From c0564ebd16224af93a69a28f9175adebd0e07137 Mon Sep 17 00:00:00 2001
From: kosabogi <boglarka.kosa@elastic.co>
Date: Thu, 24 Oct 2024 20:07:15 +0200
Subject: [PATCH 01/19] Adds Trained model autoscaling page

---
 serverless/images/ml-nlp-deployment.png       | Bin 0 -> 108579 bytes
 serverless/nav/serverless-general.docnav.json |   3 +
 serverless/pages/ml-nlp-auto-scale.mdx        | 132 ++++++++++++++++++
 3 files changed, 135 insertions(+)
 create mode 100644 serverless/images/ml-nlp-deployment.png
 create mode 100644 serverless/pages/ml-nlp-auto-scale.mdx

diff --git a/serverless/images/ml-nlp-deployment.png b/serverless/images/ml-nlp-deployment.png
new file mode 100644
index 0000000000000000000000000000000000000000..640bf19669539335cee4b70aa007f2f968af445c
GIT binary patch
literal 108579
zcmd43g<o8|)<29EFK%sdcZXtwQyhxBOL2FnNO5<E;>F#qxVu|naCaEwot}HoIeqT$
zA9(rfVZz=?)=pNEW#2`pg4`z*#CM1g5D+Mm5~7L_5YQ<Q5KyD=Z@_o_a|re!ARzV4
zMMM-NMMQ`d9BfU^txO;wsAG-v^?ymyQx6&#=<5%T(a|6}xGH}87NMx`+uhYi+}$;i
z)7zDtqOHA&hqei=+yjySxx3L4IfD4HhH>!6wp9_;oY+(S;*NbF(*Vc~q0j@XD=|(p
z1Nr+aMD=F}A#?@)e8?CA1zkl5LoSFVQZd0#kV{H}$4>B`u-`Ia49+0S5D*;R)G$FT
zK@H)D(g@x{6Khc+Nb(VcW~M8pAv;44-td*U=XMKfB7ZZ%)5hb&<HPsD^TLntn?(PH
z;uixMLWo&KXcp`h3>7RDOd2c`>=cX~jPV9GXB>)n)Ib8Zh9CbDFAo3O&!7~7bK+Y!
z4A%U<!8f$@yq>PjzP_%m*{^7|-aTDa!if7iD1j-%+;85juOK6P(jsAmV{DO(+(#CC
z99x9%*@Y8l@%;Jvn*E4}_c5A{`N`+*_4RD>_4T#FP+xz&yUWfG6QW1kV6<xx`4@P~
zRZY|-O=V>vXu)N82*?m~2pDh)68w1w{(!j?7YqRh{>A`*M0242ehQtE1O0ayYV>zS
zAtez>N$|Ilk%NhejpG+vCmkXZDsWfx=E~|$>asFCMz+?B2FA9ACX8;@cE6iI@VW7T
zi`FJi2E=aGRyK}2Zv3Qw)!+e_f8S;zCH||5lO;c?x~u}Rh^>POF()H4BQq%gk(ij6
z&%xM~M^RM#zs<pa@soaWa<b!LVsdqLWprg@v~@6JV&UfIW@2V#Vr69j*I;mTw{bFX
zW3X`~`%fo-_akcJXyjmS=VWecL;Sm614CP9Cw@}W--Q18`OkZrxS9V;$;R=&!vYVG
z>Gu;R7Di^KfBFVD<@<e?N5S09#7aZd+!{PRU>X1}7EZpu>i<7a{-yXwOZ9(Sa<el3
z+47Gk|Jzc<(ZoT-)*4LG3Ggpx{@eJ^hyQKJ$Mk#Tf6T;xrunbC;CTih@-h9xGXP?d
zw7x6^gdl{ZsF1Q7<Z&jP7gitcfSyF|-6StOJhcSo59)7*-NDqf0s3NnJ$lvQus_4v
z>BgX_!@}Aj3~^y#3hvfDQP5tl(v3MC)7#Uen@7$rxLsVA$Gu$7YHT;sxe!iP8U;wP
zFo+@l_ZlSe|Nism&kvm4hvb4#|9cU~CB(!CzWv|Z;EFLuP|D1X5iuw+5D5Qs`3a&M
z{=bBX{lB2)z_Pu=T6*)p4Pqb|Qrh4o{9yj)$`kQpteM&xRQ;d(5;>EYIpJ2hov@d0
zx?;(+Z8&TtuIT@H$)HFGwIFIMwKn}|t`dFM==N|iC}m2ci_+?ef9h0Absa5*OPT%9
z9a}u(!X91h41VB&swArOIzf|{`(qx6btk?k^SB%Bz(S69k0K4*d%UeAs`yAi8zS~^
zEf#sd2^$t)1S3O;I{p5iCbS<jcTknOn{kd2{vB(ebeB(3z8IUf2R#uYByCEr0_~2;
zF;SjyR6sA)KYa0TBGDJv{=5{CK9Y6@og0vPmYlQc()BV$)6{M}@#jQ1NOkeueV!Vn
zk(@6I81f{34|xS0&+rxDpD}_9_Oe}kaEenAM<W?!U-JV}%iN60?61CJjOp%)r%xRd
zt&sjp*o_2DwqTO4SFCo-ZEMsXhEfogjX7I5*3w5%n82zDC}N!q{F6Q@71}tPW^q`)
zIos@I>|GYeb`S*is9SR!dPMJ&exM?U!jhnSZrvE`pD)xb-BnJ!JYjpvSDBS^nVWD^
zI#Jgn;)iIu)jSU%;}F0Agc8D?I-D9s)5`q&Zjcz%`Y>%-TM;#1E=TQiD)OzUfl&3`
zZTG+5v7@qbZ+|rUo0+L;Uh60@S&xx`*CpW%$~y@~4BK?U<CIVTG72`cD_11Mcqop7
z)qI*9pUpJlw8H%bsZdaHh<;y<%NLY934TuYumYRK&xDXd6eYM<H=<mrt3&9%gKoas
zWS>JjO7_o`L1KLS426V4A0zbj&CqhaWhOd|66<H?1Xj7Pt2FE|VOVLg4ml*~%)=B7
z_X^F(2=9?%|Jf{pqZoc9kP}|J52(;A6M-ze)GmA*@Cx0dT-WZtzZ86_Ww@LArX5o6
z7DCO)|DovLM(Jk;-~$Pb*6LRQ0yP=cc}Js}pfY4|^%iE%z6d`++9e;%mc(z|rU?Bz
zYH~C^nDb+^T_ZzdH%{c;H}9oV@AGze(|SW@DERsLdmrgC+~IkYnZ3_qoA?~)qWs^O
zCyNOZA^y8|V+6N|^#wpd`5%LoICUjx*R;^6C8dP}mdT5KvCwkD>Cd2LyHx38{fxxJ
zfn$G=4hn(DdTz@?KtlDO{8XAKU7$|;zS0oKqrFEV4$(A2&*UyTaKTO%zxIZ<m;H~k
z$-hUei|>7!b<R|XBBjE!`tVmRdhL)Bbq_rb+LU!|MA_1XXn!@$T*w~(xPRxAglhAK
zem_xFo#lW#Imr2(BqANr6h6kVUP8_X-a(N|)xCP--_@wFh_JJR&rIcP6w2E#c~tO9
zw~N06r`gPAK3AHak@0~&E<r`R(Z69eE=EB#E}mchgz?ZZ0^nXBBD{4wR<xzf^U>{G
zpHtxW=TFgYDlQ5J23am!-=X&V#fRo4q6XPN82|z^0FS=Z>NuM99zMZi$PWO?>~7(%
z)o7C_7LH?9Hs9h*Cn+g8lysU1A!fH+cc81mb8^4E5?t_QD>#eeh{%6>m!e#=J|Rym
z+`|N@wS2tf_><bR5L2a8Irj19D7ClP*1n_jH7mA!Ef^&_E)1K_RN9pauVBV%?SYT-
zp9KWoNq8{2p3m0iD}{-5pNW`K0p7tfO4&YKx<`p8sip}MwKw!XMKRz7i8r*3xvZBz
z<Y+C`HdwFhsx^=5wwxD!FY*kjMjVKc2iBag3w+@IdyNPlqxm_wk2yEOV@Wm=qqtLJ
zR_yF(fG4;2glYs=2I`?uze1}MKu>zKU<+TyhXhBt`Er?VREwgW-uy311fd>eA(BR1
z*FCt?<IEGJcOGhO-W;{MQ?hF*t$SNdH&Z~05N}|OK7($D#xr?k1)yy1PF9$BU5??^
zgHySwx`-k&*o=|>?Jy=8W>w0RJwbZrRU#XOtlVsT#M%n;JmYJsnn-|3jZ7-5kbI*(
zS$VP$39t&n7v2FrW=`@SK}4JlzbHl75`#HX;0f<0Zmdtlg-r-Y8dK-{>O=5>O28(A
zL93zIJ<-`2M%_14v?>SXJ?G;{6!r7@R<8$uDzBNe#X=57j|XO6NJ~<3YHopTSzIBX
ztR-74Y$vLFNtLnt{q|5ah=gjv-}zsvelU3=UFFIx(fN)z4K-$WHU)cDF>9W;5+!jI
z1;!<c0e(5vjFN0-<I_geIxTg|<(guON=jEwGVi%;;@)wTg82867_?v!4Zx(sz=GdW
z47)pERk2OfzC^LPM!8g(hMWTZgk<`(#rep1p<+$W@mrNvgOx^!QjyEgup@tHI2ajD
zo7EHsD24=sYxHc4(upZFHXP~;&)d`3>)r#Y%z{59{!c{66$Tt{al7m<nk+^^BJ?+y
zVHALvH#bKkaRt)#d7<0tYODnDD-}Ae2{m=nW8td>tE$pQR_>%nDplEsb7c?pDI@|^
zP(ju&Pp~zvXS$tZ$1AP!NlbdV0+H`Hkef0F3<~@hH0vbVuvFAvIIWkzCoWOSM%{V>
zeDN(8D&n!}RPRUTRTt_l7Z3sU7IV{K?k>_s`-db6+kN4<AnKajh1(Tb(?8#N2OUzY
znH|zwW;6RqnbrJe)CYPIN17KmNt#UjAi{)#Eq#LGO`=l4I0T{zQh`-#HUz*kq1mmG
zmEaSt=k0N#?@KCyLbFq2yWJM4M!RVw?-F;GfNw{U!vU58uX-8gNuZItPK~j-AaL4-
z2vnw8t~k7f;p#9rpwH%ds!arB-5I8r_~U$tIfH3Skail*iVtO-LZgmI7=#zB)^0Hl
zZBQ4s59GBX7J5)mvO=VMW7O&`?rE1&wx5LJlSF)x=6Rs9u|^ohl5Ce-!ycT!lj@>c
zt_{TB!Wq6BM<d%Xw6QGS3!cMdtlk>l9ZRoYv2D6Jhpx7`DYYxB=KFi^f)7wZNK$55
zn&_7w3F1Sdsb7#1$Iid?JY%a4Kcc+bFIPJOH|DLVPKn1a?ZRtR#UD85{vM%iAj#Oy
zE+LGyfQU`|1$VA|rC~Vgq`CD2{HlRkx#kI*OH+DJV*Qt?JUN~OHX~9Oj5#?ldyH=^
zHXdC%mh|$8jcBl#O0*iI5)^Uq=augGq?CN7^7^JrRip=;arM4~8Jl)&wE|uuNA9tp
zS#ER!E@j{)SVH?S#MWCZ#32GlkODx02YTSe+u*XLs8ym^${1g+M2LU~H-6xp8IH|F
z>3lT%d7C<Sc*knJY@j4n^p8nL6N7rTY2(zlRA9XbRX$%$@#aw#xAeax<=5kz{|bOE
zGa*i1SgJJxA@ipM%`hU1U>9-==SZs{vt*5{Vq0PBrgK=#!hhtS1gI5*BxTb%?QK55
zJZLm{HUH2{u62+0my7Zj$rX$Fpv6gzq-)JlIXPG;n?6=i8}v)5#VIMeCVp$V8|BR8
zS+v?Hio3rk(g8R;(&p}1t6`n%)uS1j!;ePorYl88Upkh?5g$h(z3*_(#8Dsj1Ul=?
z0%i)Q^4h$mBwFapcyHbphCSnGWHy>&9=6Z$2m{-0*cS#k@PE76b3~Cyr!pVLbx%QI
z(#Y;PXYlS%u0Nr{fB*PqIGhM-iU0vi2Iis~c6isuWbT^kHbchOB<5{CdV(}w{$_iL
zbASvi=74%N88^3N&e=DQj*ahaM7gb9lMQT2G-BQ`v|?QgWTkKE^O#exRa)SNo7+FR
zmR*7piYPA?aH~DR@`y2I`5UV09Q7|By3a)}Csru1@6^fN!SZWd6-A1{Mw9G_F94)m
zraET~w`y3WE2Rt^PV`EUW{*oz7quv1@f_M6)5_R6;cegr-M%H=Tpp&lMBBm^auY^p
z65~->!)tIbU-oy<tu>W(+_$o43nC4!>;%g7*h9VrdeMvg$@fl}(OQ$IC}tiqN4kiQ
zw`NnwrYCitb!===C{A#A`M0!v%nT=fjByna3g@${HQc?$uA=3b{zmG2kt8Uv1`54!
z3l7bf5wrc`1z5e(4xxx|Jnqj)KnX4M%l>8*ziy8=Dpk7_fXT(&g&%{9N-)LdX9F)k
zi)pYfRA^Yn<#hyI!pP!(WrLFb@WZ|8r$LPFY5CySOFWqzXHSHryYsl3VOczBzotY*
z-b<@x{~sO8*2=0Q>23%KOTgiGA7pY4T48+_URM3uhqrs)B+~;4j@w~uhLM!*^04}#
zY+0!P6@PzObW>FNSTe4HMDc{f_>CndcqEf1IlV&Nz3fd=KG&Etls}?55-oq6kRq9=
zXptwr^)vVQ7F(;KA=p?<LDqhqiP0BcmMA!@7>|HZ;{cTEX4y?SRibk>R${jQ^=o)E
zH65yqX4PJ;RG+8ehmxucbu4%|ekYt$%g!gyb?mmdlfNuBU%m%A*gH>12lkkdpXj^5
zVm|vD9#T+Fl-Dhr7_{6vv<5kxWEEK~IG&~ZyzvM(#t*5NW`8PYk)M;EPCz=rNk*zw
z<1YMHdxk*xNd(EK)YAUpIg<6ao$<09-gqr|MN`jO;`HI0+HtBNeU1@c?+jtVEPf_*
zR%ADpOr?v#b6^64n%sf*aJg!qzkh)F`_*HQl%v$VHoC|Q)EgEtv|spQI2|p`47&58
zPIe*FS?C0F47zExtr6Uueq+^3JyeU?-^tAy50_3p@a$U>iu1lCNIOR9Kb)sW1urIT
zDy)39v<*TWCQ28E`b-Le8fbTVU^nuqB;6_)<}wK4MGyoc%VZ+>CHY&}6T|xp28R=?
zq*|l<sZ4y+sQ77JFY42TNa;jL1{P~`R}A)TdiaBHi@D>f=!M@B4l}nsuyq{>9}47C
zrApM99?563FcHhF;!sbc%&K2p%r^D7OR#4mx*~DOR_Y6Y?UK=XtN>5~Ew)=bUTYkF
zab~c0#0U{N^hSk|mJEs$?}n=2iBl~R{f5oP>`)nlO;~;ukCgnIV4|DTwIv<rt5T+0
z_jA}(UfYhh>~mFHpB=8JBKi07mz|r42%szzw4OYGBB<TAie*U*A~yHyo<uR9^EXr=
zD`<zCs?Vh3DP`-NeUdao<|Slr6nmSz-4eNgzpRm%V0}M0c>VzEDvHt=<!?RGl!4re
zMNxO!*Exws#eSWZTTJ|`xnoL4<Y#N?4{y|)=?I*a`$hUbZ3`r=*u2k4_dMv^E6{2T
z@3H2w86i^kO~x(b23+lqFI@Siaq?@vT3_BZF2AYPX%03hKT@G(<<Bc2a@9uK<LYO`
z9r9sc?(h9&VcGIImK0(7vT2+2;O!AuGx+%t@`-R0ngaQXY0Z%}HpGnlDg$fSP8K=o
zk=74wRf*(1TQVT<QbVIY#64A=$G$m@$-OHr?OTmxs&2}cqLdoZ9&4O{uCfp|=r>y*
z7wfoT?)|d@X?zOi_F_CvFO-?@PLwol3E=~toAQf2nj?bWtqwSGPks>b^{u=0j9Aj;
z>?<RDL@bz-eh}6Dj%Z3yuMDsK?Mir<?urXg6+o(Zn$?bvp5t<g3|Qbudj{>q%LB7K
z&#si6qWN#}+;0+-XWyf5z-udnxt@(5QvFzcfx;83t^9&?S6=<0HxdGstD2F>i&;F&
z^F_=IZzQ?op(*vs89Ea8<wb#!1U}+s9b{xYxcP}AyF!P_X}R@k9Flc2YUTO~qViI%
z`{PQt_pd$@reN-uZF%GK&4I63rM4SoCg(=cFbudyuunQjz6V=TuA4e{gBxtDq~48|
zu5B{+J7jA=s|+T1Ww&_6VU@0tigEQk#Ik%IH9{iWzix!B*mllcozcCV1{_>5wHd2T
zf!sb{E|)P`ny_h^!Xr#sy!U-&I<haISGGs7dT<?=d3IQn+Xf(DIYBu&jy+%pK!rLp
zk~G|4fCj%VCdp-@?FjtL@P8X;5I9n$N;*C!UO;T4B1MKJLw^RpjY8V5sxr_Y<5!sb
zF-cVl$T$ZlnqFRhjY_Z9q>7(4`DVHQgEYPQ0R_HU;Ac{yPeVHO7Ky|XbXk41G$~D?
zlV5b<i3RPE3cnfXntoKQ`@%Bu%dVYcxmgEd%Pn`UwJ)nSRgt|FiD8%w1CudJt8LY5
z&BRVyTP$NO`JsPhr#F(e5{tr7PF-mvnW?A;FL};3mPLS_2`22mH=2*~tc|Vb>|l5F
zL2Vqc$LGFx!UarFl5-gb>%W@8d@q5He72{@e6=p?)2}Y6tUxMn8}%cfr>*xzL>En%
zU)tPO=8Hdz><N{6gIpld1GND9Mx)3Q`n8YVc{>9BY!k=yw(U1jr-LQ)w(2i;^u8Vt
z6sL!Pg~9KVN>48Oi;hf#dcLypckAzgZTrFovnQ@I43}U>isyWv$J}0qXGa9yOHzjO
zd}@}-6NPx~3D>&Y$b;;)E1)nQ0s)H<s!ID^9|mH9Mjp(88>533y9kL9mTD`#6D}`-
zl<{_d<V`I<mJoHOxSh7VhQP&WWfJz(kV%3~FlyGsv%99Mn!WaUHLP^nhq<cU8C6bL
zOn<!q8ZVcx+)D#prKJU*RGR~y7p<XMIw6;V$Imq;?HtxH=%t$VrJT6wmkPydmA7wQ
zls-3u@M((lJ;mmOv`Favn1$YR=a7z9C^GWqN?7zMYt%1rEyjM<TpP*GsdRgfXo-<R
zPIshO<I$?1xR!BqIEMmsYC*2Jm@i+sCb9@51bBTENwr#kQXdHVSc$cYe`wvQ)%^4Q
zzIf!jp|j4{Q+0NO@(*rz?Ux{1fe=19oYD7C_k_MI7i<r=D`&IO>SXovR<(XqZyUY3
z`cv>@Am&?Hq3(d<hXWhZhYMwUiKBJY=x*o!pzpmUMThfp{2WIV8VPC9&89EPVfp3S
zzt-EHE(sqQKJw9wVi<HjJ*Gchj@V9DZ<~+fZKyri05K}6?P}|bJgz%$TvFLQ3w51V
zOFkG=X<Cyxy>5$SBRg=K#tT}Z=U}t(EFgs$Kg+1i?$_aI(z)GUQt3`SZ5`?sTF<CX
zgS`nVJM)ON{6ch$3O1LDv}FY%P;@ElBYrT4WN+a5W>YdwYF=MJ*|U{;@^mIC$B4Dn
z7X}R@kXG*4Vngtwy%Jx=ZIKe}>q2j6!f}}jaVXa+lkpH#BnOC8QxeuLH+=WayscG!
zy6L|eaEp*;d9+tA=srE1a5}M>`Otf_K(`1p-%(ON57Wm=;ZCvfdY(T~qU#M5Yd-Hm
za(cH;x#0qWNM+t)^N!$}kxrtY8@?XZcr{k~WmZzBcql6i?d(0GvG6wWGb%~KR(zDR
ze6-?e!)d*E_iPrOnzW7*Jr|qpI$vtpE?m9MnxeOwqRCdCSil3f_$_Ut;Wo}fDbW6r
zWN0U-Ml2hr>;<1&mFA0`tWHHzb87x+InHoFUIQYoHB$dq2_23(kFvtHA=kzQg)!*b
zGxXI@w{1D|<xdm!fh$f`7r(|90)mhVY}#9vJBgTd&_su`{c6?Rxn5fi#h}}8?JDKc
zId#w&@NKuC#;||h89!ckItxUZ&ntH-2-jI?O|gP4Rf<tI=4nO#?`cHh-?tj4=!Lb#
z*VItn6PkqatSFq)71?3Ao<mvA3%q_AA7CG8(rZim^v#~E-BX9+6|_OH%Ur6pswADv
zIFh25r#@G#E2~GBP9^v4Y<hk403KT<rW$PRA8v4Jvpro{t~5kO3;XT6J$jT{Zq%78
z-S8fP`ezA6eRPmoQD1IxOEw~EhIwIX+%7Qq8$W<_`OaIpJjXv^13k-MGJ8fiL8#v3
z8@67qFVD7&awP>}X>$6tftIrlHz14eKJ6`!eI?wtjgfV0m7TXY_inD&66RB+z0Su0
zR}=AaW={5U4>w#J9}y|FQ$vvP4|lfC_5$Vmj}9>JpYizIirX%BwRnhzccTU5k{ri}
zEJjN9t1-(PF3^1L*6!Rh{3XHaZBQ!2HOA=Yh@Xo1;j3|*!}z7VW;V!ZsMEULK6Z@i
zyXT>z7IUQ~13>~j4T2X}h~&n}-9GIe&fbKa7w2CK_;bED!CL={G$Dr?q0+WO|1yAc
z{T`JfSEA{PV{JH@dtjV^7xuKcDr`3~{P?W%*+#Ci(R!u9Icf{!ea7z!;@j`a#7EV=
z&EQnpx}>X^Zp|0(Aj{;<&b`Zh61bo2z1++8O*I_9t&wgWA&Y*CL5+%b$M+*w1a71#
z+vjwafPEj*zYpKG-`f9|<s;i&r*92Sql-r6^HS~Jelrgx>S;scb*G4cR~ma;?yK9=
z$8x9Jw$kW{=Kx&v-VJx9){?xo*7j(?J36XNw*Ye~U5^J9e1O+?$2ONSM6-?dt2NgR
zI07bJ^^}YJj#SG7n~@K#-07F6*9*GdlBVNNpVqyOtd&%3J$2@k>=>+YUHB4ud-z?B
zLjWdIw~QRwi+oO`GZC(%Db+rE8F|{-9SbxCG$I!R&w-#}iH~=QCGj&19q|(YP(}p7
zUF^ior`G)qX(qtl6)EI_W#U3`?6AeSXB{l!qX;U19<%eFRcoe+iAZUqaz<WYZ0P;R
zbKN}?<O-u&qcl@bJT4)-Q{PPzd=|q5C0t(DVs3ZWwHtx#x~C=_00I0~7hxkm*Nhxo
zG^X4aDp!=YfdLG-0#IVIo)cD&MCZnW-o`|;t(|qD%*%S}#<#Qa1wG%R2;<#R^KqZ5
zOo`>S`~(55G<NP~?!fRw1{eB3ibK#BEgB+k;}|wBBB^^FSCmO6otzz^^8jY$P}LP8
zA9h`RcMb>VGhJ3i*}WXFYKUb6r9h7q?hs0sIVOe^+L!RkIBz&g*ZnoRdUR<IS<SxY
z$k2w0{_WKsQ&qgkvb?OUijY>#t3heS`mpVC8_i0FWTp7nd$Y!XBNd!^nzqE-8d{IU
z#+maReW>ci1OMCHra2`>`<Dbm(VG!E|G5$MS35aA?KRH{K_Pu2`1P9BuhCr|m|>e@
zJ$-8}AikKzn||;6*r$uVY*yz(<8tK{_eJ058!p&kBi)e`*VTi3?O#u|3C_2to~QA8
zzRzo5tKc~RWA?1@<?@6}?;~miB~wykZ=wU_NP(7kCnzgQ6$v;sm=i)YR;W#SmTDG!
zP*bSQ-oV{nNqPbuaMSi89_-w|)_=G0O3crAMq@P4wwG@V_79qL|0EqvfxBO*Gk>Iz
z^~4U`1#6{AN^<;5S1+#Fz23+{xUrv$<ZfKF2E5%(Ce|+x0xuk%F09w)OwwvJe`&VB
zGTG&@w<Jm@i1=lIJs{%a01&+nLuW-$>%`Q@EN;srDpW5xn+~Ar7)wmJQVu{+viEGC
zV?nnvd}~?e{+jK;N0Q5$q#<zxj;Kvfe+yZR)Vi@<@%shHa_1SfU+PSkzaASTas$<u
zgfxD3_<>P#A2f~aNPn?%zjEfgkL&MN?R6<hH5_^vly*|T#g3*=_0HOMxlV)7%Pz(d
zr|l=-^YvS0W&+Yx60qBx9ww{_E{w>ekvFSVrY@SLbBa#NvZ>p9^|@zt&~#XKoZ9o$
zt8)9DR4z*{aP8rKdtIj4JjhVRx_G^3*;aSWdTGFpthcS+ZeT^Q*_{-_mOOx~)n6Nh
z%$zNbXU~;?qUaseUd9|yoBJVQ1_*f*@|djWGq#%)!QF7T=GmCxaYzGj2HkaP6!j*)
zz05!as|6<=z@+qq#69<cHAIsfLAg5xLm9#byWW7gT;(Cf={gNe@3-kXC5jZz?;f!Z
zIX<<%z^EFl*BjK=nLORg*Exhbwq9*o;Fk8S*sv}7=teB)5k)wZEYbBh5JSN8!4ccl
zy3mpFduY>E!3&LgF?yo~y*ha4iA}?buGJ?>x2g9$oBH~m=b(hApsNR-fL{B3z}|oR
zQ9Tijx`nIKby~$e2!5iKStMr=P)!d+oviRE9!*G6MZ09E6@B8gT8riteL+yuH_Lre
z_jc(LkjmUmlSZ`A!|k5ndfpyF{3YcX6u|C<yGfgY*?q%(5qK+yDbIIN=fdkd8paxN
zr8{$HQ6xjn8(X(>@+*8cV`L;_7p$8JnR`BkP8XgaJW}Kj`33xrtT^s(dk3CDBq+mN
zcaPw<8T*@wZ&+3=_=>a@4t_*&KVHm|jcs^7CyVCU;9i|U1xToApw9k|KT^ZMd_ba1
zU=GadTcz=pRwgGmp83=(zLORI^$qK)d|4R~sb;;wNVhoIa9?GkZfg#odX!<V?W1VF
z7|3n^5`$X%Jo+|_)iODR6xjsL+Tf+oeUQThtsKEOL#lYc3chd~$M<<a9X-@>r9ldR
zD@cWyh)IZ>ekGtI*b?ks4&lrRBs0G77)c!0VdRS5T6Vo#a~+PC2W&gOYcm*tLq%3N
zf3xmfbbag|#Q4<`(FFZ2(we`yWCI0oMSEeltX`fU`XE;I?Uw=UbfK&z_tsB(pi_;a
zg?trZn^W%StFF%hEcffvl^6Y2*(iW!Ckxm7m}b4j(aYkELrpZ?>D@X($wVl1b7k9Q
z+RSS6UKWFG+e@j-hR^j>H$ZCO5=eK`TcY$4kr&Uh>cVi{$Y!XrQ)X)3B_n@@BH6re
zFaybv?jo-5c@t$}8Hw?u9xm@|K!rnSDXwg2zqiRFH0}&LHrA-LR5By5NlX6e38@JF
z6lmM=BE0b`5eW6xRdMk9P<*C@y{_YkGs%K)%C>f)6TZLnN0=-utp>N(7h<BU5BCHP
ztvBEx(13kk@Lu<Ir`eagD~r-L0sj@Q4$pIa?p^hb=B^EIP~uB@A5y?BLrntmXe0;5
zn5Dt;=)?&U6Y`9zn0g42x6`+1&yTw)Jvy9Jm;3psBlwwYyX3zfb)4ts0J>0vT)t|X
zriV-rgpYn3CfkLCa;?QkOld_~A86}mKU@}N4d0)*ZpcH!t<yAa%oFtyRPV+9I_vem
z<KN#|;=Q*%b&*7UI4qmUq(O7(Sn!cs5uYT*WP&&mV7azu{PYVp&74IDV#Alu=(7Z|
zB9x??dN&c))hAe#ZsDCXf8JS4p;*Ks6m_e%B#>P;>af_ylRQ9z@M!i=I7CW7N-b;<
z{x)MNI37@R-#ig*l+Y)FgT)Vn7!ViAM4BK4Lc*s{D^y!ir-dP+S4=B}V#5o<^NpEO
zi|7M{w4$g_wk3r!J1WY!J4SG#j}{ghI-YNmICn%iQWmr9Ca$_z`X0<y=Occ4_3z>_
z(9)TN<I^{cnndJy#a2xh+-?LR)8W7@m>o=;@C?(eS{k@l!#c{h!)!!`Gj&YT(W%XB
zHUuA`983f0iC{6q;+xebxc97xV6VC$1HHnCxbu}hm1`S{y2Zfas;@_E_6b23=)dIS
zXg2nZc8a}Q6*Q2v+fZ_Ou)A~g&|BJTJ@2vcS*xm6C8a;L(+EMzL2tu)$L&zZeT|?T
zXwRtEq3zgf3#t2aG?Q0tVx-UYILi}GY6yb@h{R!bd66ssab8DVg2-dvXY64>G9DQG
z-o|0s_3Y_~_HcK*^U*v&W3BZLtqsE5_37d<Xnajd==@TKCGp(T)~AdawbN)|CSx$q
zP14LqAa_tOmVE^3WVwEYon)#FM=t-N?ki}v>FY0EpCJlx)G1|39q_|^hH|;aRy+Qr
zIY58)Awq6^-@e1;MBX84jducYm0%QvT%=Q1`w6yXe`;GQ`vCp)HrGuqi%+cG>V3hy
z1zkaHW9eC&*1E@)^TF(alQ$D8o?Q{!D>~x7tpy{wkBCTYMqT8EzDEdpRoi}EguXLn
zq!z|EDkx>a?E{P5arHQ<#Nqp`@yuN$-7<yZ^Uz?yDSwAB`8o<o4fr^z1O+L}&@Wsk
zjqN)-<OPf5haocgevb?vfXI9@fQz4Kz{BfrS-v+jvn3vKnbjD-oT_%Z>kq$}BqEFH
z-}DHtGL`hTk;SsHLBonEa#1-~IYr0DRUFyEeF>=N%L0A`y6a05)ZbSokEXEnN(KiF
zyo_%OwJR0LT`rkr^O`uc#Ls|eXRp!)wUD&01`X;P8SYMRO((MQv1L~2+(@=56h`$q
zK3IwE1m@Xb0V;VY$DTOr6h_{4FWcM!rGY--s4$2#^$XXDkkB@+x+y7~Rej3HAQN=$
z$w*^SPOeY1iVfMG{FN1As`3uDm8o>^?5u$mY3~Chdc?cyR(#!J8aXBzjn2|RUEjgc
zo%y-?z4hMKooQ*NDW(ZDa&RogzD1+y&4!nw?}<lbpa-pRlg`-?7106{M+{uqLocep
zj{29_+XcNDVO3_wY0d!KC7i_pBHz)yb#cAFT<{7h@E;j#g3KpBSauWpq1~y&&A@91
zvY=vLAzd`rzcR31hZgOaVBgFR+R`3c6(gFSt$SxSE)@cPkW{tlkM--vC9AILdk;yz
zpaX5pq{+L#A3wrcch)(avExWFEsQhJD!Nb`V(}=s?2QwcE|{gyfQD^;c(=MlCC<wY
zyi8e0;(P-I;hk==nhYQERlbH5CoAO+rkc6jerDdI5yL^GER^bk+Ln*dBL8u02lTlr
zciqv!{xRT#AK21F>DCQB0WB&th7!74KN$G}6ieUo>9iIbJSD2Y^!Xau_4r{QVE2jQ
zZM_cth4Jf4kLL4ntvq)YK0G{!<-8Tu<PxfYoZ^f)=fSrscsVnq4uOWQIK+{rgOOGD
ztCM3tP;3LrQ31ceJ0{iBrwze9ZOd8ZcUm5_jmB<<ksmG^CLWhu@K1mHm%N}5gmHff
zs)Zt&67!6Xs6e@W>1yL4v!kgY5;a)!!=6%Q-eZDipk-dQu)br05dn*-*6%iJEz2iO
zHtj#)XdGYMmX=jXA`l$6<N(~gb0`&C0!{OTkvH~@DvlNc*(ofa&~5D9n%GZo6$(#4
zU4(Q$(i^&-=he?~7`1nrqL)qby!H6+R@q#xvOsp9>VY*QjESJL5r=`=t7C6^0@h)N
z0`J{as+(uq;l<N_3Cp+xd_QMl0XKmW*oB0<W#OTpt7cwk%D~Ro-2?ZW6kcnXtLfmk
z1iuy#&4+_pBX8^B@17o108|^cZK6Vg1We!ofPe{lat;ZEO$;F(M1eD>e96-b6*^j<
znngRAJGK+40uZ2r*hR5V7cbaL)D{-PAx}5Ko`N7!&8on}&y2Bt=5KSQzFW7F^=<cs
zolKtsIv>`zVCj-*@_)O!DIuS=emISes>B*PqBSpCZczQu=0i`Xgyez@Oc+?vC38db
zetl+n6vFncLsz*#FzE){a(U=?U+2LBUnWU6aGhI$*Zc9>g|LE3?rC~__~NBhg)XzX
z#{XUKYBjE|mt;lji+HD$e9mL;E<SKL!53_}*O%dGlj^x$n8^1HC7_tRKXel@r+a6T
z*7)UgW)$qPa!Pu3KP)lxzC4B=9clmq)pX~saI3v0;^q0R&A36eFGFe2VLw;fDn}*J
z6$To}Bj4ZUzSTClWRXV^DnpaK3ZwHpct{%Lm7*zho8-vO70Gs>Z~vC^Fu!t^LV1nT
zJ=qici4gzCf`HghGf^ML5^Zf>7d~005CbTS)Z1HHI`E+=P$5<~fF|k5{ov#KGz3G0
ztK?ZiL~TmxVcYbKShQvob)oXO`PTB2S6?{vC_)3Q&^@ub>cC6-#bnMW<*xP{9#p1@
z)ftlG;^r0wzgLIa6Nn-c9B^no0%MY?q(tYj-J?$bQLHzjjd-%rmdO1x4ZrxYAOhvs
zx(olA_8yv}jn#pVIVyHyRYSzXrcO!4>US%5Q8DfI`#hw?Ad`xTM@yGfyW6AL@id|!
zu6gU_hU8rfxpz>C_wiGQxd%g3Z1k!rMV!7bou3;IH&;_uf<u_MIm6%vkeFJ2>NN^W
zMH{z~aC`_7-YW*;X88gvCDWn>vWV|tFfm2@6GX5EF{vJ0QN)^8J1S-LR#ym<om~S~
zHzA)EcgtZmVB3sY=CzuNf3>_1kJ32pX{zYS3q@TeHps1qg*55WF)WJn*l*nHb#C2J
z>~0Bk+iYC;7${a}G55p0@7CS1CR=ZTl*&6wd<|@~^SYj1SL+(S9Q*Ow#7Dm8<Z)!Z
z06ATVf!$71X(X;3t1e(SR>7IhYPh&&1CMTJ26Z0PSvSQzjylQKBci|Oz4x$(ZQj7I
zb(^`C%q+!fp~&6_-8`$62J4eTsMAl&oJkDnd<UML*1k74T$k4~8cZ!!j8+=;>{<rl
zzFW9xY!bxc2;`JF=x(;f0=;#H9@-+B%=|gbG-vRuB~WsU^QnSn{c$9>U*&l@^hSNs
zQGyU{RBHD|V}lP#mR+I)v@q6P(&6-4b#ojaM~<d}GTHQe*CIko)xDnUgosRfu_gw6
zEQH4nBFwIbxo|}1R`0eO7P7DCcVh8|-1yUm62F68akOdxz3>oj^dI~$dECO?>?5fu
zm!{#Ka7YcU6pd$ecMqc%4;IxABqD@tAs*pyPWjqt$W1Bf+;I|~uGfQBmb75InORVD
z(hhdy&C&*o+Xud@QDy>^Wg8?R9|w`kBh8(D^&WG!>9e~k0}UnvyN(dTHoa_|Zg;D0
z&8RI2@)N-ABH5D9-391CmC?Zer?%KiX8T^E8vOiv9c<@WgG1;IU+1Gk-oyr6Yg^~D
z#By+m5ZIOf<YQ4*pl;R8vqBy10mc*iFH-uMlTI>*Q&V7uQBs715uo%qt0Qnw_YP5n
z9NOz{+bY>1Dy?)1XkZTw$x0lF3awg_{Wf6oNNX8!o3)-?vc|&o=Xqld|F^EvtE~_t
z81(m%<fVhnXpJM~LwCzA`*x*mFUunV1?*NDrK@z$l4rVV2VCcaopm6J05}whVj&Hu
zEr?f&Um=W2=jQ^)OGTMiB+Xz&ZT1qUckU9kgV6#U9lBvtta(9Y*Y#b^OU#->di7La
zaXOTb-$BO~!x%}}WHUa`zwG;oFzzK&x!kL47kgCMIzQCo9S9U!cL&MR=upy0G3C*T
zIDk4Fcz#vS>>cU`B-2-6EZ^~l8Mf(pU6iaZtJp^yjE$Cg-c27}ExRk^5GhObk1+8Q
z*>R({Mzu!|RuGk*Z4fkWq?HTo6FokF;|3!KMcKQ`(uzz14zx5h_|~FVn)j`0)$_FT
z8Yn>MMjzj8Brblue7nw#l(PF>EnD8PUlj4xh^nOtrAfs@9waw63me1PzAwN8lsZiF
zRy%iZ*K2t9>sTQV0<}Btss_48c9&z&Qb*@~y8ju~i|L|J=;WKIt%jF_BBG<>8oP89
z;BxM0sY;o1-L}Gt%iYpO^~1?|@`TfGwo*V8V9+t|caqp8^T$<OkM%xLzd<12aZ(Va
zz7Xsp;9Mt+2dl9C9o(Ft#g~gu^UIztTFuSTZ~-?*T!R%lyehL@!ydU7zNo$~0t&X^
z0E-36)aCN?;P^_jWwMom^{?DF6_QgpqUfvNNfU*)>dQar?scaHr*l~k-%jqRaB7;T
zb6U?M3RZOld34zIzqd%|W(YY##pM?D>63=hlH+|w9-zQNr}KSqugNUG+u99qex+3@
z#i}tOX*?^l%B;i48U3N#F79nRQx7sP?iOZ7#Y-ZjAA*2o(1h~&wi({+q@Z4^T!K+v
zC}Kg?Skd=-`eC0w75xVpo7q$1-nMvl-t+TY&hgd3foJFiiywW+OxP=i_{UzLyR!xK
z2r#58?&OX@$|$Fv@1!9&<yS!hga2~Fh<!H9uSh}A?MeM~;2zTzvgw&o8|7NW8W_9S
zKD_L{8Z8es<=OL19`Heh5VLFLAfh?3<BIO9GxW_Nt(qj0e?i{{kB2U>SGM=V2QOiC
zzS-8VPJxO{R3_(LM6w8GZBu-eoegLyFpy>uU29jXjuZ1tvYWydt${5%M=}I_lfk_5
z%;4i=hO3fc^=R`AUWNXP*1_~tbPlpl(?p(kWe`utQyWh1ElcAt5cld!6x<zZ(m<%_
zxdSX?Q8P*Nl8eoCvh{C7t)LKzAZ%@pHp1;!FjRpQ%C4^tTSh(e{34D~$7|2VJrN8f
z!h_P8ZC4rL2pQjLZv2)fO`{GoOwD!Y4}FGQqI8<0N1|d#@*hmEJd{)H%}<Q<(EhQx
zvb$405FTiE@z^Mxh_Oc3+w2_z44GfS$BK@YN=)tr9YS={4X<quvrK7MK=ePbNLKhs
zu_RYvIfTi8RHXpe?;W4f$PVP>4#gb+^CbRXFa_&|G}wxcOg>go^X^iSPRis5qtt`6
zJ+bYZz$nit0K>|Ef!qGVLJ$*S{<|7)0%GD@{FhC1yD7j!-PC=58v!0FG8BbW!X$bS
zQ)`_kUcGw4A4v(sX)sMPlI-R+41XY%|AN+1|AhI{P5%>D04MmZ>j1R5c}p2^7K04s
z-R#7b%|yC@-tFmH(S(+2vqstt-BlYIAITpmEU=Di2TrCyx`(AbUTZyle1G%KY;*Fa
z&GF{fviq3ve}d|!hckwd=Lp$e2@5*kmaYm)lA*!iVdqD^3T<u#=P~gj4A(!w`G&y|
zI>tu3ia@zf_?uL&O*M1KP_<?gv8v@O+}YW_4~bv?scjeok&SP3o(EJ)F45n`bjwQm
zW5|C&VShz}JKH`9>pypEuv#MZedf>^PNJhzziyE^m?nufohzMVJ<9g|E*^R4AX94^
z2mUEPZF9deKI?eqX?H47oCZgyml~|o^|SU~o<Tq`u31WhV}ealqt&H+G~*#!xkOPT
zjZ1)2izum5okXj_xH-ImuObMGChjK6Y@@@Q)pgx#?;4vfAsH+=!`5u$gKK7rUdu&-
z!BH9UFdT-%*_B$e=y6`VO+R+)Wo3)`A6q-_CxJb0M|`Xnn?9>Cl`f`<4JFfS(7rz3
zWzQCCD5Y_7Ylh)6on{zz2LuH?lr_+6)~VAf7ZWEcDiye#Xu9Cs{j>7GGxf_4EZ9Ca
zeX9lpM!x4l*wqBb^VaYY=b623=8`jwDl@OIo!oDi_Kc64qK(&DoB?7yt^g5Tj|=YX
zdQ`Z#Z40T=?&3vJ4qHLEgAC8MS0%2?<^(%83wj6|%|We$V35SfeYAl6ZdlqIdHx$L
zd~krs91OSNIcRs?NLJJH>@^;7-S7cYTS#Q^Jy12@w!F`HBKb;`2=G1(TLVkPn*`GV
z_Qck+jtQEO%w%0J-4^FoiNC`t^6Ox+L;IhoirrEqKAY-q6v&!*CQrS|bus$%_GK1e
zRY{3%_<dz3EmbiG4T+le(g+oi`BylBsK`~fUgyh65!8L}mp>9I$k7B~YDTFp*eIjO
zgW<a{T%EHm!C*Zn-@8-(=`yu!E2(sTc4|5WkJVv{VXxO`R87d`X2&9(zNL`*nR0DX
z$KizRo&NW>Ng^R{xlN!?_-+@`+*N5(=v04<s#+b+W`nJ)iWuNX@=@EBatx_f;mcFF
zH+6iG%mp`KFM|;<@G+(EM-n&=e(-GD4$jUnw&Mh1xa(hoMWW$CY(~}Bf@ZZdz*3X#
zdYhPs-kMs~|A8!H$$|IV_GZh#<@Y0liFV>z)8}n%SoQ9T=ECSGyd!dt<A`_96>Ts5
zT5@vFk2k;$&$Fbf)~vnnxjO2vZ9l1nlC>SGulMasx(qE~q;`3GKc$FtuEgexn;rE+
z-)+!=gR;QcZub_QX%Pm;8a2(jFN{?$_^H?1<h687!5coHOvm4m2U@ii+VSco|5kdK
z`OmEr5}w;*$b?lrEmvSyIHXwJP69!KFZFDb>HrOwo>mN;CDnJ|*K`;hH7zs*yoD0!
z<l1B7-xl(JfXzQy{MCI_BezH5EJgCettxUEKBD06Fx^ykzU)FDhKFWzJ6^q`173mS
zI}bbQUqiEf4P1bWM$xgq{jC$9L`jn{{L83~r+0=l-K8ZYCI@bP<pB>(+H+S@nJ@+C
z$Br#LOW>U^Yt!_?v$truS35!ZNDlVdV$$e!w8@pJR|g2X2Y2=Gc;0?6-{{D6I`au-
zE8OsT6>&LTE?BeQD{BrDKM>2bk-E5{B8ny|T2rWg8ekb{xTY~(wV6`A!YR+-8}(+#
zKyl&s)Fp{Vb(aAjBRgf49T)Z|I&~knwbkxU3-$t{eD3%iEt<c3avgYyQ|+bj{sVTb
zmWP-~)1A<)7GRAc#(=8R7Z;9n&J*6uaS*95Pli`Dg7e|N%c$j_`0AlB9wAur9_%)B
zLG61I!HjtGP*seNtD>}SHE|jT3^;6HT)q;Lb@vPhJU#5`$zkkzy#r!|nl&X9r>3ed
zUybt*lzhz;U>#rq7G--+w`MpOPE>OjtCpv!D%(fsa})4p72dMKsA^++!Oe@Qvql~-
z*K3`Lhp%BL-%SovygTvQ#RuRhjh6iHqO8-$89ySln{_w2&@Ls7S#~(2p;`&y8K%?b
z&m~{3(SLygDKIaf6Zt|N8MJ%cc-7=a2qDx0v7BG9=>U_GRS<?127O^2CNl1>MEaR^
z1*u^HmLt~(@cK^egi2sQo|Ccq4q4R%*aehC<a<9w<8!-Y_7G``1iR`ZqF!oDQ$+Zq
zu)$4i%lVEM!{y_rnL^nNB?>w^%KP!ut4cX(0k6fO?YDnRF2Q5;95L403Jsw{7VOag
zCk{jny)YcQLpDPr6Sh$Jx@}d)*gf3<m@vas)|X^lnz)oNX-*Ab<InDPY<!%c1JYkn
z(F<fYY;bf69e};$z|1fPNC4xOxd+w~)A#eE-_8%{`P?LebsQZ4&QdsI%OuVF69U&x
zG6UQbKsEm#BG~^PtgEp3RG2K14be##bVkips7SY}zN2s$mKP|=L*?mveK2dNwLN41
zqObu_j@4uW`)^tYs2CO@=@aZS`*`Bf=vVw7(!|jsikuGUk8&irw641wDqen&_PlK(
zvmC<Q#;_c2G@t1|M%4C=3<&jUWApkRACf6YAb-alwJXHFC{`<A75=x}BzA$Htik79
zpcU{O=1zH#?2O57m53#sep)mlKt#f2jLY^(D~&d2|G?sq5N~!yP8<H6zte8D3ALWL
zjE2qq5(^xtQQFCHoLI1A=@FZk&RN8#@azek@cL0uuKDn0!hYcsQItSOd3?k}%9U~s
zA$)?$^-xlhYTX<-c!Ri`{;nDtU&qDVl}5L%X(;TGQ}bB(L_MITzkyz>VHB_f_?v^j
z)w~S!99Xt-#xFlm_O!9D``<1)K47aI6jyy~Lp;h7II!1nZ8xR3{Co=b0$fj1rexng
z{A6<Cpf7-=GwKcg<j#3D&P%of$8lFJ8X7mAA)$0m&%wl+LJV~R#+GFmX7XfGz~9#T
zJbRb{NLjV$OzJgi3+1?vosL{SH`EKn`fLU58O4`@H+`;4mH3SXK#a)eo~6JmdcmeU
zg7WBVFx)@en;Z=n#%m18o`w7^X%f7)w=+|>W<mq8I&y;eDC*Av>2ZP1GUU%A#*3^M
z;Y<X;>F=GFm*H7BLNkmB3bhU2OB}aR77&(EDa}eGDku1!zspyHV};#dkD6dM82Cmp
zG&BSN9>X_KLJyyZT0D=eCS}wh75S=<U?Mg!Y(6Jzz6loo6GQF`Q>@>=!4X-4-u>%-
zHeLi+m~MkL>d78Z&oJQ2E>}e#g6>+C1(4$Qekj0&2_sBk)6$oG`V}r;FE%B|1wwhp
z9>di6B!eb9FOe^f0vzXs8}zPt$dF_NP9_Z8hS7Np9#yu}e|xgHgAAj<nqwhqJHkhA
zP%Kc+rVyW;FBs;>LDyKr*nc{$d5Ka|^uk`XTcYr|9@%I<cGBKx-0%^xIduIxaTCPL
z9`yIJ{Eaz&fhVuCS$UCoLZK7?gte9UkyMDb^w`QIrVytuXrHgPa%RRk-0>wfc8_>V
zH;nt6$;e|g;?Z15uZ4NUT-ib~OM&!Usfu@AL)xWIt1h+gL`n(jN<FVGrPpQb`bLX$
z+4qv-%Hm{9A)?>;J4u`+#9$OZ((>$B<0@?m$b9E&0uak-t(o4-w-VgLsF~T#A*DKO
z1C&biO415d1zCms{j$VnFj3Z1QW(mZ2nb@cRJ9ECUwzTq_{LYqMt>UIvQx1+4$Qh1
z!SQo?!)wOCA|m>KSEP7i>QhkQ8S;Ul)W;I@Eoi%pHJu*Axv7pO=6IlprFDnbHPRC7
zJ@5@e>^S-qthB%w4rt1+TYgvsW1wo=?q1-AVclw@9T38;tGa|w8}ghpFkQB8b4|zI
zZK)c#|9#;2^N}mmxP<}U2e4R}uwJd25^2@ix{lB>WO%jG-x;L~y7;y&5~6y_L&2tX
zeV2^i0h{?_^8>hgZgS@$a$qRIzq!%=>Xcnn%jz8-$DX6;_q!`+wKw>V3Sek-0z>Bi
z$>SwQvjiI$5%qNaFXA$r%a)+wCCmB4$Jmx%;$D}dVU24Z;ZBEcEg$&Y6ckl*vKuVM
zO?+-Bii`}r|DXnmVTYL$a6s9!-9%y64cpw=Z_MmAK(HWup8Fpa;=jKI`7prQ)qJ`t
zQK5fx<1aG+tOsBEgY{r9zj391lP>?GfBbF}fFG@vX!K{+?(dQMdl7>VXQG^|gnzg4
