo<ykAznSBGX#e4<)-bN%qc6^1ilO7oSC;xE6;D4=8ZfkPP=Ob8<qSe$XiXS~UB
zu?_6?dFjqa5Y9Fc#Qu%U35_ICha#P?n!>8bA(ik*+=q>ESPXzZZ~sG|$zzY(9)gy$
zYSndLJL6?7r2RO-KPwY;odX`p0AV2`;S~w1I>enZm@rA&?-D?Nd#=LL*C#|)9+Ls8
z;2#`R4-8<&C*(9mI-wwegJ2$8#vo*KD4X{c1HN6IAE&G*n8GIMMjj^!SMk9);e^Lp
zRx(ZCcE62=^A}K`V3pKJ>7VOj)9Kmhqw{;LFT@%W`fq@Ym=H;G(6fmeH1g$3g~3!*
zsB)=5RH+sd)j{|ZQiHiUt^?F1q}Tn-`_uQ7)~H(rFNU;+be^G?J6H|~Ydx5qA<nL9
zy-yq40XIM~3BNbH#n^eY?{PYRY?uJ2!xC}j`R?jE%3YL#9hJ;{?6^bd$>@0kyBXox
z6Y>U8oVA^&QxmMGYU?g0Pf^LdC*m6yv5WAMP+LdK%HpoXF|8aaObSs^do^fHN!rSp
z?-bT&OTn20C}l`og1ckja(}UE!_w4SvjJ``Y{hklAFupL!;rJ;kuSQdo7S=%%fZ*0
zyeKAe$%q+q!@tNijg%)|#G|C+{EmRB@!YTs)8WTgBwF$>5`#Z}^P*7V&+d3#1sMg4
z+~qEds{^!VQi`m_*AeB!mO1Eb6=Kd2A`zes3`qYYv0h(Ujv+O`VUSoS_jJnl6FwUE
zKS$T6+tsofVEJ8>v+FqNc2i~vB91J2#sV{K9&$AiuY4ulO#A`^Cf@Ci1APK1ESI~L
zwgUBi5YD=lrm0X)tbS|jMeE99=TT}F8iQGC@r~@fLZXYXZ2-jDZw@&8_l0Czqk}?&
z+`MF}{g&g)AKjPZ#3PNZlxK8V=I8BAXe=ZnG=_+AoOI*2^~>iSFL!w^esk^7fIf|=
z>LPG$X3Jnx0{@DrZ`B0xv$4L0S`sd5*}#XL&nv-%>ZNVF{cW2E5AFFkc7u_;B@AVY
z-LDK+{n=jI?dJo^-&jET>v&^l^Le(sN5JbS79>**lO_0jG~@d*Bm2Mfv|j05Lz(%l
z_EB48<{5{|k;+<ChbZI6$WX|W-9hdybm1t=_-7T}{+*3|YNxH|7(kN@{Nn(WEk$!!
z%f+51k{uM$%&cI>be=z}a&Sz*e5EA$ai*{*+f^Nkz_$$2N<u3!616EJ_cjz}p5cQ<
z*%@u{VSreuGZd-Io4Rg(D;%ae`VJ9Un?}QswX^@%3FVXQf;#k5i2o*fw>-<m=W3WT
ze>H2uXeVQ2-rVITw;!pGgXfPe(s7I^20$s_fQaAq(WVdx(wpIpz7EW$#IcgGa@_y%
z_+A4a1kteO`#zpS4g|1u;${4zu5PsCPf^mBGPoA+!sC3n6o+|ttwGA;)+17aDaaT9
zx-TZ07tJbcFujNNO#GR^uo57z)jvMW#g5(8WI^|Jon6dAu|r-IJm#AXkAXQLGwC;X
z;xRi8rm^J-UJcyoGE!&ADkQ6swL~t<CnsFzCf6j5e7YSlZkoedB0<+~vG1^PeK6qy
zdz%zGZf3V~g-5HIK+cx-SI&dBWbaceh*^Xgd|j{0(Y#?`vA4OP#DXhZ_IxxDJe0|k
z(Tdf)`HxNYzP9MhP!U4NDhFXAqH#Gb&C+MGQ1F8!G*Oo0Y9rffHNWh&AoQP<xp5de
zEu}Yp#4bV0Nzm{7|03=^qng^Hw$T6rN)ZJC0Vzrc0TrpCSx|}sN+&c0>AklAg1|v~
zml_lh={-nE0BKSKp%bdq2qE+UA#nG3zw?f9@6Y?=j=TOOBV%M|XYRG;n)7*{IoB;w
zgICu`t$lR(E(;%#U4X=Z8L^(=`_ydy@Xl#x)|$i*(DsZ{wmLkzJvB`gbeiz0nc_Tl
zE;yZb>cWPBHS|cPDtN=6>hk06C0u1ElR*KQa~Zs&vl?~@N;v39e#5Un*<;`zld!VE
zccUpulXSO=h^V&-@M}!DO+5~wv)lMiVG(lU<OH`}Ixz4jBh%eangg{g7`J_DZbP$s
zxwE<Wi;tr`fzhkkNa&!L;~$_Jc4-2%flj&+{YKv<Mv$B!vFdrbhdI44+_iWf$?FY>
z*m$LP3pXy0O^Apz-Je1w6T|>K4bVy9WN2q+DLp|%xlCV}R=^08pq7<L8+1yx1d*uw
z)>Gci<o7Ex=QmCB)=l$7QGliVTv8zCV$}+xTrhaG9CT{S@Cja>%N*x?=bh(Y2QNIe
zrGAKh!Rg;kXWQGd+IuO3H8Y{qEM$$SD_Z8bYdJD-vX>gfN=*Qvo;&Vz1llvWNJ>)5
zR;Tk+V3Hix2*rh;pX5x8m6Oa39G2c_A`AIz99P9ttXiEwO77s@@$WkuLZ3w5TRch*
zrA`OoIPXxeJu-6s8RUTqivck)fA5L^;#N-Q2Pk_fhY_!TQz8#!H(=aCWa9dnL{BKc
zYrn%sJffqi&%*!4$<fY&XA19!7v`(3x>UtCS;B=q_cP12wATX^FA=$T3Q1F{MFOlj
zg-juIBF93F{DR1N>3Bm5PN9vwem<{=pz#0Lh54vmwLvmAX0X(w>K&y7s)2Sh=}kq8
z5#b+y@9~BSKAPh)?tgCn?l!H>Ea?0K1WTwP>ru#Salzqzn?h{~6dRT9@oj|J8xnVH
zD36##16X2Z+G~_Yj|qf*Jms{#;qB#vR9VPwH!56gnq$z+Gl2Y3l2p$0@XK`j_4?C!
zTBoY)KCJ<rvORz=$0M%?1zTy~>!NGPRlVc*{UO;w?o?=VdV8Cse0Tv3bRGrpLH<V=
z#CFgzvPx%o^u>o=gnKL9Poo)PKxgQYH#P9Voyz8DN51F%>0{fG4OyY34p*+5QB`{T
zeqQ&6Sun32F6=n`Q$_E3Kr_f%9sN(&zkL8AGaDq?QWf2cY}1>RnLL`N*Gui8onOqf
z#DxMj8Cj$1KXzQC>Hd!t(WuE={(<+c-F4vwS6#5g@%PyEr$IQ*_Avn_MrrVd0YhOD
zb`n@6yGx8b{7IL=cp3wJ>J_cnD>3#xGasuIM!*l)e&>41>Rx1$a-!JnPiN@5Lhy0F
z#rIkM8S7Rx__v--RVFmzNwPXXvWZd98Te6O%+rDJ=9c&Vc={f0kF90x4*aTB6hXMa
zADlo;`#^?14wni&fa*$hVSIGGH|S@tiYnyAqumMWP}T&u#}N~R?-H-CFA<1DBpy&c
zPjyC3xqbR==J}B`SYGAn-ny{C;hE+uvR9?$u+FMq9#pTNRkZH>by*4ddu^$V!6P!z
zoj*KC|Ml|FcPc-pX8u7DZ8rJ!_&Szd6S)S$?FMRA3c*Sr!Q$7g3BZ=7bFsy66#KMk
zW*eZtC;xzgTbcd#-33te6tY2V)4GF}*2O7gZAB@eN1ms_ca+IbX2XFBCu!e7PUSH5
z$V=$d<;M0o`BK&7LqBQBuII)%oPO`+62f0<1RCaIoVLZ?sQeY^cOOoBu%8%(ig)WM
zJH``@Te?1gASh6vghxEFwr|@gc3th!bm(E%0<CDqVr)=x$6@PI$))8ra*Mf*@VXuk
zmnKNZ)nJKa=CLH3(3Kr@OL#Y+8{F7g4-^^aQ;A|a!TJP8773VKl5CGNAx!y2ylyf}
zhYtZNbNNg6fK(87{4479SPFCS<k&T<o-Azp%gJg<S75IzsJOaC>Ci|)+ma&AAij=7
z`rHxP-&7maB10#iF4TD_jbc~#h6YVUK*}8E3goTOX4jkOtzLiRZjzKDKF;lhU2_p%
zZJplp=l_Ef^|-7|M=9UgpFehP1UwKqm-(;$R`~T{1C-i&-|@$oRyF02T(RZ6&p=7*
z${nDjHE>q6*T$1fJ2X1{=g+*D0xiL6Y-*FkkE6PZM&t_<p$cSPTb-ps3JAr$-v%`h
z38%L0pUK)unD4nMYy;HBS=G&)UEmBWNahcKb_8=lOTI*+#0kR#J007HR~<hP|IjQ|
z=yo)=0QI0NM~`1|Gbaf1f{A(xk?*y9L!F3n;^p(qA2Ic%@XRMWFVo2{j}3MU_1-Hm
zQIYuhlUU;tB&an+j3+i8`Q3#q{?d>vqrE&O5`9?r8V*kB>12akD0cC#l=$%n1qCts
zaGMmt1KLo$((u&!o@h`W<DbxIUN)VFlIU+0;hYZ>?ys@a?s&7nqnc#-+uK*ldZ%AY
zj`TW5Bg=33<U7mU+-6TBE6<xH23pmI62z(Vz;;Rt6Vp9!oPZ>Ya6a&Ih-7XiiqbXE
z`p~NqZAj2G7HfmUR-SIAysQ~p>JtowwP7T`^`sDg%0A)*`hf%uUCtxgOGFi@%&M$(
zl$5gx;3}Thi|Wpdj~VfY7v&8z(=$N#7#=YFNr%lrBWpPikBOC*3Fo$*ybj|}io5sd
ze4YmQ7(KuDd()YelLwG@?}~`sB3!O73-=U8EQ)P-+}+F1L%i(q-my-W`|B_ZTIbx`
z7W<wjSyH0lG}_^_x&Zc<)qefzAIL*`{VR3#7R+wuW%ukUbtD^tT1<R%tUNkC&)QVt
zZkFGmS+Z*0db(8eU790??|Xn)JlH-Jk0U+0eeR00B!4(3jykFfOpwjWMO7VZDOm_d
zD;P8-SS$v0o2)_KMBuNec!E;`yP3}+fIxCq0Lc(kyC<cC8T4-6_Q}Hp`sy1c<@T@m
z=5@Qr-Md%w>R(7xNa^~?u7kPS9kI0&y6iYy#()4tiF(fHUnMq*+vAe4Nas0Z>eSw^
z1He7xX5tYi4o~mwW*)T@7U|~9m7sJKK}QwavTn_OZ&`<CT9E+Z1EG_Bh{P)`NWvzb
z156O&f#dr-6ch>KYkeM1u7}@4ac*<yz!C@8*1j5C*ZLi|{ml}zL2C3ptx8RZ%y~>@
zTW%0iWHfz2pdNCCEZTG~i}uL|@t-em6g&!bpP?I33gp?&rqzT{zZ7V7<+n)q=;cLb
z;h)4DO%SV}%;Szvy|U;?u26dm=;_nzs`~zL0_))uG$)tc0SEM2SmtXTFCzUkU#~t4
z^ikT#)@JXc-7c2SY>T@@EkL6hy)6jbm!P>@s&DV{zX^QPnfyZ$VLel;@P`u<Er}`g
z<^bEz*1Knacb@$|?C?j{9$n}@S!JKwqt%+BuzDAQNp2vRkDC|%{J6*+b%V7B!K1>+
zik~{zt)Vmx&ERu$$IER54xfb|OV+bErbqHVfhg9#Z)|Y5)-ZRqF7x<KV<wZnw~uMo
z16d@9t$7MjS4a{psJhjUf`j~RA}Jde7F_D|p5p2sowLMxuzw*fg?iP>sk4cJW)^6-
ziWaNlxhxa{-vn6K95#UOSK~AEaoTJS`t=nkirPmNVMoJQu{w#M+BBOQgY2fG)xW#y
zSj!O;e@<n*5>J3q58lOywUv8*xWv7TXKMDcL7JyyBhmAP`q<0qt-o<Y$1`K{?ro6t
zR`<YRkF8VzvC`lFOuH1hN|dy=aiW-(<LwR1xHW~`&X>H=k<VCSt4f`lLR7vYSgLMw
zC@rUD-ZfJmMUp=Wq<002KvAmu4xLX3RfLM6YgI-amy?)Y$*Jwc0qKulJwturzDk*r
z?-DC+&<M(uSy|2r3Qv8owwn<L8|KRXEV^yq?cx}{>LBZJu=o!1Ml|tk)6~94ZR1*~
zxm^11s};yE@`H?MIrm{_3Q*NOGEiYn?YDTrp<Waw^!G8@h7rZkY?*>B>kk}-<}#gV
z&DzmrUAYyr$@j(55}UZ%qF;~%ub^#29WQl8^78&emR8^QlfQQnBZ73v0&_NGCrle+
zZ*&ai+%Y9X9pT}{Cj1U(CPk*UrPQP$=+33B|9g|-MTKsI`0|C7^DvQ{F4;E9n*Jfa
z1HT7|y$dgD7Y9CGIr8`I6l}S19#zapY~`oiYcLY(Z@gT1&?jO9Gx`wyQ!R@ez2153
zy`_hZXS<fIxG`I~GU;$H$y367-db!h$iqL0BYdpsniRbO?tEAplF9rwapl6uPyW^A
z%JcfSG$oaFce^YQ4OTjro5omHeEsWBk+0a_i{_lI3c6S%n$BdRGhEXzs%J|#FoMYf
zbjeEo5tNjU?B&ebL^%H*AI!%<9-qzqryd5i(;-ky&9*_4*W=;Vr04@5g`uHjDuTB%
zfsaen2h*fq-@o=(9?PGBI1KFmZJfbS{hq3^tR&!b*d0iX{@jiqx@{z0Gy2Ph$B-Y^
z>R?-Hf7T|u;s*8uTbaJTBR?YR@N3`e{^4OA?_7lK(+P|fl@aJzW`L3P95r0c*YhKI
zIbo=R^20{S*S`cz@yuz6cP?j(W&rqdn?U3{or4v8BGSclJ9CrzJ$>S_UN3sP<+0RN
zn3}#g9b)!S)VyM4gSi}>=HNRGLDYAya<Z(XaiSy{aa|;G+ObB`9CKgni>3P4#zr~K
z?S+sysB2$kS$tl@)K31N^aDen-zNxE0W`A;IeXsVYTVrEzf@KJo2pJfukNd11HHk#
zVsJi*h>?i2fVwTEHY}Z-C~UbdGFQUva1Y<>mmLi|V~VA6z9|qt_XuiwJ>Po5o<NzQ
z4o;A1l-zvIl3%XY?AWZ>eLDLkTX6jHeqH}^Sr5Fv{_~O5E6(*G9wX?peQI1}O8K3B
z*l^=1gG(4^0kN^dp#1jxjv1?sUxK*Vzzd2ptsG?~By=Y>--(-vB%=_#Ag)1N?VTEs
zV-yTmM@^|M`u9afH!8yoCX8Ed%FI@wlT89j1vQ`rl$~77O>6M*;v!eX4Y`lC*u+DW
zHb$fPNjLmnKdexiPc3Jgj+i|!J;ZoyD?Ynt<#n0<c`UyW)|UG-Z*AfzI%EhA_kmG9
zh8S89tvgmKO(Sm8!EbSQY1A7|7?VYerQNDA0Q7l2c}JUnR-sD#&w}9%q$Y2ZRqH>B
zbxuVu6^fr^28+@(9oMFN)~D#BQ_2{nGt58cYahj-xhEsF%hMRD4_2RNKJL0$2nnvB
zpN(S)6%$QX^3)$_BMmuM%Ja_IzP;C&@m{cX7VO?JyK;#ApdMe=ql$#f1sR;3)$Pz0
z7p8lH8tWn|*<3>!zbOR9`<eJCwoDpC^lpVypwF&KD$iZ*UzZHRzHr1JBMiS)MV|b2
z0!t{m^(QyrQ0l#D#$Tt^8gXAh?cr)5(9hL4F+kre>wf%=vg|$OPl%|fL(Bam>ooe(
z*vSB)&AR;c5yrAe4S3EgY?Vtj?K1DpWRU8qf%Bwb^pk7Fe>p1d2}lfx2RvmJQ-XP2
z2HD$jJ4W5WKN#)zgwvh$*Sy7F9T-a~61HScg2Yc#j$~uLxmqMTKA|hFm1Oe0xi}w`
z9GPL{SXn%&CgJLJ{1Z~_(>f#=h>+#vBN3$x9&(@}0{Fb?(fji()(TS!#n$I4v^}4w
zuv`jhU;2mLN*%=VA1geGZu$zf5(I_SBHO>NXw^>@jZHZd^SYB8mfV>XyxwEplx^P^
zuKqNry%>uK`17=a)S8U&Wqtj*s`RE=m0^5CF^%okP5b$aE@;)C)-D7?=d~+1S+qWG
z@SdF8>~M5WKg!)48${}w#^}){DS0*=mie{_J`@M)uP;`a(dRWaOi?M3vW~7(=G`~2
zx`tl&h5T$G?-O!ax1a&54fk<iwu_7BRF_Z(V>VM09?}YqCDyE^ZN=xHUu=f8MC^@P
zev{P1*D-i=dAS-pINwThB+(6~J{tR(x0*x4uSu~DtSd%3GjsGjerf3aiX2$J=q~Tw
zMO&1<0UwDFh_=tT7MeuLK{3$S&WK}Yy!Vuq_}*;vXD3Ye$4e0lzh9jS;ngf^u`wZj
zB@omvk8Dl|^T!%k4hgl8M~fz2WGeE!WT2&F=66&WZ-Zxdm>*Oisw-Ejm#>HUpijry
zv98MFP52Ex`;Uj(zG|*6HB1|wwiWkWRlZUxx_r4s{3bL5=WO`9g?kB^h0zJsr!mrs
z*MVRtTU=n}ZC*v>+gYy%$aDcw9|*bScKgn?FR-d416<MzLdtOtodu^{_4N8{*@}yd
zZ9)nx<!490d`Xasrt)~8jb_tKO_a^T;~6TJ$U2XN1H74;ptHiQU?Won_??<|H{#MH
zky!p2r@f8|#`})4^jGM-uL}xoY2sP7J`qd2Ro2ix$vd1PuC*9W`>T9V=yy_Owq2bi
zqy43D;AO=;9=YhNU%|Q3cqjd0KeE!kF>F<BwJCBDkoSg}W=D)7Ut50Td;cTQw#m6|
z>{c~$ZgjRmeUf$2-g}$b>b?q*)V*Vy{`01N+k6h>{(@ZunbzEczU1l<Bdd<TiLcs@
z?zP#pa@L8DPdhrFx5J;sgg>D*nv^nWf0z`Me4M+TB@loUEh4UB2IEs$w>L^l8>Anu
zP|G=@;CqROC;RcDF|-~<s!T;&g2r!NJMvSeo6h(Noq0yhTFQ|)8Vw9qw*#NodPO13
zS>T#EB$u><<M+;2iqKCTqc>h`1_e6h9JKRgzuD_5J>iDYb>vrgDFiF+Ys8%2(OlKb
zt2^khQ*ZnpdKlA)`}wF)`_b`j&|3;%b4cE{-f|10uKr--wcYslg>L)2$_!^3d>g@|
zu|#^0pzK0T7=;FJhn^6h+1MJvhB!pF39NL-IV{Ym$~qOhz}{hqXuMQlk%V~d+Fo&~
z(`leB)h?&J73<xW_!&=wUQe=d^@$_&jsfeAzwsnTWBGKX^e`8*#oF$6QjH-?m&bq}
zZLo{{c!MOPiK8-pWwONQE2*nA{^McTow9P82An;O5u<RK{)m);P6?vZmN5l1ZEjyR
zrN;b2-y6+fo^s*w2D2O!{rIQ0W|e81_F2~&{FP!}o3nZPp)GBWz9PKpSFE9s9Z-jb
zY44bckzTZ1xm)J->QE$fWYNk~FsvREZ=!-tsC+f3^mVIA(dW5ji0Qz{e8N!L1x}7y
zmBM{<3PZADr$D!W%ZC<5*RL!uDM>$D+V>#9UK)$1m>aVh+}fLlEK}rWvPsvU^z`lI
zH3C~KZwOHp!j=?Uco|Mx)L)lAuIRH6uiB~Mv{W``y|s66Y~snjOFLvE?>V{-wze?>
zXOHwZxH%{PWp@_n#9-zdo!v{}HCqCkuVQpm&1W&lIiu!hXGiGWs<=`UYzd*rcG3jW
z1GcFrfuTWphesRVSS_xE!#(&HCdFaRzmQQFf6!gv5urvV@dNteG`uFFzu{9kkSE6X
z3M4zr8x4Cd!D`=ieo3NhJ{-K${HZq~r#T``RcO1=_=yEnD9Q)|ned&^xa=qX)(pz^
zv&n5)($z62x+PAgGi4WkZE<e0Cgg8<lJw7^*Ja<1gOmu*uBNOZa)&xDg$&og3C7)n
z;k3>wwzgrH$!7nM+I)+En0UvZvE=&``8WI|tQ3n^^wi4@lZKY;uTY$sxywMEjDVdQ
z3=$&{QMKCEI%iPs1b+W9Isg9BO?E!wK2^4_d*LZEncRN;DuiuGdOG)?u9sHrBCGEk
zU{|S%EX2~3psE{fwz4t?^(4ZlCuo>NHBH^wRn~91QO24jioFn<oL5{9WOT*hggL6K
z`*oddF5$IlDuf~XCTCnQ?JWUbKX%5y89UZLEclevH5)OD`Q9zEh4Ih%pG2jp5r!R3
z2}?@p(-(Uj(Pj-&)kf?~`wc$(Vf^<nf&j8LEfT>GDU`UM%$^N7WL}*n?7oysa064@
z`IRW*E#@s@r;Y^7Lq5?l6M-Q%=8)`P=T2DBMkDgmTpZ`+_pz5oNnM}kcbKk_XvOl4
z4eWY{Cexc0dS!zbpxv&pAlnJthPcKel)vU$#ql|^>w0y<ht9k9@5tBfIhA-xZG>`F
z+j`|-9=oam`Y5LtZ+@udhkQL;%P!<ba5dp3=$WP4?wa#cq8At6<2LU#7=HfVJWP`y
za2*~pCDj!&SMvGbyw<mp|B*cJThTGS-F1VOVXuJL8MoiVqef4z@|Yz7RB&I|6vbzG
z?yGRYx|noKaqs0V?>O{U_MEb~X@u$P@WWhI-bM;zgyF3BoVbmqS$z~Xnmwyf&`{=*
zw`XQIG8|1X*Lan1xLuyde9ll((iqbc(j*yR$)8nDhvjyy)2Q)F1{EQ2h_v5%Q(9}*
zCj2494}U8@*c$OPdxtGqr#1N5g(%4C^NTH5s|z%es;PZE9^_kB&7@q%XyqX6(%vB$
z=YH7_3uQ}=ER-ir#+<yRB&iB3YqBf0Q#}{>7V;BAvqYIOvxN_*I1U`>uXgAILe*TE
z>VmgtDE>a&mf718bTPuv=g_&}8t0^v?8w8L;o~|6w=CMl_{KxdS27ywrD)ldb<Wo3
z(T>@;FM+lFjIKqL=sc!(F!8^HYZ}dtFIJ^*Esu{wEM-jw@<_#nZA3H7y{KObw8#ac
zn@GQJt`eHk^3D`KAIet7zmP@4l2Wrjbo92^O!wz#ii!?6Sx`KI$nBvc<0Du{=j5gK
z@nFek$uZIqE(w`hlws(E1vQsli{zfevZ>>hTv^Lzui5U{XghOL{LVpPO&{Y3i58+H
z7@4Ma>C<?)aNQnyU>l;94jnd^nuI>RpK}r66Wq9$NEx+=vHuuUY|_~A_pkK3XG9HH
z%cL1(z%4)x_tupBiSGdmd$yCem94=mkc!B^aYFplm+b$m1xTJL_|+KaJ|KZjsH#sH
z^TYk{vWU(~VY?{Y3CZZ*p$z_Ra@NO)<#xfzrJe6h`;5=OUEsXN&gBbl-nc%X7}P`F
zixa&ze_MwSN;KzS{OM5DLK_eHK8+VhqHxRIJuP&d!y$jwAs+_qwl$Q>vm!j9#|!1R
z|F|gsn!(mK?(eLiabtzOhCc&Mf0ho=wNb9`<ZzvaH5Il;-dot%!N^IscA0VHk65;!
zuH18Fx;i`3VNH4Y44Q?{8uf7E)LpMcKxA=&8ob_vTx&tsRq%JPk_SZ;f<?10N(wXd
z9zG7x%nLKc4znFIqi-d01^iOl%B9h0_Re~2_35Bjjpf`6UDZaS={=B|WcfgPg4L9P
z(@DDGl%kyDMIw!ugAIj#(<|@J;Op8tCRWNv{*XxMo!c9C4CCoxmh@}}?<<sZ*e9b8
zM(hmm4X6z3zP?nYM|E1o;*Wjo$$4&W2tS#T%Yej@Vf*yai*w^15T8i?T2|A^Q_|<I
z4M*XyY4>px6Zy26(m2DaJHi7JQhX1Sb`I1H6Ou$CXXKn)Bg87`HR|$CG8GH;ML}NF
zfD;!temt6P2=j6OIJfZ6TWo=+yiB4-t%g<Odb)3U7T9OOKi?jIt;H@A9j)T-yH#1#
z>QYL~*(csVN&csB{Cj7vCP>Z4osd~pq}<#uGP*E-7g4QrX0;xlUg>Zhd0nCgCj6r^
zlj4EB8$d=ZO#gA>*D)(l45<$_5jqJZ)_2i$`+y;T-XrMi6ukt_Y0`AiA~L#J3WH&q
zfMdqQdK1O`ZzVpi{!NbEgAb^sHm#xSxYasTwyhy=fF9QGXP^pAy>XT!jDiRK5SMJp
zeyTI`A90udP5)?{=|&nCUBC-!8WDl;A-7D%cek5#*@D3`r+tjJeSa?^&{4L{FbE@^
z27V{jS7K;qW%AedM@R6&M&ZzC;axP$0^=zXs7bSZ+s!JXV$MIvzZWjvWQ=wmbGq$X
zPeMABA6!xCPCT(y=&qF?HpgoXu?r8ekNyygN}f&;i+$sMJmy+adG{hc8j>&_6gpy>
zP$}IyZbokDSC;0b1mna^ZhKz%HHzhhJ;CyNGC_16!3RyubJfvwG1dcp9!Q4*o?;?u
z1(Xj4{HvR~Io!-hsrKi}gE~9b!SFM+DeXjYj&bx#Q{7zN+!+bY`Of5eQ^ivg*q>{m
zT*hJeWNq}|8<xjhdQPmLz@`SljTyg}gMwj*VSJW7Ye<3w-fuZI?dJGvX0mx{#g(@Y
z3Jl{PJEqcqiZ%QZCT~FAfv3F;cKh}E!*K+ANl++#$aTTnsX&s62w>{K>FoUp3RHxi
zX#mI=J<c)L^pGf~?nx+*msEL5kr<d9j3|N>Por*XR6gIY@t-{7ODE;pbu(&SJ}o<k
zDVE}JfEl?h_({Evd)&2*qZefJoq~>LdkDiqiZ=!uP`l6ip9K5Mlh7qIlZ{HrAG?od
z@60PBL1);%U3BB7byFrR5wySKx$VkYSX>=kJiR&=wmLQMFDnFEqs;nby}wYd`|6nx
zJ9<JJK4m@lJ>$O!1R|!=28=ua?xNI}y!}3ABy@8b%JDeATD^C;cjrUua)u3EqyGB}
z!wgxo+?lqK;T5nbnD=-?m+#3;;>Ny<2FO%yauy`_u=P1R+9jgTOcF=H^ZkG=XDB_w
z%+qwwmB%}y<VD9Mpw{sX1Fuvw=<B<_F=ZsE50+j7sp<29K=lvns}3c6)ld~F>3Fy~
zJejgzfIewG4sN*-!Cp4PraTgsAFTxQfJ8io`kmZGx)3va>h|*m)z^wzzs8KW*v=%*
zBmz+<S1{6(q|fvQ$~7E&(;&0?=!e&Cs#gu%yj#wj9n)q`L1ga7I|hyh<DEEu28zdn
z-q=JQ&XnOUxdXu-11vDq*`tKmg5)Y;ONx+O2fAs>?Mf^37GxF~0%YDG*Ph>9oZJHR
z>E<Wxr^jEHgOp|>NX%9hrI%i5W@-##*A{IG1Y4Z5hNS$Ze#PEuvpDNg*hm{lie!2D
zhwn>j<~y>5lFoaVxZ>+6N^Lr4Yi|jX?EAIbUlPVAz*AIi;tST{cOVRrT=R?O8_{+g
zlM8Xo>2yIC6kA2yCJ5>3>Etw8@vXVP$<L3;Eb3(Tn{@D|*qzMdGzS0?wu={>-aeSA
zSrgQ}CrC!)QG~L{_|hafdd!>obi*SU=0!-QJ*y&o2G_?n*Gtq|?MK<pZneNW7#+X9
zI{=FmU7B%`rf~$#5U&rKf4)kd2Seyvr`87J5_evi$BEWWI1KozUP3>d6lt3->>X5;
ztDV(Rvb;)22MG^1F3&wX)6j&QpgW_Z!&B;NZI6Y(AuW%e-T&;D0`40T?Y+1jg{sf(
zy`iQ4(Q9D5az+BQSponmoB0n=T0eHO9R)wdk%8U&vA=j<Tt4;f+;Pdfb8+mH(Sdkw
zIO)}}M}fm<{6o`a-8X!d+w1VO+~lOkCuHCqKa0>gmXS87r=U)(<XhihwTgcCjgrz*
z^M;0`f=uXsQIt#i#mdod)}T&}^4?$}|Fm5D4m2jq@VT1w$o{2)&-N3A)-I8zKb*20
zbUlVi&}R|#KYrOf(VNoj_U|+PrTl%6%6P}?LCbqT^<EZPCG}@#(}tz{r{%TYJy)?u
zLo9Ou(rUoUL=B;(m-$#_hZyh7lzvreU!NwoeAr(z$*&a_I2Nh7$R$)U&F@@QZ!Fe}
z>eYHQrzq;LuIQD!{fgrxU_cMN0JQA0ygw6y*(YNOa&$@_o-%nVEb+v7l6Y7Pvb$&F
zAWKj4JsXI5HD%O92}YtwAUcYCQDX6xQ<vrSt6<v94roq8f6U9sHr!k^ux^i*8sg4Q
z(Sy}}Z(-VJyRqqMK265XDF4XAhR=tK6}89ixU~<O1eEq*{#hdB%8DlF1j_OpuNC4|
z#7||YYIl&D!cUr!O8-G;JE;bKNjz$S_ZKy^I%$rZVjRAAM15Cl^AkJ{KFw-;ot`a#
zso9Ho&^OvtQ2zS8JCa{517FMbo^@**eivFOIOPR>zf|Dqv~-f;k{Z10uxe$~wt%(}
zuVqmIB@P2K;2)t**w1vOy_2pr^b{<P^U&;ZUuoPgHs<AHHOA4)dhmx2YJ@=#pNC08
zr!ziZb*A`Uj;=_|3*qF&o#m}bD+iwC_dGj3BuSi$M5-|?iXm!in%17AK_vvI*i``=
z+#C+cv*s|mPuj{w2exe6%DH!3xO8fkYd{A<#HP}%@~nEHDRPa9xE%@Tk%1%0?IhM&
z`Nz)#{L~<w=CPdFwH4g~!eL*SMc;PS-;~4RDK!RVRw%RTlX(5*qjBlm(GP4wq2809
zo+PDsEDzq{A7T?jv`v|~Z!5k_z)fKo^<xB#y=!$(XV48y+oD2PhO(jz>3Yl${TEBa
zY-lO<J6EGCH&Qr#n@s?MCxKB|P&Z5VcX#5U5aXLToW2sQ!8IG)9EiM}8-+jRWpgR7
zp8DWjb}FAiPqzsnsJXh<Wxn|8S^SW~X-%U(zeyW?#N^!Y2ZFh8A5^noFr;<oeS)@^
z<>CIr^n7>8YUA3pp84Flq5kTFDF1?@M!05RJjA26ust{`Mx^`;Z&nhiFr;a0Sf2p~
zPpGQ>wi{AB&};M*Y8<OiO&E+1sz?#Cn&p4#S+_0OjEG>ul{Jh_z!~dBPpsP#M6;+P
z8H7?VPOn2=*Sv46u%^#tMRD<q<s_a_nu8Lr0zUrG=z}UcBfG4Km!2K*E;;wgi%0JN
z=>MtbmSxrQiOk(CZWisCMrHrA+%#J6GBY;9NAgkKF6ubwg?yNuuKismeGhOo4zPsb
z8hj0*@UVgv_mrOzz${YAa*v+kH8R=I=kCYM@!G@Uzcm9fhWCc<2#YOjI$_)?<|zZ2
zzSgUpMPbUD#ZBGbnP$>E`aov#Dg)NkWL~A`S+E9TI-ps^m;rM^Nwe!}aovL8yjPq8
zh~}BdE`zCA$L-pW2FqF7psORsjoOVo6=Qtfe9wqI3smz~vu=>JTlV~Vt?Ox!WXT+J
zIrNb3wkQ{gwu1hh;V!X|x+*330)Euxdk;KDV-!FRbK)`r68kikX~S~-5wP`SZA2Ws
zg|;-IBK#*E+|B(rj=SUhk_!mKH7qrFyysrNPST6=CWE2uTkBF)Ugev9L4(6iY|>l-
z@(fd<AQjU9Beg4_S@t&~&;u82E+_ho!XPsz`$O89VM2`J)kkBl6e`@tcLMGID%Mq)
z%+b2Q1n7m+FC<@ZqZ4E4Ajz!(+L{vUuiEs#G`I@hqB*DjP{HxE^DRmb5)sTd7NyAz
z<LG5zibUB~eUw<_BTZ!+_>Y^v%QMq^eP?b@>A8<?1duK{KD(8K6926<_nK6!NmJN{
z98I5cReYc3fbcBb3*Y|It_lM;YB#X4j8b8lHz1;w$7OGlzXpXu)qQf7@zHvZN@0h6
z14WCitU&1>6pp4+)U``IDu{75s?(ieLvu=&Ycw5!wr>`(jOQr!gL_-%?qfS5(CM9M
z4oG2KRMGdVlpJ!#=nWR5WI4{$ARi^f(j$nX*I`cTavcw$t3$EsAO4(IZ4H+YB*x~J
zD#xz)3{C1xLD7`7s>SJ2${}k^;peZziyY?u3}h<bPQyC#9`#U~9~_QpZ!Jf0Pu6}I
zNX^a7UoTb3Sy&V#opY9mzXIQFLzQZ|{@7hmUm4)Yp+9Q&j68aOT|+)()Fj~6281?@
z;4|9|*|c83XY^zkOQ4UTNEy!BV@|M0#YVX`E1Xtv_U~uwdg~I`+OF<&-t=;z-wh;S
zwQpsd8|dqoz~KhTdU|?=<>lk1o~;L8j&u=5+!w*Xd|xRuNw?rVq73sHz4&tmf94NW
z0#f%fl|zlNTbnf544Q2aVNQ>SMvAN7`xJ}W*^{KCjWJ1$aMjG2kvsE)gNYYGPX<5-
zF-LFp4`_<rB!x=L%Jg-um~Tf!L<on7dDoZ+d<bt)rmOfnGyq@hbsjTn*pT`<<*Sum
zznstPXlfUXNZ_`+HVm*F4MWnfdvkNCIj<n2a-tn&^rLlR#<kBG6b}>4{nwb&(rdSw
z)c6=dYV3@lP5a>~HCLwwheA(QXSt*4+D*iUu9*%%`+PH8m&g&Cml5=FLsoWZZ3b@4
zMmRl=d+qZL1{-VWENIC#WCSG-v7U6YYvxHwTYNdSb`IIeOFKJwbi-V?UBm=sUQaWe
zM7o_r-6bkivS031=x@25uasULET9pCEbGyLA-_2qV0hde*vobPq|qC6bBcuRo;t;k
z0!fbJ)+*3vM@>gs0shc(BPF;>f1$l$WkW0pUVu)fQazQ1ZO#q|DX6{V1BI@t7gW=Q
zdKrEZoR^zNq9e2?54`tvG|2(9jrjn;GCz7gi^#PGC#9^lep8a<ZR)g43MxW4-!OL$
z*O!*2bQ&>v!b&)QIn}=_X!G*{h)Bm}<YaNT@pT=XOTsl#n!9E-z$S7?qAB=WivjDV
z91t72%2p13=8w;<vm}K6+TnJ>q?~l&DV&ca>QBvq`vLfEkiF%(3Xw*m)HSiT?&^c~
z%(?qiRpcu7;&xmmrNU~z!e%mq1OnBA;T3u0K*%Kw?7AxKdnH@MKNO(Q_#llcX@B5k
z!%Z%)j4S`1bO?<9A-n$1ThP)Iz;pjMc6or5_`l(I1AxE(8w&sY?>9jpuIIpy{5Kf>
zy}SQ2LaBi??Z43;{-`C88q{8*qB0x!R^FI@%|?UwOkLG~9sy9UAZ+n|j@L=-t*aX0
zpNpPA`7C_#w}6l3dB{Ndy|fx8wW<LZ6xg8G7~*&OjGb)WwBwz0Gm1o;<DA6F?X>vZ
zj7KE&2ojw$`c^rCEBtTv{rrry(fL5~{Nic<S7&W$=mjQcS8k*GP_KBshlPR1WyP|(
zVinV3IIytP<C%@l`epRbk9=yO$Fe9;Xyp@fkV*mzokpAAOB1;0Ice)y69|?8Pe?C{
zorO{)DmnZuvJg*OSe!aOo(KeRWpx?t^qnu_r;0b`OG;DBXLu^+cymi@F=t2VK-%x}
zBXPejS5cdX2dB)RTbaIJTa;9EYym29Tyw$J_->Qt7e?#Q1XfMLP+BEy`iM9=A69F%
zG3Hp*^Bqn=z>TvWs5M`*mn2S<<*}sd`A^bNo-5g_S(La|?JSh#6_pwU*oWsnix`$X
zW0i*4zWQTI21@>T1r*AKYZyujzq~u)jSguQS3G|woL#^Uv_f=EMiM{kx13qC5-KRk
z(X`ytHJd3JzP(%m!xy^ff+~#-aX5n1MG1*$aLPE$h$-xV-cx?bBZNR22e>1=zQWgz
z>DtWZ7A0dqx0m!@hckn>!FMO5W<U3qbUGcqF$4z~oOPpj0g1<jYs6z9KZHz55hrvN
z>I?#c%~pwxcao5#<U@u=#-+njdg>xU1NE1~a0K*R1;Y8PCb=T^w4)5&<Wtt10L#dO
z2#R5ZNfY$&Z&yKyvUFrrBUVz6jobU8b@kC;OiGI~2t78i1u-$JXSirQUN^8mSQeit
zDwH*}nrOj=UseE*!MXJBFQD0!+U03v1GPQpIm0|*K?`dEYxqPSx@oYTesS!JMlGhB
zFWwWIzpF?>9ySyg!S2_jsheR>8RQM%QB0|nonw8-6acEfUDh6-mv&M8UDM~9nbY12
z&HP4&3r#fOHaby|FrM68#B@;J8z%p3fZOJcsrW$--1Ps1LEwZY@F~QX$Dc^fqX-z?
z5t8G2IlI1!)2vl&tX#=SbJH@^l&D|`Sv=rYb0kEyNu(;snTe53=pn)&o>rMg4dp;l
zk6Py=^D==d$L%cAHL4;8(;Qpps#I|w_ponZ1eg3VShf;T_pH=W?e9*0jI-Q5l8gSK
zGXN^XT4&Jv0-TC1a4UBe+2Yc|3>krt5D-u9$bAZx1aI_*2qEU8P{JMA4^NJ8FN<is
z-@dCndSm`r_+oAOW4<1SXU(R2XC~yN+-p?zo(JYl%w^%MP^e0KKp?mwD8mc24Oh1H
z&OO|Z+;vW#1-tc=fAM_LT~oJP*dtJK*AhRV7kJ4xyLW)Tto#Qt|4|pM>sY=KT&q7d
z<33hmITGe~IzLkS1HApi3+Zf?LZ8~^q_claZMKCa<T@KpSt9Vd2s29sb_!|8?~*le
z5d+GL;C~~v-Un8Z?x6T7@XQU{eC=)3oeVHNi0kRUYd0-z*@ftofhpMd;aUrFU6T_o
z@V7KM1-a7csRo55o1eFfmqMshv>yY}G=Qro47$tMd5qn)-xCO|kfJIVB>jAZ?V+(p
zC+;T5>Ec}4wm6Euho)w}cf7A{*itsBXvk~!!yYpbGH^3c0^H!Dd<`T4h)kxs`b2<W
zo_z*&X3?lrF7sGj6AmD0iPvE@7#fd4`_3ZBcatRnP6S-e2S!SiX;k+eO7y5rZqar@
z^%Wjw4cxLXdhreZ66Ia_Vir1gdh7^rR^&^1X)SzzGGh!vmmk|%V0cJ}L-Ee(*K=yC
z_#s6=pxHy`hGYOO@!NZuQ1MK}3ccPr-VA=C?&{pY<4TNK27spoREhutF5T>zOrTZF
zp<e3eW=owj98JrE)>efxPs(ugis-f+#F54*)MMZ#cJ>;&cS8f|qh4%B+U?vUzLhwU
z=$-<AhMk9k!!3adW$;$v8;N-&6_eJlR!{AommHReZnteX6uxA|EMX^0|B7`qdL(#V
zN*Od|y1=>I|9zzOTR$JWyX_ypti96N{(kEncZ&AkG_fO5wH>7+K2`snuxwSjrv$o2
z(R9?rbj`zgnv5I)DU3qTa+A8w^1qAN^<wa}7gJPwk3b{`mBa(St<ce_7KB+yG2p~h
zHR?TlE)wnj&d9(?f2k#2j;Lgquyu~SR}=Qceu4`-pKAaeo8~(bO|_xRV9`p=Gsl#=
zx_|}6ryn$ETL=$rxMmvwVM?0?Cp3FRYT4CLZ7fpVe<rCB>W-ki<ZbxAr*2Z)P5!7+
z%Q?^@a=;R)Oc{QeJYJlke<CegzJwY`4rrvc{LrLO7F+!7wjzXwhqUWW`|A^=y#Zf%
z7{^-JW@OiUdE61Ap0Q8}=v=fj)IvfW9RyAq4QJgyY&?X}j}~WCJZ^sXcKZN!a6Q1f
zab}@oJO24khw{Vcu}T}sh526HU&a-=daQm~V!w3q*}0@YF0}wwc6On2Z>{&P)At}Y
zHg+9r%9}Mn*>-@7@<5Q1uq3_V&-7bHAD<VbJXi5#tZ9FDbU6zhrn*1%Snaa$?b(3p
za2_kD_0zw$oORPV!xPF8&mEOul-|nnG9Hw%+GxRb^+i7=A4s%lkDAp$4&G1)-4l+^
zf}(3YSynmFQzFs24@X9r^rzI7jo<gdCZB4$v%1!%^S&WQ5DORYr}Psw!-;(UQKMGn
zvsN9-f~8wdAiqL=kgbFU3Tk=DjrC?}CY>@AiuP(i_aQru+^g|LvZ`hBb&m>5I?E;6
zn8sduIgLuScFhyKS`-72KoWiB-95x2z_whiPg!n=Bcq)>pO0dM$EdRc-o4H=z!8GM
zh@xhuFqlG=DJO!Yk2fvpS7prSkwHSe9nai|+wl#LO6z@ltxKoGvaE4f@+aMQ!aPWP
z!jc*zek*seQ~33*SKgTL?IO<&IqZaaNwKIX+>xzVY&Y`gb=9hon$LIsn#mpCYImJ1
zxo<0_gKFtJZb)v9>$4(s2FoBd8T*q_iu|(jX))5$M0`g`pGfE8tbpSFXG)5DYzFtB
z?r#n+5;8J0VaQ*n!MRmQ7bN~N@etoA{S9ZZL)zx$S@~gE(QujNO>-VQvb<wp2YpKF
zDKHnL7W=aQ=j{0oXFzkce#e__Fxx3oDzJdt$^=*jIEW~Iv6ZHMS=COJrm?))s<|%x
z3V<i$#^e{wmD*=&&qgRoX;$+%?g03ifF?-Rze@JnMb!3>`i0SwKYbINzzUOktgalW
zIU6LqFTIk(^7v%?L~nsou_D7M>0-HLy5sU=H$bm^3GS4pzGX71QJ9i7;_m8{>9v*m
z8&&7g*dT1reV$=5oZH5_GRI~n3VxO)IeU7HJ)iO2ptBVf*spcI9;-V#OwlQ~w6DMf
zY7PB&*2(IB_;_ek!Qrb1*V}mEF~ft+Ca3{%@h6q4JCGOho`r`pGU_3Zlmqzn9RY=t
zuLV3bL{<r=pQLc?vkiS|E$aL7g0t7E^T`Dlw#)t~MD1BBjv?2bU>c>>Nb)uJ@uU@@
zmQZQaHj(X^Cv2%xv1o?CdCt52h5XV|?4KQKbkf-#5uK{v)i2Kj)4rw>CPa_R{z7_z
zs@^j4WH{uzQ_j#5#O<5$%(l=_258ZDyMDqsDd4?@^20OYkG1ISlhlk=vUXRe8h!fF
zuwg*mHE;`Sv6FblD&QLuQ0myD32TAl+zuz#6H<^~&sIGrB`1c5bA7o|@*q<sW~yQH
zQk}-Cr@w>I=X-sNusjDh?ra<*u31cQU|4dlzQqIbbM5$ljMW_uO8#>sHmubXlx&W%
z*5?`((Wdn%mp_Wf=3=T?CtosLjf^y@oDl_JsAM`f)X-Z~LGek=BYK}VAp8~};lJe^
z*x2dZ>3q4InLZ-VXmzrtSSR*RUOx~_UtoQ(4378rSk9m)P;}}$^nBen!&3|C-^o|o
z{J=rrPnMqT`@D52b(fn;w!g}cHB<c}^?aiUKJuuLJIWF`pfM8n^d9nKw~xZ<eE%lu
z;g9Kf?LMz&le64#<Yh;GJ%GmBCKHN_v!&3pIZm=dIA#0Zj$lMZw@KTB=&?z2%yxMk
z;B3mWvr)NCSwQR0{h~Ic`q^?By6>uJ`U^q^vU=ofskn}fw-7iua#&KD(?JetdxE>O
z+9~O*8plBjXes-vJ;lc#de|dO5n4DM^n1qViLq$teD8g?)K?!>4r<STyxxG^pV1)a
zF(;adcgJ<(J(VqQl>iO_%TMhLdpfZ?7OA@MeGsV1X@AgxYjxUsa&h4Miy?DhC|T)I
zFSxUmnvityCxF8=V7rx+M`V?+dXy)HrZPn~_XQZkO5ls{PC?k&fnU@LNoO7@U-G@t
zCZ&RI#F@rR>s3N0eM;fnD_Z1~v{!4$z7AHqGF}O!cr)F_cfrMm;Er7qhE!w@dpyF)
z$_}i_S=YmL0P!pHU`xCvQ*zJ_-A3eL5;8c$Ou2l<R#N%VIeZqk?_s~NdALpD9v(qT
z|3i3>;C}{HqCM*+$~}Vq1SmNBk3&$3P_JjQlqzp?D%=}Mav7N3mEUq4xBre;TCtdV
z$%oR2F)BWLXRRECyRmew;LNyTm74mSD=Xcss&{=Uk&U-VxplD?%7<L<S~|SBjF<S`
z5Hm5~>)fVgUfn+cKr*Y?(z`ik!V=(%ubY5Q0_s^rWNgvg?09jcJ$!H-%S*48bR`sn
zLK<OE+^!R~WdxIit+N-7ZBpJ}^qG7ka(m~T+g5I03k9@$tV$eKToldbTIpV6*2Jey
zr@68yJ6!X%bU!)#lA_ntt6r;IP`fUW!m52XpLwfaaezBWG3^jAmMY_jBWsbh^=}9y
z(G(RtoGtoUvsBxq$ojU^wkjQYknFYV)vF5r8j|qWR&k?%O~($Jxj4rTZLu7I$LQND
zG=~7~-9z=zf{Vs~cF@O&oCf-4Tc5vI9M6v6d|$J1R>Hq=3#j(22W1T3T=4s*v|>OS
zNBQ7cfOkieWFifKBf{FW&My9J_Lg&q6gG)dNmGPoR9~f6c^fO{iL6+hk|JwR%5rkw
z{9i4==?)x)5;)<mvhQRfn5IO{*4h_l0N~*wka=W{y9+t{q#u42MnB{BkY@dZlslIH
z{Qdnuhc!)mcnJ_7q|T{`1VDnrDH#(+9!A)<qVYq%cP$xSE=t7OM}^VCyLC;9z8)=#
zCO?&Yxe{BUWN%*svy$7DAp~h?V%Zs$seFYe_vS2l;;pIIc3fAs@)s{!J!p^lA?mJW
zI~58WPDDK#`NPLDuaXxQSP*M(H;ZoKTpu|W9%h!=ccxYXm~p~z(oWW0oZP#z&ol?+
z{CdPfmB*QElVcY{73e>4&eVr#ENTF#ETicBJ#<)DERn0(_-u4*a;ixYwoURu-4ASo
z;(V}~4d~DFV$dCEa!;zd?{H#I)75{DuSmX7>(neHT@QG(Xmpp^tD{NV_BH#xih1t9
zmXQrF5k9Ga(k+4M!HE2D&yA)6-IXH>FCZLrobs#eu>tg+vvdV_p}MyGY#=^b0HjcV
zSpK}(yY%p!dbCkgY_r*c4ps>HuAv+PKK9uMvUO}&?*5O3CiDiYMgme~Le=OqN2(X*
zG4|LVIOn0fcdzDv<re$w6yq~G+6Ze=cX^`8F7f_P58M^(I3@OJ1wu<R!1K*jK6IiI
zx^12x(Wq&%s35^BsL5_DVkR{lkG~t?k1npP*0{9VsKA-q<HzZp&Y);O*XG(Llw11Q
zl;WZDPv~(`*uVhJRNSEDyQNOE!+lYZyp^2=Hr*b5u=UN=@U3O|2H7XW)R%WzLG>?z
zv<5x*@L7Lf{a<8{N-yyxAw$X6dFnbU$57lEEP~+p;QdsGCDx9!Ff?$bzc8fv^}n!3
z&ogKImW9x~qHdNP`a;pudf<|xZ#9MTwkGG{0@V%JAf5e%EYYKxw4~HGUtIYX(B6*W
z(0D`50U!0(6~iF_rO5C)y%LrxFB)GfSwvl#h9uVNP2Kiy`3fycK)<LM+xg<MF|6IT
zc-+9mDDS(2>*zQDrn~>YtCbyA{yBKquI1{2FFO5w`f>Vq`V^5!ALv?zHo$cY)89|d
zBj#Iq2f%``$fhd=Odl+34h?100O1z37Q}la^r^iLcWEUVXki>|2mqk>d60xp^gXyS
zP4+kUPIcYLF`E~_(h&i4zeInN1NG+j<+<QBxi5WTn?cDgdH3~G&BklZ&|m8^X@qo1
zG3QzG<7e4pEH?p7x9Y#vOuJ<S&d>X`k>Zkd-%Qec*2_rVN2`Aq67qY1$^<xLx=hHT
zG+zw(GjXdQ)S0Nf3Fdq~0LG*|Cx8Rr#mO%ukAHdad-x#H2pUM~JwMH^{`TFwueEh`
zE&d4r+V-yyh{t*D118ZJ0J5HMq|m?Po~ddf{EM2o6fWJ<1781U-1v9C{x|+#gKB`8
z{~Q0GV0QoiUtY(?7q3#AE9dzFM%uPzrlZo=(aB#tUzd5;2x;~z;qqTi`hhyCaD{5F
z^u)#ZUCz|m9*yg2+joCvmiJtb?H?HMCLK*VUYwuy|Ni~Y*<8tq(SHV0L9Jo0s64nm
z?u$I&l#-U#NY2X2`sTucyBLv9a>Br0hY);2idA^ru<)U58OMY5p@8KG*25v!U|l&m
zxjg^-|6M2F_P~qW5{oXHZS}9)`MGjk@nBFgp6j3q1=W(7_ev?ldoKT?gszENyy3Yq
zGVi-s5r{_^0+*2Y+5Upui*vP`sx%8&O?r8lRu$CEltB64bq|B`a>;b`^^0C|L-lA`
z#mA4PEbu4p!EHo~K2MZl@%QfnX5JI}XXC|6MRRker$I_-X=zx<;#oRycy4{rsRyTW
z@ZA67a<4%OXPX86{rzYXVNn8x%BaUP(NRfMi2o!k1cK*aFj$DaY6rSVQ}?3gzncgh
zU+TNkKHK2f;4mQqJsq$xaB*>Y+cpK5S>Mo5P@Vmewc51v`d}unx=EAAhkj+2|Gt>W
zt{_8^5AYd_7sriK&a;n)a^%L2R#+v>%QGy3jiat}zIAu65zxx{5kbr9fG$*jXULc%
z@2!!nXJS$s8yj1Vl5+zNF-7Mo2EM(h!HGfOuX^J-X#*FLP$J*VV&PH#JJ4?FWasME
z)|S=v3~CU_uBxVH4}Cc`HRUNr$s_jo@nf~xN@ft_fakSP*u`OSX+_0E2J*1zCs&)}
zRHf!+$oYvoU?Njj<mq=25s|}&D(LH(nHguea>zw%VS1>)tiQj1vPsyJula!@N^pc=
zp5IeElISu`6wuDQuU=YNX#j;nyYQ4Cb3I+%vH2F@pT9MAc6Ro~lY_jXNA>`eBGE5T
z1N84)(oDZU8qT2f<J-4;h6zI`)YOLZr9WU2-R|!0{nMakzrAk;Rpu4kq|KMJKivzE
z1Oy%{YH_NO-5c>J6__hGVJ+Uf3OBBrp81L>{oiby0<yES8|h?FE?@49avc4pTFK2_
zX{J+Uj_?|in)=b&YRaIn8=5X*QQ6<G0n87xSpmth;E1nzXUp|2WXpucSTnWmUbrvD
zxzWBSmhBzB44|GO={Ztiu*G+4S3rJm%*^JCJ`kIln)tNIK%tIDM@Jc-;j%DLqWDwQ
z(&brKip@JHJ`h*PxqB|yS8MTf*q1T5`t~F1e{6<xN~9bo?J#g>XGYN1G%DcdZ*p@x
zP$Xq$&aAJBF>_-U8m7!S_K?M}V3XtqCV)8L;e0G5Wu^gIU0oFt6)npqTm^v&vw$@y
zOqQ6NJJfJ1wEl+tB)}>#q9m_{dT&jtCv*8NeY{SHh!oJ**T)Wn@u_^!OXA|xKR0>r
zC|_=l7BtulvF9p5n}=}x{QSv<OoCd^x6`=UlfnNNdtVh5bsVibptJ%8AxKF{gOs$C
zba$6@Bi$(i0@6sQz|aCihae)ILk%e)-JNrPoO|zC>pa}2`*82WaV=h)VKMWIy}$i^
z-`;<liQ>e7|9y!mxh3{?<MTiS;P<T~r+nHpSl;S$E*l>i6@{2H)YtcAkn>vo6IT^$
zP-l}wW4$-m$bkIMKw@(skGeD}MF9ublIU!^N93-WW_y#OX;9%bz}$STF!18ya0M~9
zethY=U0yF6pXGbhiyK>BQZgFF*7jPa>8OwCzudTQ58rX+-kHQ*eZF7P_M0`fI>*H}
zo)BkSTiZHtvHcFCj1p-SvV{#o&>wFdTnOwnhzn;cIkuiA`(Gm#p>1u3@vcX!GJdC}
zMjB>j`CunzEZpBQ{^#W!Ilc@ozh3d2$Q2Ji>3D!woRjk|R^-kz{CYS`s8q>sOHC!)
z?B?R|>T0Up-<r|Ohm_Z0LaY4>UUmMMLRcSpw~wp<A!k@{GU9&k@I-E^l3DbAFEK)_
zbD>b#y1Ke}yCr_{KyYC6;BYv!q$G}pfRNDT{$4Ww7pXAW{d^CS_4gZ9N2ZFte%XDD
z*KPA{)GXI61nUn4@x$EQJdMjbo{(8P2kRM~M5ctHib}50?Ut7JR!Px)m;;-uySBEb
zTWb|p(Y)7G72>XT`Jdq{i-koKlH*ea@_m!ZePrEW)&$5zstmMo(6d<luV4j%e0BJL
zpX&Z^#I^ta^bpAZQF;9D#rgkZalXyBK7c^>`TPyNKc?5_t3pdsB997W#0vmb*6Pya
zvk$K068EXXn2$gK9<DS!t2E!YvCsZvaI?a=+y8iZL-<~p#(I`m>zRLY&8<o~;^MQ$
zXYEsX()WJz&bA<uYQ8QUyyGAZbhsit4fwrscXzLqWE>q6Dl#eOd8PVtFi;_sk3j~3
z&<t#+1L;w+t-VR{YT0&7oL9~a-F4po^f}PSPB99ftrwO(vi2;Y75{O`g&vqVHBDpB
z&%Db=1?QCoT@0(as>sg01<nqHqTyG-HDp?TT1lw=5JgbryA{bg&kcciKNQ1-1cm~k
zS*UTA^+&J@%oba^-CyhKDl=Y00a^ugxyADMg;_?l*;^N|5J0A}h}p^VdjN4=0Noz}
z?ev)y4nEgFj!21(6XsCr>)iq`w(K1e08Ll5-)mD3QQ)<<sS}C{k#@9x>Wu=SLHG!i
zUvcOjh^%ht*kUE(f}ZmHN*=AOu*<cK(zv^i5hrJ76^pzDML7jsmIVaOg&rQ$WigX1
z&5#Db;+}+P37{isecD^=)vkv6C$nD~tn5fy4;or;;##jzl~hzWpJVpAoaLR~cx(-p
z_acbEmt+L-6!O$`80#Ei#Pgc(dMu%()^OIF=J$KyP2T<jXLD5@m6RRAv5Bl_V@TGN
zGnh|1;TjOW5-VQXmU=?oegypw@UiLM{Z{&dqD9~xZkBUrvIi<Txe;sm{NmZ!j=c3K
z%_huRm5$aa>*??%J#Y(r8mEA~6Pw_n@wir>ApT0|sJipnW`wcIXnumgXU!g*9Hw*W
zeY+GMitm46b#B*%j%_0x7i=@KiJ<Mtzy8PeIgYe>rK7WZ2G%L1QuuJ&p36zuN#(Hn
zqdPfO@%)ocj4|p;ZN0Y1A$7UZ^I6%J#~7ESg(O2FD3H(51Q4*R8;Jh-Gg`FXnb@ok
zi&&ect;f6dzjPS_KZ8#4zO%FTt`_D^akT<cdbv9;R(la3<nFN8$`C{-C5Mv0mC}Tu
zrDu<%?bGoD5Uj6p1os6kx)$TKg91-hcnglXm-q3?N2cA?uWoHXWvrr!^FbPiUdBT{
z;SFbe6G9qNzn-w^bcY3b=L7OsAm&EI?$B!PQ2nV)S|U2i*N;z7pTuRfre@q8Gzrlc
zREiJ@2dR!z-ug$MWIKjt|NP2oxPae`;l43-JfLe?BM4~_8OmOHoBh1>Ojc@$AK~7s
zX+2f_rxa}~z07aXqpOVY-Cr)Wr^h;+P1zGnJ(p(=Ls}@118lJI@0j#rq$}-{VH9<=
zovkhfSRn=JbaK?P%|83yuy54BH(=Lb#jF*RBs-pi+CPY9Er<Z>^jH5A?;YA(+jBkq
z8*#NJi+1m{Q&w~u&8&o&Ofzz{B?-}(<{aUk%T*`NhcE`p+toqvVZ_|kWtA%8g?TDJ
zUVkFQP(oN9e(wTgAn>h5lTor{ScOp?GEejmR1KfeYj!Wdv|?(V#9l4}o#A;8Hc0T#
zAL(9y-EXrC`k~BT-86L7>3^%U*?l_S94nN$$R^~OIU&H?i9V?1MlD+yJGlE}eoK@3
zT1JO;7UK((+(MftTn|B3k#R8N^OnpK^<l_Z38P1>bF<Lbir-Cd9j<-HQc-#lDX#y{
zQr*7K{(LWN-OF=cDu};en_d{;|6aFR4+^b8Y>@)rTERXWjF0QEqRoIyCXFNSlH$n^
zOq~~F>lL(E7!YYCu%i%};j)B<Y164kammz3lW6r+dIN44(>b%(5PZ6O`gh1A1~0E1
zBLyW3@jh$LsIE@jiBi2BvZlqX+xe4H)4w;HdpfF$X}(vhA7r&+Yo;{0a;N6q)C(K`
zHT)2=ifAR18qdE~2U1^Rrg<rChHT3D>k*wPg|7><bPUAJnY}Hes-guneEhTwJ!Ls{
zF$h=bCR6#RSt8M8b1cA~SxOpge@89j?Gaw<9h(_7L}tHF)ZXPH**+Y6&1PS9rm(Ap
z7PCd~cZ6;2DP!pm7GISp#RD?7(2Jw_;w8%kMbmAB#^e?qTeIyMi;nPHW6;E9M5!^Q
zgErqcKgPXMs5J1=e;KYT6LiwOT?l*>L$W1JhFb}2(WF-8PgQ@GFxtNOcv6O@1W0)x
z=tdsRi%c1Cmeoy+Vzx$zS2lwan1AhIQ8_*?lcJIBP=6j7SIBcS79roxRqN<-j<F$&
zp``cY;1`-L-<u!mX*_dJ7c;qmYH=K({p?mSui)L)pNydZkx<?rMi+eS5nn5^Nh4Gj
zU*EbH*4S^8fG_y3D_Q&cMy%xZ@Qm55@MbataNw3Ft@q-nx_@uwK|hVCY{iE_4DVe;
zvwGD%m+>oG{MNUkrSh^G0<kh8-qcsaH<LVijDAM+x~GYR(uHe>B5eX(Xw09FxBkVn
zXD^l~I%L}TS@^M5ygL3-*xsshRasw9ZW8>=zx|%1OfJcsJ*`dJK9Z+<g3_+e6_9cK
zup?uXkOob4zT4ZD>FBO4=A2+vH_zv+cqkw{y+}Lf@VvAOFPepd^7b<LHOtY6HR$PI
zI1<HR;&9Qg`0R4c7XsrUixG09%9!sdlp>M~2q<lKkL3II=n4&zBcx~vIkAdX7(0_C
zWhQ+j`yi_DFr>hahzPa%rQn*w^@Ac+UKD&_m-1qB@9bKyIeVw|CGJyUuL~G8qbXUo
zP+Dbf`dDF1ZC|hbQb~VruMy|CpuMs8DP;RhkHALr6N_ZFUKaU-xCHYLBAZ`BevX&d
zD@oNz$@F{3k>b8zta!UP>xxoR0TaDCi_*NMP}C?t8RsobLznyN_DM<8P0+UBly~uC
zC+a|9Yhg*(UkV58b@b=D@pm(BqxJm<dX*RE0+}KnP5urvk%;4+$+PSqJdwPbG!6mx
zm4J*4<RK~w*;}V~+f$WOSEt)`9{Yy(yg>X!dq6-mF^8DiclpuP-UM#bUznFGn;a+|
z$C`W9(CxDa!l3)y$3mx0>lJHpx<2m9kIjBcKlTf1jIFC0efQ?c#(MtIdSrV?p1bp4
zAPbAvLO{2K`8UMYn`{s$cW@#&16pcbrA{VEsa7y_zrF?@R$yk8j%~HyE$?Jtif>!W
zn@1f<5UQhd0j+&V13HD5AAVj9F%Hoy&t!rTGwW7tWq1JrhOiUMsp)EmZU@U!5exzZ
zHtw&=n)N|BkApT;NKVbuWC$jDh;F-oE4j}O$L&!{ySlx7d47+%rRDZ`_Fe5>JJNqJ
zl|x*Z*N*_=IoVGNI#9#gE1{FgCln%Vex4TV?OFjTD5CX<a~dC|hPk^BE@%1l-sbto
z_C^aGM@Lzv>LNh;V&-_JpqPAe$R%ST8J|Ea)G&e`&|W_toc~xM`S~}F&UonPEVF)D
zy#{j;v@B_C<L*S`{jj>)qG0M&(C@&LE5gnElXWF8AfismTmO^Z1d2k%iu)Y6^%{?=
zg9dY{2I`fPvJ6X$XUsr{%}LUv9{C}u@<d|~g^+Cx%3rb*{qd2bnWRjr=nczKy~`R@
z$e2vgGf2?8ca^PYX>A8#Y*y<-e{30!kL_Llc1PLEc>&_+blNO-v$}_TyjUgwq>JKC
z8Sr(_N>Ky3VB6ZOO98fQ6)C@yXCw_Q2VPqLG@>W?<=@V-uB5T-{fyq_S{=A__@A0c
zJH7?I+as?NVxi<zT&1s6pXR()@?PR6{g<)l>(U-vSlx?iUTF_QI@%9FHWRd@#V_Xb
zFtEt-Gj#(39|op?X+#}<T=nwT$eno)Pj!u}j>CQVcp1k;WihN)4}8=yfv0E$)lsEU
zk@j$n`VGf<12;Dh`!cjC*1W&ER@kpJ*nx{nPmKPZn2ZwLx651SKLrYnyx+db)9tT&
z9kGA&vG~U>^DoP1$ZL@(vb^d;^KG?}b?rNhnOh?>A>DVzki^6Wa!Uczir=V#-KbF>
zPz6%SZh}}^u3!OlId6KZQ2bM=IwDA!CX-wW>ss6O-U_%2uxkO2y@B2O#i6C2Au`g^
zl&q{V2L}h(v4j2nD?2kaF9qDctge3eH<;eD(CQ<jpkRKq+WV`g=NJ$)>|RG7A~5mu
z$X%W{*$?OKavMRc5X4t7_YA`PmQjNo6SASH3f}IZ_O<WB11ilOtZU*DRZUf~v2de|
zAyW;sTw0AtM@)8T)1!!8lU)>Pp(1x2o-cuhEG7~}m`QmsxT;;yR5F6i__0miX_(}}
zuZJ=TRK8L$*t-wuIE`7GQU<D~D%>fG`pVD1u;pCa7NoyBNnnXzZa`;}RZxvq9$Kfy
z;3pM>0;`&4i7@DMVL5m8{li>QeDbn=4nn4u#;5Kt>)XeXR_He?KAx`3u2GeS?V<Q$
zpPLG1aUj(1zbvh12cT2(@Fch2?mEu7jw}4??r!w8)7B=opRLVRNay)N%n>YdyQ%cy
z!-oSv>K^aZ47+SiFjTajfjQ^TWE2%E9Eig3-IS}*&a%c9?*8y`KPf%$%=xkSC?u-C
zqTr7+KK4irt$Z`oqEZdf#Dk%hGb>YQPNbPVNxmPiwMy8zPHM8U3`On_4K7{zUu6v&
z2RbDnn&{^W3^J~s?$sEVuJj^R!@3V989g`OCs{~;p8d{iF6SJP#$HxgV(9)d+>3PL
zm6sQlD70!XlzgjYJyw?|P#}E5o$Y2ZQJ67lcryQr0#TZDx7cpZuys!IGS(wXZYfja
zV667Xgt8bWRcJdDboxIg<r!)Aa0Kk*(B<*E*G6tQm%^RESg{KAeb5=o66)_uV%nUl
zq#)(7D?U5MqT2etgX(8;oikQHp`MDB)rKB_rJgw*jjP7+p8SR0xmsIjnz5;peFAdO
z)L2;{)}wdD1$()hw(955sz}Ej=fQ8Kt5*WEJta==RDulDdTU!oFu7YjFPEY5qbT)R
z3KrJ&pTihJ2nwp&$j2V@-vybzYHh?-T(s0-7@sMfi5Ou=YRLY^o6-<*i^9)QwHhKR
z9NXg>IXr(soI{v<@|rb>LbbMTkSRFH=TxZu;3X3yw)L+6g)@<zqS}flxGCIPyO@D%
z{VANj7f$s>8yqCjUr4wkfqGEmzN=d=T#w&*9;RV-ZuL3cJtGZO{q2QY>+-53_tF>K
zuzwtPb;m3vL>YfvvyNW#KPM;8S1p}z`xv0oJe1MPGEv{aS>ZEyMH9-Ia=cvg;aY$e
zqgG`zrX?=Tbn%6s+dEaoj%kOi?O1%>&2PrcYZcWJ1};}m2NZrgKPg~3zjcw8?0>wu
z)IEgNZpPm6hn11V=!%y6aIe@MuUIm~C%oh_gLfifz$D`ujsBF@qAljRfY%dYoyxrC
zejF|+>cZ4{IOU8lbtpbuGynO!tPkwOf_Efhn2XO|7P$P#ZYnMZK>se&zk)a#8ko`k
zhD1YjVy#(&YHiL%!1fV(1N8A67rtqi%^APWW!3ug539wU_Vy&Dna2zI@_Zj<cFD*7
zy0l$2QVhF}EcHF?T{;`#{i#6qJTD%n@u1VRVS|$X@*h_cbZ1zGk-vA`@9={Um0+QB
zfiBIsE>4acHTMhu1LjpCG5H4BZ(1p~cUFg8iW;vqq9hR}P8iph6JaA=foT)%epRyj
zoJg(Hb1lR4-%kA<Tkzyjzr7<D*UaLr107HEZ?X*9I@M_YbLCq+f*hi+oWk()nxL%^
z%252QH_-khGHVh<Uw5(p`;fj(0T^jPmR3SEmFNhMB0*1i#NVeE-chqN8miW&RKkRN
zB!62?bDP~J5+F%94fbXP-uhmgAO1m5{{TG^A|oNHzUkxc7jT*REXxG#uKD-jYCy@C
zm!6I+f*hkY?NQ!Q0h2vs6P>&-9@+R-IW@!1E+r4<f@Oz;7@7TT^JJfG#GYhO4_kK4
z^p9}Oe!P8hsuW8I-8kBevKjw0TKqdwArFeo@E(?Q*@GH{S!hW8Sy&D5!Wdgdlsl=B
zRrb(@;@3$MLPGI{Ej{V$`rVK()_0$h$S@|SVY7>2@GAZ$(p!R4kRf*z4MI}_lsCh?
zVXlkY{COTk(}jX{%~p22)A8t*tfdhJ7up0bMqiI{-abcN>Ui-swW54?Vf?H2P>p0P
zf@gB7EpzBh%yaBlR@H4oLU9Gt7xcF&8mInDUww&V%sVOQuNPmvi;$Gu+H_`60-xvl
zh4M<-h4#5%_QJ<F`r-V}x@<fszgPuvAc4<vnjyYn*PyHzrV0gI)_vWqC+6^9y5rFR
z=P-M~o$t}{yz);q%w^s$?>#-4PwaOa6zV-aeS=qnPq?FC&4vvO)g8^Ylu~9kpVCJ3
z)LxzRJyfb8*1-4vhk3E<lschChYVXU%ov%uE3mv$f$BVRwE>C|3h}~B{`?a$BDEad
zk=?lg=k%p_RSm@sK3lH_5K9Yfm#FzFaN#lPiD5#~(11wXkN%HUpj^QYABbYzr5bYF
zcFZItUo<xj;frCKWFx2{M@L7gf?jVR^UWTXXFD?lF#%^YmUGTMWCVLrk&&iA;;OYC
z|2+@>;)F`B=)nl_v<C?fxeC#eW{wV*KAh0Iw#;iQ>UgH=_~8eO(uiQ5=v<$c)#S2@
zPE!r74i%Q4M@Lb{Yr+i`v6j~pik7Y+JE+g<dNsDP`&#1aFNUZN&Lx$N-N7ZJxZ!S{
ztM^gE^oV#AL{Bf;=S-mE`R~W<Aq7Qgj9<GN;@Ix(Bem<RDX}V)q!`D2Yx`aCU`j(v
z4oxbXewGk*1w5&Ag-7UBx*E``CZ4nY0hKAX^&0pq!|nHs(ec1KXq4--Lw))w6hBXQ
zt_@NJ<U=5EJ?;Sdr@``c<HvQV{M~~O=H?up2gZsSZ$*uBKF~fDOWYjO_iI!$;qYxk
znWQM=SDB~OSMc<r7K(1?OT?9AJ@QDd-ws_Tg7EXIHKz{iyUg>vl_t0#$b*G$%o%J)
z&o>O3F7-wn{FA*V+C*PUTd;9a$29uzj(}tr4*h$(Uom+u*sSyuQzXje#f&5i#7d`u
zVp4`j`g2ph$wNUQth<^{(MYyZ+4HxpxN8Z#DIGToRTf6@KZP6Y4MI<@XeVT+Y$R-W
zK!WpV5JDTd2a@&Sk4ZI`r?2<h@j68A4)K6u)aZNp=Jxs=gHlK+vrsOjsaXJ=Sy2_N
z8y|);0W0K1n7*eO-;MXL=G!9-BJS1wGMVcwc9JDtaVq7mnY}ZKI%@T(yr<Ld@&wfB
z$Wlhnj5vduHtbQubyf+-U0u6z;ydU3E7HwTvE3{OgtkJ~jXk6OqSEg7=bc&~h>DMI
zw9=Qh2W>C{3{|>$yglN!<_mR+P85_?k~A}rk9ngGZ-QI#_XsI)hK=w%49W>1D`xOZ
zJ9Vv#{^ta=#$`#Nf367hmg*=>;D*=M{UB%RF)gn@SlHK7*d*)UAY;O!3RRwKoA)38
zt^h+Qq^S?I9?aTSJ|QO`Qd1HN8?~ZgWQ;st49Et^@4*3F%P1}}adnRs{MKM_xzlP!
z)2>y0I>Ch%nX4Uk_42}&&O8zHqf|^iHX|%5q~~{O(}Oc?@-)<?NeSXHQN*#m0D3lU
zsp;o6zqTh&5$GQkro@83RGg_|t7x=8<Tm+Mo-6#`mH5PZttqoP6%}64TU@c=qd;}?
zL)WAiSNn(BZZZoS7v}e~@KFPJj7RGeuDev^0%T4=9xmoo%aP~zES0-Y1c~fKqM>&j
zC~lRqUYitsnmb2NO)twcTFGk}W2q-z&>=YQ-r|2+qNY|y$pEmPNIe^*;QRMk2Mmmq
zB+KD}0UB!RkZ^KuOtZ$qGsVolq2Il|y?c8x_+rz~)8LBsERnJk#;&uF;j`ZC+Np1j
zMvm8M8)GjG)4q=h?+O>X%Z6c;N@kg4QqsvIj>|+aDJp6L25iDssYNpypLY-ZbpJaT
zw%Et2EV(meh#vAp_ct{2=qDW0w*BMh+3xi(D5)G03Vk?K&zTNJy-)e<mD+R)k<F>4
zL%U>Y8%!<pKcm)}9kCw;Hitt`-uH$t|3cHSs*$PHP23sp6M5wu$pWoVpY=+@{^4A3
zmEl0Ff)IMv2_CF;(?B}a*i@!$s&U%jK(SLe)F^sPEWPzO3MPPLOOA;H3!xs1U?ao9
z-f`3uaFJLWPh~d|3%eE<>pNn3DdatIAIC7$Lo7@8FkLN%8FU~c)O3%oV44D}nG9re
zu_rF9YhNje#})iEq~7!RUPW!YTYnQRdyC(v5CT2t&)cigHaW!AZc<|*NhW&oXdJ^P
zmYD(0hO$SBLQ$!epq>==@TX80^dBctUK_&o7s#Pra%wqxg8K6)gpT2_a9HazVGhM(
zhL<%w(eM<JSwGdxc1}hDMPA~Gp@VdczF_Eqf%Q67`yaPvRUbvK3brg410y%TDk}_k
zkyz5uu8@55^LBi(iBLp1|KZ_bL17`v*YtEbiO^ZiYUh>rkS6zXVgZM`WWP=+NqDX{
zH<d)D8QX0XXG-hW{DsB2G|cc}KL1KNdSvA$S~-b#jU{)zZ|rQ?S|o~g3Q6AtO7}yV
z+#e(YAF*%?#1&B_!o;TjjlBK=^C|R?T{OZWC7ZzMWeCOd*DT4a*eu*{#DJ-(Fxlc2
ze+3VUJPp{N(^t^6F4v%v@H-NZJcG%pEk$5C$-VOZL0*JVRM2w&bKS*yab@+9WI(4p
z?}V{N*JRNDt=kiSkte|gf2v28^a*#GT^wRC4}En>TBq1jTw&*E=o7CxH0(9VLa5hs
zb1+FH-VG<){M01e1X0{#?v8^WSTYc(9X~@bHWA<k087k+9P|EUf?1;mr=RXV7YED#
z^UQ~0G&`szoQds@czHi<X;f91)3jhOW?yX5mRH`6VPsNPx1&653EE>Z;32~O@UMfu
zy*M*H>9JLQC&KpQtDBUN7<y#?wMA6gz*Mu(TSoQUa+=c$Lu6>N(vKsJEfSVT#l6j@
zQM$@Cy30|~XVeePx|V`e+#G-EzD_xs5G!l#MIPa1hJ_)Pf8Q-^rWB`x--@{{mntRS
zo>46<vR(?BHEF6e+Hz`*vEX3zRt4ewy3r41={Kq4YCl{1!=)^rJ(HioXm+!;5ujXi
zUeEM+O00cID@XkR@?9x=Z*LF$hQf8j4>I2hR&>#|aH1>jsGxR*Folw;<51a50E$&{
zz#&X~Kjn^BD&;64tr=uTR9;b|bMV)+r`w6QNK_YHFrk31uXPaHw_VNWn!>$P+|1^#
zshgHDN$1KOll@&3x{ZmbJ6+V(-j5ivO3e<NpgFP6xEhW7K6SyT7!!+HhH+^Z`wOV(
z^(h8LZZnZW${*;H(&k~!@L=1jo0c%Ll%u7vZkvx{=IH7|aqTSXD^ES77+}@~Hg3;t
z<;`YGz8$Zm92*CCcVm9i#$(n8-{^qFf#NR>E$tH`q8^Z-r>o5{u&}UTW=bNru1ROZ
zLgz1YY*6Gg`QiWbT7SluwkwL9<5S~L;b|%zx9E(3P}9J)jP72&l2CYLP2ezA)kg!Z
z(p}Ef0NW{N{-HE@pq+F*(?7|BN{U6*%H*A7tTw~7-Dp2{nLth+7Gpb`LjE#LB57Rz
z^A)4mI<C$Gd`*WGJ3C?;_|HFF%Nbev8<Mn67iGT5kx*H+yP5PGHX&R_G2<bdi{m=D
z$Rgs|%ZFY3olw6aM3TiD(v_tv+InS40sFW5+$%F3FsJ^U*%@_ErF9B?+4C1m==}w=
zM{JDbmThH@TCI*NiIm491Lm9D>}GzT1!@_36^e-(gGCA5e2}XWy0TIjvR-maDLkpy
zdW6>)=5&;omI{ArqB6A+fyqIVqhKcD`>e%yO6F}=Nx6L=)~@~hW_*klJ`*(6F{lxY
zV1ZXcUsttC#+~#ZY`wWyQ$Ec$`rO1yr@vT%;EASP^zqACuW91n-Hy*dO`>Ds>@gp-
zRgszXn)cDvLrff3O+AqGtlZxh42GRtnVTwIcem6WEniC7JWFiNMpl*pSm!hvi$n+a
zwCz~>tC_E_AO`NsrHK-?)wQ)zs0LGW!Zj5Y)eY!vhkSX-H}=fniP&fCSK>>t3Co#q
zn<6dqAQUI@0Aj@awO?uJS=Og8#Kg6Zx5^U&>@NPF+(Pgn+5q+@rZSPw<EC+i=(UWL
z>J}+b+HS?60Il8D%xnR9OMip>Ct^^(zq8)Y{pWUDq6zBWjlilcah_l=mp%fusf|Y|
zp;?nxtw5(#DB89hUg3Go{FF|Ry_7~@Q~nF{b6(7F6$2A}UhX}Dp<VvmO20m=bE){y
zZ%doDOY`t5q&$4(J+(dpgHE8B^kkWTDlY09-@zjMV1j!u+8Ds=Wm3W9wGm_XkvvJ9
zB75{l#7iY9L}HlXrd{w?9eW8?D-E;N4%q4f7EckS>JYJXT7e}*xyJr*pH>fyB7~mq
z-4}|MqH<E`^CwaD0*+JcEuj(Mu^X=|Px^^g3B?q1b&X3mANF<Y55qjd=|*khnB-P|
zsfIUft#2ANDCjjmjY?Z>^zb}2l=|G{pibnbxVUaw<bdR1ugdheqOsBH`R?>D9tJ~y
zQ0w@2=FT&LAhjMvbGO`kM~v0CRVR9soOC|`hvMuxK`CbJ5~bVS<VV@IoR1mHCDQR5
z%APLqk?hm{FNZ8kq`!y0XD@DD!32JG8x^sr2b)gp25_|d1u!ZYh{Mk_d;<~EEoM^@
zN>{CvzY~~w60r5Er4jWmv0&`)gqzwG6g3D3CRSazb?6=~*HE`_-nk`u7f`AZ7SFa=
zK8BqrYxVG%cSj%$ju0h;a+=cA64J4LLFb-nc(=0Z>*hgf^hd+Gu6E2jh%HnEo4u3T
zY$?IWcqI<n^SI+PE8}n-MMg|0dVW*K@HO0C%W^Hq9{vi}u3g=82ePx&P|4g`?3Yu;
zY@4p!nkp;MHvAx&X$AKF340mmmu+0S$4~IRiKGSjSUGa?a84wMcP99FroHq*WmsIX
zvtviXQ%jL%$4>k?Dk|USd`~`&E5dsG`)F;+O*Phj92x{%?j{}P1Ne8Q#fAo`q7YuX
zTHS6#E$YlAblO3lIP^mMit+(_Nm1JCX-|n=C?n&i-*F>F+_jNIMO7U`eSz5wuwwR;
z`$Qp-mK(XW%a7%O2X8+ZZ;RL#n<tp<+4S-E_%^L(pk9;-dN+&(Hqw<_ry7=WmU5{1
z!nV#wY4QbBl+ZAK4m`0jbyR~w<4oA!`F^?5OPm!xmFq`$vNxKFQ-9pK{4Z!>Gl6$C
z9sVAT#r>rggE>F%u^m{G*yaiCx+8XOYK*pdRK2mK!@~-kMVRAXegY$}T^notzn>SX
zLl-_{7$rG=>i4oip;xt>2*>9(;SpfHaRY`g#Vo;IFt)-Im`N`HxBu&SSR^*LDYsK>
zgEHkapygZzTnrb{I=^$R<3PFzwHq-scSDY6wuB#y<TB8lx(Ber(`<-QrcR*V0Wm{S
zWldY9M!QOPDd+uE+$C@7vfI|lrwQ9;^2V?m$KF0|no@nPq#P*gczUPH9i;jVCCTja
zUl6=~<I$*|EqLm6Mi_)3B5|Y`=0oNCv(9gVb$|b*#9p2Cqnh`PSr04Z+_TjavUTlc
zPutM~?R$*x=+@U4SH2ybF?xE(A9MK$=|P>1zJKB{{DAq$D!;jzywHO)y3)RFlxgiD
z^9+_$ADqtiX`zi5c$h?9cjy5DTieAeXJ_YohilzDfeN1b^~ye=_E-K&k~wek8#KAf
z++l*)FJ(PgQd;O|_~hmo!!1{n{}Rt;#GjqaQ0F`AT$rPD`y;0FMh8ns^7jtO%tj}1
z1*diVsogK{2~6;{qljEg=#u!C@uoUbD{MIOEag%saPHiN=90B*a0b7RUa7vSv)-X7
zfra9us@LYW%J>L9D*jv^Ag_iq)~Y8u^^d(Q;p)bpPDWwxD#10=toEcIu+jF_+ROb+
zXg>`K{rOY(yv4<p1J6GHZmu|M#X5mS3n3P?v{jx^RI-J4Su)-23|stMCtY5Y%YE1L
zHx-|o!*-InOtW;i-?aK`cJ}a4roanchm7+1XG(&i;2+DG`UDZYjoA;_4y{IVP=FT{
zQc_alu&r!k!&qZ6_zIBJz{%UZ>lKDC))2e!x*N5oiL!J6`^F=hq*%b0#TdYE59LK$
z?BVJ7??=dHQ%Qw6I{_$?8@*P=S}>SQUqb3(SH78_WsfNcaDbSiJ!p1zG7C5_Ai?75
z98sJ_#O|mbPdyj=Jgcr<hMgvI{P^aKI`#62;B6A4ZhAh|n}N7bPZ;wK;;Ij>BiL7L
z>WG(Ob%;*xvTt^U-nu9jUXE^QJyv^RkY2i&d0>3qBOVNp-bj7j+0<#WaJdGG2}@^)
zA6u|!Dg4@ahIPGt7;s!6*{xIbD(z?w)gSe96F&+K4IQbmR7z$wkc}l5g1ly6FxfXG
zlGF=z3bA2>v|MeK5=XxW4rcLY_g&yy#l#_GUOF<r)x-Sp<A-*w)swDIkMj+h-CKZm
z0p4gBaC^Gtt_0u`M%@TsqZb37p#l1(B*Ffo$E@4(*jVGgN?vi5h_)$$t7*N0lAYS2
z%rI|63dsbnVHWz-xt758Gm?_|Q(nB#6i28CT1(|_zSxw(=7Jb_5(1`*R%U;2JFBww
zlA;l#05W^{6NT#1jHCXP!ILYZoZE7qPYbU#%2yJzSe6T|*>4%$1ajl}&ljeRqUH1z
zGHhR%c?B~lOVU|6UmFh-a-1cK)j+o)6}%;OLeNNY9F_Dh@2Npy-oAUi(zCnzjD?<!
zE!OM!pL5q^-jOmLPBC%u&ACQSa4t*%vS&3Kw&OVlbd7UAOZ^DUUN1N#BqAmT;9uc^
z48DZ!D58XPJ?{4Xvt0zoH~Ca~I>k(Ww2r&mYp<Kbo_k41P%uL@1f9d@%&s}sgNB7A
z+A`ZOIp_h#Q*v@eJ3Ho;p4fFDAi47u*;OhPDQ5m14!Gka6L6>D;z~4Z^*-#xVRPDy
zKgPS*YGX@+pRBL2pqFa&3kzSf>Nj)%-|Qcz$`;k-JMiZ@o&QRBpKOQ%Px<HASb7?o
zhk?PtsEXOb;8`6QGp#<n_rldkt~dcX@?s_S7&x!Lp+6>3uea^Q@I9aRXgTP_p`oWY
zdjtOoD%wt}Qi8jaI_wg7zNMU}tG<rnWbEW7Vg}jlyv66Mp{N+c63WTXpO!;O5%r+*
zS7aYSiwoOJSuMVhe6d^~LHP~qfZX+VTXmEV<?r2t1kjsHdrs9Xa9iqB3AABFj<$en
z?0NDGWb1A0frB~JI?N?`3CtTe>|dXJij3-Ld3jTAZhRqmd3mk|as&ij`|IK{zrQ>k
z*4OhGZ<{_oM!E28Jv##=R~V__c#2TbQjriNTscSwj=4ctIi^~7a-`k^;NT%Yfw`KF
zav({m!Te~mpG!))6#T7Bsfxk!nMACc)I<kVqtp}lXSHE=FQQ4{t){}j`7R+@k|}WS
z<sC1hb8Nrm!4tlqgoxZ6JOG~QT7*MyJgt1TaAu2OkC#S5e0-`%fN*ZEMwj1E2A?=^
z<;y53dCZ<uiGQl{_v^+RwnxB5E3+9l+=0vf-WxpdjdNF67%)RX7Ho3c{&&F5Oo$V3
z(M54F)qb~DrJ$^gpB-?=$GhN}Kfkhe?fUl@miyjZFVGryVUKZe%=(jATsKv@<(ugC
z<-E9WFFn<wBO$z2BY^?H?Jan{Q`2&HwSB*pse=}{6v6`mcgGY4!vxQt{{qN<@ID^O
z%gZY?(SjZB0)mA3*S?w(sXi<KF2d%;Pu&x)k5Jzkm{xZb^=^qZwv_ow3qK-_=_n6T
zAFVw~+n!^X)>qu*o@5tO7_pPpU`QQfE1gAUA$@1+vq55XE7k=z6eoG#T?nKT)$3;a
zgY?|2eKiAfZ1v^|kqezn(v00R(z%6pI;1GFq3hYo^C+}`bX~KuW@S5KbrFB3)Cor4
zP96$=Lp-)CRi&?3D0<~ogEJAVyKce$<Kw)cx<GjiSK-2$SVMZZsR>wc*#549@&n(C
z<?uP+IcLpwJp9}Jg<L4@bbHE-E1A~2TDSbm3(ntbH1&k{JIU6165q9=qoZ@OIUWvp
z^U<<zw}R#8b06_20&c3ur#R9F<<ohhE8DIj0iNQszwV{aJ^;M>TfwZ9lpbKEf9tmO
zJSfO~d~wg)$mZ(!kf*-I#nqK705pXB85}f+LpA0=QwvOk&GHpB^*0VP^!^k!dV2cs
zcI55;n@hG(d^d^VWT<xOLSBAvxX?I7?l6)^EAytZ1N)e#SZ4hQy>>L2Z91^eWRWcw
zJ#&JR$!N_t^Y&TGjZjN7x<B8;ZvSM~J}YGge{zxs!;6rBOzd-9Edea^BO^?eDQhxP
zV&!$(nry${&4{nK>~)pLy~r`=jCN~MjObT&^%T>Sw|+pJNxP|hWB4Qa_PG1~1zc33
z5RQ$AEs>rHxPw&QVlLXlVbcHPU0=&@;sj(oVq3rrNP9GOUW!OS2JQuYZHKqwyM@o#
zS?n*gf)csfc8V2puhmLR7muBRG7HwbmYPZ|e7X7z#GU9FI~c+9g(f#9YN?oq#vp;*
z><1uA4_<C=Zo;lmg}hIU!EydzX=zriZ8z+EvZ3;CC{qqt?=#PC3sVeVQBnqiyB7$;
zmPxhtYi8yT;N389bc>pA&HHVLy%k=EIc<|8<mGzy`ds73#Kc5no^(FX1Ij4_kIdZ*
z5L6Z2-C^Ux_kBnbl9C)i9MQ`ju;3JM-^mdV$Ax4Fc*JOwYBt{bIfK!Z2%ALIKQBt*
zeITOqc^%PM4ri@(hT#N8l30xuV6o`d0wtR;#P{loMFxdN;<oAohHtZm&-M4wt-J>`
zXv+3{AF;A7HT5_aL7C#kpV|knzB0|nw)UVciA@VUNHoIzX-TTSH`PZX*;%H-C{+n6
zl|>D!PZ+}R^VD_6gMG}Jv;qj@zHt*Hv}K>`x!!~yR<yA?QJwTsWy)Vf%+GWt{rc_c
z=UG{+3+K76U&7#*sE>+kqmV5!lJ<#x13?~|qR>olqW2Hv_a0B39xi4ZxNRt)a5XW_
z(J=)2#)}j}P#N`c(groyv~}xjx<SovHC<(rWahEx#|utBM|E{IU~>8Am;n$=ee3)R
zoZKR{BE_A>_C@!dgXJz*!~xjF5m<Q}!SK<hJ497g)h^+7?c1Qm)%m`XtSolU7+Rg3
z&S9S4M6pVlVJk_XWE3Hwyw%sdfM<f!a%dGaTVX7f*Dsh(mg(SLe|Yolt^dtig*2{`
zBmMQE%$j)-0OA(Bw^<O?X>DvvS=q7mUd(45ppwlIec*kz(+h~0Z^C~3<3&oAJJZ!C
zJ2MF&ef?Pl%;ZH-D!bpV9NRllqtUTq<&81@MO(fEy<+#5{;iUJv=o}>aE#vYI)%Fj
z`a|KZMT`p`;~EuF^9J)O207k+Y~kwO4}21kkQXxtM|e)Moh@Pb!UK_vcGq2s7e`v$
zI_cgEr@wMV7)pKzz<&L0idKF$7aJYU=ke3FzBA*+LBLeW$667i$aDL0)t$H3Y;;_z
zq-b-u!ghjrzqNJ!%WIBIrh0$uf5cYj%5UurP#$i!YL5K-+`Oqr+_UQJEbH?XekXa&
zjNLVKXOMSs6=Z04dgP!5XSP37^EdLeGJtO|(UgnivK@IX0>ORxRyo(o9^3>hOy%YV
z9{XAKd}Kd>C-xpk6cb`S7!<z7*Z!ABkzT>aM@M!ENO02Y1VA_@zXTSG`odl>km@tf
zHo^V+0sb4Gh)BvbPj?ehZ6+0F#_eY`cY1sXo0Su$0SSVQD`Ik{P5)EF&|t2TEfHRW
zi;4TV22}bR0Rf?qCAZ*p`IWb2((G-%xhewtra_P3k552(n(L)WQPb`9<-rOoGTW0c
ztEBR_T6Xur7boOv4NsUxA%|`H?}Lul4L-$pC(63!Yh9MT6$pZ~<u127<neQ{35E3^
zONF$6$+4mGzI3>-z!7Yho)3V1+7#Z3R8C-wfhLXSuysd4@>L{Hl4AcId#p5dWyj}8
z;cz=Jz+XRgcnJweE#paPP>Vow&l)*TT_Ngo2&jAU#^q%_3)wI;V$CxyM;IG7nDoll
z)<;9n$}yl#8M87Cr19=<cc@D~_Jl68P2qJA2Kha78yQhXXgN8t|NQwAU^{ubXCR8z
z2=P!;cex9^eVYpSK~yII4B3r4APb%=kNs~?wcc=^8sIN@t!FC6p6}`rp0xrDfZYS)
zp?$L-^7dT7FYDmI34tmqD;v4%S7kZZ;H;Qkszgyox(mPo!CvL&_1_9`_~FHvRUk-=
zR&f0K^i#4YY@cbv$YH$!C8_5odW}(jp0%`3i@!l=^?7H8M%B4%H(e2H=RdBEiRTv4
z?!s{g-S4%lF`5yb*ij~TbOio>rRicJEQF^E+A8~Mfq2ZV+w*)nru0)kP&5L0s)bgG
zO0pF`V1{~9ZI51wZ&?;f&U(G|He`_p0KXDcgKVst*_mfA`AZ2=?=lKh*DpCA*w|Hh
zXT#3s|58oxa_(SZL`E*FW~CKJW1kxCrnx&aqUIGp{87>3FY(#Z5b4WvvrhlEg!8D%
z^-GdgL8am(v~|S2p>67@m~BAme|rHey97!Pug?X@>NjU!?zPSO_&q(<TdU;+9cw^F
zNd$hPEYO~n3i{nROgbNeFrrq-Zg+`gEA`}I%yR+)jf=ZvJ%?P6{e_$3Y~($a2}pSs
zm>KZjtlpQv3qJd94GC@PrJCg#<h&%&wT_EfUEz2iSEajA>DPx-%C(cfFF4;LduIIx
z{6GLeJG*v?7@akIcxWUPuoI^nQ_Fj={|;rx@?Lgu{fE_Qz$%m!-O3gr_<k+pUSG?~
zD5jh1S8^eKY?E_0LDK+>W!$>Diq!x-zQ9-9N>()@ZwR29yA3sLmu~A07Ie8U9U<3<
z&dRxZQ?{MA6}!mQKZSx1vdhB5*MBA)FPB^%99S~iN7u|7Ea0#EmZ^u8zK&IX#uak{
ztUkpN974ETb^umneW>dUn0lLa!<M9n3Dn)JALXp+b8?w-`dI9prz9R%*ymMwzYUG-
zRom-<iukuqk_4NBf*+@nIYj5<G9EIdM_6kmmsVgRFR(rpbz%r9ji+!29usKV>P?T#
zwfmCfZfH+3D*Ky<;DxdZL`gnyXpr%{yAAmRk4K++dP*fJsakTmy5%qT5p}jbpilP~
zna3PT2f(*GgrDkoO$?p^-eEaERyeP?_;YP7R~z!qhur_d1RERM^WQg1tJzwHY~Q1Z
z@%54T&Upjt1tzx7o?EBT>$Zk#Q13iN0mhTO-{|#PlEXW@O&mF_<fQLAsLiNGhAss=
zFwL5vm2+Z$cuUDl*TmskdvL-%UuLyJ6?PKiWgd0}pNQ=@xl1ZEZ-K>)r3O5r;ERo^
zcSnpY2SlkFdsp^7OUoQvsD(Tya)I0=+jwZ`#RpzS@kKS7ApmA#F3VgGb*Mx?S?YOX
zGR&~CXO`WyM@Wi`7xM{O=A@6H?Z7;L6f&tXPK$%bmx8ump^`AZRz6#lM-sp(FZp5m
zc!jmjO4(~5>Dz-vhcFf?cxYMk_`$@deHav(LXJMM?>fxn_U3dlH@-UBZ~iY_OQl-&
z#S6kmCh*c+me8IS(2jg_(1rKG!r}<Ac$dte_8tGFzUH4-tgJ)sKD)DZ1gd`~B;LJ)
z0C|FjhKAjGj0TkDZiarR8lbgO#$q7#XO&5p)%b7!+<H)E3J9V@r*}cCBLw;!&SEse
z$IYFD@*rcMoH@HjglpoM9;<afs>okBi|z>okYsazKi%1=Hk*e;o~NUls1sfEyUn_1
zg=2{4e~wH*Y8u!5#<C486j<IKJfla}@w;d$Om8P?7_a6>t7y<6`}^5c6JFL4V2yAa
zN@`hg<TmdcWNfntme;r910zh9S_MMgYM<t4FvWbeICCX@2zs2w1n7-wvG#`q2HJg2
zYtz{Yz}%*Es;QnF_kfMLg8$NJ%apw3in?IB*Q<RuRkmbf;85Pv)_gwJiGlIcs`wR!
zCED*(<+>N|e)H^EV5_T80LbmD0)qE`%P}qlRCci7ABDQLuk}umf1}t0C-L$}|E$D{
zI9&S!kqZ|wl+HW6`LVE|VA>(&-vS!}g%M2k=I(kQ1Z+@8>s@7&3wc|`pgYlP?MK!7
z3-8h4{KqnVSF*j>R=+qOJm~2?n0ub;Com_hy{2DY(IQrdqfmNdBR{8QA-Vf<%y#W3
z`_Vu}*OTKY<DZ(h!~3}COKou@uV6bqK3BBGS2%6&_Z79Jnx6r?6_5U;HfK@Ot}UV>
z0hGr{TWgDQ%M^0z{4E}ujfNKvB!09{mv}tSbh$T?zwH5M^mXLVrCFGdUyEWP%=x#C
z1_NP#Y%{awh099P3ajDDhz<dka!O`H4hxy8xu0Iw>Dq5?dr~8rkV209urKzZ#fXab
zJMT0ed+ZKSBdSpG`btY<-n)PyAO8)Q05u5i9Szc0HE1F(Q`FQwKI%h%vAG!&c~nvy
zc6+_M|A&c{^&g0@b$&*TZymL@Q|(i3A_<tZYb>I>ro65;i|iTPLHq4ym4U*<iJ$Wn
zz}p{C-eE3t5SGf~BzPP81Zgi*kc5hd(~Ur>9-=ik?8~~j|7`{UX`5$Fr@U(AL)E{i
z`f{mUZA(!@VAWB1#`MfveQ+L39IK}XfSSO=<SJ5hS6!BYrNXAP{a+9`Wbq3hqZZEU
zHp4RyW@BwUJiIiS>=Z-VS)E$)TUr2;SXf}Kupq^hIL&BSYWb(?#TQ24(NUB&PPBZn
z{p9+bC>6dxdMMvxo$D)V!fo?njz^lsHHpFugot7}7B9vSRAYUQ+e)U;12Pm8l$OKa
zq=C1u13|r=4MPBsuyUVqI)WZNCrn;LGBPkMgLwjW3oYM3*Z;SBRju>mz3aziRT@Y%
z9r3hsfByWbQt&1sB#e)b?@Z+|v)lDK-CC<1ZuLF|hqaB8mv`#y3Q%*eI5^_U_25IW
z(R`_K5Mdc({SLc{?CYgsUiJdVaii<zQ;24161eXlK{NY3?F0!94sHOfUFKO))7)WO
z!^--Kz3Kz7`0iodTc}Q_{(l(iKNd?e)0_s~*$b)#%3V}0>AuzDjvW=*j5n~NH0s*B
zcgZ?Y6`j!)hlDJBFI5-qP9UPJln%?|jXm9MAI*U@1kl{rmK?tdR>;>li2#T^K1N}=
zNa$*I3QZukYf7)r2AQbU)XQ3rZ&4XY<nVRkWt$!H#qF+?herLd{^a@ecnZ`Xxl1_h
zoLBbRA?JyAo$4lC`ulAke<fS|iuT(LhqG9Pt-{FbJ-)D8kpd=<MIrU;_W+#iH!Xi6
z1v<zIr~n1;E>`9M?n;&4TAglkvlYI(FatplrDMCYeF*&Qom}=6l`}YQPnLt)mCw)a
z!v`AB)ZpxSbH7)h2Y+Q!&(Up~4%D5o!dHLxssW305ALQ_|J~E+_dmGV+aQg-V(hCS
z>gDz_n!7wcB&3*|q7N<Z|8cb3-CgA=)!iEqm_Vd)ao+NHHnIv^#m0}@rF3$F7=y@i
z=;ElLku&Ihl5Vc;$Is`7z<s>?SN&l~zUkKov&Ab<+^p8i@bmmP;IeDqC-Y1mo|IUK
zdOQBeqeIF$3Vh^KT15SP&}Ns(fNxNUJ!Z4+e1SmK5bw}s75g!r2j0~5F5vNFCEKUv
zj?_}tqeeccUk*03>zWs>EZHYtqx$n2+L&N2voT7cfWCWrf|y0ODT6ZY>fq_=X&WIV
zC_BD!TH>0R3`KncM(t|9Yxknv`&NmfXLE6Jadz`fNy#i^0e=QF1yaH({HLvrP*G7+
zc^%2?=iSrC-MN=5&I`YbP&g)F0{ID{9mH3vF<r%3dL}kryrocuz3BWG6&QM~Ul_A$
zCcN;|?dpJSq5jT<6*Qh->TiysesUW_HheZ_G7;5O;D)^0?Qo<NXn>ZRKe%Y$Ui!oL
z)>m<tIr?#k14~Xs!R-^Vi{f|Gz`G>Lk@QyPfhAXGYnO%%z9XYhJH?=~il*++z9m~%
zNz|;78A1hk0Q)WSk|jQE*%`4as2vmg#YVTkag6tWv7lp+thBfwrfO2n`0}Q{WR_@D
za1f%8kb;Y`6K4={<6;L!2xW>G8AIj-@#)S26BC?-N##*&^53uj+d3A<To!R?IedQ)
zWWb~m&%Y7$4cisXb}En@)vbb}q8gVq*|wV_x%h+x896y>P%Qk8CLM3}@%V&CF$B1z
z4d8tYTgS!3T)J;204@8b%bj5Z*&^9sScKE%k@?f7Pl>s#@bAmrzrP|LqM>Q*>7t>B
z1Oa%J2!>o7FCy<!&~ONy_U8D!&-Vsd>gPH^{SfiR_3r0OeEbd&3&2bmsf(o$yo<vg
z3K<z0D)12i@BGHR@A>_BmRP0Y=3~E-UJ!~OX<BOFg%$^au^ur*hMqxS>;yK2ac97;
zAg<$0LAiY@mSLlgLSST5w`ou%PuJ1>SuAgCQ26f#UWx0tVf7dtN8&M*J>oo8*hyb+
z?{fcO5KWPdoVH-WgE>xl<Tkjb#n)%qLCA_AIrl9Z$?%_!;iZ3=9jmpM0<?zq&sg!(
zyF0pYhxQrhmrmDajryNMK3b^P7u~Vl)4;xUk3|?6`?jZ4=~+9B+Qt{|&LoK4XwklQ
zV+Y)1r!`EPPQFeeKbeOg*g|1F7qb_=*aNp{lhc*4SjA+woKnTJ5@MLwV3I;Nu;q0F
zS<q@inz4Nf5Fmh=H4=f`yDp;$S!^1m^5|Ngq5%C!>HT{^GC>!{VdWX8p$AgwZaZT`
zOu<4$xsLEF_D&2y&IAtyxhC~7@XITRxyFVM^8&X;j0az4+e5YhJS4B=?r(Pt52@#g
zTXT=(wT40m&5Vr&dKk=WN#;3XGwBsp3HkP2Q?b6El+V}#KVWe)KW|9N_Y#?5zakBi
zq(At8z}SWHdhuyqM;V!{8TRxHzvXM+7E;3S<b#UFqSnarf|N;+ZI6%AstcX?cq-YC
zKEh(C*Od7OpLi~J$0uKO*BCjBJr%n<{rrmdoimo5$oJL&L`>(rmKJlKUizcrL24F~
zj^J_)-tu2P&@_q6EomV<pc>zSDUz~)1LIxv?E-{q512XweaUnR3_(u+zChoc+#C--
zBYv$DaU;1IJPZcb=zki-L^{3UBoU%uvbH?3BFp9MEk0<-{}$w1%^i&6MNl~uFozcX
zO7_nmm2-h8islTwj{&T!FcdzAO|_ch*R1Qd$d@wmuZ_D#@ug13AsZvp17A>dH0W5~
zaF)(10OBa_r_*^3v+x|Tf??cGhbQ%4DVaFYEyQ!b$+Jn``%e!P3`n7sM?c&MRm)3i
z2i|G!xsf$_@KOA$VEPNYBdGU~BK|`B8Lq!l;+6PqV4%$AWTM7B>&@Zr5T@H5t7R|l
z#@w%glQuqyhXTEbRg{?-lKQdc8MpZ_FZ(~!6^$G+H;+QbuRp}4o72_L-Il*vnkA=V
zULe>z@W_*t=cqsF{nv?x^Q{r(d6LyIS)${od`s4+WPj(g!<btC<{&&Dl)MP~u!@dv
zT=kSfP0ZJo16_RZcdFO~Mq#!GVU**oJ<c+o;2VUYvT{d*^NM!8LuNk=jN56sIqDDW
z^MixVJ0(|{iQqscJs%)T(V{5}Jt^R1Ft1|xTvgSi>`;66)_3SnhEXutKN{A|n<>-v
zO=E}0RZq$FkL5pxAoOMeH1!ZM3e0mn)3W2*oN}tc>?!Jk4v?n^0?zUPj_(lOu-E!3
z+S=02&K%{YQJ3SPg8p+v1CwSbu!ztRN+v(uB=!<1j%pLpK;{^1a8mu!J*+57_XVy0
zCI>A`Bnq^LhXFyCc89OrK7+lX6>feg<cLTm>99;EYsb4DMn4?7BF=hgB%Rt@qvHyk
z-FE3EKtHpe(pY#-h^`dP_Ba2-GwT-`Jr|EV%jbBiJ`XAE##;V94~!Ya)@c&>db^1N
zt2WQaWD=hx;{2h;A&SCiZW^S4Jbfpj{n~`8?b~33N7$7jN}xUtPOc<C`Z-5~I8<}}
z)Ho1uglp`N0>H2uK|t`|n|<~e7r-1G`}(Oe9q_vUFHc%JHA2W%Z1UsfKVG}L2rjHw
zT<Uc1{K6U^3eRt}Q09_rGGz?m<`xZ^I-Nb*d`LQ`9{JEx9>ob^Ol^{zOLH2wY+&54
z^43eVklTg?&F@)`wJT-TTeRfFL}h*bNAFgMpLyWntg@EQazqg1i9aTe&hoOSI;qNw
zbT|B?vH3r!dds*dzo-jz2<dJqrI8K+2|-dql<r2lySoIWML-(q?rs#MySux)@0tI7
z-`{)hhnX)hGtYCLbN1PL?X}j)*ZAY!Hj<Ck=|VbL!8wbPGK7+;S_BxwzmR-1v&XjH
zjb|dG9UGh_T`*?|qP8s@G%hw+!l60`n+^%3$H2XuExfS^S39gJV<HH*f@%NCF*#SG
z6fWnw#rMt@!8zSurU8qf*gM!u<0TRrlJnh7Df47PgIGk62rPxbA78()aD+ks$p-o|
zNzJw_kxaW!NE{#O1AX2+7RLbs$-?j<CzfWNwGr5pAZR4q-|-n$^xhI9BO#SHK}G;r
zIDZJXsgwKV=cPq$91&L_RT&D`jVC0*Nd4tjD<qQ6Y!<zn_WeEgEekD7H3yMWOnse7
zVOHo>AVLS+Z>6O<1^G<2!f=l}Ic<r6dvgjTc|{eyM`5*QQCaDD0M6-?DQkzxsaFwZ
zE%i>nJJl#`AG&Q(M-$*vNiA|9%dWFnh9@Up=Wm#oDVh5i|4rBf4FPnj#u4(_gQ@r<
zyIeGt*%uZ3B$Odn!7xr1T6!NQi*~dE>^$TkZ^LXo7fh9$5><}JuRH_hDB_->LM;qy
zUC0fnwXN!OLHRt-V*8PjvQGlG$>rGD#U&6+XQDNfINC;FEqgTHbEghNfl#E=<=ro}
zvQUK8BahM(y_-jPr`49D*QA4T;_Ji`BIzn@6?A_lGEvPbng`8;0DIL(G#;RwdGE{j
zsRg*Co_+@A+YpN>F>CF8Dh@dJ*fu`tL@&1_wB$zs@&VT}Cn~U|20Nbsz}WYZg%V9B
zeoBZh5#w=w#-_@apJ_qm^2s3l?qSgj#al>T9R3@}^p*VIget8ch~ek~y)DgSG+?z(
z-(4MQW(F7K&;O6~VE)j0v2oU#A7%B8?#1<k8(D)W(7u>;Pnu6(53OHCk@#vN2)Du8
zTuGb8Xs;}7qhbbgM=ai_0LIy(yqCH>;hF>?e|xctWWTQ>VTVDG@#u))5D(q!z?1Bp
zHvk6cmuW+@Df1W7p=x3bvZ;cl<5+dg{iS9_ik$jOnxSZKZ&aiuK_f6J!zyKcT|U!x
zTzRkG`~{l%!*>5gq#tR`a%?W`bJ>AQ1ta4tW7d$iWk^RX_6RMIfstSW>PNrZ)%>5}
zuH$ZDv2QN-K)fR|;bi^F9!jxxZmRYTQ3(p-t4&3s@Qi(fMgH=476kBmCeHWPSl>x%
zYw3)mYVZ^v>Y#W~E_$+a737)x7C@^{%lO0#1MDCi)bWp_&_i}r3*<a0_5vO#)tYg)
zs}xj81ySdtM3<+Qc9O8}$oNptxs_7OBau|hiY1&cACXOTIde&h7c7Koz8J;1n+df|
z#|QoShlO8~5#NGr%6M}5vvPju>qh%8uR^&}`te4Jhe!7sy_PpMOk5zW;Ts9Gf<w9~
z!Hpn!fk%Xei!18x&X*BJ@e^z?KTz`X^O;t^?oZk0ya^OyBYIuYa#yLAQBEZC$)5da
zA80=coXoQvtt(uL=GyYLTQIk1=Y^cv3Ac(>P~qY{JnD2sCW7DRX%TKQ+3d@!L$F%S
z=GpOc4Dzapc!N%65#G6B9}!UkqrNOr{Vk;UwzfY*9-4(NjEO5DaZ{<uMY4>gKUhp1
z%{4A(LI(;-Fj8jitFL;bTj?p)Ce;$TAz@*YUbLY<z{OLT|2g#F#Kpz6s;kMqc=L*#
z?dv}*7nhGn^8Mc>l;spyI=ytCx#t|Nc*6Z$`5P<(V3jbYLnN9$53Wn~a7-S_<!Q6E
z*J<jPbvL((^uAI0+bX1icw8E6Q`;WnP_xLm+NVwyrvMDoJ#t@(Oo$m{0AkJ~U^h=4
zglvfF?8fTuE=o^ADiV^H@E}rxw%*r2<h3V=y77zf5$fu6A!^s##Y0z@F>mB(pS3Fh
zsi1YKxZCJeSx5~cDd8FcAe;y_JUY4qgG4CN1Ew5g?$&LOp#E9*y4IdQOtxvA{g3OL
zuK(f;Wisgd&+tz+s3>wx&rdB%n~KiV1MrwpezfxDbqk|tYzg@%2c@Tbu2!1O+MD+-
zdwr)t^BEQnK&8oA%#WmNt>h%wckfMiVTpW=5!c+@Ds98Z()*jmb2O@{&l$d{(v~0X
zNUn|N^^Zt?f4G`wSr=1SpVUJ0?d;U+mh+~(w@4xU2wfrjDRtrD{>g}QLb>-2D(2Xs
z#C?4s*HeY`ZD-zskmq^$I@-b)j8K;EAu0s)@o1>20|AN!d?K*{!0;_HvNjo(o|iXC
zRo4sOcV<RSB8vPF%I6j<0ZTaCr9E<;0h1?x@f^@z{|&WrjZRg(5#U8$(zOsGv@Qy^
zT}Cn`E@`zP%xm=w{R`Xl#oWf{hG_%;_WDmS|3fum>&(uLYajAW&-KZt*swe7)|{N7
zImQxVSLoyW6Nx)FZ~V0@NoGW#+u`w4nU1ydnqO}_v@+$XgrxIih^_c{*pOqlo^Nf$
zq^g&_)5jXA^{TJ<hbrpV1m>CZ@)+)m=~W82G7}%HDVpp5*v%C*Q*x!TG6Z9%w#@T@
zdv5IkH@&*L>Uq1F@F|HSD4dXWtkuio>G9adkUPPO@YCndTL4p80qwTx#p}%k-8~j!
zCh)PNX{DPL6MB{c*XM$8a}K9<?3=}R*Muu2p%}7&X`;Vv_b$4Wg)K{DlI8a3v{iS6
zwFxUsTC0xHNPH4L`URwCjJ4}QS9?0@y#w-7-OHn5!<UaJgA98#Ofx<%s4nIx=u~|u
z42>N10VelUfjTNH{@u&V3Q4?|HZUvOa3^N=lW)OtSRL10nWj29QM*n%M;VkexQzM{
zrCJk2&a6K#AHP&uuIA-@Wxng)CHfnwCT2~qv38o6GQBDbq4hyobPzc}I&}6uIkC8V
zjwV=_o<BZv>}k6!IyK)PH7$T%FVI@E8Fqh1>0#Wy^Eqz0CIUP;x<(LzYPJAI_;a60
z)o#ws)!~JJSA@;x1o~S5$6hK=2P&-(5Rh&>dEdn0dvIkdH5x1mE7uQC|AwFpmV5F%
z*w-Jx@=+4q)%f+cOaw*=fv$s9DBC@qR_O-XF+6fa^CTxlHG76&kqp&tHBx~|`eD<T
zPP+Y$y{`gxueB%T0l{dHb{}=%WZhcXBJ-^Q8;4i~NvykQarZBgR{FP+vO}sS{!dfF
zKm;4l*PB#2nCT-@(?|zD`y$#?&8n}0k0f1va&4i{3y{IeoS>{U%rmvE4ghG_1Kv#@
z7X*rKMw0S&wp=gm7g$55d3l!eO%sC>l6j_{Vqs5{k_|i%bDX}!M?v2|5ewoun3kt+
z_><+G=QVq4d9yEaW3Bmj(SbLp+*t_pJzL)@IEf#t0HWS@cm&*hVGt~ZXH43Uw|)|)
z>QM^n>Ln+)-Ha9<ECiN{pW9i;c7HJNC%hPDmg92l%R?Ptk$aJteoJt*`_ASjP6d56
zlYNLVosTOBQR3EHEayZUaeQ*QdA#@8{^I*O*cxu=qJEN_&sNenY!Bs>mq&vJ%2b)|
zFfb#XZSxTV)q>~wAWyZ`{8XGaP~U6BcuV2;OqzG;4`1d85|>g<@1A_N$3BvF_G`Vm
zvZuQb%QA5FYLK?rko0@Gq=lTv>Q<(^#8Hg<;8i#`OH5XHgo+VpeOg$5JrL8XoFMk?
z>|&#We4^D`9FlGH?pdp-xN$vFr0NdM>8Cq&8<KI^$n-0pvEMJ6ZQ;J}Y7od05J3);
z+ZRPwdqX4lk&rUQBVR@IgS&7I`KSO<Hs(l@1e`buiVhN1Ixy*uHpGW(35OyMV3Ws{
z{B%v$4D}Vh9A6Z$griK>A+5-8F*9h^e3@jh&B<4r<aBi`-oD?k0<^C*{k_`1YmKk-
zv0AfiU@xrD=OcM-2uTolF0Q2$JiB{Q`Q}v}7Ag$NYC|7II6bRgkdtb+Oc(4MbgfQ_
zT^d#RcBozNQz0#5zbve3!HoKUFHq|Y6yu1dg9$o|SDg(Xe*daJoLjI<cD4iF3BAu{
z;)Ha^hDLr8HT|QiyR(<S>^q-b275Q!@@3H42e1LbeVrSsx|A})X1NTi>n76>5=C#n
zTE8>rzG3|R=?Rr*e?}?ieEbd^ltl{ey;I1I2KECI#a^=UXqpwP7Ra^eXgebf%TJL~
z8uP~}dBTHXtn@#SEHgA8^mjW3P2MX}@TxZCd~}q5f#(%7iIV4-eOaNQ3_L7e>5Kru
z1N)0kFDj+G+DpV$S6V#H*DTWFh-gO{1eE*9xqLm??(+e2y`?n%C=&&p=JKPkR*n<a
zfbB|vjV8)p&?sICkI<gal6D*iC8qq>+E|5QCJ~9Ff45Kl(b^4e&b3zPG9G4UT*E7L
zFFaf!NMGpjuUpek^p#-mrCmiQwNV|m;VYuhA?!{%C+pnSz_CnZv5<~U0jt9z&ZF`b
zd-VD}!wpU)q^No`2p|exhM~QCns2s{RA&Z-CDvsEa8ZN8Qb0HA^uAk<4U3FS7Qo%0
zVnc*O#Gt07?OxOde8Dny$T(u>LAq9*6*-(Jzk9Qc$i%;ILxmYz+aw2he|MJ=jI|Eu
z-^71MsA0>?V>7}97bG$P10w<Ql_lMMPY$e!$cLi2J)ejVRBs~#&u8c6NaWHRM3Amg
zxjJ6ac75E}Yf5dTkD;&yMB1-_1tDU058M85dk(l^T;6FUHq9}BDLh`BWO)d_SGGm7
zLsqpzHJP-|3=cW#yYgy-$W5d0VTLNBQ)ZI04t}2%9!!g7(c-gHhi`4UYg274S@Kky
zz#4GLu(0)b=kb!@xZL5SukgdTsjI2urQ!;@DXP%H#8j<}!?)<Q-y*J`#MlG`5P(K}
zXgxaVH@88x6`GgE?Nu%s7_+6#s~5fiqXx96H~+(qx|gNJ1>G7rJ-|f7!^4}21WdT7
z*jOB(u{P8cxVhUeQ?IwBc=P6s<Iw`hY4E^+CXT@25O=s}u8z!B-gkXs%Z398=w=~d
zRw)v?Lfof-`Dy1bg(g}&YCaGaKAAXuxj4gd1lJO<yQ~r@-WO0_qt+D#%_(5h5=|kO
z?rK-@F$3KME8?ns>$x+GM7_MAXwXU66#8TWm{_adz*7eLXIXuH3V^)BO!&70zT8JZ
z|6wPZP{AF>)8+YJE&xEGK&!GIK$4(|gwD>+4h&NRS61`@(X|I`Xnz5jk}Y`<)b+7|
z$fwVdw6LfQ(4a72^$``PJof{Yx6oIAUDC;pDewf8XReN<pmX?@H?f@qDlnd;{sn3W
zF2JoXEB369E*h)*EABsvz8?8a55M7N&i>z<@^ou@Xm?kU1T>$)Mk;pbaDSmFC4~$U
zCN97#paN}}8jESr)FGS)o%obzK-vAe*%uG>G>K<a<p2snX=&-5xf({GfOXuPK=!@3
zaD?`~U0f8AD4JO)i8Q#BR<A*aQtZ%d5-Tkkwu@gGEsBR25LsMl3yvh-f!{31m*YZS
zemnO*?m1`fZE)LXzaoASxrW29Iq8|hIz<ZIAYMV2>&g|9FR*X!{&}|z@>iPI?REwh
zptqUA>#zfe%;0H56QKp~`$K>tpI*(|aJpZZLRxRuP;)=0?j4mv*I!mz7;->UX)rjo
zSGeCGys%e3Zh{CT#=3ukizEy95sHRKaBk7?p{4SPy7h7ZpUslMycKi9JrF?78>l{2
zsA@aKA6^09sKp6>=v95=@Pip)lE7#G84A+S5juLvuYQw!=uifJaR5u|7ng-)9UPoC
zBGHd1Ec)jeC@1Ey%4GPzV(7oy`tR)!S{C6c)m~CDGWHwY1Gn9#C}vg$hQci}5DotC
z|0wpbff5tgjPyBsfNn0!Jga);pCiL_;B!>h<hZ+*yY&B#)&G9tG5{KHfc?t+A=H^0
zpyfUK^8tvM9ERO!^Vfj;UAZXeevyB81*mU}b07crWB>hR=Q2=9>H%cwQz9D-q-yrl
z_0eJjXt1fbdh*_1A8&)MAGEImb!P`pJ`|+g5RhI2b16g1;VRQn%K5_r#k|I*<p1CG
z7wQ9}VL&kgtX9Q&ZETkM(y^kr4!spXb)<UpCKTW=5H^=1Ex?f6KVb)?E+as7C5kZG
z$M(9rSS>1Rh0XAJemXkJ{cnW-=k1$0SR#poTXRRjBkdq1h5+pwlWuDYn4o7NLJxi=
znsswl)4-bM=`_lR&HKR>4M8->5VQ`j#isw8_5btMxvj=IY&cs3x8G2u$EIgGhGrPG
zL}WlR;XExywgm)7axoB)RO*9u0oa@lcNb<rfr0n|IMJIeHS@~L%cr@!{eMUB#EXUo
z&!^2db|byRyu{UVml8HOg4XtdOV81R+eIGJb+e9^gCiC&>De83<bpBCq(Ov$l3f4q
zOXS<sOA>{izs2fK`A-hng$KoQ3CTZPZYvWadj6O+H-fUX(*&Rdcd(HR0^jLd5c?|B
zn3U&ScE?`C^@=$=JI5OW;sh9S@AKqkCElO0vHrbZ*E$eiUHp6Q(!Cj|i~%HLa1S+$
z=>PY-&=?cT+-5MFYQM?fgZn_E$)i8pGR~h`=m$K_B_fE$6%=rmeV*R~hj(g5#&D<$
z4sd-(z#x^>(a{O&S+MOw2h{9Zn`NEVr-6Y_k!1YAfKCg1`PCSieZjPWt(qP96N>x(
z>AA4q9EfoNRk`_MOy;fSO!S{Wf99ZC`u`gSSMlCPWxl!6E4?oZ>)^`24*ueLYc=;H
zf`rEaYziza#{W5KoIf9=x_BxnFLk$Z?!JdtpqRjNKbGZ_l!xy)I%;g)n;UV`zTZ1y
zxT+~NB2?8zEyc$xP{;=ClaOe5$>Suq=NqHfMtv>4)t`p0&qMztadMi_Ivl^>R!z&p
zPv)`N0<5coGoo<9gDmmz!i841b7Kg|PCp6@!{mKQFSUOLSS;KJup?<-I(xz<xIt!q
z*ViWko-HR>Y}}SJE1_@b@lb+9GUaaIjrFe0W*pJRDmzSljoH~U(5pkwFAl(887oMi
zlcPF#`%jmV5VG^vbu;DY@xtzWd07e1Z{H5C%hE-Onjyo(H(Dz=t$_X7^J4TP1gcy%
z9{i=<S_ZI8X!pIo9*|B?KLTd_CqP+++K(WvfW>0~Jl_*uae|lz5qvlig?8vbe5&8g
zO@@TB+RT*ufz}-^VIG0tm8~XhyCrv(&9b22%?cd!1&&rIT1bzMZU=m`!_r;oon#T`
zY~p-a@WHSTf6YNOM<;Tug0-@}JNpI?#bN`$Bcxydc($<a)u#X>4}Cq92V*w)LPNF8
ze6CH}8F$#LG{0gIB-FqO_?>xTe>ef2vSx#uY5N(6+`U7@alKhVDcA>REO__q4H@VV
zpajvLonJ)oB(oFZ>3usJf^P?V5)Aax3*n!c!IWY1x^rCeJeS=XP9Fx73Fb5kbiSB@
zNQ9D~KRG|2{sRuI*Zq~zzi5h^+uKvXhtc^AIaq2T@ddfgD49)(&idJ=p0hLeviChR
zYx4;#9XPAN?imogI>5K>PL+@#1awS}%R{dOA0xM;%_k`QR@p|^3Cnv1)&(=5_B1Uq
zNhGseT<T?WD;795>AmrxK0Uo3;zvIX<2QJByiVAHYP5@7@#-E#t;=X_uV5E{HTme!
zY@S5Du<aZ=Z7W^7NiU7`fld$l=7bX@uXS_;v`qcdq;TAs2E}Hy1GBRz;4TH`SR6PV
zL8;*p%*It4pL-H0d<~@dpyG{AO6raKA{d&U{tB1&lO71MVATT=5IBthMh`SsOFcvd
z{bd9o;8Ci#xN`yiIN9jv=mO}Es{}j)%YRQG<#9Y-(th!bj}p`=O3H^>#+?81q0W{9
zQzseo<Jcm9*`2g-sNXsv+&kB$56+nYTRbt^I<9{8P*?s)ZTW8R?v@^}wR@Jyb$9z$
zhYxCl_{k8f{r86CApS$dXXb-<$orA?-hwDEBn|FN5f1oKcX_WYnZS>U=yzZdL{d=5
zpSv2K|HNA9AP_^;s(+38gI4y}3@HHI;64}|8^_Q0e18eo?*9T6GE;9y1^JI|-}mDU
z;6&P;Wiz$5KHXVB?k;xMi+AstJ|9A$GKO^TyEniRkQm&a@=-FM69Z7Zuz`K0h3OfW
zocw4YK^6K$KlgoNa|rrJV)~M2-Tmae-mBTgo(0vo1l9Ps!FC|txw!zDtt4v9K;<1_
z+{>TG<AUD1Iy=kVPUrIUp6Hw>>G5((d#(#TjH~e^um8xV-WXR&KA0Mr1(7C-tTNc6
z<QO_G>c{-3jwS+}^+3TA0UjRkobxen&|eDA*6H#Q89@eaw7JH}6bre2uC&r%>qQh7
zGeL-XthF|;O0}C{DeuqsiptJ_z!PXUMG+yvXfI)bpH)$4RFnxgZ~+h*W@WkQ1NlPi
z!otE5-+OQaH~n0o^@1v85(1u~sSzM2B^ddFjUsR~pYkJkzK@B@$g#?`>AnM6J?kGB
z8l2PY4?Xu7J%-wsy#wEyJX&~8W8_M`*!?A%mXzGdBQ+eeXIMfy2#E_2OBczk?~FOO
zWfGmAX*N(h4rcJEQpdq~*??m^*>{R2wZVv2`kGmD>H{OD68cjB-bc@JwAQ8kMi?;E
zo#fsvcLoP@T|?i(R(#T*i1?lCrmM}x6IeKe!UdwZK*EdgA&Oji*}>KI9mqB}!1DVa
z=RAi=WmRk<dSwEz0I1p6qCtgH1(X@S{oW9i`-uX$6~JQK28Uz(Q8N4Zl8ERZIDz9I
zNEBtw&COHk@<9-(*;GSdnq%HdbQ8xSfIr01vn0$hv044_tM&ALSU8HHG)cy5Q-r*^
zO=)h)N*`8G9bJ}0SVlP|!xlD(=B_(=P)-RUb7>bpSI(8py(|iGOwSc@dw5;J@86>(
z|C{M#;D{i$?}5Q7r-kw)Jd~cErLU;M!l9O6`A<?qts3>0M3){uIW(B<6QRK`T#)o-
zk;u2+P5+AL_m+4Fj_CH8nAEX3FpSy11s;K}my@cYQDwxQy+cD=Ku-Xx*BOY+MeZ{$
z%X13;d<IB7cDrLVxw6U5w}K!I&X$Tx1k9Eajhf0^(y(6>1*ASe<_MlX;QpF`qe6f+
zz$r)$idNc0GEaz%^^Uk*pXg>>l^D(;XW{z$$OYX`63Ub?s)8gX8~dcoSc!FTC}fgi
zdOx`qu9FuzXvE=OYlIvpz#yHuqzi<KtfKC;2yX4}{2T2%8b29uIq4_go#IUX`=A<r
z{m}^aQR(c<DzVC|%EK7>)Z@DR&u_tuFAq<D^@LkWby8t+nr=(x?S;59?eh9*`V+;;
z>GsJhPVzm!LTLj7O2D-tfk2}*$cI1(17a3nZvajFe*FB&;e4Q0RIv_dVQ?paF$T+K
z@aG%D^%sx?jg2Wl6FkUBkt(ZdD_MHS8*+%BpWk;f{+}RE!(I6+0j6J;{>i>zV$0+h
zb;zXTbEh^IIsAH4gPXkyJx|A6SNUyXwKPe@NHI2=q+c4dR`N;|$|#SZ!nUh!51OT5
zsn^3&m7H7_EPGhvIK2{|^pKXlz!aLJHSIRAbIC++neNG#+ird8MLju+{T0hllj*=&
z#!8?xp;${gEy&#j!N?2eiG@aZOG8t<fMsPWQp^hjKp5-=H6W0wz%4dTqBTO3X;2Ri
z?HMu)LI?`!kinvL1d#KeKg5EtDEkv{oF)Lu3juW(l#Byq6$AIF?VeXY+4=uo7_ot9
zy6XF2aAO_!aFpNaGsJCHa!sA&c5?k{HBPyFzxu()J<^e(xeiZ9H@^)<9+NEYBI0;r
zg6llL4DaVo(cRnR8V?qFyvX|8riXp{auW!xt8k@aX8srOy!XhCzm-k#)7_MTor;mS
ztm5wZ?1prDKoq(~FF`xg;H;fZz?;$=NMsKT3c@BO9RS?io3nn_S>P`Muxgc8B*et-
zyvS_E14IvxMN_3fH-<~h5KjngWLqqM!<5g%@CE0f*l+y9h7QcUO`EK|JT|b5Txr}d
z?J|995A0d&fxck8#zGaX@IWk6uHLVN?(M${B{>z0<jBs%p&_@!HI5luF~>89@NA7%
zL!1jU=ELjBH}rnfgp%#c1ouN>wR`L99m6jcy;GBg?a@*icU&#QyN57sCtS?WdF=i!
z%a5|{j%zsfBpg=>9d&#6d5C0=$eee_UMx9upU!lz0!i|@4nOCYHbg(>+{+-9iXpes
zU<qULVj(m-mT$vfEVCVroh22>lnEY6ifs-fRV`kF`El}w$+&fjjW`0DviHW(?RzhQ
z05{V-tj89ee%$UUAs@h}>px#@)&*>70C<l90DWVywL`!f0AnEq>Mtp2=`~<RK=J;4
zOPGE+m`U%F75SX(F86lk42EC2u94h0auo|vKDQmQJ($o4+%l6^SH`3_4SJs}cdnk7
z8?-rWsgwPNIU?UuBN1Sd>@Lu<R`!3OU{RP?3&-$i?mk5cqC#u8f?Lh?o{C!Ck-O=#
z-KOEi4-4)-4*6&IH3}lcI|S>9Aaa=L!feP)OooqgjTTbW^v*~c9i{5!mUY7AJf^1D
zmhrTiTCFMykB%;cHz+MF$0eHdw~ePm&RE8OsoY1~Fhs})!07({G0OO2XB3Lo167oc
z=c)J@u<@3WS<eh1UjwhgGorZOzzT9AHE0qnS};Uvn1pK|dq32f9Nzv_)L}<Ku$kkP
zLFa2qqPagEE42nGx$*djdSX3|!;{3aBpsa7?F0C>0##d&B?hadnbBCAzoyHM(ace+
z>fVu?NHB<9s`S6XyJC9{^Hek^wZ_h@iQ_S7kRh6N*ZQUzI6LWW!i>VOGQ{s2On<Fs
z!1r$+ktmiYdf3#sG0oB1FN(uMf99FECF4$VyX!_IqLH%UeM;tzWpFZ~r6I;ywhDg1
zZ>4&B%#X@C9AHwaoF*7B*O_ZL*IjFhCKaK@{|d|rFdkEO8n780K$s?QF(L}p+=6=b
z3^?bW%$lT65=~c|DW<vay|HfCBruu-<Co&Uw71u&VwmtlMz3AvWZ8cfjqq$^+m)7S
zvFO<8vo;>Ho~qRr?(bFBxD`3#gDOd)7ozuKrpXx^i-+vFagJUo{gpy29PqT)_QNT7
z=us<>oyoEBDmnUML+<e7q2c*W7u4-aYP%Xex&<KxoQ)-k-)sO^A{~xyFvgcw)iQ5}
zpN(Lt|3F~&gQHsS1KIP?nrX1-tB%$LC5>7~#0Bnl=J7L)MrzUPt@6k0M>}tcdM96v
zFAN~~imblTiiMQxu@UV|77+rmsE-?1$+x?C>8{|cwV5VWN?kKEGw2ZPj^}ZL2mw83
zO5%I|^5<y~xvxlv1C8#*&T`AexEE8CJMtrO1SUnI2u=<qdJx(P_8Rq<+_e}U7KNpf
zEQ)T?KA8JI&{yN)L3~KzY?y<qxMLxsnZRPPM*#;Vs4P~O2_|ctFmnM7IcB6KNXeJc
zGo5!{>}p8mg0<+wr{nYniNfg?bht6}Z5HOolh<DBa9WMOE!3E1e|sGi!hvk?-#wQS
zn&fpaV!s$|FR|Yf2QAG~1g-;>5v`ho93RDyk>4o&b_L9vk-GTvP0n!XO?US*|C^-0
z64E%d+?@?&zoAQExznmYzPy~xy7kr=`idlf-v{W8Zot8To`C^Y51=1RI?Y5?>J~Gf
zLEVUA%<0*LOmI}_?%|wC{o?0AOavI_$aF?QoKo{LS<W__{YHxzjqxj&7Zfm9g`BT5
zY+uPzuFUT1b<`Sglb^49rU@BWdA_=@{>kFUc6qJ%_QSw8snhFZbK>gZImmDRj)|o{
z()k45WSaPAqVnWaR+gaQsdijk<@oZY4{1EKvU`l~0|_$yt7^fHGTdMN;ef8Bd^Nwg
zEy_qT#PSo7qXou#qez_Xm*Fw4uN$5GZQq*JjaW81h7|1|n|!h*efa`vLRNkDDaz{i
z4xi^CqVpYtkMS1?(1lRE9^)P~Rib%5Cc}CPD0t_9%X|iQjw+x8|ME#ZQ_dBps%`@9
zNi{bg@2LTBCGtx69=DzhN-bo|rg3~t$xz$KwB}#xQnbyS5}AyYG(7oAO7U901ASW=
zCnM9KuN@h+bNmc=bI2C~F^4iQCpZYJUQ+(fLPEI^k+%ryj2M{JkiWHh5%-edF!uZs
zjw-&lzO2%~M|t*@2dITiSPOie&HbsLis%L{N2`J?wiVOq852HS9F5x+d_WshWN1#v
zdh57~zoRwXU_rKXNQ5@h7V(f6W4MHWrnV&JT|p#bvN1>Qz=po0!xSp}%tzbkGYh9%
zQs<pMi`I4&a?X45Pc2>vT9Am#3e1)793Kw@r*a}LbG)Raq=8>t{c+IxlN?yN4b1Pl
zru4$6mj$)SIyYw`UBvx8&eAS7t=pN^L_)=r$~;|6QKa~7#Hp_O@SUxqskMhxD#7_6
zd~zeyIVCW?w(Ag1UoY%@D?dN1W96PiH>eXkWJopEtAYShT5mOvtk(KU!qIkp1J8Qh
zG`RKj<Crq+>ycNqj<}o}AK#+H_vyY!sfQf1mb{E!w8Po54oSZ67`eJir(#fwjnxQM
zc!AHwHo}I}=ukJ0`K0;o7e&XdJekF_h6#&QeB1BOg<dB59W046C0AuR|1E+)7I>)1
zZ_PSEs{j5yItUZSDzC<LUb=3@!XwO?k6>MLN=HBMm7dk7h%}TGqlNYTQK~aC9NjHO
zO^v<S{k!*9&zC>=g92b}q%=dyn6ZW-7I+Dr9~n1X;^F0P7-*|@26g)qjuN&aNNPV<
z(bX@^=B$#o+0vmMXJqSoVz&g>PIa5$EA40}i;?H4e?Rt7Ej>a3TBv@zc|#G{RX_PP
z@paimzw=ofNz9XgjqW=ybYNzPs*`HcWRaCk`qU(`#XayDC&Mtd8{QX21TG^d*^>fq
zDX16T0>kW8TV*7&Z%|Y9aRR}izn!)cp)Yi{7iQIVf@<hD>A3n!R{=$UxA>2~w&w|V
z?zp4E!K=n1XTs<B%|F?yP$0uW-6JWISxe33AFJ?hpvc{$O2W3^mvOF=HqKJ(ka=g&
z_{KH<aK*={RSi}97w?w@rOXqA0b{jB%q=UzSJ<d9-fZvib1xLU(T>{D4&JYqtDSNr
zxkbmo*~PG=C3YVE37nkhsr|Svt+8rSmA&WW!sRepFInHp18g=^c-^Ta;1>8rrdWlJ
zeSxq4HR$ixVfY5p4-X6zXntwGPNsQn)1l`}KDZj(TFTo!*q&EJCP|wF6Cj!IXu*<-
zYb@3yRIc!GXNIMGFC#I2{=Hk0zU6jfe<S@K-EPC(M*3&n*kcW=rl@O;IIbFD?+WTq
zi$)ryU}A*b`AI{-YKT%OrjXM%Y95mAydZToYE^<!r7M$!MGYD1r)}bnWd_&BNve~A
z`hq61N7K+&b5@MJ{|9>2=<CUdL4%FQp+%P7@M@;n@NRLDwkft;hrth9tB&dy7Mb~Y
zQ{=IOrJ$Bc9H8_vDlw?7)*86EO7ZLI5&g|p_oYl&UZL>$Ax%3U2hu7z{dLA^?(`kp
z%C-y5Zp(9usfp%2dCvB~3hze#Pe+>*yYBcB0Yv$cd(;;3sBQ-Ja|~1bvdTUSC_);i
z4{XUa?B&m-4|Yt|p$<p;X^pSbqeKVnnQx4LI|h^Uv8KW33y(AucuozHbb<7%t*AsO
zh^AncLW)n;E9tvdgRwFG2h$9^mFn-8k4GfTJ1wn#c-s@)FmeqU9r)`|2xav?A0EuC
z6MEUsODCAOp(BD*U;TYpD##zo27a!eoIBUohFoozeo}ZV;K2M~k|T+?OTi&_VTz8x
zPHU}Qph3q>vZSS?sQdQ)$5_}ROyE%XwaR=|?pN!ykfRO_1h2L0%ltvVkuRyxF>62`
zEjL<(!53^I$dfmY)_7PEjR;<vz9hrI^6rg>r35ucr$6<(`z3`Q;o64zP2!oNoGRn}
zRQQ|Fw-G#5mx1ulhfa@pKiZpQ^tvyH4*&4w)7<EWqJJ!W@&REI(MAA)=Q3_+r2NB;
zz4qXT!j9MWMm$%S+M)3e2D><mw$)o+C5Ga^Dl_jL-%cg@q9Wt*cFy5oXS3#)mF9t=
z$jpz7V$lR))GPBNs?|E0WU|wC?p(UZI=flJ<k`Zx%amo=w5SN)f{=9jORWr{1tA=f
zV$*AaAMGTlhsojE4AJnyo9KQ^EP->*%%HH5@UGbTXR#{DXUV@icl*{6?L!m%;X0LQ
z^OrucRZ`f;RlgX82J5J+P|DF4<*rfAZV|^k7I8@m^%$Hdi@A(#^Hg~ww@Xq%eA#LT
z`V!Rs+++buB`fhHUsJg&n($;6xaF*ytUeqvHngNZ7$mZ*`ae&_ZhYii32LD1osCAu
z9HWz(YjIoTqf>Cw`U3ZuDIUkyoCpmfo;dLkuu#ZzY#6N&>Ws5r2x^7sJWk=xbtlEt
zYQ4YvHFfhf=V~vd_Vna}tCj~7)p_ayzFfjNQ|Z0tx2ZQ4IAn!3%_6HM2ZGGBQ5v)E
z3r33aN^FgB4@EiS21*|%V^;gC!{R^lT1Zj&2*v2QeJy^#D3@KoB+1r1@~m~(D##}@
zQ$KSV3H$w-xn)QDekPmJgWbX_h{w0hG2zoL4gNC~-%i;>sf}mrRp0X1&s|Hq0>`o4
z-IFu%()|Lfls0*Ro<09f_rjo9``zkM2INBMvJdu)g(!f#`fD19DAB=qet<a4&lU8h
z<?_8tZt9+!MfUa%iEdIHVYL~lCefn$KtPGUMD=Y=<toy`n6Ad?rK2Ro(K`3P8_V7o
zSUPl?eisK|DhJ7phO=-A2BjgB$2ldgv<vY@>aG#z@-^S}3dC{h#Es+ibSY8VH^k)0
zuRGSdcGpC4acMR7GX0c$jCY#J^S1LbaK2=@JZ>DE+w@Q$2Cr-UW5L2bw;oKr2)qou
zAe}K4zv>}QY8azo%P+Cm-WaTw@rAL{Tg09>{El$*s2rW2Inzu2<}{XGJ0oX&lt~=>
zei#_qdg@ZjLd$?TIWh>?5Z0;1I;y6nnw&0SM=~IZ69A>8!hh>NGM`2!7xH7rnykYI
zOOZ=+gD|`~N(DuF#UQBGUgXDBkhd1~g;mL;6CR3<#PR7-<cIU_-zPWICzlAjl6=v$
zbzYSyX?Fw|`x7OlHblPjh-TwqOI`?1b)%b#3a&(OI7lS14EQB28dGRb0Z;pOdO4@J
zf9);FJ}F8h9*G>^PRwpEcdxR&{@XC@CLIu%T~;waG&iy5nrfk|OUX=<mTV|=v)s%s
zz)gNvXE$W=`H<?boW`}N(dNmq?yA;L2c7>K8J8rJFrb8K_Ls;6+`7@=_pum3Ih0@G
zaCGk(W(gwxZm2Lalkk%pqP~amD7Oh#vmSfCi+^=KFKl~9nNp`1Lr4m_jvPr>lBaKu
z-{Ol}CUAPUR8zi7<XOLP^+<TWdimnykC^Fhmqtq-Yv;=dziOj<YPN&DwusJqvr)A|
z)f&CGk{=fvm~NbL4+LK1O6ROVfn7`xk_WZ$&u;e%+Fky#d-p@$*$Lm@c4OD5hflgx
zk`g(oemtI=R;`alWPL<y5T81(BSn3j+ZjJ{351EkAx?xne#nwt%c~~cvW({f@R}Sw
zF*n11HyZX2>#(~$k|+g5=E|HTnFzgPAB(hHrNcC;{Pp~`IDEL<V(`rWask+F^$TjZ
z&v-;W(a6q$X8xVVjb{WrMbS7P0Z<*4Figj7;C1c=C~&2!G9mPn0o`z+)B0EOu3T-{
zx5zFBsrvadNp+ENALO;N&j(4*!@({>7RC4`Buce&nfAW)sfDG)hO?RPRkxGh%7<@J
zfzF}8Yb!gyH(Kcs*_WJ|`sDtIRZdIg{3*Z;r?|=R+jU<&UVLY3mHiXS8R2lCF}o8d
zx)WUt#YheB!H8+8;Je9N2PE`BI=W6m3;h;wUc-srzbfqV8VQl?^DPx!q?#)38a|zE
zFmg8rsfQ~<T2g^tZ7<zxz&TEwp%@jq8;Yr9bQaPFn}q$hv?G`XR4sY0cU)<U-!t__
z*vEM~w`hn&ZM8Oi2C)FeWwjRGch)jDU$XQ^M@I+I#7ixJ{3LEg{8h8>)lq*t$HB=7
zC=US}%X0?(19jZ%5ePf83XWYhBU}x}zDxp=g<hq@*bSM*u(LBc-J<<!$CwPDG)OPS
z{@u`LDTM4$*Y&=O=4FPQ+6iYT*nkiJ))vVzZ^OWyJSj9V|9f>;9o+jqGbZjk_4F%j
zGZyBQR%)rztI#1gD;v=$cKQ$c$PeY4ea|5;I<m>xD&wxx>+ptt>yk7e9KE;j(Ne?u
zeu!qnQcM0&rw3qVW<mH81K_O7r_a&LW!1q#U%>sWQc=lWz)Hr>a}@-)O`W6<c9W4|
z@w0$TKig}z#08lt!GtpkYAl7djfJx<1FK=&on^N7u?mg^ognVs!6r4#;cC<3GA$$W
zTsa((R-#yxMjgs4uxN+a#l&Jg>WAOps82~#z_4YMXjxGCr}C?o{7BN9;=6J3nVgTx
z?rQ8B>u`FO+{O5>{5Z5VE9C)paZtA{kd!0(`1k-_4l+9WU(>f^ppg#k9@;xBZgl2>
zsZPyv6_9yoh4>8jNF%pY8p+&97FSJz75svCYAJm}^Z}D}^_Y%CTl0^2=LrUfv84$z
zy{iY7)j0djbXhGA4`>o?QO^FM&n{%7QiUa>HQ{~j{&Mz3?7#c3_@31wu*Oqg$s1|I
zXLTZ^Yr|MuHU;Ev4H@3VeU57z@-uunlvK8KoJ(F%c_*P=(|R7RKjnU7S#ypih<`-I
z`Q=jvNWa<+Vhlg^FOQn~-b;k(*D_~`tj1I*Fl+44uNwc<)7spK&0|kmNtapev63^V
zakFEB(@EKXhrGs2Ef%V)Sf2_pxEDiSYiPt^;)V^bNvIi~)E4_(=}>vq=@+_q;9}kw
zJ7&+mKFIKPMyshYjN5Jw_p|U5<>N&c!D!O+vAJ4>?$B)O_oe+yVM1WoqgZl$j3{%}
z&UpScLgBVZSq-KCs&jeEwVSp5yQ8QI=9B7{WTPXJIWsf^28E#36bzfm5Geo$w)&Ny
zCub{-P&I2USD^q|)v(|?(5yL5DuI85>mH{6b@QkpgFm!-UO8fghoGI@k;t+pB&Yga
zqW+vgT`hrkB;s*Vug7X!mhunoYD*4R5iB#2Z;?WaZr?}D7c?ic#69bMvL~kuiFPdy
z5hRq`m15IQJyHcnu*l75rf}1ab*};oSQGt%_&2?N{#0$Txgv-Gy)r#zxDjEi90t2p
z7Iyk(wby=OSql51ANpk*#Ip+SA1=p)6$vk_<xoA}>BW-5?WddwFk(nz4QtY9;VA=9
zmUUWc)7YuHZzKpo+Sg~CeHQ^l-yt|V>@gai!rzkOEjbm1B-DvIa%*H`H0Lf-sw6X?
z!+!kiLhN7JG)qI3+^67z(MdrGcj=m1k4?F{j>|Y(QDPS@k*PD%W_!IDP{s;G92I67
zY-}%R9Uaa$uzKTO)pe9A{&f0Ek1gDBVJ1Jo_5Sq;`m%a}aO@7})|`%q1_+Kk1j;Q|
ze_?KnaTg}Od~NoO@l2g|3F+<Cz`|y#)Z`_W9qJ1B6J;{xe!tGL)z7zMqqxA>Qn#M%
zG+VltZ6w2SyY*C>9tZY<8%h=_2*7kg0OG6KU<K$ez|#pgpraj7ypn*hm^LmHFxRXb
zcWEJ@TBXNh`R$dpPs0uHOe)o`wx6Q4@#1chkO${-Q+4x98P1(Wl&QIBSPrV(BzX#3
z*{HdpIo`y9O1GA1I)pCf<OvqFiqrh7y^0GIx8J3`ce;74Rh#un3(A;b+xSxPhoBPZ
zskjv4y;!7GD<8jlw7}?fIf^MNQu*^&Jj^jTXqwQpC~J@<P`qj7ZmA}nbH_&hdh_R1
zYsqTJU~F}f1^a%~yq=!s@v(vV4&IM%TdmBTQ@>xHbZWU;rWo>)2MD){);4aT$Q)SM
zsrFbBAY-=Vt`6l^lM)<^aC5h*m0nIex*VpYh`VBe^iXI?0`UHv&<Ex*`iqo{dc>tS
zpEH!M(+#IcaY7jsYhYs+45<1sg3I?)l0NoGnFJI`MfwRv#r+u<XUuJfdBP$`qGQ$Y
z@zf7>+{cga-1%IZTEt1uDOD%y9HeBtucqEJCt@=yrzE+!Wr7|cjH8U|Vn!nLS{!Yx
z##ZnnAMe6#oo4Nq`<Zw|J@Z%F9z1}!ATHXzi|m_{LTX*V8r=W<+_d}{3Ham*;J7n}
zb*PaG;9J}Ru`#6na!T_hzcUl)pbw=%dph+ISD`0o=s4BO9xAY5tHKKSJoP25g$IXp
zAT)w?HmmuUp0g&BL!zbw?V%CVdyXZU4pjo~!t4dl#bjPlX<g0T$?aVXI3ibyf$|!-
zf+&lXd#ZLjPQA(&aucSG?d!9OYDi?&xbaC=@IjBm%$)_;%8U;N!Lk1-Hv{*1*dHxL
z#^lJ!_b{N2<i^C5fhlCQknIVSTCu;HY7|m6X2~`RHEZhh0B5KluCxWMDY*J1DTPnd
zL@Op^u-WVo^DB);<j1Gx?R|l5pGQ6mdDF7mm}bkZ$yc5w`WI;uxA!l|3)t$W1%Aa?
ztQ`J5ncm-HDasAvwmVsStL9VR30!I$jTP2&P&{gaYj&qevZMe05-?V8a748fKryNM
z>Zw4ri}f;Pk6d)l9(Q2*jqpXE^4*g}QrBfm31MZm%#TvtxR<i(R;t>|bdGVCeL({E
z>%>8Q?hgs?p|zX*H7x`qwQe?AF<^~j*Z72%l@MC#vA%l`18p8i%NU7!4027M02R0<
z*8@5_`X+E<+XgHVs2d-!!bky@!TyNkuI;mm(1fB6XM?rzzKeh2&~3}wMWDM<?a2*M
zU~LOz8EDVyZ|?dh*^~6U<Y!zk$Vphc!!<_c8MHH~eAnI(9L4q(^__8w=TJtJ-(h{)
zry~BVaqArS2*z|k;KqnYYk&19daly6?y4gEx_wf+L`EjtXun9{%WzEl2W|hr7R^x0
z(Oc5{^??s{TT9CY@^X~I!LNN{=*AgW*(;jh-MC3h?zLF9ir{^D;*w0XK~X_?CU%v{
zqCnyIL@B<uqgEs!$WnP#w#~$%UQk@yBaSHCdiUvbCZ0^-3#ks@CjS!&s<-F<4#(3Z
zHMiF(gZ{dh-!9_$kuCy?=6a)oI&l)8XfS@)7=InC`pUY%ow4WjvOyukw)WT5L&w`&
z^;XQ>sNim83&d1uvj)kKLw=bER;?uU>YeM<pdPx3gAr6r|J6hNU7emXTmotx(CTNB
zq#j$+V(PWCc#9Dd(*3Vd0@G~PDLq}~ld)HDcLuB5e`OuxyZ3E^9#;qbZEe836Hrd#
z-2h*c_4x_3?B99tfx2CuwU>3uM6&U?*zg#}q{=nw5`bFZ%(q3|>KDbrx^<9uJ#8KF
z#qS`?mep1wR^0uMj0Z&N%Cu!jS81b9cSA_^?qQF7g;x$JRwUCeok|_iCf-9fP4FD-
zJUCz>vYo4u9ZsI?gq6>v%>tO@52R_y=wCchF^6NRf24g)?SKFtPd8tbjm5jWPxxs@
z!)9LVF8ic5=<i#xyfi&R90Qk*?aB#V3Ess~y`;o-x~8-L>}Jg^a3|z?;G<BWrH#^~
zDm6S>H#NVrYH_#fC>3onSg6_-N%*5(hyavTs#_D{rvxJ;cvD?<B#cf7Dn6|E&pv{P
zf?C#(C7@tZNVq!T>~k89Um^<~J-O~>Ysh=BR&N9KU^cnEZk5Dc+||*=E*tTm5pB$+
z)+=bDX7d*WE5L%&d1`U8An{L)m;BtCwJ3L60mXw*KIJk77c6~J@mpt)f5WpoE?mk;
zv&rU?TwAg2Kw1~BxwD|OEYRp$;VA9R?pY-~KU2mdl7?^`;J{ZjJ6>6%OyEvpMEi)Z
zQ``C+Dt{LXk&3vWLoXf$Rh~o14zlL&kEjhoV=wO7hxAriT%B5#zN$Q-Y~qK$9X}W1
z{Yjqx9SkT&ypM&9Z~BfRV{?KGUzxT$0O%Y|_Bp5~Ly>-dNaE4kZ(b=rT?*?(1I++V
zZ=BhvJ)pNgTP?c8fjEB^*+-p`)|y`{hBISU{q2&f?)C6MHdi=0Y_z)%JJFTndHCxb
zalDw0E8rog$>|$<+qXZa4a;N}6DyEYm}I;61I?n|AT@|{7m*ZSle5W_;pEXZx%B)T
z(M2Kb1Oqx!;w83?@N!bU@v&?|LOnFx-O^15S(H2yW#CT%nZ;TIODeDM79}~y#^#H1
zc7i<&X{1b<3m-)m)n0n6q$t^&8O<YGS&c0IlJPzMf$q@I=O}yC9P|5qZ5o?7h5=!G
z=TAen3GtG(v=zxlQU`EOd6SnV{Dmcjv18<h(?dk1-K3%sxqg=}6*X2PT^GP@(o8;(
z9G2~V%QgkoU%`n$BCS*=aC-c)vB<%#`9;zs?EP5f>LSfaf<;X~P0g3UrjSk_6|-ZM
z!h!Z$%GInB9T7*JCUb|9qR-K7lKRa?;wpbcl;{QCYWO516`A7EPB)R!9Y$5rmzc)j
z4KQHs#`fuOlErPz?K6jV&uad`?7{XJ?liGjc@A#${)}=s(>OGjb#^^uIi-UJ3*cM$
z+EnH5I$8{=Mg3R1jNfP6I~yZ?Y#|HmVq5k1)zp6i;HcH7Svq)w;N2@Of5}p2ST#GR
zGn<qKRI`<aOZLQDoG2&GEOP3QLFXsxJ1>$+pt0t*SwaK!F4~HUD;br_%F^PACdW|<
z|6-5}5&vU|&S>vN<+9OjRQ=?9OoE*1sFK$7Ylp4CepEw0nG{ag`V(YkjHE_9xcgKh
zKr*lS-L$v3XB^VaBbJwgEe{>*-h38m8zj<zEL46jVvK2hakdW~XXeqf#BzzzVJntM
zlXd0=W~9AV#XO5}x5Zzb*LLQwW*nv}vOkF6eRr%pO@%>5nn*y=31^BY_^cBxd#^`E
zeDpZ?&29N1wS4_YVh9OZtx+Wt2z>h<W<F#=CXx};Jc8oeu9|N{eNE`3Y;6|x-^I}^
z7u-mmZV=7S)+mtHtWG@{d`-7m*D&E_k#KhOd{T2EdTMGk4_W6eO)OYu7u$#u=xt!)
zJ*ug^V*9>(uo(K(7@J+iF+c9jZnaJ$T&8SaDY_QF%!>a_D_w<UqqhS4C~%#dS@R0k
zz2>Cj^sYv#e&6vawO96W<!?G8b;`jA9|1ycy~$?+oZqWF0e{fu$>9_V-J$znkjSgg
z;FJ(D{wsUf6=3?a2F!q~`JaGpF#k*qJf&m+0c;s6cqyncC|_&;l3HnDiN(62=j`Bb
z`>z*Qu2iU&o|?UwKSf65i%lkGp@~Yh@D;^GxGcD`<+MRu;n1Zb%e_#RRDcs%P?Nb%
zHP3~y-Yerty1FqM^k5HSIoPAWM>uNvDHT?AHYf!w2)nX5n&)csG^HmO)ep4y6#bbQ
zCjh00##V#Wm6NXK-#r-!NGJ@gCa|f+q{cuUQ`>~HMzy-i4h|lQ=Cu>njsCjM?u87A
zHr#Q+UqalZW!HYNtg$+u*&uS&`{DVFza-nQ#_LSB_lxB92$JZ*&UJE3X$#e^{X%!3
za8GXS@T1?8LB6*WA#zIcU|K=vq;woo46>U`=w)`}EE?#MMtB_x4R09&#^B+7c)wV`
zK>MpT%0H6))E{8?j}G$Hz5dx{mGO~8-8#d*<`!8$RS*^;q_u;#ImT`+d97m&GrklR
z8aH>q;2h{=bwH*~1Lfm!_G)4r8JOYr5+U?VemsO`tYQU`Wsa#eVL?J@AxUj>KSiuY
ziO5M0lg_uImA;_`vrdO5c^dgY<ed3<swzL0fz{&Fv71o6{QPtc$Tx7ld1>xpM~h9I
z)(h{73{`<I44}iQZGxSl^LoY^R#0185%nVBy(&R=1@@57;v`U0<*wV&7y-cW=6Ay8
z>yIw6Q|nZ_S3OSV45%h9W1W7t>Q2Q;Iq^TcT#~a5s@%mD_SJ3<^$K)^Y(K-E091%V
z%0J|YlJm>JS}keW9LG{MHc0XLj%rOwOIgP<FNqSTK2npDO69)|jXsX%pY@spq!bYq
zKkB-JUj7{}h@)e%%O`)r-1ddTv+L@Vs5Q!*Ijc`5Mr^2eTm_erWs&LF557&oodB`j
zil(;e{}lUUcNXWa?&|16=^6?Gb;J3Z7P7!R>DS=o`}xnXX$|nt+En;w@$AGjyn;2D
z@N-h&F!NGI7F*h0_N>Jg@Ero2_B1n-d<4POalR?<QDj!DYUXtITWFNa4E?+*i@0*?
z914rcGL7D=q*n8#>F&YtTn4Ca@&3#)oBPK#IZR_TR#TVzOJJC9<$jPYGnuGQBQkH_
zqs#AYk3vM<MOmXxIC9*1%Wo7%R6p;bm{N`UrmGGg-=Bg%eD@}CVi~7|=fCLe3QC9l
zJRkgqk;Wx;Jf*eq6C+z$8`06aP*(>u{3m?6Bgy*DHv4()*5E@!LnB|wg#yp6Z$RZ5
zH7xX;0_84?a~Au69C!pcfj!;y2%UjLGGU=22PnJy)zb+ga4=>gLL=q14}5;u76O)^
zCui$NO~--?ii@NjAOf(M03=2)>F_?vsQr6G7L|5G-@li|0vTQhN8@9^4w(5GK#wQ(
z$IW$F`T37a_D7Ab<KTMTtsM35@Rv2#J4LEU2|v4p2*L#XxFS_NS)Uwh?wd%MNc*v<
zL@dtGK+AV>D?6={y;kD-E2q%$$dmh3EOI=k<2*t@bt{9Pz$KN@y?WBaeuQbtI&E5v
zSpG!6z}J42``M-@V^GhKhB!zn`Svg%n_*-)pW&(KH;?Jwz-tRAxnqx~m;Ge>6d-g3
z+-09L<o%y(1-?XPckQE=(80(Wt5tq1eIFHBl@6cJ3V$P|>}2&-td<nl(3)8ZlP%##
zxJmmF=gE!hc3DTEVlv9fUG$#<MyvJJp`vV$Ar)Hz!Lb2RL$n+I+j|{r|2iK(FOw3P
ziumM}JgZo?Hm{5Y0=(C$jOnF6zq&hh`fuEnvQXjfu)bidovNDLoziM*`_4S@ZNkLr
zlr!w$J1aKfn`ItOIXNaD!Y>GqMe8O%{39ZzIC{=!$Ivag!<FO|s3u=z$i2mrmXUeG
zJgJVY;@A5%tGXI5PRekq!EYtj>e!;JMM6TN+sG{)h%Kr-Z|#%U^JJ5|0YUCCtN;XZ
z%ve?UbiNS>c;sfE+HN=en^{11EXkW2_@0R9t_TYen|{ayEy&{P&h-A+Ae(Usz)m>7
zh`TiSz#YK)c%X*rY+Gn)%Xr>pJP)YbG)6cd+|7R7!_5ZKoDAX2&kYcGFEH4vNR3v8
zhUGgd%SN(~=fzWZkd8LR3nm4SGq04z0RXN1@lX7U%hzv~jdYIZ-<K}sBRCKsv~m9i
ztLx-PM#W52qtjR@;fB~>U5a0rf8seWr>*JaHzPx3{fw)gx0si+Pxn@?R9#eoVoqYy
z$L_D&vN58ZxbNAC+wSY8oHm(UU7L4V>nRrD16na3J+?Ha32@(r+AfT$k$tgPY6|~<
zXgbTNsNVPM4=t@U0)ljRx2Q;ibax}&T>{eG-QC>-NF&`jba(f2_&&e?i&={oxMa?K
z?z*nMKO2fBYFTv_i%(W7`h6G@%S5Z|2Tb=nOzfSf&{-yT@;mgK?c{`0|4;ualhy;_
zx}M4MC0gxvC59>GjjZUZy|rX{h`A4(fC6wmX9(SuoMOp}<nYHihq+FiVkW#ZL7liE
zLMi{s*<s#lR%rt9xv)mpMjEV9%JwY`dp6}-rdMuw6bCfxe;n2RKsKKNU0e6AjW{%T
zZ0x_e>VDE3M3v#<2LNSMiMRlevbNhkMIf4^p*tW=J|Lq*mvC>vaY|2sIO!8P-m%%O
zZ@bzIO3Y-F)iqGG#<Ir#f?c8tehTNBpjCEg7Sd!)rMI}UhUvsty>58YXVIM%amr2_
zY|$7p9{j{YHPc;KTKV(|f-5s-=-t=AM1mCEEjCu%>{~2#*=N(Q`)n`u$Im@*@RqYw
zIP9MEE58Pdn&DR+pbp8m8)WtGdx~Ufl>F_^gSLiVx-m;lm!Q4)(%p;IO|slHG4%7d
zS+gRvg%}QmUV263;oGkI6@mt?-|hBApCU{`URvqG{aU$#P(PAy(Bq+0Otdj^iQj&;
z@o`qd;Ibh9a?KiBCL|Gxe8z^W3lh<Mf5wu!p188u&$Q@C0i~PmYT((6%#ix0ZL_e{
zl8P`4zF5v#^rDb4AoDyrgKN~Ps7wHbxUw#Tn%5q2RrOsX>hE=;Wg)JB?wO<!N^Sld
z9c}I#n%;sEf>P$r_b;nEj4$&WlqlMcaQ3Lf2@BEDa??nOv?XYt_>*QzRbo?9hXGcc
zYP~r&ol1q_>&x@j&Q1-G-i`gFLhX?RvwIc3R%!3yC}mu%Un+e@Od?ofRcYN7X6ets
z5SRQ{_qch%=9bFl)#Pj(ck)w3f7%q0gLDGYv#pj<Ee;_`%BX+_K=56!(3>_@Tk{^Z
zof9^yYY63wWuP*@ZK|=Gx8YA}^bV(FSP}*TRCs<$(UcMls-e(jrm1pSDhTC@aeWJL
z(`NiETKM~$RMPR|2HE#ab!4I#a`A#9KKQ(nMP0C-13kwVDN<}ClWK+BE*>!w=Sb|_
zyTG@>&@LNn=8hF)`2I<Pj*M@S5sAe<UqTysn^GL_*>W`AapsB%l8YUcQ2y+??Cagj
zAwW8RG<hcDR&irU@Q1Ng8`H!BMLa+#vi8J!f)?jfCZz4JovGaM69Wo7cP5PY>$ybG
z?{0y#4%o~xyJd=tUfyS-???fS27M!1kJ=kdKVG!R?$D!jv<EWLI(o<u+9U<?0pkTM
zJbX-g`oHtd{^s2bchlK2d;q)nW*0kJZoa&j2PCf-iJeaY)6<IG-Q926Cjetv`_Kls
zUnu~|zx{3*LNKsKx&lmxH`l42p5E2TN^0YpSL%j70>cw3UL*BdU{FwON=ga)3w8*+
zXvV00r81F*#xEYm<AM5YBTQ3ygl7%5s_Q1j24a~+VH?#bJeBa%0?(;b4OnZY3Yz>F
z71Dw}f`~tA#@0+oE#CwFupHM3F7|0m{ciYSiOK7j-4Kj3>_yQ!Fd&qO@ABe?c(pc;
zTX9Iy0o@$%{sSq2md<!(g;b*PqCd?<c0^?861<p6lz>_Gomi85lCzTCtTQip*J%)#
zs?-S6>YquDQki0@x+HJ%hvXn>oCVe%?1dU9iA^27pitgzUVdB8Gs+gq9>4Sm0MS|z
zxEjIyaXamNh`--=ceQ-IuS3b2Ep<Ds#v>E<Wk_oBd7?hKtTyHsIr2s5db$o|{63|z
zQh@#<#2L-k3$-j&{4}MtOq0AN?epNg;D?M>8NZ{1Q!RN1r72w#UE$f`&Xc)YZpob7
zB1NRoj7-u`B1m(WQ=I)o3p*jNPfgMx;aEsI>r0ozcXuOWsyeR7@R0IY9kvE>?+U9U
zBHlknMn-@{{{lep+XDlnJhAc$*l_QG&qAFUCZKSAleRxS(S<ewImtDEtnbaN1pwhn
z_$vj_-%@;ljD<14@4~p&U^I-=@N3}>^@z`>oK9G_eud@}gOUFH3Y=d}g2xvUbX1<7
zE$fRHIR^q9O8Ox69799BQP=JH*fYo2N@n-HRqNDGjuDDieBFkDfC}crQJE?X&^74v
zDqmA#a^s3TnSBjP(L%2OOg0A&kw_gC7DDv8v7&f=P(LHi%l%SXd{Qq!=u8g_2R6XW
z>&bIN2mo@=Nyjwl#Hq^xP2W+W<%Yn;3gx`*Bp(Hse=w^7sE6itcV3ue;xRn5%F0xE
zMU2rU+&ppNicemMwd~bb&NKOT-X<<gIIFO`TseIWaDhBPr1L1}Z+=b;C(}nS+f&#^
zpB<5l!>jYX_Wus9TWRkp{RmEo8~%9Mvn?NNwNE-y`ffvrZ|PzA9~fcS9R8I~K^cur
zh5Lr$g5#=U0_J%u<wy!xBR-b6+wTH3XQ?CYM#f=3j62t$%blqx8n!|0<0a{umz37_
z5}nk+yvjr8Q~df2xjP&t;>WTLt|bQM1b1zoo95x&+{Cx->W#sPjLWFYnmhpTE&+Zg
z3<PkH6aq#JP50yaGy!)Uz+np5AA!Ib3=l8}TN!&+TTfa-f&8p!0PhRjK4|FZa`W<j
z11A>ygK3(#=zzqw{)dPuK2Qpv_j@zwc;2r3m?{`b|DJuOGy=|;(tn9hrh>|ut*$Aq
zO?q^cnrw_JE$#C(W(ey*F>W&Q8L4mZX#T|csh$mtQadXiHJPE=*Oz<4Rg5zvtsizh
zQr}W|fSvbupI}XN*TrO<AO40n``G=G<@OQRB}P<_^zrJ2^>MKkMnTuq-Es_nk7|V?
z)=kV=S|;0J4)|wl(>g;zk;>}U`;nnoE9(b=GpNbN;gHo2F=-=oF_6f1SXr}X*q6wK
zy?JJm{%>EZ6_4??XaNg*WL@x-6Hyy~hc9s@{8rlIh9(Oi2c9|8ls;$ZRp0~(qOi&b
zY+5ai#;ZibU)t5=4}6p&RpIszg{8Ws^`wy@x8|&h@xHJBrv(^JDtf;kQUxqi4HnIV
zz&Djta{IbZ85)jN{JEU6`8c-Dle3U<$Zd9mY^pBWbpJ5g+m4b&Z_@SR1OpLgsiL(0
z-;O~J&zf8ZL51!s&%||Qhe~Su_m&^t-tlSv5<_)Dg@g23y^hzB3e;`B$oUJuhG#MY
zbmHsNNfi*gM0W}P&LylppE$Q&4ACXN3XzifHLkja0_M4>SSndo;Pmpwd_}>=e)C%M
zpZ~>t3nxnBw)K098!8CjbZP@AE^l#lPft&8#ugS9JKyumO9SBNGWGfp0MFLEBRy-M
z)^SZsOS{pFtCOczYy3Z`z`@WKaGX8e)fa5LVr#2@mv9_Q^15Si_D|I1iQvvClD-0=
zP$EkPSwLgD{eeo@<hx_+yNQoP3)MH3t{c!wC2i$;3#K0RFA~l^9J55TFD*aV_)ig{
zbWkpx{_J01u7xz5Bpwo`22qMyl&Z3mbcN}Cgy|hMtct}t9Cc4N@#ouHfc`{m?O98V
zL2W1Itex*oMzqHi-11^+SX;z*rvqYoGSnF35<t*S*KuaT(D`c`g~sq>g(%c%3GJmP
z-hO?xw(1~is-W=fb6`W_X==jd*Cm%-yntEfj{)5}74gt8CZ=`X)iPV}w4`0CU=I-r
z1{`~o&tAt}RZj=_=jnG7ndX$F1uJK&D|lU@M$OrUZsAbCO^l2EI>)-d#GA>ZGEx~V
zdXA~KO;*u>H$myagCI@3pWMX?<H{T{Q2m>bU1*kEFyV9xQ6Z%Mz5%)3DLoM1-onIO
zyPZ!lyQne``*4bdkt=n8-nQB)Q|1M@7S1&>kdQ!XP7eWhuP4u)PBNgq{^X-jSf>=h
zsMA5{1K=xe?an$`ftj5T1W%;wu0+|mK|^G{A25A=Yy30-Bml6tQwR!X0Cwugv^01;
zM8FyuwoK}BZxZcLC92j8+p>G=GcG()%EdcB$b}ZPhEnpltj;;X^>X_)&$EMZ4{^~1
z8!X7z)vJJe-QKWzK5BTX=5nEN=T%$xuW(n4)o@?itoPNn^cksi34u4kikXa+C8OwQ
ze{PYd2FXi?9+!?G(Qs*ONpDmc@z?YWSC3L%cu=PdPI1s0S^{HgP(L`8OY(gerg879
z*0HK{%QCs0c1gabkny)v4Z30WfgC{ibsoKusNd&>K|(D&i&{de@&_f_Dt*p6#&J&i
zIdAACWBp;Qu{Zx=mb1dRJi<^<w89}ojaw^nRw+YQ6(M>J>al;asp4kt8zoy?=XA&K
z=@Ng^&8>mu871;>N)FJ=@ZNY8ujgKle^T;{hB<)txp*A$*z{U}oKABR`m5%_-Lvsd
z_pL@Aw<KNk=TQz36>5I*O)&LJq#0J$rjIR5aPTY>7O;(i>dcL_kg{zJYy+5RCmX8k
zlaM0uSh{zTf-$Mg-6{eDr!YZpYcgPZ9RWDj9iF}IG;D0KK=aE7a6-S~XKqfES_K1g
zIxeo^ALaQHq#w~W@-W&gSo!VhVeiW)SP(Nh3QitgLH<kTYkQR6zLa3*!8jlsACuo+
zTZQn~2lJ*tSu@laTxjZza9Fn<?}~*53Pm<Pw*zSucCz9i7ij-4oFR_V(Cqf|!;e7E
z`7hVmtUgbJ*CIS|xsTAZYz$P!9b;TVX!WW_=z8+$`l(R&t_8&hSi4WA3*Q_bemKV2
z`xesHk7`ab{z%YzcSxv|N^5SzBAfa7usBqox>!!pM1(78a_Q!ppArRd0+hU#Dri#1
za9N=DguqOOyOI-CoaXCT9aiDqqohwakCyOvbd;6y2TXu&!+-xwFC(`8@?v8k)Gw%3
zif-Q8cfwC$ARsOVx-!kS7Qi6LURf(*<q6n0zMgiTRAX~)J<5G~$#sx#U36KRN*a;B
z*Quz!G#F;tpBxX#@XQohKmLo*H_pkva1KUt`{6_~GKL6+MWtU?JzuqKb?9ZG1$EQp
z@{{puw&ayTH3luDq{)!(6JSsU00_xI2Gv^}B2c#DvTi$vhl7KAv+BNm9FFFr-!RpP
zn3S3)ctCw`RoW`L0i4vlAvAm=p6?9Ic(al^pHDK0oQtB_<IVVW5%$_}I{I~rH2>E0
zbP(oDqg;8udaGC37PX%v{;W|Om+*TL@#Qt(l%#M>PQTT=pWpq3y`rKqDZ`Ko4l9;l
zX}pY6M1({5llWKZnE&#H783_HG9p%DKna1ON=qH>MfCV%K8(VFV5H7%s9Fk##4rWE
z4WlIO;&S~5UdNxD_4v(7)<PLHN|0P$3Ahp}j#W0he74>-^oYH2EQEcV^WRsBo&{An
zTA1G{>XA#RJqTc!t5d}m%lJvY9Cc&-LBzeU?6QnM`sx^wCkf|t%~)uFf8oj*I-ybp
z5pTwfRf}m*gR>)={m5S@=b9e>K3{8dm9-;#-5N_MkR|=E$!7+Qij~3eEE7pf<F3x5
zi5E9O`oobyg?JWwIs7qvg#yN6Jvdt79n(nskdp<WAk(c7&(9|(k7$mdD@naR(Scpt
zMPEV{k38~FY$Tm7AK##qH89NMYqr_b)y|YAKiO&iPcj}e?H0qwZ_&xgW|tj-VZ(`I
z0QrXh_384Bze}&t5cW1?K;Bxl^Qm@$RI&)*iF^x7Xa?3r%Z9MGjSX_<>iJNEFjet%
za!RE{<VMS9-V>)Sr*7=dhOx(VeD26W*zlfM?b?%<(Ava)p*8+ySK^PIwqx_B2xC)p
z(xed#?#|T+Se>j)SoL;}d2^Ppb$FyHzMJ^+%|=X!&`}?)3jiC?_bnhi7T1C(a>)T@
zk)rz#nxh+!wBe5U_r!9v_x?}`M*()V<crJyMioe<mG0e3ZY#5>J`}$H7;r#vYK?_z
z6VQ!w9<2|ms2T(8{GQpC`>bpat=I&v!pc2{IQU)ne0i9|eyiW?Jo0k@F|+O%zGb8Z
zVTF#wWTGhslmYLdx29LM!D-=8S~Gd%#KDIh+JY-Tp!eiZtNb1p(ZJ%o*`sFB2`B*T
z1(Tm>Cv7bFA*!60!bGxLU}5zTR|YzJN9SNzuv@~ShtHnE<;%J7-Lj{S$l{l(dFc6v
zf6zA&q_DzkqzrT1oo)+Z)cPA|+#7<e2@n94`vJK8TDn3BSJSclCSaifcr>cbb`&O~
zX@JK}n*=z?00ff^?-y495Qzr}I7(G23pA~ZgG91zknhoO24`&Rik#p8Pr<T|VE}xA
z7=^fMlwz#r3Y>Pv&NBHo>qB`bWyFib3Iq+5nV;8QUv<VziHYkC&dgBV1GBC_z>L%4
zyfAlr!I_69$)DgWEANIp!SU5gVvH=U6T*o7@TgyXskTh+x$tag7t!o}A|e%66qwJy
zH9)+z(ajOWz#(7ZQ{SJjvq~8PdNZSzndP<}*5bH)7T79sRxuSdm9r|SFJF}ubbkDX
zS7tZ^cv7v3)2N<A?#eXgInLH{Oa0ubqI!VH=Y~RTlgvn}3TCY2AgzUZ5RO5ou<qBq
zoygkd9D)?(5|oGgZv@FNT~l*lqOW<O{IuU*WR5ZZK5jQ3Ub(_iV_m~}`xDo;wvSlH
z_Gz@QftHtkA=S1#5gAw#{`363-qSnKn=p3N;rTdU)uoJ)ueTWuznGUtp5gUicG7k!
zBN{>Ix`6pV16$IcKkweMdfrGxq-2(+KY)4Ah?!s8%{Js;<6yb5)jtE6s8j!Wnh-^%
z&ueiOHoQCY-_KnqMm}yVTN}R(F-vWoATTTnrx)|gGC?k$Pie$#NRJA!^@oZfNZw1=
z6wciF{u^ONOA_F!92+ogOHvfkIab#qZ9SWB&pk(Nz*eF%0E*u=%ZBN`>2Rl#&(+FO
zM`)){`Z|8c7pgNYm2%U`b@ZNEf7Hc70<{^^9@8T3^pJlEsXg{<jPlG~m(kMgSjRce
zSRfy%7|5Q`#4sa0?_Hm<PH(VO02lbOZJZ-N=^vFYX%wP8KAeDp9xvjschWJ(k)h_X
z?$q;DV%S0@Zj$lK%Qahfl;ZtQ>W>)NUR@rG`%L`PPof<%8r?5Xn1Er7ri_D&Dy%@k
z#X?qgV)Ko012B{>0Eht#K-w}M%a8_eE<iA(bSmdtG#KzZ7PjvHky?QJ^4smI1IRa~
zy+y@m{SA~Eqk!txTNnV_7=KUa%Qe%>>Fdkb%9J8Cma-`U)0cx9JIl~T8EGv+ibw;{
z2|j!!Q&pu34>Me!v*VKo;?J{X@}yeny7N$4xN(IIBm1QZ@r&>EpZfyMDj8Dt2z4@n
ziQY*>E<}O73WB?b11KCk+xP!=QsSc`254<8p@ppx20n0o6!+>>dnva#6QbV>*@_h*
zUP+ZKoaqKsEz-xE9e<-l)`#~5`r+@ybw_dZzUU(z%J9S-L94d4T8-&(hQ>h#WH?Sk
z!=oH$WS9Q>G}!5hl&Y$!G`DVcCB8$8@*^l;>z%cIpgYhKcc~H!mp&wWB=FBZSusnz
zaExHG;V<a;fJd^6DX{YC7?a{YM~*t{$2VF5QbqU?93YAd5y;5Y1%f#93Y4GI78@*w
z0nEY>-)Y-s_JS*ul_cmbx(2wzye$@!lNE6#>{iQ3K+!4!sJW-_?U^DWBfB2g&$*m5
z?=mM#y>V24ycHI&wrfs2pm8eb2c!G#hw2td6HLyrU`g)b?iT8|K(W$;Wg-AUKwV^*
zfs2bc=YR{tQl7rO-et1{V)-eM2^dwSN$c>dn^Q)mDVTLxb7>|QITp4RE|8#~(_Yiy
zVSdBz9p+Tc>208dI5;`%5$R}5+Fw~{YV`QjrL{iTUl|o#$6v1Ardn7%<%&T1ynrgd
z$xeaX-AA{vYL<1<nfe!-f08ai<l?6N&CU02yZa@5?Ga~ev+NJ;4G32j2JY$DIOG>m
zQxDo|DjaCH84_0yqjwc&Cm6>D?bzN_<=Y~TL+=9bAt^l|2g$(tp!#sGk3i2vla~Vc
z&R>5A1-&6TR=r<6Uw}wQ+q*Np#YXEffIvG!7RlcYMBAP0)3{N<*Dg4QbOFjoK)Z4Q
zkWE+tSrF}QI6%33ON|C{oDJ&ajW2+^2n!JQ1p;Vm{lG_yLpmVz4Q>KBgLQ#eFj)8p
z$WVc=ygYp6NMq%(73ky>WomVxXtTJXDv~4#FP1);K<><`gmHNg1*cR3UF>@}?@ZG8
zbJT#gwzgp+5CH6gg`F-RIWF%dwag930=e8O?^^1XR|E2(k{*R(C3?W<J@>D_8{k{?
z<+KR=Fnsr9RkgcS?-3ruI`6K075;@+ftS4H=CN3goxLIB{jo>6gE^9S%SXnYo>jIe
zCF%?+%T28%0prM1&QR}AWnRtzf(NmS+@DEUVt0$Y$Rt8SncX60FkxX~R}byefQWuU
z@X<ij!Dk7+#Y8Xzxc}^<%np^q=;op0NCQ#fCMFh=XU|UWQ7lg%cxc)(f!M$Ed4tGm
zn{^SOngSj@LJB~(5k~+63Yt5B5CUv`@VG#Hwud`#1_1&Ex!)3igM)A2ochBFobPXe
zMWk6j1u$0Qz&`XCY(hdENq;H`_~T*1H=_TxLS_H6WAb|oPj;-kUg{WzifMk12o2ru
z@dsxaq>~PB;io&at}5Rgf{MaM9&d+TVyM9K3{=joC&FD<lv7kZQ@oqyX0~<dqS-TT
z8RlMP3=L91safWp>TDrvg1wy-9?A`d5mfu(Z;+7b>O6=2&*=;e1PPU1*`=(evIEaA
zFO&JU^|&gpe6csdF>-tcKR!;2m}Z?i^-_pRs%(2JFW$o7tx5lXQW4Rc1Q2*pynsV9
zZKLDyVm$DkW!P@^nN<Xq@sAo#PT&FpG));fpE@(`DI%vd41@YCkQv?Sv?8Yo(i+7$
z`lohvsmFvP?0?fk4rY|%S(G0#k~5hmnv-G8l!@!C%TlmV=WZOHJ)WKob8pfE+dNdG
z9%&a-dQI6f>c?r74I7rj2kChS3mJu*@i)jQ<lgT&{h)l4xcr6>`MLiD)NVzCx|NPK
zg{a#1P<H?Xi?cqc&oO@MAtF}AZw=@r{gi8xQA>LWn*af1D5f#{w6w=#T{WhAS{MZq
zayi~_CGKS30=fZnwzD|kE?EwkYD`1M(cE`#1&)=jK<ULIiPdx)-TThJwg9lyY6@%S
z{o=<)MfVnJ%C02$Mavt^5hRxj6YyNPn_Iz)YS@bcE!9XKW3Hq37=`gQwS>Y}H>Ot!
z^sUM6Y*nCoyS%?LWwgUnKl^KJKZbv2w}LHBh@9=qTuc>qu$9V+1Jq#aHRq6lX3x<d
z20!RLI;H{S*zok!V4*)aQ|`G?@OP&5b(QZw<w3WQ1i(i|_@gJPaowXL`mCV@3YG)i
z8fQJ$t}#CHFKY7TbmWW}FQ>E%T)>eapj&|36F(^}Z3K|E4+A;Nh{(vLQ}&)`T?jHM
z@o&Bw%`akEG50ncUZN~aZ6)bR#p!J%TDa#w$4m2JWtZ`}IX^;PKGYOzRMi@ndY7v1
zgUt4g7e>uI_14t-3kGS<nW=jP>}#rS06Ip1<Ju8W8#Qlt33@V)l7I&Of6qID_4R6I
z6S}9GR^wh<bHB7$f&BB+Kf*#yMj($WsZnnLBvo-j`XT~Y3$rZw+t^WOiHgsbsiVwa
zoe*Ql5z%k1P_%*uQBIj^wUVMU$TH;7BN9gS-Z&3Up9-l!0FW&VaG_Ve$)7@#qNPU)
z4D?To*#A69ymF3Hj!xBHTBdONnR)UHAa|K!39PZC_u_H0sTlrJ?^PZ+chQwOkH&y~
z+{GARs1Yo%<mYFW*g0hc6k$_JH~#7%x|`uA_!b@>fC>^jX{YK93v?4ls{kz=8h#q>
z0|aRfVvR);ey+8h=J8op@R8Ftr|0;XuU4>}#2GZAZmlt7sp<~gsoO8*sShNwgitK0
zgiwROw-P!VB_u<7w$CO3-HGj3+Z#pa^X3(-aHG`<04eDH#c97c{@FyR{cV8wzj?7d
zJ*prcDeee=q+g-)9oB#c_vHd$9azC|?1BLp1>kyiiY~&=&bSpkXd_x7_O&h~AVPdm
z+R-1TtrRBDl7A%0&1f?@gUq*jC!7MrL*3in<sHQs%o?sNDrD&oMRUOYb6QqL)8{rp
zLZpfn;8nGy+k7R)Q;W>67%}EfLqnL_a>K)hCQ2RR7}Ws$Nf`|M0-=j6Ib?N`zXMvg
z_+quFPpLS*JQ?y*DiK+;4!apVf8cf(k0}$M0Jts#u+5K}fSTqvYb>V@hEF%fIcDnQ
zaJ~J1S}9}<5I~g6DH;&nPk;Ni2NZ7rw}2eBU!nCudc%m+6_i~u$#ib#=_w(MZA_`@
zGq@5%5N`tAUf-0Ngz%p4j5!uHCOgV1-Tr{Q{=%T|va2t>ur5-1EIRW}4A^UM?tAqQ
zybf~qZYV`1?WlOCpMjN8Xw;~y^zxUaO?;w?qO*cfyZ@-BAJEA6LOq&9Il<e=Fz7{j
z5^~=qT2pqseI1McHQ3O0>y);RLS=hH`q^Jf1U_|g@(U7C9F8e5b^4VnQhr^h2rdy8
zFLka_EbKeP&!647m&A`;`keST+RGiiy*+@4*!)uaJUX?8vBzsx?wgtRxykn3j7y?S
zSu+90je4ip-(oVX%sc^M<3W68EZ~I0efRwH*80?zc8LRom$qvE5G+9+bp9rcKj$*x
z$|Ubygv|v0(Wvq$&|rX|eT5<DivMZOaCXNs5kLbq^jki#(7rw}xydES1iTM)H85XJ
zo%+?bA)ToGkWK-(8StqSRAf13yzAHCW~OBO;?et40zRKLFO=m@lq8{%e>g#I@cQ*N
z_!^5-)9Y)nJ}5{D?Vf~twoTqT0^d@WQcZ&0l_4MT!9wWNkJj^k&DMl)#>S{jwRa(+
zkp1I|Oly-;pI3L{yK~WPa$!ka8wiHu2L}RQRTv8OUE|ofP`(N$(3m&0BDX7!$zNz4
zt&gfJ?^ZX=pEy%|*F3%rN~pwffjSqUo=jozNswWoUni#bmyY4yzA1hQ6$nr`dN4qZ
z&&~!G6@mU}BEmv{@iB<^jT{9n$)cm>pw#<j(!;be71dTs)`Q&(eNs(|YlWbMPSYKT
zo*F;HxR?<$e@^yOgdU4xH?%RUYg(HXEgE&+-3)7toNC%c35_4gI9&3rWH6B}8UWD}
z(7ZFL0|kE|Kh9S1oTYgOTarp+dyIxpj#^#$zR5L7(LJTyP^L6)8GhgtlJ5y7vq*b=
znL?g++%_F|Z~C3r+2INkRdE(3^b_rAU!jF6{pSKKe?7L#ul85e`R9M5(7>>S9~eV=
zYU99!Nc75Xv$J*MoeBm%h{2@Ew0QTwZ75enzQ@RrynpxPcOY2}(Apryg&EMGUuzvi
zfS00iG3fm}SicVeN|cyon?|GFI*Q0oPo9wI`aIhc>xu@La;<}7>qnXkSUjZU*zCWi
z1qW{Bfrd`1w1P8v*)Ri4WS&MiQaHLQP&WPiYZ-H5qR`BYh)p&DmhPTW-k6@a7McjJ
zzAq)mx%uSl_2m7}f|3WmiRjRY1PCwu?08LP@Qzr2cE+Q7DFPv{KF2HMBrw5GtTEgA
z>zdeFx5>6J-Y<i$kod1Q$zMt>;WAd?pIl-ucQ*!y!Nkt3oc;9^76r|JV;dh%?<XyD
zDkwae-<zC0w=?+X9je}!y`N_veg#ZcL)7b2M{)~<7NC<9^pU}v(|biuC=<HkN#1Wi
z`j!=@2FZSq$fZ^zI*c{r*Ho}DP9&O;0Z#>*KrkBQ1|(fp0~fltAOqDl7Zz4Fn1Ro7
zB=5O<7gdxfX=9(pB5EuAUpd|W8CKkdE<NddE+apLAS&bE^?N}k2!gb8dWjV|U`}^p
z`-=aea4I5<5KH<77s;)CTKfg{;vUz+T~=0FEr?X&X)lu-yd67_2@19M(G`U;=D3UM
z${TYU^l<6(goa2_?A!6N37l|3SrfuAazWiPVT=1je`}loupSg}VYess$E66AppSe$
zuY~)LLEU#tMzAlApu0>*81L0(#HWDIch<Qb;V9#~58Pl)MdUVjXzH?Kp~eBZk7QOO
zms;PgO~!O%%_!7c_UQT|wJC|y;8*!io}8uA!I!Nx)g4YLr(^lK2zQTO7kYN&f`oS$
zF<m=D(7uh|0-<aBK=<3fmY6b+hHF0n&&?Ng1{1`=!J*pf#Nc)2PyF`Ai6|@MeDb{d
zx$JgWE+#JiMK0>%vx2<*v>*jAFa4}A&(3*DPRfoge#%4?fVWDDb10^WlNg4un{MhH
z72Z30W>}96e$c@qz?lW9qy)*#4(Jfl!D!D*X4VR<i0Y1gSC12FGq&k1QdcbppE^g2
z)>EC&;$ZU$iPF{9`YsqOGyTRSd!|$RJPZ!_cynx_LDPhXFEP^*F{fa`r9{k!{IX=9
z_q#;aI;q=p$PYx3qCi<zf`2z$2yWT$Ab6biYkd4?JMvUE{w$!k*^M0&Djt0Dy70r~
zp8PL}_1%pTv~Ite1TEhV`C^NipnVSX-HH0zK>u}$fJ^w$cN7V%nHIE{r7BmwFv%-R
z6%!MtvuVzct?3ivU$l|d)Wri&p@F_FB`u-7a?@Ee-1PCxP9K8V{Z{tA6%H^vC;l8D
z`T!8hBq<mZ^)dcSlnmD${DQgRG73g$Uoc%ss_(8~!2!Fwxq4BFh4+WE@sA`Ni}#0c
z2R6kKw6Cf>TKliC!iv;zW`q<1ZQJrI12P~*PH`8#NV~Jb`cxF>*IOKLmdY!kQ$`^~
z2)r6%8ObziR`(gI$r?Xv37wGEspDUg3&fY;>Cwv>AAzc1`TUt{?=?C*Zdx^Sj(F?w
zZdn&oPvw`u_fW6!J!rfSTLtnz5Du|c_2DAopnnYe@a;mC7CziOwS5k|%y4K6a@~#y
zIe@1G$*5eke}Zs*+Fdt<JIl-+LflTlf!y*9VyPWhOcpVdhz$!mnXwCp5qthx725YA
zShB3Sj<Xnh$Ki5abqH)nr6^C9OlQM<e=oJcqV`l(&nSbr3w<eZrjV9%@Q|kP9IjRf
zhg{()u~Tygn#%Shn=r5p!oGZ&uSSH;jf&8QaPh%g#4Z|?qPT4slW5W-c?23j_YrY0
z)ahP2Lyijg%{5eFE3_SN`~7B+wp5~&?Jc<87vv7*!fj3`D!V=YPAI;pZAPBAMQi$Q
zq#)Hirn*RA7c;|?I1wR(yxfST%8AJB5^QKQf<r<R0=;iv`|K?n1V6=Mlq|$EJu6*M
z5EMGt#?BSK<?W)aGXXNBviYR?KI&Tl)ZFfwSuKpial6-V3<>1E`<>-B3aa21<)lVu
zJT^X_O#6tj@2fn}Su!Mzc^N0?8*T@xp&Xzo)`2=a8zEywcIY4ppF>I)Z^j!~GoNBJ
zaPjc3C0TK$eh*D}N5+hzESVx*hA3>=IZF?2%P9*=9iYkp+O3px4jM(MvHqvUE?7Hm
zl;?iiw||fsq-$vVf;1S-JJd8&C$~G?rGH=^e%MMSpIZ#xmSdqnC*a##9h4zyt`a^4
zVFE8}bRh7u0`Un&?UKI2S*%MkGzhPj?2a*dhv}D|-ZHt^0)->;zy!E*f@pBbjsMjd
z8HHDtqC_OEhs1}UB@9tQWLptbNQoGsR!0#!+LHXL&Bx8+Y}S$7zulS^KOp+hF=VLt
zEslEhGEW*=p5*+MF+g~aAw}u184F!H<RKytscW%dzk0#4{n)fPWtnr(L2OM*Cim$h
z-#2Joa@$>xtTJETUY8aqkOcI|&_N|3DmhMA?k<cZi4Yl15^x#&gO+Q8P8Sj5{y@^f
z_yv@-6`;06>iB<JfRqv?$}zumIAB#>{rbi8Q$#{ZCJ5vc&PKDo$LLA}HP1fv{9-64
zQd5+$dbr0dK}dds&)Rqk@?K$^$E4~cbUB5svW3S_Yjmo9JQ@HQ(ouZ1M1<vNwh3UW
zcJELp2f+?a*G1*!#pKP#;~X%)4i<Q+Ws#u5J+8RE1I{YG;M*d>nG8PXip5r}cd)di
z!;0BmLtXl;beMtR9eQO5P*oeeDY`;lv9Bm!DL|Pz->`>IGw?K@F|j2O5fg-ij@2&;
z&o!ly&#pM<0{hb26=$x%s0ZXwA9?FELr)(#V|vF+s1qUKD8EsVYEXg>=yQ|FX>PO$
z2|dKa;XCTMZTwEm<=o;OdWS3Id(Aw!naM(Ft6X7<N(<c*L7+WCU`vZ(<PTi`+-+d|
z?E?mihqu7lqMq(6MaNu&qk@g2af0>eKLQa)uu1~RdpVy_ZpO^ly^@$;B_(B*sb>dG
z{Ns}*VX9cLv@+>eA0F+!*G|`8HfL)J=IYZ6kz@uNzD)Gc+D;1gH_ZWgwUvM4aUiFU
zMVnpG%?se;OR<8=mWBJna%D(5Ci{)6{sF;8ZeAkailcQSZkXUIci&Fo&lRbaufF8R
zvlCz8q~K;?TD?`Ipkhda9**7~t~ac1dAKmi83n@%`v?8}_8>~h4rV`Wxzt})i;y8!
