From 315d19be8ef473a864950ab497a649a69e37c6a4 Mon Sep 17 00:00:00 2001 From: Darwin Bautista Date: Fri, 15 Jul 2022 08:33:04 +0800 Subject: [PATCH] Initial commit for public release. --- .github/contexts-example.png | Bin 0 -> 135256 bytes .github/gh-teaser.png | Bin 0 -> 102483 bytes .github/system.png | Bin 0 -> 149721 bytes .gitignore | 147 ++++++++++ Datasets.md | 92 +++++++ LICENSE | 202 ++++++++++++++ NOTICE | 18 ++ README.md | 193 +++++++++++++ bench.py | 57 ++++ configs/bench.yaml | 10 + configs/charset/36_lowercase.yaml | 3 + configs/charset/62_mixed-case.yaml | 3 + configs/charset/94_full.yaml | 3 + configs/experiment/abinet-sv.yaml | 8 + configs/experiment/abinet.yaml | 3 + configs/experiment/crnn.yaml | 6 + configs/experiment/parseq-tiny.yaml | 6 + configs/experiment/parseq.yaml | 3 + configs/experiment/trba.yaml | 6 + configs/experiment/trbc.yaml | 11 + configs/experiment/tune_abinet-lm.yaml | 17 ++ configs/experiment/vitstr.yaml | 7 + configs/main.yaml | 47 ++++ configs/model/abinet.yaml | 26 ++ configs/model/crnn.yaml | 9 + configs/model/parseq.yaml | 25 ++ configs/model/trba.yaml | 10 + configs/model/vitstr.yaml | 13 + configs/train_set/real.yaml | 3 + configs/train_set/synth.yaml | 7 + configs/tune.yaml | 10 + demo_images/art-01107.jpg | Bin 0 -> 182385 bytes demo_images/coco-1166773.jpg | Bin 0 -> 3398 bytes demo_images/cute-184.jpg | Bin 0 -> 18355 bytes demo_images/ic13_word_256.png | Bin 0 -> 59337 bytes demo_images/ic15_word_26.png | Bin 0 -> 12271 bytes demo_images/uber-27491.jpg | Bin 0 -> 1278 bytes hubconf.py | 130 +++++++++ read.py | 49 ++++ requirements.txt | 15 + setup.cfg | 43 +++ setup.py | 14 + strhub/__init__.py | 0 strhub/data/__init__.py | 0 strhub/data/aa_overrides.py | 46 ++++ strhub/data/augment.py | 111 ++++++++ strhub/data/dataset.py | 126 +++++++++ strhub/data/module.py | 102 +++++++ strhub/data/utils.py | 145 ++++++++++ strhub/models/__init__.py | 0 strhub/models/abinet/LICENSE | 25 ++ strhub/models/abinet/__init__.py | 13 + strhub/models/abinet/attention.py | 100 +++++++ strhub/models/abinet/backbone.py | 24 ++ strhub/models/abinet/model.py | 31 +++ strhub/models/abinet/model_abinet_iter.py | 39 +++ strhub/models/abinet/model_alignment.py | 28 ++ strhub/models/abinet/model_language.py | 50 ++++ strhub/models/abinet/model_vision.py | 45 +++ strhub/models/abinet/resnet.py | 72 +++++ strhub/models/abinet/system.py | 170 ++++++++++++ strhub/models/abinet/transformer.py | 143 ++++++++++ strhub/models/base.py | 219 +++++++++++++++ strhub/models/crnn/LICENSE | 21 ++ strhub/models/crnn/__init__.py | 13 + strhub/models/crnn/model.py | 62 +++++ strhub/models/crnn/system.py | 43 +++ strhub/models/modules.py | 20 ++ strhub/models/parseq/__init__.py | 0 strhub/models/parseq/modules.py | 133 +++++++++ strhub/models/parseq/system.py | 258 ++++++++++++++++++ strhub/models/trba/__init__.py | 13 + strhub/models/trba/feature_extraction.py | 110 ++++++++ strhub/models/trba/model.py | 55 ++++ strhub/models/trba/prediction.py | 73 +++++ strhub/models/trba/system.py | 87 ++++++ strhub/models/trba/transformation.py | 169 ++++++++++++ strhub/models/utils.py | 78 ++++++ strhub/models/vitstr/__init__.py | 12 + strhub/models/vitstr/model.py | 53 ++++ strhub/models/vitstr/system.py | 58 ++++ test.py | 129 +++++++++ tools/art_converter.py | 26 ++ .../case_sensitive_str_datasets_converter.py | 28 ++ tools/coco_2_converter.py | 126 +++++++++ tools/coco_text_converter.py | 15 + tools/create_lmdb_dataset.py | 78 ++++++ tools/filter_lmdb.py | 57 ++++ tools/lsvt_converter.py | 107 ++++++++ tools/mlt19_converter.py | 15 + tools/openvino_converter.py | 116 ++++++++ tools/test_abinet_lm_acc.py | 101 +++++++ tools/textocr_converter.py | 142 ++++++++++ train.py | 73 +++++ tune.py | 194 +++++++++++++ 95 files changed, 5150 insertions(+) create mode 100644 .github/contexts-example.png create mode 100644 .github/gh-teaser.png create mode 100644 .github/system.png create mode 100644 .gitignore create mode 100644 Datasets.md create mode 100644 LICENSE create mode 100644 NOTICE create mode 100644 README.md create mode 100755 bench.py create mode 100644 configs/bench.yaml create mode 100644 configs/charset/36_lowercase.yaml create mode 100644 configs/charset/62_mixed-case.yaml create mode 100644 configs/charset/94_full.yaml create mode 100644 configs/experiment/abinet-sv.yaml create mode 100644 configs/experiment/abinet.yaml create mode 100644 configs/experiment/crnn.yaml create mode 100644 configs/experiment/parseq-tiny.yaml create mode 100644 configs/experiment/parseq.yaml create mode 100644 configs/experiment/trba.yaml create mode 100644 configs/experiment/trbc.yaml create mode 100644 configs/experiment/tune_abinet-lm.yaml create mode 100644 configs/experiment/vitstr.yaml create mode 100644 configs/main.yaml create mode 100644 configs/model/abinet.yaml create mode 100644 configs/model/crnn.yaml create mode 100644 configs/model/parseq.yaml create mode 100644 configs/model/trba.yaml create mode 100644 configs/model/vitstr.yaml create mode 100644 configs/train_set/real.yaml create mode 100644 configs/train_set/synth.yaml create mode 100644 configs/tune.yaml create mode 100644 demo_images/art-01107.jpg create mode 100644 demo_images/coco-1166773.jpg create mode 100644 demo_images/cute-184.jpg create mode 100644 demo_images/ic13_word_256.png create mode 100644 demo_images/ic15_word_26.png create mode 100644 demo_images/uber-27491.jpg create mode 100644 hubconf.py create mode 100755 read.py create mode 100644 requirements.txt create mode 100644 setup.cfg create mode 100644 setup.py create mode 100644 strhub/__init__.py create mode 100644 strhub/data/__init__.py create mode 100644 strhub/data/aa_overrides.py create mode 100644 strhub/data/augment.py create mode 100644 strhub/data/dataset.py create mode 100644 strhub/data/module.py create mode 100644 strhub/data/utils.py create mode 100644 strhub/models/__init__.py create mode 100644 strhub/models/abinet/LICENSE create mode 100644 strhub/models/abinet/__init__.py create mode 100644 strhub/models/abinet/attention.py create mode 100644 strhub/models/abinet/backbone.py create mode 100644 strhub/models/abinet/model.py create mode 100644 strhub/models/abinet/model_abinet_iter.py create mode 100644 strhub/models/abinet/model_alignment.py create mode 100644 strhub/models/abinet/model_language.py create mode 100644 strhub/models/abinet/model_vision.py create mode 100644 strhub/models/abinet/resnet.py create mode 100644 strhub/models/abinet/system.py create mode 100644 strhub/models/abinet/transformer.py create mode 100644 strhub/models/base.py create mode 100644 strhub/models/crnn/LICENSE create mode 100644 strhub/models/crnn/__init__.py create mode 100644 strhub/models/crnn/model.py create mode 100644 strhub/models/crnn/system.py create mode 100644 strhub/models/modules.py create mode 100644 strhub/models/parseq/__init__.py create mode 100644 strhub/models/parseq/modules.py create mode 100644 strhub/models/parseq/system.py create mode 100644 strhub/models/trba/__init__.py create mode 100644 strhub/models/trba/feature_extraction.py create mode 100644 strhub/models/trba/model.py create mode 100644 strhub/models/trba/prediction.py create mode 100644 strhub/models/trba/system.py create mode 100644 strhub/models/trba/transformation.py create mode 100644 strhub/models/utils.py create mode 100644 strhub/models/vitstr/__init__.py create mode 100644 strhub/models/vitstr/model.py create mode 100644 strhub/models/vitstr/system.py create mode 100755 test.py create mode 100755 tools/art_converter.py create mode 100755 tools/case_sensitive_str_datasets_converter.py create mode 100755 tools/coco_2_converter.py create mode 100755 tools/coco_text_converter.py create mode 100755 tools/create_lmdb_dataset.py create mode 100755 tools/filter_lmdb.py create mode 100755 tools/lsvt_converter.py create mode 100755 tools/mlt19_converter.py create mode 100755 tools/openvino_converter.py create mode 100755 tools/test_abinet_lm_acc.py create mode 100755 tools/textocr_converter.py create mode 100755 train.py create mode 100755 tune.py diff --git a/.github/contexts-example.png b/.github/contexts-example.png new file mode 100644 index 0000000000000000000000000000000000000000..e2ae2da8f3944dae82031a3c3bd614f3f43485ae GIT binary patch literal 135256 zcmeFYXIPV2w=hgEp@-gk7f7ff^Z=pvA{`QX4?R+(cL*SgD4kdUsfrXq3>^d&q>6|X z5k#bnAPfq;ftlx-nRCvc^IhNh_v8xUzV}{zt-aRXD|a%@P4#F18~_3W0vZE-9ZLcN zGCcwU!cz(|{79wXqn`u>j8xIqcHx$YC_YqZu$OOuCtvtAlqa7j+SiMK0KIVk?$uD? z&Pd&JFo1?1K|79e6O<>WarGCJc{JASTI1-!94sq^8dubRcB*k6w-~>F%Yjd+C&%keEz}WQ4^YI`P)qKW;87G-{$Fx3n;N;fKK0*!up#zC*rl1+8?Tg z<;HC59iEiCo4yNuElkA}lPTPZ6uxcg=X-z7&g$u1|JtW*3ywYv6Rky#zE7Xds=V1= z>dUeOx~9v%ObY;16b7r{&?7kLsT-Y`qUQ^nDnpvitaJ zP#3S^H=U@rErn*vhpkib9^VRQHG0mqW`BwLuXq`X7V+A<2aCJ5?z1C;DvO0>k z0Elf?ma5G>WV)%mzU`|;!OC_|+9UL5A&{(ojkKBrJ>Q7@8dyFz|}X{4Pv*%Bcejz zM{dsKs4IVx<>i^5-=Uk5^_MG(T`Y|mmw(CSOYGWETD`OmkLcTqItP$q>b9%|q zz?)8FWVh|8rB%<^LyeWkD9KH-J?St4S`)d7*_ChjZLF;B?V-|7mHl1ponYf9%z8`! z?ei$5y7xCeTE7y2`_ryhGi5ZCa&@Z3olG;wGU&Bss>7`-EhoK=h;e? zWgBFmI1!zC3WsLE)g0Z*bK;ra$F|?Y^_;g(MPKRnOw##rHPBIUHb$0 zy$vJ>f?>u2@m{tZG(yKbCx!{nc+44i3!3%~GAx#oH#R!qI4L;zx+Hm+1Z(;~b++dV$4i_Hs@Je^c7%zg5_##KJNVs;~~$w?5t*$q>h zie2O*iCE3>Gl!cb+wyet2oyy(_}!LdBCoxJK+QfcTf2TwU`p_g_ioGVPT)Halyg~& z&&c5Yr$k$YJA%XI8Ar;zR@S{K+tGX`<(9FJLbOktw!R|RgiZ)j zM|mA^)~l(25=-h}MDe{^rAbxZN7hLa ze(RtgBuCo*uA{fjWjH5!`Zvu7uf=Ne>?J=i`4>Xmbx=<=g(XHx;=kT&d<(2E-O0ka z_@rv?xGCM*>gqM`g{WEknbz8}U7U9FBT9y}ke4QwshVd&`8A zdB=Jqr|bJRsUiCcy%9rz#|_X>(|h{sKHkp^StRbaZ#HXMBU5a&Y{lFe2s8>J`V>CB zg>@gS6Wv(bZFvu)R(4`8`F5E)h8sE_$Wv~i%=Zh~ zY>X^v;wIF7jw!FV+BVb(xSpLTL7cv|8}s4A2TJ)728B~cvj^LB{xZ_nUw&m_o!zzy zf3Xt$8zev-01VS6d6- z{(_np*KL4Lsh&*Cc(T_rLui*s>?So6hd+}Lbkf;RQ~o}6JBMRieK&WwT9Ejaa=&iH z-S5yS%SA3B!k{j)8A6f`-(;7*0Y&Ni)GAcb)W~thul)N$zQ+@ z1riG;%hQM+O-(I->8G#mF(1?4&5Cs*5D0xIxXBD1$Pbp=IMJcJ#hdIMOTE;PExJ3g zOWiXRU-13mzQ*0&%3KjyA8r2$Q&$nO3JO`h?BqRc-q&LGpI2(QPn5M)@0=p`Z3w)j zsc)Dklp3Jik_g)>35T9LtJVnf^Fq}$EEiiLuVhnM>Z>5_@_MX;w z>)d3sU&Az*N^3fPKD-%q;zA`X+DTE=PH9i<@~TiToPv^cRB&fye{(FonwPPbd`4kW z{FUQetl==!Ld$s3DxcTDruLpR^!{R^hy>V7mBQV0go)oa*FmdBlepKIc*?<`!^Wv> zgQrpan7J7)4Z8o1uyQ7q)xTY zL>i5pegYbZTc0_Kj{czHG~TqmHV9i3FHW4gM~dpZFKwv#;FiS((5gzxcSDynO0-v( zIi^|hxaGYnmo$kgB6P*CPCr}B&d*wS^w!bmDvj$XQKGwAFyuE|+;bM@aCWh`}|w?B=r8gdpP6$xn+g+izdz8Z-1 z5U)`LkbpGFGLVosR5W3mK*R&$jyb)vR30JH0ND) zamM4t%H&yHDnkixg*S2%s#_;~Bb0|Uge#D{^2tSJ?H(mHyes;wK?})27#(LS5gGBb z@ys3VtS`Y@<AMI79Zla0BNat*mTfeC$KmqO!D#5e3h21s#z<~?VxEN z1;SRM$n)Y*j11@UH9xkz>%`{QDWp8>&bwPdGvlLF&;VaL=XBGt;w!&WM?`!)nY#5z zKeuq+~g1yKmFU#J<_@=A;`~xLI zOBFZ%3=Sfl$Gk(I&5ha#BzhuOc{A3>QPXyA8vKu_nJHN$*6N~*SlAAFnv(B^(iHYd;lfGy&t9g-7C0RB@xGG zIqQikMvWp}8Ld9PqW#fNs&_yu?lSOdW!j&fTibWc#h6w`+~t`AVOf#^_wgrki9SL9 zcY2#APOKWWp%9O%u82FFVtc>t-p#mw?HY?o)~U51@$wDbYS{)Y-#BZM=1-ezfIXR` z`tf(U1%xZsWYI;>7&5C}4>{!ew9A+h{1T&+zO}5~PaPC@}i8R|GJDkZ{eocBs$dSL1r0CK?rsagIxp9vt1-i!8W zEa~DkAmP;Ty}1`EV|I39tVyS?@79P?E@Ouw0vtM$0t-|aSB2yJbu?s`Unm$ytC^5& zbggP(*Oo(*)3Ow#rc@Jk@}p~rM$^2a<#7jB%Q}MrG-y~$5OwzCI;RYb0bAK8=vz6X zS&$J|WJmQYf=8-FfX>%^C>TutRA2Z8d4AwalE{`&y|qnn^ukwL!n2qF6ikSlkE20`_j`!+BtfFn<3#-=2Z zG-^hcEveD3>ip1v^=jAFhnOwGZe13BW~#zAFRGn~S3C_^53ESramuUefh$)Ey=H~( zPsu&F&4?al9pC#IrA$8hd!u1x9hx9S!(p=T^wPvF77MQm)05!D?Znf>kcqNF=$+p?$ZuC&mCPV+ zrJq!%T1Xcld3f)${^&>2FZb4C*{E-F^J{xIcn-QR1wXv5*1UNonudh#i^#p3nr{yX z?`vhzZ#}Xo-n7ye<_1%F#%#n!>O~4|bS-_~Y+&o-umY$&NFg0)9nDz2@^Ij$t$l%a zbXUB$FQMMX90{j~?xfx|q0_{)44J;@6XN8sy>8v0G+75FqpOkr4H@D+V{CaO#?EJ7 zYkvON`@Sj6MfQMIUYA5$f^@VjCe1+X+Q^fI8co&tRL zX)3DTcTxX73A>pAASh|9lc0{!J(X=B==Ahhbw1RtwIVbTzfaM;6w%X6Dt{w%ra8_c zsjGC4@?E;QdU&;x{h9aieWM0}qHY(Ynv0RTO?FjlOf9X`5{xzahj7h#Y<~rtA$)&b zOxeIYYAw4)=l~pevXP|Gr-2k~-0ZHh6>*qJVyq5|9s5{0xJkdSg z>@59syRw!%`)s*w@PWXZNHtN_Z>cQMq4J+wkwhp=5<}K0+*X|w_Ws@B=IPejitb~z zp!&Wf3eZo_oHQbFm!;7UH;Nw7YyqB!zYQgRb}wb@>TJR6d?61isk6^x-av58D^`WdFZ@S^qdDSxBu0_G;@GGtvjAHEw{*Q&L}OF z5GhS}!@^XU&%d<|cdxxz*AYXrEvvn`u*lisb*I9bsf02`Bw9*p!W34A=45} zRj?XnhX~NMFTC=;`j$>yq;pBh8re4FtlrMtPNZ6U8T)-}aivjj zR-XSPsIvFGJ0nNy^DNNI(g1whm07!qf>ri2o{S4i{C8R`T{q@A>jkK*Q>%l-o=OM9*a~-@!yFk%_kbAw7BA z4TVR&%@@Yt4*X9N_&YTLpYU)LR7NH$DoQ#^ zPC7W$TL!GGtSkePm64T|!b3=fT?-0FprwMs1TRSZK|{wg%stc>7491x#CJgxfeela zR}&Dx&-49DzCe_T$-l`5h5cy-ygg*l2$T$18YB}KDD!uWuyEZ-Jdr;I`X4dEtnuAm z8B5Qw;D}InPu)n*pm4#zlkjl=H$EyNG~iMl4|f^Q0M9@?Xc(R=_#Z6w3{1@bjd5WD zZ{I-FB^KV;|9}kl_4>o(<106Mi z3;s}#V0T{+=;fumyoaK^lCrXtvJ%8oN?u9HT?!)W>28U{y8>k7$O8)_RIbt4w2=~Gxs0kSR21TI%0=4!H^t1{` zTv!vVsGy_-0?ErM%Hsdzl>Y*<^$ZQeN8$xdFi2YN54sm_f#Q|H8y0a9r+5OF7C$K4-GAi(!>ijVIy zY@rDEKbV9eB0W7W8R7B%m~!_)1bKVn*Y}?R^{;W?|G`^!>R942j+Ec+@;SXp1n|fHVS9la6)KkkFuPI&|d;nc)!zXrWO7VZn66NE0 zVJ8rNZG+^bKoDz?EEKE=l~a-gL7*UzfXx5Bibw^BlDwR}l9Z>cy9Yjq@Ii_2kOfI0 zk+O=4a$tx&1g!K=q5r?Ds3ZlJv&QQu3xdii{!>L&nTy2rFQ=-?{5QA%0r0m>hd0e1 zWB3e;PslR=&d7h->>^VB4_|+V+y5a4Jn8>1`Cr-hKj!+6x&Buc_+J74C%XP)uK$$< z{#U^NiLU?O%mw)COy?Pd-{?i*Pi&;o!k_V{JxY|meHZ}&i}1xCp@F5)9)6HK+`vSa z{4+TX1(V=w`>H_v5E;@)PY3@W|N33KE1rPxn}UIkmNgsub}wadSwx)J@{!8MxUtt` zV_EyO`V@K~kgr~g&rVab-q2CExdC8vB@6kD`RQS3l|!$eV}p&~)2#cmelti?iJ%eF ze3F~5QrxQZH5Fk$#zzA-;wzSR)!$sR^ty3pR7m;GuO%f$7zIBU4EU-z*DGfzko_C} zHiiDLkF=zLPXo6ygGF8R@97xg-^PENyZ}Rij65DpUgg}y!ZQj;;a8kN+k26LHEBor2}_O|C>zB zMzLZR27d_rORPonpE~uK7aV&C2wURcNRw58f9MON&^A~V{8Nz%A&1m|s!J`LsU#|K zIo+Xb0sN#h}OH5$W~+s z@K1r*{D3r>8<&V7jsKJ;rAjPB_Y$o~_`h(gF;u;SxA3MZm(<^82*oJz#QmIBg}5GV zwdFkgLpxN4(G{F17%$Uy%(r9d##j0=gPwKBLlXl~hm)(;pCR&T4;G@CVyctelHqvO#x>IHk}bauzK66gmt8Yx`YS8(>mJ9nN~`fe#VVaSo-IxVrR zC;*o}h@44KS~*3N&^?kO9q(ljtL4sHb37JnWP;7}0*?94C_eT#QFH4dfkU<2oO^i? z)(Y+C+S-KP%MJnk&SIP#@MRxk9jzMDqt%TL+84P=+WqKX?5_uo_PeX9QrL@ACSj1~ zg*{%jW&P)x50dazwlRkM<}RlTEd?(}-l5Qc!L-dq%(1f^g~k7WFKb=JGRJJ^3T8DTI&r zDNUxyz7)ge{QbkHpEjW1@E0*g5hd3G(11T#n9g{GCW`Pg;|}VT7-#L9KvDB(a{gB| zF)WQrGmY6=w6V6A+v|SX#he>e8;bS)U;NAnuU6BWdbU9$HZNxpa zqTfyXLK6lfjZf_d(@ObhQ4P7XyzZMplc-o|kvHm=oz)Ht)c4lx_XvtH8FE{N_$mh`~x@U@?nVfRG?k@ z)J&sFGmkfis;;~{hv`uJWMYc6Y6|0I79Ne%9n9T?<6Vsr!zR)GV#Q2rm>j##C-~Et zY`Cig=~&b#h$5)2XWeiSYP(3~K`q|7#TnQT1zSP>Dq>ayaHyJ8%BNR{EB)ip!xn&Z zjhli?y7Bq=AlYL2WLmlHV(wtEXv)H5lG&<#H|ZTU{nC=l;x%@Warjf>L_^{(GwTD6 zXYiF_ef{~I=s2r^^cF(q@D;fHW(iTFHu;%lp#OlEY9jP$Z9)`M@Qfn@#^&&OMOsK` zF&lj1MWUW;Qd609R9zFiwNJA>Fqqc(zM1ynBWJl5$|JY>xHWe%L4=C*T-5NIvQYSn z`uL<0`)7Bx>lu`355Bi>D^-h4SDRH4@9JsM{ps?}imR*aLPD{wA9IJICNKcSH9GSN zC7KRS|9wAgE6JT9{PQrdKxJaOWm>LaI^c%LWi7(zFT75Oj|NT^mnzn#HBRk`0f9kN z8l(-~Ar}0_quo z$6Pn)`_5+5y7q;nhg9*2hT@NeFD2$C?eOW@Uc#9pCa6Pkx$n#4JW}6Kph9zX&1io)Tg(7LV7xh``Ao=HQZPZZK_-7)TeeL(~q*DVe>rt zye1!jHnn>@GlvywzI z8_(^Du0k;9dz9$26o-jaVKj@}z9X3)w`qWKPq|PD3 z^fU^Z70O)S|0Ft!khULs1xV(o7@bDx$~xK{hXNrfQPQ zPR2tr%?>rf@As*U2iSeujvgQ-vO?c1UprMIU$Y6s$v!X1nX6>f0Kx)mmMvP{p|4wx z42zRlI%%&=$I}!y)#V6pPCOAr=uea?&S6wy8??nTA#%trRQEGf9GAryyqAYNK@Q+3g?~L+ z!iNzUNM*s|H%hCTM#hzk-G0tEnPa^3Q2%6Cz$4VzE=@Q3NFqmDl!qSn%`V*)LBvw& zwrkwCQgho8vvK%Q5@aB~A{=YBOFK+i({F~~3-;31;}lDi!7zx<<<@RRQ-ykJlCPcX zQNh6*(&8i{VEMp{%@0h*dfYrZx*wgA1Sc}3!>c&iv&5=s--pN)MR)F_MHeO$gFI^V zh9GNoxy2ClUdG8f`Tf+y^`VH;OCb-GozM zW*N$rWqEcz{=yg{&clzX5b8UN4ab>Fulj_Tn%Pun#y!S|A@w+9z?>4hto3vnJB0Ps zrVi_~0&`8u!_e%I3g@!a!gc6GY6@`60w=H-{AOW~X#vbGx1aGmjOeb-huU-a+L28| zpHhqI?rSZ=H^}>mzBmI2P-UOZmhVdmk%XTWVIfKa?xtMcXB$_@udVxy6B7m|=iY>| zN{pDpxVM_g+AB?FRv^6C$wX6Yb4|6a7LImgmzw`Dk}LfSg!&Q9;uT{Zkt2HO!N>bG zxBZ8|9p70hOPCx%1$M6}bg@EZ}lO=-|vj}Dv}KG2$)Kp37hGv_TU z$6bZpPS%!`)+2{brUnTyS>W03^V%h5IZ-}bF>b{&I)@5RV|hLfhFWjGvkbr$Kmd{x zQYrk6LKk&{d?DOeTjYo$a;H8jdYGyo7__@6m-p!d?Qz>w}D#0|C_UQeSp!(h?#aZkl z`P9b5m|lubS(dFA8iP!oua2+&A~1Mkh$fX)&6GY>B8yL~wafg1R_aM{oUDq%qQqYh zAMMH1)Nn_1(BEDmm0t!d=EaRS|v+mDJy77Nt$t$ z5Xz!z!rk7yebV}tn1?HG_OArtvi-)w7x~(G!YqbzJEnGWhbj;vJ!$W`oxD~ga(ewr zli;yL=blIK)l&%ZdCG2W+}a+YwYe7Qg~izc$I90m>N2h_`{9Rl8xx2CXlluWS9mvD zK$xSduKeX@hT;c)43V2%D;*pUt21Vf1sC(~MGYh67g~r6KB^`Kc|bU=@@pN^WA?db zH{U5PCclH2262}6eYZ<9B|f5>uF9}2O;4)hMgy~(XBs0{8H{xvIkDOzloeON0h3A} z>~d|2vwUF?$Sn|!)%wD|QNT=8A&S{llOb0r?w%1f>9L<@1}pC=E?@rz-?1r9F`jW)%=YeHY-%BD(%}Z;m*p%{uIOvmD?5D6 zz=WVDaa@@OVk4(SC5LY~Lxg3waU1*g0aIP}Y%)hT6Mq-Gi;_}YIXsr^W|eWPD)%hT z{R$!r#Miq=+~46XJVH^dwK-#CZ0ozKUkLQ;F4JhtJ$g%EA6D=xFw>#=_=ICur<&H$KdbHBVi?erxUcBtbU5CuRh8mWk5s&Ij- z9JzuoCGUt>SgU=}R?H(w`>%oTclN1$Px zbJDuL{&7l=f1d)HkU0Ry>)xRjL#!ycLUXbkXP2>Gx0G;uA}wmmnRqD~@GOLpV>{-L z1I4b$GJiKsM5x@piS?co*Zrs`d-A~Rn5&QC*g3!eEQm%!R}=_@qR2(~pF9#6TGN=s zL^`GIvn6{`>pj@@LMXJYke$0C9WUBHEHR42A9YD7v%$I_Jvs>Js=1vzC>itLJ-1a& z0R+Jf&Px;S+a**lG8}`S?b8@fcZFdyUjF|6%i;gHzs%$>1{VDyd!2;k<{Nqb2|tg zO{BgL;cx(Pyb}9TA%o5Fo?MOp>y(4am%Qi&^(D$v>Q#%ilW^uAG!hQ4(uFWQSk{dN zoM%YsM>eJG`{3Z#2hJsRO{*Nn)(SLDeGQv$p8(ocVK8`_3Y|o7pJjk#jE2dm`!fp2!tlS3MU#Z8dLSdt4_C$g-?`41(GSJ7s=^QC?kfwE&~a&90)o&1T?7mQAZ<$ZKcX|~AyUW{n-hL@G1#>3?zZ)8=B=tWCAT-pU z&k*(-_D4yZ4$^kQjf*&wS)NNSQ+K{XzwLD=#0<|h8_yVfh#}LK6cCDOV;{aQoA=t< zM7aqB;ZU8F9V}*R;zBkJ9b|qQveBdp91g*BE%jxL~5sHBi#?lt`JYc`U}df|1MMSnhUrY!7F$Md|z0&*4$~= z@RGwvEHiV= zL&mQlI_yM_$%T(=qnUehgSuXVsMo*@H1kP6_icWWc*f0T!8@h&Pj;xs2W)Can87qp z@|)R94t7*M<`ah}8WlY{@T(@d4qPnL&R%jDq>9n%gRAZ=gQ&gTq}Q=r(8Y zaB#DojofDh@Yy~ae%k^n6=^3NoA7*dcOpRwzp2aKA3m0Vk3L-m)b|E<(C_x!{bD@h z`S!zwwe4jt&L`llMs9IpNT_oM)tfK&Fm}#8t|GT5iR^h~mu`L4!OLlriz5|7suUq| z%RSMz7HO}{KB4Y6&jSMR$CL$g@G)kX@;IZ8y0Dpma;(L8_cfgGgI)UikNOm^0KG2a zHJ|aJGW-!n@pDyz?9SURcd+RWL%`G%AcsXB(X)xn$nZKEnboo4oNNo3BZ*ipUQO0B zK+~|?vE;Am<@qe>&e3DiXZXhWWqW=_({~`;IRERwi{-A($H5AfLhUrn_(zhJR>_z? zrL7s4W~4t~WOGWDJtLNxKax0RS5JB|NYYt*Y?U+283NhqXZBHoJe%a7ONfqXBmOv) zerU!Z6)IDO{3uq%-fyA98i^!-sZZvLFTPxdq1-*@(nl&1=0-_7%5 z2wnTX{dgWdJsPOj-Xo6W;u(y#IXu?xyuC+keGs7Ph0Or$JdNoG@x&)wU47;AB>vj~ zes|z+o=zk^kiFR_>@$2mbVP#*Y89*^JS6kNAGc4yXJno~mrx#|jt3u%yEAxe>xQ|; zwMp~bYRj4T44Eb}S@-IlJ=wAgL&&_`Hlm8BhJjq##6d$eW7+zbG+%Ur~@V#)>|XWWg%uMp~MaU|BD|LdQgz zW$~X&Ag1-U%`O3l{av4{3M@_d;q?5203Zh5r>MHReiuyh5D6W^GL$s4MGyV8uSNNAcnr|p`BY8LCn3eFMp#+~c6*7HSAyG$bKVTo9lk6<)(LQ2V!B0b^8 zMW=*ItO3GtXYmP|$N9%|vYVgU@EJLldv|E14l}Hb=2H2#WobMFh<$Z7g?SGqz3N`S z7;@aPsEyW`Wmo;Mng2*{ba@O!HoWo*OGsjMk+2x`pLr(BP|@g9m@di%?e$$U;=ooTPC^&SyEn15%(P|K`} zclq@p*f0gfsYEi7}@>{RADYD<~aC40WH#ARf9Fj3%^dDFSSk-a zynwb++t)7Fm9qIpKd#q$+%A1@THtSFONHA>&8i)32=TH)@IPpAq3(AoO_tmv=}F>6 zCii~#t+%%6yHc8yqKkHTBWpY?FI3+z+r^Pd>C;CsKDfIKyP``qTc=hr%+$@;ziv7x zzByoqBDQ5$W!}E|NIUtgBumgnElui29s*x4wB9Ox=-`o@5WdFs5cgSKMuy#z^Z=ck?L`8Gmls-ezHKPJx`;yt+)E77Dh@ z4m;TAs`=(}p6nB=;;o18c17*gpZd$4F*Am+!$)TjX`PAhpSxyY3$ynyX$&?ld zLsM_RWAVXp$!?Ny(8T-*qV_~A0Le&EI0S3CDzkK!~VFyq)B-)uF{_k9*8%9w#p&5HFj136888igD!S`+*FnTH$kXrlcfXA}-nmueX@AH5!^`^C z*0<{Ii$)ZznjW`JZpN{Dk$I_YbbsXCs5NDLbPhQs1zc^dNL%O z(%b1N2EmlBL&|KaJ;yW>;TYk&Te*eiX%gL-sMaf(LIyqVr6ITpH~s-~3ble>`ho8D znU${1^t?)fkbxG`7yY4SlA3TaOiEiO0s{AIrDqNm%jTVN+8(WG>%$b;UiE!G?Jb+sfzD?T30L`|bXOmKcp;m`hcO0WAp zn6PZ(2b#}MA91bUg%S!D;NYrKT_J2aY;ErBZ(-mZ+@TNdc8iXc=&Dh1CS!h{2A}77 z4DfW1IY*c48bvdchAw==P>}v5YbF=T4S@oEAzDRjxlEu;m=_B+uVH!A z>>PA2A1I;e^z(4Q9@e#Ay!{)9T`i@2(uwr5JE0;{xhHW_tT|r6jdo{932cnj2CLZs4GP>)kj?ZS;qazYeu~)a}M!9tE9_b7DC1{5-&wK9$EM7&c40yu-V7gMe)uOZ@;kb|Pkq zU5xQWMZ!~wHaYb())E`*lwi>Tkd#*@6)HPMIX%gfJ||6(5w7F zScE`9vZfQ>gE|5Qr&Sd0lX^aMOR&_%8f}LfCNMXn8@`=dY2jL$SO7akiT}1QTL1Wb zyH}coV|^{uDN~?Kluj2$qyL5L>z>fgaQhxj{d9!%;QlcWi(ioP^^8va;`%!mbWAEWreA6WI?UGPt+b{aJBY z=2*^Uh?ZnYg~L42&%^|G8%4be)dLP@R1wlF1j(oQ9{~J5nbHsmnb)5hhCq??n8BL- z555^yJpL^PSTKaLKZNY1Jgb|+SsJwnlI&zN)*xq)gv2h<^ReU*7^c#qANFXE?<}da z=IwK9tdo^z2Q@3!c*CI5S}Zj>P>z0$fzcs5SAQ%QdM`D4tJrrlkF0*s@?CB(ilnAj z)r1+#M!a|ZV8#hhl41z9n-lN511s4B$UZLNw@$yN?_UQrTOm|UKGcs1-bxo2gC)Ee zk#c0$}5t>f8*><(fHd^K_Bfs`78u=Puf~|BTy){jv+&b7G-d` zx^#Y|s~2Y!PrOHy`*vF7Mps!5L@WlWrcr?{v4tVjv&%0~kjPb@pJwq+S@=FJm|BNg zxP#1bU#?6L?;eGkIprc`Ej)b_3U*3HIwgIZuj#}$Fov0W6BAT8E}OPWY3&oFv;<>*3F07bcHWQ}vOd5C)#JS0aVHW|G!q+f*+s7zj^iSy z-8v@RoRVLmNuENb6M=XRJ|$XgO_Rug+bS{nz^z($bw2CV-{%Fp3iFv9a%t{5Wb0XE zUz@M%Bsrm&sdk!U&kdnyrlj(?nz5F;wa5G6y$-b&5$E2s?GtyVXui}riv|n294)S^ z%O<}Z$37z!Z$MZ=Jhoj!#s}6nQwNI^Nbm!$15$Pm!k7fbt@(k<6hgX}>F*uyiQ5gh zWQ)PhO)UIk+1&0m{KZ`{SHt_d59pL(`HLv{orYluzk0I#jy7{+Ulr_hKfakEek$a5 zG2lX~+Ovw-98G@dVW(dvV=P$UDTtKv-qSbIi;38M0YzWwMD+}+hW2>sK2{VpFjO!( z;#j$SO5wz<8veD}bjRAQ{bK!yrZLj%ai0b5ieQa5Kk9!~M?#of`{c)aPa)hJRTSV- zjAMwx%LrrOAdIOx$6lw78!S++G3 z0QWrR5};K=t58?)1B$>wNBz+q>jCrR_T~?XSb{mEqQt(O2TF>Tfg`!3g;IvU!5t{Z zypRlNU>&BZeS{L*YSxRip-B1dxsq`| z4Fq_OYuY0nrqno9QB@=?l%RvZeIPJ7k)KKIF0)siY1m za~CbgU1O7G>SrHW7PDS)`Kd>!p22C6;=1NCEx=nb%vZ&5EZv60U`DkICTOj(uhR1e;&eWAkmyM^BQ$6Ou{3DoWAvgiq3*`dX>gLbhF6x z#dnE-DodfKW-`=_FP9bGF|ShG1gPa{9)aoG-9(C(W6%uQ6TA=k=_pIHL!I)V-)`%) zOh|9PRqO|ITofdX=1#qyLrwb+12-;UO@G%sKYq>S5-{V7-Gv!wO;8c*;|JTpm& zh#nZ1L@V>x?GL`NE--H*@v!%vJxK@akw7NUDY*w$h~o5e@5C5XlK5`J?$1gNBE(qg za1z@^DkT@ylb zRoE#>?6Lx%wOfd>L^CJieUmc5DK=*<+}NvgWgzq&jW!qItbqIZ-7HR&fJ+}Yn!6-K z14Az{sMpLKg}99ohXbF^Cus_WCKzR1y61+!E`&=LcFQi;u(Pnwb{YW!LiISZ@tv)S z8kyHx{MU1BLS-<%$7CmKT8p`_U@r#fozlu-Yba6m#P(KK{+B6{m4-)!Lk~k_S=d{)yD#FjJV225Wx&lvA zC7Mkl??d1JOu8nYtEk_J%p~ zLV8ArWSWtYvR5?>9BIWr*fWHPrrip+Y4u#I9DeD&HyM%|Sw8BZVt)48t%@T6zrF+$ zsa@cF?Z25P8QzOwtBT@>Uvy$UTzKZUpu`|4&oTNv5(rUpkTW#7C<~n z2)jW%Cb?WXWSN@F2z5W>sul$XA(!ERmIhtDjcKCOhoqYV-PItuVg^Pl(R>)>VE z50D_FT~6S$JV%tc1d5zTyTKeFv*y@A)z6?!+1N|=%kU?Yd&h)Gz}skpoB>oWOEoM= z3Z6WHSdbB(PWbRRdo39jQJk$ACVp@GgMMi)+vnA5)0-T=RSwo;JK;?HIMXXPR0cZ; zD)fLahGkl`hszRqt9g0)8s=*!yN-FtR<8&Z;KCQP^RQa)2Peo+n+drcOzpM4!#fFXYEk5GYYXT+S^orAk7B2+C)k+cYMySM z)=x|^w!oPo^QA?K2#%RQgPZm72Oi8(E$Po{G|MU{XP0+z$_WZeeN(QktAEpwAQ9NB z>iIASWCJ~cp|8Q7U>FHB_L`n#%gyxgR-cd9a%TcQ;yY zaH~*bxa0&DPF|8}XgXTf1V&Bx4E+x!od($`l2Zm0U7s%|oKOh%jd``JII(wUOZ0~3 z%|l%ncy%VwYgkxXR{(xPF#4LbE8nfeg$}JG=p8HARae>)n)a*`1dAqP;0NG>x;#}G;ajQaL>ZS<Ym;?zl;yr!Y02FvtG={6e$I}~R1BO98=Oi8yBf1s-V zxIFOs>Cu?I1dD{nR;AZTTKk~nyvbnJ;VDURWFL0n1 z6-k?b9wszCL5@M$sy>^7SSnqSWqhKQ6?eNp%1_}f+st$-Mos)>`~|-phUT*oiwWlR zcz?!!>Wl^y?_7z?&x6X-bJIo-T)!rEGql9;%65kzCShe2XI86SZB$Y8#4C4OM92ex zi?6!Hs~2sy@(P-h6WF{7E!!IXIXO?ww;X2$qzw}O`W4Xk1TbwDbuXc6YHN8alJCDm=VW#RrZjwq4b(_{sDa`nr|;^i$2oOZn+{Uw1pfH>Rqg zfimU3y3$=MtyzBbYg2ABRlhwOkQDC`e+H71C@Rp8XRPnj7^b)iGY3mx=W8&a1tI!a z$pXS%dsXWVGXt)MB@Y%RM}Ll-K~U{? z1RzT7?~n6TDc3Ss!rkYIPPvb-$ww8r4oo6QP;ouB&GqB#&3&*J3<>u`L!%7k@Tk(` zwKY$vQH9zDKW|+WA&(X8;br}rgIOBekfHEw^rGp^r?33by#mRPGah3i|7r z+H|a3H51`*Tv6AGszY}X8NqoCo zIi>>KIQDJr=|hpnu_A7-_OFLmLZw^R0+eZSOWM~^YVt>e7?Z=Iwvn&Dp(#41CJjtdVD zRwAX^w={lUtvFIMnwuv>{N|`~j~@=dm<1x*T@t#4_{dkHpnJHP^Og1zPukP>>zBe} zI^O2;LIAzxc5+|1tnW`_W;c|mY2d{661ykz$HTc;l*W9!m)1ln>zw~K(q+Nc7Nz(=MzjMHl%kS8j|Vc>a=^i z-w$XKlP2jnL|;|iR_mt=tMLZ^XWU2Jv1t(y+qYg-1?=mQodtG82Z@+!&lbPAg%ay{ z`;gI{B9%t<#wELG+Ix*63LaTq>gq(LCd*E3h(c(>?#+`jy ztbHq`GBY^EQ-;OpzXFy?&s$_XS=q7gwyt5!+LnUTBdfVW=a^FKrL;6Ibd8xPw4;OL zW*oORP=1cyWJqtnO?40fhnj~dT#yFd{umkBhM;vtSzU|Zn~U561&;SkXL7GZRhPE= zy4(Je#C*lTOWXN^F`x|$R^um;Gwv_$XpsKwHbKLY?|kI2%UvHBu!`a5OaBT;fSVv_ zPQ6g=*i^|lR?@_XuH3-W{V|pGC`o>=+R&8pEv^(rHCJBoaS!Mx(AZ~SM87+wA;ooB8u_scvj8Jta)dud3meiJjsfIJkLD?N11jn~pd+&i*c z?NzhPc3ku9-v!1Dp9OdkS(WDve((uIbc`FhBhm*wU)(oeDyD#K+vk)Zy}$gB)8=Kx z86K29um?PX^wpx&x~m%Wc6bsSqzjW?9 $UCyg&|& z2<|6t7bD`;YpS-$$ciOzI&v|JCiZi*BDx2v#or@xH_>kk1&=*m?+`VvugbYqr8HEX z*OZ$0EnQP>rH06@BQW_nG z7y)%o&5$Vm9=C-O*;Q-7dNgK~8`b&7tvz^08YjwPF1OW`ZDk)X8#nbt8_3|aJFbKv zJPLaFW_juh_70jpeR2y~jYhdGn&Is%TZ;!XUhpb3Qds`INBMsHtow^4gRmv(w7H?7 zf`@gL9|U%Qjy&3#dNVjx&Ez*x6fd;}!TD1s6J++|B#)7=L`PIlVuozwFt5MvK`zy?jya!l-qv>IfAD6fj;#50U^dChLrV^-eAYY zLXpa`3a?e_$ib&`L}=AqFW9Vh+}ptF_bMJAT}P~qT!f}S^W)nw6tBf=CY(WznblF8 zNsufhD-?L_byRWUG;d^DGBLsN)RN|euV2d#h6L*E@ezo~%A*Q*dEs|lRsm0XBY=;$ev48^KI9Y(X#towkzjpIPnisYMzp(wNM)hJ@dB`pk?l+bM92u z*7*YNq5Ox8P02Ck23-%dPV?@FG`^#28=-fW%jKop)tyz` zsns}>{mbIt9M!MkC{r-E^VgTqx-6mQkAg{qBkI6Si&{k;l0F4D1!6xGzVKs@L>4R% zlq4ET^6})h1d$5O3!KmKe7hTUmqix1Zs?g!eqz(xeXrlOlRVs_?3kc@to1sMq2f!@ zdFEUEy~Els0g}CPEP)|IbSv~#0_TrH>MCE8o(#tVt(FSmQM3yVPI`7zvKhR5Y|p-} z8gk)HA=&PP@+p%}OyDcD)Zu0Hc>+3aCI)er8xY2S6_x%oFy0>28@NCn`6Mz05lv4e z^tXWs7Hus+i7969TCCY!jdGB#6U8uQw*qEWRXT%RYrg)|jQ7B$V^K9Dj>OVlWv^BKjdR|P~7dre}0`2ow zxiOVFlz$7fDzxX@-{lIu61TQ+3UPMpBS#r()(O>X2_ooi6(87@Gv+LRly)k?^%P?2 zGBv8mhv4}V(RDo!PM&V(WRgN(Icpy;v7K+(w|TkPbo8Ad_548Wj+PRdc8Ou1=DU=e zThW@O7>DsBWLd39qf)4GtV;27x182;bXbB^_e>aAUIV9jtou^}1N2kNHOuoT5hk}dwUjD{Va%%R@RGcri{|lp?20^lX87dvZBb^5YTY0u%=GnTI{_cG(aCA?*eg?_DSVzS?ASzs4fKh#E1(IK zytd`!tab^oy)0(GYw%1o_}r+ak@rp$%%FxU*|k{lls<@ZLSb7V(ZkBieVxN^KmK&T zTY22Lk9C@FU<T!Leo#(a-DXgB(1I;F-+FOWNq5 z#NRc0?@nwiYB_&o4U_7d%}AblBLL_OaV9-YZjq)6tt=*PiXin@^K|CrE$2r$wtL$# zWv>P@G_qN8Gkh<%!YhxT3xo4?kE1}TH3JtpLNq8`WO=MalYT8_1TwFcxO~8^`>6cA zCEu9Bgk$B_B{W}RV9_3tt?u7T{pbKf3qJ8+xFkP8>@J5D)iD{LCzsd^i{#~Y_9b_bRGi_3ANs`Z;mj_fAn&-@ zBC`V{hF3eeQ{^23SGSA7Wk&_!aLR6?cB8-dtAl9mci!fgQ1a%Gx5ZV{1@LuNcvSKO z+cgTF;Mv%hqZ#k94m9C`Cm&NdL2zpRvMXbgbiJd*9fn#FP8o?y?+6A@}T5 zH`Ar(mZcFNy-)t`Ij&qB@f9gI6ebD=soyo#1-rl;afaR}Xz>N2@j4rJ^d2^*=wV&O z*ww~ZqVxLzXo+}KB@S4fJHmht?m_BMndg$!bClSpDEg#W4CSB7ODA7yzdfRNjZd)0 zqh;c|oQ|kG<7^dKY4VdTSYm7BFY%JYmAeHurWPi-k;2_E z$9Z1)P9NBipzl#)CbH}v!zWE&*CObBHGdjC(ZBDz%sciHNpZE7)s2R|KIk0>tZeC|wpN648DNpL= zl1vQs!gMomv??8Y+-A=!ndLn~w{GwI{b2StVRj7Fb;_W`qx}!Kx?W3=wE=y_hg*Ew ze}|yGJ{meN?ITd@W~N7#>_58QvU`M-2jBkq#>=HcMjAC5#(T#B$nOC5+o@%>4T*0F zQqe=ZBty2er8Z6vpC)=7E6@%wosZH}#(9|F4pW%j?Y@S`DBlEbv+_sBe9#cju~M{F zu%qLqSNy1kve0%`?qFqyjY3btQ1`B?{5MC2OR9N!8}ePbJ^Mt)Wy@g}qp^>_9O)bZ z9^D5!a?q`G6Q9`W)p6f~W&1P(b(hnJR$zrvvq;jXm&6AtezE5%+(e)8@Jz~?@VI`E;3o1Y-QSd9;@*6d^+ z7g;rq*&4MbkECCvl`Pokn18Z(h?Y9X+(WRO{R4h~u^DCeyE&g<($xhbU=O4 zX#?l?8JZ-3Wm+)*$+4bY=8nI9B9=zh?KR}!&?&WM)+gKVe*m-Z3B*48hiq0${RJU* zdi{495A0U7UH?I!H?*o^UUdNvO*wI{RwS#(v_IRyF{N<#28|?vy+U&S!KU81$RSWz zdxO~7(2|Dtf6@K!9cLYy0J@BKXc_JQL!Gb}@saSyb}V*EALf~hU>?Q!{e5l`3!l|A zaoW)*0W<7L8$E8il5?roJ-YYt)lk6wiRq+8i|h1y%lS@<3}=_pCb)`^uPLp5gjdyY z)n4GYH&iGHUF~1%fJh*%C3qg(lfCLVTlpfF=4ZjwF=J$&A4`e9Lu4qU1X@p17Gbf{(69bY4UfAQWHzMd;@7ZZ+@gwp!t^NTGzdu>` zi*zBNg2=o&v&rzpxlm~9AlUDUB_7$xX}#JtL_{qnFE~qS;hh#ZfAsSg-Lveq`&azs z_d7~Q$3MnqL5q+>J`TW{Jl3kLDO$^D_7{ft!2SMxP>Odh8SG;dBKOpqjC0(X;0FK= z42(lGG0Kruz-H7s zus?SO2I>FrApIK8{5{h(wSR+$zQaxp68$8rh*Ut76C@?=s$ZA zhi?7z=huqgpJfUBz4hhfY*V=0?ntHLEC2iv#1!@48LE7}{N1h*!i+dkSEBXjhFFvT z-tgZsRle3V{OaTUuMB~29!8u0$FE=dRQ~4|LgRm2wc_*t$J;!PEFMC|=3E>>p)om+ zw7wZ^F6r%hxFiu20er zSNFaTFc{vV!(ppB2T7ZU3tJfwYA*T5yu4wp`t4C&hO*XjQrL=)-*Vm>xxN?s4qG7) ztr!IcTqWwhA;`DM(W^T&era-jM^$oN8M6L@e1`z^viYCg$5}Xrm=o`GQrNMB<27cR zf*JCL!r)GJM#+@xNABXhJmiKIUA#je@v`O7?!zv_04_{xV8#7l<-yzj;@f%P4ZX*hNpO9FJ9XvF?R*l)cqm-sYl{^<`6+~)pxeEr6no$^8#IBTD*e~tTz8>xG#V<#0s@lQeb8chyT5x`f_>YP zO`hzBsS=4emMr#S>gSPz8Swlvk zF(8V(9r^SCNl!lKN!8PnCal7!%3JXzi?6Yrtds33x2J>b`DtNE{t4-T7MD$8Gi_cX z5KGm8PS9K4hM*|w_9jg7SN;JD4tVH+=>!wLW>xb3vgbOdYhLH?;DCwWtXS-SFe^*H zO*7UidyJ62?$>{RVENz^iW^d(nirP5Nb8A595WdVm)-y}%cKbv%)85moTN$wlF{<7@-#NyC}{{Tcl1mvZ(UCwJ{q>=ryN9z%_bAEadlIvY&{8^o% zd|!obmM!8?wRx{bVtnN8MLlTeURFYsq`xI@Kpqh7 zWfbVufBebyy0WFmPg+w!BAxe@U->&%G9J7kIDTD9j50}+PbEP7_=Zqq36Kn(webrz z*W1#6ziTut4I4)7S$SgGPy~x9Gdgr3VgnY8mdWVfN9F8&u3^cpy7AK6A}p-xs6gp# zeW5l9mJh@eyDHBmpob~ZyF;WukLdg6Ox$h`ci6~Z@;5#^W8RG@#>;sBLD}Nc zDe~`thYqKVD7Eh#5VL-Lqi}rS;)sgwoh25uz!QJ5*VvMEuGf9B^ap0;KPK@8w!S~v z(Xmcq&gm)B+qp9QqtMFLX^RuZiVSUs`AzES&fl6tLYc8QfG|Hj#b&DQE%|7#M!q`c z%kF{&8#vdXf8Sc=(zwP`YEidNd$3Yg`@=q*NDN6(1P~yHcah1xD(|>ah>wWhzRXnC z6$iHURc;EMds_rbe;-B1A18=i66oNz?XG$>jM zv_^5LD;>lty(cMsKxx0bye|xRe6tizT5KeF>tp+6#rn_MMsDSiB~#x)BD=4AU1-Lg z-^(ZY!(ec|#B7TZ|D6(^;`pV-cTTqenOc6~z<+`cZwb!uQ%4)i3bA_*G5_QCjXmv*>=83Mi7E}uEooa&e08HP+5IoIp= z4JkYT=A_xn?FMX7=~7)89kX)13}^e*EEl3sGv{z5-TyXXaoAI;}N(q|wtU2^lw$?0VG?Rn{gku^_=m89)?ZlWIyu9Jse%BzcE6q7kovTXh zBwu-Rp{#Cw`P0zS1t8!al^HB>AUVA$o$?R+9mgKlp;oj zhG-+PRHVg77b&*hTLk_}m!CA%CnIO##Sd^3&hDmOAh?+%z)vaA6Jb^VjBB{l!f&i6 zQT*CgOrFs3J9~8xlAu(_d+0TCElIDxFC^J^uc@i>aiIs;+wT3i`E5GrVE{?2fAA4_ zt)pE&$0)2B2C1X&a=9kHoi2|@X=e^P9KqE%UR%A9+i-qTVngQ6p|z*uOU4$xIVm=} zu5{IEUXIb(J2mtsw->$cN+V(uEPhmrJwWep;?@t+rZ&2^<#MBbWQ*S8b$BJ$i*n7J3_g*@Ef3kH-CS{ZCWdGAc=iA*{zo2⪻iG#SvZ zjxY*q0)@A8vPgQ;z9&hWmHgDt)JBF?q|ESSCVaeKHL?x{aWn=CA`l-B{>~4ko|+`M z0b3D2)cr_rGrp<$1zJ`iD-HIhc8O@z>a6z82ZO=vIDzm{`XEF$Uhc!8b)_|GewA#F z6@8oD`(aR-s}EcPTtNmO=Bo~NPm1NK;X}QEe|(e<^NAL-Kx4S(WL<%f6ej2XAS!6~ zC9(6q!vmtJj9ISQtN0;W@Mu6M6Z1Ul=~lG#hp7VVWz`tU>9oi8F=Zxs@(NkEh$7Q$ zzh8B)GO)9KQ^BsqG|z@g7oc{tZ=9|TSn2B$r+CIeY@@vaj@vC4%fu^Tgj?ONI(UsP zWsbzQ@;# zsW_e0De*uUZoxTO*D?5T3ENOEp$2-qcN{@}#L(6mc%-yxGz^eH4Iq*){GMOBB*~o{ z3s|H0IzH2&OusxfuyRHlEd?>G>;$u(}xVs`SUM1VkLu zH>!Csh`DbZkIMG9v*PAI8pw?)0A3vSqbc!`bYg-^+}JU+>@)VHW4g*+j<-nuK-dqU zm0ttgp4l8))HH`66rwI}vF!Ze1o_%+ENT=p+PEM|rs7?7QPo$GVqesL${0> zy7!PS8ViOGUOH^(-g`Xn=0+|KEiC~UV|s87FOg4E+=^puR9Sr!FaM^X#?5l2QV&ha zizIA2xzz;c+0}U~D1F7O{*~Huf=L8FX+!J+PJI30FxZtZ0cHn2DKfsz9Y!lY<4V?* ze-LV0u*0Kzjv5N2;ne>yv@%a#GjXG&_}+}qm`Qn1aD?b|X7CZnjb;WdgR3=;ZHp@2 zJl$>&7=24xs!U1#XB{FGF-)VVhV|oowpuXDJ(Tq4_~vqm0jRJ=SW#&;=%d zfdW4moT3+e52ecd&K2q3{zs$_(?V32$iT~klG~PLI(@|t<6#F~4ntrzIVK_tm{z3CgO+Xw}pUto))XO_h}0o%e< z%Ts*l7*X-GS3}yZ^y$>5tajhl`&1?*fFZr}yXXH7plN5zwQ6c?cE0|~Pq6S2zeBLS zl(kOk^kbrKtXoWQ;?rp?ko|JMOGL3Xy&z&k?cvp7pFX9?W(?PFNwuK_v9J6Ase%SGgasWV)YO5Qgq1HY^X~&ok3Vi#GOaf^~#qZ|t{k zR=veZT}+cu@-dITVog`w+^lB39NpdeE$OHL5igG_j!$pR6P00}D!9k)@F9I$4Y2Kg~DgH4G^-iEL~sdF5KfXYh*6{{^*obCub1$tr2 zRSdn=UO)u14wg}sKcIyTll_t$!<`>(V70%>AlkOivDJ72kUcGuB>Z_*B%fD)Et_mR zy>{FHKVxRt$V&TWL94+&*dDwrT}`@kr2jUzmZUXI*$@1GAxZ8{mIYZ7(#2HIr5(n> zegMW^<82ApIcTcy$GrCU8p3IyD6C42BJW12uhwW#*~WF16ZPrU+j`yCnZ6{A)#(Bm zjYE?M?Qs?P5j$hHPI@dJKxnv`Uw^309SnP;~r_7$pee^GLFr9as6j@pM zhlDD&@%1O}_t8c0h^50(xYcB)yvRJ;yA0M)GA2@;ZC2YfGS<$%BD!(yfMf&`R#r%g z<#PB7(8BX8$6~0Zm`E^-OJp>IQw8a6DR%%azEmuJVH__Roqp616<&XheQ~fwX|9yrthK^?hvg((87GKM-;6^FnFy^^$tXFRHZ&qtS!E#88|+UOzV{Rj)M7w zj!10C-5OB{2)yQw)EJ_t4wo#T{OeWc6|`Y}aO&g8s_ziXF}ug7eH1U%_?W*>@?M-p z!8YT{9JM+4OU-kA5#Lh!HrYP!6H^%IfrYnwl|^i1eCD~`y`8 z(xFer6dfL1KJxZdmiacjVDp86uh9d(y9akUS?*4=zT&WWZZR@O4{4@|qVw z0zrNH^#IJ|pWOv}^gBABfi@t1%@)aZ)>)D3<4=oGCqNFb>mQ)BBkydB2Pyd!xX*Vb zz5-prY47;l1NKd_J*37^^W4%h2x~AMkiGda&4T7t@Ra2DAkQy}NkF~N_xVC(-Qwv7USeCNj z1V7+wh%nXL~ftJ+5*#b#B2@d#i-8xZU;Y!u^fLhl-Z}K zQ+xbNgwb?`R3zw`Gw4FD8!1JEvszNNomM-Bb1+ z3a9m<70Gz}GMrY5(5TTz(pIU{m}Gs70;8iC5?4faiv`L}uxG5dEB3%_HdPRzN?#{hM>4rm87kO<$}y{CUd^=v{w_$tddE8(IJnXjjzUXMreC zB+hX$X=P$8Ee{h1Ci66;NaDXpf|kbOd+_B~ty)*$FJ&UF#t{>Dl=V{fA%Eu}jc3Sab7lTI2SW+)9mQd!po;p|dC z#UFa77X6k|4)L7Y9X=i;!Oo3Z*6G`;ZepSb#g7XP=ZTpW8&&~;;mWl=I>)NE&)-6) zU<=jQQ-zxL26zDkXH}kv+Tlwa0DFFQ`M{IZ z6-Y}!yY#Lhb*_yYH8;mtw4HvIvB?=A4P(`MDgB&!Xfvpk;b*A1{BE2V5L2`&h{hl*R2WI*1Q#W^Qnf^##FSC_qW zjL~ZSwXB5B*&%d@KRh{&{auYmACU0^kS)^Najbp;E;iNn3-A*B;XGc@L&^V%z|AK$JE@b2*Be?&(zr2++u!1 z1usqmc7Lk7)u6bU?sBaa^U4Q!X%TAXM=Elp1pJ*Gxgoc3};w^uaCQg@+@B-tdXk!aPgJT7{20j;MGG7dpr-NFcTj`p6 zKZPiH`Wl{jakd<&YANVV_wz#w0ZxRIgK7OeH&0B9r@O$=}ecRaXyjU+xV#>?K6tBjQEa0qx}hb5^mIoH?&w`L6`!U z&QqsHb_)=Bl7|t-@60SluMy<^mNTQ)N51KqET@3gUGzTGFu#ke(;P?#S2U2AzTx^3 z)^kh$7N)6pOuK5~;y=sgQPmWz8o-fm3veoZ`n~hbXJ5aG+R)+Pr5@WrP(JTM-72zk z42kq9nGJ=rHzG^yO3(B~#eqKn9EUu20J&*OlV*r#uq^;0%;pDEy9+&t?NCx8_0-0d~TBTM^jwngHJ+7cTe9Y}=XIwv-% zCbGo-SV5K?xKYfRj$W4Gh5@9N6&61V&n>?qDP35Kh^c-Q0Am;mLkFeLlH~H4Aj&9L zXc<#1U#?vT_2X+a6kl6zVQ9Z&*P24KQw)hG2DN&FaYiayfrtxV;+O|wW05n3`3W~0 z&lZ5_`TCX_V5^Zdcbl(RGOz5#B*v*e?7IgbYKflPEz3_Yv zmE;tGu=)6=Pr48*(Ql*W+&QM^USDPgY@1dS`D(#o}KqicF*#iy~p z5akw27+91hPd?3#^p~mXZT(bUpm1wU`5mECtmSCPlk#n$wCxF1KXfWqr^~GQyJ9Ct zz7ga-ih;&yzwcVypg|oXyT(UV&7c%p#IlKQ*In)1hTB^MLHfr04Ce#yJzzxMt0Y+5 zF44OmF7FpF^^&jWhR@I`3A9l}NtMOHwl6I(RChiP9+RJ^g%lf>B^cFsM?4E2+siJA zz-^W&t;G}oe?9axMr)GwsV7Ns_x2}9NUSk4fo*Mnf(K6XJKilZSm4Psf>Sev_SxEb zc*2b-1(Hni(<=TUx{MpSVGwx&FSS^{+P6FM?K~o^lGyG=mdubH23`ZPeGw?XZ0?)n zZEXrNtr{5CE=z9LOc_fVKvL`vVqTOZrabY4vMEis?!+;4xM}I;x9P?aZ;*ue6TCc) z7IE`l-+P++WnjmIs=*c_D0>+?)8ZSss}tz@+7!aR9z{FBbk%x@$;b3JPO?e#)}&Kf zT+3%nPdOmqt?shdY_jK}lnVjSBevNG)2CW^h8GwFX6aAlT0qI=U5|7=$-6k?_)=Bi6mT+TSzHXk%lfyNhun z>2HjQ7_i9?UGY?Tlt%^J3>}(XoXXCFl zhA&IepGMI;qBo?IJN=7?AT5EWWWPgb(xGOb$Z*o;>}4f;D4ZAuYOS4h!lM*W^sM=^ z2Ik6^i1ynf4TYfrc=yJ+e(kz)vA3khqg@g`kV|6*&p)oE(L**AcIDu0l(-05qOf_$ z0e2B=tOOhkn08L1m0rjRT-~=N$Z+J`sS!E5EVw{p7Psc=HPryQeb{%rT3?-m?-T0E zwzmMUTPGoFG*W!xpCPeJCtn3VsrD=%T>fudrjlZ#gQ8RCi#nM}K7s9OM9I0LTg&Nn zCK(zb_K_oflGk8xUm&LUzddpiQGslzU<^651h<1* zROUQUF{L<$+O_zqp>%c&TDnCT_oVgb!TECP`!;7Z1ZUf(f6?Y(JJ)3Z3Ok98Ta-x) zL%&_oxU>cH^DG&CIjVfWdwtbXIZoH{%_Z?$tR43Fmi4n{Gv%_75H;@7QE2sKyBBrJ zEl=qkH%0RFmg`tqmyf{`4{W>_4F=2`Or$U@K!=wIm!?eIQlH&bO&Qzq{Q29F15!!e zA@_EyctI*_Y(%$-Zh_N94($*6A3;mST>ExJ28&R7-tK4KcFFk5r;h{H#I^OY%$?7FZ1XCO=x}Nr2mc;>9kBjJ?*s| zwyHHZqU?C7i{M1}n?7~*sY=TFzG(yNiV|jhF|izPck1;cXppa&ZB$Wqj(7w=Hw@=4 zal}=;6;n;09YxUQmTbfD()$eyrD!bfbonbVEzqbGHh6=oniKxtph= z+A6HHuJF#1^oHu(?T*paUQmWNKqPq@4h<){c6))I`Kou12GMI7O1p)-B_k;Z=`sUQ zhL+$N8ltSfFC!c{uugEj{Ph9=e5$<0-!$KR#C)1EjnV_jSc{<9l zJk(75)j`GiCIW2Q;@5_&SkhqEa@0S+jS5q6r*#fF0Bm|^DGzFycu_yIe2I`Jvzbra zkm?y5Yxe>t^2|xf**bMb$kH=Yof&>O7{K9xSKI-*pt*~_^mQ{W8M=8uH}zAe#Af*` z@qOL1%bHKB!spwLpu(C1+w~@8_DIvzFR`k|R~|{Qr)XE^(=$G`&kY&Tm=XiaJtR<} zWRKLQ(VJ5RyGuPvd`gwZHuou3>!-%6K}}|gWQMf3QbDJ$skuV#h8&U$N3`W9)LGAz z?E9SV@t2$ceNgGZ<`y}?@xIN?*6=g*GS^&9Q9Zx}(%kH)<{T(p?B;a4;AK;El4|s) zG^B}P+2ni)r5j+(xOa(wLiXwI@#Bo*#62n>F=Rd2n+c6r05(DvN~fU+kN{A+@7>|9 zEdba#1#bQR(!Y)G=cksTWY&u=?=I$}Vw30CQpRYX>=*uIorMPLIwo_v+!YRO6*h(0s-kl}oNs!fNw=eq#RKR^Fs{9$!)GmL zD6MiIzH56HK_s?F1`ii+GOWOf^!?-~nUf@~2<_k+d5?+mFoyOUK-Kl{Va_4y{W6ljVRZn5(+M6|4WH+g+*Vq8o ztMnV;%AhJk#h4+fqY)>O$g67SSQA{|lpX_e2RN&=ZCQupI=gbqHca8I-*G_WM}UMc z5%2g$crgz2DWjP z*D2GFJ`}DikVnx*4--^jANQ!WOwrDczaD~3{9+Ng3V7p%X@tEzph-=6sU-id#z{^9 zUCHeQ&nd25a$#-Zb-5uq-=Oze$Tv-HSWE$2a~UD^xkrvQRo+?H;oUctD|I81k_`#&qi^)PF_g@&G+~#><}L^5nMte}L&g+(#!Do$kKV)rv%B~+ z1xx8#(Wn>%v458*4idJlAXi^xKwlzE2IT8iOgpuvOHfdpAdUMBYi?b3aA4Vg1xv~# z$a0p&pOi%BD2f0U@(Y74!eGjDet2DhOqws59$w_M{5FZHC{lL;PgIGFPtkYY3mK84 ze;dpN@@}OWI?!?9qx%L+cKLsQP4eOZx;A@%b=kM{Aq9AMlg5cR=LQBN&VO730G{A` z2^YTWR2iaFTToy=5q%qSH<+Z8%{ON#o6#*AnOeQ_g;?iq=8;c(q)p`0eYzI6C4iJ_ zufc;stGmvZ?tgP@01N8M)#R39%-k;K`|&K%Fdd@SDJV?*KIUxFgZ>ZLGO?RN%J=$$ z>vaV%Fl$TDn9cK#A;#S!2^OQk+gJz<)w@HfM+oY$=MR@P-O2~B5+8)>8n4em{Sq{? z`Nl_g7CXP6ZH~ab9gR4Tl0%Nspo5!=&egC5sQ$)-` z5c3qO6-q_Tv!do9h9c6cdFY_#B7z`64b>JkS8J%WN~)@+h(0<{V^Ni;PN?XBwyO8& z^SSr?`QH2c>*kf0oD(Pe?7jBhYp=D=+VA(j8y;=n_Ccsd|89B4`tJoZI61&9RYVQz zIW3HXDyW5j@4O3T0|8%JO z&%5?U2t&N{sglF7|H6)#4E^XX$GIYR^f{K&+1Y#|*BhN(F1F9ZrygBAG`)nl4b?gCL1AHH1_Dpj-oa=M(Urn0Klv=*ZODkct;KY zhLyv{Z;1dhMBS#xvKcVzk5!-H`ru+u6rzm<(WB zTlZ%VoHD3m`B$PUSpil9Uyfmq!-_7Z2mR6$*95W%Bx59)T@F?}Q^68|C^_AcJ*yP%R^*Jy=@1KkC;7IdBNPOH zP38Y)Ax}f{`x}g5h85apV^DbiXA{Tr>jT$CYN|W^ab~b zSao}494xXNI4mC z7tK?5K3aP(r^Fe0dPgr=M4X{ek!knKix`R=u#V8)VZ!m0ky@F^a{O16P^`rrm{%ab z4+}dqrO^;+TEj&)rzu@8$F8)A(xoZg8ar#kFJWq3>Xf8u=A@}04f?rN( z;!Vj3=e+M*iWZ3Du;ma9Q->xoX`qdnKukkWzaov=JGykdCzDZ5)IO;kb2Q4F2c#23 z=p`=4CyPJa^j^p0Tw2p0PYmsyTtR<6rrVr68H*zYB8I2;HF|tvqDAoD$4bP{#wT+ zHf>u)f}>%U006l(WTN%7mllJx6S9C7$ENpQCI9JD-hriNuexN!A<_p0ySiNKS(!a( z$38x)Fwm!yA1XuP=?DIp+x?Gv!3ghzacCy|0Ehr+kvsBLrUn3B2Ptt^pcnsV__|9Q zDmC5FFkb<`8nY|r{WeBi1y0r0S(1RgnpZN+&;Y7K($O9X;&hyOWKu2SeF}hB2|ZP4 zCbk_|_WOFiK~aAsJopb8NYwTCb7DAh7QaFjY{+udY8y&i>svt{0di>d0KKBJ`Y4>j z^*+87vg5u#8#&l)z?GRVmap$r0hr7ccsFK2#Jw*v2K5e?hR(TK7t-MiIiG6wK; zAg!ypRGHbB{B`hHzBgB+NWn@-G19Yn1s4vU(OT$RAnIuE91Gf7 z&ug*~tDMC$D>x-;FG-SU@$EQVmOa|1|Rxul$>zom;TkfeI09BKwB#F> zPydd|=eGz~En9+CWS-Ue_|UV5{SWthrlVnjLjIry1|ObTR_(HVxyd|D<`0LIT! zQhx~0AU3re>r~OotVR4R;m@c{1^t!LEH8WN^N3e^^0{(BCn8*adA4HQ9)Zm@u6T;J z&eeK!<+-^YB`)uQd;UrGnotfe_@=@XQq|$EC??8O?A2)-vrOiq~W~>E)NA0-wq7q;s1h zyMo+H0z2T>?V>WKUdikUm>lo32hF1=YhmBd`ex|Q^85mf-#c-H?S#yByE?aP?Q zs7K(@?pnwcKtqr_#;M#?+RA?pB|6n)&G)IO00gT6pL1zW_>F=F1YT~GPGdA#=!!9p zOolqAfg;FeBZJ}Pa3(L*dSZz&jm%LaD zP~Gs$R)g?bB9CdU@H8KX>9en6720SMR9Ig9ilVG8_<;vA_&n`%d0yKG?D}t+gRw`W z`p;|7Je!8r(J~e;SpqKVT~!~*9!-HbqvYq4mU5};67Hi$xE;r5CATF5-CE@-;S%HcR#M1vD)CjM7wqFu7;}G&w9G^Zq z=st<<=$zv9ssj}S`9{`Bc`luV`8*b7a=0_~p;OuxEaSF&Th9lRIsE$?=R7Y~q+JTq zCYu*+>B0--SaO{&&t9p84_d8`_P?{g;;E5u_+8TeC|6 z<%6=X5dRQ~%JneJPi9!rIh*cWo#2EO+UnYq%JoJLA3qD`;uf+L?8mI$x$c}tM0Ug- z9q5OX+$iGRSk2`0KD&sz)7=2uE&T?{v~!)M`)#IY&?RfWF#6f~eloE4J&@z5m5hv; zbXZK~LT(H!wV=6O+4Zbc<}V>|{vZHMxw<30HL112c-Tba{jl*zs0E^sv!u29Sdjt~ zogJ7{8#s*SKeGgm$d!qQ=r++Wwmu*>4cfJKsj1-984s= zIvk>t#V0oH&lg3WL1cXPi4w^l6zHhz-8tqscx#sXWDGJ_C|>l}XsbzT=EE3BcgTXR zyIeao$@hrK;P+!~C$)G^?`XnCGv(ms+Hd~@;Y#!LzhGz60XCyL-nFeM z=~vp1IMN2e&*s5WI9+3lM-fR&&hA)+cdN-inc&GF$>4B<6bfe_{gveQHQ8eGC8972K=5P@vX#&M=z<~D3hpX75@I~m3n%*C_Y;rEFGX7D1_x$2{KAIAfQGy2_J$nL%)B*i&Lw5cD`qv7VYmJ{im z2Y_ru{aW_n%LcZn>`R_AR##jafyV6Pojs{dQ&E zwd4~=2F$V8GwYIlM>Y?fWC6%(vO_vXc&ks{VQ3i~ni2!ZM#+H|R%s7m*FwlHTTO;Y zrNqea57Vi_%4dDaXZgOgA6m4mygwnW7{j&Le}h%zz|K19hcC2ksho-z&R)U*kIS#k z2-x5PN|u~3a=yk}!)-uTsep^7uW@3CGf}<%ZCUoZ4wCk(iNg3m6miZ_!&EOPiT@cg zgA)Zho_;iC>!8^k5pV5sG8={?UyV zsmOV)#l0O!jNn8=edSOxysdxuY2bPm!oP6dIa%~IN~; z@QK2~80W&<4g}D~Kaj*(+P6>5BeLrRWf=h&jjl!88nG2HX`Vj;{FH9v({|t?LtfhL z(i*-uUXoW?xY9O#p>24(qf&p?v-S|*@H1Pi?6=Etn!S`FT29D5JrC?A=Y2Y4fl=ym1yC3^F9Nt%57tRFP5nB%&$1K=STy_yOHV4R@5HQt zVvOv&P#-Tpk`Q=}bR47Z2B8NbUEu{!u1&KmZsIgdH7iuzAPKXuY%Iv`ex+h8`%mV% zmOlF3NP^1;vOZPp{#oC&0u)EV(v7B~RhmjJCcQ7#=cepW ztcYQ@shl7tGyOw0DF(HNH=k!N6)ECA=OAHXcY` ztB-xeGW-M-?O3uUy32nW$m<*H?8b5~TGf1?i7O{|wLjK9!m+V@?7@0ACs<>sP4Xfr zSRN%{acVW=cPtOtEPymhSBrtJYmY)0-Cqba%;>|<+Kryf4T8=b z=kx;BRl$B>(^64+LD!h?gez65S*WRh{O`T@B zM_2*2p;vFVgMay65@iT!3@6xzTB6W7av9)ZpT`&ICovg$I?>=yf=#r%4~OQX%JTte zs(xU&oWXL1Ig_!a#Naz2&`YxzD*YwYyy`Cj})sBU#8ay%^ zfL9=Hxz|NbeL_U_6-gQvT$S`wURTJsDD6CQAluG>#|d*z`DHM}_Ub!aZ^tDphp~kK zAOGMb)&-co&dup7cRW>L6`AY}kBWlsS|(j1tj6#j9ws!EwMsSv#7cahD9-p^v~-{nE%?pur z&n3qM2l5Q@GYV$#BMM}Lgo3g1U)uS|1HJKr0?W#od~~55F8_f#uoT&Dfov43p~5gC zjoQeREgLP6U9WZy3T%4x~j`h$mw6L!Q>Gn`!VX~;f)czQooT!3+5GVJQJP2H!1zm8QMrXpUm43bX`Y72?PKpp!6 zbXo;VF`3+2CmR>x1Q8eF^-V+1@I}Fj^wz=w>koOSZ5cKgAa`d77jHeUHW|p{PnSJ4 z&-)&yYDo5_U?#Tuy=NuWU?Gdz8ZWi2X2sdHWbRMVxpAG=`zpjZv0G);GUF4^ruk^%>l2lLG=&^lIeNfiOg3rxfP>Ag z2rGD|;WuElzuk|@?^{LM7`?;ZAIiE7=?V<+%Q;@&7%Lm-wUBegS)uiy_0r8**hR)+ zjh8)=4s-=5jXVpn7#HCzC;i&wGB%ryiO8j3@>km@vy@rJQT?Sm(fL*n7TyK1H7+GS zy&MWY+`m~43;@2G&>KOB#Lx$&@d5MKG9Ja~`wyCjzK)eVY&(JZqTQr3LfFI(;)2 z2k@>hp`-yjBzzZkF{qP#Mk?Q$^E2I#XsPO6{X02*G*&!UNqF_jMB(L52RuW23Rly4 zA=lwiEJxLsm2&ctQ*%GdYhxZe*kB43T2v&3S<{cqNAm*Fvg@o6efc_)_JRd13whx4l>R=R&uCnfv&l-LtZq6>S>SPB9GZ7y z7Uh9e=%iE1v$6&6mM$yg(_%`nT3kzk`2$C(0^2yQFr8?5F`sRjzJ4^%L1&WTTE=p$ z(mSk%3YJjTJgB+@i>@~4u01e?A(S?apXEe@vvG#OAL8TsmE(CIn_uOd$ll?MqH$6| z&2P61)-v2&$Uf)$Kf{U9OSlE%6^i-jq2Zv_=MKAKs!X?*0lNqi1I_a&Rskr=00$_IApwU>apVd;if=%W;>ViIP zdZ_?av@?A|>dnAssz|YmV$p|8NXMPWOLh?|MVQCXqFe6x<7k_?@0hmrh;;lZK#~7A zt8sbWKmLRQg`d+asJ=rxIe8b57EWLw*tIzuVQ~EcHs5l09VorZFmw%GHIt6pXs` z^5p2cTSa;zhdJI9qQ>VsGI&$CvR|iy^iYc**Y4B#f`tQAbD&&z4eZ{gfy01!)8w)5 zNUK}V8NwAAqb!Y2k<>Egiw^}~DT2jsj|$nxN`4-&ZlZ~o=QeAxA9T5c^UloE0-i97 zp6`b|9me!z{b-s#mM&a^NlJH$;r6$EQFjHaG2XvdPm2bCR`lQp;m8vfx z4uEwDOdX9#e?aB&0@&#W&%g3q4k*t$j5L)Fb0Mft`aT9s#IX)17WYD)`P&+u0VH4n!K(mPh7LOI^>OXJOcI4bumf zLC+?c-X+Ut%JaBaJPGR3^yJB)n(uV4SYTe@nkMuoaVn|ggE1u{ySsu){-K-k4<{u2v&$FO?fQEg6l&bC4hxjt*AL8L9Jt>YK-+hUe!N<>!|FdXAK0oL)(iS-VBsA z(W37c^v_DY8EpNaK5@Lf={H|dE9^zsqc*B-R9790Jyw+8QYLuGr$;5y-jDRV9dd`6 z_;f;BieZ>6;#&*4vuy1!b2cdmx@D%jM-B*m6v<+T(ycM%_c4lPAqHys!(9|3F`RnT zpab9Y;Tx^087Elch(i-NDkkXUZMu6!M%1*X2YTV>8NQJZ-TZN(MJ)Z zvBrtm8q1*9K3Q(<4hlsfl$yW(l)v`(VPRPK;`ENpUY4mG$_byR#iqEa07}=NHDB(e zsg|cbbID}alCTG6BAT3;2mm9_wXm4$vQKrxVu;fE(FO#L;gy(It>W3E)X)_MD^q zXX`+fICG)HR!ip)8eNPXzEu2appH%qJ^0Pve(BoZ`il3fEDQ^+yebnb6VH*}=eky#kR7N6!Z~MFqC~z^MF`g} ziiHX898g^{?#VR*+&j=~o%IO!Ao)0Jr(Mx>t68s!MJqW$bU;)>yx`Z^179dWZS3;t zJV=TR_)Eq;yz+h+RIon)4?rnDj|4B$FZ3D)0C8u|?yyU0(1GYpY1ogme*ep}BU&2p z{wMHBw5&bB(VPKRhzGjDwgVp+zaOZr2xS*S!1rT>@9mWKW}JccmB$3~uy|RGunLH;Z0S_w8M*!X z`Z`$3@^x2>rT4ulDo`CeiLYt`TEMk~mQ7c>ai|MoBBdeSTxU2o_5;U&sCJKz6x%9g zEwL}8fnah0ztjdccceQvS3Z!>#V8-0=mlWhfgNl#cv~@~4e;mXGUSDcabi&Z{X+_% zPOpPKIfX1)ehG;xYBWk`UAMK?(MWFFpS&EK8kUG?8ipue&~Y$47bs5IN%gz zx}3Woe3Rzl}@!X1wpu?8{2I`@NU6EH}ePw=-w&G{knk97tf<6fst<7SRe>CY>eFi zhE9cM*DJ8!p{_sHf%i4NS*3yQF<#EXe+6&pESz6rtF(7Ven-#=3$tb%`1aB`x7~>A z*O&8jLFi;}caUcO0CFMw!T(>`e}H}ISX9!WM^rRNLi-TO@@`ifho|rA~)wsc}a?MtU##DIAicHBz>8q*fhT-pp*W@A_9cV9) z4u`brflW%cj6a=9^qL#qA$_N-A!+v{TZZ!PO6QqJ)=8d2aq|{sdr-w^mJ5&*d)Ca* ztL8pJsbr#d>64C|3vE`HLO1?#&%63EmE3yTK%BkdMDa)Vo?9zX7U><{+M@Ob{Jx=& z+GZW`(KwM;pXo~L8gPkx0e`T+KpBH~SM*HXorq;E9on)map;c+Npl~z5#dNvnX}Oi z#*sChID>Z%rMQndwQr7AIz>#dZIs z@qp~OTjzuea{b8FvKrynI%xQ#cDso$oAvJ=RVPS=UJdI}nSf3zI^owT;)`R6Z^}Y@ z9d10)0bLBMJThkGLjj&bR_61%1iv~f@7R_K`M&E>GG)}!CrBHY@qyo)SgZ}13w;PAC!g@IfM zU60HXk43z*SpXL*1!Q1&>bosLUJqyt4#KYThr&i%`(8p7n7@ zt=m3HBsY|a&yJPu@MG4j5%Nm?)SUU(+Z4>#-my62E4~jLEIJki zU#XwIT@re(wYWg~^zFh{zs}0mBvqWgf8i|-%_nghkoyJR@lHPSpRF|4&-A`askxB# zkEgsZ!NC9sWs^s)-16&uasR3FRrS-&C850G@AFDA+wTdcy=(P)a+WH~<5gAv_GWuy>saDNvitsE_ie;Snp~?>c;3RO!XM@6znIsw{GkJ;pggt zU4wVSd!3MZNrLsqUP^8XkL-w4tFuXc(S&YYp-YnsMdPgP=={s!(t%hi`?K8Fxq~oC z^6!`%i<@F`m-V1F46j5wq|&z`X;^(SBbYXgbSD&H6exrsQm!uY8JxhiWB_^Q-t_qU zSN5kvsxV1Xb1eV!VWY(3xrkt_ryN13Jm=#Ml7Gk(LK@JF4O=>*wkT%h0{fr$qZacQ zi5w4+Ix-gHE$3RNUmTp3cs$&jE1mu*7A8~V)rtZ3-hVqusG51;+3@6}{%S|7RgK+i z;Cmy_5#RJ^WXa15Q@Qum#R@Ha{O^-i2g`GOl21~wb8*nm69#v4*U14NrS&UEKUSqe zRytZeD5+m=!@@mUyhCq`{(~i$Z!`T9cw26WB@2r)T~EQT|5?(d$QSqxDFjj$YLB`%g|F^ zv@o?>blYh=F>xSLs(MWLb(bwAfT?V>VZ>4Yb_(f9fDpu9QOOo!4{;YGo9TfB<8;kx zT9(#7jiE*60B@v6jSt~$ywjpEEes~rq_d$mnYQsMML(JM(!lnn5Un@MiaN|yrVw~E znEUW>S8PiUmU@syfSHK=k9Yx5KT>Kvb7<&>=rP+`?~P;PdiEX6o8|eV6rY;@o!dPc zi*8MJ^sJ!vFFNM)VWNp{+mCKxO39Ig{I<~`in(s?l=@Z~Nfp>x%&vkJ{?D5+noH{I zJzBB%#c{ynC||UiXB~!QzESJF=nGumNqE_F%ejH<5oG^b4jau8I5jKV^`bDsN8Jo zfwl1?H2;m^T)bl+=;;(ifRz?xQ)~Xh_a?5ceW8&N z`P4I{Hb?~4?M*Qdz1BEeW%m{K?nyIkdag@+FZlPf0e;ARrpj8H&09VX6<3Y^eyW?l z- z4$K|idSi66w!1LmXbO2f|6;_k z-dxWF17s(@=M6)R);6G?V4$W0{iOzwB#X{@&HrS;7J`0zELpoFJB4qKCfP_FFQUa# zo%~I9jZvomHRWAk->W;GpYAwRmy*QPvwafY$?zklbhMf=J!#K}&53(xFG}jp(Zg>3 z*J9=30F#E`laT#S65EXLrO_A_A7x{*G;E$vwx2Zp;uS_o3*iR)+1_)Il~1aZ{6mn; zce|D_7boL-?23y~5Ova@4**F8fk2*t>ehDoL!C`Frlx&XO%|k{bK{vCO#YVsk4+_d zY|+bF-zGXY<9XOxse<=~FYNHPd9-j{*wF(DZ5eQnkk?|fQIwOD3yjV$hT~5P0cFRi ze{yaZ6WA0{x+1GS5F!Km5{-z)ni%22 zd$Q8rC@NA!t^rwFoxsJ7L25w#4qHoiP8S z8zbiE5+WANh*6<Mhx_6cfPTuUyXNI zlu%1m$UQ3W8(1Br=FoSvY3xEBMS6Hvhcu)!Ifqhj*KNcVE|w0&JD4>zHLJ(yv<;JX z^c=?YpaVM4d#S^9{AxzH1D!M__mBFb<<*fEC_a3oIcO^V$9|`m^_>FNyYWO`zx@#a zP7keZy2APC->$v|XmHEwzrB7PsF=SXTJCShtA11j9by0beRs$`=b8=RDIFW{`8Pot zogaO&P9Loh>p49Lj`y8Bd59|7^(n~{Tl#&wv+BO&oev3zFx8O{ zZZ4Vsw{7FnQT1})R%kN(4LZ1za*hxV1qKhC)tQKfqPM=byQx8sR+dHcWp za^Z^kue_u6VNd(o{;`+LUY1)q?X{Lwg3LYIu{X;;@m=~zjDmV;0R;{I9ogGS%SM^L zK?&OgW;OY(5kEGyCrIWB;4*q^AW00cdrmE>86e+?ae@5~2;IM%dv9=^)nZjGUjvuT zax&D#Up-d-xhwEj-}qo|E#EEK=Y0=sR{;{(JApl4#xq1>c)}w6U(QO22XVjs##bC8 z#4lc&Cq(6twJ3cy3Qv}xE4D_%)7)Yut@CP02k}NR$|}TKTc3U9UwRAH%*yJyRm)c+ zX|o(*NP4WM8R{@->5i0r!R(@P1PEG-DeJYG9!D)_otmCxtR(J#2Z-Bmtt#(+DaYI<3Zb9 z15c)2ZBuj;4eDTSy{!Y}EewHld@ltIUELA;Y$S7 zK>lC%-#-Wz!Y*{3VAx%(BzTP$M#pjqmRo9l}r4c z=M(>0L6H&Z8=OG#=7A->4Z?~jp6q2=FnEM9U!Eoe&%(m908*?1rlBE8<7~*VRnw%d zM>&4G{o$7>SpYU%z4YdIJhxL|JyKmUo%gTKK%EG^4J0GQNcsco;IRsFESwA!f`t=8 z(LiL~N=Gq!d(*rA!ZUOao&c|j?v}D|zxn06jRtM;m{1dMrYShHuVJpNKQN!(wqtZn zXe>jOa>V*?30XLWtv(*A$O!-+;@e%bj=3u_sPyP~3VLLp$F zl!eA~zqPpLnT8tX!18NMr2MBEc;l(S)EphG5cFkrd;HS{P{qV8&nfYHWWVI4>ENe6g_RMY~ zw>d`Kr?CG_Y{tvAoI6WKewsc3`U`}aSdJ}mtCUft{Gx5-wbEeS!}R44GXTfP3&1+% zUJhMQnJa&{>@~}FrrP|49ZtYqK`xJ8`q{f8U+h%7)N{INrBr6#SjHRdE#w@!ZJ2up z()cOC09)8&a`2>4Z<=1=fSh0cN4s-6N6tO1|JyI_hdT#t z69iVTA@$?^}|Q|)1qhW0T= z)=Y%^s($FxJ~^VNVNh~xrmno)`3yG=7@>j21k&Y473`I*h}ly{6njev-@02#q3`u8RB>O1Vv~^7^~BiTU7RFe7dl94qT4!vK$|t$xT8X_!U@DMUl21Ny2PL zsA^_g>x8}8D^EMUlLwd#ooBRU4T4nNiM7uxeI~E!XxA>flCuHR5 zjQex`eke_*??~d}6_azBm#~K4w_PF(5>xqdtD{lgP zP2EN)YlWqhg2P+pVd);D0+SM)Qk}*yCVv7AB`+HQR8QJ;PsMX+svQ-+ zLHB8~;B{@X)_uoxX_+#z!D{=o^S6drNrnO-i@ci$%0{Fzi^pW2VX{5zU5H^CLwHQf zxq~c2@{nZ&;jgUi9jxt``pMO+dz^Jem#HG#;cwjYtt8JCO&R$HUCi$@eA*w@a9o~H z@x?itU5n#ch}l7qc9Utt5CVv%1gvGH1$W3*gFHr-qRU&)g#(OSLH2+9rcw*D$le`= zG*;e_-j%fPh&HL$l8s)i3R}yFEToxia%H@%FLJAT3C}Y^9k{d1+2VQ)<~pKw=|1|v z#CX0~hg>#H9Y`kE7-$UvoOo^=^4fSl;$@S$s2IbsW%|MWa=bK3s16`NJ<*S;SE{lE zfV+7-gu{W|;V*kDwK&uaTYi0NKNz+Nsj@1532=uyCIIpCec*1Vp1SiA9dkNUfoB|O z>Q8B)hGx^e6NAWPnQ#jKp^#GBI^oi$p+n`_6_%V^LyIBSM|_mjoe5zJxq5{Y4*(OQ z7zN;CtLOftLR^v@`@Mk_hL+jQFTFXdN)1mV^3mJ7W9KEEx3om$Vm*KmnGp$3E zFKt7d!g=l4oqU73J5chlDoc5x8aG81m|uo1DBY3y8pG>2SisWqBfSVDH!T?D+9gxIK0b@sd;F*sRbV{T`78A!4VGhKW-JS+YaH7O2CKHd;jG zC|2to?%+o5Hs@troIZSykk9A}>pkYUF;!N zJb#qGQl9^M(@Td~Yusi&S*AvnIk(AI8D27zdjVjC_{}*}b8c>SY*muiV^yPG_TE@Nw97MN+aotu!Xf22_JAuHa*h?gMydc5 zqA#we=Rt^jw@*d(>!N;zihR0YM|2Xj-=~LIt>aWDs0YfMyM~l%z*ov z+<2OJyWpL`%Fn;J)r-9~A1~Bg5~b0bBQdfhQ9+!!uSt3WPxorK26!EcH+c;SI-P@B7}D4A+bLI; za3KSWJC&qp{ov&j#NS$CY6*D&zP?+8%sZ^CyF#^)W}ucQldr*-cm>5Nq6Hc~|3WF> z%!et9xfM0yNK|PNRD6zQlrl;{$zr0-We_DAnoAl8$B9TeZ?W3AJ}^`Bej4H z{XMx|V`=XTJ+1{C=3RgjBR2~_-x@fTUpT_t2yy8?1|-!e5MH6Bc{?;o(k^)Mv6fZx zfEpE`jzo0Q)i7D%MH8JmZGFZ-Zr76W3hqGE5RS^DpA7j!z8^1jC(Kt8fNXXTlK4#d z2q$O;ibf3uA@Cc^CM=D`>E6Luh0bC1g_>O;?MOMS{h*5aH}xvIFh_#E2B4WuONY+( zWHp2gxV|k3mI3?;BWI$D3oJ6uH69$Ub|@BqpN*#C{;ZPuoy{35N7qZQ-= ztDZ~mM-p#*nymw)twWy!7fRQjc;W~KqkT8#1lZVaj8nb}#x#Z$3+;IRbwWB{cDXZz z0bS3Nq}ZS{&wT7s0AQdW9{I4~G?uA{a{c%dC&$Hf=uiNnj2_9t_nXE#b-1uzIykM~ z+!}ah*!pTgX*_@F+_V6@mY_w96cD`hVW~P;_|(9-nD43is<7i;3b>~Z6U^5HZ%7hrU zE2D2nJPufy8}iv6v-4T8dS!u5k)zdOz)fWVFbEC~jefl`Cra2_egnWHv9@2e^d{GC zmmPV~=OLXxH~w$;d#V?=$pr$nxpzw@i{up!p56BkE-HB{EQxe~jxbCY`a&N16*BN1 zzz{y8kPTo3cqG30>s5{S9%AVwbCn9uFX@diwt*}-H3Opr-FQv#KqwkIZ-%l3LSDr;Avnm z7?Y0!#q}QyJT~5_% zA#rBRu5{}$&-9|EVaVH(HXgnZ05c9#vpFjpK3*HCud@{3lCBpk;u$D9?PKB%0GJ<) zRj4YuGHkPH1@%NO#UrxQWS1ctgLW2z$pN!l8E3kacf(^$1IlyK|Lt2Wo zXUj-Ac~L*Azc{4yLw0N7urqY3_mN9_ngHwamssQ$T|cDcLqV*tg_hs1m$UXkoK^Qe z&Po|D3*!0Af~4vT`=bCgl3!QJ}k(+mr$Tzj!nb9HRiOkG!xN! z;revS^s?VlZ*U+`R?`{y5>UtXT!*ZwcY-aD$PrtANf7Dx!4 zgb+gSolrv;kRk%ogh)*Q0jY@vB4VM08kz-B5h)5NU1dys@vlh%$Dxid>b%;igvQGPr3CB>>-Q&-pD zJI)mS>c9Yr{X+@yZRN=|L!kxWxdPqn&xD4N1XmE6Dso5DPg7^sNz|BE5p0NhDTbZ5ghd(L+R@kGviCZo1Q5jtG8j|rm$Tgr zw{Q~)nQU8s1F%3&XH;;j@}X(?mc>$6a1=>;ZOcb~u|$CDsv$)~g^_LSvY>7(%R#9p z7y*5#kMMh;=ZFJ1AlX)3@L*VR%jUUVLRwLdqn%RA)HZA(W#}>`x7P1>7!XF1JB~%= zA#6GY)-r#va8g_FM8zM({$wrPy=LDCDQ${tC zoZS4TRk{jW%T(&&N&C2p8jS>iIz$b# zrJ7AmZkzC-si1AJV#Wn$v-h3+iBku0S}Lk@H`evId|`$QF(_?pRxl9eFRYSd(>ggI zsOOulVY=chJP$1q+DL5hKE+tPVXJk@eAH>IPEpEh~A`yL}IQ^axDgNn%% z4BZH=F4XtUS=cZY9uJ?)PO&Tal5xRF;eD6Vqp5ftz1C~8fHGt& zQHIYSCsgMNhHRK7d@(8|XTuF<2LAYo&v?{{hR3_O{Zh=MaXZw|ULyd7-6EoLVRwN# z*$}(dqv=fv6r4RXTh&16lHDvUH&9#pE8&?y#JW=xUs4`zUr$Ab_i5tbyyPPw_00zG(l3@gCjK%T-rr4_8Go2 zcfE6-dyGOmsN;%?t5T}U3}Q~_4#2<(j~}tI%NC127b(ojzvWH4EV$Y!`y?Yc zODNl>we$Gvs)BU9QoCE*y+<9Lj_ouNszE07_J*a`n0-H=*eCTIaVPM@&E=TKJH0wV z%hX~3#x4!K_rfrCn_p4w2P-+-0-H-&@*Yl#naqycG)D=u1yM!ll@ZIq?EEgcZAFEn zFl%B?#T2Dok6a@u89z#{tA*{E((Kc?NtAWSyG*=l4}|%5dG#LlRVASkH;wYFo29me z!f$Dszz-8J^akY1IWmj{{45eCLQ6tTK-zVI8xwP<1sw`Xcv>fQPuWyRLxTf1B~)&# ztluQs)X??wmWa3PYg3<&ifqe%AxSntp3;J4pK7YL5H7#`q4*}s6wcxBVE?48M}*a< zZqWt^B4@BU%eeDHg{))@lC~lH(J3SbX;~Ngdmj_`;!>BtiQ+od{KN2XHt~fgL8;UY zV$3u$e?Nf7h`QHFD{${)>raW(<*SOG^Q6*{RSDHO$8^R*w&D7(mcw~AOJ%k&*qSwm zAQ_eYg#>%~ha#V?bxrK)UZQSeAu3O%OLpzni5+JCfeKEjx`qq2^uCg^lba7^2KG5Dy2! zQ_iX4C+4zOO&|=_EdQ`v1ad=c8x;-=0izh-Z2zBfGy{sv{#+w68j8L)S-&2{qQy(S zkVtg`_j2}6zYm9|N}IEVuI*x1<)5OO8;l4v1Cn3V^ZlshvAtcczcK|BcAcprtCv%w z2Ct=~y%>QpoiO7^LFBIHVZvkT0#Pfd7TL4q(1(LKt`rEHTBduzR8%zflq83FOL0~$*P0J8sur-`__Z-^!$D6!sxKvv*6 z)g0vq$>!oaHUA;;Ur_@;9uq?3zunl8RtsukeU2YJ#bLzXdTUQkjW%C8eT`{W(4e$T z{rW0c_@8XZs4VQ>Q%%^{i-h-aejE zFGs9jc`t`QspmL&G4~PI?&B!Qu4vEX%~K8qm$|Y-`u+Tov~+^dKcC+2GZ-t>amlZ* zw~t1}77rF=kSlhEO=eN3H)s+>MRs-tM(Qj6#yBD@7y5C;hTH?VT}WZ4rLv{F+3;3p zw_h>ka1F@|${Z7U*qlT@$J}7}W%I6_Z5zo8QRX!a+T0(;*jtWEzwqypqvMEvL?b)% zF-^-ZOSTc`uvEC#udvL>aQ)V+EH1U5`~;$&m0K%w@I;^6UO5q)PLV&*0tof$l;*;b zLFKc)WN|O|LVNw3*E}FGNI}xYzf;iBrFh#@y#xTrr^?3GdrlTz_s|6Z@+-SSmV59Q zJ)ma^^-Nd0VhXTYBAf&&A6g2`)&hL7J>vy-h2A_sw@i955A=;&pVUCKa|1>au3>46 zXvba|&Pb|E7Jo`51*($41uq7UGBc)`FLhi{#H6t@?wRd}DuxB>bpQZ2`<>dJsb&-D z0#TdI3DH~US>f4=B;=Pvw!0?_o2Cp+0r?*w8R683Q3Zg{d_=GiA9>pua&r1X<**Zm z|KK1c?p3ejmOIFs6x0h$B{bhcyayg(v$kI>zI+Yq<<5VZaDOot4^TH5kZZfNn9rK2 zP29mGi8{p#9Jh1K*q9H;zSee}B+=BMa|EyWYR=mF!|ZhpXgeLFizHQQ;Js>M|Lk ztc_qgTPPDE!BO4j59Z)2GQzTKi5UMw1Bhj6DmUdCP>wwbRK~j6>9fw}p_lT72{v~H ztyp~k*D#!XdBb*4MwfbzBz{3Pr|dxog9ykws23GMjPnXHzOP&Xt}al->jH&pqix9V z?4Pcsqg^EMTJ~YWj$XF#7A2Udpv6%v5C!o4cBLWT&hENctDyT0tsK`NxI4x-tE$eg zbb=)kR>X)oF73@jwePfUcPh?O$dX`5{&7P{U(7Bh$UEc{`#}q^*L!qhajmtAiM|~s zcc%<>To#8*Mh|Ag7xnu98Yo7hO6L9~`XioNZ0+Nk=VaG@p~z9$ii0G9gehIe zC>>Y+mmcr#|3nL+sfUtW-0(-#a{M*QO-&A9Pyv%Ui1rCaqxA|g}B_k+?%{!O$g#^+~8Y$*M;B!Bc$4aUS$ZE8% z9mT%-r`S4j*CyLerTb@osiqa#(U2~P5d_$ z%pWAfwL}1izX$*44WGg6h7EpywQ!#GFCiI%*vV4EtwGt3blFUP3pE)c>f?pT*B&A^ zvap{51^V9yXysJ1yr;Xu@|7Vq)XQAT&@nmb{(x(IUyG~t9UF0CGdQszoD>kMS)E1f zhqrzIm{MJG1^7XoAc+`ro*`ONm$Hmi$SoC82S?hnPmQrMz6yBJ4)=Ta3U3H!@?Aae zI3W0g1sPOUR0HzPm^D5D@LZDq##M~u{aw!MMk7DUsjwC4 z7o~d$IYavke(!`{1*EXW`b3Nv;7_aaR6JlzQXr3a_0+;zCYw!c%B2HUV+p3%vVn4` z8`Myf@J{8Y?z$gGk;k7JnX9S$7m0321&&{p4ivdNY3Mk^a#+mjVD7;lTzNcl+yvRw zu&Y=0SMC=bnlSp=@EAeu>FC!Dg#WGN(I+Iy@P0{U|8sk3Xl1mM;Y{5g{E_-l6BJk= z9ZFZv2{EW#i1Bije#yICW61(w#>D{H|9Dx7dA)b<3mz1%V&r%RSlq)`{sFdot}*C& zSsw*ckJ^8R_pxFC+f{B-r;l7vCzI%>zu<9>gBz66%t&y*hjW}pAa|oqzZhp8Xe|%b zgo~sxJ2FCY1_%WM0!QJt9<`-i47w+xM`QlE&f7{k3``Nb#tN zz%0AuQzGU}cKEoCUdEL@dfx~3&lSOQ0*WY$mo^l(A?fPZ*Pk+I&F9zTFfJwv`K)6( zpCPdiE#)sTk&lOi>>7Tb%*o*jg`GBiAwDX`Q%hmouiL$z$GV)&=-R(mf~{Y`x=kue zii@gaipOjoF-xd8cj_}iKJDE>^ir8s9Wr(LncFmwu`{B)d2TehfHN`IoSx3ST~$`b zUbwZ7{V02kw0rNPP9@=mdm2NAB=MHb&D!xeZq@cNS3>!pb=CZhnijUK1n9Z&71==5 zl#LLA@l)E73YpFw9ZcwBs(2=&#tfu99NGmVQe06qu3r-E)Ppi$D9ummI#x#C84q-6 zfoLM~Ia0IF#JffH&`*cQHk7>Rx^o2wGUf{h2zqWT|HUFV^1kc`RoT}@JQ|b$13Dhc z$gDVuZ6eH1sW$H#8U885_P|SBdX&{BpRtOp|^VAS3`FCmbmZ!50=FF0#H>b0o#9NvK5bYdVo0jVpUuS>b7~DP%yGlcZ zQ5XdH$JtY%fHX>(C|C9Ixi(ZV5KEf3FT+Ohnb7@b1Df&f);=!$x)q|ke$}UcAwd`q z?wfy{8d%H5)HvraB{MKyvjYJp?wuYfkGcdJh}J`f3kxPe8*+DkDmA^hN3K%mU0G@y zPMc=NKwU0xV~wkBHpfpOO75uX`?V0#YQ25QRn$GvnkP%jil>=v*wW9w;1NJ`GkvY;ZGdyM@V?k1$!Czdm z&k?5r9sC)RXy-r{^3b7}&#;2*;^QgwFkqx~5Yp;h|4ckGw61~%fwC9(XlPGhmfBJT zM7QFIcE7T(GG(g)2sOcUh*DNdWor28#yg+=b%P2G6hUrC+rXZ}fvKU(@Ap`jqn1cQ zmwyTj6Y^)HxM|Rf7Yu6H(~Hrts4!=1I z9z&3C1jjPXn#a^V!wR7@>uY14*VNT7_(qa|qy<2=kv>hRT^M z3_>f%?|-eA`MPogd}{bb_F-a4-RH=D#OrgLpt@iiws;K^CyW%@=Yg#9tJ3g@ zE?j&_6v-3)MWLKvV73)}sKx!@GMqder|MZgywIhn?4F;sr{2F)#`epE>>qyu|Ce!{ zYB3M+O))S2+vPijE%5`3;_eov@uQzJ-jC=XYN752jv}@}m&HmBp;Ve_HUoXWz}Tta zekY=hTtI!S0grl!vTF&`H&b4DbPq^7i{*jWzVsy=&^7R%EU<|8!-P&6AquOqZcsIx zL@OV)@qez^i7XqR$-WvT5V9d+v)L}o5h|I}`W=M{oXkG&@nDUlQ4ZK;w^J9C&jSi| zNK&NdeJLULN1rJk7ZQ$8d;`C2uy9n;X2!|5Q(va1j8dIfh z|6PhIeVaatHhs4w=|%ptw5JG9!U3 z#{l!vDby>xy-8A;axw%yH<4JD0xz!BobVc_5E&4b^!%0%Wa51UQPv;z1JpH4>)cg!WJ$}^k&OUrrt^9<}pr^*9#{H$@ zM8CoXqCQfvFWT;@yHFGcvP>VCjAx#ilNbQpbwdgzcM8A}RIO{kgwzdgREfpW)eE8X zzWRNDx2c=6`^7(4vY1ERo*jKwJ3&{c20ULp9gA{_GmM2Sm8QP9=W~vwGKl-y%d}af zth26U>l4+}3cV!1>Dp|j)T@vw^pcv6DjwPsg={P^aBUdL;0w>rfrj-Nu1`*3C<)He&7hPya&V|0M1d8QvA4yNU#Lx!@D zi)7>I{BfFN0kc&uAsbK zVHsN+QHs#wm`<~TJy!2SD6YDdb)kM!j;F8H-#;i7IA?pMOzTMW&|(DLDf=+@U<|ax zTg_d)w0qG;Xnj=ld#vzN>Q|zw8_UB^ub{#Cm6~1h*N`dQGqUftzVB1}C;97LSI?L| zMc1G{*T-R`$&Wqt0lxwvA+j*ok_bxk7Ft)XKah=xP z{W4~qh^r$R_Q}-qRg`Z9>V@Qfr3($-ze`2{Rd!H^1g&$2sEL?64sx=19jAHswR_V` zg{;rMg(;bEdDm|y;-7a};(JCO;$XdaJw51ScsCBJSH9D8Z@`9+7u>+P=>+XL4`V|R zJxMWoahG7M#ai|&LD1PDNuxi%cSr|JQF`L1)1%P}yQ!9~PWdAt`2g`UA5kS6A*F%# zN>LuRbJDKan3t}~>C$1C1mb~A1O0fjBQ~8%4Sq$Y_Y?|^Rr18`%EBRWTJ2+A)F!`T z+j2{SC_$gj@(U{ky)ea+jiU@oouQf-1qEWyuRT#((sOwiHpBVP5>JnNPV2P-uBrzZ z$=yfZ-!F-GvCdO1?XHPs*#&}?%uUdb8RU4{37-7j*O8zWK``{}!ajud9JvYRBjiBt zQz?BnM^5yOV@8Y_or~&UI?vV^6hgOBW3PRP05bb++-e1%SC#B1?O#42-GoQPGT)K} z?m`#i!*V4W_^j9xE9VT!j%+g?{7gZ#W%~%AN%vbTN&`4n1Wq&l`f$AU=g~7u+Ie5u za{Y53uaF}}8`%g0Q}f=!Kpb-BVVi<{rbwi#=W-0e&`6d7+ub0qE~FVh=@{zjrc#)4 zulPlJGF^lkKVTISb2cdy?csMCZ)rOMB5{&f(-d37ArML!>MnR#gH7ko&uSD<2m2Q(K{@nd=N*@XJ4lK^WHu z0Llbb{rs9Fb@&5INHEJ>L+F%K)4ho!k4HSbMnk!K0U_lBRb|>=%ps4szhC4?kPKHc z@XY16*2VXnB>i^nc2b49Ct&;{y=jIQ)eUbSiaAhu$uKzhUTlK(``Cnocl2(f?mW2? z8l&T>cKk4Ac>l!jpGU6teO){}K|j1fS^oM%E6s(^6j9hh00-4V%r^FVuNJqND6c16 z?3z=MKla%#fwPAM=XA=LdW1AWCZdci$KquhoRiNIbnmL;Eh476J-73b@=5uL8=;H& zpV2N0SU_V{7)qO|YU};T6Y4$Nr!?`RKOVUUWfjGyPWU zp7Q7ByjtbhgNhBWo=7_{%u5HLMeNaV+q#fH$M=0*N)v3S)@QwNoOs%B%6NUq>M^A< zj>)l8qqgG*Lvx7WtzFJ%f=iq~Wbt7X_ug2|MCCfaDu4Bbu8R?p-_tD{K$PA`CugZfc1$Z>>r;)>`N};cnPouw18PrHikSHx#G0K0 z53mW3N2OyrEsUJgkAYXY)g_765p=;AUmy^sOu~F3K8$j&qWsnCa;ofy({Y`?$@d5j z;;zo(yCcv~FI=E1Z)h4k>zDqJ*Hcw;0NE?I<(>{5E}a8C{{-P4{ZLBAmf3FD9Gaie z$Ch|^Y1iDw50SZZ&?Xi?aNYvu!7W{_i+qdZy2DN_W_fcJGG+98+JXUQfwC|YKX^LV z0Rua(QjEGqnxpT_x*yFC@hjm&yr&(5TCohO3yyCf53P`Mh>$abxmI1%dB>C@$6jJY zCZeP(6S3mPQT*T4^7SaR6PMP_`XRd`oSnrRxZwuWAJ~rGHm<@k>tnl@V!SFZP&Mpw zt`QgTCJw`>SGkiyvXI}F0t?algKa-Ca8d5L#(hkqYPEHbOhxOq)Qj^DdBU%}-698T zdm*x%CKBS2MWoQ7s_VT?Xs0C+?;m$S2n-YLE4;ME(JM%$Ib{wz8F3Dh< z*6efdb(&>F-A0vmP!8WWCPK0h6XF9WVKy$TmNpnC*< zQ}6r?gahDfu-2JlPevt7WX}>L#OVqlky2^YZ!17{xBX4BSdUp_t*{;d?Df;^*g>9y zHu#<+B8Yvvm3iNhZ78e5R{E@FS>nIUKEtEY$5}s!h>6S_UfA`9`-AR!~od9 z72cNix>SiAoWpIvX zXNoV=9S|jbQ*t>Q3Q75Guj~sw4bAxPkM{QSJGzm3`;=zuAPkl^Ngxr8d`8r5r{V%R zXoXpuZSA{w;m2-L8Q(VAM?5pZR;&XC)Jh#jPgQoZU*1LR0ouLO?nHxLz=ngL$qUbd zZqbipvs6Q3g_eeG&NX;Opo^x&9#u^DC_VGdIt0iyM*Qhe>DK4nY82{@ZHPWu@@{p9 zD(iE|M-2^)bBXpZyP$muL-)$d%R8Ii{4kQ1j!j&Y1^%_|RZ`4#J!{s zZK;Gq%gUK!n~+tVzXbZ9R9@eCqwTP=q2ag(SAxX=Z8}23lbHR#S8lg(>~0 z*W)brm4*ZDRAgX%pIEvv93Q;4jPPO@pxTrQ{R+={6rST{ZLJ~$SrM||2_T#?DxcG3 zc9$blRhIi4as8zH=2+uOJ6;}D0@1J|U$IDbY^C<-OqyyUa%efbtMqU-~vb36d@Gg{VuLpx9{ zKfodq)Ic;|8b&8Psugs-)C*Rk6(c$hn9d3Q;0U6Mi8=sQw=B2MulB%}D{O{kuDpZ` zD|psj^J`96n=P0rtvRP)R=$Bsd)7IGD6K)JgJx9og%Gs|hKe5@4O?JIwT{*41eT2s z{cS47PrG8yzCs$PQWGTHIf&Qi{%Wo5C(jgwyZGj~bn!GMXo0jp`M7$Hf%x#A_cw?2 z493v?eEYqB*qHW*g->u!9y=agm8V$nysEA@5??u}KE3`K+r-fuvjQf()GxMSOLk$+ z?PMTNJ_6C)j-Ly7wlt#nDUCXV+|u0~DGqU$0xV&Zy&~ZYnRa=HP{out?~45prIhd9 z!w${EC#qUj-{t@GEeS>kFlHncaJBR&!^mR+orN>c#IBC)ZyvTHNX*=r5+*yxeU@%i z%>f{BpJ8~Dj7NK_3LV+OJefB*jb0}pLLJ;Hg$&3A&mmJe`xTeHn}@YKdn(_RP<77q zp~?V55@w83)M{uNY~r?mFB#SC7Yvwm`6)evJ3okDX*vpv0*cgU99hh{yb)(Giu*t@ zc(@^~eD7B?-9|A$s?-U9K=>;z$W)a`qZ$ts0pI28^PHoLv*OBDrt*Q6qsMcP*=}JzP&bdDR zGIs1@_FKT{Y49b#1u7>-m+u#Juf3di;&d*wsH>;)1xoZt*Z?kZTi}_WX0DN%{`P|T z;}O|&@fUw3$kumAq(Y1;(j~vg*Fb}siTr8Pc0Zr5l(ptP8GDpKq1y9g+}IXJ_c0A9 zBW)O=`>v5@AK4Y3Z2XQyIwcR z@B6>A48pRWO}f+I4Pg=R=#s@#B1k+-bffj29t!6w+1q=!V~O1AIjGrJQ+ouLvS zH}fZ?Q--d}I{A2)PrR2gs_;Asa`&C&Wv{|%BXfqnQ&ayvtpWc&zDqIJW z%nYYj^4X2FIllq{S9B!Qv?+viy8{^Pp&XK(LUKi>}YGArBnO zve_0BofKYRAu^G2?0jUe{A?5 zzUPGE7{(X)l2O*%+Bc`yNgCTBBCN;=x&TH#lKui21SGQT+dSK++{g+)CZAN6(<%Oy z6CS=7EIzEo6RUaNRYE(aGwMUyVcZ-Ul>FLZFlA1F!BSkzeeoVi`Ale;l>aGa%-VrE zy&ON!*Dn(K;BtxWv<3fJH2-aa0vufO0B5%~YQETT*rOj}%fT%cNdSmbBYIqUr+&?- zY0rTTU|$FQI7~wZQPebAcb@+ai<104e!1{l$V5z7$*-z13*Pm+mNfKA7T_QX{2nhh zQb|&Hbi__R6SjXFmb!xaI4XyhI?nF6WC?S0c03nlqZByX3y|tJ^irI!(+_mX_>6=t zp1$*5Z#WiFKn|IB(~lacg)Fd%M~~8^5QU~;`7azogb?JY!M4|KYy%WI1u#uIcP4Aj z2_45<`~pN-1MVx0*FswSuH%(m;NZjD(l-h@P9#@Gw~RyKK_HNTJV^$u+GpPVJC;YU zAis*nSWM_W`D~Li-5vaV3!XA<*YX^juV4{*XI^PICR08%%HrdQ*!!;iEtg=ov6oKrXzpMAhR zj|e7d3>HrngtM-p`(SI^&+CjN_j`x${@FWRLE3SfEcxp?WESU$DpAPz<5D_g;)I?9Ceqps+DbtB zWe2{4eZpYmXjQ{WKnMZiSCrW4|2c-B_bdCgO-DJ`ezucqlAl{*PH#fYsE4U48b6h69XN=x7D${!rUh#uzZHP-sB zzArsbQgoef;twmh&J!={d5rL)0hoW$Z7Fbi)Zr&)5P+@_F&tO!x1jRJek*kTqEHap zC7wvQ^xb>(Cl;Vh*qXlA8~sIsm=Ohx#;*6vUSRu&<+X54mHz4czpceR>D$^fEU%JB zO!ztWiLf%hgw?zpu3aTt8v4v`Y)&vVO5sDXwr?S7D4UBQ8%{=kep&1~x9`VM6*xy6 zI#n>*44AUJJhylLsHFOk#3Z6y=_M{BhwR8V-p@m3@tQUC_VM_z$m-Go?lGuuuyS>w zD>i!%+UxU&e2acKu!b7X0A9S|#sqx;iJS$T6umUxi?$rA-95rrgh?cAEXLMD8&$k8@f;XMUU^ZjPpQ3Nj9cfsE^+>umqcpwQH6ihNa>*<>i^ z)*S3>J@hpo+Vaph&-5H#QLlpbWToxkWhQ#CfQw8seb^~@YE0wk+UWrx^&bbl1*~bk zqcd}0M^_6yqFfp+o-&G)jjxG9<%OcDdDi@9ov1S0DcP3ffUtZE68~|P7UCFBPiYv8 zxYm9WBca~~m#fQvv!K`!6{@%wF3hr6Bqv*_cBKMeyPBrY^lpu`y??Fr8_pMeb+%J~ zXPpHCp-VmS(KltC7hdp10{J3?k|V#*5%kl>WbG)9HxuoZs40k|+{h>-JzB33SYNW; z{jRTmll2*n?w5znyQM?jLH1o{fY(gSp(L&g{=BLfa`X#L+eG9i;6tFBT2!)=GUY;} z5pQm`SveHjzj}Fk;dv?`$)@t5=q1<2^Lekf5nbE2MC$!Rl^o|7;`zyoZM5fe@{tS~ zlJ`&P#%Q~CAVoKz5{ zW&D_G*e@a4-QHv(^_C_7IJiN$lsY@95X=t7b{unKqLPYoBM@2F$C-!cM4x-~^PQUN zuE9vs2#P@~q|y;;zfk&W=0k!moxO*|cYE5HrMWr!f+y_xe5JU72#0V3w#Ba_0n9^G z`o#{+i5`#ET5^#GMM-Zl z6#9iujf6b8ZT*zifp%e-NeB34g`z`6!hP#D?i-F+_e;HHyE}_qaYCnzhGcq+)H$?p zT^4iA&|;}At3H%b-IZcM`(=}1;zjErp|0Pp;p<1}6Q$tsw$a>xK-D^lj6Vt#doXSh z;!W)2qlUv*iRL#aMSpgAB}>%vyEEfR3TB=BKRflesF`7u`#AFVdnNhYQ@&YQD`bC* zTNniyKt zs`}uOT8_b?s;0x^Y`dglpqH;a%UCR)!T`}`J{m{cs4pwy7ogQ@E6~NQX(5{4|#Za zjH~28Z+(hA7^0dscCueIb%jJ{u9C#$>H^=I3pkFR=Rt<@{l5$#7g#t0S8LBs`$nAU z8`cGq;3JDauQGl@348#5n1*`bHLlXt#sS>`Y~yx;g)gL{9FxV z;4R}(=!fRXl0!wu`bA-ukIlE`e?+TQ;?SVzbYRoTvp-2b*gS2gPt=Z}_s&5$Q*vY2 zvOOE9#)h!ZoAh+FXx|DcgQJqke}AN*S=L!DP-pt>1q@1#>9-v&Qr9pbXv>y;40Dbj zx8+&3j|pKEil6riUSNSy+31yg!0&j9mKnH`ahHWmD&=hC^;*a@sjI50-l%W_$sbd0 zUXBXORp>Ocbn3Z#b3>;wde4W`PfYm6us~u9e{A%XynFed3D#PV9^Ej^Pi5N~nz^q) zY8W^Am6jRlzU}SS7`c4b5vXf=E7g%GphOEhckrVBVC$;2z*X9ve!1s19b-IA5VHlF zt3AE0WI*-fLBFf!kC6Zsg;i{6g5hmYDob(G8SGi^TlyT^vi6u^1?(^iophXZ-!M%9 zJvW_M)9a+5RnMyY%Nynoj~dVjU4v%atfwtV2@Zp&^@gJtcr($)XAI3#0@aDMpcNeH z&#KD4T#B<_8olfgvP!gQWSL}2o$xCq^Awb5p;+>@Z_RlDNXHkDBya?Wz)7PI|Tb2sj9 zGHk2L3;_(2hX+FKhz3I?KQvcdo9Op<9(EH+ZWB4nRcqgA{6&9PB$TR!IGr*dk#4|P z)tNjQ1Aqf?60&oIXcMsuFL*#Q_ev9S|0ZzXsLG}**fivgIV!(qe#y#2=EwVb>9>ED zA^#{!X$v6gqBWF5$@?i3D8}ITTMG$ZEBYKl_v0UF&3(R zmjFLae<)*GB95(DeHwTf-Q>^9bV0SdewYD)DMCpd2VD^qDFj&xksUFf5NWD2l~H6O zNkc;}N9%l7X*uzrb{FI9m=8%jIJy(=@!>L1phuz1QR%{$e;hQr#G8|vvJkQTS$jFe zZ7JLO5MUI)5Vzu-%=8nfv)1Eco{sc!UBj-o>$TqEYCfh(2PQfWoennyp%=#IoT zkLr~?s(YLtOflR*%|Pl7=n<7?*hz!=%Po=e509XLa4gsPGB8F|j_Gr56?v#CPnnHE zmFai+hCjzPasX)$sCffl<;PtXY8Gh)j_dipjVCh;A|-&et?G_-5SpGwTMuO~VO4U= zW)7js4mgYaZu5#_7m-X0y7i%+8sm!^gSqt3S{E6%*L)9dt&-y8IPUOH&iPC6*+x}-yfwSkN? z?>B@jiTrH2CF3ljnv^LY2QizXvQRJ-qKy8UbDzG#9r@GLZes>;hNY4#IUfQ0_-G?r(x(4!N5_j6o1zSJCl6}8 zi8A}s65BH@`^^1++8?6v_a*$T3yw?ycl2vm|9#LoCOvc?M^m z*2($McO&$2%YxuCL~A!u)$zX_(A=g_O1=2|f<2FN`?is!&^92S{rKYU^?N&|fnGxL z*ZqMAbO{u()!e39vgKe$rF0**C*oa$U5T*ZBCMR$K|cOmxQIPiw|q%qcjM68Q;vhm zPp?GAo-sSSJNC@>RbexefY~!N3}q*|J=*eQp*~m0Pj(F;XNbu$)2rf^i90}bSKFfv zPo9?Hngzaad__}bq^V>I1Z=n)^rSWq>p-1T z{ig5qvV#ISHWWG6A9o;s0gy*46Hs}(o2MO)&YiY0k^7S7EuP9a-w(5@i#XOK6}F5N2_U7hkY!Q?-YG6 z5uIIn*re1=LSw!ie7OsGyX!@}Ei^bv(hNwZmFS?w%iw_wlgC_m_;E1%_Io zS?Isy$o_|0(%TB}%m459dsNf|DP<>`Cmm=6+Zw9wmVF#fG-X%B^!aI~lmSCh@vR~W z_P;Cq-&X<_(WLv2S47nrR_K;}4q9<9hu}AD@ro!8XW1O-yDD1!UmopWbe>J<|Eo`Q z_G*6T?5rloNlE&SK1Yy$trQQf6lWpw&u+Nd6g`N`59bfdgx;-c9I~GOqQX zHAje|Wn5f;+YX;w%iE`I2wLxETK9MJMOI=2So$WXIKq?;8!T%1LLL`*o+v7uR_zx!(JfG?u=BxKz#@sH(S99)3rXJ_S44Xtj~( zH4^80Bl6%(JL(KD>=+;R<<8b4z!oYX0^e z@&;VoF4;9a8}Wdey?8w(dQQF0{IeQ{qlx@FFh_M?een02X>O^?=0uiIi>IL=jdt1hi44#%?d%W5;by7q7FSET}rL1tp_*pKa z`{k;~fkTKqZrTS-b}+FPia`s!HOM zD7>3}fM+W{D(qLs@p4PoFg5d?w&W8%52b6^29+2`EBwkW0s8M{UK>Wy{XIqi4HCL| zy{c&quJbNW-LErR4v(oBmI21FfiXv(GInFduxLx&`ajdUkN{C7A*K>n*SngvYq!#g zZlnCqlLG1Zo@Vtn>hCDQrJ|4Z@#|JdVU>zw@YI9>rE%RGZdPrpC1z7{eGN|2uq0Zf z-*6XZC*xu^^5W_Ip73lXPsqwaM(G$3z+zwD{MnFtVx{bG`3gb?y`%6kRJJF4ucq~M zZOH0(&WUkOpU9tjg>J}JvK0|V|Epdcv~1yHndWC{|A((HkB72*|7YxD?1RDBciF}| z)?^uyM1+u-v1Q3jMH7;3tiz-tNwTykd!czG+sv4XCqxLC)Gezc!^tF=Q`(H=Q`K*e#@K>e4d*eY%S>LTNvW1yt04PRop!S(1i`<%n8q$ z!GJ_-aRKf5cDLoyLH>lHoLK2(f&G;^Zb;#z7YGBKMO>sJo7~+r2-8lUjino-1}@@V zpviq2Uz|@@XZqR>Ye9Uzo&Ry=#*KoO#uq)4rr;**b-g0n`R{n;e5M{0=tYfxdokQT zPyY<0WbU}ZH#4{90gVGc;K{gbmMQ;8(krQ5gPjW^HIfw;l9XdGlasixRWagE*R>h0u36y=Y6dhakBTx*dQ zim3H+sM*udt&ofHRrf1?*nW7#01HmV=)}1_q@a3pw<>cpfI7ZV#Z2svMpV@91t!3W z$0Pg;%>bSpER}(qD(blaz~LGY>Bxs)JXTQ}-_^3aH+ZD4phgG%g4eB9T{Kl>-Hy-| zM=3mBZqHg3@+KNDT*-e2$RzJEKVac^e&K-we8d28MQgaydyKg^>L@T&8~?CG#3*I+ z`DlMv3{Z=`dLKBb1J%%zI*&t9bs&<)p&C*ITyZ>C$3i_HZs1#J&EW~V_h=Bnv5s|; zk%syVT&yfjUHKm2*w89j+TtQvsnm@%yF{gU_@*e5NelnS$sILhmE2tacs9h95j~^mH|0Z>*DM zm5*yS7zk4s;iM(_oltw7)fel;SM<0+TELK^%D#22iSCC4g7)+(j`imtAcdn|7O@XS zeb3*aUuR#MRwlgmV4we?l&`;(b(sV2)<2NV?{>rw$l~{C9=wlt`RAL*%C#OQ%xfUp zKFjTCV3tl=e%&8Q9=x(GT!V$8GkQ<`8CpP|mqqQ~TOa`_cGvs5*biZ}U?K%S>TXc< z11J0M!Y71=N?*C%acG`6pgZmtZDdS0Yq^@6;lT!Ff9vIoQ3t3K9&34=XC(J#T~26I z?aS0C5;n7W%a*av13E0FSKbCHbqRQYqKOyWb!l}#*AxT*u)f{?E)H~7bZ?O$aahk@ zx9EpA1>ybV`)qAu>7c@r_IC|{dOY~FsUU{0qHgc}wftgJ`7Sd5eXOk9T*qFn<0qU* z#iw*loTE-{L1wL1K7Uv2P)^m=?6G$G{2Jvn0`eUpew}iF?LHNMBq<#n!;&D2xE*nqeVt$2 z&LiMOfo#HxJB>;DTyO=GUfC4iH29G#3@Q@#lzw}jA+o73e+dGdLi7J=8-{E5y7MHE zRd#(?4Ni=jOR~??+nzh#pQy83)wW^6QW?HrgQwz{k7yB;X(ZN|U^z)0=f>HuL2yU^fJy5h(N}y*ax> z;ZW4mDOOZMKp&F}Ds>e*`Wr7gsdkqwOt* z^rleF054b2cu0B^6w;NTc8hrpuWW7_I1@mM@zFNtb#DmXYotP>G|nv8jt=ixBEdmn zH0Olj;zqw3lsyRrq4L`|p1o@-&KDWSF|4|jWY|je(W)#hZVz2PfdshIL5f|k@;M@* zd4ItOE^J)$kc7|FJ$PU79T=ejOKNS<&B#f^-XA?fRtezQx(jDB$QKH3MBc`w!TiV({p;v1jO0W00=2U$~LR@wgPD@nt{` z$}tspLx0G~vhB^wdKy>=0+V6eL@nS&Q9JG5=+-*}pv-59$``q;g&s&*xK2#al?c_%tV7G)J z&JW`Dt}5=Z_#wU8Ojf%9z|YdeKIA_aCZH2xS%eqKuUES3>lF5nAZxc2JJ4(Ow+K;W z+q@k(6nxrP4nSd4e79WM)FJ2AJSfoEc~W%%#orq^e^$YKV&9XEcJX9x=d>*Q476TT38GXRRy9gcs62k_tzBPp-zfZ5Yja=l)PmsgCe#FuE2|C%hb z)<}0BR|c4hZ+0t?qGq$d#`fd5=0(DIX-c{7pdCWABz92#T+1VOu~sSx z5%45`0?1sND}9w|B+SivSmhfvV+ zHs2>nSZ$P^LdkvhX=s;WE$;_)k^os?{uB}Q_mNEB%545K)kXU_V$FUR7#isi{o_=? zV_-C`xo7t(HK0YQunFPwgvB+gD4MHRbu2P3PZ;jUbeCF!XBdz4X3eYEhOL9YQrh-Z z7P!7Y)a>%?hxC=>NJ05U<)xf6T=6YpOG>M1s^*}Mh5pg(vBkso*@9Q#;?CK}VY&YV zr$FwOg)2pWR4j`1`Nr>2+L=K=_`SxWV5pqAsfm?Tfol6 zdqnyz0{oTc2FuqtPP-Y@V&8E7^5NTwc}zQ`)YYhOOQb%tv%V#^^G!n?9_`{M0mOE$ zRTyo>^JloJG{jZHXrC!RMyeh9eMPxyTGCVx48S07~G6BfhJ#TcXwLEkpNJsS)^I3+U%P>=OdQDJ%GErGpRQG*_(-~KWc{c6 zrC_%sk;h%HtsJVC&#Tvs;~2Ekt_3xu^NlNuST!`BRb(fTd2OzMbO0RVnVvo}ObMqu z{qpQ50K;#XUrUb>YDkyh=oqjwLJ8o7>|LLG5@*#WRUJAGE~hg5S81p#@AL}h?n+Z| zm0l(mj6mBkspMH=wnKx+{Q0?Gf)Uvl9x!xc+aDRM(y&mrPV}PHLW$VZ8Hq=c-D-#W zG_j^zV$0Wqq7RHe${wp_!w8TuQaJ4_GE;XD@4YyBy0_9kyPyG0{RI0v|uTjBpCyRWSx|9T%~--+(O`_3XV9^#V}${XyB< z!2wRPCxksF3Q+vll0=LLUcs~`l~sq(0Pdha)0_+7Xvt>8%vho=rB_{5n`crmrU#$g zf#v(36am9B)f)#;3nd|JwYc&m)4Net@mDXJN>5&Ti%?`qw(~A`eK~a{5ukQSX%#G= zO;~@zLT`evURtFI~|Vf%RO!;!*y-)~SA_eT|$Xug1OW59@i*teu}Ut!P2E20UDM_UG~6J75;V!UT} zZ|3|;MlnLHe~{N>T{5-$@i#}plgk8 z%0;$e;tswgo>v4DC|1ywursTBq^<(X z%;ca_n;@W18}dixZah70Tfq{8mZ$T}0%TCVtx+~!ZZ}6=x&sIF;3SMG)UA0~i7e5E z(ae7JbvFV)pi#VXhYY#eF8yX$A&6AW={z!%YaNx^aQTOguuIuv!=c^2x1-#83oR@2 z-D0O4UCJ)MtW5{9FpMV`Jm4nZ!E&*$utc{3Kso+`|759o(R?7zd{zugdt%gVJ?5-* zTSUc$*GOrX@q(8@bx{Iv-TQJU$8Y=<2cgg7v;^x5-r>a#Ff0>Lz8|Ux#W?eyBaorn z?b7Ec4>>lg!{?Z&srlsnvxbv=XWPT(Pf=iy^T()X@ydHf;yT3lnF?L)9OHpTT50IL z$>#`9@A_DHzl?shGN(Maf^x?!gwv_qG{}<)W}i>f$rf6hDfB}<>V0Hr&c2D<O0<# zXV1li(7?ttBY(GgSK6Tl=lA=eU2m)w3eLHE-?H-_n7}Xm5JW3}!U_e<{wl(Q-wdpq zhDIZ|B=Q)qU|AcSGZsrQeZ@#{;}kBHWm;XtH?E{V9d2AR6q?{Xj>toUN`doGl=vDD zZ^QvlYc2)ZtPZJ@d6F2>%~C7gXK~0AW+Uv~+=s8)z~yEk=M4^sJ8-n7_3jM+Wwy+g zfOaKFiaV>g**E_chF5aTjk70RMgU<8d}QDxJ&7%y{i;cP1}*%W06KwGS{~0`#{e@- z+9@JbOt{gx{D%$9rF4n=jG7Bqnq4!PQM@}F`Y%BmD=dLGA z$xtA@6|#;I6J&U$iya*r!|s}1LN?&;zZV|}iIR)3>W)2p+NSBst#Y0pE4r)7K z4Sv?jjbUo0uj$+uxIfW&0}OC+5$WG6OEW5Szhm{k(D&n^Au9?F9WayFFMN${xy4Q(poJy`$04m8yBH$grLb{5Hd8D(;I6EN_OLmgep7 zWmq`0G|Ma>Z4Xizd>7Qfx)FO`zt))!%M!)fS1|TkL_kW1ggSskuk|2e{SkA)>onos z%=`{T2dsu;))*sTAy->?&SZ}&(bZCL(X>~>oh3e1kn*4p8Tur>xjn|e&N*82;5+@r z{tl5O0~^7MKGcKuHkyo(UK%qp%V?qe6dN|Nm>aYQP} zHV)Ct!j=0-cH=iN{1zIu(&cl;`8kN7C#=}cm?^bAGTjDm($nQa${_x!+#lRqC zL|!SuOWnCJ!PV(u{ze;TU!b@1&nz}I4(*@4f!P!ux#S;iU+O$CS=kUJB2M)H=osNZ zH6*?c8gVs0m9J86vk!-sc90cMN}g)Hj^TF%e1}PczJO1h{S`RAhWeWL6&fU3gD~Zz z!#ZgwV9DMu{Pf~ts-f#eFM(U(@G4WGv170FfvLdEaB9I9FOgSo(#>?zO6kw%@W%n= z+eL-ZFE-BH^XxqoEgG%A7Yj9}Et6sUq#{osWq%i~M}YbW4}Fr>(^Nz|JohG5XMsRl zA zOb$oK-ET>U;AG^~mXN$|SKud7j+ms&3N6}I7S_IX@&Xd}fS8iBpt@gRuFfrnIo|&y z2?GH2jZBCIqMLq<011DE2GH}hx(2RkPkG>X-&z_z*m}-cUlYud@sM3z$H6@# z_C%C6i=%Fz^CqI{{Uy$5q89aYhI; z?(%-WJ!z`OH;Z3P&)}$XnD2#fq<7gFfiaEP!(ycuUlzQqvJ=w{H!mG>w$OVC#AmDK zvwq>F%})_+s+RL*CPc3yME7+Wl4QzUl*8y?Z8{?d2;_2p^nRLkS&z|#*Og%By>`Eb6BJngDoi|RMH zmO2b+L)yiZ{*u|ozgP&xGls)_!PY7+W#N|Z)ykhRvFXC!5eg=nhn{DKU>)-o=ee2z zAJf+xcJ~!>9yfq^PvT^XYy*na{I&=GzrN=@eu}C0laZkKu~=sVzZ`1oG3wC{kCPX1 z_Hmceg)!}*{t253md-+c{RM;Sye!)Q0kGMAkCmQvhF53_UCQEt@b1A!!eq&>d%A!# zJ5?9eY6AFgOJ-PaDaf&;BW3@9r+sD!POO!-uioKiW(Q2quUH@`X6I!k)jyVV%%i#& z&vms*y929OhV~&Okbx@huF)3NKWui25`yb>XplOvs2kM$7osvdyYyl$_>6`6ixeRs z!ShOByWC*SL}sSJ+peRQr9)VYMeWgvZO39}GT!&(S)URY&hOcrugEshY#ih3*B+K2 z_OlpBLpYnvLm?FXWfoJKa^+pR=lx&y2Lj@5mRFRMj$MhO+z%bLm^;nO|S}9(;{VfBp5P-uRSpg0vzrt7j}6uGtA`(`-3|gB$r50u(F^ zD3RhtbPx3l;UpQlBvqI0O_7l0s@FUZN`e>G^#p+2oDWBK9!Mg(U-1!>!{pbLp}ZSm z2xV&+UATo}{%sev%y|v(?U%$1TgaDhSI$~Tse7Y%?{aLXpL|79&(rMifjP%22^~oO zn^jL2jqHj1(F}elK55kL%J`d(BN;LXuO-=p8MRU%x^Hh~N$#y_P0cBJ&Ce?LxT0T% ztx8I$xPp(tZXUuf3SEEj;{K4wzZTV=RgH4C49bghXb?YK_qyO?HBUTSh+lD)5=xX+ zBo_M=q;)(^yRFitxQ|I8!?k<7!8GT!$4k{P$1Lg4+_I_%atAP)Kbe@N2G-N_L#FWj z+*{q@mwoi}K2^g580%bJxjNJh-ExKxHTP+U;xUH?6@@x!ty2pdrZAgk%E&jRk_ivB zoOAKxqM8UGXwGYOx;HUMZ%SV1qs`&Z&M1rwJ8IN9bjG8eA915>I)i2Ene)%=oextQ zPN18%tx#Px0GcQ_?jL0m>{ExdlI0J373xL2^jei?9ceseYDop48g=Cl%6{-Jut+nv zdofw;0<09lg?PJUy-yr4tK-1X>%RH6C2NE3Y^h5hZE=^tH_N6ST}-7Ab>({Q-eZj} z0ga?*HiCTThdm7|-fY3Lc)BW+)&sKFMawacC%)N*HHw>PG!3djXnVQvytpvSFj)gy zhTi0NL8cE>bKuvveQ1!N=%ZX^#mElHb*6}YN$0Jf>gtyUbsmxowqNd>W3LbTheK}F1<{R$4>SC{-$Y@j?Tbe}NINH((bjTH(*d1FlKKxaGn*A$B}uwjjSx1hUeP?qrOP>SIo z>Yx2IC;nNYwdLaoi|oXy|mvO%g8eFuN0&p#>@IK`fZ5EmCi}pY60mdk-{7pE~%_I_l^0pLaTlJMX1`(Y6+-on7`{_MMD_~iq zOGgD2-0@J$qOi5om5Wc69MmEhxM^%;-tf?pE_BuSO4zErJ9zNYvQ3EZt|=wNwaXN< zO|hBSan;+4M(M(5u}T4Fl2!-ZxE3kQ4~@g+m3kIDGvUf#CeH`}9p+l{>-PZ7>gvd3z)I# z^4Es~u3dD!<73$!e;tT8(5Pz(%1(8N;k)O2HrWf#Lb%5ARXl0qHz&lkJNX?mym<6n zI?peT)mCY8kGqJt?A?JAk#DVq5L0OV?zr<52fX}4Hu8B@VD^^tRo|6IfdwXvhS{Q! z+}5F>Is_Ly%@T=B_6FT_wrreA!2z&!Vy~&u4l(8_B5M;k#HifeVP_*d@LqnroXx`r zhdt`Vn>zMn_?ZgxrZA#P{-!6xvi_N^JtF|qfQF0SUF|)lCsRdI{-#tsQ4nD?t=7;1 zsTgT9dFG=36vtDl^@U_6~8{@WbT{Y7F`nS^@j7sZ} zPB}OAW=?k3DcY>=37whEbH}QnL}}RiaSMN@_}7v!B;Qdd-Y)~!8eL36b0VBIv6@zl zW7jS^cZ|y=%-9@G7bBsyYc9WWx+L7P=g~8L_rKnkFG7ifsD(K1o4+(BvKKd7IZ$T)86Zpd!k4yy~5w=R4a(KQjhSeSUrc`iD7(jns z&Hy6ZODW3$jLxDF1i&yqKk$UCaJk-_e$z*zQ>=V020E;%vntoEHhywiO>^o1T?@-2 zJxe&<;g*qKZvygPOkSB3{QwY%QY6?gvmSSo&nEs^FM_SyW9`EB)OKj&<&RB`B*An< z9ud}Fr(I^9e9ZX3mA$Z(vkT!9t8tu4XIDJi3q*taM?CCa&M5>Gl?t9(vQhHc<)2qs zs!+fL77xRIG(4OC1pet4qda}*`3`=Yf0POQAj2WKl0+J%ph4W`@8x1jt`%A#Zmsv1 zkAa11*5)v}b37i1YSJb{@B5ZEak5`gAn7XRrV`dH%LFRcOV(iU%51^aj{P4lX%q%0 zxT;CN9MN{u*j~0i9-~!eYt$JuyLQ!FbQ4I4;Fs<}$t5u((!FyskjezIa?@Hbe<=+B z4t>NGu@ZSjkm5prf=g&dWZ|S`vc@Hm#64Zl(qx+2MQ#4sKZvcn-Ql;cH{_1B1w7}K z`2aSX)uzfh7q7896!BLLfICXZzx3VXPc|Vw+7?g&+i10e1?oAc-&(6sB;A#3X8bdc z1h4C_x`?R^D)j`4w9@b3g{&79rdURyq(LMU2VHd0qgzWZRf1OK=kh#P%)u;6yzHUW z>EK8LYPk|ADEb1edE(pE9AE(xE)SNYKYFuZaKdsTR^=ydHH7XLmO@Bdd-J^OU5VS|~O5xx+dLD+o?- z_MTs(?}lPTC@9`p;+{{YvH-gP z0I>x*#M^XZB!7OiRbuNswJfNE*|GP1e*XQ|VhUs!!xguR^(j!=xc*m zf3!o&M@74`i*JylsB`{|sR3fs7~SF8%cXnc+xgbM*}*B26eYl~Bt9TtMguKmr@mUAC-a2h zG`9*<>kPxF$w4_zjj+oMY30CdU*~svGbwwCgC1gX^}%DQe8=hvZH<9NO8Hwc%c3sT zzAae% z_mdtK0Iv+6=}ss6Ael#zU6dqpJwSrHvtOrm4k!SSoTZjG)W8Oce@6F~)Z%Qf9*=h~PO z>UD&)2}`kgawx zFFBWOI%XYnPR55VKj*6V75yICixhZ-mH0B?`s%jIS7F)O%kRsXwp5>d!tBjoDAqn$ zi4aaR063gVR~8C~w+?f~AdDfjPSQz2cAJXyeM9kF5@PZzk%!dK@=arw3i9OvE# zj}BtJC}%xBR&Q~ExZCJ-d6Gs$2mC!VE;4s<7DzGGQ#J>km7)}T6#sXB(BlW$3$RqG z!b!g@XY&++kD+5hxrxH2e7GYqmRp4Vik}A@8lOKZ!a^K##~5=F_m8K>v!fG^T}CXO@t~B*GOJ zHPVUz@)UJ~bZGVsAHhS7Zs5JtHR8?Lf-t0{z+9X7Ih36K=yVVyhvPQt2I8%83EqIR z#Nk=DP37ZHwQ!Kis)N8`_>88$f1ajgl3D-;anBh+q#ALd&xRFXw-cY5s=N!jBdf6J zQ70*c5l1_}YjgqYYpoG^#?VbBy9Va4lxO}VrFGy2khsfe5Ia$zI^8!AeNWD=$OvCuXDdX02U0~(IbwhL?IPF zuuhWsep5ZD?aH_7eEjsG5xfSL9*g(MQZcNAQP-^Lo2FvCR?NhIXl1ToAaVI*ndDu1 zz>P9zX@mny_Olb>)^cU=LOxGaWorivwWtFxK!P?$YM2?HosN3Bhhp}I7D+S+BQ}MS z=X6WWv3wPaW}8wrE&IYjiv{XgF1X5PXW+|{449ThK;}BD7^c=Ke6`EiN$}WI0}u@0 znLs(rmG}wF1ekGA1QmeYj=j!73aQd&u)F)RoTt!cKY(4W8(>(9Eh0g`bM|4ao}tbq znv4IEWVlWJna8CIdmn)MWF-0ddG`$HJ-m|{1Io@+3nz?kaa9+J>^Q(S2c>u{VMvp! zkWp~bw>f50Jh?LeQfo5LR<7X*u2J4{uLfYL+_10&AJ8aba*)Q{CyVO1`Lau$(^*QPQ_BM1Hz;kE%88tahS?X-KvTx6_H<=AJcOXb}aak~9~Dk#^4@b~-0 zmIuBI0u@9FDn#OT3&tKfXJ4$$?Wnj`a22bRQT$-1`W>Ki)B{)yI`(W_qqZpQ;09mz`e#`k;~jvf$XO7qbN~VDYLlrgmz= z!hqI|6@8Yuuwy0`p1R$ElyUG!u%hPN0!Jz!Wjw9=PBXs(T z2Dyq`xz|#kq3yHoz;cpU>2FamB{g0exMofQvj z`DQu;91VVln+Ntax;Q+4te&L>_Q{le3}}{@yUqJ!F9GOg_$QxIU}vsfM)e`0;3f>q zx2M|H@KP)s=(ua`FDr%&S%|X9(Tl!0 zHyfvWU+bYP#4tef)l`S3Tn-4*I)bTv>$*R7U0?@|z zkSxrc_ulZy1Wm;rQP*kM{l%&Lk2O>FIL(A%w>^VM@)Pe}1><$LRW4xG970Kt#Fe$a zP_ho!-OA$pKa&oY2pds?N?8L|oEnWdUG=Kaai4 zZx;lh6zk(lC})MSCaCy_O&$EYv9~k3w%=nqreeTo`twG3?3k&HXou*1$w?f-Yv3bS z&zk~+GNY%|^ONcyOw-y`Gl*x9d;*HQs9vK}R=)>O@MG)N3j5mmKg6L1$eC)}mF+vW zzNQ(;mON;CTeWmZfGn9`C!F5^=QZfK{=Nh_`IJY>)4tisUX={D@qXU(SL0x!N9=WH zFRB7p*qbJ)g;l-h!AhGu4zSvYBKP~zDvS9#$8cOJ@%u4FEc3?<9Hc$QPUNajahdIH z9zU#(WA>U@2@0lIG~Pm*S9;?G6% zpn~6Sb+>LYkkI&td$?A{Gs#b}hQ>n#B#3}i*(b@c8GYW%%a#5zM9{nR=J0GGPl&0@ z#-{c0dmpNbbbWHn&v%kUDwY2%LoseYME3sia{h>V2LQm3CHviLlYb#3lY@$@Z;21W zKM%R!XZmhKbk%bRLATPhuv(5JgLKih*k;HF#(~D?8E3Hyr31xW{WyyxnNPXWmBbQ_ zvS)`2+$wV?IcCr2Zt?u;ka;284jHkqQn=`oX@pm_>om;ca1DWQ4|`JQ@Kbg~^FZ+r z6#*oa#TKr4@Zq7-Wk$HyK*(IHNm0A(+H78orKLR0c`Ur^{Ug<*Ls)p8Kv2sWUQnIW z%)SVoIA4OT;h=Ph7+tEMPVn3dD;c&ngcgN&^V8(`E|6}{weB`+?27SZrYF!Z&Z5VD zKq?}1o(&!7JfbdG54LQiU&Z@j?wbnkO&GxZfaFNhqa%H@&H#Q~z12OLX=Rf=8Dz;A zCY`KgQzFZVBvPB^TCKr8Eg1AtG1Q<8r49J5O}wZv$GkV4sD_cu?(G z6w4^>7AzWo%wr`+x@v3`{Oe8G=-JJ5h8Od-yN>8Ax>S~~;3d~J$7@YWDHfxY^r@E* zr~r}%T40ON`2`1%4;ySkn=C99R4KcCZ~nT)wAJD`+GqV!B|jMyto8g%GS{w@c4Zr> zu~pTZ;*~&C4!V7ZzjVN#A1}P`{KcaXAqGotXg^-;a0$(nN5*vQ^-bspv{KEmZCc+9 zCkbA1l3mel1N%^`T0{HwI5y7oeLc($TfCWC$HRNK>_smS2|yY@*gTV5gp<03G1?-^ zR-Lx`shOwS?zDW10gl8eyq@rf$eee**INX6Gh)Up zc+7qBggW*ByXU<-4&uo5is_&%kHCEo`2U$ zS}HRVLZ#q+*O4lh(Y#-VboSE_GO`}=cYzTq9Dw1yYZ=fu7%=rT2~tFuUNqV-%LHO(P)OhxInUvl!s!Y?Ea40*f?u&W>yQT{)o`LW11a~bU1s-C!2bP1fe5~0 zwpeeb)AC}0C{0S{WncpBcZ6Uo^{9_gKg^q|M=je^e#?`cQJtC0wFz?y{*H!`;;|~8 z?bqMa&O1w+M3g*o2VZv)KSu?pk$y8wxN6UaT7s?QbT#wfz>$NWA5LA6u_s;P@IOWa zNeCsv5+S%`ZCC**7Qz!A1FrZ)5;m~;VXEhe%EI})aDy{8~d<*mf}A&tjbTdJVwl4g zNEep~zv?jN1Sjp`YF8Aw{y_X$F+RWUK9-;tCi5{aiIv6`>C4>Z+jb{4!Ch`|9o$N| zoY}yYtYD}#cYfW=KtljhqI<2UZTku-KBSTDW-kI^0+uf%%n`6(2%#}dlb|dY+zNy! z4L(#{E)m7T(Le(xdLxkhIXFl(Mha*>kEWEnNHmM5?hcaoyKgc4wcDe|uRrN86axs$ z$19FzYJKu8?FVQW)`kj!nBaxN4Ilk}ban3c33t7lQ40RG&tJ2CRpy*yjtPHuh~t>U z^UOK?ug@O^0laI=T!p)qR=j~sAN%~}4$lz-oJk4;HaduIb`=wh5cCFe85k$uG*l8@VX-?B?MiG?+D+Ss69#XIOE4d>!+QDR*>0Plv#9HViw;}`V>c4hlK1*R zlEV10*!T4_-4%<$)j5uzi|^o{PPd{YUK9HVL+!IID~XdFMMHp#7u*Cnu5z4uWr@aH zS+Lff%nuw)`BxU*3Cg;|m3&ReMsQt5Wmn<=){RLH0&uD%yaBL&O%_qj-86hJ@AV;G zE})w4QksRV84h`^SvO;2C4`vHh7|esgK7Ge1QV8XuY^3hVeoJIMY6&y0M za)i5YAs_uAfMG`#)n#F;i}aUVcW+rAr?>>gaDRbJ2F?v%3yLci!Am#-7SC5Q^i@i; z>SdPP!}n6n4^D0{#E_D2Xu&NXULi%_P$jyq8>g#`(N!7ZqgLeXqZrk6sCmbr+(pju z&~SkJLcj`E<;HdV;l?Khh59%8xDdkZ0eFZ%9>CruYHieo0(6q=FU-lF)Lg2{^qi#h69obn$*?%CnZ?(^v`b0U+_?*La~| zU+z4c$4aT^BNw{Gd$Owr!!}N!-mEl1L+|eNR&lz9bjC?s=p?^5L_oA z1(K-X1ZM(^4RB#EvjOgF*_2lK`-u4^e@YD2@EPUToS${;;7PXJLh(7*juzw4Oi@no zsR$c!>IlxbbXdDGOK2i~X_}QWI;=&cAfX9E!3jZjL>E`i8!QG5@oqv@?5D> z`|z_hmdXN#`+{TEEBI#Eu;qj6rqq|=5AD8bRmEeBT>v|F;hNyOc6GrzV|j@ZR873k zF&L%HU}dIZJ+eU&U(djP^L+?cw+c5Z9cto;f6g~yf=Au2UkfQ9{2=nNeeLshO~kK{ z4gpcgz3}CBh|r=e-lV2vAAA3(TU*_HW}2z!MN^e)db7kAOU{T>8%*bg{G@n?4d#9T z$&j*tNd^Ia;pXaU(NXk{a0WS@`9?17IVc_U8)&k7P;~qOW*d(0b}fq#^rBybJ++kObRz} zB;F(sGFpdE^vb`XTC7F@=twaL(fEn>w4RQI$02#5SgS2yt(S>66*^aH$JR>yd{#h| z-H+8x?W}O&oLOa1g6^KR1d@l890@eCutWEn-ZRQqX8)qosYQMJ+yg8tLwC~j^PIMD zoeYT&7&EH~^^8m2lcFjhSiG5W*pWd9IU*Z{DK4C1Rovl6P(Es zsmy&+S%CP_%Kiu^-^!w;Q6@uVo(QW|tkJt1>kk6%?*ty#y}X9v!^ z5!?s|^$6GFeF+ETfTBH_{4GtT))B^zB~O2pP$EB*au8x0^?oOR>Kal$R*l$o2><52 zme+{&+Np{0l&t|AKvHt4PS}W>m7l~)iVs8*-S>Y#BR0+4)%fx+TxB*NR`M45C~+b^ zZfQLA2hAEMlp$I^5pgtKc%7NknD7SgHOPw;_CrV||NX&f6MmZMYB!U$cUDyhbIxlB z>0h;vfXF+byn$8}AKt1YavCkJ_e}dyK9ZGw^Z8Nx^{RaVLFf^_(;9u`rKBb{TZdF< z7teCLX8nF*MIDJ_jP%gRkY7t?-@tMa+CMEx-*E=dsG7l(g_T65aZ|If?Bn)D8t9o- zCVgq>dW55~Ir;C|=4AUUN5dEP-ZWLF3|Hl6iMYUhjzYThSGrqJLWo%NjI9-OZ}T1? z4tqGdrLNBMEa)OTV{{L~wwbQIG&J8Kx!ar!L?T+QVgO>Oo7a}>sAQeq8`K59;I{ z%x?b_UuGrngtLBVqh$#_k~pDai<8tTj7ErDL$!f7B6hFT4QauR(O;V)s%BA4;?A?x zCMZG`pmI8a6ZyX}#f4-iVs1&$oqr7JC;pByN`+hp%i4}-RXN<2X`_|_Vc;ugSRA?4 zMa9+GlLn`1`79?|d)SwLOc4I5+NU~Q2YX}{A(q;euVP3Eo-mxQr!O697hmqo&+fuI zQ(w;v7)Wt`Z%-4Qc_vIuW%$5#Xte*R)I^P~sdAfDxWfDT5q6WQs5xTnaU)Z%nI`k2 zx&tYfuf$FuN*y?juM45pB^O8bZX!)YAjkmliNPFse?&h@WJpnNXtGz-fYMi^s^xpiKaB@4?W zbO`UFwFZ-#*yoLU5O0=Vn4rHd18X6^5dqI&H;X6muUogGLw0}`rY6WA|r<>;@ zf1JLuc=T0v#;Jl?J4Nozoq9fMG}h3lC-{ekkcH2ood`)#MOkxKSE~4-g#f%wSYFkv z>=T0)b5N%)GJ7X!^ zpa&cSgde?@i(xW{VmP~ZScH4q^@sb)Zl# zu0%Ppe;>NH^H7jm5={VaS1~`w%RM=2wBSc>#2#n99hw?~QI_ST5Wx$7dO#yM;0RD{ zrABTB69w1+Bt&;37VOPQf{Nb_Iu6i5-^Vq^Mt>bR*3H!_DyT~Xz4`N+o%$GpCq;iZ zwbuc1_@5taE$k3jUVeLStJBdy3xku40fAi_ve|xL{Y}|tdOKhLLHjor0@k3ptf+xw zZVo47lc#BCfNslh^L`H?6stoWyq6cBI}6nm%dp{IG}lxAN~@s2_a@MxXlUaSoe%G{ zBX!^(Vw~lF>n0`D;IaQdcYyz?dIx3x*Ih)>)c?BaO4Rtj|8z(3q@;e=k7xDoxt*vS z75_m#EI#BcxwihmPB|7gw`&YsTzT1F1PUj>Ze20wVfUZA79~R=h~e(i`(2aZucqX9 zgPxtg>8u)}ZBL~K0YTuahFu|0U)B5jXI`wBM(&+xOqr z9az_FzLz%5G}!5&glK$p-yvVu`E6NmU%%Vxg#zas0s~I%{380K<@D^?%Z1(l5zg;v z$itYt?I&YxWv}9HzB=hn)2wg#S6I~kWTp2nEk$wncb*6^|6HR`zw)1(0R~{>`|-w2 zvD-HbqSyA>N6)M>9{%@%@aS4tPtr~&-a6#VPJH-&uBiOKx2+6RXv%r-XJjRImfz^p zB1Lsq)#ncW+l>ih5EpZoZOK_YD*B6CUU#@8-ZkXzzYSaY3l5mm5|?*M4?s}H(;w7_ zbl%*aRDlcTPxB)`D?M_w!)v>M&Z2X!q-#Lg0sof$`SHz^gg?zzkZr;~QVz+Scb)%F z*~)(j0y_%jR=I7Nw!A-RmvYQaDzQoD&j{*n0z?X1kd-~Mqv0<$0S28nQS~h^w%-IO z-m_6tvck@-cm4^(vVy16Xz(3PT&c(6c2;#@z~o%_Kwk#F!z`o)8oUQ(CzkhxF2dZPVYot0pveBIm{41D$9tG>~MBg1+$Av zg6Ri{ipf%lg9g>b67@q~qr-};^;vxEd8FVgF0Wi|78rwrg=KyuD@sJj?&!lT%fNWM z2hxHGXR#`;3rfhU-#d68=^;T$R13pFQU||)C6SImDhn+(kyVvkH+RckEbXl<(oRRNQM;&us@;V3-6nE%p1A6|BgM`_&Xj|esDf5CHPaMoK=5-k1I z+N|su=6+cI1`dG!sdZNN!&z-fE>GxcpdJBD);AeMhUTg^hzii;Hqd5qWSNs4L(>%v zS(gC@0r#Cg2L73FH0Xsiu^*6?U+L)qc&IWHIS<=;ni3NgGIFia>OHR1#z1)~%{o34 z`rl*$C7*PmQo1mQpy&Ycf==QnFmfd+YTc&%rl1TM3=`&|29yg{%my~f~)`(zw7}o zMBw4Y6Bw*a2UuxbQmP^Bh$}JI&s2!i5P@}H6rSR!18_3Q3-YopLy- z#1=Jlz7BLP{#IDl!w#rej-Av-10$@t>#ScpzmnU|91auNWhzlD);i>i1SQ)S9O_O| zK2A$t-BG+FzVt$X@RFu6ItX5^RaM5Tds~FdMWY*dlRp>B;OqlfB8I3xFOXyM+DBn! zy{TJzWQpa3H=d!SC=HFR!`oF=@*PjYUPFR~>f|LbEh}l#`^KGiRQGeyFSfJSz&cJw zg_62j9}X9%?c0t60URrPga8(eO_5s|^Pk%T6bo3K6G>PZEL^8SR3f5bodn(cM1*k?c;Ent+JYCYGoPhHg9hK}m#^?opnJQ0eha zSP_QjWdk_RD$tN+`v0Ts&BK!3zVJ~8oDoG)Q4tlKa@L%uR20oAQ8Owv#WFH8H8V0R zXAp2YO=eaO4Vq?BmYlL;##35y(xO?UQy#MucCt9-sk=Sj`+M&5+&}K~T>dLzfA(kX zz1Moz+H1Y5#FgW)a*S5iliOS5{}?=0!xzpPSs#Fi$pJyI!2b>fvhe%NuWdDfb9mO} zBc2+TURss~uLE59dH?-1KLDoH&3j37!H~?Rq6DvQ7&HlfNNN8N9ak2!esf(zfaz+M zh42O_S@?GdZP249oVkz_xPp`gv2DyAmd6<$3RnuM#$NE`L_Q1ITt`v6{lBfiUXmrf zXin7hH+3Xo(Bh1Os|WWCsk~p2Xlz>(KUs_;HUU9o0%&vKxhlm}l!HFrxdr?=YC*?E zaj>)p`C>5usXs1``~3XQ&Pi(6|K3)ZGsihdbOSS!lci5K zy9bw9T*zLH=4m3HkE{j5F^|Gfjksj~K_YhLSKvK;;NooW$1}wPG5^~N`o10X-=wwR zDgb=lhoN2s7Szj>1guzQnlb#qqpc|FB9T@%_n!?ta38Vq&xU`bu0Tlb!qxzzz2m$R zIQ$mCGhdGDrYcZOek5>MRu+&o(otEHtk6#HvR(OdL>;ymBFZ*Rr_-eRK)d}ClpLew zFdKzpBn?EE1rcx_7Auk`*`2Q4F(*zKz^#mN09OZ_B8azk)Z!#n+sN4=-y|zL$hMb3 zYxPE7;ShS|Hh}(qPQ+4Xy2gNXD(u|K&sJ2%LowoWj>F>>qj$TDw{NCU zQBDvnYK`Blqte{@bIzKC2>-n96(9#r!?^+ubPDQpWBt=^vxEP=$T?JVO|ndN0KYP! z1NXg`L-WD1Tq=;%90GO*EIsyo_W!^F-Xw{ms`;YkKSe^??c@Z+OEUOBxFjh=>(M0j zq||0*RLrxK6-){?(p*Ycq*gw%=ACTg9O!_9(+ZL_YyP+KX=J^-V$C-;JX zmz}4$eLLB zY8#%#_&=VB!F2xtj?aV7x%TMQ+0G)1c63O?^xVjHCSTG|^$@+Dd3xhY)!EsS7Y*y@qW%3=juCJQyNr)4e)Q4;p-u1!Y5IGpIrLp1 zTjPJ9l_-{)LnjVk`t<(?)PM)E9H%`^wPqeF>{h~sx_~q?ej(5#tLdrJe;;D?Pg;$q zsV`homX9HA@SUvjRhqhs^C$YM&H`p;dLmHz|ND5ASj>syDy-Is$|w4)L5(j9^^y#{ zv{YmG!zlklO(yM+HjVQzpS%8#@TI}Jk7rE>raO+y!UiJY035nw&h&{V&j(%>>VEEC z`5O74hxlm#>e!b>$ROv61i+2%gG)KKemj%y4{7SdC6DJfZP7QeaBAlx)xRbCP<5@y zowoIl@;{T4PEFR;D>qnym)HbOz$CREsSmvVDxf1MH~EXt1JJ%7*zhbU6Jb1&=Z=EyVvd6)I2f!=RCQyYwr{RXKg|9S^1;TaO(P?u9nQlm0u$* z_-I+sKq(mJR`mPUv zd=&d+G}XWLgkKT=)>a4R$wiJ@QfGpI9+mu<572){^7`w^rmQ*xk{=p}8@^S&NKuFJ z_Yxa{PATMFA&7}6#x+Gcs?O`i$NUP7_z&jaH;J3zxO6#r>>i?k$P_UQT!R* zLcx~Vu!3-FU{AgbSTu?6`p?9LG}ONxGH@Va=G(;}Pr#aUKtjDWiS53Q%_*IHqQQkvdF#OVa#YvMYGL?5~7T)c3BdJ9W!hrt7jD-jdWU%GT9=l^s)B z6bkW(iLS`H_P5t`H{@8nYP!zWITZzMws=){#OuM}&g|9ca<3PRy15(QFUW5rn?Gjz zl@o@WES#qVWCFN}%|5&gdn|t^1)$-~pxI7PTvCNytSBXi#V6J+BVLaO zcXnF#vsAw3sNN(joWT{LFUa0g3`3pTaofVr|N1!>J;q42Lj|W!axlx<4~2n5x|YO+ zL_syYBRRu_h*IPv=LoUB5V5(G1bElgz74_l&3r8OxcQ;|UvQxy{-doS(FxPOIYOOL zCX9yC{G8dZWdYY%Q z489Ga_sza%8;6!n911YIC<|trok|uwSTXg3{KJB$#5$o9=FG3f0TvJWKY&9)SlwCm z53w!`v6(4GB;XuxvH;Fa38~Jy^){e`YB4BO^BF!9$Mn-UNKtt*iv>|3xlT>*^Iu(x zeVv)Y*D_ZoAE#71RG;W{16(9o#mw#su)G2;~w(MN8U+g$ep^2&+su4KryRIKHIR_h*jI%BT6azJ+B9r4XzvZwnI##hIxyU@dM|ww>D!V<$ zTTiU7r#*gD6RaSO>Ke$YbCGaKMt<*PxCG@~+HIf@&K$o|la6!;3hB_q4*hb$ZgBn8 zRn_gA)@wQ0T>o}5CtN++r)Ehrjo1uJo8xD)t^1RqgFIkKzb_IJ#P6VN$WFq76amWU zl+V*2E$<|4L-ke{h)-Yi2Z?Q`B-7RYwuI zABg(fWeG3nw}#d(3|7~%b>xcjgqTEu!n~%Y*l?a=)SSDT4t^xGsLYE`^MaW)3KMr4Nix20H&Qel@?j$VVI<~f^ zyKJC5S6D#y)0XU9sHkDlD=BH+?-zDF^O;!t?9qi?&&~}VnN$ia8`N(YTN>x`NM$W5 zwzfJSo|B)%wntxkc|MJZ&8}LuVUkTuEge2<(+Z(cN>Guj-wb{ra+MlOI@illW10n4_KdB{(0e6VW#_5q-~Bu1 z?M&(%M~@ut1iCWc%Ta&D1s1ZH#L|li{EO@81q(E5GN-8Q>*m${k^X}EXDm(Vj-IAK z{IOM|22F7lns=*z)6PKN95UJ>P=k=n=*4W*HVm~0xv@@10{XbQd0ScIeyy6|sLzSk zO;#os*xg(-Uf}=BMy&!fI$6Kb3I3E>j(b!&GFeM<75X|&-^IA{Ef%=X$uTQ4nipM@ zHXJDS{ryk`Z!&|ih-s8Iy4UGR=w-E-`-H*6s)d4-0ucycHrDr0%v1}7K492B&Ne5L zz{*_kqBQ*0C{d2rz0V#Tv6ud-A%mABmcgA|DJ`T>IHHkyeb+k>myv#2xy5F?F1`J9|)CM?wvT14ZM(@Y1Mz3nM6WR71_EjQ{C?@*aV7pXoEZ`Lm z2)0QZ2c$v_r}PEYJQjJ=byKl!@DMz4%<>H4@l4hzV|$TBmtKsT4#XX0`%JMqL3EJJ zGbY0sjZK&V&aU3b>^X0;$g4mI8@BNYiKZCtn<0!c(?P#vHM`dEN=}_^O{3ENhZ~x^ zp*U@!LN~eUKA{>mHS6=KJ}=~Vj1Ds<<>!SNcZ5992ZONGHy^FMr6>-mDt{v&VUGx) z@ulmhryRz-$DN7UNc5x0z6M{bS{~Jn~Nlq-*#r$ zF&wroee`qXW1 zFF#f3&}4h1NcH$^j*n~y3Wx5JW=9UsXc6hJIjW;{Cb8g+Q7>;>TT@>f6Vqxq5m0Md zTW*)!Xh-CAYp!i|BTtXm4RSLI z1Uheo_*6Yur^$%?z>1{C12UQxF)>BjFY7NL8D0_R)DKKv*uClny+Nrn-)hD5?Co>9hskPB7VSb}C8oBQAJickdCG>aEr`4ZHkZBN(3GrXJcb31U6lVkfqG;v z+@?h3J%JH4)dkzz%hy2({VMeC)}OTjb(t~52GHe{=lr62f@1YSTUej#f2cB0iMn88 z2=+0ptOIYAgr!10N$d3kYX9#6icwsL<-kLFy8#^xHOA9Gr>0-9$(3_N+Au2p^>^BC z48tQEZA5y9v)GW^G3GVZ{e04ur-onvD%0V)=LxRiI*)rcf|puEKPxu5fVKZmR5-85 z1qE)@rEAl2-6I!p>Yv;*Qj0_1Vs>R~EowGl`eW3t6@{Ias(l0-D34?rWe1`rFsMVi0`uY~4PEC4v79dB!q0 zO#6pzaepyzYS!t~x8t0G7mTve+VVHTdKbWR>-Lw>lA@DIVo6pTOl!9Y^H#=keUNSS z!{U*b^;wQzNBHWeA_w5_SI5jp?W;>?hq5oJ2Na zZL64N_3~SpzI%k@jh;s{Daw$1=%CIT&W_ueRqwsvI@{m#@&;l~b=`uw5bDE!-c%U9 ziW!MDaw&4V_xa9+tj=SUSvTugIGXG0bS5z<%oPqA=!wnml_R|oh{Z5ynUVp zH6Bu@ws@7UN)V%lvkf-EZx*?68r!NloEx)=KA5Q<DH*<*%tlDU5Bs8Glwl-r2CMZ-{z{fr|4cYOl=JhzR_*c#hU=hY?X*t+>s(= z*PRJ|K?iTg8>l;H(~G)uRns|U`K|kI>`=O<+gu(o8$xe~zC5o~GsW~BKCL`Z5OoR- zYI?>e@80Dn`PhPW>BXV%CpT%bq>6Ja!}YP9Fs;eDWLdflmc|Hab}Z8Ex^p?)_ePJ- zeq3t9b+$T4MHg<7?Ju3*qPA}aJz8@({5FeFvu@V(sD<=>K&|oEJREe^plN(N4z$Eg z_BcquW1jo5cUGq$`Wjkq(ev$}uZ^X3VE3gnOFh z1-Q`2y5Dj<{N&z8bL~xKh@hp*$el$dSv-eQdgGMNvf*9Ex?D#X1_3N)bT9n39KCiE zhp6DPoHcnMn;0+-M<>0$AooiU^9dFxsfK*RIce9MafT|JSAp^HjsR zEvqe9dM`0+2L!vjH~oCZvO29f0%+vv_7qJw5=+*{$w;%|grQg$AlA*sV;$jIBiv%3;FfHbDk zN?PwP$)hyMHWuK`A7{hVS`EbtHNWorPn%{l6QX_9-V{M^6{%TDE_qr>sjJzVCg4q7 zcqvC29pP@?y8ilS#m?sz)qWtKw>}#2dD{*BE~!tJiDZ7R;#`WsPDQX1i{jtm=|)+ zYez+*`0*4os|RIJOE2}*mT*(YAw9<-RX1(@LG|yF+VgF5Y6x%rud@-ZX8*h<_1IVS zwbl4+Vrfm(J5V*ZMqesg^$LISm0FrSsZ_4-gYoc7*&o6HEO7D3R2Z<3_QmE7*l_hP zoj6%!X_rQzou|QxY>$%ic$w}1udpCmE0V=hpv!X12ly);dlF{n2TFn#UYtp9tfe2r<&{tePRK&GAR-#i*!bk*`@pZ^#vt$RQ zZ1!w&Q^~To!=J5uztM(+V}G`a4=o5R62dE?uxx{QmgP&`NAyf0O^9?fwY=l(3V?4` zM|hUZU3ga%IV{(rAgG4(n|1QAf}EAD>|6ocrZk=c&2x4whyBnmd;j{naswMGXX#9c zM8zS^R&2zRuE=J;QEDbd%dPx9-tCo0e?>$?I^T;7H=3OdrRW+bH#K+Rn^8A%t-jdw z$H1;j>-ySgD&{CbVF%9-I6anMNam)b(X;X98I}E>5$E)uOLDSRfA1A-SC}O59x4ti zbf3ZJH)}Xx8-oB|WTSqVHe)%blpbw^F?xq;&&r^m%KL>sdH0nD#KtSLULBGCH=z&9 z7I2j+BXWac4K+wC$v-Q)|Jp9z$6U3)IK|5vhnV;`paQguTNhBBJ}!!~Yh4|rcCN+e zM#-3Ml#`#S8X{+v2-_?VU^0JbXeTrOQR@C>x*z(}_*TpGU5hkK6?QI#?Q=ZWuVMV>UgWwoP5;*9diX zB@eE(EV3;PW=-vc500gbi?-Sk7`bzJGTFHWu{IZF1k7BCs|Z~XL6wpNt4E3q3VE8P zG$Fpw=xuk#*5D-D?c@|BwKjGY9X_|&Bt6^v49=~DlxrKMmnAI8puUsTj!KEK*?}V4 zC@Yvv&tX*{2lqS7b=azpv+EYaO1}AP9E+VX3^H+`u89HpdiE6>>&vI)gaRD(&g~jE zit%CQMx|Rwjp^3OU6d@IB}Ns^`+MM$# zpc0kWiPCw~m{F;+%8q&3Fz^ikipNyew>2BR88W+QIKQP55?7@EBQ(7S{Th^BY6t#% zs;OjaGPSVIg?~C$_FY<6*b!d2jac79el+6S)}V4xvTdP2oCUD^m09QnPn)ChS@N5Z`*_AH6b<}D6-$A@C|7Sor~-}Z0u#eYPdUI z78I?vj5r%<5*KBfvnneB*7d4nPE;^1Di5it8M8R)dbK>H&SKd;&z6ID!T6V=IiVdL zw`iZyoU62^80iCV?$KPh(FvQz4v81sTIg!WY((WR!sc6yEnSX)%2Nq5$j+1d6qLc@ zy)Mi;RG_X6SvSiZ1FAZz5@YIOSKl;o00TE2q*#B@F4l3dsWW#Hngctb;v1WdwP-wz zDk%@CYR=YB1CfLKL;9DsPrF5#u5}eqMjN&lK?2RLtW2;me>r#@r{=3fe39+&u|=#B zs1vYQt|e8O%Iq z^-!KZ5WeY6`8YrC&C+E{os5Rxl<+g+;##~z4HJEsu|v5xA)->Kgu-U6N^KFpLJVe0 zM(ft_UC7a7%6vOYf$A@@J%~NU z!G(Og(T2Ocesa_+PF^nRDjAoQW5m`yAx}hw`$vvPn5Xncz@L3S4lX65^lo%bOMk<- z{9fe<8b{F26`8!1t3=j5FvZmL0o=H(eWgpai?Kn|yjA5jAkPu`K-j3S-KOs)4Th>0 zWUGnQ`pUZ$_=L9p!&RkSNaNNd^B3FNs#R>`imf+wxP`BB_Eh~F@JlEUgw%YMREI>W z_jBy*ChH4vX0y;UJlzE2RYHr=N_;7H>UqIHg#DeG&@uCa<{|a*3j2fzwRlBz2QR3! z6Wsjq$7Ts3b>+47S5!}9PIMwqB7gw_O&hNj>qeFJznEYvHDM6#ppT zZ6?;1wDz ^K{}Z`*rnjr8bWkp+Ar-F2!F6KY}rj&SATdd27kiW^cnK??q2Ta#Qc@TS;^%2MlB zm_7qrSvSE@$IIBDoBz}ZNsUY3LG1|CE3*9yIIYNgtI0Njku@rs>FNYjiZMFZ@uEuldKrZeI^v{2^~oQA?%n3U^H zR?8`R2UvPbb5Q_v9wniS$T_kSkWq7tVy$29A_AE^rzw_h4^F37=ufi^ zj>YEs91EmQz-g+QyQJoyAxxZF7wTjf5zSZOc(tk#!3G3($2%j5POY8j(o@zW9+ht0Xd|3W% zM$9&kkD?DDp3EAy!-F=Phtcb#%n=Dad0|UF-#@udE4k4pU2f|`G6-wBlCAbbvsn)L zguh^Z8Z+(Wrx|%cb`5Z$A%eI3(pJ6SDH5A6sh*syfKFpRNUIOdJsw%#%hOrFpOBJ+ z8#leE+^Il30PgG+?Y})WfX}a5YTm9xtkc1W7t!-APo77);#cJtLpQQmh;FM}EDJ_* zUFulF>77ryGR~Di*ji)PIV^@ly?U9CH!{Ns}Aj!IhjL8U#q zRjgKi7RuN%{tsepN^5y3Ddf`wQyXfJv>{j8z$Zuhf70F1fOs}p-@(;rtNN8<-8jZ@ z<&lnNuldrAdp*e6S}UanZ9Q#mYq5wrp4&7Q@PcMmZ1Gb2^e&4+9%8i8Y7X!8g%ODo z`4a%2bCpqr(M81xvglv7ftwO`24@_JeN6i1y))jTf3q}hhM{7|!Oc{E~Z`w6s` z9_@*E)7N7jhuW+P7Mw=U!+RTO4j!%QoKT#P{ez zQ;R(=DUf#>jK?g0O?ngM+~Qr{zUXZvN=FCQl5)=@;|Wb6H)Q=FLUN~W4BLz<(bmt@ zNx`^{IVX5ID>U1q)G`_XQOn*1+{i1naUs->WpWW7boX(t`rFQgt)}U6-(k$fn94r3 z{{GQ={#e73yV|B$C%m7)Y@nzl#>#R^<G^MlOtu>a9S&llgNt7VDX^?3y4lB6UXJ1g`Cv*5|6o}*lQ4x*I+qmv+61O}(W%L1cj?v^<5dyapKIq# zZy~`&8lubOdlA%4`}F4Xu?@pE{hp^OM$aXyQkq=6j8(IQ#kJ)wTvQsi!L9enU7ue7 z|H|aAl6vbLUsL+gY>h!KVR*ikXW=SP9l))QWc_%<(t6E_dciDitG(44qjRSjM2I^i zk~&v4gl!I@&aPMw%?U;BE=B1H@ezThU1@Jrnry&TDq7$hD~H}`*gF^AJEFrjik3Df z)oLCWkSfd{K?kf}^Gt{T{NO$iowEB>5sKRC>l3Yn7J9P%SvrnOy}Xxf9pSh1<|jpJ zT~&MK2TQZP^Kz_%1SI}yz16KoPir;n^gLD13M1`Us;~KlTY>u2J(2DHESc&WiMwWa zx8YZ^2wK2>!ZO}?P%pXhVGd|cxp~ghuTXrYbBdFcs}7q7z3>+Md#6T%rsOI&p##PT zCYisrHIhncr{k}A>g^~c88mfibnWSf>}bu~s=tngrKp31lC`bDip&l?zB#1=!eF5- zC4auqm4H@jR%zL*es8Pt$PM|Vv^c|A^`@~Y=}l|6mk~aOW@yebxaISiruyD}TSs5|FV{|F2Lm>M(zprE_r-uL|3cG2uIsX1Rs=la zk7J6EJC(4E+Mv?-%VK|)CF{>rJmy#oK5^nrCKJ{tm(tb1Fj#6OT<{=4bsZMpvZ>ye z%vCMmbvEqH@oM*^`Fn09O|an`KuK}4QT3l@3TPIm!YBT8%|7`?L`&?1;PRNHa3*{eaLFtql?U_OR$ZPY&@;ca$~%t0q?W7H8L6S#g z?R&-hxk%NYmajPki1B7R>?q5)C&`2BUSW2vIP7t^gzbHYaa&qTFn(TXPBF;CcA9y} zHh#A8+P^J?d5|__aia4nJC~x_Jj*uLC#`-U{AUb{Uvned1J?EMJ`(T=R@ z9x3Dg5OOKsd6!jQtiIfYY12Czay5ldJAKG)wF{6`xsue@H_;CD2v8K7f9) zSi9lgP8>W}`_LT-N?bD+8GIbe5Na#K(3~*Ag>~kt=2N!!00qqy6>Ba;7LfFYwD(f1LY+=YjVe_9bIi`**8cHX{++EB zf4;p)<8h}}Ps#>)1&Sk|i%H%Sq{^p2GUqtDThG5Gxqosx6}Ertd6M0Q2*D4|hL#!h z_#$I>eAMi_7|lIHko_ng8euqkT|00XWTcq(51Acq+&$}Qw0oDi1wo_=7RdaRBFLMK zI!yoCndMB?R|Yk8l1W$Gpnl3mY&+s#?SXN;j{J!p=~{ou zEs^2=pOl(+>R&oxMV>|%&451jvizH{{aX|OFtJMzrfo!Ws+&Hs_URZ&3)cB&3iC$N zWH4*sskd(Cqhd8G#*o1+>f{f(VtBnm?kSDmehP2^l!VQd!gAMi==IFt&Uaad%-wfKv zoc?=jht!+pYNczh%S}V*8_qpbWVv#TDR2$QF<-s2!ktRnrhoo1E<(P>n}sz6(hNan zR|~G-8wjR_Gh6Zho=NZhBZ3})6YL|YCDLzbZ) z*^r1O2#T*Kt<&f|v59t=ktwA~Ka=rUmnw-shxz2{Z-V!j9q_ymBr=Tuqcx)7Q&185 zrlA7$paXhzA16Ku6ZI?y<9@OHCgpBTkBr_gEWp; zfxz80B0WDWjB$|I{L7z`+8_wi#xD_*B|NL~!?=(uF4D3{c=^?H{&bJQ;XmHRt3QZg zOy!_rEa#aFLtARlaJEJv_m-sAdUg*5j*IPhmf~Z>wIN+J^Q`ykTcRM|_2{)ueP2}h z0%Vv|n(a$hTKJI+xli?w-i64YW>#$7#WOuQ0P*3HZ=QHz_y zp-+$5Mtbg@AToT(_;W$J6iBe=g$tW*vLN~?yD_2!M|w*HZ2vf?PPuJBnL$Y>^X-q5yfbU(4K?5G0S)K0AZd(wUh(`4Hw6@ri*GBP<(6)erE z$*Q2f{vjCXvR9GOtac{K!06aM$G-$Sk$?Z~!VN1BOtdu)OgUHRQWehkF2t-Zp%J2f z`TOzP9NYQxesXQ2X#!qHb^lhN)j)YeV}SL>KG4Mx#mjg>j|gePXP+RugRyVi{WW8@NIn|-n7J5Ap=pyWx&pyBI=oT8Jiby zbBz1Nm~sB8xd>+G)$&DLMs;*bA{(&KYYz%HeU@DesGd2!swN4VGZhTP4X&jvPY8ud z18`SToju>)a>62_pgsDiBDBI|vMG4cE2DYr=|5>v&Fy_;PBi(Kd;ws>V->?e9LX}T zM?E^E?kq{YoPaMX*OSmxp|*beaTHDeZsWZvX7bxoMuhQw89zYFL%suQ+m^L$l=iSk z|9y)Xx+0uJs7>?N*+)d0mSXQhq&fHLsvZM)UMJ8deW$867rC{=`Wu0 zmJ|8n%z(e5;PW68XB?-ltArh5ZYYL&Rj1z1g?(}&AsSXcYSbIDdJe}%x}ulMOp1)#Vissv_{S4V9r-n^AqS$Rmr}cFn&&*QrPU!&qfW7wgY) ze+pC~sZZ!ABc=@@b~s!QpO$i3x2;Cw^T=$-)q zQ0~m)MmFrWa*MPsdwx?Zj!Cc!+`UW3)W&fsMAI-$>{Q>{3E!V;mWw^Z^&XDBF2hRe zPYoe06?XOTv<-S;wAbPAv;2~bM$IW1ep%;-9_)Wx%bhLL zx1Cw@DH-%93R_ps2>`-AW>|yDdfR3FF9*I~m$_GEGVup*`n{teaW_TnNCZsH66fXNuP6&`yq}lGfsZFYhHyLXc8jBi}He6aPVk5 z2M=yM1-^6Ab{gdOAaB?fU){#GN@QDy#wOgYR72y;HnEHX#{r3nA)0GZVbQJ!E)w-Y}*{Ftr3)rZVwSmwIU1y#_99@@{Lc8|$g9aq8jzhCzuVLa3QidXG~l zb7uU_Bwf45$u{7&wpk}SIJxCsw*GM}FKhw#+i#eRN^_@dvu`bw??)A=JM^Rsdik*I zNC5M@r=_)cnqsm=q=$@Nd2FP)8yT0Lcs_%c)Mw<^RoNM&=|!(hf0OLzYrhQf&w*W~ z4XKQD*M99%f6eW8vh5v=RzeE-un-zZ0l#E*w_06^`j(JsSpqbe;Pcq*Z6(e%8o&3b zV)e3mS%2R!%00Y)Dq^$&Q#~J0@3mD@%e=P7{Vlczuzsz;G;~QoWhT;>1u{s2MrmQk z%qDA=*gCt$*1oL1g5`ccj~1%hIlJ(#%w{Z`zltg<-?;4c7w4c-v#I>Mn;?{JbqfS{ zD)**XvGqcgYUPQUdY@uF>2}*m@Z;I4SGA>;w$rOC582Iyt79PBJ+gs*^Cea>r!oPF ztVyoY9A-<%)Mx7-u0;JMG-GM%v(@&$QRFP}%!-lRr#tbdr&JM%XG4M}>jvhXa!oDZ zeSO1Fcl^aE!yDPeB~)ZYzUe2VSK!9+NV5`YRTjjN;i+|BQke)eU8T1E*jhJH;MlAR zev*ld7C2*PR1N`79*^nDQk#ZC?+`SQVER!US0FlqOaX>Nsrq&X?HZ-o zzqoN+w8Vx!l;5YAM?+GNbkacI?kLe`BsXug9$>hmK-C{Nb>$^&9WvO$e+X=-j*-C0 z)@#4HkBvy1wB4cRq?B>m$<~Ik(@X1LX)QLk!?lLtsV-jBtcQ;tRMVv(uT3_sAM3Wf z+gR?(y(|TcL93wHwsvH+U2hE5JKx!sS2uUmnPiMldzb@_pNO!+Cnj?82E@^GNArcd zklei#QQ5lc~XyqC)LC{QJ%C2jgiv!_iqHs`ECb5yFYJ*juq4PajV^AYXjTqM|k zp2mT`(=FmU^06S=X=G|>s1v5yd@N`D%~VcUo((TPAddu3&qjMoYkRb5EfJwtQ{Lr< z-IF#T-JEiv4fkRdMvrsNpPg@4YWWQqXMF(xH1=0^koxJ2{FmzCyA>~a~-_rxJiih4^_UwdyE?| z#4z6tsp?I-PSbxf11YfPoY=j86q)=$Sg{SZa8*%}{y{9=jl;4LNzGxQ7i9ZIjs14g zn(jaU);**iIK7^y9hdKXS^Zt7j@K8NO5CM)hV}3qzfTU}@}W0|6M@1NU4`CcX@i_i z1EapY<)8e|6V+!;@%L`=e9K-KQG2kz1b8Un0Z1uX>Zi znQ*%YFB~AKsFrr>{mFVaifWB^lApTEWNj!#UBj3$S!)!h?d&*%yhC-bFw$$~#x1+I zv$THFs32c#${vXWEwyP=X^~nFBCbtkEyS!{LWvC3ZnpY|-cyDZMrmQT7KPoE)KeF| zMQQ75&DnZ)tGp`EYa#)awT3dY)RylvJ!M>63>&s_+3Ql@GGn_Cy>Xkb95yUdUD)^5orhwQg6nS|#Ef%JD+7!`yYWtg z*Sv~l^jj=1?nMpo2l?6$l3YrV;U1zA2S0=O))VDA2emJ4SBW8r3Vg8Mp=C2WG8%+q zHazdovWSJ4f%2ork93WTWPh6A>^t-m})b`pEU1z&slf428#-EFMIq?GG%S&B$opIZ11zplV({+dn5#$r86b>K8{2%aE{6 zT6X?bB6ST{Hw5sHB?@SZSLS%uBwLItZ@0_|EwHKn?Nk;1v)t+4D?Hr3!tAyFuso>( z`UIP@IuraZudha0^~*g*ax=6ZeyTkyq}GL{20RDIiB2T7SUH z(!a-ie+Pq4QF)lH&Hb!RgSSK{0sm=2RN@nRyD1BD%*2)vVuXNqoP=yXPlEJ)&DOa` zQQzH7CiPF+?=?I?@L~BX4Zp|_p;9*Llq=S?)iigi?{|%`M)Py$aHMjGASLk4RNB3` zQd{hjnkIZjw_&nX=M&GGvC!{;6Rp!o(O5U*WHb%$((#u)>J_ha#!HY~d47xPo#Bf{^w%V*xYXGS4qV7^Xcugz!|qmc0Gb)v-8zk~XVh zZ!U;W)lVQj3n9V9C@VHnpQ7o}X|QbpPXLe(Z8@avma>b?)1#pNb#|$7DFWF)_nm_G zyq~N>F9>LSoC{6lQ=&Jy?!8G9KWUH^(OAfA|u%5vs4v&?gy9jQ-=H9Ly}fnK+s zu?I3E1$Hpp3nDy77r(kM18Tv%3iv~k;OYGtwdLRp`gVmzLBvzG?XvOP8RNDFFe~tO z7rjU&nDLn4SDa$VN5hqy`wCysxLfF0G}zvkFZ_obW&WLIy=@wva28}RpI|}u_kqHg zhn}FC8`WV2;DZyyo>XHRfqB>3rc*ttht{-*aB)K0AU43>aWDw5d4tUzn8E! zHfded)dI!CbCwq*o3r<*ut$$V_v=YykUuLbibw5LWq-AUA@=dXh{(D1iTra z2DJe4y`OE1^OxA>Upw`hWjiTWxets&WtFmR!ReT3S*o)(&LoW?N4h&hlOjbcjkL3CxjqzS)aJtgA43lmSZ}(K! zCxe!)HT`fo92KbyxBt^H$smnAaIzm_S>k@mU^BYTrKkqp)3~JjuT1gP}#%!H0U(C&@*GOq3Qm+S8!$-At zvVhZ;ywnIvSUAG5wby503!)8@?d1S8o8-AJUS_vK0%-SZawooobf zM@3d+pHW_s4 z!PVR&Jv2S5FC+2E8*i};%yMik2;QM0g-szIxCV7+X90Q~n@y=uxi(4vgIJx1C<@hF zbo(4E3CW=P*@-4NyXHh;0t(2gX-nNqdz$jTI3aOD;62BQ9PZRNo4rjwdQYA?)>+Ay zG6;}kOWxl+4gMZ7kKXRiO%JIPvQY-J0dZs5=7r*=wmN88j`h8yNlFv*2&#aTztf0rfEoL>Axc=(%cR-)wvL=+$p!ky{OA-bafd>H+oA z-`ZPaG1sF=L`FQd4iH3grA|}O zb&WRcb=8<<6pezPNXw*~_VrNIqao8?bO?wqWmbQjed1{i$G8ghDKLiZYOaF!PVML(rDI0T9?$v{i>|ub&ONXZx$|(hl?zK{ z|F~#ZKs}XE)cprrH!;b_Y+tT#6K3W)MHid&A!N9xvJsh{x|-#B6P($$DwiVL`Ez*y z&5|G>iE|s|SgI(GnGMU!^^g0%sC)B&DBJgcJY#HQXD}Fh*|*8Q?-a6!B+N`nh?y29 zDU6+A@@k=?EQ#*KC}r*}%`h{%BT{OZ}`vHw|}70_m85b1K7 z)*C9iNIlzYJAHg&FTFNwc%9+$`sd{9hVkX}gJkexjU0Uc<*wU1M`@x=@dig`%Efy=niNkWifylRVZq*_ zn6{XT?>HA$lDQ?*&zK3%)s_X>EBE^l4$T@~wI2v@X}pri6Yi;=j|-4skZRNTasj;W z-84zouac@OLH-TzN{L`3uS72z^43D%{UnzseOuU4DS4}rtf1_b5fz^yxV z(qJUNYk7+AIQ~pGyX-ctVQRJ;qN(@%3W~ffbvMZJOq-3)r$MNno0_gm1};sVCLm^x&XEn zrtfRd|~=xUbXP*CK38X8ZFhFwVlM=DSCF;e~Nq9oOUeO zTRFN(ovc^YGEhch2a_%9*3fAg{xb-z`kD1Q7Pv1(cUD4=KAF9C z+p!2Fl(`v?7KsL2m2egi*S zDGD|`cR7P_Tcu!zca(0WPrKjprV~+y)r3?e9zDyh6_^_87fV+1Bp+F;g<>Qy?_;Dg zK=O~%m67c5Qn)r9agH5bPcumm(5AoK*%IL;9d`zSJQRi1XBiv;S>)RyDoWy{u5>#wrPlaBL!B5E5M%f4Ci#yOJS3I>|=oYe<&owF?vgN`te(&{J+MDp!9SSC8$E08p z?LMN!rs;fnoLdGXG%S@}Q&?o+4x5ZZV%paggujITc^B7KP>y^^_R?GSd zFyC!;jZHu9lW-qGT?4OfpKdcQgD?B{HGwTmO`Mu4HzoyT-4I(|7+dlZ?9^-UG)c_o zWfbYsn(4j|b%pS(rtvl-aCi`wFdV&KHdw=k0*O@okhm?`EQR%4^1k#;B{<@hwd_{} zJJ$@?h_u9Xt%>dS6g0vIP7&3Ob`-F zkuy@*98j|5>=IQrK;%fdGc4x{PRtNH=qT|Zl$-~{@{Q~*=y!bBFQu(HskA`f5G$|d zJR6V-%}c8fi<;_nU0=+%QB!d7!1U*rb|cglb&rh4dy1dLBc3CwBv(@C2V(OBbBh`t z>Mq^TkggdPvDFi-jWJu7XtOh=?B!N#;T;O~f9B^yp8<4;nUc6S z4HuJhizIEVCH8C`4@lM|a?qR-4BE*-hLgPy;0nc>9*UweyP01xWuE2nc)zOEtDiid$Bh zg1mMaB0aqFSM{pi+fAty|I+5B=zJ__WuB0nTKMAIhqB~|4eL~MY@}h)&N`Ja#W=Iivc)yqH%$(kMtU6*Oh^F$#JB2dJ1!8DaJJvEmJwN3KEE7M!Q=b@y>G*d_TT2o=mFHP2u|+!_%V#c1r;OMy``A6I zX_P%IP)crrf5J<8ttKbQX@nQbUF|eONgpSOjVQ_l8@bcm&wWf(`8;Al5M6EDz3L27 z#Gnk21?LDNGP%T;3DHNyiHMZnQ9k29=mZC;^B*ro0m;cn zbU*1lJd;1$4VlF~y|`1a7$N06S-!)ImSWB{@4+gs8??3AUT0%$u^!Hf@Tex4#}A#d z548JhIx8NFDM`cG*`u{yH8e<`uVCTjLZTS*xzJ`;D{9CyNaI)c!Cv zA4(Ckb1drrVgwGtHEM8c{L_=3?l4L5l297B3@{0cRhNA)6A@o$3Xhn3i79o^{ohQK z`7^|0V9)q=t|b3liPgqPy56SbS-cgi5h=e29jw&YDxLvn>3O-u)fhX+2>CqALwz1z znW+~4Q5i<;PVq4!UKz{}3M+J&BqgPYG0GYmpyl+VYL-KLXNv2an+L(aPT3!zsP~~j zE^lGbe~>$IQW zaH6l z|IBzMKxA?g&VBfdnX8ECuhe>IfS^ti5P75^p7N&VrqapD>U{RxCIg}C{Ady-VX{+>xXB)>m5$t&_Zh zWdI7`*y$=61leT_LtP8wpn#ztp$9{|jd!0JvY={z=fOWJkdTUyE@)OPNOG5xc1+Ul%s2-@aR$O46 ziY-y8E9v?C2rqx3Mb0Wobh~7Z=`hncK&x7}R8qPd=@Pa&7f%0sgK755-u~Gl zD+0;PW;qc>)x{0G5ipzpkyt3e%yv(=b%U1nrLl8pl;x0d4;vAuSnsh1>2JZpXavf{ zORqP}K1Yz1URMnr2S!