ze-h2nqJQt9Z53e3Keot!3&cdfGhyNO`6&K4s{SJMlMI|Wec`6UP5FNY1|~EJ2iEDX
zAbI2^|JbH~5xS57=NJF~bvXXt-9BIqV7-TVOTa@O8oEXOmVE2){{^rb@X2^Atv)Uz
zgFcZ?EeRU#?O@BzQGrXjrs8*r{?<5f!kZL0mASxbspehJ=~}DFO2dg_nX0U}*Z!o?
zQfbuP=~~$ejZ$G9xPpqsd@%_)!?ebr_YL?f_7XU9bV=Q4KSSBBSLDL5)a;miyj1(C
zRIRdN25f}of)w-I?6X_InU&`>AwA&qz*w+}Q=(iFSK)boZZNklx76v|i6tqMP-8MY
z=-~{G0F+8p8}{uC#yYLtWwyFV8pevw6*<&6A90p|bwk4K?yqnrU<aauf_-2Dt%g)|
znQ-5w#|A%+&fvB{hFPtlpS@65qjqzRVvVR|5qp2F(SV$!n#mOIPBNp;^3Yvld{R>3
z@;d3Fa9AOh%64V3I#t=N7k)tzQBq==s&sv+%J(l*hAmIvB%#^cZ{G_G4;^jS>%?wP
zSIVyJGQ-Hp_Zt+I=pO6LHr|(h^Q%wZe9@|QDA?nuTVcM##oO{JDL;iB?05zlx3*-z
zExy23Tln#%I<Lla!KMT(vK)O?RTo-CLor(TD$dZIQNd|SvP7+~z0cJiH}7avS`w58
z+u$0{78!Y6kAJ6NmYqrG5h9%yw1aLB4x0#@qR057MOWWb*XntHBHn+G7XUjRtJB3I
zd0WEy`7Gv_3VH6gmI)~dSm#cPVqe_=kZW>_fMxL+Y6WJ;jIN^S4Z3$tXQl4l9yk^5
z=HqC4z3Um_QoCp6yDJi!!VaTD>D*@3a!r-&e*SSovGBu}Lh_>hr!`NPk=qmA44*{B
zP82-U{x`oZ@M8rdLH?P-@lD^~yHP<9Ve<~Ha{&v3dJqPK$7$d0JF#nwH<H}uH;;Wp
za@_|Bwgb~}&MI3%!PookXu+}FkG8M=>;<C~{~vL06;)NY@DD59Ap+7N-60}f(h`Dn
zw@7z)cXuP*-QA7S9n#%wYQwiUPkEm6-hCJU8^#`EZ`N9Ku9?4@0kh#aMW74w0kd2p
zX(b7bKp0m$X)3B9@gS0VB8g7GpIaQTjU<e)<v_h<cieJoeOTE~<@b^2laYk;9eSzQ
zty5ouk1=emciP5dp_QP3n8fu(0IhP{UPb@)X;!yHqc~hzqbME)yjgDX{kwTBKB>-;
zqq@??X8~{7RA@KWXv9?`OuC<-7}9||LRM>Ju&*W0(vnh{a&8>A!@aEfVvX?@84+jV
z?F)=P5R%<vI`gD)_e!Bwd%AOF<Tpy_WxniV7ph2VVb)mTFNy3HUhWm&<|@2o+)e{-
z?xi=lykV<+N#|jGPq1GrdDyv0_dq@mI0Hry#`~a>#H;2+yk%gU(`t^zsB~7p7&zZQ
zbC|+y+B(rSSU+sdR|kLN6digbPzkCO;8%+a_bt>;&MGwON)4<N^ZN-y*b(PN@eQJQ
z%V8B<K(7OM8K)FSdP5S1+zkZ0YX^C;OKepiQi-to`EPNOvcTo5!FL>}s&MfqCv?d~
zERZ!v<z-w1X2uG9J#|B|^l#}SOxDdsh)nEB>Ne*24Rsqp@|ksAgNXJ(jp#U)7t!_1
zLxB=|1?oct6#+)2%=cq%$L&UFi{*mgq|MHl(#+EIM2U(#bs1d;0J_EtBVrF*^|Ou7
z#8)!ibBO|Ozeo4Rz7t$r)MrQ}vjgpPCrVzMb09p@lawFP5Pu1#4tm1m!=J6CX>Obr
zaa(#z0?tA4zY+v2Argx|i%5MRkjjE)W#7XJuQTqX!=G*B4}WE`9?L8F`^*UE1w?|7
z@asmoQbyX|FlC&AHT-&d=-O>O3=#=Byt)(Fdlwn4Jxl%`>tzezc8&4{*>q9fV*v;u
zPKXk{^j!&k;uO=y<~Qp%e@{J)Ql3~NOB5s5-vxk4;T=^~SRfp67rBQ1Tfu&bzklTy
z`64}yicIzgyl;cFkN^?PJ$49G#T3+kEv4Q;xY{FxVMuN<0#nlMFtEty=?v0W8xXL(
z`TIg-Xp#IPQyzV=@Nhv%(B8o031NkUU_==m6SE2USp95H@!9^F;(P&M@{^-|HUPd-
zK7+4XSs#wI&VC6S0Kxz(Zb6i=f2QMXq5tEjfIoa{cOovjkv&{=FzX+;9RC=cJ~wTy
zWM8%X*n+Ovs{u1$s<cUAS_DaE0{arVJX>)NM<yxz!@0sbSPsfbM*4ru(yu8X5^aKv
zzK0s+yaSK;8c`zr$A7cQu&YEhcCS?P^>bljDqS<^FmAcop_oW+WKiFB{zLmAKH*>G
z;+KVhcOD~O($#6qvBr7*2Jto8Guk=eSJ7N)cc3`$K`b?aN`!?G-Xta{GF1UgGtFPc
z?r}c%S7FyqVqjp9bC9b9Q~Q-GSL)6q?&wPRlUK(&vntVGk9Xz$j?n>Y2kR7|hL<St
zMeb$kn7z1!shtgmW3iM$nKuF`v<4R)UXWEr<AmapTvSn1$zDy~1Nt|I@;d=tO36c-
z#OnW*>;ZK4Q()mT2`S-DR;tx5ph7!IUMp8rN7{5B85>J>6HBKer^1d37yGk20A*(=
zRhX)JvHrx;!K?qDm;dK7C-Q|60&GtVdy*`J4qJ(0Z4qD-KZagX{#^I<=WFo++cdj`
zb>_cn>n}KGlg+CFbHryT%=$i@fCfqOtC#ZHM5D#iV?h|Ac{adx#?V_*F4r?|x_>N%
z{A~X903d{|RT`~xD`JJI;^Sx(qI+_XuotKtLPq~-2!F2C=AgRlqn68Kn+%5I#Y3=0
zh2#QfO0~+d{OB3jij+{BZH^HR6x6<s$5$B)OQdnykui&Ozt_|FE-^&^{*_m>N|qq>
ze2tYnztTMLgMmjd>?OW*`Gt0wk%UC>;ZoxyZHn>#RKhE-3ZO^VCL{XUpUfXoB@|-P
z0^+2&D;0Yxll}P76r1(X##OSC1lhyPHDMad1Yta{P0wI4M8v{br^xtZ)t}|(I0YKp
z|1OKC6tOX&jNsHpT9T41e8GY#FQ4mM^#NDlZd)p48rQEm$1&)YT;3;~m@QK&JeOED
zl$jx5sj^!t-?1Q&1dz{P>s4&pqKh;Ju8$VuhK-_Q8RFa;gL%HCeS{m+DpsU#^-!jt
zuT-H<Xr8(Yietr;PD6kDM4~LiVW2Tm7)zrdH-VBj7-efwzXcu=cKTjr4ahyP%=^%&
z5-QTo<z{DdwrUtlWd3_eD(eXUTBSulh)PSTuw?rpmw=lp1e<>1vxL_sziKkC|2YKo
zapaH|@G?>9m9c)pcu(!V|B`l#h_zeiickbOvZ~%-nZ--wJ*6GiPacAOu*J-jZ!>?s
zRb_aE>%iuQ6k;&E5gr2sFmT-(i+o1V?X;nJT|rP{lN9}`@oEA|w0^x*Etyo+>zxF}
zYe|{}6x`$gTK~U*{}{wCcYD5-D}A)k#4)m?R`7ZRs459S?4~~)PahqdN_TZIJ&B-S
zRAtVf(lljMwz5!rU|W%t%=S&cCTmIQDUDTh*6)=u;<klgky3FUW-4H9hkrdq`mMAd
z>Z<&6xBpmEggs9=7y#3Uqz&BszR1NlD!-<h6aNf1;qyMcpw1iqKS=~mi^(`83>W|E
z=3>NeymHPZ23fr`nLmj5D#Ahhh*e-HtdPd_3Mqnzfh33itcwxn-C!sczr`#N7oe7>
zCP6c+RI=5;ukwMGM!4q7-h`a0TX4aJU`3xERh0mB!n5I4IO_o`%o~#__M!^u-`z3N
z(*r_W7RJnr#YwEON^q@fFV12ppGcg&dEsapzvGqaRJR#sus(p8)!Jc5?Dt8fT)K|K
zh!T*CONb*=RQxQJ2X^tF#CUSmNrO#G5FGJ2Z9nuYeyZ_Gfb1+uv<zXpVCoFLVAQ)B
z(`o;&(q?ME67z<x3aui=73fHnOA@e`C|Kiao@RdCjD+`n;P_MhmslCP=mHH5`B=Qt
zRwQ_J=2p^_yy8r5*lBL*k?9_zCoeFI#fs~EZmd+pKFljBnPSEAnIY<-PHUXuA@aT#
zvuiaqkP@9EaMr#xVo+wgh_5>&XA~Kv_=%uE%o2GCy&TbP6@R0Ce-0t*<dMF$<XNu#
zNlS(C5By}acS&BhLvf6hDPLk|t`27QuCPFb9I<NkW*1q0PDz##!g#VbXyo=lP=TD9
zu(Kat2|ZUVX`?#yxtFyvkW2!|P3^G5Sen$QeYUETlfUo9kACB&+V;im<On4z<^<Hq
z3;2B_SdyWKE3Kl{<Kp>0{oCQHzSaYif@+lQc%yGNJUPa2EK{_pblHo8f9*~^`Q;Qa
z17Q(Jmvm#|$X^AMY-<Qg#L?w?gQsRL2xg0Og<r+p(?ML8C?Mol8h5>dS5j>#E&F_2
zoT#TFOfUaSk>Mxdw!^W)Ru;Z)5V{yZX6O1Qy97~HC0;lw*KUzlxG(7LzojRc%$*&J
z@B<@Y4h?_9WWx?7e0?bfC7G5}IMC_^>`q>{zjU)a9!;s=y}%nb%&hx8nCg%55<q8C
zUC#@LZZ8AzXKY@^a1VE0yh5PLjw4+tPSB6#C>SEUr;buU+~Km7txZVXoFtqec^r$H
z7F;<br*Mv7OEv-^m=t?V|6hb<9cGScRs22%#Aob)e+IOm)yOg|fnVVEUqfgQtQRou
z9e5ElAnLllVZNVBc8dupRZ{(MYp-(L|NbSxIxGSKUBn|)y<*i6YR-*8EPR>zFZ{2t
zGja|tAMEpQ59C{Nx`S1K{YxFwf`n$pS=J#1CK)^<%)dhj$YjyrZu}$Z;dp3#-3nA|
zgFB$#%!oC#z+l7*%tOLXc#DK1&tUy;Pgeyn0+A4Xltr)TpiI6Lg&%2Q-*{=W@4M)6
z;w4-wBHnFgAuZ{4S33Gn@|2jUmZ={yJXPrr${Cf;SoRSZw_1*^m(yu70eWB=l)hb8
zU5^djCJ%^2E*_(?b{3CEh$p|q-?AZ?MSxY|%O-7Qz8V1aS<rCmAJeFE2gB698^#59
zJ_0#N22YQW3kx_$=H0@$QEtP<1D($a7FPO;mUkd9h;IOyeTgg#mpx9Bq4ATZ+Yxf9
zY`_GJw;-5DFRw(sQ3AXFt|&j%xd{p1(Sm9^jU{Tv2KRI>iQS`3)!!&RhE;d->$CsE
z;9#ayI(a^c#h@tlXJbex4(klr)*_Zy{*Q5oRTsXmMCR>akGo`O+M8*m)!p|-<LMYB
zUppQE|NZn)iNecXgr<h__GPV)?z|TzHk&03I}(72Ye8k{x16~|iZzd42A@!A>$p&b
zS@Gcw2QP<2lL6#A732Omuyy^Ce5nfXZpqOm6gU9X<#~UG$B;$)6A3T@!~n8)T-&Z2
zl-e5U*)r@Q<XJ6B9~d5DgD+5<*qU!BMA^;p6_rR!v3JS9cgv(fLE`5dsH?lWi&L-o
z0BP5>NoUgmfrqOa4$CI<p7o5%aS1mcukCK;F_2p{kE!Eeq*$U(4cI-cD2=#QYF!e{
zF`}|<NOv*b<P~1oBL1gntL;yZYJfhIjD2-U##v)}a_rY{uy)ToU`fvGcB|WbVF~NE
zRRzEnEPMSYwO3dOVoz!o37}P?x7eDFUN!Q;vu5?Ekr*-`1TsHKr##lzm`}?+vw+&*
zsb-j!?<B4lNA;>4?rUBy1^{CTWl~4za>&eb*ob}UezDy@w38td3uL+sg|VwsTD08O
zjmR?;?xao%KZDVlg#Bd>7xeXpJ}!%;;M~vO{hdqaQ2RnLsNj`ZuZL)z#Q>t%ZHN!_
z5aR4aM3!CJ8Nu^wvm5hzUT0h~T8_6|)-Hl0FG-W;>)qv^1hY9AfB{X|1)vg9hhQ;A
zVbLcK-)fA<d=LM&R^ychV9Ba&cXuX@=R>smhjUhPO}563S2Jpx2^k)7CvEqds`bX?
z383Aq3CTTYzNd!}SMXJRl~k#n%d-Q@;92YE0tUD)mgDY-%V-XxjOE@P?uu6tL{F8U
z6s{x_Xy|veekf=}gu>bILTkWo^m~w2NGr*0bXtF`({9yZ^tXftIkW;6U~~2?5H`Ik
z9R`If4I&mptRlTy5wk|o%ppCa!-`$K>MEPUaWlwV;}~!^#W4qe2H5?nrnZCbZ5YHu
zd%LGJ`-MX5`P^l@6#u8`R+qs5qOe&>@S5h=iJTxuji+bS%DLj#uQuOLN+8y`%f_Yz
z$1rG<DF)>VSxwhHgJ+jP61P)u8<8`pNn*BtIFnzm*P(aSjvegb_(RA!P2CX!iHKLw
zysa?9nK^nFl={Icvk8dq`MJ?`wOZp8lWO}b$oi)+Hq}XB=$TOE<428ZAfIi}ie9zA
z#JV>2i=;wh?`dMbdSFW7`|{>v+}0VEp>^^pfSfPadXGPQ{-N=B!Q|4dO~&oytTl<*
zChv}?kowsfdSW*PZ%gEHwx#q9r|cug{o#X0r%#*QUw~&k{14&x@#UJKAUA>SC043v
zOJx68s?UN=S`$(y<H;P7v|&Jv=3x%bt~$z(Dy1++pt{MyJXmBh#6(9|x>M2l4N?zE
z-zKpd?zwWSx1SH0SOcW{5YS5~W%$4)_LUEpljEBpzQ;&+_N&>8HSNY;r%$aKL&Ec}
zk>TU(m))u=t@}klqQhFh-TilwFhzvf-f)&ejA;*Q8i^|O602_k$8GC<DK*wya?`C1
zY7V7SW}nTgKlFdArJ%o?xOxaNxNJT#W+0;8+NvT{+P6u0U+E|3$H`uFo#`9;hFOAo
zWCyUq@=`oelV7g~5KRVC4hfR*<F$hyn-!<B5?9^E-3BH$+IN2)7aax+S+}T{wYOzR
zvp?}%&uv)udp7G&_DL;tA+Z}$ae6_ywO`e7>7U0^ORQ(?p9&ICy?lI=eBk-$q+xNN
zT49~xvT}QV%?0QdBKtFb=q)unWEkF^50?)Cv1qm{_e7!7CHpgl1X%1;@oU%Ba<cWN
zD|8M^R{75o%H_Ou4})<{z`1Tyt83Hgw_M}mka-YMaUz12hBwT2i-)LgJlg7&z}!yT
zyRWSpe}!eH16K9WL$+O(dIB9Te{Y8D#RFl?aG=;qzef4u(5SleW;IC59V|toR78i$
z{Xy)w74*_A9FR5)J83Zf2&Q}G&wCbo`DK+7bk|Q6U~>O}$hdbVKmx0_M8s+|I$=J4
z^BB({uc{T{ay;6QV8yTm3vAosRH`4NBpF+mtmIK`iJomCcf7tkisi69Pk)2yRp<lG
z6NB-+P_G6wAS(3v7YICB#)DMr%}XP87?sy@x`%Feh}wt7<ta@L0855jnGYc{PO1)|
zaTB3zd}>27$&(*kxIGIv*<7?}o1(7pl}dW>LGsw4cjF!uQo-eM_Fc0%ZZtYzd3HTL
z`^R|Jo(gwdyO2SEuq950(v>cS-6D#JQmSqY&}s}_p~n{j{N!+jQHG*G%n-+I1S$?4
z2Bmw^>ke3GgdU4_@bRV&tntB=kbAY^XwDNoJK(nPpki#lY;kY6rcQJ^y_A>Pt?X3w
zb6)iT)!w=0+O&hOxmw+}re|Y?%d=gtwsy`t@4rt&-t~TK+xgJSjiM~r2Ph3Lh+I$I
z-9n!LAp*HY+f|*q`>nzGBkal0td`W&Q79fLO1{s;XsQ5Ydw#j+F*KFoK6N~vr%r{`
zTE@OTxWFPp-J|ViayqtXF=W%7J~PxK!zZPHRgGx(nW`Qx&xg4yS)o!_I6p|Xq4i;?
zSWc8W*V3l%ra9fAZ2P5f|G<y_fUo8bD;)eFSj9p9St6_hlk%hD;#J2DWNU?S+Bm!E
zC*=_*hK4oQ+oRb+X^j}duO(h(&4ol?4^C3O_DhCRQIQisK<1Z{`^_U><MCanQ(lGY
zMzR~|Na^k#vffg89Gh0Ml|FCDSB7^2bO1Q9%gcla*cjbfzetply%lYt&v3h|yhiG9
z3&R$V^xGFkJWe}!y;`Za5@V&ent-sM4BH&>Oj@SxXfz}KvqlJv1ke<?&PvrW_^CoQ
zQFd>!<ho~H712zw(zrP9UQP^DFo`#Mn)Uu5C&BUeEI2}bS_O&e3)ZvK`35T?jZ%>?
zr20|tJ;>!fq$E6@gQKi}Hta3Hx?Wz_Ey5JQuE20Ip}tzu`{DQSFe}nnCb(9HI1j8}
zSCXIBCODtdXM7xrXO432bCm{k*~ZalG#HItZS~UBtATBKsa$TbA$0H8@+zz@Oq`ii
zjD>oKg%5i<7<Ti}5<yQ&Dyzi(^=9AepB?fmzxebF{8@s}-hbHtS}NPEFrU#eirE<^
z4M?r1zc1PA7;tUHhMM?uW-%dcIgh9Fh|*$M4*J%ZB%EbKfMj6p%+*aA5QMzE_zUaE
zznB9WrC8@HyuCBMF{al!013k(^&Bs8LR9=(6wceqkJhVCcfs4G%nY7=@39Z}uqoxv
ztL||Or$E|ei+35w@K8k;U1MlCJ)Ub5QtJFbLYY1#aN%q8vIoWAPfxzg?jZC;qTU>P
zupsJ049Ls7#6KI4ZFT_!GCA|i{d@O6T>}6cFwS)DB#uT-uT^QSq+`%8a_^OX%dCK3
zGJm6}102tR#T-I)Xe2L3ioMsez{kxeQxfQOLW!u*oey+x6C(`t=O?zD>m{~^D5&=r
zrM^)ouC(6o!!BTJxk`_VcZ;8TB6=yD`;IsT$`-6eV^@)pLD4%0$T2ZzSmq^5UeBnV
zQ1OC{{az~$Swfp9?t+S@i_?lKpae2Q70c6^^8g=EQNZJZcL$FMjS-6oak4wu{Kw!M
z#wB+!ye32rP3}qy6+cJA3NE~#4%I9kBV3*|SjGNmp>$`}Y_8&S2E-91E%w&&GUHG4
z?p&6g%wNI+A0W)%z<Qg9;U=y7hg4G1B+LxaOjNzyc1|vhLBEsEtV~oiomGwTFKw@c
zX+G`?!yYD_R+tf>^V-Dn*g^V&y+gqKV21f%T=>9dmTa$c<fLG+wenbQQ{L$v#L>Ms
zbc8Qq(;HmKoBtlP$+=ah2GBgqSel;zVx)-s;*jZR=pM)?+DkyNv|(qYALFjBemJ{l
z-1i$Y<=MeC-&MATUnK_u5-Mc^9RI-nfWK%gFIAa@M0&k{QplDlGdt~O!PclzmLMmx
zIy(1{adwOw=iA9DmdB*ZlUnZrHluPPMRZx{U^K!ZbbP3Invp9X6a9Bk5<OS!my_{P
z;RG@m4N2jE`9<3tC%aXUsUJ-ddnBT#imLtbu2{PMskoFbn9663<Cw$Ho&&@izgD2o
z+?L)9!@dZE%;J5I&*}I><K5;4AlndCV`hfM>{QZp+F~265b(L<ghRw_5{7<@3q(n@
zq2#RtiN<^UB5)MGw;q;HJ&vFGB0C5$%eoT-f;`zBf3Uwo5b~H^tnoU+N#oq>UOah|
zsU^Yvj;dSZE+>YCt^&KTu@y~O3q27Q>(ny07@-VG=@|ao7{lwl6(y~=O2sl%ug0!W
z-^Xk3T(Hl0&Vo0QXs^Mj+&MdzC~m(7wezjcWe{Ysy>j-acqC{Y!|>RwHHXN^DQJeb
z_nN|HH9dqb@VWI1>I}{2tLBaIxYEWy2obGX^*?QHL{JU@Apwl+MFk_Wd8kEfo+OU9
z%*GS1`vt>J-q0najdP0z1Idt5DF!Va5}z4`yqEjo=-#h+IPBTq68#1I0GK&ILF4ev
zjFG(jXe;ciFrv_-&R2pCM08-DU#pcL1c*;^XQ%7JV?f>WZz?K_Y2SL1o2_X$>=Sc2
zioKH~V?LOXmX>B7<W*yI5SsECF0v~A_)e{1@Ye<C5(n^BT{PJj5%as1H)_%*zkwwE
z*#wQnr#y-F#_;wWQ`$btH?++uU299A(DgbgKUTL|7FN<r(Cyv~oGX<Q+E^$pl*E<C
z<Iv>x;F*jFOO7~17|Z*SlYnd6%`B<)<`?4D-9g#3jU+l5^aWB_X2cuDSaohspEDJs
zw8B_$Nq`k&Ti~bQ1SnZUEtexZo`A`kpa8Ho5?I4x5~eRSFhe=n5Xd0rWcr!T<8sxa
zvZ1;xq2dx&`&!Rdl4=R25SuuGMsinPhUdHY&r8MlI)w8Fn{k8^Taiy(#|L{Sci+>g
zO+wQ<`@A`~x#t{0fX>)-OyNzt)LbcI0xbBKM(=oxGMhA-QVZX9lis}Y7S+)_O|fXp
zTdp4G0^xV!YmFp9jYA`1?HTy(&y*h6n0M=C1-}b%L1{-kX=j(H|9R4PaAm<?Kf-1c
zra%;Bk>UKU;7hrp0_@8!MB0%o5#JSU(AiSsMVz0VmkiNweG$jY7jVzB(W(mToAsEL
z_VoBBzDXgl>?!X4o4sN`(0<Iyez%=Vj0U@XIBzYha-@y?bLgkdI90ECe!8S_LMQKb
zN@o{#P~6b6VnP3|#hzR)zCmX(p`-3tx(ZEC@u$ppbUR*boZv<pn4*I?*aJH6701yX
zC=+!TkK^*dM+A_Pa+Dw@9hcAb>FEi%xqp_XMKe9qil2+96^pkOaj9N@b!;!Kl;Hz7
z;kbDa)dA5waAv<YRRkdQS|2s_;HP-c1#wI=55n(XN<Mxh_1e$Yx@lW@S$1f1bI-77
z&;hyCqW+`OG)jqczuT-yxiUK)R5a^m%8m4}oNu4e@E<h#OgghQc>YCab$}B|uwHdU
z>AYwYdilGU0JDleWh^r$H*<urE<~_UOGCM{k*4HbV*$d0bu{x%jkQD)9%p*Eh1Nmt
z_bwkIY(L4Oot;H%2kVgyH7AgoWN^KC9vaLcLzlNGq#pzs1$?Y4(r@vqm!&ho{7Vqi
zh@d-{Fml=Z0YrlmuR9{(IySIHA+Rz0pz{0VES`GL06y{wEzSkF%e0H8h(hJ2!HdK{
zxi3=6%)H(n|2T$ok}5+wj1HmwjR-ypQG+p|(5=dflTP@L2hQxKbF%63SSRnl4W_^#
z@5Az0Atg|K(hFN3?kW3IFUy}r5(yA+uuTJfH-HOVh4}0b@T11dt)Wifd_h<-BlhWc
zvSAbNWz)bm-dzKSl_FVe1I{bh-{oL`jP{X409dsRYnLAL4^j8WYW&ypQtCrX0sgDM
z#+Sc&9@uC0P#|sRpOE)I6hdAf$N=383L{Ox@0qp#xzX)2cj+#>!|`88@w2=O1zxr@
zv^!*o<o8E^2KV{h(K5Z0h^CyR|E-$|6oL@9`app~4Flf-+W2<s-e;)NVjqFe=}HD8
z+QIO*zBrQq5)K%D2R=_rO6({WPE^Z}?y`-<ixDRP_Z%_kFy&?Pnoa=?g5P3^<;NGE
z?@o(K{@f&eCAn`<YfbG1JD~p0odZ4j7qV<`Ep8cZH)pRV?)Ow!a%A)hlT^9A<CXoO
z&f@{U!9_g#_%`W_t@Av@ees?5apdHEv68d0j=#kwe|{So{pF)7p@Jge!Lst0sRp35
zj8NS$4TQ3e0_t1ccQ_9c!&8lpE}@q8!c1LTZiM_V^%d0!RN0}j5Fg|Ee2^dnbQ<t;
zZ+E|64IYU$^I5{*@FOCH{>l0G+C^))$tmL06t-2>qa}-Ud4A1=YIXR&FCgx;$H@PS
z7H0A?w39s04)l^_PJlYk;GgiSsd+eoc8L@q?%I7X`binT{VMI1N{+((TTvDo?Soj^
z`#mwxqM{-O$Ffk(9QOw)>E^+^fA}*X8k&icm67x)KbPROI2v_y=3}?0IK*R;TiS6*
zckH>;DEH97l$ql;tmH@uQOTPNL=>-_SbdRBI;w=Xx^}ekk{4=P2O$g!D8_^cd5Z5R
ze?}L(3V8Ln|IYbr;(Fz5<{<f11?GN%QDBP0+7(6jwJz$Faj)?DaOH!k@9ssy|I`Cv
zLp_K7&ue^^cL8T0JhRDq_ji?`e;)de{vfLj%%OBl3El*MuS);;muK4`(75Z$i2#+{
zzy1r_A9xr4&u=7Nx$JhG4&@3@r(CN>F}gr!6Mr0wyoj<pLbDYzRCzTl*1sW_YgH=O
zhUxgP{)xs0)W>)KXyh$+uUQI)1dDvLk(sbswQEG)0j^>*gT2{W^E#GKf7Zlb4H8(c
z1LI37i5k@lC_<dp4*5n~2V@C~p}!EKzZ7`cGC<$IOseYTrg$;DHPXu^$V)qvjq-aF
zd>#mzo?-I@O0{7<fWF_`!on-TEAVN$r3W~Aj_)z~`!cEvoe#&-Q)3zF=?nl)7azfQ
zt8NHp6$NByujA@Gp1_Pw7p3UvwRUSAFP)BWm2BHdzY&Y&b_bvy95eu;u6A;zGF5ue
z3aCPT`hYWud4s5Zt7~u$kdg78bo#7-WEwV5W;8#zA=cB~JOHkYb;aSU^V>nPxi!JY
z6YtuW$K%}h+1jaq=WN;C{4(i3o-=@Wlm<$-d8l0d0*#-ah_SKQ1v2U368_G|jqyxr
zw%F7>w^XfO;P{PY#N*){dK2dR`a8Y)`<|F?{nN;NgTZJ6d+Q(I+~Ii1{4tHKiDBo#
z>Tb>06{S_;(#3Qdj|UF=uNuuA`_m>@x1huM1G1-gSJ*fl`))xWaRPn?G_e|qh*sKg
z#x+7hqt05N{mMP2+qnS(L<T)<cAm|eG<N$_))O{WkSiZf+T8AWQs5nyIu|au#i4>U
zXF&l(6vLC_EocHJjcVlOiJ|wW7(f^#Q5WtfDv(A6%+9zL)kzz>;!KGOUdh*ww?MfY
z($1SeU1mO8e~jY~&+5z!WE)-27umf{&<I=P`CnpQ31eKr-fK$aDQP&U`Q18>%$=S8
zyn(N>+Q^G!HU+|}%nu!!r%Kh-DBK^j9<GN`&$}HmhMjioqwvgI=;;pD0nXy`c5-4O
z^<tIh6eWps3lEV0*-r>A7b=z(eR?{e1T^0Dvc(@;mpq<EWAMrj-n!kZlLOJ~fzJ1u
z$XLEsE?2_`cg}INpN!@6A8BkJ(^7VXc&lwh!OH+BYgLxYRlkT@KiLB6BJVqY?#pO|
zFGZq|Bd{g0nIH_QUp1@k;A_qEPTh;S+<hnK-P5#lL)h%@tlFa8%nE>!1oVqreR$<j
zQEyjcfWs7q{IKogY{og{fxz%Qms-o8x;Sk6kYJ{Tyeq4TCOb)-;@4ftY$gDxLS#Sm
zDUs=Oo=gTc0E|b=OLKf?Xui}H2dpc%9*=Aq?vJ<Zw^#LiwStZl@JBf&v(@Ar94dF}
zHZcr7PyRHLh5_<2(ND;`Cpou^NvC@sy@a=Jbz~#}NrP>rM4a5#x&UZsD6Uxl=Je<J
zCPu_J+EpOQFYXC^qUf|Sx?bfpi3RAtN-vM?{BrUXUpI8|#u*RpJ6sS<9N)aR?muE5
zR?>3Ir_YMux$^{|OK@4t=kI)xfG~1!i-+N-ftK?v@~slJ4MAM)=@2bL+%snw^;UFq
z;oyGwsQx+zKq?1>eEbJk48S(Sf$Qbsa&pnfjC2z1g>As?WX&9AkCM&-6-dI*6L;eF
zhx33~n$oga<k`d$Q#BinsI&b8i@`aojjd94t%+HT+4YO@_UOh7EU|;9Y6n8k&zNd&
zc;QJR`3tfnPK0v;Ju#t%tGzsxKLa~dv`Qs3E^jyqbWFGX>MZT7!c5>Z)JNx&R*G2P
z>utVaI`-iieIfgpr<VFghn=yJwA*clm8W!8o0NCa2b^^=oOa)g2>|tGY~#h_XN#SB
zLprWiYaiRgAJ?&WQ7_K!;+%*poG!2$AFAF@WVxvkK#se=J=q+T-@GU48$8FP-HLvX
zeQ?95-taNqZJQZi@j@Yk=kaLwgzcTfM&GFa?Y#B)vV+*!uii_hE_xDErE*Q$b{VkH
z3eH2`RFO(I))hzF`|cZfzTpCkwwSrOH^wfban@Uy3F2+8w@Z2}rHf`P=ikS3taF3U
zbu)~hG>>KY?z=*JZK)GwWf_`IT^n!nq*Y<;vU!YBKWmnwx`f{o?UDmDllT~4xgUjr
zAN(7K)(zf0Lqhyd5601t40vp&#O#RyIDjPL=iT~z9~+<PE<f1q?b+;V-u{P{=_iG~
z`2*^t>wa6-l?#J5c-j-aJB8A3nEJ!oUS(YBtbte^7Cj3;bchmH?NXQKOGA|W0X?_8
zf1Q^Sz?NUGzxv4NyJI)2L2JZTU{Q>;k6V?*b5B@y2iEWb)VZ+^D{eB|PZ#Bl(li5@
zPmNy~+8?4lXVcAC>;ITolPLL74Y%hj$>4p{Iv%IM=i}afEN*>td*;5F=KN>i-nR0=
zU1q-h+dA{*yp;~3e(6X6(%R$h&(GNVej87xQRZYRG<$XJvgMKCL67#m!OHEb_VHY#
z-OJha`mlZL*oAPS;zL@u0hH)!!YY&Zo(-n%MQA~#zYQ=t&AVSU3#&1j+hwV#FN~m-
zCgA*ptb~$WKG~<dQ!e@Xc~AS}<ZGL^_dmKLhJk~ZzQ_v5cAK}feXlW|lmOM6FH*qO
z`^<9^dong|>SE%J^Uf-9rrd_1)e|gez0L4At3{{Ld&l>fsYOj))>yxjHmhF7lCI!d
z_HIXT5tTrKHJd`w{&1;K_w{;by?vWNFd4b=oBdeo$r=XVU0%~JIZCO=<r*+9wN_c`
zxoyUPUcP?NtBKrU`eX!w+3Can{(ii(6I|F-jGg%^yW%A)BL|gg!?bN-k9FZk=czZ1
zb;}Ak=BGT;B#f6%!oupqAq<h|CmuO^78f7x!HcnlG=_lAFAT0((>>|3`11IoC|D|m
zNX_Fa>>VoaUCT}KJMM!|jobTA1qDq{otoYz@csv=vkU&ncw1#$&AX5&JeI#uyL?*1
z7*RHy!Kd9aUq+NzNMZFjUB=Yujr;L;%ykK0X)jrh^=CLrd%&K?rUBb{B!X*qe_6E}
zqv7~c%CX}G{(+<UBLqoQSPUljihjL2?d%Ks*#n=f=3{q$rYAJRMfb~HW+gyV1RR>R
z>+_Oqt@Gta#|;tWD^+SU!VyPRrrER$cHQB(J=i)~Op7R#LMeueFq-8pcX>e{l*_ta
zq3RyDmIt^n*j2O~-d2Bq<ey!0TXq?W&K=$y1L8dZ3~6vfhL_RiF@5J^6>Cw^{ExY=
zO4^ffq=Z!;nL{k|(;38FxTFzA-{?XfNmaV>*vXveyx4n3yhS|5*XC2}#`x^!+0NST
z?=Pkf1zApziO7I{5hLo%v)AVJF=GwWR|z(&aGVe1Tf(-|Ee-U#Y{>Mw)zCzrb^q&-
zanYj8+tnSdB)oy|`&UVPAQ|Ji8}3n7(*b@C(2`*j(eC)_Hq7xlA!TdE*aMr1B6i>g
zPk3XVMzd{v^T~eEJOc~i3~hPaSl~hXmz*vW4rPDI;dtx@9=)-<L*BdDDM!o2+SZA{
zDdUN?5$E%Qv(T=!BOJbUY@6^!GkefAUtfw2vNzI9`x7k<lPHb=Z28Wb0C?&<q6IPc
zCn8)Ozv~A!m+;_eqPBZE8XQL462($nmbFp`+tmx1C{po0`8kOgJlt_C?&6BhzHW$<
z)T{TkCN6PH3tQ%Cv<T(pwV<$i49%||_8y>SB%k7Uh*&CjR)>y_<Bpv?W^<>gxvgh`
zY8D3h2MdzLGfIwz7+P2BFFQKw0|q?VwX6`IE<!=~{wYmGXgv*`G)iS;LIEvX*NI#*
zvTd#$3HkH=H@f)53_8Ojx}Rexz<IZ31TuVA1?t}jX*$+k_Cdx1=Fu<TrV8?m?Lsp>
z=X{J%b`bPc40I1H47nTM`lHU2Yr9_1FK^OrK-bTHAKfu?>Ac+mSo4TzATg-99CO??
zOVa^3%VaeTsS+(AKaN^n6$aT=$OAP{yWt`Dn&t;nTj%+C$>c2$?G<bj%(`)}4}xQT
zdDFf%5B(!@v1)D=uPeg!DwdH;;~yIdH13zqOg=`vspw*v0&Gfe@~VyD{0P|Ws`16R
ziii$f{QF7U>#N3`t(~9ZQ!IX@9GWa@luCI8AlJ?L%5@RjT7Lhs2#4oro@&U$n>Y?F
zgQqA{TQ@+*GC3EB|HizFp|Pa2QRjglpbzQm9qgRuO;=>=;i75?1;?vkL)ZK;=`7{u
zGEBO3OG$hy-N8-fsJT+8HKi)uwSgEdIe>)VrzNi%OnOymAMNfqT6l;oCmaRsH!M13
zGGpVt^XtqkT+N`!HiAWm0qbr=7)F|&M$7~;lb@s?(8F7pN=gvRvALUP&|rd36@9N%
zx}6|`BsY1UT$Ktaukqe<%M*PdR-ZREHsqx(BcipmLb^~fODte5FTabYckoNz5O~qB
zcF`TNQCo1;@(b}}5OTl$4Bl6wuBmXy^Y?6o;=bpcrSlv{&F#OO%Ye0^n))wKv*s)N
zW>%MZcvuWI`j6Rc>JivDFzc!EIN=9lb>o9S@;K^C8O;lSy)p(kN#^#_i92GN*P3H@
z+@NYXk4zIL5)y&a7w|f#GJK1oXzxvXbB(xe(*6UE5I;ceb|Ue~&p6`3*O<p1Ua^s*
z#jbqJtv05@8&qttBr((M^rnQyd3Bi04vjB3IyTy4<yc-4fMK^Lf(;{p5W;-1fMy-s
z7^yq}OS+SvL3v;_Rlhp8li`*~t>f90^3I&BVI5=$x7w@DV5F9IPvwKx4GF6*IHg9g
zGQsjz*OVu*eD0t!U&?Vm%-jM${Ai_Gw3`1^#YA^o^cQ5?Y33$K`72fD2pF;VP@+M)
zGwi&;Xgug5Ku*i`(b3`BwUTo2NctUDzto}*EDGUNDcFZ2KYhEH*VjtT=3HQ|T>cH|
zAv>#Lhc61(s2a~0bN`iGgKTznpz}InQVrO>jJ%oC>O<mOjEg2DKz@d}Rp?%Ps1JMt
z*URFv?BpU98Ix82KJ#dW?zQI1&|*!=6wwG<VB>D04Tt-K%=vjS__Bs5)RDa#l4=p_
z&mpx-lGw)K#kv`!$y={a7zG-Y73%KNdy&K6k%mOR&&LryG3j^#u_ZHpA1FShZXP3;
z51<#p`iMg7m_x$07;;Nsi>1uaUy1VyBFK+qe?VRFs4pYDor-CF$Je5!T+?d|8<vPL
zll;#3>-Htseu#^Z85KFCLfbucl8JH%*Zc;XQIizT!gk@OUe>ZN*YMq90A=2+-nf|E
z26$fw>(L5#8!InA-?)Jh3%d&z@sM~5%hd#k=(gEOJwZ;j(M`{+CiJ)u3WDEvx<Hb7
z7^WNYja)rsf^&BKSa}-GHlp+Kt5t=+thVxGURWIB95Ed(Nmi%79M^c<^ck=x3?Iv|
z3y0nw`{i6U`-;0GWTe($bS$I?WXV){HI~Y;c)X8w!s(|(=pSh^Hxx&yrKk5D{jwv7
z@A0D%(=34^rHFo4Ck85d#S3&oQorG@!C|!&??S#GY=Hs%K%GS?ZfB{svuj3YP`a1v
zTstzBPgXe}aUA@qIqDFH$mgx&bqaMML&JnRZq5rfT}-2}QEI?Cgqzy!V+%$Y`dCa-
za|lM|ox$6^CTI<)(bXi-RE}s&M6=p6Hg?7cXJQ|LUTQYVXw!RzRM8Nqfe**t2}lGK
z-l65A0r7Lr8h&(&nMnGHkbw$HM2-asDwS2m{@)nbN#ZhG?;2GKYpQJ9PsWfGTtA#+
zXfW5x>SnjF>*f%etSqqwFy8i%Qo>nucL|JHP9eqVpG&s8TNI1AZW+urQ=74&ohl_d
zMJW^SU9rO?J5|iWaZ{Yy>Xxko;YH;TXeH@xZRNJ@AZZN_?BZnqZ?g_fiop4S^j8l8
zf&9B*eMJ&CqaP_*ikP<UB^`IZDyLB8u^jcnHaHi{5zh3`YA>OXTGoa9+sY9S-!WY*
zi=<#0@ZsMUx^nHa7HYPBlU49DaV6=q($J$jZrwo<XGLUdHQ%P!Bh;+@rcgs)3eaB;
ze(u_e%>Etj4!eO!W+OBhYB+>tpO{W6dz5v@-eDc!@*bM;Ae}G>Ix)+D?5h{(2vOWX
zU}dvrdf;3z<BR*GLT|L$ELMq^PCsV(XVZ`;d$~-@8-B^*7;!tk&+bg8EMBxcJD(2c
zf1Y7F0q}i{GvZA#S~*iTVF>>GZNqxgwQFt0-t%fYau!}zd;5x(IN|ciZ}Hpk526`I
z<ODl=^z>n|R|zX!Po^$ra4AL&Z&#Ru{66x64<$89tF1PwFR#eJut;^ncGBAd)g8UC
zhHTO$Ul2^pS`r*q=M}V~Jg3=bOEgSc!sKnQr0^}KcIlit=@Ru8eW&@@lhEXr;PkGh
zw4o@5+VnLeJZDMy8(Mh22Cks#jje*0PiD}RQ-^Ui$$#b^@Z4dtuL`ZWNovV^`&t{W
zV0STC3Gv>}u5N8ab(=D0EPfCbD<0i+iZP$8<fKKSJ>@)V{;5^)!D-mhtuhf=G22fk
zofb>POQ&p)%jaJ5WTkxf*L$&U=8_hi;Pq)r7tnDGJ1mkM`im^zUbj<;q(hbwDWR_i
zbCnpj&?<ECkc@d6>+FdaZI9oGG@NFty5~COUzarWaImnptt|LM$FAVL7gK6&-P`7R
zhxl<u(k;WLJ#Uk{x&u*DhxO|zQ!62n>l$Gi*S7;iSd1yZ)qw+xmRvn}lAEP!8L^OU
z${ifO$FHc>6_b{j&4DZ(kl6W3$mLU=v?5K3n3<KH2^vg;b*za8D8CNQ`Uz*0K-#&+
zbXfiY66-b3FE~$oAJrCF_kp?oVVR^X0ABfYqz#f?eKx)Pn)N?((mAJBVhF=GC+>2g
zz}QEkgik2FLAt!dDI59tLdo!H-Z~}`2an9NAQe_4L)zqRRH+zdM*lh{iA)%JjSQNY
zouf$T;_D!{adO3wUT@w+8=T{cIqTk*ae0KPu@Zby$5nul8f^$_Yl5)TVMcTmT2m1e
zE14wLMhiFI(dCTLA-Vv+3Wfb?bDT07aGp54mug`Y0lh$KiVo2+iH3!s!(lh{Q9~c5
zoLP)PWr8)uU50f<Nj(8BZ5Y10sNQeXMZ9%X{-cW}H#5oS`>6SAk{s03BNsAOH5sNY
zG+d+zEc2>TyXJ$J7fp$z6RUK#uY&~=NvCjWag1iVINIPsMiUrkc51zC)~}eZ*Tc+g
z6qDBOOs6(oaUSgX8rQsS*h~pf5xK|p#seB-AIpp$%%h_b?&%_&Av7G`;HfLMxr6HH
zljaTMlrnAnpC^B`lkOgqkQOVrom%Vy8E$Eaby824{3IWB6B^%Q>2c>|HVX#kn+%oe
z&PT4iRXyEz#d(eP%feRpPas2QBv@B`kYZobHp9s$z~X|D3Co%t5D<nK_X0}4L0G@Q
zY~c8zlDU$G$|{|)4L_H#-wzS{z?erVMwryR>_LUt3ay}}y{R=wa{Qs;=e0xi(J<tm
zij2KmzY8Z=C3E9(XhHAvAsxjk=mOc@sKJ@>Y(MlII9x4a-t@{E<g{+Qn3`GdxQn)X
z2meLoLq9g+{SB+u4gw`cD7m7>mC<9FLoDL91IBJs_niKAR%E;VqTc$Z!GW#0(DS5`
zn0g5A#2ni>aJH=TyZ>Vwo`sRwfs}?L841|k3S_fdT$z3mAeUB$R1>#c8Dc;LDqwu?
zT%+$jpwb<k{JgM7?s@vdRRju|`2vlZ5qZ5-hWDrZLaj}!(8XmP1<M!~3PnSDC4#kP
zh2`maKXvl~$ibZO55i)A^QHfQVVcg|w+fvi5~bZ>Tf2b9bUDKHifs=ZB;&x!cX>LM
zl;H>NuLA}QcQ=f0qA^IS#>{nEOmBi&)9%U8kY4)AE^yR$ZjK{54UpA*9y8aR*YNn-
zYGD3qzzG5=YV?*_8xaCy%|rr%RTt$|pfbek1az)<pMNq156cGxeUdi##Yo=lWZ(dP
zyzA5Z6NGrn@0G-1PWRiH*?)XNwkF^SF$oFR{x8WsKu+j~zkoK}Zm4|yZ^|sv6ZR#-
zCsg7GiGOnkWRV;YuM?9t`sDt+`X4;RfBq6eQ8Dp&p3?r$)%dsJWVH>$@g)DvnZ1B_
zM9b{Be9=|@zgJ6oi<aR2fyahih2{Y&+^B45!#X>2{QMZ7Qbzlnknfnrvy>}fp3mPl
z!!E7F+vWEQG=_9Ub&Gz~{iv%qLgX6SyXIl9HOYECzde-b{43=kb7uX@F|!9_3({cY
zY=8t!iI|tyTD3b~KiCs9xDCODt1Nqhgvm@LhltC5#C_}cur6X$^uUPm^474hFuM>5
zvJcRC%MQ9{@mNDvMw-W>>_EE^0o_zy{r)h|pV63ss(EWjb*?G&>zriGTjaNDMGNG=
zi_-IUp~Qv%$oUdsM2sj)eW?w13sHPQ$A4q{7m4BQ-+IHYkJ{_WuRB{Y!k|dgfOnS@
zA<JOCnkyA;F!Vss*bs2|+tHK00Uh-orszvtKgEbAfJs&44sPq$wVk=V0LADs$w3{#
zNN=nDKbICGHi6<zY59DTV0rTKt(O=j8ulKt{=GH3SAs=3{dlN4jTtVb<govs9I*%8
zYGibCKx*sX{2iS4i*0pMI;?a_?$CZ7;OVlxfTw#GD#u{1=eG%Uds8WLzL$f1{vax;
z-<t=v-lWp%N(u=`)%g7iSeKotlD@syM}k&3P%MT)Xh1_lXF+pn#euuko)gt<N0{^Y
zjT!U$O4-4Y_><S$l<(gS?jsyfk8n6}?}1>wzgV3=K4d@FT1^D;;-93Ld5+}MUs(YE
zyc2XmPY6~)81K=4F-XL~J^8(Gj@}Mj|3ljTK{NCF0u4Dd(o385Uw0HEUU;tjd$=m_
z|FjB{=T@OlJcsmKvHl?Z?18fX|18@7i{ktu%^@%qA*8G@GYUSOci~6L%N;)3aKeXa
z{+oo1+=1`u2_$<Rn4l|+k!w4c!#(cvB0B(hEtjDRWv0fecjzk3I{B}~B_a{FB*j8+
zwce=+3OxE|8@J3U{#j3fSP=cA7$7k0q(Oh#7nQni#b3%~<q&!<+TN=6F*4T2kW@K2
zUB1mcK}#@@z>Lcl5Bf~W@b(2!%HM-kq$GUL8*Ib%9tgQ!I7vrlt8t*se<cQ9F7o$S
z$;H45Wk-gMiwk3Q1_E9Ce^Q5#pwLM1ELsv89I~VMbAb6K^b&oj_ozEn67JJ)UGLe-
zHb{<e;Eefbb4CTf$qh8eR3ablMF_fG3^N;si$m?ojlWw1heW2VU<-qP;`gw%L2+U@
z;MEK8>&73FbWH132hf`Q&8QE!0X%gKNV?HN-9|5hipgf9pi1WN%l!Ev5)x=sHZT5<
zkdDSEppwn;6|Vm$bHv&c_y&c4hx&ii4(zWoJ4V935Aptc<mN{K+WW}M|06$SQv@hL
zqt%yr%73@&ftSEH952%H|3`?iBnEUi_Hdf}e-XX>5FGr#H?aRlv`Hi$e9u;&bft__
zLI-ekvH@V?ezp<KH#?MjQ>oSzd?`^D0F+IXKse5Dv*n!4e?82AA@C&gNQu)Kybv*z
zGSPsJa$sHxSF+mBND?^8;Bu48;V3$eR&ns=Y$Hb^jt1bxQYL4q(*L{W&e4GJemkpk
zY*d4Q)v%_gr)R(ov_ej=`YuLpaPSo#UNO<(Tjc&oSf_uRb=RMV5%;{3vv(MtJFj#q
zlr*i95G1tn1J*tgZVKc4-jtq|C62%d;-5fi!ZixT1pK`6sqrB?b3C+#%BLzy_Z~zP
z$>A{`Q!W}7ujP?K0|beL1?GX@8x(zdymFYYXCxm!5|@-on6G|{w-EJU1UB388~qVR
zjnZFE=bzf5fv8G0dAm|p8@ts*J}Nr!#f)bfmjg9$5<|EX7jV2n{&+Mjq4O8>6Z?!L
z_Lgy@Hd&Msm1=2~0J&}3?KE}|uM#GmgT{GpoO^z7f9=^yV3b$YR`SLT&sZ`W<?Lm^
zf0PMcEfT<cEDv{O*kk|CIPXa5!v>TzHVo?Cs8g`9$&9k`yPAX3-j@5h<v8W$IqkwQ
zWrlmy#^}M>DaOP_6s5s()0N7%Vc(#3vFh#8oJ9@Pf=rXuGpwTv;6OpaB8P5&sUbM;
zrG*47H`<N9C;fK+8sK>FB<qz1D|xXk@6i_5UWqM-n*sE4VL3Uut(9Zc;WNGdT&V-a
zzGlrHlQvll$|p`?O%FZof?jF0N1bloNFsN9qRU%N7dKDslI%f<FgGT*r1SfCx4RqT
zWISL?rupx@<BZz^pWE42!|&WHAfJ06uz!W}P)eT_uY5ume^K14SpJ!Qvn~s1XuGDf
zdU|8M-vBrcUM@?}sj=0>h~uPXdx6ez28&)ZdQxiS{MKkZiB^XDwq<GIRwU{C;KQxW
z0wc>^$~YNsn<u&w4>T#Q@?xt?y|dF~82c3N{{HHT!982fkwHRT{Css@BBsAdz!wJf
zB*3nnufPNT!no%;jJ#g-es=oe7W=#Y<Tia{nU)vlWzPoB7HE|saaH1xU&Sf)&q<Qh
z|AkUiSYnp85R1@@a?;A;UVk}4gF}-~g4Ec@EyYUZ3RXeR87|lG#y=nMkqM5#zly8O
z7VYO-j6<f1)kI~u42Is_?(Ga#t*r{nwQCYI?iUSj0SD6Au^TTCO>Wh%<_(s#Z;Rbp
z8vy)cX^v1p-TpEGmwC3Uj`;*08>eLOYjKv9@9+~le)%~-gwFZ&?018{P?akCWf+cX