z=Qyr$L#J?Js(B2tN_%m%bZcp3C6&T{=vevUl;h|*1S$l=XrslT&I%fQu;k;sUo6vc
z?t!>91(pk@F$O=|zEZcsegz3?l`AK{13a%N^4Vb|HO<b^bL%8D|0UGfu@bn>VJzb3
znkce5zOV!qem?ql!D;<`NQnWij3+_;_9?fdafBM7cxtj5IOp6$T+bw}9J-)T0T&)Q
z1~%-&%qv+$lrB<&bp{-NtN)0XTJbNW5m=D$3L<)0o^W-ojL)47<d80H%{|n9056dO
z+~NW~NV8DuKnfJDcepl+x1^QW!pTb3dwlH{$Voo?Ih+t-q7ME9YhC*8;}1PHEsGJw
zSF}k+#7}7cD2C}tz<SuDXSP7mLPiZ+kqeIY`TW~bL)&GSd0mhfy=x$=Hc@uxRM!?l
z4wBJ=*Pdr{L6uttdDwkwvSh3>UK}4ee0%0=e_(xtnTM)uQUsx~Trln;#+YDVPCajZ
z?SBZ;U0FkeMa8N@X<ZrM)V(i1=*l;Zvg-S`;@lk^!7Dh;xH@LCwS?E%sKp<yuwZsG
zjEf!Ffd(fIQYi}h<=$say`yCCDUyQ((tx|qpC8J7NBQoto)9!&gK_B25@mf0Pa%>a
zSaVR7MA?V^8DXjiAw$xadvO9g@n69ip%vb1^H5I*Fy`&DhGnL)pygKBKkuLf86mNj
z)jtweQl%H!xHMdb!wmINJTV;85JB!C=DskA)M6G5^6@)mEMULtgUERLk)fQevvvKf
zssw2LpL*<QBg2#i>L$P8VRUq4IwU3f>MJrrYx~BlWfiG2Gex0Kbaqyn>MF<CYF$+F
zdW)LasJ05AggUW{e&rc532W)#+n)Hl8q?i>Y-VFTWu5F$T}{-2i%W93<nHOqvh>7u
zaEQy1cX#jF+mJwh^uXDpONk2VaDUQA9mHzkbuN`63j6YrmHF*l5k2}|DjN>EOAf<Q
zmy0k~I0hpE+YH)KJOa9-F+pdlvmjtg7CJb1kQ^rZ1K*u9M$zKk;~@=o4rh8h`+|@L
z+;sgl)4VH%Ok4anD4%g%e1)lbHj@r!@U$s)*4W31`$fA+mQRdbuU2GA*41T9tzpwK
zqeIackbQJ`gjXme#g`qjj%U~5eBQT93?rG+-;1n?D6hxZM)~cVBNi;1Ev=@x`f5vD
zd|D49QJ6AuYJ&UixN$ahpU&g_p(D@ld3-u6Qc!1|$8cUZpKR(^s0xn~n2E$iJq+N^
z+8qs#+D+SlV?ww%4S4gAhURdfes@E9G&2-~x=P5p?C|+VKn!}&Bqx0wm&>i&k0-q&
zdf&VQU@xeY3WNxO2<h@qXmC_FGXm<tCB@`Fw<3;Ub+@tlh}?zs$)gen8_H#n>C?Ab
z?LVM+y8v#7;~6PRH@{w_CxgyM`$XXj(K4bi;LgdMcJ7KG?C8iLc)x|YwY!^JR)(Tn
ziVaA!dI3FUH4yt}XxOwr2CU^J7<r%D_j2=38son(+<$?JqCMhJxYkVaGHft2rUIc?
zz-BBD-SX!0oP}Z5V){^V!t`D%eqXVK)<*?Tvy%vlJa1}Q79wTDV#8296i<N5?S0VT
zFGby5$ZO)792>B4EjX-V*@L53oe7)tJwH)=OM|O9R!crix80JHQ!Ul}c{tEatJ{Ne
zyQF0XrTr9EmmYt2`x%Lo#!O0uLm_KWIXwmE>z9^k)0<29fi<6aws=Fxr(v_2bi(f)
z=$u}62@fAv2Y-?)V3EZPk!A2K{1T^gZ^$6!q05gbAEJ!8e%LQqr_hK=2R$7bOYC6J
zDQxo`+W*A;rUg4E3wwim?fjkmn_RcmIqa?YJDh2H>VZG`qkn|A2@8)%k<;6^hOY)b
z9xyhkZhExR6qDoU_n8#@#l?ksb)RV`9^b`9OqO*wOy;G!{?c)~V1ANQkrN_%kQR<!
zSdp2w1y(;Kx2JT7B`u~iqi$C1N!W`0a{AmQ$!_xBGzc5(7Ik!VoCk0oKuAVK(+WUy
zKL@~<pBes*wO>JZQ5#><LlohZ7HP7lpi_-$tv4mzq{$_1A@Z-}jGE}c0mf>GKrnu&
zjQmG+V+ss;?`FAjdu0?k>-<Iiz4^Y<WQtS4utz@J-7_Pr-qx4$h{P>weDq~uUyw?l
z+4%?X*HTZDKOf8TPht{peg>42siUq)5(J$I-p1jKnlsEERFqTEG!^__zRh%&1+8iE
z-Q7|mi{E%CbD!<rZ`ZvV#rlr@GeHN%w}?Zy<(=Y%pdbUwzaY=CtH&@|ESB=L%--}>
zrXsrJac~_bb|oe+ob+~<;Y47KdvIdK<<W8=l6VeSJ0i%APC*_a6}D8|0~U^SCaVRG
zTU%yRjAU}T_(d@Hp@qBWtIun5bO=U70<_9QUe%<&&;4c*>l$(O1Rl(pKPouWiSP8}
z-3*kDs#uLTv;?CSA6#1TL1wK{Uo?9EHKtid9uDh(_2%(YR{#nhvguKqHI27PRplR~
z-210|-iG6L7E+Z^dC`4Yf)gYoU}w-Jhs46ZHBf-`Koq=i7`SpaHnz{=7__tx0rWyI
zcLqT43z?7*@l=8IW$!+4??i~n^?6JLN(R%8aF$gG<X>WPpWLlDKQ4;1&HYqq$xSTL
zi;M%UX?77WqZ*>PA)#QgN{o*T5W0E(%g!5q_vBT03EV0WpW#Ew2C+Hn0SgSwt25i#
z3k~F(YiF`<7*vxLf0YyOpu}B_9MKsVwXTUr8*RE?9>T+hlf)H4JKFXgetc5gTgM^E
z8$+RuRi6H|R*ABQTbWTaTSEI0HbM#h<u2HwhaaDuczW@BCi1Yrkxb#@(s6pV8!ZL1
z3noLv0LV^(#)lM7Kn&E;fO0nz<bVA<1A$OT<iM#TEZZ217F>R-rULG`$nN4d)B&}d
zgT!|m%@#t=QNr@(wSn{Vyq}?g!WqPd#-;^h?<60#w+U846cz~U>;!0}Z63N+nj_ny
zkZu3C0au%Nx4D0<R7H4TY@n#eBn0xRo8XE`&ahe1gd{GlNFIC(Kn)j4CKE^_k~SYn
z1xAEq{HN#Q^2|)Y_fQ-y<d@x@osIlN?QS5+D0@b}IJH;53*f4}RZD;vD0s?PM8H4*
zOe~J>o*$~>mBR0{fSvGrS89I|_M-6%wfJ<5*OKNg@5u|;6e{CcbwRPFANFY-leKLt
ze(Iee=NNp(zD1gRiUESz{Fam`Yb(-nW5d$wSI_r@V$EcI!lC*c!I31nGnMr4i|$1s
zzo_%dYB2XCrAKf|l$ShxOYDv#>-FH_;yij`M`;(+I;`TX$QW+M6y<IB6o0S^4pU9+
zcH5=TR}GOj8VJE3)7b&_z1?qcHbuX$I#)j6j^h4N$>pP!Lhvx*xF@P}rwLUer9Bj-
zSpN*(B=A9a*>3Ktqm#mspr;nMLA+CCM!SQAxlQUABg64_R*R%R7Z3g>zwI01H_ZdT
zb6X@Ej{Cb}YX$M(WeDUI?}p~STeF~4IGBoOZ+wRc-$yA&%EKnJVvt@A{Zg;vI;y3L
z{IsnA$>vxcTNxsF@O!l?kvCWOX^e!))Y;pH5&-MxpB(!43M*miMvKpy7|-f9OS^Y}
z^*Jp=3i~ThXo9;Xc1Vj0ObL^s43G;@!kU;AVe1EytkU>hF}k{76z0^+aqM?SzIb^F
z?)*y)0(8VV<>jbvSu3Bs?)2(q#Yx^bk&xlIxS%S{%-t~D#pH?`n^50h@h?G7n^GR%
z1}Q(jq&@ZiHH1XmgS{h~Rz=C6ZncS%lpHzTq_MJC_B`kE!3xFm76n5N`6T952@VNr
z(lxHpd0?|$Wq(GpK_;HqBkES9Co3wrZzoBosfItwd@)eyu^(f;)$rk3h@o|{WNA6Y
zqXf*!?h2vzA8v2jA5}iz5o$n_F!@V=gv$ayyTtqDsM&%sxng+^CnmkZZ9WyV3M<Zr
z8;kfANvy@Aq;UN?3`rl=c6$kcotoc2YA%}rt+n{lMlNB+9LD_380XOv@A}5EtuU=S
z0Sjr5{WT{iR`qy%qK&6&&srO>g^(^lT$(z#yCX$`FXjHQkW?<L!bE{F9p^*`{NZ|Y
zx(gi_@U)z&eBmK+)8k2rMv$4WK-yNV5_myEw<Yl=pcUyCf^+Bbs>p-Jwbzk@`EkL0
zYA*Tcw&+6$N_KZbP2h{Ia3S1I1e{f}bO5_AxDG{__uE>nN5`sH#oN(=Q103GV$q^f
z>+uD^raWc=6;gi$xOL*!JA{ji0o(0(73~j*AYa{g@8)1OJ<DcBd$KWZEXS08LK_st
z;+WSM)8bn_@9seu|FE0`M|gp`_cs)j$Z$Vzv9GCYMEq#PTJ-`Cd(UT_vWgM}fwTl&
zJET-JZ%@9p`xQ<EZh48hFfub1#r#UAfG4r<@Mm)ZC;pRsqfm&^;E=(C%$|50B>HOD
zT&}S^$aO8>9<O5v$}c`?aXJb$8u#vN;WDimRd;lq>!WQ^)<LJajZP2N`|K(_#WnV?
zTK0oU%q?6eYOvGwKS5jJ1|E-RvysBgZTOh|jF$cWO%==b4Iv}(p~rQtxjjBcvD8rq
z)Y)+u((uET#8455sDv6{9l<e#9pr@*Qo%vOB^BjnpUHOGq+((fMGE0MTG6-_QI49K
zn=W`*vF?_kFSrrOwS^7@dA!QP1QYpos!vtmZ5~+9Ce&IA%2?qxsK_U-*;gdJ{Znva
zP!%MaJRA~U#_U%ss4cQ~Khgz-w4-BPh6MmqdxI+9#aSw2^ZCYLH{%26q8msI{JJ7?
z+o%jf{{6nfrB4Ba)aTG{fO~}~*>-y6t=vRL1QbJoOsbBXc803TZ&p?(^{2Awy!ki4
zj^(EpNDrg*&tqw>?r!zh?An&mq)16g@QYyhw=V}<rmvFyRpX1&T{FXI@O}c65F13Z
zF>5teIqkyY*g7RyK6-hI;F6qEZDA+`6<7;UA2$O?f|kE6aP(Z01C9cPHzp=Iu3K4p
zcKpki=mX=H8~Gfd5*j#4XHWzb=%shyvkuM=`(WI#9e>Fx%JFp!>v|EZquN~3>so#~
z>pZ2kwhldXz}K?g9N0+nokNxMS&bgmzC{H2XFNe?tW4+x2>Eo-4;-=Nt(+qJ)dQzu
z3R0I3@s7Ev6qqxzpua!HFXbR9&KERyr(%-0;+JVfpigRSO>7K(!ZiYxD~7_WmU)@s
zvU%iGve2NcYO?o5miavW0d&8X`tf|M>8aYEY!SgNX6!n})`jF2%7o#TFJ+^~*$wf$
zW-CtSV(b}f%mEBvfYyyS5#BE=>-hZ_#<KjslLDdqr%xXDBpNDxKTO1HhBSDxXl1-H
zGOmNkSbaiO?mQj&lT}=XGD3&Cya_B<uJD&z|K0*}AZMnmG1YzR5F`p&9u8iK|Lt}|
z7{tVp!-+I0xr)!$Jv+!C5F%pEl{^XF@c?8F$;m{FLo-^KRr>38FF0gI;&?Hl%iDbf
zhZQyqwa`%7lK?35&auKEF#epG+T?1U#rPKnGs?!e1w|sz@aotWDq4!tmM)9@snij2
zx;k2jblj>~B&vYid@9rtqr<B_)M07DWK%}iWaY;@SbAXG0O=IObNNNYOzB%ek$*V#
zNv*&q3BYm-&lnvN9T2)uWRS~Y@Ay@pEbUNH!U!#co*qeTimf20B7S?Cv(mT6q&J$_
z$+(wMI4~nPz(A$IZJJ6s<GP?u!BEydx)Xpn{}~d<s<--th%TClzzWEEaxlVU0Dt|N
zPVv_RJ2LScftJ;KcYJ)_`F;vslt1tEZsCH_U`U~bA5QGZ&PEUzG1@$><%=R>9;)G4
zue7+mbd~HixvTU;I!EX2BNA`Nps=FNtbp>kYW$(GMwDTU!P(j;@f?tyBHl`7+jK*W
z6fvDUBeYIi#49(*aW$QK_mCT&U&or=Np3E0vo;kjAwso&-aFWHDVG9_i*v5@qEW8M
zoHr2_7ggE65v2>Mz6NXb`^YlM&lL&O*7%Gu<;eW-6(+$$1-8Ma?F5apW&1HG(Bk6a
zd+dBooP~u2*W<^3sa&O=yufiYDLc9-CFO@G8PBBiE}6|(HU_7s9WX*pNh5`?bxV#u
zGYO!UN$96bkkStKW((pGqR&3Q;`SyYt54@?u1`oUbuiF59D|W20;}M-IH7PbYxoER
zfV(7Nq?dNWpk}2ExBv)z(QtNE0__;7$RM+gv2y3=DYi`PK%!tS+<XoqIDD#|?T`B*
zw}wyE94sMCE-a`Lv7C<#5@)(A>I^S1=uoXV4&`9kZj0rz%MEwR^-mrR0`x%FI)nK&
zhCc3Dhec*>`Q*k0F%{`Pwf-`#_l$RhDtKi$L&pwhPnfQc=K2oIQt$K+nD8F%9B@!N
z^SOkk8obec7!>9j;(dM93N(y_O|G19&OgjQfq(5M?%Uy3(w9*vjH!i*^NQ!(diART
z7hQ=bH-BmLH6v%0wbW@#SXKi8bXo+YHp)@@vCBR*i2u~k(g3UH?2>G>8k4nLecUC_
zzt;Bf9ttpB(0u;Ok@$JmQf)r6i}|6hEDVl_pYM^&%bY7vVpn9vkpTHjh}eHf@Jk7F
zk|5bmymfZ@7AS9bSHRDTeuMo$a<Mnh9@@wEc%t)}jOzV6QGybtK>(o8VX@W;a2Tu_
z*IFD)r{<7=p2_g?uNzDIm}d<lGJo1NUbj0b<{zv<JhVka1k_?$T}=br+B<Hxyw-%w
z!ClCo$?=+|LJyuWw>`VOh0nhVIpT9GmgJ10BK%7Sk2S_5Yq`QCUk<Xubv2~Jd)?tI
z={Hy5xICdA%-)qn(Zbwn=l%W{xq84AaG7UbT2F9yhH_TxjNq<OHE94n%6)EQ(>pQw
zGItJ{tWwPJnZ-<{N!kx{{Dg;?TCQ0B82NnCkBu1TqspGaD_fBNNOgx^MWKC!pNXg1
zmjwMfZUOyL6LfNYTxdo(##9LD6#W{1_cJ?=d02+TJAulYOg#SH!1uJ=-fQLezL@oM
zXlT6uP_HBSwS8B!psX6^Zb=(*-9vVY&*+aq_1*D=FX?D~e-D~k7k7*VKNe3MaV>kt
zCiw7gELkS@g=8J<$F@k5gyYCW^ok6pKoqwOBBo7##4~~bcYWxNT~ehDcZ?^5yICsE
z7e{a{P$wx%DXA!o#4t|wmNHDnOe=-<dP4h`dn{DjU=@S;xX-sLPFJlzu8?EjYZya@
z-R12otd)|Iat3JSK2uR4WM^lO?>o2vN13+OQy~8^YieIkL80f$-j}t63BOP(*-r?_
zkph}e2sUe;w|Xbb%}K1iuAPQJ3{l5Ld?g6vbzQ2XH(RD=o*4nm^T4tUq$4^9HoXLD
zf7eIN*E9tS+^TDKI4n!IwCHTCGS$p}re~HINP1Hy_3qNkFN&x|T$GqEGa;>GvB8=D
zd`4!v*)`Gmt0r4Xw$RYpH&+6>lHzyntNFVOp)wxviWJ9dEDtYDUGy)7%XRR{hi#iG
zLvtHu>S0xYwSM?m(KiF<Yu~?&@$oOakG21vIzZQ!{uD`ZrYCt>B5CT#0?ATjbJC}s
zltjs~(xS<acwxJQ3p%h;vW1?oq!lrrD#YwBKGi$HWNK%fq|Y*{$H_`TCCF~9b&{Mh
zYxtT7ISB}=8nJrm@J=|Ac*`T5U1}5!Z`+X_v(dy<U^?4jYKcP7Go$_n(UF7rkbTNr
zB=S&I3_*po)Pou;QWxT^AfnwS^`7J?jy>NWvy<E>v{e~|aQt=Z3;q2%XXN4Kta5VF
z!K>>M>R}H+mig$VOC!P^Oif$gq{>q9h0af0NnfA_##p#u@0se(X8hyVDk{$)82;8N
zq$GPv{tI^Gh>pkai_<Zs*y5`1>(5pv&>2fmPe%ZJQ=(Zmg|;(oCtn@aQJC=~dKj@@
zhBb7T7}$_i?_z7PsFIi(QFmLdj^=2|@FRb*{WUQDpQu1_tPDWbr>Kz;#ha2HzyWLk
zzSkas+hu<s##Aik_t20;8z8cFdAZ+(0=<O^o^(7~O;1mgQ1r@j-U9bkUBoG<H8s>H
z<Df%Y&Tv={BNk@HiQ8QTdGsfzU0t$@1}h{|LWcZHqRm3i6=DCmTHm?xvSL}9(O*CX
z4263`uH1#;jwza@N~u72V}xG!pW-Cl4OT21_4@d)a_hOQb`}43?T3hC`Mm1$nwDht
ze}QLC-aO^w+*^nn2dDyg-NglWUyQC!ffDG$CA~~IPDw`lE-!CI;dCQZ>nQr{xN~6G
z3rRLrIzr4hx6zaii=L$#omy;{SMsy#ZN>-$$Z%+}h|`kUOUXUau16w>S*-}^blS*#
zaPD>xnLb%;X6|HQc+Z^}fA_Ms59R6hoV6wBm?EK~L1Aua6dF=-4<6FXs%EhH*Ke{y
zj_}wfeuFcKcRWO*9^+H}|9Cp9sJNQ0TQ}|!2oT&F55Zjn1b3I<?(S|u0t6?xOK^8-
zNbun9O>lR&f4|=t=iK!LG>lQTYu8%qnRB9C_y4Yjy_-90ri5nu8Ev+19-vH`!{@yH
zX+zIRa1}Zp&dnwH0@I<yq71u4E0BaQKGCp~^s<n?9an-mej<<(xM5#(boUhxf#tW`
zwteGA(Zrvw4Z$s!Ya(ahc=?rMw$r;|;@_Lu1hD=n-FdGPsRCYOJDzy8?rgl&G(`9F
zvH?Fx$&^BEhrs!6cd&07Z!1635so*1uA~?uT(W1sok}3Ys(kE5QOM@JYcuC|hYI#&
z^0yr`{3^waOHUWoB~e}wds2)o`6JJ7`A?1l+0ADR@?ax7Etj`<RHQDvff49@y{|I6
zuwV)}^)9x%God5G6>aji5g~X9a*d9Rn2fh}p7){smv{8_^%Y2X>IYCqG*G-yIva2|
ze<x)@qJiAw3(Na;d4s4jA&9}6Q|4L<($$t!zM8xouGjIaZ^x1c$%9T*kH@L@X9$PB
zBr30xLScV9&-a*Qdf_&h^*j2anv9E$2%z6h%BCZoxi%LPmNsti>g8NI<qew$@DwYS
zq2>&6$%zJ6Ss$Slqx1}r-!Pan&09N=H2G`qHwQYI*K?QRaDJ9ui=^PHQj$;3*`P8<
z>fbvJ8jY;zGb;6tz!<6{ovn<F{4)%KK6b4M4h?DN7g(UYL6G=ry&{KGYfM|~h7)U8
zLvFDVXf34hc*BS^!I(Kz`DF71$pvNy9)=XSLjrpYM>m5X%KQVgf}ge&WMODaZ{MEm
z@ML7#+H=jAWV8`%_x*9vMa&2nhn;>bl{^Zd-G_Re@~ZT9+nXY-5p<TE=a@9X(y^$8
zd&?O@dFcp$@-)Bk4Yucs9AC1w?!&J~I}<-r7@crF)~_uDx32DaJFeLWzM`J*DIpQC
zC(pu>6U1lDWmt<p3ICA-*_OhRXQ`|!A`bI?;TW=NKlE}9r_)cP?7d#M7di-7DN{bA
z`q!I{(#Fwu1G!1Es>UgI(og1z&l(G70^<_B6Ppk<=qUvHl8dq8Y0?>KV`(5><zv5Y
z{VC*UZ=PK1%gyf75kvp7V^il#cK@XzBpLco!<6D1qUdp`@Pb1W4|6SXSaBi%+~WBc
zt@HGYp7TARM~FUm8o&J*Y={!SjQ-!XQeOSTd2v5vk3e;D#ICqpH?TRm$1_oZsj(@u
z-2d#9BbZBv9%_!Z+aIo$7wY-07$4~q%$qRp;_9_D%1i9V*G+FR^Hc3xQ5RWGR<7)I
zV$A^>)%eSXzSnuH((9M!6md87D%4Lrn?5{)Zns2hP-x%}6t7*l1s9ombP^^cR09OE
zt^^?p^QKD~)^uid8CBggo%0PzV+^M=99b})x^MH3j8x$n(Fp0`UK1$!(|$n9)RJ6O
z5=YAWRVycClMa*K*4Hz~OsVhFa_(eIAdr8}5f_}j&?y{ULO!IC6Y4(x;&~17!P}d_
zIDym%gZRJ3m?ui~YJW%fGtR&4KK&_|o82(6cH3Pdp%xK#Qplwj6xe>j6aVCK{-{ze
zhQ4i94&igR_q%)--ZVZ4ax@|(w)<*-rv#f<o{nP_7hDzE-#GL(f@6ee`p?HO!KXLS
z2+DiF5d7i3h3(U@cWD^YR|G@3nnd-er7w7GbEwaE^!gLFC6~FS8H<d+u<$$9Ee^oI
z%%b^a?DUAyC>_lj?4f2Y#XX)^e%YW=ojdbr1aK?&xB@3xYo#p{#bIHpb3N!rD>A@<
z7?|KMpL84b6I!#0uYBp_2!I1m?Ml0EXzv{gJAV-u==#w~JI{TKH^udqJ@=eV1pArd
zVnBuXOt9$xu>g!AE|PjY?j}4QwLdnx&7N*gS|4^Zjq8-n=K(i`NGURJcn~%TNzC}T
zqIQ*LSQt=aPg%JEPsH29#QkkcPb|;67hEQ{{pG~)4^=@$5DBl7=-AbN<XIZLb;N-5
zw8;*~pnz8t_#<lh@ZW}|{MhFnsb4maa^HVlP8c4uz#1&?J>CH;^S|D0Z-Xr4$W=ON
zBSoi1$a9=pCc^V_34IlZs%&fPhWgquwck=4$(?Uei`-W>K2`Tg=jL<^Lz9FN>$6%Z
zzA>e)86B+afU-*@jq-*$d_W!P{Q+&Fwo`=0u45CAUfI7D;$>f}cdzF}mq&|Z^x|ee
z(A%{>&8Y;5Fec+)*O04VzHdOhH^ER(w-GO;ah+<p3RXhBCn$}p^|0}x`iwWO!?ho$
zh1G%i3g`b>d(ywtDXYFwH$JtoON3v}@D^p7f?30SUC`I4V19mh6F<GZD=t}8D#XaE
zm+8!+S()cu4MC(3dr&!TIpox%yIMVLN0?hrdbGVLrxU>hBYDy?C9BLDCZKNY=Ar2h
z)UzRjL(M@5Tk4Cp6gMhnKpUS~z5WWDK!Y2ErPwgq1Jrcre@`k}%)OmVmKktp+;(G(
z7s^_68S+Ug5&EtV!D^gDMx&~-j3>NOGJ%B2Ss!wxQ|7(Xj|XL=!GJjyI;k)dtuNW1
zd7B`|8rXFagkxU0txzH0HUzXEGAks*l#Oyy8b9m>Z;u-=mNS$WduJE)a#}X^Zc}QC
zX-*${e)H$Cuw)@5^PM4V-x#esvBT3)1o*kN0QivnfATbARVmRE;7NWT6ZY6B@u-iZ
zq3Xt84{n%@>$Cv_kT9Yc@^fAoR!?BI#RGt?Mv+#B>fAVP|IIL<3ez5*H~0LSlD$FX
zcF-Dw%dqBC;TaB|>kObug$rm+2)rOHDn?kT5jjK~=hr!x2A%3ovfQ0Ngcl)HH}XzP
ztK$(C!7yg46EZlr$z%icz~}7(-Dbt0Z;BWm2FEOD8=g^&8C4|w=UueA5<)(OPxhgW
z?_*3*t8d>Cj-*zqP?gv4>rt!VZBEOGnhKon2QDqN;|%Tn){>Va&$2?BSH-dUjb5H-
ze_AAZvj`9UWq%Cw_Z}H_wicUAEErx>?awhhYNW?C=;q{_tW?6A3~YaDcI92`?~rTY
zbXHX`7Z2`RwpNE(oVMH*pbe-!ZJ9&06X^FGFV=c!CV)rm)F*DI2>7k*Tb^lM+4*O4
z=AU$G=6Fkp&7py{(_}*6)dvl}K<_A*K1D<t=o$^3-HK3B!Rn4x9V{Nk$die=3A@wV
zG^Bc6EA$>2I?tyb)++|w?RTA_atk+am905-D~cs~`@wRMf%K6~O|GQezv&~dqMl;#
z-u0E5lrW6J&t|)3Nr6h1Ecz9l?4jJJ8BzhLr#W<%6~AmbGG}^zf766p75tvV63(T5
zK!hVdaj<PQQL`Y!a7>y+6zkfsv}F$2xN`*)^O@*88gYvt8Yb=!%q}Dpb&I6{nB?e*
zAL*|+iHNDMeR>}F06JlGU*F`|UD>$a_+qnS4Nag_1g*NwD8!o+0;f!%6A&!E>Dlu{
zO6GUon=ix=OeF>@O@SR_)vlmk=T;Dc@BUery`$3smN_~ZA14Q#kH-}JSY}fazE4o&
z+EdkX+{3ATJBU}T_3LIi>5HBTSfk2O;LBj627(+80wMPVk#@1u8hoOR&rPkX&>btO
z(}t>UEzI-Wat=`!A2mx7?5Zz|4=t%F=T*<X2?oM}4!F?bM?L};8gAk&TZ(Jy*QoD#
zxec<+=T^Uc`li^YLor;a5b#QfKdRJvg&d{oNu_R|BT&vO9?5yJv2rAFI<XiW^w@=P
znw9w;k)UE<K#1k{mJBLX$XLukZ-Ss?XT*=bd^j~B5K78KvqzH9o$8;7hu*9;$BgHH
zBi8kqW;h;`(#Lt-34vMcf@?4?9{AVoZ;a}WN7<uau0i4B28niUpL>y4hOFWbBfGpk
zb<q)qQWcB(!dtF}7mDga-Ci1ShbFms{I|hK^YU<*cnzNAr1S-5bydotHyn+d4o{)|
z>Ta97T5Zdio&H(vN<GaIAuVGCv}kxCVFy7U5UT=dLK<g_FV|5}1>4$L4s{tmJTLl8
zB!NKIrwt<}bxz`+YM-bc##%cfswW8jl=jk+_^N95xnO?uplp{Uq=of#${Z7BiY~}-
zm{J6IYzQ947Oo?_!DO6~f4XoW2Pv<}W|PU9c8Fp}y4_w1*#e_SuEw6RzU=%14(u6v
z!YCA2*aQ&pjtP)LoHQ*u6yb7P)>l~<wZDnOUE<JJcjn?Ey&rH{nDeMZ<d+)6H*-u7
z^|@{Nt73Tg`<><i5pmj~pkXcHH6;5=hNmgNaN`N0VPv)hE}vf4lipjJ67R-p`}N+5
zXP<$>x$(W#p4AFsnZ7<V7-aZWVonA?>=m_JPtea<C;flLF~c}*xuMe1duvGhzqPqz
zViStpWtYk%0!3!x_tWT=Js73#v#!oKA~#|gqOiFYfLdTaCslAEmH7^%>3S%@LZsCD
z=h<cFWPlFlEaw<&c8OK*K$utS%VRTcmO5>F$QPT@%##%QGAexo1C#K;#aaniV1D~B
zR$J@PZaK)75iH{OPybwm^C$S}>(mrnXYGPNrKkd$%z$0n`oo!bZf|(g2uct(>6nh!
zPFz51&S^zF^wpst#Af>Nw0=b^?VVM1-u6I<^KAH@b-xz{<!2qlv3CtAiSo^6Eg4%@
z#>_NfDb)b{01bbkydNm2{ycD;-8HMAtEw~@#ig9zXVww6+)u<f)T%AOdeKn8a<Qcl
zvnvt)^wehs^V*Jd|4`MPy>}J=9>ve0`>A>{*(zfYKpMf;N{&XqoYKsF#tl1RS6gDt
z#9yK$#ch)aL|jE$6rzh9QUJEKf$FR?rQmm&vREu|hG!e24XA{O{h>zqjX)N~9yITK
zB4kGP3A>i+e;N?3znV5bgtWeV1|DIAM*LX1ld(e!)V!SPqTxn(n;t{{C{EqEtRgNq
z(AydHPTqbSWBIPPS*$B=AMn~nso<h)q70>-3+L)~yxYORar4Q`KaQ0AUHwKxplIP9
zH^hE^iqdY-3Ok2`jU&+wV?!kaDM@d15;nm6YFz@jAA+lNlZLOte=uo_bd<5O|9M(d
z49^RdXm8^i1``xv5q-Ee?$0Y7N7vHG2L^sGqd`5V?(0jNj$WL@7(UPr+u)5;H+!)(
zLMaz5pYf$6a00_=;1Af`Eb`TtyeSac0BnR@iU4(cO!DmWL)EbhQ(YDs92Wsj$A{~J
z?1Z%Zbw1?h(My^4+QK?8d++kQbJsXSSrW`Ag5^SmJiH>&5xU(&>nBEX#BSN{p)}=-
zu`a5{W4LaJ$$K9`UH;yLepiY2PDb*55O2jV)IuEcs9V>YIEfTbhpM-x2<-Y<x#2Gy
ziJm0xDSwhbqfn<}&O}RMC~8I+c}Tv69y_7XERy7I2%zl;v<DKp8+Y#+$)Zh?JOd;8
z$k%z<2{U@M@}-^Q!)5vlq1k%w+f~-^oQ)b%=YT+^2~fAs3$!nfy7=@=!BbwQxG$~H
z9NbjO*}|n4OmI8-A?4lCnzOeKXRxqoaoo^8YX%yE?XSF%I;-G_c%w_ak;p>#LONG9
zDMigmEnW2UF+A}tpsRXb0DddowS_Mv&ZPxKM@QEJ?ipDkfn)$8BVU49({-D{^51;;
zK5{hny}V~KWSFHieCXNPOwv|lGv%|!<IYnsH4P+`Y0f+T29V_V|K-+UDJzM332`0|
z?D58Tx~z8uGk9^l<Y132U7px~Z`JK|iHtFC+d+DSmCF~t)QZG}(_GfDG$@CQ5l&><
zeE3LP3=&y_?{tnSA~_Bn+QJE5*-d<}=8KiC;!_V9QOtw(2)O>^TPm7)y1FYvb)w&#
z11rVA=Pb(WGYrbge3OA?rXq-e!{75V7%Q6t<?4yTL~fb;Kc$w9nmTYFpf-Fg5bz1c
z8mHG{S>X_lsxfK(g?m5F)JRXBCap@DgJ|<zXwm}syrvVf83k5>?2MU06zPVm<0%d$
zbn}&WPU9Lh{8ntaX2;U7Kg+zNv{eht*k#pNZ1kW*#bU3#;m<z&segGCEC*RC#BFeU
z8l4*=A|#}Qym4dV!zCC)>>DBiY3$uTAQXEyx3GoJs9CjtXRB-u-yN;YDw{Qw106F_
zEm3HFEr$zubTBLs&nJ0$##^4?(TuknrQ7>OQQVMbrdCpvfYQN%yM-2m6UhDPF@G{V
zt)I`V*}pSp`yj?bHgU(i{kCWGWZPNN1T6Fe3%_Bl!-9#>KfpKTZgBTAEFj-Wg!9?^
z0||+wU<QDpSO)GWQpb6A%3jEu1LD7@=x-$@YAuDj>O3=9cXXLUXc832W(B{kR_)D_
zRpftM{K(BSGTTgo7uDA%<+fiAqn3(u>d^#n=8#P(1+)YIcJ{v~YzYDFI0UvqJ;#*Z
zXpWqgM=L|x$zQ&247^e;palJqE`NWm=>Q?@PJM&1V)w87`{L28x$!lHX-_CWFC*j)
z6sYmyjvM4Yqb@>|*66DKvw;dGp7WV@U|1zt;^?>-THi0rynKtd_QuyITnAq!PR9~D
zPD7XevL3?$FL^Gq7CO>q#7yI@>GLj6or(>omCX@4!S7SHlbf<-4h(aQn1?00+_UVT
z=7pm|ZpjkG`M-+Qzw=~)qJc)BXQHTB`HE2q1p>mlQ#R6=4(6XUKnu;PG~N~@);`3Y
zezzAtbKp4fX{8cq+u~HlzwwmK?V9+mL)b=d0a0EWL6UWC$8tFaN3gdk(krmIN|yJ9
z9NXsSP}wt)j{Jg77^}T9XmfhT%6(X2CIg?%xvxWLlh3-Oy?}l#vp!Bz<9yR6B%X!-
zI(7;-k=hz3Xvv9C^7}>qVEOo(OxAPvPFFI+zX`LDywX$C`CWXj@@Y$pDPRB2EPG!R
zNIIjgIL4QFM@4Q{0piwWN&UG;3RK=h(((S8CizMjFEjf<z)%&1j|&$B!czbyT<%_|
z@WC<gj5sD-m3JlaRFk#ZuzB`=$$36N*QyXstyJ-gxOmT2X0H)sdpIe70c%F5P?W+c
z3cdWRJKpo&raX;PbE2-3Q2rpi;3(lU(kij~E<mxapg%vT=}MG`cQI7g{1l1fXI(`A
z#s(q9ao+*uT=|F{OxCqFg}{27S)%SS@8Da;Ju*<;mmgnRp0iR(y!elRQkAxANScf%
z1!@*g_lu3H^(I1uT@r)OYVM9n{Ep8?|14X(jbS2HH(zyGDrBp7jh2sfIq)MqE9&3w
zU)Mf~1tP&Q;gq?wrcB9WQ!p@DYLtd-%6x6D&*>`l!a)bX4nMDNMXTBX=WD}zd&bC`
z^G+`6-W4F6*${m!R%!!ZWpNNT>T#{Ub~wP=ztzBv?+Ml+La)#m_V**{u6`Ei9$r4L
z>uw0Y`G59?10ehEqH92Qzh=->1sO!Jt4R_Kz^>-+Z_SPMXIa!MGCo{{ZZ?Q5^l<1&
zBMGZZ>2q*<?Fn}Oq{cw#Tx*Q`OD<*l*^;G$GasG(qr-llZy<fV=od8E`kYSNbYLqR
z0*#KM9wF0zi-U4xsO1O39t*vU*yvnLrSU!?9pIN_b#^9lOlglKz~dn%FRzz!zrhlZ
za}d6~0Pu?g`kh_)&q_*4gK?w+tlpQ~Fyx&;Pm52T(~I^B8ZF+RLkIH6+Oj=sHa1~>
zP>3mt?x0)W+34ZejlBu@+#8}OXTRR_IN6o%tbu#J7&v_G2%M!Kd%XReKrDRG<-z`x
zGnApaQ{G-)T&!HCR^{Iswtf+f4DG{U+dAdO5aKFXwA-kumR7QyMOH*>;)}V@lAU6u
zGg$((bA@Zf?qUu!su5AiMVwlNZf1VwJ^T(AcN%#5XSV~ahI0cuSM0z$e3QN)60G8H
z@P*$mWO8I&j)j$?Oj3R6=+e(V;NpcC8S027T8*I&*HSRM_bl?i$HY%6%G|o*VZdqi
zrDSE!h6VMs#$G><*1*PQbDk<HA>rV`>gRY^>bHu`jlOcz-XK<_{LS)y-L*b|6F~Wq
zTuU<slYunb(w3SivolMj6qASBqt3(mD91ux)RuG7o|RQziE)PPNN1YI&ZWR-{4ZIU
zu*5lPRsqKAwsHZ{3c23~A^OnjS>{JESC{}N&Up_HoDAyLOzKfET?;J6$ze8==vJFF
z2pKk;g4x@ovo;NtI-8A-c(LB>e&|vrF$@5>-oJkjeSWz5DeV6kw9(Ss{4W=9lE=<o
z+ZVqX-35T4w%ZrLA25PJtt1lKBgz`mR%CKc4q_yvjj!e%`e5Y(4xXyzrdA}if6#_Q
zn0)DPm4G7gBz&$A@R0`qPPoA5u=1WPjkEb?)zbcMs+)<Xg})_oc1n>B`(W{qM)l3Z
zPuvtY{{EI+=TOGI$Jay&+TTCch5+R@Y)x+i&o@t;uW}T%|9;JPe~RREjTrs&0hCA`
zOc8gb5W5=~$VpTl5OC^{cnL5$sNd?1&;l;juZ~I@0x@}ZF?vQ!dMBOy<1FtyeX*1m
zMw(CRWjMPVSUKc$4Ebr%@24aeE`=GU0G9hnpk<v8=GngHYR^%h_<R-fk^yEvhY%O*
zW2XE%DwAtU=o`qJ(hHEqOc)Bh&-&$Je|Q>`u5%sJ+|ymoYwnx8Tlb62PbR37(+3+X
zC0Wsyz<qalNWrv5QYo5*@7rlZ{e7(2?~veFj$kRr<1<}Ul-(nj8Kkyu!!gavq*iTl
z8M;6ZgPbZ@vOrv*9h;bHhVlxx&5{o&3+SUBwqk(*3tHNw1OA%pcp`R1*kBRXee16K
zuK$Ro3}Sid0`K9wg`|nZ!uMKv|5@|Gf&kMRvaj9pCqn=k4tQhrh9ICBu|EtcE3qQa
ziKC~ayfnZM9{)Gr^mg-p2X;eL3a=OHqD`DS4@*zCThZVIb;XqUR)0v7)(FTM>F|=Z
zAUS`@43aXmoar57(R%Irm{7922tQ4&8R~$GAmnv%=X)?TkgiUs<1u%`&2H73_)gA{
zgWdWm(Pvub2N$UApqtBh_k{PNS=`*8jL^<^S|$&xq$mXpqd;nQIOcOiI40^^>M{yk
z^RY_$Pv1$NtbExL648`)>iIZC#<Tu0?p!vaND>YJf!S$f|B{Id!@<I!MI}I(8-}fZ
z63qXxs0^-N3q00Z|HO!X-$!-#^qmdhnx#@65#|8FH+D&BbsEF37nnT&rJGNRCaaEz
z>0G^g9y^EZf$5@P(R0AwOnm5eaKUI<q~ZK0P_vH}OZ6DV4lpR&8<)qJ8Bz>i?9~b@
ze<sN!ybUQ&@PD1ltIQYqP>~SylnSF{uXfu%Z4D^2?*tRbt*kg_6`sD^kF0a6<eTMX
zl6<*x80;~G73yipF%dWg#PhfbYqZc8+}Q*pmY$3cGBPh;aQ*b&yVid2a;~vNi3nko
zEo+mb`VtloHu5JPtI_LVfQqmc2aY-Ce;ZKMEcoVCMS*=z5xk4oDJ&K|!qa_$@%*^Z
zGmy?<8K1(XoauCXyyUf?AMqcVf-$ue`$;|Uqf*XlPsj{<>7@XmjeV!Wi1bm&exlLq
zCok)M@SNJ3qh-CaMrTSPzdkigLj(O{hYRZY5D^rK_?zO^qdZM6N)8o!(_42hb`YCx
zne4CQa@!NN3cULlMW`9VhV>t!Z^DxmGpZ>Aegt@D!1a!xhS=(vF*t;$!i6EYmUpnF
zHxLB0JLfJc(Zn4w@~Opx&nk&3FAn3j>C3;3MHYYg7Y4osWEG>oKT0Dg-#GjnfAIhw
zOAN+g;u{xPm^x`B9doDz0)k>-0~~K)I4s#X{zk7Zs<xkLTiGi?YiVIbQQ2!_W{B7=
zZB?ZN@v5Wg)?*YzE3N(Xc~lV;;D!6N$b8eaTSA!3*=o)`T(QWKwl4Jf&6B6NZUn+T
zKv6*bS=Z;(3)Yd<uT14PZnUStVRfIk4gUiNJ%+bT0_0`k?yH=nydqgU&${#KL}W31
zSH-^{+w_dQ7u&DsX49r+?Eal-zuF1Y0dy1&ZE(i_l<}d?gGi4+(F?iO`xcVh<OU-W
zm<v`S?OHfoDRaFeN8g3!sBX-+1Y#Mogw~u>u+>kn7|{~uEE~G2sV?49%3O1)!v2Eo
zbpms8zW?&&%Y83aKoBsZwrbM-o5hFQ^K#<o4A^qSke_`qPYObt69D};s0P#^aG;FL
z%y$HXl=w$*z^DJv>7n|V^K+-p=gZOPm6pav=t&EZ_53?xYIQXR@M$-lE7#C&cVlSO
z-7$bp-RcH|lmo@M2F2}{lt~p*b;g=NsNVuy5%drq@=fyULeH|e{{mFkKi^mum6?j(
zfh9&%D*mKO>r;3@Tp5>oJNhV1VZWodi_TfTrYCLxx3e-^*zHo}HVjd%Y2`BOQ`Sc$
zB7QE`u*3N5vfg91=c<uzQ38qf`xQkFyNcrNuog2qkVr>*`5f?FEZXk6Mn1IhIA5{_
zglJK*m1GJveV@WzrE6WdDmgNEV=Upq?~!dDUVzTqmbNUXw?P@ryM88w0<bHrvJ>=S
zZ5)@iog#4*An*=C{zy+V?kQlSNYX__2M3aqkX-+xWJnZ&>9G=gz#(OZi9WKfjK(On
z_0(-|kE*yO+fihf8dq1zn@DoiPM5o5sE~@l@a#Pi{r#7LVR38wF(aYB+Lr}p$Q)U|
z`X{Vezgpj3U&`z0!jQ7j1^i)NLx!2#o%<y7aGlAZca_OpSB=wc79c>ed}?Fle`_Yo
z*ppW*_WeCWB&t&~{@$OpW4KEMkV%pCMDd59+3tU({vEWc?h%y-$O6uZplv_&>B<Vk
zWJV^1+vrD=pWW!Jn`$D*%}+RYP6Rg#8?K*k5giw(NQK-sjxCd^lgtxUR62|lD6p?x
zpH5oDJ(9f9v9Jb{l!T+`a{XK;&&a%>FoNe0I7P!2-#druT(o2rAs8p%&G8S&SB#jl
z`qT6F;^?<0u%YsV{TJ9jmO0*t791&`v_FnCKmZ2+-_ypMaOM0fogUy14&-Lm+OKdw
z|0C!LGB^{A=2$jh!d{e)DZ-yCjG4xhHS32#BPlN%!`D40z8)Fy#|h#yU1X92#;m4C
zOFOpXnsxS;rhYiJy}wZf3^x^^fAH$jj~Vm&%Y}9@X8)a*=`q34R{~BAn5zwu*XS@t
zcL=Yhd*+BacG5P<cy+i3KjNaJ0V7T$_dfe9nX{fm&j@(-s+i5PWDP3dB|mLa?qz=n
zrv5A54~cStdJ)>=l4F|1&=q!lX1>Bffc26maTh*UK1ae$49eYr_47(yC)S&0V@A~G
z1*GnC*AYLnWY1ED|E3zH=ToH3&&t$eSA%n7@S#apW=8EU(Nd*MABss@B~q%^T}*Q<
zQMR0f5bU(>@7wJ?OtH_TYoa>!1!Y*}K&M?fF~^2Ne?Rr`TB5(zeZyG?mvXSA;se5K
zL}hg>Fe_MgfChF$UoO{FjV`vo?%QjaV(D5IGnK?ZD$RS4$F0gz;`Xr0Y}ehe_rjpU
zcbIZQe+;lP>|a8nsDQo0JHccgWvS7&@>k23pI$}}X6(6pY+ZL;IQy%e2Tsy)usMt9
z$AU;CjFgVdLAzqEgNZ0jrY_I3<{leOcS_UG{1WEbPU}ta99c9=$?@hydv3ZJBCn6)
z%^@FU3KmNGL7op&p1_7eh9#U!8v<fcd0n=V@jrcyj3Dzt_V)JHZnjHbjxaXa==P_l
zr-uL%fD3@4V{}+qUymCk4<rHZAuVej#Q(Vy@89^$&dqK9XW`)N%ocE!xd8=s9r;ZS
z9x};8A65A`m1(*;*%5XjKqlIvsX7<0K`XwoM)1g08rsxYrjsl6oz`-dGnBJ2<(dzM
zf53C$!g}HA(_hyaSleFMGJ;9>ZS@@pylC3H`!z+L5_k6V%g^PiBhjXYG0WPCaI3ER
z(5}H?#rV|fmjx`fg>ch4$*;V_cnWe_Z;QGZ%*(H}wzY(M;z-v2$QUVqeUuswFQOL?
zJrAQyTv81kQihIuUz^uYa-A}f?o1l7ANrq#J=rG}N-sn3YDiqR?kLot+agjygoLLZ
zfVw)UG1pLd|0MuFhWb~N=m4&oh$Bu24*jop?_dG<mAe%>N!SIY%V#g~-bOd!;e_{e
zSAB^--KO{pYLgjw{;Ya8_gQv_I8S@Qy{@W+$$4F#C<t+-#cC~vGWvmi6vD8r#_Y9J
z1ovYR>gdq8FSuo{18K(}^Z25+aPm80?w^DiSs&*yeotIalT!Ao(>EH^YRrw+yDa>^
zaJlur`zcMnk&;kddR`r#)ytT!xgvYjr-t`}ApcqSwwg=_arZ}|3AGlx!+*+w<#U<X
z)fhq!<*#330Bbqphszyc$q5}gPPoBfu&B#SsSj|{RJdT0l9K*snOF<x5CJdv5qntZ
z%iTr<AmmA;kvs2a3Ak9ENJ>h&xSmnky2^bM#3bbpdwlc()6&Kr0+eN3*ikUK1N>{-
zMw_>t<VTJ7Egmpu*(6qA!V<Vtd2m}6E9rlJyu(mN+@$bwLq$VGhupo+Ca!tW5|y5f
zo#ia~3Yg}vN}nys#0VSPPK>86Z+T{BhQBJ?AL=2!K5e-csg?gHd*k^5v(Xv+_S3e6
zuVxmvRjRa@;_XqeeFDIoc8om)WqkOE#(#bBC#Lz9T;4;p{EPXuG!yC5B7Q8=t84NJ
zZ$5uwy2}?S7Wb;GcKN{>qy+V;8{A-P!eB}U^WWq+$i$b@sDIMmzEy>MVo%0ttUhh~
zhROYgNmxYrdhi8$+h&m6Y?NGE6F=tLhsV!c`{~E-hhR5H0qWkUjsMRAP=+XtuyQD7
zvzeS~=an>V1VRYejJFR8&_s3cX=qEQeF2d$(YTxH*y+z&j|e)v@rNDAtDL&LY4daw
zD;L*?D|R1joLO?B+#Hzae;&}k=6TzUq)b2(#zbT0!ZPDkN38$18UDobx-~#HjQMQR
zQ7_fE2+!VkO9On7ZA3~7Lm8sMs#Ib&HD2n2oqolWJ)+$i+|0gmw&m{Tt-R^}2R>cY
z=aYus-hg>%&u6S>r_jA_m;d&P>IIOCWYc7_(rgbnh{1vYH5U?K0cq@0C2&Hond5uP
z2aklb_4;%sB_4_^MHgYZj05kv>O7dY5fL74yw>i%*lbS><Y!0dId-|vc56~NeQC`J
zIcut#?OR!Gf&Pq_Fl=5mG@#YE=5klQku%wb9L@(GyKa~~ZWOAIo&I%mmt{PkQ2@H{
z8Zx`}AZ8!h(!?nX57@U29?B?u%E}llNaxa`giqkOp^9fu+}!Zs$fBX1suJ221x818
z89arqmA8+=SFaxUBoQwFoi-}!j$QtDqQID@Ks6)a8AStXWtebRJ~qLi`5<tNU#l;f
z*A^0EyJrEXmx|d#k;Y@i$?0*%3mD9vU>_&`3ORR0qwhV8rurtU>Wlf6D%)0b%&d36
zI8u$}@iD6C#;!^A+|mb8qZJE@6DRu(30;^usD0~hARTOdsVJ;$ytkQ;M)gpaY1!~V
zgFr;t3x4@Cz4M*`^J5P0{a#%bsd**dI0#|??V4dM^W@+dVT9Q2zLFl%Fk|7{*A7O%
z?9_G`ZNUaH_i;4sg1({2+a9T4xc(Pc^OaY}AsQ?uLt!*#%{w5vZr5*cbpJspYU<t{
z+_!?V|9!j{MMhn4lVOwcbg(MX2U_EcWRTS9c029YumbPr6O8(vXgL^4L9~~;9Z%?)
zM;HE|C6+zl{ROZoc8j$z<bLPD%N-WGfDt;c&ozC(Wl}B$_|HpaPjN9qSJ!(CX(i8U
zU%)@x&u*m|Ln@Ah5F4A^9(W6^CXnoO|D$3|?j~tdm%<U2T~rMNR5})xd^MWQ!zk{=
z43x$xzrvS@8-{84W);U&{Q(Vd%mu|T)E2DlboHxh<T3pnrNg!_$x!6Un?dSBMI6z}
z@bis4&C2aoB5cENF=&PXu3mU+tav{hF6hyyz4UD(V#EO7wLzbFLxW4rGgy`DqK+Lk
zGSHszsVxqk6GsN7Dg!lnK8eg-npGpH9{#1AcRL-Fi28W3GCS5}QEs{S@dN3(X669l
z;<Kk2Xs<%XrnWV17wGXND(Y~=8VVRNJB$?8O=nN-{V~f(D~qbK>1<1C_i_82Tts||
zGKJx-t`m63{*`}(`1(4=`L^ZKL#bYNf0#YJgb6RL@y6=BMO@`xC5w=uuDaz4J~XG@
z_3WqA=Y=?fjLaWOstdLQc;QXgmCg|-Tm0&R?YJO9=VLMu1k%=~`!V;}8<s*_PM_p-
z_pFBO_S3!NmK}|IHa~tGXQP^!wF9?%7?#!qJ%FE`!_T??27KxVf}n4Xflqp0yZs2W
zv$OXP4xFEMn1zId?sr%MqkuxwWIT&czQ)?FNHy*L-pi_K+_H^<fpd$Vne;<8$$v}e
zH=J*BYp>T4gvK?p5_U-cLFCg^2UJ8fIJjfCZ9|a&-p|_^s+scVTEo9>8X-FYMbMdJ
zxVN@`$lyrx-@?v<75#qi$Tvd2CjxXKcl%>fvV~?(>E>7kP(WwYGZzWUj0*PF(dbi0
zEQ$9HMP~9)0?ti`83^^Z(L=>aA-2DY1$jE&Z#u-86TcPhbu<sM2pB8<Fi#oAGGd7k
zZZW#vtkf2D_WgEf`#q%{<84Fgy`AG(@x_6S01Z3<<2$wfX}G#c1>_&TN67+sfNB@3
z`ZAkGIwqLWzu_Mu;vZ&BMl!1-3l0pd_nbu2@jd34*oWXn9_H-bFV#hq_G?KW<)vY1
z*&KWuldBpUSHPQNejDb-=svQ~ABmJEv;yJZU;@tQ)=#8}78tzFd2t97WyG4&GuF=>
zdpYJNqT$GThW%knyD{iH4lfQcjGhR0s;5lN%|3}_4Y2>WPWW@=7eGSfbJLk&;EVgJ
z_g=OQxmdz~6d15pz5jXfK4&y@Gio&<bY;IIl4Zg!oI?KI@51aP3`AfK0XeYIp-9;s
z4GKAegrLdU*)XWlW;eByQvV$NIl5e0c}F}^`8NNrPu?=3gV)%0LY6`OG9VVl*=XP0
zW@>rv!aC>kGq|V@3s8k?BYihfSeyX{x)0&jKZQshZKD%JZbghfa^deTkDkjM!o6Y6
z3HETW>&Zj(>f8<faf$F@x|v3kqpfT_@5$fOOF3$8yHRsBG8t_c`VtDPXfuH_^g=-A
zQFYv$T$I`zPS*J7&{5amlCKzynW09Ua*-$vR08AZRg#iP-({}{6q!+Nmu@UPvJbu_
z8ssZ{3kCicVFuLxC|RNQm_{>9lteb>u}BTv4s2Zfyd@CgZQ3_z7YCx&mnQkn00~JC
zGY)j0I&=bb3wbwWMSYJmBgd->X%t&Y)51%kn{`6C+q7j31m3Sc7HH>v2BO@}lynJd
zllnB>-N?LZRf7jMq=y2HlhUFtVT27Fu|I#BNq7oCYnGJkGx0l6e>-*4exPNFHCCb2
z{5y8oKaH{m$9V{s9)1AHQb*Z|{cLqQElLii<`|GStwo8xiq<Q&bt>jVWnQap_yqjU
zYR*D5Ezd%3>>5HsUTKD}qK1uATL}u{+PPPaj@2154D@xq*_bnZYH^4mOe7?77IanL
z=iDfE=2O@maWU}g#L9Y?v&P8mx0z`32dCOuZ*vWsWWuWkpDEkayrYi6Y16;)d44}&
z00T2pko$7YDxRNo9vE_cb@DTLJM&H!N_cW|0+QB|pujHs4tOvDX+O3bj;pOICu>YD
z7yrH-uXhn+37&%r3KrooXmJ%zZ+sT8)Sp1Db2eSKXFxAODn_z!bR6S*_y~`=j^^0P
zXI8f%cIAcYFM2=lDSI`GBz3ThRPJ1xTu#SLZr+-j=sSqj5`CR3so-}MdiUAnB2nw8
z=%yl<2)mHyM?QQ?^OY|nbmc%AkGJ2;!8h+C*lXTQc}D&)Hh$z+za@X)m&FZ&d)w6^
z?D~rm6cr^Ic2F5_ZnSj&Ng_hG_9P3&UI>n02{9XwXV=5n!D6D;JFvfXLCNkXP(i(r
z{lILZlCE!mt~3>6jzl3{q{_SY&}ur@P||FM|8@DGZ;jFSJ%Va#UEtj3lueFTGk-9h
ziQmuKfp~xpT_Jvom_L@tPg}G)alKQc%L1je++92Q#GjzdXq!!-5UB%Id}^xV>*=Ra
zNKB)_pD^9`HG@o5pW#B^P@jCC8e!fV3J5bOS+mkW+r>8&EMvp|lutD_H|kAfZ^J)=
z`Zt5BgoutA(ffh@*4<u<!DbwDnk6nD4>t9Ch8U5L-s9!BKv(4S&c_Mi#tjHN14b0s
z{_P%+0*0<Ht1d<5;r@icUrxb4srA!rrDDrE%aB2CiFx#%1Ab<o4aSnBrnei?1BPjl
zKZ9%+8RU+oaN*Q?>mu!%-$L)rh066pjp}40IkZ`s=OS0P*L$;n#+asxY;`qX@u^l1
z{*hqGj{OfB{0YqRDA?HEnVWBeh6{RX@BzWTlJEn#zP`TJ=f<Xb0TE$p#b~DA^BeO|
z3=mOA3dG8{iW;aTrRCjDJ#3Ds;=wY`<RK8r)L?OQA3xRi(V872blI8%oorp&9Bst;
zy@!9<aj8F+S(%WhCQ|QEwlxDzB^eAtAQFOI99XUAh}>0mj#s*NJl{{@|5j6s8gH%g
z=0Mge7JrFnzI&&4O<EzojiC)*|FimnY@?eJTFqY-6Fbw6<oB(CUH)gza>(?6(QLCd
zWhj2s7Ven;9JKqTvm*xQ(>y|9z8YE;qI481HQn&vtmAqmuW08Q@=&rN8O}`;QoSZq
zgLaj<y9v9Um9(XQFRm30TEK2WLV7zn;F7stn35K3?8Lj+{9Q49aEvDugYQ1jaX4>e
z#aRHW#EM;k$E*4d_WSWG|92j3r_q>CEqC?-G*8P#b!y+E@r4HZe`MuVP%C5;y`M)a
zY|;yjHzHhewlEv`Nht%_*`?_8pb}$?_i7UV-a{VsM|^Zeodp$zlS(|8W$a8sO^Zji
z8nLiTORUvkV39Xz?SAZa6-p|?TyG&U#zYI&!N&hsIQr*S`0roV*s`7tTsI<U&O<h@
zXMflWjgLAfLa$wl(c~~7IQkPEB_thP3t4bA2TA{^qY@g>Ay3>JRE)k3AOkz&)2KJY
z$37l0PR4(=_C|$=tpOTJwcb}2f=|Z{prGpNPan{atNho`0Xz7DrY5}qJO}UcjF0fk
zs;b=E6g+u^pdle4i0Z$8G%xU9{=rpheO?2s`a^-CKLs~8K7Y2;F-e@ZIdBD)E}eZ>
zxbxQFxpAQ}clQP{+>1LFqJWoP^o8$3G)Id@j?}DjZ)&P4l&i4sX_~W{Yu;F1zL<M#
znAfd}##et25T!!D*TcZrVu)a~b@zb4P873MXGIrJgNZ?EESkF{U!}xM-wE%YjodrE
zUM;#pr-F~DO6%}o>-~<S;UX9aW^AaKDOC{L%hvOJMPs3?42+K<p_Qd8ND@K$J?{3&
z@0+71C7^_D1f1*naY<nW+4Xya<o7z7_#eC@>&r*t_sqSkS}^DaK!)UWaK8nARabYn
zGcf|;S|<}3t*(OSnZ-GlFklI|mbfUfPJ-{qKKn7uo{H=gG{ye7*WHWN_CGzk6ndA{
zWGH#OT94!JPUYAnsc8BY^xzJi>^z_ohE_)jIV*6jjs@?WRffqHQmsDlFRO>skQMr~
z%5zNGFhc|~vnTk66jL^9N3r0_zj>`bFjC*@OAo|+UKbfRN=}5mp5onK$W}zji4fwM
zW8xrVSQc0c4_g@ky{&J1uLXUA{Q2x`tZp(RoTL^sEM647=71zGi6FY2@?$b3iz;;R
z=2;)q+78!em;AZE0@LbEHM8(#&v#qj+1VMSE&XL1&ET!#as95_^X6@##_;T<!_yiF
zY5)aAkOzD~!~bY?doe`)0~8OD-`cypx~d7Vw<(wRXJlr!*L+x4mytmM<!RLzj0tZ{
zb*J*DPTb0#Z>*u*J;*{+xYn2Ago`q-U~(0>$VNlHzF{y@8?(2P)h?^Ld_12OB>cPF
zBqLADhWL{k`F>37MEu520pa`y21Kd@RFRaEpV9=Wunl7A`fz?6S+5$5R62BpZ1vhc
zegEd10Ku|=y)al8UF=Beej-Vj>mF32{2kV`oQ<U|Rv0><!5qQaB9-B(`>Ogur%_+;
zJkW^te8`PLaciSxeXoiqcdxf#9Jcd5{pj<c<w)>e8tkB4<3?%Df&dP6eN)ZJ`x+|h
zzW|fM>J4JGEqmfxQwus%z$krbS}z2#J~#+BZZLpOKNQ+mEqR?}0JiEsXR*9aHH|9A
zC%@<dNqZ6!`G1c#8v<S)HvwHq9}v740VINffU@edOZ{FniK8>#q=%=kswxf@6_rt)
zw$te<pWE4b*IdTQN(&rlAeML`zbBPHE(2f>{3khAFC<XnwLsgA99AWu`%`u352m1J
z_B~s55C@VRc6x|j*t?%X>cYpE50S`yMPpinLCkd9Z~8(Dc%U<Tgh$Jve)^`d{Txxr
z0<hjKo_@tAuFUObgtlpu(UllPqgQFD{T|0rrc4|{3A)wB)|^g|T+{M`-ihdvIR!Ng
zuRdb37MzMjo>azhB(mfI%6dml)_f`DY;cMOoGq5Ep6@-0(`5+wp>aPMl*Kp~@G-md
zUJwiaZ3#BZy-uJeLBo;`Q|V5sYK#cBQiN)6CFd`(am2ru(c2(f^!wW*biekNhqgem
z#!i9_UU}bMC5`M=sZb2**rYpUI1VIAb8Sy(wKL>NPd_)Qw0~G}tC82c%I5LesjQw!
z|A<dJE>42vol4o&W4Z*;!L<1MGq^N2fCva-shR5X^?csM3WOvxsKv*}L+gZ;(mv@g
zU$?UKJlBt|0GT5H?d<=KG@#jTP_csdrFXT>wa)9p<iChTuj9JGj5Px-tq36e0R}{?
zkTLteHG#9#YY}s{)CzdQ1qG=)19n$NJHkd|#(DKJzQf1rMyJR|V5NDnxG4(qUdixd
zx2bx}a&W&`m4;mHK@OG1>y0MLHQ{>-A%pTZKkE@u6m&bmT^|KI@c9d&sF{EOPm}rm
z`g{`PTOkg_sz|h4;*5mAPs4_~Ay~<y<t^Zw2i&8Z!`qAU{i3hr6U0e$T%?f}Bx%-B
z?dK>P45k%2kwxd)nOneFUw^;8_U2D5EbCG8?41H%sh>VygW0xZtiF5lC(`brtUL)~
zpR6;E)_BU#+=1TNqF^A9AxwZ<80<}LUJ7#s$~WCLYTJFXB5*o>ZrZSxhi9#0aT-PD
zGlUBo!e78g{1UL|=)d#e{I6irSU2t6I~ZYU5MbH#+op^AV?{{`B9_o~{BDhwKWlwA
z+X3{`!+*Vnk2>}6K*+E-une_8l!k<aL>^$6_Mb0}9XGLLEKxX+&w=^YG|Tq2Y__d?
z#P<iOakA*!7#|KrWrH-ARIO&8>ONx?tSpPIVPM{n=iu~|M0&8p>k0)j0x%ea>EeeW
zKTDlgRB#*}Mm!~9`W#UrTW*#%^h-6bmr)~}PY3eTpgKvHEdj@?%#E9TVyTW{99M-g
zXw7mP_MVPMmblnxO6W`TvZE_IwPeL78AuDA@7Yg!!F{3J$p+z;2z7*xW%EU81g9BR
z6yES8@jXA>$_ez}x16u6d1&Xe%#f!Q-Ih+-C=E7WBQBSAZV>YCDtbk*#F5B?h#Kaz
zaobh)`xJL}Q>gl{zfI>JIC=Mp@A0<NWTw=0pHr?vb3iL;ElL1YSn9fcr)$705IM8e
zgomrYVHiMn^rO`P*BX&5JAgSsRATlWeR&%7>AQFD#&AGB4^%M9>gr+to!7vb?sd23
zsb9U)6DSgJ@!gQOi&;Q``0c6Z?K?ezIEujXg3ZO06#3@EUhW_;7vO{Eg=gMC@@o};
zdW6vrqI0}lyHK{OUkb(;Ec$n8hX4gl(x`-#&LARo(@aiXLC#VSo-!6rJ@u>B-H@G3
zOsFD2NDL9ESqw?=rbQ@2ib+BKYUQXO*U<dh2fevm(!jsg-SF;x9GTAhKcA7h+K`+-
z_d(vqI5jL040KPsi)BqXl})YJ<<_h~)kjFH!n%kdxLh~A9!C8ECr5+ViDa9)DO#gP
z0Xj`bT!p@}a(zy=<d+9zP->RrRqKQCItYe%XJ8NMaKCO|9(pYwcrqPFBlMAfn2HVz
z_S(@M*#tFb!E=)$p^g-DnU=!K#Eg-U_|yzya6^CfU-Fy`FAVzDEbD0m({z|=l{HL$
zSulf65sL1`CA6T(G6rEtRloOZq#yW@%wYem{jm-?yiu>g0^`4UtxD}WSZizRU}2Ji
z+J85rLqHqe`{~n*#r|Zz*VPzLo!Kys-BLXQ9v<HQOevN_>n5Unt}v-&;n7JN>4`1F
z^FbBo7-qqUIx6PQ9K^62gvyi=DBL42JWmP}z^exPE4ZMgeZ~&GCfHV2zdDVZI2;d0
zbU%xp*+{HOr1>+9PswNm6R_8F{#5@9>`wzn-dA|zWI0-o1D?<|myKwmqj`8k8#CVu
zAu+u<3y*S*gkCOj#|GN|$~dqJj+t!Q9xh9y)1_VYipvJq2k(7?X)npU_Q-Yk_v(uV
zhBr?BW#b?>aUFQ(e!%!U3{!$NyB+`rD9lQaxy#Gs1bXKAGXF%=pl#YJPFtJCz^eXF
z4XujFLq|?7A6Y9ikG*XDh_n{q0$=KHoc8)qAGCdUvMQyJ`&YP0=J219f{HuLOM-^c
zQXa^UL7MF{sySWA&mM&5*MdvE@~@?$x+3}uO408}%6CYF1(@%38FAh{d|&VcVRQY%
z44SlX1&O>VKSkZt5IFHN>dzR#<vhXjdJey|Cu>8~AEqNXu#YjJ5pOUy%**pwBgvW{
ztHV|>L3j`Xq#`WaQL_GMDuNl$xEuesfIy~7P|gFgGM7Y;OTAf1Fh{9eaGI73Z1Jyw
z4z5g4MVt;397izSs^A;!@NT?5UfL1OP;T~awkVprpt~k?_{;LQLLv-^qoZd|*i1n$
zBeR4JGY;oT4o-l^n*v@k+=O4H|H}{=j&^g$2h=PwEJ5zfl<{uB0U?whLjZIO*c&d>
z3n1$f0o!XHryyv<f1Uq+L-i~8{Q9~1P3U3eEOl5>u_~ui#^Ynl;DIW%t$2C<a4lQn
z5%_N3aS*|$Q9daTM%Hq=m4R>Uryz<;QiF_l32L-(PIvcAZ|6>Ig7(;<Z?mfeNP<Xj
zFnI-%%a!oaQ+YKm&bmm%_|0CW*R&p7c}^$B(C~>O&${`5yL_XQP2131p(Q<v2~$g-
z61a&^@e(pKYqS=)tkpbDqfY*V#g+!)J?ZG|kz90gucJ_rCXGhIR@0a#uGqN;SJ!2N
z${0?$nuDy5C?w&o0GrlcYD3aqC2ewddI4c^OvqK5={X3qGP+e>lIev8LXU;10qZfq
zmBT_<@QvZTpK=1=Q_37B7hX9gSDP+eB#(Tpx-=}#j+v4}f=YH0Z82>FJ4tmtU{o__
zb%J|pQo?rWe9c-P223FWdO>s|pl-pT&20$@W&{;F9V%3AJ{@+>qm4{Z$M=S^DF5J&
zfycX6dcY;<>(eqbU68;ayo1s5>L#QcHgCV+YK6F>Y48?N((LJK2=-{v8+s+Uz)V5J
z5;>6!!%`LhM!t&IYFZf>@$}e-f)SipqG$k8_2KVfW<zZCq&kniz$tPrQ?<5ZAF@Li
zqUy6NMMz*udeldebntw6<Z!J8gQNuzAaF@R8z1f0>nrA{G<?kH28RZ*EMzm&<+!F1
zS@>uL^}ykYhltQvjB(#d9u0#NzoAa1j>}+vIjctNnU1H)8DA?6(({dM$UN?tUTE<j
zSXZebCmrUfWJ$)N7kS0h!Q!IO{!gK~N*zB%%<5py98&bnt|TAt79%4=2ysCY+w<-(
zLEYKr^yk4;P*rAZJj&cmZtM@sYg>pghen5$kZ{J6BU_@PXeQ;CZR#-qqJgQYO0j86
zRR^gd$+8wv3oQ^wI351V<M+b9AESd%a|GqD`MGEZY_hZNhNR-FB8nf$ns;5Xi2dT>
zQ1KljV`P@sJWgBKE$64;#@Ww>`Ab6eM)T4R9zKrK(iEtPjAI>oTHA(U7Is8mYh&58
z`mKU8Ynxx3tx4S>j!N}tv&4X7Bs3j#mBK&Yk^bbGFS`uZcdIX~DM0i(s$k;4=Y0K3
ze4wwC<KMjCe%_M+yF|z;_<c`NLLzoK1w_xp)bR)u{u52W2$VC93dl6`@WqUZ^rN6w
z@qh)XTK_3$agkC|x~ecVy?b@ue@BiW7(ne*i@cbQm8>r7n|JyPa=-o<q56#<53*@8
zc4eJew-~Io0P4}9*Kp0J=hJ+8#uY6$UimkR;D7YQ29+U|*JM1;X1IJ4d^^uMR);=>
zvzPe=Ccq07*1Y$WKV%X}qi-ekP_2bBH-mDMtlj+sPXrDg&08I`%sq#K%q)A~N`UuO
zqFc@aIAY&hxwf7!XbwrJ##Yc%-yOQ_i~LSSYM~ll)WU+cQj;Ocl}qj(nBfJF(N5(7
z`BW1J8Evk6KAq%LsF<Div`T<Kc0|KGqfICd3EbX$ka+vBl$2%Du+!h>@)p?9%OGD`
zud|=e*3w!h_CZuJ_B9g1s+Epi^Tu-$n~{6u>*nmT#Dh`(aQ@&~eCCKzUp1e|<LeKT
z2p=}2bY<FcjcwQ${zg^N1vh+J7Hw89yxFS;snDiJiLd`3dvE>JX4JKf1}|PHt}R}q
zSaEG}DN?k!6nA%Tf#O=+-Q6u%ae_MpiUfBEo|8WBdEW1r^B0`?_6#!_W-@o~z1LoH
zt!u5Nrc&*YvS*wH`6em3H4dRnA!gJ00R;z8pY_1HYp5gOvKW}Zf0Qg%%JSOmcKssG
z-#k%|qh+Ytw}wJqT#<av*Rn>4J4}|mta*wYcwq(%3lIM+qgUu~%Ev+|iy54cRdt$!
z9CdW^aJwYC8O9xS`o4PcShV2zqS9SJ=28~Lo}xTJ38Bs78$@N-I-3s<PoFS1K2lPW
zVA6hanzwc?7l6s1%d*cDH`;y+_k=&$<}rQ1heuXcNV`g~s9WsR$k>?{(N7R>8X62`
zB(9q(sJi8JYp?lMDKDiGFf|$&q82!0j?gAjcw=N+NXKPQiqPpP#4*(mk_W9Zi|a^`
z>19%Xe4z+)MY1QWKJy=GxU<YU^4gG;bR`S}yupiKG@~0uS+=n$GA!YaKwJ~T+dnX8
zB1f65d<X3AyIl_}0eEweuVCLDd;#ia=(M16kbQskU{`>kCe%F?lm?8MU0FQBzI%<Q
z>Y2VqeJ8A}bLy3lh%!!r^|Xtb+bd;a@Dmvp7HUL3aNF!Sr&|Q(#!j7*nZ%7?=joZS
zjLQnbbe4FeVA3`#8tqIFWci3)cA50_j^<D{74Vkkl2!vnDMc>eLa;)e8T)`NTcXd|
z2LR(H_6#6*Jodd>^)5v4h_drJbG_^eO2g;4``lm4W6@3@_A=;@GYM?K3``{6Icwfv
zzG4n_MZtH~%oE`5<~zTzUuN4H0glJ;MWu&_f0sS!Z4fv9iM4-{^tSi(kr5%l@aV-R
z(WG#R<~{1&uCUQG@20rw)smHY0$rm^z3+n^a#f}J+a5*mL-qmCrDgQq*FWvpQ!z5-
zDuiD%>oIENEu>fh>S<6R9`P+aT$TetTylpIWfz!19>l#nAy&f#ym^Le{N}EdRw`aJ
z6$5O33jA?fhZ+wL`1^=uAgRyE5;m11>L&zVdxAw6yg*r|W{=l1UHGn3ra~8)YX9vE
z@Kp><k=XM-f)4S)_4dV2COzm(5#gov>kF0suUJH?)>9oDgnSjVx{=4&eA(Jc5h2TP
z!C(mpZS8d+DIw2<7`0Ub5qGR2+Dq5SeN?JwR?wMJ{HEC3pciToBJ)vTKFi3D$zPY#
z@nA!STNJp*)ZA=SpSg-S^QYc0)He>cCQQNas+7PUL9<9S!30n@>SZEQdX2|3hz!7s
zr>0;>Fwnu#&O}z#Ak+>}$43&&q9u0gcC3>wk8a69C9Ae2<s%K+S`X57q31Ybd0%(=
zDNjoEdv^ao_e}ZT4vLXO5LT}XBMV`w;FQ=9$swv4qVn3SC(|xuY2Cx*4_B`~TQSk9
zDh#Fs(su-Q2HV(e(M$kSvw=_XTWy#zrCsg>l3V>ElZH>^ws08dJK)rz2laRoO-T=M
z$aV?+*qG56?LxGq9EFPC@o+sQArGE(&TcqpeRM-bas(j@@Z|$JNrQ$g5F8v(d2=lN
ze3H;x{XTO1gycy|PW(GjQBx1VV6ymna%B8mYs@#}IU^K<ZzsZcV0HgA+lg+Ed}!>J
z`up=s`P}xrd&r5qk(Ov}&+EI5#+Z)hj5P&#GtSof==NH~yDR3#a%tN+TI{=C#b*&W
zq52n6U}Ot|*Uy;`6~zd$?K4!B0Vy%}?Q&bDAks86(>lI@4}97|(L;eED8N|zxbMVx
zzc5p<wLAG9ta%z(bd^VahIbI~2a229Jo1mjm)uA1Pw=!tv7uveeOXt@&PrLY+1S$V
zL{ALk%{NyQgUG{1e8zu{)fcb|ebI^gBZd0#>Q{I$?mr9dcB8r3(DilRV3uw(Gyc{1
z1j3)218v4p^4Jk~8)brFqMdu%u35uQRgElUTU}IGkj-Kek<dAmXNDRsDp0D7K4mDu
zX4_BTVMskoZXv!3*5y6AU*Qezs`tJH*Qk_@eXKY&dOS3<i%DdZ+yh%6(o}PC%yGto
z*b~|n%ar8jwS*NG`mO|U{*B7<IWWe}{>09l_B^YszQE)JZ}~zyz)W-T{1jN3CvJjO
z%c^Wn%agU9NXFQrOwME%G#t>*aMk`ZT$Y6a5I$T1muNB8l*pIglUj&1HE1!}Xr~zj
z?ZnI)@EH+9fu{80RZ`v+IHwp<XS9T$*-qGJ+<TY{%lD@Q+5i{7K~RA+?A3(9533{8
zSzBF0I4C$8*)<w&;<j{CY(EifmGdWN$oBU#;je>-(m>xK+=7L)bK`L@N{XW_3CuG&
zS=}G4D4NhL#?=+ztzR<p7w;mJjX~f__zd;!rbhvuv))^l^)fH=4NbOmS9OwkdSE*4
zv!0({$GG5l<zd_%@;kPx?{yUZL7=TKiVOPnnkIj9y9pT>8}MXGf>i}Rsno2f=8+_5
zR|nKoJ5fD3Sof>uD&(t={Yd{Z6y(9Ha5i7;o{jqZ6oD#z#any7NXIWW&!bC-Iq>8|
zGq+Hp789|Xsy~S|g(VIe-1jva?A{alm`__6VPhZTfJXNQ?`MNIOKr`RDOzGhbB+mO
z;Hx;3xKTkrytzo8?k|A3kFxQ@T0`_pt4tzPt!Bs^c@B4Uh?!nCiC3EeYCf2OxGhK?
z(N`c&l!$V~oKuwFr^zk{(nAV`^cqkCQl<6x?;M9@B_*r8*cvY`$Jl+$u<`CwB$CVX
z#%{O>Dn*ioy->21W60UoX`kgK-cNPjBcK04U2HL|j6clH0io2Z)%b}Deg<z#`dnDd
z7|o}s)CvX8j5%LYP1@pYIvC13mKO;|+FooJ()E8c4QY+F{M4rB_}k$3)->Fqix&wb
z$pQ(`&^WMkv7Wux(XYHQK-58T%tz)|p}UgLrw!c5DioYwR_b!xb_76}=`Axzuw9j~
z&NgjZRSDjX7UpjhKez0SL7y+JH=#LWVAnBP<}OyqpP@-4s3CDSs3F)H-PWe=A#Pkz
z-c7VafYrBASEsA}t2RPsE0(8s_HhCO{5XLt()l=})+r~qhExeJ9nlG<@T-Y9i+TX=
zZo6u{ZGy*no6A!N9{L(+on6aUf{~P-E(Mv_%v%2beu6g2Q6H<Jznk^K-*0)|EcAt~
zW}(3{n3+-{SE$FYL7WGwvw)`~7SheNnS7Rci#HE<&$Z#<D}&@`V#GE<i;D;{5g7q5
zV<Gra-kX9(mMynpN*wh|L7=NZp0IS#vs^UXdq7nro#Ncs?*CNd8n(wP9?xK9xQ~gW
zfx#fpKGvguvbXH17-PI=1g4OAdosS*y^uGmvO;Qn=K|PSThc%8Q8i;b-_YF5v+5RO
zprM&5)<+MVnfdBKiH3KtnnifGjX~--KQl<Sxn2eQt1o`$dJy8}C-(NMgp`3?a`!e<
z*<6cg9B)pXwKp5aXwVztz-}$Xd$k877N||dkK-JK@7_+S8in6JJdIT}RlzNb(#3|)
zlaabAQl)66-N^>x^32*|q+Hsqb*((Tf$`6)RU5LvfJ(GOwuE6?8jK&4=$F|hvm8hO
zz|3E|xMeE1R=~_@_Xl;rN0@cudwxgd;L;WKD5hBQB;eO)sHw5K5LY7NsH?$3#U|vz
z1}eaD@cM<8WGw}}{ydkN@+mnPO$@F@Xp^-PKyMN30h*^Q*>RAJRCY&NggU^$PPNlM
zUCsuO8x^K+6&FvWaecMrf;b+GG#ErD-abl-iq{KHuj8q^+NQnEd&*EN^q8mw8700`
zr|q?It(RT}e?5B;7gWu6c)qw=42C!N2ig?`Nv*CyKgYzJPdM&ebOnJf8>=QJaA4g~
zAIx+p>&<s9e@A)%(HXq`xWnPP9dOeg|HqLm1+&WP+W_UfYqD7*6pFZ8M%Jk!$JfRQ
z&?GxD$W+wP`pb~7PXpvf3&!&LJcQQ@dOO&=*OJpgCp7T;42dcWPmt|mV`&p^_U5X6
zjYjM@n^pY<=$t=|9X;nN3Q4;=08jg5TOcu{zu<<sTR=(N@W!VKh2MGd{`I;_F`uzH
z)->M*YDK1mlvJy~0oAONNa1IUe6qwUxQDE5Ottqe|F=@gq0KzfFz;lFCcKUgWq+?&
zuuC_hTa&MQrTQ4N>Gy3_k8yaT`h#P~b&kV(5d&0nv%9x7f;OecMx~`rJs(!ESch+3
zw$U;$1z8%dtD!}3MRpqA+yqy2lSFTMG?{BnXoKL8<FrUW_lV1NXJa?NuhPpO0x55Q
z5CvA$A!+8Pn{?D=%EsNDrd%@<ERJGE4h#N@J*I26;bbH4tT0N~GN(&}d8(q6tQIMk
z;z`6OyCym)p;V}jU`G!)-~-$9Ot7!T+6BIxy-l697>7$-=YlH%b6~;XOnOnJ&Yt+S
zEYhFbTf^APbX22ze<R(*#m7PtBSm&A(A?#X0fM9Mq?u&8x~%}^?zQNnD=Jm?-s%g?
z8x(YXii(4>uC@a$w(u)0ZGz8&MbcFs2BW@pk&uQW?qS(H3+6$+OO^fw50_aSSO~A<
zg}j-)(yeO|9Q;t8KzfTojc{{(R!Q`}P8{~ZPUyGzluuA9!RiwjgYLBe7x~>`s`|j3
zuGwB+t^L<~%ggE86L^#})DS3E<(h!p`c;$Y&Aa57EJ5+6RyhlLt!XK%qom|Rud13X
zrs)kOBmfQ(4-8to4g8x03`q<MN_(L0Y5&w5a3<hVL|uzWCymfwA9#R-#C@%Va(?k)
zbCyWNN<+infKi?fczH7N5CH9Vy`?$7EMJ%(O>X)c-+5Cwp0mIHo?}R62RjQt@C=kh
z^|y(Le;qXiAM<HH^cGjoVfhN2_=a8!XGp_}htBfxbYU|SWi~0h4<`d@N%xiF=~>4%
zIs^Q2!*m&ibTU@lF|->cR;ZLnUn!o{RIQuo66hx_sIh^1Kg=Um;K<_N!9zNOg9!&d
z(m<5{a!$KAyw$4-rph={gRD2-Lp7Ban$mEZh`a(ew3tqW3&zFICV_JZCqe<!KK5ja
zXAK)%#l*GWQ$p7ZFkYJ_7;oq0-r=7QB`Zxz9DX++U3hc(u<lB1>l~F?VmbH@C@I++
zf)2>XDHidADjNVInDw-pujNy++cIK0x?}&?z-nS~t0I5AgHNw4@$%Dx<S0iY?cF<8
zG_&X@1{t!)xhv@^4NEZs2Bc-(uGc8g`&^ON-3(r(=%Z%eYcpRH#b6u*2FGr)7d@TJ
zh&-+6&+l0M<N}eTfHp^TZx__RbTy7^w;7fX?m-Tz-)>MH;k4kq@D%};{4v5K5xY4z
z9Kv=rzl22o(0c>NA3vQRp-@5`68mQ1iQk5nQSia!?4p50f=S?o-%e8#e10#WRj&kQ
zvo`Po2e#ja{7ektk0YGqPF*QjK!k&=jvMpP3i+RpXsh2X1`4=!(6&}99!8Izr95h&
zZuPUbe3)5cp!Sl|{j9BrqgVo@no5Ln-Aq~nedGl==@S>lo_f9cDy5QOS)|g-?=^vf
zr>;(BCb(5-5dbI!Ruyz!dGFaY5kcT6Mmc$4k1v8SH%upJD9%1xHyXkDl+YN%7jvS1
z3=X9TltjcPKnt$lf~Tb<cUTkwuiZFKPJa0v20vC+L6wq5-L+QF2~k>o^!9H;C}uIW
z+htZY_oqetW=Bi9OlWrz&zq-bXxH-)o#KPAEz$bUPcQwn_&g0(WrOIHu_E-=P~<zW
zJ^<PUD7cvJ25aP4kS!|R%i^j^;|qI-I@WRPH8s+Boa%_~_HJIR+SDxKDX~vKqTVe~
zCCCp~8&K4*YmPy^>_<>!bFcd;fgUAyPjTVC%=AR94m6Hu$X+KrOyY`#9)^s6{g$|n
z=1|I7aWuU_C!fmz1tOwrrwn)Cz9OB?qFo?6#0yGsdL<0)g-N>+ZHi^kqr=%s&T<^u
z>&G~Hb%qa1EoQg<Ke~Yr1uF~YbckJj-xqaeU2>Tku=u+5!0p(kvtPcrxC9=AJoy#c
zHJJ23yxs?9hWR!sxT_)pqdwIu)b&c}@)6Fwd{kj$WH}1-;1%kjnm9r5H4SbQr1_ta
z{t(Wp1qiuL0xZin`q6WiUkMBAmp?5~Bxv%vhau;LxhBf@s2oNw4}Xsnt?1F;d$EX%
zf3?PavC>0Xew^N_&LSb#poRnv(Entq&Ui1M1P6R8u@vXZFaFvuiyn|tl)iO7o4nL>
z<(2td6WIi;p8XoMnk0dud7OCG#MY(DN14)${PODe3Cr9@tme<_bL{VkB^Uh7nN!{V
zt8zg56n(WddF4DM)oXe#_Jv^_TEMVkLbF%(wTIffb%X3lKg25Ch(_~S`7@<fl>$U<
z4_Tkh_J7_Bh-mX-c|$0MP0&5|^nDz#;hX6yB_XhfuMPJF$$6ti7?%VbdX>_hRWt}O
z$34@Hp!uW`a0<9a@qmc;Hv~H9!w0y<5RHLE`sg~|J5UhaS$l|FXNDaMW3LIxVMycW
z7e()|RP0*76=Bpc{KBo_Z3yc4eOLgn`|b($*0C76H$9nDKSBv;@*Vq;0jg-+&2XF}
zn}yL%=eD~gE<bOm1+grcdR?tNm|QtdLLUwd!n1@K<VBQU6->QH29@;+eZdvDNMo1j
z(9tLk#Kjx5y;!X1x1HtkHPCcabkWeUUzY>j^%^fBCl4=J>V!+4SgkIVAe@OROK!(!
zz)1tI@NydNwPK3sI;xaAYG*y_QWb67&iE+OiyVMJ|N6w+u+%PYCz49&iazb3L4eX!
z32nOUtCydsd-px$(H*s$_s_j;loI-kkFLVowcOTM25qm8?OOjpQAJlV#c!UniHOm$
zPT<8S7@1P5g)RQN6v8Za5}llPZN&zLJ+2zbzun$ETI(8WS67KBZ@ZZEtHP7@muHvO
zRw4w{Y$y6_gC}Ph$lAx%7GF}M{$byr>{8iw<gy5vsuGYTAjn_?Wqt}Etjg(`EJ5pS
z2V!=sYJ#OExe@QZ#ZUp?yl#^QjIVxa*JCPGm@EzZX-8U4v}w)QR3*?V<ol4*TX(WD
zfPvxJ{bNOGeDiOytT*00Vl?^97$I>Kcn9%R>>77DUWcX!KSpu`982?m$iV$!^8~EV
zqH3rV>}^CO)Ja}!91ur%@aEQiGG~IONuA(-s62_jk=@7Jf$XG3-OkGn0JPH~xP&Uq
zdzR*KfXJ9ZkQUy9BL^S(Asxe3yCn7OteejqJ|NP2xVA37P61o_!0Xc;>yc}hd|@p$
zjDd^4Z@M?_H%o<5tmy@=D!lRxjpcSGo7M1vPP~p+p6KU)xi))`igNRcd>8=tMcIkA
zZ6b|lp;0k_2V^fZ2Wc7(oAAaP$D%ro5lT}tM!nV*_xcmKGASABOaxet`Aq)8TzLcV
z2FWPt!;>F$+&<~3va{P}O28e2NwDF#G~D%f?ylX7Bk29Ov_s*p?NR0rb$4#B{fv5c
z%HDpfW5hKIKJC*xd-B(LU<J7#))jvWO^59@H)n10Qv@Of+4ax=hM0}jkWQi8O;iEv
z@CCw}mv%h33R&VvfHoQ1kS5$8V(tbD(5tOePMUlpd}(6=ni=aC*)o6o<)N9xgODQU
z$^tdOwQM4;h3N8;H|pnt1m?|o9v}#UY>4|Jh<RS3@pX6Z>jnJ*Wx#&rwHNM<_;=0n
zWA*7o!1)|>ZZUuSYuz;nn=5mMy!>dlZ&4Q{Md%!Ve1t4coaL8(Qw{yOBrDh2NU=~V
z^+}*s$Pm@z&Ch}c;vhafWEi;s#Y|t74aQWJlu%+JiQEjV$1Q=?^8Gs<bG+*@08W;T
z$lO|5MyQp;2)=Afbi$d%HYcBrv<>cuFd9Oxn&Pg``OEjB^1%zdxNP3(mvfUPWvCh5
z+X6o5o*oKEpcDhtjSbe`+prQLTqTyc(X&6K1~~H`kid@1BEgPb(1dvkGN}7A9HQ<b
zZ?&r(^`&APEf4xK6j+ADv(bVL&&RHO%IaVO%{F=pz&f2uGvC~)x&G)R(zGJUtY*6O
zap&b<G>Vf127cHO`_sH)Sm^NFO#{bOO1F&=151&(n7AtW;JgGm%J=Smz0=y93t|$?
z_QPF=s0WnggTL_K)~Gxbp(XP1^5eenHQoPE_gTl!eaYT1@{n!ns9-y(F0{4`*|qr(
zol5?%i338MAr1x00yUOpS=Iaj+r7=?z&R`WxD|G4)amZ(_Vj1fKvj`zMkDeQHu&@V
z78!jSC;00G0^#vhtfGUTE<C^AkYQFb_d;`kMrrn&`DM-a?WO$9qr_~WSdH_`!?fpV
zVj}(g1iTI&L04IxYqDBiAz_WGclvSF2yoh?6ayKPxXd=HL*m2(ZdPCVwKD6yB8ITb
z1xfjM9r+ok?&uw4#!&irn|Bo++J4!6Usu-xxHHNa&%7AW>V@S;PKLm~!0Agp{^-IU
z6;ShSiUlv6Cg&dH7?BBerY@|`PI(N=bmk6L^sp~OyS~+0?JuVG6@N6VDtH<b)isPP
zp;OB|>UqYS?rJ)yu5k1P6Ta1gwY~bg5-5(%6n7xaP0L9C4ry8r)7OqR)#{w_K4=lw
z8V)C;@bP`XU-<A{5d|R?V_ygCCh!qx!qp1uWXn4hj?NBfY)O@Ou|n?OwAfYZCus3T
z*7NKGizr5`RVy;02<iH45Ao9J_(Z%YNC%CCPoucNfnAlUPM~n8S@RKtUUvU@O)>~F
zI`=6J2y*y63i;gu9A`jJ1yTCj82}UX6E+=|F-+$Xyr^LdH$gck1>!pOSVtcq?!1&p
zX8l2Y2Gj+#of%VUoU-0UHasdaJU2i^KRHJ>O=?tgES$B@G<u<?lhjhHxU5L<R9xlM
z=Q9^bpV_9CLg0HOv+Cdm{WsW}o60P@&=H67avV=D`txx_*d$q-S1TR2$Vo64Y0%Gt
zCl|$y*HzrAc^)Eos!XPcf#JD;{ol-XQW5NO4j{pb%y682{5#AQ73WVD;@{4^4QE&0
zkj9Pnb@ol!Fsk^X-<?=oQY^BFP^G^94n8aK$pBfsLTw1WJFDB<8+L4;VAQKQNO_lB
z*GnTnboZjT$zE|ShD+<UCvANAwW?dZc>aUoY_`EB99Sx`Z{t#II+mwC;x`gEzXDDP
zWvrjAEb9U~8$YAEp^dH4qENEj;d(-DF7~!Ip4;GLgN>3Vbutg-X+Bv#Ui?<$SGBA)
zCqs(~<?`=_kMlM<l&cP_W)h@S8&71{yvSO8Iz7z4Htn{-JrN)6;5%Gp2S17moObvX
zE4Pmu7;0fW5({mXkBZMi3ow{(C&_a{OB#9Q>k$Fb$LTFpwt#cRPL8s$li;X>76OHs
zgk*220zO9>HHS;BEv?15A$3;=UGT|Yy|JJ2YJ-A{Xz3W6IpnKEPqER+@W@}UhdWGy
zxi#2|=4zlDI8628BOHS0Yg4bsZ^6LR!zMrk$IH(X$wO*JD*{5*pAIN1H@IcpIbVBa
zv|($}Z!D#iitqI*p0YzkY3nn33@C&vs?4<f(>##s3Jsy^z@?aq)io+0D?3_uH0q`*
zvENNb1Q!yn(zIBl#fI$i1ZJx!;nX7>X4jk2g#%Dt(F^WBG~8xzhRIj1HriPX&EGUO
z-nZU9E_vx-*gTavSP0;fg3-b)FDW=lzH80a?2UjWnMmWY>7;FRGhUy7uQX`##W54e
z2x;=q<Fd*f{C-c~bQeSWW=SN2EZCG%MsbL{LfsRY8^rB?c@*VquD(_X;2`;~z4sLV
z0+L|@FkHvSDl}c<5u+FWLDR4=OuqKqy!Zx$J!!3=w-$)||H*6K4wbF=VOb_$HP>43
zo>Rvsw?w3yztky@Cb{HFdKk=nZ`epmGK;zN*IXvGq9!AeR^zOD-ww>Jzw)K4d&A4I
z%ZBTGy{sxs$HtfoAlClLowRa^y+b>lY~_o4(edb_e@^$|*_oVf0^f$<LiNZD!?}7H
zq-oOE={3pX<4O;kyWbTPtNSwH+HZ)sPvIZ}t`h2L2HfAIEYHmNArFH6rulN&^>l;z
z{=rEK!4XDxT27ZA^ON06Qqo|qC-V9``{I;>`ls`z#Dm{7e#tkLxb#DlJ_g(4S*9WE
zcet`enPNQlnAAz{{Sx5V+7G5&Eygl@%us|8InE7ojD6hb5FuD?r<=n?MRlV@D&OZ1
z<G=di^JRmNT%t9u-#i>D>$LpbgJNqKyovmPYUtcU+h*5zyi!{3((?hH1)8CjRR%nm
zf2;|-{h58h)#z^c&aqH?rJGJt0R(%uzCt|AOGbC^zs=#*t*MTriZ`=C$A+WrC2oia
z1b^C>KDfB_nbJYeA%k>@Hz96j|ANDN&346o2Vd{o|LEu235Yd3-e*>qy)w0IdT?T7
zR0!>axfERWE~=Vu1|0?4ppxUoi~T2)C&dtgMfZ3`oV_?nLeUz@s`UrDK;`=@I%%|C
z<?H$%>-%Q^tQz{E(FK1tq^s+2ULl9X!q2E;0e%!LdR<oH#0S(*58k9?O9}aid=S~6
z%&PnyRwT9CNsWk3jBExA59#Y$!ht(Vwa24}rEqE!Bb+pT`SS=e9d5fRlo%u{rX@g*
zeYr$i@8<%GXMjM=3VaY@>v?QITVBLAk*)w+9e7f>H2c)z;UiCXtOvE2kcPWjXM3?7
z#IFgvJ`z0+qWQTIBB@R*WgHC7kiJREuf&!K>W2i5-H7`pHR^YE_dJQgbNlcp0xv!V
z8-&>HMR~r!ut==C2t_+3*#>XV<lEFEd#>Y5wSPtRU)`TtrCD2W$9u8lo+dIcQ0F+A
z<Lqp!ivTC9iUd6}jJg;Z<{g{=;u>4HD?<|Cn9!0oN9yUI)b@L|WJqThWkOTP<+dVD
ziX%sGYC;n@WP!w9<GdawOpAB7h5eGI`u(G<5T{5laNzYy5J7c20}_Z0nU^9ysF_~l
z(myR3L-7z3>x63UWUG^no+y=@Lc4wFYDHHX`8Lzy(}y^%eS!GjM39V6gK+v3u>%?$
zvlIoFoY)WmA`rr(tzX*UzmX@baeXLP(d(J^>{o08<T-OINEH7c9{+Cxa1{?3gE|Nw
zgWxF1Aelhj*(_<Q=>7_@*~k^bkM3?Owrdi^ECFGFLXV8Sq;&mHtuAJ0-fB-!qwf%U
zZq!p*B}DEP`5=@|zCn;^M*z1gF)zZc2->~!=SE~Pq8-!+*>J=XwP37bJs)jUo9qs{
z9FQX7pW6h<??0$(j8HvLn%lg^Y~Kl>?kDVcQQjN;P<7XVi`_mgI?k1}EKhq!&;mCa
z5+4dbTnnH*IWnB5um>G_Lu_XL22q1II|99I3kllb9o#Wkgln`Au*NM#_Pr(@jgHWB
z7y5HeLJ+k1%O}DVBqa;c;rTvE(HVhH%w-bMBX9fn;fZ!bDr6<SxR-5nT8`3+M*aO&
zfhn0~-ZZ(doINe#kBe&`SzB39xDx~6S!V-XNIRj*57mb}oTCp_Xrg0@n5(E2z78TY
zMWepHIY4X}EsZsSF|s*!e5`ne$g$4nwe{2Y;@-7a^hv>RB@p|V$(n}lc-?tj#z$6&
zy?&h6tCl%a!hN2h8)W7Myq#S{4JjEM^BO&V#}zibXivrf(M(?ul*^<W0>n<nxvSr3
zLVgmXh7cHB%i2JC@$vAy{+%CQQ@*(OpP1A4#Nlq5WsR3ZKq<ox2Yi)Q_0`<HHLAS2
z*>^2y=W}{S7fPGJZ*ofj{Bi0P%2vC%rY-w#Wx)Q~9h$~T8X=^pV~Dckj=FofI#NJu
z|8sjLtI^6^7{wb^U8cCtfyV1*`@UIHhzGr?cgo?Yr(}%OH<|*lsB4!U6Gey3O@Yk{
z)6>Ouenp)-er`BF_u4nDx3&bJQqzd40l*qgx@yQiFc7hYGw@6Pu^wnzuux&wI=3kU
zmgv9XN=Us|$W%S>TYsRgOsLWfEeqk@1*;WU-MJ`O=*<4VTmTPu>HH&b7UFFNJ<ZJD
zKgRfRz5%*N8#~e=jY{A6nfZ~*BqvFdrUY6r=Q`j2hx>KSzZ`93RX4^pinW%GsR*4i
zFLyzNmnO)yX4WL_D3l=M<kQp8z~H&Q@NmJ*0TT4z#Bj17W>6x%c-NN%q>y&mj7$~E
z9TH;n0Vfr9ym+lI$1ELHw}gl&gn{etg1WZ8%^Ze=_r>M|Tx0-=Xv2<w|0*O1k0_#s
zdMu_${)#bb62rqo!TBe5bH;G`2O1t;Jfo^P9QVF|n;%2_j;yuXFgG?hJiPQ|=_V!o
ze-{%LCp`WBzl$LBKYIq>y#Fi=2{?cUrz8Kr>*qVT<NSYC-|K&m``=|8jt=<m7innn
zIR^fT%zu|E$$wt`zhC~pkNrP8^?x_{zfut%{(nX4f92x;w-=}|x*U-x@+l>Jm{0?y
z^0zoO%0DH6z=87cALYUnUdBv;2JP*(5}~kALUI>+O^6}^M&S68g<fO>6{miOhk~T>
z(+{M?-|UYfvqcKr0+^T>9gC6tDdWy_oqTnT2e(&vZnFQQx#C^ig(@RALTMd5u0V^6
z{dEECtZ0`;?N(G#i^`be7kH|UwI6{GIec6E&ZYMd8yjPA1JbQuhp~<$T2#Rh%HG{S
z;S}Wfr}&t5%8H1Mz}=+QpG4<EQJLgMIIu{NO;5Ffw%zU3wcW@8%%0RaDR$Dv&nR>x
zrkwArn5Yi@S<DNga5g=}=iEEWV>Wk#rL+T_a7S5Frry(#&Z^05)m3(I@K<YfoUS&z
z)*-_P2fbFQ32WINN#s8xVhI$i+-OS@WLlrxcvkJ3vNGS}%UDkWbD2o+7s)*=#`}=c
zTwWV<^gnEMRKY<Kf#X^>e_Z*`G+57Q4`-V~uVs2k`Q7b8ysKZ8PoUDXQT}<?Z%mSo
zMAFxDb+#~ybZ65Ww&ZBVKFXH7kDnHrR#tGKeXGAuoJhFePYdTrGD&GH4<!3;!@C*u
zU4g`6;Zrx^|E8G=laSn!)i5^f!f&2uDV!I_ly}A$2j4Ly*?-gaHD)^%IA=XPzWJgB
zh->M6<c4fE1&K-h=v_PXZ!E8_miFsf3xL;{jPEsriQ^fW)rEZyb3FHU#4<PRRj}^B
z-LZe`jyqP|VUWqbO9<FXr($4`_OAu#qe3g5>b{VsZeW{FHn465v7{w7{46Xsk@_<v
zpu6JWC8^O>Qurkj2QSV(Ha(pT-T(^$fRB&AW~i`iTkH4q?pUaqzw6t<3gw<pHwwML
zAaNL(BhGwKPL3yG4=7;h;Dli?ya)RZy?fQJJ0~_O{95P#qbj725^nPd>W?e5fVYFr
zDJ>PmEJj|>4ol83+cl@ll-dfmMC51zM--S~3)Xa~Y$w+_dJd(7cAbt9I0x<KTCUWs
zN#*Ywqy1ItP=t_Gilw>=oJuBeFb`peib<Jxx&V4&qRi&r^)C(<VKb|PUGB*9@sR-W
zAVk8dBLVgKprP1jl#xA8;W)2`)5n&x<vgz{9Y4~TZ5bnykLu9v*Q$=pEt)2K)zrDU
z-2=PnA}O(3#R?$ff;>hOifqHj`TDQ?7$9KW`fNtnic=YLC3UAu(U;Yx{7`d+Nus_7
z*bBxLGysvF;r0-sypT2cH=&7NFA*07BXr=V&n~Ov9ZPuwEC!5e!2e9-p96~-gMQAi
z9)jY)aNh4BuWh8A9Qw*@Z1MXc#`L4*8B_-1lcE>-@J97ZBXV*Ei3-x><3wxC!uKy>
zF5#-yudQ!eP{xmvO7z-iiB{+AwWYHIfS^;S5gB~!%`Ht|D-tC1@68_1#Z*{WGW<5W
zy^xY7^^IO0o>jQo_uE0w^Nbfa&pmxXhiVg|Fi5b}{+n6sGb}Gu(>G`>k4yc7UJ3#)
zkO44)R8Yl4CNqEugG%U@cx}ZB1OJOtJCy{RXi0zw+=nsVZc}q;$9~}xw5$XgHkh8z
z&aN0h=l?bU-T(~XilKHSifEDiaX!L%N$W$1RqvvahMni-68L<v4_B8^`sw(N#WR0s
z-Lo0$VWb`$kRt7`&zUW917az|g+^ohJMP4$Ch{V1{$YYq=%E#8YV_T1^lb6Fb>3qO
zwe&VWsioR~du4i>3Z$UG!+UoJp8plzr<h!YDpZbRs<4ZY^o<?m!SVBNzejV3$cXMz
zns!J(UNY4pj{!gIXbn<jU)HFx{d{G(sl{kGGz<UBzcS1p)hvL<^#^Y+;c3I}p_=f-
zg0&ncYfU%SO)w-^wE$_pgu{3!q$(H|tu#kud^s6b+jJ?O5hFn1=jhLKNQs>dLschC
zsZ6<$JCJ_H{absaQ!QxKFD02bQoM3;_IKedth#f)Nc}Nu^<HO-?Uv2ZO0NvqZ9`7+
z+k36)Ov|&)C_Yg?{?aF4>st42ArB&gIv3L00nzF$U-?LR06!PaB!`=O&L*v&O&F`p
zxF$Zk@6=by{a!rXVo+Jx*24{Rypfc%JhA^>BVtx!xAia`HS8mn0BbnmM%@r?73~bz
zt#Ndf93I{oEd<xiPdk!!c>ex>HGF-IMQqEF7YLYA(-R_zABBH#{(>WEKk^2h_8N=b
zI<{|64m~s|+a6%d>Q6<RNKcnLN3Cl#n>m&dgI9jtVPPHYKM<uBO@FW|AlXK4xC=Vu
zr9B90i+Q5q#R@a@o31;S;`nX@xl5bp;Sl3y?cdX9(B}CM=n7BCJH3y7+1C024k_er
ze$fdF===wNO*YUablR&Q8S^)j4Yi^o7sk>Tb0Vz7lSg8ZGp4t7=(B1FH}Y@vUa25d
z5RTo#h$C4R+P~2Y(SQY^0(-^EeSJ!5UCx=^GrysIx<X$rCVCsyHg>n~mM^qw%T0tR
zv=X@CT1&v{&4md#jDw>&Y?1mZBw_w;D_(4?v#4_!LW^xKoc-+f)D8{dc%Ka`w#)et
z<XzWy)m{{7yf4ON3sHTIhr$i7GNQQ*xgmxc4K({)fD?GMLSTV<mQ`g-zkUgIhj>JY
zv^3-~93F|(AgDOZY4vKJS++jFVkw`w{avqjHUAa&G%o&~re)u<7&4r8gXN8Pdl1iA
z+abJbI*%#_4x~qR#Cd$vjk{N0|K_XL{%Ll_QR}Ss&Q%0J61rz;Xd4rlCzhB>`K_vp
z$}1(-)MH+?N4+Y+p=p<=pm%fYSrZ{?Nt?v`#%fztmAf3{qookBbz|ec0-FaG!1KM~
zsu4u|a>yr7jJ3O#$h0{ClV!BU0Q&{0o_0HN277qjy=T#b{i*7_r_~g!2ABCj0{u?M
z2e!4kvoiCc?I`ayjF_&DMv>BWoX6k1$Fg)Z`R@Oa?%|<Ho+M-cjYNBvOuKSQT#6@b
zcTLLM?f9R(^zXmTwo|#qx1Y9nxbgvj7DXB7&b>5?m>Pr_fnZ2ah{o=!tB7=|bEaUP
zO;@?`^9n{ZakQNtb+xyY$9z+F%j6BfXnz(r)E{Y$y}iD-|KkbjI?@Y8Inraf@@B)#
z7i|wGk1)Fm!b7FStp4@K9O=<^^0<Z@gB|-0>arvAtzwZ_<C;G^R46%>rIT9TK%3Pq
zA_&L3`M-HzwNauP>`@ih26qV~5a~JPj%r>3YQM2fb-LT^ewtgkp(lCza^|C}yMgI<
zVs!J~*AF&Z%U}6nac-I1QFjC6)IRxaoBjK&qB@}ttz3CaMh3Ryl(LtAd30_)jo)H3
zH_OP=wA1MwCFo9sj~5U94~%EZp6!pBQkh-x4!%)6{1@reF-*cF4?a|LM}ZNsLS7pQ
zYjaOZi%0;t;b4!oA!ql1fB0QJ^y(YW17f3f738Lf$j^o2x*AstQ}D||>(>vbkRrnx
z*e3#%;*QghgypIOeV!7;KWm&EN?B$1_8O(0+s7sQ7xPsl-Mdud>yK{~*Y97=EKL0I
zL(6v4iI(lP+8*p0UiN0+5b43dK5W`d*4L=(w8i{m3nL=RTKwvy*vrsM5P9&vV9UPN
zR_#hqAI((9qQxdEu^7m29!wk)Lq__5JM+L2%*II-X1dn0Y9s$M0>egwJa$}+47A+H
z!Hl|=DLN7+Gs~--^Axta8BjpxBa$t2Q#`v$2G;Uv_B8xr=s|lba?Tn-KAK(y^Lr8E
zLi#MR`7$^QbMJtca<&<8doRLkMzsEC2P4@lue6P4_s1eA+r3ZSzyINUlRLoJs(;my
zJ?JSW;e%*ospqhk#`#90nap;Y)6spqs57@F;R==h!{pJ?ENz18ddbF3;Bnvm{K%ry
z^JEx@<&}PF{gtt>$6-!;*cwVAToHamz0p`p@p|(xWZrb(l$5G1Df~oBQUVIar(At%
zZgfqx?;$T<JzIDpEBQPa5@s;ag4-3?_pld|UsX;C`(O|guX{E$O?IMDPCzn<ZI+q>
z-D%1qpZF_#iCbaaFM^h{s}%V9#`X-&SRG46_{mR(Z2Zow=j@Z;9UA{f8q6j(jHe@^
zBH&ZBL*GZJ=0pu)VYJhwD{#lS5}Zuk<~I7VRM7tkgP<^VJ;wrNWiCliQJJ*J>IB-p
z9#P2zm17R5aQV2|P+#-$^sh?oPG7rbe?~@k4H3VWp<jomfOE@wEG>^kwZ8#qa=rn(
z2=2O^J(QZ<WRw3n(ovR_+~|eV!gj=nFps@Da<hdInu^KfZ7GsO{fsF@@i}CWANGCo
zR}2=*5C2ex9RkbR8-{s&EIU4lf2UJ~JGYfEPn6o`(*qsj=4}s=2p+J$w<P{=EEt<&
zgsKNhy@cCl@Gi~F>kVQa`J5NvDP{U1xi|Ex)!bUXx|)(eP0g971|=|nn0?3t_qlr6
zHmslL{$D8Ip8~^q5x671(}5hcpK@z=3f#R{T)hcj-8}qJn0MEMAZ7(dN~1Mu-(oQO
zV_RB8oVzgpSxzy5S5Gj+E8zVdz#;RQ0RP1Dn->%8L-!0Xo5R{F#T~_fe+H}>@}sG|
zuUzM1?$_e^a=6$X1UPK-hXLmg3sN=^?zC;}k~Nt02}xuJeQ7eb+-PwUZhDQkNIumg
zh~eap6p_!$6ji^e>LUD7iWBs0YuAFdsT@B>L}-gi7)hHfm#zrN;mCo7glf}`OV5K{
zym}gJH!U<-(N@zrN9Hp0*6Qga*2W4ZT7|Mz`<c_P5IobU8DG|*T=CuxiJZZtEL=nM
zuR=Z!2*pKFebf(rU~m{j^^-n2*++F4CSh{3b&?J{4K2;hp#@CmJTDYF?>r9j!GHYI
zg;Zc7`AZ6Xy5xdwMnbVTc3>x8(;e6g;8$miJ^SuQ4T6M9O78dxBhk|22<WH<24v}m
zQfh>$sJWmoLq2R5|K(?8l?yGaL$(6pENQWY985m;IWuL-T}<IFqBp=>`sk2)U&DuP
ziG|vYJUS9&TDh-YDruDRwNvXY%&}Q5<|#)t3d->w)Oadz4I#OMjsQkyaEt750a;RW
z!y=lG5S?Ckw1K=&&EQ2xrT656kcPzu`pt8drU+|Il6&6-%}lt=-#6~)?F>NY)!(hL
zw6p9TS(-KMKc%#u0UJJq%<<d`uC<+|gpd5{n>cvG19qsD<sI6tn6BAMOLW*w&L5m`
z&l|`vIWMRbx`@n%($469at%R{hhT}E6lauqXrGUl@$Epj@=k#9>D-<jo+#b(ZuaEV
z(Y`1B2|VeG)$n*jUY7^|%LB6%Qq_t<Sj!|^Kk=@R<Hpw*0vE(;h0)%Y_DQLdS9UIz
zpVvLwCuDL`?LGhAfonSsEgM!D=&^?sp8W%(hSmRBBQq_X#%}~@w?m7(x{ob=u)h42
z^A%vu!^x6vJ%6QV&yE5s%lePt2mArj0?rRsuc(GI3w)~OJ4%_4Gx++NpWXmhgzs#?
z&FJz_Tcch|@xtcK4yOT2CZ`K{%w||>Mw|{N!9$<DUgj_m!&V^>&h{hj_B!{%g)#;0
z<XpGv!f9IKSu_cCX5&YmhXXcuL(4@IsLu>(f6qN{7cXYBp~z;JBes_PZl4p+BtVac
z)9Hwv8czaL$AY-%#oGEV9tjiGn0o*G3z_Jjevd>@k9+_}nlVK{*4_{tvOQB(*BBZg
zqM;Oh*ryeSJzXtzdU4ax^fk0+WV)2%JUz^jd%j%v=-8>D1orYwdy3WIJ}mII)pC*<
zzf3}n_qRIR>5^{1u4pmy);1P+X5r%ZNy)uhPG&9rPJ0A%nDtHC2Wu#y$Jt*m-oTk)
z))xZ&7)nUk*eVGL7h~aHJoEPTCp?@a4z(*O?l0#9j_$j#e;xgN-iC!79$L>WD!k@}
zcLOtbxdK*bG;IyZz@lI*Pc^fTq&fV^Fv=GiF520T)Ga*%m&&~l4`4vX#4Z^<FV5Z>
z3|ZbeqWDo*+Ye*gLpXe_StrGY2HU)n{08F>J<hai4*UXMKUu%~V7+gcNtzbA{YDI{
zyGMxfkOF(@n-_`Mg1l>t>?iv^z!}O**LSmf-tmoR+KD4b;g|38o^kEA6;gdVenJ-R
zu|Alze<eyTUDgNJTV234Du2JrXjgKQ)!_IdLS5T90Ww7VjMkBkg?m_75y8DqU2&Aj
z2!K|$vw+&WKAkVnI{&$fR#AbWl(2(2c$lZ13Uw(tJA%8shEV_4VrO3LNxUWeO${}M
z*sGA7R%T7&3v^gc*A-T5z7Ep4kZyvlcMV<Kl#jz#0+ooj%tWlEuiR1U@)kFQ^e{0Z
zOr+R}YI9BUm}Y#?gH*mU9dtRQwTb;k{@gsykP!#;%8{Ywx|$FGpDGDlX~@GP#K9rr
zfPetEv#<40nb{G}e@Rj&Wb=$ISKI)&AT^qZ@f{K6i!4gC>>odZ`>Lkj+_fHUe}U6o
z61fymbR(gO4$KUr+&{WsveJvTdf=+}o$G{l$AmYCZiPF^v_GKyUjJo$b$tH>T9NHg
z+0EVK?i-Ju=5oj5(G@{U_vDJ|%&xb^Pg_%uxcvx|8X0xvPk(vHn7;n1RWIl0UvE9n
zY!1qvh#+Kng>Bb;CV}`#q3~W58wK%G!(jffEz5?o{2;n-BkIG02i@VvklpPtqH~=?
zVs}GQd_U&a)-!)z+k?-pyCcM8ktkOQ?B||$a#QV8*t3~!@2y#tDPL5Qk9qJ4(zEcs
zZnR-v)rjR15YDfBKtU<U;6VS<LS-S>Fq<47mv1^HofEM+OS!Q(^LmAw=B#8$?@s|5
zSk3J?YRM*ZRDIF%`FNnl-HeyuWH-MGTI;n*gP37eSXul9@j=MK#Ng6wT~<PTR*-7l
zjjw|D1<Cd1i;*Af=UOeq>Nbh0x}m4bgVoU8=Qjd^zIyDv{(`l}pMcupW?1GGdm=5n
zrii~wf9}QirEeVNZGBsprqv@_tqkmzf|H$w<$0o{P8N~TQa+?zenbpHMFOKc^7b2#
z!afSIQ>!&dRh@mLo8a9{_HXl(9U5PX_0o!hlkar}^!%fWmTlByFM8|DzMKO1zdB`N
zwrtq7<W))4oe$)p3`dM7qPOste~?$=-6LciYvPaSXL<j}W&TXvz&-h)mxs*eZFGnc
zS#;r6#VH66LfE@{pm2zMxKGE{u&TJ#*4)y<zb&Sn<wA^lBf4r`D$Qus_Rjg^RV<qZ
zeSq3rt4`0@R>by)36EALlx^J_XJr{>|7Z=DAAR~&)vo>yS6rJ;i>!*^fS;Wn9$D)N
zIb!W{iPNqJrwQHgUcql2srwM%J>}I4a?htP;5gg4jN9Rg@+<Pb0HN{xB|Z0<JBR4E
z^)kOpe#xwqG<br_>c!;!shq!>TyZdU#DA-**^Z1NR4Nz=x%Tf|I-O*gk*EBv5n%M<
z*EZoN|4Fo%UG%fZGF_ttyS2`s`qX-hUDr-S)V*+maBK}%fQR!B9#;p&Ve#T2MOVQW
zhc<F0G8pC0r{3Pn0%5;>iH(EnT<i4=vM+k%!c^WA0s`pw!^WjdNIig$Cqg>WxHp!U
zOA~xSlp&&!*v>DN<6!yUoH;6f0AdvA*31;9Ud-mIO5+gyJ^V~{8<BG9H_M#Z$fq>C
zKFw}b-Q-ujlWw{gQP}>pTFhEE17M99ZyZQDCW8f$^`9J-87|mj%sgILW@LD_-b6Q1
z(h*_GD-n&39DjTdypi6U_yxbUAc~i8Xso~TpjDT~Tz#IU1Fhw5>f<e_&gKd2=*_p!
zf758Goiqks2|hMjebi?qUkI+O7S4MCik3^qAOt{p|IRB2?zyIpbk%8H_$`p#rj*dU
zo1SM3s-Thn*dg5p5mau;`U73I8Fv4g($l0x7u(!cw!o{!{VZ&D^kA#*OsM9z<x7@)
za-LO{;ee<-sI11<+QRpcVJ|VlACs9B{xuFaJKNGkx8PvrB=ZEy03Q*{hOz~C{gLc)
zWYF&0H&TYgl|AX;y@=qxXPFIi)nYc&+JVbVx5^3dNL(Nl&*h{xnSpq7*e}QJvOyaS
zXME>Qo6yz07-V1YffNJ1$mAy>ysEqh@$Qpr#CJr7qoZ?+JGP32bN!YlBrxT~zD{4#
zWaIcLc?~AjM9#Gi#$LgI%ZbT+U=`}UTu{H<xX?gp7qr4Qrr1$`X|OYqBXaZ-12A5a
zJXls)IbbQQ5q6u^MZHJ?iHePZ4RCX|{GDIJ&aJbngk>qUJ#&d#li2U~o~N!TiJN*6
zO6zfdBvdbte5r}AVYcU8%oZUa_{zdP%tiP*_a-moU9~~{l11dCH$>!}Qb|~zT?6dP
zqy<A*%-`LDoQi}ln(zGbML)UHVP|damX9Om`PHgM%v<njD_0&H?%juXtW?t6Gh!8S
z%?_W+1?O`l&|@Xr90{B)+;+z|cbW{7<=~R{HjDbCQzt9grf2EvpQ2y54GnW%AfgV_
z)Y0_ocBL}mUWYa%>qOJ%7(%J7^&|&R+LsOKSs5SxZ+?1R`ME;BmCRHe3py2~UJr{H
zYt3y~ep>U$AQ9`-Ubr`_R`L$UBBvh$==ez2H(rf-lJF#vDnE;6xmJs4`6S-~-oEE>
zFr;{?&-SX~s?{L$mK3KJJvc{BXZ~5wOgxDW6DRSMHl!~Yc`&3kQvHsjK~(_M`}{+O
z!%mo1_xo)GP+lY1t&5#G858+bh97_&E!~8aTV5`|1_U99PTZdqUVt^P>u=6z6i5RI
zY&AuSX?p(zaxv^>MuaAg$=gspLGY))zN!%Hnh0AKyZEwUoxl>CpZgj@+DrJUzm?Ay
zKkp&X{AL;3=H=_VqCmtMF_yB<!w2cb5kgm4VRG*1G%E*z5n0t#BHnC?!J$@>8x3Y9
z8!C%fgN+kU+EasvJ68a=dfL$kO&fy+LX=pZfcj}ifI6zR#;}Ghi(-C`GNT*XtHakt
zm`-w{U&+@ht9xcuRqh1O*!Rp?-xV<znES*{#vT~INhS2cMvX2B!AnP@bgSCmqXcbS
z5-=t#K*pPbXwy#|Xaa>ke=v1gpe`h(T=d16x<9uZz-HTPd1^7z3W<TS1<<ln-T8jQ
z|M_1oKp^o;rMQ~PVk)DV1&v<^i2|&bL?+5Tm6M)@%W&_xS2qoupy_t#gENof=yPW_
z9^Of&Ty`*uv+lApfl3NPFcJQoajkd92W|qpNK9`?&x1Ph`<jrLSvb=an-H;(Ayiqe
zbx*7Jej8&tqcg!haw1+`Emsj+hnsk7g#H5lWM_g9b2R;^QAl~Ru2LrRa=99vfpUbo
z{IF%Wz1pAb2wD=eH+Eensjoc|h##Pz6N=1z&Z@sEj*mTM<2D4)Ec%u!TG@bP&877w
zCh|<D8gt*G2PGdu8I8s8U#-PD0$!0}XH9y0qoAQhG62~q8ceosMCK>dDaE2DA_j}<
zBHfw{Erlie&(Spvg*xf+u!O0-i9s&J41-aXOcT{|7TC$j>Xcv97+HdNcVG7+AWT#T
z^ha%u5FCtB-Ws*?nh?Zq<S7bVb{;xGHHxRX^WXGRH7N*WZXhp#K)787v1V7Oin<GG
z!ZS@a5*G{jzA*-vUwSp7cDh*Ic-#=GV<@F+1Fjys*&@d028i#tOr^5?Q1<XQ#_!UT
zi&pO%&V+vKBdzh)R1RmBjuy#t;%$@eHQOek=u>aQ-3xv%O(w-(WMe-T`q#Nia-Ci=
zq>4TCOthZhIu9tl&o$~26T9&aF$l`a(C&CJ9d|o1u#5Ea61!BfQVu@pvX;vQ*CR;i
z*)$@3sC{pu$-34GA?+T~`DS5QU6iV&<U3McPx-U~VM9#MwHQC0k$6(6ZSLLHBjT)-
z%dsOO4p|;6wN-ysL_ipp`dPk9NKFmz?>x{kC-6fmUcewZ<QwOkZBm1)53dzDT>30H
z!@j|fGJ&p)IpEH?WvI<@Y(14=NGbn_TY^M$Q_I2!ltlAjcq{uXI{qzxSZ}oOB(}cJ
z(t(oET1SHcw+Lh^F)!ME`r;dTC`Jn~G(5UeEWn#24+5@-SIED8&2Ofrnm?MrRQ<-T
z^7(UtRe-;w@DaWK70(Njyj=YDo@>*Ox&{6ZY*y-7hhuPmrrsa;nIROr;#b0gJXt8}
z?RU9Y;xrP`s1!9(7c0dATNA15s|3W<32lGFm_qTS>82DY35X+Q2c@ZvsWtYnxb^>`
z>MR24TDmAra0nVS5ZooWySuwXg1fuBySoK<cL<Q+aB-L5?(ThffA^r7!r*dFRqeIE
zwf3!TA($CZk2?;fzU1SW#fxj)cy}RbI+bURs;bBM*T>?Y_>wu|JN(<>x}Kb;Z(SL`
z=dhT~7YtA6cxV4zhC<%`TfwL2wsP^Y{zJ)ka-|Wt|AmDQuXB5m@qjtU+EZKBQNzy+
zq+s$_mi#iH2K&om{nL|(X9yRTgHSSQ1Jk15*ZRcNa^t@awP;JPC!4WWIPX=1+AxbM
zNkhOo!YL@X@ENgbz1+~>?;mTmT+560DA4svVY_QkcQDSbm@){rUUM6P%}BB`{KR|@
zB6~TvZ`AXIvVQN%$eT;GTgxfoP*CfTPYV^RX}_QP=V#-!a!nQrnyaV}aN{qFKFtn}
zzK&T)laclsws?W*IH%b#uG%t(`6K{nGP`afCNM~Fe`O_VIhK*ub<FR9pXEiXB&)V2
z8?84e9pQBT^uI|5@sh^wX$KhD39$Q|+;Y_uD;v@E%kM|aP5)Fh+Ir7csbN&6VaggA
zq6)SAMpAhF^2oxqFpR3E<sLj!gqn$xgDfFIdv-^QW!08wn5@Z0CXio66e8cjH#d7!
zMhsH<Gg_7lk)Ry98qpbDIkGWO`p@S!hrW#`n{mZ7o2NzI{aZOY+Gzc|iTjxc-{})6
zyGTs+{Wmc6AmshuP||M9r!+ppy5b(jm_Ayz(9ll16Kd*qnu`jGJuOh1HP@TzL5Dnv
z4B7B)@R7vZ6|7m-O-JL4alJf!ozSs~Kkf9`gsEsm<v3L+9*Abr264?JRmZP|@Axa}
zSNoy-*N%%yw3MiEn=xlCZ(eGiKL>e2wn>sWN5P&&hKj{XvQ;=F#316$mOh!mAoth1
zuwrMY;<DTwg>D@_dB{w@EUG%h)p6sKHk99S)e5oIlU~#NjOmRg;L|fOT4{6*^zc}5
z3=4PvkXu%#<?qgLGGyb!{0<T6)pn|Zow^4_yu|&gy~x>|Z$>ZOu)vX=7?+lK$IAS(
zhG+yaMi&s(W`|)2V8A_p=z;UB{nZ14yel*Agx|pf|9apibb3$f`$qO_wFVRmU)_tt
z!#dGvd?&y}3fZ&0n3BK=xba~<{GAzjZ$DdXzsPx49qA>_=*XPP;INO7vwovAeg~sm
zI5~VYaXz?hi|Q=BCQzvksSQK7brcrwd?)!kTG4Qk<zkzhU7~EaWff6HgIaNh1E+Gz
z=C){6X*dWEId^Z7%X%X5on@HZX6c#dee@;K%22RtAG{**xGJWBgJVa#T<lf&OqAxK
z=>8k9#nA@dF{1>xmkOfa!PdOcT6rorcS{iJ+dW$o5_L_18|9p_jc*Z|dq%6@n`$s?
zpl`k&o-X(|z3ws7IoV}<i!J>b9P~ALi3vTZJ=1EpZH>>o;~N}#sX|n22@%MqzyGUO
zHS$1EiTn>(A??O`o)SK92Q}GSPF)oue+|Czl`A@P5-2JRm5%HRXmXM5U6H3Z{8s6E
z$VOgZd{eiCRG?C~KRu&ZRQ70Hzy9ME!r0RP`t9JJQ?%mQaqE1Yc4aB&vJENf-EX$9
zUS_u1867>UK0Md)&Dtm$*kgIdg#X)XPfu-XDW-M;u51`&%axZpYf;W3-TbjdVi0-e
ziGF9xR1!qQT!l-7l4dYSy)T3G<$sz2L)hwtqJ^)%QY};)9JfDo*`c+N_sE3*eNtMQ
z`i7)|M};@i<!0-Pdv>zbs<`4DpK511hH*PDt&j;DYtVgq`Q<QK^F_BdTgMP}6JHQm
zyvHd@1kuAl^FudG(Lw3Ub&mI{H|J&Yx9`?72RbiN-qPf+?w~az+7x_f^1p9Cf26+0
zweG@(j>OJ+Ms@#PKZ3NL<&>jyJza7#d5jy#%k8nW?6o8obFjDbeS&eyPu2I#jOhy8
z6>s~wUwA@O@v5_`?GZ%JtY#<sM}qJ0iIfk~iMng{jE)WS)YyA7=V+&f>wwiG<;Knb
zH~B$H-r@S^*TxP8dffmeJF!O$NuiYz{sHNMLJgetSRsQVCAKTIr*{oIr?2M8^UxhQ
zlKmII^>~u)Xh?A5c=w|?&Gqd&>#GaTWy+1vVhJ)ppssbRNiU!w&8S|Bzdv&!hok&%
zgLr=QT1R9_0_-A=R1~C-U6jNYLMaAyuZ$XPtSiBn%zzj;kl2x+M^Z-HloEu5crOXz
z-fPOBR|pZe>>_}MOF{geNwmC4QhTx%?9f;DNd$ler~jUcXmhQPR<J9(T&=R7u6(gh
z46fcv>L?w?n|ZsN3P)`?QKixpK@i!~64&Lr_<e+yZjI#x@-CigBJb?r_IjwR=&WZ}
z6@o=pA{6GT*`C!oxlhn6?a%u2xpV{#J+#rlC)zH_YVumR-p#<r1s8N2pv+YG{evKf
zAy$A5L1U;((ssF~|J+Z`7#bRL&brzBGZl651lvN*>ws@j2h$5NzKl<(kYm|Qf>Ae@
z?fUr(Y`({HdeG7UxR_N3$p44R)=H%+wlVXlN7r)R7$JQ$jPI>?s(8YF-M>X8+M%Yq
z`x{FTOg4eBfxv<XXp=evDf`lGhiWJAIX<>nP!&T-fgwXqH%_Rpl$s_mB7%HoUx$1_
zWeAWu^-UaQ^Ddh4n##4=tSimeUMnl^)IYQe$KHEnnn3E|8xHIWW^~rIBvR}@z7!|*
zQyRy#8_<}+==hQ6>jv&wSY&L!Dgq2a3w+Cl5<m%;E`u=GweDQIDN-;>@C#y5Kb?la
zNJ;My^v2rz)~vqE6|5>rS_+dI3PH23IN$~_PkQF%`mpcOV-b}s%1O>Qo(H{!l;u_R
zY+hPT_k(Uz@vDCQ@Y!E#(i0eq2<E=)(i(k0u4T=O3S3Waq~}-FznxwMqto2TZW1Ir
zx%oa=8gH(B_V=&*ry%(*ZR~fDpq$F?6@aq5a+@-?RXHUjoUUaJEN<M+!n@$<qfeoy
z*1UDO`Hvm>)EjDM?M)&`e8zP_QLl~<rK8)o_`-x{G7)1%FYKqjUdfN*vc=K?L?Oq-
z^n3a9>?oD@HI6c+d~syFdOOJx9gj4&pT7?9yiqL<m%mi8E=4s@N_{y$vyE00-KC>q
zL5PmN!sX4lSXP(Ln@D|kY0cF}({F^fUbF*Ofmr~V^Y`ap68sAn{+uNuO5^t@mh=4r
z09e&^iX-C~2srY4&K_YDXvJ5aXXMi~%$IW>%#^jb5U(9s*`_sWk;TgDEZG?omv-Ix
zc?|CN%Fa>mB_yBZ3G^g@&x>0)(CYj~z4QklMbt$Yv=M#4A#-zozMAJ=lxJ|`I`;c!
zNhs-sCzENpw|`Pst%py;G!ei1^nvJ^;Py(^&KDgPG9bkd=msFcRX>oM==r<DZuNVd
zDRm>2a(1~cUQtmp3j*@d-wS5qBB<{$!wbl-+05~1u7*oZcx2pYadE`NQ+H>k-saZ5
zOSnK8IXGO%(ItaX9riT40c40(wpazQCu`FaXbUd{2AzD?m-5wKRl8^znLxSm?j=!j
z88*C=jmH{2$i!yN0O=X|r5meg#fK4Gl2==D!+=cI({D=Qvzxl^Xts-?x4$?=1Oa;w
zOTLn(Z>n!yY0`pAa%|gbX14Yk8~Z-?YM*!r1nLC0VIj{<^ff;iw@Qiqb9}^hXV%r%
z+?-Bm&&Q?~cQ4?{acgN7-i${C!ytA3?JV3_O1ho*xHMicZDrVIaOkBQjF9}&9<IJZ
z`-z3m30Ba<7hdMC9A9>jSOU&!n)Y|)!a-##y6xVbmbyhdKZ%5ph2^z=ysB@hbu<ks
z7or2U)FzD+AK(zt1xy9)5Vi07RBF31t>Cq(`vjL7=deLFUm^h~M=WrNZ*3BC52;S<
z!D~iD4Q%HR)>*AYWwpW-5hCB<n`!+mIOT8sJNH(z7c#`ynRZ1LQNVF_N+bz+<n@)n
zLvV(Q0$@Mm*Z}*1g)IMyIUBh{B<*&8(DIVhkTM&A(AlI1tcEe5w0y<lT09-cjT|eY
z9evA`6%D6ofGtSFW=rxxjEX{S^&`zcoJW&w$uZSNYat=Jv>Al?T6d~f!PqaT8OWxm
zSczBbr<UW4*n}H-x6FH16Gv=wp3wL5i|yc)<<q>=L6pyx?Dyq#hwenZ^uPC-6_T{v
z0hN?acjm(MJ$E{7$BHB0jO0$~c>fr46PRrK#U3Y-Ue0>}HC?9o#Cu+E=_s+;H@HJv
zwx9$Lq8O+BVW6SP&9^HS=7HR@2w0sN7(nANVD0PY1T(%aud>VF@(gi;^TLUW--|Rj
zBdN81dV>OB;Jw$vjRruyiS8-=<mt!8;!4xBU(oc7l);Z0HWJ}8pg7W~5M?)K%7|%R
zVDbS@V>4fZmQ&9q!Dj#4YCGRmBd$l5Y$98e)ut$<uR{Q?)kyA%wld`VfiI|oJ{$VH
zU0qaR8!*G@lKib;B$#|KGXm{#=Z`}(I+>zHSza_z1YEu_Hp*lX5rL>3Qk)^}&hyEf
z6+NFovemC_SDI?rp4vUl8|OWO^1+eP5o7DkYA!*1#9raT65zLUuOdzfDM;&eHkkz3
z_@}>#06}8Ek!G46!)MQf;)>@u-QB0Bj0^`Ss_54B+Vybkm7yd4p-E$4Vf+Bz8xAzJ
zAX1rqK!8W{Eo?<?q}v<VWW%;%#nAo~T0{Y4z^@r}NB`en`vW}cT2~xbvmSWF{x%D4
z73A^J9^YoHWmFcK3_5%CM2#3%_<n6rCi$9)D^V*6gM*l%W^V?KD7w;-J5H&f&Iy3<
zmg$rW(>D{QiYwT87zUnFEN0-*wCIOI$I_QZipIO|T?pyPzkW&lS=|IuOplvc(sYOm
z9`xPmUaUAi>0q=#NFU$<`oK_GEI;rCalJlmJkz;*5->aAodalYFdQW@iODNTTHd~u
z!;RWv18&X%@IIaZ?60e-!24t^^M7*ElYZ^XzwF7NHN*!IX#Vlc##duKoBW-FfavrO
z{r*CP|5lEcyP4Ql%e3l}Zq(pHY^98588wwqRD>(b&|}Ql2<O`EY7)kU6x}AgJ85BM
zqeRNob$Ymj^_m-0eSc|;xnQx9nn9DE@}N{@P5({v&%-2<*RpIS6i6H$eGl(7W)Nwj
z<R$9*@UZ{$2etpq#jjO@mCl?nqO^7d26lXX`$!V0GOUN=@#&N7WLSIAwZ*gN5Ie^8
zaQN;#qUpmYbaXv4PnecL9UrwDG~f{43;RVW*UkY4d26&(Z`^q^EnTVKgO7~7atmAl
z<mqL10&^ziF`h4Dn#*Q|U8_Kwlu1KQV%GS*nEy9$<aZ`qaq{oKNLQDkME}aK(HB26
zbA|#ADQ+t+w)aJ)Gf5BOTOC&soe`L|9J|P=o($2!RLA?b1;*v@axdy1IH0Dk@l)XY
z!kOXWXVCQb646Faah{zcf&}Lk%Ioqt9*&-$T}`e9o}mM*|315azy0h3S`n#hLqLws
zA@b*KrN6VNM(C;S?XhetO6blkuJ56(S~t|e;0rkS=w2n{h5Vk#K7)CE(0d;bItT=x
zckpxM8rykm9hNOm3=kUQM;k63Q6(cNy(Zf)z__(RI!DojZuojshrbTA?UrTw7pChb
zQr|J$%x(}qp9}+idmF8l+06RPcg)(fswImAXs6pqvK9RlvHjbHHqjzEzN#gLe}&NO
zveq_9$;pcL^>DR4Jw0xzT!H(sv|H)@pYcWz;19%tb&A@@NPc(9A=}p=ntBOSGNn!%
z_M?|fNCeI&RzdLRN6L3<@AInz#F5OD*S6?Awkf!SVT9bP37QV>>+cgbWo-1xQ>zNf
zQkV+0s-Jf!OTZV&2TV>*HdMt3*vb9Z?F%-L6m3B=-+}C@Q&mz%087kJa9=!AfW{vu
zseqW)@=avD0j=myMjH{3vOMJmJB8{OJnwEVKPv_dugG@3Pa1`2T`W0fs%%i-u$VNQ
z8`*y!4CULfzR{2C_1}2YFRju4h)%hg!%15+iCEdQ9#$C{$XG|jPsqvfImQ0>lVr$v
z$u~EOui|y(7(F~wE{N&X4cj}SYl=`Yui-*l<5Nw`uu+fv+QGtFQK*8#3C=ZO6mDhW
z)92?Y*p&UryL7Ugk+QAA8ubMI7;ApkHPg_@tnThJv*)|&ts`(Fb<fddZiEWfgjpvR
z*dLuexRfncudb75x+KV()ETK8v0BD1SgnZ?Q>NeM|NYyqn)dXO8Xn#_MXGbn>y-Pi
z)j+*DUW)8{fzceBfs|mc*)yULw-uzDM>&3dq5L;%M8@0hg{+CUdKP@%2YiC8-KM+%
zp`YgPKp9dpDI=VqNr><D7LAi4sjYJ32Yu5!a9V4sGoxSAwJKf19=Hd~rti7Q8C}Ff
z&a(2-ZLx}OnUA#f>?DnPrz0d{B(j1<>kFB!j0Vai@uVmqbx$UPDu6@Pj_!xj*f(1K
zk?wkQDP5Uemw^uPo+@Rje)5E-OC$fXKULSCk3zjDzizLl)>Xcj3_Kv`85tYA-HJzD
zuZvL8azx^k4~knVSiiuS8HR1mV>^db0jjH&_@wWtTrJc=n(Uqv_@=!c6F%n$;h1vD
zP<y(<JO*H+{(MUxT0Ct?YxAXZCF>gaH{LR1+<*I$+#`CcF%Mj{q*wvRVEJh53A0Dh
zl&@choRLUhAZYTP-6*KuSXtd>s1$d~NYQAP)L2U5!3;6_CzzFG*pHPwd3T;SOiMgR
zthl(eZ|-#~&R<w;9b|)fye@Qq%f-Ui-l@ftn&7!b3OSl59-62%#Be~UTE;d61qIoX
z=sw>mNJ%qH_<p{Y-vC;32ui5OwD%W`?UkDR9g_-tb)hZ5%IuH@0Rl{<dnRN!daybn
z5Mjv+W{rJy==53d$Gm?_TH`Sz@q5ni#8AFcXT9AF_$6hJG{fYj@<3?Fru57^R=@J|
zbBW;g<!7)DI~He_<F&f|k-XM&a-V1M$)k@~F1hjwSmR$<Q`Q1|c9f<no7YupqJu8=
zKVBNiH!yJj-(>YskE@9{&_65(rhcq7IqCXB2v*Q<)@LI(%mkxri&su;6w4`m0zmgh
zpRP^1x`J;~7OhO~_aLd~2|_Yvh?x=sKTYrLo|Hc;cm%&%YPz_d=dlv(fR6t7v0i!t
z%Je0rMbJ5^LXqt<`7`3>1S>ka$yx3wucU39Ej5sj2PlSNGiWO^YU2pssxB`nMPt+r
z48s8f7v7THHwFE{$`%|8^I>I9kY3s=X)y%#vB`MZy~Sd=RF50-z{|aCP4^q_8(KQn
zTc6=2d+zG?L($XOq9n1yCC^Sr<y=z6X!d7nctF{t7}uCPWtYR<!}2Y?C<Y5kfCD{a
zogQj*wR)sr3Zri`%C1JcvLYEh)=UAEf_cy8WBMuzYli2~8VBkULE_n8{To(2EQzA=
zpk*yP>s~o%+qP^(QP1O(ilhoRs$FHel}m;VqgT-lS!FnF<`{h3Vw8?Mhph#<p8z&<
zk-mqsYG^m$F0Z!C8&ew2^Q%TOE*H;O+4Q+l+QyK8oq}N~>6c*&p{LmH+x><7>>q&Y
zdnMXpYGxT`-st{ujV&SAe`=Mkw&Mn+i*3E{#Za1xDRs_@odz>yztZq`u{rI;Ls##$
z5w33J7Z4*Ro12_qKg0VK33f%+2-R1LS}UCeNE}w)ZY>C!_KD5rNjNqO2^tv$+<|%j
zNV9Xi<9T;o>(%xq|7bmJCTXVkpX*cS-$sTgWj5~Nhj-NqQ?s&y4*T6zcpp&)uyXQ_
zhC>;;-t|XZZ0g0>atB&F!3OrYY{kk1e@^&w{F*`PA^(|lOi+p%9~X83ySMA&nq3)c
z#Xs$!^dR~IASv0Bth}Y#CXQjz$KsB6Gcv@)KKj0?Htj%g;Aq(@LUo?6sI%G#$jT|#
z(i<NA+v(C|;SLw>o{j$#=zN0~&9+wRrbC0Dw}v%lG7Ww2$vtHW66T%v`KhKJw2r#_
za-y3I>e`|zT=?aJ|6V@r-*?K&tiPhKL?X{F=cuI*_Ya>tT;8Rl=QdYka*z*d_>~_&
zk=_sL4FWI~PeP(|{f)0W9DB)NZ8k5SMZ_Zr<rrg%gI63NWCDb>{s;d>1Qd<vjQe6j
zhbzNbsyi;&s7#8<y9GSYHJ~A#%I<Xl?VDwnZ)2Ml>KK?=0fZi|R-OO<?qYl@b$Xd*
z`t-WVY-BY4o}wJH1^AC$4n(6%K$?HFxBt@Axn7ygMxUs}#2HUeNSuQJq4uYZvwpTx
zLnEUrgrMQw{*c2oB*GB`DtJG}t5Qgz_g9+R>CnDc+%(i^kLA%x>WtEbwDb=}BW;^=
z56Ejt>9<=psprK)>2!a+qV<yt8j5M=B$%z$8I+kd72Lyt5QMF!S$Yli){f#6A%gdq
zRRZ>Os(8Z;(Yt`W!?L7kx?%R)*Zr~}?MF6CGGola_`@6=-VYJ9*WL5k<-&TNI$?BX
zN>a<?p?Q0tl+o$ldIg)B9e3ySXf!{Y%Y<2nNv^3g=ns&@Fw(k#oW}%msgI02EL<U`
z2XL=%Se>+XPae$OeUYX}gAAO^Y2jT7lI&Sr??SxbI`b|0v>eSRzZlzpRDNF^2BHZW
zo#9VFi|Q3$7f|(AQbnUYj9rA8(F%K8nBHvu{)Qozkg5Di;TSUPKIW(M#=RR*CosQh
zkN7iH1nSctVq>hvQlQxwOBlq8&1wM0A!I<qhrI_h^qtAwBC4l$4L^U}^}1Ic-uraz
z;{Y03mwFWLko3DSgl9osZ4FEC_rH=!+SK9fB91>4Y=Kfp2g(Ij7HbE->FC$^Xh2qu
zMpEToS@l`-VEQJOGP{4_q$!jmZ6rw=Lt|yl2dRQPckAFYAb69o@~*3=^s#p280c$8
zj1H1x7vKsO?tSEs?ox}&Qf!;Z;6~r9w7Z@vCG5OS7&LRl@GX7O{qV`w=?-VB{$$=5
z@bce#WD|DfN)NKJ&PCq=18O^I@Nn~w+a->t*YFd(Te_XOEVr)|y(s;zy!t<;s|QcG
z2zXdL0RKxU`->V>@fRC8ub5rVriOW+(Y4<|+QUKqM0M<x0S#I9+S+CVB}rxh_lEk>
z-GHTi^<q)D&{V~aA7GHeoc0LY(`^Q2>)NhWlVxMSr*-kLxd`K>q65^X8E9)$N1xb|
z^c@#EtSiqO)URzf=|Vp4;(Y?lXt%?m5+2tb!?B2>`R<QA10Zs|X12*%VCY&SQ3btK
zQkfAZAE$w!5k3xY!xlL+jSUm<Ii+ZKMcg2_m%yv_6Mvd`Z{h=wC_kRG{ju2?NDH^5
zBm!gxytnQDzDe5Q7wM*x4Jhs*mOa!%RHd4>v&gOt_rr}civjlUdrW|%WOp9JzT_tM
zs>+{@EdOVtkEF_01#dXXc+7M$rQ?K)Tc$pI5P7^ritVa0+pPB*Nz_GAkEP2JR-8J9
z*=khx*2c}v{lSjia+ZlhtE^klxKwfwVpwpV|0KH_p7P6>MC?vsnhuh?`acWecZT<~
zPL>0omIXZe29??-H-4Zj74QlJ{=k%V2soe?tD1;EOBw85)&A&o+d-SVLhBD@EnsW@
zipvyc4Z*+553_$Qeb$Z>SRyvfKyE}>)7%pZ=pzd}Wg>M(kLOnx>2Fk&ovt_BSF)nC
zXKz_+etrPV0D3FHNNIoiGc!W7+C@-}s%V6jdpb9J-<p`_`<P$65JA$OjFaZWlXd^I
zAq~^}Yp=(1mpaa8^&A?D2XaT_L=LbVjz&LqCB7EVbsK7_oxUp92KV0uV#D>;l*rB9
zIW<i00zwYUux<bV(3O|a`F~je=PV~dj~<hAr<%QrF@87F$2G3>DL|Tbt{r|F2u0$Q
z&Vc){9Eylm2LK&HRZqMy+pi|=>|yOK!SLrot^_7~#QDm6ymhiyr*`*_Yrxm<i{sZ=
zJg?m2d!FdZX^B_6s}3Y)JoL&sGoYFM#U;I|BMxsrm52JpMNa&koz0b^OVM`wT~yE;
zXn_~Tahecb8jlL+C`@2}#AVaRlR`DZ>m*_X1^p4Uy+vgBRD7Rtrnc?dU6$67s7GNm
zj!n1hft-4tCHpQr6ZU2b?l{vuuvOk5M`$O+|30CEUp!IfX8os>E99=Ypg|tj=inBn
z4XeuF`nRYR&IMx|R`tJybw}%?8?mFG`>%+k4_w#sDaP-*rVIQ~++SPWxh~)3wx;or
znOvS}e=g`v1Z$sHC+aTh#0%-Vs!4SF{5#Y<=S)ucxg$JWzFRX2n-BOMdgpffUHewB
zXA=QZ)Y8t!Zq18oojn5&=R<BeTJgOnf!8>?xlfY|J!rt8yjFJUt02IRAKOzD-snJ3
zPizQWaln1W$bRNNV4^wOk1?2DyZvL2K-&Y8<zv?fN{awV7jS=So$s!N<_7Lr9a_4>
zSZg^%lX5{5LA69Li^tA3vR?kwOFMRzw)JXPE>G-o`(vya<x*5?r7>^M>t!d8e&{HE
z#m=3My@{P$9f6BYsgADG$G_B!SPD$`@nJ@$A4zNzEn%`d(N{bNubgA|K`Ya$y549P
zb2ziZDdmY3dk$^9T#}>rs7IE9(r?we>=PVn>EnK~Ga7bQUsWAW50(Ij`Gal(ymOA~
zo)Cp4NmMu2&SLR=dm|pXgJ)g@XNf{!U)uM50s;2mCp-9J!O+K!*pC5F*`Vu8PgRrq
z)a7o}q6>Xy<lP+SPHsT<hv)|U=JM1x+3;p@+tk>pn}an%n4W~g<%~S>bb8BYdP0SL
z?Q>4tK0vqG;MGtaoq@QY@cfjwIVHT}FL9xY6T}wE_&&Uyb0-fJ0&dVPF0uHvcy5cf
zjNv&DjH@a3k&8uCzR=<P_BxW!<M}|MjZrH{BssRoABsnIlx^~%D>!^n|MPdy0Tcn?
zGU=;q?-drebt*+e$*=86u^l|U*;UmBQnMO4GOfw75PW<4dbaf(+ph4kMH4Uyq<T`1
zv*wcdx51t&rtn~u-&f1D?sUYtZ3=+@%L*Coym4Yj!<g-J%=g04kYj5xd3Y=3Tt~Us
z7;bId-r`RL3mvozZGI(UARvjWGTeUjn@7h7NVohj?4+QQt02=DoR7(b#kr^LHxk{;
zSd+JFEoe7svF@Lc=d}F5RxM*xC@N@`_oi!p$mtmDZ83G@d6zqYCvYL(^9Y$mMp9Kl
znp%=*O{<_=ZKJz+meJJTy!E=8epBAx6eP*WY<%8;IFe)q-xwG|4ApTl#^6?nXD<n1
z>Rcs~SMrNQ#rIBXRtJ4&J*wS?uCnQPbQ!$q7#1p$cx9~C8c9N8EUYUeDPLN1@=@^v
z;k1b8k6c^Zn%@7T*Ia#V!<;GEGk8?=u<Z<f|CR;HneI=+_$+n%Iygm^ui;ym8soK$
zZUI~o?4K7*;_;sZSViH~avsiPB!XDG6T&RLHr5)0EdD#oy82qs{FqSyaQ>f$n|dKo
zOSWIs&D1t!IHLD(Fr3*Y|FnE?(!C<-BW&67QS#kmS)HY0La26S)FF%%YVvcRgq~**
zV5_zitF}XJ9dTmWhW*^nqrX^qWJwVF9=X4~K2g(lEYF;j8{V{C!$*u0!&M`wJO1e$
z9vvMBshL`}=**v=2}{&<ZDTZ&h!>vDY^dIz*7LY#Tab)rHP)nIxMn{eag8!LXu*#R
zoBUI9Bh6rg%X4lb3CJS@as?L~F%52YYuQ$`QltwS5^3Uuf#B^x_>*hps4b|%@>Al^
z;xVQ0+=<K*B;Qi%(PAmKOQ8|s?vpz!-X<$>9-c}0F0>i(^L%tH&XilGyq-rk(sw^-
zS*qYRwJ);p{GPlRe7u*RTh2ZIRvE6Na4uJEFZJy<Rkyj<IQR2_w>8XR08c&`xEkBp
z=ipb$^$XC!FxfmqAkC9aJbt%@z!b%l^*>eA4ENe01k$o@ZUy1i(dejDBk}N%0oO?X
z!}TB73^V&-$^1`7>OkPa^2Q1&SCP6&JEd(;GrvGmByfFqPCngm?(v5uhmINygj#lz
zgjr+!q@u^Bhrn|rIR+T%4YRjpBGS{J_v_vN4endJ<21E;OgLY}aMMS>14bPmXI&Ox
zij!3_q3Kt#Lx8v2X^)Ah-DM5}$4*_>gI_cW2D!N2*lhQGU(CX9;zDv_XLAvs@t&E^
z@`HI(<{pfwk_CUE(#p~$iobsfI*h($Oq@Bpv?YitlVv{e+L9<246wQBxu4UG@zaY^
zRq8%Tz{3i-9%7hzWaqw;PnZJ1b3k5kuUN3asxN>8En7DiL{SPn5O}YAYRz?+XY(w~
zMn?^Qdt1_YdjKY3#|^j&cM1U4OIzN_keo1>#Y>hH3OYaBG7#vXr1VxA<OBCBkw0AC
zr|wi$<7nu^7fOayLdlp!DVyjV0sVJ5o7rhL#T=rY^eo_Fo&NcDVZqH0`!@TL;#o==
z%rZCnm_*#s{h^jzcl0%-!yj<3FI{_)#v-5qGg8BV9^cBF7wFtro}G7b%OY`&fWU7C
z1xQ7tRXi@UkxRT*gQ4aHZ<N2{x61RY@JC!dIWfMV0F>YWoP3iv|Dkw=V1P2|dw`5r
z?_fAId}fQIV&@|PC-GN@?F#wSu==qq5|9eZE&|s_S;Pax(k`ZRG>>>Ff(q~bV?ePp
zPX@PXUQ+VqV)6cCmmLj^H#OA)hmYjCR+fm8Cbh2Dsyf=Mj**d_KCsiA%gV1Wzpd$c
z`O&!YY7K6-G;xFj!DUcjRgEc-p7%;^iU9PgW9<v2horcC&)w=b))-WZOw!EI*gy48
z@>!bJe328zHMXKv)=*rZ`}=)K45wouG1LmME8+qI;$L3Y-u-)xw*`fx5y4WLDj4bY
zO&oKTRDuqVL6f`t`FKkYs8!(p+~(F{zXDyq4euUF-2l!>SdfT^aj`Tno7C7Ypx-mR
z&Dj9oaP8Od0lA#e@VNi961t|?)vYKzB1G&H`)gD1-SlWYzuTZW961#S<SEn<jzUK!
zOyY*x>o=d~HR~OmNVm-);tr<NV8q?73FHoG-p~(!W__BZ+{n8L8ag{ZJKd7n=<2Kl
z@N1gp?|6$sAIUArNCKX(WMi@5%4YMjIFK-zl_V&YT0KdsjYap~Al|DE%@v|W|7VUv
zt+es(IqbRStXXAS!Ncd8zU-yc&E9cv*~S^m<kE|*>$hNi_RC5Mxa?{>DA|Z6?>1l8
z23W>FtN+lG04pE{t?(Y;0FP#?0j(#&g$bNrVShAv6YJ8bPvr{&qiit<sJOpWvJ}2=
zZ1gnrzw&88+z;xG(op4c<BjIs{uEnfE#K$ci1YH&IC2jX_Sr3u1w0c44zMhg-LE56
zQ`?UZ4<I>X;TwOM*tSZzMhm_VJUL{2uVh5m(54T*SL_{KIrP%9O=}Z{W9Uk|c5q4g
z=swiZypsG#^D9gfZ^r73nVLLD>NV`1i&u&LZPKc*-jre_-zaHXse<Txepl0BvEKQP
z^88<Ahg0mx&s}m>L?uNAZ@Vz%(&S71FV7CY<}b~bn{6k(zy+0wi7qu>WIs}Fn^bh!
zFYkrcHsGb|4T|z*#w?Uj0wv-Rjbfv0CNfRr-QU|Qss}r7d<j>&)g?cDIw)D+blDHY
z0s3oelen+6ION7qj0L}Q6%u2<Av0G5N@?KNhVVmof`<Nef4CYIS1t1#5l;`UMkXgs
zKvtt!5(f-oEFyEvQ@p1{ijmW~9sq<-TKmpLRz^S;$%?JDdhbu2$d`AY071>B(F<Tu
z=U_lDq`2c%DXaQI>S%uq5IkQJ)<On8$G%JaGKB@Tbgg}NY?o@T^8>VCeUTZ#t^>wp
z2qhgd?nK;ge7~m3ZAFTuVT%L6SpJ)=WoboQ@m6d3Qqybn=Vh0e2jBfmXAlIyj<cTs
z=flDIW+MZlHQRJ$T`ay0q0ow0h2Y;CR_X{m;KNUv0YMEboWCEnUElPr%%XL)KLo9)
zz-xpVuBEQo0P7v|tZzNDM{N4Se1RhT5<!{u{xL#QO<%Z+?`ogMk$V-iSKxg@#cQQ5
zMt}IAAG@zDGb8J2zwyTD68&yhPY?`gTSUJZQpy;U9e{)QbZpx@UH3VqM_tQuHFqa0
zKf-mJNpq1Jrom{#@9HspBW>jeW=@9kJw4|w>|PCjg)&=cRc7tW8*}@yjbP&B#Gxr~
zHZZvKK;s)}@dzB1Q|lcV?4r?*<~bI$6+Y@~gmd){!2oQ~`UaRy4NSf?k<30VG=*R_
zjXjaTqz+ivbp)x#xqNq-#@sh+1VzXrP@-q!R9HglXG)3Q{C|gwb~UH`u(B#Fd;C5?
znf#YT@_W>;*_FRBHA?u3lY3*WdwP~@B?S1kw!?MEpXC<ZPBvVqQUj>W4vRnNBF_7P
zmI&};u4F_3f(nY5Vw<N4093TWNRFv>=D0kaw>aeeZumo4lS^0TTb1zk`=k)R&+61C
z0_inZa58q7nbgQhJs!+8vQ)MMW+Na?(ibl;S8nmHNlJ8SF|xmk$sEr?4_vs*0tprT
zD4u#X)O%0aGCP6dMdHi|q+C9@W?dei;Sj28b9>+S?JQ>O=(KmeLt3#<U`36Pqs7i<
z^qsAeDj)t{L-+@i3Gg5yLS2L%5Z!%H9P6R4dF?A%C_G*Ne3}L);a_s3tp~Wwhm5J5
z(IH6l&(|9<_Z1l%%7y!}-L}{N{97I`9%6?~vD>x~r^e2CG8y+3%>iNtIC76PzMnN!
zAAv-4D?=%}Fe08}JyWlh5ImA7JbbmY4)?hlt0u9<2OG7S14+I(^HrYq_<1I*{{!0D
zqm%QKnUL6ao(TXMf&yGjW0NA#UgBGZ%XDtZUN`Fq4mSSUtDz(WHJ^hudc));p))(K
zI0D0+BXGzh5fLbhO43lkCS{^?Vlar#V*j26W3n<qS5yl7A+LH~`BIgpZDa5U0TiP(
zs4&%#r&GPP*9B7kQMQGojtFDYaQjyI<#C9pwA{hs0u43;?aP;3XYUN8b^DLqpX{Jo
z^-A<401VUgexE8qNC#S~am8&n8t?FOPN3jVo*XB#qm$7uv=>XSdLMO{OywQUVyrE~
zNuVma#3pO~&ChjVS5ei>bUkvC@E4?XZ=<A$@$X&9h|lBhsA&*}e|>Nk7r&`L-C!yg
z3N_#40U|R3GvuF;qxDn;qzoN|cgzBud{(n>JhGiLFcDn`<+h@cm@hcS-K+90dKp)Y
zH^78sxw_ZpCW<e6py6nJs|3*IA_`(IUlKZEfHI`0aImu$GTIkXZlUn*eTP%N)6QEo
zAtDIy%@6X3+|=D$*4g0(ET>p}FVNZB#x)zG*Rk%)%DXV|me4at6=!URFac86{`QT3
zFkTrcux_Nypvg<A;J?*$4gQ~!9EWz_SCiSGg&zHi%1Dlua;jZ>A}`{5tZTNlJbRL{
z;m+nEIS9^8Pj(q%mU8Qr>=yV$E)N||0eIbemk-*&q#;BZ##c8>-a~@P;CMlA%Py4c
z<6o*KOwQ+@E2%hCp^H?DVCq8ErYX3w*UXn5l~fSg>(C~hj^KZ*qLOI1i$!aA0vYUi
z5k>ecw7hF$`}=*j?lrs@KwbeJ&%<G78(|)If1n5ZibqTIaB+c1>JB^Rsk5Zs)%73>
z8t8KNmJY9_<tA+IlR=$<-c^zAwl5;)=RGq0f0q}D<%rt4UR&<9x3^=G&+yJfw)3yd
zNMpuU2L`FcB+1hW?e}IBj#Q$-5}UGE-c%^95LLf64};8C(YV9)9og;jXc4o26@}_^
z15CJYP;C74gO!@!WpMFIO#A|92cY-74a?SP<B$pd+$@Ev!TdXHw$~I}Z-Ftq^0{G`
z6P@`R-mpA_Mn(Ju*v}ej@0#46BmHQnzdujv@-z=ixj9{0J4&%?Zf*Z@>lk>)k?X0k
zY>F|?471M<l*GS6tblK>18NH30|5`84zcUENr+`m+0c*ov~=@$$n-3p206YA2Om9|
z=aZ@=4U5$9n`yTzx0i7qRQ|aiCYO3JU*np0=WcoIV8V;UA3*t!a=Cw=M}fbY1P3hl
zIFQ_nR!!SL5W>#!8GT-AkmSBoQUUnBP|0X+dOQ^XgZC-7%!yE7GopZm+)LKAt?bM=
zzRzhI3Kt|$r$*|WDlYY13^b_e2Va(iZjj$f$Zxgj57U4p^U->K1bi*+zwBaKW{-#e
z=x=2;0f&yZj4zSWqyp4nxWwxCEKt*qr$hz|PHTkGW_k2hGXelo2W`_0#AbZPo7c(K
zU+TLcdda=Bv2e+@If~`^1lUc$m*Wf50X}D2Lcb}H{wfwR_o}LI?u*QI`0{A8CaNur
z0ECCVo9E0hlXBus&ZE9GJiVdM>0sHz+5Vz)IZQi3_kH}$m(=-9br3Kkho0~94yUwl
zc|Is2lXS4IY3X+$3jA?iOFeg}5|FsJ>(vWUQGY=K6O1HNAs8B`9yJa=fB<5EXsfmN
z<_^)p6fN^?P%ih@qoA%0C)*jhNT|^i#G#cILenfTQ8D<>;wy!*O^P;XrvE5|q0+h^
zhJbAUN{T|{*BSNPEk?)Y?dL#R!>5h?=I1)>irPxuXfFeTCvFoKt-{+G^K(3D+a9D7
zN=XwTIYMjdS?`wr1}%w!;2Vyn@GTH(h(i4OXE9Dt&N!QwnPQJ4cD07Qwl<*jOZpev
z3k4P@8deme>4d`{TTA@}@k;T%gSrGQnMB(b1eqasUG~AQu|5^;f;S)UYBW#&fz1-*
zBcPe^1qW`~P4vg4#}qJ0iAyrbKx8$<nJcP7K&tK;IBTd&0xXX2sQ;ZG(J_*Y*T6(;
zXL2)Sq7wHLtHSwU=BcUOb%c|uVv&?NTnejiKCV9ZnlpgmprM>BYpbu~h0}T-u!~+A
zxwcH0uQrW#`iIS$ZAAEuxv(crO%&nb*(vz+M<32jr><Tn?jSQOzoOjfaRN0@JR<B*
zgC(-h$s=@OGw7e2F<UcnA_kO6OM1a_D>2F<$^74^Zu=041@cnLlEDn#RR#1B_4u}O
z<l}=i91czRL8u*)P#p&_V2l<xzsQ023Ahubd}lW&_^Q|l0E)LLa~S6&L1`2L5+fg6
z3fLkZVpMcck_iU}VcXrU6@b>i(()F>iBM7rJDfqTGPal7@Y&+nLj$sRTA@Bi*%I-0
zqo=L7qvad<eZ$T*lm<*c-nOxBl#4mBkC8*i7e6JP(2hm&3;OQM9KV(U&S2WgrH`0(
zW^WuZe`-v!G7>8dj314NM@U{`BP(JQh%|qdwa_w^38Lj2^7h^KkB3ZoC_f{&47Y}H
z%oNDiVdtqrYy6cdiix5&on%|sY93|Ti1e{?p6<g-V)avxRU1u`u3x*?Hu8?TJ7i}{
z*I8m*-<5R`aj<rQ9we{G;cX5Uc$u;hl|wom&v5G!x>nak`9S*OM7M^7KuV2&FDLs<
zr4E{m`|}J1!7STbF6Vu|kxV4N{rJNi_>w|vY7<mb&wfX22%5K8Ah&0l?=u{FFc3Sl
z!hL{i`78717$Hy`E7uM1!I^5BdKRobS|9=W4iqp9K<=4J+b0}JF!~#YK;1QJCiLS2
zww1#73UrL0S?n;Q?Y?gGWvqpc5LpDIl7L8SBQ^57wS~FI$aHL{)Rzr1t#!<w8Dhq0
z;|Vn?r0Nj9IXVL}*3v~n)mvNE%nvtz=-Hm_C?kX%+_?m~_*-z^8PgfO>+@nRd*)Yx
z90H#QNUOUdo{}xKj@Xt&-{@r^VighnL|Od}8E)rSmqEQUsH^kabw-a0B3a2dvGsE;
zs@5k5MFj%BqbmZTD7y8PBX@mUBy5(PD1<XpSKp%fCg)3<;X0lq%<$Z{W=`UADNq)K
zJtV!R2|I%sVEIx=<8is(Nl576@lBJh;RsbvHJJiQ-nh$r;YB7pRABgNfy$eOsg>Xu
z+KcidWdaqNcQ`2-O$xT`f3y(z0SOXNzk-4HK1NK18Nc#txLAu1Pk_Vy3Zrc7dd|n=
zfYf2VSp4x;SM(D1ifmBASyV7tz4|XfP?u_`Wq!)Vahv`6s;E3eY#8RA<b2(~*p#4U
zVk7H;+1D}KTHOe0DXuX(Au&23I~*NV6c_{j-0>3H!p-*^=#`VSM$8Svh<j1C<&rab
zGb6sFyBpAbu;b-3qQi6`2@0i)PbN~+S4&$P4gd^sU`8DXw`GrnctAR+SJ2EMR#K+*
zykL_pXmt-7Ox0Sc!|+8<nCMxtNp_bM=oD7)aBP~32ni_y!8X^yGA07Ll;Tobsz_Mm
zg>PV+_(wIz7QfiwD2Ja_;iV;}_7|GyINsf!CWT78CD7rD&pp#W2YP%MJKwLh=HH_e
z?EA{Hl$4@&WA&$GGfq<6E2!8~FCX%uXnEd9#FkVVbom9g?g}?~TsffmBLlB|;0@Rc
z(J={w=f<}Z5))6jVXO$x*P@FZ`x^x10e<tR<8T9@i+JW3=?_N;2p8vl0E5m`K1Pi@
z6jG2vOQN)p$&qSJ=D%WFs<iFtuZjBndDfc`DvV!~Ygi91^E2};`uiR48(MW$13h3q
zde<#nG)IK20J^_Gk{OJcCqUPsQk_4r<GFbkFiVD6F&y$}@igNP|Do+)G3y$Lx<0aE
zZ+N3opBfQM`W58iv4NMV&SVn+>%orgzxCiAmXIBMS*6bSewWB<F=hpx2)C=jwU?Xv
z)9W5F>7_FS6xatO;kc5SEssz?o$jIA7X8A06iBdeADx%A$?j>3*Hpc~x7FU%iXO8F
zG-qAcq6hr@A4YaPINLpnZ1$bhC3?4u+F$sM-uoxNyUqbY9&8M>4da^PIsViJF5eZ+
z$+Oi>{>v5yKBuAK9;*!!1TFdQ^Go<oT>7epBjFwKGcy*}Gt>;PZXDcfF57=Lx+HZL
z0#8`I7aO{{-%P&4!vh)6s<~Xjnd(C6T+x|z)<+9Egrn11_pS6O2b3vuWb60da${D!
z4i-sW8~@+BDkEYs>bnX=+{mzCmWR3S;yuoErIaeeV(XkwvJk91!N!rxO<_UCgKGP-
zL%M{(JA~3c``1;t@gvDH)-vVaYb&kT-HN>bguA;q=bB3mVUxGA#g1cxZi4FokkW6|
zcl{2+drSk`Bj9x;0Im|dCe7C-Yv}iInO{<ce13oD_5~ab%?#xAY<+XbQ;>2+D|~Gg
z8Uob8Vx5l_r6R1)e1V*;4nEuJCN{9Iuhj-&RE~F%bli(Q%z0V~YhBPz6ub$A!95vh
zOH&!%*uj87fLQJ`VfaS0ojgC62MMQCz4gWjz#8n;{4k^ft`Ukl5oQL@3wM@kepDG5
zYZ=!!D8vwmQaN;qKngf4DaNRoa3*sf=!Q!gGV}i?SJHqqTU%k6SA6P}ecTpb<MMSR
z(a>Ko*sm`Xv*Z^x)ZMh9Te~FmsSuw!$_;<0!50OrnKWOkvbCrB@5z7+vsrv*G5K4!
zS)^3`^W+W&n2xIrXW_n|l+^n0_<*qRU|yAWtQ#<C$abSYbdjf5KG=A&v}@;ED=EGL
zFaI*G<wc8Aoa(yMkwbqC41ZXNvRN!U3q;ryS;$c_;mC$glsofe+r<(Vi|Ln3&1+l!
z%UP_@$q}CE!pexDzPY{<a!ntR`P|DHsH9}2%e}MPoaW#T5C4%UxTtd7Zm@u?dP@%j
zji4edd*N;vRW)Y!x+ORVca$|McdTIg?#+$YR$!5%NyS{VreHutsnX%j;xT@|B>N&A
zn}yReZ#Ca@|B&pco{C;#n4bEIOvYSzhNQmWfi#a-UEi=n2&Df-{D+%A;gIstYlABJ
zj=Ghp$eY(=*7p98SljFXbEb2L9Sx*RsnJ9+I7?b(3A0sXWC97sJESY=tbkcB?b+QI
zFcCXF(Q((nx{fjq<U}tWYuNEpR+Z`-7-}BA&J+L@$P4Rw`dza@4##ezz30~GCZYZM
z;^Q$;>r&bwQjufEhUN;m{$(}eauS1&6tKVub-8$7;y!<%9=WmA2aRPVuA?cH=HKf<
zLVkt~gHox~P}kLc)^b^MUT|K})pwoK^_|D)D2T*ME+*`fMC^iuB_hoay78DuPrr|q
z6zVf)rV3+PAn+O*<7Jk2*c;;k>2}{^XCc3>A5c&%oiX6zW;VcRgAWbt(F(tAa^C=P
zZx@eVdmZ|h!Z*|SMd-&D`ypWvW45>NA7eV)QW>>J^0KrF8L6?Mz|i#iibAOhK_U3c
z4ogNFIdyHi$TJxut5w3BVUM1v=BOd+0gQ~)kS}H3#{4<F51)L8yw)tAo3CD18;&{o
z_}hb)fQ##rzSvk#no_>=JP@gtx5+4_p6zw7(bxW5^woA(>Cr(ax>iZ$cKfxBNK3Um
z)LZUFylWRScvCEI)UMz5$ydWUjW^W}&mr@3CrDW<0)^NukC_Ew5V3{DQt?UX9!68Q
zGkOhU$;TQpz}4fu&ireL(7)DAwpD_j;}!=g5wZMLCvM_8<6m-C6LwB`*4ofHe>V#E
z_vdXfrx=Gw1_@)kz>+X$yKMzobhm+zb@8-Gk88a!^;l;FvtaXz1|$Zz*7&BTBrgA7
z7Qj~sbsfoEAH&g~LrHnD@^aqL9$$&%(Cmy_t8cBf+X-vqoBofHyJI#{tC(^l1x-KK
z_Lb=U6?(f>cm2X~;yInk#`W$hR!`r9gVm1bHbQ^AZSyx9D>ioWL8d1&KOe9kn;g|W
zMRU05xpc*3`a`p9`*rPZxZKoHgbL?jPGZ?BP<&6nZS`vk6tc(Ss6$dk(z@01Tx&^5
z&n(!J^;<IgHjfOyIby``L=}3QW&S?*knLwax13Fu#^rv#H!XN;#VK=SPdaQNCaz+2
zhd%f+pOWgy*SFU3_I1)m!bv{o?8<U<zojm$-@G$WEatIH{HLLz;dP=R449SG1UU<$
zv*8RA=qA9>jqjacIY)PZctAugrH0nQmuKmg0Zs0Y8!H4iMtoY3*O|K@$aPkAs0V3c
zr=c+NxB|sP#OdEak|fXUWqV@#s`I|A1PiYk7^>~~n^U2>nt(Y~P%ueB;dgqj!9nCg
zR}n658ab1*aTyu|o%}YC%F#Ro>UNds`(^z3qwwBwh*&#;`oUHo;%evHy_fTZbWhmd
zdQA#RJetUs+LkMnwSH^a={=O9AV+tlYb-CJ6j^nOPu<K471Z`1w_$0_(a11~ELZ+8
zv+b&2>tGGCgyjD9e|LDGi`&axz2_OBFYbj4+d3I6%Y#$oaEO2HSJuo{m=xG6MftM=
z&+JLI(BhwN#J_su2fZ@|xUmUNI`cH}3+r}Fd#Y=~<(%*3ovnsk-)#sirn}1XB;IzX
zUcalD?9mg7Mr`WCEeJ}$5|cN-y&4%;&!Btu1O*ZIk$|%y(5tHG9oZ#mQc)~o=^=I$
z&oI&iTPh7)cF3_O9E4d<7gcpH?2%)@Ac27`cXx@=`#~q*DJ`6ZEw=s(r)es<Q9~1X
zsp}xlocp^Xv*PQfN4#GhE)2RvYL~*}%GdGSrl-R$hfS}uhv=d?I%~zU9wncR=Eq4E
zz(*z~LTv+qsX5&e`kE+$kBX><3QY2`ELBkV66DnIiR!lC`2G}G_Lo^@)*EmK^`ltG
zzI}=6hOo9UVyk>0=z=HfQw=}4z}X{3Zy@bXI##B*@Tl29rPJnbiT?PJRjz{oV`DQ#
z^{_`(VU@SkT}kn$pj)TRL=dV8lOyH$DZ|>oOrSR2V2J)!$2NP8E&B6IjQ$#;oUM6m
zNG09+!0y*rcU@QczeI1HzR(0aAk+0cUv}!Rn8RgZL4W+FB@pAE@7Rr9+ozgs|2FJ+
z-t<%4I*ID=Tpe7;iAS^Q%F%xgzkhy9cU;(zm9>UmEg7bGP__AD7!hD3!kI;x>M`mw
zYi5JL_RRFJgvbfPB-PenbI{lLWp)q+b&vV)1dC2aW+#s0y^1hf(pfVEm?<oA72Y=<
zl-1h&S1LEmF}SlAm{Lk{dU{J2;GpP5?LYB<CoQB+Y=f%cQ1MQp<jg{4^loJ_XZfM@
zZ@-i(NZ^Pcw(LD`FL!j$G_w7tWYzVCxaH(B!EKh+Y6IVI`(s^G6@&Y*V@O|+2gp=;
zCIt2$e<=T^dhTg~V-YG?9EcOkeGDop;xv?&RAnuzF&qP~-VfG)wuS$wot4k^9InRX
zC{L!o4`;OANS$D_Es)^SXWeTL$s;F!YJPVBN$oN27|w{mq#9lh1hTGN`aVjTsvrCu
z*b?v!_G*aHhX%Q-K7ZCrp3)Donl|g6FFGYVxZ9WBLzENvdibd#T&2Mq=A#^2=O6(b
z4oLHVC^fq3hd8;#w1MD`_YlBDw(L&R*cxKi`&-ZL(-est=J-8lE@D`9VYOcG&ZBht
ziA@%*6DZ%PoV$_C<IfuG?ml-*QyuX0=>9V@XK67@wlU5sy4|6S>MN1x(CWsf#mqh)
z{|Q$X^7|+E*+nr~5@3Cd1PyZ#Jm3)%QT^jYQrmTu9O(ZPh?JhpDYO2<uxl|l94(Y5
z(_VQN>bHQRI5kuls7{on=7a}Y$lZRss1lp-1?wtB0cF+L^*q7@W16N_hGtImDG1)l
zG5J(rrQp@4J+Z2F#uU+Vugbw8`7&KNL6Fr0Mc2}iZ1J|eU*GCXNzT5~3r9Z&Z>5J2
z8|5aT_Li!Q`^vFZMe-Q6nroxx`puR(Zb#E$9fdR%gYv8ICKhdxAPU<rMV_iUr-10!
z#;Z@YTz9Q;);k*;*qkcxlxkjy><rXbx&XYGKDWJ7VK`bkr6XA3<rL-mZ{Hf>5x?sf
zHptb{uIr8Xk}eomnF6XS<#!yd`Qo?EwjnGTu#%P=nL6R!t6tBptan7tJ?fpUapXj8
z!!CNixv^clxt)ZW&~JsaCHy1F3y*Q4+?=1{VUUH}PNbv_;e#}ny1A(5aQ+WX-yBfq
z`~MwQt>s!RTZ_xKtyQZ|wrwvi+qPSFty*@SY}@v8@B4Xv=db<<r*mKT^?vcXdoNc@
zS+1I;sFM{$2sx(Qh%)5H--#uKx59pPiGAtD5c(4n$1-aRD>y%eppZi}hQc|osy^Fc
z7@)12^Va-{*T!+an&-Wi=)&u|Wqn&1Wq0fCJRFkN)yL+bPt45IHQ)G5oY7iyof8ju
zrOypiLHw@a3P)wK7l&<L8l@~x3vM(l|4Q|6&a?Wr{pV~2?c$-QY){O3d)?`0N{S3<
zKW1ZGkWb>KU=q1q$~@BB%o_)_OEO%;NX6)8--*Uc>9zIka=}|b%(bJ^7vJ;uKfmqp
zn^xd+P1R-?oOh>lqD1H|1<oe;d+cXUizaxOix@q{eTPjYeD#zzJ)-75U->*^7rYpa
z4^@cn?EWmnb@-U8FaCS8E;=H>7^l>eF#$fAczTK549hG^dsFp*=lX&!!;}yD7*Plv
zT_oSLrqQUaaC^FUTGL#VT@CG&n>ZRf=xO_Cqkj|a0A>Jf=<98I{R+MepSTmkEG#>_
z0Nc{qYSPd8rKB-Q?bfEP(2APXBhRn3Wm6L#Hpp4gYnzW+=Z5Y|uo|lFK2fYj=UHbC
zRNkaK&ownsp$dQOdn=nwJC_GHAF`1jjY44GL7zZ<ybbU4YbUR8GV=tutMNEkCm}$4
zz~zo=fEsu&32QLUi=_2@ZIdI?B!A60S0p0Ej;7kg2u{`3yASnC>0M5Bxk;?2FCqw{
zTdiV!^OttKc57+>pyf0{2|BZ7Q6NmsRSQ*@WWFQ#gIj2?#NuP8+gO?BWxEWW?OG<U
zu41^ISY1V}RW@6DUzaem1T%T?c8?Az2OLcDj7@XLGb58$o2Qz9A04JpK|4K>R{OUO
zSm1de^b8%@H=RhG&x0CA(paXyJk{g_>1_(Cb)+`|pN8#oFK4UNc^~*+1Kg{F>uqbZ
zZXu|s4$LwQ;*Q(XE21>|K6sw0WwTyBPWHdc);#zJv)8F1AaZ2eKYy(x_&1S>1m?_m
zm9yXa1KIgW9)=XCRDpYLoW;1z8c3^O__-!IKnRn=aTLj0WNa>=r1-_!C%R(ax7p?p
zarIs1mYTISZZEYV46t$J^`*v3ln8)63g%R42QgnN%lH%5eFy%u#|>Q4XPceowiNbp
zPKHpto`$d6@QFX^xaAxqmPhD6Kzy)|k{pjb3Qwv*qjyB{(D+Kq3pLe(aOFo5+4+UN
z7|wWj_JEC{J^~5y^rkn&uH%X~D*asaXs&m`Ovy#gwn#QTa_sQ$2`Fdnt(oolY5VG0
zvB^eQz~h4xL2SlPUcHZ`4es5_+D+iiUemeSMuGST?+^|C8|SL^J8}RQxh)1tB4?et
zprE)k7Y!NShJec%$32PqS?*~?By|5snD60XOmNr6sC)8<sW87sWzOvh14Hpsp@rBs
zWtq-L13q+>r~#PYr411ZAE`IVmdLZ6QY<v<a0}rKWmg0Z5+tR6zWKn?PucES38y-J
z{jLuCu>xUdq)2(Hi9>!&E;QRyg^{7>b*{YMt7CH-*h211^9bDFcjHP24H-(-<((5J
z93T67cr>&dM!~N?A)IGCc4Uf3dn#ohRn6JFJa&N3*P0cVh{vFXcrB$ZtLl>g@GWGU
z64zPFmT-5Q9;X5TdWI?0MUVo<2=HRX``@}s%9Txtu$++3#$cEE@x!f;Cf18<Lt`$%
zYX3Sve5hiyo8&XBsPpzyLihpJ+S<PTY+AUwBme$td&}+Ow$%8KBFfAGj7;0g<f2lt
zI_41EF|Se^6JzArw2_f1+FQKPx^oBs5+6{>`-un@G_s|g?iy`t+y%tW9=qVh5qUq6
zq2O9*>UI?ZRr%T!|CS+T2x4Nk+2(cMnZbUZ#8$SLqOCV4N<Q00%5-o(Gnn(&a!^wG
zjDum7vDBux&C;7X<72@6_1@+FE6YR4xk`jNWaXaLH08C<8Yo+y#MbEz-<M)b;JwG&
zLLhAHSn12CP>Tt-@&x}FVR9sl5zc@!M=oXY#-oqT22AHjkmnu`C($DrDK{+I=_(=*
zs@kcwm5*`-w|<|LXGb3X>A<WR;He}h0}p*c&1@r$#`eDQ+_2w%d~&qS1RnjMm`o>1
z$c}D&#HyHvMSO;%&rsJCb%xnl{tl*3H=lntNkZ|=JXfgLUf!nV`jM#6jNkM7U`<vb
zQ&KbZ{Nyf4u%MyOe&73`kmB<$(Z)rtr!0|Hm0GN7qC)h?>%~_4C~BRPOX9Aj%jo#R
z4E`y1l47cwGl=W_R;zQV<kX2RTd{os>>*m;GcPUg^*li*M7hnV#^k0WHy?XRbYQZP
z02idyf=hGGr=&_)R6KU3oc>)}U|4<ked{640K!9}z|nGvVr5=`NeZs1Y|F>Df5?bH
zw<)J;JM|iqlnK66T^AJ!4bxhxsDGG8V`Ry8ToKL*iH`?is=WH}8k&xAc%%PRAy#FT
zGv!I>&FiiH9LeO)GwgaBR*Sn>^`;%fo9&C}|CD8hd!|9OB`YOD{#uKxw+yVt7tnt&
z7zaR)!}WL(zqD9>LXh@(v|JCSbVfulaZHVsf8B@!X13DY55|Y)@)QoouDxxt^-HPu
zE=Q@YgC}po)z<phqjPK$RH>oRf+drTvatOT1w_6{*aL~HNf#Rt?Kfy(&k!WfdXuVh
zISHkBE;7-3>ZvpM>LFzxg8)@FypRYdb$~kp?~3}YS;d9xdliB;C#|hkpZY7Eu&44l
z$z-UMf4Er=<+sydIe;+dD=fbX@1^@s*kA6gUl|TX`89I%Weu!m8|Sro!PVY`Ijo$I
z0f@Ewl=#n~_{aKHo>Aj_RrOp3?Ldb|)zEsE(9n-JpB(RhgI7I7<nhw8Kb1TBw?=qd
zKVz+AdxIs*wct2^qc|_3P^*K&dCjQLQqN&4G{4VR3o+8N6+EVGeD2H}bLVZnH|A}{
z{T>fGuv6{5SB5@XD~Ia-8;B+=i>>cYNADjRC4?XoBGPS*w{>~VfcV$iJXv|KbSJih
zakJZ#q3~ns7f;^5)*TWX9>=OBH}O)H0bXxnlF*v-=H`j(F3nMLg`N)&p}rb}oD!dP
zkzR^-vp%xJDmP_Gu6p1K9B^ODS8S$LDkv%~{OI*H6~pQp@d`?PL`dN|fLRJ^6k4q(
z<YSx%JM%hr_AJ&#D~&C8P#pym1rU)7ii&G__G@I6r)7Qt|6}so9DIJu3biUWWA>C*
zUNK@KIyNmP+x!5E_yUqwn~#Jz%X|Ev0?_}DcpDlLw!#079McHpHR6nFBNVi-IBf6r
z2N(N}AeTYftoHsRMqKZB%DHNevq=KK3;pGiFI*&nQXM@#dF;o<z@)CK?MH{C<m;~3
zp((QFb+<)VN?yO^$W|AMGuQX<<Mnrx`*OyO$34dzFJF)DJ+vfeTP=Ai_2t^3dTYr&
z$gD%V@=~-20kmE*DqJING4W@PM#;#%y`)p$>*JT|-pZC1_h;RJrBsq}w7#@W+bfG|
z2`PII8A%GRZMVXK-u4z6tfA=ZE<){r>q7~F<Gn!vAa9|}n5otEv9}xEkDMgiH@4I;
z#(1aZ3#)qsJ9{x+sTK<GFimq$_esTiJ|6oCuL0kD`6j%HVwTI%H*kVcZlh-iS4bAh
zQV8X>-N3PB=DYv8tm_WttXfg#;$PaSzdWN|ECj;2Di~}hL<e?AJmTJ?9jC=pJfxZA
zrgQ*ya@2h`Kfz@S_gq#Gske^4_FpHtCECF(K$EhzTX~>O2xqjMzCE^|%{g8C)oO#Z
z)EI$fE=mLc3+4GO=>7OoD=wo(=ed|X(zrL&73Z__-c9`jS6zdFfV#~iG-la5Y^T=9
zy}f(CVKXbolG^pPLdOQD$C^IGPR(}2^S1@j@730|*2Yl;n8s^Zs7iZX#SCL(a?9^W
z^N+eATK9I>cU_ugp^bsdL5-$VdvoYe*847Go3b;LjaEH;*90I&{_x|%{R^+VOPqP^
zwf_G8z2pdONB}%x<(;ku?WZ$#uyl)vyfO}M8l~PngrmjIQMRbLQMn@H$+*Hbluxbj
z*4|TKAu&nU8zxFJgc0?-QjKy!Rml^Y4$H6po=b2PVx$+!d}bSV=3raP0uT_KZSEJ$
ztQA;0YCPxLPr!eH$OooOst-2`$4MFb+8sHXV-6tOpUZ3;E=++H@*LME-6$_hh_sO#
z0jHU0&TlD`b_I>6X&iH~Ufy0@i?e%Wly~a8%+-2qDslbn&#h41e5X}gSuX0ax+;Hy
zr&kloRh>S`82@l$EP#fy+pqrS%m7R!=MnLp&HOYz20X`uVrga=*8Y-+OE*EKm<hi1
zNwT#L=q@>FC&j9Etk@W^vPff%xi`K%fE~6TohZ9;F9t<|$>LP(1@4<R5?O_Fs&p|H
ziXe~53I}^7(i#QabB5PBvjYrRvJc(@km3;x+<dBiTZl{HRbwlI$jSfda1p(Aa*GP`
z_{I!4Jt(WaJQvav5>QK_&2IQJwgF|*^)D+P@M5tGMebvse1yCR`}tLxpU(P`6-v(`
z@;}k`FT2}fNlEffZsMMq`(hqTyToaG2TF29=reVA!CqUZjpuvn7KuB;G%SOFQFh(A
zdq;Uggl4@AylKTE+1~5fVjsH5{S<)UUr}3j5ENALDaxr)<WMIS(C=ea><K2eIGO@P
zKYg3Fnfp<AW}hQmM;V{@`Msb4En|cyV!vEr3Wb9~PTA4%pM(Xd4+FC^-HXf}O;-P2
zUjM0k%wQsQoOm7Ga9iK;_g1zJ<g1o_vTW0F`?c@<quI7gcOgAUJF}m`md9`T$E$P<
zE7+`-JmY9%+lL%5^;5L87hX)qMwGVSK9p#&=gGEw>gkDe{ER=qwryN;HSZ)Vq22h=
z2?@Y@dsU8~l<##=32j(uSaRh_QE-<Ej@?K^QBQSsLnL3iY`XL#4*sq)zl}vR?cP2w
z&2m0E%i|GpCJ2|P9ivVP;W+w)->Lu5Nzfni(U>xkfrQw!C!?X6#Dp<0@Z#oB7C~>J
zJ6sW823L3|^%oK?ptwM6)RTM;SD_@4hrK0-9~g7pc(h?S8m7<*H$}q>0T`vSt)qhq
zfck={=4x`Cz5rj{{AOj4!WU~xQv}bBOfugKYJ3e0A#*<Z$!lRX0`q2NB`-wCj4v&7
zLH96M9SLqD8pfQ@J6-eFF(Bp9nwRUCI2w!I!9!=++$@R)VD8qV0oK#6?RKi3l$fPd
zU)MoXiJhT)=lfwr!a@1X#&+@@i0xQdWF~=ynLikZP0YuG!qR;H6yh0i0bQ;g$E=Z`
zJ|ajQtyGH17q(hZr(N<Yadi@OFWu2M*VWmryb9Ls;$V`n&@>X|n`M#fIZmPw32V2Z
zpXy<Ozmpjl#(LZc;j$G;MH)?zM>Wg;zWB>MbH{T1#jfqQW}lY>17)@&11Tx`1c$Dh
zdN?)n)dGSNLz)#ACq*<r3UJR~e3<A=mWc94y+-m({$kz8nzSpHu0;XiXSM3R^Y)<4
zx!J|k(&)@$)3hApU@0&78>^!hs;oTHh>fR2513&p(FFfv`IC;07o%?xIpr3R%F<f(
z`H$j@MKLk0PGbDFE9O(hYic&NztrfLNABn0Ac95jcRD(K_vTI&o{^wLY0<p+HnJh>
znr?<atyI6^b!;^2;++RTXp8W8mFUcXuWwBuGu|5ph!v9=#R*wI$Ib=H-p8OONF@W~
zPfbz9eb7#n=Q{EYiu6OC+@~kBcxG{7;w5Z8OiI^|lwT5L)G6=qk0cCvtIXyM#j&?2
zw?b5<G8~RqU$Hnh8!3?#5AMj0ojxtS$tIldZ*nFWf0ux$`b)kOYZwuv6GI$)ra-h<
z?LBfQ_8UIJEFLP%!00x;M!|ZyxlAHSiBdO;_YMj`Mu1{M{26AdkpOT(`dj2K#x5}d
z6Tlr=oujI=<<1Ps_VnbZi1mdRumSVMML%uRB>YMy%@*hs!<q*MBdM+(TArgim;xaZ
zB4NP#z`E(&kJX_9%?+!kJwgp*vYZB$5*lKu>a<zFt-SL~+@!vT4%EywI_KGq-RL3Z
zSJGeX5+eZ_@C7>F3P&l~Bu*dY_C_Elm?^fIy5eZ$hq@)av-4~V9_e;if+7ps7MYIM
zLFpJj;-na9`(8JfIf-7p%Q=v3*jSOg-#D7p7;Wn{fH%sMfJB?TmEv=Z<BFoG@A(KR
zbEh?OM-m6eQC5M}U#j`0+~v>dc`G!s5&#SnZB;aQDM>z0tvEJIr^p4)gz~E1>!A?;
zQ@~eDb^BH5n4TFvr|(>4iG_w?dBav#@eQ*8lO*>5xVFv!0nfKy3}yZOOG>&d4ES__
z8&iQj7z^ra2f|elhR@tO%0DR~sRtQiVIf1z4U7KU{DV3iDe1?@XM}@U)zIG;l=K<H
zyopxSt70|98TB!uz#<bx>t<=GWWQFu?eCgF`igOYw&pKz_RJXmX+D4U3?&vHG_~jT
zHg(Iu9nyfr8@!mU;rPZ;pq8%9$m`$8(VU>Pim0jG_Gx$6GTlka{Xz03&Hzk=_0SOO
zaVON0XTE>$xhf=bIr`q8Pzlag&!=N^zxff@)BEfB4&f$EC36wV{V#RGNR`#piLc-1
z;&aNVx6^>+$MSHj4GqrfSW=7FyepR-HLTrA^7l5=&}iF<7WjbX@>NWyOZ#_?sh`$i
z9da!8+_BFn!mIUb2`#)}R}4iooLyJlZ)+3j1C=~kcT4=2qKN0#=Mfj`k@?5sE{(nZ
zUT2STtu8XT`}=t|AkzaTPu|eK<OTe%kw<VmoR@XtM=4QG$JIz7#;Ta4tl;Oz)vykk
zz#$zb4>1^k(>!IOP_%rG1AR(J1~x>i5A?t@sCr@nVF}=w4G$fjB~R4AJ)D1p%SXmN
zP2b+J>lhJEG+gAkwjZ7Pz6;E#qj*vy;P@5&4--zWNYME&rrYyoHh^F=Fh8IcqnapD
za$*E%u|A0|U&#!7S}AV;>EfIB7}*K$Bmw~enrJvkATGL*emlaF*(xapGhBbw;zim@
zRdvh97Iu7{7wiB}Se&Y|f7yL;P&TDtbT6-RM@li<1S;ARJevLDaogbZDt^EX=x%0O
z9C*`Zae?5vCbW#p*b)*H!eHiGT)Q75vI_7ulw8KMYpijBtZtk~VH8E9&a+-b^v<Vc
zMKBN+7Is}6y=~ckRrv%9Auf7E!1`091Q|@Q0Mm5b;snAka%}Nu834$Pmq5$e=A5Q!
zF)%6z47<_KI{#FKW?a@>X}iA{fU$E0Kv-y4D33r_J;z0znfYVF&5jKm^ws!t*e=Q3
zwuD#J7HZ)lEjLe6xr3FXAS#RWW!O2EyO7XGmT|{H&7yphsm@C=7(fjN^g|q+l;``b
zQ7c)?5MGN4l9|aK9J{NMnv&yr*mIZ-yF3tYpJg4Z8TB~->!i!6;K&%$S<AzRg0)xF
z%gQ>p7HWM?M#87Rao)t>$YX$!%$qIu1S;E{HD!|)cJ~xpQCSYVTxA0yYx%M_;NmTb
zM<D=U{UdR%J&ke`VUe}9Can8bp_!*kp8AKXPE-KCL}Rb!&5xH2Z_l@7-7XR4!Rsu}
z9va=sLhxR9i5+o#M8m*XX)aV!1&EU{A4PcL2_5j4W4Md*W#4BNSlH8(#NtrVIw_<?
zt!8c)pml0RsAe2+mzFRU338aLbFJmsqjX<kooAWrM^|sJv`M-8O3rZk%fte}c+#$y
z7i-Bs?8CmV{wc8m#CFL3J}t*6XO#0l`_Ngrqbfsd!y;w}PDqv#8Dw@u#RR)>%?2ae
zMrwPkOP!@oC@jbmdsQwf?xh30oD?C2m#oRs4Qa=XZ>b2;LAk&I^X^Agrvf7oeqJxw
z<F}R{xruJw?0g^L$DfH_HMluF1GNx9H=zfiVBc@kX^CcV0pC$R`a3MDR$jSI(b8PJ
zoO^FCZa{zSpw07!Q{)S6{s8RCX><~7{65(4+?+G@uVR<}`&DB?opw;bkQbu;CThj&
zQS(THbUN9+q*vZBG5mTsN<0<_bF_7iW`F5&PWR90$Nf+30<@ddq^fVolT$Kb#}rzl
z4$V*4`?k#?ul%^GnHLFH&!g%ZItktC(Aj+?UIIV=VSjFjZ`5UN6IMsuTLb`usxX$e
zO!c&QpZ~D$YxKP%BHD&+rUr(YtdL<9nAgy%VvLLufk6#bZ-q0zw$)uXeV(1)<&-l?
