From 9e8fb924d0196970a175d0116a682de4f8491757 Mon Sep 17 00:00:00 2001
From: yuhanshi <shiyuhan1@huawei.com>
Date: Wed, 26 Oct 2022 11:02:26 +0800
Subject: [PATCH] docs: add guidelines of neural network runtime

Signed-off-by: yuhanshi <shiyuhan1@huawei.com>
---
 zh-cn/application-dev/napi/Readme-CN.md       |   1 +
 zh-cn/application-dev/napi/figures/02.png     | Bin 0 -> 43311 bytes
 .../napi/neural-network-runtime-guidelines.md | 491 ++++++++++++++++++
 3 files changed, 492 insertions(+)
 create mode 100644 zh-cn/application-dev/napi/figures/02.png
 create mode 100644 zh-cn/application-dev/napi/neural-network-runtime-guidelines.md

diff --git a/zh-cn/application-dev/napi/Readme-CN.md b/zh-cn/application-dev/napi/Readme-CN.md
index 99e16d5ce9..e6626f0373 100644
--- a/zh-cn/application-dev/napi/Readme-CN.md
+++ b/zh-cn/application-dev/napi/Readme-CN.md
@@ -5,3 +5,4 @@
 - [Rawfile开发指导](rawfile-guidelines.md)
 - [Window开发指导](native-window-guidelines.md)
 - [使用MindSpore Lite引擎进行模型推理](mindspore-lite-guidelines.md)
+- [Neural Network Runtime对接AI推理框架开发指导](neural-network-runtime-guidelines.md)
\ No newline at end of file
diff --git a/zh-cn/application-dev/napi/figures/02.png b/zh-cn/application-dev/napi/figures/02.png
new file mode 100644
index 0000000000000000000000000000000000000000..1e31c1ce7f4e93ed9bd2e8acda0facce7a9a00f7
GIT binary patch
literal 43311
zcmeFZcQl;g_CJas5|Kzj2tfo1(HYT=-g~c6W(;EtGwSFhQnd)86D38B-boR?hbTcr
zlu<(T?mm;q`JVGV=l<5ZcisQ)S&K2o^FHr>cHf`9Uk`P(l&MZLoF*b7qEb~+&?6!u
z9U&qj{!T#-K9Nq$-~m5~v3kmKL<L=pcp{>crtXR+?kI1#qce<%T}b}VJ9a^S1RCqk
zE~LONC}`vA%4_dv<AAY2VR>C(?%*TvcNE&*5$*`H|FcGrUr>-oK!Ar|)PP@*T?i@!
z0WTr~ytl=K%>Jymae%oTZz#gc4>q`A?C4^T_QHa<S_a@nfFFD;BnRGsKR|^4y!D0%
z-vpmPU0t1F#xPrTM{su~5g}1ta09S~x{86O2D_j<c<=0pgn=K*FgqlgaEUwugGPZ*
zlmz%8ykP6!_tdw6+h833c?(5*hkvXQ5EA9*6&C$NFU;M>;m_H$6){RYsyb$3de$f&
z5mjSH1$&P_Ydm2XtRvdxFAI1@cm@9a!QICd_UBW3n5Uy1AOkQ(5wQ1|kAObMW6r6o
zYv~H;W8mt-W@<1a16{O|GVaeNE;fMv-x6R~6lI6X6TSeOz&E?17&}yeT~6fBzM@7L
zJv%o|K_wl28y#a|6;*v52St0VFvQSHNYDo5!Y>8{CTFIh>!2)ZYU*LDY6pH-5`xIP
z$~l^8DQW7cx;tr#Vm&ntTul@`9APksu!pLHoR*O%R@A`V)=W>tRv(H%XnQ!=qvf4(
zCiVh25qB?DZ3UQuoT08J7G@?asHos2e;e%T4R=Gzsi5t#w_Wu-oX~>uSVbKd8#y^G
zV`nE_JvpSQuo=z+Cn&6HB;qV&Vy1|6H&uh^dLeDWwKyehJuPD=AP_ku{I-jSt%iZt
zZ8Kqvyf;SG$pfQg<KtuKq;BuXuj;KUr=_i|?dV~IaaR-7N5j?OV*F?y!X~ERTR_B4
zQCP#y$X3qXM@d0U5$j_IB&Gz%5-0>;o{FvtYFcJKPT)?$j;<muhA5noqq~WLBU(!k
zY7Dp4bTcv1fQmqc1q6-dajtMAMBhi$(ZEF!>#6AG3UybPzion%)5T%2ZW@jrVxnHo
zFszOOPFuuO4`+yixVZ?sxVfr3Dhrr-*?QT#!0k0v)x8~!46)8?Hr}4L5CJuHxVoM_
zzmTwoA<W}8+E_qEPC-;%!`)HC8Rl(t8->Q$BMnhltPM^H?c}bns-&Wa74iY5B<F!}
zfY~VBhO63ugLsLE+Pj#k0ZG}30j*JjyV)Xz;W}6wloP*>tv5u|Lq!;7V&|m7uVDt)
zH&S-B(b7YjC>weBAQ1-Mih6>!27m`j1obc#^md19cxd@x6m<;^-34`BQHn?vArCmh
zP5}<qC>cWdJvHpy`7wev9?D*Ddv$Fi2#R0X!&S^3E{L^t^3ZT|mJ<<mbLEFSY6;>r
z6v0KVHr~#1YA7A3vl4+^elOr01eJU|jXe<vWfwU;6%nk6vj@!Uwz~FhAu(e&h!E7-
z4PoaD@$nIMmpAhfazMKp=$fh#q$+3VZDS{=p^FiA$LQPZ3Fs(lJE{tJIPz<wbRn9?
zKB`bJWpyQjcs1qJlpI8K(GXW#w3wKwla~k-eH$gBh4xT{su`e75inIDb$5NRQr*<Z
zQ6HkLBgTJQM*%n`IUN`9Vy44yVqoJY05y>pb~7+=Q%1o(os1x6C@;<17<(^yZy{4V
zJ5f0kR})t_T2Ng8ivVYpv(rHu!0m-~bhSjq+|+=k$r%a2agJ(=-mb1#0TZ;4vI0s0
zYGx+rrEjd~rllz4Yyv}>8H+(&4deyXwZM93eh~#xd2M+N#Kgc5=4fMMtEJ$|udHIA
zrJ@8uc&ea1^i}lLoJ_UtQGj;vr7DU;xd}UAO|g1VeK8Ghb$(+-gsC3d#NJue1q~L+
zBSe&KoY2l_4-Z2RRjex9)JadrTSX9slk>9ES1}ZH5Kz9YAZF|8zz^5*G=`~YBDMLI
zTs$E@TKu+V_9i}H544a17Ah>~s3?!swUbxi*M&LCDeIyPU{F^qQbicz2vJrPgNmp^
zg)wS!4z8|Z3K%zc5hVo^jE$MMs(=?3r>x+n1CFPK)k8bV>5FK(>uVvj)YR2XggxX{
zeGK>=^wjM%-3-LEA!Z(K>N-xEV6ldPql=525)7hj;Nm6b=wYa>ENbh*@1!j%;;6!}
zuMY!mTFy~N%}t(PUc?bDpkd>Ka6+r87zuf*=-<|Wn81vE9Np~|_!S%s^bo2lFx%UP
z;0k>cA7J<VdY<y4u3}ndYW5xi>TWpTQ<V&X1e9R5u5b-3?zWN`VX?EEA>_8bvYwNW
zE<_KjLa-FHfT*o8oZpz>qZDovyt5E+%1VFT`fnc(em{<)LW*m;1Cm5UEJUgbPy=tP
z#j#Vhv{Oe%zqI%2Z}8l`pr&T3uW$|Nsd|BCY43&qgy_=a;?f%r?Pt|h&KF-AmTRfc
zR=Br$G$j;(J=_wTwVD~9sfW*a$Xu1G$9G^XySqIKurmjBaE|WD5BJYeGLurv5)uE~
zi;5b{IjFf_z~U1k5(<`}|Me2YF5B}-_CH-G%R)v%in=NLgZkgE0-I2m3XuQj`vnt0
z%}-N{k^YCf5rG|_{5y$++v~B&_Qa7WoDKe81pfIIdd2QP98H!e_!1x_n<!fT-wsB&
z4jk?Be~<Rha`>;*{+SyDrTMQy{%?Z=w#LFkGM1CC+U_QHgCyX>^MM698IR@?f1CPM
zXFQ_vF-eCpuPv7A*ww6*5<!z*wmY~?Rk?~(KuS9piFqwLUy-fF(tD)Sz$y8y6D*m{
z?jPS~1{iFDKBaZp$9mVL@$Y`t8;`qKqD}`Vv|Cl2{csP-B^)C|zGk=@#`vxQ+~+W#
zM@s&W=?AY9Y`s}zMS1$Y$cj8s&<zrc%0v;#=Y<__PUZ`!2ko@w;d&U3sp7W|O7Glf
z@F~7$QHd*d79L>-R9S|2dNa&$-`8=bRUw|qm5oc$V*BF}WiJ3tQ2q86X%CjsGf^ne
zetSQV*Za;7W+a6OHwBcy2SNvry-d&!SUmHkC5Gq6`+*o(W-sq4*U9+{r62nFt;O!z
zmdKvww<arIbZ0nrbx)~T5b7k6mTZ)+lWlXdEX7201@Q2*Gu$QfP-Mr2JN1IOS4+u{
z7kALHAcPu}Qk_}p3Q})f3HMYDqBM?AXG#s*X=A;VM5Y(C!xaU4*?;VnqwWHCd%=5=
z!4xc0I!O{uR_Q8&bO%f~JZEMXN3<4ex{)q&y!cK%5Y>*zsy$_u>xTU>qEsEQu$pvf
zMYzNGY{t;7C~(f;TM{Ws#~i;-N-^@Z(e6tNSSBb={oq#Py<F4G&_9eWA^pR7=&@&B
zBcmAUD2(PbZ@HJt*{+mnc>%DH9dZAqO+AL92_AguPW@^7VZ89M1zu+b4xjXDjS}E=
zY;UkEf$^OTOn(^toW1ck=c$etKj)L}fd_<34$JXcNSQ(ymYxC@@<Jsfsn%lUZ$K!s
zh}H~Q+O=Mu|KsjiP7#NcUtbv#A#nQ1S<-8O@lJzte;EDg7V#g>ABV3-KjIL`8uCXg
zEoO4QwYV1%Pg?2facX4J@p&Pm+*?x?vgd^)rk>*A#~Q2()Q~;6xq!fFm^%wKf$=3u
zC;u=Se(DeBL;tEfB~W*3Htq5#N@MM~aHX2-fQ9NkXSF~0O;bPWB)Qq>N7GW=?|7`W
zDsn;TJf|c*!vUwcGfxB)7@sI2{$CaUuZsVFrQ*BR!_q@5<+W>rX0Oc~GLk)K#8UHM
z*TN5Hm!(#}mqfOArATD8dWfo5t<qc+dxGAY1dgyz(3%WgFYf&<$a!QtBG1qM?f(9h
z7p?44|3X`Mr+KSDGZUZB>JI^A6ud~J!<hc#?B`>b^Gy|4mv&s|z7lqF*E&4as!HSM
zJf`NW)GV}W<`eA<9d9e`RdIKf!9ridoF`XqG5fEiCR>!0cT{fKh#lUYSG_uHU4xx;
zGLlLt3=y$j8miaicrwSWU0{-kWQpUlR8rD-&Tecq`lM*-$Q{^8Vj<!`IgJDCq`1TU
z<pcr8ySD;&MZ_|HR@pR9fxLOTKT;;Ds{d1)dbdnM85Pug8*L37QR3k9!Vn>Aywj*^
zCV7ke<Db#$1+`k$n^(C8v<hz4Y%M-*zG}u&@-ia2FT1@Z@8s23GmkD|WIFs6;rJo5
zrGJla3Xb0$*R>dion%O+G+x3NE3VnEJ`XN-*y61!7)D|CQY>L3cQcBYDy&U<X=lu1
zfj61qo5}swVT`m2u)=xN8qSL%u4i&guqnsoMu}YB@a&_HQEscHpURfsBJL|WQfZCn
zOEV1>IgxKu-$o?<QsLMkYXjfoQ|&hcNu~Vq&7#st+hbsK-mqIO4QiKd6{@uQB`r`T
zt*hdEsp2;6G!Syt;~lQsYXhU`f>__2$Qa(b`z^y8UELzo6P8ESUvZXfRr_zQz!pFI
zW-GMn!Knt%l!onaotR6$a2&ZLd1QNL@T&=<-n<r>)|H2Er@VH)muUZZKJ})hH#Q<B
zEnwToYIj0^wadIv_1<Y^^(pRx*ph`bJ~B$$HlrVf<<4;_5}qdA(%gEb&o}s&J&;DN
zGujZ<PJQioz<n#NJR|wlna(6sga6)U+e>EuI|rjTU#7+_RPHY07AGnW;A2W23JGi_
zhkFUl%yG)`<9uu!9dQ@m3=~D^zS5a=ZXVtpnAO7%ahW!c?KFFal;4<0CHPk_0c$cO
z(C(G=fl1fg0Z9E?j&Q>aNmbFWSV5mI+na{-6eZ+Ha<dP9vBd6L5071YDlH3Q&3&f>
zB}rG%%V)9lxT>{;PW^Cy2lg?@Wp%cTY54TW8%r;|lxr!?t6^qKvCZ$;$o5Ldt#*Cx
zxVG<m@EHbDCf_s9w4})GN4w9~yHDN3#;f4j+Ga-hW=n<^kIss=b)nrIKTgeUlA?MW
zL_vCm{ApW%dee_aoWWRX>#lp<8b?)`2PZv0rg$Tei}c`wAI<uC*`vQv&<aFB@1~*g
zXb=UvcrCOg))<F(n$J37#yT-$nmpMizb>q9UP0Q={Lp=o`*D3RJIU;8G!LRC@aLu1
zhI?mrCTq}H9`~<eh7CncxLc*iOQAoLq)gWbU)6*?iII@ljPu)BGvDL5P<l}3P%)=g
zNw$__-hHe3XmU1d)6v1Sxjja-@YxMrt}kt3H5sHRQd9)TX?e)+6Tr@W7^^vasvGLt
z5);WAJ$x=NN<GN}-9pjnEJQIPxX?X(e27M0kZX<SCUBZ{+{-PYG^QF<(oEzwww$n7
z{m~j@XzpuxFAV*mxx1L;_Q6nY2N#pRDMy#t4UYxuD3qjM&4xODV(RS9#!KdsIuX9;
z^iJubn2LxQ#%Ba_QlGy%ER60jewSKUy0ts=<K%K0yWRcG@^K|`+^Upbj<Xma)l_L&
zY&s9Bl=1Shmiuu5E$@F1uQs=+6t8g|psaT5|JlC5m_8A((~A3)P@s~8>+Z{=l$Kbp
zu0x69MV;k_9P)VrSFGp<3rp7Ng+Jl2sR7`O2Taii`#}8+&&}lC={9C4#;bh~mKrsq
z#`zx5dvuHZpMB~4;(No;qSOQB8Qcx8EE*fn9?z}hlhcV;%<nQ$Fgtsjk9_s3vhe@F
zBZ&BX>83>`1vJTKLXdDf>fg89y^we~Hqru|{rdub2(q+#0wl&;onmaq8JD*?$e%|y
zgvvL5A9JY}eb*f?b={cBjtteF7QplU5v6~*dE{n=N>gS;D-L;;d!%F6h!87qAo=yG
z@2lWBH|O?pgM2o*yIdIg3|;^nV?xPn$6v>&2^=GOuyfl8JDE66`Nfx3hLz6R(__eO
zW#=WcYDIacf_3Q~nThc9&O9_d@W^jYt{L*r=A!GzuVs>Z#CssEie1Xk{cnwjMt^lO
z<o~Py`LRE?>%p;gjuWiY^WX_eimm`(=gM?@)y1(t!(nK*;q4}DVNmy=x}fKVPWqj}
z{7hfZ<)=5dZqfKB$j!}udlxF`|J^k~1{?Cjf~UEiTJH$;=^k=kGe+#%n&HY!F7Pv4
zA9g*T##?)vxuC{caWk;!7fny;mFmahciJklv+a&g7{vk1Pxu}AL6Jpe&enTF<C5?M
zLF=}wrqZK2qZQ~v4F-`$oUhd!_Uqx!ESg>@D_-*GbLMJe@hoP{b9bRzMs1=X-~QkV
z@)|j+*Qj@zak<t*c=rq;1sA%$W>_?n8+bTw*BNt;r&@Wvv+IPC#tc=aZasU#|7{TL
z_nmyByxf!D;Jzpz`-q0mCS~zk@QKy!%yVs(HvT_8i#xRa7=E1nUN<(c5O2Og+nL5Y
zrMw<Xn=hd@!8%p4*>C6)(|qO`?D350=!`}0=8sJEEbA>u-`i$ce6LsK2FGU;<`76o
zxryD!tU@7cGKc!pRKpC`0pU*){URAn3qn*s^j4Nrm(m+wV7PlMr!Wwt=&e`^A(50{
z)~zZ%mnAHWq-$HmRmZSb)S5CQqeZQKhjHl@<Fw*#gL1FUk3yV3D5oiB<qo<hO=-V!