zca6p~|8Tig)3iFh9qGO09fhQ)@pg#MGg)9Zz<h3Ogq3kG6jonkFpRhvm5$oLGYW)g
z#buqKM54?tlVW%zlg-+h=99#@eQI~Pau)M%2W1`3wN28E*@oLT9vO;DGf=}}yKeY8
zY_#3GNL(07<?eKYhSeZf<5<V27}=K{(?SWjK5viRi|$XAQGag(O9R=ahFA_w^5OWr
zq(op4>wKkNeogDP#GdVPZ8W16FNh-mnV?|lgw=a?^4n|~IdB?DB1Lx$V#HoX9{u1T
z`3MJ>_(2GxHMUp6Q3zvuRJ-B$PpjrN@0g8^F<V_&&E@rz#x5ckp6gI!8~Q!7KB()_
z`qdQKz17;*owP-D^Tu0eYL$sv8WuwCSk{dYlO0f@#{ielYOE@oNov9LJq+P0K)VV8
z-Z=(n41^(J>~u!<)hTftR{*)Wo%R(#bv>}74-Fu2+_g&6nJPO{NPh1;g6e%RzAcc*
zDwxl1zGT|BZrZqW%oWSVVwC3wXY~kB-UgF-qjHsk(0lg%X4YdIQc$xR<LNTr@a|CG
z=(mh!3=g3C?>l(td7Y}MOY#sljbGkN&g*h-9W7`Y_dU{3f1&z3tVZ!hQkm8!Y{mXS
zY6!|LFeD@<z{Pe}@5aaSEQ-jDqT9cP93CG2*tU7Y;t(BW<tqGsQ)e>0FAVP+`Xwf+
z<_hj<vgYba0{}kB8IC+i-nHG7+^b&AQgT-adRQIh<Q*oYSR9klDwV|p$=hP07ZzUZ
z-4-w_`o5RH*0Bg}V`<__ROsKTvx`bCCuq%xl&3u0YItL}=Y#le%12V!)WQjQXy_R^
zwFn!Y%&BEFJ)2vXo0jvI0Cy1;a9*>BY1&G|;&oWYX?5t`HrK_8aOk5Zvjbtf(jUV%
zzbzWDmFDaVDjms%I(3(J>E=2;B4PO%7|!Ezje7>FLD#kr$9wxq2Q3gaDdiDhgV7vY
z(Ox+}LAEa0pg!FlSb)PhQmLt_S(f4Y+8EnOh{7K}4sUFuXEi*5H*K7DmO(Y)U?8Z9
zw)`3hu01Hf;ALh!e`fYa)0zh%VrGqH_$Eb(St5-9Zfdi)je+^s$c7mads_8qi{w(G
zvp;hJy_!h!do$7}+)gCP<#w&bx#hrqT|*Lzzzyp5rz|jF6pj2+rvACC>6~TMW#fuq
z*_4>#Fm<|gv92({=GS;kXrQTud6g90z8lYGBlexdcYmY8H-a7{f@wS4&X<WRueI!^
zJwh!V+$zq2v+C)T`7Z<ymz+R8&fV@NJ9s%_a?$?nf{vYNhA#bS%eC|x%}d$8c5ks=
zT@*>E(!h;(`%hnaPCQO4*YA_3$`s_w%p7XNM}2q*V|-vy_pL5LVeXIy-Tq<LvuPSR
zwtI5P#|Nl=YiE?{@;`$EAa~wu3k9^Epa=tI;OKP7i&brpB-q!x>}6~w4GhsS8rgcA
zq8Yen*MD%seKk0!aK|1M2?$7=sI_hY14j{o>!krub8c@UBNMeh@DT%{BawnZK@+_d
zz8VwTx^8oxV17NF@?7k@mh6>1n^B)&Kr1g<ZV?L`Lz(WZ>5t^L>O0IHeKBT6@pk1&
z>=e&ep-^r<Fk40e$=LfAveO66(Bc<MVr7O=8ApEoV~3Z|(7f|Tt5x@vsgMuk75q_8
zcl;wH%z;mlVkyHEvq%9LFp!(EvzS^P@q$f>+<y3)Tr*y%^*X@m4#Vk1tU~p<bf}#S
zWOjDf{C~{7bySpV8}_XTGJwDU(hY(j-8FP6B4E(n-8FO#-AJb(-QC@i(%mH>UDEH(
zv!7>g_uk+7-hbb<Sj$;UW#FFczOM5=kK_CuVGri9`IWq}yYGl%DFdPzmII<GZJSHD
z@#aj$G1L;DG>zRGs%13i-&@QaV;`^hr~U9rlG<GAF-+T~3-JE(N0ldr4k&zI4GX6~
zFu|8n&@b8sDn`O|fMSEk9OUM5Q<i!xau7F|{5|Wq_Q<Gp`Hkn+Ug6U-1dYHr&F!3R
zfc%(Sm(Va7&_pT^Yczyt#c9}-9SIaG0yEHS6CD&1wMaM19l+Y-eY5{_QoMgwke_6y
zSmQObCz)^4;mzk)aB0vTgaqC*SzAg04T5-ERUgTdgd7q!Z-;W^PtdoH2a9XxDl@LB
zs)TSm{T)vu5e^c@>K(S}9hWBLKLhp>NZs+yZiv?1TSiW<fkPFv*p(d&U&qlnRDxfA
z$d92w)<|9f<aF6>Cqeu!p<=Vw(ttcG+mLHWjz|L^YN=g|(cR7mN^v{%a$z?g&t>_L
zA!dya$xS1(^glPxGAzki7o`+iPWPR{p!m7xRZ^hj+u9}isPxzXm{RA>5(JO>Ea5W(
zT6yN&8q8huJk4~lucxJ8iojm@IT}YQd!xh4h?Pwu^Ix00vRU1MRyJ1tM!r48Z3$<m
z<UNH66$o6X9*O=~Ke%vIy+FJ=o8fo27p`?*cZ<Bwg<+`B6YQE+e}C){YmyLj`snkv
z;IJb&QM|?ipo)zY4~{h=fE+Qen@${yETFUFB!0W8yGXT=fY(gd{1_sMuAvL9A5l=-
zO1pD}UGtVGf33gacWmBlT8~l~y~vu<->DQlyMUPzUtTA<yHv>aEUFT}M)6B35y%2y
zZ%;B6HJsbrQd;SP5ha?<A}_fRr%w4Dmx|Fn!7QHs9JR%p0|_~9cgF2_j+G)*YrdVX
z)7K44`t1&DS&GAmgFas61j3%SKbg_X?`~`j0GI%v(O+3U$6VPy?ODM2xpa@+gmdIJ
zQ;zNJ4?dqQcdGs4!6!fssMRC+IL(y7@XL?Yj3K<~$kG7O;B_e=EN87EapftEKS@QK
z&s;fa85a1&gx21t1)TSVi`4T!Iu0vId%)#E-={7O>}6n(*(3<mIe9a2%PDwKh#~~*
zDf%mj{vq?~=JV9A@K_StmAicQ%o-_DHa<s0?N5+ZEg?pCVkF1U8P)Iz$GUIp&u)B9
z+u!)IzUY|KLrpIM*Bq`286oI`D<=CtRTt8Q)HoCo>}LRkpD`;TNAEadSg|({^6grv
zgQ5JV0!^kEKt1Y!r$85~2&k-fkw_R~&@WXep+e_W%;qFp4l~CwS5@zJKWo>3#3g9h
zH`gfW3H;<bwK}PeGt@zEssSZ<D{Tpma07XGC9AvQvH|1cA;G;xV|P=Uhg7O_b`6Hw
z$%}<ke7^|!IRTLSE`nbTieHRr5eSs!2J-0U*5%NG-7>1zOYhDZLXMj0zbj4KY!M7+
zR&`TDT<3oUUSUg<6(*-I1Crl!SFOV@9dp62A45uoduiQY;$DHA(0ZoV8r!T+CUW$t
zlX3h^hX_cXPlJ$>Hs@Eld;?&neDm~*Nv0?oE|78W6=zZ9EgfBl_l(enBR<pbUsCIu
z4dY|dJ;0f_+9Z11B_88rKa(rU-lKF9Cy<pNed!>`Ale*s6epTy|F7msk@QGZkJ80;
zy8aP7@H}UzaloS3<odY09=(GyfR!d`%r9b0b71VU>uqcwagAeaxFp)=GQp71B+be_
z+%*<4nu02CiN;o!SiN?(m7}&TPoBH^6Q*yv;erJxCCm0Eqn#YX{jn9Q(CI?vvJ7l4
z7^f>XMV0eeK0&cjzgAN8I%W_8iox^LA;{u6MspzHCDq$F7e4xXMl~}LMM6F9O((Pf
z@zWzad<cE~ldE!!Ic&P*6@k5deiu7Z+qR!Mfk-8=9L%Zboi>lxCae>4<Qz#(s#^pn
zs5<%aUnG{kH-u&pGg;b0v%S<;uujnSSSm?Z+cZ=enCgYLlyo4>h$lpp$g`Ce`kR|~
zRDi!$w2OaMZ_eMD{>gy67&PnSbJ(Z=L}`0Eui}#UCVvWl)BR~Va6I+*Yzgv^lwHc)
zZJ|jY5*|+t6}(fT5qW*I6JNIGK3eCQy812kC7Uk#aB0q^Y53r7fo!z^p_KW__1t9M
z$G6lWq~_iASwneEXVnex+Fi^pO*7WrIOvqjleIX@EKPUa#mB%DwQ(<N1_HSIr}6m`
zVJ2e7!U@;&XIR6w$9(_k$O#Q$0+o>A<;hRw7ji&|!kThHM!KaEhyqVEo~ZAOnG7LB
zMCr$q^FX92pNrOQ>(}r2$v_Jg4!cEMJ2$UjKehK&qmIBh%(n1*u5~LNi<4`IQe=fA
zY1+c1c<@~!l=RDkPGbX0SDgc6W*&PS`v`_N6Bq+5P&H=U=i2_?sNU8GYenKgO%Isq
zws=two)jsMO&A6&cXfXrZL91tmJBDhcSoHQXe^IU@=r374X54wqWzwj{zk_#ab@ov
z`w6@?)-!id6HIcLMDn#3jUS;q9?r+|MIkqj+`Nls7dL<2D?Ul}ir3>VlmerUZmW;s
zo}sjOGKlDFpQ{}XQ9ua2aWsb74xWuME5qbFOi~Wvxy)y$Z}*i?9rtF>Ho4^AoWQOS
zPtHr@!8o2?!i7#r4agXl48RTP+FbpED_*N{PZTy`I@xD);sV>ke2Vsq;qKt1#GZr&
zXX;4btD9u%2}vc{>LU${=~4ZDH~F}{(^&&;y6Lnr!<_@%YY<Zvdeq^<7+I>WG{u!4
zI*w&~^0$3zNK5LGdj`Eo5Q`>(-vKyK2_6(|K@RVJGi^DKp1DZODspmph4su)Q$IoG
zKRrY+KfH{<XlN>DMb?W5FChSfW+$!$tKhK|Vbf#Pw0h=L>P+7!=3Qm9IKEc|yI5Z3
zRD1&%q3RW5oA#=cPO%ZV)Ru&kT!x$Y?G`n+=Jv<CoJz~7aFT>+h<<(z7Z^*jEu1Il
z-605nViBSppEI$2HTOwvwE;D~(_LW?kQo)id(<5TGf$qE{XmDnpe@uE8hi76E+hdX
zPSg-=uInaQ=*L7tnbLKv0#%)j{=(*@vD9G;42PRuN?GFoQjrfMPhD{mt)!;0Hy=M;
z><GMg%gDjC4?dh)n(m^HKFua+`8mq@D&GGYak_a&U#^iMjbDT=NR|HBXF`<xwFW!a
zjj^#Y_2~$ev`8Mh6U)?d?doBHCB+VstwblkVaP>>zyJWI4^luR_x&yvg1Xt%PR2W;
z3Gu957JJR0P~HXAj>Msd_?Gqx`}j*WoO9(HO?Z>vwe4gm;p4BNMo!J0PK|wcALytk
z*+XK-N!{RY=@1R(O>?iWL8b~hd18>BVVx~I`4gfFDHJy1+04Hx*Xq8@r(>F8)`FG#
zEC34bMnS!FmhNK;=vdh-8$9pkvCass$N7H5#Xy7cfnxD5e(go7LSRU({t+(T8Osy}
zN%XQxr_OuzgIdL{)mR=3_b^0pP)b+h%!NeAsUz=M2;%B~4IIfIZMpztIO{yb(Tt3A
zr;NzrJZW;rvs!o>Beb+AMC0z0M`OBxwi1qz&$7;G;OJ#qVs2?+MV&CgA{j4FPPp5^
zvbX!TG#>T8>cD#Fb6~VP{y~|;H-boldO8Ln&rTvpww+q$8oV!7x{~B3%XDA<2qkJH
zKV)6yATA{_z#wdFe^SsRLa%j%Q-~Lwq!)jdyB&w+0S@BqY<vYzf-+CE`V{kQQCEY<
z%hx9vhtPJXf(9z!%`#xvLHCBa1fwL;6BphVu3^b>ixvf(F3>3O_3e~PrO7&>j>1?M
zgkOng?TwcsMuGBbHx7B1`ye!|F`=fkLB3#$=E-``tL({Etb=`+BYsJAD+SksC6)v@
zQG?@K!LiSnh{8Jf{A>=9+vxDVx>F3su7uY(G+s$rIyGO=dHW+VLm+sZr{;Sue(l2U
zUQ@?S+CJ>3N7+dS-#>uAGo86Hb`+2@TG&T2wvNs>zCwIjW=blgbobihc0%BlPH8#w
zGu`on&^Y!=6)ce(pIubC7n+W@MJkw!(xtWpBp*(lPmSG#*ln-YJ#|;;cg;zq59=cZ
z3^>+Z8UzcK6FK-)Q{B#wp3Ri}8qVTHEYtOO0A_kQZJBjF+;Yy#M;a61SJ4MZSWwor
z$kVN6KgtT|Y*xg6l)YA=#7lE7k8{=6C|K(!aX!>=n+R!QJiI35)cYc!C(K>iWU=48
zMP3AV+iCfTynG^GVMILoD#63?P*_<ka=9&l2_?3n<X9R&A#wA=pOZ>Y-J5j9-=8<!
zvza*j^yoiLZC?{CaUP%?tf;DPG<0q~c)WwEl!`CI=M@jQ^6Ni!8w+yC;xWQ0&|KvY
z&sk^A*foed{ZwUI1}IIg^7yf*PQjCJ_e@q$KQ1&nRUqZ5z1R&k>}uY^Ea=h^CLO4I
z=BQ*Q(sn4Gpu6z_Bt78RiJHvRK__X&w)KvEpJj0-Lp%t|4>-v(*XE)8L2gqSEse@N
zex@jJV?*0^lL$<hu#ItAtO?WLDI#uBlN6_wk&ddJ@Xu@>wHIx5R_P*F#bU$tBa9)G
z@QGZHw-c78E>2H7cIr;M`xFy|wzblR2gIKta;dQa5Aoe|^7P-0eR#vHbSwK6f%XI$
zWQ>tWpYDtchcmo|VUu2W03y!Nr$q(C5;@6Hhg>Tx@$Zz75TO|ST%kor^sqE9BcH2g
z0&xi(c)yi5n#GQ4is?8&ntFPteq?Zmm;RFxt6Z~tpayNj`9d+LhLaf}n6`>Px{wLo
zksnYXVo<5Y@$Rskbt=)mD66_2@`Jub*{Ee>y7-zT^lLxtYtj?dw6`E3AwN~IR7Q3s
zv5w~D`6Zs)Kp`j!n=W@vi!<BDoEp|&;W12)R|diftvt9^nsHRECnHm?@3_|lOINf4
z6{nZ6WGdPziR@Psk>YB_T;SBNYaJk6RDA%+g$P|Q-g3tntg74djyc%CEh`cZXvm)p
zte&|r>mD4`pQ{*01cH~LXVUH!KSrAjni(_DnmG=NeP_V7n(`|b>jzvLp^-%>8(?K{
z{$&B%czt}@ALhn#KvI^NgqCu=h%iJV)5A?~D=Q1}{Z@f146}0$O+dLio6@xtcqIb;
z+D6ujn0P1;)6*3<4$y<yw{W>ed@B0DLr~w!`G+RHj`<+EBRmN#rT5s};~q7uU&uiF
zDQO!}D2O&=O`l9&J~JXP36O9jE<npmM?&c)PT|Nd(r|S?Ky7N<!@YT~kz8*hX>q`E
z{l5U8y7)(3`ivxM4*wy_W;_Ti&!1r?{We<phh8)Cpl=XJ1e*U#6QOxfqY!*LssE|l
z3$Z^)jehTS2>&J5AU|A`HnOt7f2u-YR6rH_svSKo@i!d=&^H6%fXA!%%&*NT{-vTm
z0c0y9JgVx*f2pYdKfewcV-z6}_s#+c58_Eml|~qWfa7p_p5NsnhzOTftu~^w3t_4W
zXyknJdV*Ma*VzeAt5mAhdO{|;a|Z~+hZvt*JY#{>s+7E~FXOrI-yLralYS!RN!dPJ
z=_vQOf6QaEt|w?y{|@pG)mO+46C~rcA-6O-Hm2XzC+~dT9Xb@{dU=ds_ALmz_y)jy
zL{PkW^Z2BMCBv#GECv5KhEZMWm3(0Y`QTg}fDG&f$Q+rj03ChE6(D3VNQTk~*wHY&
zf~US}((aJhOB`7@K|pyP(HBD}<?4D9_}t=OLN7O*6R#MM6s&Gxtz2z}?^OyjpQ~0s
z|H+d29sF!@!uf1RDrL6n1#O-ch{~3l4icxJhQst;?Ue$zsQ6KOB|AW7nKL&21V-B)
z8%0Zn{dLtTQd*DJdfrv4XVygRQ9h2WJ?r+V60~oLmS&{Qqs^oK$tJ%;nEsq_SJh>f
z6|8st&S+a;t&8qcZp5ctsNcTjQkkfZu1?mcZ9Q6r%G@NjMz;pXgUIpmPoD+f19>v9
zsDf}ENd=uxbIn*agvW6;$uy$g#@dlq@PRhwtIhp?zMaWQF9q9z{o9~i=EM%cU`4#E
z_bVaC^&UUIlcV;(V=<>Un8gh$(x{M_`!oU-97zQPH7T=ZB_Oyc`GJhzVf-0&3sd5|
z7ZW5ETmdw9cQ`uh4vDwLgI*Z7EVuDHoW>tT#71WL=?CF|1&F0@q|HLcb2SD{TQF1_
zZ8l6E%R&|2<6c!5uNg2XHmS(3>GQAe6LeC+v-X+KS1N;9U3ux-LXNCYTJTN;e^<{u
zE#a=wvjG~9(F^c4{3;ncS!Llga&}wwKwA|e$BXCsXo=2fM37utd~;%#i@nc&h8)bP
zW!%{B0D@<kKYVI=io|Q;Vk@P-XxgG}u=w={HCx8_r0{uTWS+4Mnh7gJMqFaJB09H$
zI)?lXEx^E&clg+-bc0p`62?0Y5n_ka>GumN+2BJ^iZ40pt!J(;?)ww&Q~5+B6bXW%
z@H?^rM<0yn-6>0Uk+J}xdEI83b{n@qlW3nqKP9m{XM5BBZq%eowVCo_8z|YMBb1~1
ziGW$47s)GOq?_ALpwFt?6-1qQBtRqn8I$c}vi@I7l57|}UDnSnIPiy8MCd6O6+{Ph
zsF1W$E|{(TXyEEBGx><Doy;}=Q*kDV+?=(#U5{%U9CztxRusYsEqR0%SO~19MAQ@X
z7CJs;VwHU~jXeT6M;B4CMjOB~8!z=@XFby*^`_r|F;_XURjAikG-YOGy>+kPmb3YW
znUL3RiXZUpso|vU1tLkG7OMW*ptmr1{m>xe-bgcnm2nf?3BDOZ->KfPa-GQMP+=Sz
z*X@Fc8HEx|$m>~fEKUHC1G}Pdps^1IAu|3kj0u@=Yzy?!Vt*P&fUWCLMr&^9`5Re5
zivQ~IRwx4>sO_*)3*g9W7lV@y6$-xu6>-2ksb4&|V=A@ngl?Dq2t8W{(9Kf(mSCYt
zGRBqL%BL95Nz<12Uyf!s@WZH2ZO?Y9%4b0@VP5SYKB7;XCCrMbS<E$y_KQF-kI@XW
zk4LlK5*8!Cn`!M-Ju9aG3B%^w3ngp3F_tewE-<RiE|HaMjRLPL^uDpaljMQXcPzj0
zBY!p>JBARgQ1}U*x-m;(po1`^D_AB@9xyFbTP)GJd_*&<{khG{H2fPUGgA7IE#pV@
z%x{UPeEP$SPrO~I0fW$!#}}{46P$`93eyt0hZKW|_%K+J$rP4_9x02yrfrNbbq%xI
z5(raf39L821)agVx)3z971*3_+QM;l?58Z^6w549W_bn2DOF>eop&oL+Shu15)`j)
zE<?h=x;Z>O$g3pj)sx8s$36P4@<A0O1G?f+`kj1J5hHD-X;8sV`$w~`(ju;gYJ}Cb
zOob{|@0YfOoO@9|-9Bc1u-|M=PKl2!Dp(%GYIz#QlhZ1&dCG9ET6x}?QfK!p30n!K
zcKIZszumxnKkhDOz@c4xu%FzZx`8lR4mRsSwOB6$&ZG7QYHgRniw^2e%!86=6J#bV
z2UsKmpbCI+=BKW9W%?OANQnlCL_A=hpTy1Ykw%S$kHV7!RuSk_$&!{rVbNC#LX9`;
zN9>AqXT>(%qJeTVX9gq&3%pg8$=JL2eLS1qBpj<Jb_5~<c<0u<fhTHzJ2b)Wa5>^y
z?SK_TvGEQrJ$G&}gH@>$(NA7)Z6LY5vr{R>G~S}o`fVJ;l#`$TU{IM1okn528Our%
zDCYJ8P(Z)KR%fRI#|u<@5;$^iLk;BR2?da~8wQf6RHBrfux&q7jF~CD8ae`o1vO-S
z9px!p@lF_5)YTjH^2+C$-5QX{9T>vai^%%CSo$VdGBDFC{H6Kl-r+H~e=dvKD^_n9
zzpm?rO#^+!@BE0>o->m1Y`@X#|JVsc$h9;atlP~uK#uQZ1q(@-piL0L+<Damp_g6l
z{LmB-JUR>=;-@5?Z8FiP@_do65FW*n`q`rR^$z$Qvj_7MmPU7b_vb)yOESjK$hfgP
z6Zy%jZDNGs@s|mydo@cdD;3&x)$Re(7h3^zW>cFN^X*T51pDyaigj@&Pq7i7KnmoG
zD3Bz+w=pA5GP=VKB&7*|HoRgvz!YiJL6~wKus4ubZY=LgN+;6LhqE1=8hB}#3`IrR
z5#@mftRToRr)t8cb-buTkYI*DK?@|Ivj4MwIebf&4NS)5bSg5c)lTnepVBX!{_IYD
zL+;^kh$%7_C)|jjqSq57x}_#Fd{rvK(3rhgD6^;X>QfVOg)z8Pscs@k_(p`;zn=_R
z1V*iru>`D|yp6<?iSH2KGW&aqC+hR*c0V)Z_zT(w>^!|2@Ic~VI)5jB^~vJ+zOLy-
z3Qc2{ug!cy-1W%2)`C7dV~vr=B?=A+g4ThyPch?q*JPDexu2!V@VMJlsVL^}me{rt
z8rIvA+LH8hD>T@XyoIb7+h^<R3w{r0rL#wT6>KZ2Hu~3g{)dZ=j1ImrKgh~?vwXC+
zFt#iCohz%UQgUxI#fu92Qlfv>Z|m4_FPE^R6rBg$|5-u;)=w0mxHD_(lloT;d4K^J
zdt7vfNd9GUtA4Qj;1m}A9@zSyAMBU}_-2h<n?L`>*#Cd}Z30j}?OvfK0R>s<b2^n5
za+5@Y&4$AQGak#H$?uzg(wYOcW`NmjqdXc3<OK^0%Lia)idM0@az8DK2N5fbRFJV>
zu}~lj0Z2fIq7aXgc{3aaRD2Gvy}XZR-gE5!qoVKaf`>yM=<pBk9r-Hx6|QYWR}w*F
zGzlo0G>`!|To87Df7#a_0Lo%sZajv=vle2rSiWUM02RH-Xqr6U60EZvD>p^)y9|HY
z$X>pCS{V0NefY3W$fS@ui&R_Q>*(rc=>@JvMaVH{S`hI&NIw*_4<4akzo4}0`I((v
zb+B&_#b%jq9Fc&l5;6|GRv6V=U~-%Nz{>{_A+>&d{K<zR6RH^_n8y3HlsN$HnH|S$
z{zh8bsz(1}zbPszF@LdpJ&KC85=b^#$WFbD5)^3RaTIXG-$H{Y6#KX#WoX1_UHGAk
zZ5_}y6H)gBwIE4lWh_A0q{*zRoK!45<0ab&ZnEjzc}kSQRk%Pj``)18ul5gvPUk;o
z3cQ<gk3iQa$X`*HH`z}gx8rR%Bss61Y{jio1#RP8TJ5VhTo|=j(eu=0t`8p&gjZJD
zF?>!<rS`={c>|ch7UWIq)Lqt>(?z4=BglFdBx`>=RR7rF+pq$Bb72uG6%I|AIXQ@$
zqw8wG*?}wpB8gV7a!x_+q?8QnBY4mbOafQC*y*uX@_@{Yj#m(LVrjM>H3~#Hj2`lQ
zoluav=?=d^C*8A^L_UZ8sLXa>`7aqM;J2lNs@3^lUVX?c4WbKpXZchc#qtYkq&SId
z4MqP%b_yIhN_vvTj^uSX;C_J_8@UP3p(hDsEP+dtWKa3*@7-s#+C#|+$XBcJ(gRL>
zb-<@uRc<{pv}(B#LY<cLwKM8n^=C;TuVqIfHajFjiZ#{<(0?|>lf+QGk?1{m%d$~z
z!~=4J35%+&Dh1`;M_|nTYtsb92&vJoZFmbCiI~Bg#`lA?6&e)m8Q{X_NA>CK(Oz`Q
zR^`Wp)-bn__%M5OjS)byV?ww_j1RxGn}U1-+7WHh90|~AcRmQyh|Bl3qbPL`yy;Bt
ze<lA)${QoQEA)%cmt*}HI(fNa3iww;;r0E|GLvqq+4)D=?Q=>1p^pvSCI%w_(XQBl
z#?6aNeBqbYe!hOtT}QBLJCsNpymZK^5@|Z8Rg>zLr~Eaa|HVds%zKv9=ufW!YW3Gu
z0^P?Yw8z47xJ(M4_+!C4htF`CNj8r+)NDVp+&|w~#t;Xw&#D#e;J>VW9*eE9=~*v3
zMEtK~L%5ZH1kIb~IiDUx1I+X`w%7{SCtgaHkN(5!kT|n6=P@t^3X5+z2OVV3(JW=9
zj@!Kx=Asgw|IuuP=u+q8%`B27U1wGmP>k;cRxG8c1~3kocIxne0vGDd8^|EFQX4^u
zsOg!U(b(@8l)iWp@kVFEu|+q3afg+)C#m$pdSih}TwlV*VpVhtE(%BgRP<ZV?`Qfz
z0`oAKBKTZ1w_|zRcY{m2lo=#ei9H{IZqU>--NRzo%hDc?AroQ;<-(9Lo%6Oi3ad+_
zNh3tnZu<x%!R9CW<H{F&LYwYN=?)gSPxHu<??;(0gu@B?@~pZglz@{NoktDbM#k}E
zd4!bpr(1&H0Vx2q?%%jMQSAkA8g!0J_6-6NJPUhA_5zSxeAZJpbKN}^4p)6$-rvp_
zPh2<)FQm4$=Uy7e2CswIyq{%cN}Ro-zRiSoLb~0k;j!QfyfNN@5c!QZ)hw)lL8;u{
zVO+j6tVPLzqsmQ_oc=jvYyB+^l@a5U83r*%3yT$fab@V?cL!LgiYUK~&!GsZ=w5KY
zZGs+bUwFo1Q>R$HW|7&(340}-!J_f&BvIlVbzKc3#t#w_$NG|#+M7$e<^o%yqeqW!
zVNk!CQDezL5y@Dl&c?A;x|%sn!1|aP4x!zuHUiv5XwRk>Hs};g?v&X)K3ZC}!ePWd
z)(yG~!xK;fdZv7mDi>}3_b~u0SEhd6p@N;eB9&gbTw37r6s4{&V30!GWb1X3)#&z^
z;Wt5)yBL@KX3;rMrwvlKn}ldK5)b47B0`+h|FP8mvn`82h2qH@u~D|6Yy;vj#7=iI
z9Xl4hiY|yqK=o-yyic8iZwNL|AoWiiioU7&*Ops0*}_ER>^Qc};iUc-`?W#spT=Z}
ze16QA+b0dhh|R4eoEjYNv=s#ST5J74+_4K%sKh_vC+|CtkN5zUUGaLB#Y7XV7$PL+
zgWl)|wqb@<P*h$VMs{THv6N3#5(YoMWGxIZb<z!X_{Elt{`7sYgI4NB`9{GwR!D?C
zNWY;L^H^&D8IOqp(jys8V(mY+&l-G7#I=_4nR70nMT;T2aflb_pwRdtQ}vzVR}8o+
zP2Fg1S?~L&JrcVDyom3~9kVgw<J4zW42Baa>~}YlYsHx27&OVTsAX&%@!B!>>H#m`
zKKtJ?QQ#%XG*zxf76Cyj7m+U2nLuo$`T1HTyX@H>BHCN%W2_KQZ<>w_w(Ik(9aoV7
zKCxWtO*xB&dUanS91QQcXLPL;^jN`TT47%}KyviOx}pN)hlRg{-5Lhf!~<B8Ny*Ey
z2Gk@#Q!mDz!Wb!jvQz1|R3nKD*xD-u&{iUC2`9;+kG_(pFCot|!V=>YW4(G}X;@dT
z|B1|R6n7k{15N;Jom=C2$+~mo-P<8m6^nFAA>wed!N(zuvx*yhrnES#K`DhoP!?KN
zeqvzFII!0Vktuyb7%e{QUXd%OjFpNK8LZ2a4VgvI*fv<C60usjtP&yK>vCG&_U^%2
zy>3G!ZKahd5N390K-uDF++sIhy_gq$SG#e-)M$3h!I|vivG8mfa<_hRJYNxRI!#x+
zArU&%XT=DeT6DVL-=4ZRPcRhWJ(~M<R+DElrLL|Y>J%{w7*4aS*ZeHP*^3(xT_D+0
zGEL6*dH{Fnfi9xbgDx1RIHK92q0W!kxu>WZKFNGMr9nyt*e5bk%qMc+4=y}~(Bof@
zykz~AvckIAKswkF$Zw+nfJ`-dD~F)nON2*}jWZsl(up@ob>}FiwIxN5S`scKTxo6x
zkB+e1SxdI^$c!ZCOpDn|Hfb#@^rvTe5z&Y!K4wd^p_5{=j`P?~n&F#>NVe(wCVW+^
zx1&|h!;6968B@Q2(0JVOe4H#<B5T19<1qO&g`GsYeHN|eu1t8*k^1d(PVbzLU*EE$
zf6~&b1NQxr>q(QE-N>c0jFsbNPy90ii1eN9CDr~K=&^hHZ~?31zH8mD)!=~x?VN9H
zYF901VK;!3I=CCi*$)`M^AJfjCWc$^^l_GFjb^zzZZt5n>I1oJ*M_b+TuF4WHcJk)
zyd30NLs&KetEa~mUp|mAb=Tl?e1F&QT7yG3^!prO3)r^AyD|FjLnPQ3e$nIJ%^Y*z
zbgzj4^JlYRy?G12MT?8Y6XJ`3;e(O4WN5*szv|__Gtj%P^)Ig!$X6sf&;FbjT=zA~
zQ^|_T{%YY`^OR-1t~6AgrgyXKgyM);J6TB&QG;jLL4=$9<~fPL)}=ZBP?0(m(MeZF
z=UtMT&i75P@1jX7BUR~Nxp7Q1?aC-@eRiKMRBh#IE!H7vBUu)H4&)DGxKgT-d1v^+
z&uv)umqk<3VXt;1spVGnUU{KhcdSoqzP13jpM#gZjPggyR|mk%Fh1vvsChf!ZQrh<
z1;_RJG}2%N-JZ{}54yieQZT4mt~%cPL7HhAr#dfhHedS<6tXYzJtc1&uwiG4Od7N{
z4x&;I%SzzQmiGAq!FS_%9>}x76W&0YgM;}Lu{vK8cH-2xw@epfh?GSt8b+J<7XLA`
zU!LH;_D7rQ6}+ea)TUZzLzVY2CpN4DOP0Un%)Vib^K3O7dQa&_C08Oz!j6vN(3j#U
zXylG7zdkv{)l2Fz@cnQ~_{ma4tU(wJj;r%DtV2X^(N0&A(gLS{;R2$UbYO3~+u5dS
zStmRf5}~5{5+=CgDb&UD^8TuU^=c`HR0{i~0ckuZ$)Z-~Jl>xYEWbCWsY_y&oW8Pl
zPp@QjwOhaFp2RBeOwcy7U4IUkVGRjce#Mkmu9DKR@SS`u0?<L=QvaDXHx1W!4;Dst
z%f(N=i%-z`O*+kj=3U3mHi>H`^?w|qpUOkR31S%%(<tc_izb<87&i%W?RAEzNM0?+
zh?fev)oi(?57YG_PW70UwCGE9<K0bdi>9brG<{@b6RE#;bIgafTrc|T{Y3vd(Bk`2
zbn{H_=C`8vy$sn<+;^lb&F0ROl2f^zYNy*Ow&)Z1zTM|>=kwD~Uu_KpyYVrNvn~9P
z#I8-<Ub)%do#*y%abJ$dPFXRN&9zs4zx`H^w<~Lqu#czP_5Q1KBQk!;@kX3F1>jBn
zgu|#paTdCT@!(G;99AoD<Uh1O%M?AFzZYV9U^4Ne$$*)c?wN-em{)vZLSn>9k9Ldh
zN_HqX4?8=^;I;-53~`X^<yzK;Js+3kf3M%}mf=MNFLHwhW^6a}Yi~|lC^Nl|RN74l
z!k3o1DG>=dOhOpP*`RWQXEKgPOA5X-{-kdCa3>`*w&S~39gdob6Nz*t_@JD}-9B6W
z&3DcrfxNF{G+&0>4zu_w;*nTSYpqo^9tZVj9t02XTsWAQM5#8+7ujOFv>ZlfuY{Kf
zJg~;=L|lcZJD3G$jPVZ6r!*-4W8e010gMgY74TK}rys-nvf|VVosLQ~0e4%de?#G%
zZLmOF5^{lNOL2ms@)$HZu5NDQk`PpVaIrY{uDEV?LPPA6bo{Si<N<6t#qBy0U=sVx
zCd0fI$Le<EuSb2d8Vry-cEbFmt)vH7Vu#_=M5J#+PJ$<PSSkfu!7Ao~ObUl-804~(
za+5eq)$Z&A>{A$*h7=t$E`C{Hgft!6$QcR~WSD{Zh!|z;0al<~2@M?2Q-AAD#Micm
z(DkatkOML{ep%r}Jn-u5sZMH|Z83Q#O(S%2$k@5MkTsPZZz!Bo_(sV`r~|QDtBt}9
zmtN#e(H^t*>S1Racu}>$0JWx8Oa-MPK2orQf(vjewK6b9;x2x~YrH!A!PC?To&m`d
zpDZX1`Uu`#tq-m1B9iZb==S#yl_(*%Hzz8!iW=wdPovijy048<(X$EIa}B#K{G9Wr
z62J?B13qwD)3c3>)otg*`osJeFV?!L*PQ`hnu<QaOtObqHs%#9YYPDm{ZPj<BFA5a
zH7BgU^}RA;9wB7o^U%M5p)-_&xWzY#)+T&zIb4eUx@L9LtD)_%v?|c1z^d-v+{n_~
zaR2>};bfaet4--Lw#^cV+wn|DZNE;2Kie--DUftQgz#XPB6{St0GF&z&FZd>U@;$2
z_U5Q?>C*Gbf#y3YsbhSJrn!M07*Es4yMbg8*X1%FM2;2=Lzxq^mJ_`5kMFpRHn0|D
zhIK)8zAF>_GNLK*geHTUX=Wsvb89c#%gewqW3J-Qry;2Yf|SJCo0H2iO?4+J1X3<}
z%AY{!7|e=JfGN`)r)5s-I!tQyB(2dvhuF4&JQ@Qn*c70DZP(i{7N=bJp$6qnRs2F|
z^ZBcX34Q@IWypkCRENN&BABVt^qW3Bahwcn3<Ug}lBgYxE!XP%t4brdw(E!W$WX==
z9<lgWB2-06Z4^wYe#|o~2X47^r>Tap7mH5F*ZQNDS3#*GORk$ZUb><FY)U?o&ryCA
za-j?>F>RT!DcW7y6t>T*5Tx2?-w2$cJJm`$I$l>3qEqAKdi4~?B3+6(6%v^(S(+l$
zmvCyyE5hsqj-axRDO#R)y28mO6Zw4*HpB7lET3shX9rzPdB1415~jxDA92__@mP?J
z3*!%CnH@i``4kCCm~oW6g$J(Lpe|F1Jf&+r?rR#Lodt4A8Cq{!CpIxIy{ZFf2{jJ=
zj@f4Qs!v&C^(~+wo+j)!JM(dk{dSuXASJ80y(CZ_=G;Pp14o~1-O(O|L%H6rVg|Pf
znN^y)WHHwIOavTx)wUxSDd}VTDH@Z#u<QMfQLnN)+;Wp&km=m+H|?`!GtRY?z*}NN
z+C48$f9m7p0O}B@VZg~jnk;Qa?c5oYy-*lI(Y|Q)Y0{hLR_xutHeh_Q<8d46wRt8&
zw|>yF)qlq25UHr)_EGln2qMe*&l5KeH3u)Ai{5a76Bp-!5;4tGx8jl@&eOUi<u~>-
z<r^ydS29zH;xPhpXr_~epTh^G@|4y=!FRC!di#cVyFEb_@wju%YYg}r{4jYH@^j|6
z=fTgK?%R5f7cO3k+BH>u4N=a+_q%~cc*>hB5wfs~D$qk<yQ}CGLG*Y~6@_bojSIm{
z^=m?eU_2nder!NTuC&}B%j7p<ZJiVDRBN*uL*K&{6VDQ22i8C(<m9H3Ri)%D8{Ivj
z{#5zqv)=>!Pj7inbQ~*S!U;nlm?zwz&B9T6+}^qX<?M`#A#@AqELeREeTgQ5<Y)DI
z9KWz_m)opYbu{N-*T(X~ze)+1$Lp`yta`;9uu}81F?;X&Cb)zHRR=eCuC<i)e*leM
zRD(s*XFY|FCg8&bm85g;lS+Qj34#|<I*@afjC!h;zBTe1rZw_tMFcT}wCMMJ>he}t
ztwg7p8k<_AujFxeybP_2F|H>`JSzR0+u`Zvy^$33yE3GkpV95<9HqEhHW!hG;CG&e
zBzV*XtLa@mLbtoPCLL)!jN(HLX4cO-z)SMrU7`V|E_xW^T?AZPoHJ6kR({LMT7L@1
z0)RImsf4^yao5XrJ6SJLEPEmi#`bVFN-Z@%vH}g};r!!q$5Ky86z{>MSKZBD$i%P@
z6y5<83dLB@ZRkt7*~gK)gXcdN=%79MC3QjEsqqnbqu#9Xih<ZNM+9WD9aWFZVu(wo
zT15nxLmj`Xbnv}D(ohZ1%+XGzZ8!1ZfaUG8mYCD(HpVn}V;+!nshMfgSIjy__qZ4x
z$;fM(gQ^7w{^wno5&r@tGO2Lfsp{t-lehh+y={K@*N{6f9N=;@eFX-wn-K|Es)BO#
zC&@r$@Xm*P9TP5>MkW3aGJ1DEx9XDc3^2@9<u}jmuIYAEHeRgG1`-+gjp{uDpb7pn
zD)a2g;{ZIBb<9>53c%~16@MMXlO1n)xsJu3rOr2XqY$RK9!9MCo4ynaI4>9iX%=}!
z)tc-({_7D^{>dgbNhqG3A_G_Z6ko#mH&JJqq7|&Hfuu|Yj&mksD6rR~38#`<T|G{?
z069B`WoMDHFj)w`s&iX4%;wg^42lfG?_Q2(+B~FniZJMWAss&B_D&EdxBH_^8$2fH
z<>R1!t5H4mzcyQ|D^Mr4%)qbX)527OD5`~k*Pu7Dk%8<=89ZCxs<=65r*3&thzjLS
z5k~&<GofdfT*d$4vHWu|SqN?s1iR&r2ZtSXyC5-T<&dNH*fi9NwR~%{J-z$`?DdBj
z@X+(E!2&h3!ZQcZDIV%exW4X3Ssgu3B%(+9uXp;--R>C9>mxk1$cwST>mUDnLiz12
zUx1Ollmfm1h5*L|fK1tUx;?h>W8|Z#+Cl<_72@7Pjb)mPhBkK+x8V@cOtYH{cfm8I
zxII~gEzykB=;-G|1<zh5<Nx_)f4nG{B6B#|96ScXXzF)q6^oOToi-C5=HeVx*nbMW
z7*65iQf(~Bz3#=Bf)mdPMiA5HzVfLaPU4b9BNcG1PaHY{02RM~t-pt}ehB5QtYlYi
zgqD2*(%8|6xXFMjJ0gne?W-ujQu2^KrFVaK8wJD%19;tm$%`@b=~7`G9W^!%U~Hsl
z3$hl%Hc%kDs`n6*nJY)`1@aSykqd^EC|5{~LDUOm|HlG|Wzu8>;81axR8qM#Fi$^2
ztPQ{dKnXY(YyiuHa+7nFyW|q^b_EZV<FBbw0W8sy^&C*R_fH|fxNsJ!3q4=;;fs#%
zhNrk|k0VX5KgY-C-qHiJWUQVq>j42%#Q;dA+wqdP=HC@=X~Y-#x{`(P1fup@M&>*=
z^q<qhFP1+G(ZeIdBXm+$gXK7hp0{IZum{M25+AZ%+kH{o6mY_x0OTO~_XsNiw#Aw)
z0v$Z$$A!W~6drBf2pmkKD7cJf3i+W^a>`JH5rcrY0rIB^Zr*aDX*dijnQng2g(}my
zfiyu)U-mvG_4-e~N&L9*^e|cxh4*M8^n9pp{xf&WjauklftA2yaK0g?8!cestj0!v
zN>)EInIRr*vP71fa`pq4$)gqDtPN9#qoHT!Emr(LhPlBTfB{2nxvEeuy`5Iw%hQLe
z#OY-7mBR9bcYs$-zcaop?elDy?GTS7e->C41&97SXDp+-3Ie6rbFs<EBGrkn$ie}-
zKz#ZKOHwg5*h>K*P^)^=cMIG0Ejc}qi&tD!2qXZ`y+yJ{!U-kd7}XUIdj1Wp5Y#-C
z_fax@Fq!8g&lmF<{llmX!9x*|({v_q6@yA8_!M1Z7KrQnIJOc4!*v9vMxz8h+|LA5
zWT|R}qBBUMPYl2%;8Vap^g))knXe<eDMtIZgYb_P3dIaKk+yR0x#16>;UmS$KuP<_
znasvHS=Cy^Pb$VNM$(BR#wZ4ES0Mk=h(C+w1NKcv&qdi*&}ug|MO`fdmdyDs?b)mY
z*%L@cj!Tw0jHbh2AU<978u8XE9++b*53+<R1A&#M=!jL~_|rfpw0GaI^k3W0ANTTd
zsHdivF~3(e89Z?(2wuS<5&I*zKeo-`hxX+_A46<&uOAiUU!1=m^*HZ>MS{Rec(WD=
zIg|za86+@7)ea|XG|jU)*Bn!n;TO+Hl>6b9ihfD<7IMJ2J9G&PXX=3o;9~UW%jwzw
zVgcmB>yS4yGhvWGWVIg;<B8~DI3tbt*ZG4K2Q4~ySV+N(JoGkCF!aS4AhF<}cvYk_
z<3VHwDA#+ERV-r~^n;h1cp^fx-`}eaSzts3MW_ZS!j^%trzP($8=JmD8b2#B4{$Sz
zC=Fe+Zll3qV%D1X!4JbMsTg?e76%rJ^%^xZ{?{k$AZbFI5Yxw`e!b2gRvrJ|@b?me
zUj;iLm-+Do#C!ZmRz+fY+U6ben*94Rb|HW_{=!?~O0)_N-WOV4Nk%1Lw?d(UKDQ>K
z>&V!fBoitdeFej5j0NF~^UzP8Olf*-&vycZ?0o;BZStxO3=)(Qs^bo_AEo)lOtv=1
zF)!)17q7-jwM9{NQ}|t~&k;`y+ThG3UYDt_;8NfdefLlu225n~@hb1b{E^PR;5z9y
z<8%--=;+;$jtWl$_5Y#B0tA)^SF8!-MZ0iOO}}bfN_zHUm`ipR$?MPUf%v#FN3omA
zXa#X1jl#X{Y=#|N<^lm3Jhq!gKTK|~&dkeQg6k4!4xW3aR}!CnMf0Gae~F^gqng%~
zb(xR626yzMMDW38`=sYRdSmLLhzmhy7It*6{|$HAmaFWr&CT!UQ$zzuOZZLv)@@D`
zZr9{KpUXZ2L-g8E<9xUU+&)jd7kmur3?yaY?0Fn?0)&$x98kz4o3BpPJ2NCV8%8Qg
zhBH)O;j*NvDzY7JPP*e20J)gA<@iP`m2yw!Nea*T+hg1xXNU;ZGERIt&Oq@EU>ZG9
z=D?}R(-BcF>nl|6>@hST;r4w5%f!wB$1CW~?$IK||9+zV=QTOF7so3weh!Z*R4B1T
zm$LzNq7sm523~f)R2fjz^jQHjxaslz<9fet#k6WN8+AwNz{z;UPh9f!o1>IDKxkY4
zwkLsB=LVY*{QJ^+EDL9?WJR6)^neJa=k@Dj%GcL#+2Bq*c?>OZ6B1S<-q27_PuznO
zJ95^A6OlgqZnHF$Z1=!T{J3{b;I97*l#=A2p8<#SSB!&~npo?ASriQJ=tH4VMl@qW
zbg*-G@wx+?bt+wxfulb;Yg6FxPXWu!2~rs};3M4mTC?`LV%s73cv>(e_-5IwV-}T<
zYEcb$qaWw92-1;;BK64xSN313zL3QWPnYt)LsX(JzSmiQvDHp%C!t6}!7qC|^ZL;$
zrtD^Zya)SHX-8d#IcI9D{BPew46BtNV?Wd?9Ognb&jz4|2p%k`9W)!God4kg`MIr+
zJ~pWNvp8rMd2}IFYuNsGgc0ec8g(2^a<EHp@n4&Z5EQhWigY2vju~M0FajtbQdjGy
z^c~Tb@oxv@9n$~Dzha2BWwbqdRAKt(9;r>~=|cTG45`fR&%gWc3ozA#NmlR6wDdnd
z;Xi)>Gt$46B*5k`hU8T{{>1qG=s4v2fOPi+Y1|`(!Ih84WP4&{EhH7!iSG1rp-x)N
z)8c=9&t8`YQ+yoW*j{|R(h1LqM}^OQo89fRBn$Mlc*)s7(fv?{0~wIROQEJVg$~!8
za3La%WvO6`2S5Cd?_Zy?!_WAV{J#~D5y42)*mt_R<jk`9d=2460G#T#y8ZtGr<w-;
z^&<iv7vL+v0=VuY{O)G_|N6C`0nLQ~d}PkwT&jP3#u-U~)!8nbRP|5d&;M>q3{jxi
z*YWd4{cDZ?5zXQCgBq|pFRb-U{<X0FOI7N8usS~=DgInQ|9wGL|3Fo`__c94*>VF!
zv9|f*_OEsYvz6lAErQ;qaqgkN>p23b-BLpqvm#?<EiM-pjtTwuwlSwaFLST?9GWi;
zDmT?jmzSPxM_z68MpDiKnQ1PRs)^h{kb99wJ1lVKgCO0X=$(g);r?(j1pS=q5e>PF
z=>d#NRGDs%;qu|Pm-l0*5}^cu&6qB0p^A88O!odIAX&!%W$BMQsyGdDTG~7nGzy!O
zSO)x^X`q;%jEs$qCFlRpLL}&}HrHUj3s5vy0rN%GUGyj13Ns$0zrV4`gF~C&)y%a(
zfg#IkMdWy}=Ia784wG_*NUm<rxpSdo8ZH%weT(bWx1Qq;blvp^mZe@`oP!Yv2aUB}
z>#plZXsEE;X05yT^4GgM#aGcgSZunwZ8&J2IsnS4s5lL0@p`*G>L-iNstCnJ*Rmh8
z9S@pq=FDq$Spdkf>*j#+@#~VFN=B%LH5y%VIDkjcbia(r=D%EL^LB<yC}I8l@;>{l
zmC_j1K100}z#Q~Oq<3V^r8uHMI)M(wh~3@aK62nNreS`9v3e9hEmx`~T(_B6puab3
zUv3mhC+%<pRRkF}IG7_)0ROWkQ?1ShfbY;bYVOx<F(z<p7pg&55cVsz93P@aze;+%
zB;>RhrvT8-SGP;P!SDddW!53$-u*c5Q?8XCkibk+x6-Y;{^jUyFojFDxMBZkf`G2y
z3+y*hTuVe`vI1w7lSjjXD4A~lGRIxGn%{unlJzS>i%DKUt}O1kK6ev$7C4SKTdGz!
z7)Z`Ly8-A7&I41Vn)<Kg^K<RPNlsF50Ln~WH5VtR#GKP^k-9wD+Hw!+z2a>3E*W_6
z@5T#|7;poC$w|VYaC;@@Ip#zjIu^?{OJVPa^#QOwyG9?wPLtWhPR6s_fbj6Rr>H~G
zKp3;(>){<-sMHF6k!DLX<d=}+<#|%cdP>_9d1$N@BNb_=t0sRg)7c&+fE3Z+88>$B
zV|m%R$n~{3+FG|IZY$NLdm-A5miBlt;9?^?$s%I-?jkUR^uj0h7KrOzaDq7A#N@sI
zT7T{aK<a0U-+ZPvwFVsn;&xwh%U0i<C3=K*hhJ~}$Wm_@F`xeEx+cGD+ri&9PAlZc
zHLb85!$724)hd%vgZ9P>U<CxFSo$Unc-*V6r&c&fz2k{qo*63Sd_o@lqh<?{B=<GD
zc@EM@CmfSx7?FJqVJf?a-_auAG;)tXUy3VHGKP5V-eL#$>H;h!#WsWzigrbomnZdM
z#8e)4Hy^Ui@Xb$>c|6ia&reV7HVn{-cwDFVcg_zZLzdWxPLHSnLl<PU!r!lflc%B1
zuKvzIxy&Z(1vPaaaKo-Q9C$=j5O79b8FGGDaR8p?ii}5Asp|Q+lC$gi?;C9w43$hE
zA*I8D_a6XSI}oKlK2q*6n7TH!R)>{40K|}&*xy_qh~V>CG=2iYD+xFuZ-JEbT=}yf
zM**q3QE%_RE`8Ncr<6teTj^xwjMS+@?ww=~)pSUR;jpHBwo<OM@oJ`a1e*Yn#ilnv