zIE&WzRDD+Mx>)t>!WqYJw@ASav{NW2BGn}T6DDoF)9VS1y@@^<p1ayedj|A!dm)gY
zz57bZb*D|exw93HI^sgO#Fh7ZN$YU0=Q~PBfgM2B8kSkv#S&*4z46c;esZW`UBXMg
z!ooJx)dag75HB?%=08a&#V9Y6BH=3&MWEN%bJ}#l-FvdMI|o<-3j9X%?R*i1_vqDD
zt^zL1_hW2zY`JHP_;veeJnL%@kxvIsdB<*@8OE$C_uB*0wRLtAZt(lw(cpW<n#Wy|
z+XJNa2?QH6G}`<6mf?0@-of4f%K~VA#{{u)+O&<hRjVbC2@89};zhNEEq-qdv8W6N
z@Z*!13w&DUT1<5m{q~ZzZ~y7GXO>dZkQg4(MuSeYy=xooXmnQ+HAM!=)E@aahq7pg
z?*3=aGny;Q>;BEfW@i-Jd!L@GA8Ea#FEBfE_{E0K8#MBabdr6n2u%piubTDRJc#>d
zk{^|yqW~U5GEEnSqu6N8zG^BpF(x#@6JciAp;<tZXu)<Y<7(fdEgZ(;pF72;#Wy+g
zl`b)Ns$)UXf<fspZO5Kiikf3J>Uou;J#A!;Op01vI|SU|i4fI@)r4>lob$B5_rNw)
zdq)x!+!)&U+We4<`0H$_e|lr08d7OBCFL6wE0XnqT75^`)QB`%u72nZ!yq)Qh?(AF
zg&=Tfjv>R%8QuYvmpWjb4GnHn0bMrjYbrHWMZTPruJ!tgS4xit98A#Bi)}0%a>Fwe
zAe3b^EI>j4>-obb>$+JPVK@f4+<ddiF~&*tu-L(58)?M2YA*O=B)w81J`xjFgCm04
zWwE#I#e~iM6YqrgY4K(MX7ykd|JJ=`-rf3s>$<IZ1V6y%3HXgzQ6s{F(ci<@^ai&n
z=S&yFQ#fS?!cwqCBNBFJlB8(S<Lvk^cz8#?<XdaHGLuo$Ap*Ymjy;AJefl~PV6KY4
z2f+$4c5Gnd^FoaHZY+=37Ia<VQve~Oey+TSl64iy9^cu(M&E2SP<dOGpQnfMv-4Z8
z)cQU^@Jytqu}z3{#|4nv2b9!FPI<UB-0L}psvyFNLNEqymNh}RhZRMwpYBbGa;?{w
z%<2G0O_vek)h%|51Sz(B_fNDv8a6IjW{!B4j*C^Ei3vuOAoP$g_hMhqewf<7%cYnr
ztFP}3;aqi%YIJSdd<pThR|V9Etn>X`*;E$~SQE!3;D*4y!lo?XDTq1@1vi#hsy`v)
z&)URKaLqeGTYq!ftW>mR90@f$zdw2_0}(D>qXUT+|24gsLw)m?j$`eW80PNxp>moj
z<o(~(q#<~pvid8ZZ3)W|nqBX&5krsq2UAQWlRj($${UQ+8{WsYq0HW!gkK=D8k$@#
zIyzSJ1sWsEIHuy>&0-?{z~|nksc^jaQ*dvH;(Uc5p(Et*&etRwCd@j%B#SAB+Sqh#
z2<q$aS37Sw?e=c$s!KzG3D^{IA>5!5qG?-_0!(Pg_ahg^ZT7WjhL6}c2dHVIucy_p
z5e0brwuGYE9Z<IevsC3u1o8!09ZLOj!UdYvSXRh@+6bBXAE2~Z^W$~<QCCF;|10|W
zHPLZyRVcP?v|2j?z5D&?70LeQV^#E;n1G9P{K-FI-7B2=)B^xY>AZxoVAy?;E(z^A
zK$<YQF_eJN|1N=q|I=%{C30_|XdDMkf9Me<$JmP1v;?oCuHM^|u7mL(+4hn8Z*2$t
z{w69Mw=Nqg{!v6`zD&f!rrYA5BQb1v`C_fB)vCrm&hW@z6S(d?qOTyz1`Osv3GOvP
zyyB*KUTOIKWuXZc{TUJb(#uR$hQPlvd_I;f_?>8gr`4Gr(CQ<%ee_Xc28(mruUous
zzBQZ2m5G&qmBDe~{^gL>IbF_-8bAzoQ+<lskyUfU60aW$503V3-8vtLd@vYSZQ{C8
z1{y81_0l{kkEPoQl=2OpFMYD94si*azt+`ka19MFD7#3c#3um;HK_GbsWItX<Y2Oc
zCZTaw{_M~@2Cgn1M{7sBg;aTYt9h@V(ilAp%fdpj$yR8B-%m$7GBi@;Xc6yg$8jpH
z%!bRh$F=G^PGDN_Q?IL{#xLesJ8o3mU|t{es|TrK>l6tKFOI2jMfzcn7EOhKRU(~Q
z0iQ|#BO-NFrbmZBF>j;vY?R^I8T+m~gFla<p}&M$Hr4VsjX>>Fy$4{|(k)RO0klWn
z@UcpIwUp3r@6-MQNhf++MR4A*@pWZaDV?W`Ry3=<*Tv+tfG6R$=WRWRWPz)2Ka|JK
zHV6`$pPF=cSVzP{b|0EP*Gt;2GU()NOxQpV?eFA3p0M$N5S*C=4#!cSQ#+t2hJD4q
zw%iZI%M0KzwFxy_uQ4}d{iS1vIG<kql$?YPwk)DePOILStmLoO-rGwrR^DLP`-yb_
zWYw>=^>x1*Zq3AgN*oLE31!n^hGhWSUg!l?=k%8P{?rF@Y@#FiM0NwX?ihuv56HC0
zfMQA--7%Ts+p69FQpc9gf6RT9mRK}43(z812qyx1o{<vtuMDD`BB-Z4r1y9qA+y-@
zBt0te-jF$L-)SwS8$C~n&}KsJpLwP3hfzV*@5|?}CF*J$ZyWUZz&rrE`W>$KA{1jh
zwqN$MEWnX65CSHn57<{&T_mj*c^Y_3MV&w#Ag=rYG)TI>|5(=`nfNA`Sk+%DZpmH=
z4VMLPV5Tda*^Us^d6_QDQitk0DScLfJ2Mso;anrk*~u&!WjDT(w*7w0xOa0{PxtcL
z@uc}~h)#<d5s5cAqwO=6(p)5yb6{KuNZ0VM%_uFH(eH@uxKgI(ps!l_H1<;5t2DyR
zC7k{R2tY<vm{TNXq7BtK_PxK|IW$yq$1$xMrz*cLM~@y*JAOBK5tUU<k>x?FQ$=qt
z?OYB*)-Gy&Ac@dZIo9bDP)&5}eB=Fnp2uEC=l*6e$A8CNPU})vU`IBv+E;JiZ?FHN
zIR5L<P^vXIs-eceVN=1_@{!+WTxJJFqqIOh6ZBnu5I39ROD$u2niJZ@ZDw(f(^(Yw
zan!lMa=B^i(E-at%y1y0^#rcqhJkv16|_#YE!$J)l<T|iO8^CCqm3|W@bO?QY3(nT
z@ZFfn4`0xZjifUIP=vn6uqcBR?rgzxDT#fQSs34%eD#T<B{C73t4F#0pO|~500{iD
z@h><;Ox6q5l|0nx*@$R3LED`nmtL<hn*FfVugdFh)GF1bwoVM$0)*xZlJ(llM+8!&
zTrj2T0$m-jP$Y$}LP&R+ww!D7+Fmw{mkBJ+BioR)Ll8QHZz=vttIbTaAF+=HqPKju
zcb|r#wEr|o?a`uq`C?tLpdw#k&>^An4Twew*G<OECSx#D6LMrxge$5uvd}PEqA5)U
zlijaezjE6idwi*gqBm4<W8<TN=E-*!ltjOZDPq&%2sgdk_kWf?fPixDL*Ee+ksG-=
zV6xw^%nDPoe{gpo(IwB;vvorXLw^0R@*TCm;IE@>m*Yxnhzy6o(ahs|45k<kZ~eZQ
z$&Jqn*gYT?Kw0T6x^<04JiA#m+<}yiQqyLgcrT>yoEL;6VeIqjV9`N+P__e_@38|_
zDao%l1zcPURw7&D8w{AF+U48c5JdQz;ZMBTqc>+bzirsE+;z88n32pCH~ci`FI*~c
z`2*PcmNk^?V-2fF<M3-hr~-jXHLs}LfTcV2JsabG|DCKrrb#VvRi$69=k?OlKk@_d
z(BQr<*kZdxM8JSv@1HDoE%X1SioyJWJe$SO^+u}X#d#S5<5!O=_q*2v6|7GF>%j?V
zWq*sc+NN#6?AhK>xN%#udgA8;?xp*^09R4WZ;?ciA?u!!HOB)SmdoeP%=e9X`xjLQ
zv}mUZO2z_vKVsS8-Ct1wkI>xHUlAB~dwv~+)IxLUxo|4lv=Rsha8M(Q{wmHmT+TAE
z=`Isrc2&hUx~tiJq4LDiMRH4W$(6GdymJ&>ZN(`YHQqR;pg3rA`PTYQEIz<qiKtmG
zWq#urRH4b;DPfGw;!@;kjiOFb#g9ADd0|0`yeQ&cZE<EjUnynCOJJtO4g=aS*LVi@
zX(If40nKwT57{KE<Kkh%#0P-FWkdR{q-}#5c2Qn_)aFm!0fa2Y*@)x|Rn*S5kFRA?
zuZ`>pL~U97F7X~!;)O7!GkTLytVctZQ>cQ|s#EVQH~(1HNJ1{h<0*#W^|eHx#76U*
z$<0tQ-WcA<Ffl25siV3msJ+QsG?1iv7HV)mBfQlD=Rk8(O#sM*_&_ID?|7XxNnd`?
zTDaG>kxca~I0P<cc`Tw*>{hNQ)BD+^Q{+jRmqEVfTc%=>8Z3u}V2hY%S-E}JE6q}_
zRW)3aj!u~AnH9i9Ic+XiC5&ZH1o!ZDjR#L}`3GI`JRhuKD-1!xIKE7zBz`l^sRuz+
zFYNxeglOOj4e^6|Jh{x}nL?i7PClJwM-dQ+MUMzC9}h0OQN4QKH+|mGeA>+9&O0oH
z_&kFm2(QI>U}on#5t5nufL6=;#q&)%)ZN;r;4(|ztVi4lQ(v&zGr(@?Xg~@UlBBxG
z557(awhC&hgd}nCHk+sLiQ_qC-@?yXoe09hBeNjiXtK-7?IxZe`gVm0(VCSi1;70h
zk_|do7H@kOivrHRtNxRTKCEiOm%bD41i(eKB(3k@`#bD9Cq0k**O*XPcA0xRYr#5n
z8;Kgk?OOheUc9WtW_ui!7W*6tExa%<c6f5os#R><E!#>oHk3^ulH-&r_}*k#F-A^v
zi?$&HENa<9EwgIo@cKN|1;$l4$Q+#z(xHh^@*~z4Yd!Toq0c1o2N9RUn#zH)wJcfn
zF$Gv(`R49UnaKI*F2vgmSEi`KnfrKH#`DVoV`9+1jJL|nQD^&7CT_|~W`4>M*?;cA
z4anpr#Dq5b&>vm(iRTEa#wt(KB{F(CC5|u?UYBDa$Ed6PX~!sU`Fgu$!dL)=3a?jx
z5oxH`#$#lmPYcH>{#AJAo1d!?oIt?;VG;Td!>pAL&12Nmm>ti7lUGfH)ma3lTpfSt
zf<UMvy8|7WGS04w?OOWZ?LW^4`HZ{`gZlp>nKfKJ7v&`<Z~9!TR(mdor&-+e0FQRH
z3(nc!&#eMkOl#&OW8tY%I-2G)0Bh6+sqhhwtLtIi==+1N-&r~DXj+{0N(y)BkSgb?
z$nu5B*}TrgfI&z|)GP@i9YR(&mGeQM!=-U?Q<&(-Z60M6>!%n~Zlf7SW8%!77gpq^
z$39|Vkw}wef|=OqCD?>z^BT)w3E~X-d>lK!>Yt%u{r1-cJ}hBGi_hrpfSyc|^{I$F
zvT*F~)eLWreW@5e)3;D-MB4kC33`o;I?hBMz%6WkglA=kgem|OUxQm+M0!?wZTxkv
z-uwOn`K9PJvFLO)+Y7rMzRo8PqKw9S*dJJp6UQN4e<P*k&5&@WFl8A+DMYU?L73uq
zHOxNB<^0z-d~LF<Bz!|v)35E*?8%OXa33-_T2^g9HTKkPF2uN;88VoOm%T1^^wV7!
z9=68F0!e2^9!H{G-VvSLjS;~0yw%Xa3YE_-jfk-JAbz}0t-kWQlklPaJUlCHLZC5-
zREmL(IvZ(Z+CINt3RdaP>$$C<UJ`au<+o2=45a(v$Ni26HWtV^U4dozqHKXHS^13v
z6UzC~;Q9_3?t4=H=>(!h8Q{0_y1B(?2Wj?;*UO&g#Oh{@&t)dV`-AhpkTQQdofNX;
zkk3Q^B?TDzi!p-C|ER^oNjM8YO@%GKYh|Q!q!TGvD+|{ZL|bVlR#eboF@OdmIo2fI
z+>TX?0U8{+P^t>Vrg4B@HlU4?W4#)YRx>pz4|Wz!|18q$vk^miXmSKIU~XEAtM+9B
zjhm4uqc1)SC-M#{`f!Yvj|ebPXEJNMCta-SjJo)0!$Lfb;Mmc>vG_vn4<PN`H72vI
zQH@YzFp6>6;sc5D1F@ikw6I2ff~OiXV1H`pxUE&LALT14SKnyfm7j_GdcHTNZ=c(g
z&{KL+%K4W?<NVqBkJaA8^L_iD9*kpL>u5F7N5iKv-X(jF;}Jv{!Rnf_<Yietr#^pP
z_fJ_0rcmEDMG=KL&g@xaX+G+VF;}qu{3>DiD40Dlyj59&f4BaJ;-VM_2p|=`M=jk|
z0`EcIQ^m~pAW@Ax-N~*6m+^^~#*N*DrOn_c&w0`p;|sMjj+dfLcT)|KVDbB{%m2*h
z_P7JWvsk(6t8c-H;$YK`ITcGfgKfp4^}BdUNhv;$+$t@@VRlAv-}s;f^7mZ_os}Fl
z!ql<FFy1|=7)%kYsgRRwNz3r<eqET8p!%+(--VWhz(WStqr7#N$-bwaCFY(WabS>R
z{0&82>bOHLYZIDmsD-+HU;^J-0xIxNfUwIwRiT8RiN&$};|nTx$>cEO-)%yD9%80`
z@cr<F{BX3CIB3Xt<Dm}V%FA_~uI-{zUl=_d1NaVc+ugGaBg(8$%7%Wan!H-3sx<p}
zq5>J9v!ZC_yoOfw325JBtM+MWe0%8ggU4y5Q}OO&6CHhm=ol_(d@%JO^pstpmHEVR
zWwVBVX2sKIsp}&L$zk(OTL#lo%#Hk^bPWl(O-(wkby4gvzL(7YE9)d7dsj}~0RpGM
zsc*o%LuZaS&tnHBH>mvYHAqz=4iv_eJiZhSgsP&nd?O_-7Khc*Fe4@9L?$@q3p9?V
zk$@1MBbB=9KMWf@B_~*2|0B`RoDeJ;C`iZ!>wdsx26u-umS&6-8y5`DjW2aJO)ZWa
zxbNss59#W!%syGBg#XoWI-dKk|9(|joCj1V@{Ua_3X1+U2P`jV+|B>dW^N#C{tbJ8
z-gWC`HJ;P={R@86sxfPcg~1G6$0P_VgJC1qCgedY85G7_sPjCZcw%(()YfifrHTF;
z*itih+2cn8)^q&3<UG2L(mq1Vm2H1JN7by?+W~63RT$;uT|UqUNGI48;wn_h3~%0@
zEymiZwY|4_-F>y*@HP_g|KHff;hQ9yC{>~RO<Al~t#!tI{gf5BqD?#d*|f6FrhnEJ
z{w`x;H0UU#tRO|fiE^0ggK#SQr2mBGIR3bXsV~cCV$-e?jVoUQt=h8k`;-y#%Xl%~
zso+IK<K99OC284VYOS8xx2{Z+yc!RBwa(&Bx1#0Mf1(R@vE+8b2rfqO$zI(V4Rx^z
z&!++*5v0e$>=L*DuQg6yN?rQ*lZ5DC`I67mtpT-MZD>${&y>;SiN3ov68nVv%D1CW
zX20SD2k6rx<R6kke|tm+OOyEQOLOd}A{`yA*iqw#3`}^c&0l(P3^;9{swf4H0zPzF
zZKN`(+wWc=!CpCMJOt<MeJGv}CeZRe>tJ$5XL;rqOECCehQ-PI(&5Np7<`&{+@jU|
z)`K3rezjw6{}!MGZx+-`BHp)An>Pkv6HQc0MU3S?e1P`9iV{8PPBr#|f7cb#b=C4(
zD3YyIs+mwzTF5oAhtqH_$l{n{ESA2gVqdE+0W6nmj+I{DT2nTMw}6_)ltkshl1-jL
z^T{s~jGV#GmFgWH7_&;HYS+Mq0BSm&$mrSx{D?;~MuI2schx|L&A&P9#1U-j`!=_)
z#iM#f%74x=E5|M6i|O)#G*$kWe?<|XR8gv8s&YI|>WZ^p)fEoh2WC4jaFdg6#C`X5
zJ&d!nbI=s*z^e=5<P*d{r%}j0g*iRZ>*qW-K850^6fH-`(>Z|})0^UYBy*TZaiZ2*
z{~y>UoviPA&iwc1Z?5^)*@UnDT_?Wb(NBDRL6;(zbXuVG!EdPlN)k!*Rjv%ry@f4{
z-z_><nkKb}JBNF3A`Z-+QXCq>Xd8LuQ>iWjA4d#j+!;J6qg`7UZ(6j_OHGNum@;&W
zO&_j+tk0pgowq>r`^nS-mc~#L^}BC5Ctt+7_2H3v+r#%(b~C5udNT~`B%RUFl+Kkn
z_SEvFjPs_y>ogwjI6x^-grPzXPsRJ|3w_0_!E7f*Nh#BgC%!yceiQq#Oudwl=s-?#
zVv3!#dpxGv7<@5Rvp&7nFmve~NG@pu{bQD7SK{GUjMUUFKcvF8(rJqW{G|Q_cEeOo
z^ACCbK`wRz9_B#a8Govjg*0zBJxCMR)c|je2(ACr(u*g<$?3z-i8uuJx=b`h9Q@t3
z#o9<Y8Bzm-m<b{U6;WZpsMQ;{zaC!@npzdoICWg-3yFIqZ;7#nT*HGn1?BSE=U8v-
zCqhiT*v34eTphLK9pS$7T=D&5{gqw<0PAPTBbz6W-UenK1}Rr_*XwM>*Tr!v39QGC
zKoN&;<?HK`F0dc-j=4R$nQm&CMco45J7wOGQDQ2X_?$`O(I+;0UtlU|)}(9NA{uk0
zWF@h?XNSII&ApK*d#KxScb_uf5~gMweOLduF`F^r&ihKKoFP=UWw33x;z0ErWmRmp
zoLMx`2pCHYW05Kryb`*0Js`0pMpy}s*TomZ!>A4M;)VOr1BgG+o^NAM(-5!qei@2i
zW}Z(?7lU<fI|gq6D+PdBAbUBoWq|F1x=y7WEOFs3yEq#IyYSt@T|M7_m0I^dbtHjp
ziGdDJiKW_yAJYfa@x{WSeK}Grds52kmf6v&&)GlQT=+)j9L|rvLO9Wl^&X)jFC&>4
z8am$T3p@fFt>%ZX4VAHdXo7H{FBeYxt<#%-$_qX<r0t{l{Pp$V1^>)vjah2oxDk@u
zK*xn)6_D8hdP?!^uSrC(U4KJw_rB;XJ+YyRb~P%O*JPmCG}x~L*{Pv&rsTsH**Q4g
zg-y*KpPz_2K)RwpK6470@T1Lg8xd*cVdP1B^(w)HoOx(1b0@t(w*ZaFwOY4Vby(=o
zuY9*sE7nt*<TU=K=g;Tg{o4Y8W>t#T@33(6;A``DU)J*O3ZS9{V7B_;NH94(z%76j
zA3_Q|El_h@^+p>TX>9^B{t!n!?&9p))_0kqVdA*_!)8urh)lIJU9YuL`;5tz>TL<-
ziDPHoQ-D1p;M_TOKOTcxK-iJ){5F@foKoJLcLW^g)do#ajwgUuMv=MCvwLafL5TZ}
zxyZI-W}tshcDJ!!Ce?8T)UX94Q|pnWdS)8=ENr6Y-|XbJj_FsM9Rw+He11Y??uvFU
z#YpGz3|`5>t=QB@L<WPE=m$Vzpq5>GHB64tBuso~%-RYw7J!qEb3kY3Te|V^pFOL?
ziaT!;X`HM0NN@PsIictrKVp*^ynL9nOfN4V@fBp`TEvj}A=^nT`TqNZ1R{!ab5aQ1
zcar)~@>Ewjxg`Qm$Y-nte-1at8v8-g0|)Cj+TGE>F0cqLr0J;)DWd^fR73nJmA9=U
z=QjM8prRx|P&}(qb5`7A+A+UVCER;$cTbs&Hqe&kTBrETQkql6ERtiR8SSu`cndFR
zDbvW}GXvesx?aoFcg9?`a>ZE(_ZB}A-~Z4VgG#MZM1md<c*^Iq*gr(Y@<A>XfHa-t
zAqp`Wk<%e5C3dt3XYG2S7Y@j`gianb#%aTMSI}NAtae>N0l93SdA=JcWYRqgkyHB5
zR591r)Nnoj{#Pkvrx$VK1PDA(&U{Hbfv?^YY<8~%+eTMgWkUSP(f@4*X~x}+g+5-w
z*;9a8#;UHktOpotO*X!15!0+0l$Z-ZUgPG9xtgvs_%AgY5Hm{f7zK`_Mv-I3&qIPj
z;2A~%I>FUHRNI5l6=2Lse#uI6hIlUN_TeMM$r@*U0x;+|CjG!eLrp{yMWA_bf6O??
z-pcP>WA;K9x0EB>ma7_^!-F5#S9sx2ynbEpLnIT%9;d4!m_e2{Rr7j?K8JPshku2R
z&3l|a_Dx_J$e-B*Io^xcjdx4scNoCr<JBF%LS8e+1#C`GYMcFv#}{RfFWCb!+O7Gg
z%a%_b%E>lK13yBd#>cPd)G%!e|EdTrPm=co4cf>VttD~onhz5zi#o>l9`0X{P!J{?
zO*fifb#{~tr<m_kH0kDRc=Ez#Yj$^nn}6bQ5DO!d>2O8v-kG9qV+3EitmocDzWXzY
zDIrov6N_S;pj4YE3B!wpk;q_?a=8`m=L=k#SzG;d6idHuCbxG|9P0VVRqw+<hD6}s
zYQxrvpK3lUuU2tB_FM|{!g%|EC$D@eWBdv6mB%#@Rk>w!pUnyK8a?chv%&9FN!2Y-
zK6=^^6tCWQ1O5ewpoQr~m%w3U4v1|u$PQCQa0}nIw9)aIc-Ib1&TM_z05j1(hbH>l
z*FeC@1gKmh#u++5>?GxHner0s#QSdweZQ**mTdxJcJ_qIsv*6MdO;(|uaiw_zEViL
zi@iiUvB2K?-8+FMk^&n=ExnrAnkdc6+)7EBvj*AL)S&|1xOF1H-@dDJ;gy7Rw}|7&
zhG;3cBQfMB-rO_^D~z8zOgTM8>!_R+HpFDNYJkrRHYpoYbd|$$R2N|RsrV+0SG(}`
z`a23xwW`?+Bai?=RU&slc<pGjprsB148B4wZ|8c+bIo*eOs`c)V&_wvC)e8`z+YOa
zZdc?%xic-u0r{3s<vbvo$afFtu-i!O=O;j&jWON7#)M?ftmywwW}*9ErGR}*{9hi4
z9@_s~t>p{dRjo)R*GO%7z{n0O44`^2+n{F!b=UpES1Dp@EeaG1qD*&%*%XxHM?fcQ
zd1;0j`{BAF7AE+O%l>m+6dj^R8CoA2uX6#9y=7K$Q7sgpC4(E_Y+V>O*W#k_=zrIi
zT>*DQ;umCZD~;=0y~IZtYZ<bCqwbwLt6G2q7nywt{wfBII#~x&e{-yLoJk$+-{mUU
z1Nu+Y<B<e{=#iOiK&_@qig4e#0t^5G{#+&K13|A8+yxu^O<15P{Vh}MM@UkpG3Pt<
zt#wIN|7&+TC(&E}gkS=z0V|!>huQ9Qbm><(lC88kSAAD8cGvSKs*x~U-=f->2j}Hd
z-uYL4QVRl%rEzPwy@xses<yt{d<na*2RZ=hDN<^OrGfv&M$QF8w@Pf$TLH!X*~7Sw
zb-c~@7Y-#=loeA;Q8pe_H%~_-AxXAOxJJgtD@TMH=|jc4KpF*;{xyrZs?vNW>!i~E
zWdSP1pY1U{g#0f-VWFZ&v-55y+4@m5{u{opZ~e~8Z=>x{>x=8~QP`dM7jzw@+`#3Q
zl^}G9(haE*!B~k+|4SgDY`1a?0TgSWyt5)EE@wN67jhkB)?dAn6{pRfeQHyiw#{{T
zS;I$`Bampaq59jN524Sm46Of<3L|gF{hhjfF0aS+Y1@DIvapYiI2^GY2%a7h{x(Ym
z^$}j=Y-VCw-$M$1sc{Ubi3rUD<q;r3@!@oL62BfgwV#JIRHaFJuGwtRhF#PX?z)l_
z9G`iK7<twQi;+!%jTr)U{%IQ6u7)aZszfhC$UnT#Z|fHor%iPLk4XQO|CnnR052ny
zg5UFK9{_oL45@pJw0bHG0YvEP+DMnd?()syk2?pX*^vpkYC2|%C_s`AZ~P7zuE?1&
z+ZaOphbCxfXoDi+IccR3YWRcsFES{pLZV{Z&G-7e+1`PGm3{xlK18j)9t!Sm|7<2x
zSG<J^*jGwNr8(MWctFukIHR{s7<}Tfm3rgFAV4VbeDnu13Ai2IZ`|c@>i`-#HbL+&
z8e=%)^c@qxh+aHnUZ|y-j$UG{WFyqrSIxf;*k(&cEJvgjBB2MFhryOHPskgc+cOam
zf)~SvDiu-H8&})*l=75<NvKoZEWz(7J_y-F+9ypd)svgH&&$^fLS*UE1Cw87N2HZF
zs`nca2an=<95dUgmio>dKFr-nj?tL`7l?V7`Kv>699EoY*C2!3u-<)8XhaXwEBlR1
zwcx~3h>rdT9?|_O$nvkFibtgzHKXcoJok8v?%6V1;+L{ZnFU+{tWF8!38m`WKKDow
z!69#Q>k71b-m({6@5}D%n#d3(2zx0{XUZDCTAgB*gnd7el0psC<JtM%*!-b%1spu%
z;d;O1(wUzEiDSl|T(-ToUUxZ)&XMydzC%yb+H^QiG_lGh=2Ao*wLKp{Fl2-PwUhCX
zcfUqkTWhTBz=S#Ej&?+#G`)bOI`lx&_Eo8ZV-l83QNOWeZvs~pW_G)-ogaj}JM-pi
zI)14-I+!`u8kY%(3_zfNclz%X@5-e|_AEYom)A|6@QQ6mu#R1ryq6RcAcp`MHDJ5L
zbLe_2cGu%>);<XP8PmQ?0qo^_h#pytQ4E`z?Ogwon~9nt`@b7y%~_J*ct9dQ$%>|!
zfa?{ZH$W-y3uYV@e|~*s@Db$!G0P)xeD;+^15JJJUVV^y#K^R-d!wt|9uvq#jySHK
zUr$d}Pg^vbZ9@aF{%HFv60o9JG#C3`76>D#@8X%PVF?n3lAn>Llig{Cs8>bY<!}f{
zm#2A{2~I=KFX`UK-QHM@x|@JgL~=mPvkvc6EN**l>Zy;UvWeJA=!Yy2-{_XojFxq|
z*T;=<qDfI9>$afToH~nNRojy}I1acVS_2yl$8+}SZzHb;GEa*xKz9;5^B;~gbFa+?
z@IsX%Jc&WbNCxXEl4H+pn)k078b)$aras-c*^m(b%2RdxtR4bjCGCJlR}LV{DCzyY
z-F#+ZhmC9wMC{GrP=LNG?u7@CO2_U9gW9mI0;8f41Z>}V0wL_lJD`J2@BiBlK;V&|
zpfSc=wF3AsL*4{D_X&u^Qr|D{M^9`#&f!TCCgF)miHV^|f1qN5ufZx6ruR2=u30+Q
z@*O<Y3w)kmFx#L$QH!PB=!ig>7u+|tu}qu;r_!)5n+RVN@3+5SQ1DEQ@}{q5y6<W6
z6WJ_1yLOfWzH6&Uqod~@1#g0PK!~hCfkDNO;_1z;m8s_+t%je6Zhf+Xn3-#Ioka#r
z4fOT(FI9S_A%6V$;e<kB*KwW3@Q7iCR5mC{qA%IOjKk+aemA`CrW*NZXeFB*LYj<l
zy1_4MPa>=uiyjvPhf>G|?vE=wUs#Qa7W~GV$7o8KfzIhqEVVG+0Rg0S0mTZ<y(P4W
zvf#jM2QolOKE2FR6sj&>e!CWo)9!sk>p_a37>}xdfeq!pfhqz!!WH}KZZqf&>Uu`Y
zMse)o&*<lSEdBTk*|46Wcq=!cqvUs+J4LS+-rNZfKZQx?y0!C}_*={~)WjEW;dqY<
z3*J7o32$$o?$Zv&!xmAXyjW0hEai8lEGh%EfQ2M4g%Iyw6t`{ku7dFvr((jZ8Q;Fl
zS}rJdg)qXtX2$+?Bg(-5u>1l;t86vl32wcap^!ADE8;i@Yg1%{CnDeHMlxKysjHWi
zGG{U2_sVBU2tS)((`;AhGnjMYqYm1(-4p)>h7R<v(`~KCdMYIEoG)16G;_*eP5FBC
z@xTSNl?<*BM|7&mvX+#N-K-o~Fy8Uq#qq1c1!$)*n1_}--{uKfNyGp3NsA9kEUyT6
zzX#>fUEZcC;2j-(0<kkBR%ta^;7MrP1rS*yTlERH%$RbhHh1*7s8vkeJ}^*_kISSv
zfw-r!1oS}i_0Z2w`JN;(aVkLa>r+M3%5=nWmjZEF`Af8t>WXb=pwn37V8F||uE(c_
zg3uK)+n5-2*#eC3@GgC=oH6l3^(<9XY&SntTLS-zeVu8hHcqPvShiYVXn}GEV}ozY
z(Il?WD(NL_?196K4v(YF&YwegojAQNYX&T<3zOjx-<mLF&vvkX;<`e&$cgDdu0Nyj
zE^iWT*vd^7lz`G4{W2{bLKWK#D-j!U_VXe#H-OVD=05q<9V)GvgTmtte;w1AVxtgB
z?mEy<4Xi30ShgS0oT%%}=L;z)DMeju?oTF~h;64gdw}?{bVbOXn3?84%9d)=qG#ux
zm?--ld#(|Ie;q~6$f%~9oYucii8L2a?7on_SNfGN-NcZqWnR&llL&?3BBGZDUTzLb
zq;wv#<ErXv2RvDgOJh$2WO~{3r^&`;fR0oB;+X5W!vri=_C9HLR;>k6vPPfr(Z9#&
z+-#@{+Mh6%o~}e$cJnOw`Xs&y|G>UW_bF80JzqR@CIn8lT@Fen-jBHAidLP~dYp)L
z)lljPqERRK8s#=FiQT)Zk<3E#XGwT?P;_XJ_0A+ced_L1oR(26j=Ug&_0?lp#N#pj
z_>ReWhMu27CNnL30J*Er)w%7#;bxma%1<RPQuq8mr0Ms?)g?$hC8E)rMN<Pu65YW9
z_e!cbCj#dSSnF?3PtTSw4lSgfzWx+FcnmM<@D(l$xg58VOF(2CM@-h6;LCo|wMzuk
z>%=%F$G_T(!0R8{<`Noag`3I^{|XRTUctxy&Wv!V^B0~+Ceb%y8q?d`vfAUYz`wM;
zgv3i&-Ht&%Mq;v)w3Z(#Xu-QkFR!QMf^jvO?v0z<I_dU=IJr_QHTUa;xaEiR$UeT3
z^V7FDFZ(~JpQoTxy%IoTk*m5nEN)OcH_h?b*k{Xs>cAR(;K-)m0}^ZorFCI#cYtkw
zG)9O0_*?xpS3}g8>+Jv0^p$Z<w(s94C?Flu-3Zd%-JQ}UCEXp;-6`EACDIK_r*w`+
za`edUxx9bRf3Nmp+h^ByUgvon-#Xr0&!UrFtkNa&A+NmwMh|DzXmt5ZgVsV|TVSx-
zLEXVO$;1ZAOUrJKnX^UsF_F03boSk(rpMyCPYQGAwP9Q>GnF>qV{Nw{j4l1YX~nbQ
zO;<>}6(fJe?qloVEdH^swqGpr2~yIxf5kNTPrAO_0|LUtqdoaB^bIKJ@Bshh-e!P*
z4{S*h&*kM@RwzW1`}{4Q&UZ7j-$ZqCf+)z*1B)E>0i9H@Sn&eJOulITbxe%R2tx9B
z&`3M7CKYtFCm~mn+cULsj;1z2MVj8b6!&uV+^O2gUBDn#H;Ft+hxxu4kg^aye&hx$
zUuM1jWLKu55k+_~&Pw8?B9>=_*=5=`9K3|H?|KZobdO!=ZM}L#eD|QQeJ^)Rd}Q}0
z$m#IaCaom<yBMR#?mk)wag{Bc=KJbBn{MPqy=@6qjGN+L1_6yoA!T_Ap@uof*r<g%
zZ@|*!s$zGDwuV<?IZOYFg6>CEe1FqQ^DdfSzwHR$ulv(Pk$JhJ^5xC^iNhuPnkJ07
zc4SR)du1$o>NH|~e0;)?(G4FNNjhoB%be)Zh$htDns_KsaRzG^7F6#w2<7NG#7-<`
zZeyM_I7)w}dqUbr2M^#AMju9OuGNM__F92P`Ytw~@3xwZ7jEDO{jTq~1o@dL8>Svg
zR(Nr!1ifj%-1Uw!!O80I_ojQ)!g$Y4*8UUDi-`fz26rL{v`6^mVa{HQmO{(mFNxe+
zabYeu=*ru2id)A%!I;WBKT7hY$yx}9JE-R*Cf*0ncMyI0)R>_s@1XHXVheg7Ai0sc
zx~ObpLGSKOE{}z^e7F(JAQaMJjX$sTl^TJO6UA~F3F#93r=7%ZKD_E>{UF%8D{x+K
z)PUCH%5zlr*`e>_`rM(bCU_j4(j8Pv8{4<?yXK=1$=b?cNt|#Gw^GyM*!o#7+OclS
z4__t-xqn6wS{HoqbqZbbPE^G_;WHVQEsC{ceKtdYg)r*OM`qyR{v9DWJY2d}O_zm}
z{mu^H8l1cDUI%@QwO(1qR*-zwDnS<e!<5USFpKX5Xw=$7^|svFzJFQwVawmV#qJ#$
zmA3O+t?^uN+0E}w1{n@e!#i@^Zu)O0oe2r<89V;<c;%GFY;N!(Za6R*d|?dyUi*pd
ztVAM<ao>-YS?mX1wmT8V;Fejs5JqhlREVu<<E*E{yq48kBT3QPD<v$B>wHPRv15c1
zv<`fa;oDTY#@DiW3VxmtltA9M6@8oky1#6V{D$xuJW;V<<9)GpEhMmYP<TCQ$71EO
zFYr!ZAzI6wdSKLro-#9i$N$d7=FW(_4u}n@sf2s_=tU{*O}yt(Hla5;E?sb$@T%ZV
zaOuFiF;5X0SF`3#GBWItTwx^3E3Q3}7onjRVr*GALMObBdmY_DCEA>anXi)GXMFjP
zM3G$^l2W?0km`weH8MLwE&D`Dg_$O3_etZH^x9Jdmo4kYYSesi4ks9sXf4sPY>ko(
zdA%mHL@8(c(zy5ZEM$v63Y|qYZ#3<G7B%s7c=>M;tg{1F9doj97@*tV67+pa$Iz)O
zZa1Sd=}!<zc|P8OdM3-CP}v?E$AOnwzo@ZyMlofa5{0$AjvTD|`f9BXi%k5gaL4&Z
z%c9U3rynGy`aU!TJugCnb*h&JW&y<wM$YW|dW^+eY%GK=2y9Vc7Hn)ivnsOsk2R|{
zZGJeWF8&TP2*K;D2K5PxKZ7B76$j>bs|EpYVS?A@GW`6v+s{APRt`Nq_IU$S)i$U`
zT*!_LCJG+$A2#YJ@6WzkKNx|0a<N3;M%q_syET6Q0-tkyrvX}(i^v_h`I{Pjm%P~X
z;e~`nF{i(EVQqQI9U-SQ31_81vpABV<8UNpLh{9FD`vjQyZIi4XYs5^-mb{$lK@cM
zSvX$JlbsBmjx~$~GLv7UJ`p80)OpgQeZlFL)Xc`_?U1~y;H3PKcNnog+-?wf<~2!y
zyn*^ukH{%+C!9g?b27KgYp#Wz+KdE)<hj1;f#v(Bxnwv@TInckSb?4HPvsLqY&cT4
zvjyUuM5#<xvMIm#AkskkzA*!5@Yf1C_g~H)kIaMD0xg~SGipl@>}w9dKhQ52wMwKZ
zEpO!0?VU_19E2QQk-)qK?a@uPNu)JI-d~}Fa9FQ7-GFlI2hv-XV8R#Z*k^P?i~sJw
zK=V(?a40A!b5TPw8=ogB^!v@aC%d=UY?t-AEHT|=Nw(CyYaph`EcU09#B*eSm?jhO
z5a+KQgPy+`5Dmw)(d2`*7j1KXl|0nD`98~wGZn?TYRrqN#1Q?x{6@zm!eY?B=nGc3
zu2YI39W8B3mb0KOZ9`Zu2%4wNo@EEmyW2OhNRiD5Q!DAE5aM&jxv=~Gu6<hL^(R7n
zw9pFRb1Hw>^i;UP?3D?TPv8?Fc|aPZdU{yzDZLWR&~-xh1yX&2$g`*O^fFU+E*uC9
z4juRI0$@A0JJQA&p^{XO7dHv0bX<t&#j$|<6J@hMaN#y^xGj8)+|^{OAQDu*5F8c~
z_e&IVdr%(oxlprAzXr!)$rI2xs!wG_pG~;=zs0B<;ftKZQ8_pY!T6-#{PheC(ZzNs
zgyoPvn=?K&MdL=tMwz?xl7_Z|O!ve+y(U5S%J?uoMLOa`d&hX-hb$>bTm+VwCQfnN
zOv$_EKHIvn7AJ3kCCiEPO_D&V%A6iaT%x~!T|1qFpjdarj>4CzKkhgGo&V&|V6i{3
zuO2Zj{=UWHUO*C!)(?t$@!OK3f#Zr_O$5;l^xZn&&Dxxs+>%3m?ox$C6UF!>oS&~e
z*A?;9TcLJ?q~=yxxV(SHTMPTd_C2g){IeQ7rB&<{k>1KP_j+Cg+?go>CfS$REZtnG
zr~^sV${!@^2BEko!`4zrKVSTAs?R);YtppRl1pyE_L+VyX*<?M2;UQG4}tfDcq)Cb
zTbbfFIHzLbO0VIxx^$hjHr77pk>5Uu<9`mAaR?j9&tvr*9h$#S%_a)^0dhmIP?G|a
zF{T}zw<AgupoA6Vd4IT;75dd>6jZbl>7PU+|NLP3wiWkpG(96F)*AmIH3s<+yQ9*L
zDTkz{Qb2{E7v-1V;Smt{nV5-q@I-Qk2HBV&ttr^y>BQ?SD1?7hGpg|o>zB(c(3Ar-
z6qct_$YNA#()(nYBM>NPLSp1-lPk7`?U<qCxb<hA^Hm$$gS#;~)UOWo!caT#v}!JS
zP!Q?}lmFiN@5V?#{I8}o%a?I`%trdUIC|<RCrOaAw14r%m#~0lDT#PdA}Wbcn|z8@
z0h_;xBO<A)7856;!#|rnRby-Onag;D;l!K5G3f8(Ns+D}^Q$csy_D>NYkZXrXVL{0
zn0^<dD_4Dns#pNbAu3MdToiMd6mq<JQ|8Omo|aOLsP9)TfreRqSuGrlISCuQ|Co)>
zP9c<vA0uKGaK=J3Z}me$GdrCtImdUBa`nJO<8biD{?I{eUsXGB=nq+QYlf~)q{Z_#
zl?C=rzd*PSI7_C@M5;+gveRGvGNnOwfS~`HH7}iaub=!j)AsoZd|zoFUgQJ69>h@l
z@g9_Jk?45*gEg6|SdH!xec3vU#=~9IU!x}lrDYSJD+<#Vg%bolbEY`V4QufzDcT+z
zt;iDW!R$aynw&acd_A@5Ujgmg2n9KqaG9Myoue!vUN@Qqp=h*01JG}}@1%dalDy3l
zrTRGudP6=h!s8`c+=9nQ*T5U`I}zj7jyg~M@mA63uB?J2ZS5JXJ)dJ3N+2beqAC@C
z8S_Z_#Kf4TuV7_^S(TGN<T&g(WlvSEO!;}A|EB_-#xwMT4gcaA<Pf?k@6xo?R~h^W
z!)9>pQ)!&?_x|Ng|0#WA^0>{n{Gk)V%M)CWAm)Zs=VdIr*1EQ9scsn6Vo42j)5|P5
zOD?K#Njo@6e%3{#{HGpm05TX(+P{~w%D?V9Uy?dsC!ecLI9_@zhi%BC0IA=FJVZW>
zPYgGV)2l#_Bd2H?t=SkoX;j~LO{a{ZrTx^wbV{^xU6)@c;ZN_c;!v7NN=Zrn!{J}Y
zOG7iIDiU9jaDUEx7d{LDopRR@FSU6BS=Q-hummDq;eyQigR7s}<x|_xN2WoG1*Z%O
z$oSor6jC5IudKpd2r?@W*4bcEnm0M$mg@6kD<DJLp>#q|IYju|+^&q9-GTJyNC-!u
zGqB6Y7BJ|%?gLsTp|>=e9mVN>$JSwAuzB&yoh^#Mh}JGY34q+;vp(a&!@mnVrfQl5
ze$?NDjZKW2@_fBgr(1UN&DErwdWfGmmQJBeHPx-Y4x%NL(J)8l3f=LcgfIR;{a<=x
zyv*8-?ej~X-riB8i~nVDoy{a&yReXr(}koe0fFw<L00~7Kb@Pl8gvNM31jT-hk365
z*c!_<^b8UV;cG1rlQO`RTI<FCE&F;adjLcE4UE%(WIeNBlQIR@IWSgKQsiha!E}IE
zJLmpi$EnwxP1KxBW56|sGvq1@R?N<3OJHSMdJ@)&vOnz3?M=Z-+4G7MwYQ!eXm@Hg
zsVemZde1cEApumN)jD2MRP%C*ET{Y8@ZnHtUGZ!88Hr2)V#lbr@2dmaGy&=u6qoKB
z=7gMcxSTxXMt<h+pDes#nii4dOAkTJHbxzpAw?!jg0yh!j#ARuWE9b=7f5w_&vMS4
zv##(aiGn4hqmc(zNGR|#y%?`T{`4f7t14kv+3^~)MJz=&HADL#1#mXI<)JY4_s?Mb
zjQ-?|lD3x{gm+twUt8H}rv%wVDTmgDgZz~XJ&MJ$eaj>Hy(wHb_Dzc)PStFffN?Bq
zOZ5^I)gLh9mF2d#+dzJrkh;rIPt$^7;UZGYMUtO~WDo`|7X_kmGy={ii_gG;diGb<
zQ!~@%$V_~+4so>WSfx7@Ot0W8M}=Vx`Zc&B(t(FM6l^_|$eRww$9=FYY^1=}x|o99
z2`<94uRXEr%0>vbu@@sawOPhl84x9n{nW98k|4ko-6H*O!M9o&|I&-z4S1}zmDt_*
zV~NWGi;|P(OozL;omgbz%I-yL`+YC|Z-ZqPYQToW!^102rlLyHm_r?h8|{87BxF;(
z7Ml~B!$dnEEB>Q@W{eE?weXjoc-rKr92c)qbnx#O_UVQyM^>ilR<-T#v{lF|25s?J
z+~@^QQ9Ba3)6WJ=!zv=|814#Fa`XF&A#e2%f)NU6b^V-hoGlK*O9dvfD?I3=?p}D=
zIC}=P2T`O82rjA0R}T)#j4jvmz`e@ks}l~Lg~;);*T%e@Ml3Ax!{?Hoee_d0Kv9r9
z6t}DU1RYF$1NO6L@mji6Inp@DWcf`;(@PlMV`l5mcBRg$_n}12<#M^c9SdT|djC=@
z&uF_`52alW4GIKErM83Kx|}zym!H8UBrQ3q_kUUFWihPmLGm`TJtF58(F4?#pgOT^
zj=Y*bQ`<$7^#oT|1X$S)0-x?mq`4RK7Q}OAODpRojecMPxi%XJxbB+sA}?xIQxfM-
zFUQpe?&Py-PHl?y76l#pr7Z|zov+IIQk=Yl7(4y0yu}Z-ZCK<TLLDEPZxH^$bH4Iy
zE3~T6+yLPK$+>n5Q18mJ^q7;Ww-Xt$H}N6ugQprH(Msmrj?4qI^WV6zFjPa$JyE&p
zTwkJ=qoL+sl41O$kC>tzavBoY8lQ2GV=OCcGO54VDU-Hd1M82!hV`z%=<7*$tJne#
z(|bxsLstmSMrut-cPPxEG>v!L|Ja1j@rhfb*r^^qc56b8K(V(06`dP#_9~O6I2|c4
zU!Y`{w%sYH^a4rBt%UO7R~@#L6`TGGDB<Wt-6*ffjP>Tjp>F7vI<3Gl)A4#daCgql
z);@uXAU*`Vtu`twpi4v%W`n8f^`f*=sUHH1_naH|Sp_nb;lSfJWE57(lm|SCmdW;y
zxNNLY=6)%pqNP3QXrQDlzpH6{a%_%omd+MX@k&l2z!45Ju{Ii_esSDdWF3D2IuRS&
z#n5v4N)G+w1%n_sWLJM=Z~pGMbv7rb2>C|pho`>XIWP}#=^Z!pq*}QxoPE+axD5F%
znWMyDm+|UGma4j4dt>U)WKE~KL8&y{%L+->%g_B!P{FqdZ8YMg82oRLqq`2U@3?)(
zF|x$ZHxQqUASFuK3Amt)tz^s6pQ!eS#=*NSooL!$VnPPnHUk%GHUa{VqvxK$aa4?5
zY&WgTbOp2OM*dxuCp~W{|Gxh=VFB-C=eH1)GEBMV<Ffn$>n!gzN=i{OUcyPml2vpT
zhI#=y)Py%M@ywX<@!z)~dbOpx<LzwyTm~BQrRhTVhk#Y9N%p{=O1U(g8Fu=2iGYT~
z`bz=4*~ib3*2hoOpR^h?u3I*hcdnY8nT6v6K~wt~-HFLd_H|8tNHS<gg4dvdL_%`T
zIJ8}3D{z>X9@~}KCg~-9OF~dyH_qV`8Z<Pp7XZn-r9MDKorGCgy=R2Knq3Fh)C)?i
zy<R~_yxJ;SuJ3%a_69#EC4{&vixY!<`_TBik`UTbtm>}uJ#z7s;YDxWou7m?AXIQ(
zcd2Kw{p^K#CeZWPQNJ5HQCmj1^vtW`&AX9^_H|24^OcF#cPsgQFA|SONWAlir!mRn
z+7{ig#Kp~)362w(hS9RAXoBzX(*kM(Ps`VFH}n*KACzQk`%;+gTASl1vQFQ~k~p%p
zGstJ1#i5_5316H44393X^R#`u2jfKR#)uxCzvKrpv;vW;<QX4jvDu0#b8BL;)uk@b
z^@1CVUI^cFQ{0OI2Qvza<|oqscEgoT=cLZxITMOiR*tqL60b+6-J)EtKJDeIGc2UH
z#b-$u=WMmKE{_E~`q}<?l%S|=wLOiw$*2dmM(Gsl#hG?LY@mUWDr3J^Q`ZXi{0rv_
zJJ)ta91NVb<g2N{lzy|haP0uwOlNk3^u11ek+4K;z3s`0e1!cO_n%_!l@rX7y)uzS
z!qUR$6{bn%#OWgcrl3EW-Ix(Zz(#dC2Wgvz|C|0uH?maib3aq_>h?vDTmw5F<_H2V
zRh5I@=Uu|?y?<#GlD+kGM3eQ2c3cYXd{!e9kehp28EikyXiI_$fhy-Ckxlh9Qrmv+
zeI_!j?}n?}O?zEkC&4vOEiIjDlh{wk==*vMkuoN9T2^ETfVUG)az*LVuu7cq-y@JC
z$=M#s&Cc?{Zjp*b)XY`3pr2o9n7g*d&wT8Wm5>DN&1mC?z!og0p8t62szaW}IW2=`
zMZ-M0%+g78DqR2d9C|7_mxo%zNy`#HHdNK#76_p>e*fSz`CCZy;y*~UW{_;}&><p2
zxW6t!vK*qP-JOmxmf_|y#-({sbB2lr9@LGbcai_|?p?sTf#oX<lCiBGe|oAqq$k$8
z1>Uxej<5c_FD3`K-ATOHUIavf#Ui;_XnELh4kyGuBhG-)diKiocBzw45bRGlZ#?vW
zS^(9*acL<LFF$+=ZTXH*2pg}hT6nWzac~L}A-VyvA%iv#k>4Gwc<{8p@Atr~HkD8=
z?cQc1@7^Ewtr6m_OxqPyUD98VR`$C6o)T5x9qeV&Lf<~>_i~e7)iYj8yG*_LeAl7<
z>?no>o4PuUVtMCsWu^bdsiJ<TKpN#<8(iRV5@3`i6dttKYiU{<#w<UfA}~3BS(f#w
zbeT^o-s_N6tcJPGPOX6oN4OfcmkeoVvETuCt(8x$k(sd~OVgC{46onvgu4bAzs0yk
zu}>ha9uvypo31EL`fta*yqsfDZ)#}x%MM(D<>{lHo<{dqpFvljE{M9n2nW-<v7dwh
zL4`_#ygxsOs)cnDE4<cz=q1EJv_P7f<jr0}Qn4qTW5B`s7^Elnr5$l1UereAWDeKN
zjL)$H-f|Tc>FAF>8D4_J0kaJ3xZpJJiW0o{oNXPsK5kt?R&sL_{v6O(J52hFxIw>t
z{(XEE@8zp#BK=#CbR`BdMN=M0<5b^3zlKK4bgAO^4gcY&6*0(jC3V{@YBmB7!v}t}
zo*JT9{c%0*rpfts_+kd2Zu?@%&t%tnZ38k7<SYyx_9M5n{>-_9Kn!i+6tQZHME)UU
zikV;lKF$u)w?i?}b=Zr^3%VhGMzqn+>(cEwy`P3qOJuXH+<Uy_q5wX2DJ{e!B&=Y}
zHvGUp)NVidE*Ws*5cATIQv=?HILu8Cd0tvVL!V~Zmz}yu+fAl*!()y={PcJ0UqzdZ
z&lS<b(Aa1MgB<#UKudxZ&i_`ykJ3MFwt(%<;OXUVmbEiL1!(+az_n(>-$?S8bAI!`
zXOFW%vX)TB^&(ZVJXs|P%*jkZXzP7`&ce-|^fWvGTch&j6f|2Ql3LHv0Pk2RqmVWU
zS}mUxTWhIWbGs<y?&~(rNH#ZbYo4av5a}`j6|*4WY!bu9g*@*+POZ3R>Rym<jgQ^N
zy`1X<jmb?rxIG$&pX(AHDzc?Br9?0uHDX`>EB+wyrJmaB&+EF;to^6bl(4gq*oR@C
z(PzP$U*H%Wq<BHdOC+z-m`AhQmkH-7RgNnXnzpJXlE(<2C%rdPDX&GW18__f8^4C(
zO0X*UG)eKE2G;B?d&o$-yW@chGN<4pIi8{2DS4hf%N$9me^hv$@%}49<19S>F~kH1
za*B+Zr}svYL#2v%>%0xY)|}&Q4b1tJ@kL1PC6$WhpwTM5ypY!kn5Wl)(I>wvUUY*L
zI&g!9E<U}f6|zC<M5JeTck=0LxU3toB0E2<-RI|b#N~x>u&~agX5F$D7H$7hDkbfN
zUz^ZdE-#2@V3FeDIo7}V9bW0@xntb>O^Y&%m^602m3%BKP_8APd*7e$iSBqqS;*kN
zX!;C5RYv=JtsJkk3f5F(IV;Kp8hOMhOw@0JQxk!>{}KCvbhU3<7UBVVDba-7udc4p
z;ug7IT1PGKr}7`@7Hq9M3miI$ulcA0<1RyQDGW|9*qnx6?1|^JMLUDtKA4Yu@pR_<
z?w5Z5GVq6xWR|}06}T{J6DVh~>S+qnz9qaw7yuAd$7tSKJT89Zb#_pw6`Mbuyp6j0
z`i~=aEv)B3-=e=bb(4IFF6@JGq_E)xYJ{gQV(0FArLB`kA2)!d9csVl9E`;K9Ue92
zIu{X(2NE4)tSCNPj_@>f>T}VMqR|gzDJ~-Oxo6nh)XN_x<Rr7a`81DqGJAI#oPk%`
z=Dk-mm>hZVcXSjdSnt|I8~b}c_qQjFX;X_0vV5lJsNf+Wo{_T65^Qmr%_iD*8ujX}
z^vyvH+`$DD;P63bZq;k8aG2#|>uOmOXyJi0eF!u2VHus|KF-sIC??rZuG-l8{aqxG
z2%6cqiIeoyY_20EDLqCcG)6{3;7gVMqFIlPGx2*(B`>GmuUuI#FPiwK4CH>o`R$X#
zcR2Cfzudct!L;chmuYClYn?RSs*e*Jcf@x}S*rR`t&~o=!ZCi2hUe;1reTA%WzS{%
zHNhELXZ5fzpk2cx_YPHyHPbGR(^utAJ@JzsPhGq%K7M)GY%QruMf_><nlql@i)@4q
zyoFN>4A{>cJTK)T|16C%dc-$E&J20&mM&ds##4{WZ;$Oi5Kf|7=?pn5t7Fd|y(F3M
zcHA2Ox6)r*(82*vP_Jz^Pq^B;@k1ZXYK_n8iN`kg)|jWO#TdEKo=73ZfeTj9W%+zb
z-XNi+&JNX-PeSeZ<KH}8?ZjiIweNT@Jq-DUGT;8@6!Q|l`fN;kx8)*VLV}?oRoYHU
z<;$~tjxQ~)peXzD`PH$PoiGpbs*(%JmAY;~sOzeh5z)|R={At2BF8jpdt^JKrcd=&
zfR46P$*Pem(0_7b54n0R?xRNk-SvV6ni1xv@==<K5!|T7g-m!@_B4l<jWF0}PD)N)
zWx!q`C7+?9Pn)}SCnB<t+eld46js>`wo-sm?Ng9m@6V6&EwG}JcIH*WX@A{;$)ntN
zu5Ku;xN}Zcw>ly84gAtCM$_T+q*FrQK9189nuonLcSr{AG)jP4mvN!DEqA6(JK?ap
zu7ED#*j>X6Qq6aT$msS@#ciJ85kCkH-A!5rT#Mg*yU_El^)^eShfDaGb|rqqdul%+
zjYNPw`|slZ;#;W@8_X{hJFI<2T+)6IK$@2Ka;0$-j?>Bt${4Oj1>x(e^_%ba%MK0<
z9BJVKaYs1rnCgGuLf32@#LZ2jhTi^SF&dLLYDD$9M>@qfIX@bt8fVy5ktX!2KI}oe
zwOMPSh)JMWcPza}c<Edl&y9;$RZ@W2P<oyd`=AuUHrK2O+<W1b-+ZrHnX+{$nV6b^
z*S4{*Ix)U&O-|tKeagtpCI?9(2RF>vhzNm6MGPO9r?mq|BQA)D9Eh?%Qsx>J3<w9*
z7f#2yy?7Euy>nY43mOS%Li5pNOqoEWCf{B)3gX@%^tAZO?)3g_#LUREUND*+(9z6@
z1JDWs&)yB{ORIaAk)H&Gg5u0O^w7C*T=K;AAUM8>2Oq+HXz27&1=>}2od@p~<eB#Y
zQfP9!+z#S@e9yAI=JlOPyH3BS^3!A)-RHg-1Ex=EkKBSpu7pIhnKlM|K=bM=6iK^c
zaqqEc`7Ch4VXwi9QyLTUa!l;f4WnCCndZRBOsqQ=`**LLW%&Q_VqC(bO+KA37>?iy
zyIh8{EIG=<63<Ro%BO4rrI48!$^YQqti&-L%!aXAoEp0p&ExM>zqc7ND5a$>a{L%u
zt`Dik#{17g?!$unuU;kV846t>)qhTN)02NqO*G(AiQT^a5g_~rf86V1INbDO=?U@f
ziKrx=_>_ooj=UfO*|82^o*^OS+rGP&fSC7nB2Hl)Q2w%P{H#ahrw&w-{oNs!m15)T
zd$AFx7sQ>;0+aE1eZSUcK<IzW{4@S!27O<`jauqA1qMNw>(z20Mk%ETP1NhgvOrM_
zE2H4bDhcE*FM!UEIa@~=F>XhHW*YNO-aHmf=KjQEJ+zAR6X=Lnpo0GZH(;MA`X&`i
za^GQ==^!NXS4Bn|?X%WA+`yhi!88ChyJ*()pV|vax!ApFY>|`kB(%WYk84`RhU>2f
zbby>I8ohJ#m+^G*afR>VMfZC8K-Zv~cU%Co&AGn$gbn4{!aLPS+Rw$TWXjk!{vp?U
z;3+b^WI_oD8lKh*$T?5BqL)03JR}XfARb0MgoS^LewF-&1OsQrzHW)K5UIln9e_?t
z{c~|ugF)X~+ytk-5s~n2Zks&HBF!%_IkQ67$LAzptdOr7u$EbSPzNS-7h9-lzjX#i
z(2#Sk+Shw?j~zS)4Cldl1Au7HS^23!A%dUU1eDD09>T=PbD+{4-@c;*EK_|8cvr1B
z7D+s2XNSAhUBvyk?1*8R4`yA0JFM%*Zpi-x4$mbfv&C;RG~0Pa<7aP&V#K9pu?)f8
z;V*wo2r$-BzIIr>Kc_`0Y9F;o1<H4zbP>dEvBLTt-X`j+tE%KawB6_V=x!A91F;t;
z{G+A)_caUeXIV=oU#?fCJjku>fdxef!E{{z$iCjaZV9_gaBSGG^}Q;?bQhLv%F#Wk
zJ8*IWm4sv?Y)9Z{mtnVDVHYk<_UX;%KJQ4aPUOqNC}tfka>cj0D;pD}9>J=us5DcS
z*Sn<jW`eL{Fj#NopGJm0Djlw&0Z0x?T^&%jE~^T-%2DhBxW?sEIoqM1AFOHOsK@a7
z&!c3>M(#GC8`&__j%%3kJq85^@mE}YaFLG@$?<0&L(<=rEIWqpaO{c+)(IlgG7@rM
zUI>h<i5K5sK;h3`5qq3{*h?-^&05LT3yKf{5HY{8pmv+57m;{O$<K%|0&k|)79UXt
z%_*sD^G*L*hlU?Dc{-EP4T3g3d-!2>k3>(4%4DyY7`bpQ$K2EADngk#>5tr@9$elb
zauU3{ab6ESmK8r&`ETPbqjF>d>mBC6WJY7#86^=zEh8aPnep#B!M(PXr(1csk!=(!
zy?C|Eqj6CaR#KlKAvmPNz|`Kgl_nR~N3Or00I*uWBIhZmMX{8W^>x2TC;7p`4y!*@
zMyu!&Jb2A$!??SV8<_-;@M;|ZosD|tTO+<900JVQjs8cnscisKNs2Gk6f_9@O&A&q
z)s1kPNhck-->ekuj1=TKg%)=rzphu1yA}l6v_H8B!?f0Fpg0c27f9#(iVb-_3Dai)
z!*Xj{Mh*arE;_n9ZAU83S&FN`GOnN~E2UD1H=5qV8eYUVf_Lv{6YOF7Hgj*VNxZQM
zQhMX6yEpGX|9W>rO=iivJej8S*JoUuXkvcukEUKRSoYTWp}7v5x>n|=NOVI9TdFVh
zhbd>h76G-8$8nYwzK;Tt#4!^dHb97FalDXx=E%DUOkk<vdvt@*71Z@ZPQoi1V9y^C
z!EHr8dt`mLS8m40R&a(Y0IgX3{{7#f8R0YpzIl6<juYo`b;AKtAu}?#ucmGrB8v;>
zpDX04&l5m_c1-%`fiEk4=VIpz2C&+n8y4VYh*SQDPg~cXA8Gm#EkQG-%@-?`!@hL)
z7+dzzFODBq>qw(wI5x(qD5KqeOi}Fp#U4EQ=J_xct66^OzdcO0#v{ec;dsgd&CMCj
z{qu<=?>mFE4qigj6xj1>6j<T|0|fsgH!wX{@YF2k@U56CU$7ivw-R6K_LrUtkOb}~
z!WsdTajo_mAj~W4LxF_3_m=$)D-A}zv~`CCr8=&zkUr|JY@O}dS~#|5x`@MJBjr2D
z=d9|$r}&3}#X12q<lu8QKC0N-U6X%=8vr;C0yHEbH69v9Jsi7f8b#s{TuFj%cQ2b3
zcYF`}^cLx>#hk!t&Z6I_LR1FcI#aLA*xz^|XEvP3+({ol<E^H)tKi}HeadebnH6qM
zy>05N0`^Q<Nw}NaK!NX{lzRoB;9Btvt!&PpU1b#E8j(2<w%YRykD<=M67MRNf3AYv
z4(n!r!L}Uu47fS5l$Bo%+af~dD!lNryob)o7n$xU;Qc?f<m$%u22(6;yeTue_T51^
zC@VN|>&ys}!+@69z44p9OJU%PyZ&vB87|uu!sn`EVvi3!ZG-o{`m2db54Q(}Pc{#+
zb1-;@bLmveWZ@AqH0s7)BeeE;^Hy42yl*RLxNV0DSRLkqkk@mXZ`>uP-B@Sfv-%_G
zXo}+smM_83ZNl>Eq^HQ7)U)HUsUgYoSoT$ZZY#db(%jiIxA78H@r{NSNiQKx4MQPU
zonFU+nu=uEJXA&e&>1B`21WH*hQs_s*FTKI+>x6_jNke9eSpefkba+O)Q*MH5ZO%W
z1~V|(6Yph!gCR;k^6pV^L?+;ijX?pS99(4%?>!pSivpp}3-l{F+p=$~JN|$UkcgfN
z-eZskIRX9feEN8YcG7k%X6jYG=V*lafz@U>P=mLbO_$hLbJ9^Rr$dX5D#!75jeBr&
z^Iv&X={lX_8xQE3(E^>eDCDIj^B*7sP*rUoukSnEng@M=$SFc~aQhH0_kz^n`Qf=)
z_r0C|#nau93%!fZj<*>23aIT9SUbH<Ns!$*oV;0D(US@12_hs@+x(aJ*E!rbU3xM8
zp@Okn|C47(F4GcI7ynq%3F@1<XX6hO#}Y1_TpI#p53Yq(g_HL6C=M7UB_)WLo>K)A
z+G)U%bR-DSA3`?{0C<xYT~VU+&DC%y6xfTn481;TdTorJtAQdMK))6_?q&kM#{?yq
zCDY%1)#Kv%`_{(_f?|Gt5BsXD(MGc$xBJcp!Cd7E_ki5U+{V5Qdng;(A~`cCHsCkM
z^Qty4_qOkC?{bijAh2qT37uK34_=Z#s>cT=3Im*Y?vQi-@!3N=^43>LLPMS6oK@S%
z>lyTONiw$OncT$86dgD5%D2+{mnGCbJL|-Z2^r3|j=te^ILFlL%f<Y9I)L2b7LZ|C
zno2oY(Zx~I1uymz&ZF6dzJ+<vsF)bZ?WM5+RQP~ac0u|OXY+@MG#%n}^z?KWkFHdG
z`8?$%w+^<48?x+O(QA<7Q}5Q-!XUE1HK*5NFP`n&yHU@E1@(53*Ulm@tfxyGaqAUN
zOAMA1-T(*=;G|y>S5KF%jF!)?KRv@JzSyZ}`hV?kXwnU4q5$R*mC(u^<@aA&fHq9!
zD(!lI^XwYJG$h265Yc|~oJg#j_@7d@)m9h)zdnE$Am3RmeMd=EIov2wM9ZT#M7HpK
zzI>BBY3#Ko@AB&(Iyc{TBH!S}+DHc(mmzaWDN^uVJ8h?*_}MNfhb+%fp)75LC6?Ix
zv8FCpfDq_MG37=XJI;G+t;w=2<qv>;AUBh1^zd$ia#x*o-Z_X3bis~Bs~s+8)G;!{
zx6U&EejHU%wAxOG^4oL0nLtwC%cMqtk%M5MQ(X~OfXm1o`o@E4TnrIrgC+~+cnz=H
z;(>`rQ>d&&2F;Y7QHbB#mx6Igb#H{jCfIgk)5%Q`aXa|1cVqTw)#%qOHZWHLUInI+
z00FC**4JiWq&%DWMM&M4L8BrC4yY(Eb*2ik+8=a%2TB$HG>FeG<8F+25H_{jj-+oQ
zy0-xFRZ$s!a74sF&=S8N#aJc##`(<;pi*gI)1I^Xr{O>RBKnHX<XPx!!tu!MNm61b
zn-an%dPwO?CCoINde-<{1Zw_HbWIX{_h}b#fy?mPWcUSYV0_?N-0~JOpz7Rxq16}O
zh~;SKKkWpN17CE7@tt6KnJzks7JHPFh{f&5>KgNdsSjqH0D5f!D8PvPf<R2sj?P8?
z8^*UpTieRoE<qRQHbltu%I4BL-OqHac?6l(4K!s7QzeKT55qmo(3PjG(U!PhrQaCv
z9LRpbM+!0hXzuPrn{1_zYLd<-A}`2~f4TE(psz@_R4q0M<ZcN!*$RX?{3|R3-2pid
zZO-QYQD$A)UTsAq>vPCQJACJbMZ}}#V}EEe4DTD8wZb8a1@<SnkP)49#DQrm4Q1TU
z#e<9W^PcBZ8zIf=l59YYwd>ck+$j(Vz&Qkj+`$E9Dx&8P+*KdHUsK|2C+49=GI%ZC
z2O>ts(K>|gxB_u(v^KU!P<Grm08rEQx|@uTWI4!1d5axSlut_iRoyTwqQfk&eJkh*
zDsry8S`-@SR%LYeHWTM#FeC;mZLZwBcUuQclR5a6BM|#;4J@vpa=GlTKE@>@OPmUR
zG6K~k2u2e+C7v$^mQ8(dTzMiOWU{_~TF39`UTiS&=LgIPA}v1y)KW@MMu?M3AMzXX
zJjlkfT%d?ZYUPuCo>0TI=u~yO{5tx1KZbiJTt<=Eb#xVLHEKI`<A_R*ETNQ%b!z3f
z_hO2i`hoe|7Kl`1cvHU_=TH-S1uQZkX++2BO98P@B@pMJi-b+WPunZ$dMA!sJ7>Lu
z=npLYtoxcH>z<V_1;EF$=<pDN=>*nY(}Yqv<nrhK&=!RV?=rmJTwctiN5?4F%{*U~
zqm2`6VzJy<#+Ub?@wQiN7;ZRDJG{4moyeVE+4BGf5vTp<=)*|p5EEm`87TJJDQE}~
zm)flHl#s3`wm5K}V;t;0KNZZnc3KBJ{?~H;hdgF&4H>(6mwa}axm)@yD}-?ivrH^7
zk=U`WntRYYb_Y6NhA$uKhw#KbKh-;G-?nA4NUU`*nhJJiWXArJ<+J7>ipkF!AM@1H
zb^VC2;m?hlX9kFpB#i(?3lf_TTcn}ddwxXM4}~{5WXKeck}+KPKhMO1pHAu<OaeWo
zRsJHzQH=uPb2d%h%|CZZ`)eatfH>;Gp@3;NBj8PS<JM0G;TS;L-EP98o@#yJ+}khI
z_N>@~?@%0rr(uX5F^neR8CJ}FX0g^mqHv~iY)@oc>shP?N>48?f8^V2<fj{s-wEZ|
zb=A<MGmOT?iUvLl;^##5fk~*P2E<&%Ct=}QGVwX#=p5up6CR>Q&x<;_YYfpmbrKl3
zmJD+9*LHdfRnhk3T`T@(lg!P+K-^!Og9zlbT@H1gGA|1xeKX56O-<-MZvLZfr{IWI
zxUQ|7Qlj`;3n?@YeFzinwh~k;o^3Dm_s@Gqfww>iOi`U)c!!H{v6iZ%JXZ3^piH4s
zj+vqFz)qw%AO+GbkmY)-;*L15)Gg1&la*>`duklg$?4krYJ~UY-Rv^bLV8ENw0R1M
zqJe7EpKS-L+_5A4`2f4X0xmG$m#+8d8fEaB!G_V&%R&U(*Tcc8GT$5E@0HU&=K6YZ
zmN(na2rG~Y+;DUfp<hCydcT<n@{<5U>_XkcI)iW>XDJiWaK0Os7+cPyKY@j#f_g)q
zvzGt9V_+xAM^;^o+_h6*kIS?pU<j-v<B$|3ibHQ6T*srz<nN(B@g&OMDOW5lJ{$3&
zV{N(NJ3!<pQRUyo=)0XH<QrB9r3B!_GA*hfRip7a8cwj)jc0AxV!ys!+XvwOfpNoJ
z>4KuToT#%dlzywjhXK{fi&RtIftxmVp&v7^5LEJ)R=9Ih+FwVyTBE%TbU^%YEM?;r
zuP9vxzcM)5kVphb;FgVI@wr8Ca?#Qk;%Wk(PVz^i@z8+!15-Cj?z*>sB(aa)AFuUF
zj&CHMq#r)h{getIt;@{+9r#i!8U?EUN69Cm{eKh&b7ks~UU&$YV7cF^4@w^b)V24g
zrNeR<?0Ws5Q*M%e?FvLz)(bAaBa<4}S!>D8EAee|m_(f5r@#R*UR0j1hIZJq1NgGJ
zAdHIj;p^}86QxNz_5D?xZ*UaMB#hLYZ3vP0DaJ;Q7hf-e!oSaG@x=LMgGl`omSXV#
z{24|1NKF&{^kn=yzMDlV<FnzLhw_Av@2-j_FO-fKaxR$ath?G&>Ko9JU+E2XU*Pby
zVaSVB&xwCY8qjAfJ?wCNp6aZ|26R*`wVL<BMe}ccD&PDQ5CFQN5>kOEfISeutrPK(
z54n*6iAPW1Ar?3h{>oK<@<~<&l#(2oWmA8lIY`L2okz|#OueL^M(E|WE3ucueq}Tl
zEpx%1f;Tm?FGi^dgchA+eY2as(I<2b55#ce{!9bf9a9N=blr;ScgOAKIN1OCbWMDf
zw(2CZ^hkn?VBCc&iBBm>R8&}Kw-E;x2oB)7T?IHe)20%o{BrW)v&CKv^kLQG)vMkx
zzXjf?j=ooXy>b-NGAA-~M!eE8M>5$m(Z^OHYfC?=s3@ogW|n~#-<|-%0U`HJ0L&u3
zTJI_%X|6hZuG<*h5{gRqJ67}7yJ7M=OZE`&0F6^QDEo!ZzgB6;82>8qxpktuh~-~*
zQt=MY*$=EJWbE~Y#TaPtAPSj{t(R;Z*n`whE<USDEQ)zmISFI`xj0nFY7mceRh;=S
zWn*VuF(^FkC1YYFrpV^$%C6xhQj!3Qfb`wBx3#sHM312zcXHA402Lh3`*cMC%8QxQ
zX9^O&`Jjr|rJC1?C393S+xgz{qulmpm%9YFD0pq+b{Xp(0qg~`CqT=|70xOPi(D}P
zm$fal7W}DVAAInh2MJFEZN!Ssb-VZroP~6p8%SxtxvO#dY$lBfLJ*}{{tuB{ryUaC
zof99H?fgt4!IO?~q41a^CTPAs(;2!WQ5-}0&UC?Z!^?90S0U&<pnE1Dlfqj(8+h3?
z5bxAe(_RPneBC`q?17w<{A&To1sApFfbJF-P|x)U@Ar1;bS3dB;5BwbDOid)7>arE
zX;pU@Ct{<1gu{YVIT=>Z^L)2PDUL%P4Ce`sc}XgOkiE*<{=i#S;SY`P491>Mw~3o}
zJ3}4D+zp+&&0SfGT9I41B$qK@*JtwBr|kbwODO+^XUjaeVsS&PKIyY7)svr&YRk9O
z80?-QMeML}k7rFlkisc}rU4cyKQN6spfe<A6<h1T3MSY(N<*z1xvV7yskC8_jSmJ!
z?zwJ%gMFLc5)JOI5%k2zeN}&cI=mIV_5kQ&U+CbZEZ}$KM2g!9f$f>Ure4=v1^DRZ
zzNc+3H!bSJgiY1vm*r+NUrSW@Z?6$yf>Rr6i5&rorrFUvfFWu74{2}3X2AC^?SPi6
z=zwaISx+g3`m2VhRRN$g^7d6FpY|aIMi|f+6#o+~xmkJHFe|RWV;S%ge(S~Z?Hd1O
z)kL-vv}dyxVmI(49qjr@C;~6%dM9S2!9ak$g(N;Tatalipg`EMK`uV6417%E!q(LA
z3H<AHzf>r=tkIe%&^!wXpn`KEe~GE_gsxNwO}BEd@i`2}cZob>2+eRsNGlgKaCSYB
z;=?5#^!C?g<9kei)!eWSow_sg(DJqWrEB;7_1g2TQAGAinyve_{!a^_9z^|nwn|bq
z1GrV&@Vjbt=e@+qcwVpN7I^eNT}V-llR?(VQB%S=9LibV6QJtd*;fBkzfVlIJ~1y=
zHB||%w`N(oH2XJNvL#nTAevLhcHSMXVCH$(-AxS*tBmbN89shrwodx4NgUbD-OwNV
zK)KMzvO@5lJD6#1Y8yZ^gaxhvyixc*vG?5Mh1&sU3_!08X_K|%PRHDlb{tmhAHGoV
z$cu0Mctx#B%z<5@E&GGHhvn}wO-1}{>1+ipu0*F^)|@YiYapBo{dwXhe8u&B>}xoO
zCG^1)6m_HU7BBp(i{#+|<A2M%0DBQ96F3BHPgCLn?!~WV4myHKN2UY!miPE;i<qLl
z1B?uCW~!?Q*X}mk<TSn;@a#$4i_Ya#;N4xfsS|t!HymBh1}nlY3S(#F=^K=>&xJXN
zB%#J-B4S9#L`%;T*@$AGND}g|&Xb_dt%m_5rXQi4i55{}*sn>+tIB2{#M;Mj%ZtsZ
zUcd@M?lCAJ<-l7Lwie~K(pF59T;bpY?6009=E0i_yU=HgpAI=f<j16xKunT(4o}ly
zng*fZE00NqM3>u`G3gR?Zv@yYy2Y%yQ@n?EOqy2>EuQrF_T3nn7o>Pe!5Y?<owp2m
zWg=5di5Ad1dt4J26lF;ubY5K}?xiC~T)ja#Kf>*V?pgj*c6nr{l9V>>YQ9la^}B|Z
zBu)-s^mbDmG?3YbsQ8c9ek~S|daO?O$TOuwr!2kEMrM>s(<hGLFjPzrZIoGSMT8uZ
z`x>E?&2!l|9bodCVpjZ)=LlE_>`fDExA^P#^?@EGE(=qljgDuSJ?>U+3k^aUbndWk
z`hWEM?;;EVVxVwnqc2X`D{1W%bYksrrUJ0~_%*6NJ%e3FX=J}2PJ;?pBImOZ{>_hy
zL1_fJW?FltNQ<Ru4(+b?;@iq@?R`57QS9<(gbQpxbP5#oSkoh48*z}l@&Teo8#_51
zi?`dycIk>8AE%a=`v<!gMT#3=IY1H;Kxt{YF+yCj-~o2IxQ{7-0B$h!r4Cs@$LPW1
zDH|Rm?WVEOI?G_VfJktCP*maFjLUMm38{*Oo`2pnx|n=l?hn56Nmc95-O%q_`VaJ_
z01|?ns%!#Zww#}|n(EB@i0d=LR!Rc~7a>kg0yO+_pt%GEEgdDssCdk*@@N1pCVbKU
zzW#sy7?jlifSA|<x>q+EiHIHxu7I3IpxTA*R~F|$4kjnnUxHROJ(JGZ7iC#BnsBV-
zmAJSup1Noj`DK<zehqt)^RifWBHT0%4#C%zc$*BNkwp$G1pbAkFj+F2KGS8ZVIGKv
zVg2(`ua>rf;S$eNG0tM%_M`V#-M*T<_A(cxnz)0!n|C_&%^GX0sW&q;Ry1+DJ88Cj
zt2btTV{Hx!+PJFL8Gsvm6R;C$r?=(%8LFnazYw<R16D1WJ<@@|$#~Ny%ChnzC1IS~
zlX`)O7v8409ZaXxUcSusvXqPWNGE{x*Kz%Yx3@oNCMY=tj#4~|&Uwt2jbg?2Fg-6I
z0N8vBc4uC#kK<Y7xUgu*(6wYY0fTGW!6d<>UBT;X-j+6EhAe=!C#)l|dDzQ9Iy10w
zGC$V87+W7tIFjbg%`9@%WW-DL$;)QqztsrH7oV;ZUz5p99mhJXy2ynPN&(}rAj9CN
zb-IL%E#D(Tdp8zoLci)G3m7{@+OS_qCcH0v%;fR?<<y<M$eKV>dZUMbuHj_=1)c;~
zc6c=Ry{5<IZv)w08bjo5pn`q<7Phux*qT`JZu1aO`-sooI7$R9b+y!oRn`r3Z#kUm
zf^7WFMLlScZ$n~d&37B>4Z`=Xo*e<(_wrQlNX5!k%Om`RXa@o{KEz9dshyrRz9>FV
z03$L{iVV;H<0`vI4J$AU!B<0ghH-EJxK_VYkVGMiNsQl$`QIz3Ghc(UD2n9|@Ohyl
zHNvVmFjD4Kab3pBA5)$m(-(|%L%z_n33s7fJ;KS!kApsR?iIrY8rT>2J<N^_HyQKT
z&D?GH0WrYH(w-q??!^Tv2zenrs*<~;AGH_`5Ugt6xn(XYiE#Y50KBZ+3(M<Z4ns{P
zC&uC6d!ZdY>qJ89KM>`A8f)0@Pp*@1Ke@Rv0Y~8iJ|p8wy)F2X>ccpehX?8VkEy#V
z<~C9$p1cQdHh7hg_kpb+KEdbdukO0nmMv>lVeSa@A4ga|RA6skoD$}GuwpZ*Y*V94
zuqsU`I=b8#R6IhhBEhT|rG-^Wq!Y~H>A!^@e9!`RaNrJ$Fv}4}gzQWQjH5F0*dLn2
zaXjMgroMj-Z2y&2GVhRQZ(ALD;@73LEhA48MR(TU@sa}$B5Ko<C%3e1k=LeZ!}Pkd
zDnAtKh2oo=<rQp*>UfO>7HMa>?`Q6uK6;6sOBa|6I-$b>KKs`l4{Q<_@+n}J*b++d
zIQ$L@>E8(JJyn6^YsBbpg;hGK{+yO?`jcqKDYmt@Q)J7DzMU~r*9{EiIR5;T`hjA6
zgg#Mvt$lLXOHLvtPK@B$e?lwbu^>0*WdCpmFl4Rh&=MWHbiaIPIiNe~xR#VnNz(Xp
zQ?1g}wjmzsH)g9z$*iRMnv2)ZNz~^%nfI03ivXFFH2><hkg(0Z>$dT;UV|`@r^4bY
z!zyg%n4~lL8~=R`8vQOF-Qrdn7m38d&9RM2y@-xp0lN=VdICjsn2Jya3@>rNH(Z$m
zDtB)_Td|(!(Qe#qbPT$0;{>8R0LeSxw}_a&PL)luAr(bFtYnu;PcNL;3N5B6v!;L$
zZ?#hF?p~!kz^s`WA-1mElGitdjsAZKPThhW1^zk;Iwj=)ylBc#+u3O}@e$UL$nnbj
zM+K>>!aWV_|LVTblv0jAuV2t?GwR4KLuUD?yn|u(n8>+o;x045$Es<}_<_i083`b;
zUQYK%cL%}L%HMZbed^usqt{(%7OXFNIf1J;1ou?YYsX{aMR2&hJB}$E8bTC+P{Qnp
zAb+GQ74e@Oi*&RBuo^4<eME=3FjB)JHZXGUSb;`%?Qzl316Lo=X9fR9Uf!EX15n)U
zo8wldn>Ru9dx7vkvigjg5j<yKjIo^WemvthqN0zl3gtLh(K-u8iJj2(Hz4<gCO%(I
zJhvh3tQRvynE{S#vq|GOldadQ4T$YjXUv0|&C|Kz*IUp+j7y;Klk7MUwf#(ZyTjDb
z{g^u~rY)#Zzqve>bK+OB7XCEOdZW<6qOIo&g329rV*+BP2;b5uxzfNDz@|<V<P6N*
z1p-zS@qKh|1W6d^6D6}<&^6kuJ3Bp1>l-av6G6TB`?%!D!ANcz0?^8ab~)Ey92ABa
z4Yr*vfYk);tfmeaVq)3BB26j33~2fZeK?N2YDSx$0{Kj^O~wHco4@0AxRA4R-x+bp
z2#0%_SS?l1mfHmWT6VJ~wJBTMq3h1`;h-_AVz%Ch?UC;m4<$07H#Mvmabggs4iDQh
zZbX#_R_$DpY;G8Cf$8x}aK1a>HJpKMgC9S9eM)@;=oFwAaZ(?J)MN@mo?;zo;3Ni7
zJ$4x_hDuP0qdWW}9ZkWWl)vezFvgKt1ZUCy#T6uO@&=qq0x#@#Z@i=+QCrtPHzOg5
z)6rAI-l3}bZIECm;XqSBn+-W~M*J&xJk~Qr%~d%Wb`G*QCRv3F|Fa}<QE}l$BucR}
z#Ad90K5xy+G?kPmNEEUnaRtFFz+HnS3=kgw>521l$fg_;!s{^fKcx(xFyCO9!(%b=
z+$^s(o;2d08sUmdOKV}`mMJ2BReO!_nue0HyY(X{h&(mSzU4C=!noI6Pn{bLCwQW3
zt^9fH+TC^<R~(mJsSCk&i*UKcUTTW}tre$)4TJggrW}0avd`s+-yCr$jOB-MOvhM@
z*+=tMtzTreXF0kj8;c8HDME@jY+C0Yjf9qIBMt=)`O^<<sYbSj4XPk~M%HaM92q5%
z>R-O();akSmlj!@|6Na){X6F2f`94IuYjIOwiC_$#zy|{$df0vHK~NsdG%2<;Z#m=
zN_thGafL+u^%LTQO$O03V(VO{7ucQg&EdO71FzwiQ~BY?6IBL!eW5vmcsU@*EYE8j
zG#C3T6cK|ADkX#rhe0Ov++>?VYNJ>Cxw(3wED~eq`<~R)*step1Gzz&Y8lhI>hrKv
z4ibj*Ti6*=LvKF`QQFY-i^=}gMxutj=XZ`f?D&cEKz!aIW!Lucvfsqs0w*}F7~~x9
zmd^Y##J>v5-Oeq$p6-WVMi?|sJxSHNyN4_{@eYkwbYjdiyi<^ODM@HmM`ME~X1{@F
z*v6VS6vnAH2^yH`pi)SgkL<j}=dYLfq2J(tlb}ed^)O;KGCk-%cXn1&tcCj}=+pnB
z=^Uf$`o6E9#%#>SXlz?en#M+B+iq+$wr!)aZQC|(&{)su_xJqA$eX;n<J`OVK6B0a
zS!?_Lu@BhesfSt{bRLGe*Uxb06&W7+mSD5tjqMF(kX&W$f@Y>KV@AO_BqwgcR+Yed
zZ?>hL6boDxhsXBEJr;!^Kz|pbeQ`1fdgwBeUYNZCq13=gV+}b<#3`ao7Zm(cBl}t>
zlM3CLDWq*kJ8Y#LTg(!?`w3COA&O*4*|7bNA<$Q-AzUp#xNUaYclL($)fjHJq!wo~
zJx5!0!lbX&CNVhRQ+R~+1`T37<5Ht7vJc#BP3&@`wFos=t|f1+RQP?5fPe(Vl&_!Z
zaeBspe?~IGoinUo59|!HO?i6a>t4#1&uzj!#eM*8*-DG}2p^+@k8@cvLS}ZlBk5Wq
z*tAq-B_U(xQn1pU?{y9mqk=AHgxo3^B7c#~aBWaCB;tof+si;}Yw78_HdNGe_9><y
z>hdQzu)=o4G2ZGE_;=3h;~jN!vF(+8zG7C{`!GZOt)nyqU<_kL2>}W#sGH?MG$GD_
z(ohJv@~}k|m3R?jkN_^3H2I?2;7R#b>)<az8akK++MwoT|ILJ_HCj>?tev-yk!HQf
zEyI2UKde2RMO;&5ZCK{x<_FgpD^~21ymSqTuYL=YvGymwMb#ti-(oV71jF5KdvtQb
ze*acx@ijRZpWD?{v}kGRV51AgP1fJ|!PwgC(12iK;PNZyr&+9DdklmOcz7|cR^Rbh
zvDPfGIZNO@iPU7wDtoDrdIaSmd?#t7CqEPP@1Sno{MbZGO1Apim79WMNK6*+;vl;I
zjh9u)l+XJye~`MxLtU|k3M-5M2G?<paE?tEjD!hx52?vutpVbxTOc7%1O=-9SiS!}
z09C2A4JifK=(P+wp^A&7$HcxUXu<lUiJ{u9=2>GOF#Fxarwpl*F;=4E6Tz)Kf!#uP
zURTbU7_|(V>y2<fy$F(~1P7g_`dTR%-_@^iJ2>8j9nCWdP9G$x>3eQgLQ*c9tGRP%
zAI;|Ctp^9(O1Frux?TenwF_;1=&3lDmlx<$3V+O{CsP+vwBF|?B=Tqr5ZCYy{l<sb
z2?txc8{K`N=MQ6PbD#3^WRmRwz-u>CBCi|6AI>_%NKg`yLpf!NOZWE~uNjrPQ$ucV
z^ur;4EWhxc%OurSzGIfrGE1kH{ooLpoN$@{<ie?Ai`f_=NR*Gd$Ihw$w9lVFdZa^g
zU!vh`zOPYiE9>m6?xP)UJtkh0k#LO$9uPCh0|Q+7GXyGM2tT-umebZ;wGb;Un_o}G
zw1%fx#TbN>D*ehBWL6(0fqj%$Tj;0d!ROCyFGgVea^c&xFPg#(t8vsIEKoyKShD@c
z$vT52JkRA4mz>GDV&*AZ<^#;@K??$$fVNbQ_Rg7#hh|cyWyoC*WXU%NOzhz)C_9(O
zsL-9@r^YNnNA>$&XVCKZ8&&oIW-Uc?Qra>s&l7O7Qa^0bP=sKrn*zrE_}lVgwU|H>
z)ZN`Fzg|-oPIV{h(=saiwT>#dIJ$u+WkHW9>MpOaHvOV=d8{i)h?)UhLF8&qO?q#%
zdW0e_@H)I)|Mgo%7lYD~mRhu;cInER*!k%`ZvoYC>MDpaVHPcdwYr}LgkIw8^?4E&
zySZ#i^}dMlkLG&|WOM}6m5n%BJUqcvhol&H(v(2A(@}s=xPCjZ`uSc7t~1{LTmB*f
z4mfA)(ef6We`eAqhQHU(e^{^hdTOQPhbk%rqu_^M7xC6qKFP2{*Z)-MIPag7AQP`2
zvxHb{`WP;$11r{_TZCo4R^QC5gcgQGC82??{U9#zGet4G+MAS3b~L5*TBx8Lr?4uY
z-?Zo|cyz@`Xq54@=CDfl9%?l18*$Ny$2N%8!I>10t4x4AHD)C`cpvsh@2c%O#K|L$
zIot{=ojGFhs`drG?7hy!LA@IoTYat<t%6&IdA^Si#^IhG9I^%|o`f_Dt1nCDUo?OE
zUqn|*m?AFmB@^AK*$<}Cmn3^VY#~s5a5WOF&0A;+Nk}hx;)5}d1O@tz1nz{Ns0+9w
z8@*w=g@f55vtM0IuRKa_QmP%#yI^WXbaC8^U}3*qI%!8WrJ|ai@4b0t;%ed@*XT%}
zN9@d+r_qcJ27MmXWR_<%i(hZ9;ljzApOTh7^_KzldC_8g8bwqssNfGg!WS#G@DV9G
zNjF++#I5~q+s}qH7|Vl>BkBGE&O!LD6*$c#?YbUfeIaR8?k!f^daR5JCMJyM?DBFT
zzA98&(S6gCZo#rP!gGeaMQTdmAhEpSV&$llxkl(jZYCydBz_LnZui$wX0P}th@_P}
zx~oMH)u*Rlv(`N$3Zpv+)wGO2{l`daS#V;U$T%=M<vQ)l2S?B+Ne$}8#Ie|ibBoSu
z43R$NAMLTGe#vQwF_@`}mJ|Yc*xqrI1#aKRm4=SEC{C>|$2`-2YOaN;`UEh>lMQ9`
zj;|}3F&N(aanp@x%2X*vWvduT$e?fr2HPdYTb8myLUjtfLbNz=-~JMux$me%-zN)M
zWvH1JdiZIZHKAPoS~#=EgqUq6*lp7Hya8^t#<4=LN<&j)y%B2tVuJY5ioe&u$$hRl
zGBJ2_0z*Y#9a~q?<tT1dZ0;b>6iltg@nNq0O3=8a*0v=M_HCraGkV1;*>fSVC)9kA
zU%l(8_vFaA*@N6LPE}!t^oOR!9|!krQO}k(+7BALI7lj*{zmKgA<bbxseIJCp@9+d
zxjhc$U|<@g-ex(H&5Q3_dGRED;q;O@QAxI+Y2YLohM!+5-vo;U-XIm)xa!SuO;Yg9
z^pvCKdL7ivId1MGpk<4$s#f@?E8&0Zn`=)zQ0M_VcalrBSWY!}Tv05sAxS78dyHAq
z8flRmR1Td>&zv6X`sU|8#ef)Jy=_+Q#;|J7h|7Mzr<!C#MzzPJc45;tvE1swEn<Xd
z;1dZw@F9<Zk(6w`v(YYQRXUP0t14Qd#&r~gbnsqKz7CA!OodS%fsGu-^<bw28?9XM
ztb0eA<h0~0OaOjkS+G#JOq5;T+<lX_EeSBhuirRzPz@w$2G48<p6~dM65Yh(=U@hD
zh{E#G#jp3be-n87h=@>5=ZewyyE?SK?hY$xp$KFY4#<+n`D)p+qC?|vN2KXtJ!^me
zGZ46a6f(^X#he>&GS^QWAo$=Oxnu6QHt4Zo+0nu$h*4Xe77-tu8D{<a_NhUvGdT#G
z=jhg28E!zF{VAsA!=wn>b&nxGQtvlM3!2*UV&!}inB5%I9>z*=y9cEK$onEaB}5Q8
z9?pFINg{u;_xVC?*~QuwpA45dgO$+y@A9FTq%-DdEg4lqLz%wG5yOO^+6EQ;;*(<r
z^oJ%XXfw=tU<zllWqZF4;`x<(U3RbE9@12$mcgy&cEMArCa!|#;rU{uL1*1#+zUHa
z7}bs40h;sk*~rn;nEKE5*p3B_w4>e6`DXhw`<9=8nX1Z>u^VoiTzWQU=LRp)rvo?G
zq(tNL<&l~jN;HKO77xQw)<&}F-9A!SwFPecDyr<q%YBStXL}60je*o|O`@GYXj7eA
z+B5jG6BB*xdKD#OdZJN#yjtjY&?b?IPFYY0&Ea=isyJ^!5Y;USkybz7MV2(-SWmkc
z7kZ~C(wjYpM3rgorc+F5<B6BbV}f#nq^A&$E+W23ORnA68-JzM%}}2lh|O<dJGI(y
zjD2ujw|&_cc<6<?hXXs4+v0`%xZsSdWq3vF*&?bn1sx7Zy5lBAZ#6L5SWt>6k5hvn
zy6rZDhQEJqOWs%LzN#L(Fd|E5PF2QbTPfm3rha@T2^7<7Klz-n$2}Z^oTsct_6S4l
zB(qc0ZdyuZ+a`KeMED&?8&9ZYfO(z*w0ZIa446i><AWs!-u-XC;Zz8EfjAzw2w%%i
zaa(S^Z6eb9+xRcPTVQ6uvpn98buYC@G>7nk@Cmu`L*!^Lw9|8lf4P$_wq>QqrAHpb
zKQL{$G-aZuVZA!5ro+MBW6pk#S5Th#SnTm}E_^0h5ubU;X04)0&sY7ffg$XN&C&pq
zP`Tw3S?anaSb3o;EUbX>O*o(7iP}J+VtoC9FZ=M#sEMySH;~(1>XFX}7xcGzs&Fbm
zu}ZF9W9qTRT`H_}R?WT|qX{+(4t$VWiMi{x?B~ougm@)fU%%8U;%TUMn|>=uVZY<#
zw9i=EaGH?<vsRPlW+z}6hDDx+7Ny+_B`yw5?o6r~0pi=g`cCz&th8JR_P1SWBrWlk
zF{rqi?+w+P!TueL$m;J=i-H_F5l_j4Vawjkr-!Q{xwR+i@w+1&bB;PaH8pAfPqkBE
z&QZ135~kU|tq&Zu(v!JvJnS%)o{bf4YUEfB<a-c`Dm~qU!-!m*^S?X4&}tqSe(yF>
ziqGP0A~`zr9hfx(U-`b^feqRJ_^p9wRVAnyy#jDjCC|#aM|O7@2ls>YPZ101CxtiL
z5L%jHg;|LE`^yGrBE#$oTj9^*#D;mB)}&-`5X%VfThN_!84un(Vxuh(T-H>ph47eL
zS02|B`P4mP>Ob=)57-IU^knNok017CKG@|9y4SHUFvf}%_QBQRJG6Gfk?>&kEjylA
z-|P+~Xk_57C;#9DEwTHUO85_FWPRio>zaavBO$e~H)T+Ao=7jwFX(H&rDtwFWk)L7
zfsd_6Z>UoF=nv~1W*MtPzF10$PbggSKYrECe2k=c)BQTSPyXm0?&;~t%xT*0t5Q4T
zRK?h&daHE(%bP}LDq<PovF1pWfx_MfsA-KCX|l``qWRoA<Uh)DuB4aJn<;kwdO~Z`
z+%ww`d8uoqO3_}qTQQ4^ZD&T1H@yIZRtiNFtF^058O_wdg+TmCjb5wSj0vxNR=^P8
zgH@}y9Zl&C4~~~WbswE|k{%zh(vQ-OY&Js{{WN#pBr$|}M<-gy4=_Y*-BTWJM%d8I
zwvvqnJQQ$n)Lyc*Cd^%4k_M{fW!2OkfIDEjD4*3c^p=KeGUavIp&3}^iv|?j_Ra@c
z7NjpA>Ui*>g6<AZURCQ7(GSsjBZMR1aJyG>1JWXSok8_QKPwwg^vb?K7!VPD_?cBR
zWt}?Y_oXlh*jo|Z<ANwXr;LY^=I$i8`O7-sSRv6+mm*`G6<{?tlwa98f?xLlDJzV}
zD!YW=_T&t>3UWaW(Da|WKKX4h!=kOxKHTbDl5(h?=l1@T@mC|?>kF}WRB;7IHf?0q
zaO=FaFYfX2&0-gDDqDJ=uAW~aA)5SlzA1Xk@vD}a<@UzsaIby9s(M|n!|Lk{Z`u!3
zvKGb@8iS5rxwTpl_m#NeRo7iW_Oordf#G52Q5ahg6SaZz9t~7sqbicnF`@$;PnqFT
z!UeI!G$ll?vTDhvyRSsPDe8B^cAJ!wL<LA+5#t$z@%W_IK1GQ4dqbeUaJvMBQkMdM
zX-71HV#JwVgpFHCEk2i6kPFg_k>Wl%J<-|EUA;{~<&VdY(N^W-Ig^V*rrGLW#UggA
z`e{Si%;)Z0LYRqDKzX43!U$w{X6k6k054zamdSM22a33#djV{RR)Nzsht9!{l*AIy
z6|~;aEf-UFC2|n)FUE@Ib{qO}7ltN^TLD%paf|P=U}8P34lSo}dJV|W|B41_*&Zfm
zXb~ae*%*D`|Ao%8IMM`v&p<P;cR0fLpMfQSJudi7LB7NOC%Ik14T}oH)XG!8g^HJV
z6#gKaV-kami@RLwJ={-vDi{`&XJo}G*WzutT_Kr<OPSW^!Ft1Et<5ow{$$Fe`s&)-
zs$?J2s!*PgboBS2DHI6RIT5S$jIqQp?N<F1mq=j!Qq`QameZT#5I-pz*H_mOpe$#u
zY+M={h2ovZW2REK5t>B$@Mv_1x>J&c`7R!&3+0Z8WF@D1tK1WW-E2BNBA|mU8Vj}}
zy)#D89JKzSYXU7cjv{^q{fcP_*tC4;sATn~`#FU4l4V`nZ;zFIKyPd*1`QO-hI%=B
zcZx#W*ZXHJOrDU{#y$h1vlkfYV_{5CjxNW<m$L1OO=`ZX7#KDaZZpnTx|zga+5MQz
zWP<)A2UWbPTc=Kn_cn$DgyVHk@d){QzgEWUb|bASmSq|mq98$3U%w$17dNHnrMQr~
zt$^n4hmr<;8s5GrjgHNG?AnxC99>*yQa+G%mrcVjLa8Yj)S+$Y?T4f5J5^f0uuB94
zH=jFLVQ+5YRA>0H+<WdvY)};z>M4350Ir-xVprQ!2Di8)IB*=UTF3cVugE5!Wz;eh
z%$kFPjiA5ZZ1?niFl4dj{M)(-ea=IvoiA{}7OK^JKaZV~xRo=pS>3z%bu52#%a*Qi
zqAw+%e@F{HMU*-g$$;qY^{qPo`Hf7oBE}dm%Uk{S+Deo2R2B>e3Sy~WcKhxeDs6BJ
zhd1_xb!)dU|Mnv3Xc4=v)&RMh=Idb)1#?xu^IDUNe9D(SEd=gWtKF(|vkS%?)C0HA
z8AXsnEe&ZSZG=WDrCy1%v8{Y2ri|Ob{*-<i^!`5<z;o_MV-mrUUT#ijwjBA|WRf$|
zr{yDRv-3k`5C&1`*^8jDGF=wB%Sj*xHZzVbn!Z*!h9GBdzOIvSZS+85Z&J{GJn)<J
z1HZrZ5XXMNOKu*%rzdKB94ZXblhgq|;@+fMKybN_vp}~|uiJ`BsP>w`L#Y3-u>x?L
zi-lFXp&ZoQvE#97JwbZRQ4Paban>~i@{RI_JfVQC|C^++5T2`m!H3^+qpiXO#)L^=
z4Aa$(11zGz%0ctmwh#J>fQgAKPY-&?o>LcAp}o$;4;;#2ja2nb4}!T052g$cl18L6
zr_8y;T?0ArZ6bd~)o|;9sbv?C1&1fr>#K@jM*Wo7)t!>@uho)K6^AJOH+_;|sbUL3
zm$1%L*}?(|86B?|u_0Bsv&2@5{_lVP9U0BE*YL@PKuLpWK5S>@V<k2s_?hF2h#NT|
zZ7qaqRZ!>`X)_SW#A^!7yuN%_Lx<eD^T>iVw97w=mPqqqDfJF>iDTWusxpvS`3%+>
zMq_L&fvI*B;%Jg8sE-9f*-lr1SZM^j8-Cn<ZM2rmHzxc^O{-!IDo~0iw8U;`utAXQ
zjHWUCW&6<c@s%S1bXbnw#X^7B{y8;GqtWi=U5zbrL(MTzIKBEHtG~28#vy%<yuZ6!
z`LK{*ze$=JsFcjVB%xdOd?gF~Ub2<hKS^+nCQ;4Pe@}C9Se-qH!TkfQWCLD`bTk!P
zbGBVs)d;9PWYMq3!Z}%3_uL)ME)Kc~1^zNrhm$p{O#PnrogUFyR2Z2%6g1)N)TEKz
z2|Fx!it_zgB{yb}o!wEf$U5TQo*vuZFqBTSZ{NIu;ZdO9{K%s1I313H8&oheS4LrA
z!3bO&r=me_`ZP3XY@q0nUUWr)<)s5lA%670FH-UDfe-i`ZjT@Sl%KYH0zy9T*=^Vd
zTXk+8$gF0a(urGTj;soU8cQS~jBvT%kge7lbM_DNmaSm*cO*JqNRYmiTHLJt!8NfE
z6RuQzCa)^FwuSm_aFFhwxSr`&_!6mkSzLi((Ey{CxJ#-sC{f&by*IK{UvTXDxhWT!
zkEcLT<L;^5$*EK7V#%RryzAMx&cPBHT8*c=b@ZND$>fjE6cm)rMC5K)>b2Y0bNL1a
z0p7zepza=<{n$f!wfzB<ihv9<ox>%XVI(dTjTYpzO+BpTzJq#we#(#Foj>iEetp9d
z!ibriRyRlf<`Hx)qel;nT#~4qVYoKXz6HjEk%B>!&2H<j(!R;Xwg($HoFc%oJ;v^)
z`wH8k(xE{j?lFp-0{v&f&PE3pJDD+}6`xBBhS$UcZOKhND)m*hAx8wNh~8N&qHbI!
zMHZ7Ta^@9(M5uA?iNH8|6ie~82kbA+R%-1uKx)_ngAH@8LZ`$%9Xsvu5HWg5mBSK}
z=9kEdL1sy4CGkd`zvme&Sr{9ewc)@%$wJ?5e49wvC9N3q?q_;}ht$<gz$1u*@rm<H
z2~iTA<nuM&94gC}MfhjxbXAMXZuvPawlI(%pDTc^-A_muPUSD_?p9{lKDHR4yCV%i
z;=badu7_yuQO-3TaBJHNt>_6&E(ch@BsC75b~n017%W|u{ar8@3XL2>`BLrvdVcA8
z?eWAgN`J|NjuGc&zzh%l3|+9?Rb82q{qcKai{dLnHO1j=8hUwphqv?jb#<Yu%I|_u
zs@97^p3LDOO0BO%oX~2l^cPHZpXWB7D7|Z?z|1;rxb=a}S$B4a+tKB4j)3Uje@ks0
z*ldh@HcO;zxTIRW)o6oh@KZM;Syl<uBIkFJxEk;Q#tWXQ4Ff3<MZ14_x#E%?q`-t?
z;Y7!wp3vogAlTKjj-Fev*iS+Q{_=LFj-ox9u`ELPu-VV1B<Md*@Q@iF{AV_{>C{u!
z9DmEnww=8u*b(}&SjWLz&Q2;ipDv2Z)cY5<C!}G$Z34;`oH;Ny;8VbdH;cQ!?@k<T
z_Z0G4tIG7V_xDe=ay}7CMUC!}(f9C<ZJYX@0k?1LPnzym|6Wq0S;f@AUUGtkho5s5
zoG}R!&#d&Lbrl@k)YmPcZ*OQ&KlF-Nmy8qV&I-75``j35$qWoDMZucMb;Lw*>vZfC
zO`IU?5O>GA;8~m)YD#hb(F_)d1Mz6Roi4rVE_kvqdFL~Q@TC`B+eYs(DHRr+%BIYc
z-oLkk@cYW2LhQ+Tec)zTm>7?!fvfRcH-hMad)OU<FG4B~^+3q7m!4MndOGKFRV2*e
zc;vUF&VQWwIdf=|#`<DH=Dg;RCF(xy4>itU<agNU`SyY4=^i?+>o2ZFCiPYxu4+)L
zQ7IEZ{kn>~@TNgWr!_FA;kc~geAA9%f@om!N=_)5ChXhe?p%qI+&$Fi*|OISN;!z{
znrL6Oa3L7mtvt7N=f7jH32REul9zY8KIQ{Xxh!p=cQ;YgH5yho{v*r7zpu#2-`hZ&
zwKN<wmF$pTUN@fQQj8HejoY_b9;+vRAXNxD9TPA_@}#@jB4xKUaR!ehmk%ZhCCNb`
z<=yLARVAmM{W4~5qzh%L|CXa-;O6doc+sE-_=tT_@hH1IxyQ`5_sSP<7YgCOYsTMT
z$8BEiXlQHbUUcT%OQ>uAIF3a}+^5MU-WlaEC5zmvnqKQ)BJ-`gL(4h0pemrC8te(@
z??aaOIjx;yrVUvZ72%D2w>50HPm<b-6^sY!g9@B0ENsv(UwqC_hZU`SgH~;cxV6|<
z-qA4E4kQhAZO=7Td!W1WCMhsjX8U4{3ai$4TFis(i+q>7hG}KjuG(A2Jksv2Hln^F
zC6?-wj>jrHb~$4&1J@0Bc4!vXO9R47&zshUA}bi(EOlA&{C*|5u1PIiFNS=P=UUOD
zH^>r#3P!@rp+70N#}``zPsD&$zLEmW!o4yXsm*!>DPyR?bQ?9VpQ}GIwYKhau<&mx
zkAfzjCJ5PHUM-(E$`bhBqYh^cMPRtdmBSIDr=Rwq#;fY41dwPSOx0TiSsh7J)*0hd
zOEu^uK$3e%b+;q1;Qmd>l9}bbgbha8cy7=J;T=OED1@Kz-oDRYpYqqDU`1)fw6s3x
zCj+n3q~K)7o$bDgHaEEWMN&79>{-*hV&ZTdhLk|3LA!lmO5g46d{tp6Re``8n-o^$
z!byjdOQ~U=g!DyN%Bs|o_y4Lm^#5PQ1p+nP=Q_bEF6H#G2QkYQx?MUtN5O9byU)Uk
zr6t`r*St8kUC4mZUP!Ean*nJ{l<;c^1{TuqDE`K+^>_Z?;0S?wWfHFbx4ytPGp5uu
zB$&Inee}M@T*s8;57sS=ihVs9jyPgqrV6dvK*msiXzaSjl}y8_Zl9A2)cnnNAtZ=Q
zr%(NyDJZCEwAG51o^Ge|T8{FHoOyxEm8KqzPC}(+{eWGl5)A!Vq5IX0&PrFQF(BXh
ziTvpgCm2W~pn?(3Xc0B0K@=rsFe}=rQJ~Zh@bGW?ccRq^6#L;AS(%Y8*)=t1V~bMK
zqC$EzVN7nt@0UKGNz!5Rxf(GDV!ch;LTnNLycgT-LT(`@!hj-SNrT#fTKC+BN`7gg
zQU$a=oX<$e3G-fha#!pdJF-dV-c6gfY5CFUMl`C~m7!;|*F%7RuWL}bJO7I%M?$fo
zl2Q9NGSiUHyM6^H5uZ;$HC>p6Yg~#)ZtwP<g7cuIqbI>z`J?hRd}QML%Mc)Usr7uo
z<ZzM{{G{8Ta3vN&(?U$j#K@mI5|ld;Hn(_-?#q+!8!fi?uxcVYDG<zN23t1>>AOtd
zu)?+2w}pLnlf4f8k&ZI{Jek^M{3fju{@Vssc&F>TzyLX(JKYn{0*2>C*`Q_43#b4E
z-#;b8|6mSz+OVt=S7Wes+Yu1P&rQdAxImc|v7>u)9^2CH_V-piK*JQ)GpPwfOyQC4
zg1MN?ilNClYQ3x^r$eQ=vif7<R-+!AgblH;Pzg1D!&RH@OJt)#lP>+quY();HWUyK
zQ~HXm5uOx%BA<^F@HJN&Dr_COA5XV=sx`0LjIY*ZwkNlRN|8>HE=>JE5c|UB{L33~
zP3qgWPCTPt9yp*s7U;D)lg{(ieV#wHcPPm22hf$@FnT?ck(;Hbw=wXVI#u27UNf}V
z`1s3|6H27nbmestj(rDB=Z`usiA>=UC7mA+UXi>s7d@aJo`Zt7E8K?05cUe^Q1T3F
zJnu;)?Q{2VFbaD5n6x}gMZu+{&(?PNZ6NYdbJJJh=pSLOL>bAxd#(z&f>$6@nJTG3
z_xaEMAvm-3M<LAJ>}zy<$@0!@q1u(k!sdPuH2?8ur9bg$$7PF}xsZ2|*7WtQmHfY*
zCbwVR_2us0iw!|F!Df#Ay(z9zD{k)ya3cd1l>+XL25=_3P{vG&EChm4(8pGB5RxD~
zLaSL7jB9GSo718Ve=naE)WJ@FTpe6QE>no*k0kmpvtA^uZbth1SsF6#uP!5GR&6QP
z$}PP&Lh;CfLH?u|On|ImsX+4u@^q9n>}MDu9!q?G-4`?>dWEv`n)bIcZM|P=zMpYX
z_0*hALcW42dZIsecqSiSYqR0Bd<qTARH4RZv{fOxi`aoxc<#XVqyQR*-!SSIH=~nv
zPKr1XH+j-r9Xt0T&#SdZ)jpk<j~cx9eot-K+LARX=+!e^Og%}vw7TTF!owDeAYa~=
z(-CDxHtK}%-)^)N7X~I|V{>usrXQB60}^+&V~T57pKTZ`3Ded{<64~Xk}trtb<+r&
zZ|i4i#|?p=RG|Qb6A<&^@lbAeU2q~gK&w&`@c)caNQq4vsO>{q&mxG-U09nV9`&5=
zNLq^)_Z@iRUZ8GZ5A}i}wpp3JZ<#adwcF|XE4SH~{d=VXuPqC@Ol+ygL`;D;CwuZa
z5I8Po*?9=f`^{7i3y=`wi%MAVc@9n>eS8uutiN^vPBoH;>+XV@?%nB0uiTBk>4Gq4
zkB^z;#l(c}U|=Y!bU4nzG~9F_<Bw!znI_d6$HqI$ZVNFyW=Th=`uel3jptc3I-<_V
z>$G%~jYlN7Pkx~z+OhTLZPaV%uQ0GokIcLtelKUL6R6O(Em4x?XD^9CLFLCMTiyik
z^q5uE=s&%YCk#3Nu$Te-)Y@tyW#P}&V?6?bSt8(AvPf^V1cEUt4#d#Iu~iv{A*yb#
zA9MI+-IbPVC_iUhw%aW$IbDe&55pLSTC5&rr1=bl1Uz92i}3zA@lu@qBz>r_ub-GU
z-cTAZG-Do$@)NKEW&&}NeYwm+rK%yB8?X%&S?Yorz|8-0wHf57Y_#XpO-!UcRT_;6
zGJKjwZQt7a<f3ZadSSpq%FI}f+*t<}7nl+g>NaRs@E-8avcmjdVbr$$pCz9wN@X^!
zMy;jCAhvn>Y_j$F)BXl8bA*8A67#;{_hY^{!qz8x5mn^{A6K9Pr--GLDM?tx+s|B6
zsEf{Jn=fDRFYUmWzAtd3WRt{cP;j?p(!=rSO#f6Y8#xG5^|fr{&lf>&@~nj&T&nQi
z)PoC7H%y?I!{O|_@WZQNcwXJv*{Wb=)K6LiQIgQIw3Q*pD(M1IeIHPKGE5@qMGanb
zSd$(lnRzf4GVk$7dz*fX>OGw3KyiX@?RR@y2)!s=RGlIV5d^ct{r1W2P#(NK3RkEm
z8TXO7E`lYXajJ1$aiG^<MT^K4_4d{$SD3?cm7%TE_kbe3g-T@Mz2jqB)8cAN>>$s#
zyZ*ZX8Bm3<p%t+)RyUp?u&HP^yj9VMCn7ugY;UZC(hkPgf1sML7MEPfOEQ24u~}-V
z0xuf`hN;f)s`GQdyXp`=BJO^%&7VKvdVqxVC)R5}<)O>=@%MZ*Ld#h<+iM$E<bY98
zGDMR&XGuYjd5ZYDnhVj>w*^!9N4*QWQM?hV1-X%yN*(iy<4w}lRu!jVvmP27o8DUM
zjJMt#NI4ET7*l_Ft6(stb?05+gAo@Ml@=x8X|%m*PCliRgE^cbX#8v!?vuC+ba8(s
z1N`&jm`{8uY?+R|^P_KLd4tzdVT~3?lg>JbC4Z8wl_bMe_@Dn0;8eGXDf#(2&w#g`
zb&&#Qm!vm5<57w$gf~a5aN@gG?G@QB$Q!y{y9ruykCL>DS95{fmd}+l^g=3uBXzR1
z&E}I$z=zCvxq!a{lb(Xr@r8svuALf2!}6hWnigctoU0xn8Qk_NI0}=%k-XE;z($;P
zv<&gb>HFSglF5i}cW2yPZoI&}!~H1Bp&h=9v$4FOwz=3|wkJBjey-*w?1>*48sV5H
zj_klUuSoBS19TOWwzm{4c(D1B!PB-3fX`@k>=<Of;|#94I$-@e;RVJ`WPVAqtyu1h
z4mtyF)eWgWp2z`K)NewmW@h!J_YY|d44|-eDB-|nocXrpTlFnH<Lqt7EvWt}BD&cG
z#7gJt@P2Z+6ej1vKRWF=JKtYL^^4}6AMo3Aq5e;(F~tkG2iqy3++5>Z2Vepjc2d5*
zCwiSUHgkpsoKl8l_@BeY$B}}^uV7`e4bImhFl3@@!409AXw1?bSR~{a^0W;|@kM%X
z?_ZX9fnzyYI|v205X62=K12C4IoQgeW8^?|CsbeU$|PCR<iaON-jX-XrNav_s6NeJ
zDG)e7R3%`Lf$S^(fE16!!hW<Bd{fiC7z~>j65k0}Cvu}XGrYkIPfva*zc=kmB19xY
z=?>neapeX3bjgn;S|YYrs1{$55WKN(V)$fHb-C@M+Wam0IO;6HE^H-2QDp?waFU?0
zp4=TtkF=j<I70>gc@tAJ6A!M{MCw4GD!MI{KJmHLq*3{DFE5mJMb|Nc)1gzrM}R?~
ztjKp-G-T88=lB!km-*9Pa$7G@c~VrOq$N_k%jVQ+o#1)WSt7fmSYvdmaj7`$mv8>A
z81=m&>}N8XO09;p^Xf=eEF><HiXRtyzyC3Pz>Qz+&gi=8jkh3%kP@apk5#I42Z`2F
zxCFD=PaWChljVD_dVEgA7hKxh@rf2P0IPwSyDDjz&TB~y*Ox$x`to-Oz$I^L=}O8)
zVG%{n7a`35st1OO%OMF1NuNZu+cp&$0>)32c#^eU9$$zc4KfpPjmKJm9P6*PXfUhr
z@zGi93c8KKj<7km+BIwGX#Q;AVH@Y#Y-Q#u2OLuY^b|I6-KyeNP$6K{<`O+2j&z@&
zl2`hseL@<rw7kmCF8z-c3F95z?dQ(cISxN1Z!y*?p#i3${+o&S?a_*gp{tG*T1P&&
zs*lKn%Oxz!vx1cisQ3Uec?s^fhHLU~h{!HmfJ_8pCE`d2?U>uy+9#KXAD7K#e5hTe
zOJt#Fa;~W(mDll4wa2b3pRI+s;>4;Y(b2>G;TndAE5vlVuxcH#5!G}c;zAX3-@L*(
zzJ%4K$XYUK!6Npuoe@@yB^#{O%`ntbK|+yaIy1~pvl4Wk>5FGh$yW|yYmTPY9+`*i
zZ)UVp@Yx*`=*YHYrtdN?K3T{LZedp)^oZIz=6qeVd3wCkZ=b#R{Ogf|+X{UTbCMCl
zq`t#Ja7?Pvu01yt#So0l|AmRc>FP%`Y)9}FV9#A$3Httje@@>?!Y6A+3kE4a;3eQL
zxe=yOi0L$ej2WkAEj?M$Bv#wUAB3>fiqFGrKwj<C1_8EKv{lRCkjWHNAFc2Hz`l@7
zEL3@z6dd^t_ILO59H|Ot!$CVb8nYJLit(D=cL#+>iKgm$3Y|PZM@2@FlGRel312+K
zALkKHlTbi-Yfo#@-EOdE+JAcGJjP>5TuR2Mpqas2@;sQRJlq8zureziWQLwQDD?b%
z#%h#+nQD12&<DT)_=_6}=uuRC*e4?yyvJZLUfS(Hz@@2S4|2(B)1qm=ohHB72Ro0D
zZXkak<@IWpa=&`w(G~-w<+3V++S&nT(2P*4t?i}^)pmX}DV=OP+NEF~G*^Z@+hTJp
zoV~bGQ^aOfYSA^2W-e5Q*-DjFu+uU8Tsr4v67WpXYzbY|!eQ>c_yrYo;*sS9`|ird
zy7$%IXIAC&*Z}L0-r8T@VTW`-JwQn@HO9Q6hynJ`6Wb{y=Q)7WyI798Fd2+VHzGa-
zN+?Nq==z7u1)^XTL2FsDSbM`kmLya_{zS6zGRF91a>rG+GTgdqv86%gBBjE6pQAuK
z{1jU-V+dd-1<ZT$l7FB$)2XiN>TdfPeSSj<Rmx-wTS`H09~j>LxQg13l1;uA?ZW0Z
zeIYPf<YAnY_ci@I;Ao%wvQ_TMh3>v()34*D70vC$G9zzLXLsNIhVPu4szW)tZu8{#
zYzViW&NXo$Q1~eTH9xQR{RZ9P&^tKs1zn#*-e^`el0YBc<kz`d*2t_ks65EP5eoZz
z#O@2J^@D@Vw^R`g)%I8t;SN~##$z_NwC?@Fk!(=}Ko7=<XC3@~Ag5&k9nj65N8TyH
z^R4u5Q4G>%e2Ie8*H%U5>e7vi#vi=7w8Sae%JH}q;E0VSd2_NYMzAG9L}%==>-o0d
zzGrd6Oa)6qGkCyl2CQF!G0n0k6nfy!!Akas&0n-H7ExRTn%0#EfWR(kTyT)vRrabr
zV7pOXq~!BnZmBo}(x?jhwVt^5YNoG7zH`hF5K6c>DZhA%k9+=65}+&CT)>0N73HEe
zBKBf~R^4kSO+j;ZafIqn?e8Z~KD3q{DlW0^{W>$;D=>8}JOy)dxHe)V9c~K_x^wdt
z^U59~$5R-UtbK2;ned{;#8mfrTzq%iP-NbCA}(qp8svKW6`{7}d-DFBd7|}54(=U7
z3)4}@5nEJqeZib27F#RoWagvn_Voe6jsG4W;BUX3uU5LRx9E#XoE6LT7XL7#c3*HA
zm*bc2Lq|4v`$Z^eufVG#LtB=qw!D?{MNFilabLbpF2m+W!ctwA13-`CC}4q=1RI_;
z{xwo{!RX|8Da&t$Mz=keGrGJqB1f{P6Z4IeMcs#iU3OAU<2f(S6N#S^ml};w{dy-j
z4Dv1>0aI6~VbF{?*mGTIX@Es@D!)5yfC!^onThmO1?4x<H7Djtmbq3-oilaT!VAu5
z>gE-ixBWd8wT}2>>UuXam0K@Iw=}^0aBVj6Ph-Lh&RwJcX24^uS!UIf?<q3sUfx0{
z`#ZF52sd`xSo2x8)$W7Vn9K%pTht53AYXs~B+elBXpZp`w%yr#@@ht3?KAqi`!>7G
z?cEin;f+NC-JC;Q0L2du!n3q97C2SPQLRju6y4rGYLBfK3HI^h7(mW7_uSax!`wg#
z*z=*(&;GKsBD5FYA3hXe0vB?0oL-92`0u_Bmh54hGgbsQa1RtR_2qh4C^#qaqf={b
zjvxi=4cW%bwGHGFffXnRpCx?2hl+Rfg<;*YFQIe6!A|zQZcV@b_3Hlm@x~-av9)Y~
zTm{DtUFUJ^nJ4{xY1nY1*rYLjd1L!3w!|V{`2>XAS~Yt%q)QQ|&smh*`l=J_sE~nN
z)(=5H<wwWlM>D0S@k^)FqRVRo>^<&;|KSydksCge_9Tmif`9Jr^Ai$Ds{uS!OA~dz
zH<;vF^km=b7q5LfGY%>*1&IsD4~5|D9^Xl0UfrO*m;ItqS14Q?e7k4dRSST#9DEMf
zjOOtdoFe8XqQ%bKUavxI&H^l~?xZ*Gh=UmRqgF*DE2A==!B6+2egyw#Ned$V4<jP*
zCl4K$@j%oa@j_@2j(M!G&uFu#UWA>g{TJgH4w`4r9mj)b=Mg&?u9JU`Q18B0_`SmH
z#L$xOy}un=-9r7odze#Wefw)-VfN-SGYVPQ-xmclGG!6FR*fjIMo_J)mmKe@3UqtF
zpexJ8X$33iVqTGjCKgx&8@SAMb?3~ZgC_v5(HD>;RDRU>%wx6Ir00F%4pdML9k6Q6
z)3EnVVRY-MB7rRzPf$1+T>@(JtEcXxGS0eE+}XtC{bMk^u}wvSXN~uI*HM>NTM6Qa
ztpelYk;)V(;Mz@D$zQDbR@svoOfN~R))aRaWyp>0xW|`oF+$Fw&eB}bC36ML!GGsW
zJAUIS7p0L)pAQEf#R|wxUwj2y>0%xpzOK+C)Z=V?m1m%`@w8=^jAc7{<a|>@00&3M
zE5E+`YWg$Ur{hH%5$^QBC0)H5w)5n6iz8lDcb{>IJsH_UpuclRSCQoqn<L|zWCxFY
zUEuPmCex|+FuddMDp%sl3c|20)nWazLDjKiD0s1h7C5k;acr9>N_UPtfXS)H0=(b1
zfuzwqDf#N^$bie%|EKjjw7(CS;n%IPqm2nf{MO`$J%p7r-R}SpC<t!hMopcaMK;U5
zN{<#mZ0_RSOl4ryc1UkZv@riEC!s_D#vmEvhV&1ViR;T!9bfxyLVOFsKunHKF`b{H
zai#u`84wr{2DvjEV?pI7JJ|Ivo#XH%?=lP4kGhc=j{fn0p@y>7D#t@o`W>m-8DY)K
zU73^RxHW1|HnEImYa5uNZga|<321jaSCX*XC_Ny2Lp{J;r3+Y$4fU6(P=n<BrrrO(
zd5g2Vq_lL;A!efUK83jINuUw=@L5q3#x;AMCwYBt*Ks;cEy_zcTW0P@wLqYCACF9?
z5v<-?0|9fP<`;-BA+$9dc~NQ}8~GmIKG5EA$N=adYO(hdMCs1Td2xN{;SUI9q9gi6
zCq#`{cUH|k&<<N(&ec`!!@N(^HFvT`ncLlRa<OMSqb!~2)#2Hdt~*EOqISeqcs)+b
zoe36a*3f@Gmk-qcHp#`;>f98yXs?mK{BjN&&pwZ*{kHAtzp}^R|8iBsdgC&s*XEYt
zX0iREdo=S1<;bX3&5{fVCcSsw{Je?=(Dne*ma^rEdn+Yp=>CBO@X0}|rDaLh|HlIS
z1UQ`&?3i7)?oy+Z@>3NjRkaV?f+xy)ui;`@h;-ojqA7K!;@%kz(H4X<G#YCFLD-dO
zW2`%YT3q*4L>}e$_o2K$YN&%I12+ezRNO=7ShOuG>Rp<?Pj}c%U4Ib)DzT?gzfwt5
zoSA#z;POTJrOw=uv5F5hSiyyV#I0IFdrn9D9wt$B=|+UI9X}vxv#xs^-e-~;u0X9J
zI`h6(|5Z#jUIfPL4ENWdj*jFg<<!}}Wp&*m3#SjR7l+P3T;5&l;U-~)zlcWu`}5O|
zDDOMiA#_C-ox;b(WsjUMVEQ65BAe+b&jj~?x+Mj_V!$+MaKkc@cz$!d^74FS3WhFh
z=s>k|K6h(<?Eh4@l0n$obSiKY@2SpyYz;6fXpA!gR1Cy+!1>pR^b|NL0CY55RF2HD
zC*x@n@V~}gVabp`_1EZMh5SXfBa{G6>e?ERqZXDkOYSc~H^cE><}Pd5aNfTRv;4|P
zolNXNzE)HeaO6>&wd-*M%;IqC8w~G~<$j1%TSuwE-+UhXtS%g<A=L~EG(aw!Q`2Y_
zQnhrpGbJ1v+UvI9tmlvkiSi1;q~isV`5DvxYP*GQV<{vT^W6%eLf--;ibzx!{oz8p
zUv`;X>oWbap2Q@kB+ED<m@1X8br#pwcR%-FF<pxURNX44k|@hwJ1C<Oms3A5p}(jV
z8<XU7y}p0YMELs+wQ3c=+rz`#dX!(#Mo|cp4{-W<f1-|rf+g6~pZhj<=VNGxSr_G1
zJ>y=5g5pMYXLI=AJ-ef`D8=UBwJv5wbYp(o3Lvp;?thK}2oUYTv~%Lmo7|1w<%=|e
zU<eGa%#hsdCK8xkduI&5tMnEP!^(8h5(ND=h`q=9j9<-Y^h)Hal+vNL1`VG>1+>qU
z&l=ivj=}4p>uvAzvZ<|nU`8Mij59*$P%aJtSewi4gD_~WygT@AC`eu4&DZ(2&O!>G
znubbarhydq*{DlhqZK98;oSVmNpN4o(28c|Tb8lHPMQnR7@cM_>6=2)g?ENE{i_@X
zhBxEZK!4_g{n~pn`xy7He<`XUK{U)iO3x`j=jvvL(e3XQ4BK#&-)dMa*^}H6u2tFc
z7@Q-V{vfe<$X`>*`IW~|N@E_N@hf7kst@Sek%j1#0q|w8+V6(GvM5YC_Uwx_c2<_E
zT&iatFT^fC6lk9``KD<H#D+COYN|B+>`xK<ZTK^~8u7dqq{4sGlys$$XXALEKcE2!
z0q|i`g$==OAPI5k2kXmOK$>1FFAlY5zjSCXsJ6S3BY%}SN_re|NgT4|g&RLfAJKe4
z4_l&YnTyEto%mPHL^Q3#)kQd8yIRzA9n|cK01QuUrH1fAVzL4@f$$%-l%G&|xIxSG
zm`~f@#UY`VOw6WoEY)AoR`Ap(<~u`5W6ryX07&p8%|$cu0T4eD<%^H<Hf$t!m%ry(
zk<QAtVw$W)N~`jI#<etm2sgz+H#GiYknu^W6g`YN%&u^M4_~<hjpE{Zmvwd|o9$lq
z(>hz1LjfOFQfcj>XXtrUhfXqaIJyuwKI!}OS*7K2_tDPQbdoTk$6IiTX9>V*Jk#za
zZnwa}5uW+_hATN^wLxY^YXu8}Ts1JZg_~>buRx0(z;Af3RYNtfni>=sI@d)X572I0
z{)5ppG#leR9hpg>7sh4fazqx*aW;N=0-}2@cwAp&MBJNUv;R7n)F1I35&3fBKw%G-
zZP|5;&;F52l5Vn4!$K@oOu?8gT(K>m#gGM;G0?H^*_2?M<-_rqpCJNA*l!Z+!vK5e
z{%VP==)V6i2&U$}6~i5L8NcLWOH|Maf%1aXKTCyR3Tfbr<9VWaFc#5ncjG~c&T+s4
z-o9{peWYY#tfpd;-xcII^emx7h43QaI{T^sBPrH$Dw>D7HhCdIHf3Ox?1;nu7h+Q~
z+Sn|ptFO7zk9|ERX~0ihH|72-r7TR-fNpP6xYzjDN6W^G*_f7Abx0)2fp<HgkCy2Z
z<1D-|+o~p&Y@&m@=Z_s*Sf24&EFHO2UBQ9g5U<xZF324u|Nbu<Ad}t67dB8EJt&!f
zLF3#Ul{izaE(1piNB7PfKHrQy{}2|~*Fj=7ZJN$%Dqpv-D)w&zS-<e}$j%@(qmBX1
zKJCxYfhk;V2eCyc&`|d^2#ppL)w!X`{`c3OQBfjDjJAl$T48r<jKD4T2;eD2(iC5R
zcr0D(4lGX#8$~?^+ajpdHd?_@B*Ev&=tm%Y8UL33es#b{7b5qpk13Yy8x@>YvC2oP
zrmOcrz(?yu%vi8KxGi)vC3aB=G;9=B22&|fZSIbkLU#~UYB&$E;|}ufb&fYZw(13?
zqk@C>>Bu)}<3f0JD?L8`^E2f7;=zFZ9$-Vr`${rpG-2;fdvCbI<8TVE&iUBEL7UW{
zaG7GV5ddtWed(9(Dip#npvSBKk1FRyUg%Ccf{I3(mxp{ff-sQG8|3`K0HQ=YjnB$z
zzk!61X|kPwd4xHJ1`@L5WiE7^X!<^<D+b^g(K3!RvC_tHn%?fkwGCb#Xqq}>G+uP|
zRZfUJ7miL2b0ia}1DpEBPT1T19VPv7lutskWeh|VP)3|D3$xQ=qD_ECKLgEraG=Mv
zYL}bpz<61}GlW_6UU8ez!G7P2P^vSkD!#h==5UOPy{h~Mq7p585EJMjXnS!sket^@
zJ6!oT7J_BkzEBE8T=`cHbI#L&7Ty-%`Kxf<PuO|Ns#iLf*hU8pxju;VV<v>dp6LKz
zn`eG0X~VI|D6qiR7-pL{h5{AK-9b208Ch*E8x-})#r1DRE%JR=rhIqEi`I*I;7`Pg
zmDTv~t%ppMosgEli4(~~wXoQuI2A!(*UcE+=SFzLf%Cy`$#2EeeOIk5#4z4$dofVT
zt!PZf1<f=7!oyDNFSn=jt{P{x(3}5OwiD)z&7XS+ByVfHx7Pd+-MA`WDdZ3G5aT>z
z77|ZL^xloBzC3je`K$Cs2srR;eBJT06$CBG96IWsTZf<uc0%;+E>00QNFZGFT?kiH
zs2UxIg34D`(RaV4NGOv`UK>206k&M<OV)+*_iZxUP5xBx&T?i5@RW8;wVC+b!Aosh
zV(m?xmsMX^6SC}?1^RKi|NJS<F#eUyO7qHHP8S1P?`II7$WEeSg;-{BnUano;Ln_E
z?aJW71`iK_Ps*BSp|?@pUDw{y7~cfW?x+C!z@6tT-sH`h|L>KQL)*VDH5B|^McPjG
zxZp|IzJ1!ztYSGyrd6r*p)giR$krzphpQ-y2{BaD8=YvFX6p|rhAGFR5TwFlAC_uD
zM{B3cC{IVYfAk0Ndr1Z|e^a{Z3L^?%ba4O61>koz)|=B+drjih^a}A{O<TWFEItSY
z<2;eDL7v?U7nz!!h>Y9ZBxu5W!6HIn2GMLln_GnxR_^|)hXCw#CX?19BR~=|E;zMb
zCAPQlnDD6${khsKEAM-~r==3RpQ4O4`Tm37H3f6X#!sbi4Y$_#E1;G4!jBzS*^dt2
z#;`U=-l85Hx-9RvnBbuYgFo-a?pp#V*!PFKO-(j#$PY{m`0+wzb&Jqzct<KlFfRG{
z2~F;;#%q4%CBDQvxTOJYw^cs=|I{+*f1DtUwpHnJG4u!2!)hdN+2?;w4M0O`roUk1
z88U6!(MpOzNxs|HXsF0MmzXsbY5a$1?@aj69vG)|xtB2?9-26Cd|3o__5$6O#H+Nq
zVbmK9sYURoA8^`QW3#W?3i2tfFSS@jfl;7D1uUilkik(S<%#8QNFv)Jv8S7&0ioED
z(YR2Dq;DlRL~}PUyjGyk?hmr;kD%A-2}lhJosI0_f_IoyfaO3yP`h>peX#5_w$+S(
zU>vwb^>Gx6BQ)J}4L-!)3y0|C3vS)Jf6mVV)o~Jcci7kfzIeI^q5dSHn#$8@en`qP
zZXb8$xE$0c5~>u&2lM3nCjh7ePD<WBvEpHtq-p_}^83Xr(at+Mxmfk2dzzzQq?z_i
z1}D|BCKGQfIBe|T5=B7<*gDB}YyM<xahapF-Gy=PcXZ&3lV?nSy6JaF!m{1IIkx9Z
z^32TC5-y+)1nm0#hD+DFPQFY<DGFvCeW2fC>7<XM<ma<Hl2`@3BW8kPGw?FWlIYw~
z)^SPKBII8zRB*N=mF|gvGxuR)0kS)q@?eI~U}CxJk%6k{cd|Sq8Kl0UZj!eTSCj@!
z#DSUTMz!yMhOMuN<QqC%t;Fz|N?MerfI8lGDc!yK7I85VwCm=D+9U20408ueJ&ab0
z>1h$Xrun>ZcvD-GGSM^Hbz};8w}1O*=uWLVZDJXi=qKO=W>MBN1?02viXT1O2b)*z
zK<qd+)VbZaCm5qy4UH%!{I&r+CPBXFGZ{Rot4nJJz$4ZJkGWvH_T|lH6@Cr=Q9{R>
zl1>b@FYtHmxUEOFfrFY%mb{1rl_A3i@?fS?E$k{eFVIf9$abj=SQ7@pP{r5lu`P7;
zn?i*R&<ZHH_{#yByOfS9<4>XfLatkognWM)j4R_9poM$OnvqY}FjC7O_y%O`<DDih
zB(yt!Mch)75{-VNrk;J+)ykP))C0OD2XQ%9>9aAI!`slSt&&|l{BpAt^Mjb^{j7gj
zd_Mo3*{0X}rjro2YXTx8VOhO=y_Fk$zBLHh7VM6J_h~osgBIp4bWDZ%q6YFx%X-f|
z+1qN+exa&YX4W6t0Kj^~F57`opg--g(1`Wls}#u|Jr+`-o^jN-JN{wEWTh9_e*jL}
zYvP0O$qx|sZsbotCoSzktvTDA6SD$$4iJJERmLW-18IXE|H)>|{}r*VDVcAWITTUF
zjVb~Q#zg_04P*5F{wV&d9QjT`?OyeV_a=qgkT)}Q{YEcewC#ss)$76Ww8m;nc)`rC
zV!FIwd?CVni{84X-5;3Wt$!Z^=_X*AP;+K^Zv9ewx*51+D472RaWGus_ukBJCx1Ex
z@M`)t=ReS8e-tzOw(I5(mV+H7LdMG5k`Ap&1N<ZP339#jFy-IQH=TV5++mo9TxDfF
zX57bK+IWwSMJ}^}Y)qaQ<X?OeYFlpFw~%)|`O!XVv*p?zqMo?gB85GJXOji7kTKB6
zLrnH{O1Dbsr11)l%D0|IVuwO99DM;EZA7k25QP#R4PKc|exv18df#uTHh!S-|Dm{@
zPJ?*A|9>=Hb5v$;w9YlTCfhYr6DE7Ygvquwd9rQW#$?;J?fSB9chCIpy{*=2t@&r_
zeCK`lvmfl;DTEtf>VbUq{7zxK%js2x4WL5jVN8qCmHu^Dox&A$lg5r7^<LO!P;Hei
zro?mSY4!&(kSWxO$zvU#)VbbT1h>9)y<cpLxpA<oLQ*Wwd5NpoPF?zRW~Q~+3%%8g
zi_aAd5Tlxrrm>mM9VeD8w3q7gy7Z<G4Ta&k&y3DoVDCP23uer|)@_Ydmf9C2^9R&V
ze>_zOUE|V26(eYZ-Is1i|BXigboJWV`bf3_8;u^ywKNu~hYKk_d(;yZJHl8>+h=48
zKzwR)=hAmNuq6bd;^~cV(n02lWkdfIH$Zp;?qB=|YAxuZ;e=X2SX<@8AIS4W2VtMj
z5&P$VynAs?P8m|r-5*8eltBSbTdfNwdhlMQ#pZAKzr|$@$r>r7w*XhnItm6<8x(mX
zwPuNFC!qlGM}_<H%NGoclCpXr?tNdQ7}^stqg!zePg;A9no{t!W(R`N4c~?4MoVgt
zh9#+{^ImloQe%^xl`7obgK5JZHnCA#r1%gRYz?RvK(Y4+bL%}wYF(EGx&XkBCJ|8o
zYmv`Vg#7o=CkY|ZORv)mCufYkm7t3h%cJ&KB6K*yd_QkEtk6*SqNQrRi|y(xKiAWC
zLN|fGa7(BNMQ}43hyICzt{21S6Jrm9RHR0ul5h>y=O;5^r##Fz=cf}BFMN9soAV7L
zs&GgzgB))fzb%AU+BS};_{Dbv4ba-qE6ZDIAVYnns;`D}oE~jAXIGp=w_>$Woc_Ew
zHnv=Bnp3!n+IjYwy30M*mKf-)k55BZ%xL$<pE=#Q^-mxlXlukwEU%t#>WHgc3Xc##
zK}mVQfZfj#-C5)Vll^blMNz;G4-bF$Ov_41=~W6wC{8vULxw9Z>)r^*)H1U!({iRv
z(SwmZJ{hFC;5Iy6a7Wp)v^$WJq3_JyaPuzkLsGKCuFeFc+tQ^eh*`L_3&gv8$?>L+
zy%w{ms<Ji5^Km>-AKCcc*3SOPkL*$)oIH%Kw=pwEnk8B*GEeg_WcATCO8GVU;OKNw
znOuYIM>o4F-=Rxi-A1r!J2)fwB19Zsn3`Hkv$zDWw}RjEtf-%{C-Z74v-8VRl10&)
zRB`!Cp4G*56}&z@!FbqxZIy`$>5`=?Ssvrv5h#b}UHB@W7?Ff{dLezKa`H>~>4pQ|
zyz=ruZjX|?PD>7@hLgRlSA}G{n0Hjv0UwwYrO44bH_Q^AZ~(q5@!oiopfQ&JOzQ%I
z02;@s$zI^_M|W)UNKT7EdKO=j`mh=U*x(&)y^`%jq}wv9ZHI;nkxcT>&kt1EBfKHC
z9O_x3DdnLwcE1ZL#ku^vyYSo3_ioD@qgOY1S``GgeDY~i2$}0=4`MJ=TA8)aZr`bY
z$$*@-4{ctu96d88$PbK5&FWUeV-sAiVP@fhQ6o&ntt}(`U5tPJtiGk)Zkznf5Vta5
za*2rKBO2aVU%8!<3v=H^ZFt!D<5CXTdn}0GQD#p$gNscqACwH~B|nl@y;-}~{3>gk
z%@`Q&oASml=2qJ;3a8ZfsiKSiSdKNn(#o|fH|VVi`#)0TD}l&~`IfFb{5Me%dqoX6
z?pPR2<Ct8&8zGz1&MYfCbk)b?UTSB-zE6R+jG_FIl?;Aq)888*MAz0bf|*Kv=&{rk
z>(K0@qLq%6z_2PQmE)rSd37U$g3?cB{PZ#dof^4?Oi7K9<FuL&1@6OHeZb6Y3n5op
zsH?EXK4wL=v@3HvHyVuvYzIcxwrIHNTq08rP?XoL{_M+cpzyi5n!Zq|cHaxn*0Er9
zw?GtSu!^fG)!r*1!G#`8RjwtIt2ifOEeB6IzU&}qCc?@xm4BS}{xxWsNdb=Wx#5U^
zKisC<uZhPM9UeaE%(Kkzrt|C&`rYl;Ahxa5Mr+KaWPikRz{wZ0gbVocg1rUvj*N_?
zA{fV>mx=vVhDmMBcy<+E4?^w_XkT&yD_FLxVo;ls>r@7>2K|u4)?6dKIjjtKr-59B
z{jmEAEVuZ1&c6~NkkIisoGWp_zxj9w?aWCYbJ6>CsB!6$tVYuT>_>2<%@FSIuE~vs
zkZX@4m=Oyl|6~ZRr2*BR=9*PQsX^<A{Gr#ADg2VORyI4+TeY?sH~NMak+Lh~2-nK4
zA5bWfNPZ--$Ddbo6OKcSrSzlb%q@(|tJSKw9TkZU2PR-S6rIY(Pjs-}1`Glp9zFXc
z1K-=Q8-0X-3Y@Y>>%9rNI91+2gOESC9P89CCn8fqWpagLPhxtlTY>B5Z=ZU1oG9OR
zLS3t3JG&Dc9EICp^eN%aa#NU^{4IjYg^RBlcGyuS-&I?@<H08r885k81pKdSMvu*+
zvIOW-Y-nakgoJo1j%j02f76Dn>5i(-(k~TUIw$?0(%y}gRJf4q__E4ZPGz8acTI<^
zI}VwCkot@2;@Fgoo-rj}veoFP#!#Gw@!t{9#25iT``s_dkq?^Z;cLdQG0oLfswL)&
zJ7``5tV0*C<tMpQKA(sgU$Q#iynSJ~5}82P9@d^KP&**`AR3cuOHYye%%Z{)BEF7R
z*^XPL9Dn?aQG;MU?rZJ)>W=y4eo}X(-R!%`6U#>s@@lJEk8|y!M>;;=e$(gXxPxQ(
z-Mpac_J{qez@_qP3B*<Hp`@H*NqlEnCDb2au8)0jatEFpBqSto5GoF7Gy4wJGW@4H
z-+D-fwG`6!_cxH4p#OOGaiBz~MlqAyLJVm4qPBZfF=zz?ex1cF%&~K~WW3%hbGvw$
z1|p5I(q;)4t~=v%$)?AAm6bQsxmKssmtQcRYpDa@+P|w2s5wGV%U6xC+2VyB9+m26
zF2g@02w9p&<GGN#eIX<Bgt0cucP(@*gwE=T1^?(L2`>A(+SAy7zV|(i@cn_<Jx(HT
z;(DntpLfEQ{0Jxpud`E-V8A<{8q5+DJ)7Mzrx>B=Cok_`_iPemA+AK}o5U`t|2<I7
zfXlF{nNlo`&L3Dda5{}}(~d$;ymT_uVQL`^hb(fs3`_b|ebz<V|Cbxm#ja;R3MKxq
zB^}k!;oxC5xF)e%=5twL3ztP!$Avsfhs&_Na>}nKO6W24vd<NJap#UBEeZ14vci87
z@PnijCZTrXO~v><g5>{pyz&f|p5>}38zRpcWyO!g8Xt7FhI<L1G9pkxM0AA1osJ?^
z&n9EAp@W0tPgDvRd+=ppzY4pK3zXGFC}wd0+}vbU*M=s63(N=Kb!N$x6dbN#({k-C
z#5<uM`T613FQsqwYT@hO2mgw=mn%-HZjm=2W{PK(SB#eTpBE<^YmUqtlV|OEP2wft
zy&&h(>G5p5!F1(Ne;7?$7nLfas{$iMl4j5Ck9dj_j@1;KLO~-=95pZBa?;Vh$KQ}i
zS7^qncYM(v4Znq5H*PhW+aqzAN>c7Uz{7e1gX#9p7qd5KWM*ehMF<syYvig}mb-xA
zi=f|2coZxNkF1VB!t28~eYkeG!ediD&k-r1o}i8WqL+t5kiDgreUsF?ezVa0M!Xm<
z7cSFfe}%b(<vGJN+t7T>z=AkC&;ENtdC|Ac8s%3?3q0rF8uZlz;=J9x?0D{U&n%<~
zD5Rujxdxk^4ilgguZZBp{_}epgvB4Hv11YqQeC-C!_NJ6U6>NyVu6KGX6Q7oXZVwJ
zoZSaErjWUdACvSJn2uKt+^*j{FAR9%?O8#IWt6oYU^;%-y%289%U+E8clcH3?kl7L
z)NpVy(>O3lHbVm(F+1F*fITwnqr>iif9f@ytDVIgHp+fy`g~AxiI94PVB{@C<<CIF
zs^)PKcjzI;xkOX=Mu)g%<I_{E+aCx>c0#3x<XitBH#hB(J9mYeMAu;tS%#^@Q6X^^
zaFYAuu&0j9vbFkniC;P*woGE;F-_b%m}9vi8rua<W{+Mjiem>5XE1!r^BK*et<^jm
z10x&S@k?KsLSP(~%}e1}!&VMh)%P61Fa!!jN=$@xiSZ0+gN`Hh(8dw3FS?c13;x*5
z;o_v06k%b!=07(XHT+Ueu+%f&g?hK!^-y{N0Kq{()ch-vJ@EBz_4-7sLP(g=iCGwg
zxS!W{p|}i#`?kVjZ5vqU0tr-4!m#s#fM9u*uo_^D^@bKsU__tu`%l8uJFecTQif~C
zMZBeNES!0(CkRemzlBd{b(HK{;<_1(G*X&Ra*&lUPH3HO<nJLe#Zr3PLy}y_Tw@n1
zujra*KIy}oIXjO?fzhD-T-xwxH|Q5efEc(b0LW?K;bFlE`NDcNqj*)~IXV7&VPFUg
zM&!bQ-0yGHa~MBDo#=8&VzoV{@@_hx)`TEtRhe(UMJ;M88@_w&@@Dcp%LT!BO&U%d
z^aNkEE!{Zs4ZS2<IQg&&bWz|avChMwD!7uBA1C7;&G5~?d6d9F2be$b9aRgIk#!a8
z>BM>TD?n8&d#AhPQ%WenMRhTn(FJbhb?bJ)xpQus8lSpHXK4+|1I6a@uFFI{)yH_@
zK!KqV*xfe_!F$|REk>5IfrOEkwg-n1TDBLzT$0YF{N7;?sdC|&?(5j{U`_lIo7r(i
zclf{o`S#H!vRN<ByLk3iIVkT09dL5b*Pk;T#+r2BXzW$ac6PJ{-p*uXT+P87;O&_3
zQ~XYw$DSfk4O%aqo6%2*4=EgpC^)J+EG%zgbGwL{i&r|MlSr^p6&U_7DdnApI{Tcy
zi&oDK@>XCJ6*16hhHwF8CVDq-VtaEqa}3N<`92JnKgUJ2B%L4Wc2c*GgI?%X>K4+P
z&$_}Sk%-qt;H7?Kr%{Gmou$&TkFWvvn;b1`^daWYI?C|20rVPt_)GL}?$HlOm3wqY
zQ|>n*dBnCxjl*p*ju)TaVCt!XqsRjCPwY(SakpUT^lopnq^h%}!+)u^$d@`{((-^5
zyd?gt8_V_{DZle%C+EG&OzuL;m-Q>|Mfc2qDkmHTcdX|<%C7(XL-rtlMM`e-{xQs%
z%qVi!04kF+V$zn#Q28A&8N57_3{bbQ9&LS|+<uo~G&t~+67Y_c3j)hy&_Q+&<ft4<
zTr7iW>adCdfKGX&t@VJpCXUTY2s`u#0@*-6H=dUUJ}-><^MZ42eM1Z@=FvHl7OKUy
z^$l*(UjPICEw;W29N{e^x~4MlgKeo>FO|#k(y^pE@Xs2#tDgOb{E5yEsI)h>8l^==
zMGvxqiNL0&r>Fb0J|$g>zdz7~p4#*{ne<RvF&plAv{rjvEni{SQ^UrymBZFT?{H!}
z6QnI!lL#7O{L=Gr6M{zU6)~F-dnI31Y)-0e0QpEcR-}FXf^LV>5h;M$%ng{@4TfRw
zGs8)9>)^k_ZJ{MwwW65L!{=M3<!s2!l)^>Nk0&|w>W4@&F$y&3;ktaCS<#U-;~kyj
zNlTXTTe=S;C7J3~Y27a^8*0mV!TfhnetT$-o=hh)T7_i^IJWnKSBkLTK7?kRmUA9g
zcXpk@YVX#kGibuVCt|z_N*j<$!0_}pbDxuTPP^O}sxkh%UWp*h#)ARyvh!K#t{;^w
z57jNhnYI7E1hMhCth_i0&w15^A7!kWuq=XRetz>TVsmG;o2?I{OQx`G>|gVf-szh$
zA5`@<@@wc3TA`g#>}M>w|Mvo%*0CY~U0CEAG25T+U2me{#~^-kMz!&fMT&CdAF%GO
zAY`WfwZQB!$V*6|Hcn79zCSm<SgW_>c6%nA6;h$AHf;u9e*EkG24CW_#2G0m-P=H`
zw6DivRIK2Paf`Y+5mTb1@$!B{IFoB#BkAFZnmiPTDm~zBv_I4G8y2F0w1UGDnK@_w
zf^E-uH}x8UdT{7~{jx<gg=j+Rw6mZ*YpUrV`hc9oiO3plv(31+i)QvK`6RZ#4ZnzZ
zU+IbWzFCtmZy8j{9+1|yrbJzLkT>QeV-~h8HE>#C=gcV@xLqJMB3b=ABp00_PYZ5c
z7Dv}ib)sjl5@2l@F}HVbk4_Joe4$^um+82H;+hik8LWv>pNCxJVQ=p=z+I=tXKg?q
zBEmBq{&;ZQW_Dd!yx@u2<{%kP`J8jCT}|$`gp+7(q<#sm8uv1hmwE$dFzXY{WY%7M
z)wBivI~oe%GchQ>tp6)FgoSo)3Idt50Z_OIcW;2QValvxC+fF0^!iG?<{9+SfFp$?
zO&C>o>=sqPCbyNM>JS?{(Zq60*Z5G7KU7w{)EzjoKDTrd@({XvS`~A9FGbP*oV|a;
z>eZ_=s&>i-Hl^lLnB=x0ZY9JtDxo<Ft3|{cA{tkDG<W6jUo|=e2O}gTl)8P8*qYv5
z=5c*oPZY_R!1>G*{qmW?$%90k>m2USi8W@lr%r_4A)J4a)j2G6Xfk~(%&G=MvjaVX
z&8HFwcrUCPGCR*_&7=vh-!a<<K0?+^yw~>VZMmAOEn_46O#a>txU;^M3<Gw+%0RxL
zq*;)@z=RV4+oHT)Y&PB!Ulw3m(w1mVv!hRoID2dC+Sg2k+dR31*A^SPxO~RV>17#4
z16#u21_X)Zzdt)2ml8p$J)OuMAz9O+vrO#D=MMN_kInApVK*QxSv`e4?Zl+liAc08
z#*3jBclAriJ+>f&6|&T10=Cc(`%dk&I5j|fcjs(a)cHgFyIPe|%Qm3|9$%(xKG~+B
zz#W!C>uAak+2JA>4O-4Y**VvVg%5_IMJ1RHwBZL=THHHNthP0XvO3e$D9WZ|-{^Em
zEk{0Yx0mTcpYbt1_X|n0-h?PEu>GKJ>tO3yGZFhk+EGX^jHe%$w3jhVi(DW_`QoMN
zZQ?UL81RAV>sS_;y%ev0)#Can_Z>^K`Pqi1Bb=y0t|+inX~ImAjg3N~HT;CO^Ix%{
z3PdZ+H8di#1f32Hd7CaM!C4tHU|{e~Xmr=AWi`ZP9`2<Lt$(U*YFUk!X5F;p4dnkS
z)P{ujcXeg9btMgwFViiH!$50Kx~uu-R(73<%Bnq~EHH8y(+v34%jj0e@Hr)#zIhs(
z1Ri0T?mbum|4Q>SK+N2;n{n%uo>(ZzPU-iFAEpuO)qaPtGy}qBNk>qO`V;SXZV2?-
z{uX7bhIUxB|BdAFH=a1xu>I=x`2)|u1<i4j>DG{%)DSBcdE?lj2A1y~blyfsyj4RL
zgq8Z8J~#v67ndz%Lj>k@0K_66K$p&*o+kealrshx9&rGJPd~dukt|cgSWiFQcp2;w
zNnMCWc(kt#0WHO{vWK^8ag1S>l_V1WdU4qUKl;v*NiG5B(i{N#tY*)6YdKEWo8P~z
z({1r=51@54kY$|Qzc&g|0&WsXINX_HJ1$gJg>%pMr9lQY2j5RqWNv$usZVB`%xy_8
ztPI8sznSlS-iz0goH?;YrCty2qy7~e)1d!~j$D9m&jB?5mE`{GjvA44`UeDr%An-d
ze5O;YgY9q-N{PKBZMJuv!{>o7vpnhuXferF?bt6#VKEPl8i+WEI3nqxHZz>M-?{m_
zbi7q+uwxGLKN-7WbbVcYBze#vI`J$w7h+)et^Rotj{^CTMVI-9ZN}f9QLQ^MyfT6P
zM1-u(ls}E1&x4*?ERgIa_xDwW&}JRS_SQF07+>~Vs~9xq1dxjsNNzOs`6em$>(yR;
z4ZvR&kEMtP7cyZ{+V|$3K&f~rGpvNCln?in`&<y3W?$ifM!C;VT;<AaWgUQX$hLU!
z)n6R9+sL%&BY8>v0}Wc9tWLxBxeXmxtfNAD?Wc|4Vzg#wbJr@QJOxYJ)#-9W)Vg<?
ztR2A7ln1S&XRT`?CJBQJW#sLH`a!_MlDAO8=2%s)&#K4dUNx*#T0LpH>$-7IZ_MpI
z$T}n(|8O1O8=Kln-AH0Cp|Bj!EY)QBq&)g}+N3)F5L>AMGQ1j!aLo3&wCMY<bpA_{
zPVSh875Vxogqe$j#HORRSjVIe#t&*zI2?u+#2gL8dnJDRA!ls;ZAR&|5S*;dZLmn)
z^S`Y`MmAfMgbin<<GU;EGBs_xPFS~Ti$*>T{P5MH4EZ{9=30#hxjnLpKON~eb@sNO
zOcE^98+Z5$Cq5>w<?jWpBC@9I9~vk~i(|Yk*g!Q`D4)P|kJfJ7MD*Z--0>croJ7g@
ziTG^X*B;~oVI4DAlzmk&P69i6f3_0oYs`bgj}5+Hyb`GQH<Q<@{xKx!eme8T!9p0g
zGSSCj&yjMF1YY~vl&ov%bCqv)s^C7=-H<L6E^LMDoFN`a8zty?nyf(%mKUDV4Q0Ps
z8&AC|dS9QFw_Nu7w7cwhtt>xlK1t}Y=(p^PoPy{a6h#`{3Km|Ho2iV5CCIrJp*Xr}
zL#<(XFKPzoLSyg#{aqHn2Ky**Q?h=sqv_tB?Ggy(tetD)oyQh2D~dmX`KS3Mqq_2t
z=%ET%L}d&BFpQ=qQ9-|yWQJf{@m}sf7#8{8iMR^R9~#@Z;UGS!{A2prBxHBs@unH5
zp1khcyhIAaH*1GwwfBO5;0j8ue|(|+Il%>ScL(<`x>>xL_ia>zwI<x14t7MM+fXn#
zL<nv!@$beMaI=}!fMlCj&m5WfLdr%gh@?3OVuQbzl~Xu~C_OS?$Rg&^s!LcqYsVwJ
zgX6$B+3R-1wg*>+Ejq7#40ud3I?kEoZ#cvLk|VX1#&ThGc`Y>r*LmKPH$N=$F-Mbf
zU%3DQt=!htV>#!7Y+slf!eOeF(_n~`f&=)aM!?HJefgJKrRu_cILKE8V9H2r8?Rz(
zkuV%mFUx`B?x|$F5H3yqJRF}V+`0nV=eSnI6~R)^5?Z%X3j^S;ie$z5tb41c`h(aZ
zB1+>56zloO-lPM^-g!goNHawrV&m}SD0QHr)lp6fCQCbS_q?F7BwvqW@*oc|7jX@s
z{#*c(>j@8@iht`<C#zd_Gp)6rtP8IGht`3%Piut*WK`*|g*@}CkZtjS?-)4J&oq^L
z?gBMVh)z_R?t%sjPhq5Rb5TUYfh^YWb`PJ`X_<S{1$8w;^F}6nt!E%%LAJ-Q)(#Mf
znC1JK;VLeaXB8EMZvUWC-mam50XY;j`2eMN_hz(}gqZv~c_Hrs7Bo}C8mPQ3g;xLD
zMB0sKbfaeNdjmDlGTM~o6iL)pBwT-&LjMH}re$H_K&Ox<(-%aK9eApM^l_^-sLJ>r
z?y2|2$!<)*msoKj5VQSxbYJhHHrJk)_7~fqvYg-_c?@~J&vtTi!oG+dGyzXTgEig8
zLU72yKs9hX4N(_QgHI-wt$tdpmiQB2Y0Bt96aM9m*MPi1B-SE-x3tab$-g!}yfOL~
z_^;8lPs<-_%e;YpF=JL^It=Kv+MF^RaPIEMk>-X3WS>E&rQd?#5e5yGq!JSI+#f%M
zp}<|ADS)P4*ea?s7%VTYjF8}FR>G0@bnXeIPO`N0v|cz>NO2Usp&%^`lw~Z9z!z{s
zK2TluyIvw~zj=_kxUoqxJ>(Fgm;=C88J*0oU_V4f2gyG;uQN3c_<sO^KIp_&iSq-z
zsx`6k2;J=m+^2q2^yB4AVOfm)ar|gc&5fMqoTWVM0Xqg^IF$=~Zbc_arF$>)5?(`I
zOJbaUiVNQORBr8F2Z8x?rd=?G^QZ$rECP6b8q7VKek;5gya*hI_Lg1<S$V_Ekyq9E
zWXsSD6K4LsvyBIGVG+IHa57})X|Bv@D%|w%X*;g2A;ty%5&QgK7VG2h%(hCmL0l}G
zymD<n5df0Hv3b2<iI=U{Ql@yk)?T0x6{^plLTt+Otd5b%u!!pA9``bq?8be;geT$2
zCJYLl<2#&*h}7m2Q6t)pUd)nYR|k2C5O1I_1TY)kziMZYzTWJG?w94@#TzVg`~2=J
ztO_`_T1>O;gBOm6*#}z$d;Rw>78Eijp1H&C!l<aKb00PZ7M*EQ2^00@&;q-AnNEL7
zzJuWtLV}(6mOOp$QAn)WY2kI<)iNHGw2X<KdU8{soxi!<YII>X`GRzPWno~K7jp7-
za%cFGlRiUs6E3qo`K_7zPT+?Cm^Ocd|37Z)IImr2Y*_FML+|^U{4Q_nN*j3hAa<{R
zEn_%Q(A73A!Hnfsyuk)h;msSZb#eUmFf(cad#9#nOI6LLUpE|cvSy{n>dOhzV`!Me
zo--$5fL=mct}`3*3JIMD^)QdzlgM+Je=GT01Vb2j(8j-(eiyC=6j|Y{iTxP~9Sb=6
zINp4<wnO}91;F~kSN)9_k%x)X0;7@G8Lb}M+N|m3inxS(AmSzSs_A3WNW+=y<};CW
zX1STW7ZzYV2%NU0yuNN|%R9n(pcvckEM_l>{pEJD^!=-n;XMKzK#4d5m4kTSL$o2V
zB!^ij$ffS7YO*(ewdASZJ#F?3{!;b9yt!l*WBy<sfHqGb10iAW#{dzWH*8I}RXir6
z_$?MVOao^S*{p&^4HjDq6V3NDtn*^{UW4y<=%+efK6QHEjP_jI;~xl}h@3+-?)fK2
zml1}`E{dqV1Zm*!8L$m>u3cac_`TDsQBF!H%|`Y|i)J_4M%T$&)|AJaT^KkV-r4iA
z|A?w%mFAV<Ff_s4NI8++6YtI`1I?fN4k6I%uC{6Ge%p)by^zh9XM^4xoQWXRMqr1%
z`n95k5uqWlpKL@x(4KsK#edcBP0hXc3OiIZdNk*G-bA~F-Wd5cPS?WhcgaROsuAaK
zcS)RNtP{KO&y3;`wi)yu%;`Aw_Xb*Y?0nE9ppPk7Rhk-<W>;0|Qu?uLj{B%PJOs&W
z=6h~>%g-$E&E!gbaF+~s^l0~L3?|l#X~2c{emMwNFZNlqOLXRO6wzY4IPe?=%806J
zRjmGr4QC)wym5C5Y_!b>faM~9;;7Vc4#y0AwHn>%TrpT~nO<r<-8XkRT}J$j+&D1)
z0}jD1b-bNJ-osJ1zCH0mf-U_r+2P&&>lthKx-B__@weYQudr1kj@pw?zH(`mPWW(I
z>bJD9WUV&u+_HZ-C_8Bssa5VovoUt{-0swX91FKFL;SNkfCT#QmxChVfdrVy>3Q3X
z)aBG4);Y87bTUiaiVYY9H*;n+g^oNw!{zR{%U7u2@34RZ82mQ4x0FQR_r7DhD};Ma
z#%+b>D^zz^b7i|>-%*MN7^=eg^Je>l^wmrkk&9|+LPsZ!0cchBARDss+TKFFqcr~s
z7kH@w56~x8Z0kQi@D#M57`4Z=H@V!}2`5cFHfFM$c$Mze_Yb@#*KsTVjJ_YZFqwk~
zo!BW6Hr8UOk3P0-QfrUBrlZ^>+7(T>u!bx=<w;eSp2uAa6}OmF`2y-V^He8nib3MF
z?96=0Ju}J7D0TJt+5j%Nh>j=0>-WmP>rxD$zMU8LJCaA2OLJFd9pGjrL^;I#r*l|6
zwK$iL4p%W_VOHd3={Q9>3QlV4viNANSNHQt6R3px-rhtldA(QfafE_h^kH1-_E#h0
z{&6yC>yfI)zP5<csJ0gfsRZcTnZWv{acMJ)zW1xv<O>(LHr^y_C#q5NUrRJetJ)*l
zKYW5hefMMddCC%*)4ozkCTU05E@13JzR3L>E!>8R9!FTiRF1V=Lw3_QjpDL>D9X}c
zckJAw_18}Tb)AiE!FR#j3<HrKt0;2vscdSCXJ!3{Z09z+o9@fI(mW?1Q;FT!l)Jq-
zZOzxUE~~eL^71;Ws>4&biMJ!PAJ8u_hug?6h?631w+w0O5^aHweMZ#NSNxLEgX#B}
z_3s0YRGjbe0AZrg{^DP79K^cE9ZkutvKL@uq>Te<=3XTRk;13kfy{8w{|Ev9ZP4q+
z^XqzF+xa3l?>^AzTCF`G@+VlC-L%Lt{_!|v$rNwL)$wD$dEpxvl3I*eR6VtIuhxzk
zyLeGeGPTkW{_I3Owteoe2HZo%F~@+|R2L(S3Jj10*ka-}kTK}3V<k1-%#^F_W^pgj
zL#DR;*gLE|(#=z59IaxFHwu8S2Xntpob*0)_F4@Z^<EAM7m=3JJg<SN7BK+YPiI4$
zbzsliJpjy8KZ=V5E^mY+r3F@6&`dMjQre6AecZ9iJ7TM}%?vkU9#WikJw%}IcH%ms
zVowPny|*X~Gav3l$#>6CstL_s(xa8%yr;fl&GsjmUq~5LlK`YSU)5yoH{HY8N(su#
zH9a2luW-J3oqh{oL5bNSD&*o~X$z#5>;AD>2sPoTc!2=nIKWicC}u`aXHS2(AD$j4
zxS(<>PuS$idqxEKG4z0zAc^_xw**btGynNa&iXs<aU>v|rIp@0z$JS9yI76eB%8}E
zYyLYUTBj~;b2pLF59YQa3GIn0US65!8Ys5vDg2~t<VIB7oiJV#CvW6U24fzU`$L}v
z1FN!sczi^IXq}21=G7C@7gfW(yYkj6^)&9lJ80E{yf<rz6aB{saT#nTyH%R$_jxjp
zY!EzQ4o)cJba(fgj@Ag+GvM8j67KQH#=5m!&;Fy2hX3yXO?PZg$Qw~VK62p&7ypnH
z6O=rGtZ{$+`Lnyj0_~nSsqqsTPfGltrKy_qD2Vds{(O2rV4%PjkKwjh^rs_#n>!`3
z@61*(D~VZnD>$z5P=b<C-}$EAL}QyfdZzxJgngcntaYpDHVPHXNAFfOJZ`f8&vOd$
z`l51}9Khi}@b;4jfx1hkpHNT@pA8IopI`p|uIZ$-YEj_CC-w^vjWxL$&pzPc&h}~D
zF##{KZP`e3PC2$Hi;JYl!NAVJ*!8QIo2?)%09!!v5LF8KxQ+$k)?Ar2D4ASZ6tr|;
z1IY17v!6`Go?K-Uk+i1#Eqv$`g=?wZj|c#RM}vt9oz{f9d#<g-u|H6L&z(WB78e#K
zX`bjIRFliEycle{_D=M;(V{DJnkO(G8YoTYqNkwP7*gwL`o<10N1uUnnYeXfr~Rl}
zIGBpXlsMiC;?{ffQh`c)4Y>)mFG;gVw@1c7w1n5I^Bn#7{tzHtxU6KC8~{x#r+`)u
zbgoGrmUS>y{`^DTC!b75MS`>&siCSquF03mNZ{bs^_jwX*RK$X+_(fScmii9ualcS
zt;FFx%(x4&J1g|Dd)c*mV+zKM%O_<TSv&}wF0Oxa?(>M1?GpArKO@jUyQHF6^Vo%Q
z1vL+c>-DImMuAMBz<1OXm_RQ4zG*Ov6Vbh`vQ*-bzmzB!7InJb_w+?O;adp}asM{K
z9UTx?^L;?sHW+kNx@5=}T4jK3THos3MYA_kT5*X`-&7!w57Zhx27LpkHED94Xvytb
zx!3jpu78!Yl*R5Rlx#Toe*(_%>iJAO-CiWT2I~!fho4&q?bd4#u#-CSfl!~>RzMC;
zro689FwTub*6x*SB9u!(p~}i`PkX*54lsyLij>2d)v&|EBqBM7x}no6B?KFYZjT#C
z;5NEMe*H8~l1KbWTdw6s2WVqOA)iZr0cic3deWr&X~gGAkyGxj_Lxroq3;nWAc-AF
zCw-4+&{C&bXq^%H%m8RL-9|wd=_b~Wm+R9%s|oOG<B6=G0j{XOzuKjn>QuBEq}U~r
zslMI4pflJlS4j#?(|yVRyXIv&T?z0@QnSr<c4kvxqIvtkTCZzcfn)WIcjE~pV!y0u
zyBiaMKZ-<a==^?JFr7W!(_v5f7kO>VCwv`*w*&z@9Xp`)N(BJt;B03N?0jM8sL5R8
z$fXL3*j{MDEdfmq3j*_6I(F^wh+BzI?4k@jt~y!x9QhkiK#qnV#T0-T0Ej)MGieMB
ziFR9&&2yv2kVrb1F31I+Y>qSN^naoFcXhE<stJEacpmQ^?d|#AL(miCLFD+lKlI(C
zVrjGpelcv1q}9iWTmDFC=g&mIKd*|a5&rBfadwaLtR6mr49Pk`^KhfILHz&}ECyJ<
z@f?ND&5VtY?csQQm&ON30<Jp*;6r)#e@0A)6O<$<NJ;W>cE19(iliCTM1h9Sv~ujO
zIX7^$*W-GW$e%hqhf^n32UXJEh6R<MUPZxywN5C+wx(o{u)1y6T#c9kBGW^;{>}cT
z7MUB8A@#AePLo0cTcK8QJQvwm(DW7IM#>GaLHY*d4U>FkX}38Lq-;bx0obi55I7OZ
z##$k$B^6ncDrCYU2B|hF27H^QHFVzFdl9_F2#}0s)vJV)Z<>oY0>F@`RaooX_#{#@
zd+OgvyHB`nF05x(@B`mWj=T$<^A_9PH!k>h!=FO`(U<IDZsKw#lWI<Nj?Nb)oMXJO
zIu5+s1zB}Eqm5U$0|$yK&MYHr9SrMTc4}GI_WRWCYPb0Hv3VhPDx*XPsgN7=%=?%a
zR@@jpNu&Gq6wIG|z4LPx6MVPjrHdl8jG&X)Zgs^H82D;hCYv4Z{MA^l07No`R~z2w
z29haioy<Ktvnrj~Y`|o9h&G-B@clHWjr}JQ0RJ@J-Yq1S&xwvgXkjpEvEiGp<K^pk
zT9oEle<yQ($yHP$Qz5nvECf>F<!|-#5EqY7PyLU+1u8r;U0LJlla@BjTp3y=U(+gC
z@lLZYT=09@wPyvWn|a^fT5fc&m8>i^N0&K7%kA(tQ^uZ3ELXQ_b~|je$QQ=Eh`p9a
z5qhU$aHunpv@=Gv?M14Fvpx#ehRR>EoK|_{<u)_>B#t)X#S^zbON7dte#&_1&M$>y
zhqeKAFBV%AT3VwDTZN=`GZ(WPJjT_gC#(@zT(pPCYY5>_`csfztVXaJ5Eq2eixT_^
zE!hD$KEt2Gu5&CGqjOU<$eX|0gT`ck3Tzx`M>Qc~Av7#zi53@ug<<RKqXYE%cb*0m
z^PrCY1h;*F@(~MKAAmvG-RX!z2hI5jt#kxLYfe5iM^RkQb#qMM$9E@ZM}CJ3gI6g=
zUTe>K_!dPmX>|lSv(6oO&Q!rn`hnQz<HsZ_tpBm4u^axTsfapKdKC~^834;KTlK}e
zR#5bnbZ3s!#ktLddkNT=V5hh14CL}t2PHMYgY3t{hh|wb9t?Bhi!a=O=ydA`A;2#&
zf%q0}WX^u^YeRnt4d!(-0I)}UUTgJVJuvnQx5z!pPIIE0l~8-Y9<ucsY$J(?J2Fx;
zo*FfcSNnT=!&{uE>MOB8K$PN_F;TuMAR@0Lb-;v(>jTu*+jSMfXNHh3yo}N*fm3IF
znJ3h~sAwU2w;Y2!3xx1PpQ&?H-yctnO?NH;W#AjI!a?VX)#I6}x>4Dql2+=(Um5VL
zaq#P>oAf${e<B(xeBv*v!e9&&3532kb=eCjn~YIXM(mWWPUj~6Hh_-iw^WJ}7-OJe
z7^_~Jy-9sO;`~?0NHBUh%PSyb@@~~!4pj^07lAF;A7J-n$LMp69s1<JR7kqC-Vydt
z;R^|sJz~x1%Y*)h4d*W_WIJ!p+X4YaRScyO@-NL8$r+V06QB6!P45irC=rRL!{n)z
zFeV;{Y`tMYsiN26X{c`ZGq)Z8Od0l?d9YyrekiC6F6&%$b1UU*+w+5CfdQa<WSRpc
z*Tv{c2l$<2;je)mAl;(~M?rhQN|lz4Npa()Nz!Ag#LcNX9@(Vzf|z(JNjx{W<na=z
zmzTd)_FW2NYj-59_pa%ja<ESix=+~$>s+VMP|q)gRmHZIB{!kIV%D9)N)15!Ge2m_
zPX(Zv$N*-xS}on&9B|WdhM|`S-IntOWVL1Lo^cy&`QP7muE>d->YAzcfpA62tkHqR
z{_~RQSa|9Xd%w9D5L8!ch&jd|Y~C_xDVyvL$)20nf7$esn(+E_T$My&fxzEQfE9`Q
zdC|s=n_$8Vdyq?@4e6Ya%vQRgfU4^{b78qTL99POER9_khdHx5Zj#)ij~eicjalz;
z)6^=VS$^6JpVC7_Ih{53p|?I;np?p%NrXnXto>uRc8k24m{g{~@Az{cFkEWl+f^KU
zNJIT?z;byp4l*du0S6qc)g}z0QCRG1jeu!FQf1O75Jn^+$=2oE++F^?)cmF3$GXB7
z_Lez_g(rmRDepQ1gzM>ES(Tv8i>Tz(e;67kB9#))U2|=RDe=#_Ok4T~feYyD9sZ9@
z&)*WGHfb)Zh|zDDHwdI0P_z9iT>)&say3~)|4rG;YqM(c%h2srfk~Tuz9iPYqc#_d
zFZi?%MMQu(ef*zkEEljce0$^bqy^=;f&hOJz^{hOF-GsFb-8RA!0FT@i`x*b*#x@N
z2-$&w5rCru8`b2#j>!F&@dNbt*Lox0EII(1`n9O%TS^IFQrp<>q{LNceaARH`MA6i
z*O=oQh1y*|(qy_Pz`}knGmARkU_wMVTz~3z@YUEczBGMDv2r*hq|ux`OJ|!T_~ugy
zpC{b%H<pysaW|bVzrK`7+tDE{aAVTgbHoEy$F8HZJ+L>eb5P?x-I)?>qeg7Gfi>Eh
z&3Iav0VF(cD{1^Q`=wjo0++#{ZBrjnC<=DvuM$7UuF|Sr40Od1!p>>z52_4XZ(dK!
z&NRJ!5<q|*+~e$=Wo#5{Q_cE+FMu4>BU5&=`F>~AQ{>BgeCWTvIm3(We<@AQP+!aL
zLH|9Bt7x2FzhX`+I~UB&g{GtYr-rnqukb8v926vLoZUb&i3<k!=JZH&JLvpe<}3ME
zJk4J+r@m|!9E_AoJI$lauCN#|s+g|s?;tJo^q{>U8Ol$ho#H-m$`HOK8~rTRU=arn
z)E|d2MXT`kA4=pqw7Jy9j&m}|_~Oap8HPV%>NOLz+@QeE$HMhDNEf9I0qptm3al5h
zH%VeW`iz>q_~tqRPdLGxas#-@Q(T+TYJclXPJscY%H(0ayp}R?GdJIGPk^Hc8T}x8
zFIo`md|PXH9{8K%@6pemT+u$9XDVH)jjk2Zc(~`-8e6D6FhU6d=%V0$;5+$toXqwd
zCV|LKtzqef+En7Qo6sM)x3_1H+_`rv8Dr+_c%~;*fB`gcputvZX8#V^G)$sWB={C}
z>iSwRIC85zSoF)fslySfEQT|O3)^CU=Au&XBBwa8#MMCO)iemEab`gZj-P&z<>YH1
z%smWJ<6bn_B5qa`m5z|M!KaIT&$F5R5vw>Co;{d5YM$X>V51P8Cvl%D(hWk~C(@BO
zKgwn9K*I~1dFd5n`0~h+0Npr#TG-B!muAql)%l`DDW7-Q5YTRXVuF^PQ<bh)45A4}
z{3K=|efp${#@j4pl-4fd5et)j-n<J#N`pg6^XMOGBhQEfL|G95yEE`^#xSFBQ=pC;
z_sy{&?oe^m)M&9r@TJg%2EHS}i=oFQ7-A}sMPl|z@Ls(3(h>y$8(6Kj_jod%LRLwB
zpkz5`O14*zn+HU!lsJWs-6wj{I@B+1d=T%ao&N|9`KO}n6S=S7ssED0kv&tZs9`1T
zAEeE%JY);;B6OeqdsYPRMNbcxBNM^CrTp157S1=^;|Rikhz8h-1h8KjTpwG`-P->1
zeX+NS(zwlOEd{$(e4fK}M<1ZXP%)ET7PGip&V3_DbMkv)72THI&<6ap#YA5?FTsRx
z!1O%<!sG?Z&db}ucp4(zg+v~_(pr|=RWEBpyI7fKVYVljylbl8vj*IhM@hsnPG>)g
zZ6sv_qGkIO0`&>(T+`_4@5S)Go2D_5ZqtU$;mw&31f~+;?w@>htZw>)iR-`#?E83o
zQ%J^iC|KR}M#TwUs>Qzc@hF)v!da0<S+^Q|pVtQrT&o?KE>J*w*SCwYP2Zl%?z%n-
zBntj2mkiRSY8t=NHgk#{t;)h95Ol`}(;!01Y^@~UI5RxpH>j925oQ6dH4}MjxjXI6
z$NJfxjB4iQ>l5<)1Q*Ae(dCq()>%T_e2(uKjY`D2ZVoUJH4(t&6pldTAJyH?+{9Z5
zgS0v411T$8RJ{neqo2J%hKkalf{ptr@c0(>C@C7q@bkvTK{}l3ua1*}U74x9G&B`{
ztTUh39s{vl@J#Q%BKifo+E;N-lv=%xkHOh()Tx}*->6K3wuNKZth?U(P3RkG1D;et
zuAgqoAISWj&)GK(F1Js7Ca>i-oZ1p=oAw@tX>f3ezbAW9YIi?IcWbeEX2kimS+h4p
zZP*l-768fTH&ChOA7F5;9{isb;aXl#5#q)f-dvJv$BUN~bhgxk8a^Yf{)$A9ADWNH
zKT?=JoWN_ww>d+AfU9Jub^{t{j|F#!LoizBL5^l#ubkHC;cKYe5%~IrdMXrnMzkGL
zDKq1f3f6x89<?$klM(lD9iFMD^T1RWyJh8n;@O{dCc4}d0dU?Xy601@PhD}5B>x6c
zFSTavyjQCeihuWV+%Z#ISaa)s@3rMai%Z5z7#E4Vaqy*qYn!>DF9?h}50x|9ACBjI
z#A{1-+J&@`-pAYVg{{v1;PDN&?6k`%_yJ#LR8($^Sl^hIw`#eR_^rqNMvKSgmOKjm
zuW|p?6D`E)O|glD?l)GEkVjX9osrk8q?HYwjdD%S#{ThLz;L>xv)V9RlSOg8ut8CP
z9ca&!13~kB=DHH~U@h>iwNkNtvd<MkCIMOB?gI`GE4F)vu88p8p7HC(m0nLHcZORX
zyKW`A41u{0g3o)@0NJJi;V&r5xSdUkGbjai5ywwF^G`(J{$EU$PIkeief*8)Qc#2k
zKnj=cg5aI7izMt)`a#N1wi1pQwF8p9OS|cbQ3QSWSFMpVd&=gxf6Fic^T`k}hAEX7
zmXxmNs2LL}#kPEqqk+d8QxCJ`v{zD~4+uq&ym<nT-KzHNWSf=8W>*NX(+$`1TG<@g
zvuE0oE|eW)=X9Sn_ielZr=qG{Crn-;V)G<SUTLqPy5wI_rctLCczU|kb#!&yy0v;b
z4XCCk1r3!z`%vdf8Q9(XXgs1Ixz3tII6qPP74VkR;a<iUm<0ou=i8h7lG;v-nw~k#
zW8roBcQlP<?vhE;vlSlzqyX`lW>Zk@X$j|=)tV<ysymVsg=YTpGC9dBY2<7tPi%g9
zRzPkxF!r;r)oC{zZjNeI7fVW2iE9d>%;>aqJioE>X1_`>`6bZ1M2_GiuEC;<<IX*5
zU+{NRf$?x|7^o9?fi=WwT#OBM`R5xl1d1zc4d30&4gjR&F?^6*p@1#M0!LPBQ1G4G
z_$y#{#k|MskrJR!OFen*|NXnG<56K!iosHOOd|2}c#ON*VzN^!mDQ^~Q>K01fHKv)
z3s9FO!*C?p*TjnJV`c5Kjf->m;hFfrdoQQO@Ogh<;YTD-WVy_wcNGSt_Y6MKvMh7=
z(%>*Qq{bC_WQGoZi5&Tw0L)?rfi(`PPn2mzRh3vsq%C%SeUBK7a_@QrqG%Vp<{wLa
zU;>XPhD`W~aNVRdd&kV61V30f7|)U5y)zDki-Ck1(5hVfCS@uotvwV9rOsYM&4D41
z%0;1`+`m2)ADjHhw48@6O2QMr57AS9RHX_;cHoL4!2TW#Fk?y?y(;Lqvh!DbvbHtR
z*LF|Ub9T0~dKV&&&+|6s>QwlCE6U!^qhwAdA~qk>Ryb)8(2&0nmsH#%^oIVp_mDqX
zO(95bc&bRCb``<Vt)z?`{ROwm+Opl(14Iy-zx^c}v|Z_;=W8nu()0J^5HG)D!SUpl
zStjF8Rdw8mdU4~T6X_3SEejOmxqTtV%}3)fe*CAcQD^_xYNY4J+owZiYN7WnQwp#I
z!y6gn3n-KG+0K#`4{Lw<=4c4T#@(MG$(AsK{;#nGI<Q@jRzlbH++fssq~0$=axBI(
zY<A|-6tq2G73&BC(zfY!NA6pfWsgeKih2O;?MuDVwGBu#ogUHu5>B$#`ky56O$nLM
z0zEc5ib!z7P18iP{A-JRa+JqB3J!Ja)I*|?MONFQeDM|{09MG%%v8PlSKCrvD2eyH
z1}Bb%XlNqA#QIFfq*~H(!W=!uxzXwM?=z|YT;SSUw35G*#S0x{V{JYn(|>zT2y^7y
zt&6ewt2nBK%TQao&v<d?vdNHe_0-s-Wnx&kpILd+H<e@m5V`fI-6W`~vO5e<JZ-|p
z$^)%^1PKuP1z5<aC^;cQeZ7;@kB1;iD`)K`RcJWXG9$o*I>DEeV8DRdo&-Ks+9I_D
zc%bzY&YYj?4=<8>#3H(}%7@Yxe8xRy=do6SPbvU9^RM^HmieV4;qKg%1leEfv&~?e
zfY_`-$~$cN6h{<5W1T*_XeJ5fA#`L~RDa1<R4zFO2iWJ}J6lhNH#y+H6$8e#niz2z
zeu3}<n#KV}R|^TO)~mK7l2NfVEvSdKc3|ovo<P`Zln8xj`jB7H`(n;h5i<<X@0G<-
z?NR$CF94Sq#NWPJj9@_KABwyXEIF+Y2iew+2G#@8`C|-9QN!XDkXUNDUGt*u>wsi|
zHcMrqC2KSk1#U<RC8~gdQF`*41Kq~<5|%s~;BSgF#3R-bb;wM6XFe0Zjoi=KEmc|t
z5qY5@ASjoptBi4C`6QjXdn#tQc*88a?5r)g6Df_1eu9>MshAJ@9z}gn{VhB3=TR{Z
zrjK`tNWP8=R|)1|mcUGk!qWFwNbOPep+RnbV?$t|%v00>d-fgRN+oVAaIu;#(zlu|
zTu4JpcpscpiVpI4f3Z3!0Txp?Jb-%K2Q}N#0jGoLAavj$k=toCzZ)vohvfYQXfl=6
zylFqxoDpX&3Htc;E_$|+6A8kbJg*OA)s;_>br|}E7p=8WU#9_%@*(-)Gq7^V<kZfP
zn4Rs15brPXn7r#BEWZ<H!Nf`$WzA}8Ija`RX`Kyqb>uyEfK@;3g9hhioP9q&hzQ;A
zSxBLv4RF}tySD>omb^)p(cIsF_<=4C2NckS&QCC#(ENNt9;Tp1ATXVE&9Bd-KK*6N
zRFEwza+lJVzI5Zl`1BPpk;DRIt%3ZK*mx}YIO6XTu-`gXYngz@gERGnXjEiEmO2G0
z8vDFo_z1GB=H;X@kIN5muu6YAEPh_1MgX>&ub)4jySli%-FdwTqqz*S6)6<6(uLB*
z6<jLJJO0x-7b(q`G~&bo){F7xs!Yp|WCp|^AampOyBN#1U!#5hTH;ALD$CB?N;Ok-
z_gD&`0Z1-Z1L~}hu3K0)Pl9mpqjcFy6A!L~qfh|K%9Z7FYd?wgI@E$lR6_x7Eox7O
z`5p}3+a<H}jnENHW2XV2FV}#}WZ54S-rk_3mI(V7{kFv0=kk)sukSRQf`?_Ih~M&?
z*aO^N(Q-vG$Fj1)>_IjHN{APhVa6j1zjn*hq8DC$W2;d}=?F-dLM-}qpw$LCUL{U-
z-1m##i%T8WK9mw@L{UgBHG^;n{ZO%xa}T54k1`<~xmEGegwaTO(0i+5lIc6pnOauG
z2`j`|W<U%9lklGQv_H4aFaP!BpN+2Q4*q@;_z^DiW2K?LeJs4&`>5ok^|iM1V-p|9
z^(@SyF3UCteIgYSyIjB;IyJLk(-?K{(XW2hyZMx(K{vCHq<itYHcy7|NH{*gV@tq+
z`*44Z$TzIh9h<AOq}hyu8hjJ_CqoxFx8`2`@563lr7EP>tNwLYPWgA8KI4JO(TA$s
z_DnX~XaFO-pq$_8LG)00ztzqwo{V|W4QF76C$q%+uvx?+z6AYXSK!ANk&NNPg!yi%
ztrAHDlZI58>A}gGO}yyY<~Q8$bF5%+Y{u%b59H+nt1I?aRzCl}?FAk7J`nA*QT^1l
zafPo&vU(n9&i(QvDfxD@2+5Uka5$a!dOItl>9}n4_;}b&K$uY>3?#JES#H6)K=^+d
zi79V_&D?elZQumDrDDOTov)>p!NAhhj+0L-5X!pM$R@^yKGB^%aHyQ8h?bvfw-lxb
zn&Sc?#O)OTvPlA|rAHc|KhwuyCK9^vT*CI)mLnP>cjOKP_G!Lyn?0;xdqC8)mLLP8
zK=pArVgAGmy3HIb`jNn;6Wibc2Kxr(vx_D|AE4aq0yI7_R<qm6jzdi8`r;k4?6V?N
zoHrrnOEd2?1k_HJ#?pNw0+=^Bj+3g^>{&fzZ;IjrTct6BG2nM&Q#}925&h@NI;lxU
z{f=SnCp1C=Yz-v^H+5HOnK0c?1Bh((6<$qjH=@$f#?lfjFj3rBW>|fVUz-|L%LEB*
zm_|60>F}&dG|dxSQD&tWXrE&l5rJI5G)&$wsZF(85c%Q3Hp@V5fj$n~%+mms2r#_?