>y$tFb4Sn^@B)`>AXAsythHfaG_c06ek%wmGPW*+pX`a87R zy%sI6+I;hSY#=cmNg!Xd)*6&tQdimWArYo*Emh^1Pin5wW$(%brCz1Zm#tY=q?cx4sF_74(GU%4WR}BMSXbngc0%ma2451k*e5oQ7W=Bykf^HSYA8eZ8B&_ zNYD2bfim?V1>p(WLka3uH7s*2ywG`T88}OWk-yY%IgRJoNe`Z@HnR4q?G_+a7fo_W z{Y__wkaJaqv4zo4Q5*3uC~?h+{G=rC-a%%nS&Pn7TFi>8Qi?oXp~*kK z3Dc~Zjud;>80b~hT~`wR5UV{%dMPkV_rdJH!Lq2U(m_lAEm=`f(wlF%DU(-ylA?N2 z;nVAs_>w?djE9E8CndC^c7WIwmK9FPYXNdQ#a~?UIybvpUGvP*48tVnNAPPX`Q5NU z3OvzYGl7}$reOg3ou9}vKP}z6nA}f(PtmFZoOX#sNgS9d=1Ri!&*}O%=bgoGoA)VP zY|{X5u}@damZ|J3dt|K@iPAWW(@8cdvsTY;6kib@kJgMkk`YTfD$-=a(YY-eOFd$l#!=ohYH6}4HVGlj9CyXz{Ch;6GYD&Jwi4$KujAdr*O zN8Jt{Dl}$p;1OQolf2p_qXZV0sy-WjGc(d!z9p>kRLb*h-eMU;g(*X1p8YF@wJt4- zhXz$n0`X|H(cj~b!I~@1%PF}f-b?*6{q;s7qLdCiFxDz9OACz{ZF*; zV;Hy+(oa@tjESX6neo$XHTMMKR5Iu}{e{WE5~`KBzTh06tKN%)v?;8eYZRO1Yh3Zx zA`;>6IH5SH`5`h+vV8D(l0h|kdnb)a6}v?cJ3ju}WD2a_OLwLCnph5FW6NSFuv?Pj z37(|nO>q~Vt2Sd7RWXgf{a%ym9w6&lM{9=NZB5IpRJR=Ty1f4?lc1HTR+=xK{dnQ*Dw7su{7jH2|xO z*ZR52wfaMZ-~f^;gDUdnjGrht5+Xa`pG)jRNt_KM&(Ly>RWx|Be>qRYv*R4I8hOy6xcMhfFfa`|dS7jZ2<{LmDXO-Z4L z^qB78>XYd1JsGB$bAyDimqQB(zLo+}*bc(xYY#8MB0jFu%D7%W4xk8tF11Hhn z3RF9XJ_+2j^xN{(J6IdtkVH>87rr4#xyTOS>t)cqr@-kKb{v3Z?uL1cptp(#cxr=_ zA0Svii5LATYR{RTDf4+K+(VAUD^_!@k4PHREn=&4Y(MVPaKJiUVJbq01Arw~Qz3}E z)C7bb({nC9FM!q%cZ7!(BZ@@# z=S#e;(dctO?>38j{Ypk##!*BLS5$n8%xMENCFC=c`-tx;3TiE?-%%3N97S{1O@Vf*BJlaDTrcF1 zbbDmfG4xIf0JVC+7?>dsfHkcn5o|Em;^1C%-wv8`UD;Lcd&i2{UmE=FkAzk9@X^wlbI>;v#0a-4*7p4rsj3}oR*fp%mL~BelbHhL z9dOdfYh$`p&j*ts#9RrVpjjnZLIyh!TT=P1)dm8iKpCsY0!zb1Gsm*3haWSxU1z6J zWkB*dccP+~mpzj&L#%0mWrL%uq}?e2VZX`N4CUf;wMV*jgj^QaX8oum8%8C$cY2we zCaHxN_CJ$uA9=!0BY8=OG;2u_k4{%lhZkF6LB)tdL}{FB;d_ec9%!MNE?G65mepV3 zLV@I!y^bxu>zE$`Jr#sDXJH&d^g$@no22z|?Jy*RmfGQ0y4VWq;KPGw-8~bGJ(Umr5115SU9hzY} z7`59%8Gw8hH_STyrdv}JFZ~W{=PVoSY2(hyifV$N0QqEnC|Sj54wQJs)LQ;5Pwu0F zOeZMda4io-{Cs4l@YHxPsWEd~Y*8f+`f0K`<(iaZ*#mknU!fYS^Y{Iea>_ivG1_~) zjwxQrV>r?=pYd?JY@)Z*Y0}CLxXM1l?O0IfYfE;Mh^*xO*~$%^fi_F~ShGegRw|A+ zO_07@){sCP5TpfrIcb1b@=AqV_)qa|ge@upjYnZ6g#wWTI(jRx$41dui-?^2($YM` zpE@7G>vnYp>}g$5;IiLY;`aFRm$Bu^13`=WV~mqPR~8RwfPF>R@?AUDni)JLiq68Y z4*O;baKy(yfTaX<8pLK3`M?x#AL=8V9^q_0v*P+I&i=~t58{#Kmf%o; z_9zbZ)Hr~2UVu0`QNg$vG*Mkxrl6o8b0xn;us59<5$vb^b*AERJJ`XRcw|6+JX79_ zGl*tO<;vachJ8QdHPM&VqgNP9KOlfrb9{Q17B0_JAMBP_U(KEyFOm34S@OD@fgRvRZ}+R_NKk z8v+$Nn8vwfK4mIZwjz5vm|Jrd+*!HDKr88A+b?t8l(|ls{)w(rS zKzwHDDgs`Va&bl>?BhM3I<;0g@Cz8!({@jzD85HH z7X;oNU~Qo?Q8s;sWs$-fP4Nv=m?*cHDeOj6L3gG2sY($Gq=7iep>FS3v-A?3XQrMr z@RtKk)hC_+?=c4m@}oGL!ICFvyLp=S{^PYtHuJ0_o8Vy9{$*wwV9vrvz9MF830Wz= z*owU+fhNKLJaKmuk?4k99hRLTniH?+lSa4%ncW1bg%RyHD%bFC7Z#1@Jq?I8Z=KCO zW!=5ah|SH5H$^Y=q%I`srJm7UnvHd>4wxk6l`iOpE5Lohp`m7|$4C|<$3fy5@cH!- zXNFX$I#y`SBPG3H*T7(VY_T}h&mi0rlqWQvD>v*f%h?J8+k5*u=*G}DX95zutOZ9T zALw))1vyNMy0PvEXZ!+tW!Qe;Dj3OfP{Iheix{ z1HhoWR-XcuNT5;N>~P|HzSI%_3lX`p=LH^MNMT6TL9#zdSExRxtg&R`P&50 z!}oouGTESfFY~D5jZ7PB_0q6PgyK(+lQ_t>fA&pCZGP$M{Cx4s&9;Nv4{pDRd?Q!B zRS&WlwiPN_Ye;X0p7A^I(l7h#3db87eO1XfCR*oj$0ENsF#OxHN2et9B&4L!|2cEH z{_LNtEq(Vt_5PiD>du|Wh7TV;+`Rta>XUGSVOEk`$+1BVOO{fuN;WoZY`@rEUg%6U zmjc;4(G%(=>NC^QT1U8yGq{H?{eYDM<%OtnpC)`2@%z&_=v<@~>jPDqJe%kxI!n<~ zfx#+n9M`hPxLxvp!O+zZKI(dt9|TZGiw_&KI@uBeY<;12`iLl1bnk@iZ6dnikkkeC zWr6mKW>C)Yc*Bh(`9bn`RFT(;%Goq4_u`BWn^U&DZOaUMnU=CT^-D8lU5)ay6H^IZ zsDWH52aH9VkpCQ;@BGv+B8LC*1AX^ zlY?M22oA=Vyl^)!slV0H0K!S(fuzwT-Rz(o6t?k~7UXG->^Gz*7BkeN;rxrf*z91&!I zfMn8D&UTg5a--slzfT+!taDW7ieDmV+O}cHKVNX;mT4ij;jZQP$9G$o83kL3kz2e` zr)>F>6&LhbK?nA$m@41gemijUGO=N%xcN@{P%^xst1(Gldx%RFyUB;%n@yx!XQ69P zRT#`#Yx_(__y+5UY4V-w3l4G(!7+>PnnPUh;9Ha{X<%vO3*SF|c!Z}=^4PqRpSCm< zz_-4nUo9gBKtTJ|pNwe5rtL3`E!6>-_p&~5rfZH=# zI3Pfk@LlVx5?KoK5{z+)MI?>?L}|b(m&6TL4lm2$-whXY9OJ_Q}vFPI{1JVXVJyVtoDT^$%^0 zW&G%2$gxA2+V^*LxKG`6P;=ufz=dSi4c_N(?_fyq6e{?;0i8@Cz{K!zI-%YP=ghk-D11o zp&45iXe+Tq@fTQ44P)npodnqzoJ^kXC4K$twU)@jJ+JnbxKB;kqM}=3PL58L1~M4! zJQ-KD#Nk^ArL&!u`f?-bu9}s1V@A|tt6Z!}n9VWziP+|WVn>xxiq^0Yo&?OKvf!jS z0Q%9M##&We5C_Gf1*##wR_h$SO%Qv_UrMfq9puFDq3$K66L9?gRL+P%bJISdQj4nT zv)-$cNc+=KtgTt|b!_D-CxCzyP0?r|igYlY;kMyO^=JUwIkdpevi0>1wo*}L=bBH< zo4YDIU_>kUb~^7{LQjt3#Rn^Q8SAsAA)64>T+vUAbAa95ODu3yVHAI^W?z{E|Ii%S z%~&qfp?XM~ue6rdCGA0$z}aat5`@4nrJ0BgE|&_aMN4Eb&s-pS&6So~OYci*ED*nB zY{=3rne;@RsxM3Q7rn~VLMY+)YYuJj-8WqTge9vm-w6j{u%Ie&6C;%Rbxny%2)VUrkx{L6|6Oys&mcvi5o$L4c zsU6JPZV_3yA{UGLDA3OsK2xW3ToG?AtPHlqbnT!$0Z6Vzmuf6nnt0e?#rJNNlii}9 ztn-@(EY1Bra!~SScoEAXsdn|3lLUhhx&zMR9@;94BP9bs5fGQrb5&&aHb;>S%*|7< z5)?_~J6!o@l11mR*399YB|3mOvPzO&uc%WO$BSlh)o}VR&}~2LQ2AczlM!KI!3_-p z`x1%@a}Brsk^r8;k^`)ZY?G**EKL|2F?E%hzWk zXp+`MrF`7I#4=qT@|U@oV}~rZOsD5x(3hK5+C15>9^f?boo7UtW5kyHDq5gnBywEl zOCKlSgYAp05Wm;lWOEGb1bcyw~15$TOMd0Tu5k!HTq2%4#k@0lhER4kq zti)hZ8b!+Q_l-HHz_ zNrgXDP}-(Z*Q14AA>5jD*ZGNpw%w0(05Lc)Hv9|+@P|t-`9z{slbdwfn%PpD%JwZO zneO9Ezq(S5RJmQnd+Lh)QZQC8xRrocjb$bP3roTVlsS_ca~RZSB*P8T*oIT6X~7|m zUOGTTK)F`{BL~b=o$>3s%1i62jkAmotlX))y<1*$+A7AQfteWb*%$OaS&}UfKSAzNt3+X z%@ml$Rn*9$l}))_pd;_vGQ=(7|8`;4{slMrzkdD(su_>=(HUo^@y*93=gmP9aXM4C zVmK15H_pslbRBzsa-7{zh!|CH@_)jg}98y)L{7w1E3w$~7PT8yg1oFR~(- zSghr@iNm?FSWr$0@zQmzDFb9p%buM0J?o0fe<=OFebUJ^%Fur@g4E%C00X!9KM0q< z`}%+VAbHi6IYvCEXpEq@6bjwg@2k1N(p9ChTUu5T&-kdMSlUu@S_V6_hEnq&FR)$; zKv7?l|GE>C8r6;!DVeZ6g)B}ST1`>}I~V$#vfXNmrQC>qk;LHu%H-4u_v(>uM`c5`UzaAo+Ib_8K7!I?+RsRs=qPf2H1(6iUrwIf7>J)%07k!k!-BsSGabzk- z=->9@f%7Ja0un*Dm!Xj_MI{^wTGPD>@7Nq6S<_1@fZ3^XRT2RK$aPudX;fC z!}h;f7pxOH0_k9$no0X8gS~vJB+QpV0^pa}$>PXvXm*ll$?qF`9Msg3OoVub$yc1m z>Bza_TPYP~b@~7ZOT#Ue=(jZF>ZK%j_z<|AvgSj&fO z>irZGN?LXLQHqRo7+NpHPs;_b|E(K{J3zQ&Yz6SGlXZnhahC6r#7>LvTVnd9B#92@ z@B6RjZo|8VQG%0mWp0Lv4$<6W$ihgxM#zYgxi^5FWYlCTUa34gS*1*Yi3RZ^0G88W z$pBysY9~P2vK7Z!Wi?mVHcDvn$W&#qK2W>;*y6GJsz2M2gG5=X3b`2|CF$$m$GKYi zXf62E-k20V zDyA6@kXCO39K)&ss@g`2#1k+1jNx8f+XW&Ter(Z%iU&3r)B?FuwrVjxk z8;O-41gWQtgnmhYIpjNL?m_`5;rzYrT-97?oqO#oNbecK$poF&XZf2RahX7HR|Q6_qHT&rZbd*di?CVU*;Hf6E6e%Op9JoaV@#aN>u(m{ZVz5&n9U< zW~#4)X|1yyXM`gdWJ=`8U2~P+NnJ$n-(Laye35^E%HwwlaM4$0#{W(k-&QTiTH3C@ zgo`Q?K~|nlXC}-126?)1#88#1-N8Jw4g&%fOavXn?||B^0NMRt5U9V7*(MaJEDfp0 zR_}9E`7%NKDzLmKl&Txs<&Cf+8T4pjWYPg*S0=23Wxk2{6y5^>J;K78q+hp_t{Q59 zb()aY6ZaOgT#H^h#*DreINLX}MV6mGCAYQcoaduRIvgW}(7deF7rh0@%;|SRi(z77 zD)%Rg9|_d`Ma6(fKKP@u%%@vv1TDFNzIg$56c66|giHYtB9fm_a3uJC9rQ+FM;8wo zbOsc+5|r0D;<@UYJu>LnAdiHCvtjBwvbmCEG>9>Ldyc{`VAMX-GA!)O!7+x4fp+c@ zSIs*lGI1~8$(0DaNFq?~l-;q_Se~P{kEqx_g8h|J`6Q#%B__n&_bpGwoM!&p!X#fu z%cMDJ1fU&0GoqJ|H&zMERhzfFUy3dC0^^wsDQ;sM&lOi8Uffo&^0IO-1OO?>HS|3o zr@2!?+(e)eJ*i6p9-nFcws(<*()?5DFSZwTaa@F~TW?p5Vi&=esro)$kfQ+BdD z;G_#)E0r~^AoMuR4bB(P+Miij7H*EpM(oQotp~Q`RfcLN`pi^f_N%1Efx`WqGJvdB zirI3qTcXg|QKW0gFjwB?E?g#w_NK1b`D1J`cCygVQEmeqOGEwZ3mR6DYaEeu8fE>` zZ|V8#o@U`pvEDImx~rNzJ8SuQ1VyuNSHDKiM6ouK+}h4f>0lY!T9ZU<{TPZZmiqoT z%k8LgMjR$VBO2@SQzUKD)-It4yjB~<*PqyI&!jV-H{M{gi`R5%MgCr-^=RcdEnGi%j59_|*>QEAVktiYSV5dnNi+jMN z%oTj%IhO2dI81=pxmL`E9a3jmo67vgE^PkNP`(X8(^$S8df+NCFz7(8%5QhI`3u%L z+=tixMv-vfB&T~mn?kTu*P4fh1#+eqJ}J#Z$ui|-27y4EDh8-x8%f$4Gt~nrm*S=u z#anEIJ`6Nm>7GtM1-5BQ&Zi_9lypP7n5o=@lGAMeWucQ?S*AJQq;t9@T&&Bk-?$^4 z81R^`V(ZIL1R&3HesqO<7)@Bkho%Am76jzBY3OBLrEnQ(Va7|NY9`V%#^|O<@+JNx z4I_lkW$b{7cPpxZEut7}WBud~VBac%sDkTyphLBv{AZS>#Zg}HF)rG+$F<1A^EC+O z+^)pSp1%MS#yA(j3c9S$@Z;|%^qoVCUn7RHWYCu5BOj<*8`vEH05G^8U_`V?4>*b+ zy^ACtm?=Zcpe6rxPr;Wgg(CvdL0kDhj$*Gs$lb%6cujN3}6`!$<&=VEf+*u&YHV5 z@CDaD`KLH@X^ji&y}#GI9Md?mv_c0znnw_-r z=T75Q+>3<~`+3%a6BV{fw+C>E{ks`sd?nXv;x2;C_(&x`2B7Ki^(|_f*@37?bAX9M zngylED~LMi;*0NnX3GAJ_$9hMPp-1I{P(ItJK*usfVS8!nJ&oVb#8zI$+gK`b z7k$BlUi3T#0f8^iP;~6gK&299d`yab(B%giJ;L{r`%OPA%6^~mTe^R3c!$L z3B`z8=YV;P2IHDH#Uz6_%>H!y`gfjfY)g22X_YZG;$U3KO}=5n$`gLLa;E!> z#-uG~cg-0Ju)*|q_YwaNTaf7gqWB=Y4>UHDKaIWOv!McqZ(bSRND#tPrh7kjKgv`4 z{bfwWWL)|02{C-p8y*(Mi&n&bkrUnps)V(C$%?PCw29Q|WUV?=7k(3b9BeBtem9<} zJi6Q^(HD-hFrEBseCRnf3Rximp2uA*|7rwUrbd%qZcAxyWFPjwrzGFf*vM{heug&| zP{*^(?o~QcTT{Tl#7kF&=#Rz(T7l_YA)rDpUf&>pu=1G?gG*tm&_={$e=BI!=~#!fgqH#|u8Y5=&mJGt}bze_}1xzTiCW>b<{h=YC5 zq^B2MR&K#X_!Tc-bT?H3OqvhXmjqzH!PQ~gJt!2(-U`HJVV7QX)s_ue__50uFSM}l z$bH@=q)>1Pk!hE!C?>XgyfG0h%jnGJvjYlZnQEmR5g|>(!Wb*CLO)%jWArmX4r z&095Y)n;MG28*j@r2R*A<|)c+pB-q=mIAMd(`CPx|7bxC8XHZo(Gn> z3SR)6aC+0l!%dUlZME-&DzB<=SDq7YoN~wJjq<|3e|MVv>zgmhYuCa1uv66TfYn}d-Ef-F?+^B6Onr|W>wPjzm^=I3*}dgUL+cl<;`4^{ zD(%a6c8(recPKkK7B5PuTvz29w3E`SF1S>^e!CvC7;*N!yF^?Oz_L%tR1LAa({NzT zo{Nkx-(W-H%~-RVE34YyUF1)#_xc^Xea8<_^WI7I0VYpgqanL#DZOj*pMW?>fzMlTmKEb>Zk#g*``Qs&399@OkB z`jxa_*20JN4iV4x1S9!$H~W5B*}A1q*Toe(fTyyF`e~M-z`Z5?{$*RzogcU8+TWLl z7Rp1OS0$>T#O)g1!=_d;ODF$SUYPnbft)z;@(S^nCFWQ{@g-3=(VM~52`gZo9eiuL zar9>vqG6V=)$qxIAK$RDzkm2m?2Q3CM@2Obv_En4@~v0jBkPkf{V+R;9P#Lj$bnZk zE6W@wCJDV4S7Pwpbs-m8W`9=va4S+{a|H^PwBhE+ABxp#ziFsNrqGu5aVIQTuw&|p zK@(i|H(sfraX4lq(A@UV`#AN%!qaEFs#_+zj+)cAPgQlDh-FWhvyMy!caGx8K%^OO zV*eO$*;OqTTk+`OH`MIR^KPXda!>TlZBHk25n~`^*=!b-MGT}N@3&AO`&iRTYmU@n zP{C59)C=xCTt)~VR?`{f&|=|>t9a*?lD z3va0-1*Y+sBj(TN8w^Jv=6(nE|EN2>{4uutbm>9GZ&g=cf-u83I8P;^VinywJWW<< z*VH*Wd2sFBV#=Dvz&`@f+~D^ghG1)#bAK$^<)~ahmrw4^I5t>1&354Xl(y*{CGFEP zn>@HV)va`gl&-4kAPaYjAl>OnS4EbFazc4!>*I`BtrWG>TKfcAn-0ig!@<&pk>he} z(*&)Yl~pI|EBCyDPgTn@n_#b?loz!LN-xN_%q)}>YK=Aw>&erHR*%@?)qlc9m%G$r zX5aS-g1Dc&H?^bD3T;@k!J=OUI?fX}1U{jkHzo?RPaIhtx<$|&&uR?i99?%3iQ(*- zsXjYV_=zX7nF(KMsOoi^$m$o_C%|r6eiXnuY2x5D#siMs3`wPZag`U6Zr2^cI6+0O z(!1BQP3F@$Zm$dM+0v}ep@Tges;*_yyjp^WX*%sH=gIOH^M0L!hyIjZWxTH%-$PZh zV*PN0i4v^biw6AP$M7T^ca2sFWYf3Xt1UiY%XAC_Afow#uFA5jAAyg0a#s)}(W{V2 zPqQvM99#8ZxR2MqK0b6_u171=Q|+hCX`M6g&)>wXZ)3%D|ux%iSgQ~*1@XY{S$YHVuWQT*LggH;<&u2w^z?$+l!(Fw_l?$?n=SJ@y6o>oBRCt1Z8lRt7|xV#bPLLxC^aS&9ru}@E8h_aL{}J(0wJp zbZ@!JJRh5eaRnf|cjS&)&X@P9dl+$CH(~Qt&;JK^zWg8V99-<;*+$LedoZuQe7}d_ zG5++3=6v~+Upnh-xABknLy*6YCgR!r9F~3@zv|S;X?-v&YbVR-OJpaikj35;KGYX$BaZ?vkr9-h=vYxJ^R2jMISZ5-1rqdec*;INL!_{@|gz zGokNq(DIt4uAWyfj!{fu9XlaP@v!ta1B4twR-A5Uc~Nx=Jautu>vVfq)(AS5NNaJF zbomQeXqc-u0ACwMLLnQNe2zF8GD!B+4eJ6m;Z@~z73x~3ue%WinI?#$Vo7t~s!ltt zo0B>*fLdsrb^J&?$x~M$*1XGVJ^*}H8l>Nsauf&s@Jx{Tdr`0veFe$-YwRtX4*p3R zkb>y?j>CzJoE~O;aWbX_1)m4z3GQI_g(W4+{O#~Vc3&Ewo zwY>oiodIO|HXk7m3jVYY(Hd;&apT!TJbYS-HVP@U$qLOPDSx?zpy_%An#`g&$`8b@D0FVJf%xM z)l~||Rr&TS^E>!V!x!yx5E%hkUt1^GqtY*cAWnN^P^t0HeT zI%N2<4=n$2xK>wzfH1&`&oh06ZB{)s#chLBqGIXO3MuqIADvVv@Jza|?dJS09(+xc zsQ7N2iqaCwf>LUDW#PGdHVNKrnAi^<9FmE3I3bD>S@6+S5L6P=a9XD;J(D{7!wJd{ z==R$aeRL_3I+1hqcIIn9D!EDZZaY)*25Lq=2n^x+&Xocz&lwN^^2t1P#9YvMt|VHY zqiSDJL;hWWD+^z)`YJi*b-r4YC{z{@@eT& zrYl0d?fW`o-1b{CvR^sgb=-mP9Ht_q`vL+zg-R1NtM++lzwv@H;WP z*aiMfjF}@-#C=-bD-59}fwC+OD8#{Jl>nZ~L00Hs5bmHu)q!jNY&YnO#;&nxV37(y zb53&|>@DCXPjYkFTs{}I$HC%Fi4jL}_fp^Fm?I(M(?bhfG?Ga%Gy+D;) zx}J`En?W{qR(YC1^8BlwhJXPe%#KPW0G0qvUjhC9GJK|~qdP%-VzQzK+WC;PCcu5u z2Yu*TPqQu(Md2H$O|b&QbhEFv5KUbD(`wK}`-mFuR3fn7TZ3BfDX3zv;dKbUWy|FlNB5}uMWt~Wb5Ei{!JnUG}s9&PYwEU zGUy!y3VGmD%(=$|5>xUQlI)p;`s4T&lq9k>Nnu`SnV_oZ#^5~YcsKvbum#NU_q@IT z_+Q%Fnne>Tmr-}`cq9;9g&Ca({+Smy6eats-Ko;MVWvD)z~r^aR~J+K*zPy>O;*sZ zhPU)Q1A4=H6i=)bngF}NK8*sJ!+{*GC^QaV^i&)4DztpnPHIAfTJZNeP)o_R7x(lC z(gn%sMn_E{1+{!drdA>NRt;H}Mp2FiHkdrW{X9Xg;KtwD{<=KFZ)Pj0fUrMby;lbx zhbm?@g>9)bwTfw|ze7B2AlX5$t327;FMr+(NFcM#9HmdTs29+J{x)nDDX6{nQ@Vim zIB-^?z_W2T-b!fAo2Yui86V~={IUNE5EyZvb6L)Did4UkXeeYM;#C(Ii{*&jOwhXr zfNASP@~Dt^rr?1mW`~jPo=GOA2L>xW+3u~Ugp!HRRZgr`2Lr(MJq5N#7gsa>RR##L z?HEJyVjnQhp0de}{_m2ZkXwSHr|a}o=mo9;7@u^3VQ$=^lNdWlLDxfMwFg!Oh|I00 z*BiwO`mTKr(8ZNMwFKOv?j=_p0S*Q<+%_IwAfYJfYOxw13B7$Q7v8o7%|$^0&3F zlmjzGLjT<XbGkkaQGDQydlpU+rSN6_=YsNO|$2)&>--3-ADt3#sC`;K1WaS(*=AB zPyk^|E9d`HoAyXT6KcOX0?JQRxPZ_lNh}$wIVPkp-RKKYnkU$@DS6>^HTT>E!+n=} z^&t4`y2*+>;K4%VTrJ<(^qM$U9a0d~I#}wN3ruUQwQx>4PgC-e>pB0|$+#4Xqu*C#0e`o;7NZyS7cm%Ncm%K-fGtl3p zdmK|f0#b0YwHipBV;p_~P#a4e2{U=7c?*awR+ITd829Kmd=e^s;-X8$9YAs?GrRSt zQ0W4hU+I;S(9h?yf#5OQqvvwLT^ct$ehZIEufYWa1?rBEAT<3%xx@LbMzgBXH%@+N zOFJb}`pDI4)$ig8zI^>dsSzb2_12WO^OItP`JJ$jyDyU2GE6tp%s90pT)nM`` zRD{>I494L0(Q7O)95?d&{eR&8)4#XXom>!czOGF3ZPhs^t!J(CXPu+IujGT*ji;_9 z44You=@{`&nhU_i4C$uRTF-zDUp;d0O+%s9fh%X@3j6LZJa0d<$M|m8JFH_ASvzm`+lRPFT;tf6 zzZ1@>6Q%El+MOmpk4S4>pEzO3y0=}rHR1e+(r54um$4dWiK|gK@pZ=ZDJls-6ArmY zJZawZj)2tvQZeOy5l_fYBz{ud;zKdD5YiWXha&WcfC(T1yG^-h(y;DI>Q;w zwnzQ?U&Z70k3V&vMO#E;bZ+FBc5o~~Jl7RCVERnH?a=?z)|baa zz5V}(!9-&zWSK0fFqD01>=iOdhCxV{h+!;cWQa7jtaGzvUt(m)b}__I_G>9*EZMR| zB5PSLs^4+%=lgm5w(-l;;&+B#0dB5N1b#jw7jVOO0*&mQc4C67AcfZO0 zt6)w4rFz~n1)%}NkkI3MD{{QQFY=$7OLhI6zwO#b*GzhE?iN>0#sCR=Q zryYXQBhc(F`!gq|u94SVvdi%ebSFY=VNo7w=e5-Y%M!dvPOrqi#2%aY;TKHWrDWsO zH=+l4>{_#JEouIwMM2D4g8VKKiJA7Hgs9mp-&C46BUGSpvb<>Zk7~P;^(L-N@1JKN zFtXdHYkIzg=^{Vz9CD8I^rcuupYls!>FcnYeA>OIbz$x#E%IyeK<%2INB{TGU`<}Yg5Q0OWM0Z*==-^Uvtm^al*xN?kvT;Fx3#$^yk84DQHex-1@_a$F zJh((o@81gk;h0=*9^Y@M+~e_wVBFQ({!=QsOqB%|NfrfkesWV(EP1C*1S|y_P+oM< zT6F%OWBxATPR9ecxBc>WT#IR1s7B_C3PFyq=oT`m88Uc;*Ok8xxYw?Thmb`6`q80+ z4))A?$!kaKpVjylcE{+9^zb=ew?s}NFj!IrB`%I}lrNj90b~ffh3>MN#b2Ib!DP#xAl(DF#Bdx89glYYv9wu-BbO_3MIB*J9_MCt&z11h zVs)f|g!r$_q|N!gjJy^eLpTgY#{()|3kb6SQ+p{}9_Rrl@(1Z?TPq!X4zj zW3j&F#?1c7%g((vpf4ywor#DDxQg{Y@f@3oO%^P8P}b?ecH`>B_Z{{DmafnXC}l%s zLs-`_+s-89BA2HG04%Ajf1r_%lXcFHALWcPIx2v1uBQfvj{1q1x&e5uR^KiTOIGM& zl{yVg5^bc`nZv28r0l|sLP2up(gg%w*83}bV$PQTY_;P4^&>OZWm0P$Ucb9nNV`C} zKoRRXIG&SK(!z<)4wFBgB=_F~J*~KRj%B8FDOPq?HK4EaR?H{RzN(oxS}&t>+eDdz zH<|k^C&6=TEIb?uk4(1?*YvYvU%nd>=CP4Q_aD1aVrXALznSg9X6evuA-&kG4psju z=&?~q|C_&HMP`V)i|>;p#BT|bsU15CjW19oy3_Yp717|bBs!O7;2|!^=nH?GwwR1Hp3tMc!-zAQ)FNr@8^h#7k*(X^f(K(9gglXMUBHv`f3K$5ULmV-*0IPM8 z8Il)Y_=osGt=<~k>x7R2!QkA;3kxs>3eAPZm_FY%N9gmND5*BlfdlG#e zU+{g1a4IZm*L06ar=o-LIc&x)8lHbVR4+2b%t7obJPvP_oHX~Llvkl9ELYM5kGTB| z>$=Z;^|SI{w&Y-J;r^@?MGPIh4cvNFFWcE0WUN$B-V$6;?hz`-m3rDMZ+UeErei71 zJrf||C=);%!7l$UsvohcXgvL9)(G8HxZmR$hNYaU%B8zenpH?5^CCWJv|i7w{~L?U zvyyfmyAqFAAE?7f*Lq+**jA+TYoZvyRQ5qj8U1r3na)LN4*aq7EAjJ6v5jEp z5xTI>_bTP9EGA-l%t3n8Xm?A?zRNAs6)w10ZX8QvBcyC))L!%}Tiv+aXEtzl-5{W= z=xT38qf*CZn%{Z9Y%Mg8NUpU(%m)gG~aGtjP5?fJvp^gTnGt4L7Q3 zC<57uU@7O;k;HfmTdgucldUw!C*$F>9I%V(p zjzOu(_p6UsqF=Cbr-smkpH-+58~g0OpsoF0p~B!)6f++P zO9r+f@fAkfRNu~dftntA_lj)A8yeDfdaeFgom6qs?m7GFugisCJO_VjnYoD7NO*(b z$F5jOtixmmLSWX>xVzIGjJ70XR+ud=1H7?V$1_U&QuUJF!XM+JLOSDTir?CIPO+$^ zA=7t1Z$DSPejt-lr$Z7Jx$%5XI)HQCS#$p{S{SoRMO6ToI-;K0bVaan4n28y=w0ha z1p`O*tjJPcAWcdGw_+}_eqLr{7A>)UoaVSVg>RE99%h6h~w-Io0Dy#LT;juyE-v+05WEBmsG&RnT0 z!fRcqBYCK)3&Y{8g)F@j<_CROV~w>vU)DtLMG9+zt;bo(;0|P4rQ+Hj*wzX%%U?pG z+k81S3&ReSK9Nl-U5<>4y_0{-JE0@!lFx(gROZCxJZ-l9RtPTqB$Z13-PHL z(U!L>FeO*8@ag0XvD%j(I9f%xz%5SMS(X|H!7HwEm#l1JEh6aFAfrpBYq&+h&cUTf z$#Q`Qh0||`&=0Sdb^eyb9sR7%o#dC3?)FTeqNaI|CuiXDjkc2-WSS+IV&6P5`1c6H z`?Soc=Bti(=z@jSOG7c~TZ`A@lXl%#zYDLW2v*cw-FynTI@JD>f6EnuttnHz)aAXA zNE0g5?-jbZEgfDyiol2|b9nbay#4^i!#0bz$~%VAB3zBH{`D<%UAj!KqFxg`TYY;a-}KH(Xg|1dR0jR|z_?$*SY$)hpqpU=dIAwStHtP&R* zqH61JVQXkpE;jP!({PIKmR4EtbXya8KE+>Bc(9;r@eZuHD zD0K*F3^^&PZ0{Z>50_mIEb9@9QpJm?UXM>zwOM8%JQzIt-Phv!&!(MBF;Iwg&bMZn z1%Ws1cW$5yjQfjeJ$Y<4>FrzEm$TK`O^fcu3HKJV7_P=@qi)Arxg?;+>o%mx*`bQQ zIepG)QrqW^EhZHxzlsw-az3v4Dp(llo{Rbx>4M5Uk58gSZ8Z03C?U#pu|hwRN^VZB zSr>iRAm#q`EjB+!opska-b9t{<;?kT>XbwUo5b=XqhwWvSU-`|FYH}S3G*%?kuMjf z;-25tI$taP%l_aeyC9H1sgo6P3k<3@e`|!r-bMH{P;V2f)J%PM9drJL{G`84b*$gkp{eJ0l)I8is|z-lr9hBoqz z?t^}#)SohoQpGi(4r7VcgH{v5tt8JZGc{&ev6@va0!Gd3i|q^A@n64pW;Djxh;NxT zb^}D&&c?znnWZrvhqArSbITmIMk46JD9IZvf}Lq8vu^3`vz9R#IkO*&lYj1-Uhm;*N1?Fmt2Ij|> z3fEpgZZoKv5(qHqHY-^nn)@U*G(6Vq#Vo)0EUHmH$!PbgPHu~vAB;Qib4}QyQ-itn;j*Pc)zT{gBv0;^k*0sMAs)K@t1X!D6}Y# z-O{P@zBm`8#faLM5Wu7BlZ{&MIQye&)xSdq5<9535(%(~rjdkg9?F9XQ%lOVRlqOI zF)R+JBFB)~+z?&4Yh(hHm}f|COfj#z7x`joB&`O zOK~TruT={Ev}ibKRl3RmfFv|DJZxrpB%-^)wqFXY(vitu=F&JP3!Q%Jg%aF=gX}og zxPS1KWX1w5Up9mRDJQ?>E$PsRUI_>S{CG@tmXu3-(BN^d?_xSK;8rAW-v(JUzq7P{Moo6b#LQ)l) z$qkWpHqZ;J^S5krWN`zp-Sk6yP@-Zi)s>6*deWeW@|0hZEdtwtoT3Unu?|4n#zt zz5o%QigB_sx|J-qll|HiAkvkjasL)ZNXxbJxt(^efI*mRVI3$)3p6ETLroB)dR&*w z*sAVb^HV(3(%}R$Uwq#Qp3{zfq(AYay#QKNz&<`J1Wlb49p1jiFbb-V=MM|xU5y&F z1_tvoU|7d|6$BPx9jfHJV>Tg53!TAa#ftIAvCD}BAx4>)vl14d#Pz&T;iVXFptxxL z>4K`J`v!9`NDB(22~K;mL%sd5?}&I`DX&&cJskLE8#1d1e@{F>2lqR{;rO`w6do`u zUf$+X&&lkm&*k~{mYuZ^)nboF8JaiO`9C?YN z)^S3EaZgb1briwArJ$8TTJHFYXABnKRmhO%mHo#76I_*?|4gJD`Qnp^fc6oo8KWbV zTn5s}V-gNPLOlbsCuw&+Kn75$c=PRai8{QLD=UG;uge&+n_oH--_8PguyOp+DRy}t zAmLtqXq$!VNe^z?Q(yj2$9nEJdiNG~bQlUY&Vssxn0YFjfM-%=XE7jqJH+pb!H+gd zTKpTwaOy|qHwGIIzCHEd<0FDsM~7F5jOACkAQki1;Dv6I`py@@^~}sP12PCVEsAN8 zc(vQO11D}G9!&1r9E@8vzU5E8mz+643l~)rVfMSKngs$8aYi^!-(Vr3MIX}AThGM4 z;h{9H5`6Qmq>KR2OHReKhz$5dq1(H_IGd z2*}Fo zJJsZhVtJRs5&yd{1m3CXkU5<+L2pqsd+ShmqO@Ebs6QFqHV8IPhy`IGnlS`~eHEt@ zk?vE+$AEXOwVXTTFmPZ7O(crEl>fm6z`V5fR$9&AHEFKe8u(;Ue(0|IZB9>}$_nZV zi2xm$XFSm%3Yx@V|Hw)h<5Ac$l4E14o$?BRw`A4}vcAv;YYyQrK5$FU1cBzpAH|74 z!V9Vdb-5{`@TowrA#BRyF#{2R0tRQ+Zym`w4$E`f#0jovQ$SJFV7UF50P`g1 zD(v5Gs#)NJW&|MG?K?}MrCfbo(>8Ugsb>S23*=aOefbxGW38CIttt_E$_%w{Bk&V- z)kN4UpWj!10XConZuGciQ*vhN*Kt9b;*VBHoQ}4hE|dLRlCMCjN25Rglg%-q3P*vE z1gQWFKFBd^+;!d+9)FeYFO+lBU8&Pa6zRZG-o9!xBH%24@@Lfz zysB?xs!6NOLgbtWe?oR8T)Q*EdcAB=z2rw{>HK80)~Ke^%R9w82R^ zBx>C;wQMvrwq`A`52M5v?3rHSob=**Wf;|#^qm(+>gJEqKmUExu&>rd(C68>eMjA3 zcs@2WTx>6SJR~B$WF_XfWS)?U!=<8`OE^`od-H8$9Gr8;uUapc1aYVqdp*pG6t6t3 zGE10@^r$p@o$vOcVcjA7D0J-3yz!CQN3aQ2a@Eb>rxmV_&kpS|PLoC}d45N^(ux|I z_h{oe?PY3wZFGP39CSH@#N6Ez*TffR*bcAtBIkMHbRPS?Jv$j~w5hrf=u8_gNBcFT z1`Aan>a^dt`}nmeOeVxuwoora3_PY@j}^Od>6fm`9~w_6J3u-e9{aT*G@cr5)=>ZE z>U2``@3c7-wsI`OJwQyq#FzR|QiBDa&nq305?SmD@v1YI^_&O0QvXJ(NFh1Z-5R&m z?wA=EcS^ZgD2U{?+u0T|oTjWQ+l5{@ftiRHJ8e+1Dm+=yP@Wdb*x<;t=)|`6!Jo-j zMJsM~L-c(+BCDX`b-g{lab5}jq^a_F|GbJ_SbPsmFKGl&j>p%UT;s$&l+LkmKlZ*g zY5*p^shCBowb=n%=dl)m+z{M$Pogl(VO7{5wmP>d(MJNTvmi@eko@sjIsVhmP-4iARw_gvD_H4T9Q!~ES z7Cc_Asqe@WnRm**733j`Kx9Whq^!!;bDSD5!l^o|P}`WXw(+s@kS@xqsUDM)hVrJg6nrGSsr%=e!}^x>*QpD z&_*QnYjm_4i~78KR(zBkx28qlnJ<*`Q=LxdSn_}>z*3@{WD)|*R hA{+kudHD1O7 ztKPcb_s_kjt6(1bS?jlYNw4l+(?yuFq7(`u0U`hZK#`FaR|NoIz>g4P@UY;I#gAo{#fw#JcQ+S zpBaQmKGap89H}2FQR*_vF>;1&A}93UF9_M)!8N~Q33>OavKgm!&cKtHzxJ*zYL8{` zf!}B4=ksCnLVwy>ajNmAwt=0tC(*azBNC|27|V0{E@YFV=b@(;$?9t3CU75W{jQ3MwIRXmeo|glT-r9Ck{b-pN0}$x4viKF}UkL9SA%fz@1)eN{<@T zRSTLL@Oajw2AfZk@I*xxJffEIL%g)ScDNfL;lGKgYnG;|x#So~xWdIupqUB63feA* zi$A>u1cu5X#pk~LL~d0o{fX|=WE!SFJ0^RIN0B)-zBIWz+o+zL<Hg9vaf|Nto>3lNpN%GkWRg&1t!@duXZHo0MfqD76W&KnwZMH^oMHA@9q1AE6 z7_5-Xa+}SqfFf_gY^jtYpU3eP+xtQnxt8m%Hd@uIKEvkoYnIQ;&uDc?YdT{}MNi_- z;{?@ZKFD!|iUUrPeTH%3Lvw1vBIZ*%4eC}m`aEr>|dirym&Y?tLfp&c8nccVDua*Xx=xgXiOGwQxO@jdLHl%}uu9svaI$ynKO~ zJgK>pwL8bO5k6(~jf*slkdqFrDpknBHlRe!=At@iytqzVKCas7c}Mn%Fz4j1g=s$T zJLF;x&FmYpuvUEI+lGgXwjMDG!cX+xZ#h)vHC;W~#-gbB01`w3h%Z6!|9}wG=B7vZNm3%WrJe2<-XGmu-7%kLq~i zJ;%-oH(NWc+BT!bI0RFpbmVKic4(?U2Yq!jo(%}RvdtyAB!t%3rFei;j4$R2ytkc~ zCU_eFcTY{h0X(^^ytMeVlC}+0ztjgtiXX82)R^VtIi=3XKTX;-OwC>5DMqUShIcPo{$dnW#I z95MEt=@&aLW_2g};>sRzh{s*&LKwi<6$a7rl>Gf-MJ$Cq^C4qAWkOb)O&SJlA*bJw zV<60EVKldU3?Z5yV@Oe_QvizAW1x%S?;?aBC@pAw1uG`K-Bh{UjY@LrBj7)G+1uTZ5hPzjbD{%f_kWC#_(K5^kq z{-T?-iK=yOtnwjLU)pzgls1+0{0U^~%C9h~Dw%GDP1OhDWumZ1I^ld2%{2&&qa0#R zJ+7ql&d4XS*a22?l!@~QGlaXcbf{+&V_)E6vlpnU2TBxoD#pI|j*2_`7!aU-;Zu};(4MX99e(^3p?9ZYacY{u&K9ZxeX#rRYO#!B!z zqdc)J8aWIPMRO+5B`QeN8)O_nZP{FyM*xo!C1zL6kTjj2Kb=b?^S_?vnh?mi5a&G* zjJW1+7fz)I|KL)_xmHC*G1#?ijI5))+p)~$cRgm0y*b8~Ncn<}#R`M|MLttuQ=tmm zEU?Qp?Qmx@Lf%IYl#}U3vyF&^>lsda#dYFTQIyk5wD4}y8{TO9 z>GGJwIO@igjy6pT49+4wtsX#C^Yh|yiz}e>((Uutp>(yd6@_D3d=L&e7(|f6B9Bld z8(HIYST^Y-Pe_qG&?$KjGLeWIn51%Ys>~-M?L(|Gt7GULt8HZYQjbjzNlJr0nr31M zVD{4*t)Z8FIL*O^(Wq~BdZ4m48U==B%L{gl9lT(`Myf2QekTw;;ys?hu^}`ep@Zt? zWWxerk{FxuT2eg-mc0A{)6n|rZ7B@qF$B44n@p-w0hWHRLZ z1PSuujRM5Fbus8dwbRY{IMq6EMTG!|fxS7_e%7ho(Ah{$5Q<@M4FZ2bw4?OlBW0w{ zQb`F=jAP`@coGuyNF5`i!_mf2IaFHfHKGLU~L!);*sz2Zt0W@Vch_x-`msFPmKq8U#l0{VTptBk; zogilq2J%c5KU?q2%)uge7$iYQPaCk zt560Jw5^tNeb(Tw2RoQUk5jj49Z6J@(3B}W)o;D8vog|yIwE(C>(5r&wtaJkoF^7X z9y(pAtR;XQ54qj$ftB#R3IY{dU}|1(ZTlkK5Cw{kKEH?%0%6kYivE0geHOb+u42;pq)%^J@%j=`%XeU+RI%fS?@jvlxNI;;Au=3tCHwL=x|u z-O=0QX{_XZGm-~@X}XLlEf<4eR)E@2srf>SI%2ZL7;+|+ODIgDSRC6-6Xp0I|Km;` zU(#N=so?-Y-f=7Ez)5#Ill2WIQiw3i?!<5&jaRkdX2Cmj^qI8NSOQHnMW(l^1Ex^e z7$}g;z@K?i*au@D8#ix}+(g}AJ-!Gp_ZzA6l;&I3ZiF%;25N>#tZ;oJ#oj%N8I|>8 zm5{*S=)lNtQ5BVkp+F){TH1i;c0PkAKn|*RFoz&Ri+N0fdiO?rkadHPre2yT)hTW- zP__zx(UeIHZ& zlqrDm(o4dkVoiWD4A1k_UFbsyO?zg88l#_Un+xoQR=SL509>-g+CdAAS^9&vu^?oY zv%59*j9>moV*aXW=af1oqwuAWk+S?zO|$lbN54J+NWC$)p)pgN=zy&_+{B zE+wBFs8%uS@PC9X0G?uLTwy}}uhnriKM$z2=t=X z>U?@GR=UNo>_?_rh;q--y5^NTayC6%^;}p=!Av`sfGj5DEgj?r)N%esEOeNO>Zb%A z4cM;ALr)35?+_6p1$bU2-&RW7i@Em=?$`wGg3=*HR@)h*S8iu(q!V`4ej&L#u9$vK z>$*^DZ#S4mBU_05g?yYSQJJ$cjmQPOw#3jg5Gfm|g{Vkujt4>|Q~41V&z`CqWuMYN zZoM=cYiP2Brk_Fe4|JVD7c|j9&4a;>YcE;h0t(fa)S6FFR9y zEy48`-MzovXL*zi5?0EHNF#?F^F$#*halO4DWI)^3MG}=Kn#@|Im)siNE=b`YqKs^L!XCg zKnva_o}l9QKoTa!2pyO0!XLgRwCLGSrx+1c{z!zRDXxsAQpj8<;cyf?l|& zq)h8mT*u(tAX92NU`ZGq=yhIS;%*=H!Kdx2Hz*cWXRx+<*u7UpI}L^;KLVNK7@|Lb z`$qNLbk=B2UsbFR{=zneF0qin3oG@;c5>YB_9Pf04Zd_~I_I27{WCoSerf=py#@#c zX>{B^ciJl7x=Dsv+C^i92(cTg$%Ld#w0))zruSqO(zM3I3%2YOy~_=C$fM2uTK>Mt z9kJ>zLDYIA+SHtD38rKF3m#d-&GnjDu)*>5)U%o#QUf-_( zgTBKXv8{m-e*PBnrr!JHt@Jn7FVk0_6V;!5Z>z)cQ+7~%5VGzT+m^OU&8R zB6${@>3vBcmBB*J7d6U$pc}{`0GY4l@OZdPKUx{-dZzp#+5@qVfb90CEc38S?+fMu zvMUWt@DO%(bgwXI8J+_V7sYg%MJdWj1EPK z6{J}QC{-puZXe(Zvwvmdrd=j`%>6Fppmn;^gBHsA;uyg5lZV15oWjB?BTi=tH#UW% zAfcEuxi@**gk_{`g7zQtATzm@@k&B=SpixM3tZdCQ79@@e&;SIU4uql5yGEv2q}Uz zS2&*^qinz1Uc8F5lsxDYPeC%ubD?0=9|wD5{zz2Z?SFP~a%Kq&56CLpoS zVdi=)Ov|3LIye(aR^eh&y7;b}(eH!&JylDv@3#P1w02{_$anO_)!f?peP@Zonf4Ul zo>FRq?#pzIQ+My&WVUzO9`45jDSY zpupqW+w)CQRb_{M26Kjmu&8s8H%HNdwBoW=JGwyK?UpfacR6fV_vxy{Q+siPd0&B* z?&zQk>Y*4H5*iS%_4+xQC(tziQ5TY~H@9TTt33a#2w*PuO1=USfZ)xY(vmOr`4Gj= z3GEejL|(R;T^JP=T!jj<6cbaH5fdX-R#lRek(U6Mn=-uP1f=_g2?sx_6~1FcbwPI+ zndS|JQNfU^6vcR#@F7_iiAdSs$#b+1L1@%Rg2>;SIh7l2U_%5>N6<=AnH@Tns$*tt9}x?7 ztqr;*-L?uzNsp!@pN9?9%X2Yop==!}$nA9~5t>oSkO|Tdv+V~A#ONjK`)uohv^P?L z6!BR!s8|t$s7ZV&iO7j=;oNy81UpDN$=H*+8FR8xji;j@@y4y_$MlMMBB60=1K)^` zl;lPtG7MPENwdTiJ-3ATj$9OvP{o%cOGd+}=ZZQ$ZTOzI>((>#Y*NE)3*ScA_=@}Y zyq~UEzPYWn;Zg#;ve_3lKF`52U&?rNR!T(sz1VDQwN;LK_UxcC-xbIO|CIJg6U_y5 zWY2Kz`1$MR*iD>NHpYB|BNfx9vh({aW>_Ho=~D&b;k3!?ZZgo*KA{Pxa{)_k}kuZf+pr3vq!k34K_+}zASMs9W<6GnC}b{0l%P7@wRHZyK^ zPHqkkQ!^Hpe?ZCDI)Q+;#-_iaz~JvK!8k@ntjs(nEGCRboXl*D>}JMnj7C6i9!4W$ zR#p}^BcKrr7w119lpHO=Q3V=yu& z53?Bu(2SYYBP`n@H2t(Fq)hzcE>u-?ROR z``azNU}eCD1^$jxFu)%iU|o2{98G~BJ4bapJ8J>*-$Nq(&G|=olk)%R6lqH*FopZ? zi2qm2tC>3d`PZLQz}oVUSEQtWge@=7_)j5DKo`?LZ2;!`^Odm$(AL}(Jiq@HQ2*|? z{9iZ=Y+P0#J15w|96&BccAyD2BhbjqjM3D@#FU$b8OUK~2K+;fztElR%s{R{M^h1V zu%=*bzyb7!Hl(zFn3C>4sa-8hf7^)}Jhz#-7@67CSvYvvn0Yz4=$W~BnVHF%{?Dsu z#%jdHYRb*Z$PJE4Ms{vCQ}7cH7bDP&i=CO%jGG4>@BcaI|2HbKfK_DSWn!k?Fvu`Lhq4LBR=`>936ZFPr_2l>dY6U*YzDPy!hGze4^; z_WiH9{%fxPkp=!o#Q%-1|C;N6WP$$?@qeT1f19}w|31^1+JZNFuHX|}`FN@%__T*$ zFRkMQ0N_yl{(+EDr920B!hvKIB;a=8P@zb8wWwn)!CkOMa#G^pPjI^zYxe~}*i6WX zi>SLT9k05&scUQvUD%n;O7C|KRvDcMffyJ0G(Y@$d1-MSF=0hOKs5Vud&$2CvAfpl+Hy-tb2~w2Xl7=nYs_!Y zXrCDYZgPcVUAB_rf3%9SM~K+ponxBG3|v%TLU7Ye^hWc7J%&VfaR9ecWqiaAS#-c| z8yySV?^a)S?70Dz6U`KqsuLvPSL^_=_>+&W2O=j`XNL#C{wH0+CSZEJ;uj$xE%kTP z`DzcQ)O1G$O#C66!IT?JSu*YpEGXeW*xBUpe^b7j0)zh13Xy+PChmbj z|7fbgV9NThPJfB-|KBb?UXH#PAl?rkeQB(VuI1PZO|-TsA`9Fq~;$4jNn8xhm*W7?zS@KF(c z54g)rZx`wLhmM5vTP}wWulu2MO*~eP8vTVJg#xSj{uNhj<)VauBpi~mV+_HKk4oq* zgQAIT(49Hr%*d<+w^_8)#CQ1GXdo(g0N#286V6>jD4$Xw>4=J3#)HFf7$$P(8xcdS zw{7@H2x!#%jJv>|?vz>Cm)40R1Ij8vP-wZ+0kAXN<_NEu5%Czk`9xF}&hKJxgmRxw zTO6Hp7>#NQ@Iq*tPT(2|Kh4Ph!Y>Ous{$^7GSPhhR@c?k1VT|I69)Yillfoa`~n(k z#J3D00!?*cVJv&@&LLiTWrGRg>~%4EV@IuJVMB@`VR|Gtm=Z{aOwS`XNy zON|<6vGpzNsjeKo6I6e~E&QA^zymu*kVOa_yCj`QmxYoLiE0vqPhkK=Qm(W!_1k>g zKJ|p=M!>_rT^mXu;-^6Z7%%7bNNhGi$_9Ih^&1NjXzgu7_4Ol&ezoX(rTKDxHsJB* zW3$mEf@GPS{@)Yn_n_%X+i(aF5jlH5OV}?^r&QV7cGwU?MpL*0QecrO;bmULn4wv5 zY}p~Q$oCPh32TxJEo^mw@=vi-y9vm003YOL#sU)nH<{Nnehx8}*bsYw`x<5pHiD29 zeZX2UPXe(l!J~W>f=40;^***Hz3nT3sAUc^tsbI;UC&-)E@>Ug=bm4XJ&U4DwKnZR z*JPsT!rpiBKmnTx|J1p3vsJPVNfAX z$>GOh>tXJ0+FOLgbLvA>8PpkyHie;pXmB(LO+{ zL;JoGe!{h023B1_7<$!Ukijn^BJ7$*ocVKQ83Ln00WbpL@Hn@$Ry*?p0A`mglZxP7Lier&Ux)zd1IThB5Q4pkWzpdU zaIz@cf~=CPZQI@yuxn!kkv#TIt@ptNDZAM^09q78djD&U?-Ui zzv2e^xppD_Sgye}WB_y|eIb*Q+MwUy3-`Po{Q3w@Um}U%0pWu|oR2@+-+BkAF<5aJ zf_z2&m@RfdN+UY$c|;QqmAaEaC`m~2LL_`qHh>R(i6;o>B|dG%LNr1F@AE+UM@fq~d#5LSphXn&YgQB;n3(;L{IK8LzK=*;{92=n2r%B`LthL_Km}-- zi)<>?w83X-C~G3jBh0(5OVTJH1B6gsy};tr;o0R&oW%MbC0CAC;c;@^lSngU3A>1q z020jKT}19~r)Y^?^b9iPpgv$CvqgK)SNj_{yoQ)T1Apm2=dJ^Rt!Z09&UrmP33}(0 zF43SbaDKHgO17v+XCev2j$Ul`boNL!Y|BN-HGeO|%lrUj+#?G*?TiA#f)@9uK5ug* z9Na#Ux0~%In;$E1H(rH<>{Y>axJ0L4?aQ91c)cHfra$j&c#^B zM{n#2?2+w3+DVX7fn+k<_StH*tM%`iHl*Dl3>jh%30+^-nE@*u9%sHHXMStt;`e~? zt`HLJtEY${-*V_zEoceIPLiZ>;>+iAr*RY8rwH05c~XDYVliHpUgNLPmZJ<+XszcxL|tLAmP$tpNg`X2>OcFjjN;qo;MZ zT;J_rHL;!n0PoS$$cQr`R03Q5sN?@F670( z_Ck>-*nlF0*)duw&0i><$Aw6&h`@VYnP7m>+oUt#iIPZ5vdTP5f49_^%w|XuU;a@o z2t^!NLb2($@6rm!(PEq^3OyI7bYZv=GqAc`ghXO0v{Nw1QjF~x3Oywx5P~0V3*4|5 zD$eV}a$JE%00famHtrR3{K61F+fq2g;}tlpaS2z*UUG`qFa3UrIO7m^`V@;-octs= z&`+v0mTN8$5=$Vk2sq}&;bqki1! zhPJ&X^isouau^~?ee5~q84*=Q}Kq9YWyo-68<>*TW3&96DR7 zGcBS0T|jR&G5-#nvMX{@A(8T^bCoS}(kWPDx(LM0VCJV~P5QbaIKLf~<#EV0Quywz z&q>^|x9CRXCMl?Vn5LuM`>4qx^?lbP;8L{>sRT}s~piJq#_5$)g&mW#>(Qgog0(p!SH zN6!p{oA!0vdNSy{nQ_Wetp44B&K8C%kNMF4AFP3?WC_SSRrzj>n*+Q)9(Sp(Xty@< z%&wbK(4EHL=umG60LdZX6_L-`K`%2xl`t}S4(fd9{HzgPQ^Ek}wpFnZgf{ds{qMA+ z!2^)jg_uEMwukDtC418e8DuBA$yHN|ev1B`@$Q=DE{0^DkIJj3+kXp~5GD{ei_jM# zqgHO}iKU#YOD}9Us#}O2|+!VTnP6f#?j-E9TOf@_OcgH$djwpp7 z`sJ3+Ad>PXOFm%Dzt(;0PU5z5ATuRrHGDt3uOl3O94M-Os_PIQXbVm;FdoCS3XIk%R5AM8H; zh$5lj6a5&ewoC)9*R)-VN%rEW5+evZxu-DKCfR|3e=A-TOz9u{;#EMvi zX~1Un^7}swMFJR_*i25Uu+E2jOzL@o?o2c?6%;AKA!+zd1219pI8JfzWnK?2TC3yW zdnSTDJ09lg?c&y9EqilE@YXgP{}hV~{?r5GYJ(TUEzA(&i!r<=og$973k^KH9rb?Z zT*PcAjweQ7sutFjUxn24@;h+$Yu7Xn%%;!;@3R%j!CMPaDxCAX!ddv>2~PMUjpy?- zg?PQoA%{nHx@`h3a`?*a7!8t=^mC8rbHY@IYzoc_fu>G}6?dn;=QPZ567EpL6303i zE~8gl;`#Cc35Bh0MS{S`hTLNfKjV)|#UYQwx2l=)=%K{8J~lQ-9)F*gB@G2yLcnlH4?5z>v|w zQ^&ceVq5$xH!Q+6#SfmX=E3!|23A3_r>>n7+-{{Oa?%Cr<9G>XN}<}ddGeI*Uwi?= z@9T(86aSjd2&e!xEu)m?%3c15kIzretifl7vpPH1DN2NbIPR1LU0fnM1vElgm4?0v#uP<52kw$0}jpM-ynOkDK5Lpl6M zs`@<~OX%{kCQ9wqMv1%I;4b#2s3ZM4$mL@5j}reMDv10))?*`)>@>a@n+W z^goM1|3CV@T%|7j^BWr;e;b`t6X`RDRm(`t}JWCu{Nx+Jx_=+Ckk zxUhy@_kdYPycm*x$g*WlFTDkc=aWcHi z*?$rOHp{PLGPhH5u{Yu?wrMtJOnunm3z<`y&LwKn-B=`2gO|55;=}^U7;r!9vs7@q zCIyzLUOJD6^ScUwM%mygj;y;g>6!G67I$psr4(wlC_~)d~af9y)FqkeW#1*|@TXKzJP}Da+5Du#- zc`vHC6XwCgg?G|~#x*tul!Qjdpg~VqFMIb&`Mh8)uoo+mi7C3n(h2!2Q|vK2s_=#H zHy(?#`pcR^j_P({J(6Oxh<(cTa{^A>smD} zE8bw;;wH0~{+v?4=S_vl+wL^_xoIS<7fy27gPhw%yxyY6_(8c;jdOsDFbS4JpHBZhvVG9jHO%`lB2g!iR|B%jus0ZyN$<6?y(q)pEyb`zb#Wvz%qj!rrc3y$Ra;hj z7NFC}4BLXakJLb)-vIYv+u-e7{_|aQ)nVuWs#RcH@?uj!2@b2S_r2t}o4nq1W8ZLR z2HmIe=dGsPux__m{c%ltK`gk=eZX|ZP&gzUR5GN^S4*cQ2!-=o#B$MAzsVg471wy~ z`>(3Y#KrYP_L&xKnuHz7C8A_GS^4SI=V%V6f_XpFAU&dp4@a$xMtH#|kL9C?X9LF< zcXayE^=I$*&o02@YqaE<{g=)b>TU>$EAAb+=ohzlJ#E;a3_{a)YE?u*0VF=YA^>h1 zBvaPu@2;i6^ApwhRlp7=$qL-(cvSoam=(I&h=IPxpge4DGXebpiLFq+L8;1kb|3t- z0n4qL^ubF7FWm96pqU@i2@g5X2&-%Hxg-yPSHkrd` z;%2!iSA4ZiKp5iaI`+_S1IOKAaOu?EZoC=TTYj(LuN2B_!Bq#MO{ejOKo^>>i35^T zyPM_abaDlACs%YpmXGtS_f6sF;Pb@A6S<4W3vAv~G6sy!j~LIol&df3QejzV;7apy zw4iUO;_%HLqDK}8iH6~bw*rX~iJH|65T(SI1NZj8sh1$hj-Yy~)O+U5(9BE9fmf+! z?X8_D;Lg9gBXQZ#Yr_6y0M2NjjTNdVk&tng(xa39yHmulSMdCY@l_GDZo6fw)jdM9 zaGBz;g^rvkooi+SZm5zYITf*`7)BUBa(&l&VU)O(mvC*K0!TC|#s(o@?1*d}SDNRY zf|p_BTEjvp;d%mkLl*N3eQI>pyuQPc^=|%?W ztT&c)Y5d?~pEcj(i+frIN5T)4rAbR^5Yq?wMeh~LV4|5|l<}u$EB^a$(_AsYMd;ZW zqYa5~ewww8_p~?)l~a6#UXMaQ&I%wve6=tzJ3eBe*8q0<_8E8&KoSYy<#Rp^X6HQo z_3{==c5$KNPurOrB!~S06O?2@Grtsev5p#k)z{VHk`JDPFLnuP5U%93BaKSW#%|J$iLU2l?9Ps|(u+S+CSHzfC^ zkYsV=usCc400qG$M)q3x`9ntg{G*RpG`^`p;^Vk$2dGkzbqhzHd@U9+rL zxOu_3^GRd8Z%v#uD%EQjJ*tM5nmAHtOTpOpDqY6Y9nBU$?g?S_+hPN%zaCk}A$jdv z+3w7bFhcubTDGONONbjacWarvM+}oEniAcqQ;GCZio~NxYs7kf8=--}DX3$lq9_sk zCS6C?9-#egDP{t~@U{-Jhp=*kA+AYg{NeMRPsFz<=>kjMkW#9*gP~IcS@{7s#uj2K zvEb7Mc7I1eht33efmE$ut&j$^Bt$PIDh&pn4T@mB1Fc-9R3SW(rd8rb?1gtk=b5vy z_&ROS*tc6Yt#&OUh3c(;KK&U<($h*}cbZmSY;kvty&@J9l+FCw$piQxE}1;^-M?iv zg_jH2up8)nADORN^>Z9sAMeu8*4n3s6}$~*T$pf#6xNq!R517RxQyF>OmVZ&4vO&< zJYi3gOAv_*eG?W;5?a(826U3BHrNV%NPNhMiDg8+4evG6NQUZ3oA*h;0+jhJrH$L4 z3=JDiBRa?qF2}g;SCHuuHIV!W%3FI4d+>b|j~mV@FmRtp&{qIiC*ga`$yF{$2>9@+ z2|Q8v?3+u}XeV2!fx7DEvx*Vg?;gD9%!nAfN|Y^%l}6{}6}4PB!$D!ZBw@j>5k69( zu$=>gi5|hK5C>@rY-CJJW=$NL*EHMFabAk`2OKx8j*S~7#~X{ z+=p9SN!eV>#%MCGt~xsmUb=GiHYh%N)8D@JVusrwm@$tZIic7mNEo*z7+E_UnloBd zy;$f?6ugN8%>$yRm9w-str8_wn3D~kosAwYwvG`wBr!2Ehf9r?P`ZrGG?M->hBYc+ z*_T`~v!(TUq%e7iP3nY4>WW8{@LTola90#D!GNms2GI1fkeoIpJnv<}=px8Sq92jS zoI{53j3z?hrelo^xw}r-{`X~BbMXDIjpN|&bS;d zXIySR9o^(lo(-}5MDwf?&bklGtIas8G@ASTyQ6(!o7L5KnM@4hxWlPK)5;64#(6=7rh)k z9~^v%ybbM(T^!%~La84rtM}^i50l+ttmSotfcb_SV<*7ksZ{L-b0OOYX33g%3w~Ak zvv-=-*X(E>fuM=p7y@R4sMNwWy~-PF^dFdet2Ko;!sH5CGjgl$Qu6N7wCk=D6(7Lg zFqi(YwM7Mdix>!$f@!DAWRW#@jpLI9oqYVEqKYsaV}U?4moW3yq|BpUUjLqU->>$z z9$N=m{B~L%?%aeq>KoR!*TQ(yuL$pJ$A`)NF_F~DMW;PQhLY7lu|?pEPsX$Jnd4fr z!;$Di^8`>2=NLksGL0k7V%{xyojW}J>od6v*e+G+=} zvtqPj{LRj7?dVs~AUIijlz2BUL@@OrDi~tXIyw)4S4P^0C!Dy|2p4Lj7qJ zWmuTZz|WM}S;9e#MNbJTw{RDeK{;>uNm<$2!s($Lbh{sx0i z2v|ybnIo=y?`h7oNLv%`w3}&C@c=3;Po#FBdP<^A5c6d42@oEZvU)0RXEseEb{hXG#R(}?KBd9*3<_M00`>Q{_~D0()4R}h5JcRJBV5El(h{8Cp( z&)-g~Jb-XxaQ~FD=)$Z{*!tOxQ7tcmphi5Bp}x79*i?waP+#Z7Up%cUH?zbu<<6Yq z`E#fEH`hUTB=3dsG8j0&q_h*W3qn6^G{97@|1Uhfw$+VaOAc3I}X ziRs^c$ne%meS8Xf_`1h@u?~?YiR8SCRXY~eWo1?hz6@d0s`4WoU=QAN#7z$we6u9;X4l1` zCtB1}=GHxBkbFfcwF6A z5q&KBO_~1CrGZ`Jw%o-uzD#p9Y1!ouIYJ&aT7}1HX)HH6WSvu(hH=#pDB)$t{0qAF zPoyNFFPP=NiTBr6ghY+8a05L@gBRX%Hm@m!D7zqsQ?adCNtw69ri9egnKZ2Jh`g5g zo@Mx$#j8hu1<zj~n2F3j}6VG5Xr|iW&A<acE+uN$vlg#ek;_lYnR^t6 zL+;#2)jj@rHSvY!0Z+y+K1Ejh)E^A!Z%7ImcuTp&*t56>#ytgI+jI`WsmPQI2j)sG zAuO#2=Ib8KOouYEpq_YfIG@X%#IR4?q#838|(vP==`~m)~T*QWR|e zW$^+>=_l|y)j`d4aXd!d+hv^n~b6Ef7HVr=7vMzm8NjTUQ| zSg^osmFcr5x1}$~9DruM4BoJvuFrn$HA}+$9oUWOhE#+IU#JGY%vyR8eJW!#pL%RQFR+~_b-pdQ ze4xa2P;A>9ZbDV;{dq3a=V#a`J@Hs#q`QVRhU!dSywfI%v z{rpz?qzk$?D2b}B(@ES0O#6WU=Y(%CDT*+R`1g3TfO#YPELz%2GMNQdsSHq{a>2k+ zwwf-keN@U#hFfIS#>C-q; zjbBppp4+#=Tf6$qO384wx5o_WX6zY(IW8Nt9NQ8Oi-mcl0#z7aXR5~MX$`PZxWc`e z-oSkJnj0^B1Lt?M+(44dEPc%_P2{G<&>~slv(m(%VD#~1s9U0%WH1G3aFck_BxwnJ z{o?tHi&}%0G3vB-7LM@?{pneU`!Yqr5x2`iv|b_ndU$Vh1PmXN>HXJK4IA)$bSB8V z|E?wWE5LwQOn1=3H!wsEMlz_`|+pZ5XTVhcgE zA%vnwS-(KMeRXMkTGxkh0R)jIP&}J z1skT{#ZGHadq7G6vB*Jku4Sr)4lgfBLq=$ln%rJ&738$#!M-*52T3NPF`~mTS*E)N zs^nZ2l$kU9-0hO`+KK67c;28?-gAET^cmg-*Z%Rce7z0tkFd*G(3SzYZ#NU7>=nrv zup7=3S;}v~t8Vp7@YN6aV!Oag#7$`tsE^oeavK0?L$5}&i==h z9KoD6U?~@ob7mBAUhC-|DWhQ`>pPEV-&~;cG}cNkex$k$VuB#l++h#_VQ3PUlpULz zmPWlx{i-Cli$F#vh9Y)kXtwO=_$O32zoyQ&Ig#1G&D-;*_L@zb#dC~`M;uMWj$e5? z*Sp}*lzGI!7=J*_dI2D=`=gNbv|Zb;D=`nCv6nhIUsn_1`-H4$%2EL?QzmBPVUY$M~pPifA`CuD3v)(60(jWZ# z`TDGP%4&0!B;aF?XcsrXDG6nvHqR~b&v^tsFL75{m2brynya9^-9%p$pfe6pduB_aRy^B)t))I*~ozSFL&xhYR|X_y)%lL8kpNkoJ50 zC2z!!{K@C%IP}#4T>;BJ!CoI3>pzWxm*f+%@$;C?AGlvFgbX0Pk;da;p%lkVDm`0CFOzJ{Cj-^$TrPR0 zlh0FgihbS9G@X7zp53!9x>y9xa`gr%0sG5Xi8l=33Gy1>=EPK|=rlIG1)GT8I$)GU zSVae1hMZQak(pl78XE`4>(&&C1H=Y`=`BDnRIw~#AnbwdRiz%0kjsg|BPE+d#eCI zP4#cV;EF?2(esD_99(rU$(WUpV9*)N_G*0ZAA;9!<)Ho0D3JiK5iTjw#QDWtJAf1I z86JS02;73y+fjA0s4LDBX8{V1WU4aFZs8Ab;48BvLRyOLtpp7ileh)#^Qm*lznw48 z!@&@EB*;_@Hyi=f0^NekKuzE|GT%TFr%Qz673}fB?lU+DLz)Ib zWWe%36L!uxWP=p$72dr^L_Ir3sVc-g&wfQw2|DGhx_BDrms}FOT&Y$CvGRV775H`^ zh_i$RsPp=qy(dwTfyT)SGw@2k{41AH&Aw(srPoTrek!@9sUmpad9j7K3XA<>j;%Zi zoIgyLVqp=fX?+{vUv4=n+-oC0l6Rs$5F1<(f&DD{Z%7g(MghVN zx))`O=w_^ehzZ5<6OUIi9%L6nkk#hs^>n}1bj6Y*zyjMwY6+nVTaUp4wFZ^i8#SAz ziKH54b6x86+1$*55Kshqi;fF;KTpw7UZv}h(6CiEcxg%!P=qXg7Mm*?e56=_*{$3L zVM-Eqa}@TXkmO?${eJ=EKpVe~L&X8)A5hDA=(;Slnhv_38#IMgrRNyjo_>UWYvJ+N z6MlL29Ow)9L9nx@%`mH3901xceMpOhr9r}L&671K;kI4Y!|E?!-6yc=8~FN1_{Vza z?JH_ltvLX+#5N%a%1{nM!{CzEa{Sg#>2|uid{}YUK_-9u0H~_S* z`GG$wNDn7l|D)hpL5E5nqqbmA!2-b;LAI}~S+(Z?(Bj*XY{A}w2LvZoboj2IiC~6c ze%O9^^77x}9zI2@#}DWFnH^avL%D~%Yg%VaduB$*`+!6?beyg_UgfUGI2TQ=7Yr4? zDpsQCdI~g)0!4(nF-|v@^-G0EQ2O=gTzcwb?pLYs{~*-937Wy4nlS*q%6>WqDL1x# z5*p1h+5@R>6>D>Ag>B}Of&&DZu4QI5Q!ennFN0O$fo+7~a={Z}_&3vYE^VC^Gl(+s z5kI(EO)?q>fHl1x$q;l2r-yi0!Mv{n-3uZ0tbdH|ckRa2La2vyF)m&Ml;tivzp5Rpq14~ zq$REhB^Vt(tnC{=RW>Ziy zP?U(wFW#?KamlL%`PnQ|#_CIg6@vW)b$wM**H;J7W9d-lFyrf9CGPVJbh`Io?%6q$d%5tY6037cq`h}IXS4e| z)8psH(oLLvU@zVzYsm#j*$M4f+Bc6g^{m0X*@FBlIp)WHrRtW*dFD2j&1%ZSOgna7 z;@n*T8=qmsb(48w*Lpm}sSl5f4FCsI&)fsWTu=4I$khUtD(G9(nrybw)hxTXM1ME&}q*O;j# zNI60+d!t#*Rre~|1`iQj<^a&FW^({&iH$>+AUEuh(ILK9hP#XZdY+m=!{t5sit$Ij z#{=4>%=>W<&90fui5;79YqwUsS&PUf+Kl{y=Iw6>E(R`+M!B?IWB!^>g9q4u<>?f> zP=9kz4AK0=Hk_N4#U-QK@;*eqBd6QfbnJZ>O}m`WF<*`OWH|>l&gRNat$8p%oek9M z{}Yq`a&r6wltlFL`Hb8(kDK!9@idvyuc_ba-?W~3BgZ~JU~5)6D}YE<{0*sXQw{)Y zR$EeAP$%xSh@g?LPWABa7`pC3`beHv+Ym+_PGeAiP-6j+P{z?4&fvInEON27^tk(jCfaIJ>XOt~!jWs7PE zb#8|~qK#hWRL!87av7&AP;>^WwE_*wNB`nx%`Npk4>ezc_Te^-1z3X`9*)jqoZjMU z4SWy{rmtr6Y1NtkFQHu?A-KqM=9<-P4gf9opjI~=mbFyy*Kn9`R}ahD2G?6e-&@6! zAvqM$c=k^$c2E zy8`~6M!M*G!5XCI3CB-+@d0+(BZKD$wP6GOkN=wU|F>6?&yn`-o=nc7)9;w~$vpezv)1O$( z3Y9qwxbRj^q2ZtR*GF`@VH9r;&Eh^z-89j6J*C6~z_WZQQpY1G{cejWg8um$g&sNK|(*O5Ho7vm@vN8f#7zP3VMHq)`+Q2T zqx$oFHaQzP<&(Quk???Y9)HUqTEvjc-bHGdaQw(a_b{v<*=L=6G*>qyX9IJW{gV~z zx+I=w8nrTU$eULDjm;0&b`)C ziUUAP(q1W>vD(}=Ek~_8qDi2+^wMOmQkmgcFVvpanjil??Gx1G;Gh@a4=%xNu zxUH@uXzYcv%xX3VfHmtOEjt|E8$YbKqhL-rFNA-v-45Qko%~tNxSDf!J#?2S8`%B0 zHQcJ*%e{9Z>&B@ZTy6j;YF|2pdzxn;Ca?T3aemJ`Z8Z?nrlRj;sCIEwk_?$d4*08 z?CdL2!|THq0>Gs2i3Nb^voD~ap#I2wvgeHUesC6kn49#?EM$E zVL8v-#ymXSQ|{|uUXZ7s8Mz1T zPLBOz?x#mM>fqkI*__^2uvQ&?hf{_FK>HQjIb~gpCNWhr61UAqBgdex*h4RIu6p=t znORkq3#{>FsMZf=nS!R_e13Zi#tJfgWqzakv5Jfj@8?MB9d+Kc0`hJd#7{KdYXf^d zwzS;$RU`oXo~FwN@b`wKZDheyuP{DVQ0U#NJ4ng-e{`X1D-mf9+>3sb69RK)? z_=ly<0pNa0+5=DpL=I|018C8hf|Yd>=W6=&L300d5_i;M*YEjw(=;HBf&$M6STi^P zOzrk0P0%Fl?HB)xq}GClzCs4SEC8%c=Rb65&4BrR7c-TgM{(c*y*RfIb&h7v5q%hQQ*L>mW#dJh zLBD?=u=&_Ok$3kvULRVE`#9vM8NS0Q!vUcEDpsPpd!v54qO(}4MeL_K@!LN^P4+}* zbAoQ+%hEre8+^Wm+B3+jwKa2wo7u0+0buogP-`IQB$yz$QE-mnKtWeuo!`!yrPO+A z1Xob-&FNZ6-B-@#kt`Z*WZap5EN_DGEd_uQk=K@U;GP5cmR1LT#${iPO$^j@9)HUu zDu!PEe)$2Q8tZwQ+;ITQ8xUE~!I!_u^5WsXWd(o|X}9zy?~r^(4ku@FLu${=WyHSB zxUoMgl?Cw7EL#3~Urrd^f^|f9DdM=RPRxM*WYp`&Wk0oc4XO4V0H$VpQb*8Mu#4bQ z!8i{gvNG$-F9-lL7e{shnZE#Y2FnAGbFO z+MJm0KBUn-`*F;vd+-$biwdGs$KI%jy z{}MGdH~_R?q5V_j@2GbUS`xX)CkcKDgptBKKVnM zlfPpN+V0wdwj-O9-?s&AN9L2?zX5HAwV=(eEoeKWIr)9-)0PfnnxYxe=au~@t^9gf zbhz{|X6>BE3Oek1J|{mhE$I-c^Y~jP5i#`AEe3!9pPWGV8<+42k=E;(c=?HZuO^T@ z0F+33dl|z{+?9Va<+wj_?o&HnbP+9XS;9H>Xz?%3_}@ZSDEja|jH?x8=5d2rOGY0t z=YR{NFpFs8Ssc1{h-*o;=KwIZ+LES%UV{1IFOrrB#tUls>R6feW@`h$Uc5!tk`i+O zN~C>vCp%uWGp}be;4D_I>rh4;x#dMbKBxBmGdS;n9=yu#x1Pw7&uf?3=OZRgCih&% zcNs9M4c$kzq5Fhx^n9d%iw^G1n{DWPHLs|f?{Z3T0B9+r@J>#c){6Vvka=3yJm!pS z&CH3Nm~-)L{nXBSuniL}yk~chuqB{$|V9mowy$9?UtS zA7>pgfY*DIeS3Yf7mRAn%(1PRHNGA5mbKvb1++bn9(xR8?k%MjDtY=qChpOJnODu_ zzI}V~Mm}w?@7{*J#*1_Z`PSd(-(w z#$V;d|IB`4xxh!h3@QD9)Bcs7L@Pk$&H?Ogpv_>uGh$Iu2dACsH(N0QeSNmyKox-Ou@y<^am}bv8~~wiHOu1$}uk4u3{mfW1$9Fj;H~_Si@%@!t(U?)oIrA?aV#8!5B0v0@qn32# z)d&O5LTb7B`_*eS|MN^1^`G@zOCJ zd6tnU9Kmt_Ncyu_;R;SYs26Y1{P^<{dpt{Iz48b4K63!CXVUI$j`_ULmi})a9?LG* z4Cna?EqIcH&Yy#n<2!El+i?Jxnh$1;1ib|V1)rue z$M3&9fpA#SJFBSo#P7JKHT4UbvFuLPsZ$vbH(57M<=_Dse92*t-Or09hXB8O1KD@Y zXTM$B@C?1r`ws=L3`jl%I3k8#wx%KhAb>ul>CO9aTB9^Ju=JYuIs4mw$pXMv=Ca>@ z^;yet&p*rqC6E2yQQVeCkFWU0x}~wt^x!N$rOx;wzSLBIMb{_}0IO!3QcKWM!OsM- z*K)%}qH0&MPn!noyTQ-xVC#lj**3mIzj6S$`Vz99S;CWhwB~8LUU*1(i-2q?0Q~+M zW}dqfFYZ9jUEKNbj`8zRNz}Qx8|U)ZA9^J@e%cq;a>tymyh@{aCn0%MMb+m3(8{r4 z2iVZBs(hLc3%HV_KA4^Ov-uByW$7VZd6h{2(~+9TZ(sNS?45U9RrUYJpZBt5E4WZV zz=eniG8`!Gy~RCq?~yZ2$|s-<8dAjIQQJM?m3_L=kxr$$5cWdC`C7BD}Cbq`^m(uZq6Q0T0YFsy%n?A`u=Lq ziVgKh=@`fk#_Z@yO7`|yo0&GUI$u$G;(S}}x3=PbWHxVftHeJ12aMr#=A4LGuEFd1 z!StG1n{D^j;|Ly89$@@ABTu3sLw8W=08oPnmjf|!AkJD5Hcq(22x?iEXhDqdt09iK zUsq@FBYfpglpxl!52CN_j~F=+D;zBA9f+}R7c0uw64x3YZNORF<6cyMENdZ;IBPz? zrGj>X#)8rX@M=oH>0$84UvRD)WX{)981>7S)QqOYdkkDO%^B(e^?bZYm@<$@YC3a; zdJq22w96fGFre1}V9r&5?RNBFeT<1q#60;uU;dgoPhcJZ@O>Kq5a91ARNF9uBPL2- zq~D6^YnzZzH3I;Ya9N$etYDnaGGP5%#5RcossdGUbg%M!XO{cpe~1kirWydS$SWNH z>SeqXN=-}qpKf#W1SJc=S=WWrU&G(WKmo=pbr=3Zw*Ua{8qFu2EAl0tiBpl> zB}^$v*hewM!TQeU9Re&&`zA7GS}cbHXu6)2?*(O?D7$=|9)S!W!uFFk-}ei9>5vq` z5r0apVXq>LU;3_>0RRo<)Z$cxzmLLSf5XMQ;$7a^k@>u z`~tv%YRQyW64=r|oUf?7_I{F1IA^;%c>%x^IG`$q- zk2Ha8U5;&v2=Bq5kcXF#q-A0U*DoYx3U9hW04QPF zI)v6UD{!8s_g`#``CdFmc4u@t0Q6kZfk%RIJxjm&TL=kD0HT3t92}zEv9F7BR6W-H zn?6|23IHX{SO;>33bn2Qz-{;_f&@j=O<|-6=PP3Ha^c<5YIiL5pcV^-%S)m~N0FQ_ zZ#136MH`tTJbn>XW(x1t!oHb;>7ZV06Z|Z?zaj$L%bsTPrnN$uDIb0j>mL-SiUH_# z0#t_wMT;5Y>r#cpjyMY<10eL-&C<-T zX0;)1JD%X7t|fS%uG9Ku1%MK!KNitwMGxL>Tb|G8vZ50y8FCHo!eRe#HWL2i2$F19 zsniL)KJ3T7ID58H@2ft>0QRnz0RRozdXtO6g{^StGVH$sXZJzQ0PA*XTR4L_qw-n% z=D7mEKPjF(l-f9bk)8@F4*)n=t(o#-177ME#$jr|Fzto_U=JIm)DjdMC>v(>b zS5DWpy)6yADA%CEV+6dO#M!WBJ|bqL7#@{zpL(A}RpwVw|GY?Y%-Gq}12mM0-NG`}pFhQFADq(|u;48&sG5y!@&y33jlS)JIBg z*LjZ;?4F3sn8P)JP$D9M}Ihyx?JSzbF^(21fXwxcd&*X3mO*4~c7_ zK|kWF`aCEecwD3w2hf|eQVc(e%IC#_0-yL_V6x_iT`K4ys8euq#i|tiI}MIqw%Cnb zf*&uzxy~SGqk537&Hzu$+ag770O{tYZ8N&;$753Nh9#|N<`u^gWhP<2vCKWaA)uq-WusXMwpotAXxsUU(LBd~u9l1_L5 z=y$vp*ClRMKgLh1&psUMKUk8IwRtwP9sFi0H504RXk%w~_YC83NlGuEltWwzYFFWe z@%-JTG60|<8%moE-(I?AM~+>F%Nya&R`*^Jid$CLmL~udcFE&$2_-pw44nr69GcdF zXBO4rQqdO&z2O0nNH zik`EhIfC=#c?>hcfDIYCQ_2niPcT6CTo#SrxA`)Egs-OaitxB1x{lA+V5~WPL{gWG zgDn81&_z}GRm7i_#ZYHz^fZ*W#HD5cO2kgpXedJ4iEaLc=|>RQ3u;YDN0B{-G_zl;TyEQ~MOxgFFgOS^TPQ8`VF4lS)=6C6n?(HlL}p^=+pV!T@iF+-U}{hA z*Lo7wtSoh6yos;wMg7X|)U8#Tm`3F+OOrt263S4ok^?bR=^n_@YZFcS4bLU<#5DG` z96#Kf_%e9b;z^arczs__p+>U+>eRJfSExJnf{GFCPQ6M70K98u06;?-?T8sJ)q$Un zWws-`1gFl!#Xff$0@~=;=L!IA=_&v$P2J_xom{J9FeMX1DLt?g0e_TW-e*yK;YGzU zv|1WLs_k5+R{@}zx{Kng64^GyvJ8%<&glAd8tlpZVpM2Lt(BE+0mStEtRP}IUL6Rt zOrYuo>O41qoi!-&BK)XqsVv>a;D~G67&_+wN_g!gbAi)f{Rgb`c!7ayGw8+4gP))eXlfyVfrqN{aL>n`1P8&Njrj`R|82#ZJe4ZI3N|=6_ zOVhqp`2vT4%S`z^neCZFpw7?Xbu(rW``8aQ!hG#_uAmFvZ5HycD&)`+F}*W}Ru9GU zH%;Ey%TpPu0H4EeX-ggnq}>lpdap&ss$mxn5f)8}2ifv(y{qfK8p%$GzR1TKBVqb$ zKke~;cy4YT2}6ZKHwFMjn5uG7Ao1kfJm0xtlvGs)DobvkJOQAnAu0|KW7ETZm5iOO zbGnOdik+Z_psiqrpb(_YkSm-Y4aY8Jw;jJ+flG}I09Z71A_=c2J5s4Q#mlkJwmmgu z$SMN>8p>!#%y8*``1Y)gF~}vj`0U+gNB#={^e1>VZlxRcsz&tq2E?!K!Q@Sq__z~Q z=F)S0P;T%o-2>3R@xsW8RBat#Sq4WFk<^fO!vcB858o~%tP9Jn-~K^I8o1%rfKCFw@$eLz(()cY!tr5TMJk#>?DS*5!t%UTr<3T z1V5j)dT|t**&qp*d&|?IV_T-zEzX|=_Wg#)w*8r(0i5l^;fvA4*RRNs`|5MTjOj9c z$Not|hjGL|&>UL&KGK7x70c73LpaY?z?4RdX@{v^XE4y#y0ohq&cJ#>Y@uv?DxDV$ zW8U{Atmoi0v^1^k>}WT?38`R8qjc{-iElI*kD%6!JX7t|9O_r_ruV$A?59fWUuiUI z5kUi6^NOP+tV)yN(ovDdJ7&A z3=>3Ttn6EWTT>s9Liujj$VQD9~yZ84c)bEe<+TLp}4gS!ANg$`(*$?Ls@M} zNjU!wq^4c7BVS&DbI0L)@U1xFkSH7ygYZmD`5 zDyus)vL!4acuGqaQr2j$&9W7@(lZFi4eqBERd<1`u;+XJ1F9!9v4)B$|amWh*CgUiQ zob?@L70e8kV5JCKCw?k09u3xztIC3oR!_gMET5YTeqO))7+iG)Jpk$rii5vbz%lDU z?z^k5_l0UENWyVcDXKd_4X!SZmBU$Z&3jMD*nK*LU{?qy<94`)6E~wOv~yq3j>_JS z89#TZfumz}EYDIH5a!@O4R8#n$@OT_auHHZd7aPPh$^zTHxQMF{S!fXG7*B!x9cp zRHMeASO!ovtQ&Eo>J~ar6f_>x5ia+GpMQlf1$(Z*AK!y28C^L;cbgV9!?i^ajhmEbc)K8GHS=e5vjAF- zj>Ul0w@(HDG-S)JguvOaEzSqK1b?T&wG=fsORJmU*|a>(8u>B2YbEB@E5$v{gK3c% zn*Ru|{{jGQ<^a4v)u#urr4GejX43z~BiV$Kc@_Z?_9x$?*ZmP3@ub2+7M_jFy3NJ; z)OeyXi~Vrz+olp7TLjawNqO2o*p}(jAFT3upvXYU#`ISnO^D>43Hg{ zMB{g}4UC3RUE$m(ynYeAhKWD^Q}S?qKxb9qbWQ*;eY?}Rs}eOdl8;5bcJ4&&$`wMSCgK=MqFYW?7YnkI+Tzoy0wc%Q;nT@=*)ydcJ15p*&xAO3 zH!IKZgutTI7Njc=0QgzAi~vv@a8^fJD3*O9_K;X!SO39FI4=O0Y1<@9ssm99CkB(tihako#1X9|#OSeS^05p`7jp+=( z{tG{(L0Zzy-uu9s1hfxf>X=x*u=f9UO{&Y@v2pBf5y;TCWhj=|aK;Yf@I<3= z8U(74-MdeY@MlZ7@T%d#VW_Yv3CxZC={+El1LI=ZKdS+|lM>jKl)$dxwKzB=iqtNZ znH)ut3kl8{06KA5v|1q?ds^*IPnBS}!kH#I_Yi$5{MTH3*8>2rc4AeB4Wiz$d;`FD zsj78M@tC z))J==9LQLCShy3X8Qg)%5Ry#NEj1k>$^YArTchDT8jr@nuyvZ+J(pkKMT-u@e>tFiAAouNqO3?3jn5Iuaztnb&ur- z0O|}L8p0%DIwO`mtXq^LNKjl5nq{4{pq!wHd9}#{GywE4XgH{R?qR(uW*N*G?%%gr zjug3P^Y*G;Wx@wkXY=H`d^skD1LNb^H?sk|#>TRLOdJQM*JtmbD8A?rMsN|kH8}%7 zM=ptVGem^|>m-VwpsZ-=D^}kWPAQz*hF?6>np)`TRV%R>pE?r??NN-LKl*e1rhooqmcQNs$~0Am|D1IbJX0=NG0>eJJj2S*|KJuWJ{Pn}_mA zeOKOK!pn_~0)T}=ZWM}~(?vtBz;~Hm$`cHbnaiT~C(1_$Tf#~6X)lb5 zIkfkBJpj`I;5W7NI46^OQeQoJO!at4%y?QHDs$6JhQ$-3rC^HSF+s;7ZbV(Jo_@ zrEgLaMXBRs8vr(8mFze#_KwWWW~6`TH#`-_y~3Qu?o=2hQbKaE4c89<>*NmrN|?2d zMA2fFYVn6?vhqLL8MDO`k~M-s0)Oj-Qz3g^|L7QYH4mgkBG+FO;5y_7Oeo?5G6Mh_%FXsH0^NJF z_ZsIzyYX@Cp4EWe)8pAaI-321A~|H84A`BNz^;Kc*_l|4M&tqLCCvY9Vd97?95fR% zhqKCaEc??0;#N1~%i=UyLjU6#0id%W>GUvm-`kjj)d@-FmF?A$OqX9`AZ>qbbTu7M z+6&AYTAi=(ojM<7j-Zn;omxWtQ{CBB+m2VMF|`{fGvppWjr+FVj0&OENKzD-!*vz= zco1`jgmZ{SYp0TIj2R1oh5!SN+fXocGXS8omM~r>wu+34%T_V3OKzdJ)*OJpSO8!` zuCCD60ARhm0pJ8G$&S{-CrVRUEBbsYKKakHt=4ef`vfZmYX$8EzM`mxa>xZV81y)3 z63EvA0}G~(BhOkW0o0bdisz`D0IHx8)GbswDjEQ<8qMFWQ=QVZ9$T9ov#!FvS{&#X z&f&4K?7tckHX@n>J*v^q0Dw0D0RO^AQM{7b=9spMn$L?bYUEp~*+N|sFQWH32EZkuQfZo=8e}%AUj7*k-P8A(H!_;NjRI(SivgfNumC8^ z`L&8X0MtHQ)sDLIq>G4ZCXRYo$5h2c`|<_ynuJLY5u}Q=5yjey+0}(i5@xyKD{AM{ z?&=YX)56VFOi4P7jbl$zLrec}SQM#!!ue`^?Q7w{gCg10tqRS^0RZlw%z!o_jO-T9 zq}i=_b3hn}YJ2dCj|bzL1~9TRzFipidtiop;;`>g8jlHQ?C3bwjELk=T~}TWa%5T{ z^q~KSDyiTmn6K&dO76qzZXxWU^;5%FZO8A?U`Dg_`-qg03`(rRgavK*sCRX~s)F+~ zu|bS$8_ejo!Hnq~#Dp$Ym^HBhuUDY>Y%0&I%aW5W07}8_zy$8;UzMff>#(=B8=H&c zIgu7mhG#4`k~)#tISrXPK9+a;ROe7FioF z@aCyiMe2`@;qw{s?7kZIJEJ~(Mnzk~fs-1tdq^ZZIt5dgq7|Z*a|rPD06<}`=JLH82?1Vbu{${fK!Hj9m?K_`*9M$5o-ah>FN6gKZJ(WfZ5I+{(0 z<`bg!cBr)vdVr3aG!u0sRI3TRx%s6G-Q9}xC1g|#J7+dv&&1m78y&-e>Gjz&Hr5gj zJgWh_$JAy=?NU^~$%l3Xs%Nk#R{>zL40b+{TMHeF1OU_^I;l7dWydnXBbM|GPK(8m zaoH-*pQn;a@Q7fUFfS7hkBG)&W%Y8_lavslA;J=7DPdACqUFWIf+fQKVZlR!hh%1v z;9={MVV!x2%q$i6jTP4luW22La^rE)c!Ka~E?6omRlB=hXwQG$O753NVqH# z?S_lPC2siMe}$hCm?6q96;(P3r&jV}hG;iQiFA_-_loX;`P|!MppiHGp=<0<({XX^ znAw2cqhmNQE{^@vxxnZVb6*vm{Krk|>rKglB>mUe!JE4g{|VFT8xfoy25EftTPBc~aq*MBLsVMa%xx z7~C#`L_%MzNn*0y8@=#I2*j(N=e1)-1mYFtjc1g%<$NN%EDMCYW=zu|Vg7tH9ud~# zSNFu*8$4;bEYk$$_VKv%kH~yo5rKF?s2j<$GptB~Q6(&%=2xGASy=#J*(_sELj?l> z*8N1cILD4YqO+>rJOC(RQeR?3r4Hh4R|%8i=`6~ADVj%K2keU#3|E$cxwX<_5pJC1 zXwA{|ywzjF<7Ow%wE+Ol;g%3%x29-uag@e*y-iE`TB}03gWT7kI#Fmh74;F1tJC%D z{{hsR06k5T>XRU6kbNR@nTkS&qUSN|01v2RPKj^}fV_aWfpx%mpdXL|bO4qCn^E1VLvX5JhN1myaClrC`>%$5%xu8!5z!nNUW)^RBG{H#j-r;wEN1{{&xQN~z#DE) zgo|b}MGn^s0CV>M)E3%{`^F2Il!~Js^6A##X94yBOmB-A4e9iq98S!{}-qM~M{aP1k&tN{_Ej*j6##{p8-^s{bCVBU4A^L0##rtATA(ZNFz_PKzK83?H6N*ie@E+kGm4^ zQUoJZl1pO9yJALVfxEz6i8N9zj)`IKh-8a5)(5l}PATHkMk1=1NVHP|&I-a)L)a*9 zz7=Cu1L!DDdWyNC)OTV)4^hus__!;H(pth2nK>(_Z4!?K2;3Ecmxj??0i=qDSBP)D zm4vgvP3vi|dVDRyE{oBj>E{UUnoC2)utSu4O|0m0-S_=#N~teNqW?m%wwy?G6xb^q zS52X%LfI=4_KQ*ZoJiB-pdmN=p-T*)@C7HQn3f&u`) zV3aUR)ZVUJ9L}i^>8eVc7IiL)4bFK2z!j_!amU25x;D2tK(OlbkZ3So&`{VnR3&DL zV#h_tM!8jx6od(8SO;U<1f7h>BmM3;fck@8v3lOQ-H9Ac!_myL=sb(H?P_>#29Reu zifRX;D+KF1cB*cE_W3rd&=hSVMuSlHYg&bx2czu;8~~cEgM(mT8t@tLBd{1~2Dku5 z&aIoQHlg@;tHz5XqB(dq_tVS<>=_xuzQIv^(lMAYie7?tTL92atyw1rE{jjc++qOe ztYyrXiF0D!B5}ewkL^v0W`l_TA*U+9I!hV(u~jteDQ6PoNn?#>vxxjryw%l)t*In< zL9j(oQ{bXPUkzF&P%qGnprO|E5w`_I2stI7zi?9;Qmd?=Z$hri&~nP?0-d!OU>GaInS8uqtmWHkHxRp-4PVT9j; zw%=s6)`8K;8~!NwaqPRRp;%1;WrxqroZQ>h*twniGVcmV@xi<&`y)tBsP!9 z=pm?W_-G&pM7eWfXu2twG%;&KMlU)Elir}O=(I^Bx=5In z$opd4YqrKc`jGOf$q6xTEr3#S({yHv!BaBYUkN8IW{g}oCuS|pSWHpEPGjjNI?ope z(nE>$C(&w6s))THLfx|2(*uMnjB_IBl=#1v*I~O&N}!fX^Qu_V!lrF+&L|Q4xfmam z&%Le#-E&LYfU1ObufqS0uEl|w8Tx<2qc||ACa-r3BUrb&|7VRhc=9R$G-T)|w688e zG%y1=2Ar_Aw{@)JPJ81v-JMDYfSSewf(@#`SAt8T_&*}x4H58|c>ksmT-{4AIkQrZ zY*ab+$&rh~?V|90Q#h{|tQOa*2}zjrI}PN*^Qy?t!tH_x`AqZ}oykz8;;5-~7J*ZQ zM=L>N+0#Tdm?|d@i)sCAHXIK9Sz8F}GFZzVBs((b0tfoL<~8~7v&HTN9u|(z zi9R!KTIJt2@Kq~j3D(J-mleP%%dyIQFI=_=)`=Fg#2dbuzBiVhD$i!Q^RsaKSGc?) z9_y1W#A`H5Mcf!+-$c+@PBc?*7K@TUiPd9l*c=Z*fHmh|5qJhv+v@SR%v$O9wzi`} zP7TnXxazVk04QX76V)q)M4VE2SIzUWPuD(Z1`YHvYgEbBHiu9LDrvJ30$c=pPQuh7 z0JcZp1=W%qsf4euH#-i=}|_DnqP>w>)YJsZ5oL(b3~K;0>F)yG0v@d zHWEWA-Md8eZ$$?Ka$USg|hIT%DCSN+i&#%oGMHEV93zTk_ij&`xrQE{o5E@ z-sN%yz5!4Y_SUc@8%xqn>pUnCyPS`6b;)c*CCu5kNwmvkP&!golv3LaG_%e#9f<#6A!g}?~$3EQ#S#tkyQ0Cfg^23igBx|{Ms z!z9asz-L zpavz8Xr^O)pvL*wr(I*zyF9u`JAKOYs$>fSf`UjilA0U3Z88C&KvD)fUlmwI(G2^$ z^#D)=D3U$GHBEfhE89d;5@z)#S)6W@ZGAMQjmk{xFscgr}1pr#@ zPprT}!cFzrBb+abq{n3=Wsg#E6D@-Qlz_9U5}7saQ$7Holc2I_?VY~v4{E7h9sn?r zqr&crxc{Z=xX*SPMso#mLBxD6w$#gZuZ6uRRz^hSvnZjS2PGQKh9N`$Ny~vg8Ej5( zEV*b6-Nh=<08q)eXd;v3%|YSuqZmEqYEpMWF?ljgo*a_{X~Olah<#Vw8({qG+3utR zs6A*GXg#RJtxC-PUz8UB`~lUefUZzozj9pps<3wgInP5=Jy06Xx`eO(M`b-uV|8KT zt<8aXo8%za$YlxINC{n{ zogU0@wfAkPMtk(R$x_DVcl`A}H8q8yszs8KP~FfSkrNPMwf7lS3aCiYoAAEn01!Y^ z*dPzm#I#Sc6%YKIP*MGbtSN}pA1}4#L%H#bsB}!M>yYPh@&JIDwq4DAD$YgWkS0of zBTB6kHM)uC0oP$Hh+XgDQ<>(kW}7r#eQc`CY9-;Q^Sv!%v(bRq%|Uo{C=L zMi#(^GWyR4YkkXeDxm~PNp2?=Gju{MuZ*b0&fa0{?_7z`TZQm^uP~nK9nP~|gGg@T!%9lF z@5suJy|a1*NqE26m!ydecy(wPdwW%9e~SR#YZ1V@9$~EQ5ysj9)p)Xh7%vAnGJ}}M zYUdY5{QD^ZANA+u()6Cs%D?Vrpj&7>0JGQ}lb}@*wk35y2S6L{wrYau0MI=%02DIK zM%7xNwV;#q)*>Xg>)H24R6hb8u*+ySP#Pini3c^DWW)QtQO_|^1RobqZ>@NRf#7>@y|r}@#4l(BWMLxRhQ3XpRv(3NtFLVOr9c^ zIbFxzUIw)_UK2xx4Bh$8Y7*w3<`CDcG!x5Uor;U_F1Rki zyZIcZ?r32vj4=llcmODH5){v}&Q0K9JWV;?HG5DKYpCzDAisie*69GSnoi@Ob)Z$Q zXgzLPMUciZKn?!GAkl6Mz!6yk_N>8xs5y*`Az zaZ(VdEY7OH+bU^1dv{KU$Ma|<9F`IPbX`)fwUNIrWkb&}j^Hw6A+mZ93)96lL_FV( zR~)HzACixCz1^XWv`&iQkeM#Cd0f{z`@3-XVKYsZHstMEIIm;qw#KB`&U-9p$)smbF|?w+U;rf=>eazzUcR)(jko!r27IO1KQg6{zc z-35z`Hwr-hpk&bNpwPSX_p$;&bcIm8Z$_&?t1xQ=0a{M1`Y>NN93Gj1+{U85RnhD2 z=w~y!Y(k=##Bxoy)doC-yg{KjDwH`2$a22sBvcL*2=b$m0+5kM(2$`U(Cq<$!bOew zSawJ}H9LL%Pq-?OSA+s{_@2F9XMpfOaw`BJfHo4WdORp!K9MVDWyfU;46OO1crQMW z>-vuXutFC>g0OF(7R-R22-8CH;>fLx%1a9bJagTYcH&LDc7reRU|qO89N;!cupdf55`ynG%(1 zEAG$T!z>4?uA=Y#gCYW`Gb@4TN@4nzrmOcb=j;50y52DWX!V##U#|0ec7k}p2Em7d z1mo$+*WI%RC0L7lJP#^R;n3*-Fi9VilHUQ5Og5K107()mHUNz;kB;Ijv0O=<$b8IH zRJs=G8;nlkq{}>~F&X#!JgU(OZN^pq5Un*#RBhuKyhBgEu=}_z4>R;cy=72bUDGvu zaCdiicXtcHgF8WjySo$I-CctOhu|KZ;O_43@XmEVRo_=N^KYhV=Ip)KUfsQVb(V%U z%-(0s_I1~H!e0M3@#NkHwRFWZQe(n{h_{OPDp1d)kOKCAGwk6c%9ZH@e7?v;{>;sg z58jt8jOaH;bhF2B{-*ThB62?Y;6>1$AVU43wzLpSQT@K8c!=W6^fmvy^-RCNkq+%l z{?0_;emL6nR6ncXCKkNsw22R|xt7dB?<&Wri1Zh1w*qxMg$4PMv-QT(AMQq0B7pyh zeYlop+Z(x=;?ZPTD7vyyQdejiO58Oe`7HR8Qmva$=QNU8RY7K^Nw#%<2>d;Q)pZ*)QjX;KSMS0Ricu*%?f{!prx&K{O)Jd6r!DVrK!qYfp{pn*KI5r za$}ZsmpX`77_Mb!L8(Q5#gQ2%{w|PpaUh&g@JnQgnebt9k4Vp?+vJ#K_TQp^$mJn~ zXK)yY`)$yP7J=Zfx%Vtq?Ka1PD@V9kbJf*DobN|u0o#{-R2%}O&>k{4d67<*rp;}PyXJLo=68hsMmMr78Lr>fl zx8{Dx+2iLX_;yz#W6^R;gOkQn76t&hgeCR+un@Q3O-}LeI8axl0F=kUZZ71t@Tcw~ z@Y~}KHJ|Z4-+%`I!^C?Nb_K=XjK{M!S8Ex&+T^VB&5eBWZt|#s0)g&@`?f>F+7Tv~a;WpdyifVd2f+ z)GDCD6A=hnGtW=3;X$9nIpR5hG>E%1X00Y6oO~R`TA~c(n*slIqlJEMv!9NcE<3)C`=rSe1XLjUB3o@UL(^7cOX$I?ax zC>OS-tFh6*mL`SJW^PKY?}~4?ogIn*qN`G(oKb0zmd3{j84H*B9pWBxruz6c*mUZY zg!2jL#+ml9hy3Z5^phmUt7`jW+{gM+s*(jT;u~p3J5W!XP*3-QHm?NH;l<+cpUsb^ zMDN{N*^AAx+UH_5a?Q!o&F2H^1XLM`^}e|l=2a}kl)A%m>^N}Ybl+cgF(+2q9b?-u zVx2U&ID&eW^323hdpA|>$8agVZE5y06m9x0b8(iR9My&dRm{_7F9N@GmE8&9c)G}# z=<>WOg7==g}tWJwEj6;*y7J%U z9lkq+OG|O-o;&i)QEz+`!8rGhnETT}ea+mp#Om2>0$>g`()Myvh_PiaWgJ91gTB@x z@26Z~ElG4Ln0RRBf-YN|Y1>t98n~mu8U0jJT?7YHhqG1ZM#M=nMuYlxumI zpHUNWa|(h3yiUJx6D|mfR9~MTb#R@)Z*C;|69#XWlzZny7{Wceh<`?GT!|3BR=F`h zR1EA;kE2}e-5Y88HUE@*THuvo+guDZ4=B?oZEIz?+}1r;ZSPk3*g4O*++QPQ{PQhO4P-p2XR2eJ(Jn-`2PR*nGbZ>%R&-uV(L2-B&B* zdrmNbQ4nsM2IVs0Cmo)E(2x$=edkm2u>wqhtut7FwKfxmrsQK0Xm1z%D%7GIXMkd0 zjPfk85YQxqkH4DXJJ6M?);8ccqh-#7>Oz>HvEeSZmI^aGcOn$}L_~4V+Jfzk8 zAYe8KU@fv&ZFRY3Y?4mf1y?sGo~ZtGc;OorUQZt$YrXNul5-x4UY8nGdrL|5yZp{A zVsActYfxk3+w;FNfWH)OTV5WR-txEJZj}nVo%(kNYpa8y-=4Yr)c)pFcnH0zIpI@(=M8F z#++Lv1$FF9AG-XDh!!m)(TT&f6ifphBG+04q;5&{b0^4veH);M@=Pg)GN9O+54Z&n z@E4MKV2UYZr$>#(i(jJ!T1Z1{HwWG?!FsfucwRdg*y1<-^iRG7x z5im6u8%>OEzyjmL!a^_kc#YrQv#GWHC9Yj(7CQ?5YOexhs@ivu-DB1`Cr0Akb;{1Q ze%wS*%6hvu={;O7X^*EZAR|CyGiJ282lfd=i@ZG9j2wM!eD!S3urt9#u0qki)ANfC zeZsP_fts~pl*J)<&A5ygwg{rmHt~JS&4vS~LmP+z^KdWvTuj1e;y7t@yylX$YPeFl zKN0r2LAH!SoEiG<1>#=#absl>lB}ycVunXxTd#KsJcm*Zx9xcu#`iCRxe2;Zz-jo` z+6XdaJ|R#2Rmhy-=Ytd3@o05ty}LBADcSM{!%^i>7zw_3JQ_ z{7H}Z=Qhftm*|m`{PGpILVhUgb=&6z1$hrVCXLXz3jc&RcPiS*m`!|v!^;`PWrFvV z3!m9yR19P0(DU)u(H#TmfSmody~9iZ@mqf(BO>bRC&&W|VpqM(?&=V}63z=*&a6K( z^ccQ&5rS0&UQ#`1dYDmzeDors2_l6g(7>$p;_!=GZQ9MzN0rf)wu?R69f7@$;77>X z)f~R=E0@b7Ee~K(wflDNmC0B>MMkJQb+E0fo`_xtQeOFe!9*6|R(>gx9gks|fEK5r zs_%89;UofvY%>)z%8lUyd1*jdIW2;;DUsqGxf~BZ!dv)Ep>^sOvE2-@WLhdRO>lHHBDKfh`F&!GeS{QF`d;!~5>`QR)yjk=ApdnP<6{`DHdO~am2 zelw54F%lz${7VP+9@_bj*XI`4f^7_^^8%GJuBY-B-BRSaR>%Ha_C} zQ+9O2KaIAVKlYY4xx24izt;S^l-2(kf8vr(4uAgQGt7KWc=hoo+H4hJ=XG?>B2R$a z0Mkck#h)S@wQ81(fJZgqWg_Zjj2xxf|D z1sa%PL=#ApwvH*EnVy0JM61GzgqO)I-GI+A6LKajO7>NGdeHJ)vYG7l@j3c`fEihT zlRAUS0vs@$O{ELppd2f>1JLVmy6$gUH3#n)Q6kVt!tcenTRgKLf?5U#R-VVfTGw|N zRd`*AR|tPL@ZI|&2Ydu}N_UsVFA;v2()bcG_^LcZ3!8!n!2)J6+29S z9KcT&GCqD=<;*cit0mn7Kkd%`J2Spp?7q@a5{;pW`|uoLGsp_UX(q({rMvmshr>F! zT7-dL@&`&-Q(oCf#A=MPS|qXSPXcsk+JSD?PuKDmfzLA^>*JaKTFcPfKapJ~X20}h z-=M-t&^v^l+;)Kyziw^k*8cupim*HOR^virXJ^vyieRFFQ&LOmRwmRtC-pEca3Az5 zo~d@QZ*9R=0o;Ma5QCVRRT2jddndtbe+ojWW~hdaNR4e22Ezf_)jV6>bpZSS*1>+&F`3D zK-a7<1yIKeCReWf+waq&kfy1)_r@+dUGL{1!>+4H&PL1_DA}(L90h$!*fRVRVN3qV zshYOelb}`#-?p?5D*flBj$0JRzt8RQWW?vXx-yfI$4LC^P^91{OF}ln(G+1qqxsGM zVN#$y^2EA-is?88H8DO!^oUComZ~6R+8aY0rPEzGG7?{uzc3t@HjkZB+r3&l-rM|OT!^FLG28rM`44X2Ap<<1Br(6awdj6x z#rIRxu)huS3wmqDA2E&imIkf4AJd$}Q6-6G6tS4N!UD?GxGvaP-#oQuMTIF1*6cx5 zs4o6KkLu!_pkp$n9ka`zHDWQe-$#yi|Nf0*4bQp&3T9cU{EueKF)!QId0M5<-tmj~ z79cfdA3;%K5zFhh?i_?7y4CAczJg-(rVnd7L1I->AbaGz`UWfjuvgfE#YUMiI{*2l zw+*7_8T6jvJpW>2FQ_wIL(t!hDmcmd+LpHTq7-WLLDkCU=j`$BaOMvcV9!nA0o-5i z6UQ)#1H9W`!T#%li88^zYDKxxw|yk?0k)&VD(>7xI)B_!6Z|RFKj4 zEZbTfPQ;NCqJB9efvBy_3s^i$-sv+@ONLu0p`91}h)&TJ!I%2y!ALNQrm;U^%x!Yj{mSTEmKFN`amg-v5PfzzEF2(X-@@ z$ceRPn-3k~pCLFd*HkKDeLj;I9v5S9ow+fmFMc2ECTi{VgBx4XZeY04d;8@~ zJ>ni5%wBIQTSe@wa4q24=1q7JKA&HFc`*q$@HNQI()O6kXG4DH0IrEoayaYYw<_4+ zuRj97nB$)J#MGvT6I^LYc37oIpS}Izu5+yTiKQ09Rd@R(q-P+0;04S&IFt+nmU(AX zq0bDHfki0RI#@ddPxVkG=3FShB_B=A39_zNl zui0VT*MI(*^PNfO?#>L)%{16(2J;{BCpR1=tH8A-g`C#+x)d55&XD<9c2tuL({8e* zh{Yk(8)+y8f=K0q6-(^*YmDDno8ow^0tmn_j!*063ik`p#AtizJlmID$(5<7UDdVB z9J1p!N5V0WR61DTtJVX}7MKS{#~{P&T&+pfwTCZ7%S%FCp7&IY%gQZ74hkz1xI@Kk z+1{b^8cXQZjOM^dT9NV8z}t||ihKyU7B!C@&VoSbuoat?_yMf|QJ-oOE|WXghD2Hs}m62VOQ)J5{;z z4rf_Z^*`)pPiz{VIj`lPZMGV!ecEL0)~Fe*30GUIQEOTv->W%f_AH?&)Iy00*P{v* z_4WG|b}V+mFX&iojVfV@s)3cUxIuD@W}CZs5s3XhubJ!Avtjj~WiJA?mSpfAr?T(w z!w%HlY>$^$S)LDcMErk(kX(EZ|ElJprQm+0-zNIppPXyMK?1tV9;zvx41umg*%up> z+8<_eARQF5YQU#&HX7un|4-y?CsV2x4d-pfY8t=cb^n%bRFEJ_=~o;Mq?alU-A{YL zYhnX$n0lhq^2g_cpY6%XRwFAhhfzr$so{#7bq8^5mU$$G;yoZj!qb;`JJ)%PQ}f!T zL`<|>?@k_!aV;XA_J^nbBJWC^^Aw5@0vhc7R{O;%A(RJlK~3w*Q8;Xb59-O+nE(g& z16(vAA1PldU!&CU!`a`9|78{}f(=EzD$>hj7BHU6y-cBrCjMDtR?Bae5fnGtC~nBE zoHtwulW>wQ%DJg;Ee813YoSeiP|VF7!4% zqkb0fTv<<6@N^?m*vJX;{4>%nW1cTZC6NZTFG-%|j*JlK-hPLDU*OnRHf1}#K<=UF zIRf`lbl)m0!huTp2xAc$AHMWKp*zg4Wxk=(glh?$7efh5AsClX(%x4;+;2o_(^rwt_a9i*jFRpf}Qa|oIlG$4T~?Ny`27W&Zg9SSX=%?UDdNLR9dJ|DQxFGA7C#(|~<_6_43~q`w0;yDX{3 z24#5Nhp}4`!jxIr<5w@o0iRr7n?s|kmS@OyAAlcf6V`{7e)&*$DNWOar}h~7y#Iav z!zW6JGv#5`ZSRp$1r=nV;b{VGQkr^*qO1|z){dGa{1f`zZ!~Cqalv_xD zvB}P;6 zUG=8W{gbv0^`rVcOS5tKu}V9c1c3bV9Iq&sbtOxI5XbT9vUsNW{DnP7hG(mpTOD`4 z2)}4Hb%rV3bJzj7v6(WxTA$#>=IvhaiMe>@jbibds1^ zQr*DIxj1}E-A8*c*Y~AKG=Y1Dj|qVB*8x>t-m7+^& zX^~Br4Mg$XaN^y+5SM~f~}zq3A|r{Qzj#+ES81$NI&nusW|BuDdFgg5H?NoZL(&BVpi3|X~w zVH?6Y(auQjx7VwylrzA8uGzE6`pWzE4wmh|_`LF|`sE!g5o$(L6H9QJaMpW%4ekTI zO?-wP{8ZZ_5uV5NmP$L-nN?Z2Cixlp5x{H)s*Q7-9TgVn7Q`1~%6Vq^`*=@NMJVSbp2a@(=u2f#SU+{K>?&EHlR`Fpi<>gt33u$||@f9rZVzBMD-r&xbJ zjFkc7IedBGD6acH((FtD)`*Zt^1~@0a<*$%7>TZMHX#+_O2gT#bk#@>-bFFg^Lg1w zjD@dMD(zt72@=y_bjm&Iaw|e<>&+X_rPneaEu>|M{l+75ARGot3QC#A)x0 zJkzQ8d1Bftn@p1U_}6Q|K?!TaIM>>l>tS~JPx}w%EHjjyWyv(baiz;d?W=mIe7DTA zzb+%cJ$S}ZrtC%F^()Dl;yZG~=Y9i&QR>V3>{)1FT}vjP0{iI8R`*Q{q&Bp?N;}S0 z7lE+%M=9$L=I~^yxC$h+|0c%)G(Lziy-)N0%KSnUvlZ`zJ1C$clABUts;>iPR3!*( zkvw%bMtcuQIVH!3-Y&nndjeEO6sm<6pw*-MW)Zkecf2 z57o=vJFh;p5LmXg(a6wGKO2OdF)7Xl)bau@D|I~$tOq1tJ|K`k;Ot9!n1aGIGYsJX zE@&pO7$-@m^dd?$^UPi3tl+Y8` z`yln_aCu1f;c!J`Gf`5ww-f8@!!p>|UkJ!Go}iK@#G<=KZx^dZd%NEVeIM;0)$do;x$0vcVpOr#jx} z>zzwL(Omr-6d)RqBnRu^M$O{AlOlD#5_CSqby*>yugA|xef)Z78M>mh z9PHi@)o$Wv^rRWgZtr|1-FLwK16N2B>9CdF^aS^k*e`m!_go(yciI4CEBBZ584e!~ zA9gfLmsjpm%>nrzcHmIZ3Z|JNwb=RWm7iY<=O%PXpG*mpRe$;2=^XThZ1f(6xcM*1 z(Z`sUko&9&7p;xRxVncCI3V=$)6Q(>3EF4p>t;%{XDxdVO0}i_XEXpLu3 zJXL57|Dvy{ECmkaxQ&yCqe{Y^!0FYJ=j|X_!M0<#*s7=tW47ivjswbmDRBNl7R!iv6OSElXM(avmCyKx8HJ7pzQM@@e%#g`}japWP22-3tc+|JgBiX;o>q zR6#!7M2H$r$pvZ_`417wY}5%oF6l%o7v9) zk)I_>LCE)^wx?>v0swwCK)%ZF?y(SCb4{y(pUPFLQYXS92{d7nPK}5P7lrEY(!a^( zDnq=@9!#rio@Qdy@E5{gvS`>{4OE3IzjWTu9j`OvR!$2qvrPQ(cZVRlAR2;E(Q;$; zQ~k5|o~*x+9f)|M=B|cO=~D3}=N!%kUq7$t30aZ;@@q~js?~6)@*jQA6GUsNStf{O zPS#`@xO48v`{BNxgu9ne$dmC?*~34XH$fz1Q57O}csaQ&S!RF-!}qEej22ID0qMB| zi8lepQt|g6(}m=!BF+pu7cA_6M~cC@8_e0P%^^Zbb1Yqt)PWp4WWc2<&=Y&`hL^&+ zxpO}UFTJH<$7eS7&>}{XE$1UaJw@EdvV@;&(mR>)rR^YlEFAD*k%2UN3Aj{kcYlyb zj#4mtJ~q=n`}Btl06v99*OCu95a%b^{Xf=3Ga0uh9*zG2TgxeH2CF%0nUf+?$`&9w z!U{U3B#cidKbw=1vx6eoBs7pS-K*0xUMxok0XVN1K>!?HySnie0oH1XzOnmRGnBLU z&4)JAS$BgAW!X5vrRuw!W?sdt`@^B0Hio1CBrkU&n~Ak!q!YwI($7`%=524UThlrQ zWV(3HxG+>>%-4nhqK26Wk*&bmi8(Rn zR@fXExKR+H0sIu|pU3roc&>9Qr@J(f+MD4@HD^h^c#umq*Xvy(Yrnk03TB!yyVm+H zi5Psj%dd2~(8$RBEp``6FELdv+79^#Rbp}Pj`oLd*lJfrW8Q3ZR!0PGU8~T#-qHN8 zNsCdtSQ_dA?)9DEqjlN<;d!4UG$P4i15~E*(#LU-mEVoz)3&vzQ*6TIBx`Bje;_2Ti0-cUsG*G7Ha+2G2GvV zLqdxP$UjFlLx7{@CLCsb_P=MWw9@WCF3rfPjlAY46odKo*ew^c6+JV{=FRH3JEURK5cJ^YAB-qsvp~ff?B!MWm@&N27YbMIkx+d0WuXzBdI$=S(1G>$ME>b z;l_p5xbHiaY@-74bdoW+gFBn`d!GLoAnn3SIHrYK27eUMOkA9g)`Q=U-~w8W=0a<5 zuRnsJ19gHT*BcuVfB+e&z=IBayxzY#kmEMu>(C+dq}%ae2FQ4T$1{@~Z8?@j0mhpY zCWWCf2%;+HRedXTlddA#uudVKj9K+@7WWXFQgcnlyVBPd=n#RoTT8pxda_D4p{aw} zS&kSAcql-3rjlnjnPF|l5VFQ@Y6VxILM_`Nx24M}2O01${z$L%kr~jtHd%OUKu1M zj}K3CW3^DhTZjp^$%)DV{oSqk&4-@efOh*?fh8Gta>yfJ=^ zCQqMYFTWkgQ^awYyFQym6MO>^@CQ%g2*UdI1?iHR(&|*yuNjR}dc7?zxH98lsu{_A zTGWO{>uF&3Es^W;SBM#4$F+3tU+@JxB7PsqBRJDpu@s^w4NvZDmf-8aU%^z#W#mB4 zyMbD?1On>>AMLVJ(6VtVmVuY&aNx`W=3GO@y0&A0COF9^%M;dYTy5sjql+^E>W)_sN7Va@zd2+W zj_3i)y;u>vl`PjQxjH*Q?HWD@`~Sz@;li&G1#O)1Io$Y@UsLvfmAh)Fol;p)SWpO@ ztEC9v{K&HK_<6LV6)F!`LL^xsR2$wbBXuPzCLpHkYSeym8v*+&p+?NKwt}!y2@Nq9 zuTQ?Cd9B*d=~3JpjjhfTNmV_bql#F&G0}Q?(q|p*d)U;m@ciN^zQA)p((zi#eswF= zrWVU`yVtp<9eiT0)$9qvXUgw=e=?muu z0mMW{GbbPd9e3C(7xHAk0T%??Qr?pIkpsE^P*44Z9`129>tD02{?J3HQiPmO9Ct$) z+lf(%JOp9d1VuO@=G3&~y5|^)H@%SNn_|urnPjz{(9`#~vc-~!KRHj@uFFJr5{#=eqkF;(P@hO zvh*61aydp>P2q`$Dqky!t1a9bhd(;s&N?FtC4C(d`uRin*&FBpJ2Fe27`Z@CYe( zK8E3M?Vs9$c)+=?@fo7qiQJLEwkpwA-%TE+m4g3nQ0$r2HJ7&*y-JYU z)`C`zuUU3^26l!(%9%1y!7|I*?w&d2*fQHJAruyCI}xR(_!b#EM(y;vNCtSt<-4U= zVxftZEvpdT)j8&lWPJ>b)j~WBUHcS7!qgUo))M5Ta_w>St$q${9dlF&^CDN+#BMew z<-e9SMA|Ld@0dva__#Q?j>ugPy^n5xwp>BE3}a~0g~Y%xOt~KSs}$fMQVphJ+}!AC zV(NGbky<4|{H54`yo7W0FT;9VJxLWBFvwWt8sNiXKzie1Mt)9b-U=*uu=Mz>0}GIr zmd2q0p-Z)EEHn~CwS#mbj&=8c7l7rgxvHZso)qIWyYIE-TF`(*hI7LTPi&D!c@DPr zLjOk43quwaic&gc7J>Oi1i{qREU8kvW0a;xw8wV3U3ot?jIe38tl!BfBvC{ z-~TpW?Uq@_HBPFW!2S9qTB-TL7r)tP>7#BfAlt7@-W8wQ^s#`@Gp7lvRz-L~Bcj+B zAnUXZaV@fuM700V78>*VbL~@(Q*GRjEB&lNQ7%x4bonD^$Tho>Vwy0UWWlwm=Im-83N95k+mz3t>W z3x1TyJoJC1Z}hPT+xi@ojp!ab20ES42jYbTD=~r?DSlEDJf9=VaA?|}(WnvQm)HSp^+gv30@ z)zFl~#js#%-j*JG2LmfsQ*?CnO(N*SL-d+%FR7vVS~4!45m!_-bXPmI4#Fgci7z}p zdEmC=p4{y8Lyrx0RB@$@Nr|1eQ6~a(5L3BZNsE|uW3>Onacg+OSGaGt^}QnfZL`RL zRtx512*HtL0?A;li-zmRBnALqm21=`TDbly?h3zV4k$(Uf~1-Q6>$N zNQp)q?lCIA+pHgV#TuBqD2Z;yhgq~A6#B9r#@^V=Dp*J9LA!E_N(XsvSbNO`phaiD zX-w{cf%cB1khM*F!#?EKaQ*;cKw^gIG>)0%IE`lJJyhalSQJSMj~9 zs>;U5$kC4J|8|IM-~rUKk@Q-C}v| zQ7~dKV!N6e8KD)y1D-Y+>yYNa8WCNH%IzE{ZZ*@v%CD(Fr%h87#LEe?HCdhlVsV~T zc*%iV=zv#;MV}2c;lqIus?drNhMn^u!YLRi@E(cQa7yVu!K9IKyyXMja&Ly+3cgj%$x_Xp;(QBg1^~q^*W>?b7 zbcBcF*r&iutF5VF8HC3Zz2c~IR7%os=nn&XL(7-&{d4O{3 zisDlFgq&UD9{THLseLWbK#?Ces~7Nvl?PYh6gOA>U7>a+m<5G?6N$Wvob2tHdr*WN z>sNUb6=fV?-sl#X<=^bKl!igmOI}7p>4J`92l*8|XV^s;5 zzX+>ESHzpebLHg*S;o3%Fs75H(=w=&V`J_ zlpidZ!N@QR#S!8fR;Tz?#O;@A$R=b0sNw!)DnRkCni?!e=w?h`QyP6Dt$%5F4>Npz z#u2|N?SS5tv~(J%tr2^}Nh)9{I5ilbADS`0z$MAPeLef&Y0P*$h;pCb{c9XMfg<#g3*I!dT-YF{f`*|% zvev=*93Kx00S~Z9PUAv>^i|)J*$xK;=}xk7nEN3m2%}zEJsorUTMS7Ugu9M#euRRi z9G!}cu7Pw4E8_9c4bpe*ZLEV|K@b$F2n~GjaGXHn!|5Pq0`OxjJ4$_}Y>?If5@$Hf zey=CnGaDP@*EIv`*DQ1;Lehh0{cbbqAwh_7Wa0IJ=}XF;06 z*a+r#$nQa{_r+b(x5c#XI+BwZDGYFZYCo;;`_$TQ3R*2A9n?l<^atodyd|e53r*rH zw@W=n%lT3;hpXWaLt4#!U-E5{P$KBNBfDn*twxiIJ1n@k&n*&fKA~H0junL zHNLXC4$Z5f5k)PnL{#INjI`cBPg*Iytinb+=|e?(%MAa5&QNJ^mEje{elqCBX4+Pc z!-(FQ9MGbptU{4Ni!c@GDS!lkwmdmIP-j%N4}p)QgB0nAoiqwq!$FQFDI=L?>3u;v z_d^Sb66@uAZ~C#C`=^&CbBFZL8F8YW|62a7uA zIut|?(Ks_`rkKOLaoA`uhVhExwNiT95NHq8b_Ccf@L;a^tcaP2R*BAviQ(wXYrY?q ze0^g=H$;2X{5zZ;Rt_E)mD9vQ;2!}Uxs_e+K+VJwT)I;#CP=>^KY<>2E&QP~K#i75 ze}Xho|7m`$P9YsmGH7f+I$kS@kr+Q{_%VQuhjKV-1lT&Wxg?P5PG?k(#9J~9_El&oIxQj{qhQD zE1V?Qxy*)hyrfWTsncAm`!BF7VjzbAZkx3<%md@p^DXloZe$hmmD;46o@M5g(8_Lo z_zLGeED*B>I+Gfn5`B|r$eUqnXA2Kq$avtQ@TX&g%ii~YZQ?Gw`xodzNLX11Z#l0G=ZHP^^)2%S|1 z(I)k>xprEulNkfhlT>gu$?14L$Tk zYY-mxqgGL9MaDA!mlNEMUKaMoJhTTY>G4YxL*F!;OgH{X=i{O92u3n4C|7}mku>f!&xyY1C ztP`a2U^k(=4Z)qK+O+*700a$DfjP;y`J`>*A2RFx6NM8h*EYJ&F?e`bhMr)X>2@Qt zzqE5ye}mLuh&Q|BxAKW zCC1PvB&z~%OHACx#@JpUl4&KSA&5A$becEl_p$9 z;V9Y9U!(wS-`r23dh*{tay%BQm3?zU^mq9f$>YLV$es1ekhr9Z{)($l+}oMDEQ10B z4FpX1cC0&~yf^d(XMNE4*-_C&Bv+S=0`3p7Sv+oQYy8JwISD+I3n~dBv;aPk z_XK*Y#M_KkpL))dLgS0VWJf69g1!_LnnP8*!Z6(QeKaPD+thD^4`7DZB^V9CD0S>|kMi1IS92@(8?5WCt=*}R4B-8}{ ziUgET?kgnrfcL3q-uD@p^BI=@&pq-MH2T7>J%gLvF9h9TUOFu5h9o@88x+@#ls}CE zJ!=Y?gBhR+x)u=>E45qU(c+iBTsi7dCWTXR#-hkq#YuChf!30Rnz3NgIN<7`_=P&_ zjbC^J_2YMKLaU2!Q<+iTDGuG0F_>&hz;3TP#aynP4Cmo(VleC(=&{WW|GAhp~_ zSRQ&f=Fflr?x64BCY+3dZTlH?uycCPBn!Us;bqzF!l98S)N!;>wtwx@REN`qt68xz zRSPl&^80T*0)GTH2knRwWA2zK#dcSx&&;+R-g5yxtfyi7=y94&lLkE12m$(pT3; z%C$f`4Q6)z@5vK7i{PF5dotGyIch?nJdgH6eY0x;A3(Cp|6O4MG|aoGup@&sHW_Lf z!2cWkD7pA?fX~5B-v!D|RM9mAgazd~9els(;hT6!Y<7xfo!GP7&LQWfIw%M42yELa z>O0(cqTk8%PNj<`u ze5(UifRvm&{AkiY~KT^89kZWP+u<_7S!C)KlDTGHQ#WaIJ{kwDgEL z%ixVNu8V`6ogRHTAOaUeBWDX2Av!@c@|w?AEkZ;pdcaUDt*KgqQ&kf=>g8~tCmejy zKU+6sKik7wv}Ti}SC>PZto;|}19#ts#ylR2NEe#z5Ei$7Qo@Yc;`U!Pa4Wr=61_2e z#`htT6yE$TC?DGzL~T(BQI?#mPbF4s9O$_at`&|_ve4bf%vFDdzO-}M5HVH`u@8+OsPj;#Bn?}!cO7Z}W`hQD8h2Ljp}v)};5Y$3>3U`?d4;~r1pi{6k0*DrM{PL8T z{vfyZ#oOHA)Qoh6|9%#JhK*ZFQ~vS_@=!6$xH>1dpb>q^X{dh?DB)F#eKrHxkiuO!=}#M!m0XDtY^ySN^G502p~;Z3Tgm@-k2{ z7X=z?Gx!T93LIu{F{;nKNfK7&UOwPJ0Rn{R6Btm}^fEk6TVmUv39ScizsC+m9-Hcp zJjF3(5=U>sI-VAKGbg1n5}W3b9@dC zmcPo!=Gk>+MI&H4pDi1?C5Ps$WfRv{4w_iOGErL9w_H)5?I3ALzMVmSAV|7FXJY3} zQDjU8$_?Mt@5PAKVr!quYC$!xr4`FPM3dIhlGVC7*-x>pGQ9G`HQNjuY7~s4iTvax zf65t~Rhv=XKyS!H75SNYI$o}2;ZIJ&c6qBWE2{;@5POV|)0FKSMcvu2)z>NTvmjfx zHL9)V8$gRtUPbX4%pL60+NWZKE_^areY2IVl}Fk22&pV`5JiLTEO(UKb?#%(r3bJ^ z%>q#eTSYUJmE@aaeV&f91_{|TlJcB+cHxNUq#rjUR46b;Nq-=B&y_fE(8`u~L_y;*?|8GcVFLhdwnnSUs5s?^0EMPoGBk48&}nbrtjUso*~%nxYZ7Bk}-sOIb_^r8Pv(Zxhrbx{s>1$ z)wdm%cpBdB+W67c&1}B153ni5OrmkOAC?+*o3}q%JA!OD0hIH*x`tu`Tk?n)fPH@A zwAgv?kYKzppN_F)7nhW^dO z$zJXwmepcX4W`CI(zmc83L~1$k?FdJ9OMd-X0#!pjC-+BPf7Ov-LDtu1 ziry9mjuf+Y2?=q=S9|?gO3=E{4`)5sbj(U5~$w-cpA1HlzlHXP&O+8{cSQ zgP#3ezt9Eyv~N-vXd1!o^6Q1Okn%*0_^LivECgb{;_EAeu4Pqz>{-WS>krM@ybSz+ zOEr7z7)n%5e)4NqtVZvRiek=z_0TfcvZM^tSr+MPu))amm%Xr61Ks!7=cr-1*L7f(J z2st>4DAdRc;1*AHrdsnz!%aijgvw1tu|z33I(m51s9Uv1W6n>1UCTB**1ua>j1JcB zaLS47fXCm1_Ocq{PGxXvHQN3WI&xv zYbvIWa!S~Uiz@vf`2aYoY&2MgQh$In3xgb-hy1xZnEt_OPyI)hX}Mb}is*H4+Ewa7 zet9vhoSOVI>c|TQ|Dg1HwqXBtv8Kl?a*Rmb>^r45PYl-eE3R}AtFeCS$g1Ka(wNJ6 z$zUl_6Bx?WU_O<;_dFMmBW;u7Ip=&uDL-i~QM>d9ocXy!i5+tq=sd`hkx*f|t=N)i zFcb`{i_QbFJ>%2f9#iEvz9_@H z6~Pr6_P=yryFuPZqp4}I*}&DQRoE1?r_ZnM*$mRg>-%elB-UXHP29qkH;7o@T_(tT z?&h-KV|m8(9HB%!+ucX=+NwI2Ve!8x$|17VHAE>_fX)|wqx(1lfC4&-S1DI4Gf~wt z_Z&`asVa;~C1tMj>NQ!VxmUJ$0XARTI+0A&n9Y5SE7F8(?1sP1^TV?Zty(a&8s7j0 z61*@9w|!*V5F1Z1R6g<`U6`FvJhqI3cZYQMzu7SS6=n%RtY7mfykv=ylQ1R?>kT7Z zLQCfk^RG%RxeUhEeineU9N4vf8Rn&?ZhXkNyZC{a>x6RJdn&%0Afqlj2y0YKM3cCW zF-xvu)`uAJSboc{VPz@ei|Xqx{&I`;X0TmbLmGGIscYoF&ZB<_GVGUNu}=Le^~w*s zuTt^ypxb7gwqnDQ2j#lwV{zD#x}HC&x0iF0=rC43h{q{sS(?;*ZY<%WbED&cT?`K^ zOXN|?0t4y@A*L}V9cTMc9nyL^5RL1;j|?`U*lAiZ{5BKuusO<#^s-hfdD*1PmohKW zWD0;i{}iJ$Pr-Q&qxan(%WqEg+k;p?MMa$WbP#MoAGXNawlGvsl{XHnx3_G4BDXYGnFJsWC13Cq&rhg`CikLOg z?{$!b9mC?B0QG6+f4P}@CL>z}gkBuAl2e1t?W9ZujKe~>O0&xfC&YgIZNSSYTc9sG zjR5b-&l8%3{dOnB4yh{(R0t@PQsEv5vN!GSY5y@ zflvQst(2_9TpE$P_{%;3EF>#`35UKY9A(HNxg!~5Tk5coK6YStUuU`CrF+nave6le zL`{fSN%)Zp;s)yciYP#G&nn+h=o5PYVOsVcVvz zR$}@Dncd?Pu^v?1&=MXN68BpQLs>TvR?NSpP_uOIt>WaVo=!Ht^K7T)`PH^uj?B&3 zO`%OcU&TNijgX@GW=QP>^ufaL4@%_Cwij*=DYNMWYGyBNpVFJ41~lLrv)_OVX$k28 z2S;-b@PsC|2RT?kmEkpw0qBL8{I(>38w?xF#bi4uWEPP1L&>#+DwNb#&v(0CQ^+7E zIz-P%=UwuNu*yr$?S>YBpM_T zE5O9<>f<9pJl)%5sk_aun~UChcLUx^4`wn#gpQ;Qtqo^xRQ8z^M#z)KSO9`K(NFmP1hAz4}JRM!Rh=|`8B zTnAb|R*3Lu-u+Y@4KUoe@kRkRKzsuUh3b{DlY}*&c6c~!?c~&&l;tiw52W#$A07_D zhZb5~cJspDHD`pb>e^D&^Ygk#?o&DV1@E*}n_M55$C_s&dz|Zjt)JLX32Wi(EnaW9 z`3_o09guq@P*TKKz+(!#>!GCC<~uSRQsRpbzyQmwd};?iYR>`I<_#@ z_{ezUy4_0g&RaUgWOZih9rVRbB|^`vdcjHwj2%3|sXEI`kh4nlxhe;7;h z%}gj8GhwN!t{+9p}+?4`>4YxMkp&VW&*yo}F28JQYo zg&n)wGdl)#PB_^LTfjy@^#&5uh*s}y?nvm_iVrk3fCgMpAL7K?|HeC@^v0$w%f?8_ zYMgI;0GRI8mz;!w_xlkrF9HD33(UzA+-eJ4QJW{@r2x<$$c(>PypxJ$!p$}(iuAw&77+FmXKz&X6Z%dLlWcS_J6>kS_SQ+l`3%{xK5M}Za!eHLKC5qM&_ac?*8 z$jcC_YS9@Esi40ZuOn+QX5HA+D_L<3ikwBwi0O&C8e)kM9=j1~2!z@rP$}CVs)%82 zh!oD(@4giuQLeaG6n=Bd)sDQv^f-yr>Vc^@H;TNb`6dOjO%I}H8*a$hxVG2_0w@A( zJROle9j#3}NS*nH*w%r3A#C(inV5vZi{uk;^Kg+*aoRS+>nt#U*SZ3=%6V~}uf=?2 zsEB!SIhK!kJMpd)nIVh28ZS_dc=ae#Ga`F_d@vQgZyaye?fG40of-p0DA_Im8=_4I z#=?94s>``%?MZ(v|ATOaS3XErvck~qM=uWdkG$^*FaWF17_HvuYF&(>Nm>;k72VYd zCbTbPc(*;>9haX@@vl6y(dIpuyOE||u zB(wNvZ~gJGl=8GJqzh669LWb@4yXTuu`*o8e7(*RUtX>6<^M-V3JL)*@o-`V(h zfZwr`@@ww7b+Q~Z^(JWQHh8wL^kbcPcdhWv*LWUngkGbk=ghQa-uH?Vzoeqe^XIs+ zz|w-kkKQ=YrO9nq3k>Jzl^@_-NtZ7_x5DUpzd(5?WG;{BFlK)-}*u6;UP zeX;g0Pdt)PUUB-Cg6n~k6kyK5lm&l!}ygSX_&Agtg6=vJ zyk+2&=t7(lo$mw`pBIbpi9RYwn)#L*9=H+Qj;sRO2wJ(fH^k#5)#LMb3lOZGkW>I~ zl&Eiw*(^O-Zj|1E{?v;PCoo;LZ(HW7e437tJra$g!s^1*+1yN0eAr*$z z*HB56_+kUVYG-|kif+gCS*yA_g%NxJG{{98ju9X=>;O!g7N1KXKUlbPClWxV;}7e8 z_IBFxT26hh*p-fvIGXk0A)CY*G85$IMurcW?7l+)ladH1e*X{2c`S09fTHCR-I1!n z6XNq~0E>}Kiw8Xa0wq)A!3z&Aby*NSDzp$QD}2cBUqN<&jSnAwx#@IC&l0KIp;F(k zJfigry|IGP*Qw*6C7Je)EhZYHNp=VHS|*3N`AGC+!fqnk;=D1Yvj+!UWSz_gQ)x|^3c;|;p zr1qnd2k&&>^Zr1^ZTOUimK}UI$8FnikZP+oK{cx3E6DQIpK20Gy6l>?gJwneUaz301opQpqU=f!UyY~xGA8*$tAKsIXy8j#Y6EB9`SKL!NtB6#?UHuX)`w# zAGb#<0GnPo5iR7XNqmt@BB_@mdAq=cs}45Ph@XZTT00?kV*hQyXR;2tWA&C(;$JoJTK!)H_Nt`*|P#Gn@d3M6{HH3 z5Jkw*SBC6pLgUte`!q{q4YDgK8ImSgXAV05Ny^MZ$BCWKyB0LXqR)<_LerJJjZsc+ z>1N44>RukXan!ND&3=hZ&&_x9G5=~#5Pn}HX;E0hXj-7`T1WJC0a285QV4G^Pr^0Q zo((rXjPPSfrPe;|&CS(|Gi|&s` z0#DUVVGwJ=dd!nweC=8HmrzQx$5CX)I#b3d$6%kDA^*4XqJzh6}Dm2IDC*h27YOVjNW+SgG+Pl;xs4J6$L| z0GO}~EKkfSyYr@m&gB%nNX$cWJt&+oRp7qF+t~%ZfVZ143L4lJKnLz8k^zh|xMam? zgr_FnfBtDXoe6&*JL(p_w_nW}*@P~i+ooX`XTJHua7wCK z{acMK_>^`S2Lw$szDb51kA*NEg5cRHiFHTy;OCZ}UCfoyJ-YXSz`ka=;khLZo$|0jFi-@ z_cuJiC_3e~W>weJpuq{9aBk_ir`*-ylK%g0JR8N}5{=#w`u>UVCL?niau;)*Ebjod zb-^|0_&C)^VH-9)7%3!0c{SM-J_&WfF~treBrQDtA~RzCd8v~!b1+<)l;SYV9|W^` zA~+ul<;UgSW+v2_OZ=0qW0Drik%@`aLWYSQNjf@!Kv152sPS=Q{Ha7@k`~H!AjF!T zgEH!UTpp^*Qzr*mnUH)zG4;d^1}2~nm`OL~qkHXj!_zlhQeX$t(lrIZ3nW2_&`zn; z>_8V;b~yJX1h%+1a{|ww$HvojQ1`7s02O)_s{BrwZks5KHx^UXIw`ni|AT=fz%q9Z z9?mBNn1eSC4(j)uO0%SX^ef3#hL3hmf6u=P{Ux8lbom&3?WTZwu86Vyy&4a}46t(l zB0}Y{GARi|+M5f&ATEnKVS`^tlzl_SXew&RU8F$d7}i?;1K@&aKETH=A~PsL?gktr zSArj) zLxs+c8O-cM8|S5)v(z|oth)?I&& z+aI$1QUCb>c#E;SOF9l^JuWg^;HS?6%B;zu5<@_)%*q{rk0T*Cp6) zA$p@EkC#*kR^y_^m36L5KSaR`8_YPe)A?D6O?j#x!L}R5jm)qF%fvy^dSg*q=zb zUkAyv)vPtRi6)JVG3&oUd?a|1b6o~#zmT{}Wo*S#;*0>kQ(24aoewzfq`3 z+@MS}g-ymj;(d@+p7}IU6TG{sHp9P^)$wpoM8?Yc&;^I&lXk^!rFJA#W}hw*{wxLX zz;Qfm4g_#PJ0APQ#|xcAWnxW?UO<$JRODxaf|zQ~?yB4*odDe}exR|jlMojG6U-Hi z3F%2DfY7*w=IEC2d*S;^^e(Svm_8yzpGaUjP!nXGk-%G4m7Ga9Bo$EkQ3| zV}dDzoeHbnLJIhyngFHAz1p5!iRyOWS3mUE;#|qEDuTVWWBksI5nddzl2y23@Pp-i z^@UUhI!!8o=2tOn3V{fb0LeCyiAvffZwirdRckrekM7Q;<=PXdtwU@Xs=zzkrhL)# zS|aerU(awpRZ2(uRH4GS2>~mgzvXQqe2iVP|XE8L3E2jP<6EI}~Ie$RPWG zTU_%d0UU2!?+_vTWG+Lu<52k;IQ zr<8HMD~f4>LmT^+zZaJ+B3C=M&e|PDKC7C;K6~_M5Goc>^?j~^*8r$$frD-r^@it* z^)*Zvj_KUxY}_^TQX6@C(fzn{y;Ef>Mqcf|jk;QJoZbpwB2w`pB#AcRo|+$%T_@;| z+ME5!tR!O5TLCw4G&}RT>kT_fk#LyA)WMJWivAK9&&QyjI}EuT?AE5RK+8^|RjEL) zJvuTDD_B-ST+vbJ-2PDkAm#Z=qUd^8pRGIGQTk47AO#b3;p9l)j~^s08Qbg565B6d zqJ!MYQ?jRm8)BH9^(7RX&I95ccv{(RST08~_)hOu4m+^ECSGA>;{e@N#xH<#!7~2)@yc^4`^uig=|7Y7l_h!n%_ zIS%f+ModE9jh%2p&Xo8BoTcGzS%rm5EL&u;BPeU&?vJt|p5DC-46Rv&^-fvojq-@?6&q>Y3Oi#gS9J?|1c}9izCJ51@O*{>lI>NLl zyf2R_*pZrE_H%?UFOq#;#qlTMA74eas|#NR`so=0<4;>voGt6Jc?c$$%H~j9>q>=^ z=qC#IRQHa$&!Fk`I)M`qdR9l8_BJ-1%9qlf6FJb!Ed z>}VkDPA@hOswA>#5OY;kKb`A4d?qcjqtOjZltP`>#ZKgGYPCtI(P_kPi|uEaBxB?2d@jHO7&{I-s5Sa5d-B2)OA6nAG(TfD zR&?%5xZB;Mw{c)WDxG2fdjjZpD9N$HOP$0yD{Qg;X{uU8lDR}uWsS+PxXx2Q9&$em zj{6@%$6QOo#KL%1MG1J4JPHNkoSe02AJ*Hj+m5aDnmHESF{jxn2#3T4EjeK8Esx_ zfdEkmz&+tsdjO`oU?#=oTzK09;n!&BEXV0-&eW}o-@#D~$;U}ZU8K~sxMM<-U{si& z2UuqcROD7bY|a@*64sV|Owo)HA4xJ{ZS@-DPZ+URNA^T9Ua%iD~$1l@lzx`u> zDmiF8e{RA>338Np@76x?Lgka)^%0FX>Mx?QUm=4zmZTeD-lXb7&iccwoJ=2;NHXm9 zCr#!H8r@>=L0`Jz-cP&ZP@N;hxq6>l#+6>YZdMb|n=a_3m>`lL1l_(xWldeQ3^B4s zLeQ6!7HA9B99k46G-um*{JFH~KF!+LNeN>QR184KG=rse@GVQtmc7hi!L9mHjK3#3 zuY|ha;$Y=9#RnYu^ouah0jH-X3Ue$zUS8)btxHw=fVq8zEIK9!6>(mL`tqPYAR}YP z&XYLM?J>DjP5yQ@dSs;|jtFfw=(2OYFUQVerWfydV(c#8=e4SOXXItbyBvrdD-dsS zYNuTwEVx!G1)r+I@7jsTdu1JD)ir49wk3*(Vx--DkI*bN>A*_MO4sZR=jSdbG-_Sj zpMlpZ&Ec18><2}uQ;333aE)|Gx{`?W+jKR-xRlelj+-UAV`ra1wkfWR?4ypGWgj!{ z7(2-xHYGi=q4P~eXlmkXOA#ogKT<|uf~@Zk?Y!Q&@fn{l(9?A=!`D`T)E%XBlxR9e zsD5KKDcN@#Jm_ASL=fUVOmh9?&D;W}-j?YWNLlC%x@x0&%FTJ_BnpQ9NOa(gQn^77 zP86FtU;dKlnJCrkw$J^&UU2NHsrSl|6#K>MO}XqYPa(ykDUS_Wf(_2ClO_v;>42t# z&Yghf^Ek7cp7lxlUO0?*hpWxXsD*p;srZ{)n4sm>tfi^7i=CvT`PSVE&rcFdmzvgV~izZr- z9FZL!8R0I}qD$d+99ea>-H{%Ie+$L|P5_V1vx&E$+&wn`ThhA7Yi>ZzX&#D~aTHU3 zQV@~$+9LBo6olMfu}@g z2t&{zC^;c94oA)gXN)uTb{4HQ^p!_W(aSE}JE5mNsWQh2$bEO0dD zjUWHAEt8iqMV5^^Gqx{&f*H$eV)w}_bIS;?VSTRkWl!qh;=?IHn$7bMA5kCKB`>n7 zWWPH`ljqe{7d@jJrh#UFDJaheUIDfN7KYTR!d1RkLA5 zyUqk#D&h3==dX-ZnXM-6lj7K`N(zjF@uOca&ctS{CS@{DV{1nw3ajnet7%}SDHQgO zh1DA(GVTV|beY!F)9a7lg&@P#!$<{hN!n_78n$ju&c$DJ65c3;0a70KLA)R!X+e8aw#LP#sG>NwM!@}z#rOU^Yx3s7zMsNF zlvF#~MGbn}ThF5JXiW8b%ZN`b)GUMbSN1Q!g21oPfz$7 zCU2c3DfE>pR-Nfyw%PiZFHZNgs(BKsIs0{fd_Rc;-PZj{bl_T5bMW`mVV1u3%VO@!hi*QZY-g=O(ru zaC_o}*`_OTODi^cE_^gJ(4KK`gVAX3rb$lqps?EV9jmFw`BUok>AMvdy7K|?>BPC9 zDjB98#Q0Y`AB1;eqJ@*yVS%7G&wir8LZj z8kDoFcSUvFdk3@7=dL9~c@ASNY{5#6+=!afs3TH%6R|j}@h=cD z_P=ct$n|@gx4-dXc|+EzIFh*+mt9{DZb^}PiqY+k?}<6JYfL!F4)G~97P{s34tZgcc27GCDtfrTEr=dZ7*=zA65+HV26 zD{^;V7K(Z}apdmQQlPCjv094bX;9ZkvZ1iKPmtx0qe(v+!O9Xgb%fYAks>jJ4FKUg6!%lC0?rN?M4}-t1`@f2p_M7EX z2o2?}jMzU~m6O{Xq+oA|?B;L|=o&863_L&$Zb-9{g2xj5Q9RzH>@C;%)3bhO<>@U&^X1gG{$zYrW&j8-Io=#Jo{ zu`$>y=>9u|z=8BP7zGaf^HvH#=Up@$({v67^;cRM%^{(Ua7{7elH00KGZBzPJoF zDv2rru-S(h5N8SaUx1f$eX7P!DS6&WrF6E}KAbf|q%7itYL+Ui-(06wntB?3n*$ z4HPK%_vYp!heKz}v{9l9J|7B}9Pp{BpT4dLofuzXue9LKF`ev#g2vyJ@t>f;FR^){ z#fK+5%whPc4XSlX;L-WvRGq5(1%~tIU}s7KaU9+MY-#V+f(cyy^TSe#y%_N&xsesc zDd(TSdM@pbGSQ}Ohoe~^*Ws2&m7H^s)=1l-+8~C9=>Y$$INHS7sMG77F!HB!9H~t3 zK9tNu1fGd18}FQ=e^YaddU23{%*G9y0tcBoP=rNAhdIgW4>b@&CBC|zCmhrYd)7Vq zsg0_iCHp;}YhMH3e*ojx3~-6U@(1OeA|#KX5)=%>iBk?&;)t8_3%PHC&d5-azML^X zxW}fm4=8N4|F749oHCCV;7#z3B|c}hJ<|_Z`}7g*2Q@{2V$Aw219OyqW+yVf`$0Be zro_K2oib*37NDs3K9v6J5r^~|-?{W9BE6Scg=VK=%7rVsi9hrObNB5VYK&xw!Og6qP zMM$j?2(UTlZU3w9^2|PC6a1#I+@LcOGC(KZ&%c@pN^`$%{}I-t{q*qU_Q(pvz(iG0 z<5)i^?_5?QWZ&UXo_*WL{I}V)#XI}*nKEuSAAAii>9%0Gs+u|$2 z|BQulB-SG`5re?ZQb~#nYORZmO@iE$dT)#Y*@!&#a8!iMhb+?XBL4~;il|F(QSR`l z@(h7izAH0tZG!3paX~T>?r<~NyG#g|59}aQmh_SLf(D&c0sctwevBofG-NX5pTN1K z{A(3s7A{d`W_ZCS1LjIPH@COmkF&)|?%Hp;54#FxV~^DSOK|q0Pf6yyS+7B@@Q{@2 zOnCcZrsDU~6H^~|9J;z4h7}vu9bMlsBWFp0t(yc~dBY`tpo0A3%?=pv6 zFO%nzP-JCE{6pvI{)|knU|-5!T||@^6>VsuT#*0p&s}?POwvE}DuTd-@!R##28p?J zsb88#b#i}RVTGdhdN{K6ZVJl8Ftz8f=cr*cW*+|s+NJ$gydthCIgG9Bf=;kqu4S8Z zFkNrTwZ#t)OePB*5VH49g;NaKKq0r~pCNZhTAR~X75jWMDfFHxS>OgiC}i0PAH{wO z#UV!`L%<^o_y2*>@+tD*THhyJIj2v|sL z=0OdwNrW#>xe(RnTha7IY_9&^y|MReTgPjI1}NewO5K0q*qQj;F(YLJkH2%G*60R+ zPpa5N<-p`qKmm2`{M`mcPF~_DcBVHV%#B+{eppD_FQJ*u&->|*WOk-eGL@R>>Sd3@ zmNTCn4MW`PXrMHiq3AB=vSg$FYH&z&<*^qmL2hfi{~=~NLr0@G{nZRMc7&p#2Qu|m z%saj;e06Bx?}bjS`yfTL6SHrx##Wb9;;N68k6j?Ta?OjI!t;(x4l49Mx22BU`m$t2 zV&hZ->(#k&(tPabUSH0&*^=T5FT)yb+|rqc_)?n$^E;g=?yx2sb9$<)Zp(!*8)>9m z%13_-LCXj)BE;<1eBzG+tJVZH(_^DYYRW>^Gjq(ACQjj_F-3Qb z3VmaH;ww$PaP)*1?$Q*=rjF@`aB7=~ktWY&s)ZSHN4T!G{U1XcmU+{UD3AWaFFL)S zWavLvzJf58@b`@KUtitagE2{ta+4E#PWRK(&tl@^-dVoU%gA`~)tfF-HP321TCVeZ z`|qga>+#{gxF$_>hh@#t_@CQ8FvmlgEX%zh7h$o?!~EzdFZpZ4^G+8#oKM>w>z780 zGCw@&(@zs#l9xQk8L7-zzhieiKe4)PvnX(tRMi;Y>)n&8K!v3>u{VdjZxLw=tOPZL zF|S@AEH9=GIOwK#b(a@mvS|GgtABU+3_o)PkB7*i9w#b}aK~u%9}aDk#4VmoX(=o< zVV%qQ)$4N^_PQy!7g6S0$?4R6^LrTGsb^}zw7rPbtrN^YE$*9q3e&L6l)IppYkl%y z&{dC0O--zE9=8)?E&Da?Ew?3`i~g5cK^X$L&ZfA5v9Z5)lyv9!+Sn9U|9JbxiYx=} zR+8FL4a~A;TnyaV=U7GkyrbdTW)#>9JnA$r!+x9u1su?aSJ{p22Lqn(Zs3;AG2_a{ z*0Tg^Zj-(Ya@;d`=6hf?BAE{OBWSzgW?;BI3-%-W5Z%pi{Kd^$j?IRX8>sLPn$%>C zW(m_-B@O1~VArCQRfl(%l)a|#)sl> zKy-+YL3@Qeb^tiWlQlDImFGA_JtB5n;ME8dQ1I~iqpcyKHD$VhTe+)b**s3;+<{rJ z{ajuKy@~H*L3Oq;pNjIUAWo;Ip{dqHhsw+NPg}H8`<;-(nmZ`o0rtwVvARsv3-zX(dh@3VzrFtNMH=VJLQrH zgj@pwW6Pme;Wy8q_z*#6h{+CI!O#lZUkUQgN%#a>CrV+3zf37%ruhY{8H*cswp80T zf`sy2w5=2`hZqZSyhOMRzRwQE{WS(pUG~z2&h^M$)6xx}Q8`}kfGX_;0R)z3dk7Kl zfnL#e>R<~(y(RT(f%S8rmN$)LxLLQZx>;r1q8=X_wE=R>=pqEj&<=$zxL?U10cc>- z{4a&9?x}2-ALcAXl7>Gel6pUwc@3+p9me*A<>N(VV*@;88+c8alLs&QvX>T-PA}kI ze#Fuia6ck?jiac=;Cf>fmda~&c2>v`wcE;hh%v=QAsdL1Ekl@(Rs?`55 zy3e0p{`cU=dTYtBCwjp2jz1jssGucUfwfOKhnn7ahMe~%)fJi)@XVEAQNVvL@1-FD zcGJh_01JV~dar5qrKRwM6;X!l z?g8{(l0uC)>%&EoR8wlG2&2;Z%6zKu{;rm#x)rEM{Xk)dwSlQSq$(5X)l5HsX(hfx zPUl+fdHG5FHGl&DC@eqT`{iIr+KbJ&xN#PD?sh{B7mR<0RO2#Wh?$+0fRpw!cFrR3 z48V7MR|Z}(L-zCuVJB*ZH}I?K!-TEdRW%#QP~-S%Jpz}yf~`5EeDUY;80I)^6!a`;C7R%8n#}X6Fu!D zWs{(@W>zP(PVG5vPBgS#BQk+v(kQtpdZrv}QDk|KUC)uV4!`1h3BxqKDyZG#a4Jap zL#Se=tZ7 z#$MEeO1Z+^IjB9KLG8VNr4y>CP(H?B`TjK+V*cmn|4a6V8P$zakyaQ)Y1y(WeK-8KI!AN8MnlTPN%&$XVvl2rd+V==d*{}$h{)9X0@ zH6>8Pm$)7}DD!L}X_wSP_$vdY^`9c^4;JS4yu1Gw^NRV%|6{ONnA3{?3%&pU5z%Y% z_DWh+{fCeEqe}dj&HO7)&i{cEF`w{%Z06r2bK4)<@t5MCHT0jG2?pRbWB-+>`O8{G z^Z=;`zf-9Ap+5{h2rTb)s9N8@mu(sMOYONvd7Jay5`Xxoo*G5NF==<-IuK&mW*(8``~U z;Sb3ZEO0PToE>Msms{Q&N^*0p8n%?#x%{LO$ab9nVuf`YC$mkYUmWD=gz346_Q0u~ z?hLKTzvwd2pq4V{kGFf(&&^i~=l@R!D4L*w@?gY)rQ*FcBW25Ljg{$mVd~80XuaXb z*|*)qf4xGX9 z+;wZqkqx;QQ>{x+K|YZg$)<{Lq~(IpTLG;j+-xm-E-N3~L*^Q+WF=PW@wO*DMG%jF z?c&*NOfrH#q4PV@m;D=iEjqW4ux`y0J!b0Ep~`oqp<8{C>iyCbknGH61es=?9Dz=Y z@-nX0TTOa}#?b6a4I()Nx&KV&|A%-Z2#J+Hd`;bd>YEDMLr@CN#7uEelzkBHe4586 zk}Vfsdig$cSs=eC zI+@QnVJz`MFVgQWJXS%F`SW(}6cn@1mwQVEy;Vk=^7fPRym{TEX7Q^j?-H>iv;YmG z&)9j>h5bpnAHhtP1;dQ=$xeF6NZTa?FIe<_Wz>WQV%4Q9<6^zXhVapS(XV-@hArEu zD#4VfE@46#BUPx1qv=NZ2?FP^Kn>$LBm@9|kt1aiD*t;KVoCDzRcO6c(0e(qsF{oj zQ#RT>8K`hC*a22K$PZD{k?9iGn!IIzW+(ho%$gn5qPds2l^s6e*)I7g5fCv)%3a8B z%V@vfreM+#4VUy;4h)qF#?-7N03@-M27>&;)NdXCp`e#tJh7n@b;MuI3 z$Et4x;MtU?J!xgjm9TMu^RILkSWWAC8z&lark|WdK*sMQ>r>=jF$zmSgJ(dQWyh-I zYiuflnk)yJJ-IhJ*FjHdK%L5ogeT%;MviJ4F>dipE92%4Z$MXH?A%L1mBxfB7icE% zZT&+QTvS@dD#n1u-8b-!>ocda1d^iuw{d82HH*&=ytCX_Y65O)F}{WpJKn8PM`PR{ zj0o69N??XSXRsdc3coulbY@nLYWd1INMN`DXA4bHO=(QwD2o_3)RXYzTp-zV5*lN& zRSuLyj&`klrCJ>oSCo`7`Zc~T+C^r<@DhkIByH zlM#n|pWiE)RhLg9v%v^ijtqJquf`Gd$wGU&7F_l9l5-RV-afV#i#zn#rSd3g?pptEML8KbsTx(fTr04-qD6W^1dDPt%xM(_Ru8Ug6EVCDE#L@889l zg;H=xXvRwrO=z8<7FWG50Ryy5RJ*W|qP4)@N8R6lpV>!$+A<*I$zw}i#PJk^_mq>T zHz<-IA%9>5b+ttkw!^I@ip+AS)!8%?sLfOH;j$kDS(Av~DC<&AiREL53-javV2u z{!%T_#$w|}O#1-FuL=Gnf$Z7NqW<^-*d=c_9h(1iph#*Y_%o8_!u(70X>Hck;!#Oe zBf98mGm7O|{we*+y*0SXEStbA1Q(Pa7~~We9N;+e%jM+Nv_e128+EG)uiU~Fa&4o- z&udz4Kn2Kg#dj5&Eu~`_;zw!9d|Ox4GD!y9C>X`x^tvSz_HqeRn>is=sZ8G{D=0jz zf3;?k@ce0if>!BEC5!Hsz&X>QC-KYU5_M%^H_Xtia0I{kMmkL?9;EpVrj54?T~ZP^4jQo63Ea@* zifHsG%iAjr?tLIt+EUq7|K{+u!`xD_7S@P9us1OCNpF|pqBuqjhB&kM`56ZR3i=&2#24c4F9KE(jL2jVOizw` zB%7=V{POZ?+2Imjj#t9!1(n_tTxQXR7yz9rwOmUU!0O<3NtN(TrhD8Df4sBh8|TJO z%g=26oP9_U#3uT+W_%}emy%|eW@CDA#nD|Eqtc9pU%mHQhmIx%js99l@@!UqMFYZp zBeA8Q5#Mt_Pdq8hoC%wS05zUgP^ksHq1AYFOPYGrQ+zaGfz?6jou=7!C$UDaS{sLN z^nr!__aKHdwdaDbBkqvH-qFYxsBg!nGcjwssAKfKWcZfUNH`w`=U;#1hms(^VQrPkF3xIH)p`G*mE6Fh(OL;tSz&R@n?UYe#8k%=qy(iWyp^mPG0c>>u9$DSAwNH>g&G_ln$7Q96^9pbg z0^|byrothcrJj-8>uqh|su|Xln&EGcuAF$S&6Y%cVk8N=3E z-g$gVI-88z-D>;>k9O(Zed8 zeP|@~-X7mS5NK}()2?{Po<5}fMe>sMxlvyJN}PIb@oh+dr1n~Ve|&+>n-HVw>&pMb z-CIS))dX#$gS)#2cXxLP?!hIvySo!S1Pku&u7eZYg1fuB!U18-oBAStNq~A&tCM`WlEreg1exly-YXRt3_fo3gDoW_ZjZv1LJOmeSyw~H2I<;kWg zn3gE}Ib2UzsUD+iAj^7j8UAXNV4khNr4}SSVO+Sn*tNtlj~tfitHJHGZ^<;k-56>rZ zsj@ZR(M+j?7pAyJ){e%!yvz8HO#Iwy*Li8%gqr!ySjW+2!nbks!}f z7!51|z{H@W)7&gg_3+REKjuV`P9Yhf;HVJJ$8H& zKhQ;nEBSrQP-K*v6RDxif0)HRQK-$@4jef5JvvPpdMt4`Rj|%Z(!bX@uMYy@FXM z?O?p!f!PyLJEi5ZnVLeUn?{hiy+)p)!$x2^o|Ak#u>kUX)2ej z9I+)tnOp$tCsOo>IpO_=46Us+J~y}_9EIo%QQoEeDW|3G&rO{yZpHhDnKnUCPJTy6 zlWhqctWJ`XFQf#(vg!FEG6@Ucuq(^5WjsRZQ zzAF0MhHT;VZf^Dh{k%!1%q-KPD5rHAsutapVkt}cks3oE)#kPzu>v!u*}tfW!At{h zn82aX3qQW{4(0#yo0fq;rV^i9Y4?+j2NdSizlYN$N%~xC2oTKo9>~f`Hje+(`I9!G zR?j=f8*2Cw31c#*joQwKI-7u+6?@a(q_r#<5O7yGx>ioOy{x&2R9F1${bZy% zDb`H~TzU8}i3U@!{fzQu-Dupn0(#)O2BHyK`6A$)RDv}|&CKTeZ(=Y{i>tH_lobWA zK~|B(M0_i901{}QCn;!n5wD*64M&0GhaKtOHg;6qnTO=qX$d9QjPEN`<9BEBG-J4P zAtq6c<)@$)EH-JUG6sHxO>|FJLp!eM}yCzLQFn?BrAA*?COx({R!1V=Sug+Gmq zQ#Rnn%ioZBNBXBKBqX61T-m_U59LguXkwYX^?~GTb~6GC#{PK>GHT(fy6WSd?_W{% zq*3dhn?1AazpPL^r;c}mP#`z`=a(*9iE=aq;au#J7f*^;71p zrglzX`m6;v5{_HRu4DIc7A%o=(si%vdtBb7pi>T#$ksJR;PTeW%z%`uCf!gkWb1wA z{Tvf%yt`#44Ep;$)+VLJNYS8q(qMB?XV8kNBpl|VJ0ES*0OMsyR6%e;sX0nxSYNs# zwMOmZ-NZKly!K1(ce826mtzzIg%A5uAF<(Pli5|4{V3%Kpu zpS@DUKOZ!#L*-uae%B(sq`9j{Ku^4n1S+Keoi$j)3Y=`0kzp)o78CKmpedz*XO72r z7vYbb(}FH-T{ko$2ZS-GDSxsZL7TZ|DvS^$7HR_E+hQc#-q8nlr^aP5{iOi-Ad*2| z0A)nYLy+k0(;(+~15JTg=dU52)D7J(ucDuqwNiAN_!Gl!-&U}QCk(JwnCCgn3=Y*^ z;A^I`2bhayctKH^V{Y2dtAzSOVLvy6tT2$TrDFwdf#H#_R09YdP(bp@yRW7R0()Wc zvbFvP1T2o!_0PX@$H4|q~&WXCIM8*jgl$JsTU8Nic9d47dUCVMu_sIMhXEj4!{4m)aq5n+XE|frBBN(Z zGy!5en2mUrwfs>9nFfdTeg7K^8h}agc3*PiYIKMbY6lCe%dS1q`640kJZiwzFUbJR zvZPuGGw_IN+dIN&9%6nt5?@0SL4@B+OzoUG&i_$R6}G3ay$eETKMMVKVO6H$n?ES- zfYqOmjvU08a*PVQ;lcl@LY9(3==tu*MrsH4w8+@XIsc+jr2J8T#3$(%M*QhN+l2}4 z_>*ckKV4fz9r8pE+%Kxe9iz(sOxA#F6k8VzmIXfp z#lbw!)mQL{V0O0`Y){<)Va|6FQ|lvuH#Q(%;KSDhuqFBLxt$j@hFI!tjOh}1KpFb^ z`qX!i!;Up_02o((y2=p&I99*JyoB=yH3$qiZ}8%~sN;8zst@yjj=`!g{{zXZA3uY@ zucU-pJ(--!9jwm>bDU6d>=IeCJHT|#g9noCEC4HIr-|$UIZN7kO!(2$Sh0NNISfnW z94!Xshze- zj1Q2xKh*-})^OH$WU`DqN8i*yFq%0_&m2`ysaPi&_8%JvkLcf}0<3>Arz7mn^4|4+ z!(OvCo5j_}&zlrI;>x<`XD)F->UmUFK`ylY{Z|LVkqO==5P4Rf-%CGP%_{dCUW9b) zD@Q3=R9@PLQPCTHzV@dC2rLfrhaFub7uv@io9x0z@xUxuaEZ~_oZ)M^*M7r9%;Lsm zbs;;=+nvsY)9~@PPb98PMcq|`!PeyGXUY?6t%FoXmRdu(#t^@t)KoaqormfFgUitN zhtF-_cKv-r(o{s1dHiTPdarrm&d2aA&BNiT5JVt=umjlNWlq`SD0b2#V zY_ha88V8&^`rdYh*t;qa63cjn{pE7{Go964)DE5dbJn^MSBa8Ehq>rsdL$bZm_!~d z&rFOcv%q}nf++h>Wf1tpidBffHO@%erj-0oYrGbS>i`@ERgmc#io_EfL`U_-)1F%Blnjjn7D}0{hvQ$HlSL+=)Ix&H>5@; z0C0T7MuR4^9`ao%UBjH)C&i$OcU>6Nra9~>grC}S0Bdn%D+Nll#k!5kS?VhY9iXO6 z;CvjEVlE1YIAbxD9|npO_*;xKvS|a9P0&Mn+*!?(g#tslh<4L{GG;^8-!;fRHtU`2R*0(_ntdq>nL;W67*)Y= zuO*5G<|;r?vhanK=N%k%i+>M9hh&%rZ!@VhS^h-|mo0Nh834XEjKD4QRZsrWl#nnw zEV|Co3vyUEIVedJ(;M0~@sI0MX@)k8Tol`ou+H0o*M=1=1Gel3&`N@?HbdeOAK|MO=g{Jlud z4)csEQKd&_6aeJ!FnC2VpX*uLs1s>;`cX_Z z5RU_ZcpMG&4@S`5B@r$y+tY45seqCP5#QZ`+ebFEWTs%_qnZNWQtQXpq~}t>*|}tw z5xkQ}j6kQ-Y#j73x1frL)8|Ie>ud&VjU8k>T<&01g_36D@%Zx32rfRC!P<7#VIz)W=mA41(UgsjEu-bDst*TwdH6t6V*Q3*?RU7OG zrYlCk9V@B;YeCME1>BCmj`B2>|30cL0MH4<@?diY8ofF|ER7BOo;lJUwS1LO!^>w; zLE7c{W24sqeO7tK`3M~fQ;GAQLvkH700C}V3+U8sfEKaU{&5L)t+Zyv{CB+mcToDA zREbqpzi)9aNOR1}v_hg?v0f1^pj!kAIxpTa8MkI+gur8j9MP=pmU!8BTKVk^0fa5s zb7aEpVTtHV)f>%=l@i)8g`^;IP@tmD6O|Wgud?jVl@IEd1CJM%acO-&J5{XW;{S;OH zdl(}gotdzd=ihv&nM+tTlhIk#RR$b_+MCBh`ck8a`wz?sLn9~vmkX2F$|;QlIh<+< zhDscdq4Ux2XN>8N?Zh}jrOI_M!@}~?d6gYTZO{|3tkyl_#2GWEbG?s{sFOs5 zvT|ec;Zz)&DNA~$#KOvRg=z8$8r4#}Xx50_DCEyi3Pho%!{Aprbzr>l zBFXdx_v#Dy@TJW6>eYzw^noXbB7Rnt7I+20GWSnhdw*e?rh>!gB{M{vLan&B@s%{( z4o#retfCo|@?;06Y3WYbveX}G)#G6AC!^`OM1isesnG=$bAN|GN3sZ){sMwhoxOF` z_rCa%Up0)*)uG}0Qd7^lr=dqK0-Q;`<=-z`x`O*wa#Lyk^OqwmaXsMeP@-TBs6q{I zjnbPg#n?oyz71CTm35Ak0AJpx$=JX-e8V3TJ=_~h9OEXBoQF23c;!bTG^-9#<&f~~ z^qlPx{(fT?)aWaCxj1WVRwS8o0qg$ssl*W*&2uve(^AO%=|-D0$pyTd2ZE`QeRIRbzEg_NLHI-7-vR)9Qh;Z^#gr`m40ZJ`rThH-$i<>VK$2 zCEyO~ga z8O?}Ei3TGXKR~1;gw*BdG1R}r2$wBp=-a=XEM&WGrw;ocUkBA6Rv?g^3Usa4Xc+Kc zq5v%}^^*cn@>}`e!@WO${70pXhH3uy<`yF2z(0Ns5_n73_wcjz>x&8b? zuw~k3G^#|03-+7!T+bZ?+NFlT5o}F9BgtyS7JT}y`|qNV_ivUp9PTgCa3NNQpVGgD zj>)CWF#_#-Fhl~MYm}NA<`A#g^S&mhh!A6ht6M=qrn3~{<-B?D<)xvK14slmdXzE3 zBNX|+aA0hn#)Crz>IBBaNfw{anp1%-O~g1{MB_}Okjq|5L*;hC_4^%zLuRlPMk26` zhHlUFfqY?SQN}9;C=Nplv@(I*df!bXAqT_)d>uU@`sHzn`J@0sz98@Ut&CFuIx?;& z{8+(R{6N(aYBvUzb_A2kU=h#(Ii>cT0T1jq7f6x_5d}$Ka!&zWAs`3L{(*x zO;E$1wZ|m#9V0_M-d^OX`%Ng07<7$j+PJJUz&u_-^_kY>ke&GM@w#&sR@n9gyx|gY zoaFwqVyNi@0V>-cc5`rTmo`3IM7v^&t#0J$QZ0QTBK z`<;8b{PUhCc{k>b7%EUjgvlkowl~7kZ@7z$&#AhTI5n0>y_F$+qcI33Ir>24 z-{Ttr5&&gd3?eM{FiZ+ zjny4p(Si73k&yA|9IL3kLT$XRdZD_ma((N$kY4mDxyRvO3e`QVXPN5f;VJr%WVeK? zXbAQ$wyvFW<|D<3Q6qItDX?(%B%Kw{UuJc$O4Ipig4(3VDKQEq z`_Y_1Q(mQMhuV#v)S8vx@WM9_q$1taf4nA3J}GvY#T1Hw&(nKx$5|AiwNa)BDEn;ltXe&70@WGGE!H!X#eqn zoag1D3f|I5iCbR+_@!ok!p#r?;3B+{wPSc1{Bpy3xeGXLJ^G^k2eB{bWpm%<=t*Li zO&}N+MN@WrJ(oh9`%tFBH#3&X=THyxey|EM@4SCGnfJwrx+S5*)__d^nntp!ZV(qZ z)NkPDW(jRTiHAB6yjK?&x}TiCN2ll{tVowpwv6;yOy;sNjZ%ZBK)c)m4J3mFZG@>y zSN!c5sF@SR33nzlQLP^hsag{*98$(r$XHSH#AEC`*+S+J-^sxs)*R&JNMgWDf&gd3 z6_2jOwOe)?)XaTy94YZYE9xLX;C?+R9*dofD;rx4)bSZQ)x9K&r){l3$lbZOsAaD%~(l z1=H|}j(jR~10#nu=|m&R|B37L7j+>1#29;wj-Yqlw)1cDKPGZc<(O_3r2G7_e}+ns zmB|DA@m*n!$r!((l-znLoj$#BhO+^q*Z`a6SrFi*GnE4UW$O*pmj-l})-L_R?Kr{| zx`&P@Tj>=Z*$jxZfsX@(o>N5^OI-~7hI0}}r0?Btl07l9w#A!M8B0H3LMZw_{gpd% zX(`7%XOL^sA+E1%wOECTECmX93u<1{VetqHGhbt?#~6GG81p0sxmsK2Agx`?k(l8g0RF|zwDhE>Icuj>iALrk?*!d~Wo;7gvyouA+wM1oda zv-{BhKnBDA_qwFdH$A>+k@Ts>UpxrLOtXlp!mP9c7`%#SwM@Fm5Q1+PE3{S!{ zq#8=U6hF>`J0IHJqCXi{+TNxmOtl)^^=7J1!~s`>KQOh7%^E&gxWf3=XlI-(+qHNp zL;Zri56T+ADfVuq$HiS>_36w#nGxX{9-ss3vDM{7!-=r9T+}LDpfhj8+-o!{m3MV0x_`Ah~K8tqRPUYnUNS|>^A8LeZ(Ri}bXDbtkP zv(JbpKG&n+`lQ60@5I{{z6B@wTMKb&3{gio18UTTTj-TWFc+5I^CtOGyR!wh&)J(T z0fF2r)!T-G= zGmf$}e51gj5qt$tFx+p%;2(>OM1&vkKp*b|4c#!E2(z+nUfI;RwAdKDTPir9f+Q8IQ?dds{i{Q9Z(i5YCAGMR{GM5E?hpZ*ELa`)!jph8{Oz&i zv`un+45bBZdk_Bl;Q&XHx?W9?D zh7)w20h4dS)LJN7Oab!MkLk|fhvjgMD@{D?3O3#pe1D?LHY@NQK#S*MMSTK5G}VNM zBkD%Z=N3mCr+e@P(O~yW(j(f3XYf7!n8a!^yBwbf0NJ=? zt}S;{3Ss5i*T+m1Ea{lpRR!8T+2(pzJq=qkvlJ}SBUK&nt9jTi^PaQATX6*Nv6dl9 z3r@Oj6~3lZq98#*yhHi^lj@mL}1TB>1!C(-lcNFZXECJ-=8yK&fKJ@5jXbM;4J=lRoT^3cv9z|!STss z#bC$!pxKhHpC&B@mCFSPo^}R*e3^ayb7x1NeLWj~&c~i~TjJV?kDyH(iccZ83%NCthL zds#Q1PNDm%$S1y^(6l52v^aEJK5F2y-7aY9WM;{Q-eYse?(iKiBn|?v9TGT}u zsF!LfO5zty{@v8{vjAxTF|DyPdP^C1B!WZ|EeydM-z}Bba=??ohei{5xn&Z+Y{DhYZ}pR2e8O*ZZi#j9d$7<`-ZBu z52)JAQYr{pIGy*ZEFHB%s5V|Zt^#m7pg)+>X#V(pOD>z{_A2RjqY$&ja8F&iA5Lfu zl+e%9dAxa(8j;#$IW!eMb=R=4nxlg0Ybjo;Mgy!wtR6=qlH0DU_xz3DuP%#C1fh{8 z%_TX&gTwf%U#1eq`>Bq5ai(<%#+)6_A1FGeNVlXruPpMR61m^%J&_2FhPMc(#&X&k z=4~lTdrgi4JGi$K-~3gT-)!4sm`0%5{X27xxq{`MCg}RnoSVOHC8hJor}4@?z$Md& zISv`0eNE{TZYp~33Dzr~ET>0&twR-gllMlZouu^kua7U8T&{-L7A_$?9GO4hvY4L+ zBaj-?1(t>AzJ9~NR|)uT@vscv6S{nNsT*2xxBP%G1JVWArJSIr*dAruqY}4XO~|SsC4jDrTKuXKZ0CWd5_k>|8UTZx`7%)xQ!sqy^b1_vd=t8 zXMQ;q88l3d4La0T@p+Scc#DGVU(`T!MgEmr{crG5i}#&&B#$TVFR8uUaZ${`td(}p zII(u9t(KFN$P4;-EJkg0Kj?Rpaqm;KrWYf8`d~l*r0`m_nsZ!7op=Yq0|z zT9)F9B-ZuF z8B;t+dqQhRwyYuMqm|kMV{Ei*GRFFRN+JC0GY#vJ0rIr1nad(~+)n5(i*)ozl@(?+ zo`kLrQ2Y2q%9_dAwFa=Znrx9hJi#dS_JJ{GMh^fk&XYsSLDfly5Ylortd+@1sN+CC zR;#h`H|gbqwViO7&N+lt<;1-IO{TqfMjty84lpgUg)*9SLqe_A7{I<~Z))4_x}QUH zmhwiGnIui1@fzZ_>)&6J7rLRa0oN$86(6<8zu#(Wi}_<}!yS0E9^{=xO%Uoh0o#L! z*T@T;Pb{tiw#t8+<;<9URVPxy-;pODUPv09(HLI|fQF&J?afn+F;yS24&p7tKadNt z<>4+~|J6=7wl=$VPsuqnr=S7wt}N{(rIwBMxKEnk9owIEEjQZco2P!?I!e1e#tzJR z+h|Zgb*mS%)3f{4%f!N3oLJD*=F&%d77khtpoj9+2{b*|nwK0Lf8#*fEkXj! z^8_#ar$=v$9Yk3YR_Y3IHpiD*Z9xOMXYY{hO{~#IsNBb$gHGJ=Pq&nJw)k1v5>i=p zLcPpxCmVbvUtbLA`1yNMj&y_giqM#Nvn3FUr#PIj)r;C2A9Z}feA^%gtQ%B}3t0p^ zG^iq{64$h~GS?^#{t|7hNXt3wENlPqVS~?3?$v%9_YzZ)oQMleyk;g1Xs@d3WkrA* zkezTYUDE5d5c{64YN3wsK%2^?=AyFh1GN<=C10JMg$r7moJ9ood;fk#FFe>dwwPDd zmus)sTm-=U*Zf*dxzni27t^;3setxo06pAaMfh0eCXQ;H=}-}oJnq6&bQG*iVqx(V zJz6bPtKd-H&i<(-^EJ6AnX)z90mviu&N?seGakryLq@7FA{027vuK>5cSohfHS8e` zNI&*M@n-pT9+K)=P1t5!Hg`pDjYJ|-Hn|`4cAGKN%r*G!?{+L0q0<{uJ7QY#_~E%m zWYJ#r1A~4F$4Gke&t>w`b!O@s6{i!PuFiDo**{cvV7LM$9MqXKT~|#&^$WiSSZg_=@U+t%-RM+l0y}-LV1`A zFBG`FOOi2?awt^cG!Omjn1Y#+TXzQx$fTEy zDW7bq`j4Y8LBZ}A`e^yHl zXy`PE$k4ipDoQs2bpYMxf$z_3F5qyTiesQtE))-rt)U0rCjQ*z-TyJLZQX*(y^rAl zX9?K}K)XaSG`B*lr%j3b-@dZ1>G|63D8=J zPgKkJVdv!No3rqxoP-mI%S0ye|Pu9q9hJue(z z^hdH>yZzY1*%~08W6NhMcuS3o&AH!R`#ea%p2X-=_NZCHpI@=-53><_Zp5Qpuv5MN z1>SipF634kcr?THL(nzg-OD`ZL4ohli!nQ9R2=ag<;|2DV*0&_?#!e2l1r{>9L)rS zF>~ebg4e8w^CYwQW;rI0dcJz+1(1chDrEd&(yI;JRupPCG*;^(d%NP45H3-jjffn&c$ zW1_#jdYON4gt&~>RiU|%3I5m~DcD?11J4vx0fes|P)=}wx`a*ChJ+QM0I$T@K1tnE zXWB~!nwu%d{m6O|NxwT3HBx(1b^=mJEMPoWD0Z=Uo`ETxGU&txrA{Qda_Vy_dr}a= zIRP%`)3}%bG|)1x+h>^+hD4w`$Jj2K%=A68Xseds&6}M}?azS_I7|Bx5!jQ{4m8vP zE6_Fjp-JyY-sH6KPg#EF%)jIS?!CeFhqY`@%-^Vrgjmp@!!_l}a2K415agb{Aq>Lo z=MiNxY?xaMx3=zo=;Nqze2rh*O~IA~)nG-4hAhO%yU24dS^wby@?*^H8OwCx0fvBC?7-zzPV;Z1H+tVMJ_a8v!D#4^Y5wrmx5Q8jbtM$I zPW5pAY1mSavI6HPVBLunF~+O)L7`_}f$y=UuehzCCvDMAbiUS$h?%EX z&svS~Yn}gL_SU?enKCq4EU?}a&l7(3J`xAG)502bq81zEljN5`WJu={Wk-wYWc2chD_B*&;0WBbW_jA*)7WwTdj=?7Gcjr$DJ%>n^ zZEtvCu`(~P*=S@wuQ$Mcvc<#?KJB6af$qrUo59VoFn16hFgkHA5mV9E(&*_XNIwac z{=-y+$Sd1-f@Upi#zC`@nC6&Bier381Db@z1|an0A3GF5pQaIhR@bQ8l$=GEZoD-U z-z<*c5;V%ZR(SGYguxmDv562S6?&VBfd=QBa^+hbBf0`UeL3j-wl9CY`X;Tk5*N>8 zm}}yeX;RAh0Fc#==fOWl3w&5ING6P~a83*sG{*IC`W>>Vlw#6=n@v*lc{Arxu=LaU zy#Fb!^09|K4mv})$?yCa?8$#Ag&Y&#Gpskj)#HlC?0pOF+s>*Dj04dRS%)0y*Wr5w zj+z5Xj_mJpDP#@u5!;*CeW2d@mVe%h4)KtE{LM}&L(2o( zYo6e~0%anLBvFM+qrS{ghY#j1#F-AWySIJ<0pv$Z%HeJZJNSaa!bTQ2TWbiiF8C!; zaA&SD3WWX+@@11?or)sQsZMT0V2BXm^E4yTjQzaQT8j;)%sgp*b`iBlQ$IH&*o)-= z+()NuMfp#31!U{=+NgHY7USRbvr;^69pK1HEk7GcSP%t?p~ zPOUa~`_rcT;xq16pQE$%1K)8p_qdj=scYqs*R@#$0_#14eWx@`FqAr50=-6193Z8H z1V#KU6_rSQ7k*Lid1YpXL-ae2vQ;`@5=48VDGjPnI~wz7m?MkPz7mFiY73sdJXp5R zIq!q(FZ%XzVKoZ{_#)a3C2x_L4GqmYNqG8#X@>U26S` z+=1$tk&2#TKKfM12v}cO{N7EU*WyLJFOL9MBG-YZOneDBs`N@Yl1GwXL|oKOz0# zS>APBbT4HK2;k~>kknUHJ35|a=$zRIIwem%zfV*A__(Zh4Y)3R`Pzp!qTP4ArkZ|0 zF=l(hOb}c7OU9q-Fm9w0mgw&U*@ubvqNc)okG2t~7)#rI6-~8c@Lq#>C88zR`#bC=iH0{l5fegb2=VLQZRycJ8G`Jq)g~buUvKITd>!>EJ!)c+v+{@K+x>b(yFNwRe~B-wqirzZqL<815W3-vSV-4`x>HPQIlhaIASImHI#$jg%lbXZVWSqn$3P zDItE|Iv=rjAlA&BM)P*&yczAB#y&OoHc_jBVHCmi$Cbb#Oh|b-G6uB7xlVkuH}mVJ|DPY_o z-^{b}^mR8W{cq%@KI{*Z=RC6o{(4!7xVXJSguwwAxGB@oCtsNf)_c&@t3X4Tn>w8A zYlit#uJh8_`6dC9;7cUy?)UTpVVc%>0AiMQxTl`41N&j$EA5||nLPvL6kW)N&e7XN zDIRNIdtb@bK1}Ta_D#X>iR+sgldpfk*4kw=(jAkX5duVR>ipk7HbTc~CM8$opOeqa z4|4sK#-@nUB$gxT(>!>(?*$tgAPT$;*P9R)ZPb?2p&nP&eXRNat`}Lk0&=o2w6Y?8 zX54vgOD2L@K0StA3xO-^mht6rcF*4Q8{#InqMD&Q$^BNPBS`%r)I2yt_aP@vyHt$$ zZuM%05l-bG^kw@_Omv0aNCifbHeRBu?A|M5x+MPgW+MsDse883j79V4?U;78?LFSk z>pA=_=k9R!Y_C;eYO(8pd|jViQW&Ex7b54dPg)}2m2@>pm&{YvPKd(yFYN*mUD6KF zzhDr;Ct|JmV)AGnTOvczLtAhlB+JT^Syt6-7>JK6%kmJF@#NzMhAZ z;oI3X4RY=l+(rxT_mmE*`GjzoyU#1Gu+7Uq_M=S86Vcq|S~;)Hq(m8zjwxUr?zeHv zm_t{;Q|}4(P7PC2Em=*#Z9c9PW|TgosffhezFz&-svb-kO}Yks0539sw^IU8`AU4k z^G2GC_g<=X+CxkpTGS_3&QG!)o=`}~&o1b6{?5sjtE{63G%zzvKE}K~Q5v*mj<$T7rJMPLw4D;0W&OVf@$Cc9ArkqoQ%z##B!W_9+Jv zo5oPH$J4p9qtWly<7x4#)n%77g!l?d<%-#pmHi*ICwYmc88N5uab?NP+s;;`v$S_} z-@GrFtlz(=Ut+X22ybpJg3dbcIiw7HA-#9B*rJiI-*+M2R5`8Kis za}5wTZfuF+i(I=G&w3`SXeA_=w&;xNAA@4@!ZW?ac$!gg1OddZJ=SBam@^xzyJ#fV zi>e4k>gqjOPfna9${IXsfBKlDmzLJyPA{RlO(7Umx>2!3k?x*iQp(*<`kR@*{9+>D zM;ms;&lhsd3Hz3)Q4*o8c}Gok3pQysRnOTKO+L7@@S1ca*DQMXT;JA&8~+@$vP)N& znqEUl^&;}okriPjpc(m-yz+ zM8z*;6pszqhkb+-->Z7i=NXp|Jaf%k(iBo3fA)+((_!G&gOl_Ml!`eFu*Fr$+Q${P zrYK8$TlQV}_jNypIgU{Xc`c9?!t!CqR7#8S@k{0h?#K(kS>I@dgqe3V*RHm&AS!|arzeuABeAF{GzAJ*iN{pbD4_cu+AeI81~H|E8?%dPvOeIP!=Wb`9e zih+)~!-6j8Q(eJyyW3X}3w|pS`)-B9dxj+8!7GPizMX#Scd2bFx95Q{U!UvICN%1j zd71a}Gy6`J(5qeF^6xTEogdJNg2SKuW7SG+uf!$SPgcldv}-0LHJJ{lKg^C+*!Z~a zj8Q}-{r}9{Rk?k#w{Vl}@b6d@!orQ^G1I7{e1`erZLYN=el!NZg!u7@wovBWR^Mc# zuU}2sFZ7$fyJCkR@Hg)xLW)G~H4AGjo22O2}%t zUYm;LV*WCfk)@^hK*z7;*;PfZVYwpxP{saRdn9(%@$!4nYUrJ8?tHWCH#Uy31AqBH zZFAVrc9)x=^vF=r)EDio>#z(s#lI4KaEP<@{EZ}gaKwfe=^*}QfZbMJz3s)2eEqk#7_`j?ymsuVsUh|{mY@zVZ^K4%kM=HG>3KdVPsV`| zebMT*KKdOhZ9kzr&zyM)&`wgOGvDj&Plg1X)jxx&1CG#u*i)ZzyKQ2S@54`q)OLk_olTcndhjwBs_ZE}Z-W zP={6PDmS?WFBea{e7!FHCS9^!NG5r69$o%0jGy?*e7@Tq9(?b#V=<}N_WXf16frsz z@Jn{nN6wSIC+0O5wWr`0b6D^YuC(Bn*KtH`)lcG#mFc10P|B`@%)csEfYV%I<+7Vn zb{DYxJMf)fsc(;0ynk%G*^y0#VmUJSdKTe7X|2XJjc$0~JCSRhQYJQB4-mmtCMCv!@Gmpo z?mqQgxEdU8a(j6KERiRSwjT^6n^ZtNr{{s0qn|ZuDp_PV%EY)zj!|AJ&y6R8 zPqP|^_pj17zb-ojk^&EA4O!Ej@0LGxeROYHo*J@}-)8p?3)-|}Enw)|t#T70R|ghM z_tHjamwsthe<(?}aYp}?K!7&dXGLASJ2>OrdE&d^V}gvjyQzz#Vr30j==Rzlr4WXm zP};FNGyNJMfA*TlRseVUa18vruTJ(Yj&@mzhOU7P8qM6ERURlova95?LG%nl#icz{ZypwG!o9F~t^w z<;w_J9zccGAz)QNDBz{=n?qN5_Wc+C!anaMBPkxQ)@MdhH5ZP4|C6gx!&{m1)$OI= zY7&jtmX)PXhd9Ln`yV7BMDl4t;qUw$?}cb$H2a2upela1e+%}ac(SzFMUZ5+Gn{@{ zSoou>OF+^Cb517|%IKOx^zdnVD8k8}aPhk(LqV*B3R3a@gafTV#orVE?YYAyghZ6& zRm0qLE6YuNMryVD$zZ1Wn#xZd zkkSgX@65sjRA;ooN|BJqz1bfNci23hlvrR5mjN_ZQ$VOUd8k*>jY{CdI8*CljbaS- z0}&#NWnM)gUUn~0X*Rm}uw`mJksGOUrA;U>RvXG|(f{v}STii% z)p#+tS(VzYr>2C%rD&O(hdtWd^(tThqZ zc4;3~*+CznH;_Ke;}=Y~HebP-cY)q0taK_8u^jvLqI;e@NP**DxLRJMqnuTFT3z$P z`MHuEV+mbpdgOnZ{~!aKBm;Sz@MyiE)IHJng&(#%o3-)ltPwWrZTi@6J+g#1%Icin zrZ_GIqWj~e$w;3gXu$#CB79OIfVO>wqs&;WLvf~(!hQG~{vL;UU9I=}m_n5lM(_tl zZ!~IY@mvu-=Y~%-_%`&HW}|hcK&jp!*RsPNcMA;SUA{f0(@8I(67~{-^+XTfO3zS@ ziB}HdE>ycA%#{m7MtJMnb~JR6hBRADq&>sm5@gRXr(FBh`*W_dRDz?~XlTW265I#S z3$a55>cwDXvXoXUD;!l;;#18*zEwS!d?XjRp{V7%@1@&i6lE5ER8Y$r9qUX?)is zyTm_D$FQ^!)2Ue*x3iq`V?ZZH)3U%_S+SBe`@Rtpa!hPAA?y3XP~dN%E@MUYkk@1Z z0IKLPlF=0~e)j{0Q|zXk$@OD@N92-nsANsE@l@jQ4`REQg*U-|?qH{bXOa+tZkv#a zZZKzb5*O;FV?laE?ky|xoOdSAYiB>dj>cUBG7A#P@C>`M=FVNiI2sTQc#4%+$b!)% zbndvh68~w~DLuTD?ZA$-p1v|Fj;`r;aCdhI z8eD_BCb$H5cNp9uxD(uAkO09Qf?KfQ?luH>cfNV@-gW!`tUl9Kb*gIbl4`BgpZ+0U zp?WJ83o)N8&QlXOk#3WrRUir&<7a^I-rKuITl=n<F!1Zpm&zfP<#Ymng#d8HUkCe^jV zNf4&W@001GJA<`^!i6l~s5X=K382T%5oT+1anhabeNu_`b@BWIe0j092`NCwH8_}I z>P|6yoLDuF!Vj5@N7T%?pV_7iaglpi+DDPZfY`b2WQ2Id<3w%a>pju(XMf{GL*E9u zIn7R5WBs3x6$oBO{m!luaf7f&ju)I341~TXCbq3nTYw##5$^+!gxUwOV~*3FYE+lI z456Kc>lB4EhK3VxqGc9Pz73WObbw2CIDj#%$@kg4Ouny1|4CiW5L-5xrS*JuG-xJT-FkxkmB8kK^niyipyI4RzE#NE5)Hr#mWk4i` z71H~z(Bq;)UfRO#U$>&Cj&Hwsyh3G@|5OsNj26?wfgm=wxJ(g!Uu*1$#r*(J*5!1O z(x!50!XRAMX@lwASMu;0LSr84g5O8(d@3_EiYP{n=YQNr&OMyyrd+!QB+HD0K`W0g*wql&-q7~7QYI-L)MrVD%{ntJ8)`tBgvEd!+#w_yV61ui_d(0;t2ypZ#*ND3 z;;6HeNlZ`aPYTFlQM>LZyvMAk;~+!>$Kx-P zg?hP1Xp1&n8HAl?k&hL=6;Ia$1MGi_i)zkCDpw5(NCpR54c(UDRr9!G=W@v0SiBbw zw7?VtQsijf1%>ng$kN*=AkY@sw_wnNwzVxRTen{8to|Fi?2IZ6jJ}JQ=G_L5~wq5zay6N@$8Qva8bopzf zN;1=oA*i!V8}KG`Ye-HCKQvVN;4tl4tpy(XI?#^Oosenj{B!cZ!x(l4TzAs|*!GRP z2hvk%PRMTl`z+Le-wvepABg17tZ*20i?rgNQVC4|S%12W1?*O}T%ixV3r044FnWb9 zp8H0a7>6i2z531T#p(9%wTi*NjFoOFN0c{fYwW?eIH6P6Bq~2Lj3pM}i#X#lcaEd@ zmaM@ccA;z(`jyU?*Z!{4qsYI9VXbiHhog7pNHDrCbg9;hgIxMl)x3(t7}Do;vAkOpuS7aiT0a49`(bO-F1^O>ERi$FT2H5;&3lH}#mQ6;yDzH*VyOir)B zXk?=I8g6UDAnI9+5cTcxI65y+m;r?TH2drq5YK4# zf*zv9Vg(m}EIiWRUZOxc=v8vwqX%S4gy_m%DVY7{z}}Cmc(OY|pLkUaeeu!h$dvTAp@nk>6U;?x9(7g$L4=* zpFD!j*$n#89$kJj+~B=%Rv`hpRdOn~DLPHCD`$i2&c-XPkSRJa!36=_FW=!90B{lv zuncc6ffl2sIRIQ}3%$Ub^)dFoUm-{m*2v#n4Drzg4m3>e+-H+)cIiz_!gHxd+D}GV z=~Ywio{z@X4!at8WN0XOn1RA!o8uC<%@ldYN0T52a%Cna$R!IrbzYQN zr8vJ6d3iZmLWMNx6AyykKOM$MeCaa!9`AwSg!&T8LHn2l*_KCX%gq%s6$2cEwQ7PK z0`c2jy(XV{8_Kek{9sfxkSe372s{v5x9FEJ z8G#U8du^@ra#_Ev5ZM^g4>0<9{k{%CNEq*fF)>*FjUVMBfEIhCn-Y*O0H@v52$zeE z>Ma-Hy+8a^1>bWz&ex1agC2XlvVy(di3RDuIF@MnkqDXS-6D&Xbs!NaXjf7+?UHyc zAY`ZeeT^qbzmLm>&{Ysx%w8r?NrcQB?Aig zLm8!$elHMD9CY1L`RDNrZGtTRt@(&DjH6^W?#X=$ znPjV2)7$-0!0$SHfg$3Vd>v2D>f8Yx$W1PFm5&)jfP~0Vh*yI0`c_%Q4TYEDHGEYF z%}z1CNl-UKIzBWv`jSZ@)pvd6$r__O+s~h%lwv#rhTk79=HM+{DN~8ikdwWw++)cU28-h;-idGQBZj$M1^ygx9!=rC?nr4 zEG;~AaCOQSc&YL}7p6gccoc3>pfn)bVClDSk{HOTJRW75gTrl5q>JtJUMgMtT@eUG ztA0%$wmljAh^Ftjz2Qy!B`X-HUbq=0gS((30a}{N3DTC&S_Y>Z!lkzXiQR@e{RP{| z%it+Bx95uJ3@TQWiV`-}L`Gt+7eCBl|Bj>vQ!~I0Cnxb!;diqgvDpMa)XA5@{y*PwZ*J63IX?gB zr#G3@O@7vazUamMV7tP(wj6o5MBgz!fd>UV7&t8Wf;va1i1L0mFdlohN{~F_7^;`0 zx+$81?-Op0DO}GVlUHfQ1S_!Y{xdGIoa&OpWZ~(-`w&x@JDkW(K(t)W7U|nncJsh| z9bHIhaWhgEcPtN8_Vtm+7b4)pKq8Rw{U}BfUlm4;|D9$N;3=?`gZo8T*p^1OP^Tge zZ5>#-80GQDE(Nbt0>rlvRhH$^NIBJ_Cq&%}CfF3`#HJ!l+3Kh&j+KW&>Yt8Y!5NPV3y=|Mmt*UNKE}c~`c$y&%aI4> zReRrJNzhIdw%3xqt|2|9w0OGFvB`Rx@#f4!DR6+sSxX4!K9=WHqYaH1<{#k{^2d9R zQU`w+ng)Kko@aTKpeRL%ZGWo9WT@GjSwVWOYBb==8jrZ~jQ#DQ!Q``z`;K>-=EhnM z`&-~Yrh_KEeO=42QfW2{=OVh>x*rQL~muuRN=JhAPKAaqKIGb?q@pfNhWTi`>egn2<0 zzJePWv=-M_4coZnQDiZxl7$xrk-ENZ5elO`9otn z?F0-Q^0#m6{tD~e{Q^EBj&-@#H{x>*GUXjXfE!YBnQzFjr!Dym&le{)yA{*SBt8EE zb|`HkCjztWRf7-b*yUus)SyprBo} zi7kcdv#>ymi3YXJR^ZL^AT%JUmR@|U3TN4mi|Z{GJa8meTnpMe0WJ5zp|+h;?wKYY zN%TCdFkkfiMP?xxr*=2@&B^lh=Iyb3W~b9O_5F59Uz$F4K0fcISUs8Ub@LM1)^C@T zYS-w%O`e>zENgrxga1O>)J4@b-f2C(m~#Ap&L=s9m+SqU0oQZpvDCfYi&)kUwW!H* z&@~;>vA?*9ht$(f?d;A&WBP@8zc%_M?95!I9ib>}$%Uk_2?Icd-UUea3F zw(-uRVQ8BrkTY10U-F!mEN1k?o(;Q+gF<1c+{**SfJ()FyvHVm;Ljn@n}CW{G7`V)%&!1D9Yq=w={QfpZ=?5#XK=vTWIOiqc18SdBByu!Z( zEHSI#kfrYY0X+@*r%;i09J%`iL%~pDqJj>|TJFu8%}MBV@{~%?lZYC3m7`n(qc=xL?6MLAd(7{d*EYz;pgRN| zK<^ap{|h^F0NhoN-=|J~GuQyQx#x<2csMIMJ6zEDB`nr_(-jY&S7_}w$@+%ELk*2L z2w?d27<$Mj=sVTXl-M8IYcjw3#C6oHG@zix=oLE}Nq?K_Vn;Qaile;FDIJ!*-c!)D zQ~T>VqsbzHTjFU8VEtq38jKJ2bS=KoC~<}P$@4KX|AN9L$;4=m1wRzJrOsf7IsxpPPuOiKQEp$A^qJEESW% zlZ}QJHV%e6OSt-kI7kAdy_mR1q?;q&X-_ALR8crd9w2VI8BS`pw2`a8K+3{k$=;F1 zn*Njj#j2|f%p4EKto{spR|I_p2Y)UNOq}rTVr;@!w~OsUHHD>?o=j%~ge`R1pjID` z`tN*tQ*-$;uzr8|dG1gIvEchZSYZ-b>^Sndl{XKr@tV6RFhuNbI}I3I>(8bPD4!OY zi<2~_lq0-=S!ALIh(>PtqUpophz2BpX=S8vk;MMcB>-cNqSThE%0U@g$QivtF8GQP=lU; zYa@JXmV24ESl{Ql>?*)_Z%5!tD7Ph5pHAY^5O%#n{g_)_#7RxV=uO~z;$^r#CGd@E zA-6*9!nJ96831LZ$PCXa+;~lt$VeV#}JX@Su#Cg|{z=?d@SP8-F4!#K=Q z41nb4RNqK@g`YjU{OkE6Fc}g&!6YY1fo2KT9u8=9tM&X+hduRX%S0UZle!xE?ldclywyD?rq2(mKR{`_owL~V_}f#%5h2Lp4@Ess zwQ(mWG_iL}=0-8`>9*EE^l+hxy0{J=&U`F0oPqT4V>8x;iR<0Y^fsq(mmFi8a+i~` zjJgBzurPZzILuTH@g|MFb}Z!Lne^96NtyEH0hbSa;p6w}_G>I#MZ_636z&OD^IG9~ z)+yk=FYWo^?fB(UIirqPNt;+|MAUtr9g=wA3FYhBnPPew-Fa z$cs+e(^)>+?L3wfcf`MY`}S$v;itjHF9=`i445eHvM zN%&1UNMDr?7v5Bl^#L6i5fHO4aG0bz)u5f(Rn_gTxZ^vqw0{AIF zOVf0Ahf#5Ho{1+~X^W3B3{8XV87Jg!_4OklVj%kp0kX@PLDy4R=p)8yTpVAF!Ie`e zJsM@@S0NS=4;n6YyKZ7Z0i^bH7QpXwNN&1CNmqBj?9$?>Dq?Bd&C!`EuG&F|Ys$;%mn8F&k6`1%n zGpo}aG-@JbK;$}m8doPbb^lv;CQO5$=h&&t)*5u^5zQ z9S6M}@kLOvBQ9-bOI2VhD=>YDyaK%geU=CKphNQz0Np` z=w^BjI@qus%-#-2<{;R7b)dfmH~tbge_vn2`;E~U@6lK5{WX|l6mdI*1Rrl5#q4-0 zBKGfjeaQ`-E-kF_tfh!P=p17eG7@*b9)3-{d&JB!a*#H(cA<%=!zuC{FY7c**JbxL z-uy*S_X6kDRa6XjRJ#smR!4PC3v1>(m0$(F|N9Q1uBZDJy$^x?N9+~}<2;;Ucv`CT zRxkc4zEcY8%(4+P;>tDyjWhb|>1E76{N*%;)aR--NsDfFm>wi*A3@{A{U^+RQw;wJ z&-)R2if)ez%@ei4tv@;5C6c1VpXKwM>@u>!q^83Wdu25nft-$SP9^xqr-!B4C<8{* zbD(aczb@tVU)xGzAN7jzB+Uc8rk~qlgCtZhQn!~wBRD_uP;*JXRDBlllU?M|<9jIP zz=Cc^GFS>SsA+IpH>`@~_8Wr9W*Tc3+*gJ=_^zHT%TSJ}8)sOomV_9_#w4lM@bZ{Z zCL*ZAW}Au~gf@!O^L&k@qZV;q7u6XfJ4#Zfffh4^j9mCNnQ}oDJbr3^%4UCI#o5>_ zqUHTO@f0odRbNM99hGgu2K0`wr!>TYrWYfBYGO`SskN+1#zq5xi!LJ5K8jxumdPEH zMGWKnLi%(RAdNQe3TLb=@oLTZ&kXP3jF));%9xnp^*kjpFAxY>Fw9v-`~ap}$N*sI z9YYt3Y4d?%yyDt>5k9P^23+IEm?g{>IFwxWYKh3p5$!)l!X%K6+7=B5lAIi>d}XcY zY{!I<aZTTnm0M<#?)S!|W(T>n@zVfI`!{t*Ny4jRt50U=G>(vcNXsZx(FvhPVuU zGu7m_5%{tKnPHzO5?y6|gcQ3hez0yz|;!`3OHT=z5B@#$qXa0LeTzhwJ%+NCi*kyo0#e7zyZ=5#O`5yHET z`PfhpG2FWk#wRVZIr}KH9uKR1E?JpmiS|(JOtTV$*fX7mhtdA~ES9i*zsp`PIdrTpk zSno!osbce0odUqgH7=C(>??s3&nq z_3L*pTfemoXz0Or?E=8f9-_`)xUB(&Po}Ha+?jd?e@N9e0c{N~{~HEPE3cjpZ#;eH{yy zkNJZn*3Jp-jFn6}yY>0I*4OfC*s?)mjKUR|E zp;-cIDfRWC#Ly)BK*JEADJ*@klQ)SI1lZR#EpQYcuP(TWYA}v#AlG8|XqUcAJ2Xd{ zzbmP1QoRb=XuktlncQ2UTEHx62a#zn&G`JBzV*zo_%`1u3I_sO)Bu%*lOzaRgT~aA zjYiGL6leYUl=CT?)7Q$%wE-)4Ua^>17H}XDlTxC9>J9)mE6ac47Ki_cWvS51c%*@*1-naJ2Uc7^+@1k=}d;GX_W+ zKxoZ=CWXiPn1HC~`dU<$lM6CSgGQm$fBKM*6%+*t_*?dIIO)RCy!La7UKgy9-}FhG z1%WzE{4z5u6vLxDcpPj&O2CCkOnjzAz6jrCKKtXS(Tjs*yXIkX_|IIj>hG=rNu2#ml>d&J|S z758gD&AEt(OST{*`2y5D>Wrw^?Zf3sLThon8R!=!yMLf8G{7foQ1s_V=J8Nmo}BK# zM?(Noh^aqlCYNJqH(ZY+zQDUG@^wZ6Q2&=+VIcO1s`ktt?|eb(aXlFSM3Z1pNYdlv zP3V_Bl0dBCD|IsWTiR|xGvf_TKkeQ)42lkeu-ykwFZA{gW{^>n2}|<1lK+b5os4Qz z0$zl5vS9m3S&CI7!A2Ogv*$-BhJYc5MvCV~c5kV(pK+6;;1S4hNH^K7scraK>29^d ze|Qko11^-8D?7E^hH#WZ0sS53^lVJOw!5(WL}C{0!UOur-%*lb8%z+=cG`jgv(iwXxL7(8P6_DS6*=Xe?$gCn$__rG|~W-hUDIlK zcu6dO?f5+Oys9e!Y6l^$uU)hal+9B=R*eyVg~oAB`?cpUN!ykIi|n29T;H0{lG4 zbdaDPhDsY7X-_D-qaO;5_vU;)v3>M48LiJIIhC$V;r{?g^{>sb>pB*9sC~enIyM< zO>1pca8Oc-PvQnj1R~5IbD;I>04Mf;av{3CKy*^_;7Zmaqs8|jE4{7S^`L5u5*yrd zoc0r?N`!RWrSdp0DAa;G^Sg$JPC+*l0K?rvw97`BAgDLSM@SZ6tB6`iawO1Q3Gn5 z?cq@ZS%_|>p{aiFy{I3apV`v~@S-MlpXBVl!U6Wf4)(qx(tvQ$|7cD^*4!~Zk`C6{ z=XIz^dJ=g=M3e6JAIMb$7pF!c-L^iB@oFyL zCbAcR3QGiV(vQLi$p<}(M^&PoHlk4IQr#nO=ed9`o8ON9rW7u^jZ7rmFK;Bb;M>;E zB5GVL+Wi$kHqc>9Xkc=-?)@93!EnGwhf~yKdG~|f*M57=lRDo%8cY46$YM>7M=MqQWdvv(-oF-mdM_ z3^|~^xZ_N)1YQPHva?5(pzyoR(r|ic3pGz!mOo0X=AKp;w{&0BfEK&up35H%cCBOcC@6N;oxc3BOE_}N`c(A!d zPxzgmKX}u?Y&m7TrNvyMy#R4Yp?$ZlU1W;s?0T#?tM?3Cj#8SqV>vk^zF%)1VX{56 z{vh?h@ck$F4i9LR+c#sGxw)NBhmY-l?xdY62mL%q_5{lLCdRIRkrI@n*zV2G?Ncpp zk}>m_pWMg|hH1N!PmhGi1I#y=Z8(3%0 zgddb??LQxN{&Qs)`B=URs_d5)+AA4QP<}2+g!acuYC)*-1=u$$FlkbJn-w+-MPgWO z>dAG`U$4S^)y<0AR+#g8$!kLgksBS&EREVSJ@c+VHtDgog}jELTyehw;S!SU;rxoN z?Z4F`uFf*zJbP%ycys@`5uvkaD!=;`43^=(H{G@0+d%fSLgP*5gY}A`q-wuFS z9TYwph0Fd(xPd8!?qW&3Mo19Xjfik{3Sf+;wu*zEkll`lH#7|MAdo$e!$YHeY$`Q5 z2HrCgT!!ByX<*Rngv!_950)S{o-0L|Ku~a;Fm0s`$Oa~;hSZaNJy|t%%nNtY6kbTn zgL8hB*SCUq#ZHLdbh!usic0AJekd|u0T-S1j`WS&{PT2yE#jK`BJy(nc*)~8nG_a% zFe&gD992Nf)LV@i`g`An#)4IKXZ@^RMz`PuwHmAHL?R^dJ@aX?_3d@H5~;1{WCy&o zq&jyAao;Td>_i-=NglK@iXDPyTE+^}!5BK}Lr7X}Wo2b~1kw>RuldTyPrKk+B0Q)> zBy~tjThN3M(=+**N8}0Fqe+YNsw&*(_OqjIDh71Uui)|( z|1mPnS+CpBb6GXsmAFf!vz!^pW2i10dKKDM=!Kzf>43uWef2E5}y%QCyVFENj^ zSe)u&;#jA)P@Fb%kMq6U#*mN~#T(F!^y>}itu?l=G#o;Ha$+vRRp}&j>XC2QygH;7 zhYN5Z-*fb9Ya6?sTfOLdbABoyMo^q?jZg-olxh4b4!qR7^ma5BsS_ZGR9cO3AFqR5 zR4OO?UYQOuyOr~qmUmGUsZ{(H z;}qwNJj=>lFPvFM;u%XGG{DqGWaNcff1&b_j0#VM^2XMyMv;0H3kzPDk|A|d#v~=b ztaxmAZin5u_Hi&^$U)oA6#kQwQ#=_;sqs#igO{aBKLwW4Q`o6Xq<$rxoAybvW5Lcp z!u!7-_L#GoyJiLG8y<%#P}H|0&#UXzW+BeS_1 z@(}y3T+*KPW5g?1(6uRX{hY04^4Z{d;s&Wk!^~D%H%4J?7HfF)mqCzOd2c9fnuo4a zS#2lJ=50c6#1Ghffxx`x7NC=@ln=uWU`vZReMVi2VtHrU>N^jxmLb_4^Hzr zJ=ffH?0rWL!;}P7neI8=*A7ik0f#h938CF(d<)uY`6{)F7jknBQ8;GS4nj`p?^>)X zV){r5_TItpC4uT2dOwLtNhfxLe8zpmsE44?-_$WuYa?M9l@5Qb2xZTy*n!$|fRLXp zh3=QEyC~DkbLruAgnqHox!w=(%~9^Ov#_U$+`5Y?ISp;5+9d&vyt5sDp(S&K%Nela}3fm2tm?sOoza@&*8+*pqK7UvaKt1ZY7TV z;4o0)wix5WZtiz8@pR_pO;exOr`hv5DK z>?iTLw7`QQ-^%jBt7>yy@)b0fwR&pfuccf|L9kvC?qTwQNlcb?=umJb6Sn;E0F=Qg zdghbD$ISSYR*BY>zF?+?hqAzb(gOyC@?C((b46hpsr8}I?LWrpv|X>1~ozMO|l z>Sq^wT~~@hyX~%i%-pb8z8uABl_+YjIK(Z=tfy{U>nj-IXbQJRzHi~%NcNo~^jGU^ zg>Rvl3;~9r&krgKrr1{;0dWl#0^jjX!iduoUo%~oyB>W0MW1X`K{LqE!lPaG4!4nq zAa$4A;OJpRsaqR(rFG$-aq?2!;05ob7Clc%*79){=i&`!*@fI$9Gx*O9tV;vKaZ55aYy=5b2r8QQNhk)xo}2p%AJ7GYm- z(cco|zikC-6&!!cr4h4~w%zjfC6w}5B^abcvbNC=(wOcqgw~UwOgZ1kLyMkAO!b4& zA_u30qhm>1qOWcw{;MLoiJsWpE>;?;r@Yp`L{Bz_VO)(q!<(YLoFrR?jZWbFi}+s| z(z(iaE!EC9tIpl;)iDC^Kgo8%I%Rn-_P^Oy6*`d@5pb=U$%Ee7KEf93&Pt=iiC7cf z{w`G}SgCS;Z(w2Z)X(ExQ4tr8ts>VM%I9y>3o>;wAig*oUwPguY&>)%yycfo0%QO-J(qL__G{GzQu6o@;eN=YZw$$ zXVMkVvtm(pJiLQ|8pXABpaj zG#g871Lig(!C#QX0EuDc`XXdA?js4)W6XVnmkZ9KIM63C1gi= zPclfjPgP2Ho{4**ax%BLO^YDc+4SFY2HXjus0~Sn zRbToKLnbO4`(L1)@^n>>HpXRmY2lN_-$KQmVRKvBx6<_5&{F@g%xpYYW%Iqw4W(zj@q`q2d)2Vqdp`;iU%nF;}R0(7}X;P zkbA1y1>SLb!yaV*l}-^aROzvAF65Ca;C!0dUZnvdwWGh$UHViz`i};VFI14cy(6u%2NpkxX_Xa(thn1M|eR^BC5u+#NW*Y54!&d*OsE z7WN;~1}P(SV9OJPdvN-}Ci`D2!(yG|I%Ia?x5}Q9O%b~8O=nB1qpa^Mc`SX;P7BNL zD~LUT$rLY$o|NVv6cdJC_5+To`6du8gwKdpi3~8cY|ZNa@=ZqVtH#z0zft?rEkW)*n&Uh+7d7xW{Y3VkM5A^^7Y)6d%Kk`{ zFrO<{xydW-9e1R-i6lT`&~unaDx1girS#!Hpuy1*-#Q{C0j62dnh!Mm~ad_j+C0(^4lZx z(d1F@d3ScFHAxm)o1i5ozgjxdz?d!T|HPBG@ZGk`UhY{%T`A~_CW@)cbd0|=-9KK^ zXlB)DpGg=7h{K6My zbAhC`=$`nWDagHf&>{^(cXuOQN~3g#gmjnE(x4zIC8?xK_HlWWJKV~RRE*ayY32=fadWn^b%26Bd|jYmsE@4`2;{R=lJhK`w5>_@ffEH2<__USxEJeW z@4%H9m`>U4x@c7^olA?B2DWv=7L+RdaI@?Au&#d8-ELStZ(Oyje%v$@Z{j+xo=UR*zrKpVh`c3pTvkgX(mo{3s(u(x1^%MK_4!M z61p*!NFZAocU_Q&RfhMIq#OR>KeEzmyFS-yW$vDx`QOo_{qnrgU9?`^xINB$_$YVo zwIRs(;g!^BfciJ#km8={l0NniSII>WE_l19c$c|(n=q+LQ6SGhG*!uePvB&BQgt?8zAww@(DQ36Zuf;kyVI~)6t6#SB)QDd9N37 zy1dJ8|Jh-3d}8cUdrBT=uDCxy-^0oD2~9d{)>}+YuKy0PsT1_@^X4q+G<(<-k~&ZQ zruX9{tJRYS$*^;rH$P1`cFT~$k515uu#(u&p3-tjEJ!gGAUbmzAj3DLQaH%!C}-i}=7#hc>RR>t4938zAgctuiJ&jk$ZD}}a{Yy4q47DI5|Z-U<1NYV z+B>P-YmshrT*HO0kMiSF`U0d__X1zDz-c4=q$!dG}xNHJ- zjC-*#_NfdNUpsG4Y_+Gb>hsJuDF1UtPJ(6Mw@WLg3%8cXJ=Wbj#MagFzT2NmZR~$7 zE^N2#oW$mv9@wr3$tQl^92Oqj|ERiw()3|31Ja^G`f=`wQ=NgbgMWsnf9|T+A(&wb z+VR4bEg1#8AlgstG{nU_Tc?MC@8ST@wEQDs2qD+3_#)laU(q_J#D5k&Z1}moNlLVO z-jY|%=SlhZzC`JH3t<${JNOqyjvvV@;y>T$51fsQ#;&L_F^~!`jxMpSTH%I>Q=R7w zbz6nQ8jN%0n)02QqF6|E6|J-->03cy{MkOZ31`UeqZoJazIm!?D%0jD5!^>=dFNo4 zO#9Z;;$s11FGN!u314To;;dP%E*jFb)0|cwX6K!tm_@-b37dC8Q}cP@;#yPTJMv=U z)aD3t%D%B~&H}%8w&2@C1+Qb;_W?`e*69Hs1+rW(%sQ-dL2OPph)NAyw$&Z>ClT4f zCdWp{O^j{qK8IW5Ef7ziAL*|ijlT_jrdo2gB=Wh1VuY(m%bVm{)RYgJ z-Vd2?R$Yl#rT-|=;}N82WsMC#gC4df4HCOnRTRgju z@K+AQUg{nctk~!`ii+3l#$#n9HP# zb>8uv=Ah+Yt)3?P5Ds!Y$aQGOI&Lk1*JTT8Bv|a?5{YKwLO(+*sS~;76J{wr2)1Ul z1DZmNqG;z!Ve5&PU%$KNYc8GITYmR?TCr?`bUu)-5?4y}<|LWX?WJ-NvqO0Gn|piw zI@rAO5njj|kDF#e$MavzoKCV`bVkDBoFwSgpVp_k%&OC~s^{ISClnsSIa+d8LMDf! z!=xBpN5G{jHf&W}+pkiq1}7|AMY2 z)HmYBDA{pSaU$`=!{vB9WZCa(;<^w@IIhCkZ7_2PhByu2#}2_tbca00oU2(Gg(0*< zq*IC{Lbhj@e6M*LPxz~#0_`b`_^83x^10U~0{-V=oT0iS0bX*WZMX1V*crt<+Cm=P ziPC*uPl<7vhQn79LN}p?s2)m;jRM}VzcOuG@jvC;%F{B#Rf3(A+fWE42`XwtQzcP+ zQX}`?`gZV~ye^8yUm8ndZ`|;mvCQXEQR58CFR~i!)LbSp#HEsG+ zw$n|kQ8b!?%r3$V=t)kt%Tw0%FHlu2fAxDpg& zaYs;UWwJRHoDasK})s)rL8^;#8 zw@9aN2lzf+AoJgtOZ@rNQ6z`|E^> zeNlYHr|KT2gh`cW=QpR8a{CU#-MvCKI8=1A9WpktLYdc?+? z&yAvym&?PLl_UNTu+bG21EH(mE5O$M+{g|+D*dv8n}{pLKu8R&10A`<>&n|`i|G{y z#VkD|B2Tq119@OKNu5s0_Vf-ah|^+gkHvA@v-Z;areF`XTE^ClOF2f1R--_jEA}i-Y2ZL6d}&*Tc)Avu z9MwYRnZ+q%CnJ?ZB}<)c(^4iJEv`|x&<s?w)~q_-#DF-tEJfZ6EfmMn#bNPZ3;sfauIOgZ!tQ`5@oUskMULFAlXkP zDfWZ++p*PyM>H~ODr^zCz;u%yV!$$>EDe&&P4()fI3m}b$I{VHO1Ve)@Av=Cxs*;XqLXTDbLMlpMPZ9*PD%q*5|~tT6#6DM5s)A#8(Q7#K}GVO+%4K ztfVdvhYB;pZ7aODlsy8u)Jp$;<7}FAtFY)t4&Dv6Vv(1hYZK;cx_&>{HiaFH_H4MQ z7c=pD0$YfkPs7E{kSQVsmJUh62IIEkuFht=} zVCHi;f_%J@`V<4Hlti_!$-IWXZw!3pu=C)Zm0HrYE4*-g-*V)lSh2j){i2N#|AlrB zY#<`eWV@_@U_4van;TLorLdk5HU}c`hy3u_;fb<)k< zkXM2{xwfU;1L3JQyeqK*MU-D-uP4{I{~2V2Ne*MzWDrX^I`^hA~=WO5JD(%N0X*I~i1?m@xBQzyD@lo~z=i_b%8 z?>iWhh~r~7WmQ~S@;ZxSa05wCPioIUbc2jz-W-r`Ea?_;fVg<@F6plu%W^!*@O zV>*Re1_>WNP;61f@)R9JX>iXzZzp&nhR-dXpjtLm_Cw6fJ|D5*v>yj83;U6^FWIK>^Fo)yX-C=j zYP^u;sgmz&dedK0_SDoicwElsTmr~rT7{AapB3IAetm`KkU}sd96nV!1JC~y#e>GI zF;S9`KH2GMFXskQv=g~U4*KTF59&`;T*%%dFD&|-W9d*{@5sbHDQDWk+4rfE(|qgt ziu61=+cgu(2g%M-LO~c;hyhpW$l87+m7u?}QCDb$X1>jvg)$ccFQS_CfsxFXF`V&x zqaK&IH!4pUf%lYSoylCK#8n`5F(Q?QUPN{J5({11hopRpP_SykwQq96(-oVtPj zjZ2VAh0SX&4X>7Qm(>a>c64x`@llDVoPH3?l%#z&MrVfKu@B;p0}KNp=~cKUFIXG3 zMy&TXBukh}%NpaRy;^)H!w4`>^D~N^Vvk=!^K5&HctNbx1=7*C3FG}jg+i%BHnIvw zz4f<@4KVn_j)mlgX+sC|