zhK^g-lFcn`up*_|_Nk-Z;B3{b<>|uZOqs2s6?XEbn$a4NmBLYpq2scVKXFyIwo|~<
z8!RX7%DlKOp|-c=A#{#c7IZM8XKiggS|*B!9b3}0D4XQIS7mOWv_xdO*qA`CTCEnF
z3^w8RybO`gpNzf`#LK%}9_(!FR;Rw1n3cuym|Ot%u;~{I-GnR16~WtPbFtR%Q&Pe4
zY<CgbqDzshT|s?VvXCN;Hm+;*@88GWEgPvlh7<DkMuP@OVMXoW9TZjV&gLq>BE7qB
zUgHqE?IANUJig>1FPl5KR)_j&vyV=_-FjV;6lGIKV=3>ef|s1mwCCL|Yw`n6cH$rV
z#A6Tmc{nxzaQ?u8H=HbaSypDzligqz^C~7&VUXa%K1{OLjVx|YOMtlWQ=T5JYIex=
z=TEg(jJw9&`!C|_LaIy@;&`h}or&*5>L?FX=NuGGzXEA_Z<-R`C8h3Jtbdk;WM3D}
zsb!Z?Y@hvD<4wZ|eqwLrQA8pNr71LT2;=2C?oQ$@ScXZKh18)XJ4`C0vcuz+h~TsT
zAl^Q*V!F^CkS|oqON3m{qY${O0@6DF(}N!?9g=pWtLP|uTJK3s<0_4d8d5CxD2Dl$
z=J>g=0>NpA%wdFr8n`%#=8!mC4ZE_&>wP)Zn`rYdtUs`DUbiR9D$P|hx(_C%d|hJk
ze_rCyM)V<hH(XGy#AU?icU=#-3k5)(4C?js^Fncx;O4VkE~DyM2|eg6HCvo{%aK(q
z`$@M{0(!e~mzcIfuAIj50-si+W>+xoSdt5jx*hBgXmcHM!PYF)m})rIJ?LDRjlovC
zqETT`<le&rD9Yo(^}yuIw>YLWGIX;NCv|v84DI!Y%2Z$(<UJbqA-|iccJoI<S)kjN
zcukR929Q|p4G-v-cVCqYy3b}WX7BGyATX&k8d=kpmvZ0Gme$gjw=y>WXlL9P4{mTd
zg3H=xG)|J*VYoj@iF0bIYdUC-V}JAcG<o=C9qV0#<9^aoGMgd1B+F|#+o^2(0t$>G
zKtATF7;9{V#CbG1pJg_iltpaMGBRflG;Z`0GXkxv*hzDhx|2&Do)2{&^rtw}p1xa$
z>v9aNcB{+Idtc$lZaykjKJiA{8kt})VSna^mYBCvHgO2Y5T7bRv8CB<BD11s>=7D?
ztD2Xg|5MxCV)e%B)a!L$T&!Ij$EmY~nt9jHE;!@6$rZ!V_9g~5hm_&G2gMP?!aI;3
z)pM3JxwC~!jVq*6{Vf*F;`{f*Pi)#QMc30VudT|nfjn4?&n^cIOFTW~N4~hvFhES7
zx#JT_^1|H7Ld?oo9g@!}V*pisD$8ZRE`5l01dtc$r_?=sP`+w59^O>gEJmOt$0=K&
z5!I%tS&MsfH@x=ph$;Bw;6|&SNEyzpCUE`JRxuo4#WWC+WoI@V=bNsNj}6(~UYgkl
zup50R-%0+#hwJY+-U()gfA(0k^?!c^h<y35$2;RuEsrG>F!aL}tKGbMpY9|kKsW%;
z$njuuqG-57I@S#y0LzgJ-kk>bmjs#cO(96XBKJt)wYSpNouFr_MHMZDyS+Ncsq}4Q
z9ZA;7;L&b>LB3Q}MUwfF6yCOnLkccJp-6?-d`E5Jgh@#jH-2?#+;~jA!MiPb?5m_B
zRl)=U+scx=Y@zVilwh2VdV5!}A$#IBygw_OAr;6jbY~*v4U2IrI+5}k$WMDfDGkx-
znbeYiYF)XH)6==7B2!#Rty^26Yo$7y82F>9km~)Zmv7252K32Wr5VYZTrMyt;lc>h
znEU7R716njnXN@A=EYxmc~DBci*x74%1)<>aMOz|%o-d;0+&it1&>QZ`1t1D*<=QF
zxJ@=u#Q%^$nsbYgou=?aLZc?}_BKAxNpojJB(?kEGBtg?o^0#itM0f!-a}$*$C_Wz
z61(-%I_fdYOYt|N4&2Bhw&-mfEPq~fufcF3<B*iAi-QA#zF$<_PY<_<Zk@o9xh}p)
z=-pK($Du+vdQrFmXO7~Lvr@^kd6Q2V2xt4md2=Of10@DNzpeBFdK3}HQio&_z8I5k
z87%s${(gywrXmmgxZqD8@$`#P9*bjPNAE=NQ^`dbWaeP@CFUb_roS8VM%elJqQ{(e
z1?h>k5lXqp7|ZwUcY`N`K25qfxWf18=K8JZ<l^7uL}tnHj7Z3fDXP@*?#uHH-h>aO
zU{@d%Geq>C=erfCH^knsn1lrTcjGX9HIgUfGI{YXi6c@a|1lxIbK#h~$fuFk#OG*_
zD+9J+jvQET#Od8;uXtxLskCk0B-`(N+1@EAiA<8Gn^j&yofI!8u^Hw+r-=}!K4Xdy
zVhW1#gOu;$l*B^Rm>M|$ylf9kHWvyKE6_k58KMp)M1e03jCv*8_g$J%v$<{e6?N_F
z&)K0U@0iKsf}>kkJdK2hL<||<Y5;ZQN@?VP0xE?;sjx65?u+k>*6+Q4y@>+-AG+Q$
zpspy%8clG5L+}t>65QS0-95Nl(BSUw4#C~s{o+n=cXx+((%<xS&-A=s{NUbm&pEqx
z?W$F)RvCO>7H+il=3_CRokS6>`+fcMoS!KKS!x*XaqvLzF7m<Y<j)#P$8HHBF^_S#
zWBiACq~xI|A!ID3f-i(4Fovl5a!5G=Ck$vpDbdb%y#6OYG{PW2RjvB(!^~j7-iXKd
zdQjncf00b#uu4sLzY?$MOzjqlnX+G%D{UjxgTVp|^!|T;_yR=q;$V#v%Dy8KjPm$x
zjWPJ$Gpe_uYe9>90@6O{EDfS`jQxhR<8jml9T5wdE4knZ#5iKKIsr)ZdlMG9lRGsH
zF}(QDr-c3}jCysy`!;ptCb@{pl@%p*%6k3bb3KfHVWvY?tISF6V#NvS(%5xhAd%L(
zxCkH3jU;9tmbaR{e}@MvTQ{tfMEjR(_=ls&ZX=3@%<*GFZ=5YlpcD2_t}Z~3h9}G6
zhpiy)C5f+(jAo=A@0R9iK3S@-ojJ=PjUY}o{rU(8*Eb^>gm{|ngi1|mfnvAhD>o6+
zf<%#*>=+W=PkN*Wqd%6Z+ZZIeN|AG>%%rnFj;Fl@1fEUx;dPv1bib!XRs}7NK}Ni*
zzv(1FZ4MQAeRKcYbpiM1QxnLCA>(xMB}LgNE+GYSNktQH!DXL0x`E>F_`O;Atwt&F
zHImgRc{0rK0G5e;dQoX{Od7SFc`1zWghg}>k*o!q&WN8&eDm}I-s*y(Y=blsG|jAA
z3+r&EaO$clG6Wr_REy*&v{gG%*{)QfF6Rw-8T1YXj1vJrQ~$Ms|Fw|bgmM@tMWb%w
z4dXBplTtJ5s$TLTGtsUlR~=EwUL;6dX#%4JPHYE<v-F$VELaKBC&$Hw(*7=Nt*{an
zw-_T|BiKgxd;f5GahoXN`5pG`ckm$d!t4k*4*Bh&Mm4PoQWewz;k6hCYj<CLi*=>%
zp#|LS_cWOz%Dq}s9d+nXajV~f2mvfa;$;F@1;dRc9N<0tsHuv<l$xbXcT0;AUw&Et
z>@A<kLnxgEA<HEBdL7qC>f~0fu(9H|ZhdvNmiIz$?4M`&-!|ZaQdBoW)B19_K!9N~
zU*_7^_Vf0LIkHPXFswEe)N)QrR`mK4H-eyA+@z#c<$6Q`Oijx-F1^V~kcl8ym^zM|
zhcLLGu!5#eH^ev|`77xOIfv>(6bEd7_YRgo6vT|f*+Mr^_*n@^n(dsnt<Ml&^?3F>
zH=xsQxzN+%Wc@cMke!E25m6myrKq4FYcm2;WtU_!TU3G}h_Brrb$I5iQ6=5WbP*NC
zVX~C-*0QHv<(SMx<-UrH{dpQ_X(^G?FHBS1S}KW>iKaQk<hbHs05+@>t=U2nLcH{d
zl~7Kc`AMI_GZ5#18c)}v^MCBouEZcdP?cnV3_m+rwZr~coEEcafL2Wuaisf;XRPMe
zmHg!`s1_t*uS_gR$$g#KSH1L5_p{}*Ap<f5$1ioUz2SA(g(D<3VFv%U@&B5Q5;BBD
zxSlW4G3N5oli-^aVAH;Q-|g8s$z?HU)CosaD;$-}m5C*Bj2GfDB3NAr0@Y&I#EjKl
zQ&o`u$2$4c`16IJj@Hb|m4*=_3lSXJirZ9Zw9Al<s*$)-qCQ7FxoBNOAD*nV_$XB-
zs)0*s<R|l_;vxK7Fz}BZPk`^Uwf<oyn;Bx54ov44ww437FpZzIw;?L|x5s1VI9F2e
zlYiTa-@k)7U}CH|=HxcaEE&OAeg`Y)UQvRe_GlLQQ2VE9kD-LWkFbI@NxRA=Kq@`P
z!h-&HSIw^gfsI0XhfnD1M7Us@9ALarjXbsao#1Juugs-J;+2RIJCh$*^IUroMod^#
zJ<7BMv&)F|a7JsM49j%-BfS^cpzv<=-cpMpw&%;0sd-aUhK)nB`WgReDZ`9llq!`8
z_=DpK&K)?qnQ8ittPm1kJ`yLBB37fzB_dyt%q)O7Z!Q<?2K83Vt)mo)lk8m7lmEHo
zawsTR^z>d)m!W^yEaC{|;9;=3si2Ul!>BPoI#O8t_+M*?1@;JhpiWgkOpu_c-;gp^
ziWJLvF11gEETy2{<)3aI@DB9I2MfjG{ieb7J;)S<5aD`>$W@6}i^2qAbc_28im(IP
zsf|mdGgkD{|BBJ{;P^g}^DaADkU^g<8Nr#BGfyE?aDb(h+Jx!^<}*PI!-KIssCRmL
z)bO|j^t;M($kx<bs7Sj-liKy398EWQN+!upn#nO3aMD9C2<ZjsP&0&^P5tZoUC0#A
zF!7WK#&Xx6{aM7YeD5iTA6p!8VdM*}zDuM^UDH%v`U=rKjQ_M=<s*nL_NUu&fh|DT
zJM29*{kQkPCoHUYx>R4|{`g6pnPxrx6fts&#e}=zriPIyIV_1P0h`f;vkXv)BZtG>
zOT@B{Ba=-4m~bsGs^uEgZGAdqvMHHmK~YIbjambdX%xV@eK{p@Dh3Q2ko|slbx_yg
zH`!sgO+I`ym(mzjw@oe;yRt&suhSSBO_*m3)KftPz=V2XH%Ch<E1+vHSDeiYS@H};
z<O^*c^N+)+s^9-yT0I!cZ!mv3{bVw!-x*EDrmAP%SY=GwbTACp+P<XF*-5AuNmMWB
z&mGMzHGpCsY1FF3_Qt6-Gb0!oVzeu1qseV8;xCkoYxJw|7N@0~nnf}SUz&L~Q_&bS
z3$3lpX6bUc%@=D@lq%vk`gnB8HJAaJ;Y5Xcb901QG9_}BZ(oJ{gh*NJ3Bd|w(g^XE
z8WV7sOW;l|R!ahuN)uh50odV$j<8*vAA*a12X8L}v4b^v<O-8KkFcCP$Igy=^q-OY
z-{UXTzbiIhGN1p|ZMsaoX9{J<_%U>s<IJ*tF>SF*-=kQwaoo64R9xwU75xgi#0Jqt
zZ4HU~D2GyI<{u`{U7*?-S38N&Mo2E5#x$NRm20NiL6qaFRk05Y;!Ok;3x?Ys;Z{yW
zdZBJtxp5}SL;((7+-^^cD<OfRP7(C2daYB$%o-s5q;!=?ttt?$>}~n+s7JV&#-=W#
zHX;6M%b_wgc05o*hMQdz&~%_utCCYO<+fNT-W3zLHJ{U3v~$c2e`JR<+znNZO;8rb
zVKpORE-56p3B#aH1f~p&C!)qaAsmU9r-t0gWBl1Me{l2AKQc0DwOHaY7@&ggK4Evr
z_iVRRZ4_<}Tb*jG)N#_@)L}0FrC2VR>riQH*M$v`1w!WQeSUeGX7w7wmus*%UtT5B
zI@>LLV{v3&Kl!vouWbPS&c&|<5PAIRDJeZGAWvM8xSHV&jSBy@Zviij1u<Vty7B|h
zq(ZA%qGX|Hb6_sK6B~{((7489m1-cuk;M7(M<FP^Vq5hy5pX^%R2ym+1A%hwKpMN^
z;7bIq#8@ix7yM*^j2Je^X4%~%Tw|%8i5tb}G%;^S(Zh!QwOg&B;w$dKAEy-iFs*Vl
zgK9Sp9^>#Y7;FgJ8p8q@v^w8`U4-(PwvT**;&R(@D`)oVuv?!HLW3G^Q`q*CEdkGM
zwl>VCLSUb<uR#=EJx+-{L{3CGO=W5=RB$X+xK*EO)cj20fvy)(3W_xw+}8+ky>`&j
z-oyp!<+e>dA2pb4_Ku{rt0jz1;07thi@7ZBz9jMPScIl#gF%r5U<M}<D0cfZit4u2
zKTYR2=EVS8*ntFzs4r|^9FZ0cX#kFP%v>7tvyRk|mNX=+Kme6)wE+LOqCnlsC7B`W
z9(Gqn79?cfVQ^83Uq^L^<1dMrXz1_mq@6|SF-;uODI5S3P4`eN7mP&lmLp@b`yjY&
zR5`v)Dvf1E|7awY-k_mXw^fCjwBJFykTAT3uuZHkBWbxY%HaqUCi$S?aY+a1{lYg%
zkC2T<Ac;2+QlF-``TjuRO*oMv!zQVF2lOvXBAkZj6*3aXE6N8n5NE1g(Ns)2TjZx2
z`;qnx{@5L!V{m>WdYv3|KBpv!X9}L^)&9f#H}Rl@DVxXMvWwdv4wj!WdO8tCFVLPx
z2L3Yy?-S7M%3yyYbwBu+NKqRfPnT(E32L%ix=D$onlIMoq-Rl`2DzXcNg{BYxGU+x
z5*ZK=$Rbnt+Z8ApQWW!Jr7=JS42{j$F-V>f{6;1k!t8_AhYv+JkRHReZ=j-`_@<X`
zlr=yaUTv$kLF73K-{`$PTi)x^5~%rOfAFRIXSl~ELSJ#oX)=Rj%l8J<6}{BR`)|1U
z$XrHO@n#%fwP=;+O6Ww<Rm64T8b=f0)(LrgD3a)jA6?G`TIKxZRa%^=<wvo8FO(_w
z+I^euhy9ZqENBJ^b%AIrattEz#%$6^7$!9dIT``|a02@Qm2&(iM_XYTI)y(>4Qmt9
zVk$R9QegZ0aoFA>cVED-Vq>J9-4E_631xXHrfghiEk<E`G}?HlSZHKYPCMQ~a3KW|
zKfRxBYmG+#z}hX6^e)v~<s%h`zA4+&a`r_cVv_$%bvo4`uTMNq=%boVNmb!!^`ncM
z)~c_8q3r^TkloK?B(ASVn~BNg5G++0(>4)#1~FbG(=J3xt@@_wU0vlRV5VV~dpxtO
zIHgVuMBuPfH9;i_b0`YnC6ks6#Xi8i9aarWVZ?%he4{?$H<{I90K-%%^J95%d5{-5
zB!SbYhbyg)4A}f~^eD+`wMyiSXntXkVTt0eEzFMD(a8vv$<z%h*kFNt$Uf>04o2m%
zF#gwpx4-Lkx|z^VzcBQ>q+B8y%z#She!zoVVyj&p@v?GZ1PD@&7fM+a6R^Xh<h)%V
z=ce;rt*In6<l{6Vb<h@sm`x*;TA`i0oM94O)dW-~8j&q?G2GU^BgF;|AfT!F+gB+G
zfhSAIxyZk_l-$@us2z_xs!itRy54TvJV0;rf3^zmZgo057zo|y!V#lF!|qjt7Q;K`
z9|^uo*d%=P!Al*5QkaOK#8&(j_hV`EnA_;6E&kM3v@2)Qn<ydPPp>*zqDVFf(Mp@n
zAM2ed+~gbF+f<$%u?fm2*!wmDWu1Pz^+iZBEEm#35s?gd*WBC}bP<AL9hczbK(}Ir
z{2E5++XYfr|75~ZAp&H{x|HP6Cl<RK+Ul^+*HM_BKB&zEi+(HL1Y(L`MuS*GNEN70
zgH5H{ZBGLhL*^XNz;V22&%{LeODG;M_s=vqxo<;JGr!dej-@dxM;I|LxDu>Vs?yk|
znkZa;b#d>(*)8H4dFEbMS=zL_!OtoyE0Lg39?9(7OET*2e%Nh4Rq+BgXRrvtiBB3x
z<!?^=#~%4&Cl&*GqVWqvA9eSqNLQ@F3b_EK_E~ot*eKQrVV{5>uvaWrPw<5s8!hr_
zeHzlih^d0X>CTpi9#zpPx*db;+jGs7d#w*zj5@;XXynK|W$1&Yxv*nD1W7fLm!SHi
z7UYjyE{$6N{6Qf|kG+2ODEdi&?$J+8po}o}bh%#KoM-B%TTTUkgnq%PuLDzgb9Xp8
za2WW8G%Ub58oLrqd|&`60E8jRPpq(4%vl9ZQLqKM%?cFgHsy#86kp=_&)FNA;CA!1
z3v2OoeD`YJbnNg6llXnb-roARxRf)BmZg}{_4Yn0YBQx63`CNPOI#X4{y4@bDoFZc
z^-H}RNdq;p(_BTXCNj>YxllT5*H}Gy?m}SI6pc0uKzjQ^$Ti-qg?XN$OfUR|_simC
zX>R+5jD$TumRbGlEQ}WNzO;~=uTU?Xw}bVf?tYfv4=*MVG#w*|G=;F369UBo<Du4u
zBl8{7w`Wlq`f6H~dXnRr0}zgE2!b`F%IG33N-=`NJE!t$enya5==e$e;MEC2a@qZC
zz`qooZ*Ui3r&jvuw?8jBQ|gJS4?ErCq>OC*a=q25Sb1eOzR8n?SGWbHK7S#PKH~QE
z#h@*l<DdYN(dRsTNrib`7K;|2x~Do)JE>_eBvWe0;2?~_Sk`(XrudhXc0S1TP(2lu
za`Kb37ORpA)N60o9Qmmry!E`2I9lB4vEBi3IgZVI5l#-1Gp_rZDH$pC>cKnTZy%Pc
zC-0iC?ek>#ev7-!146GJskvWGjx{u{dRDxzxYqtm&(OR0xbFRYOLxC*egE*ee)K9i
z^1k=IIpibXrfhy_Dp@iB7V!0pey~W(zSy?>n?PtHf~ou1@Ljl%@PEi6J6?H+t|m4x
zcF(4hi9}6|D%Z%D@wyrd4e@LKhG_os2ZK@BxqOL-b|-=b$v(t*rn2*9vzz#kp0WRu
ze0OZ_l37M*8zPflJApEMwHRreGqDtslwL=+EX_YyV49A)A3o8JciYJ26BEQ2HQA%2
zr^lAxc>b0|Je_?{Qf0aq@43fp%$*b=7MQ!E6CgVXY)gU@yRD-V4OQFQV?k~+C*v9%
zkuNr%z9gyI`I9P}e-7qm>>~PrOWIFAIv_YRs$T1epYiQ1;8sVccNZlppnv<ezZa@y
z$jCzAE60(AF<lix;cvR-4N~Qd<OJKh{l_EgEHZ8Fr3SOZBgD+MSyR~dQ#PYmB$<R4
zSBKk+od9(ljO*!5f&d*B8i#F8P}5^yFgCo>{Y_*DTH%A9S>Z7qC~Z9S5G*5Tzw1f4
zZ`qw3%~GFy49BEY{Z`{ulR5%-c51QdM#ry_F1V~YkS{<eTYbBQ<FRq6yw@_%V|`v%
zWoXbcf&oH-XtGA#qwy{lCfz2)cR@ZLKbi>!PsN>oW}JZR0{SJmil<s}&`HK3uvS){
zH^=b`oiZ|h_Z(|;CRf~mYy65{EPH|M>0*PS&R9B)%?QSVA4N)Mc`t40>0y(&Lq#sv
z#fvihxpwZOjM|XN5z{V?xkQ-W(-FFWrQgsMLlZa`SQdG=HpyfW>O9}bm4Iht>&|`<
zLtzC6LuhQ1lMm8-WQWcLOqsMIY#W#nhJR9vf^#IysZZF0V@>fD#rSgq6g-w3A2&>E
zK5B<Tzq%S5ts09_KAO^azkC+gZLl(D75b;0D4E`o9pK#;f$RVD?GaO2{gwnBwNGu#
zn!6;TE=Kn7&_u)E*I8tH_V~=^`gA=3F|cy|{cW>1NOVWVz42oDPo+uP*2tti!#@*j
zAbrdRBbLK4D13W;=7hjTg9^#QkF&qo|7t!Ec{nh%$+VM2w<~qj>h2UH8j<;>QbZ)U
zNxVJx2saW36UjQh!YTkn`-Qba+wS($elK(CpAE+Cudd;^yyA4Lx1kJYNqLqxLU3>!
z1xhq8x%tDCq2BqSX-7O8TW6Sqf}=mO@r3lQ`NfqBbLR<Z@M;Al2>trWClu$1v2(Cw
zHy%K@f&IyqMn?aZo&v`?4+;f5c`)vIfjPIqYPMIb(&fu*Zn!FAxvExM&&I|vPEuCX
zc9h-ecPXA7OtK^@kRV!wk-1rnq$<*SQ>f?8$4wR&&~(1Fwn+m=tA1s2ZW~itC^}83
zVhW^<`C8)Iza?h>#8jMSO-uF~^aC`+iECq^YV+V$(jT|ma`@i_R>stkI^W0k%KQ?n
z{>N?p1%}ZvKc2@I)oGfZT}t5EQBM5lm3=;idZUbfdeSRb|Cek2Z%jF!1x$oa^8fdn
zeAOtUK=N!|nY`6y=6kl#BysQg&8Fr@zl0;Cmfk|r)jULL<-hXf|Ask*MY8|E?6uq;
zudGoDSBMi3{%U&Q#sO~}c{z+Po-w*jy%G%{;i^*M2hLkM2o_>6fF^@_$Il29KYXa+
zJ~lLIR{!WvoHMm3|BI+789T_{iTZqz&R`J)sJ@A#P6cPU%|KE#n$E6iJeo{8l7x9;
zn3u-k7&w~B?fe^G_bP?Oqd+Q!LEGLQgY23_y5bBZu$CvNQlYhbSCoe+3ARg)wwpDY
z$;jI5MVcybC2A2|B~Q7qBD+vPDw){Q6ox^17r0o&!;zevW0lEelqy$%uU6%}=WaFP
z8xZ2OPo+k2e6#K6jl`h2+?UBX8^-0!V51W$*8tiR-dH+*`-boS0NN|}Pp}5osNc6s
zlD6}E5WcJLn~K%L=@quu)gdOsLDhEbx``KW+%}%`8YQ43&=Kag_Lwv@*7(91XNWNf
zyqIBFuxQZ^q#2V@D!FZdOQSr_W)a+ae_$XBk{?W0BH%ju5gCaN8u<&dp~5I-vf6YI
zOD!%s<^wgf$j`3VN<*Y1RRHBkp;-c5l<zParq5zCR|_;=u?>UOHwAY{plW%bUa8YC
zlGQY1p8=c7ZIW3<OKJu%UAKp*)5_<|6-D<<)xMvS_R%|Eq9)0%U2%!BVD9TGlHCHy
zZET(e+O8t&ThFaekw9p%ERg);$=?Vix9sguMen*&=2gt&`5u8BUSCdSzeJ|m+#bz+
zQNM?90VXsd1_8ZX%}>jnV?kX&YW_(o&$1&`4FniyRWZy3E}hx5q3eA+66^5Mp-JEo
z$w3{F`oiek`au3;-fidzO|0U(+m^`VG6G;e6zeb-9zX|>w2w&kO>)U6q@CbiXpmLo
z+TYjXTTSem^7lLi^C=?(26z1yK_I0p4V`(TRv-#y{{R7N1}M}8vfA!Pv%L|>mo;0i
zcSOu8T`ls3U;!uSN%M5G;;&+Whfk!flL=I-{vh{gGT9uB?^Eg%M`2K>Hd~0{r_NG<
zDm2NP#~lu*xst4l=L?%fj0yPIVk@Mj>&Lp=Dg15;Jq8g^^EB#(R;T3LP{pzOYNgW>
z(>Y?fyxzdf3UHI3JdA<KKly)@oi|tfkY~u4lGIxk%_49@=rFtA64)UoecKpL5U?+g
z&!^ipG1*Th{qogo3$FGw&s*>bJuRc3&)%~-$WM2({h(ToHEmiY-_BPU<Ia76V$EOg
zJB#dMeK2BwBOgRQlE!9N^YU_C39^tmSnmaz$<Cvqfm-+oh#bMG6Vv~4If&(^as55|
zjnkTT{Bz)kge)&?LICrL0jR}8)QxYNk@)U)bymv*0k7vdgNFMVPVsB)znZInqNX^{
z!v@;@v+a_ZSB$(@cw!XKn^cSQ`O(qfd0`0PBjS-D3MkB8ln24v-JG1!->yb%cG~ZG
zr!TMEM0wWT_PHMi9aTwjueYOkN2(Uq;(^lL$Yu~;?)FXthz>v`8%Q9&*BaZ-pRhAH
ztq8rQZPdja%cJkX?F_-?S(H}@s@HQ=k0bv$$Zs&`ZRZ;%&;jtk*eQbnP6eD<iTNRN
zy*`tIBI9y|M~IQ@$4BCw2$?ZAtYB-va<n`a%G8hRQr?0%wbQs<?;fF!d)t9Q0Wgt)
zMV!Xv-7U~FDymj_3kOeazJpLDb2ntw&0(0B1oho6PC6{eU7w!-IuFgd*LxiNij9ql
zit~W_KBtulfN}1IkY_od8$~}2{gxfB71aH?0B9k_?FV+mr>jxmmRD>1Be)Af<N$!f
z3`$~=mqg2(D%JI9;wUVV7i)r>bHH&7#EkfW-^ZBE6pWaer0;u9OzLU5OrbL<fCxO^
zw`F-K^+YCKL5c+}?rZNG+j*o2tbtzlezgai=~c_touW=lX3+H)>Li0-oom`I_gZTP
zy;p(v&J&xY>zZNukL*<jIk|j$JIfPi)BJgHUw3#{Rp*#2E?KYTB-CQ3eqrtVWGwXi
zeey|DuamWw$DxrIEwV9YU{|6)-~7#C?JQG6fw{pFN5pm>2$w0Q*t@7Cl?@{915;Ko
zc}Ds~8mF!bg`MScLjr)hYJ-erd~CN|2x+e_Yd8!`HqIS{ehujnijLV4dPK_~@yK42
zkH*d9)`6113Hw$W9RWPF)L6Q3452dr#%>0$T?LJrOsFMd&rHDy1*;H^s`Hgcf#>Z)
z+9Sz)*z@^}6rggZx?hjkE;=q`WIV!ISdq)5lK>3G2s*`-w6tTW<Ymhl)o0mU-S|}u
zjWGZg4f)Us4pig)%?WhlJYS#K!ct5!J@_+!-O(3XWzXH%)TN$xyoM@Gk>Wp}v<Jce
zx&H9@+8ASv4P^4|&VkJ#xjxUQ#LMSw*<yOyF5S&F3z@BwQD>_=lf6-KrIjsxW;~#i
zUH!_GVyhUrdZ9o*5TWe@%Aid=d!nSi>tX)F&I{rGklo<dLtiY1HOG6ePeQFPHE{VB
zN#iy5I^c#$t9K5^oXd6A$^?<xfU$sqsUtHpxg^)_6{F$T`5pS-#+8w#|4vN<bO9_{
z`rO21FDAVi{*bLRAEC^rg^29J!KPLCIn{ZXicMh+h&!#eST{6Tcd%25&Re;ilA480
zlj)L~5ti29^yWZ|vUsSq5I6CWOj*W4E69%J6yg8TBRURns<EC)+j@F-T5*Q~0x$lj
zyVKJl()tZo>&DZ02ICVK{_W~zYt9kp^>>PG=CzLJlf$*<bFT;WjwiRKx8xIk)PLmF
zQZQdC_Kt(#IYekWOoQDo`xmB5RipCMag}qu46Le_z5UfW?`>BG^5jXiTcco0b=SI_
zU4!r%NF{OraS{=lh=ueT2*^#=;YyG*>q*vMmjVm%_1DeE!^sZq8k+U@i@2BPis!h6
zH^bQjdi#Mx@@L7zWs}mKW?mmHBFEW6F+uGg4Qd%YINdh~f9xa$2;dlJ)Q_gnsR?RZ
zqO&2GLxXYvKVn!;Vq72nEjahSVM)zP?h+;}<eNS+4ee%|_#lq#mG8=&yrK})U2&fE
z>XHOqKRX1GjJ9-T66iK*g;FXFI1H<{qG96nA*hWD1X?eN(hB!l%Qfn4K}W1b`)$K4
zMOkWI!_u}rXw-WRB6N(xjXT@b1EC=A?2QA1Wvi!V&0lBLt?rMz3k>ErYOV*J_bn#(
zkNdhHpr3^D-FY0l+tBH6P>~<bx6D#*w4m)|f1Gdk;Lvk9__^rgYud;4<4T^a^baIp
z$6MHPP|rM_!T_Kan$9MmIn25SC$lT{_egjl^2SH$K+pjo^5+>a(a|5GQP}%sBrwU@
z^HP6Z0y01?!C`s(^NXQKpV+OjBI8q@`8v(!taq9z7pL>l#t>>B69qE)bgW6jcL_Cc
zD#V$Ys*RsPkW<I+0F3x!i1d)6RP>$(L{;Q)k?P%3_e~3b$O>9J9s$<>3p9u((#Ln)
zXHo)+-j{;avdKqGA?fyrJJJ0j`IvMC%4<C`Kj=eb*~^ofM0tgu6w}wTZ8v4i1AP07
zk5}8Wokg{gQM)IMvX1D>j!AF!y%CLUg}hWNCiWZntSdx-(?!=D(si2{mjE(0Lfg@)
zppi#RoxMZ_JtO4!%%ePTxQxcRZ_wkc$+Iud>o(E9&YQ=~ILM__sVX}!8#MhMODak=
zisHQGYtm9jkUTbxw85sA)R`WrOLDiCxPgG9hz;h1*Ve}MU~@1d0d1r;)8%S0bcW3$
zgivl2hD13)-#*2RUEdyCYQHY$qdhh}T$Z4~Ow5RF>dysXE40Vy3W8=S)kLHqsjDD8
zf$&EgH0G46dv`8QYaR$dWS6LM-7bnTp;s%i6H~{fKc$M>_Pui#7+RKK)wr%*(RSq4
ze42=x7blZPPzE}f`klrYwTlxK8!b6+|2NR-+T$m<w32prxZDUO+ZrN9bUgm<$NXcY
z4EoI>3EtxkebKJ4edol1CRz}YosT=2bz%3@E%V9xRTlJDC;i^pIHo1J_tp7Ix2qQz
zpcpL%+V_FRC*>^)anD+7kL{<sOf3?>O3e5hQelgB-Cyf5x6494btH8U;!zi7+UK@C
zOw+l_bIWA?+|t9(TMybP{`*;PA~~}z?~k}z12`o8p&y<NRTQnPt(c3yi+86}%H;Mm
z@1H*tyV{;{Zv(Y`E}CU&L|l6FmYd^de##CIHiiL|DXZVHTNu5?$nf8Eff4M+CQ(ob
zf$H)>yHlsVj@eMxP;)C9gDmVH<(EGWnCFw8?PM58Q9O5bZI?^e`AY-SSe&=#>+AKf
z!jc!9hZF2i*LvH(&(O)ZuolKgSix}H#^cYO;z3K$V`d|UkmU;BQ0VMaP;uaZAe<Y_
z?v)eG9`=I>&B&{T9s(W2r+_ZZBfrf(&9l!;DJSW4mVb`zpUetBe@6<l{981{>RwUE
ztgD{^ZWh>W|HYuo7K7PH%P|PLyE?r%&v2+}ee(!gzOE5VQ7012_9fPXsP*O5v&j4(
z*;H2n@kRr$LXh`AHf0pwG)UH+36A;SuKM5fJUWOi6gpq*sHgy(|E3H5H`zoF#xNMj
z0s+Q=>Hm=gVu4)(w;Vn=DDWR8{=b)J5bA<44&+m3`sb7U+i$}AxIokZ64&VXzmo+h
z&3M$+mGj){fo^(07&`$VT7t6rm2{d%3xXrw8S%_h&qeH-?oD#}F!v@_UvCfXsgf=<
z^8e#Uvkg&(N)$`oK#9>=oJz*^v0}l7$%ghw#rcjt!GGyuvY!?6LqF|jcAiJ&?Yjf*
zYsED+&;Q4R<q2iaN*NMH&MEy0No1f}As!tv!ovPvT(N(wQT7V~vQ?#qz-<s8SvUh6
zMi)165+r?f=_qn?Mk_~cFviHgt)<UcIA7S)9E$@&nTufd4hjV`f-#7-Ujl}x%6@lv
zBR>Q8KUY9*3+8i(-gpM5s^{Zw5^AMZlQ+;fJOc2jdm7<9+5u9mIuMEG%%GEY&lW2h
z-5g8_i?_(*0Ew*ClE{5?wIK-gaWe>CtJU7YVfXpG6EpI!N&I&n|96x9JrFU#go%<Z
z7Ag|j+S;Z|3~JH=LHH4%Pdjcwn7un>3}CBM0VUJ#5?2C+K0_EuEHUmXm60s4{q#sX
zX$C-TJReYUJ5xLOqzU$aEOuxxpFP4JFkB-KcuX1a2Fpc;!@1HVjvaN%T9b)HfH5oy
zlrHm?OH~wpi$oa8N7n)Ug-un!*kw!@Ns;dbXu+hZqGGej^agoA8#0Z_aN$>Vb#(ze
z0#)Q{P#k&dld*|OD-$646$3mp3GDY%?RxWs)})fqc;Pow8Bn3)@w`H@0;zhFZ4PBE
z5UkO(0ivZeSt*PrEgUS?D}XGYg$jY@6$k{1;cz?>8%d;|E0*tpS~m!XssS=$nliUr
ze-qs!=r=B;EBKL@$7}6OF4m)cNQG9tZm=N8{W?6$;3!*}3>27qV07^c<KEYaB#>W&
z)pCpNv&nxzFM)goFiLl+Tw)|R0<yjYUja4H&r)23Ny(aT58F{S*4jOARD-Z+WMV+C
zFdqLm8pcT_^g_gFv0!WxR39M)=pF(CQ!>JwXp$pfhANwAb}+cQ|H@-sp3h`H>Qyg%
z@=GvuHy*Qje3*<5vQjYa?Z@Be9na(a0p0@M;`@+SY-ZCWNk~H1jq9EcsUmc5K-%;r
zB+oj!2ZdOK?T75E*ZZrpsNT1%o1JZ=wKiAK(?i1(ofzfj%KbhU{?UL#p8K!}nN0e@
z(nn)|A9RSNe<SMr(`Whg0E|9)^T})<$)G*UG^++XA&4H3?s|s@q%>YCotVm`kV9#C
zgZ&QTDK<nPU1bWTrSNpWs9#VZ@{eFLbVa#&;vc|FGBd@b7k~w8ufZW^m_(y(_;9gV
zp@Up-+ORy{Pe`3tBL7?Dj8?h1j4@#H?re2=Q@&D%Tl9gzQpc1xIfq_&BxD|QOT_iX
zl|j;d7#jCTf}1cb^Wv{F-}N)TPP6H-Yf>#%=r==cTy@eE4WG*Nd_HaLFcm#A13xX4
zlD&yT5g^Co_-U)5To;FImq~s5I|Gl?>8FNtEMJKg{wT7Gli7c)EjD2^<kSV2y#EbB
zvJnw7B?K5a3hXCCq{~#kAysjdQN8B$-@-x!4eC$&@Bh>S;6MGAhwSs9zXuD8km}#k
z1^?#iutMAMkjrP;dG{IYKw90bb^*p0=&2z=IL#h}lqLluyahu+{Pf3<b!$5=)4lCt
z1u)wk+EtPd1r^18``cOSwpK-~Nu<;h%NF&DyGHl_;VFU2l>osATBL0@xvc4aFu0HJ
zXV+Np?xplCSWRHMGn^?AH-F%UKA8l!`Y#yl>m9~8w;9i+mfLwv%wWXJ>z!6UzEMal
z1^jslH-w@}UeKfhRAq53t6~6*0@)#G5N-m3QHPel=jc42^NE3QOlSEB_Ge2S#@_(+
ze|aB1ET3Wdaf4ug7PMSH_HRpE^Vl^|jHX1m<7Ii3#Ym}~m79K_HzA<6b{F*AL|Kxr
z1;z<%kR|8oh%&teHhlf5wE%S@+!ou;M3if-QBm~;wa*_#kH$Du{abzwawUHg<WdI*
z-4R7|;SZ#WU66ao(M6Hs9kV=w2laev%C6p7oaUkXRDnlr>4~xaZ?5|PoxTm05Z?>0
zFhZTem;+4*)M}wq^tFjo8D*y>a5?4)1xVC;gGHjA_&EqYXHgUa*{Sww6=y#pk+sh@
z0_l3Bw<|gOiplBbpK1d<eX@xSd=eaAaSka<LJxjwQ~hP66bR9yh{tah%u~_6cAcKD
zvA@&=$jpdO3DRBq3a#?)HAHM});ro~R<?p+Mfs6sXuF>vCp=s(f?0K@1@Z`+_0Jo-
zn>3dmv%D~7agu`WDktyrsuaF1tN&rqc)9Z7lV4lEnyQH{H&XA32wA;6(zB64A>)?B
zARTFMLZ*;B*R-n#UHwH)w}i8&y+4AZo(UOT1-N`3Xk#LyRBv!km@QUr=Zi1;1xe&j
zGf5;s)y~I4sBkzGA2opzr~5`=qhBjGA8+0HdIvh_pL=U^55<API9xhg2qMG%Y|5*l
zJpJrghpYmz&%arpLz9j59)ZQI{d8+tE6zSX9<hcwYx4BgdOj33fpD7y8Oge4A0{O)
z+sEt9szuZE5DTWS1@EzE_PCMZynDHyxlBPd7cuhYl#}kUhJ1*nhM)G`TK~Lo`1w-w
ztJCuGN%K7qgXz1E{Vwyuuy_02Fx#^pr|PjGT=LJcbtsB$sfO&}w9qM9j;Wi2_c!uq
zJX>#gju+dYFY8q%neGBSF8A8SsmjWT95-_8SIGt}PX(8VUtCA4Yqn)L(fMBP(%wni
zpKlUgkvPN>i~D=S1cQcc9kCjOsCgO|JvrG=M{D!s-&91M*WAKRstmjqt)jSNUFnLQ
z#?1;rR?e#l;YsjR193JTPv4Eao(0>Db=}{PU7rFl^Tj*18;WvYpLs62{BnI+N4PWB
zvB|Tj8rMA<nc$`F5=4x8n)uI_8Z1iV;&JCsMCXXLGZZ~+a<8uLhpu00Cse|cvF%6g
z{ERNd$se<&dBzne)4X1{@$KIQk*w@><gq{Jx641q-?XQ2EKZy|f@eU;k_UvOeXqQ@
zr+Mkfmr%tx%kmmuv_DKx6;ESmIMoPy3b8s04VAYG&zMe#xEA-%*Y2*a=apqXyV)$h
zlEuvYp0P^n1w(S*9bW2Y7c<ht@CglrBq{GV_-7UOeZ2^yWbXh@KD&BenylPFAK|9A
zfNoza#mT*`Rg5_Ox;9g5%22xKvz6vBi*fGv4#lUtWsSQba$WczqNR7N<X&ErwrTgg
z2YQQ6&v-qJ29<@<Dc@fRlKk_AIh{9&iZu%>loZSFPw!WVGd?}O_@4GDY+i-Y#;UG+
z7LugO5nio)4kt*&@3_yLRWZ?(OtiGEx_e=q7`k6aOtP?P6Sto;-`vBsvHi7vR(-z)
zNq0O|TS#YyN|FqRu`@J??@r(+%LP(5T(Lc6{@dcit_}U<cr+j5yzVY3!}~~kHG4f#
ziZ0YJn#F0e?YP6S$l@TvPrC{EiKYOK3BULWHx8$9<<Z7$B~RSa?9C%8q7ECZyRt}<
zf+fF2e<d*8v4HEQ4^4t$xp0i7V!XJ6)h=axGXEReKvbNn;}68+oUPAHzjw416yom^
zm9)36JZ^H{K~9_}1khjI#k80(Rg#;E8Q{Fmd(JgWej07Q{jBE$cyH3`MWJ)u$bYtw
zhl2G~43&Be>xmkfORU&<dxR3}lSV@uEEZ^r$HGH(*Usoij!oR?>8%UMW;14IIjPP&
zuN%x&I^w>XC$^oxL9;RLV$T$UiekLP7OYpbZfrVU6gpmB9R7AU#m=_p+YEhFVv}Xs
zK5u_b44?H7)vU8575(}v{Ku*q&*Oz)(0z{Rm(AGh_Ky8X&YeGKGSMpLMA8rUCCdH&
zPtm=IvWfWPyhl$X6FiF7XChDV79-&IYsAZx-EUj9vW}<)p<fCgqi{JzIus^`qR;%G
z5EQ@^RNXljD-JG$@KMFO%2`);j+UDxRY7^H@%0hb{nwA=S>`F6!#dArl%4NSrS5m2
zLF?%s_aA?eFdM`7=;E_T^@<*h+H)3yELO>KK*j^Fe)nq$DI9XAzvDid?sR=r1g`AP
zN7H&_zF>LPw7r+M+t&Fy7E?n0)$>nG#r+NY+if@_(7+oh{>8NtX|g-nE3!Bf4qJA*
ze%z-0E|pjheru8e%BUkXR#ZQ}WUad{-DF`TF(}JT;j7z*k0CPnev*!hPKDyA2cFHt
z!kBoF9w^w&H}UCYMrQ8bB!oE3|I!`n#%HtDrad2mS@q2IqI>5<r%7eGVWN3F4=}c!
z{iW!)?_xfeY>*(^rYL?Jt;Ir_<3v~$`BzsVSQ*ayuePrn9`|OQI77v8MQ?%3kP{tQ
z)(HXG)dIgc9fu7M>y>TTPl|}=V#zkOf!@x9tDEhpW7NRnnwjy|p);TvEG7Zr9R9qY
zruCc#zvDf#&U8v4lJgS6^}2q+sO`=V&v{Ax4!bF9d|cHjb!{QAod`|od4cbFB4+T*
zDGbg;%R0;Q@O(=mzJAwJt_IV~Y0U@`ISaWA7OgDTQ?Q{j2$c-lorCNn<LQF@EA{I0
zxn1k=-Lf_}{`~$#s{pT17~UKUru=(;NiEWTviNIvAW!JZul)@Bjcmbml~>l)<;QDT
zUOTj8&eNLGladK=07uomWe<0N*x1MSj89{hcCo+zbz3E)K^EhDpetvQRf1pWW{j!h
zRy;x;x6ikBlMhrCexf;nBeBu3Uv@WIFRVfj4L)TUGuSPBYz3MQNSAKzaBZ5_Bd1zS
zSkS&ff3**O=JvS0U07j#1Z>vfHU6%!q|?jB)yw7maSX?0(&y~e<1``3X!e}gXf3^&
zK~*~X-EAPn5dE9H-F7DNIcS8oE&KdI>7LVNGF#B7c?<b%joEyKWKq^N5N#kr)ykFm
z5wN<v_4(zYwdYgvHOAK+aWrjQx}-ham+kv?ph))>Aj~p{mDP5FZ0IMx(b}Y>?cn`<
z(67kywRwfQb<EXodlMOrUWJ0UeO;!+?%92OIr&SK&X3~TRQ5`f6ScHMOUwQ|>i^5^
z^d}ryb?FZQZ~UKk>{h@16z?`~mQy|}-t7|MdX9>`yoBAX8(}`){EmXOLhXrq5PSV7
zOdx3*NPwr^>X>x=?pcrfp|va63ppD&R@_8spii2Ua5r?F^nKxT*o}}sRr7h`U52^e
z?qW{23B$R*Et>2=;&q)B`F(@Ad^E|*hp=@Mf}LvYc(`FyFVQgdzkXly>a=WKy{ln!
zj!FMDa45mB<_vx{!Sg7z=6WoRInfwr(ka-X4(|s6g__ckR&%C$-9;lYz*z#7z`t9s
zs$DEp>GxBy?akxa%DRgW82R;FHnVlgY3rr>=BTQ@JM=lVDJ%#+CZQ$JZGY8CvF53I
z#c|n+W*iX3i9KxYZkdwheG`K&r(D6;Xjj#7??!kzVl_RmHH=;@|Ai!a)NR*by^VIS
zMKHgg>6yx8KK&;~KNH+9JzGfsv+B8?#(0!UEq`4we@Y<+gifvV7<$qO@6GBDyk>Ip
z?xs`6SC^p3jGg7m4w<z#xvaxC=&uJAg9i=L;CD-m;xrz4oae2)2KFhih&;Mph63B9
zz|=VSbk`_CW*G3(h1)i>nZ9fWWiL1$1G@Huw(Im_Qizy)^w^`_1Weu8^>E%fqMWQ1
z=X-7IjicU|m=_n?<e%RDp5;O>vuH9uJ$%HGr2+`SV3MPz9M(YJC5#VB;RmWeWuNKr
zJRS)8tUP@z2cQq<bx%NU_>`kAiITSY&n<Wl!|6itlO{6m6<p{08~5!zHs@huS*@Fz
zs6zc1<WZ-R<K5;L#$sOZ_l^7C`Dg6h0!i6g&bP|eTV+XS-#j`~^*ZnGmT4?HIFYz7
z@I~d4#XgNVx12Mkx<A?08jtO+H?F3Jqb@|zBSg9xOOcsOBVN7j-!6bEJ71{4@BS$6
zjiyp3>Hd^gIG$*XWZ5K$Vp+|Rq~=+;N5f_{4u77fmmj)+`BV?p{GQpGFT3t}gzf3m
z(R<(GVt0FgPA!R<T$je+s&K^_jO`9cVC3zqv>Og)I6JxKKVhP?R=l#Br4jmq-)b^Q
zBn@+2twXQ65IjgdO2c%7arM5x?t4B8Zr?A!rZwt#X|GL!1pV7+rq9{q1k}K|UC!ID
z21+YArjtv?SN1bDIsY_UV6w(=-W(RL?qKO}{MxQxXdzY5>pCCCC-x#2+bdzl*j1L_
z59r|PF!~F#RJkGp7!MhQ{jK@Dy|o(WVcIBnPl@tKC`mR?CLV%MEMInkc@s^?)~CXc
z{Jk!v=CR{|W)YY$wUTUmxkjTM?}Mq3vpiUQIe55IkJQoZH^m)L&O|G|pmB55yrU!h
z4`8O;wy}$P*kBYZIL2rORvoa;zpnssj#mEfOVDB{j1S~5yE{t4{cF6ZWwnO;+Rl!C
zV>9pevFU~2CdPdzFnjcVQL+WlK|1ZvFrvXsS7{uR!*#2L>7FeowYmb5{n>Yxq%fl|
z#OlhmTCR1lS064{oPtjAVz`d}7~tKR>b5$alkLN^ujlLAI;L01_sue*H47~))lG_{
z&inXmOE<11T0;Aw!+5MX=x+pO-wMYy)tgUu5mCrv<$M!Uw&^^hPZoV}k8!ILQKNLP
zF(-G}+aIb|)!7^sZfBuQ-t+cD!e9zCRh>#_ILIxBIbAkx^>3g~E}+{7CSy5HZQH`^
zL#rffj%RX97RY4gF710pYUQABa<7-jW*K^Y?EU!~;hbZ3ZESz*xX_wQXC<n+Wu)oB
z`gXOTxJvr9n=tA%K}ZAyiK$I>`xDh_EOqYc22FMddq4e{$>{eR*6vYY;Qcx4N)b*$
zs`V-n8p#bpFuevJQ$7`*UJvy39-&t&MG2Pkd0yw+oH~@P&o*0~ULE(0XT}jHctuKZ
zR(!S;(~ww6!=p1}$v77;Z3l1uGY!2R>YFDYg)GPJXZEcF11%?a-S<*7$V|B(h#vjS
z=F{ct9aU|w1IP9y8pL}2K1zzziw2L!wA)Ygshx#iy~?^6w_B6UIET_gf^sV19W)b8
zJNCaa?3C5+7UFkclEK9YBJm15tP!dK81_-sPRl)bJe2`n4+JF6P;VsLv{@`*HI#*4
zc-7X!5@gUakM48CYYS-B9d0>?{8256WaUP{pHSyn%Px0$&aKk@w+k$I3$9I#={whr
zmro7b=d;R}dM19}o%DU(>9Bmd7EaPW^W3Kk^OS)lw6N(^u)u%%*2ynsQvMw9u6p4-
zw~xYgb;g~aszjDYu+@2FGnqWa@mTBiLm;+(n>^D!RMI!0_oA}XF7>Li<6%X5IHO|X
z)8U`yxU@Wb%)7tDBndnpsxK$;f!X=~Vk)Yr*cfN!3ck%+$2Gzh3}Js<jIRuY{a8bL
zxnDo>YbEnK-!Bx7&^I()Vf;}X`|*mSHl$i%{@hdJ_jbS#jOV_CVqVdhfEbM^^R^MN
zZO`qxMZ(I*Edi)h(4`IY#o)_9sUZ(~nj42TZ+LxmV>@)T3pQVFi!M+(U3#mT6m26U
zW_G*?*Mkbv!)dke5A;x89SimfFmtvb-Ov#^-gsTE>~z9WI-81ox<xa>ciNHnRZJCh
z?D8(yJ?Qq!FO0nnITB9cPsZG;jf8g?o3~=!IqmHg9U|RoR~6W4ce(n=@6DFyMnj(E
zmO6tYSaLt>b40Jv_|y2t?)J{lMK`L5n5=7PU3vR-VT@V{_3dyFpL|T%?~o3zS1QU$
z?=2jZvpx^QsBU~U1B+9Y4bh1NfAT2|+)s*3Qsm{PmbiH~#yPy35J%ERKrAKYhLvvB
zGYBtES?IWPf#;d5vb7kEfi}70$5N^UK@yCHN}J&?Mkyw_QWY}K_kmJTJF-u0B5@B=
