From 22cb62db862536a67162c56b7374367c643d033e Mon Sep 17 00:00:00 2001 From: Travis CI Date: Thu, 15 Sep 2016 02:09:31 +0000 Subject: [PATCH] Deploy to GitHub Pages: 703cce35a871c78a5e3f1c33abf7fe788e2bbda5 --- doc/.buildinfo | 2 +- doc/_images/feature.jpg | Bin 0 -> 31204 bytes doc/_images/network_arch.png | Bin 0 -> 27822 bytes doc/_sources/algorithm/rnn/rnn.txt | 6 +- doc/_sources/build/build_from_source.txt | 5 +- doc/_sources/build/contribute_to_paddle.txt | 4 +- doc/_sources/build/docker_install.txt | 16 +- doc/_sources/build/index.txt | 4 + doc/_sources/cluster/index.txt | 1 + doc/_sources/demo/embedding_model/index.txt | 2 +- .../demo/imagenet_model/resnet_model.txt | 2 +- doc/_sources/demo/quick_start/index_en.txt | 8 +- doc/_sources/demo/rec/ml_regression.txt | 26 +- .../demo/semantic_role_labeling/index.txt | 7 + .../semantic_role_labeling.txt | 183 + doc/_sources/source/math/utils/utils.txt | 4 - .../ui/api/trainer_config_helpers/layers.txt | 24 +- doc/_sources/ui/data_provider/index.txt | 4 +- .../ui/data_provider/pydataprovider2.txt | 57 +- doc/_sources/ui/index.txt | 2 +- doc/_sources/ui/predict/swig_py_paddle_en.txt | 32 +- doc/_static/basic.css | 13 +- doc/_static/doctools.js | 26 +- doc/_static/searchtools.js | 4 +- doc/_static/websupport.js | 2 +- doc/algorithm/rnn/rnn.html | 23 +- doc/build/build_from_source.html | 26 +- doc/build/contribute_to_paddle.html | 25 +- doc/build/docker_install.html | 37 +- doc/build/index.html | 63 +- doc/build/ubuntu_install.html | 21 +- doc/cluster/index.html | 17 +- doc/cluster/opensource/cluster_train.html | 25 +- doc/demo/embedding_model/index.html | 55 +- .../image_classification.html | 35 +- doc/demo/image_classification/index.html | 21 +- doc/demo/imagenet_model/resnet_model.html | 99 +- doc/demo/index.html | 18 +- doc/demo/quick_start/index_en.html | 34 +- doc/demo/rec/ml_dataset.html | 31 +- doc/demo/rec/ml_regression.html | 103 +- .../demo/semantic_role_labeling}/index.html | 87 +- .../semantic_role_labeling.html | 317 ++ doc/demo/sentiment_analysis/index.html | 21 +- .../sentiment_analysis.html | 101 +- doc/demo/text_generation/index.html | 21 +- doc/demo/text_generation/text_generation.html | 103 +- doc/dev/new_layer/index.html | 17 +- doc/dev/new_layer/new_layer.html | 21 +- doc/genindex.html | 1063 ++++- doc/index.html | 17 +- doc/layer.html | 17 +- doc/objects.inv | Bin 144806 -> 155568 bytes doc/py-modindex.html | 23 +- doc/search.html | 10 +- doc/searchindex.js | 2 +- doc/source/api/api.html | 206 +- doc/source/cuda/cuda/cuda.html | 477 +- doc/source/cuda/cuda/index.html | 21 +- doc/source/cuda/matrix/index.html | 21 +- doc/source/cuda/matrix/matrix.html | 1387 +++++- doc/source/cuda/rnn/index.html | 21 +- doc/source/cuda/rnn/rnn.html | 975 +++- doc/source/cuda/utils/index.html | 21 +- doc/source/cuda/utils/utils.html | 202 +- doc/source/gserver/activations/index.html | 35 +- .../gserver/dataprovider/dataproviders.html | 101 +- doc/source/gserver/dataprovider/index.html | 21 +- doc/source/gserver/evaluators/evaluators.html | 173 +- doc/source/gserver/evaluators/index.html | 21 +- .../gradientmachines/gradientmachines.html | 203 +- .../gserver/gradientmachines/index.html | 21 +- doc/source/gserver/layers/index.html | 21 +- doc/source/gserver/layers/layer.html | 676 +-- doc/source/index.html | 17 +- doc/source/math/matrix/index.html | 21 +- doc/source/math/matrix/matrix.html | 3969 ++++++++--------- doc/source/math/utils/index.html | 22 +- doc/source/math/utils/utils.html | 220 +- doc/source/parameter/optimizer/index.html | 21 +- doc/source/parameter/optimizer/optimizer.html | 987 ++-- doc/source/parameter/parameter/index.html | 21 +- doc/source/parameter/parameter/parameter.html | 353 +- doc/source/parameter/update/index.html | 21 +- doc/source/parameter/update/update.html | 37 +- doc/source/pserver/client/client.html | 37 +- doc/source/pserver/client/index.html | 21 +- doc/source/pserver/network/index.html | 21 +- doc/source/pserver/network/network.html | 41 +- doc/source/pserver/server/index.html | 21 +- doc/source/pserver/server/server.html | 78 +- doc/source/trainer/trainer.html | 55 +- doc/source/utils/customStackTrace.html | 97 +- doc/source/utils/enum.html | 21 +- doc/source/utils/lock.html | 77 +- doc/source/utils/queue.html | 81 +- doc/source/utils/thread.html | 43 +- .../trainer_config_helpers/activations.html | 35 +- .../activations_index.html | 25 +- doc/ui/api/trainer_config_helpers/attrs.html | 25 +- .../trainer_config_helpers/data_sources.html | 55 +- .../trainer_config_helpers/evaluators.html | 35 +- .../evaluators_index.html | 25 +- doc/ui/api/trainer_config_helpers/index.html | 21 +- doc/ui/api/trainer_config_helpers/layers.html | 225 +- .../trainer_config_helpers/layers_index.html | 30 +- .../api/trainer_config_helpers/networks.html | 35 +- .../networks_index.html | 25 +- .../trainer_config_helpers/optimizers.html | 29 +- .../optimizers_index.html | 25 +- .../api/trainer_config_helpers/poolings.html | 29 +- .../poolings_index.html | 25 +- doc/ui/cmd_argument/argument_outline.html | 21 +- doc/ui/cmd_argument/detail_introduction.html | 21 +- doc/ui/cmd_argument/use_case.html | 131 +- doc/ui/data_provider/index.html | 25 +- doc/ui/data_provider/pydataprovider2.html | 121 +- doc/ui/index.html | 27 +- doc/ui/predict/swig_py_paddle_en.html | 249 +- doc_cn/.buildinfo | 2 +- doc_cn/_sources/algorithm/rnn/rnn.txt | 7 - doc_cn/_sources/build_and_install/index.txt | 21 +- .../build_and_install/install/index.txt | 15 - doc_cn/_sources/cluster/index.txt | 9 +- .../_sources/demo/embedding_model/index.txt | 1 - .../demo/image_classification/index.txt | 4 - doc_cn/_sources/demo/imagenet_model/index.txt | 2 - doc_cn/_sources/demo/index.txt | 16 +- doc_cn/_sources/demo/quick_start/index.txt | 6 +- .../demo/semantic_role_labeling/index.txt | 2 - .../demo/sentiment_analysis/index.txt | 2 - .../_sources/demo/text_generation/index.txt | 3 - doc_cn/_sources/dev/new_layer/index.txt | 4 - doc_cn/_sources/index.txt | 16 +- doc_cn/_sources/ui/data_provider/index.txt | 21 +- .../ui/data_provider/pydataprovider2.txt | 12 - doc_cn/_sources/ui/model.txt | 4 - doc_cn/_sources/ui/predict/swig_py_paddle.txt | 34 +- doc_cn/_static/basic.css | 13 +- doc_cn/_static/doctools.js | 26 +- doc_cn/_static/searchtools.js | 4 +- doc_cn/_static/websupport.js | 2 +- doc_cn/algorithm/rnn/rnn.html | 110 - doc_cn/build/docker/build_docker_image.html | 17 +- .../cmake/compile_options.html | 44 +- doc_cn/build_and_install/cmake/index.html | 42 +- .../build_and_install/cmake/install_deps.html | 44 +- .../cmake/make_and_install.html | 44 +- doc_cn/build_and_install/index.html | 45 +- .../install/docker_install.html | 64 +- .../install/ubuntu_install.html | 44 +- doc_cn/cluster/index.html | 45 +- doc_cn/demo/embedding_model/index.html | 93 - doc_cn/demo/image_classification/index.html | 94 - doc_cn/demo/imagenet_model/index.html | 94 - doc_cn/demo/index.html | 55 +- doc_cn/demo/quick_start/index.html | 57 +- doc_cn/demo/semantic_role_labeling/index.html | 94 - doc_cn/demo/sentiment_analysis/index.html | 94 - doc_cn/demo/text_generation/index.html | 94 - doc_cn/dev/new_layer/index.html | 116 - doc_cn/genindex.html | 17 +- doc_cn/index.html | 77 +- doc_cn/objects.inv | Bin 203 -> 851 bytes doc_cn/search.html | 10 +- doc_cn/searchindex.js | 2 +- doc_cn/ui/cmd/dump_config.html | 17 +- doc_cn/ui/cmd/index.html | 17 +- doc_cn/ui/cmd/make_diagram.html | 17 +- doc_cn/ui/cmd/merge_model.html | 17 +- doc_cn/ui/cmd/paddle_pserver.html | 17 +- doc_cn/ui/cmd/paddle_train.html | 17 +- doc_cn/ui/cmd/paddle_version.html | 39 +- doc_cn/ui/data_provider/index.html | 61 +- doc_cn/ui/data_provider/pydataprovider2.html | 72 +- .../data_provider/write_new_dataprovider.html | 44 +- doc_cn/ui/index.html | 39 +- doc_cn/ui/model.html | 116 - doc_cn/ui/predict/swig_py_paddle.html | 270 +- 179 files changed, 9780 insertions(+), 7596 deletions(-) create mode 100644 doc/_images/feature.jpg create mode 100644 doc/_images/network_arch.png create mode 100644 doc/_sources/demo/semantic_role_labeling/index.txt create mode 100644 doc/_sources/demo/semantic_role_labeling/semantic_role_labeling.txt rename {doc_cn/build_and_install/install => doc/demo/semantic_role_labeling}/index.html (50%) create mode 100644 doc/demo/semantic_role_labeling/semantic_role_labeling.html delete mode 100644 doc_cn/_sources/algorithm/rnn/rnn.txt delete mode 100644 doc_cn/_sources/build_and_install/install/index.txt delete mode 100644 doc_cn/_sources/demo/embedding_model/index.txt delete mode 100644 doc_cn/_sources/demo/image_classification/index.txt delete mode 100644 doc_cn/_sources/demo/imagenet_model/index.txt delete mode 100644 doc_cn/_sources/demo/semantic_role_labeling/index.txt delete mode 100644 doc_cn/_sources/demo/sentiment_analysis/index.txt delete mode 100644 doc_cn/_sources/demo/text_generation/index.txt delete mode 100644 doc_cn/_sources/dev/new_layer/index.txt delete mode 100644 doc_cn/_sources/ui/model.txt delete mode 100644 doc_cn/algorithm/rnn/rnn.html delete mode 100644 doc_cn/demo/embedding_model/index.html delete mode 100644 doc_cn/demo/image_classification/index.html delete mode 100644 doc_cn/demo/imagenet_model/index.html delete mode 100644 doc_cn/demo/semantic_role_labeling/index.html delete mode 100644 doc_cn/demo/sentiment_analysis/index.html delete mode 100644 doc_cn/demo/text_generation/index.html delete mode 100644 doc_cn/dev/new_layer/index.html delete mode 100644 doc_cn/ui/model.html diff --git a/doc/.buildinfo b/doc/.buildinfo index 6b7cc993be..2343be214c 100644 --- a/doc/.buildinfo +++ b/doc/.buildinfo @@ -1,4 +1,4 @@ # Sphinx build info version 1 # This file hashes the configuration used when building these files. When it is not found, a full rebuild will be done. -config: 5bb206a2182263ffcb7c4270c50bc7c9 +config: 68487318b6dbd705fd569c9b901c758d tags: 645f666f9bcd5a90fca523b33c5a78b7 diff --git a/doc/_images/feature.jpg b/doc/_images/feature.jpg new file mode 100644 index 0000000000000000000000000000000000000000..0e3310e4ace5613917e7779d3198ccbb3cdc5ada GIT binary patch literal 31204 zcmce-1yo(xvNpPL4^D7*cL*Nb-3jg%Ab})EaF^g7+=FXyf=dzz?