@Ncqz$8E)p+kF+(TL(t>$Zgffjdy_vASwJ0gj492gc z4}XmhMAc-~KSU*cJ{2ny;)vx2DyVn^G4%PF+xSd1dCzD7^)wLc;;H{3lZ6BsVVizo zA{;yhQJik)UM^3rT#bQ_QRzIL4X%A8{CoClqU{zEqkReXCDItQOOpxdNxE#bO;8wx z(tv-yQA*ewP>q6-Neyolq>!pU3swIF%U&BvAyjrc-C1^v*u5&@Gq+%=dMMiy@hw58 zVH3!DvTPmEBVnhLG553jXKv`RnUhgOcfDWG)EVqtU_z8( z7+e@f&~Y~y;BhYVNiv?4pW?Dgt|p4{wA&9OZU}FoVkcjJKL5ZS88mIy*0_|kTP^_k z9_j-dK%72HLTNlbI~QH=!-G$E@m_^jr!4KPHj9B zO?ukKaG}A1}~L53(Po|^^>S8c9b|B6K zf7w?V*02bv&RTijDI74r^c`|_Med@5pZRKl?lXCd7qMGn}!$ENbFqkA^y zG`VR(R42l%m0L*(7&Uatf6`TVZYEaQO3AC=~`%QcoeaeH$q$ zX=c(^-wD)6yN&g{^s-+IX|_8fL~jgMj`$X%N=#>ho7<=vhlOZR$YsWH7}y(TKqT>r zRxC#Z>)V%feZwPK7?OgdfuZRNs4(w2dM!;l$DGMYD;JgMMGKQnvMK60)WxlPxXDzM zk{fiTP$lxNdodVt;AnV`;8h4iZ>C9_azgr!Bo!_m(9g$CcQvpi3@7sKM2TOoI)*Kx z<&aX3DfNK{g~*#i`qMNAsI%zIT!s&D^ISknDQz!8+aU z&3I4HbTw!4u?F_gCmT5_M8g zNSn=JqSd;3$>qgHQ%=qrEdNG@(5zbo{-EXXgbhT13bT_>VFdN;Y2!LXDCi_#6rGIF zdxlNsjW)sfb`!5!o2DUBSOmFGB$35a>_}v=wxc~3b}u4DjIaw=yvYF*11}iBN%FoB z`ulg^eLi_St(M(OB`77Qtw*e(!`J+r_~RTwrN~!HEQ-kNO@MHp$!Ap*lu6iE@<&^C z+iA-9(lRv!@8kjkiCvSUqXn@smPFx0qsf_AVwKk2lycYADB97m6~jSWbD*CN>(VpF z0$~Y$o_ph6ggmf~wqJ?PWnS7soWl_bUQRL`zL3qs;R(Q7OFETOgv@GnXvnYjL4}Nz z-!-BnYIcYj^b#DwroUI-)iCkk%_x_Zkz!C{a8wST4K>Ki8TlbnxhQRnVD#d>KXX$+ zr{)uA4XW%EYLzl+>DwhcThga-07%b^2%Ziau? zkO>))B~YW9Hz;3v(r2d_3S;%NJymI2?&zak8NvfM^GcLeUBn^d@GG|PPzg){7$=lj zul#1Lgcy<;wojp{yVm3|*3m3GZfvvE)P@-OAn26bkDsi?UYZmb12Tmjj7^tzN>Zvd z{md`SJ{pj%fDvk?C<@@(a=G#(DF~|26GHiH=E{w{TIhM&vpzjbRar9-5OvEG+fW@L zi8U0yXnZOy6WQmLX3+Nl!-X1M*aMlw7?@PW*i=0juhLgvX zawiUkLX<2kto}!StPYY6g(K}cq1f|_XD?pu^3EdeM25qYcY6!AY3Hb%>25a9-+w_x z{0Sdr(oPDM5vd4dN%5^SkJ5qF=@_J8KD5AMk1+56qnZFQSFg;ljF z?eID>&L>$e8K`&Q2Fbb?qBMS|$cg%JiVfc{r;WsGF?nTA_%e6$(kpJdMVDj-#|gd( z%P=JDlIl#_os8&H%;uMJ149&xrmC-X(j5l2o65bN28#)E_!9{UC1UB_wKwU~jI+|S zFkSI}Qy#JW3Sy1RB8J!}Fz_BcFO%majKAUQ8fC$35HXT&E!~v!?B#=8<&oNootH)C z%;pyggvMTJq`+puzo|f*k=l}EF=PTQq+lXyQomB8Xg!sp-X@>Uf<5XYyEy67_b7Qw zI45P)Y&xKqYL-Gia$Frs!uBl6lR9FjhyZTG)&;xVgBh9Me=a7T2YyB#m4Cu5d#Ni& zA19M8`FoAP4BaOwh1aDmlT@Ub!&ZBH9!IovMB!E?X6C&*IR<1_CGAMkQ>#Q_-h)3E zn!b$}2P#5osg*;0qENU;Vb-3TN=Cr0`XflXDLh{m9`4JR7Q}oJn%*o5Man5wDDfu8 zkfY(zOCGI(Cr|mj)Wy(?6dijOp}#&JYtW0KjN>|1B+Zxg*-jdn8LY#|7+T#|ui_{O z>n=pn_m0y>nSpM}o^^MX$?xs7<^@-EN0XlJkIV+67=jte_CsXG1=dba{iDxQ5i*hV z$Bg|jb~rmB7-8Rpny*+_2;en87)>|aP6;c1z?{$~#*ELQO01{6Ls!IiLabpp%}BX%$*~ih>b(Dfd#zV3v%5JuImLyjEO*Xk#0V z#26|`PS|$5zc&dFA!iM;%^+tq@SFjGlQs!0a~6KI{8}k^tBCkhy+0p&iwWF}p%HA@ zb60cD`4LuFX)~oo;;Yjrj|C~{4CFRguno?YvQWW7#T5R*W?)C#G*s{=(@IJ}FXo7% z=`GyDIO4`<;r`OQ^ibu2jW<<}w&dc=Tm%NxG%<4zURG@$(Bv$vpRkM-zyK^M6yW7*1=DXlsyS|yQ@NY+!Uy{BJ zA3hffX>-I28rrr3{Uq5xFJ96zvY5;PT9>vr%*8X46_x4$j7XztaDhs@&y+3cOqt@uS{w;sB3Y8poMy7U%+ zmo<)4Kq5zOiD7D8%JB&&D1*u2gW4|+146i)s>2774URsaj3;5>53?k+NPN$mln(jQ zWuusQY9lnI{BoLI&01SXboB`Ct&pYS8+GXzHju(f`0XckQwgpJOB>Q>54aLFBzzZY zt&RJ;O*Cj-KP-_&KF{3@aOB|q(9e&E;wE{^VN>#4pn6_k3BHL(VoZrP?jS{e2A0@K zl4!=w2JPtKMiyU_79RaXxbkBl-XW$#)g99oMxpq(e0)~77U>d7ZAQ7~BhN+$C4Mf_ zxmX1XmRVz>^Ryf_7I0x-0orS@ti}=jWMcniUuYu@nWOun^?;d^hy%oN<|V|QntMNl z`eNnXkTh<#0;f|hd0j8eZtpV2AT_@NuDEZJ#+)vLxw@lQY{=NPjm^fJ72%{vhBa-P zPu(dlq*_v}n+>RbUWnv<2a47@1U8MHPKbxB&qYYG6FD)+3B&ZbwNo;hMWex<@6wA$ zyl8ihR*bS%BC^D?l1y8tVVmr4!!m8ZYS6oBL|@;TP>y#`t~3cDB^RED`^!r~PkX(i zF+AlH(L|)oLt|3a4#ux-d`&&?OmJ4J+#;%4>wY-;Fb13Zk$A}pPu>>8fp}z5gKM<9 z-_y9m);*DMBJl^*D`$C^E?G*|6qCvyYawKB=U{wB1dAb__hNc zx&>1x`56w*5-h$tO(Qt&Vw6unzv@}>l zQ(aC`MHZL=zYa_hRT$8~i&WPtku@MY1aBr62tC8Xcb#LxE08M3H}K@}{YJ@StDtE} zhOihGE*__YL}+b9ZipcH9cgfSL~&YHsgs?tG5SaU*TC1_lN(*Ppn_-6?0Iesn|4Pl zO-0TO6MX4fWU5#WScTznt{qxN^tGGzGO@}I_{r0g+joU_edZDpw7Jk1wz$q%>ds4u zyMdcT%wtSH1b<{I64C~3wQv3)-F-O*I~kw#d0fObUZ2IYe&&6(gWaIkpsUU-qsg#u zwcF}uwa!6xp!$8kf~3i_9ibYt;<{)$1+0SnM3l<9lzb!l_fjuDz7o5qT4YaArqV-* zCqpfl3_rZ-dsx46sM*|#$iP#C?P5)1wS{amYbCG5J12Y>R|@&DqVGol{Uyke!2zor{YV zSi$P<>*QhP!|LQt^GMArg`2I5hpn>{_>rcWxwEH-C^a>3AN;TLIl8E-{+qm$ z`(IK3Ie3|6}#Pc>lc^Sf#2eB;#!1`FK1<8ByxT=L=anTi99({r+fW zWzJ=0#>>yjD+m=}<*~HnW;M4ovtZ@ph6-?4Ld`8W`2_w!O3}&P!_3J7`bY{O&SneH zu{5`|6tv>xVdb;rG-Ks42UIG+Va3PF$;riU$t`Hk&CT^s66$WY04mKK{&}iLQkDQI z9!q|HGbj%~D<=;x7b}kduK=q#fKPxj7gPYsVPVDr<^O#eOA8@6XE#SPK%KUZX4X)4 z7bolAHy#BpB&ne&O3lT_@%I%C2Qv>VV1p>Nimj8U&)+w+Z5^Rn9%hfyQ+N2Uhq#BK|MTYe8NA{PxEL9BhBz0)v0UR>;ia&qLhJ zyr7o99|Y|CbIZcU%*h%G*!N$6`qy>a|3O)N0(=}+0DG)voZLLDJiNS|tmfQYf~-(; zL0$na3rlkg3&H=S?(S^m;ceyym9z#p1-JnK^qU(n!*5YC{^x3M8|b5)H~`z`;AZ9E z(dHBo;^Yk0MF6SX+MIksT--ujJpaj}2>YYs`j=8g*#BQ@|Fhs9o(>SrpKE{z1spQ_ zzdiC_GJ8bI|KaN|xcwhK0U-U~hy1VP`(JVWS6u%q3H+}X{})~V71#et0{^SU|3%mT zZ{kAvyQYIW0g0YBP_a3Ns%ik$-V+yvXYL>n0o~(27)4E5pjrnZc_^yNBJCidBNGx1 z_9Bu3mk`aBQB*KSF0W7?DcXMG1KZrPn}oKPj`ac*@V9y9v(J~v2k?Vicg*e z5Bi^sf602|wtK(x`s*fac%LapX4=lTZ&_;t14XGVE#A7u zXI4EU$m|LQHV~m9zQY__KR!0rA{s`=Z8RntHp0KZdxjMbMh^Se&rDqqw#UTmtUJww zJ@UgsHWyS$PUWKa_wI1pUJ(%y64dZU0ZC^Lm&kwZ2?GyHo-Ygx(5%kRo@w3uB;b~K zd>C*@TN@`Qr)QBcpCu%(I63~gq3>dNV8FtqC+M!hLKpdQEe8<`e!`cT$k2>D4v@c- zT?oM5NLW~yzKjffTwI(5v))_F96rZ?pA`5o6OV&~qaP3;lGlTbfw6N&JOZkCz+HlG zZEJ(}vta)|0!L=T1R$ekj&&s-J|a|P8QIg2IbL@5+83FuVSh<40g8pFZvlT-vvbYkf9c?L{6+Bc^5{950oC6EbVKg) zze=Qg`m{mNeTV2Q^+uinGu#{Rue0DkMJqo=U8+UQ-tpUXw-{=D8iGTVDj+1vXn ziK7bsw@T2`<9+@832_nC*M^a_0&mXEh7!pKLU8{fHHE1xZfku&k*OT}kx^0d&&JsR zan(GRl-!<3uF5}Gi^wq2pp@*YK2(NEhKAXF`sS>wqEVOsKCBSo8fDaach^i!@($qw zD-|ZAs~83&m#;0$uclbaYLvQ<;JKbQ5<2up=d33{ds1-3H$`^oSX)| zqzTH6BGTgFyxLgM^5<5TH$xGjGX&S`w!I-Rhk$~J4F4e{_=>&zyIVVs{oYhYEiUs? zK)ALmD9qrw2gdz(C_=PV^a5!#2~^|a4M*ZUGl`vVg0Qsv74A&m2PM|gAU7E-?t;>8 z5Zc?ndHEb~YyW!6?;-~1A*Zx`!vgjI#O!k~Lta(we$q(@Z%0=l1ee=E!v3z$R5iz=fzk^YD^+2_;gWA3Yzl=v^0IM=5AeqN<80#|l;VPI&;-9B~ zmVeUokQ*5hvDe3diHWH`E(XBR9}zMtqFRh7w~%3?!H3#x(yOGN85^cHdIM#~N>goN zXXHISzwV9QK441pj*n@Mf*#$wDURNzk;uhh#_jj!m88EP{klQb% zpca^$8uSW;Kf=dj-OJBU3uAh+&%8(TtOLK@(`eeZvpVtqUUIsHyZxv(xygR(MviIsn0yWQo9%16jmK+w@DmGtbZygmg#MNX9p%hr$+10qrlr5Go-^;I zR^M^ZEYttQW8bS*)(qlW3QEOZSM0%!H2K14lXBehn!UW(Y5Cf9q$~+Qt;b8u#8S04 zv63|KlF7u>lhY9q5tbL{-?Ucm4?n+?TP72532AxO6C_H?V}rF^XBIwx*WqK6F(dqH3K(Wm->{1pJ^pR1p})0dC{b+~MV&At)lclyF1pcaKm`N0$6w~)6a zvN~QsBTGX&r4!(w*5}8+`Uof z4x&j%=y>8Ym_l8K{Q(spUDlNO$g1j`kQHu8@iCE_g>OqkRgLUNg?Y-=-N!^94o2@G zzvLhdm+m?X@*GE`8O6W!-a}npE(KT21ce>9TEFy#pf6r*Q@+W-V<8_0B4l}hX=-L> zG|X?w`Q85#M8&|cT50#Rc=pZ;8fz_f@TH9#bL7W*AKYrtSUPKG;N8td37_xT4ivCU z7Cyf8qt&kGOq0n}aKb*{^K~{Plu{X}YLeD<4Z_M{;1Cs--)fa5N_M2s$SYj{PQ{Ad zxYI=mYraI8@SDK--Zb1I{_m)f4EOEk;y?+@6Z@S82hli%q_!%t29*z9aXBU3^XAvp zBHGO=g73M8Jf{?iD;*o1vf7_Y#PWjoo1S3lsO>gMd%nWo-sCFMUSLELHG5KixMha!9z@V;g7FF7IlEhrNT~R zam!;$K|j9E>hil#YC6i{ch-+Y!PoB&xc)R>rFYStnObXc{H2ZY7_*Q4x}AjI>2>Ll z#nDpJQ^`5&yui}Mp0Nye;cZ{;z#ESchCiVji9nWKqqm$XGXV(p21pS2=s^wJtr2&5 zB)$*(;P5YQ-$bC9$0ojFxlRq9PZ`qv-!jurlKA?xun2)(QGoZApBZ^0pj|8K-J7h4 zk-|aP@j#ZZs@bOc*S6E}!TYfZ5~;uV4Vjp0>6uNeU}a)%kJxS>o&#Wcwi8l*+0Z`U zaGC_&w#VC@>}fw0j6LJbb6N&LzI{dKl@tRnb$GKQ#scx05p7m}RrS>R2nbFI&CYY> z*g!OVEy*D2N`8H|Yi$yEk){)9x6Mh^4(u%SH;#fJK-@>&LgpqMXu z^l7+iu~qCg#mN2_)sS-ZSRuh^l=|=9ZDxSNEorctJ|fdr(lYnFLW{24c!`(plC7+n z`&KUYvQ%OYh31ZphWX(Ge6KK5d3yER$W4)ml9VUm`7*8?b7kt3WvB@NC@qyC(W2My z7FJe_n#g^Xwnx+ARyuvP=3m)<{g9w3AZ{gumdsH-fY~Kv8KYP8s@$+S#z35NqkESJoZpwmGZMP}0Ab-auD^Tu z7Lv})c5%)LWQQXCpgm_-4w+YGM6^IODlvo)+?s616#FTG{~gz+(x{_ZniADr0uLVR zFGq$}n7qt|by1s>L%P!sgo%NWa{bS4i+G-C_y>2|5X!@sc!(OSJDB7=Bgr4Yl?8>A z^Y~*UqOIxKAcz?sJcLR&gn^$lwA>j#S6*JbaScBDW+?uS2cfXrYx=5HYgsHZVVvRl zi;r)J*-oU-N4idbxwe&|M(?wDW3Dwom9x?N30MsVml|ycviY4cB(aK#Aq*YS(a4v_ zD;;<%%M4eqs;a6e;*}H>IBa&0jwYm}rE5kXJl1<*ylyWYb5=4&gzeY511pUU3}7Lt z%632&ahT)W%~>TH$zNRe4i9G=fSl&F_gdi4;@Q>+ZjRX6-WS2(dy(aL)|leyBavi{ z1Tv7s`j8ONLW{ErV125#GsTLw`#;+KF6@q7Lo6^QlQo%js};Pln*6Vvy1obG=H^ml zrU7=D2wO{5sU&svD>BOP;(VLCb-0T!kfd)($QlRz{B9P+ z{xO3+uNWv$h_(tyI7~s#{MH}I7OM2vF(!v`<3>+byO?1({l49QAHXc}jEsz|xV-9* zLJs(0=gH)YzjZjvjcAj|&?f0?x zt`kyN#m^9skjh4G-vcFuub2#(J91cjjD&_?Sh@ez@wZS0Nq0_Wz-y2p$oRHrX3*>t zN(x$fCK;ZeNqD+VPsW_T%dh4GL6FAbA)h^$@eZl1i%gH~!OqLn@YqAR4BE*W2QH578fM(Z zGf+lAc6hmGL#^viS`!1b-tJ$g%4iW($TN*rKup;5CxgMdZf}5&mF{}JvtML``%bL~ zM1vA2yHR_w9w8CY@Ms!ynVy_rx((@^F+8LYSuW;)MA)m^&g4lYR=)4~9yYuG&+k+w zCaaykzJYeWGAU6n^-UTq6~k|Is&o}E;8Pz9Fn1-$%6eb8ub`?><;9Pw0_lLf@`XY(ZQUBA2p}|3gF#>jD!=<-eExJ47StUYY&qf`sMqWnNJbNd{L{rOK z-&Oy;Wg&#Y&*`gQBv@h0ZoX0{yVFEQUjD0(G*+tJSO4d}6SqHq4ol9ljVl3Jtvn2* z4dHI*bQM>gLbJBY;f+21SQmbyK16QHG=9vVJ%3X-j3G>!egYn1d0^?eV!7R@{xF^P zP`SL@2n(Tuh8#91ztkg+eBwcfR|{u(jnFmQm=6h=tURI93Xe z7!I;!Buk;CeKkU#w>Lju^BgDyE43Nlo25BCDU8Kq?EjENSxabPQVk@O$$brtqV>5_ zTQS?@LLSMZ)|0st4-=`3xXLOjsY|I05NcU)tVn5jr|Z`bGbhq!*D?wU-{VX(StarA z_%ii@;uXA))C>ejGH2v4mC>aiU93EATMB~F?McrViO_Z!z_AW59X`)6ZTEE0ZpUec zF^p2q=|cl^JX~mBA%OQQs=L$QR;R?72?%R}D;v-%Ra4V4Qdqeaq01Fd3WqyAT#lW| zRldv*JtQNlHvx9QYFk zvI{n}$3kVWWWT>H9BsxiTE|Y?0oowfDrE36Ch=uR(FQF}*G9ORM-SGdQl6!uhmwcbKL_p3s+|J@DmRcZsg46AXceq~Xz$0(Vu z`N~T=cyZCDQ^i-lEjNN#XCr6 zef~6!kV#v?lTs5-=HlBd(jqs!S%>9v@32moBE#eeJ*&;CxM9d@xb&|(kPd+XI;p;9J(y|cDz9aLk`n{N$2Pd!N|Te zkiTmj;h$F&t^oP^l3YG!->Y|QCtV$|o=4sXmpZmT*gm*FcTAf{f49lMUnQLGF@7^V zy4cZqC3gU3&a2SR1x+X$Ei()C0)H}*rTj7~_0y0o(=r?TUcJM^;8&?!{aZsc(K@BE zg{!2v?u^&j6R(<*G%MIy!=dji4~j(ex9D`HkfYnHKt)-e2Tum*@|vHb1q5LsNq&$* zd`z7p2zAC}uk*GJXhP0`6dgm9!8*ADs_H8Gec^UVx% zx|?6L=bT!u`lvq6UTbTsHXH;`Q{9*z337TLb6qdPPCFBWOJ6Jme2byt^-ZP(T~7I0i%S0=@B5n%MI%S)2al>Aqe6-{ks+A=qC&b zec2%bjq*Zub;?ME2E?|(qjZc(>Y|eN@HGU$A5a-=UA)p-@QTc@Y7kq&Yy7+&NC74! z7ohRE$QZkQtxaQjk>GKVezFXalw=CDFF1o;Dp@A96E=QN%JAwfWsBuOagf^Ld@YWL zyskXeg?Y`-E)o}kFpdr6RFXcBrj@u#5s}w%Y;2-1dCAGic;SyJ zFDb1>S)Cue=iCAmMP{U6VQ;!9!Ovr)?{l-`a{LQnNErF^uyTVYyDap=)MQ7X$kOx$ z-b&m?YoN+lbg`YQHaR8%KM)QUYDmXYXpkNs1%fWsouKEB7;f09i zjy~ebi8crMRi0ZEK9{CEh$(OM2gAn8PHO^|3yD-7P@$seV`=ie(5{0}#~=!aD1~sg z-eO1-txy5zPC_XIzK5>1KT*5Bz9!xRb0HKmoV?WMH|vM*5<*#ITEJ(~=U9={4o5_1 zfrTu-HTrtsk+Place?dtvQ`!vj)ckQ_p1g5BJ6dz9#SipK%&_t!bABtmxAokg|3|e z6mmpWy949DdzaNsf$dQ%^bHNifF5l*h$QeHmt)))`aRqSGM2p}YmF{AXF-5$*}4N= z(fg~39R6pMxuWQuMX9-}r4rf~`|~C3*fer`^V5EV?LdT6+-5iF-b$8N#iSIQLyM+{ zbXNl@&uIW;?!{%%V^%~()wlJ7AFxeJJ-lSkUy({>4owGu^eO-vbK=S0bdYK8bha52QOkfBI06BuQ-kNGX`WWD8R^oIzEiyw80JX1X_+*2xJ!H^i+>r!v1~;;2^lQez*D;+AAc= zO%T$C3PNFi7kyh~?3I8@EIv0tzaA97R>3o#G{BOrQ>3_2)dtY-8$|6cTBrc;TAnEN zX=Dt@X2-{tTb%PY1J&8F@QH~>#;UX{`uotf-{SC^_Ja6aH)W0)q2tm@B2Q`ei+m2B zd>oNx9hJoE6RHWDvuk4@iqs)VYGMAZQO+&|r^R!BjxqNjdNj>9k7YU*pSjpE_3 zp%1I8JnmwS(02t$zl()Xf^E=DvvA1eQ7Xy~7Zb^aiw)I*Cg|=RiI4{aU>=!xK}j-H z4};KLvA|IjJO=PQIw@cLN}GFB0-4}TAeE70hc{0k;r)3+$3{v*!Zh!F{N+Db{XIB& z_sI=~iH3n20|D)i|K9b775Bk{<)DbFU16P{@9EY? zuJZ4)|MAT%0P~L@0QUT@wXqPgN&x+zPDPG1Rq0d_+@8;>ANiTU{L@ta%vSzVN@uy0 zZ8^|bd8eAgm*wDo|I3#mJ?WI^H+#P${|`}#zveiA`vck)5_p#BY=*79mAW;fK$-08 z#OAvK41e@}d<2+{I|brmV(M=L{}HwS_yvHiKT!W(h*DCYD<1q2_-fcQpgHvjexZFH7XfRZTO;>o{d0<<;$L+Zt~K%Z%EC74AA zK-Fge)2=L3k2e3>fJQ#Cao<-7Kpg&~@N^HpCW$xi8>a%r?B(7x%|B#wI}NyWz1 zedy9d34^lTeZ1B~2_8+9#0zY#x0)c~^E%XjY;k<4qEfYA?+p#;cL&_>;kVf`HSz1` zfWq;_)o(8jb{h+>fm!VkU@9LL-WrYFiULJ{5M~6}oyT)u`~FysBJ!h9uHJrMI1-Ed z&NyQ?Wyy9x1Wiae2Dy;s+4=deDPX*rm7m+c=9K(ZL4bYP-`95@>;7ZXGp#@7V=Do4 zfY)!S@^}uGYcWxgi3jB(F)Y4P%Sk1d@&p8Mkc!OZ`TF)ml}JiT9`=U91xO8M8S8$l z8cOXfZf|cdy&Vg(v0;dc8O5TdQ%YU$p`S-UL~H<#d&&_89ZQG4v!1kC3x4R54rK_t z-|TL2TqXdIVo1&b9m(s6Y#xFihZDbG13yo3N`LxPLF2Oaf~RC!>&3hL6c~rz-k=#l zP3FQLt~v$LZsv3DYYiJ2tYB`R{?+q8>!qc?@W!JkL{>p3}h`J*3!i&jFK z9_-3n>n#rb8q&>TzCoMr!tvs$<6Y=cPYZS!PmK}(LId4$;isRrMVVz8^x{d3lW@Lt z0%LJZz9z+6vS|`Yx|g9($+iN|uCDlP+^>#}zS&>AKqnQNv$O?f&=RR{!Wgon6QZM& zKTy&4WM)W3UqtUzK?(O!o0aTsC|~s@{A=Y zH=WIBX@|1UhFE5rmj4aSCXz$o?WJ|WY~iPjoTPUv^e6RS^-SsqZ+q(FOden)pQim> z_x9Rcx^h^{igK8^)M5$HH&G%rBjf-<%GAxDn(d6o;PtI7lYoved{%5=f`e?ZD zceH`y0_Ea4!h#Vs(PRV7EZM_egK96x6smR;EI6@gK1G<_k82^8M7aMyWW9A*R8jjb zOiIJh0s z--6jAfho}Ui;AreEKid0_#`Kfuh@1dV`M;@+qg!#!WAG+eP{2DA^Kz~f3L-S{W4?w z%Puy6ArrL|9V6POkt!cPqjN9I67F|>T5venJK=ViAe=Pc061=<*K%=Y{~2KKtIvHU zYzFS9pFJyqMoVMRNcjoLC|#2LvyL%vvEA6erq;r)O68ruOX4x{4k0?G8yK$s^XFw2 zGM9pkP#<2YIaZl7_G||6O07=p3s)BUl`kUc4*Ml!*;&ah1M0Lb~;;9N+YH z?fyo*7I~#S+9QfuQB=2?>=_I+k8|)*RW77Z_nj6+(7Onk`@8*>)1jF*-Rpsoi9$RZZed+iq)sM$21Vc6tfXpeke%16+~{F8Z@Z;hWRl zi+TTtJrLg`B6_KJ*>fEcB;FFP?~=F-j3p2Kw? zRn-C{{Wq3-Dv9sZty8M~FkVqtFQPoEh^#>K76=Mxd z$mg~!tjw=J@~i@{iPC!gfTM%?IN4Xcog(`cA0m+|zJ=GVe*!AcjBhiyaEyI+ z%2mRDh#{V#PakU?cy+cc&IdTkjemT_Km@GYR1lVwoOh9`%GBJ0vzy0d zx5;GRk29-FjnKv9emnZKe~OW!`)Kicl|}ckFTAZGetS^*aLQ#a=g=RKxs&=n{)(I&-Ak zM}^NzHAN~v5-47J&UOkcb0!>$$Isl^aNbYt*adCZ4GdKmHLoVw6ZGy zRDq}a(Jun>eo6k(3LY`@*R9i67Z{K{sn1P8@oeL0Yu`g2826tk9yP<(c-9E<;s8-@ zH>u_Tm^NYRl}<8HBI@gngbeJhs{A5Z(pdo~gFImeznfo4954RWAdIQ?i0oRu^%6bGB4X z${hbRWXOp0tv}D&k8hXzPA?x?hMhM6<;|@i?K*E>=P^w_hvK+-i`j(N}#?yi5 zvJhsiv>0UE$c57CGxr|jSZMSsZMZyM^}Yg)fYl$f7&O>DcSjzMYvVhx6{V-?u+K4Q zHrqB8d`<8LmA{9_ zzhT3H+c7%hr!f z3en$z69_{-PZkKHf6o%}#*|hW2*jJ~%#FHaOqmW=n!FDr2G{uH7~Qi9fCiynOY))y z3*M=WuVbs&L5{B~I{49OPsq}*;px+-1C5Nr4&&@EE(@$7>!8jWZjL=ct~J-#Q+*|{ z(mfdBws?kfc4u$~Yb^lvUs#j)e=tX0c%iFDKs)vGPAq~oidk=4OoIDq6soLJtz^&D zU_Zz%X$2an^+Fk9vEcaByiUKnmzJ`ph-ihBkyuR7XV#9BrTli9a z`?{NH_CzC482wK3&xc!m-m!1baO^V^$v?$%tUo93-@K15Z^U}3jwf_^d}uMpThT5> z<%Zwol-AJ`3t){`T9i2g3}}*R*suq_FzqrDV{}*?gVYMcE6jVB)z$TPm{5$obh3lH z?)QZBeLcrh73SV)SD(ce)imtI?=1MBs++?Y+hgy;$*iz;Ge%Ys^&3#2Gh(M&~?P7y_k3DNJ2gM6F@;0RHhn&#D%}u+YC6h)k z#+SH>`^Ifo&52Dzsm*YQ$x?fySt09w zLls*HP-c1n%+uR@%Brml8;13`6D4&kd$MU}B_cQivJ5q{H0Io z1^HZoL~h17Tkp!p^X!kB85?UufeJc)0vkym8}C?+y-v=enkqzLQ46cQGcf5{4e&`E zxrM$uJ6ICMgBO;1Y>jHCrrF+?O|<_=|0G$n+qLCOh(FhwuJLypR>o%zQY6JZVE?lu z4C^4Pl4SfJkjjrJIUVmF({BzBEnQ(lu~dUA3US)1|5#X_R~wdG@ajCH;GG@6iF_(m zV6aq!kJ9Axb%Ia%2rys19hfX3=O5^qOL%%LReObJEGk+*npmK~!$=7FHStI#?eD}t zuF&P}UD4?1==$T-uZOU|=)TR+(E0muW8SRdG+d5 zt{b9~DH+fR3uv$JW5STH{f_4lJM2p-l}q}>@Xl2g@{+_m2SD|uNs`nM!f`oAI}QF7 z%FgRNEEt&Yqg<27TMfr9qjoa8c~nIKxGh zDx2{&F(9?UeS;PocJy~JVH>&D$Gk_PLMWOUMlxxdWtm%_VWUBm1!vuntz+wD%rFh@T7OdC!2;V( z9G*~iSJ#7R$#<)pvLTecujq|wctZ>O{_ElaLX5k(JnON-UrRI&marFK;p9Xto9hDh zx8xG8${=UFS1k;UZjS0~1da9f1gRDL(+AZNwdxSs?4H~r5fHS`>$*4_9>_^3{$-KQUKdcb?--A z+7g|e_R@&CE@;F;u?c9K3$w)CyC2Hx+SwJFnc$1EO~N)6d>4S-HDRGn(khD#qi73g zzkYLKZ&u!!1%mpl_b_$dU&Tjbn}?i+#=eGxVo|fSKl4fbOb-PVQD@xX#?ltSy+rpB zb%FTlqHoPHk)H~NpuQTiwf;@aVeGS-mxIF#7kuEVIEaoJ2d=NLGq=r-k?U&vG!@N^ zpL;}P6I=0CcOQ={$h{efx%Ks{e)s9j*q9*V z*Xdd0Ys?4R123PRtIur)c`IGM5N&8^pnQz{nZ6dX@sA>e{R#&L@Y`ySb!X?dFVEj* zXkYES{CVs(UI^1Ttde;BM;7okfKAHXx;4gtuU#oJXd)X?hUW}%chwULMa?||?+d}F z#Bb9?oQoR=0l|GbVPLNEHLm^ZkKXpXzPV&hJ=J2AYtJE;(^?z+IA{(u8Vzok!lFhs zQSlG&z4Kfh5tK7&80HsSF5QTArrXOQ3`#`XJ&i2ME!~7R#G9b^gfydA-S+76L^J~L zPIe9m{-N7zJd221U%dpyTw>xz`Sy^a=xI3q)*>5N-plimUt1r!WZ-?r$Wy+#@*BIj z?ks0p3#U^gUw-wE+U(1V!3?cErnWEi%3Kqt24cL7Vw<fYv!uZW7Htx=4KBf6 zKW}d6p@>_#alrUoyrB{2z)IlwG!|`U2w3DT3ch0S$4C>Oy_C`?(2%h{0E?k#j)GG+ z5=AY1RR{Mp@1)ecZSiXmFdX^4{0j`f+M9VmETtW6Vi}Hk`tYpN`M01fcY&!kK*<~H zfF_+Zc^{JW<9V5J?eXN&k2jVHW=49&x}Uc}iLus_u^)ZpwfP%RV1NWum4Z+{>EZy$ zI0gB4kZU1jd!i(Tb#4M95^j5lW-b##qU=s9BGC6N0Ujhac4G&2uHRWWu)Sf*7Y!2( z=CqnFQQ;MWyAy_Uns;>J@1OlCZbbdJh~QY#cdlAZL)tzw8EKodMn5m|vAy8X*<0fK zkD;Oyn)ja!JUmSqZ!>3{-6+stD0U_aM$1@QfwXN!|1etdomV-LIQw@l?9&U45RZ5k ziZgb5Yhl5?$lNVU{~whHT|O_9Q@^zLK!rxo#tIPJZ*UIx-S(`^2=q2RXr5PC+TLg^Fj-%rv>{Nj`e`*0R+r7D4z{Td7Y6>-| zSc7U}oho20fCu;5o&I&Dcp@ZTMA)T_SDKW-|38`=#F}oTMY)*|Kxt~ABA~%X6^`Zf zqsUlt`FCm^r_fLuPYC;5XyKqEwZQ2Tmsdqk{aGqDJpVcE7UW6|N&lM!s)_^8T9En! zuDN@HARZgm@&2CJfVku`|xO&-fHt2?XqP6VuzQ71-J^A*b|;{s9rgyeuVup^zADjpR( z0v}YB#bIcOz+-YJy9}P zha|%@y0s2^fD`31Q;e#%r>7WF=6SaWB)C*SCc6p(K2iFcEQ6^>Fx6QLXo!oiug%;;1nS@;h<e6&m8LF2HA}qY3Gv)QW zzUgMh!6I3V27X@pw?CK;rrLP~!83$io$jrkJ@U&h^j!foHAGg|*naxE#dX$iAa1B1 z<3RCIg4Tb3Mr)MD_e~s)!je7nEcO3(XtU0FA55tr@3EWbaYsnk?P~hh=i<#p5k!bw$EIh?PDi?<(g*-3xbtWb3}P@cNuS8OOSG zlRikPY4GiQid-YGLI%NP-H)a;=t%dvHuQOz>~R-@@!*L*TzYi7z7l?yOA2uJe3Sl5okF^hFqObRcxOQ`gG=+3)FY#c)6< zYxzTc3HMz*d@tRM2tR*6Y!@iGToK(5JWpAZEOpzs38^1zr=E8(_7$TA-_172AQLp; zPLe4dmG(a=<~j}g=s`3bP-ByRBC}aB$}mrnx_wD&Km>Xp9{m|2c#H>+(^z zLf7qIM5|(cl#qnW+XS3K|Jw}X_TusDPl3Ais*P_8zq43A4=qemf*Yl~3>s-o%a&YW zph!P5SfohWgWME3gwc3CAERV7P>{D;9>m+ZWyfN!KV~AyISzUo4t_}DTXy6%smwce z!$)l+w)et)u3pJ*cmQ#I(8+;5jP-c+#x#(L3P_2Z3&+7;_J|*-F6i}0~i!hG#SL)9rMcv!O{P+1u_I|2HG7=m4h0jv*c%)6Z=p|q&0L4APqOcoo{ z^!qmYs|#u|{0%K^?U0NoleO)Kr#WS^TKp|fPji7rBSm1uT&O}{wjK;{Rhq|e4)t$J zF8zNQ3T?NSTxvtN=c^>G#0Xo8`}f()Y*#v?(6k}4iS*zSrwpo{W)gnJcz;iqYhv;L zizJa^O8%%R3K6!^cwB)4u6vXzgsFn{kVg4+6mlQN+a_~&EjVMl1rDZxXG^2<+;7Rc8Opc#I&5C>G^feL?o>`Rc5VfB}RZ{h-D(T3KG>pvgR zs9G3vZH_U)wud6*{v#cNm4#T8P5dol=pAs4s;>jsu*xQ;B{mat$yqcEO1?eB>AYDA zFVtaY9c!{oQ_0Ns02)p9L(@e1Ezi6_n5Jo(gPO`b2cN#rwOHNm&uc57K7Gh&eZ>^B zL=84VH2KMK`^NfuSA~6=>=WP(zS^m1GgRFRzPn;Xr9?=!9xguHis=WYC-hjbG-z>P z$KK%Bzc?(uexop98+~^!y@dhaq>*+F6ywqo%U%cr=1JaZ$xNA5_5wY(y{U{3);g;_ zNZYMsHLk(!v;B(6L+63#Uk1EP%~9VEk9vXhAj|e$TlUzQSi| ze8aW@@|j32-Neg0uEaIZa+a8XcL}qnc83fuSV6y}%gsO5bLh4H_$xd9H->AvfS&en zRm4pnx-)+zsNY+Q7Go%3Rx70*G9pdr@>L^9-smORI4GLG9pfi6H(sX42BZR`JQ2~E z8U`ng;6FFtrp7^ssvj-x>Y;^{5mZ$+dvsNy3IzWy9X;6HS5IRVQ%l<3DO`%M9G?PA{vV(H#&A@w7z2*IPKQ*?MKb)fYEs0@%JQ_8lRQ#v)C ztE+=H-zrN9vtzer_YDnMGF+%?2Q*q1vJkfQg(zFvkR%#1|Fv!K5cCjLWb#pIHFmjO zP}M|n23V$)OtpBV!(2;4pRJuzTwPyiA*mpyk5nynKh?p9HESLTQGUC-22FnkH^X7( zi_TZKgg3h}$;ESDZ5_+!#f>`jJ@fb>Qq@ly>pwRypuwG(@dH_;=Qa<5Z!g#;vhpZc zQ;_wM%aynxC^vvCMU*59$2+z?%(ri%pMOd_-K$?xT})dGJUKtK{D1GU24L5e9LTNK z>}!z9HAc|rMDgl35WOPBaFjdJi%nRTA`s@Rx?=FRTQ+g-XL)m$Wcu2Af~8_HLqETJ z*oD>1CoS_^%I2z;tFJEi=AeJ%$=%jV5`_zMR73$Wq3}&;Q%zA79<xXeTYn;qpq#y8k@z2Jl8T|K*U+wHXmo5;k*(c?aT2KF z!Uf6yNTgL0d2F=U{31Exfa7bz;hdDjv1IDDL;E_lWheief$mnxUK^!1O6JiAO|~B?bW1l@=@3Wzi%FOFK*&Bk0bXoN#%{m)8bm+QA_1dX1^#ls9&DfHp>Da_uBzi zt{Z+En@jqiuK@xH2;<<}qF>Itn=*J<67tQD~j+4^J-xXMEuRa zfhQ784Q!QYEW?C*fZUoOpAm-(g4pAHB zWO?h}jo5Ch;nm3N2n`(I~pBj_b9|&4+{-Fe}E2 z_m$!JMfQthy;=_o1`zoOqow@$5i$V#sPOdLn{+=knErO}XXfh*g)A*#$55`faDn%TXIeQH9eZxczYs(41vK^IRl1QM+^N19k zQ(_XUblgo1^tP5Q*U}pdnu+yKiR58-c2c@ZDlTIU zklVHUe3bG##{G>20LlGG_|+SWsnF{mfr|y06>R4ZC+D~#k-yOyg=zXt>_ED<;^pF_ z16GUcZvpS4Uq(H}E%_TbOWlRRda~inL9KWD^g>rI(PyB`)J(N@a~unbz=v&eV90pz z`GQDIql{V{{gdvvKW z_!I{&^tu&bUH-^fpCB3%$_=<$of)Vlb$hCzgR%RUDK>`EvH@OUF)`eKB&a8tLF3i>3>WmCxlo(8^NM@Q4LDJrE%p#llPId zdX{7=@Ou&eLER5XB$Ma%LNa6hO5Tf$9(1H_SPO^^j9Kl9t3I^TA(xMO7>385de!tB zN2p7cdi~qQ<_yPs6`{YREPnb)#ab9pL^ph6{KuPBpKOjPxIQZ9c@}kxIx{vvh_anS zl8r;EJ*;@fwe6-G=yK<8Bt5rLDE)sgi2nsvNg|o_rhc*FN90^1ri~$b%^ym1)+1@3 zxE2cis@ph~sIGfaZ?E1;D97FMIGFY65hZv0(xyCpASp4#x`90f>m;f5%PlZq3Jdxw znrps~h)_%LLt_J*p-z^+{O1+jQlB^{;lS7FDou#;7_0^CU#{))RT54YX z$i2pf?$X}@Z@*#sh|@FIdd8m#0r9xTc9>yt2?Y0$8!7r9+}c0&2uf|Z&i??4kDLY? zd{bKfM(Xt+w6S<8z;7(YY-S~nT7p&me_tGUq%4teNy3#kyQ1?QuOJEunhsEho)bN8 z05n7}VE|!N&y+k9CW7lf@RND_`L)OAZDgXQWd`r(md%#V;bp-$etWnAeSnbHn0K?s zzPlPD{`etOg7hUy1v~^V;~~ux&Z!XNkY`lam@zR80)Ti#j=~prMDTGRok+o=67FA+ z(%Gj4C8~HL8LFCUpcMlx_U-9z&U!jdYoY`iYNFdw5NuSbPJHyX6<#+?DtOp$OB*Un3%9^#+G1rBd6(MGcBvMV~!At%dBe9i4Y zl1p>`qc95xaMGf%qnHhbHy=iotQOTz)aTv)%l_R zG{S&Ocx--B|1q^*5R4Ajt}AzKz3Arn4+=l09<}C7wgSwT2yS>ny)1+DAq!uHZVrY_ zYb|Y%l}C#UW)o6GkeqdAUCQBLa@H!ga40LA28fgyrXF~G_Dm?JQ>jseqTly^fI8d+ zqc-9zuYrx@>-l;U*=FBW7LUKktwJr*N4qyacHjBj*8@jP%zIYHa{vL-rn~JG6&RUxQ)`x^nVXC0w8V z*p`^eeu1d68RlHrQ5fkVT)6ajjsaUpHz9=6))n{fY?KnM-o;Ve(u;MQoP?kpkiB|2*zdIAZ$6-^(!ZFWk(J( zr2Iu>btQ_8c7{rWR`t1qf3=TR@*0;uQVaOkG-8b_0Gb7ugZXWTLjb*8PJx>9``GlZ zeoozD-rD}pN0|Ec7Vc)!LB^#i2}CHHvnfU?pq!pHuS=;N_1b7ljEWLH zcs3}yO$TL)+WT`9PeuG+Oo`&H{`f+=8>iNf_7M&K+zc1E%_LXC(P0Yzd&?^UQEcwz z!r>t>FwQih1PfXryPtoDvjK#O)#S5HZAEbA4(3h$dr+kw(<9@O~xKB z?RQyq0T>lIQTIvXA^!?t6|{UQm~4kT%VgFuhn~dgB9R}z(EOZINRVdgh2O=HzZBwE z`WH>s6!SIww#{nTG;sJE1CLK&10v72oMSsDAauW)&lu^Kb`4%2XG=vTItgEW3W+?8 zc;qL3NIl1C?UO3`+~Us~dh>HE#{kF`IMOx?Ggw>JcJfvDwgrDNNtg+dPWQLbIZn1d zoC|VJGBXL)@Cex2IoJ<;Qgx4Kr|J`KF!3K;))7}(-`qs<&ovB4-5g69hf@fZBrFkwm1D%MTX>Y*!dMZdb3AA4f@sC{*a- z^C=C#Z!gErTUBKY&`LG#6-SPL^91y1a&U&+un@wMf>tn^Z^yU_H&{bkxxv|>Z7n#4 zk^U1KS*^ie!bc?Be#D=8byD02qK7$=oZ9}x>FIM?6nipzI$hfb*&l194re&uXXKGs zbXJid*W#sI`;k?*%u*+lYK$*D_HStV6Zk}f7spz)h`ZzHbT`Bs)|@#AdFf3gg4E}P zU(!p>A~)E>t5^;!E?@Lke~@lB3VQwa?S@zRe8Ug6!_H082LmIQU9W?WKjf`4-Ikvc z!tL6t<)g^1E#Hs(ypOa1v!~!r$Yv8DQ~#3v8h2G$^&#d+KS{V>vE?}9;R6@a8tE3D5P#RroZC59 zk*kF}nQr~q;DP~$rT?(xy-0y)=;-K0BIVQv-s8hycFOcx(0Gm#Ny|~A0v~mrK-{Uy%Aa^p=Q^*#K&vd;K`Wq-Y3xJ5UF`KY&}!Ox}fu_`w9)Pb2CFI zX6k=F86>D!siLta&`t~igJk&hDaIY`7|4hPzIomdIsQmPqU#`L#bQO?l@)jyh4Av{ z@A(>zJR(+CCAA})QhTgws*!;<=VOt&?Uxjy`0iEWSZHf>a>S;QyF=73WR00xEP=04 z>a1p(A;lc+PBpu=;nTx@XH~dHLfE6j^WrGURCW$2^&fq0$N7>bSg9Zgc&l&XwRriO z7=hdQ8HegbVNp~{}4(U+a%aSOB3AxBLHTn9puMZY;@oZ!+-uu04q3zic)PWYYJfQ}Y_Oc=j~#LJ50Z zBj6r;c0za7i{oa!@&v?Y$G!ObWcto6@OT3C5{kXHBHJ7>s_$#QT-5Yje)qsk1wd#ZfVff1N=Agiq(lcY5c(AdENI`dV?2^h8bD)+vQ-5Lx2Adk^4F>7?>Zt6FQ~{PImp7SjnLBjm+g`9Ss$WCmK>{CC^yi+jF9m|9sg@sSY_rAvXp3q+)xBf}(ioX3#79cT`-mx=J&x z*HCsdEgdrACYq;-n!kJlCq z^%uY*Nu0opAG%8<4~)*UCn{!E*@34t5#5vTu+#Z1@3m>z3ux>>w?V#T!y?Wuo)Hau z6hBfs{V|nZy~_&*j2#e3A|_2E3EicXAHgAXn6d*Am)z^HNgu&`fOPlo0Un7wUjVgQ zYlhdXod_T-j43sCBcWa0-Nk+;=<-Ejg|xMF-`@|;A}y7JOKNjI1?D02ty-V4+7Zit zzJB&J&#Oig4VDk=cs8$ZtU%;Fm{9fLeDcE%YS2LVQcY4GYb2RWBuG)1`vC@w0XL1i z@PTY%oZ7|ZH&nz3b ztW{u!j~%!^gp0k-e=P-Osl0-FO5}S4qszaoA%?G!*8pL1gS>_!`X13^$Hr!S1e>?!Hu1@S#li_XQ0x;cIYQg6!a1ag(;7 zmac@;AzZkkrV$jK6@i`SR3CJpH~GWYU&L;j03sZ${9tSG65SA%Qxrd?nfXYIHd7;7eyELs^1Ep;h@Tp%WaVFa= zEiUw2`g>cGCSVvI<$D@7O@5p{i;EK@j8v7pWjQ)n3ezhyja?s3iNLJ4c$?*foPJxH zZ9uJ`r)}i5wdN4QP83%7jz#BO+~kK=(CriHrX~Xk+jQabL5UjYfkjF2?Z;4u&6gyT z4S|9|t+B*g9d#Ay9T+$B(}{<-rKYJX+|SR?ma=&<<;6)Ofqe@@&{CF_Wh(zQEAYmt zPlhvqJ6Q_Oj^S+4tINcQ^RIkN2_|P#=D|FHl+Ijg&X8J=M_II0O(u-Q0FDe8?>)70 z7)DQG%16n^i>-kKp!={#LuCn_Mhxp=If~bpf@~g)Cm%0=SajTLLW5bsMUO!gNC#WK zZymT#fwiL@}ge6bL71ltypKyi;xSrZ)#{7WHXplC^Ff&c>d?|K~lv0wNV`%4R>w;J>q(9>P?9~PG?_UOO zG}!EEBonRzTGEiw;dO5CmT9>`fpMCDeIA$qwa{;Jexdrwd1RhnTl}qVbh-wF7^QW@ z^EI0{`=EN!A~d+&3Txh!Er|YJSS43N7;9V_6@e4OQ4~xm2-nW$NCe8XRUA~c{8d-< zn>cyPgk6+Gn&>P`@j{KGLD7A!_m|(vc9&G_$G5A1{yLP-lJG>hfM)C@PkK!7Nv&L> zDTsgAE-x>$vP}RkZ)Ay*YwAx=p?Xi&y**kX=MEA2<{~4=`l6_^L(2C(Zjf(ba{ZO~ zVP~$!-qOuf6W)tH-jf}4*~cqOF$1PfG~j$;Su_VzPuBdH?LL3KH*^o4#Tfkxhbr#V zF>({5Xun6W($CZUAgimt!Uflo9LLK-j=|aR>ACEwBYf+1(18YzO(pr&cSyysVdp1? zUF0PP>zuhM9fI}B+xl0_TvfPb^Yqw_yWU)&w_U5l(r{T?r{c-pPo{{Il{qk@?HeBE zJMngnYiJ?N7#v&^T*!p$^F^^1WW!j*A(G{&=Ltn56Ce3(#qT#ke(MGr70c@_el^&U zqhgW8`gxLfWg7H`#Rk$(0@(S?X{AOiWXD^$OcGdSO$lgZm+xfm3o3tXjEa1htMh$&_dRXB#*aw#Y&9d#leS$xkE6@x?QUS&3< zNe4=tWw0B5e;Lg-vi1-Q4BPX$k9Md)4>pYm5|4ZNwzK0V5}v#0u}3oo98$SEiLH>D zt`n8B56>cqw^}+sE8TmMqw5lN19%aT7E{&hcuyoaE!|>D>!^7~ptcJ-(^4Q+IRX2| zE@%a|m@Wd%2$=ZzO{5)5^-N7Ox`uPcDSZE>H7;(N#%EW6gew1x8HWU8S&O0yWAHV9 zFmFG3GA8Qe(J`#H85BGJB%@?sH8Cx$vWSVeLJO~`J*DM=!LjbX%hTki?fZIUl9^U^ z0^hUs@l4_55*Ttp#5|%m(}=;hD;lL)9dIL9$cQG2ucew z!uU9U{R%T@w+zLD9^_8MzWt=BO#V|h92}e@{UVe zPPf1mGn}Qrv0ZE$8NC2Saq1Ro5eRMsyq|qJ;AQ^%k(_lVy8ObDcmhw>?+88K9Yo4f zz1bxF!e*k*z0^*9ovlvs<_2CCCsuWC#eDrq%yc`yk1ca{Wa>P)xy;05 za8#EVGI}NlA9T9e3?tg_dFwc9qV-iMxqDf)mC>cS-=lb5v+=n2R%zWR{rvMCx-Wj_W&9PgXyO{n z&amSBvru*xpB#v-uiFOMJq#ls1_qE42rG1&siO6fhMg;1I;RVjZoP&g%KYiV8u>de z|26swzMsR6L>?Y;#EZHvQr3OJ3Uy+RuGBR58B+uMcWqadmpH(<1RNy~-o0oG%Jk

<@{7sm>?jIHYi2}ZeO5+1)z;?@8CQVOf}Jm9*jvN-TytO2hgTc%hP z$4J-fQIOD*7pukm&0JR=8Vt;SV&90Z%5YO0!TTic&U)H*XX+*fIni5T4Qs?-S4B1^ zT%};h+dq(RpyIL9x?MYUxxFHN{yc7817sOk+ue&Q3CcC zBt?X7e4@hY8h#nWi23+wWY`{M0b5_F4sHqJH>)C zJ+{vyO5%Ny_K`sK%oML%wO5o$;S8+iU z8n$-t%a1fqFvz)t^1P;qV@KQ^Uv*hig|QD46A^9qlB<^Or?-55zzKMvU(g7O^*p$B z;cZjLQ65rlC;vNls^DQ`wYi2+$P3kBPL{~+J72+~-L{RD>7u078Ipp}%!qHYx}V$M zL$|xaFv%(n-pjgUK|HVE4zIz?Ub2p~DDl*t+iB ztsLoto0x`A>nPuWeKDuiD)bvN0c_>0(JTBWX=-6eT|Nx@u@dumc52K(i|1#AY3$?`utuAr zV4yzM(m^~y0F$9TLUNG1Z}r*XMr~q!%+D9J9V#?y z1HSw^qc0V}n6|m@2h?;lc>e2(pCFN)77(jlC~i5$`!R;YG^uuqLtQq3`Vizqrokl7 zyq(0XGLOvB6GuKkX+9f$$FFW|g5;)DIOBG_dj4>z*Z(DuPCvK?ogMda05dCCdqRM*wlWHbV!Z_Z2BXD2 zW(v%}eCD%n=Z^ssiglQ+;h+<=`7m<^^&=@Xs?N^^10LR|ej$l>%p*EIGMBT(A|gv% zB6XJ!9QjLry3&ggBRB4|G^8k}TQR)7lf?FpWNkhCXhX#C{Z9%x_S)C}-u1)3!?f1k z*1j%H0SYzpnf+lsdSunZNO-^##6N%0Vb?!wKb0mp#Q?7>7?DgjPRac#MTXr>N`erz zfs&Y&M8f|0PYso>wmAdj`hZs{vB1PXtWm`>z;|IBFKic2K5yfqTYjVy9)QtjrrY4Q zPND9k(+j1w(_ts8Jpr!B--=l;Gvk8`jxQyFsXN*{aeK0?9}GFOf84+m&ANI1L(}J5 zU8h}pOgwFgJs=(L-9T}zN^gKhWHVBScTX$Zp6NHkLH(!IC~0<;z7r!bP)h`)x%TCp zoEA{9b(|uIC_~bBv=}_JaE=BI@Zbk}nouL6L9C{DhpHuJE7d0I2HDT2nEestupVy` z{1UU=;}J&m$Op0x7bZT?q@6QN$JD&F_+UCj2=5i-@Q=-XAm70@V3Aa@>hhpXdMk%t?U^}UQ=3+= z47Gx^ou%`^LX+*`O(a@kN=gdDA59I^@EfG>7k~oS1>G<6b12caQ@z&)5?#Od?PLUCI!k?&Up_Z$|!(q5I0DD6%r+xzu@W+V5N#cr&( z+27l)mXk7hjb=?8IZG-GjQ>FNx%r)P zw1+(yQi!xxyMG)vr01vd!{0aNms?wKaMppVTb4!5(A3fc4H$!%njC7NyTY>B25X@t z(9W#-42)1M11TwKz^BxCKa_m#2dVWwFpEi`8dKW@QQ`O(qAVi$jzSp&X@gnQ`)=@@1`bh22%r!6(PMqN~%dL^Me!sAgeNhSD+L=q|Z&Tjw zG-7jA1Uc|P!lO{3=P{AoxQSd0wy=PuS?@#eV;{hp%y~NtI3v?2SGxO)%>V9pH0ULW zJ-Yht0rT>F_ND9F#i=3*PNb^y(Z7F%>f|-;u@tH#kR(1=7m+u|Xe~&p!R;!6yy(j8 zyMvl-x3*@94n-{A8J@`Pg=+fYFqY&51C_~Aqv@nE$K%Os`>G@~pZsdYsDi;+Jdp@*bC7u7FcFwh?C46LNC%<8iZJ2q7LE7o7vfQ8!lx6c z1DKm)!*a#|Mk}`|4yMjiDe#SgsTEG=|CRxCLAhE3*MaE~zM6?c6z7hfr*K=@T!`-iqa z8khy8dXx9BlNJM}y~rkK#PfyP4Hta*bmoS$cJdspd|B##Hif;?eKxKmSd^cy2fBY_ zFz?-!jIE_n6w}3pKt#%iQO9~uUfD#cFs}5_Lbw9{>w|dcotaYyf~K^>))qytoSvNQ z-Pe@-qY9i0!?w%*3uNi_qou*1QtnIOV zp#7P(IKzCC{gtE3DU-hZ0R0?5)>D-|hut{Qjf*>yy$=?EVY(MN`UChqY#yV5P5Ok< z!v&PHt}thy{xrFkSvR?q|5}42+CUuuJzUJE0!X&yEXqVRwmfo`TrfNixszz-R!a=O zz1U!6{jl(*t*uRgZyl1D2F9%mp=9mfk%sW^5vg7t4zLIyg1A*c~qhnA{6kMgdZ2<2f6Jhjb$bNm}tC{@t**)6B$ zx-{J^vkX9+a5o;a=yg6VTqq}zoGxfzbt|$>8{7_?`#o-QJ?JP%-(D(<+w1|-Nm^^0 zZ_iyKj2&_1p`Uuz*Y5kU2?v&#UZhQ9kj!i{sfz>CJRM$!L~pmwN~qV$tO?2g;=(0P9D0ng zg*yIxnyH@Du&Nf7|Hl0Q7dR8+Jc3a(QiWX?CG`ZkYMLD;INxNCzS`d>hTBn5yx18? z6-1YPS&sz+VFvPQ@-cxzXDdkU)+@>W_!G~%{Q>P7lKGwCuBtaj5#I+>l5-><5W-Pe z_!A}^qSH|c6?=NEWJ!=Wd6NlpJ)Zs$mTp4t$7TX235O&YF560{^QXKVd2K#}q zI%Z~SF+%x~G_BUREo5`$7~$yf3|v$;)46uPO32aJk-jnU3)*XC<;cyL-vp-MH|E-I zfW(g}HayoeTXn|5w0Bi*{P(BiC_$u~m8V!FQ@tzlt#bNgTaNl{8((+T-&(F{;cuG< z35QWoPRsYlyXMr%#^_~kEcnqi<<2`%&h8QE)80z2(BRE9@_C&TCCNuVkfia}np5IE zlM+Q1H9NsH6@@u?0U<`D8zv+v(b9&HGSdaWfsLd?`Jyq%rG+T6(aS3n&7`l0u^bJi zlKPTta60F^|ENfsl)_`x%d4x12p=U*sixOXc@uty2;o9B`1~-)Us5?DOk4aW$~S~* zBWK+m>AAFo82gjh*yaK)ougx4cse~yZ*w+|zxrq{IbVh>^_J2}j;oa%&Y4nr^>aWD z8f_yvlAcf#JC>sqw;Oaf=11g_tv$`;jRUK!tTbfA_j)pB(oFxv%gUtANo8zT-W3go z$#(L5H(1f5*3tiw?`>U?)5&4DOTy9@9v#E>zOJ+_>m|O-A5_o7itYI534uBYH!RUF z!x9U+y>3{@#w43J0#e~zSRt4}m~S7*jwcWLKI3-+HD?DnuIMnKqav#<HRXYdF?(}y3zZW2FADZwJsJEm>wfw9Z^K#Eh2l@n+4``eh!73!*+V|Fs_+ux z?-E4D#@qVKuUPbkpY7b=l}1$&w3R0X75>R&6@dUzIfl4q7blP#wQ5HR6W#| zVk~(qewNGUpRen65dbKNLQgW4h&Cz0DYmHh-pt9?k51cabz1>4)x=4*7OK{^4AmNh zm8cOf{stdXb>5U3E47n!{G!8^YMvn7Bci8PT`9MPYzZ`09)q=q=BJ~3vi9G3Y;>oe!f3zGlb>1%e*dTzi=vm$iXBUw%{zO2a%I&}$zkWi^LF)( zPd?Aiid3>vG=1x&$m1tvBDHu&l*mTkO6oyJQZFId3!~ur0T92;xsY51fERVZB~sE1 zU}Eg)XS#Qn3CNyenHM`Jh#(<$N^_=lJ!Q1b@X592 zDaMygtJ1-T`5JlHGtx-@G?K~R=7$*`pbnmOk$Y!1^oK=8)|pA8kmS5!Hw>9^hkIq# zvkxlhNL1q&U?2;gWl!lwOVgD`zmvubF55!O?f8?a8;e=SE6%!CO(!f+3&8_usfpS{+Jf#?YzI$2`-s&l3^y$scnC zv4tfa?MfpQ`?tdy>JH(+Idp0L2BS!3nnV~YDC%afR&di1BH(k|bHLIk-6IlL2 zQ>rSUsU!nli0S65ppRqEa_t z)qz<=pNFmFKoVe^<$G0TvsWpd2CSYA8~pPgLfp+noJ~TU&BT~j&a}a9d1@l@>-{OT zrBxP7knfh6Lajm;F}bxRuYP$H%k}p%AtY@{P``ZI7 zo&1CPl>UrXG!5$;pNz;jv!|P zdNa;gWVoh8tpaXPm4KpGRCbrR<-Owf<_w1m93b@~su|*>VneP`^l4z53B)8PLJ{g= zA;6Sc@5t_Q&N&Y>3yG z2e$Z?ke?CdA8ilEdrn*!a=PsaS3Hcp=I$kDL|)97)wUJrkjAg!9URz3F=(rYw|`S- z!8k$oVko_4dqnC3&76~CPQRea!^5ND+5KwIgS|w9@Njv5%}U=G_Z(9^V6c1JT|N@U zox@)N;Gp%9;R)dAT*9z+YfqW@`K1aMG5C<5c~&TBwKv9px3NNkf3(2!b%J23DfLN2 z)~U}7?+n+`c4Em)2*!GX`b@}els?1YY(t0ZW)7xV`g3>4o57}~ZjT8owG2}<_m8NL zm_JTcYbR4{4zOA5id9(kt*#cIO({GoMK25?00DB*7H&?DN znk1SDCw6Z_U=({C;-vN~>?Px?kS3jx7h5X>0)iZkf}9v01Rug+i3{f{1t{(8jhDpl zfBMESQLGbnv%ek|e`B}at6xJl*0Eglok$U>cJPb_Fca8({M= zHXwnCQFGm5HqQPi^mp0Yby!es#Xs@Dr1IThr)v5u5@gG%OG?&yv02kkl8?Vwru}F{ ze~87FYI=vc)xvyBg_($FO*jT=<8!)G*@@xuChHS4l1fOo?=R>mcIb$HI!2C>{b;CH zcg0FI%L}*2BJ;+S;^TsF-q8pVPNt~Lx1(A)TzE5Z-+KLg?3LiqOqZaG6)0NSqG`*3ILu&8@<_nXd@#9?*<(0c6;dH zQb@jZz}n^-zC(HVnk@`vWP^0lQ>@(7J-@|`anwR}clfaVh`|DPWKn5K$lN*G^Qobg z3DaX2h*K~Sscu@CEeSW(jb$r%iuVwAQg)p|jX@|(2!~Q<zGmUcKi9Aln51+#`_s>fjhm;*b`ma8^Hx)A5p)4p#VMtmyg$ z>kVON%MDQJ)hL*Y)NZv|**125xrqAZlarCwUc)F~9`=lbZ2>5tn24M?ll=L*b#ZP} zbSnQzrhze*Ze~F$*ITDFTiTty7ZB?Bu?=2K#u% z;!=3n7OJmjCA-kEyJ|6Rr`8Bi-Vc9dLh*W?tFCq&-!MZw0t(I_bv?&Tk_Yj1ctv(S zNt<$deEmE^^+cV)XFi_~35YDSj>hlmYh_)BXXj{9LXZK-P})Xz6o%+V*dY!I-{k}< z5w?tg8v8pNbZGo^iZ#BHstN^+dv@)^N}ACi2IR-|I$$j;;e}~>)bQQ0^u}r+c zvN?xCmvbpLYgiP8qLZAKH`7zzhjy+!?gG-rs-;^Xz$W(_kPv)#1hg9()h4Cb7@tQ! z8ptE}AYmcEi%l;&lz=*g$CIg_KK)ceb_8_B2Tl24n8SU_d#(#z=m8Wi$#Z*z3RZt5 zR~=shlv>^XDwfrGes^%A4k8LLii<-0*_Sukd`$z!i^=$0l^ymH6>p@eXFk*(;Q&9Q zZS3&se1a_5)7#Ta$*1u=V0r4u9ZBh09lE{QThjhfFID zfJ$@Dn$L`5dH(&vge{C|v_}?}Hw|h4O$8&(6BE49IibNKSV9rHaGATyt}N~P!*6#~ zTqChJhh0P^LTkDN@%lT@GWOu>%vcx*$@cmA`FHmOU9N!NwzZ~FXCM?7YVldnvvD@) zk)J+D&kxSCtNW0+5{v!P5O-;H*!(U-Y2rhZbRcbo221^JO4AIl9Q*PW5DPJ0&`Y*g zsu!uG6vnBEij@$SN<7tXHxF?QCoe&TR2wHNlKc0j7(W3OSQAh(QBa0hQQWeDN+1qD z1doVbjxOvjHT`EAag?-=HPsOq>I3K77Ue^u-6|^D!-sgSE(ZolN1*T!9>5A3a^`v1 z)3eaAR+~8>Des(N6bBrD(uVf{$I82^59CLna#c9CrRCsd3c)pmBVlDWxyFVD(e1lW zJltn)R54sS9pN3yk{b#xU#eBj1v@Npv=;hELcyZ+S6Q&!?*93aN#6O%>pMJ)Sep-W z#)MT;jmc+7L^A_u!Ur(0R?kl$fX(as4YWe=6lgT%-6-S?CQKA2vFQ7C-Ao06p0n`A zI}AatqFa>MhQY?lG~sPA`h||g;bLwRrL)<7$Z3Xi8#l7xT9!{23?Y;-|DCHheEn3L zQN~6?80RbXjGU_~nxyo$rXp{^=_s6{`L^Mk0&PV$48YCfb+Cj0RV{mGMP(AWnI%}y z{<|4Su|(1V;3n97vHq%aKi+s!>J8FaJ8gG2Wus}5Ky%d4*Vp$OG>s|k1=!<(1lAjn zv#dgV@LUHc0MC}_U2&J6Z=Mz(O4DZ@G~Rg_G?IipAr&)yXCfcSU*jW>RMcg_+l z|Akfd8;LrVmbQ65G^mpF0bb+gqh=V-dCq~rMY5}XL7Fn?o=6A0wu0bKivn~A)GIse0cnkiDfQ}o_m4f9MFg5WsGv-u6YRGb6P2`i`0Q4Iem7M_8 zVGIOXAGv09!&l=KQ9?ZgbuZ~5w$0}K?2~mxW&)+bnH9;yrSxw%*HxwFYJm7JO0DfZ{1|*k( zoc>!U-7Le4e!=LO&?(O<(ua`3@e+oSO>yManxTrZzzH*rc5@5~6fs&>4vt zox+zW`yon)(Z}3o1b{%;>7!-3LQcCg98GKR z&~N-35)K#q0w-9*)8k2Byg(5Q1#2?(?B|If2rMZP#i*vuS27AZ<360ejL~nTqcN#S zZ+&QpzGsQB)F}8yo4pSo;X5%$!divRC65HGQP1=40IrlbTA-&21HgnaM>VdL=7Pta^2*@3XkOXyd2we1$lqyw(s8u zI`B&|0&~;HCuP#X{t7M#c`F-aKr?VL-akixJo{30daTrEBCA~>m@oe8L#)5_DG*`B z^3ce&?S}lgRks^x6C*ARuSp<<(%|mJz0HxCGU9 z2vZu>rcuZj)GZCNzx-RT_W1bu^rt8&LMLP)Colz`qceG6$G6m=Q>+q&<4^OhDGL!Y zk*aPE%fWsWGGewCGqJ8PORal03^a;n>|8g57_r2yIIXUlK<>c|<_A^1}pgsb@$U|0YD=*I;{UBB=M z!t9pQjuJ-(8{fL<50vH%W*t~Bgdm?E)5hS5c7VI zWxyKGv>>iXy?v9GjCi5bcbudaX4ro54(!0VgKI(6E*=?~wd?6H#T2L<5hpnZT0|jr ztzhN)A$WIf@3i)V@L)Oi8@3|9o7rz!BKEA^V}JZX83IncU?Mgf7WcY2zQ(IpgC2)V z^K8lwAhA^+fiT=;rY{~Ahl3kA+FVpSMFZ{F-T?C?h~U%CDllTz+-en@Qudgy?-5L> zlB~~`fMPJDvnbkUCC7F_$4U}LM%l>olDkwj>0=Q}?ozvSHD&DQQvNMCOY*!C#*asT zRw(MM-H*1EZ2&O9d1%2qcOCs9l z3IdjtBcv^yfUXQIsIBmavKQ@o-23@Bawx$`V5Q=#q+VEV(LYHs@&z-K>!nvusz0bZ z`axknnnWP#k*Ay`@f-!>_Q8=uhIDT_iypf6_K~&-Dxpsl^VbSH3!&5x5fLOabr$VW zwoF@Yjea$MV&|qj`(nz`wL#5Z>Qjh3HL9QAizv8u01U z-Pvh?)nQabi+=I72o#@&D0o9j;bv;?-BRm9A0MClxLeO$o2wn4P?}NN-E;GcRm`GE z*ULNf&+AP(4|HlY zKUbcaJ8+!)X6H(HIUf5P_THF2aw5Wk1WcO^S@rCx@A$jz&+fbhOKd!dM{-i?^Ur#i z@mTzRO;PMVsT|B6Un82llvshYU)$DFb zt=Z@-2kS^=oQhiL`U%;rbW@%7N73_XR!e)T;?x9H4}R%s!5$o(MNYEKv9vcOH!lMJ z+>@M5jcPChYICOF)Fi9aX=sL&M$No^o-yt6L@tF*H?z5L^2F)uclBX&R}#*uvR7Fa zd5K@5Dku?Blx8Ixv{8R))=HTT8!OG4!Lsq|j@6p?PYuKiByUJ}M@DHCNEpz3znIMh zs~D%S4ZMZyo`UoQOXn;l-_hDPEl7k0ZekO~C{0keaAN={48F)qaZ5 z1QAvpL{W^>)YFVB<*bKXFEmIe{H(1}1z{oRa1aoWGQJ3FE-`o8R-cLdc51pyt5az}abDieZ+Ica5@PraKEE7tj;duKq54Q$#Hgq zsNkuM`E<|>b4{Yq*uF9N zW6|N{8I7wU;vYD0@?^F0j+_(yj+Px!7fZXgEk`-y4g&017QjDcPYn`&yeiG}eggxv zqd!F~dc^bX0LgG_YZ=Y)EK^-(Y+@vg*zIbVfy!KT{M+pG=-BNeAEo5XSS{t**H$$B zg{QUqB*6Jc4FvrM9YXhVcgWDq7}TF5fW8ghyuV2B_kTX|_jf>_-+%DF1?aBj{!{n& z;{O>>IoFPmyOJ4ES+9&eKq1>*Z(}Iah3X4SiUBjwso)n?B~kM+q1%Z|tEPA)mdfvn zJ@|mxwoIo*En&zShP-^X;qGVDk}I@w=BExqaAj0m+nq{dyfkqhT;Z1jJ2g|5@%8nz zhmgP{8Lc%gFnu$|D^=4cQ?&;xxI zgi|SM63wS+7{kNEE@aTV2f#7D1HwwX7ik6MZ{515V3)}1fvip%ip~3ZVnKXJOx7)b?$o_2xPc@`+x$mFK(T2Zst`-afe89fEb%q4|9PR%*`*y;CTlFHXv4 zp;8yZ|81=vd?aweUa82O}R6DpQc0hBaULF>rg8S%ERNBND3d+OltGXj;*tb7% zTt&txfP?gvhWTCxN?3&s@EmI^a(V_GcV?jOyjgI1hs%bLq_*hgdB6Z3>v*n+i@D_T z*+5qn?kSyo3Y*5*4e$njss9xBI$Q}+3ksT8hK2E+@1X3I!ye26TS@3Gj}Q}w=4|t{ zd7=J9O|g7jzq9cCeBZ}MuI>Et()ind&P~J*WKk3a`<%M;^bk&oBnM=R!NsDuJOq!{ zJe}83ieDNLec#`NeurbhL!So?^`Q||ZhkObsH^HrA3qRsJD8iTKYKxeykoa7aN~pU zS*92us_TIQyTj`KzHUOPiP0U)>3&#+55xP)X2W6b0w!k8UYOk{F+D1e+QQKja_b zED;XE?BDDTIe%esx*Z_*c~UMdK=;(Pe-b1Df8IU_mKr!8B5TZjTtp39`Q`QHc!7Aa zWzL1y7o12`RdrT6+EY^t><}*CRl;rF+CRXD?LVe`!f(aHCCM#^4HO!5zId% z;EkNPrHY{q>U3<$bVOfO-X);JFERY~?VFDBp8RTQ0{$2{oJ*On3E!sKpe^6U!v@>w z$|pwGpD$%!EVF!pBc5q$)oQ=Qz`#ENf0= ze4B8*lNJ!}-vK#z7eto}o1o`KPOyPrn2wT2O~eF9I4udFdPjRedi|OBW07-?n>x)` zLyL=KNk~(07MF|o2^9rdS5hb3y;y7r&#-c;XEKBqxs-mcHFSED=FB{%&{S*#l+jo<<7 z5>0X_tT%R=^V48EL|mu^*x}1PWxh~3JS>pYiMIfp4+D-T!-=Scj@mWs!7au*3c3oq z{F$vVV8dPpw7U3?QZF#29w$G-4}is=de4^GDMOl8AK&K#3zhh)-`6LoO-8XC{LI>p zz2!6-u9{bt$t=$nSkia>j$+{xFqL9NtiA(XhK7NJ33v z^T+68K2WfjhV)mdMgHD{FewvoCc*k&k*Hn*HzXc^IxsTAs3~F*k;Qj+uN^XQjXxYz?GA-)&sjLDp_iNKI>PR>$( z{rc6F1%e8e5>RRI6{pq`l)!!DYK83h`N2;Q1^Zg=C7D@{{}Q)co_1l?L=Bwp&}1$% zA|B|hsBbx%I(`=w!j=Jty?+D96skZ?i-9cIt)#92d~OB6gKJ(=2wXatC?0iI=J2go zI_AY!Fea!jO=C30aivD$TL?2wu`3tny<(w)B~(lBO#CTPXLu(`TO_#qJd|i{>KdK1EMeOhoLZh=J!JFEIZNDicsES;1=TB z*ep+5dS6{#Ws#7q00P=d5YGzZ5VI9(@9|tC<*^eh?w~D`))#1>&d@jfZb?fZ7lDLH z-sN_>M}2vuBmlwhGZn}Of`?09@WFOCpUiv8oDK8jZrj|H-4}tB^#sU#hJozGu9YWQ z(j(`W|Iz@zb+N=b}5W`3P4cQWMn=f+e*M&uU^lIh{;I`!GZIFeLwICd$%k9;1rsDP*Gfjtx#$ybj2fSvxXNpk|;K5!!_Ij(8ZJ zx-X5c?jjeU2O^;E%qyN(7^HJVKyxoEfO7JXIZTMNf+DBR2A~wK1jtFvafA{g z3ej)Otl;5c*d_n!dYc*qNX=+L3Q3?lc^ z1C)xN{PQIV!tX_-{V2+!p>E}cCJ%UU5|YE>r9iHy67iGU*2a~Juo%4gf)*x3f`%J1r%>@2K0{!i0TS=nRM?xHX|{Nv9E&p^ zAn19Su>&4T%NU@~1XS=VuUUg%{;>_I%TU+$mPDj*WmDA$7pZClUP z+ZJxYeU*tN2Xp+a0jMh!m~(Zb;WLR5egl15?ti^CP7f_809Sr{dkfEc34B{J;K0IJmxbTFJ+ zegMs9BfzmJ$MJX1kgfx{Y8uhQ`D{~m0?)t4*IQv>VU7i!%QAj3C?(@ih(iVthe_Fj z%?2}~in;9?d({*REDJe}ObWp{P@}Yp45p?yh+{ex&SDghi#+jHh=-03OI_U-w>Sg@ z1X>Xl%8$H%nRg?+eG=Iy=Xj{_LF?rfy#p2L3Yb&5W>~aOpTdE*0sRrJtfM}VyrYo> zozDEC1XKJ5hNI793W#0n*b6nkzc(+ZCtx3?26)a&C{%zw@c792*9o2}L%ldqU7sHB zob;`|UHly1fo_*vDbJ`Uy4dAliPP(-pR0sI^s67jn=^}H2tw}&3`qWC0|oVj!Z`f| zyOr+DMqR+Icz+k=<;}aMv$c%951^|=>QBIoUIlhv^iRg(pw3{YX^f@?&KWjtnKB>) zC%TG4%OH3H<`hl9i$s|K*PzXi=A$ua?hv4B9hkIna(AKYI6gDS&8MhIgdYkYK%>y8 z6);CPI_#wErk`GrES`a}@|?AIiS(AXAe=UlR44H~kE!`gZ)B-FM=pnZrOCq)^YAuEq!3+RRfo=QtxJQvZTmzj?7>b_cPC->pM;O z;n`W)H$H_$+Jzl`u)<)6y@xCVN=^_d1Wz6sECfTsD)SuZ(dDKfh%Ppqf>XJvlM3MEa=$#L<-M;m#Q|Rj8v>|J4Lk+$AZQd>4Xh+Y z9~v!QB=+j#^iZBqouwDRD7Gj#K=%wrk~P3tr{7oUC~&A&1A@ew6$S`R-gkn$x%b5^ z|7tH8R-AdsX?sEzLpW+OATZFlYa>e1!X5Pqj&D(&f+e)I6_|s8H|~2jh7od&{FC6Bz^DGu-a_h~ zoIJx=KuP|}r|E6#@cjcQVfYHlLDelSi+z-Vc{K#33-3TlnSadtHm{pvUTy%cc-i1m z04R@(x4wbxvH}z6iSkmmcfCetTtr;fA79*WE=bAwoIdU@w&{M<96Cp6A(&}$w#nsM z_Ptk>>oxOAg|Xm^TE4O>g}K6jN}k zOF2w$OH9b-k&iI7%x3Du24M2&_8Fqx>y9(NV+~uPo?{lMzjp_38k|!4nA~K z!-M~7btSc2Vc`j1=Z%nq%>N82jHR{)dD}B&WaRE}MGDdFEF-^saT+@4K(J$s?d{_; zA#K37_a3+*<}LZ3M@F{fZL7N)u^XH3L9ly3l(X#h2_5SVvG~0`w5Cd`9JtX3M(go{ zuf^cqNx+GAzCvc)q1*rxxd#i*>QbuLc1c}r-}Xx|PQU%>$+XcN>9d3}Yoo!m8-_nK zyJmNPf6=#-{>L(2#}5{`Lmk;ya4!X3lKMiQU^fj>_QpnYqefVEKh@`ANx^$M3sOReEmA#N}fl5gCyeRy)hSa^*N%n&NCYkG#0HvcYLsv#W5 z;R~7d89|jjauSjh>X7E<0{%{5l*SrLGl>>G%N4i7zJ8>SwN$80_{k&Cdk7P%rU)v&iTz0@SSrhR~sMA083JDRhwdjaYC`l&w8=XJg zyrbYUq|>ov{36ASsEdD+CpVy%LW#7cc_<@9N~d#UH%;XlUp4N?UxdN%iFP-5qa7`$X-1NM#Gjg9+qw6>M^bvamPtFk!>6_A` z4+FQyJP2Pu642a!=x-89!f2NF%bQQaoc>5LnTc!E`IuN@G3^*H7AQ)YD)!K`vQ!pVtl>36Bk zE)JfG^&jlK<@1^OZ*b4fPQ6qWZiZo;enq$m4j)D8Nc_AYD*Z)8)P*t38|YKgdn@iB zXjre|x!g52K9in@U^<8NmA7kZ&gQr>Xu!3<4Wpyrx^j5)@yJDr z6y3?XY)i93Q|Te$xI=GIlSJ~w0@=xPph|RiFbt!Z&z#RZ;MF?hoH?O7Q^D+_zIxF* z;hR1=eVtgi=K4^v;4#lhM5dE+jvV_W&suP2P53Bppf__2`_`EbG<c0#OD zS`-5Y2^cG-1FZfWhu$LF1Y0KgW8;{|hmc|2?JWBr2I1Qp;g^2kSgJS)*n*?iX+B3A z!|$5Fzc)QvS*mly@0cQ<4wIm}gdxs}zI>@8X(s&X$=;@k1n~Bz*@Z#~O<9TF{kBOp z)F|4~)!nxrGrm)*f1PR=v5h`u+W&<(YcAcVc)@YF?2;do~Z@ixTl`%HuuUbyE$kK}1&21-4eUyj{iP+-sVt?i?x&Z9PVGc?tM6iJcE99U~eNL+#Q^ zY=3w=^qw2WHO?q7GlLLJcm5$^jAH~(k{@sApW7!qc}{u;)xEvNKk2AALAW=WpcA^y zUSJ1DW4Oq#32l9gRdshCxo)p1nz=Q`2cda3afee(W%uO~{$x2v@0zcs_FgmQr0snb zp;Cz^BTY;-bsGI_@HG)A!!%`Xyj`aKs3=1{ZT&x99Wh9@yzmmuY3|+jkdO>-W}oDx zdVW6wye~|BXrSSKY9U=nKfLDfLyw8Z2VzgE>4Dju434LH6=YD}mcRdXSDN`Cg9Q%G zPT2$}02XSM%`U$R5{#BTL6%xOGVq_^RWggB7m1ybd5CORZfCpB$wwRtZx?sYRtvxy z_a>Ui$~>cCo-FX&xe1XQdLbh8&74MooSK$fvDd;_D|Gj(O+z-OTA+|g!qDu%!)x@v zJiQ-YaKf?0EPpI2&F(HOv`Hph=>MPeuM+<(FDr>w-2e5df_CB9wl;6=qf3+vg%PX= zX7@{0cliz2d_o7@;SWv{3Mo~BKVOeMJss1h2u*jcXecQShz_K9B@N)O*K_0w+;pW} z2N=R&vrwE4c8k!PPn!;e7wbV4#TPpukC#2i{1LVm5vk}Azr?+)`r8^Baa=mldseon z9Bz=7PCK_+b4UVBxaDE8D5dq^Dy+K?d^u|F*BfLxR5RIK2A&lO-G1xMpbC{dUP&R}ZDbKQxM6^?H9 zl41>QCJrC|IKnN7S#LqSP7{)c*v}6IJ0`o=4BY4}WVi}uv(Y3hX1KPl4p^bJwRI9# z&s}ZJf%0jr2hY#a7Rn}_a#ZiTB+!a01r=Nj=&DLh;)SRy$=l>&a#YGhiYuLUP3-jCSsr zh`;$##dP`5Tt<9USd%8GLoW52nf`KnFoPj)^vEhiI))ma9>rAQK9J< z6E*qJ4=ZbDZ_9^!=ZBZsI~v-fSwtcR>(PjR!tT^inCl!r5Jk_pyoY)FyiwYA|Fs$RF;@4*ppEogf8&{}a+c@%wR+_<8XB5p4KG&>+g5qMuZPX? zlSh@WJtz71!(&2)4cM_djx>5RhA|_kqHJ`t<|kEbXj57ygayKx2^iXnT)fbp9|-8O z>Zs()32lc*60%E;w4F?_tPbw&B8J1P)e5KTYKr*RlH!x7TaPyc*~_ooGIK58Cb_4$ z^0^Y7Y46?HcGZ89$9n-kt4^*4F^}U==v#)+OC*cyjoYD&i}~;-^|~}0 zjt!W%+Tquy9Hs*DcG{{o4-~0l3P`luKRq&>0%<~l1(aD#sdsL4A^VppkG2~o!f@6& z^21$pou!w&t~hCEKFg-I3}NlfFv_enfjEPDj9aHnV#^mS*S+DQ$Zod%^wzAHb%E-^ z!|*1I0}^6Zr>FL3tniof!js&3-N>S`7c<(DF*WQTTaNM07LBoRA<=08SGhc%z;yiA zp+bj54yh;qJy_={cW1f?u{^8GJqRCHrEU zbAea{XRz;TPFADVGEp5 zpC+|gR7=Jp=C$5c5_d=(j}yE#Qj`2-i_>-Jw;~v2Z(DDtv0TZBc{rviwpJ1%F{0MZ z?D1UDfjEY3=#$>-qcU$>y5K`LQL4&`>VdCo$9M17AAq`(e;;6WgF=(S$C`E9`A@t) z^e$viA}F#8>wUN$@xc&QITXRc@FM!gh+n-$f ze^+t;^NJ8uZhyAcFo{`jbJghGk0NC0pO`i&Mi;|%?vfJbdd|T#XZ7z=UZ8-31aY}E z4keDvHDvPZk%EZ8zN1*H0g)Ias6)_V!hsYj?{O{xW=XEENVwvroL7C-23ZMOIdJMvw+%hrXUzN3vp7xTX+T&#!)j?lh=^S;VUtd&%XP(+S% zHlaG5bSOUR|9#M{UUsjipJcYKdFh?L6%1j^Rvj^Uzpla)Z!GZ${XDs;@o7oPMM`f# zSYL8o8||#>5A+gKciO-46Mz3i_Txt+SVc!c>3O$*6bSr434aNxrr9b`8voZp*Ktx% z)?3vQQ&b8VasPDypt`vqA{L2YUR&4*6!A7AnA{n8n}U>ohCcMSQdq^aA6TYdbP+dSJa!D?sy)}(y{Bun z1-CLU7mioq*{xe$s@4oIhx}jZM7f~7{Q)1eqGN$epmY1JCJ*1^MQrIJJIO_R{7qF~3c#~wX zc;<<1BK0o8s&6NTJ4=d@LHFa%r7HEA9}to+Od{tNO~rr!cU>K%7@?`dlG=1`+m|bb zAYS?J?k0l;M*QuZx4{O6UxpF-?>7)m!vCFI|9Rb8PZ64jlypWMuR{1{Stt5~kgQp< zHKI(>517K7fz~%at{y$m93Zs_)HSE;^9;C3A~bkj=dZY^MZh}nKoR7m@X8C`6sZ<& z0)UpGdZp1Lw(NiR(Y@dC&l__XfE!OP33t%j*$67&`tL)EG59d>VWl+KAHkqvaR6e% z`D2i%hvMV@Ilwdj=S2f)@C(x$g#T&jY%jD*zKO%1||6Iv@80LLg5@fO01$ zbd=aqYzYbJ{N}|v0Tk+$ZQcz=udq(`$pprSG{&cHqH<5**{FvyN zdbgxW6;ilT6(XK8a>T;Uc)>F*W)IOY%-hXX=m!g1bmJ$>MlHuaDNGcI+N;*s5l#rK zupQhGqljs9*G6NIPbK2gD+%)_v~pLeuGmjsR~Wc|g>#+$m5K(s1O|f4iLp!m{V!eg7auUs? z`+P5I?x-QF9=^ezYMGlZ%YewNI>L+mxer8-n06<2*k9ygB>vhx8$}OqSKXdeL?-cbV zgEA@<$g?_Yb@^D1o~pOe?$^?82N3VdWdRh4}8c8`o{ zfE?q~?#~6i0bQH)?;L8INq3P=z=ax3zv8nddY^rwdDAeM?3)9KrYEjlNB+Hy>wWxp z@BjUC^$~?s(ymym+ezV;ASh?9-Khd8rbvwb$=WyYQ9-vYyy{{zq^V!88+biN=NCJzxAf?d0^7Vu*w;^Pl2_gKWMXjX;28n zXMNQ1GF-%+wMTlqU*_eWmn0HOC@y40!zjvhoHZoZb9)okZ(iJ0 zND+yo-6fVes$j1Znw28O;m(>#j4jUX^ZiVN?nM4sRR)n10R(k_#HXdl$Q;xfdjthy z8e=5a8IXI8Db(BcaWa?>C`Hxkop4TMOj8@U?_SEP1$_Zg5rD{ohu{^@^j?Oin64_u z7>89;rVCZSo^`&Zk3>?mO~#yBvG}<`6Xmp4v~M|5OWL$_Ktbo!t~1d&Y2+F>UOXFI zbAoBSzSpdm?tSE4>=@)dl2m6s7+<;wD9K)8aZp>sUuA1kOY=+=PJX^Xe7=yJXCts| zb{k(EDRg9LahBpBjU_&)=C(KVxp>m~lWV)u2mH?WdsG05jfTfij&7vY?4szHUSH4M zCkVt!4_yHFir>ntJn5?x+_fw9$AI@7^>v<>A+_cN1A})fznr>I&l`%ZJe$Y@xKKpL zB0@Z|`PRw~bH0Go?A>>Z?p@^+TfGGjigt0_*relsFxaojsIuh%R|6B0Hn(6=le)(^ zbE}%BW#Gxgg`l{4A)q=gErQcunYb6jO}}blf3*uGSs6(ke;xVhYaB1)_L`^Vtz32W zj})6E0hO0XVxB)^2u3~`NBkeI-U2GBsC^d)X%HAm=|LqVq@)`LLDEsWyF@|J7n zrIFn*!;lYqVeLqyr7{1D+FZ+b20RRs02NDIdUCe(SUWzaMf*#aoX$WKoz7*@vo3dP z0n!DVpa})2?Ld;!Bcqn%HYRNAqN^rb258f!4XCX%%E4C8n!2N82%77VuV5zT_aC@Hox0FYWM9lNoy$Mx2;DB*G8bhdprJRRn7;b zwC8rdM=b>8cF!)m3d+s*;8f3+(e-NIgoNv~f8(6J@VA0pAG<-^R|lNt=G#kVMGrby zjuzWga0br&*2hNdmpUdY6b?yNeT#kkPfrQD2N6h**rO|*l3@WGOgOFroXYrx`nDzP z!Q=-C@l@yL4q1osEsO9U^w&MNOYSeoQtA{67{Q80oRE>HI_#(6r_n$XPhwX8h)MkzT+actpmpF~vpUhwNBtgqwH&U<6!OgObCAL zFkUuwg`n#^jr%Ec{Bp$|=M74Q+JA&3hxfI$MAp~~GvkB6R};)}c_t)bU@83pbVZJO zwe>~?Z=B(Vu={bTUgSoo{(U592kZvrgC#T`T)e!m3pc+fM}kI;@n6y@E;V_|KSIW7 z+pA*PrBYn5A7)M~A{zio1l1kV61;!!IjJ&rpUdBQfvS6OOpR2V7T0x!u)dNRY7&^g zziYY*^$^;JhITzu0KkHVoZKd_e6aq3_b~to*n8d&h+5zBS!=H*3th^UB6)w*bQ>=T z-Q({(e==-d_WqrreA?MLJuY(9pJXODdhOYVcgyN&oxjTHP7;7tU_&lmobac_A5W~*`aT!<=^%b zkAOp$L_gm+5yJbeFUW%BK9%6TA2dRAgoumt@6H4W z-1k%)TCD&8Et#ZJ=%gkOxKtN_n#c@KIUTB8G6dNqZ_Qt) zghWKcz&M%mCx5}k|Dq|EWT=|f@fp0p^NPH{^Vx_JyJt!aNtg)aNyfHI zqq!5(#Qt)HTjjJWkAwI#=dlZ8Yz%nqMu$70RVb^jQCOGA%fN%qx7ilHk6UpMUAKUk$Qp2BL#_XT{j?n}fZ7zQpS8oF!!Zc;pe>VLgopXVC$Y5jgRUrtL z%<%97!9vW!ux((U2_UXu!zYSW zNXc1+gtYnF6;s(h@fC{^!k#jQ`MpY=p5y^Yp&gECD!R!L{IL15B2u6-w=N4CfvLvR zewAnA!11B#!O3;EtmeYgLTU$=heg<>v1rJYny9V^;}@?|Hdi`CK^6c9QWU0A-63q- zlAdyw2{1Q^lUsvXvxe&Au%A$~WlQPkx`z5wX11uEU)2+(xP0q>rK2lH2jDNhtG=y1 zdzmBb{@rg;gW*KSa+u^gqhsrMMC4Tu+3)cSqd=E*P2qq9VFUbw^zSUr6dHDM*XYQ_ z`!ELyKe&4hi|NI^k{eui3Iy_tViwQiYjLkdvLqvX_l?1;d$2hi*jNP`z7h*V4YIj( zK(nIjndH^?M5toG$m@p^K6ZIOLazv2K>r*Xd>pM3oFneuhsf;&j2_!hk2kI!@EQ5D zOe?rdpY7p;sroz}d7iA*aawQ{j`27c6^An%btQ1U#y8MpP*SBg-Ny?+OAeBcmVEy_ z9B87*zo_fh+NYs_aq2V=Fq^+EdDd=!;qc-fyy4lyaJ96^d^eB0IDn74q+F(T)bwTU z?6s{l*=^h@cK9SANRX|c(3w-CHp7$jc7oSfK{$8rs)>5Vs0nJ^vKv6STnEm7>Jua` z`l|AOD?;1>;7<7-7d-&^3sY&EqQ|Qh5K-y@IiYX>U&bt)=%TAqg;!X6N z<{2!2iP6p1j@n*k${t1$@HaVb{Sj-$hp8m&S~Xtp{`@Ki0*0FTo1>k__g+6e@DD;2 z?o9PFO^>WN*jYvIVRU2I$20zWX#xdsp~wrdB1mR`I(mSYn~Pge6NIkICBJFaptjX&rYMmX z?C%{`@>jEr#@hT;mk-1)5Umx89S7Ofa68DMp6i9NE0s9+W@^wv_2Ohxipwb4iq@hh zIx>EKFYfA1(;ANAfm6sW7;2M*ILtIdWQ0z z-h;bL$eg;3y#B}UImjiI;tI%)sfsS7LZgu#oQ4kd0a;^GWQ2LeBZ<<3xbK?AE?CHH z`w`G+{W>?DQ=hbod(RcQ)W25D+FL{1fn)RFU5$F?QQBLF6YA!-+sA`*z*9I{AJUQv z%SyFu9TSll9;~kin$+W(c{QBGMupXCb8ZV7ePwvAnD)o7>SOwA0;HGEd+$z@E$DH4 zVGFlu4j!@(ylYE(TNqg{G)teu7a z1E88#VYZ(m%}b?#=+7P z^6{Y{Qj}(1QoABo?nD-kLktT>xKCr@dq9aHg7CWJaRm#s|EHKXqZR<+A+*(nkLg8qwk>d9Vn^6`|YnS$h;s$<_cPWNHHSw!?J5!aR04s@*<2I^5uq1@_o^ts| z`O88_x_l{+!i?x&dvE8Lf7$jS=mquj@f=TGQFeb4Xj#RJtmibpo2cyP-c-9XJP(vF zg%_UBZ){D*Ne*5P{tDPO6K{>NL|kWIQk^a2xg2}5Bi>C<6@1}5zu+6D9(2V5%7chO z&WnV4w$(iH3p|Jw_o@gxhk*a zc7#Rj#5n-mic7sFhie@!XpE{5?-Hz)ewn8(B0Se}6McY5mx?s37zwFGQ|*)E9L8 zQ8HdI1EMT|*WSaXQTBB>U-BSv#FZQzPPNfoo|dH=tK4)sLQR}R-N zPRe~b5J>cY*9#lh7!eU+(flcn`m@%Jx5I>E_Y`R78WEgvRGia8+biX7vaF(7Y~xV= zdS3^gox%FObT{~Snlpl2?#Ys4Cen6ysQEa9l1i8k5wM-bvt z^tM2sujnnNlBW^(Q>YiFzM(bJPcj^*f^L7`YrpqTC3jRS#R%XRMp3C9`Im~7sO(xD z4-Oq$s+Lo|%GOjd`;sg1fNy$BwBT~`LBTm1QvV*(ZxAgi4_&Y2xa7yjgrOOs0TyS% zC-P`)Z}ly5uLl>s#rO5uE|Wxd?CL!(GWw71SVW6K}8RPZ}Ivx zspx%t9j5EPakj?>;rHp+Pul!RYE>ufOjFVIm+Aow&MMLG^eaKq0(h7voC_XqwgCoM zNXutB$q%1e{a{76Zer-P^@23|?Jy=;ZO6cG&3l@{q$&EDbhc!T#9VSk$Rj9;f`m2mptUc)=ahAQE)mt+xa~b zP5$&Vn+EH8Ez}mI$Ls;n?X>WF^2@?=`3lRE=v@%{XTR0yPjuf&XBj2~h=I}aA*>gY zUz!ddchAP6ZcMoiS`JDv24M6y3^p`FqtJ|>Q2@{<82L5^ZsuG6^w(vLuaW(P`2Oml zhAAhL8aCW$Wzpii>gf~OFBXm0cs?J2Z+A@=TDdk@_g6T&p(Og3C)>E1Cwmr@rl7f~Ykdmns z6V#Hoziy&@?Rnp0ouC^sVcb7oIHVrQ9#}`InX(FN1xT(q3pRDP+_M;o_*u7?x~J`^ zG_-j;eLv&xdq(L-g=5zWN0J%U%bVWKeEvF@@?BDM)!Q)^o=Jc#L=(d5Gm#dMm{zSn zWLtauMliq5CAEy22!u$OAm`9i*PfKS*ZFpVtf){$>O$VBr*Tpji{rV2zy#98a>N0Dd zbc?L!X-Uj9%Cvg}(w+&v1R1-yJTVrX9g{iNZz(_~)`REwNQ(LkSyLVk9X0h1N`aWc zHWIL#^)C-5Ypv->Tx#GsMW|&k-lT2|PQ`TROa2jpFjZaGVHE$G^}698WNP7gw}I#V{oxSbzm!~G zcg4Ji)(g&$GH?Vl&Iw<0ramVQ8cmq1P`w~3J-+qCUqSN_g{&(B&*&v%u{9?tzEW}H zt_Nx}Afqs`hOH2S9at?SVrT^G_e)MoJY^r-OCX0LU6f1o3mwaogyLz_i{69fM7E|K^V4e8dag><< zoIRmwZ%%9m8#Z?HH+H^OLf2KGtTlu+V7X_1xl|bo;nRCRarY`r@u@YMWIrPohenR-;d-H@|^YcH9k;I33MA3rPzvISu*V*e6sQD?KKKMcrx~Q!f zze@u$4{gISta^(PoWSy0t`Wnmobl(Z5Cl% z(Z`j_IK=khiyu=8vc{@G9NHjAK*Jp}RW_r2{}ljAq(qP;hRn!*VtJRvBeTL2Zi)Qj z4|Kws2w`^ArHs=njZe+LGjCps29BM=|5Z0sMQwoyVC2a~ybU2Eo&?HpAvSz9U~k$J zm0==tt7dDwVgMXFjmw~~+_&E`_bK-~$#xXy%q%Ma>Qko71UwI~$$GE;WXnS7QrXVv z^#Uc2f3W;8z)R{|D?Cy*%&0Ly6Dj2^ybQ4%M29i$fk-a7EgjeKo5Mwkq>LTWYf10Z zx)Xs;ysfi)C!|{k?w6|dY*ThEv@rMQD`c2{_{_Cj?zWw7<7eq}Vn19s>nK`v2AoIe{ve~vpFRm)q=G%YUi~6{YPnAN z`Hw)#Edzt-98(g%TtoDuC$O|#7tWdXrjMuoEZ?7*t34@7;b465C@H0`fi0B-K5u>gI-oX2bRT)w?syPn?IV^*bLM z`}Q^RuGg-kpRdN6EOBT0?wtstc04+AJMPb*F?;=lzl2!$=BttjG=Z~mta=g{VK%qH z0QtH)j#1;U9JzQNec3}IB$}M_4Tkc<)NfJ0ww1<8Ukt11#~jpvJE)bV!5;_GR32)_ z&;p5jOV2pMd->qybLmNvOqvm*vTVXlq3vEigpA%;1QU@18iJbJ_T2_um=%QtZ}iNl z--Z0^KKJIob1F^li5PX}zw734WH0#eHU%Zp`eY#CXC=wZJEpAFrf|pDZRdPB@AC>Y z_(Nio-3E_dz4$yEOX5=Y>QB0%^WWAv=RL>vt1njgcP3}P$&O*e+1G=n2#vee1}1c# zQ?zL9&(MajhM>GK;oQnwkHgaTnT}=$f7)h_Gd&mjWdXr*BJ^^GRy=I$5)hYUnYe@i zHufL;8oz(yl48No+rA8p&EjY?{QQNB?7|U>n;}0sFc525xRe9~P zeOzv4_nfj`jZF4G?1m_LNceZclGCnO@4`XDHyNY7!VWjoDNr^MDE?dS*~72qsm?OQ zn@-)uoQ)pZ4)xPZgc!_+{RY0d8&3;w2|tF@ifJb2B@qopE)uEk$tkG(f|1V0a0^>k zlOtsM=#F7z9WwI%*E<`%`3n^)|8%@9gqCnf=wMiH0FeN*69et@ehmY!J}&@O+sdxC z>^C3QW{{iZPJ`(OLrds=#jI?G1^WVNy3g=>RuvWFOIBs(_ZG<&IwKR;x*x9x3BMGM z&!pjMvTEMd*jZY5zAv5Pk(bt~c?tcD2a5|>gR89}6mPeX7rF5c-*|?M-Uk`Yn)}(O z%n^snV8XFDH=-_R`km@ooord1tl)-Ef!3{J4EyKyDUD5|UFY&&5AL7Z`!nwkmt(AO zk1~G`H0}Bt9lwhfK!~u5x^>7=5e;fYbUO3Vq{xlj%(--CnY&2ILgXOi{1A&_I`y-v z1&(939uY)YfF60HaWK37jJ%h`b&gb3%}y2P2UXf7_A05#Yn5%=-gy-S-{%Y>2b!YD zqF8wd78kP@pN$a>gLZO(!;g=feM#a^&`7W$;nlhlyqRNn{BS;ijSs}?!j@_ zbxNiybs6#{AI$q3?@jPMf3-zn%y0YZO(@(18}7w(od;zDXIgFQBA-w*hO^Elb9CEK zD}@By-CwUvdXQdX5!(7QmS2ALu~|(%@aDK3`5T_|Ox%81Fxqv$ICe{-$KY9|b5|2~ zX}w{_RO1_?{o87RO+a&hV~JVyo8Y1E16ccm(^Yo&TwFo~f_k{@1Nr=^ z-s^qXsQiTuD5Z1VqT$xcwEer}Vrk42{j#ow&sX(pABupX~#|0RM3|AQSS zXdYSB-r$O1LiGy^##p^jIHxpzEc$b9qG1Xne&+`ULCy$ns;Ff@ZROqzV;RW0(^j_p zhIrDAbG!#| z0xKFcQx1HTyXKdbw`{4iAK#Wd~4h+pzB+ryj6a^mqI*Xjix33sp)lK=UXU; z!!R4r0Zt86udoU_AQblde+w9$tWHzD=yry?#KQ<+KlQ(b-NtFK**d()sDEjeJCge# z^pmZ7B0JvndqcuT_3x|mw@Ixm)6!&Jc=2@0OC!um2hUi1O{U!TW~tNHhqCn?=20G+ zAt-AV-Uv+OsiW_s_X$geN?X5JC}IT{kBQ1Ub5_!C1o!3o22KG6bTOCRTt!|MpB2%J z3CU6r3RlhY0n9>`fvp4Lip%qA%={tua?9h-g7bIK-~6+~hBZ6t3jj8n;jGS3HEwa&sAW z>#(}vw7Mbq98)*sW%m%FhT{KZ$!^+Dlt`-f)irkM`K=!;NX88LMF?v@|L)yrkiMVQ z{K2{FW?1Cxp9*$*wO2d#2k!4t9l>WB+3`=~jVdDVA-{?@B~^&#L;Ks&gH~tnG5@@p za1C!6@+GT=;=g}bz;kk?faaW$;c?%&=@)0?;kCoWm;3v_tS^I-hM5{vWsGw9Mb`6yD z!Vj|_Igv@H$nAg5F1J~bzu~-!yR<@m#ywN?tN4H+`>;!&omc5?lnf8TI-Pf_J4nsM?zTFym*@_H@=bXe(xb|#Iq6?W7zF(p81w4ua3$y zvKY|c6Z{_k%HiXTeSnGMq8BXOlz~>ie1Gl}#}b;UqR|_f%iDtXk*_M&9j|uBXPm5; z)wI?I%DrWNFFOehn?<-kM&KuN<_f;>w`jlRkLC93UG|Z%r*^IJtJTdNE)crKs{bjB8f`i*k zmBC>`DxCckeJ5$K!`WpznairCxo?~#mX&;*@(Mk-eF?*loK;9+KOpTutxwmm1+B9}y!Bmv4hNT1q41!*B*a+9b3XN*YR&b&U6Vqc#OGp!_~JA8Sv) zi-#C4)ZRn>zzREU@`DOri}aZL*PK8Mn|y{uvNp}8^c=M)#p`({dWG4g)q_LA8$5o^ z-pl&H`6Wp5=m{b1_1lX`tV3)h+VzP)wzdp?v$1gg+fD(W!>%9amP7`W<%(+N4Bi=ti%SI;3bb9kwqdCYWe6yoW=;?InpJkSQqD|F%eTvyra z=`H_wo2r*wjrM0)d(y?I3zEz5^q6)-RvDfpLfN7V&UtMQsk6A$7h1e)DTYVHf;){* zaqnd7C0l^SgHAE&$2cX{%T!kJovXmvZSop=KpuU4F%-ZL`j)qY4cI5r#T(t;r73an z^@83Slu=cQouWa}lDu1A)a^`X>nER!=?y`Qdxqfiz3FQQ7#EASY^_R%>#Q>8wlVH!Pt(Om2k}f7B1b)1 zxzXv?Gu#TUqzGm{(+e*v#~gX5lgM1=ytg*M!{8~EkcvO(Bs~oy*|K;6A)if4b+I={ zhO!ArlNm1Zz*kR;vY>1?(qzn{hZ^D?FFea|4nj5@L_b|DY2(28D8!4rg{e9{ro}r( zUnskE_^Dujmh76x5Q9eaFn5oOM1Pd?`M!RXFiJwuO&pHnWC9P`nO$V9CxBB zFXFOWiKIyfp!nPv%cS2VM#beIjyOiBY|`9j)$MHpJgZ;ucUw@!{PzIX(k@dWg!Liw za;SH-7aoi;QaKm<%*N1%uJy!O{E)W4kxmJUZ;1QnN14lngYAdQ1ynWdzrHCSOEv1S zGuc_elEv4xm`LN1_N$1Ele*gv)uGc%4QGgro-&#X{4U-0(mWkaLZjm6_EX)Qit1@{_@EKM6-0Si6lU)XB>lKo2 zr3=bvuF~z)u40;Pj&}>4FJw?{5qNZKfj(mqx`v6M$&Fra?dTd4eu#&huE;4{e^Khw z-FO#OPL9d8md0$v9z8+Guy&Lyk-dqbcuHp6N42g2!MfbarJSN7CVs^dyIrK}i|V|X zONz@}wiF9uMMHP}_hN1Chz${vD_JW`K!;kMtJRUC>5hd zTM%nBQcO@<YS^HO)wg zv?7$TrAiRo$I1|kr-YHR|V#gOuU%3o{)qXg=t!rydks)*85rKPyYQeEo(f2xGI_%lQb9Zxed_ zaa`+EPw10gZE7+7VK@Bzq$F5S)oqEUEOC{fF~sjRvC_jL0@A2X){a+x-_Vzg&le^* zoBYc;WIIA>R4mBN=U5oqY7Ch~9={Ybc4R?BOCX}ddg|oW3pu8x)t_+W$sC6dtfIr@ zx^R(o%Zov*a0$8DUpYP>So?bl*YQo%9sr9i=Q2)GV9Grg{Kqb<{iK!xIjAB54MF__ z^=gu#uAoW88=shFjTk44{~zg*X_xo=*@hc10MfP=h;-8!zafObA&v*Z%7R! z0|$8;=%*?=z4ys&MhrKR&-`+sSDX+Dk--->hago2^%opv?Y_|3sd^+2wRBpp_;q^l zG)b$6%;(A(?{jYt3~MU`+50~u)db=$GaD#$@&c~j0E4XCXG{iBVloRQR-?BNa0mYe z=4VtWTIJvT%!(wC*+?ip>4A{d;mOBGvfnGboU@JS&{)dN;6_DuX!mUK4rGBDvd z@qr&2J%4OPg5m^8Q{<^&+KxMYR{LySdy!)L@d5?&A3cf03h!Wa(?PwPwlu+^r)ny! z;L3Jg-=Z(hNMLfyc`HpuqNgvi#XK`hq}7wt_Bcn}j|RDDHx#ZDZVM+O|w5jfAo_0r$WhBfjJ*jGURRGQLG4kI)PE zXSq>x0d+k!tQ|N715L&)hB(3s72DeW8Xe9fzPcIJV+mqq-db3#L?AlXagzXBVbqik z5FOzl`1LH0uzPqBAwSP!qs-PO(8UA{FQa}uAXkmCfRI0pafguOJx28V5U?O@?f&)u z5Q%XK`d}Z?nWTT=p6&2^S!LsUIpx|BfAMu~XT7E3(MVjR>3}Ie*vYYXnJuokdfMaKq|zyD{b-w$Q2U0Sb|KozpG@kpmU)Ke!vyZQ$BFF0X$ zvVZp?{@)RfjdbHGW?2;b5Pr>B(I%&pHRMtDG`;(TN@GS{GQUP6;z%*teVHwY)64u? zYnD?WaoZ619RI%&jch9gh7ME*_YU~hmyR%b*xX5>M#P&32 z59wa`NkEO&D@ZsuuOVFx>f%nT{fE% z#sux?i-T{htDKwKbbaFOhyRwSSq4q7s}Wq5-JnImMCm{5itUtTcbE%zrR}}C*8A2 zO-V`Vt{Qs!^l8x_jaD2MxS-HtkmRYcZ(u`9#Q2Cq8wcXc6h0T#k2-_wiScolsDRA% zMgh<5$x`n`%A(dbH{xVXo-`xXLphX?z@FIjro&r8IUC*>=-syazlSM4O#Hb;u}6Gw z>t4VC=6^PcdPoRr&&hu)>T-5cp7JXpg8FQfOzrPE%|U)YHv-9e{KVGgE9^8n3jOOI z04D6);ZEdOD?050@LV2DBj&(^;mrgu9f%fBscNo>MtHjLw|500s9t)bqAFkaz{Q0{ zCbduZ;J=tvPGG+Ib2Yh8HcUu(lo}W98D7HRTP!5KLBk^HoBSwJilSmy4n{ zo|r(M$n^aVLtCy`26-WZf`K-VLGV4Kwf?N*qrIp ze@^XsR@ea{!Bb--)?tV|HN>cv;%cOPD@?}a_}_c|=M(i&9T|FeXD(1~@t-L~I?IUv zc3%+rDf;0}iE65xd<7)5cvy`8Khx{x+HDQ@AI^IZoUQ9ea{S{XKKR` zfxAn<$;g|m*rZbm`2Kq-w&RgEA>ghghsz+|fs?Jb`-_v339Mf5Ek{ zj-Q)C@n5?4G$ruzh{?fuybeIo18elray#gb20tLXj+L1Zyl%iI+lUAy$%bTWO4NnlhWq5r)E@2waUxM+XsyX547?pf0R zTrKP=?3->Zf99)O|R#8QTGbtx02f*c)EX>mU=ZjQR zEv}ooU7CnmnTfE2SOE35%xV}Wh#>mi^Ct9e#=Ojj9DA%4o?%Tup~9Cx8U5!5-Y|g8 z=lH~ReaOn`)qiK8o>Ew5i)HeXXj)U7KB2KiJyAdsS`pMcudl8h3Ecw;fe@q&<()#? z&pHL@2O!Kd@WPnYmUF!>EG)c7bdvn~&70%wnU^nLwk%f#0PL?Uy*(`eO1Cpx`MBS) zSyw)#u~|rvz3fOg=H!{N6C+V& ztHr<^Z{rwqbFfx3OqOjX&~1{7!!casr+UdWZ}ffj)5igmTq}p+!{?FiFJ}K9{vx`Z znq8&kA}>s+uA8kcChDw3+}$XTOU%ft!+h7T04eCaBsXi1lMUq%^kv&mI7a9Hc;Qnq zL}Sq4UDrF`_Ej!B-MeCFJ!}Zqg`xa=tJLhJc!HmV(21{Ag33Lx+$qgfzY#$wdU2mM z@`dX;dk-$FpNU@Qww$|#ZSp!gMltClibN+wvyG4xV{~lAcx__16X!JI@ zo%rcFuY2`vN!7t-s%(m!Vn^9ody_LeKmt5qH$^gec+M$7G8M(;Nl>>)CziK_&YY5f zgD_q2aS`|3WAMBU%q0D?`As6`^0Hq+vKqG1M!W6K*nji8MwKRJ1-0om$ZY~Z#ndC+ ziNbsg5Z75`@jEhCFCt%66NK5hXj^?yVaRv3R?6yGv$ZH`LinQ~tL}$bBk8k3Mp^uJ z#R3lV>h-`iIrwqjbsT<{J}YU2+EQV@8R+@yE}PKkji?bPIwfhd5u4sH6G{YK###Vq zcgG6AyUrvWD}X=1q2@RH78@i`t=p*RbWEGjcy9CD_HeZa-N6vBvn_8}iBE6XJli)} z@NOpd?G$twI|%eC9r|agFssee*AiicCfC#X2n)G3ld8!Tgzag{*W-= zXOr9)8y5gxc!abtVL(Uu$>J`^({TkY2v;PVPNeZX5Hd2ndswFavvev`opW3X!Fb)p zm6;b8-b@$ePsBzl?)y755*g<&iD=f4Gr(u3(hZ&FI4(L-2L7#wylrH30VSxG8_+>o zFSm&FdPY2GWXNXzfRj!vd`f% z9Sb~+zCV!3x2bd|p?_{jE;k9t+rNM99L^Ij7jRq@`Yhn6?8(|PV&))mV0w>#&dgz2 zD#eSn>;mI;GvMI!rG}+GGZ1+{^zU`g0JQ8y95uQ_m1bp@mh>dYb6Y^Hf__#5Kwq~V zmU~+LXD58CoI; zl2fr(pn3XVpuzJ4E^Iy*^4l-Bc^hKESFVl*xNJdNH-7&!SAY~ISF%D1 zY=08-J7Zm@i$d1uap3GQy>H6;hCVRIltOvu`lQ6&-8;G*olshRx@pc8s<`W9VKRGO zD{vY|0BKw@xeo3+`l_8hfJ?JiGm_tm26qyvkW|+G^3cxJ+2k?p={bS`;y!3{ozzto zUGJI?BBblQw@^iU&lW$;nGGr_cuMZw1CxG>yS1Xmx#Hj;ZjCtl(MEXpu$my1J;~oi zPwF8v`_iN&tes=m&_g9+%4l{Ms>%D}%s;2bb4~aQv9Fo26e$m+sEm!Z8yn|x0aTz( znE0zlmnZjqfZ8Tg!)`BTIwt|RIZi_hAfhQ#YD}mVkfgVdyglo@Ed??_VC~xObF$dv zEfw2%ZXzgZ&}cgqBmmIv=Nz(2?w$_c_uEf;w=OKi#jweBW0&pWOR~`e;c)SdQ70Vz zAk;0E_l;Z3=i`(+u|okrT9KQuks1frZlJ0_OvnkG+QU;)agkS3`maYZ*|mwfuOaGXm?O6Yb&_l0k`0hO7oP#6I1wNQE2Y+}%4zUv(qjd{URTX4Md;Rz2F;IM}V|*Vkt&X@jGLm1i_2`VkThAU zv4|z#{$d|kt$+WskJ8;cYXS*cp}Y7oM;3MYd2b%!e4wZB5PETGrSfn%iMg0(Mu{gt z)s1)Wz&hLKIBaiy?};7u^CV2T(JdfGoqtl9%BZge@UA&oNPddD0@T#`s#d$1Xa2m0 z#oF6ywUUv1_$z>Nv)~4>$2u~Tvf=HG26MV2aS(%uM)WL%4 z0vPh02;T)Xc<0rT(kx4*3BdY@vw@_tsO8l(djptgpX`IUTCo2gz`UtpC;LmHyAIe8 zMnv77Cm~E6yuAnB0avFBP7l!Gz%6^0ot5R|5XQ^^^gzGtoz7FXQ%&6Tck+0!Uc31p z*Ar~hBb!K)#hs#tf?uZvH}Cb=fJYo4Ih@X-GYHq^DX zhNo*;!_P?78{V=WGhTA^@kTa}rW((V{KMs3Udy+f`47m}Q~zfb zU#Bn?^GM-De!AtI?DjQ#*3uSML(R<2HlV1NLe3kfG)>cVSz55nAK}B#!O`l+)qBNo zL`7aRyXiu)^IN#4s-AU15;gd&`8>j8#pB_&#hQU&4z%pu>aj7yqh0#+ogdw^=I0#} zQnQ>9b**7>5RcHS)*88$HNi@sE-T8l( zD-#6u>d%CQTVB}O+TK7k%b9omP2y4pK(`h`_r3mb!J$;51GM;vz6ZymKMOk!@x@=B zthFd@RY~1wrRIL9jX2(3el#D4Z}(3z;b#@m@?1P^4Q)?xuFlu$a$KZtiRW(Cd74JN z&M_K(ucwAe0ri?{+g;^8H1Zh?ncA z4tHT_`pNTj#VP3n0V6K@yQ<{gJ>u>g>V-CiLvQKd7Aw>j#hJg)4&J%4SFQgzo(*QcAx8F!bdZ z^L|qrTBQK${%S!k+l&OJk}WWFYN+1gVzFQc!1e2zc8UrMA+Wi#R&1ol?ySi|pqIoC zVUwu?xo9$Hzx}JjUIuVB!VC4wJuaa26INM8XBPG*s)YN|t8cvZdx*NtUp;P$j^Osf z;$X9TQx!OeS@U|nm`G}}ysIM_<4khxy+_R|##_TFF_UYDei53gq6XIe#~rKPl!$fm z6)h{H@BYBLKMG-GLl5IhymdRj`TMjr=4kXk`N{D=46;ant zL5zSL1lsh627;Sdf~JI~6){{D+B zXiF*BMVOppcKcQwc8?E+uwK(%D&q}4j{7AKAzuwD?x6k3L3Gxnp`k+2)b-`&>~~4N z?lSAJvd~$V#q4>5Ei!;mI0?a?K7Zv5z(3w zj9)`v;6@w?tXtgQ^PMVI1tNd_(d={XQ1P&|>Z3PF9=KyL=NirwuT@qD;qTf ztO& z+e--`sq0d#-=x$DWb>iZM^nWvA8U6Q?=LrLGe9>3qq#aIKLB_W_}kw2N%LspMz`F1qG68wL3y#-X1UDpN-BHc(g1`Mr$bTf#G#L&_q z-5}kKs9;b6f($kE(A^@^DoA&C_YnV?=Xt;X``&-8UtDWgEag7;iGB9I_O-7a7ZSi8 zO|aY!0;tB~gXgAaUrkSn8&6?cjf4M8g1P;dPp)LzvmIBTl)8a(LGwRzFiPK3o;u;y zVms}98^)W^o355aWnFsg9w}gv{beEdxI`en$wE*1yA_K7A^bz$reNEsRyTZ6bs^c0f!JDTh30DzqyPGlmI^N82#`1$)10QxQUU~{Y zIY|PbFo?>SM?0SXV4tHe+k}w*ySV0ssyG=Tv_saX-m_`i0wi$ylpc@*qY_SFI5bT; z)a|?%EpME|d2myuvZ*8RS1NPNuV7m!o=%2cia%){Yl1=;_2DJxGo_ajvP$#(=)JFVv@+3 z*HQ}4+6}4%6SN-^A^o1r)sP$C)f&7ViXa1-C?7pM4quJ#vHMekThcvPhyMVh0UtxG z=Zl2QgX!F@@}j*K4nxT2ZeKb<7@WA=Z^42%Gs7T_n>zSq zr3%=vpBX}O*DWEo(=oXqXNQs;!TXyR^suhAvdM%eHVQ_E35(#ps4X^Uy6c`}&T^7o zPWy`tYU%i&$h&dKDipDHBq67@5}w>`O=S|BhwHH6@bT(RP!vAu>3)Q(f?l?q+c&>o z=5vfBn2e5^x8}bXn7NSdvVZ^UG_Sr({1%h%w))YRg)|v%K`xF-U3MJ)QtMX@8Sj3X zUI?F;YY2sV;KAk9wS=fAZd;!_$neNfK!)vo7cnn@vPB(PU9lh>Feflm7iW*lmv z2+(P+lx$`B`)^7rr($(*B22%IMF1c`E3Gh;LcipD2mY`~D$zExy=LCKg`1DhIp2zV zvTkr`&Ng7f&OTU`&pQi73#{yNeH&Yi3*C~97hC!YX#9AzM)F&$>axQ_3i1zc;Wll9s4wdWMMi>i)lYwJdK#A1AfoviaYWY(D8Ac|gw_ZLcySP?_})5HYlm=hBd0YR=V za8z~}v<>IYV43wlRH>95b~4LQvYDL6fSoU4FroAwKb@{28q9k{y?pvpz+0Q@t$LFA zhqwIiF&bhT5OEtX5@ntmsf9vZc}>Vp?(MEoEQ?U} z1$y?JrqmPFkYk?wFRL?Z*puY(;+B?{ViqiGy>)J?3kq&tKXRRJRlXu7J6_EQDzNB^ zB}OtycmiC#4cn#$DJ5L3t!D|60K6FC3aK@O4ILJaPOJ`j3aK#9(yj@~{!S?ei3}mL z-BUlKBNU6Voja{8C5~ehJA7_@+hM4mRNbWcGuzg}gFcI5nVLpDFb zClQ`n6Yp?tanLiCkl$8FjyiyeR_$ri%Q7tV&r)PNb>USpwHSS$_reD_n7m#Sz3@r$ zdiozVRgy2o(hVQM4-bMb>KH9z&%-e@VZI5 zEp2WZp06=Ku_CMLhLeD+S)JprdSslE8fVCP?*`q#q<-%9x>#%A=iN&s;*)e_f6Sm? zesI6EL49_tzm`=BVfLSgSxre3Y0t>8f5qM2?Vvhk8`Yu1dF6TT zADFPLaCdkMwPZ?)wMjv8{16#cco|Gq6DK-I%00ri&_g%TA_k_^2I>uMyLZ(CNz>VT zICghZ7xFepLS^3vs^aQx_Uj*z@$-{pc4$GFG8(}oEksZxXGEsACPn;YI{Qg#Hd6Ar zcQ5m?`~(k31$zzd^W(!MpFdHTkd|*)yDz z)C^SKtVSFsS=yFMaj1Co)m3x#h_9^{E?+x-+Kt-g36#)w#WmrFT(gueNKu@3@X>D^ zCAByYle|2sk+c%^;_lz$tV+{#N(|-IC%~C*GAOp--Qa$4kIu zfI~Rm3#rIY_3&ka$4Zv;Ki_+Ho_%|Vr8WIodn`oSCHU7-Ayo!P6$4uzi=?4v@9_~S z%|}hIqPD&OA|--g5^p)S(cybB{F2Q+v+ug_cP#^!8tCSfDp*lCrV+ow{FeRv>FWIB zZD;#@V)yr1Oc|x2&oM32=q||9?SBSE#Chd$ov`d^{F4xT@NRFJW@=_r zeAxCbu9}thW1t;b) z&u->q;xpb{4wDo~rw|^>WaNiAE6#Zpb|+=1OSYo%ZMQ#=q1)eiw-;Jx-P{YE9tDt86GpN$o(68o zaff5iNpG7W?ppyV;iD+>vE8Yd21zx%2Yb~DV$Estknm<|%iiVdd%2a`PPtX(mx7&1 zN}Opun+KhHe4VF%ziN*)%9n5PrFt+xnzKuETCkjX!kJtMSbiyvF_4Q^N7FDDjEbPK zaaH=_Gqts;zJC{%uaVzKYfEnUxhJd-P&Zz;p$fG>h~w=?eMsmUGkR1402__&?rvMa z3@YxO`3Cj^;$6OS5}z93B%MM-cx`Vq=KHfb2A(Y}v9oQnitFhZ{@4@Wx8>GUEPXbN zIn|f~0o=IgQk&BhFel8fuH&2Cy!klI39nUN!?#~VGTmpy zezyU1JN8I69wZUJH%C2>`XyZ`scD1Tm`)bBBi`Zm!u0M(Zb@x=Xj#Ay7X}l(_#j(| zNLTHS1*GDD)^EtNMT)BPq{D#w*@S3^wBsY5(vAiaMOX~~NX!0=$G!iI@ozOgeUHA^jBrS?jneD(o=5?YNo4L7l2n&roaDStcpP4*Ym{UM349K zXA<6JP25Gzo%J1tCkDNc>XwwZSIYB7QE*Udpgtj*shMCT5BA=J=w|g>a>GKN%6$Z4 zANn!KccrMt9>>Z<0{mrT5IkqmCpeu+?zAi{Y}uE5=T3%ijHqz`ceM9#;2>Ak;JGv9 z6LD5^*DnSA8h;{g|NVjgmFu7Lz$JHt1YnAjY0dW{E6x#?F%_%PT8j3Sdq>O%hAd>& zO8d(S9k}sg<5I35v;_w!hM*VCT=-9l$4X5T?ZDN~b=su`7yi3lDWGRh_nkWg_oO`W zLE^61l$Go8%P$lzQ2Y!Gpg#TIb@_17 zvpQ?iUSj2?9$ypA|Le1Wjjq+)X}N(zo|7oUZ8gmOyZ!{N-MRjJ>@8LUCIx&ZhZ<9p z+4&Wf-O%49xZ4-mXrhXgcu4)Inf#IgY4vF-Xk2XVB|4y@Uh0+6yn%!*F3jMNGUV5w z6r+S^BgLgywzl60lQM}taXyr85axLBAOZRnGc3w?ff+|xbtgtRh`N_l*)=ZU_LX$N zjyYgLEq!tKz!3dDFD(wp#vj*!`xtYO2rMF(gTQK6ZR~$uB!&Q0gDl=6h&f14q{FGp zAV}qKXheSQvAiW^LK5x8uyNPQ8iW>3X76)&6DL#B6NpFThF3;swPi$4<7#X2jQzy` zfoWMJa{%E=n!jm0&2n08)oXw8>IyST_-3qPG%VeQzi2jCPCu1F$i_4ksfr6DOe{V5 zUUzlzo8Kr5(u&DndH!{<#fuou{vv(njc`Z3=X(~is=zKx@DdgL+e8HLyPW?!_>`iv zFhtD1+qbRqnI5humZFcE3Agl^km|k<_qsSMZu6$faTIq_rnc2_^3;gb%keoV4os89 zXzq=03u-y8L{@9JGBEuMf<@YnQ7+!&pTMzvV_c4m>In$u<7+56$^JI_97>uu`1M6H zdOiBJ<*+cW9{+?eZ+YOWU$r>~3_!6ZKLRr&hQ2pIL-f3zqETl8hfPLXv$v&ToG}!3 z5FY-95O`zGNCI2>@t#mJ{-Re!5-Z4Z9PzAO|_0i)27hh@$;%thk-!8}SnxL>q@jR-%-twSE znX`E>uoP-`@c$VK7#AE}wtfq?Afox*ea`;s)bDsBCzer6BVg&Hl8q2OZ;8Xg_ct3K z*c2WHnKqU>p}9CA=>G-I>eF>5GZ6!xQFoDX6@q5&t%(=co?hn+VgHxXhyES{-ZADe z8R!lg*SK(flg4OY2tX2*FtS9S-_(9Qr_7Y$} zpr6%VOwlep8UOXe2b(gL^ENQE^yds_HygPiFD7z#DvlO<{D0}accXh~!Di(Mzlu^| zkAWM$fK;LaWfb_eX;vMK4)V0RtNcOYO!=SCMOMA+bET4(F?eQi*uT}(jsENvd zUiWUXF@^k91~QbbX!O|_omoft^nHg&WL$)iE$meIoPnFT*9qo-4p$}U$kwI%PZkdH zKc8q?WjSE#%_4{gg8|v&>iwQ}T6XSln>VJQhwAU$3KkVW@%x9c_Jwk#^uxrmlUe_A zi*C6Apbs3w!=bqHde*_nmIx0)o{n$Jp-3v~W2N5NH5%L?nkk=-gzG=;%ZHagKG38@s$^wGUTh7e_R zjmmlWZZQT~2;nNj_CCtGE%({5MAKmGCVaaF;-@@gW%{eIz#jyR`p)2hxFe-}pU~a?NP$A_Vtl5N)7EMnHw(=aTt3zWq(6)dJIQT zSRL7)!ankoR4ji128{~C!ICV*_rH2Rf@g|)Bs9mH-4`ufA;xMkQzSjybcv5Y_EO_@ ze9LV{TADEfu~~F)x8OSSBGPAbSgM{wHkHYhAW}!@8oEZ*gJuMi z;s733-)%ZImf=SX<*5~VYz+Tq4^etgZ`rU@ZrNMuS4^8V`VZq!E}5{|WDM51`Z^d- z_i5Lk=X!o-gb`o0i<+`9iYGXm*9?Jg^UXL(Bus^Gb1>=H1PN&{PMo*;Jc#r!F>IFG zUyi%=R1bc3pb()5qNoBMV^!Vz92{Qk|1S?6eX0S^w}5wR=IId!IFsbi&WF(xZ)Qc^ zSBN+oZH4sv#z50!q1{|SbtnYHhaq-1fIvo*-$R4HI6Oh8(_U2`?khk&ZOX)Z9y=hG z?H@KN!!r4Z?pTf64h?yQ-IrF1l~Vm4aVxT8HM1~EuUImGw7l^BDLF-SQDjm2^4%AW~enCsjL}{my@%g-~=gy8a!o$>g+(LJ>9l46^~JOdKuf zfzs|g+p`IM4AvJCZTBbp4Ipu7^^*C6WZmx(6TnGbuE#s)*sWqol{r*I z;FK?0O82ic>e2)W>h;HnhE7VzM|9-5|7QsQd?G|~VGCZHZ=_Z3pj;kDnNt3f$Itpr z(G7@U->PkX%NMI>hWd68iy*+#{Z1rh64x?_lvD+Bv%Qqjpr>%D{x@07;+fBC(qEb{ zH%nV|+?RWDB31;kxX>2V5K+I-i=DF?;H#{c&$>CR?r*I5gzUHqX%|yPL!}5{e5HdY zNXc*y-yRRU;hMMWeaB1jQ8SuWcW%&xz}ufF0U5i!-|h(f&|(zK(Nvm z84eI)Px8%cJ)Hk~;Agu@m7$>EYH;(#b6= z{a6=23CaxR6`Qxm7|3)=Pvfx;K3t*WWmc<4|EKH)oR3JkVGsz;e|Q5&*nmR`F2m`G zYRAHe9);#ykU5XyRf-=!*`CacxB{)S6@S&E&EZSD`O~>~-y{SsGtcf?*1a***4YNx%lJzNhpC(bN2-XeYc8tnjGf8^?(hfkc z=L>nfrbjXS#BXnFb}zC2Z~X%(LZXNA zCawB9#u}oXCOEjV?IgT*w17cIj1*`q^J-)RrR&v(fRN(Pa(j9mV^e4L?u|Lx@O=X2qtSBv1J5*?3HZQf>Hg1jYD~Rr&UA z+JVx~VLmBTA!vA~OH-n+Mn+{uc08j?^OyXyRl&1-AP3wnHB7=uS|BkiC5N6xU=S(a zKgr6U4MY`BGza({gAS8g;9O^~416^RmOezV2)#zIVn^2!M>+WDu0J7*)}9HZsu&sBJcrz!(MP&vzhI=^INyrxFS~EzIzBC z6I6Sg6nQq6`dGI3&Yg+aZDyA&?DkAl%`1{&%W783D;|I8_Y=Y+j#kgTtlAh*$mrya zP7$^L=hpQc)@9wnuo^0hiO;+<>?ULVnV)MC?zv<}HqVIbK zk&z9l(<%3A5A{fKKz`E%Ea_eA1a94tp1%2uati31{UEl(CMgiR%VZ!Zsod@EZyA6M zl!W}r(5+eLa61Fn&-~Y9lfS3L76V0=-v{^bwZQmZ0Y- z)Grzt1Sp!++#Gf;k`w0-g?Qf#H|KqvK=^DP2}@U*<$01w3i*}g0ihwA^3|W2fML^% zk>$41mouZiYdjc&sFIlZGhroJ5?lltS7;VayXqQ-B0MB!M`Gtc>}y^nE}mr;PtHoK zl2L&cI~dqw)<@|9%dJWi0a<~ID`;1ca=B{%6z$;rp;ik)_t&w#^BeSj(yjpCUOyI; zon>NRy6-QE zj&D47XN**yJ^Ot@vke$RPy4n^c4wNZ0{KCIX5SzhE-lEFcaeTo494@Zi{nke*Vifz zx;R?TYMaBtB{4Nj4CISr5K)x^e!_(~(7hRrm6G44f~dW>*;B}`iA{mM z%87i1)I3^x0h>e4K&-$iKQJ#v!i)KM`_3$ix25WdO0nA9%f$y|+7+q+#1qL1&Lc&W ztu)sQ30HuNL_a~ntuc8 z6@di=&YyXzxQE@v{UgBAVcK*WL&_EE);JmU_bQr@ltP$<+x$8jnZIH6d{5Sv@;P{Q zaqv4!DG%$R_wiaUAd9cRRu{4#DI!u?+jIowB}tZF1+52cKA-wF=^e24y!)H4$+03E z?}T@rdwL2yaQqCO_LVP9M@5Y^?^o>}?mjtmHmr4NE0ZoEY`8)f?mRE}2BNSlM+{g) z1eL`oD0Hbyees$Y{(%%%ME}yBkJ~UCp;LX_PkILj>ajMVY=2n3YI0&A?R8G?=rg24 z>P-U06ryf8;Hr{sliw(-XgN>FY_buV*ahZAjjoc^?7Yts4;Ed%KeXzH18oZ2;wC0z z4gXrV6|$7B?-!+(#;nV_v9@k9+<5M9(wkzo0JNnlmrsT50p$!N`+No#5AH<{U|MQB zm*gf$?YHZoy@7w{@xWgQ*ox6~g*^?o0<>;aF>|9(U*pq$Sp^xp_;>a zHA_ON3kGZ+@6<~yI1bv8RhdtF7OT`&_iroOJc;+%ln?@M&#>HdsJoHjFNxtRrT#E`XCMaP zIEh-}SY0FljdzZ0FYdScnHaqrN9jeg67YYgQL(zbYhRzgb!X^eFtw*93N$NaoKqH> zu6D0U-aZ7nYn_T2I$Pk(7{|T<&H`{3HXRm!ykED#YQ5j!3#^b<*E_m9f2zr=Nm+qA z+f0BDZTrAsz4Y^=88!?<2=B_;Fmg)qN17hm`U-))2ItW>9OLVh;hBY*C5$}=`+yt+ z_A??p)zPgK6Oi-C2Do98@DJQi-?$im)5C(*>wW-oOgj2PvcrF@MQk0TaFv86(g4To z?+>GOX4r4>fq5L*?$=9BCf3D3OYPOpued-jix_t64&Umdcz&UT0DJu#D$VmYmoP+` z=IZ%@g^;b;%fP(~!j0-0X;qtlJr6mp*F-Z)zOi|}#~Uw4>RzvZTFHDx(!0vOs{%>~ zczZ)b5WpWX3^SR(4cprgSZ%kPsx8anphY{Z@6m9VN)0~h$I(us{{HfN;8nqwUYWtn z-fL=TIQ<0uZ;#&XzHI}dgnEe`ngl+tk6K6r_QUE?Ka=#=>ccVuXb!%!j<- zpn;Q)2@A#&b6e2@fsUikRU1Qr_07C1Hr({D{YY8gHcp4DW5!g+uKyE>8m3F=bos7c zWGUSg4JqRdyO~^k;4DP*TEF+emQV+wR^s=Uly)I2vfr5z?F4%K3?Lc&uAafB$WTV`ANs9R^Ur9|ZzqU_VQxK)cG!>dU6nnV}P}pC8y#l>P$8 zb*Yy0XH?bebq$a2(c}}pZ5T7w83DNiUj&tygdLI-SvEmElicJv(B;kl%&zjMon~XM zn>bqFdIZ=ho*?_#=79A8r2DF>Bi+QiX8Ba(!1HB92m;`xO7&Z_$y4 z&$HU=8#7<8oxF!KU>EagQ@bbARUVK}mC94Mcp0!8GjI03ndfQ~fUm>#$ARk* z^Uwufx-4bU%hdbDO?%;jNh&E#&_jy1B4jy!#~xcF9N)G+`l#7}_G_icBxB7A;Go4O z3uy*;_8L2m&ub=CqNsPC)yB=wM6~#kG8YRw9tVpb%)iuUvMAxjcz&5{P|8vo0-<$t zF+(wUqGar2)2|gBfhr=)L;2dFRHwqsh>l^YQ5VNH?wYX#UEA~1U#rNSt4O=6W)M)U zZ|4 zn_jrP^u%*wE_Xzc>Vj5q^hicBQ^2C#VDg&EEGCRL#}VMxDf&qkeXG{{WZb#~sE(yR zPvoJ)fcKdaZ|9q|G*@3MkV<8@(NtbljT#TR8G-9h&2gr&Zd(|`7X0QcKEqmdEcamb zpB>oImfAs}%xnWxnv|mb5Oh#iDXfc7ZzCVUkVd;@lPp`hH`iBxK}&ONY?uwh332ce zkjmEq4AyLFP7Np)=MUd8>I?kI_GbT^tJ4iEQps=9=S!bA1P3Lvcug8S9TUS40Q|PK zfHIDfN}tJw1w42juWsF!Jg>V8cIzKj1K^DQ3Az(d8F=MedZ4eQ)&)4kS$ z0s?}wHFBQk7wQ!Mxxim-qtZ1bcaetAx-_w#a>aImrVW4N<0*+hur1{n1^#2J7mmIz z03jBhS(%7wJ|2b!ke3vx*BsL`(~P*%oq5_!E&P&i)Dhk|e`!@aKm2|!kcxanZ-V{Y zZ`vH_v7)t3d0*BSC6eMXs{wm6z%V__&YDdnJKPb+S2ab#?)9;CrnbmEu9-fq{d>0S zxk3d^7Z5CJ{&I;;V25(xORe)x5YBhJe%@j>lt_<|Gntxbh z)F|Q9R4q0xEMYOke4#i{>rL37XE1?uwcC;OxoV`w{MNTe%Ua;_+k9ko?)&RI>(6+&6P3^iv(A$jj@ zs~szjH01r}h%NXEVOf#Y{68%BUbI)wU0E8&`UW##u&r#f#wAfbyls=OkrTrk?Yohm zSLsO4MiwhE!l{S+<_iFmmWw@-A-Q1yrvJ$_21LlpWl&?+qd^vNSPFI>X75KOqU~ER z9w)VB@p|n{O*jW$pRZyr1K?x0vG83l070@>$adGQYPIuJnI5a9OK!|}L}689J;f$_ z41m4YgAiH4OQ%#}M>IHc5X$xQ^wpx@A&NlebrysRt6pR5X6b!)oEjt3SQde1ZEq*h>=#VsCKcN)sSxxT61bhhW6`wc>dPoVUejZ&nFgvu7LVgh6OynHuHLjue)_UaZ|Dr#2{d=qJx%kY$+g zwf^cSL`rOkyBF)XddPCQfjRun+a_`aA6u>9U8Lt*_UJzf6rRAv4}+ltrB&g^H)Qqv zKrV5st{7E`qM7(pqU=SCOkwjzus+AykV4t^Vir{3LCxH%YrEg|ICF8^aq-)5@iXEB z^ZPk7!?k~|1M5hp5z>Dwj}6-!Q5xg073_ zKKy2>I(wv%(n{xd)FjE(Jppo$ieQH}3x$6R=eTn+sz)6j6*GWw52*z&04Dh)r zeFV^N`Xow?p|O7p{#q@s+dV1Yh=mamg74j&%tqOKrZg$MEk!}QQ|#(K`D_$KH}zgq z0D!3Ke&zkzlDhIcVYEQX4wa21hW29KVB91lEVEi<26WEo#fxs2!svAkctsu^6C-fz z-$$KJ`s*5IjudIiILXD2?YQ}rjK=O|a=o^ z<(^ad!~jUrR2{IZq}*QOS71q1*v@nUzlm3Y58g+>cng}@M>|8xYz3+=O=E$N-3Tv> zbbUlbUQy3Q)T;#3!l~4lc{+uWiyy=P!bE4yxX#4=4kpy>E_pis&l>X*AYt3{zU}S( z7Vq;4F2BpVe!J3;dy6E{x~fyv>-hB5PJ7m#0Wz*4|IW=S{A1t>d*eXG#y<|p6yCAs z8)m;DuYSb%i)#klAG04?@aDzeKLkRU<*Wqh`D-SIUbkT-KdAk(w*?zt)M3K#4#%XP zel=^yfztiBEYXX?o#LM)kp0{nwm7~MX@N?gAG5j*ogcp_F=GH?@%e5t*W81TDpY?S zh){tL>kg;}d%T?%g-dxbZAvF!(*%mWTuI@kv?c)5+-ObxNdSH60~&Z00HShWFQzV% z5!u^FVoDLnfP8Mk{$o%7YNpB0pjzSG?yWhh#s=gT$T-v(Dhy!KDYETlrAft7;NyL+==RagJ%n6yqCvrI{t4s5_VN-?EtHVzG~~8D7DDeX z%~=`@@s624>o?dBO;mE2Y*VAS1#R0w7!A5J!&FDIEm22@-g{$8m6C36FyKF4e98K{ zc*lxI9gXrXNBJ&>3BLStk82+E)n>V16n=dvRJFldV^2102#Pfj1W|bK4@n$)YpqPr zU3L$i+I?~uba`9e6)}JIUNyMPvHwLVV!wSxgU^p%kpZerb&@odbV$AI_iB8fHdLe` ze48yghQH%`W8gMMZR6>TUkYwCx%zj(=+M(RIw)Hgndv*KXy3KWD8n!W&`z=I1qDaT zb~4Vj^Ek!QZul<~@^W%)m+?-mZ`qv=Sh4_iHj5j(ek45-dD+44WlKlw8S z_y4+(F8gNOZu!W#!A?;+pybwWGPCmhcib>5N$V6nP03F z|8il@+{CnrHp!X7_`(pf`+Vmiq`Ai z#87g=mx%@Z2s=UP1%v!R=2yIi#Fu0_9VB@tjfVpgfy`ovIHZq`2j+YAnHuNWBg-)s zo`&_@RGv|}pVHjGQRm@k4c!X|xgEzB!4N~|A8FcSG~QNbMoV@ABOXAeE!`yz_pg#57`WSV_{`P!Uk3t@Q1=gN) zM`UU;3lTm-?Q?J3e*bPWZ2!Rl2~IhgJn5ZCd*a@%XTOeOko-h64gqGVR3`Gfn(cqK zR?e0J-+!765zz#>u#61@vGD~lvZ~Xop=O#?)%?3E|eR@@KiduNw_8`fk&Gy}a2&B@7GrZ~+Xw|cwO zbPqIvY)@JrRVy>^ylk)XA%;4l@onp*dt8)c9@m#Xypjw(pxV^p+>h6w`o+zSrkd0; zt<`Szq|;E(Z{~n;a19KxWr{i9x#k18=LgZg2i+W1%fLuzfegc|b3ah}b12c3g$@7E zl@H?Ll&<7vhtr+u3E%Mw>(OtYwDA;}py&iBKTW(?5!(Q)$RaWh395{3Jq}g}o&B2f za&rp+edd02Oka~Hva_j)FmDUORUn^@0jl!CVdnnC8LPZxorpKYLM27L*|JIWD~T!$}dc z)UdH}*2eo0F1ZZvP266$WNJ47Y#clR@-QHt%&g)Eg*EuIc`KeR8 zU0-6I3t&4R-J{T}jQJ%!05k7$humW&V3EAKk9yuM4?1$>CPh#YYw*`Fifk@gzZ)mT z_fu;o>xu_TLf-}XGyM53zSMdV-3MB`ZIcSPT;9wLFF61QAU2Q0PZ}sTyatQCIn5t$Yn%$TOZ)39pg|c+eZj zAMUMmzn!wn0EY&*G?MwzQ>TS}KJnG_XOylcL2&axbNz2iM# z;+o>(SoLZ7d^pTxRD$b*jT<=_am7w#wq%XRz`FW*pN;uPmU4K)>#`<|82-_ny!LCh zuIVx6A75OC$2ewb*p(10ke;Ce3-4#R5AAq(6^-?3*WAR4;t0j!E@c|i2e_JD0tJaY zsC%}tvn$ajRX6M89@`_6cClmkM&LNs-8IY_R`CJ zP(|yxqRXo^98MR~G9L6D>Ty-#C5a0c_fR{;QtO|%ZyUrScIBQsp-e{Q$4YSC8bjIe zhg7m(|Df`L_c1{Kd68|a29C2-f@=I9fWhVteTzK;Qan0^J-BON*@BIqA|IhNG! zL@Ypud>Y&Gk)t35>C`^J-Awv(2Q5lUkhE?#59-TS3$6h^lgbpU^|czZHaA&Li+ z>Dw&mw?#b%OEX}}xY)jO+O8h;E38p9fb8sa;)6eJp$!VIUS6Z&v`{uS&>2*|><{=O zzg<6@zALrcw>6+@CCo7EaDB5XrY4+Ky1mthN_8hmn#6PWHra6&tXy37F9HZ}_0x-p z-r}0mO4}hfOlA_ymmr!m&hL2`1H&u8}rCd1~uESr~D}T2!pj$bVP2(?V}$iuV_|LiMmIkeVMi{*+FzHW(p$-^uBmuFgsgWB2S)B8jI6A zR?n3~Q+>CEHKyZ5*Gry3Lg-Zq9?CNhXaJ>J*C!1qU&{KcaoaxxKD!o-c6Hip?4#~i zv20M$vYkCjghh3tI)5W`eP+?S4OlNcH32(yt4k;Kk67oH9f0~MHkUhKFf3m)rmM>E zMNtpO6*6S>dOKMpOPoy19{{DCUbpkN{LPS?7MJ-@L;fFMi0HaDX(=BKKU)2@P=6Re z6vlz`f@mGBorIq-Moa(L_D^`m^wLQqFUu)l?o@lIC@X*Vfzys4X>^kKj9%?z0fD97 za#sVFx;p6}SE7|B0RFg-Opy){C#DmCuH06B`qX7S7L$?1z?nswFUo=EDd_d?+8}e- zC&=e`4J4EdVzH-dFIOIQqma6kr*a8g+KltZw9=d@Uef@(b?p69sLNkx_Z~nEDH{=kTS6GJJt+cqg-Eliqa@1_Qw|3Ng8nPW~PT7)^-*vMp5)_ml;FQM{ z#i58pTOFihFp&95hc@vr`}<=@Ba2k`Ngt`lqazRe^tMy7+9m1b%JU$2*N=_Nf2a&(jVeW6rUJ`&La zp($`wvv;YIRv-~;!J7Qxa#svsG9Ujozj_o5i89t07C9YRDl_6x?0@@`((2vsKttDK zNrFd34Az>7-i}wyLkC-Td9s*KzOWcJBw>P-(rvmsHW=q#?R*0@kfq0b+tL>UYIjC)G4zX0i9aw7eZe*a|aYH?kRR_75>UnvKP z-#1HZ>mlGd^=hRGJVa|#g!?`zg{@QRg?iVAME!L9QAZ<3*xq4{n%mLOal;^v9 zO9tJJ%AvpjQz!o_ulg{f8b2i6N2M(3CWDA9W`1PRsX9nDHSAh{GU{+%+PPv|BC^gJ zN9VG9v?JGXXOwlAbiy-<^6~w8iElshZLW)?a3w=CZ^@;rHD4-~1#%~LmPGERB<`!U z3Q|v&49B`Zam{J-EeS1E>9Q;4ZOPYbc$90m+3>O`wO~n5DZbVxzjXRL4?i)PHf3h_ z4^@*jLDmA%J_9)|S9;m_(@+E)yC#PeE_#}fkY^S6?(iulbeR-dmgis=C;0mE9UKRy z8-7K|+LP%x(X9C#*~gf&U5!(-Wjw+rf?@NyTDl^+r>ibT^l0`(1Ly8JCTvUUhqKlb z{p4%U6`zjs-<|5l*Su<4T5&-Q592PsU23L!v_)Q!94zC~2196xbkr@#s3<5xt1+TO zS${SKmPiR+imz_ZHSb1pCTd>0Ju-WMd(w5Fg5<5@q8xe{!byU2of`DECtk=|eoO8awM+^+t!o zW#Pcg1t*e2(X;LkEMqXuBmv%k-qI?Is zf8Vr$Z~7HUcKqkjgQlC1JG#rw{_o80Slt4*4*Dwr5*;J+I%MJkyW0gWnB)Zc^@#Qd z?$H`YhmchoijuVlD^(zfs%o9_e$x|+KWfW(^9RfjsaSa;^QX64=u5Mk=c{ZqODrP6E~!0-%8u-`09@NEp{#X9Piw_3Z@BN8$5N zb%REBxam%iy1$JSkM%Ax=C8*z7x5@WDp!+>+o?G>-ulDnE6wMs^t+inr+CsujBUMf zl!xaMzj~{pu+91_Ees-vOyHvKnqMoD?VHsE(+iz4`>h4X=rhM!SaIY&L;$_=RA#3^ zHu*?>o*+5bPm3G?!<*JVkQ&M!461zF|9PVP4G!%Iu4Bd>>eYh6V&$GyVwL<5#Gi-g z;e#(kxzjQeG$ArUIGSRcjy;x%v$>-*k(})DFQl^;etgxNlV6Ft(S%Z&d)}$XUnVZh zI@Nz{=lV650~D6C^q0S4rZn@BQe&oc9?!_BP|+?a88#VqSc2E)ctu_#doxIE$07p( z0@`{?%SsT+>Ni~~wK(;-k48iUgAqsWQeGz+-VE(bhX!dL>^`?6Vuy{p_u-Eizfw+M60*Cx4Hw(0 zF()iOBpQkQ_5mjY+UC-V`5vJyf zn&=1=)FV4%JOY206Tu~IYwk=mi7K%n35=Q)U7KAm{LoK9ZxcWaUUHTNS2Ae7YSj-1 z+>k6s3^*%ZIdxdV3H{nUM4PcEwnI_Ip+buPrphyD;Iw<|;a(YhC8odEA|``lT*0my zDRO}6K@CyeeUQ$L(E2IvrIl9acvC#>oM)5B!=srERGYe3s>%Z#@ zHcGNOz$CR2nfmI@pID1Xg{;cum8kZS+-I0GB?ar|Kb&GwDya4I+@hJiZAZkQ`hQIe z7(D6GD>m6-)sE&_pK*nyXWKK|bg8Hu{2aYz{z`#|bLU~U(_{*%DJJm?F0?@ct%)W} zuIU?*7Jqq7U+!|VkuASTTU{Q8ozc5H6G148pSOMH`fI-!=7Dhuf5i96-!YGs)}L9IC-upo6s6l z@6sM@VD5{^_4`@VWqS^R%o`2I{vP`>ru9#^UzBQ^Cr-DcX3aKWk2Y2_OYdEJzixf& zFd^;t{eoC869n(80uE1bqIL{OlfRu{GA-@HD-8Z`6RBD?TLanTfM_h z+F|}KAz@GnV5Of14D6F+Fe)H;aQQ>jUK1vxM;QLNocD>};Ai`>S2$!$zs?A36y{+F z$ysz4PFUe7N_T%@zdArQuGA*Em|to!igL+3wJlTlfM<;#|CKwDvdt#imGU=F?AJ7& zaZ}ZZbu9BGgUQX%9a3YYBQj+d5cm`A{xSy(h|B7pWJuuI>T6@t7P$v}zULLjMZ|p# zIvMwrDTwRy0BbtGm+@B~M5gS((vC-@JBcj1(X<1$@N>~O*c3xZwpAGimiA}M>}AnG z^w-@e!#(MqpGmHNXIGXQzqxnc0D;fXbbU!wH^UKKiQGk-d4G^8F<#Yb-=rbxa3g!o znfSc3T{p@w)jG?y{|;8^z!%`{dnlT5#k7o1>Nz#HHjSggMpJj}yal)tz?Zs=plfs& zc}sYl4l99xk7ak(k%$J!)#<_^7NHcS)MQB+sPBqaZ*L<>uWoMqdPdD2iG}zB(vOOB zI^28H7=@(6wFe)T3s#wYN4q>{C zho37*S)^2`B1A#(qLSrRRaIfU>UmW%hBZHb{*1M=vq5?=@}UFI&UbZnb>b|yZ@;14 zzt~J-P+M9192^`ha3tqq;RLxu6cu8(LI)v;c3fp4h6i>TvHF+((fb4oa2-dw*ysy# zNd##@9Im!(I+NXPI-AZaeDq|{nV2FtpUC2?xg}<3gkfOPbnG$D=RBK z1%(l5kX9Wce6kPVt^mwJL|v$|9=fp2YIMeKxVR(^%=7&M3t<;B4izhJB8kdY2z*ao zNuocDem+Ra-!UGQ!?a!|ojN3SehCkq;oqO+FWzVp=BIpi>m4yDR)E;T%*4*F+Bt`a zzg3S1T_(5936BFDT#{8BO}e-vKNg$VD*NDa{OinKZf~?ubz#(u{LD0YQ+AfFPW@0yrfMhCpy;OJmfy!`->LyiAW#m7^6gx6MRl zV@%|k8ywJIZriW%{5K(w8Y=B+-NEf;rF?}10uD{>72OTU53&o!#+lt)uQIkA(%KQb zg>bGv9xc?@n?6cuBN*F;jN)BaR}FX5(htxfh{k9Ad3;=-wzHNh;QG6YKe6GV@*Wec zXgtzK^4xW1u1E%6iaa4EAqaZwBk* zxaz3{sKr*9cE_SFQy=D8V-;n@W6Z`*m>(*xsBuzY(?!bJz^8kR( z$bsinWAnd0w-x7ma%tdSw`_hmrhilIHp9`)_^ID?ytJ|29JvcgG`^pQ6FN>&koxVq z8nm6PMiFd&vHVxZPTqcefmZ-%$|WxkU1xKs%XMY#bkv;?HgsQ+19RDh9S{LB1)7IE z3o>zoa0v*Sv`HSk{QREFEUl0O57mh%B57jR!^Sy2&Aq)pZarb;Ll5d23SslcL_xVz zSt)s>sOS`C`OlCKoypT-0&{EF9RLB2YLl`iXcS&Mt*Tm+_WGOInh+=F<@?qnF0Y|Y z65OK(eDi}(7Zb}}qkQShha4+*DQP zZ{Hr~2>_mk7eac%JP-n74d!Yk4j!MlGgZeeu!p2eL7#hQt62qwRB641$i|JiqEI0Sf_3Q-}Mb`(0idEjb{f41`g44BP&4DX>)`Pwm-`I55DjjX1cj z?bWAK$ytwdTHW|G3xH{R8TK7rX3*JuP$~tYC3bIemgP3x#4}6M@0N~FP7Gl?OF{4` zU}R!{3xA4Id6|`8bz6-icSp_{SFlUEyUp?f3msN#25-Gq?7VDq`bQtaojNxxa$qvT zZi`(X!oxONxF0U7mg#tTZzZFS{J<5=hkQnp2Fa_3(_=du?d2Ihj2VsdYXbF5sX008 zl}ux2ff?8PR?`M9hu2;143lTwgZB0eI#@E%_|f=!hHXx5;VqgB#qU1>GrGs6wk0yDwgCuV9^yT6(T^}Q8Q2mpp!xX7<(>VO+B()tN~NqY6MY~;`B~B z`Qo^uDZoOhSJ}9i&se6`(-H%j(eFigAVtoD;sbTK$=%)EV_>_Ei_#Rp%E?oibZNz} zdwY9T#Kc~pAqD}3*&-tH0QO~#gF@x_~1$GO?8h;AtE9^&FY7%6a1k60mMyNB579>hlS&DzoDky zw)*o$6tqp#Vx7j49=hZImks4^PT1U_@!4;=RXV` zVe+Av_+xy85d$=^Buz*B+u+Z>gRx&@Q}u0J zm0_O8-+9@T0qCn%u)egMh;m-<#+2{Oae~tX33Pk-RNm3t=)Ica69Ha<)G4kSnyZb7 zqjv(J4Uq$fnv02#;|OKigK44f3)Sz(q44(~;!8XY7M@pDw5!@fECO@SDTReQAur(I-XI|%J(KIe zL3=FR*I6N7S5_DK%ltaZSbHD!CObm}gDjw*QVd`#jBND<+Tx|71P+C@xE{4=5}$fC24*KYfw9D6zL6dTR)qriKWIB8=iVUVsPW*{xZ*&Hb$lqZ2QKGbx^Ca9 zW@&?nzw`B0ClQ25)12JpZ&@<_GeK1@Ogl#B?{Pc-X>_;Z=GjpdVu^yxSjFGkEc~Rn zKxO$FIxZxx_h;Tvs>J_0)#JL?MiYx?YQ z-0D>XB2q~;1Hk*3(d9#${QxurM zo~jGR6bE`sZtAQ}dfQ?mBU?tN1mrpvX~;f7qV|E#p$kx>Y6fff7HiT5w7vYFS>V46 zKfk!}Qg794(iRcK^+6bhWw(x2%GaSx`5nk)r_6y$fS z0JxKTUT#_R`$95n(i=Jgc*11;aN?+GsJIDS`oyTn$l7!^rYhw46sYG84&8u{cAIo| zv_+7&DD<3FTqRZcF(Tx!L`q|eLE9G|AoO0aQn_B!roV!YVyF`6mRxJ~#b1i`M-VmF z`P&Y+7{{(+cfH#;z#5)QGu9_ z!FT|yaIRyyUH8sX+F8iXuw8=}(kL=?>68ul&0DI08zPed;6OoD&Az_A<9UGSRD%tC zMbDf&zlSB`b1C$&3@E~=#Is=9 z7yYJ;=)k^4P6sh=-2?)IhyQ@`9|`0&@ZIpqP5?^*;2;zVe#u51Ud7CdEpjI&|ncf2dne|6rg#_l)f- zr2=bdaNoS?kxfuZ&6dE}bC4`F-rDEg5f)EzuG!8DA8-S&~35#*S6$-QAX&gryVnSRJ0{k+F93%1nV)rRjvI5 zB?B(W5PLAJgoFMwolRnI%d*a+Ff*84is&$QlP|o?j>976KBfxIeh&@0IqU%OfyJOi zwg@F*;3(z(rAkp6%4^U2mCx;WMOKgAwtQaM0*5r_28B2z5)KuGalteUU^ofe+5Lqs zlU3UClVwj84{V%qQ5-K(^g+`Mp~!tn2rje;R3o4}#^(z$i{PkZ0Gu}YC6w6BsGw5< z`ihi?_MDt#fZHIiZ6D>LTYoU2!NO>7EHVQ`=jQL_-wK6$N~nmRIFRPNp)w!C1d>Pm zFd@5wtWVK-KfXi*zEJ%CG=ZXWB;1d;4aWxR6+qcJ{_bPchk8{Z5x~`QZ}d9e_>QCd z_-2Y7f+PVhLg^kt2!{psH}Bu;t-PqBt-T-~5D;(_e7C4Ao!xUC4kiISGByh zzTR1+(Jb-!Y552n1X3O>W+q)dywHvE_3qQx4<-YO2QWpaej#Rw6F48}zmMq!oH@OT zw-B}v=2L6f3?fxG1+d*;h$U?uskZ17&>XIJ_u+5xS>qe!1p3oIVXP{v!xYieAm#5T zZ5|oem;0_;KXb?f>UU!&GGNE#D~uFJWBZP8)S3T6<&XggjW_tMSqP=;}dfPwT8JeUP)eM(K4wi#^LPG;^WTp7#S5xWTHJIb8{oB?g*7!+oAe#J8~5ORyG{zY+fwMlJuH&&`Jed z*Z<6zKrIKERZ?=7;3wg|ITbT!V7K{o&=pyQ`a@^W8ynf37mCP59sC?t@wdE^1c8=Fp zti?3n5^r1NeCGy7I;mCQazFg(85OcErLSI)02(=$GX)|j3eZTp*Rirn}8bJ7l$o@wN!r~#m zP3@!ho26wSoJ;z!;bX05MmCcF>UDlwj}5JCM(bipyMq3GGlQ;Vl#>9U>HMMLJlQ%q zdCt!8>9*OQnglj|$$L<$iEsQ39(K)m%lDUF7}^TVt&2!d0{#0G*#wo6b?=~>Hr>9H zyxKMfvY5Qg&>5i!5mhfnl#Yax;KI5tsw$C#m8j-sEW4Rxk-c@buh#4Y1Qw~ZyYjje zMCB_>w#=Ljll=V9mmWXhMQ6myj}7R2!>c{x8GaCPTAt?4xKjS z<20RaTKG>{4GSd|87=Tjbq?VfKGxuyyD&kfxBE|?zBzuZ2LJV(ld-_4aNF;QL+|2k z+(b+ol>8sRP8rlPwwJA|v)c#KaRRvT`wYDQ;LXTLq8C6E0fU+M^zFB*8--y(Se5#y zQeQN0JCp~-1yGTDZd~#X@qW`1py+o$6_%ga=*%~T@2?qrzrl|rB02^1d_8SYuSYen z#nXkseO3q6Az6hTNZk@TU$%&Ln>!l5@2UMzsRwyz{$l;0y92)=Pmc$BV*rwZ$D@Lr zRS-R15WG+Rc&3+4T-CIsnOnV+cV+3?hvOu2P%n>n22+aHi~P;VD&KWW$Yy5%&5YX? zn#-Ye?I@w`qw4hemZY{DscSVvqZSwo0mvmMKa~7WS5)UT5M#>X`{t4%ZG2gv7gwS^ ziu=sNizK!2O9I){H)?@O=%5DrAO;+nEqk&jF}7Q241Z+-_z`LWKku~h$}PS z?fL@P?}%E??y`YA#7|A7-GVo6I+4uggdu!W*d(vKYM2z=mAB{=0{=C9$U_lF?%G={&@dd$#Gx`?MSNOj` z47!U&_>HCY2P#@a?>4U|xNsoe>95GQQ5Zn4Xy8!{9o7qpM*{${;O|N_tA0w2bb7HX z0#Ko;v|ni{s7uWQHJ_kLqk*lUQCY6%JO|(5kym2H+_+rUIJ_f&!K-$Uzp&g;8x8Ui zGbAopW4vDoxb_Jnr@xQsl1sqkM;h`3Ki)*pr)G<%I8-!K6o@^ zBcaQA5PbTT*huE}CqnpuHbwvKA)6qhp@HqKI%S`z=zU|Zk9z!Be2$L*Lv6{yuV+l* z^8uHmdm-4=rp^E8S-6WAT= zkKsN{xgL!dW(_KCJ(s`Lvw3r)FnPV2v+Rm89oMsb+VG0oyqaD5?v^Cf>A;*zr>TKy zo~O}#_`Kz?R(RN>?eW94E5-1&_Xfk5EBBI}czd~xR*k1(y+x zv`h^da$#3Qg#HS(qt zE2$M#eQ5cakU{6nbzZ5%`+j7Lo3y?PLa}}Gl|;`0N1LNb`9QvQznHDQqQkNAos+^R zK@Q)}OKublxNyez|2cmNg3tf_1yXmvLk>5gxL`3B28^6{=^e^5gE?0Q>N_+#UpwCZ zVtau@SLa@VbKsa$2v zY`07gVQqI5L8^(lrVmSVT>x`N3;#-C>&p)P=(jLKyroD1}ZoE>qgd+ zXd38N7;uNZZ7+;|>|XxdDp>^Pedvw3NnRWt+P&NIchiI!=qs=;K~xyA9WnihhD=r8 zk({kNWjZ_yH;M*-B(NQA&fzJ=_b{{q(FC0-GygQ5t=#Eg?^b%_)J@^!MdG-pwWa;# zv4pUPKJ_GvQ0!}8Ye&u-Q0;XK>WgqjmH)$9f@&1CTk?0T$|wNCEGh1+T%>3}HI3Uq z$Co)7V<%j$)K^|&d%rEp8#0+HL7@Smw&V1Qmajf)8?@u>yY?&RXek_o5)WBjTYE|@ z7~*`??loK_d3K9TFJY9^_A$0c&M3Vq=h;S`UH1J}C2W1Q;mO4D!;3!7RDkNRGPJ?^ zD8`W6q1-(iU2EJV8D6FJZFeX51syZ?79SA~-w(yulAVR_6d7AA!6~$Sv}~1Te|cF~ zJCNDNS6t)zEQD<~&;2HtM0F0M&&Rd>`E5I=!OdN%_01_Xjs1$j3Ijs(TV#&ziQpFu zv@Ap={jteNX`c+!#eeRtsW*#b>>T5hMpy6I_WO_~`UdCsE)1Zv*E+#TV&2cTtZ6E< z^utQ8X7!+rIHy@%3PjexDhup1+f}d*@D~sW3=AZL%*br=)bs1~#QZfkSQg6WS2I3DvV};Dizzp^{1!ftn`h+) z&^Nla$Ht!Z^=ExUYbUl?8*y)d)TgdmstCJXk>%89R-}x(efMasAt!wJLj7{^aixDaD~_?-w&Pn{_Y+M68*zhpCZ=BPw;^dRF)reZbwTl(L4M*pLVzj8!Zz<2ZF^%&;Y!eba?CW;SMM0GtR-W}6x?kK7pC4oEs2RJ(W5U=6&&NUYRc~4RK9lZ7 zb&_sA;2NocAL>3SC2m@dL-hr9k_QGvgC-{cl0BEH;LBG%_gALHWN2Mc2;t&Do)EwL zVB3RzTD>Q2Y2T+}g%ue7Y>I}LRs0k)%KLhrxG{NnI*wQ9An{RZAYVL?UM-vA(@AmL zDECM%WnK2GNSwNu7z*asy+W6ngbPjX=0yu4rFsV9iXd|vFH*8Sh^ws@E)0t?V4RpoRbHSWTByRECM{>9+uZ2pc z@XnsOt799h_jyc-_e{B$5JRQj?eC&OsBL83Ua0xBuZ8wiueGlm`O~u`HA$D#A15XT zlU`Cr-2ldC!gk$bmmGqZ`tGs4sc!+RN3TFOc==u@D=08<%~|W?Bw*k_Byv@r@hau< zYyQP`jjl2#LH_CHS6iP>bExr5TWa zJTUSPVIs2Vl-1So?S3aVmpj{2zlp01tSe`8&Zo1r6KY_)%vyXHu^MX#XLm2=TYjv9 zE#ozgEb!rom2tY1E3SQZ<{j{9O?~K_=#-+sH7DZOsdDMN*ABxOskXrmgJjY}b{430 z>a6e;3_G%EnRFz#AS@*jFMs)r=g)GwEb^x}TVpe=VJQ-ue4&os?5M*yfbdtu*qYh# zI4$1#T{JmcouHLm5x{`Jqc@Ko76MupS`gr*N|ptJ(JJvzFwflB^^WjtXOQaq;N z?QF+7q>$xzQs5%AZ85g{QAh`()PC0}o@8Ga*YH5FLWl#@7W=&M;x6#|WIcy2dbybv(n%Q z+vRE4@M)6`xW~_d;-e42fayD#5R{yduz_uxotB0S#TC^NMUzOk!)Mxg7HWGlvHLmm zD#MUCLNRL*DZ=yw`O#o{4oM4?s^0m(f-Cu9ij(S-xml~E{3j3loQ0(p4kv+D*#vDA z-_)Pf8;u)9hy>=B6`GvhP?5*OHZI<{8j!fo+UYxM&4_|pzHzk(hPOa=PW>M4k8agTWJffHOCS&jKQd5tn{gbLI3z+(1hZkzT zMNLx;wZz2O)_(5cMkQv8T|#gz*&%IF?wLC({eUFw*(t}f56lGNid0^lZoDug!*#OZ z7bi-ABRH8oCQA>@<1Xd678H(?n1C)Q2aywaLx`38oeI#_M z($Bg|O>K2K*uiFTud#9w0vtc=&t4{Us2U!>tKfjxI;cVT@AbqyJw3nj^en`>U64GZ z+H)9p!{?Mb0aZ>7AlR-h$BVTP)3#A)yfzSCAY=?cHhK_U1EG(cDa%=qA3k}b=uRhF zpFhHYG$?>HOTDN$r5Ec@*F13|I9}1Wk+;%-UHreEKnj+ux&zvZL`utuO(Yc+Jdjw9 z2#;cVM!UiY89TFb&kMw6&jQ{Xr_g>5?P6E~^KZWyC?UbYSZ9mEVj}s~F4_32d zmO52-#G@bT6D;29%BkU3<`2G1o&Mz^XSwnNfzxw|B%Sr+D}@hrfyGQ8j;i+L<|pK< zA8X2A)qNoOsO1jK7YR5jhUL_uSq9%U1=aaB?2|sO|fRUj(I}zd8 zx}Q6UDkVwOu=w30w%3eg5W&uaPbA4Ls1!&hfA@=nR>{B%H6M>zZx=W}Cw|WdufjLO zsXOukG{+MypBu^EQ0^}Vg31VL7C@oW#ByOcIR;?QW@R z@>G|br?*Q9Y0L-JN^l>%1>7ixhMa~sx{~{S`RSLiQpJYddZ+x))V8%fx3Xj}bwb>B zXB*e@8W}b6WM@CA^=R6DGyTj$B;df$wAPm4O7n5(e4nr3wON`FlUqtxRK&8X{>hcVL?4!=7*!ZoSHQJ~m2D`5RgUA)TH^U-ImmcCeXTFTDua~GDm z5Gg0~U-dpKY&$kA#bn6Yrx0Bc%dw>5kv@L=v_P@XW`=Y1s-INA4q#wxam6ejB;Pci zW(R>{O+5Xhc{X8_v)N46w_I~U}G{lvqGaYB;cfC zN`S1U;l5yu^=_`Q=`g7aed|puIfsX1$)27gP5y$q(CuCdlTO@zv)9 zQL?pbJe928eVNMqrB!3XW}ymW6M_5#pU*7WJq65pFI>Pcl0!k>p8C$?S-j|Cdu!$l z(&I8GCDDnH-kQWZCgzYG`xS+D>FAP{QQfGHZ{oNTer->#lAc%Pjo|frrZu`}gGa0V zd|aH2bjhzJ%(8Ke$8dxV3x#dirGI?QsqCA8OzP;wlPi`bc7Yb9&sA< z0@LO!8Tixt#(MR5+p-*=3PFMR|JwtChn4v1cR@B=fy>-?^8hB2AQWIUb7v)dJ%aMq zWW?g!x(0O+&o)6^=Ehs~La5nOgocct0njcF#|zOD&N%%Zxrqc#g+egM_D*HL6~Y!o z&VCve5Y#x}j{hA|ZrvWL-Y#ldrdgQ~{2q;KfjBa?8cDspZ*<#G2qTCwRK>pS0m#t# z=WE>X@Ld`j(SDOP#`~t1EKk*oBdQ26UXMCTq!j~QG^KoFYUMn`vl$mn-(kEv4O4)m*k;IPW z=$23RokQ%DQf8`w0x{LiV5UJ&$w+>#Z_qb$2n-jJ+9>~}U41$ha+z{f_N6;9rvoC> zT2s&_JBFM+W2jDS!~A%TgZglFQmfzvAWlHemq;??o#t#(V7`b8p7WNLdCZ*|Vq6Uj z^YY*X4FMZDwa_eQHFkf*uD+de8&tJY>mSLko+}t0$#!Yz zfjQdm|8w)+;78(z|G8t5Wk8;Iq1XJ+3{Nlr#ZAFJ6tdA;-+?NU^52&){ zB|SUWRzTng#HwI&3v~peUlCexC5c2*;++h86t!HD%NAISl$h&M(X(^u&ZQoJN?>0h z6;7T@FNXe-Ys9SF!CD%JuUH8vCJ;ZSD474EVVKsQrgtp|08wmvVWb;ZdLznO?e;D# zY^$H%GC-1omC<4`vhLz*6b+Md8hdFcgPWtwy&BFB%DP1-*A5qpiGw7MwT>%)!}eI6 zPu17iDUr~aktd2o3GQrOfgCH;xG{_I|<4t6EBXGm5N1YY98X`-YpS4ot_ zSITkqL|mz+QG(#~`kMm9Rj=VHAFAT6-A{Dh*AD zU77P;h!*WEgf(z>Nm1#|^zi7wN;|E?{U@O2e+wjcRsVyZ;zMD1V8=8*gNh)~95>UH zy#_thjY5Ouwh3yCb1=o%>HguD@oes5d!R@K^lI1O6npfN$!Umc93|B^A7y$jtL2D& zE^|59-=kqBplzakS=EbJ0zFo;Aq!z9saa%Yv(VovP*WKk&^ zolMCVe8&e*9#>Ag>7+aP9AF$|#@i9`QftVOX_b(Qa;7ECY^mu6#x_J`2(OxffnAdSt`~#CHBaA9ztR89lk!h);bC{-s|KR2Xge^B?^d65K2* z;Uq{MwyB25!}(K&u_J9OwHx_p6Yy*o6vb&33$6oJzZrjTn8JR475$Bc5(kcM4De{_ zF4dWO6PYL3Zc$F{zR7BpCEio|DJK92xXZ7Igkrm;6biG&rH3ax+uk})Afvv@3A)E9 z7O5tMLwH&#`e?X6QI1r9j71&SI}thjvpIqw=`)|%Qz_#tO_~WoSATZBBixa^eBx0& z_5K&{?K-$04nC%OZB7`F7c!pjzEAYe%E^XFVf_60u%BWK8_KEHKIS|v5Xw1G2-%zx z49m<`?9ZYkM5jc&0(rIIC;4JfU4ACf8qVn5z$FxOz@QsLNOswIoz>(V!))-J1n|{@ z(?>#1t0z_K#V)0tz)j`wmEwBh#(1(%=(_cZFf|_1e%4@o8PXuH$2Y#boI0&^vBfOwB9HwFMc1N&W*$ThHgEdj zyrkF`%vcyOum_DFXjwi`IfXH;x5v#%(M{gBBUuJj#XP90gOcM@UP|x@tXzV;ERet2 z9UYln>Tq253&DxNhrjrLK7eIU4jzDBUg85#46K_a_(H6G=G-@UfFyOz28Dqh93M56eT=P1R1 zXI|F~DbH6o`m{EDc`b<1K#H`8sfL1roYb3B$ryJ~D! za^5b%FV-}N(F~Vg=;8dE{xbZ#2_o-bx|D>;Ij14McD}`~gnVarxA^Az1xtdM_*qMI z5hctmz0M}coNs(%V5{isE9*T2)twsFa9JVmbk(?=q7gOxz1@(w-Ie+b=c_qe*&!j? zB|l8`bk^kbhCLZTWJ#jDTbPb+kI{KRcz=pB;gyTlKX1BAurCZKjoQ@#+(+*yU=`1e zxmL(?{pnHi*cCQj zT~(W%7VfXv1I`FQ5*mj;5rFY{Jf7*_k}fJn{x7E`xJ)9HJ-mPOXsG`aV0FN!7y^|k z-GsCI(-*jfa+=EU;*`zKNsB$l$0q$ZJS8JG`B zgPAks!rPq)qCd;4ijh1^6**bD==J00SIp0w6 zSBF+=bR6@{77EIluW!hw!5KM|G2@+u3rT3{HSV|}7=)6e%MU8$09xRCF*j4EPlt+1 zEYm|bn*07DW#WhVC(Ew%(K`1YZk=kT%!VVml&!E~rUPRbp9Cw$@JLp7A;JJ9_4H5f zzv!bfb)W80b>k8Zr9NY*)z$kr}Wx#>*9lTFc^fS0TWq7 zV5xT!`hUSde}505hX-b2PoFqIUVaBKy#Mwj`+U*O#EJC8Vp5N_HiZq^&N0Cqs zoF>|{)sbtDt!GX85TSRu>LaRGk-T1AB%b}MIB`Cfd1>P>@qUJfOG$p0yYyhPdEtjI zLLHB|KIgsjmwTz*$KDD#J}#+3&-dBZhN}hR$16p;CaZLY-ZTx*V#jXx1qaT%Ba*UX ze4?)&ILw%%qoJcfhB1^M`btDS)k_NLYL$5fR(o^Eiw0;$vHER{0ZD`s z&A~>)tYLb!$hE2Fm-3FDoNR||M2&PoT(~t^$rDGpux)#%3O94IBYfaf8M&#qlf6zu zXvLXi>MM=b%=9H%84g~|FLN?3#IoAY*?cYsAv%65d8-dt=C_j9yj}o~h*F^Addzqc zP_4C65v;y={h?%RTBZ8=>?Qf+S(Fy_1i$w&)t+(k>;0dQ=wVH9G)LXzYJX)5kU83X z3Cp~=l%r1*ItB^2LE!z_X46IllGRWknB8#3-$BiC?VhwwGrIALZLlt2V|oq8D&x9H z))cw-x79~S#3LiQq4=CdSMq@`6pT~M)Q0JAWFpxi*c`EWmp!4TyTh38mW%-?CDikO zir_FR<{E#D%DB`G{|C$wP-^1))9`!DRr+7+2?gUb#gO^byB_$wX#YWi?9nv*q1VBg z$AblMIBT(VVEiA;Qi6za@f)|iGh{&}jc)){<-2<7nlf6hHMh;{c5{CCn?gW!nAjz} zFlYXDS{UT~dw^kx-zOl}&s`(bxAlAwg5}Ekx#jVS$NAxy=bS~nMAz+D_Ox)BY1Emw z*YD16EvE~G{A8i`4c2d)sQ0;i!}H3|30tKaB`{^PP0b_3y?j5c!_Ut?j zVd(E>G0vJdy~FV8&sG9-;-vTz%`FQoF^3MaHP^0C(mE;z1SV<&ob&A;bAIP*Ec2W@ zWltoU!8ZARQV*Z&4<2y<#CZJZEOy-gIZQPE7+lyxFgA(%HJz%fr(#=vDNpp(^-q-b z9EMa57~`rgpUqRvDS^T!mZ{*gXKx!sRrcpE7>SCRB|oO|%}H z@;UzDbm!=u^WK&&mbF9kWt*~I3E-@NJ)hhj&<(h^xM)_l-_B@GpbXYjAtKJ$fXidr}B2b=h#*WCnbrQ5<5q(F+52;{TtW_)C;nXlU!qT z!gx~F^o?+BaU67Fxl)TVDQL6Z^7*O^%-UraGJSR!?kn zEb|i~%iyBI?-q1Gpu6)qXvpE%|LHUEXhAn{PcfXWhBCge7IOM5R{a&qvd| z`KCW_RJK|e0v~sHcRlB&K@(3T0{G{zWOEQrig1q$-ncmcxkkQ{V1wUa@m1N+dHzH9 zxJCX9^B>iO(G_&H>b5=!<)oy7GE@zj(j0wr1$p+l~5@C zJoJ4lVC?|lgg+ER8@*-#Ic_5i)r6^Nk_ky67lnC8y?-a)`i0%{&!ObZ5{vSVX4zzJ z1lq9HcNLDZ>AK`gZtJ;@msceG=@#W{V!E5EV2O7jB1xcwt=*6Da-kY3|l4v!N=D_;O>N~g9kfWq$l`;?*9YNJ^HJW~%3(G;!s zmLrHjaDGaNV|QBoz-rBO#o|FCN21tlntKxQ)9X#M*!*gz6y=4|h?$~~!q#hDaez&R z=|WwjLL-iGwp2X1Y(6Dt@q6@g(mapVbb^ixM&a1%gL37@mgROd0;f+eqD}{uEXf1R zk-~W1OhF05!3+XQ0D%PCPngJye$SVzor*WVu2je@H-m7 zNpEs^$+7vH*q7@M64o(rf7O&eEWm9wu^POlU$i)Ra4P`d;PWNghPO0@oA`zr8pOU z%2KfxhS`*+gEykG(-k-ABdiNLv4m|)Xt?Xq>%QC^P7KNg0NVMm;~KC^70a;GXGnwHYPx~nql+g{z+msFaXlrm*Bo8q5pYK@N~BYo ziomAZ{>UK!7=f5TE{6+jDB>wBhKhcFw#Z@g+Xy7kf!Y?({HFJ-C%)3a6_E$jgKhqS z`SN9f>(SC8QBXb*|2mN^==Iik2(aL9*8`a-BQL)jK1M=)-aVXjvqfiss$_f6)E9u^ zxA8Je_)kDOx2^2tA0MCHavKoBS!39g6Gtv==Mf8JECvb~#Owl&@+m&UohoPu zQ7?_fTsdy>LYhi`T737%KWz_*hz8%d~zj%JI^BWT)VpU6~6bdI~P zfV?Rh-hep;0CB&&^-m0us?>Qelj+oWG@_9K()yP~|cWl%=g6vwb+7dOeBQT)0Cph(rNoW=J{oDx*`Wo|q>Urc!4$bgFg$ zS7O#Bm~Z=viI8PC{XN~E{A)08igDpVuYnm-Z* zvdyG5x&r)!J#q;lRJA|eV1};_v;K%kjOiW1#KN8v{fE@r+mg2-CaU!HFKaD7V2Ae-J;naE***X_%^A5)IO)E6e^-*cKDvFaMO|7hE+l%RcT8mAr87c-{n zQ=J!rE4B46Rf1FiTIxrX{-#l0`9jnSB4u>+FOD7uoK~V&>GdgFbzc`c6hx}v)dVGo z(I1Tv6?9udHN*Oc9De!BkDt&8!d?l@w;_`@fBVt4<$P*3f|-0Atj&D4!U@8)J#8=n zygwa4T;dPtR=P_R(Du5c%(;|9m&@N6*3Wk0(M|k_b)@neWC0P%R@L zn#&7F*A-Q_@`n`SVbsxg+SUcs6I9-(h^)j0ie%(}sI{2foh+~dvLiAd02x!Mm%GNs za)gx(eea5n&g@rPBz+#Pi>gigr+|WSRzP0PnkO(|z`F&KIe#a~08U@Zg{wf$C-1L| z`WdzQ&5j{J*!4#*3?VY|)VOiT#>JV(-!piI4zo_OB#NHx4|$0MM|nj11xWWums)*> zE@fl$wjD^Y>-ajYbH()@NNV`aZnfo{0P5($K@rI^fwlr9=;XUgrBhmW-$GAkH;?S` z`e8j=2BRVbpHEP6t_>|hLLI+};>O_(AAU{-nk`l!_R!zxsL^@e2qPaNXA6Dbwt6ux z%t9V|NzQ{f0u)2b`q2LV`%pKC=&3?|6j&T;8^1amPQv;u)e>E z9GV*$lN-Y6Q(6ye_g@&M=Ys^Jec`{$BN8(-(jZb&(mnJ@s-PgvP(zE-Eva-2T_PnZAzk0&d+%NA`v+LC zIL|r1v-duGKd&$T{BCv93`STY%*z;4XPee_k?XpAW4IA^xkgMw0Ju({vqb|0#}@`n#xVakwbUrs;Ta!vX&`RmJ0 z46w^;gaTf{46kpYuJbR1 zUW$WJW*oe-S9B`OgCDK&Qe8=kg@E-cH9Oeu6J2DRX>Bx9s97QrO3SkBfWP%9kjbL{Dctq1hSTdXbzOuYXSBTOI(PRW{A?8-sRwai;OlTRD@%N|HwFTANy!a>U=|&DGA40YTX}DCcnle95#T?n3uI)k zn-S@#15U&Jo4|?RG8X(=y1}{&KV2@gE%P%0^2n9^WX`$jhrRQmChs@Gk>+h1f(XFG zVIbArGGL262AFA2`KJ>sDMczR>t20YrDLLS1~LUORhDuAD^av-jea4;PZUzx+uL~w za9FduUUpcU$+<3w`zKu9{0SAP4+MfF3XfF)H>bikJ|8_P&D^3*Hoo&+GmHD~GeyP3 z=m8#BYeIg5AMMG#uTEWNNrB+1n#`)rMviRG@#!21Pt`iUkqO|58T{IGgD4>5YqcJ+ z)NqZ{WGoOXFX%K%@5j`1vR*TwwV6p)X(zkE8rg4ctz3Y03s9 zr@X&o49x`E43)AFlSlm`kS)LFGYLc_nvDdH$RU7s_ZgSoaQ^oy7Gyge2eMwv7ub1J zcWwA~GNA3-&yyfe=d(IuoT@LlfTtL3+-j&mTbF@Q6xn~rXW0?ZXCp-|^7~T1_T_`Az($~+T?}9{lq^%PW;DaAzU6bmcEh~3 zN_syh%tt6N-;GFA`!Vs(GbrHnYQN7`V)~UBMj80hpln&4Md@VD2|Dn7fL^)o_l!ya zrwAtZ1IzQbU97JgZ6;p-UMqC=I_xbLE~c&L**w>f^Hvzb^0_X5wYmG1Rk!ztXARFG zTjX$>p8qmFQO>B}T@wh~jt8DvF5MQfT5{$7_F30h$1yL|un$|KQ9S^pwiJ-gItgC^D|JA1znn%XY5TH8aaJ3WG+g?AK*T zcor=rTB_D-;0SsBhBanCJe(Q?A>N zDwULf8Yc7o%T>q#SqS{e>TyubxS%>oDR! z@y@6lmwxkHBdr_qCRNOF{^+NnNEU6=EwZ`5&@gL*<2N;|%i-cW*S(TW>n+Cod3wgcmwfkx7 zInyTluou4kZnY#88d@Ks^g>$PckH^}?)*8i`tWIa2y*3lX zm2yAFRpBTcZ(oy?paGoyjx9!UI3CU82e`%x)M%Uk~sC+U#U*Gl7}=#AAQ z4?kI%oKLPnnWZP-Jc_dEBHvQ|)NG4|wnc)Z6brk5eA9lxlfz(%UkLb{=SWq8>!F8# z*}I_3l8t^E^zzg(A27%NoInN0d)|3Av;zrl);oMX+0-p0BfvHLhhXv^PYxzNWIlfN z%(KY+B(MDidPx6(0!AS8VnDin)gOG@CahdX)s@!A5*Xi`#Ol_V&3V;Is9fEb-Msg5 zB%7~(wb82Vl7aIQZWs5jhaEMXlrTbsA46pwCpEQ%9cq z&j5L>br?