zV%Y(|JH4K7Ra@M%+h@kDOTpiirn%*4M-%BK@6HX#UR&s1sg?11?iK6D2E%~BZd)z+
z33|qE+<D{AwyfA%0ab_3NYZ}uFT8jP;TTD<!3b>1!K|AkLyghA_Ppngz^p|lj+BJI
zHtsxo{ulx4nC0nV6x$J(sdqQ)+#{J(vAPo(!}=`QDC7gT#t4(-{3PVXH(&J4RqK_z
z+h2$%+TP!Y$}TRIMq#w`=P9owImEdN+VOs-+pw8v3f2WypW$)U)%uWeN&PLFah&q?
zCl;MHkyYJ7jvwupN0PLl``%86IF-KZd7AWRBZzd?g^s8YRtBSKL;Z-BBt0&#iAcb^
z%elcA25`3|>P$gO`WlMFO!!uxeE$R#8GiRPVWN+7d#&SpXc==fl87^R!T#vt+blV*
zaL{0va-kbB*f<&s^0i~P`O_3btQ2(+RUgR|+YOANaNaN56#5<f_&v{e(b#6(VOiUm
zQ#xH%Nx9xDP4d&SwXSTPEP0VZ3O9IOw*4*yqRUMyP-)A!Lx_#|wiDJugJ>0qpB{}w
zV*2*t>eg9(T{g@Og+CkeYO5GRJ~e9XEa#vAQPbrhbFY=`%6l?IY&2WMyAuQbeq1N#
z3B{xAYIiG>K<0v9hAi8~ZbHW)S-&9f3)SA4I_Qz3!2>gjee`V#ZqHwrqQE8f){m!$
z4*GQMci~JK9n6dH`X)FM_D&o~>~<sc;dWTK+vnDC9qE1EO5Ue7vqZTRbYasOga31N
zPFIGTSzU$fHG-eng0}sFX5(=FmxyxKCXZ=clqvZXmah9FdWm%Fe0d6|%eqJGWA3-N
zo8GAOL)N8lorp}>gtk4NeV!4l=l0haV{vgDDL2PTuVcUR$9G<llm^W%x0Xa!ia4OJ
zFJ8aJ{WZEs<qtT<v?sTth7hMx<Tr3kEjkAiI@7lXT<``g;VetMwNlIcYuEc@_rlG@
zDv;f<l=Do#HrW{om9aufJJ&~A=^l+xS+$=Z3%GvwArg(HUL03;a1cke7e6tgc^WvD
z*Ct6Nx#{7AD-YDOt{q5~>biLrjiy#_E*2g=>*2h-CKuwU|AWZj4x_Z@F02U(<a^}*
zun%UsdSq;J{1kbb5kiTGtN>eE2o%5UBMAtC%W7M1eBpj}#ba6Km>pHrmLn@meMN!N
zF=eH1^hM^sv6;Rl>nu_!s`JOT2&U1X8;$#iY0Pi8Rc4Z4?(DFUb0=eW<Sxq+Y&;Zj
zOq<QMBxq9ha?jW6qjE|?v%%AtzK5CXXtirN?!&;K4icQ9GiL}m+@GMy#!@M8MG9tl
z4y0@|&~fio2^#g$^0qFXD@_=sn^%u>UQ4Mu*)XGvSfn;uE{YiW(}$Mci-Ztc^HI4L
zepSBaZNTC;fKyqMVkZjWn8qsF%ZzGu)aI8R{8b8TZ+-To`8jggXleGFbyq(Q8B^oh
zHI>FTZ9@7Fs8B%^GAX)ELvQirEnB^4HmA{pqVMecGN$DyYc`|Im)2o3-zwmUH+&es
zH)!~%0n{C0YdkIet0AT{4#%1RJ&`dS&9%quY>(YlS5-VG{VhClQ-0TVmvh(m1{FoY
zcOK*EO2ym8p0@Wk**J1(`*DZq`iKUhIrcvTJ+r*Ykq&F@Pdq{~&d;F1uZ}8krjUU9
zv>c^&6)lv`IQruE$aL6I$$8$Od6n`gULm#IooYwcP({`K5tnRD?bF83(Gkw0loZe5
zojyic9+YKOkJJr%?N>j$-hH;ol-!AhUzAbDaotUJr~7FWEW&IP{4>F1lbToloz9z2
zc`XBeoQS2F%7w;i^m;gWpTAI4Ly%Be)OE7XF*mw1V$hCDS9HP;!s%}71+OD<AsEPv
zYIyyZLFK&<+jO+31142X)FoBQ8g-2wRQc)=*<JPVC?9p&ame#|xcX=c5&9D+Fcm44
zb{;3U+K_vica9%H27wN;{U}y4eQ-GYzU|d9F$3S}>dt<*excbO3g52ji^BoIzU=@S
zLRMo`8domi*g`DSHpE1n9kC1myOBgYnvf&$$Ur}2wYH!atn&Qg+F}UiIIb)l%Jz9K
zhTZ-{5@Rflu49NadHREf>;4P-?I`1_8!nwu`RC<4?`sotBDd0t&8xkUGS?pc5$wWr
zs<rU&t6bx{MYzv}IAz)#w|!{od)Bz5Ty_%yPgl^`M!Zvtj3%I|7US<!`r*@aoat-2
z&~}}C>qzr-KUzj~qD{x{6m15I4TIibujRJ@0cU1xgt7fG=ao$<_6a34R;iZG*F6gp
zUj%V}f>=x)`(m!!1B0GlfyA+UW12&|s_&}-gPRxQX*Lw<9_#&Z@5Ekhhz4K7^sLA8
zJ{(82LoA=1KN?)H|M>pu#X;x?`g%T*BoR#8^spr~+x2`lBZBAmC>-$o%M`$V-&|O`
zcx=zx<89w}{8mt;oK;O|rBj(4PbrbHTDm@BoiD?R;_rT`cByQ)d*w(LUU4RyxqlZr
z#b;f&^<$KkU0@lOQQqIUC!TqNUhZMHXluPcAsj4))>}Kh<uGFwdS*ktr$Od<$V1Br
zVK_fXSt27_ClD9Y;w4KD0qZc3vi>RGvqM^e*U`>pGtb1>9YB`G{HS^)q}zIT>cLWz
zz;WMAh_!Ay%mFM4s(y}hZ4Czo&<yZY=w`#M16{oPF|vJp`nE$%D+_*5dd}0<NOtY`
z|Ksc}gW_ts_3t}KfZ&jz34{Ovg1fsr1b26L1`Q4g?(XjHF2UX1T?V&7-krzpbDs16
z^j1+*L)F0E)6>(vdR=S%uGvjI&mT7%t`Z{M>n*y^9%-6F&)BWMg`NUaZ0u0{R5@Ep
zys)3Hh<V2HvZa<EJayY{_#wKxSJ=<tg#ke?z3d2^k|4`Lf4Zyyl~qaF)(Ov^`Em`7
z&TqZK)!u)$Q*(GlJUU%IeIo6m4`#3*SfTha=L=hN{#oT=#(wv;x@T6{LtMhn5(_9M
zauQdz%V|~P?8Sm{O&C#3Y%*Ez*@&=du00X|8DbF0l`E-`y2+N^**lrS`yp?21`hcz
z#;Go!fC{pjSEMly+!vm`NYL<QYEogxene|ZYYzXFEFqSYf2T^BGV@d^{j7uDdTm@Y
zOGI;@I8IKZTmLR5QV%pGo}bh*wiG@O9K?xZ;L<p6)3$sPrOJ!3xRv}PT|;jxXWCcV
z;?nFD=xL*3VPif(smbU;ms3<z(Qz)Ct|zp+o-`}r*LgEez5?XKfKzuqm>27>(7rZy
z8_i`?In6Xb<gwb9E~Zp6Duy{>eT;DY_I#;!bgwddHZCPzlzv2QlO>ePK-;&l{RZ=|
zR~b`T3Zf>vT}x}yQQ0#YURCC^=67Xc^JIQ@p4<v`88(zlydBny{rdp661FFMtm=o}
zGi^KvJj$TVw78#%j@KeOK>tt}FJa6Grj{&(+nZsVyTlBYHxgRiZVB?E?og6YOda*C
z5PmjKQ4H9^;mj-Hd-aI8yl&fA=*w`p-tTJ7{<|2a*qZ*2oFs)#KBW7Fi*GQOO6>Pu
z?D<9``Dy+O63TaEv{r*;TTc26$pE#)(7@@yik#rGh?ruex4TY9mJ{!Qeinm2Lmf1o
z-%K;cFaPnj-J^qMSO+dbo^8W*H+z^CRfKZVfiCQe2dO-{C(R0&4%<sc`YB%~DO)Dd
zdZ2Kju^nTKS1GXQa^U(IRaQ#4j#|~Z_|n&oZM^nn(=p^R7p$=9m=v2-Oeng$3pQE`
zZ*1*(Uf<VRFwn#;y3E$OXUYAuPV?|)%;`XgcRv5r#L!;2l~JkLqoi%?HG54dHqpEl
z`+c2uGn;r*ue1&=zY|?uD7w-R!d65Q1J*9Cb^fCw{zQE<t&2q%pF&M$4A~<2mg@e}
z&KnK4pWhWKINc4jDqz2vn*voNYCye18y;+)L%C{Ega%W;fwS0t3!|zs(rf`nvk)BE
zW2YSU#2BvwB(W7WiH(3HZY!c6rnoAls@mo;qAB}>xiqs(iA7aakw5)@n9)&=)zrrS
zLtvH5hPM>o!T7A^g01PUlD(n6w+rRdcDGz1HCsb<H@J6HwFEZ6gF%i~Q;$dJvOit<
zbIL+}<8Z0qrm7+SUkdG?`Vd7^fFz-O&~Fj;9{|R`#eSeF@EUDEz1~CVFwU5?j~J~Z
zmcd9}OFD^R1lIH%sM|^CaGnl$aB$J#cKW?<nx#PpC9jRppvzWz*!SZv_eUl9nkI6a
z`o73EBl)*zo%rXWX4AP!DRUaJAd>Xb?bNV1A5jA(pn6N-vrV6Q11FM9_PN+8%y`v<
z`)@Xzs+V@uuT+k-yyT$s!x&ZuZt)nK{6*E)DZ$SrGG!wT8su%}YT^O>otS-!X8$S}
zJ7m9;@%p~2vIge8?(|;qa?Y!p6_CH+culR@ouaIWX0ckvV*aHsc?W0<@c0H9j5{76
zZPv<{e<_ia^-t6}(V{=Bpv(`wT=I5PVNgp^SKK33o-^Pyw|__A7o-y}i@4>q-VRZI
zy1=<?_$BcXzTS9Hp}I%p>?=O8+zi;@93Oqm>9$nO{tUOErHpt%$8Pzw9#xbA6Daj_
zue|K?`rSLY3HOQoWSEoh?ipZv4Ix6+S$wVkUS%aQG>l2<e#3;hSbZl|Dnc5{H}>>#
zHP}lHA))$kQFeaO>-Gy3B9u0wy}hQ=Euf?|T{a&N3DlLM1n_N42ECB2dtEsZeYaFs
zyyfjfEJ+|TNuzMsAGrV|*!@)R_K%GB_fU5dM_VQ1yfFnCuC2M%w093!HxQgB5r2N1
zi>vXeA_o)DUSGlxRa5TRpWAH%Zr5wDVcj9Za^1UM_x(R{I+@+Xa(SMR+j*6HWp!x8
zfFJXw>0XbA0TL<;0}gw%QG3!yA6V3U?uV7jDQ!PFRjIOC65MY6x}3TaT-_VJ+8>WG
z9!-&KIOEp5UiW@h(*F3E>T%r#8Efuw5A%S8u^N^)t9iesxvc1DtwyFx7iPF#qd(jv
z=;R6y$46_?%6l7^>3}X^L6DnyE!<BWP7=uw=_xCebgw77hx;DXS!T$)!<jruMP{SR
z+>py`bJ%sSZKqkL*?1req4}1;HS-K^P7c8?z4c+K<Z@~vtGnIgl+)REiM*EnwGGU}
z+*M2w!F?V1^s-#hoF)KmHq%%|>1Xl~5)bKi7#{)-ka54IlL3ncnyPuCo0rGo3MMUX
zv#Hu5$^Ng4;8x%;QE~V?oPavcY1<Wk50GL0h=7=M%f&7>Dr>CC&X>Ii4A+Y$hIVKA
zW@UIP9TaYj(iN!g!fp1>kxNcT?>m3Ec|C2ysgLg)&J3&va5H$~yP^Pxn7BXL?KRN0
zbOK75wX2}X(nHEziF-t^L*IuNt%lXT)*sb+E}Acmk9T-wz&q`(OL{|ZG@j0`P~|G<
z^>Ap?5TQ}Ir91FqJI?)F>v3^dFcw>A)#6O^Za2pIfKisRUFQ#GB2VdfjwaMm_vs`?
z7C{|lErGBdB9ji<LU`C%buk=_CVljBcCexvaZfzr1_4sZ?sNo?b4uEH8soB6OVrsa
z{5%-n3i|ZvFbJ%ragK&d2Hbg!EYG7@IOirz6Ce*6ZQWmV*TXKXPp@TqlqJ@%bBzD6
z*KyzdFi%D%op`V&zCP}O_qtY-v91uo{S{M@bE#C=fw`^pIQvHrentB?+T)hdba%iN
zCtAJEVi*{c&$P#TXO`RycHJ<G)<K`%olGBHJv%(Rf$lp{o4u5EAPU7i?|aWpM)4!k
zU)rRV;B?;Zb$Un^x;GH48ZAiD4qMQ6?ARVG6d6=;ySF3Qc`}_ra4dE1A?SBu=i?#7
z;L*wiZToS85%)tEzy0`PUXhzHT`xONn&OV%o+;Z)GVIhEb<XQwhP;*U1k04mnjwL2
z?GSkG3cYTJ5NR822-caV`sz4m@7NZ+()UdWy|>ysR!lJVcd^r5Gzx6fiZk*bSt5Vu
zI-(b9<P3fNp!(6E!106YPL!zTdoL5ej=C2vkM;fio@b}~L>jJ->*d3=049m5V|>g!
z{SB1qUp9?<U<T_G9MSOhmrKNl%bM|M3U7RRf%3K^3A6PqKgZ)at$#==bmIPY2EQ9E
z#NPNFQqr`GoptP6A5J%9P-!;u5wwD}6UKdE+3hbPiK@R3ZjZH`=Hp9vp9CZOly<P+
zT~FuA$Na8dCU9qi>eFSWGIgez;!+K&KfUD~0A$zWDU1l8m5vv?w$P_jR@};cZ3c@y
zq^6VZjzRNTv@*>mfi)EY`|(qkMj!kXXC1G}b(mAK%hqTfvE}e*7kfYO)i}>Eaw7s{
zYkGWVjT7u>RgvQl&b}I(ep1FDkv)h&guYI5UK5U|Q9BwaL7ecu+8dc|embIIe_i)J
zOW+j=kA2}hrLt*zX>O&rOhojPn)v!wx?Oq)H;<d3xYy`ldVIin>GtHr&Av$=nTX#N
zg;m?VjQE7R;b=E~kWfOP2MT1|*={T?a11#P*4d%-T$B2mJDgFcrwxQ4RNMVwP6@G|
zS7EGd0<+%xX`Zn==tvbUkKc+4Fp-f`!Hv6Eo-Y}V=ldBT5pB?>@FHk_H5U+icakY-
zXn#CEyQ~ASRi~J63B$>Av^79#STv|;)AHnXd-XH?<Vri;7Gwt&2lII5frhc5yg2t7
zh*w+CO9?UGv1+=95gTgNWoW+MmE3TrCiN1n4iY@<`2dpXullNEk;$$_UO(&^7)a0E
zT8t;$8F)*}Wmd_0y;b+i>zd%&VvOz3j`$!C9Lj6!3r8BvquSe(s)CUuW(`QMa5r8j
zP*usnFyCy2uh)W5FA4v1JV87`%Z2dsjgvhY3RyRS*Ev6{THo8tYuXY1?W2uhS8!T`
z)~o*ICLl`93?>OzKZkP;r8TU5$q@~+pB=i!e}8wf{b<?@Hqcz&Zb#>rT6Q^6{p)H;
z;(Gg$*z?={8SK>-SNV7V#A2m^U2B=bq)6u4^<lRl!|-LL>5O64A>zoS%VX-#w`F18
z$KV_?%uMnCNN#&G(oe!S(~WmMq_uxe6uR*58KGdkwSKWX43D8LtiaP5@HqUI0v$bY
z!a9(`FA+gbQ%dJbis{)jLS{RmQ5^3866GMLO&f)m=iVE*Pb6^2css`Fyg+df`$;gd
zmMXa2b2oMyuB2WskGNDzX}R}N$o~!4E{Kq(0Pu<_)ySBrwC*tQJ}frWKenn5alHRg
zA3eGv%X@5@Alg5c_HlC?3g{V`kwx)biz?NEJE0LJbSn&<<Ts5mMhM#mNql>5{>mhp
z;rUdRYI`|&xswwyL@djF$|P7D(gohn@cII%Ifg1@QT+tOaD;A&oP;l<tl<ZA1H#LY
zPT~j&(OIU`me-eXxfU_>S6f}XNuz|8=BzxyVdneXm>pn4pXab5D@b#Smz-5&Zoqz7
zZ^Ga0OkM{x1*tR7^Ny)HcNgsR1}@n(PY5-;YzPw+JMYDg!U}t@@%{WdykAqM<!nKO
z#&#c{ZP3f}DqpNgKVG!s1Kbrt#`H@^x}EX8^W`JZzZ9u(S@+CrJ}0NoGg8*T;dX9k
zf36qw4v%gPGnh4Z(LzGJX50%}i|1JB_#xJuBbmi&ZvX-Bl0HY2{UjuE(0w3Rp!hCV
zw!=%N<+m<jHR9gkTA1$h=TSfr6St=8_$8C;U-wg-?~9CU1A2PY{JDsbL&IlSse8+o
zo?Gkgy+VVC)nNF?+I{nk9O-4tj<i(U8Tqkp-E{L7B2z&f+&w0v7I0G#(xg|wUfhf9
zTBLMsLeUO}79=!4;G@i+`X4>l>9PjWD}7%8#`<Vk%yIAX<ok<8xy0uY&$}-9gonr^
z)uDX)edwb>tog9F{dmZu)lm&ugyKK;g!$X;STvKzFw@0ir8gsthu|)kh{qsOr3_;2
z%8qP#sT-@Y!6ZV+G8)aaj=N@dI_DgPX_Y7Ml^L7y)L9|eZl+h6ZheD29!LK9eMwBC
zjUXNBzXB>YG&*0kkL2nEMp!;(3i-w2<|d*vMTS6`a2`-eKyu8f4(G7$^<4LK_X1hv
zW(r4xq6J*@><i*%q+l+0SnEU-f`Rhnxuhk2wzkN2tCm||Gb+Q=9R*p3WMr@VyA^jE
zEejSlGoIsVsU#*fx9gKb+E2JQ*fC@>Q+51TyM_Eg_Zl)S?C!UvBwZ(K27>CB51TDu
zjJ@Dw{a|yY)(IXd*fz(rYW|cNkvfjVzaUQbpJtHe5l?$!X(illkK!z48E3C))Eg&+
zQ%cqIXtY{LV}#7g;T?@*+>inVeJ3zneNx@c*0S^T(<iJS-9HjVLkL|Xi6@@x*VIRo
z*PTh&W1jpT{KuCMYeHkvHLT*0^4nZOTNkzbwrga$NjqMiekDuO+TTrZxSN%M{jjA9
zbd(so!9m^yertzk_46BcQG@O{cj<g|ep~b*?ySBy%X$TAsY&1s;)qD?CGELQhDcfj
z_LG5}oj>({LFO-}ToH3GQ;h5Hxitx(K`^I8)j+*Hvv*qVPH6&h!b(Z5HX7vmm@>w(
z8o@UygpVeQE!E@CA`z{2s}jcR`<6xN6IYkN=9k1CR*N|*v%o+WkJ%w5no#eQ$CH|3
z!=`3e)V3}jygSe9(_e0)?;v4fEF!WV?RL-Hz#2IHc%|qWey-~sz*q3dvVU~x+qnK#
zNBREXnDAz$$Ifq_M@zFMr@g<6Ybveas372hC6}_N`yZvg126!GFe%f}ZipJnbtIM7
z#;f|J)F+7nRDDtnu|CtD-Wye%zQbO0$Us&f9g0KTN5!9(k}V#Hwc=9UUsN5{N3R7M
z=s>xlcsD^@+3Fl~$)K|IUzx9XEY^yjBR~%O51+B$nt2Kh{tOq0xwDulEZ-$rVsW37
z+Q&`k3_!$ccCtXmBMH6)Dj!d4>3Q^&W~Y9OKfev}0qc{dq+5=s@(>$g1Z&08i4m6M
z4{EC-Vw1L~C+0lxyPRbQ6pu=pnuzC=@MT9^JKCU<o)^6Ck`#=dPE^HAcM23oDL&-X
ziRA1WO?C@QC^J$<FeOsCZ*uE?dDHX|i2tbyXdrxb0CxlR5?eV|j=eLOSM5OI-(*sE
z_NU~$<hyQouw3ySNoqtt*R+e@U|h*e73PxH2ngH)p}8xac#qvs*?yZ2N8orYJXvqh
z{JdpRVNyyD8>E|2;KgQxfZ;23{4r?y(dk%=#dHQVk5-#-9kf;IH_SHL(P;7toEV!q
zgpJVeJVPxwzb1&~uLtWc@~kU9)8nwtf{J;t!wK)KjHl@C)BU~iUH7YU&#U>?iF)UG
zkDA;uz|$}hJgfP&DNl6AZ(V2;2i(MGWw+N*Gv@xNO~7VG1~p0CvqjZ)$?yk#VA`H9
zowmjFE|H@BVaY<=3S(A(bpz78`xeF0R?3f&U~cq-2wg~D2rP`xo4sG=L+U`fe^om-
zt^D~T<$f9{zh%9}h0y88^lz&uv@yc6Zy?<vf2=4cQ*G0p;@OfJ_THdkpV4SgGs?jf
zB8{!kR6ppVDa);`v+gD?K|8j+xT}W>W%R?>lo>?KvJ^*Z;xY3rn6P*1-zxuN{FqZ2
zUsYM{mkhgY3$>49s%R(jo^&SW)}T7%YebUQh~?x9mNFf$uprDEJlBmkG@#}u3Vb+O
z(^R$npP1p@1DrPPyHi_<`$=ROV#$)(W&Tw31AUtI{^bZ#q^_D)OQlVW8CtDuDo>BY
ziWfO&1@5w{5Ei2;l6}R(O7Pn;g^6!l2j<JC4KDcu!?RH+LeSDR^QOYH!|Q%eGIN*H
zAd&o!k7gxvWh+ZsQx!9JbP9UZ%$QuJm3FoMO^4VSp2ePe774-tv<`Mu%l12w(wZIH
z1Kk{o!5rcyn}~EaXzcJPvbZ&SP4@0YVOO$Qab0x1y%5}=??F7^q5yFRK4)taAap4|
z-SOL>XLTB7ThF;$^U|oq;pd`}A5S~i<=tMAw5v^Nt^Xdz-od=?2`O{m*ZvO2tNr~?
z_>x2CxsYHW`j1YUDI#4M#!muR+U4lFX*d=}=^iKYCiZt{8aC`(-6zcUrEZNo{dc7u
z>0kWbo)s8p8teik6Ed|)Y;VGd(iL<3>wc7e=%q(K=2UxiJ=St*N;1COi}2ace1<C)
zol>)}Enh`Kw3t!dcR38eZf>-gFOxP$J)Rql_bz5t{J5jb;yO#3Fu<cepKq?bAK*;j
zJriZWM^V6Lbx6~4dgfc+ddtCjPw3IRX+M(Xms&&Aaf>vnRq3g_W!LnSY%vpagrl-1
zUEI_yE}7>cofjeart9jQSqKcIjO8mTiPK^-wcwC~HN7P`hElTfvw|ooxJ1*woaW)}
z`k0>`kz9Wyuc=Pv&c~RW&aPY0`UIdBfoR){dnOmdL@H4N%%h8%LFFU6@o53Aaudm>
zITuzZl{C$550?gwnt4nDWx+zW=_+0$ZLwDN!|xZMmd}L%gt`QWe4;Z$AH}0lA%tAV
zJL{f!#?5{Z$$z;C0Ux)#sHlW>x8g(3xYLnD-$537P#unST$Y@$S=D@Nux}ez!~`1C
zT!#jWXWNL(!nHqpWrpcqG?%Mz^Q|`_j&qA%a7~v<Bmb7oo7`LN^ho0w7aH$^oDnyy
zImT!fVN$Mf)~+*b6$+xrShsCog%7c!klnKOZekzv%5q=8G-$iypAOqO1%K;F=$NGc
zfv6`8x&@A!t9SI}r*`YXVN2-E=<|d<p2V>;TKPu~XYN?8;H~!2!=*7WU%5G)n#FkX
z!I=*!Qc*f*9O7M#uKT-b>vWC##VFprjs9cSt@spc$0g5!<^;uOrnBYF^hydy<@mzC
zmKO{k3Fx;-as?M(cX8iHx&M#6rD<~{K<a5R5P$w>e`OLK8I?J1jRR|`nPdVS-dD{&
zQG{faTCouZY$dv|QH}Whw+WkE6HBEtPD*HzY*_?}6AXw1q{EJ3xEdb!K5#Wh-he~`
zlI14A2#mpPtq+6+S0>t2y{)4YNH1NKGJ1=jfX%qQt>kwMXKaaWyChCFR72a5s>z5g
zgG%3vtB_ix#wLA_cQwW9^{Q3=y5G^wAMTnzL`ZF;chQwJee?QMVBIn~X?8Vn0t1UQ
zAvUk#VDviTyV>RNNC2%~zXb2jiu`+chwCZ*o~bnlV93an%kFsTt%dEHGYlm2&|M4Q
zGkR+4j@<JeeYydd56^Kg-(y|I_urs+wblNS3oHT-!FFs>DzoNF#LFL4;B*qUxbod@
z4~K#wJiqJGE`L6TkxO{X&hKY0;oez&1zec6+YyV>-4%He=Ss*=QS$dw*JBl*$%``{
zujRp_dXJUk{o%r*nLwUhShkenm%&MFIn?u20`K`V@{99>L--3<qRLELw?EwW66isM
zPYLHMz)|o4D9+-?xDz%<nFJ<Zir&^B!5qzq%brcm5-Yvb648Wu?*IH^2g)C|d7LS-
z-_1*4lh#+baUKQ)W6Czc_`>S|-1DGm1MOcaubb;$#4c+oOCMQN$8@mcH6I;6E+R5Y
zeAX^?7#6zMyFIRD2qDBKn|0X!gxrC+dEDsH1C!jC6fL=h^>(}QMM67C?Au2%jRhBK
zl!C_At)obB{CD^Tg}Z&5pXP6~<5gl!(0acpWgz~(*tc6Z-s-L6K1-9nS;=`!szB^l
zJoL2#Vwf)pxxbDIcsdIAl@fVQuBvi?Jr4)U?oW%UU_X#GX9)j`fGp_gG@C(9lFNgq
zP$TW>eAQOZ;Q@1|)FNROGrxsYUeEZNo|#8})PIq&nrSE0VayI<884!y90>fcVC#K-
z&#^V=f2^xR98ktzH0Cno-rgW4H;M>H{HrZ=w>3L?Kt#{r?(h6Iz>1-U84;?ljjXq4
zA%iO7HnJR3S6lplq{(rhc2Ru9Fc*sy3I;ECDRUk6!=o&WCv2(Js%Mz%wcAbx0@`;`
zLK5F$=e0GRDCD)f9L*GI71nvOxvG_044yMJlIr3^x4;7E$3GkX-_25d?1H+)z9Aty
z07}mzEn2KqTZG!y#Jd3jldLJ<dbL6}E0wEvlhMX2d>O7H-R-edr$bwF<~uT_ggcp(
z*v537g~_0DMdTsO%Km5y?KD8awyAMie<!;(y{W(XU)KVZI4`t?VBM|3N|SK{YXj$6
zn>-3hELwCXZ>8BB6SXgic|Kj&hm0`3R{xh$9mod~2JTc`5P&UjjQ{`&vogrVtUaV<
zG?`oU&Osv^lA#m>Ff*-Yv~R&FthA!>R9ah8fcvUw^0D#e+Vf3fRWo`+a%jEv>UVoA
zs@Pr7@GPmfr`bwl)O55!W&Dws`hW+ix?}=PjP+`>rJ~IF+hpZhuM>O~>ATOvgK8U>
z02#T7qXZpj?kvyrw9GYPn&j=^MP0_-*`adW&B_`bD=)zZv5E)QkSR9ben*Ah)F<sN
zanfO5k|eM@Tk&&3&8X05(8fE`1|{~%q%$yow4DDx0|Ye8f?&QN>@loT5vVnIrE$un
zvZ;j__&$|QATMK6P9nXtc;6qEQFzRfKxRWbJnTS%jw#P#egg`hcTHUXQp#*tW>O&Y
z$uO^+`}%I4TuqB`fHKMKz#lo10<QaDy4i!7nq+c)I3+)&K}DESU1z&ImfqkuY0q^a
z)f{qL5#bHrR9ym_LV>i4P@)5jCKkEf*<?;j^ayfFB$>eUtIce29`PWW#h)61|Jh0Y
z=eHqZD6_T!fxoz$ET@~}t`FPdrmJD`5vzPP=L)rf?yOVlXGg8C`v}oVpYwUVte9I)
z+mvL@7PFgL>d0B=qq=C)IK@3zr%{P4EwsE7ivZ?_v-8=^B&k|iZ%vl9^;&@oZIM<<
zd`p9^UO6XFmbn}?3|J&d0#+&ryI*GfoQ<cn;i=V?|5%1tH5~5OHqQ`~6_^^Ao9bd_
z7pwd~@0bWG3U5T0bArLmY_7D3v+q4qVc+$Vcx_+|8QBOYyC{GdD))Ev>q!U5Kax(Q
z8PZIsBXUtn=KAV@f0m1Otan|k5P+mA;F|)FR)r#!$r{5+1v@Tb*4^Q*$a(<UDUIE#
z%}puKI1En)I;p_8g}*?gXi1WbmeFL)z94by3y-&p-X9{;e~;#ZFPfMn9lf`3abZ$P
z!Up>tSCU~^j?TWS1~X8psQkg@`n7mRC+O;YQ<xn<h?#90s$=)g_ltHuVO>}V`t2Xb
zy|1W<F0`ri>T=kE+}fXA1u<PE<e~hOl0fy5Uaf4rHa92z@3r6qp--)4q4KD|G5VEk
ztpi()S|hoh_0eY6$p>%++h6)%DVTlWv{_0W4V1j3T|Z#&qX+SvY$m0&sE(TAv)MEy
zUVe$xmqjyJPLWL)pUx8G6ji1#8=dx;NdAA>O;nP8iA>imJ~U^(sU2$mUA6tYg<7Z;
zYXW+XFKvHB-a$pm2=G`V1=2mXFyXdy<i_8eVp|0i_s7x-jB84wExs{IXNb4vOXG3+
zHrY;o8N^QlY!t-%ljR2(@dh1F`)X*6n@TyKPXRWL^tZ1)kyFlA%ZA^P|Bs0Quom(K
z2xc+}jlFKhF_W90A8~bVA<S_OcNx~#xp5)C(^hD4KK*i%62hul>sK0bGE?;|f+uw;
zMH=R$Wms>`?i<-Z6o-GbcYF|0ba{TK5a&mxRvTX0Vt7jSotATXmas;RYc1b`mZa7K
z+et#PBb_+T)NxW1XjT_UCig*pccWU8-_X9cxt(!R@SX4bw!-IR0Ic(WmIk@sXoZR`
zGS1PPx;#sDlKZhrhNpVAr}~OQ9mbg1%8D1DPy+5b_Vth<lNPIn@Mf(-cQoEPl}D%8
z&EBNi1g4WWN75EOAX{$F-%<r}VegLD%1Vg$<n@x)mJ@Taq;UUPS18`s)b9=SL&GwD
z+bcoO$(QU%&%dBHR;A|rdy*Dr!pHtzDDr-NLE-*=*+pN?yMHkL+R=9PBO=?h=N{1?
zO90$<w+8d3FYxCF`6?nT82R^oKef7i+jIr5YxKaG*}m_BnGW0TFgSePV=S?bYwj&j
zwYa;8qtLjF4@a!5oz1+SoQ>9iM@m1`Ot;I0YxLN2+F%Bk6e>F~E31NTTB#x%g;rAG
zC1V=Ip()#z+KRtQS$q68qB+=(SEjoW)}*xy**-rRnlwnQ8dv4N@sjyJ#@9m7K8;7S
zCBl~6fFUAIl@Vf|Zeq^Q*Z=K%_{Tic7;2HQA8%AqmG!?(MnC&R0#=@Zp`qyi?eX~o
zCKWJ~<&fXx`Ty}Z?U(|5W%j4uN<#>Pkx73W$@3AA{!kgr0t*rQh>-!dqleV*`*tbS
z8zXabjfCuw)tLXj*1#vI1IGV;r4oa-^O;Qe7GAaa-!_PmoG=q60_!0nP*E~~$K#X+
z`OjgMEK9q=0XC9lm(z$X_5by1MyLbIta1N#%rKZQX?&Ww*pa({yDM`^T-;5Hvv@Jy
zt$ewZ?|Vvhu~FI$7{3E+r@3&LfBT-Iy%}lhtyFcTA)=Wy$&TXF{><l-gso2Tb6B-e
zE|sZJLEfFdf#fW%_8O_~T=u<6q7v~xT~>idpU?0cAkIBaeN{{9pTnPn>BVx<f>nJ&
z3RvJ!awUDBU~}cNd>8k2HSCUi-Fm}6vVVEFBE<LfINu;*cRkO1Pp#fN&FOM~j;^$B
za8a#w$c2M0$isF*jLP!ymlV|B!@&oI1@$mrszqxChob77);OwEGB4hIsmi{1#C&Y{
z94K9FViP^RnR!>H)gaGewRVK#1zz*421#abSIIf&=Y5JK7WJ@Zg<smGR7RW%Bqcpf
z0RF=f1D|WNnuJ7gscMrtvJcw<7Rk<YNXW9$-B}w^zX8nO9~>~@fS&;rA2!$daO0mj
zCnS+EY(XeRt)G4i<`)!DMywN=Eo53_$?7^^7Jtvn-WlIN`M~9|Q<;gxUgvboDsj5R
zoGZ53y(FlX<EBLO)d2P1EP&H8p=NQ&t$5N@Uao%MW%^=oKX#nC66OAYhQ;{C#HQf)
z4cj*mYcI0@jQKyG_*A0Ve)`g4F;76DN+%-`dtRm-7h5X}It&*xBaPSPbUwpyU42I>
z5Bo-m?W&FI5Er*v6f}xL9;HgfqYX4hZ4+YsB&uj}0Y@P*Yt~97O@)8(6?GV%vwFng
ziZD`Y)bHCOWhWJOhb0P>j1`Og<2*}8E@-Zvj8!~1SoB>Kfx9HH(6cZI*Z7wkDOrIS
zKl2F-4Nb|tJ!K1R1hmA-Ud+xj#dj(qZ3ptqz%J{hT49q0rMM==qUi7H&+8?seSLit
z;9VXc7e+yPcE|h5y;{jxLV>A;+f&Xwv8Zk+9|Rk#i!Fp^kKlqUg<a8SApR=~LL?>-
zqCfJXV+$0vljH4mVuXKKb-z8TMQ@MYN{Ez`f(t6ZT!!3{4WoZ3&oy^I+Ryvj{P|!0
zzvM8fTJGOqgISKxm4bT|NlkT^>%kNH6^;}c3%zp6Y+r~x=8A}laH!O?UG1oASxiu5
z`zB;vA3t0FBqM<r3X8eJ$R39$4Q7T{f$x@1sXscs*B2(WTSj@Ekj-sz*tcenj4y}L
z6bHl(LJRKtuh!FF5TR}=F+9-BenZ817GTU2QJ8(#>|3oN6W<SMM2n+?zqpr@54`;1
z8O$<MU^<ztK9NO(W``C?H`3(4L8Vw2NuZBNgy{FzdXq&%Za#q=)laj{5HKAbv<;F4
zjeK?EiNYcO2^^k!i)X*gXsJI&)<rQF+67vBrC}WHQf(zhGuo)eOr>7MW-ABA>)*@P
zKgZ)o*pCvpX)Kyy6(pSiRbMo!A|aW8OQSg3_tYFj+TT=)Gkbav-^wG%`)6ef|9T$<
z%$&yve)2`+Uyq%XaFeF~dt+KdF>3X-v9$Xd=}noJC<2D@6x*1Ms&WM)Va2z(;gE1p
zRGqof{O%7j=hJ577do~*WoH;)x^nHdpC}PefFT;`2g=wh_c!v!t`22yD-1P#*iN&0
z?Qn$1xKed%Y}UO_5B=e$Y0#o{0LF_``@<FX#FuLQfJ^mzLJmh&^ZDYeAXGtrT8;Lh
z<g#s+g)V%g@o&Avt$Aq8Xbp7dKgjT`V7ro{llKx)T%!N^rvLN!1K-CIK`J;d%-r<g
z@<U0j+&7)}R6G<7MfX{{*jXCa&*>X)sPVil7mUaS={Cj7GB4?{DgxKK1x;Hmjqgj?
z&#pr@E9inX4RE?Gu-Li_G#1g<klxHKFKr*R>qdyLA7c8VAtLhogQX>=1a;E7iFkvh
zaE?rllFQk!zNt>|vVPDjHFA(Zglv3R&1y|F8=7`kM38b#qW?IDBUDb0MDbSV|7?){
z0-Z{_0*$J|Z}-sf{4$GF8sZ5AT2w&YEab9=eveKt>B7PF#qsQc!G0<>8#r7xzdyVB
zG9xi^o2rGv{hH78m{akAkwARtk$cN{Hx(>%a`(;MoW|{Hq*c_Fo~XG*<q9M$AE;ZJ
ziP1pL$`z*jLzZ!7mXaqBr=QYCM+0Pw_|3=sX3ec{oBqF-NBH<3#~-EOvq+fzG1Iw6
zkb7774e(j5Cmmpjm`3T8T^BHBJi20M6n9A_r?W8x`G|P*INC;pZ9Z!@*sPmHBi@Iw
zr@%*mqJvwg?Si!5lj#ac{uL~b1Sh>xuLXjR7<$<;)S|`gEjjrbx=c0|sOrCf_`5O~
zWiPY)*Ds=yESx=%fU5o*l3Ik;tS-Y})q3!>y`+I1>ihG7(Y!tX_oIKK;^fz-TlwxA
z)@fIyIAahRf*ubE49Xr+G8m2KJ|e3WjT<dBQC5xAVT=^!>(&Dn497QwzMrtCKBbv^
z9NpVZ_Ot3~g#Kzb#a*Bm4?ig<rlPDpxi8*GrvYMuNudc?yLP}@SPxmZHyValj*^!D
z1zBiokg1X=>I<fWo1%&~L^YtUdB#bcD^-(5JS&sP&DQe8wMGgIDGudSO4(I0>-=Rg
zN~RNUbF2A~_V4Zb$G@IHVGhjKS*XZ1q~QKMjD6pW3lkC3^vSQ`bU>bzCvKirQBVzy
zumU^~(iAtEi+j+BVDikgCJ8^!7f93qI9$(xACSJba&x=JhHZ$pQ!1CsBcjC3BhoDP
zY_(9}^Uw?7pdwXTt=eqqWh8;wJs%xwQZoMMc1+KgY0w|u9FLP_a$1mV<&kVfiBjO1
z^|2vv*Lm&h#!(m3>?McVN-ybF^C!QTZ;`Z)b|FljWK8w;#@0rRMQ=H*yVZEx*wlLT
zf~~(#d9{79uCyz?AbL6f8O26=Ld4)%!6HwfcVj)q7x?s<#_WGy`p<JT6evaf&CH`H
z<4u8p*<;b4$4V~TzD)$()49Oe^FSr_qg=`H)^o~z#wYocWCYazUiSD{fH=??T3seP
zR{4t7!5l5d)h3S~nCKq+WdH3<pNa&vh+pEyATh)x{_oNMPhj~a88D|8Vnsz1{a?`v
zaF3v%zk&w#(&~REMZa(vq)4gNYQ$TM#h;2Fnky}@l(~i7fBIfA<3)kqptE!SCe5|I
zF*~|L%&bVWE?;v!N8!wPVCFwp`Sawn2yn~4#RT^Q%mq}#AD#S2&csnA?TDzBW0$|g
zQ7XmWmMs^##4FDtk`UDZRd~YyOl!BJNz(9s*hX@;xs05tXX>3)?EjcYpOgPM-a8!c
zWi_{5pPD}yU(Qv=HF`E7PgQyZYPwD4ea~^L+rFewD$d>Tp3aw^;gnz<l;xF;R9Rc5
zxw{72X{w(6GcfwkqwUvD_QWyG#yc1cHuSA(tuz6K^EsaItto9KcsYM+OmDG}5yI1;
z|DMjB0e`{fBjTL#(8JJCv81L;BK->3uO6@Ch6SS`&!J)2y(3to%y?mAUNC*J4MHW1
z?|4b40+fq8@YT{nokatRH%S9osyah}$!%W3`b3WpbxRYjG1PLsRf2{4U0j>ExZW5=
zb2RJx#M!EQwHE!cQ(ScPH`Wy<zPVDZgwt!BsDBU3XF8w8otg6y7sOO2Q=-eF4Ldv7
zTw9`j`L;vcpq+`V9($%*Lt;ut7DR6*xiY2Z{(!FboOYxe`kJykLWEfTa=V_?VOzX^
zM8xD<Q?YP%G;IeTAD)is%~ebhU!Pbl!jaldf>W0+-h=Ma=B1*`GbmNc<~B{n5(~#J
z1j;uB*y)}gRC4?Cu*HEqg=4MJu<}4bPDd;vEt~8pw-t`Wccc_(s0^>$xS0|SQHi*b
z2TofDxlk_bn~g5N$7CF4d;e5>iDk>UbvLaKlvuRX$*ge^JdaX=JU3$XQwAjeUbKBW
zVYV=~5aCchydnA=zz`ekA6ffL>cp?Fq?s@HcQS6)fwE1(_JYZ6ifnJcGh3VK97DHz
z(lo5+(hMzy)p1;`{B}5Y@wv7>KOqumw4BH~@8^eFY{uAGBkzY@7yck;z^o`bu~GV4
zk$tKuWNX+)slPQ%M>l=>lnOw-kx-~sZiLWCjO@XImjU7z5urB|gE6kq_deQVyC_ZP
z2dAlyb?NVA>TzP<@P-Bc_1%&e>|FeW<1_1LOO@N2P&Bk{j-7a}`MbC=v@`!*vzY4#
zaT%no%QViEVJl)Rj<0?yQMz>tsp^n9>79g-D<y_55Ix}WXgIQ<Q<nhHir3@Uy)r(t
zuGwnE>vyKy@2ad1de<0B0?Eo;$pnR(kFo%f1fX(>yx_bhkQ8Ob0Ye$X`Sw<Ya$R{6
z^^JAw1IwZN^vxxN(0iy`STa{>c+Y-TDM51jYnb7FxfX-#Ia=D<a7AlYYrM`t5Me7Z
zpgKeq1o9Mc7h{siL{;%d#5AmVsJ(knp4nr)noN$$b)U#}->GXhCq?aCtC!<(b92Yw
zv?N8LQXZ=5a1&=gmsDUvQSbo8&+W1XRiswa?D5*8dg!^TWDgpbgvzfHi+lD%<&B#y
znU^@8quv=Bc0Xv9ZeI7~6?sW#Q*%CBAx)Dx{XDO=EIAZUor6j|e#rO|qbEqn+_jtL
z!Yc>GA!B{$?g#U2dL*ePPaxpZ9>5P87tMzz)fkQ>>M+j-0=8)#UaS&+k2g=C&svo|
z=fg5-90$$uv^rznE6o?<@7K?djHU69SM!kp@GmjpH(B*jyxAL&?`BF4FXTIb+5Z|e
zOLurR#;sA_wl_qD3f=yy7fK+~Ms3#ukisRbJIXS=KNwm(eG&dOa9H3pcQIM-3h*O#
zAJ<fluV9VXax6@m>Rv|~F56y<xC$fzb6zG51zDuLa=@3>a;aABC5=N6kGYigZrvq8
zDw#{z>*)|%oLpI|TzP-ltl(42*VZxYX)J0D7KQzO9E$I`!PkvD0_@;cho{EdlSRr5
zkH;n&S%C|EhK;%Q7X#xSN>qOYN3Gd%Ey<x+in4Kv^UVL8Cw6#zHpudJz1D~d;G+)l
zXOft|-S%bgGf)P!-XVjVM#|Otc2nbkcFyy3$=07F1%XWAAK|?SaC-4i0IzR3K}Fgb
z-(mH@(Ub-N<G)hQor0E*<{!a(5y~9gdb^}?S^Vg&o*<)6#s7XKjRFtdHXXykWzDGr
z%eJlJIImBp<HhCRgYB=r{$b=K-`hX#;rNf5CDSs>-7a0-efGeK8c3k~+?J0eL2sVa
zTwuP$&sxk$c#;;&<!d~y7X_@}v{gi7^!+;C{2BB0;mCL^aolRTVVHIOCG+LA@HjLp
z*k|eoU(4B}#>~VW8)>|CROmI*A+K&EpJ^um4>!Q&uxteUP<2pmp53>2o5#$|%pTb;
z{WB}^_U5gYYWuub%A^X9NwHm2(s$pf9l;Vj+k7Gs{)MH~U6o+DR;^TiGAl|Kff4w~
z`*ff%^&P>j_EVt1c8k;dA1H|ra8SrSx$>fkW_|^MC<!cO8bh$ysU=I>m0b4|+mBhj
z0oduZ)EZ1NDgc#!4d=qz_}iCPh;6MK8=WR_ND~`R$F`VUzl-<b(CO{>dlPKuTD06?
zk>h_0x!jVhfNXDjfSy7MW_v;dIgqAWa{Fhit)E2>XO9`ws$G+~Uc0NZR@m0l@n5Px
zvPIIgm7in<Jk8Fpkz{zkglSGZYf>mSMy}8w@X%YK7nY4aTnv)w96Zr$-un<mK$dBV
zw=-AD0%Knv<_)zvzQs`<JD;oYSpiA-#AD!a;W1Su=<c*?S^ag?>OU*7CdtO}O(RLL
z4CY?f`=LB`J(@+Kn=_={knTt!M@!uxluzHa7Zjgi2q6wBv`a`JESirGf*B2DP2;bp
z^KF46+HoW@Tl6?~MS6rh8+Ydi;Ke6}Y#}>#w*fM9U-7MedfwrOAw~r$5NM3^?jtn{
z7Oh%-+pl0$BDr!+cF=R(z7wYwsLzPuJlx%AsAKwRP-pGx`(Zi9C7O9<wb78N9gtWU
z*C88`?0I5ZFyf|(x2QO#)S-;n_Pj%x+QNbm`HvmA;jz@Hrg7BGQmWxuL?}aUt*)!8
z4EiZ1owCxQAG0uio?nN$ZVP+9+_s)z;Cg@sFFPVcBmwYXb*|C#<%>IUtqK}s?bCyg
z2_DcAJ1Gxt-B#ufZCYkG89G_KVrqbfLm3EW6&>1+K%PvOgW<Xgo!(iwav#u<@Yi(R
z1aVl?d6ydb1~Ia;tG&D&tGsH?e^dHKd9ihs-@&u>>GZ)58~bg!*7!*|A;t<zW(Ad6
zv*!Ha+zaa=sO=MH{XFWsksB*oR2WTRQvK80bHO`=BSb&4n+$J2V(xe*|1fK7sfxV2
z*!ywGr{QjTUB`Rds>$U1(@PNSr<sZUh}nr{trnN#pugB$+9AfZGbJjbK{kM21dq<6
zrgxZPxu!VS?{oX%?=?5~z9A(BL6EnOhgTw2hpsK8)4jHlR;xJ{Wm*PdR+O+w$7UOy
zVCZTuIH8uc0rIe_$SL)`+acaM!N!kEM;o-uUR2JNa~?>m+9(FXc%2k%+#C#oCxA?=
zG@lLB`nXNJ$+Du7Bk0mp5ku5ICY4M`GJFvKcW4?p48sMx_jq#*``Y4oxJ4*pa@o(-
z2GC^10WfZ!<KdLC%Rq+fV>nSjUtCacSqzqaHj&6ey;X|kTC3Vp?S)gVVCWg%N2e4P
ztK^wtB}wE@r9}a|sq7B5{n>yvr*7GN;Uqod{F$3u3vmc$ckzeSLM4BL8lVhBSTJhL
z1i!qqGXoA}!p0L@mn5^+CCN4jU35EL?l6KfL^4D(zr!Fsv~&sNRiMYiv>Gm$PC3Dn
z&QgVf_>m+=S@(+#t0Zod^76^{XUH>DjpV*TkNHI<l-XGFDI&Ji4Lu>U_7ogpqXr8_
z#W$h;?FjC|92(6=p{Y30iyPLp3$i+gewowFYjCBNt}7`A_%~GEWV4dw=-!cq^?P{L
z;Cz&!7R5q2eo`Kf^ud~O4i*yZEkvh4Hh5*-rC}KP>6>dBFT0TH+qR|%8U}(Fz8&3r
zvcO=xo}Ogpp&CTLfdJdH$t(3Hv5=MUAx<*_sKoO*S=U((HYzy>35db3xlEqKieIpa
zp|}wg$}a)f=z+J)fp9QD2}tG3PdMm%>t0eoRa=oK^b4AYmvZ$DQ&RH9K*b3)scTNQ
zK<bfb5Yg>kHX9is`mHZa=b#N<+`c-2Gq>mk!hk@mgoVI5oB6r%54%Dd{HplS8YQrQ
z)|+>So1P6Cu_nnP|DFJ71W@UMhLk0Ge>yrGK77L=2#`^eL0a=~a@aSIij8r!$C!8h
z;1C_CpgH0V)gck*LEx`tzCa`dXqRc%6Qg_{X;ZaeNhzsVfZz*pA+E@mPHC=B;?;$*
zR3ukhsC$nVm#Bddg&>cBs0L`1(UcWC3k^0MJHIH8)Rm8Y*1hm6I=Yt~USPTNAOQ)8
zSZ-adjIJemu9^Be6!nG4tR{7$1fRZnLRN`}6Dkz<C5%qRYY6)5xn=a2E$Nj;cM<0{
z_<L5kN7z4B?70d2rF%wdIj+4<r<B^W=bge*2D3J&i|z9+_O1U8twG1E0|Gq>rXUSu
zWkIJwO~U>s4cy+58Z}RQ;mJCfr<P;lnOv<8(CPh|#a~%qn5lGLJYG0*&Jr{|d~j%K
zm3N-4R#5G4)wFIr296=uF=Xj(a~Xl2Ihsw?#jWklsu600`ALsBWB2}|=merO>y<yw
zoftxc#pIx7&BdUznT0<w8jnt&Ts<Ng?e6kfg?XIpd-y>0dwFeC(G-rsl>lr;J`jw9
z7SXezlXcu+m!P}_Zj<!CKPg+TH?pW-y((SKR*JA@$x95*RMs%z@f>4`!cPJXT(iRp
zw$M1vCAxvwiw2-W&;N?cB$m$-fKB@fbJXl~?3frwsZuPX^kH}MoO&b6W$%)3Cx`(r
z!VGZM`J|&$rQ%dBZ{tr}wD*qaBZmo&SF_pl*V#8~1)N@mK}9)n(t!D*<MSlHAjiE>
zYMF+o>?mMinU_8@FGrAlP318u@6S3LTLHM;cx46K=V+o8>V~uusWy~=OvOwVtBsZv
zelk_A*CrJ)-X@UC3nU<(`myOoI4@Nh1bq!-qA$G?QrpHuNG};s7v6)I$1MQqNwr{^
z64B|BQV7b2`E)Z^DtrBN>wQ{`)P$EHLhm{K2_!3__Hz-v@=c~XRg_k_^N(HA!B>Nh