(PH+Zu!^V^yxl* zZ=d_W`^Fn@cQShKIcrMQ{OYT!nYDh;{9XZJ$VNL6K$g)SUv&Hr6? zNTaF>4t6i>$#b9;2m9=67`Qc)o^DU7ZdqGZVnH7w1~fd%17CKN|0aKPV#>s#D!%yl z4)BamR-oS_ZGv8>r|AyiJ734YRnT$RdG+p6V@D;sxS{deJZbh>vOsEqgI?gj9ym$w-3z3m;OMr}ZJKQdKe z_upYTj0dDt{aDVw!JZ17Q?GU2j{W8JaIiKxd{JXL)wC-i2AE1EN5mJr8DLjjP___N zy8=np{C&D2_le+<( z4+{D-aIh!ho|3Ium3YMM_emNLEI3}9IKo|z_Jf#9HQ*gYE1@qfpq(!uY(Q~tz7^@u zYQ5(dmdGHaGr9yQ!M@gaMeCm8+XZNV>r{_#K6Ep~g=a59R9aWAOSpJL zq7G05G?dB*!Gngm+~cn5U~jO82n}j~6$_>vobJ`J?``WZ$~>K(5D7`!HDLkN&JuqQ z0tP@zMB4YR~V6d8e5{K#yT9HDGzpM-oI@DTViM{1p_ z@v|014;pfyAH2864WG&^*gDWdOu8x1s21IA$+dF6RlTR#P3Bf}TN=Nkq7U+4iXh^* zS!rq}1gw6BKr!7J=Y16QlZVr!C_`+o>9*xtmZ!_FH|(RGmg9e@^!2Z3_D~M{h1N9lDod9uShT1CjWQb|61~b21fIsPh$!^wInvcy9!^2s-{u$k^@ek> zh6Q3U3@6#p>x=hX5nG3N4Yn&ICX8TAi~rc+UhiY6PZak9pIPEw{@^!Py~MnD+CwL< zDaF#(zKNF7^k}@fPigi|U|c`zsJ|O*WdW}n+%VkkAmD9tgjsMV@M@(TD8IIY8>47E?#FOy7($R#ut8l=g=pR33*uyg}*uPIU z@GSwb0TTa3b7BLCBp|t{k;xAy(1UdY7XIV-kd4g?Up^Z1bq4W{U$xk98=2n0H^`}j zLf8uW3FmGz_j4hu5O2%yDk~JtcWN}$9BlnE6v;r_;y#Qgec8~y-JEM%u>l6xHs0M} z9YnPdp0PXx9(%KG8*cETf4UvW^v6jUYY;?7Sl#Ln1phcGPbe{nwknQib27HWC{iJG7L z?~GrA$3P#Ya-x8B$SG(E*KK}!SwU-B@8_-@V&5ZnJ^s#+xy7EF{5%yqsr7^hp>U)) z) zyao&-_RA2NZP2l)B7DTQgQjr1Rc{OqY(sQ>UUCvd5cunBH5Kz^p4U_GzF_-!o>zktIUY$3 zM)n00NMhjaK_e^72mBHJ#GO(aFps}b%U&q!ea+r?_Mk& z+@xms z614BXdk)uddwzyk9fXGfl_4uX@x;mCt@Gj;4+2{NJ&ySmmfIn1xht!^h~qZ)^v`Dt z+k+lTHdppVkG9*I_?J!<(=04(xYq_B>i*qV(17sl^=m%?`ztsgBV$}56TbU1I`<+S zaIU?(WDRG(n$>$QREOOaDBKljm{X9&LHC0u!`jfg$yI&zq^2e~lk-RBI4fP$to!3& z;D_m?G?^aWzoOV*K_4085PX3D^T!8BFE}tj&`>ZiFtAX-A2_f)3@kJV3I`5@2b-K7 z6APD143CPNf|`Sq5}sGu#2H9D5P@_A3J&@=Xo|RP?E`)(s#Y*#$``KH@Ty*2voa2C za0^YIHTz$`f*HGUP5@*Zs|@&^$kURJWyE#BC)AnE$W8|~mS6?-S!ji4nM-sj0>(r4T0*aBis zhiED>CG(50hT*xBY|LiW(7jl*(y~#Z($<*voKa~Lo=fW&6Xg>h@?I94qhBi}$T><)@XWWx26bd5LDX-9^EtiMFbkUUp; z*-~jO;Mye$e;m5o3%2#|%XyyPOW*SeYTPh%tlA!&ac2;FPnJARy*J*Gxb7JZUUO(P zIE;Q-=`f3;&O@cX1erd@BFuCNz+sG@zdVac*h#H+MnL(6O3b`6H53-B4XjF38UTa1 zgxZ&@{D%k5*<$?$X1*riNMUD5VWD2T0HOmJoJ4HqKGzE=SXW!{H znt~JkWzVwg)p69Qk)Y9)sHvx;4n^NK{s!S(qf;&TgnNao2=`xU+=ya=o9{n}&;8_g zXx}|0CHf5t)Js<@YJ6DA_sEFfVEheQ#k2ZxCoFr;=NUb-?>ZPuQ9qY8swnRa|FKhF zJxzr3E8(~DXGZK&>5Y}Kkl9wBG~_IbNLa4Dsl*yAoefSFi%1#^dpW*N0asjrT?eW0A zTtCKtt~_{3M9;-wwmTk|n;ujQj7(5+^FX+wIbdQ{@d`W1zdBLsnwDLkWUwU#pF#?3 zMQ=A(6WB1f740clNGOPA1r(z8U(=oQYzOJuFE0wD@tl`D}(h|S)ZoV48P8>|}FjpihrLhpfd&rrrt5{zM zJ9>(B(QK)Ow>j><4xd+yA(2-5Xkv+M8z18bnIF=w~x*do4PqZz8{JrD9VLlfDKZRpg{{ z6QKUNsbb10O#ya!-v6BTu+Z#c>gL&(*ZRbE(U!3(hAn7Wo_P)B#c(YXW=aHmbFtY zKw#J(6lnG<2>uPS^O3ScTP=jmah|DB(6~c@;MFn=t-#=XvZZLiu~2`}%>sWH_4cN$ zGqxsmrE3x3XI`!ig2qI%A_7IBZHsHPx(I{3`{&R_1WIf!VYq#tjFRrzJ5JJwFf9%w zKjHEZnQnu9Hul2)Ceq`>o1!!AMh0wU)P2>am+hag(7|!)2yPYL6HpL#zcsHgnDNf} z2-Rl+BNy`m2w~(NX}iH<&{S|6DI3$kYS|y9-02C*V5L!#RWR}UKZwAx%RGtMK-Vy2 zA9SWv^LqL$JOHoU}z~9*kHVz-75fP=Ri^1&GSj!9PAKz*O<9P=4t$jb-G(qxD0= z6hlRdRmwr_ad#m+g63h6#vC>ZR7=Bw_|nMpl3BW!>HxdXp(}pY3T^ycvsY3WQ=rfU z_)*r63+o?GU#^)UnCX88REpt{h7zGHeRD`l?Y^}3hF19|p_TN_;|BU*K~Zd9 zK^yQTv#%u;qO1{%0NpGf*KKRVC+3Bl>!&5&=v$tAJ1?g`Ygx47$yze>Sy#D-IU1v5 znp5iFN6~KyPg7iG(obB+FN(4O%mTQZoS)=$h>jM3_J)kRaE~KBI z^QS}!_wT=Rt+!gAo&d+7kyTrr=pI|r&-BT5u&oy3PETxVHu8c#i9hGf7fA}S!3cW- zhat{SrX(w)q4qk%Y?WBIE_s64RMvy0qZXhtfHuh`_3O)t-U23M3dUg%6%;ck1G1kn z#6eGK-bh%E$)0B_fE$j1EE#=+1yX|L6R7yt1B&qbFk_?yZyjyENRYgcNZ4bKWKjVK zW-2ic@d;o$T#hf*o$8bJ`n<)k;<>QWrDaIsKZ0__LMNfVxO^m+{o;)Cp5-SG9(9%G zJg(215l9;4IX6YH_VW6ai1o+p3B*OUeuD~7;v3qiVgECmhEi$cIQQN6bu{%X2fh07dhK032(^j zuX_7{@Kft-7G+M24FN}`c_aI_HyDq?fivwEsWEs!A|TPUSq5Z6bsf`EYquIVOfQ&2 zis11uUUWzp{NccMPqDWz8kfNVhx#c+GEXjD-}e_YgsJ$|xR?EBU7za`061Bb*TN>~ zcm5H_ft4FR%0g4W_<@*3#0~EdK*QD~>gA)z?artI10C_W50#epBC z!dv0~i*u?8U}2r#AjDeUkkvhz5+DqXC$7WlwS-?3ooR59Piqd>q_OIK`H^nENkrY% zD!erE7%|FkD*$B7SXawY$$n?Td!I(la&ixx*8lrp-f^?jY7A(@cgG^iwdNWI-x*+! zrQgbic)LGNsslk}#-?c@k)P4-DG_@Uk+ztvHQZ#PZLUD)A;52$x5?{tfeelMyTl$v zfKw+BBviZZ4yq}2O-Rl;%F{Nt!Q5e8Bqr0fOJ0q6y(?U`&v)5NsEeimTyp%pj=LJy z6&xj7l;!i{YUpXj;>c0i_;mxuHJa_OA%Obl@Gu*&#H<^L-8K%k_v#9 z2Why~>Q!2RdFS3H<`h=*g88fv9Zm-=)kNIR`85#vAt3b~vDngD1VFms89lqZfU211 zpAu3lj#vQB0cYW02Vgq`i6v%pZZjIcIPWH-T4uxR#E{SmN-|4lQtbcmXR!AKNZRLo z^j$j;Vy6#$qC<*mRUev1fXog6X)Vf<)sA8raO5)_!eOhxZq#_9SvbcuC0=Q{MUlt= z4*ZqXjlS72G&^{s6f=g$Ii3jhH|PoC5GX;#+*q-G3}YeD&|F5XAotye%=L?`miZ!9 zEgUd6hr?q84s|1EMgrl|`%wQ>hFvR#@f)rcaxI*me_5xtNh!D?o}(r8uU~)f)|}^{ zbpNML&D5xczl;9U)Q%~!LjPLkFKP8}J!XIX`ePUWv7vA_4Er}|iwvlr=`le-f5WixQYFA8%$vqdbevap-b)K>2C*{m@BxD{{5hR|3;dhtm`asd>Z~arHGhLT0)2Ss*MLqm z)ing?^~8R}yl8a{HrG4Z-yoZwaJ4cQZGC*KA+7Dh)J0p*L{>btv3xD7;LI zvTW-BOv4PA1{MSdg#Zf!zE=P~VADK@#-LyaZWF{*)pACU#a;Y5O-y4U=79#A=Xg~C z&D+PuK2^8Q%`Te(`Ia*J*?sGMO}DYo21Rs!-c&IYnfH;>3iJkl@`#ohp1&|_S!9`* zCdXNV#|Mc(ezIspnpm`E%4b?J*AF>27bDxBs!^VkCR#IL;G!DRm~EmVKV7A+a+P@b z!AN4(>pXfy%efAdB*FdJ3pfe995aDtLX4m*IqmseiCLr<;#4&Qr>`23{j_g$Wt7#> z8h>H*=nB-vu6EIO_2=b=AZIy;poY1>kPd5MF8f#k4a;xol*s8wU^Y3r=cd?Eg48~_ zaU<0GrS#<%x1`O`U{nvKAViS&HZN7=L#Siv;F)_wFe~Op1-g$TfqO0$@*Q7!W2FECx0O2j}bf>QVCN zYLry$rY<<*>VB~`Tw)Tha?Cp6xHZgIsmDy5UHt=&PyV>SffohtZ&o975@IAB%|sI0 zM1^ms&h9Rc490&gx5;4MDs9Z#W-i)$cza%!?_M7*eJA_Sv#y8+kdh=7I~~JFRxXJy@6t?ne^}kHn#QGinFJYplZ(0ccD6O4KO}Q^1Ai9F z>3)=VzOKN2u%`_k3|4hIt$&}g(wWy^Ya9Yz`^4fWRxHiFg8Ig_yvz|o)!}YdrfmT(5AQ9QI2|4=x+J3 zE)+O4gnXde5tIFU+YI>~k==%c!FXl`YxU`7}aOr(UEQ{yh^N4UuD@K9v@Dr9Jm+AZqcC4Zv6_(@v@+MW_28D;hX4=v1FazmkB3ewb^RZ*>PU-uYZ2^APxDT%?+ z;aZ&5%rFmqKgFr)>&9vBVAJ=x@`^6D4~^g4m3kOWD8EZGKps>eLL;&-rionpfbl-3 zW_!gFSUnRT;oRcs4Vqj0#`@7<K}Dc~=9-jKs@V)l$E0 zYW#JDah&z(4$(?iw=euT9~lCGHWdt=wu&c}FfJDN+t)7uAP5ntAT?|MNoZ#7v>^A? zGFu z!l?h#ivNXPFnalpmv2V)It0LUyLSa9UaP2@2qx2MuK&oz|FDC9X!?guopEqqnp;A@ zZTbxY>J56C`u*{Sk^e}=PIgXfqUTvBWEo(E&c=N9bm^PA1$_>4I6ISVMlw@=gAQ}3 z)xUl1|ATn%fRLiV?DS0IH|YKC@X?s=2YZ=Ypku=P>94Xp3c-$^=-p#5tcIHe61ADa4|I#Vqwz2|(4>guP)iu< z*Kdo?OQvNTG>LS+o*#FEdCsS{%%{)shErfAMkKOC)FU@4KUgzi?&N;_W+FWPIe~=J zyEnfd3x^!hNE@Zlb+D3vf$^RZ`NW&{PYR0bf%=sXidZ;5f$Um(j7Vyq72)%=<@3b6Y72Xe zL#6)h=Rf?5Mj$?%^*q6o8ddhPo;jI|IMYGiYtaUggUfwUw)?!Kdcl?o-_7w+d)GvN zk<_SkEne+Vm-l%}0V8No@U%{XCHUU;LIlb*+A*z8_5mKL17A{EW<)LcDhCSIiv9V>-;H2=Ubv-;H>8}d`bym3sLeQ!Rc86HF|^5c8LaeCpqR~=1e z(Whlu2=&Aw^X#)eY}jCL<(^4FH#sQQ+Kyx=YOJ5q;l-0Ja55n=&Oo@8R-Za`f64fd zT8c}VXkJelTEU<0O`mZ7jY6)P09$+Uzf+m?i5+^UG` z{TUKiPrCe12WuXT1$1e;4W?yJQUa?H`!ZmL7alVrM4mmOj} z(?p({)sv-txUb&=>n!B)>SugxYtfO{9tUx-1ap#f+qt(JwgmAU=lhj7F(2Xjgrk=X z6WiTjFLJ-Xi;dzkfU3uTv6f*qYDW4uLB8r|T3$?f&Ml$>@dr=kv@aa>cVBhX$rWh! z7VqVoZ)J~FsOphhUg%=Mvt>vZZ%M}N_I$rHRpA_@$dtGetfa1k$5Pw2+_4qF~RQ>HnpZBnz0 zHE)*IDZ~JjJ)X-u3^ggqgagb!x6ccQ79E^dB!3mjnb9jG%`wWJdC2NjGePsbw{v0i z`yjsdJsq6VY$^)K>UkUZ2v|g}xUeRXdmjz?e?{)TN)bLfD@vt8jU#{BTl;87)h=V4 zpa2G(8C}#Oxq^Wp#mhaQWT^rfw*78srWmOT2WC9dW z(35YcqKHrXO%r62p?e~zqY9Yazxlq`NG*?(k zhN}W6!j&ZdDm{Yk=-Efgr)K%O>YYOTvTa9ZXm}rRJ8r~q4uo_qteV=QEUu-Mp3o{O zm7#>U!}xS7;8Z7?sH8xg=N;IguAIS9792}OAnq{$PD+l!6N?UkzpjA6C4xyI`HQKc zFhKUS3kSnk4`Cca#$GIpIf6Xnsl?S->`i4%l<_-gs2`gHf;Zx8+BU+I@1!SPR>8wX z8}>RhbYlb9in+yQBc*?!@v1~USF6hsRY3)3ftwDwj`5uSTNO_}QBX@$*}wyXvlbk6 z+CLc#WlkN{f98XXPxK?r4VA+Tor5YIJSk@6jo*IUoX|5$o@&F7(egnY*f?|b=`)XJ z$)EZXBrsX7HgI@&;F)ChnmyvGe7|TaPNJ>4z0Mf}y#%L5e&%C*<*&bNF=9WztAL-l z#L&rW?nwqVN7J13)EGzaTuo5ynajGf>}loI`()%@X133k=nq-7^^g*nw>nBYU;x~jJJu4O&b|kLVU2wSm+u%lf$l(b&yy2!s9>8%NTkcuV6LtU2 zc0m|bRjW&)Q1`P;)GAXny*$X1U5>JcW#8c4y^;co^l1LW9DKtZ581@|U&a}(#8b;_ zm}GLcm*2*on?8fYri?Re)2Q!5xNYr;tcv3tI{cN6!8!ej*wjkSGoo|x^m5_+FX(35 zdbwvsR7amK<2CX)NQKSyk3fv{o>3NVs#ks~KWw;(h`_Ye5QX_UV`!)l}%JO7vf~UK~w-Vin_ow_#4Co{th9?%_vI`i33xWua7?US$1pA+%bPD z@df_HTEPqb`8O*OQ`nw!7ofB!!FyhM;0yzlVLgB{6~kA(=cXNvn>CscF}`tSG{n0i z=*14W^H#l5VcgPn!_RWY=9G)Ia#AMIJcg&0Hdz@7>E4NjW3169MhU4>>89V!J#U~q z#)2<|l|PNH)!uN%dU$1_%t|jHI6<~{Y?1Yd2O?KUqgYQo_3 zv@Dl?FA`*<9#6-v;W6Nf5iaG2sn{5{=o(18qdlVp+P%a*oM#?{BV z`b0FoN4o~UZ;_U5{LyvXrv0kb1@=un2C0m9Zyiq`Wd4luNY=1xanpCLU_RYWD61i% z=y7uFD3FU!dK>k9m@$%omX^!S>}IF=D`gzBqtj&>BF|&NQ-CE(`{sZ=BNT7K$Dn|*=$H-!A&2V^6?>T7vWan*LsWPGLhM=OO{b-N^ zy`W0Q`*=Cp4^yWd@|sQjQWZ8sH3Gmwb~Xerqss2Sqq{S$E7a{r_YXDycDQfiq`M;OramBX ze7?v6vt&E;HwXAREXkReYe;zxi6l)GHg$^lFpkDV;Fx;#aYju`9G*vCVnmhsVX$T<*sYj|3`BsNBi2XfSAL$&9!Q#5$v zYdEULPgtzxN{LH?s!hAl4(<4hqMyX8v)fNp>mlbx>2{o~h6LX02;DG$9CH1>Z8r{j zv)Z{GF_stp?s%7n3cig$nEi@}sJ?g;G1B?Km_s2EH8oPZh1h_eea2pb(0|p1C|x^F zvdbyt$>Q638Jv^VF*|8Ak0d4vO|{ae50DDi+MlY0*zz|>U?88=(f1x5l7*)E?DJqG zDd^hc6)`hcY$m>kcO*QLbq>!WciR0ZcVAB~2s{$Rx4@rwqq%+91p-Po(FcGQBW~gN z#&S~(Yvg?UAL8`{VVt$feB=i*W*B|l(sc18V+I^8?ILKLR;3$fJknb1^xX#YrC;c_ z#n3|mjLjXQdLjQ) z{#rcbR_MYLO_2eW0j6mYIrlul+=PvA?%|Ihc%e9Pt&OVsu}Cgt^1bb!GM$NK`brd* zoKR(iSz_VUs{#|cJU&;S*!FP`H#jUgqLkYciN~`YQ6EmbHK4X-O=U@jceI|C zo0|TnPuBz0-~hfZ>P#S+(jX9KqKWiT+$Gw4r1h*@G>PyT$BOeLn~u$s+m1U`>#vFt zSFR>TDf!>=LuQ-joigVFfHwk?_(@kB?n=^b>g|g;caru)P|#KF#OC{Y+dmgeYS?_h zJb*(O%2Em`@i=%8Hc1LDgpn953HZMt{@B z0FfrRx_nwbj!=CSuIoVU04eaYuqA`qpbv}y@>@% zk?GZT7>@`!uEV=co#g+JdnU!qi^XeIL$7v90H)j#D^RqAwiL=4Dec^xnxx?@I%O*HJ^d5emDjrN=DaMhx^2#=*WlU3@oGS{c10nA2 zl|`j0|G8nsJzO!(##QPwA|JwIbgdGr{tY6jZ{w2Bp^+I)Y%iTqtBVE+w&+kw~m5;UkG@& zz*4(AVvfk3cvWJpdx!VmcbMTCVfnn?tHT&vSf+lLIT!X)>W#nrIAy3ik4DMbi&yqR z>Y=D=S#;wSU|h)ENhWH=v>38*Yr9+Aj;v)=buDWr#MUC(wd_%mJ-%}ljEeOb@kLm$ zzZmKNyf{cRnIkBZn7iQNic&M-wVL!;BI|CL3k~owc|5l4A>lZ9YR|bOh13r!yfh*& z>tOLyTMzS1Zzg6Axnm;*9#gjw0r^fc#lG*K(wQXr+eZsKw}X8Hu=z43B1QFq>sG8% zStGvs&Njsj6GhsYdhO!}l7l_~3v!qDFLs zQ3rk!Wjtev)Ec=u69cT6qPgE7ZX+T7y8yI0ju?aC?!U6h&IF|m;kgF`FGzwBhg9&c zCVMWJ66&m80YgrYs_N1)#Q3L1e*%1`&l?5%&zrIjljryAj_sL3U~O|6x1PlAnL_u}n6`{$%($BdWNTlV|jlf1w&S z$LSMB<_6&_BIB`~?s8n$qWkv0;!w5Pn}!_OsJ8m{c1R4`DF|4Knq}qj+8K#2PPI@a zQ7|wr$4h`~?