zk(xs7EjP4EfS|?JZXAW}s?+(IB6ZWa2i{;YS)widOetPC+KKbaQ*D_5Yw{<R=2~w8
z)P`8qdX#!*vI!CI@|xkm#oYGyJN<bB&bwmMpZi}F*e^&S@;<-u4~q}On2o6Q#L*g^
zeRNE)G(oZ4+=PZjcr6GF&FxzEkpqc9TzqPvVD9I53!fDCB`E`^&u=c4?<dKOU+?EU
z8)IyZ<6T(zB2uG?qq@wV7O)=d?&RTLkgjR`p{^bt`-G{_T-d!|`*k18{zsM=W?Dq_
zd@1`%I8ZzigyYwH1S$s<<E>!%+m?sl>LQf%ODiNK#DlbfGhJ0>#giW!Z6ycrO$;7I
zS|9Hraq}r_O{5GVOXC~yjFR56tJCd#P3q5{N&2iRNcqf3gP{BLMbje*r@ZmRYTs^N
zF~7N}OT2mpr|EdVBV-_LJilwbkR2|mjXU*fk-4MFuzHMn@*3-(f;s35DEDfupz@>L
z;9jR=TCT|U_g{!ed;(%k>s)q#^-S*Pr1|p+g?&@=o4N^wUukFk{+hlAF#ST~o}Nu$
z`l;_-Cljw4S46hm%N_mR+uiIJn+iNgcJsherl55dPuI@hGHn3_0;|IGjHTK4B2xDz
z!ApS)Gsu=Inof-fvUd5!p4#}F7PG^(k;8cY*jMipGLq$8c%Bz23ZFr^?*HK>)@!S4
zt4$8f?b*gwe1>mQpQXmQel7*oWTNNyN-a>k(Zx@NCloU3)7P=C#r95TxE-JIPCBTx
zgB5?t$5gs*$kX8Et3gqz8SiY{469e`2(Jra1s6XN?^pj@_@#g{e*Tu2-UGnDGbjIB
zu@W@;zo5+j|3H}#ZrAM<o1KNih95l{@_H{6PPL}V1irugTGHr!_#I_nE{)B=5lHj8
zaMS?z$gv?-fhT|@hOGYrr$w8sb;}&JfNglSr8b`V?M%lH!eMRCqH_!D&3J1T1#g;{
zJAaJGvu>~&EU@IJx&nEI+n85eTKPFv1H<=auP!$nsjba)X4^JBayv0RTeCTWoc<A`
z?ddK(_6pV>r&sIkhTH>nNNZaxH*B^m1+$P}q1O78J@cwVdmPW|_Ee;LzFAq8Y7B?z
z=*YoMlcMymn06k$ES2-k2g9X~F|{l2DI>lzG7GnbnN3;M;U4Zvt+ue~)!{boo=a-8
zWB*FwNvPxeZUGNA^&MV}w5WvZrD2nJ32vFj_BUjBjATWZh*S&W^DD-mrb_v?)hyRv
zDSyjQ8O|W0Hd2n_zIU3Yt*lf2oRs}N6NYq01kF;3|DGo@$vR+%eX^)oI*lB)HrpM4
znube3GTUEzOm?f}guKL!z5`B;=Pdra>kq5nFNBMZN0C!8n4Fcrf3{OR&3|8X!cRZ$
zmg%(-4$d^0!)<wS_c<1BqdeCiGe4fcEmwuzj%%}PauT&x21f|FA=^Xy@l8#S0{~!U
zXD+4E5TMmQfqzCQDNx~I&eg@Q)#G`K1CIh62j0Y%BEM*H85i6lJAEn2I`Cj%cNgUQ
zu;vIRL6@($@nbb!1w}mi(4<XvBbAx!i2g%ftJ)zmQ(F8*ZiMfzCc13hOeLcsqt|-b
zuM~~9K#E{JQh`RJX0NAfG~H19O<?v5G~)>nDOQqzub$v?QV9gmX%@Bc2NePo$wm3k
zh+bsRpx^f7{lTRqnf;Z}juZ)Fvr+`B573Z-Z$NoWY`(rs3*Z93mbgTB#Bgd4dM2D1
zSvR$dl-hi5fx~CpWTcRys_pWkI#hddn(3u*rDrblv&_#|b|#6Mhv61(K$EH_m`l$9
za$2%LDB#%e_jBR3@VBt2jw=NrLG@zJI#Gx5&FLR^q%Q&-vn1Vj>t}4h!ES3uf)LX^
z`COy?r#YHQYD*(!`7|bmZwy+!dj|7Ng?#pwD)MJ*@oL2}TH-h9JoC~~8}kF0`PVPa
zx~&&R%5pUlg>Nom2W`nvxP@2JX_`lf74LDM*?T;-)#t8Cn*6NWvO<nd1nh;=px^<I
zdkCOEOC~Tt`suCZR)8X-QFzl;V1OLx=WT2tZtKO^?PvlhzT;8AYAYwEUsKeSkJuZq
zZW@?G<VcT{@1_a#^9#kt7&7?;v)&w3uR|UFbTOzy%vnv}+BOw$uo?B_@R2{4_OISh
zmalN--dqoQ8v+@d3A^GFRr9l2ows2hvoP$Y%;=`Tt1nUXIreTf1A6}R>+_)ye~CFD
zX+B{2wKO+UF>N>dD&6eFj`r6w>)oNq?7cZ=fBfB_nS5nEYfU)%ke@a1P6p%zy5Yd6
zOMKzh0_^f^GFGd6AmXdI)6AAgz8n$p5deF4UdAbHfOdt*8(m>?;t!<ojq08vTCx%`
zP&wvF)J13mL|(ek9gVtg$51<m?OM-Icq_dr&ETY&AgBn^Dd(l9uR*3aXz4Q@JE%^!
zzLYX4C&&p+lD_ZqrNDCVk(4D<h3ll;YoM`^vxLvRP~Sg36!hJ6Mm8nZOI};2HK!um
zn5aM8MVi+8Rc(ZQvH1{oj|v#~D4gs8CGD*iL)tn`&reV2Z6+$K-KLcA<8?JOXTk#3
zsWR=FBWV2NzDg|DUffi?#2`LjKWX)-H{V<o8A!FTF5g4DP;;-4&;w~b1yJqT=JE3&
zPk|v=sLg{YzYHn4IKT6o4ilVRsU|cMXg)hUbZp6-d`51lCC%^=(SrS1H4FC*b>IN3
z4}E=*C$<zyZi4g8J#b?vHOw{QSs){;m<XupNEB%|TXG0YCPmGsSA>*1_J1-{&EY0n
zYWxYxt-*PbI!j~YZpU;Hc*pZ1Xi}K7Sqx{Yw_GXDu3zm%t=rIb&FpGDt)sd5^5o!e
zaXjWf%12#Fj2aw+aLr6!CTRPCfycXN>OB?)P5t|V3_d7QO)N&DUhy^d1Zk(SzN{z@
z*cdP#%x~(J-g7QDe{yOh*{`{X01`%30!#bux%e>1fpqam*=XWf(@9`}{H8#Ef6Vtl
zE|M-rDV#|19{1kanN4f&6rFG$yYM=ruUDBF<<We0<U8IxMcm_k!5a+bN%FhYLzfpB
zM!-Skk6tgco^FdRIXf+k=<iWQCw*qokZ1ZhE%<j9;AS0_;Y3j)pSASDZLQF9L0Fr~
zRR-^S(rZg~TjM209Vu>~=uMYOK6LdkAgtG`e1OrM@s(H%K5;T{;4VYh2EM=93}*C7
zhqadl{i7w?=rZk3&Abf%c1QFCZ7pWHttgiyaQ~Nbu1xhcbkidmy%M;lQK=aHD`!Gm
z;&0>S+BX{D=j!xsjsQff!x;4CUr=7ZQaat8Hen{_`TZd$oN`1<=3qT9h+B_xb9n+?
zhguu5i_|pj$(cP(^;9l2$2$Zf6;XfQz&OkCRjKzcTgRay8w;rUa4~GUu;K7R-GH8P
z7XFIv!a%OW!InB-ojq9+ouzxUM~5qP6xgV9zHPs2``I##S=gvj+RGet_&_61pU~Fg
z>?$7hK7)J2T_h<isLW~L>B+0!EMH2dh4=eq(Q`c+rra#1y`OFo4wGdJ8YXr1j7-FA
zKygK0oDn7=CS{}meK=X$7+Ffh<;;xM?O#(<X1(x}f<Prj<_C3ao=^C#g_k-H_u<GB
znGj7^O7U;khaIa!7D6a_i(EJdX3NDadvl;k0gE)w3|1nD9$E6UjL#RA3MNC!3C^C6
zw<p#IJJZpM<1m{GqRLMrPpxk2sn0h)D(FUYgTPQAP|%&xKdMs}kpWFQw6yL@uSj%i
z_~5@zcB<Oe<?}1u!68-z<@zPL+4HiEWS|<NixMLvNdTQNk$#$Q`xGqS41fy0zBEc^
zdS|56aWFp`40u!=-jyzudD`BuEyMH0$~T|p)7>!Hga9RTgzr)XZhllS{+6k&RHR1Q
zR1P!hCDUC+%17&)8Y1+fF3D#Ygj-HOU$1gtcs>k`@$XWBTFnU!yN@b`Gng*DHh8UP
zg#AE<@=Sl(jp%2<WBcCwV}35!H>mhW%nn+Il(Td*tTPdle!awk!0IPqS)*8i;s}v;
zJD!(iaRGhQ%GM#)AQFZS3%4ek!M4|~6_hjAPx%8E{@I$j937}}dPFYGCZ}6vVWOd7
zBc~gv&6tt(>2#Yf^8>m0q7UZt3x1KJaA)>-G7IZA6KRSaRp@sGco6+Mx*xtEnEkI>
z{QBNx+DqFg7Rt*O6jTGE&rci{LMnKz@g7oXNbDxiX_iL6SR#fRQZ3nuJ3g~1UTy7s
z4U7=dwo|qZ6wrw`J0Nva8>{vRKQn>}v#(}Y0`^>C<@C*6F=^~=8S;x7h@W}p%H|a=
zJiwwcwJ$oQ!^S_of$#2%Q@sYUv3yNenF%&9-NF{xAp((4d;1dg+1z4j6t|&x5?6x+
zSq5h5C0b};wi167HZJ}KnlAmAcaW>P=LBW}@h#_!I2gX+1xQELnJzqa%UuxA6Le;i
zfd_g38kg6c7B4Cg`_~)qZ7xke)7Of=M@gp!(zo}5ck2->>zeVKpE&XnC8PE-Y?{^=
z($m^`Yi|8a=UJyp-*ihE7oWG6yl37^yG!2Vn)_|N(sM#f9Dm(s<=sg|`6`*C{eH8m
zm)taE1q}7m`$H82^Fo_)q2)dC+;jEd++AhM<8{8-vcV)zMbTCZNK%v*R|kYpdVL36
zemMG;lO**#NY{k2Ut~~_T7f_Z8t+8+Niv8>h8@ySQZX764;=5`#9j7(7c{Q-3SIo!
zoO!l8jgN-Qpv9B?Vkm^F)g2a3{_r6=sygj}T$ieJJ?3Y+W_N~7g0yjgJHBtQWTi`L
z>ua9F9R73_$h{vI$R&xQQwh1UsdY(c|L(?uIiJF9AisP_?!7SV?u@?wU~&PDl4?Zm
zL=%T*VhhNB6hGYF%S6i##NT+vXKlrym5iv1eU{!iDN3J2Kr2q22BD3j*@eQJ&`Res
zd|ZBw+8L<roiAzQ&7A4z7XTh9BE+bXx1a{ltHQt*mKg-`qqjD&1BS*ArIR#`b*J~A
zUoLgLA|Luu!Kifj;A^c<Nht_pr(b`i7!ls?yHf%J?Td}7?eCBCXFvGw;$N<|aTm5k
zFiF^dd~m+6(7M5yPrc{RiAThqTR-PXSz3XSMTJXBnKh^!^5f;_CgelQlZ~ni$z4Q{
zgO@DNzp5CYKhk>64|P|x1097^Bj5KRc1S|*<pMxOU~k04%Y>POy2M^v(hx3yC46q{
zUsfPH%Ork%H;<31FV%l@*!jK~$Y=5#Vnk3Q29IdCX*VzBkfTOSB(C=l*!+^t?ELAB
zSQ)55ThoiXBjV(<Dk8*RD^{8-%XhziH$L>%u4F*L`0Clq{IUgAGYs4-Z9Bc@lEwuV
zj$|RbwW0?I*PM(LU`y9@<;#<yNiHw<qz|{J1OxWHcjhu9i_iLl9J8%4_{22G+gfm4
zjNBv12G3Cd%nAMokfRF~cG{$pY@nENP*apYPf-i(;M;v7B2@}R(CXGipt*7n%S-##
zr=}Gg9Bw9_nmV%b(w3&L(E0=BxDED`85NsK;(0EGxs$>QJi0ILUb<z{<3{Pu>_3xq
zh?bmn_alM)?t(qI32Kj+2UOH6^}y!F3t6RPS#E$}2YoGk@V8Ayx(^o3bXWIrlNJM>
zwgrJ1f<h2Me6saQF9sixE}Gx4^O-C8)CH0#lcZlBKnkF$kpvS7Gdmahh)Hb$9K<mC
zfCp5jXTW5E(7Y@+wJTRTpKZ&IbEWIvyt!_-<eJ9I`!I>)E4jgy!+pK3q{-mQ`LB<7
zfEz6VL?vv@E*`^2|CEqXc*_B;)qeI+=~_^b%3$c|2J3gfV@!<D=vzqy)yo$yCExQZ
zvLF@`3wuI|f}a5TVRutvuKfY)2!T=s{82HAk(@Y&Fd<SUUP?twDhEazE}#3Kds@*_
z*!RGXE~J&XM!V?S>>XvT>~z0-JjJ^He&F%aIj{3wTj=f4`B!{vv0oajpBG||%v!sx
zmk;+hcDLA;&UI~N#`7T@B|vI*2T*0MqPVRXXKgRv06h@CTIovF#n(n?yI3{w!|MTX
z+23tl7%0<sc*jqs_~}roy88%-=vH)qgqe*Oy#b`(X};0D0{yG2#H4~?9whdrPuF82
z5-81`aVn>2yeK%6IyjSMF(VyuDCss?Te@w}B~t1Kz#5u}2tz4f!Ag%WNhPd7K}SI4
z_O0!34`c_i)7Qo#PMjyu_=t$kFUKIn6SzAa(&t{GdXBpw6><TYa20!}4|PjAQG+au
zI!&J~NY(<#$Ib1n03mBEC0cXoyUR#mehUCxKLyFM5RvSY&;(uUS0wej*624KBiBa-
zB=dx5t>CwM%Qmu+1XGS|L%&Bys|x9JHCAfh<UA=)6f6uTg|g4b%O6i65#(eE&b0I~
znJBn}05k3|^pKEJyK;a{J09l$kJvjvEEX!^Ej=Cl>@--Wr*$@rh{O?$*r>7b-aj7S
z0raWV05Oa_6eA`{paaRpnn*esNG>!$a*=S+{20jvH+>E)01mfA(2(2&#koj!G9xHN
z`)&Z9D_NZVJNE)k?g^Yv=|wCFsURN+y)#HzN}%)!pfnKTEC<EXAN{80*Jp8sCMCeq
z4GHNPfc1AV?tqjOVB!GbyZ~7EO(DqV%fkz!gn6niAev<Jx9lVU7ofY*9TySCMtvK2
z=f>FNvw-noWY4{XEQqusiAd@%0_5FN6;Dl4MYc`$NEeKXTAjaBzp9pv;3T2gRs?(U
z#bz=87A!RBW^6hWMV4-mb$Z`}_6%TR*+7JZ`k4^jG>B)$o+~&Ltm7k?fTg9QyTbNO
zlQ<xF6+VU#A%PNRm23y_LL?OSdO<#-F}ulEh)MSWkTKKzrjL^pL$N}6NgiaHwFGI$
zs`Y9CX+H+iw&qfm0W7%`+(?GnXAywS^_~?ayS(M%gyjVsPy+FjUTRbEADn6dqjTmt
z+>JZuU7^6pe`kRdSqfko`xC~GuwdVxxS?qwj5sNV(!F|**5tSR&wyz%p>zsi;$U}z
zhPc>70N%U?Q)Ftf*PGq+LEBCf)kUKklpg+;FiiU#%vr&em=I(njs)4}*D^AKX<LG9
z<uuMtCk8Q<id1DrBWHgoPz;fD^LP*Ja)Sv#a5idwd9c?L@t)J9eOy4?W<FxKfqv=$
z>h3*fY!d-#>z7wzx|hMvFfxKU#Q{h3d)O>X5lFqX2%`he915h)pd)5K0>s>>)p6<t