zk3^TD{MXOLT<POH&;IKQL4B4dEP}|gCr(DPuht`kLQBe57=CQB<MLNR{bBgDS9`Q_
zg7WV7!+xM2F|Q9QXYI$%ukD|P{BSGwT;5W0LFM^4*g|eUxkY`Xdgt#Fc-we_$J}8X
zcx#+ZaD`2EX=AWSY}+HJgo6d~ng6iQwkPFF#ua%k*Nh9pgr(JurrJYbrr%PAF@d%i
z_f0vr%LVnCtV(_~r!<;<<h@`l{larOf5YLctL=}bs0P%=KHk<@LU}7)^o$$xDiT`1
z0yHbH(OH$k<Em_wpWsI0Dp$0<nQmJYO%HjOwy+=qRX%tJNBk1u>^1zGD6GA6s_sft
zjHyn!MNB=PB@Q`UJaTYx8%mAU+C4C}n#-kd=@M!H=hHn(-v4AS?2|gHL<dk1a<+|o
z;W#*P)(|3y_YBqBjDwqjSJp}Fep{Us>_^ulYSu*)MZ~6=$$q=#IGHW{h?8EO)AOze
zZQOS3Mkr~6BH*)NIvVmVN9IQ~JD_)OUI4`x>86`bc>!Q+U$fj2bNdT=o)g~m@xaT#
zvmz-4WSwmUH+SLta1bL2rN_~3jRCbFk`M7zYT46#F96@QI4CO;Nb=7L03%2^1;gw3
zM6@BDqK9v$|KU4)KhW28(0CBsGwo-KmVcm0jO6NyfT5MB`81uUwh@FNVi@?1Mt(%e
z{nxO!gC~;1cgtM)CiBrB;#AtLifM%SJ5!Q06WxDp{q;&@P8W(n8PPp<>RgW{VAH`U
zK|AFKYf?FsPw#5I;psckjW!%$u9vPS%WH!I;U#-wm;Xfy|GD(-PnGOc?-%x$ZJh6q
z1@E?**OAtmGQ1z8X+x*m+*!RfM_zmN*;Vh29)2nvJ*CsOmACV!d!CVcC-UL@OeHcs
z=rmndml$p3de1mjOXnAYg6V>(kfwWW6{#+v^<atE6%p7qfmV$<wID`8{|v7^VeZ!Z
zsQy56n<}1zOE|=t&YY(MgPxhXO3hA!U%REMRp0v$oOX@es<LHgo<vGjfquh|VPR|?
z3Y@@q?#HKhu_mBQ<Dc7!jPD%Wg8^_h^Qu?l-U^b#r%7emqUJ4r=%b&)?$)iKAMNQx
zxm{2W-@O%4>42JrvbOwjE=cx!n(wNBk8asE@8dn4RynXiAgXSpmgEGgXTT%Q%=rxc
z;Ls&!Kl-s*&9+}PR6Yx}i+<3|U`DN4smi@^g^a)gx&J%vaEIaIPMov+Eb`e}z234Y
z0I8KwB$GduPmUR2vAfTq*#M_6^-dgA9pT*%rZa_6w+v4JxPEC^S#ng`e;%nDj$qQ3
zMhZF$j%XoE5RQ7Fa^PL-uiJ3Kp<hjj_d;!-*)9l;xkw8UTYsUgqWwH)>yG@&<y#nY
z5q<!Dgvyn5AHjq+Z*9FRh}-$rOpqiS<=nrU_$Me$40G330#hGvt`+34YMe-kC?6UG
zlkZ^)?J5-YwwHV=;0F)7vwYA`hNF4pbZx0v=r1)rnG{6WW`gFR#bh>}iqhm&7d?Fx
z&Fo7)ltWM9&ZmBgaU708Rn>m$1oBA)qU-w?Lnp7g>yB<1ZTw(y*Iv4H`i#dQIM|*H
zPl-UTCx0w&vEyQG?i)%$+Z7z`zKbWs3iB?14?_2A{x*Sp_Lt6Qm*KWM?$Wj_#{6d^
zXOo9s3zkUlUwfn1oPwkL!7(a2xHPV#L)bcAIW7hP->GCUOJOCT52@uUe!AXA71TB$
zA9g#y%YHa%GRcVT&+t&T{$!i*vZe}pT=U#_@@6M$mFth-eb97(B6c-wD{U&CAExCl
zQ8Dc+E~Ilsp})p{EBmc{A9Yq=gtmi0pn;6-PVP!c&+qJGrqHW$)810bjf<`K-RZn0
zSnK?!z(l!Pjm*(pS=si&t1bW<1-EAj6DwB+Y<_&>%_wxgXp%nDKWeMl+JkOKck=#K
z+CzaR7+es<oey@pbx$4R<5sk1IDRnVKEu$w+`r<^L7JJHt4*!H;M*KUryY(dZman2
zV6VqEUy91~kLf0meBNE5R-^p+=<fvaCQ)Ps;A&7FF}Jzv2Z#FW2?vbT`wodRPXA5H
zSLtNah|gG0Rro^WFo(I~G1vNhI8cawkv`Z}{C;nj90I|Ac<-mVYSYrn;4Z0PTFX^Z
z8AvsZoQ|BA>!}<;i`p^58#r0gz{hS4D8_8eEkteMlb{7bPpPyGJ%2HOR`ebP;NoSR
zL37PLm?3m%O>`FR+Pi*aINI{Axb#9;fkgEK&}IU*$xK9Zlg6q$aAw&KA?!(Xma<6<
z!VR!o>E{PP9h|y*8j)dvY({xRlHSlgfi(p5n+5kaY25vNc&m9YIC9$J>tLq_#B>u&
zf<X6jZa_9yardpgXM{|a5b`&wCFm{aG>Zbgf&s!F3nKkeR2{SmrIay-tsCqc2qXQ-
zdjh(viOyv9sm>ngS{G+z#17pTLV|q5`PL~u+>0R$co8xb*8te^rmeoT;~y4KW*cu*
z$phg;8^hDKeCU-;`;OXVH0249_=T5Y7=oXG;{-?h{3HP@Gqbc#0ot?7TdX3pBHwur
zHryfqrZAC2jsS!^o{D;uuQt6Aw0ITR?)He_xS)BbfFM%LmsAqMP>y<zG7W+i**~#Q
zmBe43_BdWKWgR}o_!-8_Ljz_y<o#uOO=%4XFslYaaPy@&K&>>U$q15FfG(56t1Z<;
zi1AiNuxu<L)F=)|9*)3Il>FJxH1t8}-MHQMK)kcG9YlA2!au-&-j(wUlyfk^AIWUF
zAcx4a<HPh>XcGZ>PU>NT9f~zrPk{$sc!b_!wm1%J52yq~`lc2;>Z*hOW((K&1EpKP
z_!BhkUUk^u<>*eC9tCV`;;>8ZAX$btt!Me^A-QfUL!d5ysvoRGfUxEl*;+=SxRlM`
z*2;g~yffH4We?!TV@FRAo0cA7+ug4pr$mlL7YtcN`t@eBBV{uS<t516^{+i+(-iw8
zhw-%T4ZL@^97li;{jICEyEJXvW3hspF(DTHB?BR%iCxJ5*WOnK)wOKh27<dpaCavJ
z*8stS1^3_tmyNp<oZv1I95(Lm?!Iw%cYQ18+;eiid*A=>RlQ45)UMjA*J|eKo;`bx
zF-JRFwN20u&I@c#PPMMTly(QRSWUk?%)w!OHp~_NcPiT*#(gctSAsZ=)ma!g^ddI^
zufecu8Oj^Sj~Z2Q)wwQ`I~YUHXTTnXRcjI<Q^>XqL_jv;Ow9^|290qhd;sSYbSpGM
znu=>2VuSpOVdP|_v;vJh;ESIL#!4$x8`_S)ezyB^2x(!WNpNjG%NN2Qh@|M*%jN?N
zGtc!xZB*K$)M63S6)Fh*nVq+kY}C$r!Vm6>bLmynb+|dB?WEV5)7TBCl2H!AJE8zs
zise}AP&Pcq?e+@c#r+S0ZvD4h;-+8aG9}l<-Z=C*J1iJ^B)+_%j4Z}leGI;`a8v@f
zW&NFv^N$oF5Dktg@#|@e(KSeX{)O8MgO+9BrrZ_gga_>l8A`pso(}OQd0mnuK9Q(1
zf2j~ggJTal0UQ)?f+auyj~8UPWSBZqiL}U7ADIy(QPAJxAc(OsR03y#n6XB^yv1qL
zcV-YJST)e0;tNMWX8m%L>obrHR=kOVYXc5p5obR2TdBmjRT)N{gN4Qd(I2#X6`6ip
zAdwi6{j6CafvPlwZA(B_{_(;;bJNf`hYuqOqTnQPsw%j{?S7!iISLM8EwIv8mbK~d
z6-WfBS6UU6xG<F~dYTiYVjFn!T@H!vv08~!LET9N_b?dQC(v;=+>FY6sr(lc$7h~2
zN)9pNdzYfI4A$8tiOxRUFu~(m_gpWD+9*_22z8FlrAKQQDA{_aVmQ6&g4f?7^TrTZ
zEA?gJFoSO)dk9QiuV^fD-WJ`Zp288|J4r#-%^O;$DoVoFE_n=yA5+#`$=MiAebT;g
zC-RU(hzI&p%RW~#EJQItV#bskEurkLU!J|$i<@JnT%Dln4np6a_#^KLQ_vY`&}4)b
zI<Wm4!m|-ZutD2IY$cdX>S*l`fYsaIq*8$n7Ux_sm0Nn`uRqz}fjL`{{4u*$8~nll
zOdnf%4#*lUr2#!-7PB&XW4a|)E3L8e*wG+;pY!(+_!Ht=XN0!J2y>BiUuSG!VElI|
zuLBKg%eEL(!wY4bkAwbPjkJ`xg3=+PbK}*&Ul}EIh|$9ajSc;V{5gm^(u)x4as}tT
z@nPx<>43i`*k52^nFt9Q^7=ZgUjmkJnN8N(r-fjT9Vy-H{JM41uStV#1skpnDe{B@
zOemL0?>#Y=EcOP;eH6}zBE|oX%Kt?A^rMBegxyzibGbybW3IzO`e1YYF+;3cQzHyA
zW~~J<i6O=>)vM2IGXA{>sR-*yq-C-JTmKhcTMsF~N#nL`js6b|G@S|{dgY`<2K)!N
z=>LT_7zwG#={fWm$^nAOjdKKX&oC!f*4lKJK7Qi!7_iuT5QqMm*yy`yS9pGo5*$iT
z0*mybXCpRgzLzX2IhpF@<h@9r0KrykL|VPkR`G&DvlS!oq63tO;LswT9SVM~Cwq?s
z|C%%`s|>O@%l)Zc4V8+p6Q<2JBUE3jPy-DSAn!>C#ECVz-%LIYbWFfKww~(c%vz%v
z1X+ICBz+1BK3c-u7xT9FlL=grCGe^7zm9oY^IXlxPvicDo$>ySon3w<7oU$tf;?7>
zc`~?NQS_3Fz`1OmEW&3<F7U7e$J5pK#K<)>07{T?klTxsur&vpjr>XgdVNW&#Z?7<
zZm%<CIYWcGH+|6-x7S^G|1r?VNsx^5yh&k*>tQ1E4DWK@Tp>^2rFN|<#W!N@<?)i~
zQ|}$ICn=C^|GFY4$UqrF?F{S0U**xwgnPMf{$_PjuUM}PwA@76zFy3GD|KO}T!ziB
zn;IOG(>u0753<*W9T0`RK22dg9d2e@KbX!yrE)F~P%pI$H(MIs2rwa@Gz}bEOsS^m
z1GXI0V7{js28}W;9Z$<2d`|=93R&*6QT;)8``T9mq;mX@;8<G4-I{B3K>@AXMd$FQ
zB|@hW1M!lBYJjo#<HrxF%KSJ7vnSaU%JQv&gm=&HTz?SVX6g2?#o5&qQ-b-*K^AA5
zMYmAuJ7e9|CXQp}U8Rd#g<F9R5Ili;$jU7v0(Hv%1iRLjn`0kVux-do#vh?^f`Txc
ztPUnsC#F?J8%}et`)tkCthOwn3CL2aA<T4A&0bz*!1>RgREpH8uQ5-{01_Ee16+xw
z^$s9Q?dqd??A@L2q32Iw1;J&}s(_#QQPX^U%1hml45aRmUq=H#tK^+ykuaTja|$?)
z-c{#?0s**1F=u!RorT1`jGpQ1uVc;u`dlY{fd!!-3zze;eAOx(HZ3K)@?~xi|4P-_
zFu%xQ(^|ZZ2}tV}7aqN@c){)APQ>&`BnXX`qp8Ljhi&^Z<)?e?o5u>_g_avzi8N|i
zICP#qynI-QC&m@9@$?hCsi=6{`oLbzICJW}fe%!Z&=2nu+-;9r8_tDs&j#bML|h1%
z@N*VIps*3iglasI+j^<Km=V6F(Av7Q#kGY;?4W^)cnBE{jQnBN?kArm%uQR6ek*%#
zQU_>#8$GTb$jnhUN8Bl@6zkA4CbOToX=Cc1M88z1vM?pQdi@d(LMUhYeNFbpDf<T@
z7s<kqfOhodLBf&eweqsZ=b6k1E$7Esg+jH4hRe-6p8cQ<ciJ3>!_4;qi(hZH(yj!)
zsxYP?jhT(t)e1D}t|%n&8P)B}$$lNmsfrn|x1Y!wg}U9@60}dsSIMK;N(a>~u{|5*
zdjxwK0wl$;3P}+Fp_mFldWnSq>)dxp%-539Mc>D3cUQ!@9kJ1!Uq05P{UIrHRmbb#
zr9!=m!_C<d?HiLrVgAD7d(dz!+{NB@&9Ue`hvzo{3s7)7bG_VVGU|_=>Q(`(GdDk8
zCS-n}&Q@kqt}1qIzIlGsl9S<8z(390bQ%F*hVFTvz*VpAWH&fY3Ny}LGkx{S9vBfJ
z;;pb|)aJ?fhW=e!DvplySjlQJHFtMOh^cOwfPa4?_^Da(X?*LQm{?#YfK*n+2m&y0
z$jiQdWmO$A!6%+~Dh2SE2XhrlmTt+OzIC5Z>=rJM^_y<HXnYV#<`$eTV|-!PI*JYo
z^_s75sg<YL%h+w}ms`6nAMQRg57!7j#&7hieqfx0TQ(6580XfP8Lm5UBqABOMNKi>
zXI{G~nFol~eaZ=+7!VOV?a$fGIeV@JsgdV)_%tEv<AV0vC7G!}f^15pdu2nAQnB!R
zY7IxWYz8I#OX7(v$?vy3Jbb2Hf}6~0WE*kRXa^g{R@2d%i(H>#CM)fBI=6NxFb~74
z1)_9RwF)~2#*7Nixh;lb0Gh4o<>xD=v^nJT^mgNWPSsCAH1vIsS3mOXuF?c-L+-kj
zHayy&hZN_e(E*^|sIBkK+Qj0eGuu;24u28Nhx#hU`yf&@>fGB(mFX7)2RzTqGrAs`
z%%?4Xky3k%66_O_&DH~0Q-=XkL=n;M(nUqY<N1KCyS`y??IjSn(!6#Qgm@hUjd@ZH
z&VpX`bo7oZLnx4oZ+PS6IRftx4p~w0C*rJ3(x@n<UQ7oy9I{d|2|SO3ztWo@pR%4|
z2*;m{`M|`dTD{=={)5rQ?uA0G&AYfKvOMMTWG$OT_1n3s>`2Sl&Yk4@;~4M{<<M-;
zKzvTiuhva>>KXKFo==(~1)~W4s)Z`(IG5Yzv`uOsFYZngh>E=LqE9yKCSP7z%qNV<
z=05kVwh>S%R(j`2aF$ZIR3qs~`?bmP4y`tCSh={Ff9lr#_$0yC&6w>15|y3~>~NoM
zj#{rD@PCROpx`-|s~v%+uB{+G#k4tYRS;_*cJ%7A*}W^2D`sB&J?`+m)v)3mID&U;
zj^XO#ZBcG}*59WaCpEyjp2USKzw*w@GTube)&hL4&<OWMf$I*+!l;6CPc*DS#K<`Q
zZ6M~RJie5@xI}ZHMrQCOiZvVDs4p7KKwbFJ9;ekSdlkLm=JWX7)T3C?&@Nejti<*5
zcjj}?bws~z6yH88&=zxaq_`Ko!`{?b`^2%|ecdj>bm2#t-8>JjKw08_dXy_g=)FQV
zo}DA$K2>y2db5@0k$e~X1seOwXpei=o|)|c8e0srg47r;n0pm%M8Dy^Yq*SaivL}-
zJCATqYG;h{BB~ACj8A`|q~1?+g1~`Mv^~QAbPCbX`lg6xd%j+x3PXmvxR?{vTpYCg
z4kPlx$hUcbY7K_63Q~3HkUG3SC)r0E8Kzn*!*|B*v`Dr<o3Bv&ZI@*#K`vJ9K?ac!
zXAYK(2VnZ{*hrZt70z8Pe9YK2kiBnp&%mO{e%wwJq9BcIy%Hpq%x{M+ed?SR%-2gT
zxOYiBd)7VhM9Ax8DgcYLKFzQ&Crcd-VB8I{yZgF_MkbwD=xBijXC|YoMz?ny%8^;a
z4ZBy;bHkYC8dH|c>A?AiF<E9bmI9QzBuJ_SXRkzB^_DLO&a6M5jB4=jUw=NIE*6<-
zzTc$Ay}J8US3Z~fj^8KOpA$OJE<&MQ*gO~X`T+GKR96YQ;DZM8jN)ANm_7!fEhnfx
ze*To{Q*og{P<Xe)qi<)KTG|<qw2nG}IQ5MgHJvS`QYXqSNFb&FFYbF#<~M?I<AEu@
z-Dxmm+$8tqR8OS6+bRxE09P}~sf~8S$>r0z9|Zw&*ZzTJB%8q|0q#xf=QNHE&IXpQ
z!fEid0~|B|u9NZfkOGMp)MPpsI21+#`CD#?D@{PwEsbEYTfH6^E0Iuz<+!jQL3^e3
z97pUW+m61o1K~vp40@K$V2Vo@h6nOXMJ*7H@bzT1fx`>1TbSgyyK9_D^(lei#Q^<o
z9nZI6MLry7lL=}bx}hM}=6g~vxzv23z>2AZFK(|z6DAvqCHZRb3y6eq@61=5ql*vy
zoVmO?T3iC|lmbpS-tK)Qb}uHreMgXFK2%V=*>pa5CAb$9xy&wHiCd^XF81gq9y9=-
zW;Gp3ODuq1uoNu0+e(K*hSdz^{298voos%MHQ@OM=M}dHOA9gByze0W;9OI8xa+{Z
zN(cRm!2>@?|EyD0Fdv@TizZ@lLL>}^-?!`D7;@I(1U5&R4jTVsQYHsYTi3G_R0F#M
zUuq$A<Kse{Hz+X2Z}AC3Wu@+4ghnP@Jzmuws&S|aPC07?uY49Et{*;1+ZH_biTdD^
z({kM!)M7rKU)u~a;X@xxeyfmHZ$JcV4yUT+=!2D=v!mUf?xyWZJ3QvH;7-)yYj*OQ
zuQQOQW#o;$)?6MMNrz2zND%B778Fz>88*2WH|U<rk_UNZn`>1b3CZXI_M<yD25l6$
z&U>hRp<@d0-C9atY;4+Mwt!N04dk!mu~)}ayn&ygRB9F86IL~bJhwJ6zc8@+BnrQj
z^45w(kd8T25?^j@p-;QmVv5v9ke+!XU@L=?-|ae(K0*1cS3rdbm+yUG7B=sClaDL@
z4c}?UnZ`U!yVtTM=dLV%gOy;!jW&xk3WzVB);3r+`lM~AW0IKfhxZCj;5#&rdx3>o
zg0nN{>8f$eH%~3*ie;jR4;WeR%e(3ngU}W=!TjGFyCj};*x~IiQv)6w3b($nGwxMQ
zyF5!D=DYD$QqaM$?;9H%Q`WrC{{H>DdX5Te?MAlw`4vz8ASqhfeqZ|QzV{2mqeK$+
zecey&>Vk%>I!T9p`1YlzV=chiNOu-H;}_)}^f*fDv=c;gd@U$&PMNcQAy94J%V>S<
z+l_hv1$8$(k+kb0iSJ=K!`6B;dzYN<%1IPGprrGmqV=fn`;>@_5OY&WiU6D(C0{X%
zmJ>ib&V#2IoD@k^xSgX<+OZvng1#@UuIA@za7oZ{FTN~WQ3JrgxG#b_-qiGwcmq2;
z?X|?<%`;(&i6U*0%>hjsHKp8CJX@&x^KWMr;#iO!T=<R1lb9byX9hZIpaXQbRnG&x
z_6OsXIuqeot?|9Fs5x|(mSvardk6cuLH5W841GWNi|GSDmn>%-fAgU|A_L}v=^Ou1
zVjK<kGgMVl-n{S5LM6+3iL3*Xoe0<nr|+%=Hu2-Kict-Yx9hcI66?d=P6amD1FNyx
zm96U@rLdJlHJrU{(_}KZHS02*vYWiBKTk;zDS&ir7c6FKIxjqt^pN(Ju6Q=|ii*AD
z4&6Q{UMj6-eXn&&yo-ihtTIQ<wQ#e&IrG6Hf@uYXj>52LaZX*1@`_xqIm);!lpi$1
zUhdBinYeN7Iq~ABj^hxrtv~vBf{mo4lAEUXCL2HUJ8R?>9~iWsj|!M3p?B<0C)Nt2
zdg8O<wI3{+F@7p(MxWxEr-H&d2KLE2gF)i$($A2EG8O%D&RFb;sAZF>FcO`4P~_k*
zTUiD$zfmjgffin#3bLc@B-mnCMAaScroRJgsk4KiP}Xa0$Bos?9UDCrLXNKWM3QuI
zpb>`Gy?Kch(z287YVp2>L>$gJ;Kw1DWwtM)dmgiZ(eGrsm%K*wp08rT$-OG9z0G2W
z8>T|h5wH)tMRy+|7&H1HJ)gy!k5h~s3D2{7Os12!NspKPh){(Jwu8V@)d(5ln$MJ^
zV@bkKM9sf#EU-0Ll9zUW=P|PAjCHim&S9D;=ZvkloteO(KHr=x6ujE0oEGUICFATa
zY|(skC|-wcKJ*$~Ro+!;4c2V_m;o~WsrW_HdB%naht-^4eE($7i8&o(SurjFZIJ+F
z;W@Fhw%*>f^(lHl#}yn}y5oM39@B|xUubacK$k&c7EEZVt5hnUA17Lx4FWd~&uv$7
z?xisGV()x0WpzU<vW^3S1&a3j?paGxu&aaFO?hE%ATM~3;dw{LOa90>bR9c#7Fz*1
zRh=$3B7*$)<J?m2?<lt7SD<i+=x{~0Ad<Q&@_Y-$<#F`Wbl^E>#t>;9ct#LCK1ZRs
z<KAVbi!M9pK#+Cc=*eo{%|^z5&hrsQi^HaIxfKWHN(V%=!ei6iP4Iz1-o=Jc*Y-!#
z^F!WOl9nsU0V||X9h4{=Eq;P1wjJ%S4Sb3vznCC=K=^DEI6aLc=G$VU(`K$o1!7J2
zU+p~i<m)-jFThO1NMMof_oggZIhRG}m@j@!NPC8UxFeufuP%7qYNHmG+#-Tdf){-w
z--AazV_9uZH^L90_ozO@y(-cjOdY@2Bf{tQ-@6UU=-QzmDP*<l#S!HklF2bgxspyC
zw<af|s9DgInY$i1Jn~~lwTgL$Xh;d<a^m2d(}bXFZVV(W6emsD1Z6SefnaT1u*OKq
z<e3OrI4NIX9S<%Kxig+Mid@85fh-J4`iFs8%|^#G_op>3#;Ot|!(6zJpj%VIbs!eI
z`>u~@jT4yKcs>8!RqZt&GIP@{ApbI^==zy1@Dr{G;qrlZE=&6ui)zK^uVnefcQPew
z0^uWb4Mr)SP%;523on5MN#79<m+LjN#Gl|HiA;&jzMARt-K<6mXF9&y(4&i}(k4zV
z4=ti8!egD9I0zmZgeK*}lx+^>AQ+9sB9yM2)&9o_x1$F8xPNaiH6JN+?n>KdkaKjn
zOJy43+IcWq?gfD;j`N_+P@ujJ-%4J)JFRm8tGIIj9+FAm!t7crXu}vBjzfv61e`1E
zE<cRvv7h(49B#0IsF3$V+^Iv2?O{@F!qjD1U}qJiay^c-L5imT#YSUC`%Cj2&JVZU
zvI`9Lq&xuYgn%7Kw3=0m^Eg*`o3cSLqd?Cok?jk&Wvw2Ubs_SSbE$3JP;tY)Fh9(T
zZkByEo$#@3r@IIBkoX#h`V7QkZgO8~641f-^f-EAYz>@uwi2l90l-$CN~T(FUM>}7
z+Gm=-s{_$qV8uOf`Z@w)_4ebriTmg`S0(6Tb`&3Wa5^-)G<=Chra6{~=ZVnw%`UTB
zK9(zWb37f0jzhpUawjm#%<*W5`O=&3FLUZRbM|&#JAb(-!V3(AmpOhkf~2OfE%QSd
z1^HszqKtFSW^K?mt}HUJHaG7VQ~ax#ewH^)m7)1nGRRKPu&e6i`*A&R=${BxyA`l6
z>^};qjW2T`MhU4vw+h<z;YBV<kaWYayhiqY9y+J1M|@8_a^$-3we=g(#bgxs{j&$w
z)~&2|chNJ^g>)BW_8%y6MWdM7xtslZUy*P`x0dfRltf^s^IaS&Cv8%B)ct9AWSsRa
zKjCzx3p5!ty?WKv;$A0gHnX%n+^p(*p-T7yOfPhE9Q(5ULie~t4~J$mit$seM#6)6
z<uVKps`H!30R3g<4rtNsnoQ8XF-})V00ql1jGV=aI<CzY2ERCbDoYOHS@_7#d^VqY
zB>&B~LwkcEx&~gLr3;I5T}FN&*(@q}-TSyb#s~-~Aw<=cc$c9=7(6#>Bd}U{w9Hfu
zFf7h@ED5NV4?9~+J3Jz~?ZPWya+qs2Db&)y=~X0ltrteEUK8B~2My637;JnPE%K5B
zd7Z~%3|M2|TnxaUy9v)$gF$%h5*~!T?sX!DSV96^UMv|%q2{`+Sezm^`N~D#hUxjF
zTez`$Lh{a#wmP<%o5m0PIzQQH^CV6G8k>9<T+_Cm8Mdy}IY0QX92C-B!vtC|YzJJ#
z#OHB%B5N*L2(Z%WY(%+7Hxpl*@ljH_=Vlz&dFBvFz1y~sKdZAP;!1)zz?H&?K=enc
zCfcoX(Hkn%P+x$?_P63|FH2iq4#ko0Y7jrJYD&p7tBfN-R}&ZW=S+r?4z1YQ^B6OK
z2gWEKhrCP4XT0JJ<$|f5AhYpFe*Ck^@T`TVhsW3a{+?yY;}JX9H3@2rBkZ&4j2HnR
z;&jU|gYt|sbv!>qt9-j(UU`D|I*DHIJ=0~)+XTqscT}5S*WS@*zfHW$<(HLha{FWH
zo^o#(+7fCuQQ?^BvIcZOJ#wb0C6u;Er{IMe7X}XS*!~WdBd3p^AO19!fifrf<p`az
zGvUWm&?5uCwotJ1y(0R+c8{UH()}Wx8iBiTB*f+Ak=OZ8oe5Yuo{Qa3v~(PMCe}}=
zp~5X|KK6RoCp+1JUz$$4-0DDZ??C2Di?=wBT7@bs_Y#bDzVEf*&}v4TiSchHeF>)E
zTT}~fWD}>a*2lryHrt6yPFK$w!(I5T8}CYWBr$kds^Qx=kDu(=IWfV{-fbZ|IIp+9
zd(`{fSX?6^aCa)+NSB?o7koYCenJ>HY5Yn4?C6+C6o+dC9(Uldv7lLFVndNX63c-P
z48hgl$_xW0N^$*>n{YQ&3O@MgXLXSmi~UQvQDWD9ZV!TX!NxwzaJM)OO#7fyh9zFZ
z+(m88L0r*#T#+CQddk&qBBDb0qxuB`o`_~WPxpt&r8PNGLg?pwVTnIsu=dm8(@%-J
z_@u#AXYlqj-ZrfTieeFXMX&9kHILhc<yaJTf^ya8FhGb4&Q;3vF^^e++-roHECoU6
zAV=1Omm-@sk1e4=Z@6%Gzg+RIX2BC_c~{md?QfhrXn05;HSVdT2?RQ|pUo=-Yp^xU
zCP)4;n8&=QiIKY!^*U#^=$}``8dYT-AvY1Zb>jtQd^Z4G7VLxsY{TL@kTnD;eI?@i
zZfwOdv?5Cg*;P*WQU&1_DCZTbDs$`v-Uf3=V9;;6x9wJ8FtHHxe+uF1PAg{9vd(eZ
zA#n~OMQPvCo2oM)-lS}|Nk4C>s9&tAntiyM&)!;0wp-?F$wf(2y<5rZs6UE~Pe^x@
z&cNmo-pshP)ZEGHazgWzj%+H4?*-N+u7vqR*2@x|jaGLVEqLqZtI{UoWxmFpK%LtL
z5gzB`w{C~fjO(`P28XPNSr{Z;(AqIvXv{k*lyl`M?bT@pYBs&S4ZCKH;_)>}uknTk
z6DFRc^R0K7M>Kf46`q{PV<BTynl4Bm1(tMt#vb52Px*O21!}K5Ko3Rz;Jh?#FdaM1
z580`1`GLRf?s&&9;?A1_@GF!v#;eg?)v5(Ohlf;~HLW%s!$}(;*5t1SK=suc!%pBO
zc!US_^4HhXFKp<Eul@)W{@_Q5mmjWHH;6rJXKEbS;krzE*u*x~#V1ssIo}z-Frm0_
z9U&;!YN8!xYhQ|b)z_~3imBCe3A=xdcRBYpkEStKgk*ciyQdbElTsqkTwpbv2<zxd
z6@DM0TmbH=c?7u^gNL;azux5<c7yFa3UWDcQCEyD1IMb8j`KcCQ19(p@Z06A&PYZc
zdmT7t@&=*od!3D{obQeF;%IxF(gUT_Vhzm71uA1{QhPQuAcBl9uaUgW=jE##=a}R*
zoGlS~z}&3N=@P{IMd(`6Iy9VfT?2A#cLy05`dsj=N~4}=GwpNDYxAw#mUD}q^X}q=
zsl`82*Dc2!Z#!K4vSM%o(fmd_QV%e}w?q}X#4xt08a`7z)i~42nRO&qEP*fIu)4n}
zh;;@0W#F}Q4rNDZ%uQd0V1MP>3$&eOftB;9*1U&hMt5axzoQ7}AzJQ6^egNfw<JxN
zYPT@{21OTXoKz7mtH$&aE_9+-d^g*Ta|U5<4w;~tgX9ij<>Pp~Qma$c3$~fMN)-n6
zXV1t_B}LyUX^g*LSZ;kgT?)C-4+oW1r5(;@Gh8k~mlJ2`Y?V(RC8eQ`wcP~mam*|U
zR+|s}*f@u5J6ONxGq7q#uXSN&b$Nel(j7tQfN6_fsT1@Ee?sH+@2X52cR4=m>7bvH
z3eb+W!PCCFtTFg^h8JDj4h`Ngocz?b!3I{gHUqsZOKePs&bGlNmv+h5Jxh{WtIoM;
zy5ONqO(Kl%h`U}S?NSLUXLQx<1XUgcX=g%&pwMRiCHW?|O6^Sh!n743y@d$3lXIyh
zRjZ{{YAaI%K_1iNH-_EwLxgS`R?`Zyr&U{e4-wq?&YvjxeAFrsh(yby%MpXJk{ppQ
z>dLySJvsGT>xNAs2RtI~;J2xLN3l(+c(Mo8vV*Ff$AZjfbs1cppg!se8?K;+p{ubZ
zkYCzZ<zab_k6x^^5$r3}Hk#;EjS`2-O}{#INbiAbX&1LTMCZ!l&f2Fk?3X1SOicXV
zits<p7qfyZ_P2s7nWikOm*BxX>T94B7>Q*ZsIA52(zSUqpDz^NPOE7@Rl?pfH$?Ub
zm&v<I7&`Kpii$%AJ`0Qh$OG@2%b<ulcqil?&qDCA12cwTN@&ytPF%=RMMQ^mP|O*N
z#k|%#mJ!>{lWjb8BWVAmXDqK{EMv^ZwDYTiI!mybK|B#2#?3dPcr#r#7-|%#UngN#
ziT|n(`6aVaf(2Y&vKk)TQ5#|*{g#dXZ*Cn>4Ex9L$}i~=&#nQVf5(RPeSr$q|2Fxr
z#eNeo{(l-ULR!)j(Q#oC&}Z=$VLf)G^1sc1kq4u}o=24T-k?@O2s^W7J=kUp;TX<k
zY=q6=ACnIX3jrAl?JFv>3YFAf^!&eOU8y)A&T~1J$_}K7{w{+4^#PD)$!tOSZ<GI8
z?60Bn=eYatfcbxQzK}3o<`3Wj5%S+WDesnTLC)T5ZeF9w_!odi5lxmThEEOj%@3V8
z{jnk$?i~-uN(4IGP+{mr7#fzwn`4^aZAA%v26@qY+KHKBjPPT=7h8=gkRo_ipg*pW
zwv=<SN9S*Zh<><Gi=?4yAuanjUsL!Gofj_5%o_V|ShaT~Z$^#ZV)Y;ccRv%RFAe+c
zx}rV+LJ9Hd{aU^*&@t~)vJO5|VC(>wkv9gY{euJ6^Bx+<O0qu%X2#K_d_lix;zeTV
z)M_E~U;Px4#%}QysA!P%woKrJFiuIy_~4o|4&DAOmeTPSs`yokQ%Z(qf)j)(UP{&n
z_e0TO-QVB;%T_`-zXaDXnPUFK4XTj=Vafk$la2sZODY*?C`9S>3>wPNm$C5|eQ+rK
z&;Di1C~fFFg)yVcGSS(3L2bNAI(pRngm`FJvMuhrzHm1~5&h`Y(QGl%IQseq5V=lw
z*Uz7z@iB!8d8njXcd&F~y&|A$mW(gSh6*@+7A(N&lWS2CCbnpU<>JOQl%hEg4_|Px
zi1>@@PL&Zp9v|SvU0H+teg$dz!7U$uiU)e}DGS_L+QyqzTPjKF;JQb|GOEAre~h`<
z$dRO9|DN%>=lp4b1o0$5R)v*}Y)RPm#%O(k@rPtMjF$`3q1wm8royUo7rpfT`GhPY
z+3sF9x`F1<s~&v8;)|gEAu+4$xU_rRAee9AKp0q2NwbFyf+3+WVO-h)UwMDP51r+7
zo0WC22nusuew)>EPQD5Ur1?-Voww>|HMG(vi?h!U7o`DK70sGSVa%0RGPR%1t%v}=
zqbbU055DrjWM#;!1m2i8>o;K4{-y~(yQP;7KRqY)?T%hLtY<wj#<amAOB98iL??Sv
zbjS4#R@iCbPzgbQ6av;wBs|vS0&nW!B^eF+m@Uh=G)ed$$`55@PtjcpbgftRPR+NY
zH|JMzrx{r4ZWARrXfkv?mi)=vYk50ChRgI%eK&;p{UPCQ8%wZ|^#{_N{)qs{`j#pF
z`eH$HzxbL;Pedd9`sxt-<P#p&r|atvt9#8;+s97l7Iegqby`td>{}CYig>C~oXwgH
zA2|2p*ff;m3x<m6jXu1X8q{NST1|huliN4BSIDDK<j(4GpNGCBfyus}V>j!3<9Vu~
zy?G{^N3~aVm|#s^9G8?Ypgd6;$}<@Snp1YLTy^r&$d9ZvS7Z8sLrx{0zjlB(=9UDD
z)(p&;4iumD2^5z*2=`bWAosmfT>di(i?Sr!p|8K|;BNiO)ZB2VPJ2uZHEAIunw0pV
zOvspNk<9IqFe+lOps$4<BDJ&np5mhAUF}7!t=%jxsu2LqzA+twGgYZ>&~kA!Fx5Gl
z;dk8im3ecSp<HG&>1e7}UZ~~uj?Hv3^IJwWca!0Ptf@kM>UnKR?GS|68$VenhwA6W
zR?nyD{`G+R^Gk2eSA8+loKX&tu^-9{Sf_KAG8d0slh}r<p*d1d)316$25HBx?(ytw
zyO9)%4fVLAY}J}FqG_xnwR749JPgQpz6H*EUQQd=9v0@8nQlS3>_8@qO?M>jMF8!d
zH$FW{?ihBXxTLQTgoOCddbQYzDKjSWG=!THOEhlLG8lSI;VIWfFl)C4cix!ozsXb1
ztEnjdBufHEOn8wY29hVtsJ2T$A?BC!%qW$dV`l=tU&tn;Ewj0>ViY*r>Zg|M#pM}D
zq}?O@ntY5No=QtOQS!5>UO}0te0_6^N^<p3iMFVNsqc6)qu}9W>Bcsg$tn5P)?6p2
z!W}GnYRN3~Ra=Up{S_EO{we<$?Aa(k_?(W`D3(P=Y3I_zvn(v7^kZ<FT^W;Y^Ze5#
z3ho&gR4Z+afc(z%(Lj(d@j>TW+>Hss(MKmB!{_<SR|-{Nt7UhP7j7km->;5M+$2G1
z8?|XpOC9u4Qm!rQ;hvuozd60DRj8MuHT@nq?0M6T<#JsV%){j=x5>0b6&a~?wVhY+
z)3p=p^kJ!gBi_XGt4NRsQ3dT8r__^4ckie2DbtTPOVLGRcPFf6hT+!!h#iy>;Kd4Y
zp0`Ec+83@T)YO|6RmBvUJVWle*gqQJ%$G-;Z!1ZzDiM=#A4JVxEn?x`^t(z*aZ}*O
zj`$PGTUOhB2-#P@v*M|)J8#YmeALz{rB!#~KI=za$nT}hxfS)iIDP-)^9r=xhbz4N
zg$O`=Ef09Z9XCOMRFNi3gwwPkJ+2s;&pdUBDq{0h7eib@T=j9Qx#lj)kd@;5(7stS
z37UpD)zpugR<*fVi(=0ZYP?=iAn`o!qKjvuGsqB>L(1nSRB>Cbi?<6t)gs$Z^7D*e
zGO>y-h<iOYNVK~k<%Vs-Fw68jR;^IGwkDky$$EoQ#52KmoU`vBfqgVx31Ns$Lp5IY
zh)<asp4rTc^13ObLH1UAz9*yeYS?~!N4LBG6I0Z}-I&pQ=++0yjYiQrRsNftut$S9
zrW`>9Z`6XCx9QSnxFmGwzrq3`B}}zKzJQWjvv<VxFF%s2&>2N&aA%4&XiXLcpEnm)
zrsne{l%%<h8=lUis(Z@y#nR9<H#x?w4`;}UMQZ-obd<$LknC_d8OP;WyygUd#p*=r
z-#oZ8IxfIA+>B-%`ytEi@))x-XU^@qO{uYe*<2Wx%DT*UaN=Jb;~3ULT<+B4V%c8+
z>)F>MqB<QzsyZ1n*ax}JJ{5iBx6Ub)k&e3)V~i^mffO9P-MQL@CUj|!czg4K(Ce{a
zQp$!X_V9rIGkA-;sJ}na6mpMy{{~c#LfF{$v_Q-F^v()!O#ae<S2i>Zv-yU=pxrEP
zQfW3W&>E$nR_CgKsM+8cX|LR_AZMxZk!lF?t#kIa!;q25mY`if%)$KA1s9M#N=v>;
z|19h*T=>Xt5*oaEJ*(p)q{kjt{lZwKNm*9jRH2~a{$z=&(JKk>4RZUjioro#P^sEi
zuB7JOnV4N3WvS~<%C<pO9c6!9ZsVk2DnRTLkuR5a7rPr8x6_D3m`@h)9v@&x&UJfD
zanmOQXZ9z)<)6|OC%lfu(5hya4E-SLd;Qk&G0&l*rsJwAoPKWsAylvISUNo?B89I&
zg@);qpi>fB%r~)xWoO0>Mvq&sr(KWD;rR^#>$IC8Ibt8Txau<DytXkp7koKo5#s{d
zqqXA{{GF9z;>fFq#X0qdJe4Fyowu@q3p4Wh@$}!;$fbopE{%M~e*7-Tku39Tviti8
zEfbmJyfKQVz(DbN!A{VH6WHCdiqv5WHX4*doYS1$3l^-LM%OR3EmE+PM^#R4a;^`~
z7CcD*0Kx!~gm=Q0DtSO%bnxJ?Z7?a&>}|wlzKZvje;b&^s{+Z{d|Ix_qmOW~(0B)5
z>S|bIn_5+bd@E$rw%uw^*69HRHKVebd8Y(z@x=?2gegX8!%cSWR{Z-fL5ypY&l}Rw
z-NBx6V?RLsW^?1t$0K@4RW_hkqVE~c&WoDyO_Rtf41MKBjDvqP{m3SssAoy#7<!YO
zty<mX8j-`#*L;Uyv6R9z>|Uzi#r;!{7MHaoDVbMYW*&wrf_z@@f)0BCoLWyO=~aEk
z7iDV^8}oqE49uMMRG}#%o1t+7B#J0=7`#cEpW|jq-+#=zIUqNp4zHJ8xc$_r#Bv=a
zLZKye{J`4`xCtlYO^wG{ncR;upK3`Dc{s0JBsq_yzgn7=5~IX9l*c>tBQ=z$!B$!H
zlJU&+g-a^PBaTLQ`f8RfN=oPlsOLObZ7h5K+L3>G1?^A}KUUDyboW5OUEm?Z4*l~-
zF`ihv?loOQWQ^G56!M*rmVnzGu3hiWg+rp-rzf`0C>tL-3u@2b8~JLaU6!WQBZyVC
znDgugxr}}DgypXHVDxAxaXr3vUSfK*X9Z)Y%W9}VV+T6LuAbPsuB6|XU+(zTPZUV%
zP6d@*>(Cs|v@ca%d@!?h>3SHSXqY3nb3H4IfW|&a9Y~HJw><=Ecn~6h;LIi!hb|0;
z!$NL~zEz`~tH_8lPrbB8Bd@X{-A|3g%d^p!;Z=VtLH`vopEh0WNn`zf#!aw-q$N<%
zE<MPiY_37=T5ndXKX>-!rP+y6gH)~ajr8Dm|Fr3>WOG}lT`dvjZ8kMVE&DhPTMovD
z5*-=@oq_7Z9fq3;!9O=2LdP&jxudj4Xgf^Pf~;b${O5jJ6F-~qnAGJxGn&LW^0=ct
zm`(GR{M1XOS<|*D*tg2;YGIP9^;2~8IIdGg!8muf2~UWx4r^EU#mrbU*<BM4_O2Yq
zdw~#AY4O@Do{wwQn!DP-GOatv;C0Wkx!mXm0wwq}I<G@8p4EJDLxYPrTT&&NB=Tj|
zg*!UDF~gZm#%!aY2v}J@$Bq?uE+>ocYLc#U?nsYO8eU6Hkr~Neg!3w>P9;#*bcHh!
zkhzHBhklVbyVj*vEAdRHUbRUhk+O65CDwUg33z_=W1#0-Re?v#fLmr?m(a+i8L_$i
zI7L<%xX@Ewb`^yFTQl~lL&Ac+PDhqxWx}a=^vGz!fhNHb0&zEI^Juk=DV#F5olCUS
z331T+0df5H0u3a^kN3lx^bwV#{#S$oz4T4wY6pzM)KFzJJ(Ilm;Y3YMak`b7)a^)R
zQ%<^b?aqYdK^sng3`;R+=l8w2;!N^KZCVtc!mk!$SYMxxrttPpZgGq%Tje!`O{Y&Q
z2|Rg}zKTaRi~WksJ00`sQJeCHC52SceC7!9#lmW%T5C?Kd1$@9Ilf=S>_WZ4`izK@
zXr<BrVsBK#ZL%r7L@5V3l3K^WZv1@{CJ#^4?b?N=L7V8^R!j!d^cfW|J1zNU9Obpl
zlgX%+%eu%C53Cxj<G{IS?fHb1M%ZKd=Y)piL9@nObe(t5nl+BiZ^7DH$RFv;$+P{U
zYR74dPCyZjpXTrqAb{KPL0sRPt28&}<g?KY{2~o4UOJx}#7)>OP`>0&hG;>|y_%B3
z0_aFaD#`7We(5#5c<hZ%I%=-ijL<7jd{U^ZIh%iNK7-vDe#4Qdio1`Ys@O!qPK@n4
z&C}gYQ0|rNq&Ii9R__uJJkI)S#*p}^k$X>n{ZX(Kw#qb)R+1U{$WObHk{v}Ta4oqQ
z7fWX(^~%A6tA?Id20h8a#W69>LL_8dZzMLVEu*jn<J^I+b8kA@a^}#DF}RTd=Z&UL
z<C)!_)slIOO~boMeczWxB5CG3vGM%P64wH2;Rfx6?H|)42zhr6&k9Bjrr?SGKv$BR
zf2*Kp!n0Xde;SA-Lf+>J{_w0tpMMJ0jSLY9Mn88IYXhAUZPQ}x&oVD)Y#ODK7Tcrt
zimo=Zvg-_Nzt0Rgn})9H>XZVeK8P*jHByahWBtn>6(Jw|wM%DtBQ#xQO=|--et9$}
zX!V3rd-QJdf9+rYnO<6!^oAbCMo0hm?fElAm=D0puBL$UA1aj;|9fqO>@mO4XX;Ez
z_Q(9nPn6B-7mvJ1WSw3wWcfQk!GN8s`csWTRZwm^%`zb;h1nQM&IgxGW-#RU|L84I
z&9U)?G!jb}vN0MMTB@(8+ok4LUiY_wN@X&yW+#6D#g{XIQSj!G93qfW36im2;+L+-
zrS5-UM*-*~r8nkbzBguKLp8C(NchZ#hqw+4ZR?U1#X$a7pMFJxhnTq<#b4arz36ct
zwy1CgO>zq7qRE2LhzA={5)~p4b>9*Hp{ef51PY5mOces05;X=#IoaPyVjl^zXDh-d
zSdxrwWl=(l)Ibu*r5RQI`TtBJ1Irfz&=hkNg}NH!=Jdovjk+2(k@t5`M46y2Uy@f_
zl9d|1Nc=iO$AJC&yFXJi>+gUYxRi{l8UDwuMS*XCHq^P&1cx@hLTvcIulNEDz0WK5
zjy&C)PF^?tS<is)f1pY}z>39CVfV#heLN4thn4I1gGDi+E@8jtvdRf7)3C@v+Z+D}
zRy8UCj7B`oNfl~=fHJ5R)!z>$@zEj&j@vEH3JdY4Vfmhk>+(_j9eEkj?{|dp6J|y=
zd@Rr*@6{iYq5qz70v6L>nU2p5->BYZP`sB>EF=5f*rpAk1l^@gerq-iNU2$jmoESQ
z?$3#Yq5dn2R=}@@_m58f@!ERIf2H9p#y4bJxd%v{xZb?{eFbnB!9LqBB~_~h8WvE3
zNxDopXE-0kp5Y&6@ehA?_eD<xzY<YVSIlzUxLLp9t6>JA!S8`Aq(j}(q^o1y8l&ln
zG!WDO<iC-ghTqxh2QQ~~lZg4-!ci@~lu>C-pB!%!sjK4$_~(J1waaitW=E!U0wq3Y
zpLf(Pes_+<M;<ZrNW2w?4j|tGFh>P;@AoB;9h*%kuU?x!pOjzcKW&H*=dHOkTm;6E
zzU2<am%c^8cKKxyunbwY=y&nRc%%OK_tGwE82`NgF@{hEx(5(Ntj7YTmqpxMxxS>9
zy0*GZIuIxJh$Uk~uVxJeV)ws%l@EfD*0*7kTQ3Cb1&$LL!O>r;yIh#R1dXS^%fi<%
zA7OI2zO_-DYfn6ylDp$*x;k9Y1XPWvfbx+zUs&W#el7;qf0)BN<Jqh8*3|x!wQlp_
zpO~>?qRB^JbQiYdTFRE{hMO!`=dRA&&Vk%q*taV7wvhs<OO~ml8IGl*jHw3+2tm1Z
zihEJ_V3V#uIX-h@1#aN)!G22Svyb#P$9c-*($bytt}w$Gf4>kSyOef4x5gE)(zH#6
z@mhN__+V}sdMHn#JV&pYmc+vx2Q;4yo;9&XHOIMQb$1HgdJ%v3r|bXoukW7V?48<X
z{l&Nie)i>&u@_B7?N>3H8&(TL05u8=bL2Ap*%Dy=+H=!b_foSx(O?RiJaya?yDK?j
zwp;F>mY*kIo_sCBG_^KmnPYIdcFA%rOAf^eOl&K|Z}x&#IAX`M1)uqitr(~*9`A1)
z_|6tw=PfkH*;Y!1cWT%zGA+$DeA-U`F-q6pLvtT6tMBc{$9o7iX#$lD)!F6?RbAzW
zTO5GHJS7<6e8q`rVnxJdRaMkeGTgzQ!FBjz3XeI~(oJK^U05<m?{`r%i2}mR9p?&N
zLT|Ub#u@9~K+Z*?1MkSMV$2v|pv0y<uYocY_fB%&#8%sfz{T-<`1doG=N!aZS#55`
z3S9lJ?G~cI=m{5y)Wn;y7Bqgasv0!O<txKyo^ZzzI0a;p#dK2q8%1lD)BFK5ReQ(=
z*dy(|7}^KX`*9Eb4|^5Dj4UBd+2KtwHbEYB`#Wt$WdK)H2jN_wOTrISzSqmRv5H9X
zAM`w1@Fci9&}(Q*Hn^ww+ie5RCEZ(*cz_7P7zUUJay2`?>V#!O4)DXpsQRwfvR5eY
z_q+EwaY6I@oJ(mXNi4G60|MY{DwN*>OcEro5Mjm>PFM9Kt)HwS_Z;T;68Q-|BV~ZM
z_HpL0<`_gbvtQ5!*S2n7j`$Nr_T>3S@7mu}2l$|lxc);fpoTlxS}GdfG_m43=?XBx
zW+Ml07G>2EN!aHO&+-Em67zY0hcWi&F6b|4(i(b7Ai=4pkp3YEi&?%7*iQ!n!po$X
zMj5S0Ld+{$v4Iz2>-KzoEM+{upNMe+Q~1CJZi7Qg7+>q*mB-WBSS#Hoi4x?=Z`CkN
z5Xwvul8sEUBTW-BO)B}H#*!w9eN@i$RFjQ_XJQoTK!QRVP&3_OnPMA3?M-%|;S@IK
z`^DM&7mDx<@fyK@E(7T&Sgea8O7T~<uA(m{0Wl{wn6xB&+ouhdC&Jzr(cqMArD*B5
zR2tw_qV8Zu#E}hs+F*FB#?U!Z5r>BXy#aJ%`zrI|X}UywxHcGUrfKAEnBQa58>S2c
zq3TmtljbXn`{-W~3kXV|f*NyQhCI=x)gLNUxBdMa&iKF8G`wL>>3nqW^jrOrsZh9J
z@tMGHjB5RUE!?f;A-7RAtT1>A_|iDuNcJ&^v2qjb#<0JiMUnyoeNpZ!$&vWh7ZF=p
f2L)Rb<`I3->-4xjM8)YD@FV$NR<u}H*Z2Pb#DiJc