zc=~{SCId#cXH4K<xJVj_D0}}~I_k{Ie0C0QVRE#8^4uFmq}V~FcbjQDRWlsGT2z}@
z3rhJ&BIEZ1UjJrM+~GTkD8{EEA4c<4kk#KB2$xkVDE^O5iyy-s+xYn)&w;&ht48t*
zU?tT|?O)t-RD#eP&=17I%)qU>5FC+kk<$&x?zK-QD%6Kok%sxGlOxCWX6a)-uFnHE
z>)F`2o1nXEZZJVzRDCnNkuAg^NyC@3Zqnq6HX<gbW{F7+26o%&ilB_+PVPMQ5)<QH
ze|kbc<@Y);@8y{}RL7t^P8$;-e;FPu_#1PH3ve^uW}q-WY@Yn{b^}>~SNV<vz(#Rn
zIeY=67ew=$wKfN>2Jg34fg0ysuI0UMEW~LGw>Jfj_$K2M#|WE_$_2&nbGPyOOibG9
zuDD--?w1VZ58Jn=WYNey4N|{X5v15Kt0g?u(Q-~XOZ4g=KZdz{bAYL<*}A*8XC;ep
z-uvEB$?@TOQ?gPo0BtCgJsqX=H!WkrQm|YE>q64VzI8K@YGjl#-E}1kBqLNf`!A7@
z5f?-c4$LRm!RXD-5$mjg!iYll_&2vjlp%h*=d%V~I*Fh?d-zri4;E*N`Ha5WZ;*iN
zoq$ECaES~Zwn(@MU^0O<tNl_}vdI2F?0scam0j2Frn|dK1PKw8?ha`wMOs2yYSSPs
zEsb<{iAZ-hN=hT$U7I?K=l$OI`^Ncm#u?-MJnJXyz3&xs%{i}YUTfWV&BjdM>7f$=
znuM~F8vbx>`ZiBfz`;7^_;|gO7VA?{F>sZ<oMOidr(6cs^74^iN7{@EMlF)~d0~Ub
z?Ac!)VK)2m_*;$2P_LR~pxA$H3?`k&3#qH?0F^GQxcZCs)!OHQbGgW}02_|^cHaOz
z($R0BxU<p%M-MnJ!uhd2?#jfeUw+43lAYudwnn-T*i23Var%4X1}e}4FQ{vB0jXE8
zEXPXPe8p9f69Li`mxLl*I2L&yklBV@I(%qf0>yV4jjdJq_g-xCrpdy-KTHXj;d_9i
z$8TgPd;jaAnJj|#H7#r57p>6diPw9mRmVwojW{=<6;zXo1IJd3QoV*kH<aL3h4d)@
zHgG&3iI_987ThIzNF?~IR`c<W_qSP{^*kHz;UlQYKg?Ro&GCSvykY8ZKQFHbMn{K(
zqaRAiHGV$6Ka*^If6C+t_^6~J<t@zC-sl4z*o=sGuUGVYwMw6QQaUc!8e;qhGPh~$
zTwy_QX*FKBOu&pQrhjnlr*{HW@5HHp+m#&6P)f!CaeB!0MZHh4q{1VV{KWq|vMjvj
z-5-Xh)ojJ7HJ{*3f3-`YbaQe#F@y7y`!R@c(2!%aM_G*<RN!O<GwvI??D~gG4UjJ>
zhY~=6eq!gzIU1SO0MwNShQjc)z_4Xse{Zl=VzE{?R?*N39NET2!Gfb4c`!;bYjpCn
zD-mb|x|9<j!kX1^noaa^YjU565At#${>hkO(xB1Y6!i`f@jxYcn6cOI%JXn@@2Db2
zO+&7{)?WV;YA36Z#;<QfK(;)u&eJT<bJ|tJ*>_8=G2HcgO8sMG7=w*Y*e4iUSIb?r
zw^pk%1lMGh>USv1i&QSf@v8Pu?1zAObi7qwDO-=K9BGULOfPTAmoIs6Ln9HiW8TKC
zq6(&gtMz}&2GM~KZ|lduuIb>D4hipg;Z6w-m@T}83(XPF|1-ah<X5m-BCV*@$!AIq
z`O@}2b+<R+&!l-ZJUB4Jt?m`}<_(ZpzdINHeULV`+-FGOC;0-C?0`9#TQ|CTLR-aE
z;@L9?3mfVYF_<%BkNl>kgThbBf^zJYC)>ss7fOGt4yGG;@ukBzm3lAQAB3n!?FUpH
zp4ea<ne+(2^6Wa@4LR~z%O&^Hi$6X*)F!bfnT@tz-0ME~hpc<cl<#DGSxB=+s07++
zpkL|oXr{4tV+WPz`L4m_6xaH}JHUCEX-y=1Y8~awtrkqU9Upxqt>uCIo{BJhWhnGO
z4|pR0lyfxV{JSwY8z?Tor4)B3vAwI{X6SD(=fTw!iq++QLGEjQp3!@#bVg}0)msNt
zo#REkR*NeA{T)0HUcDZ+-dM}3$iFK@jZ&S;cQE-mYc?OMpi9SUao03$qN785`Mh#+
z5~#Rx624;0;?a>#xpH{BMk{9%)v4=j!O;0bp+UFMbn|v(KZm}o5L9Nv_8cf3ZF8&)
z7NWWf)XdvQ5ntUU_Roj$gwdPx6J=;dvfKARdm^Wv8JjnOKkcHntzI9`yZf+5<IjpO
zH2jLVkT<>8y}L%C9hFMM)(eI$@J3UgSu4BT-SW?f-l4kpOs2^SKq!SVo%b#KypTy(
z(MYblC(BY4qukDhFfEu=#4NXk;tg%aUyd6+TnTUGxtv^C;UDx_gbH#&P~l8H4f^&y
zru7Owkf|&1%9*)nyR{)To(d?Hr`9&|Dxb;FH<7VH7f{yL#Ro?aUlHx(pO3lN@_J!A
zKj<^_P&kz!UG80b#Ir6}6e86z?z`>P--}i%5O0x;P_RBI&Er@!x`@%dsGHic_(A@H
zS|zQY_x@sm7Kaq`3TCWu#(c}Wi{c5j&1tOVRZcr|)_)_$I@?@!MuU0YTinZ^_l$QD
zajy4}{Nn%%U5$L<uO(a0aj9F!KC7O(@d&CQv+ahSVnZn@<)n&ll2X%{!mJ39nu8IV
zIq^#G&`c_Ini^zZVO{Q_^YW@N`c;$%Je%0#OU;N}MS81x{yjF~m)7OUENrj^RXzNR
zqs^i1j9HWQJcRK5LQA;;A5L$vosq}BG~fNr+vMz4>bxR*wz~lxy}c1Dc0?&vte+F^
zA#oFU1G#f7?6Xgu_b^R6%1~n#FtiYf?OLBR`+6P7nm!dRascyJ8g-9%vW#bXzY*vI
zS=)7Ajg^=0`!z%~dH2MG|GV{pjUx$e(!#OZ{L<g2i8k)<(Yr!z_HoS|oRb<pHh)c*
zL=6cs1Li<RU{8z@mz6_XV`bp^Y10cG)z@^uE^UK@r+XSleV0eZJxXB6q}<%J<Dz}8
zeCeV1m`vTx{*r~iMq8e4yjV`}&1?0vXlSf}^Jfdoa@VG~F>Ly$KZi`-eY=KKxa@_<
zGJQE4tgmc+&`R!qa9|JjTs81nfl)pU^H2OCx~}gL^^Do~E*6Ky=gx^LvVvGl(4p`E
zw*EXAVRmHSo()EMp(5&M@@un4>PNp`fttvH7qYBJ!@am<=I+?VL6FI@#LT8QVsD3=
zDz9CkfM#B+Xl~ul9_ZV;`qpcFeiFk6YvP~lx1oolcAU<ASxs+roZPt~Vsq6c%wuM@
zKi@Ajk2l}fOQJHz+VZ-P@Niz<N6#`zk_l1zs!?o*#U<KXO^kgjP(Z?6wik(PF>drj
zaAS(5TNkhUs_@$K4|%qd5F2z6LnlHba->HV%$BORqg}+|rPL{LrO^HRzy%|2!6HtV
zlCo}PDX9o5C<OPHri&nD7>R(vBvZg~R6d;deKdY-CMN0oi~N?YPf<imPL8sl!5rRH
zq_!Dg`}i`NPsOcK;Z!58TokTcy-xA?HAEhy<*gQSY4=&uZR2DR$HsvQF`z2+p!!ez
zM*jrIhL|(MX!D#C1m3>%9d?l$hJ~_Ys`ir_YcFRpgV8nWX-P{m6z6jm2I2mAR^Mzf
zFbao?mS_PeQh-=R^anP<#*qnr_HSZ|uk(;^k*-yJ-lLgzu-BZuaiud5W$u)Llaz%I
z3>@R0towowzhvjpb@uU<UZ=`t?ds3MEL*)rmbXXVH|F*RLpnplSH!LFMM5NogAgBd
z!sjYX&Nfb?=flwjS{{hiqx5m_*MDveasfYpb6%ine|sw#*wL&WtgPCjY(0Y{m-fq<
zkJ!t$7`JM5jxjpsP4PXMmsi%jH}78cVu{mtC8<~brt?pr9*Pn1Y@H6h3zfsSB!-SH
z_(U(@PX~sfabBvLSq3jMf^Z;bj^+!>f@`4-x4Q!fUDA>`&a3a_p6Sum6xF3D);7MS
zojibX7c1>FWDFy9I#cb`8c4$KVr!P3@qX-$>frQI4!jI+_$VwDo3Cn{24AH4K5tX*
ztI61eZ99Kh{nFeEgG=aJo)54;<CUW+dYoG|&IA*7iDJpF6uE07n(49%a(=K*^xE>W
z{Rbjem&VJ>mYrL(d2((-*xOTQ8A#p2qK$-@@;*|%hsu`usF<KL{A|ASSS0oHxshp-
z0bjak6zpf!%Zy?dowv$9orfk&#`+uRdo7-b#pU$RGP${|%`5+wUN0aCb0lC5Rp1PF
zFRgMNIgY<C+|@#WynDjkE*wL1Z!d%o_uPt*bPjJVTP$!wglvlG0nz#Va?E?dH1}ov
z6FGKcy7M|}2g?ctlH)o_L`1QjPt9|O&k~))cIG3lAHIaVH-z|xV17fsna9bZD%2U8
z%tZKxU2nw_r|Z4yV4oLGiZGXjkY0qee28Oo9+>{-Ji;G%ve#>pb<V2>s+mODXW!#7
zF5Wjeo?v*oxpTMd3n1wE;U8J=KD*kl;LvnehJ&LLNVoN3g0xlLTGfVqm_(=Q*`53G
zEZY0`gqE>ev`1{!nDf>)Qx?@cjxZ?D_&}uI3+3`0Q<Le0^)at{A(B=Um1^)E9=@B`
zAcPdcP@1?)p7{+d%v*eHg{(52%CzZ8Kw4T%^JlB)3zOj>$$|0H1ci#0_8z&&wH&8O
zCYJ@OMz-gMWIPl<nmuTt(3ldf1UeitIJY)Ki(YIB!E{egx_3R|0RbA;hiFJhV;B2D
z315wJk<p`kCq(|3lx&<2Ia%~ZeX}YpR;{)6<tmMV-;~b1lG5?MWTwQ}6c=@D_<Nex
zK-$hz-q4sIlBA<U;|_oE;EgPtV=ShpAjhRO7!;5os7gv8r<qI0>KIahp1H9pxll!9
z=4f9?F(rTn38{SgJTzR8sUuVivxMabf=5;wA2K1aL_m<Z<un{U*T7l*eAdCNQA3%_
zU%s<c4C<xs91wN(rx(UXGr0ZkIIE&#1jvgRN_NJNb1Hp~>=K{JpPg<RJ}fU)-X2sb
zykV2SxmfvPqL!4PdVZ+*3+di|$^$)lcXPD;{(z##7{h#YQ}6TLc$8uNDzyom3}fz;
z4Q);6%Lm~%PhOgEEceV6KhcbC^tp7_V6JyH&!YS_B=<7O5f0YBI);HkE73vQT?0oL
zHrRO+l%Hv5p^)7B^ZjfXkvVTN>;t*`bJ8;O^5;8l)tbcVVWMcw^Y)i<t3vnS%|-v#
z0_a;HutnDS+POTT9nd<kQ-+K=_ubjOxPKZ(gN^0yFh1O-#7l8W5u;&c^6?txtvY?H
z@sak=D^Z_5|DefRAt@>SP&f!Zyy5&>L;7v+B+)0qpArhnoai<Dbl+-y*Atf7T9(Jj
zQQ+1t`lBljQRZ$Q%AL+r!7;l<wPyRla#__vB9$?1N)dCS^zI13^sa;5&6UNZUsw0H
z#GdKv0;2iYR6bT7H15GJe%|~fv+Z}>FgKxt0iwR<8%e!~R)v&)FXTGSmtv#!5D>G*
zu&D}<QbTO92iF9_^X(fOpV!yPlT-adsg=TN_bWYA`}0BRR02a^NTW*BeUT_fuDbt>
zaqV5wUwv5NQ{qu#KcZn1U#dw|2xLq}TMcp4xF|ZKiKkGPr=!ZcN-ezc2crS`b&qUh
z+3{LnV=Nbp^K?H?>r<u3F;kAk=CR2b#tEH-1Rdkgikkk{mRC|vjyg1VIbZWP`{=f=
z-EhUlqw!<j`>4p_OKCVUK64#-4;RfCVwUa}^|3&?j>vUP?^a2EKQ=2_ftNm9{_x6W
z44cMxHSBt9H1YM<pVpkY%6f*$_N>c??X=&@B}HTR7wOEuTM<Jl|74~KVZW~j*&4ZY
zLajAhb<P_kT6p-Z@DvR<>Q9=1ChY~98w|mnaJ_!@kDX1-$bxON3AxbYiw5g;3><~U
zD}NuES;GG!j%8RC9L(2}dgyS{8h~5r!%s7H)fFnZF;Mdcxayl$1OIgGZvNM3gZ)re
zh^kPKJ+$;0cSVDYsts<I5jS?~IjU_`Z+7phL9&4+>>l$|pR38kq~Y}R@bl$A0kOF8
zQdqzJPVB9wbFMzUG5(Vl*K)5a6CN~Q9kuIk#a~G8=Vz4c$+UNBrrmNnXNmNR$i8sI
z4(;~Vuzt5?R<})DY7L1LLQ8M5dpVn-$g4elyT%29KtM{Y!$7~HQ$;xYxmnxpC%YDE
zSoge=f3{eQ5ucPRA{%rG=f=F6ML87?cbML&{?JHK@I<1h-)EW&eRR2I_4LU27U7ER
zx<c+Ks=K6Dq|DyBkP-Dk(Ur6{f(!#PR$XnG)yGK-?@zbHN8HB^k2%|};P+yE3kj)`
zFz?x?QrJ)9S)3A_xGV;`ONtNosUS`Axp7Xs_B$WK1o!d~OrcPh)ks)%BO;1yUzuMi
zXbtu29Q;ta8T8hvW2GLp9`Q%*kerJX&aNEOp(4iBftE}N<r?UqPi-WPmd$aLhRIfN
z<T)X<+~j`~?}jbGWtZzC_0}%&`KXU)mS3a6{=AeK{`{@U{PeWM_D!h%vc8903?5sN
zzb_g$y@Xh(6-HOG0XIAJ5KXp+o`Buiq-kX}i_pckWmE4iqEGKlR=NJVldNL>OR)o9
z+}UG>%F(+%(<AZMH%LFn-&Ts8@h$IeYSFMV_8)lJDoozmekbavGA5gd*j3z*o2Ymn
zbSIi2*53#Z+=KF*g$J~eP0?V*69nG52i9EdwD&*NoIZ27^sHV7ym95dctyExrpSQ7
z?s_NoT5-S@{s;6^jEv0afPF(!j<MvzOx90lg0*{IjaiLXAJI)u(kgvUTDHB<glSMU
z((}-Kzm9Oumm7w(R#`|NjG<v0h-SWM-7!7>Zk_A%vFm<Y5F7U8+_fa{v`x?YGZBVM
z4}-`k2_zGJIXPR8eO>z@v6gcSpIGG}w7<T}7ds?GPVUP-xs<zV(HoLJ0Aif~XqWqp
z@}zu4E_cJ&3cfJ7d;f+lS|Cr1%q3@g7E?-UngtmWf(q+S4Q2h!l0_wIiP^-G+(kHS
zVpzWk6%&U$!`Ot#`~9@Tp&U$1Y6#+TA_V4lR&5?mWT^b~0ZvRbE#kI2!^HFqIcec{
zEA6gF6X^rw1Ko<9&EY?Z$ab<bUo;dxH2T-GN_LsROPNY_ccQpWriZ*-Y}Kdp?%cu>
zs+GEM5IlU#HHLj3gtZBM)ysWOcp`|D2H{$6<U!urm@`AStY$ayKF`eNaL{?trj-}{
zoc8y>v5`JKb|yW4W6$ztT$Y&#B_vbeX~G4mg66(|>&WJB-NmD$OHY$4ip+k+x)a1b
zR?QXB3(%$n{u)J)W^!#^mTP7=Vw1P0i1kxHYu=zsab3~T@(u#JELG$vUz2sN%c!!=
z$!&L-JYS3MJ5xab)CzYWQ0focb__~Ai{n`|<c1|QWY_e)GHn+Sg{wpLAl{FA|9C!d
z;f?~$<(GbfQL;mZn&Op1`QU1v3l;4p8D!3iL#zH&Y;?qLHKihR-m_8H%Sj$p?$@s|
z<s>#SHmP$Fj?+^^kB24Ahv%&CaqQWye=|RyWBcAQKKs?*HceU3q0bbO>5B42Gx2EW
zbHY{U7iKm)cRbl!jO85pt(e`YEOW`_3NpAE*!|)W;eMq@VA`a?PjNwaQrJfm_;*q=
z;s?K})C-=Qs<(ycToU`9wayc!A*@#Iem}i;8b0rJlr?^__a4=g8?N-4#v1R2_Hqxw
z^eRFcsa`MqIE-4$mUQZB=OT_la(NPg{ga0opILWm`_E<8y3#_6WV`OdYNI+q21@Up
zNon{wKaSNTgGp|mP_O5Qoi+%EWfA4m;S0z%!%>=xmfQ7-v}ka<1_qo5x`DnH_}Dt{
z3v}t}vr8VmA?@h(TEpuGTkBJtS6_*{rb*}pdUV!d?Q>?ETUn;2Vp^JyknZkxx5C_j
z$Hj8gwVG>{Qc%$W&4`L@Rt~O6Xw?qm?^(n*CM+3-xGO|<$|a01PzT1o_F&mbRdcDx
zqwR;D(C@xDrI+7VU{^c644yBOjcCX>Pvxpr$COci%Wi{``@D#|A^&anwK|o@oY=zm
z_6bh)Y3lQoht^EU#e=8%Cw$2_S^IMcHV+l5G7%*XHF}a`zj^WPe%>HZ&)$qG%t`BX
z9@*ZAPASj;a}#fMAbSK4G#}16;f>Dwj!$v&p*TekH*gLaaFgCkp-1+sI<eESF#;O9
z0#e-hI$-y~&4?y%a}h<Oq6Rojdr|9ww_01E;o+(}E5!Z1h`)^pG5cC6bV5kE#8#Yy
z-8W65@X#KOf)xlm#d!*)MlGD<5vS&yH9?BXD;HKof#e1}aa?V1*u!{Oy<ZbG#B_8r
zL5br!o(fel_~dBqzW#l%0Aw`wI$jDc3OpQ@e`~k#l;05(Nhxg~EY?skLw}z^WZD{>
zY%uQjK3aIVh+8fo57^}=G>J+X;!2&8#(r#q<mq{M#@q^?{c#XInfy*Lb(t30mU5#j
zIW-vRdH%9sMF&Qdn@~oJLy7Esu0WE;qpmmLFI#tB%_G%)iKORncRVFB&CMRVZ?m(g
zV{{mAGP$i)_6BPGG88K%6U_8ohg(SDM$J4cJ^kUhuLA?FS~HpxgKUHE*Vfwalv&DK
zf8O@6@W*0opZnir9n5133yy2t6hulhbQyVbH<;v|#&MO0MB~bp&DA-Fzf{+F1$Hj0
z;yr>4wPJnmIWUGV(W;w)N>*)Y3n(%-4HaK#ai~7-G&4zb0`RSlBFvI^mRnLTdGl9v
zMEk{#_8eaVd(ei=*B;oJVHny~VP^&|t>BX8FGgm#UH7J_gcfS_gm@;zxGDmp`Ukua
zh+~H6K&4vxaFHICZQgl<Yt@JKyt(N=wI~-jOWeZO93abrp6;2Weif9GLoIFf!5^6U
zg3DRe6UjwK`%OlSnvXDnsld%UG46>tHQ$8PCZb`LMm>XUKv4ThN>QcA15_vfmUv)I
z5IoK3>h@Y1l=vwcX_S1lcY5H%Iyc~t&HAR#d5DcPx{W_Otn$^4R%8gk*^HK3p6#({
zXMCma8#a_OZO8EOS-r_r5Lk;$i)eVM-?l$JhO|F}H)B4GxTg(;W9cS1h#wKhSZ(we
zd*Oto(6w`LNI!D8)3qN`opp8DG8G-#Qk$BlAoz(A)}K>(e~*?U_vskgP(cIYrXiRI
zW)r5VLq8lGJi07fuZ{lWF9&I@aJf=WGJ<Q1cT)Z({W4M$a;}ru5KYZ0=%K##ViQ@T
z%f5t+3^Jj#_}ovMU#&hqn4~WtS|yX)RaR4_?croES%e>0E1LXV_GjxoMVOdS!id>B
zdLk%3t#lxD_4b-<jb)L(FodkJ{QmH&$E4pQWEmHPNyqNLPGXi6hc4+mEaL-lc0M2%
zulSfhck>Y+ISr@huNg_P^&HO<$^dL~tnnfsgs0U{ikcgHcF3n*Q>+r6W=Ub3=~+&%
zq@Z)Q1yVxTKE4)foZul6+lc6B@#|6Nw4OFMk4x`@bXp0Ctoc(2rK89Tcm-vBdOULE
z!DgI^{;rC2544>J?|=&{asaO~`_ON5vR6WE;r8ZCb|@qB!VLFk>eVKYlE_~oo6y4^
z0ljryQqX%@YGW@3PL?L#9;vfHHcKD9&l5awQ=9k5H%;Je_XjS2ln(?e&9<F{#5ETd
zjq(gy)?8p{cZxiF8N=m2blk}vyIJ5cCp0N4&8c;GL1u_p;nZyA{6aDM20Osz%zoVj
zlP=uv@+pQ#Xr^IEWLla?fq{$T?dk^fQyyV=Va_X$jmcL|>u<db>rc}y%L+9YNDJ!!
z?rg^YCCDo0G{{g0g?dFHL)24T-w5XZi0zGVXYoEyl^8?F{flNC+i1s{Q;-qRVCyUo
zQBqL;JZ~;W)Og~=%p1v3MCPm~hqAYS$4_kb;Cnb~+jFL59@e0J-SxxeiuUX^iBbJ(
zdN#_BcMqwB94CdJ#uj}_20XCvrV!Hva8SJ8TEph4;B}`fp*39#9#SgaGWd9Ul1(g-
zqbXgxm4XZwY(nV4{FgLAsPR<k+pFo;`#`f{mhz@aXDUk~>F;L>$l*5^`<>OJA{b>8
ziRbCU9+l>!)NFbUgdDB6T%PX<Us#3r#$0|y!3<grd(ik*Y+-fg+ish&v=_R7?|q*i
z)=CsM2;K~QFg27fv$D#*n-!k+0d}ii!uLoV>K9ld!~GC$G`zXAs+{*~J^g}ILIjfN
zK$C6lM_SUp^J$GWlIdTYn(ybDb%I6*n;VA{lE7OUy4}7lENKn?>F6OJ$_xgpKCQug
zhcewC(dc3?3SdS2vnG~T6l<})bQ!Ob9m#r%?e_cV7?3m=#sCZ0tGf`|tHx<6nO3D^
zn0%u97!>p_U!DzRpBYGqSrVbRjj76=r@~Nt@g|Os;n0USx16C+A_$)r+e_e*0Cu_F
zKcvgfUTDEqE5pyW)jC$dXKPJH@4Z|SzgxmF)#66*MYTgV?f4*Pv!JWD8+vpKwL7RG
z&wz{97kz~}6%3a8z(-~wFuR+Yo))6sOW4QVmsXzf(m-^;^Z`CKN;tRc0$1%<IRZ)E
zD4o2o`n65j>YSb8e)Mk+y#SM8m8Ve5SJr^Z&W%j7YI>amcxmA^I7x+j<#<j4+75<r
zn9zFYSlBWyv&mD^l8e{JU)kzjx+5M`*7&nuVx3MeR!uUaV!Q6Ii9#3_es|3+Upcf4
z0-K?PA6#gav~+#)_fB_Nb@ADo%@$$HI;O@Qe%nQ#IxiJ)G8WRr2EoO{`;0>E3VHVI
zSx88TFL0d2#eB{?gp1xcHk5q!kWrBvHj(=)6I2pjQXG*hTA4`7?I~wRN5|uN$5HX|
zSBl?Wp~W((1Vu&Re6OxP{{5O`Wh_fFiPHo*-Q$42f-j*dHg%Q&AUw@nhF^4-RH)CY
zyclg#3Dkxo{7M6_^R?*YKiPa{O<ZXB=#q3LWd#zf9H#KentY7aczS3nHi%)AtO;au
zQ(-8ZwrdMhkCm1!Jsldifh0<{(H#U(8Hf&!^m8$!vFAiNIuD#+!OnIIdzV6s_vzTS
zCULfIdNCG+@3>#9pMR}J^lQ>VNCcZK{nO*&Zc=fEE9%O3JpUw%F`z+Sk;Wi`7%4Mb
z>gT!TChaW0bL3?hL`Z@&yk?7gbt5U&z|Uz{O~#tMptG{gj`aLXu~JMXCIwR7l9&a`
zB?SgH8!_Or7?6Tdm`J5lT$g9?2(XOyzK2`wR#_!h>d@g(!n-ZwGAw5agj0Rtrc^=2
z*`s&hzleUKW7~)UZi+`lM`sZZF8>IilL!cnzdzc)X?5r>d&Mw2V&hAK2!V)^iIQEM
z#f~`^FVg7}ffCi0n>H<csCqCuTs_pMuc$ebYSs=hF#r*d2&+g*B2S#Gx7hSRZ|;MF
zsvT{?#j68ut>-U)FAy$SUBM&{W)V)J@noowA$Cvh;Y_^UiU;V#9x9(F%~%`mLtDsm
z>s&1whf#_w*I&kjY>_W(uAB-UHWB2=L`WzqVwP&L27n_HNM*XUJz*RzsPZxP7kh7W
z3mTv6Jkd01YyHrVUsxRKeFrldh<kQ)ejbPmucoHvx*o4hJIKXjJ>xrGpyJV8wU!s)
zyZ*B=JVnyHNRCa5Ij4s#wtSLqFxum?#NDrWOJCza5I0-8R5!7Xzf=@nW7?;a9lh+z
z(){n_J2;-x5{E%2&Zoo)9yGqA1`lw%tE26(UeQiiL94iBhBb^5mTaBFS2&2=phl2+
z;~Z^AiVWt7LDqr2JTmU@EIAx9fs`2S+H<U2x%4QNG!>XY0^7;Vny8y^MTQ#lkIzc?
z2==U4Up>8iF2_sKM9p)_xpWN}FY#4%u0N3f;_e)!QQs^F*(f(LUsuif)sAt!c?OGi
z0m6%?4}wl77Qaji*oaNUgHU=14kE_~sI6}^=FNR~pQ;UKg>$`4X*2UT=}%P7_&u7=
z(=3PJi-%imZ0>482mF!SmAo99d^&$tHI<FlBNzuIJ^1<utR9Ihg~d~MGvpuswo?s-
zM(8;`GCwTPK&b67hp|dul8<(7a&So2Z*P=zH#k$}Fu<0V<W(f%at%wbpbNr$Xc&U6
zh(*}HUugx;-wBKhLIDZJ_$)6Zq)!;@R5z3O>gp;}8pFhRcOagXTCk#gSEp&FubJ2F
zWDR}v=H@1Gk6EYMOese+p(*u=I2k|jlzF}+8E?PN<hVH`?n`Y4Y~CPZLX^ACNW+%Y
zOwm|d24zj;j;=<@H+6Q<b*PI``ok;O53+zV3c<gT;rXZ<708_5{;*wj9y41n&frmT
z7*H98j_ugEajdK&02{`mHnQ$*-W5TuMv6zuwE3gD5u`CX5@50-G;@D1gr4Dr8R+AM
ze4!%Z>tbj(4Ew01HnPgXq!#GwaJoMN<#>+(h~F_Gom@PKloUJBNS-?r;}8G&61(_)
zsjvi9<YZ>$cPnj{PQ{;z;h&ty(7zgDR!7%>6)^evWre@YNT{TX7H}@R*AYWm`z%RY
zW3Bh`<$m<_q3-?>90T(^ednHXeN~O0Nnc_6bif;-4^&dn0UQaoxQ@4D#kFrDgmw_O
zy`B3Ys==#zt~bPNdW)}rx^_!HCs|FybBP*cn^G$V_aPnTH2(XCqpP1I0(EE}aOO!D
zQM9%#m-EUDFz>$}d6e!Si+p=KJo&PS#QVNdv8imwrNz7FE|}!b1|M%S&*~|}uy*dX
z)_rf!lwpkmKWNNs@6gmJ4Vf#<N7Es`N+~?tFq7?BkfPCD9^imd5y(y@TFNgc_glD)
zsyf`44)rSH5})+>%Zibi%rl2=u!o`}XnoR`_v7;B$_<d}`s06eTwxPk)4+;dA>V$?
zt!KaI^UQVKknP*dV7W}ptOt*59UysO>w`}zDBR4>2W9A3L1k%X@izEtMr|Fbmbb`{
z=0;E+Z&q;p>>dMX-d%(!@mCS{_nu<3pz5_bWs72$W1civW0`=8`U>4Dc3Qnh$<MR%
zMIr$^sDoABUtCVI46uF+KnT5Fm6F@yrf6a)=5QDl>FJ7}yk1HV`4YI^E}HaQU|u!E
z*H@kOY#}Kw)|9zZR+5Y}e;^-LL|}U^uOj+(h4Wc<vaX<DK*~M^#;<I~wf4>|#X6*c
zK2Qv3e9un!o~XHS>^sF0?{k&-gL^&&<?VOQ40id&_P3WWXUto1016h%s>0A5ZANt7
zed^@)fZ=JA;``d)yY1LJ?33*>I^|v?u1akSL5q6l9M#+Oi6RWP*6XrC0Xcg5-*uuZ
zeq{97mZ!z!kddv1$h|#l&!ct}RZSuSCglJh#y~TzQk~Cst#`x)%4rf77R&;mXiO&=
zM9a_r=i`E}qjYygM`tIU;Oi84T7lU<X}2}GKerhX$2!^Mr>DW-ExuV)g77w%%Aj-x
z355ssPC2~>?HO2kI5iI6o+@t{)LW#4C_%R%Zp}aL2FDM!s=(YJrkJ%0SAmMv*wESb
zW`GF)l}C8O1qP+TUF~lVbLUg!w%%IUv!r=iV`C$)6O}QS!vGh1nzQ-WxZZHMHnS4S
z+avvwQF%zrEXBeOrLwk#MR2iCz_e4i#rXU|W`9Ygsu^;GD%bx^X5@J1yACa5f!K{{
zNqxDK{Q3@mwmS$(eiJPOKMa0K2znDTvjw1Hfl66B^Sed4hv)0*UcMcE+4POd<fumB
zzdSu7j`hQI<en`IlTf=jTIUH`9d6<j>7fAPGDRSGo^i6wipT6E`m0R38<l<iUm4N|
z1{|`L<gjAw+HYY152rP{ZOlf*t)@t;*4jNgm#bfj&u_2Va@6tj8~ZL<Ah4fOAlIK}
z6rtkV;3#;OBb&LoDUKK{rl9CIQB|#tydDy=6Qhj`iKG$?>ZS3aY$%sxeaAVMEsb%z
z^l(Sh;Iys2T@{Q?#d3WKBHQZO_W0FyLB?3o8z#jJ5vnPNBF^V}Tv|KCDoKK?+HvC}
zX4T@FlQHs_J+9Vng9+@N$Ezb4MH_2t5Xi~N$+CSN?^uqNST*x9GE*S)=?RMnk1?Cn
z&4&<n3OeO9Uwu^w=vtIYfVY=up?JG?)UGh(_pu3Xe6GbWFV8Hc76LaHkMpPddTMMv
zOQB};G_?@g{=Q)B>s^je0WCKVpk<j6bpEWbU1ESA6Yit%Msu?@$0`IB^<?aVpXyz!
zIP>1j^*0}-3Nr8ew(~m{5&iKcp9c!sLnX)bl=>gbhBNLLK@o?%*)Tf9v6|pxm5<lg
z=i@Xg+rKEXHFZOjtmq=CK6`gF6PEPWaQ1v0aUuG4OBvJo4bm;hpu`N`$n0**X~CT9
zf8$$xATQBrxVuQ;F6^V6&;n&)V=Ha;C<>7a0YZ$j(p=^WYzqdN9NMO2*&bGPxkBA9
zTN5G(U$v#_EbN|se@SQaA~9jQ#g%SiW@==&o|`zWf&3xP1?%Cyg1bAJ#ff+4C+=5%
zeKJsMsF_C}1}kH-)Z(=@Q)Ln;@7dPp@o?0M!@$i=l$Di*D?!V_fusA=x_!>BHR8vQ
zS55b@UDt#9%>hbtj&DL|MIWTCxBDHH#ue@xIjMCQi#1CsUH0D<j5;68lR=_0GDabg
z$@fx(FLs`8j>fl=kwIUUQcSlsm?B8kqj}dj>k?v;!LdIn)q1LT_xEoW<*b3TB$a(v
zcc=Fe<aTI&hPZ6V!FtZG<DPfoM~Vi4csnX8TA&kWuZAP|uH5E`7~w;Dpf5QUg6<)l
zeOJx;n&@O|6vx_eZQEf967F+9h`TE83DZ2bCd!hd)SBC8@wVl!BN-*_=8@dx?4HV$
zFDJDwam4gv?Kh_}Q2n@UxjYluN?-7>%|@fJ^C4an*adIcYfv#4N(O2oL9JNw1mk?g
zu5RnHJ7#1suq0P9jqJN>PiLUPzn{W+3)ZGsmaw9q%s!d&5X%?2wq8ISm-l?Wk!inF
zp&I5QL%_|$qpRl53OIM65mp4qW0Dxd_!>g`**82Kb0}Sy;&gMgcecjr7@$_oq%mcs
zvbxEaw6qWe6l^I^Pa$geEjsI&O1QUg--d)XG+4e>3s0>4{`u{;VSt9^?mH8qghI7w
z(-|0Mz+J;)v`V^fUo)ysnQBU}1O{5!<m)*Yrmw@5Sn}T%Wi&Y5!+X_8N~hi1-`hjv
zXOTh@s53B}phI4R><?3INWw2ruXz_R^{T|y^!+M{ByceOKqd1eDM+Hh9$OvuOb(R5
zGieLibtNy4%MNY7Vpv*Ry|2H4yRz0nFny@N4eYVx7pBF%xahAu-$7ZZSZO^B2~ntq
zZo}>V9LAOAj%9f>12xS1C_!oS_bYVN|LE<*OYmlA1^9e-G2FfWXO-t$SPPf4x9N^u
zwf;Fu@VTX!eVZfjkcoV}vFsENA>sHO8~TY2cG{G}q7R3EHorX_#Hg6pFD!!7+Ml-R
z<K8gQqMu};Aa>x*p*Mti-bqPKyl?~5`Kz}HcoY<2q7T=jSF1D+Az)$+_h)MWL*%>d
z_`~8Cd{I%+OoKD)#r~Z3QwT0TK0Ksvz~Qbt)!;{&wptw5=io=jbUE*c(-TRxdsg4<
zacMkVu7`?=>C~|6SEw9U@=SHt&(HT0FM<Jw6BQyZJSp_$?JB`%0@($DW_Vg==E9A;
z+;7{zFzVtEZg2f}dG7CVTa!cE{$3i|+b^7z<&UJnC9uIqQnTFDWkZ@_aJ-l=@6*WB
zGieQf+w_>LEE0*y;d*;lq8oXi)8eT~I`#^;rQ)!tAjHfPw0I%6)fV^ffX|+!`R0>l
z#D2$InL83;!oa|It<?d$zK&;GE_Rl_{8Doe>EZkKWdvt%NjD*wZEPweR+60hZrQIZ
zt3@7x^RL@ZDn1K04v0LL7n{~FydqsD$NHwWf5h%C%3sndFU9RRJl~E%_rm?mVkn6d
zLATb{nHBgFo4KFeQ)RlJ?r*QTqIhjA757s49eYg6Z$V6lh^6p=x8GK+=jG<sc^>@j
z4ugbiNMb_9xYuDVWAqFwR~xKb>sq{DHrO*kXfCpWhltq_BSCqR2Z!mi-tS7gQiQ6l
zX4ZuJn{@MzshAshy^k?evBt^ZBrrsS#$49HjgPBzT*$tVBKfx#pnXp4bFAkCHj&N^
zL2me)N=2_|ao?aw-ZiN;_Sr5;pR<pAEpIilqAO6bKMOe#u2DD78x$@2Y9|2jVOgcq
zbf`bT1n;``j8BW*zrl^cur^+ZQ9^<iha9<Hdr3x@axRt<bnTm>KN2LNa3fmCz9JJy
zDQ<4;#9N)J{>Fa8ww}ZMwDY7CN5pJ`kNyM_)(!6q2g#nbS6XN$AwevU%#n@IY{<$o
z9S;r&3hJ<|=`gFNV_-n2s;VN2TxfE83CYy1(2uy+q34pdjD_sx_yl-gV4hi<)JBAa
z`=LfLJHBLKbbj$vor<NnoDx&6zuez8A~Ik6&XnSZo?tkd6CG`KHXT$caH7gFr*1no
zMzC|XiU5e6vNA-&Ji#@8%c@%~Ek{KqnojA)wVn6dy(3HyaD=U*H4G?^#pi*D=avq*
zc$(7VnftSF&YLc)g6a@@E^>#&0{{R%5@tMT)Jl!p>m0PuxF0YA;SF30R<`_`kPUVg
zr5j1Cqzsz4s@TkMaujw3s=m=+RJ6|_;Tz^+&(s~zC(2TWxPR)<F+QEV&2GR3>?L?~
z1ai-VnR!Qx1sof6{>`ZsKAu`xgPF|hhVsM-Bspbem))7t($W~T<J{cu289x+fc!;<
zRNAj}p!aTUZkqi0ZVK^j@w(xGnSANPk`M#>uef<Kbq8~xV<fV*cto`KyloR}K68k2
zXw;+=ybxG7<;-F-5D}<k?Jr>6&691Jt4Pi@697qJ&|N1qb5pPA(DoYxBB*%+rfN+0
zr|pkx|4Ew^u*@JzK|o1|V)QULgq%?L5`XWo8SUnlf6QBn6?Q9%U^Dvsv0Ul3(VCgk
zcd0!66ps+|SrJ3-h#uX$A8Tu^ysB)`?|ueNlnYD6ZCYW!a#2^8l`CUiiZ@Usd)38I
za$Ixgk(D82TfSdT#Tb?r5>mKw@(pi15QF6Ck9o$*T7N9#H9lOMZS5)u+E;@H_kIkr
zk!{n{BzHHL9FBY4VI-VZQ*4xiE~Sb_^L~Etk6|N9_<|IY2{x<6Vv}2-)soY=60U^d
z`OajLsE>F5`DnUu*vIkw_7!_tZtfbzecv8WHaxckE+>t6)*I;$gmH6+7m+F19TGF5
z?kE5|JWsI*#%Vk5M-eOXThnO#;nn)huC#azhpMo71otcS`ya!HD?N+&-0Crq^>|Gw
zw%JxRdgI}7Ye<OIpX!^?sWIlp*#ef}w2!gI%VXZEbt!`L2_D`gM%He5u?Ej`fYQ2~
zFBf$`ruh0onqWI6MCyxcYiovpIEklDiHi#$(RYOtd495;mzQTW`{Sc>hSg*-(f#$P
zPXghq;~<matDTaHqrYJs#&@vu#l+f}zHuiiPzhGmuJH4&nsCuA!C6ah&cY<y=H_pv
zTJk~wBE$Sy761YdR`rJyU{2R3ri6q<&oI%@p7opjAqBV%<!s1ld_SJTj%Bu4Lh3^#
z><$V<l(xlXTn=mi3tQI9N@0!1vua0Y*yXNaU}JAI6T%@~+n2(KcQ&wm_jj_TK>^@8
z$SAsEcfTr70wDB4Rl6K6hm3omX*}tx*`B>vjS~F}GUOiUl2`E2(Glzb`>XcxH*(0z
z%1RQy<5M2n`4D#9pUdD7R^RBTtdtZYK0f|ffeOvp&Ll=?Y<3_>*{+uTaVm{Eksr$^
zvQgC1;^J`G+1b&lso~=O$XSxX*zv^l+Y?0uW~1p5(qY6#ok3W9KKDX@S64Z$XI@}Z
z3(b7YTKH%2lW+my6zu6G3aV>kfRjDC<ozk(@uo^s$^WMt0<GcGUAzgUT?$P|CMSW2
zh?zH|b`>c%UcKaWmaV>dLw&<Xsx>(OW|K_{JT;|jm^j%>IDjcZCH`L^LgH1}+3jX-
zR7zN6DJ20u9=DTzHG(fW^#O!7`{Yq{cGLS%?;JjgQ2`gR_A;0A5CXK&=sVvs6l-fm
zE-<A|liuy&;kgmq{g$55zpd8o)hTGjT%l2YElh7&SBK!c6WefL&X?f%v7vdr{D#(f
zOCwA;+X@MYY$vHJir~b6;#RCkfYof9=Q4`!vcB5+z2zTt`UYbqN_zT4g~=_hGV&!d
zDJhMY>WK$E$|?k%5|0eL@y+ZZAuv-^1i+qXm`hm{dd$VKYn9)ZGGxDf0-$q=twogt
zkTs1v;C|JWvPY#&Za&BLWs^iinzqaU8<`#t7u9q~ib2oZv{RTaoHL6zkCl=BL>&__
z&T0$Cr@<2}BTkObiu@@EWHjCk0O*F>*1gt(is@&%<B$7G@lI08_WL`)9=-UZ!>bkN
zp#CMKP`^*uT^z8W(eEda&W=B?j}!sb@Xk*Y&_i}gO#(c%v=~4IgELfx-|;J-h;vrc
zrwPfTGI!br%-2&)4DV)Ai2{|NSU`Y;!j45Qn(?s#eeqMbJ~@kEY)E`OEkV!2(P>6V
zp?7~XOakZ#cY12ZqNszs*U_a^adQJ)qTSB}QBg||pX1|wmH%fPX)&N8gh_+d8=6;6
zfd*%K`=R_*1rt3B`M`|)`2rUG3D=_-+O`QTGS8siHWy)_?h9^J*-km=@md_J6};)%
z4D0|IOf&93)4cguMGuf|tFq<<+?8oM0p3@zV+_SP>FGJ1Tv8zZ%6&%py4BD7{(bT$
z@YaCSrmImYFIdETY;H6l;RbUAux`h<RQUb1Pg@1maCEfN(vn?W-`m*mzLLZWxCGTm
zGCiJK5KS``>KXN+2QB+uI|KsyKqJo;3g4(kniLY!<ckaZHvBuTNu0-@R<nuOX+)6v
z@3!XlIQ;)KH@e41|ED$mGn)T=T0aD64gb4wbUi-p-|Z`jG0@Zgr}bfQLV(5oyR8L2
zDkc8YP9*;SWB)x>|A$G=oq(eB-FRKC1V@|Jw={}NY3y3_P*7r{Kpw+qOc$RZSb_Rp
z+uBasS_Fuvr>852wA6G!j%?Hvlq95y7y+VT1^$XlOdJHF>@-vxFdzZ>AIF@jAhh!M
z0l4S$3}9I-t)29O^FKoP(rL4;UhC+PlBnTL!e^0;Y5?m{i)cFg{aWXT1s=q3V<=fp
zNvUj;8a!3a<g|yv&JaXrh%fJwCxO?^#lQd2Q%U}B!@vXf`v1-}qQ{)|Dh`!FuiX64
z&YV(3eW)S0BqYIYZEfxckB{W$Nls4Y^SEGxWNMaZp<-i~|5smo{x@qy{~z{9|9@}G
z*;b3=Dxwj`tGx_NOa(2Ce}8U|7ns~$9oN~fNcanofUdYHOOlq@r-E%T!S~N@m=;nR
z&n&j|g66!V#18I-9rQbbkA2;I6;OrViw5>e1i_>G_)pW9D>}q}T!8cDtG{7bC-493
zKN=5@yD9tKiap=ae(jQRcPLTc$cT)!H5)P}sjpg&3@9OccX%8&2Q#I+!h>|e(ft;5
zw950^!Fs8MisbowF%1nu*x9!KuN+wh0P-Ymf~nn6!Neu;+EJsUqaSsX7#NP{E7dt|
z<Kq($aDMtL<+M9hHpGd7L)`_2t@l0c?x6872seA}8suPfAOTxz)zp-J_l_I@s$f97
zqzXc@+uPga<mKBz2@$N&=9&4kh1$VH4wIFRz(>pq2(qy<UEX<zL3WFU25tx~FE7dK
z1h#f?v4Q=;JpU6$#+69H12J{=m~lrHK)wNGMMvq|YGBA(E7yQ)U2Z&V=>$BU{l*Yk
zKww~&$K?Sjr?FH&_i$>7Kn2I6d9I9HGC}}H3y6!qfDY;g*>>hfazI!FgCrsgdbnuX
zL(sirT&nIpbUdkTU<8KvXgT+==~jA1#=O}(Cg4<izC3!plRA)<3f*6=%BiSuxb1<;
zD4)|7u2$K*K-+}|aZSyw2K^Nsy=hL53u`w0rn;e3Cgs$&F9b{=-<Nny21BBvI8K|2
zivdr(J6+KWdV>WxsPxybUrT~X%U4YHlaGC>$yctkZ4!E;J`1?X@VwWF32u8s$uF<&
zZw`;|E|&y_eYRW5-c}<Yw%phzpO)*^VuK!c0$hPhNf{nY<4uZ$gv2;{IA2d72tDWm
z<mE2__|Z3A;-j>uojrjIA%dl(rA4d{CYoHItS1RVpG!zemI!HYXS6R&CxVzmS8{N<
z4!-v_Awg>+iAvDLus4zlbiH?wqZN~eCWC8qx?3mjio8U-q6c(rP)OU;+WAHWnB~7L
zckR;yn7ioo^hjd%cgyvg>5`zv=%m-%?vHZZfx)4XG|yqP{X0*m$(7UC#N>D@CyM_0
zb8O-;r0eTzP?Ex0blap70(8X{IIzF6yQ{l$RlDep0iZ$`F!RV|z*z3Pl<GAS+qGOj
z-55y=`^q3o5BMfXbab>Y<LiXwFHGreAY*)CK0;YgvK1yRBlDP?$~ZW1`T6+)cXx7n
z3WgRmZQw)YxH-~&K5Y<{ltfz4+j=@8nDoL3{;6_mNL!m2At9mRNGiWhizmOn_q8b>
zCH8yhf^(99GYgn6quRMDP!=6Z6C`<QHq>UG;brD7y_jg={RJ$Tgc_mHNgu=3pK7Mt
z+go64kEL|5blO(V8G`>9rKEP+;b8Jhb9#1mEF&YMB)~Gv{<OiMLyV*mjZD;cM+5A0
zhgC&$5ZG7z&1N={nhKYnj+Wb#xJ=Q((+6)cx%WjM{oHdlctt5GYUle~w^CpU-`Q4c
zru8L(4|pQ)$2KUZpwLlL(Xw+g1MJrG?3XMmI(i3%UDIbEJZiVBkwAs>0oo)Q5z0p?
z06GT7@#(0jps?)2{oOH0$LXIuL6iVooYp+J{OeJV1Tj2ZY$l5Q`eb#tvKa07(`rwa
z-BJrG7FHnJAmiN*7;e5w1~x>e$~bUv6*!4-usJEj#l>+Y+D1olE*GFQ0l0jQ8<_MA
z499D+D)fLK&OMBMsw7cuHXNOj61KOur*Naq2B=%GE=Z6^(w6j<1@7(L9bJGg+aN%I
z{EvGiIvV%QCTrQhwzs4Pn90;s%F_4s-`A45x}^HPW6H-o1_QE}W@0f<tY?3q0hgv+
zp_^h7ZwzD+u3#Eo_YkJ<{cHXnQ;@H+fKJ1J*sn(Lb}TKmcARo@ay~!(a|r&}P;T<K
z>r3UkGUd6mg9Qpm<;U@-1*1!zr^CuLhhV^viP>RmV`F-VA&}xcpHobFk=0@HosEH?
zpA7QC?c_rN??t!hDOhR6^L1byaFE>m{D#uks(R<=Fnkak-Z{5$0Fe(GWNW&D29$(Y
zg=h!~12|1Uocd|IuyaP~e5~hwvtKJn^~N>(0c^2}x)(@~ac=^+a=6S7>T}7!^%wyn
zoQPV~^@tYoXtIwRCyiP+b@u{5ZsiQS-??ALdl3*2;2?p4fzNMDd-`LUsw^j2Zchd|
z`UVF47Hm1jE8Y|+*9%QeCfxmv5B&P|t8xa%bNBb2WyO>to}_J)ZJ(JYFLqyp)Lgsx
z)XQyeh7)4{F0nvIK>@u5EK|&BqBhA=-f#*p6jsLs`^OhxJVvoKwL}gBWc%}{Q)Jpj
z<_-=H5YpJ~^3Y$ztOke`No<44G!@D%HzymT8By1#o1li#Ho1SZPINh}Nwctj1NIUm
zQ?XqFydmcw&-7e#^M(sV2ncX>d*78lEw^0Vu1|JDi?aDzCoiSId9IMcPezX!s8Dg8
z#R34e?|*)BmVE2E_i|LoU&IUM+TzIl=$OE>IxTDd78RsRfOW$oB?X!X+5q3;Vp36W
zPcJ9|#BLt^)==TdrRgzz@ZT>^=Ape}-J2CS<$h=pd5`O`#$O5tVmkF?gR>PRp2VMh
zq+VAa(A^M2PywPo)K=PGY^H#`aNf~)1i5r{bc}(%p9k!6ci#a?AY5H}Ecmk2<V_13
z^(NuwZ)koUwL0J$i<j!({hs;9mmZk>YaD!+4}fl01#RWJw3Kb{Pj&Bhm0qI@66E%3
zH3}?WakFy5o2JvYw?1L(o0|ys&3WF(f4kEUb-10ke<ML2L3LscK1ejnohPqb6y@dN
z48t`;-;g9EBqjz|wBD2Zs=c3iGZd~lR9JCO03znyzh;m{NQhF{<07DQr42~T!oah)
z8nE@IKV2Oe37rnTd>(~`irS_JT|k07ZmBBZQewV9!7&(XKDU!6Emwar3Zw%{Sb=AU
zfWnMuU8!(%QW7d8N!WuYl9FFt{frkZq~pd=d$x3#&{W1Zbnt6b(0MqB>q$RT6%d=m
z`1yUyUjcDQr_vDaxjwi%b>Rs;eX%we1SSSXXWM6l*OB}#dpiAmOv1ucD#GW48k*9z
zwhOof1gljo-^fam)C~TK-LOvbM@#&qB6N%hu215m<aa>4T=MZr-ZVBgexxCQWNzZ4
z_beL0(`7*MKE@?|uXD}BY1frN5}~Oj5VuTzWlKMHgP2W6HD=le5I2k^OS83hQ6Tj&
z1{~B?C(XU;3MG!v5yWgHb-PMrasq6<Par7)ehI&#=_GWVr!yF*VpadqX|sVLJO*O0
z%{~J%fBBazP1#tB>0pBE#UFFwS(TS2MAbVvk5;DP^ABQ*Iz2t@1cDuZ@EN1WJa1@7
z>c8^5DwAI0{uoAx@AC38dM`Si{bBQ^aaRbQPQ3%;Pn)?QAm-}m?L`3|4P;^Bz<<Bq
zL2{J<Uw#~-=W#d56^MVI0BZzce!FfkU(xz-cL*FfQUZrTB=~MNNJzGR=V913Y(MSl
z>ES7a_~OFjh~B=C2`BvwQixUH)Y`#r`t{@E8>jtQA_kdoF$)U|FKFx|cm><_H`rzw
z54TK@91zG@ut9<rnE%mYx|}Kqiy{atvv`b9<DbPLo$gZD)}u?Fq{-4Y?*h9+0|Dfo
zk9pCe$Dyz-2gt(yX~m3~yN}$!KwqC2_$ScVf*Yp2puzuzr24G>NwTS-$A6`VQh0Gd
z*@-c52Y{uYc1w{3>F(%|7p4P*|KsPR72&`{Qzyp8`agS<U*{<TN{FMP4?_Q)gV+3n
zc)>+pkyM>POH<`_<MP4GjQ+(7-2VQ4QV~zS2A6$wIw>#c*MIk+<bQ*`Ngkj74fKFK
z=l?w@KB)r?DsX6kjyDFu`+OnooiqDgX?n+W_~Ua6ZJEy(;2vl1M5Lh{K*Yq`M&8B*
zZ!MJ-D;Y3&DGgr0JpJGH@N4e}b93;}Y?Izd#wh2#nX24nt7B2HXsnL&U|)c|Rz(E`
zT({s(__&Z@_Et7GgCC1jb^7}H)EmPo!5}rJ7Z4!#_V#|vR3veP0{~5^@n<QFgu@u*
zrmla#FpV`r?E->>J3#{R<;$1H>`SXcf1gXXnaO7gu!1Z7Oc_$ZapIDahTPxXtbzmv
zAXp;U4Zx(CfJa0{MFrpi|Nf%3m5!IUb|9lBzap{PyP%-J1Z-kCMa6)O3>tNH^+(E;
zo{6b9r-BUFAReA3P8bCELXa8|kwDq=KfVk6Vy&q;cnjUmuKbp7bTD6Ew_Dyg@Z%X9
z8{6s9>%`~X?6vcLO-)T)QHM*dG&h%ry^qKX1aiZ6KcpjYC~Qn6WjJVTOtnC?6F3x*
z%!v<u<Hi6EwhP!49*Aeh8Jg41k0uLUbTd!pc~pu66m$dzXb7qV(Wd{X%7L#@1EFBU
zJWa$)V9|Y-JvT3pPC#IGdlX^f_MBzu@X<q@ZT&Kx-2>UnBU*aR(HsDvj{w%J#wPb8
z%BRoXS-}Qi2H<^F;H=&nozRFS%l~~_ZDs`OVnk1?2X8h~>|VGYEu#|*9nAdrNPvt@
zx#hqlrt~OrA$t1oS{L|-$4v=#!C%;J1;@*4gV}n=%@<}vq=}oyXS>|yBdO@I4|(hB
z>qg6M|ErfPe`@NA!jIZ^TB)|S2m^?z!YC->g0ch>1VxQdRs{mVWJE<!77YQ}5)mCc
zL{t)G(Zq$SIMRd>3dk0yRM~?yEJ7gK5{QJw#E^twNP3?94?REKnM}^R%XgOVocmr*
zvfG-Qcb4ok-eSqDq}##wLG>Y1GwqOm8W#16*3|8V`pGUTT4qpawQay&^Gq$Ac8hib
zc>|1@Y<n&A=VDhQKUi7ixl<>8hOSd;gUzi+>%6=$5L8?5dFLXCwae+2=a-CcC8HMy
zf~(IU#@+<Z$VlI{>k+G%-cLh(l`+A=+bE`qd0|2FeYf&Wqn{mcU2Yzo!VZi4()FuV
zPrmV8p0_^4E-%D2>>=^l;|um?`uNS4PBM=qeRMi_`OOBG8g}jExHi{xqqaN!gm8~}
zk*b?B<RPnrKKK1bU=Z$E^T`2sJc*<E;7M+xo&=3GtWfb=e}Dfz{2-jg3AhmSq1i~M
z8yB_m1?}wU?R{f?fodQTF5p^|12}F4z`CUE1-6qZ*tlA;`#A-L-)va0>j)=j-vrci
zX;-l-+Hm2PsIxCthf8_2l1@kmopCsBY6BOX((jSgn+PALDw-#tGSA1wB1p?~2k3PM
zHHeslu{9QQ_(}Lvpl3ljzkZW%u>Q9sNLmQ2oSd92n9>mVa(<I$cy4a)fqmZ2+dU)M
z+R>8Y;`ZF{@x^xO5HotQ`-z3=5NUwrTst=2c{9%CwnN@}-{50lwE(1F@H#1b1Pp)>
zfGY{olA?HFD{Zhh)Ki0H&CSjA2Kb7hBOLZYR%n`tp&KvK0SBy1S517kV#UM1Pd+Fs
z+lbl#Uax4~Bq=eSrA?fwsdQ8U`#UlAGQaL)=jKo|-;9imvu!qMQpiopuVw<dkgZ|d
zH)7DG&R4HMT3Dz;l(|qf)#8~(<x6V<3z$`$1gB?*ns#k%RQRw$KaCibZ5X32(kdVB
z8cmegbwG8%pqi&zVlcfBQ90Rwr^myfojkyg_K=-{3KvyPzc7?Gu#s6*p;RhK)Sg{b
z4KLzp>84x$0RckmWe5_HwH!9v(cv0?;eBJa*Wz4gwbvpPOt=PvVNR_+d{-io%xN@_
z9J_8DF9Ab{pN0yI>`FSEI9Tmwcw~FNtE5;U5X5H@3;WkDLK&f*%vd};=Fg^04fKdz
z8@4GR`RmOj1)J5))QDM};ndig7wzFJIUz?kLK}?k*O?&anPsx0WSB{J>L_gx{*~!n
zH=0-m*}mH>iK&I28gDj({3}4F($OMIUq_<r##NdgiB;^$lkdr6FvATU#H}BA9^^7E
z{a*^c=Y%ui^<;{979K(F(eHp_3V~R2dZ;m*=421>oHE8K8QVEJdafyM)}L98Fa(3J
z8+rx`z2pwaDt!A5Kuw&TovYsU!kwEZPeme8zr+BF8`3@q0sGhCDnkmE>mL{xJ_c9T
zXN;dbr$WL72y)(!{RL=Y%DWTdoKTNhu7UR*gdRfJIcKmlyL45Lt1}jR3tCQ&WPxY?
zzA;SNGmKj@<<B!yoKrUP&cUOR@$Soz^NFRPXZK7bFRUN0dhv%v>CkBuRRk8~)W;!s
z%2vl}j?9;7%kOsm(|WlI91kPEI)cjE6Rmq94iUojBlwaSsEBuUbBn01O%NBprGMhp
zyb;$XRSetO+FDv$cL7S+UPgV@f<QDWomY<OAi7k-MT07HWrbit{sadTg@`SK9Ff2R
zb_y*|$BHIRB!srtSJJ?Xec?#h>wPF&Xd!c&AWa4B7szY+s9z(<sE-Nk#jskdm1)FD
zfvjn9?bh7fT(G_4M1Yg0QLmEOWS%?c2K&i3z<Cx`WvJy~!e+h&rS`8yA~x7K)Xz5b
z;<O~XUfIU@%f4#|<j#j!Qs~KAT3Qmy`ws@2#qG4Rw6f}C2jWYbjRudHm>7!$KJBqB
znHDAgyovW=2U<$csr$ow7x<q+RVJDw1x>>js=>`2qfp*8#!QdYRYhLB<5xA*EtMVv
z`^gkWFx>~SeFYWuThtsAWVdS&sG$VIs|Pq8R1$Z%xD-YuHSH>jP9^e`;3<plG7RpR
zEQPmA!tjmTiS-WPru@qLI~9uLfeP~78F8%)XaLx%*QqQrnY?nXC(sLx6yAV10Lvg2
z!g@M{Tn`l#uB3%giNqhNyw*z-<%vZ;LIimi65b6$nr6ZzK;+}!E)CxMpymtm>XO;g
z;GQW8VUXPxAxnevP7js_Nb~_X5>)E1qX-2_S%=~S<o{8$fC&<L;ry)I6zDSvz5#pr
IK9mdp0gp0`bN~PV

literal 228234
zcmdSAbyQT*-#0ozNJ~nCq=105GzdzAbcb}OFf@Z0fC@-=3P^X?V35+CgVH_Z&~uN!
zXFbpR{(tYf@4aU&7#4H(oHP6E{oSA1Q7<)>i12CgK_C#3in5{(2!sRt32nId@8#PV
z-@i9-tkskhfmh(TRA>tWUhq7XjeS5Mh7bRKz_IH`b|4TlNJUXz&p&f-3FdFMo+OJp
z{1linwZscEJ?QDdUR?&+Fo&PEZsT1Ka<paGK6E1=7qzY1es?OzoQM5Hw9%`#qhCMS
zZ1UsOu5!WHB9cidqG%+fScNu3cBBCJv89?}L7L>D^k(}ewXnv7oZ|mFNPj(@<^13O
z(cOy$BK@C38_OC3kNltG5kvs|znx4Go(f`S`JW?NP91drf0~j$XXXI?Ps_+imiy$O
z|8-qO(nvhe|FmSuzxVV1#%0GMLLd|^s%mQ5Mn+F(i{%w5L_<PCs!b|D@U1OJYQp#O
zGOWP0Xx+;Vc&@IlK%Ua_imW3WLm8w@OtQfF6xykv+&FRA0e!ar->7aGqW``zS{#E)
zvwkd)b{z1%A2x+yLVx!35T85UCw*t8q=Xw77zhHDl$1n9MbWGY0gdwi6xG#{FD)&B
z;E9P8%W9-!k<~yWsk@8Sh!9#u(kMI-eoJ%nQ$CX@N5!rE{dYx9ptbq`>hY3VqsJai
z6l)z6q@O9;r%B%te*lsHpXkm1Ll5!)_m_=k+#_es&)YVT_C(Qoo`{Yv5!kXpPkNv(
z4iLPWh!DU38vP2ETKt|Xv@eZ_h3lbS0&=0Ep6J(B&IeXG45Xx=v4C#)0xiRlf<b0|
z=8jH?805TotFOsWX}*Zl_HOa6KLgH5y`$FHf(f>NWObZC!%5%|AH&CS5aYPE#UKJs
z$gTr4O8p5b>Dc{gd&qoSc%nbE$gAoRj^k8b#2w4ZDQ|8i<rCWQBN{|>2_j1uQ~<mg
zr`(lsh8Q`A@S4ZuR5|0sP(WSmA%~{KiZx<TQqp(-rtaFxAt(EV#}iQc#ueI)V<ij;
zVr*lg7URCz6kB}#8d{YJjadRo1g-hJV14-n$KO-n373oK2>)VhP&kqvG@K$C+F7R*
z5B9JCGcJMD`)Lv<3gjn8!DdxZb$cs4vL`meaZlcm+bIgyJJNc7#hUdc{pLh-akKT>
zYS3(@(y7)DziHk8?@2nZoW0-^R+31D9+>|WFGA}a)U-e^8`h#KLrekASzzveke^rC
zNa^p*X|;0}{5zb5E9P=^pgM~R=w7)W{sGF0-3HT-DDSf0U__;<R`@-@mYpQjs?UuK
zh?-4^%*wzLfnT{j&Z(_TmRB+zGzd^@e7PkP+fo&XU#ctKBB(z}2=fWtKWjMQIOy)M
zAfPLh+Td00xX0t)eeH!)Y3IYc7@>T6PeWxS@MoWxlGzDpNBFoaD#G`cucu^IT+?=u
zQ}v**r;jV7Wg3VhiFGMG?1WG=1+|hB7Quv`Bf%@#0Y9pAI)|`sJd0fl>pS8lRPG0F
zFQcTV74HJCQqe12L)|QMe*`PK7+#uZl&9=2*b)f~5Ewm?w?G18?9K`(*~@;_%6GfG
z7aqDgoMIg|;|f8Vfk|QzErvuo>!7O41n1+}3LJ0UA21o93wqnkBQWRXw5SiPa+*7R
zjwP=ve+QwEJ(C1dL6D1^qR||3qRLYHm1{}1SDU}bIg=X2cEkBy!PaS5%=z-iu?Yu2
z^Wo%vuR~+1s}ru01lxevUuTD{yois*;p2fTLc1Rf$yN%5mLo8tV(!on(Qw~7g)mLU
zJLj!9bDltk#}E#C)$tYO8j9grwU~t6!Q&EnX$_!YKLpm7{HCQ+L)p+rY#>!Sw#co6
z7u0cdzc?+vG(P{M3fXIeOg3FVRZjSIE2{~PbtC9CCaP4JIb{mUgE)$Sogzj}=U%;#
zm$o{{(XO_FRM&u%qX{tL&sjcFUR-AUyX|O~@J%hqp*dsr2@6Xttc_?k=sB{+!{Upv
zsQ5!Om(<_h`=|9AHP4AglEH0!eeUo-xlLO0>^r{AsdymTdA-9j_sPNGKk*W1$rQ;z
zIFb=dVyISuEH33b&2&7wyk6cyNt_FjABDJr3{L?o1bFeOilWDfztFaPQw(oK-n0XD
zXa^GA0z2&^1~X*Pz4^PQaofAxOeF00_Q33YbnW|5MkYdODw@gjx{mSTyeYQs<l`4}
z8plbS<_o-WVVcW{%gr)0@Q~-@v~MCGEe<DdFVNui4R;C^K9pwI>ng<rHW$2vy!j!o
z@$t97Xj!|82{##$zH~HN#m$`T23HDMMGcY)GPB?>vRLW$0CmlBz7ul0b(yYxu~Jxv
zuo)SY#sO9#1h72M^{W%_ALV&gs$bdec&&=QBr95R`<#s^VlWRMeKlN8{q`Pg^cfVj
z06oq;{#rnHfn17#GWW~AlbDwWgM{&p&2(m8Xem-Wo{F9cnnH)^zoYDotr0jQ#-jKj
zF@+A+NXW8PR7lzh(TR<jf#awMA9OzQQQ<QPlxdv8y4fO{37=O~6c+pkK|C^K<uNs0
zV~FyxJjn+gm3kk}{lka~lG|P-WRcCULiLRf+IoeUDa}XdNp3F{C8o^zSzo0Pk`hUm
z9&YA^Vqzy<9FqSAg*rp-585{hzttGjcj+#uEm(Jb2TZTe{f~Vb0O;~xqfeTlPk9`+
zm)s&}rtUbRL^9RL{KuJ9&`-}XmMMlxb27{os7ErauQ&e=f3d#pwWx%ocM$q>AoZ%4
zjuch`6#%I5GDz02!>@-ZfjPhg@<iMebHwd}pWl6-0ei7Y4q1a2-Ek$^cE)6CG%Ome
zsQ3>G%|(62k!k84WU22ece}%~${5CW`&L41t3f;gnz7S8W0Ld{j?1QQ^`YSw{f(pZ
zw}Z*1EHl_s!0O(KXtd504ff|e@z^ic(a&;mUlI_Kl41`sT*@U5OBOj6&@-u)&V26|
z2TpkD?^)Bi`Zqs3X!G5<S>&5nC-lgfx6RZmd6JWzl>_QRL~_p)=4Uur^80=OHzl5$
zh#h<unrl_Qi~RK*O2P+WG$*0{&^KjMWKEY(^UB$@KJa}fH^Pw>)i*wxGx9Qyf)kI!
zBWaNq{uvwL{SHgD`rGiMmox^-l?Um3Go;c1*+(?Mf;YC(rwln5qj6o9(86e{hE=3-
zVWxccA5#<95`_n71X6y&X0-Twn+kkxf5Q%ML?li`Hz?S-LaJ(s(rAdDNUD>_NQ5Zz
z^EmREdV}(N7XzSlaAZ7-dX#xQ4(E3V)6Uy@x;J6>sY8l3i8?~dKAr$`e%GZ}(2wGY
zi~80G9>_W$KH?Go$21$Kr{@JY$B0KI<%_imAF5#=fsAQ8=NnQ_-sxJ`#dUc9%rBbo
zQeQm#CM5yi)$(5Z>ldNO?9#dX&xj8P_?VdcEG!xOAK}S-E7d6mUS&mtUU{PIxJF_G
zv#_b_2BC*k{>bCe9Ia}%EZk3;xOZLyzPG~CB>w0uR+lBkA30;pBfCV{z>+^Y`tZ4a
zLY*D2ZCxNo^=54>|1>6lL=$I+;{COK@s-Sb08lU7Cs-#Pdl{c3c*;|5T>y`*!U=62
zyhfGJI*Jcd_b^-A3!)mX!^QZ(N~&ym5M}I}?@@?bg78JsX66w!=$%^~mhJ#uAvc#K
zWon#2v;6Wq@^I6$s~CnPEOlEDixrwTZf2ina}g7wlc_gcbCnZ`lQ0Gf_D`yp3tnaM
zycNxLI?fgm!JU;dTvqsgMs=QT2O>mt@*H47^~g{yXUCzE^t`FevA(`SFBiCZQ+w<P
ztCk2AhpqD$533skJOw-(Xur9hA3qniT6oqn?QbF}KB&S#Q`Zw2*(V8b2glhAaoUT&
z_A4a{2QbLmSJnGo&+IQoc==vHZWpK^OZK@<B+4-)9xT@1RJG_$>PV(XAj|ELdcRAV
z^Soa-3^);B9f2Sv8PD8%%@(++f$ISRu*~|_msuwOL%_!YedHKW@?aXBW1T5Z{f5d>
zOC<Hjcw4ROt*nTsSfGvA<YdAip6E}1gl=ZYp=^%*2=>nuM*@hjWdbSR`^_)^T%iTD
z<Jo4V9&GJ=F_&BRtE452{}$|dw6a&u&XS*Jm&#8R75TGV00Cpd8D4}$JFoRfiO9=z
zIp*JiL-VZRuDx2+ixbTJWY*zIVj!&ph`RkquUjfnc`8-s2`yev3~hD=cci0jbQCQu
z-rkMtC3AjX96cwmqS_AJ$4)M!&CG+t9s<*(X+B{4)z6p1bcoX$v=qjx0d4$^eUM4t
zc>+w_ypI6Sil(m79+wS$g43sCo0>@!93qL@!^^E+tZDLo5wE9Rk=mom{g3nDe|Bm)
z{xftcwNi#2OrHv<G%yfE#YnO_0i+M>$v^71#>`21G4olj4DXRkoB@e3SA=ZAe+Xiq
z{p;3Mvx;RIW=U-7oPKLc78dXSaJnyaPI!p+qlgfy|KQ(jqJ|a|pp2gqdK<d_H_!Ta
zY3%>E*y8^oqVZoO6=0yC+_tvseGDf{v@O6=JuqE!u6`Fnz#997k?Q06K1M8Ks6B-x
zpER$j=>aI$X`VR;PkZeZ1pcgt>SAYa@9mr`<67NA7JUAMe&N~Te{}qZ84QND%7Kt-
z8_ib2fbWORL=K`56c7*q!GHf&|HoO6=K#)Xp`_HiKrmBZU*A*}@$YE_`uh2Gnk&bp
zrw;^ivCxw0=l~kV3)Z$Ia%?(F{r^0#|N8a+BwPPK)c^mVxNPoLw+tAN3BG<Mo((d-
zTd>0!Cl7y9gV+!UfcHNJh;8b_=~+SLc!;Az4o@oxuYcPu!vid5r4m8#eA)nvawK-q
z2Bd}c_2+Nk=9k3u?Kq3&fDRG<+t%t1%UH7b_%<;z^3mgey3Dxz1l&Oj6vQVPdgqrB
zOy7wU?8)vq^BRrmY#?yl0$CS?k2%S`*4r9Mi@ce!hh)p(WVWtf1VG{Iy5Z=@(D=XK
z!z>!D2ELHxwv$$$g6B#Qe^n4WG~S;d(w1D6%enDb2|{W5<z9?i<!EnFOo2XBf}3{2
z-3j8RnDoqL)!qzKgm#1=K03g~_@Pu~tH>cn*4>|9^d4IuKto}pXCd{i&t-_sumKX3
zD(UVy92*(w$o9vLh4c$Cf0i%N-9nvFIFbn`el;AM97iUshRpx6d@*<7=W{jCQTH!7
z)!?mR;KnBw@q#x+a%-SZsM)}d1>mQF49#A27gcV5LWt0(kS_6inLIucNWR==s#zaA
z|JSSlR14lz6sAZG3^yE6k!v56LS5Drz^@-b)Q)iv7vBt7W8d9Sd48QeI;k-CHlGio
zx!4P|GC5euL@L`IK(iOQBo&uhbck-yvf=zolHQeLVN#E~!=P&&6KN$6F19UJx(=Wm
zcJHqD00jAhY8ZGD$9a`R{vG$daaVs>#;88$1GW1xULeKjBQT&*8!*6Hk-^5?l>I_5
zu}{LmX}=K%YKT}E6dCy3f~o3(YEV`l0IGMP09XfFobSxTkqr1V#{B^|-d!#zq=f?R
ze;BGz@aA(0=OPO{XN+yEnONv44$35NZyfv&i!4-u+V~`JFyP!$nu~Rrm9*6_C!xm=
z#QsBH13s0e|A1*?NresgppZp_gg35-2nnyJ+99`Q6*w_S)U5|^2*NV<#qe$d@QA+s
zB?g2$rWgy{a*0FrbjiYj!A7D*n>lheWpsixLW~K^dX{X<jc6Y+{1NdEKj72fg7AKE
z#Jn-$dsJlrHIba|2dpm%tA$PFF=j;X<^WFJgbR;hoEhmdxEH*|w=&!x;q6RZIi)Pp
z!T5k1KK>r1$Cu5QbhxROfp^i%wUXfSEaZh^9;}e|;w*HA?N6bqRd2uwlm~wMTu$R7
zx~dLcTaPrA<axbGOZ=+=@ru7YEF;@OCGe(q8sM@0zA&V|zxm1BWD~U49aJV&|B0Sk
z7*y;5#{CGVJA==$@HT8{ff|RC8j{77R8PaIGR}@^850;sxpc8yw&P>p8ZmP|_j4d4
z{X+Y1!3uul@=~(1p3#3d5W^I_cyGE<bs;8b<J#AmbaPFkr1YSnEJo-Il2W#2bHui?
z>H?pAR1zuUuTd)F(L_X_lSJs6$r=*(d2v&e4VXcgrUf6CiZ;VH!DBBj8AytrG06J`
zJ>dSH0DAgb1L=hmQEzsEz-Rsvf^eedDM>JY7Gi)7L)w6|b!Ok4{H3v4e;t6Kt~|BC
z1(hH=mqcp=he?uorr=XQPA0Ryd(U9Oi@<iD$?l0!!?nF$*#0EZlvjv-zML@fPn)TM
z&cdGi<pu{gpwYWJzZ!guq2!=7<%q=x?S9K%O1jRFYImw?yJb{oHaY1#x_@G>%qh3~
zl>1&HK7WVBQ(m!<Lvx(3?4==9$~t<`n+-U6iONAWe3liVWT0Y-3)_E5D*^9fNEtXD
zaa$Scw))TX9&h@%p``?$vk8BjI+~NhslnjfE5^-u$iHb|o$Z(&tS1Z>I(`<UshBrD
z6Sjn+xIrs9#O|SJhM`IxB~)y}Pr&zyOAn#LZu5M4eXWI+Lw9v}eJWAE8pX4W$a;`@
zI05_OxS0Jk-tqVYHMD=WNM5U-fpj)Y22Pr<cXi*{;k>$_<$ka@>KCt^j@hhMY^0;n
zKgs)9WkXq4k{V%LhQv*2t&a}848OPRE|~VKb<<bki-8b1zv%g{SgdpoX{$4=s91B{
zK;Gq`Jm@&f-!vP<heP~fabjNvX+(G#qL6!zBa50xY{SI?9N-iIvv|P6##NxCMB=%_
z0ia7nS5?V_Z-*pBYef(*i3I$k=ziR()!Qi>ZK+Pf+n~wDf#CRf3B2SYPKZwr#Kj)m
z-`-vHq*Z61o{%VJw<%nF^}R>N)2W#T<D*tmy!1ABkzejbW4(H*3wHh|L4VjUy<BFj
zE{xo=*tx}*hG@yAg#y&IP015u-q|NgGR5F#Gp>Cuo^;!##0>n!FC>pM=r3rmog2zk
zLTXIHi(|o-VTjw+v;-DjhBMkI7t7<l&maHECaQE_J-1FWJ{*c<MLx>>N0j5l4RLNx
z=wB{BMYVi=YIR1qvW&f>-?_h}v#-6xLeL?wjd&Ba6J9dRcqySj<jpF=!H_Cz(U{m;
z=b2xhKE^t5Z_NRMJEe&S$8rn1;{d1^-0N+Hts31Lx^!k+$GpPN>w>mAfcS~T_ogJv
zZ*dvSo8Z(!B1DEE{KEk-Vd*mI*en%Dv&_2~j3BwC#(q;~-{>9t@4F?=94Ww*bAM*^
zi(du0E-*0>&v7{`{HUXN!$k`>f3W6Nr}cfwJpnXd`4Pc6Lwq>kL@V*wU`3VEbG5M^
zGoTFF_zY{;v@~pY-#Xh7Uvb2%4&bYH-$3yw_o-^r9q*TE1%0hPD+^oC-2Bt4KO~sw
zH>S8^Qx=iYo|U<eJOTu-u#@{YP3XOTNA;lu+dEy{@H$QG`y#xxl}jPOilkuG1@nLw
zWdZVBl199lSEFM-+wrM#47A9rm-b93ZM~t~Aup^(C*Gj0_wDNM$jJ9%K;?O~B@Dtm
z<&tN0$g*EKBS5rxuOOQs1+VMd$AqnVu~!;=;B!tNJr@O5?%s{#jvIRFMl8~I1Wz|a
z^NFg)lc9%T439-%?Rq-9c((43Kt-|VA(@=(pW~m>?8$8XF__-ErgjM25L|czS)eKn
z3(-%O$sMwVYYQ>YfAevEeV<|mB0IO+f~nSrX3Mh6qcnWJ`<ZwMaT%1S|L%<SD(;}W
z*cDhg)tK3Q|BIXQ^8;@7r+@cfc&T(sS!79P$JVH*VNq$K4ukcFazK-<Ek&&f38l%W
z^FAs#xIyD~ICmZ-s2tPqZNl-Fph`VxnABernwv$Hfc+59LJItgAcoO`8U^l3egox<
z=G0e{9PZs%w+aIZf2Eonj$TZx1=8AKk*&TDwX48c*)F(M@b`UcigRPlgu2w|=4nh#
zBvSu9*_13miyOa)k=hL6Eszi8#7J@S*6`^AVpBWmMNSsN@DSa_s|v^;N9&&t$U$ZH
zAP=+g88V-17UP0X!nC$S7=tLOn2wO+yADT|YQLhptdD_4<?)N+0d--Htzc>~d?Qh8
zBN;nKb_ITY6>p;|U9q%(s4N2DWB9GCzsPRHsMm^q$zyU{OyD_(*o+Gsw4N}Tn-L{D
zx^Hbi0w0q#JMUmEH!WTVNW-3yfet$3T1fo2KVh-Uv=1EJhr+e0pq*Vmg4Eu$tioZ7
z55)v7=dObMBqc2wY4AWqI$p2>%{K@dm0CHi7?|`l*-x2~;o(G;H;?q>vk52>!vu0(
z47pA!Pnhw@1s~4}&cF+G73idDOxRO5d@gdq{&)5D_xhCa|KPQJ$$L8UOiO2Uu22<k
z>!S75-geUNaG2~`%QHa%gAK2n>0VJrD4e7M%lb!s{JbFUQ@u6mJZqlhAkQy%o94R3
zQewfB^K;oQ$WTY-m!nHO;H4|SfDnxV$Z=h}<%H9kgeCEfYt^QKN3naqW;Ki-n{T_E
zp7`f4)$O*s?xZ}7MR%=zQMrvtYcNxr8~A<&s&i%*+#YCQzduWsOQSL_>M+=zC#e^F
zXm<;&^=e7AbXLFgdPht}=0dEICJq~p$e94B?fylUVF<c5%qkBnAn5TG&l)(ObB9Ja
zxJ{$97+I%|nQ1JZ|A_qi=OXpd%GhTF2~2gwPs>09cY4U+I{Qm*+k=ZUFwH%;sv)+K
z?v9l3u%&zIeS|kbBCkQRORMRNPWuo(!~WgPF{1Hve)@8Qx&3XF1f9^6x-x@TN>4zO
zkM?5NSZMqIRm$AjQx!Tqga!)PKNRS+Udx6&5pYnuj=&tm<M{ys2F0hOou@E9gL7OA
zxJtY7$FFgw*~?6TDD)nt<m@X)trBVJ&{zBNRoflNS?zTU!VmXR9`HQXk;%N$Qseh9
z@x$~;LCCH>_|WVxbsTysnf?CyX9UbL!C6+zE&C?VX&f^xj>-i;a*6J#B7r<8gTr{S
z%RxGSBarhY&6*CT)Ee1!N^0<HDl9a8aGDUY|91KO5w|LxE$-W5=C?w+Rah<UL+_jO
zSOXk{Ax`qaGMDkm_kx|XuwF`II|_s72;h>%TbNcQwNqKYh6+T3iB*r|Tz^BssIS=6
ztTqDIpieVghG%q)SO3$W^dvb2U8N?rso<f25@($3U{<M3+b4Zj))SfCSqb?pGA|&$
zvWwf&G|jHjW(23LX6(H8vJ$>&WFcJMkv+zIPtHR>Jv$w!)C{A4<gE@eD^En9E@-}E
zRPoZu>kMr~|J?bh9`GsjsAw<O)BXu);lY5ru3Y@nK**IScP}F!4x`16rGAKe>8{Pn
z=cb1f0VN2`0^4&k2UUidqXM#ZFLM!_twX`F!A0EHXUb}<9{pk((9jmGY@cjlUx@4?
zc5q0L6>pZh{Yrn<GC`{Q6cls)p2F^3>$gA8-n9>B!rrs}QaSoQKL?jT3fOgWbKAj=
zu}m@7&9XPHqx_U)o%*GD^=;%oFC4$rC56hS;J>o~4a+<snOUvt(qPmT_GML-EreN&
zI-3k6$-gCA%<%4Dgo{g)8ynd`(LBZMCNNIEUc^7(^iZ4~_hQlnK-6XnRF`DGdAt7P
zI@RuXlp4R0G$8Une|!DvU5J6zh#lVi-Su=h8{*U0@x_i)+uzu6oD~NTh-IfC?>*!K
zr^=@*KW07?Kfl>{N=f_>Lk7kUJ|;MpJC#PFpJKPF|6+56e|t`w4&+M)M?VP+OzMzO
zJ9+inB~CG68PqtmZK~68SgCVsAYFYJeHeWq!-GmGKo~ZC;P)4R(W5LEzJ^5kc=Jk=
z`a23#d$-Qd=!Ix5w(|;UYc9WCU}^_A%H8g9Te&N(vcV^nlvm$f?2mhb*d$(3%s0y1
zRafu~CqmG}Oq06W@j1LRSq?F6jS(u*35}6-_PfX^-3gYc=N=I~bP{ZG7elxo3<M9o
z8!#hf0FI{zIHJ{E?~jA3%PeW{p$a$fvTw1T!%UM}?rUp~xD**ruFNmE!H1+OkIhRH
zncbf~ej=h&8FF4R_l(AI{~!rj);2BG`8az``iUDq`*O(P9Z+mS`70TeuQd7+vBE-=
zf7hnQk<zsN67IYHldV=8I?60Nb842>A?(q<+a>O+hO}86<88~7ZJ5TKENFl8$PIi>
zUVU9#|BNU}VaAp-Pcln-yr-s-GF94Of44O#JBJ5)=SjWg*;?garsx+O=3HT>78Ut|
z3;+Q4Dma{Ywe@&D4ozY?O?jgU&lvN$6}Z8N3sM^P4Wbe=yK6Y2!`lPX3!G7S9RqZB
z-CIu0o@Sq1@z8chY*sf4I@|eZdwgmOg7HCpvbW!~7b{J5I76o0Oebv$z6u_st#)Ge
zQ(w=qk1Z5?5W|PgBwAEViwY!6CceV<T{g{ecQr?#j7E?3*SK4Q*l9dl_1vYT*vs^q
z)oQ;ndNk~!GuIrt95kTwt?VuF^s*?C-yw*Ewo`13F4>xwHr_owQ5A0`{S@Qj{-v1v
z3m_}$k&1Dg5$JsdJ8KS9()HJ8b>-UeOyVWpODKmADk;|E`_2`Fp4cl#^5eoYE->GZ
zSYw(?iH$gS=fy>NB|~i8S%5y`R``>ooHc%R$|UP;#!jo<qdqlgXSMPCd7^0FwVFXR
z1=BZ<S&OIp(tt6FByP8O^kecvv|K36fCsg22rr+!+j-4zY#AX>b_diy_u^h>)tJ4&
z+6HHp$S(|erW=8nyX`0)Y-D*sR)VZv;I{=+0gY+dz-eC+KEL2jr)rm`?XT_HGE`a+
znHCPgH<%CuK)Xgn(Z%G<2)7(I8}RfuqaVY}Oc`2>4tMDYb*e{OzIUP3H4@AR0~tCw
z`bS^hIG3gu7M4}*M0Xg%Z|IG4ws{}bRqTrNh3vh<UqOG63d^|iTfwbTlYjnn#zRpt
z;xis7pY&kXcSXwsY+p>5KB%uFa$Gaz9Im>l>TlUjl;x~Hs{-nKR<(rnv+!wXXMrh_
z7y_8qfayUgUP87PxQ88LT3YZ4G2k?wQdchz`DWaY_*uTv6nj)O!ae1sku{DMlU+DF
zLKz4Y_jZ=)5P4j;?h?QU><`6Wr1jTk-AwYTFn^N>{Q-l>O~3)CUvu#_1`ROpSicg(
ztc6tG@v6Hcw{W!I(91<CcO<4OkHNpN3t3n3##=9c@y26n?y#kV-2W_-O~kcLOo!Yy
zcd_H7P$XrUZJAZAy@|}THwSR!if0mbX4>^-pNhqam+Zu*)YaJ?OBj0Tdd)=D#Y=1l
z-FRfM7VrO1hW#zl^OfMFuF@Nn_*(!g>&p3{`xI#!!7C(=I%nSGIc4hFWvLE07*$!p
zzG`cmVrJ@`osJvs4wLiyD}7X2euhuMokA3cH77f=Eh!_~8m3e-{5+*=<fCIjecDw{
zmYh%V=IZnij={+5o5gH3vK#bt<B2*Ks);;a81KQgh}$#5C*ljYQr!AsvfH#bO%8CZ
z`iTU1&<0Vk#i4}4#M+Gc*w-plpRsQ7eC6Ji9~b_;U%Np{d3T})Rlw4;i)1|Qpz~Q5
zW|q{)>1XPy#B>z(LT(P?>W!k`+EM@#F(6}pwkmcHV0t}ghZEZR;NmY;$=Za?-lGN|
z>dkfkFbGN`OU^#p#)@k!wu=1sFONj+`Oi`<gFYCW1qjRr3<M2yI;{DD+RIGXp;M8*
z*uMYhwQH}<6ObC`;2SAvZOB<z;9$_NOnzYSI-kk4u>;U<30r+IiO%}k_@00^S!$?b
zjb5^Q7N<(QbjnnxgMQti&AC}JJh_Q!2)WXA+lA@n3UI@=K4=c7dmyb+d3oF4{_$49
zCkyH~WPaQxn^%DJ(f(Zj*>m~fn0@?CH@VoEe$sqqGnbe*w_aywN4+9*|Bs<AAHm57
zQ&g4;zW^L~Ss^CwycH&YVXE(ZsKf4=7rb(*{9|TLaAoFN8q-l?7?{7%Btzre^LYh-
zbl#w0`IxG#e(GZJ{b*J~be>Mf!UCuCQrC5Lx%TP2pdr~-=QPLa3n%KFR|o<AZu{O<
z2J*Yo<8ihpSKpmhW{G-5v@}c0l@2riHZ$(;ZRem@ZR{U_=$pAua`g^r^a9K$HlVSK
zT<=KkzPTl>bc)fg2?VZZ5=u5oizUv`gjV|AT*5YFnBm1SO5BopW$AWJ`#;p$A6!4q
zyjRWx>b%PjlXL>=aq={wm|HpE<B<Yd3aeg<aiGI>zLR$ydI=WD<Y*_vuH;U9N_^k+
z5Bo|AJ-<9O>0A)G@Q|QIDOuHH_OFTYDQ8bB4!3l_2GO18h_MY||AFs3v7Q9o%UzOI
z2y|)Nbyyf@r%VPYGOj@@2iVz^Jx_NDWAs@CnyhX>H|x*<kwtQb2xU2?W4-(64Ch~B
z)kkv6L$%T+=rOKvtZ+m1g!7lc^!M-&es9xd2K^fH(v_lG;Z<iu6*!S<SI&hg3?{ne
zTd4YGJ)DqZSRLmeU39duC!?mob~gkl&!HVl)K@!bGbHiN>sa)|MBZpIn_nVOvB5yQ
zYvVu7!O^?unp3dDz@nl5`@i%!2`yD&=$P*bW)o`?ej>^hW-&-Bh*k6M0ruTmUQV4f
zkzB?1mt$5agK$@V300hDSNY&ak5l~Uecj?`nQ|T~*ec95YDKU)+vHEejCO9XN{5qG
zo*%!w*EvGXFfS^>-TGSxTs6+Z#rGn9Uv^<PplPxBHNl;4#Cr=DLd~Uk-)R1Y2ivl^
zT7d2zTMk)nQx8Q-Q>k#zj!0F*I?cjNpP3?Wa3A1?W@lfRO)T$OPAu;tJ$H4*&dhCu
z`vP}eX{rO?2lMefuL#lii$1hj>{pIA7wnww6dilPEOJkhH8YKON2oJhqHxxs$5Ca)
ztTM6TV>h$p)9DeMjZfH}N8VOdn7{utMaMNDz~lH<!dDk#U0wrie#r$@4-ueUv;WiE
z7Q)o#n2(j%K4+B6ul?%Kq~g@~Obp&ugBS_hdSUb7vDrv6k?RkQC|v9O^$j)vAubzg
zrqIb*{@5d?P$BVLXmkOBFV}c0LuHj>Ec`i1yz_^xx|UiihQn=L+O;$D?A@#Hi4S|X
zDG&38K7r$Jj`T%ZZEH;+R`szN?HNr9Vv)0HI4cd$tXx~`A?b<hNYOPkUjhTovdR8e
z4p5Fcf7QsQ)^-_wnMcaf{ROJ#yQPl~?YNlRR|g5q!JlFj3x<A=k;!e>$e{@UkJvMI
zBCc)M7vVuHaUgNj&wTR2erJ(^($yra@n;T${JZrXvEP?-h@v+u?4Z-Vp1iz$^^(p@
zMmbNi9&Uq?IJBCUWQ%JKvJnBp_RLW6Q>T<1BIB-15_N!Sr1)`g&+u8gH}zltxD-rN
zu?0Pm(Z$C$7`bO5h(HefC7=X2ePdqhA2K(r<qq}I!w+x1<Db#s6_5eJCP--g!tJ%c
zhABmiMU(Xc%gT?kuWnIIh?9+$_&t*3W@;<v%EMK?s_nc`5`X=X^l3JiLl2p6cj={i
zQW)p9)d%WvekOr{VLkqf$xFaH$G2qwT>}~|R(a#2jzJSl?}~G%nshoF+@kIAZqc|2
zb%pFnMJn;UGA}|64a(Z7(%b7x#(xATKOxm-v-vj`S}SLJF9WkWu|S~Tbo>=&cixjH
z8e!c=Y5XfL<^gEJ5^itzsMU{M-59`2N>A&Io%o@fL1~JxeZ~Pu9^*QjXV35|caM&Z
z=Tu#vDDS=#Dr6)UR`$$q(S6gkOm$>#Z<R7E7x=_uzGw~W{IJ(Sbv=1mwX`sy0~1nB
zMi;o>;POh?FA?iY5sv}OuU~zYpEk9i*Hx-6>_G$^phHu>Htg2l&Q*o0VrJx^4ItuH
zGB3XK<$f8X+{ce!jG_G5)3D$d`KQ@U9{hf*qNs?h(0bqTwIE3n29m-eO+4C6ekiS^
z9j<z4qZ#L6W;y1?ZD7_snjwR4MBt;o``-xC-$ZA`bokO!zKY-G-dve(sP0BO7Hm>6
z(6Y&gB7@Ok9urGE9itr`hYVcFYqt*5xW&(rC8h{n+`@`a`vlH$ew1O-+;tgu$Rflh
zg>iXp-6+%fVUqq3pn{Oux{CBS1e;V+=zcr>iJplM&b6Zf-0`yjt?r^k#RG?f6Z#um
zxJ((CKj-7KrssZ5fD)_4<g>((kq&BKC00=;!kPLYli0<bKr<UK2E+Kv`8{XkOGb~q
z9ER3}xx1(B2J}|d0hPngJizM99iYLhjOn}9my-tu+J$zvd+B^WJY0Mj0ZxtQ6@s5C
zNBZ*43Q*(B*JbnLW{Xp0iUhUh@a6)dA1hEAUObZis2Nlml4Cd$w{j43Lgp!aHURO{
zX~dc})SPG<6{TG(y4o(dn$N_t{-w7P!P7y)F<q9AcX4TPc-uWl7oM{RW?aC@<e9-V
zm}&^tS5^+Gp_rgzbPj{?f|(AqVV4=@oyEq4H}nwW3}3ttI7x?e9Y-{sLBHEvVB~!I
zLu8S6rzU4R9kY&ec-0|!dvK@i#JL9(-VC_l>Bb!-KCv<#N7>Yoh&FPXDHXp~JQOGT
zAPqNB#6OIenov-7^$UB<kI%dYX%T-%)IU5Y5C3Z9Bx$e&k%RVpikG{I4_<Npe4{C%
z;aB^c;lZrym+#&yhPd?P(b7~-uU_i0%6Ds3O`z!wI<jN;J-R6gtmbT6vay^Wq2F?b
zv{KHHcG_|bR*GBh2LYxbv6v*urx$?wi3lNByItuREF78y6y65ffDR?P!or^$;)qPI
z&XMh@ldJe;2h~F9IJctDU6p8({!q#OjSa76FVhWfz%)XX`e-4?k(Fb{!5D_M&N8yd
zx-d(DGtP8=J>{D3AKuhg9^2c5vyH$gRF9r(RY{2!OxSMi68(C$c^7?RLsYZcj%ElJ
zZR1lNIhnJ8d&si!KIF+@-_FUvS09{cS`gCfd1Nu6aLMRdXyOJN%t1o#0tr*I30%`z
zXNWBMd8bN)uXDsiz!$$IoC>$x_B-q<%enM~VXBtb!;DT25?th0f)8&qVG<%R)!M%m
zWKMVu!KL$;v*lTZvfvfBl!;EB<2L+cPd2Wo$oKgF#?UvyU!7A0i|TK8_7mRajXg?g
z7#8zi1yWCDJTAlTby#sXbB&C2>>r1TuxrcY8UppJ1h!*BHlRo7ge%BSAyNG7H^XvT
zpu%_C2kv*SNnb2|GkQ)=M``q(8CY9+xoZeCpqk4NVoT@%?fM3vB(~8LZSJoYGOL%!
zjY4p&`|0h?;E2xCy!+MXxL;4|aB(ggU}|waj&V7x-3kkAa**EyoC8FQb2{5zbZ&`_
z+Ji-<hDy$!9S($kQ`sR?mC>RJY?AXD2!@2Uz&dpFM6PV#=XN*Tw{_pW&p<|iGUm?V
z+4-ze35a&CuWs8^MkO~p0s$kdU9+>?rHVkOj`a%k31B;l;jRT7><*R@f2jBPa*Wb(
zf>Jn+>=*ilkDt5ajW(v)K0P^=!r%(&Tr`_L_%@*VMRA?svNj0CR16nuukYy_Nj8@P
z1pCEGj}({1uaB+fC}w5?x{WSJg}D?XW5+Dl7aXvOskSl3=axi|!GNN=(cH-(`ds8V
z^cyXGJxt}*qvE<ocNK}dvv26#p+D9tt)07?n64BHav}C_&Za)g?g<RUfAH_1<CTi<
ze(eOsm3K&|te;sw>2rCQc*3bS#KhlraY}XX@zkY>K$>0AOY2Om;j{K&{h3L4`++Oo
zS<(bNc9C87qnm-8Lu@c+&SuTGQY`!xyj=Q$#C|>>HG!0$EUIbs_-x;fO@*g~62JRA
zna?=r=>rAo&-Vtt14ekZ1Y2x)c5Z-A%IVBub<pC5R-A8&Sb&ob)=8T}e4mq0zk4xF
zL1EnA?N!xF!Ycfp0e4rH{4a0*aZ=}lr@iN_oxdA*eat6V+#H8fNFJXi`~{>z5SrhU
z!Pyy(&Shs7+kC+NHO$%O!{S%1;pE&xH|E0&!jcXA=MV&F<=uo1>C=bg&Bi43U3^pe
z1{$4*@%`g!6Ej0fE-82WjqpXqUhYPjG2uD+T*K!Hi|ziUa??U9a~3A~VPZybtaF(*
z`HCy#B#(4x7L;m=_xKA1r$YEhvdYU@sx8f!6#Spp8`q)45VFy9F^i6My(ekBI@?Xb
z(e-CsrHsvE0<_`p$l)-)FVV6N|5Dx_%`#$u^n{VZbxkYqTTRT6h!b|&kDpP9#f}tI
zu9~$PWp*Pt#SsLe8k?G`BIzN|et-f#^W|`Zp$j5E7vk1kM6xl%S%nGV=qga>bvpXb
z+JT?(sgdp7vk*E|ciSmvmvEm(M|OAY=HU89V0i4GR$a;k^&tjX6vhcpQ23}g=+*;Q
zl-JN5rO_eyW6iT>*QSxplDj!VIxg#t!?OQvNV#;0M~W!%ivqUL<pzq>bXFI_>&=;O
zY*tH;ONKv3#a}WBDeYopa3WR@n50Whzq+Mc%<N9$rY)4YeaQ!P#F&!+BUR$t(q&Li
z^Fz$xZiN+^eJ;Gv<lzB76rOGX2oKk?s|d?7N>gP6QLeo{k!GCe5Al;&#x5PNr_nV$
zMOT2R@-0rB`rPGikHTrG0Z?F<h{K0h|H)SaNs~ms4{I@P#l$<7N=~50-#$oZ`(PYY
z4!TdIx?TIA#9L?5;n`HvVC&B%7m3spLs#B6ohQ}h&?R7`rLRx&_KwgI)ylNlka@OF
z-gafhv&vr{wgc13MWlapD?u#cNH<bD3ZZmJygb9oLQ4#Y?X-CtpOWC;mg!<m&uX7Y
z0xpn!m&+N<A=Y(JN*6!)Ro(R;&um~LM%UtDEB4hSZ_5||86B&k$1Nqg9pN*Bjl+up
z+Efk0ynN9vb_9EiUI@-<e~<M%3F`8E6*3a;j>mO6lbyIL(%@W^OPR2Cy+6B4!vuY(
zRsD&==Me*%1Vo6cMJ&v|@3}H$^XR(JyR<~NQ^~$IP-_Oj&2*Yc2(9X@cCHG_Opit3
zvi2T6LolXwOaSE9MszW+mJMYLF`!A!OoIsTs4qXxdOR}Hm7}NTTw)}w9k{(aw#+Nm
za$qmx7}vYM;kMG%xV^hOU-|Tgcm7e7u19y6C9y?9^!Kk%RRYVi!M_=I#T|_nSnyOj
zmbs6zL60hv<{k!W4EZulXD8qXuYxbdhm-n!_&HSHP^IJSp>;&0Wfbh=U#As@=;+t&
zj<`SRJHzz2t?-~*kR9?79sq~g_|{I+T;MdW)YWXzxzSH|8_{vSUZg7Gn!E3-SJJ3j
zZ^RBSP9JS5d1)r8CwhqvaV3LhKFn>(%EXytIO<#v50UK%y#8v%yx5Ha3<DP)JFSwq
zXHbbYsG1-+-d|d0QMN8mZTVx-dVi31?(OKIDQFVbH-r9aRlWy0Q-q>`jRy%I&-M{M
zoVFiUH^l<@_U^rQcng*>apexKaB3O<fXpbjy+nN|>OYZ=dz&2%C}$tTpwq*NSsCYU
zRkV_}pQk6c=Ld#L`>$-(TYm$=?48-~J2AmdOJ~<V@-Q^`J*uc2NunsnuQWe3!xe>7
z=i8>P(08|APYrI9GFG}erul%q6ATNU{`vCmWQ<|mJ{Vn}ocg)yeS_2@qn6Sg*An#I
z^aGx7PFO$~p&m~GuL0gC=#~49eXO=x_3X_Rd~>I!f9{I=*o{)9%w8E3cusxTyWnz|
ztcrN{Yv-!2@VNPU+~!NgBwo#Noo(@j*eEX$zeHDbJ9+#$)CkPz51C5dFPT~*03thb
zbX2<hsG+bO0)OGgpvuKsF>dYmk7bWkx87jjddJ&mAAWS=%80#I=AoD7|1s%*i=+eT
zkbTA0&utA_e{p%r+mhX^Ryv=^gLBE<PH(xyf*di&yoLd`lU60otjM?xjlWsJC!JrH
zb7w%vCvk1z+7QHK#8s8E9S}2usaLeRW>FaiNMKl?OYZlD!XZ5jXSxLYWp8-5LIODc
zK1ojlS)AhdAD0+FYui}5@y@}z`+jNm(k%%a6ejc6GU%}nQ?*iF1ru|Hhy+DK>Uq)n
z6~;rKfs5Eu=XOowtnqz#x)9aH7Iy<W*oO(F#V~bzuJgJPakQo?f?DqyG}(Ukj>6ll
zB>Cjx7#XerZtrr6b3rQ^^vC?REk@7$wp9T_!yJdtnQ<$xmW}mvIQJc;`bY3xd-R@E
z2?>nlG4Pog<u&>|qMlc;Ej#V9r17tCio5<)G<+u;<bS`+bS)6|MISJQQ>@=;RVA#=
z`EV)(E#vh0pgx*X-*D!CiIV&Wt#+syuqy$>dpb}U2JtY1Fq-ULrivykj#`$T1rvhg
z60jG1I16`zm;FmZJy3PGpZEg<Q<6nNZTzZ*36eO!BtOhL8ej1Br<59c@S6O!{>u9z
z_><n#Z5nz9pEAaaoSF3=SB3BiaXL6jdECSL<wgRR^m?mmcIM7zeQtQHdCLAE5ySHJ
z?}hi3F)A}*K1G5zm#Pg?7zmLv;J0?3XtvPE7;?xOn$=?oR6WA_0qDNuhkISd($ur!
z;=6)sItRCd;SO4*`ehrNeJ8UMsD*mPl^?MF;mnuB9~dQ7xkJw$xy4b51yW3%8j@ct
zQxDU!N++ESj2tiLELw8954fPnUCc0*T3*+k#V)CC=(o#v|0w*XD2Vd&xYD&rg_PxA
zYxo-h#B@L{&+jY%lN1a-uIjy@#w(g9i_{;AzZqS9(b3uXLinj*LU`2@f7Df-UEZ|}
zPG76_YY8APGaAAal*Tn4J8w8X)&t^VZq{Rm2A*~HI)OsNl>OSjJu!uQbU~yVHT-5v
zOa$~zcuNjmqm*K?{kPTq{lI4UufESG%!?j|2b57h^;4)ko4Sn}*?M$FBh3B$!La)x
zPUy}1O+W7qnU-lfB4D4_S%?Nv?TdNX^Ul!N=Y9h%*NmSTY85=r2uCO*r8^KGEjGmG
zBn>$Z&kjt&X$WfvSFWpYo3LYKPaAjl_P&T{+SO&AY`u26Ox<98Sz)3_t}enmPK@%Q
zgFD3TdW;1Y>4gq^`OWm5*j+{ULh``=9WSF(Z<b1)D+*_N;ey8Q71HYItXF6i%&~+F
zgy3MVo#K9>?#jGvj`(41#Pq~XVpV~}ugeV9Xo$R5FZf|<WC-0GDRv9)V%Lz@P29v9
z*f{SAElT`u+qMG`4m0BJo4$>HRH(lAa9KKLUNf0Ci@T2Ohjqy^cvxHdMNf7SIjho9
zA8|yfl*6W(q#(h&4*?Q6+4;gfeNWNo=~x$dwllMZU2NWi%V2-132cev>yo^dX^pBm
zmb~kM3a&M_i8XKT?iIjxt3K&w4X(J1(lg^d?~reu^v~snPs?`Fcm28ZO)Ho^7+i9V
zWF$b{GB}IP0N}>xy0^^^H<H(Qa)Kv(nbRT(U(SR7^5@;s)3pAW+fVvoDS^;u!cWJ^
z03|3wiOq}Jk&YDRI#TywWGmwXxX(Qmh$>-jwN`2#*e)GyP~KL&B1C0wq5Wb2;D$Nh
z!pNj>-yXV!v}ttMy~o)uaTEMd#TyH*ZDUQuv&)!%((tTiX7oV({=zu6)KNO2NQWL5
z<+w)H4v_1F4~3RrgNzk0ch#LU!xO#iul7C`YWX?dx>mKr=KK!)uNi?LA*OD^=la{=
zl=GqPvW3`$clbyJy0eJrg9XzJm`L#6m27sxW*^;y!Y6-5Jo+~+r1qtw#wa$H_=X4Q
z!h3TJwLq4hCIuzGMF<pqQaaBb0>Tdvv&<*u=I5=q#Q0?_(l^ubUH7nKr~(_`c7Kve
z6GCM`D}S_nFKke|vfB8niv9bDj?A^%`6_~3)a%yp@2umB?k+ruQYRp^OdfJ?j!WmZ
zHMG%&1oP{1Tw=?!X5PwB?){fgx3=BQ2bx%&XiY1$-HP?Q$BC;pUr8?J)5z9fy`?3*
zcl>&NBSj8W@M7*WD~2i-PJQm=)8POBf81r<7eblg6&WNf`tR=la-A4s-8??91MtxR
zk!_-OX}wz%zLvuLn_rF~j!SBYqgF#<ME2_ETYWS#WC&JUP(cS+KJ((=Bwi0SeE0Eg
z#L_!EN1IFapZQ1*4_5M1J{MVLB8Pr;Q#658zwkoY*%Z$oC25dZoj9E1z0)#?*4xw?
zO2X29#67s6E~iqlYla_Wps~D~xez&7uMG0r#8xrc-6ro0E+r(8H6tt(n}6?@hOVhW
zpli6>78ki5r-&p~<&{cz9H-K^ijEdNn!l(1JJFnYxFY3^N(2^WkZ$erPx%y^B%a6g
zpf4w@3Ryh?$G0Xeys7d>mX%Yj7GXh;$Nh$o;oIkc2inCGMlrQx8FqR)aA-Y2Khq_Y
zW~XEm)KnhCZ&*?#tG2_2=$IRtF$~E$_Pu$7S6j5?V{D=_-S93<^lcTo;(?}`eybz^
z_1>Bx`CLSR1k4l`Vx@=uzq0_ni$}+jmCnQ#zdB^2q&I!#Jg8yLW>o|jbsCoZJorDh
z#Vusq@$QqL7>_>DIC6sNoL~wN8xCb3Pbn%@?IQhlzU(llV=iy@n}2iO`)cW!V(fmu
zcoPB}ipy9ISY2XrnOb{ayu=n#FI%nSo@mLVw=ktW&+O*5NW`F~hQ(;X!W7s5E5i<Y
zY)}H9NQhL8GHg@n_tOP4s2`fK=#O<~3vwMn#YP|SL@?bnaiI&)E;*DJzho`}()qPQ
zFMCO&bdF`D=H-BjZ#h1i-QtMD@bG9==qvxN2Z{e-4ioV~W$z1CV5iAf!O@gVy}5b!
z5$AoTfinNkCrS5fLP|EZ`l}t2a!uQr%0_(FuilghU-W#1#(E^Mg;e|5&Xu{GWP>~(
z+%(Nh;@x<1qNetb87IfF=^w?*g(=0FwK>Fs4Mq|%7?k;V%<6ibbN3IS!mI4MLV2SP
z=HEPXK%>kv*P;pKLYc@9j}0yvXowK;{c^F!F-}|ehT5@VJ|G~XJ(BGi;!0wkkh3kc
zyEb1mbJ{-ISaAGvNx}~OYXu7fp3B;-qKkG}10tCkeqjd(*x9l~^pCyGU9L`~=DaMX
zmH415lwC2|s}p2)OXtJv$0nHA&1YQ%^Vt>me*)I^NOzc()yhd!hG6rwxl0j%3Qc#w
zh%007*rQ(7@S)&)i6^ugzD+jF>JOKfR27BwfNhY%Z=~B51HVJNyT%4Jd|i31Cl<Gn
za4AGXg`C+Zz3j?NoZw5Q<A&Q;@cK#t+Nr>I2kqY)x6XgsCUch#>W@#{N)NK3tVXPp
z+`8NN#odQAH9*rt@x3n^p0!p^g+LHYclAwU*B;gs*RRjtHU9#3@p!Xd-f)fd{_J(U
z_&qPfcII(78}LwgGd`o#<H0*I!_sh%4wg1&SA~Z^#)@8bT5F;E#X%HTXIkf@Im=nI
zk6NPKcM_Rnm@W~ysMF@ER1SGIn7qb9OA-5kmHTFAqew;Ge%XRua(JkhK&=qBfM$H!
zs*p6UgI=hZsBiSXprhwhed*!*i9u6TQTrGgq`fe6v8!Ez!VH>LR(0glwzy*ko$*%O
z(s)MpWx4AY#;!Aodd?dA-ERxz=Uw(XSf-}!Kei(S2?9+-j7M&+sZ=U&Y-8`zc(^k+
zPxd+>Qa*%?{8azKQWgme9|J!K0=L+f@SR~zP0jPB4rdOVjIc&qP6s4O=gv)qKl|YI
z>vLs-mH~S#(ePPKrP#gren0%HQsM)K0{`H^{5>guX+;-7hyEkP3o%Is*VA$$H<vFn
z{LNaR$2r{lSEpu5wK_obc^#wg5QDW~Tg+fYz|ofrqdsjQC&I!7k&h|=q5-VI+23P2
z@{Hx!9UXSMi(xWx*?$k-=lG&kld?{p={#2#Mjc3ox2&FGGRoJ0k(iZMfyh~o=jO>^
z{HVv%={euP?~<WXN~Rxyfn#9-A&e>Kxq*<H@eto^ASC4+0VB5wSI_D#b0-Kr_&R@d
z*X+QE^2jjf;ronZHN3?b9I8RU;V{VXsZa}N{QdW6Cm<Yv=+e~UN1ce_hGO#OE3B$C
zH?1@>u*uD!kF(i+Z6lXMcBoKO57=MY5G0bkKrx{*<Agx|wp^%f-^w=o{7>;%w>{_K
zm(1rkAsS`==UAv>R&U<G4QkZq6n3jMC;y#Aubrqe>Pn*zpMlC~iL%m{k#uoI-@v&0
zr+KDo>w0uoeoqqI1Xs0-LFDrP^09S{%vrJhb(^XOENd999ZvUpX?SU%H04M-N0l8B
zBI4s2`?mi4$yas>;#*h86nG986*%T5q*D8$YOVRM!7Y{YlkbgBTfcw+d1P8ag{tJn
zUKW@BVSsy%>*bGEZH*t~gMD7leD^u0#!KH_-?sRX@qr2oyD(zHGBE-x2aSwWq-+*Q
z8fZ)-1E(5W`CqH5ff!pR&|PA3enNB6b<d<%`!lpvDkZagbi&&l-pgE&d|NssC4EkI
z00<O+rhg|!XK=!>tN85=YOj~`;Ckj4rsk$&Yy=7m8lyG4<8PFFw<<O4YYFvC<M|1;
zxWv5^zHZ(;6V)pZ>pDcnpmC+h*L%cf!L||Lr<OQ3NcY~(t{1G;-lbS{2~vyo&9{Ld
z1N9^Z!oWrG=l>UV?-|e58~+c77FDJFQq<PgXwBMtRn;nL@6n=Ws6B&ftEIIGY6LZE
zgj%tMwyI_{h!J}SF=LDTPyesq?|ybaydT`xqmu_Y$vM|KpX>7;pU*|&cF_|fdv3)4
z=H^=Vo<p{RzB16T0E_oEPUpl@e;wPaLbPkPYa%&A+&W9hsnNflQeB>bCCJXJ{tWl$
z-I}|@$~{(WUOvNrG!5j5%1!h@!OL7vhj()aR1)xfR3Cp>Jzh$DZ)ZgT@`bnDw|`YB
zS>Nv{F7h+1^?szlOySqtE^bZDu1w>7Ln+nMYj+4&oXk<^-daD^>b;1gj*?yZ2zAnV
z6w<1#p1+P_^R33)`o5FdpBpH(AmNY!2-^Kym1)%sgSfNI=pYCSnmpT32{M|xcLs<Y
z-@wTjSCX&H?&w{V!hOd=r-T8C9s=*3?d+Y2BQP>MP&_Y3Gze@xF(U{Q^+;2zn0&Zz
zv0AL{FacsIVok^erPkWa^+(r~AZ)D~2XJpy0>Vn{9ts#b(<DJ<VybeD;<8mEJ?pph
zFW)BQN(040fx_iLBMpOU3cC38cLl^ptAr<RSSi2y8-hnNT6yV08uz}=@T`Da1~yuY
ztn{`&fiXkm`m><8_WDpXH(mvd8Zp$qWV}hy2Wmd{m$|>K)W1AsVZzX|qvCPnQFr1R
zEszEN>*erBm^m1*()PJ|!+ph5fI?+Tdc~`A^e<PQHvCG=BLdnDR~y|0Qm->P<r2I^
zrj@X1|L$ncSg&_)W{M)|{n5hnqrp)gt=COgEo8|6BF^i}TT8~&JA!W`a%uV`D{k<#
znbCxw+KnT0XFYSNa+>bS;AV?>0juN&sRc0G;8o~upJsp_WjR;<YG5s+G|)@RE-p5<
zv96eKD!y?eWOIFZ7+<e@KM8!Ew_MyiSS<Ws1Ne=}4CO|6GY{yolIPfwiLzCp3K>W%
z>Vm<8hkH$=iJpU^EwZ_zT$Q$Sq$fsyS4G&+KGEe$(V=nGP~Jq=!~O6*llYzOO%x}s
z&2#Fa(BK<-ujZM($>&1<G<*p@e>`tNJzRRQhoaf@ym>tbeB$*jL`Oespn&>FI!MQ-
zlY>a6*}j>PK0<kEe`R--(M@k@Iz>>%bT*YZ15eGmF~G*JATry=eiB;HqCuX~U-IXC
z=b)7RiZ7)WTQcCUDCC}6S8xc^?5G#4FSVL@ZbRdwmrhcNxDvc7JG?-MI6qG9{Gr|1
zW3F4E)+HMP16p3ZSQPM}ZclVH%4rjoGhZ*|?Q>+%5)`8C#7yR9=5QPw!E`ryLuOy4
z1Z)AO%QpW;_4!e~@t!{!Vz?`iP4&N&*OQ7GN<;K2N^~90(*5#C#4WK-P4iO{k(LrD
zreLv}ow%FmY-B_7EWC8&#PTV;(_zH&7xR<PMgBVe{Po?PNwu%Flza8rfcpBcL&!mC
z6ibLNQo_QTLkc9Co4kzf*$x!?78g(6@m+&nqE(av33&k&h}VqWojlQhA}RE*ycZO8
zSodb6xFhWGtKvV8C%6379YK$zZU+0e@8DaHMbkK_E}MX7<-j?=Lupre#yP%oNmMI6
z8y>fw7h)uKKCL~D@1+*&Jks#}`Q6?pX|D$Z>Xvf9(pWW4Yl5VbIPsPY^$K`-et?&r
zH$Mle=PDu#a_$Jy9CPgiR(|_OSkcRt$y+yr=w+Y7H`=x}y+;mKI(Ke1Z$FR&?mbRu
zx2)mPG`ssLmGo-`IsKtt5H6@S{QCC})p8^lF+`!`r8~E#I*}sN?B=fK7<nq;A2fN}
z0psdmpX>?qmeA&182Fnq+&?<0|Fza__rv8B&o7;dsCT!vgv3=-<PA7RgW6)|&F01s
zIy7;`sT?Y>tQ(#WK(kFnJDV5ycQ-s%<Ijgpr*5@D`DcS!uUU_=w;cT^v$4{?uGC+|
z-(Jlje(+m&A75N&WN1ITE_Y%x&e!#0;=y4Eh*CY$I`P3IV;Pwm?UvpU<>diq){fzb
z09pTET*4R@XcGW~DX&@?va?ZmX7O(-uSI=xe_dTcI)RugHC{z?cMXe<kr}k9&#K4_
zCMSL!KeVSJ3%Thj5u$?Xl}qt_oskqDOTPN~K;>f|ef7fZB!`7$2bJm#Rs|kz{PvmR
zLtRYmkP^Z46Zl`BgG#Q60?VD^88dFo&{2eNq(gFBDKgM5!p=G=eCvTW4?M3RlDSNV
zQmdh(Rocc;lgz%|p7SSQuAFW5uyzuideF_Xs_e#KWh01!5uRv&5x%4Hde-pi!;k3f
zYyDa&9&rlQdE>`Ypvi&;FKFbZlZ%+4YBfya*-)C!-JUfWIXkw(c%xBYlwz|N!e85?
zRu%T?6F(W3ML{45sLOW)Nc(%E4cZ#5f4Sd0Nc|vCUwPY2VZm+8&WJ`alJ#WRc9!`l
zwT?H*67Lc@??CsqZHBM)dPkFnmo$jwTZ1fDlMyc&MFpQswBK*-DE+r4mEY>MLoSP^
z^UK!7KfKy4el=!3X=Wps_(q2pV#m50PZgz<!TZ_{JEyD6Z%Q7Oy|&>o_|p5-r0k0c
zwNRNCRz!#@qNT@gx~SP0@iubBj(W9X$^rzGx1Avc@6cRe`8uypUhLpv#$B&M_2?AI
z95%x|q6kr!D;@JY$~9{h;zn<6_*th`q-2!U%L!<D=W+DgPZy-wf*Y4T8vTxL*^~cq
zP<+<D@}<S5LToJ|HNVi$o8(}nY)XD7)OtJf{KEChGT?z5zY3bq{I$RG8%$04%t<i+
zu5Lip2c+j%?+{^|Z|?(M$=L|Xvl-DacW^$iCAP189G^jmgy^L<{Qj)?(Es~d*~Hzu
zW&*C@oxfczCC%BqDbNR8c|w`oWPL!6Rr_mNa7`vGhev;U@Vb_PiLQ9*As1=DXRZg4
z6;+KCf@gcSDw;B!9Q-Cb9?0NXX%@E>Wu;Bo%1oj0%n>FXB4*no@4ha04DSm&b9Lxv
zcX%1fd^V5*aov&xng%IR<?FIV0h|uhV@ZNidB;s@r}qbS0+RH+zI8x{Y_?#jJ#5wW
zbRC}xQxL-(Y%e|+yAy>A*{^w3QqN5=dEgAVsclRpcE8%9t9oWTD)uKLLT6M90oDET
zYUgkAV^o85YMabPK-u+mGgkvfHeuw4e$y+$t!Vl;rX$Ozq|}*xxkwQtz{4}h%<N2k
z7N@IhJU}}kl5}ccTn1}@FXH=is&F%h?;dOHKgFm^5(H2X^G}0h`2Rkop4J~Ae_Bc=
znzzS(qaLy`EY-|%C>b>}O?B})gVaTOzFw#E`UtxN0{ZX>)Wq>or>&LrqF&NmUa19%
z-ua4QZYZXCQ2AV;`rRFf`iaSMkrpzvUGT8ZIpw3Qy@=XG<^u;K>DS|uD6;;OJ7u$z
z2vsW$w_%c>Ym#cDHFHu5H`#mFuLS?`k@Fco^%vYeK;He~eg_x5E{vdliFvz`?h>S*
z+_O&*hL8nZaVqT%%jiD_1l@Y>hS@)jx<<miR9=G=fwng;HbWQ7;^QgZ_z%%6J>{Zb
z)uOc%g-p*mS{W+m5p`{n^M7gg=b$3ZFF4Y9syQkcfwF??;tAIf;nR_?_UUAunr9zK
z91r$GWTWPlmA5IG$u}53j&Ejy?(|I``qLpus98u)Pk?y{7is7lbAK&w^YLu!5~o4}
zo*)}PtZr@Pz|3i3u@9nY<r$@@ZT;T=s;vKt5H-)d34e`o&P2_UtWP|{@vN{lv{L!t
z!bkG2))(KM+vydk^F#Ef{ALM1cviFDj@(+%&^t2BK2M6P;1#2SpoumAJ_lTfLQyF~
zyP6A8e!O*q#!Q&i5qxepxuzU4b?bZNPkyL{YGYMbbaDIjO06Y$D%4~v%LP)&oaix0
z&y!^)arbSQF7YKn_$G6jiLpyhQ^xUB(DM`Iu~B@zPj$Upwg0)N2g4oC&YVZXlkWFn
zTi$Y2NK<l654FjuQ=JNKWbwlZ8_ZAoDt)utXApGC9@7@m5lZBFX;DUab1MC*lNwe;
zEQM=)LK>Evjj+dMNGzGb#z(Eum$TR}D8uOpgS-x>wN7lpM6*RXRX5!R7#=X4qwehd
z#BMU^BHGsP9lbr|Q)OS+k4{46ZNa`eOr~fDuTq`5d_dXCxzeqAq-`Y&dylBrT!hKl
z)F2{-PX=rEHJCzV=?oj!tR-R-c)a84TF=<4KPIKcW$DzLL1SuUR$`JfZqq>OqCk5y
zWJ>&3`zixcz2>5;ahh!3df9z68Q?dQ4!b||NNSHr4iddb2J&}I?C|O$xw4W6zC3Ll
zDV}R{Li+iNou6RVsiS%t`DU;4=8D)LHd5-~&`eG@cZU57FArOW^2!#kxbrSouZnM<
z4vZB}#7Dfx<V$UZzRrAa1h;;*?GfJfA(~WlCXB6(;xuNv-EoojHN0I5YM(1_=t@`S
zxG^tgupiaticd@irqtt9FPvxY@{4FGB+A_U<JZSevvGSx8-<T+)AQ*teLyFY{E9AK
zNl|3FSqC?lr>(ac-@RmYNm8i(PIIK<?Zx&>_~EmNsOidUJ+b+o*N*BG?$06IkPVlo
z=bKh$@-69}-R&rsu@+_#Y?$=za0}TMxm#1A7O*+*H@V4>mnuhDx<_^g=S6SYjZ=Ac
zPmLN0N*MM`twK|93qE8U@r4mDn1%YvK>U;)3fF9IH}8NT78}H${aQAgd)U4YUKt8o
zn+ghB58KVkJ>q!Ci|}>F8qy#dlmjdMpNq{j{z-`M{nPgmiCzs)Z;+nyDFUbaNk)c_
zk%#O4=x~I7uCtrlQyy2d?9YgE*S*oYN1(&4x(%+~<Fer<9@sk)y1lKue0n_B{LB_D
z#fMTzt?X^mYL-!fJ%(;N#)X@^r_b<mD?&w({X*X2S9EnG6}KdUXPRGv7F}MzIN!rB
z7oLI_-#pU9!;c!<ymXd#!W#w3;bTslvlrf+jedvJSBebm7(@QNu6_Kr6=lZPAzs@u
zYb8f3l5F=nDG8=vbT%Bl{$A3{d`xHCh0D~$dQr?MCEy}zxg`OwAPhM~9nBRhRVw|h
zr3N$f)tGT>EWUbw7v#^(&ick{;}-?4T8z^3y-ihW$H6!MUX5zZ&$UXIkj)pS@nmlB
zPc+F#82J;f?eO$L;geNirf(dRnl0cGc$`)SSvPv?1zJ5jM>s6r6r7p<;!Uc$2$Wj)
zZv~RPwy)*vDQ?ERDQp09l#4LEn`|X8CZmJGBb@qEd@yU0vs)i>o^2j)reW3Ih3U5#
zfV7}dr=5!DK^I=_hv~G)Ypt&*$*-nT%H6y)H_@GgD4wdeIWcD?Q|=FOIX801I=;aT
zQ@ozL7iN1&aDQ1xPI-(#Xu#VEfy?vA(TmFt6G!3f?e_L&aI*FN^9c(02*|AIht1}B
z&YyE`=nTK-N9WDX-%Xe3)-@h`;<n^p#za6?cEU5<!`aSzb-vupe;}U`p#<-qyS&YR
zS|=_93f~+!5mxnr@e=2$AKqE}RH#|SE0)_vD!Gv`^F#FZ#mzqQq28w8NoD(3f(%S*
z`t&-?^$1oYr+e_!GOS?tDrr&|?LFQK>tzUK;hg3fWW^}*mp{@x`}NX#hWQ~ki+m}|
zG}i9?oHKx{sj@7C4(1&})S-c=&lzAQLSB1hMN}NwZ!fanvoVF|x{OIpoO?X?R&epb
z@)HMmwe_f@lw77Yf`d3c^g+)!ok**74`nq!uC3yrw|7h(DqKbk-$xeNw0NfzAzvp0
zrmeyv4*9H;zv#TlC-evORwE`Oe^2`CxBI~l4R{KguZHROjy<n0;0?@OUHy6+eDZzy
zl3gBJm$4;Vvywz;Ucu~(IR%#yay|i4MdO&o{XS1qhJ&}DsH>o1uu>y7<WT4R&+^UJ
z%%Y7-FQ$F}5QiIwiVCN!Qo@~*i6pjCON~B~HSFY8rA1%d)}D9PU8UP!+jZC5NEj9)
z(_B;h2Xe?H3wrdfnUZY2Y+yr&px#dI7#7^n^YQB0qOQ@Y^JY4$cixX~cekI)ZkTIs
z2NB(E6Im)#DU$ll@W)j;h1caC;9}w6Fh%6SL<71SGr6|*+ij0)r%SJl`Vz(F{FdkI
z-uXTAD%2eTJude*X(NnGiYz8CNx!w!E{6NftX#pqR#eU8WNX?Db?ImtZw;hu0{>6%
zt{?VQpO9cLsOcu76fIkxiu~b~FxVWpfBk+tJP7uEe!R5078z9d()L%Wt{0THD#0n`
z6E9I2))1(=c61qn4&v0+Q}|M*UgW`VJ>TP2=xq~;BOVC%vf^*^t5~0HiR-FD&JE8w
z<4O#Y34b+)^3JD=mlJy?SyDv1Fp=#v_$EwdG?Ez8&h+kb|1i)-<@M9Eg^g9Uw%BiF
z8PZ&y&DR|^Zdnf=Zf0yL@Yp<Z?OnS$)Ckpqdk2*t_H|u$WBPW+u3*0Gg{O0rK26i6
zd8)w0!?kTw4B<ises@kOtbg9(T*6y3eQmCz@b?VcJ;gHofhGNXGY^A$%?S}5wrVjJ
zC~rDu<tu8otFmm-T)z=u-JPk^YVbs#y|{lwp{U}w45}t+OA!}TU2mJaIa(VV*s<VW
zY+`w{O!n!%QkLUm$uI+L5-}RyMrVr{VCN(Hq#kVkF+ZA2bgJ;!u^JB4k_P$9@PMMc
z+_U>52wn|zl6blFFV5W@;H3x}5hoej9AVb`?E&`!QK%<QC*b@=s5%B(T#JGBY0{_p
zFPY$C+yV>3*F15L8%8Hw`SA<Pbg5L9o6#0zHybhADNlv;w>l~i6TYteLw2<S!zqex
z0iV-(Q0;EiS&2fvW_u$e&emy*JX7D=E&>lIw1sXWU-g$eHpG=x>Q$orJv2hlb3*!U
z@_qceBJXd~h40w7nQ&P<-;Sus$#o-xTEBWq_jE_avG-=FokC_XxoQg2=bF5><W=@}
zDU7`D(HT{`-jREdQh2R2H0EiA6%b8XNBd)>^x4;|jDkHQaW(K>e1$bv=A6wQcIx6U
zBIS?vdAWw;l#3mCibUH$jE(Yco5|+OVW*~HF|eyQ&AS%bJP{7>H+d@UQ{sQp7aEIJ
zIO?*g9!EfmdH1aO3&U$Cg?F;o6n)IHqE#n8D#vnsdaH(evu4C_adR@PFM6tRs9u0*
za`lW(!@@_b3h6(G$NX@fb7OmMW#j<Pcnl5(uH(e-X(;g`A`G(jL|ebb=Ol+9YVb6I
zh%%dM-Wk3lsCAEePFV#z^T}#>V=g;m@t3z@Ub|+PB=4{4yv&wB?Ha{q;isD=*DzZt
zUfE;&>=SUM(n|b%EKQ(IIe1Ole@y_->iE)P|GdDQ;pU|QizfxwWR|q(k66akt-~sh
z`h&WhXW4%l<@KAvgG-}``{*jiLDEs1r|u(_-8LI|aMD&l+p!_;)A)pg18BWU3j_js
zYH(*;L*hJ>jI{#V2_sTv6)DHP0x4`uITz`Nu5zl9j2o;%kGN(xWjVkPS+5M5k}O#Q
z$uEmHx1v`E888w+;hjiFRhH+N@Lgqlf~uui%wZ?7(Z#5-I!vE85cxZb*JOIaU$xx(
z`7{G5V-_*~NBec~qG}MBfsb$Eh!==H543Z%@}^&zmjB&SSn1$+%)>Z_j;u_dm5~V`
zGlLZgKGmwzv{+oVDK@pB+hSD6TzjSbo+nrP`#s5D{1vy@Cwo>4*lF1G^U~e$iC(!`
zd3HY*I70?4>G|)}=TFZXALR}1V&_gVS!Qmms1q`%qU;o%`h=gj@QF&K2?1OyQPf(E
zC?B%uTD;3^_9Cw8#o2QY)Zwx;8+T+!xm#3t8E-+}G?T(}jx(Xk(4P@!%+~Y#MxSb3
ze;?H`aWjipuLQ@~vkq6_`*Ih<6^8P_*XvC&7EyjSk;ilRvdkKvv6ElhjhjuafmZtP
zN#{s~#<x$umVNT&%v@+>R7))kN45~XuJxi2OPV|lw9*e`C4SP=#n#)FgTF6pwJoCx
z>Qcf}MDM70Sz><9BBk`wFYfDXtULnhG`HERB2x~$tTx9-Vipjk`8kXR?<&0OO#`@m
znxgq_fZ6P4PSsT`?klJpRNHl2iEU)bRzRDDo<T4op4)%cBS)Uo2QI3*X?f#%D!`Yf
z--!0w1&sXgOxHEWUZ%n)b#9UU%RSS{h?Mmpwnh6Z2s5m|@L6t9afj4xkU!(U$Roh`
z=BF!E&(pH&=*RZxVICbLWZLI_?nmX?ZUIN@Y;lXTiI+Hj6c1;9LdoGlJeR}Ic!0XA
z2?^d`tFCb``K?NL<FNOP?H*)H2PY${d5LWns`Qz8%wN<daQ69)YeKrhj!MTyfQQtr
z>QMi)TQra4EHO$5m}wK=VtDa^-KnxA*Ia&~`CGoTpv2{dE9mVdjGt8cn4z-eqx&Xc
zwnVJX97cVSUdk_%m;<kO($If(;iOS+Q&m@FcaQZ7rhF-09qj_`zL|6eK};=S6XiU<
zCVfkGM!RGvSlb#~xsolJ->j#-ejaXiN}823qGR5&)zL2CjZ%IzmoTXuc~Il&ltr*d
zDX3h+nj9p8FIROc>;k^Z&ROa<p&ys|*SGd~!Oh>8K8EQ$5|e|tE~*;=>)KYt-qp8E
z?{pbU=P2O3p+X&@FR^~bKi7m7T8XX)Vx`<YR_Qa6N)bOA+xnuvD<oJZ$BsC<>6cL4
zb<A}<brMeeaU5b|*-m+zD&XSrMZ;aCgQ<=WC3J|mowJzbn1&etUyqKy$h!!3TUj3?
zD3Yo_KJtF>B_@hcZD8#k*0}Z)_nhmf{RGV!bS}}<DA@vQZ~SSVcIGrrrV%3D)E2O-
zd?$@5p(Z_d4$_8iUI%c>!l3l@8O^vBjV5+Zek?&#R@7d6$Zo_X$V#^)&fUXwlZVX5
zfO3ZWf4u-TQ&rm4SvhAAil{2?SPdUgORWjTkX800RV(?bl^q19{>%~;FFePKac^8N
zx#L)*Vm?;_4L!1#uxwd-iaQW2;*>*cR(rOx%Uc%1MI3sL6P36hnFew+LgkDWyGThC
zUyFC9Uqtr2Y4?l`pSwqcfc$Z8Fex-$qP8_@ZZbB7*DW|jAR8)Q1N!m2rS5N@6jeWC
zHI_6}Cd?>SV`tXbtbv1*Tko)Tv)sy>C8b;k_2fy#=;z;H>-WJrG{U_SmGzKlQVKjt
z3*lSUVt3YCHx%l0^I^@#pg%&g<vCvHB(l=xY}LN<x|oqh6Urrh;wq_C8i{n?b=|my
zYu!nyg1M8G+fc1&b>$O2F^?Fp<OwKikLzo1b@*xOz}f`%%sgXnNCa&dUFQmW8ZGgR
zo+GNGek?AIHmukiUYq`Kp(m$Y9ye(ls3_0m_<pp4swj*z#cbz0ewxR~V-tMMxBe%@
zEQYYh3pQ6ffuKLE%KReg@1F{N%O2S{Fv;|gcyGk`XQ)qWaO<v2%H**MIPjRG^_BTS
zv!f;ziMPyPLQ1oU$*N|6ix=Wxw;Y(lqF2YZ-8%Csevs4a{Vf^&)JvS%d8Kkw*{Sy}
zoHwwHnT{dD65t|fMzS74za7&Lzni6!{43zQuMf_=^!#XSK<fh*00K8w*RAxum%f@<
zdsi2pVaRXPP#2zJV(zWB6X)I<#ezjea7EmmsOXHj`gEjG>kKJl)v&POw|)GzL5w(M
z_Te-f-IHDJKBBZ7d`PDozU<6TvfzE6jCGfk)QJq&C=VQV*Np3MQpvNUz>xD;=|*W?
zL_C~-c|O7@oN-fA;#&rxAzQe!p7e}>cV~ld;O1F8Y!`!k3YMh*TRE7F>f3dAVu$cw
zxr1B9nU%p(-N-(g*cO>#rut%AnTpoQY&Wc(@Rd1LQ<j-d-Ciz9<xkZ0+D#iGUTR-}
zKcr7-kf7=>dM4P}^mEtq2VbvMysWlM7IO0;jnpvczlkCoFykS~>wRqg`atT#QRjK<
zc>Hq9lfsPq0fa>DfgiE_AZI7{bjqi`EI{k?{sL$75r<+^Bkyt#;Zs|#@##ZZ{*S*0
zk8Ok`J~SKb#|W`6bx?eE!i%W){(+Ph+gE<7jr)!Z5U8%**oo|bgR_%tdx~L788$xx
z&z@WJw!X^XJrW(CY#uIBGfVmpo{k+($?phn2cpxYP6H-xU+9DjAHEgy(CjoPu9J8v
zfzG&e9a|90Chx&PGr*+ZPs|u#?K~WEJ6r9{Y4xv4&)2jvDsN=ZFfJm}%Tw{^hxLAa
zue<Uoj&djb`VvCyoBytsJvqzalh08MN0>-@&5gv`q>@LOCxJjMWs~=fGC=F&BhmVQ
z%jai~tZp=fYo%Xyqh-vdeDNw?pWBKgt4e0!jf0;r+3nl4&vuwAO$}N`Wvp70o>Z*o
z{2BNz@LSUk01$yCyw3EYoJdK(Cevj()6Au+$1{UE8NjKSt5Po)62<7{^q4QL=`6W6
z*-GRiFFG`~MI7FCA<{oe1wgM(h5mZ=xH6MR=`7{|^0D8AF@Y-+){>uSy#wQ)Id;9E
z8D~)kYdM*ln>ke&<k0<0aPC(-Z<W&)a5@ow&A{0fLJ07vfL2ZqoTLw+w*3PaIImeS
zE~?b~q1@^d{W5YWNb2qz#awVipF!!*5Ns-nAF?5ULuD721|e*Pp{Z_T2thu6J3Lo!
zW(1Ro{)ZIKZM&~pF=;%QPfAV~AYJYzd@pAN-R^M3kkz$3c>U5y{K%I`=4)D{1K<6M
z262h41|6OSiP}U0USUf?zD@SkGZnY(j}wB$?rdAr57ULg3HIOBYQy-nc5*YOj{Q6e
znjfS1?&PT0E(npmUhrMM`pG)N-rWjrCoyTalvK+pOqJb}>e>&#;xfZ{#dF%b8pRKt
z*aKq?bxT%WO`Zqhs}OH7xtW^|T8&tbXJ@1<>qD+&SN)C4*fIV~^%KIcpFXh&W&M<$
zUv8`nGyVIERg&f4QjO->?-o}e3n~9YTA-{bu@sj+#D6F{Dzj^j(7EUfL83y%`6XnJ
zpe)cQuL~>x%-quIg<zwYgA{DK7gF;kHoOj_QT(<_T{D;UOlK2$Z4Zb}@eng3`FHZJ
zq)U^*Ev;Wk^f>Q~Zo3unfRFGNTt2&oj`P{fu~cG7)!%Xs2B{#Emv0|g`}qz8jhJ7+
zQ*mIQ=!Dci$Ce@KgkS*%*hA<-Bj(*4(l~TZ(G7RC;_PzwG46Z^k#AR?Xgl&@A}-qG
z;JkQ6JKWpZWlCT$%<Gv~dbB~{KoCP)f3Le2%y*%-4jOeAo@h{J#gXnR_%&(Z#(Brd
zL^_(({QHvcQ_pYj)WDV>0^djeqE{zv(SF_~3vaBOU%2->G_mP_QWGh4VB9$&L$^0t
z3ZkE9*F71yBQPW0j^D}Y>r3L}YkGVf{8`#!53ys+Bb?3M?t$<hkyP+DaO%c<F2ImK
z;%pqxyl<g*EuYPi0yM10Tbm}I@}|9V{{ghy5sa&0V4y5&mI)vy4t{`Y?X>)`+@0k2
zFqK3+G<@Xz@zB)3c``jUqPMqiR~Kge0`HYKIT?uan!AU4OEXQiY%RF?qH8u1!nd!w
zm}EzRX0EOo_iq2mctu^Os7pi?G51=&`kpHzHcVgML+rLM-YGlwKlor0YE4ltX#O+a
z+wCUJ$B<qi<lve@TZwEaQ!8#ai!$1%Y_S10V4=<~)ve}A=~H-p8SMn?sp!#0@mOlN
zIoWJs0ciE^=bMSU{G#&Nf>Bkxe^2Syjx3Xu(upO;@SryoKc>`Q#ZbH_tvqxw-PT8G
zn>|ZoO3biR@l^L)Qm!;;@v7yh2wS$?tlGNWAlE;XR+TNFR699&<fkb~6){ugAnQ}(
z%vNVIkauT9w0*6S1D8Brsb^enx2YAnstvRDf-g-Yu>P#?f+Vba`PsZ8#>3ejTW2~7
z-%#(`j_5|Gb)lmT(c_$6XH9Y^1B@${wM!e+VDpzlcV5AS`FDCe`1w0@1jc$yOW!Ns
z4VgsuFezXW-pRbzbE@1EgCfX5c2CJc>m>;}2A4qN1?ydDyYBVxMsBB1Xn1;3GMr71
za3)DhRV}C6#pp_&vuu5~m!y(8Bt5Zh4s1W7^|K9qx2`O1Z>krgj2qzfuzyr(!a5=&
zmHs1zVtA*0@balb0>VZ*H`NU?>z(b)d=<cNm?v#KsxkIw^Wn#8FB{vR%y_}l??M)=
zvZZ!Ii&ln3QlDhs{H>qUw3>STODI0DJ}>0xvh^S?VK$WjF!smYNaa);ipapH*t;uU
z?JiaIr_UF)i?e<TY5kXqBD&Nu^n@h8d($156VxhST{so(#q<74{H7>mffazr_hH7j
zS1mRR3?V|3$2+LsdGV4#MT2g;+GnFiI2R8`ezKU1$<Dw2Lz@<2hEq*45rRE=A&4K3
zet}i%4p+U*&NWhUlP?dj7H797Lkzj~DnRA##f|#)Tjnsj`{x$t)iC4(yz=&z@)DJ-
zJ%@4Tl42)knq&dee`R6<dnk%XM<;evWyUrR+^{xP@CYU_eU5+)!p<(2Fy#-Xe(G@v
zz;iekC_kJjthaNC*u6($QFEOyTlxGq-X0W(nRj~^2M;d)T1x^&G5_Dy>q(5*`xBy>
z<zbazj(s?g_~e3=+wbq=+pia)?C_UGBXBWd(I8b0I6K#k@uIgSG+I~=LZGE7`aq5#
zhE8pKtRWTwKA1p5=@OVJV?$}7B27!FG3=hTi0#~i+_LX4>YbH}7NZf&i(FO@q^=e=
z-s<g5QEuj`c@m4A&5t#yUUjiv^h7<}TnXDKgh&H~ZRDn-b*>7gmj)eW_b^A{zxupg
zjxiEafo($f3<JjWoR1L^MIEy%KB2$AFMs%lSDZt31kQ7s#e#|m3mSWG50xihtCYsN
z!qB9dsD$LK7k-g>dJbdS;%j1dQZJ$i?LperTa6%GF)N#_qE{3|+3!cyMTgHP!W0>9
z4-UO{E%)sW(bTFer@F$&u$4BU^Mr`QsQPU%O%g_Gq8SU&!|}$c2FmiuscvxZ?Z_7V
z=NurMAgUL}ruAcDW7vi(R(yp1+HIsj)*?gQ7U2#?cV%?H$jjC!Z|Uc@WdbGVPJ$jC
z{&2$2QctGGLX%nOFq@&<tv+7}sV=$T&3b2?@CYwOanJ0>@GAQ>7Ya@NSL$g@I`;Mt
zi8xci7-|(drq!v=wUyKqg+w+x_q7#$tex$DF1#n)8NyMN%@>sfQRwBReArAOyr?a(
z*1yLyD3Jzar~dCHJ!90?v?T;@S@Xo>y~lFzpEvESY#dK&pR^gGKY44P8(4P(z}PlT
zh!wm4L(&$i_x);As+r)Yo)Ry)gI#gl4%Au|Z`44M@RJlXZqV>)*TT1!zQTP+r&fIH
zhr+BhtWD>Oj8SL@uHD3V?!wMWOm46F>d*c{D#m--B^m4M9Y+k?As^tikHs%A=aYIW
zhZfiUDmiH7pX!$k&NW5EQBX&;4@HCZmtQqh2J8J!qF1mZu1&2o1}OQ0HB8N7z{uCn
z{5B={SL0I>Jbe-PcW;iucT3VGDy61am<7vyYl~i%lkWHR)*fVZ`v{~$YTcJCf8xU{
zYs=!b>wh6oE^j)5yE^;@AEsUU6hrL2COp<^yYd-LZN#Qmh3oU~8mGDRCI^!(C56Ca
z67^YKEAHj;X-E*6RCA@DSc#SSRhG9xptLS}xbm(?EO=SV;c*jc(ufo~+t6;L&6B3Q
zRj=(Cm%E<=X4_~Wj+k8{w`s^VHD~QI<+ApE20dGHfSIL$N?Ixo*^=ob+OU-q^P16J
z5=Tlif^$~Tl9vHh)wnPCxx82;e!as?mcGA<Y}-shf#niA1>Ah>P70=-7p(MB(BH3a
zOOe=DV{&;V>IpEfh*z{hd7wK+_aFz*QN^8MdYYF(X=21L*#ElS-zel;v^c-ofT@gd
zg}82lCBXK1)2tz|xPxMpP3{D@u;_j-WtYBQ*Y)_<pL%pe+wql3`|rqlO(~u(9vyOw
zVny2WCZOTg)fYlEXH5KJqby}hskF3sYaZfZfHen$o*k5s<y?jr<@oY*cQkO51UX7c
zOLt3u^Xc~lMW{F=IwPUcX}4Fl4pb}NqRIAcyOLY;Y@Ze_!<AgVku&4+dw8^=;4Lqg
zANtV~SzF*avhh0owYswr{NbxYb<#FOG6K{VS>tb7F-fZt0<{Qs54(V{S-oVO$_4Y<
zl3rT7omU%_meM{Ct9}jCfOiDL>wRF540c7AxzytKOuvwUzAuG;gx3b@0f7kVRig@N
zie~qS4{#=LU#2>$?V2wG)+pfw^oyxmIEY`dN*h}7=GVjXS6Z>Q*}oK+|9G8ol1^#%
zP}MFsgAnxhQgWW1=ZIuyP7ujDH~(>pqx-RD@9DkQiwtIII@(+k(z3yf$6eRPY#1oD
zU>1v2B&XVN<<|U6i67<WN7C@5b=%<%L7tY@WQfqC=A*$0=7-I0U(Cg76YHGE3n1Al
zebElEW~;N=rK6BQ2`yQtAF6EJS?hh&MITTu6>pgm>$rP#!zzyhY>knGX(8+M2H()o
zDR06+;}Li#Z^W1Tm4K6rRdxawJmE3|HI$|tVA_KiQayk0Fm1cp5Ww+^=9X(teB*Ph
zvx|f#$)YD`1XFaR(p>lrFx)%2LG@Gp*#W}ROt$X+Z-AVBe6GCh3FI~p-{LQh-Y_r1
z_??`8=_`KNG#-sODEn~=L45JFGtTZaff$@RfxlO=J67=N+kLmE6GFrhpM1y&lIrKP
zawe9KgAbp5n|bb5z^iKZ%(ufR^zs2NyqK`-@wYMSk?Prc;qoyX=vJ2K9e*}HP)^8l
zvZntW<k*dPcGx!*I4MerUmTXhX&P82PH2Uv%qpO!R)hcIuH(`nM?tQxX?<~(<*zS5
z-i?!Lo&YvIkJsx*-hVKZq1-Cudk*XC(-|9`pelDk<@NRXMy59rzxk8gAKoc_=8k!7
zJ<I7^V|BLl)Pw4_vQ{&94}HXSE6}N@ad!OHGq%of><u?=GK|RmTWNrrKCkB!agMOn
z`mhmNQErPY4fz&m2Q_^?tEkja!8&}}?lWO!ph2Vf%Ee+c@!?x+5OSWQR8T1Y^sHc&
z$pxVKqyI|tNSf_waytla=S_>4et?_poqL*&=9(8(%@~-Prk7X1U33f$4RxxE5gav7
zv;y8S$|!3af6$4nvp9ogEY{y4Xm{ijni<3S+N-JzTY@L)zboz2BlUr@-NBYn388{V
z-!d9(TYPT_wsOO5Tg~h*_cIPhjSb3xJtn>eQSIKQ2Whd<gKX6Sw6HcPJt2G%9cE(h
zp}SO%<eir~M@<N1U2!S;LRk^~MgEDExOC{Dxd1&V<q$#|>}T@)=(NvGbE!F%SZ;+q
z-jMf@k!l;TGSG$^G}(ACB9wd;ywmTsaLs04le6-=A`-+7)_-*dVRC4LFu`||Sl{F`
zb~>FEU$6&}4aP(batR<k;jT~9ENhk0(OIA*&()uP-Z}~ilMSj735R}2_wa+^<ta+t
zKVSv=Q?ZK(oH|XVF#G&d^-kr;5X<dVuFiAd^~0&FQ|D)IngX75E*ppTOg2_{`jdMQ
z2{KU@L+EtUoRbi%V?T~-&YgMIw1El4X7`GD?cgQBUi44RRW%;vVr4C|BJv%*cQ-bi
z@df4X*;w^MyxSWyZqtj7U#I{HV(jt59ii)f#n~lT14+3o487Y@gf7#Tg5!H?3OQ&}
zYs7VGIUc?d1FSa4)td(d(yC3pDB$I2X9hi{ysgV+?@6-7IuL`);wXo0Z!cS*8B}}r
zq#9(tD4?xSV=`7WY9P>X!|ow1NR___<O+O94p)vQfk6JIpDoFN%}M`_&bPpY{?C|K
z1AgRx#x6Jg{~m&ZZ&Ck0$BG|;zC`~sIDysSf5%P0H}XFtiYobk4}sH*D}+)3ALo_U
zDvH!yiqD>w86->AKiNN(53L4q*S0)$Be~%goy#7Xv;NiW>)%*!2TrQR3H;(izx~2`
zAEZB_?e+~FVqEd$PvP~EWOufr8?dyjzcuG%pqZ{3U);@Vt=5Fb3&;*R4q=IIGo%jz
z`MdCoOq`-?_E!&z1!Hv8<xj%PTFbrD%-ThiT1SRh)WwPDA7TKtRD7|OfK~Aa2hgU#
zHgADcsqT}3xZUE(a0b;n20QgMlR@k$c@qMI91kI`1!*G*nmPOZN7LCit#~td4n?6%
zpP^8_q*}Q%vM9my2D88RSnl-OI_jZ~@bx*!<FoNnYs$Zmhs?EABGu%5^+S(yjwW65
zCETuQMNU6}&+xv7iM$`mSWQgXw(6AA2`dLDU9M=m&EXKSy#UL6;iXUoS)nX@P^}*L
z_ki^Zx~mTTz8~t!V#l`<xR=fhh&!p+BJ-OQK`@>};o7SwaoV<|IV#Y4sBHVyq_q4Z
z`>Iswt@w^X!Sk)6%TMDvEcizu%Y#d31B}*KfN*5wVl=OS^j5ku@&~yEDCxSI<>p7Q
zX7Rp&2b)-VwPe7@w3+UT%OP{iVc&x;sP?SgDSA+%NU+ChzOwe|8ScxUF4F6YAT4!W
zVYwUrPt~-;^nn*lii%hBIK6~Uzb#JY0S*akjLJuqnz4ZDBC3_<n5&X6p1FH4Be4!$
zkq;Mj@AaBB0oo#=d()ylO$P|Bqk*_yypbvRU`K+8j{Y}6MbSA4SmYDHR6@i`po_mi
zkt(KliO!!OLn_F}8{>73M>j#kO2G4$lqO8&NJUZ)gncg_8-J$1bD0CsAJ{r0FoXb3
zNVk5!F>bD7X{-2%Du2ng$uk`fx1Lq3Gu|-ehBy^Sx;^jNvaU^l>Z$}6g1Oae)6I|}
zz4TmMc05Y$U!8C9LqFK3u72jpj2T+9DJIN`YIy_IBIl(+TE*`Hem(8?VGyY)`$&p{
zH0IvmB-omu>k))?*C-k<yN+9F%moTBn+TZLUv^>jW4Dqni7amCfB!VLf5%=Ay<DKw
zl@4ND6DU51B@k_cTYex;e+v&}f<X0L?z$kJm8{%tM7+;nHvLVnIF5^(a#kWr$FIHw
zQAVwqn&=)YO-Y|{8QJj?Puy^a>iAasvghNjWM8y^so9)`0^+Bg%GN~<%C6Q)XgDj)
zT>Du0-xqulMrRqNMY%oGDuL_ay}cpBb@<mESKT&W1Vye@1+(u|a?Y~L0S+Ah_XIAF
z7L3b6@^}vK4+C>818)P;B?xru2e8{QwZ{#uFax={%kXsn?G%%#`%34=rh;IF#qe|5
zF!)V=`PnxT=(<*+bt6+j#x-3AN*zHy9Y~yf=uXsnMC+`;<x6VHV*9oSkewkyy@czv
ziKMQ!tx>a4BmwHbno<Wg2iK^5@ycoEq1{-CK+fB~$#2-{??)75;ql|5QYhLuzFXfJ
z$>$&$2g<Vn@kdI4sqDJ7<!HJ9thKO?F423(lWWc5ri+M~f++V?kkwS+#1yWfu`a`N
z1(o24ScG$1v(6Z_gtj%Dc)&Xhr@}i=m#XIk;W^1IL<$f$6A=FB&1}7yH@!?SC8A`n
zFQQwCz|!-k8KW-rHzHaMhpL|3{rQ?-bP7e&3g%uc6yvCiR26a&(XQt3N8nfFE7FkC
z8sK4aGF*2fUcz@hhbSW2AI&54ES<81`+G;XMuUel{M?MP-a-9O)e?p#{Mbj=^^}Ot
z>_AJU7Cls-Dm5Lg**{_Iy$ihQXG&<J58%*pj~R0@^RYl1a_$GXT{&H=Q4BT&dqO5n
z)1_N$V@Dly7$TC}jks_2WO)>-j)#FPkmK~xv@c@uQ=x84%IB71+M&mc!qs>6%`IwC
zWfB>luk`8uV*#lu%x_Yv{@AYrKR18-DfJ{T1c&8n?ER=ZEaDqW3d-q0AZRZ?027Y1
zaxoo|h(XLnkm+wJFL<5HB$o1#y>1?|I%pd!*V7JYr?rhdj3O-jOGtW^5=S;jNwe>V
zuxBUX=wbTf^Q+#}I-#$>(#Fs+Uz=F-IHqP3hm1!Ps;>o0=&)6P!@^zz-Aev65N}fE
z@{`B;&fpEo*%bwesQx45Lan%G!bl&^C`R+kA3DIFbM$cl-$&Cpr_nnV_~fjc=C4YH
zG7Ns>Pj8x6?&L*S<M`d!St?M}6M%|HdW1A!l~J8q`2MoW-hl06mI%=X5r?y0R?NSZ
z=D+*dcGn!Er4eFAD^v9D>f^#8Go8qpMNjNpho`zpS;wY@nl?63OSH}Tt-S}U=Y!SJ
zE;Cy5wBJ|lT}%vfXYu!`b<V2SrYK22r6_@@TepONM;5D+fxS-LBj)mhN^T}!S9s7F
zJLGs6NOO4}{+DX0;{?iK?Ul`PdSp`eQDU=j@JK`%l@;OU8M&q%)77r98r-tkM9UBU
z_`KU@?t#)#jQ@Hi;DJKu+-EG`#D}f8HdClm-PSdx)P3*fnt_>0>pBWbIJP?Daeg&1
z@5_V22k8o!eF7|N;17@Cd||8D<E1$5KnOkh)^|)q&EkGn$wom>qt3xosX1i*2Z#tc
z9nIIc{hY@Cu7J2Lfh2%u<?P`e14wx?mT+i{ZWUEJDE<p6R-#*@A7I_~Y&B*?i`sv_
zXj-EB=Q34$-c+|Oyq0djTv42>S&XYBUw?EMe<U~2gk0H4Ovj>t)C*=MpvO5Xc0T^!
zH0Af7jW;R5uG`SJXq+w+H@nfjK`v(!+aTuW&{W#hx9A;dt>al}aCY5`VvN;bS@#lt
zj^3L|S1p0CEf<StX!}}db=D6p@p~sbM`7Sqmkz>RRN)u)54Hl&H%l$Zxy@?tFSkg-
zL^XiG1K|6A>l8N8YW|=E*Q;dz9P?w#pn9sj<TXHz#E%rQl@vAVIqH`X-h=>N3DbjK
z_cTa57&^$m@p*#{d^nD{Fa1Ye6dO7!%6k7$-*G{&Fb1*I_bK0Ves>9OmYrAi9mYI|
zIPOgGIRf5=9@r*B=zNrCo`ox;svT6)E&m|IsV>I&$yJc63a4c!b^h1Sn_Ns2iM~v1
zX_7AQ>hn3yn&!l?lie>-ikrghs8||wdqp8X<E)}rW2^1Z4Vs9Snl=B>qboxftHlW5
zO2!V+NEIvs{4OJYbI0p^PY;`65A7gk0`v|XwL5E!*;vMLpk+ue-<-P=lV!F7F2ngn
zBvDjazaBI9Z2jkK)_h)8@$%SRWGDkbvnJ3I%SY?-2=>~AG5!5cAn_Ty(*rZPgGJ38
zd8v>XsHQ?RAL;LOoqm-sL5Jf-bjg3Mq~=X;SB;18M1~An(KgUs>KZ}bQh+0gDd&^a
zUW2S&!Xx5^ZG6hwjRW6x_J-EyEpZi*wSlqK9BD_e<Ur0kMyP#=R%oF487amqo{KwO
zR(ti-vv|ABFji%J$r{f;r73c&{&=B6?`PP}X2lO~kRd<pGu`zb<D~mDvSTxTWT2!Q
zKyXe8Joo&Esi&`b#EBCFqtExc@y09SE!DPvy4~@{nS#U_Uxm{I-dkxAM>ZEUi*fF_
zH+R|}|2Ns)zl~5?DWqp9JH%+ynDmSc--*PvWe|jYl`|j7=vql)qHR@zA5SsgFKBpW
zPUsvr{2&}SJ9*jGivI}^wSxh-J`M~d{WC7?HCqOT6Su_Orfzwtn`8>oil%O7^3DWN
zLF!-FATw;eb};kwG|%J>lTs%>$1AL9?PM{8th<=S)ceK|{t7CT2n1*U3{1GLBXx+*
zy955S7_wPJSZ5Q1Jx-z3pQ{($E4`s#{l8v-HJ^<>ppdswx1Udo_z8cAH+Etq%(a=e
zgoLhDjw(l}on10ujKorYKgAuh6-g-o2ywBOU_Az@M3z;%Y5YDkwnD7?EZG#-yXy{g
zbL>p-w$ehA{Ie&@y&H)CWFe*=eez}J$si{eTSIctE}s>MOzVpQx;ECV;NbT{iE<TL
z>&GwBMMQ(w%<(lkqblCeS^2-=Xn-6CwC?_<T%p)m9~1e-N5P2WG}{}AK9*_|xpMeq
z))Z|LiSA#3z;enurmtOz8);O@?X{&Tid1G(DL7nsj}Zp|dQmF#lFD{}7~oQ>2bV1$
z`K4|}b=?%L?U}e;d?qj2Z>92{>JU;}Fmq%3sSuuMm(atHyH8UbY3E!QQRuansvKC{
zV5z+<XZ9>*H4zbLr`ZqIP$lJ|>kAO_HD7b8eH$Uo{rI0sHBy(-cN4wAFvkBNaYe_2
zx}<|5o#zPg>JsH6X=Ux3UbRed3FX_^-qE{&7GYhhV1zAmcy6U^*uon-$yL2omv^ae
zVqzU8*jNEHx{K_-rfYQ(QdhrZ;n!dN3@PcD38!CQ^7Uc?fqDT%#O+KE|ET*R8B!!_
zFKUX>aS<rGfz18v(ki=wmtEbK3u3=dHvcqHu}t6xngrn0VsHCf*;r|DDpVlss%mvO
zu0m$gVOGnOsN`FGQ@GwC*!JroTY_c$omYcZ2<oVaZt09)Wlse8)dBOBft@p0J8uq5
zowu!g!BX~jZbj_g<Z&%SvC<b6RPt2T;XvJWqLWMR?ZCo=;#-YBwHP?;Sg*-)#RyrZ
zEWZrs;nS4=dVeivu<d$Z^t@<?@QnO@eCtZzg(1TC7{`G@d~4I`LnZ$vq(k+;V^fm<
zH1`8e%fVZ4T6>SbuZ(FQCSRhY5OF61yw+LMjaq##)<r@2Z-KRU-tE1d(TnIej*Y_Z
z(_V8kdaTQp^zu%9^8`H2OON(nN@Lid03eShwf9C_VH2fkUkudV{Vef8imERmr;h20
zK0f}|be7Vzz=z!RbkX>YrHj+Ty<X+WaLQyIi5Q*8>GihSNX+;8?IDFTCT)A<BzcO$
zm*^A<xucLgsR7%&YUbkQiSSupcc<tcXrI-Z>EvBz(NvHtxVpbtE4jl!{%tKM`h28*
z|Ht4l+-Xfj>7bHvbc18Bk!!Mtexc^}@-J_>>mb#z-Svk;ARf`V73x+jzO|Kisq%;1
zI`@eL0(S0hSfuY(;6J4?76NQ}v5_kIB8vEg!EM4>XzXxMnEml;LQ>x3f33A*^*`oq
z&9x#*lnPP>9@e{KOF6DiF~uiS9_f&j-=Bm-i}KQ(iX+WyopKJRk53mPML(>D@!V6v
zo(g;m+WvWJ%2tImu<_cVfY1Zhe<1!%P<T+VdSF{Kqz45+pnRJ7+FPK)J4?Q}nty*0
zaLenBEP7{TJ3}ykeO}%0>j10K!QQ)0Qx3!(VCl38=iuOZ1^n}SSGl57yuNIEMu`q5
z&q^s=R(=Eu0FF`KFd}X!Cc{#<I=GAUgB^&r43T*QH^M|pLz(@nW)E{UL^BQ=T9!H|
zUGX9J_D(%T-Z-_NsUU4yAZFMN;tgbda3$d|J(j-<1X{mV2?9kuctw}W8EV=xpWk2r
zyXPj37qhXJn_0}LnElA^zi#W$rC&}`z!Zw1H}JqlLwV^VEEDX5oiAyo(iA+Inc%2k
z?Sh5}Aq3ir33xfU100;yPFcUHW}dW=q3qW{_+gDU##&cN0=G7bFt%jbi?-Wt`w^O6
zJ<Hkr7-^6U;m_b}?(>*@0w`^{-nz*F)(lkg*6{yQwx;8%Tx~|m+idAx=?igsRB)&R
zoKQ>|x*-)Tt?H`WwJYwiM+H(f2V@RV;U9qb3@j!rjn#zRb~ypJJ|n+f1sHDnSkR33
zi<sw9j&hUTH;QK8Gj-a?PPG!Dttpi^0M%WWv|nUm$N_16?^sm94`k2*Qpzy@pOQEd
zJP`?wnIpgXciGwc`*wn*^zEjSS@Mz_%NpiO2bSkq-AhFW>4Bed$%5$w?p^Hcg|b4`
z0Go4fH$bONh0>fDaNHBVlm6xEPR@U54=OxARE#J@7KMmAz_Yu}VD}pF>2s$s`i)eu
zA_F17CH5rklu57s+wa4SSq5iQ-58a<5N3cB;(KQb;*N>EQjCHpMO;SvmzCi0-KT**
zVqc*A+Bkcm_Aq_T9g3GMqq=83>C*Ln04ZSlq5IS<AYw(9JaErO>K{>9?-seGJi%pk
z3$w~kiG(qV?^42qVl`Sb5@ec!e8CZmd#6PPBFPb?IKsvDAM60D2*m%tZpNCnpsK;Q
za^WIp?onIMaN8BF-eMYtd4LP4=!1RT<J7a?tJwY8A`d?8af1~IC1|(9D-?L(hx9mk
zAUpa1Rg2zcFmjF`y^n%2qFDYyM9YVX1d$}=c9ZhTbh=z{l{}Z#4{pWMaAjRqIw2_M
z9N(B!eOr6c^Rv#jRl8rHd5tU<>AAJ?J@jt~`mZmI!||qD1$kLZf~bt{SsyK$%P!e2
z?mN9ha29}Upu@PK{o3MV6tLQZE35#8H7^AHU7VZ*irRkroe?CPl<h2h9r}55U5QNI
z?nMGTLG9&*G8?+pKe~P?^H2%ZmlTtx<#=w{PX=12qdewQ&HX$rguDN#GIw#v5mTWo
zYRx*<Lq5H{n!Vy=)yf6<fbRcaGL9pkyT|N5ApiM1OMwRPCO`y&^6LtJ-}qm;?f;d8
z`~R5Z`~S!3`?l6Ov>NyKeIZ+pT*i;?nEW61-ZCo6ux%S1P)Z~e2?0SwKw3b$ySrPE
zknZkQ5fKpSlpMM{1nHFSkRD>_p5eQA_w&Ac|JgsjwSVokKG$No7Q@Wk_Z8=PoX2sT
z*KJ~w<$p83EVW?2*ywisH`@Q%Nk!|;BHVg!rb5-B&2crD!vE5w<+z`bft7W-K=oI(
z@0ByNQM1nT=g-+Fn;4hC`2W|xQ7<n(6#U2NX9#$JF%Aw+Y)*G~cjA0^qhPh-yCx6#
zly>VeAsT9shku>RmeT%0gL<^!d1C8H8vG3%T@S|v+J8RB)jnMbQmkL6#P4x1!CbeH
z-FCad`t9IkM>#-m+R!7dXsf|>C-$IWivkT5WznhUMdQgZe|%irFZFz-o%v;C?fJoC
z5hY?<x^XMrN5IMWKOZZywKe-35+5Jmc!#{Hv7KdSHf~)w0`B^YfLP5Ut-^{5)kq@d
zF+-2V+=T^$D_^0r-D%}z#DPbAOiUII8Q1pd&Ukt~>wi8i^W?-9bp?3?*VfYdDRlH%
zxp(U%ci$`c_X>Xz_CfQ>Fjw8y)>d0p_Iog@hRD|cye+WwcN|tAvwrQz@jS)rv4qx~
zg@uJ`q3JMeVhw$LMp@E#A65R?Z4M?;%g)t1eMo5gpYO9dTzDLqoJ{Hm_gNkXm%?c?
z#Wa0k+~Qr6?t3izLeN{ItE(%S$0_?Kvwogz9Id2*tmDA~udJGtRUtb&`zs-fJ0yZe
z_T~Z(KfbWD1G*2N8%*QRJT8U6n?9qsL(xviisowP`YsIi4>tjj)gMc;vR2dXN{fot
z^d5=SsTdox_BCojWSB>2ApYl78HIUyQpT4Z#7`UVmRfw4ui)eP%Ddpgyur-8^Ex*3
zTO(Umad&r*+-GE9c+H71NZmIu5H~^o_z~J-(?36bpTpl6XJhiLlFfGgKj>p~piqX%
z$w^5+7RX^X3<mqQ1lGf7*W|>6>nRIFqzUr{9Bi<{t-^=x2a8QL?)!Rvn;aRcBk6*M
zi`>G(#u<n`%lk1QBqko%I@vki81O!u)IN*LxRsc%bC|{pxD;P>84+YYAnsS$Twj;)
z+dEwGdWrA3b?4D_^WO}9Bl^f|2yA!P?Oqz&9?jCLFb&(ik+vq*P*KUGk^lK|qCmCr
z=Afy@q!T?&z)KB`qK{?Dpu!D7xHnAt=@M!O+<~HLln_2YH^jYMoWx-%)x5vn7dKaH
zmvVe^GGCkufv`eLO8hKm{q5~p`qTk1)H5_(<d@V>RfcS{h)F_Y<{Tc#e6xHCX}d$f
z-H!fA`rX|uD_{LK9nBO?=Ct|A<*-z|*yK?s0}0h4O%?EBhatr6Ub1s==%+Tjb=}m~
zy6%h>)CnH<QpXmj-AfjWq8~o2@nZ$1rjnbRo4XxSpg;mkY4I_e;~>4m!<+YWi_bR&
z$YZq;4y?dw;{d~0SMWfVx(4QV?+Pm`)#{eKi(c>a2qP=PamY3=kEFkm7Y?Ox@7`XW
zF#8*;slAfrbJ<jwFuKa3YM_?-x_$hAZ)BXmK25HwQBhHL{v8;fLjAX<f*xSJmIVtH
zywkwoR_W*Q<<Z&`c$3vQt?bxBgG<w`-+v1zh9O`KE|vq@v`X{{kGArK5rwaJcPOvZ
zQeby+bjqy+_>|oC29fi=LK*grJK*i&c!Cp~tzZdx-yZeQ$YK+-?EOD4C0!Qs=P8Vm
zs3dx(E0XwZz!<59fk$D>Mp6gXF&UTb=$6n0uk6^@=N#|ew+C`vv~B#<Uq87zy*(M(
z1&b#HEXmihK;WFzZoAgxeE06-RIx6z1%~U!|12IpVdJ;W!r!Ft_oBzLU)>{Y?6aHY
zyVjw7Vito@5LXCK*t(<0@*Vtk_2CLM4t6)^``0k-HqxWoEb#~!hVW&Ovx|$Mi!mLA
z*ZPx1@DX}#{(>Pz5LavvwV;KCF=h{@pQD02w*ZAg)!f`_UKbV=e6pd3EOyM**qm*!
z2Iwjap8b%Mm*@BPfG8gg;;)!JBPQMr!k`!hb5{*U`2~-o2qz~eY`QFiRxvrJsI*az
zfL7rX?`PlR^}g%LGXLBNN`IaE;kY{4;<TQ4Q=(HLrfux!_D8hI@alh}*l>w{F+`;2
zJsL>%;e;Tzz0?K3cJxIBDnus3p6$VZqTT(g|C>VD|38xG|C`MBf36+^`9JUrtp5LP
zf%yN!%K0(-mmhqk+J|K;Q0M@GRL$>~#*}!nkc8e_z5c=N?Un~CZj)}Hn-=i*avPdz
zMz!yt(bZ<D6K(asE^xtHx6|*xezc48Va{?%UAulg0dbC<bPiped_a=|=}ZyO7XvT$
zg{Tl)dOK>01Iv^=)Frc^<A!RHNAcKg@ZdgXFA{-VBa*M}uL}+k;M1fszs{wi8T%os
zrlche*WEfyNytgKkP*sw_PDkM0VK5$y4(N*`ZJb~cZdfob{Hda!g|b{@1$A<@qbWw
zDIVVdPP!J}iUMgat2LT)*WmLVZjHP?|B;6q<GtZ=nvNo}B?7Ycs1>+oK*FWZ(xag=
zJ1`E$FL6jG3qwDgmsHKTEd<}Gt*->gNY<S8M2+Pq)Q<S0Y*PT$YMNCN4;gX!TCh?w
zS<nEcKEcw24%tSZg2y90ere2Fqa_D~RQm>lgD@Chq6Qv(1wkxR{n5kRjM9q!I~hT0
z2AiCWw)gwArAZS9P19Ba7kMXk+?3UA`X=^s?ig4F1wYSIV_BwU1y6E-x6?qYE;EwH
zaf*Ru4NLA3dQQU9Py3Yk^F`FrJ~Y&^R*qUZwoX=@17@6U$&|PCRo)SV^%74p=l1DX
z<O1pzw6ah{JS6Y$mA+(l^4V<ylk>B#5VBmX2YVeA{Br@_SGAyzVwV<AZW-C$5IH$^
z*ZRtCZuyuwJ&DRtP_)r4@5j)gg{$+8Qq%m>5F2*j#Y38s<qwPktOoX8^W|!w4>~8t
zkI!s+w2PdVRj{7E8(oo3d;Xiu$iWXS*R^0O$a|`1kj!zQQ_|J)E@+L#;ntmH&v|>y
zPyd;1WqaA;r>1vNTWwP2S}b6JM8o3Krw4~K<VM}hBmKz8awq3=DFvt|xm<e>P`)(?
zqH8uSPz2TqRLfp>McRej#SRf%9juK3nXb)vXB592Wd}&MMzfYx#ozgq?IRbF?0@Z0
z+9N?X5M<l+o8fmBsrt8HdoBGAn~Ijurue!GV>^1=@Z`L&?23AubQX_mvuLzpxS<Xm
zO`qR_ZYfHeoMY&lI&K8Kb^N}56>2VDbjNWl@%ZsU?%m&SqL5~IO$gviXfjy6-`29d
z6g+Hph3~6b5epgt4W_CoN7^`n!A5@xSZ~wR-KOJ<3fDJp?@Vh9jUU6}TM;OM!(T!A
z3smZHIy8<Ao)T|q(mGRk{?B(lTfj<3c6}ZkIDl|AoL^~s{eq9q4~_-bSa+)~kuL6M
zqsSP6qK|0h1!M$DD3ZUqGBi+Cc#+!bhcenAc$|!IXm+V>4F99$fgb!K0JPARI9_%#
z-j7#FWwS$k$zzd!!HQdV_VtY*Cr4BL*XTTQUa-r^u=_oB@UcsRo^`l=W;A27+bZUP
zTNy?9PmYRr=(cZ8I#Y5RyF=Xx6_<F(1IYSV>_bF%4Y6Idz)GcAOSPNNmr3SeZ@GDp
z>B8qC%jf_5>RhRfk<rHP?4KZ>EahGN>5;Mg@@?elDQ;owgJJD-A9cMp?sG1;sM@D*
z)q)|zm%i>uc=-G-HstUrDAxm(EL9xnN4jJYiT*qD3B#ptl!1wxo}-)FTV-q^W^8he
zdigwoc{rVs<Ojt(gWqnMMlRjjUhb(qUUeP6Wsk!Xt?$%+o3NQ42FQ}z@P-$#RXxM}
zv$Wa%e8(mLP#D&Tvk=jykyV$mJNxRB_!HH+{ao>vFH5`^XKUkY#*))^7AebIT#y+F
z=F`NGQNw#|GP(myQbpbJ%C>H`h|149iy6cA$1iu%$#R!V9p?3DS@j#Ol3j;#dW&8U
zYsk<UXspX}kzDo^-3P1<N7_|ArBp()l8wd%p(npH&n*0{qS2H{NpmY-S#7H9K(UUr
zlQMlk4*@U;>r+*`C*ZU`=61?F{e4x7nps3&7Ms3Azt3YWE`L&aFDK@%<$~P@>;R>m
z`&VLG;p>qxZw0jJ5g%(b2!DM+ejLB*2cC`D=LhM+PcViXZRAh7O{_M)p6*9aGiQ7Z
z5YI^7qLXx)^n5xBY2;>^M|s%=Bg-v`$Zg5bh4GPn=Q02OuDSTvMX?R=K%yDoW<*ws
zj?W^;j>bY!_8TnI#t+$>+6A45z0aq)WpV{`oDj@B#8LBKn!ke_VCgQ$t)#ezT$}~g
zSU2L6XJdYQ1*LKR=c(0*j|0ov=@qbk&Ze9;LlzvMMEhsj3=Dr?;^_Id*rY78kx(Mo
zSs-M{amcFcQ*GS*5z#9SwYJk74sZFX({l3VU_Kbv@xyyoY#6`Y0BEpRIs-v<8dp|&
zSUxIBZ8?4HG;7Sb?hh_9f}FAL&w5Q<+?=2E0yx-#)7-fm?>nI1r@6I=-phP;9X~x^
z;DBuVf*dCDXrG4x0i!7*d3-U=tc7yL9a&<`{%^nv`UY`@9pjUELt`kc39PzJaJ~GE
zIK%bFpV_y$pVm!ouTH_KP`-b<`c*6`5)$%K#tN^gT`}Rk%5`+j*zkwHb@Xjah(r1v
zvN~XP*_kKy6c_#W&P7KWE_n?4NdM>*2W1y|=!gJc=YE~Bh<1D${SX;V%jM6i{W5Y!
z!KQ)ho9U^#=KU;1YYNBp_5q(cI+X9b?2Tg-jJM4PK)<_X+WBm!mh(+j9ew_cm{AMG
z)q@-z<)tLhB&iDj2BYyu%U%dF_#rF`rR{GH^tCk;sr~%JPS0g+l%1g+KjFCa73t@O
zHPC8~=j)ZPSQq*t-P?-Yb+q6#E!a8pmOvkP=EI+%3=XV0Wzk)(7Ir@r&Vf#$vSb(~
zxB7`*sO#Fm+g##_QlG0@Fe>sjW(xs_&ep~Z@TK<D>8ex(I9*?0Z^P_Utcr&8i2JL#
zoHEl+1g{%kVO;wS3uRFC4CHnP{)pKmdTX!UU%|)Y<)1-gLi{a;>!3=r90#Qkw6gqa
zI!S#C!A$u1P~;{4=xS#~joYr4-D0CA1}^#N&UikJtl#zN-*S`AcYu~qOk$Vp49AIE
zR_5S>80wps)L?z`U~@i?J#E4n%P+OgKpcNzL+}JC6W0-Xm2e^h&JXy7pbQSTmzEZ>
zvLT@=v7=%UDq_+iIS;?n78J}hl{QXEFKT8OE}rSCRy=5`rF1vb)FP%weMR$W%w}8g
zhX^L~>l`Kalbf~gQh5t03rS|Q%NM(xth7lS91*hY;2f@H(*&GAC`o()m-42qJF@l3
z)0|gRja(N80$TgyLhkC&Im}O(M{}+cFI3GUcvCc<R}65N=y8Sml&YnmaGE>tnVDIU
z?K-x@F;-~%Q1Lr{^}fA4(Md!`K6>StXW0zfKRthC@YC&Y)#XTTL^*%^d<3KFsx?wS
z+L}zbI?iKDJ>idpToNT!sAEq#4kV)UcdgxgBHIV_48QZ|S65fb2bMK@4S?|S+|G*F
z9?!?iOUAu=BKGiW<W>32>(l}IBJsTv%ogJdDBql3vGBvVYf)<YTq<j<k|$mXk&oj}
z&!FrmYd&J|DX9B?t7|WPDEjJX&GWC@uoP#Lh!IZ;^^OA-&P0M_ZwH}QhlKMcb59Ac
z;ClskGrWnFkaW@w1kPJ+7Kgk8r&}xdpg1P=Iw5pzwTgH$FJPW_G&`&G5nuS1$lw44
z4aF_)+>SzQ26o}r7>1Z>K?)%qcdPQ@Fn{fAYG2}F$9c$1K)*{sFv`C5fuIs5mBga@
zJ`SX5ahSF3wu3-f;5QO+WB3gl4-exdH+Nmf1vWNz5*P;#i@{90`C5exp|pw*J#t!s
z_}TrBB3Ey4%1!I^KE7eiqF)e{N%JMKy1?>Vn|=}Ov>25Zg0HFhH0EL?7ury!6{ND&
zl*efU@i5YL87ZqfRll+*5Yy<oEipIGHz?UBoYh1irs)dR99y*tPT<)LG!vnRwWj=o
zi>UIo+Q>Bfd5ITzyF1U#a$aITmRBLRJf+a~8LO+Qh~83)yM{(g3J0i)$_yLK8kzH9
zQHA2Ss%>bxgK6vt!?v1)5`O6ZN^f|pEF<$C6yi+At*PP>xK4omA;-nw<>g&pU;kkw
z?7EH5s8zf&scoDqpTGhk=X3pZwAMpFNjU<z7^m%_eFCuwuQ18ODKVW!f-h;!VSBt-
zWrHEVETjCEiUs`ExL<IcUxkep&6Fk^Ae$c$*@sl09CmOzDtg$d2AODCLC7(aCoT>=
zsrnh;1ZULQ_(b>72NE_l3(@4CBGb8lwqdIdJOeY&TK>N11gAtif3@Zcx}a&NZX00`
zrCXyX5qc@Kou^6dc}5$L^L<pq9KmImDaRtjtrUM*WG@x`z-_PAl$&WLkMu>;oFl#%
zCKWUl<*n_^TYtovHYEGkuPeYZ{RYIDf`Y==o}2K92tbmO%~e}<g0OH27@Lly`fDBl
zcwRzeB}6qiEH457>a~-bAR{loHfK|2zS164Y}C>SRevTnAzC%v`TMK9as7<=AY*6%
zbBMOXJC+wb5?f6>g|;8kwLd{;yfx0Ud3-AVU5JOQ&KWtvH4iu7(Dg_Ts^5n*nDJ_?
zxB6wC`?{1)s`zwKN|Ky=%GxOqy<GCLq~zhn<f6@aN%xTtqKFq*swsoZkNCeeQL5_{
z%k3*9<#3(ZU`Iah2&Nj*Li9XpaczD|XoTn)V|<NSdH`jS;*Uo&M&_ZW^mAl!J-4SU
z=+Afr76c6#jB<-1Qt$J<zE~Q09>42%(Zb03ysiLkQf~V}>S%$ELiK!|THEI!P?`)S
zb87kvfX(4Tg%^AAF_K9j4#x(U<YJ+MkGRVjyS>!OL=>}S*hnMhePmsy!_A6o15SRt
zkLF3s0P!8lapVN;9P+UzGPyr0VTg8n@)6RV+0|-u@h;>rN_DZC`z&oHf74njvfWW=
z`Xuqun$4SJPId)Os>y6k>dApleU|l(xeyq_cT%Y4kS>(po87MRMDwAS-MFlDS<+iZ
z*H0yvf>l`Fb-`_h#GZaO-!dY!H?@pzF$iy}uLJuJ4ktoH?svA@#^XRhtG;#JR_RPR
zPad)e$D<quRFzJHb8c2Nb10@Q|G0rP6MvlZ<8W8@&4!!0pydlIO3J4X$6yj9`=+g5
z_KZCWj_q3X3cHq)id2hGhml$=e(*ZKQF^8LhmGI!Nl%j$8T<@)O)C!ZVoVOV9W@y`
z4E^q$1T&4vmWSjY(zYme+ORWRg|}K~4H-Uj>%=59v_0<_)-j^;tx~evG%_OOtO$cH
zFBPO8jWFIWw=$T`eJwJlK}fl&&wkPSDjSL=NX0?SUWt0+nLATj$t4gk!xw{PToxV?
z=gYwg=C#H+J|P3#lf~t~0596O&nBJWUi(L3_(9SRi}KXFX71Kd(qyYip{WxRs?*+n
zbp&%NGxC8gFDjh5^@mrs$uC=^7AF&ff;nn+^ke{gnnQ?UA+)Fc-+noQ=b3EPpkBTU
zhi9Lpmh_tU;o;l*!m^Qk8dQ~&kA7?O+JX)BbHRg$!rGlUehdy0iHgwATg#;<N{G8u
zR8PC0xvwp$SZ7Ap(GB1Ie_w79A>pZUkA<mwZp6gYNfw7r>TxUHa;X|R;2WK-5e8Ho
zm7llWxJA7FAOq$uiNO*A{<HhVU8w^1^ay5t%OuK4>7M`kkgu@HX@vP<dw!c1)7jeB
zDyhIyGqXZ@+Z<#Ak4gZxu)UUQ@dhWRw&Kz2<Ysp(3Y-AaA}*=ubWJi&wwTdMDo?g0
z@rfty7`N4g{&Igk+Dtfab*;EMFVo-V@sszf(0uJE8Lj>7K}^ziZOC~M+W+^C#@J%3
zGh$#BKJ7#IK=iUKS%89TYzRU1>e;596NSLr%ohtBGRBZ^ZSx0NWt|)D<PkG0OG%L+
z0^PkLgNS_!%;HuX-R|w!_31qSaPZ5HydZ4EsN}1ky=q-<`riPxg03ra1Q`VP{gk!N
zL3W;_MX6pj-?DM;W#nA0NoXUc$+L(D7O$Izm7L9OJv5EmxG>qZ4?h26rsFW_G_m^-
zou+QC<~_g`_9j1Vcx~Rj<{jhY!>@3Z5We<sach^xnXGT}*VS5N6K+vUdmG8HU#~*(
zv(F1-a002Nzr77W`i_l{%PJ_WgAhl<#8gb1_bL?gWeKVHZnrht1VQ9;bTW;*Hi>$|
zbTW)PJ@g1=eoBsd>IJXdcVo}G<#;#KdT`3b5c^RPhOEehi$%Vg%&Eo-<QBfzW%RnB
z_0h8CZp=WE0>}cBu8k~XO-U)v3S1Bg$*5iN=}jp2PomDJ&HNd#FSrSTK~KG-2dpcj
zzu=2~n(I2m2?Rn}cQhrQNS*y+qL80=*nAuE+UB5fxAf6+<kzoeyHh3AR@7&ZP<%~G
z)Pt5sMk(;d3E4i>KYxTq|6sfM{Gwjz!h#-L<!Qv@;G2P_oDjZ(N|YcLM!eb=>Z)If
zII2H|9~|Lp$*T`O2#P^4$%UeebEKU$q7d0mat%9K?X6Lt2}j^}2=kDazn{<DCGD>J
zh#CD_v<lrQO{*l2+jN=?e%Eog=#~p3ScU4^_8lE=|Cx)kq86~B#7>F+NLj{Bw(fkz
zRnzCWNB(S<ish;IpCadPATixWu986p5?C#}z7Rqd02NPYKhWD-P<@H_JwCoGUfZY-
zfDFt>u}}%IPtZt*ha=y@@P7UdYwoX0QGjMG&m1hJg!Dc1n=g<zH1%j-U@m`?>s&JB
z$fy9r5V!IAk<@nFz1csi1l<dxAa<UV&DP={n?TK)NF;vsHm!G}E;CJCrH=ScG(gaO
z)_{2UkAzXboZ%x&gLhXi=3m(VK!^Kp^q;@=;CS}Ti(+1&(K~m*e;IGNR2@FBp*vOZ
z^()zRfe3(#_{iUfljnfRPcWinH<JmQ7rI&}fEWcJ+>p0h!f|nNou8gi&(u3n@9gX_
z=~aKiBA_J^q@j{{F-}Efl<~0Wyinu%gqH!W*6*ucV5ywC`fB=v#EPd=FI@e7?5TXO
z_{}zdyS4c8)$nEcG2J$I*7w-VK}z>a?)<%}Rk0%{;NjdUR#RMFVGQD%m%N$-JS>mA
zyTyNKh&@A*frAs*zg-Iy4)_FnQ@hWTE{ycfljF{$;NFa{nxaUA6*+T04YO~)lzMwb
zd+C>Py};}9L@Zc?3$~)<Ru$gb7Wp`29;0Kc$WdRLxfk^Uuhcu_FpT~XfxcBoIImeg
zQiu(bIMzR~#{*0*(c7=jcIC@@y0!M+oG$r*Cs9244aC5RSzB6m_E=@1tI%{oAJn%#
z=eDyDd@+(&??~<wb;phWd_K|L;*=*>VL>j-&;)JNTcwMi(AytBW#+x9tZ!ith8-Lz
zv7>Rz^fU^W^r~o2rJJr9F}^pGwJsNMq4cMi2UUD_G=xJlim_G6G;@vI4DSfG7ZlJK
zQu~huxL%W>*Pb9#zdG-c`&L#u=gnM@P0+R@#5GzxOvsDSV)=?9F&~HU2QT8@8DVqL
z(I``QYS7nkX)^IAd!zf$tTZ?bY*Lq-azBQv)oeWx!(7!5ceqD|1*hIJGP*L$vsIS7
zE}M@5ir}$u(Jj@<S28!`(JnKxJ9Q1k7xPSdpi_|BeUb~F(Ipa5)FM#GM>~UhVTcB2
zR{a})u)v9VT@&=g5M0sUDu!f6u1D9^;YLTA(VJgBYtW_ZD)l??j_1iP%2H-DJqKad
zU*Cxhz@X>4hbpnxORMAM7dc<PM?F?2vwT`Yf2%kk^&SURKc<!fvcKEAd3a@@seR*G
ztCwlFsDp^Y(4Fx4=NGa%BxxNpQYg>pQirDk$3w0hT6uoR>c4fiuq0JT$7VI6OM!@a
zvcD5Oo4$4{>(A?BW*W$Qz94Fct2LSZ%nm-C$&MD=*xOwKLfeB*ZU|6Z)BNvNL^X*`
z2NT%{TPaCN^}Gc+IVZPwFsVWhD_w|dBpH6r#M9DHG-v5!$A3ZWl*LmIiZXXBx`0C~
zC;s_p{^~SfOL<8zdKNsIT`Cc^FYnQEiDw>ww~JdyTc97^@%wB4T4Rk0c1R@;?QPg(
zcD821{ur0PXP`7aF=6K(PPwngMi(OdiQ`s0^tDb8&t*o@>e_n$S)J&aWrJU!w;e)M
zmCd)l<LqC+=aYuR*|Nhul8d!ae3(u<nK`FN+NjXE0d9^ZbSwr0C2MYQQ_0*8!=QRb
z>Fw*wGHUUL6dN`X1+o}2r0?_hhcQS=$(h)Ld*dEXW^Pi~d3+UiJ<v5#T`WMD<;j&_
z9Bw~PTT^sy1=kuqve0S3aL-V2DJ6b{rKnk*$O;>b-pAGd1|3B8@cwh7z4Ha8Q@wbS
z82c*=l(XV?mLP3vZ+1jZL%YPNZhxP(Ou|cxqW;CQ6KK~){Q23Mn&be=LfF&kNSiI^
zm5jFb2jY`9e@oDsovOYZE+(NTH9(E!dU!~Ft{B5n$Z$K8Z0$aH<Q0xYd+uQ2kM7~}
zILgJR&@+Y8o@Lr#Yv!p;q`qo8QIz&zS?@y+{LhSDU%-tUQOX^bg}g7c3?Et!C8LQ9
z{3MLt8%bkiMCAoA@>f1}D1OJlDVHN>N>M`co{ePJVW*T%qdfD<kLg#W2Muxo<5CV#
zV$|tVm4xK?IKjeELD5?EM%$z_M&{HLc|!yaPCOb@ky@ohck*VNP|~>o3Oe0)9l{c!
z?Jm>p7z3%6wLB^@cXc?Jd&@gv^8(USe1WW5qCGkv4UErJNa6}ukhb~XE<{wy;%p%g
zO9kmxi<YSnVqI_y-S@8CHC8&>{ST(Nh^ihr(;t2>QBPkwfq+2G*$J{~ihbi}hc*OJ
zAO?kJAZ5U93~-=S`6Ij!T5s%wgBgU(Wi24>zz}2lf9+pg1wT!XQ?b(8qZc2}xcQjN
z3kkZ%i6=kHrqGTG%-uvvO1gVqSe`8mlMCZH+&)w`D#_*yQ)7LW{-P+`pPO8Z?#d!=
z?(OrLR(HEZPyO^Euees&jEmajxV?=d79CYH4Iep@-N+onPDyhXB5G07%U>YA!g@|L
zYjT7u;4#_!&b<$}-Y=8YVw2391C4`|LK*X}>gv<nW7pbQLygD6cIt}R8RUhiQlSn>
zD!Nkh(Kck{CSjXGUy*hE6GtOQE-HJ0^=~+rUtSr4;1&IN8It(c1eJ}A&CJZK9FRkL
zb@qnd^Cu(1##XljtZk1Aj35MbO7Fb+z*cA;Gz*R~)cDa%MBQW^VA~{~_LxOMH3hO!
zV~6U2cYM(Cw|S_P_)h-WtY-DikCYZ}yZ|%Qr}`ir6Ok4Zv5~v)E1p%)co}d(LIc>6
zG?FTpO02zUTRkqTPnMWU_eP(y8PD%9@4kA{6-3WLZ1LJ>*j1hv9aK{Fr2S^ayHf6&
zWDwf8epA=JEN|W`>t7Z(PKh<>>@19Q#Lv;W&%%msbLap{nt0Z^Tw0>Vn8NLl2MT&~
zWmB#bP`+zI@7k;<8My2<7^UFNCrV-A?f00YXWM=hlmYc9`<V&dU6f7YlAniWHN1`}
zS4OQo-CCh;a2zx)ol5ojnFTSOtct!|B3?9xrj<cSJUZj-DUOs(@vjw;cbUoNkEm`#
zmOh<}L<Yi1@*RBX($>NPbSlMttdhd$8QQZ;>%Zr-Q?k262^)<T>0?iF_m3Z0?sm5_
zayK}$Dj*t;{|OiPYiE=M!#UAkVNppKntP%{3Z&Y7E)HmznWF&c`?t21#&Qr59)1ia
z26P?%SdS{a52-;fu>}N4g>DGBa<y7pUdyj+GFB(->oyzaerfu^7>@U|^M`D=^y345
zv@kyF%sldl9NhGthAIx6O}ijOet#=5A-ed}_XgPkXG-|yr3dIf<57MJpT84!1Z*RG
zaH`wjDEku@NJ*R_$u%-j3>*}+j=Q~4HLcCq1u>o<Spg@#89afLHzul&QhS+h{B2)<
zRSKm_J53mT7mQWED>oUkzO_Hh!R6iE`9Qqg`UvR@>e8g5->X@na~*c0LQ+GDprSE1
znuUNz|0q;{Tor=18MpZ_d3%E$Dnvx|I66v2kwPBUsy_%lxCwM0)sT)F%{Fy=SthB3
z!nBM=;H_!5RGK>Qm%Fv`Ns3Ed`@&FsTL2Z_rb7^OAqecB9&7(qLgV^~Uz&XIWv)b<
zvA$BUSm^!22lV6YPxwK3iUem~o?g^m+NgFY0)vk(aKZza-6$8ypXLw|?uvz6Hp)0`
z>M5Ev@{+^q2egX6RRp(Gx7*cq-sxlQ;%T<-Yy7m1Z9aP=AS8x1dzm~7IKJO2-^vd%
z`<32Q77cxFwe!89I2+eG^R{Ahh_nUOTPS0t?Mo0YRrXDx%j;(%;a`wr*7Iy@+>nlv
z>+T3mYQBJd(XGh4jTgdraa%>ZmaNc#A&c#WXQlgup4d`Be?s+czhDOCHl^lIURa^s
zu)ow7uNtg>vpqL>X!VGy`85+sDJ-*J=VMmB_O~{`GEh)rWt2ImW8o8E0hm4z-kOne
z7*+nBusSP#X{{XATUFQ|w|pMYS%S-6<jgaD!PUAvftQjB>XroO$H*3~vnD9LlPW+|
z$v*xVFcRaU2@cp6G2vznJ=Ut0yE|7DDQ`e0n7coJ{*0FBa|5?cPmXNdr^vWZ6s!D|
zDBxbRr~Z=p1!!$~GSBTU{23j4@H+``ct&M*wtXD|SWSlUICJsn+!xv{B+z~ogI>)u
zuU|A(?T$no3tW(Jn)W&I@9s*ulRZ+RZCubShkmxi+CP4ww%^ex$YD)+K4O?37AElA
zGhB^2NhhST9+1S5)Z?`J)C7lGFl2i=@U0USpr;$A6czk?8L~Wm{Akkjy#(e&f|d(s
zsb4Y7E+?!nhuzY5Re`IyN=I@jLEB7Q^^(44wYBKI3g>{pPadBq{!5~>lr7k_?>8vw
z8;e?*=0swRF|U-a0~=fpvTY%slFh3ckhy~Xw>&-3l$J;Tx&*G)<A8rMzy$D$iLcKT
zF9TDTu&F`|+{9e{`NF(c$C>!Va-TQ|!n}&HYRdHqZ<L>pA09QijphiQPR*D#NCe>!
z*+m#%{2c1E)q1iyeTKUbR**bajiTWCL~*)>M!=~Cmb=O%#>GcYp%k0(86WxP)x-2Q
zbRj*XiV=CNxaXz&g)Bi#Oger>T4x;Uk<8K>1hXN}&6gu1)shH1`TEeU6`wEZnhRL}
zZDXD|5oP_FZJKy4ztMGfyoWn!Rx9iGp&xE11P<C!>|1IN?_?OSVyQ&;TA(X@t#NIR
zC43C3;wJ<Io$AVhn&awE+(hE>0@D4jt%fs%9o7vPdItx82&01g{WS;s+FEAcN2Ps0
zvMpp?{#xsUK{PZ!BHy@R4$;rZ%XaOac4A>ynmQ7sy;+!CxWP?D?mpvBZj@Gb&4{_Y
z^Bm7%;jss<MiNR|is+p?G<0+qJPmOF?p4#dUYpou%C@7YiNNj18<J#*VRB*H)vkBv
zE;p!-a*%#*Jb}74D;e^PEXa`SVuUwmbDe~FBi6DFWl>49rYq0F*z(N@3gPYdO)J^X
zEV&E6rqYhOH%wD?Zw>#M<!!MMc9S`#ZkI7xe4Nx-dmlIVF)Hd5_9Jw3@2@mv*Q(<W
zqdj+iFE20K`PwMeUsCbkd_aSJ-^fVW@1!q{+5K#B@U}Z4;Emmgk)~K32|#iHnw-;2
zECTn%zx}%9%}zYwhwlN&8ZJ<JPl!BuQ7U*%(M&8mQ;_~;=5|_b3;T?AJonR(wS>j<
zhA9_XGTe-*S}N1JBFR_+B_)X$8>PN2=*bddrq~bY7#1E?bsVz5)QHVC3YmUHMQB<*
zjKfOi4rJsnm~Pj5fP)`)NJPLepYJn}oG!bfTlwOr*G%<N61p)G9S`TE@5di=`~5OG
zK}7T7)jzL+a8SZ6$2;n7W#I;R6<SVgO*DcB3Yn%W`U$z&Li<lY3M#Sh3LF;g`UK0S
zX~OZN$=^z_`${mIR%i$WawaK+hKG-}_;?U8>q!8g=xnPWKd>I|&DBIMwb+!eh8GY(
zZ_h6en&75fDOcOSqCZZTz6e3b`PV}ku-<`&(4WxN)g4UXj;0b1A1yWH7rH$ZAGF{A
zBUu7F1qIY)!#sL{425zmpLPqfm1u~xosqksaThUHkX((WXky@z==<x-Lb1{xi<6IQ
zn~#QxPt$b$MA8Nu)rG1u8C30&9YL*wSSG26{FSadXjf(|1u<@fwYQ*;U^9E2GaQ>r
z;_Tr+A0s7{+j=y`qFft1=#cZOj|IVkFWT)II!PL4+P?{>NkY5TRg=SuT@CS%#(%vB
zHU5Qcl+$Z6BSZR8XPn5FZglOKtgf2Id={oVR@Ni}I0{<&e;(&pJs_c6p_!0ad7CmM
z0KB5P#*~6EAs90^cY<z}h0|IT_gJ-+dPhfxjEu|*SZ3xxjRTGaiz+tDVRHMXeJ<c=
zm6wsh1ShGS!X_glD7-%bnjtPuy`gXk7?8{LN_0xJ#>U1VGgX#&BLXLnOZ4k#U%d)O
zd4T@b@7gv0!eirbwX^wjO#T$Uobl-sCSd=c3wpn!^gRsf;WW2&0Vc+^vH%3*3x4<4
zOiW)X1icSoMP;!`J4nS=XxYF9^ycJ<Fs^g8;COYp!DY*Hzh>HVE6sz!x6SY7{Ob$;
zpMU;H@VRb_oNNwZ;b?L`1GONp)4zwna`~a>+lza$195YqU4O?hBG74kcdbpzWveq;
z_*w7$a4K)}HLM8e2&6@mdgJ3y#4fXbS!$lP6y-TlH@4d$Y?P}<fP++Co#$6}HYV=5
z!KaCXz&iw9>M)lDF38!xUkFC~SXO)qoq*lZLVKG<R5c#=`Iwyf^SBpMIRGi>$ws2R
ziZZE`_u`JN=auwi!Fbx{-@`GtSge;4ZZj`nh=XZA$Mvt?c(yvqoSQG^4b7LXsyH#;
zcgyXsNz-*OYkN8fO#;8WxRDVBBqk=tnPr1U*|&r%8aiO6mNLEO-LB}j7~S$6W5f@m
z)^sG=hU@ok+{BTr;ZMtu=Y6FdA@lEodZe%KZW04G1`<vdTt+H9k4#Cqr}XX$&w(YF
zmR)V%ROsjr0npMGH)imJE(3cgd>#h^k2<^io=*N65tGh+86sw1x|`<&?4v79@K$Q3
zV(06LzBsy_#U^e*C)sZaeH?x1f3sirXnDTbivwsDC=;|@|GFZ*fIykLBn)_$$s+CX
znxb7j2lu8lVk6HW;4~iCNL*QYj|3eAF5`;aye%!aso!4kcLS}HMA*N1ykd2A)pnsi
zW;l&syRn3?Ox|k}adoLyGe^s1JM+D6**_)0$on&pNshr{F|C^Aaa@7i0jkSQP|MNr
z?TGN5z=+Td6AtObU#<&)xEVe<JdAS)UOxX_`^%cT-C7mnI)on@fJmFoseF-lm)-U^
ziDT%Ue<&=<o8L*ukqa`s)lp4j9a7l*l}V=_M{7TfuI${%0V~V5?e_@)&DVni9={VM
z;3Kl1M1H7ea*baYxr1o@ifIw64IUyLe&bPC2ldUv`db`ylDFU;AF9Ek6h!r6(UMIf
z?02Zq;c`?urH5-a6Eeh_0{LgM{G`7By&``Z7w1~G#)?!M*R?z&mTDY`_LP37syoxe
zL<J~9H+viuo+~KP(cFM2mikKa8&r_t4F+bZovZKHdt)ew8)`0BLhusV&A$xHS5;PS
zPZkja3wM^)c%D*^z+w9X-=i*K`-~p2wXMcpTa2U+0mQWM4FQn77kI^RDLo}+0O)T(
zW21gkV0$%RFrC{dOli!qo&?taB``QEDJkIss&wl&1|`$tnRC8B;(Eac-om6?85&8<
z3i?LMdjGZIcUT_VmHToxw>{3g%|RO07BQ7*PvB}{KZZY6jt<v$2yHXQ^T2!bwd?Mw
z;F{+6G>pjGPR;mcH~Xd#b^Q;tU<Nb-dd*%jq}_cL)qS2_=}TrSl<jv5L!|eBiFD!7
zn2iO>*1$kjAr7hsJhCQjDf-Q?6lVTFK{}I7Yfqn;{>qe=r8eX7#>@cz50oPZfo>Rr
z(M@?xTSCHsz#zLJ;V2V%j|#{f1V7Rtn_=W?26KB^i|8)wu6)+^K4}1UwA}B;9fFQa
zE(4ft#wbO3c}7D6`_W7kWB)4}I;GThi>i38XG_&q1hTTSkYe4c@cYF_NI@+qm}c0d
z;gSl15|{mg3_z;xO)aNvYU=7qye_oB&hNp0L9d?ozC!2nbX>V_aPW_Pow|xjI{>hN
z_-8`g-Mhh+;BD^vbB`kL?eDbI)FJ==nc8xZfrz#S2zSO7#4tbnHUB)KZzhtpRmRNh
zHRxe5+(`w0qF-m93#w1ZeO(QDcRG!Usr`jowMw3f9Pp6i0-!VkWMp19*t4zSv<Dhf
zBDS_n%zD)?`S?iV{z`+&^7~VfY_9pFJhiT>V}@ld)@KytuW+tI{Wa-PG>l-+b;=2A
zDOu%l94JakEzP`I?+B<hc`=xm5W15klME_y5JV+Gt8S5<Z5!8N&+1Z|vZ34MyJmLf
z8_zf_iv>EmE3xewB))I8*!)ar-r)ide7Hw{iX!wbq1f*>Q9?6v36mEMmdL)5crqVB
z)4u=UrHu4*4tkVzGC=JFxM)g%GuLW^q&X@0bI3y^=^3zpAE-nw?3<dna&Is4y*FhI
z9cB712SFhmMzgE!R}DWpI<j4APLhfu)ok2vy*XF|!v*d&8H`ZF<Da$j`}f%ch?1j)
z>U8IOv+CaSplLylxC4(uKn5H+;D!k3PnbuMk(H%4++zh9@o$|2=z>s#XtETT4}6KF
z?~K5+?0V!T!tSsnK&z0bvf~6ad_2KZ{0!e?EXX|;ix#>r7~s_{(fURqC}(TC5grIy
zsdASCPX9`K-61aI9f}TiQM+$_eM|Ic0vFhl6yDj+sK73<sjSlgozz0}zK*{QVJf6>
zgO>gr$1vr_-QF!|U#+uxqUt@Pn(FSNzSu8sPQj|=>Bb7EiFKX0J4z_Js7(1>fj8We
z=ijT?(T59sA~*kp+XLO_0HMekNQ83iglzgIb37!iVP>1y>PMa>S`m3+tWAzv!*j2Y
zZU5R5bQqD<B<dy3=#g}viO=JDa1uG;a{f{G%$aUcA9`V3qMNF{Ej;Z3E!UB}$xn9@
zr1grPd+ljf&$K<oB}(LJ{j9l(WX$k1jj0e`7N0&+cfdTE+l$M4g-?O{dkil1DR9;H
zZ`+?X<QXgSFhZDTaiwO~))%z!K$_2{^z9jQ6NfC2=XGrrW_?dVgO|LLQprZ|$bWkQ
zZ~>zaskWVc8rdc0<ixf+T~;<G=J6S{>Wu7wz#R2WSs3vH1SXVnv%c6zxlW+bLZ0{z
zG=S{60-~4ic%o4KYOiXfe16{-bdSVyLHbibz#Q)i2uw^Q0UN#-wBF1#y4i%yLvI>M
z&Q6YM^v1YUF{yli6_5xaCINkm74bmmMl)>&Yt2>&bz~$wfF<lo|NW7BJj>m!?WK3k
zP58&2k<Pa1cHzwq<5q3@b9wgywO2viwpyH~;w;Pp(&4}FlvJWNNE7tQD-tz_Hro^e
ziwrOHovlUm?KXlH3MkNm%0I3;7A#YjeA#zYvoy5IifX#<9e!7eA?iHA&|FZU!#vS?
z!Et+rF`0QCxxDgTN^zvv`}Aifulq3n$r}et(%TK;w{BR#u_ElCodIVZ3?I=F6(tm!