}5K4hmG~c$+E{=CI!3BP>O3z&0WKT>mb3GHA7~xf2zK%gTePTr#XS` z3@GsG3{_PG%JB+5Oh5EYb$zfA+8z=af@#u_3~;M;OG;c&2D)98hi-WTcb!e^d!-4S zW3n-oqv4bA$k;B+33&pws&o}R$DenfH& zaDSAcm%UwQ8$=NA;^JTwm&ARf#BNjLFZrjYU;}1#46|VT)tnHY(@T0!d^(?3b_an zk_NrpWA!$I#Adq^#eA$9+zAm(U!vxStYk=CunmH}+qI0F2@@28(9d_S!*V1e4LlZ8 z3^Zed=~>PQGs>OyBOge?0~v9$0X#--?n-rd_jCVRx0fl}%&A9M2n=P&l3e*Ul9q#2 zGc(3W=fyhIChE2hKM5;DSZLIaKtjoDwR@FEUSV(evUb)m zYc~LW^x`24a#8YFU~BY=#fcgQYQ9zqv^fgWTz*U~T`S50Z>!)tkHJxOSdX0(w*im0 zC>jsJmfN?Ym-{J095dBJ3ZI=rF0UeMqrubUN4PQCEE;}YNY{^icR=x;HvV4oPr)4_PvitiTQd<3n<&Fp&T`Pow+R>%7&h^Axj8q}VW)xR7W@SbcDrMT7JzY96||FTIVmwGde6QR*ZrRX<^ z8T?=Y69(`?2MjbEEbxHJ>#{=V}%sb*>9R-EL^n{&I#Qte_RmKtg@YuW3?YAzR-(JdbI(hwwj8ime^_gTIpc?lyILuLp*JvxJj|mZq zvpS1;Age7BrmVUbL@!{L_Kl3v;LZGQ(nJ(uRYwZvqL8PWIfEKZGjLH98E;o?^BXh- zy}u@j(qlaiOi$omLVn=X?EgHunzp1EPBqQ=A}F5j^*bb%{K@-I3A3L)$l1DHrB@#7 zOg*$AX(QR;ZB(+Kr>nFnB#?Zwayn75bqqev^;>yjR5mQ!_^iEC37C!YY-DEey*ugK zx7eRmkl`X;ZZ!`n;Fl?7%}tfjL~7HH996Jo%w+}hJ3aMKduy~oC7I+K{B?T8%HPbl zj9;$QvNhAu%16BNcs+glt%uOzAzgUbjQV6!tg|?apxTEkbT@IQ?xx(f87_K>RcbYR z4cUao@v#Iyfn!s=DB+z1PPBv2G=(6i?-$_$IHRs^E{^P2=P~kojrMtgc z8Y?j@-qaIdA*M=V6`rV~mP-bPl8@T?u=I;Xu3t;K_@bQZapxsSzLLpdSqMWaxnj21 ziItU9A9fgDm8ZV!yvcT-?7k1TR#aTs*Mu6=&e-9n(%14O0c(we7W9kkllTCu6b4n1HLh z^a@F=>#g9ZyU5$OC8xoR-?*+hybh0KG8X&FHJ?REk)~xAvG-Sk;cJ(+V4y$OwqQ5; zLKV5uIen1l5ub&6IP#pMEy-XRc>w7p<-PUgrHfskf~+m`Cn;o_h9BW?H_4=3uE88m zLtq0;u2-<2x~~N)%7Vv z{zZaG(<|1-I`vOKDM^yTimZEi=DxTf(`lsgmVwDl?`&lEwjfkx(PHCzaY5H1``Z#Z zb#c4%ux0v6#JCN6+ZpZdO&LYLCU(GEBXQmR*v-%Fh}lYJHg6r3vb!#Sq!;Zq5B4_C zVYrntSe`|C(Pi| zfw!UFMq{cZKw2oCf5F$>f-n3@>0G`=J}_05NZVwKz#tXkYexeL+i$(ij0IbMtz9oJ z0UHku!rF;0Gm3w8&EWZftS3J>W-CQaA|!Npq#Yf3i>|LH553ESgw`fhPDXG+3Ji#d zqM3o$!V+d@*Pu_1#UaOb#uuQG+S+k{8l~fFIK1+sjFXtxCa&zcPa0QDj1|-eCkVD1 zJGq`N71W0;ygVoxp<#N>F;#snNaVw~nHfu%G-*8-PQe(CVo9nWe8qxhVB#so;g76g zWYH#nASb}U`hI8Zi=N49VFHwFTFSTcd|6Yb7tVl}r)3ZNDyVsBd#8L%tsQNnVTJnU z&@TMx8o&DK_00wi#bKJVcFY+n5<&W$SzIz`i`STR(gW2dO2Mx7Kw0*M6gl&p{7Zbe zpoQtBWKQ#i;(hJ?r?F!>y6jkqUv0zXU%;fydU?Y1@6KN%YrSV00Ym3#_h9Ut)}DKAkZ+G zV~?dQ=_Xy+6J~33rG+di$C&&E)x%=m8x@2QRJ1Ts0f<`UPzy0A6#Xxyx^Kzco zI$L8eAbyti9e_i}`h2uFtFQI!;Wqj6V*Y3nGaNh-{S;i@;u_zihDbIEU$~PdEsfeC zl1~T+VCzmoSAU$`*k@+`QvJ(#!Rd%Yc2-OpV^6#Cf&O|BoxUx4qpezG;5ZP0l8eE;H{R}#{CuY#6>%?$=ch%PN!Jc8&7djo z_oZ`I+_bR!-e=ZCL9?S6BK@vqUXdhLLyuWCEGI9;&7C>TnSq%L&A&6KL=*80rO;Rf zAmoE2-Z85P3aE{;>>~~5z`|9>jqi-=tXj4?nn;boui6=3Vk9iE)G45Ln^c+FQL=O4 z7lbAyWz-gI^zLTL*W9X!dP$^a&xCd0syVurG>U6DH{4hQP%>`E;9lN3HjaHd&vd`iAuVbKjJF;oWU)j<}h*>RMo4v_JPkSTN+Ny zad0zFX>+$R{p{rLyhM8@piW21+v}Lb z13&yRup~)=oYge`T+cA@=3#G0lg-?6Zd{wGJOqvYWmOuZWG620L{7y7+CF&`34Ly1 zPC4IT8}+m{ti#~n=E1}=XBPnbZu$!>T0WAY7-B=H*S1ar;k6TAiWW-RTf=?{)^nxU zsN3rE@Zo064TkaGGm_&-NB8^&+2TVdKec}QsNR34vJa%+`QGGL%n<5Vy;9GQRqBCb zBEPgTUyU0T(|qCororFya(qCi57ax_u9G`^X!Fc`x$S z)n624A6x-Z9f0UMLp3Lg=k9&J9~HFKw`KE$q)z6-N^e1m?9y+PrPv$ZN;Mhm^RNLcb#PVzPvP^yJkkO;rmXFIyCcTgIWA9=B`@< zEhB&R^@WdW7f>;S945sXY1QQ~M*P%PdIWj|xJ8j26vMZ|9*2*M3buq7`M)J7A7ZQ8 zkkXW=C9=OQD3-^{JeMzwM(_O9(gtXi^FtB0dUQ~sfY42xyMH-FnArhk$H zCtKDWpPLY*)4X{O(3bL;P8eIR;KUtN;w6JLsidHM6mB%4$0^o;{i1l&Kv6w6*5&)* zds|4Co^=-21SJ|=;xo%s`-OF0-h4Y}jFCxkbmTD-%%#FS0)srRr3BLriMrxZ-Q$o- zSXf+_WPq{y?xHLO6LN>E3`{|C4jJ|LjsBgq+(3SpS;x$Y^L0md^c_@UY zuL=kV3_o&5LZ1QJ@X&LDEbSbaTbh>MzQxap6Ko+Gdp=Nwwz?tp=8L?pm+G^r>2oGh zaye2u84P>5fm3oXTxdPP#JGwbgIjgY3goX_^+<{x;(nM1G$oX+Vb%-uW6s6<5uZ3a z<>1(~ljn0|aGs&*T-lVC(kCec3xXNxbmHPd(;l|Yi8#NO2S;KXcuh{#qS?-(ppFu; zVfjzDH)n-`q_M|x!16mk)x9=e%4k!?0a1?rph)=Dgq^`o zEi#Gv4Bi_DnMEqguXt~mze+eKJwr{97866GU@l%_YQ}jQ3Uh`$XVG z=S>f3sd{D5&7D76{A`X%=IUf=B}<_)1`>985nDC)2%n)Ak~%2!tnyI5c`08o@DAr! zi_V-jf${>EObN(N-C3;_POy#vL;QP-`qI2Yfv9FV=-b-I1X*Z>e<@80FX5Wo^gQbL zK2zNoq2Qh24moSbr>N0ZK4I^-Z!x8zr-QtKcYHT3NRov26j#iOykm9BG-HJVYeB3x zMm1%{a=}CK{P-oy`D%G2^3^CB3~T<9hE$i`r!DIE-nE(=6ZHld(^wM7k3w2Maj8|H znAeqt0izQu&)23|UuE!Su~D+bxkuqS9`@;TO1|Yz{|;Ily6!4xqi1;gWtYqquStA) zMl0@{Ok;6}tx(oQ>Kn@zCY53BRc@mK+^!fL*W4K>;SzNE`P_-fH;_ZD7U`Xl@+sey>`SU+*v-fuUiif>0j&76XHIX zO)-zf6-)Flgin$5#8Ii2h>=CjELwR?4I*a^WdoEap}F!8*d>M(JoDqK=a5qdiQg}l z)ZrW5V7uO{Vdq9D_)=%3iAlJMgf8U{DWDSeP4JJR`^W%V1`nZrUxSv9*hbybYe#J3 z4M15If0^~Ei%{x0KcCwM&4)o!1J9Lk*eTkOHxqNHZ10hpy-UkJ8kFZC=+~qX&h8c0 z#C5}0*FWKwSc)xq>5vO8&l!Nk$%}R@wfn~gv@ymQkvdx{!EP#oeCpYRsU;9rIQLNe z(m&i&idjI&2jS5yRy@?SPXuWr)9Ugmdk#%)n994PU%vSxw32c5*4WRaK=DW zgtKA!a`wd+xHle1&|8pZ{{K`m7^@}XY>KPDG#rQ!jSQ#>zzAPBTb_&kD#3mfyd3Z! zN`i>oF>Pf(ajh?MP|Ky?akMBu2ox+vW+>oMN`8IuswZ2L@K z;s-mmmD67-`wmAg56h>K`ezg_IKH*E42>puPhT5*ycmprl~-+0Wh5gXWb#E`hoz8# z!TE`~<;j80KyPTr$TT<#pEC^}Zb+Saavv#O79{0cFa(~@%YQK-Hnz@v!sM-Dtq46qxK zqb-a@YR!QnWf}wT5Ap!j3sfbulL5u$n1o-LuR`3VlSyR?KG*5{Bl4h|^i1)Hm09z+ z3!wGPDu( z(GMOI@(royD`FOVV_3*?6~Oa+pN8p5Ddqx`S74+*RtacEhPaP96&+>#2P}l0d_(eO zcgt7Di>aDn@#z7|M~^an;wz7_A$0^Pdb8xJ=R?&tuUElL*Ioz7I!FbdQQ)Io_BrUkMHNV)~wujX0E+w-;>N< zGuLGA!1o7hE`1bHvWor0fbzuRDe&g1N5pa^F?QzrqhUow^EY|!_EiXza=L;KqPqwq zJx2tcQhC9Y`9@pCxbu=&#-jvgpGWvBh9TD_3f7afNu?*E5{%}q3-J^kxF_V(L4k?( zOAm0@PhAEl9_ZK%*$z7y=8-|iQevWyfIoBCa?SK@y$HxSj^@*}`X-}V4wsr?GJFgp z^BH=dTzljlgU4c7Wc0-(1xwvL70t@JY1tF;4N?-Y@267dk+#XzBkFyFob>9)Oux(W zVba&hHlX+g=jH)}UHz4?5AZt{4*zTncn14Zj#mD2e|ISOr@sG*+arL@^P%JH>kzXEGT%W5yut1f z;M+v&IQh5e=x@0MxHR-{A>VZv5DEhx1@Qi(0}i0Of`fYr&R{|%8MJaRpa=}O-nP5) zw}wAf{4M^cXE^>5;Qx;bnC}^kkXNqpYz*NHCWUq=H{dBKY%{C1~>+8&E^@59V&VUH-a7pEO7X%5?&AW zJ%ho=0DhAVe{EUG@qhtcV8CC$pYZ>wz{!4A0yx=^fP(&;>;D~-36CNq0MO8K=N(IG zm^lT)iGHqxANVKg|3<9h|8lwse9t12xQ)>l=J4E3=kxXD5_i%0&p7+nuf}HaQum6G z9|3^E)x71buyGG7ZHH+1;PVz1gjXQcE;AOKIEX9IU(hqwTA8sl;f5|@|G4ihLQ7b_uUm>Sc8bn^&uZD)E^b2gq*^hG zgU7P{jBLu#r8r*>s`PBT#jA= zVIFJmuMN9U;zBJ9X0Z@n5QlusEQk>xI$4bpq_`JPN-Yr_1L_QLw|74Hq^d84{A#Eg z#abUa=W$Q70IFPgxgPBeA_`nvb4X*|)g|d5AC2fwt@)&nTxKpd4!~GHtm~$#{;D+D z0?HjdCq+TsG$xx5SMzw7_ymKVR1#S^ZRINK%r(|>>|+sBMXj~Y>5;mHe)v&6ds@qg z8OU2l;dZoJyy`D9B%2EF#fl|t`GA$J1Z^kct}Cb~<*+N?E9qm)QX)SWLEZ6$1sRVA zpYNtF!w;V7sh)PUp+kX=2Z=L|8P?~1^E{?%+-u@?Tv)O+6Y`srcnwsp%R<-w{A+H} z`tv{w#uH7MMRiiQu$}FvV&(D~izJMWMF<{xdO6fbwg<>{hPc%1n-n4`hCOTUeiL-$ zKDC8ib|1WM;~JXiV(;qT_Q^ZGV_sqy>3Y9alU77D&!dd`<76-Mb}7j(yNuWaZX)s- z6@9ZiF(2zu!sMpLvzfMeF7kyjh=OJNYv!gGSZzG!(0YMEwepWl%>xJ#iw0*4DJV-% z>&#@1fogUKGB|YXx3j|A(y?3JR!WU&v;Ls`>?Yc=vy{4uNXLxt8qHr+7CEjg8HC0Z zwz9WL#%$=FzhI@x$4$D?eHdY~V+coeR^yHGU%acyAlp`vz=5t$Gpvsk%Kp`gkt4os z+fax9vlMmHxiH036MRw0A|b+t*la88&*D>47lIEs;(0LRCws4Q8+y($8&^JC2eYI0W9h39cJk{71)A15o)}VvHCn_3_>uMQHF+=3!f@%m*PCsBN!+#HYB1D z_~`}N*V=_kJE zIfa>)3Dvgoeu~?1fPlYuwRamEhl&M1c~5O2BR&!)*&7H}T%D>*D(*slK%fW`j};8dAw{ewpBA zIiVf(6OwwDlMl%G;UE zClk8<-D80lG}7qtolA8z`RooLkK{zPXvK7ksITTkg?*}6~8S$-Ch`gXe@_l{!x{Qu0vW*yo z(A`dz0jqlx|B2T0tw>KiTS+P8xfuEs6`c%Dq(6W{kd~{8g~?}l`zPE4G~G%J#B%RQ zLKQRbk9-qg^5->EYjy2nSDY!%lKJxZiYgK`Au;$SPKG_{0E2VLmgJ{Bv9)P*1Oc_@ zKXgMpB8X_PjFXsDCz=c+ky-SS<=ZRL*cDshDFq>H*rx{_1Cc|3=o&eg>xZ73@lv#+ ziKx=ZUcT_*xo=v`h{>}_ec=vjtzqQ_4L^R6p{{@K}J7Z zt^fquNBoa~x!=tgPf*S}&LL~(gk#4Vc57Z40S#j~QjdV8?I*UT={TJeNJyJyjz!Mi zV;rQkd(jJ(?M3)+eMRNIW5lC$AK(KrOOjX<+yHuhRSxhGTu!2ZqLtnSz6`eyu?^hx zWleAu5}RoWe2H?TNaV?3MNr??`Gt-A^vLB#LJ9F^*;OX0*9H$Cz48{|NE`P_lH>T( z9R#tW@X5J`>CF@yqU>rjP$p3S959D&@&5H+&0zSmp$r6XtCz~37^u^R;?#Gi$ejA5 z6U%%F=0Q5L2_wPR^Q>HWI6-`aCyWV5IT;RoNg}sHB|{9vW1%Hvr(8oYTWhdwl}JY# z3kq4z8h?o$l?SM#_AyD~CDJxN9Nq=suYg(t$P)AOxS|PFT($GNXP*Qwma=xV?;#hx zd(|U}c$cECB#_(fm~508C3!f7oWqOHXh%3)!6l8}%P@N~_Ab_WJx-kF#oH#WWhQ{r zxH?XUGmQ#?#s<=GPFbJhdJBjeKz*U;j=n;IwtgtQtmg3qP^*ek7{OvVg;KP@E@JuDY6Mg=6 zoqsL;e*ezHjE2i8Facj8ukP5iSN+?`kAKsAisL^5qJrEsBV&gTNKiwu_eD`3s!&t3Yy^km|K6Go-di zT!u~qjb}_rHEPIY3ZGv8l=DW_Z~Pw%0uhTP6x0h|Naq^l<5M#HMBK?!*hcz^f@Lh_ zSJT1Vp0v{MtApQTF#1Y6?+_METHY!_%aZ8R_Yura6ioYUwrS?2V&f(2u0H@j)=~zc z@t3l}EA0$tr1nptL3JrAx6O7;x9fqqV1b=TyqJxQIkxChM(^m1WN{AjIQu!(y***<;)wuX)Q+FtMO>WlQ_s3Pfvd%^r zSHRB7tzS;~Y6L*r+Z>s|xe&q_W-H%W8{%gD1}6-6F!fUwd03w()ML!OC)N4su0z)-Lb7-e+mz`tG_#`LsCSePsimrsQAF6;E<=_l0i%H}_EW|;zE1Xm%H^SYo zO#I+#$`p?|#&=i_vPlyj1zO(>OpSh_Xoe7kwW@0 zdGMMl@ipK{Bn#zt^(f3~h9|-~925}##8(4mBmS=I-HwWh60D1?ZS3+<6op7Lwi?6x zA0*yq;+Dh-Ll+7Zp3SvAp9SD+$ozN&^ss^MgZ8j6^9J69$?1V;`OqA7SG(z{WKE5F z!s{zy;o0-SmHU;?Fqy+ky6&Bmhu)&7)%a<2R3W@UjbVX@-~+5RArfIq>Si$+oI8Y+ z#1}!ort(9SFtwY#LKY?c$f|@@@$05RIvJlI>GR zEK`=|^Hz7lcts(CIuUUQJvL@_A|jP`%tXX}G`6T@;F2RIw3e`UQ{?35-#!T5JH{hm zB!S+Z<~ZWwq8t@GAdc7a^foQTBA(==+<2Y!x#G#B81?QI2}WZT!?)7Y_HoB0F-y0k zx}*!#S}Mm`ES@Sfg|%N~?v~#H5RC6thm&*uvX)vak`J>rxH2&p2KFbT#^AHCR{d-% zVKvg@(WW32LNJT|h=;-4q~R`G53hX2ew4{8*DukrDf7Ldgk->8wIgKoj6@Kdg3bkB zbiHXMh*>Io???5McLunL1NU49BdrJ$M4t>7ORBV$gMbdyy|LwDJ(AQ^pYukV+>Dy~ zJiIOQL*NNv-5!Yh=_$Z9F`8FT6%2G^(~lnksx)*M+k6k{<)G^*c^nTXiX7TP*2%iXy zVN?7P7;`o!-*X^a;AosK=tD`}S%jS!hjI!6n86E!`?Jsb{P9BAFoFnE2HI`rr)a|!(3i;>fm|l%b!;d zP=)HoZd{=8BAeQeU#-Rve14Ox)Tch?c@*AIq<7O>1VPMZ z{XTnNrdhI=Th3di?Q7T0rKfF&_$Cc89hD(aI0H@rX04(_pqa)86sH z3^8M%n88|T_7!W{DU6N8+KeT%J18`>9XpaWBZlIPyq>d5emX@hhrw$yF+LVx!s9pi-ZzLpz=C&E{9NY45 z@#Lx-%(HoM$FAIS5kk41QlN8QfsG{xXz3hNUq3v_$bL&V&ZVP2`LBY&gXspPuf(S% zaRx=h?S8XPIeTcskARX|+Xt$9wA(C35?8XRQJ&bH6uZGyzg%PuGlwkkl~Nu_h57)Y zd)URD-j`%Xlgc4q4ez}Ces_gpB2n(8&+GI4$udDfx@RHl*ZZ_`az=GKm|f2aNSlR= zE$ktXPCJSd0CWRDj>aYAVQ`o|;NK?0ZcW^)%k|m{cOYJZ;)u-j`XbhEo9B3j?<1bU zbsxPw;!AvBz|Ahm@TVQridgP|%1on07Pr46K)~#7~P1|}+jqqtQujlSTXsu0p+WXlv0!&z*P*;kRovtEie2o&Y`fa;;shT7YlNJ{ zuYv3uGC#%o3M*^b4s@|J;^TzoD`#E(RU~KhF(XZhy>W7Yw<`sa@u9$!^x7vq>WI*t zqc9EH((IqMZQ@LjeX7@kbO#!a>LfieBFl3LY7*3&zo;R>UTpWeoLUJJE=2m(JNFsE zmGvaVJuK(sK(C8(3ByP#;!SofQy!t26}mEPRK^Tu6Voa6!)OK@iZ(Nl1v&#Sg?}b1 z?;w}QId!q#g8H^K3yQMQY~0PuZgP)bm1F{WZzTrmy{;Cb+$d0C`PgAV1?4Bl`orPh0}F@-egkb`~H z`tjVgKHATtx{TJvYY5F*JNZPsE|d|`*fbX0DnJts>@-uE%i&3^;0T!vPZ4Cvu|+Zc z{RbAJp*n+PbAV{^*=q!K1yboxG^iI}&XT_(NW~~7g|iqHe0cqIfu*2j@<6=Ym*fsZ zEjf||g1}$Nw9Ksf@)MXvqf9e?6RXkbC0EA!BVhdPE-!wN;a_tF2QpKK=7^fgxL7P(ueUC>=>D+a<8Z_*97 zUiLH%1_@>jYOu$Qwb`^?kUGCZmwk&JX7kkPWslO?TeM-u`?QUTbp@?aJcRgAn)GSX z;Md9;ZYlyoTerR0uA+V2Jn(>iP}yzCUJM&3xM6%D8etlQ)SjWAVSv{Y1vBdOOI?lZ z&v^TkLY0O5)dteVdP)J!ps~eE#;t`YLCH&PuH6bj zCCt9e#J32&DqHc03^L0AKV+m7>oa|`A*l_=1YC6ov6fphg8chTp4hR%;}^B~)haYH z_p`-+aC%^KnOFpZ5`_b?7E1FeMTUtnBHiNClW2wARAZ^br^xcify{uGj!G_G{rv$eLB9TU9w zn72bY8JbKaHHP4im)(n3Y{^(wyEGw+E^S>S=cUBPCCY^b!6l|@Vd-WguRw`*l;1nO zJ7f^z{0L}!gRB5e&^AWfttNonEevWc%agwl^^N) znN9&__0;XhOB-{XtVKSwmOKAsNl2fgs9Gq!