zD5wP_2?r@5C7~-#XwpVfyDEdjdnJoef<%rF5LU~o>`EBl1MR%+BeJJdBhS6)H3?#K
z0-NY9wDwZ->jAkwxkVW+M-<Et<a&mZ4+MFiw@!zSLeQi@Csl1A1!j=K9KK?!lVy2B
z0M)61j2;19HRe*3kOv!GK9>NDs69)1mS8U!kevx|QBwqgrNlP`EfpF$!M@Tr?HjA1
zJ+wt3R|YLZARkW4VL=RD0UHN1lMs;tp?zU7NRYnv?Hus@fu98sbu0)<v!>2@zGGGZ
z4JXVl2NO|%Ao0%*f;{gEDT<i8(q@buA(Y1=3<^&YKI+Q!aO-a;WM)pBYOFl>dsr?N
zsN$;qFv$cWq`dK5@DmbHqQOdkB?BQ+uVes@pL?tpPu+48%!?6(_E#ahbbwes$#Z~n
z3cl{H0?qW`uYl?&r%6dzfGmj~Y0#Ztc?h_neVV?=g}Rh1?^~Ehe0pb5NDm_^H3-L0
zb3!<VQUjHtFwBXh*p>%QgW2PZ86U++h>=^#u6hp~j7Ad5=-*@3zy(@VlBc)0pC?WO
z!DlBcig<w>YTkI-#l-8>$Qf>mP7szF0h7rBD2S*@K>-n}1KWKNBm<@h_jQ}oH$u{R
z5HzB>4Wy{Qla|0^-Vg(}2uOHf__CYATOW{+<UxtcJv9;{O(HmYWFHEd8S5B~ECFS~
z5tvonH`d49mE{F?Y^%c(6l4tct_AfiJ!si~@;=7Z#6aL+i2~j2aNS^8B2Y9^$wuR^
zkw_Cfi7M-fL8_4QAeyg5KQENh8&M~Mh!hM4GAZ||iC8EIC(LMDKTjM;IALZ4bb$<N
z-oxc$6nJWcFoQ}6IzX>ZP>{+3#UQ$rz<cf6MW7gImjv&OflW-_Q7!O-K0H~57zGQp
zH*ltn=ZJ}Fd5f^KTD>v=w0k&2!%0^_s7{~pT8`{~AfOQ{_kjh88;IqcA_0(M{m5>1
zp4maj=K@pZu3WGB1;z&S1jmHmR(LW7qj}RT>(#<+q{sVZFlys@Sf#He{W=WfM*M>_
zsr%y4B*IuGs5qMm!6=#ZtaO2><=X4UdvsPenwbM@&oE0%I)9@@eML4xFh5j=t37h_
zOiMn!lyj#fCe;A^5zM|&$`(tKaC+n{v-GW!+3u(s2hC^g4A!hM%Gxn{?CROYNZrVd
zEt$afB7so*rkF^ThSU?;>NP(LQbt=J28?dzt}_f-4%TfXcGUZ(TKG4WX?d=Fg|w2R
zQWyBC2xfSOF#TF9D}0WalY}8k#OZU(NGSkzsz@q8NMS1CF)M8uSspO9rl}*E`Cb15
zo#0BTW|Yt;kR=zI1@QWIe<8Xocf-+!qeJdYj<Yf0NszX$H`lV~e{tHg4HiZw37XOJ
znX`O2`Q#~tXY|p5#QiH}+qebezAEjLVwqj^f_X<bz{G@WdB4lVTTAkYMec=hoMn#o
zbM|;E_o^$Mm;**qR6z_@j*ySP0rq`@6zDucrQCCAByapKxl0Dryeml43o%s0nmg<#
zcy^AQ#EKfIms~8uPLe>UmG_PhzGq6S>Tv!FY+#-)dw8Q0rjwR&vqn)6pr1<Boe?tS
z>N~46*8}VKe-*LUq4Lr-wTpbWc!$@vuHN*r_h|)HQs1obFCtPDhtWL<YmjUdF!c}%
z|J|S5!99l8GMM@sJFUuScei>_CY>|68L|%4{Gfho+rH4-c$-Y{eCg6?MYoye=DYh>
zV4tJh1oy`2R|mdWw@LN?jB(;LiYz%fDe85hH2a6bsjDAVd*Z%jNR?dybKY%=b|97#
z=H>}r@=Szmm*F)nT$dAsQ;mQwy`}|msO&~_dzLImNI75qzJ>UP6DJ)6as34DTy)bp
zzGuA`@3;3HvSlwVEoEXDZ+pv1Z?E|K9CVhgPI4yw$h!EZb!USQ^zq-I^TQ;(=hFQS
zn;cg6A|_z?RmIh&cx#pJF8?gsdyfX))Z=cj59FrY+F6~6;cFPbwkvnni8J*UEaHY;
zo%gpjo;1_kX06u&kLGytC#83vhNJ}?>bAw_-Z0EHD1-DmbRBH>pIco`G%TJB-O39@
z>MZ*?ex1|1X+G1E_hE6k#9aDdUmrP&$G%i4yO^umK9iq)Czxaq(rhHI#l_^WVI<?O
zXr5KR|90(M`(&BuV9U6_?{r6h*V9jQR!&Q4YY>xB*z$N_HauK^8nm)jyFOJ+XAkhT
z6gC`On@#p8uyAXQ#jNYbS};Wbb;4i;3F?#s9IwVTh-{+-Nwm>Jg$3f=;aYubvqwQ`
z1V3q@x%>2C`qJUa<=EFHoDNd)ooC0jUg2y%$6rI-7QuYy_xUw_G|goZ(_T=X>S0uE
z;J7gCd`_zRy0aS^<!L@B)C=!Q;?Rd{K8;mRi!0HRF_*KB;t%|E0^>8qjb-{^b#rUe
zccH}S8~$`1j61F+i~gcQXlYNR^c9!wpx(ZfTFU&DcB|I%oi*+E<>*fz#V})FG_0ab
z3HK>E43}%=1Le?QMl$iOXHMQ}xOyV1NwWf0L-j>^{G3bxpZsokkzD<(l^=((IVTvz
z=n~6KMNV7>XV=AE17~lHmi-n2QRZ^K8+4!SyD9s7C&)Rv#2mJ7Cx8oOz7)-CLgI1~
z6?nMua-C^JeRFY`t^ZT^v(;&?2majnOzm5d+j^vyJ`h(08Pl>|vZ`MrNm-T)5qfX;
zbywTbGVT74XIk4{8mZ*cdbKo;%l~K$J}91DZh*DrJsf;fp_=)!&8^v>d^btPTKc9{
z?Oht&G-%)~Y?e&Pov7-`u)QDh@{9C`>ZLN3$!o=a(}MK;`@B_^lk9zU`h2hboN2_o
zSAz1a_KGzc9Nv^Y>Iz^DEokYl4_q5F!i$zx5Zv{y`vM<m#uaO2Huux9DNv521{|u=
z^S>xGxM2+yFUQ}a6GmJoqrCJ?lerE}hDee=lqFxk>P$b}omvuE0Uyha^T`Ogm(`zQ
z`Qvk<N1mbo<EuE?xO@{Xs(I@uLpsrVMUW}Is7o0S-^89J@DmM;G$8lJ!KqqV$y^1K
zPTK8B5+V|MnmhGvH)J@qGlE8;qgNKJ<2}cgqqMx-qYIdD_G8}>CoOh2yOa48duS}G
zPrGzvoyY9)<iqZyW$2Zv^xh(~e7M=t2E(Xkw*Ppp)@*k4^9Ma@G6b@RdeyMMiyd<K
zgjcJAvh6*r`>r{<**X(B-RfYh)+|xe5)EsIm2sDjO@(P+pjuwk&T}YLbZRJ2S?Fze
zyuZKDE$H!8<F%LPfpaDleZLOrs=c$cHDN!#zL(F!_gQ|l5Y5jdW2;wQK0^GefYt&x
zfe}0U8g`Qr7JW{!XXdmJCn5SWf&@b>$5x)SmhGE;ECs=Mv<b!o>fC*cv2W|0o#~os
z7a{&#myl#b{w1_+jxSSF$(VS)J)#M?muvI_YL2*_Q(-bMG_`1`KWQ}8#BdY$*s;Xu
zi+OYJY%k4rgTq^PhEm7$A+!lIe*A6x`eI)*ZqTAjyAtJmj%wuE7%lSt1BP$-H&9V&
zQk2V$KZ2+NX&RQhT%l2jM9b)5QDUdibXEq%wi$58U#6)vX1{{Cel#h8AFPOBH2j4~
z5O#V5#wG>ZO-~dIuQ{}gm2jAKQ&Un^O7H%VugYv+>PtV|5rnsFKj6Db!DIFUnsoI;
zy;P~~z0Be<rJZ-m%jz{Qs4V{^t>KZEdx*Ua+74~$J%#$A)r5g(7jsjugX!h+-Q+c+
z&k7;s2YXxXcVFA(4fyvOrJxSK>*GI{q%c}%ztKqJH&=_j6<hXxxwB#-$~R|%Q{+v`
zQ{)qwoyDm^rz+{OT~lc46)^85`b_L3p{BC{8G99t`L%C1gAoLEn|ZC*XF}=+68h=A
z3w!mUF61cBsEd>v#h;AreQ&<^U2fnOEwAM;HlbO%-Aanu6q=&tt&`#IDK?&xYM39k
zsCeeGcObf;ND(Q0==zZ(`x=7MG8W#@$olov5_REz^tAn_tSXEGo{e!#O}&Q}X}?N<
zW*U!x)*luMNS%2lTWWYM4^BCko<2%(8K43%gbgCjju<qcdPH%boRaIs*6I@|f7Ie;
z&6BE`O7UV#4y__J^;H`kl_`OTtaap`@iAJj;+E(!HD1e4ZKlZ0`RsZOy(`B0%-Wbt
z$DOtSy2?|S=hIp}5R=a1Hvk~&g__{s-L#*lNqG@d`Qks%@^{6V2D=jY5jULZ>Irgd
zz>ubvmEs_CXp;XZ47lp9jp9%p8DB%gIrZ~|{LcO7m=#xC;f;PUyez2v?4fFE$0LYT
z<$+{*Vf|sB*}F|(7b>F#`O}BPkMrVrhtjr0QBmfjBJt`BmQG`X3^)nr(kT9D0OX+y
zQv}>1EW8G&Y?y?f*Vf5XR_#7Q!ag?k4kl@&cze4kns)>@lK%#P+yLy+)YSBe!dso?
zE7y6W5}(|#Qrh5jkm{0U_Z=?2yaF|!qa?p3^<ZMKEN!YBxmA#7F1$40+TF08yur~)
z`M&l%7NE&`MR(ri(a{aP>r*c*cuI!7`b-}=*k#!0ie{o3YYBMWXgRjqUcNM{aH7F+
z%0X=hLsJatk45wRbmowf*m`?7bktl~7kmI2P~jelDD?9ym>s~^M~4XbW{=&YVrbny
zACo^@crReR{EMp4CxTZuZLZt-L^~LNYGi-sK)n!@mca(V4P=`=n_|85Ty~Cd7J4H_
zLMlHDg6L&xD;ifeM9BmL?pq^qf|z3j@V^4r;ty^KQ@xY3ezY<(x{#%ORG^1{r6!13
zF?nrSqGj|dn@|0+xMvh%`!l|)Ko4}@7N$x2XSLoQYFzfI5N>#@SU`Qts0V&OVB%mU
zWxr<28b2^yV8I71vi_{I@Jv7dPJ2DdKw^7JTc#yS-~?=ABpy2^UhUuPi61X4*<S_T
z&$q4Cf{*jTr6>4?zZq)EsAo*UQjJ3$qnvy>XR42ek`^QHY)ZfEr^x5mmE(Andu9gE
zS|bF2#-4!v{hq}aC6hBI_!v!}u6slzzW~70#QkYAhobU)Y2Et#<3UvXE=PN+<*rJ7
z?am9&39B-XO}vA{7#Hr_H)5Z;=##MT>%Ru;cgy;n(pZyZjuZy-tsM?Vw<j-7Cvq2d
zpZ+`^fUa||%gtQe@r@LA@qTEK=l#rYZRv|3e2x@pXqK1GBfQ=ZnmRmL16p~h`AZq^
zj8n{Bw#LR=vByS8D@pI$$4G8{7t#aC$?)bLWMQz+=0I6>^8LO1gN>oJtIyeDaw03r
zznD91<Y&i5=3g5V-+Xqkx7q&c;RpMvoNz&}How=n{Zpp1154x)bDX(G4R@A&d&%h3
zdtWr!kC_*?awf^d^&`_*Gw(m?E`JnBgUa!1?)@DQok2}pvLL@et@1Fa;6iVt)yGvr
z(9i%ms~`YGd7jEvheGzKQX{3`N{&*VePQ@9KzgLQ!m8UdcP>NzzJhUiXQjiN+LFno
zJb6Up0Eqlmb0MEV=VUxD@QbTfngj&*-F0|pR%>A*QX;2VBki_&m*mpobAXO7{}{&L
zkPF2$>af9DN7LC}I`OaLwPnIJleF{#E=7#`_#tQD$IVBDVS@@jE0c<*sKIoX#iECV
zp!)!9Q0vg`Vp{C?R<hDBx(uTwgX_xq0*Q(6@{G>pzM2yLJUv_t90mBlhalKoWT&gx
zSv-{FO%2!o2g>*E{Ozix$x?U77}Q*xA;03@{HXLfMo9!2lKoM93vJP~<m}<W9&9Uh
zY*2=d&&cBRr5jNZaMd}Vbu!Y;`7#sEei|<Z6>+?JX?+v6%LIj4SZJ;H_?+*FxeD$J
zV0BPbYVh5HCgEvb`2Th(ASf6DmKPdDcq(61fgnn(r86eEe?O-l`~#FLPenEG=NHqL
zesC;juU8T<$3Flk#hXTl?i83uN;=Ogw#JDA4#b+5Kc0z%!dn}}n4-k<e@pY<cS!{$
z<C@<oyzC`#uP{O=I!2<X{{l9D7_$ZfLn<FD6AQxb3;}pdJr7FFJOffx7|q?8zfX3~
z0<t8n<;4F!Ixfiu&@+?Q6vqh09}2sHd-18i>kcpk0Rn{RtwaDk@#q0o#9GJJS*pJP
zUnIb~e6n<eg;;~_(|Y52REJLyHeh+-7bG3`cNPGM%O*7RFY-cYEl{rj_$3Ma7l4!h
z43>e1E8m>tljqO7O2CSKGpd%*gE(O53vB^M1CKcZ>E9x#_&gNQEQL4=Sn-=Y0+zNX
zl6CU<S?5M>usqE0k-0L}Z&>DE55@9`h<oJnRd8vLDOd-<7#dXAK<Ga>v;{a+Yb}oh
z&poDx_x?s0*#U0VdF$|NzK!C)2%i8E78*X;{KOXJ+!M#CeRaY8{yFA|*-~eK;{yZ)
zN|0s-fmrDO5NicKTw~8Vp-X^L{+SFwZth8dBK;wEnIMCPuyyb-E%ZtzVT_9(bOSo`
zaA2I==p4b>|33UUK{*1b@WjXeK>#3tfZws-2X+5odH<>!53V&n`WF?tqnKo~kq*!r
z0v}nv`GW&3e~G%P0A|4L+zB7TPk}%$bdWEehY(u0a`zuk2><0$RDC*?IYC;%p8-Ib
zJt&0F1e4vGEyRB`9enOg;zEWhVvAA@*7BNo+S-qG`MvxIfZamQHwdTsjY9n<;jgL(
zodATKRB;XgR8<1}ZH5FR`H}_kSotyP_^}zH;2{$H*%8}{Af6Yeo*yFU-u!t0TUHXt
z^cSM4hzFWP*^*3vc?l{<0zh!MXBB|43eM&Ke%|^4Ff|_O!cw$mPn@T%*6;I!7h(cg
zHQ#_<e7QgP+qHb;1k&e`zPeSxfxu2A3mvzdf*OFHbQJJG&EX57f$#${&=uPLW4`}D
z*8XuOp_;p-QVIFz4*#^kzR7^&eQ;9ycg0w|aADoPTzcDKFzqkqcP97dUr6nbkbJ<!
zCF&|YF~hsx(o6%pB8W&12~_9h{JW`l@UlX+|KAWu0Qhr`Sb)J#h|)_P_dbF<fQ(^-
zGK0gKM+cHddpqKPlrV!}Z{a8ZxFbo{mF@=L>5Yns_?K1|+}17!<4)LwKKAK(b-t)}
z-aI~7<PDm)%;x0*oRKm|Z$7^X2oZMt%HkBbwKT!Uoo3E6aj^aPt1FC648J|~ebcmI
zlFhUxAk)IvR9<Z7BbV;hMBOafw0bpg5W<75m3u_T?Q}5a(%dEE)tWDA!OE_F=`k+Z
zd-rm#eBjau8()EsdaCt7*#3Gu1i6;U9T$_M$s0%6Ev5g6hUda2b%v*y=<Ut9(-`;?