z?ksUc<i3<t#@Yk|9F7Mylaf>0B_8a=rPE}_b0%@?S+}(T9@0GxI$+%T*|r?byG||2
z@9fj$)fGDIgqgtr1jzz<u$B0~eEE{P<bJ#^yO%)3guLt#e){BzsE&@#4$bS=q2Isb
z7i%$U6~zNQ{bT83FsTwR7yQm*I28-97>iPE{)j{_J5mr@*Y<S%Ffi|XXy)eu>162b
z2VfkbJ-^dGiJdcP3a>x=YZ4L?KkV^ya&msO;tqc$<?P@8a0Z*iRw=Xen)EaLdVO`7
zJk1V-$84@ScLm8hD<>ghu6(GBm8xIS`R}cw)MX2g0{_Q!?5jy^gv8h6xQSKI{pa*5
zMD4H9X57Sbm4x|F`FNIZFZ6LW)&qiYNJB$7Wzs^w?PpbLJhGkS*L|0w*_V0Y@7}WV
zf$O)SHo6k+*Tq$@6S%WvCDt}!jL9hdu%F@a@pbjlfv?2kPj@3b1!E)Y{cYV#H`d3B
zf&A!=%c1@Bfstq`!Wzs{(?s4?cFl<Nx}&#13~-Y;bGVrpS-WzAxhO}+7_KfzUV26B
zvMuU&7>1k*mNL@xu%!RE)lS$~%iWOJCm?-H;L~~5LdCC)2I0h1e~!B4GEtGYDBATv
zmH4`UL<duJqmlzuJ~?@L?4U8P_3|G!XsYT0?lBo9rP2wK+S}loAq$rjS;q!|_#YWB
zyU*L5?#;$=FZ;%owO)UTA{*fQ&TT#MDL=zM8H3Vy)cm#I=gUjay1JAhi@7?75fC7G
zJj+cvet;Ih1S3}ZmNlg^qn1Q|4;BchOIfHl1<%q!LnzvWO+W*znMDfe>nm$o2&@t5
zpxll3tuvbRNUM>ORBMPN@-WXBmV)0X?*!ng>Pis}-Y%ZhgP@mhN7%+~Sp|)R!%}y}
z(&ZMX2xbja2#Gp_?G6__4o}-w!|vLD85e$yK;&z@HbJPwW2r34RG_aOPGdQWXMCZd
z3}%9|9m!7@hLwq!uxSrwj*2<tG>#s0x6a93Phf{t40C)7E2wqhJ=_$#@YB*d2r5hU
z4d)G<#8`p5tb8t84Y+#{J)$xDzy{A+kMXo|<_-MY^qN?De}`+$rSb88XZ~ti%BgTG
zR(-B8pg{S$H3uJloW%#CcX~5vF}kF#r5P`rX2aSZgJsf-xF!LsMs%^>NOvTB8<S*`
zP%&WCsk8`KVqheZU1~eKAkQCNZDqKCc2|^gz!c|4nWQo6XiI|xeFYlQ=$V+-LHTqG
zuz~=<>FZltC;uW@pAr-IBpCbeoU>J*I&eaJb~nggEqS*>TUn&;e0?PW)s0`Hca1jr
z=6Eqf^ULL22tzQea|8N#Fg-4v@>jD#8d+%r+I2Vfmj7%aEXCU_m}Y*UJH5BKnQBEr
zG)6D==;8#n=j-VwbnEe<OM$RS+?-!QH?yDic^~P+HJ^@@y7p@Rl@k4xkfw=OvPJ!f
zf?K2IskxroNY#|r9sF}W9fH~o%MUT@pbGK5Z-Mt_g85#tZb`qb#dA+EF9$BN=%1Z*
zjEt75F9eHFi7a04j_|STWq^<9mhi4y9;h^go#;hwG*wTZv>BLUd#;jR9@?MHo+AyR
zF^KO<-RDpG{yKf#7V-P3o!cB!;KhADIb}&NNI*NHzEQ$r?P>f(Y(f#(1mD=*-b4uB
z1_cLyp?m}6hTLiGQiIQ(tE0Y7#3l|EIv*DsIoG<QK3t#fn3<a&P3b$JIKo>loG;6F
z_1vVRDco)XD(Cn2mdi--42{&=e8so!p@ScJrq~d&{4G4=;kKV)Ge{9!_geCV9#BMK
zzSA|DBSxWh#6@AMC8sQ#eg87p-1Dr+w^itFBUsI`N*8{$9L)u(Lv2O8B)ec2meW{n
zDe1KKt(%VRHv<OT*UW4_g0^y;$SZkvY|{8hZ)}MVPqU0ZMTpX#c`r<i5OYB1(LFS<
z-O~Os_%EAkDe~R<`>n)ZJgt#-7qT5wO%U`_o;`ze7uZU8?DpB<X)(tx%E)5PrCck)
zwif+alcUj8f*?1p=w1fHq&cq|I-50OWk=Yuu`scMNA)x)pm{th2sigMl^!3YiuH(T
zc#m{@nzr?gj~p+0+1*1gPVF60Ox4J+(9pxx2ujrNX=#RYDzy@zpIWXg;ADu~C<RFE
zn@N`G^oVP~uDJ5LgYHRtN=6!*+*5DaeZ<qDl!VDR#Cn|af{>EGzrWR027t;#vk<A{
z)5c5K)`xfhHI{O?8p-wf+YC*Hk~3COZ^rcpn)R_qK7?8lJD)|3G`~ZRT3xKMcx<ax
zb^E61y$NvsX-#^2{>pQw?Jys%(XupG|D^ot)7Dc>nsJcLemp6F_P+J>eI1I{MDfA7
z{e`xc(GQ}fzd}TSzRX)Hl~Y6P^IdMG9m+g?_gA-47E=%~UkSWg$4Nh8a@pq<JnP%A
zh~T43KD}gO!J1OqvP4luXvkgf-2tM4M<pB1a%M1cxJ?vHG_^Zc7I%WJfSX&#njfRw
zxs1W-a+-3YVwUk_<+JYLkRC#}vDJe8_u=qieg!l-mJaxa&e7o`@|dd`_Vugi|Nc$%
zbytlD>1T2Ie*E|bgk=s0r~sc7X*GPz+4{XXlp?F7)V;d74*alt5`uz@XMNkCj}=fa
z(c|+VC3}KSYx%a;W?lzRPfsiA1>hcyO-<c9_DT;xUX4BF2jKMkfYYSv;&fL*Ms(S|
z+?uT=Cs@*gM1~0vn&(UA=Ru1jh)T%B6D;Ps(O+RW@{!?M6m)%q6?02_eyMk76Ee1h
z`nD=Ezis9-j_+@{MQ*vQ>`os3K2ByGaO5p%>&zsI!?o~qU<oSykotuQKHhjN>)S+f
zdBWos%j(O>uF!%kAu9_1D(0E8G?KU=MwnR6rfJvVeQVG>;kiJjtM6THTd;IeYFqXy
ziM4INuV}T=Xc%tux@9Ov(86XCb<M|{LcMZ3KoPG@DOU>=gAK*30Y%MJTJP1#tmC%g
z(E=wd20W~U!wqVbjSJrZ=b=f3&OO*z5x`ynhEt{%A+uimd)gR4OZ_YZ6@Vd=l9Fs^
z%7eI<JVQK|yz)CffM5{w<A(%*DLUovQ7ng3=?za=-n@B91D+{FE&cNw{BI`$Lo=E}
z@NzT70m$?8e#rg#Iu6Gb5iUD5`e@J_ovB&W&pB@7vtKi|3o68Wukzra0tD0tAgpYB
z<3D&0Vh-dX28Ii0z-~%Y)_i`{NOA<gW+thhJ8*uRfM*86K|TACZe&fNy|eTEvF}C9
z$1}0j)toOl>##>n4`Mxv3c5IE&6gcWP<}Q3{=^bbj4!Z5f?<jKk&MhNyjHP4{p7e(
z=-^#31|^+O>qwqPViOwe`CsJBk~_Ql@+Co%8fh+{>8sU>C)T3yKVF*luecs%=k6uo
zyM<4U@z5MUTJ0;NlH0oZgOvt{FX6wy#lzL4Q5YK*H+k~zvj&Z`Z|(<!4S%fCM3q_5
zn#J~WDEA-ujbXG?rd)rG)xRpMNj_}#_fZjR?WfobB5MI<g^NAQ-q+Ey`%nq<XixZ%
zH0ElXee;o&si|piZ|@2S8LNPx?Y7oz+9=Vh0a>nf1^9Q+-oqve71H=7_t)ZC_0^a1
z6jOk996WzBRc4&ATh_L;u@3<MI}o)g04dZoz(y1q2H2m49lLvew|3s1kvZuE9NK&Y
zej0p>!Aq&Vy)tz!rKSIYCzAKru_pFsxG!P%t_*>L0i{vnlpAj*C{XKcsHU1SxNFp6
zULYHuwHm2A0fE-w8-w7<(uhhymC^1e7ML;fmW#F<QzUZ#uiA~vL*0or_|y!6>5)64
z&GOY@sQ%<Dbuo3_TjK0et8tw(p}(C$0TT30Pf3f}?fAG)_Cdum$WWr_c~cYiM))Jr
z4)uWYG54vaAO?%Ju!ddp*WH7Y!bwuEFyO=9a>^tcsH=x_=&n_srvK`#$iwdl8>oV`
z-X0G;s#^voCnm$jAKTm841&m&-y!!jHzuA(Kr8`}OaM{4@XhCXCx4N}aT*MqVhY#a
z%G_OoOK@(%@gD-^`n1R{Qxg!W7r-GK_C1;rs%7$S0Ya^`_nt{TSvZzg&T!lxsk4<?
zw!;8|)mjt==9?+x;qe*tR(g$3&%^tiZ)S7#kIghzVvXMNTCEl8ns-4MfyWlIBF6E<
za%EW=te_H7UEj#KH~RPesMcd(5uk+I4Y^Wqzq46M^8#Y5p$S()yyKBYyOYZu59?`T
z9?nX*e3Hk35quld&32PjftF<X!BPVu1u6RUdGM}Jkw1sSVzwOXRSE~O=7U{s;*J5~
zjtQD_+l2FU{f&K@agVm4!t0FAgjX%T-(H0l=-;0m0PNL|H4e}q&>SUiZ$5Bvz&coR
zd4NaOVIu12?7X+<PaC-4IJ92Tg2K`goK``2U{j&vdAS<F>vPVu?0>@yj0vv+Us5%z
z04Rb6#QA&@0KY(i%QtLtx8Jg2Px6En=Fe1`V+Y)>QCiJa(}M*m<ad2|KJO6s^XFzo
z=yx#EVDs-Sw`JT@p}^S_k2H>rjlB<2`N|pYDdz7VQX41004p%G?l^rW^2s`p!x9(3
zbA}@AG8V-TPsNHl^EE%bEdh`-{No#x*9QD)IIOqQJ|PR*_M>$Ls}TP=s2uzme_`4j
z8!j?YqoP_&DTr_9R0G}$48-2m{Ys<BD3G^V`4#4YEJn>T{;iMBK=OFSqvJz$qF8u{
zWVic3HXTCGZ4;gI44BPQJ5lafpHICw9$wQQd}|(ybQTZlM*&plvV#iQQG(6*LnP0_
z#Vl5hBH#zTyck+I+1bY|x728aGiG<wD2pV+I)8PZ#lO6@BD*bDy}5K}gHrk0oek+t
zR<%<T=qSA1$3-AmLko(2vKkMAb=(8EV|n5($)VZ*qLT~aHUN|VH(N52G(DyOH{%r2
zJ20>|n8Yz-3mw-2v*34BEruyP)%aNte`%}niC<xXBgu;Y2d$6gyqbXgC_u9j1Rf5J
zoQ=Tq@^eT|x+>kBB22tYYc$uZ>r;ESX-#IpcJ)n<9Z41`33}4m1uZ62(#x?$zmWZ%
zp7p>I-^;f|ci55dPoD66&yss6AuV!2G)gtx^tWCNQ?<iPO?94`tWXgqkQNmr_&n|n
z5YdxVp8EhtX!2o%a*@R#u-A_K9Bnb*FTJDK{&}E*y-@1^kl1-OP4X}S3x8Q={sT;q
zYn*>+TFO0aqbX}9=tgM;mn2pN(FMHQZpE&br@2cn&ax~SKjt2t5d$*2JrX#SBSI+s
zFM~i8@V~fv>!_}tu<e&dK)Sm_x;vyhq&p>~rMp8KM5I%Y?(R+rrAtz}yWtGKXT8sR
z&RI+R2lz63@0r<me6DL!GX>l=Tgv-@=>xB)YPp+2mSAcQ*`7QRogB9hD4!QHAZmWW
z;BjkYJ<`W&4y~AgQM72P*TPxAeJgBM8LY}d0deFxCe4n>>wTSfyu$_wtyd!NxihZ$
z$nWWo7p~;ZJmSX;kK-`BUea*KiGT`Ux1%eqSIg(!@cZI@%!q#n@EY0ccqwlm_4%rm
zy_t$Ff~5E@Nts}kDO~gFoX>l9f$}yEoJKpkf+6bW*U#}3x|E|xN!WPPE)&d3yB-cE
z80=HJ_z5p2yb}DyUCVt7cb~!DR1G*ci{<|5p}z6bp@MlUb84QdiFtT@{914bV#j%G
z)R5G`Yjj=Cgw5xxeqUlQv(;jm0~*m=D~B>{%!G-BTE4Wwhcx`A4sSXUTUWy$HUF-n
zq8KeaT9Gn@^saakW2D7XT7`mq^!l~Dj&XKp>*8hGnK8ieT__Ze-`Z9ej+{`)LjJVX
zEM{eAPXH?<?Rx8^!RqGb=C-o|!hjOr8BhyEy+M2KF|CF`!@;4U4v<I+3Ga3OfpqIr
zpG0ZN5Dv{1>F7AuCC1+e16?z$ZZK=?ONubNMiv+241j|&#hDKL>SGz+A1$f))7=v1
z!Z)-P)OkHbE)LidRsBtIJwZWGk&)`TOY!mXlQ*K}J-zG7KkTSZTu+w*$;S27X?-j>
zTq2n)Y#S*Sx!Itf|CYNgeMZVIuWotdBNiqcedy#~Yi5I$2px_nN9%F-nhR%{Gm1Q(
zq2r$HE-tHqZPAQ5j6T$tsPqt^5%KTD;rdZkRmGW7LO+P9k>JBR|7n5LGS(9AGfgE)
zIe}+~c;ol|lqcKRhUr*QVg*Bi4$hHNW)+dP_yxjX23UH&fMce}X6JyIH#Reb5M}d(
zeK+X_*{g@VpSFAqp|hG>0#x{rF(4_Zs9;pCwnNbd&PU+7${Sa7TW+(zed_~kk%Fq~
z*C~TGbV{jjhRRAx-;o0#Ou0*8^*}l>gH8(P-)cDHx(X`Oag<`}o;f5;iZD82)?Sg2
zKNR1VEmr#a9MItO^t4Sa9EvgTB0S4=8uyqW{IJWA8*|f2ucu-^xDKSDIKUTp+1S`X
zB!tkd$yTxiU}6=VXjE8<&T_r4u^8#luvP&&MQ^utHN=V$#A!nmEUR<*j$+!VAZ9I3
z==r}H;neSFV&AMukfph#j?xuQH|l9dEv~W*Sg^!`#{!5RB1>wQjtU4NG)re_cT|h{
zYgu9q-jIu$@$Cv!gzR+&6kZ0IAp9<F_9t3G-Fy4An3S~hj!w0m!K)7t{xbp=@|_Sb
ze-F>LD43>3^9)RukEIoBLmdnMwJU*?o;3@A>gn8erpL?85O49>LqKp3+S%3Gj7mHZ
z{GeUuCB(;<D)O|aETt^_W_?HDeX_+qMO&de!>t?fe54@DI9#g6lc7Vg9P)OXn1<!m
zcZTV_yAt2=K_CejgxKBbzq6@+hrMkZgqdo-id56p<>n}E^aW83N%pFaR_r!Z!+T~z
zE~;F5T^*;-x{?SWHZyMK-baKGLb@egFafL)J=lJXMmyEXvlC7MS!b^*s~PL6R^Fk*
zRTQ9oazNo?8bt(U$EXCqg}r;%7X5RQrd-sou;7Dwm;mnz8eFxL2PXi5{DjrSaJ8T0
z+P_$B7XAk`X5Zv~_=u-JhxA-4%Jj9h)Za%D^@N)%8u5ncO1&p!9~Ry7UCOF$m^XYW
za-?~yv46{=%9ppn_G2ct?!IGD?XTKE)`Ih!sK}Xt{_D-to$B9pfhRlLcDiXqT6q3q
zuLYBiq-mXY+`>$E5oW44G*=g689?-BwL^V`Oem6%>GCC1a%txCc&Ppx1kF{#6&=ZO
z`O(U}9qr;GJQR!fbK6;iK-opo?tN47dsFH572gFfGefU-o4JaDIfH`JZw~j~_m`jk
zgI2}D+5VC<p!&Jhk_s?FRvc~p^$dX(59ggvKK3kw-krn>`em-09?*xg4@Dn2qAx#A
zuSKd7^>WOu!|tq6@+e|obo+t`vv9XUd?IU~%U?Y{s)0{QY}($MN*!`4(v5SQ{ejgr
z;y^K=%Gks9<C^U>KT*6FTx>RctHD#}Vt+utFJtl>vIr*QO(0oOq`_9t)tZ8Hd@*)y
z=O#{{;pXx!qY6QzIK}ILj899`^=i)#4A+E(g@v1t0w5U)&D!_p-|f_lD=4kXB|~{+
z94^*~KU$KQO1zgsvQEcMZ&D$s8x?o=o~x~YA^iLcDPpTKCj_gOF`~NV6y|>R+~!l<
zG5%G`Yp_*S8%D%71iLkNw<a<~XmdR_um+thgl?*cq|JrtwUbI3keEDDFBvAfI!lkK
z<N)JhSHH>o=v`hSU}zZZn6%DX<vTVie>>ad#_C_XFaK{?gSh17!ALK5%$l;D(BbOw
zhg}d5un>H_g?kUu0U2d2@^nxgEkJ*+kb}g=`HqchJ5!SM-vs*ro0(o5)thVG69!jm
z#bUI-NJMiPbf>2e5}N%j_vnj-x{|HaPrTWszK;XbT}u!$Rwb%0TyZ=(H`1c$ZOtm}
z&*Qd2C6mBpOt4Q<<K->{EwCF@%e;!B%JCwP0<|o?XM}G}lP+SeNA-cSVzSoQ)FE40
z@m3oUWrUCp&;L4b)k>vgrgT`QTKOnaU(r`Vw#Of)Yh5+tLb`JI-zM<zIIZIUXRydL
ziPnzW2cUxr>cP<BBl{1@xY#JD$iZhqKEmbn0!dqc&z_APX|DtdBquz~_Fw#4$6vv7
zP}K8JrS~N-nSeaxhg>ZWkJIc*?(98z$jw1S?g%YH4q^&yy!C%<yI*o?x{Kz?<=(im
zU7po+g9jZqhuNO+U!cRP?Ji??{H6e?KhJb`0<kSHL6sY~;zRvI=*_MW&L2$uS5cEa
zyCC#h7@_?uWdXsc*AOjm(l&K;`>J!iyayY<^?zjZZL=uxT4Y0s{X>L8J>DNd0{;GD
zD=z;$9x(<w2+EfI(4zZJQby<1e|-Ez$J}JhYx*dUEm}+Sbr<vp8<7n&{wpmbkoo34
zJDr*<E;`N9T=<=qXqR8}mC3D_gUYjT3FKD_{g)40F#=ZN-AIHDbGd~^^r{I7q`HE|
zHNw(VT<^0xrP?jvm5NvR{;dBlNJZ@U{R<MXd(mm{4{ZK1k$1n(x<12sR|pnF&#TaY
zMkzxE6lQ=7`<1p+NMQ{uDxxc-4$!PL90P0w(t8&dPLTey{)a9@y{{@TRWtI(URBw-
zm0C%aRy>#@$RdEveu2ZykSKSJ4i<T<aaCAq^YD4%31>u}52RZfpV^GK8V#Kxn45I!
zhDL(@nL?$*#5=h=(zDU>DB!D(Qm)THRH^Xp$9>K1<gkK=9ppc3{>~?Tl(u%fEc|ku
z=1{~2EA(<^jU+!&J9axh-Rni3IDo2{cv!w5ZJZW-M<y4S#<9vFVF4Q!YwUV;C}^np
z0D1!=L&QJp`e^B5E7{1jHfhTT6!z0-vpn*|hO+@_py@tk4B*&-T7flF(JiUsy!#D;
z_vti?gQ)v!?akA5$4u)nvL(?|r_3e8|C)~J_fPL^qHTt}kMlatEL~!8!4712UujKr
zY=71n0)6X~w8#l+rqRn#Pt?uh_v;Topl>g==AvGx58Y-?a;k=Crw~1Wc3^e8Yabfh
z=%MbSgmKjdX*$d?=Uq+d1b1s}7qsOi(M$cOyjYe{fU~*31ueoRph&yD-Hh{mP1t}S
z=4Qp)A#{nQ@=C;P;4SC_Z~X@^|F(L0huz)EfgBmBajCV0hqoXpDKV!`8t~o}>KN}$
zC5y@!)!)4pAQ5}`sLM4#00X^P9M!m6u0~|V)zN+=MxtPl=TA<%3fFn8mM$#zW)-Qp
zM<++5LxhAEHe9f=09svYu|=F-%s~BbpxWu@nul}MV#_t7U+N-|?Z^%`gz#1CXpc3R
zDCR|#!K1&MQ?1EXhGj)ym4PH}vkMDVhaacbdgyS6B^O*4>8Y;k`YUwxLprYI2-2|^
znaA;TO_apvRMbIH9PGz%7Y#*{dv?+@Ako203C7>2Iu714U&Wo7YCsU=?s?jfpW8ll
z{F?C_Kes$KI;teP@?hJp*b8eXD{)|4@=dNmyrlF38m2_Zdj;Q0rD2B4j)|cM)hpjx
zB|Xd~=tAQo(UXb6^vJ{0KTb1;&&E<Z6VVi{^rE)|%B+@&4_VUVDJp}32I-)#3KSM7
z;cjdi8Es(C<?m;}hoIJNupI`8+8t<LyfRszK79f%b(n;N@;*L-CMG5;KCiqrLD^zq
z+Ji~Y7XAn*Qq^m6onXz(agflgzKLomRZMoqoiSBNr(w-cjQnzQ^QCDq&n&~*%7cCC
zVxG*|IN5+b?Jo*|qMx2sp=7?KaWO)6+#B@155Ko{nw%`3{WW_hOnD+d<4j-b1nuEI
zOBBGc<n185ay_Q2aXd6e&Uw)&<`Q!6Jw?fgmqk*}E$jEU)KE*W+N;d)jJ2Wl^A-k}
zTrk3w_7L9y>Q-m}H*L0D?_pyPP=QwJ#=#4(Cz+B9e%d0Dk&Blyllr8NdQ4Qr>=sU6
zO=a^uz>&VTQo0rsT0w~+ue9uN$Ur;gl`#>T5md6eN}3*lh%N)?9^wjb;?VN&S&^{o
zHmf5w@ehUTK>9Fv1Vkc$=wI|a8STXd^g4FyjH@#G^QuF@vmN<p$qC;mO}(g-x%U+j
zN~^zpA21py@;pBSx6^h0dh;IV^ifEowSoQDMyHnRm(aqBFh^Y46`DldN7AI-CC|Io
zeQv(P<<S&Sc_|q%4Wq3H>%*xSth0aXPw*oz;pxN9PyWGQ532RMPx?j2J2Hy8g_|k7
zjlZ3R{FN1686Tj9ypVkr>wH<<YZ@P7@Q<CwaMd$KN^#<#Ite5G6rWnk4a46jrfT^m
zs$>uUK703#M)165erLz$CJh2R#1Diadh7mD;v*llomC_qmx%X;&LSK3Hzug}76*!M
z?w$$-<3K+C+?ltZ#{HXTC0oRM3r`nJcM;UkQOEzo0wf><MNLufzH0t=v@0=nd?bG+
z^AAffsk!b9PpP6bn1a0OT-exR(}I}4DGd#vVh}ux-za3k<Np092x8%>8RMvlpE&Vp
z(l#fp(2kOX07pr$<K(YkHH^&#PkovUZ1`Y9VFtAkm%oD_+0n7FC0pH&WE2$g{a>Cr
zt#UYgKqD{2UeQ&`@Xi^Gp~qwHCX7is=ly1`kqgYSJGP6nnv=vE@@Onf1QtUiQd$8}
zFu_DXMG3sF7S7h~VuR$l3I5dC<PeQy{R6zNNfk$t(zsO$IP1TfS%F8%H4L;uARtCb
zAJaS<Y~K5s4z#!Yis}1|%XDA@f<p#05aPsQqTHVzc~#}#n6#zHc(}l{b&g&=d|7xa
zq$l>`_M5|y!ZsAGL#qa%>W1o2AsES%9;sL@gC5;7VR1if4;{7fGW=w=nFK^#f}ZvM
z{v?++(DH6pgf{gs7GW{3obaotGDkERuu)i8SXG}l2PRMS$~9Zuju2n{Spj~`(q68d
zEeuw6t@F<K<XA8uU$OwH32ItmVq!smra%oTkpC|8H>ThGuBR=iyReNvIAe(2Cimfh
z65_t^$E6Q+E$TSrLuD&G_Pa>c30wlw;N1j?pg1GxF$bhFCFqY$_}7KW=1?IzCMxhv
z4axMsb|x14->-64omK5r&s_%G=dNTYg3+!=f-DJmXXNDM1c34|i3%YUXatgxlS3xT
zy2d;461~>lfb(U_zkeTr1#ednEYceU1e2TNzkrSu0>mMD`k0ieDh2@Ox~=O8gfzfo
z=xRESPk=4G1~-XWs$dW>^&Fo+0AQJ}N>r+$<YF~nI7FuL9hixU7->C7b44piJIp0k
zFDXM^BMnXdtRJ_)A*pj_aIWzy;vlNC&RQq!1W~BI>x>4YVOt7lsn<Ad$?>}%>$lvZ
zmTdb2YE=h#KL8HbY@x>dm6`?oSBuMD&;Cs5-n1+>jDoWA4rtJ`d!>PoV`fWTTj@Ui
z%{tRo-nQ%5RLvqNq9kbt_5%U(ZlJNR`|`ctA6x|eByeFVNhOtudV$F82JI)u3+&&l
zc^h_PKW3`0p2VlC8^G=R09*mM9zbVLPX6VUkI`nVKzc|Wvk<=DrNqF%04&t{l0)a#
zD}ct!oAL^bf`lF76f{;?0kZL<kUj)$0|S#aC$mL=DD<MaL{Me(exDosc`#B5Nl^>D
z#5X@M*AFF!-UYy!pkW(MabSY)jNo;^i`dMTO&;i4-!wX`uI%|%Hf=;s9&|okr;FBt
zA6b`G=dH6Z8*%?{^8t$v?Km}H;7a`j8r0Uou$Nwwm5dC1FCn@i=OEF$w}Iz3F<<C@
z<qEj&e<OE({UiU4;lKa(5y8R^3J#X9v@|#OB+5=q+`R1Al9ZCt^f&$g{GbG}DVY19
z|1Mwo0-_3%K^E<*u07w2^K-dM_v7V-oP}3^m;d`KuEAjP#R;Cv>|KdGuA?3<H&1QN
zgX;2U(4P42_XH~P-wl-&6`9(FfqqeZ=XDML-*Lh4gHeGsgU4a<7J@rH2AFhz2F;<`
zR|@KIp%*wW>$JGM13tMTnf$I4A|j-(x1QDM-_Yb>DvyKV-c<x9ErF2Y|7SoNtX`K^
z0L7tI&MMn0euFCLzBe=ck^L3z2Eba^WPv<n3cwOMEXQ3uv%7=gIRIHYx?S`11yrTk
zfJPeP{`$z~lJhmb`rpr04;-`TwWyuB<P3oWDr5+N-q(OJf)>L75=&bbXwYQD9xT+-
zR(3w~gP>*LBfCbE-R16|m6Wvq_s2~0$JQ2BJFG@hZepb<q?o3gDji0=YcWv-IT2f>
zEkwWw3?82o6(Ns3m4HAxsJh1|BtX7Tq)`NRy!e2H{vWN@|BkEZ2>^Nbd(x3wFNi48
z{YyAwz3d-hM1NF5)`Fv?C5hJ(SJQwpd<vj`aHu4spg|%lBNPAM%J@H@A@O(H(fHwJ
z2kWH+gGkA~ap|4Xb7<bli-D(e&UdhLFltu?0DN+)%0xUbFAone?5qASLICfayai=e
zvwB@novd)t15{JBZuzqR%TtoQysRwPMb78V;ykxr(CMtl8WR56TOKXea|t2;?+pdF
z_B3K6=Ai0Cruj1I0QPx9@u2!-emh-OL)^rK{FO572|;*8KwcyCvNFc+j5<94-Q8X7
z6afWE>|<vD@dE@JG#%taPqKx)zXLM~<JyH+y@R8E17wZMo+@BNK@I)t{^eKp44Bej
zssKzFF@}H(<d=b5Ld5^S!D#$)jW$^CPsQ7;9weazWu>Mj-q{L+s4<Hlz=yc}CA<9s
z#i3O2;grcUuJ3HfDdXr+Xcz^*?>z?l{12=})e)<l#4Azw+V?t4KSES<sIkNlOg@TY
z)PQ)LcdB}`&s%PlG%m@*D#J{^##ZFUbxV$-o!^V;AmgHA;`XV*z5pzmSNeO&zO}r^
zhRCoAv}(#uZvi9=xjR{a1W=<4;D6q8@oI?$5e$!dJrk$Ka+(SsAONhGb?U6pU*T`?
zAWCE2(M4{+2moU&TVAL_mVkb!2t{Knb{A{IrCK#C$q)biL(o5ToA*(8&Nmm8L92@h
zU(d{`zjvnxzv!*ReO+t$Hm;Fa1hHQd4P(Dmb|#?Ka^f6xCLsYT4F}dMv(Q|jGU+Sq
zTWc})N<snFkcy8FKcK_+0qS&^Hv1IdhK62apgoDy1+w5?Kx2VwGIry#6B60kmg~VR
zIw`3#_}$0ziy4i_?+^e@JY1*^e`VZXzks#^aiY=cPY}q(dW@z%U=?6Ahq5~3+bWme
z-ad1j_8Q_XLmrbrnvmCXi1Z6d7F4aX9tt7zh5qIxYN77wazi6h|KE#U{=9a$mBZG+
zsw*MBnlCoZUkP4q&<;$2lsojo9XFheiwK+EU1P!IMYFDWSds{IPR^q-jZ4)#TP6_*
zqcB~dGXvQK{Icm`UWohp``JKRW$1gQ0l<Voz#%}K#T`u-V*;PIHSmK#ng{U}4NuU6
z70gE}I=U|)CNPwP0J|b)K)OT60;z~@=D-d@06@pt!CsT)xs4AL0GeBOAdybT=dkz<
zZYv{BRAu20x=2hpx%HV)$J>kpwgXi8uWMTGjAzuia^}4WES@1!v8gY$t4TY)gn19X
zmsSj8h|Hsl%!47~3q76KsyYJ7i{L<GoTC_1(`?nBm&xP$)Q-bUA-Fr;%_S>I0gule
ztFV&Pj&pBb#uYz&gP?6mXg<S1`aVRrI@*t9yUJ_`cg%SQI(!qQl^r8sd!rZhTe53B
z13u*uz#Sb#7kFU#OM`wQIOG6NR8&+UgEs*<m|>)kh0R!RJ)CsH8&BrTc02^`BM1)*
zC4c|^`nvzAlqm-`vnJr<mI3!tTU-0e1Me+0R>Hkt7vSjB{qj9s?XF0~P}hVC_qANx
zg4GCqL78;^$l-LGq5kcNk6J@kvS57he=c<zwgxF*&GiL;Xr%IJK|QVq!{d|T*4m4?
zt=p}&4O%VFt-E2Q=X2v9*Hcmnf;Yx@!t=f(BF?H<1@^`fkpq)?1?y+C+(%?ojs^JO
z1OW&j1#;x0VOHY&1N>+h1cgm{U?8?9vQb~XH6ZbL?1>-%{#OVhtj)o#fxErJY$#SB
z0dY|5T-lgG37DQ%7rtGCHuTz~uBl~Md3ps~UJsU#iX-ureWvB-UrcwXq?6;S0&ylG
zY3E0t=;MvrJvtgX=A;q-bl^pV+Y$KHX;Um%GRkEU;tR^f9w+vHE46Ll>}^WBs88wP
z7RgE08=kj8ZWWuzkb}!(@m?!e7qQt6-6*?rnnKM`Nj;|8o?W-M)U8_&C13|2v-+nk
zXW@Tn@(s2MlhZmw<fpnFQMa!&MHfDBY=<Bd7VhZT1jSo8iR+zyT%&(rdP3#^m>wdU
zhz}zmi9u_!&C|KTc0mj*(`6b$tlR-&2v89Psckq2-zI@~S(ev+v3&Onh$_Lg2g|d+
z2ix|bwMMsqBmxz?^SV#P*;(;Ufv#+^FE=UIb1OsfH!3>jzyj~J31{-ZiuGS#XDppu
zPSqWr;6=Sl!6JiD;{%r94VNviDubGHRQvfybcFE&?-<g}2ZNRLxDA6?^P!U|G@LVU
z6)eej%D)4;wY_hgIr129MXcV;lC(c^m0SNQO&*;fpV*8Bj61~dN~8X4K0v>NnB94!
z4*|^CK@dcafZo;{WbDph03xCj5R7E-InP&{!FN%MK=cc7MGlAn-H3#wBpXom$hRow
z=H@<b^K_c50&twyF@W^tc92tZhe8rT`#fs2^Jfl0<%hP34vrcBax>DjEM)(O6=&SQ
z|G6|fr)`S1)@k3I`aU?0<nCbwa0)gy{6tTBl*xeqTH(pBRR@+p@lWTE)o+W5c;pmk
zeIp%)Icj-ov>aXv=;VY3P09a#7eFB3!pXo^g4aPHaOditazOA<!wc7+kS6u*Td39G
z$iihY{8v9`2yML47om~m4s&VbeZ&QhOlSsOI~E~eoR|t;dq)0GT;R=-`pb+?qxb`q
zF*t-{ek4x{81tE|_k;omzEdEHnQC^ntW;3>fVDbFHMx9i%pei|V8I;8=Y-hqoH2b-
zbzS8iTk!hOA~%pa7EMQci}JBLmJc89ntEb)yG3r0ZTY=C+rO*P{Y9AYW4ido%<8^O
zl(>ogf^In(RV%yZ1zs?!=`gV9f}iPgJN+ZqBGeEF1i$5U_rA#z%=|6={Qfk=CX*H(
zDjGyqPJ_hm5Ud`j#tjauu^{}iZ>$B*7-4VmUlI6s@7^hy4Qd&<Vna*_-4H>57k&BZ
zBWMTHF*9$!D(fJi;ohEZ#i==*M#8@b)l2XTuh4Hi1b3baRssD7liwZ1c)bL<dK56Q
z&q9TwXqfBmW_J9Dy<gki+gkcU1-0As{=op2k?&74Z1f&nxv7n}+kaP{`eTpc$emS@
zALy2HwJ@TO3`N@f<m{@pQp6#;IhNG=q2Eyr&2f~D^bnsOaq~FW%dl@+CnBP(=-<3s
zp?;_2CxVNd_Tfh9{KyAOvUo|`bTc0_2ZaNiGJ!(G+tuQF0Q@bLlXko>ce{78{f{r}
zQAi=+lM-RT;_AFS0XIhT3|%Ed?FXyjT!jH79X<U9P%svNGhSZ%W8ih83eNICj8xT1
zS)RMVlTKWqf;Pi`SyQ~Is>N-0FKg)wqdeWUT__?|3?CKbO+b*p?oxAJOSc<<LVmqG
zRl@nh(><@VOPy*`C8K!ng<Q^Jc7NLjs-^>e+>8F;kzL@rrtVYmIM(lT>y#KSA-RAl
zfAe0w=*-OBL8Ev0op7&pV1&k|o)SWXn4Ix1cd%Per<|TTzE+EIjeNR85HP<0?m|#7
zgQ2qRE2xd90KqOgG4au5J&16Vzd``X9T{(v>5&(RWg?=n-bLv-w9hn($Oz?05kr8(
zPIi9hijQXSC4#d&GC{u>l%@uOz5#@vkC%laTi`T^9QTtIZ7BU8avuoTeV4jIvZ|~K
z57^v{?;&=lmmQ|qL&i|_;(BA%+4_)C17pXGWu&e2Cm0HV0Z<=}<G$Ef+zeVPzgt#-
z`*(hxZRjo(BinEL-QTkOGSgmHHQq*4jOyp($Yn8)Cjk%JJQtr?jiAzMrQ(?Su`tSk
zkCE>&2>vd~SqAGdepUIM^VyiAqV<;3mVe(X8n(%Xiqr02rOvy^yjz^`)JhJJrT`O2
zGoVlY8P{sJ#GO2$!-(VLY(xIbXt4OKrU*>yw$mOYz~)t3{h=6$#)l3VvoHf|_wr~d
z88D<P9-GJ@R$}&qql?7Q2GZGOHidu6^i#Z0e%$=0B`*^evi(b|o#EyW#)Ak89#8JX
zg6w+q>-!v<49fyiUrD$M6rm5&7^=IaOP66@n)RD+`Ue(^Db!y>ME8u*kh>)o_fIW(
z`B=jG8jS*r1d{LbBdqUsS-+ISW?oIdb>1WlC*0kw>jlA+tT_TSK@{vUA+*}PXc6XK
z@z0pG^>1Me&BfrRO*0#wK|p|_!+J?+#m4IwI1XhrjVlU-0<m><oS<RSD~>AE2fozT
zxwzAjZHQT`RKA4;z~U_-htc*{s5x#238r8n8DM6ze6&f7Ch~YD9CFI7*U^6aqD!7u
zt>t!gvivA}!0TbkF8G9vm8AZ{+biJdpyV3(mzG>;aDIM6RofY@f@^;9hzT*u<j(I8
z8}sU4_Le9wOfbFxs?R*6!B^pu_XDO02b=M+zf0?Px;zXl?}Rr+pEmIw8HpFEZ(Zbs
z&6Yk7v9JsKS=9{H+@@nB*4uLn2y#LS8^FuoXfi2$d0Xu~wWccNdbQYKa7M1lps<Gu
zV$1PFM`om;;2gWQTY0X&SUqresaa2l<??_`QSkD&IPh-@24U630HMPvSj-yub0r!K
zB>s<|U$w|wb}HW>$Qfbo8(|k`uN9UrPG%~CUy`f3%IZyna8L9ND-3_r<J{MWx`vnB
z+;<b_*vOxZzx&N*?RB(PHGIakr9#lwto=6fm-9a&#uZr1hn~7;j6+o{$xmNCZuRQP
z8^5GR#ua9T*N5j=Vk~<XT#OIg_awMDZ=z<GrXvJAntJ$$o_&<o|HkIeVD!7J`tFQe
zbTZW2)ghr?;TM$mC$_O4h`Y0SUw(cr!Bw04f*8!eKYh*kI)k>@g~vqo7;)g<;&p-Y
z6Odz|fLJ0AXkL=DL_nwFe~MIr-@airE`9);56WasKK&x4+b3_MhKU`u5?45N4wI=y
z<j0Ts?mMcubes!JCT>28yr~E*EJXy5?#7K>JbL}h)#M4^VYACZ3p)bBExQwqWP(_j
zN*gSH(nWT@qYu==(L{buw%kGR*c7<HUT}@*F?HUwnS7^hP;+jwVA$;EU!ytukpzjC
zzWNyL9~K^|h#iDPc)B;L16xk!SvAf`r2UA@m$0+H>kZ_r^$zI1iHHbD6~B5E<$O0r
zC2uDPbo_+`<!8;n0Gt=<g*d=<_rLKwppK4=+yHR>*6SCjru?^K8G6SALFtSE`QaDI
zC>#@a+MzPK$e_CEv~GidoDNVyN_oI9wneRZUx-5y!&&6pTSlSGx`6(5t#Cs?U_)_@
zyT}Zl@pO!PZ^i13(|h?FnX1k@eKHmXBjBid-DRiMHGf=T1j5WG({+@Dy0gB(+S6W=
zgw3KjD$cL5@IP*A2`BB7AEtTqB;_gvvthdP{aC9TtGT``xax<AMu-^2V=<hI+n{)n
z2;_0p*={`Wy^~YCvrisUUl<9st~p<qSl~a5&1G#Nk*}Z!iyGcqGPDAy0|Z|lFFsmY
zUH};<oBeXrVoF#ZoooD~>8)2(^V~n2dkf~kkbWrC=+<SgV`~wiGRntI``f2_3_<%Z
z-Tnb9H1)X;4c80m^p5muP?{vw@{Z=7(7X0JzA99mat7*Wi?*r>4f_H9J6j-l7`-Ev
z?=SpUFX=R<O8H6h48?6NynSGhg^R7B;!k7DSv9#R+^9u+2r(ORR&BBMDe1pz7VeSy
zzcPEt;41AZAC8IOGSgZ~rwU+~g{MU>K48V?MNG^B)o7Pd2X<xMQE}EpH}*rp<f=xd
zMX|y7jca(ju=(pljhZe=R{=Wf&A~X!$J?{~!a@#EzFW+Q1s5AIG21PfSiZ@I-jOPw
zgN_TUYzXhn-8=Gxt@pI~LlV{fWAeR2q*X|Wl&WW=Y3EM$QMz1D7ir2B0q?i%bWa$>
zJIR_o&K_DDYCO$!uhrt6^&{#@8l$YE;?pbcrcbQV2NgRxdD9`I`a>?{3ER<AxcMGr
z6n*Z_B}`#%Ar6iBiDK!$r~Kyz$E5$#79Ghz_Mm$Am*#fD<M$5G%zb7{*dFjsuR^>1
z%f;$1z__TvMvhKQ(-~FjEnY>?L+_U}O%wQX;;!-b`b2>8V2zn+4c9F5B-sgZz0>@F
zH8~=vhE~{+>0s;Lqg(^TEU44gP&tAK2nb-``q5CBONWdj;JDGY%L}KMYv*(1kVBtn
zV9aQ&mr7FJCX)P~nqFbi(Sd-mbK;Rg_HCV2p6KAnGPin3E=qRd#K6J;4Kv~~=|QJb
z1XQBexW(Wc7dJ7N;c%)J)uy!T*;cfE;&)h<ac7~`N!65kDf-=t^RGm*er)*`LpNmH
za{06b)>?6sj;y8Ls7|rdgE{JsXo9?k6+*|H=iW5k)6dW8b8SaigS@~S*5M`tV)xI8
zezF!>*+WpUv6D#3y!1!a8l;>Lyj1;x^|AVe7g2py{FN%lY)2SmbODJto{|*En(7px
zED_Z$*RC@174t?7!j3ibf8C`j$Y0zIW)A)hP}a<{iYbxX^1=0fwc)pUoh!I73Q+Gq
zd|Nd8YjV3PB~$rvW10cMzL~=DQuZdZqDbgdNKsWJq-c1w44H#*o2V6<i<RzD?Hk*k
zp!FL}$vopTd$wpdV@S{_)2n1P^3=xM$QzBhDa>-VbENwFc)KVE;Zrx_ThoJLoIu~^
zXJ@_5kdTC0%L5F2unQ74^441X7WaNkJETdpC@u0+&LhUy45Q7GQuX(!sZn(|U3v8H
z^oXsc8DfN7&$ZBY4yBML1owXWI*s6*n1erW7=04s!hdN+Ey)-dG=pddWhE>%#2oKE
zx*tqYyF*5=&>@fMm8NX^?c0b=neZ)1)<o#Sfl2S5$P2q!Z44tV7u%+rzD!zXT;=9k
zT|r;SZTs)OMxLV?zaXS1K7reUMXmA&CLoW~QF!&uMFXF|{UE|0N`DV17yI>9Prlx=
zOH`aD%eu#lD8;JhDq9jJ8Og_Z6h{r^H<jTUfst%t!E4*ytOeOVb1c9+kEc1+ANmEQ
zrQusi_oS*>cex==LB#mZh{dT-h0loqcVae|O6kCu=Kc5N_H)`omabzpIJiYg2KNQw
zCO_k=l>=LyLQxC*nSg;i{aa&dB8A{qY3}8KaMjL;H`tWkKL)CVzLt<aGFQu<byvIm
zDR)@Zj-(`j)fOWsC$P|`o(mQKq*_ggOU-|o>$h@b-n=K_7}kEsrTDk3KdAgPiY-si
zaCe)Zln_A#-~9L~DKdX&Xy<TLhObfO>-@vSmy>UW7yG6iE*`f0^n<n9R3=<yn9+MQ
zHdSd~y+{nRSY)F{Y*OnOJ#OF;Ojp_SO;#{*Pq>BMz~D`C%EXDA1XeRl+}d6fG7(P5
z96IC_WNCf)7h){n%!b?R%=5+mlbz}H6Q)_d>S-|x{XXncLcrgQv8r>3M|xP@J`dLt
zIbupdi6H)ZMfH@S-ccU1Y&1exN&!LvdjOf%D3x^4YXwmE_ZUz@*P7)Q57Xa;=<a`E
z{kW_je1;-tciMQuOo-Sjhf%)F(m979RC*&EiLt9cryS&$r3=aU<wenTK8v2FJ_$Ov
zTQq)BjXnyxJe`$Wu4Kz5alZ^Z#sjlXK46fxS0p)VRNmaH?(3SrZNRYBEVZqhhfVdx
zn!jfQNwW3Xo;}9G!+6(fXXQM;Wutu8^OS3qwTNJbX~#vipd|I_fIG&cJJBls>QT__
z=Z&Z7->kJObd29ul^>iAL~H*>a_!l_MzlWi;Al#0$unFxiR(e2x<CD^l&v&~7RwCt
zV`jtvb3G~acnoU7yq5t<f=B%WRN$wl0?aPrb#hf-Tnn-oBW)=ERNAoUaqdB&s||HP
z)~tF+6}J25@IwP|&$qVp_H5oyLmay{Dp^6Mx7J7*EylKwT6#G}(-*vw#ZfNGj8k<t
zzv*3n#qU$|Xpx4@;qsw|eE$}CcC05q`U`^_Ca)()qwMUge3OiQ`gUGKArQEs)Iomz
z*Y=6^@iYD7wl(te+_y35^&-)|c31&}2c?X$ARe6N<bHvKcRSjjk&o4WU3)5N^4^-N
ztz}g5!8)Fnxg6X&y_hscQ6=NkPZU-TD@`rxhJ)mTF6$2pgyg53j1kVC1^dPqYzwwC
zG}Rwii`re7Jt&6gqgEK9Yee_>R6iRS$f8J|(R7QR9$@_uc(&J?8T-70<GAA}!?`<l
zb_X{>c>h(^!}8tZmQ=#r8>~g@_ni%<^gF<fjb5|-4zpIYkcNhyWU5xO#BS<Oq}NN<
z5Cg{E*Y~@&AFzxZ4MhIjKHIhrH73yt6W9}PEa=bTf6_VaMx0=`)r%eG;+GV#D`Cy0
zeY6QZtG1}p{wIw0X7bxr%oyBpJ&K&1$&M`EWc9vt?HnOTtfCOA$YIj>;C|I!(OEU_
zxddX7S?zv%G7oPWvrUaqA+wQD8P4us0(+u=z7t$gus{Tqmo8Q`!dYCGC2P-o@P(LG
zu+z=hTXypG{sczQKwuRFIx#O@q@+dPU^gOT``U!VgfZB6h>aS=)ZFLPvGYjys!*T4
zK?x#+i&$96nsOB0sj=uxMK%R)d4HqG2S+oYihs7YXenA8`}(nC7!Cy`rm8>jjXY=1
z_q888e?TbXgK1Tcr1|WU0uvYJ&Q2q=9$_C*6VbR7<H>VrYkt@iKDow)s`2fZHSwRu
zx0a#!SXDg7wOv`~ivPm`l(`IQ$M0YKvzXx_G-y(*dx>FnNQYq=BgmU|mvZ?EBWtd<
zF>>ip8kQRvvezD%=&fLwu(RszsYzVq2i0{b#LV=o6#*js>HJ2JQaZY9iT2B&sJU#9
zYxsJ*xd!GPcWay9#5)hiHW?ky#H{jv9Nc2I-!3-3X2*P2`ZHzEkES=+9T%?U5c^s<
z@c;d~r>l&n^?3Fc{UYp>=-N!o*S%BGo!RLX&)=_<M@p$5xc$uGEcyM%&+e`aC|=(W
z&wttB@}NiQREqT^LNqIbX1cScj5$7VbzDfSQA6bQNZj1{6M`U}?T;(k$G{H;)H)dJ
zJ@3!EbUZ9&HQrT?4NB1LRG@RsjoK3dJ(vvjk3ZE}L+{Y+HuvFPa`O_h^QN_e$|%ye
zW%<YG1d!ER_=lL$gznuCRy<GcgZiOv;zD(7wptAM*sLL?+(VbbU@8!N#rqCd$ohtF
z)l#?f-Bc%j-R@KU8C}_I^Hf&&g@Io+;)k^q+dHRMU~`|ga&91!1^MEF@yMcVg3#8C
zmzBj#tgy0?wo<j7Q@uLeCp<neD7Ajt_i@iqyQ6cXg?%1T3s@%Xv+_u-fO{r)^aJ5_
z8P>&esAKs=G(grh$HaXPSgXLSV$vwF?-9=t=fIm34LUw9QP($-GW@|l$YFd9SIp|>
z^AjgW`fJ^lwS1W>q5Aj!(-48q=tU|Dm}5M|In?+7Kes?z838VL79E#=L~Wc#`G^?_
zT2zF$@B`feZd~b`W5MgEd^_9UP=bO%sQ1rnD*{%4g^*A)x#5h<yFHInV$=|x!<h?m
zEqaYCp_Firc$PGwi&Z9kAm@792MX)uH@6}EpO}q9ZwF1wRr~Nki{AVW(@7f|RPc?D
z=cCa#aJc9B6e}|XrPDOs=U-J5=+kQ#w6#8nR#TxK6Rc-?2B}}Jvb`17-SS)Rx@P@Z
zday$&zU9+l5`T;P5m!c3+{t+k7Vq^I8R!0Yi#QR(JK*Z;A*I}|k9<Z=A-!mZNiF-N
z<uxMza6b>J+$ogwZktUesPGM0Nz)_TWPQ&uQlxKg4;*8L{4;%_+CC*H#ZTR)f$n<B
zSSCa@V9(v^^8T$gi!2(`X(@3u0fVM{+v^LzT%Qt@zb3Xog1vQTT#ynu+xJ4oC=vb6
zvOHAdp_E1|(RBSdVZaxWo8XlHv$U$MF||E*zi6>yi6xS}1P$eH{w)gem^*mB)39i5
z-=7+c3RA}ZnZ_Qp+D*OMpH*rHem6H?430T5>j_d`XNG0RvaevX*VKL{4Nw{@<>9gP
zshQOd>|Y*BzPdohHqiOEa|e)ew*mbHjceZ8S#?FnFqrO{b`v=fvlIMJro<NCIdIKc
z5{wvE<P0z+Yd(P1-WiIYJjDEw8zN%<Hzq+rqNr<cdbx7v3u7JI9qeY(xgSX|P6rsd
z-8JvItgU7{4X$vOKv2}G6!fz7?p+d_e+C?r>1by-&-`a7VM8?ymiiA(bw{y>s!wFT
zA0l8>=159zZ@X)&Ognx2kBNUo-7b<X3~w(JXDY;^e!R}WP;{jHhZ8A$l4M`98jVIm
zkp(NxAIbJ<-?P{rhBuLVabpVJ14V59kn6&rTDM?BSxZ-rEJ2TMrkvqMOW&P}C%pKH
z8Ugo>i?zLA5gt)~1b4wRC~l~Fbq5h2L(k$MJzAwWXUEVaI4O||+B`pj_h_GXQ$u6*
zS-Y;d2z0*!xv1lV%EiU!fqtJ&1TO)nRUXEEtK@@^N+wrPU-j8?h%Y}Ow<>p^UDWV=
zTG7V9s?Q8j{Ph(j^lws&a!?uT&vX}Jm`JR0F!2n2rd?|f0HNzTNNdf(@x(y5iiwXe
z3#^BL$p`3woC6>vK9iViJTGz*;?pdipKtx{erulP*WISs{*1qMkOL)4z2a*v9TnM)
zlzz}gwju8R@sj%^Pw_@D&W-OA0ueX!REuFskTnm(=WzIW<bN^8uRR@J%?LF7WyYXS
z0#k6o6~j=+_|RFk4V;TMfV@>Un+D%SJde(oAt|us-RYAp<9e3}O%sgzFyXP|+ei}5
zgA;)j->!d|pN}<58YbfR&Or3IO1s2?o`B%(O$>UXnIr<%WbOKd^H!Ql^X}T3j)$kP
zPiwi`wq`0gF({7s4?<5}wr;WXGV7ZGH+SMkROBG<N|(+C5N$3GDy$ALex(|{QAts^
z=M^D6>Yxy!fbruaS@e*7-dWSxJTHcTFR5<uaD`-C(&9iRvxI>kw&W^~qcRAm@#eV&
z0hDF!nkhJFgzl<snUEg-MYN?Kv>{o>5=|*V(o|q^Ydd@oKwPJT@<zjw`^r(!cbo@q
zQ?KqquYz!(46m*H30^gX{G6ed4gvkG##3y?{KyCxnWN~LCdRET_>b_#AMfw15Efe#
zp&}y!DEH7-i}o$38AvUd6Up2>{pOLfD{$MDe%SIcU%a{gv|mp_%dy_8VeGCpVqPs%
zujS-?!7gco(tWKyu^IcbH|QuYwB$BfApF<%$d**iVe9AK_a}9Oq!aP^lYcHt95~7_
zm$|1!ysESK)2;be@C#YKtXH*u)DmTTeDc1%aQPDDBs>sw8>asJ_V1rM^vh3_G~*Sq
zrF8Y85y*xw<Gxx^ux$7C=xWX)v1ya_1ytffsa`xbjKnUHs@YY#ea@DS-r6<NU-#}L
zK~yfJ*Bz3TrD!eC7>^R@Upb&(w~P|nU$aeORdvD32t$$vksd4a`_Za}h%y)b(vFfC
z`4};C)YPp!KGkyE8F|eLj);k2yxETKtvvTl*0aS6{;hQs77wo=*~X9lU2ievxtRy*
z0#l#EJr1F7GzLlTZ!eD9KogRHKv)*S=WZ)mA)ULt$865fn+cc;4w_cKx>$ESUF!kP
zLQ+z4+36P(L>i3BH(d&LESj4{k*k=%BlE=P1+SeMYO)!62NrauKdn;wN*@1kdYo5z
zD>K{$9m4W3iX+eJKA+R9+FvYt9Tur(wJ!KTf&6#VEcrNa*!RF*Y?$yP?$CB!kwgc)
z-KOu0C#N0N;-v#Kj_=F}jNaOPca3c6JNWS)qa6X%fTchUcq-A;c`7rhCa&CwVPyZr
zp|sAO<n$QVmVel?Nt<!utt1?b3m#C$?*-V{HNU<8I0Gpw)!K_r^kCb*9B%y}{CSYI
zS*m$b{A6CCg%VohMWj=vrHUwU$qj)5R63;gmh&t!KZ3WoSjLMT=c<YOY=6rVHqOU(
z2hL9_m?en{tb0*D5p5J64yjlu(MOtN!M7=QOD5N0C2^MjWOgopQl6G#vAwMrT@<gl
z(cY>iLs%I0B_BYt;gbCDEi7^l69U+BF1t)ACO`WFf0VPv&Zn5&-AMGxwl+b(ho#kX
zps9s+1n7DqAPqq?y-kQ3JR)AZ4QhlnddM~7em{f}{*vjJ*75nC>9^zDYyAZj2{q4E
zR^Ro|kf+Is3y^-a5JS5~YwQTf2(kRMkb=&o=H~Spzdd8wmWM5BRG8n#*yB50t<O`_
z^P|KgL{Ph*8x8J$Q*OGxw|7pw<OemrBjZ>q&XG`cDy3RdC4uoN3u>)+r~m!QaRQgW
zviqe%&;(B!o`VYgFt%jD<d)f+D62QCe50;87b6u_HH;=3v0qf0KF1dq(?=SynOXTV
z6ULg$bhU6->#qImHjQlA&Y{|rv;C@aw5mxTWg{TF9$n_sd|WQ9&2~qMP1CafPAK!B
zVUAkP?yI`4_Zy{;a_HhwczDaA^syQYf{we>`iN**Pr4;dF&IO`CF4x_YAy3|HQ%$#
zlDOn@%CmV$K9fAMOjN&3K=l2$STZ+yL>t{6N6a_jp7qV925c^>LTy<PDO@%<{r!0n
zY`{=x#s9^(Fsu_ar(UhPK3Z6uuf9AdPfn^nUu3_W0`oyHZ@EZN%}}!FVMi^0MCn+d
zA(TO?32e}&-}jgc89>)y2p>1^J%}nvOF9;z^aq;|M@30izVwIcJI$?VqD_Q=AkzyC
z#XB0R#C@be4vAI%LD@kWi>#q7B0hNc*i$<UvLSk}N^#$5xen`2s`zM-zAF%uA)p!k
zp)njia>JW9Zm>1&cqm}UUpOA?Vo<TTOg30+qeqb_y~#yz@^wst`GSaPxA`6kkxtH2
ziI8)XS)`)@69PrIKa$4FfJ_Mf`Mlh@Gig~2VfsjbaxVHkN>$6I7)jfi!OU?HXir=2
zk?3&Le<{wqyGv}NxHAZcjapW8^)-7GeSmRBfN<EmdI&m2J44W^>h(X-u}+L#(hI*m
z5fU<~jle2Fzs?RE8!Q=V&E(lbb`_pm_ITavDGZIC$5J_VCOd(-pm|~U%lT`cEBW~e
zkO&^${2=nP<uw&lF;Jz)W_nqo3-ot@eG!GK++vCFlsuh#1w*#@vLKZwxa_}ZX2ly{
zU`JYf)X5VoKpc77@%9#uB9t1`CNkk)e%0ENK2{BiIG<w+D0`7?cG-vjr5Vsl&W^rZ
z{)>i=coS3ZsR6XICwpAg2_R5kxNMB*BwCaGf@830S*cZPM19IenDX}Ddy5>?+TQe4
zZ*INSEP7RQ*MvM(57%7pPjw&yGuYb8(o0_OhcnE8?|kHWG(i#MCnR~+{#R?gZ=RUM
za<D4mhrl5mCM}FfS{`g^<F;%<Q_`^c1Nx<d+b-bN;x4hh@ae%Xi#KKa&Py%JHM~^V
zLKAUvZZZMfhb!VrcZf10e?IKxTyglgO+aXXRw?+#N{{W;dd67%&eoWj52wPR+FA8$
zAsDYE1=ar<F)NEkqg$$4H-CS>N;EZVNX1#h87FF}%$Co>)+gO#Zf5){$=o}}A--ez
z?T;21L-U`+XN+%<ow8L%ICWB!K2_=E(DX_4ycocfR)mx*{qyAKPNWi*XDLD*kdDGP
z$?axNPLVOYUk3q~2I`#|gS;HvqPxV0<peC+nLUdEhV6MLg?#SlD#t<@NXhUAH>DMe
zGv<W&3kfL{pLXi1l7jaBp3KvQmJ&%OzEIVKjyt@^YXVt}bn~Up2-nFc)1&X=m1gS0
zjcwhVjw1J|TbHQ2c9~TV*S@kBE-!^^8r@GCg}`(>RN@r>Hh*Z-WBeC+2n8CQ_{-QV
ze7y^XWo8@`=VeL%u|hM02M7UqdA|zaMMbkHHD;YkzXSzAwAQ|hoFpq&Iu)6xzwa8C
zv4TNW4%l_-7H%chJm&Fxti@g5tUp#1&t&_Veg@MuWDS!|dlgu2jsT(jkUi`>M%*e~
zwP5T7^U>Lzj>(aJRdu4b?Wl@UqXe;BQ^p?7tR-^RY3t+rd|8gsA(fM{8Dchc3WU%5
zK`zAlqsh+I&PY~s)7~}m-I{pIhHVnjtaU%`ZW?q?Bb27k-lpEHJAds0L-N(_JsmG8
zET7KE9UjMY9Jj(`A>qBs6#rBMM|Z*m8b%&3>a^XK+lFeY^u&7N{AvE#i+#KG*b+Tw
z;OwH+lN>?N=`K82UWJ;hS&NZ8omO`iy$rToKZ%mG9>B*j9=7M(X3|PYy)w5arJqa~
zr|%X$b`T+;c_wfT{B7Li+*B5Bm|=OyWzH9Cjcx_YxuP8zeaLvAptah1UKg2^NAtRt
z?>&Sde*+_GZhNe`_RC9`G;X_nx5rSV<f-m<udDNG7f^eXxeRW+ohL!VhO$AGN8*3~
z3$G~;I{E|wJ!w&MVkfO~Tc-n;=8{#bT#!?TzR}GyUhfk+^qdCe`_+zj%e>mU6Mxw%
zyfuy{!n&GmFE*jcA*)<aecGhx$BE}{liT0Tk_LJYs=q~{<p~#}p>O8DYxWhcjyS1=
zRfGS6?id^HEWnLeakS|dEmE~0iR)uR&9te8D?$J9+0Dbz$M}GaA#kG0?RP=w?nG;^
zbwbI}eHnzhD&mWw0auDIyBbv6MMIQSrw+3B)C~dj-|#>Qle4Y*By=aFb$t}hO4?^S
zH>%r{+eQJUkCC$HyUcKLc0gp<jM}r9kpGgGY1*^cLSVoTYG@a1tVO|8m?%GG-Ay*%
zur;+e#QD!2to(FF+WUFtvI>7k)GJPfZpw}SyWX6Oc9(LL#ya!K9+RN^60!#6?9wx1
zBFS4O#}@^xepiG_S_ojw$xICaxJ)4u??bl6%@^=SQuhGaOJpPdGGxahl*ganje_K$
z^eP_YmH1$QlGZZC#>{Pvql+R?b!OEscmHTAGZwF?E=FlZqQ>T7THgh|xz=u8!2q+w
z6AC{Vf#c=c)RJtK^x`h0Uq*F+fsd&6(Az`;QWPEL?cV!dM@9qTO^cZR<@)vq{kjQO
zJt4bUwQgKnVH!w!chC?aEf2|#>3Ezk8fW!Shd~$2w8%`GQ%;SWm${vR0sO=tRW~Vp
zEWr%bYRaO9TibRR7yX1R=m}rI5DdFcCuRrb@%E~^k$OD|%t_^RYc@A#j$)}@{*1}d
z%X{GuE&0_iDEozkq4b}_?3??#1aAi=a!@GX{;V2d55HnWF6*h4(E$PS3&dkQOzcjJ
z6Jm11Rml7A`(>$-a&01UiyM8ONS2CJcHBhVSrDL5C=0F3J39eVQ6t0Qpt9ln#-xjG
zi_(X$ZkkapOMdPB^S4Xka*cm$+Tm`faNg8(-5ztL{hZsISIGYvJ<@8hp0~NXzvATQ
zeNHpXf&qP6_>=Q(q61hWMr&7FtIB@om|4EpIuNgoeZZLi;+oUnbUBQ~X1E0h80f73
zx2OL!$JEJS`D|8WT`cOfHoguD+s1uetlTAxWX5FLx+LLY78WEv9il;hw@Mpy`((O&
z9r=*HpQ3pfx2aoJ3X>7GtENX2&y&E_S2pi$U=!#13wXvQrhbJ1%SlN0Zm{kbicceE
z$2^@>;MSd6x-GXq`^Zaoegi_vVqIhD*7de3A*eyj&(FYbwK-YU3`qAXYIT)y9UI}I
zL%GKkb%Jv^(tENCWvNfV!_Q@WheOKrl~C)3{pZt(O>9QuT4$k+(T522ov{@%V#P+(
za__};xg$(slsJVD(rY=Mbjme6bSUDq?mMEh)pyEPX!)06?|h+^NBU717a@&oO9(Qa
zE%f?dq(dXN{Z!i?);fbn1ykBPOoL`X05j%lJagSX!1%%B++dhzw;CAXE5nPZ@Hzik
zCZ3AEnL+3IIkJLxf{5fQa(lA2xU!M!xEzcs{nEh7sMA7MaR3g*>s=nJK!Ss)>Lu>Y
zAEJ$~(p~#betL5v6nC_UabNHHoB_KDD+AkKbK~jQw*CpwYT>}i%4Bb{fJnd%Kxt$3
zFKvM5eg)khfj9Ivu)@0j^Y(zmP#MBy_jjPt)4tZLy*h+GFu?+BaG25K1}+sXty#4;
zFF4-z)h=15%DB5=?l1TN7a*os0%)@&<i`XrHd1b4*JmHspqD=;r$5#)-t)+7SJeqR
zD0AKLmr6n%vW&TRnD9EIujtSd)}_N4q15OG97miFYK6dvp3I&PuKLvDZ`*$+QEwpA
zING~dtCl@qc!)~Kq<&)?0%BmZT<!(;fu>JvG8j2+@0Oa5&prKH4D*@k6MxIl2ezBE
z5Dsz1qGkTmt7ux@C*2O0NWu4;XYeO%zt&;Q`|EQ{B*?bV=+(#Fs;-{>>-&M+63kwP
z%073uX0a*pszK1UBsB%}`OQh@uQ;q^ISo7OFc=@}w4!SoBm1DrMgnls6pmb8m%`k{
zV((X54gx&FKX*0Xk_**Zbj)@SEfE^hOsi!eK3|ENZMN2z`DvM%GyjuK@-c~U`B^7_
z`r=+)vlGiRn11xFNI4UGg@O^!&zbAdtk_0v2|crav+L~tq3JAuqHMo5zAUlQ-QC?U
z9fC+pH&W8woeKyEA|N2$DP2<14N8g9Dcz}r)OYjE{2iTf#&L#y_POJnb6vlyEjpS&
ztyuc#w<&v3Lp`6y8+XcQ+FfxEPFC3$W%w5MF!0XaTP&Nh>rwJN<h+jmE+Bw|#tM36
zoWPPqKukV*r`K!UpkZVbJ7(Q&S%mOAiDy`uyQ3e7>9|I2p9oU`JMonw1XvEv{~gra
zngj6z3ml2+m>S&yOL4lQP-!Skn>y|o?Sxw>xP8*>y6eE6s5X1S6hn7xlM1X-^vXqh
zS_Yg+QB}4N`(pAx8Lv;2)<@1J&oDlQeRWHC+@9(9q^;7%*WrOf#*x{+u-$)?@u8VU
z5!fDOnbyoQNS+PJO4l&V_oKkzqaUB2$-`cG@z`S>ce^=fe+MkyDAOCM#vgHkE}csZ
zI+I&_KhT`XH6N4YF(VRaK6?84lsO!^4K)TFxarqQ@C5QpeYNRaVS5zxcpPW_PjbdO
z+=5i+Z}pKvq9DH>$j=^B@Of#Duds)V?%gsw)s#$Hg<>clfy?wa0#jam7y)!8a~Ha{
zPf}Bm8)kGlf4z%_v))@R@79P*nd5utuvNX<Z4eSLFfhp#Wp7t3biE!EsJ%dP6sJQL
zk=FevMzQdtk-*9`?K+r3CCaHRA%_z+;!>r!p{YJfKaV%RlprjyQ8tk}?!d2M(A7%M
z0RL+eOT*#^MRk$d`Qkpt);LC9!T}CD+4R65XT|{cc=MhY6T4WioJGT068z1b3xA=3
zb_EhL^4UqqV<rH#rvja9GEhf~PznJ95?})Q$;Au)*Z>M?O%)Cz)OXm*DT3!qP=i`A
zHjmUJuPQa8^a4Cz&n6E4Qq0dLRzYn@WmAv8!b7WHnXrv3IS!6JJTyx~t3$N~?2yDX
z1t`fZD<$p{BU?l7&k~pMNKR8{I+0rfk{$I0DcmHP*f5UCUrT67xLEMdqiU_wx$d0y
zsY6r4_6dJMYB$~@vA46*o6E<RrXG4f+*K!??$oT!vOWkU=#FY=vPsjZ8FR0m@Nw<&
zxwIi=KxspF$A$Vblb_2UbuWa|aci5b0BuE>W75BF%y0(b&ba{*0z3vyU$D`8E=U&2
z#tzeU+=d?%eH`E#`ICVHi_Z%0;z~gBw&L1;Z+HB|^SUl~^UzSR?x!f0CMMHfR0hk#
zZVJmm)w-qyDyN<%S4<$WKjR2>!x!bx0aHd;paLj>uWO1t@0HnHu-sg#rFZ{q&GI)i
zH?sf<O|b35PF&zHzr1ALr0y9L0U<$8uGmUn&B97p@k&gYi8V-lZj^}dk5W5${>1VB
z_K6)>hpd69x>CO`Z0mb!gU^XceM5smhkqlm1IU&U(!F>YR}=_0Ux3^!jvoC1DmHOy
zB3*u(gehf9b8&Oxqtl?$57c7(YJzoM5kHB|w)w8Nx_*boukAdBJ)ij0AKP}^5o8nF
zCI2Scap7Lk7U#~KLtI>1aGdw!`Mm=rGvSAc)wS1lO#Ep_wi@(JUzJrDy}Z+sxe@1g
ziCdrXRV|!|UISCU{-(LRinb>EL;mNNVSm)_u!A2EmacHc#Q~VbuzWhZ?*8rB#tvam
zKBP7GNWg8P3^mT<mva=h?(ltGeN%w*$Ll5WTY*gih11(Q#VJ1`V#d&x6qU7ds?#OR
zb6;a?^#ET_P1!vWSk=bbtX|(7Sw=AtnBI5#Jhds#gHFTAU07Yr5uC~GUoDhsUBpPL
z=e)xptihz`ola4K>cz>G1-%F>l3`ddJEbmC7&%;MXeE+j|4Sg|eLR05u14kGtaM6E
z)BGO*6NQAEo}hF5O$IKzC)qoQ`4I8{!vfATZg9x`=e(5m6wwEuj;3S*4|je!f#-_A
zp!rH$o8U?EG)A_A;to{Q!20I?@5=>50jRK_m~p^Cz{JKD^VARrd!v^AXL4k{YgyDM
z)jg2gR}6L)upp>hy)4VS4GeFcHNd(5p29o*v#ZPDEl0+3Isx_x(ob0T$%Di3oSnhA
z(6`XR{qTZN!_M-FIHZI4lJ{@N!#A=sueS-(0=OGCIzl5BF64P+;$Qwy8Nb50cIBzf
zRxACP2U=O8E%HB$W2)`wcV|mRixy#9WF3Zmp$*d^_c!7b78Ddp_;!iQ%*E?s|2-3J
zPR_HeVL)}&6@+`qdLsmOk(by;DRu)_Pz@(uNXEJ4*7qQ%G3k~Nqv>xK7hHIV)vliu
zPK6#?qW+6+Um|RBWv-k%F*h^#b@{U7`tk1A5EU}p%tb0@UjOf&tJRq>^_xEpt182e
ze{<wE?@HOA>DQflzc%J=XE4LOv#?$)M!xg>wbhu&Usmdky53;Xbm5_7dbTj<zw|e4
z5chOVc=fT;&fq-gc{0;TS_@07DW>wJ=~c<PDN0l7%M#qg%<q3SH1_EmCb)jzC#|lj
zj!r<^z3;R%@giNl?r=g>Cm6vm;7z^&HWhael%O-Y3@mbjz{F(@9Jc>CEhK})nS{&$
zKj~(WjY7n2GTQEU(XyT4z(z5f4U+=ND>9K7Z-6H^m1+?*;GY0YKyzGFbS+i!g=+2Z
zYhR6zJ)eQcK?&mS<nm+vKP}_+7Kw=!Gga)l&;h6qMiSZ?7blm?5(m+I|8rrJh07|U
zXMC+Us7S3yYKJ0>8SB-C{}3*7h&%*ra%&9uHBB>8XS{k}a5*psl>c|Q1pQY|AoZbc
zey!_ZE#Bj$q2(j#-jYEG$-^1qon1B3%l5H3XB<kC`&sg7#s~lR+RQGZ{szazpS3*`
zO^sQeZvgRXkFkw)w9=8f5n7EH>O(4XSEWax``@nyP6twFKW&Y|F9um=|0X%O2G`j-
zkiR+E+xOZ=5-k?I-r+lYz7Vuu4<5otBkt$M%nN{tAv3qm^voo^GVmgpP2i0_-<?~z
z_H|SW`-a0jk~0;@GWE2<hJIE3O#2)4=@;neiSHiz2v+@T=R*A!*Ulald3;ao6yLc>
zqJT~cvm~)82BT{oy>Tnh2S!z_g^+KVoG6;`6{psnR1H9dF#degCS;`X`#mgg_2?cn
zC7g5f{bR^&prmyxSg1EgE+NWtawxzm3p|m>^R!6*|I-3=JqAOnyB_?3L88}d)_4fS
z!~pmD=GN93Kx!kN7E<32C%-w7xLWE1hBF|=wFb29<m6;^ZS8(wiB(il*#f!GXCo8(
zvuFM=sTVJxT_9Jj_3_~r03zdJUa`eY<?YStLNm)qPiOEpdYY5cUAJHS6;g;pJHyLV
zFnSrGi@DbPfjFF}@Jn>fiqTyJCGWh4^~&pXO6DSHtxIN1ypJ-$hB`ffZY+wH*X?r;
zUL{B^V+Ul58|WrXvFQe)?|1zxI=S(!($4xy_gDPRDhJ85J-s~!gV5s7ERmAb3WuEN
z`;?<kQgAV?DX%EAiV=9QTwLo}%*pke8eYVqUc75ZL~+mQt)awB@~ba(5K5yo+kctP
z(Nm4O#vEK22b37fMBxT-W9HEtFGpd#>uhW58_(?9iA57ZCZ=Ao^Z%@8^l4V#gm@c$
ziHR_c{g`L38cH%Kda=mD5!o0eE6P`@E}45AjW_C}OJUxqfIIp&$5mKG)RXQ-f9wo(
z1pcaXIJe+e6TPCH?)R&`Nm|;$JUAlY%KlncfcEl8@Me64f8lQqgD3{jjmaWZGdyq*
zfes5nE{jqSL*X^u>v~LQ+fYK{CQUPTx~2UhiO!Uk1K=I_Zu;Q9_XupO>%io*k?phc
z`}%AHgHkjZcqd4d4L%vXe2D}yH&{T_CKgz-8^a&zJr8$p<bnZu%X%zBR$UzjI4S74
zR1KNJ!@^oXVvRwaIWq7zosD4|3_X3>($WY|CK_-EJcWaRvpDnHCJ@CVL5VE$Go0$D
zA1jI3yuz#MH9VT6$yY|M-4ZilovPbs7+0SPKN)N=JOfxwUSlJ8h5xsmik3m>ovrXv
zfYr2R3gNr>6o^iq_Qs6WWv(0bolinKdM+t1x{%^*%IUS`Md)-3%Y^jc9o=3GbRt_L
zf^mDr)QScgH4Tf6U(R6*Re<TGAh&)n>V!36XB}51n@p<(%K`3ysb*!;&Wqxh+tNZV
z>ea)HUUG*7DG(;-ptbSs&L`Vf;}=20Z0e-pS1JQecU58Z2N$&nd28jZy9F7|h>m^u
zwSxFV>+nGf-5i_r3@v588u??Ej}2FEYUUF2dudT~pi8*D`9(BPT+eiU0QL106u*f;
znjZ5to<^~0DEc{}BQy#qM91V;R7R8;N<QBwH)C$8(WfZycVCh$4^;frO@^<(%5Nnq
z)olClrvN<%_V45B{jrY@6*?+v>*no0K%X^2dq~{@2PTY0U|OGi<H?oY0;m^U$dBKD
zw)wcAVEyu@wejshTodZL!K`UN|5kkY9#kQT2?^c6fhHgPr;z$biyK_N5H#1m;aa+<
zo;W!Dl5{G9u4JpS53k<{3f05C#VB4SshF93MEc{;6z!M#T*Pf<)0cT9tVqAb4kvJ*
zR%)S4<kgfpL`K-~0lQcy`3PLG|M_2{{CE@Mnxze6o9?qULoZ6an4!fX++G#|Heu%K
zBoDVeD&lPFty=xH<BvJvmapRj1b<X=9esJvOjB#Y73$F3S2p^cFe^KwujM#rB76|n
z^*iBoPRXntC9@`LgUU6s2sBd&i)i(Ijw0wEJrFGB8PF>ZN9Zo3ryFafO$XPY+Hl3x
z3`GRCa(Ovp9cB18^E-}7Z%oA)+|(WvdAiNK^+p{O(~=6_mNFkg1zn_BV<s@NniA*L
z!b*17s}IVtGN(9{;5@)H{f0Y@!puJq_|<*(Y~<A*oHQeFo}VM37Lh;1)wuN}@AJzd
z@!Zyxzh0b@hN8&F6kM~ZiJH96oHt?f0<vXJ>oQ}_f&-ENo%g&dSNopAnJu~UrEG}>
zn2=8X++6$GVY9LD==$yP5K@VGSAORO<2ZAG){l&vO*r-cybmG{n8n000TUntU<MEq
za&pGQu_XaX1;b}GNT|&xbMEX(MNg-wll1L)NUUuCf{AEtM^6dTL-?B(FU_m5PAmf}
z<PI<AQLa3uYo3izCq9tP@a~3b6U9E}2WBp!o+JWm2LsFAC<956$NH?^ET-DqcDl4K
zN|Ow1wL7))a>>^x<dnWXqlV1V!FoMeqei}dp^j6@`)-pegw|G$f>+ZX0zF|!0w)w)
znJ3?*6=1me>mN1eN2f{4n34jlNs4vgp`pun0`2G5uWR7RIn_1%oSEJ9vo9RA8?ATw
ztM@p2g!Qf{Y+h08*B#^=hbi30o^VkLjE{>oIK&7t&2miBwn>RoQaw^UG|0{1<W`55
zc7^S9R`35+#~(G3Z+Bbf^$6Na*?sIq^&v<Z79CHjG}kDL<qvrH7Zc8exi_{Xb^uLp
zt>5-&9)Imod+2k}Wk<T+O}ZgTV$%@ocl{Rm{ifZJHQAjW8P=H0#x8-KQJ;(1hVx5g
zWCaIHQr(J+)A|iJy$P`cG}ob<3-<;e-TRa7kDlsm>wmEz@X1MaCY6kSfE7>x#~+Y+
zJRO0w+}!a`jE|>3z&3g@%0Kah`FJ6NDBr9Ey4~V^iNy=U&x~LHV^+X5VWO>lc*e2H
z9GNf6m@o;NkdZ)B?^tv(dn|@n!_necphqP(MHehC!`3%S_Ma}SLA3gKF#Zq&abaY8
zT|ZJAOTdYN65p#>)A5QGh}vY+F%p#SWZU}hDw)pl<wTNEfL+*jh&wEPw%eup^6c$C
z>kSk8=2%<}mP}H90K0xC@nsI~hrK~uyj`;80bf{{nAS<gWLK5v3)aT?0lF(n{*jSy
zv#u3{&YjRmuQ0+}#t|KDI_c?PC)vm%rfe^fh^b{w`y{2Q5)+OT@@d(k=q8FbqY-M$
zkNanZd*>@3`nHx7_ZSg48$4B#y&^z7TmH!QY9EHdO=7aQ4s8evX3mY$3{F)?LM%@J
z<=LQDX#hLu>{p5%1{>+WyH)+ZVAl8MhLZHRCDWmZ6-WMo1N6*fS55M@%i&eiL+zn5
zbY#ckiuGaxCdv)fO-A{EjE>#d8V$R8n4Y{Qb#nf577I<C(=Uq~9`#&~%*Z(|EH4v{
zQ^SL7*?}C3CklGVbsx2Uy(Q{X3?pFRm;tKI6L}su4uh}SR^7nux7CppblJ){BN>$)
z>cu*=6H<&qlnC@W>ZLmh!tYk&cJf@MdD}bs{Z6~YfnPb}q#Ac6tt>!P@ROyibxz$`
zxVR%3olNn<YOLOiJXaPK3<G5&Bpye>c?Lty^?)_yl%u4s584?=qO2v+sG-vlhRp9i
z@ab_L)UfTQJxg1{W}00i84EI-nyX`7@+i(wgP|GRBslTRJzH1)*)Qw<)ho*<yv)wP
zy<^F}-CW>{Oy`g>aW?7FvPbMvssyg%r-`kxfgJpLdeK{x#G~*}Ia_Y$-)G3w)M{Ee
z%r8hh5ae8pd^oz73pX1~PUui^D7mszssfk(+7UGvp52Zk$w6rZ(NvMwIwq-sG|R79
zBY2Kn9~DxFTDl*nmmJyOq4@K9e@ThRs?vk^g)a1Grt!!Tr{<kVBfLPb#^nDwhTtwK
zO34orzd}^Y#@B?{`$3A~6HDcM>$`~GIm6SV2ltZ#V2%gw>I~pZ1cM3vR(HnYVyXfy
z&?^h>wI>`w_sQI^1&!(~Tf12zxiAd&D{@ocnmKG@Oq4B$uxYl8;k6qZTYPL8v44R|
zjbUS4p@e!pAuyt~iw~|VW&*ZeB98SP4+%Y&F2!^mDoC`2eNg$D`DgbF!#77NSlQtx
z-u_o<N`G%w<u(%Ch|rgWhnp|lT-P}6?6@2yb#>4f|BAPN#UVO8y%%r)^qFiD@LBgC
z;D`%7@sWOen8vPrKN}ts!ug5c6h`Nju%mq8mlR|eUJ{t5%$IgOA-UFx8KvdgKNs+|
z5Al77{BA-N`j(mUG9&Ja&>FfRt-DuV#!aCj#5`cB?L1`qGV>2(jIJWJteavj<2#Wc
zleMAfS<#>s=u+x`*#;dCvD53#>dxJ~fTI9e3A|9LuP<-PKL!8{uLC1S4Kl&T^D2;x
zw;9cxWA^xP?|HIvKeYkkM$AFvVgKOZIk?e_-CKdf<2n<Ri*pR~a4@xQhs8(ZNUYkF
zz;p%T*u8T=XU{pxm}$|!;(J0IPixKXw?XR<jJO*|aPvH%4d>-o5u<bZV&CPmqor@|
z-cB7vzmJda|8-`uX0$OV(1q_<#k|n$1vWSJFQ};_AJtp!K>0UywwmOOb<t@L|1z?m
zyi<P|aL?!cd6lc!V-K)Gve=A6aIL<TN$aA%SQR2Vrrrtgn9OE+&du+DrRS+k&}Zo*
zJYzX~dFkx>Bg`{KoCnLK6M3ZtF((5$`}d-gw5|D_+mAOJF-!<_)Dra#m1Q4znlf+l
zJZQ9LXI~q?LKF7X>4C|*DO;)wQ~&Wz9FvjU{OJAMT<NFAD<{Di=7rUPxctZ8RnA@X
zqx-Pz6-Wu_`Yi!?ASi_S9p`v}yBfai43Z%0oR>xA;wTIjwJ`Jpxe&BT+m^?H)XW#`
z{?{O*g-u+@eNzEMNXddALvY6ye!L2K9D4q6h8=v_kb4u50bXdrXHDNw0*!n=nzPm^
zN9Y@`hmFJpH*7cn!M^$B->vN!xl-Kxdfu!hAy=WLmF}c|?N4Wt)sXk^!+_pnV*_y*
zsg4Vm9Qipp@(G8RDZ*S$)rnPZV9#H~VzR}<t3^hu+#fsS{V=IwxP80@X^vRRw~d<H
zk+AGUB%a<!L4P&0Vd^h&N#&km>q$6d4B3Pa1%fj;;POy7;QHW;qUx1!s9<jC-j&g1
zYFsK$$Fb-T7<dIqctkjB`WixU=haC<&{kl<-?E~X{qmvr@E<Ez_sr{Myn)KQx1n??
zEi4|!5;0Q=d$XVU@5#KXkL-{*9{<gD%{4iZ0~B0=$4*QxhzZVa0YLzMpeJr`=N<Zf
z)qaG_sHGctwPnrC=>T~4E5TiSeEiM_p%of_{v=RTBqb&F<v#pjh5)_T)sjd2z7T5+
zp1OXO7xy)GfzILgGl60shqCH`zk<H!qQ2Yyz8l_Wx}PV2LlBxoXod_PAHPLC87?#%
zs*aXl%~LA@1Oq!Lv-#3^zx>TAeV_jl2&*(l$vE8lcA`bRb|`xSCYn+64hb4%c^UOL
zz!WgV-``bK{RW{o?4JBCYT>oa%P3{y?IKAWoxL>FAEz6*U_R@0LAK`v20_8xZgQza
zllkc`I~I|uW53Tv#u<gGtYn3|+M1lJyOw`2skYa~PBC|j##Iqr&)ZHc$LdzZUU-!P
zs0C*D?yjx(;am?exW7o{GJb}JRsa%E#KBeKDa4veqJ25i=y~nj(WhzNjws;gzW?{P
z{d)a9AZwKNdVmY$Mp<c&dH*|H{cKy&r;NZ?>3qm=_Ue|fMu*vd_8BqpPuU1hv>%X-
z!1Uj*iH|R3HOhHx@B=%hudho>VL3g<dEjZwNQWlp4)*us;F(v-ddy2+yB42akxt<T
z2U`g+*#dv2uVWMM#JdAnZ@*+`rm{?rH8JaF3sTMH#V;Sz_-CRvQ^&YwbN$4+oeS)j
zD6!Q(lTSJP{gjk`x=B1Z#Qa|vvCH{#pcGrr$xW)~9_f~l9xM{o-q`HK&iBVll^G`)
zv-Z@hRYCHsc8H&>Im>dT+x-#Jx3MwC=cdlRAIOCS+fzkz5GFznOxQ}?-VzuhnP25l
z3a4;oTObc;3ak`<8g1>YcjOLi`F7C)2AmIa8}5n^o3^5N7OYub1#1-K$XBttH`D_p
z$|)Sbueu&3(1|!`X!EhsO<9#FyP5H%K<s272%QuMh;5J!YUD2tm;)eQ@9oo2Acb>z
z6`{IyJ45+#rS?rYcL{K*PvwzJBX7G1F)dCu7Nb1gD^BR9R)~Bk+6tn)j%DF*3bqPG
z4_e?OjeKzWow3rH4W5`iuEdGde<_u$y6+ic<a|Q^M!mVR<ISj6WL7plmjS-pL8a5l
z&)R8S<+|Zj>Xvi}dDEW<_m+BFe#yUN!BDw;ps3XS_(Id@Cll3!3*ViGpRA%R%=mS$
znmn@gDYd2F53YRY^p)9U4P)5_w_@^;(lwgvfwXRnhw!`3oK~tPity#qB6yKTi!Ty}
zZRxBb&=5*>1GNiCTRolDGu82}1S|yc`#(R8JjMhfD`>yRH>9&}f!mNq79KUGkTF8T
zObazJDVh(~)kGEntt6S8XBZ*MXwYlHQ0b$f=cPr$s0+c(=;JzxBhSzyqdNxDbghU9
z3W1&Y1RN4Yi9T1X!f8yARf};l8PXy|=wwKhP8{z@_q%j`smRHY^sdftl51&tCc<q#
zd}ub@_byn(DzfhW=s`!Hio@x@)R}nFV7SIyxOU@`H}^qRY?!P-(vy2y$8Z3&Pu99g
z=X~S>dClowUUWIVxIfDHT%iwdlaHa8T;fc4t$%iNsq#B6e|+mBG@7Pjm<;PrADIsL
zXuR=2M0v3@0bm&*ATu5q&s$<x!!?4>-kbV57=q0kji<BCasB!0oh92UKMP)S{&Lz1
z-}uEyseQj4!T_v-zO{$C&Z^+1`Il_=4_F@ySk;V@k<R73&ML3US&nNB^KBCE5Kxsm
znn&jx0vfAV-Eq7kz#wvS-gvrv8v2~_sLOuicXli@<X=-+oAkZ~>FE!)ryeB3GoWKn
z4(@DK1FpAKV?-D3h|#6;iKwXV_^9`+HJWYupSz5${^Fw3r3-wmFw&a<XfB>FA5!Dk
zRTe-yhDBiDVDSpzE}o)&+yE07hmcUG(4qtE!X_D?faYV}%u$Fp=5?cKxeI!XY;J=d
z9>y^Z9$L&JDtf3qQ|}JB+1^UI=!`_t+V~qy*&}uUO1d=%J?9(ii$JcxCh4y@+Cs0-
zm}EqI&x2|oV{hGl)4E=mJ&r26#TVK)dQprnb9_DAvLi(I!N70O2;r?yVqm~3Zlxy1
zMH8C-aq?s3=jr~79i8ie;?ptRAL=2yY>Hp}GB>viWI8Wt4n<Lk&lx^WwMKZ>%SHt}
zN)Q(3<KQ7J<A?Sy<U~K|E>}<ep=*Fqr_Yg5s0%P$NPO-LB<{6Y>4qmWuCnrJznYtD
zO0}6dNPtvnIx|33*^H5oKvaOZLB_PftPSZ=dXq6rLlX@Ub%>pRNyV>AR1*2RiUT3`
ztRj4uhxBKCbjVH#=uR2w=r(>m`9dNUsV^H-YTRq(f*<mxu{XA2&1JRi7|E6ZtBc|!
z*P*-_@UohT0#erI-Jn3z$*0mu^Phrf4|os8&cJ-?NtG<Qk@!oyh=C{4C-H3u_zZZ$
zPf{p=@IpgF14J``3+i5s(gUjv!wlBB$U42br4oXTKpyw5|2yJ_w!D~nG}cx8c!(bW
zeJvgh4sH|nT(!Ni4_O@_Nd>|cF*PMU8OwLtjox;EF@Sm49jC4D2Z^&$fYce#^X{M%
zD3<Q>T;B`S{>vM^&=)0Gw9L6*jQOHSuOf=fkS{$YyA3M~i3frh`I=8BljZd%0nsYZ
z2gHTmw<hPt74JEg(M2M1hC+v1jF;JY#+_s}pondhf>aR#uz|@%Yy9~M>7zHkJLSgR
z-faqaN6!&kWhWV+{tVHy*I)Zie=D)_uuY^`)K6UBe6m``0Pm-g)Y@Oh-(V2L!oqTT
z`R0jVg%4C6xz$6(wzh^ucU!M{qa93O6*Pn=8$1>Q-!6ZBO@|D#CSx9y6zb$EJJK<w
zQbPM}x%si1sK7nExL$;{$^YwC#K05yO;UI7x>d8W*)~;Mg&7%6TRzRUa8Cd2u<-En
zWt}-yMWcCyzwr%~)wbg4VdZ;|VEOVh3qm60$?@L}`7BC(Or@3KwkLZr@wpQbV9Ttv
z7}S~k^2+|)7p1-H#Y1(s{~438?#-FI<<@)cd63}Cr3-%g;>$CCXdVC@9IkYbL)w<B
z-KD_C3})g)w|3#Ujc2DME{eG=)G^=HBh54p!+9Mq<KLu>)8xX&*`X5jS(@6`a!Ywj
z%R4kJ&f8hThh1L6Z6!LtP-qASZH+ixn#EK187xCT?QI6$%H4&jE)VwjF%(nBun4jI
zFNwr2eV-XU#~Jh%BAWcNI8f`+mf8KrUl7bdA9!T%q@bwAvTQ3flU?%1<2Uc>J%V5P
zT+A+yil2<t<*VB#y}h1)Z(g({ie)$id+<@vSkkZ$7wYtINet%0&jI4lERK<|uF>!U
zyd8^W&~fbAx&8h-{UFFfsTyoFX?@@X0VN;c)&cMgE<XOoQ8NH;X)M(Sn(81L=z$52
zCd-yU<IjfQV*@(GS6C{)S%cnacKz!h9cP6<Rwohdor@4uAj^5<UJSqUFLAQht!U)3
zZwqc8^g3OYR%C5&Ab4;uB;mg$zkO@~{cN}7orNx&_x>lwDO+^PO<OA;tTrOin|QOO
zAk@!LLm@XWP?;l=CQ&0d-zK=)(tZ)v_KE#~ew^~&(V!muw*0%&=+*h0iqp_3J$pWz
zGjV0r3)RwYq1t-YZ?$wu2>ULGThVL2Ku^A=unN}lrhH=~n}rzbA=CB*L6EXECCK9s
z%03>C4nwJ&2A}I^B}U#zrXvaMzY8-zZS=`cOk})R(-J+Kz9vHQ<Y9LeC;ZPT2JKvW
z5QRN)uP0FR#^~G<@lcn3zmai3kx{)QM29`9cwR!)LgXH_F2f!*KRc8jhoM&DNB5ZV
zw8tHy!ZFD~O3h9VBE)g|D<PmAwk?&K@sRNStSVMI`6ijYbnR;9PKHCde#CAe+$W2R
zANc)91r^>zwRTg*fN73ih<m-xuotz;GuC+UN5`-&bf+1DNjcNCaq}SiD~Y{Wx$c@T
z9nZ=|ZlHBXbV*i=*|kV}sp&|fh_aJqv18Wq{XOGl_~W*t2xhEWNi_k!KRqzS;J(9R
zG`<WPSf2>0#Axxr!r7JYsYN9I3_Nr;Yu;yXx{n&>!*YJ#{@I0xP$Od{tJ`FXBu&Ao
z#}EqIm;Tm2hTX4qiDarFw%TbIvlErMTT6)gKirn>29c^pcd87*DQ%n{kL}S<&jr^y
z#m%-A;!TLcY0+=F-mU7oAT3yFdv@yY<lMYaCiY^_{LY}NN^R)jffC-6fZwbDc{<_d
zYTqCNw5Q}~rVutDs=Mb#_LvPru+gxbM4hYd??-Ug797_~NM|C|F;+`nhsCvRUFK3X
zQr`*Ld?z(CL`fjxJTmIU!d{~eZ~|XX7mNjRag4CG^IpR94Hb>}yI5ULuXWDfQYU@A
zZaFP%H^QczPPZO9IcktoNYnfO%u(P@$jK1fJ(SMT(*DqqSheCJ$j%nZ*$_WjXpD8a
zJ0V)+bR%NT5z=uE_9!*|!!<@E^WF@U+cAHwic?i@a%W?<Vwh%*{4{T(PQU2-%d1y}
zro(P^SioGWt}PSo3?EJ2iAT`g*3@)mNitR4pd9+!6|5X9etH>9Lq8aBh2D3yhCqVx
zXY{YmAC$cM`GB3&uUu|#d!7Y)huyvB5X(V~Q}@|y6p?AuN(Zl5mh>f(l!^)_h+~N}
z1*MzsaqA9fKgWS^QVNjqCJ!n!ZMNT9Oicb7o~b<~){Y{w(@RY}=iSvW(Ef<GTl(sR
zVtUhs7wbHeL;3M2<{w&^I&jil`m7}&=Zk{}Ocaq5vYFt(TGl8i-!L@7{v}W8gZq7Z
zXRlKOJW<php*ez8xSoZ4n8K10Qia!8yXa5hLNBC}z4@CqrNBP#3zctB>fhB@E)dms
zy9~9PwyZjLVrY?P6sHf;<vgmF2p499lhTEy0TUAg@C8`ZA^66=TtQq8X&=#D=28T_
z=G2523)&I%1#{Y=Z=YSlj?*}g(^iR~)YNrN82h`$n=gY!PQM?((`c6n|LLq6&8_&_
zyl{au<sclV1EqoaHJ7;B#a72a_m+Qo8a8HL(c(8Fnw}woghrH;oabx`ofc=}BmZru
z_F2E}(!&n)d^>oVzfd*`7U?)$0F*{Tf>Jcwxf}3J_ZZv%Oeq&!?of`klVC)07&+&y
zb(LkzQXzlpKECKqq);;^?b8`#M%<~6S=8-8*ht*Ho$?=gZ|b3BUu>FbWcw#MPt>!{
z^rwV)!dE=qZ78lPm*BXz!fG%Jus_SdnROu{H#6Qe89R0yEmn~`HDfwOO>bd-`Rmh~
zJ`1>_i0S|4mFI-}R?HKhLn&fm;*2rPIdzwqW<0>3hVCdP!>3)s^<_@oY=*_`#&?f!
zx6vti6i-GL-wW}TYN|JuNJ56~FS38)zQSfZ28$QXY3ghF1AiVw3i1-etX}+TL=2fn
zzOx>iJbdP~@0rEIw-ftJjc{;Mwh;EvG_>%#v@kU%TVA<L%z2#U#3+{&dT*;r@59)s
zKlq*Hv0=i9Z)^rK{ByzkJE`ai$@U_6F-e}&1VG_EW*Atn`0&lwJH*i71C^iLCtb(%
z%NH7c9W}|C5aAFhthUinZYq+A!9T1f_|=IxNhDW_21Xb8)o0nVIP80#T!x2kDy+o|
zL87=c)GO=4|K#zTD`J_~ev?!*Mq1zxXkUUHil_~?P<9aDpkw9TI*qf4#y414DZsxR
ziNsY)JA2}lBcrqJiS*`RW7x1d<Z%&wx6ze>)!q=eUelfK)`wdEjT2_)Lv=35n-W8X
zltn8RL!W6T;ocu=Y&xmWNKKjY=1V!V@%a3gNAyFY2gG=!wOr^5YPov!=-bV*GdnGB
zZn(FGhSmO{W$+M`9B4d~-}h+mh!RE6R?6@Re64Wx+JhL`fPVunyMbbwN+CmWa@Y(`
zq(Y%YZS9vh3E7*i2Y6wKLAqGR=Xb5lCEtWx2-uSVBPPQVj}!wXTzVMjm3>e1mZB_q
zNSvi&sJxX`T|@Q%rv(sFUu1JS8#|>!uwN^E57+Z~20NGQPZ$bEEz`Le*PU+*({&NM
zc2R9kDK7|VP8n4yUSmEhV~ILU;4aYAq}wCoMhVzt5=jjEW6qFsr#-W9DIvO6)salV
zE06wlhiqZwKc5$9uB*_!o<PVu^x}z6`mJ<?u-*10;rS}$|3t}rNHEI2cp{MdK(oni
z7yt8y)*K@Rcj*Rpw;PzNHygaY`XKfWoBoloZuz+d{=ecF%|iI1jY!CG#aQozrgEI=
zr79E-V>A|`Y)Ij|seT7N!sZ}2@iB}3wbP<0ZEC{??>afUGDi4JUeTG2hD)>qX&$%4
z%aDn_nBHvM_#BY26N-(k#Ty}1G~ifgkLP;G$H*cs`-;W2Cf40RlksOvxj`s2yvHn!
zJ^=cZN$XCX2u_pO6`C;xtUNa66o_j!yNYGIk)ax)qehQ*T{DgH3S?mpX{}7BJ76`b
zJye83zGrOF8v0<Z7OLU)$EC`|$mU%ZXNUQGkcm-ufgbE`NBDHDaVjpCbe21W4tT-s
zhHQMXu{I6PqH039Le1zQceaAaTabSohukQVqG-E+;wC&U*-obf*Y>8T5S1z9@su;;
z-Y=T$^hC5v2&?+9$M*1?Z7Kd<>xwQ>JulYh*8FL(fAPy<#3(ND%zSHyNQ`x8t_t~*
z2mx7-bq{Y>K4VKWA(0~?DcdGA1l^M@c?OZ&5jof+C<IK56X|V_3vqcn5jO+e%c!ou
zJ3An1dUSFYVbx{A@-3<Q1fyP^`_eU4*&%-;+Y3HJ#_J6Ys6kj+U{*QSdiI<#fdDmG
z$nep5|JmBO$aoOb5v>lNiVeiSQD(J23oP>f-V{fSL7tao$wa%zSryh+e=Ii3{KuwR
zYC5yo5`D4|sf>_fV^UAD=*@@cPTdhX69PPwT}g6Z$@QdMm%)oSLdQxesTWO=IYv@W
z3<v0%g3=LP%fs-DHJKiM^0QpCF-a3qDR555_Q)8KoNg-d-G3n*MN>z7p24r)cJ)5c
z?ESlmw<3D4;oNE>JD+MT#qC}#Ud-|6P>quzQ+%8orzXbw)vVaXM9XJxRO32Lt~glK
zn91;;++aQOt*ev^%~(hsMd!<sICJObk9b>p?X<JKh}WkH*K3dMyi(Qp6IliJpi~@L
zBV3Q*TUZcO+7>G)OG2ZnCPgQSB{+N)L*O&Q#yG-s#ER#mG8>Eh;HNHp?P9)TU^Z?`
z6G2*xjauX@3z4!)I$?l7%ygNm5KV5a&30lW=)w`Iuy*Ttnd2~voj3lF<dq}0sLPBj
zok}H2#w5*!vZItaXcUKeG4p}5MuY>Q+YCcS7-E4RA;2+)iV8{8|J$YPDC$gkCsbNR
ze*IH*txGRSJKD&V0$H#hMMzhzq<aGyA>1?$QPAk8DeVz!tX!U<B{Cg0TnH-QG(}|E
z^Uxg=-EG56;S#X36iP`D+avAHa2n+$^*~65__~dCPMj_7zs9i&^9em=?1#*`*GNk|
zI$1=`B5cT~yP9hq;Y}*9M6mP)sao8~T<dh(gBz6;ji1+q&>#6z)VXF|+>*r>?X=w=
zb{A!8N-L&iT0>~@1g1#CqXtzI<MEU|h?kkoF}|uGu)Mb3MqGo0F!G3L-FXcUjJTgk
z?--<aFMAXZt$im~6jbBxneC$*`KT=tSROW5lA%i$pGX&KKpz@ycfRxNl0X9?f<#ph
zBiU9tHjm-3Uwi1bI{QVmp8STLD&Ml64jp2TIQ&Fe0qq^{HLt?q*QBo<%p^$=@c{bM
z=cZ_SzhE~A@tpJqy?w?1t?o`uc6lU*?$O_>^k~sZByA?LB?&Y2_%NiAV4(C7o8_aU
zSftf6$OF%^o8V6#uZS3?a;V2!Iar|UJGSJG2qgsZrmU}oj-2fqz3{ee|Dn0qs6S0s
zyoCPSLy!Lmf>3{#A_a?Mr>!IRDfn1yvev#S?^k`FLe))$!9wush*v_!SIj2-5i+8v
z{hx@Q!C?(baH4$Z1HP3ua)tUNG{D-rt9hXEmDvHQ+fY^_*CSMdIX#hSJWnX$%;61T
zWkIAIeR?L*^rSC|RiQ;R?=>#@oR>9qaTr6F$^5Yybxv4=`#0@(vag!6qn$ZFAZV^h
z>zuhGJt8Es2ij$Ng-TiRGa|VsD!w-@udBnk6wU8f|Ect8ym_>!PWxQ~WbKS_liHHI
z<vA;63%6RvD;l%cR1v(w(Wwb{2=HsyE93X<FHt_0ehYv(`OtibgB_=<W&a1MLL?5z
zKbXZU<BL|#N`7U*LlT=GvxpDT(b1Xf2oO5IxX{qmMJfF84&;uTpKp%nxBK#HYRM*B
z@sME5k0_!LStZqFB&6$DQ@Ad_R5M8r_${r&!CaH@F*!GJLC?l|>g^omVH-{GVxpnp
z7yq9w(p5;#xtNv1j1m`vSc$N2?-*SWiu-Ba8nF=ma_5+V#D;$%+T<*S*QlcD@vsqg
z%P5CZkfY*c1MI%5`e}I0`tWd4S8o$PB75(h4uemvHye)d=)&hll-V#fxJ3<ueC)e?
z?f{Nl2nDN`Q6t&s66d$OIHEaoaP5~ollDUtUP_{;rs!E}SdTj|O;4$7{UPZO^K%Gy
zPS~dhI^ie$1Fo+;pD||3rfa%Oz03Ww=B`V3I%z)De%wU+CSP0qG(<Y(+TxpRKz!r#
zc$}RIR$eAXRP_~_BZ?QXvBgDMlV@lZENc(_-yXi1>AC4Ny+b_R)1Z;aIEK-h%doQe
z@8U6<94zcnDZZy8%{^oF6t7aKf#7R4+gx+~<g>k}iA;c$OQqshg&vohh0L}&^@hVu
zIr0VM-w=LK?0NV_n?tqHHrT2>sD+TJUdktBVl%E#!b^fNs6Oa~3jg|56vW6Q08=XA
zk~6T$`0ix;fJ~CZYUMXPB;G9+L@Jj}*>%e$q7sNL0S}X9kmKEAsqWSa#E{@We0za4
znP?oDG>=7B?j2u_zLN#@xr)%eW18&j|AS%nI<&=Uh$*IKztRb@bHCz_n2kt)B$Qhf
zGC%t>$oxO(tup*1<}(`Jvq(eIx0Sw}Gp~O0Xv(R<eBSURGCi=8zc>6qHK+3|s{)#%
z|2vz_z~#q}4Wo#9k>ivRv`ZM30b2Xlq*w{qWa>QTIbv{U%X3oi4bL^TMatW*x5&_m
z%Hr+~Y~yDX31lGIaDHSw{*%*kck>QyR?u<bmz*MzwAI1MK2zCqjsmE?IPz-Iibtzm
zH@XnwZ4B%1^o&eybX~L*xXj|OkA7;4(?><e_eb1wP$muM7E%z4k^o0)B|`?gXu+49
zoG5pll}OmfKS>Q~k*VD0p{2SaD<)-me({q_%!1kr0xtvhs4RvjVnJ!sK}T1&9gCF5
z`A061V96)H>y}1J#bR+^v0C{Hlic-Sh71QA@>sx%q?8B#YgY<F9(dG=H8?I%`p{b#
z5rEe7tzr_BbOoipGsxveiQ919QhI##2*~f!KvW7A9xtHY`xs$`gW7)x;h+R^A+Lqg
zUY{No3OWG4aJA1Qw0D0Xa632F(AMxn9wp^U^c_lv1Drqf@7I{%qI~K5w1p2p;nplp
zG!{nFUJ*9F))!X`MgEXOq-P%9kxD*1$0?Y+3FW9r_=AFFQ*p1MFB4hsXY!frQ*SCW
zD9p#o?2uRPu&$?LemA7JPrnspfQM}zGl}3`=XxcRhy|8CuO&H@NBB9Ngrc--$&AoE
zh9W8H`23RUQ0-G_K0^2C$3KkB$a=*i9@$@e@gOpvytK-GzS@y0VH%D<wz7G42GiZ1
zO0B`!eETZy7q4PWR~U|j;aURHJh=!yL`2aN=a`N7;PW#E?Skmfob^d8wFkR70=$0D
zz9D1j5uho3?rJxlGqrS!6VH~-?qK+&XRo2~tAc*M3g3!hntw=<wX+AUboV6ii-CMM
zVx}5`1lhi=HQ2OJXak^mZEfTp+oLzM*C2W)+-JoP21ck6nu>O=;XQEh*Ec|z4f>{V
z)zUto#a7qRiAccFW$H2W%}7&UH6&>Z=T?2kldz@m7^Im|dl;!*5LV`|%h8u_SSb%X
z<Svmi^I%36<THmfmaCzZVXC`JDG{XekCV%M@}ouQ?o)=(dTR<4aS~9to&?@*cac%n
z9&zZr&A>y|3|}b5?B2O)dYzzxj!TbycR_Q1KahswAAU*2S4yPu?!!9SeolJAdNlim
zq;#;h0rQOSLgE3F@3<J_9CeGGC}YRh{|Jn*F>ARq8-z3OAd+IdJhge6>QiR-|6Rq&
zm|L}HE$8t1D#RfjPar;Y&UWes<z<99f0P+UrxSf^348t08*16FRKpaPxW~_7Nnrne
zK`_w>&tm5Ao$SUigAo@0^dOcW9t3PfoNR;iOuR21koH|=3rjbr&iMSf*oFP$#=aqK
z$A41u{)YR57RaQSM}`rJVN45&bZ`MJid*#P(4Ek(<0epES@1wzPD=es=%Xw0`)By?
zqorSa=xP0aY~Eo-5E@7Cv;w8o@Asg#@&$m69}7wqTUl8YS)yKO_J5^*n-yzD170&N
zNFf5?;%?9&om^RoYhLk32kP6;OR4>4Jdq;c!Rp0gg{*R%FTR*+N_oHZU*Zo}Bp4bz
zo^n0uI?*sOtFCgrxmejmQ`z*EOI_C87YF$glwgMc)RI3ku}SreA2QgtHO+QREnwc+
zMKwleY8R)Tqnt4<P*c(#nVJ4KB^jWTA-y@mE<=wnIfLnG3$KN|QT#9<lHC9C+cm*Z
zNzySKQAg}OEjzBSzd#4^UrGXT#aJ3U1fLCN-UJbC%UzPDbkQ)VeeMpT4f}0doi-ZU
zN@Xt7*n|~NcW7Xmq!_L9t?WioE<JNP-i|xcMDbKb5=c_mFDpphP)JcME;@MF{Tw)Z
z!n{%LGiG@%gOdS!3rmDWGzTW3po@RLm>$-$iLENrRAJZYeXYA}T3N{blj4V#Vy{_9
zIY*Zt3M0M0SF&r8mgeEhR*UQaI36-JH9~Ngv#+NuS?s3i&S&?5KNZFiPuE)>5++ME
z$WtWO^T0>Y{WC;<gk>2O_ksa#)RGaK#KUPHw!u?7VOZ6m@C!|88fr5ExP>3GxEk#Z
zg)=yd<H9e9K8+RWaZMA{wBycQV<K}*E0uTut3mEIzJPxfG{otTL?eGm=C{VZY6yGG
z&X;>3FCQd!BNdBxifDV7cU!G2BOPAJ@UN?_4Bz%mbA<Z$wi=7g(0gLI%Q)uYuwq=%
z#B$b;nfE&mgX)Qjub|j@3=;*8YukHMqoN&(z6rM-XdH^Eb8i@J<ca!(6h9O|OyX1U
zG$|C~CPSV*$4-r7_G6bfyvNtdPP%@E38@|{;hjigAzbf3vo~Zt^C&4;#aJ*!enqWl
z48Q!ZvluV0<W`X@8Ex(NKzCdNLzlHP-&m+<nRB+RF%`z5r`_&7*dteWX#azk@93$}
zQM17<T_3*4%%K+)L~o~bwVJHFjHlODqTyVS7?KQByleV2_FuC~wC6<QHJ%Qq=gGUR
z7awk`BR~y@LSFxzZX#>X=N%vNWyJ8v9&xmIkl{;iMo=~@6gB9rU*a*Pn4jt)U~{3E
z%&Y3WxJ*k}2U#U%aM<a?)l~Q+6I%vcLKcy|;y!uZYbM?}MP6CI4Dyw6g<gjfutzwJ
zyJ1oApM;1Z<8IwcKf0axjb{spUvTm&9F-9TwEiH6cOB=`Br)o6{(51mj(W%7Vzy=i
z0(-}whoBRnG9W;I{o+hW6;OA{eUrJ1Q;v1)z2?&!Dv9nP7qD4RA}N(eop4+N&#1}Q
zg~oFst}NJeB?gYOD0wa>UAXuEv+MpOH;YcLt#lSaATqL^Ls-fV7rOf6bCy22A9|u(
zWLk*RApgT50Y3t&|4}(QYYN3b2lly4StG{a!nNot;ap3D>Co~^dG4``IFk~alQg8n
zZ2}7<<LMWa?d*~7iuQD(2iWv7VLnj!PTAbL5uw)s$}K|SM(7Crzc!S%|GK;4mBpOn
z<1BGu${L`}ZPtQ6j}PJH<^B9QkFZJoZx?D3Hw4l@h($XD{`t3VtXv{DYjrUsM4x7)
z%aYlEJGq#ve@Vknf-i+FNmBTw;r{k@#b>DBUxp!k?T(KOVs|=HuJGH%pc_eAm9ibX
zqY`+>xYVbcG7atg&CgRma92}>xvSg{>m7Ck>uf31^%<S0U!(M4Qg(Y}9v!fUn7+tH
zRI9=}c65!<jt*_=Axpcx{(Q97g^E?gPJypv_;1_OmIcoAvL>$lZ#(TRf7&E@s9L-x
zJc>BKZ6KUYdaH8DiYGAIEM$#XIdt8gEZpAE4l4-{b(G9%_4^<Ny^E>2Y86s&Yu4zB
z4`GZ@bt3~~y;{jc=phQsEdpb@sww;tN3vb(B}$TIe49FJXdUXY*n@iEYlo+%Rca(d
zNN|9JtNIJZyDQ=-RjAY`|5R2ClT!P@TTQ*ND&9~SA%9j)L90#eb4CNE694Az!fMLv
zTb~)tynvY9i?-inhVtTFvKgghQuXgT{J);`Ld6$pd%<?o8F?tPhAFfJ2aqdF$kX-?
zWTZVuZD0Rs{R5%qpKjg3y%(pb{Uh79J;$g>Gx226)q^<SIv+1zN_qX;@{_R<?<8GI
zbEMO2w2WPo>e7Q~to(>V`t)#67KX9${B(>*rS;1%-q^~X^tKx&xqL}#ToG0neRZhO
z-OKTt?Q4eaT2f(h7B4f?ODfFqaw7lxaqV-174C5!rqEd3%(kNRJ|=x}G(W2AMXKAL
zke|IZUkR`KX1lBS?LwKF<E2#Vh3S>5Po6hqC*+?B3RHEk(EWDug{<v%n*QgUhj6^Z
zOrckfNH99*mcr&tI+J#Mqovn>(%=ZKo%s+dh9bkpx!hf+nb(dwxi^w&<}RvJE{YEG
zmsIkT%c&`W2P<-pFh!tD`gR6}H8MmiDNa;<N>acoEPgM_2|7=VUhCt{j!?z!ppV?`
z%bfh7qryW&w|8&Bi?h@p@$m~m#qbrScNQTyh<ATLb14cR-xoK12<(as^VEOQ(sDJ@
zAJ=fdDv$J$^i06w7TtG$AvEN$@%ZiiHU;7S-7Iu=kk`g;7m~B=*!8-t7~R(krih?~
z`G5!yg#k}y(1creny;A^m!H5>b>b>>X><~FwS3ig1qegMeMdqrD~=wfV6oN88X%tb
z65aMJ#V}<kWIUP_Bc;5n9MxfA$vakE@B0_Fx54uHz&5}B5eoVN9}z12(1R_Q?RF&F
z(%Fu;@+uL0|M8oBx681-5Z^mU6-peLqwXvjY2j3|-@8fy-$^4?LRqotNfB*c10EVg
zM~;(R?NhQnqus9<M)7APP8yW@>UN8N=7|0tb?5pZOa!v3;q_58CdM98^n{y)n7{2)
z$qw5o)@zI2df7T0qnRW6UNr+5RGVv-t=Arute?D{`l9a{*TWwz@P<z|Qb?<VoJlcS
z*Q61em>)3%{V{|Ml`ID08ni$CTM#<`4qNCQ)?)t6zF1Z<LO^jeIeC3Cs+T<cMK+Ht
zEm_VZ-XNWUlYvP+BO7USI1jLCH|J*><FOE^+I5%i%9IhaaCs)YQYIy}u5dz>UMz{?
zU+0o9T`idlM-e#PUI=a{#ZCJ4{8^wJCE2^{tXgJhy?MC$&wIKVaU$5s6r~{_(bq5Y
zj><M}O&TvYuE_9%xEGFcI#ybffEvx2c*mfg4>U9G`>IpF8HP<pX66|%#+U)9{4)%U
zx4F60l6R*vsqt-&^L*DgH~A$ca=#`>dxuvL60);DbUpm7@jdN<Jqf(ufFADf_;^#R
zlcFLj7<ExWy1#RG#R3u3aF7TFiVTvtEv;XBY*=SZ=jFwGAq&n1`|gMC)hp3zvV$L*
zpT4`y{(;5o?r6ThmxWC``;69C&s;L?RNOM3UVe6D@>nd@Yb|k`|E1_O|3W(Un>CuJ
zcu-Im6SLIq-HG#WF7!ALiD7cQwLmL9_o6TR`O`_eff_why6aM;Id-H{*0kz|E+HXF
zn!PGNVNJRYf4Z%iIb7H>LwA}IlC`s96Irz6AwFZs4_;}|wI=-pS;8Xsf*%NN<61Oh
zmgQL_SfdoE93vlV!qn`i`U7txz3p_$RkDkO$c2&0vA5)6a9JnbSI}AWY)Tcvuu>p4
z6)ZiJze*(}LO8r?5V})udN;hK8Bf=Bm5h*E&MqEhB~P=f;z~Mg*e8-kb#37m8*4?+
z%HtjnoNr4qHWJ~vT>H7oW1~eos{^9@>KjrMGa7DjJJb<6mT3X`gH0Sc6{A$L(jlgA
zVXpts811T?CL08K!(vdat-q}<?^M`&PL(fs$=P`fNj;y6qedGgMRt@>DcKzdt4UFU
zirp(K?-zM!NFYo+<f8kgL#%a5VL-<GBcG%2f_DG{4kZRhJ!OeG4+-Anp665O+WrBI
zEB3=BQ3_ss)lUN$$>Dt#ZR}4}NdBwu3ylsM1_sf@CcX$FA|g-bDOb>ZpyTAkq!`Mt
zug7~rMy`N-97uOzykrbWaHSLw5CHBQ&&x#6KgIuYmB0~hH7#!^QF{W_$y?&jEtkJE
z&fCXT$@NrmCgX954PcJ>G}e9goq)!1ujBxy?-SG1f>7hHihrK2&qbjWEoKtD7?^oV
zyXjooOT(Dh<zI>Ue{(uO;*7l8*NE2}@YmT$*fT><#*-M}(Z)LzdZ)u0za<2g{Z_*b
z(xrl1Tb%HFLmRaY^{Cl}%e@1w+JabVsFGjuJZApqqmfvwRM46k&4qai#?)t@&i_5%
zwE9Re9oImF;)vSaIDh`#eEN$fU0IL$o7dAXj5%RSaQT#PTCJ>&VM~DlHeuSw`<5i5
zU;Zj6_ZyYdSGhMvh2g~h{yrd@foS0UgK$<IzFR_SxyHq02mZU^?J?Sm*~Xs&(Z7=l
z|FnB1D~u^2<rrh%?I3yWdOGag5Rtx04EfHa4=eqIw!p2tnfL)#yEi8ljpxI|JKZtl
zlLCIuZJ(JHX|E&8x9t@P7%uctp`q>k6Tf0f&K@+hrlRC#NwC)(h{3(E6NlB}*`NQT
z=`6#ddgHA<LwAF8w{&+&cZ+mMcXvxjmq>TFl2U_ohjb1h-Ce@j|MR|wiw}Gn=9<0t
zJil1$UP?I=Bs)*O11`N|UpjzR<fq)vBJ@b-qM1|^ITeH8AENI^+>wQt0O|_b?d|RG
z*jOJBw}?$jisr<KHuwRn92azWcv!^E=G=8lu;F*i+R?Mw=1KzuQX+w%%cGgS7lg#3
z7v*Gl)6GMTV?38mP2VGQTT|x7_Ou{0W4ru_^GB5v|DZDai5lA6>RO9x@N1%&dYs8C
zQ^%`J5FToTVSK8A9os7X68NugfoKBW=^~g`R+^e0ScpXxN~8Zm2SSlSOusfJx9h3b
zRjo5qRKci9xK{4UI?t_>9-Yx~NHDL@ex|ELiV2?I=_F>UZ}qt0zl#*TaaQxEq$U3B
z4aR+E7dlFBpKJY5A+ErZLgGBMqs-BW*sh153dZ4}z%i(2!4q~DAV#((qL!LwNLqrG
z^eA;Gxe#WzC<SehF@LF1JIKd#5K3K$0K1Wrjw9iUXJ%32@7!6P4&#+JFt1JPK^cj?
z@(K^YCD`BmZ&p7cih4s+x(!>A+qFxyD_^Ht&A^0q-Tp1KTyQg3Ro#4dkBA3DXzWbt
z9`Q|XXl#d@mOrw4zgzv@)|cjD6eCLwt|BzHnWwzU5gjyCokUtV^umJS`T@F(0|i>Z
zm3#3Y(v_4ZxgWhi&@ZUwFKL#<;&@9&^2hJx%2v&>NM86gw0GL!Lgv~-*F5Y)f2P-s
zuyPW*^}#NERQ}@oZQa)_^hVjM6goxr{cGQ1y-8wH5*DHXcPH|-q%}t*=jtR9paalD
zc$=_2rZO7YtW77k`583>Kuf;p)Qt>!EPXNsij4l~dM#E?^|g%;;4qp}i1j1TS&V>`
z(RgcN|G|@L%rE|SvU_Z;+5JY6j^krWg@vp5wm=g4yBbpJmtq)$f<RxQhv+lDB>=UK
zt+-k<<|P^edrxAUC=eEunNF=#-U!+{9m^*Dcqa+?P@^#CQl7daU{u{)3Uy!MLIlrh
zmY4iwbKoSj7##ht(H1B9MZ>E)`#lP)(zSW*f19X`^rlkd8gg^3J#V1v{-&j+0Y8!c
z%6c$E6VK`h;rB8~JRjgKiR}@H3ZP3v(Kqe6?`r7JV8Yg?H(OgE4!J_OA7;>13<%QZ
zLuLSt@9VcfeXCTR&*9Ie+?^bbXQv;IRf|5@1VLjh&b!TrW>Y~Ez1`ysnZJW5=;%On
zsvjRMtT`K;P#OY?oV!&UT?$N(>!e&))s##7Cv)9HFfmp(RZBr<Ci3J4njFLmcGu=)
z)1{>)*AJG=XjcTeU3d?vG!N%!d1#|bOxkYm?tb|Yp*d5{aV~DazHEx=)70LDM(hLn
z)c^byglIgOeg}n77eEX!7&K?c`|H;)QyyY$EC~v)+66D!D<D0)zYkM9P9>YqY$A&{
zul>R7>mTHha76gL-@v$XpYsZ`VeXqy$`8ZtjL5gK*Jk(Yqa6M_Gcev<^MLojKUB?+
zSz%HErsho2pJ1DV`Moqf{bgAW8xe1HeA~Is7Y`WQ3b~UaV(H>@B{X<dZa%<1O)(LW
zYtz{<qol3;%^+7v1={8dItj;9nxu4j{Kv8oZ&&n-N$_$1TyTO``rG>T$=m4%m=K1K
zFdeZrW=>ad>hGCC8~lE_ud|>e;3!F=(D?Z~0((}TXiwVK-D==Dln(<#3zc>Yp^EQp
z9T<rSW0NQKL}Qe*CqD6~!*G4OucD)Rg(y;dk9(#z3zd+joM!%U<nAGo`er6ytOVQH
zG|{b0IRxG$NZFP4!NNx}#Ku<N4}M7W+2w;>AhH?H*p`EtUQ>=19r#bbMao)gU0v)1
zV}PfP;t$(CdV>Z{Smqf7-DFK^zZ!r_|4$1rz1gnyA^7-&7UjtAL%hVOh`)F8Kk4QT
ziCkh}sUKN+WtiQnmyAlv<|&e>@&4Jw-k3mvJIHVD$z7N}bT9hSjP`nIT5=skV_xY$
zw6Xcm4VGwHfAmja5X`Uw0-xbdE?d)!+tKJP0TsD&cSo_koZ=!qBrb%gTSk%&{_P72
zrC<`p^b*{;Kkt=@b1902TD=KUcrPedb!g61Bm}tRhkrhM!2&V{^3raj4dz5(?092n
zSNvmRIQ8uH?AkI2VLF^uG>iWBCznDVVnNJzyr1$^{YQWNvL=n8r?5@(yw=W~BM)nQ
zvHH&`meb;IyNo*O19v);t@lYyDSE5{{?Z8dhzHDBq=<5?G05=bCX&4`du%BxBXPy>
zhJpS)M4oPVShg*o_Z1~xe*RoY4E5>&>vsO`eYB1u3KBbsTC8n+0*+=KA!BT>&LM@n
za&JPftO4eFxVi0lVW6Q=M14L=lP4+z1wq13R?c)%h51tFAnP|VOkl}iEH#UOX8((4
z))HyYUp^>=)8j$JWQk_DsjS9<VDa$)MN`=LL_<%;NS&^E(zYBmb3iHAKG1y<_?pa_
z>e6tf{aZ4lCFQET4SDPx03;ZJ3j<)`DNS!EEb8CObnMuLY4Lsh@b7HRtmUC2w%QzA
zddQe?<i=>7-DBO4P5Q_$mRpd??HcJLKE4TT!Gs_RY<a&&VU(IwOQvK<t~*nugDTL|
zp)4N<u-3r5KJZ>-AJ@jh7=A%iUtBN=!KL8iP|(=gyBxv(M4mRQ7+*#!q0t;LG_J~Z
zUgR1_YR`)BflWE+gOsbwpZb;@F!{9`jOl!p4%C-CKj?|csq0qv<#8%_5J>ggU5_<K
z`b`@`7-y!KD6JyTdD55+nm!19Yo0Nmye{#7+PH^^W5WEDOHM<fOw3WXOM4kCb(=1R
z&)c-y#QLf9p*E11EQvpXpjvaB8#p#XMP+FHr)SYcT0iBye5EN%D0o|Jf?;<`*L8%Q
z4m9I{Np<-CH27s1!_dRl#`T;LOLJA$YU-ZVC4el{ZYnc)<kK89_P&LXRHtOLIgt(p
zzQGL4!G!)$hQ(NWoO%ys<ZxMs<Kk@{+VwyY2?z>HQ7^N<(gOdHRnL&%N$qSJ8+kul
zmC><w5nzsH#N(HucP&IJswUe;n3LYe^RW_({EvP|#9uO2rdq_HOKg0nak6y=Z~CI5
zjm-WpqRd&L<v$#hplNe3^$x9XofQrUdF#pp;5<bboAtTJi&J0MYZ4PHZ4uv>0ACyd
zV_9Q7iOs-l#VKw7A4Pla(ll!Jaewez&SR@t=ZJ2g%{MQI2Zqi;`+9dj+EIP({h`UF
z3q61`vBfxr90eJRVa$?8vv7{z+E&bonhQx!2&|W);*QVcmU9NJQr1j2hl19GDS$D;
z?AH%~U@?SZsyL*32~Gm&5WjD)_yWx(um%6A%OW$+W4xrLeL?%Rn_1{fNA)2gTOSSx
zP<qfY>%Cemv>J#<ZwN!%2G)jC!RyP_fMnFL2P2r*#U#&rrHicbu&piRRZgw~{&67_
zGMMY4fO;c^erh+1=8<k1Z2Gr92>mZP13~+K9Sml${QXuZC*%s5_Kl|h74@!qfJM*^
z8ITotGC!ciV9(Nf-3_Z;JKz^rUlb_V1qteODYU2mOS@`@x`WC-PiIU@+-pYnJjA$b
z!d+Ys;5j&)6bx7?A6z6?*`6W*6T1Y8E^stDO%EJrdWB~5*n8sK4%xwr{!BKWN|!*`
z4CvB;7Iwtw!0q|jbzSxz+X#^!nR5&`mkgQc&54+Idft$@!R}J^hsh)TT5@PS$olft
z4jm$XC!|W?Z6Er!s%F)3mPKxN3yE=4l=rY@EO6u*eb4PyC_5PLn|y$kvfOUZCaM)L
zf`u#P?$dZ8qSPA<dG&YTnEiM8qB0lg(w3HIZO&Lx!bYIQRRoN#p77|Jc>~=yqbZM+
zKXJ}<=aemgUOs1>-l{9{E0E;Y0#7_jf0Tjq4yS-3$KSFU;(qCJ1c88<vt#nqk5VWT
z<%VSEMIRVB?-RS~CX%&^ZAHzlG6lSrUt7w24^}GYLdQ)XV9Y0GR0!kd6n2&DjUvzg
z|AiHa=9&M?MaMf+3@h~vW1exuXiEF-jvTsET$d}(6t)tnWywWLfI9UjT#xblQYrW@
zwnmeOkO|3Ul?5sXloXi4Js~xAg=c7j`VO#j+6wsO()S63?!r`A9X6bvCtMdxZu&3F
zJDm~7`(V8Jq_mWJ4vP{%eLidoJ!`)*Ok_AXD3!@2%(XQrk?!bjYx=VCr9Z4kQLNA1
z-Di6HNt(#Bkr5wQb#Vy)5#3L9GZIFr(EP(S{Oc6WHhh!YL;$NIb!mV~ZV}PqVVd>l
zUN@^%94%`A)gWyWrhH;T*yne42kjID@?7t@xtS`Nq!S4qwj=Bpg_zLw#8rGqK7QW)
zT2*=V^l=e+Y#;bs?H7m)F~j5#mc<rhxU!U@3v0k_ddKAJyT6%d)83I7dJ(jX(s#z|
z9~x?eZy-6~Zk@s=VGh(uF{25xyNbDWXXW*l*MWlHCVyquHfQI38DfJ4geLVcN@8N$
zKI9(V#19V~4Gu6D%dU&!>?o5l7&+ait?vNGWI|dB_~+-wsX`?l_J>32{m!4`G3f9?
zU0H+U?mD!~dYf#H@vf!VKD!@|)Q!0__ww%+L`4`Cel0~hu!e83<|xQE!V!F6c2WTW
z#2-_Dxh4?&DspsmB%LYr-*)A!lNKS|Z16|{V&XmKEt2kLiggej0Z7eSob<m7{Xxk}
zM|$ZwwwXOfMw6`y6IPe%b)iv1og1O4z#0BAwyl{f5dRU^|Eo<e2U`sD>75A(rMi?1
zw&2TsR8;9HE;f$~b~V(7D)`UM>%asZ&;#zxwSen`VPCxB|9w5;i;1=SxEk}<N|NBh
zNTcVnLHkFl{GhCpsJ?e0uxoZsEemx}n4|Qqt(z?^9Mx7bNeWXmeJ0|%>i$em2QM=5
zjzQjfj)1fR=SL&1j~zapoC7v{fs7L>)u|goACC9Oo`DD$lD$;z{vo*<*ap&&M~lkv
z#tUZnP}WvKO~u|Hhy?n_$*|gjL02dWM1p^G5U1Urf5q?^3uGaFLDW?LgCGBj+KdO6
z!vI~zCYDf0X1~f6Pek@#mmR=)za4A0s|L#HK<T`T?@W=qBdj5paC2`r4*19@fq%!Z
znri%cb>jn>V0Hqkk|3YgwCKh7uh6OAD}1KGjcsQ`z#6G^$5v!Hxc9Od7`ae~2EJo#
zFP)k3NzazOw`@;c_ZVny8Wc2^l_9;WaAqKyF8|nJTmcDPaHW51d2u_GktOTeE%$1}
zdpMM0R)1(o65?0398BK)Lko`KLFglqhSQLXTXOoBCO@=(jTTAgqyU<mpZ~Ai7Vy*d
zV`5hVaS-NN93fQ&J+!Fj?R;iHH}udIw6*-}FmTRmA^Ps+{D6*If9&hoH>au=agXFG
zbNUjrH8~Un8ln~5R+vqE*w(3%qi61P$mU(!D{%y2Hqx3%IITe%%27>2%|*KT5mRIp
zfvHL3zppu1F@s+dN{~h3^Hbu?P1YsBS_ajU{;1O8bgT!vDSel4Cb@2`518_Q;wfru
z&>{I=NGx;D`d7ldfwNZRoh@VNK+P-h)}=oK=O6~ZCa`|zl{zc`XDxt&s;oq*U2)Kb
zf$l5JiYBbTInv3opBB?dwT+vQLW1F8D%n?bI<1Aa2C>v9=Tk&vgI^&Tt*~k}1(azq
z?jJ+FF$6veiZPSx34bvT){Nje>^r7U;VF1e%80EB0zs+_Q4+{qX3DpA@u8bT6<ED5
zo{G5!o{DQ5*u-RXE=I_&)oFpbYSanI-14F!yJa$#i&1~?sBrcpt(p6$J|C@yu781i
zU3<;{n-N>Tw-4Sm0XyZZ4B;p`#V3>}s0Z@`1C*zG#mPswkQsT&>tJ$^PRDu-pL@_n
z)A+3Z^2p5!kDt6#)bs(+9{%Aq%2rXx0d!gbCM9dD+NI$6c5|rTv4?TBHe*KvEz+z3
zdW{WiOg+53%z=bb8;6tNZ6LMAVy(^f;{LwT1}(f70!Z{DIH0kq3+X%>y-o6>#`I@;
z@7)n!SC^aBt=U<dFimiJ)B*@?HT+;w55FhnC;{&#kfFg3g*7N0r7FVxDMuD+5)}%<
zR565|pPQg8p*l%WyeZB&WdIG*%U9pH%f<Fo(?+Cbp6vbE^*4JOXA;mqSmNGaAg;J-
zl0!8VUv^(SX7Pj7F>5z90!ute1#<l8lJH4(LZK&TD&?F^?Y%VG?|1I#h=xgSv7E;F
z%_PrA4-#O7CB&wA?L><VD}Ef3liL+NJMe*?`NIBHfSH5TC_iz@6&F)>7h73(g^r)(
z8>qSQ0TTS<jvW3NOl@Ij6mfq%I*NoeA#vRd{yU;7REIuS$oe%?L2H+3!HuN@Bm7^s
zn6#l)R|56w5KpbAV0urdzoUlZ1afS`o34aJ$^!(=wP`%w5n;tBu8}tqb+)e}e90*a
z=E?fQE+rnOFp$nCLo`oE?VHQ7zmQ4yU3b&GU6f0P$E;&|l(7UhQp5nDrca>8fE>PW
z`i9;27vV^P@URH{OEA<<Up#QHgf?qIB6QA#t-OwCL9mYyjb*Z|q>yH0>4X+aj5y+L
z63RWcbbRYqp}jzL@YC9eN6b80_6`m*7*7(6^jcn0VhX_AfN|+sh`302?**BFI)d*8
z;cPyStr~-zCPb$Y^b0^Q)EV**{HhgY3gf9QEmBY7)Kg|0F%g&z$Iylq?^qDk-)a#E
zP%t#8!iIcg2u=M9=-@(TkZ(4K4X0aQACPPfevw=sN&8ze+_FFYJ@R{(LIVt|dto_s
zws5sd#}a(Lng(-ucl!!W&&lYNAS&7<Gr%FHBz^yW^bDRIwYo%=qZ*|D6@+FoIGnl|
zrPs}R;{YR}_*<p8xXQ-gVu&mdKmHwD9x_qmE(63={_l!IK}o_5G<U#61>wH;Rr^yu
z4qbQAtxx6yrAweNzERCVHN>0m0kh5O#~Apb9v&}Z(-@j_AGsm1cy5z91pY3V(k<jy
zX+X2o(^Ku}P30>g7{D%IdTg5_$~l5<MVI1{0#IXu>g0L<<_d-V`hZIlGOm|(eS}U6
zz&@4EOS|s3T0Tu4{>7BxGXJvqUP*bXOt9F-iOu_6b<9yqXg+djKUTEDky>Q|r}-mX
z7K!(b=7l(&=8|`xv>v4=`s;b?Y=;)t&H%$-<+QocY!K++_G7cU<^a*;$C<YAu(!D0
zFhyjMA89D?D9nz|+^pf2$OCb(-Jl^?AaY95%#3O>Upz4emjUA;(7=G?zYG#~qxFjq
zcdL8A)Z}IBxxV`eWqEm-<zdPc4dz9g_D@3qjLtVe6LlKM#V^(S28c5^9wA5F1x>PY
zZus?qG!sqpIdhCvZu<Q2K92+_-{@C3=3y8QFE$tdh~-HsLNE$U58kiZN^)!tI0R-y
zT1KuZ=uvVwM(msxeW(aiR~<1Zw2*;Irusqi8zxj(#$n$P4+W(Yu40%3f?alRm7$zs
zbS9nf#nXq9$Qk3a-D=5*hGCUEs~#hMrB$kwCTS2eG-DCWuiSigf^<Wpmc^i}DV4z~
zayYMnZ>v!2AxN44OFmo-*G^Mx<X@ACR+A;2{<9J%><vUmQPOaV)^Hec99icm3b{52
zvF@f-D-VBH3V(%9+Tg@I(R`M5tT#E7a-e=V8(Rp*>yJY#x?h?Lq_lI`VfyZSVg8m_
zX07MXwrpX^f^RV#qm5hyRN*zGWSbNa4W<zZuU@^S9HcJ~u=o2KMX{0*2-BO!EVb(%
z=Wx%k_r^O9*QZVU2?Lw#GTTr6ZvOCcI~ra%fTN7{fS)P8bH;%P7qyXA?nz{l%<6<x
zP9d&iPiX)4cb_MdB0}FE;3azW6!X^~Ee<|$3Xu<w!Q>e#bcei#I{@*tUH34cv(9T(
zZSAD3%o!VL0K^9n``~j`Bda(g87<F26*VVnI-jfnKgA%luoBnD>*ocv$P$H*-REFl
zO$AolKOI$l!q(bv_gj*4XPq7*hh^_{XL#YCllztInyh&n{XHT03_QHYtD(SvAlqvC
zS$q9xg%|EwiCaANw~_mR8Vhh5>GaeQB4_PXWVvn->ZkoXZg6bRxv{x2j>bsd_gfn&
z6@rR0lz5Ne31{kE)7FiCa3%EwFLS4pG0Xo$JC3&`II4^Z)+^x2{p*6L<HZaQulG?I
zj@h7si@iU8&{fUTg)<<rOf|(<K!cCNCl?zP5E~h9%NBwbl&Of?ffM)%gDags(`V&3
zUcjV^yNbSy-9;>-FMRt~-_IxM^eGrieh3eb=C<qLElE7X;jtn?)lzxP@)0gVTfqJm
zK&{a|V#ucgs0n@~Der$tp^yxkyQI9&nVD&`j_I<q6vw~V@ufX=E5oU9ngw@~k6`<a
zlm1PdHl;Uq8e!$ORh?s_Ms_cZ6k@H<me2J7<SZM{8(uhKOkpWpdu}@B3QZT=-@i#}
z@(5wCSKHBX1V-Grqa*_b@g>61!84RneGx1<QU58^V?wi|a*YZVWx90lvxZwuC-#30
z^&S7OY0n8$A5;vUdsp4<sl|swwcxKGjeX&PA9*#3rBM;jBW;nZVH%&)=EY7A&py=U
z3<CfGaxDH}dibA)78VMAv8dlc=>%3RuaYa>oyl=5M(D4J(Tygd1h`VT5Z+YV(EPnE
zsJH}u^CSV3oUw4UpW&m1PQuRvw7Nscs{QEIM_5djG<r^)T||-QQGLv1X*;|Tg%}y-
zf&X*|O8kIvLS;WUO%^}9M-S%33q6zb2}U*#N3`@756ac!jFARnckooDR<dg@1aO}&
z8w~;?SUw(19AX2PE#YKTiRcD!6<~2p-G9->J}x{2B{H$AhJ^)|>k)V;S3jU{7p0xa
zRf~yf`eQrkyz2LB{5YMAAjPDq@shffzqs-i|F-B_s{2*ANW^<|{pdObC*gXU{Pw31
zm-63$K7rk-bU%T+CY#x*LxLAx%c!uK#K*q=#y9=1Kfe(9!#v9K)LWCyf3BJLg##R_
z<{^VKZN%?n@t62j=gmK>zN4qJM=)wGFONo}{>0SiAo{^g2{hqp(n<7iX~--xszOsa
zf>;m#YMpyc2KsirbRjV`n_K4lc$>P8EB}#kID99QB&5CA>>75+)nn7~;5Z>k2yslo
z>ukj*F91jJxKeL)Ll5FAp^|;G!@x1Zg3xyihk#I2QUdqif;t%x*#fG;kyM7A+eMS1
zt-dHuyE(=oM6=<rlV(rA`YQ}*Zowc{R@UeQf_7pLT;Q+&oUT53a`N&>Vh>Q*-cthu
zgG+!U_#f8O#@4pCLZi6Bas-D(uK}c3>Y{e(vYNr^(XLLTx<g(7vQj|LJdde#!lo9@
z3TsnsUH1#&_PTlKpRO$(aWsd6rx_&){e!=YO<5|#%NL;>R>uPqMN3GR9gTr-H3*iO
z|Cd#?lTtyqPUPN6DQ>4?!k=EnX7KCKI@1+WAux*_==$E58maaQ_cv1#d!%XbRh`~$
z`00fZY}7d$!nR8$v5}P#3i6irKhk#yoo`(pD!RCZ;SytMI~9(e*S4;!i7F*uUMd7k
zCkazzh%uz2X8)<P&(d-*EdNxP0ve$>SH_86*s;OqyQPT>(O=D$A@(V~jh@-AATdG3
z7)xtdWf0xhA%@h+r*v5fYq)JGCS<xOF%LlV!o>eE2JP&86#Gff>n6o|dg`5$_)rPY
zQOUzm$v?-Jv^mFphoZ&)Iw^@AX=+;fxhHEtizQyiu)d4Rt6hRaAP>IJu^bYEpX9Qj
zy^s-`W+^gYevd7G?~0k<D{CA6;Mmd3!@&C}0awLz&_iht_40$ILmQ(dB`G-jGa$Gi
z-oy6pV0!7^Z`;>zBP$!R16CX#{s9p8ZKh-a`x<(=*+7Rp%Ir)9i;(Znen&`$BfR0s
zVn9S2dKzCm2DmO;;uvqNq-LqPGNwU-nH(@goc%4!c7i!KZN$RC3A4@*BmihXHS_1>
z0T+@Y572PF`_-S;i6!~m5y^oufxt1c4G@FEbgKt|n~oiboajk<#w*(LC1>xt!MX9v
z`b3QOduC*N+g#*vH<csn_A5RCK~X~kmO+a{@AZ^8W^izD3%kXz`&m`f>3v4D^E*WO
zv%;?IU+`_+!Bb@DEaOGr3`}IlrS*`mznn!AK4pDJdC`zkvUk$YmmY;NwIo$wSW4+@
zjGC0kPke8wYa5H+Oa`GKQya`XM}h0IkYZ%ouJQU=KVACwVi9($!s%yF=R!;JmW5B!
zb#E|;kRK_|HM=7u@Lj*7FdI>YV`<1lvZf;OI3aS)IN8&nBk6%>w?E|4E952rlU#``
zm|unA5e6tG)k6S-`0qNYf(gm;LfeWpkF{nvki>N?GSdwmf2yVeUq8ODxSU9LO4B*!
zSDsE$-!IH3xyZBinaV={vJ(s#7=Pl7#OY$3(^X!p;NJaW)Iq5_xhwhfOCOmfTLSqV
zl|aH*g2*cFyvb>HjYCC#0P#mVId<wfNazc_pu2ix_Le&0rj8yLc6?lN0Xb!Ycgs>p
zKe?iNHa=7yTsJv%-l*byl#h3NdE}*gG~Ph4ftO-zETko4hL0!r`A6bM!D4$vn>qDU
zXD;4sgKFvhOooKRM`}O}i1rTX;iRRdp$9X?-vMklx$iFa?%p0XJG<K1pXhB!C@doj
zOFtlgxCFv7QJhw1W<tyJ{7__zvH3TAXq>tqtsb7Ai|gtz{)76m8U2d*XcCC8CVD&9
zu$jSmlT5r@m)r)8M}x`xq=c)8&+lV+0$$`DakTvtg*ZDk&~&UuHX~m67sNu0*dLYr
zlA>2KnlRkEHq*59a7F8KXV)|L!3M-(&C|e~K@cBKL=Jh5f>mFq&lzyRyZ8}hD|-4>
zwYmIZk!p6G;6w745g!9gc!b0>O-)8_)z{-|y1n>9CsKD-XC)sHzbZpJsoKu~m_WX_
z!10xzA+Q|5EH2M1^gDe?33%~`U+D`9!^2d%qopK!e{2TZvVNwO4#^5DO8_=cKXByx
z3{XmMd7ND$D``TQXhNzd(%{sY4+PoDhZ4AD@NH)%d+5CjctOG_q|QUvkl=(==ilmw
z-?n{^aJVtEs?$7r>PBzJuZ1Px<uxLa8?Xx&k3#wH)bFFY<6;=pz;Z@SpJHwXip9au
zG33H=0chnsMEKe9ou7&V!zN1Qdf%InB_Kzf%tXoZUK6W+v(}6c;nr(K|L+1qSXJ%-
z80>EV*xO(=LCj*<hUM$)+jcoff=xn#Iw|?~SA~C_c8nKj)c=Dz_iYcvHUeMjQOF?x
zES4A{91P%#hK#(|Z~LZjad3cOJa%Y;@EHU+KQv5C65HC^%(~bu4sdB_td);;=e29a
z2&=oecS<~k_0H@C6@Upf0hu`r2vB<}IFEdf=%x%KFN*MBDEa!!a9JU?ETm3XEjFk*
zh5kCP2G5;gzdB@F|9*W=B!mQz#urs;!2<qd9G!Twv<JW{m05vY`8)q{5MyOZ2k57y
zPBJ}b-vx#vV^}H^Q6M<HZ}_H(id-<-+g{w|<_nUav|T}hYQv9|A3X1B(U<=5<R+X{
zXV82am3GWNrw4&jX4)+HF4V9k{r(X84V`qK351qN`k21`>@hHhule5o1&g#_2US?9
zAfJc(1}nG<<*wAXZ{#WzhhA!~6&T9D7gO&J>ndp9kmz0I#^<qiZLLwU%4uq}EQyzX
z^4}Duc**}J3BdDO)kEbP#3#ny?W<?hvkJ5stM4++7Kd)u(>>;<2yuxdqR}ffOEcWu
z*0-oL;0_^=p}K+YauC1m-6#dA$TvD4IH7ub`2Qy)G#zA3BnQlBtOhLwB@uqw0ADl~
z$UOMEvY70Zrf2Mb%3CpS_g^58jRW4<&W`VE6kIQ`+;+HqE-0Wh^j`N|`tcd{Klb45
z@gnAGAW>hq#wCOtS-+)HLn1T6>yv?OK=>E<Pr$Z2M%WePV|tnfoJ>yeuimXlG!htk
zX)>CmEpgB}#o4hJDT(?;=>m#JL?2x7bpN;qk@y)T(st$aRUOtQMz?pNb^rlG!uZ>c
zqfu!af{%rF4;#OPWcj!Ev`n0$;k)H0bQg-Wc1<WVm4$WK1~K)85n(hFr!vcU5fzTd
z^(|!i4$toAwlM#O8C@~156y=oG-sTLnIe0qKmRagfp1kj{?fW5snFG4Tv_yb`mdz}
zBsT<m|Mzw(yC|&GGx^c)%0ru!E`H!@*KSh@CVftE^;PP0S{~FHJncIDBF=C?o^v|5
zKN!Yx8Kj4+RV$90aTFnZagvOM5vK~mFexc%iy<eQ_KB&p)nRLJ%5whrXfQjg0#CVL
zmlC0uMA0>A|MT#+?Q7-rn8t9_`Rd+o`s?Ik4Xu8$F%IVxX$G9-V24K~VBp}~O$uao
z<RNbKyR{R!fWRIcBykj{CZ{r}Z7(%iJj~YzeF*^#iQ3lI)=t|n&L*dr3+sru?<s>I
zGsSXIx7~4Dy=fk^Oz%GUVA<v?X0G{F!x&-A7AsV1;fT2OY1{wXRvD6kdwaf1M8?N}
zsT=n+JS@@%iI2EX9%OJPwKzeSSxXLs&zLj0ge)u%q5%=~nJ+}C)rbBexEbE(G&{66
z*j*sWb*<z1ls{X(au2WX@<(W_y|>mpOs6~OAvhm&--9ah#0kVwks{vgQu2hrgdLY_
z{B}7)+C+NgCOo55k<t-FtZKOT6u|H?Eey;g{bR^PL+~);@FWJfK3-+3986L^SOBw2
zXiPeU&ePiI|Fi(9)d*R7s%h#oNA+c<ne6Vh5n&OTnGz|O1eo%Hd+k5<AX;e)sm9%e
zVxsglROigaJJ0)aQufKFZ=r5n^d$&y&RH0pKHLpKLN3s-ix*3Pt#V9cJ9JOaFd9<M
z2VO<3SZy!P`1QV<9h~==A~IxHJ71Gpv<Sx>&CwgWz?l%B0?ZPAx^d7M|7S_V1&lWT
z<vX`HuHN<am<0dx)YjH6sIErs>gvMiF)VFueaGJYKuxcf7YF!mVphs4E0IB1xVW&T
zvt9}o=G5(M#@zwFdgo(=<bZA5e@`$^!1X_WlN~shMl+9^Ee=ggR5Am>=A`M{oJ)GP
zmHO-lG)}4#)K0sU+hzq@ey^r!O-6ZR>(6YSaC#}k@FMLgRr!Q<h4Zigu3(gSEw~OJ
zSHWvY>lKNDj6hW}b0JWUJ#{yk_4jAKv(tI-1oaE&uIWdri&b`;#?>_$)aMo_zpg2#
zGCNt{Cp;oF*%mmhP2|eD%rFw)Dl5lBuZajFwEs0IAzFRbpDIpMWTy5p{|nj8Q&d3i
zS+xR_)PbG8%kHgwdp5~tH7>jSh1|{F{GCWnlPTmPhnpIy%`;MtO@T{O<=X(k(gT-f
z*xsq=g0Z-p`;fy?!O#1d1PpMHCt0l|OZy7-zkn@CNK_PX@_`?oN(2(wowfAfUrvRe
z-6b&6#+L$Cny=+=1AnfG3Bf%~5I-C7XvqQcA@m+@Kp=hLX_*1?1Hhg^53i|eV!f=b
zYt;8laK6_Q{4=You|wqk$~R)Cr?jLtD1NKd97(!l`n3drO(6>M=(4()@*iX-3S&7B
z3@{!sK+|h$Xw6ya92k5l;UOJXPg{RdsSf5pv5&#bi8y7>#*|2~W@Ux^wGcF{L;axA
z@z0D-Vq!<`&&tfgcBJN^*o`dI<r4@a1HPaz^~~T43?Y7T-*%(uaI}{OR%_xwri51<
zirnZwjhULT1qc=OaCC`_==KM6br%VTXmBOAp9|@9hWB&m4|;qb;{YZDo%hpAOK^yM
z%)?_KD_Zxqr2<){RvbIJ#5wt{^<z_cYi|Wg=Ir&JAM!}trfyOsSy>{uZcb;qMQyU3
z1;=OZ`m}I*-cJLFj0I5Y&*(pRTbrbqyy>UHaPeOsZb|~Zl`tSJ+#rv2JptL0K2AVu
zu23!;D!SiCRYsstyW8lY47_4VBABcTCBUBwtGw!ErnnT*rXXvmpBDX_<t?Sibl|DX
zF#~99kWqc3MymDO2R%?C_KXl*g^x#*Q1IGnS=Hr6G6x~t;8NqI(72*41>V#PT`{PO
z59yh0W%7bHg2?559OJ+r&E~4Nke_uNGsMTJXl~kzhgTsx)~~vHgZ#6n8K}Shlug%@
zhHj}Vxbt`rv}2(opnsM)s*CS|fRQXIYsp~4X%T^7cflu)sc$4u5+N<r@yoSsIAp<T
zDA=96iEY_vI^KTh2|BSF)r-{GAr2$rEN)>#?9?rUZYk4BEwt1w_{l?p5_IW)z}Uma
z!~5x7%mydCo|tZoQoNmqtpMCJ5vP!9Vg)6v$GSu+TqRGWvl;PuoYbBG@U_4MRKL6X
z`)?QjfS#HCehZ$7jcqVb)c?PzWoi+_08^FnVukNLJ^yf&u0d@xHZNP7eU2tdP5QO#
zljz~0^avtTV|%BN3l>3SVv%;e*d2Uu+gyQct7@pJI1QJD$M?L2SW5BNC|&hny2Shm
z=e|YjaMvO(NR^C-WfR6RgC{j#Kh*8HLe$TLbVC#;@@g+M7H0z~ZLV5FvwrZ()ku$4
z>w=9_Im#24jD6do%ee9!JOwM8_in!BGm2B2pK2tpNJ#PLx{Mvf%6KTAM8@Eq36I!#
zh`&{fWGRQ=$+_+_b**;O4;fo`2h2NeQ$Kz71mk2mdjP_Mjcwy{>|XM5%Ix}+S);?B
zWp4sRb2r6$1K6muvJb@)fr$UQcBIgeZYetJ0iXTyC%otcW7pDxR4txp)R{L9f~+%>
zJVfCBM?!B1eA8{JxHwDZ@UX>)9Ub%cAWbkvmXK@~Kg(NYGcXHb4f}^StUyS}=<dl#
zLQ*~ed#!)@n2A+ggjyOl#9Qpw=;81V4?~$R5FP{a^@L4X;ZP>%GpMdbQ$t2O2`q@L
zudF;PQFyVw$R(s3-rpe63qy!|sNokEJuAeCcvJh1owwI{SM-`#liEW*)AIS+;rynt
z*DnoN$i(zV9=JBk=p+5Ls~+8djodS|Q!j2eSR+m1hZ%bR(!es-j`SSSZ|v|M4aXDn
zk;!+L*6yU+TSL-sT%JTvd#}%#8z{1U$N<ZftE(&Hdm8U|oKMegg?|n&PJT{N19&E-
z47T*9dPN!MaD{i91ZAdtAjWaOnaI`dZ$Mi`$;LLhcdl#%M>>Q|9=^320S7`IvT1OM
z`}^;A?+-KKY0S3IcKo`#H$^3q@hwgmp+O>CMYqwI+&7M=dH$h@7b7F;)zF?pZ+b60
zcpB#KLuuNz26Z1;F3o1GDmZ%)D$=?i1*;rCZfk}UlM!Q30R|xp4|z0xG=<X%HrA+b
z<cqxc3xB;VlyfkExQ}h+W{NB!e03M?LjgLS@b*mN3jB}`G4>XYi`mIi?j?%Vk7_59
zGO@(Gr;09Y2!(?ThtTpL<j%V#ZIhVmd_uVnW*lw;T4W)eGj(GGep%JE%no8C(X)f0
zJC;^)7}X3nlytuX*FT{}B6nd6`04~uH+OzQ{~{GyrAyb6gZg_QGW7m@m{3$V@m$u<
z^6lU1Of^JU?)S|4$6IG5k`lS0t@UEFhrdO9(g^5vw0g2}*omxmL-V@$%=+4ZS+2sA
zMt0~3Yj>zGbqb?h{wp*Y&#YRauhk?L29nGD>ta99j8E7~1@8H>El{5!je1df4_=#e
zDAaup=16k8FOJdqTT55#Vwk>Y6zZ+rT=P9c&(Pp82a`<H4*s?|L8u-1r%MDwTkOid
z`KXP9Rr{?YFWdU^DxK3Cz*+^aBU=qO8c5(>|MvO*=JDI(ng37{g$eO70F;OXKzjf#
z`M*34P{`St6C5Jy$GMRKAPg{~rG@W5Y*z{*x>l=%qhsy5qo1FRRF!5Wf_5KfYar|6
zzw=<{dvB9~OZWtG(XSO1j1P|r*?JrZ{2HLyR1yEl-^>NWudJdWwfi5E_`&MFu3|*+
zv-e$;S6SBay41{yVHQP*HfJy!&w0u&tV9`bYuv#FzFX=c-Wo&curAli(3Q)Z!52pC
z2keS}Svw&3a=2z$f8$P%h<@s&2bV8U(Z@}DK#se-fRca#FXT^3!th)fl5PD+#k|&a
zAM{7Yc9s7Ij*8jN1$msD9rm>`@r+kVk!1d1OpF)sSeIG*yv!2z#U1II)nh9i38%Hq
z5qx=DwO)0(b}L2+UZ#kACIE3ii$u1NA^eiref0Mvy-Sw?wnf|ZLmANt2yR`0*l5x+
z;2tB`&aTZCvzH|}(>@psPO9zCcYumFNH6hAWaQJFzMV0FOOxSROK1>eypam&B88X*
z)6|5MXi1-#4?w6Km%#`!NfgGn*U%l8xynb!O55Rrkgp?47@Maiwk9+bUsf-2?CbXz
z1Vb=@=I43UkRMx5SQIJe-CiHsIS!7LJ*|mq7AIjiKJc7l0);##a!BCwo!P)we7F4H
z^Z<Vx2J~`bE&lT<S3$hy=TBT9<Oo12uC-y)$i{&{(sFXJAYx+TlcV*;`FSu10KOj>
z3$5J&zC?VcNK1g_@P(LCE(}~K+-8e#u|@z%?l-f<%0zg=QOSi+^dPL9R50P|zy?2{
z`1mzD{%P)2^ufyHJ)%?0N2mzBWU@RGrXGB{@19-*tbfXUt1`@%Dyr{Z0}Cizg2@}K
zP>78Ay}@8+u0WtMF|2`o{egPWnCf$(1r|4eqmBM}<4&C0zim<(9PF&Mj?(#=;lcU0
zX^CSBO(^XpCo{qauE~I)jYixTels2z=lk`Z5o`!2vU?crXGr4y)1bgOVJB?_Y>zRz
zVIAUE`;bXvN**l245YYdPixHV{Sj3_#tyQUfty%V4P68_Xz;JCKulqW6YIs685h>s
zSMVswDTTEwU#1Hz05Q8Lj5w2(yQXu-rP=uXyYiG3P`}r;L()7<_XU|FXGw;eMBXFo
ziGhcIIO_}pCP_4gWKl2&kX+CSgo8a?CI-j(pS2?bNvhFtacR#yUYr?q*w4?=(MW)4
zOR{>t=+`ge%F0UJ)&IO4AP)P58en)4zE!QQ#4}gzbQ=WB4_Sbs9hwV!mkKfqg|L6q
zOi!tYz&;Ea0}e-AvMxyWCb*}a9AhFD5tSL9gA-oBS7z;QM~gA~wEphjK#!&*>h1sP
zCiEy9Ol#V92BW}AZZ*Mqz>X$}44CG+{2-ma6V1~>G1U|T)bB|Lg%Tb){R0j)v9Ts<
z*YgUlk{`Q5TmBtV-<#lCJm8Q??G#ol5g}Hzy<SQpbzwhnQM9?UP1~CMo&9P|M#qS_
zx}UY>?@~bPUFPUHM+J+hc+v9<m(jkVXT5nDVoov@j4Ki{N?%ZyoA<AB=E^I9KzA9S
zCH>qNxtPX*+Hp6ON27dEE|Gz)W&mr+-Mp~~!2wIb?)Pp1CymHy9G9PgYlre>CyW-?
zjgDuqeEQv#mfV5q_xIfY%_4xM|9@OZAh9oQ69|;@*~_#}%F6oRg_5uDTYlGbLSohU
zE8=Ur*6e$_cpUwMLT{{60}$$6OWpIT`6q8rBkzx`kDlv$1nWn9Mn)ZB=vWR*2Xqn@
zRLBunDOj`qWow%h3^YtyXSvE_yVpTo0hN8JnNgN*`|VSLQ-T?kR8+kdzIitu=fAtu
z#2!n=#NPwOndtpm_|lSI=gZaD`m3P^UX6Y;3SgHVIIizWKqU7m_><4i&u`Hei4F>Z
zM<uv9tfxcK>1saw)2Wft*yfbDRw0mOE|XcB)~oY}_E;n>hcWfUkySMpF(C&%s9_*O
zR}KE|UF%!TTIV?)e4Cq{3-{*DFCP|l5Auhbwr5w*2}qxPXmuG3>X2V>A0LM?ANGoa
zgiNjoA0>J8&Ervj7t@DOz-T!AeP+>>hH~!HDJ1mI$`FMBYbZ!-%D@Oq5T67&P9cNE
zHc=abaS9F(w+5jae8PG(a-hBFdml2<{=?RI$LvpMbys=FqBrdRk&w8U5PF+$f{2`2
z?Ph)8Nm)1~Q5WVtfO#Yd+2f;RlEop<X*c6@X<3nBnz}>PwS&2+l_~70%*BIcAx8}G
zS-M^_gAJ+t5cvG`b+%(+R9d<#eag9!o15FL(NJJ&b7?8M<LsBx^<u}-)dp@tL3jtW
zPv@i27tiB`EubNj^z##1sQDBQJYqTG!Q}tt^qDp5b6O0dm6Vp=J_4Vv43WZF;+YT$
z1v3}0!LCozWZXZdL$x!Na`DG&+o{>>DZkvp0}-y6jLN*+yWf*&@=rFISt_2d$moF9
zx<Q0uSU&b9`y@|wcG)5OEu-h_r>K5j;Us2h5_%0KwDOcYTKu;*Wb~M)M$5UpDl0$A
z$$(E>#ohGTTh9@47lv9|R5WoV+HK=8rG&0-^UF%;q8c5feZM$^RNQ@EKTl8BWlj@1
z$Y=AFSUM|o7t+L1ELJiP<1mr2n~+5>I3g&dQfIF~^=)a3^(VHTAbguSiuxyZI?px7
zEaA>n8hbSBtI?35P53h-=FGzlyJi?);#3_FWsjjMtW0KeF2Yw1n9UBfr3MP0H}!qD
zNP2kS?AGJ{P$$#X>&g8WE}{eC(4!|fPs~>C7Zd$NEOUFA;13jr<)9GbS0qr5pgY2U
z0bSeM+ZVG+?Cqo(+1ar`Zb1KG{%0$zZ6K3H0*FNI(JkiR48Z*_9|V|R5NO6%Gm8Ap
z1HZO?T3w2;fqL`9iTB+a=FWACcpy|b^$3nDrqlc>G<?1i&j%ZL@`z{iN^uqZo=E-)
z850gkIhsr??_X`OGpDioEYs(3*2Sa}?+C#(K_oU1N$d*jCXG6uSaE=)X0JL28I_P=
z9wLf(_@1{HB3kYohMq-HtUvXsgsv6{86m&?r9lqz`ph$oF#7dmda8i-pO_%Ri*xm{
zs9!4f!x7A(O5qSLKZpdQ&6#6)rc(p*)}gm)uYN(wo~r!!#ckO|H4aRVk@9P+aMqo%
z9_``z+hnOFDRmr#^_mt7syajuK}JS4b7%L2@?6Xz<_*bYh7K+IP~=osAO>Azb=N=6
zd!{tkXUrqu>B`%VFUczuMW1~c6R$Q>RdEob1+yT^Bx{ZEPu`KPQ5{wOmtcq~dV|I1
zxDo+|&EZj{p{{>?frBj65`RUEsyI1u_%FOcuhogH#d$NKs){*|m=|R%gMBE4PAP|P
z?6lfM2@mO(Rg<K%O_89~Hb&EXc6mEs+jN9W+;IERw>}^LW#Y~8F-3BtvC_=wCuz)v
zzapQObMAXy!4L5pL2!k0J&=q+YtWl_JRE3JcZ^<FQyTY4hPT;S_wKBtL@m*F^#!rS
z>E-uK;bYJ|^KT>(7p991$&o$#6ZH=wvv4fL40O5F{>v(X=1tvulgKkyyrJs4s4rIG
z$FkFW!WS!SOs$@eC6MtLnEjN-tM=EGx-*7{!OtN6@suoR{kc~#ab<Cv{4$3jFaA>C
zkL&YB?ip-NyF{scBY5K#s@oM|ArtoAJoc-QdKS>0HN2JB7kr7j;@-gL%2K8epH(h8
zRO+7VnwCSX#WM=_U@#>^T$-HfTU=QQ9!0u+>%8lCCRVr?#%`-d7|VtIgth$HE5;aL
z;s2)O;E{mil+*3sV`JYsiHU<Hb|JW+T~_>*^=79;(ydeZ&^e|a+XPZk1j}oibA=HU
zVV9M_>z?=JBlN}`WGo1Iew*%}!<2r~>0#U5<-xo@CEs0|^bmd*9(DeM%!H1PZeyEl
zblu<IZyX2%z4Q(Zb6*(}@5CWYKk9jeOPOHHJWO!tHiKm|boMJ=^a5wQHOHgxdhWN9
zG|z|f6NZ@7A?KbLUv+);Lq;OPqog5{KXZZkyiRV~6)Bq+!b1wYwU_%CH>es#8wQND
z9B9*+c!lBYAZ*clEV6;&o;5YzC%Dq_TWZqeC>8y{F#XXZ!qgP5EliMm*KypDiwIbK
z85HoQbGq|9G!jGZEc%0hv37=!j}`mjZ}UcvuCsfK?7}sW&$}hwB+g;svv1F?!Epfx
zN`8ZDn;^$dwst=A{L`kCtX%)q5BmBSa;61|U0rj2dXK^(q{9U>Wc^?<xB?tAsH;u9
z+-o`SGaY(s;?Kr+sob=5E4L8%IR^uA+xW71avkl`8$UeXp0QlHY}e4nu3n;V;et4b
zcaehI+RV8c_Dl9vD#5N?d{~P1Jba<>rmadAN5@uA-l}*auscSJ>)j4G4L{Yz?$-|&
z^B&KW(otSE=OL?tkh9;Y3@X$UkT|mxEUaPUce`D)#K`7$wg~^Wy#(=(62%nJSt0D3
zp8t&J{50W{;4&N^aXirZ?ER-Zx{0C_zfG712fT^=!|>KwrUd&IXSjdjGG18IXCXgd
zx4+?Y)O{p-6sf=7F;J6Fa-6q3JcbCpNlR}<fkD(LObnyz4zI8-IiG|KsXarz34R54
z@9IJ2K1Qh0jO>)IaP@)ON7lbl_+=Z@(_A!vft+uo(md8NqPXnKNJONnWm(0h#Xs(T
zUyBpyml>J3D8(g0LgM>5i7(|mnS|-xn5?;RFmhiCZ0VSgA>J8?bv?bvWD`W-%6pYS
zEf2VhuDg5ZPqo*au-kM@1@@U=f&ZNV4{^LM$${7Q9bZM4K}m3Ye76vsr_BeFo4TtB
zt$EMxwAd4Ksv0C1LWuxk;rrws?9=AX)N+6%sNz0gxf#?VY=HU8g6qbC)n|l|iOz4g
z=GoQaPpC{+XJRseS4YlQx7&XW%K$(|0a@2si5(kT<6!$LBU)<ID*iJwXe#>+q`qF!
zwP8bvmwPW##;^nX*+IzV4TW+;CjVkdo~agkg3#Fw@AE=d`Ez5Tv|sB3>cyZFbOh6I
zajFzTSa&^l%8@sVZfhF@DgFq%IN6!BZp$Tav14yNnqNwX;3PKT_%z6WNUF9)M+bg*
z)X~fvy?o_u=?PvO2a$IW%d5HlyOj;Me8!C4-wSr0YnF>(kV)^(|DaRc;#p2!_mmGO
z^;5>Lp|hjsStN@UA}pv`Ry!Z%PMTJ7ZYjAJlE)juPiV<1WFK+N$sMV*T+!`f^Z;#m
zHjPGk3<~y<Hc;cGram(A^MHZAJfc=N=FRP|boz3W)*`WW-OT1G$3*+J7nA?+=xqEc
zk%dejDPg1(yMFyWp+;y3aW+Cyv&U-{8ER9>LlmEw0)@P=j8Gmc!e1n^YWmORez4wU
z{89Xk8-62fW>#7&_L4}G;H#8Di9*()vDl#$QM_sHmD<;N{_gzAOS~Xz<x__KAK`DW
zrI6WdENCJeOfB8m4U$#Y?S88wjYauNa4*Uvc*;4|ShX9=_S?jR<KxO8JLyjL;>!ZH
z>7frYRA3>WNP#P(hJ(rIAO=ODbA}7!)Pe!+YCkSInsM3>4>UYH95O&{3$F7+-t<~^
zY|y70xCWN+UP4xY?TtJ?D*AW+{9e8TBR-^&%d{mMtw6$6XJ_f-4lL^^i>Fref~u(3
z!nKA&RAR!YtQr^pYnb`AX}jQt&O!H?TS=zefPb;6b56ok(ur~hii(N`{ZnBM;;nIS
z(3_@dDJ^JdtSRw+Xr;1y`F@J1=2+h~op#8kQPd=Cu=_~P^HdqC9zz@7#w~>YU3`?k
z9$7xI$mLLLH{RXz^z$^2kgngBGzc5}3t|t2mp_mybvgRJ9{OMA^Su(b2D>%Ss8x|T
zPt(J`@9ii#3-3*H*0;l+cDTN7>QjqTFQ`Q0+n&@J6X?tkX`J-TFPKkE@yN-#jj2m=
zS2+OBWkk<r<wY$Hfgs`TIW3w|+SUk_DT()+_|e-?JEQMGZIkq406K!GJ*V?vM=Bny
zh)?D-!rBg3i3g^?<YU4H=-a^F>Xn+bG+uxMjIuP<htRY{L6?>LsnaR|7dIfM5Waeg
zD3VB)38PGO^d<<OQjJj3O{~_9ZwA-hOsMCA@NgpF@<Zy04=p<5elO)R)Sg&rYyT<p
zDo^Os)>E271a_N!<6BstG~N?Edisf@+5|)gG{yizp6Sl-`|hzR`q3l8a4}5ByPJC0
zzOi=%VqqV`Di;bA{8vJc#%QOA<6|&Z=!ke_dC%INlbbO};`E<^=iG$@Bq2@pc|aN&
z#E1f|{%V>4k*a93BUaah-EAAGP*eb&8IuO_OIVSWQ-A`?>4wGe<oQa)xtJ@RV>!T>
zMi`yHv|yBt>L2?MK;A4#W%bf?-b_j8*P9M5Jw&3uj4rnxZ{ubXq#S|Y9rXOo`9=-z
z@=nN?NOS2In<e;+&Vh;@_*ID=#t;=9e$&J$$xsEJ>Pri5z1OH{#<l<V?i_>w!}q?d
z4vl6Ace+G&q^z^M+#>_YxB&vJ%M=c@uMZLlD-jrVB~4G$K#8iyO(Vix9d)<pohFh>
z5{w69mle`SC!qt2g%6G+WPFP-2w@|nR)L&ehrTaYG&(qvJ!IwLf+yrlBEgcTO|b~c
zuR&Nig%_JTUOV7tQ#_hu&qpGB@tJA!%;TA?y=#c^8-n@dugfRDS)4^r$p<s62#au!
z`s%6Tte$Sqxvvc>e;<nbSwf2>i4azKI7us|P!`-kP2%Gl)pCh8j)nb&0s$1wxVy%o
z<g~Xjau^W4&*6;cvKlZ_@V3uzmj9j(+-93#QnpNkAmgN+p_-H(btFqkk3+byc`xS#
z+{Xd1raxi-ZCm!S>H#3ljb+RVuc^`=+@efR)!uo#|6%Vf+u{nEZsEZtNN|?~x8Ux<
z-Q9x)cZUH2L4t+g?(Xg$+=9DXa2;IE&hx(aeg4C_p8mpzVXoO-yQ^xg>eba9KQrvl
zuR=(cjkvr*P!T6qdlkN*un03O){!4|s1NDpwkRBxhnixF+EO~M2y8y;h)j89*WgG;
zCrq;j7UN~%qYBg_-w)z$2FE#zJE;Ix3LrRbczGIx!uxLg_KTnoAdthB>3ToP35w9)
zb?1AhOcQQ)8S(g&jS(QNvT@&l5t1Td&@!N$Cwbmnr06<vmU7&jQ^Uf1gZ?E~-XGBA
z>5OTjZSMlN>6h9S6+nfkFd;W_PA-xoHM?(iFh}VeO(UczzIG;mfk@6qTO0T{M&Wm>
z#ggqtEOUxXz^wK-DWiGVHx*3^z))Y<F6^TZ*p!;a5->lA_x!Pd1GlLh&G5UlV&2br
zZwAT*vRVaZ5Mjgy^~3u*;@{s(A6$`Az(X?>1sgGc5aBA2LO5()VaC1RlHK&AbF9|x
zJeacJP2n?be)cGq3xl5@gEd?}YXwrBzvG(x2DixBsiKxFZ>a^Ig0lxUNFt%Tk3sC1
z^tbWP{SoGKSRvkdKU8}zXf1^#(7!-MW?%7MWk@zUmeI?j5FLgZ4cb`|2JRpLH^NL`
zffO2r4F9K9NIv^qHyPJ(&2_Y`yYZXV@ax0!XQW)%Wev5!M6RW3$ARpj{kO5LzzjSU
zDFASHJOH9PDE~M7dx)EABlT*@&&QPph3P*6OtI!a!@iZ&0PVCa4i-v#kpXZyWU&sZ
z`)1@$ozCMe2ia#lq}mBA)C5jpR}Sh6l4x^$a&&>-RsU=iRj!3F5UPo!1gc&R?D9MD
zzTU{Gm&A<kY@lVio*praSL0xymX(>9CX_)VoK`w3`F#4MATHk72TT*<9c!^Pgf?K|
zPHWu=QtH?MU^Yb>$^D(>uUO%a-GP`Av9<Cup-^O`P*)>nSQK>d_*oy^q-Nrv`Dt>a
z33UUQ1~)%&s?KuCg_A4wjI7IR1OkEU&bgqlLF_X}wQKwHM6C(8igC$Lwy?7jnCRsH
zXaS%Ten8t!z^QJ-bj8Z$C_q(v)3aEG-K_EH(hHVg=2BMIeB`fgcw->S@9H~?%NX^K
zR)nccA#KMZGdTU^x)IizXYd_UTt?yIJiR^k*Z&|uKU5e%w`Q{6Czio|s9&>E0DppB
zdGVH>PLP4i$L$)ZsSjt%|KY8%F{vq)l{lFvStm6INd3>?uL(ZXzAzxJwsdF5`S!L-
z9WGnvNzh%@-z7RkoL}MEYly$arMW%^A&Xa%&^Z~5uJovEo30re8)`cGfK-e%xS{vl
zoJ*aRYmZeqfKa;8Oqs2Ayl)0~F*q(FW^*Sz$^Zkd!)SK1{|Z|p@G{}%_7~+R-@33%
z&bxAVq~_hmq`e<(FnOwha5~d8!GN>jgPMWbritUU-|&~3zl37q;(~R@`L{KLqxtDl
zUn1Z3#qhLGyBYWcVq)CE0su=EP^vxaYzUL84yrIEv13}zXgoY!U)=G%?#jd3>#oa!
z@nD*NdoD%x^(NNt>bkxYoJP?XfNyrzJufa%!2a5X+z?}Dn-I`5EB`hg`nj)0bu(&B
z6?#NeC7)v-QkU<lM{8x9psF#!_12XS+$saZ_A6+^Me6*(&ker{{j-D<&c#AZYu4D1
zAWVB3S+falwWk1<Eie&WkmuDfjUQV`*brVRQQ!}6tFN52E$19@8XDR4ni~u8^K78^
zi1Y!LLUH?!mGq;;3~-sA=$>W?ueL{AL_gJFvE$`t4!|`cx9^7owiRN8STP<4q>hO<
z^wc>U?~eW`BU|0cY}Q^s5~i+b(3x!8Qq;|TG^Ybz9+2E9*=Qx9louFvW+Z{j&+(VL
zR@Qyzl^~PqwBo+bDEAxlXGu0139m|0a)IZJ?<&^e&dH#y)7D~~p8X6r;GR7jHi~u=
z*CC=PFpFRffK_|L{8`<9o~+>7>_4qt*XH_A?LzSpnbXfqE;8)#$0~bwtDuO=-uJ;V
zN#k2-<MA7z>Zyvq8z-EO9z5h|>51e1-7kjLGMpER(Pl?h2|ki+)om7q6~g4)k49a|
zQrQsvgzb&Xj{et=eg!{o(`WHOo*rGcuWb>`@$=G1xn%l1BDfDqJ-oxILWb^JtU*x8
zeAvzI7GLKt_ALe7T-#AR!D&yd6}ZFkul<}kb)@SqvQLY>$Vl>c@3y@y)JqT4oJh<e
zQ>saU+0dTgvg2`o1UO8A)Knytuy?=5!oohQ#Xy5_aO_(ZF)-j3A74}t8W|!`?>{OX
zeHQXeNk>G~Lf$?1Xr12q!#r5GWpC{3D!#%4r;2GljxziiVrCS(`lsxZyK+ndbWGQ~
z=UrI~6`O=iO4O2=bcn!wHn7JD@==n(^|gwM*Tauzj2iZFF$!giiUMxCtD>N2slo9)
z4|r}0?SmU)h&#T^|M0jk<|k$qrO*Cu!hq+6!wS)H>Z=`1N#_vE>(h6$F!%iBy}L(g
zB*%@b%_P!8YvtPbjcNLzEQrZzR#rD~=-&Bs&GS8wi@Pj$Q%a%t_>^kn7vQ-;K(@ZP
z70RAF@;dnm-by2Xk1T6r7IhqsOxN?U<YYywAU>-Fp#}X+#{VJ#8(uuQtTo&Da88|2
z-$>wj<hNPFxwVGqFd=;k6@U~Jcu)H_5xUD%;+7}*-Hbg!9Y*Y_jcjER`T2l-A?*!K
zBT5x_bG=}Bf79BK$?g{p(L|RpTGKcJsgE8P_FBuAT3gcn1UK-rD^<0i5!c5qr|{QW
zsDOC4TJ6B(p-otfNIoNVx$cn<cahI8^~*j#pg>3RaMeNxo(x6ToYnR4Jo?bK0-)6n
z*dbZB7vk?Y6Wnhkljxvv+)ZY(5Y2fG@+?Re@eCK&2SWU$t&Ioei(PD)L%E9EOynpK
z2wO*O(ajh4??d`!C{Qp=Z0TmtY9q7cf;Q@q#)M9<DXZ-aV}mIqf6I8T7^ffB?lc^A
zHSq9q{(Rv1G1qYh!G0^87D!)t8Ah)$L_pzLDH@ug_C>I4Q%69tR$st;CpD_ujU9%)
zsB^+TiHKZKz)k?>%H`JeX34}TmP1-@N@pau5NXj-q;}<a`S=Dc3}HVS`cuYYi)3R*
z$HYYr`;{9jM(qQajriAvX?JQ)(UA!r8%Bg7)Te}*{9pyAgyREVZAMhIF8`FqsCQ+k
zi_Y@kMyk0hZP(!Vk2<_m)u_awo`@oDL>o#^+Ljw?9wTniMHfwA;fsBt1%Ja8lxHy3
z#f9ub&uoKFy2Gmnj;ziMTHZ`RO5fh{$uMf&E_)X<U0K4hi%i3F<P26|3dM~<k$1#+
z%j19+=ZIw#-OxD_+|@NtNlkG;t~AA_u^x-AJ-6ZNnXg?EHq2_nqL)lN)}AvHul`42
z7`W~*Y~x{1<6Fc9sG#-2ad=->eBh`*tT_}p7iMxIolQ#MrJz4*#jz0r)3G&R-+}*0
z&6cG%c4()(IO|XWh*%eEdjQgCTQ>us`-bjp9nn@DXLqOCRR=eLS)Jnm%W(9O4X|k;
zBnY2Q8zJhV$MNgSN%AkN?rEf2FB=Y!lZuEK0|&_z!Y=$Y7L!y~TQqP@=@0|V>vN7R
z*v++*)RP-wF&2Hb>oZruWaEOhhpjHZ?P!uhBT)J)M4#c_FeaKtzwakBwuExb)@EbO
z8*{D$Pkl8$M!Ex*x|jLYtG2-UOLMEm=J4I#%dCn^nlD)sjx<3Gpe>~`g>e6vx4C*x
zzRXDQ$DU3V*=io~bh3gwf_;^pQYNcbm#%M1CQZXyT)4K?XWvSU!BJi;g*jA^h93ch
zJTdH+QwkNv0Ruq4nULXg`RrA%SpBTE1O`2JG3Z9W@8f9gjqmO+r0-^&j?tx&l2#5z
z`DlrY|E}K2zLP-8$b9c2WrYrsCx!kk@#rC>YrsL6%z+Q&OfCqKp7|P=Q+4}bdPmp#
z8hZOufqDsYOCol9oXr?F-gn9B${d_UXLU@zQX(oJe|7rmujNmwr4_iqWh34DVr!$6
zoQ=bGHkLW7LGlF4b-dO(G#<YVwZ&%_z7J0WWISpOZ8u?bKXnZ)pL*od&9XqMy@ZVL
zOcgT#D6ra&z%^!o=!teO{~``|>TpW<Q#Xs1;dq*)*@E36%bzxn4!f_yJDSc~NZmCQ
zvuuTSJ;mMW%F?2EbdY>qv73Er%&aZY6MwPj{v)03G|`d16-T=DF%`uKnUCFmAsP~b
z0RHF}lnbQca!1GJ;i#9TeldukGHu{yB?x%Xxl)oUE$w)QC)as7g9f>4bV?a;zwvj`
zIyUNpC%Fz}Wr||K<;=uHbg0U2O{kwn%jkaiGJi_TKTq8*Y1dF1RP%fiP}18MA~y#s
zYC)$Mtc&5erS8;#0*<JMt?jH}USCpdBi-uvXfKWw>B#mE`Lt#z({7{VRP%tXtSK@p
zcv5XBUoiCj{m}P-{9$sN!UVO|BoSn8`h3RqXy^sg#=X(Imp~j<aA8y3QvF=s`B!88
zzt3yT1QdQe0tuCo954%$$YbgCvXQYHtW2FlapV3T3g>q?!zJJNaqq1|)`taQo=xSF
zA|K!%2<&*F%QNB2Gnf4`jI=+<P&6of%4~jNUbldIy=A@F(GYyHB@_PJODG-@9@Os7
z1bARK;DO)MVT#P5d^>ZlcUUpHHD&^+u=tqyAfzT-AtDwE<l<?8Wkthx<s}-m+uOO`
zi4gOK(U7XL)->EQ+W?M6(8AcotDAQOx4Xyc^Lb^Zj^Y=AT3Ie>58R=sA4NV132_Z@
z%G++4jelm&P1Z>52Fn<Z0)_oIy#dtr=1a|eiN*fe^14S)Bg)?1IfJ$&ajaPyZ>fj>
z-3{kWxg9*X3F&^=`i<+Z-7LL5W@EzX^`I14UqBMt-#bz+$yT7kt?Wq^p505+QE}tG
z*JTX`v_6aJU=IWCScjDKg=q33&3VVhBOTsCa5wizFh$N=tvEly+>iLa^mqk{f)SWW
zDTsQQ?5=dkPIM1@fMW2|xPK^xWnfGUv>eU7zv8O}$B0!mt`bcHUoz?EKu=(MXTo`N
zc)U_+wqG1{$B1N$Bw%FSAiGKG*!8N0P50LHRsZf-BYQVVZs~K{8YH>tZWv;|xBD(8
zVGg>>U#Ty=$6b@?VnlH>XS5^B<_kx)V?9UU9pwjIMZ5PH5OY6AD{1kZlbM}gZmVwB
zWIyHOlia0my$0cKj5Ew_duv|H_>n2&xIeTWNOf}WwK}x!-M5qj=n7-&#I?XzSimVh
z8$TXpRbUV@p0y*zZb|7lg>UDApH^tMnzS=xFg8UkTou%>Dx1CZQ31N4)b;A==*F$w
zIqt1Piz-M<>DI1zs&46OZ=rd3dbQGIrnF2mg2UsM25$Ozb1T8wdkCYOg(wgyC?9rK
z3-@nVV{+uQv`Pfk&eOVE+Um^H(}At8b#5o@w@6B)M{7FX`z||7%9G*})cd)%Uq~lo
zQ+3^@=dR3og;M!4wZC9$9vrmxuD6YAV3{EtklQ45G#j#~*RAyzg55tBb9k%VwA~46
zFWh7fl!G8)L)29sy-G-QOx&)v@e=IjC1*!lIOJe(MNY(f6A}6BkEibJ!U_GwswNPs
z>30vV=Q}6NC5=`q{jN8*v`f;<kHU#8e1s&7xkixGwUF?|#r@!vK<{HM=AxW!`M6Y5
zxgT0pG-Mi{RIviu_Y?D+^Zh2Iy=+C&iL*SKSl6a}Eq;0TS3;k+4ZELPUw<gUJF>w|
zTXoewy{d~rRTW&E)VJN{QFByW--sBLBj4X6S#$6nqu%^6d=)cqRPo(3U$iPQt&&cx
zpvF!L0526Cn4ZUrL!Y=*JG)+0S6~I6T5iq}3kp-c|5gavD6cMj63WX-3Sb``rM#Zl
zNQniAlaJco^r}U|<AC4`B9KlSnvE7e6|ASHaLaOi592$+J8m(PmKv^c?kFpca}$^S
zW}7&PU&v%i{F9XW82u{)uSw0YuDL>JoFl+(Ch8r|+8!SVZk?@S46s@Wp51ZoH{V|v
z+8?mFVo*QGw|eV3rgrL`J}QQ!S0%aG36sskIed+UO+<|IU4MMCTwH~Jw#ZBWJ#$nm
zle6^JZ5w|TJ4!Qh5U79}rF$=!U<70<y^V$RL(%uuG*_byaO4`GQUgUc13`oJ>rlCN
zpYggmOwg5PJi&brPPg1j!LHSU-0`1RRn=h~F^iA6oz$?xui>=u#Xnp3h|oX<oOq;4
zKVr+raIhdrn@;9^Pzz}yMiWJT=67eW@1Q^;To`2jTG8>X>|6#$d9tn9`MUP@4bv;x
z^KnPiOXcf2o+fp=D*5q4sRZ2@ft`MWO7&b7$2?$1HWt%n+W|7B1XRW}11@B0w0W|z
zun6c*AZVmJHjyP@QxH{<%yGhWt<jGAv`wo{6cu4hH|~c8cAvzOhK8mu-iCPUDzz?Z
z+(iX6%(Hv!n|+4`*}6Y;KGwDWf<EZbZg(5jM%wK3P_`~z;2K5s(=7EJFhf5idJ&h-
zk`W7dDQbw~q*c9p*~l``Hcdw9o|X5C*~`529Uqm3p2lieNQI}&+t||2xD^r8VXA&I
zJK$6-u3if4@}(>wYy&VfI{K*OrdX1kSG0(=_1xVxl$pW|E6BKE7@~F_UYKV|MbT<X
z(Dap^g;`Ef`Jv3|^jZ%B*E^6u?E5*R=RK)8a_Y+NUH?pnjF$JtU}vs}fu8o(+Jf6`
zTtkdlJ;D9*XF4zMVsU666$@kaA2%{rZoPyUDh~ZX-e@t?3Ir%E(Lp6py=?H=g?|pj
z=^VZt5!~^L1UQXZI!ajF+vExQY8&kR6E0oEmb@mt;~nNF*ba><Qy{dV;<T-Cr^QMn
zO#5Wp%;;$(=yndSC!mm7awfe<I3WFqK9`GANrkpVKYI|cKtbs2K)FB1n@EGUEP&o<
z&UYDTMcyRK;@Nmc#~u{iNP7y<q3}7NK0l(H-VrTpatVkS<eNlRP@3JCT{5xEft47w
z-g`<;gdHP=Voamu9?1DbA!EM|^{{$qRl<hse_cyyC1UeXQ;F(R;l5-xDM$ge>v7li
z`Z6m0OH~UY0y18+W4Zq!W2vUa@-sWA+CAoa+<&u|v!h;;Fbw>5)or+7N@3oyMHic&
z_xmf|T>T~aPZwECRFofPsLVsqxX}s5vHWSy%(EWn#Y@lttDGN3H=KWgHB`!|h_&g`
z#6lS5TbkJj^|c&TITu<8nmg_`Ncmi}!<6xMifZ&~@H8mHjQ#PWb2Lz)bhhdPAEdpJ
z;H1Y5WZr!`eQri>*;@D4jn<J$rJC`WkUl72U$%X$i!LK#h6RC!5q*6Xo}%<)pIdrB
zD96lES-6pCYZQ=<z)Zrr?(b=rnbx*&z+Ur<LOe+=RA%DetXi=Nl(1u8`hhsGQ|6$j
z&4}_Q_k!Ho>o=EQZi&PUP$8^`YuqN$@vak!sD&<WB08u)&juwMU)h|h^(U=4=Bq0l
zVk27^Y0jroDDu-|xcThmOMl+&={@Do-@h<;FWe=jz2!>;tE7Gts+sab;(%GTOm5T;
zaPtWy(?>g@kIxp7=%7SKQB2&Qi5MI`K72tr;pwyy1_4exh?c)>amslM5b)5+Ax<{s
zNch7D7x5iSJxNT*R$0YAZmZq(W!)v$SCfsto1NyWt&=G)MHv9~QC}jrNtUX*wUZ&z
z3dGl9`}Mi*Mu}&1>Ve%knIJ!I1je;ejb!J^r=^pGOGI|$HW*oFavyuRVZ|1f4FTjL
z<t@w%0H%I(5C-qupN~s!x+e_3OedoiDoQ0=a*lK(+$Lq{wwzDpjfaCL(tcS4t+uz=
z2+V23$BoT{Yq~33_Q=XeSm?*@?-E3ZRh=bm2*|ZiEg{}-)gh>w8s9)w=kVbAY3{1y
z;oT3*iWRg!A75YXjq$90c^_VL+%zw0?QXOHM|5*M>{LDS@9w?LgRVi{y;HxKN@<FB
zhacd1b1tsSzAZ2W4H<=?{v618Tbb|JO&<o`#2}2=2ns8?q?sLo%>&DoY~6LZT~Q^Y
z6x*sZ|ET8X?QjZ99>@?|9;X@eYnMAjH8#;!0|E7Dg^ahXLaFa0w%bV9b)&tx`m=2F
z^TpK}5-Pob!z+LMoGgX57S(3Eli|)GAQn0PizU*Tb7u)F{dq8%V`EYRz?e4~xhVqb
zuNevSG(Bl6^<9b3o2@CWMT;^9b?FwTsefwkurDTrj3FF;ehpz^y(BIs;X@@b{bvq8
zT9;i^Ukw&eo~=8MC2etuE~e2+?p9-Sr86Weaw#a&hDqWA)80DGl=FGhG`$B}>+}XO
z?y&Zzjk(_AM$RW1<O5fVkmwd@Pt3vjz-K(v7W!r%ju*RfuA9Mj*2(+BXVN5~E(f)U
zcPM!OnoaZfjk*zQv5HPdxgSSe9)%Hwx=nAbHnDY8ff4$k3tg_u^`;<cJk|W}O<46{
zxL$>CUMx^G=h{W=+|wwTMOw%~14q&0g&2LD0Ew7+u5n_V*#k^?E1zI{V$<-iqN)4$
z<s=*>hf~@Jcztn+K%cjW7j`@*Ow#}f(}pMoC1BSp>5YphFd+`oF&|aulZZbv3;*mL
zbUahe#b0e7^9<BRzx{LkQuRRvRL3pqzX=&<4y`12<2HA$ldf~Ee_;c@n`%$c;C~uV
zWR%t-<_2!Z<)3t(AFTvxLfjZ5(DjG!O@2jDE4?t+BFI(TE9W+$h+2PpBOE0}?fwaV
z^*R&57wY!dQ3E)eAsl1c<_8#UOJf%wKx~AJ3HRCiSNx+M7gVwk%2%tIDNnToRTtgI
zRRE5Gl>}S-3=j5ebxNFnOi)M&>Frxu;S(8CXJiq14?c}R>H>%?_&0nx2?G+;0yx<l
zl8~X(5kNP#uRiSSaTqX=D6&cy0Z97#`GZFi6syqRl6kC!@vC-lOw2DZCdtBghFVC*
zv*1D>7Txh}>sy91qZb$PU!AxwfG75zvVU}zbs({jh&_4BfLk6YqraG%nur-E{2*l{
zD5C_esWitee1#e(YW<^aIh`eKci`!Wpwmkb?r|d`u>6CNE$Cu~AtQ-5j0H+dF|Pbf
z`JE=}X_dn>q|VjMn(i?bWrFxDje5JsxhA^X{MPtzd{a?B-;DL`L36WhG{b>h{Ag9X
z%x-?;cUm#`^?qs@)er#@l0f^kn^jt~^rl`!D|bEsU^DQn7O}OchJOCd*HW!Md|4fu
z<z3gp1~E`WE3J}mT<>8~vA90@ZZ>g)7UU5oG>?5IwoA`q4h?MZ{o1)w1jJMEee$Jm
zyD0S1prw?=mXU%2;>kJRPHB5-70Q^?b_I0IMwE5#+?iIsD&H713Gt)VAKCJq{IQO$
zS~%*XSnoi?P-#Bw4?PPX%DS=GMLR>-&1(DUlp|r7l-ltF+Si8lVioUK&C%FU{B5}n
zsX1DzG!~#3aV9i2ik;QwA_&ZR2_Wb*;2ymW$(5epJ(NA&fiaL1;)L=v{yw`P=-Ud1
zI<86CzsPFt$?-+(CW_AkyBgfp_y`GS?GEng?h=~2-=#O!pb7$RqI$|fn)+oopB6Rw
z^Sr3cCQ-~a7;q7D4r$YC8L@mbHIwhFH@6XSW7}EfGu%o<N+4yjX=X*msrqKC_p2(Z
zk+(twNK?+^%se!Em<&X_E>331EOyDqkl~BKD>o$vr}f!&v-t5hIagR57^`0LQ-w(!
ziOPr`Y`i7G_dD4|Nr7dTK^@vOPpZL4&2FSXHD3TlJ&?%a!QD=BIZ`1hIHiJSH70@!
zhWl8X>^^OY7V@B_7^|wosZbk2kg(l(V{VvoyHLGI$-a>WHu;Xjr|5Z@j)$SbV+slg
zO9%Gzx%;9j&IaAei%Rdp3hEA`S)h%&wCw|$_T-{8(nzEgU@cH!3e9UP{fqJA%WLvQ
z(Z@*&1VO-t(CQ6BUx25vcOecF(=Tfh6XdnF(^=Tzvt>sSaOm18>INGE3*9pJH1Ve_
zSl`Tp)~$<fPawI8I3rrL9=Q!dKM3(LU_lKPE;S{|`WZRoQ@VcQ|I#pw+A6Vw>X=U;
z9eH)u5bjM8C#r~JJ&u{duIisR6ifH$jR#e`1UWw5UG{7i=bHcWF`%4d829X8D_EH&
zC90drWraIk(AQBv@$?1q>AfmvywlOp(gqc$=r=yyh0a&RFnKa+B!Wphx6TjN%4QVY
z>Or!8JzD#tT?TMan0bLdN$^5m5)ohP8snDGGS3Kn_|{jdY%Z?cfLQ-;6KhvJUdNvB
z^$Biu{ci~fHwFft21Y}d=&ZrTL!E<Rj+=A0<W7$R?fxS=?WLoc@HFMQ&h-A#VQtA<
z?9Nx$qz#uKsAjjy^dk^ub?yPRwK25gq$+vS|CAh^7`c`>QWB9E0kqQPXv`(M)A^v&
zZRyorf}A~0FPc-5Qv0yoSlFdVowpi)KEJ5$>|OZg&GY_yDMJwyuSDjk5hMbx^3PIt
zEluv2>Mx6>;O>sXDqy+=u7->2d$2m8lQI1}0XDws@PCY^e^CbZ%(V8JJhB|bY;X|W
zLnOP;w!}p>OJVb$O~8q;i9&oI{ZNlm_pZI=`Krx)zao5jgpO1B#&KOoN$yqK_h@{j
ztg{pH*F<FvFdH&S!%JMdJZMVqBNLtLwNqFaG*A==P~R5*u%eGm7N8Ich2d@_(;JtK
z`(8P3nzDlc@r}lQ{TQkG?j<HIu#R(L5GdNTC4mS&B#Yvd0ua<?0w2xSQF;Wz(KsCS
zPY<+RVk)^DEZRNs*={m}7$%=)B=cupq=FI4Co!q2VA57=!s@sB12s@A^{sqPCmpY(
z5I<0E^BN`*8PW0eXt})-ri2==IqWXf=j~TsgZ(WD%9}CkTT1&y8dxZT37_e^alQEJ
zbam%!p-z7>E*IeMfApv;;F}#^d1L9H`|31Z7W#Q%@n-$9Z3+tu9ttSnvnWG-857wD
zof}%-gLLk=^>RvX8PB=H!(jidl=-8Q+MI?<a>smedq-@EBau%T<Y}SNKB-^p%)lM~
zL)X`m$pHnQ?2E%%kH`>!eHlP;;T+jou1N}HlCky#wa;y?pJ|EWsNX!Dmne<3avFum
zk`%s_eDar5g6?dQJ_>I`Zk5CjY{T{VL?%9)ghGZW;2u?O;*JA77j<<X0JPJgMRFau
z*{0s@`ISgpFzH<<BW5=9++a{SNGUE?!skKLY=M>pKYV;Q8Xi~3<>l*GOF6C2rYUx9
znacu5L(;$1mpOK6s~4y=4Ufi!11b;!BF`0tZ-gL`*!>j-YVSKfK-?Dsyyq)g2Aul|
zHW7`z0qW(G*7V<;GgNVl1`6tMKrZp}zlnP+l#YwH0(1p`EaKMRt603ar493oJ<0VM
zdZbYW#24y}Z5oQmj{?2A2t7@?9gD>)rvl7UEzqGVGcb~cNW^m(g=AB+*IDEUtr&h#
zet2(?%9Leyd@F6Lf>5U-d#lebrk9mE3rlyM3;oq9uIivr7{Uq2;_O@sXuk9Bumz0z
zXuF^O7@p3)*DgsDrFT5`r51gU!Rz=SC!R|8<G}O^gL?BR3Fs+>(3tTz6ErQXcr<8%
ze!Id^f*!^Ky*cjI4g6x7Rn^(aI(qllpK78D6BiC8Fh{o1!ke!+)7Vwznw2vpNQ_pW
zcje8JXP2or2(q9ejhLvI8IH`ag(!yG$YY_&Q)y(2(c3&3-(%6;1scdz5n%ITxb)fT
zu-^s`gM2-{mrDurx_^M`{lNP8IYsW{M@$r;)-x%QLWN~LhrKDn>ey*$-ztRyOO5tL
z6I;cBOIU8?6QH?O)EbtFoQ1WrD56sP^-RHqiJyXbq61Xs6lGWiQe=1E0lpU(3Tm+o
z$h}q-ayRyw*?D#9^vOV^0Hi8({9YFvj+KU_qr~Ll76OM7;>^Clk_M53WW+_)X9<`i
z#CN#?MP)5gPbM73gKoUe2*ns7#!1S13_#~Z8_ptAJ9K06>R$on#3(2uo-cupZUlVt
zpP@qg1xPgi9keK{fH3O613LsCME>!=16d51@PEDu=sZ3Eik|;Fdf)#O5&U<UgaHfh
zzay9BH$d<IJ5=!h_h<iOssA<df0xDoE~)=43gGYmlMXcG&uAC0M*1)O{TnI{Ii(_c
zY0qh4VcuidCrIyAJXOl*O;)a;PNbZ3xMpE~KvS8rM6w@r^VpCYbHMxj`2P%qEvt@m
zn-^;Qw5_(NgahItdQX7V_t%NVpdEJP=a}*2&#>Z!e-Xxa0DMo*$Slm9@+SKfREpe?