<^lPq5zrZO_(Nyrp7pnZP(*T*`y6NYH@q0_uL+tGlde71Kp!Vzb}<*ZkEpgZUpJ8beft5oje4d>yQO$b=uWx_S!)dnCEk zuck`*GbW1TkQRE+ZKzZQC|JbR8*@zQf<&4D9IR00Qms?Q)NrXQh`DU=J;y z-Shs(GE7MMvhP~`?CAUT#s~1e%RBbruy2aocpQOFkTi>Z_90Ykck&wsi?HPUF5!71 zbdgF!Oh$9LPFci_x(kYLWi}nD#Az^76vjTOEd|PSof;VaUiYps<#(##I>#t|udXMLAU~zE5h%81)yyaSN~=Nk^~3zx=zezuiZ$_j3t+9tw4Kyi zB`ji96jQZI{$@9?H@HPeIB}zLPTuzAV6yevKMe)jGjGymBQ@e>PEYX?>))6lEk6A} z5bK8=+Ii;@CVy8*CIM*%_1RdAF!c(_C)PMN#;*RfnyxRy3!!E8YOwY(s=^q`m>;X3 z5n$Bz6BmVtoFlpD6wRS$yVC4$JR0~!-x<0R7n>7uq4ue1mv5Yy|oGG#LpNitJ@ zK>`b(vj76TzXdK$dLH=eI|WCVNcL+4l=)gxK1FbmhkJ^=Uq=L(s-rU2ROWt#c@R#At=iPy912!FA znAGH%Mk>qiep0Rvfm07+)J)GQ;5sinO@9YHwy*oQ9{@IIZI|0TTiZ>xeWXnRCr*uc zaB%ity1wg=EQ-e;z(@_ihThCgA{vPPz*SgR0l^y;QpSk8z?r)WDMIx_LJ}sGI_1a8 zfdJV2`gO356KTZ)4Nn+B%oAH3%2Q!$GOAPKt+_I}A>$_1Xj-Y)x55BDZbT`+c(qT^ z|tL}=!o&X13Ph71h0{Sy{1&&oW#wP+uj~+uyaVl(cNERM(~V{$F1Bb(&hLS3zZEyP&7WHZZU2Uj|Ko5A7-W zIby$z0JH|ab|ya-$!3DC(z3{hpPZ}b^jii!cI&I=4L-@*A$|rdNjv^4m&KGmhoZY| zeJn6efwJCL)KxZJvL&CUBYcMWD!7 zP%p;R01iJ+IP_8P0!fSqJ!ww!qsv{u9`SxfA~OGJoS`m&yuVOaVi^c}MTk;1_2B1s zZSJq#83QWZ ztJuwYz$zQi#v|&3HTWzubBsGl-vk%FkOP)IKfku^l#^XqSyB0jGhnJf%{k2^<-;oj zgeUy{nscKW^3v97Z2%I>ihEvBQ(n&7KRo&Z%v5VdJ#@NaxvYncj=NSJai z;z7S6_M=F033~_0^VW;qPa?}1H(7UWXZnlW2845bY$Y>+6|EXghnd0Jg~w%gOANwn z+WVF8F<{bt{quHDB%R!XLUorU4>EG1z&{PX$#CiBOf+@_T@c+`^tWP@!#`VBQR=T z<~H0w4Au-<1A$BwQleEs@IY_CM>mky)j-VTPE0AhD@&!+LjZFbh1Lfo%7mL3BFMs zOQSz_SefIixq*o@YyEgpJGM1m2DSwOW|sTXK{l9)fFr!V!EN6a?jqk59DFQItfQ2r zhG!R>0(v&Xj*{feI>gg1cH$GkKVkS{>UJ^J*x(2TqPIl-tmv;Byyhxz=q0?iHq4ux z!UdrDu;#7KlX-C5!M%g1K(^Xn`SzB28vBmV>*Z2_qsFHC(gT#~uyt25+P6dW66!H>MW?@|URk)7pgou~I3`)tM-h?BGNQ3o=3&P2vl z9X5de2i<Q9wzZ@jfyu*Wsz4#0=2sFC3hdWcY; z4Uu2itFV!PO&r6`Z#B7Jwua@#Y>;jCFf@=;adcb16cATw(aH_kwqtlP#jXBhyIRj6 z3`+g5n}FmzhBf57zfgvVvns8CMbQ3@6A`kLIDP+593~{3Xtyf#rc4b1mA}XP-T4Z* z7duzU6AsmSSaMN{udJYb1E6kjN5>ME41R2fEF9pk$PEAtadRSYno#`IOD#mi#ETY6 z@|a~Z6s&uS!E;au^8)9*Cy-p$p`Dlfg!0!`#eOT!YuAgl2OJz6knJ7?{Y;bmXzP?7 zlQ-}78DyEgaq`@m(R$k15+faWTzgZy#O%u+$@Dc=)-0oVR{2U*w@>t4mTSb)CMp#q zj(={Y{#_fWnst-&>m_W8<8gdS5YoW!HHTJaGjdzRQGqq+d4_D| z^e8S|3tc&fZ$L&Z#TbzYE0DZ@7z^29Ohwcp{F zp?{r+e}8_bTT)4h@7RLY%ke2+U6al>!q_wSBYPS%77{eKc{`u3YGA{B&1oI8B z(Wjdd^;5u5DNv8UTTyEU(sjNR_IKXF;cylb6lgx=X2P-t{FfoAimm)hEA-eu6ctljT7f0*s-1fjX!64$4QlULKP*;izzq?t!8w&1aEc0eM z%)>tHs7l3E_QywJQ-@#fBZvy!6Q01@xG};lOu^8AaG0pg$lU7Jb2n#MiiZZ6KQ!EfacH8MZ zhtnSX*alKEz11{AD9X~jnw}t_In~L(^*Zi9UStFj$af>(OgIJ}P*h&vd_GL0y=mXT z3x$RV4}XNc&>-bWc*6S{AASWNASB{^oRuA|eKq7lz`J#l>-wc6UDL0?n%Z(6KX3Ol zw{lrdIVpRJ4>2NEaN1T+uv{&s7n! zdx3uo2|e$R4Uam+K7FivANxrD1I|fuG>l9?{-EBFbW2ievvR=-ruTYJ6o)F4paVxR zvO!jGg(#OAqlS?BWIv*(tSebD*OscPlQXadn)ll20pV?(=L&9P`Z9K&RxH$H2_KDo zeyv`7hl2=L+~OUdHE$c<>j+t(Itw@*x+_Y)h^kqUrvuc?szWh9z%4K}!dO9|K0fbf z!0r+8SA>Um#;!~h?gRy6D7)^EBpu`9B=hJi#5d_GcP(grZv3`{A6;2)9c~Ov_V$=tk?OJ`bMPc54_jW!WgP0*x(>%QMemu#RsM2yMmbl(z?6l9ZPVKJ6gRa0& zVCahlbZrG9%E8<<#6Kt9I=#{XVs}M%4@ck`LuxB?e!B6Nw;>hFIl`eX#!o6wVhc99p3_6R zrT(<63Qo<=&NAV!va%-O!PAz>AIPVG?nSr&*z19<$vyt^L^`X?VPAn z?IClHsJeDFNgM_nC5@x(xK6cnkV$gWC;L57hC;dV+0fYoe&CC8R>SBPpqh&B=AK*^oHibffm&ckt0=fh*2s zO&g#TH*5OKsF`0pTqKkC&TD^L3!edYCh!0H>YFV-35;B#>>a5@;l>0vXTg}QT5PL0 zPtbUOQL?BA)wI7x|BLnaIPn^S<9}uo<=*Xsa>3^*Te%_O$F-RJN|HhatUi?<>d&|x zD>FZsp4W4x^$sd#6Y6GV~zvV}&r zSvBAq$A~LP;G?nD^k+CSu%{0b!_hTlunT!`ra*jM*Y3fxE&u!wNg{sJ3nDaF8P?1tznS-r$qnM3ABm zQ~9Qc5h3d-94%ijD~U@L6LFk=m+t#p(WUvUXi2uqgd_@+R88L)#PXAZr#EF;Z%^?` z!)>XWt9*1{kCxxl(tTvtvsWfhbpoNK#49hY9q7Pa;I3_~$gP!KuijIhwY5jTp@210 zWa_3p7Bi=taf+umHG6EPM7$cAZtZt#yiz$p{b?%^GDENNEhdrvjynQ9J6yHrG%pDB zzNTUR5WmyXFm!l0_P!J~%y?jIs@B=aF!6di^d=PuV*$bU5mX&N7~NXFZOvIc%Fxv^ zQns_R8#`;iyD`$GI6?g@F1_P_#T9_3K;40Nx;k|KeNQ?V1z1nJ+*F|FbVTr!6*LlQ zHJx2}Nl@~)c>fylIJZ~}8z(2HlsE?V0gDAsnGi`y&d$MMs62c8lP{@{j!E7B6_^x z(=iwH8fxC)k%5IfUae4qV+D#Lx&PaHuKDwtCXrVEPkp}_O$X}CTxU^w#8aFCVaShC ztIiY3J)$Ca{HOSu*e^OmNV1vq4kRWce^Cedkkf2nD|-y_(oki6{G%R3G-UiGc49$+ zU9TDb*l<|h17rJ>JYC$4%VmaDqfhrMT@C0jC>=z{CaRONgKtr^^yyC9YQVPq9g<(;+%tt-9Jb<)jY z+H!H+R?Rjeg8GsyH_G|fWn(FsbKM@wBH{)eEFgeyn=nr&QU_wdI>t{V%Jr5ygBQh8 z?I$(lk*Mwo7H?+vxZDvVOH2MZ(vYH_*Em>52npD*v$K~A9^#TmQIfC*|D|ZXIh`ln znhiMJ{UHRr9&R51!CUev{L5i#_DTtkyI(&_ALx_g>7DW7pcYQd#obRv{~XlTN? z`u5@EEm$v0K|dsFVmf3RYm5*`y~bE-Wq%#6^a);jub3Xl9DPE7Hn% zd1b`~>Gzb!9ISNtYv#RBh8x(cu^1vAVgZ{0?!PLfN>-5q20yX#RywR&yB4(@%c(}b zeycR_s~e=rF-6$lCv>E?iPC#vdvdYU8ls(<^l&bH`UI!tTc z2i3;S@=R>sDh<)5{Ly2>p2V9ft~@x=pl9cJt&iY#GZ{6)ln6?a?ymq09u* z5t21>UCiqe@_1r5Ms&m?F$Hp|nD=0Ula&Qmg-HJ~dbKo6B!or_6F-A$$=D0(P}V%C z9`qa?GoOm5BsVblWAODuXknX@kKZhWh`5r@8T4tBo?7wldgN|bC>F}Y6pr(SS==2l zO#?hLNs+pq-|Idi=K2iXWC=b|i<|U$2Yf<-`Z9sGWqg>jfA3Ry+)p996#7Ib zXmGZLG{KYa+}w5`kTIp#lkDCR3;6`odi#5mz!ja-@aQIim1&>fd5--#(M#4{5U(<=sK=fjnUqm^B zKK$})lhv7ylH7nXmcG3T0|f91MR}_nV>xnF&sZlb)mW!XbaJNLl)clR6uaKwkN5LQ z(f`EynX63+8m8rl#6AFhUTrejwm{f68BQ&wJ&I-@hB)bHKESyT{XXhigVph{>;M~w zS^g0wBj9p!nkf)LgHC$O;NJ0h41>}Wi;iAwi5dO;^s(bcX##j+IK{%+?mo3DNV?&E z2L-5Be3Cw^>4vaRY}pAIkxgFW7tm<`$YYo#_vV618_O?p z8}6OAO%zL;ziWGM-9}QIA7^ zr`>`Yx_`Ugp({yUwv^_Lf>R_{|#R0l^y;`Fa>++uyku zz#@PI)f6SJq-Ux8hfN3s^nW^Bt3G)Y?bY*_PO{V#Wj`Q9(UzjQ73vdb((Zq&?k=ln z44;*9D^PD>U4o-hC;!G(BcLN#Z~`)xnlM~n)*?&XP9k8C8vwhYh4bLq$FJC^?(S|} zWdm07^)w%Wc0fG)amDQ6w!Xtd!uYo+b7X>)qksrM{f89Or3Uhj6bb z^^xg``IxD_nSr0Ug7~#xvO^pThH(zC4nRMjf30UXtD8oUCXf6EIpz13DE+y`c)n##$d=8y|7t-c#x}X+ z_qd2_@?)e1Ja}C;hkxS`$Eqb{1O)fs{OT|kz4zqxoEeVti94_mu~k60YvqeCn~P3o z6%2H>Wrv=(WHQt_IcMm>+=30;g+v^t-uG%}Xk9%>9Q*d$nCQ)$;#?%%oR47q)V`$r zGts>QiLeaE+qsHTEvW``;N~D7@4(yCQT9qSy*)&V4|YZCy6`kkbPv_jpZaP^YoKm7 zj|NZ*Rz6hY;t+{jnc%{F*A?msvTT=&oY409z134fvLoi!#l@8utGHJo!7~N((%`_6 z%D+mF*+&4v%r}2{m~wKzju2 z)lQ2i$4@>I$*~WEmYd{N-QM2vpFJ#lDdT0MV!bwQuq8KP!4gK+3yn$-wN1w-fYf;Tuw z$$-CT1eyWZw!+T)I{cgF0Nkcw?i<+@k`_SP&NJ}9#s(-? z3npa=95gK!yb?m)LCnOR65mOwJrbjGdOj#^h;ZMo8D1<=7^o|c{kQmD_v#cT1V(JF zeX_ylLz{YCjp;>Gz%U>u71IuK>ZdUxNMUC-`hHHS$S(mYCA#Cokp3^rrEkA{ao72R5po2~8?y%KrIg+ctgDOh$jSOPo-!vReF^p;W2rAF4%`kS&z+@ztW5 zHKC(^c_LP1I5*KrOFRRdUHY^fTQ3R5s_+JP6tBbY!NXLn;WU#thYn;l_(gi%pIs1& zKOLfB2~_l^{<6SjTcZJ(t6_-B>FZF+I+?m&^0+0&d&;djox^%);+m zu2uEql~hM(Oo$!cY)=#^!E2->!+Zceb@|aJL?SChTE#)3f zaUBF7kcFP(RMq+sApFpZld=i%VsNB4H%2olB zbNjcLXvoCeT!9+w-EpwPcO5(6L}xFZ$}?9=q`cxyWEl@5Bkoi9mX@5X6crWab#v(r zAZZ{Ji)_&M!=97dL)N=@Nj=W@>%JYb{QLpY_{W5#%=qHK`!W|vkg-49N%FlCpdAl8 zQWt*3d2Nh~^h=s*0z2kE4y2^{j69kj{VAfxEjTD<;@63(zcusEH!IPv(fPRutnby> zYz5sDwHb<32?1IwFK^IaVhfoVq2qBE464>SLwvd1P4VO1z#D5_=x(DS?GQO4VOyFP zYeHW%@y&(!%_B>oQ{VRG!1q+O=yhwh7qTY$v0!`Rerdy{X;}O6BzSaD&*8L#DgM6& zAk@13c+5KMlB!aCkba^(A7ysBtqV;9Afa46{GB0bumU4|nrj)?5@ex1NdRe(MhBuY(8rM{G+j=o&&3l)kHlr088n{C8FEc zdiJ6uOW|cc*S<*#A>c80fnB%7TFwzE1kVCvk~rH^;Fj-tleCwtyAm ztr~;cAR9h>t5P8a{;;1y-x0zN*`CCP2i&X?$eo7`T+ZHI33TLKK_Egm2*_uN7U`hJN#m&fn`O}A*@3^_7etdlcLJSTvi>coHy z&l_DE3BC3Fm~ji7z0?LHW0A;QGTCzkZ4J+ODWx)Sbdu0V@Fk!{J|-q+0!Wu^NCIHm zKZmP*dKMPhePPhS6*i^l(UFkT$Cq0=p&D&%ZPM*op|fkicXOYHrstUM*ztH{=o(+# zar}{Ku%HeIp4KHRE6dNkYB40QPqn_2c^&xiW-;q@$X9ZC#PO~H5a6r-SmEC|*esdG zReoatTrxFQCExXbxijKFz3cvRr^A05s?!OKiE>`#c*zCRj{K~7S8*D+6_mY6sp?3M zckf+q$t?e4d9JqCS){Bw1rpA6@7+4|ofa!-rw2-=TSRL2mNGc_%oMI{-WxMxpA2Tj z=$L%c3tHzWK8^c=PJGKPu%P?($MdouCKYQir9MZYWujyoP4?v4_5ka4Fz)xy#T;G+ zO3(Tpe8Yt!aS&jZYPY@>D1+n_nqu;^3L`!v3}1n7gI4g94=ZUgK&roPY8Gg_Bre`m z+T+EDVT&sUKiEUt*6>J-#Js&5#H;Zeu#kLjMF4l$1i&jh|A_7`iT z<3r8Bg%XoYcGF+hZb5X&vEdW9S?+#5>5((ps>{%(zeIh!Q!V#%F(%%8(E3qTG%^>O zE*wf20mZygWRg1FA8~P=S^{a6t2#}7+!l3F01j?8pCG+4%Lqk8o_a>iOVEb;eDOU8mib}nwhBIW2%x(qCe7{n`E+k$TFK=J8swb?t#P^DjjEB83^VXVny;)g zma}wwg0?OWq&tJ&c%&64jy&Um)D7F0P@HUsL5(#v!8iN9?W=!MjdePb0;<(-w~BMd z0S=w}I>ev?+3D*z=-H}@eGnNhf$oXlL(|rV-i0*z9io78yq2dA?k1a-VOHJ}$xM>H zL_V*3^%S_D+HgVpPc?elHw9gM`M0(5UBl>|BN#}3_RW2F4d0t%gZ;*oBx`*Z1eii# z-#f@uF&FDD{AltTUk=k(cd|2WBu0LMvwk=6!}9Cy9YJxhn^e}Fj`r1ZDG_8?&M^1}uk0L#2fi>qr$UWf%wVMgGkx-A!%`R?nLOZp78Jp?}niGG$Ri}_M;cOdj6WS^3VC@GJ4 z+Vypl7}Vu$5y{r?S5AC-Y?O8-^VD|U&iyUIAnp;niY(z9O<(NS;5j0&*v_eVlU;@i zebK7>k(W?0IzkbQcxhw|PRJC`l69L&!BU+LiICu90>vhUg7(0&iYi8~qX~gzjV~u| zI7f^j_>QCu^t|uAGG1$mXafQiaw77Rzgv3sLNT_tMncF-JdKAXeYZ!10D=9C z04^|5%AjlzMTjg7AU_cX-ZB4{THs9&wDvc_&g35~>i982tq|8J>5on73>_e*N8D;R zY-sD;%%^gCih;K89k<`hZ)1ExWd@%+l4`sV_l!p^t3!y_X;Xz4Kx^_EV!Y!!O%j=v zyc00UR*&!Ir8p3mYsb1!()OzFgMmL5_P4{(a>9a4XIqmbC2;O53I9uJ9DH<#F8)cC}eJq7@Mxnxjrnlbiv3Fov;XcE<*Sp zLT^-V{5a`B%{(p)reJOQR&qn$jG_2gNwQ0jN}0{@Hvxo9{vL!(cMB`KVwpzy=0;Um zd~42)+&>ANOL&|*@!sW%Yq$HUA)ko?^<1%DKbsws4MO(e(9g+cvfnGTmT=ASr|Gpp z%p^xS>dXOW_Ud|7COoCf3qm5S@)$ruk~u}5OE#_YT+!^Al19OP+|!|=*=IN1FGB$-MFBK7v;JuYNOUMNf*(m?i-_WEdH3C83JYk;X6S= z#Lxiug-(%-64SeH<>Zdzl@H5Q7;0VBv@omX>^#4fr?|Kqed0zGO2rG}C%6VcfyN(R zghN`zS#2e66Ma&31r|jRuYH_L6+l>i;e@GAkisn;8_|~~Z`rVOKSje}QEaVdFRcmf zpfN{KX|@kVgQlG~1o8?RvgXt1;xntc&(5r6VdP>3Uz^KpU;z{A!Ko-4#OXtsJU9<` z7SK&pzhM>T2F;P->Axa!BTh8Rc7M;7qCG%xkH)TuU)gMQ-}7au3<&bqZMKLRV~Zf0 z77bUtH)^+#*;8XLszUv%&_MI|V+6(J(LtO5m!!~#m%Qd$H)?bm zWlg{tZ59XLISa92d0iN^U=zm+`#W|}gvPyUoJY;azu&+gKVwc!_l^EoKCfxsf}ft- zhsMDZ(N#MUciz$QIr*&UmF}XL>gN zeyG{;u*kf|VO9HigEhfb_{qHE(p<$8DR-(N-Ne4xGO>B4`InUo9ZBM@-=g<+i9d{( ze~o%sF{cg~%~~VHd`@jD+1KT4W%-utQb|jy6?Q2eLdQX_6(twHG)MiVAGgOdjIl=o z*jI&?y#A2b|Ez!iv09MrLgV(3oBy;6?;zV*zDs$I6Ga{}1+-Dh0#0u+KzEG8#>Rf7 zvh)8HnLco^c0R=7UB!JwLZbT$^S;x0jrHza4GVDOj+*u=u$Et1L!nSa2~x4w5B`Yj zMztOGGSU#CEIMhg+56qlSE1ZFX8%K21bUF$X^Cu>laHUg?Z==zo?qib-VE`6E(r(O z!yevq`Xoh`BiQg=!brpP6o5<*_2^mT8^fPV!n)F9QaXcSyHM4x`!3&aI{CHzE8gWV6j1++WYOBaDo4qRVRSj)5}_VO~J=st6^M=J-$#py*j6oVXxlpY*;XUlE}2Ymh?zG&zUMwt0q;5A}8hpKFDiI2xk$&?$u&`6U8co@a2H)C*%#x1=6Vos*4Qu1fT|4iU zY4globfkgiZjXElKJ0YwLYwQ7*YR@vm`QKW?ml2874C%p6V#v9y&E*gK&b$D&#zHC zz!XIsZ81w*t5TW0!Lmz@`qXUeQVM0$z6RoHjekAo%FcbDm;f!PQwn~BV3DU>m?uL0jM4fT zd5Wcqr5Tl*okFPN9@OCUaIRk@Q=b(}T2K|1$;SNiqO9Jyy4LHMpF!AFSj1rOvcVO{ zkPnL?F-sP%fcJX0)S%76m(P{v&ckmGM= zwZF<(C4F1Y6otb751##>=2;p5?1`-a3NcMjj{Y;itB(FHqIHd=oz;VJ?)}5gfZSI@ zE~H}60nP?M;J&^a`x9%{+~B=)9}sUQ?i8&MjH793Sy(JgAWpeQ0$JBUC3$9^E_z9Z zrcT_}^n>8&57WgutWNWFe89zR5`&nl*^0AwgGDlj*7;=v*?!Cw5wf%wQreX9e;moA z*WE`*smAVnIC^{kKZb+*{MeW3L~>UD0sI2OJ%#O+sfPg+zxmCUbkhd}vAN;nWi98S zt4j~HYh!x+JF-cx6dh(FOUkVjf>{efcQE##>6uUBzCwG_Oiww#5F4|H_XO}Bsj&CV;ZxUk_z%5@!M!8jz^+Vby3RW+gSt-E0=pu zXt`z#-B1_KqGPC&d577JtQwnn7VX2jEh}+m@$7Cvkc+aDV%IZ?MT;M`^I)-YqF8!T zpzGI;#02tw_G-RL)fBS@c8c)S6lF3DI2h#O70elETdF46u~2>&9Qq2GX8*8T(PtA@rM-~Tm1fOcg?j-RHs zUq^*nnU2}uVw3?K0u=U79K5`}(2otDTx@4YmX?a-yWycqn}IF*zyADR2mL=m0aaOI zoDbRI?6U`lNAf8a*FJct{!ZwbNP*4ZF&6t<3zF^o4`t?4GhdQ!=l)FdORsal^S?+y z0SFDOZ|SBaSVzEVzCz4DiAn9NVP!F2%mXe-k+|1!0>B0oASa(3|8j#+;C^rNA_deSqkOj+nZwPa}W1S=MytEDiw~W zu6+lz7tzn!!=6}m)1|jq(Yp|m+E0FXr?HP;&nQT;hi`@T7NLkgYH$5Bz-h5is(2ko zersXc)ZCsI@W;xjV+I%Qc2Y)DBj6xNRE)1F6THSA%fauLRPhV`G|U4*${E?hO89G$Q$#mQFB zlE1w!|Mn)HMNpDp5!XwnHsh~`bmN0qEEZ1`^k1~ROLzPEn-Ejmo)l~DYp68oh>73e zH`xcAP4@*!`yLb1fq}~di#CON7TK5NC&m);bQ`LGn0Sa}p`*4hz4lMJ&1_bHA6Z^w znW*5~E;^hBEYOEETSR+pKjDqNNV3XFx9hC^B3Tsm9J0vTe!sgoEUlkA9ZgP6<4=yjaK1K5g@oGB!#=X6SQ!b;1b*&f(9o@2oPKf55b)z z$cGao1a}DTPQhKtKIiQHpL^T?e%C`i6tu=-&N0Uv(t97Hg-ZlPFG-vi3#R}e1_DMp zV$>OV592OhrqvmswW6TyUExfz{nlPw?=gut=PU5GAZ3EF?HFlhN)VWknU5~M!sHJ$WC0`v+*v3x#g;!3#%&G0xT42ri=@!$r|?n1y$Y?*)emmZzm_)a zAk@V8r=?iS>SiR!s}l_CUkg)3zgjuB2hqUYwLNZxq|v%xhG2qlKui*ORd@KwV;PtW zSB)iHl@__+DTmKABPO-!Zxdf?l-twf4HRcxk?L7KHC`bK#9`JNZjg!zO+Ag)dM27j zjgCv);*_#WaXV0kq@6X+25(s7IBRpD2LIF9{O1?tq%x}weF4sb%>QcpXk7q|G`8W)S30zt z8Q*GHqGR~9h4gsDtxeX$^_ekwqn*Tjl)b4%i~ZGdYTZy`J|fE#+V%kz6Ta>bLKi2E zU@}k{UdNs0a(#s^BgGut< z4hnwf&1j?%RMaG?Dl!R+s{OY0eQb9eJvRUl)o5o6)N0E)Ah?K!CE6bUdR1?^Cnf~R z`gX-JW!R*TiU#|(a}&eP?Nesk>ZJ`Lr*{T|g|AcX`Ll_6LSSTfv-5N7ABZv@^5@C}Q`!ZfOml9G9XmbeUB_8_thHnm?cEPO|h3Lw&|%%yWE z8SjU8FqIs~ndrsdNltbDSSRb`JsOYLmGqp3GmhxLc?%_+TJrK<;R^qIdKVg{ZkStweMmam(R+Yn{qlp96f>fzZ=c}uf)JlhQ~v`$s%pC|0!pcNeW$$j-20>z z33&lr6ZMN$ztFD3@4n{aNyRh`H63yH3^215|I?FFA25k`XgV}1`i7%CW!JrH)xvf- zuB9G;>C6Rjq$ZSFapr-i2F=jk^+*O{}4{in>?5PDKO4G^)*sF;!~EEOWbrH_IDIG%s`WRc)ZbY?8MI=PH-$18wSMQSd!iK-XwKpzQ9 zl_d6j93ZKT`xYQ|P#hgjr1LpGgDZB=Fn@7G$iuoB4YIMxpDCyIN!lHV9va&4C&YNS zn%S})oERuB8eyhGV7dGy9gP>vqj5x0UJYcaDhk*>$39$Di|)}cSoKtqr!piZqt4^w z4I0)n8uu4I&94YMaj`D9X4QoG#A7bU{ z4H*Q6{>82dG32tw{~8cMmixHN1h-5YjGLVPc<~lQ0OQ7LNc;@#623)I(Ql6-ejb7W z&q9GB)yz~9hY`R%@+6Q*!NVEE&B=UG)Wse0YOvq$QnEGJ`9wcF;ZjpK_X0C)=ai(j zbEo?4vJ`KUOuX8>7JlWJsM#;`!baox#M0oiH8(CLZw{T_;tN(%0mNzt{tkbz&xfG9 z{@}vp&O7SI$5Qd&Z+16g)sqqp6~q+{0Fy_B0-rL!uxvKl92bFX*%oj|J@W{T{=o$7 z*ON$m-NS`#A$bs!f-DI+q1w95=z3?<&RpzWVyX}91=upLD+ruohd88-(FN$)!b%@) zVs&K}+-^n$vc!nwxnF*6(nCZ&iOOAt0Thfu^qZ`*sl>CljPKKGs}0$%jg) zr58OAvFA-{U}UdakeB&aWQ-ZMzITvu4P*#`dWxQdd24u3q;QhI)SQ>7 z^JACYK=vHUac5;oNMEqdw|?031m<>x66;n8d6xnR?(vtN6da@g&_k)PWaM>(v5OJc zSQZ8^{9`dUhlk| zu$aSjp^KeKwD`Z;d=}#{tc->bukdhpMRAa<#H%TUR98ZU&+KmX&bE$qr}V)~8`4cY zWn2nj$^VbbvNZF5*`U4TUvwqVTL>19{!*UJ6CLC3hFe$(*dDJ@>c^+5rI8i@bBgG@ z?;pHO;p-`Y*hQ+a7Vv#bV|*ekOqToa75cL4Q>GniQHO33MBj< zL^moKqh)vZr6VQDDJ4HX*Afnl1&iQPa8qNS@=%I7(TkgBcZrgGEM_Aw)SnSJY7!g_#O*CR9``p{MYPLip`Yj2OXv!rC<)UzR zV6+%Inxq2nOmR(<{T`H@fa9H~MIrmub{(3EfUF6ixp5tgeBV$c#6%5W zohovaw_mQHnTa-_p&SVOn!y>}dX=PDv&@KHpe#!?fkCfA7eWv+A1)XRZ#r^0q%m9C zj?dHp(nTEER5W#6C$2oeVNOx12KGy6_TqCK0i}?DV56(_7`llYy2iHe2^!s%H#omQ z`APPWyo2>t-0NDqwgfLW4;qlve$wA?}I8QieZ*ZU6NsDXeHTG8vm6D{%36VKiBNP4f5T)cR**006@1Cf(S^( zJ~A!=>?HvQRz?%h7`tZSKL(}HIgRDW)$~?2jV)rM=E2w#Sc6EwZE=2r-)S5$^OAnu z7NG3#knWuLT@~^f`)kzU8S7#8#!wNet~yV#Pn7L{rs2&3A%T){lFk_`G9C`TCkl2f z(JbAC1U;5s?RXz`_9r|pBUZ)&a6$+vW74li{9qv@AO;XO?8T}HP?~og$)63SurpLh z-PIJU$N^bbFLjP3!U<51`x+via6?%HQ|LA71Ybly+vRbMY4K0|eNtDWM;z&xe^OTO zgV0e;o0+!}V9p0h5MOsT$(#6Y(VIgX!5#q`HFU_&i!jc~x6}$t#@30Vu=x&)ZuY## z7aHbmhbv_P3I2_j#Q1Cfs|K3@oic0g$2QYyKy%cW5on$SFz%;-uKFw_j z;X|PaloL;OIzp33y`?C-$=m-yi?xo+k!E#CY(e!T5GWC54DHe`l!lTIjZyIB&*8?C ztG7h-;#9=|961KJ+%Sh~}Ks`N`` zs|!L}@{k!DTyk!rw^Ep{ciq`90X+Viff?Z5iUFv|c#JW+{%XBbFw{|qbkqXK*Q5}^ z;d{!!c#s^qatrlI941jC5-ht$tWmc$sTB`N{^#2n2|A%>HWatrR0-6=Hir-6qv3y^z7Y>Eu7G_VL zlW{POnYha%GeQd|;v+5!T|99lkVYP}Kg%qriJc z<~&fY7)Kh5P!%~XvX&G`G=zMqu+JTv83<(Y*{R5jpW;iD}-T=Cw zs4ug5UTZ!54wXPIsthAX+g{K_be6e#F81%Sy8P}S6eBKlx`7jO}J_T6kVd~SOX*2 z2fCnZae&GGHjzPNr5c&X>O_&8$C}>rkU!}tf^;<~7k&qRN{5$Qogo;iAD-=iB$KQYAE=OKkt%|Z_l~Q z0Ja6V;dw}j2I}Dm`N=PK^Z6JCpS`gEh^eniG=asky>NFwrGILh-+e_FpS?OV>%h;K zFWnJ|gU$oJKIUFO5}&E=|GbmZimT!>9Zf9J{M1Z!ueyV@@0?2LeI$x<{)zjC&l(OR zhh}pW#AAp*O?YXjfJT*7E;H}SD~+cW`AKW>ueI)9sh&mpNk@atB<&vr0gX;k1$K1O zC?4qpiJ_lGh%txyh0@DkZKvksE6JN#fiGrBLOY+Zf6?7e9I%*g$>LKpttpZ8) zvKxRsYsoNh8Aa0tse3SW&~p>7u4eck%K+F8ypG!w&t$yX1pE1Hh9Sn%agaLYP=Q8{ zMXO1Q3IO;Z>aWSmprg$rK1}ots}DjgmJ_De{9?^f-qQT`u#uw;t$F{HCa9$PQP|XU znP%|c^lA`JHWlOdl~ImCeca+bnVnlcoFbOkUB9eI&B*dz*bqpAwwPifO_I@Vb2KoL z6QluaB?5@YSY;M=vqS*Bg%5@5HSa``EVF*gA6{HO!HW99ixK^cFrGIq#_JgKyL=9I z-DPqkJ`@@woJMBd;)X2w{T#yTH)*OIbvM}pIew3yP zpYj2C`^9!jx|4lCP+~G0!QjEen{_GEHxak#0FR1cH%E5-(as&dO)IXW2bSU8Ac{Tst3_bwv4o;N0|n?b zVIRUy>^v~axd=BRN=X^yoF2x@r;c;Iyw#BJ zjf?*7kNIzr-n?Z}ESsx*m4c_2~ebE`{SFCqcgc}@FK#2 zf95L763f{&F%=fDDl%3mVqx!wsN0ps>c0xt3JigOp7I}wn@6N3 zyB)BFm1nQTX1`~pXAJX->K!jXxU*#n(ds7p)AOFYIxX-L?f8?xHyX$TL?Vl6B8@#s z6%^YDd3cD>`L5ch4m7UeUGat78bV-F_DqmLb#(_opFaR3gp#g#bV~BJ2v-Q_LhJ28 z>>${$^bFm95RvvT^{Bdz^tcYh@ziPsSBO&8qtq;6%|du9U{1MDwB0Nc0?@b>0RP@K?iDRDdni85>B1oW$Foa~RCgkw2Y z#*Aem2Aotq0$p2ZBwkN3IS12Hxca+~l;Y5L5ceLj!eb=ne2!oUe~4(e2)rJ$ zOgg<3R?GN@X&A;ob$xnDq4Me0&wT+VUSTqso~@ermD>K*G{1(3l;q{BMYKq_(X~o0 zhoXEFwcYZ*@BSNlnQ-$zY{qTWkU8RM=QvjTg4JLNDCn7p4P={(b`Voae;Ve~(^9Eg zFCxzw#0M$91mf`}TKlmaQ5;wM)^+);KY=`ewEAiA0LV{t^xb_D_6Cd%Xix%%*9Fk_ zNzl~6V6eyiwQHhFW$-jWiUGI)iP3_$bhPD25~*`X#V`7G2fHq5gV#1Uwfl{grzn2h zRd@FIhkU>;5UCy842AMGIry9BYIilJPi`;6EgovnFNMpqWm}Fa5Zj{ry!#H*rN-0u zLvAFSrha9D`T|*?V$=-{1iz0{wyDkEf%iTc~#q-ZWer z%~#jh)%hs>kNk(CmY}x=itk!%pyxfW=KQXqe+CN4seY0UBD5J#!klbogG|g`kn*JZ z;awI2=m)YOkAoxW#g2G#33wNZ_H>3mwjBVZsb_c|b1C<7JoFb4mWr~?br`7&*4J2F zeY>zJ`(tc=r?bWsn^Ep)Y20T;OJTc=zkRU6pa*~#m*IrHQ(8|A%r0P`rbdUCLY7b5&KT8e$>EEnEVi&;fFv5?=N$VM&m3|5gQO)h4ANT>z$<6iN2Z*kd8z_v zA`~pyivTsut9>OZ|0`WbN~wyb{UXh6HkUb}5J3kOcUKyjEj7~)wDdch0T?B4|9$1>>|0RKsE z;WGT;TYntcnGs;>yr<(0l1;SsE8U^N$KztTZZkHPR43z}UzmY;j*f4mss~h4or{;j zG9BfnYuZK$Su~d-ebeY=KQ;kG>v{bH)AAY8Sr^}goo`IVL%kHZeY^cv@W!h^K7^z1 z=99DIUUUGeX`XYDol~{cwgq(D1Av*8{*iAEU#OU7YZ`Q(nx$5g5Fe!4<5ZI@`QeFWuy4D8gi-$M_CD8JgL#^1eV zbOB~Uw7#ZbHEBwFS~*)8K**}s@$PBW;_y7c4p!N_LYuslCIGY{DsO!=U(4$JKy^aS zL(9XEK$`<^;#1oF{A^m0CzZlQ#s=m={kRM8zSGQq#&Ah!ea6(RY)Ayvfm~0O*d_tI z#v9(cq;h;L#_l+!bRlIunU-a5xoLjwt3RUy$@y(C zrFoV^7;*lIOvshXx}qu2>@60#1hn_jx88yCv@wC^CMN3*=~P7f+ygVh5U`wa5XX-_wz76yd{9FhU**>s8B0S(GdXY|Slv$)tA@Wl(V2Q(jV$nb_=gkK zONUR;JNNJE#KpPK8vPs_Go7sVTH;Ax(v}Oog!qw;IMES0T`;b+6y{_}q(_7e|%m$G6;RN8D?bozGb8NUSXA(dE zOJz8_2Y~t1c`DNqqA}WFTX{Uts!|H6E7_Th{jH<72`XZ-8!!jD%Ft|)z?|3k5)j&N z0NngH(A@cvQq!6L(lkgqvFkpQTJ$akJRMd#_J(l{7KRQ(;XuB~%F~m&?OG?hA^Kni zSI8R&A1jI*O3lg^SAOHN*IF#5xT4{@{>tkbxDaKGz>UD2s05>}8dL_ws5BHW)qQ(u z1Bs{)Kxg4l^#82?sK8tV>G#l_MMfpiQV($k$s{({n7Xle1cQ5F_&Zty43SifH?oxC zxGWrzzuDJ9I+1#aOgg{jZ3xlQRMZT2GWCn*bxVyD@4J(b&4Jj!L0?l&vB>;r8R+KF zHwh+2!KWa;)gV$s++Q3eeA@au+Cn9lj`)ei(|1yK13v?KychX0JIzG_t)Ixt$o_Fg zPh4Yn>?_oaALBa^WH?VkKJjkE3+pfRD$M)i*6<-nZd+n*`HlH7@p+1a0p}FAJ#W5o zMh&|8q^*r5k|U<5T&O?(So_8x-uu$Jfy${~^7aYh!w$)_X-w~rVAf8FO^`>z| zrSkbRT10re`v+cU=Oc0dIed}n>|0qu1R@Rn@=A>>rMIWYWa91`eR@a!xwiG+`xIa` zch+@Smp|Flo``z-FV`h2BcZjKm7K{wraYtz+_tJewg(r#LZb}$4LFNQ7Di-!?h^x> z@sJDKrRsvFWUFb4r(cnCh=eE(dPty2fV%179xlUx2JCNuj9!_|)EXfDDL8Boe>J$f ze1bxbm9Cqi``tWztf<^) z#BEcw8zzJ{4h}>=|B7%it=DEQYF_?bVgKx*kFO|K8c`g}e*_!2FkF!De^}b`h$3+& z59~dOCYUMEz_b0v+2+uLb0Rq>34agLfCP(t1GjO!_!ZSm8`K=7LoPXsnujJ3F811p z(-id7RT=AhF{Mzp%lkjcQQ(i0dj2&5e?YJAlTJVA-0G=`87e`76(r3Jo_mZz>N)OV z?7pX<@$@?!p3))`tTB8Io-o#N>=F1PG`N4psyR^H|8)Eb4C|uk$+VGpdZz5^$p8^h zQ+ZpMFiE#b_oJ35zmdT*r{J7>`{tTuTD9j}bQ6Jh2+tdIP>&HfT#BWh!X7%A67MAp zEtsGrR%e@r0ifCK7h7>B_q;|DQ$wigyLAWVXQD+F?e4KaT@|1QlwP@l)j9rb7jAW5&0t!Z{ zs!AnP*vI51RJ@b7V5gPa5K@IK^aesL0T4D$LXMb)0CK}P^89k7I1f!ez1!wh4k;|^ zk^gnNilfJRLhB4~B>QjfxhSFl_?0E93P2SGOM!LlFo+$b;6Cyg#hT|}Wsn@WB=7Mz zcOPot&z&XGZ=!zo6`wX+P0C`##s?w)W5l;gn1?F?3Hecm{ekTBMjgfc zG5j@ABeDdsY0E%xVGiLInuMZ{TjP(HI!v3#M%}+QPg_Vx^;UyJDrpCg4!$4D1m{?gfyF9eB`K*}F1ejXQMsao3LoZJNdb??HPTW4)QYn$bT<&& z)=D2`ys4yR`dSb)0b%+P_cCDP8sL@xp>6zmrhER^-u(SR;Lvt`2z-#ABKoZ~88E-# zbuEy~7i%RHY5=pS8%(rX3tOMb0cypUM3z5MeVO4CjCXPF@l!z|PXR@z+V`6a(^=V1 zt5b=^y~)Z!8atV)Frh^=xS7aO`(p-8jU;C=Ad0=mmRqAO1FKt^17=gZFUZ%wWMNF9 z!&pJGf3B~2I-3t@T?q9GwrZoI3J|PqD4dk+q3sCJs{Ur51Gylh;_Y&Pd-6L-^pZVh zv;;--D9Y7BpawSp?jX|PH@}tyjFZmfH~+*W7zD&Bu>RzlESsVi;k%gYv95B<4H}BfBlQl`L=#V!_&!;bn#Td6 zSrL1m&G)m786qzW*+ZORg0dM0AWCkH6RR)tW*>>AX19EER0Ms7U!htleYbfF^*U?G zu%-B`hy%Dm;ITiyYyPhrle6~7i5)D1P&U5v{)1#nwT*q}+e1PCaE_e;MropaMS&4e zU(S8P8bm+C>1e92t<`A^wo>=$`DHIdRy}z}7JDjawW9U>L5KRQr`MNjqk7pH-?!$_ zJD^+TnHMB!&kuSE4an013yhltOv7rjJ3QjC6l*xL$C;?QiO$hMpd^-W%4KLU00S|t zGdTd)Otsb&M%_vEFPRwlG5Mx!lTLXDh4(m?aa&aHfNqpNbK;~QT14@B<`=B(E=)f@Z?^>meE^WTYqHul`V+ z!|szn8?@jZ=_0rIvKyMUt9Aiid4(>A5##+9rbDdtzV~E- zlCkXOe@@cypOf@27iYe_$VBnyg!wM~YkPo%cAEX2YI*4?k-8pq27Al(&@5H<#=M3w zZ96Z(4Ww2!%P0G;@JkQugBkeI)-Okeu6UZ)Yco~d1}pZ`*5*lRaNU(k6UFwbkb0Td z@1p#^*Vr6L5?{(cu0(PT7PxR}_UqFT`WJ|@W>sfdF1)P6iN@mIL*1!&Tz%m)nGpCR zT`K|<3W{=(AUG#rdeKc+)HoTWQTi=xjde!My1y>&K|b>sI*3C5{4)w6E}0p1P&<*DX|q2u(hNShSd~-6JqkDI+-(*Gh7lWn(+*sx+Sa! zpWInrR17MY2KF>RL(9O11p|7xa!4GU<09V4bh)Yw(FmWQY>s@w>eruY2+I5tGA+_J z^sWYUut!kJbufy-sjBj2+x6~Doj&}uOao+69~+j6^lgZbY9($8=VQK+!EA5vVuL*ad4Ud4x8`9AjR|CG5|Z!BS~uG*+SN>gj&!0Dap* z0rl7nRa{7gV#bI|Qh_Kejgn%8(+t*HwUnfqYlE)*QU#WJ2s`Qz(XSy`BTCBmnrg|2Li*w0mUNV`Q;5zXMnG#zoN(pW*|o^vXsiD za>spPO%RrZ0v=7}q9h+^@6Pf1jKg~M88$I;;{1ele&tOo^U0>5;)UChTB@8_m+@N` zjc8IkpPTa?)hi`T()r?WuZ_Jez_fYZCnRCfdY9{c(@O}1%AaxpuM-`NRc>2Jr; z%1sP~x3*^9yR zvn5jcs|bwNr6g9*94|u^ z`Zk(Zqb%+#Y4+dF9q9JtttZ^0Ycl)7nt*BG)gDgIHY$UI`x6Kmx(7A*9U;GDF5xC%!v`9C84)&tx~?C?;`(Nhu#JE1@)q3cL?~|l>a6Bb8^5^ z4gsQP*sLTGyDdq2sEW|R2VTSnUX=_bh?OgIwA=QjfrgcF7)I53?azf;CMzQ;dWxxm zdN7&T3(m=s`&yPUG&TT z_pvbN^>=lMyJ0?j)D;cd+`!FFygEJIU0D-2z&Bo$HWnUV6trK?cN4R15$+%BB{>Ra zEj1R{_GUX=)f^SvUZ{5n7Tm6MpL`TP9Cb}|)_1HhHZhIFqq;4L^-^Ko&VJ4x_>Ory z`{QW;-RqFw=-A(4fD{)5e9%ieKLQ5O$#;v-&rL4uBZys)MpcFXRIO-3`0 zjFV2e+vh`k>5(5u(DyP7x4$;UT)8jsC8XCLI_wzE-dLW5R&T%ZMR@b~Uz~Msp4lqQ zZfCE4r&=6aQ@wjQ{~9IO00>ZFMr&NzZ|x#v{r{E-*0LEkvr`VO1IzQW-L2cM-#mKx z9%w9+>^9r#F1wB145~bOr@4|CauhyY&(71T9uW!g>t>_zP)Ei=P5p`Yg({3-VNZSGUu zhc51__JUVrzXRGcv%8qKeHOO)*={Flvfa5BL}vLvxF;4n-D>O~QnD1aHut~WypZjX zxKD}ok7F)1PIkI|VlW!`qFEo>sjg~j+4;}kZQD%CW&fL*mJc5H4jEN?xLo+nZWH6k zm$jI&vaucdiro;SsH1-}onH(Thxd3EavvH6MqPKt^4)!KA64aFro(F%mcQ_Z#~#^< zv(*^eKi;v;O_yZ5Pb{z~j52O#Z#wm{G-kWEwl()V=Mb)YF9d9Rc&mnIc~vpL?9*Jx zPB9!6RQd0>k=+_<%4PqTcL#jA>xpC-Tqzx2&$;FL^M>JQRBqU&(Oq0t?_TX_f5lx; zO*_6GY_a7U_7k4%e!5ui;3jzBcI8lh>~=W0d~*#9yG4iti*WE*73=T8SZaLY)M>=R z`qJcV@tmO{`>x*oB=o*xJAn4XjmF4U?5U-%XN6WpO+agNb94WV%%*tq2Z6u!oPU4R zV%39@(c%u5yB>oLixn37%5*y!@( zAz?H7%+zVFp*q`rZ2_oCjtW|(`jpHi557WO+0}KD9u(Zp$?}Q-EX0I=TLd^<4{YZ5 zzP^FUHyW(lO7GiS|K>CQzxmJ(M=0DxZ~K^HYLeY)-FR(ors_`8O zJLE$`+IVShYjdJ5L+{Z`w!>*tlgvMN4cgtWFMe!?VaN<>^pDqn21^l^z@X+Bl-dkvWxE>O1JGlx70*M zwMPX!c<0=ekwu|y+1NV7{mi<{9MS@`mI>>~ev&ArKGJN44Ua`%|8T(Gv`O1$JgO4W z_1$NtX6}CbvjNES5KHfGSZg>JcRCoq1F~!T^@yAn(I0kqeN6k04qJC;iN^bP{M>wk z>5Zyod3O&Jr#IVWnj#4qcTDT@2alGQb|srZy3CA?kG4w=nVTj}8=J3y*t637P19sn z)Ccs7gBk1GrN@Ms**7NWrUk~V3e^ogH30K{Ipe^$hCIIiv9vu!WE1$$hA#oU*Ok}Y zqyL$2`PEo*()6k3L*DypL^av=a?9+|9*>Ep<)0-ewh|WA6~D~uY~Z~+zx8e(wJ0S; z6e?5v-QVv*-*vYfHq1!w9`n<|isyT_mEmz!hE;&Z^+;VdMbSO5!l2TtO6@JL?p)9G z1L5+|%a7mRNN2evU{-G&5xi1$^&`WoebK-~BV*?PQ~Dbk>P5^~8{}XYoKy6BgVhgM zYX>~tal}4v%#@qZmcOprMJ)(=-6#+Is^0YMlEG#e+(6aT<DzDNbifVq=TvGpO5S)(2L#?PH^l9~sNT;_3)Tnz z{B$RF^n`sma|0zeQJkyan1R!5)RlC{WnB%{9ipb8oLWYZ84x-Ad#*)iM95@RsxKBL zh{Pw2sovD@i%NFcn$Iu}UlqV@n*hOolkp8(P@y_4Zs<1?Jp`T2{)ROMq=$5}{v z`QXIyi-r)(dSX;r0uPwsoBnKKOr}1`uP86#GBla3zjm8-$6+>%%ti8rR+*y^8_x61 zUuGDfH}vyZ+m&3l2kIgkmIa%frYp)nJUCair4l+V`tMH9-t9l4ydo)h-`%sT;obO* zuE=iUeFU6LiY^8sRX&0_F5an8^ybNW9zU703@h@{5>Yr*G-Y2qA0n$@ks=$Nxj!hf z5HruLPhf~kSDw4kWU8MWoEgsJU{DU>eeR8TPE>$FSCr`i%7xX)bH!D45o4)i;t~{` zV?>S>Jz&+FOQkyfD5d}BQ~dd;-=WtR2c!u2uH@ermUv31i2rsUQd}eWlUv~y zW!KZ=;(XWAm%^YUDAdmpm@lGd(k4MYN;DI;nZhml-rHbo(38E^an#e%-7Wc4D1}ia zdg9|j)@F5Wi-V*87-W(DsQM0s-*99rk(bkX7u2$6;j`t*-R`g3c66xW{kgu>F>l0w zZaD&FqU~FtXUXW^;Sn>VJVqI-emGpo=%WLY)%lh1@NL{zMuZ}qMzgK<69v2;|NI#h zi(1nK<)^kt>D-jh)fb!!lZQOtrjgg2b!Z1t~Ssu&uWR0bb9BX(`;Y{DUH zWLvlWZCq0=vb)&o`M;DXjz4-UoT#NuyUnm|E%yer9v(g{Y6c8Dk@ce;#PkS=D9yLw za2x)D&?$~7<&r6|1BRvPJ`+f`A7EMsMrf`%ENZNo85`g2>8-Q+G*W4YrBlV?swZ-hT=Ae2seRvX65@2mPZA&3seiG{v z+;7P1jWXIgxeKFhVTXz~&_`}>dnC?y35Ei@9L$zvAr-u8X{Wc>@N3Z8NqZf6qku#d}8;6;ks{vb)7!zQ*!;!rAu7i{4cu>3FHy z^LKa)XO#;L_J^%MPU~48XVBnu9>QP!hfAn3+f0&#emE7(5=t3711yI6=k|(Bepr!{ z1I@t)=f?VAgAml&`CYb<)+oJWhs>L8)6MFW(7c47{);y&uO8isQMKQ*05J zJKNb*?+$4*5VjK!NQqBgsy#B#4#X!M9$i@3wGX-QbFd&j^!#~$ML+@gs6`{%Vr=z! z#XvfH_>H?yjmgjL4(rF9m!99mzf|7r)MPz+H_jHqf{VpYX;vR-B$tX?_SBNDX6P-D z4`%!g9?J+rx_BQ2Shw6hr)ipcm*0%HDc(p;#e?q{->XE=_06$bZ{u?cKZf7TEKBTf zo6qR5>-)WI8=FkJf(C54)b*N5JrJgQ8Lp6ZKpOHp&XkqzCi-MICt!5tZHNxjHtDB6GFfIb&gn;itBlC|%I^*c2*sr}#SQ)Xn{Q??Z6%G(OpH_B^M#xG@W)QI#j*olLL?nHlQq zg~Zq*C$1J<|>8L4lbQy0(wG z(Fb942!lNu-7oEvbj~s>WY)m)Gx7M4Z=e2r+g@LF-&MH zq{mZ@(^Pm`aD$aTA7^6i7fnzHIPst*TOD;0KMXsUi6c_Wpk;i$2K_)s7#qpTIFsBiaDhN`fft42rKvS zgvXaY2aQfh1I4c=&trjdT1S^s8}=@xqbr9|Eidxm8YhGubxod*j%l^#NiI)x9;BFl zjMr=|Wnhg)X;pDq!xj#?2)PlZTi(d}hZHjcO01k4x*hyvj+^L%(8%c#?zKz)LzWZ~ zUAR`LTR41dK>{EB%N@#Y0i<|Tx+^55@atG-wTT6mx)2DVu1HuQ1$X~2tk=@JiCCO? ze1ajgzc=SEwXrB&Cte2(<61?ONZgjeX4o|kSCCMy!oCaZt|{#>$`DNO980Cf2uL4H zNZLOUnEV(iBnZ-F-My5@W_?sp=+Zltge^xELM|F{e~w1gvuoU9Q4u|#U%7}pA3JD4 zMazR~_*yG^M!+o_(y`p$xfCL9*pJ)ur*qRF1$+_Uo0Fs^2?XX@z+K9Fs3@jE!Jt)UqPYmqV z&s?Y2J#5^b-h@6Ntp99zi-suOF4ox@hoek)<2_4+xs17RQ1UIIDKmpiSz-WWg9_JI z)3`(ReeUyn++WWxYF&vZQiyL=zhJ1-9l9OWcs&>^WMNu4u*w9oMWhT^TeZuR13^!E zq@Ms|;A3hRvX#Zv$0(_#bEN75wa%Hv+|pHvhWmUfd&OV}oxlHMwFuL90)T{4S2Zct|cu5z6E{U5Kv`2m2Hr50MY++o4^o~K+c}p+e zuow>F|fS!%ayd~Z)!^hA)fP53%i z3AW}fmPl54lmDS3oijV5k>}<|#%bb$&os;6iLQwA@kS%=<%obM~EVWBQ9L~VE0 zgCQ5t4O{G(F%KC58&&qvjyLv=+^nU}NIhu}pX#GH-&_6e#tAJIb- z5Bcdz>0wPR`s>^+gB+cj2la1NVaEDACsyo*7mUBp)Rxg@|+uR|Ua(=2C}u?sOv zz!hox+WmkIr!ioigV8DK#;)F|XSMo~4vP-|dAqIbwbA)j%wh-dkI@HFv$-kH#0Y70g+Z+>pB|*2F80x>vaz~QKJ~eGwD)~t; zR47VGk7s``r!cpoDS4=$3g1Pl-}d26CO=Vu0_w?jU7wE3!Tw;`=M~*We>$ewH7DaT zVbe1F)LzFcQN2&TQJo%7|=phF2K~y-pNB}HZr|vl_t>>9mmVwxfRCA z(0+P08v*TF9F9xvY!t*+H6>uz4LO;P;F+hwKu04T!?nVqkBU2AGdNtx=^5MGJhBYv7) z?skL(oxM)v*t-j2s0?Cg+AbOrA^EjZDiuu}y){_vbC53jfH4czOS)8g8Q(uaW6)k_ z3a@cTDMOIgQ!I6xVjR6hOp&WEz?_?n>aJG-s7@6T|61Z*+7JBlU$E6X4~rSUR`Ss8 zk~6ABlWWOQq&<_tjW`z@J(7vRRBvV7SFskqY~)$UbXMKazpTxhl_FKjX1&qwL?C(M zB}vikUjNg;7R)z7fu&ybcAt3~h=z&-4o99Wvv~!2)5Mh)VXA)zR`KHVkP>0RR#A_* zOR=8jlY*EVziCLp`NB^;GQ_O`5RMvr?4`dHd?&j4GXRL7yV}|}7th|Hu9GNty)e)6 zR=c4Bb~HQ*OWgr@(+}QA8y(xH6X9@1jU32@0;VH(y9~Na#LLCyrynTT6uppyHn+Jxp-hX(XIfjo!rRQ;X~t0)FIhY^0#UGDrzkCsz~7F z`-{)rJ8M)33pmVlVj#S2aeVeHbc%O__$)_c3S;y#J2GL&BZmQYjukd7*(|GdL}>@v z9roXI<%fW}K^ZN==<(XD$f!61e#Q_KG@gDa=98|O35l=~)Ygdprp*=X;_I(8L|AZq zrbJNSQgLh^6bN`?hiqe`r+rVFGuw#h>_dkr_H(xR(+|B-$kR>`|2pb$z)Qpr=s(Ft zq`Z(pNDdMfD3XadNF1oVJ0jbg6@q}8b=QXKoWt5vj7)gLp#hEZ`NKiUSMMBKO4pK| z?s%k5(UosIypr`?^grc#{r8%s>c-svylC1VW3?NU(h{E0wI(h|=WZ*j#I)*kx#)lt zhk}j3P= ze0;u3v3q%yrAugzG|}K=J$3%MeWOJr?fq=1Y%I*oy=WRXxq4$0K9I}%ebouZS8zYa zJnkW6EiKteHUk)Q@LpDzhuv8G230P%sk&L~-fpXdMW#l!oIaLV%Fa+|%>LwKDM(Hi z0RMBYI3VT`grLGVbzlicz$OPxKwwI1`Yo2eG^=Mo|HgDN&5_h@cA%kF9Oat-Lh#WE zM62?c8MFEa7=?fp8205?jRLqx*OX%`l0Mrh)56|M{9k8sb=N+}WhO6c3Q2K=7me}p z7OTbVn@bt;n4(`yD)PQ_^VI)K&IVtK;>k=hcn$Fj9<{<>R3t_YxgQVPt}$1m+X8ui z`N2P2R60St^tx-l`%oCLu&MP2XDh|QS{E~l5eCbZfnk$AsU0p0z)HwNSr#w{kAgP% z;q4iIyErSrmRbPx@nA&wPn5+Hr?C!!c4$SZ&l-nL3k71bfU9haE8zB8E*kII+NkOo zAebUrDg5_TcH3ZSVl7<;z!v$tyVI9Qbe93^{#C=&24Hg=AN=YT%nY&vAM6EFE9)U8 zi5CpwLZqd&+1ZV^MUIjZ0whj(?%`cU?DwWe{+(U5*R?M>p|36v=IH2#pVPe(T|okx z(uY?;*MFF|262z-r2Numq%?F)Z4d&0QkVQu3Est}!mo;yt?&3Ee$%~4tGMpDpUB}9T~g-v zuBHgyMfnvGx1S6vl>uQ-gwmH>JEuKtguIn_u@hrtA{Qg=RV2AM? zs9~MA-mX%E)H2n`HiM&&wnkBPlb`{iM{mBX`elf-+LWK7yt_o@akGPt^Q2#mHY)iA0vIrqLYDH8PV+Is#?KZGjQ+G2pd@f(^%z2;R;vBk*vn+r8~qr|e`k}{ zkG3q|giUwnj7@?%gm;5~WvBh&?=$rP{l#>MzK4_qa|8;vUN<3M+m0d=IxLW#FBuw? ze{N#@bV6H?L&M2%o5`;I5YPMb!wjXds`Cf0{aRgP)J&NtZ#~TdZwkE~BhDgbPG#?n z^D8#)*HF)xeF6d7M$vc-{G-KJL;mEHL?DLrmwir-=*3Y^()K{QOQhzN5-l~cvni~4 z2cX!TNMIS^%Dyul15@(E1zjm!v5YsGT~eDTC$4*a@UF~tRs(F^NZ;4oghmh-h<_h) znxMbh*v7Qv7?dzq&T~$;d_BzygWn1yyGwbY+%6k8m@JFUg|-l6q9@9{c^qA;s;gN` zgca44WSn$AS*D#VWu^Pk5=l-o2MIm_9ZHU>`M0`|2N$u>_6rT>Lk}Qw;R^bP8k+AY zMmRJ4JZod)c9LwI(&!l{lRzHAQes3`MT=BHK2=sRA_MLo9EPnzo!z2zI;nuE_udetQsmbRjsg$5<)jSxXr=lmMsF#*?b3BbBh(m&nj%GyPDkX{F5Yfp{R zgLqFjTx{q@2klinVKmDP^OmIAADVHLcVLYPtPxZJNyLO!tvt4XaLvOTI2 zi>oq~$oSSf;Q+ppitJR}}Q*s4k)bGXLfr?Jf$ zuk<(o6Yo2x)i1|XyATLnpKUnxZ6xVP>G=iacbyn(Fki8iY;Tjcbz(cC09& zXC&*{^!ubr)x8CEjXdB_&fJf> z;V{aZ@}^%AA5nW}|L*%7+8@_!EGMnI7mE2y1QX&n9<5i%J7%rM^vOKU@1L?PcL`!# zpnjuS=c`o><$Bz?E5%QRmprb~*?oOS>Z%VGhPRZh_kxKXoWE|ihOGbUR20_^bbo@p zjAj+BYlF{wySoGOu>0VElIMrb?LE`{M+V7=qP(jLW02$ z13O$O#%q0&6qdSevv0J(6R5E3wbxen?FB>e_rvFc^$AIlTW+PFFu;Lra`YYlK0UV2 zJXVEp07~M8N;S8Je;+NLzf^fJFWV4+`O0-b`rO-w0MHBtXFV@?++8MfhSG3S4V5qD z`6Wy&ws~3{@;kOwO>okZesMo~I|)}L5^0HH_KPBiY;6Qq9)+hWnMyTmRsyyHWiqNf zzDe-S0XIs`-f~LY*$rQLT;RQ)qjZ!)^mEFD=xsDQ$@%?x#YAtS5^W$oPw(LWAK=3B*TAm%5_F%lD_tUcT7 zhck2JTIZ-zoRs$mi>!yr-yWhibElI8VTV;Ift>4z#Ua*^zy+CyB9bj~eRC0^aBN!| zAh`9JN6wry$RoK)J|7{!uCbhrrWOLHiY>3*CgR0T;b8R8Y$L%p+h|d7&2=wmftb`6 zdE1X^MX4GiW&AbSV)U2LzyQpe#U@r^iF)g6IW}3q1}0dWS(K=4HFq*4j7$F@^s-SD zoS?IH*G$lHeFxuu^mckVeP@?9MJZ%c+S5L6uAZt;S^H-*xb_fMbPJt=4K>H5S)Q;b0KxfG;g(_#FS&msT0s4tGv#R%33dnBZ{+(PS(L(a-O(Ff|Hs%_(lx ziM75ntFa3^jo{oG>hW(+)NJg(}K+3(-zX3p4&ehDG z4|HjHHl{aLN=(Wf%on!SGb?0HE_^3vPvO3~3GpXmv4gg#M>7Q`ffe{?cWr0QYX9?p zen-nn1q|Ebbn1*fE*`gO#%$C3!81rZe{Ok|q{}+mYa;|4*q%+W8@8Px*VWFd)H3+b zanQeu=h{!0q9XRrzmXn?3%dYfqQ<3vN6qXbCLG~+yv1aaf#dhzMd{mPl+ +\[2\] Archives used as-is from [Baek et al.](https://github.com/ku21fan/STR-Fewer-Labels/blob/main/data.md) They are included in the dataset release for convenience. Please refer to their work for more info about the datasets. + +The preprocessed archives are available here: [val + test + most of train](https://drive.google.com/drive/folders/1NYuoi7dfJVgo-zUJogh8UQZgIMpLviOE), [TextOCR + OpenVINO](https://drive.google.com/drive/folders/1D9z_YJVa6f-O0juni-yG5jcwnhvYw-qC) + +The expected filesystem structure is as follows: +``` +data +├── test +│ ├── ArT +│ ├── COCOv1.4 +│ ├── CUTE80 +│ ├── IC13_1015 +│ ├── IC13_1095 # Full IC13 test set. Typically not used for benchmarking but provided here for convenience. +│ ├── IC13_857 +│ ├── IC15_1811 +│ ├── IC15_2077 +│ ├── IIIT5k +│ ├── SVT +│ ├── SVTP +│ └── Uber +├── train +│ ├── real +│ │ ├── ArT +│ │ │ ├── train +│ │ │ └── val +│ │ ├── COCOv2.0 +│ │ │ ├── train +│ │ │ └── val +│ │ ├── LSVT +│ │ │ ├── test +│ │ │ ├── train +│ │ │ └── val +│ │ ├── MLT19 +│ │ │ ├── test +│ │ │ ├── train +│ │ │ └── val +│ │ ├── OpenVINO +│ │ │ ├── train_1 +│ │ │ ├── train_2 +│ │ │ ├── train_5 +│ │ │ ├── train_f +│ │ │ └── validation +│ │ ├── RCTW17 +│ │ │ ├── test +│ │ │ ├── train +│ │ │ └── val +│ │ ├── ReCTS +│ │ │ ├── test +│ │ │ ├── train +│ │ │ └── val +│ │ ├── TextOCR +│ │ │ ├── train +│ │ │ └── val +│ │ └── Uber +│ │ ├── train +│ │ └── val +│ └── synth +│ ├── MJ +│ │ ├── test +│ │ ├── train +│ │ └── val +│ └── ST +└── val + ├── IC13 + ├── IC15 + ├── IIIT5k + └── SVT +``` diff --git a/LICENSE b/LICENSE new file mode 100644 index 00000000..d6456956 --- /dev/null +++ b/LICENSE @@ -0,0 +1,202 @@ + + Apache License + Version 2.0, January 2004 + http://www.apache.org/licenses/ + + TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION + + 1. Definitions. + + "License" shall mean the terms and conditions for use, reproduction, + and distribution as defined by Sections 1 through 9 of this document. + + "Licensor" shall mean the copyright owner or entity authorized by + the copyright owner that is granting the License. + + "Legal Entity" shall mean the union of the acting entity and all + other entities that control, are controlled by, or are under common + control with that entity. For the purposes of this definition, + "control" means (i) the power, direct or indirect, to cause the + direction or management of such entity, whether by contract or + otherwise, or (ii) ownership of fifty percent (50%) or more of the + outstanding shares, or (iii) beneficial ownership of such entity. + + "You" (or "Your") shall mean an individual or Legal Entity + exercising permissions granted by this License. + + "Source" form shall mean the preferred form for making modifications, + including but not limited to software source code, documentation + source, and configuration files. + + "Object" form shall mean any form resulting from mechanical + transformation or translation of a Source form, including but + not limited to compiled object code, generated documentation, + and conversions to other media types. + + "Work" shall mean the work of authorship, whether in Source or + Object form, made available under the License, as indicated by a + copyright notice that is included in or attached to the work + (an example is provided in the Appendix below). + + "Derivative Works" shall mean any work, whether in Source or Object + form, that is based on (or derived from) the Work and for which the + editorial revisions, annotations, elaborations, or other modifications + represent, as a whole, an original work of authorship. For the purposes + of this License, Derivative Works shall not include works that remain + separable from, or merely link (or bind by name) to the interfaces of, + the Work and Derivative Works thereof. + + "Contribution" shall mean any work of authorship, including + the original version of the Work and any modifications or additions + to that Work or Derivative Works thereof, that is intentionally + submitted to Licensor for inclusion in the Work by the copyright owner + or by an individual or Legal Entity authorized to submit on behalf of + the copyright owner. For the purposes of this definition, "submitted" + means any form of electronic, verbal, or written communication sent + to the Licensor or its representatives, including but not limited to + communication on electronic mailing lists, source code control systems, + and issue tracking systems that are managed by, or on behalf of, the + Licensor for the purpose of discussing and improving the Work, but + excluding communication that is conspicuously marked or otherwise + designated in writing by the copyright owner as "Not a Contribution." + + "Contributor" shall mean Licensor and any individual or Legal Entity + on behalf of whom a Contribution has been received by Licensor and + subsequently incorporated within the Work. + + 2. Grant of Copyright License. Subject to the terms and conditions of + this License, each Contributor hereby grants to You a perpetual, + worldwide, non-exclusive, no-charge, royalty-free, irrevocable + copyright license to reproduce, prepare Derivative Works of, + publicly display, publicly perform, sublicense, and distribute the + Work and such Derivative Works in Source or Object form. + + 3. Grant of Patent License. Subject to the terms and conditions of + this License, each Contributor hereby grants to You a perpetual, + worldwide, non-exclusive, no-charge, royalty-free, irrevocable + (except as stated in this section) patent license to make, have made, + use, offer to sell, sell, import, and otherwise transfer the Work, + where such license applies only to those patent claims licensable + by such Contributor that are necessarily infringed by their + Contribution(s) alone or by combination of their Contribution(s) + with the Work to which such Contribution(s) was submitted. If You + institute patent litigation against any entity (including a + cross-claim or counterclaim in a lawsuit) alleging that the Work + or a Contribution incorporated within the Work constitutes direct + or contributory patent infringement, then any patent licenses + granted to You under this License for that Work shall terminate + as of the date such litigation is filed. + + 4. Redistribution. You may reproduce and distribute copies of the + Work or Derivative Works thereof in any medium, with or without + modifications, and in Source or Object form, provided that You + meet the following conditions: + + (a) You must give any other recipients of the Work or + Derivative Works a copy of this License; and + + (b) You must cause any modified files to carry prominent notices + stating that You changed the files; and + + (c) You must retain, in the Source form of any Derivative Works + that You distribute, all copyright, patent, trademark, and + attribution notices from the Source form of the Work, + excluding those notices that do not pertain to any part of + the Derivative Works; and + + (d) If the Work includes a "NOTICE" text file as part of its + distribution, then any Derivative Works that You distribute must + include a readable copy of the attribution notices contained + within such NOTICE file, excluding those notices that do not + pertain to any part of the Derivative Works, in at least one + of the following places: within a NOTICE text file distributed + as part of the Derivative Works; within the Source form or + documentation, if provided along with the Derivative Works; or, + within a display generated by the Derivative Works, if and + wherever such third-party notices normally appear. The contents + of the NOTICE file are for informational purposes only and + do not modify the License. You may add Your own attribution + notices within Derivative Works that You distribute, alongside + or as an addendum to the NOTICE text from the Work, provided + that such additional attribution notices cannot be construed + as modifying the License. + + You may add Your own copyright statement to Your modifications and + may provide additional or different license terms and conditions + for use, reproduction, or distribution of Your modifications, or + for any such Derivative Works as a whole, provided Your use, + reproduction, and distribution of the Work otherwise complies with + the conditions stated in this License. + + 5. Submission of Contributions. Unless You explicitly state otherwise, + any Contribution intentionally submitted for inclusion in the Work + by You to the Licensor shall be under the terms and conditions of + this License, without any additional terms or conditions. + Notwithstanding the above, nothing herein shall supersede or modify + the terms of any separate license agreement you may have executed + with Licensor regarding such Contributions. + + 6. Trademarks. This License does not grant permission to use the trade + names, trademarks, service marks, or product names of the Licensor, + except as required for reasonable and customary use in describing the + origin of the Work and reproducing the content of the NOTICE file. + + 7. Disclaimer of Warranty. Unless required by applicable law or + agreed to in writing, Licensor provides the Work (and each + Contributor provides its Contributions) on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or + implied, including, without limitation, any warranties or conditions + of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A + PARTICULAR PURPOSE. You are solely responsible for determining the + appropriateness of using or redistributing the Work and assume any + risks associated with Your exercise of permissions under this License. + + 8. Limitation of Liability. In no event and under no legal theory, + whether in tort (including negligence), contract, or otherwise, + unless required by applicable law (such as deliberate and grossly + negligent acts) or agreed to in writing, shall any Contributor be + liable to You for damages, including any direct, indirect, special, + incidental, or consequential damages of any character arising as a + result of this License or out of the use or inability to use the + Work (including but not limited to damages for loss of goodwill, + work stoppage, computer failure or malfunction, or any and all + other commercial damages or losses), even if such Contributor + has been advised of the possibility of such damages. + + 9. Accepting Warranty or Additional Liability. While redistributing + the Work or Derivative Works thereof, You may choose to offer, + and charge a fee for, acceptance of support, warranty, indemnity, + or other liability obligations and/or rights consistent with this + License. However, in accepting such obligations, You may act only + on Your own behalf and on Your sole responsibility, not on behalf + of any other Contributor, and only if You agree to indemnify, + defend, and hold each Contributor harmless for any liability + incurred by, or claims asserted against, such Contributor by reason + of your accepting any such warranty or additional liability. + + END OF TERMS AND CONDITIONS + + APPENDIX: How to apply the Apache License to your work. + + To apply the Apache License to your work, attach the following + boilerplate notice, with the fields enclosed by brackets "[]" + replaced with your own identifying information. (Don't include + the brackets!) The text should be enclosed in the appropriate + comment syntax for the file format. We also recommend that a + file or class name and description of purpose be included on the + same "printed page" as the copyright notice for easier + identification within third-party archives. + + Copyright [yyyy] [name of copyright owner] + + Licensed under the Apache License, Version 2.0 (the "License"); + you may not use this file except in compliance with the License. + You may obtain a copy of the License at + + http://www.apache.org/licenses/LICENSE-2.0 + + Unless required by applicable law or agreed to in writing, software + distributed under the License is distributed on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + See the License for the specific language governing permissions and + limitations under the License. diff --git a/NOTICE b/NOTICE new file mode 100644 index 00000000..e5c7919c --- /dev/null +++ b/NOTICE @@ -0,0 +1,18 @@ +Scene Text Recognition Model Hub +Copyright 2022 Darwin Bautista + +The Initial Developer of strhub/models/abinet (sans system.py) is +Fang et al. (https://github.com/FangShancheng/ABINet). +Copyright 2021-2022 USTC + +The Initial Developer of strhub/models/crnn (sans system.py) is +Jieru Mei (https://github.com/meijieru/crnn.pytorch). +Copyright 2017-2022 Jieru Mei + +The Initial Developer of strhub/models/trba (sans system.py) is +Jeonghun Baek (https://github.com/clovaai/deep-text-recognition-benchmark). +Copyright 2019-2022 NAVER Corp. + +The Initial Developer of strhub/models/vitstr (sans system.py) is +Rowel Atienza (https://github.com/roatienza/deep-text-recognition-benchmark). +Copyright 2021-2022 Rowel Atienza diff --git a/README.md b/README.md new file mode 100644 index 00000000..e2b16d97 --- /dev/null +++ b/README.md @@ -0,0 +1,193 @@ +

+ +# Scene Text Recognition with
Permuted Autoregressive Sequence Models +[![Apache License 2.0](https://img.shields.io/badge/license-Apache 2.0-blue.svg)](https://github.com/baudm/parseq/blob/main/LICENSE) +[![arXiv preprint](http://img.shields.io/badge/arXiv-2207.06966-B31B1B.svg)](https://arxiv.org/abs/2207.06966) +[![In Proc. ECCV 2022](http://img.shields.io/badge/ECCV-2022-4b44ce.svg)](https://eccv2022.ecva.net/) + +[**Darwin Bautista**](https://github.com/baudm) and [**Rowel Atienza**](https://github.com/roatienza) + +Electrical and Electronics Engineering Institute
+University of the Philippines, Diliman + +
+ +Scene Text Recognition (STR) models use language context to be more robust against noisy or corrupted images. Recent approaches like ABINet propose a standalone or external Language Model (LM) for prediction refinement. In this work, we argue that the external LM—which requires upfront allocation of dedicated compute capacity—is inefficient for STR due to its poor performance vs cost characteristics. We propose a more efficient approach using **P**ermuted **A**uto**r**egressive **Seq**uence (PARSeq) models. + +![PARSeq](.github/gh-teaser.png) + +The figure above shows word accuracy (94-character set) vs three common computational cost indicators. PARSeq-S (our base model) achieves state-of-the-art performance while being optimal in parameter count, FLOPS, and latency. Its downsized variant, PARSeq-Ti, also achieves high word accuracy while being comparable to CRNN in parameter count and FLOPS. + +**NOTE:** _P-S and P-Ti are shorthands for PARSeq-S and PARSeq-Ti, respectively. For TRBA and PARSeqA, FLOPS and latency correspond to mean values measured on the benchmark._ + +### Sample Results +
+ +| Input Image | PARSeq-SA | ABINet | TRBA | ViTSTR-S | CRNN | +|:--------------------------------------------------------------------------:|:--------------------:|:-----------------:|:-----------------:|:-----------------:|:-----------------:| +| CHEWBACCA | CHEWBACCA | CHEWBA**GG**A | CHEWBACCA | CHEWBACCA | CHEW**U**ACCA | +| Chevron | Chevro**l** | Chevro\_ | Chevro\_ | Chevr\_\_ | Chevr\_\_ | +| SALMON | SALMON | SALMON | SALMON | SALMON | SA\_MON | +| Verbandstoffe | Verbandst**e**ffe | Verbandst**e**ffe | Verbandst**ell**e | Verbandst**e**ffe | Verbands**le**ffe | +| Kappa | Kappa | Kappa | Ka**s**pa | Kappa | Ka**ad**a | +| 3rdAve | 3rdAve | 3=-Ave | 3rdAve | 3rdAve | **Coke** | + +**NOTE:** _Bold letters and underscores indicate wrong and missing character predictions, respectively._ +
+ +### Method tl;dr + +Our main insight is that with an ensemble of autoregressive (AR) models, we could unify the current STR decoding methods (context-aware AR and context-free non-AR) and the bidirectional (cloze) refinement model: +
Unified STR model
+ +The dynamic nature of attention masking in Transformers allows us to control and change information flow without modifying the model architecture. This characteristic coupled with Permutation Language Modeling (PLM) allows for a _unified_ STR model capable of context-free and context-aware inference, as well as iterative prediction refinement using bidirectional context **without** requiring a standalone language model. PARSeq can be considered an ensemble of AR models with shared architecture and weights: + +![System](.github/system.png) + +## Getting Started +This repository contains the reference implementation for PARSeq and reproduced models (collectively referred to as _Scene Text Recognition Model Hub_). See `NOTICE` for copyright information. +Majority of the code is licensed under the Apache License v2.0 (see `LICENSE`) while ABINet and CRNN sources are +released under the BSD and MIT licenses, respectively (see corresponding `LICENSE` files for details). + +### Installation +Tested on Python 3.9. Requires PyTorch >= 1.10. +```bash +$ pip install -r requirements.txt +$ pip install -e . + ``` +### Datasets +Download the [datasets](Datasets.md) from the following links: +1. [LMDB archives](https://drive.google.com/drive/folders/1NYuoi7dfJVgo-zUJogh8UQZgIMpLviOE) for MJSynth, SynthText, IIIT5k, SVT, SVTP, IC13, IC15, CUTE80, ArT, RCTW17, ReCTS, LSVT, MLT19, COCO-Text, and Uber-Text. +2. [LMDB archives](https://drive.google.com/drive/folders/1D9z_YJVa6f-O0juni-yG5jcwnhvYw-qC) for TextOCR and OpenVINO. + +### Pretrained Models via Torch Hub +Available models are: `abinet`, `crnn`, `trba`, `vitstr`, `parseq_tiny`, and `parseq`. +```python +import torch +from strhub.data.module import SceneTextDataModule + +# Load model and image transforms +parseq = torch.hub.load('baudm/parseq', 'parseq', pretrained=True).eval() +img_transform = SceneTextDataModule.get_transform(parseq.hparams.img_size) + +img = torch.randint(0, 256, (1, 3, 32, 128)) # Load image: (B, C, H, W) +img = img_transform(img) # Preprocess + +logits = parseq(img) +logits.shape # torch.Size([1, 26, 95]), 94 characters + [EOS] symbol + +# Greedy decoding +pred = logits.softmax(-1) +label, confidence = parseq.tokenizer.decode(pred) +``` + +## Training + +The training script can train any supported model. Default configuration is stored in ```configs/model```. You can override any configuration using the command line. Please refer to [Hydra](https://hydra.cc) docs for more info about the syntax. +```bash +$ ./train.py model=parseq model.perm_num=12 model.embed_dim=512 # dual-GPU setup. Set embed_dim to 512 instead of 384 +$ ./train.py +experiment=parseq-tiny trainer.gpus=4 # quad-GPU setup. Train tiny variant of PARSeq. See configs/experiment. +``` + +## Tuning + +We use [Ray Tune](https://www.ray.io/ray-tune) for automated parameter tuning of the learning rate. Extend `tune.py` to support tuning of other hyperparameters. +```bash +$ ./tune.py +tune.num_samples=20 # find optimum LR for PARSeq's default config using 20 trials +$ ./tune.py +experiment=tune_abinet-lm # find the optimum learning rate for ABINet's language model +``` + +## Evaluation +The test script, ```test.py```, can be used to evaluate any model trained with this project. For more info, see ```./test.py --help```. + +PARSeq runtime parameters can be passed using the format `param:type=value`. For example, PARSeq NAR decoding can be invoked via `./test.py parseq.ckpt refine_iters:int=2 decode_ar:bool=false`. + +
Sample commands for reproducing results

+ +### Lowercase alphanumeric comparison on benchmark datasets (Table 6) +```bash +$ ./test.py outputs///checkpoints/last.ckpt # or use the released weights: ./test.py /path/to/parseq.pt +``` +**Sample output:** +| Dataset | # samples | Accuracy | 1 - NED | Confidence | Label Length | +|:---------:|----------:|---------:|--------:|-----------:|-------------:| +| IIIT5k | 3000 | 99.00 | 99.79 | 97.09 | 5.09 | +| SVT | 647 | 97.84 | 99.54 | 95.87 | 5.86 | +| IC13_1015 | 1015 | 98.13 | 99.43 | 97.19 | 5.31 | +| IC15_2077 | 2077 | 89.22 | 96.43 | 91.91 | 5.33 | +| SVTP | 645 | 96.90 | 99.36 | 94.37 | 5.86 | +| CUTE80 | 288 | 98.61 | 99.80 | 96.43 | 5.53 | +| **Combined** | **7672** | **95.95** | **98.78** | **95.34** | **5.33** | +-------------------------------------------------------------------------- + +### Benchmark using different evaluation character sets (Table 4) +```bash +$ ./test.py outputs///checkpoints/last.ckpt # lowercase alphanumeric (36-character set) +$ ./test.py outputs///checkpoints/last.ckpt --cased # mixed-case alphanumeric (62-character set) +$ ./test.py outputs///checkpoints/last.ckpt --cased --punctuation # mixed-case alphanumeric + punctuation (94-character set) +``` + +### Lowercase alphanumeric comparison on more challenging datasets (Table 5) +```bash +$ ./test.py outputs///checkpoints/last.ckpt --new +``` + +### Benchmark Model Compute Requirements (Figure 5) +```bash +$ ./bench.py model=parseq model.decode_ar=false model.refine_iters=3 + +model(x) + Median: 14.87 ms + IQR: 0.33 ms (14.78 to 15.12) + 7 measurements, 10 runs per measurement, 1 thread +| module | #parameters | #flops | #activations | +|:----------------------|:--------------|:---------|:---------------| +| model | 23.833M | 3.255G | 8.214M | +| encoder | 21.381M | 2.88G | 7.127M | +| decoder | 2.368M | 0.371G | 1.078M | +| head | 36.575K | 3.794M | 9.88K | +| text_embed.embedding | 37.248K | 0 | 0 | +``` + +### Latency Measurements vs Output Label Length (Appendix I) +```bash +$ ./bench.py model=parseq model.decode_ar=false model.refine_iters=3 +range=true +``` + +### Orientation robustness benchmark (Appendix J) +```bash +$ ./test.py outputs///checkpoints/last.ckpt --cased --punctuation # no rotation +$ ./test.py outputs///checkpoints/last.ckpt --cased --punctuation --rotation 90 +$ ./test.py outputs///checkpoints/last.ckpt --cased --punctuation --rotation 180 +$ ./test.py outputs///checkpoints/last.ckpt --cased --punctuation --rotation 270 +``` + +### Using trained models to read text from images (Appendix L) +```bash +$ ./read.py outputs///checkpoints/last.ckpt --images demo_images/* +Additional keyword arguments: {} +demo_images/art-01107.jpg: CHEWBACCA +demo_images/coco-1166773.jpg: Chevrol +demo_images/cute-184.jpg: SALMON +demo_images/ic13_word_256.png: Verbandsteffe +demo_images/ic15_word_26.png: Kaopa +demo_images/uber-27491.jpg: 3rdAve + +# use NAR decoding + 2 refinement iterations for PARSeq +$ ./read.py outputs/parseq/2021-10-28_23-23-10/checkpoints/last.ckpt refine_iters:int=2 decode_ar:bool=false --images demo_images/* +``` +

+ +## Citation +If you find our work useful, or use it in your research, please cite: +```bibtex +@InProceedings{Bautista_2022_ECCV_parseq, + author={Bautista, Darwin and Atienza, Rowel}, + title={Scene Text Recognition with Permuted Autoregressive Sequence Models}, + booktitle={Proceedings of the 17th European Conference on Computer Vision (ECCV)}, + month={10}, + year={2022}, + publisher={Springer International Publishing}, + address={Cham} +} +``` diff --git a/bench.py b/bench.py new file mode 100755 index 00000000..2a7b897b --- /dev/null +++ b/bench.py @@ -0,0 +1,57 @@ +#!/usr/bin/env python3 +# Scene Text Recognition Model Hub +# Copyright 2022 Darwin Bautista +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# https://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +import os + +import hydra +import torch +from fvcore.nn import FlopCountAnalysis, ActivationCountAnalysis, flop_count_table +from omegaconf import DictConfig +from torch.utils import benchmark + + +@hydra.main(config_path='configs', config_name='bench') +def main(config: DictConfig): + # For consistent behavior + os.environ['CUBLAS_WORKSPACE_CONFIG'] = ':4096:8' + torch.backends.cudnn.benchmark = False + torch.use_deterministic_algorithms(True) + + device = config.get('device', 'cuda') + + h, w = config.data.img_size + x = torch.rand(1, 3, h, w, device=device) + model = hydra.utils.instantiate(config.model).eval().to(device) + model.freeze() # disable autograd + + if config.get('range', False): + for i in range(1, 26, 4): + timer = benchmark.Timer( + stmt='model(x, len)', + globals={'model': model, 'x': x, 'len': i}) + print(timer.blocked_autorange(min_run_time=1)) + else: + timer = benchmark.Timer( + stmt='model(x)', + globals={'model': model, 'x': x}) + flops = FlopCountAnalysis(model, x) + acts = ActivationCountAnalysis(model, x) + print(timer.blocked_autorange(min_run_time=1)) + print(flop_count_table(flops, 1, acts, False)) + + +if __name__ == '__main__': + main() diff --git a/configs/bench.yaml b/configs/bench.yaml new file mode 100644 index 00000000..4b5872aa --- /dev/null +++ b/configs/bench.yaml @@ -0,0 +1,10 @@ +# Disable any logging or output +defaults: + - main + - _self_ + - override hydra/job_logging: disabled + +hydra: + output_subdir: null + run: + dir: . diff --git a/configs/charset/36_lowercase.yaml b/configs/charset/36_lowercase.yaml new file mode 100644 index 00000000..ce2a5a08 --- /dev/null +++ b/configs/charset/36_lowercase.yaml @@ -0,0 +1,3 @@ +# @package _global_ +model: + charset_train: "0123456789abcdefghijklmnopqrstuvwxyz" diff --git a/configs/charset/62_mixed-case.yaml b/configs/charset/62_mixed-case.yaml new file mode 100644 index 00000000..07db8445 --- /dev/null +++ b/configs/charset/62_mixed-case.yaml @@ -0,0 +1,3 @@ +# @package _global_ +model: + charset_train: "0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ" diff --git a/configs/charset/94_full.yaml b/configs/charset/94_full.yaml new file mode 100644 index 00000000..186bf42e --- /dev/null +++ b/configs/charset/94_full.yaml @@ -0,0 +1,3 @@ +# @package _global_ +model: + charset_train: "0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ!\"#$%&'()*+,-./:;<=>?@[\\]^_`{|}~" diff --git a/configs/experiment/abinet-sv.yaml b/configs/experiment/abinet-sv.yaml new file mode 100644 index 00000000..fa2b0118 --- /dev/null +++ b/configs/experiment/abinet-sv.yaml @@ -0,0 +1,8 @@ +# @package _global_ +defaults: + - override /model: abinet + +model: + name: abinet-sv + v_num_layers: 2 + v_attention: attention diff --git a/configs/experiment/abinet.yaml b/configs/experiment/abinet.yaml new file mode 100644 index 00000000..d6915e0a --- /dev/null +++ b/configs/experiment/abinet.yaml @@ -0,0 +1,3 @@ +# @package _global_ +defaults: + - override /model: abinet diff --git a/configs/experiment/crnn.yaml b/configs/experiment/crnn.yaml new file mode 100644 index 00000000..32e6c028 --- /dev/null +++ b/configs/experiment/crnn.yaml @@ -0,0 +1,6 @@ +# @package _global_ +defaults: + - override /model: crnn + +data: + num_workers: 5 diff --git a/configs/experiment/parseq-tiny.yaml b/configs/experiment/parseq-tiny.yaml new file mode 100644 index 00000000..9e5879ef --- /dev/null +++ b/configs/experiment/parseq-tiny.yaml @@ -0,0 +1,6 @@ +# @package _global_ +model: + name: parseq-tiny + embed_dim: 192 + enc_num_heads: 3 + dec_num_heads: 6 diff --git a/configs/experiment/parseq.yaml b/configs/experiment/parseq.yaml new file mode 100644 index 00000000..bfb8da13 --- /dev/null +++ b/configs/experiment/parseq.yaml @@ -0,0 +1,3 @@ +# @package _global_ +defaults: + - override /model: parseq diff --git a/configs/experiment/trba.yaml b/configs/experiment/trba.yaml new file mode 100644 index 00000000..59907202 --- /dev/null +++ b/configs/experiment/trba.yaml @@ -0,0 +1,6 @@ +# @package _global_ +defaults: + - override /model: trba + +data: + num_workers: 3 diff --git a/configs/experiment/trbc.yaml b/configs/experiment/trbc.yaml new file mode 100644 index 00000000..a66ba1da --- /dev/null +++ b/configs/experiment/trbc.yaml @@ -0,0 +1,11 @@ +# @package _global_ +defaults: + - override /model: trba + +model: + name: trbc + _target_: strhub.models.trba.system.TRBC + lr: 1e-4 + +data: + num_workers: 3 diff --git a/configs/experiment/tune_abinet-lm.yaml b/configs/experiment/tune_abinet-lm.yaml new file mode 100644 index 00000000..6efa5283 --- /dev/null +++ b/configs/experiment/tune_abinet-lm.yaml @@ -0,0 +1,17 @@ +# @package _global_ +defaults: + - override /model: abinet + +model: + name: abinet-lm + lm_only: true + +data: + augment: false + num_workers: 3 + +tune: + gpus_per_trial: 0.5 + lr: + min: 1e-5 + max: 1e-3 diff --git a/configs/experiment/vitstr.yaml b/configs/experiment/vitstr.yaml new file mode 100644 index 00000000..6a8111ae --- /dev/null +++ b/configs/experiment/vitstr.yaml @@ -0,0 +1,7 @@ +# @package _global_ +defaults: + - override /model: vitstr + +model: + img_size: [ 32, 128 ] # [ height, width ] + patch_size: [ 4, 8 ] # [ height, width ] diff --git a/configs/main.yaml b/configs/main.yaml new file mode 100644 index 00000000..759f2806 --- /dev/null +++ b/configs/main.yaml @@ -0,0 +1,47 @@ +defaults: + - _self_ + - model: parseq + - charset: 94_full + - train_set: real + +model: + _convert_: all + img_size: [ 32, 128 ] # [ height, width ] + max_label_length: 25 + # The ordering in charset_train matters. It determines the token IDs assigned to each character. + charset_train: ??? + # For charset_test, ordering doesn't matter. + charset_test: "0123456789abcdefghijklmnopqrstuvwxyz" + batch_size: 384 + weight_decay: 0.0 + warmup_pct: 0.075 # equivalent to 1.5 epochs of warm up + +data: + _target_: strhub.data.module.SceneTextDataModule + root_dir: data + train_dir: ??? + img_size: ${model.img_size} + charset_train: ${model.charset_train} + charset_test: ${model.charset_test} + max_label_length: ${model.max_label_length} + batch_size: ${model.batch_size} + num_workers: 2 + augment: true + +trainer: + _target_: pytorch_lightning.Trainer + _convert_: all + val_check_interval: 1000 + #max_steps: 169680 # 20 epochs x 8484 steps (for batch size = 384, real data) + max_epochs: 20 + terminate_on_nan: true + gradient_clip_val: 20 + gpus: 2 + +hydra: + output_subdir: config + run: + dir: outputs/${model.name}/${now:%Y-%m-%d}_${now:%H-%M-%S} + sweep: + dir: multirun/${model.name}/${now:%Y-%m-%d}_${now:%H-%M-%S} + subdir: ${hydra.job.override_dirname} diff --git a/configs/model/abinet.yaml b/configs/model/abinet.yaml new file mode 100644 index 00000000..a19757ad --- /dev/null +++ b/configs/model/abinet.yaml @@ -0,0 +1,26 @@ +name: abinet +_target_: strhub.models.abinet.system.ABINet + +# Shared Transformer configuration +d_model: 512 +nhead: 8 +d_inner: 2048 +activation: relu +dropout: 0.1 + +# Architecture +v_backbone: transformer +v_num_layers: 3 +v_attention: position +v_attention_mode: nearest +l_num_layers: 4 +l_use_self_attn: false + +# Training +lr: 3.4e-4 +l_lr: 3e-4 +iter_size: 3 +a_loss_weight: 1. +v_loss_weight: 1. +l_loss_weight: 1. +l_detach: true diff --git a/configs/model/crnn.yaml b/configs/model/crnn.yaml new file mode 100644 index 00000000..9c109e4f --- /dev/null +++ b/configs/model/crnn.yaml @@ -0,0 +1,9 @@ +name: crnn +_target_: strhub.models.crnn.system.CRNN + +# Architecture +hidden_size: 256 +leaky_relu: false + +# Training +lr: 5.1e-4 diff --git a/configs/model/parseq.yaml b/configs/model/parseq.yaml new file mode 100644 index 00000000..16c9b09a --- /dev/null +++ b/configs/model/parseq.yaml @@ -0,0 +1,25 @@ +name: parseq +_target_: strhub.models.parseq.system.PARSeq + +# Data +patch_size: [ 4, 8 ] # [ height, width ] + +# Architecture +embed_dim: 384 +enc_num_heads: 6 +enc_mlp_ratio: 4 +enc_depth: 12 +dec_num_heads: 12 +dec_mlp_ratio: 4 +dec_depth: 1 + +# Training +lr: 7e-4 +perm_num: 6 +perm_forward: true +perm_mirrored: true +dropout: 0.1 + +# Decoding mode (test) +decode_ar: true +refine_iters: 1 diff --git a/configs/model/trba.yaml b/configs/model/trba.yaml new file mode 100644 index 00000000..717e4642 --- /dev/null +++ b/configs/model/trba.yaml @@ -0,0 +1,10 @@ +name: trba +_target_: strhub.models.trba.system.TRBA + +# Architecture +num_fiducial: 20 +output_channel: 512 +hidden_size: 256 + +# Training +lr: 6.9e-4 diff --git a/configs/model/vitstr.yaml b/configs/model/vitstr.yaml new file mode 100644 index 00000000..a8207426 --- /dev/null +++ b/configs/model/vitstr.yaml @@ -0,0 +1,13 @@ +name: vitstr +_target_: strhub.models.vitstr.system.ViTSTR + +# Data +img_size: [ 224, 224 ] # [ height, width ] +patch_size: [ 16, 16 ] # [ height, width ] + +# Architecture +embed_dim: 384 +num_heads: 6 + +# Training +lr: 8.9e-4 diff --git a/configs/train_set/real.yaml b/configs/train_set/real.yaml new file mode 100644 index 00000000..786042d1 --- /dev/null +++ b/configs/train_set/real.yaml @@ -0,0 +1,3 @@ +# @package _global_ +data: + train_dir: real diff --git a/configs/train_set/synth.yaml b/configs/train_set/synth.yaml new file mode 100644 index 00000000..0a7f0719 --- /dev/null +++ b/configs/train_set/synth.yaml @@ -0,0 +1,7 @@ +# @package _global_ +data: + train_dir: synth + num_workers: 3 + +trainer: + limit_train_batches: 0.20496 # to match the steps per epoch of `real` diff --git a/configs/tune.yaml b/configs/tune.yaml new file mode 100644 index 00000000..d58ad9cd --- /dev/null +++ b/configs/tune.yaml @@ -0,0 +1,10 @@ +defaults: + - main + - _self_ + +trainer: + gpus: 1 # tuning with DDP is not yet supported. + +hydra: + run: + dir: ray_results/${model.name}/${now:%Y-%m-%d}_${now:%H-%M-%S} diff --git a/demo_images/art-01107.jpg b/demo_images/art-01107.jpg new file mode 100644 index 0000000000000000000000000000000000000000..157b4072f62ccee9ac1564d140b3b720975ffdf9 GIT binary patch literal 182385 zcmbTdXEa=I_%=EWMi(sxqh+)q5#5YVg3)`45`yT2(MLolktN`0G^w9l9T+e0Q{dpLJA}!rvOn>QPbQEXu1U;B>@6S$$;eKWMns^!*A{b z$Qa3)_$1UQn2n!+?)b4tMx_-}Le=ZLSxqJm_@x|C(Nxs8*}&`^0)j%qFcE1PSvfdD z{=q{HO)aFhj;Wcsg{9RaYbWQYF0O9w9{vG=LBS!R=$P2J_=K0Q64Ns>v$Egh6zKN`R@xq*9hwyo5WvR+lNQTC#PrU z7nfK6gNpn<~x#6 zlq~9Lh4tN3P$`oGR!7tX^=*FXp8|*f1MPny`+o;4`u`WQ{|ngvjcWlw3naNo9*_~B z0@#3>gOasJ$fE;ywvb387VhN=i6;M*POHDTGj0PdXtg-zpe$cYTcvZ8mcq_+#OrQ(qL4WS)-0)|2#kANiqMs&#h&wtIRQaE82gGxTl9+ ze}>pR^F@Wj(wP{*(x=2Uj7#1(!Egp^w@OAF;z`CrYW~`oa;60MYTOFoua_&xBjs@W z`d6PJe?$^cOwt!#0)&=#72KIrXT+)8oup@jg17oB#EY=bgeX0Jzr}?}NLfrqijMmk zP`R;6?J;N#MzTsT%#{vd*;9BWw2&IO8t$Rdg~6aSw6Lx!*~0vpvdqA0&B1T$z6#ih zFBtM(IWC*2CQ6HGHW2RnZ^J&DS`Rn=s3o*furu+f;Il&1CcB_rkcA+ z!{Q1?bXMwJktJcL(z)HlG7o?Mv1;B?lpCGqDqLtXleq2I}kRNv3DzQ!l$px)xLY_Hzt*kV~F%MjIHYP^-6|DQ9wX)@T4S zu-I^6K|KR;zwwh1ny&7$ECCd`<^#rL*n6Kic>mP|u0429e{qbf;sIk9B$28EUp`7H zRINz&{e(-xxL&B561j!saPw6z1I?y76EE`sT?jxvTK*W9$y)9jm1Y38WxPt%YZfms z(uG0mi!CKRhW#83g>G`fI1S9xIOTbF9}eKDaifo9!v>59BE4vhTY*tivq=T1YMaqs zSM;D1!)n3Zut^N5fjiZDKtZ+UI_r6v7%H?^x_n#ZfsUlF{J4IU>X_|xTf4-`m;rN9Y1Nqj)~2$-O>W#*c3u~zJH z#@@F}o3@Y~{Ivdr6~+!{>9Y-=*~3ohWy@egBS{uMMbJbxP}Dq0U3d0H)V%##-~NU! zs!5A=?DanyxRhuZn}0h<6mM*>BsHdoHof%(#b6zx=qtLBd*EG}NT~!O7yuPl!=~W9 z$m*kL533rGpHEJ4pH!N`y=lh`XsJ7`Dtutazyj~#{v<7{%$YfnWZqXgGLphibpejJ z-lAs!7MvE-`Vnd)(0Vw6G)canLAR1bqB`fCa_gQdp}x-z;#yyrj2w|Tst=noP{6@R zz+;drYI5o+J6AIu&Jni<(SC~ay|?R*Ri>=%P~rO(GpDR zA($>>D@C$z$m;j3?rq)1aAAi?6Yjz-`0Y0{m$TlC&|}Sv=hSo{gGjvsrx&|>cQKcnn0W4!SD9#ER;lnLpX) zHNqa*m{w=3$$sIbAr{#86-01*R#Er_hegF$(SaXpQVH=+Gl`Oez+)v|AD2YzNu>fg zY@&qBu2sYPR!rb2WWLhiiNK9@gr!~+oLo6V{&}m2ZNfV$YF)G!d3BQ-pq-uF*eT* zb10fq2cYlDD*LwSyypVKMrAj!c`IB^3(2bkDQ6!m`2Ll$VA3wW1_~G z9^uNOsrFmzI{vN81cM9zi|F8m%UzsjH6 z6xl{jEY~WLRC9(u9yI>8t+th6FI`jzL0W#-SLSRhdX{YgY6D%(@wYl+JJ~Z2aavRP zZb@|yY7C30Gm&852jQZd#dD{kws+Z3B3D2Qhr{HSXLstpV6R9|@uu-2Li;1EkcW>> zP)sB7r<-Njw_<{jpr(-Kq(dH)#HVrw>Gx4T2zIQ$do_{;NTAP8IV%NvP6=#FTp`V9 z)mT=tw38Ay@xp}T&Gzk{wUXHNyAzFCWkPpa3JzCZE4=(S&zhn-l(?R4^D?oqBJ+db zXeP;LR=FHCb~txa%ClW`&mEx+#Fgpm zNH`>H&ax>sJJus^ag6tc85wv>r;4&<%%fz;k<%R`f#IG3mQtvVicB$)tB3CNKDD5# zrZa`*N_2x(-lsoQlmkwizk;r(C1>fV`lnr5CcQ#gb4l;oMsJvIJh6BkuZIbjvP$5G zxK<-SiuZ#c76E<@rB%K1mQM#Z*lD6-%$_cUS6JI76|bMI<@qY@aLjBNFX#0^%H6!n z$cTq!KG1xkUbm1EZ2tGW1@M{LpIrQ~2QvQ&oPQcjhC*{s|7`uM;E1*9jQ%JCUAzOt zy=W5Atj3b$mptv|PaD7+>RDt4^QJ_Upf9}iyoaRy4d~AwMlWlrYZ3HAB#QwIv&Bp> z#0cUgG;#m$QW~3EZ!1Q09iZfACI6jt1`mRge)23?!9&6|{L#}6ixLEx%@2O&J&Xzy zL2yJ?wt6@EtdJ58g8ryh)^uD=jtu^-&*_(nIII0w0zS%r=g1kzh5xDY@<`!oGM&8} zN)6KS@oh714N}_TaQW$%j?S#bV;(xVf%VQV56UhRK`#8yPTJ}270Og=F zAvqN3sKl6Y-aE0Ff=dIxxhB`!ig2G7S^P|tue>%I^bK2KRt*`@EPf&v9fHxx4xi$Y zteYYc!;qPx6FWG3-J%@bXhQM3;zJwpqQBJCf(n>d14ev*FJlZHg@lk3Qo~vw(>&D{08cz8T=h1QiE*J-s+DVa&pK z!-*N%r7Lm5y4kx8v?4BGb73tLPc^SXI2l1H&N!LvIoXc&f!arb<51Qn)^L3FiO5*# zN5KQ&eXp6@O^&&${y!48y{B)53ZSV&o5@&x%q(@w*!eyLTf#pkaX;@qIjxYrMM8@X zV4c=3(U50VFuOYc;Hr6a^x^w?)zkRh`!2&+xi{l!$*mtoJbwB|YSq^ya#B6%_si+u zGctVAEK4m(i8sg;;h^?QM-nL|`Y8~7z8YSO^BSQpj|I)&aQ(4(p4y+(H2$t25-;8S zQP&3*7dZx(i7kRFMavpG8oTm>_)yC4-{;MH(IdgzQrM~W&dfpchQ@Ol`lIyhE>R5Q z!{VH<^0)%uw{u*rC9l|DeT~OgBPXwy0`?NUg2}0%hij$VN>ocx-DIf}dpbFN^pq9+ znNpA%{~s)g|0 zl1m!TOHlq^Szed}BRBLY{Z{2Rn2#{+d#P&Ffr1}LH{WwTF5)y)J{`cL zti1UN^D~ROxwMHDP+`1?JjV)aTRWl<_1O1u+>wsuct4;f#S!Nh zh#6}8hilSn=Q35%qD839Rz<)DMHi!1jmPpwDjxq@LgHrMdjXTGA}+OFuSeTPBVb!> zl-MuP6Ru)lo!@Ue*jC{u4`jbLA7mp6Us3)a z;NvWF{$282PrHqW2!7YeLfoD#7B@rJtE89rdTFq|j#PYCoN6`Zk>xFxr>K0fkm{Cc zK4B+*#$>9hOX@M7E(q>>F%c+7*^yYP)Kv%O(JB2sqk6i2Xrb-GX-0TvYpxARE^tyB zK~1-bXe&`V=4Q|{z<*mp+rWPp0oq$D6_>8jMM85dHP7308K)gT>L7C!-qF3gY7Eov z3GTJgqLyP@3KxAFt;P+8U>uvDv_*a06>UccSMv$B$dvBCK3LQevzA3DX?pB^k`Cqm zVO{IBbN&ht*o&MyDDobC@vibco%AyH0)if zUAsrU62eZps+j5&A&lo%qG$)e3TbthS!mWuCa^W zFOlY!{95+m|k2X{YG?{i5w~7rjwIZ;~ zaMs`J{(#i}p3(Q?zH}`Hr6QkXK$tiBMr7wVX)aCr@g(V3h2P(R_m-Bm1`ljIg2RH$ z0hnU#e6*x_smjZpIWsDitVE}{2Q0PO(WKbr&fJGIv!{}jbgx1J%DjLnXwA!4#5^&v z_Y4NHCKX)rc@N1Tfu|MVxK)x7z9scu1Bux@yimGOycDI$^d?KaE zH=G*+T0<_!D=|izI?2Mc1l0ZvQmFRkqN+^hG^G&)Y`p;x)Tk*DmJF71tXk}%ZNj+f9#|`M>>#zE>n{4&ZxXEjqmu8X5 zKJt^o#5={*;26DyGWe+fOr#?;ye;y$rud34Kkr&lHt z#Qq7q=S?wJNPde*rPba6`NAC!UaJ z(v@;pVBtvE5c*OgP1;x()jcr6XFQETJCB>$#kvf1{e^v%eBm`o4$AZ0>1CsdA6@@v z*2*85z{2Xap8pkBg|cY| z?v%!9n%R_|0wtA9F3NoGgwfab!0(52fdfSDClbuiDas9=@FD7fsV4=22LZo8@lk37U9PQ}p%Q-EXVhOO~=y7k7`3c~%=b zXd)@}g@85a^I|e*hfU0xa!2LHSWNm%y%v<$-w=dT9rM6A^KE9XGp4$+sTL{MLQf6|<0Ws~{bDCuR)tf+c1J8+5fA zR7|Po6zN8m7K0oGAo`S0B2^1n)gPJfebWI|(#~iyW%u8A{OQ2(4DIQ((59q?zY>U7 z%^!L4T_1%`=ndlR#CtLP8N2CMYBl3c6RQvKza`#<2szEX7P)#MV?wZ~Zw8IFX`GFO zJpB)l(p0$=Bi~ncVcO(>E?+XUn4vUWV^~r9IE8ll&7pu3V$Ed!Twe>nAdrZ)WqWcB z&D(sYd{49OcC+)9go5Y5F$Utlqj^NtHE>#Mox<>h?=Y;-kQLh0e(|nqS>h;{&;H3@ zPm15Y*1bQ6-%LiLKQxm&W%Pkcq}Rnnr%NJxOfMWTPpu6r96j5f@(=JmKNT#X5AD-F zEsUgP@G{l(tWXzyh*`qBrs>XH7Asy)#{@qk>htQ01F=^@C1j@%(^zM0WM~Pa@@nx5 zV25>VzG?&A-%q%csE=qT!K!=3Rr8cfH=lElZya2weTtQja)^!2{{eI+o@^_nHfe@A z1^N2IDB)dOJ9c(ukTGDE`VlVz;y92ake)@Fw;h^7s$j5X;$X^kSC44IK ztoGUK)lzKhk5JJR!#CHWkAM6Wi=tQjmTiC-V=tOl3orcR{QPS8VqVK2K{^uIBQyoB zEt>k75N{T9#?zJ8#M3iI%hx}ZN~qsVFVU*xY0(QReL$k;VS{Ut&mk@3t}1kIB~eTmM#H@krU&ywRgf_8;Sd^-Ym1 z1{oG?c4iooUIZNe&1L!z&?=HO*5gg#_~>3@#=I=7Pg*uuV8Hfrz)3ZBJ?4k>*ITsv zH=F#B&lg%H!(~G*2|o}rDgFl@#*BYtrhOQKIo-~r7bi8s9(>D{r7!zp{gN29^SIXk zz}1o+Ci7*S`5$Xi=s~4@vqH8&D#K#~$qRa52?9l7>FU(mQPcO!=it^+Nd zWMQj1p`afnPeo*)pI}PT_m<$v002?xLm}!%Z=d@yjEpv2fjY91r>`N?SwPgsrpYw@ zThwXg&N)iRd!?^o(jrNV1>3V1w5%?j@)LQR|T?yiSG`>FS)tXo-H8X2G~3rxiACh?y&-cK@cK`H;5 zd`ZsmSv!ddp}fz#ECp;E=uFq5IV$|)ugrZ&d|I!j3wuZ<(mlJq z{@wl2b0^pF$F_D&i@`;OlYgV(rY62(9{E=^*-Foxix8jy^FrS(VDb(9ASuSfKH_3? z(V)|=IuJ^K^UhP$;6rJzi_2HuT5%jnSpG3~Ti^&}qx)Q~@T9A2YbC#xCo|b^j%D3G zpEl&IhE?Ye2W&pc(ie{?z~jUt9&hgxTJ-*~({PmYx)6p}3oph}dxV-Gx?y5pe(;dH69{$cNnG%ASCXmIKK=gHE zN|@Y#AUG4I^ELyQ{$?fO{B3@-i(=Q$DC$o87CpY3zzH#SW{9#!Qu)koldfdzI3%xW zOw`fdVv3G7jbfy%CcMA&L#kB%*UmK>4X0!Do{rZGq(U8g0ty!D|E>~uL)zs$s!us` zx8#emZx>E<&nfM$ZyOZ{Ul&bgWC1xF7FlIUy0`sT;_a40SSyoq5FS#)0-CzFxf%?a zX-PTrbsS53aK7R0;)(Pmv>s3WigV?4SbG@LEJ5>S4%!a$R~8OJ&}2}Abr68-3w62n zdx1iahYvH(pW$o#^PV{3%g=i)j_Jig8zxj{P6huO5vBO*s>p|6Ky@@|w#cDt{kmr> zv8jRg7Nv}2S2}FSTbY>XXWM(E;9<6UpFiMNNHMj20pN@k_L8o!4Rn#>(cw-0p5SM< zQdDi`m@#eJ0=w(sTf9n`%yWeDx-_ozleu8looYOy=%$Il#L}!+nby;hk=4-*dYeVy zWz0W#B{`Ll(&uUYI4|1p@Etj4#-#BH-hQQp)ZSxA@b5em*#=g2_H|riA$wg#e)WUk zr?PuRU}o^P-}Lx!w`%vb$2VUIB`Q|YhgEGxnhG<0Tu5Su2p>P~i#sKpzVs41QF`B7 z{CVC?=b>OvMOv|vAaL=wr9SJ@ zuZ?rD!P}cO!S`dPN_52DXLY)hrcG$~nzkpJM`XSJCu6_qt8Btuquu)pV=A|nPv+eX z{zd1e1s9Ixkk$u`cy& z#zuH2AJ#kDX%##Y@YBM%k22Q@-2Yd5Qx=djUxOiQZsi$fNH8=vklnLha)W&ZXXE&t%Vp)9KYF$UDd zM4oq#kgNk;>t?T|bEBF2^D?mJ&f-!NGb%Z~v@!yQ^ev!5!w#>-cVaPc2&+K04d*Cn~%jo6fc z3w$@0Hfz|R?hx48drCvZ&TR9u=PmZ1q?Ymq(q8gMB`;YjAvF}T*0I!~M5)vHuCY(~}{Vb>(-#8r_xY_u)`J}C)ajz<3nkgsmb{Hfh_R;22u_v>C@o0B=oE4As{ zo6UFT?mVS79N86pm2y9O6+P4RmeW@onD~5Bl{i#CzPuf0gQ6Q&Fh_{3Yz^l;*=F!4 zquq)|A5T_23?^M&N4qqev?&gJuD7PGSwphgVXJa!3zf#K8`*Oyf%$sVFHBjwe}>=m z|1V!p*O>Z)9;A$$$2NIp-EdHgki>d#H=Jc|_v6i4Ydm=1H3EvAok(ht!EnwGg(1v? z8q$jjMm^ft_SUV1C0_;@V8 zOUQmcvzphy+ROk{Jdr`ek=;iZRYf|x<(;uXaV@E<7>rAC`n)KyLdihnM`l3)FL#!s zWQp&0ehX}~T--u~V&-Zaa7P=P%7E6`X-tGt~|{Y04SuX`Vl?ah~jq+ z!)R)~qGRf8m0nsu=-C(PTu1a;%G&~X5jLvN{mqoB3BICaah#|;z1oUc%Lm_VIfV(+ zlU}1;M&$RB^GoD_?Avt8b|o(#1*H&C`1RG$6 zPp15J(y~0Yk-<`)pVC~lprjwJqg>QGfsZB~J!M<(BQ9(d0v{-C8|Zq!kh~iNACP`m zj#o#AnSM!to)(mSxCP_IVx49)g&tr@4B3HP=OjI&9b$Tz2G13GxKM1EhMKjt2k z-tWh3P6;EDyrZ+RjrVx0)Lr)Q7((qtFG&JgiBc#WUi?Jl&O~#4Ns0(ibTNL4bM%iW$k%vqLTX zQ=PxOR`U!l&$iU)<=DvR(SJ-Fa9?suoZ?5dbV!fKqY`Q2dyqRZc_IHEl*qgR|H~8> zT3m_0J#9LMmL&O_mQo^QFkSdDbUm1b<>Zzu0~yvNW>m{MNQ=+V%Q4TAmPDFyng;R* zZL-dd>=goH30G4I4{-k)S#sXiEzw44w_#I-`vAOCt-u<0@^92g-c=*D$07YYI+lNH z+&}9|oV;Ia{SxA^I(1tT&4}+nb{&C$N`89VAv{SEmDD?ZQdkpVBLBa}_DGf+jn0T< z{p5s&W}K1^!&AmGQUw1O$m^pDEg_eqho4DPTRJ<_I+aZ(wXNQA?bwR;8;})_Yoeq2|y-E+@a;Fk_@~ylZf;&0sgX15cvWZ?E>!#{pu-wztekTIN z)^!fg(y-)43VsJt3U?x4^%#$~A@A>#irTT0iyLQ8X2YJ>cpa$q<4Xc?x4koXj6C16 zUEdDOA97H@a2%Hbm8^~X!zeuW%4J-lgyFV`#f|ucGI(=hP9%mWHzMY^_@R8B-UgA)f z3=wSdYFCOtlSHp*jp>LqZhHjS~gSA31b{%tv_xLk{UUOrj5NX zlqqrjsr2|D<<8e2&|hqX{FTSh*9E2MsApx1H7ozDMmz)>jWPVjwG-~;6Dc|G4Dh2C z5!BaeelY@{dT_%LU6vY84j!MW(v3{}MJ=NuO2R@^zXUjD@*+-8Kb)o6(;C065KG9l zy5^ktDKf_<$8?n}?is`rzrr3%>wYM~U9djq)B;Qd`Td23xu38}xoCVzkbdiIo;WIP z1n1PS)#XGUHm|7sSvPN|KZ!|($$t#{7}&YS?@<)T-35RVaP!=m6Wv<|AlA3fT2E;i zdOG;1!)~i``S}glq8bFqVHt7}BEdzC_G*z*l zx(I5M_Mj028Z+)o$3`IygX~|!%5KD!59%PYG-G0(>bq_O`KzgcvB}iHGH%$bc}&GM zO!X!yJElFPfUzyjiSax!j$3s@0b{do>(Q*BZ4p%Ko<6ld*3V$~u4TwXGP@f0su_cm zS6Ai3V(wn_OwRzKcqj4sULc5b+kZ}(c3kXjz$d%<2)pQyt{3hE0&lQJC9e-GW@dR*PJ81wpETrpMM_vF25Z4U6wnAb)BassS06&@HssEd) z?fa4b=Ie56KIR3z$g9E+UnwL`5M1^4*7{`&w$XykNypQKYNy8+oT{Cm;i>l)D&LCo zEDG=K$A1Xc8bqx9O>aal{t*CMlfC}XELaemb(oiXq6@g#*n8lIyOYnC=Z@=PHPj;6| zJlZTw%@l_q+?aKg4x@G+J&N0__vT9B%wfr1WOE9tVpN%7Q@wf6xoQ;m$2SlAhaN_t#~2b%&1U+$z3Z>fqh&5WzvvO5$%a5=cZK zPI+&ZmB26Gxa7B=-K~4|s*iH#v9gp{rIwBDR>>p1EoTCshPli7y8o7OO$iGDChQPK z9m*RS!jqzNX=~81fG?yo$~j**InJ|n1=apE!26P;lCPpSZg}1XYG2fsca|Vfn&p$r z1n#Vfgrx7xPS)aO`)hSF^HaOYMNus{&(g0^$ALA$-`p&Qv-dhh@8=S| zLKi>}kVq5lZMUGFr^#UKY1w@QVnF89H^Uftz4^hwyS|u-B7{4$KAmBZLaNUIwo0$1 z_Ls0w^wgGT1$Yryn&)XOTw3F9-OVzn+t1jM!h&x(>H!}6&&jAt33vNre3eAZ(-NJD z+n!bdit=yO5^&>jWLzS9+{scO@2KP_$)j+ zm0}&+=f5qgua+hYyQlX@68@CGGU&s4XD`a8BD-Af#;BspGqjl8q*$@hIeWeJR@@L0 zW+{VgtzWK*u51huX(MysVEz744w5|mS)A=&B*^0>3nZJDd7x0pG4m}tnYnqS#t>Md zED4ULvf$Eq_yZxP>;MNCwVAVn&!rLkQKijd#gj}FRuLkCN&gPKFho}BdePJu&veXD zRg(T==vF1NoVzB?6ZzGMYO59lB@cf^!#V2t((sD+4uf4OhSsx+mEqdUZk9*&Swg=w zID=N}F6;X%1oB6G2T4_)o1=7jI{NoaMFRxaD-e%;Ew@9;E_V%|Y{i(@X|EoHAj&8g z-3Gak(>W~X+aQNRLn5KLwmnlti8cnN=dCB@ch`Q1^5+ZpX?QNE8t&tH16U`t zW!+g;He{l>)GN9n$JRlUycqR;l)UR~LF<$ODuF&C_PlBfnhhFQZuN^cAb%!2C|}ry z`WztlqCYB_E>}tIx3)|%#r-!!CWuRYfBOYXYBhSl-$uIOexHN$%<_L$bv8~;mkFNy z@={E@&%{btNvpZXB-Ps5zg<+~(Ac2MM?L^;@t|onshWq6Rr^nWlPk}TYZEe01?xIZ zpQ^mm@bevxJ^Z{-iRqVLCz(Ditzi$n@n8|NwC-379*FFyDu~Mxep;Mne$fwVhE-vn0)2^?sWk(tRl;m$ zd&0p>HcC~*mD&3{tcGe^$VJl@54&q}pR)1soGODz;T{pfG`)v&*83n>bmcVTY$a7D zpZ!|-WRTrRs)46Vy1K5uK5C2Kh7f3>6tbBohO{daQtlAJ=uxoShAVw=LIm;xsNk$d zW!Z=U|jqXn(Eo?=OcF1AK2iF zTZzQ#CU9(+)d2%Y#^r~}%g!I-|703Ht!81?o4+(flZv>uNJ$nd1qLGlMCoIw&-&pa z)}nL9Y5^uSNC}E`kMY&xra+q}lR!)|+CGM`9)nIkTB@(^acW4 zBZ9~CX?vOlGx#qcM0YXZNNqE zG^}Gz0tlxiLRx(0d{;EZLLpSJvoy(=(Z8i*)j*OAQq#Ljf>7Gp z|BR-e9@k#?buXv`D~=;8Bj=x7PBw!KGIPP_r_w8B3(@Hiu+L^@gYegr5=X$+ShXCl z+{o7SxPqJ5k2m?XEW3Pq_XUf{AcXSs5)dwvDUJqv7!C97!4g-Jx*&Nbt3hk2CpNE} zv`H$8otZJlNk3%ar={5h_OK z=?RE;T&AG-mN;jgdh4%{vl24qv|cg@B%pwQzSs{NEld^wm}HKMRv}S!fuxe(7BY(k zfVRVbZGX$qZpB6|WIEr6@{jZ9ByhG}O<8UVrHtYvmQn82J8ocl9EgG0gdW zIT9F&&9qk{^hm5c9r#3%57ATi4Hr!M&d2u|BSA zlNDmDw0P`FoDmgnm~Y}j;q@`O`%VvjSy?iExHgFvblY`wrz8YviVlk1-@ng)U%G1j zbjjcR2%(&HV&%hCWN6)aj1YSH%&B-*(Yw;4X7Ct1S2`mzCO!K5xL&AD$wQp2LfQ%8 zLud-itUIaD%4Vz|UH@TxX_{^*Jub)L+ZRUacJww`>ZCeTI~eM<9y4x4!iDC}lDARl zBz+;%P#U6TK4V2=qs=qD$7^xWGiqH-zpO{gB(Z=?Ybc>;%ok!`l|u~ zYZ+Q}eRooN!&D+Ym@C-H+ox_Kr5AeNljZqe?;?9UxVX*W>fup^%&faT?dj09M8W`B z#L$DpEBd*IQ&pLqo?s{Z`_sw_4&JZP%JF#tMU6g&&B5bGboW==A6ay^(}sPBxTMgTJ_8N7 z3-=vgrMSRHT!siL3=gTib$%#Et9o|7;0^t9klOiJzx7WUyaw4%NRC7YR`HjywSA0E zSI&uesmuzM;nL3Vv7w~CXr+td>6P6Vqxg=NJ|!;DeqoUE4*s}2 zgJ6LhiT{|v)er9t(eB0{m7>aGPDUUpt*D}vguZI2Vc>2$GCwiRkcTs%gCNW7GzU}} zZKrK!iJDEV?>{*&>Hm|untS49&wgT&;+Lqp@~#*D{OJV4a!Z>^}UurwP)9u)n@%YB=)kZ#Cr^3rW{q#F$zx zy=jPMIA2VeeLbzVn=yaYZ`=3H@gekkoE{}K(r?FJ;!((=O(S3D9F)-Tg|)%+P|N%g zTfjC2nsxcF$MmLFL$Gp`uLRaH7n67_YM*%bUKTN47QDXSeO>9Pdy!H2LtIc|l~tR0XxgwZ^n6~Q+yj1L<-jHe zp=YYq2`SJ-7vz=kiQ)wAoBv7K6Azdo$x3R<Q zX8F|cs{2&U7FTxg=0O|1_i!rn-xc>T5lOTH!(y=3=rGbzOYZ=-CGiO83a&MoQKnt_ zit%!Jq`K9*nm^h4PPg0oEYE7pN!Wa!&G)aA@mpWGm9Y02mTjmUSDn2{b>`I<5*XCM5#Ed)=6$d z4L<+zs5<2_dK?_MIPyS$Uxt+~>Gl0v9=B=Vb5=M34uF1_mQRYG>vps_w1niKH`|x6 z+#13^Tdzf5Grsw`Q)!kc#_F6>7?98H&K4=URBS>h@|Ci?%jRcWwU|ETKe+in#`UnmBs?}6Q45<=hDv^T zT}s~}wPg9EG~~113?;HO%?l_qa`SX30y0~`Aa9$ z)bV{bclYgq5Zh$VwpN6%x2rpf1UL;OCBxl6#Jz!oK#;sRs~X?%PgB!RU$`JxYMefB z$_@2}aQ*cN><{cGKoXNOLik&TwoFUqU6a4)%!)0nJu%P=|?rF?xK zq++Y&Y_ylv{~x7$h*FE#l-%daQ!%CyU&pnMhAczH)b zdsLOkYWs;fAlFXIQIf&SY_JVHHf~CTb)aQlGLl3H9aQZ`4V>`vejo=BXLHnHURc)O z226Fl-UOD=2efG7T!)EvthIH0WZk8I_b+`Y6I;xM5rvKW-ejpN|E0RaxwL=_m(R;r zOY%$-1`oSQx}aPlgOBtMv&{zom{G9V#*;D5pH>TiJPWitbQ$q%LmjQBrs@lO?#tluRo6u9J7=jjah@=ln(_uqY~gbM z({kX~O8>P8G2fmZPAAH5lF{^v)Ef%LntUt7rlt707lj3maX0bK!-08X>PkHRMu+UWaobP;znPOabRq*pGb zL?Ec=iR|9$4I^BVgnSOL*+ObRj>zKezpOtRzwUbr!2 zTg^`$8MwBV-){|cFW+`;&7EdW0L`TQ5@}%O1JIW>!L{T1B=9;Of>zm_TC+`?n_|^k zazXf)0H|UrVg!yqrcHS>ZS=x{1kC-ugnj;oNSbaGfBV2fmgcitANtCa-TV`JvNCHF zaaN`brlnUys3b833tY0cr}GcuK-EyOUIb^di1~mcC6KFE3X@Lnw^H)?&Xm`yc&ahz zob(dlPXiWhnpoLs2^Wff?+#ZAv4N^_flC40*Y1ukJlOoI%a5<*b%5p%tPLO$9&+v&gGpA3SQhNOQ z0HCr`hDqDWOUm@e0K!az>R`cQYUN8d_i3prRTd?so`3H@t-?ei1mC4} z7{xAjMmwc~Y|mLv)rrZ=lJ3cz;O?ubD09e*011Cd3&1UfD-IJ@P5;(uz$$Y_o}nBt zCJ$%K{F`Qg{e*1VM0peOdouUSy~(^AFXhR!4l>jvTti3=d#s9ox_&NAP9$&rAZ$7a z<8p8HQSj6Gqjsq&aWyfDz)Rm+o~OlJIaQ0vp|qvH(vy_!?;3JPr_L<4Mmii{n50ri z->B;D4EAd3R}Oj9iFTUFCr+3amcLl%Xc&wCRJDW>nASB)JK8^LgnbnKkA*qijoFep z8zgFAj5laps`bK}X8luTV=v6AV{t(AN$nujmisB*1j(;X*3(4;u3-#M;*eaR4V}W* z4#sIXE$OQ7tK3>860*3c=q1KPUt3ow6hiHTQXBiJWthY9@N)odeV;_sLQ?T6N?1q; zm}rZMQLa7=-$%&{a`lgI|{HgU3u{MGs@dg`N`(}%0^3H{>x+?hG-{nFR4fj-b|j_Q~aj1fl1ORQyacI8c1 zaEWHCp@%K80?;$0+_CqE$sb<2EcwEgNFu2=^9=UO%9{9ftzW+IqR)J*#uJuk8_#|6 zHa{Z=I*8woB)8?5xozW4M$85lsJ8MX7G-{!e!bCxH~b#eKp77^6sUD? zp*PbSKBh7)YY6MjZVTlGoHB;&>98kfgO$>fep_F&X`gzc+8CtO3?G?!PKSh$Zn3%$O^yd8w43v$xh$Hvx=o(oe_aZwhBlZ2=-yd<#`!ycV$1D25n)BMmV33$d!Z^dHj0GrP z6X1wBus|I9cT<&S0<}AU9#JUR`K}urb8OnKH0|51V!J8w< zkhwmQUz*#SyGl&EoIp3VZ{y3Xw>17*-kn2@6XH3!g3@#@83F{;Wn3vgGCO1ha{qLY z%Z92G4fJ+cTDGz6#={-sCGR9;4H|FU`A9cT9u4{EW3caX5H9zf$fhya6x)e+KkCr4EJiAe6RSUgOx}C;g7n5s#Uy|M<>yMBon-r|<*;Czdc>GV11H9D6wlOsLVLl{SYP|&8W37tbHDz9e z6fsV$FkfIwBdmc9dRnJI{K?HVdd8Ql?P<2~#boH7-?M_Oy={D3V?@=B+FcU?xyk;E zX=j3$Lk?U8``yql=(%LU0OCk1VXi);mBKa4eKXtf`l6yVQx_RG4gF|%t{o+``s!{p zzhI4hsbhKzs#_y=@~xNSLfY+};>3-k@~RJ9CZk9LqGatp^4-YIH|#l#;JQ_;tJ_=M zs3)@P>3 zwd;QEhgMyg4#N;${%K%3D{_Qi^67rtvzFVh_7Xu?IT(8<`~Rt@I3|G$lg&4~8>FW4 zw8SLfLwm2WQ>tdCm3o4q?0#_6K)7T$n)~TH4}Sv<1S_^n+cbeuL4i%aqb$C*b371{ zjaOSO&S95k9}R|ru8liJpqDB%g#2u~1a$~CMV=pHFxHcp;OG@fymUPE9qBWQ=4^fC#cu($ zOH_CK<4gzZY7e8#LeNFDSjFv`-f_D#n4bvRGvr`H-FE2zXLTP(1I%xkSEx1FWv&C4 zDz+kB$z7(e;0BLp`uGH!{f0$s0lC?TyPl|P3U{wmaO_L-0L<|vLy;r z3bdi@?C}8)&7#a|zM8ZU?KCscfIxQ~xyiJJzb##*hFB#7vxhl(PL&_fZF2I7^(f$f z5<@N(;SZlioKElti+vPQV2iGc&X(8gPO)3qd2Y9m$MZrcxl3kkHYR5sWs)mxYK02i z<@(M1%OeZIw?AsC%5!UU3W?VhKNx*6mn4`Pv=DGq=`GD^{O@;+<*!!EsRjPi{nzFr zteeyJVIbicOCyG~zT>)rx0Tmq)&?)b+%X*>2!$?p+qscl?#rPMbN`>AC_t?s4Xj%3 z1f7&%IpMLM|{=42R_roz5^A@!}aVNYSqw*XS8gG6FkRN!kY1brT?Byna zmL@85eZIoF!U*!lm93n3a~XJ~gb+n8NE%K?KunI|5&QI@?w@IKl~&WfR7pJQZhr5R z@D>NJg^bFJU+rQQrlB-_TLE^r1uH2zilnxknkJ*y`bO1J7|kgZhkeH!vJG+u_5C=v;pZc8e? z)m8mUs`{21_QT^Bxn1_Sj1kl}Catj5wivqst(~*%nrEL+BKKP-4NxNG8CJTO_F(FmB0@q>Zscz3sa?LUWB<07gO z4OPSc*@SV~E!s#v{KDiBuZ~dpk&Q2r-cCg0PDjBs1pptQl0>}TMF!Gi)^mV<1wBz@ zt8R2s-j&Ha7GLAv_XcFoICQQ9O7s)d(ox*c9yPp??kNfKu0-5;EGgniThzIrQ^NEe zhmnBOh-Kp_aTVYOsi4K|(%MafLb;!*X|9k_CI{nwiP#j=4IIxHY5*#ArwziZ34w17 zwnXT0#jxLa!;N9qMD6*TSJMkO81nkC^`8uw#?uo*q(adB;~VCZT+~!)ZNy64`vPf9 zf(LKXKBnY@99duW^S7iU{+tjWoMBz?-VMw01lh%Nwkd|c-l!Tlo*Iu4~TqAlnL zYpoYK%bW^Z!3{%bK;$KC0Ihze2zyj5SW>cz!^)V5V$7)Ep>` zTX)TXnEOI@zp{eC9vePt;6I)e_qU~{g)fP~s_2?e#h|z@@fDVuf_D9!B?n#UboO^s zMx78~llwNp0HXIPM3I!fXD4CQ?l{e|UDHz13iuSJ^ysK$b`+v3b7N|Q`*0b;8cmn$ z5YAoX8~OP)pSEL0UM&d6A2ke*5t^2}x(nSlc=ctW44QEW9$HjK-auu ztIkt`_=2!LNBv6rUw<7QnicQo!5qDY8(BcdBUSBM>vyY9hBr$$jWn|RDs9Qy4YGiG zToJ?PMRs}3cyr~XTgIEq80&H9uKBRY(fp&^fVEawAK@Nqynz8uHm5o0zgRDfjF><8TDix$a^W>)5skvJ-a)L$OkjN5br}L;Sii}n*4Rsv1%9APy70}iwtfv z;<)71rf_v`7?qMmcIjS$9JG!yN>>~kRhIHzlc2hIkj;}rlJBCm zrhX*#Rpk$#qp`>JgGL0_hKUN$MmvD+aOG&}IUjG|MuXG}h;==QZR~5R2DWYa#SuQ< zZ?+%ghCe0|q1(DDQ;|@uwxst$Fs1IKA+XT#_m8||+AKhnfKXY|eDSy?U|iRlH8cc* zaUX+jn#;!zzLC_L!d;tM&0~+xbTqq)z67@`xD0Lx#>ChZ|LOtRD2#A(!M~=@xXd(h z-;<)uWD%ryP^*x<8ri7`Wqt!&&A%qEwi{fb=r{B;f@8IsA9>jI@Y4oIIlL6iJ|THq zaUIfQNLu^&^6y|13L*lbo3N!5MHh%q@ZQmydU34$^Fy9ztCs6Kx%S9WgzkF~izbJ& z(D93LWs$T+FIEgm8rPkn4be=Mk`%1J3!eyQLl|J@=StF!n~W#4*cai~T?~J?^Zu#D zQCpiOXv9fDa_<|B*7+Hy$5&VbDaR}d4^4M_w115wr!j5W#ZlwdVmR`q>*UKbu+dj5 zR=Xc-eT;^DH{bsEpY9_X9Wrgl-$j+_!dSZw=r2iP3ATYJ-cS>@G(AEjuMQFU3X8+5gF?{B`&y&LLDF*XKRb)Kwn zpzY_Mr-e1hD11YP3>rO<_c?65V|zAGF&})_cxmm4{t!=Y!!-jScv3%{BJ7N|Z9H*O zr@xIxfppj-RXDcqY5P-^HsSTB_uu61e8nfJ{c9gs+D4up^2s$;<_}X|+r*j}cTQ@( zY&%HEtyyFWsy1?q+23{=>i;G?bkdaF=Hg8@>ov(bWFjjc-%IJSy58cEic=&5xqsZm zm#aCt2x6rcmU2Vf2c*Pto(BzkUu+-sgEop>m&dSLZN0GVT{Gn_BZPX;JwTZ95ikR; zp{w@B(n9QA!8ZbAGbi=Wtae7+n^kFk7w@&w?l;zxMTM>dlZ(XNePadYQE6~GCIHmI z;1SfkhveHQ&JANQE5w0|~bmj9PAjH*WVe3DZ+; zUQGdKT>=7LqUfV=p4EEX8){}M-`bzu%$JYMR|#IL#@Wc7Zv$k#v1RB z&~lCWV~M6 zZ?2^6Dg98K&d#R6ip_iBoM{?h>6KlhX+$p=`pAQr#HDb0Ob^P zQQPqAoR-&G>0GRJm$Q`hX76_SwMt&tPxCP8`i`%h=V>QA`c`ds3i7oH#IFF-(=ioQ$DAlGGi6`8Z$En7j4 zN^c~Y4ey#V^%DZ<1SeK;j2yh9M*Da1q0;5P+KD~azLApsn1^_{k8WA*dG=MQr+4-| zo_LT?^(H90E+AWaBzI0ZhqoW|P_(bbdFBM(_BwwH6RiJo`y_V3@H@lDK{T%Zs)4)J zuy*)!-tWRL^>pcxo2EvixeKIsL8N!m0W&Nu8ubbh3SLFK$4%)SO)YnJf8+NG2lN^w z-b#~pVrC3u5T}g+W(V({)aG-#)_8;DjLyF46uZ-4|3DuCy?6VwjxVOFuu-S}BCTfT&ewXM;@2#hnlxv$*Z~g5o5WUQ?S@!M z5%RfpUc}7~`!orkIvG`?&-}!cb{8=_lilxm54ohl=XSUJ#4VEfjG>#Di@jLmYM0iR zL)!BoG)os%kf&-YkLMwdb7|95ya}>3p@sb1{;p`h zUX&n5Oj_`f05uieQd`UW?r|uA<*iNH^xPR{iA=NLDMU>fbntg!<;6mbW_lj$(PceK z+P1r`e7D$y1~Du61}SK!-n$yznf_7%S+x`Mc_bKXLQl@#w&F$Q(DxJX16o_e3+oZ`{ZMhfQJD-GwDQd7Zds}pkKhp;z%1Sr2>rcc^W$iO zHD>bAhRk{G6Bf4@X8^i^?&itVik5Ig>4_5lzVFE68-b`M6zmf0Lucjax? z8hwXCg8Mu5x{j6A9F2`G-q&hw`|nK*sQiWreW*2FyxKM1>nS6nHrrp-cK==hEsMNJ zh;EHUR*Se>rjFlD7gDyU`%eagmP8tW$YFG&gK|i$uLv;0A&m~VUKD9u96h{itrR8j z*gnR_A63$qjeqf~WKYwI>!u-W<)k`DOb5jJhxFKz$k0KesNVb6tN+#`_)8RXdecMy z$7aD~!FcH%mBh}z->|A~x-3Z21m=28qxUI)x6XI`WxBO~q#$bQwAw9bcmUE56+gf- z$R*qHM1~W87#d5WMEFBwlW?(?eBA&;6g6+j=+bjS#)PimAjEgl-QPsQg{G7WgF~h3 zH3z)=s)PX)VMj*N0{Rf&Nnqvc+%)Az57GawaIX#;nmz1)0kb-{!oWZd>(=v$*`Q@)`zJLK> z;0cUX{tCcAL_9r_L#VCJDUrkVL+)gWq1EqABSKB_ykS*zM~6&Ag67;^u0F4% zA?3;-_ip1N5%DF^5ne|A(Vk_*>I=BMk=XO3Aq_IGW?inyRr$R|J*-7cz99`mE9Bhf z{vulqlqhbUW3Y17cu&Ul)ti7TQftQQk6~4>0}~jLq{?vf_fZ=y1M-P9i#E6frA2EA z(?40I-MO>B7{DMbgCK|YSP|T!da~Cb+r;H5$t~{Fr18-Awhc|R?mROCJa)aLpr3aL z+(eL_D+-;3-@tio;J&_9(5LO=4>r`f-Crh_%6BO<+413Pu!ek&bJ!}LJANnkr0Rin zA|pqG4A&fWzF&MSQNLsMC^@28c3J5=g_T{=>N=CWE|`y&utc+FL|am*;)@bxgpA6p zbimat%Ryc#x%<{nOlTuzH?iOC$b7#n_7&h3ZY5K3dVxI!66WxJvlAC_{~>_RC`i=%4CWu z3SXL5CLpFAl?c_8?qjRCor2HnzUx^!m{+Zi&f;*4ytxViq0$jthB_Ncs{jP^zE$dZ zEAuEFhpC-lR7 zlA5*bbXTi;rEft}sDhC)j=W?;b5S;}HZEp5Tx4IZN?pyqF;5~8`jULL48(50RDqX1 z7oY;+dPkWAS?mX#Cjj8DIu>1>eAjC|@7EM9ukQAkrk}oNyOchT%a|aGpW3^-7~q!v zPw#m!BWf((k8v@zZ~N{E4BF>69X0yw60$PRgb^fQxAq z2uvM|_XULW4VAS=JjwaQ@Q>4s{(h%+J~TW1v|M-GwT(V}5hlsI@)fHiY-a7KQz4Gc zhi_OXFM8o{Au{F6CUao^*{6tSc_q(+#0J^zIs8Yvre8h+scPk?Mc11^XCEaUzak&0 znB45;hhC}wat?~*3Vp-E)EilMH-yO6W*6HcsgCti;Pt?%1fRc1l>M3t?RS(qH$8jX zzO1x{>Le-^o=7g`cTh`5o9w}(kg}Ga0G>8$AJzOpZXmX7a{5ZxSbRNULW8p|AXMt{ z4O{KOWhtI*SAjEY_o_$FEw5;F4g-=EU9ND<1b+wu6O}Ajwu(NDWI44AQ?dtQs#1Ox zED7dp1$Cax_I#^y-0Udutz~#?2s|}#o&^vs2m!mI48CAK({;PT^34k~-JshQnb>Ya zbALIY8Hc+u&??7IXM^CqA=xFjqqZ?&HQ|Lgp(_dq;i%blSTu1~O{Fg#Y&{Xa=2zr+ zEj;uo&=Mxr^JmovbLV*Au)#}yj~t2*7y8gGotF<$4cbu;& zs=dqUsebJcxZPP|VjUOw*^SIyc2YAHG3K`w)nF1hazHaj7>j+>l&@UC<&R|LaNy9+ zh&NcR#rPrqGHK_>wZyAzcQKJ~neZs~Hu}K_w1BRDK)S_=MXb4K|DeSW`^DJ^2rkKd z7!*-d-A~i z(W3uVLq>i1e;zeu^C1D~kDScm#cmcfN7%pqtL^;L!X{^kiFv=4n_m;517W=Vn%j5u z1;Wmr#J_BR0_7XAl`hz>J`EknGskbF%S}PIN+vN>u_&AQ7LCSwZgXzeVwBI1-SM;_JNtFc+AZ@i z#W?zmAb`}?lseZs3(1~s+H(E zqVxq1${Pv#l|Q9X(5b)idps!YmvOg)3B4_%Q!^wWCxjBAgxE43eFR?0xyskIX&y2D zc~sNbhN^Z>K2H<(wo0Yr{-d4$`PvzO-5wmEv0j~?ehqIFBL^+M?x9|o7%{J@TYtl{ z>^#ICOYZghC?iRK)p(2T9(IL6I4aY5ra7C-aWV1w_cz|mT6u3U*s^A&h1&NeDp9~ z!{BCjt>XEU&mcKzli85A!nEkH)8`5%iaGYVMX%6PLJW#f=DJCJYNUL~wv^t&;P zz#3vC?CAY3`kzgqAPu2YA!^)d?45X{vby#ES`nmmi}mfhHR9F8m+mkXCLIODx}48 zyr`(Po^Sj88gu$IENAwVOL0`SywTd=;CDv-Bb%-{QQ?qp7N7&OE)$-kNuH2TdX9W; zlQ&l`K+^($uaEp?D&o=lyS0_0DKGs$K$I8%l45}UYn8@m3rqrEabEN*Zs2y?LDAtT z^5eG)>!8jzUhX@nd)Ce4_hxGD`9;K}n+tw{3~`+1@o)ZasU;X3$tj?;%yVg)jg(z& zhsA|QdKW$kGY2+){Oq=WrF4ZwKHzhBDZ|#hQo2M(P)Y^%@k@KJld8Rpa`CZyesnn~ z3oj6tx;6fWrT$}JC+*W8*>e%);&=wMXi?iE!a%u6~o^;(4`$BUY6Xn~FkB@6rJl(8l!4Dcgm$@qq@@}O( z3I35gsf^OFW?`#NSttsmEIzE1LI~hiUpJULeb30jy;5cGDr2p2Mq#H8 z2}uCX%>gk$#jOxDRyN~i;pnJ)OIJFr!n=R@<%|{LUJP*?U3IO79HbZZiRj4=X=NW2 zY}2#i^^7O-iXz(az+Z*bXa~0_Z{!fk(s#r5Op>CnHu6%ZIC?#+0r0pwVvHR479d1M zU##?P-_d-h)kBeBRiCN=PqW^TA$~+3KEDU05ZF2OZXI8Sy%7Vs7b1Jzsq>J#Q2;s` z7D}Vx#g4YWaiN&#W%-b0Gk;B>4Cgdk5y~?wA6L3&JNrmCB?~&lUF@KLliX#bl$zjb z0j}9~Mj+l*-;rPWt2@ZitM}K0-i;-KJ5={lMza4;7cp)|VTz>E{xK9Hh*dC$X?fyk zPVZD`O;kb2+noEN&uIj8?+s!~X$5>?+n-DC!A)YN*OkvS5iFUY?7`9r$~KqXgNb4}=f)_c`OhS}vW$Q4Ehl%wNBsKUYdEtjr~BEvgR*d{vpi9fvn)5DgLj4XqzzYY zc=OxBi52BNeP{thZ651SQ1)68wfQ(&k;&F=#drn&N3ysUZvSjl$kWi+B^C0 zGqWEPh{L}-$X$-4Telo&!ebT(fQoxmz3DByEAN3`P_C=xmgpKHW3DxEe=rHmK0{Eh znrOrzGdwTyEKMk&h$_mVJo@e%x<;V;-E^-F0-dMf5z#*(B55AWEcCJn&Zu%^xEf4; zbC3hAf=J>H;6I+grD?O+jfC=;MV~Oh*`z!esm79f1zECEqv}aw*~`MIvOfSP`zdDT zePP4ZG@a=Quv%ZF3;k|VAAt>v-eDpn^y*Z33F%NKTtB?_5pJMof1&Od!qePCLF^3= zZ7~2E6HPFw| z3@+kOn`X=S3j9>Vuy{gP(72{WQ;eG`Fln4@@jWM+n2DCJ%GYfFH&_5%g zhQWvB*i|;EexXE5m6NOF>Ojk01tad=daQ*|+vEqb5}UEkdsRuybBVVh3 zEPP@|+XIGZT()V%ZxAVm*^y;YH{AzBqDz^<6N4>)q!$1>4NImFAQ-p)WTftG0lCNzZ}kP=MV)&>45i|}|XU;?JM{)0XlV5xY=rVW1Z8A1FqF?~O{ zq&6nmP+LYfdl>q=>XLXHD<$2sokFCW$-Cm!&_l2e%SARok7UR!f)sIH{N;8RHtT$J8 z7eq?T;(j;IL!~Vogf(9rFz`zc=B|5}thbW5T0DW6kpcu|&%>fuzGbtG17#c=*@P(J z?f1=mHn7gqB6pNk!S_=NIu;viwR?dHEr<4LoDPJEFA*8`9<#b-lc>Ql3jOzmq@qnF z4u-85@|}D>Tg{OJ*WMpZsWAz-XA-{`($sckwqfBxWP1WtIMqM@8Lw3>X)0ZXodq_4m|(KFkv{V!bsdkhw5UMKsIUSEi)HS*tmR zn0`}i$N(XK>rUUq7!v0g8ou&N>VkC|d%<}kl$SBGB0PvQRQs2AmG)pMD|_|*4S@Y5 z9WZ3PH)>N77uFPb*2Qtb^-jy}VM$a4&6{(N*R|fGzUlNw$wm!Q8mq`xBSX(%`a;9ei%vDMZlUO z&tl2;CrsDpeg6LZ$rksInJ6erC7cMPeMHpkngv+-&f9AkpKpF5_vvj4$KkRWzarDh zW|V#oq2AMhm1k?|eZW-#{-JMSYZX)d2R;Q#X(=(t&o@3P?|ivI7fqa0KHkXq{Lp=o z8?+;SZ5wl*d~|bFFBZaEU;4pTA~Lj#TFZLkjx2^&IOlWAUdcUZF5E0q!%8!ex8Xwz z*r^w{b;!KqN?|Y{#+j97KBMQ4<+J9Q@dqSa&l^*C;h>a2D-*Oz4xyKdgl+dXcK@gYh-k}2HJP5>tA6btBsT?!UG|IT3=`T2ckI_K-?G4-$3Ka>cvKdt-r9;!vjwy_^moxYYdL^`VUCoIUF z;>JED2T7hP&vZ4XPnkTZ6Z+5r0N0+>973{SJpQsKH!1OxW;xdc+~l;{<1#QmpJXZG zeOVQUtKL{I6aifs%?)dKF1R`rRF1h5x%!%v`XWwie>f}rz?$&UZ5uppjOX|@Rk~wr z;Q?_?ls69%X3J|#Wg-e7&yjJAGOK8tWh=@7sqXt=>%!YC7-RKXregrb&}0*9;NKRh zh|~%`_+2bm)eC;PqOCmeHhB@t5p%+U(PMa@4!C=SOW;{Y;DmJO2s}l>>&G)iSqzNu zGa`2vUAfTl$#yz=z9K8GkE)|}K9w!_!Nm;9?W?s(4Wv0Mx=IhXjbw>Bw|O1OEq1}= zWogU`*R1<=jT=Bs<2?PWVkEh3cirxI)0|Cw6!|u-LUaYOyaePP+DoRb+_nCOHPB^F zFBwmdkFja`9lugR!BcGa^i1e_Humn5H5FZesor=t#KOZ7yvGJMMfpV&$c`wfd0wBu zdn9-A6Rh6Tx>)KTEL)cQRx5wfL1@%8CzQc4*hteKpMU+c)L^6=#~QNZi1f5jGYkxF zrw=0pUL>?l7QJ*$gzgVW^TD=F{-N&XFb&AEGKq(+7sbR*o?ew2Gc+!{^>m02s4->z zrFqb03{YE=c<)yu2XRU2+1=(aBZ0XDEBFQ5ZTO8r4x_6 zlct+~_azF`B$e&+Su7GhwB&|*f6bIi@e}WeL17Grw)}d5?|U`4t|n6xA?{0LC*PlF zQ`j)k(Z5-8{oV*p07%M8nnfj@RLCwR!-ll)^zvHH(dS-R^CPbTTP~#+1_24i)5PWPznb+4*ak<7 zXG4&-&AXOD$EC>hlzt#4mHAF;4*A-4#==2L_HjAY3BTH-s()5o=P;NS#pQKEyf@NI z^%nihtW9N85bspd_zFcd+FV~cE{C`r5~nuS>EEyQ!O5ZvyD7X{YbZH zn;to4(b-U8eeNKupEC1~f73DBPxRUF^JypF%leG>zXW_&x)O~m&L(lljt9WU_eLtB zDqaL-b5l>YOoQ&u`~MUi$2U3tm+AGOS(;q*p|dAiEL*(rh0k7BQ`|QVwrc8Rf&U!- zuoF^r_<%c3``skJ$q&06(OEf^-Aa7`@bD%mZ7XCnWL=?9<6`>1F5lucjiT@`p-v^Q zn5ZV18@zSJ#h{2LtxAa-e&7~pLxz30v?&{<8RS;f?bGz#>1uL>bm?27Sl75bHU(o> z9BFlKqg13&YpSf|Q)wU2>>`y>V)Z9^wt-Vddx`HDOzNo|+s?>HhMhKKsD4ttzV_c} z-sWU`C_LA(;^QYO!tS5_GfITQW#sRf7^NXpp^3Ip`tCS?U@2bOD_Pwh{Pej->IT7G zDpCXXwM5$SK23V!O0cG1#I86yKHtL_aZ6fTck?c{khiXz<|g!ZwqmYmTXdSh-u;){HwiMLUUamwo_0u zz%{l#J0cl^*{XIBY$;lkZ)ft1<|FrsA2~P#{`%ep?hW{E*8Tg%ln~|Q)m+w-p9#T* zoAxtBfg7PSn>>AcbJd;SdFQF*Lp4?bE^Rx~IGKV!8LRNt%RlA60Z8nbu%-{G0Onfr zU8NiIzdiENcW4Z5u5H!!mNnakvu|Erdx&ta0Irxlabexc&B_W{AtgG^rjTAQA}0#@ zT(yh$*-sqi6k?7RBdUW$aGpiMROE}D%)MCBF09}g3Av|>Rfw4iY)HE7uYD;T;OV34Y%rvwzshic+FqTZxr1BDVJd8YJzcv?Or%Bx{Y6As2Tn**k*pCmgLY~nyO2>om^ zdD%foQEmHw_hGkjhw=dvUL9D=LE}N@O;U5XxtP?ysJzVb(fig{Zv4tCTWv%?$xV(T zuEF~#rF~N`8Ikj!Q-vEO^ZL7aj+Kd0o$6eZUV&W*z?k|5K~~C^H;_w#QTotm9L`|) zF&noMrj=LO)&=o|8`HCEnpCCE_nQ#8JJ0X>fd9040tNrp-Fo8wWm&|Q_nw|FlqEKS z`}=Z_Kwj8S1Da(Jm*)Qf7bmxrW1rgBd)6r3`E%EVPewlGT9%BSjZo+!R`s-JDV|Y;W^}x+S)YLuszpJ4}4R^DU@5~Lq%3ignoDNmH0oW}B zQ=zknW}>Bq0gzOjR8pP~S1Za(!lSWWN;7JDIGof8+W#e=%&cL&QAkX|I11) zftjbngF1cSUO<*g(y|6Af9cbLrDzJRg;v?rrg|BaI8$z&+u6ub?Vm6+u7AclbvT6R zdUg~}J>mcm$T42W1E8n1wCuWrRORxtJ%J($dBUkTg?iX`0^nkRUDKxm#2) zoGVJsO*vsKf*X_RHZAT$De0xA0iuC)+4DtL)SayonaMx>6+T5)%WI_a5z+sl9%w`8 zf3Fw07$cBVkX}PW4uTYj3w<_CuqRw0SJa2j0A!QF&05{lqoqn6^S# zYE01`m67khOmJAy z42^~O8Iq?^+omBf>T*Y(8J9f{KuECm*{%3YD%Etqy_iV&z3O)b5T(R0LFIQW6I^YM zHqC~KaTgOW#w>0=<9W-I;=_AElIn+nFwv|Q77f%!>2mgBl7}ROIwZ+&(53?nV5K;K z7PToT4pH#4uN_?CXcwi&wgu6Nb1~4jyEAR%7DY)u2;}fL;02wWl%vysb1UR!ea#-z zHVvWPqkt@>biOmx%HuN4|7XLnqpJRV3EN560Eh%*XhbmgrA6agQM&GfjJ`js@xJ|w z^3rrJk1S1&FQyhuwv508TE+LbmI+W+9Qmknw(*f{I+YZT?%&;#L64Aojgl)CNckUi zf&Q*G=rd9+kh{E-N*}WN3A7MRnF6_(23VKn096zpnAjuRwi`{*O&z5Q@jr+b!6)3z z8QzkYepl{?u*YxC4Jt0ZaIlGqvHtIkJ4?#KNu>`ngMDI5F?9-CD>iBx2t0GsOnj2%bE&{5Cp+fl%i*Mwjl@?=O<6?y8%1PX%8Wh$@`x z1_aE+L4Li+$=!nMYWs^3tHd98uM2+*jL{e%x8Wq>uA3)`>ADNu%%|FJpiZ? zb75)0r`=#VrCfho?V7}=O#u%`&w%B^E9nH z{obyEa(9$39%#_GS~Nyg+kBJbzuWy%_h;dX9eYJA5AQwn?4|xe#77IC$NgGgrN|Lc zO}Az=q8qnf=bQ#{yX1%_IXCVkTWGCkU4FJ6jvviK7}cI#BXR`EIJ#&>3imknz8$#` z=;WbFt1pm4oW=)3jX3!|wwR{Go-hyJUAbcHr=rSplRRt9#c(3khk9Z+cBo#HGT5KQ z$E%Er24Ut*e{T4H@e_oQivG%L#!QvyZb?%EQP8RzMtN}#hW~~=07L!jJ)&dko;=4! zpSmMT4fB(mMik&`$no1p5OK;y5e74~l*TFehlKTW0>zMGPhUN2e&n5xq^5p>H%k^^ zRRWBfNtLimsd24id}pQlQ^ODg&;~(jm{&=N|3wojp{*yJy5Fp*?gc~$F&%(^&gZ!o zQzRQ>kR&yOi+gk8BmJ#3+xJAkJb_r*Z~p|&K3#So*^L0~es{vNZKc5z8V~Q}m$`%* zTK=PM=Wc&owM@TLe~2zR*h_w%Q~ zrF&nj)3ej_XWj0=Wv&#`>dk4^6s$_+6=b3IUBk)u1c`q11;2xvRKJT0cO%E1u61qP zl$+=w^)a14B`@J@oeLZtX*-RCKGD;Z>YLKc zqRy=)JZ1rhRNrx8493$uD>NI2?cMxUQi0l@$V&6OZmG9^#tA$ta?OV*pTg*^FM6+9S!f^;M zOm1*=%{q!-5yZ;kg!wWu(=K(qEiK_a57>%oDvtA8We)E}AQ$7>J+5+??k7k^;acM# zGppfLHPw)r#!Qlu?_<)`eILLz9RxzE1%E<^$XmXWu;X2mOZ)w$R;3_kmoj|UDZwf? zWt<=)!V=QA7ga?YxbnNwOBNe5`8jd%TC$p@i@n>OsmVnIUc*aeI2+FP7QzK#g>1cn~3P>XO(~S4#CWM|JQiOR2ktFj|2pnJE<1?E+17icM>O*i3C>}nMZdEI>$XK@ zEMOXT6vw#4wA&!nPzH2C?$i5+gQh~W4SBtPu83#Bn4)Cu3X7lhex3O_KSQ{mB>0i7 ztfZu#)T7vBQ8(~s?L2QKd*WN>{P&(WX(HxzAN`u{{W5R;=+pP38GUA@7V@31t9y2H zM5t?9ufcv+Q#w^1ZS!K!5K8r}OrihFDM#w))*5dlSv5n+1pbE9@{duOKL(#S(1}Ur zLjH-U6&>EW7N<59c0Jmb-xob?^FM&-J=1VrqTrF`co6Hs3rdE)5R-eh3un_P!5{>I z($a{H7W3>;$V6Ew{AfjG`d@$tv{_#%Lze=epU@}gjEmngwi7s}!~6=pJF*7VUls_L znEl#vJyAvbPMK{WaUhTC3vB%ufTIb%we52Rp zfkEJQLx~YagyK)gP_1p+iVyFr`*IrvlXtDQ4EOVXbce`9ZDcbl7^>OoaH}Xzk}5rP z&rQT04IlkCt~hkQWOB^I=V5roD`lIMjY-Oq|2|Z?kf`%d#(89QF?fd&!5AMy68zUI zW5zB3A`||bwaV(-sL}LOne`O~SHbs9Lug*?Y1RUq(Kw-%<1zr#O7p(0n`9=9@w_6o zt5Rxo^UMmA8;C@JnDRTMT898nRXDw5akB|44XFN(z z`YV-Zt@qQUlWNvevt)@DaCnUBr+v}TZcGO*6QBIB0-yoEGtldte50rRJa~O*=$K$@ z2S~}(htGcE(!GqA(Y?~+l+UySu_?d#vAFfxXMMigIp{$)eNoO@kwXqM+wo?zmgKtV zElui44=K3KF~yElt<=sSj(O>}3<1&2+70H{e)NyL-bZ!btIK_97EaX%@9_<n!L{!{nd>;WWg{TG=ffwPrl-D!a|yrpWM3+ucc_HgO^x-WIh;MBl}VBE z(ABG+Z!(W!EpF+nX>^68WhFyMFT=%_s@PS{s+v_k9$h~k_pg_|E+Kyd|JVOZYsW97 zh>hazt!m}*n;nXA9dBhGNjYf_03(cThqPW+=V*YEK>kJ27+&kLR`kUvGG8H z<5Nd6Uu<%8rIaxu9~E>tHNu*-(0j`c&00dRVQ;T)4V^79RW}wFaw3SFGi``WpSI<* zUvgE3qDSotA?T9G5;8~LKI9hBt=2C`30 zhJg(-E&bNHJBHHdt|rZ`}@$I5Ru!S6dt-C=)W%AO-346m&W17GY zzgJ9yCYPHgQijyVgLIkWM+aZ!cpeE;O1Yh$ZXM|f3g0(0@x)Hs3H37bWzqFtN+%`u zgCNzc3=)swew8U%jHu;arYD_0lBX=wiRxJzSoPQB4 z_2E#R#g{kkc=RUZTf-g6YBNh~J33-Q?@w`04DVH~Hg=95ODRespMO)2KOTbqg$o+}kjh%CML_>s}J)gNMKQEqrPILc4085=ZMEepV zglCeV#%Xfgc0)!x51QvREDafC&FQn$z(-9>B?HCD{Ac?CSnwjdC&s_SL_b^~pACv$ zy9V!Hh1BeA;R6)VR0ChTHmrGEsO83R3zv1%d_mvv8d7vB40R+TU3dXskFGKg5mTG| zISA?P|0Xt8D~?