%!ZSvA!ZYpWD#?(O*B{@EJdfmH5FrH z>_z<}nytL^o4`fy_(NZjE&%(5FNr_)97+N-S>nL&&f&8;P@Tl25<&*Z6e5VY4zMY^ zmb8gLrSf9HiY#dC0dH}s zOM?6SBM~dHxsN;xqS0Jb)kqre`i7V>sIk>LROR<=Rsl8ZfTT#rB#G#=msX=XJqrT? z(biN!WohDk{g?ZHb>sQUXh*-r8}OjSg9$_i1_8stI*+958Y@F7A=gU$(v72 z4*VEjy&rtol2wYG$!G#8TdZq)*4Z68ZgX>UEpK}z-$_@qr2-=iev(3bive0VEwVq& z1Qa)-Af^B`{T2`UFzD;@Jh7-lHhTlXo^rYxg>Pl$=t}uJpPvW4nEaG_mN)TiMrA5a z$PLS7!m9$uidobze6#Ge;0|>UUIT_4fwTVZNi*a2j$xN zNV=fu?QlH{3i`_Td*EunuZ1%YoRB{5*(L*jTl3aX0^AQmIdV5RoIhvQjBVZXz3?+hY!eL!6_>;a_Kc~Jv zJk^Bh7y+^6Hb*rPW;Qc-!$dju0r=-ok@Xq3D$b-S;T>--Uad8jgv%#dbj&y8RWrQp zt6Cu35-Pz#louna0|K4yP57dN2r#xQ4a+I65NQA!t zaSP$#KHYRkEQ*N0ptcuXej&|OHOI|me-1Dm_S9ItTel= z#b}Om$BIKYEzSPykc}mPMB0~i^S<@1Hzrc!n)$Aq$pX!5RG64!P_8lb_-zO3LsX=E zJY;tgr($7K-ul!2tIxWh0J?I~gqS z{Pq~W*8aIBe5(~}G8#c#Ql+*z;JfR8HARV{wt^MgzzMs?J5@(a5c_wNTrlg0E6^XaX zAfh$L3#a4sP@j!15^o3o(r?w*t!Mgm?JeF4D(hUeS#MHF=)Ov3JiSi>tEhEbKVYVs z2nRcSDV!C>*^@;yT_Mum!4tF%O`i!aJPE*N#Zj4CxFs>y$}DhN^$I1^Kr_4RFE$^!1CeMoQ+WYqR-OM?^5Q zb4!bw*$qjunjhMgafcA0z_fMC5T7^F*s~vR-5|R_6Da=-IJPD&rrlW@ab$+)XNwRr)FM$4ysen#`LK8m(_qaaIB&LR5u;o z=QK%-la>6qw0O2_&zl`UCw6?7hy<|+m#l$%Auhcwc)v@ zoGi-e5C*@yCRfw8N8TZNpS(~dX_MWxpG_z+TkxfXh2g93=ruZ2z4=`7t=Myd+vuz# zwXFQ^_jONj+Ji)u?ac8MEtN#3eYFuC3ZRMEPb=q8v^*8Sn%#jqe!r$Qi{$J*UPs(m zEQo8DE#QFX=@bvQJ#T&bTdV5;TA7o?R4kK}p`9}=sTj%+EW0ei8=Q@$R=_A$dEA)M z#u+e|JX^DdL{QOxw*UXxy_?;Y6wS%y{Uk7`ek|=@ z1$b|(GK@`2l7ZgS@Lh?IMhq^pOH_>;qVeJ=;}K%YxggIv?*+S|5vE&QG67B-^_A@B zUoUYJBp(4F|7p!)*PY&~Q1$0a!`|sCMJAswbL%kJ+JxzZH|(yLv5tKU&}S(gwla)9 z93f_`hDI^2#aMyb8&Dp+AWt*8)#!hAxo62U$9}^cS^)Um36ugs%2t>OUbUQ7lJ-yN*Dd^Z+m=h% zv;*!Ii~TMwr^?m*Y5P23==wLM0=-8-Ju=OQTZZjOuGxKYC~yb;P9R#J z5hH;GlWUIn$DsH@ARb2`kpq3Qt@)x*M>7p2ZMZXAtk*J z#XROoml?g^d~aA<`o$eVWgb6Ju{F97X~xE0%Y>E%PXqznwio=l!y%{x{42RR;jN{@ zYW>Cd83;hnh3qq?7%UcKBZJy?s&CPyZB0m1I49EBg6?+bOpQAwwfBlV%px)cKLWc@ zlPXB6!EM6p3*5=1IuP%`E|xdDj?#wyEe+8OUu8`JlAenrvy>jbyUSHwCCl27+hJmL zZ}(wwQs#zy85P@TfxE#*G=pcmD|(mFma@~OR1k+qGBaqyyF9)US)P(C*_4e`Go0{8 zfUzH2sdapE3H)OlWiGS?(rwT2IuJQ%{K~>mq#QyPT3h?x;i+SL+I;;`Uh>kHkOs-7 zEobhZgFY4Ro`i1#^VhyPJO5Nk1WMB|M38(egEjv?3H*HqKmSIt~~A<9@PV&=@)6}|=C##32XZ6Jw`?}U8XJY11X@o_ z*SSZ-N^<|-VWLg5{`to5x2Cb8O-~}@ZqWbvG6bf;+yC!?{{EZbfsef6tN$7nj&MKn0x;}Q~&)gQknnX0qK5Nx}yPqCUs5j2FUSRWLbmBV; zqq&?7agQ|kK4y@=xQ`R}Mf*oBWVetUNyhBGOuw`}#gsiW-qMfRthXd)uKyk%9t1&{ zKfNE3T|w9p$xK=_W3!y_MlpV2L`CcIrcT#MZ@f6Nb5CJtt{`52ff~FLV2LI3L-4Rl zboC_7uC9EVcJ)xY`~!n^8k~=JkKaZNEuYaUzH}k_TH1i-Sd8Upo+<_R6f4fvnHmYB zNuMlE4QqNd7%m?ME>y4bWWdXAc}}E2prgwx{dZFXM=N5kS)2JZJvYCswPDk(F6OHm zE~~MAz#CJtsVw#OIU$rlf#7PCMNW$DqHWqWB8yjkotwe1>$%;JklnoSRR2cIfUv>u zRyB*B>vHI(t45JS(Kgcz3_S)r@h@uDo!e~(@|fhY7wvws{N8{e#m2b$nj$q`Dp7da zF5@JK=-+xZb?}VBgYC%QAgGm5a@Z zQc+>F-)1|7*(c~VM`C7$?7qD`SlI*%GCkL&`Dl|++%$X6=VZ+xA3=-U826G*PAOY; z*dcT7>62TgYwsqWxps9JC*bvwcCLv}9&>8dpRIaCph~8`50O8(+!9n0@xKnE86a!q z1nk(^Y|^U8aljWF@t}J>0$)O7l&P{_Mi3Gxbe%Dc*_4!%*8TcCu;&rWZbwS7WbSS@o@eVM{jhRp$?j((&+KDFe|xl%81(ze^s|w+ z_lc{;Slq~EN|NjJ4^WNKWr#e5+;#SM2JiK$EPnsA5&u6wJ+n*^DTR@D%3u+^P4q8L zD?Cv|_2$aOHbqTZj zrJEvHCKRNqSFsE;72@OcGo;#Dydzfc@p`J#mYgsmVLly42*I24}d(6Yjgw)dK7_P&%n53RRj zes}5^T}hqC7lhE zPT!Att!7~;ag$Q-l`s#-k=p6ju7}ul^quA0mPyp&Xy@2EmyW}83*y11;~RnfRIjcR zJV!vzYhkiHcQ6Y*oI5C5KO6P(qlEWONBxcrj)yXOV$Xwv7DA^tDw>0nq1^-zu+<-HMlPRD&*!s3i}PFa=D|DvN~acusguKGoS>>QYWg ze1WwsiZBoPV8wO|<~-!so-XaUzT-p_oWPaLRA!;t8)ehhqVzfB4%!~oNC)#^Nrs*8 zBTj1(Y3G|7Q3Vmi3tJ!OdM(5c`*STiv%g)?qn_hzCBp~(>5N_&fbn(~->2jl&n{g` zh$J8XjQupzpL3~gF?gfIEzn~Ht(mue*vtoq#&lLSQpE5#8_a+ zQ}pn9s8N%a3o4>pFm9qLVy7RIpu$z(G|tS{G?PjQ-XJt?41r)RNuV$yx1M!E z2)*8UX5|mu43cWb@;=>Fl9lJULnGpIF`2@Z$Qp5G|0opyR~C!EZ*~+KDNIS~)Od0r zqy-0hCcw{#_t+RK;3H9Y>hXy-KL$*h$=_Ec0>||pT{kyKfjf7#8S3Rg$GB~CF(AG-NjewRgZj_@vPsssYBy2H(DqgWy#OfeQTnHa6P9 z*S+i6T=4+kyogc!I|c-H1H=>3fuDST#?>P3^Wi$A-p8Rzm}Luxuz!0{6s^MMuw?@09-P2{cLAb(|^R<%*ZA-ziMgV@nWU4PP#W z`&6EdVsOPpap-A @h5ft)joUN{^~+1m;#qv&xp%)icz7#2`b`8fGtB6wHl+9wkT zj+$E3I_s?cWfEwKv;tC#xqTif{G4Y){ts6o!hXr;y8D$^<0YTNan>YSKwcHh`E zQZc8;m1svc*0yK3ot>=WUnEXAdNkN$xup`eVrE)67jI1Q5hH$-`OsX@S@ z?ky*hKr5$_*!`df@5D(FTK@~Z17N_Hl%=8_0GRp*vc6OSzGO37j`7z^iGYRkq1|hf z{rAWPfPs7n9lZ6|WE3Fa2@FO-_7`kR{ebGD!*9vF;;4oM+|7H8I5W!8d*njRd&!_FvZCOlCCXV*~ z;B<3zJ6L0{e67533SQRW!{4NN0AS_$gJ9=qv#|o1LQaO~vqvBF`R&d7h3nz5jb8Hj zdcTBDDmYDJ$ZfJK8PUufvWJU1*9=EijqF#u6qvIw1kG{LwJ$Abj+fk7uq+j=qvxgL735{QD4VZ?;rj(Q&PAZEz{rsA;SEoFZ=uim$`5 z)(yFOavPkX?hWO?zRvq5#P`HoCC+x9>Qiwb`HFHJ>rWIijn!S3w`XQLC59>WAK|jl z;1@eStBizPt&@BW2^n+6`a1hGfCVM1=Ko7-zw)c4dZPH~bUISg+BMCrIB9nfL8=s6 zo<)@2uveK-b+>O6&au8efc(nzf$=+@v!aI8&Jz4>k-E99&S?C=;0QBA-GFZNV|NC5r&2q1@msK=?O8uF{qX(? z)bi7l7mL%h&ym9L)e*dqrGxvuEyAP0&gh?#Uw7_RDG4;##MZxXL@8%jhUFo~)@$e+ zu8I>J#mcb$9V$^R#rC#vQ}j?5Ux!zw`D|H!$r~uW;{{K#s0@p>z;o$}5WyLrV6tW#eKdwMy~@1xi}rJ0!Q527 zo*4)bE+b@`*F^?APq~WS-d+C8ateGF_KHSL^c zfNv*SwbmrtrkjxpAb1z$2eF@SvHC_@~z67e3x&>-$hRI7}M zxR|OFx8JAI`l?-TNpUBW!z;3Zk&b;4 ze&G*~_`mJWSw5N2R9A^*=cLf*V0|!SZCAIjq&s-lwq=%E*7+{7M=*C&O^b`ni&Yy=4vWf<(nNfH&5o=x=F)O}yh2 zIyg!Y^<5h0uAww?e7J(oac^YP%IzbdnSK3(i%`t*LLZ(v%vIW&99dP`w=F zO*LVi`q~qd-iG8mtThf-zJS;ssz@;@yxT*O zpcgzkto>Pn&n+!Oi|D97h^BP2@nOwo%(~WCRDUP4zi%wfrI1%tWlyNx2SncPLE3<= zr(%9ynKgwhdQMrv9_x;;740bU?YcmsHTmGKS?JEXfp)2F^mJch$640$xo-{M;jR6w zyr7&@xk;@8l@m{`uU|_{X_W`|QLW>YB$HprTCmn6#n^q+HeCz7J-*S8R-R_gkeb|r z$pIw(kkmVj46WQ&ODs@@m&Wj}2em;}(pOu;p-Q&PXti3Co{quU?9(UaheOn(%x{ah zF8lh9jvNtdmIk@wOc93i6+~93jGm{vUe2VWx{j+$uCj)mCn+1zcYCw9!RO;A+O%7+ zCU&<3j)zEgsqbi)R6p1_9u;WeXcNN>2T3ZM{pcna)N;Jg9X-V~Pt-KK%uns(V2?qd zteV@Z$$0JggI8IzpHn{nL~P9_v-v0~wI|2ehRbnsvEFhlbdJq;y5q(6?m<3M_*IEXQ!65(C?U;)6bl% z&H2FMpX{kqOwPw#gAzPf`Pp?S!IV+6;@v;LlWZ@>w%Sl}-?hO*rv9SjTjykzw?ZC& z(3V)5wKXedz*Fc_RT9SZtTe^0p3GBkr=VJF?dSR+y`{=;zAQty6L z7Iv>Nozee{UQH7KZsBM@i(n1F$e3VyuYFT|JN|I5jl^nnZx4&I*8w*x{ew>??oe&=x6iS>b&WGYGkqLHUCXi}imO-O>LFZo? zn5pZR21o7l0wH4ufmxR>+y)zx8qfZPxS>}N z*BuM;pThDA;T@>~1nZYKg!wmH5rHRs+eDCmuqgQHRt6xkE|m-Z#R>3qz!uFiK_Guo z!B0s*gJPA#_VlDsqzW*hl<<2b> zBmTZcw+)^UbD=^0Kwt>k7w|;+ zGYsmd!Dn)(B7^J^kT_p(`2xfBp8- zD7O2;6_KkxFd#lZ#emNQa8&Ba`03>2e8uNax~%c0(0eQ>g!${uCSs|ECceB^mYA9A zmeq-slr~76u`!`(Cxp=h#IjHi)$43jTk7jWw|VxHrEfQPED zEEKP+Qj)rZm$jcb{O2q~IMA|@-9OY_XOd!#?AM~R_B_p^On1A{Mn#sgh{FCg0%^>b ziKciiN=h@HEeUq4UeuQJ_ueR$wVbcFT9e#~XEoKeKOS;FsM*f%%dC|U?3&H1W&3IV z*YpV30v;T6IuDb$CLG=EV^!f`2+YlPV``NN%K4Ow5WxuJ+N;d_I?QFRbfocQsD4Om zxofZE>ug@`)A8bS)`G&75nr8O>;w7?yRUD1Oy#3fJ>1nUENM+M#>STIJr& zCp}pu8tUvcqWzUCC}H=SuFbxdrV2X!F5c=HUpp zmLh*$JS}qH+9jkMZiIi$A3NcDRd^>RBkQjQI9D8&1GLytyre}-Y}aaH|Zbyks;<5>r@7uBV`6h zsn6$%6jk!N1bbwwh`LU7eW}r8ewnse5@Tqx(V`qTf}8kUlDYREHq3^_$Q{W#FmD=g zZNS51J3+oy_MyfyGDdPV$ac0Uk} zM{vGLhZLdBI)}=el#_SLDztNIYP_3FdltO(UQbb^E5_TW+|&;V&N)4pNPb{IE3#s- zL9RXiqrp*+4sdr6BZep0TOU(1g zz}Hu}x^Zx!luv;E7^DL5tI#sHsGxuYt>~HE+D5h7N z%};I$oH8?LOYCJNc=;bc0?GfKd$lD`0J(LF?`Rq)f_Ks(D~$GeYA4bd&Er=R4-{9~?xb5M z`j&-PjZ@Ne4(&@ZIIhKM%&(*P)(DHFi!P<>j7uQ%(kCr6#i zW`zax51C-E(tgw(hI#)It=CjyBEWnFIl;Zo7a=1Bv6Y8RrZQ7>^0SaWR=*Qq(k>%$ zZbf*0+$Y^DEm@(}>O+8;27XKceO_(e2_C(+&)G$5o;$ZMP3=_Mk?kF|0 z+svYdb<_F#_Jmy1I#Pq!o_-rDr9|*v0r>=iJ`P?18-met)=xeUJI(6vH(CNtacRp) zp9a?FO0zZVbjO-^j<{>yBTIwF@HDx@z%9mznd{`hEBwDeEaImQsEyIZ1vvdSy%=^E zKce#qps6G=E(sc=sYQRX6us9}yumiO@csDtcTMO?;tq{l$X;33+XDLLVm4TxJ&BNh zX)YGB|v`hgPHTYSn#C(~vmC$XCPbi^0er z5C^2+BAmpJy#^Hk+btl6n=(ltHEwsCDG!Gc6^i8&PW}B*R8SyH-{nzwYIFP-fG@z) zS!&twQ0!xOyx8A#^&;WpGP!xGY3^^F>@WPOzy{zlef^;Tu^lI`kg{d&#JR~8f2t}a zXfWc#Q&K;6_^yBe8UWHt0ApT22uLd-jVY7th`iD5rdoCA;WU?H3-0Lh&VaRdu>UgP zB$o-B>vgd**L6EvPb84tkAnq+dY7HMBLv&}c=Q!vX!s}ME=?~QW2jL=Onm~B7Ckr0 zec1#<{4s2)Oz-Bwl>PI&uAwXMOAxtl)YzDDPOB9+ZMDs0sV#oV`%9 zIAU5SRv>=2{dQXEa=TZm=$3@<{d)I9=Hrc?#IEiF+<;2i1s;~4563GHd-l3E_EWEP z%fnVytN=``r8{sn37z?r>F~ag{=8!*@vkq`I5EldrcY^Ukbaj7sx>=V!aLv9I2~pn zvUJ?&Xf5R`(_(U|3_?B59nwex-a=e367d0oUNsRRoW8wkHk~# z1kw5;Qs}X*3%QjW4#hsJNVv3$VN9=(kbM=G95=^R#JL;pZ#F@pMPYvrx*Ro&haz?|pWH=m9m-Ix%B4srD~~F3UPkecAXJOsg_kVOmTpGfNmi zSzI^WeZ`-@B*|(HbW7Z>tEU?d1Q~brHPb%5nf@`uj2jhoqX^A z`u)xT$(-X{Sha{Z2SnTo<9j{kUqj%7-PJ~q=mS{N32Jt}HuJ^! zEG9aeF0I@CpgZ>k?DxV8;X$_2p=GB@KBwXEPLdKQBB&YI1pyx?9p`YNCi6A9muQUd zh}`E1V=gY<$u`%%ZOCzX_HiKQlu0}{=CX)w8?h!FxF?$6>h#h0HyE#F07tl-Vj#=D zcmGTP2LvNeh7V2_uB%xdqw88?lT=d1z6cD+-QGu(dDBAKZB@}&TyQP8>N_|Z%D9~YmYz&-i{7KBZkuZS6*wPtWreq z&XM*D9cP!s3`N?^Gz)pW;zH1a8kT5z{y4NUmTZ?&oVl}*khR^c3;YLFS zdC;13R~kW5Z6q9&k^zg5`MGYB03YnFZk>LX{v_Nld~S{YB*Mu#cAaWW)%8r=Qw_OH zNZu;V&={wzkZHwC0BBf(Ld|CoP>@OytnmQGJ^e9LmwKKll4P^mBF;5uySFd zGv|At9J>w=dkI1tnKZhpQ$3h;{cH|zAlTo)7skP?aFpv8Rvj%~n%F}3_8z@Gs;ZD# zwMfWVZ_LkpC14Yr_a3s#K%GtLoAX$wu`&%od%_!fKo-*nd}zl(DtkSRr&e;?xQP-M z6~36Y`V5NYE3zLj2J3(^a0FV)APMu0SBgd>oH4>q_id4FbU&%k_ExDB*?7YtV*>1S ziz(7bE8lRW#mrP49M$>S(_2igVLOy5{U)*0c&pfv04J#X90(uMIYUL5`0Sr7^$gI# zyn!wGG$}Bl6G@Guqqo|`f_xmQ##;oqAJaB*&L`Q^)iOkqieMJ%4BoWC5X7R8e>F8P zNbRM#0s=u3+&%XiF+lbKT8l9p?4#*pWx+7`;2tJZk3?GJ%$tlHS#gPrtm&gKL~3DR zl&816o&KyvPAcEzaTEom9pw%Ra#2xuq~HKGl2sS+A-*9YmSExNB*b@) zYCOj)s4}+LKlzNMvk;3)@VcDTrLM&0=f-#WPcdPNkaP5~#IKBS|4PP}_V$Cs>vw6o zOb}9nr*-p0hWek}*16^}@Ff?N3``z5jvXg{-Qk)L9$u*ZX1HsiG+&vbJ#2ta@)A)0 zqycgbvO;SYFDiGprAIt&&iNX_o%CgHV#fPlRICjKO5S5hGJdXe*~MYZFHPl3_(8cL z_yF%@!K2l)&t8_@HN9MU{N8=Jk&R_uO;k8gL2HbVq<@ zH&8E#D?dCn;jZ|%Si?+G1AiOzMxWxKUP(yMR8%i;rFT1sELovB@ztNwdqP4;h6^|| zTs*wI@3Iivkyl&=f+;Dci$uaLK~ksV8pvlWaRE7O%$5y46U>a38U?f9ea_yxAkS97$-a=2n+ zB4|{8W57QJs{%Zr1KIs%gY;JbozLx#A((wJ z*CFw2(txv+mR6Cb2R=rBElSOa5RQ^A^5w2^ zxxdh5v!;+3%k`E;yIsn@1V^FF8Fz7IGghF^$*8EX#9FBAvq$fLolqQc2wf*N&(lne z4fY?GER!D}vN6Xj2t=j#!<@jE0ep)}D&f!S5NtmII& zGuIODjb6hJ0A`x_rHb&-7AAZTnS~wKgkpZuLqX)f4Nq4+I##t5z$&54l)14?Fx!rOTC(Lj7xd$P>CpJRD4b!A2l8LXvqiHSBh0x6T8kjq4QaIgMCTTDHlF z8V%y~&~PK}s$rnmz_r-9SeG4_I9o5kSfs>rO|NSC5lhqAk6a6-Z;2@sb%hwrhe3+| zz6_u}hLi}+c^%LB2$5IcnQv93FZsCIytJ9$&9k1pAGJB!S6Owa-2lR`0F7YYWOS6?z`UQ{N5YogZ6OV}vSf4P__177;&3F86gT zIP!`~<^yQ`T56wfIRCB}K+DIy4L@hh-iz7&y2{s!J$rSkDRWc;!