zk3awG%c9F|_f3L>NO%<LkwEzWL@-+MMceAmOHA7*cMouT1J3R0%U6C^P@$YwkLU38
z=2%Qvane6X?)#bP)3?2TODZ)OP<r}&JzhpI#Q!biA?*(8$F!da;!w$CUz|jhu^dDw
z-JwiWR-7&ICM)nMbUirSiNg1m&svS*q^+?pGYg?1hg8;<5rsmshs8%IzS%RYBf5%n
zG4--(yTP;L*_7CU{u>6o{13A4dNa)ZBm`RB<rk*kfoClqx9?%MT-fQK9_N`=2Jx#$
zMvec5l*l<QY}&jxjr?3tfUA@V#yL-aBK|&36~-m`#rs)%mCu+rad0f)+aG=AJNR3*
z@73oka1HgarQ>Mk#>kONYR1VdbY|ybk8H;wO|3v>E5Ee5^A7EO7HH>2Col~kLnV1r
z5f$Sh7_#Pu$Mv>Woy8dN(rt)eIEL6=Uayw3UDIoMS@+G~&AYw00A(`1e9(W={<lNj
zS<fAt!`Y?HQp|4e)$%(L7njflgp&uaZW^-Q_WgmMPp_cvoRH7u%Wve0oJsE9&vi#H
z_vPNguU`9h?SgwvLt(`4^9eA4l>)bslIWt{@%98*`1t0&a+zXL!v|I=%L_}G=<@9%
zW5-`4q2O#(*rA%!RBaE<wsbYH0Ezvg$^`ub6jER&_}s!O$o!O{`rzd7rdn4O<>|G9
z#SAU<aGpPhk`mtP+iZfaA@WM&B)Qbk+U&Cy_99n1lIg7(`A0V889=P`4#<2eF(nBH
zCd_;#UhQ@pXFpdqUuO;%E`V8!213sb_o``}da|5PN^?<dPKYvefJuDTS+z6YnvO#>
z6S=ev!K4;J7$Guhzz0SQ4rRSpes?+b4bywcf^Jey5)&|sLdP3nY~rwv?%49TxpCEG
z6W?_Wkos*0N+NmMQ&hj_;UZy2#Sn$qsNtGwyL$h$lm)<DWpxXzL%c)J)PS0Dw(*x$
zME$f@e6c)L7_Lxiy?38RTgXBQkqx4U9r%J)&;^fN#}6-n4O8EpSu9WP2PA*_8Gbvv
z>b<j2?^2Mp3wwMpEWTt+kjdEnpP^DM4Q)ejv|EanPoawuioI8pUF6&A=TIstQ0c@i
zUq+YmDwe8BKik;`zuc?f*B0ITe*r>PFp%?)SQ~E2jRv3ey!3Fu@oyZU=5`D<C=rzl
zk#)6ZY6vD}%Jxmy&pGI0BIXvO&BnJ_$oRs0R&V{G)#T^$O(T0ii*vh>)xGg!VLNav
z4Vr}knYy6{YF^7I;N>0z6riDU8K**Zuck_8E0Hny`uF#SlmC9y{a$gKvFajXI-s2!
z&DFhBUN^cp*j_MEJI#1K38i@rK?q%o9FzoON%n8vwFtO9ay9|ti!-h_5~1RiD31yM
zz+<+=`!CjlUxeLZ=eZ5P=^8B@I<3r4W}BgWd)VjCMnTcf?k9K(^!xXGvp<}*g_mJx
zaJ1lzHFtr-X<|NyI{(0s>(v`Fz(ljwTn@7)c1|+~VgDTOk>xGTZ+m5z=FTq%G)?#&
z%^feskcQ~$<yn<4&m!vCM{8B>*i^W`Gh|#V_j(3=`^kxGS5Zfgs&Q@?=(z91v7+~K
z_nC>Q`gupm$o{nqIC1kk?%vdb?nDt2WLLcvuEy*==AQpY#ja#f&is6b2JY?m<h$3X
zhJ(iJp81o9cj4ZwIP{7fI6K!zFFdQti*^(#zGZ-){_`I6gEx{cfd^ut%4L|cQWpiJ
z?#gASnVzie;h8GFmlPchBCy+)m+$G&4%7pR_x?$nkYDFpqTR*ewO1eb<~`%vi&PUJ
zh=bT7Ex+DM6(=0P2-Hp3@^}6ba$j&)*ps6J-K-N`9Y~7j1GN{xh3H_^ZlcIo{wT3u
zPOP_@whnH@I(<J|m{Oxg0?p>eb&s{Q8VxQ04|}|P&%df1rr8D;3tvZODM8|=-hdFZ
zD;$;45ZdL-xcojmZ8~WSkD@EwmF&*E+TI(_+G{E4?M>F2%1|x=>*GYDraDdlbi4j7
zlqPCBB6k7P^`Y;O$4Mzk7lSua46aa6;is;<;eA+vSU2WQuCE$DEu?nI+~=bG<J-S)
zq(77W(EX&p0xwCqwf<pYE61d%wmt*>Uw~fQ@&C<Eb=+n*6cbMJ!$0Z?<CE-s=Gv3~
zp=7yWQ~zGps12>>*@<L1VKc}^71xX3E#{~~VaS1RXXfB!S&Dp9H?_bcp$L;nL7}?y
z_2bw^J2TH-0FcjSit1w=tU|miiDkV*HVVa;X<-ncy`&&fG#@`2ctQt+dCSX;{%I&7
zvzFr0JFx?kycKAo7NPJd>%AoRRES03rR-L$gJYzdHt;%v3w*lPtggvGsi$_Hg=T_<
zXU#AOL_o~>R#8!Xx;%fFBW#h_ioYuQ)Y)_X!|-NJf2c^k_<nwBV;h0yb=eO1W+{`~
zy}uMVT<YB$R`3$WL#RZ<c56_~=2@PS>+&-~z>h}4fBxERHn~>Zzy)g+vQ;oOAKCA7
z*5fa`{=L+7Ls@9ulQ80LJp%Qh$r?NXJXN_#eEa;lVn_aiSH~8Fa4Jf3Zu|ORwy)6x
z@6T+$Rr!k<)NEZRYQR^EhQAXiT5gr&!@?*_YAQChA80y}5Oz}ke07DmsDt7NLTa#}
z`Go&}wFx9Wx7+n0VL9ZM3~}e=0i5|$h$0ta{+cfdG^8wDTLko-m5j_(tU3$FwBa(X
z15cwDe^F@vWDaw7rw4ngMUq*HqY+Iy^aXkOyI;=4jz?cL%9c+&bWV;mEh=R_U%|K0
zmf|$Y(|7lHo=$dDo$+u_f1b55C)+Zja%dyT1KoS|bX|HuUc2Rg3{8T6y(yLR--WNR
z^4p;AO!UPdSbR1)+j=glX{!{RS&;bLqS3Imznb-BtwpK2^9w^XSEjR0Yf)|md?@Ga
zipQdk559u#h>&+l7cUf2p>~$g!A0?=Me)iz2o=U(l5_TbL~_ZEM|So3Wxu7K5a063
z2Y$9s!#%4@p{zC&CizC`K`V9doY-2A&r6YasN}ML_am5B0teLz{?$CNLXhH&TW9@r
z5UE&^_W3)ZlEun6#^&hojtSp9w_LOv?}YSYuLaHQI?&>pHbU+u9PaMjLTYy0I-*U+
zcno%43c7eP@TipL1zt<zzqdle?~>rqlk#0I+`hS20TIQY3Eaocyt;@8T8v!`SaLx#
zwUbRaE8=k|fs6wy4D5E~6=O<#&dFpN{&{U_{`y{QYBW`-Za(7)cEtuA2&P+QBdVcY
zcvUkuFO_w`1%rWPCUdU{tl$*|{5FVWkvW*2ajtbg`?Ev=nQ#!w4=61({l7B@`hv!R
zN?A%xl76B-8V8<6Y>`u)^P2F5pJmPt2A6))IK=x$rtqKt4!PDdIj&B%V=wk(YQU_s
zB<sPGffxzdCkvn!QYFEgmI8Jn{bwik0f|xL*X1?k5Ht7JpfE&y%gLaJ&T}yt8Mns<
z+_@hIavObf2;cyj6^H(@43BLhDR7h>XA3y%E;87unZExitsw~Yp;_Pn>bZ01rN-I(
zLK$DuW@VNeg#^B@nALD{cmd3Mm8Q(GT04lJcX4r11#710w{|%&(^^ZqLF-{arjyj{
znI2JXJid!n%<8oeY!|>g#qIJTiw=Mefsa_r@C?%)-%wTQ;!Bh-=UXNn)y|Yy%?D4D
zTX)=7-T$hzdw`#kmUhq8ebFMU=T`CEE2=8<{AP>hcw0L+NcK201{x4Z2b7Rsmkuuj
zEi#H2a<aR!goD5W+i2q6Cj>c6Dj)k-(LK!d*REwC!4tN}#uk=oxJ)o8nT<TVJ`TDg
zQM<-`i$7!Z&|Zk`+!v@SkL0=Z`S)N)RCkP9Z>f_V9)E;n0VgVX`>4Ccchy3sOfgu!
z8D}5)iZv3H$1s#tRY7sb<!|5iG4&hXJ`f+(rnD5D(Y|V2adoEFVEU8Y{WMA&5n_KG
z<eBqDzvW*8scq5?>Ox%@uV=CH3cZti8h7QLaD4%C;%SU>8NESgBxV+9X50g|F)+2X
zH`AZ^wy#CGjG>@&e=~<gJ_Cx1jsRG;Jy`=n0S;tO9iv5IEXkL$x2E`&b@`ZHG*yT7
zMuTl%@cB@>h`i7eFf^+=!QQ>vl&RakY<O20)0#93lv?W93JuxQsuA&serXQkfsU0G
zgbQT8uvwZ1uHL@>FG%gmtw>3?SwTs)Z|F3sKimYX{@HwPY5oVPUk5Z?7p*UD+>Yed
z@@zjSU*$Tu86y>D+Xa(`d;otg&&!jI&w|!)?3LG^?_Nz?{U9p=Ie_ZLVjGBGpR#Cr
z@Ppn^{VD%urG%*DnM}?Z-j1_ayG|PZu@8N@5M7KR`>v}8BQDXyVjEm8W4fB8rK&n(
z()#GFz0kRB5s$o+4#y&&IbY|c8uQHaM)nWFSj%$>IC=R$o1d(L#(sp+`u;2Yv;vDs
zWe&JP{1zjc6Q>6RhqWWQop^(joW_F@VEcvP5{WeHSYMA5_Y~+G!?gElP{7AsF!i*;
zlkQ2SD<M&aYjEc}MQO(-)c?ENL1TBR;gBsC0ZR4qnk659FU-H>T~Mcw0E)kdO+J4i
zEqCaG0sz%o=5H4JkMdd@ASzZzf@y05rgCQCg$j$Zmp@ROK~XGBUC;v&Nt-|x{vto7
zEa1A$yar7b@u)C(UfUUEg_PrB{b4l)JFoK60%#KTfC2O(Dqg?l!Rt)&s&m*E6-vl=
z36crKyiuV0Xqjru)a<(95v!JU;5=41LZaR9COW1tHV;|Ntjv2lR@!`+jv6W%6-%@3
zDgN8@53H6GR)gTb>y!;!OdfysAf`beo$x2LQN{CXq4s6Dw(x~qFbnuLnCbXBXd4R!
z&F3@r&g;DWh8fK2#p6Wads(UvX7&7dY{+=V4a1z!5kq$H21AoJ>7HEjZS-lpzW}N3
z!gr7Q+33NAg-CuVnkf7q*((?KqoVQ|;jzQDC9?;B=z&wY$qf&!)jG|$;yv@~_Is4s
zs|nPWQv{cQJMM1ZIS8=8x4@ls`=a4W0g#M_H7xtwJQH1a-pFfQc&@1szLF*`EBY*Y
z>~_Ofj8}nMs6>3Kk4I1dUX@Owmm{&~zoF&I+?YSSR4Q}1SlH%vbXj(Oc6mm}u!)?%
z@f99<OIsQD2LVt@^%lxr!6=Xc-t#NDZCwwz7Gy2KYs?}dGa6tbT{yUt+b|FByZ6V`
zvu(T1-hbZ;kYgG&NGVV228zbQ7(}al*)i^XKr2qF2<g~_{iBTjd!U!)3wHc#W!D6A
zYMQavqmnX%f!W^D$v^Hn+sKuiXcTS!V)>!y?V4vwJ#r_i$b=1_mDL><UaA1reLYyl
zDeopKW`TP>0U$fF+039#9x)fWN95Jy##x$c@{N&5#lFpA>~t&MBCF@I-%en`1c}`W
zVeF0*_v;N4n{m;|NZ>SXi-Ca@Qz_0JnSZHXRRR#jJv(m%$K*x^j^n`LkwZ|ehRmyD
zN~ybX2fV;TP@vMVez1oHZQkCDR$o4I0XD6fxkSb;R=%8zOlYRw!^|2JjYvV-xp){W
za%wFg1_+q3mc~DlUW|6$mNt{PQ&?{fF!!+DeN1tpkFf*^A#E`Nfeu;Tn4mwCUwZpB
zE~}doHkAfqVp%QVNi!f6z<<G9<AsL>91g6Qc<rADj6+hjfmeOFoYoE|Xt)ws&IDtA
zwRRTDE`Uqa$_yR04|RHEY??I3=f=P=9q0oC<nB<X+aidK1ui5uFAz~_FJ)*$u`&HL
z-gYnZhl|}DfZvW3HOW8bzWZ129oIdvYi2wEpOgx2JC6a3A`i&P5H)}FyinUFZyPK_
z+O?CGUd5-ZmyUadA}COs^<d}Y&XDjk-N(Cq<11dcAYEq0vxHX|6WxvhbF6i)HaG?s
z(v!rTy^7vK%qFM+=7WS)^Z|l8zpK>sT6^C?$i6UWkbC2HiW;BBOpJg77W5N`9ae``
z44OsE+1ybb85fck&7|SO8Pz5D!j{!usd6w317e|-6iD`B2eTffjjbcAAY=Rb47uHl
z`j}cNbqenC2~UPe3**qgs$3=RU-H|aeHbQ=A+&+XZZ!;}RnDeGwpY|UyW#8t?`r)6
z>&?1<Y|DoGa|Q2=4^G~04pGa;iaE+jJu%ix?%M|^HU8*dnS;8s@h`sI=HX)|{3ufF
zTCNZdmtoxon8Ef`tP%Q%<)1&r>L^jp>oD(r-f6k7N}CNviN{sI;hjF<-p-Lo)sF`Z
zBE#>bS*jX0dL-Y)(i1OUiph7Kn&894^L)YQB;2pscz0~W%)?}Ag)f9zNaPUuD|vVA
zT`_&RjS+DEaU&2Q%EtJD^Vc$_K~NzWI^AK6=lin~bsaeT)@V897057S;mH|1P)O8w
zRN=kh$T(h6{XDDMh&c2EgJ^?`i}F?AqMOVwa3G2mEldmuA^>#qm&hUvo!TDGt4o{_
zGWpvuJkWFx-SrjyG7Fbu&{k~txnVVPwjb)ww*XSv?)bwv^0}^_4Wnz6=vz(HlrjT@
zvDd9ta%Bs~z|wWA;JH!6!lUgqSwP5u!-w1gg2oPhPV^?bCzQft`^%kR=_=vKSG4_;
zM?zh}XVP(n^f0{{znPOy)5ZgTlXFsSOo(tO&?fw#c1%Mbb|J(5o$NR8?FN@AgZsx?
z1&pq(7_!*5mbw#GLjBO7!8FAkB~!9)zroiF^TcL$dUapnYuuv1bA<zs=Cx;!AsfCZ
zpoz>orKlF8YoHB`(Eu2o-x_^KUr!?@v*=iNv^@FZY3lADxBGJ1!o;)ei7e|AKEesK
z;<`qO$<>9q;IE?+{Lm3A^0z}v1r}CC)Rxx$eT_$rY1eJY7p;>oFgT(rqOlfG`JsyB
z!PwJ2TuW>&{<et0w(gd{6htR{$g<=tI(e_{(p)CA@+4z72S6LCo*y7CT!|;t-6op{
zQ3m*`#DA~!=D0^i`_m5)+pR3}WY7MJDjtPBWQMA^pNq(>p(i3EqKTeOlknE~<;flu
zFX>E_YcnmD(;mO|uoQaj=g9VtV9}9XJmx&a^D}Y|j@O1CI$?P=UXQ<d4X&5=j;}Tx
zXaGN0>DI1fOk;TL$X>LvlFNske}R|}npBKM17F?`3u(vGE8IC85Q+Vq>Jn{z2MZn8
z9qtH1Ohu7!lH7<-N;0jp;=V%GrJ<*H$RA%|eDhxBmbYSGu8_BHi7t%eVTT_X9ksb>
zdFx(jPk>~|Ep7Bh5U}RQOQf~3<kBKiRx>|`ln#BJ{hLh4i<7W`4$hN9_eeIwI^FwA
zD5%G-I%0@oui6qb7+afEinZz1K<WN8W#ONYkf2>LPu*c;$s5hwDl{?I_9oS$8qNSS
zo1Ja%JGJLjN>4x4IDR!2f5G}(wQZ_Uix(V|PZCS>+u<|lU8_((>XV_d(`(kJNu?;E
zw4^xuK2;MZuUG96V%%Hp>FP2xzQ=wnd)v-+>Thvz#1WltaNRnAVxAnAgqiYB#fB**
zQ<m%ux7JCMHM}KK^Y7182{PKVUSVPe6#DMOk+8o)_ad5ERcSSUS-`Tb;`x)!N=ey@
zzij{}RS9-f`~Bc(+dzd-0Y;;*O{!V>HIQEK017|2rDn-rv{J8}EVr+$SAh)c6LYU2
z!M|Y44d&+l(o9iEWXduY0~tJH9>q-CFZ4WY)O-zF^$2bEDXP4;U+EN7M}sKi&c2Ih
zNRVF!liA1}!;9$P5S{1yPo6&<KQt`Z=E~<NR4U0`NSv|RFW-I17)!myB|BRxWVs0S
z(w;j~Qu?L5LbZkOf?MyFMdvGeFWot0P$&~BM={v=xKZ8Budc!C!BCo3E)Xd4hb0(B
z<c}G#6G&Y{BOKx0KR|4rn61Jqb&osgw4k{b+o<GK!9G5pS%GJrSf`HmzR=jK;j-&B
ztj#&7m&)(9=Jb3$cquZQzPJorWIer0n4<54flylOdbkyREvD4}HLmz7wWQ&YB{4DE
zvaMK`-g@0X&?ee%o-=93CEi~ce9V=0jMhGMI|9>Csx6+@+x%%;{~L~Q71(`2B7O9L
zx6Awa6qexGi)A1AvV%A9qGxFElx+L_!VDL;(!#-<ua<JAuSzhkPjYIoi(*l|X|j2?
zlreMUIi4)ywYQKb;|nGx#W(M3>1TC#I9Hp>BnqBkM6IAz$t$uL87|D-i@oTf;q;@8
zG~Q6W>ReYWbF_CtzrwdfJDORK1O`G-dhHubJXi-dsLt3eqWg2zYjVCG$h;9r|NQun
zw?YhNe2ggQKbCIVS3N2=mg^T%ScrK?ARo5<mX)m6sE2l8HZaJ@ycFgf83;+b_^8N)
zobQwJl&PyKN*c{XRC8ym?GJ~O@bKV&F|t*)q?cmo(jLY7(+b1V#p9#^cBm7TP!D~7
zTGap$vA37&02k7z8o5|K3nz9Yx>hbhI5yn9j)b-L9ozOm+~+<e(U<<(PUP>n-EyMV
z0=!S;G&S;PU+{&&p`~@yCg*Ea>ow;wE#IiV`}50Hw|tnQNhRu|d7%NHLr)Sf2_rql
zifsD*&tkxjnyAv!*2cG(vONqwP%e2b=Hnd_Pv$LOzowYws$Z+>INbeRj1m0(drmR(
zq+OPl-|Wja`G)Pmc(5($5*fNQ{@pnhIY&LU0&i9CORgIFR^fZYBPRrtPDt1Oe1L)5
zPnCoS!CvY3d}GZ9yB4#c*y`%}mu215+FCYx1_mfZWaM5m>qqL66b&!^x?^5<9khAg
zBs+%95E$m#ac}_d+S%>$00)XT-&eE)+t#!WoKXFd;M8vqIp^0Xh^L1MKFFy_NG231
zf9&(mj_0xMOtvdiuvbQTx$HG0%oPqbCzkw}U3>(7*Eb^>&T5i7VoKr{h1{CEZCNN>
zS%#oeQTOT)TIer24p5;;N%U_Q_fBr^T&9%ds&4<mdnkV-`mL`!&bbxW93m4?mo;-A
z<#DCg<K46XOZbvJ<3eo$Mf2)&?>8a)X$=xK>YOYrfm<Wc9$eMv`o*{N>4dIJtRz{H
zsNm_9o_^unGrEN!;wYV=HtF(c>*3k3F?IvKQ!&29DsP2EW$F`&EyM&FwXbWRz7h35
zT7YnWDIW&3OY>I;PkT*x2089J1-_x;KTry#YA)ADiFn5x$5Hi!5B4sW%ZnK%;=jt8
zBvmdR%rOFGNYeLm^xsS3%RO7LTU~eT>O|;|P(Jk6<i9c0qrdkXRK$gb$w*PEW%L$8
zgk6_>E(LM~^|qfI2SHrA8!LtU8O5`?%zkXX4F~EW!;_<ukLAllg*hE@xj?y2M!$*_
zZUVi!hUnhZG!E;&V->!Hso%$Yx9&3gwDtxL;9agZUweJjOqY0?!}~*LM;-T6__fop
zK|NbD>}F*?IG0|2{ffM81gKEOe7qdUfopgBt5>z&7U!?Wh4PY2!*8esoSsfsf^DKP
z?7iK(=3%OT9b)eBkFm9ZI<V`kTE0iGM4?90Fd2xpyf5Nu@k+o-=OEV6YG#o7v^6H{
z)!fH{cR=}!+L<~JkBR9dgh9F&O|@R+mD-s3b?DdFbLXU^<YKnBd9Q+<_VhVUD?BAv
zo`pi+#FhG|mHp*=f7pm}N3SeD;)^m@VQm$YAXm?aO}x2*x5uvIl@ZnCT8PQG4kxO_
z*=ob1(enJR^FiTcu1<Pe;iH?4&*}PC`?b{S68FU2m;TGF;jy5ai7{&>TI6UW+%_Bn
z(&>oZt?TBpwStDsD3X_LukyYIThd#)HRxqdalb$j^7q7{K&FO;RuE5KW?|9Z`J(r|
z#w6&tio_dbp+PXF#Oa_pfmNC<3*sH`-Z`=Sg20%vd{az4ch?-RQl~U7925l=`E{nc
z-b&}!A=Y9zV9UV?wa(es*LsjmMAOW{IfECs-xlitZ}QKgJnE+0?31fiSlvIqzDaf8
zx&#STGl&+XRM(4ow|c^|<H+-bV_cIbGBGxcc}=MbOcUc+oms2gIwG_)Bw2gHC*PH1
zsTng|(jF+|kS^aTvVW!IAt1mL3LfOsk7zrBqQ6j0TEWMiE}OdJXZVtAvDDb9lKY{R
zXc-?yhi{W}9|Hcy-V2Xmjohnrn`3E=U_qr|x)8?v#vZ*VrM6x)5_S28xCj5eh~JY|
z!)vTnlY#2iIdk#N1Y&fCy2^metx5DP4ULWirR5wH(+ZLf6vc-}wNv#Ic>|mJe;0a~
zpV%t(rN?0y^ryzNu$gREzC9OX^2o9yu|wNQ_~h|Z$Qc8elA6;1@@+<Iz##tw`QqP}
zoNC}5Al8471tI+(qP{Y!sxDd^L_kUDF6r*>E@|nK?(PODY3U9@X*hI;bc5uf58d6(
zw|Vcq<2!#C!@)SR_g-txr{>!4-;IsRW%P&Y5<BV|-VQd0Clp%atBGy&5ZCSz&Bs&U
z21pD!xsm$7xv!1c-#U6*zGj&C!}$>q8EzMY)FPFNr)lgo!mYB}&zX=vcGU0V%%?Ly
z3dpu8Q9#4pa_ZE7Q#<y}xkiMTgT6WRk&V{GcZg;;^WE6?hoV|J7K%002S#L`l0h`|
z{I>SWj%+72!Rlkv8kMsD1Q-VI#cOSURfGs5aI=u*TV?EIk18<2Dq@>5gBe=$zma=2
ze2z;uekUsK*|v$q_b}k{y4pBcTDsCnGs&Hy*|ORtsRFzs2;{S*KU9=1Y@#NJ!;+rN
zfvk;wu6gp4GJV8keCr~OT~{ZWsz9~6hR35mWta`>>67|kD}MsZ3_1O*nsFb|2*R9-
zIoP1X+W+Q)N>`Y$%k_m)iyh_H&i35vDlQT(b1t0FLpHJZWGmz!T_+@=OE<j3M17qK
z!IrZ3E`*)SmW7*IZ)WRH%GKJz^&Jg*L^9u;PpO>2oX}@}E0<+hZzEyKPrz0He}E4Z
zA^kTBt5W3Rvq3p3``DkZd-}OeQF41<k^3OBG?g+v{g`RDzX9h=x=TuxVgGK>9PYiF
znX+ReU*`6UsJ<)bJ%N@phj%Ya`N__9`I*r&<b^Ir>G@0TI23`r$Mp3CEiNUDNe|qm
z=KvlOvXxf*<#P(sIK)?YY1le-OO|PAqOIJpnZ+72_sY9Y6mhfYHbBq8PN8+A8zA!$
z-q972g4<Sn+o-If?{^S00_FPTbj%vey|9w~u=XWFfWJS_s)_E#wooF@%xx0eJ*DU~
zbo)DgwtNK=1-wX|ef*z;{41xsi$qFb^7T4X1AO+_!_A4yML$?-rf|AiPRUo2Uo&f@
z1GeN12X`G}=(wgx!Sxou2#u7Q4_^eYRSY{xKpUrV`#Y`G0k@QL7jt|?e+Dh1aSUUZ
znM(HEpy6OWz_h2l7@mrY+{?AQ*E%RYnDTG|*mm37b<B6acMJ4zIpsj53Sv#PArg>1
z9hX+|cj~1@Cjo-KzIPTUvwOh7B=6sePLmNry$ce-0lBv)D!$lxnL=9Y<%Xan(tA8h
zMk@5MeB6WkQCRL+S*KiFRqA-7BjU-mvRb94o2J>av6Y%laFBxa91?PleKC&xOq$(p
z<$1nJwodV69=s&HqRJ!Aol7;QFVEbF>4SF{B5^NIE9@Wf-|EtEg$DMmkBa71UgR)%
z;hLLe3H#gn7b3I3-)Wv>HQx)RL1-fsQC6M2K-ncMbyhW##HG+qHNpN%5pi0OXT>@~
zHU5EMtD&-uJB?U4UdDyMe(BmdyLowHtta>xv_XdU%v@YZhC-9LL5KLYkH`O=YVnJC
zlegPh;WSd#;WUCYrGE`@x^O7wr@;MEJb<&&w0h*HV3OZIe6E)lJDG>gVEtq!)=LZA
zh{0-X&!0(W@d|o7WH#F?SE0+*KJM+va>Ga1OsjlRh7S|wdUrqJtd9jLD9n46HYSs9
zZDJQFle1s2y_0-UWK^MB%P-t#@lQmr<t8}a`T+-F^(Fd;_boqvHZiUnS4LzGPCeBh
z(Fy-xkiA0K8^l`=K1<9LqQqGPN?KK^qZ$!)@A;QOo1Wjx%3CMxEc)JV?F8^2BM2~<
z@?1o8`yHAP;ZaujXp5I?E-3-V<YL}*v-msi#K%xgZIjX~;9UL2p=E00(ynhy>(Bw^
z*36ce9koAiqxm=k^V5&z`s*8B9=v^wnTTsj{`wJkcWW^_MSD8J<TfWu=v?#X(VLoQ
zD2h8N7nK>B6Q!0&To+G7oB6sl3Ua}BiPi?KJ`_NmtIV&?T@&tYjxS;p@}((@<f<((
zMOBYX2G$fFiBPn%nzDQuwWTERHJkLAWrt-xRQ=Xe%l^v^`YEhAywE*^?D)UMb4mRU
zp&cqnxRQ=G?q??MI0G1g_Apu#p(f_-VzXh&6pgQLLXnbh!YVP~y`6(h`FCxAV8z^d
zXiocjn9vE>+WS<E$X&G+u*>Zhv6raa$a0wK>n%hRH5R?xY1^CZLK(TOOZxrbg-&qB
z(r(}O`PJ(CII!Gb)c8<BWrV>-tf+(lU^o2h{=tF?{2+aY><?_CNT5qxh&-01<MQS&
zM8861KP_Dx27-UMn|#gG!tdF==u0btgjgN8tsOQ46U;_{2L*9?k58Z+tg5b}((E%q
zIHppS9GR?ClVaJ?*SZfmR2@>QQ*h1;^;k5wi5z16(gyW~T1nzE;_qzrEOH&EZ=kHd
zQ`S_W$26U5f^*dcOE6e*_odxoZjpkdbaZF74zzy`HY=xFUjB~Hg{s+?^G8K>`+M=U
zWQAc@07|W`vF-e~6T1LdRF9@Gd6WK+o1J+munp+8r8gppPC|k#u6y4x(L$5{q4Zc4
zutJ_cg-vcA#8BP#?)`(VyN;P+fc!gJHSyYBOGPr^&%R$6p)EmKtES`9z_d5LNC|JU
zT`JIWOOmoiVTvX0@ew3;%&fkV7&vUIUBPVJ^o_4iLI{_f7DUw`Hn)1=)A#1Wf`$|J
z-xtn=?$Wo3ZO{JMQwcQz`qR6&HW|CGP?Gc3oLx!Tpx(dh`Z)6MCvXDaIhF(QLrN$w
z{er_#FR-Kk=Gp6E2nWf=7je?8TrCf(Ew~JGx24=53-3OU{VKI(h@~5g*I635Nz`k|
zT@4}<9HvVBqr6@`Q<(N1Ku84#G6l%&uo!kI<VzY_4!x6p`qLllGt-=T%6EhmUnQ@F
zDL!s4B(fmkmMbWU*0QqIk)-xVZ<h6p2n=_u9@~S8&U`X`by*Q>UcZ^_+tU%LyIwhr
z+^dl5y6t?IEHUmbOh2L#RtA0fRgn~tX;EV}U*vjaqlL$am5c>Ta8LP4!d7-+X@ayh
z!{wKb>84Z52gF5myZd|-nnSX;OYFDBqM-$*_xpEL<VZj5eeO%>YFr2)96Lig8U%W}
z%vp@R)$9v9!?=F1csL9xM+bXhb)$6(M5iGxl($T#r1+!+YdLvnvcN|2Tc#PtTO}XT
zB`$3QfzjSd%g>+08K<X_%qYxeJ2|}e8X$B-l9w1|lTQvFogkIl3odY9?m$|VXq)#L
zy+6H{5`a-yOjwG-uXk@k<o5_>7B+ETF{uw*b{7+lGluiI;O+bh?kooNgogCGal%t;
z+K-@iZy^J>nS7J&IjSO|RQ!pLCnlYp$uurB>r}_vQm}73isQeA&-3deE8K<h)K_A;
zjDuMxfMIx6UXBTX@6Dk$G_nE8o3nH1ky2#s43%UR4fjqYlx{Xm-_5czn~}^(dz8mO
z;jzU4uk1HVg#`jc66?&=-J@lPcFi#`4zF3>Bf6mfbG^b$Wl^5?ecph05-;5y>+Qtm
zI1dqz)X-nul~Xe@iNkrB7BR`~?5opehttfY1$ygjABJte2Y9zW1_GIq!Co4?J)jVl
znM~|Vr8(j1G!@=thGSdO`KVcg)`Hp{SUZ*cV#E$DIw^zc8w3po0?z=d1MHbM0I4I{
z{w*MwS^lAOV~z~n9$`J)pZVXfpx94um#>9IaC_nx%#Fe*!dr<m_3Q)L3cAre*9y0h
zB!OXG8#g+NoEc_Na0cp~{qntlAu$Q&Pw3ktuk2ku$SC7SoZ0vBdYGIU;MTFN*p*rp
z>qyZ#_xEqLsj&r8CJ))HrcqyRf8;D?f2ig;eN}w8UH*6T?!k(sl$-jElC&?4!WkM?
za?TfJOl{7IF~-QUv61_uvu{vCJE~t`M6)n&@6=gV@(?QCvK5h&u7HO<R6^U9nWAig
zBa;TqSS`3@pI_<v`r+}S?)V<L-_Q>EG<y|H>Oj`etdm8QkT~#=9SJu|XaH1!|F6}y
zqW)jKPIfzdv>QotM9b#`3S+-+H)czcFB!wYqG`4>E|fcy&Px7^c-lX}jbf`|(VPoc
zn3DTbtKu6A&nth>8TZ|VI0WbJ;*4!o@in%-Z>hQrMNA<mi%2L7KCy6$zDiuaY(Hfe
z#kXJlakcXHsc>p#;EB?SCx8(+*#E-5U##Qs(Kc!>`I+O=Vys4Zt31EgK(UHd<dV}W
zc>&4L!>_#)ay3jizUYb*_1Hehi&kVw+*o3HiZk-?i)8Lfd|t&J3M*LxVo4<(xBhaY
ziDrgkXT=Hou8WnXv^&WTwU)&Kt5QRNt8YB<QJA8ZF7DP_U~M(6h-UnLW?*+qM|@8N
z#QHU=O|oQTG{&Q<!IQrd>3$mpvGLeD{j)ADP?`&O?kYYe@>VDq=Ey%s=U%=P&KraQ
z&!tU)*<Lnnv~SV`->ywRr8CwxF2=QF(#v2DR51?yPYc~ON%x)QF2S@!hG+O28ANR^
z^DObhX0D&;{w;&6{xRxT!zIfU;c&35rqA=sBO~6IMPI{gsOY2b?^1A{M)>z3wN;SI
z-baZ@iv*zfekK180_Fwf1Q89+#6$nnjp^;!(@nC=RFRK_{`t9p6$v+beMaS*``nb$
zMO1;0h)*0PdsgumTM<l;B;F*-XTzN=1$KR0wCv2@ZzfQPk3s<)bl_z^xkwI9YZbCs
z;gnw#4x)_ou%!eO$b{e5lj6TwzrfkL&vrp1mt>eAartzvh?Wypb`LFfaMbc_n!T~R
zk@M#~Uqj{oL<VB~?s(XlY9C#^jYV{k^dn5x(QitZ%{(#NUTf5)m>T&-XA4FfvQj$+
z090{VQg<}bM3dWU(6pzA`b8ok&$X`pcxYiaWNpOn7_`-KqiL*m%<~o2Q2PoM?Sl?I
zi7As%|DEBjtnY8o<Z`rR7J(OC;%6m(MG;Ba)$f}V$u#h&JjnLu?9Qq7xQZm<Sv2_m
zB=Qq~DfBN^o|nKrm2%b9HIC#U!lzO|!nhtQL`RSN-8KL$r<v0IRqkST!yA8etmOaj
zW}GIwU_;8uG__%&(u>45<E;;?lam8woIF=6@mn|p`Toq#t2>W0)u|?lT~hj+uS(B`
zREvZ=vpzTrBRnoWM8CDz4w4s?+Y(a)d;=|9yhRtgbS5Y^i=AXXZlqIu`k*Kzfl>rK
zP|G!d=y8rHmK`Wg@n1SVHKW(pHSIKIWX3#$Nvg-apzBFV9PRsw4;}65LmrDUYi&1P
zewVo(qOh<u5$J{o^aIP%W_BHk70%J_so9o5C~ndShJfQHBC^H;XHGMx>wF>ue^WRx
zBRkY@2vrN>iC5oLtnH?gRvS4UU@$ru;IOfi&Fp+wnTprC&b3(45vWvKCz(M&(JUGL
z3gD22<Ufv87gak{^H8;@{t=~zk&93IS)-5p!W?G@{p-p^ZgOUx-Gz&sBlh(2M(O=h
zfv^q6`pA8mh>50{{Q*IF$(I*h%%!i}_p(o_F}9p2@D0quT{rjPH1MifVO&sUSd?D$
z4cvzP)!i<+85me)#wGw62Wpn}$-b-oR>_@BRuLryQv_eZxD!_L?G(<G`!z+sLtCQZ
z<?A`mTzq;%N6k-kbO5!0QoU*p_dK8;c7`xQYkR`kd74_O7DkNwCpq)3Q&$PG_yX;E
zcAbE`6hesM`w6J|DIr6)L@(((;PP+_6yKw{I|mIjA%&B{+QWCG*CN>#3D;wqj^$5d
zhu(k|A!;Md1Hhgj5g5(>oIh4*OU+Yp+~Cmfe_Gs&qT1^+GJ}tMJhIH(18G@Jn#r9a
zZPu-k5=N;7Nvs@UR01tsa6ulKvg7;mgM`g#SMgDUB-iwSF#5X|b21_hcsR$?&C2TP
zoulhtY)Lv1Z*~@bjLSZTAvP>4mHx4>fb`7y7E%ch70{k?M=^Qd(Vi-Wfm&nFur~KE
zZ<Ff`{h&XVjc=AblNhNS1O&vLNw4=y9iEBqY)HjTegZH#MJJCw*!4!~AH$<`7o5Hc
zGxMEhY8$ZpOx3rmTLlzqpK>L2OH^$+1D0?C!fVV5GN<s-Z*p-q=kvzWz4Cu*Cl4U=
z=TaMQYrEcGqI*BN;r*f?-}=sYn|Sn#`AbL%vXZMST!1GXh@EqT!5eO73^%RW0)_j?
zH%FjsRRp$VVHZL0l2`;M`-1vi4aKEfZqbj+2v<hsXU;v0`I1>qQ7Ltmk&$el%*_l*
z>IW(I#sTrIn1|4}`-Iz^b~K%O22j)9w4V0LNOAQ$noa6PHs({stqoglZTQkizoE~B
z1YjR+et@r8uKWjHK3V>y!PA;Qg>8)Si%AH|Mad2NHLL*|MVnc+B@C!6z4vhk^Lv5@
zU?egQ=h#lM=7S*c9bJD|H}MOfoCXuVlq<bI7m8<G6{)9G9+9i|yRwgg>94Rghl_O!
z;A~cWO*z`C6a+PM!pO)xIlP8HWpp+ST~B_G9ES0M7;hB%9B>qk6%TL6C*AVcJ;!V<
zEE7;WSA@pQnYrv6te}Oh8arQEj@6KhKkk<X<Q}fK^XCZIR9lb%7r+D8K7l8xPE(f(
zm1l0YapzNi8KPtODfqH!v*<M2t?3oUmS#b+GoW&({VKFRqh1G#0MSK*)$y1EP@g>b
z`$yeV-&7j=OUSqtdvwu~f#L3*2QPh^-&q-)HR#}Flwf25t>>KmF2BBDZflQKp-dU^
zt)g{`3vcBYw;aNe!`a*VK-<7|%Ku=kf~_ly($R~4esrjMIVCGs5|WwSME&M>qjDsV
zEh>3T)h9J10tjk=#;efMC{reI<gUL^G@5%jb?8(>7`{1h%6SY6Y?`&s?|aYybvOY0
zh!H^Xss&oCSu6s&Zvnv5zF(3n3DY9?O9N|SEtkwCJ^KdBg&liV!uoO8d^{F+9OGOw
zzRB0#kJ9yUZN7=Pr?@(By2(fX$?HpuU^u6w?!0b%tSs9Ub20mKa4Sa(bmrfPjGQoF
zJ1UQia2RgRHvvam=Y90DejOl|A0lK$7ifNk=?_N*q{_N8(SBn<h#EZx`-u-RF|9m&
z7=arvx#cIcLfKX}b1!gJ`3hTR+%x|UIqTHUgObGTWs4PInsf+&!PsM)*b5^ZH*mai
zA-uh{KY`suDE@n#>Y#=K!x{<hZ?&usMYE1UY>frAGpM8G#3KY~EQZ7e{SL5)Lk-wK
z0W$h(8BiH5Sf-6!c;8@5W#0j=xLhoqZxXn7;=8AwSsK+4+o-Sfa5BI91pJLz!ch%#
z<uCyJT`x>S<Y8XJOx^a(k-5)91BpyMr*o&UsIk-n<oB$@LvT12m1Gzow;7^kDP(ih
z-(mf7$=cc8?iIkBF*Y&A0TZKM&b5JKI0(}=w+(Q`F=fqUO<9Gdu$}*bBLmJmk+Oty
zOA0Ke$f^x%54LG*ucjD|wc)M3@P>2r2L&HFG#KC3D%mF#s2~yL;K<=atIIc({*|_#
z{Ik&bQ$(!)`Ii{}aIM@A-f-^UWE|rX$H~?l&dfBXH~>B&7gX_fHzs#jMr%HD3L=+@
z`(V~&xXf#(stP?2XP9A_bi=Xk2)Er^(aubvuk;5vro0nVUrFM&By^uu(Qnnv+}#7^
z+hNYjePD;>ju6tRqEQ#9y0!aT+usSGCZ*;8Km>DxIh_0j!*R=990!!&7dVI%kgxbW
zW~ViCDmlL3iX+*;mP%3}n&1U`3`RoT|6?@@P_(p;O)CUD$`)6&3FIJ=yj0)<0Dr=o
z35Ej@3w=v7fsyD#P|q;}C4H?ZN}zgZX0z{7f$7)nIlc~3AKtja(e2F~R2QIewG7yO
zVE*IZwO%R^=gigjhJ4LsUp1G96@K7$h5SJK%0o9)iv_Hd&x3nPUS7P|QGa56z0DNK
zGq*|^*)#>_U##6zi_X(C?>hpNaR9A;<M#(|t3!hA`cVUi0_-i5$2-B|I>$lqu$kY@
zxr3t@V|p&lM!G)trNp)?@sYD#t)KZvp-Ef{yztTuYuLpSF^G=x*zo)HH!YF+BX#Pw
zBI&25?c`uX=Yf-juO*f8^8jY?<k<GBTsmhcDsPQt3NYx7YYyP5O{swS7~!b);bX^S
zTAXc&8}4g1PO-h}4(w|G`4V$W%<JFD9U}#aoQfZEeUXKA@z$~2-#NGgFl|W<r9TFN
zY7ZS6+%;QAe&RQ~2{4yhv2*<0CshD6-Hl~V2Ijle?<UcgJ>uTgR(nNy#4c1q2_-vC
zA-#K5LFIT)Ow*gx)-4nH%d>h-+t4fhuPkaG)J!HO-C}^%$4U!i0ov9Kv3Dzt<bk?6
zHW+7N-a1nD1~vq0SIneem0c^1@bh+s`-Y-dJbl_P4T37tE*qC8650xx!SE-$wVU^V
zZWY>~s!;O+XNc}s@R9AIHJSfwa>4z}0J-|>X{CEAmuGa?XL&Y2c@By57Sb!C76DXR
zX}&KOV{0C|>X=VMk$(G1%2s4xTZ_@WY`5rFX~#Ua!!>w;n$vdkX~w}l!{fis&qvLr
z9tcYAqkpMve7H&!+WK)my#GEIGWhm1ppMF9b^*f58?4ER|L0bhjmP6sux-+gs*&Q0
zy$*2v-R~bV7D-M0w@L`e!Pqp~8wjlje@Ta4SBmK+yQ=>fjhX-XQA`!}0{?-n33B9f
zi0@ES_LK+SdW3b`DN*ERv4QX-Pxbwe+G-C_J!O^q&o;Dg&=*?Xe9S4B!o_HY#^Au2
zKcZu@?l#KI;FQ`Ed=BL^JPioCfO~=Qt6)YfqCIPWDc4CCV-XAJ;MySRk6dtjEx;P`
z4ByERXD@Ybc9<aWO?>vL=%5C^M_|qf4&19``4a`)pFNV$)~S_Ni=b)s1s$D^4q&pZ
zv=?;+1LU8LPMOAT<3H!S?we^6eNlKJ;p+2qpp;Zk>v*<$bA?2Lb1QxJe#uqOX21&o
z9ynmfvQqpG2fJF1K7sNtI%C1p!o^YGtI6PyH;?(P<b0LK1{}~jR=6p~yl2@|m1XPd
z1?2`i=(Qk4E5eJ4PN@z$-dt=YlDo91d};2b+k=EXcCD1tmJ-k8^0P0hRIhN!m&bGC
zX_&VY;=HO5c3GfrDVp|vq5t#&%9H<@|EN^FXyK#C#?4CYf^xne%={v4z2;Bjp}fIL
zW*t3*4-zN;WNX`&i#iQv+vL2>P~J$Sd)dUm)l$jbvp*mCoZ*CYy6}djDTk4bI3q2b
ze5i#e4(<boo!>N6aPHF1)BwgIhG@QnD~9zJo+f9+%1A3DP%zDv)4{DX6)HQ7AklVX
zWL0Tum01Jqrt_v*-rDG;KY0=r#EGk?<+f)*zkV@O+t10^41yDed+%XpDWjwhR8Xmu
zgQ0RL3%npe1mDT}q2{Vj{pka~bSY(2rbmZN_rWF0vpd~OXPog?C-=Gb+(b<5phsWq
z1q{-i1->DnflkrBmyQ`1_{1}Vz+C{e6!c--;sM;AyK&PiuO|djsEG&m#BTP<E!df5
z(0fNZ@ndtQkoeu3SYGl^BZFp1=kL=I#4QGt8rI|?Fc#nqxGpl8a908!Snkfv8-3Dd
zjWRT^tv;Zm19~c8p@tI&$!;w~%G8^xY<EqPTYArY7pPjIVOhAR<AT)sUcYvFS&B~6
zX(*)kbOW&1CrKAyWwch7ILnQS-C6ULcNc&idG+LPQYAE086r3LnH-j+`aR?&4uFMT
zpo5vL;$q+2+Js#-VH5rH*o;lTcl=J8t@3XUpn{ZAk9MN+YF^#O*Z<1`<Q!MqMgtQ3
z`V~8Vg@?Fy!Yxj!ZMI}{LCua<vhgO+jh@u&VwBhec3L|o08iDLm8e|~MxIfYj0{(i
zXj@-j&(UNFvB?Z9pJ57$6IN=3UR}CEI9gNp`WTV9XfOy!2zOjQuz($=wqKkd93Ws9
z4GsA6k9<;jXm$^0)?j`-%w_$?7oB?wmye)AznIn+Bmgx4CK(C8;=O7k%%-OXj{PKM
z0I-*VZt>1=VY&Db;r%<mw~M4tAnDoQS^QZ8xQQ)}?7;1~^qK>Wo?h00Yn%a#7I5|B
z2yKE0Xd`af{qqc!E$kJWWNrlk<^XP4?X&xcb7?nD^8Zbbt-VQ#CP#r0+3cUKw_*$F
zM~aNfT|C}|37s9c;-~CRZtK{U5z`v>6=Z2<82V2bocxqu65^R&A`E7`DDcwJ0*IN3
zvWjH=zwSXX#o!G{P8PWG|H({H^?_w3&Kd~Ei?SK!zaOL;kKhRftZaFRcLp_LM%uyk
zg)rGRkLz9K)r$tfFdzV!TJ3c~v$U=loyf(I4X*PoE%NMqIdfQu)_G$Bia^AM^b^!x
zth;ayz!tkI`B>yC3Vz2B13p7o_5?e(iAuJM?#PYP*iDgEQfmgt`xR8U8sIk=am=FF
zLHPKc`M~4d-J?mCyP#^|%LgWhs%lei99u7_vP_vWtWbjVnqwnCDsSzJ8Cw-WK!L7B
z)v5<CroMxz*7Ks|!r*W-N<Li782}87>fvDWN+2(IV&}no@GF=`8*j~#R8rzIwcGoG
z$J0V?yoEv%079EuFNf%(u^S}n0gQ^fPOx+Qd<mnds)8XDa5vLT$=hN0r@EL8QpdfW
ziV;vxeHtH_Zz1V*!?vpmhJz~ubN%*Eq3(WeuYgn=a&Aa!Y5u~bXg)j_UR`qIZI-eF
z+K${BcmqG6B-jU<OX(Z+-#$N`g^Gw3(nq52WC7L_-=4(PPpWTI3phc|ozLovA(S7d
zljy6*{SZ0Zi3^$#!k~j$qbIpg38Pjm@(u3893_uDiKjmWA+fMJOC%<l`vB1o?A?l+
z`6e&lJ7xcaSGh8JkPBdnY;v!vNWap}_TTc--?QE-t)8Bfs%=UbX6{HKAi@U+8LK@{
zaBFf5IH0ak6SeO$)O45rQ>EIPb@Ro8L+N5^apN+=1{!9e*c%6jhN>1?KdRuFSv#c1
zqHmrPGxq!3maha<GLNZFi#Lb+h!P-+Z%GJ(#hlaY4~0djzDHAc0Su7c@~sm=7teyF
zFL9f|H;cx`cD;-kC;9orO63wL;CkQou=szkeF;Ex0zi>{%&W8(py@fdXgE4urJ4UL
zjbO!PxjKn52|mpDHIPVPYx?xJ;%HtF*<IY?^?hNV!dh&vJU-F(F+!s07N96QZ#N9w
z(@tFP>;P`UJ=K=Z7tmOLMxy><7GPzrkAB1ri?y8chm~d-p)&VqPo&i)amOGy-2&X9
zGJ*z9!jBO`LD_p4#qGTsCfbHI;+nPFr!E>?i_o8TF-p`{`6qM%go=!nyyV^vh!+7N
zBrdO8Ma!LOfb}(b`z&Pf$9%z87p-S*?3=E(oWi$$t6$Ge{O4>`5;>}yKLP0q*a2w&
z>&l2l2LTXx?AmUkKw2chef#N&72*jWo7}m6=Ywp$Fj?5BDuD@DPb^4;Q+ds@uk0$h
zzT!>nP6#9xFI&|(B7t407W^yA6nr7LF!z3I1OQqK($Ov(a8{YI6gs(<DPw=p#Uj|X
zZyBh+%=cymhKRpF&N<#HJ(ArbTlDu$mPb;LL)kAHzbZ$J$Hr=jWy?ly7+jI|$^Nod
z)2W&Q|F?xPE=!a4+{l|lk?|kzfco21MMhi@Cy2K9r~kJDvCiKWpPd;t)l()V6SH+H
z1DWTn32#A-vXA;EJho%I`!Kn{;b))UVi4OYAyIq8|5i7fRF@>au-gYkB3KA21a~bk
zOw80wwg570Iqg(V;iK$M!fP0ii;os0kiNh!E_J5YH2=y|KOu#JA3J<p!P8srcgn(i
zsOeA4(kg3x*2oH1)ASkzr~xSxwf;sv3;vkue{4iPs~Pt@8+b#-q6&K~{MN0Y8F+c-
zF6Iqq&GAvGwY6r75DKni3J360X5c}M+&JLx_)c9;Qi1=_0xritYKKYc(?}<yU@vCB
zU*`AhJ>eV=IlxWe=O0GdJMg-aokIulKwSyN#U9HWYZ5)FR0Rnma5;)EU3!%*wLfXs
zB@gLZ%+k`<ReHjMNHQi0hzc()!}k|!alTL?nN0Mvm<w3@Gg(W*sycp9&HN`dCcW7m
zZW-GUzT-s&M*yQW^@=sV>WWj6!jQ@)DizRie8~X34PEzTS8K!X^S#a4Bg3G<`1dT~
zd42y+1*UmQ%G8ySf|>02`<@PTLW6pv4K<i?&XJ`oE)#@vsLxe+br+vYBQugCbkb0%
z*j0VssfwuW#(U#z-bEn!*zmugz+l0J?;-}Cqlsr~(NhNj1Bg6i7=_=sQ28FdSa(4&
zyp?vV|0(T`YTco2u8zcgHy3dFkQH3<pVj9mfj|peo6~0o`*idgj05SRH83>D%gp7?
z1W=Eht<5`#e#!)3k~IGpus=*d00lERzz#`Q?sh)`;v0>&E+LJ8IN&;qWSct<TV_Ko
zlHS$T=Xc}u4}5)3v<;uuWJPrt=08vIN0WkL9D$@?d%ee+Nb}OUn#r^9#Q<(LoG74D
z%pMF4grY!)JCQz%`duY<4;NQKaoRvBp_|>WI|OoJnHWqPi+3)%Z46q?J>B1rgtgS_
zLdA4xfi4gohdaEm-m{X@cyX?LQdWIY$s>H=+5%f=qIoYGR|BMExUF7c0M?z#XvWAM
z?rqZnB+AH0&t-Se>9D|;S*4`{Iczu|Wg*`1A}R5o3)+>h28XJC=472rCnu}73J%{e
z+?sq>m);R0-^2bFtI#8~9SZpO42L*V^e#{zIANUt1VOydlI`BX%6Xw483WXus(eTC
z=DE)RdB(-y`f*7v?ajcrj}lD5PHE1dCMLBuUjt3)CalFEL-wfLlJ?W*3a;7B49eHb
zGu?b4K;FFu^f?=7DVfUT%)d%`xXOE~Z9iI1vQAzmge;da_kY~YW$Cw|<qKP*oo;>x
zx#rmHmR#Z94GS&jILfkH!u5!GS6BC5Uaox*U%C=QI9vT+O#VNM;7v#Y7z<Cp<@3C{
z-6ZjV^Dzp$`(YgJ0C(HHPHg>8evi%_mGNJ6dYzlR?l~7`1zd_y{3B{^Lk`S*)yNw<
za<7C@DQxi6=+3*9X{T+dk09Y+wh*|RgWDccW3#O6%#B7hYgd*kia@r4V9ZXsBUQVC
zM_^U2${P+IM~u}jKs%P-Armh8r{0(eo19E{E~sXPhh7;wb*2LR>B6oL2vukP{-P@W
zCfio7ydW(MNH%x)TMGFF^If`~79GU=Bg9=+0Jj5~DZ8p1X%;!Sj+0BwqRP%7RUgTA
z=kpE~-{r%2DTor_IE&vBmU>#C5Pe!s23BJm+nf&`qP}?ok*K7k*v8N^2@9pSZK71Q
zkC@m-aCSCN{jonv=3&5T@G+dw(jr>DEuLP<wAG{t5CQZmoK_H6*_Vp6HJks+k5>4T
zFLt{#vcXC6!o;lx7Jw8o(w&|id(RZ&V%{8DsAfy%wh({Vltq^?lVd}f$?CMqCOBHY
z!kn_qTq3iFj)A3-+eUx*z*6H^Bk0&p%~pXT+ofnmGiG2?MKsc88}tXCwpuW5>4e&D
z0?`*E!N7u1k376rTF_#0j=NO*s)gxAV3qHR5golW{N+ng_)U?nNUQmz{2TjvucIjg
zM9?f%s3fA-Zi5{DEeMN(-FhNnnuP?>2QL25Ea4kS?M3y-mpE#144pKZp4lEy3uIgf
zc^vyEFTi(BVRHHioN#qjDy!ZEN@m7Zpta}Pl|fuNp=brrr?;oiyq<VR*)5E6c%wGp
zPj|Qft@tJ^$BUOeT<*3I+uoOPhilyIHn)GQtDcQDeib-OHJ{a9s3)pem$$a(vNj5`
z5M?$=hz1~sZp?Xr%8B*(imL<B<||r5(uvvK1Ku^)+?R`}&8vSeOcSttreT=yY^5w|
zz#susWrR)B<)zlcdw1jGpQyO_Q`6h`wcX*>_Ji=f+~O5B0E017DZOK|fUUi0?->P<
z%kw|qxcd2aS^2SF;9iBZVV*p@m4f^YWh|jVR|rW-2_U0ewhD)H9(uEI3HSaPt=bTo
zD?b#Ca~4k0=RcT8viVFwA*LWIkuI)}|0?a!^k-F7lr{xP@%5JKg>hja&xc-@%lPrB
zhp+#Y97tMeLHw~e4aJ2CBt^%DB^GKCIX8dq<iSAm%)RQ{J%Iv#6>21zJZ||_Sr78r
ziza0cY3;WCxDKdYi&DBjgWx*PA+4wTMF5lNpwOkqz8cfN^m_QyzGJR$(!}bG<_r(r
zA^G365y0@b<7$5siX}u$f%xrrooJ-vfmpQ;Q^|FyIcx{&GP<~A_!rWJy~Q6~RKjT!
ziTuVIS((I-%u%ej2TOT-o1IY&`w0O8+rc^kkoH~9H(If@a#|3>syvM=`0n~oln4FK
z>em|FP1rG<=VOHjSU0L6X2serK=YndC|5x=W+Rj2wYg7acmWnj>kQh*dV;3QZ%7y}
zOOA@KE5Gk~x$TF{9tG`%+QNE*bae6$(d&z$Xc^ifFs+|47e>+=o@oK+4N#IQSI#-;
z;o7$^YJDpX^^tVBgNF~;jH=bCa^z&!Hp-Pp4^+e_esQ78It8s!X8PU+-aH9z`2p7m
z09SLMSiAvO0--b@8Po295w!Mx^D!%Vx#f@&h^1H)b*gzM<DDm_t!$$J@oI(A9w36#
zdoM9F`Adi?@@AZnkIME_YG3rUx=<g%lUiH5_Niw0#%HKSjf{IIHJk=GjWcPjkGw)$
z-Q`qL*l?`H9Dx~YrkS(ahpkvMdX02km2&>wCUY`>jJJok24sm}LMs!zti5;}96DVq
z8&`M*8L;Me1A_L$mh?emSg->bSUegp24z9){{FaOxnWMDo-CXly(?%f@O9wXC^=2>
zgA@@JM+SY>=kQ0)aGkF|OpbGL6>YL@Y=8&n^rUi&w`ust(ZiYpj7@#srF}jh41<fg
zoDGDMIr%VJ$pn`6&A(z9@bOG+-)rSM0wDpt!`G9yBqaNPHh3NikQJG)>WW4tKoZ%W
z(CAS2s3L6%)Q3(U6p3RKW)dVE4{75f%)~0)+0NfF<)&F_`eJg`$=|`(7JY2?sLg(-
z|2-y$J8)MzW@~%}WUO9JoN`QB!otgsIndR~n&U!zcgLXv*^x>y%{laCS8EVDV)f%1
zVvZ$PGWZw0SdTQE5<^kGurZehDd|qV*X3t79z;1xh=lobkC|k~g5Ooyo)oe*T_;u5
zpO=!drd3a9Ce3E80-ZKG_%Q-EEbIHS$nN^NZ$@*7SSDA|5HMY4OSjMyxYk1H`vB)=
zuF{Fq?0G0&$EbSeWulp1UdY#(8S)wHv4|uumbX^L>|d;;n6Z`&b6Dur#jvM0pp;be
zI3PdLvz@Wuv5*$i?@DA%ZS|`M<p~Lm-o@K^2pnHzeed?!M_ka(8LokLX`6~G5Q<I+
zvfzh|&TLT&F6y_!QLqV~=z&%qgL7)D=-fNT?t8Yq@b>g~!sB{GoRwj*oHut%rr>Tu
zFYpH{Y9EsBr;A4YI%$Un9-qDNy8)H<2s}^d@8N~tx%FI7W>j}^Dy+QS5+_cy07F}>
zoi5{aUavHNJTK%&Vs<;;(sh`OZDD?zqn~E6{LX)M=%kgwobV8#k(H7B`$%o~zWPp_
zXs94`vM-MMU)>1aBn|EyBHU$U0CrFx6QF3Z0VdX)3vLmP>cDv!pO8{YFC)ItQLoMi
zMMyBm8}o{O{lF@QK%ds!A9=r{t5eq)0*L+9KvEWsKL>D<YZ^@X|J)z|s&D!y*6#pS
zm;OdSB!%^#F8kjMK%C4y?3ad6R$g7*OS*be5Fq%!cL_PM(W?e1xzBA!SL@2@Ee(MQ
z%I>e<B?2iI9WGsm{WqR|QkW~l&rtAkI&{vv{op72k8cS7nCRwm>jm{Nr`+I9I<x2g
z7X-1wJb2cW2H}QM7|Jtx=T!G#;!k+by#QBS<x9jE#xv?Tob$(HWt88RXmGV9>a&l6
z*k-qiwuZj%lar-U7msOt+<zS{gf6$^6pj>qZ+055PY%s=9RkrT?@d8Nu`v05@Ch3g
z&whSqS3WMI2+nN}F}YXWN^3H=={~Y!ki+sg^^dt_^A}OEAwVq{GiqD4qr#Dnvm!fw
zPfZUnpyj&5&<EJk81*-`2!O9%HJfe<r4zH&7KwT(-BBf#`AXS|mTeB0!dHHiW#b=&
z7sBuTa?OGEx2j~=ySQ--gt{!L>y5sIMLga}q3dq9<w|PY2{7dj@SFZTEp+doM)%Wo
zXNni=iGTM_g??B7R+E2wbE9ki32ms-a!}y+2y~v60^q+3Ar0_C9m)GU;=lendDMMw
z1k63V7lALD6aU;gADXxvF^gUv`Sqi0GRYoN9;O>-a`Es~^yWR7mxAddIHmttjrF-@
zom2~<y-DggQZW~iiZP;$gJ9}h5Tj4AwiA!zlb-&`NYCjPLDx@aOFJ6X<NG)>+DZ3l
zK&c`6K@4QB|FTnMcXWr@c9fR#92%a^FsA;zBQ-$-AhE&)Dz$=%j(fE3=IifeuC}Jf
zw8c&=MnWvGOwOJsk?@sI5J2LsiEiV+you8>Jwif8@-at^rd5Dllxy+a4e1-=nRzH6
zO|JVolP@-AqBM>Io28hv)64S25tBd{79_441HGj+6D>Y3l&!mT7eF{XZfa{7Ft>+4
z?!{%-=~r>FPT=vca3&FLtfcmNS8~iCc#!~zuvs{?z3Cw!lPCDW25d(lciWcP0T{X%
zS))LF&}MAR!%*R|w9r&j?Lur<SNq>>Iv*vKuNelQ&Q}}lNh?eVXo_vL4aPsszfw9Y
z8H!m=D%jrF6V)beS_8S-j>u8B{rIQ)N*{&Xgac^n{>ko;A8KtS9sZ_xUVF3|z^{ZE
z6>(Jv@eS5u9VUUf9rpFZO9zI^LSe^SuJJn%hHolptV&u=cF|jr_Y<1|-((;`*=pyp
zLVxcFV<gVYMp>UDcdT@HGoKOl!l*y)IyTXU(b(w(UufhZeH-@GSf$QPeovC^V1>NC
z<4&HEQ&JN<*yOOEYDNlxjPDgB4>N;+PEjTvt0c0}s+(CwF{{8j7kJzX!tE?bcoBuR
zMX=6tS^?f1YzANqg~?LD)fiug&Ps33TsTIp6}sDd_DOwkS^8yJIf^7L<erL&xlBWp
z`Q14@v~jv~v-Jq;sP?x>%9ks|gYLQ}mnIhLuh~1UKpe?Lb<66vP7Tkup<c)ukUm18
zwp~0*nQC9|c*N4cnzNT6+QsNn%eMcfvmC-Fh04h-4Hj(qdKbS(Fb1%RB^a@jX%esL
zP8$BimAS{{?rzqc{dV*C)wh^el@Hej?T{`o$}Dj*SI~7E$!0U6!6#M+??tKY({5Z}
z1!Xp&xK)V^?xghOx>i#_D{DOQj$k!{apM-p`hY)N^W81OO#T?FoN?8_0mi}MO`Tzi
z#2@pQhXd&SJ0$-rl%sjubeGspjdiVC&FoF}nqg@tUWOmErj4Dj>ItzaPins$l&Jw+
zd&%SP)2US%%U?wM6&&hx#gjy|q>S2X|J&z%vxx;#a5#6OM$S10h(jYMk5qqS%IQ!8
zXnr({BNDVrT;$XZV7G})r>A;+&Jh>w-_1_EN0sF#z*)-oF4L7bxL9+!DE`v7MB2dE
z(K35Ds5St!J&NLp+eFV827l4IK3z=^HZ4WocK`tq#?5Ht7AS#yUGh8_%QX$&o_z&O
z{9D*Sz!;Y!8oeO6$DE#P!4EeDD3q#!4RCGJ%C&6DTn-{q#;Xmn5}M6%1~)2ol9gsw
zwo~xm_GV%bUAnc}E8%(R;i{e|giI@N|31_UHuJIPH=IL`g0tooFXzX)fCqZ@bm)At
zatRlJLk%p=&joa;xfd^68x^%YKGQ1zCe$-nT4hcF;#he2Q<ou}FRvO^IYx)(GpsK#
zrW)I<evi}3^^f%}f~c-@EXfq<eb0tdGltz|GUz}#C%fG@>^v7{BA&i(4C;J*7-S5}
zNR3KpF|2Jf=~<LoIgPFoZP^;|{0S%zZZQ^*u1v#~*!DRk!9ebl9=>yi!bX!Q!wmMF
z!%y`;px%C??2DK`!G;2FwmQdei_7G(TJ37)$L8^MfX{}ZDC~^eTa{a>$yF{W<_u}D
z--hA`r$KhLrNVAp_<gHb105a(qCm?C2m6_REMrZu&HB&Cg4-wUS(Mf~RZREi(H{wl
z<qb~TC_Sb$!!A6%I1M3fbMsT2DC%S`BM1mFC||ipJflR{uX}qk`s?Vp3<U4o#vlOg
zfUC$JPYq9<9Wtfo=JBMX4R+x#FEXCiDF?L&n#^9nwWpZGcX!7p)>(|=!y(&T3AV?s
zLKNs9vVlMrpnz@cGXj~CzS4P5#$y|A&9*Ny#W8gx&7nV32%3JM8gR8lPTgMj@SOXu
z+&WP>RNQ-nBaplpK!ROd#19AYzNY1%4aMwFX7nSoTg6RrXK~>p!9jgLhU3%m{ZLP!
zH<IJ+T}D@(V36L^6yxDoRr7*84cJ+Fb=#ti+m4D~Ybs!YIm*278dP-9AHOjiZFw7_
zWRp$T75WZ82%lME268RlG#2<<%ZmmQOgL#@AGzcJ>Fr+3`BWJT#GP5Sc>F`RxfpQd
z=<kaG7qcVni~`yT1d<&F53>@Hv0+E4242GY(5u3uB>*sGM`h#`p0;3bKoYW(_AGCj
z84P_S+!--p%?_!KAilB3D_|g`#1sn&m{kvxcf3HeN0_7pS)2%~sHolq{mGxJewUc-
z0(WArDe`PKhj9Eyuj5$7fA9A5ZCz`(S=@J4qhKxJs~+#i{JQt5$}EMMC%vwHoP+I;
zWPQD#`})amNkXh16w(R};5#bl1J?oj@F3|&X`H`kHXZ?&*3#cOIzpRQiG@b^p*8Pp
zZvkhJkPKRUJ*stEL)XRZQHjiE=@#_|eb(4s`;i|+_3=#P4QHJwc${PYalpIuc(>t>
zZDz^w#8A}Nfoh+iy}ep}4iJ?%cR6^hBVoqg>K76}=0Vv0ORC_Z%MOu~71U<=Mk&aI
zZt)+=@#~b%jy!ByDafMs|7?DhcQQb-;rCyS?p_7<`&ORjxpO*gGZj-bAl-5~Tp5CI
z(uRo}a(;j(vmRd=<}z|VS?`o=zT>bjk&&CsX(va+!uz9=jTNBiGG*~`n*x$#yqRY)
zGb-;y5aRxNzEIA0R#4Z}xA0ajcfj<sk538MyPU0cG!E5!bLBAtc?PfWfbE0%+Dy^1
z9g~|b5Y3Hco2!(tP{IXq>OhXR8iGuu8~)_1t=*(KbZ{N_o(<hrto~1acPlGS*I$dr
z4^02)0|)`@7LZ=vTiCTBX6n3Z0e*)sp(ka7hoLlY)v2X>Xf6o=3|O;RiHC%tXt@kJ
z;}_TezV!Hh;f5SDEX_pn+Row`+08;gl4B!XHw)p#;p{d<n(lVtkbo}Fxn@OaTlxUd
z`>~0^fBr{}o=gPqz9mE>A@48r;@1njO$W&iZ{>k@`F7uOZUxaKZ=hKqi6a>pcLV}l
z@;|$B*U8%3=UUf(sIA&)zfag=2nQGKkMMiJ6qv3>ttzn?KVK*DA9xJECxA%<KSuDG
z;p_kUebo1|P2F-(FaOI3mpM&!w}}v((z0{=5Wl&2@YDU`ZHyaogPS}s3lZM3+1^J<
z_r*~AyP=?OeO<z1I|uIBl<A)Ul5F#;E#QK-klswY+VpgM@ub<1=b})+xv@)vk&TQQ
z;JEg-y%Z|lOymE)T#H&LUuk>KtEiqGdncTdZ%)KcGE%dXpd9vspREp55K>F^|Kfth
zK!Pc$0u{{qLK5|TID2Kvau;Wy`FNt2WyAH%Mr%L*0pF5RXH?HVqS$4b&POM&G>e2Y
zLBh~3{hc%m5?<`qZn)pN?@Z0qW_>S<DQ>?k4yP9kGZ5W1yM#+BN2*!^bl9Ih*Dl)b
z#egOa4i6~k7hexTWy5~nC4`D{*0f0x{5C|<Z2LM1H;@qGG-Yb*0tq>I7`!I)I0;*B
z*W669dpj5^DJ2jjtm^1E;y_Fgh-q-^airj-2^8PnQy9bDjANzvjc!j*Z_9uVM*i@>
zsKS!G9pwpq-%Z94l8$v*fx6xEmu&n6kxyj`C4Xm!RC3=3o0Ho}$`5UNosYbHHebxv
zp>t=WStn-W@5{89BM(b7agmK}S3vF(I@8YPXfDT?GZ4q&3W#V(`K!f`zeH8$i(S?(
zF#?|+)eKBj1Lpz&C@87WCJBa?W~q@c8&!(j*;jYVKltY&;n+L_FnA9{f$!LCtj1ip
z0TvBRJ2@-0(kn(DMVsss%%pdQk8Tg!#nslA+e1Qu`J?+b@TM!O<*5?`pe;9`u9jy0
z+JS4<MU75mmPROFViq}B9ci`)4Se&%{D{QnEbK=a1DwoaKHMN()oiua4(?1qd^p|)
zxe1=R@DPz2x%=51RUIIaRoIAPVYMQSe3ona%udLF#p6!Jy;#iXaT0mI8t(K2oK_PX
zmp}S?_n4l(awUgPM)4k<lw)5;{ovQj|78KtlqJx*RtZvSvvP`Zj<06u54<`r$1let
zVG+6BBSsbn%A~H6iIC1vh{&9_^0RjeQ-v@rnTuPpoeE#|je6vGnb-1PjyC#sAO=W{
zto$(0`Cb)HIyDs!kIX7`n%M3??WSj_c)rO0J`t;1)m`)Vky;VoVE5ZcKm&N4Sa)}@
zLnIDck!vXJl61?7+0#G7E{UB4=I?g+ZCt~zBXV*w??CQ;za?Gby>I$g>j@T9#w;Ii
z0;jF^e0HuZv5I~hav+0`(u*&5JWbb&F9XC5oUj|p92`ef7nR$49KtRq%gdL)$DT|g
z-MUhNxOke=^xY(--$4Z?&Ejnz5X2myk}%Jynp4|yuX>9auGYmI#voOgv$1_<BK8_|
zBww5Bc793>N?L_R!?NSEQ8`#GQy=$H)N|lMy-y;MOQ|XLB{CbQTMVUe7R?kz;v?P6
zc%ZtwXWTRU!vo3v@f<YplW#ZAZUosI#e02KO$dsGw2*)-AqH+NG+o4)Wmf$gN8ru=
z)9xUxer*T)Z#s0P<B5@t-YM=DL?ySjnB5D=CBgea6o&~!?c1Eel5CDxdxIACx*&v1
zuYt3|1K)!>xbELS0~@^UkSl2uCmU?tmntg^VQS6uvjWgAnhO@RY7?WtPeFp$@XNaz
znuCb^!dt=Jr}vc;o%=+|cHTLiaNbhx#$yX4ERD8U!l{+{^D2r7g0YtykF0E!ppwym
zuUOz+l62=6#7v_5Wfd9T_9qZXTwH~z0|{KG^n?Vr$9K_;^PZ9Q&T_TPKl?R&D5iMx
z*uC$`Kh6n4qKp=yfC23jOSI!C9F^@B3eB*xuu!fe#XJvR@pG2x^JkU*=y2A}IIr@}
zpq4f2)A4{aXOnCH#@UO8wL7rOV2{-2T*1);_-{r=Mo#!fuuu#P3<7sed(zVl{09hH
z2F>YRtc(-UDc;vXr=8FiYdEk+WJo05?77E&!8-NhCLW!8mA<Qoyo;glnI&gJ#_ac8
zZ>}H|pn4~T+^`nc_}gr;JUhMc+8V;8l}?8PQRO07rdtQa_8k$hTkp`gm^CF8j$fQS
zYT*wekOKa;KTQ18;${Xvf!&ByG57Gbq&5GRsCYa)l8_UMhd%HIT%2NgY|j%q`o8NG
zu$1|O-L1%;lP!<7X1IY0ihJw%24LIc$%=KH8S72I<Br2}Iv)%U^MsTf?mb5RqaP;a
z82w36^_ZGkCC|kqD-bS^KH)?#z`7sW^LpmfX5jnD?OxTEdpXkOqScbW0k#dUaWhV3
zd{5#NZTNYnd+NpFs+Iq72+P+5|6Ko={p#xrwmOAF7=m|}_y<#cG@Bq6!Hz!Aha#Bv
zawLkOrd^KK-It2#Y>8nbf!a+wLW8VG^38_}D^Mqz#pS5$3+Lzh2@wj)0uqj_V>eOD
zLXtv7PBbV-j<u}2$Z%IpjMB)%0X#x$8nZ83MVK=rSG-@1#92)=gnL`MNHR_)o%!-L
z)$>o8-|34}P?2wJ#E10WpTmqEK9P*wR1{_cZ?A^$?ah>{jEB!c=A3Qvzq|SIK4<A0
zDJoHrv%lk1sw8rFnr_Qf-E&e>LLrn$G#&ddnjnwGd+>$uQI61)073(_N7Le{lOMi7
zQCxb9j-HD@hb{{eSKqErx`Fct@RgNiCO7Ugw{4L~ts^4vi3-$$h5~+~a;cvpO$oiG
zy@UEphi`VJrN>?bua`z=($7xj!-Hd%H6bl2Tvcw2F>6h9ES}iJV?={Pjl`O&;1>DI
zZfRrb9#w(ad@*MQC^@9L4uf~5VX)vI^yD1dbx*!+DrIw*XAQ5gZqW=B4-#JB37jFj
zYW;wyb8Rp<NO0@7=|Wl54&Hcxa3`#trqFl^iBao4cZI({qcE!5&8;QB2Z!IOgem4!
zV6BQ&ti97e(USl!ZxynGje8+G;NKfQdU$nwIp8~}=__^#kN_3))aa}Hd_<NYW+z6I
zz~Fyg^Pl!3fvUSMiw{Q`xXJW!dfFw~gC8SVO6Z`W4&JRjF}L5BbGC8!T7jf}7{*9%
zWFh4nxN3T+e9GJTb-aw0)ZVur=b!fhTz>k`nVV?H5#^((P#N>ZUiFVX-1pxH_?Jho
zX>rbmUP!dHt3&xJf?|zI^fEZIRL-9R)RXO?(cW#>lsR=g<L~huA0B@oa#bS<B9)3@
z!D4V4ds%I~>|F3<##H=8y3%<0$^uVFDUkU}4Al4eQ3)mbtNeH5BDsyTcC9k<(#?0O
zzhfCezW+R68=1f!+v2aHg|^%AyX0mbUG?9XyjL}@GORx!iVu&95p$uJsGVJ|9&Pe&
zObg|s=m-9Te$VXMoO{2x<WDc)wzePJ$X&4yC$al}Hm}2Q8*O1~(_q~P0kE5pP=NQq
zsXFTxpf5twymX!Xcwx5}GvnO6vEM$_Q1nnh<yjAsa!faEDfx>2I@mKI4yusM=3Xj_
zVSD(zMJ9#q08~q1Q@w!ET|sqaoH6$QN7Gk^MfH7cQ-Vl?beEKLcZYN%-Q6*CDBU65
z-KBJ=v<%&yLnGbr9)JJmoohbvftfjHoxSgM$6l7SBKjRr4;@INKMwFxWShg~xfarV
zcsm-raIUSvanxh9?nwxsKiFR=VQtZorndcYhUEQ-Yj`&@M<=JbYNS_MVd?4<1NzC7
zArrlvSA=n=-XQFwiF%dP&k{nGRF|8s9X21HB;O}pXpt?K7J@S(43?`ds;J0sPGZ<@
zRk;PNcXs)}6Q2sJZ+<8oxa2eG2BtJ)ujt=prqJ?aaoXZ^-iyo#?Dh6LDct4K1>j-e
zha|0&m$_ztHc684KYJtKw!e?kLgdLb+hHaS=qF{5!F?&lCQNtVgnU96*H&3tU3EuZ
z+B#i}ojDd{Xne+0j$7Vd2ySu~1S<(crOxQbROUm29+<5%xs2LhNF*|*(zpaU(uLhy
z2JDeL+dkND2{d%J=jZnhw}200XL%vJwhpAyMT*(K^gWP(2LkInI3HW(bL0(W<6JP9
zIsyWRn%0u_iiJfmHF;09hi;)3CA|CwyR>Efxq4BexH9xM|4Xz2Uf%$sc!E+r+&hhh
zGF97%)s(nL=bgKI+G{@_r>-0N(Y=Mo8V96rWwl5)#nlL2q}Zp+rtSWL7ibFn7TT#s
z8%S<bX?F1`v23@Pn0z3M?}f|@wE>}C=9Pjcr$1r<L(o7Wv4h>RyY11y^G3~xRU6LV
zlcmOdWX3ocxcBjGH^??!rIgFPJ<xpIPcW{R8AnIh|8}qCMl?G+VDmC|R{aG71=JMD
zu7sPB+_=Nct$x2?<_vrcIxTLD^wB&lXlvqt|8mmS7MOh!Ud2S)KVj<v^?Wo5YRJwR
z-|AQ)@a}l*eF;xehYJKA<E=|j93{HHVOv3x6UnKQ8iT^)?oM@nj<8+dU{v<GuwuP6
zaYJ>h!lii0$wIY5Bh<GhXZ^yfjX12*d=d~Ctqf_uAi5jI%P1a<OdkJ)3^bw}!+p#Q
zoK(%KD-XzRz~Y7ne(Qf^H<*-@b5iujeL5%Mg0#G!Z=L!>M~y}l-DZ_T#|!_^=e%Ai
z!+~PouQQ8cJB3CJr!8ue89D1graDx$xj!FCwX%~wci#PnL4SN<0BN|EflU5GzM~ti
zgG_7u^`sfS-rQ=}p;wmx&&25>00b}XUa7zlr{m3juafenl$sdWdNWEG#4|dnrp}Rt
zxY*mz*DsXXZ|fgrY$}|)-f)`2azrFNH6P04aJMopIC+TARsyuclXAYh^xKQtrsgWj
zbUqn7=6>gUqj}ZKa}_<2HO<}}Yi5RX&;MKcM;A7>`g8>eRa&Hd8y$DR{R(iMfNe-t
zAL~|{BA&WRJk+^z@4VM3!#ykV3$N@iC=AxwLqg1Loe$~BGk&)U7CiKyeGE<YLX17;
ztEnv^Q(dMoH-irV8=V)j4mpw9zf-tl+Ul)dXb&yHxLnM|xUkP?DqdV+X@0Bx!>B)S
z9rW*lP|2ejH_U&p^pHm9y@CQi29TMUB(%9x`t&Rbq<b+;>?>Wilcaz=Ik{max##&V
zd&n&%-hGP?L=JuY>gXk|-wi+T0qQI8%DVi&{jEkpD7O~%G3C%oQO621$W)ivRcdrt
z;L@e&xclg&^OO4Atem^+^>RkZYUkLG&@13Vc=<z~gigh&_<`S>GaGL>As1VLpjAFi
z!<QL7nY6=^IiF8A^USmKBr)UUcXzX;0=vRb+|qTAoW-=8-ybBLiypq)1iW8QkB*(k
zv*&oKp>$r}`Al(DN%B=XZqAg#(PUGrW{j)-GR}0O0XG}k+S!~48d0DTI~*{O3Qk<I
zWhHKaY|oJM&Q^ljOz2e4{TFN6UW=@cKEOc+?{-%2{YjpVI1&EAi|NQ1p*X{@T`AEa
zsdbkpcOxI<*v**D@YPY+HyF!Z7fjQ;2CRdR!yFZ?9gj7V%dQ}K(VesQ?QK>hHICia
zURh0+++c2lLH9RGQG^0T8i`D}B!-$j-<mzccA>4U4lUnhdKurv7yLGfG^fe=`82-n
zO6&>|=y`zXbQ{pG!n<Cld#Z9jr1!9YML5R>JFTpfSwF`sgV7N81llSY16C5|X)R}v
z^`*0Kh^Mo=kzRjmXKz8YOU8>&5`9SRNi+V$FXnuv+b|k^U5<8k03NgVaE04X?DSpT
z7Z2mv=IUQ;q-~OKU?r_qTvP&b#@V^oWyKK?-d1lC7OsLT*%G<T*$P%$NM(rYRR#^0
z3Ty?|-n*}1A@^!(==@g*M2fY1asc4L>YE#Zyqmoi*5i8(c&9)0dZm<A`$D+DJu?{v
zV?6-h`-|s|Ff*jWRx_{WtIdI|Z7=r-KRNA(;E;Eq;VE!5r0#6LK4-%F6p@9Z22Bp3
zt3Qn^2RuBJ4i>dQyhRghe%{5{_2^25F<b8HJo_Q=eG(eF=Dh2}S&I>*(a%;nqp@Pd
zqP{=@Mz!r>l)605XGbbw6=&roj^Yj<s_scu_IGoy`x;}Y+SWn0l0Gp|7B^0>q`Y_y
zjg?OR&;H6w9yF1-9)7zE4*_W(v``>59jP{rsC|9$2)RM!J*rlC)0>8I^64R@N5_**
zdl_5{^Q~jw4yXsl15>)D7)G7_t|TXC`1K#e!wB!2?z#^Mw!8ue(U%jv{0#S&ldcqE
z%(<+7oczQnZ&K{eGLh!J+wWSWH}#TSKv}z-5bnF^#&zy^M$01Y4r}s{Dfc<cZR%b+
zc)oi*!S0D7j`S29Zx>fzNm!srXP*-4cSD1oY9dxmbr8)hZ1iG%@>lzzNZh}C(SECi
z(4ZEO6JtB?k@rvrS5+b5cc<`*;f2otudqEo8#RI|N15ORu87!7(6?XeXhBFNjt=_i
zl!P>7{50NaRR7M;zLZiGLR6@w+GB2tG@ooKR+kbI5;9N=RSEO2gNFSg<Lf++q}TE&
z%BAk(I9yhEl}dct;$lh67qeXC<4#Dh)i0eNjlV!|{+m&Br4tLl#s|`eT9IW|yR16^
zdgR-~sZ`a(Z9uVASc**hDk1h`=+;&Z3_04p`l_2eblc{)7!tFtvgGz2Dc;;K!$qaf
zsBc&KXcqnt4{oh2-`%~^5uWoH2e&CHZO}RuV+GYDw1%X98=m<Ibu}AWoLXA^l5@SV
zy4|uSiO0L;%s0*~A%7xaRd0Bg4QP@TNqOW^5O<-TJGPFBapG-@jO?vffb;mn`Wq2p
zY5tyrgl_ksI%j``*c_+968~JoOV^{xYcs?(=TFu*HVdNVv+M&Ok0;GEOc>zvL^)lX
zc>Jd1P~TxEEsTxz9NM|emv?_=^Xd@M?S}Vz#<q;^m~Hu;ucw|oyiB@bX}e}%M%<<q
zmY`1<`SlumS(0AWapFU6Ax611I*9{M)r)`$w!YKh_pd^V_Sr^%|A@;i8nnk%{cX_1
zuZ;+W&&pMjS~(!0osaqu^JFjeX;h*4={d))c5SHJBVf8r3eu0bcK;@9dAjrXVju27
za#K-#3{V5%Mo-9y-(5}>cnp0IJL~$Iv&Tbcg<wH@szo(xN=F|4GImOzKmn2bT!L|O
zTAj-cXbKimxd9k-PfCGPXxKX$f{EY!K4qcm8Tw1T`MmY`tmX36t?h|1tRggWh7Z&3
zwEO*7S5YKz=f50C-OhOY0wqpN$3E7<K}YsOX*`nnUps{4Q3dr4C?l=qP^zD1n=8+@
z3qu_mEZGI{{TMXp$koyN<sBcTQoKe`XJVuZ4hTjEQnNR;xw$?*d;ZWG`evXJVS7Y#
zfN=YiwNleYVL_tPSMx|gvF8N9Y!>=1VRt&G(0HJc0F8(hF6Y(Y6ci-XwNs$slPJe&
z;In=h3T;fYE})6y*(eQC!32B@?UUSXay<_@XAQpw{+fm9dU_A(Lfa30LN+#gF5i*A
z37KuJqzfA}Hp5aRQX>)<k>5yP;B);Oin+=r31`e~Zf4T%4+Ewn^|D`mFJ&sRmjc9_
zK8W@IxiNNXRn~eo@%BHOmZTAaQr3+!FboTiyo@JXR;_0^C|M!Je(6a}$Lg^d@?Usq
z>3V?y4XAcRbso2LWd%0bBaLD&c+L4V-(yex@KLN6Z}6=L)9tjEoR3OhA7Ne7f!s7T
z9gKmdDw24QYH5duxr_9rpdH?!z~{&0eARDGb2dS@Y?2mk4WJsG{2!rRE<}^2a34rz
zQdGdaXbW-d&nwVMr#3?i)^oE~`wUIbtlf4B$M8@bD3D!`smrE&uBA>a2!E|{_qp+V
z#EZqZrPrmBVuVWb82pI6Rn-GYcNK@F<L2;~{@ZqGC%+52<@xdTm@LKFJvpNX%y{!%
z@_rSjw%<O?8CX4xbfyR7s`ppxcMrczI2G68pCjufo#j@33Cvw+V%q(J9#R3W6rWMq
z=~9e=dbo_ehgo3(S+R%J2z3_E(XmP@qu5FZMrReA9eMce;V^7igN`Fv>)VpDim|%t
zy~xoz(u%p(%-?O-?+h#<7rC0cQUC2nL|Xr=G$o2Xme=V+Zh@H#_ggNaqGb)a=>~iW
z$fM82yy&+!BPN$~|8T6ukFv}R0}o4KuZEZekWT;yVY821rgH3QbR3YT&eC7#{MF(s
zJ(eqprCa8AA;l_EIt7OuY)$R&IzMqi7ru5{8=DL1G|QHGdQQJKBzeCj&A@v7W$EU!
z#8k3SG}1ja02rO~*7{%8b^tdde|7;;n+GN#3&QOc$RYc^Kn?pLkS#Zh!8(Y1sJqPL
z^|t@(0&js)-pl(furby9rgy!G_-K_>h1)Usk<T=%$ZQunC?PBhR`z%p_5k5dJm?D~
z4a$C^25SaEQUkbPQu=j&WvIi^Wzw~WA^P8v?BTHFvXe#8vUTf`!uB~}%G~)!lNTu4
z-1=`uo}>x7NkW>OiNP-+i)T+k8Yq7mq=4~rbA)0IU!2|;;pqwsGH2@@!Le;H4lCD0
zNMiz(SE0c9_?>4;Q><0lsgz12-WSkWYQm_rmWAMl^{)t^8)j_EIJ}T1ze;lb9U5VB
zpV=GvQ-Z$a<6Nt%QGPBRm+p(OHk(#M$eZys>=OXpLR_Xi<NUhlP5gebULKs6eM5g@
z<;=onWnFaWM(_(rxbRrYIM$Vw(@?>EjW+OwS_y7fv>WuXXROJ!%&#^Oy*r{Wf#%GA
zp*u+tUHs`fB6Iqk;tvA#^%{Q=UuP4LA3rL;S-B2t@}a32Ip?~?<_q*cAY0Bz9)6jS
zYTaJeMBE)NWRh?jm3r}NBCkJi!Jl{rN=;k}Z%$M!Pmcv>A-eYOQ~!dfkdIZRM&bYJ
z%kaPYdS_(Tx;J-bt~+jEMn1(rALULS@hw3k2O}fUKr^zx_Nv>O1!~qjN~Nadj0e0-
z3|c30g10JJ;o)AWN?-K3g(okYA;4=q@=wvQyKDomX=@FqaiU##)k0dPqoRR^$G|i+
z3r#wEEaGYg?Po~n<D7Z2wElY)l@<%dur|KC>>t4bZovs7pps<9Cs9mH(reEQfEIap
z0;4Tz`bg*sPT}+eXTfa$LgmKz<YjS{T`>#`U#OJ?+^OEh%{_iP(5XcSG8ac)jZP@g
zv3FKrVvxiprH62A+DVut9lAy~ouYpkVYJzGZ%}L~v6W{w7`A+}Oauz)d!CQkC+;G>
z;eiP8zz@G=@YBPCC1$3DT;4=Q$B9L#KIQBAut(MuF3eHbHSJ7))8amUM%Xz&b-`D>
zB&#F~EXaq3JUQm?k!lR&A6q?m0q=h|p>@+lRZM`8BSX~obq~a&T&i?+`0U8{l5e7C
zs5Ic-U(Lu@I$-@x5UyNTX=d6(HTTuw4M$~4^oA58p_dbSU9&x`ckKN7Fx)`7Ddqdv
znbBd7hmK*zKAtiJ{NetIFWUB2{sbBZYM8BQ<-GUiRI6+Z$Mw2jMnWPlx*M}xDS!z^
z6U951rn$Q}dV4^jjyaMEPY+xCZCZK1q1;SdeRESmR;HuemL&>ca;<gkA23_~n#47x
zFyj;ELase$52Z@n7*=N_Wp$~(uss^V(Gtxy*-7bhpn2VU4fz_d9}FNkbv?m*YnQLw
zhI=QU%su=xW}XIN<w-MNnR3k{a7P%Oe&55+^lJXAtEsuKTFMoVpE3$C5>5aK5S&&9
z<3KkExoO*gg9bKk^gKO~cx9H_(Im!UQ@iL)G)T!RP1;EmAg4O(U_iZk8#~u0F)kzN
zQxZ!6X4_u)h7I-e=g-ZdALas0ryBEG?fiEWG3P^a*`#?MNmHKk8}$rj&W&rmOQa0T
zIOnLoj=<LidU14nOi%A8-Owx2jUk)1K}Nz{^MjuTMT>Dfeb<pLdcs6Zad*k&rgk0x
zmGxAp^yK%DhyY9pm_y+7D4bH(*1$lFkl__X|K%(c$DfGt{!>xxUdn9LqFTVWzEJVs
zbSh?H&D?X}&XUC6Crz0q@R=8avtyb}2{}el!)M)r=aB4<Fuci?zeY(&q_)O}PaOeZ
z8CdS5fewl@V>Uw&Lanv!dl!?JIY9zON8mZ2-8s5Zs@p7!Ub>lEd2oBPW3ploSMqs{
z@yz^UnMIYa=iXTEH4+p`F+ClW(M~8vv=Hrzf%SF2XIzNl@}BPv8Ui8p;(J^B1bw6=
zOtHN^Lg2(@P5$M?mdDEb86XhEiYa?Pf^J;i9_Pj<=c}bi0UmUtt|w@qJdVvmAhx;{
z+WeD*p1>&YS+7UJ+`B&(@8ndKF$pqioL@pMVJz}fc^F!Z``JGeUZ0h#7?&KEwPyaw
zJYO+|i-qNU(y*gVH8fR+-F4u2<fi?g+I}me-`+m52T(YhW2j2XZMzCh$-DVT!xbph
zgXTQAhKs1Rjlw;o^>^Dz9bB-YP>_MUySmztvI=P>+JJ*bPM|jl|83TMs!tK<(`B~M
zLPHV`VDKr^AcB6<qrlb@bLZIyVRAG=uV*Y7A`~8)&cm%|kN&;Bhr_4EbDu^1k|8vT
zPpF+gd#*Voq91PR)I2VFQr`X!b}_FtLFP|P_VcMbT8=5Z&R&$fCEPg#Em&Gg0*ov8
zl=XjoEI;qjOoW9I`4Qb1${j4yde2sL!t{^kxh#p<;7&-16|jSIIe#tHQ{wy&9&Y?&
z6QbC&-p<ndf^-ex$<(n{RnvxB#DAI1hAJ~G)Z2~ZmXml9_j?-z8f^3-6NSNo)@-h`
zSEg)j&{b4qw6hbuuN?#tQcLg^V1h)L8H(vYMQbT=HC1UGdx9o7a>m-8hOHN>{+5%T
zId=h4MF%5Ln9qcvq4=>Yn|qNHR&asQlONf{0*J?li(;36ZWje23F4&qjC>NxuCdZN
z;}rQcdhN#J<i~|iPNg<Z0$p@cL#w41c3erfpXnMJOqOjgqdI4e60QA0W?Rwpz>%v_
z7wv3IVqx&Mpl~zCBfz`@;@-)6j7Y#fe+OiolH_XHxGnRhHTp+K_2~6Di$Xq4(Nwr9
zyTCDu{vM#jOjO|3OL1u<U3jp1SefFgxpFKrg@d;64_a*}RWe)e^V5Abe<aeP0Rjr8
zh={R;r+cE}L41ST*P>KZCjV%Uk@Ii0Hz8}WNi#EF`GZNUKhmR-gtj>`@xZ9R7{ehT
za&NeS+z&r!YB_M4w->ObwiZ|7&1Ik-7Y<jhjiqF)$%su-)+L@>>WPd0ot7~-{|*Zk
z6MoM%EG~V|T~R&2dl4X!LR7$<1Fb>a;^@eRQSnD`t>c}On}dyyirPDy%1tLY;>Fwj
z_)hvm0fWT{^6zX4a4_0I7X#8+lYrj2QtgrcS~Z@xXV02@t*RelpzgG4&O0Z01*^4x
zz}KSJ(v=%~8tbDF!?5QjJ+@}$fLRwk%(a1V+_f5T;oj@!>z4+Iay%Y$epN)*29>qU
zH+9o_IZ;QK)YXJgEZ=8&5z1XrTl5C~cmz)F)xT(J<umF>H~BGmYhY7Q1a=Z*jhtqL
z<6wv`gd^AWD`izx{nzAT{Nv!A;n{2LIMM81m|+f7u)&aH{se8@Ba4|*dLp3=@uhF7
zCT$3Jfea=)lYnrTahVWSQw*?36tDb%T(-=|w=ec>9A42C><P8z07)lyoXJZUw$L15
zyMyXk7xb?ZoVq<jBm$OWtJu%&+zIm>dH{s_d_@Kat`(vwa*efyqzX8(s6M&r)msoS
zue@e`5;JPw-!D1&mFgaF^pBk|r02SP>|*v}dC^gGg^BmR8_LOv07%D%>P&!$Am{i<
zS4ereX*@Vq9`U&WU>cO{0<zJNx*ct#-p5<T?M0wsd6K;l6zynmQoxNqnEq=N{5!c|
z;>~swAf&lo90k?-=3r%RHj~DV>eWP7UNtnZxxwMlw=n8OQ{dcpg)>5ipqZQbpoV6i
zm-sq9(|`lS*<61K{qB--G-)hUgrgt-C{hYE(vWG|af*V1U;3_C;JA$|pIls8nBD*H
z1;8vnRUS0JoXxUG{fso~uQ^gj2@Rzn&b+wd8TUb4))s9X2Dg``*^L^_@rl&7a<$Se
zbnte&7>vfkvQPU!l387T7o#m9bIf<CCk$qxhx5&@y?ATJ`NYCm9Gtto7O<mFG5uh;
z?<Y0!y-g=PQc6}%nCtXvQn^qzzxO+xBk#A=$m*^s{gz*+A-;Oog&X1^HxgPaX?{+a
zgp7aby36=2141NHByaPO!!hV+n>fwv?=%&_fn0dgA)Iqv?30<~cf{d*9x3G#qz@;?
zC4+E<OATAk2mz$~<kGwWi;sp~Jjd_UfP=X<+MimMo9QFByKw9!0nOjZ|5Zh8)85<-
zCC+1U%%t^0#@g$n1ZaN%d`G(Qt@6J@#Q$F*Ud`S{92}U<hOyX!aJRF%91!`o7&HqH
z-6>(04q0xd$$>EeirzDpT)zM>#LoDeN&Bfo(dVlrzICNf@zc{Wm>TS~Tn1~G&~z)@
zm&Sh5^w=?mCKYKpdQy!i{7<dJ&gvG45hvmvPo|Esmd+N1_8XA~c4DED3;A8Bpi|lY
z2O~^?X99#kF$I)Z>)brY^-XIm+&Gs#kj#>Pj9UKHepju0?#?(FER>O1E`6#Y)-3Dz
zl7`mr;pDrzcK>ea_E?|*0~T)$aZ+mct_&tfn{A*DtGb5UFI57x5BTBpg4oVJpgly)
zO9{6K^E|;mCA8m|X>Po^JhTlx)ObcHt?=l6+_?5npz>kI50$FkGD0_Ibj{MqG6-Mf
z0oG74Vdda`xq0asWypBwGC|>61iFl%k7@fUiJqat1wAjt%~{;(Wo0i()(!}i+<NlX
zzq^1#aHT6idIQlu;2bAF9b<FJg&sC#W<5=U^K{+NTFrc-XiLrfy?uFutn193d;QFy
zMV%W4Twj2fyU0XELa0ZH6Yoby{89q>hz{&S#Sfb!f<B3>Fi_3EV>cTG1%LY834r(K
z42+Fw`=!a_@aF17zM;$Lx8+wGnt}y5$1O`<e`gIXX$8bTn0F1eu^sBYI=|Krbys};
zOw?Jj{0wcpUumQhT$5SjWLqcf36)^I?O|LGXt$_Dj-A`$OhcY|viRvzwVxB^kxkYv
z=XxKA+>%;_7h*944?M!jPL<#@LjCiG%k*NRISAmM5^|=S_O_p`RMj>wD|gu(&bnYz
zaU*_mYOLUO9^G~1T+!YY>HDuT_5OzxBmW^q9&z3ST2to-Z>ZPC#!ga*Fwt#FIYTNN
zySVwuSr0QOg#eNL*nW>H1rsjV!$5lUD4Ok-ZN@uqJststlNhS3xnEE?8OlDh>7|4$
z0k?_&!M<7FH&yxUriEvq^H)VUQjJP>v8BCpl~ud^kpXGbMkjk2ln53E#v;vX1kmT%
zk6BP&fgPk}&yCDhxm|yM<RUUL?DVp~ujcfj_0YhoO!EjQMyjZB-g~4$h6vvy8~KBr
zgaGwoyEgkwKnM#1#(Z@T*f%H9?%k8hqV~7`j2OF|nso8;<PtZ&3zl_%l=9fRopsK_
zT4<juDMI>+uRs1H&4wO3*&W4QDp?{?<$TTB2I~X6_gm&U+DjkhiUE1EJFfRmluwx%
zV98VwNCh>ws|QwWKW{R+ey_BUNzc~t)yFBt1%Opo_M5_cahO6M=Ayy<M%7fJlQ@Bp
zCPrNQa|b6L01E@KjuP{Qzj;i5PJZ<cTbM{kR#cpoRhCNJ)aM<DnSHSQ_rnwwV5SiC
zu%SO1Ee6~d#_h-cwYJ}M8grJ4$ClJPv51MaWuC2{B)UVy6xRk62;an-llc!Dtb)9j
znk<H25QudxUR10?p5)~yV4^E*`hT`KMH0w5m!opwRPS9}GCqHzN^q#L=<R;w16EIw
zv59{7<2A+~vFhM<sG&iSlrV8F;^#qxn&_@!2oHt}Uj5^1>f7l{R;q_me%_W$(4(U9
zsi(k|#hDCoz%%_w#spVAmNywbcR%I{SCRWrt2XE-o17?svwyvvVE|8}=$Oa3cIanv
zG$91@Kc0n{gH+w>X-U|9GHzcre??qG^vjcCy~lWC@D5Yfm<@4N!F@i#NyFU208~u^
z=#IjGb|%~!W+B_hGdr|eNRz{Z0pR_adbirVXPAYZJidunS+ERLF9qJe-h2Rm6x=2T
zrAp+^ojolwJFCUg9@BU5<Dma?GH20FCMdMvRbXW`uPqkndxvb_2Hkh;W!AjgPQkUn
z2)O|Zn>l#C;9?ez+s|Sh2Vru6leb#im_zg?3_W<>JK&3g%VJs+QzoI$+2$D4M0DSl
z^I6}sYBLGYP9n635D3#6bNs3Q$%7M{GHhl4REQoiG7MceRjJo&?oK!~oAnc_Clvvb
zK7aiRsfqI<d(pXyi46LEvxD3X4*13h@@cghLD^lZTy3hYT?apfYygB=RNCTcoCAt`
zkR%m#eU~R}_i0m<#?%>pzgQSy<nrI?rs$IWY>FlQr~887UcDriG&W1@2hh7ks*tLm
z9DU{<_`FaHXu!xGX3~(^d`gNN9{#G&iYNs25Li>5klBH&48Wf@72l+*6xSFeyq{FF
zxqPGK#qp@GjlX*Lk<2up(*yFBKv$TWyq)#Tbo)f=JQkX*&Jr|qX;Aq4Q1Gya6eS}O
zBHO(!D0shL1n4gSqzDv3jrM1b8lC^u{7N<@%VRloO~%MUckl}99yVG~N13~F#_~1W
zpqKCz?eFgz^&GHc*rfF%GJrP#dPBN(cQg<3dAUiv^~C0%pU<?I5>1$S^KO`F$&+eF
ziUi#99e9e9_!z#RM$iayJOpzErG&-4Umbb}MeMbSuKb68P&b&095ZneEv<?P3nt;C
zUU8<3?4bL?%K-u!9&~@Gw}kv>ib{40V5WowS%rRYbzH@qsjd=nOw!a;yj2zF-3qTc
zwL~^*R^q3EJ&4!I^s$Z8_#>*lE>jb~|FLF5vHjY~r{im}FtMW-bz?%<BdY^}PCp9Q
z$IfY+)c@>bpN#(5$H2;{{IQQ(a9WtF-E3I~LEM+6+2sN)?Tq;uPh+uGx!+GCZ2_yr
zdw<o6$g9M=U&8l16yy=(ocyf;V{(9MduV}U8?{{yNT#$ee0QH2$bYVR@`ERYB(5lQ
zVel}epk=K>P_ei@j12Usi@6Gs^`XRMR##p>aO;-KaTi^qTKTv?baN2{w4vhZ331d6
z%$8*4Vy&N>(=8)8Cl&Pfzw}yL=2DBn@^ykOf0|9Vj)Pvaqc8(2?}nVkRbT`w`jv5n
z!`OLI%C$&$p)E6AvUoUX<mkD6*4M<d&;`Fv@tTFS^Z&f582uItE<mTjz1`^ZZ?Bct
z=?#<I*qPJ^PUEO$bmw6li!gSRU;_Psj5RTB?#yD&OD?c^n4_}ZS`}3>9bjXZMM3iI
z$IMH2pcr-t%H$$TO~J~c7TM>VEVla@g6~npTaUce{D?m*23&6k9n5KbM<JZL^uWU}
zwCE<y2yL^VUS3W<K#2T`VK6eVrn!yYYdP*yXs_}P{cDXZ^J*ACa;h>@CTueEuW!{p
zoRiG7T2iGH9`dQ8-$74tVXrWHz%wH5OltsL3iR9R3+tTpJHH`n)V-KALc|h~<M*{8
z9r>OOcBBNiUbTmEhxS=R9EvzR>}yx_XnGY4;-KS9W&wh^wmva!&LXULg8heD?P1`p
zL+Bx__i;YY5fVpfAp&kZM<#E__6b>mY~Ck`>50CY!g_!Q)11K^geZ$siAat<LN`1z
zv>xB=XN4IS_`a(+j3kI7x2CY#FsRL&_wkiNSwq5vO|=wkMK}+CvOx|=91{+mM?v$y
zIbV=!!2z2xP>xzcHO#`&t!Mis`pY?#f+!qrT?%Y(&G)T#oY*xsCW^`w5&wZp=>NcF
z=rlD&kP~kbEoaM$NRnJH-Qjbw8uZ)9;zVolRMEl;A|i19YQlu6k^%zxq=H3#OaAeu
z+jhnk?AXxte%~CLa^bvBT5@}Zm)QyfV0-Nw5`5hzfqReG>ptc<*|c8(FpCCXp(>CP
z;)91z!uf1V%0GFjFr}}2z`FMcO&BSHodC5(Ibfupg^J~^dJ+OEIal3R0ohwZ(3cr@
zWu{j8`-$Iu+q6YsUKZE3LyI`o)T|w$QY-4fE$7acBMMT{f1LbcCr?xyF3<`3p4smF
zWHtAr2~Flsyt65j9@~gb=_dY1(l@W~sWIqELy`s69&BH!WV5jyp8he}b2V|3?9=ge
zfRPn-vEyOnq|-m^Z4h7dFQ9t-8|T8Lh<D^6;5b`xqvvs;Pz!$Bw~MaY*#i4`GD8p>
z7&8GJ&)Q0kEwE^gcD0jXu6X_$vAFiOdw9a{#nLbMuXfkoqYU^@oX7c~!9}SdtW{aO
z80#rI@x^@e&-=x|;N<_8>#<Al+m_j*{la+l{xiaX+P?P9HWFyYyU4&VO{*!Rd%v5|
zZo&Yq!Yo-eQ#Mc-Us_UBVu%bgG2Z)D^o*a(>X5YbY`d;8j{PeBZk?1p61Tp=POzQw
ziDF?k;|PnBq1PwTaR-&glf3?$?d{M^Yts^>(%MRwiyZC4YZrnPd}F5^cf%V06za#i
z7Bj}$+3}tBGK!-CF)Q_8e9V<Ztu_JqjRVsYB~X^ipguHtKEUC4CV&4?$5*H)+-nu=
z3lkp}Bcb0qOWCs>icU5r$_)`B(aYjV0??)WQq!*iPp);Y2m`gC<3bO?qNwYN!3_}^
zS`N>ya}Es5K$k0Gnv-)pWI#ulO_ozw0=xlaaYP|9*2pKE?fY!)QL+QiI-)b2?hEIo
zT4tKcSB(IP$&>h>x5<(=|3@~VDic}rE+j8@1<SkWK{KS_5;VOvJr0WuOo_!zfZnYC
zu44puRC1%2c-g825uUtmdIs(-a`f4tIQ{}LXRyf8^0+<)6mTMpGrtfz<!Hf!$Dn`k
zw*d3E^(%oeC0xR=(^qL_OW&9_cWMX6uE;=vU7{o;CREwf+3kfdLHoP)FZJmF&qE=N
z_)C_mf9KK3b%hns6w?UA8x$mu=4U*T<y=>Gn?y%Uh>#}SI<i$N&+ce+*U}2<J==Ed
zY0MPEE~t}D7pS{*QU`iIY~Y)DUx;Qoc88wc#gS92Qp3Bs{nK7ZKZ38lPXg;2k#+nL
z*q=In54Ut36o);7>sL$IIG5DC|BO{ZrL~+N<Gh3Vg!^Ko19>i-#mtGB{qpeF&p)Ev
zLNAwb){}0Exily2%D~us4oAoaCk@4IT$-}DNxH$_iy1F!CT?>csn6*-(<_IM7$5_`
z8>lZfSWDV55*Fyh=WPb57vO2woUNmX?3`~oJ|updPf$Rcf-!q**IOj@L>ALsSAPQj
z?92Ct<~pn~Ak23HQ6hwW6C@Vf*^8#geFC|uL!UP@?4Ns`uH7ZyJ2Pd9{y<MBZmsnv
zaMbWaF(5&c8|7v(8_4#Zd6EQZ+k-{0hd=@q;7cF}B2;bO)ZSY{POaz{_VM0XAxU!&
zzcp1uf{og59Ckb)oOJN<*metRz5qf0()Ggm>^K2YZ-YS~&jmz#g@ztbK}E}Y?m@R|
zOFjt(PH61k_~{>Y?Rk1G8j~(xXg=7_Jmxu=Ef0c*W1yihJR}0SBv86YM}7^H{S$n$
zBK}8BrdigXnQZnN(l)k+->tyizu%zIykBhwoV%%1Ik0%z(tzU2n3f#t#FJO?0fxHm
zeXlG3#tv{4f5zA+hxw#YWP74POBd@tr`LGr&bM+2seC;T_RIURDUZE>Qb~bHJR5)8
zp}tNQ(3Q>|ghhY)ZYw_867YhQG)r7S_FEKGh;>sLbme6+1kAR*;#nw;=HjeznY-aJ
zWDsM%*wI5l8*-mHb?}}UN(AimS9YcC|3qyed$-MEa5v-xYqh1?&`qm$4$<R8?(gxl
zaI}FzC?CWpNp7vubA9otSP2m20Fy<H-W{Bmcqt~9h^&>p848o}tE9K~z^9p}@rV3f
z+?ggmK4GSZZ)0YJ3)jp0a}OI?N_P_$M)Dpc^4&O47*%HtYafzwti!nw!&^MB*<5>>
z>Dr^{N=w?Dax$+-jkMjIj=za?n6xJ;kLv|sKz-ylB+43^D6Z7md(J)*3uoKnnE?o0
zMntID1r_avr~%v3#E&*Rc^gMmo<in?-1Bk8{q7skQu;lPuI|4mt9nbvjZ=m^o&Pk6
zu?_gV+qjtTs(o4Juu`niexqqZ$(H@G=385p_}wb;pxNspLV~K)$Y+yvwzYmzvzBKX
zecnQR^zb{dA=B;rQEV7uD_(HbYrAD70!GQ)YK%ZQ1!p?)j|n!;vip6WEbA)#X+Rl4
zv0H!znJ3HV!wM#b%%_i&$4CDRLbH|J-qv(|4$?)%xtG)VVGAw$r*T!f#Y}6)``lz?
zs>#DODAdfWO4JfqfnTWu|NNZoDFOVMI5-ObS6v@{2rX^{(wjZrLWZ~z=G7Rr)D3LU
z+r7-Mr1Dhn45Zk9nz(A0NbkI#h5I0~KnK#SD>8uv9Q5!_XsWi?@oSTB-OOtpZgJt^
z5^m?%vtFP5{f3{Arn;X$O26Kg{3$!@(OZTAAkm+g+FeItq^}ri3dxnB$J_$2M)dE=
zC@6ld-Lcfv3QOyrA1<Z<LX4glae)F$<R4%NBe7$lnkPY=gHK2x{9S4vK$9CBT|#M5
z#R%lP!lcFXrTpV1$(x;R^B*Rp^*f1x)>~m3pGlvydPD!HYB#x4&5O_er(wGdHIMd>
z6b?gpTD$j54Bb5%qrW*sop%IT(f>8>>kSM*5J95-_B)!D)X!U-ZkS*Z64nn33u?eA
zi<DCDxTZm~*g4;8Y<S=E(hDyJ_hs_NxoV4RWu>7(okXX$c&{Z^vmty#e7ApQ+-p;P
z92V*9CroUAB_+lYbIxk^2WLR@qO_{>bJR@^YjytB*Eg5011F55XWmeCHFo*}@w7>2
zN@1$BnaA7PyC2iBrzH7@1OX!fLrL}_m7#AZuhF=50Y35&ff@8IjKQnR?XgJ)1{6B*
z0iLp=-~wP(v#hu3k?_xaQqt6La9+VJ556)BLll_b`sC3%)n7)R-2MEU@yXEAJNoHq
zDnHUCMaynKphmf5&@4liOL?J61V7-N+aIA4TfY2Z(N~3JN@%%+X`FIP$3g@cCplvn
zCCgKG)^PB+J+qCrWavK>haU}EX!yy~e~7V^fXdzVWgQqx?e3qh)DHgz5}LD;nX&Cn
zDOjZ~6-Q+Z45%P->_Cp~(Kb+c&x2Q>T9D?-0ew=z0E0E6t*YsSbmZl%u}Dp*ZA{ZG
zzf=Iy)@TiB{!mii;7h4t!}x!(1dmdE_dh5DcCCbn*$<p;xPt5b!+g<95f<uSKBBp=
z_3K(o<(-_Z2C%h_i>SGXP*A|E#5Cs{x*_VUYnIgn19QGnv(f47gXo`_fn)be-QKky
zM};T#<KwNx2u)?{&Kli5CQa!j`!Yq1E!4Mq(}=xqlXf!$yh$H1d?P$;CmGnH5N2xk
z_5{^3nwCHj^%a3uf!c;~fS^y&$_T|k@5+1d*Ka}~!!Zi?0`l5SVX}j44aeM2i&m4h
zcA7&7*XuuIs*z6jW#61zHaq!Q^;ZIhvmZcTKXqDgQz5pV_vGzcTpagsI-!SgF;bH|
zIu^orh{ja0G1o+%Zr#wf<tZ(=^3n)qe*-CiKn9@UnV{zw;u!tKmclj9e{1>_l;#XN
zV#N1)BfyQa5j@Vd^jVMvXrb=ca(eEd5>iN#+*?v|!>(SM^LkxRxvB&rTPxcxR<g2v
z(~bn=(+mxR0gr!DJ|$o-BnUu|6fG*ba9vdNU$pDuYw+vzc~=}B!AdU@`yDj5vZQ;?
zpT1{hu*M%_jQ}ZlpI%>3*V=lDH~0zJ+8OrNP%y0UL<|qb$22g0nC}a6IM{d~?@Lwf
zGI(^9an|O>EVRVM!Pk(&j5L`Y1%+olcX(W7<4U36{M$TqagMWJ(=RhG?m8RTXVf?y
z%Pc^O-K2VoO}VevcP5SvdYVEzfpm3__~z&lD3?_GS+PyHwFw@y{<l{i{?8Olk?iEX
z`d|sji(8LgDMNA_kp~1M!%O2wv~L9IrdXjZFS<od{r(%`Jf-qn85bF%VdfEvuWghu
zw>Rz;ACrq&Iein=m#NaXSpy8Ct_lGaTEVtdYf*jO(E~lK<Pyk!$)DbIv_9x&O_#L|
z?s(N+oUi4!*lz$P3IHTM+uqUmP92{E<R1U#SAZTXr~}!Ruf+sQ#0|$@^L>;(_*2`-
zwR05;_1exEM7(wL{L%1Bab?Dcr0_i>r=!>cptgzYxBi)Y(a@I~?C$@*S`tHdNEf!h
z3y3*ampKzwJQdHK0dA;(Eyz&?OCnPl4}W1n&m4@n4KVm>m7E+SLcy&%WHGyL1*fau
z0VnWEWr_$o>j01il-lVZtKrJBHE2li(F~cY=~WaB@1ITE!3<5$6=4Si4iongmzmq3
zu>#MjID=+x_c?4<zzs_qxnax2;yz3{Ax)LFSb-gtI00}&7hl(HlbV?!j<N2~4vrrp
znLMjspmR>Yib!4!#y=vs=?R2swY~P5j!zM-9lmh8`G4W4L(O2BE1eHgUIHKU#*r+M
zeW6X&54u`gAXGXrRVQw`{$gny{;-iP?9E{9`>O<Ht1b8G>(03XzJE?RYm7zCm@+OH
zcFqGcw5p%5HGNlVww_{8P4&zB-l9PKP%O@BMoOQYnOzCNiG*MU+7`Dtb+X+8an`d1
zmytR2s;WhmlbRLMB)>+Tv0I&f^44MtZnfJ9W4&!<p|cCEg5+>0>0EN2?6U6968h~Q
zUcOya{iDx1p8zMHafPvmZ5}SE$jVRD&LTXjrGxa@`sQ*WN|;_CC6a}0e%5=I^~k!R
zS-qPX`@K>dpmK-!&bP)IvlWm}3@wh;#et~ipZgD4sBhdF(6n4Yf3?~i7G~OgN3OTw
zr*I!xEN_@Pj%~U3YhD_ij6om<zw=EYkkM(c4OnYKLiqm|9~Jo#KS3NL;0CATPvWgj
z*H+*p_dq2y!(E6V<JR`c>*E>)n_yE^lKM$^05Ycs<&kCL6cC8xr{@|x54j*3@o%*y
zB^tGz@YDBT6&zuYaubo~j<XjE>)zJM3VC|AD6~0Ponp+blPc}kLV&0oRZ*?aPY(`S
zzR!SXqFFqlX*$$HCvty%x)jIZA8m0O4|-<XwC`sA{?Als>Q$@amXtJlN9gTRV<9x7
zYUG*!+RB*y8pKtAlxEl#T5sB>2K>i?^-y57NQh$+tgY;S81b=MSsk0Z$slMDY*$YG
z+{3lS)Sa%4IzH0U822;J^W6`o&9dF+<m6UlZD*Q)t%ujei-O@yc_f9|r+oo1XasWQ
zyw2h96+H_X^_%SjX2YPx_Eto1XAvo?x-74ta!o7Z0&Z^L5tAc1`MsJy{oVe_2Ofi9
zB7&dX5zGL4JON}xi$sx#tax<6c^)91EgdK5M43GqqxlWaW0>F?N;E=i4!LHY!hdL1
zAoVTh0u3=|czvky5Fx^O#4fh%%3I!FA9`nApl@X@XEWA)r*j$br1q<vxV&YT@SLn_
z$N$N+-e>R0++o}j89fU;eMICDep}q}4Fgicd&%Gx@)jNfsKb0Z^_Gf+;5=zF8VC8-
z_!Tu~DWZF1(D;W9gd?&$;>|-Reu7dgz|5|!^_g?+kJ|j!(&;JQ%!JyjmG~U$*`hC%
zHH8@J&d47`f%Pp5%+p`bOfU4W^-uNRAj4HidxHgVJET9NT^<@`1Y`A0SjD*BqHhNP
zqmf?g8KVEeQyt{0H35EnWVynAuwI}DrIVuGcR7cxMK8~GkPl4dJwl#!t{}ne{-$s!
z`K~4M1JJSdx<3hq|K2HkIzQ(ZaOE34h>8!3v&n8OpNVMlz+Db)><CC10R^M7W6|@V
zZ&S6Bi$F2l&Yp2Qb5KxV?)5nX=F>Xs%nLn7hhU5zLcw_-mQNhWd!xq*%tX>sukQZk
z^blBU!QAi|yQ+<#3R}0B;mwR)mDNOd-mU7tgiZfCajgHPjLsHM0ZoJZ<?0+ny!dNV
zs13JG#{Q>@gW3aiQwC=O=Rx1sshjYVVF0l`k*K}jbr4KA^BMO1@^|h)DaEf#fQIK|
zN!5gjSRAqZ7had=clPiJ7#SvymlEn6O+KV1ubFXdEwfjWT`aS`YrOpFYbC(<uCT4D
zUUnDD&I9slKuR0vhSBQ+tx`ji=JsZ1lr%H|N2+q_TcCkDR9kF#xX95w*25|iu#qU<
zB4+|N^M!+FTz_`7<5~E#wM&X9c=>Cc?KcRpyf;68Az7i+h;T&>Z5h!@EA|25>D%@V
z$5idCh)3PGdHT<EN%(w@vsnENFbMhY@%7cTUmv1A*Q`irODE656Bf!Aj97h2&Pz?*
z@`Z5Pbq9RhIyJA$|L(QXF)WvPJY2)sweTcHEHB5jw}CIIpmQ#T^}BRQ$d3Nk5N(WG
z1q)jdgE%Clz~|xRN&NR9C*NDllR@|BCYx|30{y>4FdKh9Q(6?@6!zA5NAWqj(O&)2
z5F~@(9B^eO5=_@^XaOOdx5G;)AYE{<z@&<}lkK2HVWDIqA{ks>k{yPReoB%i=sO<!
zPWa->h~9eC(#TO6!4Y)~YLf)hm$k9NSj^?7zZstug)Ndl@3cPO37n0*BS-xJ^O5E7
zU^Ve606F6Hi=qUA60Ek=QASXcbLUVE?&b7XLXH0EG@HOqW73ySp=dWv!xQ}UGNb^m
zoVDb=@JSX4sZW>juJL>pYHeq6`%1IJlZytE^ou+H-wW{P^p4(rBx>ew1|c~Y4&Yyl
z>&cLt_Tt@vB5ls>rCzD@<k^A8JT+jm(7=W~VNEqTB6&Mdjuu9E(EAsYL)*;qy*9ik
z%&Xg*%1&5%3Ekjm;?x-ByM|4eny80NxU^#uqz2EnGF<Xf>Q<L3Y#cDjh>6XBbTko6
zSO6fJ7}W)tSHJ5qkI@-2yxQ0s_cLt0vHqf7;PsUR92C1wi!lZGffP~7Q$yirC4~!X
z3?lr8<0(I~{*vi@#M97DQ0a=Rt{r+fBa|BKM7y5G%sTIgJ^em}zBQC@LNV#Fxc3`~
zR$gZiB>hTlI`{~r=Z~Iq0sol0PZ+)d&Ol#_>>WL|BgeYwyu=(!Nw3F#The;(8)QwE
z$SLX4bc7860T|?AY%Fbr;wC<gJR;O+#JzRP0<W%rZTAPXR0~?Hc2s$qIWiF3ODDV8
zV7zu8hM_AGfB<iS7OB|xcHY!>T@J$kY*vO@COK3<R!1$eO=r4f!nk+&mIR+u&a~n|
zzj+P$&;^>=;ti~h-yq`<^^#Acj61<S8cbLpBH{AkAn%+RnP(8#^q+|g2SnQwX{oc`
zKo8=RsnjMWtu2NmmdQ&kKdC|i{gnP=vN`MA9~s(GaMY}FHGu?akg2GKGoYUUsY4Rx
zYl=@eKx{Xm?MA>KvH+rl{e#o;+n~0(_vTgCs@<c_v9??kaFBt<fsu!6(X;G7v%nFX
z*@2Ew@*}7~j;Uw=DBCUW<e<q9u++S!0HfW~KZ{7|zn9HZ#TRy~Mx^GhL)GwjYsckh
z@cewP;m@QyfTZ2Mmy<m4lLM4Gbg^*m*e|Ec7Ve5){_L&5S##b!;Qqqs;Qp9zLa^?a
zjZsY9lkmBs!J8lLOmS<!502?Pd+1WiIt<g2*YV`#2MycW5U;+oANzPEmnYeEWWV|A
zAZeF<P|U9|leBbY<D+~275(j7M~n@4;74Y@`VDJ?e?IIuAMR%)kvQGIFUr_$w$>s;
z7ihULuT6#`?ptQ7Uf4J!S;Jn~bZs$peV1FDl;M&_|Km^WvR_eHo29R&M3@ukN%Bv`
zS4fjz*cI8^(hL6iN8hmqg9%sQ6o?@=dz?~)2`-D*W_7HM;>u@JNSN>B4u0)Z@)dL6
zFR#KdD%~{ti<cv&f9!s3v+y_e@T-k<_lOBU3NEnKChzM9V2l%r11Ms6vt8)z_Usu!
z+~4Yt=WczNN}~gvtBa*mkILwNIA50{y{Tat>=BvAg4sG#8d%Hng|t8JN3$C4xNX7}
zX*BXiEQlN$7Ah>=-H3^qSh*#@pObx-=XIVANPUp@khccB7Oo)gHrB+&3zlFL2HIv^
zOs1GGFPgNlSyQQ@{Sn+Z-dMhC)P>6Xt=(IuaaUhMA6Mv@G*Bb^dXUkpM2(bot7a+2
z@z2=>>{1;FM@mq@J<Cd5zjoR@PTrGZvAuCgLi8Xv_td<HY1yB+IDU2{sO>-ooqaB3
zbz&%F$~k}IuY!I5&-4g7jh$s>D7&}`?3B7x)hHh7+BpBRe;MSgw``d?5m~%HEBm@f
zzK1jj&)IBLzCiTMdJZyJ;U71xo?@g(j94SF{Cc7-8dF|)CI7l2K(iiX5%(<4g{zh*
zm!6?F%6_5sH`RX!QAltQ5T*MHIRCdzu9sI1TnvsKAavLk=jdaML~8iG$9ICIz(bjh
zN8i0-5)nEN{Df9vpc1<iBD^C_))wKUu?_Z%kRP?)tgL^HW5PHzI7Q`=P4}`w<hMb3
zpI~<#an$!<506fdZ$6H{W&a?#F~IyQ=dp2p7?~ehTPGv*-`NeVgcnUR^FTliP#Z=y
z7AVp1R=pu}poZ0$LAlu)UrH!+9E*;$SOL<6qhiFmnw^a2ny{-4l(~?9#&gM`dtWbs
zgwK_WAB$ez>OIykRFWP@1BBVzGwr;YgM;Hm4v0nHVT;C_h@$_(Jt@dBafD=hTXypM
z2=a4DNML+~gA<iVSog6)B=4m8--#}MD|VP4&$-VR#`xF#EcTbiPyVoPTt3*j%0WHb
zp9ers&?7Pi;eFrc^icNr>4@ZS;XFl<RS{+s1)ocWT_@D>0~lOeB?t3<cKmoNFp2<$
z%+|)K>x!UH8cJZ^daDveAD_*QAuRBw=Wla}@LLVyH`(G0F-!bk^Y72$eXB(n7$w(Q
zm^M#~sz}P(R~fEySLPx#RCMVs)LuNaFPEZYZ|oh~i~e$FWjJ45y?a;L!g(Xsza!|Z
zHiX2(?ay;g@(`*~oJj}Lh{yamvAC$#!r>*~e>2VYDgfrm7)$1TXAmQtxvh!~UUK5L
z4egpCLn*bPSK}7^vrTdSgL<z<7QEraTp9Mf@DsIuBtj{nJ{tItMQso|eni)nJ`Z&r
zw{UWc76f55Ffg~Xi{^*gPx&sqg%hP!re$}@?L+-?fuG&5u7gFcakkLOk+>0kZN(#!
z5rw{wsP{W3i}@Yi2T1)p8^<?4$@?8R*?u6B8~$Q*EnHxg1hajzOY3zMU$@ocw$3oH
z+YC6|!!CVqCf=I}E?zCVrjfyZHua#tJ3H@*n=&UtbV0At!4opV)P|*Lc3@+4ZdKN-
zmC$~&H4QUS=KvzJ&reC#pBZ(e^JlvK+nN0tS3)bwWkAU@S`J=SO!sK(U9;z=W1c^c
z$zL1UQ9ujsS<HfAphsj0dArPgOqRw*o((UQ>dAi2PKP5Mb84{1M$c`k55^QL<DgZU
zZH$jjnQcJG1tjRBXR=qZ>bOTp<&s_(?z`lwc=f@dY}qPmvGZ>iKB&iz^ukOiw9t9U
zsfUO^B3)`tRc-BjJB6ai#hV`$LW{M09PR6(`Eko_J^EddK878ct+w%NYj%K5@%(l7
z{K3!me6CNz<H0S)p?6Qu-zSdvS_R$^fwLzf_t8CQvsYA0J@Jf9fY+su@h2#Gd3k~D
zEdCYSzuhETPMc{DT%sO=t9;fOe2j{qjhr^zXh9jp%EMm2QX!cZ>ffh=rTg0nf{vHD
zLAl#j#x@y1$waMLvd|gBK?ZUMZ%6IKhsGL8C)rQ#5q>5oj~xZRZ63Ef`SBu^yMCfz
z$|pvx*nEDboW1Gh4Crhl0PXM*WzL(XvtT?luxEBwJLqiep6|a?nBgbbAMbTP*0;~K
zB`jBe*0SU*7PEE`(cJyK1DR$tkm`qS4|35amru~u*lA$=^Y$zN|CUc~*g)A3%HL*r
z^_n6t^7cAc$SZT`P0`=DjWB`s#2D_bbHCCRtGD7x3KeFfQ<_Tro`Aq`_ob&B)wl2H
z)5cZ&aL(%7p=t?L>^1Gk$4S=PbFM096tXJK&#8JL)0cND(o!1eNb~NdvTuS0F+R02
zBtw~k=XZi03jIy4H0O>ZMY_OI&p!pLB9*YB?uD9WO<fM0###Gw`*G<+F9fojW6ni!
z80Kvq_Zy3|7Y!fT7%0{l$aPhVXIV@kNWR>Ey`c7qbO#Ii44)me3i(=0Ay>~V_t@fB
z+`P^zoXMs2mY4Uh9_0@IkEU-7%lm)-&Rketwy|1V#uk@t+qRdnY}?kAZ7tijUH50-
z-~I2&o^`aNuJb(KFMigHt`BSiU3bzzPoSAAiVD}3Qe1Dw0e@-D`(7omerW)k8j@uT
z&f#*>O|ys2<<rZHf9qH=vl*K|-L`g;!>$!T*@u+pq6(uyd;dg$XAt>y73o&*ys4c2
zitx3ns7GmhtyRZR`{Mv?>Ke!vGH3S>%Dhe7Mh{g8r?$m83j9l53sOaw>n@=Hjfv+=
z?ytT+AG?6J@aGyK3+C5XJnPWW&6PDeAanQtz=qKt%l5*1&t%+%k>imhezEm$W?_P`
zLisYTns8!e3>#g+ZraWjxXrsLw^cXd&?q+c`Y`){reUk+h>baZakZM+DR<vv+V~$g
zHB#-lOQY&aN^ujMAErtPrUew|p-}v7N#C+sUD{6*%WFovd6K`Tw2nx|e0pb`Yj?5$
z3)rZuedNyS{&4?3zu3&podVlRc>SHSAFbDS=v;Q)Ga@8%kmdc=`Ksx%-{56kPpw89
zZ0l^~<Sz!}(krz`@c2FbmwA64(a$JnUcCo=Csj8nr{0m_=%RSbRiVJw@YNQS3n>dT
z^Ap)r7=j#!E3#{goTA7X3=yT{&tg{P-#Rj)6nzGlI@5}jHlDskb^NzD)V%U(le>{9
zm-|~SV~39La>@p&`_V2fHNQ~a_9|(SXYucFeVxr2B?yeDN3rluqLeV&w@V0Y$J6p)
zKiUaZdhf=Qit*;zvwFVoDCsQcKq7%^eaIR1#9pU|*ZVYtW;iwX-@1{zNED4JFdL7h
z)QGhEty_r#ECW0f75h5=P{t}v40P3-UtSb~$3^`#P`Nb{3iQNx^hnD#rJZSf51>gS
zG{<^{eQa>V`#0#7GzdBHQE!5jKQA@}72^O%(_JMF(<gcqqnXlmo;te7nXq2*ikg^$
zep|O|Rd1YE1ftvbvV~3ZL-}75P1?q<{uxrdxWXUTs4(%twr1A&IyOVwqmf#n_9NCS
zUt+W@R18!8$e~wLxN;;~5BK!w;6~JLjH>Cnwi@-HaLVp*gIn3_=jGU~HB!0RPFh%Z
z`}oQyePpNVKrNEf31*Vpvm%w22`a^~)K$OjX<L3H>~ddXQ)Oz$UCDSbXuSc?x8?;q
zBw9~6EAk%k0j+h19Qx%$Q~hONjg*WVsCJ!_^jtk)K8+<ZfZ!hz+OF%6K{W$P%7Ix^
z&&F{jbuZOvurZoq<BEK9qY`s51zDo*Yf)7Q*Oj>#qjNLyWA>Q?-Dad)*9PV|bzQ63
z(yj~$4KVBJJi>al`RT-FR)<yHK6=wObtXL#@kK=4mh{U#GB^S4I@X-6_83>=*c3Z%
zHe52%0?yPtGJ(L_j&-po56#9UBCNl~+sOjt$Uf4h<(12goz87XURUGO!TUpt@d-L1
zH(P0fI^pnJ2%fu<$3?bfXx{u|8zyu?|MCKM(*D498OeUAbrY)R3h&~%H_pl4%<ajP
z&UwG{WbYvFqfUw9O;*5rWRu`UpU2bGLD0a&gvRmJ^SKrBPaiVMe_w;Jb&y?seSJtW
z^4J}Vjn?3hFinE7l&oJJ<hG1n!(mOBnS3ZVH56Q*qYgNE-NEW!_=a2X&I^}=J3h2B
z825l-13q5${+NaTmgz(E7}#LMed|WX$yJN}kgqerb=B+2Zvt+EpE<f*OYn&rMNg@~
z#6eJ|xYZeohpLHde?R-Rt;glWaqXni&7u14cx4kQd++(d!qg4Oj!-p)JQzSR$?-!l
zL<LwLlb<L;bhILBAXFd&)xTc<at8Cw)tK>MPZa&)@*+nznshVM8a9%@3s=`piiSh+
zXu~C-+_zUx^+g-ZS&207dq{=eTsqoZ>l@{PpXft}5_aFUH;Z@j3&X#;y>z#)Oe7iz
z?OIWScQHBkT`{KRO_#KZXf{qNF;w9X3yhE)iMkas-~gc~oPaM<;E?6n=51kZBo5=k
zYqh`c%HT{JlFC4H!*3Vh$tD=@2i!>Ilgmg0GB)hWrzBj|QAm4!M-evE8<>B(TZcpY
z2V25jIS%la^O<%~kfcM8*XRYS2>3!yf`2^o{@ns20aWlYFvvfT|AdJhXMYg#4gqSU
z?$sFf1n$3ugjr9yCzc=IZwU&m-hy@D$-!5fUNHZj4Qs}8j+`{#fxkAgva&0k>M4%t
zR$0*G6}N~IKfvkrvR*^Vnj<QBE@U-0dF7QqEgi5mfQvFDUUcl-0|6xW_2IbM=bg_Z
zlq`H5TvXPsy`2{&_c6AYUF(1L{bo`|bp()jgXv-i8<r#H;5INT_e;Dk>T)-J&Vn|K
zFA%t!@FjLj(_IZ^_*II~E4~uN%qA0m-rCZwJL>$_3HOIWB;T$oV6H&J)pYX2nsO`h
zuxeb2QP)t8Gie|9a<O{plMB5?0fnqVn*-~x9>u&-PFKLEsfrS8dnH({4V^wJ;%%Re
zyUVcHNX0LoW&<+YK0JWl!$D~Kv*|&iU~&~Js|=Li0)72{6;`3@=;-KiA4X$PzQFLt
z^Q*swh{xP}w(Nco%_A(T5%7yo*}G58E#Yau@Dd($FhiImcntZt_xz(1)Hbf;{2ojc
zma6iU4?}ew&B$T@<C;_W#WGC6!LC2C^JZy!U&#^|mp{o`@9)5iIDFWSn3chalf4z#
zXSUn>`^MaI9R+^okHIo((Y5xxX<yYlE5rSt**npY3yz$@CbcH+iLRzW16oH(XpIg`
zQ8Ij{#o02wKmB`0nA4h$e{h;R*q-}pjjjFp5$o4kLJOaXcdBnB!mZfLPC_M?l_w@8
z)HlnE(X?%&J9kUm7fG<P(Oi`&%;v@u$2&79+z|zq$$j~mqd<G=xU2dQKe@7PXm$`#
z1Lu@$8rzcjfzCNJb;hmZ@oJa<yr2mEOGNv;^i@&qDj-K-u7qMs(}&YfW#$?cj-13Z
zHKvhcc^V(T>K)eH!p;$eK3RjQU%AZvW6hOx>WY#I>tj{Nho4Wn`{TQ{xYca9innH%
zMYqbhQ3os1UDx}*H`_hxE=iI87p}r-qCFRKI`YV1mP+GS&=<bJIiIjr=k_6JcoSXB
zOrjO_5rXHu^;-?#m)t`(MLV*5Yl16}+v6QTU?~(3H}dKzFspd-pk;9~9}j-?_hK<j
zSuGUKDt%^yrE<qV^5IE42Ep%_YICPU9=-G*ILi(}^Bdw$Mdb-4jVxW|bu_+<muTO6
zBPgUu50~~H2UtFP4;x8Uwx-*m{jxxPzNN+AsB^*)LRME+VrcS{eAWcH>rsqOCrh#-
z`hi9FK8s&%e0bKl7Gxm=M#fke>t@5);>y`KTv+sYrgxkxk6fkGC_b(9#NdAX9RI~k
z%Y1lADS7n!zOBl_P1KIz^=FcF#fQc)3-Fy)C-@FrWNMRvuh%bMDSUy$d==H%#f?A*
zV|c<BNjZ&xi59oy0_>H~4*{GV@9*%r(IQ@>w@l`!>$bjJ&b#j-6@6SmfVT0+CD+)|
z14h=ZTZB$<hw%L+R|Ud*fOgFS%i~6DA+S32AsH5b!Qa~Kh9<fX{Tw`)#j55U?{~4I
z7%wpPv?ZnQ6W-O`zx5cYavW)HR(LG4Vx5Qv3L&z{NSe>Z+BSA9;i~kzXwuX;sFkW%
zD|{B?)1Yk1mGk-D;NM5RzUs=B@kKL{YChM{8%&t-c*~-a(yw0fX$l!+wFi03FqsH6
zd;G0vf81VVNlzWU5bg-dY)i{naDSrv{3UdDWAnkPfh4P3#YH0HTMP*AbZVM)HsB=<
zmLvl?o{fLUb1P+FBcm{<T-(g`eBq#P0Xm=cJ=2X>jU*FkYK7oe@6!>iY}v<rt0J_r
zR+8BQMF1$!^(*>=Ahw6@&M#^FsM-Cxiod~vp4`mLTaLh`4<)cxjkWOFG61nOr$jx9
zz2o|%V_M+O;w9P7SGn{IRYEs02y*2VAEmOLE(}&xs(or6dy8AWK#Jzobcn$0$7|~#
zI<y&;*U@4OJipl4<8~0kO{mRvDS8Ug()L+=e0ytqHAQ=)7-YbjBvAM{#oGRj#P>>F
z+X2Nl6u!to=NXevhT=<iYdfK}Gr>7gww+t&_5MjSOE{s_TjQolp_SDKRM%**@y)7V
zv}aHb_CTV&MKt8OjTjo~fL7~<{XFNJ%^qDbBbeF%i*egli@B+%jv-G>-M9Rp-fq6{
zpO2IDin5Npq(*<`=0e<ayqNux;oC^>A2aBE<a(;Y|87At^F8w=80G`KH?}BpC$a()
zvZD%?a~<#ao`F-_9n0yYa6c4RNpP*r@9ziCMDrviYMtdL^!bhcg(te@eun|zLfD;*
zS_0SA*WV60#7N#co0Qx8RLrJJz7IM2guTt<2v*TN^g{fr_xlrC6LF!k;?UNGi69s)
z1y{p(nI5p3@>k*@Q))$>RS33!2h_bPe`_)n7NMAnm8do!8<Cn#0X0JZf~P;Stq1v*
z&94u{=57#wc_X?FUZi@+XQ}<#v*S$4jvQmD$?@vSLY>PEKoq@J-0x?`Vz{&`EmUtT
zgr}aEZC41ckKV?HR5wnZ%o?va_s-$R(w%Zm`gKcqMACu2$L6d>3Q37EPFfG6rIY(t
zUKiP>60N+M>FQ`<JoZu!T52SJo^Kb7vOZOlH%1(3G3|to;jNe{XD*p{q1HV;HbP@D
z9B75y=r-zbk&o_{gj^0dEVJApq}z1tYP?IpWl{F3|KR!u39#=7)-R{}f-2k<7enRC
zSLdu!y`$cxi(6e-U&amy!i|Mp6QUV#hY`drzRG6Bf9)CH=*By`-}MbJqohLSveHDs
z*}C4XiQHdiq^h*(DEAnAR8j=oE^IfQZT8X#K+PGvlEo+TJ+81v9vhmh1BShG>7*;E
z`CDhCZL5k;r=X?^mb4h=0NxG`8ofXggDrP$F4kb?8Y)O)6e2&61NZUzPnu}bvT)IF
zC^z205csh{<$n3_xtc5whAz6#Ieo?~%)OF5DbD1XF7l3FkZpVdRD*w;xV*?{N@TDs
z8wcT;?llH`yeVke<cOMnM>%z?=pM|^T-W>3Hv7i-Qv}8G{@G~Q<mSQ4CeO5V<4&i%
zw`b9R-?+Gp6p#0d@gBn#P=<(OdiK|TMp^E-bTfyyV_f}Y<CU|;_b3<*XEt9!_;{JX
zf6XaV#1mp(vJyx|^&spN_4*U&G-3o^0V%1K04_ccwhO~`sJfGbn0Fo@FGCa>|GCSn
zU;oGPG@eu4!>)$HE;v;T^F2OGZ;yIb*LP_s+Ao~+Bgs_VeStG>ju#q158TY}x4!#Z
z@D-al!1Qx?$oCD3Td9OIww5xv9M(|V)a>{7R|1Zzu!lESmXv=(yUM^B4EVR1q@bXV
z;je0%aYO#Kpgs6BOxce%lAfNLb^9Vwx2&LtDgddutE_$0$Ss9L%=oFn16#H-a^NN@
zhs5*RkS`(c##GQ{>ncb%fB)mQYQ8JNjfnb56Xm^0*T5AM9p-emL2>e>E+T8OG5E=u
z5?GwSA+t;_E|)xK4=&oKclfYAJ#x62j2%ozSk%PF=bUex*sXm1v7P3mmaci+o@%bx
zM{I1&^Ot&oSp1n^NI(pZl%<yB5yk!Y?OR5bNN+N1J^!=H?{Ai`48TSjJmYoS#rZnL
zn=3YT<FD~(l9Uz`m50TQV{!4o+R>-fE!u+U2Hb2s^#@MI``)58GoA6Va7c4iY+7xa
z7DBz%+I&cEUE<?9DDeHxqZvac5&qf4O;#Zh*I!<3l6}nJM`T9vXpbTFetj&s72nKt
zTCnASidwMlZ-qn2pRSsVKU(<b?h=Dq*3-WaSABMQTcq*{zM0s4lszuhH5wl9Z3JR*
zRo=3;!t(3Bh_M8W^fr@HMY)}{USI7?72<~*`KESFT@tQN*^$!8xCfwk@Gt9jl%6mB
zR@R;`eVTBr+?&v1$~Mak7=Y*ocE35P*({$aOeN;{z)Q3AdW>_8iW{+vr@|#V^j%?T
zZNyzFUX09l=)g^}KLJ}*Ry!3@Pn^fz{$XH6`7t&Qb~f{ZjOVSeDu_)p*v`(|1HJIB
zh)gb&slQJ^amr_8kIFg)Ugmi7972Z~E`FBVMz2Z)0}h+@Olo|OZ0pBQU>A4pflgGN
zlZ6V#Am|V9?<?Gwx(8)bQ*DfH$2#8DD(c^u0u9)y9MIFK0{O>n^1j%kcQ8Dr*sBzL
z_&K4cC(P#Kh$hxVIO)jDlHHvH_{pbysfSGv5a)Qe&;19+Ra)DPOsWy~GyrwQp!%Sp
z(6XQ9oQHmA62;U9ERh-Aaf#H&VlkU|w~mT9H&>>4lV2b>o0F0v;*wh$?k&~DVNg8i
zeubkPF^M9}%JdZY;b-W|&-x`Re;}qh9k`Q8#kiUJ<eBuBw^6J%pOvr^ZTz<zBL24<
zqEN@U7-A+65vo1@g8vTE-HWgeriN3-!JRHU3yuuwbANNxs&YACr=|(Y))gso+|~Bt
z#p5oa1&Y;f9e(SoP?Y{Qrw;qhv#jbQnGK~rz8c?YOlYjA2c@Q@%^G@AmMqQ1P)$pT
zVp+C2xp8Wd9PML<jTEV_X5bz1aQ+D?ImXGiW6wD#ft50V-eA&7`?ohljGhchZ)WZb
zo~356`=vvAlzOM&DtT~Zfd%NsNFTlq9bxIqmOzwDE`XH9<xOthZgzmfnto*35g1iU
zX!CHEF%ZGG8t$xp(-}b9qG%@g9C|l$V%zaSIXmmhI3I81U!*N)><Kdxdm&r*aFjas
zSq^aJ9r|5?IeNZLld<w<U#$sqU%hrux*A@jDK@97<OYv86OOiJa&v2)-q-;5ZngHp
zuDXI|xg>c(0olX(0oJBX7gnIr>yzq$l)yL_#9V@Vl7Q^jC=<eOuV3QFPwvPwv!eVc
zzrm=He(OlvGN{K(7>}ckVN6-V1K9&_2Mmv|+)!_1Z@f3|jAER5TMb{ACV-ER(Jg3&
z#@6-Rb85;J_!MfcBJ4QD@na!HCLw!I^m=TtVwqa@49B!{GQG1-g!3akgGo~z=U7rR
z1^+>16#w;5GFL@_G#1~HX|J|*UM}o(^H6KU3U#^I;nGa%mcBurDCZtn2y2*qbn^X+
z6_O|xi}ZLhVN+>|B_n{VFgrnGD$^w*GWq!hw79-8{K|jh@u9of4tFCZ3+_C#DnD#2
z^~ZL0+`dldT$iNL)l$S{gSP{Y1}@;Isf;3utUpt#M_~Hd;`BgT@}ZlrZS{>a>G@W4
z_+?^Bb($IJ^BJR+LbQ@9CohP<Nc4fDzEkvDWx=e{Fd{7fjVBM(<f+77Y+emgl^KNf
zdCISX)i!@`3#FG~Q!6CdSPKs^t)x79gA8}*+yI*+teDuayP%b+k<qCr0%#-=&s^=x
zDFk|<{rDlif!I)?Kc<m|D}8CGz1Ix>`G5XT3n15IDMH5yQyU>BA$Yvfsm77TU!TWC
zoi?AhM=-}t<#8{#Rtq1G)mim?7dIYscWvR7DM|&w<k%XQK#|6|f7zCZ4afRxCuc$h
zde-m3J^E!)Cu2u}@vQ#vz3uPYx$nVg(CBZ_crtw&Ou7$!YMO^32RS|e&|9Ke>uLJD
z0Y0F%rc$(v{RdWF_@#h-WoEt_dU`8lxj(e>tZhLgFkho<G^a@TvW9$ZEy1TkC1dIU
zD*!1(Twp@{CBAujL?WjViyLi$FzR5BbpxhKXojfKS~5f5^5e-3<fi*8@?7ap3bgYB
z`(e<P6{;Og6LrLwwzz0^BUZ#*4$?o9ID2(yqtrzEo`Zjr<*^16?Ui#bU|Ib(m!!Ec
z+~Du$^59+TM?YhI%oEU5pkE)ti>a-x@GR_vUXRLv90PlO*yI$*q2g|GQ_``KoVf*k
zJaB!(!Vt{OAuKdt+ad0MTf49T^uxoMcY#tSE<phLWBDZd?E~7Y4X5%?Y|EyNva}+r
zrNzZYCMfxZbJc;&0QawW2j8Lt%FL*yz-Jlo<Pz!o?3Y;c>!rabr_tlkCwAZ0s&ca*
zBY%$y>6`3<uaBo(T)hFD+rMY2UuxbzByWQo@lt)}Q+UDBfr`r+oT&rD2M*Lq%y*}G
zbo%$EDGHoa$Ee3Xx$s&u$Ynn-Q#Q7c5HSFSy!C>5?!wzXPhoDRkDu$j-OZ^CPDztp
zEyrz6!X4#@T`z)l(3s+HqdN)e7<_Bh*}G{g7>7vY6`FC&9gPes#rP%7bH4J1FlpSH
z7LMC}=55&kqG)L_a4Sw*$Rp$qLWPKlkC=4<0|W{2^C@FYAxw83I*`17c4ad>?HxZq
z6zFWZ@BncZ9lWMLpr1{|#f8aTOz&4($<00y5osBkW)2J5r4#OPgH@m{x@=zlp?{>L
zPh$IwN%*t%=0dt<)tTCF5o=9=D&e=iyc#trBlt>Ra9Vqj7C*rP6;by2^0{6##8dp1
z>O{uw9d1EFQqI|0q_SKTkTdRhe$3ou`GpVSE=kE7Bm@Ol9bu#M%}HF0XX6YR*fE>G
z1G26Ru3*A6jmSC6BSugR2D%O1l~^lDM(Gh29#$t`IUn+iB@v|xr?J6Z(&7Hpy6b-N
zKqjS_yL2YyA<fopWwMrO9a4o5nEA(Vwp-AXd<ISzZ-%vjHva_tV(p)cTRJ#sOL5}x
zWS$-ZE$Pi2sgc(wljj%kej!mm=oqZylVc6@3EOKTcpm~UTdqY1H=(@oUvc9&csFZQ
z#nJplSt|EpCKjOTH-Mq}SFq|*@gBC#_Oi}6Bbj*~?p!Qw#Op{!H*Q&*GdCg8HSzt*
z7mQSs1-?PqKJ`(!P34`8s?!AyV8IecE6tj2=KkA_zW@KzxEVs2zU0HwPG)QmpyMW(
zKmW*E6fBlrIv7G<I1N#842c(Ql4l;dpAd0OQ-ars-WObYY0_{D^T6stJ0&Pzp|_bI
z0n_AQEA<<qla9j^B0biwp2?wAH{n>fdl1Zi)>LdBJFE%Pa(8mB7$sr2aqMc6sRt`*
z0U3~oe0Kk*ro)pe)wj(3%%uC7p5q&+&@)xBPaz{xN%b8=NpK1B0gcFM8_$)NE@>;3
zqiFmkX8Ef7BH|FdZGJ}^rjot#ig2g?GSKLg3Q~q($aqt?9WU@CJ1I~J$iu22CQ`)w
zj0by<8b3#8H~zM2K^3ViA8}KWRQ)T<PwaTpO%@Co=1-E#itkRF2Au|zp?DLmTsL0D
zQYY|b*eDa*<;igXPwkp^U;;-Y(1I&2G^faSM$t^&+A*@26q+kQXjDF~z1H{zpp>U*
z(_pXT6Gl-9W~#NFBMDjAzR!)(#}lnZCUMU}F*`U=QYinf$WfXV!_BbpKV&~L&Ud}&
z9+}+~;XA%@(LS?7Nu#@$ub7RsmMN4J8{}W5p3Zlor+xixQpbP&)aoco<fXSQR!Et#
z=aoDfdUC==-C1dWk*M~FbZtf!Bd_k#xr9TgTGRuo&FFwN#T~h(X!cgkcTr^%%+J?v
z-y{_&&E!l+^`Rv*fT^1eDA(>~KioL@rd9NUl<8FC@@dGmr407l1XDD_x3+fSzp5At
z^X0MaG?}})B4X6@W*j-^Ymp!~;W}C?w=P{?@on&mQZ%D={P?@hmJCSwPCwO&wAEF-
z5kmjFV?0}+L_j?usj~)bv?Z>6M7lAdPzw?lH@7h8*C?OSWDQYqEG9xZW1?u%jbJcD
zm3VRGnQ5T*MZbRhx|gWInVUl^K_{DLvMVJvMU*}Olx6J%iQMm=-;RO}xy0Mv%@3-{
zR+la%FCb5DRcbL^**~;4_bCLY0QnJ@R<i^vZTLTg=@@4r-%TPb;!-6p*BM;>{}C=o
zR6bq6V<#o4)Bp5XQ^<&0nAF4GDHz@ni8=GsOyZB>9`JB+tG-dPq}>qis0)O$VD;*4
z)%^^Eh#i4?sE#PBK9pcpe)CoS;iCU(esr;L<F4q1I~90%)nk484}?hYacueAir=?$
z&Hc*JueIT2YRF`K_Xc~_NR6WL!530(D`HBmu0#P`ML#TB9o#P3%R`(OAh6+#@sEaP
z$SxjQWX>)Hfn1W)doUE=(ciV6@BqW6TGigq0RNT34#Q{d1uJz5A~cYd48Wq7X&k8K
zNoyxlR9QerpxiUDic*y+jOFwQ{I&p(+PLN|TL|WffyufJ?cSLT2RHr}uyI{YY2h7T
zMQL4xwoZO^LB7f-2d^hg+@HrMmC&nL&%d33qqL{vm@EiDx-!P<A)my8`F+c*ziLw;
z>mr)ZFIOcbZl*pId|dQzb#c9*Th5p?t5}%hs!!<6k_C4i3H1#ecN3SnoE=cLu8-xU
z@?@brV^c<1el@sdYj|c%r+Vo8Xr%xXk+%o4=Od;lvVEEOxRf}vbT-PryyV*m#Tb3O
ziTMlM6)J0O;0I$aZ5{vEhnoZ+gZ4A)=mm_Tef=L-FaU?pU(hmyAa0<zFHjyT4#M|`
z%}uu#9*hl%6h1bS>LB}zHgH+&9W;=dzR1(^0ZBY1Su%h`Ky|I(^kp5L|J~odG_2N>
z`Iaaj))v13{8z3rg5|*7!!Hh2B)}}3eskcDZD4aYM1o(%8x5S*F`x$#;>Yrfod}>t
zdt-!0sqchn?;>Tv0U#VZ9n#`C{jFXiD8Qb7yHNC=TMvO2HhGIOf0krZ!oB(IZjtfW
zkVDBQ(N_`DGfdNt*mXSfsOxnMKFk5pMTBDm?xPtYAp_Q6z_oY+P-J&&JZpV(b2x@c
z-<^IEdPJGf^dHl%DFA(V{sBOw-y>t&6l*<MuxvCwOp;_LjIpzsb0<0+T;q(<Eky*}
zd%S)Tb7z=%jmaaC;h!tJA~a)oySVY2fmw{V^qHwe^_r{%3OpLSgc%a*%sPGPTwz(w
z*70#Xi0<HaB8mar?vvooTh9H}$E~&vNe*WQJ0;s?8;;t#>+Hdr9~1J7Yp*J_3FeiO
zX6sKm<!ZpQQl{7G$65z;!v)x`AC()ChrhvRmL&4X`*b)nGG#qI**-uwh8ZL0j_vH!
zkaP9Kj{PJ!j&Av@>P3BBPBXl-pG8l<bz<h<J`(Ad`UhdS{y;<X8BZSA17I4xbC+_O
zHR%R*NfQs2gTbtcz-EOH*Sw!3e_+xBYwf>)g*GT-Rh&$0C$sERV$@A)3;&^=?&JxS
zbTBxYwqdP3-T8aq_-`gmc7IS7zua%CKSr+P)@YdGHswq2CkWTUvx4yRCE7X|GQw~4
z&8StE?wOGVj=8}A?2x_6-U2d(god6@8&;s8{IY3`j3*{W_+dmhR};(L68V0`0txm#
z%?o)rcz6Te|N16ufc>2PU;e)g;XLv;a}hK&8K?H&;q1Km7PZfu!GMlt)m9Znx|q9Z
zM;(MZdU#dA*9bCpLR=!cFbzh|Cg6k1(UVnz*V{7f|MBlWNL(TPgJC(ya%{OfGjU6O
zp^J&dclN@GZXbN5c`s^NY09d1PLS9Qu$@NoJ#e|+U*EexAV)gjb-2OsaTW-NC<?f-
zt1+hTD{I>9580snsaT?_`J01$&c!Keey0z_5W{j$o=!H;h(50sDFYJi#t%k>id_7K
zbeS4y3+=ThrT6Gpqd^}|*%_WGA~X2Ci?txz9cI)HMiVT$#PBKvdT^ubqN1Np)wzbk
z0xPSj-+u$S<9B7>DqcY&F0z&0t3eGwLsEA68(X%UApnWrrF_{YWL+2;wH@~yOpZ^=
z-F)5UwS3_5heM>A!0iu<(>{0#$huz6;leMUrualbVyCXiPdJ=BK+V4vcDI`1%xpvW
zK2p}fF=fl|R(irfxPgLvxlWMZ*qpe);6{%=Axd5!t>SMnFI5EyF6y^QHM2Z@?}9Ya
z)^q^r`+Rs+pt!<x)LzaU^dYZ=cXN6(*2DtRa>D*!Zh2!BdTc2tNvPtNhwW#3EOl+`
z8!zaUnshd08GqIknGpOj^c0S-UV{$a7iYjT&ct(!W>yiFk>)UzS`sJPNOLkmi|LW~
zo<cnD3+__5mJ)1>-O3D-MZY^aiJepM)hpFW)Q-1LgSfKxi5(M%Q=4yeG^4)46Cs9T
zt<c}(SkBmQxv;De?H&u^+5sQGQaX0Hehwn%DA>yPT8xAdeVP<kYdMw40OR~TT}qqN
z3`eiL7QeFcT__gFe-~0~|2yT)FHS3BWmnnMc23t>*~4<L>2UT>Wc2~@zqOR~$pazU
zSo1|xrkCz|h`Bgz{=^&XP_T6R_vZ*k%lAK2J!G#VrqmWbf^;<Owu*k&GxodRM9I`m
z4L`U~oRJsscgkLgAfNG>P7ik*GPKv~J8DK4YR?F;vB}fBS48*OJjN+l8v|QZ<{mv%
zEhUOAsp!Zn#&5XSo5i^sAb^${;3n`=Zfr`V2_<tQH^7a)xX~UbIH<qu52_5Li4^BF
zdXlF%iMX~*wgS5o(50LQ>I8m#lMhwWh_rautWPIXEz~||h9^eVOyP;5iyM^Nq$k;s
z6KZ1!<1F*;xG3X8`!g66^fNQBcnt!Bs{xvxUUR%Vm8Zv$TbRtlPQs9A7eroc)FXob
z^0B9qrR;b()GlX{onND|m52O$Oqy2TRh)B;_S96!KtmO&jCAzKblilA<2hkaZsGNm
z!*1!T^MP}{fg6$_=-<hacsA*duW?+ujl6<ld6C$4!_DpX3t|YJ$&y0{E{-4NDM}xM
zq1y16p(O4#&Bp?FBzhpgUR@9TdP+ggTW!zBmH-_gx4CGAgf&1Vv9NI7cOCxjzRa`*
zJQ;OiHoIK25H@`(;QXOTXBar@ZR;EVu&7`X_;15CxhC!1h(xgYSH1A}F5bDwZr_&g
zROMV*&R0Z4@-FSXK$3xK`hv;jBw^wknj4q>tEHO3+zv~$o0gQE{!skrv;)4Ess%Zj
zKFnV#`k9@(U)+vhPDKajA&JE=&UwZc!pG@H?}ZFv94~i`H9%ByXI1X`MmW_Ox6_3+
zJ`;cNS<njZ?d9A!9?<ZRc_8GUYOcBA8S2cc!l7bq-`UFLgw|S$?8XKUbBLZeT;A~^
zz`3G=b`~YAFkCsj*B?;V7WD=5TXK^-U)R9(TKd(>c)a`9TV*-7gCUMrWl#W-jC-V*
zq$;ov(BFDm|4xebR8OgmcyQe?%e(H#X0&<QvQHWxT^FbuL92Jzd{h^jiXsbIon4UP
z<zepUm1_hz?|!9b6p_N0uj&9CG;cCEi>1K-i(Xy!VWo_0F^cDse7A0ZL}+cn=333b
zDenN`LHeQ)0P9`9^Ddm-g$}Aa^3Ts^?-BfPw3WI1ch<dD{ZFO%ehauR%4e*508@hd
zN@Ef8)@RkKmg%acIM7=0Qz=ZCf1?RDVE8n?05m}k&%DDPT!_s!4|Euuypts~?X;(w
z2`8-u>U1;%KvKA>O<Vgo;s>ytS$ImbNCFxFLgW)HG~+K^VOee*F9Mg+h=y*CsoNA5
zYB&JD9?BbR&=3ux`iIR4eqM7uT=1!#Q(nFVDPd~5(g#sFy8!r{V3PtzRy^pzd`-|I
zqCsiFfcg;yX(OZ78%bT9gp}k<%%70iqRzoU<@5&TOC6bd3fsKK`>#}0$`&mi3zBVi
zkCIISCRNp+{vp9dBKJSZox0rn8#wvhRZx&!Lxe;Apo&RQ8m-Xgq8o-WL?#qA#16zJ
zvn2MhCItg{P;Yl?oAF#&rtHHx{Gmh1i!&J&d2NY4O4He6QgqC7uxs89a}aj$HwvbK
zn|Ec#;BKl987K;=2y`o6zD~)C@-dq|7lm2K7kH5McX?y1$^yjcC=jf-dhB!!FfE5R
z=h9oe2j@4wrRBahlUHy0ylJ{v{sut_@InaM%i<4Ph0q{Jp7w)sE*v(_RUWSwuZ<~d
zfzQH~VN%)`bne-cdlic4gBHv7@F5)u@-^Jals6^;T3SI>tB|M?F)tX`PmAFAP8jwx
zjV3%;KhwW_`=`0+`jsfQr02+6m}k9l%muy)yU=u#3Ps%Rd7g@!PSg>3RvA1+3uMl)
z&s6QEU9Xsn6io~XOPjg=B6jUqAxI}&%No3I?;W@v&=2Iaz~FFB9|9;2a;ZjzqfcXb
zJwy=Qh->GDfDk5g^#i<2vWc}V2{U%U%1oiTWCyx9b>w#%;VsT9EN(~cFHDbSt1YXx
zomAh$<RBnP6AHHJBt5b=*VRBEbciTyf&D$0BMXiQkR*oeq<~GaYKHIU(QBM08p<nI
z&9=%jwu9`DO3eU{FZ+B6DPa#4Ei8Tslcr3x%<;wtB_5N1^LZekUf~zpNGXwAlhpu#
zg|AuUz?1UiUZqTiPd~)y!X7+q{K%<r6Z|<T!L4w!ida#fYChvP1RQN2(8dsePHPuM
z6p%oz)guEf@*HH5;iBi?h_lT<0r5;{Q7?9H%}`Fg>h=2VNZKMl$nJ;&JqM_pB)p;>
zXxQZ6HVoU6tii>_q$Ih5>ApwowcR~jYHgWUy&PIb0)PO<#x<@Ju$GU(wNYR^{IX9_
z_`6d&f0?KxJR~07?8qGS@<rP}r=~5+p0}Fw(NISXf;2Y@-QKJNe+T49qY#YF(mB=X
zlmQ^X#WdKnD=@f@B33{VUP{5kjoj6AXuNc1BF$}0w0I}FS3-uYw%K<pcmivkDEA9a
zxb%+*PZ4q;%b2X-SyKzn_8yi+b-i*)fLR1-ppnRy&Z$qRd*GExuzPs=SoHHvJf2~7
z9DUFm=m(QJorGEq1W-$2=UA`pD(4g`DRqO`+9Ysg&Lp-n%LCdra=I>ft_k|R(ZLvg
z`Mk+5FyHOIqkMvu&Kw)MN(GyW+Rr9jjE>LzyvQT<J{vgf3=?Pq;3P>oS6V{fV@+>7
zk=gQSyKdBGrp!xYo{{?IZHU#)XyOyHl@jyHUyz&**xQAF!ih-GUeol(z#?zyG%u<-
zESki&+Nfgl9lohF?a*|Vn<qp|rBzOQA6i~?0n<7gpw#BWyk<+0L4o<PSf@I%MJlQ=
zF3zB2Ktb2{LQH~++xGXx`_yio?kz1=L8EK9Q+qH6cT#DdAETn+<CxAZXXVpGDBk=b
zozixvYc7p?sLK;s)rz9>VXW`Z0W9pC@)so;MlOP&M!7`KV9!_&eMjJO1@JfGrbgf1
zh9{As97p)0#VI*wo`3K<0QN`9t3bO&k>)ad9&c+iCVb1}RSHHrvnX_eMfot#n<*vM
zBq33q>_jc;WL_$Dzb2rtEL{|?I3}K1k13y;^+npc$?$tCye5d;fn?l8h4}33*-8kk
ztR-TrV|ZYvrQ5WH07O-NELwV&_a<d~@oO^%Fe&f0>F~a*NMp>N0sQDkT6O})j_1QK
zOP6)n#QontNGUK+B4A|`83fmw8@E#3iJKE;;=)}!6s+qY5PhQ>*!ec5FfAmHSOi3V
zg4&J$onf6J|3Pg}JF-fe8%&p)TeXg#?X%uh8A14#Fciu{Lbr|2H3!EiCb$Dl6~a)^
zu=n&AMG!-@^q$`~b=q8Qv1Ckq#I(`9)={~VEx6N1H;`Q@>*t$i{Xhm5Y3N{Jz;W<s
zbhyG(dw{gg#}aMcFg?Wt6czvjXSZ>dW=Kf?-PgxqJ;*~<V_$8jO~DrJ28Ks8cDycd
zIy7ju0B=O^L%jw!TkKLJwx@cRUu>D0I%MK^w__!{s@g(ODJxA`t9A0DM@}*D_bg}z
zpGcGRR6*Fu>x95z?Xx^pX(O4Y?}`g3vS!;ofRTurqX&eulA^`A8BV4F=j<J)5|si6
zf6uyJU1(#_O@#%wjfu6c#h5goCZ^1W{G?AbToNE>_>R6;dnuG^60qhTLCzxLDk8AJ
zcsHa+&I+)MU-IO}ta9e?k)5Y%25y{`xRa5LAvQk<nhJjDEzkSl;Cgs0{}}N1@PpYS
zFHwut5&0=(NDlD_^e|u6U}15`CKzXE@RhxT;E^bE{5`~n;4wLu5~d=jHY1H&dV6}z
z0zU3+9JweVk#z_eCd{-zM1xpFtp^8dj$`vI^Za`TZ9&k53zM#*9=UjfvP3zfcUY;e
zmMewwaF34%Z<lKTAZxN{<lAC<)RAc^3p3Nm5+F_naJEUO_KY#G3nNnM%VP(t2WgIR
zADv?E#vB5+^}Cz|jJ&-)(pVn6_7g3l0VJpJjJ-C`Lx+wD9dTzT#vJ>vR6p)_pW|<E
z2)N0!yWH|A#_~KT{ULT()Kh%{BM2ad;!#xDdEJGiXJZ?pLV0LCJ1-SpZ9k#d->b5m
zxBE-Rn&IQzk^;zl@G?c)=GJD{vF(1L5iSa&(v*PeqrrjB#^~@{1K#8JcK6GTNh>w(
z#9)i_1E4S4HQ{3>YH!Q+4BC?8W3NvfXrQfp()og?Zu@C3DJv$JjSJh{Le6Dw7zYN{
zfXjtN{@@d9S_0-E*42GS6HTSr>n{zq?t+#Yo*JqI*I<uB$-V&+{ix28hJO@L=>I_H
zCj($jIhZ`no+z=~2=FEY+Y|`JgKbr2ocN*s9f$d*Ojttlf%_*G+Q0F$G6Ou8`U&mL
zZ$cYIz7mrnwqgWYj8ftFy!Mri$ytQd242ZOw7P;$9OsiRau?O4C2V=TdO@sZA}6RA
zgOkGR+iPSSUvCbD*b^?g3uqh>=OP!ae`RcA5p%`F$nX;E6&ZoHDxzWO!P^q-BU&3L
zHj(c5DkZrlUv!mzU|^E~H6KAg1h|zj#7mB-xkO)4ELl1lh9G5x)M_uAm5bR$KduuR
zI!T=_e$NU?mr*PfrVp;Z5}o+ZXHI1$Peu`p+|eAo@Vnc61~ONF0O2W%v==eMqa}53
z=a7IoZvs#=*A(az>EO@AxkX<b9kM-c*m`abYZ0j7D~u?kyyIo7+&b63RKCNm70a)6
z^j|c)&z^cIa?;W`K`WbikuM&xf&1po-<(i>%xEU?WC94TiFSvL!GIidKugOGx$q0d
z)@Ptpq~2hlzNO$+YJ)OL$Tfkg1_2Spxg+eKDo*%`F=`&`vqiX}fNP?Q!Rj)5`ianA
zbTq1L%|uP5@o<YBS&mjJlv;HEYMB=Qdc&R2-y0*1b;Xz1y*q5I9m5}?fK)y?$&;+;
zaU&mGPnUP!xy`qgFC0QUz>K0_8y(EAI@c94j-6M)Pnh33Vn)cYBEabfKlA&cd?97@
z)|oa6wB`P4Y;VIrL6bm|1~@KxjHZqc_&usDW*{3}OzJ-6-Gjm~<v>MeoPAO&{tn>u
zVO(@2@|kSQ>@=48#u_}6`$B>U|6}HB>=@yfwKmCamTh{ij!A4Ma5uL;wV?f3_4usB
zlMj93yedo86O}N8BFKZm3xl+)E_gl1vb{Z+-Ii=?^&%>uxoPhJvl(kV80$KYlpF&v
zg=li`Y8Dox_a>g|EOETB^=SEb7?A|ht@AhW*pI=xF|ac;GjkCp!-E0JUVbR2gGSJ1
z=LH&eMLpCY;^eWo?}pwMZI&Yf@Qi<DDxMDb@Z*^s7Sb1R#e(dqOYQt@o1pQcf)X^;
z?Og_4(bEH9lmLtwdglGQfd@zu?|LMGIC+weiO$F^C7!=~sHKHKoJr9U@VuHD!g_vg
zXaFGoz~_@mRa!b+>vD5Vo{3!y0PG;;11~Wl>LkR-{3Amc1aO%+C4UoH>QL>jz=`s}
z!!ZFtzHQsn^#$RW6N8TbxD12<7FylQdz^+d&*%#mB2UB=z@6h)Jv3v7UwX9z))JuP
z!KUh`{V3wReA+h4cQ0+O*B`>Vl&YAk_SBz|#!Y?{3MB1-&UDc@?p81C#sDA$Ny%v>
zAZR=Fqri?UHr4zTn!-sCeP?-iOPUA`GsaIzFZI=i7(P_Z>!bUYVDfq*#)c=Vh1Kkn
z&3)U(#YHX6^--Sf`1|5R_H*{?7c$7?8WVaUV_OjvVS^t4Q^v|tt!x^qEFTCmA_KZ>
zz->HurfKPJE%b9t)M{@H8ZP%=n%y6QFKz#Yd`_<O?O}G(T{8L{h9ju3#nG+nDJb%9
zOS~qhA^HK0_YX%09_dsN^o*YcyG;cJ{hr_&a=+zJtk6Duvwq7^q-9H2!Wr&;`EF}(
z2K1Zz&GmEBtTd0`$*kjEHn9`X3EeX#EyY(S#s-G@pX@xOwv#(u0QFwOKx7+2MuF6-
zYu4bV53f*Hp2tS6jw|*=^uq=`szv1PF;E=vGQ`tqU=^F7F~exd*U9$=D0~3(sd-yv
z1M43dAShPUrgP8~QC&e;voj!4%ctJo03`bJqN~+^84~ylb6PR3_xG&N?dfV9=ZEnH
zppv(zv6<v!2~8bOe}Lx#vZ0^-z{gboPYaNbQgC2PRms$jrws`}bHpgu=i&dE@j8S4
zcc2KH&Nkk5y3Qfo*fG2#k1nA{ouq?_2$5bF*HDLd-tkVE_p%&qJD1=B7(L`M7^+qA
zCA*o(F+Yxl?ANEdcS*$r4=_h%eUhd-PJ!R$okga!{|F02SiX&~Kj|3ypEi>2Xy+1(
zgarA_1>+<@RmOPptu|qYQ=gx1W2pHV`%Tb@IdMjz2aIgm2$>qt8;`yLt!T{0-B{jW
z6;rM3Va|`)+Da!>-I%1XH`-OA%?wTmS%g|+jS{@J>%hhfkDFk@>Y8B%3|t3nk0^i8
z4W)KkGM_j3yRvJ)3SsW}?0oKD-;-`s!M!!CR)c(B5o2Tf6nVO*vR|A@og}5KWbn0Y
zHuq9N5xPaNL4<~V;xIkL?f@(9gTkKz&*s51L_?wI@@V>ENxg;$DH4b@llt{NJnj?B
zd>SbVVHn7Q8_337hn<7nQ@8M3SNA^?JNoi#%m7jtK<@%2!J0cBX91@pcoDkMTk60^
z<5G15VYKIR3gLn<SHnB6ber5R?N0ya93J+jfroa{nQEzYh;Ba@65OM0%P^o-8v#LR
zt!8`?kaEoC7<Ap34-<H3XUmst<#-Xe8fLvHu}_$7t)Ch;9uVpM7(<Xgaa1phb~0Ba
z+Q)u1^z<z8;q`PHWXE56snO!k_?Y!vIrTmYeQv&oY}MImdz|@1(VaC4$^h7bCAA}s
z^{c&emlls>I6tFn3ZOk)Q?hfRRX9z*L67Uia`#)5%^kn5GJ$|S4cKgZtM><#Z|>&v
zy+wPOpp%Y?1w><@QVmx*La;BRlHex>M2pH9Ch_8xro(O@n*~46LOjLB(f(kH7J!sT
z!9{o7-C>_=RWuUQ=S}?NGl*Avv-yWyBK(J3{xz=uD_@@r@g_(`GHj9NwF83-=m)@|
z(GxHwYa;NyG@Uq30GJkQN;Z-HR#l4(;7CNH1pDfv*V}<Sqsk2}?R}4IRNW@B2i_Wg
zm_Y(jhU4`$(;q){>qwlWgvlwICUe_8WS&u0CWhUxV3K~L?#FVPx>OTa#uMZ)P^dvk
zhRw_Fmy-9OApnWe0Zq6Rw67VXVgA7V4&Zr!02IOWJ<;yK50!c}$<-6MJN4SP)kWHT
z@<ho(@(s(I#so&WB2wE?%u7i~B^C^{8k@DE09j{kk8SX1cF6gsu*38BpqWsROOiGl
zrFh6LGKfbkE<YS(g1t<#;dck8a*`wl-{5C;BA(QkX@jUNSN0t~wp@D<3?!bpc0P}L
zf%qABgyDWT2pw#JNV+Q*zSJn8(*%EOZk7jMJ`UaItw1xGGxnf$O;AD>@4KATd%f$a
zy#<4iZu_Hz<}|bWWswosu>M>Cr%9A`)(&)t=RD_ZleYC-<FVfx=)TQw;wL}|SWQZ=
zEiMA>0h1!+t}XK_Dn}2ae`7)9KTOt|*Eg82qjIg+FkE97j=I=<7}bIIfy3Hf29Mbo
zY>%cBo{Y9V(*zSAcyMrj`CQnZe~H)|Or!Y+DHn9i^6<qQ7?c=<+YpWE)8H+XuOBih
zfOlD4?}2k}a*kZb-T}V7Ua}W?Pv^wD;`+LaIxgFXYF`AT)PI~UPvZrFxzdODWd|n7
zCkEz4W?6+u;5U@a-Q)V#(!+?ol@?{1VYU-iA)X=vGhk!$xFXu)mN}&H=jA`PnAI%e
zfB?>q$DID}lwPq0M;-xpsY?<$P{e*50LR-g(1|nyL_g9YuTCu!bs!KfN2Z$3w&j7y
zjiu--nM~XgC||is{W~4Gy9zJ&;iLQ1>7p;rhA>;~;E+V@VdMKZj%;vaeyD@0yQzL1
z&tYpkXg{+tKfqBV06s06GLJH}Ai(2B>EryL5lFmD?%`bM$y)n{Ru71QSwk6ZzCoR4
z&g3YsKpJvHnhJ1VdI3tME#3knNdvW~=Us?Ls>);^P6K;{Sc~4j=FiP;s$c#~-k6~S
ziDs-F!!ig|(WiPcPdse={a9InEGX2=aFy!{SN^v#X58yn$xIosfeWF<g<s?vUc0QU
zl$zf&+G0tcg9|ri8U6$syJ2azd`M|Pt|FoZ{wqiK24qrDt{HwYmqb>z)NJI*ZRwo=
zDg<DkjxRqS3Xoy%$ROAI;=_0C+fChBF2w5yW^VC$dRlDFgA9eM<t@%Yh*(}eCe3qD
zT`p%mbTcQmObhu)SfYn1HK^9Tpb}+8C=&=_*5$JK=R^T3Zf$&7%ZMrcX@jvCAqfkP
zvD~JEVxb7j)$rul_>zPd#{xPQ<OwK-y>}QE&wHY`snL16&rEPWA3q0Zy06xojNXyu
zmycihxNT_YB-v}O^B0Eevo;rS;?RlI=BvDlGNV97d6|nqjy<4ODzhmwoLfdLk;DQt
zsJ9l<*`~PaPMAGOOEq~|3miJfwd4$yE&u4DI7aJuPlxY-Z3f0@F53Ep4sgH&oY7SZ
zoM|&9%<@(Ncy47LNg20%Q|9<OA>kT!E##^R`n?j7;3vEghB~^D<E*Lsb0NTe6&OUq
z0+gu|RE{z?V3AZyci;@teSjBL`8EaZmD`0hLy1`O0CI|AyoINACq>rlW(N)artj>^
zXZ^YG?=2dg*e_*4lhlf`92KI+{KZOx|K2#Ab=D?;6VfxPuEc+#v49L(m!~`-v9{@N
z@{%>LM;Qv91Q=M~N^Ua$rp~K8D8UB7o3!<N5+Ef4T<1~}IC|716g@pL!MGeP);2(k
zt(-TTb)O~94_@naGIc$^N}GeD6BNjkUMBSLmbM3Nj4ki(VMKd-;mJ2k#ESirwGv|T
zS+MzKU^2VI1{YMC3MJn8`XmPXfBw(zX7az|SUKml9)F<?D04Fh939QbgOz7PCRqU(
zNxouYmeycc7*NV|oEr6u&w_MxR>3KFqU4k$cjrY+eZ0IlAbjBb_`&UodZM(@N(Fx0
zi3b^O1sq4^N!pP&H_^NnRT_r7dSnpaT4q?|3&L#~o+cQGv+3WF&AN2U;|l#N-gu#x
zBLI2GA;By|+G>Fl*x3{~eidKzKT$QPzGInH42qDos;W=fb6DOr?>rrVcaZ79WrM|}
zrwfqZiGh9tilR@H$3;3TJ{vS8@2$a}o6q}_TIypnPx6|5BUPCPC$CWVKTA~W$!>pD
z?QOTI{w?Lrf;5|BAHWu=C7yV28dT^$T^^DsCC6+vr)RC@S$27l`6H+xO2bQ54b^pD
zp&Fb%QshN=5OgMUC3h|qy%Q9Cx4o0D9oTyBAQ^4R`-no=Y2lmK0Ik`41@4Xato`kq
z9eUh6&<7P09Bq2zT|a_QpV#6}Ve0AsAOzjX{nzL2y14Fh4G>J05}<5d-J=X9be(Zo
zUeSN`jA+Uge03|V@NhnGb?vZd{fY?8r2`G76D{b>f%A30odO@d)+_d^&=2;Cs^dM5
zL<VO`u!v9siKIVKrzPK4ff8a%rw5JCgnmZ1oH)FdcQ?1r@w*u13v2!s``nq_%)h$;
zZ_sXx3CC;GUGAu@S_O5d8#^af=&))PXu`#RkXT6qNm#O1tc0yYTaLXFkq2B>At2QQ
z3Ln`z>Pyu8FQb9;B-`S1i#uSl5?M_c%BpROzYr_QqIX$!lhS*-S#lvFs>CU12-H)g
z>&bbwaGKSxM;`g*#I&!zrK*o!SS~VT@x{j8@iTMh_+rXo#;a?Q&n@F<;dFC7$*I=Q
zkN;Q$eLm2r(gnIPoPVX40<^E^#jkdSS+J#h^~$4`NrTHynn{=MBY#qHD9>1_o94cf
zU{>Tpd^U1N`{%c)@B2@y02KiOs!*#syBzW>&Oge^<o3PSv#pq%jw5*VNP%tvVkm&k
zIF70SCs0}RIdC3N#nt?_akc7<jFhFcR)cADL<;axLRJib5kZc<qHN}O5>OHf6pNB&
z@s!t)_Cv~VfG@IuvMb-klfKe02Bp&0ADi3K0Jlaqy^g=w**`6|*cUBN=}r>jV`!&0
z8&1iv8f~AGD`r(^&M8glKwk3AU@D`2dEk#{oeoV)I-RIJ;4D*AU5ks*8XlgQlO&`*
zkX=!qL+qm-0??4Ei(h3OgZJlrZ?RYiZzFt#u5uJRC{5?g&H=;tCGd8WE_Mzk^jBj7
z2_})gC}Zikr6i4mHS>gFB1u2I2dVx38&d}k{>ml}DpXx-m%$fNSK?MklvBxj)&9rV
z2}kcI{`(1TzA$^Am4&fWYj+afKSDwm?UFIM7>LBG)dB9ONFECYT+Y0-?uw_-+_bb1
z5=qjB?a*j{JvP=-4&Dr6|I`9~kQ+DL=Xb$GDgVp5nc&&5-^gjDu~nGX!G4V{tBE~x
zp%s&-v1rRk9z8p<V|*wRo+aFEAb(hpPiKhbOzS2t7`{lM{yY8%f$76{^n|lOCAAY*
z-fzjo9jH+32rr@kb4+z2`b#LQ)VAuqHs2!KplrfrV{L$ER#Vs!vw_t3aBws~WK3GX
zP0?@Dmj_l_%|FW&`YDc?HRnP_=uW*vz%_9!4wZ96;MD%2*SZ6K`r%#KGG9MB?7LbM
zH#i}9fU@!bXu8Jey1H&1+jbf?Y}nYg-3E>A#<p!YP8yphX&T$MZTs%N-?;xWf-^GC
z*>kNqAI({q4=Y!Dadqq%#eE9b<8Oqo4_})w1_mw+GGF{gUO=-5j(E#cJ<;@b;N`|!