*peXWG+G(42tkhc4pBl1lGO@DK_rL`oS8m5n$dwuI0*6pa#hCe0# z^KafyPhK|i8qeLtFIV*YLARWkKWx0aAU|1bgV?f01*<9~B)X2){3dO)`?p+XKkTnm zIWH64kY%#_fZTv!Ki^xfR4|N~D^`MG8bdydsxP&LRFSjD(fMN; zjh4HWAB)b)DKEEUq1&ANL;5aJIIfAHG`81Oy%ANT=O}{exkYeK=Rz>%UCzXe7>>Mw*1RHSSZ@i#u zl-G(M{0|_M;Pk5Lmh9-;;Y$#OFjboFe}e1cUq>wsKeI`lyTVGDbv}HGBF)tu-4{ym zXl>=^%onnh={t~+c_6YhMSF4t3yLb(UjB|+BdRJCu&4I#SkD$5{;Wo9Z+ShHGpax4Y)~CC_V4x%Us7Sk_?beC%O% zTP0~a$Gt8s?`}N7jWLXlAWDi%Z%RI{C*i95>V=CB+h*eS?&xYxEk*C#v5lF#Y;e6e zhWc^%#~lruHsnWao35^|*_XDVj;d2-HIrd?hcI-1U0npfD6<04G)9C~*XH8zwxsBV z;zvEAlCkBx!lm;EgU}a5SG8rTk^6ZHJ9yvJk=H}65F`I{^zjzmgt*)*OcemXhHXAhystv+RJ}2;2 z8A%4n>@ZtMf+;$s%p7g)r7cAs6hUU1Q$UgBGKUL=y~CoVk`>w6ZmBo=<)`u=uV31Z zVMde(DJ){A;9?s^QfBz<@T-dR6!rHH?sh9xFK*niOfY=&NAK}{+UXmylZ6j4qxDnP zh*=cNL5Dh zT%PMWW%&u$n-8M8J{ty(1v?vgc^#&izr(Rzj!$~{8YE-@SFNfCDvl0N!*fDa=iK3 zlx8?_Bc!r>-N_^WT?)_`QKxelyEL<;7b}(D_;uke&_i1#B5gF#WDtXzdcB+^o$u)P z4qeIB1y#80;I!0ykebH>*vU|`j^=W-S{9edk$z4*lDIqs45H%)uH%kb%^OsnRo+A4 zyY2iYtj~tCqA*fpieJRoQ5#nGnwUS1Z>!whV^cjdLCHsl)uJK}R4rH_K`5UgtT`7J z=g`qri&@HgI8HKWO9yyHRXvp$f{;bgtsr80uCut=#+lg5_vN1oDZB#koVazxx%_;P z&+8jvfWCcqccR-`r~7A2)*omiD*a+IYon>_MQ;`4iBP48+*^K_ip#GCNI~9 z#V4YQOwV4%gCDsBy2X`|08vI9j zS@>^tb-WAY%-F=xZNt558{*|3B}y(zBomca9bqeId_~#MKh^%Y!h(C4@nGY@)Z7wR zXLL5Iq+a`>5=60yGP=ahNVe#FVAOoxs8A>Q)`xpO!YLC%t%xgk(Sz*1`wvyDJTq{& z%GUckCFAjNfSbO1cFDs^IOO-Xqtf~_>0cf_@bvh!;go^H7hfhhO}8)65REI_=+&TH z7BU5}ue{EjV4KXJ*fno6n~d)=K1PQCv&bBI$*0v8G^N7|qUijAmEMKux8U{8FbiGj z+s*UbcG%|I-JwgBEa`CWoiztlr{sBJ zXWsvl`%tr+*!i!!RqcUy5*|9c>8~FZt$&Rcmsy+`pv7kU!N%0JT(1RWu{BqSCN{qq zysVJn6S`<2b4W5Ad$g!i_;N`{n=Kg&ThCu9xgRfKuBJQy=aOsm$+dXMQ`IWzz}ENP zVfUr#^WaR~N1;eW1n{s?;cJBIpxoNa_9x}Oi*p}ZAYG83Ywp8OPmL?r-`4+pcGq80 zI$$sY-CfgScA%n#IxuqEr)mznih71X=J#$VFGxSWFmBf}OA1kdS=d_sqU=!v{NR-6_{~jaj*7a|JB*eIM8%^|#YAx8L`cIQ0yp8iFS8Dn-QJeE&|@X|Wsz z<5yT>P((CotgNjM zs4Jk_ox)6Pr&Dps_T^)Yz$Y}N{GAO`6JY{t56EFjhBh=KZeqfM>f6c)G~0v~eB|}P z>{<+gr!|xK_+9Pm!ntT40#qKj`^Zx&gE>ZLY}6JnGA){>2H~cRN;=^a?RH16u8%jW z-H2ejFiE!(h^7piXsGy_$wT`jYZUCI(+#nvmEC$Nn!}We<|t3PTMkLtowcoR%f;ul z%>1ILlJ1?M>(nqeNe=jZrRk=^wZ4;S{(Knj+Mb9)dzxfT-)iD>M zu6Wg~OD5*CN4j=ZkHvz^zFu3tk$ukh_@#%O((mWX!*R8h->K%`PptzQ$ZWovVjvf5 z{eDS13^Wod>0*^5xLNudv(Mpak!E68;^R?JtMv{i!Rt>(N1tO;x@|-Dx@Feh)75gJ ztf(!$pRL|nq79u7IULlS_goTz{;YPKuc9LLC*+?|qMdcsB=e~kR&89T^4+P5b_J3q zbrJ^2!bMM(nKU0a?rLtK`vdQzcHA5SwR{?u-JUuX$-yY=){dGM=Uy3KiaT-c?#@-7 zL6w-==v^wm=NZy~y*@&LzAMC)_i;<>>5nw=X1gp)hR3;FJ5MN09hE2QkJXc_bHkka znKY;XgH@#)!Jy_>m@p(PNJxwemDN}uC3vf$`@>D$XH{t_Dz^o~v+9Gb@OT4O+pFXJ;Z;u3F}VO5Kt1Z!yK;CdZP)-|xqJKldi^ zo{rr+?<3rx-HOlkQgNefAfo6d(8B+{Rv?6p)2ZYDTe}MJQo0Hqr!r&vnebk2D-;% zR2-S-^U%+P2Q5Zg=w#f~i=hgQ%*mL#95o!t zABT-6Jv}4T33Ab`3To5@OPqt%20>yS{Ywh{B3|L%w=$PTw;@k&778Y+78BPkXP$c} z*2vw^vs}p)gAB&&gng;VO8%@JG)W2CeAGun+~xx2!sp^vib3&3qCUg{Z4VXocKLzv zDPDQwp~kfnrxmiw$B;7?ma8w{HFBTA#&pL84b?%CH;BfCjiI{}sda|NT>Fc&U=T%Q zf=3|ZeLotLXk)bdygQKdQ}Hbqu|Ct`1Cfma@yNF&ti2(E)EW(d2+Blk_)gFn4p2!& zK()(_2|C~U_n$u;0H&w7CypMZt+snQ)Cw#%ld)vh|>*q6as-e}j9E{s#~X!XHsLo}~*&J$VdxAd;Erj|vW3 z3v2UCT(5j+y5_!LsqlBPLerVkPv(@G+jwUvw=vG0Z6_n4=A(;HY8MR51*{Cx;{NmU zc@~*XV&s5cl5`>0dj5*MFzA8K2&cIf3jW>7ZquaPFk2J;!;>{Hhm#KsHT2Q(WTEK* zO>9ZLxen7a9d*+`va3NUDgrbFClGW(572?N|2WR<>6fhkObOG+AGImo^5K;Tldn%m zz8kAf*w8^Xg(HVyV5e}CnrW}-15 zftGPnaw~#DcdwMel3YIF2Si)k@{PYk>hIQwT&dNzSa}X&%g`jE5voF7s{pZz4rK#t zMRSMoBj2#vp5YnM6LP$FZ7+A39SjS(U zEyEy^JilkfF$Cx*d-Kd+)`<)?;5&T5s3v>O=zd^NS=KSC0;4;AC9UKM=fq zy#D32&PWA;ojY~`np_c+=2q$nKX&&uFF)}T!VME$wY0$uF{I33S(3`v?xl4qn>OGayW-l*(FN-=Dv^>YjJgLO9w_$BG z;il`rpidcg>fJ9yaUMnm*FJPIGs5xaiiXe2%rRdIk{|fDJl2#)`km) zr-xNDNoC4JTp-0%355MRnPVt*5%Tb%WXARaAj$q!wWy5lRX~&w6iF)kT4JG z9;;VBXjWzi`Nla8aIXt9st5YjaNTzQ?i%_Y)fuF)PdjGoqdm-KFhe|->E18nIz4^@E~hM!DLicg?XiGjmjbqN8olhA)I9s3ULB+)s7P8m2hhRkCDfmVz_9MNjHak{V=HWFHx2x1`_(PryI1k2> z3Um$qSA5{k$cP-HAAa*yPiGq^_Ix8e?im)L1^bS_>i^Tv3HyV-Eu@e(bKj?(a#VY= zQfX9MJ!n{myt&oCs(J;G+z%XoolEJ^p>rA%q_o$pMNf#oE4NfJ0=YBdKHx>RqpH{H zGiUtSKs;8*#%aQfDdkwR3sN6rhVRrv^Hk0cgB+Mwq_JmxVq&!WuAFcaC489EG;Yix zC3BWTGhIShusCfai5VH?U1r`+Y_A*&`-D+}i^$&Fi!MA+s8rstd|*dfhBk`I>j zrS$g4E?e+p-Z);m=hhN>HGqd{oBLLpz76pdWa>-7v{&1K+(!$zFfAJ$*=Gc%oGGcT zAEMNv=-6n*1OOA5?CRIdJ{$80zw7B* z!X?x!h6P`sIrCA$pl2Y_6Kf`&tAT97S#kFn2pf<+damlK4>mA|`dZc=QVlIV;XAy? zRsYi>$|bv!FP7k z%Ss2f)D*h;=Wgc@QFrbuM-Bt=CyVy<`bY`C^aX`TuaS&>Y0*=Sm`Bg`y-zz5>LN~@ z8;bS6^I=--A`S6@&asv}!e3F@+SE9w_Ccd~gLG^}q}ShNxtgshE&W8*jC!a;6t3Ad zp67M*Ju$v_UzkwMUQKinv*B@fXDVF-?B%>3hPVJbyRRBizLfAT#gC7h*O;bUZHu!omSn z2j8U_iuo3X1&fW6s|GjjqWk8(-yoRwlSt4b?0*se^0nk%11x~!wFDxHMkhR4Dmz)C z`{Hf8-y~fK-7%f3?JKo#;^W$|!bWBG!`UPQ?`D6V3+68xbIw4FOKXWK$-4b&8q|Xd zUl}~f2u_?UJaMoMg@~PuwT88Py38HU?)x?|-huBZiWdrUD5u<&aN*VWBb(4`J$Z4~ zI)|(Ua}b77hdJe)VBBFXvP?LcNZ50FmP9ec(Gbo)nOzs}8tQQvZy`5r3S&S{PGV3t zysaL_)BitR{uknw0j?LcFoIj#;BN2Phx%#eCKmli)#uOKTN;zy9kc@OWsS)I4wWcX z1IsSh$7DnplXqT%0H|p3i;-)5>_;9;Ujd3S6zk38eZ7P|BSbvL-h@(3w;Z}lE^Yr! zGdltYH;t?<`#rtpBg!n0`6>5!DnY4G$>6+{=wy+k>ApVQR-an#C(bf@DEwymC67UQ zh`bXQFf@)Yy%-Y?gzZ@mypjR^8SBH8gAP2ZYXc(84(`ESmb*as;s4&;S_sL|`ma6+ zv@MgPp}bH1-k!Te~qw zGD~}1y4=i9G8^J>(?DIw}e1+Te$*3o}izWojuqAL%V3MfYF7U%>Yk ziHcSZk_+{yueXfVH8{ok*LLhV6}FU0T(4re0gFa0hOLvN!!e&NZw+Q8D{1TBxd$!s zyZ~*S)!mQ(r1HX>?L9pZ93ZK-UVyfy(k={)NKM5EZCNa4RyL7n;i-C@nyDF?Ht3g1w$v)#V~6iPhiIO#yu1iN(FPn%-8{K@_5W!GgaZ{ z>yMZ1xfl~LaA~^e{f`-v=sv>}2a~>vl))z~Kqcuk!?y-P*8i9dhO4CAn9}AtD01cK zG_YseD{XBG|NgS@b`*m`>Uw+3TGPo$B{rvSpW~O$4l-qaKTHIZn05j8=O0qwaE`4Yo7o|z#Ds&#F0mZhJEiIJ*=U@RAc5Msw z%}tAvnQ*t2&J~jOePd(9(|6L#SK&`zU2fag;9yS=1p?`Ku0_BBT<;*a1W^l^*K0K7 zr8AV6Q>p@H)qaLJ#Oz##>y?h^ZfTw#*-I@ZAYwujjp5Onb}HdhI(-%=ThuN< zf0v}!U|b(j4aYZIRoZO{N(5WfdkL6{Ubn;ekgaYi_YW`I4TfdQ1~S{Ox(uUBzsf?$ z7h2!k*Gf3Ysr8alc}sMkBKe_y;f$o*w+hM&VU`hMFwM|mFKByj>N~=Y_m=!;p91-5 zdYgK14?ewB`$n{?p3N*U`A`H)StP=&y`t5vy^zzyl6m6`pLN;Jj(xsCNU@HOY$qR; z-UIW(v{aYcp`_NfTXNxCP*LXK=-~n`J`g1MYOb78FEgb=))`M;yv#zY>KVPN*T)$K zblp>TF{jYCrJ%UH)FkUfzpFFmIV|kn6l0+n{R!vR`?Cq(5o^x)u`cW05=~qZ46KjF zOgJsv;r3Wv0^(h@6w4f|W)A41E=D9|QvrkFP^wa1cmQ*&( zK{Netgpzv6>da)=9|v?-XXO#C0t}hPpiNvCz<{v8a5SB8K@M}?kPi%0UA+P#8tdD~ zRwm;ztSPH!HQuc25H4A}HT_pGuUy{ zMTS`n3ya7coY=CW+7c+jy!@^)?7P54YXu*G!x^4RoT-u$nfV@C59E@|PBTW_SD0VP z6YB4kWo($xlDL-$cEyxV;V(I;6Z+4U!Ukod)RLe()eCs;Rdbys>uIR!E)J?M?O(Ea30P`KWl2DGtM-u*gYT{T{G zEN=24XRc_caM_bZVT$z9Xa^>uZMK}E(*|JDv+oyPLE+t(Yzya-X1N>Hf7Ra`BsCLQ zMIU%6-a_c=Y078oJN3#rL>on@xIi`oOaDA@Jhpc)1?>mPcQ&}R8|aNi zX=8ZlJ|(6Wu<*dnD#NTYlEl-Y80KiSQ*7xA;b2fYP3BdYg+W-*df7D2$@|V6KjJ^N zB5!~X@x7D6m8GjRMyIaYs6JI#vrDwsq!LszRLt6ZY+kw?Nt>Fno9z4XzzP5s3(~vb! zVCYt0O+dMu@h;Kv^X-9oMVl4^_)*1-DMiDzkUwN=^)jyK`@^3KyE4g1@Mxfx$~=c@3?YLT9&6w@W~KBer32#+NY7P@t-1-0KU?feVv_U3Ez zDydb5wH$I(zbREt|8RHr{(91_@*wRa)D&tT3}S8jZIOtilN?eDLC z5%w%D+aG0^lNi9h6la$LWG1osOesQ1z2Ci;V4xxYzm_w33?_Ehi-s^Uw;~2uTzrr! zdO*{e4}%J4b}6t;fJQ{ajUOG66xx%h{Qm`c^zd=$z*B^`8F?g6eIp^_g|_}YeEw>b2*infZzYs{?T?-4>7I5s0g;Of2W;~ zbn!`>iO)SG&v5r2RpcPGkA&yT;BIAyb#MJL+XHxGAqrUz)-#BG<X|D9*RxRQQ)($(i zHx@~f4E(+0Iy1=KSBl2piL)X$hk^qj(s2M1C{bB#r4Y^4!w@YmsW)!bA+XUlOk7Vg z4YU2pm=^Zv$8Ck(2W4gM?>=4FKFFTz>`k=p+Q7t#4}H}Mws!tjMtP$DY84(_ zA(HE!e?ApM%EH|WkGWmErgz0j2K8&A=GVjz(o}fI5WZpLNZAIlzvdlKF*RU1EHf?= zwyposF#=q19f_h1uHg%Sc6LUy-Os(%^gIu~)$pxYJuZ z9){X=U($gn`5l)$Zl2RlVuvc)a*utyp@}je#TV~#F-X!orf}v>(Hyz#yPheFX}?s5 zvinlf%wmVj>-HN}<^|ug;$C+Zt+4-d`WV&hKaAv4ujXo#;XR#Q?}`-o@t9u9MmTX8 zvov%|8e)pJ9iEhq7_dIr0r)yt@}Q|)4B;9k88HanT|I=du9ii6OrF~`Oz7NmsoVCR zZB8A&{k0f#?qj7popuQzuckd@gOFF%oOVDi=$s`88hCyclj}E39m7J(8trpA5d$1Bk7E_oV9_Ce z@|T1|%7jyHru>Hu&V^rdUFAuQTF_hvTbS^FhC>%JDfTme7791{s|Ih3xV{{)@_kZ$ zItYJVIJ7!^t-mTy@CqBYQ#{Z_LhMxAT(|eGkN4T|n^`g%l4!T_cZU}Ql1$ES;1@)g zCphe5V&30v)?~tB7MZ6xSFh3`@9<7@ugbhesDXI4rn?lu1?XSZ(O%~7U5Xw?Y(#9q zC3?GXrbj*=I(sUv+Anoo988A}JLfZ2j^giTzd1e`UwY03N*aL2rQUo;je3qXV@B{? z^&V#4HROv;p9kKfQd8$j0b?-JIg@f#oSK66LB^pdIvnt?t*uuEOa(~{-^L)rVz<+ zNj$%^3nEwa{`|Zr3Lg%(zyMLi`D1e8<_zJFkUcLi4r9wTJrKPfq@%J)3vFY|u%D#z zLzw8qrQfkA(Bh?5$ZK;oFbjnptN&TfaQO z0%bD7b<^SKIlM4~iM6&DN|T8wt$lSn*<=gpjY{I`y&?mOWRe$ph%BAQB%?>An8Ooa z`lYa&$-hS(l|_{lpFa>XDz0j}@>*EWP0gm|N|Ytzu%cElvHHw1^Q1B9-j{vORK`L3 zd)Xn`nGM@K;r`ZU%>>d~71}wZg+>C^w+<_|sQmXvpYa};+MWFfkr|<7ZCFxJBaHD( zQcj$$Lo~-dOav1;4>}xPwC0{+j6YYvETkhs_|*8VMzktL`$^o`Xr8cyLlVp!;8p>w z2)e?qIS9^6UBjEu6k4rp?Gryr&Dir@5FH*TwEf6p#f0{|+af$oH&I)r_mr%d4F&ql zejE=faZ>cSw6ZEQ=QJa5rn0omQlW%lJL_Q{TJ&P9#nB2K zna2QD(>45hrRja`D+*Ziu7|v&HqF_fz@``(zRJVKyRSw=>SP4mDtd1dcv?V#;%R16 zKojt5Bu-vi!|tfI_Q9(1z2{U&UG4oUZA5H)}CEEb!q%L!x-wfEi;(pC2&o^cLPf0N3T$aMB=w{QFT z-FF|o5yreKMv?L}>)xFq{49OIEXY815aooGLjJ4pnBLHs`|i1xWAb4=vNF;`uO@%Q z=rp~h0VO)BS&nSdTdG=Ab?$nGb3|}lTg(pbd>|21IPI=pH=I1JG}o~}U-Nq^W}%QV zF6gH#@8`SmLAv#;x;XS4C=0QBUo=LAVP0JE7SnvEw>R<9NBNS0*O4A_XI90pn5IAn zLrJ81d0kZ)e~u&%Mj(q}&4zKQi23y6^ey)Id=+X{bx1L7F}pm4YwH&B8MbQO-daF_ zCWmF|6b2DSw5j^z;@F=vD7H#dOI)g{8f}NyA?zxS$V|h$2L6OI^`QTRzl`|d-h~B7 zYkEn;nh@nB&l|@}?adLkh*l*$Fp*pQ%8<$L__0fmr1GP_A3q9ZPquS!m;mkJ)ywqE zJi|<<$MYiQ={aZ4@zZnJ(bFBWtCzTLQ%I|&*_wzLwiA2lb_@fr#>zG*#vrR4l`dA5 z;v5gId7`ZQXp#Z5Cet9U6FMh@>3hXC^ZWcBku?P{vV$C1&6n_x_J){(SlO;2Ye~3A z&Zs~xG2bEyh%`eS3u=j6LfIa<6VEzCi}V87QU{e7M`S9qaDmj|=OlSYDuL{ooz?8w zSd!|{rmd_ED?tlB!|MaWI!3qw{EKB^UO`R)<#5**UC3C@^r*VJo15 zKNWui#rY+*-#hyLs9-BSu~LF!oE9Hfl7}9GHp>69mWtS&Bh5b$*%+t{7$%2ES_Rcw zAb&GW0}wK3$uE<%4-~oAo*l;d0lV5=WMXq-AWQqWo;Pp4p zb3=@78SUC?61l5N?LG+&H-MiiLZ_2BEbsLyTA>cmhgCEvtlDdsPzei+JQG}Yc>m)@ zmiDz)hN$~P{VnE#q0Q!T&slMSZe^`&!~y{>-N3IXN@o((Z4{WisW&Zq; zywZ{TBl>BwvG)s#q*gGEH4VykQ2##w&z--i?#W%hukhv~?5+mhJgMf2G(4{6)OL($ zPouoXnxJu04FV(MO+hsEBk_;->CY=+iPfFFXgWsfZw*bseAwMb*T(ig>;NaOPXZoW zo0Y_^1d41_`&G1{+dN820bAt%V#TIpe0W?Ov(Z!OGIm=5n#phXOMKDvXW4A=i+(5! zZB3JCJKf-%_IG!%v#`r0h==C@B-hoJ_Ie>gqhch3tU&IKUkimchEmK(5O6Z0@5zie z-lwVEJi$9;re>HyBvr1K9`>@@w{-&Cll#h}V&(Glwy6(!>h8YDx{KysBHnipn$f{uX=W3pJnwRBFW)Zxb}&C(-q#uB<3#A5P$y~ zl1(Xh#9>VIk$0Wj)nOgqNHrDRWlb8TK8x>G(V**`3^&I+x@y^>dGfv952>LUZBjQy zzNT%ASFP%m_|bpZYo-C>_FU7#f&(u4bk!LGH~3Ja|6!hXD)rfC0KS>`0;;QNJyD{sc; zbAoTu!6&EX)s1FzT=N-R&O;_E@4(}EFs|2DzvW%NgdWCMrfhS<-HZaoHoeE$Zo z-xvI4jGWk#yqMSYgt&o9i%~ zvprIAF2%|NT1MXW+#~sZukdesMyfAkALXgLei4B{Z0*Uk%RKt|7EFcZ%N6 zquSiwbBQlRX--uVnpIQbuplMAE-kSLgq$u5;u^1LX zlkh7*uM87AEI)!prAIIfh_=v>z##_nWg?aEL!V0DVdpqC!fX=gDdZ&Mr&X3yoQ`qm zDT3AoNS~y&qE_OxmicHs>s?a}3abJ@8dYR|G!z;(wI}sk&J=oe`Oqq1DiLkhczV{k z*-~IZuQk|+{ACq?XR?91FNvw`6uqd%&v5iO?&ANJS4Vm{^IE(xk&(n-Ik;cVL zz-1>?v?_4>yoc32bQc*ak4;=F-)@95a7cjaIqlS`_Xj69x6;qV)ZY$g1cr_C5iW!b zo@i#(3?EPwB+>RGRwvDpVR&N8;;fVr0}3Va$pfZ`;5PT{lZhWFWVM@V5LRY&rc=H$ zj-~#rwHlh&&?b@$WOBS zIp&_mUPB8-L!4V~V`D86zWvq*3$!)n@iUPTjQvSgBBFa`&n6Gbcp_)^Xbk~&`4D;G zoHlm0-y_nnbYPJ0th+6j=%8X)I6Pb#+PBJ~eAULdn4@au!9UH^-=l80&c!M0`*Sfn zH-;O49rBJj0)T01Ko>~5n#_?T_m>tl(9-+**ER!V=F@1fz*5W-21PG!^FnP9 z?9ak$nZQ%RNo3_;V-<|R2wz9q<%DRqj3^Ei7Xzj&>dq)Tt_%%n+p@a*JDh8fqvIT< zZweqio_9{Dt)jYbWB9+Ihh|+XDG~u*_dttT45S*@Fl*sfv$U17+0W`FStps6Kx9$T zYbXlRXZzUjr}~{BfhpPkPg32K;M%p&JGs;#v2BGHpEUc8nuoaiMK*1_Yl3#r2dATu z>Zi}uBwrc|<#5j*RS-ZW1DIo71zW=*0pT}7NgPqVmhJLAp?`2S+OGiNHX`3*Tsm8) zHdhBF)NU<5`5#j&ac#aaed5B7$k#(Z2n`~0aQmxQ#dP4VLfCZ?$0*T;fw0Vrv7!Fp zi@PsAfOb0yQtv$#_Ur?ynd`ns)a+^m`15F4hdt2U`&;ptt(W)N#UZ14m@2W}!|H)z z#Jk}qb`TptsUZ#oQqHnxQnvHUw>sPc`VD-N6mGnKR7&Bwu_fbcjq5q&oxM|UkLY}FPMQL zEFLyI?w%SuaVATYpFT=5gUqr&*D|U%7nG5u$l)mLktgFu$yLv$xxN!!>hGbNQLeRw z*O?hE&YJF2=j-fDv~@Z44&S)l#a@yp$Uo!fRam>@m2+xe^1TS-xaTOPfo3Dq_BZ%} zrH?8iOyTNb7iakAsKvPs8ucHm+vHrlFRQI~Z_HbA<;>c#q8Kg-*jBgh5MN#)xS}kT z=f7XM$?k8QrunAC^yUS14yiG^cJF1#LK65;B#l8Yezm2hxnnIEcUaDTK+JnwI=`a5 zpEX`!X_cXEmpf#$hA=rQ|_ zjJ<*#)AJs^cH>AmB`M&}vIQ75^Uu)-7U#fxmzyJQKZXUGi_I@2p@y_K0|CWPXi)MH zoYPnwAPNP2>0cYKV#CI+q`>D8Jl0D}LYFfdz zTt35uPm3wY(AVB4WJCjYQlOPJ{_RBJlmG0<#=!vGxz(Qro~RC$5y7DL+GkJ9^?W4hh5-QYbcXr|RIaVw-{6r!QqI zOTsvAm%SpfS;dIbE;kZIjttI(!AsWjBd zNULx@ZXeq&V2!*BZasVAY>r+HChz6BC0Is|C7kkysup=s+BR;D{mNYX)_b<=v`Xqj z-hSH~1e?^m$kFyABS=;FK!={Utlx+Txs|@e`&pODD*F^5o2qd9QLa0;qJ&qs^}5 zq3gm25NOp_>IH_`?0;v6+i*4a;ckbq+g>OcFMTiL_90HZshQ1D_r-N*Ezn}>#G1?b zY;w~94eQ~%$D72vwc$P*9@Zv$s9UyRkQ-SyW3-tw2;Rl;dT~aS|)DpK_@&{ghd;K36bTa0<95?^Pzz=y=J9@Bl!?JX=wB6?R zd&k;ttkL~?vtI`@_J8RRrW@W6U=Fjs`6lGKvT5hpUdQ#s^}#pyJS~2s`^!-!$QbeE zhKL*CHfGWT?_P3C<@YA&hE4IZw6mPnwtCS-$l7?z+l!oW_PPH+BU%P1Yg~$~(#D{3 zO8dIs#+6!sKf300QvAf{2baf7oX79xLe1_eHzI#kF8G8u)cG#pbEg`u>R;yck>7} zJS2NK`3!&1@5%+(MDc={!z)AF9g84)(GcJJrgzkTQI|svO~XDyIGd3p-jyV5SoN^Y zKz4E}0?A!&^glo`s`$zXkeSmxURlE1CFjmPX9Go+1}*qwzp4c$A22(2p-a3@Px{y@ zjq*k|;a^h1jJ56ZySgOT$I@;t!z~wu-wtFd(%9Cm?JWuvoVR0D9HSaO{a7w5>OH5* zcfs&q%V+huu^3T)pW9D&EBeXf3dIv?qQFryIQ)=vGb^^$M8*`?DaiqQ3W};4%PwVo zsLy>=FW)q9$(pk$8*a@8`l=N6YPlSAv|-&R%!<@7cJ+FBj@ioJG}cGSCK~e$jMO>+Fkz#tS*AqaUCEaH=;k+Q8n+-t4a`%NzIeFkLzcZ0OkFG^y#@t1PVc*@2vm%=wvJ;cM# zTTj+eql#s-_=}>4`p-%vHd^bZ8%g$Y?2x)|x_=Q?KuTU~0;R5QO94mtDO{E;8Sq-V z47IbpqLffK*Rzsb?OwrZwXuaKp}U;LcC7muIu!2{R(Ma`zj7(V_O%ggkuRja*m!}D z;<;}9CGTO&{{ZU^h5$~>6#^IScK#*qdh@bPeHSX!v8v;#1(3a}zu}MNuhf@8o+>u4 zo&R0ncvzg!{h2LFevX*?iBHT8^i$WcK0yts=|92_sRoIhGKsJ~UVxPQy z+eV*9Bovb(8a4297N+J|>{967#0QNWhkz6;Mte;B;6L$GXzQ3<&U&vvC@L4LeH(N4 zhG5;HFSzXugZ^J5!X#?Ruen81g<;he6Dl_~KOWCV(e1a-t=JrFz$>;;;sdXho!~bP zLlS*ip-WToo)hPchlOk9!2^-&Tn)EIF}P1pySv}J`ZFIV?x5;nNi%o7VTD+lJoB*n z@q=Q#r89WQ=SZon@%q~dfpTXv=iz#k?&d5h>v_-o-K)3v*xdR)s+5)>^TpEM3Z*j3 zqKUp2v~n7R{zGJV>^8YCyLIq@*U_pM%Q$DM@1ZvA9{b81+6IjwdUsEX`89v@N@!|S102dYEGsq#05s`@8W7_&BeN6sRM>-chAo7F5<9H6d*W>l zt@&|vQQ;)n2AKL7dMHPptX1?#`*;6Ej67Mr^XjCisHhCeY3%J>Nj=oU`-Z3gT#G}i z{axSW6b1&p)~oX}cQP6A$MTqTC(l1M16nlm*~zqBCPEV2ZUSWwG3uysPvQo#t=XkM zmjBfOUe&oF91A%^)y7XYH^s_ubPYjjPC}ae-4B>%i=RJzFPSdV8!)jgdTAHs@s#Gm z{XV!p&tSH)LI}vkt4dl$E~;P=zZ}{fdg*B^bE+_)9v}jS>z5nLV&nOmrNRM=zem5p zE$mv2ir11Y?=rBb6LKy6&O)JX*u<(!CS9}<$9DI>bFf|w);ddo|55eIn2`0^Pm86~ zV+WS)7vY+{7FUT*CQ2DT=1o+l>9wM!FE|el@#GRat{wo3$xAU>uIp_BUQziKNiKp_ z3}b|4eTwCw*cc(4A8id{58YMxXo=a za#%w$Nnq{9auU#Izln;M?D8Rg~^UF-9ta1*!z85H#%q#XDkqZCR*hvRrZgsi&(`CVipc zOril((C()YxvKoCaq_`Dr#qchf|!EW(CGOmJic_oA5;s}Ryfd0iT~rkqBbLyCy=B) zj>*kd2H#4LdOIJUxNg14vgG#?dbdwfd3MyW1QA8i`jXL;+lCQ$ViOxy#IuPQWJw)n zX0t4$f)Sh9o~)>=i%NJ5g`rfnWzW}XA3$e&>&r}cigKq2obi8^Bjqvv8D{yz>Zx=h zPzP74u#aA|zt<^MQT;|Ysw-BMCXX4NOeXw`l+yFGpU1Wh-q5w*RLM3LcQ{VovbK8F z1>(yi{Ycl3v(=S>(8Px8b>^k`b#`qa3)zZ{(%PHXrq>XFK30Mwhx&wY1^|e)mWxfF zPtE2_3`A;a5{?E97o}~+)0uqGq?e19H?x#PhaUvGv9YFTg>kT$iVZ6qS6q-t;sYY! zkxm;JdrLVPSkiiM@mJkz$MT^qEh_d^n_0#Z*vhojyl+1CU#|z*iJmuf(Wk#@x)hjp ztf@xRDQasBO5++eXHe1u9mMQ+%qIB)|Bs?`4`=#+|M(m-#1JyfA%vlv*_;nKpU;xy z^ob#v^ErnQ&78%WA?AEY5pu|(IaD)e62jz|^Mq30-+q7Xy0-sz?RxL^ykGb8@fhsk zk0dGAR`|BN5Y}`b@x#5wq~EZIgd4peKDz7j*$Ve>)iZ^+C|gc?A@=Iws?4Y59S&%~ z4|~+wyKaFM>HF8X3jcE|I*tIPcJ;kdT48Z=Ae$gMpZ-9K8l=;f4J4vKGgp=QEn04ar;t^nM$#A3I-Iujg@+0%;?Tv}d1TvEP6{CG6=m55XVuwz z!bOvNj9cE(tb@**N(XT)ALjPCZ5W^bS$BM04GnPk8S*QfbJ@6MkSSb~k$Xprcl6;+LpO=6D83>+N9?T|i{n)*;Z>9A)2i$2 zf&CkKH9}+9V2~sbfL@+%XMajJQfaeLf_1kI+ke8W0~M|I!*H{T&P-moB&<2Bx_h3Yw)SE-dSgEjHPJ{G6L5Z3u32L0aSs#rf&vAQ!L?#SyU+9 z=OKOVkhavi8j8}aca9I>ab~f+%J#!8Qb)!JIr-|%G`0K%0d>!&mOi|BW!^^<)^Msf z1`x26HjkPuJaDN~iyYk8h}QtF|9 zea9xo6#N|2IW_r}MrnZf>tKfMVXgc2Y*`k(If$N>p^~FsPrKT?*k4>s7=PN8uMG+c9 zJ`crkmWX5l;1As^hR|g7&g#xOuS!pv9h#p~#^KN282YL73)>ZCcIoyhrTOPT6giE7 z@dG(MYBV}God8G61%n=|Z*zfxt=Gk}Oap@yP34Tl69R0KkXb|3?G`5@O-QUg0Mywi zaDAysSv_((HzIlJ9pcr!7D`$|;~hSMUMj@-!pUPynsYFG=(|~p@wz1UJ8kA!b+Ec? zAQFoiJQArE`W9ZR@mJcaKt zRC~7{3B^tUG$=rD8Ji`J!XQ0<7@YyrK#j8iG`EA~anory3p)NFNe(eT7af}2zUFc* z$a(nWH;asQk4h;fiP_$0mYPpd4DaK3hJH7pw_Y!*7HX$y`pyJg%bZ31E|)Xp;UZ(t znnfnGZZ$khk3)z@Ua8>a-coZ)r!3jfq&fosR65kPZ+S?z-kmFb_(IVqn~mJxToCqs z+3&@z>0-8%=IU+>B;#bI3^{OW(dn5L{wops(42GZz-8~LeYqF6jh+iGn*EfZoJc2u znhRGlV_h5~9!%EWKlhX%<$k33?PxpLgj(_xp}>`G4ap;1PrYP-+!bM7n(5xbmC+ex zOuVjwarI1kHsLC$Jbg<>Qgn1LgW8`dMq#p0iP6|kTme5@u%~l!hR!)%5UHh#7CCi_ z*AUXpa&ht$(^dpUlEBU3=SF^ZW+x90w4$_TR#kR4Ecg;X%MGc|P4&{hzW$zz?je)z z#gys#bZ*NWsKj?zi4AGiO4r{w`)rdl+lY($NKlzbxB?KBrK3BkS=Qxi326m&+h~k3 zU#I>FO%36++NV|uuSOMAtpnJU+(x(pADx#D{Zf+wvDA3~+tQdzjZ7oFM|Xu@6wC_8 zJ1mVmYE{&_`l-q)9bbRAiStW{rXlsEd3>&yCa*z*aWJ}MV#S`geFvzDCDp^13zp-UgPVAoDZhe7%9 z65(ij$ojbay*w<^7$!hd3HI7HjDMUzKGPR;k@nal!`!D`bNgP-LbkWy*9!N=NbTD? z1T)A>q0j(*P-WKGHC{Ha5acVYB1wa z^U2@~jKF835Z*owIUqxC@?jS0=jk8sziap*Eiy;0znAOyE#=P=siCX|D>=ky=U97x zquko(8_CC7=3siX0o8-U;?*SGar8LxKuNj~$r}@p%Zvujc64%4h8UvxKJ^OPf6wU_ zt9&*XJ1A5V%6Gf>*BhB&Y;})&0?n_gV+-UMvze{*c#5x!QNzObTOV~N{+`#rTJKmt z){Jo~La2RwXG`-zk`2Bk-SMUy{Y!Wvd$~h-(;hFW$@XH%UCR2bInPDEtLO^i za%K)LdM50mh;<%b{rZNPU{8-I?8$c%`14lY{zI4X(1iZ+MLn{AY%jOh27Qzc0|^8^iz*KAL}BjYtK z3G^UV6%G0yI(r!1r^4Q)HFa${*i$BxQ4gX>QpRGgf}*WQkx;(-pQ&)1VGx@iC#aB1 zPAmwi*Xx`Vby{xu1>MH~_XOR#7nGsg+C^sl1QagUSbJakPGrFJ!(<0nu0_W%e9zId zGhdYTr_n6d&y;1B8zPvxw_G+hocSH?FTTRdoJJ}#0>-O7n$(>PPeh+x7--Rm%~)Js+$A5J(_tU0YbdUk3TBlNKq9G#^cr`DM0LhCi z*YHL!?nieHZrK2A%QmwkN6T=E+osfcyhFRySbmH;dyzeJf z?8CLYyQZoP5{(uOirnasAxXPqGK|zF$qU{Toc@~ujt0#)t8Jc2gJJ8LOEtm@hl<;l zKAUvf`Zns<13}PN?1E9l@1R|#IXSa9;~|C+!yse&e@a7lE>++pys?}Sl}+xEOw>Ss z-L=Th3$Wx7Frzc&3HFA_!>isguth@xt*__c8=%yGDNkIPm*$Q_ms7Q6F`ktqo(gpa zHw7C|clhC4+&4S$$MZ+o^13BY!}>S5p!Du*)ZM7vZ9S2!p1FBzTg$cNyzRY(Y#dv2 zwByJHa=4Kuas^_mCnCIHo6KgH)^K`_ChN^3b7LE+jx^jOU1Vfh?tBcHCHHhUfN zy7rc^wS6Fcv~5satyyf39JCUbPx$;0{+f-7XE#|;TrTd&22;n?QIGmg44%6@GXWN? zWvddip}c8Ylrnahb_u;adEG8I1PA@NP9e`$_>Au25UR9P%lPDTE^6}{MHSQPDw7H; z^Ag}h75|jXt3LGM;+=6;c2VmKC_QQ?1{`67bNdftH!w99{K&`&0j}fB@*DO+@XLIy zJR7qRM^0t@s-4JG_KpvGKJ4((qO(OSpTdfZwtJO{cYmfpBt!^u9x5+42jPO>8!Q&= zNH>*8NR$sO+$;Jga7ftijDpxe>6(o9-2Sdw0bP8rKX0QcJuhlR_-YBg8UHyc?b|0D zJlyEt`Yhwi7cQqeb%x9n@J*F@RJiP9M=4L_?_r&k30D&>Eqfq#_HkCc$hU2I<&x(I z9UhtUf7(`g=1#sZDxF>9$w3GEXFM7$0Ea1uQDJKjcF(MCeL=BP%#WOf12rBdP}RCV z5Xx^`o0b}Gz+o`7n~HXZSKRENRB6%)wiXiv-jB0^xvqR9eV)R&Q7yMxt^WX3Ick~x zyOUThNEnfBPFZB=pwI1ZEk2GzUrvD`YR@S^Y4`cXEPv$JdECJJ$5$h-{c$UCt~Ksz zfjD)ZKk3JWr?Gg#a`1G8qL}SWFv0B({Ht%&!P9u|3aR^kriM}TRu2Yx+y+hCrdFIA zr#!O^Xq)%wz=G4$oY3zGu@Fyn$JY@ zeP)3KET0`djQTvULWll>DNSwp8B!sb<~t#?uUo1H{M zGYaOq}{;Ty^Ip#9ajvnXzFm zTTe0boAq=X4+J9J9*NrI+JM>L855929&^XdHJAUb&YwQgoT=b%>E}KPZ?Cez@5s{E z1yN()WW(*w9DDpk@k@}ZATIyuCe{wi9%mqjI>M~`F8!oO7iqfyKTkC(`G}XGDd}9a z*0{mh%=$4l7yYwHt`azHNoo-bxs5I_Pgipbg7xCMs{0+PG4ovcdjxY+{9#?6N63K- zpXmkWDKku+z@(l$A!L4jTLhy4m&m8%p7x?N2wg3^f13jp`>OpNHYUb#WTSA6HYj8k zh3mN7(u8wWl-0dUngD33*=EZV;SsziR|w0%kx$;C_d^HyA~ z+d#A+&Cz}Ej_$9Lj+MC=R9y=9&r-1PN2CcZWZVfY^kg!F$Uk*%AplyEm|im}3Lu4K zY-?C?<6Ai{?m6eVrc|hz@z-#i^c*_Vu zxn1bvbe)Q;tU3*KB_5cPlFwAoA*!6~RS2C-6GUxFMS;L$>&FltzqA$36=)d9F&eWQQf;Jr1Z%@-dZ@ES~zu{nUlX>(m` zuj;t1o9oLH1iRS{l&F=(*0Q`&?Gj^<1`^j=A=tW1`JqAKa*k%ZLo$z7{u-+_xp3x& zzW+fcQoLjRvz!iGBI8UbMGc%fx5p;P+qWb2Dn)`T%jqr3ATW8d8b%U;mU+l&mlz&{ z3rfBT(+2rJLWehboYy;AsG1cTxdYrbdcEe9p~@Df{>7l5E#I~?C=W8+#82=l@~V$t zGIW<=^9aySBa_(ZOkpn^Ww2na5`rtL6x|qY;B89{j$%lSb?9l=s7-lHEkuw}*z>K+qwee5d-lS;ceA}p{TyFP zevGH>x^eRn$Z#Ehz|+02V5!ZgD|f%oOT6ZyECqP8XIpi(UxGfY@tR&?({|xQv@~AC zK|oUqQy7mBI@O2y=*J8qNaL1e0}_=JHX)$zgB*Ckk0T_ zPt?9f)0eG{&f3*2zRi2@%#qy?Dxuyo<=SAiURkK_W&LJz>Voh4xtWq+_T*C2--gTb zT*kPB&}y##Y<8>?SS3Y@LNGe;#s*Y6ox7cPk<@ih0BsSm)f*?Baa5Y|YNDZ1%l)}!jn%tLWEtL!zBd2Ch~k7WTBaO%ivq02~Iu(ElRrAdGW2;oYXH>9l=^k+G3Au&mFo2Gx;lnh~{GQ=}rU{@p# zp`f`vFVTyr9DQs<6@CmwR@YOwW>77{s-m+m)D5FZc=-xDquIdLm}%VdLw+mK+1|uM zqs&{Ex)1A|Zw(ywwZ+b1*<->k{jKiI?uXu^;FjKH2gmL5>Q)m4MlrTpbswf4o${+> zIM2+S!?{-l;pb6WIZ7|Weccvwjfa<-?;l2F+&cYx8-Cs4Bn?zKryGIbUTQ&mJK1(8 zNskY#b-oaL2VX!&3kkmGj^!To!YfHqI})c>g0bo<1_z;A^UY#qdv1=ac{g5AEn;(J zs|WgC(IK{awW76`b^l6hM}2Oum{^|TswDt?dXar%e`B0vUg1!7)qGUa$Tr5X?8QW_ zt03HEIy-((N1*<5DADvD@66yj=qn-OcVMl+Ptz3?t{%JAz4JO>MeCkWVkn*{@O-t1 zqxG+)0@Q@YNGfCuQT`%m*vgz4?veR*NJnr6<-MGSne0HDK=RrJG@(2+K@5@Afj7c& zJ+&2A-VUjs*q%y}vsq|i@B`DmlO*|VbFU$J6g%QB6HCZ}gTR%*j{#qmQrLh>`w5`V zUCTjQ(9pNt_6)aei=N~Z{=_j$q;cdXFOeng9Phd(DvX=CCqn<{V+!4OrPz5uygg^T z@9waYU8r8qV8~wm|toUf0jaJ;-aw}wg`P?Unpmyui z3dh`OCEE}qujXj2XMQ2T>xZ&x#Pstsr9HiLl@6%}1*UDM3^DUJ(y?hxtR}v<_E(%M z`5$l#oW6g>j?%K3;+~;z3~^km2vPYb*W?P=i4pqO=2OynVMs&Sa}0!#Yv{un>JNqw zlZoa(YLm^%QWWUIYqGw4M_JTFD%vKNd9 z{+k{g9!(^{m6-`Q5@%ewN62{Tq%t`PA=4<`R_ek-wzC(!e`S2_H+Rf4DxmzyWTo}| z;^(~GJqsbI*s;lhDI{gcQk~DGC$V>MXtI*rY#v^vaK>8d7mP~M70*rpoK-j5swL2~ zgu(Hya3un@vv`WXt?x}6pTD2xCFx;27-s-0&CcnsiCF(_GaaOu6@4P}#AaxcVcjP} zt13H;B>tiXVhsWTsu19&vc)5!p&SF#y8TNk)d4}pWZaRL1Iaz%GknO3c_8~^ANtJ% zCMvOOFrnVU0HfO_!g4Txn_1PSD$UIT=atT?#N_$#s)S%md6KD{VlwtGRA+Q_BUy9) zc<}Qc39oFpY@rGe^0V!CzHDW=3W#sTxdRfk`Ski!>BP_!EFi|t_zH^&Jtk87MUx*3 zZZIPoBIi0j;X3#}`4n#pgkNqqRI1t;1-L)VG1JvQ4hlJ~TwEe;n$mn*z_4k{ty}YO zB^fdil-93>GfnyaGto~4~MmCx;nfR|^vq7mx2`H*k4)d-$Dp5i~mZ%-QUS)mnv(jHs+ zrkT&nz*M|6w-Sus9I+PvU4ugAjzqic+1sS=V z320lUS-x{#)v05E)$TnKdHc?}+`6s;Kx6HvhUI(o)YHYKS@g1|I!hxPrP-SZ7`V1gL!mTP@uT_!Xu%STWw}h8Y9Oa!w6x+xamo{peKBOjy%d)g&G#C z#jP>#{(fUjTv_XUM=wd2{J=gkYcSg{JQ(^6f?n&5J2AFU$e~#qGZI(ggu6^FKsIZX5L?XvBhyi8N)wT#tZUJ(4nd z*hcnR;xJu?d7chPjC%+QGUi=NM5kiqb8rT7rfY0Sf~(asix2-ny$2wo;|7p zHOnJo8Do2bLM9q7`Dz^dr3cu}QSTYyhu-P&2{Eb( zxbg02vS2PDf~{|A;^pHYD>p>0$|=lM$WPx!)UFGVmvQ%=-(fJ~XwP_5N^zm+b1ox! z7NhGvC=&-gTA?M5(t)cwfj*`%6!fdi!WViSG-Rwz}(`;x?qA_ zyb{}p`N`b2)%{^j`CW@)t=3a8FW10;7l6&h8aa2|NL10hy-$4AIbeYX&&qrfEiNz? zhR~`{xOWTyr~%7n+gu}p088#TqEbl2=uA+?=YbzGeGOy_0ZTbM(KED9gW0~B(R<0<@8-K_lWONeid_$7)easP$C%O<2Mz&PyEBqZAr1T>GR z)n^B1GsOLwS<215_oD1kb?WB$uli@)d4E10hTP@o9AX1@*8e(Y=fV2SmRAik))y~~ zr9%?&PYeE|dO`EN7nV`1(dwSo5rXnWCt z-<+;3*%TG&NjKyHZ$;0E9{%cY2nFl9UZaEjnR8bhdcrqEZ{`Rd5ic^-=F9i2cQd`U z`p^?F%w0GAqH2RDy=^NrhDpqa4zMD4dXn}m)&SXfzjd;+Pr_F~{=U)2QqGsVCkU3O zuTDj}mx|s7KjCw@C6I|2uD!it?;T{hUZFGKnSSa*a=&*_-DabNr zBMCYBFLZLamTTPEH7JHTa4Ip&#hR^p)f5;nzh{z3v%8R(%L=D`lWoc7XIp-cuAB=r z2+pp!u&zdtccA^yKEyC^_D5^O+(7u^svym_^C!x81*V~` zv)aZFyslWtLNfw`E#ypMf-XL+c?us>8~@UEES+R+>ylz>S{ZeB`}f0&E9=Y? zSQWVI%-TArB+mbf64B!4L)(V%P35g{m#3bKPGVu)dH^1~=8)SPS zQLY3PD_{ovw*C^$jb1peT5{Jq)ul}6EZ3s&8+cp0my6w#zbGSW#9v zT=DOkP*)g~aZONqJMQOxNp+sr9d(iTgI)ZT6vSQz(X35xf?u4FiP7H>*f=@L}GjPgpLDKGnC3&sp} zs=shpynLt7_v_<}Z&H-MevO20z51oyYO|ik{l_4h9VtCRp4eXtmzQnn$QH4*7Gcov zYy~~YWV3@HZOm(;olte-Kd{E1Xq%Woi}Y}mJGqhVS(dHhqMh~TRZ`2aUw&4en0NTA zGag$8;n?P}m_&>Jn7`5=?V849et1@`kN^nqJ|jSCW}0jjRt*6=&^yU=*&kB4Fu{y!V`rY#;h|7pL;QsJHeOiu^`y7BKHx>i9!tN1WdhG}l-8nPG?tOx+yv z$-AJ^L#{$%Ht?-NRs&5u$Vv>(C!BL~FICClxG_=QGM&iHaPF@WI&LwGuSV zKl`(+GL3kpP!DQ#QM+d9^-{IQKl?hJbu>bE-;mmq(4o(cJ|&16)KqW4|6WmIVR83n z4uV-T(P50}Ig8wO%qR^tXKm-q2*%^i$_rK28PQOWD_Z z?g{zz_+O9s8Ym{BOmaWO@XI==)P+@Sc_7%*DQC}eL4FQn@F}OA?x;G+N^H{0^joHI zVp3E!OCrfs7abfFqU%`!2rG+ju}_>n$q3XB+cvJi{tsXp!ft2`h?8uQVUq$tra8x2 z44rW$h{-Q4!yQ;p>P?J|BK(d9-Ryfw>oDFQ8@%~)trs|I>@jHazG3y*-a#L|hP3Q=4rD<$2xw(x78A&iuYWrje83BF*o2Pg9=X8PByksjM|+s?{d5J1V5v zwWk~SUz71byGu1$me(yZ>1e#`R)>acTd+}1-=%-nT`I=Rf>{s8`&!!tA&N1N*LG1^ z_x_i^=Q&B&w~?o{ks#k+0a*{NlGlaN3k?*c0}`Gv`!2?WR&NHadymsaqYa@wR+9xU z&`p9ELY*9c6udZUND3E5GX_V5|xKgb45u80)%s#Es4hm|QeZ@Jhs4 zL`FRSC${YOR!PNHL*-(xUwL#l{zmU?po@Q}SyKFdWJ6Ma7PUg}6Mm+esUB#2qJ+LX zwxN+btPG|Ty_gF2z4CVWhu#~Rxrs2J$y54berA4bM016P#2OXZEIAS;A=tC~W9DW^ z4lz8Q`A8HIjY8b`_+MS#Th0uhDgPwcR-CU##^;Gs<(@VZjzs#1M*Z*YlWq6Fu=0vz zOgOxZ**_?n&r{C}`wPDqV>3mM(I?U%`!IlYtZksffu$o0663^rXS#hi_dQ0~)T1gH6d}s@#+x)Woez{6c-AKk$gVKC* zS##a8<|o|)w-|CW({Rn#QIQ)8H2sY{-JQL0A9qex0QbfR!I|(adY-5zjvj5ptWK4>PQxku74LopYRdX0tJ{6^?5_j2z6O zj&$KGl*%nwS>iw*SN9n~*3av85id~=6=L-*XM7D~6fn*M`z}b>^gTg^*_06nPgipL{Zm_x7}`P4>tOcf(yUeeoTmLG{(eHVeysRxbROu*SFvf?!+O^{ z!}%!f<^IdM-ljk4CZePq1{1Nwy;ozJqz<+RA$T0?PZ;TiZK=<@_iOQ}I!aRx7` zxE+ZsHk90@l&jc#(5z^_JFR*~7670&^-2EEtNgU>*Z&+?)sws3Q?5_6_8{0kC9trF- z^GrvEffYEWPwp-A9bPHSy7pE$*IMiG3+`_Q%mlut)Rq54mkvY0DogxwlK!=Uj5@&Z zaqXbH56Uja>tgDr2gD#9;ECsLuO&qex@4J6-OTw0rV59}8>7PnO#M^gx4*srAE4Ba z^8LBXX0gIof~<_?3R^(N@q;xO{Hz9>Sv$0l8Vu9i?s^@hqcQo~1BKH>R5Xi)aETC8|ffpI)UOcRD(e;FPkkiw@;|mk<42QB3~Qg>Epm$?pFFoX$g<5DoL~3N;#B zaH8G7_Z3F_=Zj=8s6vSn2E5r>d;919BX6jZ1;4DdffMf-V-vzIU$ACLQ|Xp86Hw%p zYkd&XjFTm}d(u0dAsRiC%WC3K?CO(9ssDw-pP&^->N3YUW*hH@2x~XqR}bM7TTo7< zVI%f$8#$%>H7Wd^N2heQx9I3?Y!9=M=>P!Ds7Cv8;zm%GLs%jJotAInh4zZNFRrs^DT1?MlNQ;+`jVjjMLUXu)#14i%4HeKm?+v(9q zT~m$KKjPe3_>x_*A-3Y#>LT6J9EA-d4w8^!4JKlFL>=QX1_j?h;|+{@5`53~_)%90 z^zPOE9XNgq(Meg$&q1qAC(3>oA{*=csXnPRQd$d+Mq0@6!8bLb@n><>3smk*QJMRR z!P8c5AW-6{ay8fpc8Z6n3VRl0RWahr@A%*wGdptNVFmgGyX43SPW3ZLe=A$5Hf~c$&Qu&G1r6`gh{fZ7hV8uZ{AeT?-= zX%0U!O0oUHN_Zc;I~_`lA}m_?dEc(22MW)bp}d=NF!>kl8g<5Bg1M1N>W+n3wLMQ) zHl;0QtV%+^u*a;`cs-^QkREMhd>3YlsTaIkG1YL&xWD(YtX8wHK4oT5kX%NH8hTNq ziP`~|%KU0Np8X*!!p6Ij<#Oydm49F1JLWb{-#Fm(#w90EnsW}}22M4OuULK=6T z!rauFJuG;K#p0FrnRbV%43M1R6<Q~yg{ql+BUE?@?Mn1u8RB0Yu>TnA|~b3L1e%3_n5KVi~XiLrOTBEc*Cdop_K zGlDh`7sb1Q1C!r{TuiaNWSeTIO8VA<0N~vWt2a=w*@|}eY@O4pz)#uHd=56Kbx1^o zD1;d~{aH7`Basvx@g-7vvLq+rD!?d^5|yH(nJHfMwXlJa%U#h2&b7=B{MH5oMgEm0 zYS-9#8TSh=)%-`l4)W6%se{yr7k5AE2cK2g9v~ZnDp1zC%nHU_*ZJRmo6PQ|rZ%a> za@`ej^CadT45*!=0iiYCQlF0Hz(S~zD!>7f6q0e5AI*5_3w!VaGG(OUnIv|~1vynB z&U?a$nyFmOdMU=5h$Rk56{$5izrc|x8J!$W**11ATtQ|8G3%ak z-B#J8$1vw|L~0fPs@B%s-L-^k?Dth&V~eMNWOQ3;N+%0mhh#7Tu(DhYx)9DO3|-9t zg97{)zTbdqiD6tY>XE|Mu7s7sae`8qDTtoDqv>ddefiyjp3|WTDJdzEX{6VVcHWMHA>N-aKOKrDNHhKJ=*H6FE8A zk=q$(P&<3FcNVdwXmqU(PTU?>{H9@_Z8$l?p}S}Wu;Y#$1prRA-I@h z@23vPz8Pj|cmIOnZ#x7J4}Jhd)x9{|GL8FT=SF0az9uC(P}4_H;g4L=h4|4*_5tkE z-~f;`;`7g8FN|$|o6tA$H(`x8m9IWbpA$x}wMBz`ccOOh-ZGiKAPosJmRgoQGktPZ2*F~hj-JtZyjlp z&BiS%zL<>N{NVTedyeb=_WW0A8c_(66iqWMP^$wY=j<9YM2ct2n^1=VMA$X;DP0*R zlBp|}`V$pjJ4#X{s0yXJ5W0W8~9V8Y94aw0D9rik`GB^pb9|4m5D9&JYuxK zP2;E?D6(*n6J>*?nmVRv80?R)(Bn#Kk=b#7CcZt!1FmM95b6I+$||iE0=XJK-hG=r zjN~V0hhTm`={sq~*t=4MCVO47Xy(LDn=AhDyYhzm-pi%^J#)vd%c%S$v2+MEWxsAq z#y|GLD(+gl-g#qae{5AZ{bsa%`am;Rxd(J+#c1$HmpPJYC&P42#X%{Jq^lm58(H7?PD>NN-cE} z5^?uD#J5-6Vg{h~?i@%k4@VvJ~ngxBfZvR`)G0jfI)6XX%4z$Da3ttA)=IZ77+jjTOXYS=9be_Ye(K_2MqE#lNTq9HN3uTE`6cV@K(+4&WfZ`aY zl}W{?+A2V@P3sZbr0;AEKPqgqV=-5QCkd@Oh)q8t3n^s)|J-4&mUz}WGB($h$X4|a z5pze&F~Fz&30~lRe8HFd=j1zWz_|5C=&z(>Wejz5l_CZi(`K3e+Zs9rEiM!Dh_}v3h&!~y(pnw4t=H%tS{V(sR=>s4xqISNY;g3%LTvS(z?rPo z?5k3;cnQaVu4t({>@VUE@()z-6=$u$-z$;;jvUf#ckAVYmIYq<}}h6sgz;Q;~;nAPT?c!|&5u9&{jC~;n2X3S)AtVz7-Yzf+Q z`^1A@9(v_VhOc=4rn+-?v$Q#>L7q}HJX!gCO)y<_?~ls~MQE(~+l?5NvFcIOsV+fz z&pq2?&Cd1kXy_T3?cFe`KG{+9C_Hx`xUl*^KyQlR_b*uwC#rs1)U#p0qzS*q;ndK2 z{NKHNiRr!S>Sv*weShrlQkLv8T}DU#2QbdU^{3c=y!}JEC^(W%FE7tGZJc{d3j2XY ziO22-L4UT6n(+w^asQ=OiN9PGkhQWFN!6ekGE#2iVjqxu2JGOvJcAGW z*>at2ZrkptPU~k8my05+X@;O|uJ&1FlQ5!tWT~TtqItnmpEZ1>@zVQ$0IB{XwzN6d zHh%=$r$G#!fB&lf#;pkJI5RuUYH=3vc;l@?$lNZ0x(9X`ad7os5^)aU1|+*u`}_kv ze&58EooxOt+)T?mzx$?^GZ9Fe*R0ti9egt2()N=5SMEy0g1?Fx@2W&>K*wGNaU8Q{ zfB#d3!1m|tVeH8yZWKjBYLrhyV*fZ=-r{*JM>3!-&Q)O5)|PGWsN5qR*Km6Db5Ebk z4mQ{CD}b zIuAVv=@K<;Gd!6s3-HQt+>y4!8V)_dD;O6VNc|6>@LLeEHyB!_&TVS}Yae%nJ>fj} zv6!ZN_|*&tTcZX!>vQM(?JmY;f;xVuW)~+xI@_37Nk&NgeSfQfU;hz$QHQtm`!K!O zkI!QJ|MiJ?eyhLR_#oi0J3zZz&VT1&rtvyUOxCWO{E?F0Z7wmtrkE`A=}&T)o=h*e z#KRzD$-L-3x#w={h4+OMUGr>R^;K$q`q`mXQQqV%8yOPKQ)sbE`muo8w^hSUS3U`) zF@%o=`H$a&V2Sl09csO5-Jz%eiH<(2S$lqKWO%Z{NQju_MoFJ{Fz+Pkf9J&)8no+0 ziUsMem5RNZVUuALK)UYZ>v*3_z1x&%{7Eplt8>1VFpC;vhq&J(Mw=}daWWO;QL0$Ga6{+a%5~apQ zMV)<{=r?I|oP_fh@_2cPG-jM6?nJ~hw%G(E$^0Z2=GMnK$O zg2p--?d$!YkV1BxU%vP3v)gyT*W!%)+^j@fwoUB;A#-gKFmUBQq2rnkOsN=5P_JQ* z6l~=n0r9k)^fek#Wzx;xG1i=G06w(qy>hiDov%Z(RE)a>eZ& z$bqDw){tbBBV`As$z*)NY1^KAwfLI7)XGUt#+PvV6}F)s@)}PffnpWoBXnRNT+z|Y z(wSIROwKWVlh=<=e^vUy%Bu?d@?JLW$BRdQV=B7<)o<@>Czk0~@WKwt@iY@ST z^OKb67YArv%a5F$!a3;>_DIsFxSfF48o0EKRi#@VdnKO1rhI#Q``5ie-RseFSNvid zE;QV(MMm`~i3vp}T3yZRNnFlufqj~MJ&?HT+y!BMw|}Q1Gwrk%oUn)9Y&3KrX(AV= zQUO3-!B@@?F%A!OeixAHb==Mv~aER8$*x;$0>(jP#q0X5*FrU1w%c z#9EsO5m4x9ZQ+0Ui|#({JwZa#j&A_NVB+mHVgj{`!jh*~s-NZOam|f(1@kaCxSDH& zp1pCTihOI&gu4sE)~qyiVfBfT{#eQ!v{5wZAN-ZVb;C?|tbsFc2mx52=Ici3mBd7X zNHcu2`r()VZ2N+_AU55(Aq%ArD`hAS^+@ z3BVXmbSy0%s-Y=|LBK|(P4)NburMN@n4(dzhp`Tm2PlXKqp zd7t~fud9H%7ix=3uF_cY$kG0SokT`Yw|{}t8;h*{Dt;(w^+52#;w4(}wuOToC<;Bsz@gxqeOPL80_cgH?Y(><3`VKTGH*N|sYx+7#RsDufy(tRzmlVzC3DNvaz&N|fhH{uQ=`o)R)U66l-&#K$Vb>9 z@IW*Iw&7(Sx@!nG8Pe&gUDQ**dRpn@rju+DSC}R>h!nJ*J=;LBi)`wt$;= zbaYkl?TUX%6ZZT;oK1~Xl;Hu*^5)1%&81{!Ht(5#R~8_891w;Mm5S&qA=dCxJp#Me zD<#bv`Lg?F7J|U&oOa`IVL7XLS>4oKyUaxX`Y5zzg8xH${|Jd!Gi?GM)wH)vGR~{> z>o$kvRaOq*8%anSuK-oQPk*^hOzE40PtiHwUl_cNMnhXgfdIBl^_j*4vF|sMduxN( zc|Ff)nihKcO!;m-cfCC+bt_OjvG5r)yv!;yRAr>4MtVv^W^**&-f{e+-%H@JV)xcl zZpH9|&osIT`8<^m@3-TUuAOVdg!c;#pe!Sd9xa9z>b(OEzsI{ZZvDzR?wK3lxQc^||6baw(6}dB2i(%TGhTaJL(uzNIJxodMI5-=Gz*F~ z_n04F6(8Uo-)_z5lXEiE+s;z7+TY85j#?2uANDGgg%3w;wrI3X)b71tK-tr^2%q1| z&9B+xgrHWB~d&Q)s>Cn<^a6mrdH`i#RTwT_{$Z_edbpN@EVYs)~w)Thm*x?+x zQYu`lWm8x!5fp`;9%d7L@m;H*taT@)sPlH3$w#=@-8h=(#obT3x^rI0=v^Py?OKPa zn-=;%@|}!jFvj#X2&F`^5HGM3wT#>#@vt&pq-mtv1X+$Fo%m$&S2k&}*I;`<~H=~}`^eH%_X+QKT;)!35oA)FIvK^Q{i99A6VZ-5Of z-W80-#Y_8=U-?#Z1l&TO_+~yD-#2X`3Wr?TlLUO?9rEbtahRyi8UNv^D;x|j&(VwV zFuVD$rts6+K!`||jNA~(D>ii7hTu3l|4wvoUGc8{^mo&Hao-r*6EL=JX-GGH6m@2z zsMZ`Q<~`^h1q z?7b0+jqy00<<$D6Hf{_X4dtxxi%O?DF4+C3I0EW(_~2gp8cO|R2*pOrZ5b|Lny8?h zBC~%?kXXI^%@}cqO+2tI8$gR-hKns4xP+_ZIwC9X4<$?gU~km9e-$ymH!?w-=^ypB z8P(XRb_dl;bymH8kjVSD4!D~X{YPafw?2pcQC8n1eKc@wS)C27Z%UI!yjW^u=v!te zW1t_bEc}Sws{f8xGLLGKef4*$&d9oPEpmXAD^+>*KR|Vo{Z_%-nkP4l9*7-RE^AQS z|21SjJmkLjzVE`+Vxzb9lB3t)rd2_X@<@J_TjH@umYK?s!4Hk(V3QQ#xDjLS z*1L3z`LGtYVL3B%Y_Fr5D(v-XAxtqj{0CI}{N}F3HqOT)YuW07#lQ!ipT>I9I;-8b z4}SF5U}ETTQhCnk9;=D^_YtZzoB%p#oy`E9^N}~7u$Z2%HjPlzIN{AeVd`A^$94q+ zPp1BI?sI-i&a?7LC7%IPU*^vhmV=Qq2XJ94346(zpc~)-Fjs(#>%QL5O$n(!$ivShF?seS0l%o0!c+~jzm@QI?y zkCwpQFJ{d8c%_I6;{pB4Uzefc&-9v0`yv-Jn|yIsqfaYKEBafL?9D=1m^LEBerP!} z0pXj!wG5-F=L40aZ!ckcvOU4GD|&kx3XfL^3-2e>bh8fqMpX7t>BEf&zOE`z3F4%p zn6&LPHJefEt8bD}6Khk+lR=2rvhlFfA@`W><3;L-0!91hUx$tSd)16EqhxlwzfG1g z*NN|<9~f@bzIuG!9QMuW(!o;qf_MM!7_tO=ac!1y*%Auq^^AUVS6eo78#gk&H$C2U zpJ}7GrZ|`oTCKcvE${IO(^q@erW>vKe6I#_H9|?}zwbJ#-l8&Jq)|y&?31jS|H`F3=gAjj0yB?Nl25%a`rn!_NbL&l$P?g|iXW|KY zT_F#Rx6l-%nDQora%g}W8Bv)-bXOO1v^nw4)A8Blk+KCznd_3q@^ht>6?P~3aKEmn z>_Ycn^d)1iBBD<2wkaPJ16cwVNK@HKf={gM zv4TIEpd$_P{b1aBI<#c5bT6&^A4LryY_^YK|+iUqT^Q_Ryv&mm+QG2 z$nl$($i@-VxFYxUkCFW%uwid? z9kvJ9J4{ocF9{M{tH$gzTptU(Z*l9ctnkyER0q`}G|QYHLe#k6bq;d4oXI{8grWLwo+DeLTA_rj0pJ-i1&h&# zImB>cO}W+DVXz+M;IY^dt~AdzJ!aCexWQzQXPnT;nvc zcnEf_5aJP~1xJlU$uYvtia_cfE4w+oZgw~Tl#RUsu;LbrkYgGKoIpM2X`{FGx((Cw zdKbEb6%xB+03GMMMhl;~AR#4z)Uxp?*@yNM%?SUSzo@TtPpCWIX`kjhd1UC3jMurV zS7+Y`IEqI`hK&I^WI|szf%ErmS%)SWynWacbX|~j-sA8u>OVjSRb>-I?hqoiW6h~^ zIq0uGGLu7Aukwt5ohKau!B_^>6h0NnagFJy`!i(0bwWqQr5v6XT}nfrCx8`JfuzNR z#hi)+7~Dv-f{RQg9>-pcXPPpoURdZ-C~WN*z4+zy=xwB&d-EyhF2yB!+sN-VkvY|5 zAgOQD2q;pkKr8!`YNOFGk9>R;=gC9C7^PgnY&xO->etk=q|k8g7V*pI$Z$j{<(Q?- zj_5m2XdeybXdC;p!a(pc_Rj-(*Q=LeWaRakYgM-Q&5kF}ivcBgv8S#E&*7J zNbDt5vS0?0J8_6hT%9Wx&zV)0XRYKV^eXCl2GmI?CcXeKS?WhJ>m5`_=f&h~vtFGh z&oWth5lQ9|mjQ9^E*7M=f{~NS`lbod1Tm8zIGK)Xf=khh%f_;#llOY2SZ05-Tz!=} ze-n6ViFg^&7=7S8oYb3&u;OKmURCl+Vd$Wl|Vz!FsyW?eIgnCC7(z^8fHPN4UYm@A&nZQ&3Zkwv?MNfc?QS zZmE?NAQ3k(kkh5=m}~IUt;gRhc)L+&g=KxZb;X}rC*$w^U^b?%6;ik;RG52PJ<5CG ziO8~=kNyHz(XS{7@APeLaik;L$uhjZsPC~Q+CjLYg|WwA(w%pTN9-0D4p>Sr@d{ZU zkDXY224vjqG5t8<(w>x=;(Rgv{PzNi4K6mXY{$>DoJ_-bzuL!i1=B&9$ut6X><}2K zZNH17vq9?RnhZ#zpv|q1pUy>S-d@6YfpqP;juB^h^Tx={UUu3H7F<@Sa8oeup7_^2 zS<@N_DRbme_E=DJHK13lFg?1HtkQo{w49D(Xt>OBPVRK;5wEsI_vqCFsU=HWXx0+= zK7;WcW>5IjDk}emN=U~s&1;)@=?JMGp-l1=zPBQ)pbRNXp}^)s;%~F}@Gyn!rIe0~ z@#xg?6lNMYJpg=~7tV_<^OYz&=!?A_w3*Sze`7?-PMgg{OHJ5iv(^0~bFhb2UREP> zho9r8L*3)%N&ULt-(<&IM=t&3AGwE97Hc(C!|iYcM=9YptYb!d$p$;t^%J|56x)2f zAP~Z0^+7Dao(ID0z}6sb9-ck7j2;r>FT`2*`8GH$ZAo~39{gb$G<0j+?nUXDpu5L$ zWs<$)yK$+U`zyVRQ{6wCmVz2yX8XBz-A~|w(ueLfEvfx=4LIj!vCMtHX{+zE;|+E= z8xmQ%$6H+Pv+P_!VpsncwBc`fD^UqUIz77frc}=@3{nS?p%QK)xG4`VSLZwDeQu79 z9lD&ZGa2K3w;7l^blbR|4H072vDRd`&PPCjlT8MRCRoj|5ijRwVE&Vm)EwWiwTvTr z-Tg;M$s6wCSl(k3{!p>oyNkA0uh2}|lsL0j>8MMHU;dn1W%+R4sG(fr?uWOet{(-d z`j@f%r-lUdV*5zn2arKV_$!1XB9v<>#CqKB%UJ7==sgM5`inwr5G0H2+Dthm15e7h z?vEJedQFf2Q7iRI6SwbJ7ip!YP_Ha9T6vXH2@J-7J8`x#gJCO|SbV$BE;-fZ{SgY0 zxcTc+uRh@^8G9|2;v-_H^VJuRJdg9?l}+ycGCy;-yZ%lBwy{iHHukf2*dn_LIh;FA z{jvfbiv#zvAPdIvnbK%izME0usO3lT6^~r~GJP>S(V@w7VMBW0oi{2^(=DY>4eUj~ zq!00yLJg7+s|DnwhHLE8))0)z7Zh_!$eSlFo7?yHEMcVn-lpvdF)i?@`)iitv~J?gw&+C&uNC6;^rVa*Z_3F4 z!3;$&kZIEPsf5fApqxQHNm>2x5+{ivK?kZU*=2S`YMzXPV^q{=_*;tW-i2qYlb#P6 zxQR!>L^jP{+;55)mfdu;L%#IrP&v&*Kns%)257qo^qJ7_Yq{Z&m$T$x5tj?soRt2z zToTkRZ-K+ckhkZ6Cv#rj+$i1$zZ9|Me0-G?3E(;CLH4|0y${Vxx) zL5FfV`zKEmFEmOuj|Y{{ZLP{be!;IART1Vr+7;HyzinBc1Hh}+48>twmPEYu81a5a z!60wu%*(4l{93KeTEJJ>E%Ib_+qW1Tme(HrZ{Uu8z{9FbGI?F3fbi*OC<42Hehn>2 z1zg2A>JKt*;ND|9hvkAD_#(tL)$ox)$4M*k-qWeSq-Ks4_o#uqh%CzYY7*sN5ami8 zO81VG<$k}9zT|j)@{`YT{hPSO`1u>u&;Q>I;%o39@tbq!=t`t1KZBkFK_UvH80M7bo6IfcOGr2 zce)iJy3ngOVQ_WQMj}Mj8rRhXtWih|pa_t8SYCBpOud%yQ{8plPRG5D-p;)g_zqe! z2bCEP!Ov77T4-BtGv0qKB`(|{6+JV-HT$iYThO}JTJ{ZT74Hlr+N3>k{CHIV{oURt zSTDR{dj0_8bm=g1UHK5fSGUwEc+!%Fdwf(Z0zWhYL0j@pi&jFa^&H=o+4?`Qpl*~w zoNV<$3Y-A8;&lF>WZ(zS4YaN#v+Yw3@(5gBZ=y;R1eXgp_v%ms9xBjujEqi<>8%jF ztm?BKv=%+JD=IqCg{IxmRHl`ULO!U9;65n+@!%7>4CwzWRPng2cHk0Vqa#9MBJx^j zHCzMZ7BFjl=a!dvujzDmyLG}x{&3SzATF-7@JfqLlo&L)0=HWbGt`}EF}n&PN|b%7+;&~(kF>G~Gu zK%n5CPgQ{vrMWFqi9UU9E2E5%n#lN^4sPK{r_6>WDgFo{E7F`PqPL3}#FboMyP306 zr8n_1b&C7TvrDN4vzQ@0uKKiQ-O;QQ+SWJR9as|ihWy<+_F#}A^XrF7S?l8ELmlQ% znLj=(bn?0)-pN1-?FVt~s&aRIw)aoQ)V&G_5v$%eIoQi}j&_OD+)m14ZZMMivhZ+> zU!?Gf_{aBiGjfbGqhgAiyvm;(rF#_~CHgruAzO<*~&Ssh?IU*a<0{o68H1 z?d%p@f{4q0=Zv|D6juS(aup%@r}SOkI*gv)69epc>8PxD)>&AB(7 z@y9)4l3|-C)v-O1e7TyF{9>1PMnKSt+kfZzal8Ir+iEyZioxq}u7R%dHi8{C}zrJAuI7Emh?c!hnGKVv-|2zVHQDne(17Bfw> zjWyxoR%4m1Q7W6?T6PiOwAnvmeHD;MPMXwI>jwJb!T_a%1mI#uukJj9{5VyJb_K!2 zlZW>Jmb9~N7$t&6>SIaY?J&MJRP~qmlbB@pdVT-7Y~DXmF-j?|85#lKLLitKXh!bnKY#7*2Evu_ z7rwh3FGTN;nahyfA2=@^HaN_s0=Y1=?mc4UzWc#{k6#l;6S1Sh4#YIsxw0Es@fQA(79$M`m$?}-9WHCau@fP* zFm>vIwDPI>*Q9T;lgGx7u@>BE)X>_z8^ZQxbJhkRW4V;sYb5Hb3BKUviPr>@PJ(Xk z+#BrU;ao+4sq|vX@jtY=5(rF9{9*WIP|GDzs}3lN7@d=?D*`QRb|AefSW8l4<(WnI zwF_Er=#?eBOwHWYSBjyBbRWHy?c==6v^^yKPv&#{kP3wbfY}k_y;&T~$uOdDq~#D? zq_RTaSEX=8l}%K{6w|BpPa#%_MYnTPmr}RiqYz67<X=GpL@I z#Lynis*jEteM8sCCh3WAefDZi%^VclpRU}ROXRut_^(8+@U?*z>IC+*^bbdDA62pzhjk+tcK>oEvOZa^EN-xVAg)715b`K}~;P zsVdbLB46qq4+RHSK8TNGF$guk*ph5n@#WeIP`%3$44~O`d}&$WVs4W6?M32lo}8|B z9FH*WAUAOl7GeHlN0RIwr^njV?)60PaPn2U;oFP=rT5f5RBFtZ&^ytJJz}uhB<1E$ zDMJ82P4fGxFQx2C=C=fG@G7I@x^t*r&= z%M~muHd&z)9onnzq&an!^=K>MV%)P|j@e5(=$RB0wbR7qwKr}r5s!N){`F+r>Yv%b zw!_}WhTRqd96xlYL>pCA->BIMM|~~g-V?d=7wc=XZv0Dk#hlF;ZQz1s)C59VuMZ=) zJb^-0q7I~b7HvjI0Lqv_238~pBw{p|Qw`|`MmT{3oLtK|$C9RXU8bsR^%J>3P|k(G zkH4qss5)W&yl4X4dtH)ZWA#Je%0S`wTWcC-aHjHSzqEQ3BJNeSXq^qI@Oy)_rd=ni zGRHdUu+w`12npK(`K$#HGs2l6^v@RGi$sd{P6z$UT3b!;_p8_X;l-pMs~h)C7RoYq zR7ynLnhx1tR4V8~i?fw!Mjy2)cmD^ttjOxKLb!EA5s6}|wl|;hQ&fu&&t5pfUb{0` zMv)M>Yx}oE4O#q2cM|dnj|J`&nB@y z61x5C8Il9ffyD8k{oJbSpepe+rV+&mruLL^8LUNuzI@KzqZ}Z*{YtUXz9GouGZB&;u8$$_|)-8$uQ^8`)a*!V2D@kNdK zIKpq4b3&-kg$Wvatwyr>eRFeWQIyd8*@7qk3*LKM`MI*g|<`&HRT8wSw zPIuau$P$IuF+|t&Y7??TCCdA;s84g>5^7BoA#djMD*%N~?W>~QVn(VZ+@=!RCa@U zT?&$HAkuAOQvmC3%tFp_lQMnb;!bSer&|TyJsTF{Z|PGwHKMkCBkCNEdIxr>L_IHw z<@xSp-GKp>vx+*@=ykq$ljbS_g9_wgBDn6cTkuo;^~|u(W#0{M7n-A9symw;$_oJz!6bh;@iJt?d!xqIcA&(~o&v4giO z>HQ;Utv7q!K^;s3Qo8eErvCEgEdxIy6UNKAk?Q{xau^rhTB%rMbpXkdl6KnEp2}to z{|({WgQ_{e#E!jg3+;4NY)bbPV)(>p8M^OoDvkh)52cVM*D|d>%bNA5grdf#?9ADx@j1=R}2mh*zKT*)#;R6I! zrdNvd%X}^R_YEPK**W6Xlu$Oi9(r-9td0W^vzzQTo~H+Pcy|L2>^!Obxch40-@aRF zP6H~0^Tl~%O|daSS`b01p=5@NB>?XL6Voy6=mqV@dkpN z0S4Eor~iPa@70_8Z&`0mkTR#*iE>KJ%b=VZrR1x$mNSOvRIL1PMe3AYpT+C#lyI|- zUJ0v{lDnz>7gjK-`#}E?AfCp)xBh0``edhd@!u7jfg#vy6$(%bUj~Cr=$)@ME(8wfEj))0V7|Sd~pu!_j0wn=wzQ%P4N^ zo;I=6W*qoTILA0IqN({FtAMzPanpvan`OLShz5k}R?cE&v z>Bd_UWeQQgg_LZ~XsBVg_cjf)0qnZWwS9W<^WIjmk%pBT5&@yI&a|{lF&$?`vhiR+ zfv(mc)`MT1j;N>D{{VNV6vAu`=w8H5;vbd&2k>_4rd;b)0pKp)9_!S#TVL(uTW`@n z-oNrhtp3A-)XTlxS|DJm**y1Cp+vS`cweG}m(A5OYvHwW4`CJz!Mkw!30L{XUY9Ks z;J_xun-sbzf1J-wT58!c%TRpi2#>D=usfaLd0rFfA?qZq5*KX>p0{@`ha(MvnY1yW zd=|TX+MMo36Wzb>KymC-sd@a<>c@yQdxHTF%TTArs}k?DRk!iL*HMrlRjXw z6=ONbrAHw>7PqJ>l({ENEaN)v>nQ_nqvK!zszS8=C?pUBy_gw!$_aiGxZHGqkrr(&{C&AB?a^-jotQA)C(gxzRUAnSP3LaAsi z3LX<{s3_9N|MUT2SRGe-^NR%E-^1^d*7fcg4RV$1 zT%l>K4BzO$2ZQJo?9^`q(Hp{vHlbkH3UBl0UYqN6{ij6&FahZv2nmDRwQlJV)7wTT zyZE7#4&%Y6o3{=f?@sc5O$IW09N-^KW1GTH#4nPybr&-v+zH^AsnJX>4>> zk*yEkNo14V(XGf&yUNBoHvqOFH(K9^gMCIBCt1R+(&Q)2WjZ?c?_-ptJtnz~RW^5x zA%xK4!_FQx6}wbl6#$l3P80+t^ZzVojh(<-g}?eFjdYlb?NK1~H5Wfyz@I$1BL2`n z?@_jb&I5mlrT2B7H8M3@dx0~?9-z7p0Oeg~%~TKaTYN|v>p-Va6@5#u7rP9JUKi)s zing4iMSKLaZLQ5sZQm97%){h3!9xbp zva?)trcl+B?^P&eUY*}_#>|;;YA-LjWV~&E>2?ga-COVX?zF26VNmbtD^@}~t4E9j zow@29^@pPt=pozo`tbXUJQgffG2$$4w9Vocf!`(6_*)CkGE3Vt2QwdiBVP6%m8v%1 z;t-6~-BICMQ;+A1^WDhFZ%jd~?A&Uu7&Es)roVEYvKnC>reukOjKf37s{|~KVcee{ zYgXd%Ps<WnU+a0$kV=QKzh}5d&EY@H7$xT-{tWxUqB?!GC#OgS$~n{$V8kl+5R;StPWI(PwW9xb=T_Cz@yMoL9q|Mv7^ToL>b@Dt9B72jH zs5cyBcZvNq!CW>s2m_{g$Ji&oY+m~mJjCM=%=rdk@Li_L# zcjp!JI#YBEHAvP|XZvUuKOyy47Zz_Wt@k68stE~IQOAz6RENP%7@AGC9Phyf zgilL2d)bA@X*veVxZc9cl=SbpA9wVqHJTieyV8mOI5}MxS~iEc8D8bZ=yd?0q974a zG-4V+RbDQc2u7#sJ+r^7=Qz8$uN`-ZRNwpg>E%&fzzdQz-7!&i-8f=D4K1< zIohLeQ;&ULA*Ho-vck$CLDe!kdWS(ZNBg@Ncqu!Av8$94;H_r<{_?%=%0r{f4MF}{ zV`(Oe{Z5PUsZUpQ`KP|d6^N(Gt`yJ48ao`MnaLF0=BU@Wc@Ga)x_Ngx*Hb` zyjIWr5$`pD!ZtoC;^;QKhf=*%=i{uiO-_A)RWFFE;|b!h9t)4IkDl3;s|G;hd_zC= zOfT|JthjcJ)bHGVpyJJ2f6<$F-o)OrvbiIXRy6;Wsu@F_7UB1R?ezysD#)){QwGoeQy)5J+!${kQkd|&3^)*p|F_M zvH^3_#7Ho@#BR9@3$5<3{$h3v4(K-423>Bcd?VQ6YDom9JFOOND_z93D)^)=P916Y z58ee5!fh}=ZbKYU0mAYv?)|`9Rce=)zeI?=FHtuS0uN#4L=t!RcbENSBb`-1^lh-<0w4923|7N6BolzVMR z5=TSFGO#L?e3mQqW+g=zUs(bgVziE65k& zE~60Mv#Cu+Tv*2!mYBGx^$Gs2BIuYq&+a-5OUVa460zu9!oO(-=4ka97ssZcH5Fa* zoR>K%!6Xm`r3%S0nh#ui0Vtw&LW)2oUKWIQQT1_vux|%vRO&>{rd~5n1f6^ zg_km}Ib1awQJ;86V1$gyw;!0`iza-pEQYaXl9+0opF%c+FGAkj99)MgyMA!BcK3U! zD6wd~-B9ETAL30FEgt?4(DU@FUl2?tWx9g-KjNU&TPm(bjZ%%FwAH?~6W z*vW~doJgZyPN2@8t4msC5F)%co&u?Bd^OR@=7rJY2THNYc3W*XYd*XzMX4@Q4$~X z?QvRQzv9Z=K0Zjw7jgVZf!Xn}5rtlu$}Uo+lXK#{lz;if zLv*x}h{{=E1Kb5eU}MiK*z{Grc6QG>X@OqN*Mzn~N})!cucb<~zW)Y~(5)6P*L*0) z){mo{DXYcAuK1qyK5)gsrk3u!vDPMTPB%4}iVyp z1u0SR9&IJ}BoPp_SKe^G`$_ZC>(7|UVDe2QZPfUuc#3=~&htql#YErqtAj^Lm4XQD zpXKG(erApzOqo_rO1u7i-8YB$4$C01ng(p4+d~SD>GPfsrSONi5ITDBz`kBoq}#WK z>%wfJ@(r(+kbw1MealPld6x1~!T`6}lhTLnmm?13BDOO4W^oN}MvggYgJs-w>q5jp zE+_kX-A>kMlJ_+qy&{oE=+R?Kew*Oc{{S?Ofd^?XyF)XNLp$eq-_YfBA$gd;s_c0s zpB2<+cWkHHPRI_AqqdXY$b|E56Rt2Gdk}3R5%W3U+*v=GS&2-N)7A92a+}wahhxxv z!XZQ65Ot0L^{%$oECJIDDu(csp^!Vh7UqhvqF(v&{}q4C!5?-pRqnszCTV4JmhiU# z!6QNpCMSy!PfklOZi8--+K@5+UhPEPv#P!Rm0-6{G8hblkp;wt>2_X`*Q+cedk3@yNt7d8O}*|oM3y#OMGc=Qm=a_aq@YD9LrvL`^Io&CL8gAcGeB= z3Wci{nMmq$K*OP0Y*<(88y*8}haqpmCbaZWas!atD zC6!-@k7~0x;Dig;d}@z3^Kr68R#*^og=Ij%{m)k^Wh&*XwpZA>X5=J@ zT#(cL^0%7@{l>Jo?fBcjlKDUA?k(KHL>%rerndW|?OyWQi_EW;4)_W7!@$+eR`{%h zBzq?7^!fsudAPXF*&eN0XT&=^bib9p6wGhq7Z6%49$#@9x z=TvDtg-IbmC*L!?B;kYjoYQ06aiHA!&3INUlu&_s%G)Ma;U2W`Kse6zD+r{{J3Fp! zROWds79S$WujMOl`4E~7d7f*v!XGA5SIqJ2BIn2mxHPELUo+wN+ac8{wRpSHuE^CO zEH#>E+pd>26t|)leL^*}D3_44vYDe)%;wr2bntVxzHqc@ zi7LN^nC9A_9!vaTNp;!xxb#m#GmaqO-pP{oWeP3oSG@ifRk<;a;f?I8 z9xeH6Q^MEY<(RKzsb5U+!FG+-;n>MzC9kx**$Fjv#!D#>{%mqTsLD9cXH(<7OP-|1?`Un;r~XN@^b=;5lAo<^vnjGqWBp>E;IsV43-{;lXuQp?C*$_e zlWm0~jgbDKWOHM2nO*#P^jyoT7?hb~E8=pb=tiBA`p$as*mSl1AHy+W0mW|ADc7+l zCgrXhL5IO4!zIbUB0}FZ`jjP7b?o6U>peM>bCaLbL?^+CJ*y`z%dX)JYafj66Qx<6 z#m-JO2igC=$FR8O*|EXVpRf9&Oq}Q&;q{W^gUr?<1(zG&%lhOq`HA*){V%Q@H%ae5 z(%1-zwlNKvyJD7;Qe9Wj&AeQw+t06xdsNF>{vNt3n5x^mDm4C9;V<@+Z49I6Jd1i$RIK0F9|V>9-Z-{rk?vZHud#pDeuw=1(^;N=-JVZr`Q}u zAb5LnhO{(|-&lYqBs4hw2WScDT;`f&VlEx%zt#43@vfifLti9zzupI-_BjyZ*UK9A z6DOutxGpa#?EJUDj!$^mpNabF;H0c#Myg;Tw`kN;XGbBjy-E}tiIydas#vuTPJhEn zUP4&du*c&CcWxhg01S5_XFVqJS_ShCDT4rtj0$wS;+;SSAYnSItbQE-|16N*F>FWUaJ*oFV6^2(GLbNJpB?m>Ro$#USbg8L*UTdz+ zc%ZUil#v1vz<}`5mhZ)kaK4D9!pYGl88+ZbP4)9_W9yZb+`}U0J`!w4>~j?LcVP4s z5}Lc8r0I5^Z5nj&bO?|8N=BOXVS|BnTN)V}lrQqlEYxkpc#H`TIR7oxLB?_DAaW|R z69u~a(*g!ruacJ2QR!u4Th@JTYqcCpLd}i{an1uUy28uCuBp;1J;6i@hgl<|MnM#% zboaFQSKPOB9UM**fsX!V9)%Ra@eCtCoUue4!&1VvZ?TI>-w5Hq#{&X%ca6>!Y#qwj z0Os#};0%nNMSx5ds2B|l$O%kI*ym3mWca{`LDMPzvcGuc{|GyuD||xF0+zp3$)TRq z&zl&Z>;Z(2!mau1B1_+M*N{>Auc?@vWGIfn|5E-@a7E;Vp{6g*?*04gA}R`NFy9uD zDrM9<6cR)1o(IlC*)+(x9z zp(RJnlV^VS9<-w4-}Rq^MotHF_8L;7M@Z~ogOi85b9y=~`D$pq&r|cT)nJL-m!z25 zFcBUBJL>`UzzYlQy2sSw{s5GN5R=;Hl#!5%AE%ALPUz8;KNDMr94;mjTa3bo5XM{$um#3R@ULjgQVO#N^? zx;T&|a0s8y4C49zyEbf~27B=Fg-?)uU^mWK1~d1;)yY`*kA%m)8R`1h^`UPXf&C*~ z`qnq+BG@Z|&q0I8ALivL)THyci(jc}GW8mg-({I=Lsmbn^x=(qbHmOO26a)+*sME) z8Rvy`uCMRu)yemX;Nq9)vU$KP9Og#|-6CqY9hHBy;1%{VN&RxuXj{s?0z-EWr|V^? z2~?AumikkfJE1l+S9%H0F5eM9uHpX4FNF|uPXeT;H#0Y(^6;zNz6m3h=uwYNy%-a_ z$*GgDlZs-Tju3SFwoA3gHByOo2&Y@MSQV8S069>hlBjTLl_F5a!dY{EHni5jGFg@x zO`3dTLQCaj4@rv!(a6YaZTConAmiAmu;S|->#()-gF*lmA;sFVg5K86XG3pm0*LDe zg)JQ=B2HliU0ibZ1P1YA`~4;xW(Fj-1Uw=H8*<{NcVYHN_rRY&9&y&{(^hlM(NmdM z;Sz4XouYR!0UFO*&8OM_*1$eQbQrx*75tixP9;N z*Ih@OxPSakTSt=V$QwE&V~MYAYLuiqI!Q_hs_pCZP#^+l*Kb^4&!x^&u9uXvYZ=7!^`wkY(~T`oQe1;nhsMC_GNSoOk)jygb1aIi=irw|0bN_PCcuHNzGD8MyGh|(1 z38s$LU#guJ@dVbiE5qfQ!6gI)A5WpDib$&F>Z7PDHu+TCdmj25=j-RdLF+j(CW~Osgc##JPu>?S(bSBJ%B8^bK7Sc?N7CZJMx=!EK9hs` zP@CLAxcNK0TOPDstc~+ml1+|hp5qr(>B3hISZT#TvQ%sug4HJHOr?`AKB7W;+ic;@ z8#UgcFn;5f8}G}0*v}u-h1H|pTPR8gAq2}CED&MDtjEVzNR_`{M>5Z~I(nHdW`a~@ zqbWAMmQ@w5dKr9rE46=>pAvu%$I?4O-7IiUCsIK3(KD0K_|tTwCXGW zIQ@~v8GYU{kaY89kyO*$keuKnn^e6GiR)YjM(Gw@g3i=YNSyI-aS?0sQXcHyN^O6y zk^XYXU=`6)SC8ksWEUpPVlk^7MMH(2w_ouwiwe`o=<4}i10*Y>udLu~_;fKYZeaK? zh3FB$-^JcnY!8GWyn1YW?IHnr)t*|#4RdR!L@<=up3A&Bq zI*XG`^VXxIb-!+~FKb#f1$zRgi~}LJttM zgqdx?uir$y0e7xIJf>6f|J@#) zVpIgLbqMjCS0rkw=MJ_w!jPEhlEf0YYu{xi)}-%ALdaO6+gwT(>xLLhT!4N2zrf0& z_GxOZd6&!aM^dkyqT-#LB4+~ij~TT09I~`WScx6`xry%^>2Z-X;1&G3cDFhqxGf$v znu(TjWu=F|dcCVNgX=vKl0Hwiui!m%Q7gb?3+qDHD1O)m{oR`_)03 zJIfHJ!M79U=c5f$iLuAd%ed0F|KsRf{F(mW|3AkeIYl{asL8S9SR3Y4jya!0DwAVm zhL{u)J)_m%1 zrvYBWg8}Cj**JcKyz34M1pKvR<~PYaE!~JainL$rbE&B;V(k2`xiVeVQIeQl?;BqD z(+br^=tA=!Yk?pBf|UkZDv&bGy=L=>OLd`x0&|Rz#5hC1hUG(VVjR*=H=p>ZtnGh* z^%1uziE_>@=b?#oUn#-l{tl#v7y4tsj{~=F0k6E$6$Fzci}U6*pBV?8S8gY}wI?_L z{y4tl>QX@N9q1*?M#z3_`5u#zf<~~S?VAz8FKB=4SsE9ydp}?0p{z?H*Z7uUr%Ke_ z0VTpNK2H+P^Y85Nul? zLxiziifK;KV3*5t50uj+F0$;6y6>okJj-l^(ndbcHp=CRb|DBOOLVFoVj|C@oimn) zGh!hQs)ICD)3y_`J{mfI0jR^Hf+ljvli7nX|DkUV-3NC95X6#^)SID&KD8&KXZUYj zej?@i;qe(u#jeU;P1dEeg|x8h53Qfn{XZ>9eDNK_2yPP8e+j*)-zpM}NOLp1?hGSU3KKFm&AQ1NU%q7K9IiVt|C>E) zmU31Hf?cYnAzFC(`rZ}{o=S^8DkXAwt>sa*9vLl%@@S=_ZN2-<4m3&|U53(USls23 zo?X1dFkJlr+lNs?91bQ5uCZ zi4-S9U6y1^0L+Xj2&&Q`B2&6D3BiC)c_8~D1%jN6LfJHcUjW&?_O17?A@F01^dCb) z?U)HBA$cPHG%$Mp59yvtoy=az4o4I{QODV*da7%fY4q(op!hCaHupGwQ5=|(uVXg9 zo-FjwR1Qaz6y;L_vhUxsIvV`?F5pvCYH`0~U)sGJ{^#v}-0kHpbErg!R-aJEO#is> z39Z^YdS1sIkNv%VSmon1_>E{nvC~e}2|X+gnz4GWxIg2NmFuX`h5P_F>|0Z6yp>u| zCW4%#6LU5KFgw30f8DX*`+0HIGJLp4){4)7k+<*P*;#Noxm`;{C&?TX*Pr@a>t>dn z;qH2#ov{|mwz1oI0KL%k3+hyoz2P+P9t?z4SIli*w(SKgdzof%{$|}ZVP5i zYi!*tu=X#TK>nq*JkC+J_L!|TOJ{EKj!qP0KH0W-2!srs>o=;SzUcbE{0qNo zddKhaW7AwkzZ*H=*XG>RFH`Nf0lX57A(>RnE>(zX{4wuTV7$gIFefAZMByPx3Yr&PP8M`Akm6uM&xv75}x*wv083ATEk=@imM zr~6hB>&9XxKnm-JMh1XZc0!F9Kr;6?*R0JS&&jQb9ux`v7XKgMx|$y=ZLxr&=R2V1 zYC}lpUU-$B(bDOmYojJFCI@9$FX6Bct~Tl~n|44~k2>xg#S7y$`jF!f*y!W72zVdzY z-@zXXGC0A}yrm#n^}ajJ8P4@?AF~aMVgciBVFN}{5#~IC&3f=b6wsqa=DHzGC_?lYfC6a_O zEkLpimdCQFnAk2^*C@u&KYp*?DM=d@>444m_oJE5{`1b6PG5bDBzTh{S$_eXKWZ`@E-Qo^JWm5h`Kqsheh(Y}AZlnoprCu{yl5UZo?t-n3ko_PU z3M5ld=2ziXCEM?E*+U{r-BLt;5gb>G`HK;J;mH< z0NAGhfYh@JD2D*ot?V0M%(mB{OqwxwDXZ6hzr0+%geAa^)g=X@dF2Hnex(q!sHq@R zBUcaSDCdR^j>nV`?lUzgc)2|Nm)vqg&QbIC5}&wMRbCwtPML8^Em;u?hgm8aHjh=G zhhBz0(ZP31h5V_}OX7WVt_)ZQK?<+Uhop{U0DA-5fY;Ke+DF7k)|c z5BTb@^2ZMa)BRXu)U5sY^e#Mg;MtpE#Hg-%>3B9?6N;F&nRxi9$4>bhcIC`@Po{|JVS z>zX%8R%n)S24r-jKb(Ej!^z=qP%8TIH3;p5eQP8(9rG-!*Iqa~6;6<~9i&+D9%(G& zSljcyq;};@#bvRkbII+Nt7H0u2TDo~;U}W<_8x1ZpS_rw!v;{7+C-1(c0Aa9|d~`9O&z&GNMzcBfBzQPB%$ve0W122TL$pT%>Rn zr>^Z`>C;w}IXb*B@Ta~)Ge^$uNQNko{jgKROSKoOo4GmtEJ#n>tIzDJ5d@I_+%+yB z-my0b?oXO4tmOpkyQI%)XJBIQYP{P*f9UC431#>uOHiE8)aq${Pv0_wdTaZimm#dY zIb9zn8>^V~d#a8Fz&$YEz=;`{DW*x#mzj-im7wl&t0(v3-B~h@9R`kV^cTSjv3h!F z!;&v&&qKI4}mx`bG$qd>kngk{hqIkR>=hYzYt;jgI3hWv~&U|yA& z+1&a~^2@>3r3UwmRfUI2@AyzZHQdd365)wDJAQB8TP>hRI2A@)cu}pw34r)T5kWc4 zi(jkNkEqgw#roSIlS&ck6uRqq=Rt`OQ}=PC&S*n^t52*0$(uQ~Jdlt`(Ul2~U@Jg> za#5YpWzf8w^$OizJEp{`;?CU_#`YM%Ri|2bLg+~=j`1a;HmaLbZyWJ)xW`6N5xqG@ zSWneor}<0qtYQ3n}g3vKcCh3vrRr&aHRfXyOB&!O`x|Krk;Ryq!bCYR z;sfcGJ=U~UhJNh$a!H8BoMzd;JeW2hSM1+W9?`}BG`sD&zVvR+?T@Jl&8)?D?dk3Bx;4_3b?tlL3df2+q!A;?(M0p7AU%7~iEtdZgPv3MIUpz36(eGuee|$do|3026`Y8ebrGCU$QiU3-6uWHEFBxe+*HoNprDc!) zi0XtHsoj`4(fj(}xks9gra?U2OhD7Bio1yxXL7r;dfw5LL~c=<#|6dK9Rtl($#X|< z%KX$Ww@mIA4XA%!dKRW=7yITdv3s~_+u#?&Me73gie|HyuWT!mbyyw1^dzZf0zP$#ZM-Po9`Tcc^}nZ2nA z1qtzZFgd-?LR26VRr|%{*fd~FwtC}n-MQovL5ykTv{*4rg9cm2ax|8lHf~yOj_wpS zKDuU&2GBWQ@Ef#-(y_ttPj=@fZ3mGS`ahVo8ZqJh-X4&>;eK<#+Uo*1j{e+T!DMgtgU=y0@sg6&fBPWqMpg^Uj6vWd;1Dj zMa%Ir(l3f{_d?QDjVR~Hhk$q}o$~bWLXFtnfQ&{np7tLoM;+^hSNZ78q^uXxJ8zr% zyV)JA%zTp{0Kq2PK}_&`*VPPWJ!Z)Pdprm}aROVqJu+frF0}bO+%NE;_W!tvhb=kI zs#{0>$$C{CWj~apeW@8Xz_q?@F&oQ#>0*|p+_-+#w*LY+0l%e}R@lB1Fhf?!xC}?|9Lmoltr4g?lY$DU*=iaq0e5b&=V(O z24&WFB0#fhFb0J?tZI)jg^8Sxj*{fCNRN7jVAOC%3{2b`&4%4QPpkQ!yVCSL$%C0F z{4Mg~^8EZxGHXfP_(NEU$CdH-?^QOmoQJ*$>|Sk1I(V$;SsK_FDo62)?=XL{dDUK( zaCIB9GGSQ~r*$a&{n2Jy^)^!;yF7FV^GSX_yd3xqKMSos+aSaZ)wEZHeFua_NsoJcD53j)DoA|ik*4K{&WvNSEx7I8cx*`?1gFA`VU z1%8s0$Jub_CMYWT|0(&TO}DW8FnLNuJ6`_{{%Mf2fj(Txkca`?r~dfar8Ua84E2~z z@%Ol&)0(f+O;J*yOA?0v6BRp+S9p|$u{Kz{6{5e8skWaAEr(7y6HMWQ!|N%MjDDcBgD*BeYsHo z15A`(kiI|4gKu2P?>Yc8ChajkIhLski!v{y0k^mCTouPS=OphE|?_KmPnj^_R-8;Ai`y8+Y-G z>3&r}(M9380@n})Z#GcgFNEaFBQdky*$g+gy8W(K*;iUaYl@ial{*^wSCud0BEJ_C zqMZ46rrUlp?P}98Piuz+=UTqX-%{nG2ET1~aDc2lfKA^rjXbEzcQbos7n|)k+fDM{ zCCaZd1>&#Evkr7_bl&5?D@L!!;s5B@Xnua5*XND*eJQL@uG0+n3R9L4HU*x~gn5RDu%EbkiKwd+fK-tVVE0n204{Wd=z=}C6W|M=RDnv4v zAhM*htLV|jZdZNPt_afCg%VMgSgW2c_5K@#5tbgNu1&vf&Ce9p!=$k?hW`l*qmC6WCtis!Nsh zDHEZcY{jCAXFr)IdRy@+tpkl7wuAB(KReHBN8A1AwYbj5C1N$=iWOS+`JOYNZm=jd z(r{`rUzvPZrgUDl zfbmBgi|=2^+^*v{&xVevi0A{t%iTBK37*wmIs7NI3ajmiTiu;#$%vvMgXiNIsR)q{A}ftWnyRe0|_sSHZb&+Ea}8Y?yCbJt9lL zpM1FIt}!8`fsd*kPz2BH`8A60Y_2^@f7t}Y?XQ~}45;UE!@44bv7)IQ9F;0d=#t9(%`og( zBjm7=U{k1MJ>8*F;lrBBB#(ykr+n$p>*c8{h9A`1O9n;^oef2rie<)oQZjEaat%BaPfiFm?wx` z^?Ew+Ah$`^g=gz-6T(uow-{t636pIWwwv4`U7wwt?=7CoBd56X>&{VoS*;DMEm(SC zr2sAZK55%1Cuw43u-{IeH)6maIsv5vM+g$%26e_fT1}Lg$J?r3cYK?eZe8n{K9}Db zRaQGJC07&6rV8m;bXpy5-wacS;|YTEFZgFwopm?jG}f{Isq;JnseiX3BvLX{1!fZ= zSr=PtBf$NjXWh11qQ>=(=!-6^g{TwzF{GHzGvST+C>*0_Aj-i?BAmP5<#8G((&pNg zdpdUFVDW3tL7AJ4Ys~HNnXq3ZzVHHn0l}WOMiI%c3Rb4Nb<1t*c73Ma5hbEvW8ZFg z)X1T3fjOriH9IV#{F`_BwpJ&&j4Z~52J-r$_`f=5eu- z4LK7V*9d7*zCgB}BN&9u?;nzkc&)Z#g5iGdF#E~!um{3x&ukAxQo>3;OJ{Lpf6`nO zqw$+_UC9YD3cM|-=))(cOup-15B-gks{vQ&`LEV!)Ms2Fv;V5neq|;d!UdHaV~lik z`5r$tW~H1z7>dbgWn+I*EQ3dn_yECvTJ*eC6V9?9BTsKWFZrI#zEwuv+N$y2dD8JM z-~d0jv+-KASI2pfOgs{ivDdl!<+@^f5$A{C=1<__%iB_3Z79&H7)ZcGv>esEQ$y5| z*u6x*op$kaR-TkoZ)a5as|5+L_pHpvOch8W<)R}+24$tw{3xI&&PWlBc7Angv16cJ zuOxGvXj6>+mDaO~9i#>pe~K+iyQ##B5ScZ9>#7Eu8k4$w=Dgbdov0Dr{9x{{=RArR zYS_LM1Wz#Cqe}U`gvGh0{{h}V0J0csE#`$334hD7BJ=**Gx??<{FPhehxG%{noBCQ z8$v;8#g~124Zy6sA1Yj~4P7f2M)oW73m0+2ji5VU#$wAw;b!F6(NsT@8FVeWyfs+M z?m!&T2&|X8yiZhUq(gg2fNg0ku#)Nfz=rS}5Kcz<;Xw1Tp}M9pOJk*oE`tZn;6${A z-FvtXGaCzUy|!$f+ccmwnPph5K3eY?!E7X}o8viJB}^9GV!ZZXlyVB6S%WP2m|+~= zAgtO(<71*IT)CLKI!_k4j|vs2BPXkt%Fa@k-Ya+j;IKHWc4}EqC?fr*ab4PM>mo6L z{+~z$k?!6Sau)|XHjZ*2D(_l$HJ}$!SEN&cl#uV=yP2_IR^9a%cCN{ux9uYqUb9gE z$dPxmYj?LULK&u3)?$SiVkS9Oe?xjTB*4dD%-N#F%tfJXo#E0*!+FMhNyv)#I?%c% zu#T~5x~vx6e?RDCByf>^cnIm-kCZ39n~&iy-|~y-iuNlEfejcb4`cKDTvKQ`#Lr8Z zp5D{5iaR{S0|Bm9hH?A5A}9^B;a4S=!X`-#U5z{_mZstP(SXPL`g*SFIKlplg8<$k z^>pX5hQtlv2{TaGR%I;#->IxG0KDtYyGH~7sPt?;zM=M z?efk^jbCbXmQ`<8Mq}xba+uC*PpRfNq;4`ORwIL5U(Q`UFNS40m7i`_Fh2NAj7;}c ziJ7fwO+4Re?|gs9k{csY$qIj1IlHEjm@Ph=v#auc%Q^GU_| zMamg5$%7al7}*q8?f9M_?JauoU)T#8?bwo7_G762oL>mrM;jtBYwEFDe3(Qm9a0~e zo^7WRP&E{$wk91CqJ?sIdFf`~^?m2_w<#mA#Up+A?j*J1dRCViL~h+yR?J{2FGngh zcCaP*dP-huf^k^)c&e^()MAT{^wRE;z1Z~|8r#wFg_jN1epK9^}*!Nm@G9Eh~_ zo>mlBRj2b^#j!ivIF9(j6UU*^{3?9C zZ6S5q!q`SjQ%*9MECK{~SNm0~shnm1*5h}aNk1s$-6bj*((%RP1n0;9161UHRTXrQ z%siBM1)@H3^M!Q}*-E5OLRVg$QwZ?%EW|Ae_hK)BXIm6)V_+h$syyB&V{q8eLb?*< zjYv|0O zPw}tq6c;Ma=x#x{=}!o~uzi|(Vq;(xBNoF&y$@C87d6% zr%q|rJbc0Mct09lvX0gWjwG`ce5)7A2%>94q!Dv(dL<7PK#2cZRPZnnUi72|%V5w! zk>FlYY&oP$t!> zqQ0MiQgF60IAzR+ovE+g6PYW3zj<#danXLjSh1QfI6~y$d6LfAz_Obc1!KCQtt+4J zb?EXnk9Z>!!q!x@9s07yhD-c^qOu22o$>(Q4D&vw?2c>GvRka`(QpOT?zxtl+f-VD z0pqY(RD;*H2fC`zv`(~yKCRp~LFJhN_$k0qu|mG*S@QGcF}ys+Pz?er$F-YX@_=;= zVT6qcmq!q_q7{OHH}#>fnz8yJ$f;DJv+_Za!dxM>_6P6)@(%q~^+DwDt;D{;XJg-D ztPQXs1d!0P%!`vJPrm+B`ZyLyO$wQ9)0uEjzY-I|C9)TuSUtmVHs93-XqChNKWI~4 zQzG=NVq!N0u4JUHV84bxx+Y0|TKwOD?5Y}&Gt`9k#}g$GWKw2m7Ik7Tji7FL506GQiN{SCii_HlG`Tpqy#5bX>z%!~W{l0fVk7%V@o_FFbo5_( z`gr$}kI#qElCK9oz=E%ztD};FWc^5K`Mt|YEAK)cHI78RU#Y2@4Y-g3|Bjp4Ipf7M z&~{k**LUeqhbK9ia-TWJA4!0Ko@FjcOatyS3zz@-k(*OZFZ%mHJN^3^w)uBlXUQqGm5@>IKX(HRe7Ovm6ABQBxpSM@rl8UC&4dk}ef~Qca9wz7CQp3M{b_NCR&f zujOd@D~+4o<{SZnu??lq>WbdwZjTBd2AfDPP9^6q@xlFmHbP17uJcq7fs4xFMvU+UsTKdS#to0uZ9-1{HkuC%tU@-a;}CZvriC0IMDixj`f> zp^CI1oj*hHjt8l#gcCcTE5f1VwD{IRFe9c2CWARBjN+WnNZT~%6$1m&$S!`8Lu2D! zF{JQh%;@e9F$d=N_@b!=FC%>@GI;l{M*r{Rs~!DH1$mVqrUyk!SlIE$-kk(pN76TN zjNp2n`hd0_pMpVKZ#4|$PK*J!lxvSH;{IhyJ-DsUf|-^j!0xMB zfrzUVPDhiM2oqY_>LrL)s$-gS`)UfWnENdo5WMhsH1gdVzT zmh-^Mn)9tFBBOxpe>sQ?&TsCbPrL8=q7~V;D3wT3Q+bA) z1$^s;+!gf!P6Z@xQOEwhHV842+h%->w?>D96U;sD2Q`#g0|bN{YORP1k!RmBoLQoiE)msDyGIC>&Bf2c2-us*EO@eD>!Yd{&b%PfFTej%!ho&0g!wLvETpMH%!F#Fdp|X$!G!#kiH!kCrS*KJAs)m^G6}!xto#^6_>_>h$;<2J)=Xg^J#`K*coH-mgA$|(HeTr9z-UV|9jQXa#n zepq;)@T8=R(=PvBkZqX+p4$OsGKKQpVE<#U2U)U>b>Ydeb|ypPou%Zx3c`|<>;C%uqwRC6 zd#7W1Aw{2>9M{Y=3K8XHw46Q`2=0X135TvUiu?U2&?ld-PRK|E#%?IM111M;4IF0) zLl=%dVOcbXAb)c8;^iN|kqEI=zb_3ig)nEJ+F$ZJSp&&s`*FeOQQ3NYC{G_hFwe=G z2TIqJ;{eRY0qK@Ez&1ax{QdC=HIaAo;uVZ;R_bouBL{r+dOZXxR&l}A|9fuh*-3o; ztjUEynFyDeQlZbQt0CQhFc=x2I%sX+R9#i_@u`66p{BEp*FlY;8go1TJC&?xo$bp) zr#fvXo?1o~-BoH`l3r8OuaeAxh>LrrLO;3z^F`)i`uN`-xSFMWf4kL-PktNdjbFPnU5-?Nh#Jk^ zdl>RPwLm?wb&N++K-Agq=DRAv#>U#z4JU&HcDBExck3!YaGTgKZyK-L)MXVmncy%u zL6`U1J!al=>-UV0Ubp@dUr-nRA@}=3f8@27y5Kb}H;q*C8?G&&rlWlMI%^SBu7bLN z=-NVAuXmvi$)D8>E;_RdPL8?=6g8eIW!oadeGIoC0>%byrMtL=OhQ_oij|3fYS5xS zjOO{8d7SVg^V|!pyIygj<4ilzzrPmn)&37)l@H6%j@a~gSj5=#M-!ZSR9r9xU0?Ec zh-rVbO4|6;YYj8jJ7ubS%4M>WS8b$uit%I0e3o)GBtU9FkrE_d8yjpgI+CXkOtT(9 z)9giKN-`{jYCTpQ8e&U=0O0knx|bu~&E3_^ugkG*57eA{dTSRtQrGKri8r}btcF+} zbh#;iIdwTF-K_HU)-Xkr^5b0gBOx&nDV5ce2$0vCjtI z;6Qu_UtzSpD0|6;vf~gdd3IjD$aMo51jsW(>N={Q^@^*zSwL`x@%O4kY3zwy>(qgV z+|Hlr8AM3(i?D+lLZXgOT7P%}->Jdhq{<4@4inWWGac3dg$6N&j@-((_g+5Ze|__d zdPh~@5+csTg2)kEh(4D_b*6ri9WL5cj$1Qc@#1YfWcJalYD)V_Yie!-RU|&Gu-`T2 zX3ixQCLw;k8{bz%4qp((lZ5Fr4!2I5fcs^5)K8H%`JW-nBA30<-ftpahp)DM zd4$4_-I5OGpU|ZL15w$XSuf}*=8p63j!3Tn$YuD0{NxNzVWd{qp>e%iE-HRPf(HDNZURxG4zw(PNn z{vNDb>PI=iZCM~Syk?e~{d74+?^)-EyL5 zCos;HhM$;1XyzX9je*Zg_^zA_VgLe5CoJbml>Xyq{@}T`htXU+C<4;ThI<`ne@2K? zi>PX+0G%=pKf%i6T7XeA_DVlkrOl5J? z34VuzlPEW|m1^@a@oE zJy!!<2Gf=J=lkg8G|zzxdafm(#LUfV5qLn&YO^(RCZCf!f=f zuo`D>!*96-B1CV+kYC8jC+z>g3WZ)Va9{ob6tt87^@d)%;L}fymU}gbameJ(k&~%hEl;!!Yj#F#YUo{A%tD$Fw3By=8pTzXEI{% zYRbHAvh7#6-k-gQG5d=DFSTkK>@l}{6T+4@p;=N@vdAppD$S(1k?U1owvo>fuzpio zL#oaAh$XuOiBF@}b3LU8LHrE;SJEl2i2DU-PoiHpR)bwL^}1PAw6$Twe=idpWok6* zB!gCI{ zHPt2;eZ)|#WtTe02SLB=p*AbuE|@S_AF3<8GuD-){`ls*g4ddZ?_~C@l!)fvS6^A4 zl_X~Jzs?wsVJbriQX{D>FffXFk%$^uyh6Bz|Fk{mwTH3?EoPCj);atq^)BPv^;h*WrsCrxDp zp@^A?>Gj)$Q06ZH_ERt-{HMdmz~OSHBbBXFSEzAB*+c(D7pkYGj8CEr_^t~4+#u0D zzP->}iHaq8S9K07JS}MltfXE;6$B~O?>@oMhqq+X<$GTcmUxyQZLECGe~>I%3l$Yf zdH-57jao_w>Y-4)KIMBFeN%0%xwmtzJ-_+!HG6^=@A!YQI;~WG4XfaqPuXoou`I?SF$hXf-$qTXm z+YOz6akW7!J8|Y40UuO*VY;Om86!C2O1zBZS+mmlsS^D8+!#mA%k`Cgtt@Sn>D!kv z?6kbE^|GSDn0#!%ZSlkcIk53u%TON_qno+$8_J34BkUgt$Ct(5zW>K&Re!O- z)pitAK!Ki{#w)nPOvh9Y4+|CKhNJx~UZmulv&|kX16O2q9Xq{9PB%7_dsj9yA3NP3 znMtGLxx3w|7t~+AzKV|&UU(C(R<7&#?8TGUo`NwU$hmiN&FS896V-~h*K0I)6jKq? zBy^~}eC3BF6z|hD>+EX=))WmMZu z-NFjs*BL?l#BxTB5KV+RI=h$c-UiZJjIr+VpuF5R$E9Zu8*R8bF}zuM>eHemhT}xd z1Vs;8c+M2{Z0~MmzfQ+4Z(0n{)lK0%>`CpW{m+?_H@AFZU<(wtvNug;16HaT<3xgS z#7e#|dj{(a&q!mPwY+E-|I;T`J~Fw(#+bZM(sduyH)$d2y(OP9eGl3es{IlXWr|jc ze`jmnoDJXn!^H(dVI)*+OvJWz+ATYT7-YQY&mPVvZyQ%#9T9eE z{(92I6xgEVN}9e83@)F!DJmvhp^_uLH0pKsY_`m=g}13pLVM<~N34pvcTU9ZhFL8h z^fIO>)B2Y|OdJCO2!*V>qERuXRf<-_LmOZ1acdzx<)|P8M6A=THXh73KVCgEowPVp z%v#|0c)4RHbfIzCbr=5fexxF!HYq#~W_nODc}O{He2GFh+)&cVTPG=X?5T#%C~Chf zp39HLu|5~S4_l&QCr_Mu37DuneHcjwy%IZU4zLikdj4i*P-FP_rRg&Y;SnDE@` z$Uz~2u4HN$UJSAwLJyHAuz*xa6+aaKh|JeBWlgLTSoQ$PnrTsGV~M0v<^tIkl= zzVVj)efD6ub{3Q6g#y74!=2zQ`#3VKz#5kGIhs}yabe6MY04aA@gW!@pC$0NoKJet#SxkD;_b#D zndGGDtRcELof+B{HH=g3dd_3WX|0z^4&o`?C+w%$1m(AVh|$*=zd9;;YTq7&V|tCv z10J}Q4FsdiM20K^MY|WR*bAObmZ}0`xacDUJ{5KSpDu9yo;w>iD0ZpHZ+jUa{I zVnzxBC# zbaA}J+>tuG*ZBQQL}a9Z7abgJB%fLmn;8fxRZu@3?uw)w=o~yD3NeWhc?^z@l6EGh z;776l(Auxws~-~F`Lg!jr`L2vztp5MH8Q`^A~Rln{pTaycWtYf@s7obhbC<4KDEY2 zc+JUihHD$!zSEtwaX78%JNvZlcP24_>@L6$WZc!LD6NA%Y{6fKl1t0-i z*w?yO@=WBg{}<-LQ{0DE9DfDRQkZr8C!>WlVK>nOB_>iXUtwdzxAJn+eg>~9&$D|Z z9{zQY_OhpvS?(`=bf*)n87|v9WcxT9{_nZVakPc^pnKqITklYyt)WSN4q&^-F~*or|RBwwU;~y z(dnFAYLKCN5!IlZE@Q9NwGa~08owWeqARgOi%jTN zu-0&)HlGgBQmPM=F=;0cd%GwM62Fe)9pSB|iQ7CEcVCjIFy1yVGYuJf}^r4pVjq`KR{Xm$c{ z1+(4~=~RAm?v{>KBo>|Rar&fGvPr}1aMQvsJyq<(Cdf1I^`hH4OIhz$tKfuyK#@|#6q&7$-nCg{oyk)QZ1}NNNhvDF*xO1e0{_70KZT+P6u9{5 zkGmJ;mir^e82gokoDIE?rGJRoRiex9OgF6=QVKt z*A%=}d3;f_z%{uy`K8P*v(0#=bt#U8*_nG6(*0%Y0M4Gi2DgDO?WeeUM;ySpNQeO| z^nIFrmdMZ&Qy`Mr7yhoiXE9x&%HU*(>8jTDRf=t^pbmQ9vQhnY$eEQXz-f7-5DI6s zB}r0WtI4}iAGgV|jb}B{<^1%rG`?J5E8jL0nNDZvyukA3B^yz_6FQ%QIxI`lZq$+d zL6T8j934+L45)oYK!7BVNjyLc&`4NU#9@g{2%OXSEVH+z6fWS!T(D%#XkB?tp55-h zc=t0s??Hfnx6mY8qs9oA3?tE5*8qwJ9x+j(nNA+DPz|_yn=>yehA$& z-smEwYiuba>OLy>f7E2(Q=&hXiuh=Wa8uw-?0FT7p-#MgYQ5_+TDS5p{(pdIl~sNN z5BBeF**d?g@0j(bnQP5#JcTbt^~>s4eecoUs&U`>nL;@KBXAXsen7{r|?>ohFRXjvAwHMY5Y0yi8^@?Fr~m^Q(|yY?pktN&4Nbkhz2Y_TXcr%WS?a*3{~TkO_0viAQ49 z(;b-k$MbIuz9(ZBSAkax%=Lp+vF^AX1=To9S<^6GD$D<+zjX*^esKmfSPzuprEA2n zjc0<0`SOZaxyN7UY42m9oxEvGg2X`yBUOIg)Jk50LoVG|0hMPV)WptoN84OH<+q8h zaFO+O(Q0A`9hT{t+u~um2WPbjv_`w192{1`rS3LRFFH8`@c|F5BL6CR9jJ_{ z((rWx5@qc64j_61N$=-DsVu_(L|8o20w1%4_I?*?4fx?VQA&HeH3-0N?#_grj4VzU zxRO-MHPoQv@?#Fc3U?aUN3v@1jhcN&HJ#U|z^^{1U(3;51yX!qGD>$>wLAXF*9~v{ zm~+=<+GlGxcO~ne)XeClYhRJs$z3R4|02^k5e6ha>3hyfk9C+@;`$X2R?qA759A^> zx~hz3m&<(r|u!e`pC(+R>! z`T+07%7w_#q^nAeNE23HOv85d=2y{t`XkS<9D@4Cg<%F-i#GkTHz3qF0o3V zS-Rw}STiSd3gK}f-Pi}TO&~UXp~_Un<@d?RwmZPm@J_e8H}TKTjlk~Ox1A{ei}xjB z3;b;-RzP`94ecuQ_g4jG1D-6;ESiYiT|?wIyogmYZz}ZP&`q zg%?sXHrIT zOb4Ix=ywo!d05P@c9$V=|LPha2C3Aa=MXEX@kG{S2O4WMZN%KrguFcP6&ktKNbj_a$#lLd*n@S%% z4uQCD!AjV1wnuhSOoyMz_l9Vr0TTl=!hvuow9uXAKw*{++b z;w@Q)*-QZt;o=&18&i;t#+?sohzt}<+6cLsV5M)wQy2pG&nt5P_mXB~EFBNdIV0Sv z0jC^%{PT9V1iG8q>s7lL0I4jtkHzUXq_aTi(R(2TRh^S@N_uyUD9d5_?q*_ro^T}s6tkXNKt=wAY4=e|WPg8l}C4@`WsZHm}*S@1XR_Hawg z%UE$0dgW-LV#B8`o@01R=OnaWv-c*x&;k4Rjm~coy1Cuso2%4I#YLF%$(%0QGtU6Y z%i1q>P~)GbSI13r`2H%fI^w~7!Cc_z0P5HA5GSASD>-4@s1h5`!E07P|FjoaE#z%K zsa*L*Ik*yP!~7jVy1;p(u;^#>()Idlz~=_|d_*GX(~Dvsjr>J5_=+ z+LQMr%P}pM;Ptv4%(6zb0?UcrA;SUy zx<5kK7(s2CM?m|PIARv?NF!x!z81#Jbuql(Bx7BFX;CWwj@xyw8Q7u@wdrybvXs7) zz8u$JwSL#;`(4%?PrGA>oiUSJ?mqqsk=@>iXC`n=+_Xf)=lK~aK z*hn_yw6&;A83V6?o9yCL9naZF%l{&}y90B7~-9th&K zLO8VH#;V36e4?QxQ^4^S_E8X_|3KKesvp|;db>OY3tbCmrTgm!e`d8N#}RiexUEy)Dqs_<;T&Mdu#K^!NYq$z@3HYMAR*jFf9_m}~C23%TW%Yh>nrDVNH}tho%aW^S8X zZk20ADA!y^$SurmuDKPue1H4>>2GK6b9p_Vj|avUpG*j5Vyq<>_oI=?jPBES5qrkH z88F>m#2?0`co7GZkDQYwqQP$2Rg)3$#{<_M^W{CVq$ zB0QsElT{8zyQU;-N?mJJ9^6R94!rx8u-nSWZkeb zL{ZzH?^}Kv(za`eXnrd75-cRNB(j?tUgGva;2MK^pyHsZHl@n>+(5<)^sj-}m>CIs zoK{4ELn`N8W<{p4E1r0*0*z5ivA=Znh(EHIi~COQOZo!akpF^-c4nC**z`QfU$Lp= za3(vMWJ%57aj_n!-_}uPLmAl)qjq%Y7@iig4fk9a<$5O8R}3cf--eQf?NdOg-+6+F z!urpM%7>w%YYyN?YRC4ekJQAZ7nUsUWPp?JLCS%U3dyO_?;L6#u}h833WuMbbgB_w zJ{dPulP0KX#x6}vsiq2fiT28&<{XA1Bg$r$jme^}mTEs)2B3O1qgaQeKh{48^r>>I zCd>IRzdFRWrk+=guf~?Dhpt9wen=0#z#RH_q3yXgWCJzCEZh@w=0Kzc4Zn;~-@CBs zT*{|$@xhM^;q1oizvmO{6dPJ&H7J>&mGHV-ZF4*%E- zw&iIt|CLu>A(FVR(ifspTW1qWx0QL1N1SMHees}qmxfP|^?Mo0$CtwEVkkmzM3SLN zxqR>v8f!gpyY?yi>72zf?4tSiFkU3RHjkOOmfXVS#ll%@C&rWv&R~TG41?-BehR+N zPjO+4MLKvn&N%Jgo&Uz?g)>J=1BnFRA9&I;5;o$Prl!GC=K<}?p=FM8ibk)Xjs#!k zzhq{#PK%#?&!!R47OSpg?}yBi;bmZahM|9Jumx7jy{5s`FGC#HWmzsrLUhYWl2TOrBlO;a zxl6J$UJ+La<{elD&m}+@%3h9?HC?lba2c$uod3^5z9%q)pbyFQT4-o=eydvD_6kah zUTgbs`7--FL;Rb!7vzw2{nj-ex35q?Z}-wl)zMI+gW6Y}>dXOW4biuIxGljB#TMYf zAK>bk>z}kaUpjUl4?b3sxGvtfq25E`eo^*ClPp)YZ+|t&W_V0m=e=OKrbFf*n|j2Vv!I50{LPbNTC!u1u1r)KY6*)2`9##k*c}vv0sgCd3;C;}*VR>~A(^ z&~VwW#^FP%+%2l;Udg}y35$b>S3gDCzCI~l&S9+i3%&xlHhK)<8q#{J9ehI2#mSdG z)8ucv5JSpbmJc$#?49Y1IW}Y~*@8=yG5$BLLo0U* zsN4`_&I)L9ka{cn`JZIkEoym7W`g^Z#CvLi@8)}&D@@QT{3;!_<;&Wsv-V$|CJ{V% z-pTa&e#>R&P^E+~qGdS}5b9q?w9eA7IWT}2RE&)y|N75)rBC8YSs0Cr=NifX zeen7ti-hq|#4Kb5E@4r6S>=pR-h-#~Ar|*_RqhU@aj=d1byK(4qp~C9g>zmwB&*a` zqH(n9tEqCq^VTLE-B&D>5)Cb(B1k_#N4FGw#}p~6%5^&wPLZjgbb6L#7rB1YR#fHk zQsR^FCgu#PuMwaz;1AGLR1SL}j`SGlZIfh9Aw!H%^HXApgYH47fo!e1Y# z*_!}Yl_(t~u*xNrhR}8Hi@p5|BYPia$S0#WN`*i(ke8w%x{`fY?E%tGRo zT}}6EyT`&-Y-}c~OIM;Gf<%GweX?7Ti8U5|Oy)N38AW!3gujFfK2d(&i=WBoc%M?i z$oRfPp>rAQ@KZyhfQM2Y)+3cghoiJ90w{e+SXKDjT(nKJ zCc)6UP**2JSriGY9+>W+8@Jnie&Eg2+wLEw|HXF_FG=*OQEeJeU^*K=xFmC}A|q%X zpCzw#w(n&>V+eZ@xprC7JJw$C@B_|7`)T-S(aP?~CB?xz@PpovmX|{hh<4C@+{BaI zO?VM5y3Ph>;CO0^N54=JNHmP5>;hdFi&z8W6R&Hc7!fFzWN zM8a(}X@WH`C6lo&!$(XNIVkeIs;tR~tw~Z|1c1ni#gTHZ>JxvKEPEgHV3f=0R*0^} zFxBPrt)28ebIDzh_ESZZth&AgT+52c17O%JPS%}g;_H7q13fi8Dnx2%Q%`2vgNCME zsNP4#UDn{vOZh=5y7c6`F@eFJ7AlM?)`+4qwvM?BsJ6O%A3}e#3G21Vc8d6>tM%c| z>M5HB_tEn!=L+>Fv-vF+HP9}h_}ukMQ%uq*lYW2yv4RC>QSf5&H~~BGtrP6kTF(E} zVyxr1^1b1oQl7obJ={RAyD^l+m!cYlrc>C2m|);!-p&gs(>7AxDct397}`^DthV(o z{(d=#%$jboXkQHR>CB3!->ENFWIC>JogLInh_BlGnq%Jg$pfNqod3e!z-KPuc+Vk- zFH6hZ;?bi9Wf~Z5D~GdQ^}VH-i9H>iVrf2CLO^~hL%a(3lr)d*Ml|hcGM(551c?ch z3q#Zl=aR>p0&dMV#5F>CxGhwSc{5o3R?#HYiF3wp+DnC!A2Zol$d_`TMrskfZ>B@P zoyi31K^sOoKeBFOx+vM^cR51ZBxyHDrJ;oQ{|ee zqVc(Ryom}eAg`I0eU-`H=A5z6G+zz)-$QQVQRU8qnFlEAx%BTDCVv70Z>M>CG#y${ zEF4{ip9$TmULH()X(^*-TLfvXO}9!5yG6sh{Jt`|PIxB-UcR5_AtXElVNCX}>Fe zF3M+zu9tHL(G^l3m0X$v<_l)KbVm9KYoD!c9gr>`h2@mrPH)+Hhk-^U)L77+G-^>( z4iu{#8ZT#?@r@aB>X0t`ZTdH-iiPHUw?|^%S`H_^1Mr(FSFn!i@!xFu1grF!Y{7xDJ!PhNrrZw^HLQe9)Hnbk#tL)b?UTMXLkLpNEuK@=1I za(fOR*0g0tJv2Z+OA*M>gn{`w_w*v9dKk{_rxpH#;5W5MKy{?OULp z&%fQ{PyozLYTWSHxBulG_s=V5;TV=DtfA8lz`@HKg0Z^bHG+Q36q!tJ(DLuHj;n5? zY&MIy9!2sh+0+d4IGu@hFhr2fZ*OH(4Z}M%38!Q*8cU#3!(-dz=vSIEDs73JFD^Ew z5=+rrEB7I_XtOM>oQ@NTo*$b(UkO$xp*n$WkrkaU;5GhsmJ{nC{j2xB>Q6@RIl!ko!3-N}+kDb7!rmtNOu&p9l**Rgnb&)|=2 zn_2pWUUueRP*_FqV5(obe&=#xhJSpozOi|D4bGvIw>B_!zc)GPcouQlSYU+t8HBJ4 zDASD1Nd)r-I)#Iw@)25k6A!${>n=CvL?|l()%02LZe#lnT}&na#+xkjADu8aM$O9h zgZkNWuMizRa+lvhprx+>eyES5n*aF!4vnX?k6dtxXe}#Y{@z-S!lE#Gg4`c-IHvX< z3Hl|@g=Wk#jjs4SJL7J6vI~KL{`x_1N9RXf%Y{MRVr)FBlR+OM$h{IySC#Ib%goeU zhxf>8a;p@^+(h-D4W3<#>Uq?-CBYWX`(i30y=Un|d|?0U9`lIrT*2neY0?^f3uLc- zq36ysYoQKWd2adD+Jb(s14T5Oc8i+E(`$l?2abvveLb+aeEXA`w+P~cxRh7)1NzjV z^8i^{$wi{hsfe2z_)Y8TiSrlv%2`84?^AM&!}A>Z=kv(G+jrC+MQsmbf)`zUfrqup zpWYSXZ)WwbYVriETL7K1gyZs)%c7)Pp54l_tf+-9<&IeTW2YNg##ZnR6AI8nxRX|dlr_)S zn0A|w0ej74Ux#FN-EHsH#z8SG)Da)}auxDMN(2N2B4s9+DBZ0TaO`NU# zA)*O9f_cMC6oniJLX6Hd%NTz~|1~K2mhg+LVZ26UK=Q6+*0a;36Te}qi@bS;iOy&) zF54(hrZOcck14aAEHoM4v1P%+l)&_>jPFo9q-96f3dpiJUf-Ih!bcdK*u{o~gsR(P zaeUmn$!($U8m9_ici`4^S9n0&f#!qtOWS_*)tYY0B3j$YhGVZVh9UC_jd5tn za*}vC@sjw`yZ#on55r~xuPbU7J&FCTVq%tZTLX)Bvs^4-Y7DYL_ANMNwj|y&axR&f zsd_q798H^fN87#CRr>nHva7v6F^T9q7kS{T;3|A3pvbRQH@L(lg&MYvpg$VV6el^q_?z&DFxi!-^in;t? zUAQ~!?8T=qugt6^xq8Kv+%kXcbxC*i2bW<}LB55Q$jFO)E(Wg|3eI*%UC{8u;sb+j z@+9>%6b3%-*7<}}>8*)BzxHFuw%0uqofj!6H&%62aXISO+hR6%sv{%fi2^hhd zWK?(BKT=>Ws8cGK5s>?eX!+zRpy}qzyMJ)7Y6=k_0cWUp%)G(QRny zu|*{aNc6jKw;y3{B;7>OP1!iUj|fs{4Rwlm`K|ab*uMX8fKO6J(hujZ2IzX7OlgF} zBV{xl%sDmpZ2JdM3q1-3o0q#k4X!JkPDm)Jp5qw0uG=7T>?I`R9rxT6U$J+EMZ8y) zagDH1MAFGCM}ogpxDH9rum$se+PcP3JGwW;w*C5%8=XC|qdlTOwm!Fh`d;H35vQ~h54y2E*|%xud;=KsoH9ZT+X-mx?Qo3 zcLKlnQ!~6&r-rPSm%slXn96?l9do*eLrt|S-~my5y1L5Fc(XVqn=o33@*L=X0k!!A zk(G03d}!py3HSz?z&s#$0=W#?Yb|RO^0yxQSVn=;A>Gp3EQJ9Zs{P_tj8t5isGtWC zZLbhCh?vcEoTw4`Z>MWXZY;N&gi6q`LeR$r5#QyK@#zCudV;*%+<`fzGih(?xqQn5 z{Dt9e@NyN@F__8tKc)*wij1IE-zJPzKq5!Z&=f0|Il+Kh=N)cQDXD!P{VrjGm*>Dl ztJ9!!=_+6^iVyLej_@AdY3kaUP>S^?dslSV@{YgqwZc5AO!ms=tuH(Sfo9`%*T* z;E6s-o&;_cu1&MCj~cNcPC^xdFNNVl;T}%bt*bs0E1P6fg-^Qg;Z;Tu&Bz*hO|u)9JNze+VpusyEKW8i+-4W%iWwqSzq`LwhNCdM<(`%o1| zS0KjIDaA^Ik+}7FkrhQZwaR=CwgrtU!&S-F$++x1=|m-^S=FN0L$gkJDL6!y360Sm zam)6O6l9}eee#k&0@Ix$sN4*kt8Qm+02<5xaVB0sAKaTeZnG_#Z6YSBDA)xPS_fyy+uk1a2++$OZ zD|JPc8Gpi%a(XQ+kz{712BSlEybd~}vR>35{WGuB4Pp%q1)4tUx&&TM_y}W2hh1!+ z4!)X4`T~8>oAPXaB3EW~yCdMch^ok@|F$w-y%&4c^h~Pz8%bPS!PdgsNTK0qIYS^3 z*Hi7HH`e!wqNDEcN*E?zcwIg83Bl)CJ_AkJ%yHG`O-5));KucHQIiTjR5&?1db0&l z_mWLllfamu#MVgfUr91_5!XTbWygH(vR)fKJ9dNv5j28rXYHjbBBt^p>_yJL0vRb;u=btkd3%0nnu4{E=yt!dq?yrWThNA!Li9tC$ zs3mpn%I7(AV|koXdp6AP>2Dnc4Uol^CJVk#w%%in$`V7Hq$WSwcUZG_bp%NaG?m)s zgQ*tRtwW<_1fPr@Q{;P$wwyV)2PUfpY{q>Or=+rvgtr4%zw^>9OE=%n_O*d_w7wp{ zJeROY({xCUa~sW-|8K9vf@h(3ku^+@nw-q8q0)CA98rcJGS%sz)hOAm7mhj<28ua6 zaNn9N7|)!{`OSSnQuM85NZZq^#vt;rkqqVh)(o+O6(~bD_G0G>F^repjB#gNNQ#A* z$vuyL0cR47eH7JKDzw?;Ssv~TG(c~HiRFs!t6}q zN9?nfm*;60_}UDypOwugCh}-vSL{BoFcFdlJtw8UKa457(oFt45jtzD;V`~Dag}k$ zx(R`Srv^!KV`QpYWerBGxZUMU4IpJzTY>&3xF*40!=-+^hV!RHNUYcWWmjs&+?|lf4IN6mt+I@g+n3 zhi;w8*k1R1$+{R@;IlS`;JAN5&zko<{eAu664rW6`^k($p0SR06b)mgBN-bKADHgd z$tZd-C&du{lLF$2zbbWy4YfRmq~fqN+!LDzLOqGSy1fCx%fQ>{alz`(;2@27m`*>i z%;2)eS+c-rI!e!|^Yk5q0m-%)q6?f-#YS+ei}*cyk}6WQF8t7meZ=F`?*6kmEas*P zQ>UiYOKyAb>tJqJ%BpQc7Hol*OgA}Xno2iaGCICoBAqZq6l+-{0?up%hLs;>)4I52 zJtzF(2nEylk8=Ky;yt?AduRxnn#S>(9qL=G+Con}CER}A#_>x~BZoEh&xW+k>cq&z z?uY*zLzLw+Grzo*a2lVy_Mtywo#NdTA5|B~VU~MITWbVW+Zb$jq@!*(dW$+x^^v0Y zkrKYZHpX37T{IKdC`mZ-*E^Uv{XJ_t^-ugH2B~3U;;p;(QFC%CO)csuNW_2oX%{Wo z>1Xbi|AW`(0J@m{!nByPFUtNuS~PbV@Vz?@xlM-JqMEZ-?Sa2fYrQ?=dSql8ZAGK$ zeP#Fz6ctbOK6Je7b>FWD5)5nqGu|e<5@Xv^Fum}>CP^`h%o04rd^(0IyRL7X!B66A&mGskaQ)G^kSTwfk*hn3`#K`GgMbivto&cjtA(lUPaK!6_)_GX z*8T_ZA%K#*&3LRs_x7WlA9>NZ?*P)Tsh3qFb z@e%Zp=c~#6-SK=yej$E8O#}n`m^B?3&)f2E5$*qd^zafZXh{VCB7`2+wK1=Pn6=j4 zhR558sefI~)hRZ#gjZx&?+#wG*}vEtRO(M1(V#IW-cygN`#gS z!!eQVpuUZ)+a?|BjSztDaTzOu-p!K0?X2pUs$rigMXvpYY`*>;_8pTC;!25I#%3`*H5XwT z%oUonwJi)boPI*czjxtR`jlWl6FM`C5$ZkTS53r`vZp-rCHKu{%P#1LZM+7z%azn_ z8{F;IeG9QE=+{=P&I(1_%o&U8>BST6zq(3Kg$O^84c_8(~=^ zOO#?9A2G_%hX+z*zV3x2t7=#euS`8vp%yb?)C4TwFL{vO&J`6*y{6 zu+ogAA)=0kBorOjWT!?!4ul4rO+BC&yYQp-J2`7sOfrpV&&q*cMwDDnSmPH}|7(j; zp)Kn*{6||Swzm9DOgsJMLY?BJ?ed8mBSo0BuN*uJoyd2D`gK1*{L+N2z2eTHd%Cqb zx-AdOs4KjiyH@96_1g3EpN|8X(+!ve=I?k57U&~JRCYM9m}~oUWQ{ZtyE%cir%VBZ zjf6F=$2AYX3-*=a7P+|ih8!gye4AUr(`~*M*NwQZF$xST<%co|skQsS1FRwR-Q1Ui)0-NVK2EJoXn42J-LH2R~Se z=XV^J%kd@7));)$?oRNVGvs(Y(0X+wfODg=Lcg6Kl_gKeG)1tz`?VSp4XS0B6!-=# z;cD|wDuZ~K;h)O)$TrsaZe#cO_w|n&!Ex{rUFT6bv!LYOCiN*-C!7fB+9mF)rk9GQ zid*J=dk)uw{2$VG}7O<4xSXx-mC^}Ei_xL1oy7HS-_eVEb9kKe;2U!0SjN^p4i@4hlUm7Leb+G zy&GWc;?G*IDbw4f72sP2_opik4{?9D-2Pn8I1gj_ssLw(b@UBI*yL2c?d~qWYO0#_ z%!M!K7uh+?p6im3A_NLReHVOE?`A19G&n|oXZ_}#f+*-`RqF*F{mh}OdHu{3L#6*MTp* zHjfw|QKs<;U&3k_{d4HSV>x>1gU>uJM|GkCmbp$6zPX<9HG=Q(ye*4`6o}XJZ$l~( z+MH1{-#lZc6w)0CeTsKg(u>QEAV?n7zrE%6D2mC6IO821ym(E^dM1XYKv(m(ovVdO zZ{row$iA5I`q@_Zx7XlXTP7dV^SmWY4X;KliX;8KAP{6#1iy$CfJw|;>OZf&M?&(n z-*cp|qFjJ)nh{mz={XI)^V3%cjNJlBgiBAGNG0#t9$bguXWs5rQ|aefXS&`io}`FD zqweOq?p)Hmi74R!j~|?ats~0K1@yX|k19e1X+?WOF|PwmHyCT*3XgqW1H9WZwr9W0 zk*`P0(3l!s;NE{CvO}muttIx!P~ti`I@g;1AeKd+pSLnuMmNJDGbJW9*4(htEk%8% z;dr`p(T+>+(>U{owT{sLHXXSQ11TebM8ZOWfhyW0^z|n^-MHh1AW3$FArT6ihSRW< zU-Yh{*GJMOUp=gaT;FG4-=phRkLuh{di|&-7y4{;hAp&mXggRU&ZwyB^DFWYeO~H3 zKO+~;&Bha>J5r;8bzf^WXBh$YG5`}cUeM9cAwX+`o)r1eM{e=-P7}QjU$OYgyb%PV zoDJR+qFrDntJ^DyGI9gZjU?TR!7;f%-bnWG*8O;6kjCsRAx|o1{^VjOEC~q4O1FX`nYcCb!@Obea z*s{chtb(|)9V7m=fG9SmCN-E&ex@=(aFbUt`?}L8Z@iXd5b}u#UT(-~_7w{(D=>hf zD;l;tQ-I=)J%9AaaOF#{vavq5T)!NHr~JQX-r3!mTmwG;n8x@~!v+Q|4dY^=LEzyV z&0JBvrnUk`xaj?F-+s}*>R}$dH*qR%{$ot0AY5&;Znbi3$Kfb{DuTI zV72GqmE)gcs#Di4mu7CuZ%P$8UjOr^ca$?1Dc&D)>wM<8G|u?^O!??(LX@R_j^Qlq zqfmJ|+t#hcZ*ZYwPWoi!FAukP15HO~u3d}@!|hBiJ$J`G?t>;z&3c-Cx{U$fiG-6! z`kU8UpB*M7JR5~)IhnUGwn0GQRBohw?dZ045M)hh(LAvHn@hUW?2nP8aI>I*nY*Pk zsSN1k;Rv-o`Qj)oFM}9kL*~0KhhhI&k?nCg5ZE|sjDH{pEPIqAGp-ueQEp?}H@tyO zk14OZYxKDM3v-cFcJisn2VJ^Jb6um0_jpTsj;2SkR!Z&2L#pBXlvM1czEyqo8s&rKm_rf1+C&@= zMcaJvjtiC+9}WU|{4%=dgkla->v|@}RL-VhG&D6~x4__0u{N($u@{~i$y8`%gW0ww zUxYqcL>D^6NY~y3tsI_En@q1o{VdJUfh1Fj-`p%!bBj^cUmDJ}b~B^;ZLy;vCLMy< zB3k4UM&z}ry)q{{W4@9bMWpiU3QOc@l*9)n{3{&|Sd8C|roa)s(g~t?Y`Up_uLt2K ztEhsm&Y6U*A-lT~+f!khn14k5uQClTcO&K^@K*)Bt^Q`XCP~TRUWo1SMj&Cb_PnBQ zuN!Na$I=LtF|GJqjb!sSm&&pknk)sA)W_c+sF?MwaQ>36P9Lm7SXfwAoRs)OZ|VNR z$!|zJYZo@xFlW%np{QQ*uPY0{G3Pr|9yTvCQGd4R@@C!*4aNe`OF^|8wZ`{aha3!j zPF~uZ%NUUs>#f*y|7@-Pb^B*-WSzTfc-0%aLfj|Gk3Uj4Nu5~=YwUX$-9)`l=n z&aI+rnN}EM|MiwqqL(z*+nR`)N&2km1rXtJn~3NWbU*XVP~4g&DDp5gdDkuGs3Dem z7XTM>g7YcJ10|!F+x*i*m(Ft}#n`@pB=!oXs4k6X4p}{0T+aAmRHNyQ<+OK02qt}E zU&~-d#sDf6|&J?@$wm;UzRb;!9h$FB3n&wtk) zhL_`F7i|rNQ&o?FEMDm`!i;hl^dpN19xg1?Z0a+-V%rP`*zsx*ZkubQ0DL8KRPcp# zOXzwnPZ_LgB@0N?6A1Vzu(ll&LnW@`2`X2JP6#|shPqP;{yH-qi?KE$_G36<$urx?pbZs^)08rvfQS1`)C_)K zPQ~6rQ;CQ@ML^P0q+GcuJUEszD@qhXq~5I zXtLP^$~d~n1LtoHMHy?6Ij!j!0DCXZ$i>PI1L*Gt+-+`FR+R4!nsHX=?0p@<#iZ%; zPL^Hi_-v{WgTH-tmBbgY8k@$NX(7m20?GY~Dc5%fUDeqRFlaRDmkg`Df2TVjrm^0- z;&iE$pX%D**CTiSy{?jsSk=KNk@8>uw_G_C(ql$u?#>Ht9O`HJ!cf6pk5U9PZ;$c9 zL2A@F=U%BEpxVgh#Uat%USL8XmEvmdWHjO#V)JQ)#G@bFg_h-H;EF$9(Fe35!~$QA z`zke2r`qj$Gl zQTU0qR+{MBsv0BKK}Y}4QWkfdy!x&rZ`Ry#_1B5j3(eF5|k%JQK| zIP6=nQ+?~5_;O^RUtdaQ?ocR)hT#DEJKmKQ+jSp=v<&!Qekr#2eqQ(Geu&MtNkp;` z%YHA#aUxj%7AApat-bG-?+ zjY>{7mq86HS>Y;!yfmy>u7v0wovWC=r}|imw$M~-c-(_*tv>%7E%+o6vr{zjLSFM~ zOr>euzR&h0w*c&6;tU1CX?au|Iov-+~uOT-K`%9t0xuet>Gj4+~riFZxJ z=riFxQvh5J+BR9fg|o-Vk4VF8l?P;RM@_%DswMNXc(P}pf0U<-RHbqBt4bt(I!BO? zbX-JzsELhnD73kDt#{mGv#McY6qlA1CH;1<=0jh<%h6E{^mb*(FsyE1+nts7Wm?O= z;2oz3Uoq%?aFYBz^Q!}Fw4))T3a1=s1+7kq^~oA6rM#ui(B^0MNNy?5E6*T>d|&I- zwpBb{f+;3W=Sg59#Pb5FcWDBZcuUDhDO>NSrm2qd); z`=x~1#lh6pH6KTuXaDL*5@Fhg zZHe+)_?JQ&%>+x7?bZc3KSRaQ)a3Naas+Gt1`<|^L|M?dep~N#iy5_Dm-K3roO2zT z+n>qQEtLsgnD^tdv^6Z15LRm14Wv+aCmnzANTOXj829q!hNH?Sp-)F-YC@MihM2!H zbZG^4@5Gdsp2hN6XLqbqkdcCme_2&BW=;C*q#!%sTBPCJ;?a>R)(DjlORZ8S`c~{2 zN5a4;UG9xwR5DJ*vS*$3{{^TB_#!tN6Pl}=*s-T`In*Ri=u`(p!&_sSo}d0KGae#E zpCkTfiP$nLl9xKY>Imyr4VIY_Lz^`Rv(pQm&WBp4+sMa1D||^((RcqvTcFt6blf)* z>sCxcWBGZ6%R`~HJ^}Wg#$RM5o}%kSAvxh4BQe|giWWLU_c!ge{1!gOw#&c2@Zahf zN!;s&T;|VAC*lccSlu=qws%bNmezaT10ymgSOiYI^E*v7dSMvW&cCmD#$8ekeR%zM zIVE1IXe%S&r_u;GVTCJkda7T?ebrL!+lfq%gNV?95tQ6n(9NSvrD|-l`k8~v& zJwn|j*w;O*a=Q9sfXqi2cR2`rBP5tP;)CN^cFPKv;OX;{P`<$Ed{ALrj~?*Nw151n zU@)IOrU;ZD7djZhgypLi1T24jA?ip|o1pE$&bW#5+a8AtIXd(GMIjckw%Ojlr#ry?EP(Y2hvP%hXGTu!!XUvG#5(0l2z&AQkkcwVg#>iBA$e!h za>>#r3c!#Q2Q21Ko;dgXYM0264E?dbi=Sy*?|O8_JZ5(APPqJpTHLRt$J{P~&HGRF zSA#!+9mXDkM>YbtI*##mdt7tve}Fu$Lszp~=!RY4 z;ZPVm%VX1f0gAD z;fXOs^uT)?VG2Dam9OpXl{$WP{GtpiWf~e;Of23EuVp?<4Y;y^P{j&zEgqVN@Hpva znFFS;3I2Gj4iaE!J|rU{NdKB&Xo6=>-?egMPiV6#?b>}67o{sr$EjVb=Q-EWhXPw} z?VMSdHQz3jxY9)BHh$W^Pjw@+prfB*m5Hn8WN>%uD!C@()FaHdvc z);xffd8HLM)re3FatdsQrg&k`eBN|+ht1R_AIo3zl=RUFVFyys2YUMjZwv?GyFC|| zA2dm6hRe*nv=O{7#-`QKBTOICnLWqEhyWb;(Fsc963>;G7Jv9LD19i>1tzr4r^sRo zkhrco$7oz_>A2#=5_|laq@Dv2yM91ifgkVcvFgQjD3!TK;ykHz0=FA zdf3Op4SSxTtU@45bw-cW2{n?VvVNIlL@PIqMhjOfhNydwMyn?J#lCo^1?kl4+r7_M zs251R+L#tN!JV}ZspyW&1P@&r#15dRm*2(n^h}*O%jFB>kstzjMv$TKnvmQ4lJRZm z8soovSa%`n-^{=AH)IY7L zg~cZWV-+eZnQUnjd}{#gj_LmT=caFk-WF8%4N^W9Ol8m(j>kgYYL1(opaX+m{Vm3` ztE0MOG!?H}RY}eWx_cHjaHl1jk(4du=5{gfW1nlDbNcXmStgA}Rh`1XfW%WypwrC$ zz0}3AtCzEkPCJ$i;=b1fI*3dbUmhGZ&5FQIX_ki;U;L)GxwLyf`o8CIZ)F7s4a3d; zFZ+YYSK>{sN#XYg`#nEf zZ;LU84*J|5ehQA}Z4DdgTigPWxOM83EzSz>fhxKUq;&3skB%svlEg<1Yv7$Bmbef9 zK-B~G8>+mex$A1uyip3LcDEC*S(u-1@l8U)8d39~!=6-*tYQM{nPoC9mEHHpB?=5J zjA_hm?4q$*N&}?cnY>a+;V=nw#IV=_`u6sW-Aee2&LMu5%brF_6d3jRcVy`FT9oDO zd4%ZFCv(!|_OX#gEk|b}53ertVhjR<0BrJ8bDXxh$Pj-W$#5tRQ1;QvMEIyoL#sHc zlqJ1_Vg=%U)G$8{br;Dv$H!CW1R`eXb0(1s-@%^5(fnvaN@BW;BKpJxcsEC;g7Mc! zdd9TZrBE)ecysNwV+Ksvm{xh_nU-c<#2>07UD&62cvP0mIE51Y;kF|H9}?}I>8gUk z{{Rt&(Y2`5pkCzWztZrwJ|HvoZHwkw+ij$lW1$4WQ^+Y-hi&ins8O6115>qu+^Q5e zW+t%$Nezt3vN(@mhac1?3mF0nL5w*0i=~^MTV8JjmW#<+mF@`;*qP}%j&7x@+=ghA z-X+`F9!kO=*vlGuol(iZm*Ad#L-H_U;uY4Y%9L&r|RX20T-aXw^B83&>IZ#HMx$n*w9GNpJy8;dh*zq?*_dGS}T)5Wj8(!5>&t)?C$|#)i)Z47} zweO4k4TL6-Q`{vR!^LNaT9KAp$3~4$hA&^Z68@n3zr}GoV!@ofi&MpHlZAyk=6%r3 zm(R(e@)$eHg^>uHR)H&tQqE~~)_HnI{)>kkKR1nzX($Sp??(JX9sDU5&ru4#N-vQ8 znr34pr4I5&w$>nT?D=zsJSwN$v~C#o-XnFrE%E|?(rTpp+XDx*;;ulvfB&*?6qu(YON=zkum@K;hwQ#pHg zbDeYUI2Rq&@OPs6)p88{{lYzbV*)mcs|+d19NJGkJAwI%9edPw&N=)&hVojIv)=2F zI@66(=UlEsr50`fyBe3SE;?pBma+#jW6Vj8%#o@Tvk<0Uw9A{w7*d<8tgf!q);McO zyub7MMpO@?EO=l?cxwWqQ!YaaFmLc^xGu&n$UqZ|Pi$kw*TwvvTX08}sU}pgJW6`r zW~@1Gy85m#Afr~1vEM!I8*n%jdj2eC3MnF7cC^-XEW3j02~g1Les=e-24;sMJN|S8 zC~Qgz*hB(@FE*{qWWj~9zO}qaJXmXdxa=dXM#lwcm|aLv;Xbw^VAf*uJ!OYO?Xt$s z3!1UWM6yh}t|YENMMfvqzU6?bJwr`ue(ZnzFrqD{#=5=CJVElCi&|GDdR0*7dz4&_ z@DtYfQK}M><4QE?nY2%X+58K3)^ z34U8c*OW;9N`!N?_vb9dCr-!*(~5na6}>LDNsAY^c1*~>&k+U8$k|eDhkc%&rhJ^KU!6}}#xd{MI^yzXoZt-W?eVhrXZ!w~1msxVT z;3?nWO|Vy8j6W8_GzB@J%sF6_>1M;*uT>R*@<`#dO8Qc2rqZ5c@Ox)a$FbT}ALdKn zRZ$|P5>p7a?!urX2SuL|$#)ee%6qTrsITEXm+pU~Fx(+y* zF~72-5DfkM3Dr-(r{j39H#Ydq=gypjPNDrX?cAkjg2^*&$xDUUpLs`Hx!}K4r;=U% z(P46Q6Y>>nBFp}qLuQ=gkAA}S9k{@n?cx5G+grinDM(9IOd}>hByQCBCeTL7Er&Bz zg*55g_fpH9e7Ha|wgBFXnceqEzL8pOF4}YJeyaYgP{L4RUv;4EO6KTGKfON;V+X@9 z|IyTASjCi#?{*(E4uO+<03&p>@8C83E{R9$Z{iAXWM4QmxanS_ZMwWWAyt0$7MgXMaWm=Y7NntQR>lYqv+?R|M<62338MMWlKdI z^xB}#xwbD$>q)u8^ABaxf7Of(s#^zUtqQeLqvSs+ZJa#_7`VBQ#~pvG==?v9&c&a} z_woN@GN&kqY|JTgEIHep4{v9NIVMRFhGfp>kVB}&oO5V~Y)&~AAt@^6kh2hyY|L5C zhvf6y_xBI%vB!NsuIs)Iuh;VlUbV!jJ_}0%*kc*A(JlsR#!9)nSJ0Qm<-e~LBjk!CS}GYkzAVyG1lBH0h@}Ucd{H=)`-

Q10>F3GApAFVCrK z7MI}E*P1f<{c_SGUE)(ehG(YjOX^}pb556rDKW8duiD65GNRC&q|P%wPP4Jkit@_y z8coHYI#>wSH{JT$$%uCR(JS_DAx3T>2O|Q3nJb~8vo9;eYgH~1@$RD^gcCbp9Nyu; zy~;*`5;lDr{uDa)h&)jI&BnUlbaP_)<(D|j!hZEh9Ctk~axJptCIos8CxO{}0Q(h5 z8x?$N?Lw11f{GX8f(cCr;>EsD>P6Ell$WmYbdlk@-?+25tRjP-fPLKP=TP4n*4@g9yIHC%Zqo zDtqCBF&AK!J>;c`>ao`jg9jo4$uK~y>c%Iw4?6n%yCv(pS#td$R@&PK;8!+Z-*N)# zE49?iC0`N^xkXotYs2lD?uDTzvTq9lqD#DW(-bL=>&Q?jhh1RQME=}R+h}coeM*}C zhF+cAo#Pl(RND*D8cgoCTk03^a(-9~xFVKn{*S9DbUUTR5fsh#O_M`_vOK=^u#fw{ zetjDDw14rlVCLS{e}2Ui9|RwSr2Qq{fBb_Nf0eF7X>Qx&@#Ho=TW#yBzdiL)K+VnK z!AulD_w;+`nvTH4^;;hX!bBfmk(098SED_j{w%&UDAY*sHH0Ab!8$o}m%DUx4R44M z8I@;JIlS@`yTvFaM7le7H`d?teumHo&q5oeD6e=z6x)4zuBQ+^TPtr$T%XBc(eT`} zDE8H014)cCz_hErfGJo2TvvJ!SP?yTybC$N+ecjEMQ&2;%)3o(Q|}k@Oq@m;?K~iqa_8iD;&MlgG-}U!4aZ=Oksx)yVX&efPb?w0dK7Q~M8H52TsxJQLgTF!GY@`f!K8 z1h&(ihHHN1#qAO&A)HfIubZR^*rk_q=uA7$Y1I7)GM^#xBX;aN!4befKxkLb?Onh8 zwy_haRrqK4nXp>qQjA*v-8@(r^Q)kD@fY1vR4Mg5A6u&5|Mx#YD^d^eC$diUjYg_> zv$oFt;B~85Gbz4j6R$=_<0pwYadEn9tLn-koVs7z&aT6kAHLjJYpX9bttGK;=KP|c z>6yOVpicrD)jOOd{$H6OjQ1aYVTa5|9daF_P5oZ^d#(5B_ZD}2)I?W{P)Elacj@Dl zLf40I5H`6kbWhVc_axva>XXuLhfEPvv4^X}yn=AyMp8&m0_zf|PSZlfVd<1;Mqq4C zCm6@Y8*rte&WXt`D3F&jlR-@+%r_-o3v${i&(Q9p>+a;$vNZmou$`%w$8vL@!8K95 z8F(O`qEZ5-G()b1WhiKl)&HcT4Ve!hsZiRViWCWVPW;#=u;CQB2^#=_5YnwRIK8DO zl@WYUqR_j5e8-4)_b%a$pS#=;v&zmmEu@Vq#Vw*wM$~9@N|VbzXeJ0C#bznGeiKyiWM-99_&$Uf|M zx`*d%`anxG9zpnQcPshJ^hB;(39XtQEauUhv-jE?SOx*7`(;}Dq_xOyMjo9UVjLW< zk~obvT>Mq&LaLMMjh8)3*)1fqM`1gc>J&8b<3(U= zXHJ}aEq|Z$O1|3M-_TdE-I7eaaAftLd1492ZbUa?_bPz*pbl5e#ppf>8OllQaDDsuvvgbIlYQs?co<0a9oSR$O^CoS&hoRO{7DM-s*$gmhD8;FJ zbll-bUB(it?R=%k-Aej zB%IH8Rfpw&009Rvh-jIk{W7s&vgda;n6hquD=!Xko}2q14Baw2`A+;nXm8i)RJtQH z^{>CA_^h2(faCj=!6zAWxL0kjbEj#r^$90;+gl3LM|F=Kny0^3FNpZgvLlDc%zpa) z(3KQK+m|G%bj<85&A_BN=#W}dhL7>9dYhr;?ja6--EES)C(cn_hIf?=4sSdO!gC(n z$Q?~<(8ioPCRYJw0uzOspAyR`u&&xF`n4Er=W>M0*?>oFf)xcmWMXS4Ch~0i)_eQE z*1C0?b z;eq3mIj^#h4=d^vdtuf7eYd@m8JcL?G9+W``mu5eB^)!Vf<3dP9%Gg#U7qYA^@;cQ zKjpp(4~hMA`=14(@%X4wbdQSLN0Vv2FeDC)IZp=J(Qe!QKV=_r(l^ zAE1`^oC#mv`Ci|#dQ?K9rY->FCR33n~^~C$+)B&<4rCBvM9=tgS z3k;r#Yi$D$N*a&Pq+={?oLtd^awu)CC)xGLjFk>ORolVL0=)-CUuYKWuouIes8c``(7j?_X zTL9mpHhS*H6}uc~!9ihu#oiVK7&LEx7fNg4TWa=L7r!zM2>!)L=*0buu~JZk(TySC z42-A!^t+lS{zrnry(A_=naW5~GA0;dGeC$axf90A(@rQ@Ktdl4(amR^EWJAW%NK_G z8HAfs2Ex%Dh#;Muh0+e?Rm&fHfr@KJS`RN~PUhE47XU|){1PF3bmFT=xpjx>>)p^3 z%s&~^kNdlAOPc9n#TT?y=9koGe&%@-THZbRMF8c!^mkEQjWW={jqss zV;JHSBS=4I?`iZ?Q=Aa#+$y(q??a1lgtKaPc|v5!}#$|m=z`D)Gw zz{zrMbZ!BjGBL0-mE=AZGWD?dqnndZ@Lf$`HSTHGjt`X%ITYrPA2RvA$n-_G@O2kvWu|DfmbipSpzHi_?w%5oKAb6q{87+Xmo=5Y-> zYg>n2DZ)fj1MD9Pp`03QB_%GjcIk*?A!;mf+l>5pEGIUEB_J5VfZmesaU#jqnY5Mz z8woL0NTm$`w9wuyNO7to+>Pl^>P+)~@G!*LehVUmVrnwp{!-7Rl0}XQ!N)8Ro!-rM zz8HNv3ScZ9_ZNTL?ED=-X}%*GQhSf5&^q7Xg5(7)VHr&tAL$uC46>;QUN_cXlTqi` z)8hoTk(-$g$tne1B^Y`?Jtg|FhHW~XY#MLMO1_#OEIRyXC}T11u?9$dEdQc736=%o zbr-3Ha$v^phjRof8%vig1aS%PGjjYLJ1Re_5JtH&X%UnyQoXU3d#O7Xsf4C!(II+D z2YwzBzM?AtV1Cl6lj8~Oz9~8uOi0Lw%yeMgR`(3nAh)cUqMZ6aijIp|Ov>LVaAu;j z2XYSvJPCfEKi*tnzM-!Q9(g+b@V>73YRV^mW(W8!f<@-`x_OZOC3rVj=pX89E-=F= z@=dxNs)%3X7u#J%4o;JIey^vEIoI~I#N*dbnkJ75|N2N?c-B(UfgBf(~z;5!_XrU>J78*lcFTHT}*` zA4>Ro%CIU7aCc=p^`Mk;T z^VdIO;*)J$?lvlZ1OU>Ldwv{R_Snk@`%NdD&KbGWAQaIZ+Bb_JYaQIjDGsWCx;vIF z1rxf+6ldL1>(z^4p=i;Y)A^tUG)F=ffU5oYFkvTQGd8;Wt&cwt1l@0ZYpjX$Tb+F7?7pk5XkQo79+{Jz#&yP=1nwXYjqchXX3e&7_s|8? z_OSi!^}8m1ZosCG988AG+wpGUbyoJvnE!}@iOcbYoN5akgnj0CtdgSM z5(arWeo%nJc1?ys@D2G-#Z7G)`zwK*(n8O~nf7`6l&rNJmztYDT)C-2ZV_?$63#)5 zM^NW1@e6jDSW5O}0e)v&PePHrZW^xfP^3G&&~o%iL7^x{;BXss+!h*PpSrc(9leTT zcby{(aJ+B_@a>z#X=&3%p%ST|G`Y;@@$z-v+Ua`8Tl(8rlV}7 zm6=dokCE62zHzHB+Hl8w^#wC{j@8wjV`{hssj}B6&V7qcg_vx?s;)M7xZNsv_Ee}b z{ho6tfMNTV+qPY~f~`TStmp0pqYPh@dv6~FRPg@w;iw-^cX?)JT~&tf-?ny*bogU0d)3W?OoY z5vYjS76 zC_SL|PO2Tw`N8|}D*x*e=KbM3Ju=`8%CXrdLPA>5gir9w@L;NkD4Y6~VLf-|m>4Ud z1k{@ZaOc8Ah=V)(O5SZfHSHAMKe$J4R10QOMxVX3Fm5k=zM zxw@+NC=V>Yt|xlUSOH}b0A;{hCF-i18vurs^)2*1A3pk=0d3ec+tL*XcHhCFGQ(s& zgqMptb@sn%chc>(yI~pV>~&M24X2`aW!zC=b#6P)&4br(9}Qj6qNRo$)OmS*dqPR@ z8~tojwO26cca5O}LOi$UG~KiqAuM(Xu0&5?(g2!uLdyQ|mSn`B0N%}#F~?;RJkz$5 z@hOjeVXKAUopG!JLP45=j>%OdLZI2E*3cdEAR|VoA_3!A)cgTi&Z*&Oqw`%$*}qGw zboJK6b3en^7h%S3Ct(j^wAB4IS?L)IFEZ>rwCZ743ws}JtoC0roGkvy<@aIT%@^Kn zP!8F8l+h2|ctCEM=G)f&<=R2%7kau+*s6pw5;N(e`L!@O%UWosJJ_E3=%+Tvb>G>e zC-eVB2}*=)-%@{}ANPRCyIk%z8s@s3+3?jp;xMfWTCR` z^!5_hJgri=X_2_=BwYS#-tcl7xuN2Pn1Z#6i9z-4dppO0Ye9d$0K6p}Zf`=G-YA(! zHHT{O&6u#xX|zE0Yw9n}3(oHH0`gbn8Xeq*$m!QL36iTm_m!#^;`Fy~wH%KA#@2&t z9RYzLKKAa`m^o|@&r-NH)aZgEv=;ij3aJmPc(wHa5afqAAIP#7sP(OWQz=S_WbS}X zB9lgUa}|`50!$kAlU})>{R7m#hC|9+D+f(8w&RVN6~}px7iAU7fMXv{o9(}+mv$|X zOsb?VB-1&9Zo4-XS+!EbHyxd_SFtD&84l%Q-_PAeG?5Y2rX~=$$iJca>@nD@smy{A4)?9~8-}alnf8M1PzMAgk;%&q7(m z2^VA)KuiI``|S}@0qAur*ZofVVzrghl)0N~#~@ZzL$GhNxR?# z<*`CCD}5O6gYrxql`C)fdld>c#GC%?QMSr5N`>N6sdmKt$e1?1=P?m1(q+XilX&P4 zGWM%+>TN?+!4HiL@u!aP0kz=l!VbPN*M?s4IeVuowZPmior6^*HT`{hPWNBmUmv3Q zPN4#Z<5;_N>!JAly2QcexP9VgXN3odcMQI6$+WPCV2Yvp?6IH#=}XYUAcAgBMrcW` zBIkwR;Z?I%icp6PJ!FovhS67ZZolG43m#aXA%x8FpbAjAkEYW@Py+GJE2l2!2pBr? z2Enx^U*va70u<@e1O-or*JkgℜHMx4x$>4%J+5^DPe*J!=(`Su}83renl!^uD<9 zrFFUJNbk-MGD7aF-))g#bL-FsJai!g)gNTMgG@FDrl9NF>sW zh=I`yxFYF!^URh#qC+pamZ|yP4<>vIB9YgdeHYJl&-90(K_*TUDY?mc0iBu^lO)?J z8uwVy>f+wT)-x2VkluFJt1m7aa*1(|*?e)!ark4QDo8MN=04%WuH9^5R4X3xpukhB`?%zLU)7VnRI@o_t zJx63mycpNrDr)RobqnuZ_@iuRB$X?AFC(}|c8LxExN97fhziQHA!cUS#)*rVM`xHQ zDQQ%J>iL#_acJE>oDChP(f5ErsXz4&#wW%amtAbC=ac!w`iG^CFag#14Mn5>oc^+Zw)N0%pcP8xY@RvslfShkZBO+c zCI)eC_Y}nAy_yd0a35LV$!J#B@XcH8oDqC|$n3UJV8B)10fj=09R~N zh!)!*i^IJ$y+u4df<~8i({*%^KO3h@y9U_kkHp2ej+UFpx}UTCrYd{1^fd5JBd3zj znDe{^{(HV#CtIAX1!qpXpZSy=0OA`Kc!2|hiF`Vbl~hJ#rlQdNMN)2vJQK+4}XHGWM7do&FglrG^J2iPE@MEZ$gT%Yqq zYNhV`9PL*Fs9YV#^ldVc>7ZM+z{Igz@=EKuS=pz9%G8~b0gYYQa$&$C_*){Fgnsq7 zsw_@(vAVBnrexXZOtx*(pUfQbKfrdiyxf}40fo11A?$jITNk}7D7BJT z%|4``pYpb4Ox*hJ{!d0Vwc{@pC36V|F=3e|O$LWS+P5*ZZLXo5hIKcsPvFhB=jCqJ zW3TckLah=KTKet}E1eOzo87uPX@fotsxwR{F&_3(h$5F+7nh1i}_D!u!*BzFsP40STfNQm;ntZQtl=oFnVT zl>qW#!xW@6fEs5qmIoA)sho3aWDc@MSebi>0`XWcgA9{pW?!qN*p;7KKFuwHpA5NR z641X$;UWw&d3HSU7jF&2O3X<;?3~D`gU}eaX3dJ+?iQ62jkoYX7wgqe6d~;yEC-vTeqJ_!LGVqdJ0&q5$VSPR^v@K__L1- zugP^F9xDpbdDWr}+pb(`(V>rsw?8PFCR3n*;PhHuvYS$habxb)Q02@d8$a4VL)+E~ zc^ALr>l!r@`gWdZ>8TT9u~7kyb&^BK zMC$(lr>kpH9z_#pim0%(za>8o>s6=D1p~Z&cjLnj&(( zZdWE$mL*N!bUd|le#BmOFQe*+?;v+8@nI;B+<$^P&E0-?Ij?U4-+TsMrfWCy=&G!l zkK|#hr~6^apP1Lp-_L-R({EMGJ^+72s`meMzBALv^WSRvwhUk1Tc4R5D)vH(#wKGq zslt>$#Lfjm%$i%={~6TNpcD{E#y^$ecsvNbZWQkv2r)QWfZorK6pAE?(`!EL6U!M2 zbHQCDyGp2-9v{&-nw` zR-I5ssv_b*p%_#<+v0v-iM%PL7P^3wGc+A1P24O?R+W+}L_6IJZaohFZ5R_NH1X`S znsJQZuNx1IYGefd09Ku2xkt0j$HgzeJI@_P-_0X1PL{Pomj@+(T%1^a#1IZ6OMZ}A zbb7AaB&DXd$0LAOlWRRrU%XeoTFVijHKnP?{PA&HYh*bJ^8AX#p6gh+EZ+Y>e4$ai z*xVpckCdK7ypN#^faf>c8GFWB0egij{{Cmg`QX+=zGBi0goM;5cZBF&U|@mpEibrW^42eQ_q#PZpgx!gWBQMWWGjsIfo}TA9$tLN=V?_kmw((DzIP4U;;LUXZCBHPIb!?3ETu|g< zYTOs-{{R{Qrn55WUZH_$D-mgTXr2^|FPOyCqQ}x=UTOKdOS2lsH8_ z@dW+eeUkfaiF#gte#?OE!(g+-sj=dWGF)u4prUv@_P!}5E#yRKd!IyAbVlwya?{uyAl|jt0&y3X9CIGxW+$>G&PzelD zh*_OGQXPv-a2G&njd%9Jh%oW__={(#Nb>RcyM&7>6zvrN{rFb^)wyob4ZM_^?rRg- z2Palskb*%+09*rftDq&4hRaEW9c}8>^}Q;h_*YRKJL+CJ0KC?nv78Ims_yOyff1|N zp%9N7L)EbArjJ)b?*DrrzDx$cRyxl7dko%xJ$V{$#Quw~+m5|_!yrze@A$DWN^HJ?Xh%dK7f!*x zMtdb|&my)-cf|eTK*A*>U*CjwE|)8zr{Cv%H`)gKy3M<_3d#LvDj}LaRD7dagO>b= zWP7bP?cuj_oU{RnHBzAQAylDn*lw~|PF8+w?nv90$n)%3G~NE)t4|H}81m!4Qt*@2 z0I}Ge5^%EN8wS|R&VW}gdK)uc0`STya|cg#m_2w}p&;f?pE&NnhTnK`Gs_x6%-scm zbUxStnGd@!bW0q7kD1bKXW}~-{`mxOA)1~t*jTIHVf*-=Wv8a-t<8;u>mS`kT;ADp zCTCg??(Pe%R;e^86~WBrP3}Jdbnzr4K^@6s-li3n%+tgLhZP)H0X7z{y;W) z3ku5%0v*!=Joj&+ar7|$6PkEKIlrcw_x}L32!lZ8$fG*`#%+^3UoYh1!0{`ZW7OQf z&aO3WRG{-h{8VMxac>WZ#I*Y(**bXG6}eY5eLBv}X)PKC*e^rNHAm7qt_VQ8^-s`+ zsQ&c)^v{Kno9XL_6FsJQy+i&$I`m`+W#$M(^~eerP>Su$ozLwz?6MpF{SXIutXs6= zJs%49E1+H2oXi$VIuW6Qk`yqs2sNU@fIaT&jlCGw5xK9MTNR$zR z=NwA=sP3Fiil3<=r;}P+N16lyHfS1jzSL7pqN=lIORsj&HaGWUK#Ux(F}-L}DoLr@ zg(UH@y-N@zySSP4$okZJh7FCsYkt_&X|K#FyXN$S!^hos(#!Kon#@92h3oC7IbJOU z*yhA09b$DK;kJlmR@`(pXe!?x+5GA*ZlKlv&X(@@1jCk_e~%T!(M|FEF87|P<3MPa zn*VNk6fAwuiQFvBJkzp4STpA)k7F&-g)1gZbwy8xWb|c~+y4+(5il)W6Sq20WkG)v zYq>MRV}9`pK;m%IJS`cbh2+`RgM+s>&2m~aisTnw20G(IRvn)QP&xfXOfsig$l_Iy z+6-d_45e`};(#)sy!nR@zhZ<{vJZfcdE(*(9JGXWc`Oo}GAA5CCDgGIkWljDB8F`E zwiOfg3T&C+95DQcP{8YqyC(==FRs1X>iSOxLcUt4NC5Omn?9_2LJJseI^Yv_#Z3#n zwbsw<6|pm2&(4g~UcFj3BR<4zMjOjR2e5V$XN#?C&BzA9SWjP*(Wg>Z3E;{h9T!I% zUIp7h<4S9(igKShRO7&2bO-CU{t0WJ3YJiw8{bL8O*Uh6uNSq*&|Amleov|VeN(*eTOhUu{(=VYw23MV1O*OVxZVh+O3!)Bu5iWr8-OJ3t;?yeRiH~C$6mAZ!8uP z(bzOVZETJHkG3S$qK}-Cd#_x@ClSNm>VaiT_@u7N_hCTAx3BkJW+9$qr*QpAU0u*F z{?pU5YaX?q2dzG$7Ma@RA9Bx;U?Gxy!|eC3MA6sSfx)I z)l`M#YDyu3#Mi zK%csZ?v8K1=F~WMI&PB9cDUAfO!h$3{$>o5Ult|(@O8kGtJ~I<q%;(n>C& zOqd<#mQR;dUK)8Ew&vqXgnG&;Z&$)D#M6D$jZl^4EJh?$(ZQIX*gXjtoF7Mx7p&;E#qUbAPy2Zh0RU)V@zh z&7_su_-50$ba&jXXHvBPFq}K|Us>HRFD}4PgmYyn)4z&)1in$XKjwJ80jeogy( zWG51|{FXf&M(p~9_ac;*#xx7Q3Ti=K*)pqii4@s>w~e$iYNcn}75-q*dP?Vc7J(v41T!l^R#wi_=ZRfIfHlOh?3vh5ef=iQ}sEuvIz9 z#8bYdTOE^1i!8ouYDk`3rB5J3k87hhVKZMD##5c9S~5(c?(a~Qlol`{^?6!$Bs1cA z-jyZO8Drg>^rwIsIGWHJ<()zF0O*S1EC>fO1*s1r{TE!wBeCl@j9Kx!7V0q;5><-6 zUx%7fjgTL|LwOS6W=(FT~8m}^JtC?bs$EEsinWE86h_QOicG5lc8OU-?*SG z1@1bNnZDmf*{gX6XS%nype7xRci-Qycqq=8QC%-Vhv% zf@LVN@Fu>~@amX;ExyN=_2H#F>fK;y46nQQb!Oemcdd>!t?p?4m#ok#0Zi|7N{QAQ zyKOt=p{aCL6SRU9gD$5!paHODqNN}+PM19^O7%3J{WmRUZuNz{s+*X5}&BG&Ft#m zAB?W6PrW?ZewX?9jg`{Q{mfp&U#i3hF}*o<6Q^dFK$LS~&uDZ0qXkD!rI^#n&v+W7 z_WHdBPcOWNoiTgV*#5zC+=_JH!q^Vht&>)$MuqO^)ro>P5vzUW6i}_EC=N5C8Na$| z>3MPW#BLgkXkrh_ISp<46)S4C1ylXDuEYLzLfn1wZ~XbrV3_j&aucp{k}%x1@P7Nt zl{A8d5j*)O-{-r>Vdq_keN0LEonjI7uSqRt40DN@xA^Cw&F$8NYq~blxdpbBlrkOU z1F?4vT}QlKO|&ezIlAf4af)-Bjb-?5j>n19%XJ;jkK>;H;!2zgj)-Cp=LyTDz5%+0 zGjSc+$CQcJ$Q7?MQgo7nbbH_5O6I|V%r$oqYtlpW332?gfxI3Q%jIbX0PDTV5e^FP z>eHSPX+zIPl8doTaooy;oH^*k(-H!Hq>E830MMr`H~A49;rFdTe#tGk^Xi9fz46<& z-+w>hWan7?Sor7gC1-KP)mQ#HpWmZq6x;CosuUxKSu$U7ov-ccp%UU7zzEK?xq><#xJanH1z%bYvrx>0Q@dcSa!;5GSuvWcfZVuW7}s|32esM#%+mzzG@;0 z$NvED9I=oqT`8reh2F#sDoNs@kk z3b*>Pjj$I~w%rhGv!N@z(qw|whMUVSut=Al^s!Ik*K=V)ngXd@*|?MrG04gf;asWz z?sNmv7lB%fgip#rv5$N`7}9HDifpCEIyp+RceFvAlnR(+p)>F(;-JJkd@b1OBnOu- zPLfDOTBhzy+gad9K`~Xv?C!BDJzzMU$t%|4Da$7=@%UgxB0xwd2Wy_7N%&c%>8*8C zZW(%M_iN$Z#TC;E>AqoGi>v}bOt8^_tpyseu!A2D2OM_+Nib*~M&O<+9v{n`kAK)G z+9?v{(*rA4qW%@$-<&mN;z0fnz;6jOjVk>ltr_3P8Z;!=u&!@Cxb+46=GqUZ9vcPv6{wkUou?+_`s2CR63;U)bC> z5Y9;XVtPkh6wjwvIax8t>&|Lgyt4rDbt>mU2YAnNfjk+qdIu7$bK<;Q zrhP%CQ(Ib#?GX9KqMnuh09Fo48R;~o;5vK;0V(zRrqNnpcvE1^y7_=@NDpOqyl$f`gYi?;WTlz!PL^W4 zJqhvZtzeqJN-WWB@@``+W2Hg z<<1)n@0ZXL3zZz+lz6MO^->w^eQGBV;Q>P^hAx#0C+M|Em4^=g!bS|QzH`oe<#UT! z=>ZqFBkok_DhiQ#^XHTVGc?%@gZxo5rZiYZLPeaSt$T)g;i z?GnbUd9EVknp!0qx?H%Sh>rzK@P*#P1DH`dHc_-zB2z=(y7fYk&ZElL+)hHRXoAUZ zc_tx6IA=V=RF?!t<4TZ1r_5$r!cikZg6Zb0*E{1b0Jgy5%Ox*sll1>>52Y>?uOl3o zc+OVUDs|O@XL1K{abOQR4Q)1#WyohHxLxZX>dTgAkV!4WwQ1`d9M_WR$9&x+8Q02A z@)#a-iA!gKdmjw?NgqfpQkixLE% zN)7qow3&R>KnQ55i?pEb=Hq%(^xR(fHk>-fm&IF0A%w*j-UEgiqi##1N14}TV6+1N&nbxIW^O8&{H{Iz_y?b~((x2jG+ z_|65^xZZ^DwO5z1|F_@_pLjhiwhlG|Q{(Xh39Bxq&sG%2Y6F;ddYIwWEjOONd?bAE z9 z!vEw3dyI!@6P~$|B@5;$OH}lHHi|<}_aDZF0VKt{$UGK>%caeoUWVvfZhpZmXP_r& z<0>^~NOb2B(fH06aq$Y57}hn_jO^Wp%Fn;w{M0sb4zbImKijUJAmvRfZH8JsnGG|c zw=)@6#4jc@5-Nm1G}{)W33Q(~(=ET(j)?Nj zAwAN?I3=mjsjIP1`M+9!CaU!RY8O93F6Yep-n8}I5U8ZOfjjHnHL zsHLx&t_oY9>EO6LeSZI=?Vs`v`RCQd%CGW)ba-1WyIs0{Bfat)E-ghkCYkMibbN5@ zwujB`7v8b{GW5MI3fGCyW4m$Bg55I|(>c1xQaPtz(SpT&bQGT~Hh4`0(7FmEMdd5=K2Ut+2h8}Fzi-RyPR{k8tiMoo?f<6ZrJJsLNegip;y3xEmJSvtmZ zPL0xN5eccds>c)szgb;OiWs+Z$N=s0%iT z&Q>@@rJ5B&0Z!&~GAq;3-%N#;ARXSLvkPwH@`XcNa!-c@)H!6cTC~)`TPP>Zx8pho zN=}Y4B2E#v?X8^jTJQ>e>DjAp4|am}b$r**Tsd6Dz*W0wvB{Y`Gk)*T{ho*QdOKm+ z|LRNT3{x8siu3P-POH?JgYDBuA<=adJSQsu;I(?73Qv>2&V%*BM97MaV@01@+Lqgx zyv{J@30-wLxqFhq+~<$o{pul@Jlm_fGQ-!~Hs13sP{D8T>leGOA?cqoWqN9XCYBz; zO~TXkjE_>1e;BY?elGFUNvd>#HT!tHCF{pUEC>KhrN74>g#FGcR(^w}M<`I^egUDU zC?m)O*r9Mhe&j+=>D|!C_%?g_nK*E!P3~{@wXVL!_g07{kGQ?Q{W=5VUt;jydTXHd zrikFzd}1x)`lAjwe^e4#^5yEP(|#q#Q7=~MTcA$!x`BF@j(X|lW7ZHfCw|KElQQmO z0lnt#RAD-+zoNb89LqDFhZtFW!ig=Rt=>9^?{NDRYGujYH&AFh&~h?D-8L*lZ0r?d z-Q?7jAA8LXhrB$kFo441^(aApw=6`;P0kL7QFXWOkz2$Y_HQEiu~lkl%uLOsIe&*)ePK zz4TLyv16aO>Tb}-&n)ZfZt)An;69J*VS9aI^t`OlY!tB9H%Rc2)g zLE19dfl^__24Y3$d(OvCPMPM>+|Q149Cm2 z^UXopRw@VGDqVGC!dfwxf8Fe;Ml8?G--(oJe$FTSKDYaml)LdCIQQ-JjmTe5W5LotZ@k(FXeGPk zSvSUxhH7Og#-Y0vr}H4I*!c-9)tHKaJ$KjvP&GxuAFpFbFXNb71RF{89`9HEMh5C_ z+fXtTZ)1ewKYUDM#9skm1Jjk|9u6BH-~GKK@jOzN5EP6(d~gR_t#QAR&afzdNiw=* z>F)S-jLCo;k85tqZGw{|ek3jyp;S85Zh`mRa`(I%V~>u;U6r}-d~}O)Dv7q)5t)&^ za>;?Y`FBisf_a**5T{qEqQSHtM2Zu<}OG9IuSNPED4><=fDsY6=)K z#D%tywG!T58A;_lADwrR=bUa^iUw7h%Q%BxrzQ($$+f zuLzvuyV276rp^91*l1lW~{ci0;MaXblC2j(|T*w>HE!*kNk!{W|>bY=KA<%`v7r zr_6+#M-l6;->Jzahr-P*F={sT&B3_Yf@k%HteiC`9~L`md6hr$0xf14aXlSC+6Ac* zHoZ=c%VrrcZSATUv?&iy!__=EannS+@E#grJ%Y5ID%2DJ*foXposQ~Jv6eQHH-uJo z*#bL-r1hlWgR(T3c)E+MPu`89HsSrw7uR+hzGyinHI24EL=Q#Qtm3cof&2D}&!%;@ zCtua-gqDsb*KZX4Xb(nRP7eaW&enA|#=bJRVm&od(?pHoZn=%TqZR@wws;Th(!V8b zVe;!i3!a(~FHd!Ui7%C~gV*cVon%Hl_m_!z)+T+aT)qQYkcau3L!fB;7S_!&fiz#; zE0r;W&tLK1eXlL2qZY9|WV+l$d9TE9nKC!A>nvxo{FuM8I?e5l!~;8y>YFH&AaL`| zA}9Ff?a3{HYxhBnZSaSM{Jm7~ddw%k%8j)9UT15vqp5i+h?*;EfqB;qgVA2_yCp;x zb#JintC!HVUWgev+U{F?r0BT=O=VpB+-Cr(4)?z)`qyrlHT66sQ|E4GA`(FnQ{J|) z-Tk@JzZMip=DnP_XHx>qFzXwVQGC%w!)II041LK>RshB3q#AmLka}mQ5@!m-I6)I5 z76x(pkQHs|6Kl$3uJP@UiWV7Vpii}ckH;ZtjwIBN5r13?Hs6 zhhE||RaqLA3fI|pP6{J8!omM!oWti7Et&kguIBkdcd$lMb0{WmYui;5Ejn$k=};OZ z9u~Ejd1-SfVnZe=+r2_)c*M?`wj`cpQdTN&Z%+3J0wUTaR(7)?n+#AUx zZGbqh!m__tRA#p+t9BBTY`r+FG)Ny1YmaqmrJr8UqROr-(P=i|2dI}}`170*vnDxw zhTekh-sK}^nO3`28ZVUt5D+%b-TEYKc8~>u*;zH@b}pJ&c3|=l(iCpOT8UoFCYO3f1!6k>dsI1VJ^UUcZ;xIAnvm3i3mwi!7UZB3KLg{&(pguaoyuVnVS?r5TI zxdXW#o6kPfjDGoa@x?>MD>r_ey`I2bm63=h`M;|nQd6&o((TU@6y1y}iFB{1iHJ?? z8uEP`b%>Mm8=52Pe#YT)VlG9+o4G*R8WV9}Ap!<3r<*`zU^luM+pa2b2b}$TT|N$N z0!r!R*SP}x9-Fw^C+QXi-T%H2bxYMm;GJqOn38}p20+fp7?zVafYte2eF@?DpRfCL zJ};`X_YRY;#s%sDa1)Dcv?`sZu7neDi6a$|{?`{G>ArqkF98P?9zHXQgQ{c%+JMvZ zJ-4LB65Khrhc!G z?{9VfU;^nT3d9Qaf9m0X`4>Hp=z<}&-0IPN9AusDby>S52%T@Jp0e+Lv+Ov0$ttmvT8_WKFd*xk_+|=iC1k*PQ!3S~*}TQAyhIlyc@LJ&^S(rnkDO zZSfO#h%M_32S zC&sXsJ9=Dth+oxy9Bv_=i;iFpxGq&EnN(8Pq~~oeYsjT$*Jm7m%n=B z#mKWRuhTmJRsxf(ASdISwnme3x@k7^dsdXr4Rx4j`Vca8Zj!f)34==A`X{>bV_Jlw zjeh)zXSg*CP{$fsi~Q|Cd=Ud|znl3fJy>b9c)~ROTAP(AD0kv>RtgL5cz;3jxv1m# z^qi)wyc>AL&ZCL;Jt*uE&+OG^nGhq?^81H4uS?a^nJyUfs(_m~6T#v*mva$>SHymD ztcZJ(G2pOV3&6E&z?g5yvRV4C3Jpw~z_ezx*Z?3 zrEuoymH4QVLC!BkapdgRjw_5!(KQs_K^D||`RM7TEr*sHZPGY~h=(^;BWa2838k$S z3)i~QBIEtMxl{~JjUhu@m*WTZ?owj~;z|cT4PIqhzg_z25m#T!3*FCAUC<%acrW%L zBGcM;%D&lA)m51ah&!Xc(5${3Xd)s={`jkp0?FX7e=7%*0gZkDWA{x*K>tef+rDk{h!$q_Isz3EI``>5>30py!>3FD$X1Y}0*l-G8CBDV5CV6s-7 zwOJ&Z1p=Zj!FqW-5GG8U*lmLuY%j1$6)B@##HdVlTAI_RY?%e=>BHIHc(46(CicPk zX8)=X#^v6Wgl&KVQJ(09Bi7s?-dnTXGuWJXt35t=bXh@eui{ZkjG(k%fk2+caHY{Q zPKA909syX#|0=Gh;b!LnFk-jtOBsN_T3;bcCs4gK%Rsl?TJ?Q)2j)pq2of%;Th7`2 z`zt<-*vI&yVpB`bf#OyN*!(*ia1L^HL{SV7Cbss``-(?Wp?mN!v|QwR10cL&<+9oE zs?4v_x(mpeCaI8U^}?-?wl3M6ji-V$Vj{fl2@D-ZN#yfx0Bzm%SBS}O>{OrNTYtB1 zd9|LTvt__-)3L9 z>^ZbDUJLAZ@N-nRePT@9-z)bJVYQNmeAAw#K>7g=F184UL^4vH1S~ z0Q|47uzTUSzX?~pI@l@b7-96LyfXd3N~+V`{-y{>c_a2Xyr}KELw!|TPMl2Dz=S57 z9J~4aL~&ib(4hFCtG4`LG@ivyDu?yxs0Q^Go&WXQlgF!n+H?6#vvu3E9PS@4%ISa( zFQE~2OkH%LJJMpHI3<~g%5--j7X4?X%%K3W6H1V-+IFfn`yX=AkhF6RO;jesM0eFF z8FN+Pho&%0FI!IdqF~7)@yxq+Z&Ij^99?qLkT9QhSjyeCwRvkE`J=wnlQ(2>7k53f`At3a&-+tb-%{mXJJWz6<6-~|f29;(B(LmNE4@a3rjaI0bYTl&BK7OG7TYdX*OjLK?9ucw zQ{S9HcB!+p9!M0x*`+1R$#()m1|gyHWEnb1UX1(v`O6c9UZU~YqX4l0TT>$yhCb+l z^^ZAms=H0+_Vo`;sk9k8IF-GOR`dY?eo~qq%zPJzkYE?Q%2g6G7!HYZEWA4}| z!$`B6dW@WzbNjQ(R)JX&_e}dUU1H`XmAtxtaFJZ91ht7g=qz8-)>45+_S)7WvTq$y27OvuQk++un0Q;8i0IMa-2bFvl{lfE3L0eJ5 z*CKcs@GC3Hpm{s3@KRst)O!MwG^wV>$j>u|EOT$`pgPO7t7$J1$-qvpXHLo5Z-3ET z#NbILnn(u4o-8C1%Cw#8Ugu6|`q{AMRbkW|_asiCDpk9g!C@YW-SHPUit3iMuJOB| zgD^WUnHjxl3m|&3V5f^ikZ8L$d;ww#I@025@x^LxMPVS`-FL2dBZgkh9{F}OVp$Fo zx2kqLc0I@rb4FXH0`u3e4ku84DCiy3s|pu-zg9bB!uVhN&Lx&LYkCepa<%T_66nzn z20UQ-y=xKHC&OeiKL-EQtOOFp!B}byM>?-wo7srd^*;>{ev$?9s8UhQjFv{*!Nv80 zwQ(VmJnbiAInO=Ky0+vm_AbY5mJzN3==%9cyBnil`b<$9Kj>a5fYeVDLD5&s=+o^X zO;yJ(KbNO8p!=UcNyL=WErl7@HQ4+i3Ce^hKplqp4s0;uFLn!$>I3xeV*i;4ck8bI zxDI?FEe4Z%uRpde^R^wctSrnOr}|SYS|cAMEc8Ak`H}9b;ohwQ;=}Str-qv|3R0dN zcM}VWYH_g^q-!}<7Qs%OJOfsUJ2tVz49gJh$;?xm(byi$iG17jeSo&ao)RT_`9)&3 zx-Z6yO5r#4$X@QEomhv=9@TIt`GsE~Mi!ywLAoofbaf(f3`a=NPSCQB;j;CrLH#P0 zfnXYY6u>Ax13|pFV?x*alo9-g|T1;0fm>duc+gF5=<<$Xid!G07YPY zfDW(UtR+Yug5NjjPKS;BG&2B>j=dxJvu_wPiykcJH_i)CazdQ&wCN}JgcvBz07k%4 z-x5E{@H-*aN7M@mQ42{yWlZU-9xXdDE_+)gb-o{46uF}%A0W}GD>`CjO;4X!k$9SJ zZayrksxbZD3jHntU!OQz?r&$X2EK~L9p7nV@@XRL>7|5?E%;V&Jv_H6w-R2CGt)e zXwczO4Uk*5W}BXd1~V`wpQBvq4|<&UzE?H?@h#WKgqFsW(CHapsM}-! zdLjERH4_F2Ct5pa0O5nkr$y*C?n#n;XIw@-;9vLyUZ!CsX=nV!58VA4gK(WOvwyi{ z$23K5{LrU{FJIyUluJ_)o}U?^bc>D{r_pW~=hK?Tw&lg!W1B-mvo?})m(DOANjJWd z-|>2?ZPB%CJIaKIhl3$yW_GQNJF*mNsfDHq_PMCkEr$E#-R9t}rShVM{{TYa6CKol zpC1Q3qhk)2TW3QQ|JBA%V@eSb=qyo_~SvCD*;-;l!a;l}F@L&1X zjZtLkbwZ)-o`2^qt}>wmJBD+ZM*TV{42PDW52do0tm;TCEF{tTRFK(n#;4=aQaKz4 zP14p^1~J4u6|XuKAFf-PH&!#!VHJ=P5=D01JO%#D9S;}Ey>ZODzv}w+shapSA%ACW z?)Pf><-X%*^UIwqVp=idF`{N0RTM!000?lvdxk_+SE$cD;r{--mBk|F8BI9!Jj0Be zwZuw?3N01zby&m@&7KEO4Vxe-_1RVGSrNKY7PeD-&tL9nu;RWG9~kCp{0B(nJbq^d z4%jf)T~Pa|TGx1EJUvoh zu&Q~>K;aF? zn#Fl~>ey{en7+=AKXIPlSVG<-2a>sAWWJE_fUa}5TJ|OBrUDd7Tfd{>@m;iiwMORN zP5k~$>cO4ob5d%v8aRmz6*;z!`sGXJRc_RYOv*y-@Mr8n*|!%E`?61iGXDhooW5l9 zplmbErtuI`Q6*{Se3(}PurusTeJ`6q#dP>PV|LGE882vg9zTa1A!lK-d_+mYKcQ%2 zwP<7|1Xxx!$z{p({bnW#qtHjhr+)IHgc|w!NoV>SOfl)6Q2x?4wC`o)46A6Jj5@9e z{|2v`cWi#Yi7tQe_l(9X1SapVpV>>LU}zlW`yX3@Q7+`L|M_K`^nv{PC!&y1*&6l5 z@I_m$%&S&{0{fv+2PE52_W9&nh}1eyWU30AGb*wajXzp`g*nK&^3u=%ELQHko8*6$ z1*CS}7jjfw_GUxgvh>6@0gOmAs~Jp@j6>c$m1Ef_4xmkTX1y4imU&h10tqMa4M%Yuw$VZE?YK0MdIFwVY*u)cL%OfcE~&nOSTOMD?CQ;jUh?8uc?8k(G1 z8!>O1(Q&U-bi}U5zVlLjGmzDxJEj->+?5;uy_?N80Fr^Q9Z6T;wfP5eYT}lPy&SD+ zGix}8;{Y$btY-*WR41zXtnMjT0)rJe80T>|yJaa0Ntq2KFMg0mCoyxX!4a5_U*%G1 zL`f#wX}6c)Hfw z)Ia-4%2Zgt`lTDfwu4IKFy92N6T7qCbfMzEuP#7*=ZbPanoSn!!^a(xO?B$Ri$_WA zgzOb+GS1(GyD_yXNFvX~;Jf8&F}S;Bz_p>{LJm;8ttN>>vzLX7AC)@XG{#pqCE96T zxyom29@M(f@x|E8d?oJM9g`;L+SUtMm0{}|0}ZZ;K}mgi;JV|53yJ+OIm&Cv!JUre z={7F~xuP}899X4UZ=e&#dZ1N&yvScv<+xwwkNv)4j(dpNqJ7`91ceQ)acae4lS*+) z+O^|b-;F5OhG(}bbN;5?jzJy!>6N3iJ z#CjM>w=uP(I5kUUHZ7m*lBHf)G4)}efnmR$6|4(9?4SDKBelZ(-zC23j4Z(rO9Kgb z;2tVO;ksuQrtzuFycZj$_&@-m)LnXhb1XIRE^622)e~CZszV?awwI~%+6lNmWY8il z<5NC&=QzL`vKX!#c=mx0Cj^WXca=)@t^sjL{ zYEqEH0lEApYD25iC2lD)T$~|@5f(WT^6Vq|2U`myk=Gy|S8@KEcP}kux#o-Xy#j2G zOuT7*#Rq2AI+@fu8t&R$w-IgQ3>~-W9*nrP-*<6P@S)rHmvODEC*y~X(ofdim(Vlb z!o7iosW+xY+o}S`7JYs0r2hI!&CcxL)oUa1>|qqIN&D*ta40vMW~w^Nx75#Q4T$D~ z(~dXcp-ytpieHmb22mrFsu1{m)(9Np8sUX5x@;!|92~=1$vldb6I&%UCjsYxB#nY4 zjce-5=A1t1U>V!kPJ=4Z!KWCnA4yFi<6-0?9oYe@dW zYp(bl!cky50Bs)>FTI2TR-4vmO--)bPgLWAmqN_t2+I}jx2vMbR<2*##+mFI@yR{b zG5z;1TJ1*Gq&l1HdPh&;r+Fz{Er{}S5N4}tLvfNK6t+;=$F7SwEJG$ww9||@oOgWF zD;1FLtJSrXaKO%Zkc-vzi8k<(H?03u5m5$L9f2>FKc4)q`#>*pP}E-QO-P3dxl^i* zrSzx!fmRXv?(5n|aQs>q`cSVa`4c1iheTbSSe}6cjR80&F11sK$J2GeQ$4C0J|KkY zc*)J7mD09RVHLec9tsIbSHXxLFZ1#zVg~9nH(UOgx-Wh`jJ_Cg=U2Xhxv}V-o(ZD+ ziLW|U*gEEf$#T}-H(1quUgWT%0g;Q5U-Dtw;GvW( z=f(j+Vy%pLBqMuYmyQd!D9{KsTE)w6!ZVRe(z6i*ogRvn>oj|)V&UEwO`MYF z2xo(iAGs=@WX=`y0z$~?a8zRdU$tAIhD(Mc0sdT-i9gETOme4qTJPft#}!9r9_QS> zK~Jg_LRAb%OWCJw6;?T~s3oQ6UiDgunqB~(23({kyS=L}7Q9+EA1?F+iI1rDNwJKY z{T&-D{IzU!X$=&j=J8@@@-vTDZ1aO3?Xfi6enJRk2ITqkMKPC7`jYuD8r5_B!S z^4|TZPggEf2(i>(Rofhas%t0qp%+*au5#2o-bCQzg~py%l7?OUMsjrw zOW)_0-#l9m)`?S+hb;0?Is+GmllX6Zcn*^T8Jsv}@v?|Fl}KqGVH+=oNnz1mIFGb0 zqBwQ?Cp~rP5(gdHBOMx5jP}8ND}2Dz3sSq;E~gHO*u0sb`d1xs98iJh22pUnB zgNYi7AMhZ)Gj)&X=gV$rGABLaG=_o?h@xyf@dtCEQQN6e|K9=l_rnA^7r1EPZNq+; zR5GD$`(|?Ah-_M_QCZjU8DTn!Gx}_LwGPu{(|HD)M-16>K_>)ySWJx%!-Mj=4ofbO zB{?02k)x44>U2q3wQ_&~V%id!ZfJ%ac1pAuMwOk|HK4=ws1(HoZ!gwRc^d-b{dc$u z;6ol76Z~GS$d=$y)q3eu9^M}7wh1Y$t$~$OanW?9*Rkx z4|0ELr0vAGD$0(1YaJ2RE9un!>}X!L@=}E_2OA~hI3<$%QVF(PtyCb_jNHVf4c(>& zO;a|1!UpPrx@|qS-G|!_&z@d1dC+9MsrX3h7@S$GTS@Wq+^J+%F{o8EnUVC=5;?FC z`RQBdcuCJ;6}PXdRKBo$;0E^_<8r=R7hrrVOgB#PTDP@HWMmJ=wFw=sogoBLnvQ$u z1#TILBfBnZlz)hBf%SwTuSv;Z0@L0H9bJk$vF$e6Yx}d!I~lgJS3|O81G?H^(RQc; zIF9uyycq56)CT@u065Zp_)|7FcCtO3R(Sg1enFk^LmZ1$U2lI1I1{;f`Lr1Hz7II0JVFwTUWl<{~0@R&2Je)OJ$YU36&pVn2Ts6 zSzR*{qSIJRo^jxkVc=q5q${A_m&M%@n=-u2a#%~$0lqOU^;d+)ZfIdWH0CdZf2^xE z$l6$L5uUuyXnK5|`AyBW`S_fv&Z$@S5punzz(n)mxN_tW_ayGi7cabm#=@QR?w3Ea zFq|2GCR_BF1vWkSc2UktosXrLvJ=!!Cc-(BgH7{%*HhV0lyf8S+n=B;LToEWj}&ha zq8E}GFvXcwgJWaJE`iD%Mf|1CQ>)H7A(m_kUN;|xsYev53z@B2p@V}uE0mLKmuzoz zt}uH&UZ`5tymO!WCTnzDE+eZhUbY~p=W}d9hhkB&81|4SrRLw)lF#!yfPW67!Q1#r zzMYf1*RWuq9NKsFb{S(WG-qX(-*dqblKF`9^X|1OT<}YE^wb552f5Q6wNk&Ut2R;| zj4rmv&i*QUeK9!NvY|}(@5bA<*)KC`KE9=bt0rz`BPdVEuCthu1uy|IQM)?o6V9MK zuiB4uFoJKB*Gpzt(cUzjq(CkUVY7;~Hlino@0j-<29`aGXXO!zcw|HBP5DQ*2B1`_9g6{cfn z=#Cx>4UI@2&VQ;9CYJdby^EnnTO?Rn!PT0cK1A3?;|Wm`)%u3enkw{ZE4yAJ2S^LK zeqD?