z14eGz?t0*%pUswx`&?A*(XZLnAC$-+DO)mc%WvwiG+5{?pK<i`uzT&-p=)dXs%btn
zrZ8}GT6EA`JXedVsuF6cTrVanUaaBx+H&gpYom?ZHw}v`GtG{wkY43Wr)h7nlxd6U
zIwiZ;8r(8A0#EuaWbF_zB{AR$I<8*$7o7t=n>6DQpmORG0-P(+jf5vT^Qo9syKTJM
zYnc5Q${oN2?EI+i$T%+L5|;`87YMqxMwP+B_hngUj}<C&9+Cb9xqT_2;fI^)*5?zL
z@nMhegMM?rj#S$V<nz-}hgt9RGAt4vG!Gp%r(uxstb<>e6NNaq`Ol!5@4g$$(!L*e
z{sh;UXsWc5(-U$_29>i3w~eKb=0}T?3}rHkpWr8kPqSjl<uXta?hz&05{evW<~hQ<
zB)H8H<)}So!8?n$OpEsMMKy6ZtPDO;EpjoTYXshe_S@ns86@n2&!o6VS_CSx`P`Y;
z5qn|K`k-dCJ{r=qDT3-V+4ZdNJZ=1eEW|Tk<l~cdv7_Yv?7(=VzU`1|o95#F!?#QT
zGbdQ-{QrSt4}-#3BP9=>y3^l=ZZh9y$#<z%8!4-`;xZO*GsCYhlryci^yk{Iemx0m
z?rO0A+Q{2i^C7>A<dL`BD+6+rS(piHLJSyc#PFmF=rIh+u)(Q+T-P|`6?Nj|*|t(&
z=BK^=%kd;TYwD5Z6)KOf@KzX^;AfMT1LK;wQ)743jm(auy!MG4-43RQ-$O=!c5X>O
zJDGDP=1na3!!WPvWpu-GlyaJnYU=hx?%=Q4C(JCO*4T$MD)@i{rS7O<O6D@7j*W${
zYV)a9{bpiYdl4zct5wGKBX#ZeK+|F4AEB6bk9hXF;-y*Z!1&Rcb20JOfy%<yeQn>3
z=^B>}+aYweB=Q8{TMLC?-uuJ-wAQKa!ng|20O$PTWbqdjWv2yZ+oWm+outrxvfM-Z
zbWF_?N9(yBPmBla*lIkFcv8$Ku4nFf6h}0JV4Ms}Coyc>XCqQn!qeH>-);?0_@58=
zoh~$A|NXBYPOz^`&hYfrPO<Dv&RTyUmw<c<J143h54||u&sTr;%13)aL_hWH+dYof
zY^sHO`6+<8UY}|1sKdjbs?oIHm(r5-KjRf-i2S17?|G1=f!8a(GhbxE-Q~_qC2j4!
z!GjmRQ|s{{c`}WuG=E~{t)yUJ?c97tPh^;Wd#X27{9N*#F_$X$O~h-=V(*KMnrcHn
zB)dtLbb&uIvvgkItN2W)kg}pB``nKt1^N6dqnj~tqE<Jx>nfGNbA;g<3?owCxr$Bd
zeYcikXu4}}$#aZf6rg@$fUK#T&#mY>%P4AIdX=GElc7)Qdw#XnZYzX>6m%dv7o9Ga
z{^P%YP^STK)6dFrBJp>JyYVCmgT?w4DSfzf(O0u!p>*WJtBNxtP+dfjozTbAs&Z9&
zRWI^~MdD3zleyyfkYp;8sav$%2=lt1QetJF9u7LaEI=i;W@SiasZ7dMaWQH$V}I^9
zrv<jA)G3?ic9y18GjR8n^a!7+4jhi@t6w&$!q0ya4D3tKGv@_QG2+Ih0*}}%Pij9^
zjFe<+pAC;mDy%YCG)MWCp>6VsK0Cxs?}N#pmXV(!xR>UW7J|&a&UFWe&4RwpGW4R@
zjJdTDqpuRWPDX_cn>||oOR`0X==VlP?vf*ECeWn1lmgd<9(3Nuex)zxxwvwk0(Bgj
zwV>7FW6+i&M8-?dq~VPWkKfAji;zV;j%+^2KQ;1IboR;k#qA$!h0{5k2l??tl9FD@
zjDZ&GuI$2%y|H=>v@mgmxv>A;{bE$j#^PI*+!!#RRxR$l`$}5ypdMDd8}7w6E{+$b
zI4yc%-Lq+7xl)}WEl{h&Ar^nQiHc`1I9SdwHSK<jy-<oj%zQb<_kp9k9S2Wk0=Sm2
z-MZxqRzq=_&7taAU4QPOq!QPA>sJK^s;Bio)VngjtY@ImN~|#<4^Oj`I_K)Wy!n)G
zB9Kqim7x@0UYI{|qf>(FY2v^eGw$t(H?H>Ls<$B|^@M(P%@bf7|2e=%L<3YzY}&c9
zE;7taKWyvi`+JAJ0wMiT3Kg2uk04b01-09f_#Sj}?aF+TjXYs;dOWNOxpmcXq@bZp
z+-0(Ka_|b5yv(n40r!TUcoGH5#A{2L-Az}9^20Z#y+a|s)_a4x{wEenqy6(7ektOF
z9J$DIrI%6LV{65!?eLRm+=5e9{I!o$xfPP4s8MS!Cc`e=yQW^6wkCiIVMnUZB(kVq
z^|dnMNn?w*aMj|f?h>G0DFcb*@HG2etF!+~cb<#`c+}wt_yKw&{XcR3$)zG=Dg1|J
ziLaianUI2sm1ql`I-S^7p7`+iH+pTAr}KoF@Zs*ney!VeDf|uy!~A(S(+cPD0R4*z
zA~|VgBi)A|`<rz-dJaEboSRd>vA*MDc7>{W-P4E@ATg*n+B}I|Uo3FJbM3@A9oA0y
z`~8viJ|$!MU(6?;*1r|$L*+$~D@KGbt|46=3on(R*)6wQ`p;WGqK+(NO0C}+=P=z^
zVSMd1^V;nssyg&Qikbn=(dV!U1(h5K9L9<1oFyCYWJ(T`!RY5cC!q||{L{!4&|(qZ
zZ}qu?DE$<0I8cGX-VX*a+=XGI3&X6cwM>>Eo|;xDekv)fEADcsDPZv3Zl%k%J8~Ob
zO4zk1!N4LPFx20r%wOLe3b2E>eO2VnyZ>YBXOG}Q`A6@N^q%^mjGo8H@u%Lq$h771
z9OJfRw-b5)S9@<ARps8bjjDu%fQpEMlqiUlgmi~AC?HA+NG)0^326mI0R;tNA)u6W
zcPc6%-5p9US{fw2c`tN7dq3~<zUQ1jzH!Dl=i7S>hAi%N-@p5pb6)eB*W`cjm8d^|
zH*8V+JY(`8YmY(OCA6QvO^E15;B$0|gjoNMsTTQFM=RGc3j3i6O2sY4hR+)Q*#rGW
z^)qOdqZ`GHte^7?f?Ic+CykIUuuI71i_w)U__;2ixT=z?V0`}&t_YBM%$z?GxN+;!
z?eJhz(5t-=)2$20gh*g~3>Ogg)k0}~LFc=#jWa6y7@k(<<Sui`$<Ak0o^@#GUYr=9
z^9*KnAreJ3$Ej4ErV7P4eB5?KSq#x)R=aNiw16(Me$7{L?Hgy6P0M;WP+{ZXG*zME
z&%P6PPbMW)TXLWy?M~Q$ZMpilb2l!}YE~`wT2#E&(4UP~S6PS@keG?p7QV&Kd#^)G
zY`amZdz~s~CRrI1%AH`Yi*flCWMqK<-Q@V}O2;+TiXNS_o8!JL>Km(3+GTw$Z-4!a
zu))4Na1+8TfdHQZzL8Re#pMvNl@*&dLmqqWn9U&$oeb9ZczED@7)N{KpO<DHVjMGP
zf8_$?$Cj_DIwheuih}#AM;`8n<<ALeR_-mxX$x89>vfLK75njthF1nH7urfs__~?b
z?LHtYh#5GocKnpiQmeVom%R7$i9N&NO6UiK?z7W77efMQG^oNc?C*vK_*&d-wevqF
zzne-nutuzeZ$Btn0t>J5COm&IYP-9@T#~p#F9_CXxNb`5mpsxF+v!3cm9dql7*@l(
zTpsT4SQX`0SmleX&$0G>>UX)7kI5Tfy-1RddREkKO{a<Y1*~fS^a}{|TFth}PWtlg
zdxW*ccM|t9)GUe3o4PDI2gh_IvZiF(4CzYQnRkYNArDO$J8Sb!D6Km)B58iRUfW4E
zN6VLjLsQ;%p8IpZ3|+Lr`o5NrY#E8hEf10^_ev-9MBKuIy1Fu<CzXfuKNdXrO7m@Q
zplZa8FUpM*tftj1J5@1fk5yR$#E`HxDj3+Abf^1<qv8J)Kjs}>MMXY@8&yeTZKz2G
z!$p7kI?@VA1`DQ2^-`Ftl3Hf*J}<}^?v7eXB>4H9LBF8q{fZ$3hpXOm=4Rhw{7Rnx
zxcg#EwxHI?rqB^ja94TCN|W|IZ3}4++`B6(i0>Wvdx#Uf@u_bO_BXjDX(#YY`tuD|
zvR|j8k6oFM{Q2ktxRFwHJ0k@uJeZM!g8X?Nq0BfH{ynDwQrMK!(>=%ZOCBsAw|cF0
zUJ^~QB)TVVm8hRE6*h5>-i_M2=VO>j8)M6pT`8O)rf#18_awX-vHngk#on<lO0{4d
zm;4I_yTzyo#BxS$!9y5IR~b7iy}mN6GxD&Yk#;~0ypX#z%rS9IcRV(FGDQiGKFisz
z?>T`@+3KSJ>P(qfJ}KhF(u)C|Vza!x@+y|_=0k=%S0r7zvptY&(v2!v&-K6IZdyg~
zy=Y$8Uet><=EVg-!J1!KkL%M+Y~AWssC>9}E(?{%&|O(@Luc)b?T}D^X<t{nMfGMe
zOrxB+>vFNb)_Xn3GrFAH#Qq#~W0!Bf3YZEG8w(Nc?-zL)r-GC19wEk1wX<IKe5U6H
z5e3_s_0D%rZBvUw51SJ22Q5Ds-j~TNiw_xeyg8Q5DV$Aqn#Ivk$iMTt9ks{8gj4SF
zo`QYNYf96C4vW?#qq#?b?++xu^^d)8l+ATSi~<?O?ElHAuU<~pQutMAJpRh^rYW3@
z>SGCGHS27vvN#?>g|BjOgv^Ppqx~<uGkPgz&K6@H{+N$#u;I8p-%f=}tRmQz84)<s
zQ@O;mwaB^m1?)0CH&S${)22+!RAO3vQnP_&MVls0(^eS1(qsGYBTE)p2^~|PRp`2E
ziG1gtCCfk<@FK3;Y}#;XsvRYX2Vc|>m#~X|)lWuOv!A?zdCnxY@p^ApwR*%DlP8F(
z>}fBwD{?a%%S#u+K~NW38xzr-VV666%qq*QV?}_le|=%V952L=x$&vs0u4?!!;!-X
zWLMVuKXriw%G-I}BDt>EM}Ok|N^@wr>;68w|F*A><k(bQjauayt_$`jLwikZ(@`A`
z7Mt;QRn{X9tUpPj7d{l7cBbZ5fNX<kr)YSUAx>V2vYF78jH$}EchcOnrW5Bq_QzUG
zM~nO3w?q<CUbqMbRLQ(V0T|=lNnXq)IrU%qhS^fb43|ZfOvYciPb8&LH|!|*r58mD
z?e-cC`}iHrxUw=$G;w}`%8l3_qCMw>7@YcYRybYVx^yB1+;i5#JLQc`^W#+Jhj&NG
z(_&jHS4-EN^+;%`yTxlP`1?q)45~;`tO5C-ftpdq!ybGQnG2BxqbJgCZ5z6hFr>e*
z*q}vl0-&{%s?fA4sS52C>s9b`p*pVl12;b%l_y;4+Ms38VU<u`@pJ4E?BLxYRXa=-
z{%Wtm+P9|B@hgO5JA1DZTvy<y#0p0l7+36e>_rU)%lPqC+_WDKea(&9v7T*uuQL)B
zEZ$9|xh6{h5QbgGj_}lT9d5fH^nw{?1;y8jcSE1QtLXo>G&~<KMyyJ<^R4v3__pvu
z=t{F#lbg|(m!^v571tTtH#>G$`)uFp4qezS%|Gc;W7!;n*r!^_W>;BWh*!;t-;?0;
zSbj{=-A6~$u<iTQ<=0tCBH(aJBb)p)K#rU}PyiSMXYe6{OaB<V?`7r72l^)QMR>f{
zZBs8*DmECb3e=|MY9dsN_+Ka8(vP0!t7x@RCh5-)a+cRBHF;WWM^o_YLt3h%tU>MG
zSb|;hfGkd;jJVtEk37zdu(KQDY(VAkEUW<o%&owVA&Bz>^}4ntq&m2roUaK$7u9uc
zk=2v$yLS-{&V~4E6<xJCFFG)E{$}v*qU$|X@%~uPc=K<c<YJoFeEsBW2|UC`jW<+6
zm8iQLi2Cgp;%$UC=7j7gWT)z{M(ESOz8oIX%PWj>zu%-=Qizys?VCHL-5GgQ_xv;a
z9s6$;O4CmxhFi!X-@?Z3qx+594iIfz0-p<QboGE^SgUb_YI0|{rAiCEGS$EgP!RDi
zWllAxZ>?V!hF6N_hKfck_22p8CAj3g&&+om+1u^&<<n=;%9c2MtGP)5;AJHD#;hm!