zpmS7K+mSC^L9OF7Xt$_GhEgAIKjT+Q4BKA}KqpQgdScS(30;CJDhR`I=wno=3>pk<
zOyVS8k!(g_8b*=qdE&_n;yiC9jyP-Pa>^6lwhvVd7`gS#vMkN5rQwg61b>|`E|+TU
zE|;1YNBJf_<%Md0?+E<+3S^u7WbXWwM6W1;Y5n)0n9Q3_L<`$cZd*cPd;X27wksTW
zTvEHs{v)qBZQ~8RxZjNvo+B6y2p!9})hcl1T2P}1ZWcaYhBUctNufqYzNUCd=Qm}7
z&d;B^e>!tK`iA~zpk3ui9VcwHl|H{kE0vBl%;ga*#+}Isv{yBA(gJ0_I0ZltKDSj{
zzz7#Pq^(BZ-zN2uZl82SJ(|hb+<%ddN-E&>%Cs#D_IP>a3T!QFwiT0I(P{sMEyYp`
zB%S6Vk?gbnIMaDGI{U!Cz+RY+%?I7lrQp{q!5~TC^N;HQD>X;eyP6@6L)b&u*|O<<
zMbDR!2dFPNr_R;T@dq4Ci6jA5>qPzIvc)YfKvbz;QA|ee@70x8sxHR17*9;Ieh=#z
zd9%&%@>XCYxbj3mC-AwI0xyGqB7r%_0H~cw>1*x2Ng6od2xrLoeSs_TtX`IphHTqZ
zpdR0ojn1yF!#Inb^SCYw)>*qJBH_DD@sAG}*`_v3wzSJG`Hrrk2lGV=%v?nDJ+k~a
zbijx!|9C2Qmp{^i>kQ;^GjXhun*$tR0is;Qrf2yvjI71GK;Sc<ad!MUhq0<}Ssq{+
z?&cyOg!2lpGL5NvG0e?<#y4=<SM_ME<OphFZOE$R@3QJPbIa2CQoR!FOM(SV{B=^W
zhoZ9Jj42a-Tfa6H(c{LB8s*lz$E^qa?Frqk9U)LG&au8-7F1PaEoM&YH1W6X3c<KE
zw}jyhSj&9ZU2>_r>S`X-k&K)b<2a7G6v#qfA2@sL8}~NX`&vy{M`KsX1ismWRy2-a
z8I%q1{pBB>7NMwz*uNPQp2%;YUIabE7uwZp7??AOib6OtqLm#zZO`0^Au&P^cOoL~
zy)H?`G|nD6?s>n|)!Unz<=N_l@|oiuPv?S0B{r1t*~ib|lyv}LD3FZb2cYi0DQ)g$
zH5wsaE6s4Uo(d$Y@Zj-;9ALtwCO1>8-0mH`GqG456znb`Q%VuoJFhq?`07bq-H775
z7u%Yz`pGnAIC`ST&)1j|%R9cP?cSI99yh>1xtN9<*6}KE+<TD&|3t@)$63AYj<=W(
z$rWPf!Bo7NVfr<}teTPUPSFgV4_Heb*N{DgQ2eeb;7LayrjlFnVbXta&p<T%rYAHB
z7YaFmOImQYs?jd|ff11}GV5Hf%vn9)v1I~0!j0ok88ytV*fY%=q4qa@8@OtHi~cWK
zC3^C);hJQQb+uhjbuS`RxM6=GX2mxQS}$2|FxAxGoEBS9;uB4Ld^bzXch)341AlW7
zD9OttU!CYjSb2V1D-`eoN~SBSnrS68wu=Rj7rqcOZM`0|tzp%7ASo-yxZE<2Uv%}q
z4?Begbo9SqVP`DxSBuWr`@G=;ACjY`WD>Y2(gl!jwLn)nYdU$KA*HlvW#91vgz7WK
zX9w?qI6(rXk@+G-xg8hU<kE0i8{SjDj;mRYkN2D*L1Q|KSPcQi?KHFoJ+G2!@mCn=
z{J4qCJkcCxwg&dOW57-Z%;ajmz$UBqg&p&O*bG)yb$OSBuX{~}h({zz+s-=QyUa<X
zvvoi0Xc^xd#v{Xz%ipjE((4asmn(U2N9x$%^acEBEu5aXKTw+y`{s>>?^%Is&|^|I
zKzyoC4zDJ*rZ&@pmnZs?jgzh7;PX4?<mQF4)pl7HOMjzI6lFmaALfa8s#D=JWxcDQ
zc<*(6*f*B=$-SM!7E>5S68BukxwpGdGBOI#dl{G3!_v*&A0e|e0YrkZviUMC`3vZu
zQMrWT#jAr(O>IVE*aQxCa$+4FR-r<I?~Fxue~?&uvhX$`ZRQwVt-6t?41SjA#Euy5
zeSguyQ{BC`(kg?qz3|97SZ>!BbA>Y>@TS#t%l0L93}DIL<)KX6nJJIbB!D0)AD@mc
zoZ%=awh=MTxwOnU+PM&|0>#xgug84UDF=NaRUC~NT1<&Ho(x$!rQ@{4ItJ2yepAn^
zt-1=DXDVLg@un~AGsC{p(bRHro0jiVjF7;bw`YfH3P#w^Zn|Am_II(#Ga3Qa)MSh7
z7^~O#UFUUE`Z;EJkj+`%lG7aa{&)kkKp)%S^}X`U=}QS&$<IkhvyyPELmmrW0YkJ<
zNw2XJ-2+xK$2c;J8n(YL_Tnq^sjn;nqFnB5JKvux+OvLbOi;S)8H=RG?gX4Y)HnY8
z(C5L;v^H@&sFy#Z=p%3waNx4{dUOdC!Cla{1eQeuo|$ezu#>xH3Rix9i)`7z5SHR-
z*BMiB6??b~9|I>BLHYOsXC&L~4F+?x*HF{k@R{(P8`L}7RE^?A?eSB7MR)IFdR4Wm
zaJ^V1<HDx1nl*b~m7h^pn9{#&*Nm63wA9o#-A=Hx1btSvH@j3ZvToUOD!{nT?QA<(
zJCb=tfV$b*@rBce_yZIUJxXZjuRF9+Tz77+_3VW~CDq-7R$@Ci`JPs>>bUyRER)hR
zJp|+)$MJ?4UOPnQ*<pa&Nq!J27#zEz8G-!8sK31o`)jEi>bN>m$94gQ^tRJ??K0we
zrs_~c-Tf@|1#jJe3ehNnpz+}xu4*@>Y9MXZit|ao^Mu`>GGf&GJnfa2ok|LRGxeQm
zhIbKEs3IfuJ>}Ln7;mL)SfXg-4^FNs0wU+Cgn;0Uq(@a|pJ#d`m$s^(kkU58!NIFw
zS^`IExy##3c|(9HxzKKeW(<@4ONJ^2*0=1^#;YpLY22@Rf9McWoJx$^U&t5|$Hk8X
z;fg*!0bljFEQo52rvr&$cemeM@h@_9#A@q1d=`4t)A_s}ZD8!u<UJZsHtT6<U0T-|
zEk<Qo83QNdux(^oYbJ!VB7#i%87k(%7$b8B4?sDavE3Ing&*a7b^K9D<Zs>*Tnt{d
z1r);al;{zabhr-!y_aO3ZlCukGkbn3!JzcBI)CnSvNI9K^;H;p;y}T30U9@l4^24E
zAK$cmgPP!_Y%q~xIZs5<tyvZ{NZlA9U&<%8v@X%dp-GKC9&Q%(y@qYjrUy1c_kzDm
z>+mO2K@*mumn5vw5oF3fOnbiVb+~V@%m~D`8K$oUUAxwb>ukKyC@eYuk%!2s(ED32
zxM4=KNd+W*1q3LMXMsgTQl-p=m<+2is1|z&DAb;8G|d08i4UMO$mSj?kjvvNbO|Tn
z4;EqYUrpi)q<Z@pTNbv%+!0>3HqPwyT6oacDijpXnMCXxckH)o4JLdLn25O+y<4H5
zv+Nwx?fw;ApYFPy+4WOl=0#7&w)AS^@8C^%+ZC!<Ug~)~k8h|xu75EZG_(QenN-WJ
z4Rdc4e$X2>ABiLc%jQ>mJ<iD{y4TDBzC^QwzBs8ZvIINYD=dve%8fxv^Ely!xg;uZ
zf?V@S!FdzcUyf)fn)+tlA&R3+_y{>yFfH}<W4sB_HOl{SIe?UkSMFD!oNRX<rDO^E
z1H_%TQBGU@`m|M@ev*T%60QH>SW-n3IQnAsbh*_tumPP56^Badt2*0AkFK4rHxT|8
zBCG4egAQy1A)|P?+uNhBLw&0tJlre2CFSitQ48vp0nu2)h4NspQ5d5wrRDCLTPn|>
z#9Oat(FzS4+c$%|s?9ynbBLHs!D53@h}Tg{aq??Lo2t4Id{9V9C4=CX<h9_Fjh&_z
z8<K$vOV!~yz*V@GZ&$19+rHwEkUKQ5+copSOZ7BR=P>Bmk@>`*Yy37`b=_kQ*9Jr7
zWN^x5E0ZN$GtKp=y+|8x_oBV1otK#|Ka9Fl5n#`0W*()7e({F+!D5HJaO&RHol=ab
z5FT6SL)MGY3mkJC9wyFH9C`wse%RvQX&sZ$Z4k}8dOw}ZbSAy$=?=PL-SmgsX=Qcg
zL}lpwW$2EAyBI@;pE)9&E_>%v|0e$?G6NS{&N}p)_E)9;PG2b&el)A8Nd48y2nwsw
zz}svKGeL!sT_4N1bmmAg*~QLAFPf&~%rS_qFoLyD3#=Qvd?o3sFN$Q;CAqMoYQ9~Y
zK<a0fGSS}*`L*<wBd#t<4!z!7V9B1ok<q2!O-+_JBN*+U`8?9=+^Ar$98eJdo(HGm
z!zLDEIB%pcr5u$Ihr5j|Mv4XQmfK0PY5zu*z0U`Binb>HeLf!1^4iNU7^G3F_cYpT
zM~99NoRl&)<&fyF&U~aFgvJY4tPUqOi&mJaIArKn&D&ye2M1!JpMo}vy(eZUxRRA^
zo|_(Z<hx3o)rJY0UGIrTf8TJr&-y~Ly?G`ooyi-@imWwZcS+_7<8xrxOpJeH`UFo~
z^#!<rq=eToJ<fe*zCkKJXws6wpEPoAy`FlDg}bY6&hd-jQ)QJHU-L0jR=~c5Fl{F9
zAU<-qLWk0^T_)PxA|8z<+fX~XzZQ1Q3UvU)dnW$Iwdy*lHSLOI^6@*qjHvxgC$7|O
z?B41j+k3*3bb{Op{=%l$hfNR~N$$e#xNdyA`+_{@bi8&b%@<+b_^KB-a7Pu-U6{Ho
za$ZEBfATo9SBF6nS2z-ax?J&Fcw%_bTo-)8v?zKhn<DxNXFbr6#<EC>h)U|Sa`~!(
zW3zW(7>N|?8ob{>xr7b8M7(>xGH$j|O|3DiECR=&Y2Q5BINU{#cDyJORh<S-aeBsu
zXSyIOO+zMQ`sa~iUUzYm;0NJGbKd4WVirBET@>j*f2M!T2N9kKTxyzlsNvD`#l74P
z@ED~?+F=8e(&J<M&VQp4r}$dK_~YUi!WuaNQscEr9BCfVFs}NZ)VWV_c4=i*g8wr6
z(Uu=UXDqE@2Sq%zjU5F!9lZ0-vqvHz5JxGR09!L|U`L0rkt7%g#wVTa!NW3y>sv5T
zIh3e8a(;ftlS?x!W>|hezs&<qWR<VBuFWH}ihb#o6Gu0#+BwbJ^d#}#K`DmMnA8pt
zDol%tN?qB_Ncfq2v}<Ut^<Hr=K5$O<>~Q&S7Tr#`?_Rd6NoWWccD_{x>8x+e`A0hO
zY2J^}5-UkW-Sd#P?{DsVGr8dxg_f>!U9m4T@x}^}Wlsk6ufggc0Dn&yJz9@@<9@G#
zW6ZQO{5a&5*&Lv9PzL=2ruCFj@b3yfVyoaDLeA+Ws<E`KvX!@gL7_IpA7WQxw=n<)
ziqUO!YG*(CE3r~kYR(U?vD5-NknK>7zpmQM+&;vitQXSeRW<_V*?bM=qj2&U-k0~g
zD1B)HbeTE_OcIQdMP`_@wGn<O*i!6DA|ub)A92xLfEm)%3*D(l?I5_ZIFfYexTBlT
zJRnHWBE+29sIh2Lm(oeu+b4+o^YpeLb(Rn(@3Z0kx^Yma8*7NGz@zqbHM!-ImAO#7
zuJm<bf3yA(BF8FMJa29YyPf1<aq!}s3%R=U%Fu(=Pk3&hbzfJgAF0rY+XyWuX`I^F
zdAMj;yC)NLU*zniht{c%zC|>@Sb^|umrWZWUw&0!#ixdrmd=E^zMVER7l&ytha#<Y
z1P6QPP%Q8B<FEG<{M{k0+PgwoXxhwfASsV*V(}h$+g(9+(VG?cSnvG6{w#^inOury
zQe<wCIi*k-2!f&IgJ|_+(#05-RrZ@Mf|YBRsupm?JWU1wjBMxRJcz(zjNJ8sgTPEH
z)riOeo)?%b(h1iWq;e#@D3j%~muB)I9TLItk)6yDK7RW{c~vL+uUgy}@MpX5i92jc
zmY87Fznxkf9|X1aT8frh#{H$eCT|Qvg54E{vA9O5*4D>gbH+?w9$N+|e)21aGjBnH
zr6s7V&v0b&G*RyxixDb->0}U#`+?V6v})?VwLNV0grbs?Tv34m<-)6?YNRk+Qm2B-
zDk(t$4MXO765eEH7OALjtekHDs2sDnb|PJ!%%<IL3vNbgRTs8YkLmoRlGNQ(lO?#*
zusnA%^_2+V;+rb(NbEsmSqhYz@L;ltJPb+uf_shw<NZAFl>Co&GJ*{#C6EB2GEelF
zY;+by{Y4;_{)dc9$_tc`IiBN#-g<Q{59CrFg-+)wnt}k|6Py0t>-c-bPlg1M>Nu2t
z?~esbLHxJH8~X+Sg$(Eg;f21gssUxkH=<QXzWKeWy-vs)J^NZlqIa+V*8(gkWC%vM
zUN6SGa_iZMWk{n+c%|L3F$J&o&qxaH0%fzde*_hPTIb{F`Bgf=EXF5Dz3$R2)aoP}
zRUMO6q?V+Rm>|K3$yhrtHRT1yE1|JP8v;yQ0-q1US=?yna4}=%!)8&X4b{K2OJi34
z#Qy!Vasef?*OD+i`gbsuwUaGaId|fE3c2f1i1Roy*FNZ!U7H?2dl;WlPWskCFck^G
z&FyOgL6yyIAd<pwv0>CGjaJwAU-`=3Nu8IN2J5RIJoHFEcmdCb`W46Lz0jMZaTKWE
zjs<a!QES5h^rsVt+3oI)T9azcTJB_Vuo2*yMA=#yC6CCW_l8izKHJ`_Pn<C`esn7{
zEPWQgy@G!{-D@Zu5yeqF^ksx5FvOkItP#MEVp!c;-t9t(5B^pw>8eQ~!mnGTj;v+J
z7LYv_xKXP%Z}bfN+r9h7U$Be-En8N$OZz~CuPag6u(9_LqeRfdy?_?ntf0o$>t3ko
zyTb7^``{35NLy_ZSx-D>qQm=pIb+fOWMHR?4j+Kt(TGamUCcUPogTsuZM%15O$#fM
zywxJuY%q!$kO9lK!-s(1_~d%H>+liMyF@NOHg?DsDv9)SjQwvJF21qa0ju#2QJtj@
za5k3mjUSG2)-`;TlbFPAzm9^;{3?rH@Z0WXG^|f8TU1Ek<BV1NH^^5SCsbrxXm|rA
zwr{ivj<_unH<5O0d(QYs<)%N3@C9q<_7DD5#K*`Z1{eUgA)})>sgny8Ku4NfKP{pX
zGk;-I9RS>=Kn5~4IHJqX-BXAM@NdAgET)Y=PhZWhcC-tBz$BQZce$~lZ|BLe>_dna
zhPAm&gBpoV%Yzedrj)p0DjH#_XHp(T6xn1Y2@3N64n<SYpYe_)ZU7#Gx-PM6T+7y-
zHU6CQi>|sj7??$NxLkUjBJ03R4)F)iR5L+83+WPXn%8=&t|_9MNUMg84bJKMC6Trx
z*A+1g_~)K}Aw&cF2J%uqHTy>vK@;-BES%N#^JmAh36iD85*^G`(}^vNMaPIAtu2%v
z{TAhFwb>ZyNH@@X@%X+Z9HV(~IZ<$N(i)G$aRdhr*!qay{Ing+W~gYw<F~xEl^f<3
z!70N>T=@G{iL1?&%sFF4T{)j#+>A7z;w8YJc!0zgk(KsxWwZ)2_fJzUi0&m*@Ja>T
zlR-l9lmOJ@!1y{mf;-H9fBEATNy@k3^K#Qz2or?O3QjNg43auj4nB*Y-7{a0Njpiw
z>QYCKPy!_h|7y%itS4vQEXw)J#g};vsom#CLE$D)oPvplz{V7~uYC3C&%0!qil$oo
z@6<5JYhxJ+#5%iz>#sw`_&_JMZG01xY#7=4LqFO2^u3BKE#sYOJ=Q&grkq8(#Z(Go
z9zXSm?~S^>$GOVC4uPAyZ$yj#yDhg>ghTdah?b3(sowBv>DT=GqM1f=Ff#aiDD^9K
zoje~!yFa9TyibR_Ory50F?l_>+{2HXyNV1<%OW|TH1qFVW=;5i=FY}h_WTWT{5L~D
z6OA;F`UBgqcOe*H$HN=9n_YyXKaHeqRt!Ma9OeZ1br3@n)#5`YWxDyUsD2U$?3+xO
zXnw5c)=3Wga*>&=0*-O5SLplw?IMbW_?&I2<jxy`Ru6nT9FO{;qE7Zh<C0uox9K6U
z7g%+6C}3(od6tbV@F{=hZLK>n;IKXnH{4BO62j-hqGcX@l%AcriMI@k>NVr#VLh;>
z7F^WDrE{*lJjjg76N@DHGnUcV1Sf^h)~Sv(>irJ{xbRb?90NcAG8HBJ%GaoFsT4}~
z-!cjLMm2hGvHfi~Fb;zOs7Cy-&DA{kJxkE*>>rJwb{&nvbYUzUH}<`|Agpzml}1hq
z*IQdtozUs}<^jKUR`akcqdVfs3dDS%L%rSpxRe`~M)fs#!)|rb)Qw$7no|dW22E0H
zb$J*ub?HyRA^A2d0XGZcT(}=$ZO*C`@VK<(7$?r06yhz_VJac9)9peTN{2KiuPNU%
zp+5-gD(o>2?Al>N?duAEK$_WrqWv}`Fk82+A$$nO7Qb-T)7aS<v3=icHOd8B-}(Il
zS>8(SLp(on@B?)HjUjR3hTs)eS49Z*-xi@2c38x^RKo=#TQPzt3bay@oxntol0P8Y
zR7jW`F_D^v9~3>DdLsZw;6`LwTn;AGGv+2JSh27^OM_ns*Fv%8iMEVc_I@=;MP)uP
z<M94iNezt@%MiT)DL0pq?aJC^cHHENU$aHmZ`~RHf-=7Iqlo(moWfjk6#q5cND>)U
zP|yPlz?ga{%Ifh6<`XGDw4Q^0ZWkXqE!K6yJ`hJsvR-pr5aiEx%(ujTv@f#5Dw4=m
zpuT&Ow=Tzij~y`71j*jDh~ghU`&gpy`xvRxML#)IgfSQ?ixa3$?-L|&ScMG^^|$U)
z6OncXIxQKlvHj7l6{>7)WA#d^S`&QnczB1v&>GJOU@eB0PWlG;IXd=V10{J(?G17R
zb!K1CVqFKVabBx?R8zG@eMY|n;c<v9H-E}s)0#u>sM;TaLDH4PEqRt5w+49IBM(la
z`S&EB3R?EoD!@KuKEmuAQ$2|&t{v?e-_h+J?%He_dIsxq7|adK(pi7hT&Yr(96EaX
zqOS$2q+#r}wb#>L`6%n)`2L1irBxFcdrnU0B{AL%Dd#qt5p3~2h$PZ}D0m-zIw9v5
zUY{gdwUu;KmfKG86WX1PIb{$W-43`5aT)qObWndvWvicy?TZ8d<>)7b3$$h}!`~55
z!#%h{M05}9Qa!pmo5*)Y%;N|&^{o;0c!t&a@N{nUpH;WiMZ>x9(0LuToaxago|9W*
z_$uab2`f&rG)A9)q?<EL9savX1iXm<T_pVRl)LBrQ%&p(zkdd==3zc^+I~q^<uwE%
zvuBgL^3IlHXz0HFtPu5|OQOr6ceE2Qcm>zeZHF+KRO;Eu1i}v3OtPqggHO~Sw$KOH
zJOH%gx8^`LKlDr9+j^`?eO}a^4t^05X>^-&{X)ujT1u?F3fkzVC3e^Y+xE;k@s&q1
z>}wl^9IhlGB}`;_GWcCGikckWwFZL}r>8oJ<yu|(QG6Vi+Rz`qFL$OS&<dHzBah+0
zy>uj9^60X-FW}ea4!NA88<r1IvmA$+s0@SbOsM3$J?KePdA(lMQS8Qx*}UF;5ggqv
zm?dK3zJ-&5YrEAJR5p6z!5KlhgGEOl8I{>^R?VkiDo=9u8ucZEGwIZHJ>1URBB|Br
z2gJ%SO4|AaCZVkJ%|>{<Xe<8_xgPg>@RoPBh{^bQ$stBsZ(I!t<G@d3){>8PD(0oi
zpNUu)5qHlklP(uxYw00_BYPb2y$B_}7$Xjpg{JVtt_{`KJ9!i~>FK1=-ctp|`-@5B
z<%ym<s0k`Aen%@8J6h^SWiiR(79DZ#w6U%&DX=P4*0o>N+2rtE{kcVj{P4bKnm(`P
zcOZE_1s2ByRythDd>U*%@sE<SDz&vg-0=-*`i8~=ZLkRvMD~lUqqyQEuib_*8H58s
zTlDnkL~C&kYF3Pl?u{ULy$B4bU(%OazZqBRfpOQ))XQpgLsb1=`{%r7nYsj=_z)^T
z(DEygXg`6KdD9JVZbf74I)*ANp1b1#-9>}9Ct6Y8CaGUmRsicyD3zZTc@IZ1#3rFg
zA3hKv=i-rw^K8y9Aah%cZt)nE@eLo;u3OLWdh%#o2|N?kqxBhq5LMOm#v5f0ROx)p
z8Y^jR=d<4$NBt_nfQvW`nAU1E4x{4Q@0mY0o7eu}z6h8R$kkpy@*xf6)DsPXr4=i;
zHW9A&o!6Wrt>EjvJ$snL^^=0I6(pHhldOO$JDR<FD+POhr}2jQ<XAU39h+g9*&vq&
zNL)q~2vV7$K##i_E(fh%Uxe;=ZuAtH0e*K;{xd4LPb|Zo4j&a16pR6A|1-#G2}&mQ
zoJ-M-L!MLKof~tcCq3<=Q&k2F>+o6t-2eIFn*-V769skSf-mIv71Z&6)%8B`zu6_z
z5T6tMrRwaCAce7PPQ#kOxXPEC5^IEe=ErvHVy>@>FY}E^YbwXHRC}`R9*i>sc)YU;
zXWK1jr((fd^x-~-5J26I-0mZK065DCH15u;ss9V5Utfgty!u_Fi@dotlVI=Fv2{SQ
z5KNHE+{#*|Mn!$~z*V;|8nvz!nCzBKjh-{Hiqy_<8mEF+&vCxZvuEOFbT1K3FOH+0
z9v5f?vev?$xh90vC?aje7l||NP6<$z-FWy!((Emi8nV8fLX{zqDFm|yCLrJ%2MI+Z
z1{wl=_=Mmm1jJgo>@rPy?Ft&v2_bxskUS&ap!z1uH7&HahreThRPhfm)rU8VUy6lV
z0Z7~y+WxMwqa{KnkUZj!zuff2oI2*@v!Bha@pO9(kGo&W+=XH+evc$IzvgOYkq-Y{
z@|8;JiqvK7)>O}hN}dNtYGn@qTL|8MR&?#iv2)V2uepZe7?RnD8`r_$^uGtcLP9(l
zYD>@!iE&lVN=49+7CB!cY%L*SZU7fFM>!X`B)DPi6lSmaL;4+UIF*+Nqk^vC)=pWn
zHQ!PBLX;?B1h4+7U!r&fc`9f#PxM0=lK{1ZqmkH_j4V&daB>zSP9%yDqK#idiCM*s
zu(`i{tUok(0k+NYG?>qM7ZPWefvS2wifx!+yw73B{nMVzHgbl{%i~4Aijo?k8C%NX
zvD#5=uKub;)+CV5<9nFkj6ZO~SrV+q%pw?-ThQSv$)ovvliRB^&Vb;s()tN)b(K9k
z+Mwjsetlf)kRSXLR1A}V#7ZZ2YHm0bIU6Kx5oFW%i~EX;b5cK6%L>Mmh2cf{^_LE3
zLABp=qf2i|E1SN=8Ix(C7h05@D{Sz(TE?j2e^7FBD?OK5y1#tWG-n}ZYHzl9YxheU
z=V?Y@C@N^|Z|6jtmoD#e2VuF0@)E%}>AX2<ujphS6Jag(vQ9$TV|ThE=U>lIALFkf
zxd`v2xlWcsyr#{(p#ZQ#E@iQF!;}yNSoiF`?mN$^HxYnNqrU2Xv76@FX{yOad}IiM
zEH2+j$>TvZ;vs+IJlL&^h6Bk5U_U+*fter^v<~$1s2+Tyli!@ytUbYvI5v8RGE|}I
zm@X8zpOH9+s*f#ak@tD!fZ<*_Bo^ZKB?aHnk}sB&XgPa8wR)@dR*{y~!aB=}E$(Y+
z#lK+Zk)WaCn93-E$eZTAFAFJr$g`E1HES=WPqrriu(-=9r2lb9v?_88^XaMEkK0BM
z;8x%%H%vkb3Ri9nHU}J{fVIrnwK%R(jB6~jKYUoG5a%=4=)lA3#V;hH2rTTt=bHx)
zc4gv(pw(K$6xxH6ejm8?p~oERm<~zPz~BgRl@?|fq6yc|;3fm*Mca(XHVfkA`Dk-(
z?N4954d%*8hr_1E=-i+JfC^o+)`RANy=b<weSUqus%|ASXe)E|{dxh0>$<W?GdP82
zXYgRq*DW_@wd0{=qa>x<1=DI|1|((5k0Eag^*P97^&^L$D4v+M2laNv`1EGE98AwM
zA!sLnU>FE6OstfaD4~u+8o7MGm3m5i71C`_xGO+8L_fh{Chy=pkjQmdc7&0)xbT$`
zkrXksy=JJg<PEm<)e*^A(~5HiuFnvgj;5(5CMTaApeXHqhp}JFXQHFl9P2f?e;I#p
z*-B?wJo7&eX}qu9DeO1p_My6$bE~Cy?@xQbDzEBzHi;&a>Yp9XRv7g&Ks?g6rNQ!K
zZwmcjk0@+Gem~i9)(;$r88tIK<(kPxuZrV!7PdjV{22=s5oP@sZSpOT68P917+F4l
z0FVGmLT~XjNKQKQyQh3MLcl(6g!^0X0>hwof2?<IHR|o-W7nuYsZfH5NbOiBpy+E1
z<8Ln=5)c+?j2<JyzGXr*EQb}{46-DohBzgHtJsg~_H^6-fw6-qN+_*kgFt`qu7F<w
z&t%nbu^wTRRIU!?2{&oU+kDjR`%rcE+)@y#yioP38t@`<i1$P20|cH8fyv+5$!v#B
zZFP-IOke*wmef(xQxS|a4<J0}+CPT=uBt~j`No5yBvC*;ql@z|48G$ZJ#QOr$Fq0p
zhRW`I?1@kCt+RtA@9SF%!7I}{E%f{(vVm8Kks3~_&$43}%#y!Z-47>ywc<cY`p=+M
z!y~-c*BZS$)COXUQo?tk5x%OCty!GUGz9}>!EI%9n(A4CzOijCCiPaDZ{G03e%)Hd
zx{HS4<bF*)$_z&zbO6f+%pVlMKMdat<s^|Q5EY9zJ6L^m4q45K@SdBBdoJ*}0a3_-
zPP0ROG%ele$P=$|;aG%x_0`8{h{W{FE2u{uX0%I}c*XWp1+YHAlq#jxjF2-D=JH1L
zh!>50OlKf)MLMko6_7>fR?+?nKH>YX=9FXqX-@_=@LDgxDs0H@-ygr^OIwv-;l={w
z-R^jfJRf+2XupZ1;i(2p3QR5c>|MZ$611!+<xreM(aM_9m^zLOOU}T^nQFXU2P8m@
z501>Se>2(vv_WWaE=i=@_ukbt9i^?W396-DDWISYgytHr&6A&qma(kXl*2XgB{-TS
zVv9~^;>9>9o^&xPrwp9db*b8k1DE`+ac7?&QFBO-JZFQhpJK3#W0yys(X&QU4XuPB
zeu+3wY1<RPL7{BzmS=6W&_AgF0#QTxD;Lq1&o2Y<i$9dBs<gM1H=3ul*Ycr+Qh!l)
zp1xsWV9XfxU!K;+j*<2vA4qUsbig%2;iJ5J{-ssR;kqVyy2k*T8o`Bg_hJtCNqdLV
z58n=b%E)JsR9Iifg%q(oCY=7k>$o@U0sHef-@XWAIo0ay8{S1DKX#Z>@WVNtzYtva
zN;;x@*^NyJKdi|k%`7{;vyoD)x~IrRjfDk9<B_{$c9|@*iF-VAq*WxVC4)Bmlbk|(
zqL-&GmCH@*{z_yzz(EuD2(U1e?d^17a?|yD*Q`LR7&o*Esfz{)t;B~o&0R$Q{)i1J
z>s_|lU6KJ5NSKI!wTbm)f*tALEwC}Xhar~mb)@BO8hkbFu@CN76r<K;nj^rlkWpaN
zo&Ih_OH(OrQGx%e($yB!v@5mJ%#Y%m`GqfgGR?y5uh&hysBJs?-PmIkFSfecIo|J%
zV_gp7mrAAS-<0o5rxN=m=k6oq+V&Zzw)VlIT+-FU`I0q=#7~<_G_+(B`^JC$K&Ob7
z?^!C=%I+PapXn)vA8EPbgno!FkyI)Ag3t9N&H=boR`Z!H2U2>hy5&}#9+2B}D|tSn
z|Iv)5SIt89waMp(?nSW>9gS)V#OHUf9?kppU&mZTrDeA9L&j3&6OFURCqHIcGz<(#
z(RLQ***A~ELFN7f`P;h-fn}a+=9C*V02PeF!dy9Z!Usr%+1l0U4YjsZ<xIO@5>6lA
zx1;0xVCF?2N~=UhxXv3}%4vsNXC`Ba#L^BKehR`TdU6p<{Ri(oUOKcr2B~{LK4wx-
zm;7jVXKeB$>G$XZ78>OSo{huStoh{TZ3_C)a4r--52Ay<EY4l$_vY6PiTbtDIy%a*
zAzQEwf}W32ZzbkuA_7+HWAk=owru{Lr{|OKv|K^r?p8<d15~%#uB{`RzM_t&&~j@o
zb~`v8(${UD=v~sv0XCfWMxZq1<Uv5TsmF6u)Arh$)Cl<??_RUc(HnRBwEtZL5Fq9M
zeSMKe;X*EhLp^1f2_$i))VFra{f=omf6^M+p#vO07gs#8zgej5Hme^Utfqr2NcnP`
z8n0;rd)hpaJMfCWD|x1Z<HDDz|6Dq`Ej%#s>WMtcy)qs+;pTOOlL<WGo5dIe=visP
zXLNpDXbpux{8}S!w|-4);2ZjPJ`v*lxuz1`N6@LkPy_+ujdkQ-rf%#i8Bu7zBArCZ
zcEnnxhM8=)RyryOJschGKCP2IJsE&*>1vkTt`A|613u1ae<oe`#y@<sNDz4MLdP?I
zT!Oe23C78@;OcsX8ZaaMitrBklTkUXi#Isd#rc=WlopY86glN{Dhx{lS6Yq<I1L0x
zsK(dd(^V7;`#(%FRsPKmc(14`5Mgy~3(XShgszAl!dSTY_urq{9yLn{!7sY3SeoFf
zrzK5!?;_-5e4h_PZ391)7M#V)p*~KOC>O%%T2(cJpfqp!PJw+L|7kTeU$FgPV?agJ
zeIUxv>0y*hpZ)w4`DJXq9CqCOD#PAw{EE@FKlngjFpgN_olM%w0jmvcD06EFd~voz
zgxuwgXnpx}LRrO=d%1^cUhf5Qrjb364Xa1BAUoAvJk1~72G>K35GWIfl)XThU#i4}
z<9OAIgmg9x7f3>7Q<0~${->e^CDh*WXk1c}{xSF4(SR}Pw`b%jwB3{($^kv|-v#Oj
z2>l*wUp&Zr*V3BKI`jlpth2Z;STe^z-L3!7hDYY$e}{f3s!(-ZBA|mkUm7?ItXlDg
z9OE~R{Q_qKWPz;^#Nj5?Ba_PF8&y$t(5k34RP6z}2623D9NLcPBd3nEcYFXgi1fzp
zM{Yn-BRS|v>^J^;9QF|T7295oGF__+PSglhBj<;54ojt)e1K<cU%i1S>OHoO*!*Fx
z8~MIrfINq5Y0fLwDw<W~4|^?FCu&w!+*^0KNc=JEtHh2NQsU`U!sYA)O_dL=qApz2
zg(YtHejGb6{|Lc!3g68+M%S+=n{L$ufk=*;w3;uVn3{Qa$(ByNe48z=l>>c7dW0_t
zA`idrLk=v1*7hV@KQ{TuEAn=u-qV4j)!-wHX^^x&&3*0s^StRP4|`;gSZ<ucXc+N`
zaVveoOj>O4$`1Tvjjli|ha~9e65nRjxy~&Cfi-f~)$be|WB++s<r(>}%i5QcWHFnF
zHfPM^-y3`}=<MuVxL?HP)Bs|Y-OV_hsV;Tv>!keFP8Z=wp;Ncv6$uRj;sERQC{A`y
z>#f;rWzHU*>Au+IE}8Cf@oat_dqcq8*4`;XS^l#4<_1GzK}epPR7%9L<xR%2dh+_f
ztF&AU!FYYP?X@`!kuCE6OXmbf%RozUrL-a{Tecv|)=%2d#iTY@bhAtDK}=y8Fk0{`
z`tfA+0xf*?p=&s!K2|$Z?}_DYVrv?wf{}{=#d=?MzWbk#3$?LsE^~rUE;dA**<n=k
zKEsl9{kBw{{W8hZVkqWg2%^$adMAtr!MBQp(nh9pyEl%{5FVB7QPEWAg0TsoAI7vo
zrG4p9!-yWNFyO-<Jou#2`fhlnM7G~ZP}{!Ry#ty<6pbAF+Ef<nj9%DwRcBOq)ZJyG
z)ukt{@Pa1vTyI<l`VusdBL?TG_jnCaDi(J1S(9h2To+ucDsJH=2OzT<<o7fJZ4+2G
zuLnBcBTL!n_?GOhzx~syISw#}Y~-$1mb3=>|Mz&?M=C-L0+WA9bkZ_>2SdXNjpkR=
zd+tLAvWBKgnV9M1#Rl)HB66!{S_+9H)R2;Jpw$tRW?5!@m}Yd`0M{CKu?aa12Q(|K
zCx5IyV>3D9>{o6BKo!8ODfY3{ok5Fz>p8?;&5J*`AO<*;kl=@}I>LNy^1eOA9`uOK
z85u9dlL#b&8BsQ)!BPoG;QK}2XV-W)7sRzbXPa4e4e;uW6jr*rDW@|(;2vS5skNu}
zzLKC~7)5UG?c}!4R(5e25=~I8PafJ867%og5owx9Tsb;&r2DLw{~?iSUAFO~8pRkL
z7Ph;-okDegd(@1kqc*D88kNuK8w+Q115U9HFE8+9dE><5{-|xBQWADAI)_i{9>Z5%
z?Q<5~Geoy<PK(4=mM<++d85wAp_Z%805>YRV?8)LEh9<M1=w5j7E{XbBC$>%DEl_J
z1DKisL4XHLQC;zT+6vs4c}JJP)kaw0CY0c<K|K@WnVrQ~d=QGTTyxs`ng9L@%FA1R
z=f}EOlgl$G5ETxu&q`!}BB_=q{dkkQI;$CPLjaERV+{4G{-6**qV<e(BoMhxKAW7A
zsPxPAbk&PI=7VzAk+P<LvQ}}8n{PvuAz^!N@;d_D<X*HIrgUYTYS+v1)z5}(i?CO|
zIFsE+7eL+hN!%Si1<}9B71wfke|fW5C?Z_+Xh#3S@%FOaUGZh)ka#-&1$Er$`%?Gw
zSp+597gMAOm$vv>u#5d_Op4I!*IGqTOgSkj`|RZ+0Gr7<Bh8wl<eWaDzB`{94QGFm
zt<k~bMG;(bp43J_-WkYK{xRvKg-wuanvQsQ!N|ETJ*|?0W&LA){K;D8M7jmGs|rkA
zBpFJka<zblK`05U{mZYRMqCzd8^$G)fXUW-kz(u2k^4N-b~zF%j#fJy?QcxVUVl}D
zJ+^ROjS2Y!RpBd<?@7yscSAeaiGGj%upBk_p<k7gaJH>~E3mQo=P*c3hybIP6VT_R
zTtXiVT}i;#*=`l_ku`@_eIfZYG2gONG^srN9e?8rSkCKn`VvrU3lj+qppQvp0CCRX
zdLD6h9y*0`D$eU8=erMQ2YjZI-Vc7P2CRS-Z5{JhLeSvopZ9H5DP<Nwcb$KWzbOj`
z&^$eoi2ng<mk!G`-p>W^ZE(u_{Bbw2=aL;7!h5{3=`#+bf_(}SuT4lq@(}qNgeg~v
z-zJb0m231k2%Se@X6HuGm$dJgq=82vilk%&t>IJJ<7Hm6f65XGN!9S~-Yqm=?mMkE
zGLD2z1=&A+;6K<E=Z`PVp5^C~wIi0V%D8yG%zHATH)!tMyA_Wla=ohv)(Q-4W@LSH
z8of)_fY=d@nk|YHN=frAhSeBKMXYek`?CbfUM<SP<>Hnre-y?Y_HxdYCmmxLzk0A`
zJXX$ymzFji)BkG$*gmAvhH<$^7PWow)Xje)%b3axD)9_6{syB5X*$D}A?RCQ4DP<a
z(B9R!Ws|UlcrNSjUul8Yk$#-o6UP>go0=^d0OUvSGWj9?YF;P#-B8TO=6e#omTpJ(
z=rP}424N(`3zO36>y+MfFZ*k1RoyS^=%2c+2D-MNgDigDOq+ZHQbB$ANY57I`&XRm
zCY$WU&u5`eD$2<wwXUTj!e+}8vcDt-EzXf8`92^TM*NTH$zQx`Kn*LIRF05u+MdjU
zsp3q-;;s-scnIr@SOwt%x`|5PQy3s5kf059cB|QlI*lHRgGzgIrz_I2F#<voUU{B5
zz~i>xWrd$B^*stGB`!-$aDRWAH<<h={`S6i0pO$v2{=vwgHFpK>_$8T0t{i_Iz4P}
z@c{;>{|rw%!M**(`xfEeu;R<eATe*jnJbg}jCm(ZZ5x@DGj>e1pybLDS8eIpGD_DM
z380x6>2=b984@#r9#`D3+h3z}tebMN6;(L$+pAk55qtq){&YO<qylwnZWU(?!6lj4
zUuw%*(I24kD7-&4H8~aZOMA|z_*>q1Ci#iK|0o-}F&sW$Jv4U(@Fp@D0OCUhwF8U_
zsL3x}!2lgTR!cXrAQ8l#M9rgf;xIb(DRc`8OekedOWcK+!1w)PyC-yGe+ppwVSwf@
z!-HR4BOnv%51{@4g6FC!HuOe!B=>J}Z67#1ZGM<@=2vX(^2!RqU00x>Kx@p@P6NAr
z)@mt#2TAf6*L*R?w4UOXVMzE3DBQG6f}xj2GS*@PlEGj??>^09ND<1?m<2s-{cUj|
z@fNz4*Q*YHXOW8{>R${4mqWyl(s|9cr92PO$?P$?>>?A7VI5y{!CeXUku$Ucj4jZu
z-xZt1zd5R~0%FAN*K^qkX)rRhCl3|z-Thx@OW|#n-1GER4l5@AdOT8r)&C@JTRXS4
zOEo$_@nas=uW6R*I69*245QOS8R?gMNAY|+K^Mx4%_U83xP^;uAji4F(>P5^c#}<~
zB!5wKY-}ZM6yfd?=NvD|T7w=fO=ERj`vZT?OdBT0q^sTd2W>#aSp@kmAGfV2;mk-7
zP>T-P+WuBVLVptE7piK;#$ottfg3;`h;H=AcsiEVAU<&vQ^<4)M>hdi@XqV7TOTP{
zMo1IISk0^|a9bS2dJlV2gHE<&5~42YsL9i2&iUG4**RrT@Y4PS(rU6VYv}1II5#3j
zZ+ycPvO{7(HazP;UtQu)EZT~0%;{n1Ud?rccBRrkR@TTT(-;!@P<p7G2l^pQifB!Z
zdGyl98>g<+v@9318%jU;zzI`0-OMN8O3Z<zL*b2aJs_)F&D67keb{mP_(E5mF4|Nd
z#4(Qsbe88Pbrl@D=b%ZKLi~*1NwXx}Wgxw!`66COL&K-Zz7{wu8cyo*e;Pdy30pr-
zH~ZPBbH^*^mNZ>Aw$GDzP#^TdFsECWDexdHKfTELeK3)sc>ypK?u?*a;gG=B3ToHp
zt|W_8z2IFnXCrCxDah3l?KSO(jvq<q*!=-P#E_V70I-ln=nb>6SG8LOS?y5x`^rvf
zVpC1;$Pso2X;oi1m1Xw+;%-$*Y2l8`6oM<8Yrm^ZbH44l5}V+7*hER&e=i`C5SJ17
z!G<s5S<6drCE#p{dAg}tbA#35mklg}YH+6a+(Cv9pYY)Co{U}qL{0X*jN?oZik|$o
zGV9Fa6K>sLhY7BAl$f8M95i<0B&y%0FHC@=d1#7mp?H6)iQ{$g!M!^B`+FUCdZSxr
z>u70nkh`osLh@(^UzSK;z~B3Pkle&k?4L9n4GIw*>d>!wSLp^%<hn5D(N(VOCjpI8
z2oFU%<GQeVkhkYKoET1S?!I+P=@6G9FAOBdiyc4uXw^E0pDJy6#%LCPKY{Gon0|H!
zw+_`KWz+YYDt!A1z?OQaUMFS9^;IHrX+bwHcDKKuVc_6Nj^d739`z;T3!(;MI!0d4
zQ8O|WWtBssgd(E)pJrUc63=t@GRPH7H{A}9t(#Zp@=;j+O-zrx|0|*bbzpMu`=`m%
z_X(hJCO*_bGaK(X4_%$}A_7H}!(^nca}yW-`UJ=G1|{l|-1<pOby#geeFuQ|rtRML
zuy}@#oqHg$eTyYxw*k@M_4O6>n4Awq<rd${k;Hxjo9Xk>>%J3k9vXdhN>1BZj<#-+
z$7jZ~c|d94-<MNpJWX_~=A7aE{Mi<iEt=flhV(UWn|i4f{hZ2mJc8)2mF@vSR+;B+
zjII(}v_*%@@7sfcG2%fyf$T4$E2`%e<lLWP(6>PHu;}k{DiF(c?t?EK9<Q<j!_Vn&
z*lkm$S?mFbhW(c&;}^U5%+W4=fD)WUUsLn_9C*({U0(4}eI2Q)srJAoX_@1{Czv{l
z{{5NC`Yok?qPx=$b&-F>Uy3ZRt*KMksI7!YZ=yeezk+uCoMZfe>E+7#q6DrsEp_C%
zD<Y^jYLP8H_o*Q73b!}sXgc-pfG@Q|o?~Ctz|(5*M9*`Au1?LH&wc}MDxi1^C9<f)
z-enU0XcCy(`u15IxxiN{)dTVt5cqKmdQg5JzOS#|#$#m#I;JDy=_$3EZwhBYtoIDF
z8>uad5O8bNqJTr=?t#kVl2nKKn-ynS<~-TMX*AI5yvZK7&5;0se4$+F{|?RJmU`r_
z*px|uPdW0U`%Y^)zO^ksyz3XJC6NyV*Us6$M9x#tBx`+#;&w5b$O7aut=^SZN~|j(
zKE1L>xi`}DhA0c4y)fR4S^<vXp{Xm&Jk#|f2avi0n(N@aC$OkkqTc2Fv(wT`FQ$r}
z{Kc7@9!sevkRLTl*FZp45!Rf5=(DuOvgQkVX}aRLbghFFSu_=*xG>4_{Ft!`=_jV~
zc6hHdy(F%Yq0;dpZy~RGO*_T5w8lQ(M@CnwPge>eB!E$1YgA;y2TT5m>UUZwkEt7O
z#Lp+)xv`=P?nf9*a@ZN(mV$+w=tw;CQlVclwdiBfQ$>eb6;8H7x*(#1YQe3^(~ji0
zkxY2*ki9W5pvJPB+tyxE*Ua`}vTN5oQ1d@2R$mHwDLq782G9)lJgpcq9LT-gP#^uf
zYCc3CHI#vBF=yXZ(5+d!VLl!|3o8gkOU9P!&KoSt1^_nWW~bbTz^_|ClczIxUo8*T
zfCC=E;{CQi%gBo~I7?bSTx@`^QuJ-iY)KYQ<@Kp>0TV8-ws{*517j_4g^kRR**~Jz
zQZhbKoe48e=WO716|OV-s`D30JydIzVG1P>fDrsPcQB@IPQVONHEqd7Z3CEPeTmus
zj#D|9)#e$y7RZbP?8v!XXAmJKlRZ8SM<lBu7T`)%9MZx2WH6)0Bi9PSP-$kmuna{D
z&KKuMg<7a1-Do;MJNvBhF#E@4l>hPqhj<~mS%hEYm#vP~V*l21RN6DtpYCDI)wT;N
zoc*e8U@&L58cEI1uhkQP9T>Q=N(38IB!>=7`?ve`t=0HXIWUQl+$ea>j1Ia00=e~b
zA3<+Xp6q2U)j+`UoBxTF0S9o%42BQ+wpRP&7cO$C?(8jc0sg)}3*23tY!22kVN`LF
zss~x)w_dM(|He9!S0E-fJNJnK-iL*6ek((Rw;IdnoM)AAH{S}eH1oE57>UHipn_4B
zHBh9zl~b2943K{SW)d!3rhwzmJS;f$!)aGq{n?ct_m#R9k1fX;$<df#Ckq4{u07w9
zYR%q`x;Vq%YhR~7Ozzc){>eA4Px~k>$+FZ}Lz?z979HT?cf}&2Ir(tK6I#pP7yyp9
zZNNe0%=bd3b9SdiNj8p=esPQ7H#={}B$A-yoP)*c5}%KofI^-}azJFgI4g@I&AS5P
z!D(lSOT&wI8x)m;?q4<48;61U={Gr<_QciCLx}`iLg-%*326!lDXD`dk=s(4pyhjx
z>|mYrHPtvBB_k+rOfE{?jt^n^JWnxy&sv(^a`zH?t~fa~v9#imb&F=-9t@@|`ymR!
zMdQYL_lNS6Uh%x24(8p$DVVrEp;sGu!j!O1Wy}4d%!?wBb-y^zYTrZAuo{pbCmPO5
z&MF;e+1;741a=RrA*rZE`RL*S5x)u#aH<L)_-p@34H*9i8k_4@a3sA4{_@?4plekW
zEOhP?c_L8_T~YQ4BbQ94wORVW;enw<vxLu|)egTZ5(|_H-<p%W@yE7g$Gvb1JL?fJ
z{5EP{0BafKQdsM7R{oN(gz-(dt?R{-Tsz^Z^q{~N*|JtkIRt6oX{R~lfDGCbwV@L*
zXyKep@dhP|ntAnI9TPd8iHm9Xn3JN}{f?L+RMFx}Il|p~-NuaFaeO7}Sm21A+cE)m
zqx9NzBww9Lu$HX32Yru%&HC`2Y~bCRFGSSVXbIQ9UPu?v0JxLp>hJ`^AaL#?9)KCg
z+*hNq*et*F+%m0Ny7`Le@m9a&2FUTOT~|Zf09;F#SHT$)3MyfV-QPW1R(iypq>GRK
zO@s%qnozdExToU7#l5*Tju7*uIbx3>a5{AGWTH{U<r(lFkuLAe-hGAw8Wk>IeF*mG
zE-MvlKEO&7uL{z{e*X{xj&>{z%SBkBb6!6{-G+1T9RoD>svASINoca4HtyV;`Wz_)
z-aSw>O|FNppq~XhR!4HU@6i`NXW0$<_QE+%2wdJg=y`($K@nU=r}o|t=ci=;BoIkZ
z3W*jw;zYB#ux&U^{AKaaEx+b5e{vO>Wi1x$;Pn0MkPIX%l<@0!ou2FpTz(-QkR8N5
z>gnuhHOWsCZ0iI6+HvcQv-|PK7lK|`>Sy)k!S-H!6!KBD^D~u{!x(*+u7l@v9|Qej
zZ#6Wb&dZw=h6r5F(NB<MdPi|X17fwe5FcuZ-WWyb#?`Ul?M0O<M(HCPPX#Fx<@(?n
zzFmGe&sKK-3(%YG4RkU!$hiLSF9@PtIkyDFbQEhFF}l+=2uqhU@vCNjou5nlo@PXV
zDf{Gu_`4CsY?<3Fzopc#ZN-4xml(9j!b|gGB3s$EFcWA(V5&-(Ch<9?k1TwNHyj*;
zmR7`|1E3U{2xU!%nRd&dO(h}r_CUEb_7?H&=D%Yh5_AcpW*`!?-r^G)*ET`d8Y(}a
zD?C?+aCCVtub7WXPq(J{mhs4C^A-6^kJxm68$o=k)3DX(zvA$6>q`auXM@w1Xs(H=
zU{41F)0O@BFogK?`8%Mg#Hh}FT<+Bi01OL6rC;@1XOalU#!KZnYs?-G0KjdO^jBTo
zlwiRbj)v{&Fgt%tEkm6<7)Lk$;A^)c=gg!{#A-e2Yr~|HiJ+g9q9As!eGhPjAQ6vX
zbKi?lQYO6ZNZoU4N0>qYUY%HYzlj}$ma^|W$ufb#_I(^Uq>KHkfXY8HCzh4l@-dgA
zYrdbH)@$KF`<(lc$1dw}*-70OlQD|;;9b1kbLg`x6uUGS*uTzwC4tdUOUo&d2N0Q#
zh^HE*=aI|E4m#mD4b|d4fgV-OKWV<n{vS<e84z{%ynT=sNu|3HN$KvA?hfhhZVBl|
zx<R_TyIYB+yO%~f_W5%Ep1<#WA?ofqXJ)SX%v`TLZyM5>Y+JB7&zfP04&~6oPXL&1
z^;1G5L>3TPX&Ss4=hos{_ap);{5Wqs&cWyL8cjv&oY?Eg*iO`6?QW81b65c^wvY9`
zhr@hM_e8)Dq;N<Xf-tbNqi8s43rHoO8?nL#<TI+B{8*o}nypda#v~cSS#|bS0YWC}
z&yI^0k-`paNBpiHd4GkpdE_OU&>1JX`xd{F+R;81Cs&piC~fj*2@4~P^6+n+v%L@u
zg(hvIp^qVIA-Z)Ys50;B3Dkn^fnI!n(?e!P%J_4$Ji6yi<;dtHzh}Bp{HLz-Cq8!)
zk7!T4ja-g+nfJ}G*CNF_AI8a=;L(b7#ap!poriS=5i}`7K<(>;6N|MsD^+3XQn3M_
z^e3JIHzJ$ox?`oRaW$H<CV&dBm8pe!)OlFo(qc#_ujCn1ukWFChz<NZn6d)LKp%2a
zM8vC1@GSUiL=jeY+_pc|Had-1A(}WO=cMz?*Hz4kDEJ%-s5QNbT0L?=_3&bD?T9j-
z=QXV6PWuKz?8j7v>I~w&8qRfZu{kh4p%vV{x>4}r-=Ej?0W<>@Wfi&0WWxs)M;?@F
z<YRPp_DapNo&WNir~f_g_v0<Jc6{e%uZCWN_3&B{IlkrZsDitd`yQ?_!;b3GTNNW@
zt8rXe069o3=Q9~+&15vTz^lc6$Y2AHEf<Yw*;Nu%zoH440?OrKJ>#p16y;<os(rVR
zW>z5&{kTm9jO9B>$jeH|2wDN_w<LJTX)?A)3<ak3`v9aWiRTLdVtrTMWd?F#_Bo!@
z7(^K`eRgL)tzCIRcLc-karW7UXnaA^b<^79-&e@KEY4#!q!Iu@^iQ1-+^G%vZoG3$
z9sQiM4vOAjtslgc?1eKmkl}St1PP^m-)C0V_KbA`Bq}m=uM|{e*iY!<^t9L~UAMgB
zS_(Gig45T01ax#Iv~Cy}aPj&7)<r|(;?~EL(>x)mlqU7I)RZHmKCg;(PZJaMqh}}A
z&trGxD;AnqJ>eq4aqci3(=kl9=};w-3AC|@Msv%OOKLz7!mb+PPyCwt_te^>yq`Y|
zkc8#bOG_rE4MtaY1ncWXo<6vsC>U#(J7#vHG2EY@-;~l#O$+obaA1+OJ)A|$&kuq;
z$2E0ZVss@j92aOM2L64w`im>-grnRIt*9FYE++Ab!BAN-?lfv!IseQ+nYP4lb`f*M
zENR6SUgOqV=>q31D=+A#df46#s-4}u7t$P(3r;ZFjRyO~Jdl@$N%F$qnn~9kC8Akc
z!jedo2}lG@!mC@K*2br+{v-#}U8(^KnHGE7T$^j%W-82(zj_x)I5PVU%2goWFraV=
z=B3Ke^J?B-dIzrwihA2ZBVtxQ7ms(YrnpgU>@%1fs9|#`fmhR$n4?NmR@G?UlQWsD
z_KE}6FsCDY$vE~!BU%zpihRa9@DBFvmip}*Q}4kOccYL2#x^diX)3V~`FalJ0lzjS
z9_Au7B#pDB1c&YeEdo`0&!0vxOvs*>GuA$I?ro%iUdLa0B+;7o!2{a?dQjh$nHAhx
zxHkzi#2lDK9GHbeCN*qKu;~Bk_%fi*ix<x{=_YWckLyI=XwlBFcEAuI_>UyyyFM#7
z{O~bnL0Z6_3g?fkt%W&$k$k?iJljr8za9)Pcx^e?EyofO<&TJc{*!4<`nK|n75cyB
zl}=;0A6Ig&cS37(g}PB~AtP1AK8XvT<|3Tt^2uS#vB=0U=$B8~*YPK(&0j*1IbJ5s
z)%?#dVmfAy<l!aUD2DJc6p>eGt;gK1Bo;KD8`t)J#(Z3Ps$#=e#2yOx>Rw<nPRg{W
z)>VW1^r%&T<2|G!czwUt;rLJHQQS>}MNw}(HJ-&c!>2-NplU51)F!i?q-lC~>_xj-
zTUX|=GiAQS|4(ISG*d8jZd2W7<G{3P?eY`J0sR%XovN&$nVa__WHc}cW5xX<uH@=x
zZr}F36Jg-p-SE^{kMQvDUH1;_JsOP&-nrPx&Nd08KNTgYR^OE$*3&EC6OT-*5!jfX
z(FyIFUFG614tM0}qb(H6reJGc*`^C=wI;9%G7@IW01Z6u#$vnj0SztsuD<&q-q)U`
zefUGSL|SzGP?V$;IH1EMo?28{XP#z&iWCJehHcj-=PrQs&o>E+73yy6#MvMro&y?Q
z0n+?1)8qLty|x|gS7a*~%kW7IwT^VJOohbvO0G$<G2y2h9nRL|A~t*?vh-H9*<(gn
zwJHZ6r8%TUCJqDZyhE$3E2BztbrU{U3k>x?t?!;zp>g<bHzu{0E$X;_f(wbsvjv-;
z>aRgCz`>aT!_{G*57xpuS6l6{{xZM({aM;u+0wo?M?Kv>I`Y-ZuS)$l;m2~S>K|lE
z{J4K-6}FcUFp?G9wBET3vW0wW<uS~A-&1vARdv>y!@cqsNnQLnI6UCdKSkT53Y<7-
zo~Mrgb9O#x9VL22Z6&q&H$}?rl1#|lVS8JX7OCp{{mAv{KP(N=mpJADJRi@6x{$p}
z)K6^aKt2s_s)U>14$}^w`;T2`*MZY^D$?lOoOX&m^4bRXI_0qBtC)R>lZS(tVg2mG
zqfI#Iwn)V)7QB?gV#j~i?Ef~Ai8#IQ`<y__3T(B0s5hjBl9iQ}$9wmV<i;mZ*xHV?
zo=$N-tLYe`lkV@_B#3WnKz)^e&i*pPW5ZF{1rs-w6^2eMTs-4ugZDMmJ;ryuV@GCD
z#T6GF9fQ^-xenNJzQg&7r3Czw+(=`K3?v)a12N4#>nEMf7C3<&rV<nIAvF$#uR)yx
z8k_W6EuCRp)|8cj+YOUy(M+@A=RVFBWWP^4X(Tw3Pc7?|?31XTlhS2n3WlsQ2teTx
z|Clf%Y5$b^d)M$}kRWR~fA3pjU>6rXDdQWM4zcDwqoo9F#zs=RGO?Bh%l!i>!zPX{
zxekU<wnKuPR|C&a0~3MoQ1#4?6RRBE!;(CjDE8}YcW*%iwyF=~OX@Vs4<BNyPYCC@
zhuU79wu}YMCj;DfqQE%IiTakEMrLOBk-$Qd7h2Cv86wqC&#=f+CrmmSbWN)d6UD-Y
zMB|AbIlwFQ+(-MHe&5i*#o<)E^BhFQ3nTj4b0oyT6q-vKJX15fJg`{!aENV2uej7&
z;#r;sjSf2i^nL#w_q!4h?=>5j)7G`RY}jMKQo^3eWi32ATfYwsV6&pHsy;ZvKQrr!
zNKRF`J%Lrtshw;iJrl%V<rEyz?-67@Bk#jpF0?$a<Qb!QTHjvj*yB<=ZLemWd5|5}
z@w`4vX`5!Rw-snUU$>-%b|_OSr;Ep764UP<nj99lSi>G~{kRsDrW1%i*t^~5YQs1^
zkk@vIyZjT$QB<nt0W}%_k}o|+abiZ8*VD*=fclffceMs%Yq1b}%t+6=eiqb_#RndL
zyxByWIg$f~J}Xs6N^8)+&C!4KUDEVJj8^uF1lZIx(VIu8kOL4wNvl-d;8y`Zhs2J^
ze>J6ii7eoAo7$@MA}@^)9g+LCJ7fmN#^U@drKwmx?+a1RjVcdow&v!})KH1c&=ynU
z0C_4~e+xX`TJZy`H~S6q|848bRL*DEvcbE;pa886Er0d*gQ5fd&O9#)@oFPu(g_5I
z+Vkf)aK(Y1f7W^PE@YgENy(CHHaKdmjJK4Znm%iH#3OS;qMcHn%%qgC`~Zisgl&}7
zR3~VAQ{TP8t4d4YoQbuTf4hU%=g*GrunIZ5)?rNBSHG;tzdE`c>~OF^)fu~65X{C{
zPVR1M&m<tH!0;J1<sU)4W!zE@m6^wzu?z_4sW^+xR`y3CV}u5qYHU5*$os1OE2A|I
z&)$$}df7q>AoPZAKJm&vzvgfE{3(97-Dwx=wN`<QD65d&^KTc4uyl0q>G{`5<&=n+
z5GLiV2GOi($rrc7;Jps%7Hr>h=P0PvjR#sVxv%@23$o0$ch=-y+KtyguX_e<)<ab{
zTgjzHy;<z4y~cuTA9x7#@|=g;hkH>51{St;g1Rzh-={=@APCB^Ui*&aEy~ks93eWC
zI)RAIs`SL)UH^EQM2YLg*EHz%dRSOWIO=LXR2U*deoxlfAs?q_n*C3Z_YymxvaxO+
z$ge8svya7?mNU#ro!*K~_Lsa<PG@2RZkBej;it;Avgv`=uNnfe9|Q1+>NSa-x4lR6
zxU@8y5}ng$kS=BZcv0JX_Vll+s$*eeS8N~T-?V%xKD$2sP`<x@{7dZY;~`=4x8|DL
z0WAt0G0fu6CE_7m=3V=ZGNSx$pAMevDWZw0rBxfD4E7oJ&ekC0?eWR}aTYs28p_k+
z{enP&;fkMA8{$k*)ez1Gpia-38}Ib{Oye7lNk2iO!mpItDfV_3xrFMx6AFtGVfa^h
zlcSF{@QeGB@IzyL)2;VPUYn*~TTu0t-(L3Im-N9CE_5dMwhYt3jFN<Ctg=H>{~MNU
zz6;`Nx31!)=`{xWXakpYw4en?cgw%reDhoT*l4d`8ucnjOqeb_s(Q5gR1nNq1dXU<
znA(19R(JK?ok@!0JE9uJ>ND^dyd@lh6#mcOMgbqDC<@JnTB}W7zV$7M3jIP^oBUsR
zN1PXS&64T5lYjrxwl{&`AukrbDaqT<o3;+w%nbZy(+ByJXdbW-f^Aj3juif!?uLxx
zyJv~gxErw5SVl)i+~a_&5e!rLd(inkf`(lN(6C3S#Jq@EgP1tAZ>(`dyeh)oQ#E<Y
z;qFxY@mV|YWRxeIuQ-?wT0Eey;9UC=m-IoXbYAl+f3;_$mTII4huPv(4lss+>i%V8
zy17iScs88!UHsB)$hIYe7ohR|NnVruiL%G}y<g=(8`YavmHx$=l6R3VQ>~iS9aGlG
z_j%~NDL3oGVA~fNeA#Q$A)Ux<PqxB2I*?4n2*JXfU{(Ir!k_9FVz2^ak+IAMrP@f%
z1#UR7;VB&5gc<!j#2}KhuClM4Uep-i4}G$q#-$BrcPDazKxpTBi^?T0zL!l=X`#0(
zw$xrMzJGO2k%K2dJ%m3R%5V0}@|Wcm9rkU31DjPv(a}a~96CLY!4EdG4p9H^t%4f)
z?r;jSY_nsnUQ}R_Xb#RSUr^5=;|@`{q_tC8-5yyiKt1Y~L3X=1UkUsJ%+dTt8dq|R
zdeXtt2A)XL<J3<6`_@0tmnI|Vcb*-J{KKiCcLxO@*}Met)+Zh#WSaT5F`f||Xi>8T
zm?JG$%<g5!<gD4JF4(_NhMaEApL{64)tpsh(>_>8>+%-G+N_Pyl0D1NB6+Ufv5;eo
zJ@HPA2iq5n93AD&t-7Gd>^uoTD3{}4P1^sD1vu96V;fp(sG8+5Bg%+f6WW$x(xA>M
z_O?l)$~30vIw&AkXpp)v5T~j?M*JevRMURN%tV;J1DKY?$?c`dHu^;G4e4NDCh%qB
z^ikC4@}P_lj7|#tcS>2I>tX&g3lsh%mMzsC4zc#d*w|&N=KOGioje5O|I$bi?t-~R
z5{XNtWdRpb19J`7rJv?P%C@x?H6G5p1q>MIh4fzt{w@_jd)#h@{$PJq!zw;DF_>pr
zK=aB;LG_;ig3~818Nd;Sm8*=`dfZK^!kdB;on9NMIk>LD0){(A#E2f_L(K73{i7a#
zcIEXd{@A4r5MZJ1hkNfHn4b_XX1gyEGe-^DX|8V<dc3M~C~+1L_+c=tqyM!Ee?rG4
zUduT8{IP}4(k5~5Lpn^YC_I1#<&#^JWU!vY!(hW#24QO38S%7y9~R`o7y8mNp*A{K
z8C&N%B1t^|Gx9yip||J>{F@Y?mB}4x<qPV?N>ADi{+}*%@4Nnm0?x6mAVQwD5OXvD
zsnEo-E})Bp*_{He(_jxRqzjAqE)S}qP6xwk<LbEfqt|1|l=7{{lfCB>p;EYtYHf+g
z3wJ~X<Le)RjHOECk~n!vM6cvfV1G1i3T}E9&3FVS9<je|#FCqRek@LXn*;qn^bKn4
zCqmX!55m}UP@sa9JO|rh`k(PX>AD<Ar+E^!vjUVMukm`|^4@VhH5=L$+e06Fmj?n3
zsvOBeD*B5}!FSSqU$;aG^C*<uY(*JH5R^n@<%Lg{i2_BRi5cC_2SWzM2X;TrR(|~i
z{cG(zdqqrH50wXGO5U>XZi*E_FjscitC$<e46m-X6z=KHHT;h6jz_Hr%jvA#l5g98
z(IbAYgmL>iO0bYmjHW1T@Z_Y2+}*!WUY#L*Hd<4#4ds67$30{EKz@g_9(6QMWXp7e
zd)+)&o|bh@F>G~}ZYp0&Hfkp*Lm@#?t+S`Lz893_qyKIEa^<G2$%O`Z%6j|!svA5U
zXAT$Z7x{NpvLW}_UC~V2iVi3oYPs&7xFEA(7Jix^(eY|N8Hxix(0PR;dQzmtTU_MJ
zMXJms2cdyN%WlCFE&RB*L479`JUpXy&86sBI-51n81Nxp=9lNrVM>R~y>T6WM+$7#
zmA@cX)pyHfwW!W>j5%XTrMlQ3<P?;Y<VCI*+4pQf&du^C<o`QyNTWs^G;#R3R9Kfe
zGGtcU;_t#^)v15^JA(lf>s))`Q0$xt?1D8~+C*bblTzlf65@NvFjx(M!O6AsG)>yd
z^=Ba+RS=vC=P&cG9;b3#E^6qdg+D11>ufNwP<CZ0g?H0;kvis%<hLuM%SxitkLi|^
zPg@G&4p676)&?Fn7X4FG`-%P_{iC~N@O5HV-tUU)7s#=n$g*Yn4gBBTZbjZ_ri(?$
zHGCzrLsW3>NYmdUa?L3p#DKN?@)Arxn3^;md{4%}me*rLztW%I$NmE$e^lN#qdQXu
zQxYVozFoFCX3Gks7~n-M&p7Yf=Zw2{2>(S0B@#;~1?ygp$8?^cgQJ_v=J77^7LkQ%
z<W5V0B8Yn!n`2Ry^TP82bg3d)D!xCf9-^CANks%?m}xUii9ux9Utw3s%UY5K`C2&B
z4p{9#?40(Wc%oewaEa`l%#2lF_w7!5BtM6AET3`l_9^&4{xPEMRfp<0ireY>_KhFj
zO;wGFKATQ|dDd-NIK52D^~gSUK;~@77K<PE7`y;Ziu}f=KTL25ZDDVJhRaFny>N&A
z3tYLC>G(5-adfVK;rX$xy#+Yyw_N(lr~^=nAaiU?7xmU>(Bl-?tOtOHABcZF8$F?F
zGIqd{4SG8XFBJ*ql_h^FAmooqUFIw5Vj_CQ)jBvhXbq}wt(+VTi>6F8s=#=jxih7<
z=}58h7B~|p9D@5yGl=H=q!lL9^VO8DifO-%U!URgb7xR8Gnz|?c!;&8YRk_{$94YV
z^h8SQ57XvNTMzt|@o;h+dEfnP6s2*4zBN1r>#06e{33Qf``Dn$dBD3yTr{IKq#&?h
zMMoFg)E$$dZLoL(YWgByWaBwg9vWyxx^(Ax#uODz6fjdnq@0m_=f?~jS#A~Pk5GV}
z!C6TeOX5%)oE|bRZ`m&_t_?3N_Gy`#wIdPzCRkI5KEN036$x(8k9_^M&zK~NkXM9H
z_Vs<Ovqmt-X+s1vOv!_`V%?mYbauZxm5H%#*+lQLjhCbeREwJOUG(9=YPN#YXtLyU
zwkSHRqS~#h7yKVnwY{JdIQlQ^19h+VGaUC~?UiwHu-<M*+2#1vY6yNQ&F|XttF=is
zGp_$0b;@39v(wz34R6eHswh_(kQe((y$1?_=}0_xr-|%XJ;xbpu6Xp?v<;HUo}YDV
zK1tUFdB#(*BoZ58^7=W6Y9UnOU`_Roa1ry*l{Su1(F8p;hx(Js3@K`f!^olgxIFvC
zeT{ZV0e3OJ#qA-G{u2_9qMgdN@(b4T7~J%g-t->r;RMnv!zb$&^XYDwn8uQKcf@vT
zka~1?qjjLltwB^4s*aH4HDf7aP7n+IHSTo^Y+I-n?1lmYtY}Kj95sO!&61IhOUGzh
zbX;n%hauB!{Z3jokwV_2^bc|BUQ4otqgJ>}>W9t9UsO7{-Ivv>UVo0Jns6Lu0j)=~
zD-?(ItfbmwUM6wlfT4ky&Qa|GNTg%tb}^~pSefexI>~vwI;%;_rb;2GZxt9_Qx2%A
z)?1}|Gs}46n2x+f@P<xr{{-c;TQY$>E=S72pR=+4iHXM|=_|P^k^>5+8-4yt!qwQF
z8x0^0LIS?o*aPr_mzUD@b#59aROI`ARQtk-`KxN4VP{)M#M3wL&qtEU>OIdkZ3M<1
zYW*hh0hjUZ0i}~_o}q|@NoZg+#xD$n%4^w!9z4UWyWtn3UDhe+7yY?lrpVy^@$8OZ
zP68}cei-`4NNrVMr%W`kRf3X>ewV<5=rr(V<2(wTI;1A|e0+k3BL_nKlCQt7fGH+1
z^lvBr7WB0U$&c8ph3Q(b?L$(#D!d9qDIx0ZnP8F@o(I1MnFgz%KNZ24n@Z)PB}u|W
z2$K2i=Q&y2RN&Wr-+#N`bFp_RAZ@s&)ADn5$~ACgMNN|UUv*h)Z93A_KJ79PC@wt5
zB)G!rG6o;w4%qR>NyOZgjoSJs(^V-Pc%Z#n!!$t5h<a~s_~<osov%O@;@mD!ppIB2
zWCFNHf9-(OKXA3vT0`=hE8C}&9t`NOF|AyVY|}0sza;NE_nf035d^2LLGw$xT{q{?
z91do_Lc%BWIG%ge%Ny&{+^VW4`D2S${+Ft~{6JZ{d%ovUqeItwn_{krdMe59gV+e%
zJv#haVA|sz7M|oam1+FO-UiIUo<Q2PxL?-lv3F7YbBIQ>AZ`9dse0O$c4e}WH``2-
zV(o<ntz#HC_Y-F<1*7=Ya8sDCbx&9V+^>mS#<BA){y+V!R7=5VSicM+7Mxw{+!??_
zxGz3OME*b-ecxz7v9Zqt68I`$<XwFt{aa=a=?JBEI2<(yoh8iK=;^6qq?`g6snUh4
zt-A1q*kZhv!+-^nfvh#MJYkxZFc$`wfHwe89t-?^R-yVq%-qfx-Z#SfoprAVT%o_X
zLW|GQpoBRF<g<m<>_2gjrvmvl;06ElKS@;5?YpK6lX~(!=VWDpfr&XV7Z(xO@f~HX
zZJu^XYuSsI1GUzhsSx9URE$l}sK2z$!wKTu;p4%`9^sv9qp0e>e*8%4YlRF85b>+E
zPgGr_I%gd6LO^`Fd~Uxs2)8A@izoh@o~B2#-A1JF)Y66>BgkZjj#j8=>v1N>vTIAs
zLK;2ogaa%bqg`pX==JwcLrnR2Kyz-~ey$WeLVI^07G%v=>bdLxN$t<tZ}njo?=BDk
zY<94xr|_;#%HX*toYVLI9-~|`szt<XjUoS8{Ums+X8Q09IC0tzhpw65cQh?_;xtX^
zDczn4^SdR<dQ#PZ4(^9eZk1>LDu?lVQ80l(t-LjCnQ%ytJUs2ezY`Xu$V<6_AJ^Q_
z8m6<XE0$KCfT;l&-js^XuC#988^u}gLlDJj;;xATR=+b8x4}bm=ZnyD@uz^(&hd5g
zVR1k8O!EnSI)AcHgX}FQg+DiszXZ#?{+<DJL)j@Y*^h>R8cyihuT2Q?!uDLFW4K4#
ztSLBIi*~+k=MsGAncV!9Zk1WTK5%nHm=yHXdo<Ue92ptsuM@MAGo5>~?DzLt8I?qA
zbWY_5BdNXi0gaO3MVt5CX68e>`9=5Z!`et>7pIFnnHF-#PZ_cXH0eLSx+R}vL!o$@
zVjH@Xr@dbKt!vtSrM~7*eH&2wz4DN^dH0wzICmSy{{%aCfQ9Z+>x&BBD|i#{-0)?g
z5D`fu!jABy3Wn#8@neNbPH9_M_C)es;f;$L@edXT9eC=2GQ!!}*Sf)L=W3UQa@mNF
z*g&T(aSLpBgZK)%)EQNWuk4J6fo41Gj0e*iU8HAN1wqy~WU$qG^1b}497#`GfPrT6
z;U15ID|y!}_8)Pu@zdUd{i2q$wO&XUlq;1JYYXui_42Aq?^$Z4g4XuXh-JF<y{D!d
zVPS7W7ayvc{m+Ab?k*zVcba6JU83Ltq|o<2oi5R*fFAcboZ==+lZBSd?5bEyc3S9U
zd&9f{RTxXr<r}!u?Vs~|!pC>Ethe3$8_LyX?GfF0P!~dFuZGRqTH)8Hkj+OUMxX?t
zSr83@8yWZSy}YWYeY!#5cvA*+TaN7coG)NvfmQtJe~CV5d9wl)g<}a%=&N;AIm&B$
zw**#dVtpK$wS85}VEGPk7=VT{RrN-Oy72b;oF(v^-@tH*n0DU>s;&Ot=x?{_a-pOr
z$aNEgvZ_pEa#VtW?q64xlA5lob-PenSUP^S@OLsT$Rnzd%{<>wc|Iv7Ru5A#3ky%f
zhPW{0D7FkZH~8^RD1jlF9cJV0?+c6WL(C_2^1XVfAiifrYJv(8!v10eB9=0yF+BR)
zKeMW%4e2KPJHmp#y}?v=cB*VLukuno`EvP>Kw$}7@<iWcazO&f=LWv`vfeu@g+_LQ
z2?HGrc*I25{1iO*d}pT>_AN#$#>BuDcY%r5AC3fPMMFZIb#KO|X@N`7hDJ{C1tw>G
zHxI-2E2p$NoLhBp-&;?<O=jLEq5*gE(pMSSUy*qS3{j%>Ngs01@!r5UG_-zcZz7}0
zHn{-V4c{EuxK~5lYtVlYM??{=d~ax{5xICSHh(fdy5srSL$5cim+TT#1@|;p$nw^#
zn*Y7Br00nX{eyVx=q=5*Z8w;6WeBjpR(J4&nN$b=*4U<J66N0IulaD{_>FA38C0mM
zs)~(MHS5bCU^hiOx5JGXF0DU#@s?mVk9W#$ync$vF%cV{;pkY{6A%U6M)b#|r$in*
zkqr#ysNNJY(d}7PEx!imMPmw;9BNJODrU;AaP(UTl#s#2C|9k$vW+8jSX#__uH6<B
zq=LL4FYU0ugnw`8W)h5F9}UwsW!^u3urI8)FR5GDytf=ZxElE3uB7<x@tcHZUkj;z
zw_B51&4uoe?dt5AQx{~<^_F-oN)F>tg?NZtJRbIoh%b8~I(E#>vJPK|^iRpV{g5bL
z4~KKGKy^X>74XL<c+BWWQ?*=!fyNgK5sWsesaq`G2W@W(IMfX%A~I$4oLiq|`;;F}
zyXaR}28y@lB5GK{Kdr($>m@0QJgeE*9}+1LHLFu&cCtW9u|zk#s;-NUQ(nj`2E;;P
zmrT9adav)-R7T213tS!E0g`htXN+HXbKM@#bUIjtnx%0b*Ox~9ldkjn4Zys(Vojl~
z54jz@AWB#I#9sHYA_!~__s#na&(XRuu01lJ2C3jv$`EWkH{<{*qQg#%`9o9eQ~G9L
z@Kz=PXBZGcK{A$wp$%q6&*%v#>A9R@LYqL*_I3;Pc8Z(u`yq|3VSpN;son5H?^ozZ
zBhKaDt=J47-<gHht8J1-SUF%o2_r1%rx6}IwoN);uC}USNu1bHEf%$$qqpn{h{UwL
zOFof5UcQZ*YO{c3h`B%CyA{jvo<jFM_h-REO>0UL7yl?AYc%ovBXL#d_Obh%R`hNm
zPBA_d{qkLvyk0|$A<?<FK^9`yFlVey0&(yKV#Tcez{9JogNK1RQjaK}B>065R`_Dp
zx7^8mMW(f5e$gy#<Mr4W=&EM(NwQonSO5vxZy8OiO&DwB<~abm+B6I<+)d;Ft?B$S
zwRrAApZvJr(ML<nJCi=>XsRmpwY{&bmD)2f)|zl{Y7O$i;|#b2o<OMPOL5G|<9>7J
zd0t|9D@e@g06)$Qrg!QdFe6|M3RJ>(SbMbIZtw!l^gEAdG8N4mv%5#7;zM*H^W?_+
zI_q(qRpOPjXncx7$ghz`J>N7~7>G)Ln9TT=&fZg+D+T42xf@l+^Q~=RZ@_$lGX+!c
zP7)|OI}54dr7lMFc0Q5aoTBBAXn1zR`vFbxIn9s!T)5y#cXG(>mnE5Qvf4+n06%k0
zcBaPZt=U46W{04_d>Jqy8*8JC-(qON6*utV6@I6Pm&k;{J}ocRuDv$g2Y001{My)r
zz#;W}R<9HcWQpzNYAZ(pH6-x5v{D5Ca=FwF4Vl|TRr|)L{oQA&n9iOH$90ZTgZ-DX
z_BwylQEc^;GefDQE~|D|)Ao1pxyPCx)p}K)wCk<p+^KKZ`hO_$)-KRiSu~6iVa%qO
z4n9U>N)G({!`~z-Wqqd5IaaM8tao_c8q<s~F9AgfKj_pSJgnGcGtg-9!@D?+l>GKA
z(zGuG@lSRZJ`|AZ$UVh<IvamRleAt9lCur`_+UPyr@L-N?_3w{Dat1PS{(B)ZlMD2
zo_MxdaL}Gghd=uqZ~ey-KXG*m{-uzb<4Xs+r*r|ghFYG@Mqr&U^4}6i%^?LVjB4D(
z*~~b_!Y=cz_i<2Axoz)3edjg$o>?2;5z6aP<hBf5IjUW|Qkr_h`C9#xCHv;6wLM#2
z8u=Dpz1EPFJ!ictk-ws&B2>W{7<UT$la?u0?f>r2$(6+F?om4C^aRD7UugiXy6Rzb
zZYe<0@RO=1n1C>x(gB&RN>8^&CVM)~tUWwiYeV=YFry>*DTR*p%hMBeTC6*oef3%j
zHq$>3`AS+z{W?U?8++}9?lMgJp}-s~$M{h5g|bh-5?a0vVLdVt#=*4|%=(>;8BEPl
zW+4?rt<Il)T|)|ZOTFSeu1Z72J|{#bn^M?Ue0l%Z568+A@9JmcNXCM3>*3cz;eMYS
zSX`{x*xg)@HE@2pHn1XZmLn=P`^V|olOw}Dw&@*ev+wM}1s(Au*Vwe9X|Pl;Uz*=X
zUvkzK@4IfkHpT6VKYVMKW_Gr$dxhl?XCm>{T5yg}+JR*QKbaq*y}R#>w`4_`6*Wl4
z&*iQH5mf@88FoDN68$gR5n=z=sN-<{K~V%1kHg#rVdG}R;N|AN77pSyoND5<VrzeB
z?c&8KRL;y~V#EK;4L_N^>lxVnAkOVGQ#=t3x+DPQC4*VdD+K+8{W_B(m0%-#C&O=c
zIa$X91hY_3O7)&(U_wdlH4)^`ZHq)!ROx5!oyu)u>w>REyZu=lcRB?Up^%zz_n$1r
z5p1uI)3387_wuxLgX_#`uCAyw?OUSJM;G%!VDcp+GH*o%<<r=B^^dYqGUUViq%<SV
zj=5?Ln+baXd67Exm|*tV`3NW7$Sz4>VJ|qkS;-g!w|UF18hZ4x6bTdfAmRwL4RWus
zyd)=i0<Swc8w=%)+aoDL6#-^#yzsMBzhu^&>#?SC!}<lwR<dF>7FZ@C%2k*Cvu-gj
zbPXwlQ}p1Kr7P;-wrQ)J0q+BTLy&tr{|T9lgEJV}^KyppgeQt-I+DO-apE*Brdj!r
zes9fq!xD|n>tk7u@e!wk_lnGitY#LYbH7Ee8}Z3$Q3giXauk-<#;WLRQ|SW-rENq%
zrX_Wzl{tnEe&ErbN4WBqa4|IB6ZS85+%YNR7M6KefqQ{n{;jFL#exVZ7Pc#tq5z5X
zFx4;2?_UQ1$9Tg-2*?AT!PAI&dFh#fqj;emmO)eTf$GAx&7yJV=hZ+-X5>N%1`o*y
zhXsxR>1iw3-n*_zgQ=;Apb{Za4I4F0w%iw;%(H7~T7rMxv%u4umz;8RMysRWkw(|L
zJmd54O0@U9E75aO7U&&>_!@lQF7o!JF8gR(J8^!E3A_&c+4YE+?oFLZKv8LazdW0%
z*8vY%1GXnsRxtb63F$a<e6wKuyx!C8eX?_1iI=H&!rx1ed|~rMuP%`?+vuQIPX#p<
zNicv?Ny7qOyz>L;uk|fodpt%@I>~{atK&F1bc5L1%E~q$k69~s^4sl8TS@<njc1+h
z1JdI=d^ZJY!qG#`AOYRUfVjBen>HQAFJBa64-{2Z1{4SlE2&b5MHCp(WXzqmmj-m$
zj|zxH1-{!Z$u}jR_vf{ZAli-tFu^2wf1)vy;mC3Zx2`jxy>wp)T=`-5KD;_g=F!<p
zVOdL8c4)`Y_;OVdyyu{Dk{)l(x|4Zg$7)oSs8_PL0TW!S(5o%?*jbjQeb;Oyg^gpp
zA4%!A*7SL*6-eA&<Xdie6!E4BbBij~10HwfOOi@JvEI^XT$i2=WwC<FafD_UVY|$6
z#AAo+;~Ok#*nanY_qSld$!d`+8sa+<C+kA!pvE#9CWgi=k(FA@3!7KpT@`^bEitos
z45(!|8Wwg-df@3k)e)TE8{h^rbUVF2z9b}s_+`Ix5|k$#lPnE_cO<Q^6ydmVm4mY5
zOd4(MNt-vbBCm}+$!4$ZiwJoFuE(c6{#9V+^g_N<Vdy&Y#n@r>E-v^>stscXKA}p;
zo?CSEwY1?MAh0ToaC@$}mn}<ZNLhuXZoRmh1nY)g+Wji#UEOXpT03lwiW0Qq+%J#E
zs!*GEEB)~Ow*$;e=iy*8iv0Hc)SVN!3g<cd3jlKH=#j9w_|E)$G<Wk31L_`6K(nA{
zT8a{N#iWB{<#2loN0mC)0;GWfajAu)8h8VAnz(qgcfmHX^abi5p&G>ek;c)F4=49V
zdI!hE+437BiRX$K=(zZ6iFtqcT}cYm)i)m<{Q`uSA-aN)fJJ+u$Pufw%%KuGtC;nh
zpYQ9U61)j(v+VIAfgTCUC{)(`W;??l?EcA0ny1Y2I&0psTZR(22U0C`?PlR)PIR(C
zy=py{wx4683oq;Tt3-2|4ZE$Kf4D1%GT{A3zDNoHWhe>zwPwT3)Lv!7jlAB>906KW
z_fL}de&%`r?@1fowS}i!uj>CpMg_f7N_T%blD2kC_OLk=96NY@2U96+W#Yw9da?L}
z<6jx#4%osttLhaU-J=~WD)#MrhnETQIyP&8@wAcMbTMH3m-X*v?!<Vc{GAUx`0QoF
z??<I`^W&Mky{t8ROg}p}U&ozXj<LjW_8v(0UX?UBgu-Qnl(;29Sog>{lQs<Kq4gSO
z!Qpr*DYd!{&o=h@pJLeWaz~^#*}|72bm?r^l0up$yf3=99+0mbeE$(ItemrVWOs2n
ze!4Szr7^rxR_qPz`Q6%8xZVIN#Kzvtd|_;AKdWXc9BXV|O$QAC&1~78H>EAJbi`|-
zOg?w{6uR+}zxum_4-C+-Gu-lV;C_c+hxC0bOz#Gv(g)MZMuigaRvj4n-n#p~W%s=U
zS3^KID$yjO3~el6gabUQh8mf)wEABzt_cU}8J2q-CCLDI@~4^ljCO1iPzO+SY09;C
zNfRQ~Y~esf8eW#R;(fog4_LsJMt@K^@J{<A7=?p6`CWW8*{k#E)%5$SBW3s(ryX>k
zXFcTR>aA`h9e%oAa+C08_d-WU{Q|lEHdiiCrRd4RNARA8XwB=bDQL{V)uqY8*?}Q5
zwh{>mYH!mo$u`3p<{M`1=OH>?$-#d+Os7R&zEdS9@$r}O3ANLrE#QHtt~=<BS6m2a
zpS&V4r2XzW@*5MuVD=W#eFvz%Foe>Z-%nu=LqMP_jA75)aY21SFM+E~pHWJvBqdH!
zHH2<ZKJ}B&w`dsGyfQ#|BCUQ_n=MvXtI<T~b=h83GHxOL&UJmQ@hG4VP+A_9*PZ^0
z)+$Z*f6lzU9C(zP>?H6x@=a+>oxJjBhvf2LDirK&e1Dw5_7nzxZ!3G6|EsJ^GVL@Q
zfw0-5zl$Xk|1&e%blYSX#To!<Q0LX!stjU(uxp61!dm#|=sXDs>N;m)4Q70HO8!Gf
z8>7VX(vd#Fi4dR6OsvoyKRjz+DB+l!f1JFBh!LbYcdk~oa@kPT?HPOF(iPtRQRYaJ
zhS#<`42EwnVcTZ;t~_dNvkozVP+he{M?qmKGVksy+#^q~#&9{k@meR$(Wc*AI$dVo
z_w)*r4QTY^Rz@J)RJ*)>;cV=R<K`BuYvubGdYo@8MZhSHKiTRClX2QEqTeNUr`gy5
zyVs*XAg}il<4tSJV=j5Ol@>%&)8arRul6+Y4^aV#>?uGnAC>!B_{v&a<$@!Kb6g(y
zCIsamgPiPEN#!Ew6(waE#^U?mKJ(i-re4H2!^r(oR8B{r!7sqe4J(ATnFsT^p$Xya
zr(%6QkW-7DtdPts)KD?L=<#0um?)8ZlhByoETTv8BL0l|uruM14Q=i-S6R+Oa`Ei7
zsP)6ZD@^ZV8fClYs&xbO2mZwXd<OyUMoIMgLc@Z-y>^9Z)wi>0pJ=V6>v)2+!72G~
z=2mu_0Y6VQQSCqEjKQ%(#IY`p??@4s`xRVTMj1BCl7H17U0?hk3y@!-6YX>el~wlT
zVfjx;2+N>PQi8&-L%~#5+A`h*(go}HAX-!>jaQ+8Siz6f2<sgTkqkg&M3A8`>p95a
z{aUu@4)?pxnE-I29`YL3Ynz-%I?Zs6hEa?q5@(<j3!DFprKA>eU9P|m*xG-nF4@<C
zfm;DcnU#YgUlUmaKw}ZUYmLGkL=8V1?T+;+@_yd@`_W0zZn;jsCGUokn(D*OPNsMT
z-I@fTTXOV<%SMrm?FM5==~J2m?W2D0ja>uD{Wd&WCKD_O_t#i}#C@YnjVfu@DF8Kz
zkzH=ONj%;(^IAY;X;lh;a8Crt1_42TKL&qtFdWtI>d%$kNBW85+e4D&b-RwliB##_
z(Htmgk+AT2wpmw93tuPUJ_n#Zk2P~Z{TFkno9Bh<0<OH=amoj!v1iy<2Fi2(J#fJl
zzob*+8JkS9bCr1~R^^3G<(2Ao?2&a%>FnROP=@K|d<_Q_-Zgrx?JkQ5*l$tnf%Z=w
zBV!2+KJo2Ia(mv32EXL8>v7F1Gj+YBtnDoL!{7L^%^Lz1^=+dMECsAa+B{V`IQhf^
zVcIIu=vae$wtWx9<jhsf>E(5eu7TYS-r8LsJs*?{ou_U*<N-R3?TMcrrH}_pXsHbe
zq#$UR<%DOH^Qu3+!q1ll^h^cQQVt1<VL%!=WcFXFO-szog#%!r=F;)*UT{zcrMv2b
zr9MwMV2tIYg@lwxMKfs*wmJsAi7Ab*+aEnw%=(d_0=!j<<Iia?j|L0hH51A0Zo#Z0
z^Z88u=747Y+Qjq}d!*%#xG07kM6PqD{~x5+9rnK@2ZPrwaMQ2`1Y1#qAjEFg@I`Dm
zvw=Pq0Do&%92p+h+qG&bzo+?>E-=h|2zx>3O%PZ7=n+rvHX>y4+aDT?mGaSZ?Rfe4
zf)AZ4cy~ed?r8L4478~388EaWDUHQdEai`N*J^#8%KG4FxB5WuHeau-p|%yjFzXTb
zoi-Gel#&`KJtM&FOg)|=#lG(_WJjVmw}L8L>JI$W#h*|ESC`EqR;GLNqZdL@O^AT~
zoF|6Up6d?uz4jN~sCzTeyBjGKS^V0d#<_RKU!kppyQ*LVHl6Z)_>enb`KqU44$}S^
zCuzQOAIT(|9j@*e2b4U%wsTv(eNSKVuOFzO7^QIKzbW!9@k|kc8<?r8g##5O!5;Yx
zKfRT!H;Jndxzxq?n)H~mXKwHM0s}X2n=kFu|M^EMe&*WbhDhlmcU^vX<rIS&ua<f6
zyZY_3`l7o&_4w1(ls8^21i&N@Xg=%#?JxXWDGRqKw*)*w`>O+?lO+Vt2d2lNq9!!j
zUlgRwz$ICXnWhvf&-Lg!df~&%`vD>Y4%j{FtLO`U>1Z!xfHL?(Dv2VrXHO(7@~*sg
z4FE;1x8Z9`X1b?^^nTg4d=d#koetQ|<$CRc*v70NTUHP;meZXRKg9rQjn+D(k~*!h
znXSLTo0YH4{bjnYEbf4oT<ksd7^a)ZaUmRU{y-FC7!mK5#{-^2R!dH)B$+Yq3DS{d
z)1Xl9O|vPW&$s%TmpyoLt@~j0g0$!KgPhN7D(G&0{80(af6FDl?c41Pw?(<au84i9
z3`*A*o3r5}94XhxKUs{>)`U|3A_8ubdO@IYJM%h@I>mxx=M0?@l{`|%>pQE#|3sA0
zOyAx#grb`C2e-g_hX)omd%7hHDE&PhK`YP7UB9f9$i%SmlrVeOLUmY*Mp8c*_P4Gs
z?JJuR=?HB`mybn%#z`Hecti6f-rUa<S+R7`lUa6vDqZ3L+N%zB4j9YWVRB&Yw#5YS
zT`xjFr5@>K5aq7HX)MDeqkjin>c1E9h%|ppHyvANm@86re_Zd}y`X(S8FU(LNYuVB
zfACDd^(n{=($*E5nX=_0vED1o#-vielT#@kGu+O4v|j^akM8_26L8K#{Yf5@a^AnK
zHXt3s>|JV#S#6>SUXQ%n+7@h-oO^}#?PxG*j}9T=b)E`Xvx&M$@=QY)3@+F9yd<8{
zibq4=yYlgEHuc1ZeH3~1Z8cjgKRBoOW>~1McmS@c(m!IqM?>F9@bEVGt~zX-J};(Q
z#-FONi+!|RLQ51$T{DIUsPFj#l>uK}=6Y|wGtfjM3z-o*Ydik{T*m7@C<$#-uw5nG
zh7{W6?`W+AJ)Q0jC*TS(RLB2@_rCtQgrwg*PKKv@6F%JSd9a@IvPgMrYc&LbiM;~B
z%2Sf+SSJxX4d3i=J0THr|AJW_HYD)5;XPlhO0Fm}C@diHLkfRgr<gwnGkFP)qrr8v
z0<~fV(~5Yb5e<JKd*Vk42gwzge`<BSStHgO;5IWAjK02F!I!WM`h0+MNWR1&HFLT3
zn(%L5Re-RAl#Z_$i`8+SyP0mjJxaK2AD=4RX9X>wU=6KnPm@i1$KGDwM!>ALf#C$e
zH)#A!r)c7sVhKUp{KWQE=Okp54f+d8QkFq33qo>qR<9`nASvm2^6<m+T*WVAO2}8}
zm{y%8s#u^D=NFv(IqU(h?#B;066V1dDv1^Gpx5<=O~Ik|7zKp66Z%ZRl|Wnh<kdro
z3}AqO{+@B?b2_y<<hXBdpwWe8?0g)+xLXNYGKdMAxjm>DQa~&F+gHDfk;LZLV5p4A
zNQ8M>Usl3W1)^{tsi)2bihXXXgXyqg?L&0EESOWG_tFm6V$FitgE%wa%Fvt-Z&NK!
z@Ndr>vP30k(*srohiWm}nmJaXl5s{OowUWTildX9cLVyqwGo(A)%*GCUUnTJy9<Ep
zJ2Bam^K2<XW5)+}8Um2HGux~oNz~_v-TF!*?rwvkm2H6`8HZJ-x)1U+0EfR?h@K<|
zcAR=bOd5j5n?LnmM^qPEUydKz?B!jfi({1G3DsJIoUxD4Qx+d)U*%%nO*%KI<XeaM
z;o&rvpf~30rflV>W^R{0oFEI&;AqB~+7S2>XqLY-LEuJG7c27}`Fw+@V6I<n4nXYG
zn|nh;sD=QP-(ojBt6zS@8`88htFJK2+9A5x%DMe~|9t}ZCV^BGt=j8_EMf%QuFOBN
z>LdM;?Ux1`IECDc0xXk#=rC(pjzw?3wAoou%0j2865Zj=3mCXd1$t=1L7)eSJ4xe^
z9QrTUyb*{^u8KtwV|#dB%rZ$jP}teKBrJ)1qkrE#XW;<XJk-(exrTaLZj>9oYk_t0
z2cT3rbW8SVzDZGQAb^bCZZxhmwjfBp^$+^2L%i+a8Us=hjB{h3cI!ZqMt3Zcw_CZv
zbQ`>G6a#uH0J;v`A{afD?3DIk8an%YYt8hk%VGkf71iJtk4mE_ld0JQdU_l^bx1uR
zEjyp7Gz#7eOj%flX{)LEN2O=|?lkxZ^lWW3gV>xZdcSa>0FZ$k?~D~Puq$?(yzbWB
zd6A47n~a|Cf`uUf(BrOw&Pln?6W-^Q=1tX4Ks&%HO+El<@}74BU6-2SjIv+-L&#~i
z+{MfZrJZ+1IGx>Op%ORmE|^)+g5zv#I+!YObibKMab5k*Nk2iR;Yb}Xhd23bZsyz1
ze=0w@{^zYS^oRTJC{hEx8pbU6-rgbXTa0+A)7f}VdxvTZ8LJS})dA`wiEMe`QhWz5
zB6)%(=0p(`%gZtH^D%&3q$htnrL$1RZYx`o>Dda?`0Kb5%AL0FnfC^$IXFiQbIEBD
zE6#{vtlY4trUl^zv37nOs$q<&-A<jz1^}EJ(0}MKU~mF(fm8_A;wkhL{mdWYvCIPV
z_*4zyRC*;tS9q==WOnvk8fb?Fp*19^T4zqZ45iW%sNS<LBXF!dCYQ?TcLDNxYWD#n
zRW3*}^XyCkj*c)(yeH0GBRXSs+tC2M@P^<p+K~e9Q>H5O^WQ>eTtK03Zkje4?aCQx
z<_UM#Zeirj22>cN_%%T^Krt=!gGu-494aZPpk2!M_NlGUB*H>CB+3(v^Pp1#6!>Vt
zGjQkCFaR8qq2NY!Fgjpr9rN}mlF#f;hb(r%G_}9}ysqaikwj~!>7`c(@U}R3cK)~x
z6Bnshb~v)ncw^s`z!QPKuu*y&{?TzsQH!IqhF$hzG;Sza<DpA}%m6YIY8~^DF@za_
zi8vcu<A9F}JglF?_Jf7HcY^PceLRHXlV&ERRsmk(*GwH=gUKl0bKG<Y@FYW~siA<u
zOB~&(>!36m_jEh{Rlj6VrIi!<ED1^^mJ0#|JVdhS{|yt$$~5FVUK$U5!LS!QgIfX6
z2uKju7FLF-^s3dv|J(WzgOyK^gdyCzm(wo@;z|oxBc3c4QT1ow9pE@OZVKLa_-AV1
z4mv+AJJ}ag5Kj+k6|~YH5$39EGjO0bmMDT62mN3&*v!#jp~qX0fD9qeW+^No3oatC
z&YnpX;q@i3NgnV{Rja8KfgBjS?|pxtNj0D&M`8FD>v{ew*3x1~Npa_on?Z(YUhK({
zjH08X@Wmg=sGCOBE{mypd!=AP`+D+d+=<l&8#f7nnmNPeK6)5)6%z^}@Raa<ZK8VR
z^gqps!)>LsUG0WkEsomrP=gRQsf{6(nq{D97g=hZoF;9^jeD1JiUDoRlSeZ}aai_>
z5ch~Cq2K<Vecbta_b06p_nnU~=+3yIiLEIr<*(CoShvM8)44g5q3$Xv05cSyrwf&Q
zngB*dfXR{EC)`c|Wk~&VF7V12^Or~Rxx+uWhIvvjA1|sEIx}M2`tFBdM5gJa(<yeZ
z1wqH(e`0H;JzpP`=lVpFV7&QrDtrqfEsr!Av%SU+%$)G^%HFaWey5RZZXh4ut1j^z
zZnV<2M*e2SCK5JxURr=-*$21jK+g1_1oC+6<akO$1yB#VS!aAEUgtqkx^{M)@=Gb~
z&?Odqx18glyco&mMJ0*p9NhsMGlxsy-$v$k`U6;F2{2)Q3cX8Glg}_*)X|76s*dI<
z1_mhxG7ocAP;eXHDV|EImvdWgy8?<)oP>8eA*UDv-N3_?e<ytLuQxg07iQcuu_K+T
z8P-aTYRv6L9N}ttBZD70*NjMbsfqi73;Olk0<*ele9xiA`^V&j9a{EG4~{r}BGYVJ
z&ASR=;L4%T{>|WXw?_nEje;R%jfnh{wIHBnF!N#GhlJ+#IC4?=x|hzbudQlC%j~~-
zI$3;&gm-H|kHjB77%_-ob7Xn5i~*6Qk>6-oo`&do+WAv`@fG6ls-m=$tf~=fzu7>D
z+zINL4)M>$O+&SSl)o`xf%UItN&)NNB}Z518%Ehg=C!y>lYeb{yqH3I{kdw9Eb?V7
zh>WbfCpWdx5`q_h>ufqk1X_m(e3W5dbx-LLwEhBl=@vU(IzXFRaP3qHUo7A>O7O>{
zG`*A0B$={l>S>f~z_eyh{%}=<yk-O>&S22y<MzkQ`i((=SW7z9@$K>>6YkCl;Seb0
zHd}H69Kr0{h{l7F`FKo%<Cy6WvSV-dd3zW5kjN4ta=;O1KSvha2mz9EU4!Ek4-3@b
zLX+T<M1^-;upuycZm0+=j#~7zD=-M>$AA{{tg&B4H`XPqd)?i~eW0hCGb#M)%WY-)
zn_A|2!C&#(2lhk_>melvar(>cYX=&6A;kAu*7Cx^#(gbzv7p(T#^j~Dvl}lYRg+#r
z|3HZ^?=h*GC!Cw(D-<^QhvNq@y&IQS`ZlX!nh(-~CFt=y{g<hnxuB3~oAsZ&ws+f;
zKfLJ274!8cc~(!H<6OJch2f+_q#tMkSAkZK!07`U4o!>`!56oqqKA}+%R0>VyFMq!
z@$vmIysTQ;&bEcn@!?d)wHkdlwU_DeCfx39&2=fRxrQpN|0Z?5m>~p&ha(xatJ-1|
z9hSehuTs}dp}SUJ0Cl+SI)5#fS2aY7IG&<7t={VkuFu%x@yC^?xypz0>lA?bt_EB2
zev#lLXWc$)8sH&x69{%*a(O8Buc^rzTvOteA|uinn7@b}m6VmPtl}l`TW&Lz-?|c%
zVU|~k=vP(qiY_$}yWZ0R!0mHvWVmQ_lnCt?SMSx&W%QAkr%yAfl~_+B&Yu!3t}oqR
zVdDG^su~>-gqFH-JSXj?MAHvg`G9zoLu624ZL_jT$rJhjAz;S=U|!v80SN~8YS~1M
zYG7!t&uj07xQ<h~x(i^4{x}+-_ZVHq<#92@Qx+=i#7RU6PfiL~mIxstgZNnG7e1`-
z`sW<e=?XH<y%KHSGTLj0vi-9OcC%y<V1R;!KFYjy@a3<T4n0S!y><zu?~<c0y>kh#
zSR9>e(IjS&CR?&Yaezl@Bqbr^n7Y0M)JSK+i=>}REvSB(vFJI6AO$JWTu9e@0dS!;
zT|g7+6Dw(9L$B9lbS`a)+AqY$EXyi!<T#B=sFd<vL+Da_|3Yb%j<FXcr9VH~;ch11
zD>feAzXq*kUz)x@MK0LMkaa)=BCTAP3<7{~ONmOt6VljgMtbRPJk9`*xedQpbJV(@
z=m)RBk{Id{Rd}9aVMs(?mC5bkZx*^H63!xBsldsnM_qhoT;RfARRe9B^>A&iyh_&P
z#y5U6(ZTlU|F!qtQB8K;+h7D0r3fmBpa=+vbfxzqg7n_2AiYYL8W2SVr5EW)uR(fE
z5D@7gT}p^b@0|c4$#?R+@AF$T|9rD%*8DSTO%{fj+}!1yy{~=kYww&J<&HEvqmXq;
zk38S7PDD!7Vm`AvR%?bwFdvKcyYgD;TmcooDxD(M#n920Pf`XI;59xqbm8fQ&WFku
z|3J)39-Ow{E@KppqtS1LU?$E*C*GNJh-&0ypkza#H%Wqs%r|pazJ&>1^ljM5Zd7L&
zD59!NYQGe_ZoP25<>;kzyR@U8zO$)ColZ(bu}t)A#=T_=QqHO4U&DXC>TQc9zWFJd
z!+0ao_OF5x@tjC9#3-%vcAf9Rx_osG^Zt*z!CI?HA-2KL=9=h8oXH|mLa;^*zR=+E
z{RY#8P^RaMqKUt|$dLDVtDEP@Gmm0l@QDWBfRvVAQuDrFX)^6{HIziB(r$i_%^@0s
z_&I-(vd|ov9~fa`@)}fCoaCh6->_C8wNAMw-J+Z*CE<bCY00!0?C;u}Y^y?s@IBEf
zyIBK9?^g?BU-G9gyElSPY6+Zt_(xB#GdKBo%~z@8F5ZR@i_4^kZ{gaqn`OsytR9ja
zubrqhALs^gj0J4x>3HI=@g|tHrpI@STvZ6FDY=R5?+D8H@`qKjEq$VoG{YV7+Qe)T
zd;1;3(JyTV0sYkEo833*#PmqR!jrb-bwv5xVxtdml+2NCXdO^6Hu|9<3;KIO_QT9g
z&Rhq#%x0|8rA=foCALthFpnvZgymZrUujp#V=Y#A0>0Uo>~T|Se&LB3k@c{2b1Cxe
zd!_v4QKzXe+i}qh&|4%^9KutmiO`yERz@F9fTe`NEkVqA$B8joWt>zU_o>8vpSYw<
zx}eFJe+XP(@e6Vh*k)XR^jS1J&6C&g_<-l;Bo*^x<uU{b&xHrdMVl6;F}XP)YI&YO
zE=~gnRowh}^^1>bo0oqvR!D7z`r(Bf_*J~`L%IwNGnt^MqX^S<pEE*@A7Az#apNvQ
z_p(*}sb9zo2sxaH`D~FmzF{+H$Qkr#>Rf3BjfXSu@?UB?9n|J=nhO5~-8aKdO{bRk
zDd;0?N2w`0&GJB@tZ_Lwf7VEtitI*f!xm^8J5e^Iq86jKow_#uTC&2kP(RX6Z&My>
z{{F)lQ5yS1A50&1)#K`gXPoIl_)qCcv0B^<d_RNkkmoF~r|9zY&rcW0v`;^?TJc?%
zHr3WxuxwH+a5uRDa>ItOyP;|6d2}w9-w9%KBVWh+ypd+p52E)LYDj^qd>Ir7Wn;?8
z<!AF?ZK!Md^-IWxZYt&=4r;o*?8+qX%J0fQ`i*(J`_~pY;k!l;atbf6w8nfz`e)A;
zP6qS}X4n>lP%w5OuvH)F+&Ak$0W0Dpv1m^S!Nll4c4jAZ6CK6$9BwV{sl#QaN*kyL
zA$wd!#qcR$_s!)XI|mBG!#T~Q#g;tSw{>fkY&Bunw({%FVXP$;m5D#SP0KQLN#cYu
zB5mn!<}C~9$xo)U;FcjDUiScIVs$gP{B3EeR`oS!=l8`9-vP+0L}pB0B-u!Slq-lk
zGwt)qd`V6|VK7WtVKni)zxSdc@{RY_+s`BVT;h8Pzv|SQ?%Mk%<2heOUFz4L3f~$(
zwSA8ayiaj2x;3C+hbNqLbXU6>TIqzS3JhOJp#0J}BEDF3DQ$RQY3kKh?&HXxoOH|&
zCh_)_7p$kU*n`z$yb5cD6Jx34O+~hzB<nJg>Re;^2AU&HsuWnh!=%8M0|!@D7C?G^
zUVvO`4fvj|m(ho}W%IPE;Es=PBCD&`IVj>Me8HsiKmA;y`+b(q_<|Sgj&)2&>wcqC
zX1dMcP2JFAi9*<r-;XfAM#K?oE)Sxt`jDy8Vy&giaw5y{z^iKhRbYZhjg#EH=qdaq
z{lOP~20WY=N4Sd8Z~Yu1gendst6doimO0z00Y$6HMgOZGl#Pm31rGIMRs>JQaj};v
z$6oW73Nt>=QymlTP%O(jyheNRdw&rNH<r%7i;t=QA{P;HUxyKGNhKF6SZ+nQ{m0>7
zQ|intN(yafCn|b&h+bX2fTHXbiX)szu@HF@^i<hdo8$p8EotxxO#c$v`<JHg?gyq!
z{aV6`C!j~?*3)*`rHtI)=MKs4&`1y-Er;Irk*b$9-SR&OV5q)R$H%sD1+?mJ7Y4}5
zfzAN}$xeIh6doE68W{uwGH#tqPU!nTcu;h8KhjdeoCqRIO;}KaUYv-{tKn8gwGoxh
z3{#;WVF7~%{m}%5>f(4JKaJ&Cv)07q!KhS(ME7@Iq||F&Ck3&-WeW9RA&Vwyi3|Rl
z*tvvL&TCx)ar9PKUueY8ef)HE4li4>ZM+qfcZuzeHv{Pxs~N1*NwG}ZT*^em()t|_
z!~wmiq^-4H5a0FJ8KiynJWBAhOPlLAI&f`piI|wuX1ojzfgVyjj+G7l`X=~l`h};l
z3{SOXpT4zjbMn~8HE@)3-jg#utI+2#aKTQi)bZwoluX>$k(JXkk(H&QdMX>MV1sI)
z;uk>UaJO&V<WCr2O?X3Ixv~%^JS2sl7oh#zgbo@y4T@jHt-Z<&Jn*42MQ}I4Lx`K&
z{G2~f$0Q9sT5R46<ae67R(=<;OaHyA*JtqOH`Lpg3TJ-H9qsso&CJ=Q;O!Vs{NWeU
zDb@t5o@Tq^Abb#fv;OMU>a%ObzfNC2^5RXoHI#B*-xnkur5QUReJ1z5@3}kj{tnj|
zSf2V}eef9VKJF;>TlS0h7cllSAN4z93|9RH{ZfpAlNLpUJOWuhVU}C;Q#|jAk)Zn>
z=J}m=h#^j#M38T6L(>SM^B<=cVwmH^RAM+c{vo(8eeoE=4DR=;3CAwby7awW^xd^S
z^!984x962s2Pj-0*TvYwIb3eq9p|_Fv6up<8b9pqy)sW8Y=p_b_T4{eyKWq9RJvUu
zp~oYE`#RTM2cx@>>8D`ZPkz_AIjX*WJum;wBeIK%T1q^*AvS)o)3CoyBZC2{qync&
zti8o4#=dz($!t@xf2z+8W!h90V()nijou!-=8Uc~&dW>3mA?L%IQ}ivyAfJg$dRWc
zXS4GC>ZyzgQy_MxW86qlrF4K7X+zPt-11IpHra3aP}V7wY##APiRGBhx#z=nfTPjP
ztuotEAM0dT<*aFHR*yaI)R3E}MYN9mbROR@FzHGrgW24gy)Pb96P&Im2w}SP+5=0{
zyY8p|ywGZ!j-$Xpk2Kz*eACskw=Ljp1{|d#o&8aqj$gv;R!_J&g}`-fq5Ig3Bv$*K
ztjJ0x&X_k(Q^nZW#i=X&y?*R(ppI0sFIkMs2*j*Wnx;Q=xZKz%+7y8_`fF?7JvD;I
zseyZG*C;wUDDq*qW~a}FtAxbYB}&vZ8O(KyT0%xo6m1(-m#_I<V~j_;5BD|Yz&^2x
zHQ5t|g_+MTA-1+ZQxbhvPwVr-1AD|!(4lT=w&^W5)Gb^zcq8Oar%1(Qj57_%@j`9J
zJFiSd_RFvv>QsNU<FToC(BBtdS?L$^P^&hJ?`OmJ*?QKyU7cHDW9@SpQ85^nwyK!p
zhQEj?#~JoF=cb&eBOaO|U)b37^=(t(;8yK_`0pszR(};7N)_0LpQg>Bv8^5LUS%6J
zo(Sv%twu@e2Gi}?ZCIjUuwsC#F)>Mh+bW!&)}bY0o{3FwZ^gzxLxLSOu*OV+Clh^k
ztzQI_?TPX={J9lDj@?XG`*d0#Jzo~6mL5;Ul~&_*Q*cp>XX0A@%;X1+EdfaJ+N&8N
zUfXCE_KpkAD+6n!l*(@6H3GTEmRTi6;tY7G3~9@5g?n#z*j&MN7}3`T_fOU^%r|66
z<nZVl1_~joZoiTX14GEgdm*M>(OL1FXeo3rrVsUm*H(moz0U`$9+ul~?^Q5v5_aL|
z6=$~FN9+C5P4LZ#Ka7vHzLxu!72h2k*rrnY;LXcduJFiUrJqImU0l3LqsoFW#Z*gn
z_RqF!A?2f4>zDKWCWzG)6@p+YG}%lwj$YGNIRCVY86}u3{nfzkseKLT<6Xa!D=UY}
zr+7@<rcZWn@D!Y%UN{~v!x)l6EBZbA13Nln+cM}$mPbso4ED`Jh9=a&PF^|QeUHQw
z6{7jO;EHo#q6yjeM?{Wfxv|Z^bC-_#Fq3!Kbm$4gigT&~s}v^(4#Y%GsK{<Us*$*8
z;BYBb$ef;Cq+yXX*U7rx)v(dp)BXu!pQU)E8kJCalecZOl{Ux#7=qODJvtv%*kP#y
zHs?9%XtkiKLk_zsigsS3o@d7yN-2I~Sw2m>v4nRBD=&MeSDQ{Nl_w|PCA0fAIatgR
zZ*=N6A+s`iVxG!xVlBM+Z!bXRqbJ9wr~A<Ar?7VUPVp4;hOGUKN!=i)(sHJ<yO~lg
zOQxch$A;;MmSG+bXOE{i@Q&qiwya7Eg1A%MujIx}TXT!OGh%}&u8u1m?}q2=<d288
z__@SqX94_L+5FD3X>Yx8({2E)$U*<O^t{Bc<9G9!_}V7jX$s9nW|et?=F5}`6M23G
zN6cyGWrnw4tK3oWkctduk&BvZHn;iLCw0%lu+2V}ugau8HXpI9aYDXx1NbZ$cx1y`
z^i=SsB)x$7{mcoo5#yi7^*nz>qyr6(8&EwiGm_O_E}wqimHTWwXmfPu_N&idv0-vQ
zdD4j-6(#Sb4x03=Pm4U-kRo-I`*``&A|1XsNcC1fE=WBmBGjs<IGdwPRm2#6ro0`w
zk}SlxapO3&id^8Mv3PX3G*<qW-6x|bug#{&lXg%wdXs`&Wu`H*Vb!FMY*s&M;L8F`
zM}@X#{a$BnLHR33$hH`hW*e0dxny=<=p)L@U`g2h1FIzRs>|1|ZH8<Xw~krH4qaAe
z_Qu^F>b<#-;8)+yj!xNGc>Bj^<EPx|R1oJz<Q1LgG)VleoLG1G)UPW)aMs}&V?~4k
zI2SX@BRy42U+BTa&{elx$qbwm3sRb=%}xm1<g>~Du}%73)c~AoI4o%g4t>`aV?}K$
zr-vMvoaV1i>zjsiov7iekKKP7&3I{5yV9-I5p!8yHe=YYVP4Os+q)XJb9CiY2G6~z
z7lP<w>|2-0u^Ui?J?zxlxz_2O8`5kSBB(z(w7b)v;cA-yh!|3^S7||~r0w_TVu}`2
z#?=(jw4>zKP|k(XJ2@YD?mG}SUHxgZvMqAv@m7T?=RHU)5n9slFww`Ka=OK@S1iCN
zqMs+vY9X?Sj`ZLhc1N*y^JVHen3wq~9<4wYW^KETde0+9y+1V5Z|@e3>Ry{M6?rxp
zGMgOh126Ay?m@Q)S}}CH%=3SJO#`7ay#Q%LRjQ{8uZ?NSOq?BU)cNgLG@0?%xlHR|
z&yIBQIPB8ZIF^9kc(ycgAE%P;Y)Sv>nR?edo|wIbZXxGM?h2!-C*9Eux@86x-h)C@
z#ZWx1y6^KBqnl_2mq#XbaX0P07_e|~aMqa5u!~G#o*BGBNtYPE9isIpP1pUibPU^-
zUP*PG>BWY(unqgY4K&Ylr#!cw4NKqeYkVA|8M~3jJSrs5xg}{3W3kU?d4J_6SgZcy
zH+&-uZh=rZV9VyS(XH`UZGx4YgnLy<A8+%moIE~F3A%gMI=!{j{J6E5#P)*#^31u#
zaurdYF64!O66_Js;?k|{tD}pI_pgH~xzZ*b=)Fhr;Y*jQr^@5uela!}TbuEu>8hIY
zdE!OO`o?jlQ0LvBvS~!hILc)b#`CsNul+YccwCD2w%IS_A<t)aJjub*VxVn)QNIgH
z*<c69f1{uG_}1Ra>lghZ3p<`Dv=?2)jEb|tTX@;D_#s@ryrykVYyCI$!pgYv>B?~2
z21cg%wZ)fGV{B!z``X{HZCr!pg<Xi7TyI_b+w77y>&l42=c0M?+=5e6eBvhZ4F8Zm
zP!4<2-Z0OxV_8^g466{G{V68UswA7&*xYxr_D1&iH;)dLgcrgx*~@GL6Hx`#+)enE
zYjnzSBzl{lt`7_yFZa*aE7%ncUC%Bmq)!`B#ok#_&VHLRqsiIrr%44cs+|Pi`8!?0
z)UG=tisg~(ZBJxKO3;`kxs)+${CIdg_1g{8kB*lgTv8By{Lrb$eOl43UpNfEi^<dV
zp7pQ+6cEpjPl6vfbMR>@I+YGgZrM=W*5eRZW`W)!L(Z=Y$l~<5RI<NYDe<cr%RhVS
zYXNz0^_T|o>Ui4zXs?@%<fSzJ+$$J&x)N}@lybBpgEu}}zhvyRA-bQElmJUij~PrC
zaqf6?wK{6<z3W_~_8k$o@wq0iqs0VGPF~)Va5eQg+N#b8_t^cZ;9HY}brU<K#fyH}
z>;Y#%pQ(ZBJM<Ac6|5A!4^iZ(j!xDGmy;{sVa};?clAncu=EFVDl0BLs17)eo2Uz1
z-7wKV_2$};<ZYPq!#$mE!?VS%NIsN@UC$%px@%IG>^CVxbebJi&221&=@YVfb?Ei>
zPK4PMinYea^i&4Nv`ZC$LxsmpYXQ@!Za#W{CbXV7r;};&HZ|f4saTOlA7-75O~*ux
z<hURDJD9Gwr24W2OW+ba+o;~{UXJRB=o4O{&|xn8na`ihU$Ggk*T-4o@^#tC@WvAM
zUQ4K!XUFWSi)Zafe$!fBthd$0tGqTlGt8K%VcR9|IOw0&uLe|~d1iYYeu@6|>LSJW
zX?8d}T(rB>;&xa6$RDMBN-n<*kOcFC%Sx!>jIwlmTNIp=$b!N&YLgRpmbz8#=azy`
zVDm&By`ZB*NeSGp+vvgm?o`P-Dpb5XQv#i&<nueNptww8z)w#q)06f2VOby_;H?yB
zo=f)hp7>R{)Irlkr0{#CgBpovlqnmMLdH)#n^ltacwOK1?9lQpW}`Ju!{z6LrkxF+
zvD0W#-~OQNi9eA$mBBv6C1v+a`{8hW(sBkZ(eWsDsGP~Smx&7@tv}xX_}%8(FS4P2
zYRJQPT<!Ju4Tz9EZOgh%Zg&~3yjF~hOQb%2@tL)xHn8^O&9|B^RVW36(Id4cgkSo=
zgTeb^T#)azpGhI(6{d#h`M^p~PfxuDH}fW>+bK3A7_-$7u-&wBIALh%2<$9itsuqb
zP2+{O9Rx<uM#<)sOEXiv;2C4EUcIaHc(FFUR6y<Dzkj`df4d;H_x&b-$@D^v%u=v{
z;DciYm&br7Z8ZPchl{Z5kts(HB)i%i!8tisHj`&GWE>tnt=uAm2aRbltscKB%PvkX
zvv!1hT&`Fw$8fj);B6TA5m)5@IK6><oc>Az2J^zkckR}mzteq|JecGQ#<-olOZ(9J
z{gx-a5zIBV&c@;IXbQ~;@k`)1?wjWZvW2LoeQb<)jh55DK5e=lK2X2K=8gan(w3Fg
zEPyyJ1ezoKJqyX4E?T77*XEfjiH=F|I*H8)$FN-*eMU2g;kz`7@3~_)@BYklzt4h0
zYuU<kxCgV3TJ(S3`f(tFd6w>o-<I`3uUO!N^WQnF_!Vax-*xV(3co%~hgUq`;m?@3
zjQdlovz)<>*sg(jxIk+}G;4P<96ie)E>fI2ppj|o)1rf$;Gzz?s{LCCm`^I`oNA3e
zX|;t8F-Sk$LE-6@RP}9``jk{tFx!Dg823OY+;a9>t?pT5!DSDhL#3M8{t5M|pVtY-
zS_5@&I!<Aoj{bOf=(*$8h>W9{Y2wTh|62H%mwVSD;`M~b($ZZCllw1eI56p%&Bc@c
z+1ZO$N1JzyF0Btz&OzxS&YcaFuU~|41_Zyzbf4e65svTrRUcG?V`dQ(E6wh-J20qj
zaEUAqcKmpmVhMN3w}k&KbvjjXE&mzF0)$QM`+c^)k``@U3^meK&{Zw_vDJWreUX1x
zs-35}|4NddT%I>Ybudk6Y#~O5cTCXdWY=nIx`uu6@qp*RZiaw;^m$0z+3{A{;}<_3
zN=sWrSUnUF&`n_1Otv*PK?F5jyT_eMl=ome>ema3MiZDe{-&gR8MCqBaU5nl`>FkM
zum9MLl+OA<$g|N-b@wsN8byN>GlQKAVU``Y!)pQ#ISsc1Yg_~A=zKbzobwC%HxqB}
zi8<ANf7%3VQ}$>Jd6M-kglD+%Y9waMrU`l|+Z#XzEK9lO+W@BQa{s!g!UV6rUm#Ie
zG#_YzSH>mvt1wV=C8l$sa?!T+8xPCy6V+#wq||5JY5ek$4$q}ZTTl1ln99-F&xPgl
z*!$agUc0i-(S?IcO@8Gc1KLyi(g#iZu;1kQNBd+>Zr!}Ew5A6;E*2w~hsB{NsUv+Z
zHD$kSTBMiMgQdHdjOU~RG_ndF66=ocYg@(U%$kS{jcmf8dXZ(%>pgmLc0X(#EK*AP
zUKS-)=6QJR&q)|8+nV<^K*N+qi_DltF9_z{z4_2doTbj-1gcobQD)z@ze5{DqOXKg
zI2YUSykwx9fhYee?whR5)jIP?5qZXMYHw(_<0NR0mp)CwvX`$gqgvck(&7blmT3$l
zGA9RJrWF<K!ue<EdO19if+ikq{yJIjpgy?izm%S+U_P+y8`V>_OAwWtw6sM(J`t<#
z4rhqzDsJqGMdwY|5c@Mei{VXW6o34<>ZL+^Je#W1IpV$fB#jh}$e#vJ<UX-+Fv;c7
z3NvVeWZOxdnaC$zv9E@59}HS?$kUjBFW8GdDma)?isR???0JP9KWsvl(}8RkDP$!x
z2|9w=hhw2DUhk9V=KyP;`far&kz(8H1D1&x8~a>n)-ZSPXv6r$=faQTKEwhweC-#G
zqcWjKeY^DZ@o+2i)bKjRfUWmB!gR_U{PYH`L+@FqXB(1J5gqH$x<m{1YGT?`rGKDi
zTf1oop*i`@*~WbO8O%ANCm1i<1mzdr#qoK>SrEgFKMJTf0lj7T4XV#)%)uCr>0L)^
z!qOG{1+3!B<)-w{B2#tj+bn(^*Ju*^H-z8TYe^SA43hT~X|^bV?Tki`ALWDf$41W1
z`{>r=M+?sxeC|%Cw<T(1lg<g;j-aMxPrYuI-PlM<Gg;qk?vHDz9(EibZnYV6ZFIZM
zk!LU+b`QV<|N0vdW7hnK=Le5Qoqn#A&+<&IGA3q5j4UN%l4r|!ee5U}Eo(TbsK{9!
ztY(PboA^7R=ox7ylFjJ%a7kcZNsUxZCF+fBZ?P%gu8Z86HRQ@#{At8t0Q{{0zNR{t
z_PTV)dE-zfBZTA1BeUXXG56y0lenTN%g4V5Utjop)5Zt>1h(NGzwu+IcU?kDG4rE}
zM5gWMGwr<VCa41+2F8;x-^0`bU1FK>zi!ZpI-dZ&)n7vhcpucRI*4%~#p}6cs=#lM
zoe;t!hcsetMKE133G<BdO~#&c{$xDnoeUfN60b(|f%!I)e^zB5GhBq{cst8e6I}GQ
z<@vJrRch1dQKvkzjoXR8efxvGY`QW(ozV`c2&kvJ&i48scfW-$PskIMnehIpl~G9D
z5|2PmDpi{ApWS~z$59d%I!%!W^43I<vGgb@SDfDrJYY$e%k--sm$e={QG<_%<VIgJ
zDt5;Vjv#Peb!{j~Y}1zL#DU)PBTb8}RzA%HG^2oxJL;B_(6VS)P8kF4V@R3q3Tbc_
zJ;j|~zzS`_qCtgbBKq5XpQ2LZfZ7o(lQNy}xO>@6^EZZ>)E}L<M@I*~K;QMvWkPYg
zWrysP&MJQ@8N*K5*F)-oWmfh)4!L8Dk&3J1i<(#>ZE}YTEY>6g?zN49^5*%zxhVjF
zSbZp=hqQ}E=Mj@!TYXDw5sTWH>WO1XGE(u#xkPqr6^@=CYe0pR)z%97tp1cUv<$8j
z_uaf*>-H^<CG6enj~~g6nccqL5xg>2?21Cbf9#OyC7Zw&NmtJJerJdq2Uryo-HR!A
zlL9gDevN;Bv1DT!NT&~?Ec@H6DnKH*lp)Q=!lA6#FI>&LaleUgBo*IyGVS|a;Sxi3
z<v{OySbJw;o8ybrI>^WGtj2~E*U0lfo_=Y?qn+N4wTsSIs|o;FrmxtXebYJx=s8^D
zMXH}B(514>zaQyn7#UJF*^p3aE3_^tIbQ#<^4|8d6SZM`D)Q3ZXY;9JwosZ7#AeRl
z1!8<5XOGaPV(&&0bgB9`g=ChjdEPQlxO>~v+iP`82x#kOrEu?j?PD!sc3urT1Km{f
zW(A(4FWOw{b_dDTenI94Y)x99%PcQvaES`YJNG5`@0?Cm#sOh6LjW*~qCSTS6N4h&
z>lMPSHm59@bFU}mA*ELYipu`Z*z~t;3rtJ+Df!#mj+e64s0XHQNmRckhHz@Dh3Ib2
zHR&$=_>{09T0Kl@C+Ie3gvYeu%Z4%~Qkr+_U1v$5GOu=XuHBO?j*-C`Ws3WZ<|`*G
z?okg@gzVcq53PFsO%mJ<%zxtG;9v=jAtOOL|3%QIiMVI^ZjoBNUq9Jfh|iy!Gds=w
zEKl_?>U6%%Hc{w=PZ{VuU0wY{E)O#+UhR!o&-h+Br(QaW5Q+^=D_gZkHhHZqR#-vj
z4NRK7&y9X<9o$GjaIN1}0UG5F^1(sHRQNfTk{auM<HjN$cl(sv@b@eAG*wDT?iXD;
zbfzMz>>54qqsSKX1}S7dp_*QvvG?AVR9lv@OPn5Z98vLCH+|!yma*d7fBQ5z6(6L&
zPjdB9X62;~vw8WI8K1?OTu-s+fyP&KEBhVJ?zj7*qEd9<Lis-NSoY1eR#}sLh)M|V
zr}P%)l@`Yr!tU_e`o*43tIz?_ptnCk$GO*WLQ!KFv0s$G@7^G5Q1fm0?fq71+IdZC
ztoy<`Gr#3o+2Sv&$tp3tT)K%WBE5~zvUbP6Ob63XB%`pjLK33ISz{3JlrF!uq@=I1
z96BYILJ|_$M|R*v_qc*d_pNT&nGoc4R4*iy^_1_u5u@ZQt#lE0gLl^$RLK9ThP9$8
zC3i!ceKx>tPz32BzJSrMFsV}+mO^I;yZp*2$__D!kv?;%!jfM0`>WC`V(7g)SHIPP
zhkd|UtcFuc{~Rq)4R>qVto}AQJzZme6_T@lf>oLSD!jQoG&9k+b*uUIV=W7Rp<THL
zFnv>hgUTyq709nuo@AFd`$r$DE;Gm^+pk$S)w{)cgxpG7@a><fhYuZX^Rd?%?7c`$
z9xsFbUXdIwU5Ye4movh8-AG6@m0)Sb*$O4t|MZE(ILI*#{y43Q&)tmplq7NZKupO0
zcYd>=8|MMWj|10<w%J0N2AKxE-VdORZSZPXC9w`;x&h=5=Y_McGiP^1v|$T+YrlOv
z(llB6M%ImBJJP!5Zl^OgT;O{H({-b{vpWH)qp{c98^9gdN`}lE>3J*a=Q^Kpx@}9W
z;5WmA9BCGMX2BU#G;2h&N6S>ICYC91v6Y+CD5}aw+j2BmC2&%N5a~Q8Dt!up6q7%u
zgE&l<r!EhqvT#sc`PJfAY2m*_GYA-mY6|Z}&9JnowQP%NgBx3PLSkb61>(!a`eg>q
zcpRoK=+tuy)n=KRn)=q&Y$fGRa^qvJT!%*;+|}h354Oq7HV1BY+dh3n)*~6B79nC8
z-D2~cN(hK2&ckl0&nQ(RIG~E=(-iboiK0(Vk%7H{KhV{@py-X=KeM12soFXe@4F+d
z6M<r1_<ptbkxkKL<U^UdhJ&U3atr2#!xH}4tPP}0?u!+_EnaB2;y$<K{i$fzu+*gZ
zP2d4rbLyv=EeaLF?FBt<zm%F9P%H<#%&8>Dy<f4_W1~y%<8L-1IelREmd1b;SYcG?
z1So=rGB<ShGPdIQDHF9#mxxv(=9r*%&-E)4(jL5nYNlLtPxswN8Ea-~;<%>jdH9El
zroDi(q$9O`k!Ly~S&6X^E%Ug=Xh0?B6};MaC(+~l^P(>SBcO^>I`tEe5Fkf$pL=Rx
zd=$6%uq}{JY(;<w0<o4?y9|k}=2kT}Ha<J5gjQId|D`=!FRQHeUeh_+LA7mW4m-~^
zRuLRxqRi0wV0o~M2vX-X5zo4XVm!0MZl2P)@}`s-{r>V`JMc@(cC)I;+)MeNuIa+A
zqd=!i%E-9uJ#k5k!&CEFtE%Ud*NIfk!dG}bH>4A84JGj2rG{lPQ!p^s^}?!Qi%Uy1
zuxb@GHSVK>gWSaU`^bQz!?XT^#ZA9_W|qWX%;$d*Y;PVWG9=G##}~FeTuS0MJ>TDs
z+!2u7i@T<gZ$pAZoM1;&$KtzL1f=CW;-7$fWj@u>Ex4893wx<olSu8jh%SH*lAoJW
z_Qk<k<UVYPzxieHA*08s>U(#WvU-}nhNYG(0B=Llo5IY@UtSU!KIY)i8M6yJlWMYO
z9gB~PWzwv@yzFXf(rk3L^Nvk-Q?M@{*2b}mFBWry>jwCzd7tDwf!$qunfe#TP$>Ff
zB!~WcHs1|~a@>Q^$PcFw*$bVIys<}9&g<hPCh!2Crxzf~PSe!|29*{SdCy{Q1;RBl
zWs}b2-(7n<kOTz@y0T8xEgrzJY5RJ6uQqLc|8O&&s3hVFB&Tn@e_U*gyzR5&lia^f
z1&^ajBKCPcbK#$ng}X}z7zJS2n%r=R@_X>od>M<%hDVU!zAa0z4+_~J`#hcITTDm8
zHV5wzb9f91(A)M3EnZHom^HA(o7Bb&9MkR(20uG4<cmb}vFT3=$=mwu81F&HK3u#M
z_SwhZZ`x<07r$d%-8OCPlAtZ43$qdd!4+ZBygXRUj<LDz^Kr<_UO3iR#YA&lO<}C4
z8M$4ZU63nfI_t5BZ8Nm3q>MB08g^SNXzL5^j<KmP@mF}uXYAoI8cx{@WZ!-0nR3;)
z=r|Kq0z*IA?Dx7}s@`66^8$qFj5ie$n&0vI+fvnw?rX|ACMG6TQ)8o}i+isUv$HGV
zfoM*yvPJIqnlf0uc<-eo!^xB2uqcb5R#1+#b)Tj-@G5(8O{7UE5$?Zm+K$)pMA65)
zAIsPZJ96g9<&|z)e6ky_;6rE1BzrA?%pk_FwrMIVN&zXoti?H=HBTdA&hK1ls3gJj
zcr$tbz;+zQ0u*i)Rm|e%EDJu;w%X4%Wq;phTqOZ07l3zCaYLM#YLo?1Wf=HYdj5P{
z$#Cs2Hq}&w4>-a39!dBn!-GwYTtAsO^4}3D0jsF$ELt{;9s;p7TXP;CsGntdZlb8;
zS#Z=P=A(lHejm7TYi(Sa<FT?`lmSrulH#-*Mu+ay{JptIA5vo`AJISGVkn-@m1n?1
zcXVf0?IPsyX1wv#lBkB2d1y3)gktRu4hu-e_nMlhqWDMT^57mqDXj3LNmI6Cslem%
z<-v4z<@K>5%}*%W;;Uv;7L4;uIqVvf`r%w{^2O8$%6SRADThMtUFCWn%VK?^he>hN
zuX=<{?JWhv@d|)*^Vdl9w3bJG{GrZizl-u%G<{@v3eCEYk1X&!-b(V?EGE?H!J8{t
zXPKK7>3L8Kh<-<L0&`6;dn$t$Z{&qXWr3P>KKy+SG^=?&$d^t?rn7V0&oVAP@8k0G
zew?nOdDT><J}wc}iywI+Mkq6cs141#FIhGloodSPf(j*P{SB(?FPjK6TH`7~mr7BF
zt)YZyHCZo5@BONJ5deIzE)c!l5%fWaiJ=p?`cXRSTyeI+-7qMq72uVj3+g7j#_Cdv
z<s|qpisr{)fB)G2QvXp+HdZZuyw2HB$mQ49?@jKldy0S^bVboRgW&acMJDLv>jRC9
zhuOnLZpdbuG6t;#^d(+j#XRmNMWo4@eAO%uddsi3*t}?6Jfhe`BRgK#?s^ST#O9#o
ze&~D8J)Qu*M^rb~cQoB#t(6ESdo-+c>vrfbfZzld$<I$7?TMSfI~ki-7FWCd8MWG(
zXJbdJootD?mS=>xoazwlMLgT2>SA}Y`@9QKG%FGhh#^#}<lP@WU5^73y#X>1-a2(1
zo%lkvRGs@vgM_+>-N0uZ^x)9Y<ck=o!bU1FVG%%aKHOx#jGU}>)Rj6K*L{0Q-P#KE
zvh`p{BH(<l`>2;aTZ<Sn(E+N1xYM1!`6g@&Erq;Yx=T|;^pyc$xlfCm(-r6T*^v&}
zjmimJQu_UCV?{|RDeN3tFX4)l5LxDxpBI*ZSZdS`e22yVhja(II7ItThBya4&}-2W
zZCqvD1!N3Qa|tz|`A__k(5n})A6ng<maW2!kB@T(Cs$7eKiC;qvuY5zk+7J0XyYyM
z=YRuW(!F?B#^6UUTa)De`M!wl#X-pUMaJO6xO@JK*WrLM^z}S+7P=d%^j(c26y!yq
zmZZ*QQ2w)2zpNHuNv2ewfzQUI{>Eg*M2nxhx&Km<XOjwI36MbeZU-4b$&$t%>Tdn2
zlK1!b?;9A<1H~kvm%nA}+);?}C#(c&&sEu)Yg}XJc)GtN9&lJ}Qs-3A9!}!?Ge=I2
zY-@5cI4td(%Z%AFNE1O-CZm&ttus-(GNfRcLxF}vvt6Dj?tLtDBJ!(Yv}ho@XCC`$
z;DI_<lOcrn66nS-%X3e0{*JcUK?Ncvm0juT1g;3?dMX6O=wDSR-l`N%k~-nLq$3q@
zU|VU<Q-;#l-XVcNs_xqvKz{%2B!xhde$*%7e?1kmA5JtO$_X9$S|uqHm_ORmMH>3+
zFLt)ZigqE9b&4X-=4ji{R~`aU(az+80NFJWkFm^bYvaT`=pR_)L|6DJ)=?rZd39FG
zGxY*jpKCzUOh`kD1Ii%~PR)UF2n1sGL`~NYZDB=dY4KsZG<aRCNXb0s>*BLt_aLEB
zAA!1B(`ieJe$d}s?RO}lonL7AYU>nt36djt5&X4K{6=Xz|M~=9P3ClUT{#!5{PjKN
zNs#8TxMW)>OI-kTx-kNE*Pu@gKt(E#$usP^qKOTvNHHL@7<~Z}YV-u`L`$Hns~6iz
zVq^1msL9RFFE((1k}aW9;^~mf6!SoGgpD!3NkiE<4s1#DvDUi@dY<!9{EO~OBcMud
z{1{2{#O)xUon?O4acCh3B&F~(2N5}>@$hz?Q?D0Zbo$O@ZyTs<%aagwWeMErTc1(4
zOZTN#ecEJf4vWPvP%hq;)>Bk{W>6WzA0J<uZwZ0u3A5CMNiHxZ;cGN)-RGEAR<HLi
z#7&E;+BIY)nH<jd6oc>XTC|i7)Y~Q4g<+&15ULv&AQ0I>yEJ~U=DvY76Qq<}c@=jP
zbXVRVe_Wn1ClL1Gy`0YdbRN@wpoD-g?m~1;;foLm-LmpiIXy^H-%|2KUPJl4q}Y|V
za$VGk)HG;Xny!(sxx>O^4x8P!t_JpD%ft>SpzIYzJK`qBAS1h$Xf_f&(hcPC$(Sq*
z8?6^71oA%Z<dy$6HWi4Wto@vb!ACaB-btADcPubk2&mn^FVcrVLUjmrmf)3`rMof@
z?5D1Y`TO%b74*7{=<I*j_d|o7sNah*6Zsgwk!Drx=aS&Hp7`<mQ#znD0AXf({sKhn
z1r?Y!$d&N8<RLX+eWfN(UJo2+mbQRnj2q=cz*+wDV7h$$zpp^o!~g&K|AjGWDs3Tt
zAIyh1$>r`Z-FAb>7W5Afrwp$ouy%EK_X~cgA%=wZyn<W!!!49`%*^f%>p*y~O89OT
zZ}UPPM**i-H#5r^9UBu0Jl<U2CWYu-ovmLIjb`l?RHuv39+UzzC-e!ZBd(@5&^IBx
zJZo!fYPm$5VSlJU>H)(c_5b^UT@aM#l}BqG$3S(f!D$;zaj|2T|1uqX@5i0^d>QPT
zCV`X)p|1dAuz54_uLZo9BLK#K9{=}WM=XeH`1lb4#Q17Ie2+mYKrep0)!&Pao!!tY
z1~`R!jXO%3K|07NejJ3tg2F<B|GcgE?;#<fx1W+S{rf*_D&VjGef)1<e`DK;dv<R`
zT?CoG&>dswXLw)Oag-J5J*EzXyI%@E+`#vxXglGg=&Rsz0fz$sBnie8jVWunv-J1i
zG<Op1aE}ZEc{<$nva%zXFu;Hs|NF3xEo+tl|LNP=>t<|w7<1{L4@wycXi~6y8WgYt
zN7p`YCt?tNDGO4*byQnh#5IOef;u%UgKzy2!&r$3##}~1r(ou3GEl^1xpCf$?r08B
z(*f*y@@UGkGBGjn#@0?7K4fdQK@;$xi2eQj!{b?Z=cT?RasU1NnOeuZIF@In&Qq|i
z<LPFfG7y2|I+iy5=KYJvJ0H$Td;EP0LT$xv8#bhkpsCpQCp&gVP>6djzSq&!O-e|}
z`_uWZ%>Q6n4y8JR$dnE?afcrz6H+7~#`tW_=r_!HMx_ZkKTsC;Sx*G?E00Yzc@@Z_
zeSpchHXsW(H#c*-vjiRECo3&0Q!GP_fz`3qLk|W;z<Q_XI^Qi5261mC@MiP*K7dc3
zDVi*Sj2B5OXG9P!-KYD(9LmuP4oB+~{|uJ@f!%a{1<>-Gc7M08gB-XD^MOd;qtf$_
zRJX)VYaV!K1KdUSY|9;5mXVR+iDvwF-n3IvQow3&KOBd%q~2D#&ep5X*1L|8Fb3to
z%v<_)(ET16H!$MSd4?AL3N`k_yvJuEsGvM|#Hk|TdI3$#HK;J5ncYTX!0Z#4*J{&B
zR)Xv7+gZY>QQQ%B82J2{UdT!B1D2E;S(L-406!d6O_T6*tD3sTAfB%fb<1%iS3xeu
zvesQSnR~<%e>P>^ZC>w^n#^NbSfX3RDI}y%P`khck~Tm=1H;+*!LSf8NAr-=CAMe=
zPR7aw(SW@!TGd1Jw=iOYJuDArSJDaCJ>=o37<fSwOc`eZ!W*yM5*CpG#v7cMmP>Y>
z^OV!p4A2Nlld^*LmAksSXl5;F(43c6j9On+6{@EjxR|yhmbPSoqo)F=hU9J!Lvc07
zo6~teKHb?EdB&)pr}$}6G`ijmS=xOsu>AY)5N)rOp^44unxv$p0+9Qa0_*T>L5hQ{
zC7Ihex9xn1FW_+G+0l%1B?nk&={s&%8dDxYPM2%y4%aU!DcKKz11+=!cUUCj_~RoT
zJ-^NS`1p8MWdJyn8Dd_7A?L3M^8;XpQalPF3tEbfdRStLf#s_xE5|ve5f1dP{3l>H
zpao`hqF6gG2ZLYYgN_V}pvr;8bBc)>19F%q9o$S9Cid(g9qDXMWQ*UF6nZWXFi8NS
zgQ)_~zz%+nm*^gyZB)LKV%Y)HlCnIM$(7Py0Bl1`D&SD-$&=SEvy)-OB()ABxt@dD
zQH#JdUsiDqqn7|o)>?I5E+{H;{`>s{K|3Jt{0Z#o@cT_o2Vh@k3yk<;0)>~dx1BGe
z-df0!3m3E-xLr_CKodww^Ut6gKLmW{eJKzez~9>f@dtVixJgnH+nAuOadLG{jgW0$
zVk~BPX=!Z-if_@;(K#Ft7%m1Q^q+z-SZtvAc*fZW6^vd3jVxzi_HG{D27@*hj7Am#
z-rM_#++XsZxzB_?0bp1i#|qWijvo(xo3b({{>!5HQ7c<II59c-L4P$VW+LwyrFxp6
zay)>g*RdE%dO;-uQh<Zn+>W?$<HlhgBUUNGYW3$A`BKX^#Jzx3W^nPe$HAbJq13Q4
zO9S*afa*x%OdC;CQ&ZaG7ZI#$Yem_Uz+jU;e}35c$i_ICUf87=)TfJqKUpzU7Vn!8
zj8<{!6d?&pV7GY9j2bn8UIjK(8`A>{mypl8UTR@U5|^P93K1+fY?=y;7qnam0)*}e
zjC43AxHvZfnzidfumQ@WU#&V5KYuo|2De%L^klilvMmIDw|fL2($e0m*Y#@?WhtzV
z03GGP%HV-0va^7gds3_=qxbA)YaLmXO=|4adxU}OdwrPV?$-f%1)DOMnrmmzEqAjh
z8N{g#EGH{WL1qq~dzvpU>6=f{LP8w~#;}qN<yBsvZK&ku=T~vjR8}4VNW`XGW>|SG
z;A80tFCn)NpVxhwhX+ft+A&|Ef-sf(Zq2YMzYU*h_XbD`lyBa>Gq7VeWnlYL&Xw?6
z-l*4R4K02<*P7vAQ#ngb4UJ^hObI_;o3@UsK(t%%ZaaxdgInhD4vMdG%Cy`P6@qc?
z36|%X=KE}H)D{AY++>mWXfw*VEuq-D{2d~a5xr@`8JFs0?ev2ksp$M{mLv@4AxKGY
z-NQf8-!q>9BbZb9?7!y;2KId^I3LxsZD!%mS|#?;j553&k3Qckz3m!C7WJFvD>v8b
z#!{~z^h>IMy%UH4OY13z$bBps%^{B-Bg+=1H-Vk4#HhW)p=4l#3?i$~2?h%Q9;-6I
z0_Iuji`!(;&@W(se41X%tpA@5`#RUhekkMW^n-C?+jay2SU^6wv+4&UBH$A^T;Iap
z>YsOv1}(l@i5VH(z-4YnfEEb1=C#80?LyXja~B|2p7|sA$R3DNljoD@zNFN9dA?4?
z`B_W%j1pkoJ9@x-ik*P!oj)4Yers_Kt_6A$S(U+4WI!9|dK4cbmKEtWEY-Z<#|6~x
zijF7Wbk-<G^(#P}jDIo(ZuAXV!XUUChdFu*7}eyY@^A5&B(`}a<rfk{&(mn|Ia4XA
zCw>?}4S*xIdZ6N3f}S(-X7?0cb5><SQVb@~(m%6Rw!JU>IhN<;@usozmy~gUY7%D*
zTtz_nP>oOKv%FVr0O&<Z^L~)!^5UWr>0B9gJ})<?h=?KREQ00Bdi-&?x?;h0cVYqx
zQmNp7pq!J1MIJy{?C{Fio2xf#q4*H7XqIk|L%<w1QrgaKdx8_<<Gs5X{FLH<L^G5X
zsHWVy)~c!#5`ysEnXdq~ObLhUfA*c`Tbo$dIw@tqGWIzjRgcdn1W6#%@Nu67u8(u5
zXG@#c*bPRSKsh-%IFuWZzUi!MU*6ptC4jDx*3WuZ22j@lp|nsVLsUBxc@dIhlC*;1
zmy$9kzD)jb&VFSm^DSBokk%1!5N^<03Luow@uI!qEQ;p*Qw+la;=3+S0cg|sCq@|y
z8O-!>Z7{|cd(th#TVDaTTt~Jv0NB1GD6>Vtf1nyU01HXuw^ix!l05$PLVII}0KI_Q
z!cUje0|5CTJ2=zK5UqN7Yzk~={EIvZp&3vHe>$vscVRwoX97Tb?qpFUp*@QbzTcA3
zI_)p^%FUVuZq@0}aRJEU1&|f=JB*|jbiW24O&HQk__};J2Ix5g#yWRq_~o{Khdu)q
zq;?bU2xhQOz)vvh3INo`USA@s1)SE_*0uyRzc&AgxFcvkbXQr&$;s(MFh%E2q#vqc
z1J<HJM$0>HpB-EX0v8uAuf`{OAsr2k=zybjeIq_5`s(UxLEAoRIQDq<Y`>q+=j_l?
znKNU_@@FLqk3*Y$eI7=Lec*4)0D7#?HPyHQr3@U?T<-#;^U>iqV#?L#4cOTE=|P<v
zQk=lMeb&ddfnj<B9>xJ^MRUtlh-@8j*n;kb5nWBG0vaxZRV*4c1&D*qY(3~!D+8)Z
z8IWE02UmIPDS-i+O^p0wfaFDze&itA{sv_7IuInK@aG5Id|))dqnP#0(FKUU)6P5`
zfK9j=ZwkGjgBB9`AL{(x_#ZU)`}_SWA?uiFgF>yoG@qXbRT$SOQp%ivWpxBx>wot3
zl#r4>MgCVpi=Ax&)e%eVy5YYG3Fq~H`1Y?RuObj#ib4Oe0|+E48X5)w-F>#|rb<D1
zi=RBNxrN{}fboz0=l4Ia!+YTWya}FU$?yOB^8e}U(Bj!@+bGF<G2o{jbW0<VC3-2N
zfuC-{=)kVEgv=)X8G`Tkx;++;0?42-w*SXZjs7RI_`f)n&|<NdzeB+>#JyLwPE0^G
zm6Qx|>s$@k`>zm?3GmF-gEUKd6v)tm1pW`26|P>r>IAZk4+l%j%j=&+=6`~CIgukr
z%*M+4P(s2Kje@p;c&%>-1y1wj^kBvL{A3UC?-UU05@Ta?gk5J=vjVqA{&q#lFBiuw
z2O>cqBY_>%IgC8%?d@Hks?5gi#Xx)Nw{BpAms9P%yJ!p_0<f5#o<2kFgcRu(X(XLm
za0S3$9|Z#BOB40TC8REI-dtQ-TwDd|(-fHBynsQ%3?x!2go2@<(!bPYPWG$b+?r{2
z2ThCF$yKm&ot^whCqO_nYG*Id4`7X#^4m5;diL}54G!`gp8>92YukT|V2kw9L5=f4
z2kPJo6Ibm4<c*GwzDB$wWW69CB6JTE$PkZNqZSBL+R)1Ag;m11fQCGa@<GsRRcq6G
z0}hIFga!fbenNSTu>BG!n*~-TJWwA6_0lqs78J{cUzS6~=<RLzShcMTrWb8gwjqqt
zk2a>90Y?h}8Fwx4<{Y{7eKXKv0-^~)AOyml9)X%$?^^0q$^9i{O+Yo~d;B^03lxt*
z5uD}2g=M93vnCyYpjVo1<lMJ-_2(_<HdNp-`YKCE9*?(Xi?XBwZ{0cPTL*IquJ%>W
z!c$BD0BiX7-%xEp&*&dL4-JyGEa@iLl~&U~R?p-Fri!oIqZXt}9aF+?a}T>cBC%>&
zl46UGvtIqZ%J$&Fdfk-87@+;Y{{gQ8;R!T*4$XlW8J`bmH6a)PO+RUBBgu7^z|^HN
zn*tPfMYZB<L6w-mf#imTf=*t-frvc=4{R99k}?&&g9Q}aq{*`Y^e%F7#NzBgrUZsO
zM(X%8Ffc(dg&^^4{)#80Vrl{}%W(acI2};h4EOZ%IyxNF+yH!ZAA=<%7A?LNX@ZXT
zmCrV7hHF7$a%1KF@)755nI%>Q%p(E$ea=rjrt5hsr^-O$TLS*Y=ZBL}uII%_N1h20
zgQ1|327wl^p61u>8)Q(*DO$^*FSG7#=8lez049Fb6Tl@t|M1Y8AV9zP@liYcGR2~2
z6JW@H7ZxT!Rf5N)RtfAfnH2<slJ8=`5ea$m9rsr4Pj^HNKy8D(GuG-h0fErl$hud1
z)<>Jux;oEopVXq9t*vu~-RH}#IwLqZI7+_a0f`;OW1x6Vo#(<>;BiNQ?l=MG>+9)>
zbxiYYG9$n`fi=HsB_o;3^M^-*w17}E77RFam?+g(jj-xj^5kG=e+E#F#{~^&-eqK4
zE>wnaiErMuxPo71dIYHfRGdhV%SQo1FMg0Kj{>l#;eXdgecK2(=pMqt5xasLNENt_
z?h;g&I{6(QpAR5G<}bFN-+$j`qC}T{wf;JbA~kNl%#e?eYRJvjfkfl0wA@JQk<OBN
zCE$Q%A=opy29TjwK60oAb(z?1!YOE-{$qE)bNM|NdtzsS*&2S%#vw4<x}Z7&A|@#m
zbVpOu(z3^Nv&3}gfql?y2cGQy$?2B#UX2A#QV5P)46<t#C8dXRsD(Z1<<;TALH&@B
z5NcG3pu^9c*xAU-m))9{l4G~XJGBUm{?h0+wLaaHs?HykFs)`#TYnWoIF*TReIx_n
zMWho*XL?)}ND|=|m8}-;zq%=eU*Uhfh@lw@pnL}`yL;hkORzD>ik`S}rO~squ;h&|
zeDCT~*7@)kK7z2gL~3yN%?(CIK4oB(R@s1PLSzrC{xK?An*Qo1Zc)pj`-f{|EZ~N9
z<g}9_#SSfG`je0#e+<F=a-9CrRb@YES6aY3Q22vve2+)QS1925=8ZtPc-w?sSXej$
zQ|ttY6G3$Wj2HA%&A9`8NFOJi$7V7GBq<<RKLyXA+OXbHD@Zfa=&{g&0&LoNd*Lon
z(uy=P<5}UjvyJA3XbFO<2E?|0!9i<*IT`HX*w>glgJ^n+k!FjeHR~2T!rr)<2|Eid
zRs}5q=h%*pTAJ&OjFlsK&s5_Hvj9||#0^meuRpYlWt4Wnt^>~2id3ZxY~fQ5?<bEU
z>!VK72sl&qhOiq*E)1lKZ`<&HD4{>WKJt7wIdltw`BYlU03XpXt#=s#jZ>tSXbe^y
zq=g{%6v&+ap=7(f2f*3%1jFDr_k|!d0Tm*iwcEpcvFAELZvmncZ_P_LTLd`h55hsQ
z|2gRE{~VMTL>w?;;KD%-s#Lcq8aQJ2LLA^WfTZ*XZ-NV12YvfI=Q{+U6YT7NoE12l
zil(L^xGo{9j*DA>ZvJ`<{CXRXwoo1ga>F<{ae$3|kH|r;3MyLI-FGsHfOdmXFWrJG
z$_v5_Yt1*O03m|Br}>AOXmMBs?RDL#1UdxxImO~g88Fcj+dp#^q7t*REVf^$Ws29m
zq!Rypa`FlUB@IJ!AN|zs(gDJ>G+K=KR#_6_1=Fp6#Cq9+)seh+2;7-7Vf26&m@>N+
zy}n06L7$Jspq+8&r}P4LAAuOa)?W>h<6>~3PQSid0fHP!sHCZ9Nj81Q0S-WbPrIRv
z5}<Z)K7OoyFXkCRNC4pM1KJ;el(e1us&YhYWP`?4=+8eRE0p<1G+Gy#`Oo+go<>0b
o@V}4$t@irgjs8Es%}+eZZDMjJ+QfAW!X+syXgq~IdH(u;0VN}l@c;k-

diff --git a/dev/articles/index.html b/dev/articles/index.html
index f832dcd3..c20985c4 100644
--- a/dev/articles/index.html
+++ b/dev/articles/index.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/dev/articles/workflowsets.html b/dev/articles/workflowsets.html
index 217401d7..315cf448 100644
--- a/dev/articles/workflowsets.html
+++ b/dev/articles/workflowsets.html
@@ -36,7 +36,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/dev/authors.html b/dev/authors.html
index ef010923..42b48f96 100644
--- a/dev/authors.html
+++ b/dev/authors.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
@@ -74,13 +74,13 @@ <h2 id="citation">Citation</h2>
 
       <p>Couch S, Kuhn M (2023).
 <em>stacks: Tidy Model Stacking</em>.
-R package version 1.0.2.9000, https://github.com/tidymodels/stacks, <a href="https://stacks.tidymodels.org/">https://stacks.tidymodels.org/</a>. 
+R package version 1.0.3.9000, https://github.com/tidymodels/stacks, <a href="https://stacks.tidymodels.org/">https://stacks.tidymodels.org/</a>. 