~*;lLEO*Z(RQFO%RyJ`gkOU6h;uehegW+$-iGlz?zm1KH%U_B?e|6p)hoG`4vGtM9PAlne21EbE z>=_hmhpJ(ENs5BLe!x1mBn~pmy4e7kLC0KAY%eebPkFeQKt}=`9ERCXaDG8~rQxjn zh&6=(f0?``Z=0j2ql$egGu4*#fQ$T^UT!s;sgWU-?x`idoC)n0eT^oNNz%M)Opx#)n2UXA z^d{ClbM#;fci|zEO!TWc9*?~B@+~LKJwkX zYjCzJz$7-Az?g!}Ma`JYzr%NfyP*~6;4f?Ai>{gds)l&qRBz2x%d`92F59S@QbHU@ z6}?SKZooL)lxwkSM6t%Fyc0xxPiwQ%a1Bhqw3wkJu6qL`T;F4ih+O7kD}`LD^|{sH z9Qs}KMQjZUlj9m>W-_->hpu5$`%O{Hgigttpmx4_Q#HhF-|npR(!IJJP6CG!@OtqN)}2rai4M zJV~UaQQ6tzR1e;~Y?8ZvOM__w{i}=IB!2Mf&B8_VYbtkLbEgv=n9{rm`F^vZx=IRT zcC+h6FY(8KYSP?bpi?6f3cwOefV98Ul}9A++~ zP-+6%aNaN?!B=gjHfw(#@0G(=EQY#o7T`V9WD1I*8l z@lM-NebCi!GLDP6hH}xKe6UNIy``yh;eP=39y3c+b5w<6%X1a+53U{85>P=sQDZM&H0aGe z&ZZ6V75~kez|RM0}=(}oq-6RBI$bedip(d}_FTePq{`wuo7oBQzEW)N7Reo3Q zi%ljQihA4FBKQ%i^KXUk&fArB_3-B62XHNPxW-)x##HIxcKe^}6PBQxnQATNn;wGW zizB1*a4JPsqVs|inrqB5VjEl%2L6Bjhe(HrD_(l=dYLy|2m9OZ+4s@KQ6i{VPR##R z@GBZ65f&y3lcWQu-^gVYhH7(K|JF8GX9UF9OqaE-8$tG?Jr~n;_~2idSVs}9{0v6i z>PYRaG1jW5=qr+O#_HJ4gnwa55?81v)Vm3nEY)xkYT4Q5U40{nH9w}*MMis0#}~GM zM-Qk@s^df&j1=p{;-X3~*!Am-we@H|Twqh+R0%{lB*=xZ{wn0FL;V@E1jTz8E~c{G z3?dK2oLIX}8UUApEvR^g7W}V&L`54M#FXY8?dz(Hw| zvdGkf9_qo(yn(H_HoNUR8Vt>D;3FN|aMPYR7nM0-QL&yVxz9o8H*kYi5L53qIhIb8 z0`g;T*tBhZ_kQcxQ)z;+SuO>pZJnWbf9xSRK{8A4bzl~DBW$WXaf-!+4dpfEh<61Q zYtA->#;G73FYGkeZ1z<082ByT4X5^u4BlF-)o4NdjB_17JLs!e*e{dXC&?z6d~Cfs zpw%d%`z%|GygEllLL$;AxuNYA1A~b36_8p3a@!r;bE`5bSa#zN^vp7KO`c{ zK)9Tsasdb)S-)kGvjbRu$xTHw+zr2xA0zcH%#M1Vna|X6GXD?2S!^vWW#0|iV#9k4 z)wN~L?iZ?l9K8wx3azZ}5KedY64wk!cT&#(d3(|J!*$mf*~g^QE$4}g?(AX@yM9q&*U-nwXVb&kD8^D8{w>ZgfQ0W^l8!o*BY(x!C-Tnzh}a-Nmt!m}X-J%$tA z;A%8;a#E;Y+od%fjtVi&&k^Jh&JCdk%KQ9h)#LS^Di{!3jL0!3#VaTzleWd+aw)An z;s{VE{~bT)EqR&%iAiUa$#1x(xV@q}nI^^FpbR~IUV4wS>S;)gdWInx{;L-4s_Iwt z-Y~6rS^9lok26h87t{sc$90pPB z(tSPDWpNCelMN<#zkwPb?yd%=E~b+#dcUFtW*0h2s*CpcpDMwsje_d*!N=or43wETB$a= z&PMoS$Ygz}vG=|$BF8 z6&H310ubGQ6;}6Ex|UyKdjlp4fccRkUP&J-wCZ#lemZ7))w#X&aEUq*In-+%x$FMp z@uhh3n~#N_8!QJkSSP^bjaMVE=Ff6sY9`wOwq}fICg@P+ty1Z?vqBq6bgO4a@xkwR zcOud!-Y1L)8#iUY6;B9_Q(Be~@TmURhPd*_U81-CwR%z`I+lwjT$ol`*j-4lf*Yv7 z%C8AP*|MtTg>@fshBD1sGoPVh-XG+3pV+kuq77gysjlqT8KFi2GTdXXQKfs`@+9R+ zVjm4xXLe&gb{xeqo`2aVv(54->871lZ9UFC*s`W8OC|9Ez8Ao~p;L zRuuj^tsJGYB##Nv_0QVpUp7CX@6?)}%sbt{_+N&j(2O)+6ZV9+6i9g=xSLxuo!h3( zvS%{mRA70yd51T=p0p|e2>GH?Jnnd!q1&hLq&xvHcQW?So1_ zfMABUy%?nB{tIaAw*;~fWu`2RNfHvG<{}anV<3ZcleaGcu%^52Av7SwxOk^0_T}HV z$icHWeRoRET*T;$_ydKvak;8$TmUIeX)%PH}%LYi}kw1lgd=PnwAw@pyRP}6hJ2qVZ$bD6Ef zJ0`{h&=|H}5OCwTj`?rEgNMV;=!l|~8|Om>9{2U1`djaW0>xHlvPPCHmG9#A|IX{^;d=K`aRh@;N{wT@78`iH&vt%ui;G^UPT`P~ zN+V;Sr7Z?={z8G(;E+l@NTKkwi}1JH4NPkzN}?7Ls}t(zFz``H?PUygGZlqTHqlN@ zyk)Mye#!;P5f3Gg3CRk6K$^FRIVJfQ!MS+Pc>pDmL;5m{2aLu>j$M8u8pW%?;RlPq zKWwRHarenbMt~5}b#PGB?9C5~^E>N)uY7%kW*@FBH6Isfun8pDyk{AdLH}{g{wIUH z_i_sdEBO!L97WVsEBII-$uuM01U_2#pn5$0Xl1S%qP^pHHj(lK*1c_>g|?Gq?_Nk; zH5FdJ$$%fQmArT)4M@{(*Uv?Ttct z>AKR)n{i($9WL;VKR5)Ov&(E+@#>m@oLC%+vcGg*hAYO=`0w&Z2Lz=eUsKcD>shUc zO{n+&8he!qkZzG|`@J`6d+6@=+|u(Q<-ZltC!7afj~(?wqdP6*u4(lV_j!i;$_IQ@ zOkKOiYV-I(sy{Rue)To8;F-0ZxLvP4P7mZxPjL?D2V6|Nn*g+fyO_)T6Rvm|-;QWc zT(@7On#^laO|nU&6z6v#5$dEgZvaz&CEv_rP6IlG>!_>>-&fSPh&hZ+i1~{!-#0Z0 z+_FD!LmO>VE(Zy>Ej3uNk~3yO{IE z_hXmd{S!g#x0QNkvrGmn`5v|QO|s*M%GSc&ljxHFk|K|ualb<407SP8aA zIo|!K)<&CD9H$bb(=nh3i@XlTAXS3-5)icLL#1wgf#7 zD|O~bfk-%1osWFW8pt?*sSuRyuO4MYpYTE=_uUZ>g-_TX$v;mqrJeF!;5SDivlC;% zMs>E!Eg9ayBhZD`-!uL|vE%BQvoEpNG7ZF(hChFud?jI#bZhC?e&3L6G%fXO=UcwmuLarr_8^{H z-cHq)0e-(v4lWw+)fXlqsXlnJF5~ZBNAtjK(hVK&JwBV#zfCDIrqycjNu?k$jqXM& z>lQ$k5~L-g<~(PrPOgi1gZccGlJTji81ly&Hc|~ZrSh`~_gt9+_gsK&4#~MA4W_~t zFov6O6-#;D_1{l+POoWP#N301iD}yT8-mdd<$%DLB%oc)C=$4-kiN*nINa;T@O5V{PYyo+#w( zpEd5#3$siKp^V%fg28VrXs5Ae`nzE#gET9Dy7sC?V}wXvX&ddu>QJZ++>WMeC2VG|mPC$2f#ilXhk zSn%GCxy3*itqFe^V|e}LHv}y>pGsV%`kj|#Onis6DXw!aRo+>xin^_}a)3pz!|ESr z*9yLA%?e-=65?M8A-6)C!s1N1M9I>%upS#m4*z~d_McnwzahawF*o1coG~jwCsh?D z@%;R&_)o1T%MUL=GoH6M$kEj9h|6!olJtr=dj_b&W zNtO8nKvJBe2|(=KWNPf~(fFuC8F#)tO-c*#cwyn=m0VLy7DJ0=2o;{U`?XjAH%NsI zMN_^`ty$7cP=`a{VFfqQIBq#vvWbRq#>Q zRr|J0wYQG>KoSafgaEIEgaj^0oL!Un_U^E{A8H?Pl6CW)+l+jP9UfCiwF|oSv4YyP zg#P>WHzfxU20^07{sTx*GTW8CM=6E0MUS|0c9t>-ifT7&%$&KM0IOl?xx;r^of=gX zLcw&6hqi-MA9__zbQs1Q5Wh{VoU(|nbVIi~%{k)#ZvtP6Q^GVr6ye#JGlC5F#R;ZF< zr|{L^RzACPxNr&X+frp0?W8fM_uhN8p}Vz<)~ZaU-gkCB_ieyky$bhCk5^jyaNeJ8 zv3rJkspFwsL3|+fg@ZhpGTWuYAUw=rF@nF14-9{Qb{5 zWO9#v|Mny!Q@EbX{#R73Z0(*5+2jLsp|keqIh;Y$d;GUyf#OtFCim7AKt_+7k?_7-vsU zP*mX3_+~HSrNnSWZYHq#DOE%&XiU-?W)CQErTnVUGUP7l8XUHE*fLyh$$1UCO#zds zc2<@OL&KJe!bHAe$ov`EhTu>8&>zoS-@5qOaZlZY0w#wu+afNZvTz6b&v8dH8hri& z%=Ldy#Y;NAHxQaI%h_AwQ%mzn24s?E4Z=)xx0H)E@zU;Y@$w3+m|?Lt?>~-GXQXu0 zEEnpMVM7!v`#3q^wbsDsih!gmFGrXfA(r>v}A z5b~CzltfgVV#wzW@JjSinG-8@x8^u0vBy2!OE2NKxBh}X5yW{=^Us%q&kF;TnH1Pr zW)XjJHmj+AZ2DbL@S>aZt&J`~NlSS#%$Ly-5^Sn%yc74zO-c>7uLAAYs9rcrdsH!N zN!Fg7Fq9TQ{PL~iydp2N+jQe)TE9(;5YbpUo*T`FkAQVp<|MY(dxK1vfxP5TY{Tcr1355inGYh>|=! zBK9)8e>vUHx>#yU{tw3<1gV;JbF3z5+d_iKD}Kt#gA7v;*anwo)~mYzNb8lQAicIp zy>pmzA7aq~0dM1Dr zn9Yo)V$*wHCR2A!m}tk=9b6);6UBQfnkYLCM+oCS&A{MHV@U}Djg01hK>nT>o+WXjI%PQOsKip!L~!2$=nr0yB(Y&E?>@H-P_}}% z#OC4C;gx+lAxhIF+k&yOQk)vD0&@%;6H9Hi*gFUO1fcVe%^W&>IWd(@ac}CVdh?o; zilr(>C6FV2rzY9L`zK=CJ+I{0%1JRbocngWAQz2FeW;pdu0wr_O|;*#Z#xiLHkLiq zdq>1ezfqc9lP*CxHw@Alztx@o&3KA8t@gT0i(~J(Uf5x+DX6@(Xz%!O<=}!VFB_Sa z(**Van|jNqB+cY1$|>^trwF4OYxHi9S=O!1grSx(4T0Dn@^WhdU9aj(KI!WDEyeIlqdQJZRj>nU3QtNnIN7up z$Gs}*W9f-2ftM4%Gn_zHuA3MU4snfJ0}Vb0t{EZMkE+&}T||@Zk0HsaCO*yIyXBHVpnoGPok zJyC5_TD6+$ByUO`z$9ZRki&AXC+5zqX?Z8|vcz18o0s_lF*-=Tlz4&iQ-@l4wjZL%Z>LXU!7H~IFETbjAyhHj5ToHc&mC8qzL*{K4a zp5?@)?nu|YHQf;TyjZzVCjDWPZ#^qrEKb2F(GE=7K12{tc~N#`@+mm3{Oi4oEOMnJ-5-y1+b8pWbcV&*riM{H=n4p7bC45@#(7=i zFv>TXZ`zMkU5-$J%CYr5DD`@qZv~pE18c@@pZiwPqoz7zi)eE9uz%-U?(h=a=q^dv z_*>ermaXp|sgN$%&umcI!vK*k;9BKy0#E5;0p5+uuN$!wF3q0SYN_?ZTYw73t6Pbo zRzi^#Xb9HhqOrf)cF5aKb>;8mQY*`Y(>!xVA5QE1K?KIa``h?A*i{5B0}7MM;(Qsi-kf#<$kX_Q63g^(%q%pSs@#-b^cQdKAlq=~xRn z8SXaNy1|0zlVQQ+&AE(B1K6sqf^-4`ho%t@xZTQa2h$Dk`<`4Z zaAawGBMMs1#)zXIq~qNRjAK$>1_e6w_@dOyh@* z5k4!CnsI#jt9}y(;Vlul#v^UDdb>_N;Ct84ym*N0wck-1(69(T zvde4wyoLOV%us5wRGocBd1%+Ld*$1-=RRA$nw!b;Vv3>I>UR>&U|xnE}IYA z8%6xg%Ksq}U>6f!nkxrn#x|gL&4Ml@WKH>G@Ao22J!WmOzvC31@vtP1kmT$};2!U( zUZa2g{kV)kTDt~1I|^okuUNnp-czW<#fg%Xs0U52&9&rn%=OdmD=gx$;D#Lp+4Z?8 z?o#pk;MtUSV?XgK+8K`&cX?jm^uHY|SJDp!s z=gI9%X%X~GFtVJ^`TD`Poz2SI0i$pJ14Q}m*Osa08Q0`<8F)%1D0Rg{rmBZq z_+~Pm$}GH3eQs1ZgZ8S9ycpoORH@W1o&7s$CRR1%6TVy`CHAY@6Kp}D%t39bPPoMD zpX+JGmm4R)*u^$zVUxg4`wk>0dC90+G3T7A<%VZH6^N&d^=wyQ) z$ezFIS~N-5PiG;~-I52TB`&|(MO>uGCY&cpZpaTSAp44*xqCq?adG|4?=bF*l`JRY z$|W!Lp5Sli_okChMHiqC8&_VUY5+97@hOg+gioNj^`)S*pC?i36<_P+wS79-MiwgA z4YHd}-I)+Pj%oyY|6NUnzYW6}E@v1epeDI1W36%4j^Ck@^SQM; z)bN{W@Si2E+4UNaNg-&Xlg;g##kP$6VRH@Z#2*JFy6}i=u%Y@3$Lsb9Q4;xPP#uWR zt8IgaxP&!rFsw&Y*YA>rtRK_2z8+0E29`vg5wQLY@x-TB12iKczPlvWzy9ot3)x(p z0GlaLle8)aCcB&PYS`bs9uOg9w06mpj+l2CQh)g2siNg=6W!>Fdi-M64Cz5sz^ql} zOyN=sMz_J?Oa%8zoaj&abKK`HDXSFkF0|D?f7jjc)hA7Yi{>Zcx)D~E(u@VmB%Mzm z=x|!dt^Ac{6XdkasVPt}crs2!oC$`;o*r$WfqW|EQR=f~4spNS2P zjA>}ve@q#f=va;u|AQ8b>o31}EX@fxSa*19zn>cWHo4I^k-l^avTn>uy$oL?-zwBi>e*%=ymrrihb8K!iu~%g*95Eofx0rrGpWIv zK!CZd8k{_bGZ%#(AK6!W>dTUU$>Wm-RO@`8_#S}Uz%w=5xOYb-R_~`N;{s*)VcsOy zvMhyL%5e5crT|PhwkirT{)qh!>q+xDTw0p1j>(-OtX>re#bTB=OPvt~PmyQRzMVin z@A*KVpgFH_i)ei2awvAIRhlkq8=2d--V;USva7}0yiy<>z=|}TDBs6P1m-51B{NzH z*1hvCmy)e0m^+(jK{==?QZ@t7dv?VzOyw6n;-(z2o&!#Xg(==5d6?@vC;+t&vIAB}+$k!hW8ia+Tz z37Zbg%$N4bw(+`#ftMPtCO%gZ@X$P1wdQ!;ukd8=SkcH5s!44c&A9R@1XP+eyyH@{ ztQXnydHw2(%lDmR)6<(-M!S6^XWpcTIEmI07=H?}QF-(ez>n1Q?EP3MC@ za;DsM^t=3c65s8ZV{^+sI52uJ11ZDzXDnbj)fmjK_}!ztM`TMh+J7*L<+G+@_fI4I zmtzz+e5umM51h6uJ?&Pk;XzKUE@(jlf0aPTQu~j6j`OW%;oc?MYodjw+yP0=u3}J) zS(bkM#86%gB0Y@8M(ZYM^9N{DS)ENjhbM1d)E4&H@6lC?*Gk*D!F~K6sRyv?C?@H- zFV~B#ir*#ee*DBqRb?~<#Ox5Gqvi^W&G>}e`YsV*&B5#MPogGJG6(7*H;$s|2=@_zEj`&R|qYVo@{L& zG;$o*L3a7LDH%bdZ`Yo7-o8k$IdxSPsk`ZHUnd@2__=eude=l!``WErb`-e%xwlXL zymj63r*v1ik4E$q1zwqO>Bw@tb*Kn)`dU(+;o%ucQ8F&2}62)A{_K zM7oZb4Ae2>zQVWbJh>N}OvYU4GpDI4nh=PP3O+C}CGCLh*6%D=LF z=ZA?PGvEGx^;1a5xVvGP%6V73&KIv~zY^hriI)3JOMmqPk++jua#uQBqp3RbrTy5K z#FzEN;$}`S1c8*hY=@eB6FPO1VBa8_T9R=)dL;YuJx^ix%~HTOAtWg}U^zYns6SJC zxMoIoKD;&$d@72fgMnRieZKWNHL4u*vJ@EIw9M}+*i|$&yr%Vk^8@W`JN+^{*K(|M z$^1?lBW-f3Qgr#ks>k&h*`eBxZEK1wunNeRlA<&k90knAzJ~y05gbKmbX<<+wPxna zF&D1k5ty5yRUdg=b@h?D&$Hsrmgis2%|`e{J2wXGyc}TMEaLXBy}W-buh%`s$O)>& zt4~&;Ow@5#JeQJkrZ*PeR<*kHNnUBytOxrkHlnv%ajzFx94m%+)(_Rl)w7~Y%SMcY zk_qE!pwpy;%s`#?f8X`AoxZnf{{y+CUXA|r=Pk4`OFCC5RyDvf%@j0WaD_{fjlTKu z4suI`loDvgBNM4BFMWXZpJPRU-`8lGX#I0IWGaJ+_ZvU`7&&Q}Ff(izv|o}#f}I*# zHN76kmd?2{`*Z}fvseBTS+T5+a9`NA$RR1m;4R(08Q7Z1kdAlENYEer8e%D57mn;( z`gabhZeJ;%GX)8_iP=ctf$0X@296xAKj(Kg?uZ-eG6JcQh}8~N&sJ9IU^Jb*7(M4u zf3@Li$6}OvTmrCaF3mLNzsPl8zoVom$sQlaxHo#+y3@veSwrQ-M4OSEJ3C<4NL9@V zG+;Y(Cga{{!5J=%*|HW1t`>xAO(c;dRVgAI-JCaquwMOTJ z_A2NG!+AaYuy5yckU~;{Ry_>)QmPp}8O`B9H-vxUW@w1AD!QZX^2UZ9RU)o7lQnO# zuFiI=W<24K(f@woKdh!mW9y3^m4SO@DS7P+;s*~B#d`}a@;?_A46*|c$i-}5CuAt0 zVHwz7atJf9!E>SA(r654xLhb+UGA&QO)?KXtD~>nd;F2$i2f{jrV^g!FrLucgM#hf z>P9GSXP9<%SDcR&o%%q4yO6TtiU;43f2Y!8-goX)*nTPvMieZ)^Dls9CGg>yjkURR zWkHY-Rh8krGW06&hLIq08l4}7;e5>Hsd(V7V>^?H*Vi+Ks%33jLs#xQBS}4E6N0ip zt-tL2;}B~z^g&6sDQYPPfO11Jmi5h7dGSO5Nk~Tb|K}nj4BV0Bn$FB+xP|K}!Brrc z0h%OE4j-R1jBz9bd;UxbD^}qnu=w+D*9MxYZ9~hZS485t;*Fuv_9sJB4B3rMOT1rg zK~-kSU)B@^r2WvF=Axe4d(YazIXT^zem&K{E!18n7Xbl49W?;PRZc5rWk{0XKoT$< z0LdD&4a##QG{@>K#lQ7Tv!}BsZXhTB@nff*m%0l}@7TtucofZBoOw{{u1}1oziz1G z_nsp0-G^Ll5%;macWc|6aW>0k8 zEe|SnY`OTvM)tS=oEJ_nXL+Rj4SeyeDT`R`^_whoF(ahqA7bp+EKTZhg?p_^D)-X6 zm!tElS1sE@%3MDFJKuPjD&1<`V>qakAKn5hT=L}x2yOddbz#ym;WO!Pm38bw7@RpL zvRrrDK4RTCaz2)gjOxWybOo_$L@@(rWTZiv5S?A2gsI2yEq`4%-jy>8I3v)JEk2D9 z+Xgj(%`A+(WT?mbrSY5wxIAd}C%uDFG;kB=86L3%^@MV$fF(+*VEaBBluM?V9`xes12+vCthJvB33G$KQ~zrzo}0bRGUywNHF^FVRO}3phn@y z9;!9WlLow$jJ8$9*%--e8|V#JD64j@nnrmKDqL+~tXe|23Q~{w~(fgqN@0(&6* zGMiRUpK9xxcP<}yQ_4kMn$~qYT%3HTcD3IHs4HXt6R=CWXNZUDYFgIIaR|6kCw=Yu zj;lO+wq1PR6+?M&Bu=ccnf8&BmXQ%rXn5~vti9S4$^I7mYpGCY8f^1r#2HbN+)*i! zB!f%CHehrg)66aoH}5ShwUGYyZzI*f*1`8W1s8P3xKEcaE)Q}M8B<)=M6QP5L2g$ z>_o{DzuEYa^psfD5Mvu`pqH_V)w!41Z=)TK;xf$7%lXv@am1VGlZ|^ul0bmgF+|5+ zjrsVF=FPb0DwQ2WZdvcVI*CVV*XaR0DpsAu_(@jN>6Y{Uo{S%4oSq65h)m2WpRQFP zP8ZhV^tJbA9tZR|(Byk?#2mX)>Mj?0j_iEvj;XEusuiy?exyzQX5{L+S#yfmPkpSy zZRttSn!rxgX=b8*2=-6C`zvDzcDYIH^(nRtjn4*fruN$|p^dpuhoWjZv!#6y^^ybd zHVuUdoM_4*<}BrPIJNRpmCR(&$h4QPNmuxI^Ts8@@oU3M*2-See{icqv08E z*$MMBr4H|IJO)9hOSrAjBR*MWD^Y{_68A(}qt8Cvm+Glskb(1s1Ts5K443e!JytTk zg(PK8`3&iknt*9YA_9?H*zO@f*l73;0kOb3yK1ZwsUqVLSISc1#8gqwMQcsR;ft#< z&#;P4qGRR5!=Iz}0riR1KSIZEay<0G|Fj7xgMG$WQSuNsp?(Q|G%X zPU@{#-M(hkqPf3R(Os@sBA4K7uOfPAgr5i|{I?tIxPH$`%>PA0hm^eJ+NH-b+eBva z46rb5cK4=h^Dem|s@}K>_VO-!Nh97Zx7ih-brsM+#ON;MJvQKWoqOEKR3mW%KVG(6 zs$%J2M&XmPv(sp70iF7~ThpnkeNu5BGzkuI0Z}P0ltCoi7Qx@VKEn z)cePrAGpc*kiT!8*;1NXG-Mrh>l7(;viEJJuBH96uU$I|UKxhhc^)Zc1J;^Un2#qk zCeVqP^--J3|EKz6%8jwj10n9oZ*mtL@e-9==kg7+23>(GPdm?~=1AvBl~vb|B9{rU7gfq`&-1)r zq;emcvMMmxsh%V~+sjNiRmB3KkFJffwjX=>nQlxGd6t}MC^X%qq5p4%)m%Av#aq&Z zW=!WGN*S7APLNHVZr6R440v%-7)aUdRl088!)LfV1%6CQ4-wGGWu2N$cBtewDM(&h zF)jn~Fx+>mC7yg5${8<;2w`e@{u8;Xzka+v_S7WL{m`;YaGoh!r``<23F0=~;~ z+?k%ObF)TbOf_mOey!lV++&_CUGYk~X{$0+W7==QcR$Rm3C0b@lbW2GZWJeU)T!80 zp6p4vcXgbL;V;cf5<5!{*~}MZ(41erG0WxbiNQNR7VVSMY07y-B5u-AfaB9pjhk=1RdYw`x0O5}mj^QSaa<-i!FXt}$V zyhEchcEaFQW^LZp9ntx%#u(B#y0nXIS=^J?N{r!`1w1o^bq7{i+fwQ?L!@WLa$A5= zxr(G##LugCx{sxA7t=ytIq;lyC)r>ZKlIy9?uWc_#I~9OwUZ%`59!HsWmoR@Ubu6X zJExYhLi)i2@$X?vyO^=Kr&hDw08lqP4*-3A3=sM#0{ivWkCI@|)lWHzNjJ>7oFpE? z{9wTzuiW;=9`j6j1=6IQZEfO4V0URj=8)0k@g)*be6hjm{M(i*@}fiLEeXV+Rf|Tc z0zcS~ce8t{h0#gcn#msmvk9zm+KIS>H>BqBjh`dm#`6l2Rur$bROgr)XC_VIGHU#; zPx2(FrjJEAM0n+_+b|w)_Mn~mTw;GeTr^#SZi05{{__H~md^(BDD$Naq#uRs59Ucv$;%M{l( zm|~VA4-qM!(V0_HGO;}!(S(h41-EY1iNDJe+dsTlSW*D_=$Fw`L<#%3YI*z4g7*yG zn&mg!cjOES$q~%F@su^5xs!9~@Ea10hSW$yP~{lc#04p=KSUcyr1w~xB68<~F)PA^ z$-Cfltq0v8E)Vo^K=-dA!hXf4sQq$r5`+>ntFYR?w=i;2IwV4>GqzcV-Q>GjdKAg7 zwGtlDt)*Cs9@;CfWP!ZL<6GZcQBSj}AHSWIG#5@2l}jKp7na#l%KZ;rM#(-?F3m~$ zm`JXyNxpmZIxHa*go2NdxPb*yUPsyIlM8|~H*;UxNAH(hvTfw-^4aND(+KV1PX7EW zBSeA2hx+@-u{%~2o?(tv5reZvm_?H`(S}436f4b^atBg_k}6oHknMv*TtIfg#8C96 z+sT)ww0S?ts(Ig_+KqPNB$hT_N|!>Ig-|Bx*XFi2p`IS-k6eoi<^dr4%l8t#T+$Eld6+hS5AElD9 zc+EVL{dK9FJf#M!N%W1=LyL(toE4GwgLN_V?ODhglaV?bf}uP%ymP7FH7%iqyZc_v zyo9iK_005}O$DIKZ>_X8JzWv>-5bg0xy<#>Z~J)L{Wo>V)l~5VJi{G-t(+psuxdi) zEWk2pGm2p_^A{$uSitz#tPnd1)V9 zoQWvyY#d2iVcl#(uI{N{6DCJq2Msh(RHp}`P?{~`zl>4SLQ9y=5 z+HMf&+L*4$v%x?>!b#gFE~~G688Z7==9RUIQF4Kx<3$sPb#rT*HpTb_5M=q*IRf?C9QvRWK!~Tm(_Tbq`dqeuGC>WJQD-rTEiCPQ&3lk3^T@2dB;_zLU%j z*6}}-vQ}Ddfh~h}t~VF#I1m}e=Hn6!_nqs|xNvV1`$%xwazpox5t{#(`e4(gw^L~= zL&BVjL8{5-TPkTyzO6qpp=kI&4O>BOOw_6y`TYIxr@Wheeqh9xn#$+BYr{u|9hYaV zO4`th9t?XxUqQ@vHFBZ2a3hIDHt$!sp|@XvgCeo&YLZZWoI zb^V5ZGavA!S-K$-niEwnmKCxRA>_WJf7-hp0JMOa2_rPq)p&i$HEl;p{uB?+OZV*U zS)=x+=7Scfjc<1b)Z@9L?H4wW9O@bD7*6$HqfiqN-fVEgn+9S)o2A7qJ@@8wD)&pg z$sh+4^qk#qaW*+J$YYleJdxpM9sQZgEgQ#6*zP_f`qOOqh{?ieMm&|i(_WS^k%n8R z{DXZS<-N}p;$JU^gFzFScf=0pKHkO50_!OCl;E$EX|^UfvRcQ#(1f^dEs*eJXz9~Z z4ie$>=(42ziR38Ap<~|&sv=o6I!4iR+I#2*)+h7ICmR6!E(0Cy#0{-F2hH!u*sYHz zJx3%Dr8M@8hFyzvvR2DB!5&!{j7(TFZ`zi5MIF?s{`Pt*eD-T^5y6ky7XCwDq||4- zap{eCqovR1`QlSysyjvj)b}dvZ=YwI9n=;x+~VGC%gkmSKWiZd!Nr&E8)v|IFg#ZpJ>Obb>!deP1DK= zQee-Wqv9ZuK3FOejjTiD9!f;>UD9b(5GvO@UX2?o;G6R;yT^VfQ;)qAXS>^`2kf+D;66Sw55WivV=G(=!^6mDxUTr`14J{G}IP1`EVhJf_ z9&GnIf3;hFSB^ceFRjmUPf;K>#F7qb2aQH6t9@=$ z_}KDr2&U?yHheirHrGoJsL@IMzk7Ci#~T1;jLk~(+$qAWR-1lr3+Q-(nopKP(!X1# z;xP|@%@DKI%t4B(r7H?wv*rHe?Sz_*?xKYjpwL0 z4ez03SSU$_!e6L`VF8YVa+S_}_|3%B+^K7s9+pB-GmiUBL++L@%yAT_P2)x2QyHT9 zEK_#KyY-{hwfAZR9xXpsgs}}iU^Ac~a%ZS5OT>kcs9HvqcHo|R(2vBxQ9szyo2bGB zM=xTlC*SxYExY^#pNpi2sn*NMj-IA4pN+IEw?1^Aw$qCG-lH4}x!w)czRJK>>O2z^ zE-qxY9MUn!hP5h3<)fjHZ{)1>Hi7Y~p-MQ+Why4atuOy@5nT}_hj$*`YV!%ciH|{h zx{~Lb7t@_CQA<4SU*h79M(_QJQF^6f_ZcCxW`1WgBi#P&{l<2-KVLCPTH1c#RWr*B zyEbdAkq2A1h6`@NGum;nv+^s6a`O@IQ924j#Z>g_GRe*O2{N9)lpKM0cIw2C`qx99 z+1%NdvXuag!T#u%`gBo#2Ds)6h3j`mQ1{!?`?K70`+719W`4NWFUq^%Ao+&(bihAgJaaT` z`n3atUdp<*Veommq)%s0;ENq;1OJU&lq%!7W6tI*C0_$0uietb<*$_D{UrD*hH6y# zb@`xL<4+)Y?cLHdXfVyCDX$v;*S`3QMv{}t7N|$dw!<}t60u+(V!BJa9VRpG;&*ez z=VFMeS&zb%2GI=+CzpO^Yb0$3mP3k>7imrBpW*u(08anFg3Vi8vbHhDa{UNMVq_%& zo6ZZp5Ggku_f2hk(A9c~xxv>*#f7ydw{26TiTBls;jyLQ*(|gFXK$8m=mvwJE+{PcLyd+<@5(IdB~*HCcQjCg=|R`Sy0xOOh`G5<6!!!1Bawx*(QK|bX|@zdJ- zM?XKE(_5n_v<<_6V@BgtRV{>l*H5vx~g%HJ&T{Xp#TB>VKbFO_3O!Zell zR8T?UEln4817jiRRdPe1YJZu~(0I#OT7)N5f4}CVym{Rx8dbRc)SM4toKzgzI|pX8*;|BrQ}cYMzPz1PHI8<6it=gF3eI{ip4u&`Zl8LZ)R?$K z9i0@(yco@kkJ+I9O^W;(Bk=Oka$0lZc296XRq)sC9|4A1!CatKJ$}*yaqp&A@P1Wb zfodK;k4as(%Tz_Ho5?*!%mYS3wde1~SkV4)Lp|t|Smrgjcf=9T47k;?5us9C!8c z=>ZC=&hbIHU`ML9<0E-cCCecHMWvTzu03Xntl1`X3MpLG|P|Iy)Q{_yK0uPots5*wFc)jR!3fKJB`{Aivp8& z|GxpIonV*L$D-Gul>uWH0PvvxVC>ErslD-F!=H~rjM zB^slpAt&z|K5uzLnw9mF24MmpZ3|=ot$FzQsv#mLzsHSv-Bg#UxX2*rW8g z;r5u!D_HtWXqlY0Z!R^YMZI9ZCkA`osh|x4oEPyKSc7k!{`Pw9B?hrJ){Ot~?!pj^jR8Bu?_KO>ZAup&RMTHIJoSu?n*8`Z?8#xU z<{!IW$zXrBu5-bYW#3aA%sLpTJs=DIRb9kOZBVTnZ3-cdK%{Kkg3(A)?gsmw&U-h3 z{$`}ycnVSu_z!%Yu|=yd)KeSYiUHh25p+RgRka`3z-w1)%Y`~5w5YMXTzUN;e7u`$ z_;pR%M#~&uzYYg%0<`ddV!!W3Jix&pr`f9wGF3bn&wiUT%#xf@qe3&DBW5qr=2M8U z?+=o)9%ct_xmh<)$108P4xE~>ibJMpxI;S|$3#G&lL`-t0z}0wPbScTGF7}LU{lTL z$tlHCoDD4!U_!uQc*o(?AgEo$EK^FC_J}va`au2SN(a zs~$O9FNQ5@O7qZz55#9li|QNE18)wUw<}goRTDiOvF{ZDS-OojLg96|S8ZqLL3R55 z?z^F@$S@97s*+zUs>0Xqtca&tLfYf1Gg9)ADK;@`&Q!6}i!;G7&!u-+aUOl!SGu%5 z*Lh`dYO!<*N+Ua*HA;KG7x~PcJ~3W z7oXGIz2UaX-HKEU5TIO>mjjyn33A`IOG}nW$Kq)9lX5%2%g5WL@L?@~X>9^IT=!_V z2#PnO9+qT%;n>zYK$oYvhs_sscg*L`g~3L86mZmUZX8$MV-#kDtfal)v&>OtS|=zm z3MkqTjaT)nE-x2D@n9K7>Q1SwH0y+ZsUQdwvsO(iLP5}wz>2L`q%frg>(cikN#3_E zY^=h~&E)FqqPC)RVU_Z@_UMIabV9t+Jd#9hi!-awk345F?&%i67 zrp0t=w=;5a9?{!|c7l=j{W}&XM5P|@Ju9od`48C<#`ABY(njUqWx2hoj1?%(9zC@% z=R;1BlY^u!w~=UpCri+v&?6Upy0p`JLWX%)Bj@6Yfza`^{@7SuaaMQ#=Z4K&usW^m zu(@RoX~TMoR7Dvm+pOE%Oeo3SXCgnwu`~q3@>fKGDWiV*pa?ob75CfTIPrVxob821 zNygLZO{<%e>siN@)gLmjAwE*xtNLZ(;?qtN3E3#rs^xMw$UeNdI~-;$Fdzfj@Miwcc)-hIDL{XvdKJSp>q zdQHPl%W@-4XXJ9y0zk^(9rZ;~4&mK0f$QwLx!zK#G&vB_ww=VUh6hs{?^tYD9K;T- zTEhiT2$3H}>Rz$|mS)8;*L!`pO@H?ulI}N-gSHf6eYUIvKGe4lsu?Ahkzt%p;0Tt! z?nS&M$0fFxd-PiOIO?csjyMFKF`nn>^H=18N6IW-|M<9=(5!?DfnL zodCy@+8*DE<JaRyvCa4!Z%#>}rLhmzrEd$^ z((vA597~xOt=Z5=j_>^*i*+DZ9jDZYewUI&22J72O`YPr8}Z$gU%FstkFRGkvf=|3 zTwGpgevhS;e)>fG#kQ^ogwuVN3^NAvB1NU82(y}{Br~JQaF#er4AD~Y(PtOV%>THM z-X30dB8TEVl~f7s%>vzE)Tr73@>{}4L8h{NXRI2TUO(LRZ%myX*Uk5V`qplmu5s4% z<(+AFz7VDRUmnF#15XVrcxoNgCX3ml7?yyGNwj=lCq>>&cKRH+sFWB%XD>tWyX0>g zEChiLep*^|YoGc^a1lWIy=xG(o;ZKf2INU#{y5Y_f?Fz>*MEZ&cIkw81|mK$J!%9o z?HUFTeX4(J4o7|@&AHwR00ov0rt(gNDv> zz4cxZR^2UM-h?iEe#K9FG0sSjHZOt`5FNL~dN@rMx*LyMbl&Bi zxj8u`Wo}{*(e~M|vdUzgZjkyi607S5^r7K%;dCS)o+|&Xd0{ChC`+{G=nQW>ow$AV zin9(rg#UKXkVBkVrm-jx)Jv{N+eJEG&Q>!cjSO$$iz+1 z{MQHe(53lateS`UoiG*{6IDruK_Wq~2ytlyNd`^~+(Mtbj9w!+t5V*)ELTU0JUfeq zWLfzx_&B=c9G-cV!=T;jt2~Ks{;OnaAL63?JT2*h%z#UlaJO#IEE` z)*jFs>xhC^8N!9FgD^6XyK{jT*O=G*W~zgg0!N;gfBGIMl*$bnps~9B7Q%fsP$x%RS zHJA+SmIyq~8IX-EZC;8uN$$Y}{FI&JQq+bM$sX3|Rm&Rwnlrr|P|Uh#3##`x@PxSx zJm~W^GULEpe+s;Nv%Z|N^j8M{opH6b__YBCo)$A2;&hKHMuDAr4UPE?8a`{wH zM-J8&K7Q*qov$EwvG!^+0yU>0z6nI{i8iP)0CS56F$fOn&B!#p#U^exjk~bhS@dk zi4CHnd$813&t8H@GU!+~X&)*kK$QPjYKxxL!+&Odm}_ZB-@KZ8_n;ddtE`sBeHa29 z943kBU1ET_bCSyiTuwgdR zFJ5(D?6%*}okE{!7S0ZiwkPJ;TOoR+6FSP-G77a_S|FkKH2hKKJrl`yb-HS%%{95i zj{L|+kFA`nO zB3s|81!Y-o)s#n$J(suI{ty&E49xHfsG<(@Nm=S}~-P#8rZ}nO9GFfobtGNf`-@hQRY>waT z24&OZe(E)@DF8zz>u45}tojAXPB~p}gQ=V$go@i(GPjGjO8jtrz}Kx|gzEp_LnyGt z&*~ZcgPdVZ`o0de;a=1Hl=k_*k5UJ{bsk&qyB=9-!m{U8WQxrGA6;TpC^uz<@c#ad z{Tp^8+4U<@uF~DNEuyPfU_(96`PYlC-x+S}HMgak^ctSkA@U;NLR2GnPv3>!Gp|h@ z&tu=aJ7q?)hz?M;pX3k}$?2JEy8tq`Z`fXKThSb<@9wXm!Ve$b;+zTxyM!!Xh zEQ652INE;+fuWedb|}A)W$g{bAWYo09}r&^-U<4FJ9vHRS8bXEDo=7q z>Lwd&q|Jz`iL-y*rp^N0Y?i71(p1ZBrcK^|pLp3^=!Zm=?^|csC27zeN@G{IOTrGC zx+Kyi2$5XJo_cx9+&OjS1POw-)<&q1t+~{Q`!oiU6cC1X5Y_P>&`6QX7%zU!-&fwe zP8?Q~a39IBeL=o#8`8GEEC5@15uiy-0~|}N48l>?vl;a2ZmlrG<8+3@%dF43gpy2? z8b5WJjiDmbh4}kem67q z0Zd96ZieG|$?~k4OMrv*hzTtUP&Vy}diIG^TO6%DpOv9&9(+3(px-H_em-K(%@+dE z_z`|XI3Lss@zz$YcR(KXQx7Z6Y652LC7WPs^ygOeUxnmOLSQU>Pv4sA~`;68rm-6pY*N-&A3!w}7%y zh&(CjUj$n-G{gM7#fVqLsiSBIYa8r!I^A$3K&3}(M1dY;a*0<@(ntZt;m1yXm(r$` z#}T6-6pAiTC85`CKc3NJY|~HcMo(v5B`0;*UVi=E%^^jtVIcK`qV4UcF`-xc6CAIc zjDkCS`FPn{QB}auqYWir%2Ltzdi6C9_v@E}0PO7?FZY6TihN>lfC*UUGbOa6q`?#7 zZZB@i2~kgZtq%GE_jC|xf)Rurx`gl}#J$R(4YCWzE=I}sgzFsm6GXDPme7y`1x$fyOw}3)FTqHV-rD?|uF}j?^UK0~gA0TYdK- zq2A%@>HStN$L-0ZiR;?``D{dv{rbX8QS%L6j$_gWKesPsaj9ghty+5zPOWRV{HIqc zCR;Eth>@3}vi0ywyHowSyVD1)Oy7c6tr+T??YQGg9y;!$T)wJ+#-(Wr2y(-H%pgYX z_ejRq%3Q~EZ)cpdW!?{ju?*R$h_UTx0z5U>BD>*Pue$bke%wg0%Yv#UVT=O5%%sBb zLvBkQ9SlNSAVFF(*t8ozUP3;wvEVGQR&AV3F%|WeVQDWB=k^9XWy_}9Wj1Rp0^=0_ zeygnI8P2_;6Hn5~bNTUDEEQR+l@`9Cj1ntn=bhU!y2#3G=5Uq^y(KXdz7h7*>TVlx z#LUSu;rD)3h==?=1+kyqj-In0oN7e(F2#nN5Zbi4+m_H{ftJY}ves%unyW7h&x&31Fr3~!Iq$T5zao(P9Jw&v9&VliDg{L}f4E9T4 z-QMK+C0)hk@tmhRUCCZXt?nbYq56#a5>G#B)787)p%CI|@Av5w?IMK<*NRjsPgi;G zQw4U^eN>S3FvRKF0Ja`frTVDh*=GlkVdnXD>5jXD`H=~=fhXt8PGqY!jyKMghujRb zF1F%-5!@EjOiTS0PSxA5mfD1Hum`a0bAwZIRsv0upM3mFvz*N5$g=^0eVeqWYz^HG z*R36Uz8AZ_7n!y}=pQS(V;wH4J-xn0{FAOMXc?9zBOoE!_1@t+?$GpnStWct=I@JM zTFQzN)-_emH^loze-Qe=j>(S+8U5tCMgTtg!_p^K(-B!KV7?ogt#j-3>6^z9-4I2A zwty**QdZ6ocDct)T706iGh)sVtgWK{O&S!+4;K4rN%uJ! zBk%9~?r|R&vmSIgzhl$=Uta6rbfy~f4Z@Lc?_b#B&Ajp$UkBM$S3I9Ql0sk-L2B*$ zC4!L*U30@$5NWJ8+SVDCH$;cAB>;~O(-1m(ecfsXUu@mjY3I`-9mHUn9bd2P?{|dm zNu3X;qUV?f*#Z6+H(r@CY6FZ_Co7?3_CxG=X=*3M|Vf~88(YjlHw zCq(XLN2dcpdBDAwLfkmS#S@qn{Xf7<0PeSB8B)WOSUpB^`ztAXFK2@0@b4ssFg{x+ zsQVTOioxb6E=|dg)oqeL1w?Cpa_WJ&{|%hcfnx99${imdUb225zRkUbLLxCr?|)yt zAGEiaUM$fr`nJ9CC4kP~t>csRIWVK>ij37})FR=HYNw{33`OypyS*h%bf05iVZIPwwuLJ;AMBOtaF&G_iKLO`bo(mJP3cII{pzva>*Lt@u z3OrlTC5EhFtAA3b`Mc@nMQ!bD#f-1-%=U)>xL%vE^p>vM;@hx*{r7zCCI3X`j)eaQ zTC4G2m9}PlB9^=tU7af6f~e2)fLK|Z2Js(QIqrK74t_>I8m0W7w4~~*vZs)L{&z>Vdgt7+5-X#p4VhY@tJZ@*C=PlU zE(Kg~rLQLUWbZOrYj-~9?f^hpj6iUgofekV*KnH`$rE$+L5%s=iHM0Z&gO`*X0^9$ zDSh}miOS1y9=f=@GQz5+t*%n$NApXcFM9iUb?mQuTxVTkOZ~j`ny8dRRlEgr^c`P* zD5!r>LPWH5d%Pd{aa+!xZcv6$bKD2I@jJ6GG`@K?qiO9d*~jrfcp;HRE=xO2WC|Or z4dX3HI=P_4<)(|Xs{C?t6=gkBdY82E`s&`qFKvxHV0uxCSgJXeD{{nS}c`p36qXXX#vRf zY~t~epat^R;bSd))l7xy^xcDpwS`?$0(b7pUcCsbN}%fhKVC0JxGVpSk7A|x0yMJr z;fS4!k9j~1^ls-nw)k%E?C&BoeDPKU4%_qcLLz!1XDjUDGWzq2e%zc4Ew5juX-$yi z`&QEulR`fg-JH8N+0Q^RG*IL%=y*n9KR8@UXB*u@L&_4oB_wxKu`>Qa>BPyqxbOi!?8<#~KK3!jYamKf( z`PFHpPI*BIoMNBLr#9mNjqSjF^j=Y5Uj|VFoK-SSS&|BRT7>)R>~DZ{P!$C?OHLFK zllGp<(e_3i4L5@peD+I3kIV1dSb0fg9f)WT&?Bcpb;`4C3W&gRiqgmlMs2QJp^xIsX;9+_n&ERi zkXkJV)o83sbFQM#oQqbJ&Y^4AQd5{hxSSoi)0Z=Ko&2+NWbNK;(8Kr5$=+%1ekk1{ z4qUc>JNP*8?csua?*>nq9QBq%!?J9VTQq~xrDfirlhxY`v62}+Jw~M6dpWD$2a{ju0 z^{!SQHQ;Qff*5-L8R~xZU6oZ%pF3STr;aOe*P{l5be#ka81@QeN*~i|x=QrRKTupv6j`jydR9 z)oCoXFa!HMoc)9070H1Bvo9;S`4Wl-I?Az{PRi|kpqaf;S&+0Cg9YcbuX>D1y~#+e zs=4ANN(8)*3>zApH0muGkO`Jvum>a|#}z#~xvvEU1|QeKJfZ%KV04usAMU5A=CUZP zSKjKQZ-l6`ham;#XY+!KMPY8>2hwsju^)g(0jyNjPw5!wp98N zOyp%lPC5%*w=qyzbk=J|56R8OW$pNFmFE0nZPT1sz&s!MUm-%dJ;&I(Bf$y{z7tFs zhHg`Ja;q^#u|5>xL*|^HG;T!wGjV|;yrS^nk=Su%imHJMssmiZk&o(8vCm`I`R3-; zD+d2V=ExCx93WZwV;UAATYCGFOxXWbbnfv?zW*N|lOd-v+1MN+$C6`hm_t5JbDm>J zLXMG{<513JF~>@4&YMF{DMTm|o62brJ z{e0C1beSP!nQVoz%7s95ZTL{U!ZOqkgxI`zy;k+{X){zvEp*-Q8LNF)`#ZrPTDW1H ziDa-X{k>~yYUuvfkGN~XwKtq!d{t@|r$sysJY_SJjm&>}=FI?n!G&MfPD4XY<%hIM zD#0Po5^X@@)TCds*B7H=Z*nE=C|;YeqpQfMEY<#=&)e;5{F^ReI($#BYt_AA@gs48 zl7>IL(JE3S-y^3WIetNNh47|fX+sfpoPWOyM@Y7!4SMS8XtZ_+@GrLs(L1W(wjTxI z0wyw=D@iI`*A}816($}Uke;0=h9y@S?xeS9SX9`nH=mV1dA?YfpSP{yRAjH_+AO8@ zc~nrv(|g;xr(-r%nBHcQ&du;Mxt>Rp8?CB8>a|BSPPCnckoT+AwW$#p&YM~P($JjJ zEkJaBCU4{r(mIxPpRA=P*-_hl0*(o86m@}JYu8S(#eRZ2)0So8mVmtWYj$z9<%64D zFovn;vlDAtKY8`?-Jec(huuYds4{;N@@~I7Kp<*)nEQJums;AMYr01AQ9+;W0YU;T zI};7&c(k8i05xYPOhwmOsa1!j4md3>);5GRIs25VrvJVTtGlw{N~fStGQ)e&O)=~4 zHddUA4mIy~^tA;qE1c!Jf7OOox8_1tu_FwFwkSI%J6&>%>I~z(94@-!hr>TIt7S#J z@1`=SDSK1ZrG%VecVF@O*u|%gb$*c=5$FXxE)$TbU)3M!g)M)fdnH@JN3p+V3k;W9L2>hB&v|taQ?4(mKjh6(OHy{+q=0R@^B`XnPrFWa_JPAES zcK$^L(fDTHqW$?qyR*V8lX#$24HE?GV^bX0zOvfT*I)ynt34bb^ozR|QASCb?y5Pi zBtwsCNp%qW(%EPrVu0ABa4JzCkmiPWfcCs$y&{V{C6ymdM`q1Q7vZT(d8L|O7M-W) zj3X_e~WHa4O|Vo~VdReSi-AY~wFPBQDB8aWu7lUxggKFJF{ zbF#D3(w7P0pQx>o2l>Iu{i6*i!AxXab^bjQvrKwe)_Y5@dp`Q}8K@y}_E^QEKx1y zTDMuStSqR%X3|xn`X;s|Sd+Y@1U}+*Q*WK}+~8bp^R-i3->V8+05!F2uu~+L1d<#6c&CIZFNh;o?Zsw}_8ec5)w1#T`zroE?e2rOr-Efv-aUjZRmv}p z#Q$>%KhV4$s}Lw?B2THdLjg~+>h7>}WV%#+e%V7hDh05RY}3&=2<&8wq=!%+G&yAl z0Or{RYo`6~thVmy%j)IXGL`1l8*`J@)bU}iucP#I}?^-=vu&enO4JI-1fu=iEUXRI_uw-P0GzWjHxU z_>>Kp^rg$Nz8<%XG7w_x789FQVG@!Z)F^olmRALB;J}5f9jkAJC@dvw52&n7aIjih zM9G*FY@qr7#^gC?;O5q56(SDCuRI>H=gz+cY%eg$<;hS1H~+jJL&E`U`TH}NN&i6N zF^|B<;;x)nPb@8P!RSqE4K2_5zr^~Jr}ggCr?gqY9z8vBynS%Kd5`kBQTGq?C$EA+ zl91m*=~9(wqvO-KGeXTmo)6j*q6cqWxr__WS`7ZFWPg9V`C%eI?EY(4HiCCJ>13lE-Y5PoL~<(&&49j)68xug+#-a(@VgS|CYFyU zW&jB!N1*yvgI6U)xw#g6EAAb6U>VIL`Pc%64Elr}%APZDoD02VcmAX%HBdHEjnTFN z`NApGwvaOL`$SJNbkOL7H#C}>wRf3|z!d)A7ERU>_Js?XmLo_@`Kd}qBK58m&CzGN&3@l7=9@)HC3(_|_TtzxEh~h%7>w#b zPwR){eDgT>UG5$=?A{gmqR>oV#u$F@Lwr>nFN@(>xQu&)e zS2GN=+41Lc{~vwPrt#F?NK5fq56(~H3LHQ{{FFP-ac!L*Azb%8YK>}S3y?H0kES$; zJ`-+Daf~)~{Fghkqs}O4BHlvEoH$`ZUl0yH>S*FMS7;swc(_I^u;vC)E6Pan(fO-i z3(H4Pi=oRG-CMJ*mUY*SB+j1uc>3t!HLcLsBe7znR~}D2JcYr_hx#t{4uUJ_!&SaP zUdwt45grIZ1U!ATF4lGQ=z9y}?ni8@OEO^okj|`32SDdBy2|+OZ2VufSNE1bJ3S#8 zd-@)BR;!8H^A;|x>JgpB?r}a%)%aQSYabIuWS54B_rVUc>Xv*gPsMH5N}S0SJT~Al z{kDnS@&mO?^AbxNOj5b@#le1rtuxjTFIp6(hT}{Q-fA_?s@IPoHOqZ3k!nvg(m?X zAjqw8Z2wq|&VyHT*%MjtwE=C=u`el}wKXVY9gqxFT2vyUx7cuZ^(+5e}Mh!)O!c@;2>1b#wjqe?rfn7FI($J9<2H zs8D_KMBz#MX_G^Or)8cf%dnAs| zkmUuXSQQYr86LqrrF^rL$IA?~F%Kb)!!hoURbW%J2dy^j2F_x>s(`$*M4jxe&Tz5(pB{qkhzR0<6f@>E?ok+s1|1_gW~_~vxdzVM z@HG;1X*%1^4c&B+4EIFb4Sgla?AUrl_NWYHPDT&O_gt|}l2ZU|_%7Pa(Gmt_HS_qI z6SI+8%Yk0=TIGq5I{9R&J4V5OZc_ank&>H7psZ!^^+{tAqp>5u5$m7>#G;)dd-|0A{s!7QANQtdFf@KG*XveZS zjlyTx7JQ8kWU2qA{SiwUoBU%Lt8>uiJ#dYXO`S{y5>UrlCmwWO@u>Gtfu7TI71wmU z^F%*l{}T2&bGf25*5_Q|aM;YnJd>l3=NtEr{Z|#V-xJ~@l0sRiyG_6}sC!DDy=3|B zKbh`GeOAVt=pa2%DsK2>%cmjn(HPu9#z}sI4M7shy@Iue!IE*4fiXa_GCQC8$SL7T_j90FVRiCRTT1p% zeV6K&*<6{3vFm)f~XO619`qWAb zdKNn)-uN60*>^or*69;eX_Ub*j|Heu=Bmjz_Bqd#CTf-dZ~!uLA@usHAVGQI!Ol6`XH$C zV2SN_@Xn!SdQ{w?NbM-=u}-2XBDaXw*!Y3?*U8_$i4IdbeE^v-RF;U0%_Yo z%rsaH-ftS$?I8rN9V=aCV#`2t)^Ji47N&VxNh>q_XdvP%)81rn zd>l_5J4RYjHs#^ThNh^WM6y8u5xsi#+#Ng_nQ7 zg%lH$GQzb@d-VZ+GCVsqY3<&m=&6qk^!g>A{SKM)Q=N?-XVn*9^n^A;)Ub>!s#Eyg zcjLbb^KpX=;jw1alTcHEdZ)$9C#L_VU4DM&y3`ZxmG}k_?~m>e z<%!NuHRAGGV1n8^%|=RvDO1l|G=Dec1b^4a_K&H0%m@*AH$AFT*SFxN))^~I9qUlb zDgQ&cz&FRr4qF=yx$v0#Tz(44wqqr(h6l~rsgI#Vzu&HqO)w0sLkH^G(3D z^Uca&_`j%mmCrq#k8(=0`hT3rd~2gDiqx01iQHwISYyn}0joeD*TY7-bF@t}NDg8k z>9_+uR!T6zi8RVJg0q*Xn7XQ2Cx+r3E-68#%Ijla8^Ie6{TRTY?U7r2-Fy36b`-#{ zB-hZvREL4`8lhY)EEd`T81V7KI)ggdT;$ecj2?Z^&Y?>8@=V4Rt#iF|li}rL5GUKY zag?t4xC*0M)zKO_EOV^!wfE@`7Z{W0#)Xc>}a^V1R{(n;VEv&KU)ha9Vn-yW8d0jM4M=SG6s* zx9LM8=N%L?*w9zy3zN6f$gNXrr1L`gU`2Bgan~?Il|i_LZi4h=MkI0UL}v=&+a0Ut zm8l6~Ox?7kxuYtRDU#v;1Nr+MxFrIH3gnTKcSjYrF*7UQNnGmuc1eBfMVkPT!LnJbShDnUc|_~nRm6L=1jF1@@( z8`L8hYw``Pk!%Vg@T3R`nhh1z&wZ0DdLC~#!i%SbV^k7wM{3KL3Y^-{F}!eTgGaRH z<`oYdc8``yZN{v7AE0j86hPSC?!1=~OI>AI6Y-RvJlkHRc`+#~^st{_rKEW;Bit-B zSgm)bfiUU1bdSd6wRf)w&bitW{^XqV5u0TSU`GJi_o!X_^_VcMD~$e>7dmM8M)lnV zVItSIuJXP!HKw9*|0i)1RhhmA7a+u0^4v4 z6P#*r7%jEGjKwEXC1KxTumcj*-(Rv|(tXaH~Sa&jOgRc24 zvXMfHUy2QgPA~k(#oKh}Q~NSlbzP_&^Fj^ui%T3rfKNY1b_n6`B-av~(>cVXUF355 zV1$JU{sNjQvcIDhh>|*rD}kr>QFWkEg~~J zak6_Q`WP&`y-s3IRu8GQ3QRtDFG?=i@><^>pvxCtLd6lV%lQxHp(@k0)29>0cdE-8 zl{p{i#H^si?2^v|^LdhZ6dV~Wthca~Jbqi4$Vvm=|16V&h54bc?RQh5Tzxy$*A)ul zM+r7;gk*%QVA?**(F{Gb!!^59A}JApNpQM(3o%D&3#y<&fS?k6d8`AjK=*W5s^Zez zUo}@P?q4Eysz?3W<^^=zj*LRHs2t}CtC#E2!wmy@+g5ZQ$>-UQYvH!+&dEI%%6b{@TI3!p=s^DR1i zYstKP;yU(izid^m1>4_$;5FUe;sE9MxGI(9qr^zWzPDm0BX$U3<){OGnCX7RYz3{nQ6BtgIf4mrWvA?*9`E zp!hCY;|u~3lFrL5hbmfqnVxiXbq;}y!P~v+4y#k83K!9KkgE^6_T#U#I-jlMS($vV zqAfw9BR>+&1v>qDA31$~UBV|b^0X=1NYhg2Vl!SLJ?CplH15^2&c%5`+`!C{PxZex zS_vFyk3e5%iqA+Aq!BFhm$;sz+$Uq?DUNGsnx-94?6Deph^1`7U!hK=>m+lt^G4q% zk^%iT7{)Qamy&9a#qqDRin!ff))Ew>m3Wq?DqWzl=ji4xmnw0`I@!F9KoSmBG&KCH zowIS=5SwzdD6cl0=BCi%HZDD2P(dF)=hE6@S3?i72ov46btw3O`lP~W13T~}^NmJ^ z@(FsjiJlWr=N#{qV`f+Z8>=*&xN*#QyRg^rlLt0=#9N``sB1~RjKsUVCeSx-2` zMD=ta3e^C4iRr{Jb&q~!_BUakdYnE@?;P0*(eUQH_+Y(e6S_EZ)L+EZYt3s&QP_TY zAW9lKw)CO-y$Bx~)mW8w$xiaY=jVcF8==|lUvfJ~|D^fGs4EC2@q#1=PHzX!?(crgIr*G7VPrUv4?ThY=@Ww_s zbAZ?gJlg$JaF~wm`m2%Lh`Za0&sDyClW6I4(riATQ0i8bsF@oR2-oenfplMY7^`F0 zzUtabW`b+lY@DWD^depi8hJDvl-?-ie&W(^Dw*;RXw0wb`iZ2CYuzsLj{t%91vV_F z?P2VKBeL=cVdVu9QTAZKWj7c4{-31HZSRk}NulZQ6LlYPKHZVm1OM)AuAR}nwrH{F zd|f9Yl-@tce<1Jzdp!y@|ETffuWXN`^>^tO*X7I1oUHNHAt&k|_(Vznt>>&Hp1peX z>Pa{zZop-#`^;5>pDLf|BA!|6d?hLCo}$fuhPpy-a_EJTExf?q$+G!Na@!@Yo-r_i zX&EfjRpoamrCnNII#UXc6VhLaID9G37J`0ua>DW(jwVnIB3FmmV zo`t(_6Zi({x{fDbUJwci{$1*%D)x)=7NCY(J`By=-fg1fIeC-@oT#Te>O@z(zXc*p zh-+M`%g@fmjVy8mgrnE9ld77Dwz#P zO<2%^o&+N0OG|5;PuKmr#@0S~-)C?0 zQ(pq77x7qlc0&4;|osM!4k?^E&*K z-PXr)yGBi2BjJxjB#2=!cBw(ZpG{XDV#_zv zrQDk$w3G!QUQ5Y^#&hJdMjfErmO8O}4>932=LeN*r3kj8|9o6G6--&9gyg519I6oV zN2QD4>mA9zOu!fd_U%@9xRxGCpq=+}nr(~^-zvn^#?DF@Ox7ehaAE#tXoyDIq3Or{ zljAGWa*Ho$_xj9Gd{(7*zU3P`{h6?PccL?^zMY)|O_^51m{{ zkbcSPIrM45PrAK(DYSGfuKL57Y^ob;bK&)ia^b%!j;;^Ut3CG6y|>qF7u$Lf$tKMw z7g|QtzWh7u$vchUpRCiQ36?^p@=g7?8$CXHd1H_BJ(%9* z?h;b^)?Qbu^&oHXR)OX)K6I>o{=w@?Z3I`xAJ?gpZ%Ou!jF0ONuU{JB`*(L;T6!<| zN)0sL;l!#nX+5ma(9`!bsK3ph$OW7Wle|1jsnrhbZg{b-G z;#qD|=9mm@xgS}z$R_^bN%9QJLA6{!)r+eagezPnSd+Tswk4D5Xo76$KU<)imPpHb zp8}TNp7mLxYEKoyF)HVMW{vG^5deaxIYdu+S?3wDL3sc|M{)u3h;g+Lf2ln*5{+7B zrSj{{BJzTsh!63>Av&=Oq#(dsU*IaEEdicN84ON12%y~Pd?cDK30kJ;vC3*Z%ycKU z3(Qe8Hd2{gLxk`gIEnvIH`gSEWXI4)l^`j0;OxoA9{G$X&DMuFPw_K2 z%SOm-YG-==ox?Ne@uv8+QF&HRGh4UrSHr@Oca{f%;%q?;2tvlxW4__riA(zKJEQ5U zg1_Sa`H+L{;q!%P6x~3~(icQa`*J^f0c`cPQK@1W zq^`$8^^F<Ho+edBWk6y0=dP&@vg!t+9OzmJ>VT!H;ogB^+3Xrkp0-)Xvq zHBIY+HO0u<+}UY*Xy+i>9zVM1<<9sqz5DC-mh-hv#fWleQZ^Tx;s>&iBz!56l0c2# zpaC7n9#$<5cNd?FI9a;4oLQSGaO;Dir86I}v@GD211p(WEduNI7Wrk#Ig1hP3&*we z{;EEQuau@`Ei2q`Il1+znXUxMd7W^~Sl>&=Th&%Hu@~VF@B^h}4-R7JLm~{*nY_Se z=qo19ihW2Sva{Jn9@#(n9I3zQAVF@(RNkd)SVN~@rF=B5_ML+X6J{TbzCy6{T?g|- zRX-K|2+@(YvjipJ&RKx)YsMSZsYDs?)*7(Ejev)#yj+LMT+jUKelOM9xguU4>0P4m zuj=_01;UjebOcu_dS|En{%ip2HDZfHDwGe`9A)LHjDA^el)5cse1^B(L?E8-V110K z^R*bz<8{Krxr{CoV~%oa=Xk%LK$oZGa1M$VC`i1&!L=cxRbixS^Q@zvm(+zfQzhsu?Dr~-6V~u-4`5rQerJTe?QL* zHcu`dj+TE5UK_*tZ&{NyOI1)DiYrKZTYQ_g&Z5?|$qDt<2*1jno4*UBNIL{6d*7r7 zP51Srz!^7GvoC%vh}*bystB9kBw-NHi5FY6XukXHZ^Op8n1Zb090YnCnzS)O^Sg_d zmBEfYFZ$6t-#uNUn53|(YJCA8b97zdA8$kLY3=c(geluc5`}n-h{p#U2+(;;ENfR% zGu&`zTKM5S(zd#3;MCyP;Ri9ltIFYr_1abC;jgllM{Oi4Zt_Q+Hpe_#!A0UX97nn; z*YY_WH%Ne=w<=9CP)+s9?u21m_%E#|`6(_agl|ur{_8#Je`v+A7>;#ewu50SIN7A} zSY}^IPwm>5!c>jm1RuXOV=tSq1&Y_E71pq$LePm^nYTTkim?TACwBw_Bt0sN@TpOq zAElG-)!M0fPCIw7eZI@wq;Lci`e47v>0TfbMp~}FXEhg zKcJD{YvXm5KFu~ODR!0~x2Q+VANQyVw=QyF$eeL4bTIqVJlf`KrO&iS)mopAQJ89Z zMek&|p-RVf7bf{Ifc0$>$VhV-vDYct^fXU*>~qqn)9uvvvSe)BXdTtMaAQZieCJ@C z;~7aGV@C)4_Wi*eTC;E}P)7G|SV(tIC>i%D=0(*I$W#`Hubj&Y{;w6mUc*W>8{oD+ z=C<)8f;XXFhRB|a{SN~86Iud|7VXbnlF!S%;_!r7v&17qJkYlMd5Y#5 zBA|@gZjxfW5VS^7mUsmEsWHts9hX+a7?&SAsFy-mmsjK2CLu;_E*>DFI}+f|>*e0FB>fuuye1LR6}Zqe`N zM+&kkqcy~zqjydTzR}zY)NS>pXgS^2ax2 z@tuSnbUjXc*YU-zDbcU}Mn_MU4VF&GGTs^K?bl=X(jHy(ODy$~`HI$k{6CPm`Y5lc z#T2F1n(p`Yi@293w;TSHNv<6)5E=qckB=Dn z1F=>g8pa4(|DqlCs)ys0B9MTwQ|F8P@uGMgq;qIZ zjv{kvQF9ua+$lymz?d z`s{D*|0-<%yY&KZU6zPbw{ z&^NDXv?OmgH1>AGfFA_TXe3B{AM_i+)YESG_8MWAt~+EsjCK{9*BGAdJmwXMpa^Sd zQfz*zI3uJ(C#Lu5k8e>-CXltVrQ6BGF}Z|j`bYvUW}U> zM7C#Gbxv(W$#kL?HAvVCW-xcb67D&(Tq4m<-WwpoZ8@Dt(bA zPNYD&vZGH)9du9?Ut(r{WW*Nc}@%&l&lz?BeD)m-@>*6h&& zCP()q&ETMkGp)?oBU}jy4kANTVdeXHqy9Ysd0cOD~(Z=r25JpmNXIc~zt3+2i%T!d#;^pKr5n5lQLlgf2=_g#fC$pz_FRNNqwd0+ZE#vi1n;Du&$Xp9Jv;xoi?8ta4x+`WzxfkQ z_eOQf{deZ`q~pO~NjKj**k-b+z8RPtR6fw!Z1|SiKO2zB zAxOOSFc@t#l|D4Asc?iaHeUzCaGzA_?|wsTtb_w03+iYE{kej z(hHBD!S6esNtxX*yLA8El)C53ViW0|dyBXAZ;cUMKD!(ayDqu-SWrIeZ#i2DzAlJw zA~fj>-!Z_atzELcEp!dQ+&B4(JpkIHno7sHUkU3)HwgyR{{9u$ra10j`m&-r-DM_z z8C~JE^=tEcutFp+7iaoI*+SBpALg&Qg|$HQrIBv2e&i#%CWL&u03K0cP_^)f$&5ie z2sSbFS1Hx`>Wgr+bemS~(UQ*UT_x+=4EckJq0`4|^=Hg9(8+?66F_lZy=f~EW zFeM-}6@D;5rOuTm)uk_N?oB8%UBY`Z)I|FsnH9HXMO8HKeV%|q3nU_2`V?g7=%RIS zZd33BdFrMmmwE!}v{Hq-Kr{9NKVa+nyj+d0Bg1Y$Z=GnaIO?;z%zBV&3G$_#88jXZ z2_6Lj$z)@TwWJP|>|Y_!?z*Y;NAwO|9$bw#E-o<|%qvYY)204re*v>f`@jT*a^5D5 z>wc1L6-PECj;iwws#Kc+KK&26`R8ag00{b#6w4mG)5y*&M5zW8RaKs+T%qnrFHyq% zq{~v8;+v}dvH>EiLeVoD%J`Lpr#VZUm(fy#`xk^GG4XLeLN*@OAiDXswHjQ!)MHwzZ3Zs~spKJ5&J(%p zn2LqqspO}E2%C(8Xpp{5QAUJTtC0Tj&}mimgsg_6TkrM&efi_ zX1(+nywnilv^=~1ihb(V4bJ#btn|C=zzNpVEbX@ic9?pDk@ak8HIKeX=+-bnFND;& z>h!wXi=}fWY}b>Fu4~y6#aqAmsl_GT3gMLE*S*fP|40&2#cDp*K=`HX&<34-OEuF< z)0d&?KEfY$&+9C9OhKxDgzz)nN5&$Ee?$6tkrwkIaKFF{+9~TU5_NaoI|ing)jbni z-?zpL0&d7jPa)PX$W&8VBjk>&#^NEE9W=EAbT8I@|k%{xUk-_|< z*dqbTM>^LniJR8B_hY_b2(}2}A}PjGD-O8+1=}Nj|LhJ-FNp=o`(Yv;b4o?v}8xu*p}iV)*GBx3XL1(WaY%MxB7e)71!v)RjkI!XG` zs3Ya6`KJ|(9QB}w&L{Art>i(}WgW_WQbSnS$L>X7^D>o<8<){MuY&zPlW^O|>$rwq za%kG-R9*0y^sLTE=j;z8UjAfPMelMM_M4%!U2fHDt|LW)AEG~6_~q`vDDGWH$`{SUgbGp~I+HFDg)j}$(aH&W&y3#n_&vi$QT+R(hqa5m>s!%tjSOIx+M zr6BRZwMIT0kY>J1T(ce&x@=ILkTTUxVlCzG#;JObkXo=8d+%RaknWxer;CDBQPp40 zy=~YXKJ{+NeV;vnT&pZO*D$^~zgD+(!Tct2i=?;pC9MmiJjB3miy0=}j^D7!`x*C5y?^g}JPB$%;NprEJ_|!1`~u%C|e8uhD_< zy*rF-(2!%mYu5KQOOrV7xO^>mCL~I0zBWq8<$zw;p=+S3HGm!lius%z+@M|Xkw?;Z zQIgMOD+pjyK9?BIPR6pRS0XhzBvX>#1#$Qf zTTC(v^LIctU=*O2@%Zsjkc_MNoL6&P#&B;DSRG8LO4XvHI!}PDU?F|bGyB2Rq`m6f$67wLoi#DNQIxz$ z^extJ%ox<8*nM203J0T?X!Retl%QWMuz17teZuU-_RV1xR3SJF%2YwTtKrmsKS2yu<>_=e?=s;~YU}VJZ8!oqDXXhU*CVm{M zp{F+)_~f$+tAnSdi*9lDaYE@lybHL(0=qcc0)k|-X)4);q~6?FS-#H71HudZQW(S& zbKfH`i>?ndGSrKNt7NeIB!@v)1p@0u<7NndqJCWQ$Ret_i0-T4>nzEk^l)=}E^c4$ zH~Swb>v(dN0#{kD8zC+CUk}vvIHcoI#4IaYQn^;bWE#bQzK8jZ5|wGnLI;q!HsBwmzs(h5><(+nyr*F`lE6XRTLmn?oBSx;RC z23EgLVT4IK+&wp=JBV6Xx_g_Oq~@1Z(Se@Z#K!43+N2Ns*6A2RRgysVukxsfaWk-` zc4^SR)8l5`KJheA07;n#4v5NJqo!t1@Fml_-Ubn9*;y=g=T%`Y9qQbZs$X{2D_QYc zHg#v`(cuY>nRwQRvC=UKYQlC?iSuAu>8zt!{PghmBt`q5Ps|M1T~2U~1?iy5j@0vkMq15gJ=?T+4gniILoCqs#gI2F{0rocJvauaSq#kzA#+5asIrC_z}_KKg)FuYFEmQkkYuFpXpE z#X&#QpF}EGrrz=aiRdC&Jbl+2p-!Z(lhvdZW>=5w!dax*RV!_#eMKc{C+O8VLn9fF z$cMK1)H&gu(++UUN)oH;$=E`FQ}9K~y8D@nsvv#WO{L(ZC(^~hS9Yf@^c$^%fh7VB?@Y7%repQU8t?@H3x z(kfy;IjOd?5cZw?DyIVuTnc~*F?`oR)C}q*Ix);6Ho~i1nJgY56wZmne+>qPICm=H zR{cGzN12^N(xt7hItKX;#?9jS{-D$v&A{Yt9HNeA?!eJZ%fIn)_UD5?4mYl{Q_P?P z2|pC#geFjsw(}7Nzmc_-#qH+dmD4<2DWum5<7Pzn>*@QaXTsh~+J~61NNNqg+^N6{ z3mhWP*$XIIcr{7c^a#U4%YKTa!WWO@O$x3i0W{3b0qIsp}4LApL6lUvl|{^-bq1~)#*<6s;4=qzF);*+D{k6c4~AN znNo?%pEwldW?z+46`Q!iqLLw>$);~cUH{+Z%w1PAb@qqW6T1OCa~SHP$h^wCzM5pK z^E42>5LHRCUA6uEomkd?`FK*9DAQeVw>k9LORF>8?5-%L4;Zd40uMPhDSRIHiLYbjTD4xbY(MbVZ?$~65np~hi2grZyy zi!R^!|0zl&6q-Olg;ONut8#H_y+c(7|34R{CO`h%eR?jZ&O%E+43bazeMy_4=@oqC zO^Z)wKbWFigJdYs+6)Q@i5t6Lo(Kq<-^!j8ZJI6uPeu>3wy4SP#ad(BLIr6-qi>ovJ_AnkR9T~dy_EVqf@Hq_DE z;GbgBpwyo7;CqK~rjl;a#ViI^x!FRb;n=T_xm_;PHh=-zkwX|e634=hlJ&OMt<#?Il9Z1M8IXf56aq*p++woFHLbhA@{Ldz z*;RpWgKlNS8yd=$mFkPm<2gTqO?qX4Gf;(U7h2VKvK-<7D}Z4}HCUeHwLmZvSB?}z zp=xf5=s22ieT4?{u$!^U)ya0PEDhjc(#vbf<3Ipz$a2j@d!_r|{(NfBReSv%dZD{aJv(6|>7`01gfS zfa72R_9p?R08WlW9XjWsxPXTa;^G2wfk9yKUmL=G1PtMZfWb$2k8txG>fpr3%fokg zIc)OxS56QR2;zZ&A%B(pPqyC%fF1$B0ro%+SpX-L0|@2V?*hOMa)SPfcM$e(ac}~; zKw!v04BmqRyud+xAn+hG=pgYy_2Ywg02dS_D09vjEM$ELA{!vA9+CRuh@44vyU2}U zy1a&KU?ew>sF=8fq{1mhCFRqaTG~3gditgpFPS0DFJG~-wfoKfrh}uK``vpUo?hPf zA3S^%6dV#7_2g-EOl(|yT6#uiR`#<$b6ys{Dk?5{U0POCTZgZ2c-Khi==y>rOL0RaAub>RPz3wprCc{l^eAr}W{ z(7^^mxj-`Kz=FoskUIfFvg#2>giTUkRJU`>Y22WTxCRdMh{|iuDKHMv{v!MDz#{)I zvVVd7n~MzK19BV;4+sSy0DJC{eQfIm)#tuMC)+}ELI#$Y>a%6nTb0DJ-v@xqKh*a2GDaIrC%f{>NkrQH@ zmCr`Yb=JzC5&nT((xX=vojkLbd6qbBWDUT^%scE5x5Q#n7f5L15(h5bI}b}cVsw`y zCm`fj-St+#D$t7LS(2(DHv`u-=EXn%8u4`TD%x%RPro(m3+w3-KP!d0VRS21A(MQ_ zDB>;xT0U)Zw!Suki8J7z zUrsGjw9_siHhQT2xZPRFUcz2+3-2xMe&0vzrKoUK&5&`}i@Vk?&l<3F>!F0~f|3{v zF!)BBQ2XcR>tPRUrKzU=0`q%wn{STeCd0NWzNn#}!)JhA7dAwPI%=I-j>&;3t5miK zi4S)+{}ao8`%gnVh0l9e5O`wv8^4!1#CG!Dc(NUK8p*{Gelj)t{_?WRam-jiu!reh zTJzO{;C%pP?EM_}*pOF|mk`admShK2bn_UDZV#9sJAo;$DL!*v7Ws^_hOFmuSyYT-9~(1*5g(Q; zkv!}OgyhY&qlD&mZ6~u~nK;vr!spziMTGe(=NDQVaH~~SXeB21tV?26i{OZ==8%B5 zFX$D)^NIOBz!R^ITmV_M$6ur>4;r}x<2yX^@@D6?e|nMrk(cbl?BqV#3U zrL|B=oSUPg35Lkg3PO zQIKu8_$-7-ir*h*-j#b+(L7w2`!Tf6yJV$On(5C*&mWE%5TyYTmO?&C(J) z`%sDyF2?H;$a1+pQPQ&ysGo^obFyrh_GK$CpwH$us z_r{O~axSDw<<6p$E?a#|sX-?xJ|a6#e%|!erqaS_B{_QtxtZn@^wGJh_Mu3}p&o<1 z`*Xf|bIiC>&#X{V&#h(G5fvX5{PE`|W!1mGr@ox{&6`4)Ds%Y!IvKCqP zU@LHQop&V^lE4;2mQFZ(2B^Gs@c!WUdt;=81DLxZZF=7OL_w*XMB8@yYzpoCZXna( z^4L8(Bn}NN^J1T2SiY!q_R?RB$=T@`-1$CN+$j3u6@RRPTgIb+)TU_HYw0&JU-XGR z$_Ou*B-*1I&lpZ}jk@(o)3c(mIzn}zF2M9iVvm4{1ID+Dnx4zkceB*?^;gG!yXjru zYIHJuD#h7oA8?!O^O&QI`P`Q7Lc6@H#_o1J@jZ)v%{mxIOd~_%HofXrS`}bkixN%G zxp14qp*-77Y@P^4*gUM;pX_GKgi)%5mFOPP7&XsOmnU_+7W(3a4UMP->wq8)+7FM+ ziK4*;vW~mb>}PtgBIW#g{5nRF88uIeT|>KI^>Ws`w|>;=&}0&NLGCdgnX%!CT!7Pw z9Pt#4dN3Hv>q$T+FEB*52xLgI=DHxW_=ojp)__TvK}Zg&yik7ac1Wo9)SAy9{D9Y4 zazIq$w8bN=P#6s#VNx~%??tDxCXI0PE&LpIe0LVtog#t0#!T`komheKuW7n)FY8=Y zag44go3aRyvQ`nSO;gI&ex8M2ISeJ0u_sR@m#jGn#<070;uSbRRB`&sX zYpQ!wq|y!Dg{R_`M1O7D2R_hmuluDUHauuD1z)|`sG^g&2NhE70g$6c1Pj9&7zd}WsBN+ln)PhkC z>J1X&Q!{52<=rFzZieZXtll;@psa=Y9H-SQyf>hDS#~Eqi4?nlk~rglmf#~qYGwF|(PC3wZTKmo*NS&TG^F}k;fR5KK%HNn7vuPr zJX3ksGMJ+MnH5Gs^-m%MnE4B4I+7XIFMDBmQT52*?sX;dhMRv)KK+X&M4}M8TAyNt zs#r++529pNdRW1d4JyZWtKk}-82#(bw%XsRMoS)9%Fu|*sr4T!%L*P08di2@s}v5l zR;{p)rLIdNiLhuw#z+wY;83Yx}m$w6q>YP8P7Qfp7dc|lDGI( zh-4`jk85$9}tPdv}0?qtIKrudKVUCB)kv!{kEyQg94u9KP7rJ{9}z}AMtC1<~NHD z9UjfSraI{sENKZn?2CF_z9NK##J$QAb%9eGAMia}$er@ixm50dmB$kY!NxV=Cx8ucvq->bFhbP3xhwc)_WuF zCjUz>t@>heALV9rj}H=l#*m*lMLM=?=}B-dS)H5D zy8*ZIGmCzfZ>F{nFdPPdi~FuSR0L(FsiMt)x{NQ{)ALJr4Vm_xIid|-_d|-9ffHJA z=y7nv{l-{1_?Nit@OCPd)Q4-(iAB`z1H7>Urs8L`itvf5LTOjd)tSlV=bj}Nb5xZX zK~dxu+(9dAAOCzAv8SuQoew`+aH|*r!{F&DG|`jc$tGnrfz zr^qPHnX)~xthi))cMc0e)j0HeQ71HRMFgDXPt9wJIiB)p2q@kY77s^UX^1|rVwMUJ XHB%-T$J&+GO4@jTDXb(tCWea|`9IoCPwv)m_tPOzX;*Uc@>K@1Ea z5CiZ7I++ETf*2Y8JN}y)|2vqO{+n5tnVFbbSy);BXRvXwv$Aopv9hvrv2$?#cL1-r zxj4E1d-LB-{`XQw7A7VZPBvDy|2gIV)pXJU;%5huL1-ohX%Hhn0~0^PNe>7N(8=;Y z*x#LU9V29Uu8oX~IzAfJf|AesfB7&to`_&i znT6~!S86(i?M4Z*+7E(bI5R`UpOBc8oRXFOGAB3h)$9BZMa3nhW#tu>wRQCkjZI&_ zHFtIQ^!D|CANVmgKJj~U>d*Ad5^i~Ab#49c#wPLK-u?mUkbHFfA72b0rvGINnE!A2 z;s^L*{7(jK|MA7Z7!C|3erA?)nydmQwrux<1f{i}vkP6xEUf9|kkPgy2tNoOK8>^^%CKVdpaJsTOW9&NN;MP4HM=^ zpLl*@kU6d-$5O1wrlbe7cOK*|iVqpn?apBhu}P+y$(0tV4HY}yfGjWk&X>*KNcFUJ zIQ!2d?u^QoCV7+WQj;bSP8s?6?=G6-)l%8T+MrdW3Vf8B1;0vWA}%g*55b7u*u*rr z`sN=`qG*Ep_7^0+-v2J@Gv9UEixW^xC9<>U1XPR0=TSuA=Cm9HCtRCI%dNmEZGQ@F z3wIgc#?^jpOtfvvcV@TB&A&vvvSi)^4x&gaa(5Ex6 zO5!^gc{Qo9#K&#k%``H~N!eJaF71+KdHwI>x;nuw@}|(;WjXmKs}Q1l3a@r2`1yyn zhw$OD%{;1jEP=DXAKhq^j=1Fa#N3{Y#2z$#@i+mgR)=wj>msW&S;joZZcOb9X+JOc zx`xp>0bwUFar_b5BzB@kZ`3DQv&dx#c<$BC_oK49`WQtIu^kbQP_}P=w96ZPD$Cjz zGfK7GUBHzD68Mkjt(^nPkEir+rDT88Rw8e8@VSPXhAiBg!{4M7XB3v9`o{?<9W11a zN?Q~-fF?B$+?zUN+x}&60zyTIu4ZVW%s_&3D+luJi-45Vs8^7z^pQHB3&p7@eyG_4 z&UVhL+A2+7rO(G9o1kb8L94SV`U772w^)OtRet?YEMdEI@arg0tN(h~Qhe&DAI*Jn z>eV|#ka5Gl_qN^#Unvmb1e6N#gNeGLYRC^d@%7JTu+~pD^%Qr%CS+*N9ayEigFX{o-yUq5m9T_`wHBmO=@zjOZ+f0K zrBW@BxV9%DYHKBtazhhw=zNO@{bGN$BF-cV??blHC4bQ%W8N)H!*2REbCZqWMmt|^ z4oLg^&9O~Ox`G6UWTlzXil#6Y$xG#`f?E+6vH||C+{<%m`{!yQ!+mrQoqr>o{rOvp zQneH9D?-{=&;9&3f{}ds$X$XU{K46jF*NJ%SpiK*kL{IG8^hMcU0b)1A38F1i=)U_ zUj#^}3NDs@@&y|hW8|vmBBqjB%|4PbJ)-T(MB%-8=%$`VP~_tyR!gzkn}*!`OwZ1< zF;<5BgS`+)f| zu-#K|2ck$0%M|sj6ifcPrMA3v8*OfMws`RGRaQI%U$*RwCWI{FZI+R#5W=lg^z?9m z`-YJwq2*d_Lcb+nlIyM3hKwEm+OMlWL%L5X=<}cXx-dNie1jJtZFa;6aM3bW1MIJK zzm#fCSSbM4eSP&T*50|OVAMJJ$aFhjco+rl1ZpjNEQobLRa2Ldh4kNz3Ey-1GqlNX=oKZaH6ayW*T|es60C&efW8!9+Ie6rWV+fRIc$ zZw&Tcj{YFnMSt$9;_Yy^N1-c!ELTs9e{H=||LH^23CJ3qZ~|&r#G4UCNxx_xsCOuq z@H_4K0aiUTFSAX3FAcobnZ4Biy(Z~}rjX`~XCUpE9?w_kMKJk-_A#{t|L4?O`^3rpJ!mxYU&=(Uk|9(yeUh!FiSIk2KsiXMY$wQe_d!t=wBe*Z+8>vok%>$ay59*#$fRK%Ia%Wu%#>|-^}i(5O>yNmZ+ecI%D2H?ES*txH;2Y zke!09OyF)LSq|B0&8UARx~g-avQbn_Z`?8x-9J)Yu@z5yWyAy{RFNJN*t)r#8Wj@Z z!X>^*ulkrpGR*+zI28tEbGdu)kWxpwAqVk{k293TFm<1R{2|k2J&X0#$yALZi_aD% ze=pk2Sy#+sjXv*quxFNi7H;==rcHHI`9wCQXu!qD3e4V}?flOc`UigRKFh~voVR(^ zy)yD4K#53(%J5%(27$c#6%X#-ezSSdRCd3*=i&PM>YuQgU7RFZ0OK&ACFEzei$?Q`G_H7DYzPuE&XlUM*fYxJ`T!o`4bJ? zA8-pEXmSryYI{A{NY*zHv z3|$TIN<0|ACxYObB6PR)RP}YDRO$CWKTR1~?(bw>jZV|WfeqGCeQh@&Z3gzyZtd5O zBs(d}_FuNPt1!;~N6xGkhkw7*X<$eL@Ea!O^bmvm^sX%pJ%BVE2WSQpti?6`O>`21u77GUfaR(Sjv08Gn#nw%Mn9`UGUD zCRsDzuDOSk*B{J#;PMqlmSLmvgO|WmVJ>WVGF?65aQOt31lNQl8o3|q~j09 z{g0KZQWuI!D~!Mk_JbvdzrBkU{kVjAMWqmUIJgd<`3kS@g8-R#

@!0 zJ<3_KHt81e1hhT9Y*o0Fv}-(79f2VI2y4#5{qxUzVW!>X_&RCEi8eYE1;IC+fY`*4 zoyHhn34DDZo*IcE%S5HwE5 z(Wv&LOch@D=HVQ^tmk#S5u|3cZ*sQmyQ4vHw9V6o!Yk@wEr!9`Yi{Pn;3tfg4rMVe z^G4pqx>Lu6m9DR~$6>F&(NsorUW(yN%50yar%5On$?f98oCp8xhqNdLhe|^`Ze_

IO*ZaH z?}7>Ye#fT8lx`ZK26gu%{YViCa%%`^&%cN?{OxE1$HnT+r+oNncs#;LA$ z$?^S`hqk}iTj%I@&&iHySY&X^$V|3twi{s9y^*SMIa8%g%*Uol78|F9J9oefx{HY@ zHTG(mj-O?C%>%i$ZYIjhd&uDQZc5*OqftNh|N6l*Z2)KfLJHA}mPY!)Osz!$1uY+- zA0u~-M1!(XlqW6}%(C9LDc}uX&}!(?TL^J7ihx|wONIxq3|incwo|L30-h*+pX zS%h00JsCQ?WVzOANLry)e4y1}9@(6xPafMmvx!q6x|hJBq%uW5!^D4F!{2*WI~Qpy zqWUZEyIJ3MNE%wb5$fJ}G1$8I5_&G0E@9NU6r3NrE^MGC359(tBE4B&q>6`APLs#_ zc6Lc{^3*bv(a)~t!ehSxk-D<(Mc?EeNA8Qkfm&>z;85Z+od?kKm^kY$^HM~4+7>F4BTpuh2+X!P30G^M{!@!`^w@LZrWC8?N-Hl|9*s~^}P+`$he?~kn zypCR(bsyqn(%~JA?s31tOmE0heqF5cKnf``y@jbKaUk4JBA=b7Y2L&W;bRv7S$bf$CF^>jf%8o=;ICqoEk0giOe z+rC8Yz7X<{sm_b-4X{F!Ag|^J*7rk#xRW|M#+rqjyl`d)m;AROF<`=zyzOjUfSMpF zfmUDy!)}r8$z_}5&pOxI&ZnhK?lEki^P?c4L|B*2tRS6Zb-CpP^t_peR-Hig#qtEt z-b}nrc(;%?F7d`7_L|yGVe)IYVzo8pm%&!7kFss}QT?Vl_9*qLJ9RQF*((PI>*gQI zKki4BEtahy<(yPH&t^^u`(J(QvgMGcP#kVsLp?2CCix*^($j$VeQe5OS;^Rpr&GUn znQ)>lx7Uz<7gcfVH=4eU91n)VhRj^64QZL}=S+z(k~0}byvfm%ii0q0h`)Lqf;Jvj z;)6I;U#)(YUq1u!E0Mcso3;7E)=XGFu^&yiQ}&|z`LF2tnM{L&-fD9GQ8y{{JPJkU z)Xsn`Dc0_0z*qaXQjWvvoE4=;?WYMDpM`Q_Z#Boa*Giv10gZaK+ma#hd}!}3hhMO; zeIwBvt(w-)a3|116?qP7_1tmr(apgB%NYu^zyrO)v$HZ9cN z^$CQ1$RYE0Z4r7OUVrdoBw!yc&z1Q?lom~$G?k$Zi$ zJo`i4InVoFxjnCq2=5%8Gp@UyR5Q35@u@|ud zsGvDuYEregw_mWpwHy69Z?1pAAdXeCh*yE$Km;n)^^IK{i7VzeN+o#%0Gcz1QE%a_395TAmoC&h|G1n*}A9ViK&xO|U1RM1gtV{(+_@(@5MU zge(lH^^X9P$(7ep$a)rZgr)$_vU7Y<%2;6_3JduI>lg$gK{Hh%vdb06{-Jd;&_hN7 zm$(-5^I&Ve>53Rn;7eJ^^)5P~WKMD1ayF z6vSYY-A~n!(Zt{rkO`_Mcd&V+7kWxT{PVZa)tctvVI9NM!P)_gFTTz7XS)vo_IeF4 zW%*QjU@j$F;pFQGWgX&uiA52Pu5QwzQ!7b#H%|@1+5&Ib<~rHG+Z2qFl)pLl1hra( z)967bK}F#n;Y2Q6bw&e9U~A?%w8`IvRH8ONK|U=}l9PI#=^00dB(F;Nk_}#yVh-mR zhwUR{z+fY9?4Y%s)9M7Cs)CF9bSbdle5bfg-S?Is3HP+egFX*D{P)gqI2;{bvo?l{ zu&pe-nUyX#8h8RS%^^bK8u&ouH@68=pCsfa6U#@+=4H}^vpB@%nV+um*j9D$|5^B;gR{Ix5AGs=H$3eqhuk-s7? zzkmPN)Mi=#35d~>sEx%W!}%DQF41!o8`#R;7QrI{Yg#$A>a;GqoPAJT_(NgHq`MBH zv&_kx>%m!t*XY%rL&CW9!=OnBNph7>vB@%+-FcCfiJ|1q%wm=fMYX0CA zTQBC1mF3iZn|ePCPs4`l=lR)nSFu0B9XmyEuKjqFAE6-Hzlr?>6!x7yjGp>V2=Q4I zsQgs=vYmDUl1ToFj2H#R*dAbC3=Y6KVN&PHtAgO|heVK40 zV7;Hi1&Q$!efp2+%6}ogK~&94RK*hzc}o|uh3xEGY(n8xC?fQscHUN6uCc63MEETt zBvp3oV8~%%SS58U!t7(Va`#%GQ`6}`+S{u7Gv6bt*EwXf6a;#Ca1uLIN)@`+h!-HWLzOOx&d1N7>lh*BH zcTD z4Qq`tw0sQqH)Ixx3sEnkm{H9L)!kfLtO5KdpvD8K*4Q(x$$_leG3r1lQwJ@D)M!v8 zRw>N%o_76FwZ4}D%U~znjz1RHsROhxbWw2J2?%^J#K~or@`Db;^PHEy4}YC(czIj6yNt@T0WFhU z&p+y?bLtVSsnLWkNYb%fX0%l|L-8xKQ)+T=JbsRZe0fIiMqHdK%$*|X#}$m2u949o z;%onxj%_@SoA7nYDRPU&2`G+YPgYqr;P?ec77vuVnRPHO`$t-%mt=L4ATy!+!mGv? z30`%Ct-hya&*7%0MDy=#=Z*(R8jp9Uoq)Kw&y(&ulk}RJJ@>^6UPq~lUKsm0d3dUZ zPq2c;&E^3wyY4Qs{vaOJkBmDO#lx2TohJ7l__`Z3JYyh56?gS3?Z!AeDnL|UQ1TBw zDQ~ajS^Dn%MMPRhhp2~6^;_Lp7sbpS9M!E;{};JvZuS#&?mcY?E4T#&FKPdMxH$p5 z2O)&gybE$DK*EBx3NbbV>xKPU??cx@Q#FS0i_hR!rbM3Tc|1jWY+fIhxrO*gD>a<& zmgI1p#DKo;`0oJbM^Bsd9z_CtXiR>DcX$3|gK%jgQN9bw*EvhI zwy7ST%bOyEih#k>^<6Wy?>p>>+8~1mk)IEj`>DcKbdG2WyRTpbUoTRMdah*NhsY0( zX%!A^YP!Y$){NvYcb`Gw`C6pML<)5mh!<+qA-;SJx^OI^#Q$uLyHC-*TBBbh&m{Fk(##F2dp`!q+23RO#rf~cH+W6`@9PZ`Zm17G zL%w_JuA1_h$o~f{plgUp#5j@=&X}64gZi^aa`jH@yW)QW;@@pNMW4J77|s_GTt>@; zzC+iqQ$-}`oQ4!v)n;TbawfY+Lh(?sK6zihHu%s8^18~O z!n`F#xaQQz?KMs31<%y6FmbosR071Rl>^w}Tzby|7?|f`v&~RRAHw54IYlVkBIMRp zZyp~`Z;xANdd19{30)hzzVV27POhli;)M%F$MON|+Caa8(AQ-MN=U1=vQFvN);A7t zM6lK+3$}lse3t$$t;`8mR}7`}Kd7lR{=Jyv?4xkHdgBe9XO#BE1`p^z^Y$;JGq^Jm zZsfeBI1C|ev%EuvmZdv?{~S{#?NzaBbDgV!(aO}GHuOD`5U7W5(%7z`MEjnAxQ(no z4f4g&Rn5SO2kd$%C0(_n@txC8lf|%G9W4(Ke?Q+H=fe{|16pF5SX0IJs{b{mc9f1z zrm2ovn=feRhrxr6-%FCptJTzCpJZ#>nlgO6(u!-Y7v@Z)oif437~0p{e#y6XM!$wO zG<9%QPoIEr@A@+RCC&)i&4(4v>7RgduTpK1?m3>m57Ld{hKEC2veypDoY^pn_u)Ri5Y&Pq_*=Zlb+zm-nl2_0N?+bz}JF_fu?+ z*?v$nB|yTL&j_1q)g%xx0udpugv2BbBQ>0xvel(4T}$s<8eTsh{d(zA-+2?sQznx9 zZuQ<8F-QR_d_+*}N`8YMgc-v&sWDy_YXAR>ieT?UqMVsJ#l{)C`_E8 zp&OuNRWwz=qWc)rv8Cc`WLfR}*IkE%(nSl~TkM?{YZ&J{=)})lbdC^Uqwv5l09Fb? z?`TE>sWx;BU4fXG42M*2sxAH5-o4lOMLfDEX4SzzS!~5zRPgSz3a)1pD*6U^~WD5hZ9%ok&F>@*5>Zd4ccFixTz7lZm-a{6?X8N_Vu|3 znsX5M^F0ruH>Q0sZkhW?s>nK>y@Dcw04lr$8)0$_MT=~{ZbKAJgu9i!gE~!8E!2h< zK6&d$c_+!HudZ0hYT@Grxg--$4U;M2x({&eL@bq;O60_orD zZC@Xe_T;_vNo-7G?Ii${A;?Yur948B72}7hQ|r|j;KBsr3hfXDxAEKJ>59=(eUG#1 zTUp9`WZ=Mh%Y1y(ww*O`cO7{FPEww3*Hk?t<(>;gKa9X^i{ij>1){WCgv2KF41&F> z2g+rIn>7q@?BeeY(#iYa=rqyK3~C$i+o?@`@UiX)aa)P6fBTRsnGINoZUCP5!lqMi z(5j6YBh;sCCJ}-@U5l<`^@DOv8A$F@Cw#hKtXI$X$_w%FI~Q~ARC|)0DY^jYQN&=M zLpTeGhpAz2?1^55{Eeud<2l9?P){@Ur_Hf$e7&NZeU|88liA%zS~W7!(UDl2WJh;d zvdE&av2ld_Vn<`R%Z=uQiVtK!&H)NG{|0j~28#4WwUqh&dl|Pc&f;b>PNhDs&HUC8)LvM z<=;xHTqAP?Q7p-8%OqE;-{fGrLMx0_{X<%9DwC2XR_~|G*6QvNwe1Z(`u95aMKJFr zm5ujD*f*}|f{DbMO5D5cLr4{rEHi>%Ekkx*Mn2Hsq28X*Ra_+oX2hBie>b1IEmgv( z?%7EfE7Ba+%#rfFyUGr}J+=2=CG8v_jzf+(ArHR1)T}=vUP<&`fh184t0N%LX2MkR{=i4rw%5PU zYM+Iw`j(siojqjcpJ_8OvMHvIw}Tr9i|Gw2QQ#!;M;x$1|0#`HjZ;^7MTR#W1Bv&au z;IhOUbUxxvtZw?sRs5W`_WfWG#OXw9m!4zbId zqd@E2z*Noj@{O5<^zu2MnW2}HO}W`RdJavT4%dorCpfUZFgYlXIHdWPHxo;$Gd`Xj zN1Xk+MxPc@;S*r-&Km@=B>d%mU1Zz!ogs(q^veLVwe=V)-y2$50vBIC-i#PFxgAS( z40ZMUNl-TDSm3kiUqhUu+LK|p>exj=BU8egei8FUXF_{ZSc%`S5K7CNNEaoxuXBk7 z2hX9EU_xmB_U#}-Dye)7;=jo0E0E5GbIYIg&$bZV`E>BcY0^l5Rdl$?=?~MOj<)1M z|7vtV{O1}W8QZW+M%^5$yjr22qZ=E)_BEz0NN}Ly@3%Xn*9k-d8L7+fB)aq3H77re zKRQ62vXq|bR$2EkcB#{@VUW4O^Y7w0&-t&>6iPI(I~&0D8XXu>st%!l+uQpS)uxGP zmOM_iDWA6xuHBfa70ul7G?M$2-F4Y*XVsn6ZfC5oY)|CJzrCBSskY zoNiH;lF+K#nQNZ+E}whO%%2$xooJ%+N6|So0p!FYOiQa@6yBOi0~5Z$6)TD&j%R(` zrjpRjZT4(5w=M?x2!@{Tkl$0Rp02|2ekFQ42Nn%4?rY6;X%{+2QOcqNror0uPBbA9 zww*^2kFZAi00t3~S`QK5%D;r3J&>?=4~y+;7N`t-^1`OxE$*mJuEXgVEO(9Pt|L_l zM(2G-;OpM<$HXEDH#K-?wx%XGCu^xDld3;_b6p=7ZC@y63QLo`Mv<1Vul!l2djx=J zmONSwfX3dRfJ8BVHX?|T#YV?8UNl>5*tFU#wXGjuV`1vR(B8qLJ7-{0RB%B2}<|nGNqxyo1C7HzwSQ zhS?FH6v}?ENvMgaG&RbrlhG7^Bb(Z&FYN_+c$MMK;q{b4X^&ojhVK}H<0z5KWdeeI z2dD=umC48D&PX27;_mo53v6e!j+T{C#;W*=336Cs%iob=hOEM1TQGz>7(QgDp)t!0 z{Rgz}%xcCvqC!vp*~amEV;`J~2pQ@1UFx|yGtw{Peee??p(4vP^U4#@Fnn()O#ANK z@INb=+$>9Ih@oJXrBT*RONA_J2iCZnFhrfLPrH!nq6Bvf&w;o=E4p93Nb7Q6gwU<) z?^)&)FAd=89lP9B0yv>UWeK_>W7tfAfNM^cB%kSFVHdX+28CxF19%BXRX(`5<Ob zU(Lw|h2DL-e@V4Wv%op^j>xR;Jb3zr0caFz04%a5pei4r+z@Orvd-ZRF}1UrCl}+B z?7a--QH^WQEN2;198u%Kw%Q>D-b!9e+-B_k1v*h4&!to`C9GZblkZ z`}4eU4xWVSz~y~A@dFc`5K0&wwU3O<1UAeyT5>&lc9*0^M0SWaG>XqPqGzN|wSB|7 zY))d^S_0|9$b3;F`i?galWiSh_PbT@lY2zC`?x|~`t-jmajlpcJEw9r2O#G$cmO%yDO1?{v~2qu#TPMl0;)lj;Vy-c zyvY@jORgs%5g(qSEz#T{^Wn)pvlm*CDP0Ek6&%V-zbxg)xV&0F=RDm*^|fa_PFwa> zDOyspuec8X^(1-}52;z-E(9uMoYoY_|Ds@wxyE#ss&sTqzE3dpb8q&xy*`@o#%}?x1_iMly%0Z;2g0!>A(W*Mx3n1RXotU z@LkjYk=htjgY^8lj+d<7Y7;9@z@iZB-)gebdXE+%LIlP@78f|5{yqYV*XFBQw7Qr* z{(+OI`Snqt)`v7ZTes*9;;Q5`9>>Nwj!~K5~6jys7?1sW1K9CIYT|5gVh`uO( z0!q`}72K3RbT$?y6pMSbV3qr_?pg;6$eLXn(>%?dRH@G-g)T&Pwt@-fu<6ELxFa#! zW+Kcg!Wh;>y-H3ZHBz~$U962sGfMyl zF(`5?A;*=mG~$t)a`-1`P9*YLvsi*#QuQNs!2+9)7`-U;muBzYMY5hh$%@GOj;}jf z4;$T|?EOw{|CKQ~;--ecmf&!NQxSb6VQr~c_ObR?;uh2ievf>K&_r6Mk zu^=*L%u5xEK=4j)W%e)ggt*1Oah~}2N+;Ml049NN;?EUSt_jZ@yd$mF>;Kc<52r9> z`g*pCj=BBUGo>3_UXGhPyIQ8aM{jYWX3#!iyflFB&YC8?7;a>`8dp-A4()0#Z1gYx zm9+cbQ@|ul>S=L})-^W6@BpI<;dS9-TY!+$^m*u)M0^v`qbm@3^ z>c3^W>YJbb{AZ@C8Ajl)&TXAItA<+~4>=#saU%S&$KuBy<*w)WC;7>7U=IP(cZ3^@>AuSvY_0iCu?LYws6GT=KHRMQ@;An712~nW zZh@?D%BLJzM-k7wk2fNhh#EZy!IC@9a9WO=FQ$8hI|&em1I-nx ziXT!yhEwVEMR3jTb7bJ|v_~(p4w})`lm9z|CSyCh=R1>T5#8`(8J%V3))Bp;{_gZt zt?jd?DdpF1Wo6Ba{VwC2n3MlpkVe-sK2Mb)?smXZ+<|@+qUn`vYa&~Mu5UfwW!N?* z`|7IY-Q{4(KnF>!+Ar9YwW>oaI`30J01xhx*Si2a-gy%;2i*iEXLre3RUc$U{Z z{wpBSR^RBAe#TAS2M-@y=?EdanDZKD@adBNsM_`3y6t*!0FfQiI$Hhn%`Pf*?{Y{? zx74&Q#hY-O&Mrd{Li~mjuFb-d(X|%I2YjdCH;E1?*1E!o@gB#735$x(Zwg~l^V(-x zrS285j=bu+Tq`EaUxJYLNW)=lG1zIqYMrLNL;5IT9_%I`Pp1MXS1^nP#xtnNpmPE` zmE*L0H&?=LL->9NH+}8Nte?fMg%R6Uo0!(D%fpmU;Ev%3PqY9SMd8&YN zfG0)5CKX8k50{j}uMxwt-sFyz*cs`S2y@Q1e!JsBe3@YUaLGdk{LQW<1=(o9K9iVP+W{2shJd;dgX}{I9A@rW61npge`9Mkowt1w;*VtW zm11lDg)o?(g@cLLtx7i$+~W^j9RlpmkHNk@yf3#R&=rwhCkYF*lqd$uaTf2zE*Va- z3bGOh9DA(Zd?(eZjz0VsVRXjWDs1cD3FyL8DBgl{k3I$gqUcCtW%}5naD#@R-}{R* zTxA^oWj+1!68@n*Ply!hv}IMNJ*X+rV1&fRVHy{yLQEEoOWyvPaU7PRYrhx|B`l`3vZSF*%AfO)QzhJ5$X6GVU%_#~wUv8Vfbp zTP36JpdO{wrco!qtZ8i>dOi6!t6W=z02-P)ZvtEEBeX7$D$u=D@~%2*|MwK!x8jfB zf7RjUt5sYh;*6;}!$O}*KGeqTU8s@e7dLR=`mAt9`Hl#f>kLwcyG1cUItOF$l)cnfTNCh`kXM_dZhV7~w85U3blCapvGOq)jGTPi*-;mw_{>hRq zLq^_cBZ_WkQ*Js(BrnAmEOr~~jM00IiW#4%t#FUlNyUcRezs9%lUe6W!w`Z;sKORF zXma3z1{2)wS1U`Yl@diAwS?iE9+tk@S=ycvd8W-tn62FQEpHtO=zfoVkHNkNb5Q+# z08Es24C&S2)~?(X$=(o|@hTFE1huu2GW(^lT2BTg68iMy!`VsBKV?RKqVx9CrDWoj z2dASh!>N2R#)k9}j2OlXlANE{xr;*bhC#R@vaPZMhF@2j)Yq;&m%UkBOuWE5 z?crtK|D*5g3>}U10Z&s%K+Yic*$L?Wyk37h^W;+2DmK@KmE0S^xalr75reaGebqVN zB&$04b5=4SlS5j`V$Y2@h0W_hPPZ&8BpWlq-9}m8;Gmp-Sx$bW%GDRwUY#3FVp$GT zkv3*s2@5<}eBQ5mJt(U__b+h|Ubu?4FFmRl^m0sA+Jt$V(}C1077ZrYL8n{joCbsvZHWy3?O6=I3>AR$jUB~+dmBEe4spqVs+^#!Q+a%nwS3&(UZ(V-kGgrPp$$pm6 z+sHS5HAi-p*P^^S=hKzXQt$?0)Y#)!a~CdVRhTXCj4EK^Y5Qcc!WRy0Y5!VFOuxN7*eInUe)O9-lnRn zU{_S@igp5EKWi^oAOdMdbdQZ(LJ9bgniJlY&sqi`??(mJc873PIA&bEXIi9mw&L{J z7yd{8ed`FY*|frdstwS^fHw&#CO_@N!4i$tBE$F?{ceiIX$I2P&@!2^IxH{442yhb z4IYRdzpg$7Tp2^}hH^o$?=f}eEfizw&E$Jg<>~#2VfD|9W{iuU9qHMw7jP(VvM)+B zKMef7Gf2$pa%dsDpMcgPzAZRP4sv-C`|r@9+D8IsGd>j8wYH4!EMdlXieqpb4pxy% z!PWoOy_|wipJJDJVfF;rB$(|bB9vrFo~@>jEQ+^Y3|%}$4kmPl#s;M6Vs>n?}qStEL zi>l^!HTy76^k0Bp^ckF!xpT`>e?YfS+ERu&UeDmHbVhMOOxDN0GbobzC;!TNlgXJ) zgB1Q3v!^ex?~a$XS=q*^2|swb>QEUH-m-~w#xNT3WBJJP8dO1l%U`xp-wtAVsknPToRSU~}`kbxg z<(w#I=Ns`>223kpvSJxkqJ(k}(a#l46vUSwmF`o*J19X$;#E+_W>)y_-hG`_3phUJw^XnyGYh>SMjZK5iq{L zQX_ru>)JJ!tCH|({pw*D+3<}mIkU=w5Y>IrB3Ab`T(h*aBD&K5{Ki;A-mLruq6yax z(Lra%eofAs+m>S(;ZW-k7s6zXKadLWJ9H2wiE;#aM0-9SwT5Y(8j!yD;?B!wCTwS) zLqC9Ps;QDKlt9Gr@eL^s^5Ok5DwMF=%BpCQw8PV0A-0L$sQl;!`UhXh{N5*d&oP~r zXQUS9fNBd&pCuZ`uAgj)bkMfeY}wEWiT7EGYD`4S>hU0 zAxA7w3FxB9PtM{NW4GT?^}_#|hgAjT255QfpZ3U2p2Hb%h~3FDg${`(w`$|~SR)cB zT?m!Y&!Kpm8!cH=wqM%fi*jdt{uK3X{!#wX5yN*yg<}2Cfzqve1cdM`K;kY&z6pQ3 z2uhzu*G?r^P*u~F`+pK`5(A{ydq_3%%$bUk3wJuTdM}z!N4x0m-@_Bo`AN9`;HrrtPxr?l%=E)smJBjp{rL?? z-T5U5P#n77m4`qgjl6BbRms$xjJ)H(W|mpMC$M{Jn!(Em83>;*W5cmjkmn z4Jh2i0P$7l4@j*eUUz8_91o!0XJ5$v%ZB9w!^afcD+_0S&y}9(+0FF&817}5U1Q23 zXgS#`J`r@9bEiRgNVM6wn`rNx=hzVGX0iRPNLg5e&|^-We*>ytl|Pmf zt9ZWq&mys3770*|BxDmlXR3n_NM2b^7)cO{e$H0NtyJo22nq+=&uz=yYg)S2#66pm z5N>3Skw$c(Yxz)+z{Qz2UC26x?)hIM)r;Qt6EC@a^FvLef6scyuJ%ayy@kE#d$!u^ z8`Je8CRgDUZJB%ad`a3$$GyDfeE0SOi%mzWIFLhOCOXUh#$byuf}sy8W7e;b10;X6 z07lgjSzay1raW10az@>RP1*loorEMGVl`$wmf*&Ti zu`%+r z+xUtH-$@GPGKzog`-gZ6{;803retdOU9y*6ZI-2McI%XD6h#a1D?hxWg>NQ!54)Ij zdugYve8|IdnZLBf$r~B*RjjVr=MO_*|bSDNw zum>kX8+N-~ZO96fWfaxkr6Vc_R`bkIABz3kFI_Gc&*#*YSFg|I-d1;({3-ACdN9L2 zT1tiAk?V?fTwwDtm?sze_$Wc$zS1VAt<7frYlYZ|AMkuCaCa7A0Gr>VGf)GpMv}o_ zhO0aY<2~Ch&PH;<9~Z!F=WYtOe&(1lgt++3|@2DKVZ0T`dR`q*lQD{)?TfMWA&z)|Bal0J4x24rd0v^ojFohVU zTSI#Z;7FpvR-u{DEe7k*g$wJFdL;ENj;Q0X2HiF;Ggoi(!S`3)^s z;qWvRxdwNS?q2xaYA)>mJABU_VX;TMyljO2beG7N*s#m76<#1lG*1mL-Mkn*IJ>xU zNE+fMjmro}tXME)zKr2~!7g1fHKw9lNC?nx(G*e4Of9}%lEA^lw)a6NlHMHk^B=21bxaBAQ z{8qkR{CiWQz?<6O0-?Eq7?bpXPEg_J!Cnjjb^@Xbgpit~n91`LNJHc6LMO`yht^nY z2bFg+y$-juU@4DFURp{QTI7pZ$j!BrRwDbUXS69Uz{Bk(6kRfmWJQK8o$>0oAkKY& z3X^jetqWAxx9^3W-$)J?y6Ty4Hb-QglaBRQpH6^to9lY<>L_i5e;TRpSk;hCBj(KH zDIW%`A+7??A#u>^3O2uUJP-rq&U*rSX$S}4;1~Cr6Hr;y z=JE4COI0=!7a8{?~V^^q$~OKsrbb@pls;o1%J)LGIz%e1Aj~(8HJ_3|S7d zy-EPOC#r$F)lVUPkuw+Od>7<@*FVRhz&eG@gm-Tq>BOC7bc6S#ol~I6uifHh?_Yya zC8~f6cCtymv>Ffc>he6h4VEu%>akyV`_ovd!jFL(9{)!G)&wd0X|#+wt0}mUbAc3U z!xIYjfA~5#{t7HUBz#fRf8d}001Lc8TSf46(du45mexlLbE>jt-p#rPSR!#F09ky= zlpH(<4;3k5>qiR>w~(^RaLX`r%qp*W-9>(<%l`lizu>;#@KH~M-?ZO~{{U#6BjRw> z?0jXfKB269Kk*E*-ssWBPT1}3m6$!KiXf4z!pK-ghE4dze2wud;xCN;F8o5){we%K z*JsrA?-6RZ8ji83T(;-9xSB?h*!m63a0dWj3;~M6J$M)%lvk}w4@Q%`V`hBSd~Pcf z9$K?)UiS4rwaO@>z6|}GD58o0D58o0DH#>HVM|2-d;$Le1$h4ef){)t{{VuAX;*(2 zz7_q4!XFnc?XeEDAuDOG1XHpLb8;WAp%ZhqsPtV3*f!8P)QeMKs;uLUSOqw~7>diYwdomjNd?Vqu4{1h|xwf%tn zXnbG$H2he*(x9GSGgR>9==)BSsc@kA_O5mpAwcs}Y^pL9)DvH;x1aEHx9qj?M6PfAT+K=BRRvd0TdpaP00paP00paP00pbW1WcrU_#5d1NJ;GZ6N zYQtI4wM$mE)U>Tq(I>XLg@zhABXCvJ_XJntcl;Fh{tbeL`dvfgKl~G#;USsXr@HuG zs@;W+iOD};x{%>H+lVDQc{^lK6)2**aM-HYY2KUv03*-LGVCT{OO~{6+4WyD@sr{o z!;c^QBk`u8@ay7-gEc=7YuZ%aR->b97Ef<;Z1@BRjlf}$K^~xjG7Vx;MSZm?xxqGm zM_LMF#cn?pQ)ty1QfP?vjS3luoVAl&iAN*JXjgBsnw3WU4c}RmsP7xcaSaqwv9Mttje2^XeTT)} z;_`)A)m80jeRq>_o~%6|#<~9d6|eIbM9SmZxtdHEW9t?h)>D4?`V1@$|E&xaE*m10 zmt^D{E*tPO5va?M7VA(+2h`I_$+a;>vA_RTh75}fu@Ci}_s=F9#l_iKoyxwgcC%>g zmwd^O+f!tXQPe~N{@{<>N;!8cDR+T6gA& z+Lb7+XyPU>`N~*#=19OS(6t-{h??G3VDy&+o)4nl%^-dkp){k}!c7Wi$_M3+G@`su zrU4mAHkKt`;_MUjmxv*zuzHJhmx-odUIHJ^;ja+2q1p|+&zm7!A?owrzOxnS-X@a@ z$}mT@oW=x1F^3e=JR;zN;F7+w*)0}}bEtD;8UBhIs+c=!#jq+?;VhQ9KEJytzk+sO z0S#n^+J?ekrmQE2YoiL^F73K;jF~a#Z88!1_0+RH9cp7Fual6k6_q z^EPgCxXRsgU)YoQL$-GQt-Z2`tS{N%U}6D6BE{07Yc8FXC%O$MSmqKdhVCoiSi z;c|-=hZAWUXBUnQr9SRk3k;&iH5{5*tl@oXm162&bTn}#)Nv(tH2ZcXvv;+;{L-j4 zP52cSLU3(pF=`RuGW+p?lDTcALs;?kIOHQ9?`{~m<&?&pv8G0{xv@xnhI6XgTrOKe z{}#N{`v$Cjy4{I}w-xAlPmvBcCLo&HA&&rt&U0Sa%)?FaHfAXrN?d{LPhF97so-5;pbiD#f51!SG?C z{Wb;JMx@_3oVGR_OH(P_07HxgDH$(M5hvhp$q2kGc1QC&EkV`p5>cNICU3RvDP@Ri zSH!U@(ECcnV1uYzdzJ>mE9Sm^h}DXaC?fmK`FV;Ai$KXgI8p&;J{7^Ffx>sh zng$EcAP##ee#lZ3DTS4YLsq|bF7?h^gw#yx?y8c_tj+zaU=C3hQGZUmlVEe?KE>Y$ z;#f`O*-*#zhy=|QO#vN|gyqs6;~C}OYM|e9%7wF=Q)J^_E}5|ccoA3bTETDVla=;m zZDO#ytb@<-^*v4CE71FaIAtS%{d0yyYQEKUkAeU(K6wPg-}2$l4;)CaBVZMSz5&Nj z3isuOdz^8pbuQad!^C2 z)Bs#a3^G=@%aN1xt_=z-^<8P5L^}H-j>Z}M@LYT?hq9WVyCY@n8*xYWL)Tk%ni8?f zM{g3?XsmthN;t-AM&X*?)JngFNpsj}BuEr><;-lvjVqDQZYUjP^f7*F7n80_*>IqE zKUsq%NktPUuY@Zgr&Nnk;fdCvEF)ajQ!(>$?A2#KzNu5xVXkMI9D$?X9}i{pn}n`67g)Ue5ydF&@(R`%UfJp9FJ}nbh){$X{$&c z%lN8NYlN6b?`ba;$xs*kK>|XAznvQl?3EiDDc1Jnp461*#xZ*s_0z*t0 z&PEJ{zvB;LQchUMEnavEB;TJYpC@3-ni?WIOVd!a#fpV2mr5$rm7!?Y_dXOm1zjsY za$XtfC$G15UydXM3(1w;Um_giugW`UxI%27x*U4P1Y9zDUx5UYh5nOrMo%`sMG9%r zxQ4lag#e=m_7&YVd>u1|djf`lbQR+^hzv^#^$z8IR@J+oLGD+kzY}$fF5^oii)+-U z_&-{9mT-isE)uN>=B&|}UK|M_%<+S?^qH@3XtkN7Z98+g6mvh$?QmwJ)zE~_@ty6@ z^{ykkP`o+f*d-Km7a*eB@=cnE+-S(Y8~flXF5q3WxnI(xCKIbXR^m|iMiMBNlftp+ zUB=>l{<#&tt@EM0onKk8L##ibrv4^*Bgwo?T`}Y@w(B>X$GM@DgGO*pfJ}iDKb0$H zPRqF9)7?ujecr25g?M<4g@=mukE1nTby-j6r1XkTwh-Dq$kRIM?x`FA(xHe17Y|^U zt_zQp%x$S!o|`-09J%|OBq+-E7la`F)PKaA-9Oy=5{$_T^aJ@nyDnjnil2a_|!71yamb1xA)m~_}srjOrj z2SM(yGGdBBCz3fCQoSKusCqnN3T;=T^f$@{)z~k2k`YT*GCh{&5LxVG#$2^n5}wf! zw-ev=HxaG5pN|zNejnVaqPnVd7a;zh*ro z#dQ$3M1swKOJZPPuVc~>?yfwj-A^T@j@oxz|KSUZvxuWtD{ zekc-yIp=OP6wb+}jHm(qSz|XiVLCp!VrTjPzolIS2$GJWSz0mo;> zBg8`Iz!_!|Rpt~``-r+IDYZTZX#;2q4rk&04t_zaavjL#QQ$H{O9#4_<)b6-o`8=` zmI%q-he^-g<>9~m87tj7Ny{Hcovae1P(q;>1TYA!p?&H|wTRpJoCHb(tD{5xR~UPU zIFp^()w_@ZMh82iOo8WVEAx{)`lNm)8c5C-k6X^WW6ru-2Qwr82C98-&bq;tX&fGY zo{91o^trk^Ovx0;Holl<%DcKHqV3s6@~&LuJhtdjXK7v$0C*Yi*@GtIu|(H)>tHO% zq*R1+swoJiW_dOLum%%<*oaSD43A{S&tu3AS|k2K>XUoN^1g$DFF?o)S%I=p9LXG+ zM2@5ox#VZ%E@XkTt zZPOZ2sQ!5y21%|I{4DkNWlP$6FWv@x=Uh7=;{Rd1)$V2rITt$u7LqstA}#up16Bn_ z9hH7m%@(8Q-_IY&a3JE3KuRsOKQfA*LMl zYaU?BJdt44pYtojX(E_}sahKusU?JbK8cu)4G`BuWP}f<3v*Kg#Xq6Hvp3_j@3OBO zkB$mMsoZKTe;LMKrl+*`F5X+FgImX1YLXWPxV!r6(b58`{i~L58zsL8KWJXTGhU;J zRm1YMl(B8@32pK)l0Cn~(7rjYgW2nRB;KDLwAYEqw8Kr(vq%I=;=9m1+5#P<6k3vH zs?Ys}zf4~mqlL;bCC%zh(fjfwU|@s4kEm~zN#rOB!;uJCvBM2n`;usV$ET>nz|l-! zs!;}Dw`s+&$cqnlYZ29L8Q6&B5S#m{~F0Pj6Mpd4uV1W;C3uUr>5VBCsd) zNySy*;hL=j>CpY912HJ1!!VqE>2f8gStwYlI7ouLpg*W>Fd{XusJ0R~@~P!|d+ky2 zvdDdD3q`GjwOCdz#mgir9G!grc#fy4*f6wibuE4<14PO+zSXCkuB2wv>>XyoqAq<_ zw4BKtG)%@h;~(y#S}~N1oeW>_NS=ezZ_TE{OKXjdIxX)V0Fxx7B&qa}c#-nrMFXPYQ2Y0B(LHY}Vu`=kCwczqThFK8) z{Nh5~=5;2T3+hd^o&hGOkNmMy9N1&Z7~wHq!g_9*oPiM4963Z=U^cOZ{OQ$T_o( z*yu)B(x^|`A`F77MjjK0wi&v6JKYT9TCEDZeIY73o`YdcU_#vX&VDVf-E$Nj)dFh# z-8#F3?9kz6S-Ozmd>%H}#0MGILS1iPr)>e?edC#aEOQ(U%5jlMXN$kmkL;yX8*>tW z%!vm|>_%ngQF|iCkfXRLWX87b_%!9xB9ocgkL*M0Ua%@ScKphhl4e!wdj$vRzV~>; zl1x0nKazsBDE~Bu36*b8sSA%-4&7}Z4zC|pZwuX_25mL|pw3q&oJZ7z!l_Z5bLWM@ z16o?oELth6qE%M=J{7jO66`ONoN-~sN+gEz^1EEg;9ml6Gfb+sFy)xD24;a8J$b^J zy|r@ry99|6ySn6OyyyWdMZGE26hBf$xl(#4Xf5SWHq8Lkip>%VrB=wJ=0tre)=T@Y zO8Fa67$p_|yNGiqoe-k7G>49M>la^qm!Ct@#aJX};z+opgs>rlzFSdCyOFxaT>KEc zrm@NsDXBXKK4Hb4SBA1H%W7`o16fI|Q`0^jcSQStK=>bd7h?v*iK#x1^p1DtUYE|?d*$j932 zcdmpM8B^n77V6wY+11K{SApSmz?&n)ws4tgASeqvbXlVOXiGCud7-x%j~|GVEZRgi zduQ}k)6JJ|mV=YScwupIEVKQsi_3LYRb5845m{G?*dMGs<|$z%_Tx(VEHzX0vJ;_Q z0alOJy@_?fWd3csI12zmLAcnmv}aM+OGscpZAPAn@FSYYIS{V7o^x{;Tux28R-@SU zg<&z1(4Utjgc+5$bMifkYs6a{t5Px`#?WlfA6fs!_;wgqoUYedFsqDJmd&he&J=tdtupZ# z_`Qrs;fk=`jHCRaX8amktmOEXFGQl~54GkP9Yyeuus@1c2MN8RBR{03vgz>(_!ECO zvIGNRCi73|GV`MBV>FeCOu>|R{8nzXLNZdH({P-|&4ub}$qMZLW>tkhN|sasE+0E( zHie1k-{nhps(cl1*^-*QELIkY)v@*Xp%30e;yh8u=2yl7sOQ&2m?W#Wsw3|5#1A}e z7h&>=-xqpG3K?JBG8B?uWrpvoCHAL{hY7Zi{~c}q8%3e|g2ThhHjufg#93coQ_Ldr zcjlZL<2`B}ryg?u)*!z}74@%TJIgwFL-g3Y#i1$%i_fLtC=5l?Xp%GEOmcrHpp^c8 zs8gQS155BB4D`F1di!Yy^jQ7-XQav*GQnzpg#TVT@JCu~^A)0wh@|SK=^XF($X~~k z*p}az9xscEZj=dzWr&L64R~n`G8!q1GP5NG1jUTO9!zJly7}(+ER$d+#Z&KK_cO=C z3a`PTkFGq-4zV0^0`QUIt(xhiQroj1g5n@aaN9VA?GK@_G1FdRpRjA#$19I4htpw^ zwc{)U)#f181IpHB6DD3%G&ZL%@ER&QZGi6}EgSZWh(s?msEM1BZ+E`uj(zh88!$`| zv;HtEA23TLx9N2#TNE2~r{M#X_d>;=L(Y6|*P(P|_h+IE2r7jknc66u`i|j;-EvK6bI3Qq&vVj)V||(bJ>csv>&RgC$DhEg?O_J zjx-o(yNp7zW;U+Qcc=S);ZuAQI~T?O`BmY<L}jNij!)E>Vma zecoZc-F9SKhIQV!PX_Ny5`G(dECpZJa%ac)5HLX;=Y%&l=3 zELikbfm~LxVS$UiuW95{=$gBq_Zes_ro9Vthdp(oz}*JP3doTu9l1uYuD**R_ZA4^ zv#2UBN;YKU`g`gVd=8^iK701xi9cgL^BHQ|9XS39HPp5wE~zVjnx69sFLxF4U$NMf_nM8v z`WqF2vMiGU!P4d{xunyB4|^v1kb39J^0EvLhBB-I(B0M-8&IBPGL2$4o%M$scqC2y zKPmFm`PNsG&mc_Mcs;wz2wk}aZpHYVemG&%4vFz|KClJ^o|To;thvWoRySaL6CV*wT!HO2eJLAvo@|I%`xO3MWpOdH7Yjzph+ce#ym z)H4ch3jiJCr5J%9-xyc1lL(M3gcS3vMZw|U+%hSICR2$K<-S<*{5J7mD=&m4T{!3m z#Uf5ne)^(AK^vWVTVd6dWa>>c`vPtu9(jw?GE2H{^6RW(n62Q+Szj>xq-4rBDVRR18Co;O~G5l!9UFt|Ap1ejno<5c!xf9(A49X}AQ8d*oV?3t8i{V=@jvq>rE6VbVxSza%9 z(p7BKF%r<91!WEFhM)1$mF+BjhbdyI#)?F2G&Qo}&SN~L9gIJ4v$P}>YrJO6L?iDD zW!?CDCJ!P8U9)n8?$UeXE2KJJWwy?|LV!v}%OtS1@eda|IgVJBWuo{2P;g1G)F6QOJf-@9q_7jg0o6+xwqL0dSMh9mqL;m zg^sSwttt0i6r$nAc^}SpZ^*k_1b>uP1H_VnhpiB}mdYRgP=mNn3!G*-0*Y21$zKR>Xl2T$$&z zUZ$TLyLq1eIfd`&$kzgmNgLklB2o%^4lHS`?+0-8ER0auBXwaJFy81rIdg8AWbZ|a zmccX`RGOCk*MqOY~a+^ zQf@4nSm;{V)(#Sq2t@t!rK6Efle(iL?H3CSCQ4Uv$B6BH&mTls{?hhJ9mLy z$X?NIWxi*K0+2z2w!hfPAGi&`stzEVfTt-4q*g)6?=bH){ti8q^jwnDZ4N_0Ex2?A5*A1s-;S?#0dJqn5jok2-9n;cSk!b#M>)Op{n z6#uX%9dFi(w$|1@{&%n+bt<^_4_^;}dcI}^w;W5~6YaLvwZ)#f8_qcn&or*J&&W6YBLHqv2Sg9k|p>Ai&K+b(Q70(k+PK@kT~st?bbE6-eRkdeMKmK5R?uY zS(QoRI@0g&@+E`A4Oyg10lT1ifselrJpHkiqi!&E=>1J}x+w@uTvE*t7-#O86gJNe z)5I6#>$^%Tp5~%#&I_X28i7kFD!;PE>pu2rvK0KJnOD5yo1{=TcIw%C=HwTV$Yr)8 zz!LO}v@?H!ojOXx%KT_&F!djkbW#`|H@eS&DRhgG>xXMQoHuDC9g|81T zx87Cd$aAJt5#Bt~!m)%O_L%AV)}{<=O8Dt6N0eG8D;=%f^u|a28far8m6xNlIt_q7 zF0^YA)jk!kf=r3$Sx!$FSnglgs1S#kK96%Uq0ikJ@V+$qG37^?-st{Fz)n@)mL#n} zr^LeCreAl&sisFCCnunh<5W%$#S9^)VgnE&FsZ-@E=8JXrRL z$K{2V&%K`}iVM9q?#>Cbqa68?Z$(&Nq^%}{I!i5y5(A*j(I-`m^Suvy^hZ#((s6Pc znV;G8a`3duSU&9G5^?8}v~PPze^V_ktn^e}s+%{SV_lC+3^wQ}Iv%jtDlpX4c%+w| zz?E8C7|Zd>Xz*|H$0rB8Lw6`N@Q@od@M;cY7(21DJTdaH#Nk`=SeT5f5Uk&KUx+&~ za0F2i;5(K|N%+&zQ<3nOYu=*zg@Uljld3GZDHFpYSNC70m|W`>Dx7HXB^1 z(%a@PPM7y8-dFL7|o@%29DH9d#_ttFk$#rA3 z7Yvx+L9TH_TPae16|LPqll+#1-(l`^Fv>IkYr^V=w9V4jdz~7$%9qRQi>)HSfGqyhAu5g_TE5(&Aq&DjDlT zNfvZVA$+jQXx%|rpZ0vHo{~zxnMeUh9`n&`b`vEl{DhzR5NiZDy0OOMOCfL7Ktdv8 zVW%XF;xDk!*sNz}*MZ#b;<@~^8DxIXO5b6Cf4!hu+ z0?5Vx%>q0U?XC?FX0?qIv~8D4H+&AyLrh`O3J0t@(H1{)o_)+}dTC6gVURgzZEaOu zP1@K=x8dsCY+rpjn$Fp&ks&y`u8G|kNd_VrPX7Z@tc#)qGRDow0W`;uEXvzM&hI`N zD75!`!gw|D@+Y>jE2O#`j{nkK9+dT5=R0C@{s|X{)Ou z#|L;OJ`=bIkCfc}%wv9QNH;FgmnO@Id`rQLYpjEFT3l(Y)}_%@G#sb^iKEJYSNSej z+?;>RFBWfxk9hLxSb|sRMIeLtIrcQ>RrE1pJ%~|M039g+t8HEM$+Z63V`WXO z?=jD9`sFnKVFSm>^G1SK@E%S(>p68U=0)^rs3LUj#h*>)tjhHXgMc@fM!tT7BM!px zF`0fTR(eZ{lid~OKL5=lB%g=d#V>&Dk}#un6pT62D&U;hW0=g%-rW0VlLrlLth&V` z@TAPkYNWo3CGpQ_rNj_%k4ytP#Aot-UJ zvOlt)W2X9T2i5&K} z;BF!rtE;>c^!O#dC6VZbf8Vd8F#WUSyBvxUP5}S)pm2-Di9a}cz7|ICMNouF4>u4c zV=9H>A@2e#c1%uPSbCi2z`0tC9-`DZiw(!GuzlHIy&y@yIjQ1;7HMa3&RaL1^=;K7 z`e6e!&LFf2yJH3O^=o9DW-p9luq)p=PoEOAQO}`Bvqxdb<;a3)lHR8T&lFB)ZF7l! z+fY>_z4c|Fc%BRuz*V)^9ZLa`kcqox$AdrO6O~23@avCId#A7uqn54Qzj30 zPBhTIn!{f9)*cK8K_!*qB+U^VY+`wF3C2^U7URlEV~SHL`PofD;d@_(S>+P) z15uFwAsYwFr&5Bl1CcDQ;UhoraYTfk=m(2QmBy+Tnar8CEc`k$aeVv+<_lR|66{Qx z69k%VH6*KjP?b+$FeSIw=G%t=rXj&szc~eOxvc#yPB8t(svw?&Drw6sgefD<`BJ~5O~r>bbEucY~EYP})~ zuhNZqC8@L$F5W3yW$P3i%%>pbeTN!msv8f{t?U3@y7KV5r>;IbTAksWia6o``u1i) zCFr0+tgoJOBb4*WR`(ri;QXm1T--kE{+YAmEEEGdX!* z;7loa2YA1Niy7M{Hizh}tcd%G6}DNnAt;y)28!h7D!E{{Hp5>5IW<6Lgl*TT^9Xum zxSE#RA3v$CWoqs%%upy{MzjjJ_=$+b6XN2UDoikgtF23IWIwY89XoT(Y}jyzP-#Um zX${elfO1AzizIUw=9Y%Ks-)I3+IkNnxC&c6leckf9nBtC`GQRr{<26)HNV%TH)kVv zG@i&V-kO*DQ3d~Cqp0q`<@6Hu`Xb?iof%#1=FlyTt-; z)}-3Ec9XM|#i_p;(}zSE2ag$bXDH>V_O8MNo!Jl`WhB!xIKGlQ!%CEMC+pD! zDjG}ZlDjC4oX@=$B@vILVSf`P3vxX{pUS|&jS-}}*|D@J_wJl5B50lD=z;bNN|%0- zE~S(c48nB{z%`fFVMbOG9_uvv!lfHFdeV5p7<9tfrd&S-H5CWdv$P zXPE{J$yghDm1kTxxd1h&QgD9!P?jxk2&M0w{JRXfC8N1M z`2syrS5DDg6!-2BA%R1il{wx<)VRINmF$K1Ej&Pt78eD^Z=%ub_dYEC!rL3md~cDo zPW7_pTI=(`=ic$cHe}lI`*0Z31!W9f+I1tw_&k+H3mUCHf2^?om)a(>rk-bDuAIo_ zCaT)Q&R+nK)n9og)?zP@3HT<~Hfe59q8#U@u%345$eHUokU=5)1 z)nSbd16Ddg^~v!u;a2AcyKRj2#?Fqhyc~n0!2|68C8ML6Q|zXQoK6XWLq# zSF)`voO^mLg%sxA~#e|37J^(hx#7z7u4CK+|*KwKw*9cEN`6BPtAQyDNPjz_IsiwXK~|Ir+?v{-MixZoxWo3wJ+P$~|s&_VMUwMnFAKhv7vJ{5#f~JcLjf{ll{S zU>O&6d`Pr2*h>#MN+6GG66g6$Mdw|XfjygQSA(#!Hc)~I-#G;omSW#pm9i~$kCLG% z`xmB6*K;#YGS0X{?*52(vASOvBbc){hd`Lw*54eh^fY9$;-2${LBLOeE9lfGo&&IA+W0 zLWgSbn)`*qj0S5!b{HR2I#yTJo);i@(q?r62(jA@X=l``rnam)hSkz>7l2LWaj*TO z|KJT`cXD?Y<1c|ksvZzpR+p999vH}gh0?$Pm0lVEG%07pb#+}%R+%>3@jqf?iC5}n zkC2Rc+pDO`I!j;FuNkG|9H6L?{zM!T{M1vJK=E)lSzHqw>)ND~Y!J%1?^~aKyw?49e@HClb4OM;FJv^a zfRQKXb!TWm*fhgUBtX({#4j&p7(Vt@`pGUlk%@6qGpARXd+_o2NKH#KlxMz{=pIwNrzvFvI zhCDtAbaSO7^kSG-1LPP~f0s=t0;j_8PkT->s2z53@&H!Lm*h>1CV>Vja} ztag{ABn;FL5~b#yUWgnr3Q&6Pdf_K!Df{fw%*w}*|@|WB(ZY!QZ+!KXfJWA@!vn4cq^mzPTT>LByQJxqDg`? z%(mxrXo&)k;{gdmLB)mptJ^U60gUM2cnDzbMERxgd~jF`Hal@LNdI@!j)HTpl`fA}}r@>^mQo&PveVn66ERf^?VO5OGn)hdBSRh(C#eJ-t)!hD{d4|4yCEz7@s5kV?JjKeQrD3GF1qHU!s2Sf zX!8{=_R#k^;|~in{G;{JO7+fBm3fAt)?Wjxj&O}E!JpEgd_{vsNVrL><5dos@@+FK zc+bhO%Ib%v)eu@2AoK)JxdygLKP2c0t_W0`@^1Yi%1v?D!LUI{7I z@13kZH&O)NX2 z!x5F8U?d7YyApFP3ycbAr(ys3KwPQqk^TgXU4jw^2ubXX*Z2n0az~>> zWZ){BgLfDq!HjlhFn$Dtf`w_e(VIC&fDR<2vb=*v(~*Lb1}n=9e}Qmp{s`P997}X$ zpIoj_Uu{t-6F-GMzyMYuuzhk5Zsp-02#)?P&BoB^%p&xZ-v;U3Dz!c^FnuMLpk%T) zc;Wm{C0HDwC*lx0IGhs!b>9Bp6DnFpN7rOi8ewYFI3N)TQ`<+o!rr@}%2ww1Brs&O zm0Ew9SSU!t8X#Ewy8;}q`M9>XFVH4{3O%91={_Y6oxR3SR=gyxO}t1SMF?Sdz~H9J zDtKIXd8ioq_d0u>Z(`OrR+~cpOSfJRz!=-^__rhppQC{-&<$RdV35Na;nZ$QC$zCf5^^66~q8&J`K3@!q# zF9FxI;=Yy{9q+-#prcl{roq@ro|{jyRdr<{)sXOC`p6r~VmNIICEJOLKRk~|i;cns z;0#;&8e~`NsQz=JSD#ZJ2qw5)uLahiZC&;4=0Gk|57l7~vTueB*bj(>?^LOuU8q2- zMU!&kW&Y!F*H>Ux;wRyo99^>Y`*gz^*M%8gY4cXNjx|9o8@#oDP<_)06x%D=nw(pH+nX#?(=#kvrx=*{ zEn?cTyO@e?-}%9w3^;-OvFE**p>^jJO*GYijcCjLZ!I747ufg*PP^Q0qUft_cSt=# za2ADH2mN9rj)`@QCOY|B9`YAwxrXL$-p=jw7MTqzL9Lt~Ha%G~g!#p(Qw>Bsm~5}2 zt3Aw;;Br0^;x-NXZHnD|T0Ay23KPK`A%y9Ea$m;B? ziMC@)wOpen*?GN6O4ir>4ZatyOkFD(PtO!I;tZZi-@Bz}h zZ*$U;@Lxx7m){6sR%4X&zKpxzJ(-B^b7IxmOaI;NYqb3t0a}!

^iI7;BIj!=*x zKtT$f8B}V1xH0yEw3ga5d=Rv(qU}D|Boo8aa_1W#aSL&M;k2F5+Fuw&UCV6m->ixt z#Le8^_=n;VLiL~DzhuRh7l!o*66uSVLmF>vx=vkyW0Ng#m$Db*oqM+I^Lu~U=d;z( z!L8_-_c$Lg6(CSADfalyPbHd^%}=J`xUJ?x-T^(>BbbM~s&lBXMMc-`$4T97QTt88 z@a{l*>(eh+rnT*bAdl7-yDSjSnZ3g~W5E|&$E|L;wbT|MUEKiR7v?_k<_HX?25e1o zngnv_!qVgf&>;Y_HPHE0MlSo`E^jz^y^Lk$B&*2oC zJR|?c!m)t-xDci1vtwM0PxeWl5cB!S>lkouL$LYR@B37*qSFXI!p+a95~f7e9!{xa zZPN++4>Zd0SXAHz4tAOAM5;np!S2>uyt#l1MOz=~>)lUI+2Gh4&PA%An%TfOLydA#Y)kYxMlya%TPR z&GV*dN=&<`7wmoAeCGM-miF4eEMx0|&&ExR@1ivL&&k^xgM>^LeG{{@tHNf7OvZH@ z<+dN!WA8T*pt?^2V_8pVyW{GuUN2t=pdmCeP8iJ9ceCVuCc$x{nAmdGrRIQ@Wy!S6 z`(bgCqd&SF$;CyXI!9t=d=07IHLAMK@t2&n*MH=YcUo3TIc2b~3Ky8k>B|>WYql~+Pc9TyCz2#)eHt0&#xkZtwrKK>dRAc~+v+^*-iZN*^W5AH ziEusY-0Uu5%!xxKgoJIqK<*OacMJl58E4&9HRrszo1xlWOcD*gyIe%IEGD=aVUf=v zQzTvnD9Xc=MV5pajpMN__UUBt^E|v>@eTPluQ+TXT;y;Zx+AShK zF0j#i?6rp9(bc{NY)or!X6LJcYlp!`gpbkFatF)f(f>xE{MI=W52r`WyvLqK;vq?4 z-RHkh*ob-f7>cb^w1bmF*l>^sfC932O9+7#-jb`kw?(a2q0>N!RoNxG?%O(A?NLe; zlsiBSfYZ4=#>X?2QDCyv)I_2H-OB7v~2;e%8)34^e^dpF#d5Z4b zd_1%f%mJjYyq0@sSl%O{u4iZ1$6KB@NDQzk?ixr4fams}FZgGp^p22+cP$?V7H4M1 z4z<_>1%Nxk1`GqaSz3fSCKDO1Jcj3+cKmsb@G&_YcpbTOeT~=z5cj|Qz5YjK@X7m| z`{T#Ma<9>*5n?-s7Z$2LA|;<|(AFoP+-G1g3P@q28?IO4p7ke!jI5`i(Z>!kADE0h zqrh(G3~D{9j-Kk}w--@9>_u+?0{xtB4JZ9xVKNvo9b!$_NF3^dxef<3rcCV<2SLET z?*f-fCW@M`1PgG8Kwi6Ehvhz!yL|(#Y|a@10X0@YhYMdJLXC8G4L^Sl@$c}h*~u~9 z=1w;%zKwV2{XdzwGarS+#fo32p5u#j3Owx28P;CYD(tuD(z`HC8vfwFQ{KQM2$O?L zP9MRJt20bO+0T6nwLTXF%^gpeWm(T}57ux9q0Qq|@Xisw`c=(U zLGvvi78m)t2-@P0kJC1<*DhPvJTkw+)f^LP=&j$8KKA1Pp6p0{Ig51M34AH%oG}6q zuiM9yu6cHWR|bLgrzpR(*Nh~ESM-A$F2Xdo(M&%}0OwEv)>`}TR8EY)fk{DLWO0AFb8(Bh@+3nLPUqmDI zX!{Pe=vqEda`QT-#^apYS?E+|V%7lF)-eXt@e?=8t?%|_w!9W(a0n>Onh;^&2CC?| z_x?9Z#OfHdN_J^=^z{$vw7fZMfC9E2Heh&-!c&A%*4N&i&-IFpGpv$$?$Qj8?Mu!J zdw6oE*4cLsBOSEiNQ$ZGowKY5lDcmVhn7=DH+6FCp?EG}axwa+i zm)ng1-1-w{PsbfWo9;_ul7dgn^;&}pqtIUGj`3VA3PvULJT>ZWhFc6h@mHpS`Wfzk zX$2tA2)8%Ye|~EJ(eX$TyG-nOI{cWf1sqVGM0%rL7I37F?wgFJU(S@8oMLKl&Kw@L zzREMMY{5n;Tw=zn>ugi%tXKbYJMzGblU)6=2P5g)?vguuRODGN;imu&59vM&$v(qg zkzx1$CfM6=34kmXZhu7GOncWjPVe0r)McSTU&X0&?K*{*by_|hrPnqC!#8ao(>~S1r0A z*Ii8c38CGYpn`NANLI1TL}Al~Ka`Hitnr*?>SW?S{Q})JO;2wbc!~XJWpiSvN-81D zCdL;}D(LV4JQk-+$ObghYjQjS*&=QWJ8no01DGTQA1M}Nd?Ic(EQ@VhD-nkIvB%*4 z13Iv|r<%Vn5!LSj8jiS9x%$c)=Qq0#9#`Hs;D`1n;C{RW%}sbqT%GO{rniqjjN5w- zx$svJw*Q0zQ)^LWZO4ZMrmus9Ayi4+tX`=(`xWPNJJ@J3x_iKY2+(!oW^d}cbQ>nkMT_Nj$etHeXj{E%*ATu^n5 zZvTnhC;%4U12RHE(xOm(XZ<*m^!C@lNVK{9|I!-a`gGn%_sK8Q_C@@a=_M+f=@oB# z^HX}$oxh0TEjK^K;o$u;VMg1?vg0R+$LU)M(=!Li!q0sYf(BaoE`421P93e(T3^9zuH3vn zA+K4*x8W9iWA$Dhk&g~HtE;NJ>BGAYs*A?zc1`JUutlZmOQ?8oiH0cvewi+tyO>XQ6#oza$sF;PqrSVnx%@o~zVt=hSa#+xFMY4Wb%z zaV4xmW}0n!N;p2OB#Hy8d=3{@$g!nl*$Xo6%_#>SCUucdm^qg+A!*4QIt~pt1pS0h z3u4T4s9}z??#4wk{j#N#*HuyV@p}oJy@$}I^)tNr@m=vKWEKi5oY_`d8;UAsT%kjO zsJM`r;gL=@i=KvBveTz{NhLTF-HLzI{lb3n-9V5pXU7U9+61J?ekMU~*Dfg~3Fi7^ zjn>WgA(R<~%*qwx_i2}mv4gWKhqhMH@s4I9(X9nsGOeZ^osH95KmL$asdKBa*h&(u z&JX&y#0AEV#lN8=y9|M5kFlC2L%5AHr-PVSAf_SQCW%0|cNLUmHpZyOGJ}Q^kauY! zIg1$Nn;Bq|W8{cEE1mai!!E#e!`NlN9T1A-5>OO6zd8FrN90Ti91mkcgF{hvl%*)*#&HH7_D zATz_;G--p2vG}M?H>OomZPzy5kE;6V1?ue&4nZ`>6J8LMH9Pu7l=wBBM$zFLOZ@7l zay@)V6%Y5|t~v^AizyBMU40noB1!a~u<{oi>DiU3pGa)nVW+92?Y8J)=5fF4zw=v+ z5e}8?=k)%l7pJPB_;FY*gcxXrE>qpj0X50cCHA+M_;#92fCp2ThdT@sY7{$1aPS9epl}<}>0UR_4DC zFO41OO~HXF^E4`&Z2ZWJy~NhmqUW;vBO3#DXAi|k&v~*8EgV8o* zcmbDmm&}}Zp0-vLKO=MrA@mrZq?gJ$IH5O*)pF(#|+I z!+hLjijIG3Pf%|iL9;;$ZLw)l;;jVDu+n+Qs_NA3AZO2X&B5^$6_>vJDj|j#vU^pn z@+pU!F3~`@@W6J}>VZsU=1c6wcqU0=uOEkhMT~A#MYX5lkHpWFZnyu`+E=ag=COY4 zVuhM@MHT$E+I4$fGeioEiATMaHpF1$ZdDDdgt} zR@jYC#fvOZ8oU07rmGB!>Tlc6Ur<1$Lt$x_F6jn=rAxX&x?@4Qq`N_C0qO2;>F#b= zx=Xs=Mbi!;o)XV34xuROx7Iyixd z=>&+S5$20g8|Qy)Y66>7eq2=*$!vI*khbs#tDHC}I(YVttR0Y^YBx71lju;E|) z*(;%M{Li>jQ5QZgX)zX^OY<4Z{Qs`BwnAB74E9xBTmb&tQr*(#Z zz!;|N!3C-QMelEVsA}X0yVYVB>gcej4AW@h{H0Q0%!<&ODs25ae0@1CLCYf$+J>d- zC-su%Mzxa%hJOc^OY2-w`$AE>NoTPOB`_{p8TU9$&Y%Vygi4Nb7J#xyigN|kLQAqr zMPd1iWq?oFYL#4h$QGB)=G}Lw7Yzyp6*O}A{6|F`+vt84i$Uk%cz0pQ$ zoSYcVlP046aL64KH*9Nxr43fJMn*MKPEJMzpixTEUUXJivV^FrQ>xSi#=VS#8Dt=$ zO^8_RajurC`%|@iPssx5QyAZQVLe7J6w4@;y6*K9^i%Rcn>yDfVK`AdYM1L=(E&xf zchDL|X5+ike_SdLTyqKL(}^>G%qHp+`=pttxn}M=ed|tQ&gE zV+U|vG!wjYMMJulY(pK>wJ#`0SZv&>szZ5Ls9Z-3l8oWN_RRY1BYzKoRFn`|A8T+& zeNy;N*|CBrsuuYtyT@N0F$!4KO?uN&g{xAv1|L)0g1E2v;q{77pvQ~iDu*iB@g}(Q zN|+Xp`bu9Noj^vNxaVQyie>diMcwAIai@gMX7kXWgDEPsT-+LoU%ehvvNr|Wjx;$D zA8@t!kD5ciD8#bcCzVc8RteyMwD#aL#j5G){3lg!H_`kZZ6I}*pd=q#iGGSYfVX-F(^$z;8Yw%Pn zZ#%n~x$by}y!f8-7fCtS~JCv2*x={-?@xDH<2?H21I9_F~S6a(lj zR(*if0ug5ar-S$W#aid8pbKS}t%5rE^h`HuVxZNmIItue)oix4>SB+aR_}^V5CNxQ z;`k46!3FK%AMz{(lRj0cQ<=4B4B;u_fJ7nIxZ+UmnuWYCPCTLPkuPU=m~B=nO_LLm z>N$xZR8Iaw2G`G8k$1rMrU`C~-*9piN<;&HaE-WnZmnoDM!3kAl;5ZjQ z3Sgqg7;U&%cBKp%m3y>x(4B#QH(GU4(AOgI#u7Olxoc9NX#-oSN8$bfI%q@TJen(i>a+w5hd&Pb-wqJ^i1$t@ z%#=TiPtK?tcTmkyw3uSnMmf~1b>=A{6kKUTm49IT#uv8Kqw?71N&FzB?6cQF1AD=1 zx?;KTxYp_q8aX!KmWf56;DfxGVx5VlK*jN4>4=cCwVx*})^%AtAA zwjr7$AzMJnCl%tiH7Hi^eXgi4ud<0wZLnTuIZXwhVeM&9<*rhy`lRahN(@pS$L`h$H6<&Dy(=$? zxZ~7Bu|1NQ*aZJ>LO*-3Z9QzcP)TjfW``E#hY#0@GtCpMdwLZq^)DKtN;sC>b4zu4 zRI`6V4#~6r)L$jk$#MCvju+X=32W{U-0FnrYZ|lBC-?-?HLi?a8kyBvO_R5>k|hK| z6kD+gy2;$|$(Fp8Jmv;S+(mk~xf0~O{0;x|pxiihT{|K+L~rE~Z-}=Mjq`#sr0Ibl zK!>bY?{~LMs%D@96g7=od2Q%ZQ{Ax)t+9kLopui^XCS1CR;5d#nq7WJeRFlgVRL;- zbm?!X_x_CPV2W6bY56vFJ=KRu1+f%vRzT-ET7f9P)R{C z2a(nqeO8?3Cf}m*)}k+v0o}9L#1Kt^IfEpsqzFgx4A`{;SjX_#NoP9bLHilvr#h6K zRX=;a3^A6;ICQ8K?a3N|Y)#T(v=;pUFwo;tJOXd_A~##r$dkRpPt`k zgoMiwR;F&?s_Ub@ht6mw>eXC=;Wn5T)_CI89Q>*1xka8#%o~O&?_Q}ule>#tzg1bU zadAJa$FtoeoJUEu0`}rqk&(JA9!TusiIFC*K{T63_cwXrtT4>x*Y=u8oX4+F&T*^> z_JJJ-ioLMEz>?W>E&qkxL@I`6VoaYCuF2QmLlQEQZ}y`)z@_rrmq zA^O7~KxRt9v64-Pvj4K6aX5QkVWu7VruSwE!#)MZhgJ@-XsNTd;9SSB$~nbW zOGem;HQZbrVekDi%GAf$@A{%F_Z+kI9g~$W)M0}aQXo%==ehu5AAb4l?YV`sfQNA} zp=4Q4OIEP?5d(@qxbeqT>L#<4!g9?XqWA#zX#0{ZlFWQT&(w#;JYv#Zx8#9xo^1s5 z0U|e%zCT{sqtl3WH_&#KZr=n?XECWo2CQ}|VmH2p;EqP6|8*^35%w_|e0(%yZPY1t za?+aPiF;HZAo79io+ISWz4ZhU!RRI&zLZ6*PBCx>QdbJom;gBSZ`UPli>xfHxv^56 z#q(Um{H=O~EaNrio$R^#1~FHwdXiUcf3@W4zfx(OKSZ%UqNw%4h=tXM_V&S+pP&?wcYQfyJ#|`bR{XT(d%s(g+b%R)L5W@E0}yAaIPT=+ z{@EeUF23N0vkuE|d-@?J9ce{U1yy(Df19CPbA0v!1z2;Iqpe*{_zW4P}-3R0YdJ-yQA>V(+0wDg~NpoUZ% zfYkyj^niRY(UkvChcG@fKYHSZj65>FME$a*O?C6hWu>psOs~-!bvsm?{CczR7m=)OEwH z9TB+HDgqdrZ0l*n+G}XN&S>W`R{gM&s|?ntejw(PD#jonOp@rFLOxPJ(hBrR_FHn% z*TsxKcbBa7&04bk8&(?g9a$+P+g>6a8aB9|O?QIIb&ItfPc~dKj$cf&J z{5>5z%R6BrYIB~hDUia=iBOW>Q2hs;wit^wrg#lKU6g+j4@_ldBkz!WiC=DaL*RYC z{%@@pK2(=iU}Iob{Svb+LgH5az0(~)UN4bXEsQa2sz_+t8;RI(jb*0~w%Qg9y)gmz zX?mu;v!_4ZFt-gash%YqpS9P@+VPa4Pa0Zo(lev_okx;o5MjTCi22JxgfFP6sNNx8 za`YuZ#JS(JTKm^vMk68GfDE^KmgoeJc;7iivE6Hc!+}}xp3iT?OHlPA3He!5Foa9$ zXK}REj(gPL(LF;^s&dolO6eJ8yBTipUv(A6SCQ_AOgrAP&&zb3>C-yJRnbOaQ|t>Q>z z$U%^N%{y(QUU&?r=$@1;jR=!z!hP==l<|VMKdN7kBC<$8{=MK^mq(n2V3_nSOhoFN z6^UX#trF7wH(3EOmMcG;JzbHd{ze^(ZTpO{Hk3eXEUI()!U#qKlw)LtEk`$8u35A= zm=z8~JGm7c#iIKb6W^?{{BTnxG%DvkW3b~PPtk;ZQ!%zsBm!^;C5PFo1g;=j`6wr! zcT97hM_u9N?C=t&jaB2V{AMKmIwIf)peumtG#i@nj8*P!pLw$uDP>l`9jm)J=$Wb$ zVpirWxy3EzRnnvRDU^d5W0^z3xF3gn^^3?VX}cU=xmukAc}LuBo`rR2OtmJ1Q4Njn zxYBUr>^|vQHKcREk*=UU&*8?S7|%P`e3vf{iI|*)5nh7dm9zF@iqi^65p9$HilWSo zIZQuoiM5|e68^XimjFztUg=q*rHP^ZV-r#sqh=QG$C5Z4MDzMP77yvlP&38ZPV3yu z=Tm?s)R{Bpr`{I%zmxwN5E)4_UV)dALe4TMWV!x*cxJYeMfy=wZP!+rL9Cv0Zxw<9 z2TPQxh0q4Zn57okhd%^G&m|Cof(zUIO_9Wm>AVvd%`CLEFw#iO23EleU;N&$!1i;2 zJ#xewv}omYZzu=K+*PU8P3>Fn%6mc==6oU18 zqddQ5OOuw(rnLpNS=cq~+dlWN>5(Qw3Stg4Pf$<@J45W$>rcNTyF#t|l&7L3F#4*n zc}kD2(WbqDKpMr)aoo;^+G1FAU(lZ~5k35YwPv9j%dGG*3n1bI=9I!bleEz(Af{e1 zAU!?+l#fY3g*+YYn3I);eShtls@R8wvUnUNZwl~!w9ts?c*i@|l87`^Dnkag+}jFe z0A}S}UcS~n&})tb_bmTHnc(Ho&)%AzdH-*`G!y=u((nM5GDa;bhMhL4*)a=nNnK+A zAU|?zFAZifaOm$^aKlyyVvpVBcepmyW<~<1I04#*VhI9w@rzlH^DFh+1O%ZxX+Jip zEOz!=^tOW126qP_oMG@q#!-|B0 zNY#^peh*R~l!odh>7h1|b{|$e+mX}!Nc~oMJ!Zj$qzdZ$ryk0rlDwFt(>oU8*)FLL ztyvSczU9Ac=iB}$B+l{xGd`kYA-Y1=E4v_Ds+qZenJx$r3r67i`jLRtiPfS zyZB@r#V?+lxnZrzc1FEjsY>qjmObFzg<7SM-P_e|=H+gYMJ%jiX)s2>m0n}Q7Jj); z#K0Hv)GMYy#|j$l2jAntJxSRs9rZ;X<+;xOqD1Q?<4UoZp1DfnHC!_X{(_9}t6wZA zTn1Cz#i#xd8QKgZHbg>o4>m*sQ&)0y?kt5RmNG@Cs<%U%x1WE;*w+OY46Iv4$-X(l#X^GYpM1_|rtT?!Zj} zfNwD`YNSs&RxjJ?6bLzDWeL*s<3K5iF1$KlHLV{_e_va&At z=Pt1)!6@K+wvlz)6_(y{-jK~iVjJm^&O(F#zBo-i&d2_U+SFYiy;$KD3>!D{fV@vb ztZF3!8>D^hGjQV}8*R%)EsYiMjF#Oau<)mc5Ksevu?rk2>hM$^wUv?BF4HJLowa9& zgXG!}>mo6XbVY8tSaU+CE=aOzUX(v~Q{$Vsa%hF_r;<;$ zOaHneYobsHVIAA7!0%~mPP@u;N82;!$v59}_2#3-A%f9!I+PV7W%iMQ(?8_i2<6*9 zcqEi`DDl1u9OR&fR*r4XTJM?1rV_F6v{Dc6S&MtP&(~fZ1Q$`3On-)tHChm($moS8NL`o+hQ`$Y}TfWf}N!_p_qhsA3UW_1uFbF8W6hC%}fAQ;2~u;YG4m$ zgsZH*m-?=+grXNI(8H@H30Cv{u&OPmyFxhIZ4zCfj6dW)9Zs&=n z-Lb`RS|F^8*QB|3OUkDI0A*0(y2&nzvEK zVB0bepE=S0YaMo$Eey~hfL9k%vqshEgfmV+PklK{D9dD20*-wB81|3$Wt!nQmN#Oa z@gn$Y(E}-@h!)WuN9skn3$32sHPhFRmiWReJaWL6DfJEBMiT(G?#Ch%;*NYh`CX1n zMVOxp#ql+=&4cHlm6(wIUm$J ze1yd>oJWFt3WY&@?}$H$F3C77y*_&^4Kw2hzWkgc^K z4)q2ME{3cVY{UjRjO;XUsTZEqJQhnnuJN47e4ydy?t-ecD-%Y-AsVx2hH#rWNWCEfko1P+Q;mWjS6=5he-aX9lv z^Ax1k`(4u(Q{Wae;P*#2IuZc)O$Jw79(~pVY5N1iuveIdT3O6whevIUXa}EFMi+$eeC%LKeKrM@oLKjMj5jb@GsOXgC~xO z#DU8s8!|F$l=qt_KfiMZ1~k2C&83CWo0FG~H0lmI%cNJz~W|-?=14 zGGDQKe%2pDlOpEpcDy64PqwO~0`KB)Q$!D>yUO=1D3i-xxfVS%D$a*{lqG_koLdi1 zkok016&%)i2isgEL5&h|&uFZ}@4KS#Z8;)0mbZIkW`osr(Y>$Rke4MwS!5~L#xd+N zC@-Bh+Ff!2rzr^lL9#%U**#aN!ucM1AyX@;qCR>i_R%k;B0_JB5t~+IVt9M&lxtkr!w{;`MhLT%m45R z&{x8z3yxSsE+`i18{YVBo0Kk1RpxfC?GD(aeNnqy)%w+4{9iu zL>sx;q0w>S3&bg7L6xrc0n`^_;II!aG=rF3xuCJ(f1Dx#xL3bIFd;2&M2bBDikSb;O{9lmn^bKvx?CkF!I@O(8R zU9bR3FRa*nw>cl=>m33}ql+6cfNGJqn8xm#0mo5zf<-3TJNC;nIo7mIcA1r=O?lX< z9ytK3B^=`N+g!~7aWp(2F{a}09Zxe6=li=1t{Y5 zdk(-N*C{*fYU>LqG>Vvc;ZtGol$nLQ;K&J5kMWL|iPKB4mADFtj|Ig@zI3bq(YF(` zm(&xMe8gDXineh7tFWa1)S)u=`t@ai4PIu2tWABP#?z=|yDckBuRm~f$1gT0Xn~(3 zjoQOASJ-lku5$YtzErnPlk$kxbd_Z_GlvFkS{XyX=;|=ClQJSZ>gMv70XbEnm&j+}FMB+@X8-?Q z0E{(3ahO|%@gvsd8ZMU22grW^#r%>k32R%z>IF|Zl1lQQ|J>dJ0D$#wZs4=`Hr*b@ z)H~gdBz1p(z=Iu2a5^{9qX@3Hyr+@sJYD$sFPfrc$Ux7oIWOkGG}8my-O;K#3MqD6 z(vRN0O06#0ZAZ0%JKN=Ef7F|dLB@Kp%^4|lLZzKWjAk90*? zUj6(Q3vzjMbJj-R(SMR&=1D2oHO@%g;)#R}%-Yj+qzvQgA~r;k)~r+cJn)z_bWi}3 zDboGfmJ(|Sm>bEpi}hp0r+9yZNdQ|Fsy%xpTbiWA4|{*XEmbSe+NR?_@b@Ii|w zb+u>k$rKQbj`fwOw{#y~A`u|?cmw!e!lNhpMH&yGT?Z9XBwn($HYh-)g8IGI0tLNk zbZPR|TRGR_5yT@{NF=so*$2R${%1JbXo(nVfm^LG73Xar50vC1!;&^f;OjMtPJleN zOaHhf_a;bc+nbJHU5gz2F5?QQBNJ;}DtmtI&`L*O2g#_oco*zU`b+~zOo_c7W85YR zlo0fW^u|7c?)XJ}^)yj}*{m*XE3sw;L=QIzQo7MY8W=IxSfsvA|L>FoZzb#p;jp43 ze?=(UTLPwpdT3F#KYil66)T32CF%iu580E;XccBJg;DseG3u&VA!s|K73NjVP#@HL z2_$E)_@tX`ZnRlM-P`hldH0L2xkqWQ(L=o?iV+>8Yt7;s!-#8iNb2;mIy5KS#u>lH zl@kc1%eLIhP@AsEhFW}nd!H*1oHqZ|6g>Y)v&G*qc|7Du z8}Q>{X>!kqwNV&LEb@|%pW)^BaO+{<3!z+fi-v9X8>0!58g2ZC0?e{DHOg!hW}!2@ zAp0Trx;&Qc@6W%!9EP&ffXC=ql^r?GI=q0LIZ7qP3JR1Wlq~6L zsY>GRr_-c}9whMgGkkgrI1eaXjXl2Ms^U?&bx@70&F~f8n>?ILJsmv09HoWNc>Qb1 z4E$D+J;}R+fET~>gw9_69CfpnS<&yO&NMM4=UCL5meXVdOfk3i%wU}*o1mOm$X}*- z{hRF%N-*@0I%e!C!M;RG0TODwDjC=inj%&UW$1Cd36#GE2C5@MtmDHT&*uh7q7&*&EY1{*(kW=ARH~;lgtkT@(}rLOP$~v zw-7+80KP?`)}yaqU=jecP1(or?`o8MCOl8k7_?^r)Q+W2zllinI@Wb0}Mq{3;`4xP=H6kO9hNF`9PRM(xk3-OZ6bjUu^jk>W$a>}%ALj;y9Dq+Vs_kYmCvdubk$KGP~Wnw_$* zEN20Il|p$Kz6Qwv>6sCFLRc+%6`P%k2Ap?_BA0Z4J$n*7LMu*p;a2B7ach67MyLR1 z9q2y&w!Hr|bNd3RdjR3Gt&^R#NJIb@)-IWCp*y#F`5vHwZF^5vl+a1*m|_g|DaP-U zvp4r}xHFZw|3Oc!qTE(!C$;*E!Xj{W zXX(KWMB#18?DgS>i-cHXc*3837C*o6R)b~#n8woI*{TObJVTpwD(2#6Y9uGOm1kUZ z*{U3_&x|7l#jG>-#JBzze6{S#)tt}<#f6R7QD|^Ud!7*hZu4N8n=r7M^__$S0Vr*V zB-@95oe+$wbi}&f@N($Oh!eVrAl#z>0<(er7KLmReCXrHt~n`@)c$k3A6U8gWVI$R z`&BSY-`j1k+W`^}`MwD+!c?zqcl>JSN%dT$2UZ~Ar}nnV<8Eqq~< zW-3fi5xXy$|737)7y5QY``0+wjyw+O(aj=oZbltebGu*3eFbY?*>>kxJgE z=1bx;O5Td~v6!hond3G~-Ut0ymLEy>-QfS0{wr5ag9774R}qNKqb0v>&Xc;>lj5iL zMk@$#kFvIJTxLpNA(K<;J=d#!fLQ&ehj3PqaN!&P>j)~a&#gpxo%7DUdF#+e9LM(( zFw^POqsz83h33%I*VW{J(0`f6^DAfbhcLaZ7%x*IL8GEgZ_AvppGjhIUZ#2D6%rsx zf1BP*0Q-dWM6<7BaS$ktW#-W>5@l^{1-xfc|AD!12r11Z9paXVupmych;_YsKbcRN z`-{V{%q8}Eg?iFEGa$nw_gV-H^cI?o#{UJ$OZ+n5!m%5$7)hvhYv?)CQQzQ$5LwZv zaIs54`QLZxXQE%khI3xU0XrH9oMYE+V?xGwo6F;%T3+gdVQb|2r z(^n>ua+lO&2wqQ&2&ldufq$FC03NHajn|5G?IsX4GhY9XD}j@o*Ni6Z9&h%t!u<3} z`0Ck5sqT&_HV7rkWcD(PJ?#XJxP}(CErb9*R#3sPCCITqee2}nlXrd4QK)|<{m4l3Z>$7mXim^_Ryjo*C$egq6C4{=WH-Nxik7AWVUg+lc_2dpTD?B3@S?lrk{ttyizaF`ZAn90@bR*Qys!o?MCg~!B zAt^nU3A%w`L8j*<0KP6$;=YIZg?>sH_4GG9qy#EJPig5oG}DeMlOvyaA_<4opkMVm zgNu5PrBo`pggYfddtvkkUspI}`p%Vy_Pili=dZ$!ubKbq;(Zf9{OjcBRHIXNvSsd# zlsU*jIAp#nQj`L6dJb@5jN?2H2+PwiY-;6jId6fLhs9cIUx24h`)^czF*bIL*H&Bt^eAwAF`o%AN@U6@xHd4*heF(-`p z$N-KCBnW>Dudarb81%KW#eMF_Vc?93S1pnjRv=$2w6^t{O79r3YGE)IYMmWWIQ z4H0`)Z!6?_|;>y9SN8r~R)i9{BAW>ZaV=Ti{*f!ZxBI*VsL& zx?Db~Ug;5Q!N9;m1kVGGCo7~}+DX06Y4OL*m@*z~?V14%UVw7+Su{736tV}dTuf}YuW3(8yKaNH zFaqZ$tnb=qqVf=~Vx{g)KE&SLgvI!XjcXk)yLl30{<8HeU|dNiCb^F`**nPc*^C|C zyWrsl=sogNceaZMKx4H>MO|rZVDz)ls%&h)hUl+OJ6Ca{a9W`_fcfM z;-GzY39V+pnBW~zHR0+H(+0fv-U56&{shi|2Mf4|+8|Ug-}MPM&!~b1-%Dz9|APY( zN&Mt%h{~e?yz~g-v?^YhDPgPu;0)6JxiZdl#W&dw@+_Z}i5uaSl4YI*wVpA1+U3Al zYc$;EIai%U34;om9yBNeUjS>nBrVxDBkEzd%)@1&toB>#C8XC}%L8h_tHLC}Ys~&} z?&eIdZ`|NeH|IORtpNWcpd>+zweEuyYC+7Lh;a7c{YJYF>_C(9bF0v z6twGq78Iu;T)w}NymI5>4JO)5@x)N>xz++f>rHUB#%W=mQU`zMab&&3{JbLn{K5q4 z#oT(ow(lnH3Nz?BzY|YA z#bs{^gl*V#ly4l_JNy-tIgS+y_e6?t?PkAVJ}hse`2c3Nrvh4UEZJo>l4XvNG&9|d>@g`spjYEEify4q)mJOt zdLr1=t*_X0?sBO9nrnDK4fNMB4fNoab%~;vyn+#ors}SEXJOR~pHu^)tAn@K_?Xmh zpXQCD1R+Z`vgZBVYu9Z$MPJ(fE6YhCh%)1H%xc6IKKgu_<)KpX0-4Tv%{G1hv`vAI z%r)wuX|WkU;=U$FV-OEcV|GYAx5F1V)unPd=DEk#k3DXbNVwg!s>r?iz zO5{Qbpz}Cc%`8vs2(B|47gj&WPxP|PcR{HD7n@%7`R#V6AdS~SR_zf5G;slgtWqj& zWI2mk`-$ha>taw~;RKy~^kG9){`>a0>9pc(`6R z&L)ddn{{C1qXE;Y-)YM8Xy*Ack{z$dc&8(c_8eu!t@m8p_jAF;C}8H^iyRvt5oM2n zXqtbWA0>+|u?H0{`j*`S8?&J?6iUs26|h?MR@xF4WlqSMZ#YC%4Oo%5y$%@^M9-S^Ci}V za#mX!rhHt*j~A12*ozJFwO$B;CYV*G$&P_q0Ku~_YoM^g7oK7dkQDe??U9kt=>fi8 zCG$EbE4m5pxm^L;`M9t3Jh`YwAu~pev#482Tt9=Cd{x-wf7)j3@BPL#(xq;8iTu9x zalZI3unGQT{RH1Uf-AcP;^sZ4n^YF$zV?Gd>z8v;*C>qc{zJK0M*Jb~SW};*>_+67 zKG~yQWGN_Bh^01mLFtL3`$uB(ttW=&pNj~0u$*YXZD|b4WdJ--zfkX)IwdDM^q?(L zzn`D@I!A}$WXD_!l-bMa09NXC(Ol(DTspT(rv!j0@yh3$z%5c9+be0WGrHo|mrTmJ zs(HT9^eGKzH!YOr`*>^hEAeznie&w z2B9qD2X(q#?*i9dbykT?tOcI_Syt>!>iP3}*642U!DB2jhXv}ulDt)#$Qtm%oD^sY zZ1FvQAVaKmS-eTlU|T6R%F*ovHfXW)!}&u2YZ_1HbXaYt{EzWBy^iCB+OJ57^c^w$ zjl$pO!SZ!@?UE&1_LB8~9Q#x^l&mPf@rl-&q%W*32Y}R2CK%z%g24*?AN9)O2ekWQ z*?BRFtQ%Et;X0Guqe^UWOVD6jLy`6q_AERT2!7+I4l*SxZQupzJxyan7b~1=$irqO z>_O@xQ+MDxKYCnV@aQw*p?K(#CY8}Xnx;Pxwz)yVPKXtEQM*t}AljUvXV`H($|tvj zL_X4U0~CfM;dk{P9qY3F$FV-}e8w4O(j0JPF8*rYe^a5vekD+6<@2rWFS>&`aqNuG z2eJ7#m!?Sb^s=5w(>)TCVy#6ZyDeRKD>G`IOA?43=@y~fS9H)_RYOJ~o5w?*lE(sM zVy5$(q*T4mH)^I`WzIkUMDbuDoyG_~Q{c*UC|2Lm(?;>#9riV9Esj!$?i}Alnx<)V z`fxvEDrpHh)g(2066|GDpoO`KWPAC4e`d&#^Vcce-@dAqUJD(8*Tq0?9>uF< za<{jVJt1;lklPQx6WJaS1}Sb)j&#N64mT`KsPzlfpc6u|x>J&9kCYA!Ie1CDAKHnw zAh+dlr>>G?`)t(`8F7MlF?ShqU4QBC8b(R2i2IdK?(~1>ePP5h)g;e%N&LQwSPcZb z2~PG5Rmmyz%*IHb*d zvTS!URYsnr`_2)_4J$W|YN}eq`iu(Y*+sH;m3qxRQ*w(+cu8zqE?zQ~3Ztn1+?r508^6EA z3AS?_>g9RDt@Hi4rY|*zT2$B~f*s zaJIV|@S>`(lOxeHrO)B8D(W81GmHtrXi0&wx-f5f(*y2bQ)7edMC&F$pfxk{mA0WP zRhdP)FU*H{uVTGjcE%qCx`0ocb3lNGWKzA(m6ASX4us%)vRU7~Q1%Igi!#YazQ{=S zb7$$Y1%U8SPBJJY`4lRhGrp~|8l=-3qXPC=SEdT%lhiC+h&J^l@XHk)ail10$ZjV+ zHX2GfCwrS?ud(_uZ>~@lN)-} zCXsM}7rUz+0OCSE)B=CyQux}>UM-Fk=MKtV^5(a)4kt<6bY~URq_@~RPn*|DK*(%<3(k5HU?7_3 zMZnD0td=-cuDhcz-dU$5^b#>Vd-wi?Q16he{MX?E6;!~gCt3a5K4M2kSq1*5ax z^lubFyX&u9mKk)0ep%K*aHR$#>{((7lzUbUF-(X8*g*t@sd;>^i z_uWDMxBWnGu{`fnv4=nz6Y2uCVmF1@9&b(37XCKZgY~y@9=+r@p?SyHQVUy5Mhz`6 zQC7-K$hKq6jAVja;7TlDC%uUGt_Tefd&zpP>%LhsYtowP9zFCn9^FE8slc#mlgKqe zUB96%SJmQ9obQYzpwBoY%Oyavb@B?CLAX20M?TRbEjCL}&d>%glEh5h0{W23H(lng zcAU&E4VVI~%z7zb$n4nfeJxU@4q1m+U12dk&&;09W!E}X$NJnvii=)CWLbkW8oznb znkJ2@&`jqz?xT13fG-E~5Luf+hiJ{6TN3SAx}g_*_yHTR+&oI)l{aN~plXO!KHwmf zso2;`)@he7hfEn$*W0Db_9I;OXDYVcA-Q3X0|gNBT*tSe?qNrX-8!@fTb5aYGH&mf z@b;la^;zr;ieK)rBduOVpW!}PS`W8!%^X=pj7 z_Ah2J$tU-th^wB@NkGdH!$O0!))5cD49r!T0Ot|x{B-L6i-e_FAEla0YSg1wbV@d% zcV`y;ErMDHgH6F<0c8iaREcwhLC-|?hc!G9dj1OqdBf~=s>AjBAbA!(I$1K zy+vj8MSk|LT_N6!SnppU$+~swMi1Y#YcJ2V^&v77Ds5v$LZw`La8-z-*r+7P1Lc5# zvbSEmZ-UVwh$POSLq_S$COfhu;6VM`;XE>1DfViM@X!VVYd@l>v(7rN!zigb-4D-S zIzv`r5a&UhW31pyd3!o8X(yVxX>hVSE&i)->b=mv92Vx}kI~IfoGLt)VD843|8vzF z<@h3UE4-F~NzHk4k^O6>_b-hjg2gC6&aYDqHcQn)6aTl$0MA#1SsuffCLGCYBE!Yr zl_!EOtU%Q53Y9InXe*R9=fs1eE+|uSDCprUrh*}~h7Als9nFZ)uf)Smeriuk5|RYX zVq@DzQxc~b(+vR%{a?7OJ`xVAU5K00wHQm+yFXm1bI~%lNV^Ge;}sO6J@HC})>p)G zzWfCD6Zu6dyB2!z=$CFjhghS%>p53f*Vl8t-o45dVM?-yX;=)cWQgIojDfLiTHc)< z%r%(2Lr?Sv%{s4u#LrduVmMx3ZBf%55iQNmXCugMDjjRG!TT5gOGZ|t8lt}a* zT*lWsK9D(w_E@`os_NDq(u0zoz0Gsr>F_5#0T&^-&}P0SHrM$maz9fPiJ~hK{K6=a zurI6Hng=p}gT(y|ZOGr!t!b9tURH{N_qE)!*PD~ae`+jZ(q;u***}uoxyk>zE{!+# z`I>zqh~kAF!n#UcXqED3jPecNCt&A$LPc8M%E=|TMb2dXYtT`7{lyHWxJ8~Bx7Rh< zF8OP@hCFUi;lf`LZ0gt~n_LSo1?8fx7JNTMdT;6tAapRmKqgREG(c9woKKenC z|4`h%48Nz$@_$KMe1_M^($nwnr-h~~Uxa2xUZuJpUca$Fe;^BTV^aNm2QK?1I%Hix zlzm59C&vTZwu8 z%3B4+qYdkxrWp1iP}jAX2+?9Mv}TQL_9Z65w;!lol1{oOeUATwa9<|}U*+}H5EzMk zPao%w^q-9~Sc8-HvxoTNw!#e4lkm=y-9N1A`4Syp9Bgz?uog0A;PRE<4#M<@LFE6t z0C`tXOIouJY4#Nj7*hhmH~R?x9P|@cNbw2Ug+N0ndl~jpoybpHOrMiClL8Hr1g?@d zcgmLqKH#p}JATvrlyR&up3fNFXyc^L?uZx73LigRr6F5_DA%f`%ePqarn_#wOnQg^ zKE+>2edN*ny*XM)SU*JWs+qfe#&Qp1vhA)}6#sh54t)jzs7^J?WZWPDQopdt?2_Xc z=rv?3sCI%4S?(wqaLMn8x26#WfIfw=Fk@S3k}cIj8WkIr!e*XznC33sJb93IP^_I$ zH*qU8aRj!;X{8eTD8{RV_VBFL;kaNy4~$g{y6rjD3)Ctg-te| z!yyxRNonrl*?SqzZi4nyDe!epV+o<_Of>SL>*!rLW46LF`&yZ@WLhv({Tf=99Us5Q z(Mpnd$=@#3C9~X1{B%ZPm=+wes>FOF{p7~>B_rOVZh!zus?zow`&G^t8^&k((>vvC z4=&{W#C`XGlwJ#B0*|PF1OZGfQ>@F%CcqH}M6>_+b z4G%iQFi-ByD~gVPORyGA3w%*nNY6@5-%20#f2*=zVfS8j#G3nj{b@@)zeHB9zSyW` z+gsG-ryLk+g)3q_#HGmr7|h*!p}37)e{H+FgVHd zz0?*B(=4a$+zpupFoo9l-?!0Md zUtZ&oxAG^DoN+y$>s(5OwqR<;)k-|1uJqqXtPXY1q&&p-tT}mV1vjY~0K~!p?ZNc? zw$gBH+j&Dz=-h99h}ccK^Da7fnIvY$C765oUDoP{JN~^oTeTa9qJif*Mq=LO@|WaExkoII`Tej*Rm`&r^#|HV$js?8^2w@>7}MP{$98?aDK2BIOL}O z*h~?jO93zpJmNIDyBx5ueqoU=y2cw0lp6=u_I`8RBj8ggAkU4%b%5KPKwizW{b^LN zD9%fXGT0asHdfkQ0R3q`jdg%jD}2mPx4Jd^hVo zRRxfc2;8|SU<Fno+>Y_;f?a*PSN-1g1D`Nx`EUO6t1y<_Q29UW0$*W zjLS#+GU@(#%~`*5jrdmNjNs-kvUdl$29?!lYOj6-Eu2yCjuWWI7cUH*waeBp=s z)StLRR4)+gcEEl6SVfs&b2gmUHH&;9MHaZVmWVnjc)R*k?weVM<-0NPZuHj^dHUQt4+&+{m@GmxiPZLY{~27mXKrSypQ92p^8P&y0Jg?^kQWDp z$Xy_TuOAReQo(j^&UJEwiwZRS#10S}B(OsRtf0=Nz6CkK=M{jKlkf51m7a3I#PINS zdU8-n(E^|^+txUc}Z#DykJ zp;bU*vWK8-n-oK7XLA7?q@|5}@6AQ_+FEP;MdoaV`;UWX_{k#e$S;{vN22u|n<8Zz z$n27{+G2cTbbbSC5|>{dd@-*K^q7;G5~$93g@C%rRej|#p5lIHB0(1>Suc=TydzU3 zUrHZ{p`S3P1rX)%01Oc7lQ5Ox#gP6$JYNjf0#wR=!my5Xf-&nUII}oV?Zb`G-(XBN zH#SN~(k)N`GKCEOMN8nvnuiis_SQF@b86y9T=(J8um{=&Cma)_{t^dlwcy}`*><2Q z318;nrI1cORr!ozOgFP(G_49T$++we1Cm#g9?z`$^tFwps{C?6qa@wlZ8-W}5%191 z8=p8nu{b>Nim2b`n?=dFS|lVh`7XpZILuZ4EPn<|s| zPQSKIoeUGq>X@GPx?)aJ$nhWX|Hz`J2YZv`e^!o3B^osolpopn^SpBZg{F@`1 z?6FN2C_goadycZ6&k;$yvy;+uO5!<k-+%aP`_;#gJ7H%d~6#O1NF4yxhG#c~|q8_@@WP zfGUi~%&=2CatZb2(HLq#tSG&uScqyTbEb(H|4|0NHF)r|y=5C;jyGVwCA@*jOA6U{|xOupgo8 z?D$WCKj;!_4dA2yK@*#U6Cs+DGM^pFW!1pnns_bodn@LwU=#6C!+r=9vZE`H2h>odrI(?*m$T00}- zsCoH*l1~~?j#p|=m~!t!NC&40lCDV=)2G_z&Xgvt%wN!RnSkO&XdEz$7Ujf9;K4^nn<$~|uk74LBa`+;;bcw~PCMiHZ{|S;a zV<_+pIgDTcnkPT-9CZ6iyb(*VMdq2`0rk46xi%K#Lgg^uxb)R7M;^tm@*%+Y4=$+7idP`uc7_B0YmsOv3L`;i{%GUguxE}`cx8A#~ zlISXIr|Y5Dx13x_c61K64NJfPdy>dK;5TgTBT9?d*Qcy>4T`@e`u6cBigi|3i3Q^N zoLbv5>1{(Cr4xn3%uJYKPK07{m0wy7yrO}6Emq-6oYO0gC=qX)tgx^RAzW|NuXOlm zAVQA1bjN7by#@k_eaSO;Rn)t0Y${ zv6LqBXf@9*Kl80yG*dc}(Jj>}G;LDRX^5#eUn1)T-{i@<%DgaKRu47-(OjM5>k z92_DZ?B;U3CjZQHMdNLzDqp)!rPTRVXB36KQ9JhM+=aZoAVYoTsixkk6AAqvPFd4+ zV)>_ozMJaz7J9S|cKd>AYZF)dMvj_SRO)e9mbw5yWVdO009d8e-66-8swmBG$XA2l znCTF*HVU0_mS-E%bJbnVRGD}VLVS4&i0t*GB(^UVr$x{=(5emT(OE7Jf&9(^sbo7pT_La`45{UhSGrRF1)Q@ajOWQmK<8>pqRHn9k z!y5g_uWu4!ajFjjQ%|N7KJQ}h8L@n5kTTg^B(Nri``S|g$SQO0uArv;t_uos|g zqQXKga1CFJl}~bDFE@*~9pS%yYL@$jmNYl(B4S)Te-21B$Za$W3_8DLv44|;8T0yg zT-%kxK-_ZaCb^gSzW;t4?_0j07I61zmR@gum%z18Sp6(_yvVt;X!Q7wW;HRUJ4OD0 zHCuR=^BhdQKRw+>?2+=I)Hcd|(F zoIonYnI#;NYV*|bW1?Mp~?Bnv{z1v3#}@ZHX*0S-aRclD971U5ZBOGM}+?^ zeRy07JF_8^D*xMCtR1HjlInVgPC8KyPx{VkgdXvwBrThbTGeQY36HteQ%-o12`~9< zSsJ?u;#((qIOZ6^5!Tl0%}@JgR{;SQDh>3H)H`lTw9q=y%rA$yv00l#+?%f0^JT=n ze40KDmSy8R&gsTCE?~{xY^d37=*yD~PTSqFm%NSp8Q|BL1Z?^L7J%K}hc{-xB-1&K z9)c?3k@mV_&%1-$hB0b~T&bTKJrx#6fved0Kedum-FtkrQTTq-OugGsU+*m*VHZ`d z-inF0SFz-$nH31<M=t*7%OG^ACnS+iWqa-dZ2I4c zRMDBj7hEA)sy3|Tk=Z{g-`q85ZyXZ2&(qHV+M4$iQrV5%cVmz30AeV0M~bjbub>kF zF><;T&#)??9cf6V>qHzGGSoU<^ha=VY)*4AFQ~cj_@!v}F0N>=9I)Lsq37@A(vQ^L ze)G~Iej`~ErVsJ?IDZLKcaZ!e5L=)Dv9x+8Qs8`!j9L;%>9mjB!#WD7ILc<5J5w{o zd+NmdI zYNVsp(nYn{_Sw{Jp1IZs)TkHl&abIpc~lj-H$5(bhgl;GYQ@=O4w4`%8l}!pb3=s$ zKT>QvRRIz2YplHRzJmmxa8TOu4O`A=VDLnhi~H932*+Gy5jkw5UY9VAhU=d)1=7OqxFpN<}sqga;pi z&IxdqVpZoZy#?*FW;0puJf+)TKgia~R(~mBjW&Cbmu4REl#zEO(y&g?(Q2sD;ER_B#NTO5;>VRnG5PSe`WP%`B?S_@&dB zN2nNeEpBKB6tIalX58SO`?VUYkC^J}5*RgiV$@jJEHuxWNlO!dssZ) zdZ!X-otf!c8SbN<=UU5s>!!4^DjHik8zja39Rj7mbmtgvl*vcSuZ#gcTZanE_UftB z^&L{ZGWE8FC%@4>{}+cO)Lm(uTHO(C%(tz5E*L|Sf`>T$&}^C4AL->uygvPX{D;wG z00j^?`sq{Vk?5JfXT;tj?GL;Ebgnpe6{EA+_BLc1iRAq$3W#wW@p3@y);kxs>KzJY zJNi{vHU}ss1JdgDH$x=eGQ+p3xG~RC@nq!JLA+lL*p!oCqmU|m2rj)tj?y{ zv5SnUftMCr#SH)!XH~El-!epVc8tJ=w+%0KaJUU$N*RIi(zWPvy}CY)v^r=)E-D-#QvzL!reqZKnsti!6D-eV*&|aMi<4nJoRSUh3F#el$s(>m z8{gUPcvk`%Te#>>`Yzfm4K&OrPZ@Y|YSHXTX#6t`{*w=9K}4Z6AklgAUAN)p(DFT4 z2dzqxLT39S(-`%H#YHf+pa;4l_jwSP=~@U^^+gH**?*8uuq=$aJN-$UKeEg-$AB3B zxYU*=MNL!*v!}dve$$}*PjVS(Q~KdZ>V3dD4p5fJEN_0kERg`UwsNJN>10zC{Bo%J zR8Uq6a)#x5v7-gF7z5A*F>LihiU{abvEeMYs(@Ywq5yK;qbd~S8=O*4M=Yw!0gD+j z$3XmEuMD4Hmw)fXge{Pv7b9mu4HADefhR&75Y4T|CT`3t;4PEF?WGtP5c)XHG5A!h*L{Fj9 zSq#d#%MH!K?d27KMk|`838x=%85>>Wp4GLI_->1~Q4kGIMN_k<@>AQMNS7ZBw)YQR z`LOLO>;qHrVZ8wg9+7Pe@SSdaetwMG+cO)9j|7UDr;ZeqJ%F3bJ1Bk4qJVW2(S3q? zTHrk|CQN^QA_f?pV0j0sI^5Au$z{Pa6wo38B$c751_jaZv7IlUa9i76FHk3*>*x>| zEm1?q@rT`;6tfrE!Im{~>15cLLD>ZI3si}}WcLTz<6gzmr@?21KYepZD%frNdlq$Y*u3ln?%e0tThsC9dDY*XXd3`2KCguuE>Puw$I{POA3{ zWE!BF{Y_A91a}B9w@V1Fpy;Xae+?Au>*W$P>)p@R1OR^!UlWcjnhNr5j~S&~^~+l8 z`n!UkPX@6F|FkL~PNwniAvk4q6+6CamO&oB0+`J;Ow+8tO57REt62rB+L)kvp>SOT zfX~3F6AC{Y&=c&s`=cVNRQV;1HP-)JO2v{Frq5zdDSZjB^IV!Co?x9_MwNYXl{EKL zX`eS?-{2tw4z*&-!dr(4i03x;^p00Kh#nZ&KS(J11Y|zVhktfctJ<~x8F^EPCWxE- z8`fnW+lT8?RvP>jADQW?d5FUQU%Lb7N&}+w<}ADm+7!mJjWmvDE10%U<@Fw|0mO1V z(7qI{Q%H7_`rl)N+0ZhXSJOMNq{Gf~57Ju9gh`+?>qD@aw-{Aip*Jsx@npj?hX}k$ z8Q`#q*rbFFNaqFq!4Tt4iglAM12duCIMT|m-T zQ$?{`j-g_ac2Uh0iy#_sm?*aZhyUi|*SYrbKC>)M^z1DV)X|KA^ncei6Q42f(VFeQ zQ^zwd+Nu?!Om;AJYKpY5cTM}HvDNQ8A`ok(4byg_(iKwOXCa`^iE>hBFnYw|L(%DSZ& zDsr^PY5k?VKfqUpY(RJ1cKEkjrR;k8)#phqWF4yu|tmsS_7c~f$pN}aYTI0l)!*<2S? z26g^g9R-YNv6Au642u-+EwFy_>y09kk8OqYsGbPU_~89LDXaE*lITGWx=0CH+Vbhj^~kX1a-JLme5o-$T+S<^^w5zdg+ zwbvh@{k?HV!ZtwnYJ&Ot2R+~2uL&ToeQ{kJaHSdT;G5i zwreNzy!!D9=;)U8ix}q5tzQk z|M}r&jKu8onUCT7E#0FWlhZ?%gCN5_jprD^S#rzR^7Tpd!(o#2xsfit)kS7J|Ds$& zH309!E4v@eV}xn{w0&=L|6Z}}!;b6eLf@G)Em)O3^2eFEK11=2Y=!YeM$@Xh0*Wc` zDpv4QREk(=UUM~@@$;FH1V&i$*&$}^&e#x7#}MPTrlW|;Y|cL=t@LCNka6;MuwLr} z2{(CC`B~yqA~EV>lIbSbfjI!6Y}0pul(s-L=Z2;Zv7gD4ejDXcUf(0u8;}nrKv>7| zcwOs9!c`wo&YG05?Xx7f9NI&Y7Tw#368p3#J~esW|B~DqQ~^^uA<>;W_y__m57>?k z{600}NDpv)275*U`)3u1)ppx0Hq!~9SoDpVW;TEcrUG`br4aj5^Iyr9)a`?r9R}NF z;xqxL-A@1jyOTzYnbfEVzL7{bHV<|bg)JAA1JME{0D%+{xs&bz4|keg8pnY~R58t}-aIre*2eWia$+{rK;ghpLl@_bvEy`G%4u zI+r^|cgtJTD!Zb0t8`=Fyy&SgsjDiSqP*S;|8ED4{@EPd&iN}qJQYnEARSilz(GK; zM>{&dme}zXpD^*ETxMu)@Zx*^ir?8o+84^UY-tJ&ka_)6TjHK(3&g142_|bvJtV}tW5|G`OxS|14lIMXP5|E2{fHdsiEZ4kF?YZ^eUrA4tqfCiE=DP~g z1bK95*s-g1jG|F2fSgN`%Z#?hhJfe}L@5eJdIGK)l4w7jg`wRtx_>6UK6J&WMLImM zaZkM3Y?jhqXgRuC@PF&q0e;I4;j{Fh@)m)d`L2XL;cFk+Uy47s$Sj!Gi?rbStI7@D zv0KmX>JR)wZYzM$`BgRQmHsYjWA=~(%*zsIaGFFgb*T%W-qrYa;Oaw}rG+R+na&Ww zv@TS=V!JgFNmOnG_QjgV7oo>X8EYcpJ87(yj%#gEM%Jtk=s@Y;Z&KPwUENG&SGPGU z{MB>xeTM@EVADW6p#Qu_|MqV-?dRQ3<$on-><7wPQa11K{0Q>~V3VBJ_iupm?aDGa z%N{C&`DpP2!IX?jXi`^|KPJ(j<_z`w-UbOhjy0T3Q3xSTbgCvaFM1N}Itm7-e}Cba zWKjJUh>#GA)v1Uf(E@nb(T05h_qzG)sg@6&>@@Qo7c_bT9(7Ay``u~URv>A%Ys`hM z&IfJHmYaRsA$x>zS5A|-<~H=aF6(>mGpcB~+?KV%e>zdKUUP2nnp$`<_OM4KL&&#H z`Y;( z5Os?!m&SyOS^8{SHo!)KMr*$LirQUlp5^jiD<}zKsGpvB?3hpzs8}b!IlNjgrI#APT@!C@NVnMXFW)F0$k`iv0+mgkS?J?AF>){L zl8yPAWMq%1feXuVz`<`iuf_cuF&^D$AZ26OEVetFMWz@?1GR$;m)_iw**QR;9W`P{ zSN|6BXiyD~oj}Ia(A^4sHEttJ%}NTWS)B@W`FbFiC?}i4 zW^1z#gn=aU3pgENUs2Z3Rkse3u~l|wVVZP2PR8tZzEkimb*pZz*mz5_8j3=~dDe7O;T zi9{iKDpT9-?!B)#Rr~@#n04C-0fkKA9 z*7$e=-*aO3?5~1oto$}eg|~Y4$ctYO_$TKo>s^o|Iz=d5t?|wvdqOf9`LVrw$1sMD zr$=jqHydTT>gFGT9G%dpay$wL4#z9~af10%Eq(GDmVyP?_MkCHa zR?w7R&gY>D^NR zB%OqnT!7u^DP-s_w5<_qZtv?1=uIn@(T$eiupVty9S=F-1*)?zC4Di5>JtP>b9@O5 zLQ*{>i!2yho8hd&wn~gbC5-F=f=5c#iW=D z-c?Z5-yJD&guu9Ai#(@y4}NA|8VTL=WxOku+&F;OrDxL6h2hPl0)?NP--}33j2)sI z+s3+JT%IGw%eeRJR?E|)^sd&ipf-2iy=sjccmbF2%eN+HX4(FPTM1yiZA)DbnS9@9 ziwptqbeTu5Q#A!*y*QdAxs;X1IMr2n@$$q1@kAbSPYXL#R#{mD2B0tAC=>6P35;=2g+d8_ zY*%D?)dnRrqh_YXkK6RLT7~nV)OK_mCJpHyf-3g0t9?f_uaO7NZRMzqEs|Z!G=rl? zz7{P;0yScM^q&1nW*iu%0;@A2f5Tq*Tll3!_^X3Dg@?R9pq@ofVAl!W`DLEjbJ5uR z?Jw+pY$KB&3#$z|yoxZ7vU)F6gldE-5A&)55@)UgN{7w!1T1S4s6Aj+X5p8q*^_W<`j5l2R;hUX(sF}OeHo}~ib4ZYWOMYpgd%l4Mz`#*F298J zB?Slk*G7Ur+RW4s(QJfO{OvDUj4st4yc5Qg?U`O+Km%GYS*xa0f4o|wNR|zS1t@Q} z4W-|$TLQsU?N+V^-kIAOH6CUa>a@apQX{@6?Q&DTkOOC!{3P%y(#yi;J%4H2qOdLs^TP!&?}hVX4qR zB7?=_MC?t{0`tNHAv24{0n$?=I^jZM>X7r)m5{U*GW{*LUdpp>B)GpjDY~aK6rBC-uZVi&FnVVE*IP#CPDj$HaEA_g>{0KjHhn zhl({!>HaDnZoz|+Q6Oug9Zd(Zzne>mDBc$=BStT|Ej58Z;YSo_Cr{xGcX&{=2?#gO zXlP!Z0Td!J|js(d~t((a%AhY?tV#wW+ceIq#?6$`iZ7j2-0XZP;N+H zSqs*9lN4mNQ$gGO_X9)U-8lMPO`EBt~@iM|kfT=s0)d?3N44Efjuma(WaXyRnD~(Oypw z?dVP!XTs3&D7esifABW{=JrEb07e)ybY)@mXbnYmK<5`*XJ7jR;Og;>+{wFec8fq` z>+O59k3kt9#^3E*%#T4o2b|>AQ^+lyh?G;2+#}ho_goR_z(>T2PK2Rq`&Fbi-Mmx5Ul> zI`^S3IG?EQ=iE@{3azhgRCI&`c}(?%wOi7lq=7KeB9*^e8jxLrD|PcrwNoZDFve;d z;kcN;l*Q@S*s*i7Yq~~C4DCHn1}RM`DO1&UzzOXeZiCzAj)Ldjyi)Hgp~nLXz(_q1 zT&<^)>W)>SZ3a#=OI7&Qq9|}!wIy@${thP#vSSQ0ehJ;!@w{znJORU2$blyZ5#gM; zde*CUR|QwOX0jY5f$Xw&j82#{9^~0gu4vxWPk8ePOf<$~#U{_1T)mkb#x_>KU=`(% z9jSoMOiGU5q8`sSBl4{mR<}{RnkSHqNz&eqMBIebLfX^KEhrigf0w6iC8Ab0-Xsv| zebQYOdQ}Bf8EH+aEQ1$NL+|QHk}AJ|eFcfjH`EQ0J%t8_i4BXLcbH=n=D7vloHPKpzm^8RI1zttzqe8_epdQBn)kMIhF0DBgu$;>jMCAsul78(?M~zm z1HYGW-B8TVDAcRML;u`@%Y8jKG6I?w(ayRwcXme5qM2fbX7BK;5E>m@2xWD zk@&t;oZ)^FP4r!2oFvAZ3jGmii^}npN)kUb%I;32ST#`g=^?P_=peB8`Ut42N1WU$ zieDVIXgqn*4(HZxImo>gD-`-L*am7`sCHvjx1`;Yq01%>{2m!K%_u>zvA3~FEQXI6mj%4=m#@FXjOe_4_1u(2wOT*s5cZe3-(%Oq0a5`1 zg6;m&0B=|7bHfcw3zX1(k$UbSIgVv_t?bC-5ZO+baL=~=yE)fIYhyOLkY=>8dRLUf zR@p)4YML`j_F&ijn6f8A^)50}URxd`L{Mk;_x6q``8%Z~q0X4cON5bQ-^jGQ1Y*4n zh2>Y^4ldSXPI~BfFI9A_gL^rO^{)1-swTI~D~h)#PZ$#e{Naix!+Cz<`&}QKmvKQ-%JynR7}*KCg0U>`5YjN09Gd^goY^ z-=Vt@{M@r&zR@&vv$XfX@e#@?CyL!>Qq6arC|DP6b;ZV86Ift>V=&WrgR6X!$k)(w z*CG!0Z`5s1(Osd_KQj9jspxJfe4JIx*%I^Xl=x;JIeF{`g~f?68wd}o>n)tpa>V@Y z+Nfxzr(;mKVxLFnZXNY4gCBRSy}RQ!^)bUrm^6LU#ZGJ(W?R|)zSc3pILafsT*m;W$;zy)aEAeWe&=OIqZ-BM7W=#Rt!qpYlcP z&q7t}&-~15kLVT${J3oX(&H-zsA-)AUch0C=Hx9q13;x-b zBr9?ejz)nZ_8^iY_UO6PF=c7r_^)hzZ;UaEOsr385c6IvjIp17IF#x6%fiml&VdJN z?h@Q8c>y{h-}Q25U`lZyEjxrO_kaq;s z7R4ogC?tijHS0l?4r2r?bK{7!`rz`iOfRs zK*)jh>Z|4le8zO|Agf*9av+rKisB_=zqyk$X`+6`$%ECo*v2;SYEA>y(uk@mP;w7C zM@{7rpa}aT^sho&+)@1H$?m`44L7o_?2$0I+Xodu$yN*~P!0nC4LR~~k3MaAP>$WH zm$Y2sOOx5zEJX*6C=n+(#I`K@yM)5vnl^3C0_%9G|~L|hML6cfCMVW37{U} zGqa&uH4jJFCyF+y>`xM7-lsPF22GIO`0mIZmvj4t z4PTWB-zlOgEWa*-#DT^ktCmabY^47XMfJ%|2GpS>Dl{YC&~m zK43cvX}Bl5?!UpE=%nXw>2J?vBp~sXW2ze1y1+ib7oc5^Py^}u8L4RVx^(*DJ{;o1 zd+>W6@fQnSSM13t=`>^XFuo*eMJGm!yI^hpDojI!KZ(cc%z`WI(%UkC$Y(yD$$g@2 zb1fSDxH!V82qUb?QNDMPNdkFc$~PjLlKVgIIEOGbwCO-YMCUO zFnRXw+2m^*8bxy%?V-FiT07H_-&5YlH>%Fn9U)tFoF5na(ZyB0fT5TaJqcUaZ8Ukz8X;J-^KNj`iI4aIzsxCjfJ zMXGL1RfGv-vkAy4)n3d4K_i<8v0g) z-ZhpKHd=hg~_h(efa=sHO$R`zZr?zT(zdqzO%RWfw$)m%Xy)}%HPZlO=&cy zd}dZh4jK8r!&JFbZ+TQiK_k7fS-!DScebOAMW2XY4a(QI!Z$540{GQ5!^Zhz66%Wb zsI$(E{^`*}#P9*xnH;Cn+!HU9aaInq|ADeM z+#K<9odfk(2%+syR4W1@ZVI6|++GsL^IldyFTL0~APU%9 zT&T0mW2zJ9I;CtUurY;py(dM+R%01&e16K2=~IfesGR*~9~mJbt)BppAlz}6v80?t z0E9iTr8ddJML#=w7U|>Y_AZIL<6FZy@yP%(sUio@B(c$k?XExxaf^JVbQaD1#G_E4#;DMpGe73WBRlu*QX@%9H5PS=lGS z%dAl9Ik2p5oK-~V#OelCud!6Lc@lvabc@=fb%5GLkA~=J32GblUd0Cd8QdMp(l7jC z8C*H^nt$m*&`3MDzsfXO>RsYMy4yE|Le?)@t3`+iE602JUwjfChw0SA7@$*YqkwUk zRTQtC`bW~?9<7kiRID_f{Hd>ikx_SlDBrPL%$fClY_Ly>i^P;o~l&R6h?8I%; z%hz9*WKDN9!!aFHW{-k#NX?M-&Yq`dJjbhXmtv8gyB+ayo=b3py1{>`@5_7P6>weY zD9hg%;~D}NxE5CkbQ1xJUv`m`XKQ|qbA(BN*FMW!u8YWz-!1tK8!jb} zGfy1mz0fQLyJT4*5!