zzcmg2ptCK+3EZfG5{G1j$F{=6<m3%s>8k!u)}$o8fAutoB4yXF+!eaG9p3)tsH!23
z$EE!y8vJibYE@549|R28zOZ2YW#YaiaU|KnDFRhw=+Ady%3)E@<^nUIl|6XsC-O72
z^R8VBKF2~#$?@v)&D7%IqvEkIkf(n6Z=UL1`GjFA@B62DZd*f2WOHH^soGnYJ1gK>
zIJe_)t8=|mv?A%)!u{k^9Gb09oP{kOiAk%(maPwkRLKnO6f07gwzn?auXxv4ztsog
zjg6<|nNWE+e7hrB89Mcl;MzEs5TGjy=Ejfj|1%7Ek`QuP_^1tMyED?wlxk4)(+W72
z=K?;9ld|z4p@ee;Sd957p(NVxMgtSpV?<{5xg(G^_5BE+C`DS|k3%628XaEZA$h{q
z0C)npeh<M=@sZKT(hdJ$qu@IvcsDJGr1iD)9|KCRdBaOAWIPDQpu>pB%-KI=%>!h(
zg&2q`ck!(IUA6KOA3!`Ku_q!P%T_N&(ik!yH0GVi|8n6j&H;`L=P-v4#6rTno5TpA
zRX>8=jeqB>zy;NYNSgws{|(or=S4kZHtYz_i(=oMU`8obahYSoAomcsSky|+{5;Cx
z^XF@S-rL*wFiHRLmY&IXcgjF~n}?031Ric?N0P+fJh<XB0|H2sqMuuXaLWm2)qe@M
zWcwB3E%LpByr=Er<lzM>z@}$v13xtuJ7~d5$wK<7;4d~V78e)ce6iBqc7pMm0O??{
zQ|HP=G7c9KmpG5~{{b60a<vU2Sxn=H{<Kem&qtn9`W{vBPi|cNSFr$t@961o13f<y
z9N~sXZm+=c{wJtUK$+k7w%xq&3(?^1uE{M1iGKxt{)ezZn=fCo4|Q7q8BI^H3|6d6
zF_aJoRs`5yWBEig&~yJ0RjF2JJ<WR0_$GQP;4sY-gtOa<bh0`t61!QYyMSwxLNj^}
zazYLQ6Hk~?&rTQTqz2{u5$ef-_r6P5vg72-7x7ULySKtT+q^^nGLBtu^lK{L2pNUY
zf%=j}J8Tn>vAz!M#FVvh)L%Wr4K`P`P1}1}^+ZLA>wkqRvJoeX<k*4VeDHjU>Xjo9
zb+dl-R``J;=rM@eNge(D7kW{+9CKe~^8>C8VMop%!7BRWTJ*0asXzbqFkWP`eLec`
z3jpN!B3_XG4(Oc4;`q1xp$mxG#D2f~Ul}z+r|QC(L!3$>`LfJa2MMPWY-%BPbNj!K
z;O__|6%6+8Y=albJH38;;5h!d@t?R}J7f}sAEZ|JAW&`k99;Y#qL44Hfl<r*!yW~+
z5l{s{8<{f0#aOa9wln@DAweDqtW0bk3CMOM!Ajx({MwKY5yu`>d=8XpkFfkT@;*Ca
z34iIy{_Bz0MOcS8#l7Cyetb1ivOs>v2Fwl}ZXu2yT+g2RCGlres*&J&&CEa7_U1L{
zE&jv=#)CmQ16zifwFT-7>`#V;-k72v=nT2pzwU|b5FyZQp6Qo8kcf&NtQUvFks1eM
zbr6B_PKIeHa+8A!i6Fi~g0rgZC_?0*M>aj=#~Wn%Hyv}<K|EQPKcUZqzb&mqV*|Dc
zFpBI2G1%!&@cdBKo1;mcj;D~h58UV-k&XNJF;uXYWZ){sxoG4MkUv)S#vS5~zu(_8
zh6j;9|GhlJ;*n68&xujWmFibf!PpSk=56oTDKQQXeMs|p4jbp6bgW^@QA!6Sh<MQC
z>K2EuNYh{Fz)*L(+4Djp=&u~vxl8*-z2gRU;?NN<tqj(FNwk8iP8?*G)Yl1&B|tz@
z`tYO%-RajHe+BvBJtfqk4;6mqK=KOxaXq-}Xm?D^zY>h#0$`r=UBhZdNl3x9ggH6n
zfA2wPd0f-M^SmX5kSU$gkn?A;Vms*>eCtj+l?|4$jlg?II?A7|1DPVvPr`?(=Rht2
zx-J-I?gL9%rGMOl26>T%Z2Qwlr~lz>VN>#!st$>_>>|lphmI`3ye^q->O<zp2*J{`
z5<5i&IqZZv!dBg~a)>fE6LN%fU{17u`X-4yV4oPS+P6|B!#-_?6Z&_*=M637&i$1)
z3=iao4qia3x1T^d*yj-Oo4u9tf&8n7GgW6w#nO5nrwE^c8aX~6;60JN5xC|NuJZmD
zk$0+py6Q0bmE9>pNnK0)3-IHdAkh(3Xhrm;9IW~%<REl-aV4Alo|ngUt5;*EoLjpq
zzHH?P*PJfFHREs}+UED=!#`R=O04vU3vguNAe?;kse9s-GBTF-M;epsw{no21!Qu6
z22cK(%_$pV-@@^oUTPm1c;?9)#s|{>3@GE{zt@38DGXaEIfTcWLn=<lBNO?=)pe|O
z_J{s}+#N~E31$3$@had==$@CK!$#{NCnJL4%4O{6!g}+@h_?{-hf*8k+F(22DXgK3
z@4t#+J0krzhXa-X5(=j~#id9I=d8Ygn>jG3(UDH?5M#jm=aHoSyS5_qBnP>k)rVkI
zV)EXzmq>=9S_>iTf<^Pgi&HG>7fSFq9T%0i&Vdl^-p-1eQUT<|BN@@LMDAs}pg);O
zmw`yxa5EWhb55T1`~Lfsagj&H6&gsA1YAluCK~EqGEN|BIUo_FP+~vviBGD?9mG!(
zAw5{V%-Ke})DcOy|8n1NV}5|S%z6G>p$FvmCL)=x4^#!j-aMl&ogmLt0)4;;UAu~|
z?RF6Exd1m)v1r-fnxeyWr9~i`MYOKtCsIH)Xs~vFcVlE?EJQrmy{eVQvMcqX9IdN~
zdTrf!GXU<LQ-er11mMBk9jwXc1tb^g-X56CuBxr!gV*IjIcF&iRCf@erAThmz*8K0
z0@jRtcEfBo{rN9rJ+~DQqBV%rM1Yc!0_d}xyQ<??4q8Ct{p&xTafYX8B<p;C`;@Ne
zp*l^zO>ed&{OMG9zQa_r{d_@BU|8HdWR|=VbeNP6p%qP#dQ4OjOqr7Tv(p}u@$Erp
z=1bnQ8v}5f-Xx_$CV`7w^{PQ*1*5|6&7v;)&orQZifaRsMTrzqV7{q>-q7*F!&Q6R
zX=?>!Vw)0L>9-j`&?!f&kmQ`eb+?TLn;~ba*)X2XzACRO>;8NN=eb@6svQ~Pt-CbB
zZw=InLNYb2`f?Ksq+0J(m~Ak32l5Gz6BXIzA8H&H(`qgJEuvD;fB&N|UVG}zG`*s%
zYgW?frO#<ab^J-FnQ(1#5{~^U#fZxv8+(uNLnLHiE5=9IN60!&fczqML8D|OXQeS#
zT>J;&{F3x7UQA$7SK6(Ha@Upda9!J%en-i4KK6Pa#`%*Z{wQWKsfcFgvLMM&`+X~u
z_ffXn39KmMo{i{tw-e&#A)8uDlvk#!|2d7Y7RVXV@m)8SMaz$B<=;1YGu`>g6KRJt
zwBy>J;v4^JhZv^~NRliMJs7FC%Bo-jrGwXIO;M)A4YtMK@n<6~OW$K^$i!cjI7}%%
zJI$V}#fzDbhxUb^<uSX>+Rl6L+3Le0`3lgSK|3`ficjI31{@&La@YP4OaFihU59TW
zjLw(Tv`oV^jpSgbf$5T|80w#bZs{mSlu}y}O#rg__(B-rba_;VF?Jf_4#IAxq&jHP
zY8Ly&bMKAV6VFeN2!TH;IjaQfMg_JfXt+(DI9ZwRZm%VR<V`52L8UdxV1;{8`TpK7
zt=kVkV5p(^{@D2YXB_rB&dMZ9QG7O+sTR_83KT(EBI2Ax;D&Qe?W;H&(B)bN(U$rT
zH{!=Z7c4?}p}0n;UJIu4hg1+wg>-#DygWfFsk@}uiT4W*|4_i#rz=CB&e1`eA;pvG
zm12k5L7TbZ6~d!E%0Z^)9H?OBt<Ded53GD#Wq-9jSlFJZ)Bq|~8Ljh`zZQrG1W5XL
zFLfBB&hT4xv(2i!2e`K`VE}}xZrfJvIy0A*6b4a?VUG3Z2V3=IHZ5;|51=!F<)+=@
zM_l3LxV?&b?<MwYjt7R|`E=k{m^EMXCo|#!&K8vK#o8hCU{F#sBIGAn)wu1Tb{HOg
zuN-0sAQAFl$SFc(bJQ;v)pmY%ydy~^><AuF9l4qa!&I!7XCsA*^Rue0N$Odx4h=QY
zt{cpbUD09#Hj|B!-RmGi_N_3rxlOXY{N+IdGq;cUMaP0$(`u+v({;P(&@9ik<JthI
zHfiV_7A*>njb>%ma0}%<_uM@Haa=!Tkl<S~I<a%%G4uPzAE(5Re#IGMaC}ej$%TNV
z-?w!nt}9}1a<8qWMc8xraWE0`SAsDHlAhT6bfIzVwz<~ApUDA%tXnn3AMx6+pl}X4
zC)h3&j&CCycq&JV$Ak5Q*ctIvjdxObAe!VqZ4*ybJl+y(06i9FW58U|I`s+jjAAuh
zdUx6;0o2&SK%-5#Pjr5uc*6Lo&GuxQPJyNAY(%Dj`)Zq9hF8PZG11+kXx2{q_tUzf
zPFH{6kcrTPR!q~jJ-8|vRJCris0-s>`@21*8LMv}3@cdB%yeg1iPB}>`~HbV)V7rD
zG&r18mnZ}GF-y-Fs|4`BQ_5A2gzbAWqZ)R}A8&=h|2!KTcq=x`D>ff4$GDUV(zbhW
z&5Un?-o~r*cI0ZUNiDVQlqUvV#)8^`X{u)w=`oPtZ}X1M!@mb|hh(MCc`&Bnb5I3O
z*OFqQSEXuVLHtRl#SEtGh>RbThos+dF6vV)du;1~Ir)51g%tbLo=$vnjN9d&XH0W(
z%69BSxu(z7&Q*4(3bRk5jk8K;E@Qegbl*3v&U@n8RI;<vAgr%;*6NIS%-Y2C!7jE_
zwZ4F(6L@}pe^dHeeYVN*Q!lgQ2=MiA=!q^}%ACHepk1$>OaJNOR9qac(TNi$uJ}D6
zR5(BRJTA+(_`(U9Q13_kuHU>y-nuUi3iAw!zDut12sYPGN-N#gnE&PG`Ad`H$a!3w
z=1O0AUp^~xcalCBC+4(US$t%6?FmmcLBve+R)NhxceIcz941Lx;;-3tyYIIanzzJs
z?)Eym#^a0Y3?;p`FAca2_<giwajTG)n9fb-oz0(djNKi#%N#LxJ<+1>g~$;<?V<Hp
zXEa2sAc_OibtUU|&fOGo-KwTY9_I5btR^JZCW5k^H%ifh$38x#Jz&v^V1*h|j7JZ8
z!v%3>`ZW6>`Bd~q4L^xXe}UDMTWb>Q5zr)O2H6<y{k}>M_dML25t%b)JEi=3HU1=T
zkw&E*Vu=#-krux<Dlx-AfvL&8o`rgB5$m;Ak|Mx%Jx;=soaDw@LZ6|MoRfi22FJtQ
zwH}pfs+rtqe=<6bwb>puhcXWF?!a2uUX(yFOhc`pz_vZ{lUp4m*j_UAL@blz@`_;T
zCq!%AA;RtkK99fxci^LW37N`jxko5>d91d~^TUH@)MYbXbyN)Vny`Ru#@TOhI<p}=
z#k82djs=to3Fl#O&xoH`0gK`rXmd@3Dsd2&@@or7qGOg*CA(m^ZJ(<#EASW#;^!{U
zr?AE&?oW5^Avav}4s0yEk}Xu=TbRE71~ia9q`c~s<~zbEn&aG+dK1z+#FTWfUWy9{
z95l`H)NLNP1M*1q4ReQ6L6G1`+m)E;wYQb5eWxFE+Xj^6LTH4|RH^jyQ7^aBKx;3B
zGd+L<b+Z$+NlX_8OU(8QuQP=3WGv~LQ{)T&%#q3{b(+N=8sUhJQ`s=;sF)Au&<7+?
zQ^A8pv}-M|MU1DxK7-W*PdhuVNr=Q5Hik)Fos4S@Gv`s{Mj?eu``SNLk)71%5Hv~b
z=)tCX{7E4glss^h5A@}k%Wiyc-xmS)n;_fA4AFjGc{JEef|z`{5L)j0KRzErFV1@2
z)j6A(Y{8*dd>sy<sbilLyS3;i3e|Z13JtSt-%6!2M*={2wL?SB;NeeM2l;CP0UE#;
z@sG!l<z}zHDI4`AZv1{0qec!hDwR-6TWF>|K#?Ci1&3iwSGpf^=^Qkz9ML1N?*2s5
z?--kp7@vrCC4h*_kRSWhpS7DX-pBm~W@MHDFIn|%sCJ&20;mHrW}-T#*tK&%&;4@#
zLUHSjC7-A!LvLd8i<y({(B1}LJmSVYV~;Q2m}w<-4cBzD1NjQNy_GkbJF>5)m9)DH
z9`gw-yuH6tzk>8s0t94~qEg0Xu?Ns``P;}~YLMHuCP@Y^gn?#eR-&sRB)yF<lr4sX
zZjmxa#j3gk_XmnfZGIhQHOt5CnfhgoW*en_m3O-LY<=EZq7s!<op>HrNZm}u)2amJ
znoiA5OtG(if04f==ps2gMOvgkyn8ku<3k=Ivizivyh+C_3M9$8v&xqPx30OZ&smVB
z)<yD|f2K<=SS^5bV9BbWC&J3D_-r&Ptxt_^BR0}|AP-;MKF0MM@xk(_tx0R3BS$i8
zY26#yhSG1Qg-^Fvw#CaS1)me-KIi|Uul)uICjat6x31l5Nk4o%l73!-4h<R5Y_js~
z2^1JQCEht~RjT%egY_I(ErMI>WOAFoM)qS4n}aGO)3MIfGTWh2)7ycTw|KO~a*p+<
zUss8k(qVpsPu%akFvwCWj^-HX+Fdnkj%FFF4T%70M^l1<U|P}3pdRylXYbddWJ-aT
zeo|DQ5rI}b7}Aj!YqnOViiAz?+sL(^zU4`az`iaZht^c)P+NDfxUg*JyhtK`l+ymI
zbiQ9LY^c0#f^IFZj%?+P`F2ZwQ5*GNV5qrE`CI$!&WPEvu&|hg?PEHU(SB!+NDF0n
zb`S}Q(yR=-FJAR|;V)!s(6-dZr6Y;vs3j>3h%Bg$92Rz8pNnJ8v7}FVXTDu?yq$Vy
zdp0XcN~K0qpENbORA;fkqJzod_~Z4Jos|~!FGq1vl2+c&oiltxx@vwcV3H&_tF*3D
z;z^<h_CL4Lf&C0A^S*qs%Dh<zg%%{vfs-2jU5*DO?736Cx-c}LS!Lmpl>B0{K8)E(
zoKv6ngXAsP4Vvtm;p9x>_|<u3I(C1@yq5u;7-uvTA0lR%I=|W;C&!w`Tb##{5!LY`
z1+2^WN-9_BFW)II=Babcs3Gh>|1+l)GQ(CI92jZJIo4F@@-Z=PR^90eqTUi4>gX#*
zQkGAWj*?HmBQLePGkWp|7|)TewykkKgcFXg+4zwS5uE%t&8b{YvuY{xnsjW{&k#AC
zk=iZ<W2@5Y=F^u{LM5?rO}~#8I+%Z3F3YaJwS?|ofQ5Wahzf#`WrcUY5alIka_V1p
zfx{hbsKgqdZZ%x)YKG%ck>9S&n!U=Y@AjeLxHLOu(_GV9g4$5E!G>=~D|@d(f=>WD
z>T0;wr-_D$ma}FOv+2Hk!eUxIh*BtwF^7ra>#k?uJq|xI`gplNU(pui`cj$-s8puX
zVRM~q5!+dkE*T3JZd?&|o<m9T50P>1>d^?=f9~bC?!AtPKy~nWjMy5^7Il?6S@D+!