|B_J18N1~cAA}N0U}uQ!})uco>ksEVIHN=j%zBuyASIylze6w`0;rB{JPty z1hU5dw&;58GKOKNSME5;`Dv_Z+RHM|-ebz+pP&^brCkTnP`^)ntaU^{iB`UzDK|28 zvt5I7%Kf1y4)mn)n}wuzz=Tg1?`Y|&e_7-Mzc=m@{52DwMs{rxHTSCS=+-f;huaLd z3bQsg9}u5wF>UBcKCB?&M86NQjS3|(hzgZ!&fugSPar4>w@I6S*DA0)m$p!>U`3zC z*03J*It@cpe_H0=H<+^ud#bp}9J(}SpBgz4X7^S+q?WTC)dw%5hFP33J=m3&+Fd-h zR_L~Vk8UVsJ=3);O8T`ylUqg=1EIt;9?0Wp0Enj{R6TXEQ#T~HQOFUY`_fzieEhrkua%cg$V;ulQbVwUzqrp)EavZ$^-%yOD_J*nA zU#f%v;&q6$EA^2c?~L?|n2I$W#*7??GGx+FsBxpi5}#cID%Lf5Yp2?1PhE}MhBTY|W$Y)@QG^iQ; z%Ri!3s_#FdRTSjy*JofsD_h3iBtZYYLCyJy0h9@$H-2R?W!fijdzW;3-I`*~sC}B= zVznQJy^x+^`#OQ5C`B3&<;2+R+lfH%NVI@HW4vl;C~|susXJBC^|;LC-|jA|#J}BL zl-5E)ufUr@Ehw#K7XQi?3pYcful7{c=g9?v%GnEWa}K`Lw(|ALyA&q!xA+ zOQ3IJ#Rm!fs@)- zeo47MI6#O<6>wzw5FS@r+jgPqrfRw0`cP4DV>kta2#3kfT4kCd{W=DN{!dieIc9^e z?K?IEqclJFcD@9K8Wxr_bW&y<3&+!k6Ife&)TclOqb0-c>5DVAj_>FVU@0dU* zw>goe&CP!CQ5^k5D`xIO-Z03?T{dd+IYtoI7`u$g)^_R8aM3_6TS&iZbn01l&IJboInuBW1yzTn?6<}4X z7D@h;K)E6?eQ$8q7c7(R-u#S=TcQcf_C+AA(@4K-_qDYsFQAA$&q>I{)OU&R2*SPh zUA2w?GXap730xKOU^epYQ&|ask-VYBpz{6tqNrOKwNSLep~ILX@RuFP(~Bfn-$9aP zHMb&&gmKq1l$kR9cuMpv-nsm^9S`D?R<;no|=Fq z(xgevahiT3?(0yEt^Mjv+E#RdUVBhO%6@DvsSTJYZy!>{U5;fHeRo)1P zo6={XB0ir{;8+A%VzKoLx-5{|5AI(9kQ2?BAWr%T!91!||*Rb|Ek;STAQwAv~lgluf~A2ki2{5 z#y4R4Af**^KkT><&Z0T<*uK=^zm&~`X}vk?wf|9fx(D`fZ%uiXWurvhl8y1Tq#Ch7 z7=1-*X3OATX&a_RS4o*@iawjm5r0q;`@||%f~C{0v~^`^UB8n3Q*M90BuDcqb9NL4 zsXMTF5R=QrfqRL6ywNf6JhuBxMB2a_G~Fvz#fjtb?_@s6#qdbAGPOv1dT^m=c0>5P zp0EDt#F2(&pL>S(fdca&?Zi);y~D6O{mQl4PM`-6fKL<94a&g5r{OlWTwK(WcfCS0@yn$D&UWjN~}&o3k{WCir>G@Ol?4D6j@r@x;^ z$9yRVU`9}K(0yQN(&%+DJ5TP=@J$SFP{##--p8yhTZl}$w z>}HLric^g@zN*LaQu$T__O2`AUZ64Q6q_>|i$(QFTk`HbIt37E6l#ZPn1K5*op+4k zadd`vOH-=iq0zKZI(XG5L>mXFZvC`h+Kk6&hlP*C} zco(qi8=~|ClTUSodJM~GC&+K3`7Fs3EBn}QMERxqjEVK37f?ebK$eqe&SA?25hc5s zu#$%qj$%+zB^9De6C}0w$kDs`$g_4r@A!A0nvfbYaXmE7M^Svcq#8#r{>ivrmFm3S zG*Q1cN*t(gxPtCQ0XqaCpdB$HQk(-r31W-UHS%JtNP%~drV5eL>#=teH*I@1rcu+c%dF-L4x$-!$~Vo6Z4K_|ni=*-EAv%S_;f|t>EvK{XZUEz2wpDu z`*Q#Q!V%M~S^^wM`{_q^ds!rFsk$d`AXACXP<{_z>9&9cXB}LqBJP_)Mrrk`UUwa&arZY>6DmB&lhe?O(3@2c{*c?18QP&*hgP7&|mZ%c_p#A|3r%s z-ROuk()w^2x+?K-rqCz9*n3HGosaT$J2(JvAqYzp3obPd1g4zM!G#T>hhO=jwKQ!- zRgrdI?2>(cTnhjcxjmrB%TOR+3wzd&+jTnEZ2AU2Fb2Jy!|IWzWd?FnI>avwc&uiv z4T)=3*kW>(OlAq}-~d3WXjT(1u#c%~R<}DP{JxPgDFy{f=Y*f%K|FF%Cs>x?Er@zr zzgix7f$;5RLKeLMN1+AuNEF7!q&KdpJ#Vf3xViEjL=wp@ z9RNI;xfWf$3RitRJs-cea=RI8EqGySJQnHo(t0QAF6yrFTwVv}lHu1{GG=QY zMr<_Pk$WJ}qDARN0Q--?Q`2k%cSmH?U9A~Em`5nc%iA70lEJb7Jj;%RtYZM)4?}i% z#S_6>glE_sKK)roCf{_gz!4Ag2XLb?Ql2#zU>%>B;;S7&*ARA1E8ePZ_krtcn#MgD z0AoY({0h_L$3NoPi}%Ku4ogRQwJ=_hqO~pzf)hV$do;+{pBoGhy@)6?M?Ir-^#BRD zt}m-l1xEDN_QdQEx=?}K(0RzK2(9H>{N~5GGe;|LCHru^rCqIqY;y`eU3csV_@_5Z zl!kNB7R2r4(rN$^3vPzbh>A zLkl0R+yK8$Wn#rv=mAnP#!u3O6e|@m2d3*R&&R5hj821uZ}f6q@o)!ASrV z@nBnWZ$z7-x5Vq7Gh`0(_-&K4LdAl8 zi75_bgd!@52^T%n=BteBU^S9LP*FGY=|}a?ji1}P7kU)J__T3TN`EPAY4mY1{9BC- zE?nIZoyJ_cCaG<(py#3q3wQwV@B@dz#~~o+gnU@H13JaVQS120f3!#U(-609P~iVn z8XzMVv8J_hvQ2$hVu#d>s5jF2_oe}6n|`1kV+=x;F&fVZHB#BhB-W-2$brD6ySwTi zv5W$AAp-DWZM5+RC4fnoZile~Lc`gJ8O>m~%m#(3G$w(y0qoGu_{BvJc)T7XS?DAj z6=Rg&?|>z?e{+?m{gn)w2_Q^Nqyvpz0N3CGRUTdkl1^Ba_>|o^^V7az*#rL;bw)55 z#%6y)Gyvz&q$2D=eYYJ^G?~8K?C6sQ$AH;@owp6A6!-+R#K{L)-I!n+IYzh83?BQg zC3U04!}l_Yz!-p=m<63E?et-9PNk3( zB94@WivA68?`LPSKp|8zgVy{R6h`aG{Y=Ml$Zoj&w9K^2>u@Dbo_=1N7mY`-zRjcX z+r}(F%d(hYs7!|Q%|iOTvCw=x9i~1bW4&MbGl5Z<0-BY-0UNL|ouIpq4873}u0i<% z=`0LuxEkb?tDZmu@MnW<-Y>dHWoc*t1tZLLW8MWB()0w_#?%jE4;R3vh|m^m!2zE_ zh|R*t*l5@l^CsVz7Mv4g`i@=@`FR5U$6eSMXBbs2XY<2- zY(~_h?dLq2Pzf$}d{Acz$RBRv-t0dyw8H{DAcrxK6OO=8->&E%9#C7dR$K$JnYSme zV!A=+6FD2&dPlG5w?k$Nz#-T5Hs4SapId7#srQ)@+zYpNdH4_$Vmhv}LO?yFf$%vx1^)8O5H z`BJ+Lz@WAu^p+JM2Cl-ACdPrmQhDu)jYNj$clEr+CanS9ch!331Gd@IOQ_baKlU>0wyt$EYa+*U- zj73_6xshc7`QHm9@I)ZNIhxof{Q9gqQ0Q(HAqdW#ic4`E$eZKDcdl{Z3S3oGN_@Y{ z{`3NX>T3NQ-7^wlPH5BI01MFHzlC9DVg=%DJmdczQrc~&FycKWwZ^{ zoEeqZnsceTO#LAHx8@eUYyY4&d4Iy+d8*Uhe3IUTwz8H`{oYHk8!xx4L>_?Mc<|Yw zBo*#7fr*beTo-vc%5D%SUdn4I!9inP`=yi5YlO!({TIMOlce5ut2c)|Yq5IybWXFi#QK6+|k1=fk$gT@cZWs5{kJhJZ!=vDx)GKzZ zPsWonNWZBnwHCN;5z)6MLCAqd&)uYI{jEq z_j=GWbL9k-2G#skpE+poF^HrV-8j(tdan@^ns_&j^YZla z@$ExNkuO}Pj0`UbeF+~sERQ<&d+;sTZ+eaO3!ax5nzj_k`+_CnGgG89D=cnXRns6B9HCJN6C~aIcCs z>n!1NgUWC#1>Hq1id_b>&jfF0u-ZU7x#@mb`JQ);4j!#2r{Ls%&e`~!s42K4!Ns9L z3f(7ud#|BJjw@MEW21gSdZ!6nEz$4jg755pZ@^4G1#G47ApgQ9eNfW+#i;M0|DJf= zI8MfgW|9ExMdjGd!g^(ik89L5=k4FR#PRKkX=e5uU#6Pkb)HmQ1GnyZlMKg`Vw%=w z

cx?58m&(sSup`X2q%$A0#8_w)Ew z0F#AiW=ZAD(_7s(+cAJbBCw2kq0QE0-iul>v)RF6E%J0lPw{AFogjXSveZm2;|)bvfIMehqbAihQeX8C#;g%j}?f!7~7}1 zG6qEUxr&#a$hzK$yINebsEMmw-IDW|J~Cns=t`ep{e@r0girK73-V@~fm+LVTVotI zmbs&HE`syg*Tewc4 zxy?+)Vynt4L)jbz>FSAs1&>81lVrW!D5rF(V|ahJIV)d#lmJJeG_8^Q^SR#73w~n! z`Ytn7svcZo7Pb17>$t*xoAe&fTO!ZuYVffaRTB{`5#mc1{Yr}+L!Ggyvn6#7GJDg) zPWfFLYT`^RQ?x1)PWz*KyLBeP3w+uJMzNitp$?N2@|e_|ETbZ^pWZLoB4HyCjV_dC;? zKR0k3UFmA$5|P8AX}(-u)Ls}f`xRDvUH763*Cej|*IwvcOP2M45Dv@pwwo~pxV&YN zkC!>IdXw{|TY6udR+fxQzKHO}9Vw#P_3k^cciTvux>)CR*^ya3-f=u0nwD72Gy6(@ zUuwj7Wog0QaFk2Xa1!1{*5hca14EeiRw7Nyc2$=*Z>KKy#=kzXuzRUF;_Rx>KhXDP zKTApZj)cdpic!Lfl2!XX!=&H~7nc!fJ+p&uxth}ivo}?=c_rl}l2DP-T_8})RCyoI zlkCcjm^(_*;lcQY`_m+y!zQd9i7+nbeoEVInJ;AY#&7y~)GV!KC${KWaoyEsTm9Ub zzlST#h|o7aj_H@?d3xDdu0OY?!GBe{+r8q4PYRk>={~E!Nao4w$4;AddIaS&4_Sgb zMFmHmEos<`IA5-RGf;f797Egc!|)*6cR%ri#M?Zz=@m{ac#=eG;IlKv2icV_%W)EL z$BVTBRmWMgy)MbQf&wk*Nqfo0%;~`9v@5#3#OVyS7)#QaD~u6!ZD2GG*{l9-T#LlU zIh{&;Xhgt#dVzew+4JHJ!{iOtM^8PR5(`-=r`TP>H;XYX<@5B++=cf1Ug!)`r5W~$ zaC}1x=)fGxG^}_efBj-RugkPp6#;}T8R*?-DJUl5z~!F2hRY$M#IDJiH5qSVtP2AmVD3=`(Bbyq4ow>)V^~&d*miJi=C{&_&P6O67S`=})oQ((Z7^ zoabR#s= zSl}X$)kRb@q7b>^({sc?D;X88-S0mVd;o=_0~!+j`ZW!2fTMj{QRQEWl1x{I5d~+W z{Z56>)F;)SO@?(p zSo21_j33V!RG33r$EkL@d)JE9eAt>00#mAvHi}jQQ&OXg)xl*%b%uXJJsJ6+mWPTo zdc*(LjD0N6<(%(1=Xsvr>-+D|>*YUlU$^_ZuFrja-rE2cCo;y*hF?9) zd?KXVjw3@9JAWghNCp_qe{k;Ff+Lg}9V0kje($2Suj6uAs6DURFA#ZxZ~xGvoH;Vx zF6<>W9GPS}es~hqAafYu4U1o!@RuspEsEqavFnAEPovmjy-}(K+Y;;Z z37>{&>j9gHh;&5d)6f-}=SHg_1S6vFYM2~2Gmu2r9a!$gE0;MSYGvg`y z%Kqx@S?gx|{?psjS-by)o(3OLO88ht2R=gQIHnXIb~ciAk~qEb)*q+eeiLYktJ=Fg z)Bdl{-l*=U6F6%-y?7~DY70JP>!IEl7uWHQ=L`FB6pqXGPl@2^w;CQArQs(HiS#6CT`^Mlzx!~tU1rJXBFT5vEt#U~2 z3VGjzw#a_@`bqUK-ZSy#6i)cBMn1>;|8L~MMAKfHd0qPkLwjR)e+E%4aijCCfvYG}WJrh-GrZ-_qpW?W{jq6Y8BvjbT=ud681EoNtnn4VLB8lHd z6sM5(^Qt(whPdm<9}(ah5gyJXfHwozn0e!6jc$+BSEQZ0u&@323+)X5h>qUJL)S!h z9CP%~Y0f}R{>wf`J_=6n_|Lxv+*?RXTGBUAF`8y?1`VhQTV47-R0HG&3Oi3$#=d3$ z7mk_bt(nb2$jaHqCfW1Vg(4ha1!$3nF#_407m(c<0q>j%ipDsX^-HF+eq4dM0%#~f ztVL)Ts1{-RLIWJ}T2@NP=S?X|>Yq;WdNUuI-csyf1`p#FEfswm=p~axrXndRHm}*r z9$4LH|Cmh&OqXCGTb%|yu~I^0SsXhqMbA?0a$r%(^i&ori+>q_IjinIxxO3-Sj1Bi zIUug`Gup#3AykQaK^b|5B@t^aY-K(p zOA*QX(Yv%Y=K-tV_L~J$7W$UIEt$1kCT4Mn+FjU8k-&i_R3gTGFeAp_h|a&pD#PFj zCIqZN4x{T!aW7S&=f(w!wl+I&SzXs#pHvkljowMrP}khiW8uH(SuMF$y;D~-;pF38 zWpMW@IiTBk&E^)#-m=hYZSCP?jvBu_!JjxdA4yzO1=U976pBOXHgKTa+i71C32^6% zW3G;yM0U7pbMeP&k_TyMM3{*X%@BTV>KS5E-A7LJ$x@Z6)O!wzI_)0bBoixhPn7!$ zZ}S~@_iy0!{`!pUz1187#hKC1Ma_3oSG8lTuppgFUEK**2>aZZ<{0<#Y>+0S2&vA+ z=H0n95eXTc$HoBkE$#@e*aqDCc8lOCnn+E+z%V{Y4s%&iJ2zP}^>7={jXLdUb&?%N z9jayfZx^3W`xWM|1V|lW*_4!XfkNO8guXl<@dfj;e`%la(p5blPd#FYkM+QpD5Z4v zi4C#w5JL*`thAFQeQejH^E(qYR|RY*>@tlhnR!`3lEMYLd|$wm%1vq9p8|X*2;9Jy zifE~S$TMrZb?%0b-uu2F*}Ok(o)=9&&$d)`M=1oKeRwF#W#)OcV$ZiG-{X=*K-7~p z>r^vqKZcJV5}S8&zh7jObjH5+RP^t?^G}!a0@s>K5QsSjCc`_5>kpFp_ulh3k4MgeZmNA z3O&;2-tn_cA*C5Vyk`FK3n&=H$@}0V+?m|o&sm(Z{)%b*;z#Ie{MRwlrmu0_lYsJs z*Z06?Y*y`Xt+sY-2Wv}pZsR^|hlL!2&KazKVv-*!gQSoV29OtD58Sx79Ca{ZwN&e$i`q(gA|!;5?+UjYgT%&C>!)^?iN zW4_pxW%~4V6cu~KeLeTRiGg_UkeJUL74Pq5H5g(1zK5I`c)wS(_&hB_4d-jKDje5< zqSm#oA~NCMu1S>;tmpxXEzl@QJtn_1B2P**zmI3)bu)CVX2o7L;g;JQolwVK|ZIu%Lh2${#?Xuz-V4 zqgv#cWz)hwi8@@;9k$ffjO?p_`;S3JC!jMx9WQh8?ZKfP!RjpjQe0`)I~ySy7UBC7{Bu_?>X7 zR&#)>#(fqa3a=G8c7ghG2^ ze}tGxe0AG0>a%DDB#Od-Rzq9RklXc2hXaxmo1MERCC}`$^Xb68iJFt)^X@@vx#g;d z3Ffz!iKMqLZ_4CrIZzbj%Dx%EC_Xrlij~%+6H%g1Uv*>(4KEgLQ3O;-G{ci^*>SL( zpLTMAs#%9YEAZu;b{+<*++sEUw!ismetpHaop69X%o9n5x~c4{q4;D?eo1YwRXN_Q z$6nEBQ+UZQ67WT$W;`5!DVqS)bb$DUtPUJE>!yRNRw*O%2td+qp%baqw#0mFe1LVf zGja!^ew&@E>t_w(*+qc)>IUbZ7KO2$R%|=&Y022cKk+!y zN)>$iN-L|6g}x(|BcdzyHnXHvctuB8gOq7NthGve)&SL=->{gq*yFK-H2?X0nzrBT zv>Uk2(za=LyH*?Rf(U;3EV0kMWd~^ra%Pz$MRR%iKHEm12HLyhXgQT{`ONJ^OZXmP z82RHsY+o3^ih|q9r)-O7kDh=t_$L**#tT)4b>`fAzxTn%!-svC-qM|Evvj>}*6cf-Ocq zv4pW^*)V#aUg8`DNQHr)fG-@Eo)> zi;iLI6YI2HjEJ%8aAH3Zr9JE!!8YECI!z7-lL4pAlQ%s7q8F14H#y9OQFFHSBho{x zP8Tb(2$R>^Zmg~Pjiin=07C%L$@c?8CG1!zRg8@^saxe%<5cTC$5|@b?Vt(?Y9=FV z1UvMkopqq7cN}$UL?87@oMZNk@0eu3y~?K8*I zk_wFBC|mF5{pfNl&}mu4P1O@99DBX^4$IwgVAPNJfaY)`&kM}N&#Gw4CDM<^&Gq0+ z^iOKR9lh9O#;t2e#4$D^=4e;OQAT)G-9uWzxOO>%b3@d0tA*nIj5zfb+9^gDNuKt^FZhMe>)9x17Y zJi=^te&vc}7r9>Y(~*k|_2Foglu0*^_U_+smoxL(9r6>da*dxSU4Zp;X+aDN&D<7> zdL>yVfvVS2Gu#bx`Mv5We6zhRqG1@(|5uO=*h`}*((Ll_KDy$_pg`48YAH3I+gVrQ zkAcsUTUCks;iA;%Bf;C^BtjN*<-PGJe5<`JrePRz4RomO$dQYnLuEPcW>yfWF4?Cw zCYeRghPWVm1+iJj&`J>;c)-g%TMr$WE?tiRJbc8(c_`;(DCTe_IgvA<0`XyEzB&)F ztw#%@x8EA})t)sdc<(9wF4GEslq=lcqTTNq6!`(Fg0 zq;UZr$i6MYNF{!eM%7U$;icEJ&oR#4S2}{-3kIQiF4X8!4kEFKv)^;+3B;~&SyYN8 z7Vx6(?7IvSE9&Tc#h<}oG<*(StkSjf6 z(J}JXEg<=IUZx4}Q{BK9`C+Q>vOz#`-mU#K%`2UbmD2%Khs`n#>f8M1q34DLoX`vD zPR6|IwuF^WI6ZRdS+o<=@!niJ%2N!U24>ea)wD%E_1ADbp@ZMc!o4v4$MO-32@ARw z8WL_X1r1$ALs?JPI|1M^yMJ4@BgM~!ILT-bKLMZ6ZQo(l@rnd#9!O_t|?D11J@ zeaV+cSB(B};%=^~nFK@*??cFowqGh5kemhrhGq$j`VQ|!Xejt+WEa!Ni$FT99FN8g zP&E@gsd5{_R0l&X@SSA`+e%m~O0j!dKEP79a| z*v7;gT3bkE9FJT@doHBnSCD5F?Rmpd{mpibkOm+B04=7*^LJQ)SOs+-+=YOo6TRm? zYh+XRV6sn{F(jza>0U+Yj{{__MB+}LxJ?kyY6VSc6!+4eHyhR8YS&0>sHvT}jzt5w z2h6d~iWB}AoT-G+6o$vbG}BL6tKV!vAMBfI1^C4^ghw?2F$DeLNYTVnQfM*ZJ;^5T z9$ps10aChco34RMl2FExQu|1M@W*RUCi(5mid7Nx_T8#nRBFR5V)Qu#&wleoY7(es zn@(Ad<0eeO8104RMCt%{o2XdpM|0mylupX46YWXpahGq3U}vP8Mt)B>U`5?A_WQjd z#_xcnD>!-A8Gx@mU8B{TVjmU;pO^WH3fzUJ_X+Uu@6WmQ{)LBSQ+R|w9%Mf{