+aOW_w;;hp#?kd2iwX>v1Sjl*ZjW$=Cm4o6WDYqVNQI93j!7XRRg9c1OtSbOeU z)Sq+j(~Lj2oo26O05q4)xsam=I{sgxodIX)Nca`fhx;eml<$ICt6hzAMu3D9FWRD{ zZJq;wjm0U4y2531CEy_M0H))J0pMLn@4*-@+RJx={cx$6Jv}i*VL;>W97-}KLW+@W zW7%@J5nmBbn4eCVhEPbIfX%*ZYKDe+KR6u28c(7}>%ch@Gv^qaqLsLZ`6c-chG1lb z`{LAFjCYA>Q3h0kxLdY%$akXKarCAqAVR;oB0@!3=^FQDC_nOksyvir zxEC;59EyombD`%y6MM?tzX_b?Qw@A1keFy3W9xUNkfT`M6o0{Q_SJ!|L5Hq9V%k9t zIUAs=ErslB-t3A)xX&n}-sUpn!8&~T4m~%AB7wi7 zA9{~>S6?#}eHhQz(p{KuaeO@vuJMdrHk@9u&e<|q$JV3CTCg;tW^I?y^_J{DWXBBK zfll%%`N<<7HKG>k_rpHpVER~r$J=QkcVptI@_J8FMs9IJB2mDXfcrh6Ft9ub7>fLZ zkX$lN;PFlx$OBDB)sx#r4+$4NAM_7b?4p4(=tpSZAN@o{JILfrZ*Moz`Ol`$e4>=q z?Nkgc-ABhC#%CK{X+gl={-KrbT_QrxMi;GO-@_AtWSvVqzdtVWj$P<3(bI4Z0{i~> z9AEqfnyI!ht(T%4Myl)2mXM0E;3cl55PksP^>^In_AJhCsdm!Z!T*YGaJW`@<}UlA z6~$I2lCA!<_Q15cBRe1QE!y_P9T%ZvX&Fgr62(ZJD=mQ6o@oaT?n@)TM5z-iS6MXxKF_a@cbxc-vN^3+Re zXWy@mFSC1*e*uZWL=w<98oq16cWk;&;1f^=)-n|5U`xd@Q3xvU;g<9}!kvqOhjKJy zCI8DkQPW-8!fviv*4p-P7Z1My@F3J&oi(OL`8wQoT!*y!sC2Mo^eBsI`|g&Ak8ds$ zNOGuLSx17`#H)FlTzEhV1?yq*vxFMa)7PF#2VUMp1SAH4wllPvi&fot?ULV-0>D@L z=4ds1PE?O~5?DU6|H-7GSI{ONUbqxAL@NT6#d?1`U>7CM2eAC61N)X*>s*9;oecgk zVJ3r_b*iktT6Oak7(lXx+bO#3%sx}{(e>ekzU0P)xNHWcI*7{j$HCgs2DAfgxymO4xo>`!BNE`ScB;Y z#5kP2S=7i{Axek2U37`(<@IZbCwAs8JK*4|98q#~_IBQpb;LM<+5Bm(;6yj~snDqp zqj#xbm03)1U3FdsncAPwc_6VXsP29?U~H;+!3xMZZW*t-md6MqmPeTZJ}o`4*9?!~ zA74rh<$z>${bO9c<5q^9jjBE7m0?CPx$M z)W>IL!By>H4+xt@G2`)Mdh{GE5=L$zD`#{rviL*9*>t{cK`Uo?ktNN}<06C5Hd)vq zAGAJ8m^g1!|AbTafTT7x+*|j!Gp(%elBC1er|I97836S3nggO2JKHlHp8U z8Y>r|Bzd_8ba?sj^=6F%%|i{2+Ow-msk#eoH>iYwQpA9whU1}5MM$TwzGf;87@ya6 zZElUKto=ZKeeGIb6CNOFwHQ1g$w9_2pSqp6sPjB^EJ9w zcf^=&^XlJ)8T9rAmosq!w08Jz>Q0*x7Y?%+idBZA@gTFK;}P;w8w(U z-DIKbc(8yQft%Jr=(6DN#-HTlGkm}iJqSPhfkR5_f?ME<|D%5QB1Ar=Av#}n;81JQ z7tyUee2oU{06#(vu_|Gyiu972n3h8R8TH6f@Yh$@XY`IL3P$gHWajH~J$~_vO_bc`k+xmm=OYD~AQMF7ozopM& zV<1-u7%X?^RKG=~Zw53HN#~n~Ck#NW882~?U$Q6SEEap+3--WC zK%-7fMqcukZoygp)1(zAeH%4>KT@$zYG|20kA}2}=0>xs2|tm7|FZo#M~>WGPvkvI z_6@FBIh^7%Gq#U%?(omf0(3VbgOV@Tg4Ns2kkobf;&W(E#;&=lJ-FWCn?_0wZPDY*IR>RNKK?f5D!R%h zcl>uO1qv~c<$!e5iCD@bHlSHo>JNNzqNdrk6=YzD;K zi4f^+3=h55o6_QFn@FnlCP$a`7QlsBT4C<)6AP;AOMIJ%lE(YM#D1y%;bU#M zZvOjV``&##0Z@U<&+=$HHf7F3Lj*2{zI~&qL-z@Ps>5n+X$Z+J^K_t zy`2Iob%eV}lFVneK-cCbW>bx-o331H%aW#fV@TiXB;YdlZ+V!#e`9LsVS4D@_wjsc z(e6f z_FhY0OuV04qpt3d)qnVEcE;wNG%-ZF3R7EW_G~49&VCq-i3K!4Rd5CONc$QQIr{HY zX8`jX(qudA(?daa!ykNl`W{#WwWoTv2gK{>=i+tNf!0n)!((j09k9_?x|Ugd?WJoZ z^e-m-Uqx3P)zlk?2?YVENi%xnXrx2BLuw2`q`ON=NlA%~W}uV`j2I1q5*yu}N`vGN z5l~S4U3cDd&)GTqW9Q!QdEe)Kp7;B%LS!H!MkS}2(hM|3#!2!8Wccc^eNIkwYFlc0 z+M~Rp_1-y9$2u!~zu&cdAYV>ickXH%AErpAuy4f~^Phm*17<)iDfjiJyqWg!OsN}u6QX&&DA4+mNA~HOimna|tT{`@hd)6r z8&%`AHPubijx1Qnki-B&s+gfgt8WuJ=gzZ?J8yFDq!-F)bPFUweszLLUdcXvI#6_? z;?J|E!;!PG>A9(?CH-6;&;^(`B7Ncjdr<+L z`u)X=0y<5!b&eYjnip}D!g3`=N@5Ug*#l3e5q~-=Q`8v!Syh}lb>bC2+>E+W3;R}PYJ}o`a*ocd zvYw{M$EgwT)Dt|vFD7ZZ(tahi30i;77o9Kv^ZFCjaVvDNeAU((scwCfvzbk~J>IqX z#$hw*h~;yh0;Cz{K2PKt1rb-eSQegp_DMgMbZ4@trWg(o%qN)aFKkfjS!a-eBxUH zD=UT2W0_XqfGzU09qH|{lYc+nNmBvod5Hbeo2pjwDy{xt{H*M)M$Ey^iw_77^0nF2&V zDK)VW*hANg$6xR?jqas9(%04CqF=dgf23g4KTEVEVMHmeZ`3apBGc9z8@&@NSI_HE zSZ;*QWL#qhO!v;Er=)wufTw4$JSZ-`JRoa6XIT1~plrXd^rsHc%GI`Hhjo>er;j0M zEdgSXd{`8Vs&eCywNAoq^;LD~{HVfy)IcM&V+ITm6C;=gfA8SnOeNNx$iGyrA3wIO zuBw}K8=DN5kYz4eN_7`~d-}@!qvcy_c zmnT7PYI=a{w`DhRKRwjt#jHH8sMxQz>C(5fqj=l5eI9x^k2i9ZQ-8e*xzkmimFr(s zn=?{VvwC#0V{ ze|!hcdW?oU($nEoIUxmF#r7GL{^4xX6w z?m(_1;H#uOxOkxQkp`#6upl2N>hP(b4yv6F_Cb~sukJyJ6~IDgNYE^2&sINY3WK#2 z`4P=WI(rVD%6mn5Myb$T28$DXs)$Q;sxY%0CQ5DGnt~)G{M#_M<|wb`=w60I0)vHc zsD+Z8OHZ_{4|Q|lQ^=Z#pJ21Sk^RW6@3w0qI+eZ*V8RZOa?+ZX5i0HlcX z&JX6VmI9*;8Z#e(6pNJqq%l7z2Z+sp=s(!k!#zD9Jy@{B#Z;pQ=RZAb z)SLP(k-S@MBHE}1bBvwc;I0i!%`o|rD)#tQk5!D}1$Fg{`k#H^YVS>bfY*oeybu-F zI=~*NSm|Ex^B8GL4T8V&1cVm?&ZR%^n={AoY#>{dzplxk1|;k7*FP|Gf3jWsz#=|9 zRXO&E-|#hjP$xI4c6xMME7sUV>s|YrVlQsjJ1u7cNp;`h%#|3+nS$jfpc~Qw@a5k< zP+6tx9oQ65SC_l8_SD+VYdu^|GC?g;Ke^3k{bM36DsQ!bK8HqN$I*#=?{K+oeztEd z&2Rd8cM$x9C(JWDb%VP&)fEt%3apB#Yo#ay+Wj&{0PV{S*oo&L|K6Qhha&d@+kMxs z7UsQtWO>}L72%3ZYBDCU;h|-LajA!X?af7pEPqzNv1~Z+QAvMHXrGe+s=i%q5T##x zS-ONl4S+?vvr1xJY-w_Vxbf3GCwb>~2BcY>WstqGMp{;`FS;}pnQaa*<84PLTCF}O zY*@s;HGy2Nse`09JtkRd-_fXM5NfPi#Eq zKKmr<{G6%`ee5TyyBsZ@L;>KbtD35-3@=uK4Jq+={Vj7G0oq&MuPWt5dX)K8ub}Ld zd~@m>6~5_O4Z&Ur4TzW4fhs^Ov7I5_Ixr1vpCX>bKdSx=l`v(~%@IcTS8L-zq}RkI z04V@v_SPZ4hqDb*^R&=r-5f5r_^$aH#5_G|J(l|VeoB~@nAp$EDBS+B)Q@0h%%!T3 zrz}`tm~+1P^?0UQs!)muI$0TIb;CY-5+PyWXJ^u!!y0R`O8KihOS;l#v=|TIHd`B- zwGYY9Uf4#j?wexg(%W*BIBVQj9)5Seq)naK*413ZSa}FKeZ|in$G+v>#`T_37h+LvSl8Q6u@z(O{(+z z0>7q&JR6PpiuJW`o*W3Pm(r;W)$26}DMT|!OzFayn-HS>2k>g?=?;x%A8F^f6k1Xr! z(KB0DfC{TwyB=)9L&*KUP|>CDz0KOtH;{bZA=kFD(zkZ2BHS9}*V&R5H!Zpwu~cDk zXVjY{MZf{{6k84AR}tYA7-^e?4`g3D3qPFq>v(5^WW^VOrN`I=v&78{zssHhQxG*# z^O=)d*Dh>t0@loUdJKySU!5E`gHka=RPrfW&7!@UED zrpwK3w8vrGWcwP&%1<2WNvhNl78?Ma;woX+>Z?FiR_^Fp5d+pe@IYN8sP3%gGg8zt5otCf)xuG?=blSsAgMjW0ayJJ1iy+Tj zBvb{d zB7zOhAKX7HGd7m5mE2E$P-)8|r(=Byca<4KjkFz} z0~GX=c$$S&KxO5du&x=$RJXfc{~V7K)-5wH36dUB&AHcxbcxg~V9*iv9uDkJ>*i*w z1L0ra1MJFHd0=X*n8(FF=?LQ_o{*!>_GG%d>}7NU;4>W+as|0Y^8&*NTJu9m-v2pz zo^^-GFHGV#8kqFIfPPd_7OjkM=gVTK5+tUNcSx`Pl+a>^CkthC_#>YJD+(MQ`Fk`R(MgaH!I_|Fh%h78`pv2J8#EM2?)ivUNxuQc_5S&? zSlh$-()i#amS2dKz&*ax8e301O8-kYjwB3Gd0*(*Wf5OsSU9wObuS z=|B)$RrkTn^}-KCl+CnzpS?P32S^ZW&0dX%n?*jKV7+I9MW3S1GrYg?5Y96Er32+Z>U{?|YmyE)gd6{ZTwo&Tqp z6&7)oQ;cOwDNOvi<;yL+k%7UC?B7~f>#acT?i2^C;|kV z(?m>YBu@-ce@=10=hgZnj4)&NGdcUDRbfd(lnaBs)ldSYHD}EJ`V%1&F}f1GMkrg# z<3tTUtN@*xGS9N4EmBfhIys&J*GniWKJ<&Bf&d^_ftM4-W~><1;$2gcCG5SW{dmE^ zT7CQDZ`j4+|D22k5|U>UgPd9Lj8YZ%x`}6G&nIQiDWZn?ToE(u&K`AJqsn{3+ZoUl z2981zmZjgrIYYVQ`o}JE;z>cyB>_$zZ)xiWA?>t?RnjwrHtNGgt|hG{6l@x4xyw{N zNZegv5drU1L4OhDg%-9=i2)X(0^@@bN26L`JRW--eJ^*JSi2V82NJ}spZRNpnh(N; zbkLr>nhkC{e;Wvq0-5_)ew|JFdavi318Pcg6V{9YLalwgK#CX8zrdHb$_Z^S-2F7# zF+;y&qQ}MywbJIVI8|Xuf&nH(|5xT>&mgW-`@4UT(@B5c|B?7D)Rp~xc zyhrwzcM=2)3t$t|ViU8i^At!E^hRU=mjUen_b7!P6)5iMG6&_2(6h3|+1R?6MTfBL|s@=kAP> z$cDeSR|8=78QRt4v$kc5uJT&5KnH(3u8{>OdT^vrLWeHMoK31 zkJo=8)WRMi$dTpILoP3O1F-zT=mrBqyrEZ?Eo=;nbmm{A7o7wTL{66uXzN3Z{w?jF zN04hTY*l z!&o51>S=eIWmp6K6!tgk2Mjx{^-*SV=7Mr~)e{XyE)#RVp_TjV1+VQlqER>Z-HF$^ z&RO8l2VmlVtF|cP6K>s{U zv;C{|cx(yZ6%kJx&}m~1GzCkjLhEuFT0c~TU-jso&MT>xtwh_jP($e*Ozr|z9V}9_ zZWbcs0xfKxo*lTDxAP84^?UCTbMpt7bI;o~Jf&$3<7uB3>urChlm?ST&{R$TX2PWd ziZRWB5>lc{{HzC0Yj_y&cLw9HVdHZx{-vG8+qT?*4kzfYT&;za?0>)|N2v_sCEItz zFDrFYFzwv|L+dtNEia;t-7j2mXp&oTL~J&SuWrw5Co;<-paGri;FJwX=*5N+-fStC zM=)}N!Kz)Zk3`|GV@xHw*GT!JsRnd=0a;h)M>gn`q=er5K0?a@(=gv-dlrJZR)mKO z(ACAv?e%x~n7t%I!h4czKuz~6A@(rGG0&^EDLn!);e`HV+GzMt$hywA%K^p+qWQn~ z>qGrM_iHscJVxDc9JPkZH0$7%lEsert)BvnxNIHEPMW8VpTd{Hu17rjx6#XwUF_;> zicRoQf&n4X(ko-kY2zCL{b`Zi2J`vu(bC2G7uMJtZiHU=jV;t|NmFBbv1BDl!_QZS z&tHZQK{JU59eYjbd zn(aLZVjUdIdaJJcc+*vRl5>Cig*GQ{52Avx%l5hfd4DZLG==g3Voup%?%}K+?gkVyZkuVpa~;G0we*)<8DS2m3owL7l0g zw**fxm!BzL2f0x@j7&2Wa%mAhwqWJKC=-fUC}}2K`h>pUeaTg`+KY!rte~T=V#dfVy zdr3KIW1U$wen6P0K$TRHYG!l(RkFwD@6UKR*15zOwteff`g%-2;VS4V%Erwpxiwmn zE(jI0z3F}_8n@TylBIt8x1rWc)(=89zBS4ecXCfO*yr@+QE;n>7h%W3Q_wo&ii9Hb z&ENEP%J)_K#UV2ICCa7NDiz}Ajzg6QxInD6JH&(|GdVAYiYC>RG#|EzN^6Ud-MZXK zJ^q#m#D%DxlP2@8jIU!7jVq3PWt(~T^S>~)Q^9qUtli74In4<%7U9&4$W{cuykujO! zpve}=Eq?y_eX4{bDx^=wMTm6Df%{l&p36xwf4D)*8{f0CA~&Bb_?HcDtNj{hR`Zns zr=@+*)u5Td?abs0SuGp0gm6bbo_I5fwpX(k_#z|J#EZmo z2?C&ZNzk6>wkA!Ct1*UB0_wd)i`V0gsrt~+ZM1*X+%vkWBR356;1i9eZ3S(PiM7A@ zD~_zDku?m3%M*EIj!nGJki2uY`>WYNZ8M*I-BTx{UaR4$MT_SwyhzpgnYO!Xd<>ih zFJoOeJ5J0tS^9G}Ehse2xz zObt`g2seof<|^hSmryAJ%SMPHCn5zkHyDagwA0HkPR!PZZyD_B2l|q84eYM|M3$)7 z=ab{lQD8XN6U{^L@0)VXWY?(!2Xh)r*G!s8mtl{~eHDK!rrno*(081V%)X1WM`H@)1dSBGJDlQsNlL=mtY{pPc6{G@+|O2gj|)Aj&kyxXCuntvPk*W3 z79r85L_(vMU~MWJ(rU6Ds*+F!=SJ>sI3M)15a~ zi+?D2HvMOdy_{Y(m{T;_J7wfiSD!;+dloeBzfz?))rm5WR+xU!wjnCj>XyUCtgNOV z3tRfW#TmbRPXBkDc?yh)XmA&weDuYaxdYh)|5`ap)uVaw%-$%Adlb}JG`~tF&xAV1VYO8ic0Yy^%o@IT&XpaP3$96{< znn6AmpKY1st#h3?eZ!kydYb5X+-f6};eXCBT;PbT%(x%8kWTx0MyD!j;nT=Y6x8HC zzK3K`^ z#QT;L7CQN)o{8)eujOEIrZ&NlItY*3Ibr84YSWDK&87C}1CxtXGsUo7H*+Q$43QdcrF literal 0 HcmV?d00001 diff --git a/demo_images/ic15_word_26.png b/demo_images/ic15_word_26.png new file mode 100644 index 0000000000000000000000000000000000000000..a638fb30ea699922ecc2dd2a17a59770728cd236 GIT binary patch literal 12271 zcmVPx#1ZP1_K>z@;j|==^1poj5AY({UO#lFTCIA3{ga82g0001h=l}q9FaQARU;qF* zm;eA5aGbhPJOBVMZAnByRA@tWdj(iy*_JNd)jfS>X70@Q-pssv-<^BgtGi1mp>THz zkOZP=AR%}lK!R&B->~5+pMCz|i!ZZ{+v`EP#n<(FTv+3er_?svcc{qMj0 z>bGBf@db?k;SYcK$AA8(8~DX9zxa}KJb>{B4?g?i%g^Bx!NczE?mqvV=PUSaxc1;P zXn*q$4}S9%eA>@A*FT3(^M4(jPkn!IJDgAaw-NR&&-L}??aj^Y?e)dQ#pUJY*%|!g z9A{@|Cx<7{9)CDKIX*r-ID+E@aC~-ief>Fns4u_#hu?hl```Z-zQR|(`|X!seevK+ z_~>8!_8fwz^coThsi2=csL^L%dni*=FQ?xB8x|T*dmQ-DuiN3X& zfsGY~K{s(`T6#F!__)~zcshoL2Bf6K7nkHWRhD%D>T3oE`W9wqcIRey78iC`mJc^K zPxkk(j*jk5PuXW@?CWa|Zf@AuH|(ox4sLGQ*KGDV`}%@?dC3`oE7y1I+dDS<_V$*2 z3qRQG8#ec*+Yh%_yVo4--JNb;?JXT`&K|5y@2$MsS(@0KA6uCoUL5OrH_-Z~qqe)M zs3t$XC@(#~Jh#k)>F!{`v>=(0FnTy;O|&dZLk^8o!oU21PyUsh#8Yv!6iQn~%ht%w zm}p{3FtZ?9T41fr&{ihubW=434e3bNaN$9y5Y*MQHPHH+Xnl+ZiGZW%>6lSSR>ou- zsv+ITz{=8$K{IDsTRPZSIoZ)&n6_^AcJ2V^Z~=V>h6lir=?QRUc)KxuJRJO3&Vha& z;X&Tf!TxcffeFd+1-V((Rpsq(v$e@|JxAM$NeZgkG}Q$Lx#k%j4yfgSmr) z>HWQ#!-c`snf}$O-sQ=z#j*DJ!RD#%+OhVk;fCVw*9EN=IrYUEmC4arL99?uM{iFj zFNV3Dxq+D;Mpsi=10k)VAR;FtAS3lkO7y9i#B<4~U-Lfs^JD%W1eAqTRD=;4l9~uU zgz8JAIv++uKu1f$*hs;KhIHm35{>k9Oi6m?WCI$NLbo)vXVBeU9DLoK1Kgbaq2uQ4 z>+a&?>E_4s2=MU=@?{14`-B1l{lfpBhzyM5AUHlcG%+qBEg>o+IW{{fE~hxZwyLb9 zsjj!9ZDgQ#`t9V}`r7I7hkH7>hW7shP{uJx4({OQ`!nuw1Eind0n;2p&)ByI?6bA= z{n>-T+rdV)7EN#DQ7ir^1i^`sAhp6hliY#md~&&dSmOjLyqFFvveD zJR}|v8WbBA90LdsiH!)2iwaMOj!2A+N{)+8iH}JIK!*z$4J=vPu+nCzfp4wj@-+Dj1I^VrC-8MVfIQ{1JcyGx- zTV8ile%I^FmZJEY^pK)x?~GuV1V8&|Z@Vy8s{p2nyCuoaNXyI+Z)AWck%@*>9f~E@ z(%PKPv|>6joIM;p!G2vBu1pIC@J_-RXeeO#e&QEb25{2(-X7dIcW(w>G2#CWR#ZX zzOE^4?rIvG7@S?6+1gn?KHRvtK45ckwa>oXXJ72yo$ubB?cD4w9|87Pjt^F$bGR|L zy)gcMdT@UI-Tu<>=3M{k!r=OB-^zH)O#kcAzKW5qqW+eg&bqYb>V&$A*w>{I<#~Yx z={~vf9_iuEiT;cjFRL(Dvp{=_H%-sklxRo6Sdp<7hFDWF&e%-P%-)>Aa`g1~@{I|P zPD@EIEhww0uJ35>9PAky>>2KD>ualTuPLj|jL$OGHIWmM|KlJ2^wXbzl6)yE`%FoO zS4H|4HHn{*(m&zkA8RY|>hVyF%q%TzZ0PnZ4?lk&;1rB4yR5LHwz9slrWsK8x~cYc z!|TfWs*0M*vKnZsE9zcX)j_|ath&6Uy0o~WxS+f+zvLs!nFYBS`M_*OGH{xe5|@!0 zn~@Te86TdK5}BTpSXh};(^l0r+&1>Ee|C0wX?^MdusLeYidi)_1TrwY&Ic z4X#f2&rkNwjJ8kqHVihEv{mOcH0E`_PHC=4sIN+DERV0t4lYjh$w~3fj&nqz?_h?vx6*O3Gw`(0cQ?Y?Xd|e4>L!+Y40jvfFt4c8 z$kfuTqWbdJO*IW&9o>V2!xLkZGgI>$Yuo3iSJ#(!7spoz+aE4Zu8()lnyOowG{+}D zJp1N9zU6-;B=$&3>=$|QpVY*EMoa#TlX;{g_e76JN1J49WJa@~ySsXMd-;V2MWiOA z73Y=IR@ODwwX`<0Rh3p(mQ=voE6yt_%qc3!&WE-zr=Tde7}|pD!rY9U?6jQBl#KMG zjI_k`q=eLjxa5TB#Mp?q=#ZGmps1*T$VmUFAkW|s&yaBM$oRm7%;>DLjPi!!#`enY zzWUMr`ia51$$^H+;l{TEjpIWNlh6*x7s2O3kG^{*_Ebjvsgl$awB%E)%oA;y$NJKb^msJTI2|p0ilMQM zHN&3a7^Sr6!~#$0a31$Hzs+ z!t{v_iwX`53knDg^a=9w^7nD`^>Svh9KF08JiQ&fS@tXz!`sK+H`FyGDKIfVF0UfJ zvNFA)D!r*XqZ#lztGPO}iHm}`vee-0FxMC#+aNa!FGpin8xqq}+m@;41|&jjtL;ox zrxE21$VhX0vRfQ0H7lavbw)#TQF}vvOL0nhW<++Pf2ymQuQ`FKrD#NuCmSNIs5n~# zG!3t4=t}cRkIEZtd9yXYbF~lR;(qbGUkwjF9KM};H#9Q5yu1o31CGnfYc8iJN1)x< zSEn~Hk1LK^mg;|ac2dv{DOP}d_BBf z?49A=Fl-&{tQa#?rRgH%sR`m*;yI2Wy90?|0W`*XPHUW(JnCqDtMZ0x@#>2vGt| zf25=qUfGag?3NjqKiWO9F~51bbNo?|+;IfQ=HAxg`-QF^SX(EzcX#e-z2x%CA@?4j zNx<_eNg?=u5$F9zL{W&SBxoT1i&Ez^px;0b~Vw`vZWo zQ2$^Th6{&KJ7*I^s)>P#DcP82OtYYxnHrkH2&|qq7S?n#x;e$1YG6z@ppvu=w6NMl z4J{6k1S|rlDx-#yLt&I~cx3`Xg-AqbX{l;!tLhtSni=Ei<^(%vEwt=x$W9Ja4+hnh zNpW#BbEjg=aEdxK9UE_kf0SojdPH_*PJMsV$eXUW6MfUe9pib)h2XUtt6JVquO99k zeNt(+*LOEpU{@SY_x8L5j{ zDZjGdL8)u1D2Q_D)WeAt>C;|aZ z(8210xKmd|DoHD;DX8O+IA9Qm#N*Heyao}is;MNaA}g*4kP(rW5RiTONKo*Ji0E@E zDLy$Feg!!}B_&Z+gaisLtEC~QttqPu9jt;bQI({vO2(_`>8Ke}@TS1KswA3(ri8Pi zYx5gBYkJ2Argqj2j&{#Z4=!)c*juYRql4q4BNJO2ySv+a=cgde;D7-2>h|v58*+8q zF&n}v8rhEb`&a6+7$r#*N(oOhu}h1~1V_2Cym@kTdE|$V(|m zi^%}9ic*TQVzSVpRnSBv9-k{OaZT z%`L=O7azas-p>Ba)b!fQ>h|{b>B-R@$D+707;t-Ye#_Cq_h${2tr&GJQGRJjVR;E5 zIdufaf#Cv5W@c)3e}Dh{{1kEpaNog)-Hol)w^QTe6C-u?)#Kx%oJ(NNpF9BM74Dhc zTi*+C4MK>aY7y|*LX4O)yUUP3@xK|&d+s9|YjT~Szhuzh&2cgX#Em*-d9&dt^J#@hPQ!s6cU z_VxMM{X_*@;t+Iq`>VnJnlS*~EiEDc^bsH5b75&I1rV2|C6%x=on4$?USC|@TwY$C z!o#Mf$NLAm#>a-MAvA2RpPhNjc{0S*oZ$LC|8RZZUb4$FN_7x=vd>lVvL*%yM+0S7 zyaa;>i^8ias>?|#fRzEKV02Q#(m<=Jz6nIfHYRjXzd9Paf=`9`9}6JmQHBJvjft%_ zm2RX>g{T9F7JVVc`@^&64_|PZRgzL8qV@ENMmUs?x)KhhLO?3x6=l&+e-ZlW`N`ZEI2MEG9@)BJ3FJeAh)cdv=+p2Lv4#A(^W=NK}bMU zR$4( z+Yg8P@D7fTPpz%&9{%{~Q-&s(YU0+vS zSddd!TQxcHhJ6Dr;+FGFF0EJhzW9bc+%{mLZ7KT#ffgr`lN~1E|?R~vnf*6)AW@KASD$~Nm9uhYSGMdkxh=2W`Prv`} z228woONH4OID+nxtx-v>ZT1iGs zK}kj(Tpsv7NkKWXwu!e(Ktw=vpjQ~Ab}*_$;9X4~a$ICrhL^twu;yfA#xT?}mK9Za z{QYx?I;4c;HB|6kPJR&q@lipE?v8$Rb4Ob%S1?H%Yc~^PhKw}wo3Eb&_+QDXBD4)i z)^uw(Ki|;cposXml)T*H>ZVgV!E&`>+L~GBr(}TTa;){&q{GSCG5F*8#RUu& zFD)ZSv#~8KEqgaRw|jVSdU*lX2dfu7EVnRjPL2OL@< zYsRrwE;t`xdw#>i${it$MvLi^lx;0FSyY4@4<3)zK%tctRmDZ*WhB&;Wbsm>NHHO0 z0*3725*QJh5D^|96P*YKWkfMkMQG~lQCwZTSzi7Q_Ut z3Py(J!M>4kVaX9eaXwxlKCDm{E5ySq)Ym^cC?tVq?II=zQ?7aSTKk&~6*(%b=2BM6++vkUGD(c0c7B`5pzC9eV1FfA){V||C~QX!@U z6?LzQ*w860E^)ADU^Ov1+C-9$HUW=yV%ioKZ*9tg-$)L@kV_CR#;VU0hT_S^}vcOOOyn3BFR+K$4st{lY@xV`5TL zQnL~g)0|yA^+`rlV+%Wm6V1xb#@4~i+?qht6PJ*E_|p@Bl(d45wt<_27cd*>7wPR4 z;_Dse?H%gD3ib(z43AFn_6yh5HRk0Ndisc81F2* zqabiNEZlMoc6WczC(xH)L_khOAuKBL-P9aMvjdgfB;h?7H#axEy}dy}0eX5m`ue)& z7RC%FJvJ_)t-XmGAYEPF3nQ57_xI=K=2IwCU0wae#N^%GT@Ic1;uLBJzbZ0LR_*A+ z-u&Fu#Q4bY&_G9f>-(jJt4m0)aby;F_G|D4FjubGg(;;tWnEc*Bu0vCgmSdP`fzB) zVl~xN(URgyP#}<##)%54z2KEqR>0fR-Gc*TfY#Ks?DUM>fWR(3r5ug!rUPXJ?j_ zl+v>&{3tb?m8mVw#D=0zwKSobkWC?ipi)dd-M!1oDi#)&;7h{t!Zq08u~9S@BO)p4 z;^E%e)pHCHB=^GK(Ry-n(%RY@9UU1N8J?4y!4XCiBRjhrAYi$?Lwf~N`2MJ^t+lqc zMk0}pj*e4PZ@E$zlo{N7aejL5H#p+#`s!k3d8x0ryS2HcsiCp0rFCM7fpBHcN(;&2!g5+fsp;M83tF(QJhypN@zZ*FSu=Mx?onFywu zl$;S56mIX}>geR|Msg}AB*J^*?$NH|WVG?kpR%d6sxw)yNq&Upf{QTUHf}czIud<&jRQGoeh6aWqt<~1j*;wB) z_Ga?*7!tsLEAU($v4=Y*9LzkWUZ^Vx5_J@pt0Oc_DKJ@SV6C8? zU-0sO`_~^|KIbRkv_nH8JK8z4a!d$Td(H%Ar)OpufOn@^78RAOu55lP!vOW%JAXa? zzQ(`_ut4B)l>o(K7|^~PD+fXL z;cRqxvb&?dv#qzIwWqwKrlqlKckAfr;9_oibzy3Ka&W$@ZphC$N<)@}lpyITTN|U@ zIkdv6rKy2OC}V(DDRB)+F*M(Ed0_!XZKAO~!^7JlFE`9*2z zIaV}VMFmw@Bn@=PmL^tyUIEsYcB;xK;2NyekYp4SozUMm%%uq=0oQt7*VdStn~{wu zX=xcCv(C?fEY9(X)@vvza$m$fAYOt=4Muu zqgir-g88}k9}dqBc8@PXn{t#c2WQ7O??9{XhQv!;g=6pS=R#`jnSX6@j*<*+xgk8M&UGESMbP;*txke14A};ws;~TA|eXbw8Q-mcd&kQGyrF+erjPEfyLzB?Q;2r{uM-;+=%P; z=IZ$9u)nu2Gd;_QYyufkSt&*Sm!irFXgbX?F)p*Vx^-mm9q_ohc6fAfwZC(=w!Akw zG+R^ImYSF!>>q1JWff-7J1{H^#-mPlsTvuA+@Z z=pYmcXf-_|#sq^TlZfWp8b-PpQ=F=inv@Pn!!CJM5dRR?8H5F~HZv|`R zdoHaYr(tQ%1X{r%f@5XcyO>+hUENqfG57?K2R^<*Ah7=YpWgzWJbeiuYUx?i8G(Vp z@$reEU6PYi!MGe8oJ>s29GH$FAz`tx@hp}Pk*E!Ulc=aT$-pQyB&xcqaq8XT`r6*| z`;D>Dcf&*DGt-MOCC^S_UA_Tz2*=3CsEdmm4ukdb@a*d71n#)B0%cbhoWLJWSC>~{ z9f*#MT9})MaUhx_Y=Pe!_UX~(;@nC{YhQXwK2+cd7(+E>f+ms#)ZjHt(5i;&$|STZ z8H=JqG2et@!_m~%ZVq-Vrj4hyg|oRa(}cpnX^>$P1uzT0rAx5T!kR*dfF|o;jY$Mc zLoGU)Xk)75U}MDk8?88uK2lXzRtgKI3St0gRZ_r!6Y%p3i;7NwSr3a3=vEU`8Z2Bu zbVg<#Fbl!}Tmcv?R|$ada&mSL4U2$@TwY#TU)P+MR{*sfx|I#n&N0wGC@v-u$Tc;w z6qk^al#(+f8+&;8XJi!gbPq2rZfvaYPQ9B4THn2$13$37w!OBx1!nvH{Yphe6~mTE zHlXu&ik87llbCjQJOgJm$Z?qDiT8kus z2LOWw>H6x0jItt@Zsi0^7MLekE6@rmRZU&f%9;V%0Ahs7syb*vX@O(?mw)|tVG&6^ zeTu!4J5&;~vJ0A<+J^?mpqLoo9R$jhVQmkZ6iSe=B01Q*Dkvf#j?mIIq!`m2U3^kf za(a43mY26@W|rZ%M&G>c?iv7_U6_0SZgQrxy$1?8M2r>+f%Nh8Xm4)V-`=`BJ>jN5 zI8*+9n(u9G`+E8)$tvPCaHbR!Cf(kd>FUDtVp_YI8rs0F9N3`)mSv=CV@`ImvtT(` z`#akPIa&uXEd1=ueI2ZVT^S+vwqDK-EKj!pUsh;vKy*w*T6%I}URFg$N>OHN2`p2E zxvvXysyXXb_G{Su04Oc2uPAIN%Wo*osRdNzH~x)QEdm*XHdIm6fh7wJ7HH)<0e5#l zSgazW;=wRg)il6?fV#* z%ZN+E3os&4LBQz~jmbI|RDCOBl8rfq35$%EQ&7_409tZGYZYt z+7WUqZmvEk6hTB>P7Q@Opjc|+^wAhy2WRi%;_AA(HppY;=9FY+tZ4)s(f?S9Ua3cip3>xu&n9akz)G<7RYZaC~g!Ey&g3 zfidX69iN+@S%JTHkA2Vv0B&a$DrWbmfR&}4_X`{MSo<%ug0%vzT!RJQ%FidQhQQg{ zIs>5~s#x9u5LW&4^HUH~2vs!1TJ9b`V5+%!#UKwry8h39`_mu(<=^;)Bn^!$z+-|_ zhFB{nvnVz^(Zk6bXf@QK8t56@&>b0eE;yW?oT9o0R>#PMj=_^;6_CIzlj#OgnUkX@ z+yGk_s7B^mcs&gy4yl4tl2sNLlHlWg`STCoYp5e&gH2~^Q+In?M@w^mch68?|C_;~ zi8rHAU3X?WSr}U~teKG^QHe3hNik{JsfCrrbuD#W?Tx){4Lw~g0|Q;7!@U!u1Men= zXWos@&P*)Lzk9ztx4E{swYt2qxxNMU8;Gy=cRw8L9YI9Esb<5hfCSe0-TBEiSJ*

!$+XM!|2>`8Nt>8%^tkTgn zG$fmVJOJSZmJ6QYkAM0r(E9zur;;)VJl??6)Yik*-_JXYK;@1A=?$OTC+2P6M(dG3S^%W9MVA`zh+>ES@}^S~gKz3+mYfBec4{vI0?6Sr6J5yprF4`;7OA zu(}#nM~6bAF)b|_Xf*NJGr{k^d-U)jFJw&A)UcerMrJk;GFVvH!T2Bl_{}%pKY8?2 zKuk;loHcARgsp~>81Ux11OOm?tC~uy0|#LJUKW$Jp`k=f@0JiXTRaE zwLfm&wOyG#&M7=Ji7&2Nm@h$pa5uvU6>l` z`fAErAk9=1@WT8GP$PNtGgQ=6K-KDK89^k*_d@KuZyx{h)61V9zIe_nEFq?dR3qqW zQ;B#(NG*Q-m!IM4x8MBy@W-bDd}7K9NSvk)2yYBpOPEg_%16A9Up@KdB?piBo<4c? zk%`M8w0e(B0S|)~Ji@?I#((1wf@$t#U<<%XBsC)dk zivNE%#=kExav6je?Sy@IbjOD5%?I`!6dCr}x4Z1??c1xZo6BwX+1BmJ#`TBQi~Yru z?fDOz)4M<`1qi(d6==nv^wm{}z%1xq2|?xOzX-qN6_WsdJ>G z^QQujA3cWz7@wesq^yD_Ud!CV+RdE>#T5u5qN8IWB+JgsN>5EoN=SsZsGzXAysWme zxU49%I6tkjB(J`@wBdDG!|RHs+UkzxhR){Zp0>8Z?%v6f(WRNW-SzF`{UeAeATP>M z3lLO%n#lLI$*Cp&{WehU4q3Pv5;CX2hn1uK`GeireZbDt{?^;w&C#8; z;mwu7_4objOM{yW0~_;wYt!A!Q=Ln1I~L!zF95BETv~~E3IL~JfKt_l2t-8zt00M% z6jG890WK7U1*C=er6fcZpndsF1Wunl5q!Zb{QRj9?_&YzaIZ>O^t&+9i#o71D(ySwbkwQHKTn!Z-)m*x_c%Ehh|16mfp>* z&MvLbFK@qJKiWJzKRoAT`nhD?vj$lg7`Z!PUqP|);OcVk5^%nAezbnHx464Gx4AvH z1DiXyW_JOw#dB?9V`+GGu5WRsdvUsJVY+=`vUzr_ZhE+Sa-d?Yx9m+v;ZSSAa7*q$ zeO7OMMt4nmS6zA+hgL43c&s4-M}}I9CQ=WWMIy8`RrRn?tVil6yAupq*s*FJ> zX)4L1kg9kUMGZ9-EG$9Lhg3fSOVZadCX-B!sWcN~EAHk_D;m?*#tF(ZF3v1_rWcQ?mPc0R`xhqL-wrp74%Ced)Qt`` zj1MZPnG)odah#mZxuYWL!}}DeP2yJNj<+?cCzb60Fx7OB-v8E7N1s16_SX zojrr?obu{Wd-rfh-$-Zw8(8Rj-%R$6zwMux9-3YpUtFD92P{o2&yOrE4sx*6|DI!U z9rKgTQ*Y|VhpI*gDn|OshI&c{`%8xh%g6dl-*o2pH)nEoc~+)0l_xcoCNvbq)aQp+ zXZV+;co)UH<^q7SIH%kghula;R=90Om`!G=b!M=2R)Be$k7+W?D8bVx-jf{1p%vKt z1T7qe3mpO#0zQV7t24th$R{i!C?-5GIyy8SIw5`$&_ZPesN@pr>k$a|z&P9!7o8Lr zn+%myXt@AYpO}yafWg=p&Q%ymPRigSJS;jiI1JV$mZ!JBZ%}evN=-#=f7ihH@Wj;U z6juywzTaA3T%UO}JqC+~3^a*U;b7(AU>6IMg~i z(KR*RJvH1i2BtSyJu*-^bkE>GXJKDUPIrB3Yjr|XWn4pXcujs_d5&*cu3tsAPg$B< zL4s3uI3qR0E;*Q<9AKU3Z<*+07Uw~Ub|HncjN$;^)IX-0zd~wQj)Ww5)&7jnh>9! zoRpQ8nwycH8ylSv6CN8892puA78M!;52-4yhON&aIs00A2ipfmd)~YonwlDZH{3ql zR^8TE(om3;pB|Q$9+{C7otqh*lOI=DnpRmDUs{k@R+3azlKQ$VrM5V^rYNDRAhxnF zs=Oe)JU6&3JFq0nuOy9C5bu&5WuG4Jm=$K7>`jSs*AMm33wPHIb0!8euzuE>ESiRw z6`EzG>1~S*VBmw8gir=P#9c4WLqFbAKk2{lsGsJklflx<@*(B;8Rq$u^P%-M$mPP+DACb+wFuR#UPSM1JruAStmxReo81Iego`_Wtgco}Sj;#_Fc(lGng8?BM~H z^D_zn#o1-~Ii(PK=46%to6rKesY$ts2^q-=*|9MxP$3Tvih_Sh2oHlmDGB})0Ro|f z=%o0Fgt+i{XkqJiUP^vRR%v}%LqmCEb46=gbqAQ?K+8~9O*iZ`E=?(ejjOS&xN!H# zNcWg9*Qj8J2!CdHutQX+b6libVwg)}uzf7-Y7elB31Y_iSw(qUgnOEVc^HSfQ$n0c z0d_=Rnx==7ZU|k|$5`2wqTpnK@S>yqY|w$$NPlajuO-rlrshjS`rB!QI^ZIm2r&Q$ zY>Xo=)(M~BOh|MgrhEiGg{7VCqm%EiUleFi5^(PF6}rMnOeST|r4lSw=>~TvNxu(8R<c1}I=;VrF4wW9Q)H;sz?% zD!{d!pzFb!U9xX3zTPI5o8roG<0MW4oqZMDikqloVbuf*=gfJ(V&YTRE(2~ znmD<{#3dx9RMpfqG__1j&CD$#!-Fd+4@%eG*5|$BFEA~W#iKTx&(db6V8^-3 zv3oxB7M)|?wENS|a5kGq+vd!&`4t#sV4BO3dF`ap`DbhvtMr%MRpB%{6!;dE^b3IBSD9>@tB?ox*G)t&06}p_}K=#C08a@6#Y<`IEA4q<$7{q z)rS}F%Vq4MSIBFM9*zCc^zZq5&&GKoJV7lxXS{MUjGD4==QOc1MZ0QwGjIQx_D_g? z_MZB!5kJ|xoR4NbnAwtcJ$PO7(G?F=@9vsno4I}Si|~wf9fF_FORV#sd~J5~yk%1| zZ@$QUyEI#9jo*O{ZrZO{HQW4l^FNK+rMF~*Fw6OEMi#jX9$Z_or%Cs$=@lb4$DdB0 zGIsA+>}zr7iNu3PeXcr7OwNSVESaRK?UE9AY3a7Ea`EKvIXli7cTHSy=|E2Ww)z&) z)l&BD_XB&EPV?7NJmwoc|By9Dc695M^a)odPujUB(Oi}6w4s*3MM|wX4Kp-c9(kde+;L9kabJ tU9!H(kaUOnY(c=O4>9gK>3qwZ!}7OU?>^3!*{L*T@ydWnn{)o(1OR*Q5_tds literal 0 HcmV?d00001 diff --git a/hubconf.py b/hubconf.py new file mode 100644 index 00000000..f8db8703 --- /dev/null +++ b/hubconf.py @@ -0,0 +1,130 @@ +from pathlib import PurePath + +import torch +import yaml + + +dependencies = ['torch', 'pytorch_lightning', 'timm'] + + +def _get_config(model: str, experiment: str = None, **kwargs): + """Emulates hydra config resolution""" + root = PurePath(__file__).parent + with open(root.joinpath('configs/main.yaml'), 'r') as f: + config = yaml.load(f, yaml.Loader)['model'] + with open(root.joinpath(f'configs/charset/94_full.yaml'), 'r') as f: + config.update(yaml.load(f, yaml.Loader)['model']) + with open(root.joinpath(f'configs/model/{model}.yaml'), 'r') as f: + config.update(yaml.load(f, yaml.Loader)) + if experiment is not None: + with open(root.joinpath(f'configs/experiment/{experiment}.yaml'), 'r') as f: + config.update(yaml.load(f, yaml.Loader)['model']) + config.update(kwargs) + return config + + +def parseq_tiny(pretrained: bool = False, decode_ar: bool = True, refine_iters: int = 1, **kwargs): + """ + PARSeq tiny model (img_size=128x32, patch_size=8x4, d_model=192) + @param pretrained: (bool) Use pretrained weights + @param decode_ar: (bool) use AR decoding + @param refine_iters: (int) number of refinement iterations to use + """ + from strhub.models.parseq.system import PARSeq + config = _get_config('parseq', 'parseq-tiny', decode_ar=decode_ar, refine_iters=refine_iters, **kwargs) + model = PARSeq(**config) + if pretrained: + checkpoint = torch.hub.load_state_dict_from_url( + url='https://github.com/baudm/parseq/releases/download/v1.0.0/parseq_tiny-e7a21b54.pt', + map_location='cpu', check_hash=True + ) + model.load_state_dict(checkpoint) + return model + + +def parseq(pretrained: bool = False, decode_ar: bool = True, refine_iters: int = 1, **kwargs): + """ + PARSeq base model (img_size=128x32, patch_size=8x4, d_model=384) + @param pretrained: (bool) Use pretrained weights + @param decode_ar: (bool) use AR decoding + @param refine_iters: (int) number of refinement iterations to use + """ + from strhub.models.parseq.system import PARSeq + config = _get_config('parseq', decode_ar=decode_ar, refine_iters=refine_iters, **kwargs) + model = PARSeq(**config) + if pretrained: + checkpoint = torch.hub.load_state_dict_from_url( + url='https://github.com/baudm/parseq/releases/download/v1.0.0/parseq-bb5792a6.pt', + map_location='cpu', check_hash=True + ) + model.load_state_dict(checkpoint) + return model + + +def abinet(pretrained: bool = False, iter_size: int = 3, **kwargs): + """ + ABINet model (img_size=128x32) + @param pretrained: (bool) Use pretrained weights + @param iter_size: (int) number of refinement iterations to use + """ + from strhub.models.abinet.system import ABINet + config = _get_config('abinet', iter_size=iter_size, **kwargs) + model = ABINet(**config) + if pretrained: + checkpoint = torch.hub.load_state_dict_from_url( + url='https://github.com/baudm/parseq/releases/download/v1.0.0/abinet-1d1e373e.pt', + map_location='cpu', check_hash=True + ) + model.load_state_dict(checkpoint) + return model + + +def trba(pretrained: bool = False, **kwargs): + """ + TRBA model (img_size=128x32) + @param pretrained: (bool) Use pretrained weights + """ + from strhub.models.trba.system import TRBA + config = _get_config('trba', **kwargs) + model = TRBA(**config) + if pretrained: + checkpoint = torch.hub.load_state_dict_from_url( + url='https://github.com/baudm/parseq/releases/download/v1.0.0/trba-cfaed284.pt', + map_location='cpu', check_hash=True + ) + model.load_state_dict(checkpoint) + return model + + +def vitstr(pretrained: bool = False, **kwargs): + """ + ViTSTR small model (img_size=128x32, patch_size=8x4, d_model=384) + @param pretrained: (bool) Use pretrained weights + """ + from strhub.models.vitstr.system import ViTSTR + config = _get_config('vitstr', **kwargs) + model = ViTSTR(**config) + if pretrained: + checkpoint = torch.hub.load_state_dict_from_url( + url='https://github.com/baudm/parseq/releases/download/v1.0.0/vitstr-26d0fcf4.pt', + map_location='cpu', check_hash=True + ) + model.load_state_dict(checkpoint) + return model + + +def crnn(pretrained: bool = False, **kwargs): + """ + CRNN model (img_size=128x32) + @param pretrained: (bool) Use pretrained weights + """ + from strhub.models.crnn.system import CRNN + config = _get_config('crnn', **kwargs) + model = CRNN(**config) + if pretrained: + checkpoint = torch.hub.load_state_dict_from_url( + url='https://github.com/baudm/parseq/releases/download/v1.0.0/crnn-679d0e31.pt', + map_location='cpu', check_hash=True + ) + model.load_state_dict(checkpoint) + return model diff --git a/read.py b/read.py new file mode 100755 index 00000000..d9e7377c --- /dev/null +++ b/read.py @@ -0,0 +1,49 @@ +#!/usr/bin/env python3 +# Scene Text Recognition Model Hub +# Copyright 2022 Darwin Bautista +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# https://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +import argparse + +from PIL import Image + +from strhub.data.module import SceneTextDataModule +from strhub.models.utils import load_from_checkpoint, parse_model_args + + +def main(): + parser = argparse.ArgumentParser() + parser.add_argument('checkpoint', help='Model checkpoint') + parser.add_argument('--images', nargs='+', help='Images to read') + parser.add_argument('--device', default='cuda') + args, unknown = parser.parse_known_args() + kwargs = parse_model_args(unknown) + print(f'Additional keyword arguments: {kwargs}') + + model = load_from_checkpoint(args.checkpoint, **kwargs).eval().to(args.device) + model.freeze() # disable autograd + img_transform = SceneTextDataModule.get_transform(model.hparams.img_size) + + for fname in args.images: + # Load image and prepare for input + image = Image.open(fname).convert('RGB') + image = img_transform(image).unsqueeze(0).to(args.device) + + p = model(image).softmax(-1) + pred, p = model.tokenizer.decode(p) + print(f'{fname}: {pred[0]}') + + +if __name__ == '__main__': + main() diff --git a/requirements.txt b/requirements.txt new file mode 100644 index 00000000..0e2110ee --- /dev/null +++ b/requirements.txt @@ -0,0 +1,15 @@ +torch==1.10.2 +torchtext==0.11.2 +torchvision==0.11.3 +pytorch-lightning==1.4.9 +torchmetrics==0.6.2 +timm==0.4.12 +nltk==3.6.5 +lmdb==1.2.1 +opencv-python==4.5.4.58 +imgaug==0.4.0 +hydra-core==1.1.1 +fvcore==0.1.5.post20211023 +ray==1.7.1 +ax-platform==0.2.2 +SQLAlchemy==1.4.26 diff --git a/setup.cfg b/setup.cfg new file mode 100644 index 00000000..268bbb91 --- /dev/null +++ b/setup.cfg @@ -0,0 +1,43 @@ +[tool:pytest] +norecursedirs = + .git + dist + build +addopts = + --strict + --doctest-modules + --durations=0 + +[coverage:report] +exclude_lines = + pragma: no-cover + pass + +[flake8] +max-line-length = 120 +exclude = .tox,*.egg,build,temp +select = E,W,F +doctests = True +verbose = 2 +# https://pep8.readthedocs.io/en/latest/intro.html#error-codes +format = pylint +# see: https://www.flake8rules.com/ +ignore = + E731 # Do not assign a lambda expression, use a def + W504 # Line break occurred after a binary operator + F401 # Module imported but unused + F841 # Local variable name is assigned to but never used + W605 # Invalid escape sequence 'x' + +# setup.cfg or tox.ini +[check-manifest] +ignore = + *.yml + .github + .github/* + +[metadata] +license_file = LICENSE +description-file = README.md +# long_description = file:README.md +# long_description_content_type = text/markdown diff --git a/setup.py b/setup.py new file mode 100644 index 00000000..05945655 --- /dev/null +++ b/setup.py @@ -0,0 +1,14 @@ +#!/usr/bin/env python + +from setuptools import setup, find_packages + +setup( + name='strhub', + version='0.1.0', + description='Scene Text Recognition Model Hub: A collection of deep learning models for Scene Text Recognition', + author='Darwin Bautista', + author_email='baudm@users.noreply.github.com', + url='https://github.com/baudm/parseq', + install_requires=['torch~=1.10.2', 'pytorch-lightning~=1.4.9', 'timm~=0.4.12'], + packages=find_packages(), +) diff --git a/strhub/__init__.py b/strhub/__init__.py new file mode 100644 index 00000000..e69de29b diff --git a/strhub/data/__init__.py b/strhub/data/__init__.py new file mode 100644 index 00000000..e69de29b diff --git a/strhub/data/aa_overrides.py b/strhub/data/aa_overrides.py new file mode 100644 index 00000000..524afc23 --- /dev/null +++ b/strhub/data/aa_overrides.py @@ -0,0 +1,46 @@ +# Scene Text Recognition Model Hub +# Copyright 2022 Darwin Bautista +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# https://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +"""Extends default ops to accept optional parameters.""" +from functools import partial + +from timm.data.auto_augment import _MAX_LEVEL, _randomly_negate, LEVEL_TO_ARG, NAME_TO_OP, rotate + + +def rotate_expand(img, degrees, **kwargs): + """Rotate operation with expand=True to avoid cutting off the characters""" + kwargs['expand'] = True + return rotate(img, degrees, **kwargs) + + +def _level_to_arg(level, hparams, key, default): + magnitude = hparams.get(key, default) + level = (level / _MAX_LEVEL) * magnitude + level = _randomly_negate(level) + return level, + + +def apply(): + # Overrides + NAME_TO_OP.update({ + 'Rotate': rotate_expand + }) + LEVEL_TO_ARG.update({ + 'Rotate': partial(_level_to_arg, key='rotate_deg', default=30.), + 'ShearX': partial(_level_to_arg, key='shear_x_pct', default=0.3), + 'ShearY': partial(_level_to_arg, key='shear_y_pct', default=0.3), + 'TranslateXRel': partial(_level_to_arg, key='translate_x_pct', default=0.45), + 'TranslateYRel': partial(_level_to_arg, key='translate_y_pct', default=0.45), + }) diff --git a/strhub/data/augment.py b/strhub/data/augment.py new file mode 100644 index 00000000..160aa653 --- /dev/null +++ b/strhub/data/augment.py @@ -0,0 +1,111 @@ +# Scene Text Recognition Model Hub +# Copyright 2022 Darwin Bautista +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# https://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +from functools import partial + +import imgaug.augmenters as iaa +import numpy as np +from PIL import ImageFilter, Image +from timm.data import auto_augment + +from strhub.data import aa_overrides + +aa_overrides.apply() + +_OP_CACHE = {} + + +def _get_op(key, factory): + try: + op = _OP_CACHE[key] + except KeyError: + op = factory() + _OP_CACHE[key] = op + return op + + +def _get_param(level, img, max_dim_factor, min_level=1): + max_level = max(min_level, max_dim_factor * max(img.size)) + return round(min(level, max_level)) + + +def gaussian_blur(img, radius, **__): + radius = _get_param(radius, img, 0.02) + key = 'gaussian_blur_' + str(radius) + op = _get_op(key, lambda: ImageFilter.GaussianBlur(radius)) + return img.filter(op) + + +def motion_blur(img, k, **__): + k = _get_param(k, img, 0.08, 3) | 1 # bin to odd values + key = 'motion_blur_' + str(k) + op = _get_op(key, lambda: iaa.MotionBlur(k)) + return Image.fromarray(op(image=np.asarray(img))) + + +def gaussian_noise(img, scale, **_): + scale = _get_param(scale, img, 0.25) | 1 # bin to odd values + key = 'gaussian_noise_' + str(scale) + op = _get_op(key, lambda: iaa.AdditiveGaussianNoise(scale=scale)) + return Image.fromarray(op(image=np.asarray(img))) + + +def poisson_noise(img, lam, **_): + lam = _get_param(lam, img, 0.2) | 1 # bin to odd values + key = 'poisson_noise_' + str(lam) + op = _get_op(key, lambda: iaa.AdditivePoissonNoise(lam)) + return Image.fromarray(op(image=np.asarray(img))) + + +def _level_to_arg(level, _hparams, max): + level = max * level / auto_augment._MAX_LEVEL + return level, + + +_RAND_TRANSFORMS = auto_augment._RAND_INCREASING_TRANSFORMS.copy() +_RAND_TRANSFORMS.remove('SharpnessIncreasing') # remove, interferes with *blur ops +_RAND_TRANSFORMS.extend([ + 'GaussianBlur', + # 'MotionBlur', + # 'GaussianNoise', + 'PoissonNoise' +]) +auto_augment.LEVEL_TO_ARG.update({ + 'GaussianBlur': partial(_level_to_arg, max=4), + 'MotionBlur': partial(_level_to_arg, max=20), + 'GaussianNoise': partial(_level_to_arg, max=0.1 * 255), + 'PoissonNoise': partial(_level_to_arg, max=40) +}) +auto_augment.NAME_TO_OP.update({ + 'GaussianBlur': gaussian_blur, + 'MotionBlur': motion_blur, + 'GaussianNoise': gaussian_noise, + 'PoissonNoise': poisson_noise +}) + + +def rand_augment_transform(magnitude=5, num_layers=3): + # These are tuned for magnitude=5, which means that effective magnitudes are half of these values. + hparams = { + 'rotate_deg': 30, + 'shear_x_pct': 0.9, + 'shear_y_pct': 0.2, + 'translate_x_pct': 0.10, + 'translate_y_pct': 0.30 + } + ra_ops = auto_augment.rand_augment_ops(magnitude, hparams, transforms=_RAND_TRANSFORMS) + # Supply weights to disable replacement in random selection (i.e. avoid applying the same op twice) + choice_weights = [1. / len(ra_ops) for _ in range(len(ra_ops))] + return auto_augment.RandAugment(ra_ops, num_layers, choice_weights) diff --git a/strhub/data/dataset.py b/strhub/data/dataset.py new file mode 100644 index 00000000..e36bc31a --- /dev/null +++ b/strhub/data/dataset.py @@ -0,0 +1,126 @@ +# Scene Text Recognition Model Hub +# Copyright 2022 Darwin Bautista +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# https://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. +import glob +import io +import logging +import unicodedata +from pathlib import Path, PurePath +from typing import Callable, Optional, Union + +import lmdb +from PIL import Image +from torch.utils.data import Dataset, ConcatDataset + +from strhub.data.utils import CharsetAdapter + +log = logging.getLogger(__name__) + + +def build_tree_dataset(root: Union[PurePath, str], *args, **kwargs): + try: + kwargs.pop('root') # prevent 'root' from being passed via kwargs + except KeyError: + pass + root = Path(root).absolute() + log.info(f'dataset root:\t{root}') + datasets = [] + for mdb in glob.glob(str(root.joinpath('**/data.mdb')), recursive=True): + mdb = Path(mdb) + ds_name = str(mdb.parent.relative_to(root)) + ds_root = str(mdb.parent.absolute()) + dataset = LmdbDataset(ds_root, *args, **kwargs) + log.info(f'\tlmdb:\t{ds_name}\tnum samples: {len(dataset)}') + datasets.append(dataset) + return ConcatDataset(datasets) + + +class LmdbDataset(Dataset): + """Dataset interface to an LMDB database. + + It supports both labelled and unlabelled datasets. For unlabelled datasets, the image index itself is returned + as the label. Unicode characters are normalized by default. Case-sensitivity is inferred from the charset. + Labels are transformed according to the charset. + """ + + def __init__(self, root: str, charset: str, max_label_len: int, min_image_dim: int = 0, + normalize_unicode: bool = True, unlabelled: bool = False, transform: Optional[Callable] = None, + num_workers: int = 1): + self.env = lmdb.open(root, max_readers=num_workers, max_spare_txns=num_workers, + readonly=True, create=False, readahead=False, meminit=False, lock=False) + self.max_label_len = max_label_len + self.min_image_dim = min_image_dim + self.normalize_unicode = normalize_unicode + self.unlabelled = unlabelled + self.transform = transform + self.labels = [] + self.filtered_index_list = [] + self.num_samples = self._preprocess_labels(charset) + + def __del__(self): + self.env.close() + + def _preprocess_labels(self, charset): + charset_adapter = CharsetAdapter(charset) + with self.env.begin() as txn: + num_samples = int(txn.get('num-samples'.encode())) + if self.unlabelled: + return num_samples + for index in range(num_samples): + index += 1 # lmdb starts with 1 + label_key = f'label-{index:09d}'.encode() + label = txn.get(label_key).decode() + # There shouldn't be any whitespace in the labels but try to remove them for good measure + label = ''.join(label.split()) + # Normalize unicode composites (if any) and convert to compatible ASCII characters + if self.normalize_unicode: + label = unicodedata.normalize('NFKD', label).encode('ascii', 'ignore').decode() + # Filter by length before removing unsupported characters. The original label might be too long. + if len(label) > self.max_label_len: + continue + label = charset_adapter(label) + # We filter out samples which don't contain any supported characters + if not label: + continue + # Filter images that are too small. + if self.min_image_dim > 0: + img_key = f'image-{index:09d}'.encode() + buf = io.BytesIO(txn.get(img_key)) + w, h = Image.open(buf).size + if w < self.min_image_dim or h < self.min_image_dim: + continue + self.labels.append(label) + self.filtered_index_list.append(index) + return len(self.labels) + + def __len__(self): + return self.num_samples + + def __getitem__(self, index): + if self.unlabelled: + label = index + else: + label = self.labels[index] + index = self.filtered_index_list[index] + + img_key = f'image-{index:09d}'.encode() + with self.env.begin() as txn: + imgbuf = txn.get(img_key) + buf = io.BytesIO(imgbuf) + img = Image.open(buf).convert('RGB') + + if self.transform is not None: + img = self.transform(img) + + return img, label diff --git a/strhub/data/module.py b/strhub/data/module.py new file mode 100644 index 00000000..8bf3473b --- /dev/null +++ b/strhub/data/module.py @@ -0,0 +1,102 @@ +# Scene Text Recognition Model Hub +# Copyright 2022 Darwin Bautista +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# https://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +from pathlib import PurePath +from typing import Optional, Callable, Sequence + +import pytorch_lightning as pl +from torch.utils.data import DataLoader +from torchvision import transforms as T + +from .augment import rand_augment_transform +from .dataset import build_tree_dataset, LmdbDataset + + +class SceneTextDataModule(pl.LightningDataModule): + TEST_ABINET = ('IIIT5k', 'SVT', 'IC13_857', 'IC15_1811', 'SVTP', 'CUTE80') + TEST_TRBA = ('IIIT5k', 'SVT', 'IC13_1015', 'IC15_2077', 'SVTP', 'CUTE80') + TEST_NEW = ('ArT', 'COCOv1.4', 'Uber') + TEST_ALL = tuple(set(TEST_ABINET + TEST_TRBA + TEST_NEW)) + + def __init__(self, root_dir: str, train_dir: str, img_size: Sequence[int], max_label_length: int, + charset_train: str, charset_test: str, batch_size: int, num_workers: int, augment: bool, + min_image_dim: int = 0, rotation: int = 0, collate_fn: Optional[Callable] = None): + super().__init__() + self.root_dir = root_dir + self.train_dir = train_dir + self.img_size = tuple(img_size) + self.max_label_length = max_label_length + self.charset_train = charset_train + self.charset_test = charset_test + self.batch_size = batch_size + self.num_workers = num_workers + self.augment = augment + self.min_image_dim = min_image_dim + self.rotation = rotation + self.collate_fn = collate_fn + self._train_dataset = None + self._val_dataset = None + + @staticmethod + def get_transform(img_size: tuple[int], augment: bool = False, rotation: int = 0): + transforms = [] + if augment: + transforms.append(rand_augment_transform()) + if rotation: + transforms.append(lambda img: img.rotate(rotation, expand=True)) + transforms.extend([ + T.Resize(img_size, T.InterpolationMode.BICUBIC), + T.ToTensor(), + T.Normalize(0.5, 0.5) + ]) + return T.Compose(transforms) + + @property + def train_dataset(self): + if self._train_dataset is None: + transform = self.get_transform(self.img_size, self.augment) + root = PurePath(self.root_dir, 'train', self.train_dir) + self._train_dataset = build_tree_dataset(root, self.charset_train, self.max_label_length, + self.min_image_dim, + transform=transform, num_workers=self.num_workers) + return self._train_dataset + + @property + def val_dataset(self): + if self._val_dataset is None: + transform = self.get_transform(self.img_size) + root = PurePath(self.root_dir, 'val') + self._val_dataset = build_tree_dataset(root, self.charset_test, self.max_label_length, + transform=transform, num_workers=self.num_workers) + return self._val_dataset + + def train_dataloader(self): + return DataLoader(self.train_dataset, batch_size=self.batch_size, + num_workers=self.num_workers, shuffle=True, + pin_memory=True, collate_fn=self.collate_fn) + + def val_dataloader(self): + return DataLoader(self.val_dataset, batch_size=self.batch_size, + num_workers=self.num_workers, + pin_memory=True, collate_fn=self.collate_fn) + + def test_dataloaders(self, subset): + transform = self.get_transform(self.img_size, rotation=self.rotation) + root = PurePath(self.root_dir, 'test') + datasets = {s: LmdbDataset(str(root.joinpath(s)), self.charset_test, self.max_label_length, + transform=transform) for s in subset} + return {k: DataLoader(v, batch_size=self.batch_size, num_workers=self.num_workers, + pin_memory=True, collate_fn=self.collate_fn) + for k, v in datasets.items()} diff --git a/strhub/data/utils.py b/strhub/data/utils.py new file mode 100644 index 00000000..7fb77bd7 --- /dev/null +++ b/strhub/data/utils.py @@ -0,0 +1,145 @@ +# Scene Text Recognition Model Hub +# Copyright 2022 Darwin Bautista +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# https://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +import re +from abc import ABC, abstractmethod +from itertools import groupby +from typing import List, Optional + +import torch +from torch import Tensor +from torch.nn.utils.rnn import pad_sequence + + +class CharsetAdapter: + """Transforms labels according to the target charset.""" + + def __init__(self, target_charset) -> None: + super().__init__() + self.lowercase_only = target_charset == target_charset.lower() + self.uppercase_only = target_charset == target_charset.upper() + self.unsupported = f'[^{re.escape(target_charset)}]' + + def __call__(self, label): + if self.lowercase_only: + label = label.lower() + elif self.uppercase_only: + label = label.upper() + # Remove unsupported characters + label = re.sub(self.unsupported, '', label) + return label + + +class BaseTokenizer(ABC): + + def __init__(self, charset: str, specials_first: tuple = (), specials_last: tuple = ()) -> None: + self._itos = specials_first + tuple(charset) + specials_last + self._stoi = {s: i for i, s in enumerate(self._itos)} + + def __len__(self): + return len(self._itos) + + def _tok2ids(self, tokens: str) -> List[int]: + return [self._stoi[s] for s in tokens] + + def _ids2tok(self, token_ids: List[int]) -> str: + return ''.join([self._itos[i] for i in token_ids]) + + @abstractmethod + def encode(self, labels: List[str], device: Optional[torch.device] = None) -> Tensor: + """Encode a batch of labels to a representation suitable for the model. + + Args: + labels: List of labels. Each can be of arbitrary length. + device: Create tensor on this device. + + Returns: + Batched tensor representation padded to the max label length. Shape: N, L + """ + raise NotImplementedError + + @abstractmethod + def _filter(self, probs: Tensor, ids: Tensor) -> tuple[Tensor, list[int]]: + """Internal method which performs the necessary filtering prior to decoding.""" + raise NotImplementedError + + def decode(self, token_dists: Tensor) -> tuple[list[str], list[Tensor]]: + """Decode a batch of token distributions. + + Args: + token_dists: softmax probabilities over the token distribution. Shape: N, L, C + + Returns: + list of string labels (arbitrary length) and + their corresponding sequence probabilities as a list of Tensors + """ + batch_tokens = [] + batch_probs = [] + for dist in token_dists: + probs, ids = dist.max(-1) # greedy selection + probs, ids = self._filter(probs, ids) + tokens = self._ids2tok(ids) + batch_tokens.append(tokens) + batch_probs.append(probs) + return batch_tokens, batch_probs + + +class Tokenizer(BaseTokenizer): + BOS = '' + EOS = '' + PAD = '' + + def __init__(self, charset: str) -> None: + specials_first = (self.EOS,) + specials_last = (self.BOS, self.PAD) + super().__init__(charset, specials_first, specials_last) + self.eos_id, self.bos_id, self.pad_id = [self._stoi[s] for s in specials_first + specials_last] + + def encode(self, labels: List[str], device: Optional[torch.device] = None) -> Tensor: + batch = [torch.as_tensor([self.bos_id] + self._tok2ids(y) + [self.eos_id], dtype=torch.long, device=device) + for y in labels] + return pad_sequence(batch, batch_first=True, padding_value=self.pad_id) + + def _filter(self, probs: Tensor, ids: Tensor) -> tuple[Tensor, list[int]]: + ids = ids.tolist() + try: + eos_idx = ids.index(self.eos_id) + except ValueError: + eos_idx = len(ids) # Nothing to truncate. + # Truncate after EOS + ids = ids[:eos_idx] + probs = probs[:eos_idx + 1] # but include prob. for EOS (if it exists) + return probs, ids + + +class CTCTokenizer(BaseTokenizer): + BLANK = '' + + def __init__(self, charset: str) -> None: + # BLANK uses index == 0 by default + super().__init__(charset, specials_first=(self.BLANK,)) + self.blank_id = self._stoi[self.BLANK] + + def encode(self, labels: List[str], device: Optional[torch.device] = None) -> Tensor: + # We use a padded representation since we don't want to use CUDNN's CTC implementation + batch = [torch.as_tensor(self._tok2ids(y), dtype=torch.long, device=device) for y in labels] + return pad_sequence(batch, batch_first=True, padding_value=self.blank_id) + + def _filter(self, probs: Tensor, ids: Tensor) -> tuple[Tensor, list[int]]: + # Best path decoding: + ids = list(zip(*groupby(ids.tolist())))[0] # Remove duplicate tokens + ids = [x for x in ids if x != self.blank_id] # Remove BLANKs + # `probs` is just pass-through since all positions are considered part of the path + return probs, ids diff --git a/strhub/models/__init__.py b/strhub/models/__init__.py new file mode 100644 index 00000000..e69de29b diff --git a/strhub/models/abinet/LICENSE b/strhub/models/abinet/LICENSE new file mode 100644 index 00000000..2f1d4adb --- /dev/null +++ b/strhub/models/abinet/LICENSE @@ -0,0 +1,25 @@ +ABINet for non-commercial purposes + +Copyright (c) 2021, USTC +All rights reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions are met: + +1. Redistributions of source code must retain the above copyright notice, this + list of conditions and the following disclaimer. + +2. Redistributions in binary form must reproduce the above copyright notice, + this list of conditions and the following disclaimer in the documentation + and/or other materials provided with the distribution. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" +AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE +IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE +DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE +FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL +DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR +SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER +CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, +OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE +OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. diff --git a/strhub/models/abinet/__init__.py b/strhub/models/abinet/__init__.py new file mode 100644 index 00000000..60481103 --- /dev/null +++ b/strhub/models/abinet/__init__.py @@ -0,0 +1,13 @@ +r""" +Fang, Shancheng, Hongtao, Xie, Yuxin, Wang, Zhendong, Mao, and Yongdong, Zhang. +"Read Like Humans: Autonomous, Bidirectional and Iterative Language Modeling for Scene Text Recognition." . +In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) (pp. 7098-7107).2021. + +https://arxiv.org/abs/2103.06495 + +All source files, except `system.py`, are based on the implementation listed below, +and hence are released under the license of the original. + +Source: https://github.com/FangShancheng/ABINet +License: 2-clause BSD License (see included LICENSE file) +""" diff --git a/strhub/models/abinet/attention.py b/strhub/models/abinet/attention.py new file mode 100644 index 00000000..eab942cf --- /dev/null +++ b/strhub/models/abinet/attention.py @@ -0,0 +1,100 @@ +import torch +import torch.nn as nn + +from .transformer import PositionalEncoding + + +class Attention(nn.Module): + def __init__(self, in_channels=512, max_length=25, n_feature=256): + super().__init__() + self.max_length = max_length + + self.f0_embedding = nn.Embedding(max_length, in_channels) + self.w0 = nn.Linear(max_length, n_feature) + self.wv = nn.Linear(in_channels, in_channels) + self.we = nn.Linear(in_channels, max_length) + + self.active = nn.Tanh() + self.softmax = nn.Softmax(dim=2) + + def forward(self, enc_output): + enc_output = enc_output.permute(0, 2, 3, 1).flatten(1, 2) + reading_order = torch.arange(self.max_length, dtype=torch.long, device=enc_output.device) + reading_order = reading_order.unsqueeze(0).expand(enc_output.size(0), -1) # (S,) -> (B, S) + reading_order_embed = self.f0_embedding(reading_order) # b,25,512 + + t = self.w0(reading_order_embed.permute(0, 2, 1)) # b,512,256 + t = self.active(t.permute(0, 2, 1) + self.wv(enc_output)) # b,256,512 + + attn = self.we(t) # b,256,25 + attn = self.softmax(attn.permute(0, 2, 1)) # b,25,256 + g_output = torch.bmm(attn, enc_output) # b,25,512 + return g_output, attn.view(*attn.shape[:2], 8, 32) + + +def encoder_layer(in_c, out_c, k=3, s=2, p=1): + return nn.Sequential(nn.Conv2d(in_c, out_c, k, s, p), + nn.BatchNorm2d(out_c), + nn.ReLU(True)) + + +def decoder_layer(in_c, out_c, k=3, s=1, p=1, mode='nearest', scale_factor=None, size=None): + align_corners = None if mode == 'nearest' else True + return nn.Sequential(nn.Upsample(size=size, scale_factor=scale_factor, + mode=mode, align_corners=align_corners), + nn.Conv2d(in_c, out_c, k, s, p), + nn.BatchNorm2d(out_c), + nn.ReLU(True)) + + +class PositionAttention(nn.Module): + def __init__(self, max_length, in_channels=512, num_channels=64, + h=8, w=32, mode='nearest', **kwargs): + super().__init__() + self.max_length = max_length + self.k_encoder = nn.Sequential( + encoder_layer(in_channels, num_channels, s=(1, 2)), + encoder_layer(num_channels, num_channels, s=(2, 2)), + encoder_layer(num_channels, num_channels, s=(2, 2)), + encoder_layer(num_channels, num_channels, s=(2, 2)) + ) + self.k_decoder = nn.Sequential( + decoder_layer(num_channels, num_channels, scale_factor=2, mode=mode), + decoder_layer(num_channels, num_channels, scale_factor=2, mode=mode), + decoder_layer(num_channels, num_channels, scale_factor=2, mode=mode), + decoder_layer(num_channels, in_channels, size=(h, w), mode=mode) + ) + + self.pos_encoder = PositionalEncoding(in_channels, dropout=0, max_len=max_length) + self.project = nn.Linear(in_channels, in_channels) + + def forward(self, x): + N, E, H, W = x.size() + k, v = x, x # (N, E, H, W) + + # calculate key vector + features = [] + for i in range(0, len(self.k_encoder)): + k = self.k_encoder[i](k) + features.append(k) + for i in range(0, len(self.k_decoder) - 1): + k = self.k_decoder[i](k) + k = k + features[len(self.k_decoder) - 2 - i] + k = self.k_decoder[-1](k) + + # calculate query vector + # TODO q=f(q,k) + zeros = x.new_zeros((self.max_length, N, E)) # (T, N, E) + q = self.pos_encoder(zeros) # (T, N, E) + q = q.permute(1, 0, 2) # (N, T, E) + q = self.project(q) # (N, T, E) + + # calculate attention + attn_scores = torch.bmm(q, k.flatten(2, 3)) # (N, T, (H*W)) + attn_scores = attn_scores / (E ** 0.5) + attn_scores = torch.softmax(attn_scores, dim=-1) + + v = v.permute(0, 2, 3, 1).view(N, -1, E) # (N, (H*W), E) + attn_vecs = torch.bmm(attn_scores, v) # (N, T, E) + + return attn_vecs, attn_scores.view(N, -1, H, W) diff --git a/strhub/models/abinet/backbone.py b/strhub/models/abinet/backbone.py new file mode 100644 index 00000000..debcabd7 --- /dev/null +++ b/strhub/models/abinet/backbone.py @@ -0,0 +1,24 @@ +import torch.nn as nn +from torch.nn import TransformerEncoderLayer, TransformerEncoder + +from .resnet import resnet45 +from .transformer import PositionalEncoding + + +class ResTranformer(nn.Module): + def __init__(self, d_model=512, nhead=8, d_inner=2048, dropout=0.1, activation='relu', backbone_ln=2): + super().__init__() + self.resnet = resnet45() + self.pos_encoder = PositionalEncoding(d_model, max_len=8 * 32) + encoder_layer = TransformerEncoderLayer(d_model=d_model, nhead=nhead, + dim_feedforward=d_inner, dropout=dropout, activation=activation) + self.transformer = TransformerEncoder(encoder_layer, backbone_ln) + + def forward(self, images): + feature = self.resnet(images) + n, c, h, w = feature.shape + feature = feature.view(n, c, -1).permute(2, 0, 1) + feature = self.pos_encoder(feature) + feature = self.transformer(feature) + feature = feature.permute(1, 2, 0).view(n, c, h, w) + return feature diff --git a/strhub/models/abinet/model.py b/strhub/models/abinet/model.py new file mode 100644 index 00000000..cc0cd143 --- /dev/null +++ b/strhub/models/abinet/model.py @@ -0,0 +1,31 @@ +import torch +import torch.nn as nn + + +class Model(nn.Module): + + def __init__(self, dataset_max_length: int, null_label: int): + super().__init__() + self.max_length = dataset_max_length + 1 # additional stop token + self.null_label = null_label + + def _get_length(self, logit, dim=-1): + """ Greed decoder to obtain length from logit""" + out = (logit.argmax(dim=-1) == self.null_label) + abn = out.any(dim) + out = ((out.cumsum(dim) == 1) & out).max(dim)[1] + out = out + 1 # additional end token + out = torch.where(abn, out, out.new_tensor(logit.shape[1], device=out.device)) + return out + + @staticmethod + def _get_padding_mask(length, max_length): + length = length.unsqueeze(-1) + grid = torch.arange(0, max_length, device=length.device).unsqueeze(0) + return grid >= length + + @staticmethod + def _get_location_mask(sz, device=None): + mask = torch.eye(sz, device=device) + mask = mask.float().masked_fill(mask == 1, float('-inf')) + return mask diff --git a/strhub/models/abinet/model_abinet_iter.py b/strhub/models/abinet/model_abinet_iter.py new file mode 100644 index 00000000..1a8523ff --- /dev/null +++ b/strhub/models/abinet/model_abinet_iter.py @@ -0,0 +1,39 @@ +import torch +from torch import nn + +from .model_alignment import BaseAlignment +from .model_language import BCNLanguage +from .model_vision import BaseVision + + +class ABINetIterModel(nn.Module): + def __init__(self, dataset_max_length, null_label, num_classes, iter_size=1, + d_model=512, nhead=8, d_inner=2048, dropout=0.1, activation='relu', + v_loss_weight=1., v_attention='position', v_attention_mode='nearest', + v_backbone='transformer', v_num_layers=2, + l_loss_weight=1., l_num_layers=4, l_detach=True, l_use_self_attn=False, + a_loss_weight=1.): + super().__init__() + self.iter_size = iter_size + self.vision = BaseVision(dataset_max_length, null_label, num_classes, v_attention, v_attention_mode, + v_loss_weight, d_model, nhead, d_inner, dropout, activation, v_backbone, v_num_layers) + self.language = BCNLanguage(dataset_max_length, null_label, num_classes, d_model, nhead, d_inner, dropout, + activation, l_num_layers, l_detach, l_use_self_attn, l_loss_weight) + self.alignment = BaseAlignment(dataset_max_length, null_label, num_classes, d_model, a_loss_weight) + + def forward(self, images): + v_res = self.vision(images) + a_res = v_res + all_l_res, all_a_res = [], [] + for _ in range(self.iter_size): + tokens = torch.softmax(a_res['logits'], dim=-1) + lengths = a_res['pt_lengths'] + lengths.clamp_(2, self.language.max_length) # TODO:move to langauge model + l_res = self.language(tokens, lengths) + all_l_res.append(l_res) + a_res = self.alignment(l_res['feature'], v_res['feature']) + all_a_res.append(a_res) + if self.training: + return all_a_res, all_l_res, v_res + else: + return a_res, all_l_res[-1], v_res diff --git a/strhub/models/abinet/model_alignment.py b/strhub/models/abinet/model_alignment.py new file mode 100644 index 00000000..9ccfa95e --- /dev/null +++ b/strhub/models/abinet/model_alignment.py @@ -0,0 +1,28 @@ +import torch +import torch.nn as nn + +from .model import Model + + +class BaseAlignment(Model): + def __init__(self, dataset_max_length, null_label, num_classes, d_model=512, loss_weight=1.0): + super().__init__(dataset_max_length, null_label) + self.loss_weight = loss_weight + self.w_att = nn.Linear(2 * d_model, d_model) + self.cls = nn.Linear(d_model, num_classes) + + def forward(self, l_feature, v_feature): + """ + Args: + l_feature: (N, T, E) where T is length, N is batch size and d is dim of model + v_feature: (N, T, E) shape the same as l_feature + """ + f = torch.cat((l_feature, v_feature), dim=2) + f_att = torch.sigmoid(self.w_att(f)) + output = f_att * v_feature + (1 - f_att) * l_feature + + logits = self.cls(output) # (N, T, C) + pt_lengths = self._get_length(logits) + + return {'logits': logits, 'pt_lengths': pt_lengths, 'loss_weight': self.loss_weight, + 'name': 'alignment'} diff --git a/strhub/models/abinet/model_language.py b/strhub/models/abinet/model_language.py new file mode 100644 index 00000000..659d4465 --- /dev/null +++ b/strhub/models/abinet/model_language.py @@ -0,0 +1,50 @@ +import torch.nn as nn +from torch.nn import TransformerDecoder + +from .model import Model +from .transformer import PositionalEncoding, TransformerDecoderLayer + + +class BCNLanguage(Model): + def __init__(self, dataset_max_length, null_label, num_classes, d_model=512, nhead=8, d_inner=2048, dropout=0.1, + activation='relu', num_layers=4, detach=True, use_self_attn=False, loss_weight=1.0, + global_debug=False): + super().__init__(dataset_max_length, null_label) + self.detach = detach + self.loss_weight = loss_weight + self.proj = nn.Linear(num_classes, d_model, False) + self.token_encoder = PositionalEncoding(d_model, max_len=self.max_length) + self.pos_encoder = PositionalEncoding(d_model, dropout=0, max_len=self.max_length) + decoder_layer = TransformerDecoderLayer(d_model, nhead, d_inner, dropout, + activation, self_attn=use_self_attn, debug=global_debug) + self.model = TransformerDecoder(decoder_layer, num_layers) + self.cls = nn.Linear(d_model, num_classes) + + def forward(self, tokens, lengths): + """ + Args: + tokens: (N, T, C) where T is length, N is batch size and C is classes number + lengths: (N,) + """ + if self.detach: + tokens = tokens.detach() + embed = self.proj(tokens) # (N, T, E) + embed = embed.permute(1, 0, 2) # (T, N, E) + embed = self.token_encoder(embed) # (T, N, E) + padding_mask = self._get_padding_mask(lengths, self.max_length) + + zeros = embed.new_zeros(*embed.shape) + qeury = self.pos_encoder(zeros) + location_mask = self._get_location_mask(self.max_length, tokens.device) + output = self.model(qeury, embed, + tgt_key_padding_mask=padding_mask, + memory_mask=location_mask, + memory_key_padding_mask=padding_mask) # (T, N, E) + output = output.permute(1, 0, 2) # (N, T, E) + + logits = self.cls(output) # (N, T, C) + pt_lengths = self._get_length(logits) + + res = {'feature': output, 'logits': logits, 'pt_lengths': pt_lengths, + 'loss_weight': self.loss_weight, 'name': 'language'} + return res diff --git a/strhub/models/abinet/model_vision.py b/strhub/models/abinet/model_vision.py new file mode 100644 index 00000000..3190677c --- /dev/null +++ b/strhub/models/abinet/model_vision.py @@ -0,0 +1,45 @@ +from torch import nn + +from .attention import PositionAttention, Attention +from .backbone import ResTranformer +from .model import Model +from .resnet import resnet45 + + +class BaseVision(Model): + def __init__(self, dataset_max_length, null_label, num_classes, + attention='position', attention_mode='nearest', loss_weight=1.0, + d_model=512, nhead=8, d_inner=2048, dropout=0.1, activation='relu', + backbone='transformer', backbone_ln=2): + super().__init__(dataset_max_length, null_label) + self.loss_weight = loss_weight + self.out_channels = d_model + + if backbone == 'transformer': + self.backbone = ResTranformer(d_model, nhead, d_inner, dropout, activation, backbone_ln) + else: + self.backbone = resnet45() + + if attention == 'position': + self.attention = PositionAttention( + max_length=self.max_length, + mode=attention_mode + ) + elif attention == 'attention': + self.attention = Attention( + max_length=self.max_length, + n_feature=8 * 32, + ) + else: + raise ValueError(f'invalid attention: {attention}') + + self.cls = nn.Linear(self.out_channels, num_classes) + + def forward(self, images, *args): + features = self.backbone(images) # (N, E, H, W) + attn_vecs, attn_scores = self.attention(features) # (N, T, E), (N, T, H, W) + logits = self.cls(attn_vecs) # (N, T, C) + pt_lengths = self._get_length(logits) + + return {'feature': attn_vecs, 'logits': logits, 'pt_lengths': pt_lengths, + 'attn_scores': attn_scores, 'loss_weight': self.loss_weight, 'name': 'vision'} diff --git a/strhub/models/abinet/resnet.py b/strhub/models/abinet/resnet.py new file mode 100644 index 00000000..59bf3889 --- /dev/null +++ b/strhub/models/abinet/resnet.py @@ -0,0 +1,72 @@ +import math +from typing import Optional, Callable + +import torch.nn as nn +from torchvision.models import resnet + + +class BasicBlock(resnet.BasicBlock): + + def __init__(self, inplanes: int, planes: int, stride: int = 1, downsample: Optional[nn.Module] = None, + groups: int = 1, base_width: int = 64, dilation: int = 1, + norm_layer: Optional[Callable[..., nn.Module]] = None) -> None: + super().__init__(inplanes, planes, stride, downsample, groups, base_width, dilation, norm_layer) + self.conv1 = resnet.conv1x1(inplanes, planes) + self.conv2 = resnet.conv3x3(planes, planes, stride) + + +class ResNet(nn.Module): + + def __init__(self, block, layers): + super().__init__() + self.inplanes = 32 + self.conv1 = nn.Conv2d(3, 32, kernel_size=3, stride=1, padding=1, + bias=False) + self.bn1 = nn.BatchNorm2d(32) + self.relu = nn.ReLU(inplace=True) + + self.layer1 = self._make_layer(block, 32, layers[0], stride=2) + self.layer2 = self._make_layer(block, 64, layers[1], stride=1) + self.layer3 = self._make_layer(block, 128, layers[2], stride=2) + self.layer4 = self._make_layer(block, 256, layers[3], stride=1) + self.layer5 = self._make_layer(block, 512, layers[4], stride=1) + + for m in self.modules(): + if isinstance(m, nn.Conv2d): + n = m.kernel_size[0] * m.kernel_size[1] * m.out_channels + m.weight.data.normal_(0, math.sqrt(2. / n)) + elif isinstance(m, nn.BatchNorm2d): + m.weight.data.fill_(1) + m.bias.data.zero_() + + def _make_layer(self, block, planes, blocks, stride=1): + downsample = None + if stride != 1 or self.inplanes != planes * block.expansion: + downsample = nn.Sequential( + nn.Conv2d(self.inplanes, planes * block.expansion, + kernel_size=1, stride=stride, bias=False), + nn.BatchNorm2d(planes * block.expansion), + ) + + layers = [] + layers.append(block(self.inplanes, planes, stride, downsample)) + self.inplanes = planes * block.expansion + for i in range(1, blocks): + layers.append(block(self.inplanes, planes)) + + return nn.Sequential(*layers) + + def forward(self, x): + x = self.conv1(x) + x = self.bn1(x) + x = self.relu(x) + x = self.layer1(x) + x = self.layer2(x) + x = self.layer3(x) + x = self.layer4(x) + x = self.layer5(x) + return x + + +def resnet45(): + return ResNet(BasicBlock, [3, 4, 6, 6, 3]) diff --git a/strhub/models/abinet/system.py b/strhub/models/abinet/system.py new file mode 100644 index 00000000..ed6eea03 --- /dev/null +++ b/strhub/models/abinet/system.py @@ -0,0 +1,170 @@ +# Scene Text Recognition Model Hub +# Copyright 2022 Darwin Bautista +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# https://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +import logging +import math +from typing import Any + +import torch +import torch.nn.functional as F +from torch import Tensor, nn +from torch.optim import AdamW +from torch.optim.lr_scheduler import OneCycleLR + +from pytorch_lightning.utilities.types import STEP_OUTPUT +from timm.optim.optim_factory import add_weight_decay + +from strhub.models.base import CrossEntropySystem +from strhub.models.utils import init_weights +from .model_abinet_iter import ABINetIterModel as Model + +log = logging.getLogger(__name__) + + +class ABINet(CrossEntropySystem): + + def __init__(self, charset_train: str, charset_test: str, max_label_length: int, + batch_size: int, lr: float, warmup_pct: float, weight_decay: float, + iter_size: int, d_model: int, nhead: int, d_inner: int, dropout: float, activation: str, + v_loss_weight: float, v_attention: str, v_attention_mode: str, v_backbone: str, v_num_layers: int, + l_loss_weight: float, l_num_layers: int, l_detach: bool, l_use_self_attn: bool, + l_lr: float, a_loss_weight: float, lm_only: bool = False, **kwargs) -> None: + super().__init__(charset_train, charset_test, batch_size, lr, warmup_pct, weight_decay) + self.save_hyperparameters() + self.max_label_length = max_label_length + self.num_classes = len(self.tokenizer) - 2 # We don't predict nor + self.model = Model(max_label_length, self.eos_id, self.num_classes, iter_size, d_model, nhead, d_inner, + dropout, activation, v_loss_weight, v_attention, v_attention_mode, v_backbone, v_num_layers, + l_loss_weight, l_num_layers, l_detach, l_use_self_attn, a_loss_weight) + self.model.apply(init_weights) + # FIXME: doesn't support resumption from checkpoint yet + self._reset_alignment = True + self._reset_optimizers = True + self.l_lr = l_lr + self.lm_only = lm_only + # Train LM only. Freeze other submodels. + if lm_only: + self.l_lr = lr # for tuning + self.model.vision.requires_grad_(False) + self.model.alignment.requires_grad_(False) + + @property + def _pretraining(self): + # In the original work, VM was pretrained for 8 epochs while full model was trained for an additional 10 epochs. + return self.global_step < (8 / 18) * self.num_training_steps + + @torch.jit.ignore + def no_weight_decay(self): + return {'model.language.proj.weight'} + + def _add_weight_decay(self, model: nn.Module, skip_list=()): + if self.weight_decay: + return add_weight_decay(model, self.weight_decay, skip_list) + else: + return [{'params': model.parameters()}] + + def configure_optimizers(self): + agb = self.trainer.accumulate_grad_batches + # Linear scaling so that the effective learning rate is constant regardless of the number of GPUs used with DDP. + lr_scale = agb * math.sqrt(self.trainer.devices) * self.batch_size / 256. + lr = lr_scale * self.lr + l_lr = lr_scale * self.l_lr + params = [] + params.extend(self._add_weight_decay(self.model.vision)) + params.extend(self._add_weight_decay(self.model.alignment)) + # We use a different learning rate for the LM. + for p in self._add_weight_decay(self.model.language, ('proj.weight',)): + p['lr'] = l_lr + params.append(p) + max_lr = [p.get('lr', lr) for p in params] + optim = AdamW(params, lr) + self.scheduler = OneCycleLR(optim, max_lr, math.ceil(self.num_training_steps / agb), + pct_start=self.warmup_pct, cycle_momentum=False) + return {'optimizer': optim, 'lr_scheduler': {'scheduler': self.scheduler, 'interval': 'step'}} + + def forward(self, images: Tensor, max_length: int = None) -> Tensor: + max_length = self.max_label_length if max_length is None else min(max_length, self.max_label_length) + logits = self.model.forward(images)[0]['logits'] + return logits[:, :max_length + 1] # truncate + + def calc_loss(self, targets, *res_lists) -> Tensor: + total_loss = 0 + for res_list in res_lists: + loss = 0 + if isinstance(res_list, dict): + res_list = [res_list] + for res in res_list: + logits = res['logits'].flatten(end_dim=1) + loss += F.cross_entropy(logits, targets.flatten(), ignore_index=self.pad_id) + loss /= len(res_list) + self.log('loss_' + res_list[0]['name'], loss) + total_loss += res_list[0]['loss_weight'] * loss + return total_loss + + def on_train_batch_start(self, batch: Any, batch_idx: int, dataloader_idx: int) -> None: + if not self._pretraining and self._reset_optimizers: + log.info('Pretraining ends. Updating base LRs.') + self._reset_optimizers = False + # Make base_lr the same for all groups + base_lr = self.scheduler.base_lrs[0] # base_lr of group 0 - VM + self.scheduler.base_lrs = [base_lr] * len(self.scheduler.base_lrs) + + def _prepare_inputs_and_targets(self, labels): + # Use dummy label to ensure sequence length is constant. + dummy = ['0' * self.max_label_length] + targets = self.tokenizer.encode(dummy + list(labels), self.device)[1:] + targets = targets[:, 1:] # remove . Unused here. + # Inputs are padded with eos_id + inputs = torch.where(targets == self.pad_id, self.eos_id, targets) + inputs = F.one_hot(inputs, self.num_classes).float() + lengths = torch.as_tensor(list(map(len, labels)), device=self.device) + 1 # +1 for eos + return inputs, lengths, targets + + def training_step(self, batch, batch_idx) -> STEP_OUTPUT: + images, labels = batch + inputs, lengths, targets = self._prepare_inputs_and_targets(labels) + if self.lm_only: + l_res = self.model.language(inputs, lengths) + loss = self.calc_loss(targets, l_res) + # Pretrain submodels independently first + elif self._pretraining: + # Vision + v_res = self.model.vision(images) + # Language + l_res = self.model.language(inputs, lengths) + # We also train the alignment model to 'satisfy' DDP requirements (all parameters should be used). + # We'll reset its parameters prior to joint training. + a_res = self.model.alignment(l_res['feature'].detach(), v_res['feature'].detach()) + loss = self.calc_loss(targets, v_res, l_res, a_res) + else: + # Reset alignment model's parameters once prior to full model training. + if self._reset_alignment: + log.info('Pretraining ends. Resetting alignment model.') + self._reset_alignment = False + self.model.alignment.apply(init_weights) + all_a_res, all_l_res, v_res = self.model.forward(images) + loss = self.calc_loss(targets, v_res, all_l_res, all_a_res) + self.log('loss', loss) + return loss + + def forward_logits_loss(self, images: Tensor, labels: list[str]) -> tuple[Tensor, Tensor, int]: + if self.lm_only: + inputs, lengths, targets = self._prepare_inputs_and_targets(labels) + l_res = self.model.language(inputs, lengths) + loss = self.calc_loss(targets, l_res) + loss_numel = (targets != self.pad_id).sum() + return l_res['logits'], loss, loss_numel + else: + return super().forward_logits_loss(images, labels) diff --git a/strhub/models/abinet/transformer.py b/strhub/models/abinet/transformer.py new file mode 100644 index 00000000..a920805d --- /dev/null +++ b/strhub/models/abinet/transformer.py @@ -0,0 +1,143 @@ +import math + +import torch +import torch.nn.functional as F +from torch import nn +from torch.nn.modules.transformer import _get_activation_fn + + +class TransformerDecoderLayer(nn.Module): + r"""TransformerDecoderLayer is made up of self-attn, multi-head-attn and feedforward network. + This standard decoder layer is based on the paper "Attention Is All You Need". + Ashish Vaswani, Noam Shazeer, Niki Parmar, Jakob Uszkoreit, Llion Jones, Aidan N Gomez, + Lukasz Kaiser, and Illia Polosukhin. 2017. Attention is all you need. In Advances in + Neural Information Processing Systems, pages 6000-6010. Users may modify or implement + in a different way during application. + + Args: + d_model: the number of expected features in the input (required). + nhead: the number of heads in the multiheadattention models (required). + dim_feedforward: the dimension of the feedforward network model (default=2048). + dropout: the dropout value (default=0.1). + activation: the activation function of intermediate layer, relu or gelu (default=relu). + + Examples:: + >>> decoder_layer = nn.TransformerDecoderLayer(d_model=512, nhead=8) + >>> memory = torch.rand(10, 32, 512) + >>> tgt = torch.rand(20, 32, 512) + >>> out = decoder_layer(tgt, memory) + """ + + def __init__(self, d_model, nhead, dim_feedforward=2048, dropout=0.1, + activation="relu", self_attn=True, siamese=False, debug=False): + super().__init__() + self.has_self_attn, self.siamese = self_attn, siamese + self.debug = debug + if self.has_self_attn: + self.self_attn = nn.MultiheadAttention(d_model, nhead, dropout=dropout) + self.norm1 = nn.LayerNorm(d_model) + self.dropout1 = nn.Dropout(dropout) + self.multihead_attn = nn.MultiheadAttention(d_model, nhead, dropout=dropout) + # Implementation of Feedforward model + self.linear1 = nn.Linear(d_model, dim_feedforward) + self.dropout = nn.Dropout(dropout) + self.linear2 = nn.Linear(dim_feedforward, d_model) + + self.norm2 = nn.LayerNorm(d_model) + self.norm3 = nn.LayerNorm(d_model) + self.dropout2 = nn.Dropout(dropout) + self.dropout3 = nn.Dropout(dropout) + if self.siamese: + self.multihead_attn2 = nn.MultiheadAttention(d_model, nhead, dropout=dropout) + + self.activation = _get_activation_fn(activation) + + def __setstate__(self, state): + if 'activation' not in state: + state['activation'] = F.relu + super().__setstate__(state) + + def forward(self, tgt, memory, tgt_mask=None, memory_mask=None, + tgt_key_padding_mask=None, memory_key_padding_mask=None, + memory2=None, memory_mask2=None, memory_key_padding_mask2=None): + # type: (Tensor, Tensor, Optional[Tensor], Optional[Tensor], Optional[Tensor], Optional[Tensor]) -> Tensor + r"""Pass the inputs (and mask) through the decoder layer. + + Args: + tgt: the sequence to the decoder layer (required). + memory: the sequence from the last layer of the encoder (required). + tgt_mask: the mask for the tgt sequence (optional). + memory_mask: the mask for the memory sequence (optional). + tgt_key_padding_mask: the mask for the tgt keys per batch (optional). + memory_key_padding_mask: the mask for the memory keys per batch (optional). + + Shape: + see the docs in Transformer class. + """ + if self.has_self_attn: + tgt2, attn = self.self_attn(tgt, tgt, tgt, attn_mask=tgt_mask, + key_padding_mask=tgt_key_padding_mask) + tgt = tgt + self.dropout1(tgt2) + tgt = self.norm1(tgt) + if self.debug: self.attn = attn + tgt2, attn2 = self.multihead_attn(tgt, memory, memory, attn_mask=memory_mask, + key_padding_mask=memory_key_padding_mask) + if self.debug: self.attn2 = attn2 + + if self.siamese: + tgt3, attn3 = self.multihead_attn2(tgt, memory2, memory2, attn_mask=memory_mask2, + key_padding_mask=memory_key_padding_mask2) + tgt = tgt + self.dropout2(tgt3) + if self.debug: self.attn3 = attn3 + + tgt = tgt + self.dropout2(tgt2) + tgt = self.norm2(tgt) + tgt2 = self.linear2(self.dropout(self.activation(self.linear1(tgt)))) + tgt = tgt + self.dropout3(tgt2) + tgt = self.norm3(tgt) + + return tgt + + +class PositionalEncoding(nn.Module): + r"""Inject some information about the relative or absolute position of the tokens + in the sequence. The positional encodings have the same dimension as + the embeddings, so that the two can be summed. Here, we use sine and cosine + functions of different frequencies. + .. math:: + \text{PosEncoder}(pos, 2i) = sin(pos/10000^(2i/d_model)) + \text{PosEncoder}(pos, 2i+1) = cos(pos/10000^(2i/d_model)) + \text{where pos is the word position and i is the embed idx) + Args: + d_model: the embed dim (required). + dropout: the dropout value (default=0.1). + max_len: the max. length of the incoming sequence (default=5000). + Examples: + >>> pos_encoder = PositionalEncoding(d_model) + """ + + def __init__(self, d_model, dropout=0.1, max_len=5000): + super().__init__() + self.dropout = nn.Dropout(p=dropout) + + pe = torch.zeros(max_len, d_model) + position = torch.arange(0, max_len, dtype=torch.float).unsqueeze(1) + div_term = torch.exp(torch.arange(0, d_model, 2).float() * (-math.log(10000.0) / d_model)) + pe[:, 0::2] = torch.sin(position * div_term) + pe[:, 1::2] = torch.cos(position * div_term) + pe = pe.unsqueeze(0).transpose(0, 1) + self.register_buffer('pe', pe) + + def forward(self, x): + r"""Inputs of forward function + Args: + x: the sequence fed to the positional encoder model (required). + Shape: + x: [sequence length, batch size, embed dim] + output: [sequence length, batch size, embed dim] + Examples: + >>> output = pos_encoder(x) + """ + + x = x + self.pe[:x.size(0), :] + return self.dropout(x) diff --git a/strhub/models/base.py b/strhub/models/base.py new file mode 100644 index 00000000..4bdd4f03 --- /dev/null +++ b/strhub/models/base.py @@ -0,0 +1,219 @@ +# Scene Text Recognition Model Hub +# Copyright 2022 Darwin Bautista +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# https://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +import math +from abc import ABC, abstractmethod +from dataclasses import dataclass +from typing import Optional + +import pytorch_lightning as pl +import torch +import torch.nn.functional as F +from nltk import edit_distance +from pytorch_lightning.utilities.types import EPOCH_OUTPUT, STEP_OUTPUT +from timm.optim import create_optimizer_v2 +from torch import Tensor +from torch.optim import Optimizer +from torch.optim.lr_scheduler import OneCycleLR + +from strhub.data.utils import CharsetAdapter, CTCTokenizer, Tokenizer, BaseTokenizer + + +@dataclass +class BatchResult: + num_samples: int + correct: int + ned: float + confidence: float + label_length: int + loss: Tensor + loss_numel: int + + +class BaseSystem(pl.LightningModule, ABC): + + def __init__(self, tokenizer: BaseTokenizer, charset_test: str, + batch_size: int, lr: float, warmup_pct: float, weight_decay: float) -> None: + super().__init__() + self.tokenizer = tokenizer + self.charset_adapter = CharsetAdapter(charset_test) + self.batch_size = batch_size + self.lr = lr + self.warmup_pct = warmup_pct + self.weight_decay = weight_decay + self._num_training_steps = None + + @abstractmethod + def forward(self, images: Tensor, max_length: int = None) -> Tensor: + """Inference + + Args: + images: Batch of images. Shape: N, C, H, W + max_length: Max sequence length of the output. If None, will use default. + + Returns: + logits: N, L, C (L = sequence length, C = number of classes, typically len(charset_train) + num specials) + """ + raise NotImplementedError + + @abstractmethod + def forward_logits_loss(self, images: Tensor, labels: list[str]) -> tuple[Tensor, Tensor, int]: + """Like forward(), but also computes the loss (calls forward() internally). + + Args: + images: Batch of images. Shape: N, C, H, W + labels: Text labels of the images + + Returns: + logits: N, L, C (L = sequence length, C = number of classes, typically len(charset_train) + num specials) + loss: mean loss for the batch + loss_numel: number of elements the loss was calculated from + """ + raise NotImplementedError + + @property + def num_training_steps(self) -> int: + """Total training steps inferred from datamodule and devices.""" + if self._num_training_steps is None: + if self.trainer.max_steps is not None: + max_steps = self.trainer.max_steps + else: + loader = self.train_dataloader() + # We have no access to the actual DistributedSampler instance, but the default is drop_last == False + samples_per_replica = math.ceil(len(loader.dataset) / self.trainer.devices) + epoch_steps = samples_per_replica / loader.batch_size + epoch_steps = int(epoch_steps) if loader.drop_last else math.ceil(epoch_steps) + limit = self.trainer.limit_train_batches + epoch_steps = min(epoch_steps, limit) if isinstance(limit, int) else int(limit * epoch_steps) + max_steps = epoch_steps * self.trainer.max_epochs + self._num_training_steps = max_steps + return self._num_training_steps + + def configure_optimizers(self): + agb = self.trainer.accumulate_grad_batches + # Linear scaling so that the effective learning rate is constant regardless of the number of GPUs used with DDP. + lr_scale = agb * math.sqrt(self.trainer.devices) * self.batch_size / 256. + lr = lr_scale * self.lr + optim = create_optimizer_v2(self, 'adamw', lr, self.weight_decay) + sched = OneCycleLR(optim, lr, math.ceil(self.num_training_steps / agb), pct_start=self.warmup_pct, + cycle_momentum=False) + return {'optimizer': optim, 'lr_scheduler': {'scheduler': sched, 'interval': 'step'}} + + def optimizer_zero_grad(self, epoch: int, batch_idx: int, optimizer: Optimizer, optimizer_idx: int): + optimizer.zero_grad(set_to_none=True) + + def _eval_step(self, batch, validation: bool) -> Optional[STEP_OUTPUT]: + images, labels = batch + + correct = 0 + total = 0 + ned = 0 + confidence = 0 + label_length = 0 + if validation: + logits, loss, loss_numel = self.forward_logits_loss(images, labels) + else: + # At test-time, we shouldn't specify a max_label_length because the test-time charset used + # might be different from the train-time charset. max_label_length in eval_logits_loss() is computed + # based on the transformed label, which could be wrong if the actual gt label contains characters existing + # in the train-time charset but not in the test-time charset. For example, "aishahaleyes.blogspot.com" + # is exactly 25 characters, but if processed by CharsetAdapter for the 36-char set, it becomes 23 characters + # long only, which sets max_label_length = 23. This will cause the model prediction to be truncated. + logits = self.forward(images) + loss = loss_numel = None # Only used for validation; not needed at test-time. + + probs = logits.softmax(-1) + preds, probs = self.tokenizer.decode(probs) + for pred, prob, gt in zip(preds, probs, labels): + confidence += prob.prod().item() + pred = self.charset_adapter(pred) + # Follow ICDAR 2019 definition of N.E.D. + ned += edit_distance(pred, gt) / max(len(pred), len(gt)) + if pred == gt: + correct += 1 + total += 1 + label_length += len(pred) + return dict(output=BatchResult(total, correct, ned, confidence, label_length, loss, loss_numel)) + + @staticmethod + def _aggregate_results(outputs: EPOCH_OUTPUT) -> tuple[float, float, float]: + total_loss = 0 + total_loss_numel = 0 + total_n_correct = 0 + total_norm_ED = 0 + total_size = 0 + for result in outputs: + result = result['output'] + total_loss += result.loss_numel * result.loss + total_loss_numel += result.loss_numel + total_n_correct += result.correct + total_norm_ED += result.ned + total_size += result.num_samples + acc = total_n_correct / total_size + ned = (1 - total_norm_ED / total_size) + loss = total_loss / total_loss_numel + return acc, ned, loss + + def validation_step(self, batch, batch_idx) -> Optional[STEP_OUTPUT]: + return self._eval_step(batch, True) + + def validation_epoch_end(self, outputs: EPOCH_OUTPUT) -> None: + acc, ned, loss = self._aggregate_results(outputs) + self.log('val_accuracy', 100 * acc, sync_dist=True) + self.log('val_NED', 100 * ned, sync_dist=True) + self.log('val_loss', loss, sync_dist=True) + self.log('hp_metric', acc, sync_dist=True) + + def test_step(self, batch, batch_idx) -> Optional[STEP_OUTPUT]: + return self._eval_step(batch, False) + + +class CrossEntropySystem(BaseSystem, ABC): + + def __init__(self, charset_train: str, charset_test: str, + batch_size: int, lr: float, warmup_pct: float, weight_decay: float) -> None: + tokenizer = Tokenizer(charset_train) + super().__init__(tokenizer, charset_test, batch_size, lr, warmup_pct, weight_decay) + self.bos_id = tokenizer.bos_id + self.eos_id = tokenizer.eos_id + self.pad_id = tokenizer.pad_id + + def forward_logits_loss(self, images: Tensor, labels: list[str]) -> tuple[Tensor, Tensor, int]: + targets = self.tokenizer.encode(labels, self.device) + targets = targets[:, 1:] # Discard + max_len = targets.shape[1] - 1 # exclude from count + logits = self.forward(images, max_len) + loss = F.cross_entropy(logits.flatten(end_dim=1), targets.flatten(), ignore_index=self.pad_id) + loss_numel = (targets != self.pad_id).sum() + return logits, loss, loss_numel + + +class CTCSystem(BaseSystem, ABC): + + def __init__(self, charset_train: str, charset_test: str, + batch_size: int, lr: float, warmup_pct: float, weight_decay: float) -> None: + tokenizer = CTCTokenizer(charset_train) + super().__init__(tokenizer, charset_test, batch_size, lr, warmup_pct, weight_decay) + self.blank_id = tokenizer.blank_id + + def forward_logits_loss(self, images: Tensor, labels: list[str]) -> tuple[Tensor, Tensor, int]: + targets = self.tokenizer.encode(labels, self.device) + logits = self.forward(images) + log_probs = logits.log_softmax(-1).transpose(0, 1) # swap batch and seq. dims + T, N, _ = log_probs.shape + input_lengths = torch.full(size=(N,), fill_value=T, dtype=torch.long, device=self.device) + target_lengths = torch.as_tensor(list(map(len, labels)), dtype=torch.long, device=self.device) + loss = F.ctc_loss(log_probs, targets, input_lengths, target_lengths, blank=self.blank_id, zero_infinity=True) + return logits, loss, N diff --git a/strhub/models/crnn/LICENSE b/strhub/models/crnn/LICENSE new file mode 100644 index 00000000..f98687be --- /dev/null +++ b/strhub/models/crnn/LICENSE @@ -0,0 +1,21 @@ +The MIT License (MIT) + +Copyright (c) 2017 Jieru Mei + +Permission is hereby granted, free of charge, to any person obtaining a copy +of this software and associated documentation files (the "Software"), to deal +in the Software without restriction, including without limitation the rights +to use, copy, modify, merge, publish, distribute, sublicense, and/or sell +copies of the Software, and to permit persons to whom the Software is +furnished to do so, subject to the following conditions: + +The above copyright notice and this permission notice shall be included in all +copies or substantial portions of the Software. + +THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR +IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, +FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE +AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER +LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, +OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE +SOFTWARE. diff --git a/strhub/models/crnn/__init__.py b/strhub/models/crnn/__init__.py new file mode 100644 index 00000000..a4535947 --- /dev/null +++ b/strhub/models/crnn/__init__.py @@ -0,0 +1,13 @@ +r""" +Shi, Baoguang, Xiang Bai, and Cong Yao. +"An end-to-end trainable neural network for image-based sequence recognition and its application to scene text recognition." +IEEE transactions on pattern analysis and machine intelligence 39, no. 11 (2016): 2298-2304. + +https://arxiv.org/abs/1507.05717 + +All source files, except `system.py`, are based on the implementation listed below, +and hence are released under the license of the original. + +Source: https://github.com/meijieru/crnn.pytorch +License: MIT License (see included LICENSE file) +""" diff --git a/strhub/models/crnn/model.py b/strhub/models/crnn/model.py new file mode 100644 index 00000000..1a71845f --- /dev/null +++ b/strhub/models/crnn/model.py @@ -0,0 +1,62 @@ +import torch.nn as nn + +from strhub.models.modules import BidirectionalLSTM + + +class CRNN(nn.Module): + + def __init__(self, img_h, nc, nclass, nh, leaky_relu=False): + super().__init__() + assert img_h % 16 == 0, 'img_h has to be a multiple of 16' + + ks = [3, 3, 3, 3, 3, 3, 2] + ps = [1, 1, 1, 1, 1, 1, 0] + ss = [1, 1, 1, 1, 1, 1, 1] + nm = [64, 128, 256, 256, 512, 512, 512] + + cnn = nn.Sequential() + + def convRelu(i, batchNormalization=False): + nIn = nc if i == 0 else nm[i - 1] + nOut = nm[i] + cnn.add_module('conv{0}'.format(i), + nn.Conv2d(nIn, nOut, ks[i], ss[i], ps[i], bias=not batchNormalization)) + if batchNormalization: + cnn.add_module('batchnorm{0}'.format(i), nn.BatchNorm2d(nOut)) + if leaky_relu: + cnn.add_module('relu{0}'.format(i), + nn.LeakyReLU(0.2, inplace=True)) + else: + cnn.add_module('relu{0}'.format(i), nn.ReLU(True)) + + convRelu(0) + cnn.add_module('pooling{0}'.format(0), nn.MaxPool2d(2, 2)) # 64x16x64 + convRelu(1) + cnn.add_module('pooling{0}'.format(1), nn.MaxPool2d(2, 2)) # 128x8x32 + convRelu(2, True) + convRelu(3) + cnn.add_module('pooling{0}'.format(2), + nn.MaxPool2d((2, 2), (2, 1), (0, 1))) # 256x4x16 + convRelu(4, True) + convRelu(5) + cnn.add_module('pooling{0}'.format(3), + nn.MaxPool2d((2, 2), (2, 1), (0, 1))) # 512x2x16 + convRelu(6, True) # 512x1x16 + + self.cnn = cnn + self.rnn = nn.Sequential( + BidirectionalLSTM(512, nh, nh), + BidirectionalLSTM(nh, nh, nclass)) + + def forward(self, input): + # conv features + conv = self.cnn(input) + b, c, h, w = conv.size() + assert h == 1, 'the height of conv must be 1' + conv = conv.squeeze(2) + conv = conv.transpose(1, 2) # [b, w, c] + + # rnn features + output = self.rnn(conv) + + return output diff --git a/strhub/models/crnn/system.py b/strhub/models/crnn/system.py new file mode 100644 index 00000000..ad40d0f0 --- /dev/null +++ b/strhub/models/crnn/system.py @@ -0,0 +1,43 @@ +# Scene Text Recognition Model Hub +# Copyright 2022 Darwin Bautista +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# https://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +from typing import Sequence + +from pytorch_lightning.utilities.types import STEP_OUTPUT +from torch import Tensor + +from strhub.models.base import CTCSystem +from strhub.models.utils import init_weights +from .model import CRNN as Model + + +class CRNN(CTCSystem): + + def __init__(self, charset_train: str, charset_test: str, max_label_length: int, + batch_size: int, lr: float, warmup_pct: float, weight_decay: float, + img_size: Sequence[int], hidden_size: int, leaky_relu: bool, **kwargs) -> None: + super().