zTF~Lza7|1aAxml*anD=ymVbvQE<j#9IxjK@wDn?4s7x@4U6wlYCXXd_E+$Dn#~0FJ
zNpjD{EV@;H>@n~Hbt>ej55vURbtf7P<QUh=Y>e2CwVOT%jq*m`?2ajm>6>uCKMDWf
z&r^une6%unZ=uaHA0&E*+1|3l4j;wjcE)F+DCsBz%1H;@9K@ea@u#AnvB(1*wZLMc
zE75_4_rB7(kWtvY47@Lncxm3rFt0>JJ&u9A5VCuz!3zkYnaZHq$xNzNF!jC63v@*z
z1nfS^mtT;OMMs!D%R!JmLKv-ZiLj5?fA4wf$eu?&yAjtENt<yZ8(~Vd1Zg2pScE4$
z+$TJQc!bbJoybH!L7p$T5`XoGGaxr}AKI{0zvv5a4fxR$e<|p^2g#O6<=1dSI=CSM
za>@lEqHIXk{X-n=KLFZ8RMgEsgZ;FF3i76bUxY35e<M6&)QCTTY|tNIrU?SV<1i^_
zg*ykY*molyi%@~%FoJCs{E*SFChC8`7k9y{`nlLz*ba%SDDFu)@L=yx`o2US8*_5>
z3a*+8rQbnjOLZ7DWI>r)5-l9O>MJ1tzkR=SpA1>MJU$>1{B-XoxK_#|BGGX+poghm
ze9vfw5`q(0*ZAPai#hmGy;w~-vf>Rj;$XY8|NfZr+tT+MKYUwlJR7R)pfg-jg?1wx
z;2RYP9*O!p#Chn*-;Y9iGejx^)mkl*rC;WJ_DGQ-5K*#4*E%(lhCmIOm8$ptKfmk=
zkmX7Z+e$FB$Xn$wNj+G?5&c3}?hzmBZ$E0?9|y>aW?r9T^4Y()8RSRen$Rm>KfvDV
z^M(Lds{l{zIu3uDCHvAM&f>c$b|#I_(g$<mp!ey05Ofa3eTE2b^Ww=7Oj`e^vy5HV
zzx(Q7RzUY(0R3Uz+@?#52m(^f_Counu2#sqsHO>pxZjdt2X|e11Z#>p>8T_#3x85P
zNk|Q{EZl6SP{SI4e@1*0$>VH^VMY+&m=oBG_*#@%aSn!#{ySb_b3^!&@#?2TY0d&u
zkBzvm;f5kr$5@hHMj-ci&v{(+CUuu}TkO%8gC>IZ073s?q>V4%gL4=>pL_7zryD<z
zq_0DV36r4_d}5~#ERy5*G+-vB^6mCE*?M&JboXRlSyXJDM6*oEl4Hfmn0y$J@xODT
zE*1G+R^$1x-_V>e%!lU5!2dGmz18q?Q!V48nv&`8vJRxEBu-&Zgp>beAuZ)dWT*0_
z_C4H@T$(wUZ88w>XgUj`D?H*ljD|3$n}Vqj5xT>4#JFk5ZENXpv#<eE@BjX`B@%{6
zkBIG`aq&2K(f5eybqw@!{%wVNx8NN~>;GW3vQ=9>dw{tgmqpfT@qbvS$1^NDl%)AN
ze)pXOJr&XinnMJLvM&>$hoK<arwD+6A1MlqRe(RlSPUv3CT$3~lj|^!B1?X;k%CdS
zdgF*_@`>rqJ;p=y39y2_@Z$6i-u{9gvI}cFQsvJNmE8i+%n+Z;Zdil`woL!SG-3~L
zgSo^4S<HECI{EV7!r5o7Tl23~k3lnNV<r4UumiutJ5Zjc56>#I&&Os4$;%)wl=tHT
zO4wsN|2dsF7L7a$M$h2Gl_m>(4Os3e+IdNA29<h3PBRxkXe0tWq@y(V0ES5f*E^*O
zOk2e}F5mI`fFTf1$p#=Sh$&uMo@!bAKu5(mia@yNs8xu!-Jr<!ExePF88YKAeeN(P
z5m5W^d8~c;IGbsWk>JDeI%OErz9J)jl<<RvXg2L!nNmm0dAZ<oghEdMolwfUQ=we9
z4)%+wVCzdpxv1*qQcx)*0}Cevgd=O)i1};>Z=dG8!v-?B4xMB{C+=Ydv{bec^dEu`
zvachWMLA%we-fH(Vmnn5nS35+oc#(xm!=-jS58Es7HpA*5b-_Kkpjq}RD#iJic^gd
zVTqRu2?r#hez3$LQT44*uKk2!-gC#f-kg*TOa&B7Q_yfql8oUH>`E<>&jgG(@@L~4
zF6qG2?3clJ<S~j{Z+R?%BmrLhrhEnR`v?#@!_X(x2deP(Q(eebbOBH_2fmV*_%>AD
zn5WBRB1_(VAvS<pn$;)ol`OQaBgoeyLh=Opm*;amcik{;@d5cZ16teb^Q?mQ<5xa{
z9DPTE%+r(YaH`iw@+9X`ZtiEa#E7zj`NO4j+v!+S)YUMEf!9q$7;u2NufxPyM_+GU
zm?gFf?Xc_1Ad+=6pGB%`kckr`;*VWVFH(@X7sA*MiVKCtwWQ?Xm>8?t5V{C*X4N;6
zs9uV>&bG!q5xe1XCISZ?n?jI%S5Q!Ri3mTZYNp&x?FAj(@5Kh@x%t=V9bm>)!ux}w
zo+R2lDbVLDqz<x^PK#hmkI~N{>*mQku1z;N**O8j)b?SQF%n(Bd!=4c94cKn<c?lN
zHUou`m&>?}zv7B+b#@=JcqH)(wz2oPN=0yJAyqMdC9S?ea^vcGBFeGx#A-xPD%yj@
z7T1P?Rg0Ix7F1>L-1f^M6>9w^P<E7JIwir)<<sY<<P)Iu$?IM@1lWxv>&z#rQ~V6a
zfreCg=I+WRM7j?qPh&`TJp#@Rt~3lFM-dXRci>xrTZ7~H8dOZ>GxbV7Q8fZ2z=Gf(
z0d&12{ps1ZfZF9+_i?AdqPg8N$%K0ndF>NsK3rf;+=j*)aorD7LyQcepXGeWAv2Ej
z1d{+rp_42)YNNCw7KoE6Z0NjpasY)k;3%Db&$g)Mm72nH6*drL=^*=;1ie3ejqCH|
zjTj*Qt_;57GDNl{pDV~lFY)ciX?|ZsqPh;yQyDUTf~j1oT#(;Cnt|%5T$RA#=Nzh4
z@e}0uzuUN8sOU6ba~!0;8CY*+HZFfzz0d)c+#OO{kxQ_Wh)Tru^30kN0OM)S2E#@P
zn#Lj74mOJn#B$lUzlwZB0(rx4g|8h<Gele?FRLafj-57u;kJ9e&2YK)Z^!>WzN8iI
zUD&EmOIy>zZqkg5=vJFh33OSgyQeF1x>jc)U1l~6-;W7kKuka{om#~95nB+UIE-tJ
zm`oD!Pwn8V4|%%EFF`MlvLYK`$NU8p0>7DzR{P{40$yP<RhvJES+2)<Yi8<c=fVpZ
z8UET>i~!kjWxxwr!I*tT%uy@>`m~n-$9%eZ@dN7Fo9Fm{vnT1TkcoaAmmA19W3$*3
zT^{1`u-ndDKoLJ&1_AH{*mXj?xZA%xqf|ok*_e!d4MHL<LR0|JFp5#tLRKz&GkNtG
zj4B{2g-aR59SDbg5A86FR4`jQI1z$*QNpgv2`P#lVbVBRjB>$sZ@7#TK##Y*!0J{#
za4!6PGW5%|u021)peFM3ktS%lH!Xtu9UgVnN)a65Pi_pph1qY4S&-fEq~qN)ry*Pp
zOop*EiN6*r%pl|=2fj4enzSaOhQhA#P}_juttm)aGpLaj^n)pkexOr}NO)K7i9m4b
zJQ!xr4n5QL03&*`jAiOE;>H?a4f9EPz)#dVF7ppr*|aApEVnChD!?A;NBXTPF<wWq
zr^5Zg)^Y=<pi^fbv&5_beCNyIw{j5n6@~tDx6NM;fO#T58Citb8#~CNwEhYu8K5~~
z8M;N^U~YrYa1oKO#q?ywjyFbjtYJ$fEW;2ZV$6+_!~m^?!NH*fr<V*)Ijjf)zqqSM
zxZoY1G4O?e^=^(#r5)5;L}{;Hw5l8-$&B7Y#*fF29|ew002F<Yi1FHUWdVRpIYvYq
z7WPFr2I?IpR4+rS;w-xbh=WH}u085__39K-<f7NTOt9nq)gue=LM3DJoM87uz_aiS
z{R%+~98a#!GiN}Q5w?TV$g{=B$|MJ%iA~>+BIX0g4=aLnwlt{;KoChM=Rv{t6+o;3
z<>E{J8p*2lfW3@E*$Qs2^cJX9l1jLhYua%7hQZ)-iE<d$AEG!S5dZiD#|tq!klF(=
zd1@4<UL#{dhaD3Wth3Zkb=V8>XWjQ)-El=Ak$5r%b2pm{iBxk5gkVjECx|Ekj~R4?
zC)gJVAAsg+KQvkoR@IMB&C;Uh;ae}C-hG2TVNZYsEBpEN_y1?tg9y;`XWazF$gPMN
z-@@J5;O=+dHS&}r75lm(jgeUBp(ENzvfeMT5C85GeU^o3$T26<$p}438+pke3qKrx
z3AVY6-dWrK;zIq8*TYy8!{&t~+Gk#Ahx6n!Vq%~ECQ_3IAKiH?_!Vo)zIcsnUW<N#
zzdL453SaXN7vTT0lWs%9f{F$GB)Ukau7kMZ8N~1>h^hsj201UM{8wL;!^7%@i^>r{
zX4D)$d|(D31dW`VB>g#<MfYA6L-yiHXxe<1g@D_5w)O2_N$KvtoVH9yY*)-LDu%yg
zDk^x9wtqvdSy<das*%#~Ne1+Xb%iU?WB&5yK0-fla==!~bHtOZzZnb%yG%EQq#$ah
z0k7WEj>8=3nzGx+Xjy`6cau$8VupAe8YJE#ekngl5*)ZT$W|8(^I4E0K@>Vmxw=Ji
zAhILb^y46Ad9#x%l*gQ-q{>*WWWE_H@5r!>?T^HIg+_#nvbTte_UvrD`pMI@X#3bE
zM2&_$L{<6a)E#}!f~;-1#z~R1qZYlrWg1E;H<L4uofVW_o)qO|_)3u_37w@BGliw<
z9K6XAl678*k60Prv3`X?RwJP?G<z3e*L!`xW2(NZr#r$LLdKD05Ztc!e=zdmJ5Cu|
zJobvQkDgN#NoFIY)v!LlG5;uzJ+<8Nz@@-C*nduH;MiR7^)~P;gpihs|0?E^U51-B
zQ<L7Mq9W9529vx=xthX>XHy`Kr#~J(un_t(%v|A}lq{q*N}`QFeRu%JnA2U_1DEXJ
z#_}+RFJuPcqo?Obd*bGINvskEs&wnbZBdtulSR*R3k+1*g&dpiQ?1-wm5$t>vr17C
z^%CApdswtay`}r}q<V^d!LbgvnNEei<*up8mb9Mqix2^RNmseA<vFzfEZyb)Q#N<A
z80}CUOlVa<Aa49?Vrh%^l$x({Cob15=nGH0?aG2utQB^tOH_O%`k5cx_5CN_nm0$k
zT&&#H38oXid~@G3WxT<FzNXlDcY}Frv6<%Xn#bJqiN5Wo2#9Ywe7jUqtmtIn<V9Kk
zTDZu<YAE8;h)w!e5M5=ph`|y=4qjPe5+?IQ8kopu4~s(v@;g4O?c0%JM#aZyxwQ9&
zwm!aCq3BH-dC}Hh(j%P&?xJm|yIqx}8=3OU1&;u;>8Vz`qX`B@g_AAGB02Izlq^hI
z7j?4(?u^9UnDNjW(YYVNW3H;;`XPD8bJ{S<tf#9=+ds%d`SKl2!L=5(=uPRbq_fPi
zUb)(|Z_8v>uLP|2q;ZO{?v{-(htYZUYp!U!=UMcgp37YzCVXwe4<>uxqh+CgEs_{G
zCRWDtxlkZCK3W^sMD%|3g&&^Mt%ia%TOzCFTNSy=v~z=s3l&)!Gm?j$efctYI~0rd
z8)G=~=xCT%!poT=x=TMjXV6`rQ;(Bia2WONdXOP>t6|T>J#kp2NWS5hkWxLA4;+aZ
z${oiDF3)M?O(ZcDHPk8OjoW=uiV_2gq1%<uZN_pYQ|ojl_hL!aS8`d$KT0xQP#Z2!
zYAK(3yV&4&0xZ65IY0EjR^{<0faK|Fx?(_ZcJS1aajV(QNE?@prYEX9`Mz1g3$~0k
zwM|!)xDo4n=W&00(~R!oqPSyv{$l$rdpJggcBww7P;We8d&8qoOMCU^!fb2D9FI9O
zDo<#lvn_w$@kD;fmY2;mNQ?sU?FJOrF$ldR{c}~IRMi%P`N`vg3*Z}+YV5n4POu3Q
zV>TW$o>uR~@K`rfy*e_uF+U%h85G^dwTbAeR7`5R!#2TDe5K}ik<He(=;%}>xeTro
z`H!ROm%Nx!#3O;i$k_$jaB#zBJ$Nv3pgw}%P6wadfCEE{U=wm8g7)Q(EwFHMdNI8Z
zEZpd>VBfDyema&I-~BlN$}ALXUfq?8PfjMC_N<y4S{Z>rB!DBo<emXlS;-9x;mMyn
z%!;zwmhk(0_Eo^O<$kZ>tMM06E3FA!npou+Pa;<^K^S2c%(sZE)sMGHH-<Jm4Pk#F
zytme4g!hi4!fSA8Z;~E&sEQ3dS*{dR!(1F5O%_EehA^pgBzj@$TI*>`rkZ`w$9c%Q
za6p<v4%NI8U`qcJ==&^N9U_UA(oMFVsWih02#OdiSuw%v*Zwr-zI%_;^*%-_vS8_=
zMein`gBKfw2U9+j%IShL!t=Z>wfWah7@5h}a#Qt0Hx;N+vabNQKM|IRog60*qM@*5
z{}avpf!&F!go~77IWi`C4A$$PZ#5)*HFe5e-AAq9JN!UJ=1z3XMVR5)L_RixE@a;=
zEqFcXx?J~)`Lx3K^Iz0s$rsw&n3{Cz-lRArN_><$`xUj8b<<75Oq9{I;if!rzg>+V
zH)^j2b{6M5`#Zb+n924TGFcu|YnW%(N!)8TBCr7F@Q-D$a2Lc{_m+4C#!U3#|AM<7
zv=mT)T*GGGNP%UBl9c*+Zu@pN*cn45r$&2rf0^`pXD+YL-0Fy4Na5e#^wLedRqLLp
ziOJ=*n3Q5sU$wK??k!$tcepJ%&ze=wiy}^cL2PaP*yW2-csRx(<~8f=omiuXc9rX)
zt0=DrQ66IeNig#yKM#ch-ev^t`F`<oqz!ce?hOqMr%j<`QhBz0()T-hVx0n_uj1HM
zmqbouS4Ar9e9bOUw`i-4>M#P7Id`I9XS{!6$*MJ}*XOW~9~!Ng<-?+0agDt`oE;rW
zOp;#saJNMHERPKD6|14lG;K4t;jJz8DMd2IGwsRo);jya3D09$iiY~s!BMgG+_#<O
zU%odIy!<7Z?%Ns(l_^xUn5OET9cfS8FvN%&7LP+nt%=_d&t^LSyE6vIFj4C3U0sXG
zs@l=KCekUrw^(M^lzXG1wlw{as`TgYV`lQO9W|*z=<yX^UUL1b9O))6HG?+irD{*i
zKta&lK;FP3_tH~cEtHbzSo7Ei2#6Vqgs=Jrg$VzSfF<n-8*jJ4(JKojOqr3`#wxA3
z-RPZJa9f@PZ-DpxyzlnfNZT6c>`2LwJGGml37m^EF*IKhNqpR$bNMWvD@Dsp=CVDP
zw*#kwl-8mO5>PomCKs<d_9MsM;-cl2U7lLhnlAkhD^AYW?{>?cEJQdrt??^n2surl
zImKV9i49)hA8`xVah-W}Zg02fm?XMD*<qI7AQ0apLNkLww{lOJRxy(8b(*Hy(Ej{F
z_xuDLH_C5rO{IF(nHFS)+f{8z4Uc4h715~es#F@`m3$k#`|_~<M8-`{^hCm%ZK-Uo
z0$*RgQ4nw)q;&GFnL|{lUNWFM6sufA)VL+13$5&b>DwFb0|NN{Jf?ff$l=hVSPox4
z@g}}SB*9@5fj0Z}21p=jZitme$ngqUEYwkfOqk0x5{Ozel#qG_@PhVfb{*go?yZT3
zGyO<h<N->8E7V3bZH^Xf&ox!z!E(EhsN<Js`)iAx?=qqEn~MEpBMX$_yHXSkH3b1n
z6V)zv$vFjqF4b5u_M>F9uTEZ6d^4i<R_G#`=<17y8$%B!OP*Svni<*>{q2L2AIAN`
zHo5_J=DxXy`mo2!Rc(Pqt}U09ue?0(KSnN8t%p!knYkH$0;5mmI=}JqrM4yIEUCxO
z6pO^9l_ioJ1ojup*w6a%?Hc^xOZU+DdTbt@H1y<YNJB=^$lJ-d$9!atVOx1g^`Va)
z#y6yoHt}$zXi1l@)3TKY=x$pI?0U2;#G3YR8ks%);A9cEF&g*Mesb<CX0Wj27!@Bw
z6RnL~?7Fze&OPm&&h$>_?PD~zwi9#i?>&s)bMY<wK)!Mai-;g@rJ!^XdA`MMUhouM
z$6v~s)V*M^UuoLceHs@JqS4u%>Dn3yHx8V{(DnHNCGfobNb|wBR0kKUbA;E$4Z<8I
z;C2SzNx%M1sy#`C8YyrYzB^r2V+%yZ(g%ySM?M0rQW4xumD@@C@fmv?z|zqB8cf5w
zzui;S7;ElKBXCawxDHH|1`l(dpW*s6h!#V+uQc%o)$_E7F!ty_csn;-F(8tq`1&%p
zO+T;TV)@Fs3jUx9B(4<XdiNCpJ^guH8w+v2vD)CMQ58<nr>i>;wfCQ0GqfTZ45k%6
zKh+%b`r#HauINSHUOCEMjz!lgrQvd~-3F;_auviFQr)?)muuR>v9&@$K$V^7<-ev1
zSmLW4Jfb9yAd)pQ?=IEYn_#CHgik>#^0wk8mU>!^5XYO%*Z~vtOSz~NE+4_6FB74I
zUncZmfelz^+&-aN4yH~B+Y8sZJCVW!>Q8cY>8oab4WdrgdTH9-0ywM6)!}TT8s%Po
zk8MZjC)Fsgs?H^w);P(@w^wC4M$=Jk?;}ZVfM434V=NcRYY|FFiQ~tYVPZ8<n4ujd
zbpMA8d1^`vd(q(Cg<Jd_UNEFv#xbr+2T_pw$y~Qv!;9lpnqPuW*Hrm3Hjp=hoFE>D
z5s3A|5D>xC0S}23an0R<nubi*p{$kHjJ*;`_O%dNcBBv^n9xy6mJo+DJDZ~Auie1y
zH2&XVy*0Q-R52E3KI|SX1B$9q>Q7#Yj-1)tw(5y@+Yd=jA}H*(*fy<BqBDRz-=Lf%
z6VP8t`?CrW;z|1LoKKn=m{4t!4&GJ>QTtktREj#KWN0;7#FnP7rX<IN0;R)z8VS88
ze`60N`^dO)GT+tY{TZes1lQZb>qfw|+c6J_W8ZVACcqx+*fcORDj;wi?nPo}|3Z_K
zh?MvBq!U8Kw(D>e0Ka+(#3Usk^(wDvjZ}IxRTfi8VA|z}(2Y7#80&Gbf^K?iz>S%B
zJGgkQx*H$eiZJkWZGBDVR)ue`lb8gBwIlAhr4Uwo0{JrAUlXW&8a~~SH5Zq5A>Upl
z;-yxmo}jcW8sC>neI~J`v-PJRnONOw2~~Fr<t=;gbT9TiT;?OVP;iLC=BaQ6Xqv@R
zL;iQmVMfHlJB}?M5ou8j#S~@Js(INuM0<ki#5tH%qvc7;eBVZ;ywl1SZ$a~C5q{!M
z<0IY(g8FuxKz>;MOs@7uv^hUer6%k8uH}{Fn`SVWd?D?A#Q7DpZdc42&%kbJpnT>Q
z1Qw2s^^DMZZW^)Zm*#8lr7)-JL}(>pD(}6^M3^oJ21(6t%X?I@Geh?l(P_3fmRk@l
zKv(s|sp2#=t``xno%2-FgQM#6Tk30YRAIo4KDiMFu*v=)IHz8uXAUFgi}x>N%_uVx
z_NOyj6ohMOSN!}+T}VsYyl0{9l&&PWR*gF}0`Rs?(l=F{Ttntqe_@F<ZncPyGnJNZ
zCsAfc=}<SL{GYo8qDBi`SIpWH3*59Ryxlf^wDvHPMBS1{sc`uqc8cq#Bb)pVjql-x
zNlDZpybaw?Do5~OS;Ph0&!fyi42%ep1#R7;0WELD!RJWPk)9cGwE98H`&8~)Yxw7|
zhBO4+Mhc`m@NKF37DD*y+9`bPqckYfXf3ZqveZwm<f5-N$<$gcxkWW4HXfG2tJDZh
zU%o=1w4V-cu|AZo(r)9jY^5kR(A64iPvy^$SMt&hIFG8*!J{lD(=+^#+~r91Roer#
z)SY^h)dG;fDCzT`a=z0p!a8;}&pg9&q!N+*+hsn4g~T2r3W8AG5=)iRLTa#fs^h%V
zE3tok^$7&NG_<nr*pP~6Q|D_N4MF`>69{jp0}7oTlfnPEET3nNRJPIQsE_*xyam%9
zfJXDq-Ha%lIjJ#3rn=$WfkWY_Rw75|(J3+mB$_byg_7ExUFO(_=Zl7%(MVtnSWlJm
z{+}vn5#m;>3701B4iZb+;Pl>TfkeHaMJ#of{w%V+G+Ca9sJ9{}w>6zZKR4w+p!OEq
zXvi8cSC#HMbPu9}llOeVD2g25<?YZ=SX*uX=yw$DN0ZyXuR`{)bxBP+AKoV_wG6h%
zsNK>0MQB(1xpPSAPY3&1U@kKP%FZ3YHEn-WlvgU^B^>J&W`r{&{DY&`_7-oElsrP=
ztH+Q!KmgVJRH1(XKs&Vomf$YIjX*W%n~+NiIRnZP2?5lP9JCqwfPdzvm2Y8CY8Q!)