 </p>
       <pre>@Manual{,
   title = {stacks: Tidy Model Stacking},
   author = {Simon Couch and Max Kuhn},
   year = {2023},
-  note = {R package version 1.0.2.9000, https://github.com/tidymodels/stacks},
+  note = {R package version 1.0.3.9000, https://github.com/tidymodels/stacks},
   url = {https://stacks.tidymodels.org/},
 }</pre>
     </div>
diff --git a/dev/index.html b/dev/index.html
index 4fe10a32..d79c1249 100644
--- a/dev/index.html
+++ b/dev/index.html
@@ -44,7 +44,7 @@
     
     <a class="navbar-brand me-2" href="index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/dev/news/index.html b/dev/news/index.html
index e5bb726b..a5424d18 100644
--- a/dev/news/index.html
+++ b/dev/news/index.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
@@ -54,6 +54,9 @@
 
     <div class="section level2">
 <h2 class="pkg-version" data-toc-text="development version" id="stacks-development-version">stacks (development version)<a class="anchor" aria-label="anchor" href="#stacks-development-version"></a></h2>
+</div>
+    <div class="section level2">
+<h2 class="pkg-version" data-toc-text="1.0.3" id="stacks-103">stacks 1.0.3<a class="anchor" aria-label="anchor" href="#stacks-103"></a></h2><p class="text-muted">CRAN release: 2023-11-06</p>
 <ul><li><p>Refine package alt text (<a href="https://github.com/tidymodels/stacks/issues/202" class="external-link">#202</a>).</p></li>
 <li><p>Update example objects, resolving deprecation warnings from recipes (<a href="https://github.com/tidymodels/stacks/issues/203" class="external-link">#203</a>).</p></li>
 <li><p>Fix bug in type checking for <code>blend_predictions(mixture)</code> (<a href="https://github.com/tidymodels/stacks/issues/204" class="external-link">#204</a>).</p></li>
diff --git a/dev/pkgdown.yml b/dev/pkgdown.yml
index 955e5a70..1f098f20 100644
--- a/dev/pkgdown.yml
+++ b/dev/pkgdown.yml
@@ -5,7 +5,7 @@ articles:
   workflowsets: workflowsets.html
   basics: basics.html
   classification: classification.html
-last_built: 2023-11-06T15:29Z
+last_built: 2023-11-06T19:03Z
 urls:
   reference: https://stacks.tidymodels.org/reference
   article: https://stacks.tidymodels.org/articles
diff --git a/dev/reference/add_candidates.html b/dev/reference/add_candidates.html
index 1823cbdb..a979eb43 100644
--- a/dev/reference/add_candidates.html
+++ b/dev/reference/add_candidates.html
@@ -36,7 +36,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/dev/reference/augment.model_stack.html b/dev/reference/augment.model_stack.html
index bdd194db..877c6b04 100644
--- a/dev/reference/augment.model_stack.html
+++ b/dev/reference/augment.model_stack.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/dev/reference/autoplot.linear_stack.html b/dev/reference/autoplot.linear_stack.html
index b407e61c..45a9c50f 100644
--- a/dev/reference/autoplot.linear_stack.html
+++ b/dev/reference/autoplot.linear_stack.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/dev/reference/axe_model_stack.html b/dev/reference/axe_model_stack.html
index d707f338..873ac38c 100644
--- a/dev/reference/axe_model_stack.html
+++ b/dev/reference/axe_model_stack.html
@@ -20,7 +20,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/dev/reference/blend_predictions.html b/dev/reference/blend_predictions.html
index 4cc43615..b92c08e8 100644
--- a/dev/reference/blend_predictions.html
+++ b/dev/reference/blend_predictions.html
@@ -34,7 +34,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
@@ -315,7 +315,7 @@ <h2 id="ref-examples">Examples<a class="anchor" aria-label="anchor" href="#ref-e
 <span class="r-msg co"><span class="r-pr">#&gt;</span> ── A stacked ensemble model ─────────────────────────────────────</span>
 <span class="r-msg co"><span class="r-pr">#&gt;</span> </span>
 <span class="r-msg co"><span class="r-pr">#&gt;</span> Out of 16 possible candidate members, the ensemble retained 3.</span>
-<span class="r-msg co"><span class="r-pr">#&gt;</span> Penalty: <span style="color: #0000BB;">0.001</span>.</span>
+<span class="r-msg co"><span class="r-pr">#&gt;</span> Penalty: <span style="color: #0000BB;">1e-06</span>.</span>
 <span class="r-msg co"><span class="r-pr">#&gt;</span> Mixture: <span style="color: #0000BB;">1</span>.</span>
 <span class="r-msg co"><span class="r-pr">#&gt;</span> </span>
 <span class="r-msg co"><span class="r-pr">#&gt;</span> The 3 highest weighted members are:</span>
@@ -378,6 +378,10 @@ <h2 id="ref-examples">Examples<a class="anchor" aria-label="anchor" href="#ref-e
 <span class="r-in"><span>  <span class="fu">blend_predictions</span><span class="op">(</span><span class="op">)</span></span></span>
 <span class="r-wrn co"><span class="r-pr">#&gt;</span> <span class="warning">Warning: </span>Predictions from 1 candidate were identical to those from existing</span>
 <span class="r-wrn co"><span class="r-pr">#&gt;</span> candidates and were removed from the data stack.</span>
+<span class="r-msg co"><span class="r-pr">#&gt;</span> → <span style="color: #BBBB00; font-weight: bold;">A</span> | <span style="color: #BBBB00;">warning</span>: from glmnet C++ code (error code -82); Convergence for 82th lambda value not reached after maxit=100000 iterations; solutions for larger lambdas returned</span>
+<span class="r-msg co"><span class="r-pr">#&gt;</span> There were issues with some computations   <span style="color: #BBBB00; font-weight: bold;">A</span>: x1</span>
+<span class="r-msg co"><span class="r-pr">#&gt;</span> There were issues with some computations   <span style="color: #BBBB00; font-weight: bold;">A</span>: x1</span>
+<span class="r-msg co"><span class="r-pr">#&gt;</span> </span>
 <span class="r-in"><span>  </span></span>
 <span class="r-in"><span><span class="va">class_st</span></span></span>
 <span class="r-msg co"><span class="r-pr">#&gt;</span> ── A stacked ensemble model ─────────────────────────────────────</span>
diff --git a/dev/reference/build_linear_predictor.html b/dev/reference/build_linear_predictor.html
index 3f145528..a4def358 100644
--- a/dev/reference/build_linear_predictor.html
+++ b/dev/reference/build_linear_predictor.html
@@ -14,7 +14,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/dev/reference/collect_parameters.html b/dev/reference/collect_parameters.html
index 66379f1d..1b3ee87a 100644
--- a/dev/reference/collect_parameters.html
+++ b/dev/reference/collect_parameters.html
@@ -16,7 +16,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/dev/reference/control_stack.html b/dev/reference/control_stack.html
index a474fba4..cb1cfbf6 100644
--- a/dev/reference/control_stack.html
+++ b/dev/reference/control_stack.html
@@ -26,7 +26,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/dev/reference/example_data.html b/dev/reference/example_data.html
index 92b5177a..79e9a654 100644
--- a/dev/reference/example_data.html
+++ b/dev/reference/example_data.html
@@ -12,7 +12,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/dev/reference/fit_members.html b/dev/reference/fit_members.html
index 5ec45ab3..573433b6 100644
--- a/dev/reference/fit_members.html
+++ b/dev/reference/fit_members.html
@@ -20,7 +20,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
@@ -235,16 +235,17 @@ <h2 id="ref-examples">Examples<a class="anchor" aria-label="anchor" href="#ref-e
 <span class="r-in"><span><span class="va">log_st</span></span></span>
 <span class="r-msg co"><span class="r-pr">#&gt;</span> ── A stacked ensemble model ─────────────────────────────────────</span>
 <span class="r-msg co"><span class="r-pr">#&gt;</span> </span>
-<span class="r-msg co"><span class="r-pr">#&gt;</span> Out of 11 possible candidate members, the ensemble retained 2.</span>
-<span class="r-msg co"><span class="r-pr">#&gt;</span> Penalty: <span style="color: #0000BB;">0.01</span>.</span>
+<span class="r-msg co"><span class="r-pr">#&gt;</span> Out of 11 possible candidate members, the ensemble retained 3.</span>
+<span class="r-msg co"><span class="r-pr">#&gt;</span> Penalty: <span style="color: #0000BB;">1e-04</span>.</span>
 <span class="r-msg co"><span class="r-pr">#&gt;</span> Mixture: <span style="color: #0000BB;">1</span>.</span>
 <span class="r-msg co"><span class="r-pr">#&gt;</span> </span>
-<span class="r-msg co"><span class="r-pr">#&gt;</span> The 2 highest weighted member classes are:</span>
-<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #949494;"># A tibble: 2 × 3</span></span>
+<span class="r-msg co"><span class="r-pr">#&gt;</span> The 3 highest weighted member classes are:</span>
+<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #949494;"># A tibble: 3 × 3</span></span>
 <span class="r-out co"><span class="r-pr">#&gt;</span>   member                   type        weight</span>
 <span class="r-out co"><span class="r-pr">#&gt;</span>   <span style="color: #949494; font-style: italic;">&lt;chr&gt;</span>                    <span style="color: #949494; font-style: italic;">&lt;chr&gt;</span>        <span style="color: #949494; font-style: italic;">&lt;dbl&gt;</span></span>
-<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;">1</span> .pred_no_log_res_nn_1_1  mlp           7.08</span>
-<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;">2</span> .pred_no_log_res_rf_1_05 rand_forest   3.10</span>
+<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;">1</span> .pred_no_log_res_nn_1_1  mlp         7.41  </span>
+<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;">2</span> .pred_no_log_res_rf_1_05 rand_forest 3.44  </span>
+<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;">3</span> .pred_no_log_res_rf_1_02 rand_forest 0.063<span style="text-decoration: underline;">8</span></span>
 </code></pre></div>
     </div>
   </main><aside class="col-md-3"><nav id="toc"><h2>On this page</h2>
diff --git a/dev/reference/get_expressions.html b/dev/reference/get_expressions.html
index 421a6234..8b10869a 100644
--- a/dev/reference/get_expressions.html
+++ b/dev/reference/get_expressions.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/dev/reference/index.html b/dev/reference/index.html
index 567ec716..0244da38 100644
--- a/dev/reference/index.html
+++ b/dev/reference/index.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/dev/reference/predict.data_stack.html b/dev/reference/predict.data_stack.html
index 8674a763..1de58842 100644
--- a/dev/reference/predict.data_stack.html
+++ b/dev/reference/predict.data_stack.html
@@ -12,7 +12,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/dev/reference/predict.model_stack.html b/dev/reference/predict.model_stack.html
index a09eaa18..c3bca3ec 100644
--- a/dev/reference/predict.model_stack.html
+++ b/dev/reference/predict.model_stack.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
@@ -157,17 +157,17 @@ <h2 id="ref-examples">Examples<a class="anchor" aria-label="anchor" href="#ref-e
 <span class="r-msg co"><span class="r-pr">#&gt;</span> ── A stacked ensemble model ─────────────────────────────────────</span>
 <span class="r-msg co"><span class="r-pr">#&gt;</span> </span>
 <span class="r-msg co"><span class="r-pr">#&gt;</span> Out of 11 possible candidate members, the ensemble retained 4.</span>
-<span class="r-msg co"><span class="r-pr">#&gt;</span> Penalty: <span style="color: #0000BB;">0.1</span>.</span>
+<span class="r-msg co"><span class="r-pr">#&gt;</span> Penalty: <span style="color: #0000BB;">1e-06</span>.</span>
 <span class="r-msg co"><span class="r-pr">#&gt;</span> Mixture: <span style="color: #0000BB;">1</span>.</span>
 <span class="r-msg co"><span class="r-pr">#&gt;</span> </span>
 <span class="r-msg co"><span class="r-pr">#&gt;</span> The 4 highest weighted members are:</span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #949494;"># A tibble: 4 × 3</span></span>
 <span class="r-out co"><span class="r-pr">#&gt;</span>   member          type       weight</span>
 <span class="r-out co"><span class="r-pr">#&gt;</span>   <span style="color: #949494; font-style: italic;">&lt;chr&gt;</span>           <span style="color: #949494; font-style: italic;">&lt;chr&gt;</span>       <span style="color: #949494; font-style: italic;">&lt;dbl&gt;</span></span>
-<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;">1</span> reg_res_sp_03_1 linear_reg 0.483 </span>
-<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;">2</span> reg_res_sp_10_1 linear_reg 0.249 </span>
-<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;">3</span> reg_res_lr_1_1  linear_reg 0.126 </span>
-<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;">4</span> reg_res_sp_05_1 linear_reg 0.067<span style="text-decoration: underline;">3</span></span>
+<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;">1</span> reg_res_sp_03_1 linear_reg 0.485 </span>
+<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;">2</span> reg_res_sp_10_1 linear_reg 0.247 </span>
+<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;">3</span> reg_res_lr_1_1  linear_reg 0.129 </span>
+<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;">4</span> reg_res_sp_05_1 linear_reg 0.066<span style="text-decoration: underline;">6</span></span>
 <span class="r-in"><span></span></span>
 <span class="r-in"><span><span class="co"># predict on the tree frogs testing data</span></span></span>
 <span class="r-in"><span><span class="fu"><a href="https://rdrr.io/r/stats/predict.html" class="external-link">predict</a></span><span class="op">(</span><span class="va">reg_st</span>, <span class="va">tree_frogs_reg_test</span><span class="op">)</span></span></span>
@@ -175,14 +175,14 @@ <h2 id="ref-examples">Examples<a class="anchor" aria-label="anchor" href="#ref-e
 <span class="r-out co"><span class="r-pr">#&gt;</span>    .pred</span>
 <span class="r-out co"><span class="r-pr">#&gt;</span>    <span style="color: #949494; font-style: italic;">&lt;dbl&gt;</span></span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 1</span> 119. </span>
-<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 2</span>  81.5</span>
+<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 2</span>  81.4</span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 3</span> 102. </span>
-<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 4</span>  35.6</span>
+<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 4</span>  35.5</span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 5</span> 119. </span>
-<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 6</span>  50.6</span>
+<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 6</span>  50.5</span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 7</span> 122. </span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 8</span>  82.7</span>
-<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 9</span>  50.3</span>
+<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 9</span>  50.2</span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;">10</span>  75.7</span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #949494;"># ℹ 133 more rows</span></span>
 <span class="r-in"><span></span></span>
@@ -192,14 +192,14 @@ <h2 id="ref-examples">Examples<a class="anchor" aria-label="anchor" href="#ref-e
 <span class="r-out co"><span class="r-pr">#&gt;</span>    .pred reg_res_lr_1_1 reg_res_sp_10_1 reg_res_sp_05_1 reg_res_sp_03_1</span>
 <span class="r-out co"><span class="r-pr">#&gt;</span>    <span style="color: #949494; font-style: italic;">&lt;dbl&gt;</span>          <span style="color: #949494; font-style: italic;">&lt;dbl&gt;</span>           <span style="color: #949494; font-style: italic;">&lt;dbl&gt;</span>           <span style="color: #949494; font-style: italic;">&lt;dbl&gt;</span>           <span style="color: #949494; font-style: italic;">&lt;dbl&gt;</span></span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 1</span> 119.           138.            125.            121.            114. </span>
-<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 2</span>  81.5           82.4            84.8            81.8            77.1</span>
+<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 2</span>  81.4           82.4            84.8            81.8            77.1</span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 3</span> 102.           116.            111.            112.             93.3</span>
-<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 4</span>  35.6           35.8            29.7            32.5            29.6</span>
+<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 4</span>  35.5           35.8            29.7            32.5            29.6</span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 5</span> 119.           111.            115.            115.            127. </span>
-<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 6</span>  50.6           38.8            37.4            36.2            55.3</span>
+<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 6</span>  50.5           38.8            37.4            36.2            55.3</span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 7</span> 122.           123.            103.            104.            137. </span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 8</span>  82.7           82.3            78.6            82.0            82.8</span>
-<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 9</span>  50.3           38.7            37.3            36.2            54.8</span>
+<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 9</span>  50.2           38.7            37.3            36.2            54.8</span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;">10</span>  75.7           78.8            75.3            76.9            71.8</span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #949494;"># ℹ 133 more rows</span></span>
 <span class="r-in"><span></span></span>
@@ -254,10 +254,10 @@ <h2 id="ref-examples">Examples<a class="anchor" aria-label="anchor" href="#ref-e
 <span class="r-out co"><span class="r-pr">#&gt;</span>         <span style="color: #949494; font-style: italic;">&lt;dbl&gt;</span>     <span style="color: #949494; font-style: italic;">&lt;dbl&gt;</span>     <span style="color: #949494; font-style: italic;">&lt;dbl&gt;</span></span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 1</span>     0.909     0.054<span style="text-decoration: underline;">0</span>    0.037<span style="text-decoration: underline;">1</span></span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 2</span>     0.101     0.434     0.465 </span>
-<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 3</span>     0.092<span style="text-decoration: underline;">3</span>    0.419     0.489 </span>
-<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 4</span>     0.105     0.429     0.466 </span>
+<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 3</span>     0.090<span style="text-decoration: underline;">0</span>    0.421     0.489 </span>
+<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 4</span>     0.104     0.430     0.466 </span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 5</span>     0.909     0.054<span style="text-decoration: underline;">1</span>    0.037<span style="text-decoration: underline;">3</span></span>
-<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 6</span>     0.909     0.054<span style="text-decoration: underline;">0</span>    0.037<span style="text-decoration: underline;">1</span></span>
+<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 6</span>     0.909     0.054<span style="text-decoration: underline;">0</span>    0.037<span style="text-decoration: underline;">2</span></span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 7</span>     0.909     0.054<span style="text-decoration: underline;">0</span>    0.037<span style="text-decoration: underline;">1</span></span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 8</span>     0.909     0.054<span style="text-decoration: underline;">0</span>    0.037<span style="text-decoration: underline;">1</span></span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 9</span>     0.909     0.054<span style="text-decoration: underline;">0</span>    0.037<span style="text-decoration: underline;">1</span></span>
@@ -276,10 +276,10 @@ <h2 id="ref-examples">Examples<a class="anchor" aria-label="anchor" href="#ref-e
 <span class="r-out co"><span class="r-pr">#&gt;</span>         <span style="color: #949494; font-style: italic;">&lt;dbl&gt;</span>     <span style="color: #949494; font-style: italic;">&lt;dbl&gt;</span>     <span style="color: #949494; font-style: italic;">&lt;dbl&gt;</span>                       <span style="color: #949494; font-style: italic;">&lt;dbl&gt;</span></span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 1</span>     0.909     0.054<span style="text-decoration: underline;">0</span>    0.037<span style="text-decoration: underline;">1</span>                       0    </span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 2</span>     0.101     0.434     0.465                        0.339</span>
-<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 3</span>     0.092<span style="text-decoration: underline;">3</span>    0.419     0.489                        0.212</span>
-<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 4</span>     0.105     0.429     0.466                        0.318</span>
+<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 3</span>     0.090<span style="text-decoration: underline;">0</span>    0.421     0.489                        0.220</span>
+<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 4</span>     0.104     0.430     0.466                        0.323</span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 5</span>     0.909     0.054<span style="text-decoration: underline;">1</span>    0.037<span style="text-decoration: underline;">3</span>                       0    </span>
-<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 6</span>     0.909     0.054<span style="text-decoration: underline;">0</span>    0.037<span style="text-decoration: underline;">1</span>                       0    </span>
+<span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 6</span>     0.909     0.054<span style="text-decoration: underline;">0</span>    0.037<span style="text-decoration: underline;">2</span>                       0    </span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 7</span>     0.909     0.054<span style="text-decoration: underline;">0</span>    0.037<span style="text-decoration: underline;">1</span>                       0    </span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 8</span>     0.909     0.054<span style="text-decoration: underline;">0</span>    0.037<span style="text-decoration: underline;">1</span>                       0    </span>
 <span class="r-out co"><span class="r-pr">#&gt;</span> <span style="color: #BCBCBC;"> 9</span>     0.909     0.054<span style="text-decoration: underline;">0</span>    0.037<span style="text-decoration: underline;">1</span>                       0    </span>
diff --git a/dev/reference/prediction_eqn.html b/dev/reference/prediction_eqn.html
index a70a3950..6c72d102 100644
--- a/dev/reference/prediction_eqn.html
+++ b/dev/reference/prediction_eqn.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/dev/reference/reexports.html b/dev/reference/reexports.html
index 8152e3af..18e26f26 100644
--- a/dev/reference/reexports.html
+++ b/dev/reference/reexports.html
@@ -48,7 +48,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/dev/reference/stack_predict.html b/dev/reference/stack_predict.html
index e11accdc..45197df6 100644
--- a/dev/reference/stack_predict.html
+++ b/dev/reference/stack_predict.html
@@ -10,7 +10,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
@@ -86,10 +86,6 @@ <h2 id="arguments">Arguments<a class="anchor" aria-label="anchor" href="#argumen
 <dt>...</dt>
 <dd><p>Not currently used.</p></dd>
 
-
-<dt>type</dt>
-<dd><p>The prediction type.</p></dd>
-
 </dl></div>
     <div class="section level2">
     <h2 id="value">Value<a class="anchor" aria-label="anchor" href="#value"></a></h2>
diff --git a/dev/reference/stacks.html b/dev/reference/stacks.html
index b66e1b99..100b02f0 100644
--- a/dev/reference/stacks.html
+++ b/dev/reference/stacks.html
@@ -30,7 +30,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/dev/reference/stacks_description.html b/dev/reference/stacks_description.html
index 15513c17..21b40b55 100644
--- a/dev/reference/stacks_description.html
+++ b/dev/reference/stacks_description.html
@@ -12,7 +12,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/dev/reference/tree_frogs.html b/dev/reference/tree_frogs.html
index 77057795..6ffa246f 100644
--- a/dev/reference/tree_frogs.html
+++ b/dev/reference/tree_frogs.html
@@ -30,7 +30,7 @@
     
     <a class="navbar-brand me-2" href="../index.html">stacks</a>
 
-    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.2.9000</small>
+    <small class="nav-text text-danger me-auto" data-bs-toggle="tooltip" data-bs-placement="bottom" title="In-development version">1.0.3.9000</small>
 
     
     <button class="navbar-toggler" type="button" data-bs-toggle="collapse" data-bs-target="#navbar" aria-controls="navbar" aria-expanded="false" aria-label="Toggle navigation">
diff --git a/dev/search.json b/dev/search.json
index 9102bb84..d42ab845 100644
--- a/dev/search.json
+++ b/dev/search.json
@@ -1 +1 @@
-[{"path":[]},{"path":"https://stacks.tidymodels.org/dev/CODE_OF_CONDUCT.html","id":"our-pledge","dir":"","previous_headings":"","what":"Our Pledge","title":"Contributor Covenant Code of Conduct","text":"members, contributors, leaders pledge make participation community harassment-free experience everyone, regardless age, body size, visible invisible disability, ethnicity, sex characteristics, gender identity expression, level experience, education, socio-economic status, nationality, personal appearance, race, caste, color, religion, sexual identity orientation. pledge act interact ways contribute open, welcoming, diverse, inclusive, healthy community.","code":""},{"path":"https://stacks.tidymodels.org/dev/CODE_OF_CONDUCT.html","id":"our-standards","dir":"","previous_headings":"","what":"Our Standards","title":"Contributor Covenant Code of Conduct","text":"Examples behavior contributes positive environment community include: Demonstrating empathy kindness toward people respectful differing opinions, viewpoints, experiences Giving gracefully accepting constructive feedback Accepting responsibility apologizing affected mistakes, learning experience Focusing best just us individuals, overall community Examples unacceptable behavior include: use sexualized language imagery, sexual attention advances kind Trolling, insulting derogatory comments, personal political attacks Public private harassment Publishing others’ private information, physical email address, without explicit permission conduct reasonably considered inappropriate professional setting","code":""},{"path":"https://stacks.tidymodels.org/dev/CODE_OF_CONDUCT.html","id":"enforcement-responsibilities","dir":"","previous_headings":"","what":"Enforcement Responsibilities","title":"Contributor Covenant Code of Conduct","text":"Community leaders responsible clarifying enforcing standards acceptable behavior take appropriate fair corrective action response behavior deem inappropriate, threatening, offensive, harmful. Community leaders right responsibility remove, edit, reject comments, commits, code, wiki edits, issues, contributions aligned Code Conduct, communicate reasons moderation decisions appropriate.","code":""},{"path":"https://stacks.tidymodels.org/dev/CODE_OF_CONDUCT.html","id":"scope","dir":"","previous_headings":"","what":"Scope","title":"Contributor Covenant Code of Conduct","text":"Code Conduct applies within community spaces, also applies individual officially representing community public spaces. Examples representing community include using official e-mail address, posting via official social media account, acting appointed representative online offline event.","code":""},{"path":"https://stacks.tidymodels.org/dev/CODE_OF_CONDUCT.html","id":"enforcement","dir":"","previous_headings":"","what":"Enforcement","title":"Contributor Covenant Code of Conduct","text":"Instances abusive, harassing, otherwise unacceptable behavior may reported community leaders responsible enforcement codeofconduct@posit.co. complaints reviewed investigated promptly fairly. community leaders obligated respect privacy security reporter incident.","code":""},{"path":"https://stacks.tidymodels.org/dev/CODE_OF_CONDUCT.html","id":"enforcement-guidelines","dir":"","previous_headings":"","what":"Enforcement Guidelines","title":"Contributor Covenant Code of Conduct","text":"Community leaders follow Community Impact Guidelines determining consequences action deem violation Code Conduct:","code":""},{"path":"https://stacks.tidymodels.org/dev/CODE_OF_CONDUCT.html","id":"id_1-correction","dir":"","previous_headings":"Enforcement Guidelines","what":"1. Correction","title":"Contributor Covenant Code of Conduct","text":"Community Impact: Use inappropriate language behavior deemed unprofessional unwelcome community. Consequence: private, written warning community leaders, providing clarity around nature violation explanation behavior inappropriate. public apology may requested.","code":""},{"path":"https://stacks.tidymodels.org/dev/CODE_OF_CONDUCT.html","id":"id_2-warning","dir":"","previous_headings":"Enforcement Guidelines","what":"2. Warning","title":"Contributor Covenant Code of Conduct","text":"Community Impact: violation single incident series actions. Consequence: warning consequences continued behavior. interaction people involved, including unsolicited interaction enforcing Code Conduct, specified period time. includes avoiding interactions community spaces well external channels like social media. Violating terms may lead temporary permanent ban.","code":""},{"path":"https://stacks.tidymodels.org/dev/CODE_OF_CONDUCT.html","id":"id_3-temporary-ban","dir":"","previous_headings":"Enforcement Guidelines","what":"3. Temporary Ban","title":"Contributor Covenant Code of Conduct","text":"Community Impact: serious violation community standards, including sustained inappropriate behavior. Consequence: temporary ban sort interaction public communication community specified period time. public private interaction people involved, including unsolicited interaction enforcing Code Conduct, allowed period. Violating terms may lead permanent ban.","code":""},{"path":"https://stacks.tidymodels.org/dev/CODE_OF_CONDUCT.html","id":"id_4-permanent-ban","dir":"","previous_headings":"Enforcement Guidelines","what":"4. Permanent Ban","title":"Contributor Covenant Code of Conduct","text":"Community Impact: Demonstrating pattern violation community standards, including sustained inappropriate behavior, harassment individual, aggression toward disparagement classes individuals. Consequence: permanent ban sort public interaction within community.","code":""},{"path":"https://stacks.tidymodels.org/dev/CODE_OF_CONDUCT.html","id":"attribution","dir":"","previous_headings":"","what":"Attribution","title":"Contributor Covenant Code of Conduct","text":"Code Conduct adapted Contributor Covenant, version 2.1, available https://www.contributor-covenant.org/version/2/1/code_of_conduct.html. Community Impact Guidelines inspired [Mozilla’s code conduct enforcement ladder][https://github.com/mozilla/inclusion]. answers common questions code conduct, see FAQ https://www.contributor-covenant.org/faq. Translations available https://www.contributor-covenant.org/translations.","code":""},{"path":"https://stacks.tidymodels.org/dev/LICENSE.html","id":null,"dir":"","previous_headings":"","what":"MIT License","title":"MIT License","text":"Copyright (c) 2023 stacks authors Permission hereby granted, free charge, person obtaining copy software associated documentation files (“Software”), deal Software without restriction, including without limitation rights use, copy, modify, merge, publish, distribute, sublicense, /sell copies Software, permit persons Software furnished , subject following conditions: copyright notice permission notice shall included copies substantial portions Software. SOFTWARE PROVIDED “”, WITHOUT WARRANTY KIND, EXPRESS IMPLIED, INCLUDING LIMITED WARRANTIES MERCHANTABILITY, FITNESS PARTICULAR PURPOSE NONINFRINGEMENT. EVENT SHALL AUTHORS COPYRIGHT HOLDERS LIABLE CLAIM, DAMAGES LIABILITY, WHETHER ACTION CONTRACT, TORT OTHERWISE, ARISING , CONNECTION SOFTWARE USE DEALINGS SOFTWARE.","code":""},{"path":"https://stacks.tidymodels.org/dev/articles/basics.html","id":"define-candidate-ensemble-members","dir":"Articles","previous_headings":"","what":"Define candidate ensemble members","title":"Getting Started With stacks","text":"highest level, ensembles formed model definitions. package, model definitions instance minimal workflow, containing model specification (defined parsnip package) , optionally, preprocessor (defined recipes package). Model definitions specify form candidate ensemble members.  Defining constituent model definitions undoubtedly longest part building ensemble stacks. ’re familiar tidymodels “proper,” ’re probably fine skip section, keeping things mind: ’ll need save assessment set predictions workflow utilized tune_grid(), tune_bayes(), fit_resamples() objects setting control arguments save_pred = TRUE save_workflow = TRUE. Note use control_stack_*() convenience functions ! model definition must share rsample rset object. ’ll first start splitting training data, generating resamples, setting options used model definition. Tuning fitting results use ensembles need fitted control arguments save_pred = TRUE save_workflow = TRUE—settings ensure assessment set predictions, well workflow used fit resamples, stored resulting object. convenience, stacks supplies control_stack_*() functions generate appropriate objects . example, ’ll working tune_grid() fit_resamples() tune package, use following control settings: ’ll define three different model definitions try predict time hatch—K-nearest neighbors model (hyperparameters tune), linear model, support vector machine model (, hyperparameters tune). Starting K-nearest neighbors, begin creating parsnip model specification: Note , since tuning several possible numbers neighbors, model specification defines multiple model configurations. specific form configurations determined specifying grid search tune_grid(). , extend basic recipe defined earlier fully specify form design matrix use K-nearest neighbors model: Starting basic recipe, convert categorical variables dummy variables, remove column one observation, impute missing values numeric variables using mean, normalize numeric predictors. Pre-processing instructions remaining models defined similarly. Now, combine model specification pre-processing instructions defined form workflow object: Finally, can make use workflow, training set resamples, metric set, control object tune hyperparameters. Using grid argument, specify like optimize four possible values k using grid search. knn_res object fully specifies candidate members, ready included stacks workflow. Now, specifying linear model, note optimizing hyperparameters. Thus, use fit_resamples() function rather tune_grid() tune_bayes() fitting resamples. Finally, putting together model definition support vector machine: Altogether, ’ve created three model definitions, K-nearest neighbors model definition specifies 4 model configurations, linear regression specifies 1, support vector machine specifies 6.  three model definitions fully specified, ready begin stacking model configurations. (Note , applied settings, one likely specify many 11 candidate members.)","code":"# some setup: resampling and a basic recipe set.seed(1) tree_frogs_split <- initial_split(tree_frogs) tree_frogs_train <- training(tree_frogs_split) tree_frogs_test  <- testing(tree_frogs_split)  set.seed(1) folds <- rsample::vfold_cv(tree_frogs_train, v = 5)  tree_frogs_rec <-    recipe(latency ~ ., data = tree_frogs_train)  metric <- metric_set(rmse) ctrl_grid <- control_stack_grid() ctrl_res <- control_stack_resamples() # create a model definition knn_spec <-   nearest_neighbor(     mode = \"regression\",      neighbors = tune(\"k\")   ) %>%   set_engine(\"kknn\")  knn_spec #> K-Nearest Neighbor Model Specification (regression) #>  #> Main Arguments: #>   neighbors = tune(\"k\") #>  #> Computational engine: kknn # extend the recipe knn_rec <-   tree_frogs_rec %>%   step_dummy(all_nominal_predictors()) %>%   step_zv(all_predictors()) %>%   step_impute_mean(all_numeric_predictors()) %>%   step_normalize(all_numeric_predictors())  knn_rec #>  #> ── Recipe ──────────────────────────────────────────────────────────────── #>  #> ── Inputs #> Number of variables by role #> outcome:   1 #> predictor: 4 #>  #> ── Operations #> • Dummy variables from: all_nominal_predictors() #> • Zero variance filter on: all_predictors() #> • Mean imputation for: all_numeric_predictors() #> • Centering and scaling for: all_numeric_predictors() # add both to a workflow knn_wflow <-    workflow() %>%    add_model(knn_spec) %>%   add_recipe(knn_rec)  knn_wflow #> ══ Workflow ══════════════════════════════════════════════════════════════ #> Preprocessor: Recipe #> Model: nearest_neighbor() #>  #> ── Preprocessor ────────────────────────────────────────────────────────── #> 4 Recipe Steps #>  #> • step_dummy() #> • step_zv() #> • step_impute_mean() #> • step_normalize() #>  #> ── Model ───────────────────────────────────────────────────────────────── #> K-Nearest Neighbor Model Specification (regression) #>  #> Main Arguments: #>   neighbors = tune(\"k\") #>  #> Computational engine: kknn # tune k and fit to the 5-fold cv set.seed(2020) knn_res <-    tune_grid(     knn_wflow,     resamples = folds,     metrics = metric,     grid = 4,     control = ctrl_grid   )  knn_res #> # Tuning results #> # 5-fold cross-validation  #> # A tibble: 5 × 5 #>   splits           id    .metrics         .notes           .predictions #>   <list>           <chr> <list>           <list>           <list>       #> 1 <split [343/86]> Fold1 <tibble [4 × 5]> <tibble [0 × 3]> <tibble>     #> 2 <split [343/86]> Fold2 <tibble [4 × 5]> <tibble [0 × 3]> <tibble>     #> 3 <split [343/86]> Fold3 <tibble [4 × 5]> <tibble [0 × 3]> <tibble>     #> 4 <split [343/86]> Fold4 <tibble [4 × 5]> <tibble [0 × 3]> <tibble>     #> 5 <split [344/85]> Fold5 <tibble [4 × 5]> <tibble [0 × 3]> <tibble> # create a model definition lin_reg_spec <-   linear_reg() %>%   set_engine(\"lm\")  # extend the recipe lin_reg_rec <-   tree_frogs_rec %>%   step_dummy(all_nominal_predictors()) %>%   step_zv(all_predictors())  # add both to a workflow lin_reg_wflow <-    workflow() %>%   add_model(lin_reg_spec) %>%   add_recipe(lin_reg_rec)  # fit to the 5-fold cv set.seed(2020) lin_reg_res <-    fit_resamples(     lin_reg_wflow,     resamples = folds,     metrics = metric,     control = ctrl_res   )  lin_reg_res #> # Resampling results #> # 5-fold cross-validation  #> # A tibble: 5 × 5 #>   splits           id    .metrics         .notes           .predictions #>   <list>           <chr> <list>           <list>           <list>       #> 1 <split [343/86]> Fold1 <tibble [1 × 4]> <tibble [0 × 3]> <tibble>     #> 2 <split [343/86]> Fold2 <tibble [1 × 4]> <tibble [0 × 3]> <tibble>     #> 3 <split [343/86]> Fold3 <tibble [1 × 4]> <tibble [0 × 3]> <tibble>     #> 4 <split [343/86]> Fold4 <tibble [1 × 4]> <tibble [0 × 3]> <tibble>     #> 5 <split [344/85]> Fold5 <tibble [1 × 4]> <tibble [0 × 3]> <tibble> # create a model definition svm_spec <-    svm_rbf(     cost = tune(\"cost\"),      rbf_sigma = tune(\"sigma\")   ) %>%   set_engine(\"kernlab\") %>%   set_mode(\"regression\")  # extend the recipe svm_rec <-   tree_frogs_rec %>%   step_dummy(all_nominal_predictors()) %>%   step_zv(all_predictors()) %>%   step_impute_mean(all_numeric_predictors()) %>%   step_corr(all_predictors()) %>%   step_normalize(all_numeric_predictors())  # add both to a workflow svm_wflow <-    workflow() %>%    add_model(svm_spec) %>%   add_recipe(svm_rec)  # tune cost and sigma and fit to the 5-fold cv set.seed(2020) svm_res <-    tune_grid(     svm_wflow,      resamples = folds,      grid = 6,     metrics = metric,     control = ctrl_grid   )  svm_res #> # Tuning results #> # 5-fold cross-validation  #> # A tibble: 5 × 5 #>   splits           id    .metrics         .notes           .predictions #>   <list>           <chr> <list>           <list>           <list>       #> 1 <split [343/86]> Fold1 <tibble [6 × 6]> <tibble [0 × 3]> <tibble>     #> 2 <split [343/86]> Fold2 <tibble [6 × 6]> <tibble [0 × 3]> <tibble>     #> 3 <split [343/86]> Fold3 <tibble [6 × 6]> <tibble [0 × 3]> <tibble>     #> 4 <split [343/86]> Fold4 <tibble [6 × 6]> <tibble [0 × 3]> <tibble>     #> 5 <split [344/85]> Fold5 <tibble [6 × 6]> <tibble [0 × 3]> <tibble>"},{"path":"https://stacks.tidymodels.org/dev/articles/basics.html","id":"putting-together-a-stack","dir":"Articles","previous_headings":"","what":"Putting together a stack","title":"Getting Started With stacks","text":"first step building ensemble stacks create data_stack object—package, data stacks tibbles (extra attributes) contain assessment set predictions candidate ensemble member.  can initialize data stack using stacks() function. stacks() function works sort like ggplot() constructor ggplot2—function creates basic structure object built top —except ’ll pipe outputs rather adding +. add_candidates() function adds ensemble members stack. mentioned , hood, data_stack object really just tibble extra attributes. Checking actual data: first column gives first response value, remaining columns give assessment set predictions ensemble member. Since ’re regression case, ’s one column per ensemble member. classification settings, many columns levels outcome variable per candidate ensemble member. ’s ! ’re now ready evaluate need combine predictions candidate ensemble member.","code":"stacks() #> # A data stack with 0 model definitions and 0 candidate members. tree_frogs_data_st <-    stacks() %>%   add_candidates(knn_res) %>%   add_candidates(lin_reg_res) %>%   add_candidates(svm_res)  tree_frogs_data_st #> # A data stack with 3 model definitions and 11 candidate members: #> #   knn_res: 4 model configurations #> #   lin_reg_res: 1 model configuration #> #   svm_res: 6 model configurations #> # Outcome: latency (numeric) as_tibble(tree_frogs_data_st) #> # A tibble: 429 × 12 #>    latency knn_res_1_1 knn_res_1_2 knn_res_1_3 knn_res_1_4 lin_reg_res_1_1 #>      <dbl>       <dbl>       <dbl>       <dbl>       <dbl>           <dbl> #>  1     142        76.5        67.8        83.2        83.0           114.  #>  2      79        74.3        71.0        64.2        63.9            78.6 #>  3      50        75.0        71.4        64.9        64.3            81.5 #>  4      68        47.9        49.7        54.0        54.0            78.6 #>  5      64        60.9        53.1        48.4        48.3            36.5 #>  6      52       207.        168.        129.        124.            124.  #>  7      39        22.6        24.9        31.0        31.5            35.2 #>  8      46        23.2        26.6        33.4        33.9            37.1 #>  9     137        75.9        78.8        90.4        90.6            78.8 #> 10      73        67.7        73.6        67.9        67.4            38.8 #> # ℹ 419 more rows #> # ℹ 6 more variables: svm_res_1_1 <dbl>, svm_res_1_4 <dbl>, #> #   svm_res_1_3 <dbl>, svm_res_1_5 <dbl>, svm_res_1_2 <dbl>, #> #   svm_res_1_6 <dbl>"},{"path":"https://stacks.tidymodels.org/dev/articles/basics.html","id":"fit-the-stack","dir":"Articles","previous_headings":"","what":"Fit the stack","title":"Getting Started With stacks","text":"outputs candidate ensemble members highly correlated, blend_predictions() function performs regularization figure can combine outputs stack members come final prediction. blend_predictions function determines member model output ultimately combined final prediction fitting LASSO model data stack, predicting true assessment set outcome using predictions candidate members. Candidates nonzero stacking coefficients become members.  make sure right trade-minimizing number members optimizing performance, can use autoplot() method:  show relationship directly:  results good enough, blend_predictions() called different values penalty. , blend_predictions() picks penalty parameter numerically optimal results. see top results:  Now know combine model output, can fit candidates non-zero stacking coefficients full training set.  Model stacks can thought group fitted member models set instructions combine predictions.  identify model configurations assigned stacking coefficients, can make use collect_parameters() function: object now ready predict new data! Juxtaposing predictions true data:  Looks like predictions pretty strong! stacks predictions perform, though, compared members’ predictions? can use type = \"members\" argument generate predictions ensemble members. Now, evaluating root mean squared error model: can see, stacked ensemble outperforms member models, though closely followed one members. Voila! ’ve now made use stacks package predict red-eyed tree frog embryo hatching using stacked ensemble! full visual outline steps can found .","code":"tree_frogs_model_st <-   tree_frogs_data_st %>%   blend_predictions() autoplot(tree_frogs_model_st) autoplot(tree_frogs_model_st, type = \"members\") autoplot(tree_frogs_model_st, type = \"weights\") tree_frogs_model_st <-   tree_frogs_model_st %>%   fit_members() collect_parameters(tree_frogs_model_st, \"svm_res\") #> # A tibble: 6 × 4 #>   member         cost    sigma  coef #>   <chr>         <dbl>    <dbl> <dbl> #> 1 svm_res_1_1 0.00143 6.64e- 9 0     #> 2 svm_res_1_2 3.59    3.95e- 4 0.931 #> 3 svm_res_1_3 0.0978  1.81e- 2 0     #> 4 svm_res_1_4 0.00849 2.16e-10 0     #> 5 svm_res_1_5 0.256   4.54e- 1 0.159 #> 6 svm_res_1_6 7.64    3.16e- 7 0 tree_frogs_test <-    tree_frogs_test %>%   bind_cols(predict(tree_frogs_model_st, .)) ggplot(tree_frogs_test) +   aes(x = latency,        y = .pred) +   geom_point() +    coord_obs_pred() member_preds <-    tree_frogs_test %>%   select(latency) %>%   bind_cols(predict(tree_frogs_model_st, tree_frogs_test, members = TRUE)) map(member_preds, rmse_vec, truth = member_preds$latency) %>%   as_tibble() #> # A tibble: 1 × 6 #>   latency .pred knn_res_1_4 lin_reg_res_1_1 svm_res_1_5 svm_res_1_2 #>     <dbl> <dbl>       <dbl>           <dbl>       <dbl>       <dbl> #> 1       0  54.3        56.1            55.5        56.9        60.6"},{"path":"https://stacks.tidymodels.org/dev/articles/classification.html","id":"defining-candidate-ensemble-members","dir":"Articles","previous_headings":"","what":"Defining candidate ensemble members","title":"Classification Models With stacks","text":"numeric prediction setting, defining candidate ensemble members undoubtedly longest part ensembling process stacks. First, splitting training data, generating resamples, setting options used model definition. also need use control settings numeric response setting: ’ll define two different model definitions try predict reflex—random forest neural network. Starting random forest: Now, moving neural network model definition: model definitions fully specified, ’re ready start putting together ensemble!","code":"# some setup: resampling and a basic recipe set.seed(1)  tree_frogs_split <- initial_split(tree_frogs) tree_frogs_train <- training(tree_frogs_split) tree_frogs_test  <- testing(tree_frogs_split)  folds <- rsample::vfold_cv(tree_frogs_train, v = 5)  tree_frogs_rec <-    recipe(reflex ~ ., data = tree_frogs_train) %>%   step_dummy(all_nominal_predictors(), -reflex) %>%   step_zv(all_predictors())  tree_frogs_wflow <-    workflow() %>%    add_recipe(tree_frogs_rec) ctrl_grid <- control_stack_grid() rand_forest_spec <-    rand_forest(     mtry = tune(),     min_n = tune(),     trees = 500   ) %>%   set_mode(\"classification\") %>%   set_engine(\"ranger\")  rand_forest_wflow <-   tree_frogs_wflow %>%   add_model(rand_forest_spec)  rand_forest_res <-    tune_grid(     object = rand_forest_wflow,      resamples = folds,      grid = 10,     control = ctrl_grid   ) nnet_spec <-   mlp(hidden_units = tune(), penalty = tune(), epochs = tune()) %>%   set_mode(\"classification\") %>%   set_engine(\"nnet\")  nnet_rec <-    tree_frogs_rec %>%    step_normalize(all_predictors())  nnet_wflow <-    tree_frogs_wflow %>%   add_model(nnet_spec) %>%   update_recipe(nnet_rec)  nnet_res <-   tune_grid(     object = nnet_wflow,      resamples = folds,      grid = 10,     control = ctrl_grid   )"},{"path":"https://stacks.tidymodels.org/dev/articles/classification.html","id":"putting-together-a-stack","dir":"Articles","previous_headings":"","what":"Putting together a stack","title":"Classification Models With stacks","text":"Building stacked ensemble, now, takes lines: make sure right trade-minimizing number members optimizing performance, can use autoplot() method:  show relationship directly:  results good enough, blend_predictions() called different values penalty. , blend_predictions() picks penalty parameter numerically optimal results. see top results:  multiple facets since ensemble members can different effects different classes. identify model configurations assigned stacking coefficients, can make use collect_parameters() function: object now ready predict new data! Computing ROC AUC model: Looks like predictions pretty strong! stacks predictions perform, though, compared members’ predictions? can use members argument generate predictions ensemble members. Voilà! ’ve now made use stacks package predict tree frog embryo ear function using stacked ensemble!","code":"tree_frogs_model_st <-    # initialize the stack   stacks() %>%   # add candidate members   add_candidates(rand_forest_res) %>%   add_candidates(nnet_res) %>%   # determine how to combine their predictions   blend_predictions() %>%   # fit the candidates with nonzero stacking coefficients   fit_members()  tree_frogs_model_st #> # A tibble: 10 × 4 #>    member                         type        weight class #>    <chr>                          <chr>        <dbl> <fct> #>  1 .pred_full_nnet_res_1_04       mlp          21.6  full  #>  2 .pred_mid_nnet_res_1_04        mlp          18.2  mid   #>  3 .pred_mid_rand_forest_res_1_05 rand_forest  13.6  low   #>  4 .pred_mid_rand_forest_res_1_08 rand_forest   6.47 mid   #>  5 .pred_mid_nnet_res_1_01        mlp           4.60 low   #>  6 .pred_mid_nnet_res_1_03        mlp           4.58 low   #>  7 .pred_mid_rand_forest_res_1_07 rand_forest   4.07 low   #>  8 .pred_mid_nnet_res_1_10        mlp           3.49 mid   #>  9 .pred_full_nnet_res_1_01       mlp           3.09 full  #> 10 .pred_mid_nnet_res_1_09        mlp           1.43 mid autoplot(tree_frogs_model_st) autoplot(tree_frogs_model_st, type = \"members\") autoplot(tree_frogs_model_st, type = \"weights\") collect_parameters(tree_frogs_model_st, \"rand_forest_res\") #> # A tibble: 60 × 6 #>    member                mtry min_n class terms                       coef #>    <chr>                <int> <int> <chr> <chr>                      <dbl> #>  1 rand_forest_res_1_01     1    26 low   .pred_mid_rand_forest_res…     0 #>  2 rand_forest_res_1_01     1    26 low   .pred_full_rand_forest_re…     0 #>  3 rand_forest_res_1_01     1    26 mid   .pred_mid_rand_forest_res…     0 #>  4 rand_forest_res_1_01     1    26 mid   .pred_full_rand_forest_re…     0 #>  5 rand_forest_res_1_01     1    26 full  .pred_mid_rand_forest_res…     0 #>  6 rand_forest_res_1_01     1    26 full  .pred_full_rand_forest_re…     0 #>  7 rand_forest_res_1_02     2    33 low   .pred_mid_rand_forest_res…     0 #>  8 rand_forest_res_1_02     2    33 low   .pred_full_rand_forest_re…     0 #>  9 rand_forest_res_1_02     2    33 mid   .pred_mid_rand_forest_res…     0 #> 10 rand_forest_res_1_02     2    33 mid   .pred_full_rand_forest_re…     0 #> # ℹ 50 more rows tree_frogs_pred <-   tree_frogs_test %>%   bind_cols(predict(tree_frogs_model_st, ., type = \"prob\")) yardstick::roc_auc(   tree_frogs_pred,   truth = reflex,   contains(\".pred_\")   ) tree_frogs_pred <-   tree_frogs_test %>%   select(reflex) %>%   bind_cols(     predict(       tree_frogs_model_st,       tree_frogs_test,       type = \"class\",       members = TRUE       )     )  tree_frogs_pred #> # A tibble: 303 × 15 #>    reflex .pred_class .pred_class_rand_forest_res…¹ .pred_class_rand_for…² #>    <fct>  <fct>       <fct>                         <fct>                  #>  1 full   full        full                          full                   #>  2 mid    mid         low                           mid                    #>  3 mid    mid         mid                           mid                    #>  4 mid    mid         low                           low                    #>  5 full   full        full                          full                   #>  6 full   full        full                          full                   #>  7 full   full        full                          full                   #>  8 full   full        full                          full                   #>  9 full   full        full                          full                   #> 10 full   full        full                          full                   #> # ℹ 293 more rows #> # ℹ abbreviated names: ¹​.pred_class_rand_forest_res_1_05, #> #   ²​.pred_class_rand_forest_res_1_07 #> # ℹ 11 more variables: .pred_class_nnet_res_1_07 <fct>, #> #   .pred_class_nnet_res_1_02 <fct>, .pred_class_nnet_res_1_03 <fct>, #> #   .pred_class_nnet_res_1_01 <fct>, #> #   .pred_class_rand_forest_res_1_09 <fct>, …  map(   colnames(tree_frogs_pred),   ~mean(tree_frogs_pred$reflex == pull(tree_frogs_pred, .x)) ) %>%   set_names(colnames(tree_frogs_pred)) %>%   as_tibble() %>%   pivot_longer(c(everything(), -reflex)) #> # A tibble: 14 × 3 #>    reflex name                             value #>     <dbl> <chr>                            <dbl> #>  1      1 .pred_class                      0.884 #>  2      1 .pred_class_rand_forest_res_1_05 0.855 #>  3      1 .pred_class_rand_forest_res_1_07 0.865 #>  4      1 .pred_class_nnet_res_1_07        0.868 #>  5      1 .pred_class_nnet_res_1_02        0.848 #>  6      1 .pred_class_nnet_res_1_03        0.845 #>  7      1 .pred_class_nnet_res_1_01        0.842 #>  8      1 .pred_class_rand_forest_res_1_09 0.881 #>  9      1 .pred_class_rand_forest_res_1_08 0.881 #> 10      1 .pred_class_nnet_res_1_05        0.789 #> 11      1 .pred_class_nnet_res_1_09        0.835 #> 12      1 .pred_class_nnet_res_1_06        0.858 #> 13      1 .pred_class_nnet_res_1_04        0.855 #> 14      1 .pred_class_nnet_res_1_10        0.851"},{"path":"https://stacks.tidymodels.org/dev/articles/workflowsets.html","id":"setup","dir":"Articles","previous_headings":"","what":"Setup","title":"Stacking With Workflow Sets","text":"example parallel “Getting Started” vignette, except use workflowsets bundle model workflows define candidate members one workflow set. train using workflowsets::workflow_map() add results model stack using add_candidates(). ’re familiar stacks package, refer vignette big picture! First, loading needed packages: example, ’ll make use tree_frogs data exported stacks, giving experimental results hatching behavior red-eyed tree frog embryos! Red-eyed tree frog (RETF) embryos can hatch earlier normal 7ish days detect potential predator threat. Researchers wanted determine , , tree frog embryos able detect stimulus environment. , subjected embryos varying developmental stages “predator stimulus” jiggling embryos blunt probe. Beforehand, though embryos treated gentamicin, compound knocks lateral line (sensory organ.) Researcher Julie Jung crew found factors inform whether embryo hatches prematurely ! ’ll start predicting latency (.e. time hatch) based attributes. ’ll need filter NAs (.e. cases embryo hatch) first. Taking quick look data, seems like hatch time pretty closely related predictors!  Let’s give go!","code":"library(tidymodels) library(stacks) data(\"tree_frogs\")  # subset the data tree_frogs <- tree_frogs %>%   filter(!is.na(latency)) %>%   select(-c(clutch, hatched)) theme_set(theme_bw())  ggplot(tree_frogs) +   aes(x = age, y = latency, color = treatment) +   geom_point() +   labs(x = \"Embryo Age (s)\", y = \"Time to Hatch (s)\", col = \"Treatment\")"},{"path":"https://stacks.tidymodels.org/dev/articles/workflowsets.html","id":"defining-candidate-ensemble-members","dir":"Articles","previous_headings":"","what":"Defining candidate ensemble members","title":"Stacking With Workflow Sets","text":"workflow-based setting, defining candidate ensemble members undoubtedly longest part ensembling process stacks. Instead evaluating model specifications via resampling workflow, though, combine model specifications workflow_set(), evaluate batch workflow_map() function. First, splitting training data, generating resamples, setting options used model definition: also need use control settings workflow-based setting: ’ll define three different model definitions try predict time hatch—K-nearest neighbors model (hyperparameters tune), linear model, support vector machine model (, hyperparameters tune). Starting K-nearest neighbors, begin creating parsnip model specification: Note , since tuning several possible numbers neighbors, model specification defines multiple model configurations. specific form configurations determined specifying grid search. , extend basic recipe defined earlier fully specify form design matrix use K-nearest neighbors model: Starting basic recipe, convert categorical predictors dummy variables, remove predictors one observation, impute missing values numeric predictors using mean, normalize numeric predictors. Pre-processing instructions remaining models defined similarly. Now, specifying linear model, note optimizing hyperparameters. Finally, putting together model specification recipe support vector machine: model specification accompanying recipe now defined, can combine via workflow_set: Note combination preprocessor model specification assigned wflow_id can use interface individual model definitions: elements workflow set nearly ready evaluated tune_grid(). , though, need ensure fit appropriate control options, just evaluating individual workflows resamples passing stacks. can iteratively add appropriate control settings control_stack_grid() shorthand using option_add() function: can now use workflow_map() function map model definition, evaluating hyperparameters supplied resamples. results section now contains three sets tuning results. Note results corresponding linear regression subclass resample_results rather tune_results—expected, hyperparameters tune model specification. workflow_map() knows fall back fit_resamples() rather tune_grid(), case! can extract tuning results extract_workflow_set_result() helper explore tuning results: three model definitions fully specified tuned workflow set, ready begin stacking model configurations. (Note , applied settings, one likely specify many handful candidate members three model definitions.)","code":"# some setup: resampling and a basic recipe set.seed(1) tree_frogs_split <- initial_split(tree_frogs) tree_frogs_train <- training(tree_frogs_split) tree_frogs_test  <- testing(tree_frogs_split)  set.seed(1) folds <- rsample::vfold_cv(tree_frogs_train, v = 5)  tree_frogs_rec <-    recipe(latency ~ ., data = tree_frogs_train)  metric <- metric_set(rmse) ctrl_grid <- control_stack_grid() # create a model specification knn_spec <-   nearest_neighbor(     mode = \"regression\",      neighbors = tune(\"k\")   ) %>%   set_engine(\"kknn\")  knn_spec #> K-Nearest Neighbor Model Specification (regression) #>  #> Main Arguments: #>   neighbors = tune(\"k\") #>  #> Computational engine: kknn # extend the recipe knn_rec <-   tree_frogs_rec %>%   step_dummy(all_nominal_predictors()) %>%   step_zv(all_predictors()) %>%   step_impute_mean(all_numeric_predictors()) %>%   step_normalize(all_numeric_predictors())  knn_rec #>  #> ── Recipe ──────────────────────────────────────────────────────────────── #>  #> ── Inputs #> Number of variables by role #> outcome:   1 #> predictor: 4 #>  #> ── Operations #> • Dummy variables from: all_nominal_predictors() #> • Zero variance filter on: all_predictors() #> • Mean imputation for: all_numeric_predictors() #> • Centering and scaling for: all_numeric_predictors() # create a model specification lin_reg_spec <-   linear_reg() %>%   set_engine(\"lm\")  # extend the recipe lin_reg_rec <-   tree_frogs_rec %>%   step_dummy(all_nominal_predictors()) %>%   step_zv(all_predictors()) # create a model specification svm_spec <-    svm_rbf(     cost = tune(\"cost\"),      rbf_sigma = tune(\"sigma\")   ) %>%   set_engine(\"kernlab\") %>%   set_mode(\"regression\")  # extend the recipe svm_rec <-   tree_frogs_rec %>%   step_dummy(all_nominal_predictors()) %>%   step_zv(all_predictors()) %>%   step_impute_mean(all_numeric_predictors()) %>%   step_corr(all_predictors()) %>%   step_normalize(all_numeric_predictors()) wf_set <-    workflow_set(     preproc = list(rec1 = knn_rec, rec2 = lin_reg_rec,     rec3 = svm_rec),     models =  list(knn = knn_spec, lin_reg = lin_reg_spec, svm = svm_spec),     cross = FALSE   )  wf_set #> # A workflow set/tibble: 3 × 4 #>   wflow_id     info             option    result     #>   <chr>        <list>           <list>    <list>     #> 1 rec1_knn     <tibble [1 × 4]> <opts[0]> <list [0]> #> 2 rec2_lin_reg <tibble [1 × 4]> <opts[0]> <list [0]> #> 3 rec3_svm     <tibble [1 × 4]> <opts[0]> <list [0]> wf_set %>%   extract_workflow(\"rec3_svm\") #> ══ Workflow ══════════════════════════════════════════════════════════════ #> Preprocessor: Recipe #> Model: svm_rbf() #>  #> ── Preprocessor ────────────────────────────────────────────────────────── #> 5 Recipe Steps #>  #> • step_dummy() #> • step_zv() #> • step_impute_mean() #> • step_corr() #> • step_normalize() #>  #> ── Model ───────────────────────────────────────────────────────────────── #> Radial Basis Function Support Vector Machine Model Specification (regression) #>  #> Main Arguments: #>   cost = tune(\"cost\") #>   rbf_sigma = tune(\"sigma\") #>  #> Computational engine: kernlab wf_set <-   wf_set %>%   option_add(     control = control_stack_grid(),     metrics = metric   ) wf_set_trained <-   workflow_map(     wf_set,     fn = \"tune_grid\",     resamples = folds   )  wf_set_trained #> # A workflow set/tibble: 3 × 4 #>   wflow_id     info             option    result    #>   <chr>        <list>           <list>    <list>    #> 1 rec1_knn     <tibble [1 × 4]> <opts[3]> <tune[+]> #> 2 rec2_lin_reg <tibble [1 × 4]> <opts[3]> <rsmp[+]> #> 3 rec3_svm     <tibble [1 × 4]> <opts[3]> <tune[+]> extract_workflow_set_result(wf_set_trained, id = \"rec1_knn\") %>%   collect_metrics(summarize = TRUE) #> # A tibble: 10 × 7 #>        k .metric .estimator  mean     n std_err .config               #>    <int> <chr>   <chr>      <dbl> <int>   <dbl> <chr>                 #>  1     1 rmse    standard    81.3     5    4.37 Preprocessor1_Model01 #>  2     3 rmse    standard    69.6     5    3.57 Preprocessor1_Model02 #>  3     5 rmse    standard    63.7     5    3.30 Preprocessor1_Model03 #>  4     6 rmse    standard    62.3     5    3.12 Preprocessor1_Model04 #>  5     7 rmse    standard    61.2     5    2.95 Preprocessor1_Model05 #>  6     9 rmse    standard    59.8     5    2.82 Preprocessor1_Model06 #>  7    10 rmse    standard    59.3     5    2.83 Preprocessor1_Model07 #>  8    11 rmse    standard    59.0     5    2.82 Preprocessor1_Model08 #>  9    13 rmse    standard    58.5     5    2.84 Preprocessor1_Model09 #> 10    15 rmse    standard    58.2     5    2.88 Preprocessor1_Model10"},{"path":"https://stacks.tidymodels.org/dev/articles/workflowsets.html","id":"putting-together-a-stack","dir":"Articles","previous_headings":"","what":"Putting together a stack","title":"Stacking With Workflow Sets","text":"Building stacked ensemble, now, takes even fewer lines individual workflows: results obtained building model stack workflow sets identical result building model stack individual workflows. make sure right trade-minimizing number members optimizing performance, can use:  results good enough, blend_predictions() called different values penalty. , blend_predictions() picks penalty parameter numerically optimal results. see top results:  identify model configurations assigned stacking coefficients, can make use collect_parameters() function: object now ready predict new data! Juxtaposing predictions true data:  Looks like predictions decent! stacks predictions perform, though, compared members’ predictions? can use type = \"members\" argument generate predictions ensemble members. Now, evaluating root mean squared error model: can see, stacked ensemble outperforms member models, though closely followed one members. Voila! ’ve now made use stacks workflowsets packages predict red-eyed tree frog embryo hatching using stacked ensemble!","code":"tree_frogs_model_st <-    # initialize the stack   stacks() %>%   # add candidate members   add_candidates(wf_set_trained) %>%   # determine how to combine their predictions   blend_predictions() %>%   # fit the candidates with nonzero stacking coefficients   fit_members()  tree_frogs_model_st #> # A tibble: 4 × 3 #>   member           type             weight #>   <chr>            <chr>             <dbl> #> 1 rec3_svm_1_06    svm_rbf          0.779  #> 2 rec2_lin_reg_1_1 linear_reg       0.275  #> 3 rec1_knn_1_08    nearest_neighbor 0.190  #> 4 rec1_knn_1_07    nearest_neighbor 0.0205 autoplot(tree_frogs_model_st) autoplot(tree_frogs_model_st, type = \"weights\") collect_parameters(tree_frogs_model_st, \"rec3_svm\") #> # A tibble: 10 × 4 #>    member            cost    sigma  coef #>    <chr>            <dbl>    <dbl> <dbl> #>  1 rec3_svm_1_01  0.185   1.17e- 4 0     #>  2 rec3_svm_1_02  1.23    2.39e- 7 0     #>  3 rec3_svm_1_03  0.0195  3.37e- 5 0     #>  4 rec3_svm_1_04 15.2     4.78e- 8 0     #>  5 rec3_svm_1_05  2.97    1.30e- 6 0     #>  6 rec3_svm_1_06  0.134   5.62e- 2 0.779 #>  7 rec3_svm_1_07  0.00163 3.01e- 3 0     #>  8 rec3_svm_1_08  4.08    2.98e- 1 0     #>  9 rec3_svm_1_09  0.00429 3.45e-10 0     #> 10 rec3_svm_1_10  0.0280  1.54e- 9 0 tree_frogs_test <-    tree_frogs_test %>%   bind_cols(predict(tree_frogs_model_st, .)) ggplot(tree_frogs_test) +   aes(x = latency,        y = .pred) +   geom_point() +    coord_obs_pred() member_preds <-    tree_frogs_test %>%   select(latency) %>%   bind_cols(predict(tree_frogs_model_st, tree_frogs_test, members = TRUE)) map(member_preds, rmse_vec, truth = member_preds$latency) %>%   as_tibble() #> # A tibble: 1 × 6 #>   latency .pred rec1_knn_1_07 rec1_knn_1_08 rec2_lin_reg_1_1 rec3_svm_1_06 #>     <dbl> <dbl>         <dbl>         <dbl>            <dbl>         <dbl> #> 1       0  53.9          56.2          56.2             55.5          57.3"},{"path":"https://stacks.tidymodels.org/dev/authors.html","id":null,"dir":"","previous_headings":"","what":"Authors","title":"Authors and Citation","text":"Simon Couch. Author, maintainer. Max Kuhn. Author. . Copyright holder, funder.","code":""},{"path":"https://stacks.tidymodels.org/dev/authors.html","id":"citation","dir":"","previous_headings":"","what":"Citation","title":"Authors and Citation","text":"Couch S, Kuhn M (2023). stacks: Tidy Model Stacking. R package version 1.0.2.9000, https://github.com/tidymodels/stacks, https://stacks.tidymodels.org/.","code":"@Manual{,   title = {stacks: Tidy Model Stacking},   author = {Simon Couch and Max Kuhn},   year = {2023},   note = {R package version 1.0.2.9000, https://github.com/tidymodels/stacks},   url = {https://stacks.tidymodels.org/}, }"},{"path":"https://stacks.tidymodels.org/dev/index.html","id":"stacks---tidy-model-stacking-","dir":"","previous_headings":"","what":"stacks - tidy model stacking","title":"Tidy Model Stacking","text":"stacks R package model stacking aligns tidymodels. Model stacking ensembling method takes outputs many models combines generate new model—referred ensemble package—generates predictions informed members. process goes something like : Define candidate ensemble members using functionality rsample, parsnip, workflows, recipes, tune Initialize data_stack object stacks() Iteratively add candidate ensemble members data_stack add_candidates() Evaluate combine predictions blend_predictions() Fit candidate ensemble members non-zero stacking coefficients fit_members() Predict new data predict() can install package following code: Install development version : stacks generalized respect : Model type: model type implemented parsnip extension packages fair game add stacks model stack. ’s table many implemented model types tidymodels core, link article implementing model classes well. Cross-validation scheme: resampling algorithm implemented rsample extension packages fair game resampling data use training model stack. Error metric: metric function implemented yardstick extension packages fair game evaluating model stacks members. package provides infrastructure creating metric functions well! stacks uses regularized linear model combine predictions ensemble members, though model type one many possible learning algorithms used fit stacked ensemble model. implementations additional ensemble learning algorithms, check h2o SuperLearner. Rather diving right implementation, ’ll focus pieces fit together, conceptually, building ensemble stacks. See basics vignette example API action!","code":"install.packages(\"stacks\") # install.packages(\"pak\") pak::pak(\"tidymodels/stacks\")"},{"path":"https://stacks.tidymodels.org/dev/index.html","id":"a-grammar","dir":"","previous_headings":"","what":"a grammar","title":"Tidy Model Stacking","text":"highest level, ensembles formed model definitions. package, model definitions instance minimal workflow, containing model specification (defined parsnip package) , optionally, preprocessor (defined recipes package). Model definitions specify form candidate ensemble members. diagram representing “model definitions,” specify form candidate ensemble members. Three colored boxes represent three different model types; K-nearest neighbors model (salmon), linear regression model (yellow), support vector machine model (green). used ensemble, model definitions must share resample. rsample rset object, paired model definitions, can used generate tuning/fitting results objects candidate ensemble members tune. diagram representing “candidate members” generated model definition. Four salmon-colored boxes labeled “KNN” represent K-nearest neighbors models trained resamples differing hyperparameters. Similarly, linear regression model generates one candidate member, support vector machine model generates six. Candidate members first come together data_stack object add_candidates() function. Principally, objects just tibbles, first column gives true outcome assessment set (portion training set used model validation), remaining columns give predictions candidate ensemble member. (outcome numeric, ’s one column per candidate ensemble member. Classification requires many columns per candidate levels outcome variable.) also bring along extra attributes keep track model definitions. diagram representing “data stack,” specific kind data frame. Colored “columns” depict, white, true value outcome variable validation set, followed four columns (salmon) representing predictions K-nearest neighbors model, one column (tan) representing linear regression model, six (green) representing support vector machine model. , data stack can evaluated using blend_predictions() determine best combine outputs candidate members. stacking literature, process commonly called metalearning. outputs member likely highly correlated. Thus, depending degree regularization choose, coefficients inputs (possibly) many members zero —predictions influence final output, terms thus thrown . diagram representing “stacking coefficients,” coefficients linear model combining candidate member predictions generate ensemble’s ultimate prediction. Boxes candidate members placed besides , filled color coefficient associated candidate member nonzero. stacking coefficients determine candidate ensemble members become ensemble members. Candidates non-zero stacking coefficients fitted whole training set, altogether making model_stack object. diagram representing “model stack” class, collates stacking coefficients members (candidate members nonzero stacking coefficients trained full training set). representation stacking coefficients , members (shown next associated stacking coefficients) colored-pentagons. Model stacks list subclass. model stack object, outputted fit_members(), ready predict new data! trained ensemble members often referred base models stacking literature. full visual outline steps can found . API package closely mirrors ideas. See basics vignette example grammar implemented!","code":""},{"path":"https://stacks.tidymodels.org/dev/index.html","id":"contributing","dir":"","previous_headings":"","what":"contributing","title":"Tidy Model Stacking","text":"project released Contributor Code Conduct. contributing project, agree abide terms. questions discussions tidymodels packages, modeling, machine learning, please post Posit Community. think encountered bug, please submit issue. Either way, learn create share reprex (minimal, reproducible example), clearly communicate code. Check details contributing guidelines tidymodels packages get help. stacks package, test objects take long build every commit. contribution changes structure data_stack model_stacks objects, please regenerate test objects running scripts man-roxygen/example_models.Rmd, including chunk options eval = FALSE.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/add_candidates.html","id":null,"dir":"Reference","previous_headings":"","what":"Add model definitions to a data stack — add_candidates","title":"Add model definitions to a data stack — add_candidates","text":"add_candidates() collates assessment set predictions additional attributes supplied model definition (.e. set \"candidates\") data stack. Behind scenes, data stack objects just tibble::tbl_dfs, first column gives true response values, remaining columns give assessment set predictions candidate. regression setting, one column per ensemble member. classification settings, many columns per candidate ensemble member levels outcome variable. initialize data stack, use stacks() function. Model definitions appended data stack iteratively using several calls add_candidates(). Data stacks evaluated using blend_predictions() function.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/add_candidates.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Add model definitions to a data stack — add_candidates","text":"","code":"add_candidates(   data_stack,   candidates,   name = deparse(substitute(candidates)),   ... )"},{"path":"https://stacks.tidymodels.org/dev/reference/add_candidates.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Add model definitions to a data stack — add_candidates","text":"data_stack data_stack object. candidates (set ) model definition(s) defining candidate model stack members. inherit tune_results workflow_set. tune_results: object outputted tune::tune_grid(), tune::tune_bayes(), tune::fit_resamples(). workflow_set: object outputted workflowsets::workflow_map(). approach allows supplying multiple sets candidate members one call add_candidates. See \"Stacking Workflow Sets\" article package website example code! Regardless, results must fitted control settings save_pred = TRUE, save_workflow = TRUE—see control_stack_grid(), control_stack_bayes(), control_stack_resamples() documentation helper functions. name label model definition---defaults name candidates object. Ignored candidates inherits workflow_set. ... Additional arguments. Currently ignored.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/add_candidates.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Add model definitions to a data stack — add_candidates","text":"data_stack object--see stacks() details!","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/add_candidates.html","id":"example-data","dir":"Reference","previous_headings":"","what":"Example Data","title":"Add model definitions to a data stack — add_candidates","text":"package provides resampling objects datasets use examples vignettes derived study 1212 red-eyed tree frog embryos! Red-eyed tree frog (RETF) embryos can hatch earlier normal 7ish days detect potential predator threat. Researchers wanted determine , , tree frog embryos able detect stimulus environment. , subjected embryos varying developmental stages \"predator stimulus\" jiggling embryos blunt probe. Beforehand, though embryos treated gentamicin, compound knocks lateral line (sensory organ.) Researcher Julie Jung crew found factors inform whether embryo hatches prematurely ! Note data included stacks package necessarily representative unbiased subset complete dataset, demonstrative purposes. reg_folds class_folds rset cross-fold validation objects rsample, splitting training data regression classification model objects, respectively. tree_frogs_reg_test tree_frogs_class_test analogous testing sets. reg_res_lr, reg_res_svm, reg_res_sp contain regression tuning results linear regression, support vector machine, spline model, respectively, fitting latency (.e. long embryos took hatch response jiggle) tree_frogs data, using variables predictors. Note data underlying models filtered include data embryos hatched response stimulus. class_res_rf class_res_nn contain multiclass classification tuning results random forest neural network classification model, respectively, fitting reflex (measure ear function) data using variables predictors. log_res_rf log_res_nn, contain binary classification tuning results random forest neural network classification model, respectively, fitting hatched (whether embryos hatched response stimulus) using variables predictors. See ?example_data learn objects, well browse source code generated .","code":""},{"path":[]},{"path":"https://stacks.tidymodels.org/dev/reference/add_candidates.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Add model definitions to a data stack — add_candidates","text":"","code":"# see the \"Example Data\" section above for # clarification on the objects used in these examples!  # put together a data stack using # tuning results for regression models reg_st <-    stacks() %>%   add_candidates(reg_res_lr) %>%   add_candidates(reg_res_svm) %>%   add_candidates(reg_res_sp)    reg_st #> # A data stack with 3 model definitions and 16 candidate members: #> #   reg_res_lr: 1 model configuration #> #   reg_res_svm: 5 model configurations #> #   reg_res_sp: 10 model configurations #> # Outcome: latency (numeric)    # do the same with multinomial classification models class_st <-   stacks() %>%   add_candidates(class_res_nn) %>%   add_candidates(class_res_rf) #> Warning: Predictions from 1 candidate were identical to those from existing #> candidates and were removed from the data stack.    class_st #> # A data stack with 2 model definitions and 10.6666666666667 candidate members: #> #   class_res_nn: 1 model configuration #> #   class_res_rf: 9.66666666666667 model configurations #> # Outcome: reflex (factor)    # ...or binomial classification models log_st <-   stacks() %>%   add_candidates(log_res_nn) %>%   add_candidates(log_res_rf)    log_st #> # A data stack with 2 model definitions and 11 candidate members: #> #   log_res_nn: 1 model configuration #> #   log_res_rf: 10 model configurations #> # Outcome: hatched (factor)    # use custom names for each model: log_st2 <-   stacks() %>%   add_candidates(log_res_nn, name = \"neural_network\") %>%   add_candidates(log_res_rf, name = \"random_forest\")    log_st2 #> # A data stack with 2 model definitions and 11 candidate members: #> #   neural_network: 1 model configuration #> #   random_forest: 10 model configurations #> # Outcome: hatched (factor)    # these objects would likely then be # passed to blend_predictions(): log_st2 %>% blend_predictions() #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Out of 11 possible candidate members, the ensemble retained 2. #> Penalty: 0.01. #> Mixture: 1. #>  #> The 2 highest weighted member classes are: #> # A tibble: 2 × 3 #>   member                      type        weight #>   <chr>                       <chr>        <dbl> #> 1 .pred_no_neural_network_1_1 mlp           7.08 #> 2 .pred_no_random_forest_1_05 rand_forest   3.10 #>  #> Members have not yet been fitted with `fit_members()`."},{"path":"https://stacks.tidymodels.org/dev/reference/augment.model_stack.html","id":null,"dir":"Reference","previous_headings":"","what":"Augment a model stack — augment.model_stack","title":"Augment a model stack — augment.model_stack","text":"Augment model stack","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/augment.model_stack.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Augment a model stack — augment.model_stack","text":"","code":"# S3 method for model_stack augment(x, new_data, ...)"},{"path":"https://stacks.tidymodels.org/dev/reference/augment.model_stack.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Augment a model stack — augment.model_stack","text":"x fitted model stack; see fit_members(). new_data rectangular data object, data frame. ... Additional arguments passed predict.model_stack. particular, see type members.","code":""},{"path":[]},{"path":"https://stacks.tidymodels.org/dev/reference/autoplot.linear_stack.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot results of a stacked ensemble model. — autoplot.linear_stack","title":"Plot results of a stacked ensemble model. — autoplot.linear_stack","text":"Plot results stacked ensemble model.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/autoplot.linear_stack.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot results of a stacked ensemble model. — autoplot.linear_stack","text":"","code":"# S3 method for linear_stack autoplot(object, type = \"performance\", n = Inf, ...)"},{"path":"https://stacks.tidymodels.org/dev/reference/autoplot.linear_stack.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot results of a stacked ensemble model. — autoplot.linear_stack","text":"object linear_stack object outputted blend_predictions() fit_members(). type single character string plot type values \"performance\", \"members\", \"weights\". n integer many members weights plot type = \"weights\". multi-class data, total number weights across classes; otherwise equal number members. ... currently used.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/autoplot.linear_stack.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Plot results of a stacked ensemble model. — autoplot.linear_stack","text":"ggplot object.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/autoplot.linear_stack.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Plot results of a stacked ensemble model. — autoplot.linear_stack","text":"\"performance\" plot shows relationship lasso penalty resampled performance metrics. latter includes average number ensemble members. plot can helpful understanding penalty values reasonable. \"members\" plot shows relationship average number ensemble members performance metrics. point different penalty value. Neither \"performance\" \"members\" plots helpful single penalty used. \"weights\" plot shows blending weights top ensemble members. results final penalty value used fit ensemble.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/axe_model_stack.html","id":null,"dir":"Reference","previous_headings":"","what":"Axing a model_stack. — axe_model_stack","title":"Axing a model_stack. — axe_model_stack","text":"Axing model_stack. Remove call. Remove controls used training. Remove training data. Remove environments. Remove fitted values.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/axe_model_stack.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Axing a model_stack. — axe_model_stack","text":"","code":"# S3 method for model_stack axe_call(x, verbose = FALSE, ...)  # S3 method for model_stack axe_ctrl(x, verbose = FALSE, ...)  # S3 method for model_stack axe_data(x, verbose = FALSE, ...)  # S3 method for model_stack axe_env(x, verbose = FALSE, ...)  # S3 method for model_stack axe_fitted(x, verbose = FALSE, ...)"},{"path":"https://stacks.tidymodels.org/dev/reference/axe_model_stack.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Axing a model_stack. — axe_model_stack","text":"x model object verbose Print information time axe method executed. Notes much memory released functions disabled. Default FALSE. ... Additional arguments. Currently ignored.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/axe_model_stack.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Axing a model_stack. — axe_model_stack","text":"Axed model_stack object.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/axe_model_stack.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Axing a model_stack. — axe_model_stack","text":"","code":"# build a regression model stack st <-   stacks() %>%   add_candidates(reg_res_lr) %>%   add_candidates(reg_res_sp) %>%   blend_predictions() %>%   fit_members()    # remove any of the \"butcherable\" # elements individually axe_call(st) #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Print methods for butchered model stacks are disabled. axe_ctrl(st) #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Print methods for butchered model stacks are disabled. axe_data(st) #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Print methods for butchered model stacks are disabled. axe_fitted(st) #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Print methods for butchered model stacks are disabled. axe_env(st) #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Print methods for butchered model stacks are disabled.  # or do it all at once! butchered_st <- butcher(st, verbose = TRUE) #> ✔ Memory released: 155.86 kB #> ✖ Disabled: `print()` and `summary()`  format(object.size(st)) #> [1] \"2465848 bytes\" format(object.size(butchered_st)) #> [1] \"1565608 bytes\""},{"path":"https://stacks.tidymodels.org/dev/reference/blend_predictions.html","id":null,"dir":"Reference","previous_headings":"","what":"Determine stacking coefficients from a data stack — blend_predictions","title":"Determine stacking coefficients from a data stack — blend_predictions","text":"Evaluates data stack fitting regularized model assessment predictions candidate member predict true outcome. process determines \"stacking coefficients\" model stack. stacking coefficients used weight predictions candidate (represented unique column data stack), given betas LASSO model fitting true outcome predictions given remaining columns data stack. Candidates non-zero stacking coefficients model stack members, need trained full training set (rather just assessment set) fit_members(). function typically used number calls add_candidates().","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/blend_predictions.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Determine stacking coefficients from a data stack — blend_predictions","text":"","code":"blend_predictions(   data_stack,   penalty = 10^(-6:-1),   mixture = 1,   non_negative = TRUE,   metric = NULL,   control = tune::control_grid(),   times = 25,   ... )"},{"path":"https://stacks.tidymodels.org/dev/reference/blend_predictions.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Determine stacking coefficients from a data stack — blend_predictions","text":"data_stack data_stack object penalty numeric vector proposed values total amount regularization used member weighting. Higher penalties generally result fewer members included resulting model stack, vice versa. package tune grid formed cross product penalty mixture arguments. mixture number zero one (inclusive) giving proportion L1 regularization (.e. lasso) model. mixture = 1 indicates pure lasso model, mixture = 0 indicates ridge regression, values (0, 1) indicate elastic net. package tune grid formed cross product penalty mixture arguments. non_negative logical giving whether restrict stacking coefficients non-negative values. TRUE (default), 0 passed lower.limits argument glmnet::glmnet() fitting model data stack. Otherwise, -Inf. metric call yardstick::metric_set(). metric(s) use tuning lasso penalty stacking coefficients. Default values determined tune::tune_grid() outcome class. control object inheriting control_grid passed model determining stacking coefficients. See tune::control_grid() documentation details possible values. Note extract entry overwritten internally. times Number bootstrap samples tuned model determines stacking coefficients. See rsample::bootstraps() learn . ... Additional arguments. Currently ignored.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/blend_predictions.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Determine stacking coefficients from a data stack — blend_predictions","text":"model_stack object—model_stacks largely contain elements data_stacks, primary data objects shift assessment set predictions member models.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/blend_predictions.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Determine stacking coefficients from a data stack — blend_predictions","text":"Note regularized linear model one many possible learning algorithms used fit stacked ensemble model. implementations additional ensemble learning algorithms, see h2o::h2o.stackedEnsemble() SuperLearner::SuperLearner().","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/blend_predictions.html","id":"example-data","dir":"Reference","previous_headings":"","what":"Example Data","title":"Determine stacking coefficients from a data stack — blend_predictions","text":"package provides resampling objects datasets use examples vignettes derived study 1212 red-eyed tree frog embryos! Red-eyed tree frog (RETF) embryos can hatch earlier normal 7ish days detect potential predator threat. Researchers wanted determine , , tree frog embryos able detect stimulus environment. , subjected embryos varying developmental stages \"predator stimulus\" jiggling embryos blunt probe. Beforehand, though embryos treated gentamicin, compound knocks lateral line (sensory organ.) Researcher Julie Jung crew found factors inform whether embryo hatches prematurely ! Note data included stacks package necessarily representative unbiased subset complete dataset, demonstrative purposes. reg_folds class_folds rset cross-fold validation objects rsample, splitting training data regression classification model objects, respectively. tree_frogs_reg_test tree_frogs_class_test analogous testing sets. reg_res_lr, reg_res_svm, reg_res_sp contain regression tuning results linear regression, support vector machine, spline model, respectively, fitting latency (.e. long embryos took hatch response jiggle) tree_frogs data, using variables predictors. Note data underlying models filtered include data embryos hatched response stimulus. class_res_rf class_res_nn contain multiclass classification tuning results random forest neural network classification model, respectively, fitting reflex (measure ear function) data using variables predictors. log_res_rf log_res_nn, contain binary classification tuning results random forest neural network classification model, respectively, fitting hatched (whether embryos hatched response stimulus) using variables predictors. See ?example_data learn objects, well browse source code generated .","code":""},{"path":[]},{"path":"https://stacks.tidymodels.org/dev/reference/blend_predictions.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Determine stacking coefficients from a data stack — blend_predictions","text":"","code":"# see the \"Example Data\" section above for # clarification on the objects used in these examples!  # put together a data stack reg_st <-    stacks() %>%   add_candidates(reg_res_lr) %>%   add_candidates(reg_res_svm) %>%   add_candidates(reg_res_sp)    reg_st #> # A data stack with 3 model definitions and 16 candidate members: #> #   reg_res_lr: 1 model configuration #> #   reg_res_svm: 5 model configurations #> #   reg_res_sp: 10 model configurations #> # Outcome: latency (numeric)  # evaluate the data stack reg_st %>%   blend_predictions() #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Out of 16 possible candidate members, the ensemble retained 3. #> Penalty: 1e-06. #> Mixture: 1. #>  #> The 3 highest weighted members are: #> # A tibble: 3 × 3 #>   member          type       weight #>   <chr>           <chr>       <dbl> #> 1 reg_res_svm_1_3 svm_rbf    0.638  #> 2 reg_res_sp_03_1 linear_reg 0.486  #> 3 reg_res_sp_10_1 linear_reg 0.0482 #>  #> Members have not yet been fitted with `fit_members()`.  # include fewer models by proposing higher penalties reg_st %>%    blend_predictions(penalty = c(.5, 1)) #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Out of 16 possible candidate members, the ensemble retained 3. #> Penalty: 0.5. #> Mixture: 1. #>  #> The 3 highest weighted members are: #> # A tibble: 3 × 3 #>   member          type       weight #>   <chr>           <chr>       <dbl> #> 1 reg_res_svm_1_3 svm_rbf    0.629  #> 2 reg_res_sp_03_1 linear_reg 0.478  #> 3 reg_res_sp_10_1 linear_reg 0.0515 #>  #> Members have not yet been fitted with `fit_members()`.  # allow for negative stacking coefficients  # with the non_negative argument reg_st %>%    blend_predictions(non_negative = FALSE) #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Out of 16 possible candidate members, the ensemble retained 12. #> Penalty: 0.1. #> Mixture: 1. #>  #> The 10 highest weighted members are: #> # A tibble: 10 × 3 #>    member          type        weight #>    <chr>           <chr>        <dbl> #>  1 reg_res_svm_1_1 svm_rbf    -10.5   #>  2 reg_res_sp_04_1 linear_reg  -1.38  #>  3 reg_res_sp_05_1 linear_reg   1.35  #>  4 reg_res_svm_1_3 svm_rbf      1.19  #>  5 reg_res_svm_1_2 svm_rbf     -0.963 #>  6 reg_res_sp_03_1 linear_reg   0.642 #>  7 reg_res_sp_01_1 linear_reg  -0.400 #>  8 reg_res_sp_10_1 linear_reg   0.319 #>  9 reg_res_sp_06_1 linear_reg   0.193 #> 10 reg_res_lr_1_1  linear_reg   0.183 #>  #> Members have not yet been fitted with `fit_members()`.    # use a custom metric in tuning the lasso penalty library(yardstick) reg_st %>%    blend_predictions(metric = metric_set(rmse)) #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Out of 16 possible candidate members, the ensemble retained 3. #> Penalty: 0.001. #> Mixture: 1. #>  #> The 3 highest weighted members are: #> # A tibble: 3 × 3 #>   member          type       weight #>   <chr>           <chr>       <dbl> #> 1 reg_res_svm_1_3 svm_rbf    0.638  #> 2 reg_res_sp_03_1 linear_reg 0.486  #> 3 reg_res_sp_10_1 linear_reg 0.0482 #>  #> Members have not yet been fitted with `fit_members()`.    # pass control options for stack blending reg_st %>%    blend_predictions(     control = tune::control_grid(allow_par = TRUE)   ) #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Out of 16 possible candidate members, the ensemble retained 3. #> Penalty: 1e-06. #> Mixture: 1. #>  #> The 3 highest weighted members are: #> # A tibble: 3 × 3 #>   member          type       weight #>   <chr>           <chr>       <dbl> #> 1 reg_res_svm_1_3 svm_rbf    0.638  #> 2 reg_res_sp_03_1 linear_reg 0.486  #> 3 reg_res_sp_10_1 linear_reg 0.0482 #>  #> Members have not yet been fitted with `fit_members()`.   # to speed up the stacking process for preliminary # results, bump down the `times` argument: reg_st %>%    blend_predictions(times = 5) #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Out of 16 possible candidate members, the ensemble retained 3. #> Penalty: 1e-06. #> Mixture: 1. #>  #> The 3 highest weighted members are: #> # A tibble: 3 × 3 #>   member          type       weight #>   <chr>           <chr>       <dbl> #> 1 reg_res_svm_1_3 svm_rbf    0.638  #> 2 reg_res_sp_03_1 linear_reg 0.486  #> 3 reg_res_sp_10_1 linear_reg 0.0482 #>  #> Members have not yet been fitted with `fit_members()`.    # the process looks the same with  # multinomial classification models class_st <-   stacks() %>%   add_candidates(class_res_nn) %>%   add_candidates(class_res_rf) %>%   blend_predictions() #> Warning: Predictions from 1 candidate were identical to those from existing #> candidates and were removed from the data stack.    class_st #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Out of 21 possible candidate members, the ensemble retained 5. #> Penalty: 0.1. #> Mixture: 1. #> Across the 3 classes, there are an average of 2.5 coefficients per class. #>  #> The 5 highest weighted member classes are: #> # A tibble: 5 × 4 #>   member                       type          weight class #>   <chr>                        <chr>          <dbl> <fct> #> 1 .pred_full_class_res_nn_1_1  mlp         12.0     full  #> 2 .pred_mid_class_res_rf_1_06  rand_forest  0.670   mid   #> 3 .pred_full_class_res_rf_1_05 rand_forest  0.101   full  #> 4 .pred_full_class_res_rf_1_07 rand_forest  0.00457 full  #> 5 .pred_full_class_res_rf_1_01 rand_forest  0.00219 full  #>  #> Members have not yet been fitted with `fit_members()`.  # ...or binomial classification models log_st <-   stacks() %>%   add_candidates(log_res_nn) %>%   add_candidates(log_res_rf) %>%   blend_predictions()    log_st #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Out of 11 possible candidate members, the ensemble retained 3. #> Penalty: 0.001. #> Mixture: 1. #>  #> The 3 highest weighted member classes are: #> # A tibble: 3 × 3 #>   member                   type        weight #>   <chr>                    <chr>        <dbl> #> 1 .pred_no_log_res_nn_1_1  mlp         7.39   #> 2 .pred_no_log_res_rf_1_05 rand_forest 3.42   #> 3 .pred_no_log_res_rf_1_02 rand_forest 0.0281 #>  #> Members have not yet been fitted with `fit_members()`."},{"path":"https://stacks.tidymodels.org/dev/reference/build_linear_predictor.html","id":null,"dir":"Reference","previous_headings":"","what":"Creates an R expression for a linear predictor from a data frame of terms and\ncoefficients — build_linear_predictor","title":"Creates an R expression for a linear predictor from a data frame of terms and\ncoefficients — build_linear_predictor","text":"Creates R expression linear predictor data frame terms coefficients","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/build_linear_predictor.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Creates an R expression for a linear predictor from a data frame of terms and\ncoefficients — build_linear_predictor","text":"","code":"build_linear_predictor(x, ...)  # S3 method for `_elnet` build_linear_predictor(x, ...)  # S3 method for `_lognet` build_linear_predictor(x, ...)  # S3 method for `_multnet` build_linear_predictor(x, ...)"},{"path":"https://stacks.tidymodels.org/dev/reference/build_linear_predictor.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Creates an R expression for a linear predictor from a data frame of terms and\ncoefficients — build_linear_predictor","text":"x object uses glmnet::glmnet() model numeric predictors. ... currently used.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/build_linear_predictor.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Creates an R expression for a linear predictor from a data frame of terms and\ncoefficients — build_linear_predictor","text":"R expression list R expressions, depending type model used.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/collect_parameters.html","id":null,"dir":"Reference","previous_headings":"","what":"Collect candidate parameters and stacking coefficients — collect_parameters","title":"Collect candidate parameters and stacking coefficients — collect_parameters","text":"function help situate candidates within stack. Takes data stack model stack candidate name returns tibble mapping candidate/member names hyperparameters (, model stack, stacking coefficients well).","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/collect_parameters.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Collect candidate parameters and stacking coefficients — collect_parameters","text":"","code":"collect_parameters(stack, candidates, ...)  # S3 method for default collect_parameters(stack, candidates, ...)  # S3 method for data_stack collect_parameters(stack, candidates, ...)  # S3 method for model_stack collect_parameters(stack, candidates, ...)"},{"path":"https://stacks.tidymodels.org/dev/reference/collect_parameters.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Collect candidate parameters and stacking coefficients — collect_parameters","text":"stack data_stack model_stack object. candidates name candidates collect parameters . either name argument supplied add_candidates() , supplied, name object supplied candidates argument add_candidates(). ... Additional arguments. Currently ignored.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/collect_parameters.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Collect candidate parameters and stacking coefficients — collect_parameters","text":"tibble::tbl_df information member names hyperparameters.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/collect_parameters.html","id":"example-data","dir":"Reference","previous_headings":"","what":"Example Data","title":"Collect candidate parameters and stacking coefficients — collect_parameters","text":"package provides resampling objects datasets use examples vignettes derived study 1212 red-eyed tree frog embryos! Red-eyed tree frog (RETF) embryos can hatch earlier normal 7ish days detect potential predator threat. Researchers wanted determine , , tree frog embryos able detect stimulus environment. , subjected embryos varying developmental stages \"predator stimulus\" jiggling embryos blunt probe. Beforehand, though embryos treated gentamicin, compound knocks lateral line (sensory organ.) Researcher Julie Jung crew found factors inform whether embryo hatches prematurely ! Note data included stacks package necessarily representative unbiased subset complete dataset, demonstrative purposes. reg_folds class_folds rset cross-fold validation objects rsample, splitting training data regression classification model objects, respectively. tree_frogs_reg_test tree_frogs_class_test analogous testing sets. reg_res_lr, reg_res_svm, reg_res_sp contain regression tuning results linear regression, support vector machine, spline model, respectively, fitting latency (.e. long embryos took hatch response jiggle) tree_frogs data, using variables predictors. Note data underlying models filtered include data embryos hatched response stimulus. class_res_rf class_res_nn contain multiclass classification tuning results random forest neural network classification model, respectively, fitting reflex (measure ear function) data using variables predictors. log_res_rf log_res_nn, contain binary classification tuning results random forest neural network classification model, respectively, fitting hatched (whether embryos hatched response stimulus) using variables predictors. See ?example_data learn objects, well browse source code generated .","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/collect_parameters.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Collect candidate parameters and stacking coefficients — collect_parameters","text":"","code":"# see the \"Example Data\" section above for # clarification on the objects used in these examples!  # put together a data stack using # tuning results for regression models reg_st <-    stacks() %>%   add_candidates(reg_res_lr) %>%   add_candidates(reg_res_svm) %>%   add_candidates(reg_res_sp, \"spline\")    reg_st #> # A data stack with 3 model definitions and 16 candidate members: #> #   reg_res_lr: 1 model configuration #> #   reg_res_svm: 5 model configurations #> #   spline: 10 model configurations #> # Outcome: latency (numeric)    # check out the hyperparameters for some of the candidates collect_parameters(reg_st, \"reg_res_svm\") #> # A tibble: 5 × 3 #>   member             cost     rbf_sigma #>   <chr>             <dbl>         <dbl> #> 1 reg_res_svm_1_1 3.09    0.000000465   #> 2 reg_res_svm_1_2 9.56    0.00346       #> 3 reg_res_svm_1_3 0.321   0.0256        #> 4 reg_res_svm_1_4 0.00486 0.00000251    #> 5 reg_res_svm_1_5 0.0146  0.00000000959  collect_parameters(reg_st, \"spline\") #> # A tibble: 10 × 2 #>    member        age #>    <chr>       <int> #>  1 spline_01_1    12 #>  2 spline_02_1     6 #>  3 spline_03_1    15 #>  4 spline_04_1     5 #>  5 spline_05_1     3 #>  6 spline_06_1    13 #>  7 spline_07_1     9 #>  8 spline_08_1    10 #>  9 spline_09_1     7 #> 10 spline_10_1     2  # blend the data stack to view the hyperparameters  # along with the stacking coefficients! collect_parameters(   reg_st %>% blend_predictions(),    \"spline\" ) #> # A tibble: 10 × 3 #>    member        age   coef #>    <chr>       <int>  <dbl> #>  1 spline_01_1    12 0      #>  2 spline_02_1     6 0      #>  3 spline_03_1    15 0.484  #>  4 spline_04_1     5 0      #>  5 spline_05_1     3 0      #>  6 spline_06_1    13 0      #>  7 spline_07_1     9 0      #>  8 spline_08_1    10 0      #>  9 spline_09_1     7 0      #> 10 spline_10_1     2 0.0496"},{"path":"https://stacks.tidymodels.org/dev/reference/control_stack.html","id":null,"dir":"Reference","previous_headings":"","what":"Control wrappers — control_stack_grid","title":"Control wrappers — control_stack_grid","text":"Supply light wrappers control argument tune::tune_grid(), tune::tune_bayes(), tune::fit_resamples() call return needed elements use data stack. functions return appropriate control grid ensure assessment set predictions information model specifications preprocessors, supplied resampling results object! integrate stack settings existing control settings, note functions just call appropriate tune::control_* function arguments save_pred = TRUE, save_workflow = TRUE.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/control_stack.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Control wrappers — control_stack_grid","text":"","code":"control_stack_grid()  control_stack_resamples()  control_stack_bayes()"},{"path":"https://stacks.tidymodels.org/dev/reference/control_stack.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Control wrappers — control_stack_grid","text":"tune::control_grid, tune::control_bayes, tune::control_resamples object.","code":""},{"path":[]},{"path":"https://stacks.tidymodels.org/dev/reference/example_data.html","id":null,"dir":"Reference","previous_headings":"","what":"Example Objects — example_data","title":"Example Objects — example_data","text":"stacks provides resampling objects datasets use examples vignettes derived study 1212 red-eyed tree frog embryos!","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/example_data.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Example Objects — example_data","text":"","code":"reg_res_svm  reg_res_sp  reg_res_lr  reg_folds  class_res_nn  class_res_rf  class_folds  log_res_nn  log_res_rf"},{"path":"https://stacks.tidymodels.org/dev/reference/example_data.html","id":"format","dir":"Reference","previous_headings":"","what":"Format","title":"Example Objects — example_data","text":"object class tune_results (inherits tbl_df, tbl, data.frame) 5 rows 5 columns. object class tune_results (inherits tbl_df, tbl, data.frame) 5 rows 5 columns. object class resample_results (inherits tune_results, tbl_df, tbl, data.frame) 5 rows 5 columns. object class vfold_cv (inherits rset, tbl_df, tbl, data.frame) 5 rows 2 columns. object class resample_results (inherits tune_results, tbl_df, tbl, data.frame) 5 rows 5 columns. object class tune_results (inherits tbl_df, tbl, data.frame) 5 rows 5 columns. object class vfold_cv (inherits rset, tbl_df, tbl, data.frame) 5 rows 2 columns. object class resample_results (inherits tune_results, tbl_df, tbl, data.frame) 5 rows 5 columns. object class tune_results (inherits tbl_df, tbl, data.frame) 5 rows 5 columns.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/example_data.html","id":"source","dir":"Reference","previous_headings":"","what":"Source","title":"Example Objects — example_data","text":"Julie Jung et al. (2020) Multimodal mechanosensing enables treefrog embryos escape egg-predators. doi:10.1242/jeb.236141","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/example_data.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Example Objects — example_data","text":"Red-eyed tree frog (RETF) embryos can hatch earlier normal 7ish days detect potential predator threat. Researchers wanted determine , , tree frog embryos able detect stimulus environment. , subjected embryos varying developmental stages \"predator stimulus\" jiggling embryos blunt probe. Beforehand, though embryos treated gentamicin, compound knocks lateral line (sensory organ.) Researcher Julie Jung crew found factors inform whether embryo hatches prematurely ! Note data included stacks package necessarily representative unbiased subset complete dataset, demonstrative purposes. reg_folds class_folds rset cross-fold validation objects rsample, splitting training data regression classification model objects, respectively. tree_frogs_reg_test tree_frogs_class_test analogous testing sets. reg_res_lr, reg_res_svm, reg_res_sp contain regression tuning results linear regression, support vector machine, spline model, respectively, fitting latency (.e. long embryos took hatch response jiggle) tree_frogs data, using variables predictors. Note data underlying models filtered include data embryos hatched response stimulus. class_res_rf class_res_nn contain multiclass classification tuning results random forest neural network classification model, respectively, fitting reflex (measure ear function) data using variables predictors. log_res_rf log_res_nn, contain binary classification tuning results random forest neural network classification model, respectively, fitting hatched (whether embryos hatched response stimulus) using variables predictors. source code generating objects given .","code":"# setup: packages, data, resample, basic recipe ------------------------ library(stacks) library(tune) library(rsample) library(parsnip) library(workflows) library(recipes) library(yardstick) library(workflowsets)  set.seed(1)  ctrl_grid <-    tune::control_grid(     save_pred = TRUE,     save_workflow = TRUE   )  ctrl_res <-    tune::control_resamples(     save_pred = TRUE,     save_workflow = TRUE   )  # for regression, predict latency to hatch (excluding NAs) tree_frogs_reg <-    tree_frogs %>%    filter(!is.na(latency)) %>%   select(-clutch, -hatched)  set.seed(1) tree_frogs_reg_split <- rsample::initial_split(tree_frogs_reg)  set.seed(1) tree_frogs_reg_train <- rsample::training(tree_frogs_reg_split)  set.seed(1) tree_frogs_reg_test  <- rsample::testing(tree_frogs_reg_split)  set.seed(1) reg_folds <- rsample::vfold_cv(tree_frogs_reg_train, v = 5)  tree_frogs_reg_rec <-    recipes::recipe(latency ~ ., data = tree_frogs_reg_train) %>%   recipes::step_dummy(recipes::all_nominal()) %>%   recipes::step_zv(recipes::all_predictors())  metric <- yardstick::metric_set(yardstick::rmse)  # linear regression --------------------------------------- lin_reg_spec <-   parsnip::linear_reg() %>%   parsnip::set_engine(\"lm\")  reg_wf_lr <-    workflows::workflow() %>%   workflows::add_model(lin_reg_spec) %>%   workflows::add_recipe(tree_frogs_reg_rec)  set.seed(1) reg_res_lr <-    tune::fit_resamples(     object = reg_wf_lr,     resamples = reg_folds,     metrics = metric,     control = ctrl_res   )  # SVM regression ---------------------------------- svm_spec <-    parsnip::svm_rbf(     cost = tune::tune(),      rbf_sigma = tune::tune()   ) %>%   parsnip::set_engine(\"kernlab\") %>%   parsnip::set_mode(\"regression\")  reg_wf_svm <-    workflows::workflow() %>%   workflows::add_model(svm_spec) %>%   workflows::add_recipe(tree_frogs_reg_rec)  set.seed(1) reg_res_svm <-    tune::tune_grid(     object = reg_wf_svm,     resamples = reg_folds,      grid = 5,     control = ctrl_grid   )  # spline regression --------------------------------------- spline_rec <-    tree_frogs_reg_rec %>%   recipes::step_ns(age, deg_free = tune::tune(\"age\"))  reg_wf_sp <-    workflows::workflow() %>%   workflows::add_model(lin_reg_spec) %>%   workflows::add_recipe(spline_rec)  set.seed(1) reg_res_sp <-    tune::tune_grid(     object = reg_wf_sp,     resamples = reg_folds,     metrics = metric,     control = ctrl_grid   )  # classification - preliminaries ----------------------------------- tree_frogs_class <-    tree_frogs %>%   dplyr::select(-c(clutch, latency))  set.seed(1) tree_frogs_class_split <- rsample::initial_split(tree_frogs_class)  set.seed(1) tree_frogs_class_train <- rsample::training(tree_frogs_class_split)  set.seed(1) tree_frogs_class_test  <- rsample::testing(tree_frogs_class_split)  set.seed(1) class_folds <- rsample::vfold_cv(tree_frogs_class_train, v = 5)  tree_frogs_class_rec <-    recipes::recipe(reflex ~ ., data = tree_frogs_class_train) %>%   recipes::step_dummy(recipes::all_nominal(), -reflex) %>%   recipes::step_zv(recipes::all_predictors()) %>%   recipes::step_normalize(recipes::all_numeric())  # random forest classification -------------------------------------- rand_forest_spec <-    parsnip::rand_forest(     mtry = tune::tune(),     trees = 500,     min_n = tune::tune()   ) %>%   parsnip::set_mode(\"classification\") %>%   parsnip::set_engine(\"ranger\")  class_wf_rf <-   workflows::workflow() %>%   workflows::add_recipe(tree_frogs_class_rec) %>%   workflows::add_model(rand_forest_spec)  set.seed(1) class_res_rf <-    tune::tune_grid(     object = class_wf_rf,      resamples = class_folds,      grid = 10,     control = ctrl_grid   )  # neural network classification ------------------------------------- nnet_spec <-   mlp(hidden_units = 5, penalty = 0.01, epochs = 100) %>%   set_mode(\"classification\") %>%   set_engine(\"nnet\")  class_wf_nn <-    workflows::workflow() %>%   workflows::add_recipe(tree_frogs_class_rec) %>%   workflows::add_model(nnet_spec)  set.seed(1) class_res_nn <-   tune::fit_resamples(     object = class_wf_nn,      resamples = class_folds,      control = ctrl_res   )  # binary classification -------------------------------- tree_frogs_2_class_rec <-    recipes::recipe(hatched ~ ., data = tree_frogs_class_train) %>%   recipes::step_dummy(recipes::all_nominal(), -hatched) %>%   recipes::step_zv(recipes::all_predictors()) %>%   recipes::step_normalize(recipes::all_numeric())  set.seed(1) rand_forest_spec_2 <-    parsnip::rand_forest(     mtry = tune(),     trees = 500,     min_n = tune()   ) %>%   parsnip::set_mode(\"classification\") %>%   parsnip::set_engine(\"ranger\")  log_wf_rf <-   workflows::workflow() %>%   workflows::add_recipe(tree_frogs_2_class_rec) %>%   workflows::add_model(rand_forest_spec_2)  set.seed(1) log_res_rf <-    tune::tune_grid(     object = log_wf_rf,      resamples = class_folds,      grid = 10,     control = ctrl_grid   )  nnet_spec_2 <-   parsnip::mlp(epochs = 100, hidden_units = 5, penalty = 0.1) %>%   parsnip::set_mode(\"classification\") %>%   parsnip::set_engine(\"nnet\", verbose = 0)  log_wf_nn <-    workflows::workflow() %>%   workflows::add_recipe(tree_frogs_2_class_rec) %>%   workflows::add_model(nnet_spec_2)  set.seed(1) log_res_nn <-   tune::fit_resamples(     object = log_wf_nn,      resamples = class_folds,      control = ctrl_res   )"},{"path":"https://stacks.tidymodels.org/dev/reference/fit_members.html","id":null,"dir":"Reference","previous_headings":"","what":"Fit model stack members with non-zero stacking coefficients — fit_members","title":"Fit model stack members with non-zero stacking coefficients — fit_members","text":"evaluating data stack blend_predictions(), number candidates nonzero stacking coefficients. candidates referred \"members.\" Since members' predictions ultimately inform model stack's predictions, members trained full training set using fit_members().","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/fit_members.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Fit model stack members with non-zero stacking coefficients — fit_members","text":"","code":"fit_members(model_stack, ...)"},{"path":"https://stacks.tidymodels.org/dev/reference/fit_members.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Fit model stack members with non-zero stacking coefficients — fit_members","text":"model_stack model_stack object outputted blend_predictions(). ... Additional arguments. Currently ignored.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/fit_members.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Fit model stack members with non-zero stacking coefficients — fit_members","text":"model_stack object subclass linear_stack---fitted model contains necessary components predict new data.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/fit_members.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Fit model stack members with non-zero stacking coefficients — fit_members","text":"fit members parallel, please register parallel backend function. See documentation foreach::foreach() examples.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/fit_members.html","id":"example-data","dir":"Reference","previous_headings":"","what":"Example Data","title":"Fit model stack members with non-zero stacking coefficients — fit_members","text":"package provides resampling objects datasets use examples vignettes derived study 1212 red-eyed tree frog embryos! Red-eyed tree frog (RETF) embryos can hatch earlier normal 7ish days detect potential predator threat. Researchers wanted determine , , tree frog embryos able detect stimulus environment. , subjected embryos varying developmental stages \"predator stimulus\" jiggling embryos blunt probe. Beforehand, though embryos treated gentamicin, compound knocks lateral line (sensory organ.) Researcher Julie Jung crew found factors inform whether embryo hatches prematurely ! Note data included stacks package necessarily representative unbiased subset complete dataset, demonstrative purposes. reg_folds class_folds rset cross-fold validation objects rsample, splitting training data regression classification model objects, respectively. tree_frogs_reg_test tree_frogs_class_test analogous testing sets. reg_res_lr, reg_res_svm, reg_res_sp contain regression tuning results linear regression, support vector machine, spline model, respectively, fitting latency (.e. long embryos took hatch response jiggle) tree_frogs data, using variables predictors. Note data underlying models filtered include data embryos hatched response stimulus. class_res_rf class_res_nn contain multiclass classification tuning results random forest neural network classification model, respectively, fitting reflex (measure ear function) data using variables predictors. log_res_rf log_res_nn, contain binary classification tuning results random forest neural network classification model, respectively, fitting hatched (whether embryos hatched response stimulus) using variables predictors. See ?example_data learn objects, well browse source code generated .","code":""},{"path":[]},{"path":"https://stacks.tidymodels.org/dev/reference/fit_members.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Fit model stack members with non-zero stacking coefficients — fit_members","text":"","code":"# see the \"Example Data\" section above for # clarification on the objects used in these examples!  # put together a data stack reg_st <-    stacks() %>%   add_candidates(reg_res_lr) %>%   add_candidates(reg_res_svm) %>%   add_candidates(reg_res_sp)    reg_st #> # A data stack with 3 model definitions and 16 candidate members: #> #   reg_res_lr: 1 model configuration #> #   reg_res_svm: 5 model configurations #> #   reg_res_sp: 10 model configurations #> # Outcome: latency (numeric)  # evaluate the data stack and fit the member models reg_st %>%   blend_predictions() %>%   fit_members() #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Out of 16 possible candidate members, the ensemble retained 3. #> Penalty: 1e-06. #> Mixture: 1. #>  #> The 3 highest weighted members are: #> # A tibble: 3 × 3 #>   member          type       weight #>   <chr>           <chr>       <dbl> #> 1 reg_res_svm_1_3 svm_rbf    0.638  #> 2 reg_res_sp_03_1 linear_reg 0.486  #> 3 reg_res_sp_10_1 linear_reg 0.0482    reg_st #> # A data stack with 3 model definitions and 16 candidate members: #> #   reg_res_lr: 1 model configuration #> #   reg_res_svm: 5 model configurations #> #   reg_res_sp: 10 model configurations #> # Outcome: latency (numeric)    # do the same with multinomial classification models class_st <-   stacks() %>%   add_candidates(class_res_nn) %>%   add_candidates(class_res_rf) %>%   blend_predictions() %>%   fit_members() #> Warning: Predictions from 1 candidate were identical to those from existing #> candidates and were removed from the data stack.    class_st #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Out of 21 possible candidate members, the ensemble retained 8. #> Penalty: 0.01. #> Mixture: 1. #> Across the 3 classes, there are an average of 4 coefficients per class. #>  #> The 8 highest weighted member classes are: #> # A tibble: 8 × 4 #>   member                       type           weight class #>   <chr>                        <chr>           <dbl> <fct> #> 1 .pred_full_class_res_nn_1_1  mlp         23.3      full  #> 2 .pred_mid_class_res_nn_1_1   mlp          1.89     mid   #> 3 .pred_mid_class_res_rf_1_06  rand_forest  1.71     mid   #> 4 .pred_mid_class_res_rf_1_10  rand_forest  1.17     mid   #> 5 .pred_full_class_res_rf_1_03 rand_forest  0.407    full  #> 6 .pred_full_class_res_rf_1_05 rand_forest  0.222    full  #> 7 .pred_full_class_res_rf_1_01 rand_forest  0.00160  full  #> 8 .pred_full_class_res_rf_1_02 rand_forest  0.000322 full     # ...or binomial classification models log_st <-   stacks() %>%   add_candidates(log_res_nn) %>%   add_candidates(log_res_rf) %>%   blend_predictions() %>%   fit_members()    log_st #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Out of 11 possible candidate members, the ensemble retained 2. #> Penalty: 0.01. #> Mixture: 1. #>  #> The 2 highest weighted member classes are: #> # A tibble: 2 × 3 #>   member                   type        weight #>   <chr>                    <chr>        <dbl> #> 1 .pred_no_log_res_nn_1_1  mlp           7.08 #> 2 .pred_no_log_res_rf_1_05 rand_forest   3.10"},{"path":"https://stacks.tidymodels.org/dev/reference/get_expressions.html","id":null,"dir":"Reference","previous_headings":"","what":"Obtain prediction equations for all possible values of type — get_expressions","title":"Obtain prediction equations for all possible values of type — get_expressions","text":"Obtain prediction equations possible values type","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/get_expressions.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Obtain prediction equations for all possible values of type — get_expressions","text":"","code":"get_expressions(x, ...)  # S3 method for `_multnet` get_expressions(x, ...)  # S3 method for `_lognet` get_expressions(x, ...)  # S3 method for `_elnet` get_expressions(x, ...)"},{"path":"https://stacks.tidymodels.org/dev/reference/get_expressions.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Obtain prediction equations for all possible values of type — get_expressions","text":"x parsnip model glmnet engine. ... used","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/get_expressions.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Obtain prediction equations for all possible values of type — get_expressions","text":"named list prediction equations possibel type.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/predict.data_stack.html","id":null,"dir":"Reference","previous_headings":"","what":"Predicting with a model stack — predict.data_stack","title":"Predicting with a model stack — predict.data_stack","text":"data stack must evaluated blend_predictions() member models fitted fit_members() predict new data.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/predict.data_stack.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Predicting with a model stack — predict.data_stack","text":"","code":"# S3 method for data_stack predict(object, ...)"},{"path":"https://stacks.tidymodels.org/dev/reference/predict.data_stack.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Predicting with a model stack — predict.data_stack","text":"object data stack. ... Additional arguments. Currently ignored.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/predict.model_stack.html","id":null,"dir":"Reference","previous_headings":"","what":"Predicting with a model stack — predict.model_stack","title":"Predicting with a model stack — predict.model_stack","text":"Apply model stack create different types predictions.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/predict.model_stack.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Predicting with a model stack — predict.model_stack","text":"","code":"# S3 method for model_stack predict(object, new_data, type = NULL, members = FALSE, opts = list(), ...)"},{"path":"https://stacks.tidymodels.org/dev/reference/predict.model_stack.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Predicting with a model stack — predict.model_stack","text":"object model stack fitted members outputted fit_members(). new_data rectangular data object, data frame. type Format returned predicted values—one \"numeric\", \"class\", \"prob\". NULL, predict() choose appropriate value based model's mode. members Logical. Whether additionally return predictions ensemble members. opts list optional arguments underlying predict function passed parsnip::predict.model_fit member. ... Additional arguments. Currently ignored.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/predict.model_stack.html","id":"example-data","dir":"Reference","previous_headings":"","what":"Example Data","title":"Predicting with a model stack — predict.model_stack","text":"package provides resampling objects datasets use examples vignettes derived study 1212 red-eyed tree frog embryos! Red-eyed tree frog (RETF) embryos can hatch earlier normal 7ish days detect potential predator threat. Researchers wanted determine , , tree frog embryos able detect stimulus environment. , subjected embryos varying developmental stages \"predator stimulus\" jiggling embryos blunt probe. Beforehand, though embryos treated gentamicin, compound knocks lateral line (sensory organ.) Researcher Julie Jung crew found factors inform whether embryo hatches prematurely ! Note data included stacks package necessarily representative unbiased subset complete dataset, demonstrative purposes. reg_folds class_folds rset cross-fold validation objects rsample, splitting training data regression classification model objects, respectively. tree_frogs_reg_test tree_frogs_class_test analogous testing sets. reg_res_lr, reg_res_svm, reg_res_sp contain regression tuning results linear regression, support vector machine, spline model, respectively, fitting latency (.e. long embryos took hatch response jiggle) tree_frogs data, using variables predictors. Note data underlying models filtered include data embryos hatched response stimulus. class_res_rf class_res_nn contain multiclass classification tuning results random forest neural network classification model, respectively, fitting reflex (measure ear function) data using variables predictors. log_res_rf log_res_nn, contain binary classification tuning results random forest neural network classification model, respectively, fitting hatched (whether embryos hatched response stimulus) using variables predictors. See ?example_data learn objects, well browse source code generated .","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/predict.model_stack.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Predicting with a model stack — predict.model_stack","text":"","code":"# see the \"Example Data\" section above for # clarification on the data and tuning results # objects used in these examples!  data(tree_frogs_reg_test) data(tree_frogs_class_test)  # build and fit a regression model stack reg_st <-   stacks() %>%   add_candidates(reg_res_lr) %>%   add_candidates(reg_res_sp) %>%   blend_predictions() %>%   fit_members()  reg_st #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Out of 11 possible candidate members, the ensemble retained 4. #> Penalty: 0.1. #> Mixture: 1. #>  #> The 4 highest weighted members are: #> # A tibble: 4 × 3 #>   member          type       weight #>   <chr>           <chr>       <dbl> #> 1 reg_res_sp_03_1 linear_reg 0.483  #> 2 reg_res_sp_10_1 linear_reg 0.249  #> 3 reg_res_lr_1_1  linear_reg 0.126  #> 4 reg_res_sp_05_1 linear_reg 0.0673  # predict on the tree frogs testing data predict(reg_st, tree_frogs_reg_test) #> # A tibble: 143 × 1 #>    .pred #>    <dbl> #>  1 119.  #>  2  81.5 #>  3 102.  #>  4  35.6 #>  5 119.  #>  6  50.6 #>  7 122.  #>  8  82.7 #>  9  50.3 #> 10  75.7 #> # ℹ 133 more rows  # include the predictions from the members predict(reg_st, tree_frogs_reg_test, members = TRUE) #> # A tibble: 143 × 5 #>    .pred reg_res_lr_1_1 reg_res_sp_10_1 reg_res_sp_05_1 reg_res_sp_03_1 #>    <dbl>          <dbl>           <dbl>           <dbl>           <dbl> #>  1 119.           138.            125.            121.            114.  #>  2  81.5           82.4            84.8            81.8            77.1 #>  3 102.           116.            111.            112.             93.3 #>  4  35.6           35.8            29.7            32.5            29.6 #>  5 119.           111.            115.            115.            127.  #>  6  50.6           38.8            37.4            36.2            55.3 #>  7 122.           123.            103.            104.            137.  #>  8  82.7           82.3            78.6            82.0            82.8 #>  9  50.3           38.7            37.3            36.2            54.8 #> 10  75.7           78.8            75.3            76.9            71.8 #> # ℹ 133 more rows  # build and fit a classification model stack class_st <-   stacks() %>%   add_candidates(class_res_nn) %>%   add_candidates(class_res_rf) %>%   blend_predictions() %>%   fit_members() #> Warning: Predictions from 1 candidate were identical to those from existing #> candidates and were removed from the data stack.   class_st #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Out of 21 possible candidate members, the ensemble retained 5. #> Penalty: 0.1. #> Mixture: 1. #> Across the 3 classes, there are an average of 2.5 coefficients per class. #>  #> The 5 highest weighted member classes are: #> # A tibble: 5 × 4 #>   member                       type          weight class #>   <chr>                        <chr>          <dbl> <fct> #> 1 .pred_full_class_res_nn_1_1  mlp         12.0     full  #> 2 .pred_mid_class_res_rf_1_06  rand_forest  0.670   mid   #> 3 .pred_full_class_res_rf_1_05 rand_forest  0.101   full  #> 4 .pred_full_class_res_rf_1_07 rand_forest  0.00457 full  #> 5 .pred_full_class_res_rf_1_01 rand_forest  0.00219 full   # predict reflex, first as a class, then as # class probabilities predict(class_st, tree_frogs_class_test) #> # A tibble: 303 × 1 #>    .pred_class #>    <fct>       #>  1 full        #>  2 mid         #>  3 mid         #>  4 mid         #>  5 full        #>  6 full        #>  7 full        #>  8 full        #>  9 full        #> 10 full        #> # ℹ 293 more rows predict(class_st, tree_frogs_class_test, type = \"prob\") #> # A tibble: 303 × 3 #>    .pred_full .pred_low .pred_mid #>         <dbl>     <dbl>     <dbl> #>  1     0.909     0.0540    0.0371 #>  2     0.101     0.434     0.465  #>  3     0.0923    0.419     0.489  #>  4     0.105     0.429     0.466  #>  5     0.909     0.0541    0.0373 #>  6     0.909     0.0540    0.0371 #>  7     0.909     0.0540    0.0371 #>  8     0.909     0.0540    0.0371 #>  9     0.909     0.0540    0.0371 #> 10     0.909     0.0540    0.0371 #> # ℹ 293 more rows  # returning the member predictions as well predict(   class_st,    tree_frogs_class_test,    type = \"prob\",    members = TRUE ) #> # A tibble: 303 × 18 #>    .pred_full .pred_low .pred_mid .pred_low_class_res_rf_1_06 #>         <dbl>     <dbl>     <dbl>                       <dbl> #>  1     0.909     0.0540    0.0371                       0     #>  2     0.101     0.434     0.465                        0.339 #>  3     0.0923    0.419     0.489                        0.212 #>  4     0.105     0.429     0.466                        0.318 #>  5     0.909     0.0541    0.0373                       0     #>  6     0.909     0.0540    0.0371                       0     #>  7     0.909     0.0540    0.0371                       0     #>  8     0.909     0.0540    0.0371                       0     #>  9     0.909     0.0540    0.0371                       0     #> 10     0.909     0.0540    0.0371                       0     #> # ℹ 293 more rows #> # ℹ 14 more variables: .pred_low_class_res_nn_1_1 <dbl>, #> #   .pred_low_class_res_rf_1_05 <dbl>, .pred_low_class_res_rf_1_01 <dbl>, #> #   .pred_low_class_res_rf_1_07 <dbl>, .pred_mid_class_res_rf_1_06 <dbl>, #> #   .pred_mid_class_res_nn_1_1 <dbl>, .pred_mid_class_res_rf_1_05 <dbl>, #> #   .pred_mid_class_res_rf_1_01 <dbl>, .pred_mid_class_res_rf_1_07 <dbl>, #> #   .pred_full_class_res_rf_1_06 <dbl>, …"},{"path":"https://stacks.tidymodels.org/dev/reference/prediction_eqn.html","id":null,"dir":"Reference","previous_headings":"","what":"Convert one or more linear predictor to a format used for prediction — prediction_eqn","title":"Convert one or more linear predictor to a format used for prediction — prediction_eqn","text":"Convert one linear predictor format used prediction","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/prediction_eqn.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Convert one or more linear predictor to a format used for prediction — prediction_eqn","text":"","code":"prediction_eqn(x, ...)  # S3 method for `_lognet` prediction_eqn(x, type = \"class\", ...)  # S3 method for `_elnet` prediction_eqn(x, type = \"numeric\", ...)  # S3 method for `_multnet` prediction_eqn(x, type = \"class\", ...)"},{"path":"https://stacks.tidymodels.org/dev/reference/prediction_eqn.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Convert one or more linear predictor to a format used for prediction — prediction_eqn","text":"x object uses glmnet::glmnet() model numeric predictors. ... currently used. type prediction type.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/prediction_eqn.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Convert one or more linear predictor to a format used for prediction — prediction_eqn","text":"return type varies, based model prediction type.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/reexports.html","id":null,"dir":"Reference","previous_headings":"","what":"Objects exported from other packages — reexports","title":"Objects exported from other packages — reexports","text":"objects imported packages. Follow links see documentation. butcher axe_call, axe_ctrl, axe_data, axe_env, axe_fitted, butcher dplyr %>% generics augment ggplot2 autoplot","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/stack_predict.html","id":null,"dir":"Reference","previous_headings":"","what":"Convert one or more linear predictor to a format used for prediction — stack_predict","title":"Convert one or more linear predictor to a format used for prediction — stack_predict","text":"Convert one linear predictor format used prediction","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/stack_predict.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Convert one or more linear predictor to a format used for prediction — stack_predict","text":"","code":"stack_predict(x, ...)  # S3 method for elnet_numeric stack_predict(x, data, ...)  # S3 method for lognet_class stack_predict(x, data, ...)  # S3 method for lognet_prob stack_predict(x, data, ...)  # S3 method for multnet_class stack_predict(x, data, ...)  # S3 method for multnet_prob stack_predict(x, data, ...)"},{"path":"https://stacks.tidymodels.org/dev/reference/stack_predict.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Convert one or more linear predictor to a format used for prediction — stack_predict","text":"x set model expressions generated prediction_eqn(). ... currently used. type prediction type.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/stack_predict.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Convert one or more linear predictor to a format used for prediction — stack_predict","text":"return type varies, based model prediction type.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/stacks.html","id":null,"dir":"Reference","previous_headings":"","what":"Initialize a Stack — stacks","title":"Initialize a Stack — stacks","text":"stacks() function initializes data_stack object. Principally, data_stacks tibbles, first column gives true outcome assessment set, remaining columns give predictions candidate ensemble member. (outcome numeric, ’s one column per candidate member. classification, many columns per candidate member levels outcome variable minus 1.) also bring along extra attributes keep track model definitions, resamples, training data. See ?stacks_description discussion package, generally, basics vignette detailed walk-functionality.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/stacks.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Initialize a Stack — stacks","text":"","code":"stacks(...)"},{"path":"https://stacks.tidymodels.org/dev/reference/stacks.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Initialize a Stack — stacks","text":"... Additional arguments. Currently ignored.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/stacks.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Initialize a Stack — stacks","text":"data_stack object.","code":""},{"path":[]},{"path":"https://stacks.tidymodels.org/dev/reference/stacks_description.html","id":null,"dir":"Reference","previous_headings":"","what":"stacks: Tidy Model Stacking — stacks_description","title":"stacks: Tidy Model Stacking — stacks_description","text":"Model stacking ensemble technique involves training model combine outputs many diverse statistical models, shown improve predictive performance variety settings. 'stacks' implements grammar 'tidymodels'-aligned model stacking.","code":""},{"path":[]},{"path":"https://stacks.tidymodels.org/dev/reference/stacks_description.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"stacks: Tidy Model Stacking — stacks_description","text":"Maintainer: Simon Couch simon.couch@posit.co Authors: Max Kuhn max@posit.co contributors: Posit Software, PBC [copyright holder, funder]","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/tree_frogs.html","id":null,"dir":"Reference","previous_headings":"","what":"Tree frog embryo hatching data — tree_frogs","title":"Tree frog embryo hatching data — tree_frogs","text":"dataset containing experimental results hatching behavior red-eyed tree frog embryos. Red-eyed tree frog (RETF) embryos can hatch earlier normal 7ish days detect potential predator threat. Researchers wanted determine , , tree frog embryos able detect stimulus environment. , subjected embryos varying developmental stages \"predator stimulus\" jiggling embryos blunt probe. Beforehand, though embryos treated gentamicin, compound knocks lateral line (sensory organ.) Researcher Julie Jung crew found factors inform whether embryo hatches prematurely !","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/tree_frogs.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Tree frog embryo hatching data — tree_frogs","text":"","code":"tree_frogs"},{"path":"https://stacks.tidymodels.org/dev/reference/tree_frogs.html","id":"format","dir":"Reference","previous_headings":"","what":"Format","title":"Tree frog embryo hatching data — tree_frogs","text":"data frame 1212 rows 6 variables: clutch RETFs lay eggs gelatinous \"clutches\" 30-40 eggs. Eggs clutch ID siblings ! variable useful mixed effects models. (Unordered factor.) treatment treatment group embryo. Either \"gentamicin\", compound knocks embryos' lateral line, \"control\" negative control group (.e. sensory organs intact). (Character.) reflex measure ear function called vestibulo-ocular reflex, categorized bins. Ear function increases factor levels \"low\", \"mid\", \"full\". (Ordered factor.) age Age embryo, seconds, time embryo jiggled. (Numeric, seconds.) t_o_d time day stimulus (.e. jiggle) applied. \"morning\" 5 .m. noon, \"afternoon\" noon 8 p.m., \"night\" 8 p.m. 5 .m. (Character.) hatched Whether embryo hatched response jiggling! Either \"yes\" \"\". (Character.) latency Time elapsed stimulus (.e. jiggling) hatching response stimulus, seconds. Missing values indicate embryo hatch response stimulus. (Numeric, seconds.)","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/tree_frogs.html","id":"source","dir":"Reference","previous_headings":"","what":"Source","title":"Tree frog embryo hatching data — tree_frogs","text":"Julie Jung et al. (2020) Multimodal mechanosensing enables treefrog embryos escape egg-predators. doi:10.1242/jeb.236141","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/tree_frogs.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Tree frog embryo hatching data — tree_frogs","text":"Note data included stacks package necessarily representative unbiased subset complete dataset, demonstrative purposes.","code":""},{"path":"https://stacks.tidymodels.org/dev/news/index.html","id":"stacks-development-version","dir":"Changelog","previous_headings":"","what":"stacks (development version)","title":"stacks (development version)","text":"Refine package alt text (#202). Update example objects, resolving deprecation warnings recipes (#203). Fix bug type checking blend_predictions(mixture) (#204). Resolve package-level documentation aliasing notice CRAN.","code":""},{"path":"https://stacks.tidymodels.org/dev/news/index.html","id":"stacks-102","dir":"Changelog","previous_headings":"","what":"stacks 1.0.2","title":"stacks 1.0.2","text":"CRAN release: 2023-04-20 Added augment() method model_stack objects (#173). Converted character variables tree_frogs example data factor updated downstream example objects (#177). Fixed bug resulted errors using model formulas \"mgcv\" engine (#193). Made several optimizations reduce evaluation time memory allocation stacking. Various bug fixes improvements documentation.","code":""},{"path":"https://stacks.tidymodels.org/dev/news/index.html","id":"stacks-101","dir":"Changelog","previous_headings":"","what":"stacks 1.0.1","title":"stacks 1.0.1","text":"CRAN release: 2022-12-14 Removes unneeded data import attribute tree_frogs example data associated objects (#148). blend_predictions() doesn’t error anymore control argument isn’t control_grid object. long object passed control include elements control_grid() output, parsnip::condense_control() handle input (#149). Tightened integration workflowsets package (#161, #165). Refined logic adding candidates via workflowsets allow partially trained workflow sets. case workflow set contains failed tuning results, stacks inform user excluded data stack add results trained successfully. Extended documentation related packages’ interactions, including new article package website. Revamped errors, warnings, messages. Prompts now provide thorough context arose, include extensive references documentation, correctly pluralized (#150, #167). Various bug fixes improvements documentation.","code":""},{"path":"https://stacks.tidymodels.org/dev/news/index.html","id":"stacks-100","dir":"Changelog","previous_headings":"","what":"stacks 1.0.0","title":"stacks 1.0.0","text":"CRAN release: 2022-07-06 stacks 1.0.0 first production release package. release includes minor bug fixes, ’s accompanied white paper recently published Journal Open Source software. can read paper ! release: Addresses re-introduction bug arising outcome levels valid column names multinomial classification setting (#133). Fixes bug stacks return incorrect predictions elastic net meta-learner used, type argument predict set \"class\", outcome levels differ alphabetical order. Transitions package internals functions deprecated recipes package.","code":""},{"path":"https://stacks.tidymodels.org/dev/news/index.html","id":"stacks-024-github-only","dir":"Changelog","previous_headings":"","what":"stacks 0.2.4 (GitHub only)","title":"stacks 0.2.4 (GitHub only)","text":"GitHub-release change package source code. update includes data-raw/paper subdirectory containing source contributed paper Journal Open Source Software.","code":""},{"path":"https://stacks.tidymodels.org/dev/news/index.html","id":"stacks-023","dir":"Changelog","previous_headings":"","what":"stacks 0.2.3","title":"stacks 0.2.3","text":"CRAN release: 2022-05-12 Addressed deprecation warning add_candidates (#99). Improved clarity warnings/errors related failed hyperparameter tuning resample fitting (#110). Reduced model stack object size fixed bug object size model stack inflated drastically saving file (#116). Also, regenerated example objects change–saved model objects may need regenerated order interface newer versions package. Introduced times argument blend_predictions passed rsample::bootstraps fitting stacking coefficients. Reducing argument default (25) greatly reduces run time blend_predictions (#94). package now load packages necessary model fitting fit_members(), available, fail informatively (#118). Fixed bug meta-learner tuning fail outcome names levels including string \"class\" (#125).","code":""},{"path":"https://stacks.tidymodels.org/dev/news/index.html","id":"stacks-022","dir":"Changelog","previous_headings":"","what":"stacks 0.2.2","title":"stacks 0.2.2","text":"CRAN release: 2022-01-06 Fixed errors arising outcome levels valid column names multinomial classification setting. Fixed collect_parameters failing return stacking coefficients two-class classification setting. Regenerated example objects updated {rsample} fingerprinting–saved model objects may need regenerated order build stacks combining models generated update.","code":""},{"path":"https://stacks.tidymodels.org/dev/news/index.html","id":"stacks-021","dir":"Changelog","previous_headings":"","what":"stacks 0.2.1","title":"stacks 0.2.1","text":"CRAN release: 2021-07-23 Updates importing workflow sets use add_variables() preprocessor. Plot fixes cases coefficients negative. Performance member plots now show effect multiple mixture values. Package diagrams now alt text.","code":""},{"path":"https://stacks.tidymodels.org/dev/news/index.html","id":"stacks-020","dir":"Changelog","previous_headings":"","what":"stacks 0.2.0","title":"stacks 0.2.0","text":"CRAN release: 2021-04-20","code":""},{"path":"https://stacks.tidymodels.org/dev/news/index.html","id":"breaking-changes-0-2-0","dir":"Changelog","previous_headings":"","what":"Breaking changes","title":"stacks 0.2.0","text":"release package changes elements internal structure model stacks. , model stacks stored saved objects need regenerated predicting, plotting, printing, etc.","code":""},{"path":"https://stacks.tidymodels.org/dev/news/index.html","id":"new-features-0-2-0","dir":"Changelog","previous_headings":"","what":"New features","title":"stacks 0.2.0","text":"package now supports elastic net models meta-learner via mixture argument blend_predictions. package can now add candidates workflow_map objects new {workflowsets} package. interface add_candidates tune_results objects, add_candidates now generic function. Objects tuned racing methods {finetune} package can now added candidate members.","code":""},{"path":"https://stacks.tidymodels.org/dev/news/index.html","id":"bug-fixes-0-2-0","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"stacks 0.2.0","text":"Fixed bug determining member hyperparameters member fitting using non-RMSE/ROC AUC metrics. Fixed bug arising model definition names valid column names. package now message case provided names valid column names use make.names associated candidate members.","code":""},{"path":"https://stacks.tidymodels.org/dev/news/index.html","id":"miscellaneous-improvements-0-2-0","dir":"Changelog","previous_headings":"","what":"Miscellaneous improvements","title":"stacks 0.2.0","text":"Drop {digest} dependency favor {tune}/{rsample} “fingerprinting” check consistency resamples. fit_members() now warn supplied model stack whose members already fitted. Integrate {tune} functionality appropriately coloring errors, warnings, messages. Improved faceting axis scales make autoplot type = \"members\" informative. Various improvements documentation.","code":""},{"path":"https://stacks.tidymodels.org/dev/news/index.html","id":"stacks-010","dir":"Changelog","previous_headings":"","what":"stacks 0.1.0","title":"stacks 0.1.0","text":"CRAN release: 2020-11-23 Initial release!","code":""}]
+[{"path":[]},{"path":"https://stacks.tidymodels.org/dev/CODE_OF_CONDUCT.html","id":"our-pledge","dir":"","previous_headings":"","what":"Our Pledge","title":"Contributor Covenant Code of Conduct","text":"members, contributors, leaders pledge make participation community harassment-free experience everyone, regardless age, body size, visible invisible disability, ethnicity, sex characteristics, gender identity expression, level experience, education, socio-economic status, nationality, personal appearance, race, caste, color, religion, sexual identity orientation. pledge act interact ways contribute open, welcoming, diverse, inclusive, healthy community.","code":""},{"path":"https://stacks.tidymodels.org/dev/CODE_OF_CONDUCT.html","id":"our-standards","dir":"","previous_headings":"","what":"Our Standards","title":"Contributor Covenant Code of Conduct","text":"Examples behavior contributes positive environment community include: Demonstrating empathy kindness toward people respectful differing opinions, viewpoints, experiences Giving gracefully accepting constructive feedback Accepting responsibility apologizing affected mistakes, learning experience Focusing best just us individuals, overall community Examples unacceptable behavior include: use sexualized language imagery, sexual attention advances kind Trolling, insulting derogatory comments, personal political attacks Public private harassment Publishing others’ private information, physical email address, without explicit permission conduct reasonably considered inappropriate professional setting","code":""},{"path":"https://stacks.tidymodels.org/dev/CODE_OF_CONDUCT.html","id":"enforcement-responsibilities","dir":"","previous_headings":"","what":"Enforcement Responsibilities","title":"Contributor Covenant Code of Conduct","text":"Community leaders responsible clarifying enforcing standards acceptable behavior take appropriate fair corrective action response behavior deem inappropriate, threatening, offensive, harmful. Community leaders right responsibility remove, edit, reject comments, commits, code, wiki edits, issues, contributions aligned Code Conduct, communicate reasons moderation decisions appropriate.","code":""},{"path":"https://stacks.tidymodels.org/dev/CODE_OF_CONDUCT.html","id":"scope","dir":"","previous_headings":"","what":"Scope","title":"Contributor Covenant Code of Conduct","text":"Code Conduct applies within community spaces, also applies individual officially representing community public spaces. Examples representing community include using official e-mail address, posting via official social media account, acting appointed representative online offline event.","code":""},{"path":"https://stacks.tidymodels.org/dev/CODE_OF_CONDUCT.html","id":"enforcement","dir":"","previous_headings":"","what":"Enforcement","title":"Contributor Covenant Code of Conduct","text":"Instances abusive, harassing, otherwise unacceptable behavior may reported community leaders responsible enforcement codeofconduct@posit.co. complaints reviewed investigated promptly fairly. community leaders obligated respect privacy security reporter incident.","code":""},{"path":"https://stacks.tidymodels.org/dev/CODE_OF_CONDUCT.html","id":"enforcement-guidelines","dir":"","previous_headings":"","what":"Enforcement Guidelines","title":"Contributor Covenant Code of Conduct","text":"Community leaders follow Community Impact Guidelines determining consequences action deem violation Code Conduct:","code":""},{"path":"https://stacks.tidymodels.org/dev/CODE_OF_CONDUCT.html","id":"id_1-correction","dir":"","previous_headings":"Enforcement Guidelines","what":"1. Correction","title":"Contributor Covenant Code of Conduct","text":"Community Impact: Use inappropriate language behavior deemed unprofessional unwelcome community. Consequence: private, written warning community leaders, providing clarity around nature violation explanation behavior inappropriate. public apology may requested.","code":""},{"path":"https://stacks.tidymodels.org/dev/CODE_OF_CONDUCT.html","id":"id_2-warning","dir":"","previous_headings":"Enforcement Guidelines","what":"2. Warning","title":"Contributor Covenant Code of Conduct","text":"Community Impact: violation single incident series actions. Consequence: warning consequences continued behavior. interaction people involved, including unsolicited interaction enforcing Code Conduct, specified period time. includes avoiding interactions community spaces well external channels like social media. Violating terms may lead temporary permanent ban.","code":""},{"path":"https://stacks.tidymodels.org/dev/CODE_OF_CONDUCT.html","id":"id_3-temporary-ban","dir":"","previous_headings":"Enforcement Guidelines","what":"3. Temporary Ban","title":"Contributor Covenant Code of Conduct","text":"Community Impact: serious violation community standards, including sustained inappropriate behavior. Consequence: temporary ban sort interaction public communication community specified period time. public private interaction people involved, including unsolicited interaction enforcing Code Conduct, allowed period. Violating terms may lead permanent ban.","code":""},{"path":"https://stacks.tidymodels.org/dev/CODE_OF_CONDUCT.html","id":"id_4-permanent-ban","dir":"","previous_headings":"Enforcement Guidelines","what":"4. Permanent Ban","title":"Contributor Covenant Code of Conduct","text":"Community Impact: Demonstrating pattern violation community standards, including sustained inappropriate behavior, harassment individual, aggression toward disparagement classes individuals. Consequence: permanent ban sort public interaction within community.","code":""},{"path":"https://stacks.tidymodels.org/dev/CODE_OF_CONDUCT.html","id":"attribution","dir":"","previous_headings":"","what":"Attribution","title":"Contributor Covenant Code of Conduct","text":"Code Conduct adapted Contributor Covenant, version 2.1, available https://www.contributor-covenant.org/version/2/1/code_of_conduct.html. Community Impact Guidelines inspired [Mozilla’s code conduct enforcement ladder][https://github.com/mozilla/inclusion]. answers common questions code conduct, see FAQ https://www.contributor-covenant.org/faq. Translations available https://www.contributor-covenant.org/translations.","code":""},{"path":"https://stacks.tidymodels.org/dev/LICENSE.html","id":null,"dir":"","previous_headings":"","what":"MIT License","title":"MIT License","text":"Copyright (c) 2023 stacks authors Permission hereby granted, free charge, person obtaining copy software associated documentation files (“Software”), deal Software without restriction, including without limitation rights use, copy, modify, merge, publish, distribute, sublicense, /sell copies Software, permit persons Software furnished , subject following conditions: copyright notice permission notice shall included copies substantial portions Software. SOFTWARE PROVIDED “”, WITHOUT WARRANTY KIND, EXPRESS IMPLIED, INCLUDING LIMITED WARRANTIES MERCHANTABILITY, FITNESS PARTICULAR PURPOSE NONINFRINGEMENT. EVENT SHALL AUTHORS COPYRIGHT HOLDERS LIABLE CLAIM, DAMAGES LIABILITY, WHETHER ACTION CONTRACT, TORT OTHERWISE, ARISING , CONNECTION SOFTWARE USE DEALINGS SOFTWARE.","code":""},{"path":"https://stacks.tidymodels.org/dev/articles/basics.html","id":"define-candidate-ensemble-members","dir":"Articles","previous_headings":"","what":"Define candidate ensemble members","title":"Getting Started With stacks","text":"highest level, ensembles formed model definitions. package, model definitions instance minimal workflow, containing model specification (defined parsnip package) , optionally, preprocessor (defined recipes package). Model definitions specify form candidate ensemble members.  Defining constituent model definitions undoubtedly longest part building ensemble stacks. ’re familiar tidymodels “proper,” ’re probably fine skip section, keeping things mind: ’ll need save assessment set predictions workflow utilized tune_grid(), tune_bayes(), fit_resamples() objects setting control arguments save_pred = TRUE save_workflow = TRUE. Note use control_stack_*() convenience functions ! model definition must share rsample rset object. ’ll first start splitting training data, generating resamples, setting options used model definition. Tuning fitting results use ensembles need fitted control arguments save_pred = TRUE save_workflow = TRUE—settings ensure assessment set predictions, well workflow used fit resamples, stored resulting object. convenience, stacks supplies control_stack_*() functions generate appropriate objects . example, ’ll working tune_grid() fit_resamples() tune package, use following control settings: ’ll define three different model definitions try predict time hatch—K-nearest neighbors model (hyperparameters tune), linear model, support vector machine model (, hyperparameters tune). Starting K-nearest neighbors, begin creating parsnip model specification: Note , since tuning several possible numbers neighbors, model specification defines multiple model configurations. specific form configurations determined specifying grid search tune_grid(). , extend basic recipe defined earlier fully specify form design matrix use K-nearest neighbors model: Starting basic recipe, convert categorical variables dummy variables, remove column one observation, impute missing values numeric variables using mean, normalize numeric predictors. Pre-processing instructions remaining models defined similarly. Now, combine model specification pre-processing instructions defined form workflow object: Finally, can make use workflow, training set resamples, metric set, control object tune hyperparameters. Using grid argument, specify like optimize four possible values k using grid search. knn_res object fully specifies candidate members, ready included stacks workflow. Now, specifying linear model, note optimizing hyperparameters. Thus, use fit_resamples() function rather tune_grid() tune_bayes() fitting resamples. Finally, putting together model definition support vector machine: Altogether, ’ve created three model definitions, K-nearest neighbors model definition specifies 4 model configurations, linear regression specifies 1, support vector machine specifies 6.  three model definitions fully specified, ready begin stacking model configurations. (Note , applied settings, one likely specify many 11 candidate members.)","code":"# some setup: resampling and a basic recipe set.seed(1) tree_frogs_split <- initial_split(tree_frogs) tree_frogs_train <- training(tree_frogs_split) tree_frogs_test  <- testing(tree_frogs_split)  set.seed(1) folds <- rsample::vfold_cv(tree_frogs_train, v = 5)  tree_frogs_rec <-    recipe(latency ~ ., data = tree_frogs_train)  metric <- metric_set(rmse) ctrl_grid <- control_stack_grid() ctrl_res <- control_stack_resamples() # create a model definition knn_spec <-   nearest_neighbor(     mode = \"regression\",      neighbors = tune(\"k\")   ) %>%   set_engine(\"kknn\")  knn_spec #> K-Nearest Neighbor Model Specification (regression) #>  #> Main Arguments: #>   neighbors = tune(\"k\") #>  #> Computational engine: kknn # extend the recipe knn_rec <-   tree_frogs_rec %>%   step_dummy(all_nominal_predictors()) %>%   step_zv(all_predictors()) %>%   step_impute_mean(all_numeric_predictors()) %>%   step_normalize(all_numeric_predictors())  knn_rec #>  #> ── Recipe ──────────────────────────────────────────────────────────────── #>  #> ── Inputs #> Number of variables by role #> outcome:   1 #> predictor: 4 #>  #> ── Operations #> • Dummy variables from: all_nominal_predictors() #> • Zero variance filter on: all_predictors() #> • Mean imputation for: all_numeric_predictors() #> • Centering and scaling for: all_numeric_predictors() # add both to a workflow knn_wflow <-    workflow() %>%    add_model(knn_spec) %>%   add_recipe(knn_rec)  knn_wflow #> ══ Workflow ══════════════════════════════════════════════════════════════ #> Preprocessor: Recipe #> Model: nearest_neighbor() #>  #> ── Preprocessor ────────────────────────────────────────────────────────── #> 4 Recipe Steps #>  #> • step_dummy() #> • step_zv() #> • step_impute_mean() #> • step_normalize() #>  #> ── Model ───────────────────────────────────────────────────────────────── #> K-Nearest Neighbor Model Specification (regression) #>  #> Main Arguments: #>   neighbors = tune(\"k\") #>  #> Computational engine: kknn # tune k and fit to the 5-fold cv set.seed(2020) knn_res <-    tune_grid(     knn_wflow,     resamples = folds,     metrics = metric,     grid = 4,     control = ctrl_grid   )  knn_res #> # Tuning results #> # 5-fold cross-validation  #> # A tibble: 5 × 5 #>   splits           id    .metrics         .notes           .predictions #>   <list>           <chr> <list>           <list>           <list>       #> 1 <split [343/86]> Fold1 <tibble [4 × 5]> <tibble [0 × 3]> <tibble>     #> 2 <split [343/86]> Fold2 <tibble [4 × 5]> <tibble [0 × 3]> <tibble>     #> 3 <split [343/86]> Fold3 <tibble [4 × 5]> <tibble [0 × 3]> <tibble>     #> 4 <split [343/86]> Fold4 <tibble [4 × 5]> <tibble [0 × 3]> <tibble>     #> 5 <split [344/85]> Fold5 <tibble [4 × 5]> <tibble [0 × 3]> <tibble> # create a model definition lin_reg_spec <-   linear_reg() %>%   set_engine(\"lm\")  # extend the recipe lin_reg_rec <-   tree_frogs_rec %>%   step_dummy(all_nominal_predictors()) %>%   step_zv(all_predictors())  # add both to a workflow lin_reg_wflow <-    workflow() %>%   add_model(lin_reg_spec) %>%   add_recipe(lin_reg_rec)  # fit to the 5-fold cv set.seed(2020) lin_reg_res <-    fit_resamples(     lin_reg_wflow,     resamples = folds,     metrics = metric,     control = ctrl_res   )  lin_reg_res #> # Resampling results #> # 5-fold cross-validation  #> # A tibble: 5 × 5 #>   splits           id    .metrics         .notes           .predictions #>   <list>           <chr> <list>           <list>           <list>       #> 1 <split [343/86]> Fold1 <tibble [1 × 4]> <tibble [0 × 3]> <tibble>     #> 2 <split [343/86]> Fold2 <tibble [1 × 4]> <tibble [0 × 3]> <tibble>     #> 3 <split [343/86]> Fold3 <tibble [1 × 4]> <tibble [0 × 3]> <tibble>     #> 4 <split [343/86]> Fold4 <tibble [1 × 4]> <tibble [0 × 3]> <tibble>     #> 5 <split [344/85]> Fold5 <tibble [1 × 4]> <tibble [0 × 3]> <tibble> # create a model definition svm_spec <-    svm_rbf(     cost = tune(\"cost\"),      rbf_sigma = tune(\"sigma\")   ) %>%   set_engine(\"kernlab\") %>%   set_mode(\"regression\")  # extend the recipe svm_rec <-   tree_frogs_rec %>%   step_dummy(all_nominal_predictors()) %>%   step_zv(all_predictors()) %>%   step_impute_mean(all_numeric_predictors()) %>%   step_corr(all_predictors()) %>%   step_normalize(all_numeric_predictors())  # add both to a workflow svm_wflow <-    workflow() %>%    add_model(svm_spec) %>%   add_recipe(svm_rec)  # tune cost and sigma and fit to the 5-fold cv set.seed(2020) svm_res <-    tune_grid(     svm_wflow,      resamples = folds,      grid = 6,     metrics = metric,     control = ctrl_grid   )  svm_res #> # Tuning results #> # 5-fold cross-validation  #> # A tibble: 5 × 5 #>   splits           id    .metrics         .notes           .predictions #>   <list>           <chr> <list>           <list>           <list>       #> 1 <split [343/86]> Fold1 <tibble [6 × 6]> <tibble [0 × 3]> <tibble>     #> 2 <split [343/86]> Fold2 <tibble [6 × 6]> <tibble [0 × 3]> <tibble>     #> 3 <split [343/86]> Fold3 <tibble [6 × 6]> <tibble [0 × 3]> <tibble>     #> 4 <split [343/86]> Fold4 <tibble [6 × 6]> <tibble [0 × 3]> <tibble>     #> 5 <split [344/85]> Fold5 <tibble [6 × 6]> <tibble [0 × 3]> <tibble>"},{"path":"https://stacks.tidymodels.org/dev/articles/basics.html","id":"putting-together-a-stack","dir":"Articles","previous_headings":"","what":"Putting together a stack","title":"Getting Started With stacks","text":"first step building ensemble stacks create data_stack object—package, data stacks tibbles (extra attributes) contain assessment set predictions candidate ensemble member.  can initialize data stack using stacks() function. stacks() function works sort like ggplot() constructor ggplot2—function creates basic structure object built top —except ’ll pipe outputs rather adding +. add_candidates() function adds ensemble members stack. mentioned , hood, data_stack object really just tibble extra attributes. Checking actual data: first column gives first response value, remaining columns give assessment set predictions ensemble member. Since ’re regression case, ’s one column per ensemble member. classification settings, many columns levels outcome variable per candidate ensemble member. ’s ! ’re now ready evaluate need combine predictions candidate ensemble member.","code":"stacks() #> # A data stack with 0 model definitions and 0 candidate members. tree_frogs_data_st <-    stacks() %>%   add_candidates(knn_res) %>%   add_candidates(lin_reg_res) %>%   add_candidates(svm_res)  tree_frogs_data_st #> # A data stack with 3 model definitions and 11 candidate members: #> #   knn_res: 4 model configurations #> #   lin_reg_res: 1 model configuration #> #   svm_res: 6 model configurations #> # Outcome: latency (numeric) as_tibble(tree_frogs_data_st) #> # A tibble: 429 × 12 #>    latency knn_res_1_1 knn_res_1_2 knn_res_1_3 knn_res_1_4 lin_reg_res_1_1 #>      <dbl>       <dbl>       <dbl>       <dbl>       <dbl>           <dbl> #>  1     142        76.5        67.8        83.2        83.0           114.  #>  2      79        74.3        71.0        64.2        63.9            78.6 #>  3      50        75.0        71.4        64.9        64.3            81.5 #>  4      68        47.9        49.7        54.0        54.0            78.6 #>  5      64        60.9        53.1        48.4        48.3            36.5 #>  6      52       207.        168.        129.        124.            124.  #>  7      39        22.6        24.9        31.0        31.5            35.2 #>  8      46        23.2        26.6        33.4        33.9            37.1 #>  9     137        75.9        78.8        90.4        90.6            78.8 #> 10      73        67.7        73.6        67.9        67.4            38.8 #> # ℹ 419 more rows #> # ℹ 6 more variables: svm_res_1_1 <dbl>, svm_res_1_4 <dbl>, #> #   svm_res_1_3 <dbl>, svm_res_1_5 <dbl>, svm_res_1_2 <dbl>, #> #   svm_res_1_6 <dbl>"},{"path":"https://stacks.tidymodels.org/dev/articles/basics.html","id":"fit-the-stack","dir":"Articles","previous_headings":"","what":"Fit the stack","title":"Getting Started With stacks","text":"outputs candidate ensemble members highly correlated, blend_predictions() function performs regularization figure can combine outputs stack members come final prediction. blend_predictions function determines member model output ultimately combined final prediction fitting LASSO model data stack, predicting true assessment set outcome using predictions candidate members. Candidates nonzero stacking coefficients become members.  make sure right trade-minimizing number members optimizing performance, can use autoplot() method:  show relationship directly:  results good enough, blend_predictions() called different values penalty. , blend_predictions() picks penalty parameter numerically optimal results. see top results:  Now know combine model output, can fit candidates non-zero stacking coefficients full training set.  Model stacks can thought group fitted member models set instructions combine predictions.  identify model configurations assigned stacking coefficients, can make use collect_parameters() function: object now ready predict new data! Juxtaposing predictions true data:  Looks like predictions pretty strong! stacks predictions perform, though, compared members’ predictions? can use type = \"members\" argument generate predictions ensemble members. Now, evaluating root mean squared error model: can see, stacked ensemble outperforms member models, though closely followed one members. Voila! ’ve now made use stacks package predict red-eyed tree frog embryo hatching using stacked ensemble! full visual outline steps can found .","code":"tree_frogs_model_st <-   tree_frogs_data_st %>%   blend_predictions() autoplot(tree_frogs_model_st) autoplot(tree_frogs_model_st, type = \"members\") autoplot(tree_frogs_model_st, type = \"weights\") tree_frogs_model_st <-   tree_frogs_model_st %>%   fit_members() collect_parameters(tree_frogs_model_st, \"svm_res\") #> # A tibble: 6 × 4 #>   member         cost    sigma  coef #>   <chr>         <dbl>    <dbl> <dbl> #> 1 svm_res_1_1 0.00143 6.64e- 9 0     #> 2 svm_res_1_2 3.59    3.95e- 4 0.931 #> 3 svm_res_1_3 0.0978  1.81e- 2 0     #> 4 svm_res_1_4 0.00849 2.16e-10 0     #> 5 svm_res_1_5 0.256   4.54e- 1 0.159 #> 6 svm_res_1_6 7.64    3.16e- 7 0 tree_frogs_test <-    tree_frogs_test %>%   bind_cols(predict(tree_frogs_model_st, .)) ggplot(tree_frogs_test) +   aes(x = latency,        y = .pred) +   geom_point() +    coord_obs_pred() member_preds <-    tree_frogs_test %>%   select(latency) %>%   bind_cols(predict(tree_frogs_model_st, tree_frogs_test, members = TRUE)) map(member_preds, rmse_vec, truth = member_preds$latency) %>%   as_tibble() #> # A tibble: 1 × 6 #>   latency .pred knn_res_1_4 lin_reg_res_1_1 svm_res_1_5 svm_res_1_2 #>     <dbl> <dbl>       <dbl>           <dbl>       <dbl>       <dbl> #> 1       0  54.3        56.1            55.5        56.9        60.6"},{"path":"https://stacks.tidymodels.org/dev/articles/classification.html","id":"defining-candidate-ensemble-members","dir":"Articles","previous_headings":"","what":"Defining candidate ensemble members","title":"Classification Models With stacks","text":"numeric prediction setting, defining candidate ensemble members undoubtedly longest part ensembling process stacks. First, splitting training data, generating resamples, setting options used model definition. also need use control settings numeric response setting: ’ll define two different model definitions try predict reflex—random forest neural network. Starting random forest: Now, moving neural network model definition: model definitions fully specified, ’re ready start putting together ensemble!","code":"# some setup: resampling and a basic recipe set.seed(1)  tree_frogs_split <- initial_split(tree_frogs) tree_frogs_train <- training(tree_frogs_split) tree_frogs_test  <- testing(tree_frogs_split)  folds <- rsample::vfold_cv(tree_frogs_train, v = 5)  tree_frogs_rec <-    recipe(reflex ~ ., data = tree_frogs_train) %>%   step_dummy(all_nominal_predictors(), -reflex) %>%   step_zv(all_predictors())  tree_frogs_wflow <-    workflow() %>%    add_recipe(tree_frogs_rec) ctrl_grid <- control_stack_grid() rand_forest_spec <-    rand_forest(     mtry = tune(),     min_n = tune(),     trees = 500   ) %>%   set_mode(\"classification\") %>%   set_engine(\"ranger\")  rand_forest_wflow <-   tree_frogs_wflow %>%   add_model(rand_forest_spec)  rand_forest_res <-    tune_grid(     object = rand_forest_wflow,      resamples = folds,      grid = 10,     control = ctrl_grid   ) nnet_spec <-   mlp(hidden_units = tune(), penalty = tune(), epochs = tune()) %>%   set_mode(\"classification\") %>%   set_engine(\"nnet\")  nnet_rec <-    tree_frogs_rec %>%    step_normalize(all_predictors())  nnet_wflow <-    tree_frogs_wflow %>%   add_model(nnet_spec) %>%   update_recipe(nnet_rec)  nnet_res <-   tune_grid(     object = nnet_wflow,      resamples = folds,      grid = 10,     control = ctrl_grid   )"},{"path":"https://stacks.tidymodels.org/dev/articles/classification.html","id":"putting-together-a-stack","dir":"Articles","previous_headings":"","what":"Putting together a stack","title":"Classification Models With stacks","text":"Building stacked ensemble, now, takes lines: make sure right trade-minimizing number members optimizing performance, can use autoplot() method:  show relationship directly:  results good enough, blend_predictions() called different values penalty. , blend_predictions() picks penalty parameter numerically optimal results. see top results:  multiple facets since ensemble members can different effects different classes. identify model configurations assigned stacking coefficients, can make use collect_parameters() function: object now ready predict new data! Computing ROC AUC model: Looks like predictions pretty strong! stacks predictions perform, though, compared members’ predictions? can use members argument generate predictions ensemble members. Voilà! ’ve now made use stacks package predict tree frog embryo ear function using stacked ensemble!","code":"tree_frogs_model_st <-    # initialize the stack   stacks() %>%   # add candidate members   add_candidates(rand_forest_res) %>%   add_candidates(nnet_res) %>%   # determine how to combine their predictions   blend_predictions() %>%   # fit the candidates with nonzero stacking coefficients   fit_members()  tree_frogs_model_st #> # A tibble: 10 × 4 #>    member                         type        weight class #>    <chr>                          <chr>        <dbl> <fct> #>  1 .pred_full_nnet_res_1_04       mlp          21.6  full  #>  2 .pred_mid_nnet_res_1_04        mlp          18.2  mid   #>  3 .pred_mid_rand_forest_res_1_05 rand_forest  13.6  low   #>  4 .pred_mid_rand_forest_res_1_08 rand_forest   6.47 mid   #>  5 .pred_mid_nnet_res_1_01        mlp           4.60 low   #>  6 .pred_mid_nnet_res_1_03        mlp           4.58 low   #>  7 .pred_mid_rand_forest_res_1_07 rand_forest   4.07 low   #>  8 .pred_mid_nnet_res_1_10        mlp           3.49 mid   #>  9 .pred_full_nnet_res_1_01       mlp           3.09 full  #> 10 .pred_mid_nnet_res_1_09        mlp           1.43 mid autoplot(tree_frogs_model_st) autoplot(tree_frogs_model_st, type = \"members\") autoplot(tree_frogs_model_st, type = \"weights\") collect_parameters(tree_frogs_model_st, \"rand_forest_res\") #> # A tibble: 60 × 6 #>    member                mtry min_n class terms                       coef #>    <chr>                <int> <int> <chr> <chr>                      <dbl> #>  1 rand_forest_res_1_01     1    26 low   .pred_mid_rand_forest_res…     0 #>  2 rand_forest_res_1_01     1    26 low   .pred_full_rand_forest_re…     0 #>  3 rand_forest_res_1_01     1    26 mid   .pred_mid_rand_forest_res…     0 #>  4 rand_forest_res_1_01     1    26 mid   .pred_full_rand_forest_re…     0 #>  5 rand_forest_res_1_01     1    26 full  .pred_mid_rand_forest_res…     0 #>  6 rand_forest_res_1_01     1    26 full  .pred_full_rand_forest_re…     0 #>  7 rand_forest_res_1_02     2    33 low   .pred_mid_rand_forest_res…     0 #>  8 rand_forest_res_1_02     2    33 low   .pred_full_rand_forest_re…     0 #>  9 rand_forest_res_1_02     2    33 mid   .pred_mid_rand_forest_res…     0 #> 10 rand_forest_res_1_02     2    33 mid   .pred_full_rand_forest_re…     0 #> # ℹ 50 more rows tree_frogs_pred <-   tree_frogs_test %>%   bind_cols(predict(tree_frogs_model_st, ., type = \"prob\")) yardstick::roc_auc(   tree_frogs_pred,   truth = reflex,   contains(\".pred_\")   ) tree_frogs_pred <-   tree_frogs_test %>%   select(reflex) %>%   bind_cols(     predict(       tree_frogs_model_st,       tree_frogs_test,       type = \"class\",       members = TRUE       )     )  tree_frogs_pred #> # A tibble: 303 × 15 #>    reflex .pred_class .pred_class_rand_forest_res…¹ .pred_class_rand_for…² #>    <fct>  <fct>       <fct>                         <fct>                  #>  1 full   full        full                          full                   #>  2 mid    mid         low                           mid                    #>  3 mid    mid         mid                           mid                    #>  4 mid    mid         low                           low                    #>  5 full   full        full                          full                   #>  6 full   full        full                          full                   #>  7 full   full        full                          full                   #>  8 full   full        full                          full                   #>  9 full   full        full                          full                   #> 10 full   full        full                          full                   #> # ℹ 293 more rows #> # ℹ abbreviated names: ¹​.pred_class_rand_forest_res_1_05, #> #   ²​.pred_class_rand_forest_res_1_07 #> # ℹ 11 more variables: .pred_class_nnet_res_1_07 <fct>, #> #   .pred_class_nnet_res_1_02 <fct>, .pred_class_nnet_res_1_03 <fct>, #> #   .pred_class_nnet_res_1_01 <fct>, #> #   .pred_class_rand_forest_res_1_09 <fct>, …  map(   colnames(tree_frogs_pred),   ~mean(tree_frogs_pred$reflex == pull(tree_frogs_pred, .x)) ) %>%   set_names(colnames(tree_frogs_pred)) %>%   as_tibble() %>%   pivot_longer(c(everything(), -reflex)) #> # A tibble: 14 × 3 #>    reflex name                             value #>     <dbl> <chr>                            <dbl> #>  1      1 .pred_class                      0.884 #>  2      1 .pred_class_rand_forest_res_1_05 0.855 #>  3      1 .pred_class_rand_forest_res_1_07 0.865 #>  4      1 .pred_class_nnet_res_1_07        0.868 #>  5      1 .pred_class_nnet_res_1_02        0.848 #>  6      1 .pred_class_nnet_res_1_03        0.845 #>  7      1 .pred_class_nnet_res_1_01        0.842 #>  8      1 .pred_class_rand_forest_res_1_09 0.881 #>  9      1 .pred_class_rand_forest_res_1_08 0.881 #> 10      1 .pred_class_nnet_res_1_05        0.789 #> 11      1 .pred_class_nnet_res_1_09        0.835 #> 12      1 .pred_class_nnet_res_1_06        0.858 #> 13      1 .pred_class_nnet_res_1_04        0.855 #> 14      1 .pred_class_nnet_res_1_10        0.851"},{"path":"https://stacks.tidymodels.org/dev/articles/workflowsets.html","id":"setup","dir":"Articles","previous_headings":"","what":"Setup","title":"Stacking With Workflow Sets","text":"example parallel “Getting Started” vignette, except use workflowsets bundle model workflows define candidate members one workflow set. train using workflowsets::workflow_map() add results model stack using add_candidates(). ’re familiar stacks package, refer vignette big picture! First, loading needed packages: example, ’ll make use tree_frogs data exported stacks, giving experimental results hatching behavior red-eyed tree frog embryos! Red-eyed tree frog (RETF) embryos can hatch earlier normal 7ish days detect potential predator threat. Researchers wanted determine , , tree frog embryos able detect stimulus environment. , subjected embryos varying developmental stages “predator stimulus” jiggling embryos blunt probe. Beforehand, though embryos treated gentamicin, compound knocks lateral line (sensory organ.) Researcher Julie Jung crew found factors inform whether embryo hatches prematurely ! ’ll start predicting latency (.e. time hatch) based attributes. ’ll need filter NAs (.e. cases embryo hatch) first. Taking quick look data, seems like hatch time pretty closely related predictors!  Let’s give go!","code":"library(tidymodels) library(stacks) data(\"tree_frogs\")  # subset the data tree_frogs <- tree_frogs %>%   filter(!is.na(latency)) %>%   select(-c(clutch, hatched)) theme_set(theme_bw())  ggplot(tree_frogs) +   aes(x = age, y = latency, color = treatment) +   geom_point() +   labs(x = \"Embryo Age (s)\", y = \"Time to Hatch (s)\", col = \"Treatment\")"},{"path":"https://stacks.tidymodels.org/dev/articles/workflowsets.html","id":"defining-candidate-ensemble-members","dir":"Articles","previous_headings":"","what":"Defining candidate ensemble members","title":"Stacking With Workflow Sets","text":"workflow-based setting, defining candidate ensemble members undoubtedly longest part ensembling process stacks. Instead evaluating model specifications via resampling workflow, though, combine model specifications workflow_set(), evaluate batch workflow_map() function. First, splitting training data, generating resamples, setting options used model definition: also need use control settings workflow-based setting: ’ll define three different model definitions try predict time hatch—K-nearest neighbors model (hyperparameters tune), linear model, support vector machine model (, hyperparameters tune). Starting K-nearest neighbors, begin creating parsnip model specification: Note , since tuning several possible numbers neighbors, model specification defines multiple model configurations. specific form configurations determined specifying grid search. , extend basic recipe defined earlier fully specify form design matrix use K-nearest neighbors model: Starting basic recipe, convert categorical predictors dummy variables, remove predictors one observation, impute missing values numeric predictors using mean, normalize numeric predictors. Pre-processing instructions remaining models defined similarly. Now, specifying linear model, note optimizing hyperparameters. Finally, putting together model specification recipe support vector machine: model specification accompanying recipe now defined, can combine via workflow_set: Note combination preprocessor model specification assigned wflow_id can use interface individual model definitions: elements workflow set nearly ready evaluated tune_grid(). , though, need ensure fit appropriate control options, just evaluating individual workflows resamples passing stacks. can iteratively add appropriate control settings control_stack_grid() shorthand using option_add() function: can now use workflow_map() function map model definition, evaluating hyperparameters supplied resamples. results section now contains three sets tuning results. Note results corresponding linear regression subclass resample_results rather tune_results—expected, hyperparameters tune model specification. workflow_map() knows fall back fit_resamples() rather tune_grid(), case! can extract tuning results extract_workflow_set_result() helper explore tuning results: three model definitions fully specified tuned workflow set, ready begin stacking model configurations. (Note , applied settings, one likely specify many handful candidate members three model definitions.)","code":"# some setup: resampling and a basic recipe set.seed(1) tree_frogs_split <- initial_split(tree_frogs) tree_frogs_train <- training(tree_frogs_split) tree_frogs_test  <- testing(tree_frogs_split)  set.seed(1) folds <- rsample::vfold_cv(tree_frogs_train, v = 5)  tree_frogs_rec <-    recipe(latency ~ ., data = tree_frogs_train)  metric <- metric_set(rmse) ctrl_grid <- control_stack_grid() # create a model specification knn_spec <-   nearest_neighbor(     mode = \"regression\",      neighbors = tune(\"k\")   ) %>%   set_engine(\"kknn\")  knn_spec #> K-Nearest Neighbor Model Specification (regression) #>  #> Main Arguments: #>   neighbors = tune(\"k\") #>  #> Computational engine: kknn # extend the recipe knn_rec <-   tree_frogs_rec %>%   step_dummy(all_nominal_predictors()) %>%   step_zv(all_predictors()) %>%   step_impute_mean(all_numeric_predictors()) %>%   step_normalize(all_numeric_predictors())  knn_rec #>  #> ── Recipe ──────────────────────────────────────────────────────────────── #>  #> ── Inputs #> Number of variables by role #> outcome:   1 #> predictor: 4 #>  #> ── Operations #> • Dummy variables from: all_nominal_predictors() #> • Zero variance filter on: all_predictors() #> • Mean imputation for: all_numeric_predictors() #> • Centering and scaling for: all_numeric_predictors() # create a model specification lin_reg_spec <-   linear_reg() %>%   set_engine(\"lm\")  # extend the recipe lin_reg_rec <-   tree_frogs_rec %>%   step_dummy(all_nominal_predictors()) %>%   step_zv(all_predictors()) # create a model specification svm_spec <-    svm_rbf(     cost = tune(\"cost\"),      rbf_sigma = tune(\"sigma\")   ) %>%   set_engine(\"kernlab\") %>%   set_mode(\"regression\")  # extend the recipe svm_rec <-   tree_frogs_rec %>%   step_dummy(all_nominal_predictors()) %>%   step_zv(all_predictors()) %>%   step_impute_mean(all_numeric_predictors()) %>%   step_corr(all_predictors()) %>%   step_normalize(all_numeric_predictors()) wf_set <-    workflow_set(     preproc = list(rec1 = knn_rec, rec2 = lin_reg_rec,     rec3 = svm_rec),     models =  list(knn = knn_spec, lin_reg = lin_reg_spec, svm = svm_spec),     cross = FALSE   )  wf_set #> # A workflow set/tibble: 3 × 4 #>   wflow_id     info             option    result     #>   <chr>        <list>           <list>    <list>     #> 1 rec1_knn     <tibble [1 × 4]> <opts[0]> <list [0]> #> 2 rec2_lin_reg <tibble [1 × 4]> <opts[0]> <list [0]> #> 3 rec3_svm     <tibble [1 × 4]> <opts[0]> <list [0]> wf_set %>%   extract_workflow(\"rec3_svm\") #> ══ Workflow ══════════════════════════════════════════════════════════════ #> Preprocessor: Recipe #> Model: svm_rbf() #>  #> ── Preprocessor ────────────────────────────────────────────────────────── #> 5 Recipe Steps #>  #> • step_dummy() #> • step_zv() #> • step_impute_mean() #> • step_corr() #> • step_normalize() #>  #> ── Model ───────────────────────────────────────────────────────────────── #> Radial Basis Function Support Vector Machine Model Specification (regression) #>  #> Main Arguments: #>   cost = tune(\"cost\") #>   rbf_sigma = tune(\"sigma\") #>  #> Computational engine: kernlab wf_set <-   wf_set %>%   option_add(     control = control_stack_grid(),     metrics = metric   ) wf_set_trained <-   workflow_map(     wf_set,     fn = \"tune_grid\",     resamples = folds   )  wf_set_trained #> # A workflow set/tibble: 3 × 4 #>   wflow_id     info             option    result    #>   <chr>        <list>           <list>    <list>    #> 1 rec1_knn     <tibble [1 × 4]> <opts[3]> <tune[+]> #> 2 rec2_lin_reg <tibble [1 × 4]> <opts[3]> <rsmp[+]> #> 3 rec3_svm     <tibble [1 × 4]> <opts[3]> <tune[+]> extract_workflow_set_result(wf_set_trained, id = \"rec1_knn\") %>%   collect_metrics(summarize = TRUE) #> # A tibble: 10 × 7 #>        k .metric .estimator  mean     n std_err .config               #>    <int> <chr>   <chr>      <dbl> <int>   <dbl> <chr>                 #>  1     1 rmse    standard    81.3     5    4.37 Preprocessor1_Model01 #>  2     3 rmse    standard    69.6     5    3.57 Preprocessor1_Model02 #>  3     5 rmse    standard    63.7     5    3.30 Preprocessor1_Model03 #>  4     6 rmse    standard    62.3     5    3.12 Preprocessor1_Model04 #>  5     7 rmse    standard    61.2     5    2.95 Preprocessor1_Model05 #>  6     9 rmse    standard    59.8     5    2.82 Preprocessor1_Model06 #>  7    10 rmse    standard    59.3     5    2.83 Preprocessor1_Model07 #>  8    11 rmse    standard    59.0     5    2.82 Preprocessor1_Model08 #>  9    13 rmse    standard    58.5     5    2.84 Preprocessor1_Model09 #> 10    15 rmse    standard    58.2     5    2.88 Preprocessor1_Model10"},{"path":"https://stacks.tidymodels.org/dev/articles/workflowsets.html","id":"putting-together-a-stack","dir":"Articles","previous_headings":"","what":"Putting together a stack","title":"Stacking With Workflow Sets","text":"Building stacked ensemble, now, takes even fewer lines individual workflows: results obtained building model stack workflow sets identical result building model stack individual workflows. make sure right trade-minimizing number members optimizing performance, can use:  results good enough, blend_predictions() called different values penalty. , blend_predictions() picks penalty parameter numerically optimal results. see top results:  identify model configurations assigned stacking coefficients, can make use collect_parameters() function: object now ready predict new data! Juxtaposing predictions true data:  Looks like predictions decent! stacks predictions perform, though, compared members’ predictions? can use type = \"members\" argument generate predictions ensemble members. Now, evaluating root mean squared error model: can see, stacked ensemble outperforms member models, though closely followed one members. Voila! ’ve now made use stacks workflowsets packages predict red-eyed tree frog embryo hatching using stacked ensemble!","code":"tree_frogs_model_st <-    # initialize the stack   stacks() %>%   # add candidate members   add_candidates(wf_set_trained) %>%   # determine how to combine their predictions   blend_predictions() %>%   # fit the candidates with nonzero stacking coefficients   fit_members()  tree_frogs_model_st #> # A tibble: 4 × 3 #>   member           type             weight #>   <chr>            <chr>             <dbl> #> 1 rec3_svm_1_06    svm_rbf          0.779  #> 2 rec2_lin_reg_1_1 linear_reg       0.275  #> 3 rec1_knn_1_08    nearest_neighbor 0.190  #> 4 rec1_knn_1_07    nearest_neighbor 0.0205 autoplot(tree_frogs_model_st) autoplot(tree_frogs_model_st, type = \"weights\") collect_parameters(tree_frogs_model_st, \"rec3_svm\") #> # A tibble: 10 × 4 #>    member            cost    sigma  coef #>    <chr>            <dbl>    <dbl> <dbl> #>  1 rec3_svm_1_01  0.185   1.17e- 4 0     #>  2 rec3_svm_1_02  1.23    2.39e- 7 0     #>  3 rec3_svm_1_03  0.0195  3.37e- 5 0     #>  4 rec3_svm_1_04 15.2     4.78e- 8 0     #>  5 rec3_svm_1_05  2.97    1.30e- 6 0     #>  6 rec3_svm_1_06  0.134   5.62e- 2 0.779 #>  7 rec3_svm_1_07  0.00163 3.01e- 3 0     #>  8 rec3_svm_1_08  4.08    2.98e- 1 0     #>  9 rec3_svm_1_09  0.00429 3.45e-10 0     #> 10 rec3_svm_1_10  0.0280  1.54e- 9 0 tree_frogs_test <-    tree_frogs_test %>%   bind_cols(predict(tree_frogs_model_st, .)) ggplot(tree_frogs_test) +   aes(x = latency,        y = .pred) +   geom_point() +    coord_obs_pred() member_preds <-    tree_frogs_test %>%   select(latency) %>%   bind_cols(predict(tree_frogs_model_st, tree_frogs_test, members = TRUE)) map(member_preds, rmse_vec, truth = member_preds$latency) %>%   as_tibble() #> # A tibble: 1 × 6 #>   latency .pred rec1_knn_1_07 rec1_knn_1_08 rec2_lin_reg_1_1 rec3_svm_1_06 #>     <dbl> <dbl>         <dbl>         <dbl>            <dbl>         <dbl> #> 1       0  53.9          56.2          56.2             55.5          57.3"},{"path":"https://stacks.tidymodels.org/dev/authors.html","id":null,"dir":"","previous_headings":"","what":"Authors","title":"Authors and Citation","text":"Simon Couch. Author, maintainer. Max Kuhn. Author. . Copyright holder, funder.","code":""},{"path":"https://stacks.tidymodels.org/dev/authors.html","id":"citation","dir":"","previous_headings":"","what":"Citation","title":"Authors and Citation","text":"Couch S, Kuhn M (2023). stacks: Tidy Model Stacking. R package version 1.0.3.9000, https://github.com/tidymodels/stacks, https://stacks.tidymodels.org/.","code":"@Manual{,   title = {stacks: Tidy Model Stacking},   author = {Simon Couch and Max Kuhn},   year = {2023},   note = {R package version 1.0.3.9000, https://github.com/tidymodels/stacks},   url = {https://stacks.tidymodels.org/}, }"},{"path":"https://stacks.tidymodels.org/dev/index.html","id":"stacks---tidy-model-stacking-","dir":"","previous_headings":"","what":"stacks - tidy model stacking","title":"Tidy Model Stacking","text":"stacks R package model stacking aligns tidymodels. Model stacking ensembling method takes outputs many models combines generate new model—referred ensemble package—generates predictions informed members. process goes something like : Define candidate ensemble members using functionality rsample, parsnip, workflows, recipes, tune Initialize data_stack object stacks() Iteratively add candidate ensemble members data_stack add_candidates() Evaluate combine predictions blend_predictions() Fit candidate ensemble members non-zero stacking coefficients fit_members() Predict new data predict() can install package following code: Install development version : stacks generalized respect : Model type: model type implemented parsnip extension packages fair game add stacks model stack. ’s table many implemented model types tidymodels core, link article implementing model classes well. Cross-validation scheme: resampling algorithm implemented rsample extension packages fair game resampling data use training model stack. Error metric: metric function implemented yardstick extension packages fair game evaluating model stacks members. package provides infrastructure creating metric functions well! stacks uses regularized linear model combine predictions ensemble members, though model type one many possible learning algorithms used fit stacked ensemble model. implementations additional ensemble learning algorithms, check h2o SuperLearner. Rather diving right implementation, ’ll focus pieces fit together, conceptually, building ensemble stacks. See basics vignette example API action!","code":"install.packages(\"stacks\") # install.packages(\"pak\") pak::pak(\"tidymodels/stacks\")"},{"path":"https://stacks.tidymodels.org/dev/index.html","id":"a-grammar","dir":"","previous_headings":"","what":"a grammar","title":"Tidy Model Stacking","text":"highest level, ensembles formed model definitions. package, model definitions instance minimal workflow, containing model specification (defined parsnip package) , optionally, preprocessor (defined recipes package). Model definitions specify form candidate ensemble members. diagram representing “model definitions,” specify form candidate ensemble members. Three colored boxes represent three different model types; K-nearest neighbors model (salmon), linear regression model (yellow), support vector machine model (green). used ensemble, model definitions must share resample. rsample rset object, paired model definitions, can used generate tuning/fitting results objects candidate ensemble members tune. diagram representing “candidate members” generated model definition. Four salmon-colored boxes labeled “KNN” represent K-nearest neighbors models trained resamples differing hyperparameters. Similarly, linear regression model generates one candidate member, support vector machine model generates six. Candidate members first come together data_stack object add_candidates() function. Principally, objects just tibbles, first column gives true outcome assessment set (portion training set used model validation), remaining columns give predictions candidate ensemble member. (outcome numeric, ’s one column per candidate ensemble member. Classification requires many columns per candidate levels outcome variable.) also bring along extra attributes keep track model definitions. diagram representing “data stack,” specific kind data frame. Colored “columns” depict, white, true value outcome variable validation set, followed four columns (salmon) representing predictions K-nearest neighbors model, one column (tan) representing linear regression model, six (green) representing support vector machine model. , data stack can evaluated using blend_predictions() determine best combine outputs candidate members. stacking literature, process commonly called metalearning. outputs member likely highly correlated. Thus, depending degree regularization choose, coefficients inputs (possibly) many members zero —predictions influence final output, terms thus thrown . diagram representing “stacking coefficients,” coefficients linear model combining candidate member predictions generate ensemble’s ultimate prediction. Boxes candidate members placed besides , filled color coefficient associated candidate member nonzero. stacking coefficients determine candidate ensemble members become ensemble members. Candidates non-zero stacking coefficients fitted whole training set, altogether making model_stack object. diagram representing “model stack” class, collates stacking coefficients members (candidate members nonzero stacking coefficients trained full training set). representation stacking coefficients , members (shown next associated stacking coefficients) colored-pentagons. Model stacks list subclass. model stack object, outputted fit_members(), ready predict new data! trained ensemble members often referred base models stacking literature. full visual outline steps can found . API package closely mirrors ideas. See basics vignette example grammar implemented!","code":""},{"path":"https://stacks.tidymodels.org/dev/index.html","id":"contributing","dir":"","previous_headings":"","what":"contributing","title":"Tidy Model Stacking","text":"project released Contributor Code Conduct. contributing project, agree abide terms. questions discussions tidymodels packages, modeling, machine learning, please post Posit Community. think encountered bug, please submit issue. Either way, learn create share reprex (minimal, reproducible example), clearly communicate code. Check details contributing guidelines tidymodels packages get help. stacks package, test objects take long build every commit. contribution changes structure data_stack model_stacks objects, please regenerate test objects running scripts man-roxygen/example_models.Rmd, including chunk options eval = FALSE.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/add_candidates.html","id":null,"dir":"Reference","previous_headings":"","what":"Add model definitions to a data stack — add_candidates","title":"Add model definitions to a data stack — add_candidates","text":"add_candidates() collates assessment set predictions additional attributes supplied model definition (.e. set \"candidates\") data stack. Behind scenes, data stack objects just tibble::tbl_dfs, first column gives true response values, remaining columns give assessment set predictions candidate. regression setting, one column per ensemble member. classification settings, many columns per candidate ensemble member levels outcome variable. initialize data stack, use stacks() function. Model definitions appended data stack iteratively using several calls add_candidates(). Data stacks evaluated using blend_predictions() function.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/add_candidates.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Add model definitions to a data stack — add_candidates","text":"","code":"add_candidates(   data_stack,   candidates,   name = deparse(substitute(candidates)),   ... )"},{"path":"https://stacks.tidymodels.org/dev/reference/add_candidates.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Add model definitions to a data stack — add_candidates","text":"data_stack data_stack object. candidates (set ) model definition(s) defining candidate model stack members. inherit tune_results workflow_set. tune_results: object outputted tune::tune_grid(), tune::tune_bayes(), tune::fit_resamples(). workflow_set: object outputted workflowsets::workflow_map(). approach allows supplying multiple sets candidate members one call add_candidates. See \"Stacking Workflow Sets\" article package website example code! Regardless, results must fitted control settings save_pred = TRUE, save_workflow = TRUE—see control_stack_grid(), control_stack_bayes(), control_stack_resamples() documentation helper functions. name label model definition---defaults name candidates object. Ignored candidates inherits workflow_set. ... Additional arguments. Currently ignored.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/add_candidates.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Add model definitions to a data stack — add_candidates","text":"data_stack object--see stacks() details!","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/add_candidates.html","id":"example-data","dir":"Reference","previous_headings":"","what":"Example Data","title":"Add model definitions to a data stack — add_candidates","text":"package provides resampling objects datasets use examples vignettes derived study 1212 red-eyed tree frog embryos! Red-eyed tree frog (RETF) embryos can hatch earlier normal 7ish days detect potential predator threat. Researchers wanted determine , , tree frog embryos able detect stimulus environment. , subjected embryos varying developmental stages \"predator stimulus\" jiggling embryos blunt probe. Beforehand, though embryos treated gentamicin, compound knocks lateral line (sensory organ.) Researcher Julie Jung crew found factors inform whether embryo hatches prematurely ! Note data included stacks package necessarily representative unbiased subset complete dataset, demonstrative purposes. reg_folds class_folds rset cross-fold validation objects rsample, splitting training data regression classification model objects, respectively. tree_frogs_reg_test tree_frogs_class_test analogous testing sets. reg_res_lr, reg_res_svm, reg_res_sp contain regression tuning results linear regression, support vector machine, spline model, respectively, fitting latency (.e. long embryos took hatch response jiggle) tree_frogs data, using variables predictors. Note data underlying models filtered include data embryos hatched response stimulus. class_res_rf class_res_nn contain multiclass classification tuning results random forest neural network classification model, respectively, fitting reflex (measure ear function) data using variables predictors. log_res_rf log_res_nn, contain binary classification tuning results random forest neural network classification model, respectively, fitting hatched (whether embryos hatched response stimulus) using variables predictors. See ?example_data learn objects, well browse source code generated .","code":""},{"path":[]},{"path":"https://stacks.tidymodels.org/dev/reference/add_candidates.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Add model definitions to a data stack — add_candidates","text":"","code":"# see the \"Example Data\" section above for # clarification on the objects used in these examples!  # put together a data stack using # tuning results for regression models reg_st <-    stacks() %>%   add_candidates(reg_res_lr) %>%   add_candidates(reg_res_svm) %>%   add_candidates(reg_res_sp)    reg_st #> # A data stack with 3 model definitions and 16 candidate members: #> #   reg_res_lr: 1 model configuration #> #   reg_res_svm: 5 model configurations #> #   reg_res_sp: 10 model configurations #> # Outcome: latency (numeric)    # do the same with multinomial classification models class_st <-   stacks() %>%   add_candidates(class_res_nn) %>%   add_candidates(class_res_rf) #> Warning: Predictions from 1 candidate were identical to those from existing #> candidates and were removed from the data stack.    class_st #> # A data stack with 2 model definitions and 10.6666666666667 candidate members: #> #   class_res_nn: 1 model configuration #> #   class_res_rf: 9.66666666666667 model configurations #> # Outcome: reflex (factor)    # ...or binomial classification models log_st <-   stacks() %>%   add_candidates(log_res_nn) %>%   add_candidates(log_res_rf)    log_st #> # A data stack with 2 model definitions and 11 candidate members: #> #   log_res_nn: 1 model configuration #> #   log_res_rf: 10 model configurations #> # Outcome: hatched (factor)    # use custom names for each model: log_st2 <-   stacks() %>%   add_candidates(log_res_nn, name = \"neural_network\") %>%   add_candidates(log_res_rf, name = \"random_forest\")    log_st2 #> # A data stack with 2 model definitions and 11 candidate members: #> #   neural_network: 1 model configuration #> #   random_forest: 10 model configurations #> # Outcome: hatched (factor)    # these objects would likely then be # passed to blend_predictions(): log_st2 %>% blend_predictions() #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Out of 11 possible candidate members, the ensemble retained 2. #> Penalty: 0.01. #> Mixture: 1. #>  #> The 2 highest weighted member classes are: #> # A tibble: 2 × 3 #>   member                      type        weight #>   <chr>                       <chr>        <dbl> #> 1 .pred_no_neural_network_1_1 mlp           7.08 #> 2 .pred_no_random_forest_1_05 rand_forest   3.10 #>  #> Members have not yet been fitted with `fit_members()`."},{"path":"https://stacks.tidymodels.org/dev/reference/augment.model_stack.html","id":null,"dir":"Reference","previous_headings":"","what":"Augment a model stack — augment.model_stack","title":"Augment a model stack — augment.model_stack","text":"Augment model stack","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/augment.model_stack.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Augment a model stack — augment.model_stack","text":"","code":"# S3 method for model_stack augment(x, new_data, ...)"},{"path":"https://stacks.tidymodels.org/dev/reference/augment.model_stack.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Augment a model stack — augment.model_stack","text":"x fitted model stack; see fit_members(). new_data rectangular data object, data frame. ... Additional arguments passed predict.model_stack. particular, see type members.","code":""},{"path":[]},{"path":"https://stacks.tidymodels.org/dev/reference/autoplot.linear_stack.html","id":null,"dir":"Reference","previous_headings":"","what":"Plot results of a stacked ensemble model. — autoplot.linear_stack","title":"Plot results of a stacked ensemble model. — autoplot.linear_stack","text":"Plot results stacked ensemble model.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/autoplot.linear_stack.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Plot results of a stacked ensemble model. — autoplot.linear_stack","text":"","code":"# S3 method for linear_stack autoplot(object, type = \"performance\", n = Inf, ...)"},{"path":"https://stacks.tidymodels.org/dev/reference/autoplot.linear_stack.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Plot results of a stacked ensemble model. — autoplot.linear_stack","text":"object linear_stack object outputted blend_predictions() fit_members(). type single character string plot type values \"performance\", \"members\", \"weights\". n integer many members weights plot type = \"weights\". multi-class data, total number weights across classes; otherwise equal number members. ... currently used.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/autoplot.linear_stack.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Plot results of a stacked ensemble model. — autoplot.linear_stack","text":"ggplot object.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/autoplot.linear_stack.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Plot results of a stacked ensemble model. — autoplot.linear_stack","text":"\"performance\" plot shows relationship lasso penalty resampled performance metrics. latter includes average number ensemble members. plot can helpful understanding penalty values reasonable. \"members\" plot shows relationship average number ensemble members performance metrics. point different penalty value. Neither \"performance\" \"members\" plots helpful single penalty used. \"weights\" plot shows blending weights top ensemble members. results final penalty value used fit ensemble.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/axe_model_stack.html","id":null,"dir":"Reference","previous_headings":"","what":"Axing a model_stack. — axe_model_stack","title":"Axing a model_stack. — axe_model_stack","text":"Axing model_stack. Remove call. Remove controls used training. Remove training data. Remove environments. Remove fitted values.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/axe_model_stack.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Axing a model_stack. — axe_model_stack","text":"","code":"# S3 method for model_stack axe_call(x, verbose = FALSE, ...)  # S3 method for model_stack axe_ctrl(x, verbose = FALSE, ...)  # S3 method for model_stack axe_data(x, verbose = FALSE, ...)  # S3 method for model_stack axe_env(x, verbose = FALSE, ...)  # S3 method for model_stack axe_fitted(x, verbose = FALSE, ...)"},{"path":"https://stacks.tidymodels.org/dev/reference/axe_model_stack.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Axing a model_stack. — axe_model_stack","text":"x model object verbose Print information time axe method executed. Notes much memory released functions disabled. Default FALSE. ... Additional arguments. Currently ignored.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/axe_model_stack.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Axing a model_stack. — axe_model_stack","text":"Axed model_stack object.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/axe_model_stack.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Axing a model_stack. — axe_model_stack","text":"","code":"# build a regression model stack st <-   stacks() %>%   add_candidates(reg_res_lr) %>%   add_candidates(reg_res_sp) %>%   blend_predictions() %>%   fit_members()    # remove any of the \"butcherable\" # elements individually axe_call(st) #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Print methods for butchered model stacks are disabled. axe_ctrl(st) #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Print methods for butchered model stacks are disabled. axe_data(st) #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Print methods for butchered model stacks are disabled. axe_fitted(st) #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Print methods for butchered model stacks are disabled. axe_env(st) #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Print methods for butchered model stacks are disabled.  # or do it all at once! butchered_st <- butcher(st, verbose = TRUE) #> ✔ Memory released: 155.86 kB #> ✖ Disabled: `print()` and `summary()`  format(object.size(st)) #> [1] \"2465848 bytes\" format(object.size(butchered_st)) #> [1] \"1565608 bytes\""},{"path":"https://stacks.tidymodels.org/dev/reference/blend_predictions.html","id":null,"dir":"Reference","previous_headings":"","what":"Determine stacking coefficients from a data stack — blend_predictions","title":"Determine stacking coefficients from a data stack — blend_predictions","text":"Evaluates data stack fitting regularized model assessment predictions candidate member predict true outcome. process determines \"stacking coefficients\" model stack. stacking coefficients used weight predictions candidate (represented unique column data stack), given betas LASSO model fitting true outcome predictions given remaining columns data stack. Candidates non-zero stacking coefficients model stack members, need trained full training set (rather just assessment set) fit_members(). function typically used number calls add_candidates().","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/blend_predictions.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Determine stacking coefficients from a data stack — blend_predictions","text":"","code":"blend_predictions(   data_stack,   penalty = 10^(-6:-1),   mixture = 1,   non_negative = TRUE,   metric = NULL,   control = tune::control_grid(),   times = 25,   ... )"},{"path":"https://stacks.tidymodels.org/dev/reference/blend_predictions.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Determine stacking coefficients from a data stack — blend_predictions","text":"data_stack data_stack object penalty numeric vector proposed values total amount regularization used member weighting. Higher penalties generally result fewer members included resulting model stack, vice versa. package tune grid formed cross product penalty mixture arguments. mixture number zero one (inclusive) giving proportion L1 regularization (.e. lasso) model. mixture = 1 indicates pure lasso model, mixture = 0 indicates ridge regression, values (0, 1) indicate elastic net. package tune grid formed cross product penalty mixture arguments. non_negative logical giving whether restrict stacking coefficients non-negative values. TRUE (default), 0 passed lower.limits argument glmnet::glmnet() fitting model data stack. Otherwise, -Inf. metric call yardstick::metric_set(). metric(s) use tuning lasso penalty stacking coefficients. Default values determined tune::tune_grid() outcome class. control object inheriting control_grid passed model determining stacking coefficients. See tune::control_grid() documentation details possible values. Note extract entry overwritten internally. times Number bootstrap samples tuned model determines stacking coefficients. See rsample::bootstraps() learn . ... Additional arguments. Currently ignored.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/blend_predictions.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Determine stacking coefficients from a data stack — blend_predictions","text":"model_stack object—model_stacks largely contain elements data_stacks, primary data objects shift assessment set predictions member models.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/blend_predictions.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Determine stacking coefficients from a data stack — blend_predictions","text":"Note regularized linear model one many possible learning algorithms used fit stacked ensemble model. implementations additional ensemble learning algorithms, see h2o::h2o.stackedEnsemble() SuperLearner::SuperLearner().","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/blend_predictions.html","id":"example-data","dir":"Reference","previous_headings":"","what":"Example Data","title":"Determine stacking coefficients from a data stack — blend_predictions","text":"package provides resampling objects datasets use examples vignettes derived study 1212 red-eyed tree frog embryos! Red-eyed tree frog (RETF) embryos can hatch earlier normal 7ish days detect potential predator threat. Researchers wanted determine , , tree frog embryos able detect stimulus environment. , subjected embryos varying developmental stages \"predator stimulus\" jiggling embryos blunt probe. Beforehand, though embryos treated gentamicin, compound knocks lateral line (sensory organ.) Researcher Julie Jung crew found factors inform whether embryo hatches prematurely ! Note data included stacks package necessarily representative unbiased subset complete dataset, demonstrative purposes. reg_folds class_folds rset cross-fold validation objects rsample, splitting training data regression classification model objects, respectively. tree_frogs_reg_test tree_frogs_class_test analogous testing sets. reg_res_lr, reg_res_svm, reg_res_sp contain regression tuning results linear regression, support vector machine, spline model, respectively, fitting latency (.e. long embryos took hatch response jiggle) tree_frogs data, using variables predictors. Note data underlying models filtered include data embryos hatched response stimulus. class_res_rf class_res_nn contain multiclass classification tuning results random forest neural network classification model, respectively, fitting reflex (measure ear function) data using variables predictors. log_res_rf log_res_nn, contain binary classification tuning results random forest neural network classification model, respectively, fitting hatched (whether embryos hatched response stimulus) using variables predictors. See ?example_data learn objects, well browse source code generated .","code":""},{"path":[]},{"path":"https://stacks.tidymodels.org/dev/reference/blend_predictions.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Determine stacking coefficients from a data stack — blend_predictions","text":"","code":"# see the \"Example Data\" section above for # clarification on the objects used in these examples!  # put together a data stack reg_st <-    stacks() %>%   add_candidates(reg_res_lr) %>%   add_candidates(reg_res_svm) %>%   add_candidates(reg_res_sp)    reg_st #> # A data stack with 3 model definitions and 16 candidate members: #> #   reg_res_lr: 1 model configuration #> #   reg_res_svm: 5 model configurations #> #   reg_res_sp: 10 model configurations #> # Outcome: latency (numeric)  # evaluate the data stack reg_st %>%   blend_predictions() #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Out of 16 possible candidate members, the ensemble retained 3. #> Penalty: 1e-06. #> Mixture: 1. #>  #> The 3 highest weighted members are: #> # A tibble: 3 × 3 #>   member          type       weight #>   <chr>           <chr>       <dbl> #> 1 reg_res_svm_1_3 svm_rbf    0.638  #> 2 reg_res_sp_03_1 linear_reg 0.486  #> 3 reg_res_sp_10_1 linear_reg 0.0482 #>  #> Members have not yet been fitted with `fit_members()`.  # include fewer models by proposing higher penalties reg_st %>%    blend_predictions(penalty = c(.5, 1)) #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Out of 16 possible candidate members, the ensemble retained 3. #> Penalty: 0.5. #> Mixture: 1. #>  #> The 3 highest weighted members are: #> # A tibble: 3 × 3 #>   member          type       weight #>   <chr>           <chr>       <dbl> #> 1 reg_res_svm_1_3 svm_rbf    0.629  #> 2 reg_res_sp_03_1 linear_reg 0.478  #> 3 reg_res_sp_10_1 linear_reg 0.0515 #>  #> Members have not yet been fitted with `fit_members()`.  # allow for negative stacking coefficients  # with the non_negative argument reg_st %>%    blend_predictions(non_negative = FALSE) #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Out of 16 possible candidate members, the ensemble retained 12. #> Penalty: 0.1. #> Mixture: 1. #>  #> The 10 highest weighted members are: #> # A tibble: 10 × 3 #>    member          type        weight #>    <chr>           <chr>        <dbl> #>  1 reg_res_svm_1_1 svm_rbf    -10.5   #>  2 reg_res_sp_04_1 linear_reg  -1.38  #>  3 reg_res_sp_05_1 linear_reg   1.35  #>  4 reg_res_svm_1_3 svm_rbf      1.19  #>  5 reg_res_svm_1_2 svm_rbf     -0.963 #>  6 reg_res_sp_03_1 linear_reg   0.642 #>  7 reg_res_sp_01_1 linear_reg  -0.400 #>  8 reg_res_sp_10_1 linear_reg   0.319 #>  9 reg_res_sp_06_1 linear_reg   0.193 #> 10 reg_res_lr_1_1  linear_reg   0.183 #>  #> Members have not yet been fitted with `fit_members()`.    # use a custom metric in tuning the lasso penalty library(yardstick) reg_st %>%    blend_predictions(metric = metric_set(rmse)) #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Out of 16 possible candidate members, the ensemble retained 3. #> Penalty: 1e-06. #> Mixture: 1. #>  #> The 3 highest weighted members are: #> # A tibble: 3 × 3 #>   member          type       weight #>   <chr>           <chr>       <dbl> #> 1 reg_res_svm_1_3 svm_rbf    0.638  #> 2 reg_res_sp_03_1 linear_reg 0.486  #> 3 reg_res_sp_10_1 linear_reg 0.0482 #>  #> Members have not yet been fitted with `fit_members()`.    # pass control options for stack blending reg_st %>%    blend_predictions(     control = tune::control_grid(allow_par = TRUE)   ) #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Out of 16 possible candidate members, the ensemble retained 3. #> Penalty: 1e-06. #> Mixture: 1. #>  #> The 3 highest weighted members are: #> # A tibble: 3 × 3 #>   member          type       weight #>   <chr>           <chr>       <dbl> #> 1 reg_res_svm_1_3 svm_rbf    0.638  #> 2 reg_res_sp_03_1 linear_reg 0.486  #> 3 reg_res_sp_10_1 linear_reg 0.0482 #>  #> Members have not yet been fitted with `fit_members()`.   # to speed up the stacking process for preliminary # results, bump down the `times` argument: reg_st %>%    blend_predictions(times = 5) #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Out of 16 possible candidate members, the ensemble retained 3. #> Penalty: 1e-06. #> Mixture: 1. #>  #> The 3 highest weighted members are: #> # A tibble: 3 × 3 #>   member          type       weight #>   <chr>           <chr>       <dbl> #> 1 reg_res_svm_1_3 svm_rbf    0.638  #> 2 reg_res_sp_03_1 linear_reg 0.486  #> 3 reg_res_sp_10_1 linear_reg 0.0482 #>  #> Members have not yet been fitted with `fit_members()`.    # the process looks the same with  # multinomial classification models class_st <-   stacks() %>%   add_candidates(class_res_nn) %>%   add_candidates(class_res_rf) %>%   blend_predictions() #> Warning: Predictions from 1 candidate were identical to those from existing #> candidates and were removed from the data stack. #> → A | warning: from glmnet C++ code (error code -82); Convergence for 82th lambda value not reached after maxit=100000 iterations; solutions for larger lambdas returned #> There were issues with some computations   A: x1 #> There were issues with some computations   A: x1 #>     class_st #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Out of 21 possible candidate members, the ensemble retained 5. #> Penalty: 0.1. #> Mixture: 1. #> Across the 3 classes, there are an average of 2.5 coefficients per class. #>  #> The 5 highest weighted member classes are: #> # A tibble: 5 × 4 #>   member                       type          weight class #>   <chr>                        <chr>          <dbl> <fct> #> 1 .pred_full_class_res_nn_1_1  mlp         12.0     full  #> 2 .pred_mid_class_res_rf_1_06  rand_forest  0.670   mid   #> 3 .pred_full_class_res_rf_1_05 rand_forest  0.101   full  #> 4 .pred_full_class_res_rf_1_07 rand_forest  0.00457 full  #> 5 .pred_full_class_res_rf_1_01 rand_forest  0.00219 full  #>  #> Members have not yet been fitted with `fit_members()`.  # ...or binomial classification models log_st <-   stacks() %>%   add_candidates(log_res_nn) %>%   add_candidates(log_res_rf) %>%   blend_predictions()    log_st #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Out of 11 possible candidate members, the ensemble retained 3. #> Penalty: 0.001. #> Mixture: 1. #>  #> The 3 highest weighted member classes are: #> # A tibble: 3 × 3 #>   member                   type        weight #>   <chr>                    <chr>        <dbl> #> 1 .pred_no_log_res_nn_1_1  mlp         7.39   #> 2 .pred_no_log_res_rf_1_05 rand_forest 3.42   #> 3 .pred_no_log_res_rf_1_02 rand_forest 0.0281 #>  #> Members have not yet been fitted with `fit_members()`."},{"path":"https://stacks.tidymodels.org/dev/reference/build_linear_predictor.html","id":null,"dir":"Reference","previous_headings":"","what":"Creates an R expression for a linear predictor from a data frame of terms and\ncoefficients — build_linear_predictor","title":"Creates an R expression for a linear predictor from a data frame of terms and\ncoefficients — build_linear_predictor","text":"Creates R expression linear predictor data frame terms coefficients","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/build_linear_predictor.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Creates an R expression for a linear predictor from a data frame of terms and\ncoefficients — build_linear_predictor","text":"","code":"build_linear_predictor(x, ...)  # S3 method for `_elnet` build_linear_predictor(x, ...)  # S3 method for `_lognet` build_linear_predictor(x, ...)  # S3 method for `_multnet` build_linear_predictor(x, ...)"},{"path":"https://stacks.tidymodels.org/dev/reference/build_linear_predictor.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Creates an R expression for a linear predictor from a data frame of terms and\ncoefficients — build_linear_predictor","text":"x object uses glmnet::glmnet() model numeric predictors. ... currently used.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/build_linear_predictor.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Creates an R expression for a linear predictor from a data frame of terms and\ncoefficients — build_linear_predictor","text":"R expression list R expressions, depending type model used.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/collect_parameters.html","id":null,"dir":"Reference","previous_headings":"","what":"Collect candidate parameters and stacking coefficients — collect_parameters","title":"Collect candidate parameters and stacking coefficients — collect_parameters","text":"function help situate candidates within stack. Takes data stack model stack candidate name returns tibble mapping candidate/member names hyperparameters (, model stack, stacking coefficients well).","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/collect_parameters.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Collect candidate parameters and stacking coefficients — collect_parameters","text":"","code":"collect_parameters(stack, candidates, ...)  # S3 method for default collect_parameters(stack, candidates, ...)  # S3 method for data_stack collect_parameters(stack, candidates, ...)  # S3 method for model_stack collect_parameters(stack, candidates, ...)"},{"path":"https://stacks.tidymodels.org/dev/reference/collect_parameters.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Collect candidate parameters and stacking coefficients — collect_parameters","text":"stack data_stack model_stack object. candidates name candidates collect parameters . either name argument supplied add_candidates() , supplied, name object supplied candidates argument add_candidates(). ... Additional arguments. Currently ignored.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/collect_parameters.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Collect candidate parameters and stacking coefficients — collect_parameters","text":"tibble::tbl_df information member names hyperparameters.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/collect_parameters.html","id":"example-data","dir":"Reference","previous_headings":"","what":"Example Data","title":"Collect candidate parameters and stacking coefficients — collect_parameters","text":"package provides resampling objects datasets use examples vignettes derived study 1212 red-eyed tree frog embryos! Red-eyed tree frog (RETF) embryos can hatch earlier normal 7ish days detect potential predator threat. Researchers wanted determine , , tree frog embryos able detect stimulus environment. , subjected embryos varying developmental stages \"predator stimulus\" jiggling embryos blunt probe. Beforehand, though embryos treated gentamicin, compound knocks lateral line (sensory organ.) Researcher Julie Jung crew found factors inform whether embryo hatches prematurely ! Note data included stacks package necessarily representative unbiased subset complete dataset, demonstrative purposes. reg_folds class_folds rset cross-fold validation objects rsample, splitting training data regression classification model objects, respectively. tree_frogs_reg_test tree_frogs_class_test analogous testing sets. reg_res_lr, reg_res_svm, reg_res_sp contain regression tuning results linear regression, support vector machine, spline model, respectively, fitting latency (.e. long embryos took hatch response jiggle) tree_frogs data, using variables predictors. Note data underlying models filtered include data embryos hatched response stimulus. class_res_rf class_res_nn contain multiclass classification tuning results random forest neural network classification model, respectively, fitting reflex (measure ear function) data using variables predictors. log_res_rf log_res_nn, contain binary classification tuning results random forest neural network classification model, respectively, fitting hatched (whether embryos hatched response stimulus) using variables predictors. See ?example_data learn objects, well browse source code generated .","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/collect_parameters.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Collect candidate parameters and stacking coefficients — collect_parameters","text":"","code":"# see the \"Example Data\" section above for # clarification on the objects used in these examples!  # put together a data stack using # tuning results for regression models reg_st <-    stacks() %>%   add_candidates(reg_res_lr) %>%   add_candidates(reg_res_svm) %>%   add_candidates(reg_res_sp, \"spline\")    reg_st #> # A data stack with 3 model definitions and 16 candidate members: #> #   reg_res_lr: 1 model configuration #> #   reg_res_svm: 5 model configurations #> #   spline: 10 model configurations #> # Outcome: latency (numeric)    # check out the hyperparameters for some of the candidates collect_parameters(reg_st, \"reg_res_svm\") #> # A tibble: 5 × 3 #>   member             cost     rbf_sigma #>   <chr>             <dbl>         <dbl> #> 1 reg_res_svm_1_1 3.09    0.000000465   #> 2 reg_res_svm_1_2 9.56    0.00346       #> 3 reg_res_svm_1_3 0.321   0.0256        #> 4 reg_res_svm_1_4 0.00486 0.00000251    #> 5 reg_res_svm_1_5 0.0146  0.00000000959  collect_parameters(reg_st, \"spline\") #> # A tibble: 10 × 2 #>    member        age #>    <chr>       <int> #>  1 spline_01_1    12 #>  2 spline_02_1     6 #>  3 spline_03_1    15 #>  4 spline_04_1     5 #>  5 spline_05_1     3 #>  6 spline_06_1    13 #>  7 spline_07_1     9 #>  8 spline_08_1    10 #>  9 spline_09_1     7 #> 10 spline_10_1     2  # blend the data stack to view the hyperparameters  # along with the stacking coefficients! collect_parameters(   reg_st %>% blend_predictions(),    \"spline\" ) #> # A tibble: 10 × 3 #>    member        age   coef #>    <chr>       <int>  <dbl> #>  1 spline_01_1    12 0      #>  2 spline_02_1     6 0      #>  3 spline_03_1    15 0.484  #>  4 spline_04_1     5 0      #>  5 spline_05_1     3 0      #>  6 spline_06_1    13 0      #>  7 spline_07_1     9 0      #>  8 spline_08_1    10 0      #>  9 spline_09_1     7 0      #> 10 spline_10_1     2 0.0496"},{"path":"https://stacks.tidymodels.org/dev/reference/control_stack.html","id":null,"dir":"Reference","previous_headings":"","what":"Control wrappers — control_stack_grid","title":"Control wrappers — control_stack_grid","text":"Supply light wrappers control argument tune::tune_grid(), tune::tune_bayes(), tune::fit_resamples() call return needed elements use data stack. functions return appropriate control grid ensure assessment set predictions information model specifications preprocessors, supplied resampling results object! integrate stack settings existing control settings, note functions just call appropriate tune::control_* function arguments save_pred = TRUE, save_workflow = TRUE.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/control_stack.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Control wrappers — control_stack_grid","text":"","code":"control_stack_grid()  control_stack_resamples()  control_stack_bayes()"},{"path":"https://stacks.tidymodels.org/dev/reference/control_stack.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Control wrappers — control_stack_grid","text":"tune::control_grid, tune::control_bayes, tune::control_resamples object.","code":""},{"path":[]},{"path":"https://stacks.tidymodels.org/dev/reference/example_data.html","id":null,"dir":"Reference","previous_headings":"","what":"Example Objects — example_data","title":"Example Objects — example_data","text":"stacks provides resampling objects datasets use examples vignettes derived study 1212 red-eyed tree frog embryos!","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/example_data.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Example Objects — example_data","text":"","code":"reg_res_svm  reg_res_sp  reg_res_lr  reg_folds  class_res_nn  class_res_rf  class_folds  log_res_nn  log_res_rf"},{"path":"https://stacks.tidymodels.org/dev/reference/example_data.html","id":"format","dir":"Reference","previous_headings":"","what":"Format","title":"Example Objects — example_data","text":"object class tune_results (inherits tbl_df, tbl, data.frame) 5 rows 5 columns. object class tune_results (inherits tbl_df, tbl, data.frame) 5 rows 5 columns. object class resample_results (inherits tune_results, tbl_df, tbl, data.frame) 5 rows 5 columns. object class vfold_cv (inherits rset, tbl_df, tbl, data.frame) 5 rows 2 columns. object class resample_results (inherits tune_results, tbl_df, tbl, data.frame) 5 rows 5 columns. object class tune_results (inherits tbl_df, tbl, data.frame) 5 rows 5 columns. object class vfold_cv (inherits rset, tbl_df, tbl, data.frame) 5 rows 2 columns. object class resample_results (inherits tune_results, tbl_df, tbl, data.frame) 5 rows 5 columns. object class tune_results (inherits tbl_df, tbl, data.frame) 5 rows 5 columns.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/example_data.html","id":"source","dir":"Reference","previous_headings":"","what":"Source","title":"Example Objects — example_data","text":"Julie Jung et al. (2020) Multimodal mechanosensing enables treefrog embryos escape egg-predators. doi:10.1242/jeb.236141","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/example_data.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Example Objects — example_data","text":"Red-eyed tree frog (RETF) embryos can hatch earlier normal 7ish days detect potential predator threat. Researchers wanted determine , , tree frog embryos able detect stimulus environment. , subjected embryos varying developmental stages \"predator stimulus\" jiggling embryos blunt probe. Beforehand, though embryos treated gentamicin, compound knocks lateral line (sensory organ.) Researcher Julie Jung crew found factors inform whether embryo hatches prematurely ! Note data included stacks package necessarily representative unbiased subset complete dataset, demonstrative purposes. reg_folds class_folds rset cross-fold validation objects rsample, splitting training data regression classification model objects, respectively. tree_frogs_reg_test tree_frogs_class_test analogous testing sets. reg_res_lr, reg_res_svm, reg_res_sp contain regression tuning results linear regression, support vector machine, spline model, respectively, fitting latency (.e. long embryos took hatch response jiggle) tree_frogs data, using variables predictors. Note data underlying models filtered include data embryos hatched response stimulus. class_res_rf class_res_nn contain multiclass classification tuning results random forest neural network classification model, respectively, fitting reflex (measure ear function) data using variables predictors. log_res_rf log_res_nn, contain binary classification tuning results random forest neural network classification model, respectively, fitting hatched (whether embryos hatched response stimulus) using variables predictors. source code generating objects given .","code":"# setup: packages, data, resample, basic recipe ------------------------ library(stacks) library(tune) library(rsample) library(parsnip) library(workflows) library(recipes) library(yardstick) library(workflowsets)  set.seed(1)  ctrl_grid <-    tune::control_grid(     save_pred = TRUE,     save_workflow = TRUE   )  ctrl_res <-    tune::control_resamples(     save_pred = TRUE,     save_workflow = TRUE   )  # for regression, predict latency to hatch (excluding NAs) tree_frogs_reg <-    tree_frogs %>%    filter(!is.na(latency)) %>%   select(-clutch, -hatched)  set.seed(1) tree_frogs_reg_split <- rsample::initial_split(tree_frogs_reg)  set.seed(1) tree_frogs_reg_train <- rsample::training(tree_frogs_reg_split)  set.seed(1) tree_frogs_reg_test  <- rsample::testing(tree_frogs_reg_split)  set.seed(1) reg_folds <- rsample::vfold_cv(tree_frogs_reg_train, v = 5)  tree_frogs_reg_rec <-    recipes::recipe(latency ~ ., data = tree_frogs_reg_train) %>%   recipes::step_dummy(recipes::all_nominal()) %>%   recipes::step_zv(recipes::all_predictors())  metric <- yardstick::metric_set(yardstick::rmse)  # linear regression --------------------------------------- lin_reg_spec <-   parsnip::linear_reg() %>%   parsnip::set_engine(\"lm\")  reg_wf_lr <-    workflows::workflow() %>%   workflows::add_model(lin_reg_spec) %>%   workflows::add_recipe(tree_frogs_reg_rec)  set.seed(1) reg_res_lr <-    tune::fit_resamples(     object = reg_wf_lr,     resamples = reg_folds,     metrics = metric,     control = ctrl_res   )  # SVM regression ---------------------------------- svm_spec <-    parsnip::svm_rbf(     cost = tune::tune(),      rbf_sigma = tune::tune()   ) %>%   parsnip::set_engine(\"kernlab\") %>%   parsnip::set_mode(\"regression\")  reg_wf_svm <-    workflows::workflow() %>%   workflows::add_model(svm_spec) %>%   workflows::add_recipe(tree_frogs_reg_rec)  set.seed(1) reg_res_svm <-    tune::tune_grid(     object = reg_wf_svm,     resamples = reg_folds,      grid = 5,     control = ctrl_grid   )  # spline regression --------------------------------------- spline_rec <-    tree_frogs_reg_rec %>%   recipes::step_ns(age, deg_free = tune::tune(\"age\"))  reg_wf_sp <-    workflows::workflow() %>%   workflows::add_model(lin_reg_spec) %>%   workflows::add_recipe(spline_rec)  set.seed(1) reg_res_sp <-    tune::tune_grid(     object = reg_wf_sp,     resamples = reg_folds,     metrics = metric,     control = ctrl_grid   )  # classification - preliminaries ----------------------------------- tree_frogs_class <-    tree_frogs %>%   dplyr::select(-c(clutch, latency))  set.seed(1) tree_frogs_class_split <- rsample::initial_split(tree_frogs_class)  set.seed(1) tree_frogs_class_train <- rsample::training(tree_frogs_class_split)  set.seed(1) tree_frogs_class_test  <- rsample::testing(tree_frogs_class_split)  set.seed(1) class_folds <- rsample::vfold_cv(tree_frogs_class_train, v = 5)  tree_frogs_class_rec <-    recipes::recipe(reflex ~ ., data = tree_frogs_class_train) %>%   recipes::step_dummy(recipes::all_nominal(), -reflex) %>%   recipes::step_zv(recipes::all_predictors()) %>%   recipes::step_normalize(recipes::all_numeric())  # random forest classification -------------------------------------- rand_forest_spec <-    parsnip::rand_forest(     mtry = tune::tune(),     trees = 500,     min_n = tune::tune()   ) %>%   parsnip::set_mode(\"classification\") %>%   parsnip::set_engine(\"ranger\")  class_wf_rf <-   workflows::workflow() %>%   workflows::add_recipe(tree_frogs_class_rec) %>%   workflows::add_model(rand_forest_spec)  set.seed(1) class_res_rf <-    tune::tune_grid(     object = class_wf_rf,      resamples = class_folds,      grid = 10,     control = ctrl_grid   )  # neural network classification ------------------------------------- nnet_spec <-   mlp(hidden_units = 5, penalty = 0.01, epochs = 100) %>%   set_mode(\"classification\") %>%   set_engine(\"nnet\")  class_wf_nn <-    workflows::workflow() %>%   workflows::add_recipe(tree_frogs_class_rec) %>%   workflows::add_model(nnet_spec)  set.seed(1) class_res_nn <-   tune::fit_resamples(     object = class_wf_nn,      resamples = class_folds,      control = ctrl_res   )  # binary classification -------------------------------- tree_frogs_2_class_rec <-    recipes::recipe(hatched ~ ., data = tree_frogs_class_train) %>%   recipes::step_dummy(recipes::all_nominal(), -hatched) %>%   recipes::step_zv(recipes::all_predictors()) %>%   recipes::step_normalize(recipes::all_numeric())  set.seed(1) rand_forest_spec_2 <-    parsnip::rand_forest(     mtry = tune(),     trees = 500,     min_n = tune()   ) %>%   parsnip::set_mode(\"classification\") %>%   parsnip::set_engine(\"ranger\")  log_wf_rf <-   workflows::workflow() %>%   workflows::add_recipe(tree_frogs_2_class_rec) %>%   workflows::add_model(rand_forest_spec_2)  set.seed(1) log_res_rf <-    tune::tune_grid(     object = log_wf_rf,      resamples = class_folds,      grid = 10,     control = ctrl_grid   )  nnet_spec_2 <-   parsnip::mlp(epochs = 100, hidden_units = 5, penalty = 0.1) %>%   parsnip::set_mode(\"classification\") %>%   parsnip::set_engine(\"nnet\", verbose = 0)  log_wf_nn <-    workflows::workflow() %>%   workflows::add_recipe(tree_frogs_2_class_rec) %>%   workflows::add_model(nnet_spec_2)  set.seed(1) log_res_nn <-   tune::fit_resamples(     object = log_wf_nn,      resamples = class_folds,      control = ctrl_res   )"},{"path":"https://stacks.tidymodels.org/dev/reference/fit_members.html","id":null,"dir":"Reference","previous_headings":"","what":"Fit model stack members with non-zero stacking coefficients — fit_members","title":"Fit model stack members with non-zero stacking coefficients — fit_members","text":"evaluating data stack blend_predictions(), number candidates nonzero stacking coefficients. candidates referred \"members.\" Since members' predictions ultimately inform model stack's predictions, members trained full training set using fit_members().","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/fit_members.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Fit model stack members with non-zero stacking coefficients — fit_members","text":"","code":"fit_members(model_stack, ...)"},{"path":"https://stacks.tidymodels.org/dev/reference/fit_members.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Fit model stack members with non-zero stacking coefficients — fit_members","text":"model_stack model_stack object outputted blend_predictions(). ... Additional arguments. Currently ignored.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/fit_members.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Fit model stack members with non-zero stacking coefficients — fit_members","text":"model_stack object subclass linear_stack---fitted model contains necessary components predict new data.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/fit_members.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Fit model stack members with non-zero stacking coefficients — fit_members","text":"fit members parallel, please register parallel backend function. See documentation foreach::foreach() examples.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/fit_members.html","id":"example-data","dir":"Reference","previous_headings":"","what":"Example Data","title":"Fit model stack members with non-zero stacking coefficients — fit_members","text":"package provides resampling objects datasets use examples vignettes derived study 1212 red-eyed tree frog embryos! Red-eyed tree frog (RETF) embryos can hatch earlier normal 7ish days detect potential predator threat. Researchers wanted determine , , tree frog embryos able detect stimulus environment. , subjected embryos varying developmental stages \"predator stimulus\" jiggling embryos blunt probe. Beforehand, though embryos treated gentamicin, compound knocks lateral line (sensory organ.) Researcher Julie Jung crew found factors inform whether embryo hatches prematurely ! Note data included stacks package necessarily representative unbiased subset complete dataset, demonstrative purposes. reg_folds class_folds rset cross-fold validation objects rsample, splitting training data regression classification model objects, respectively. tree_frogs_reg_test tree_frogs_class_test analogous testing sets. reg_res_lr, reg_res_svm, reg_res_sp contain regression tuning results linear regression, support vector machine, spline model, respectively, fitting latency (.e. long embryos took hatch response jiggle) tree_frogs data, using variables predictors. Note data underlying models filtered include data embryos hatched response stimulus. class_res_rf class_res_nn contain multiclass classification tuning results random forest neural network classification model, respectively, fitting reflex (measure ear function) data using variables predictors. log_res_rf log_res_nn, contain binary classification tuning results random forest neural network classification model, respectively, fitting hatched (whether embryos hatched response stimulus) using variables predictors. See ?example_data learn objects, well browse source code generated .","code":""},{"path":[]},{"path":"https://stacks.tidymodels.org/dev/reference/fit_members.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Fit model stack members with non-zero stacking coefficients — fit_members","text":"","code":"# see the \"Example Data\" section above for # clarification on the objects used in these examples!  # put together a data stack reg_st <-    stacks() %>%   add_candidates(reg_res_lr) %>%   add_candidates(reg_res_svm) %>%   add_candidates(reg_res_sp)    reg_st #> # A data stack with 3 model definitions and 16 candidate members: #> #   reg_res_lr: 1 model configuration #> #   reg_res_svm: 5 model configurations #> #   reg_res_sp: 10 model configurations #> # Outcome: latency (numeric)  # evaluate the data stack and fit the member models reg_st %>%   blend_predictions() %>%   fit_members() #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Out of 16 possible candidate members, the ensemble retained 3. #> Penalty: 1e-06. #> Mixture: 1. #>  #> The 3 highest weighted members are: #> # A tibble: 3 × 3 #>   member          type       weight #>   <chr>           <chr>       <dbl> #> 1 reg_res_svm_1_3 svm_rbf    0.638  #> 2 reg_res_sp_03_1 linear_reg 0.486  #> 3 reg_res_sp_10_1 linear_reg 0.0482    reg_st #> # A data stack with 3 model definitions and 16 candidate members: #> #   reg_res_lr: 1 model configuration #> #   reg_res_svm: 5 model configurations #> #   reg_res_sp: 10 model configurations #> # Outcome: latency (numeric)    # do the same with multinomial classification models class_st <-   stacks() %>%   add_candidates(class_res_nn) %>%   add_candidates(class_res_rf) %>%   blend_predictions() %>%   fit_members() #> Warning: Predictions from 1 candidate were identical to those from existing #> candidates and were removed from the data stack.    class_st #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Out of 21 possible candidate members, the ensemble retained 8. #> Penalty: 0.01. #> Mixture: 1. #> Across the 3 classes, there are an average of 4 coefficients per class. #>  #> The 8 highest weighted member classes are: #> # A tibble: 8 × 4 #>   member                       type           weight class #>   <chr>                        <chr>           <dbl> <fct> #> 1 .pred_full_class_res_nn_1_1  mlp         23.3      full  #> 2 .pred_mid_class_res_nn_1_1   mlp          1.89     mid   #> 3 .pred_mid_class_res_rf_1_06  rand_forest  1.71     mid   #> 4 .pred_mid_class_res_rf_1_10  rand_forest  1.17     mid   #> 5 .pred_full_class_res_rf_1_03 rand_forest  0.407    full  #> 6 .pred_full_class_res_rf_1_05 rand_forest  0.222    full  #> 7 .pred_full_class_res_rf_1_01 rand_forest  0.00160  full  #> 8 .pred_full_class_res_rf_1_02 rand_forest  0.000322 full     # ...or binomial classification models log_st <-   stacks() %>%   add_candidates(log_res_nn) %>%   add_candidates(log_res_rf) %>%   blend_predictions() %>%   fit_members()    log_st #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Out of 11 possible candidate members, the ensemble retained 3. #> Penalty: 1e-04. #> Mixture: 1. #>  #> The 3 highest weighted member classes are: #> # A tibble: 3 × 3 #>   member                   type        weight #>   <chr>                    <chr>        <dbl> #> 1 .pred_no_log_res_nn_1_1  mlp         7.41   #> 2 .pred_no_log_res_rf_1_05 rand_forest 3.44   #> 3 .pred_no_log_res_rf_1_02 rand_forest 0.0638"},{"path":"https://stacks.tidymodels.org/dev/reference/get_expressions.html","id":null,"dir":"Reference","previous_headings":"","what":"Obtain prediction equations for all possible values of type — get_expressions","title":"Obtain prediction equations for all possible values of type — get_expressions","text":"Obtain prediction equations possible values type","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/get_expressions.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Obtain prediction equations for all possible values of type — get_expressions","text":"","code":"get_expressions(x, ...)  # S3 method for `_multnet` get_expressions(x, ...)  # S3 method for `_lognet` get_expressions(x, ...)  # S3 method for `_elnet` get_expressions(x, ...)"},{"path":"https://stacks.tidymodels.org/dev/reference/get_expressions.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Obtain prediction equations for all possible values of type — get_expressions","text":"x parsnip model glmnet engine. ... used","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/get_expressions.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Obtain prediction equations for all possible values of type — get_expressions","text":"named list prediction equations possibel type.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/predict.data_stack.html","id":null,"dir":"Reference","previous_headings":"","what":"Predicting with a model stack — predict.data_stack","title":"Predicting with a model stack — predict.data_stack","text":"data stack must evaluated blend_predictions() member models fitted fit_members() predict new data.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/predict.data_stack.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Predicting with a model stack — predict.data_stack","text":"","code":"# S3 method for data_stack predict(object, ...)"},{"path":"https://stacks.tidymodels.org/dev/reference/predict.data_stack.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Predicting with a model stack — predict.data_stack","text":"object data stack. ... Additional arguments. Currently ignored.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/predict.model_stack.html","id":null,"dir":"Reference","previous_headings":"","what":"Predicting with a model stack — predict.model_stack","title":"Predicting with a model stack — predict.model_stack","text":"Apply model stack create different types predictions.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/predict.model_stack.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Predicting with a model stack — predict.model_stack","text":"","code":"# S3 method for model_stack predict(object, new_data, type = NULL, members = FALSE, opts = list(), ...)"},{"path":"https://stacks.tidymodels.org/dev/reference/predict.model_stack.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Predicting with a model stack — predict.model_stack","text":"object model stack fitted members outputted fit_members(). new_data rectangular data object, data frame. type Format returned predicted values—one \"numeric\", \"class\", \"prob\". NULL, predict() choose appropriate value based model's mode. members Logical. Whether additionally return predictions ensemble members. opts list optional arguments underlying predict function passed parsnip::predict.model_fit member. ... Additional arguments. Currently ignored.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/predict.model_stack.html","id":"example-data","dir":"Reference","previous_headings":"","what":"Example Data","title":"Predicting with a model stack — predict.model_stack","text":"package provides resampling objects datasets use examples vignettes derived study 1212 red-eyed tree frog embryos! Red-eyed tree frog (RETF) embryos can hatch earlier normal 7ish days detect potential predator threat. Researchers wanted determine , , tree frog embryos able detect stimulus environment. , subjected embryos varying developmental stages \"predator stimulus\" jiggling embryos blunt probe. Beforehand, though embryos treated gentamicin, compound knocks lateral line (sensory organ.) Researcher Julie Jung crew found factors inform whether embryo hatches prematurely ! Note data included stacks package necessarily representative unbiased subset complete dataset, demonstrative purposes. reg_folds class_folds rset cross-fold validation objects rsample, splitting training data regression classification model objects, respectively. tree_frogs_reg_test tree_frogs_class_test analogous testing sets. reg_res_lr, reg_res_svm, reg_res_sp contain regression tuning results linear regression, support vector machine, spline model, respectively, fitting latency (.e. long embryos took hatch response jiggle) tree_frogs data, using variables predictors. Note data underlying models filtered include data embryos hatched response stimulus. class_res_rf class_res_nn contain multiclass classification tuning results random forest neural network classification model, respectively, fitting reflex (measure ear function) data using variables predictors. log_res_rf log_res_nn, contain binary classification tuning results random forest neural network classification model, respectively, fitting hatched (whether embryos hatched response stimulus) using variables predictors. See ?example_data learn objects, well browse source code generated .","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/predict.model_stack.html","id":"ref-examples","dir":"Reference","previous_headings":"","what":"Examples","title":"Predicting with a model stack — predict.model_stack","text":"","code":"# see the \"Example Data\" section above for # clarification on the data and tuning results # objects used in these examples!  data(tree_frogs_reg_test) data(tree_frogs_class_test)  # build and fit a regression model stack reg_st <-   stacks() %>%   add_candidates(reg_res_lr) %>%   add_candidates(reg_res_sp) %>%   blend_predictions() %>%   fit_members()  reg_st #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Out of 11 possible candidate members, the ensemble retained 4. #> Penalty: 1e-06. #> Mixture: 1. #>  #> The 4 highest weighted members are: #> # A tibble: 4 × 3 #>   member          type       weight #>   <chr>           <chr>       <dbl> #> 1 reg_res_sp_03_1 linear_reg 0.485  #> 2 reg_res_sp_10_1 linear_reg 0.247  #> 3 reg_res_lr_1_1  linear_reg 0.129  #> 4 reg_res_sp_05_1 linear_reg 0.0666  # predict on the tree frogs testing data predict(reg_st, tree_frogs_reg_test) #> # A tibble: 143 × 1 #>    .pred #>    <dbl> #>  1 119.  #>  2  81.4 #>  3 102.  #>  4  35.5 #>  5 119.  #>  6  50.5 #>  7 122.  #>  8  82.7 #>  9  50.2 #> 10  75.7 #> # ℹ 133 more rows  # include the predictions from the members predict(reg_st, tree_frogs_reg_test, members = TRUE) #> # A tibble: 143 × 5 #>    .pred reg_res_lr_1_1 reg_res_sp_10_1 reg_res_sp_05_1 reg_res_sp_03_1 #>    <dbl>          <dbl>           <dbl>           <dbl>           <dbl> #>  1 119.           138.            125.            121.            114.  #>  2  81.4           82.4            84.8            81.8            77.1 #>  3 102.           116.            111.            112.             93.3 #>  4  35.5           35.8            29.7            32.5            29.6 #>  5 119.           111.            115.            115.            127.  #>  6  50.5           38.8            37.4            36.2            55.3 #>  7 122.           123.            103.            104.            137.  #>  8  82.7           82.3            78.6            82.0            82.8 #>  9  50.2           38.7            37.3            36.2            54.8 #> 10  75.7           78.8            75.3            76.9            71.8 #> # ℹ 133 more rows  # build and fit a classification model stack class_st <-   stacks() %>%   add_candidates(class_res_nn) %>%   add_candidates(class_res_rf) %>%   blend_predictions() %>%   fit_members() #> Warning: Predictions from 1 candidate were identical to those from existing #> candidates and were removed from the data stack.   class_st #> ── A stacked ensemble model ───────────────────────────────────── #>  #> Out of 21 possible candidate members, the ensemble retained 5. #> Penalty: 0.1. #> Mixture: 1. #> Across the 3 classes, there are an average of 2.5 coefficients per class. #>  #> The 5 highest weighted member classes are: #> # A tibble: 5 × 4 #>   member                       type          weight class #>   <chr>                        <chr>          <dbl> <fct> #> 1 .pred_full_class_res_nn_1_1  mlp         12.0     full  #> 2 .pred_mid_class_res_rf_1_06  rand_forest  0.670   mid   #> 3 .pred_full_class_res_rf_1_05 rand_forest  0.101   full  #> 4 .pred_full_class_res_rf_1_07 rand_forest  0.00457 full  #> 5 .pred_full_class_res_rf_1_01 rand_forest  0.00219 full   # predict reflex, first as a class, then as # class probabilities predict(class_st, tree_frogs_class_test) #> # A tibble: 303 × 1 #>    .pred_class #>    <fct>       #>  1 full        #>  2 mid         #>  3 mid         #>  4 mid         #>  5 full        #>  6 full        #>  7 full        #>  8 full        #>  9 full        #> 10 full        #> # ℹ 293 more rows predict(class_st, tree_frogs_class_test, type = \"prob\") #> # A tibble: 303 × 3 #>    .pred_full .pred_low .pred_mid #>         <dbl>     <dbl>     <dbl> #>  1     0.909     0.0540    0.0371 #>  2     0.101     0.434     0.465  #>  3     0.0900    0.421     0.489  #>  4     0.104     0.430     0.466  #>  5     0.909     0.0541    0.0373 #>  6     0.909     0.0540    0.0372 #>  7     0.909     0.0540    0.0371 #>  8     0.909     0.0540    0.0371 #>  9     0.909     0.0540    0.0371 #> 10     0.909     0.0540    0.0371 #> # ℹ 293 more rows  # returning the member predictions as well predict(   class_st,    tree_frogs_class_test,    type = \"prob\",    members = TRUE ) #> # A tibble: 303 × 18 #>    .pred_full .pred_low .pred_mid .pred_low_class_res_rf_1_06 #>         <dbl>     <dbl>     <dbl>                       <dbl> #>  1     0.909     0.0540    0.0371                       0     #>  2     0.101     0.434     0.465                        0.339 #>  3     0.0900    0.421     0.489                        0.220 #>  4     0.104     0.430     0.466                        0.323 #>  5     0.909     0.0541    0.0373                       0     #>  6     0.909     0.0540    0.0372                       0     #>  7     0.909     0.0540    0.0371                       0     #>  8     0.909     0.0540    0.0371                       0     #>  9     0.909     0.0540    0.0371                       0     #> 10     0.909     0.0540    0.0371                       0     #> # ℹ 293 more rows #> # ℹ 14 more variables: .pred_low_class_res_nn_1_1 <dbl>, #> #   .pred_low_class_res_rf_1_05 <dbl>, .pred_low_class_res_rf_1_01 <dbl>, #> #   .pred_low_class_res_rf_1_07 <dbl>, .pred_mid_class_res_rf_1_06 <dbl>, #> #   .pred_mid_class_res_nn_1_1 <dbl>, .pred_mid_class_res_rf_1_05 <dbl>, #> #   .pred_mid_class_res_rf_1_01 <dbl>, .pred_mid_class_res_rf_1_07 <dbl>, #> #   .pred_full_class_res_rf_1_06 <dbl>, …"},{"path":"https://stacks.tidymodels.org/dev/reference/prediction_eqn.html","id":null,"dir":"Reference","previous_headings":"","what":"Convert one or more linear predictor to a format used for prediction — prediction_eqn","title":"Convert one or more linear predictor to a format used for prediction — prediction_eqn","text":"Convert one linear predictor format used prediction","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/prediction_eqn.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Convert one or more linear predictor to a format used for prediction — prediction_eqn","text":"","code":"prediction_eqn(x, ...)  # S3 method for `_lognet` prediction_eqn(x, type = \"class\", ...)  # S3 method for `_elnet` prediction_eqn(x, type = \"numeric\", ...)  # S3 method for `_multnet` prediction_eqn(x, type = \"class\", ...)"},{"path":"https://stacks.tidymodels.org/dev/reference/prediction_eqn.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Convert one or more linear predictor to a format used for prediction — prediction_eqn","text":"x object uses glmnet::glmnet() model numeric predictors. ... currently used. type prediction type.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/prediction_eqn.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Convert one or more linear predictor to a format used for prediction — prediction_eqn","text":"return type varies, based model prediction type.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/reexports.html","id":null,"dir":"Reference","previous_headings":"","what":"Objects exported from other packages — reexports","title":"Objects exported from other packages — reexports","text":"objects imported packages. Follow links see documentation. butcher axe_call, axe_ctrl, axe_data, axe_env, axe_fitted, butcher dplyr %>% generics augment ggplot2 autoplot","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/stack_predict.html","id":null,"dir":"Reference","previous_headings":"","what":"Convert one or more linear predictor to a format used for prediction — stack_predict","title":"Convert one or more linear predictor to a format used for prediction — stack_predict","text":"Convert one linear predictor format used prediction","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/stack_predict.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Convert one or more linear predictor to a format used for prediction — stack_predict","text":"","code":"stack_predict(x, ...)  # S3 method for elnet_numeric stack_predict(x, data, ...)  # S3 method for lognet_class stack_predict(x, data, ...)  # S3 method for lognet_prob stack_predict(x, data, ...)  # S3 method for multnet_class stack_predict(x, data, ...)  # S3 method for multnet_prob stack_predict(x, data, ...)"},{"path":"https://stacks.tidymodels.org/dev/reference/stack_predict.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Convert one or more linear predictor to a format used for prediction — stack_predict","text":"x set model expressions generated prediction_eqn(). ... currently used.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/stack_predict.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Convert one or more linear predictor to a format used for prediction — stack_predict","text":"return type varies, based model prediction type.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/stacks.html","id":null,"dir":"Reference","previous_headings":"","what":"Initialize a Stack — stacks","title":"Initialize a Stack — stacks","text":"stacks() function initializes data_stack object. Principally, data_stacks tibbles, first column gives true outcome assessment set, remaining columns give predictions candidate ensemble member. (outcome numeric, ’s one column per candidate member. classification, many columns per candidate member levels outcome variable minus 1.) also bring along extra attributes keep track model definitions, resamples, training data. See ?stacks_description discussion package, generally, basics vignette detailed walk-functionality.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/stacks.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Initialize a Stack — stacks","text":"","code":"stacks(...)"},{"path":"https://stacks.tidymodels.org/dev/reference/stacks.html","id":"arguments","dir":"Reference","previous_headings":"","what":"Arguments","title":"Initialize a Stack — stacks","text":"... Additional arguments. Currently ignored.","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/stacks.html","id":"value","dir":"Reference","previous_headings":"","what":"Value","title":"Initialize a Stack — stacks","text":"data_stack object.","code":""},{"path":[]},{"path":"https://stacks.tidymodels.org/dev/reference/stacks_description.html","id":null,"dir":"Reference","previous_headings":"","what":"stacks: Tidy Model Stacking — stacks_description","title":"stacks: Tidy Model Stacking — stacks_description","text":"Model stacking ensemble technique involves training model combine outputs many diverse statistical models, shown improve predictive performance variety settings. 'stacks' implements grammar 'tidymodels'-aligned model stacking.","code":""},{"path":[]},{"path":"https://stacks.tidymodels.org/dev/reference/stacks_description.html","id":"author","dir":"Reference","previous_headings":"","what":"Author","title":"stacks: Tidy Model Stacking — stacks_description","text":"Maintainer: Simon Couch simon.couch@posit.co Authors: Max Kuhn max@posit.co contributors: Posit Software, PBC [copyright holder, funder]","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/tree_frogs.html","id":null,"dir":"Reference","previous_headings":"","what":"Tree frog embryo hatching data — tree_frogs","title":"Tree frog embryo hatching data — tree_frogs","text":"dataset containing experimental results hatching behavior red-eyed tree frog embryos. Red-eyed tree frog (RETF) embryos can hatch earlier normal 7ish days detect potential predator threat. Researchers wanted determine , , tree frog embryos able detect stimulus environment. , subjected embryos varying developmental stages \"predator stimulus\" jiggling embryos blunt probe. Beforehand, though embryos treated gentamicin, compound knocks lateral line (sensory organ.) Researcher Julie Jung crew found factors inform whether embryo hatches prematurely !","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/tree_frogs.html","id":"ref-usage","dir":"Reference","previous_headings":"","what":"Usage","title":"Tree frog embryo hatching data — tree_frogs","text":"","code":"tree_frogs"},{"path":"https://stacks.tidymodels.org/dev/reference/tree_frogs.html","id":"format","dir":"Reference","previous_headings":"","what":"Format","title":"Tree frog embryo hatching data — tree_frogs","text":"data frame 1212 rows 6 variables: clutch RETFs lay eggs gelatinous \"clutches\" 30-40 eggs. Eggs clutch ID siblings ! variable useful mixed effects models. (Unordered factor.) treatment treatment group embryo. Either \"gentamicin\", compound knocks embryos' lateral line, \"control\" negative control group (.e. sensory organs intact). (Character.) reflex measure ear function called vestibulo-ocular reflex, categorized bins. Ear function increases factor levels \"low\", \"mid\", \"full\". (Ordered factor.) age Age embryo, seconds, time embryo jiggled. (Numeric, seconds.) t_o_d time day stimulus (.e. jiggle) applied. \"morning\" 5 .m. noon, \"afternoon\" noon 8 p.m., \"night\" 8 p.m. 5 .m. (Character.) hatched Whether embryo hatched response jiggling! Either \"yes\" \"\". (Character.) latency Time elapsed stimulus (.e. jiggling) hatching response stimulus, seconds. Missing values indicate embryo hatch response stimulus. (Numeric, seconds.)","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/tree_frogs.html","id":"source","dir":"Reference","previous_headings":"","what":"Source","title":"Tree frog embryo hatching data — tree_frogs","text":"Julie Jung et al. (2020) Multimodal mechanosensing enables treefrog embryos escape egg-predators. doi:10.1242/jeb.236141","code":""},{"path":"https://stacks.tidymodels.org/dev/reference/tree_frogs.html","id":"details","dir":"Reference","previous_headings":"","what":"Details","title":"Tree frog embryo hatching data — tree_frogs","text":"Note data included stacks package necessarily representative unbiased subset complete dataset, demonstrative purposes.","code":""},{"path":[]},{"path":"https://stacks.tidymodels.org/dev/news/index.html","id":"stacks-103","dir":"Changelog","previous_headings":"","what":"stacks 1.0.3","title":"stacks 1.0.3","text":"CRAN release: 2023-11-06 Refine package alt text (#202). Update example objects, resolving deprecation warnings recipes (#203). Fix bug type checking blend_predictions(mixture) (#204). Resolve package-level documentation aliasing notice CRAN.","code":""},{"path":"https://stacks.tidymodels.org/dev/news/index.html","id":"stacks-102","dir":"Changelog","previous_headings":"","what":"stacks 1.0.2","title":"stacks 1.0.2","text":"CRAN release: 2023-04-20 Added augment() method model_stack objects (#173). Converted character variables tree_frogs example data factor updated downstream example objects (#177). Fixed bug resulted errors using model formulas \"mgcv\" engine (#193). Made several optimizations reduce evaluation time memory allocation stacking. Various bug fixes improvements documentation.","code":""},{"path":"https://stacks.tidymodels.org/dev/news/index.html","id":"stacks-101","dir":"Changelog","previous_headings":"","what":"stacks 1.0.1","title":"stacks 1.0.1","text":"CRAN release: 2022-12-14 Removes unneeded data import attribute tree_frogs example data associated objects (#148). blend_predictions() doesn’t error anymore control argument isn’t control_grid object. long object passed control include elements control_grid() output, parsnip::condense_control() handle input (#149). Tightened integration workflowsets package (#161, #165). Refined logic adding candidates via workflowsets allow partially trained workflow sets. case workflow set contains failed tuning results, stacks inform user excluded data stack add results trained successfully. Extended documentation related packages’ interactions, including new article package website. Revamped errors, warnings, messages. Prompts now provide thorough context arose, include extensive references documentation, correctly pluralized (#150, #167). Various bug fixes improvements documentation.","code":""},{"path":"https://stacks.tidymodels.org/dev/news/index.html","id":"stacks-100","dir":"Changelog","previous_headings":"","what":"stacks 1.0.0","title":"stacks 1.0.0","text":"CRAN release: 2022-07-06 stacks 1.0.0 first production release package. release includes minor bug fixes, ’s accompanied white paper recently published Journal Open Source software. can read paper ! release: Addresses re-introduction bug arising outcome levels valid column names multinomial classification setting (#133). Fixes bug stacks return incorrect predictions elastic net meta-learner used, type argument predict set \"class\", outcome levels differ alphabetical order. Transitions package internals functions deprecated recipes package.","code":""},{"path":"https://stacks.tidymodels.org/dev/news/index.html","id":"stacks-024-github-only","dir":"Changelog","previous_headings":"","what":"stacks 0.2.4 (GitHub only)","title":"stacks 0.2.4 (GitHub only)","text":"GitHub-release change package source code. update includes data-raw/paper subdirectory containing source contributed paper Journal Open Source Software.","code":""},{"path":"https://stacks.tidymodels.org/dev/news/index.html","id":"stacks-023","dir":"Changelog","previous_headings":"","what":"stacks 0.2.3","title":"stacks 0.2.3","text":"CRAN release: 2022-05-12 Addressed deprecation warning add_candidates (#99). Improved clarity warnings/errors related failed hyperparameter tuning resample fitting (#110). Reduced model stack object size fixed bug object size model stack inflated drastically saving file (#116). Also, regenerated example objects change–saved model objects may need regenerated order interface newer versions package. Introduced times argument blend_predictions passed rsample::bootstraps fitting stacking coefficients. Reducing argument default (25) greatly reduces run time blend_predictions (#94). package now load packages necessary model fitting fit_members(), available, fail informatively (#118). Fixed bug meta-learner tuning fail outcome names levels including string \"class\" (#125).","code":""},{"path":"https://stacks.tidymodels.org/dev/news/index.html","id":"stacks-022","dir":"Changelog","previous_headings":"","what":"stacks 0.2.2","title":"stacks 0.2.2","text":"CRAN release: 2022-01-06 Fixed errors arising outcome levels valid column names multinomial classification setting. Fixed collect_parameters failing return stacking coefficients two-class classification setting. Regenerated example objects updated {rsample} fingerprinting–saved model objects may need regenerated order build stacks combining models generated update.","code":""},{"path":"https://stacks.tidymodels.org/dev/news/index.html","id":"stacks-021","dir":"Changelog","previous_headings":"","what":"stacks 0.2.1","title":"stacks 0.2.1","text":"CRAN release: 2021-07-23 Updates importing workflow sets use add_variables() preprocessor. Plot fixes cases coefficients negative. Performance member plots now show effect multiple mixture values. Package diagrams now alt text.","code":""},{"path":"https://stacks.tidymodels.org/dev/news/index.html","id":"stacks-020","dir":"Changelog","previous_headings":"","what":"stacks 0.2.0","title":"stacks 0.2.0","text":"CRAN release: 2021-04-20","code":""},{"path":"https://stacks.tidymodels.org/dev/news/index.html","id":"breaking-changes-0-2-0","dir":"Changelog","previous_headings":"","what":"Breaking changes","title":"stacks 0.2.0","text":"release package changes elements internal structure model stacks. , model stacks stored saved objects need regenerated predicting, plotting, printing, etc.","code":""},{"path":"https://stacks.tidymodels.org/dev/news/index.html","id":"new-features-0-2-0","dir":"Changelog","previous_headings":"","what":"New features","title":"stacks 0.2.0","text":"package now supports elastic net models meta-learner via mixture argument blend_predictions. package can now add candidates workflow_map objects new {workflowsets} package. interface add_candidates tune_results objects, add_candidates now generic function. Objects tuned racing methods {finetune} package can now added candidate members.","code":""},{"path":"https://stacks.tidymodels.org/dev/news/index.html","id":"bug-fixes-0-2-0","dir":"Changelog","previous_headings":"","what":"Bug fixes","title":"stacks 0.2.0","text":"Fixed bug determining member hyperparameters member fitting using non-RMSE/ROC AUC metrics. Fixed bug arising model definition names valid column names. package now message case provided names valid column names use make.names associated candidate members.","code":""},{"path":"https://stacks.tidymodels.org/dev/news/index.html","id":"miscellaneous-improvements-0-2-0","dir":"Changelog","previous_headings":"","what":"Miscellaneous improvements","title":"stacks 0.2.0","text":"Drop {digest} dependency favor {tune}/{rsample} “fingerprinting” check consistency resamples. fit_members() now warn supplied model stack whose members already fitted. Integrate {tune} functionality appropriately coloring errors, warnings, messages. Improved faceting axis scales make autoplot type = \"members\" informative. Various improvements documentation.","code":""},{"path":"https://stacks.tidymodels.org/dev/news/index.html","id":"stacks-010","dir":"Changelog","previous_headings":"","what":"stacks 0.1.0","title":"stacks 0.1.0","text":"CRAN release: 2020-11-23 Initial release!","code":""}]