literal 0
HcmV?d00001

diff --git a/serverless/nav/serverless-general.docnav.json b/serverless/nav/serverless-general.docnav.json
index cfc371ca..f0591821 100644
--- a/serverless/nav/serverless-general.docnav.json
+++ b/serverless/nav/serverless-general.docnav.json
@@ -63,6 +63,9 @@
     },
     {
       "slug": "/serverless/regions"
+    },
+    {
+      "slug": "/serverless/general/ml-nlp-auto-scale"
     }
   ]
 }
diff --git a/serverless/pages/ml-nlp-auto-scale.mdx b/serverless/pages/ml-nlp-auto-scale.mdx
new file mode 100644
index 00000000..1c1cafbd
--- /dev/null
+++ b/serverless/pages/ml-nlp-auto-scale.mdx
@@ -0,0 +1,132 @@
+---
+slug: /serverless/general/ml-nlp-auto-scale
+title: Trained model autoscaling
+tags: ['serverless']
+---
+
+You can enable autoscaling for each of your trained model deployments.
+Autoscaling allows Elasticsearch to automatically adjust the resources the model deployment can use based on the workload demand.
+
+There are two ways to enable autoscaling:
+
+- through APIs by enabling adaptive allocations
+- in Kibana by enabling adaptive resources
+
+<DocCallOut color="primary" title="Note">
+  To fully leverage model autoscaling, it is highly recommended to enable [Elasticsearch deployment autoscaling](https://www.elastic.co/guide/en/cloud/current/ec-autoscaling.html).
+</DocCallOut>
+
+Trained model autoscaling is available for Search, Observability, and Security projects on serverless deployments. However, these projects handle processing power differently, which impacts their costs and resource limits.
+
+Security and Observability projects are only charged for data collection (ingest) and storage (retention). They are not charged for processing power (vCPU usage), which is used for more complex operations, like running advanced search models. For example, in Search projects, models such as ELSER require significant processing power to provide more accurate search results.
+
+Because vCPU processing is costly, Search projects are given access to more processing resources, while Security and Observability projects have lower limits on their processing power. This difference is reflected in the UI configuration: Search projects have higher resource limits compared to Security and Observability projects to accommodate their more complex operations.
+
+On serverless, adaptive allocations are automatically enabled for all project types.
+However, the "Adaptive resources" control is not displayed in Kibana for Observability and Security projects.
+
+## Enabling autoscaling through APIs - adaptive allocations
+
+Model allocations are independent units of work for NLP tasks.
+If you set the numbers of threads and allocations for a model manually, they remain constant even when not all the available resources are fully used or when the load on the model requires more resources.
+Instead of setting the number of allocations manually, you can enable adaptive allocations to set the number of allocations based on the load on the process.
+This can help you to manage performance and cost more easily.
+(Refer to the [pricing calculator](https://cloud.elastic.co/pricing) to learn more about the possible costs.)
+
+When adaptive allocations are enabled, the number of allocations of the model is set automatically based on the current load.
+When the load is high, a new model allocation is automatically created.
+When the load is low, a model allocation is automatically removed.
+You can explicitely set the minimum and maximum number of allocations; autoscaling will occur within these limits.
+
+<DocCallOut color="primary" title="Note">
+  If you manually set the minimum number of Machine Learning (ML) resources (allocations) to 1, you will be charged every hour, even if the system isn’t fully using those resources.
+</DocCallOut>
+
+You can enable adaptive allocations by using:
+
+- the create inference endpoint API for [ELSER](https://www.elastic.co/guide/en/elasticsearch/reference/master/infer-service-elser.html), [E5 and models uploaded through Eland](https://www.elastic.co/guide/en/elasticsearch/reference/master/infer-service-elasticsearch.html) that are used as inference services.
+- the [start trained model deployment](https://www.elastic.co/guide/en/elasticsearch/reference/master/start-trained-model-deployment.html) or [update trained model deployment](https://www.elastic.co/guide/en/elasticsearch/reference/master/update-trained-model-deployment.html) APIs for trained models that are deployed on machine learning nodes.
+
+If the new allocations fit on the current machine learning nodes, they are immediately started.
+If more resource capacity is needed for creating new model allocations, then your machine learning node will be scaled up if machine learning autoscaling is enabled to provide enough resources for the new allocation.
+The number of model allocations can be scaled down to 0.
+They cannot be scaled up to more than 32 allocations, unless you explicitly set the maximum number of allocations to more.
+Adaptive allocations must be set up independently for each deployment and [inference endpoint](https://www.elastic.co/guide/en/elasticsearch/reference/master/put-inference-api.html).
+
+When you create inference endpoints on serverless deployments using Kibana, adaptive allocations are automatically turned on, and there is no option to disable them.
+
+### Optimizing for typical use cases
+
+You can optimize your model deployment for typical use cases, such as search and ingest.
+When you optimize for ingest, the throughput will be higher, which increases the number of inference requests that can be performed in parallel.
+When you optimize for search, the latency will be lower during search processes.
+
+- If you want to optimize for ingest, set the number of threads to `1` (`"threads_per_allocation": 1`).
+- If you want to optimize for search, set the number of threads to greater than `1`.
+Increasing the number of threads will make the search processes more performant.
+
+## Enabling autoscaling in Kibana - adaptive resources
+
+You can enable adaptive resources for your models when starting or updating the model deployment.
+Adaptive resources make it possible for Elasticsearch to scale up or down the available resources based on the load on the process.
+This can help you to manage performance and cost more easily.
+When adaptive resources are enabled, the number of vCPUs that the model deployment uses is set automatically based on the current load.
+When the load is high, the number of vCPUs that the process can use is automatically increased.
+When the load is low, the number of vCPUs that the process can use is automatically decreased.
+
+You can choose from three levels of resource usage for your trained model deployment; autoscaling will occur within the selected level's range.
+
+Refer to the tables in the auto-scaling-matrix section to find out the setings for the level you selected.
+
+<DocImage size="xxl" url="../images/ml-nlp-deployment.png" alt="ML model deployment with adaptive resources enabled." /> 
+
+## Model deployment resource matrix
+
+The used resources for trained model deployments depend on three factors:
+
+- your cluster environment (Serverless, Cloud, or on-premises)
+- the use case you optimize the model deployment for (ingest or search)
+- whether model autoscaling is enabled with adaptive allocations/resources to have dynamic resources, or disabled for static resources
+
+The following tables show you the number of allocations, threads, and vCPUs available on serverless when adaptive resources are enabled or disabled.
+
+### Deployments on serverless optimized for ingest
+
+In case of ingest-optimized deployments, we maximize the number of model allocations.
+
+#### Adaptive resources enabled
+
+| Level  | Allocations                                          | Threads | vCPUs                                                |
+|--------|------------------------------------------------------|---------|------------------------------------------------------|
+| Low    | 0 to 2 dynamically                                  | 1       | 0 to 2 dynamically                                  |
+| Medium | 1 to 32 dynamically                                 | 1       | 1 to 32 dynamically                                 |
+| High   | 1 to 512 for Search <br /> 1 to 128 for Security and Observability | 1 | 1 to 512 for Search <br /> 1 to 128 for Security and Observability |
+
+#### Adaptive resources disabled (Search only)
+
+| Level  | Allocations                                          | Threads | vCPUs                                                |
+|--------|------------------------------------------------------|---------|------------------------------------------------------|
+| Low    | Exactly 2                                            | 1       | 2                                                    |
+| Medium | Exactly 32                                           | 1       | 32                                                   |
+| High   | 512 for Search <br /> No static allocations for Security and Observability | 1 | 512 for Search <br /> No static allocations for Security and Observability |
+
+### Deployments on serverless optimized for search
+
+In case of search-optimized deployments, we maximize the number of threads.
+The maximum number of threads that can be claimed depends on the hardware your architecture has.
+
+#### Adaptive resources enabled
+
+| Level  | Allocations                                          | Threads | vCPUs                                                |
+|--------|------------------------------------------------------|---------|------------------------------------------------------|
+| Low    | 0 to 1 dynamically                                  | Always 2       | 0 to 2 dynamically                                  |
+| Medium | 1 to 2 (if threads=16), dinamically                                | Maximum (for example, 16)       | 1 to 32 dynamically                                 |
+| High   | 1 to 32 (if threads=16), dinamically | Maximum (for example, 16) | 1 to 512 in Search <br /> 1 to 128 for Security and Observability |
+
+#### Adaptive resources disabled
+
+| Level  | Allocations                                             | Threads                | vCPUs                                                |
+|--------|---------------------------------------------------------|------------------------|------------------------------------------------------|
+| Low    | 1 statically                                             | Always 2               | 2                                                     |
+| Medium | 2 statically (if threads=16)                             | Maximum (for example, 16)         | 32                                                    |
+| High   | 32 statically (if threads=16) for Search <br /> No static allocations for Security and Observability | Maximum (for example, 16) | 512 for Search <br /> No static allocations for Security and Observability |
\ No newline at end of file

From d8dde76d6cfb34367ebf16d759323bde76b6ff29 Mon Sep 17 00:00:00 2001
From: kosabogi <105062005+kosabogi@users.noreply.github.com>
Date: Fri, 25 Oct 2024 11:46:39 +0200
Subject: [PATCH 02/19] Update serverless/pages/ml-nlp-auto-scale.mdx
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Co-authored-by: István Zoltán Szabó <istvan.szabo@elastic.co>
---
 serverless/pages/ml-nlp-auto-scale.mdx | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/serverless/pages/ml-nlp-auto-scale.mdx b/serverless/pages/ml-nlp-auto-scale.mdx
index 1c1cafbd..85b56ac4 100644
--- a/serverless/pages/ml-nlp-auto-scale.mdx
+++ b/serverless/pages/ml-nlp-auto-scale.mdx
@@ -20,7 +20,7 @@ Trained model autoscaling is available for Search, Observability, and Security p
 
 Security and Observability projects are only charged for data collection (ingest) and storage (retention). They are not charged for processing power (vCPU usage), which is used for more complex operations, like running advanced search models. For example, in Search projects, models such as ELSER require significant processing power to provide more accurate search results.
 
-Because vCPU processing is costly, Search projects are given access to more processing resources, while Security and Observability projects have lower limits on their processing power. This difference is reflected in the UI configuration: Search projects have higher resource limits compared to Security and Observability projects to accommodate their more complex operations.
+Search projects are given access to more processing resources, while Security and Observability projects have lower limits. This difference is reflected in the UI configuration: Search projects have higher resource limits compared to Security and Observability projects to accommodate their more complex operations.
 
 On serverless, adaptive allocations are automatically enabled for all project types.
 However, the "Adaptive resources" control is not displayed in Kibana for Observability and Security projects.

From f6b650d53e565c198281e32071054a9effb8e3e0 Mon Sep 17 00:00:00 2001
From: kosabogi <105062005+kosabogi@users.noreply.github.com>
Date: Fri, 25 Oct 2024 11:47:44 +0200
Subject: [PATCH 03/19] Update serverless/pages/ml-nlp-auto-scale.mdx
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Co-authored-by: István Zoltán Szabó <istvan.szabo@elastic.co>
---
 serverless/pages/ml-nlp-auto-scale.mdx | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/serverless/pages/ml-nlp-auto-scale.mdx b/serverless/pages/ml-nlp-auto-scale.mdx
index 85b56ac4..8e5d47dc 100644
--- a/serverless/pages/ml-nlp-auto-scale.mdx
+++ b/serverless/pages/ml-nlp-auto-scale.mdx
@@ -22,7 +22,7 @@ Security and Observability projects are only charged for data collection (ingest
 
 Search projects are given access to more processing resources, while Security and Observability projects have lower limits. This difference is reflected in the UI configuration: Search projects have higher resource limits compared to Security and Observability projects to accommodate their more complex operations.
 
-On serverless, adaptive allocations are automatically enabled for all project types.
+On Serverless, adaptive allocations are automatically enabled for all project types.
 However, the "Adaptive resources" control is not displayed in Kibana for Observability and Security projects.
 
 ## Enabling autoscaling through APIs - adaptive allocations

From 4683ddc96da9c586d656dfd18db9bc6d36c1edde Mon Sep 17 00:00:00 2001
From: kosabogi <105062005+kosabogi@users.noreply.github.com>
Date: Fri, 25 Oct 2024 11:48:53 +0200
Subject: [PATCH 04/19] Update serverless/pages/ml-nlp-auto-scale.mdx
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Co-authored-by: István Zoltán Szabó <istvan.szabo@elastic.co>
---
 serverless/pages/ml-nlp-auto-scale.mdx | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/serverless/pages/ml-nlp-auto-scale.mdx b/serverless/pages/ml-nlp-auto-scale.mdx
index 8e5d47dc..48ee39e1 100644
--- a/serverless/pages/ml-nlp-auto-scale.mdx
+++ b/serverless/pages/ml-nlp-auto-scale.mdx
@@ -39,7 +39,7 @@ When the load is low, a model allocation is automatically removed.
 You can explicitely set the minimum and maximum number of allocations; autoscaling will occur within these limits.
 
 <DocCallOut color="primary" title="Note">
-  If you manually set the minimum number of Machine Learning (ML) resources (allocations) to 1, you will be charged every hour, even if the system isn’t fully using those resources.
+  If you set the minimum number of allocations to 1, you will be charged even if the system is not using those resources.
 </DocCallOut>
 
 You can enable adaptive allocations by using:

From f91c731b506a1f3a385aad08112766907dde737a Mon Sep 17 00:00:00 2001
From: kosabogi <105062005+kosabogi@users.noreply.github.com>
Date: Fri, 25 Oct 2024 11:50:12 +0200
Subject: [PATCH 05/19] Update serverless/pages/ml-nlp-auto-scale.mdx
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Co-authored-by: István Zoltán Szabó <istvan.szabo@elastic.co>
---
 serverless/pages/ml-nlp-auto-scale.mdx | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/serverless/pages/ml-nlp-auto-scale.mdx b/serverless/pages/ml-nlp-auto-scale.mdx
index 48ee39e1..f5507bb2 100644
--- a/serverless/pages/ml-nlp-auto-scale.mdx
+++ b/serverless/pages/ml-nlp-auto-scale.mdx
@@ -88,7 +88,7 @@ The used resources for trained model deployments depend on three factors:
 - the use case you optimize the model deployment for (ingest or search)
 - whether model autoscaling is enabled with adaptive allocations/resources to have dynamic resources, or disabled for static resources
 
-The following tables show you the number of allocations, threads, and vCPUs available on serverless when adaptive resources are enabled or disabled.
+The following tables show you the number of allocations, threads, and vCUs available on Serverless when adaptive resources are enabled or disabled.
 