zSQtzCsF$r2Vd1<uJQN+8s&xJ8^}Kb{4!?7S)_vyzRpB%45*=9OBx!RjG2d3&u;~~9
z*uY|;tqNA(h=L#AytJB$2UPM^d668z7cJ=U_4zz-_`+nLosJw?tD03;&#*6pdTT=o
zaap*~ze;iShL<1&0=hiZ1_aRj-%N=;ED>@<)t9e`Od3#b|8%_)?gq;CL&|CsA}&jD
zoiLBqynPy}pz>OyW;))T{#vAJG;q$zMN2f(njN>2{jEW^G8{JEs2LubXIgO>h4-b6
z4y)5`_DxrN>VOm3n3AG3f$rLX3Ex<pe_kc_U6)NM*%0vXZ1#A13DdJGw^Io>?|5#R
zYP?PURfB2Q@vwa-^*CCEBZC}CMKenF_kj`)ZU_KHh@fdmZW;K92$*VJ3A6U6z_E}0
z@hW6|1R@zc=kQ3X#CBITG%D5s1&^#~2fop_@*rydK>s^bnLyZ)$*kIQBX701o}B!6
z#v;zHYPT>&pd(DxX+Qrk4!%4T09Nr%YEZtK&@mD328U1byF0)E4Tb;A>Mcg?yCGtW
zo=Dw3Cwf~lB6?;Q4&nOa<BwN@2ggXDg?5eP@!)fwU_g!WKu|x(4%u)estu5BYH|S5
zOj~M>pU2vvR_nK6<6XV(3&gCbFfoKxw655Z2IMt+2&P4OE93B1IBreD*n)`%`xM0h
z&_5<H$+3J@BThZ~li4^KQXTh{7fGD(ILr&eLo$Kb|1E;kj#RpJrZ6qD^s*-Z&Vms%
ziA;@FTJAva9({k9|K#Cy)H9i9?g!pJ0tj0t0sNsebQR29B<4J=@E*vQkNb+!p;)S9
zgxt)s@=$=XUj~pLzu{5p8IlsUiq@a6n1jjuC2}El9kM1OFUIQxB`>}kYpsH6ew7Gg
ziH)vHyeBcy&HE!{geYH0_KBvw8)w{8S-G>U6bYv>QVx^Ju*0$hhF=k>%!ibY?u*a3
zKlXL96-RVjf^)Y5&XJ@*)q&eR0d$_19t_?OfWi}UKrA-@n)0jc{zb@1)?IF9xAT{1
zN6x&VSIbfyu=YxxDBQyCYlH}tVSO21&PPq;-l%uq2wR$Ig4=QL1&5>t9;cF|um(8F
zTRN{Ov$nS{W+Zv-LHYw)`owCs?}K4iW~Z$W!MbD|t)ktl?Ne5DNxy0)n%@pQvh{J#
z!nEJWaM8^+^bu^W782APun<V`iP!p?65)iYwAGPC+^#2Dg+oxMQS}LG_AoLC$|t2|
z4*66e^Y`TOI>2gzvi`BZ6-y2v=9_#1a{&IG(%8Pgg`CL660274`9P5Q&Bbuw3oEVU
z#moR>wW%{T>$=_81)wc<Ur~m|7?~8alrq*qF24pxvSSNx#+8o5Xsstnzp|iN8cE78
zp;Y))l_#eK%!aB|(t5=<M{zXTy2~Eekr!I%W+XtwEDAG1^o#dH$3(gty{`^rIt{Xk
z10UAp5|QUdHh?Gr%pva>NBQA}s+t5AsUdQqdqHOqD_9E!bCFmMg)M+!nSG)bym>c(
zy4m+g&23adAc9&Fp{^xBM0dGULoKpH8qI5QH$3<bRDHGx&$Fc8fJoA}ovjs9NGh0=
zxgmN%l1v5#1VU{ys=0P9OP6Hk*Z6J_X;P$hn|{YT51b@(-x8{;mT&yX@yTed!#~V~
zYK4dRugJQr(84NgrjEZ>9RZ83GTQl$kTRUjSKbw4^S7%tkfoT!IvB^^tC9ar1?L^)
zZbb%Wz%0e5d5%_+E0{?GPN3PASweC!J}FX!8c{TNcp^EP$S??mMGlBThl8VyCcows
zf&h{CR3o0_b5JxX8`KmQ4ZAhqT2XalOldx~rf4vezt?6!?-fD;>%0*(2WQte@eI;V
z(NV`m;mmOeCo5G7P6Qp7N3vTJ-s;?G#8$gKj=>7g(tn4!60P#|ZJbJkFtB59=1VC>
zh=QRu%=WA?{Zj*!O81Gtzp3*x@o<E{Mn|$*1n^0otm`HzpB*>^z`{wALEhW$kO9~9
z@u*u=3Pm7N=n_g&(`tN43J`cy>X^k(GKqxSH1LueZlJGPnXOxE@QB(tf@#hq(Qz>!
zoMK-6y(#g1E(-h7XidM*K<urDnnwv>PchZ80ZrzE+)7z3v{Ho6=osZ;aU{;=y6(Dh
z9eViPqkng55V9M$Vf7($CX2`2kO!0q)Q?bRwHW22K$hu9&4zNji5T6qa&>j}#y#`y
zr!n6zPZOq49h>oVpbFHESWh{6b-L1$DjidK|KmEkVNS5rPB(y*rqKuA&7yjJ)Q?ON
zd?a-od6FM|HvJ}^x1>c)HF2)$mr}V%_;K9mQY?luULB!z0F$JtoqzDbROM%cpdDRe
zcmV>35a$U}!}VzdYhzOgd%%Pu?5Mu<YoOj&Y3jBq3BX?hlBrnk^792w>t(JCnX;8;
z;Y8RZ85YUPKc))hk+QZ<0e!ve(L;<H8A<XOgeaAO0<8kvSh~(k{Q=Dei>}*d6t3S_
z(<rPU@|`YtNpd6)>@DqJH5J|qogUwP^-CY!+<4JJu9ku#Hpee_B9lZHO+IToKZ*pu
z4b;Rr=?)r-g}#29MhO6JCIk3|tBF`maNr=i+>?<hcjN9OA!8h&5G)G%q!bn~&)xrD
z#2LLQ@?|@ppm_)!-wvLnfV!7Rr1E*3>#-vQ0*ID}tvp#hB(Q{+J2arNoD;+>H(v*V
z$@>YExSwEN5X(Jv2h5!TruWIaRA*#-;bgfu{rL(Tf^{Ih53zSRb+#-^UA1w`fxj|2
zzh%7I2HcfrnVc1W%P2n%;IpoP;vis>gbaWY5^|lv_mS`=YMj;(mz4XO{nzavO~n&7
zg7H}6;p6|x9YDx*r)Bj@v4pOJC({Gfkvf*(Z$U=vafkiE0hWG=w>;&)KKa|fz*Hsq
zD=_)D&*?{fVQLURaYR%B{_6#QBg6jlkN@Xk|Lm{-(~bW&d;YggkOx!z`iBl3mRFFG
p)cm(+Vt@VL_Tc|7KCF5lCx^V)RMNtr;Sl_%Agd~qDP`#UzW^-Hm74$n

literal 0
HcmV?d00001

diff --git a/zh-cn/application-dev/napi/neural-network-runtime-guidelines.md b/zh-cn/application-dev/napi/neural-network-runtime-guidelines.md
new file mode 100644
index 0000000000..d1db02f381
--- /dev/null
+++ b/zh-cn/application-dev/napi/neural-network-runtime-guidelines.md
@@ -0,0 +1,491 @@
+# Neural Network Runtime对接AI推理框架开发指导
+
+## 场景介绍
+
+Neural Network Runtime作为AI推理引擎和加速芯片的桥梁，为AI推理引擎提供精简的Native接口，满足推理引擎通过加速芯片执行端到端推理的需求。
+
+本文以图1展示的`Add`单算子模型为例，介绍Neural Network Runtime的开发流程。`Add`算子包含两个输入、一个参数和一个输出，其中的`activation`参数用于指定`Add`算子中激活函数的类型。
+
+**图1** Add单算子网络示意图
+!["Add单算子网络示意图"](figures/02.png)
+
+## 环境准备
+
+### 环境要求
+
+Neural Network Runtime部件的环境要求如下：
+
+- 系统版本：OpenHarmony master分支。
+- 开发环境：Ubuntu 18.04及以上。
+- 接入设备：OpenHarmony定义的标准设备，并且系统中内置的硬件加速器驱动，已通过HDI接口对接Neural Network Runtime。
+
+由于Neural Network Runtime通过OpenHarmony Native API对外开放，需要通过OpenHarmony的Native开发套件编译Neural Network Runtime应用。在社区的[每日构建](http://ci.openharmony.cn/dailys/dailybuilds)下载对应系统版本的ohos-sdk压缩包，从压缩包中提取对应平台的Native开发套件。以Linux为例，Native开发套件的压缩包命名为`native-linux-{版本号}.zip`。
+
+### 环境搭建
+
+1. 打开Ubuntu编译服务器的终端。
+2. 把下载好的Native开发套件压缩包拷贝至当前用户根目录下。
+3. 执行以下命令解压Native开发套件的压缩包。
+```shell
+unzip native-linux-{版本号}.zip
+```
+
+解压缩后的内容如下（随版本迭代，目录下的内容可能发生变化，请以最新版本的Native API为准）：
+```text
+native/
+├── build // 交叉编译工具链
+├── build-tools // 编译构建工具
+├── docs
+├── llvm
+├── nativeapi_syscap_config.json
+├── ndk_system_capability.json
+├── NOTICE.txt
+├── oh-uni-package.json
+└── sysroot // Native API头文件和库
+```
+## 接口说明
+
+这里给出Neural Network Runtime开发流程中通用的接口，具体请见下列表格。
+
+### 结构体
+
+| 结构体名称 | 描述 |
+| --------- | ---- |
+| typedef struct OH_NNModel OH_NNModel | Neural Network Runtime的模型句柄，用于构造模型。 |
+| typedef struct OH_NNCompilation OH_NNCompilation | Neural Network Runtime的编译器句柄，用于编译AI模型。 |
+| typedef struct OH_NNExecutor OH_NNExecutor | Neural Network Runtime的执行器句柄，用于在指定设备上执行推理计算。 |
+
+### 模型构造相关接口
+
+| 接口名称 | 描述 |
+| ------- | --- |
+| OH_NNModel_Construct() | 创建OH_NNModel类型的模型实例。 |
+| OH_NN_ReturnCode OH_NNModel_AddTensor(OH_NNModel *model, const OH_NN_Tensor *tensor) | 向模型实例中添加张量。 |
+| OH_NN_ReturnCode OH_NNModel_SetTensorData(OH_NNModel *model, uint32_t index, const void *dataBuffer, size_t length) | 设置张量的数值。 |
+| OH_NN_ReturnCode OH_NNModel_AddOperation(OH_NNModel *model, OH_NN_OperationType op, const OH_NN_UInt32Array *paramIndices, const OH_NN_UInt32Array *inputIndices, const OH_NN_UInt32Array *outputIndices) | 向模型实例中添加算子。 |
+| OH_NN_ReturnCode OH_NNModel_SpecifyInputsAndOutputs(OH_NNModel *model, const OH_NN_UInt32Array *inputIndices, const OH_NN_UInt32Array *outputIndices) | 指定模型的输入输出。 |
+| OH_NN_ReturnCode OH_NNModel_Finish(OH_NNModel *model) | 完成模型构图。|
+| void OH_NNModel_Destroy(OH_NNModel **model) | 释放模型实例。 |
+
+### 模型编译相关接口
+
+| 接口名称 | 描述 |
+| ------- | --- |
+| OH_NNCompilation *OH_NNCompilation_Construct(const OH_NNModel *model) | 创建OH_NNCompilation类型的编译实例。 |
+| OH_NN_ReturnCode OH_NNCompilation_SetDevice(OH_NNCompilation *compilation, size_t deviceID) | 指定模型编译和计算的硬件。 |
+| OH_NN_ReturnCode OH_NNCompilation_SetCache(OH_NNCompilation *compilation, const char *cachePath, uint32_t version) | 设置编译后的模型缓存路径和缓存版本。 |
+| OH_NN_ReturnCode OH_NNCompilation_Build(OH_NNCompilation *compilation) | 进行模型编译。 |
+| void OH_NNCompilation_Destroy(OH_NNCompilation **compilation) | 释放OH_NNCompilation对象。 |
+
+### 执行推理相关接口
+
+| 接口名称 | 描述 |
+| ------- | --- |
+| OH_NNExecutor *OH_NNExecutor_Construct(OH_NNCompilation *compilation) | 创建OH_NNExecutor类型的执行器实例。 |
+| OH_NN_ReturnCode OH_NNExecutor_SetInput(OH_NNExecutor *executor, uint32_t inputIndex, const OH_NN_Tensor *tensor, const void *dataBuffer, size_t length) | 设置模型单个输入的数据。 |
+| OH_NN_ReturnCode OH_NNExecutor_SetOutput(OH_NNExecutor *executor, uint32_t outputIndex, void *dataBuffer, size_t length) | 设置模型单个输出的缓冲区。 |
+| OH_NN_ReturnCode OH_NNExecutor_Run(OH_NNExecutor *executor) | 执行推理。 |
+| void OH_NNExecutor_Destroy(OH_NNExecutor **executor) | 销毁OH_NNExecutor实例，释放实例占用的内存。 |
+
+### 设备管理相关接口
+
+| 接口名称 | 描述 |
+| ------- | --- |
+| OH_NN_ReturnCode OH_NNDevice_GetAllDevicesID(const size_t **allDevicesID, uint32_t *deviceCount) | 获取对接到 Neural Network Runtime 的硬件ID。 |
+
+
+## 开发步骤
+
+Neural Network Runtime的开发流程主要包含**模型构造**、**模型编译**和**推理执行**三个阶段。以下开发步骤以`Add`单算子模型为例，介绍调用Neural Network Runtime接口，开发应用的过程。
+
+1. 创建应用样例文件。
+
+    首先，创建Neural Network Runtime应用样例的源文件。在项目目录下执行以下命令，创建`nnrt_example/`目录，在目录下创建 `nnrt_example.cpp` 源文件。
+
+    ```shell
+    mkdir ~/nnrt_example && cd ~/nnrt_example
+    touch nnrt_example.cpp
+    ```
+
+2. 导入Neural Network Runtime。
+
+    在 `nnrt_example.cpp` 文件的开头添加以下代码，引入Neural Network Runtime模块。
+
+    ```cpp
+    #include <cstdint>
+    #include <iostream>
+    #include <vector>
+
+    #include "neural_network_runtime/neural_network_runtime.h"
+
+    // 常量，用于指定输入、输出数据的字节长度
+    const size_t DATA_LENGTH = 4 * 12;
+    ```
+
+3. 构造模型。
+
+    使用Neural Network Runtime接口，构造`Add`单算子样例模型。
+
+    ```cpp
+    OH_NN_ReturnCode BuildModel(OH_NNModel** pModel)
+    {
+        // 创建模型实例，进行模型构造
+        OH_NNModel* model = OH_NNModel_Construct();
+        if (model == nullptr) {
+            std::cout << "Create model failed." << std::endl;
+            return OH_NN_MEMORY_ERROR;
+        }
+
+        // 添加Add算子的第一个输入Tensor，类型为float32，张量形状为[1, 2, 2, 3]
+        int32_t inputDims[4] = {1, 2, 2, 3};
+        OH_NN_Tensor input1 = {OH_NN_FLOAT32, 4, inputDims, nullptr, OH_NN_TENSOR};
+        OH_NN_ReturnCode ret = OH_NNModel_AddTensor(model, &input1);
+        if (ret != OH_NN_SUCCESS) {
+            std::cout << "BuildModel failed, add Tensor of first input failed." << std::endl;
+            return ret;
+        }
+
+        // 添加Add算子的第二个输入Tensor，类型为float32，张量形状为[1, 2, 2, 3]
+        OH_NN_Tensor input2 = {OH_NN_FLOAT32, 4, inputDims, nullptr, OH_NN_TENSOR};
+        ret = OH_NNModel_AddTensor(model, &input2);
+        if (ret != OH_NN_SUCCESS) {
+            std::cout << "BuildModel failed, add Tensor of second input failed." << std::endl;
+            return ret;
+        }
+
+        // 添加Add算子的参数Tensor，该参数Tensor用于指定激活函数的类型，Tensor的数据类型为int8。
+        int32_t activationDims = 1;
+        int8_t activationValue = OH_NN_FUSED_NONE;
+        OH_NN_Tensor activation = {OH_NN_INT8, 1, &activationDims, nullptr, OH_NN_ADD_ACTIVATIONTYPE};
+        ret = OH_NNModel_AddTensor(model, &activation);
+        if (ret != OH_NN_SUCCESS) {
+            std::cout << "BuildModel failed, add Tensor of activation failed." << std::endl;
+            return ret;
+        }
+
+        // 将激活函数类型设置为OH_NN_FUSED_NONE，表示该算子不添加激活函数。
+        ret = OH_NNModel_SetTensorData(model, 2, &activationValue, sizeof(int8_t));
+        if (ret != OH_NN_SUCCESS) {
+            std::cout << "BuildModel failed, set value of activation failed." << std::endl;
+            return ret;
+        }
+
+        // 设置Add算子的输出，类型为float32，张量形状为[1, 2, 2, 3]
+        OH_NN_Tensor output = {OH_NN_FLOAT32, 4, inputDims, nullptr, OH_NN_TENSOR};
+        ret = OH_NNModel_AddTensor(model, &output);
+        if (ret != OH_NN_SUCCESS) {
+            std::cout << "BuildModel failed, add Tensor of output failed." << std::endl;
+            return ret;
+        }
+
+        // 指定Add算子的输入、参数和输出索引
+        uint32_t inputIndicesValues[2] = {0, 1};
+        uint32_t paramIndicesValues = 2;