__init__(charset_train, charset_test, batch_size, lr, warmup_pct, weight_decay) + self.save_hyperparameters() + self.model = Model(img_size[0], 3, len(self.tokenizer), hidden_size, leaky_relu) + self.model.apply(init_weights) + + def forward(self, images: Tensor, max_length: int = None) -> Tensor: + return self.model.forward(images) + + def training_step(self, batch, batch_idx) -> STEP_OUTPUT: + images, labels = batch + loss = self.forward_logits_loss(images, labels)[1] + self.log('loss', loss) + return loss diff --git a/strhub/models/modules.py b/strhub/models/modules.py new file mode 100644 index 00000000..a89d05f6 --- /dev/null +++ b/strhub/models/modules.py @@ -0,0 +1,20 @@ +r"""Shared modules used by CRNN and TRBA""" +from torch import nn + + +class BidirectionalLSTM(nn.Module): + """Ref: https://github.com/clovaai/deep-text-recognition-benchmark/blob/master/modules/sequence_modeling.py""" + + def __init__(self, input_size, hidden_size, output_size): + super().__init__() + self.rnn = nn.LSTM(input_size, hidden_size, bidirectional=True, batch_first=True) + self.linear = nn.Linear(hidden_size * 2, output_size) + + def forward(self, input): + """ + input : visual feature [batch_size x T x input_size], T = num_steps. + output : contextual feature [batch_size x T x output_size] + """ + recurrent, _ = self.rnn(input) # batch_size x T x input_size -> batch_size x T x (2*hidden_size) + output = self.linear(recurrent) # batch_size x T x output_size + return output diff --git a/strhub/models/parseq/__init__.py b/strhub/models/parseq/__init__.py new file mode 100644 index 00000000..e69de29b diff --git a/strhub/models/parseq/modules.py b/strhub/models/parseq/modules.py new file mode 100644 index 00000000..0038381d --- /dev/null +++ b/strhub/models/parseq/modules.py @@ -0,0 +1,133 @@ +# Scene Text Recognition Model Hub +# Copyright 2022 Darwin Bautista +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# https://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +import math +from typing import Optional + +import torch +from torch import nn as nn, Tensor +from torch.nn import functional as F +from torch.nn.modules import transformer + +from timm.models.vision_transformer import VisionTransformer, PatchEmbed + + +class DecoderLayer(nn.Module): + """A Transformer decoder layer supporting two-stream attention (XLNet) + This implements a pre-LN decoder, as opposed to the post-LN default in PyTorch.""" + + def __init__(self, d_model, nhead, dim_feedforward=2048, dropout=0.1, activation='gelu', + layer_norm_eps=1e-5): + super().__init__() + self.self_attn = nn.MultiheadAttention(d_model, nhead, dropout=dropout, batch_first=True) + self.cross_attn = nn.MultiheadAttention(d_model, nhead, dropout=dropout, batch_first=True) + # Implementation of Feedforward model + self.linear1 = nn.Linear(d_model, dim_feedforward) + self.dropout = nn.Dropout(dropout) + self.linear2 = nn.Linear(dim_feedforward, d_model) + + self.norm1 = nn.LayerNorm(d_model, eps=layer_norm_eps) + self.norm2 = nn.LayerNorm(d_model, eps=layer_norm_eps) + self.norm_q = nn.LayerNorm(d_model, eps=layer_norm_eps) + self.norm_c = nn.LayerNorm(d_model, eps=layer_norm_eps) + self.dropout1 = nn.Dropout(dropout) + self.dropout2 = nn.Dropout(dropout) + self.dropout3 = nn.Dropout(dropout) + + self.activation = transformer._get_activation_fn(activation) + + def __setstate__(self, state): + if 'activation' not in state: + state['activation'] = F.gelu + super().__setstate__(state) + + def forward_stream(self, tgt: Tensor, tgt_norm: Tensor, tgt_kv: Tensor, memory: Tensor, tgt_mask: Optional[Tensor], + tgt_key_padding_mask: Optional[Tensor]): + """Forward pass for a single stream (i.e. content or query) + tgt_norm is just a LayerNorm'd tgt. Added as a separate parameter for efficiency. + Both tgt_kv and memory are expected to be LayerNorm'd too. + memory is LayerNorm'd by ViT. + """ + tgt2, sa_weights = self.self_attn(tgt_norm, tgt_kv, tgt_kv, attn_mask=tgt_mask, + key_padding_mask=tgt_key_padding_mask) + tgt = tgt + self.dropout1(tgt2) + + tgt2, ca_weights = self.cross_attn(self.norm1(tgt), memory, memory) + tgt = tgt + self.dropout2(tgt2) + + tgt2 = self.linear2(self.dropout(self.activation(self.linear1(self.norm2(tgt))))) + tgt = tgt + self.dropout3(tgt2) + return tgt, sa_weights, ca_weights + + def forward(self, query, content, memory, query_mask=None, content_mask=None, content_key_padding_mask=None, + update_content=True): + query_norm = self.norm_q(query) + content_norm = self.norm_c(content) + query = self.forward_stream(query, query_norm, content_norm, memory, query_mask, content_key_padding_mask)[0] + if update_content: + content = self.forward_stream(content, content_norm, content_norm, memory, content_mask, + content_key_padding_mask)[0] + return query, content + + +class Decoder(nn.Module): + __constants__ = ['norm'] + + def __init__(self, decoder_layer, num_layers, norm): + super().__init__() + self.layers = transformer._get_clones(decoder_layer, num_layers) + self.num_layers = num_layers + self.norm = norm + + def forward(self, query, content, memory, query_mask=None, content_mask=None, content_key_padding_mask=None): + for mod in self.layers: + last = mod == self.layers[-1] + query, content = mod(query, content, memory, query_mask, content_mask, content_key_padding_mask, + update_content=not last) + query = self.norm(query) + return query + + +class Encoder(VisionTransformer): + + def __init__(self, img_size=224, patch_size=16, in_chans=3, embed_dim=768, depth=12, num_heads=12, mlp_ratio=4., + qkv_bias=True, drop_rate=0., attn_drop_rate=0., drop_path_rate=0., embed_layer=PatchEmbed): + super().__init__(img_size, patch_size, in_chans, embed_dim=embed_dim, depth=depth, num_heads=num_heads, + mlp_ratio=mlp_ratio, qkv_bias=qkv_bias, drop_rate=drop_rate, attn_drop_rate=attn_drop_rate, + drop_path_rate=drop_path_rate, embed_layer=embed_layer) + # Only allocate position embeddings for the image tokens (none for cls nor dist) + self.pos_embed = nn.Parameter(torch.zeros(1, self.patch_embed.num_patches, embed_dim)) + nn.init.trunc_normal_(self.pos_embed, std=.02) + # Delete unused modules + del self.pre_logits, self.head, self.head_dist, self.cls_token, self.dist_token + + def forward(self, x): + # Essentially forward_features() but returns all tokens + x = self.patch_embed(x) + x = self.pos_drop(x + self.pos_embed) + x = self.blocks(x) + x = self.norm(x) + return x + + +class TokenEmbedding(nn.Module): + + def __init__(self, charset_size: int, embed_dim: int): + super().__init__() + self.embedding = nn.Embedding(charset_size, embed_dim) + self.embed_dim = embed_dim + + def forward(self, tokens: torch.Tensor): + return math.sqrt(self.embed_dim) * self.embedding(tokens) diff --git a/strhub/models/parseq/system.py b/strhub/models/parseq/system.py new file mode 100644 index 00000000..9577df7c --- /dev/null +++ b/strhub/models/parseq/system.py @@ -0,0 +1,258 @@ +# Scene Text Recognition Model Hub +# Copyright 2022 Darwin Bautista +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# https://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +import math +from functools import partial +from itertools import permutations +from typing import Sequence, Any + +import numpy as np +import torch +import torch.nn as nn +import torch.nn.functional as F +from torch import Tensor + +from pytorch_lightning.utilities.types import STEP_OUTPUT +from timm.models.helpers import named_apply + +from strhub.models.base import CrossEntropySystem +from strhub.models.utils import init_weights +from .modules import DecoderLayer, Decoder, Encoder, TokenEmbedding + + +class PARSeq(CrossEntropySystem): + + def __init__(self, charset_train: str, charset_test: str, max_label_length: int, + batch_size: int, lr: float, warmup_pct: float, weight_decay: float, + img_size: Sequence[int], patch_size: Sequence[int], embed_dim: int, + enc_num_heads: int, enc_mlp_ratio: int, enc_depth: int, + dec_num_heads: int, dec_mlp_ratio: int, dec_depth: int, + perm_num: int, perm_forward: bool, perm_mirrored: bool, + decode_ar: bool, refine_iters: int, dropout: float, **kwargs: Any) -> None: + super().__init__(charset_train, charset_test, batch_size, lr, warmup_pct, weight_decay) + self.save_hyperparameters() + + self.max_label_length = max_label_length + self.decode_ar = decode_ar + self.refine_iters = refine_iters + + self.encoder = Encoder(img_size, patch_size, embed_dim=embed_dim, depth=enc_depth, num_heads=enc_num_heads, + mlp_ratio=enc_mlp_ratio) + decoder_layer = DecoderLayer(embed_dim, dec_num_heads, embed_dim * dec_mlp_ratio, dropout) + self.decoder = Decoder(decoder_layer, num_layers=dec_depth, norm=nn.LayerNorm(embed_dim)) + + # Perm/attn mask stuff + self.rng = np.random.default_rng() + self.max_gen_perms = perm_num // 2 if perm_mirrored else perm_num + self.perm_forward = perm_forward + self.perm_mirrored = perm_mirrored + + # We don't predict nor + self.head = nn.Linear(embed_dim, len(self.tokenizer) - 2) + self.text_embed = TokenEmbedding(len(self.tokenizer), embed_dim) + + # +1 for + self.pos_queries = nn.Parameter(torch.Tensor(1, max_label_length + 1, embed_dim)) + self.dropout = nn.Dropout(p=dropout) + # Encoder has its own init. + named_apply(partial(init_weights, exclude=['encoder']), self) + nn.init.trunc_normal_(self.pos_queries, std=.02) + + @torch.jit.ignore + def no_weight_decay(self): + param_names = {'text_embed.embedding.weight', 'pos_queries'} + enc_param_names = {'encoder.' + n for n in self.encoder.no_weight_decay()} + return param_names.union(enc_param_names) + + def encode(self, img: torch.Tensor): + return self.encoder(img) + + def decode(self, tgt: torch.Tensor, memory: torch.Tensor, tgt_mask=None, tgt_padding_mask=None, tgt_query=None, + tgt_query_mask=None): + N, L = tgt.shape + # stands for the null context. We only supply position information for characters after . + null_ctx = self.text_embed(tgt[:, :1]) + tgt_emb = self.pos_queries[:, :L - 1] + self.text_embed(tgt[:, 1:]) + tgt_emb = self.dropout(torch.cat([null_ctx, tgt_emb], dim=1)) + if tgt_query is None: + tgt_query = self.pos_queries[:, :L].expand(N, -1, -1) + tgt_query = self.dropout(tgt_query) + return self.decoder(tgt_query, tgt_emb, memory, tgt_query_mask, tgt_mask, tgt_padding_mask) + + def forward(self, images: Tensor, max_length: int = None) -> Tensor: + testing = max_length is None + max_length = self.max_label_length if testing else min(max_length, self.max_label_length) + bs = images.shape[0] + # +1 for at end of sequence. + num_steps = max_length + 1 + memory = self.encode(images) + + # Query positions up to `num_steps` + pos_queries = self.pos_queries[:, :num_steps].expand(bs, -1, -1) + + # Special case for the forward permutation. Faster than using `generate_attn_masks()` + tgt_mask = query_mask = torch.triu(torch.full((num_steps, num_steps), float('-inf'), device=self.device), 1) + + if self.decode_ar: + tgt_in = torch.full((bs, num_steps), self.pad_id, dtype=torch.long, device=self.device) + tgt_in[:, 0] = self.bos_id + + logits = [] + for i in range(num_steps): + j = i + 1 # next token index + # Efficient decoding: + # Input the context up to the ith token. We use only one query (at position = i) at a time. + # This works because of the lookahead masking effect of the canonical (forward) AR context. + # Past tokens have no access to future tokens, hence are fixed once computed. + tgt_out = self.decode(tgt_in[:, :j], memory, tgt_mask[:j, :j], tgt_query=pos_queries[:, i:j], + tgt_query_mask=query_mask[i:j, :j]) + # the next token probability is in the output's ith token position + p_i = self.head(tgt_out) + logits.append(p_i) + if j < num_steps: + # greedy decode. add the next token index to the target input + tgt_in[:, j] = p_i.squeeze().argmax(-1) + # Efficient batch decoding: If all output words have at least one EOS token, end decoding. + if testing and (tgt_in == self.eos_id).any(dim=-1).all(): + break + + logits = torch.cat(logits, dim=1) + else: + # No prior context, so input is just . We query all positions. + tgt_in = torch.full((bs, 1), self.bos_id, dtype=torch.long, device=self.device) + tgt_out = self.decode(tgt_in, memory, tgt_query=pos_queries) + logits = self.head(tgt_out) + + if self.refine_iters: + # For iterative refinement, we always use a 'cloze' mask. + # We can derive it from the AR forward mask by unmasking the token context to the right. + query_mask[torch.triu(torch.ones(num_steps, num_steps, dtype=torch.bool, device=self.device), 2)] = 0 + bos = torch.full((bs, 1), self.bos_id, dtype=torch.long, device=self.device) + for i in range(self.refine_iters): + # Prior context is the previous output. + tgt_in = torch.cat([bos, logits[:, :-1].argmax(-1)], dim=1) + tgt_padding_mask = ((tgt_in == self.eos_id).cumsum(-1) > 0) # mask tokens beyond the first EOS token. + tgt_out = self.decode(tgt_in, memory, tgt_mask, tgt_padding_mask, + tgt_query=pos_queries, tgt_query_mask=query_mask[:, :tgt_in.shape[1]]) + logits = self.head(tgt_out) + + return logits + + def gen_tgt_perms(self, tgt): + """Generate shared permutations for the whole batch. + This works because the same attention mask can be used for the shorter sequences + because of the padding mask. + """ + # We don't permute the position of BOS, we permute EOS separately + max_num_chars = tgt.shape[1] - 2 + # Special handling for 1-character sequences + if max_num_chars == 1: + return torch.arange(3, device=self.device).unsqueeze(0) + perms = [torch.arange(max_num_chars, device=self.device)] if self.perm_forward else [] + # Additional permutations if needed + max_perms = math.factorial(max_num_chars) + if self.perm_mirrored: + max_perms //= 2 + num_gen_perms = min(self.max_gen_perms, max_perms) + # For 4-char sequences and shorter, we generate all permutations and sample from the pool to avoid collisions + if max_num_chars < 5: + # Pool of permutations to sample from. We only need the first half (if complementary option is selected) + # Special handling for max_num_chars == 4 which correctly divides the pool into the flipped halves + if max_num_chars == 4 and self.perm_mirrored: + selector = [0, 3, 4, 6, 9, 10, 12, 16, 17, 18, 19, 21] + else: + selector = list(range(max_perms)) + perm_pool = torch.as_tensor(list(permutations(range(max_num_chars), max_num_chars)), device=self.device)[ + selector] + # If the forward permutation is always selected, no need to add it to the pool for sampling + if self.perm_forward: + perm_pool = perm_pool[1:] + perms = torch.stack(perms) + if len(perm_pool): + i = self.rng.choice(len(perm_pool), size=num_gen_perms - len(perms), replace=False) + perms = torch.cat([perms, perm_pool[i]]) + else: + perms.extend([torch.randperm(max_num_chars, device=self.device) for _ in range(num_gen_perms - len(perms))]) + perms = torch.stack(perms) + if self.perm_mirrored: + # Add complementary pairs + comp = perms.flip(-1) + # Stack in such a way that the pairs are next to each other. + perms = torch.stack([perms, comp]).transpose(0, 1).reshape(-1, max_num_chars) + # NOTE: + # The only meaningful way of permuting the EOS position is by moving it one character position at a time. + # However, since the number of permutations = T! and number of EOS positions = T + 1, the number of possible EOS + # positions will always be much less than the number of permutations (unless a low perm_num is set). + # Thus, it would be simpler to just train EOS using the full and null contexts rather than trying to evenly + # distribute it across the chosen number of permutations. + # Add position indices of BOS and EOS + bos_idx = perms.new_zeros((len(perms), 1)) + eos_idx = perms.new_full((len(perms), 1), max_num_chars + 1) + perms = torch.cat([bos_idx, perms + 1, eos_idx], dim=1) + # Special handling for the reverse direction. This does two things: + # 1. Reverse context for the characters + # 2. Null context for [EOS] (required for learning to predict [EOS] in NAR mode) + if len(perms) > 1: + perms[1, 1:] = max_num_chars + 1 - torch.arange(max_num_chars + 1, device=self.device) + return perms + + def generate_attn_masks(self, perm): + """Generate attention masks given a sequence permutation (includes pos. for bos and eos tokens) + :param perm: the permutation sequence. i = 0 is always the BOS + :return: lookahead attention masks + """ + sz = perm.shape[0] + mask = torch.zeros((sz, sz), device=self.device) + for i in range(sz): + query_idx = perm[i] + masked_keys = perm[i + 1:] + mask[query_idx, masked_keys] = float('-inf') + content_mask = mask[:-1, :-1].clone() + mask[torch.eye(sz, dtype=torch.bool, device=self.device)] = float('-inf') # mask "self" + query_mask = mask[1:, :-1] + return content_mask, query_mask + + def training_step(self, batch, batch_idx) -> STEP_OUTPUT: + images, labels = batch + tgt = self.tokenizer.encode(labels, self.device) + + # Encode the source sequence (i.e. the image codes) + memory = self.encode(images) + + # Prepare the target sequences (input and output) + tgt_perms = self.gen_tgt_perms(tgt) + tgt_in = tgt[:, :-1] + tgt_out = tgt[:, 1:] + # The [EOS] token is not depended upon by any other token in any permutation ordering + tgt_padding_mask = (tgt_in == self.pad_id) | (tgt_in == self.eos_id) + + loss = 0 + loss_numel = 0 + n = (tgt_out != self.pad_id).sum().item() + for i, perm in enumerate(tgt_perms): + tgt_mask, query_mask = self.generate_attn_masks(perm) + out = self.decode(tgt_in, memory, tgt_mask, tgt_padding_mask, tgt_query_mask=query_mask) + logits = self.head(out).flatten(end_dim=1) + loss += n * F.cross_entropy(logits, tgt_out.flatten(), ignore_index=self.pad_id) + loss_numel += n + # After the second iteration (i.e. done with canonical and reverse orderings), + # remove the [EOS] tokens for the succeeding perms + if i == 1: + tgt_out = torch.where(tgt_out == self.eos_id, self.pad_id, tgt_out) + n = (tgt_out != self.pad_id).sum().item() + loss /= loss_numel + + self.log('loss', loss) + return loss diff --git a/strhub/models/trba/__init__.py b/strhub/models/trba/__init__.py new file mode 100644 index 00000000..a574a8af --- /dev/null +++ b/strhub/models/trba/__init__.py @@ -0,0 +1,13 @@ +r""" +Baek, Jeonghun, Geewook Kim, Junyeop Lee, Sungrae Park, Dongyoon Han, Sangdoo Yun, Seong Joon Oh, and Hwalsuk Lee. +"What is wrong with scene text recognition model comparisons? dataset and model analysis." +In Proceedings of the IEEE/CVF International Conference on Computer Vision, pp. 4715-4723. 2019. + +https://arxiv.org/abs/1904.01906 + +All source files, except `system.py`, are based on the implementation listed below, +and hence are released under the license of the original. + +Source: https://github.com/clovaai/deep-text-recognition-benchmark +License: Apache License 2.0 (see LICENSE file in project root) +""" diff --git a/strhub/models/trba/feature_extraction.py b/strhub/models/trba/feature_extraction.py new file mode 100644 index 00000000..17646e3f --- /dev/null +++ b/strhub/models/trba/feature_extraction.py @@ -0,0 +1,110 @@ +import torch.nn as nn + +from torchvision.models.resnet import BasicBlock + + +class ResNet_FeatureExtractor(nn.Module): + """ FeatureExtractor of FAN (http://openaccess.thecvf.com/content_ICCV_2017/papers/Cheng_Focusing_Attention_Towards_ICCV_2017_paper.pdf) """ + + def __init__(self, input_channel, output_channel=512): + super().__init__() + self.ConvNet = ResNet(input_channel, output_channel, BasicBlock, [1, 2, 5, 3]) + + def forward(self, input): + return self.ConvNet(input) + + +class ResNet(nn.Module): + + def __init__(self, input_channel, output_channel, block, layers): + super().__init__() + + self.output_channel_block = [int(output_channel / 4), int(output_channel / 2), output_channel, output_channel] + + self.inplanes = int(output_channel / 8) + self.conv0_1 = nn.Conv2d(input_channel, int(output_channel / 16), + kernel_size=3, stride=1, padding=1, bias=False) + self.bn0_1 = nn.BatchNorm2d(int(output_channel / 16)) + self.conv0_2 = nn.Conv2d(int(output_channel / 16), self.inplanes, + kernel_size=3, stride=1, padding=1, bias=False) + self.bn0_2 = nn.BatchNorm2d(self.inplanes) + self.relu = nn.ReLU(inplace=True) + + self.maxpool1 = nn.MaxPool2d(kernel_size=2, stride=2, padding=0) + self.layer1 = self._make_layer(block, self.output_channel_block[0], layers[0]) + self.conv1 = nn.Conv2d(self.output_channel_block[0], self.output_channel_block[ + 0], kernel_size=3, stride=1, padding=1, bias=False) + self.bn1 = nn.BatchNorm2d(self.output_channel_block[0]) + + self.maxpool2 = nn.MaxPool2d(kernel_size=2, stride=2, padding=0) + self.layer2 = self._make_layer(block, self.output_channel_block[1], layers[1], stride=1) + self.conv2 = nn.Conv2d(self.output_channel_block[1], self.output_channel_block[ + 1], kernel_size=3, stride=1, padding=1, bias=False) + self.bn2 = nn.BatchNorm2d(self.output_channel_block[1]) + + self.maxpool3 = nn.MaxPool2d(kernel_size=2, stride=(2, 1), padding=(0, 1)) + self.layer3 = self._make_layer(block, self.output_channel_block[2], layers[2], stride=1) + self.conv3 = nn.Conv2d(self.output_channel_block[2], self.output_channel_block[ + 2], kernel_size=3, stride=1, padding=1, bias=False) + self.bn3 = nn.BatchNorm2d(self.output_channel_block[2]) + + self.layer4 = self._make_layer(block, self.output_channel_block[3], layers[3], stride=1) + self.conv4_1 = nn.Conv2d(self.output_channel_block[3], self.output_channel_block[ + 3], kernel_size=2, stride=(2, 1), padding=(0, 1), bias=False) + self.bn4_1 = nn.BatchNorm2d(self.output_channel_block[3]) + self.conv4_2 = nn.Conv2d(self.output_channel_block[3], self.output_channel_block[ + 3], kernel_size=2, stride=1, padding=0, bias=False) + self.bn4_2 = nn.BatchNorm2d(self.output_channel_block[3]) + + def _make_layer(self, block, planes, blocks, stride=1): + downsample = None + if stride != 1 or self.inplanes != planes * block.expansion: + downsample = nn.Sequential( + nn.Conv2d(self.inplanes, planes * block.expansion, + kernel_size=1, stride=stride, bias=False), + nn.BatchNorm2d(planes * block.expansion), + ) + + layers = [] + layers.append(block(self.inplanes, planes, stride, downsample)) + self.inplanes = planes * block.expansion + for i in range(1, blocks): + layers.append(block(self.inplanes, planes)) + + return nn.Sequential(*layers) + + def forward(self, x): + x = self.conv0_1(x) + x = self.bn0_1(x) + x = self.relu(x) + x = self.conv0_2(x) + x = self.bn0_2(x) + x = self.relu(x) + + x = self.maxpool1(x) + x = self.layer1(x) + x = self.conv1(x) + x = self.bn1(x) + x = self.relu(x) + + x = self.maxpool2(x) + x = self.layer2(x) + x = self.conv2(x) + x = self.bn2(x) + x = self.relu(x) + + x = self.maxpool3(x) + x = self.layer3(x) + x = self.conv3(x) + x = self.bn3(x) + x = self.relu(x) + + x = self.layer4(x) + x = self.conv4_1(x) + x = self.bn4_1(x) + x = self.relu(x) + x = self.conv4_2(x) + x = self.bn4_2(x) + x = self.relu(x) + + return x diff --git a/strhub/models/trba/model.py b/strhub/models/trba/model.py new file mode 100644 index 00000000..41161a4d --- /dev/null +++ b/strhub/models/trba/model.py @@ -0,0 +1,55 @@ +import torch.nn as nn + +from strhub.models.modules import BidirectionalLSTM +from .feature_extraction import ResNet_FeatureExtractor +from .prediction import Attention +from .transformation import TPS_SpatialTransformerNetwork + + +class TRBA(nn.Module): + + def __init__(self, img_h, img_w, num_class, num_fiducial=20, input_channel=3, output_channel=512, hidden_size=256, + use_ctc=False): + super().__init__() + """ Transformation """ + self.Transformation = TPS_SpatialTransformerNetwork( + F=num_fiducial, I_size=(img_h, img_w), I_r_size=(img_h, img_w), + I_channel_num=input_channel) + + """ FeatureExtraction """ + self.FeatureExtraction = ResNet_FeatureExtractor(input_channel, output_channel) + self.FeatureExtraction_output = output_channel + self.AdaptiveAvgPool = nn.AdaptiveAvgPool2d((None, 1)) # Transform final (imgH/16-1) -> 1 + + """ Sequence modeling""" + self.SequenceModeling = nn.Sequential( + BidirectionalLSTM(self.FeatureExtraction_output, hidden_size, hidden_size), + BidirectionalLSTM(hidden_size, hidden_size, hidden_size)) + self.SequenceModeling_output = hidden_size + + """ Prediction """ + if use_ctc: + self.Prediction = nn.Linear(self.SequenceModeling_output, num_class) + else: + self.Prediction = Attention(self.SequenceModeling_output, hidden_size, num_class) + + def forward(self, image, max_label_length, text=None): + """ Transformation stage """ + image = self.Transformation(image) + + """ Feature extraction stage """ + visual_feature = self.FeatureExtraction(image) + visual_feature = visual_feature.permute(0, 3, 1, 2) # [b, c, h, w] -> [b, w, c, h] + visual_feature = self.AdaptiveAvgPool(visual_feature) # [b, w, c, h] -> [b, w, c, 1] + visual_feature = visual_feature.squeeze(3) # [b, w, c, 1] -> [b, w, c] + + """ Sequence modeling stage """ + contextual_feature = self.SequenceModeling(visual_feature) # [b, num_steps, hidden_size] + + """ Prediction stage """ + if isinstance(self.Prediction, Attention): + prediction = self.Prediction(contextual_feature.contiguous(), text, max_label_length) + else: + prediction = self.Prediction(contextual_feature.contiguous()) # CTC + + return prediction # [b, num_steps, num_class] diff --git a/strhub/models/trba/prediction.py b/strhub/models/trba/prediction.py new file mode 100644 index 00000000..5609398a --- /dev/null +++ b/strhub/models/trba/prediction.py @@ -0,0 +1,73 @@ +import torch +import torch.nn as nn +import torch.nn.functional as F + + +class Attention(nn.Module): + + def __init__(self, input_size, hidden_size, num_class, num_char_embeddings=256): + super().__init__() + self.attention_cell = AttentionCell(input_size, hidden_size, num_char_embeddings) + self.hidden_size = hidden_size + self.num_class = num_class + self.generator = nn.Linear(hidden_size, num_class) + self.char_embeddings = nn.Embedding(num_class, num_char_embeddings) + + def forward(self, batch_H, text, max_label_length=25): + """ + input: + batch_H : contextual_feature H = hidden state of encoder. [batch_size x num_steps x num_class] + text : the text-index of each image. [batch_size x (max_length+1)]. +1 for [SOS] token. text[:, 0] = [SOS]. + output: probability distribution at each step [batch_size x num_steps x num_class] + """ + batch_size = batch_H.size(0) + num_steps = max_label_length + 1 # +1 for [EOS] at end of sentence. + + output_hiddens = batch_H.new_zeros((batch_size, num_steps, self.hidden_size), dtype=torch.float) + hidden = (batch_H.new_zeros((batch_size, self.hidden_size), dtype=torch.float), + batch_H.new_zeros((batch_size, self.hidden_size), dtype=torch.float)) + + if self.training: + for i in range(num_steps): + char_embeddings = self.char_embeddings(text[:, i]) + # hidden : decoder's hidden s_{t-1}, batch_H : encoder's hidden H, char_embeddings : f(y_{t-1}) + hidden, alpha = self.attention_cell(hidden, batch_H, char_embeddings) + output_hiddens[:, i, :] = hidden[0] # LSTM hidden index (0: hidden, 1: Cell) + probs = self.generator(output_hiddens) + + else: + targets = text[0].expand(batch_size) # should be fill with [SOS] token + probs = batch_H.new_zeros((batch_size, num_steps, self.num_class), dtype=torch.float) + + for i in range(num_steps): + char_embeddings = self.char_embeddings(targets) + hidden, alpha = self.attention_cell(hidden, batch_H, char_embeddings) + probs_step = self.generator(hidden[0]) + probs[:, i, :] = probs_step + _, next_input = probs_step.max(1) + targets = next_input + + return probs # batch_size x num_steps x num_class + + +class AttentionCell(nn.Module): + + def __init__(self, input_size, hidden_size, num_embeddings): + super().__init__() + self.i2h = nn.Linear(input_size, hidden_size, bias=False) + self.h2h = nn.Linear(hidden_size, hidden_size) # either i2i or h2h should have bias + self.score = nn.Linear(hidden_size, 1, bias=False) + self.rnn = nn.LSTMCell(input_size + num_embeddings, hidden_size) + self.hidden_size = hidden_size + + def forward(self, prev_hidden, batch_H, char_embeddings): + # [batch_size x num_encoder_step x num_channel] -> [batch_size x num_encoder_step x hidden_size] + batch_H_proj = self.i2h(batch_H) + prev_hidden_proj = self.h2h(prev_hidden[0]).unsqueeze(1) + e = self.score(torch.tanh(batch_H_proj + prev_hidden_proj)) # batch_size x num_encoder_step * 1 + + alpha = F.softmax(e, dim=1) + context = torch.bmm(alpha.permute(0, 2, 1), batch_H).squeeze(1) # batch_size x num_channel + concat_context = torch.cat([context, char_embeddings], 1) # batch_size x (num_channel + num_embedding) + cur_hidden = self.rnn(concat_context, prev_hidden) + return cur_hidden, alpha diff --git a/strhub/models/trba/system.py b/strhub/models/trba/system.py new file mode 100644 index 00000000..655b4390 --- /dev/null +++ b/strhub/models/trba/system.py @@ -0,0 +1,87 @@ +# Scene Text Recognition Model Hub +# Copyright 2022 Darwin Bautista +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# https://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +from functools import partial +from typing import Sequence, Any + +import torch +import torch.nn.functional as F +from pytorch_lightning.utilities.types import STEP_OUTPUT +from timm.models.helpers import named_apply +from torch import Tensor + +from strhub.models.base import CrossEntropySystem, CTCSystem +from strhub.models.utils import init_weights +from .model import TRBA as Model + + +class TRBA(CrossEntropySystem): + + def __init__(self, charset_train: str, charset_test: str, max_label_length: int, + batch_size: int, lr: float, warmup_pct: float, weight_decay: float, + img_size: Sequence[int], num_fiducial: int, output_channel: int, hidden_size: int, + **kwargs: Any) -> None: + super().__init__(charset_train, charset_test, batch_size, lr, warmup_pct, weight_decay) + self.save_hyperparameters() + self.max_label_length = max_label_length + img_h, img_w = img_size + self.model = Model(img_h, img_w, len(self.tokenizer), num_fiducial, + output_channel=output_channel, hidden_size=hidden_size, use_ctc=False) + named_apply(partial(init_weights, exclude=['Transformation.LocalizationNetwork.localization_fc2']), self.model) + + @torch.jit.ignore + def no_weight_decay(self): + return {'model.Prediction.char_embeddings.weight'} + + def forward(self, images: Tensor, max_length: int = None) -> Tensor: + max_length = self.max_label_length if max_length is None else min(max_length, self.max_label_length) + text = images.new_full([1], self.bos_id, dtype=torch.long) + return self.model.forward(images, max_length, text) + + def training_step(self, batch, batch_idx) -> STEP_OUTPUT: + images, labels = batch + encoded = self.tokenizer.encode(labels, self.device) + inputs = encoded[:, :-1] # remove + targets = encoded[:, 1:] # remove + max_length = encoded.shape[1] - 2 # exclude and from count + logits = self.model.forward(images, max_length, inputs) + loss = F.cross_entropy(logits.flatten(end_dim=1), targets.flatten(), ignore_index=self.pad_id) + self.log('loss', loss) + return loss + + +class TRBC(CTCSystem): + + def __init__(self, charset_train: str, charset_test: str, max_label_length: int, + batch_size: int, lr: float, warmup_pct: float, weight_decay: float, + img_size: Sequence[int], num_fiducial: int, output_channel: int, hidden_size: int, + **kwargs: Any) -> None: + super().__init__(charset_train, charset_test, batch_size, lr, warmup_pct, weight_decay) + self.save_hyperparameters() + self.max_label_length = max_label_length + img_h, img_w = img_size + self.model = Model(img_h, img_w, len(self.tokenizer), num_fiducial, + output_channel=output_channel, hidden_size=hidden_size, use_ctc=True) + named_apply(partial(init_weights, exclude=['Transformation.LocalizationNetwork.localization_fc2']), self.model) + + def forward(self, images: Tensor, max_length: int = None) -> Tensor: + # max_label_length is unused in CTC prediction + return self.model.forward(images, None) + + def training_step(self, batch, batch_idx) -> STEP_OUTPUT: + images, labels = batch + loss = self.forward_logits_loss(images, labels)[1] + self.log('loss', loss) + return loss diff --git a/strhub/models/trba/transformation.py b/strhub/models/trba/transformation.py new file mode 100644 index 00000000..960419d1 --- /dev/null +++ b/strhub/models/trba/transformation.py @@ -0,0 +1,169 @@ +import numpy as np +import torch +import torch.nn as nn +import torch.nn.functional as F + + +class TPS_SpatialTransformerNetwork(nn.Module): + """ Rectification Network of RARE, namely TPS based STN """ + + def __init__(self, F, I_size, I_r_size, I_channel_num=1): + """ Based on RARE TPS + input: + batch_I: Batch Input Image [batch_size x I_channel_num x I_height x I_width] + I_size : (height, width) of the input image I + I_r_size : (height, width) of the rectified image I_r + I_channel_num : the number of channels of the input image I + output: + batch_I_r: rectified image [batch_size x I_channel_num x I_r_height x I_r_width] + """ + super().__init__() + self.F = F + self.I_size = I_size + self.I_r_size = I_r_size # = (I_r_height, I_r_width) + self.I_channel_num = I_channel_num + self.LocalizationNetwork = LocalizationNetwork(self.F, self.I_channel_num) + self.GridGenerator = GridGenerator(self.F, self.I_r_size) + + def forward(self, batch_I): + batch_C_prime = self.LocalizationNetwork(batch_I) # batch_size x K x 2 + # batch_size x n (= I_r_width x I_r_height) x 2 + build_P_prime = self.GridGenerator.build_P_prime(batch_C_prime) + build_P_prime_reshape = build_P_prime.reshape([build_P_prime.size(0), self.I_r_size[0], self.I_r_size[1], 2]) + + if torch.__version__ > "1.2.0": + batch_I_r = F.grid_sample(batch_I, build_P_prime_reshape, padding_mode='border', align_corners=True) + else: + batch_I_r = F.grid_sample(batch_I, build_P_prime_reshape, padding_mode='border') + + return batch_I_r + + +class LocalizationNetwork(nn.Module): + """ Localization Network of RARE, which predicts C' (K x 2) from I (I_width x I_height) """ + + def __init__(self, F, I_channel_num): + super().__init__() + self.F = F + self.I_channel_num = I_channel_num + self.conv = nn.Sequential( + nn.Conv2d(in_channels=self.I_channel_num, out_channels=64, kernel_size=3, stride=1, padding=1, + bias=False), nn.BatchNorm2d(64), nn.ReLU(True), + nn.MaxPool2d(2, 2), # batch_size x 64 x I_height/2 x I_width/2 + nn.Conv2d(64, 128, 3, 1, 1, bias=False), nn.BatchNorm2d(128), nn.ReLU(True), + nn.MaxPool2d(2, 2), # batch_size x 128 x I_height/4 x I_width/4 + nn.Conv2d(128, 256, 3, 1, 1, bias=False), nn.BatchNorm2d(256), nn.ReLU(True), + nn.MaxPool2d(2, 2), # batch_size x 256 x I_height/8 x I_width/8 + nn.Conv2d(256, 512, 3, 1, 1, bias=False), nn.BatchNorm2d(512), nn.ReLU(True), + nn.AdaptiveAvgPool2d(1) # batch_size x 512 + ) + + self.localization_fc1 = nn.Sequential(nn.Linear(512, 256), nn.ReLU(True)) + self.localization_fc2 = nn.Linear(256, self.F * 2) + + # Init fc2 in LocalizationNetwork + self.localization_fc2.weight.data.fill_(0) + """ see RARE paper Fig. 6 (a) """ + ctrl_pts_x = np.linspace(-1.0, 1.0, int(F / 2)) + ctrl_pts_y_top = np.linspace(0.0, -1.0, num=int(F / 2)) + ctrl_pts_y_bottom = np.linspace(1.0, 0.0, num=int(F / 2)) + ctrl_pts_top = np.stack([ctrl_pts_x, ctrl_pts_y_top], axis=1) + ctrl_pts_bottom = np.stack([ctrl_pts_x, ctrl_pts_y_bottom], axis=1) + initial_bias = np.concatenate([ctrl_pts_top, ctrl_pts_bottom], axis=0) + self.localization_fc2.bias.data = torch.from_numpy(initial_bias).float().view(-1) + + def forward(self, batch_I): + """ + input: batch_I : Batch Input Image [batch_size x I_channel_num x I_height x I_width] + output: batch_C_prime : Predicted coordinates of fiducial points for input batch [batch_size x F x 2] + """ + batch_size = batch_I.size(0) + features = self.conv(batch_I).view(batch_size, -1) + batch_C_prime = self.localization_fc2(self.localization_fc1(features)).view(batch_size, self.F, 2) + return batch_C_prime + + +class GridGenerator(nn.Module): + """ Grid Generator of RARE, which produces P_prime by multipling T with P """ + + def __init__(self, F, I_r_size): + """ Generate P_hat and inv_delta_C for later """ + super().__init__() + self.eps = 1e-6 + self.I_r_height, self.I_r_width = I_r_size + self.F = F + self.C = self._build_C(self.F) # F x 2 + self.P = self._build_P(self.I_r_width, self.I_r_height) + + # num_gpu = torch.cuda.device_count() + # if num_gpu > 1: + # for multi-gpu, you may need register buffer + self.register_buffer("inv_delta_C", torch.tensor( + self._build_inv_delta_C(self.F, self.C)).float()) # F+3 x F+3 + self.register_buffer("P_hat", torch.tensor(self._build_P_hat(self.F, self.C, self.P)).float()) # n x F+3 + # else: + # # for fine-tuning with different image width, you may use below instead of self.register_buffer + # self.inv_delta_C = torch.tensor(self._build_inv_delta_C(self.F, self.C)).float() # F+3 x F+3 + # self.P_hat = torch.tensor(self._build_P_hat(self.F, self.C, self.P)).float() # n x F+3 + + def _build_C(self, F): + """ Return coordinates of fiducial points in I_r; C """ + ctrl_pts_x = np.linspace(-1.0, 1.0, int(F / 2)) + ctrl_pts_y_top = -1 * np.ones(int(F / 2)) + ctrl_pts_y_bottom = np.ones(int(F / 2)) + ctrl_pts_top = np.stack([ctrl_pts_x, ctrl_pts_y_top], axis=1) + ctrl_pts_bottom = np.stack([ctrl_pts_x, ctrl_pts_y_bottom], axis=1) + C = np.concatenate([ctrl_pts_top, ctrl_pts_bottom], axis=0) + return C # F x 2 + + def _build_inv_delta_C(self, F, C): + """ Return inv_delta_C which is needed to calculate T """ + hat_C = np.zeros((F, F), dtype=float) # F x F + for i in range(0, F): + for j in range(i, F): + r = np.linalg.norm(C[i] - C[j]) + hat_C[i, j] = r + hat_C[j, i] = r + np.fill_diagonal(hat_C, 1) + hat_C = (hat_C ** 2) * np.log(hat_C) + # print(C.shape, hat_C.shape) + delta_C = np.concatenate( # F+3 x F+3 + [ + np.concatenate([np.ones((F, 1)), C, hat_C], axis=1), # F x F+3 + np.concatenate([np.zeros((2, 3)), np.transpose(C)], axis=1), # 2 x F+3 + np.concatenate([np.zeros((1, 3)), np.ones((1, F))], axis=1) # 1 x F+3 + ], + axis=0 + ) + inv_delta_C = np.linalg.inv(delta_C) + return inv_delta_C # F+3 x F+3 + + def _build_P(self, I_r_width, I_r_height): + I_r_grid_x = (np.arange(-I_r_width, I_r_width, 2) + 1.0) / I_r_width # self.I_r_width + I_r_grid_y = (np.arange(-I_r_height, I_r_height, 2) + 1.0) / I_r_height # self.I_r_height + P = np.stack( # self.I_r_width x self.I_r_height x 2 + np.meshgrid(I_r_grid_x, I_r_grid_y), + axis=2 + ) + return P.reshape([-1, 2]) # n (= self.I_r_width x self.I_r_height) x 2 + + def _build_P_hat(self, F, C, P): + n = P.shape[0] # n (= self.I_r_width x self.I_r_height) + P_tile = np.tile(np.expand_dims(P, axis=1), (1, F, 1)) # n x 2 -> n x 1 x 2 -> n x F x 2 + C_tile = np.expand_dims(C, axis=0) # 1 x F x 2 + P_diff = P_tile - C_tile # n x F x 2 + rbf_norm = np.linalg.norm(P_diff, ord=2, axis=2, keepdims=False) # n x F + rbf = np.multiply(np.square(rbf_norm), np.log(rbf_norm + self.eps)) # n x F + P_hat = np.concatenate([np.ones((n, 1)), P, rbf], axis=1) + return P_hat # n x F+3 + + def build_P_prime(self, batch_C_prime): + """ Generate Grid from batch_C_prime [batch_size x F x 2] """ + batch_size = batch_C_prime.size(0) + batch_inv_delta_C = self.inv_delta_C.repeat(batch_size, 1, 1) + batch_P_hat = self.P_hat.repeat(batch_size, 1, 1) + batch_C_prime_with_zeros = torch.cat((batch_C_prime, batch_C_prime.new_zeros( + (batch_size, 3, 2), dtype=torch.float)), dim=1) # batch_size x F+3 x 2 + batch_T = torch.bmm(batch_inv_delta_C, batch_C_prime_with_zeros) # batch_size x F+3 x 2 + batch_P_prime = torch.bmm(batch_P_hat, batch_T) # batch_size x n x 2 + return batch_P_prime # batch_size x n x 2 diff --git a/strhub/models/utils.py b/strhub/models/utils.py new file mode 100644 index 00000000..2e05d776 --- /dev/null +++ b/strhub/models/utils.py @@ -0,0 +1,78 @@ +import os.path +from typing import Sequence + +import torch +from pytorch_lightning.core.saving import ModelIO +from torch import nn + + +def _load_pl_checkpoint(checkpoint, **kwargs): + hparams = checkpoint[ModelIO.CHECKPOINT_HYPER_PARAMS_KEY] + hparams.update(kwargs) + name = hparams['name'] + if name.startswith('abinet'): + from .abinet.system import ABINet as ModelClass + elif name.startswith('crnn'): + from .crnn.system import CRNN as ModelClass + elif name.startswith('parseq'): + from .parseq.system import PARSeq as ModelClass + elif name.startswith('trba'): + from .trba.system import TRBA as ModelClass + elif name.startswith('trbc'): + from .trba.system import TRBC as ModelClass + elif name.startswith('vitstr'): + from .vitstr.system import ViTSTR as ModelClass + else: + raise RuntimeError('Unable to load correct model class') + model = ModelClass._load_model_state(checkpoint, strict=True, **kwargs) + return model + + +def _load_torch_model(checkpoint_path, checkpoint, **kwargs): + import hubconf + name = os.path.basename(checkpoint_path).split('-')[0] + model_factory = getattr(hubconf, name) + model = model_factory(**kwargs) + model.load_state_dict(checkpoint) + return model + + +def load_from_checkpoint(checkpoint_path: str, **kwargs): + checkpoint = torch.load(checkpoint_path, map_location='cpu') + try: + model = _load_pl_checkpoint(checkpoint, **kwargs) + except KeyError: + model = _load_torch_model(checkpoint_path, checkpoint, **kwargs) + return model + + +def parse_model_args(args): + kwargs = {} + arg_types = {t.__name__: t for t in [int, float, str]} + arg_types['bool'] = lambda v: v.lower() == 'true' # special handling for bool + for arg in args: + name, value = arg.split('=', maxsplit=1) + name, arg_type = name.split(':', maxsplit=1) + kwargs[name] = arg_types[arg_type](value) + return kwargs + + +def init_weights(module: nn.Module, name: str = '', exclude: Sequence[str] = ()): + """Initialize the weights using the typical initialization schemes used in SOTA models.""" + if any(map(name.startswith, exclude)): + return + if isinstance(module, nn.Linear): + nn.init.trunc_normal_(module.weight, std=.02) + if module.bias is not None: + nn.init.zeros_(module.bias) + elif isinstance(module, nn.Embedding): + nn.init.trunc_normal_(module.weight, std=.02) + if module.padding_idx is not None: + module.weight.data[module.padding_idx].zero_() + elif isinstance(module, nn.Conv2d): + nn.init.kaiming_normal_(module.weight, mode='fan_out', nonlinearity='relu') + if module.bias is not None: + nn.init.zeros_(module.bias) + elif isinstance(module, (nn.LayerNorm, nn.BatchNorm2d, nn.GroupNorm)): + nn.init.ones_(module.weight) + nn.init.zeros_(module.bias) diff --git a/strhub/models/vitstr/__init__.py b/strhub/models/vitstr/__init__.py new file mode 100644 index 00000000..19e98567 --- /dev/null +++ b/strhub/models/vitstr/__init__.py @@ -0,0 +1,12 @@ +r""" +Atienza, Rowel. "Vision Transformer for Fast and Efficient Scene Text Recognition." +In International Conference on Document Analysis and Recognition (ICDAR). 2021. + +https://arxiv.org/abs/2105.08582 + +All source files, except `system.py`, are based on the implementation listed below, +and hence are released under the license of the original. + +Source: https://github.com/roatienza/deep-text-recognition-benchmark +License: Apache License 2.0 (see LICENSE file in project root) +""" diff --git a/strhub/models/vitstr/model.py b/strhub/models/vitstr/model.py new file mode 100644 index 00000000..9543d2c5 --- /dev/null +++ b/strhub/models/vitstr/model.py @@ -0,0 +1,53 @@ +''' +Implementation of ViTSTR based on timm VisionTransformer. + +TODO: +1) distilled deit backbone +2) base deit backbone + +Copyright 2021 Rowel Atienza +''' + +import torch +import torch.nn as nn + +from timm.models.vision_transformer import VisionTransformer + + +class ViTSTR(VisionTransformer): + ''' + ViTSTR is basically a ViT that uses DeiT weights. + Modified head to support a sequence of characters prediction for STR. + ''' + + def __init__(self, *args, **kwargs): + super().__init__(*args, **kwargs) + + def reset_classifier(self, num_classes): + self.num_classes = num_classes + self.head = nn.Linear(self.embed_dim, num_classes) if num_classes > 0 else nn.Identity() + + def forward_features(self, x): + B = x.shape[0] + x = self.patch_embed(x) + + cls_tokens = self.cls_token.expand(B, -1, -1) # stole cls_tokens impl from Phil Wang, thanks + x = torch.cat((cls_tokens, x), dim=1) + x = x + self.pos_embed + x = self.pos_drop(x) + + for blk in self.blocks: + x = blk(x) + + x = self.norm(x) + return x + + def forward(self, x, seqlen=25): + x = self.forward_features(x) + x = x[:, :seqlen] + + # batch, seqlen, embsize + b, s, e = x.size() + x = x.reshape(b * s, e) + x = self.head(x).view(b, s, self.num_classes) + return x diff --git a/strhub/models/vitstr/system.py b/strhub/models/vitstr/system.py new file mode 100644 index 00000000..c72b312b --- /dev/null +++ b/strhub/models/vitstr/system.py @@ -0,0 +1,58 @@ +# Scene Text Recognition Model Hub +# Copyright 2022 Darwin Bautista +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# https://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +from typing import Sequence, Any + +import torch +from pytorch_lightning.utilities.types import STEP_OUTPUT +from torch import Tensor + +from strhub.models.base import CrossEntropySystem +from strhub.models.utils import init_weights +from .model import ViTSTR as Model + + +class ViTSTR(CrossEntropySystem): + + def __init__(self, charset_train: str, charset_test: str, max_label_length: int, + batch_size: int, lr: float, warmup_pct: float, weight_decay: float, + img_size: Sequence[int], patch_size: Sequence[int], embed_dim: int, num_heads: int, + **kwargs: Any) -> None: + super().__init__(charset_train, charset_test, batch_size, lr, warmup_pct, weight_decay) + self.save_hyperparameters() + self.max_label_length = max_label_length + # We don't predict nor + self.model = Model(img_size=img_size, patch_size=patch_size, depth=12, mlp_ratio=4, qkv_bias=True, + embed_dim=embed_dim, num_heads=num_heads, num_classes=len(self.tokenizer) - 2) + # Non-zero weight init for the head + self.model.head.apply(init_weights) + + @torch.jit.ignore + def no_weight_decay(self): + return {'model.' + n for n in self.model.no_weight_decay()} + + def forward(self, images: Tensor, max_length: int = None) -> Tensor: + max_length = self.max_label_length if max_length is None else min(max_length, self.max_label_length) + logits = self.model.forward(images, max_length + 2) # +2 tokens for [GO] and [s] + # Truncate to conform to other models. [GO] in ViTSTR is actually used as the padding (therefore, ignored). + # First position corresponds to the class token, which is unused and ignored in the original work. + logits = logits[:, 1:] + return logits + + def training_step(self, batch, batch_idx) -> STEP_OUTPUT: + images, labels = batch + loss = self.forward_logits_loss(images, labels)[1] + self.log('loss', loss) + return loss diff --git a/test.py b/test.py new file mode 100755 index 00000000..8a8d32e9 --- /dev/null +++ b/test.py @@ -0,0 +1,129 @@ +#!/usr/bin/env python3 +# Scene Text Recognition Model Hub +# Copyright 2022 Darwin Bautista +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# https://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +import argparse +import string +import sys +from dataclasses import dataclass + +from tqdm import tqdm + +from strhub.data.module import SceneTextDataModule +from strhub.models.utils import load_from_checkpoint, parse_model_args + + +@dataclass +class Result: + dataset: str + num_samples: int + accuracy: float + ned: float + confidence: float + label_length: float + + +def print_results_table(results: list[Result], file=None): + w = max(map(len, map(getattr, results, ['dataset'] * len(results)))) + w = max(w, len('Dataset'), len('Combined')) + print('| {:<{w}} | # samples | Accuracy | 1 - NED | Confidence | Label Length |'.format('Dataset', w=w), file=file) + print('|:{:-<{w}}:|----------:|---------:|--------:|-----------:|-------------:|'.format('----', w=w), file=file) + c = Result('Combined', 0, 0, 0, 0, 0) + for res in results: + c.num_samples += res.num_samples + c.accuracy += res.num_samples * res.accuracy + c.ned += res.num_samples * res.ned + c.confidence += res.num_samples * res.confidence + c.label_length += res.num_samples * res.label_length + print(f'| {res.dataset:<{w}} | {res.num_samples:>9} | {res.accuracy:>8.2f} | {res.ned:>7.2f} ' + f'| {res.confidence:>10.2f} | {res.label_length:>12.2f} |', file=file) + c.accuracy /= c.num_samples + c.ned /= c.num_samples + c.confidence /= c.num_samples + c.label_length /= c.num_samples + print('|-{:-<{w}}-|-----------|----------|---------|------------|--------------|'.format('----', w=w), file=file) + print(f'| {c.dataset:<{w}} | {c.num_samples:>9} | {c.accuracy:>8.2f} | {c.ned:>7.2f} ' + f'| {c.confidence:>10.2f} | {c.label_length:>12.2f} |', file=file) + + +def main(): + parser = argparse.ArgumentParser() + parser.add_argument('checkpoint', help='Model checkpoint') + parser.add_argument('--batch_size', type=int, default=512) + parser.add_argument('--num_workers', type=int, default=4) + parser.add_argument('--cased', action='store_true', default=False, help='Cased comparison') + parser.add_argument('--punctuation', action='store_true', default=False, help='Check punctuation') + parser.add_argument('--new', action='store_true', default=False, help='Evaluate on new benchmark datasets') + parser.add_argument('--rotation', type=int, default=0, help='Angle of rotation (counter clockwise) in degrees.') + parser.add_argument('--device', default='cuda') + args, unknown = parser.parse_known_args() + kwargs = parse_model_args(unknown) + + charset_test = string.digits + string.ascii_lowercase + if args.cased: + charset_test += string.ascii_uppercase + if args.punctuation: + charset_test += string.punctuation + kwargs.update({'charset_test': charset_test}) + print(f'Additional keyword arguments: {kwargs}') + + model = load_from_checkpoint(args.checkpoint, **kwargs).eval().to(args.device) + model.freeze() # disable autograd + hp = model.hparams + datamodule = SceneTextDataModule('data', '_unused_', hp.img_size, hp.max_label_length, hp.charset_train, hp.charset_test, + args.batch_size, args.num_workers, False, args.rotation) + + test_set = SceneTextDataModule.TEST_ABINET + SceneTextDataModule.TEST_TRBA + if args.new: + test_set += SceneTextDataModule.TEST_NEW + test_set = sorted(set(test_set)) + + results = {} + max_width = max(map(len, test_set)) + for name, dataloader in datamodule.test_dataloaders(test_set).items(): + total = 0 + correct = 0 + ned = 0 + confidence = 0 + label_length = 0 + for imgs, labels in tqdm(iter(dataloader), desc=f'{name:>{max_width}}'): + res = model.test_step((imgs.to(model.device), labels), -1)['output'] + total += res.num_samples + correct += res.correct + ned += res.ned + confidence += res.confidence + label_length += res.label_length + accuracy = 100 * correct / total + mean_ned = 100 * (1 - ned / total) + mean_conf = 100 * confidence / total + mean_label_length = label_length / total + results[name] = Result(name, total, accuracy, mean_ned, mean_conf, mean_label_length) + + result_groups = { + 'ABINet': SceneTextDataModule.TEST_ABINET, + 'TRBA': SceneTextDataModule.TEST_TRBA + } + if args.new: + result_groups.update({'New': SceneTextDataModule.TEST_NEW}) + with open(args.checkpoint + '.log.txt', 'w') as f: + for out in [f, sys.stdout]: + for group, subset in result_groups.items(): + print(f'{group} set:', file=out) + print_results_table([results[s] for s in subset], out) + print('\n', file=out) + + +if __name__ == '__main__': + main() diff --git a/tools/art_converter.py b/tools/art_converter.py new file mode 100755 index 00000000..f61e0b54 --- /dev/null +++ b/tools/art_converter.py @@ -0,0 +1,26 @@ +#!/usr/bin/env python3 + +import json + +with open('train_task2_labels.json', 'r', encoding='utf8') as f: + d = json.load(f) + +with open('gt.txt', 'w', encoding='utf8') as f: + for k, v in d.items(): + if len(v) != 1: + print('error', v) + v = v[0] + if v['language'].lower() != 'latin': + # print('Skipping non-Latin:', v) + continue + if v['illegibility']: + # print('Skipping unreadable:', v) + continue + label = v['transcription'].strip() + if not label: + # print('Skipping blank label') + continue + if '#' in label and label != 'LocaL#3': + # print('Skipping corrupted label') + continue + f.write('\t'.join(['train_task2_images/' + k + '.jpg', label]) + '\n') diff --git a/tools/case_sensitive_str_datasets_converter.py b/tools/case_sensitive_str_datasets_converter.py new file mode 100755 index 00000000..7ce7c0dd --- /dev/null +++ b/tools/case_sensitive_str_datasets_converter.py @@ -0,0 +1,28 @@ +#!/usr/bin/env python3 + +import os.path +import sys +from pathlib import Path + +d = sys.argv[1] +p = Path(d) + +gt = [] + +num_samples = len(list(p.glob('label/*.txt'))) +ext = 'jpg' if p.joinpath('IMG', '1.jpg').is_file() else 'png' + +for i in range(1, num_samples + 1): + img = p.joinpath('IMG', f'{i}.{ext}') + name = os.path.splitext(img.name)[0] + + with open(p.joinpath('label', f'{i}.txt'), 'r') as f: + label = f.readline() + gt.append((os.path.join('IMG', img.name), label)) + +with open(d + '/lmdb.txt', 'w', encoding='utf-8') as f: + for line in gt: + fname, label = line + fname = fname.strip() + label = label.strip() + f.write('\t'.join([fname, label]) + '\n') diff --git a/tools/coco_2_converter.py b/tools/coco_2_converter.py new file mode 100755 index 00000000..2a2d614c --- /dev/null +++ b/tools/coco_2_converter.py @@ -0,0 +1,126 @@ +#!/usr/bin/env python3 +import argparse +import html +import math +import os +import os.path as osp +from functools import partial + +import mmcv +from PIL import Image +from mmocr.utils.fileio import list_to_file + + +def parse_args(): + parser = argparse.ArgumentParser( + description='Generate training and validation set of TextOCR ' + 'by cropping box image.') + parser.add_argument('root_path', help='Root dir path of TextOCR') + parser.add_argument( + 'n_proc', default=1, type=int, help='Number of processes to run') + args = parser.parse_args() + return args + + +def process_img(args, src_image_root, dst_image_root): + # Dirty hack for multi-processing + img_idx, img_info, anns = args + src_img = Image.open(osp.join(src_image_root, 'train2014', img_info['file_name'])) + src_w, src_h = src_img.size + labels = [] + for ann_idx, ann in enumerate(anns): + text_label = html.unescape(ann['utf8_string'].strip()) + + # Ignore empty labels + if not text_label or ann['class'] != 'machine printed' or ann['language'] != 'english' or \ + ann['legibility'] != 'legible': + continue + + # Some labels and images with '#' in the middle are actually good, but some aren't so we just filter them all. + if text_label != '#' and '#' in text_label: + continue + + # Some labels use '*' to denote unreadable characters + if text_label.startswith('*') or text_label.endswith('*'): + continue + + pad = 2 + x, y, w, h = ann['bbox'] + x, y = max(0, math.floor(x) - pad), max(0, math.floor(y) - pad) + w, h = math.ceil(w), math.ceil(h) + x2, y2 = min(src_w, x + w + 2 * pad), min(src_h, y + h + 2 * pad) + dst_img = src_img.crop((x, y, x2, y2)) + dst_img_name = f'img_{img_idx}_{ann_idx}.jpg' + dst_img_path = osp.join(dst_image_root, dst_img_name) + # Preserve JPEG quality + dst_img.save(dst_img_path, qtables=src_img.quantization) + labels.append(f'{osp.basename(dst_image_root)}/{dst_img_name}' + f' {text_label}') + src_img.close() + return labels + + +def convert_textocr(root_path, + dst_image_path, + dst_label_filename, + annotation_filename, + img_start_idx=0, + nproc=1): + annotation_path = osp.join(root_path, annotation_filename) + if not osp.exists(annotation_path): + raise Exception( + f'{annotation_path} not exists, please check and try again.') + src_image_root = root_path + + # outputs + dst_label_file = osp.join(root_path, dst_label_filename) + dst_image_root = osp.join(root_path, dst_image_path) + os.makedirs(dst_image_root, exist_ok=True) + + annotation = mmcv.load(annotation_path) + split = 'train' if 'train' in dst_label_filename else 'val' + + process_img_with_path = partial( + process_img, + src_image_root=src_image_root, + dst_image_root=dst_image_root) + tasks = [] + for img_idx, img_info in enumerate(annotation['imgs'].values()): + if img_info['set'] != split: + continue + ann_ids = annotation['imgToAnns'][str(img_info['id'])] + anns = [annotation['anns'][str(ann_id)] for ann_id in ann_ids] + tasks.append((img_idx + img_start_idx, img_info, anns)) + + labels_list = mmcv.track_parallel_progress( + process_img_with_path, tasks, keep_order=True, nproc=nproc) + final_labels = [] + for label_list in labels_list: + final_labels += label_list + list_to_file(dst_label_file, final_labels) + return len(annotation['imgs']) + + +def main(): + args = parse_args() + root_path = args.root_path + print('Processing training set...') + num_train_imgs = convert_textocr( + root_path=root_path, + dst_image_path='image', + dst_label_filename='train_label.txt', + annotation_filename='cocotext.v2.json', + nproc=args.n_proc) + print('Processing validation set...') + convert_textocr( + root_path=root_path, + dst_image_path='image_val', + dst_label_filename='val_label.txt', + annotation_filename='cocotext.v2.json', + img_start_idx=num_train_imgs, + nproc=args.n_proc) + print('Finish') + + +if __name__ == '__main__': + main() diff --git a/tools/coco_text_converter.py b/tools/coco_text_converter.py new file mode 100755 index 00000000..09d130db --- /dev/null +++ b/tools/coco_text_converter.py @@ -0,0 +1,15 @@ +#!/usr/bin/env python3 + +for s in ['train', 'val']: + with open('{}_words_gt.txt'.format(s), 'r', encoding='utf8') as f: + d = f.readlines() + + with open('{}_lmdb.txt'.format(s), 'w', encoding='utf8') as f: + for line in d: + try: + fname, label = line.split(',', maxsplit=1) + except ValueError: + continue + fname = '{}_words/{}.jpg'.format(s, fname.strip()) + label = label.strip().strip('|') + f.write('\t'.join([fname, label]) + '\n') diff --git a/tools/create_lmdb_dataset.py b/tools/create_lmdb_dataset.py new file mode 100755 index 00000000..82dd3f4d --- /dev/null +++ b/tools/create_lmdb_dataset.py @@ -0,0 +1,78 @@ +#!/usr/bin/env python3 +""" a modified version of CRNN torch repository https://github.com/bgshih/crnn/blob/master/tool/create_dataset.py """ +import io +import os + +import fire +import lmdb +import numpy as np +from PIL import Image + + +def checkImageIsValid(imageBin): + if imageBin is None: + return False + img = Image.open(io.BytesIO(imageBin)).convert('RGB') + return np.prod(img.size) > 0 + + +def writeCache(env, cache): + with env.begin(write=True) as txn: + for k, v in cache.items(): + txn.put(k, v) + + +def createDataset(inputPath, gtFile, outputPath, checkValid=True): + """ + Create LMDB dataset for training and evaluation. + ARGS: + inputPath : input folder path where starts imagePath + outputPath : LMDB output path + gtFile : list of image path and label + checkValid : if true, check the validity of every image + """ + os.makedirs(outputPath, exist_ok=True) + env = lmdb.open(outputPath, map_size=1099511627776) + + cache = {} + cnt = 1 + + with open(gtFile, 'r', encoding='utf-8') as f: + data = f.readlines() + + nSamples = len(data) + for i, line in enumerate(data): + imagePath, label = line.strip().split(maxsplit=1) + imagePath = os.path.join(inputPath, imagePath) + with open(imagePath, 'rb') as f: + imageBin = f.read() + if checkValid: + try: + img = Image.open(io.BytesIO(imageBin)).convert('RGB') + except IOError as e: + with open(outputPath + '/error_image_log.txt', 'a') as log: + log.write('{}-th image data occured error: {}, {}\n'.format(i, imagePath, e)) + continue + if np.prod(img.size) == 0: + print('%s is not a valid image' % imagePath) + continue + + imageKey = 'image-%09d'.encode() % cnt + labelKey = 'label-%09d'.encode() % cnt + cache[imageKey] = imageBin + cache[labelKey] = label.encode() + + if cnt % 1000 == 0: + writeCache(env, cache) + cache = {} + print('Written %d / %d' % (cnt, nSamples)) + cnt += 1 + nSamples = cnt - 1 + cache['num-samples'.encode()] = str(nSamples).encode() + writeCache(env, cache) + env.close() + print('Created dataset with %d samples' % nSamples) + + +if __name__ == '__main__': + fire.Fire(createDataset) diff --git a/tools/filter_lmdb.py b/tools/filter_lmdb.py new file mode 100755 index 00000000..0d1b4451 --- /dev/null +++ b/tools/filter_lmdb.py @@ -0,0 +1,57 @@ +#!/usr/bin/env python3 +import io +import os +from argparse import ArgumentParser + +import numpy as np +import lmdb +from PIL import Image + + +def main(): + parser = ArgumentParser() + parser.add_argument('inputs', nargs='+', help='Path to input LMDBs') + parser.add_argument('--output', help='Path to output LMDB') + parser.add_argument('--min_image_dim', type=int, default=8) + args = parser.parse_args() + + os.makedirs(args.output, exist_ok=True) + with lmdb.open(args.output, map_size=1099511627776) as env_out: + in_samples = 0 + out_samples = 0 + samples_per_chunk = 1000 + for lmdb_in in args.inputs: + with lmdb.open(lmdb_in, readonly=True, max_readers=1, lock=False) as env_in: + with env_in.begin() as txn: + num_samples = int(txn.get('num-samples'.encode())) + in_samples += num_samples + chunks = np.array_split(range(num_samples), num_samples // samples_per_chunk) + for chunk in chunks: + cache = {} + with env_in.begin() as txn: + for index in chunk: + index += 1 # lmdb starts at 1 + image_key = f'image-{index:09d}'.encode() + image_bin = txn.get(image_key) + img = Image.open(io.BytesIO(image_bin)) + w, h = img.size + if w < args.min_image_dim or h < args.min_image_dim: + print(f'Skipping: {index}, w = {w}, h = {h}') + continue + out_samples += 1 # increment. start at 1 + label_key = f'label-{index:09d}'.encode() + out_label_key = f'label-{out_samples:09d}'.encode() + out_image_key = f'image-{out_samples:09d}'.encode() + cache[out_label_key] = txn.get(label_key) + cache[out_image_key] = image_bin + with env_out.begin(write=True) as txn: + for k, v in cache.items(): + txn.put(k, v) + print(f'Written samples from {chunk[0]} to {chunk[-1]}') + with env_out.begin(write=True) as txn: + txn.put('num-samples'.encode(), str(out_samples).encode()) + print(f'Written {out_samples} samples to {args.output} out of {in_samples} input samples.') + + +if __name__ == '__main__': + main() diff --git a/tools/lsvt_converter.py b/tools/lsvt_converter.py new file mode 100755 index 00000000..e16c10bd --- /dev/null +++ b/tools/lsvt_converter.py @@ -0,0 +1,107 @@ +#!/usr/bin/env python3 +import argparse +import os +import os.path as osp +import re +from functools import partial + +import mmcv +import numpy as np +from PIL import Image +from mmocr.utils.fileio import list_to_file + + +def parse_args(): + parser = argparse.ArgumentParser( + description='Generate training set of LSVT ' + 'by cropping box image.') + parser.add_argument('root_path', help='Root dir path of LSVT') + parser.add_argument( + 'n_proc', default=1, type=int, help='Number of processes to run') + args = parser.parse_args() + return args + + +def process_img(args, src_image_root, dst_image_root): + # Dirty hack for multi-processing + img_idx, img_info, anns = args + try: + src_img = Image.open(osp.join(src_image_root, 'train_full_images_0/{}.jpg'.format(img_info))) + except IOError: + src_img = Image.open(osp.join(src_image_root, 'train_full_images_1/{}.jpg'.format(img_info))) + blacklist = ['LOFTINESS*'] + whitelist = ['#Find YOUR Fun#', 'Story #', '*0#'] + labels = [] + for ann_idx, ann in enumerate(anns): + text_label = ann['transcription'] + + # Ignore illegible or words with non-Latin characters + if ann['illegibility'] or re.findall(r'[\u4e00-\u9fff]+', text_label) or text_label in blacklist or \ + ('#' in text_label and text_label not in whitelist): + continue + + points = np.asarray(ann['points']) + x1, y1 = points.min(axis=0) + x2, y2 = points.max(axis=0) + + dst_img = src_img.crop((x1, y1, x2, y2)) + dst_img_name = f'img_{img_idx}_{ann_idx}.jpg' + dst_img_path = osp.join(dst_image_root, dst_img_name) + # Preserve JPEG quality + dst_img.save(dst_img_path, qtables=src_img.quantization) + labels.append(f'{osp.basename(dst_image_root)}/{dst_img_name}' + f' {text_label}') + src_img.close() + return labels + + +def convert_lsvt(root_path, + dst_image_path, + dst_label_filename, + annotation_filename, + img_start_idx=0, + nproc=1): + annotation_path = osp.join(root_path, annotation_filename) + if not osp.exists(annotation_path): + raise Exception( + f'{annotation_path} not exists, please check and try again.') + src_image_root = root_path + + # outputs + dst_label_file = osp.join(root_path, dst_label_filename) + dst_image_root = osp.join(root_path, dst_image_path) + os.makedirs(dst_image_root, exist_ok=True) + + annotation = mmcv.load(annotation_path) + + process_img_with_path = partial( + process_img, + src_image_root=src_image_root, + dst_image_root=dst_image_root) + tasks = [] + for img_idx, (img_info, anns) in enumerate(annotation.items()): + tasks.append((img_idx + img_start_idx, img_info, anns)) + labels_list = mmcv.track_parallel_progress( + process_img_with_path, tasks, keep_order=True, nproc=nproc) + final_labels = [] + for label_list in labels_list: + final_labels += label_list + list_to_file(dst_label_file, final_labels) + return len(annotation) + + +def main(): + args = parse_args() + root_path = args.root_path + print('Processing training set...') + convert_lsvt( + root_path=root_path, + dst_image_path='image_train', + dst_label_filename='train_label.txt', + annotation_filename='train_full_labels.json', + nproc=args.n_proc) + print('Finish') + + +if __name__ == '__main__': + main() diff --git a/tools/mlt19_converter.py b/tools/mlt19_converter.py new file mode 100755 index 00000000..665d497f --- /dev/null +++ b/tools/mlt19_converter.py @@ -0,0 +1,15 @@ +#!/usr/bin/env python3 + +import sys + +root = sys.argv[1] + +with open(root + '/gt.txt', 'r') as f: + d = f.readlines() + +with open(root + '/lmdb.txt', 'w') as f: + for line in d: + img, script, label = line.split(',', maxsplit=2) + label = label.strip() + if label and script in ['Latin', 'Symbols']: + f.write('\t'.join([img, label]) + '\n') diff --git a/tools/openvino_converter.py b/tools/openvino_converter.py new file mode 100755 index 00000000..333a12b6 --- /dev/null +++ b/tools/openvino_converter.py @@ -0,0 +1,116 @@ +#!/usr/bin/env python3 +import math +import os +import os.path as osp +from argparse import ArgumentParser +from functools import partial + +import mmcv +from PIL import Image + +from mmocr.utils.fileio import list_to_file + + +def parse_args(): + parser = ArgumentParser(description='Generate training and validation set ' + 'of OpenVINO annotations for Open ' + 'Images by cropping box image.') + parser.add_argument( + 'root_path', help='Root dir containing images and annotations') + parser.add_argument( + 'n_proc', default=1, type=int, help='Number of processes to run') + args = parser.parse_args() + return args + + +def process_img(args, src_image_root, dst_image_root): + # Dirty hack for multi-processing + img_idx, img_info, anns = args + src_img = Image.open(osp.join(src_image_root, img_info['file_name'])) + labels = [] + for ann_idx, ann in enumerate(anns): + attrs = ann['attributes'] + text_label = attrs['transcription'] + + # Ignore illegible or non-English words + if not attrs['legible'] or attrs['language'] != 'english': + continue + + x, y, w, h = ann['bbox'] + x, y = max(0, math.floor(x)), max(0, math.floor(y)) + w, h = math.ceil(w), math.ceil(h) + dst_img = src_img.crop((x, y, x + w, y + h)) + dst_img_name = f'img_{img_idx}_{ann_idx}.jpg' + dst_img_path = osp.join(dst_image_root, dst_img_name) + # Preserve JPEG quality + dst_img.save(dst_img_path, qtables=src_img.quantization) + labels.append(f'{osp.basename(dst_image_root)}/{dst_img_name}' + f' {text_label}') + src_img.close() + return labels + + +def convert_openimages(root_path, + dst_image_path, + dst_label_filename, + annotation_filename, + img_start_idx=0, + nproc=1): + annotation_path = osp.join(root_path, annotation_filename) + if not osp.exists(annotation_path): + raise Exception( + f'{annotation_path} not exists, please check and try again.') + src_image_root = root_path + + # outputs + dst_label_file = osp.join(root_path, dst_label_filename) + dst_image_root = osp.join(root_path, dst_image_path) + os.makedirs(dst_image_root, exist_ok=True) + + annotation = mmcv.load(annotation_path) + + process_img_with_path = partial( + process_img, + src_image_root=src_image_root, + dst_image_root=dst_image_root) + tasks = [] + anns = {} + for ann in annotation['annotations']: + anns.setdefault(ann['image_id'], []).append(ann) + for img_idx, img_info in enumerate(annotation['images']): + tasks.append((img_idx + img_start_idx, img_info, anns[img_info['id']])) + labels_list = mmcv.track_parallel_progress( + process_img_with_path, tasks, keep_order=True, nproc=nproc) + final_labels = [] + for label_list in labels_list: + final_labels += label_list + list_to_file(dst_label_file, final_labels) + return len(annotation['images']) + + +def main(): + args = parse_args() + root_path = args.root_path + print('Processing training set...') + num_train_imgs = 0 + for s in '125f': + num_train_imgs = convert_openimages( + root_path=root_path, + dst_image_path=f'image_{s}', + dst_label_filename=f'train_{s}_label.txt', + annotation_filename=f'text_spotting_openimages_v5_train_{s}.json', + img_start_idx=num_train_imgs, + nproc=args.n_proc) + print('Processing validation set...') + convert_openimages( + root_path=root_path, + dst_image_path='image_val', + dst_label_filename='val_label.txt', + annotation_filename='text_spotting_openimages_v5_validation.json', + img_start_idx=num_train_imgs, + nproc=args.n_proc) + print('Finish') + + +if __name__ == '__main__': + main() diff --git a/tools/test_abinet_lm_acc.py b/tools/test_abinet_lm_acc.py new file mode 100755 index 00000000..00c26665 --- /dev/null +++ b/tools/test_abinet_lm_acc.py @@ -0,0 +1,101 @@ +#!/usr/bin/env python3 +import argparse +import string +import sys + +import torch +import torch.nn.functional as F +from torch import Tensor +from torch.nn.utils.rnn import pad_sequence + +from tqdm import tqdm + +from strhub.data.module import SceneTextDataModule +from strhub.models.abinet.system import ABINet + +sys.path.insert(0, '.') +from hubconf import _get_config +from test import Result, print_results_table + + +class ABINetLM(ABINet): + + def _encode(self, labels): + targets = [torch.arange(self.max_label_length + 1)] # dummy target. used to set pad_sequence() length + lengths = [] + for label in labels: + targets.append(torch.as_tensor([self.tokenizer._stoi[c] for c in label])) + lengths.append(len(label) + 1) + targets = pad_sequence(targets, batch_first=True, padding_value=0)[1:] # exclude dummy target + lengths = torch.as_tensor(lengths, device=self.device) + targets = F.one_hot(targets, len(self.tokenizer._stoi))[..., :len(self.tokenizer._stoi) - 2].float().to(self.device) + return targets, lengths + + def forward(self, labels: Tensor, max_length: int = None) -> Tensor: + targets, lengths = self._encode(labels) + return self.model.language(targets, lengths)['logits'] + + +def main(): + parser = argparse.ArgumentParser(description='Measure the word accuracy of ABINet LM using the ground truth as input') + parser.add_argument('checkpoint', help='Official pretrained weights for ABINet-LV (best-train-abinet.pth)') + parser.add_argument('--batch_size', type=int, default=512) + parser.add_argument('--num_workers', type=int, default=4) + parser.add_argument('--new', action='store_true', default=False, help='Evaluate on new benchmark datasets') + parser.add_argument('--device', default='cuda') + args = parser.parse_args() + + # charset used by original ABINet + charset = string.ascii_lowercase + '1234567890' + ckpt = torch.load(args.checkpoint) + + config = _get_config('abinet', charset_train=charset, charset_test=charset) + model = ABINetLM(**config) + model.model.load_state_dict(ckpt['model']) + + model = model.eval().to(args.device) + model.freeze() # disable autograd + hp = model.hparams + datamodule = SceneTextDataModule('data', '_unused_', hp.img_size, hp.max_label_length, hp.charset_train, + hp.charset_test, + args.batch_size, args.num_workers, False) + + test_set = SceneTextDataModule.TEST_TRBA + if args.new: + test_set += SceneTextDataModule.TEST_NEW + test_set = sorted(set(test_set)) + + results = {} + max_width = max(map(len, test_set)) + for name, dataloader in datamodule.test_dataloaders(test_set).items(): + total = 0 + correct = 0 + ned = 0 + confidence = 0 + label_length = 0 + for _, labels in tqdm(iter(dataloader), desc=f'{name:>{max_width}}'): + res = model.test_step((labels, labels), -1)['output'] + total += res.num_samples + correct += res.correct + ned += res.ned + confidence += res.confidence + label_length += res.label_length + accuracy = 100 * correct / total + mean_ned = 100 * (1 - ned / total) + mean_conf = 100 * confidence / total + mean_label_length = label_length / total + results[name] = Result(name, total, accuracy, mean_ned, mean_conf, mean_label_length) + + result_groups = { + 'TRBA': SceneTextDataModule.TEST_TRBA + } + if args.new: + result_groups.update({'New': SceneTextDataModule.TEST_NEW}) + for group, subset in result_groups.items(): + print(f'{group} set:') + print_results_table([results[s] for s in subset]) + print('\n') + + +if __name__ == '__main__': + main() diff --git a/tools/textocr_converter.py b/tools/textocr_converter.py new file mode 100755 index 00000000..6a9d7150 --- /dev/null +++ b/tools/textocr_converter.py @@ -0,0 +1,142 @@ +#!/usr/bin/env python3 +# Copyright (c) OpenMMLab. All rights reserved. +import argparse +import math +import os +import os.path as osp +from functools import partial + +import mmcv +import numpy as np +from PIL import Image +from mmocr.utils.fileio import list_to_file + + +def parse_args(): + parser = argparse.ArgumentParser( + description='Generate training and validation set of TextOCR ' + 'by cropping box image.') + parser.add_argument('root_path', help='Root dir path of TextOCR') + parser.add_argument( + 'n_proc', default=1, type=int, help='Number of processes to run') + parser.add_argument('--rectify_pose', action='store_true', + help='Fix pose of rotated text to make them horizontal') + args = parser.parse_args() + return args + + +def rectify_image_pose(image, top_left, points): + # Points-based heuristics for determining text orientation w.r.t. bounding box + points = np.asarray(points).reshape(-1, 2) + dist = ((points - np.asarray(top_left)) ** 2).sum(axis=1) + left_midpoint = (points[0] + points[-1]) / 2 + right_corner_points = ((points - left_midpoint) ** 2).sum(axis=1).argsort()[-2:] + right_midpoint = points[right_corner_points].sum(axis=0) / 2 + d_x, d_y = abs(right_midpoint - left_midpoint) + + if dist[0] + dist[-1] <= dist[right_corner_points].sum(): + if d_x >= d_y: + rot = 0 + else: + rot = 90 + else: + if d_x >= d_y: + rot = 180 + else: + rot = -90 + if rot: + image = image.rotate(rot, expand=True) + return image + + +def process_img(args, src_image_root, dst_image_root): + # Dirty hack for multi-processing + img_idx, img_info, anns, rectify_pose = args + src_img = Image.open(osp.join(src_image_root, img_info['file_name'])) + labels = [] + for ann_idx, ann in enumerate(anns): + text_label = ann['utf8_string'] + + # Ignore illegible or non-English words + if text_label == '.': + continue + + x, y, w, h = ann['bbox'] + x, y = max(0, math.floor(x)), max(0, math.floor(y)) + w, h = math.ceil(w), math.ceil(h) + dst_img = src_img.crop((x, y, x + w, y + h)) + if rectify_pose: + dst_img = rectify_image_pose(dst_img, (x, y), ann['points']) + dst_img_name = f'img_{img_idx}_{ann_idx}.jpg' + dst_img_path = osp.join(dst_image_root, dst_img_name) + # Preserve JPEG quality + dst_img.save(dst_img_path, qtables=src_img.quantization) + labels.append(f'{osp.basename(dst_image_root)}/{dst_img_name}' + f' {text_label}') + src_img.close() + return labels + + +def convert_textocr(root_path, + dst_image_path, + dst_label_filename, + annotation_filename, + img_start_idx=0, + nproc=1, + rectify_pose=False): + annotation_path = osp.join(root_path, annotation_filename) + if not osp.exists(annotation_path): + raise Exception( + f'{annotation_path} not exists, please check and try again.') + src_image_root = root_path + + # outputs + dst_label_file = osp.join(root_path, dst_label_filename) + dst_image_root = osp.join(root_path, dst_image_path) + os.makedirs(dst_image_root, exist_ok=True) + + annotation = mmcv.load(annotation_path) + + process_img_with_path = partial( + process_img, + src_image_root=src_image_root, + dst_image_root=dst_image_root) + tasks = [] + for img_idx, img_info in enumerate(annotation['imgs'].values()): + ann_ids = annotation['imgToAnns'][img_info['id']] + anns = [annotation['anns'][ann_id] for ann_id in ann_ids] + tasks.append((img_idx + img_start_idx, img_info, anns, rectify_pose)) + labels_list = mmcv.track_parallel_progress( + process_img_with_path, tasks, keep_order=True, nproc=nproc) + final_labels = [] + for label_list in labels_list: + final_labels += label_list + list_to_file(dst_label_file, final_labels) + return len(annotation['imgs']) + + +def main(): + args = parse_args() + root_path = args.root_path + print('Processing training set...') + num_train_imgs = convert_textocr( + root_path=root_path, + dst_image_path='image', + dst_label_filename='train_label.txt', + annotation_filename='TextOCR_0.1_train.json', + nproc=args.n_proc, + rectify_pose=args.rectify_pose) + print('Processing validation set...') + convert_textocr( + root_path=root_path, + dst_image_path='image', + dst_label_filename='val_label.txt', + annotation_filename='TextOCR_0.1_val.json', + img_start_idx=num_train_imgs, + nproc=args.n_proc, + rectify_pose=args.rectify_pose) + print('Finish') + + +if __name__ == '__main__': + main() diff --git a/train.py b/train.py new file mode 100755 index 00000000..8a7fb621 --- /dev/null +++ b/train.py @@ -0,0 +1,73 @@ +#!/usr/bin/env python3 +# Scene Text Recognition Model Hub +# Copyright 2022 Darwin Bautista +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# https://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +from pathlib import Path + +import hydra +from omegaconf import DictConfig, open_dict +from pytorch_lightning import Trainer +from pytorch_lightning.callbacks import ModelCheckpoint, StochasticWeightAveraging +from pytorch_lightning.loggers import TensorBoardLogger +from pytorch_lightning.plugins import DDPPlugin + +from strhub.data.module import SceneTextDataModule +from strhub.models.base import BaseSystem + + +@hydra.main(config_path='configs', config_name='main') +def main(config: DictConfig): + trainer_plugins = None + with open_dict(config): + # Resolve absolute path to data.root_dir + config.data.root_dir = hydra.utils.to_absolute_path(config.data.root_dir) + if config.trainer.get('resume_from_checkpoint', None) is not None: + config.trainer.resume_from_checkpoint = hydra.utils.to_absolute_path(config.trainer.resume_from_checkpoint) + # Special handling for GPU-affected config + gpus = config.trainer.get('gpus', 0) + if gpus: + # Use mixed-precision training + config.trainer.precision = 16 + if gpus > 1: + # Use DDP + config.trainer.accelerator = 'ddp' + # DDP optimizations + trainer_plugins = DDPPlugin(find_unused_parameters=False, gradient_as_bucket_view=True) + # Scale steps-based config + config.trainer.val_check_interval //= gpus + if config.trainer.get('max_steps', 0): + config.trainer.max_steps //= gpus + + # Special handling for PARseq + if config.model.get('perm_mirrored', False): + assert config.model.perm_num % 2 == 0, 'perm_num should be even if perm_mirrored = True' + + model: BaseSystem = hydra.utils.instantiate(config.model) + model.summarize(max_depth=1 if config.model.name.startswith('parseq') else 2) + + datamodule: SceneTextDataModule = hydra.utils.instantiate(config.data) + + checkpoint = ModelCheckpoint(monitor='val_accuracy', mode='max', save_top_k=3, save_last=True, + filename='{epoch}-{step}-{val_accuracy:.4f}-{val_NED:.4f}') + swa = StochasticWeightAveraging(swa_epoch_start=0.75) + cwd = Path.cwd() + trainer: Trainer = hydra.utils.instantiate(config.trainer, logger=TensorBoardLogger(str(cwd.parent), '', cwd.name), + plugins=trainer_plugins, weights_summary=None, + callbacks=[checkpoint, swa]) + trainer.fit(model, datamodule=datamodule) + + +if __name__ == '__main__': + main() diff --git a/tune.py b/tune.py new file mode 100755 index 00000000..222ba868 --- /dev/null +++ b/tune.py @@ -0,0 +1,194 @@ +#!/usr/bin/env python3 +# Scene Text Recognition Model Hub +# Copyright 2022 Darwin Bautista +# +# Licensed under the Apache License, Version 2.0 (the "License"); +# you may not use this file except in compliance with the License. +# You may obtain a copy of the License at +# +# https://www.apache.org/licenses/LICENSE-2.0 +# +# Unless required by applicable law or agreed to in writing, software +# distributed under the License is distributed on an "AS IS" BASIS, +# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. +# See the License for the specific language governing permissions and +# limitations under the License. + +import logging +import math +import os +import shutil +from pathlib import Path + +import hydra +import numpy as np +from omegaconf import DictConfig, open_dict +from pytorch_lightning import Trainer, LightningModule +from pytorch_lightning.loggers import TensorBoardLogger +from ray import tune +from ray.tune import CLIReporter +from ray.tune.integration.pytorch_lightning import TuneReportCheckpointCallback +from ray.tune.ray_trial_executor import RayTrialExecutor +from ray.tune.schedulers import MedianStoppingRule +from ray.tune.suggest.ax import AxSearch + +from strhub.data.module import SceneTextDataModule +from strhub.models.base import BaseSystem +from tests.test_tune import test_train + +log = logging.getLogger(__name__) + + +class MetricTracker(tune.Stopper): + """Tracks the trend of the metric. Stops downward/stagnant trials. Assumes metric is being maximized.""" + + def __init__(self, metric, max_t, patience: int = 3, window: int = 3) -> None: + super().__init__() + self.metric = metric + self.trial_history = {} + self.max_t = max_t + self.training_iteration = 0 + self.eps = 0.01 # sensitivity + self.patience = patience # number of consecutive downward/stagnant samples to trigger early stoppage. + self.kernel = self.gaussian_pdf(np.arange(window) - window // 2, sigma=0.6) + # Extra samples to keep in order to have better MAs + gradients for the middle p samples. + self.buffer = 2 * (len(self.kernel) // 2) + 2 + + @staticmethod + def gaussian_pdf(x, sigma=1.): + return np.exp(-(x / sigma)**2 / 2) / (sigma * np.sqrt(2 * np.pi)) + + @staticmethod + def moving_average(x, k): + return np.convolve(x, k, 'valid') / k.sum() + + def __call__(self, trial_id, result): + self.training_iteration = result['training_iteration'] + if np.isnan(result['loss']) or self.training_iteration >= self.max_t: + try: + del self.trial_history[trial_id] + except KeyError: + pass + return True + history = self.trial_history.get(trial_id, []) + # FIFO queue of metric values. + history = history[-(self.patience + self.buffer - 1):] + [result[self.metric]] + # Only start checking once we have enough data. At least one non-zero sample is required. + if len(history) == self.patience + self.buffer and sum(history) > 0: + smooth_grad = np.gradient(self.moving_average(history, self.kernel))[1:-1] # discard edge values. + # Check if trend is downward or stagnant + if (smooth_grad < self.eps).all(): + log.info(f'Stopping trial = {trial_id}, hist = {history}, grad = {smooth_grad}') + try: + del self.trial_history[trial_id] + except KeyError: + pass + return True + self.trial_history[trial_id] = history + return False + + def stop_all(self): + return False + + +class TuneReportCheckpointPruneCallback(TuneReportCheckpointCallback): + + def _handle(self, trainer: Trainer, pl_module: LightningModule): + self._checkpoint._handle(trainer, pl_module) + # Prune older checkpoints + for old in sorted(Path(tune.get_trial_dir()).glob('checkpoint_epoch=*-step=*'), key=os.path.getmtime)[:-1]: + log.info(f'Deleting old checkpoint: {old}') + shutil.rmtree(old) + self._report._handle(trainer, pl_module) + + +def train(hparams, config, checkpoint_dir=None): + with open_dict(config): + config.model.lr = hparams['lr'] + # config.model.weight_decay = hparams['wd'] + if checkpoint_dir is not None: + config.trainer.resume_from_checkpoint = os.path.join(checkpoint_dir, 'checkpoint') + + model: BaseSystem = hydra.utils.instantiate(config.model) + datamodule: SceneTextDataModule = hydra.utils.instantiate(config.data) + + tune_callback = TuneReportCheckpointPruneCallback({ + 'loss': 'val_loss', + 'NED': 'val_NED', + 'accuracy': 'val_accuracy' + }) + trainer: Trainer = hydra.utils.instantiate(config.trainer, progress_bar_refresh_rate=0, checkpoint_callback=False, + logger=TensorBoardLogger(save_dir=tune.get_trial_dir(), name='', + version='.'), + callbacks=[tune_callback]) + trainer.fit(model, datamodule=datamodule) + + +@hydra.main(config_path='configs', config_name='tune') +def main(config: DictConfig): + # Special handling for PARseq + if config.model.get('perm_mirrored', False): + assert config.model.perm_num % 2 == 0, 'perm_num should be even if perm_mirrored = True' + # Modify config + with open_dict(config): + # Use mixed-precision training + if config.trainer.get('gpus', 0): + config.trainer.precision = 16 + # We handle NaN here to terminate trials cleanly + config.trainer.terminate_on_nan = False + # Resolve absolute path to data.root_dir + config.data.root_dir = hydra.utils.to_absolute_path(config.data.root_dir) + + test = config.get('test', False) + tune_config = config.get('tune', {}) + lr = tune_config.get('lr', {}) + # wd = tune_config.get('wd', {}) + hparams = { + 'lr': tune.loguniform(lr.get('min', 1e-4), lr.get('max', 2e-3)), + # 'wd': tune.loguniform(wd.get('min', 1e-4), wd.get('max', 1e-1)), + } + + steps_per_epoch = int(3e6) if test else len(hydra.utils.instantiate(config.data).train_dataloader()) + val_steps = steps_per_epoch * config.trainer.max_epochs / config.trainer.val_check_interval + max_t = round(0.75 * val_steps) + warmup_t = round(config.model.warmup_pct * val_steps) + scheduler = MedianStoppingRule(time_attr='training_iteration', grace_period=warmup_t) + + # Always start by evenly diving the range in log scale. + lr = hparams['lr'] + start = np.log10(lr.lower) + stop = np.log10(lr.upper) + num = math.ceil(stop - start) + 1 + initial_points = [{'lr': np.clip(x, lr.lower, lr.upper).item()} for x in reversed(np.logspace(start, stop, num))] + search_alg = AxSearch(points_to_evaluate=initial_points) + + reporter = CLIReporter( + parameter_columns=['lr'], + metric_columns=['loss', 'accuracy', 'training_iteration']) + + train_fn = test_train if test else train + analysis = tune.run( + tune.with_parameters(train_fn, config=config), + name='trials', # required. otherwise some of Ray's checks will fail. + metric='NED', + mode='max', + stop=MetricTracker('NED', max_t), + config=hparams, + resources_per_trial={ + 'cpu': 1, + 'gpu': tune_config.get('gpus_per_trial', 1) + }, + num_samples=tune_config.get('num_samples', 10), + local_dir=str(Path.cwd()), + search_alg=search_alg, + scheduler=scheduler, + progress_reporter=reporter, + resume=tune_config.get('resume', False), + trial_executor=RayTrialExecutor(result_buffer_length=0) # disable result buffering + ) + + print('Best hyperparameters found were: ', analysis.best_config) + + +if __name__ == '__main__': + main()