 ### Deployments on serverless optimized for ingest
 

From f17b2a412d7b0a727f2fb111a6d0aeaa6d7005aa Mon Sep 17 00:00:00 2001
From: kosabogi <105062005+kosabogi@users.noreply.github.com>
Date: Fri, 25 Oct 2024 11:50:34 +0200
Subject: [PATCH 06/19] Update serverless/pages/ml-nlp-auto-scale.mdx
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Co-authored-by: István Zoltán Szabó <istvan.szabo@elastic.co>
---
 serverless/pages/ml-nlp-auto-scale.mdx | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/serverless/pages/ml-nlp-auto-scale.mdx b/serverless/pages/ml-nlp-auto-scale.mdx
index f5507bb2..7c9a7e00 100644
--- a/serverless/pages/ml-nlp-auto-scale.mdx
+++ b/serverless/pages/ml-nlp-auto-scale.mdx
@@ -96,7 +96,7 @@ In case of ingest-optimized deployments, we maximize the number of model allocat
 
 #### Adaptive resources enabled
 
-| Level  | Allocations                                          | Threads | vCPUs                                                |
+| Level  | Allocations                                          | Threads | vCUs                                                |
 |--------|------------------------------------------------------|---------|------------------------------------------------------|
 | Low    | 0 to 2 dynamically                                  | 1       | 0 to 2 dynamically                                  |
 | Medium | 1 to 32 dynamically                                 | 1       | 1 to 32 dynamically                                 |

From 26fb9e08269d26278f6fbc32cd4650d3e3b2a548 Mon Sep 17 00:00:00 2001
From: kosabogi <105062005+kosabogi@users.noreply.github.com>
Date: Fri, 25 Oct 2024 11:50:49 +0200
Subject: [PATCH 07/19] Update serverless/pages/ml-nlp-auto-scale.mdx
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Co-authored-by: István Zoltán Szabó <istvan.szabo@elastic.co>
---
 serverless/pages/ml-nlp-auto-scale.mdx | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/serverless/pages/ml-nlp-auto-scale.mdx b/serverless/pages/ml-nlp-auto-scale.mdx
index 7c9a7e00..38407ae4 100644
--- a/serverless/pages/ml-nlp-auto-scale.mdx
+++ b/serverless/pages/ml-nlp-auto-scale.mdx
@@ -70,9 +70,9 @@ Increasing the number of threads will make the search processes more performant.
 You can enable adaptive resources for your models when starting or updating the model deployment.
 Adaptive resources make it possible for Elasticsearch to scale up or down the available resources based on the load on the process.
 This can help you to manage performance and cost more easily.
-When adaptive resources are enabled, the number of vCPUs that the model deployment uses is set automatically based on the current load.
-When the load is high, the number of vCPUs that the process can use is automatically increased.
-When the load is low, the number of vCPUs that the process can use is automatically decreased.
+When adaptive resources are enabled, the number of vCUs that the model deployment uses is set automatically based on the current load.
+When the load is high, the number of vCUs that the process can use is automatically increased.
+When the load is low, the number of vCUs that the process can use is automatically decreased.
 
 You can choose from three levels of resource usage for your trained model deployment; autoscaling will occur within the selected level's range.
 

From c8b8b3bcf4b876b9a7c07e3f1e79b7b1a8250577 Mon Sep 17 00:00:00 2001
From: kosabogi <105062005+kosabogi@users.noreply.github.com>
Date: Fri, 25 Oct 2024 11:51:05 +0200
Subject: [PATCH 08/19] Update serverless/pages/ml-nlp-auto-scale.mdx
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Co-authored-by: István Zoltán Szabó <istvan.szabo@elastic.co>
---
 serverless/pages/ml-nlp-auto-scale.mdx | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/serverless/pages/ml-nlp-auto-scale.mdx b/serverless/pages/ml-nlp-auto-scale.mdx
index 38407ae4..51c3284f 100644
--- a/serverless/pages/ml-nlp-auto-scale.mdx
+++ b/serverless/pages/ml-nlp-auto-scale.mdx
@@ -53,7 +53,7 @@ The number of model allocations can be scaled down to 0.
 They cannot be scaled up to more than 32 allocations, unless you explicitly set the maximum number of allocations to more.
 Adaptive allocations must be set up independently for each deployment and [inference endpoint](https://www.elastic.co/guide/en/elasticsearch/reference/master/put-inference-api.html).
 
-When you create inference endpoints on serverless deployments using Kibana, adaptive allocations are automatically turned on, and there is no option to disable them.
+When you create inference endpoints on Serverless using Kibana, adaptive allocations are automatically turned on, and there is no option to disable them.
 
 ### Optimizing for typical use cases
 

From 55117e6e50fcd7a120d1c441fb9814dabb2cf96e Mon Sep 17 00:00:00 2001
From: kosabogi <105062005+kosabogi@users.noreply.github.com>
Date: Fri, 25 Oct 2024 11:51:23 +0200
Subject: [PATCH 09/19] Update serverless/pages/ml-nlp-auto-scale.mdx
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Co-authored-by: István Zoltán Szabó <istvan.szabo@elastic.co>
---
 serverless/pages/ml-nlp-auto-scale.mdx | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/serverless/pages/ml-nlp-auto-scale.mdx b/serverless/pages/ml-nlp-auto-scale.mdx
index 51c3284f..67c01ffd 100644
--- a/serverless/pages/ml-nlp-auto-scale.mdx
+++ b/serverless/pages/ml-nlp-auto-scale.mdx
@@ -104,7 +104,7 @@ In case of ingest-optimized deployments, we maximize the number of model allocat
 
 #### Adaptive resources disabled (Search only)
 
-| Level  | Allocations                                          | Threads | vCPUs                                                |
+| Level  | Allocations                                          | Threads | vCUs                                                |
 |--------|------------------------------------------------------|---------|------------------------------------------------------|
 | Low    | Exactly 2                                            | 1       | 2                                                    |
 | Medium | Exactly 32                                           | 1       | 32                                                   |

From ac3a20196d3c9312cea14dc0d2f2a7916309fe76 Mon Sep 17 00:00:00 2001
From: kosabogi <105062005+kosabogi@users.noreply.github.com>
Date: Fri, 25 Oct 2024 11:51:42 +0200
Subject: [PATCH 10/19] Update serverless/pages/ml-nlp-auto-scale.mdx
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Co-authored-by: István Zoltán Szabó <istvan.szabo@elastic.co>
---
 serverless/pages/ml-nlp-auto-scale.mdx | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/serverless/pages/ml-nlp-auto-scale.mdx b/serverless/pages/ml-nlp-auto-scale.mdx
index 67c01ffd..77234400 100644
--- a/serverless/pages/ml-nlp-auto-scale.mdx
+++ b/serverless/pages/ml-nlp-auto-scale.mdx
@@ -125,7 +125,7 @@ The maximum number of threads that can be claimed depends on the hardware your a
 
 #### Adaptive resources disabled
 
-| Level  | Allocations                                             | Threads                | vCPUs                                                |
+| Level  | Allocations                                             | Threads                | vCUs                                                |
 |--------|---------------------------------------------------------|------------------------|------------------------------------------------------|
 | Low    | 1 statically                                             | Always 2               | 2                                                     |
 | Medium | 2 statically (if threads=16)                             | Maximum (for example, 16)         | 32                                                    |

From 18487de52ca707388ff517f3ee381a3d747ea41c Mon Sep 17 00:00:00 2001
From: kosabogi <105062005+kosabogi@users.noreply.github.com>
Date: Fri, 25 Oct 2024 11:52:08 +0200
Subject: [PATCH 11/19] Update serverless/pages/ml-nlp-auto-scale.mdx
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Co-authored-by: István Zoltán Szabó <istvan.szabo@elastic.co>
---
 serverless/pages/ml-nlp-auto-scale.mdx | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/serverless/pages/ml-nlp-auto-scale.mdx b/serverless/pages/ml-nlp-auto-scale.mdx
index 77234400..0ea61820 100644
--- a/serverless/pages/ml-nlp-auto-scale.mdx
+++ b/serverless/pages/ml-nlp-auto-scale.mdx
@@ -18,7 +18,7 @@ There are two ways to enable autoscaling:
 
 Trained model autoscaling is available for Search, Observability, and Security projects on serverless deployments. However, these projects handle processing power differently, which impacts their costs and resource limits.
 
-Security and Observability projects are only charged for data collection (ingest) and storage (retention). They are not charged for processing power (vCPU usage), which is used for more complex operations, like running advanced search models. For example, in Search projects, models such as ELSER require significant processing power to provide more accurate search results.
+Security and Observability projects are only charged for data ingestion and retention. They are not charged for processing power (vCU usage), which is used for more complex operations, like running advanced search models. For example, in Search projects, models such as ELSER require significant processing power to provide more accurate search results.
 
 Search projects are given access to more processing resources, while Security and Observability projects have lower limits. This difference is reflected in the UI configuration: Search projects have higher resource limits compared to Security and Observability projects to accommodate their more complex operations.
 

From e430ce7e35d31075389229f37f3f7d94e0a303c9 Mon Sep 17 00:00:00 2001
From: kosabogi <105062005+kosabogi@users.noreply.github.com>
Date: Fri, 25 Oct 2024 11:52:27 +0200
Subject: [PATCH 12/19] Update serverless/pages/ml-nlp-auto-scale.mdx
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Co-authored-by: István Zoltán Szabó <istvan.szabo@elastic.co>
---
 serverless/pages/ml-nlp-auto-scale.mdx | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/serverless/pages/ml-nlp-auto-scale.mdx b/serverless/pages/ml-nlp-auto-scale.mdx
index 0ea61820..7184c9b5 100644
--- a/serverless/pages/ml-nlp-auto-scale.mdx
+++ b/serverless/pages/ml-nlp-auto-scale.mdx
@@ -117,7 +117,7 @@ The maximum number of threads that can be claimed depends on the hardware your a
 
 #### Adaptive resources enabled
 
-| Level  | Allocations                                          | Threads | vCPUs                                                |
+| Level  | Allocations                                          | Threads | vCUs                                                |
 |--------|------------------------------------------------------|---------|------------------------------------------------------|
 | Low    | 0 to 1 dynamically                                  | Always 2       | 0 to 2 dynamically                                  |
 | Medium | 1 to 2 (if threads=16), dinamically                                | Maximum (for example, 16)       | 1 to 32 dynamically                                 |

From 27b445c4ab2c1220ac9b7f14d98db37de3ff4af8 Mon Sep 17 00:00:00 2001
From: kosabogi <105062005+kosabogi@users.noreply.github.com>
Date: Fri, 25 Oct 2024 11:52:42 +0200
Subject: [PATCH 13/19] Update serverless/pages/ml-nlp-auto-scale.mdx
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Co-authored-by: István Zoltán Szabó <istvan.szabo@elastic.co>
---
 serverless/pages/ml-nlp-auto-scale.mdx | 1 -
 1 file changed, 1 deletion(-)

diff --git a/serverless/pages/ml-nlp-auto-scale.mdx b/serverless/pages/ml-nlp-auto-scale.mdx
index 7184c9b5..78fc03bb 100644
--- a/serverless/pages/ml-nlp-auto-scale.mdx
+++ b/serverless/pages/ml-nlp-auto-scale.mdx
@@ -113,7 +113,6 @@ In case of ingest-optimized deployments, we maximize the number of model allocat
 ### Deployments on serverless optimized for search
 
 In case of search-optimized deployments, we maximize the number of threads.
-The maximum number of threads that can be claimed depends on the hardware your architecture has.
 
 #### Adaptive resources enabled
 

From 8f7a9d207c774819b0071b139b1cbfcf5408a420 Mon Sep 17 00:00:00 2001
From: kosabogi <boglarka.kosa@elastic.co>
Date: Fri, 25 Oct 2024 12:01:29 +0200
Subject: [PATCH 14/19] Changes paragraph placement

---
 serverless/pages/ml-nlp-auto-scale.mdx | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/serverless/pages/ml-nlp-auto-scale.mdx b/serverless/pages/ml-nlp-auto-scale.mdx
index 78fc03bb..84122128 100644
--- a/serverless/pages/ml-nlp-auto-scale.mdx
+++ b/serverless/pages/ml-nlp-auto-scale.mdx
@@ -20,11 +20,6 @@ Trained model autoscaling is available for Search, Observability, and Security p
 
 Security and Observability projects are only charged for data ingestion and retention. They are not charged for processing power (vCU usage), which is used for more complex operations, like running advanced search models. For example, in Search projects, models such as ELSER require significant processing power to provide more accurate search results.
 
-Search projects are given access to more processing resources, while Security and Observability projects have lower limits. This difference is reflected in the UI configuration: Search projects have higher resource limits compared to Security and Observability projects to accommodate their more complex operations.
-
-On Serverless, adaptive allocations are automatically enabled for all project types.
-However, the "Adaptive resources" control is not displayed in Kibana for Observability and Security projects.
-
 ## Enabling autoscaling through APIs - adaptive allocations
 
 Model allocations are independent units of work for NLP tasks.
@@ -78,7 +73,12 @@ You can choose from three levels of resource usage for your trained model deploy
 
 Refer to the tables in the auto-scaling-matrix section to find out the setings for the level you selected.
 
-<DocImage size="xxl" url="../images/ml-nlp-deployment.png" alt="ML model deployment with adaptive resources enabled." /> 
+<DocImage size="xxl" url="../images/ml-nlp-deployment.png" alt="ML model deployment with adaptive resources enabled." />
+
+Search projects are given access to more processing resources, while Security and Observability projects have lower limits. This difference is reflected in the UI configuration: Search projects have higher resource limits compared to Security and Observability projects to accommodate their more complex operations.
+
+On Serverless, adaptive allocations are automatically enabled for all project types.
+However, the "Adaptive resources" control is not displayed in Kibana for Observability and Security projects.
 
 ## Model deployment resource matrix
 

From 1beb41c6a4b99e8f20f327662c8c7ae234fcb9fd Mon Sep 17 00:00:00 2001
From: kosabogi <105062005+kosabogi@users.noreply.github.com>
Date: Wed, 30 Oct 2024 12:58:52 +0100
Subject: [PATCH 15/19] Update serverless/pages/ml-nlp-auto-scale.mdx
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

Co-authored-by: István Zoltán Szabó <istvan.szabo@elastic.co>
---
 serverless/pages/ml-nlp-auto-scale.mdx | 3 ---
 1 file changed, 3 deletions(-)

diff --git a/serverless/pages/ml-nlp-auto-scale.mdx b/serverless/pages/ml-nlp-auto-scale.mdx
index 84122128..126c8861 100644
--- a/serverless/pages/ml-nlp-auto-scale.mdx
+++ b/serverless/pages/ml-nlp-auto-scale.mdx
@@ -12,9 +12,6 @@ There are two ways to enable autoscaling:
 - through APIs by enabling adaptive allocations
 - in Kibana by enabling adaptive resources
 
-<DocCallOut color="primary" title="Note">
-  To fully leverage model autoscaling, it is highly recommended to enable [Elasticsearch deployment autoscaling](https://www.elastic.co/guide/en/cloud/current/ec-autoscaling.html).
-</DocCallOut>
 
 Trained model autoscaling is available for Search, Observability, and Security projects on serverless deployments. However, these projects handle processing power differently, which impacts their costs and resource limits.
 

From 031c91f1fdea7a2003be7e8f4d307cf7d421ed8f Mon Sep 17 00:00:00 2001
From: kosabogi <105062005+kosabogi@users.noreply.github.com>
Date: Thu, 31 Oct 2024 15:57:23 +0100
Subject: [PATCH 16/19] Update serverless/pages/ml-nlp-auto-scale.mdx

Co-authored-by: Dima Arnautov <arnautov.dima@gmail.com>
---
 serverless/pages/ml-nlp-auto-scale.mdx | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/serverless/pages/ml-nlp-auto-scale.mdx b/serverless/pages/ml-nlp-auto-scale.mdx
index 126c8861..07162602 100644
--- a/serverless/pages/ml-nlp-auto-scale.mdx
+++ b/serverless/pages/ml-nlp-auto-scale.mdx
@@ -62,7 +62,7 @@ Increasing the number of threads will make the search processes more performant.
 You can enable adaptive resources for your models when starting or updating the model deployment.
 Adaptive resources make it possible for Elasticsearch to scale up or down the available resources based on the load on the process.
 This can help you to manage performance and cost more easily.
-When adaptive resources are enabled, the number of vCUs that the model deployment uses is set automatically based on the current load.
+When adaptive resources are enabled, the number of VCUs that the model deployment uses is set automatically based on the current load.
 When the load is high, the number of vCUs that the process can use is automatically increased.
 When the load is low, the number of vCUs that the process can use is automatically decreased.
 

From 9fdee403d1e6d6478a9b1490285556a063a937a6 Mon Sep 17 00:00:00 2001
From: kosabogi <boglarka.kosa@elastic.co>
Date: Wed, 6 Nov 2024 08:13:04 +0100
Subject: [PATCH 17/19] Updates document based on feedback

---
 serverless/pages/ml-nlp-auto-scale.mdx | 66 +++++++++++++-------------
 1 file changed, 33 insertions(+), 33 deletions(-)

diff --git a/serverless/pages/ml-nlp-auto-scale.mdx b/serverless/pages/ml-nlp-auto-scale.mdx
index 07162602..e127294f 100644
--- a/serverless/pages/ml-nlp-auto-scale.mdx
+++ b/serverless/pages/ml-nlp-auto-scale.mdx
@@ -13,20 +13,20 @@ There are two ways to enable autoscaling:
 - in Kibana by enabling adaptive resources
 
 
-Trained model autoscaling is available for Search, Observability, and Security projects on serverless deployments. However, these projects handle processing power differently, which impacts their costs and resource limits.
+Trained model autoscaling is available for both serverless and Cloud deployments. In serverless deployments, processing power is managed differently across Search, Observability, and Security projects, which impacts their costs and resource limits.
 
-Security and Observability projects are only charged for data ingestion and retention. They are not charged for processing power (vCU usage), which is used for more complex operations, like running advanced search models. For example, in Search projects, models such as ELSER require significant processing power to provide more accurate search results.
+Security and Observability projects are only charged for data ingestion and retention. They are not charged for processing power (VCU usage), which is used for more complex operations, like running advanced search models. For example, in Search projects, models such as ELSER require significant processing power to provide more accurate search results.
 
 ## Enabling autoscaling through APIs - adaptive allocations
 
 Model allocations are independent units of work for NLP tasks.
-If you set the numbers of threads and allocations for a model manually, they remain constant even when not all the available resources are fully used or when the load on the model requires more resources.
+If you set a static number of allocations, they remain constant even when not all the available resources are fully used or when the load on the model requires more resources.
 Instead of setting the number of allocations manually, you can enable adaptive allocations to set the number of allocations based on the load on the process.
 This can help you to manage performance and cost more easily.
 (Refer to the [pricing calculator](https://cloud.elastic.co/pricing) to learn more about the possible costs.)
 
 When adaptive allocations are enabled, the number of allocations of the model is set automatically based on the current load.
-When the load is high, a new model allocation is automatically created.
+When the load is high, additional model allocations are automatically created as needed.
 When the load is low, a model allocation is automatically removed.
 You can explicitely set the minimum and maximum number of allocations; autoscaling will occur within these limits.
 
@@ -63,8 +63,8 @@ You can enable adaptive resources for your models when starting or updating the
 Adaptive resources make it possible for Elasticsearch to scale up or down the available resources based on the load on the process.
 This can help you to manage performance and cost more easily.
 When adaptive resources are enabled, the number of VCUs that the model deployment uses is set automatically based on the current load.
-When the load is high, the number of vCUs that the process can use is automatically increased.
-When the load is low, the number of vCUs that the process can use is automatically decreased.
+When the load is high, the number of VCUs that the process can use is automatically increased.
+When the load is low, the number of VCUs that the process can use is automatically decreased.
 
 You can choose from three levels of resource usage for your trained model deployment; autoscaling will occur within the selected level's range.
 
@@ -85,44 +85,44 @@ The used resources for trained model deployments depend on three factors:
 - the use case you optimize the model deployment for (ingest or search)
 - whether model autoscaling is enabled with adaptive allocations/resources to have dynamic resources, or disabled for static resources
 
-The following tables show you the number of allocations, threads, and vCUs available on Serverless when adaptive resources are enabled or disabled.
+The following tables show you the number of allocations, threads, and VCUs available on Serverless when adaptive resources are enabled or disabled.
 
 ### Deployments on serverless optimized for ingest
 
 In case of ingest-optimized deployments, we maximize the number of model allocations.
 
-#### Adaptive resources enabled
+### Adaptive Resources Enabled
 
-| Level  | Allocations                                          | Threads | vCUs                                                |
-|--------|------------------------------------------------------|---------|------------------------------------------------------|
-| Low    | 0 to 2 dynamically                                  | 1       | 0 to 2 dynamically                                  |
-| Medium | 1 to 32 dynamically                                 | 1       | 1 to 32 dynamically                                 |
-| High   | 1 to 512 for Search <br /> 1 to 128 for Security and Observability | 1 | 1 to 512 for Search <br /> 1 to 128 for Security and Observability |
+#### Ingest-Optimized Deployments
 
-#### Adaptive resources disabled (Search only)
+| Level  | Allocations                                          | Threads | VCUs                                                |
+|--------|------------------------------------------------------|---------|-----------------------------------------------------|
+| Low    | 0 to 2 dynamically                                   | 1       | 0 to 16 dynamically                                 |
+| Medium | 1 to 32 dynamically                                  | 1       | 8 to 256 dynamically                                |
+| High   | - 1 to 512 for Search <br /> - 1 to 128 for Security and Observability | 1 | - 8 to 4096 for Search <br /> - 8 to 1024 for Security and Observability |
 
-| Level  | Allocations                                          | Threads | vCUs                                                |
-|--------|------------------------------------------------------|---------|------------------------------------------------------|
-| Low    | Exactly 2                                            | 1       | 2                                                    |
-| Medium | Exactly 32                                           | 1       | 32                                                   |
-| High   | 512 for Search <br /> No static allocations for Security and Observability | 1 | 512 for Search <br /> No static allocations for Security and Observability |
+#### Adaptive Resources Disabled (Search Only)
 
-### Deployments on serverless optimized for search
+| Level  | Allocations                                          | Threads | VCUs                                                |
+|--------|------------------------------------------------------|---------|-----------------------------------------------------|
+| Low    | Exactly 2                                            | 1       | 16                                                  |
+| Medium | Exactly 32                                           | 1       | 256                                                 |
+| High   | - 512 for Search <br /> - No static allocations for Security and Observability | 1 | - 4096 for Search <br /> - No static allocations for Security and Observability |
 
-In case of search-optimized deployments, we maximize the number of threads.
+### Deployments on Serverless Optimized for Search
 
-#### Adaptive resources enabled
+#### Adaptive Resources Enabled
 
-| Level  | Allocations                                          | Threads | vCUs                                                |
-|--------|------------------------------------------------------|---------|------------------------------------------------------|
-| Low    | 0 to 1 dynamically                                  | Always 2       | 0 to 2 dynamically                                  |
-| Medium | 1 to 2 (if threads=16), dinamically                                | Maximum (for example, 16)       | 1 to 32 dynamically                                 |
-| High   | 1 to 32 (if threads=16), dinamically | Maximum (for example, 16) | 1 to 512 in Search <br /> 1 to 128 for Security and Observability |
+| Level  | Allocations                                          | Threads | VCUs                                                |
+|--------|------------------------------------------------------|---------|-----------------------------------------------------|
+| Low    | 0 to 1 dynamically                                   | Always 2       | 0 to 16 dynamically                                 |
+| Medium | 1 to 2 (if threads=16), dynamically                  | Maximum (e.g., 16)       | 8 to 256 dynamically                               |
+| High   | - 1 to 32 (if threads=16), dynamically <br /> - 1 to 128 for Security and Observability | Maximum (e.g., 16) | - 8 to 4096 for Search <br /> - 8 to 1024 for Security and Observability |
 
-#### Adaptive resources disabled
+#### Adaptive Resources Disabled
 
-| Level  | Allocations                                             | Threads                | vCUs                                                |
-|--------|---------------------------------------------------------|------------------------|------------------------------------------------------|
-| Low    | 1 statically                                             | Always 2               | 2                                                     |
-| Medium | 2 statically (if threads=16)                             | Maximum (for example, 16)         | 32                                                    |
-| High   | 32 statically (if threads=16) for Search <br /> No static allocations for Security and Observability | Maximum (for example, 16) | 512 for Search <br /> No static allocations for Security and Observability |
\ No newline at end of file
+| Level  | Allocations                                          | Threads | VCUs                                                |
+|--------|------------------------------------------------------|---------|-----------------------------------------------------|
+| Low    | 1 statically                                         | Always 2               | 16                                                  |
+| Medium | 2 statically (if threads=16)                         | Maximum (e.g., 16)     | 256                                                 |
+| High   | - 32 statically (if threads=16) for Search <br /> - No static allocations for Security and Observability | Maximum (e.g., 16) | - 4096 for Search <br /> - No static allocations for Security and Observability |

From 37ee519ccd53192bde769036612ae60a4b9ce4b4 Mon Sep 17 00:00:00 2001
From: Colleen McGinnis <colleen.mcginnis@elastic.co>
Date: Wed, 6 Nov 2024 08:54:11 -0600
Subject: [PATCH 18/19] mdx to asciidoc

---
 serverless/index-serverless-general.asciidoc |   2 +
 serverless/pages/ml-nlp-auto-scale.asciidoc  | 210 +++++++++++++++++++
 serverless/pages/ml-nlp-auto-scale.mdx       | 128 -----------
 3 files changed, 212 insertions(+), 128 deletions(-)
 create mode 100644 serverless/pages/ml-nlp-auto-scale.asciidoc
 delete mode 100644 serverless/pages/ml-nlp-auto-scale.mdx

diff --git a/serverless/index-serverless-general.asciidoc b/serverless/index-serverless-general.asciidoc
index 08dbf303..7bf46038 100644
--- a/serverless/index-serverless-general.asciidoc
+++ b/serverless/index-serverless-general.asciidoc
@@ -32,3 +32,5 @@ include::./pages/service-status.asciidoc[leveloffset=+2]
 include::./pages/user-profile.asciidoc[leveloffset=+2]
 
 include::./pages/cloud-regions.asciidoc[leveloffset=+2]
+
+include::./pages/ml-nlp-auto-scale.asciidoc[leveloffset=+2]
diff --git a/serverless/pages/ml-nlp-auto-scale.asciidoc b/serverless/pages/ml-nlp-auto-scale.asciidoc
new file mode 100644
index 00000000..347bf1fc
--- /dev/null
+++ b/serverless/pages/ml-nlp-auto-scale.asciidoc
@@ -0,0 +1,210 @@
+[[general-ml-nlp-auto-scale]]
+= Trained model autoscaling
+
+// :keywords: serverless
+
+You can enable autoscaling for each of your trained model deployments.
+Autoscaling allows Elasticsearch to automatically adjust the resources the model deployment can use based on the workload demand.
+
+There are two ways to enable autoscaling:
+
+* through APIs by enabling adaptive allocations
+* in Kibana by enabling adaptive resources
+
+
+Trained model autoscaling is available for both serverless and Cloud deployments. In serverless deployments, processing power is managed differently across Search, Observability, and Security projects, which impacts their costs and resource limits.
+
+Security and Observability projects are only charged for data ingestion and retention. They are not charged for processing power (VCU usage), which is used for more complex operations, like running advanced search models. For example, in Search projects, models such as ELSER require significant processing power to provide more accurate search results.
+
+[discrete]
+[[enabling-autoscaling-through-apis-adaptive-allocations]]
+== Enabling autoscaling through APIs - adaptive allocations
+
+Model allocations are independent units of work for NLP tasks.
+If you set a static number of allocations, they remain constant even when not all the available resources are fully used or when the load on the model requires more resources.
+Instead of setting the number of allocations manually, you can enable adaptive allocations to set the number of allocations based on the load on the process.
+This can help you to manage performance and cost more easily.
+(Refer to the https://cloud.elastic.co/pricing[pricing calculator] to learn more about the possible costs.)
+
+When adaptive allocations are enabled, the number of allocations of the model is set automatically based on the current load.
+When the load is high, additional model allocations are automatically created as needed.
+When the load is low, a model allocation is automatically removed.
+You can explicitly set the minimum and maximum number of allocations; autoscaling will occur within these limits.
+
+[NOTE]
+====
+If you set the minimum number of allocations to 1, you will be charged even if the system is not using those resources.
+====
+
+You can enable adaptive allocations by using:
+
+* the create inference endpoint API for https://www.elastic.co/guide/en/elasticsearch/reference/master/infer-service-elser.html[ELSER], https://www.elastic.co/guide/en/elasticsearch/reference/master/infer-service-elasticsearch.html[E5 and models uploaded through Eland] that are used as inference services.
+* the https://www.elastic.co/guide/en/elasticsearch/reference/master/start-trained-model-deployment.html[start trained model deployment] or https://www.elastic.co/guide/en/elasticsearch/reference/master/update-trained-model-deployment.html[update trained model deployment] APIs for trained models that are deployed on machine learning nodes.
+
+If the new allocations fit on the current machine learning nodes, they are immediately started.
+If more resource capacity is needed for creating new model allocations, then your machine learning node will be scaled up if machine learning autoscaling is enabled to provide enough resources for the new allocation.
+The number of model allocations can be scaled down to 0.
+They cannot be scaled up to more than 32 allocations, unless you explicitly set the maximum number of allocations to more.
+Adaptive allocations must be set up independently for each deployment and https://www.elastic.co/guide/en/elasticsearch/reference/master/put-inference-api.html[inference endpoint].
+
+When you create inference endpoints on Serverless using Kibana, adaptive allocations are automatically turned on, and there is no option to disable them.
+
+[discrete]
+[[optimizing-for-typical-use-cases]]
+=== Optimizing for typical use cases
+
+You can optimize your model deployment for typical use cases, such as search and ingest.
+When you optimize for ingest, the throughput will be higher, which increases the number of inference requests that can be performed in parallel.
+When you optimize for search, the latency will be lower during search processes.
+
+* If you want to optimize for ingest, set the number of threads to `1` (`"threads_per_allocation": 1`).
+* If you want to optimize for search, set the number of threads to greater than `1`.
+Increasing the number of threads will make the search processes more performant.
+
+[discrete]
+[[enabling-autoscaling-in-kibana-adaptive-resources]]
+== Enabling autoscaling in Kibana - adaptive resources
+
+You can enable adaptive resources for your models when starting or updating the model deployment.
+Adaptive resources make it possible for Elasticsearch to scale up or down the available resources based on the load on the process.
+This can help you to manage performance and cost more easily.
+When adaptive resources are enabled, the number of VCUs that the model deployment uses is set automatically based on the current load.
+When the load is high, the number of VCUs that the process can use is automatically increased.
+When the load is low, the number of VCUs that the process can use is automatically decreased.
+
+You can choose from three levels of resource usage for your trained model deployment; autoscaling will occur within the selected level's range.
+
+Refer to the tables in the auto-scaling-matrix section to find out the setings for the level you selected.
+
+image::images/ml-nlp-deployment.png[ML model deployment with adaptive resources enabled.]
+
+Search projects are given access to more processing resources, while Security and Observability projects have lower limits. This difference is reflected in the UI configuration: Search projects have higher resource limits compared to Security and Observability projects to accommodate their more complex operations.
+
+On Serverless, adaptive allocations are automatically enabled for all project types.
+However, the "Adaptive resources" control is not displayed in Kibana for Observability and Security projects.
+
+[discrete]
+[[model-deployment-resource-matrix]]
+== Model deployment resource matrix
+
+The used resources for trained model deployments depend on three factors:
+
+* your cluster environment (Serverless, Cloud, or on-premises)
+* the use case you optimize the model deployment for (ingest or search)
+* whether model autoscaling is enabled with adaptive allocations/resources to have dynamic resources, or disabled for static resources
+
+The following tables show you the number of allocations, threads, and VCUs available on Serverless when adaptive resources are enabled or disabled.
+
+[discrete]
+[[deployments-on-serverless-optimized-for-ingest]]
+=== Deployments on serverless optimized for ingest
+
+In case of ingest-optimized deployments, we maximize the number of model allocations.
+
+[discrete]
+[[adaptive-resources-enabled]]
+=== Adaptive Resources Enabled
+
+[discrete]
+[[ingest-optimized-deployments]]
+==== Ingest-Optimized Deployments
+
+|===
+| Level | Allocations | Threads | VCUs
+
+| Low
+| 0 to 2 dynamically
+| 1
+| 0 to 16 dynamically
+
+| Medium
+| 1 to 32 dynamically
+| 1
+| 8 to 256 dynamically
+
+| High
+a| * 1 to 512 for Search
+* 1 to 128 for Security and Observability
+| 1
+a| * 8 to 4096 for Search
+* 8 to 1024 for Security and Observability
+|===
+
+[discrete]
+[[adaptive-resources-disabled-search-only]]
+==== Adaptive Resources Disabled (Search Only)
+
+|===
+| Level | Allocations | Threads | VCUs
+
+| Low
+| Exactly 2
+| 1
+| 16
+
+| Medium
+| Exactly 32
+| 1
+| 256
+
+| High
+a| * 512 for Search
+* No static allocations for Security and Observability
+| 1
+a| * 4096 for Search
+* No static allocations for Security and Observability |
+|===
+
+[discrete]
+[[deployments-on-serverless-optimized-for-search]]
+=== Deployments on Serverless Optimized for Search
+
+[discrete]
+[[adaptive-resources-enabled-for-search]]
+==== Adaptive Resources Enabled
+
+|===
+| Level | Allocations | Threads | VCUs
+
+| Low
+| 0 to 1 dynamically
+| Always 2
+| 0 to 16 dynamically
+
+| Medium
+| 1 to 2 (if threads=16), dynamically
+| Maximum (e.g., 16)
+| 8 to 256 dynamically
+
+| High
+a| * 1 to 32 (if threads=16), dynamically
+* 1 to 128 for Security and Observability
+| Maximum (e.g., 16)
+a| * 8 to 4096 for Search
+* 8 to 1024 for Security and Observability
+|===
+
+[discrete]
+[[adaptive-resources-disabled-for-search]]
+==== Adaptive Resources Disabled
+
+|===
+| Level | Allocations | Threads | VCUs
+
+| Low
+| 1 statically
+| Always 2
+| 16
+
+| Medium
+| 2 statically (if threads=16)
+| Maximum (e.g., 16)
+| 256
+
+| High
+a| * 32 statically (if threads=16) for Search
+* No static allocations for Security and Observability
+| Maximum (e.g., 16)
+a| * 4096 for Search
+* No static allocations for Security and Observability
+|===
diff --git a/serverless/pages/ml-nlp-auto-scale.mdx b/serverless/pages/ml-nlp-auto-scale.mdx
deleted file mode 100644
index e127294f..00000000
--- a/serverless/pages/ml-nlp-auto-scale.mdx
+++ /dev/null
@@ -1,128 +0,0 @@
----
-slug: /serverless/general/ml-nlp-auto-scale
-title: Trained model autoscaling
-tags: ['serverless']
----
-
-You can enable autoscaling for each of your trained model deployments.
-Autoscaling allows Elasticsearch to automatically adjust the resources the model deployment can use based on the workload demand.
-
-There are two ways to enable autoscaling:
-
-- through APIs by enabling adaptive allocations
-- in Kibana by enabling adaptive resources
-
-
-Trained model autoscaling is available for both serverless and Cloud deployments. In serverless deployments, processing power is managed differently across Search, Observability, and Security projects, which impacts their costs and resource limits.
-
-Security and Observability projects are only charged for data ingestion and retention. They are not charged for processing power (VCU usage), which is used for more complex operations, like running advanced search models. For example, in Search projects, models such as ELSER require significant processing power to provide more accurate search results.
-
-## Enabling autoscaling through APIs - adaptive allocations
-
-Model allocations are independent units of work for NLP tasks.
-If you set a static number of allocations, they remain constant even when not all the available resources are fully used or when the load on the model requires more resources.
-Instead of setting the number of allocations manually, you can enable adaptive allocations to set the number of allocations based on the load on the process.
-This can help you to manage performance and cost more easily.
-(Refer to the [pricing calculator](https://cloud.elastic.co/pricing) to learn more about the possible costs.)
-
-When adaptive allocations are enabled, the number of allocations of the model is set automatically based on the current load.
-When the load is high, additional model allocations are automatically created as needed.
-When the load is low, a model allocation is automatically removed.
-You can explicitely set the minimum and maximum number of allocations; autoscaling will occur within these limits.
-
-<DocCallOut color="primary" title="Note">
-  If you set the minimum number of allocations to 1, you will be charged even if the system is not using those resources.
-</DocCallOut>
-
-You can enable adaptive allocations by using:
-
-- the create inference endpoint API for [ELSER](https://www.elastic.co/guide/en/elasticsearch/reference/master/infer-service-elser.html), [E5 and models uploaded through Eland](https://www.elastic.co/guide/en/elasticsearch/reference/master/infer-service-elasticsearch.html) that are used as inference services.
-- the [start trained model deployment](https://www.elastic.co/guide/en/elasticsearch/reference/master/start-trained-model-deployment.html) or [update trained model deployment](https://www.elastic.co/guide/en/elasticsearch/reference/master/update-trained-model-deployment.html) APIs for trained models that are deployed on machine learning nodes.
-
-If the new allocations fit on the current machine learning nodes, they are immediately started.
-If more resource capacity is needed for creating new model allocations, then your machine learning node will be scaled up if machine learning autoscaling is enabled to provide enough resources for the new allocation.
-The number of model allocations can be scaled down to 0.
-They cannot be scaled up to more than 32 allocations, unless you explicitly set the maximum number of allocations to more.
-Adaptive allocations must be set up independently for each deployment and [inference endpoint](https://www.elastic.co/guide/en/elasticsearch/reference/master/put-inference-api.html).
-
-When you create inference endpoints on Serverless using Kibana, adaptive allocations are automatically turned on, and there is no option to disable them.
-
-### Optimizing for typical use cases
-
-You can optimize your model deployment for typical use cases, such as search and ingest.
-When you optimize for ingest, the throughput will be higher, which increases the number of inference requests that can be performed in parallel.
-When you optimize for search, the latency will be lower during search processes.
-
-- If you want to optimize for ingest, set the number of threads to `1` (`"threads_per_allocation": 1`).
-- If you want to optimize for search, set the number of threads to greater than `1`.
-Increasing the number of threads will make the search processes more performant.
-
-## Enabling autoscaling in Kibana - adaptive resources
-
-You can enable adaptive resources for your models when starting or updating the model deployment.
-Adaptive resources make it possible for Elasticsearch to scale up or down the available resources based on the load on the process.
-This can help you to manage performance and cost more easily.
-When adaptive resources are enabled, the number of VCUs that the model deployment uses is set automatically based on the current load.
-When the load is high, the number of VCUs that the process can use is automatically increased.
-When the load is low, the number of VCUs that the process can use is automatically decreased.
-
-You can choose from three levels of resource usage for your trained model deployment; autoscaling will occur within the selected level's range.
-
-Refer to the tables in the auto-scaling-matrix section to find out the setings for the level you selected.
-
-<DocImage size="xxl" url="../images/ml-nlp-deployment.png" alt="ML model deployment with adaptive resources enabled." />
-
-Search projects are given access to more processing resources, while Security and Observability projects have lower limits. This difference is reflected in the UI configuration: Search projects have higher resource limits compared to Security and Observability projects to accommodate their more complex operations.
-
-On Serverless, adaptive allocations are automatically enabled for all project types.
-However, the "Adaptive resources" control is not displayed in Kibana for Observability and Security projects.
-
-## Model deployment resource matrix
-
-The used resources for trained model deployments depend on three factors:
-
-- your cluster environment (Serverless, Cloud, or on-premises)
-- the use case you optimize the model deployment for (ingest or search)
-- whether model autoscaling is enabled with adaptive allocations/resources to have dynamic resources, or disabled for static resources
-
-The following tables show you the number of allocations, threads, and VCUs available on Serverless when adaptive resources are enabled or disabled.
-
-### Deployments on serverless optimized for ingest
-
-In case of ingest-optimized deployments, we maximize the number of model allocations.
-
-### Adaptive Resources Enabled
-
-#### Ingest-Optimized Deployments
-
-| Level  | Allocations                                          | Threads | VCUs                                                |
-|--------|------------------------------------------------------|---------|-----------------------------------------------------|
-| Low    | 0 to 2 dynamically                                   | 1       | 0 to 16 dynamically                                 |
-| Medium | 1 to 32 dynamically                                  | 1       | 8 to 256 dynamically                                |
-| High   | - 1 to 512 for Search <br /> - 1 to 128 for Security and Observability | 1 | - 8 to 4096 for Search <br /> - 8 to 1024 for Security and Observability |
-
-#### Adaptive Resources Disabled (Search Only)
-
-| Level  | Allocations                                          | Threads | VCUs                                                |
-|--------|------------------------------------------------------|---------|-----------------------------------------------------|
-| Low    | Exactly 2                                            | 1       | 16                                                  |
-| Medium | Exactly 32                                           | 1       | 256                                                 |
-| High   | - 512 for Search <br /> - No static allocations for Security and Observability | 1 | - 4096 for Search <br /> - No static allocations for Security and Observability |
-
-### Deployments on Serverless Optimized for Search
-
-#### Adaptive Resources Enabled
-
-| Level  | Allocations                                          | Threads | VCUs                                                |
-|--------|------------------------------------------------------|---------|-----------------------------------------------------|
-| Low    | 0 to 1 dynamically                                   | Always 2       | 0 to 16 dynamically                                 |
-| Medium | 1 to 2 (if threads=16), dynamically                  | Maximum (e.g., 16)       | 8 to 256 dynamically                               |
-| High   | - 1 to 32 (if threads=16), dynamically <br /> - 1 to 128 for Security and Observability | Maximum (e.g., 16) | - 8 to 4096 for Search <br /> - 8 to 1024 for Security and Observability |
-
-#### Adaptive Resources Disabled
-
-| Level  | Allocations                                          | Threads | VCUs                                                |
-|--------|------------------------------------------------------|---------|-----------------------------------------------------|
-| Low    | 1 statically                                         | Always 2               | 16                                                  |
-| Medium | 2 statically (if threads=16)                         | Maximum (e.g., 16)     | 256                                                 |
-| High   | - 32 statically (if threads=16) for Search <br /> - No static allocations for Security and Observability | Maximum (e.g., 16) | - 4096 for Search <br /> - No static allocations for Security and Observability |

From aaf7ae9b7df3708a2c1b39b47468cc72da46156f Mon Sep 17 00:00:00 2001
From: kosabogi <boglarka.kosa@elastic.co>
Date: Fri, 8 Nov 2024 13:32:05 +0100
Subject: [PATCH 19/19] Updates table

---
 serverless/pages/ml-nlp-auto-scale.asciidoc | 55 ++++++++++-----------
 1 file changed, 26 insertions(+), 29 deletions(-)

diff --git a/serverless/pages/ml-nlp-auto-scale.asciidoc b/serverless/pages/ml-nlp-auto-scale.asciidoc
index 347bf1fc..33fc2e0f 100644
--- a/serverless/pages/ml-nlp-auto-scale.asciidoc
+++ b/serverless/pages/ml-nlp-auto-scale.asciidoc
@@ -103,11 +103,7 @@ In case of ingest-optimized deployments, we maximize the number of model allocat
 
 [discrete]
 [[adaptive-resources-enabled]]
-=== Adaptive Resources Enabled
-
-[discrete]
-[[ingest-optimized-deployments]]
-==== Ingest-Optimized Deployments
+==== Adaptive resources enabled
 
 |===
 | Level | Allocations | Threads | VCUs
@@ -123,16 +119,17 @@ In case of ingest-optimized deployments, we maximize the number of model allocat
 | 8 to 256 dynamically
 
 | High
-a| * 1 to 512 for Search
-* 1 to 128 for Security and Observability
+a| 1 to 512 for Search +
+1 to 128 for Security and Observability
 | 1
-a| * 8 to 4096 for Search
-* 8 to 1024 for Security and Observability
+a| 8 to 4096 for Search +
+8 to 1024 for Security and Observability
 |===
 
+
 [discrete]
 [[adaptive-resources-disabled-search-only]]
-==== Adaptive Resources Disabled (Search Only)
+==== Adaptive resources disabled (Search only)
 
 |===
 | Level | Allocations | Threads | VCUs
@@ -148,20 +145,20 @@ a| * 8 to 4096 for Search
 | 256
 
 | High
-a| * 512 for Search
-* No static allocations for Security and Observability
+a| 512 for Search +
+No static allocations for Security and Observability
 | 1
-a| * 4096 for Search
-* No static allocations for Security and Observability |
+a| 4096 for Search +
+No static allocations for Security and Observability
 |===
 
 [discrete]
 [[deployments-on-serverless-optimized-for-search]]
-=== Deployments on Serverless Optimized for Search
+=== Deployments on serverless optimized for search
 
 [discrete]
 [[adaptive-resources-enabled-for-search]]
-==== Adaptive Resources Enabled
+==== Adaptive resources enabled
 
 |===
 | Level | Allocations | Threads | VCUs
@@ -173,20 +170,20 @@ a| * 4096 for Search
 
 | Medium
 | 1 to 2 (if threads=16), dynamically
-| Maximum (e.g., 16)
+| Maximum (for example, 16)
 | 8 to 256 dynamically
 
 | High
-a| * 1 to 32 (if threads=16), dynamically
-* 1 to 128 for Security and Observability
-| Maximum (e.g., 16)
-a| * 8 to 4096 for Search
-* 8 to 1024 for Security and Observability
+a| 1 to 32 (if threads=16), dynamically +
+1 to 128 for Security and Observability
+| Maximum (for example, 16)
+a| 8 to 4096 for Search +
+8 to 1024 for Security and Observability
 |===
 
 [discrete]
 [[adaptive-resources-disabled-for-search]]
-==== Adaptive Resources Disabled
+==== Adaptive resources disabled
 
 |===
 | Level | Allocations | Threads | VCUs
@@ -198,13 +195,13 @@ a| * 8 to 4096 for Search
 
 | Medium
 | 2 statically (if threads=16)
-| Maximum (e.g., 16)
+| Maximum (for example, 16)
 | 256
 
 | High
-a| * 32 statically (if threads=16) for Search
-* No static allocations for Security and Observability
-| Maximum (e.g., 16)
-a| * 4096 for Search
-* No static allocations for Security and Observability
+a| 32 statically (if threads=16) for Search +
+No static allocations for Security and Observability
+| Maximum (for example, 16)
+a| 4096 for Search +
+No static allocations for Security and Observability
 |===