+        uint32_t outputIndicesValues = 3;
+        OH_NN_UInt32Array paramIndices = {&paramIndicesValues, 1};
+        OH_NN_UInt32Array inputIndices = {inputIndicesValues, 2};
+        OH_NN_UInt32Array outputIndices = {&outputIndicesValues, 1};
+
+        // 向模型实例添加Add算子
+        ret = OH_NNModel_AddOperation(model, OH_NN_OPS_ADD, &paramIndices, &inputIndices, &outputIndices);
+        if (ret != OH_NN_SUCCESS) {
+            std::cout << "BuildModel failed, add operation failed." << std::endl;
+            return ret;
+        }
+
+        // 设置模型实例的输入、输出索引
+        ret = OH_NNModel_SpecifyInputsAndOutputs(model, &inputIndices, &outputIndices);
+        if (ret != OH_NN_SUCCESS) {
+            std::cout << "BuildModel failed, specify inputs and outputs failed." << std::endl;
+            return ret;
+        }
+
+        // 完成模型实例的构建
+        ret = OH_NNModel_Finish(model);
+        if (ret != OH_NN_SUCCESS) {
+            std::cout << "BuildModel failed, error happened when finishing model construction." << std::endl;
+            return ret;
+        }
+
+        *pModel = model;
+        return OH_NN_SUCCESS;
+    }
+    ```
+
+4. 查询Neural Network Runtime已经对接的加速芯片。
+
+    Neural Network Runtime支持通过HDI接口，对接多种加速芯片。在执行模型编译前，需要查询当前设备下，Neural Network Runtime已经对接的加速芯片。每个加速芯片对应唯一的ID值，在编译阶段需要通过设备ID，指定模型编译的芯片。
+    ```cpp
+    void GetAvailableDevices(std::vector<size_t>& availableDevice)
+    {
+        availableDevice.clear();
+
+        // 获取可用的硬件ID
+        const size_t* devices = nullptr;
+        uint32_t deviceCount = 0;
+        OH_NN_ReturnCode ret = OH_NNDevice_GetAllDevicesID(&devices, &deviceCount);
+        if (ret != OH_NN_SUCCESS) {
+            std::cout << "GetAllDevicesID failed, get no available device." << std::endl;
+            return;
+        }
+
+        for (uint32_t i = 0; i < deviceCount; i++) {
+            availableDevice.emplace_back(devices[i]);
+        }
+    }
+    ```
+
+5. 在指定的设备上编译模型。
+
+    Neural Network Runtime使用抽象的模型表达描述AI模型的拓扑结构，在加速芯片上执行前，需要通过Neural Network Runtime提供的编译模块，将抽象的模型表达下发至芯片驱动层，转换成可以直接推理计算的格式。
+    ```cpp
+    OH_NN_ReturnCode CreateCompilation(OH_NNModel* model, const std::vector<size_t>& availableDevice, OH_NNCompilation** pCompilation)
+    {
+        // 创建编译实例，用于将模型传递至底层硬件编译
+        OH_NNCompilation* compilation = OH_NNCompilation_Construct(model);
+        if (compilation == nullptr) {
+            std::cout << "CreateCompilation failed, error happended when creating compilation." << std::endl;
+            return OH_NN_MEMORY_ERROR;
+        }
+
+        // 设置编译的硬件、缓存路径、性能模式、计算优先级、是否开启float16低精度计算等选项
+
+        // 选择在第一个设备上编译模型
+        OH_NN_ReturnCode ret = OH_NNCompilation_SetDevice(compilation, availableDevice[0]);
+        if (ret != OH_NN_SUCCESS) {
+            std::cout << "CreateCompilation failed, error happened when setting device." << std::endl;
+            return ret;
+        }
+
+        // 将模型编译结果缓存在/data/local/tmp目录下，版本号指定为1
+        ret = OH_NNCompilation_SetCache(compilation, "/data/local/tmp", 1);
+        if (ret != OH_NN_SUCCESS) {
+            std::cout << "CreateCompilation failed, error happened when setting cache path." << std::endl;
+            return ret;
+        }
+
+        // 完成编译设置，进行模型编译
+        ret = OH_NNCompilation_Build(compilation);
+        if (ret != OH_NN_SUCCESS) {
+            std::cout << "CreateCompilation failed, error happened when building compilation." << std::endl;
+            return ret;
+        }
+
+        *pCompilation = compilation;
+        return OH_NN_SUCCESS;
+    }
+    ```
+
+6. 创建执行器。
+
+    完成模型编译后，需要调用Neural Network Runtime的执行模块，创建推理执行器。执行阶段，设置模型输入、获取模型输出和触发推理计算的操作均围绕执行器完成。
+    ```cpp
+    OH_NNExecutor* CreateExecutor(OH_NNCompilation* compilation)
+    {
+        // 创建执行实例
+        OH_NNExecutor* executor = OH_NNExecutor_Construct(compilation);
+        return executor;
+    }
+    ```
+
+7. 执行推理计算，并打印计算结果。
+
+    通过执行模块提供的接口，将推理计算所需要的输入数据传递给执行器，触发执行器完成一次推理计算，获取模型的推理计算结果。
+    ```cpp
+    OH_NN_ReturnCode Run(OH_NNExecutor* executor)
+    {
+        // 构造示例数据
+        float input1[12] = {0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11};
+        float input2[12] = {11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22};
+
+        int32_t inputDims[4] = {1, 2, 2, 3};
+        OH_NN_Tensor inputTensor1 = {OH_NN_FLOAT32, 4, inputDims, nullptr, OH_NN_TENSOR};
+        OH_NN_Tensor inputTensor2 = {OH_NN_FLOAT32, 4, inputDims, nullptr, OH_NN_TENSOR};
+
+        // 设置执行的输入
+
+        // 设置执行的第一个输入，输入数据由input1指定
+        OH_NN_ReturnCode ret = OH_NNExecutor_SetInput(executor, 0, &inputTensor1, input1, DATA_LENGTH);
+        if (ret != OH_NN_SUCCESS) {
+            std::cout << "Run failed, error happened when setting first input." << std::endl;
+            return ret;
+        }
+
+        // 设置执行的第二个输入，输入数据由input2指定
+        ret = OH_NNExecutor_SetInput(executor, 1, &inputTensor2, input2, DATA_LENGTH);
+        if (ret != OH_NN_SUCCESS) {
+            std::cout << "Run failed, error happened when setting second input." << std::endl;
+            return ret;
+        }
+
+        // 设置输出的数据缓冲区，OH_NNExecutor_Run执行计算后，输出结果将保留在output中
+        float output[12];
+        ret = OH_NNExecutor_SetOutput(executor, 0, output, DATA_LENGTH);
+        if (ret != OH_NN_SUCCESS) {
+            std::cout << "Run failed, error happened when setting output buffer." << std::endl;
+            return ret;
+        }
+
+        // 执行计算
+        ret = OH_NNExecutor_Run(executor);
+        if (ret != OH_NN_SUCCESS) {
+            std::cout << "Run failed, error doing execution." << std::endl;
+            return ret;
+        }
+
+        // 打印输出结果
+        for (uint32_t i = 0; i < 12; i++) {
+            std::cout << "Output index: " << i << ", value is: " << output[i] << "." << std::endl;
+        }
+
+        return OH_NN_SUCCESS;
+    }
+    ```
+
+8. 构建端到端模型构造-编译-执行流程。
+
+    步骤3-步骤7实现了模型的模型构造、编译和执行流程，并封装成4个函数，便于模块化开发。以下示例代码将4个函数串联成完整的Neural Network Runtime开发流程。
+    ```cpp
+    int main()
+    {
+        OH_NNModel* model = nullptr;
+        OH_NNCompilation* compilation = nullptr;
+        OH_NNExecutor* executor = nullptr;
+        std::vector<size_t> availableDevices;
+
+        // 模型构造阶段
+        OH_NN_ReturnCode ret = BuildModel(&model);
+        if (ret != OH_NN_SUCCESS) {
+            std::cout << "BuildModel failed." << std::endl;
+            OH_NNModel_Destroy(&model);
+            return -1;
+        }
+
+        // 获取可执行的设备
+        GetAvailableDevices(availableDevices);
+        if (availableDevices.empty()) {
+            std::cout << "No available device." << std::endl;
+            OH_NNModel_Destroy(&model);
+            return -1;
+        }
+
+        // 模型编译阶段
+        ret = CreateCompilation(model, availableDevices, &compilation);
+        if (ret != OH_NN_SUCCESS) {
+            std::cout << "CreateCompilation failed." << std::endl;
+            OH_NNModel_Destroy(&model);
+            OH_NNCompilation_Destroy(&compilation);
+            return -1;
+        }
+
+        // 创建模型的推理执行器
+        executor = CreateExecutor(compilation);
+        if (executor == nullptr) {
+            std::cout << "CreateExecutor failed, no executor is created." << std::endl;
+            OH_NNModel_Destroy(&model);
+            OH_NNCompilation_Destroy(&compilation);
+            return -1;
+        }
+
+        // 使用上一步创建的执行器，执行单步推理计算
+        ret = Run(executor);
+        if (ret != OH_NN_SUCCESS) {
+            std::cout << "Run failed." << std::endl;
+            OH_NNModel_Destroy(&model);
+            OH_NNCompilation_Destroy(&compilation);
+            OH_NNExecutor_Destroy(&executor);
+            return -1;
+        }
+
+        // 释放申请的资源
+        OH_NNModel_Destroy(&model);
+        OH_NNCompilation_Destroy(&compilation);
+        OH_NNExecutor_Destroy(&executor);
+
+        return 0;
+    }
+    ```
+
+## 调测验证
+
+1. 准备应用样例的编译配置文件。
+
+    新建一个 `CMakeLists.txt` 文件，为开发步骤中的应用样例文件 `nnrt_example.cpp` 添加编译配置。以下提供简单的 `CMakeLists.txt` 示例：
+    ```text
+    cmake_minimum_required(VERSION 3.16)
+    project(nnrt_example C CXX)
+
+    add_executable(nnrt_example
+        ./nnrt_example.cpp
+    )
+
+    target_link_libraries(nnrt_example
+        neural_network_runtime.z
+    )
+    ```
+
+2. 编译应用样例。
+
+    执行以下命令，在当前目录下新建build/目录，在build/目录下编译 `nnrt_example.cpp`，得到二进制文件 `nnrt_example`。
+    ```shell
+    mkdir build && cd build
+    cmake -DCMAKE_TOOLCHAIN_FILE={交叉编译工具链的路径}/build/cmake/ohos.toolchain.cmake -DOHOS_ARCH=arm64-v8a -DOHOS_PLATFORM=OHOS -DOHOS_STL=c++_static ..
+    make
+    ```
+
+3. 执行以下代码，将样例推送到设备上执行。
+    ```shell
+    # 将编译得到的 `nnrt_example` 推送到设备上，执行样例。
+    hdc_std file send ./nnrt_example /data/local/tmp/.
+
+    # 给测试用例可执行文件加上权限。
+    hdc_std shell "chmod +x /data/local/tmp/nnrt_example"
+
+    # 执行测试用例
+    hdc_std shell "/data/local/tmp/nnrt_example"
+    ```
+
+    如果样例执行正常，应该得到以下输出。
+    ```text
+    Output index: 0, value is: 11.000000.
+    Output index: 1, value is: 13.000000.
+    Output index: 2, value is: 15.000000.
+    Output index: 3, value is: 17.000000.
+    Output index: 4, value is: 19.000000.
+    Output index: 5, value is: 21.000000.
+    Output index: 6, value is: 23.000000.
+    Output index: 7, value is: 25.000000.
+    Output index: 8, value is: 27.000000.
+    Output index: 9, value is: 29.000000.
+    Output index: 10, value is: 31.000000.
+    Output index: 11, value is: 33.000000.
+    ```
+
+4. 检查模型缓存（可选）。
+
+    如果在调测环境下，Neural Network Runtime对接的HDI服务支持模型缓存功能，执行完 `nnrt_example`, 可以在 `/data/local/tmp` 目录下找到生成的缓存文件。
+
+    > **说明：** 
+    >
+    > 模型的IR需要传递到硬件驱动层，由HDI服务将统一的IR图，编译成硬件专用的计算图，编译的过程非常耗时。Neural Network Runtime支持计算图缓存的特性，可以将HDI服务编译生成的计算图，缓存到设备存储中。当下一次在同一个加速芯片上编译同一个模型时，通过指定缓存的路径，Neural Network Runtime可以直接加载缓存文件中的计算图，减少编译消耗的时间。
+
+    检查缓存目录下的缓存文件：
+    ```shell
+    ls /data/local/tmp
+    ```
+
+    以下为打印结果：
+    ```text
+    # 0.nncache  cache_info.nncache
+    ```
+
+    如果缓存不再使用，需要手动删除缓存，可以参考以下命令，删除缓存文件。
+    ```shell
+    rm /data/local/tmp/*nncache
+    ```
+
+## 相关实例
+
+第三方AI推理框架对接Neural Network Runtime的流程，可以参考以下相关实例：
+- [Tensorflow Lite接入NNRt Delegate开发指南](https://gitee.com/openharmony/neural_network_runtime/tree/master/example/deep_learning_framework)
+<!--no_check-->
-- 
GitLab