ll z`1%@ZJ2Yx-MAVh3SR32C=L7KA99Yk!U>4ZJQJ42`+FTCR)ps0y2Nr!8CgPfQya|wL zoIDS$f;XPpIXarkt}CMmBiqgtrU$rdqdy02_SYWTpE`9>|KvLYmPlP&1Qozke}2ZG zlLmx(NiAfDNg`$IT0cj`iy;Hoa|rr(KnzIvBof>fA)6yb`Ul;^SB3T-hEps%ljt+Y zAUcCOnciYpU|)rbnh3i7GOJhc7F(;PYqXWnXN~AGrBRbJNb8Z`2)9?zpE@fbKsrNN zqu_RqtOubxlAB6DO|J|ZFm0|Xy4CM^vYn9^jd@$KJ)mf>$K~SJ8B!mAy9!Mo5pe40 z!Km4*&*^zbyfY|VX-slq5MDpN;Vc~fvPA4>=xNcwa3B#8T&rg!(5dvl%4I|EW7y<2 z399>pa;lXlz}3;hpd1gaM%p_ll4KM0u&?T-PW&*QUmJBg_ZW5&ow#M4nk&0&q@t)Q z;8~n4KbEs>rX_j8k{dF11|aTXMCyiyE6wrVC8QQtWq`~1&PM~&OT;uat`h3oh!)Z) z$0(@_lLoh>j|D2XY}k>#2*y0Nn?1+3fD%22EuU~mCZLD_qo7CUOZg5@r@&`?ySwPR(~IapGJ7Ns5f@9nVLIGdkskKi>~%R zGV)TZHA{Q^`XaieR?^x9o_Zvt|?4FVbDh3xKzW>f{GSuWj5II#t@UFl$! zJxCa<+opezubUna$mBy5lUhIW+bFY%m?8*=al_p1;BHZgD~JCnsOgN8gCXPnp5Ms!1uM6!zL$lC2l z>*(3Z>*`a$0G@}@faI&X)nrD-p~I7)aYFHZWg}Z~csqfwPgHP!vM-6C#L6Q^<+Y|{ z@~Y5w6Qj{S-TQj9&c=44_YfA-ep=zU%4+OgbD`Tt!eWlXmgLduWuLWi@0p=&1!?s$ z5rNgDodSPFep!-{-Di!Tx2PJupST=09xK6#(?!wTQ8*oB2)3Y29jTK<+Y)NCQHh&D z1;JVoP`Ezk;^DnjQnTGjY-f(R3?44Fjq6G=`+hVGEydovS)`08d>(n_eeXd&6*fR9fvyzcM}fk>qmQ4t7j2=bekD|FfBC* z$t=Ck^HG&X=hB0pNyKD#AhL@lcn_iLv(%#hYi(o>f*= zaQB*9a$jqTMMGp*a@b038U`E_Q+JAwXX!YB!Y@zRzo$P_ZxQSgq- z5G~Z-uJ_Ta#EH-TZnIDoJ<1AIIesDhBZiCcyY159K0ROQN^hyew#8l!6y{%)xZNhN zGeRwnWkT+&_2q@xVLe=Ht)h?#)M#^s+T?sBnl92=dMkf6ZCMP(1T6Jd;xp~R8n|WY z@J04$mjFG@v(mF&uRn|iat>NKVzZ@DP^+-#G_@wu>8#;F?!O&)I*g?wc zdDg*_abps`es|>Y10@-SPs>~WeV$7zK42X%TpJyrAXe58QOZWDqKI{^FXU+Z2kThV zp#lS1mH4HFVR}E>&T7-^IS|CNhky8OE@uYt`S)V{8u~Q`zJY%tnhVR6P~3G&qqK5P z+`ju5U0-6LJZK&5JSV?VVZYYts82nUM$hb#P^CMP!LV_e`Pru*eYxq7geFydVl%gh zvMM-fab)_3X9$a=zuA^!IK(~GJ>W>^{Xnp~lOmPnan-BM!A=T7$~-V%^{`fe+A%YH zdKCh-djtRD6}E*Hke26F-$R`3AOh$&;9z8Hho>{z=!vPmdy(t2fp13PhI4u0vd|Wl z8lO!&`9;Ah723>k%xF!;)5yzTH2NLye+VssMeg?$Om5j{NaGT@YX^*C2z3_T59M3$ zV=8Fkqeg2_1{+tA_L=SR@(`*<0j!Z#Lpa^Ktji-At>01X;{l(`671I2l+YZsbP?HE z>0a5_UORu*KNYewt&|=013RZUk1`MpvUuE6+bwhaJITtu$$E$8+k$=mJB~Ll^jEjy zH0e0SMQep06m<4`nWr;sJ(VH7-*X^gM3fsj-nVKof{|Uz{xh=bfZ-Bu$_CfG|J^I^ za>E_rMtevuyr!ZL{$O#hT)A+wghzKGjn0nw!qR8+3+FEJ$nVMVo6yKh!frcqI~rQE ztKY1?>31PR-fD(&W6G)|3Vys>$C^qj-1)NjWtX&wjPTx(-|9p>Qfj?H2s7F>QDsTa zYcHd;$@o^1xAX>7m0odlcuTbw2fjt0G&^j)4EeSG0# zw_5O>4N-GJ5nY~snwGO3xi92GVs3i%H~Y1e`@IxLNo=3iDY?6m9ZOgVP}K9jcJhoo zHCZs%-Vzq>JIRoIu@920!>tRVNEz=+hp5I^WUIAfElh@yapi8cTl&N5e5j;oF}Bg` zQATy|EYDXOn2%V@)`k=A9?$iK0wD<6vZ#X30E^q2g$}vfVuOdPCz<4a*GjS_{7TJ~ zx*a-k@Y(S0j>7bdY6_o#^WH$G!o4}hp8w{e)17^P&g>PS{?lom`Qdh(mgh~g@3J7R?yqt;q>^hN06}a z6t@{P7M&ngHHtnn{0$=NIlJUjsgMOyBMjqPGgy`n zdTwV`qHga~nF!K}g<^95NFuB(-}dg40NyLufFaq={kO3%8;}R<1=a1$H1&Kf6L4IO zN8k4_;i-2kW;PzLe~+`BoBfiR{GcG+lW(r^;BZUWYJ@wuUwL~E%tl=3EkBD8yc~xr z9Vb12E?nl?})p)mMmVxkRvPcXi2-JWSTJ&$x}NCPka6CY*t95FzaXaAG;qc}b>o ztLOXU!J&gK&rad~mz>ir(^fL1vV#l6dN6&k6vwh6zRH{C_0D!{iGbZTewOerynE?- zvf^vq^4>3v6=!7jqT{ zm@_gr2>LJPYydFlq_fl&|6yz>oR^9p#eymaqN~ACzP+ literal 0 HcmV?d00001 diff --git a/doc/_sources/algorithm/rnn/rnn.txt b/doc/_sources/algorithm/rnn/rnn.txt index a918f02ab1..9653ddbf37 100644 --- a/doc/_sources/algorithm/rnn/rnn.txt +++ b/doc/_sources/algorithm/rnn/rnn.txt @@ -30,7 +30,7 @@ Then at the :code:`process` function, each :code:`yield` function will return th yield src_ids, trg_ids, trg_ids_next -For more details description of how to write a data provider, please refer to :doc:`Python Data Provider <../py_data_provider_wrapper>`. The full data provider file is located at :code:`demo/seqToseq/dataprovider.py`. +For more details description of how to write a data provider, please refer to `PyDataProvider2 <../../ui/data_provider/index.html>`_. The full data provider file is located at :code:`demo/seqToseq/dataprovider.py`. =============================================== Configure Recurrent Neural Network Architecture @@ -106,7 +106,7 @@ We will use the sequence to sequence model with attention as an example to demon In this model, the source sequence :math:`S = \{s_1, \dots, s_T\}` is encoded with a bidirectional gated recurrent neural networks. The hidden states of the bidirectional gated recurrent neural network :math:`H_S = \{H_1, \dots, H_T\}` is called *encoder vector* The decoder is a gated recurrent neural network. When decoding each token :math:`y_t`, the gated recurrent neural network generates a set of weights :math:`W_S^t = \{W_1^t, \dots, W_T^t\}`, which are used to compute a weighted sum of the encoder vector. The weighted sum of the encoder vector is utilized to condition the generation of the token :math:`y_t`. -The encoder part of the model is listed below. It calls :code:`grumemory` to represent gated recurrent neural network. It is the recommended way of using recurrent neural network if the network architecture is simple, because it is faster than :code:`recurrent_group`. We have implemented most of the commonly used recurrent neural network architectures, you can refer to :doc:`Layers <../trainer_config_helpers/layers>` for more details. +The encoder part of the model is listed below. It calls :code:`grumemory` to represent gated recurrent neural network. It is the recommended way of using recurrent neural network if the network architecture is simple, because it is faster than :code:`recurrent_group`. We have implemented most of the commonly used recurrent neural network architectures, you can refer to `Layers <../../ui/api/trainer_config_helpers/layers_index.html>`_ for more details. We also project the encoder vector to :code:`decoder_size` dimensional space, get the first instance of the backward recurrent network, and project it to :code:`decoder_size` dimensional space: @@ -246,6 +246,6 @@ The code is listed below: outputs(beam_gen) -Notice that this generation technique is only useful for decoder like generation process. If you are working on sequence tagging tasks, please refer to :doc:`Semantic Role Labeling Demo <../../../demo/semantic_role_labeling>` for more details. +Notice that this generation technique is only useful for decoder like generation process. If you are working on sequence tagging tasks, please refer to `Semantic Role Labeling Demo <../../demo/semantic_role_labeling/index.html>`_ for more details. The full configuration file is located at :code:`demo/seqToseq/seqToseq_net.py`. diff --git a/doc/_sources/build/build_from_source.txt b/doc/_sources/build/build_from_source.txt index 2c32c37e9d..a191d31318 100644 --- a/doc/_sources/build/build_from_source.txt +++ b/doc/_sources/build/build_from_source.txt @@ -37,11 +37,12 @@ PaddlePaddle also support some build options, you have to install related librar ```bash # necessary sudo apt-get update -sudo apt-get install -y g++ make cmake build-essential libatlas-base-dev python python-pip libpython-dev m4 libprotobuf-dev protobuf-compiler python-protobuf python-numpy git +sudo apt-get install -y g++ make cmake build-essential libatlas-base-dev python python-pip libpython-dev m4 libprotobuf-dev protobuf-compiler python-protobuf python-numpy git # optional sudo apt-get install libgoogle-glog-dev sudo apt-get install libgflags-dev sudo apt-get install libgtest-dev +sudo pip install wheel pushd /usr/src/gtest cmake . make @@ -125,6 +126,8 @@ mkdir build cd build # you can add build option here, such as: cmake -DWITH_GPU=ON -DWITH_DOC=OFF -DCMAKE_INSTALL_PREFIX= .. +# please use sudo make install, if you want +# to install PaddlePaddle into the system make -j `nproc` && make install # PaddlePaddle installation path export PATH=/bin:$PATH diff --git a/doc/_sources/build/contribute_to_paddle.txt b/doc/_sources/build/contribute_to_paddle.txt index b3d5fa7c9f..10d5d86311 100644 --- a/doc/_sources/build/contribute_to_paddle.txt +++ b/doc/_sources/build/contribute_to_paddle.txt @@ -25,7 +25,7 @@ repo or just head straight to the command line: ```shell # Clone your fork to your local machine -git clone git@github.com:USERNAME/paddle.git +git clone git@github.com:USERNAME/Paddle.git ``` Then you can start to develop. @@ -52,7 +52,7 @@ To do this, you'll need to add a remote at first: # see the current configured remote repository git remote -v # add upstream repository -git remote add upstream https://github.com/paddle/paddle.git +git remote add upstream https://github.com/baidu/Paddle.git # verify the new upstream git remote -v ``` diff --git a/doc/_sources/build/docker_install.txt b/doc/_sources/build/docker_install.txt index 997a755956..3cd9d1730a 100644 --- a/doc/_sources/build/docker_install.txt +++ b/doc/_sources/build/docker_install.txt @@ -8,12 +8,12 @@ Docker is a tool designed to make it easier to create, deploy, and run applicati ### PaddlePaddle Docker images There are six Docker images: -- paddledev/paddle:latest-cpu: PaddlePaddle CPU binary image. -- paddledev/paddle:latest-gpu: PaddlePaddle GPU binary image. -- paddledev/paddle:latest-cpu-devel: PaddlePaddle CPU binary image plus source code. -- paddledev/paddle:latest-gpu-devel: PaddlePaddle GPU binary image plus source code. -- paddledev/paddle:latest-cpu-demo: PaddlePaddle CPU binary image plus source code and demo -- paddledev/paddle:latest-gpu-demo: PaddlePaddle GPU binary image plus source code and demo +- paddledev/paddle:cpu-latest: PaddlePaddle CPU binary image. +- paddledev/paddle:gpu-latest: PaddlePaddle GPU binary image. +- paddledev/paddle:cpu-devel-latest: PaddlePaddle CPU binary image plus source code. +- paddledev/paddle:gpu-devel-latest: PaddlePaddle GPU binary image plus source code. +- paddledev/paddle:cpu-demo-latest: PaddlePaddle CPU binary image plus source code and demo +- paddledev/paddle:gpu-demo-latest: PaddlePaddle GPU binary image plus source code and demo Tags with latest will be replaced by a released version. @@ -23,7 +23,7 @@ You have to install Docker in your machine which has linux kernel version 3.10+ You can use ```docker pull ```to download images first, or just launch a container with ```docker run```: ```bash -docker run -it paddledev/paddle:lastest-cpu +docker run -it paddledev/paddle:cpu-latest ``` If you want to launch container with GPU support, you need to set some environment variables at the same time: @@ -31,7 +31,7 @@ If you want to launch container with GPU support, you need to set some environme ```bash export CUDA_SO="$(\ls /usr/lib64/libcuda* | xargs -I{} echo '-v {}:{}') $(\ls /usr/lib64/libnvidia* | xargs -I{} echo '-v {}:{}" export DEVICES=$(\ls /dev/nvidia* | xargs -I{} echo '--device {}:{}') -docker run -it paddledev/paddle:latest-gpu +docker run -it paddledev/paddle:gpu-latest ``` ### Notice diff --git a/doc/_sources/build/index.txt b/doc/_sources/build/index.txt index 6fefa7990a..d6d0d19e11 100644 --- a/doc/_sources/build/index.txt +++ b/doc/_sources/build/index.txt @@ -5,9 +5,11 @@ Install PaddlePaddle ---------------------- .. toctree:: + :maxdepth: 1 :glob: install_* + internal/install_from_jumbo.md Build from Source ----------------- @@ -15,6 +17,7 @@ Build from Source If you want to hack and contribute PaddlePaddle source code, following guides can help you\: .. toctree:: + :maxdepth: 1 :glob: build_from_source.md @@ -29,6 +32,7 @@ state and your experience of installation may not be smooth. If you want to pack docker image, the following guide can help you\: .. toctree:: + :maxdepth: 1 :glob: docker_install.md diff --git a/doc/_sources/cluster/index.txt b/doc/_sources/cluster/index.txt index cf1ea97715..9062f85f98 100644 --- a/doc/_sources/cluster/index.txt +++ b/doc/_sources/cluster/index.txt @@ -5,3 +5,4 @@ Cluster Train :glob: opensource/cluster_train.md + internal/index.md diff --git a/doc/_sources/demo/embedding_model/index.txt b/doc/_sources/demo/embedding_model/index.txt index 45992ad856..06f3ff1f00 100644 --- a/doc/_sources/demo/embedding_model/index.txt +++ b/doc/_sources/demo/embedding_model/index.txt @@ -93,7 +93,7 @@ where `train.sh` is almost the same as `demo/seqToseq/translation/train.sh`, the - `--init_model_path`: path of the initialization model, here is `data/paraphrase_model` - `--load_missing_parameter_strategy`: operations when model file is missing, here use a normal distibution to initialize the other parameters except for the embedding layer -For users who want to understand the dataset format, model architecture and training procedure in detail, please refer to [Text generation Tutorial](text_generation.md). +For users who want to understand the dataset format, model architecture and training procedure in detail, please refer to [Text generation Tutorial](../text_generation/text_generation.md). ## Optional Function ## ### Embedding Parameters Observation diff --git a/doc/_sources/demo/imagenet_model/resnet_model.txt b/doc/_sources/demo/imagenet_model/resnet_model.txt index 2e5c7f3244..5403ab9f17 100644 --- a/doc/_sources/demo/imagenet_model/resnet_model.txt +++ b/doc/_sources/demo/imagenet_model/resnet_model.txt @@ -165,7 +165,7 @@ We provide both C++ and Python interfaces to extract features. The following exa ### C++ Interface -First, specify image data list in `define_py_data_sources` in the config, see example `demo/model_zoo/resnet/resnet.py`. +First, specify image data list in `define_py_data_sources2` in the config, see example `demo/model_zoo/resnet/resnet.py`. ``` train_list = 'train.list' if not is_test else None diff --git a/doc/_sources/demo/quick_start/index_en.txt b/doc/_sources/demo/quick_start/index_en.txt index b537d8c834..ee3fa2a216 100644 --- a/doc/_sources/demo/quick_start/index_en.txt +++ b/doc/_sources/demo/quick_start/index_en.txt @@ -59,7 +59,7 @@ To build your text classification system, your code will need to perform five st ## Preprocess data into standardized format In this example, you are going to use [Amazon electronic product review dataset](http://jmcauley.ucsd.edu/data/amazon/) to build a bunch of deep neural network models for text classification. Each text in this dataset is a product review. This dataset has two categories: “positive” and “negative”. Positive means the reviewer likes the product, while negative means the reviewer does not like the product. -`demo/quick_start` provides scripts for downloading data and preprocessing data as shown below. The data process takes several minutes (about 3 minutes in our machine). +`demo/quick_start` in the [source code](https://github.com/baidu/Paddle) provides scripts for downloading data and preprocessing data as shown below. The data process takes several minutes (about 3 minutes in our machine). ```bash cd demo/quick_start @@ -157,9 +157,7 @@ define_py_data_sources2(train_list='data/train.list', obj="process", args={"dictionary": word_dict}) ``` - -You can refer to the following link for more detailed examples -: Python Use Case，The detailed documentation on data format is: PyDataProviderWrapper。 +You can refer to the following link for more detailed examples and data formats: PyDataProvider2. ## Network Architecture You will describe four kinds of network architectures in this section. @@ -425,7 +423,7 @@ paddle train \ mv rank-00000 result.txt ``` -There are several differences between training and inference network configurations. +User can choose the best model base on the training log instead of model `output/pass-00003`. There are several differences between training and inference network configurations. - You do not need labels during inference. - Outputs need to be specified to the classification probability layer (the output of softmax layer), or the id of maximum probability (`max_id` layer). An example to output the id and probability is given in the code snippet. - batch_size = 1. diff --git a/doc/_sources/demo/rec/ml_regression.txt b/doc/_sources/demo/rec/ml_regression.txt index 47bcef2d6d..0c14e4f5bb 100644 --- a/doc/_sources/demo/rec/ml_regression.txt +++ b/doc/_sources/demo/rec/ml_regression.txt @@ -219,9 +219,9 @@ The network structure shows below. The demo's neural network config file "trainer_config.py" show as below. -.. include:: ../../../demo/recommendation/trainer_config.py - :code: python - :literal: +.. literalinclude:: ../../../demo/recommendation/trainer_config.py + :language: python + :lines: 15- In this :code:`trainer_config.py`, we just map each feature type to a feature vector, following shows how to map each feature to a vector shows below. @@ -257,15 +257,15 @@ In these network, we use several api in `trainer_config_helpers * Text Convolution Pooling Layer, `text_conv_pool <../../ui/api/trainer_config_helpers/networks.html #trainer_config_helpers.networks.text_conv_pool>`_ -* Declare Python Data Sources, `define_py_data_sources +* Declare Python Data Sources, `define_py_data_sources2 <../../ui/api/trainer_config_helpers/data_sources.html>`_ Data Provider ''''''''''''' -.. include:: ../../../demo/recommendation/dataprovider.py - :code: python - :literal: +.. literalinclude:: ../../../demo/recommendation/dataprovider.py + :language: python + :lines: 15- The data provider just read the meta.bin and rating file, yield each sample for training. In this :code:`dataprovider.py`, we should set\: @@ -274,7 +274,7 @@ In this :code:`dataprovider.py`, we should set\: * use_seq\: Whether this :code:`dataprovider.py` in sequence mode or not. * process\: Return each sample of data to :code:`paddle`. -The data provider details document see `there <../../ui/DataProvider.html>`_. +The data provider details document see `there <../../ui/data_provider/pydataprovider2.html>`_. Train ````` @@ -283,15 +283,15 @@ After prepare data, config network, writting data provider, now we can run paddl The run.sh is shown as follow: -.. include:: ../../../demo/recommendation/run.sh - :code: bash - :literal: +.. literalinclude:: ../../../demo/recommendation/run.sh + :language: bash + :lines: 16- It just start a paddle training process, write the log to `log.txt`, then print it on screen. Each command line argument in :code:`run.sh`, please refer to the `command line -arguments `_ page. The short description of these arguments is shown as follow. +arguments <../../ui/index.html#command-line-argument>`_ page. The short description of these arguments is shown as follow. * config\: Tell paddle which file is neural network configuration. * save_dir\: Tell paddle save model into './output' @@ -303,8 +303,6 @@ arguments `_ page. The short description of these arguments is shown as fol * dot_period\: Print a :code:`.` after train :code:`dot_period` batches. * num_passes\: Train at most :code:`num_passes`. - - If training process starts successfully, the output likes follow: .. code-block:: text diff --git a/doc/_sources/demo/semantic_role_labeling/index.txt b/doc/_sources/demo/semantic_role_labeling/index.txt new file mode 100644 index 0000000000..ff3035059b --- /dev/null +++ b/doc/_sources/demo/semantic_role_labeling/index.txt @@ -0,0 +1,7 @@ +Semantic Role Labeling Tutorial +=============================== + +.. toctree:: + :maxdepth: 3 + + semantic_role_labeling.md diff --git a/doc/_sources/demo/semantic_role_labeling/semantic_role_labeling.txt b/doc/_sources/demo/semantic_role_labeling/semantic_role_labeling.txt new file mode 100644 index 0000000000..05fbc8278d --- /dev/null +++ b/doc/_sources/demo/semantic_role_labeling/semantic_role_labeling.txt @@ -0,0 +1,183 @@ +# Semantic Role labeling Tutorial # + +Semantic role labeling (SRL) is a form of shallow semantic parsing whose goal is to discover the predicate-argument structure of each predicate in a given input sentence. SRL is useful as an intermediate step in a wide range of natural language processing tasks, such as information extraction. automatic document categorization and question answering. An instance is as following [1]: + + [ A0 He ] [ AM-MOD would ][ AM-NEG n’t ] [ V accept] [ A1 anything of value ] from [A2 those he was writing about ]. + +- V: verb +- A0: acceptor +- A1: thing accepted +- A2: accepted-from +- A3: Attribute +- AM-MOD: modal +- AM-NEG: negation + +Given the verb "accept", the chunks in sentence would play certain semantic roles. Here, the label scheme is from Penn Proposition Bank. + +To this date, most of the successful SRL systems are built on top of some form of parsing results where pre-defined feature templates over the syntactic structure are used. This tutorial will present an end-to-end system using deep bidirectional long short-term memory (DB-LSTM)[2] for solving the SRL task, which largely outperforms the previous state-of-the-art systems. The system regards SRL task as the sequence labelling problem. + +## Data Description +The relevant paper[2] takes the data set in CoNLL-2005&2012 Shared Task for training and testing. Accordingto data license, the demo adopts the test data set of CoNLL-2005, which can be reached on website. + +To download and process the original data, user just need to execute the following command: + +```bash +cd data +./get_data.sh +``` +Several new files appear in the `data `directory as follows. +```bash +conll05st-release：the test data set of CoNll-2005 shared task +test.wsj.words：the Wall Street Journal data sentences +test.wsj.props: the propositional arguments +src.dict：the dictionary of words in sentences +tgt.dict：the labels dictionary +feature: the extracted features from data set +``` + +## Training +### DB-LSTM +Please refer to the Sentiment Analysis demo to learn more about the long short-term memory unit. + +Unlike Bidirectional-LSTM that used in Sentiment Analysis demo, the DB-LSTM adopts another way to stack LSTM layer. First a standard LSTM processes the sequence in forward direction. The input and output of this LSTM layer are taken by the next LSTM layer as input, processed in reversed direction. These two standard LSTM layers compose a pair of LSTM. Then we stack LSTM layers pair after pair to obtain the deep LSTM model. + +The following figure shows a temporal expanded 2-layer DB-LSTM network. +

+![pic](./network_arch.png) +
+ +### Features +Two input features play an essential role in this pipeline: predicate (pred) and argument (argu). Two other features: predicate context (ctx-p) and region mark (mr) are also adopted. Because a single predicate word can not exactly describe the predicate information, especially when the same words appear more than one times in a sentence. With the predicate context, the ambiguity can be largely eliminated. Similarly, we use region mark mr = 1 to denote the argument position if it locates in the predicate context region, or mr = 0 if does not. These four simple features are all we need for our SRL system. Features of one sample with context size set to 1 is showed as following[2]: +
+![pic](./feature.jpg) +
+ +In this sample, the coresponding labelled sentence is: + +[ A1 A record date ] has [ AM-NEG n't ] been [ V set ] . + +In the demo, we adopt the feature template as above, consists of : `argument`, `predicate`, `ctx-p (p=-1,0,1)`, `mark` and use `B/I/O` scheme to label each argument. These features and labels are stored in `feature` file, and separated by `\t`. + +### Data Provider + +`dataprovider.py` is the python file to wrap data. `hook()` function is to define the data slots for network. The Six features and label are all IndexSlots. +``` +def hook(settings, word_dict, label_dict, **kwargs): + settings.word_dict = word_dict + settings.label_dict = label_dict + #all inputs are integral and sequential type + settings.slots = [ + integer_value_sequence(len(word_dict)), + integer_value_sequence(len(word_dict)), + integer_value_sequence(len(word_dict)), + integer_value_sequence(len(word_dict)), + integer_value_sequence(len(word_dict)), + integer_value_sequence(2), + integer_value_sequence(len(label_dict))] +``` +The corresponding data iterator is as following: +``` +@provider(use_seq=True, init_hook=hook) +def process(obj, file_name): + with open(file_name, 'r') as fdata: + for line in fdata: + sentence, predicate, ctx_n1, ctx_0, ctx_p1, mark, label = line.strip().split('\t') + words = sentence.split() + sen_len = len(words) + word_slot = [obj.word_dict.get(w, UNK_IDX) for w in words] + + predicate_slot = [obj.word_dict.get(predicate, UNK_IDX)] * sen_len + ctx_n1_slot = [obj.word_dict.get(ctx_n1, UNK_IDX) ] * sen_len + ctx_0_slot = [obj.word_dict.get(ctx_0, UNK_IDX) ] * sen_len + ctx_p1_slot = [obj.word_dict.get(ctx_p1, UNK_IDX) ] * sen_len + + marks = mark.split() + mark_slot = [int(w) for w in marks] + + label_list = label.split() + label_slot = [obj.label_dict.get(w) for w in label_list] + + yield word_slot, predicate_slot, ctx_n1_slot, ctx_0_slot, ctx_p1_slot, mark_slot, label_slot +``` +The `process`function yield 7 lists which are six features and labels. + +### Neural Network Config +`db_lstm.py` is the neural network config file to load the dictionaries and define the data provider module and network architecture during the training procedure. + +Seven `data_layer` load instances from data provider. Six features are transformed into embedddings respectively, and mixed by `mixed_layer` . Deep bidirectional LSTM layers extract features for the softmax layer. The objective function is cross entropy of labels. + +### Run Training +The script for training is `train.sh`, user just need to execute: +```bash + ./train.sh +``` +The content in `train.sh`: +``` +paddle train \ + --config=./db_lstm.py \ + --save_dir=./output \ + --trainer_count=4 \ + --log_period=10 \ + --num_passes=500 \ + --use_gpu=false \ + --show_parameter_stats_period=10 \ + --test_all_data_in_one_period=1 \ +2>&1 | tee 'train.log' +``` + +- \--config=./db_lstm.py : network config file. +- \--save_di=./output: output path to save models. +- \--trainer_count=4 : set thread number (or GPU count). +- \--log_period=10 : print log every 20 batches. +- \--num_passes=500: set pass number, one pass in PaddlePaddle means training all samples in dataset one time. +- \--use_gpu=false: use CPU to train, set true, if you install GPU version of PaddlePaddle and want to use GPU to train. +- \--show_parameter_stats_period=10: show parameter statistic every 100 batches. +- \--test_all_data_in_one_period=1: test all data in every testing. + + +After training, the models will be saved in directory `output`. + +### Run testing +The script for testing is `test.sh`, user just need to execute: +```bash + ./test.sh +``` +The main part in `tesh.sh` +``` +paddle train \ + --config=./db_lstm.py \ + --model_list=$model_list \ + --job=test \ + --config_args=is_test=1 \ +``` + + - \--config=./db_lstm.py: network config file + - \--model_list=$model_list.list: model list file + - \--job=test: indicate the test job + - \--config_args=is_test=1: flag to indicate test + + +### Run prediction +The script for prediction is `predict.sh`, user just need to execute: +```bash + ./predict.sh + +``` +In `predict.sh`, user should offer the network config file, model path, label file, word dictionary file, feature file +``` +python predict.py + -c $config_file + -w $model_path + -l $label_file + -d $dict_file + -i $input_file +``` + +`predict.py` is the main executable python script, which includes functions: load model, load data, data prediction. The network model will output the probability distribution of labels. In the demo, we take the label with maximum probability as result. User can also implement the beam search or viterbi decoding upon the probability distribution matrix. + +After prediction, the result is saved in `predict.res`. + +## Reference +[1] Martha Palmer, Dan Gildea, and Paul Kingsbury. The Proposition Bank: An Annotated Corpus of Semantic Roles , Computational Linguistics, 31(1), 2005. + +[2] Zhou, Jie, and Wei Xu. "End-to-end learning of semantic role labeling using recurrent neural networks." Proceedings of the Annual Meeting of the Association for Computational Linguistics. 2015. diff --git a/doc/_sources/source/math/utils/utils.txt b/doc/_sources/source/math/utils/utils.txt index e00dc6229c..3df721a47b 100644 --- a/doc/_sources/source/math/utils/utils.txt +++ b/doc/_sources/source/math/utils/utils.txt @@ -1,10 +1,6 @@ Utils ======= -Bits -------- -.. doxygenfile:: paddle/math/Bits.h - Memory Handle -------------- .. doxygenfile:: paddle/math/MemoryHandle.h diff --git a/doc/_sources/ui/api/trainer_config_helpers/layers.txt b/doc/_sources/ui/api/trainer_config_helpers/layers.txt index a09d5e3d4d..f902d1c995 100644 --- a/doc/_sources/ui/api/trainer_config_helpers/layers.txt +++ b/doc/_sources/ui/api/trainer_config_helpers/layers.txt @@ -82,12 +82,6 @@ img_cmrnorm_layer :members: img_cmrnorm_layer :noindex: -img_rnorm_layer ------------------ -.. automodule:: paddle.trainer_config_helpers.layers - :members: img_rnorm_layer - :noindex: - batch_norm_layer --------------------- .. automodule:: paddle.trainer_config_helpers.layers @@ -251,10 +245,10 @@ addto_layer :members: addto_layer :noindex: -convex_comb_layer +linear_comb_layer ----------------- .. automodule:: paddle.trainer_config_helpers.layers - :members: convex_comb_layer + :members: linear_comb_layer :noindex: interpolation_layer @@ -286,7 +280,13 @@ tensor_layer .. automodule:: paddle.trainer_config_helpers.layers :members: tensor_layer :noindex: - + +cos_sim +------- +.. automodule:: paddle.trainer_config_helpers.layers + :members: cos_sim + :noindex: + trans_layer ------------ .. automodule:: paddle.trainer_config_helpers.layers @@ -347,12 +347,6 @@ rank_cost :members: rank_cost :noindex: -cos_sim -------- -.. automodule:: paddle.trainer_config_helpers.layers - :members: cos_sim - :noindex: - crf_layer ----------------- .. automodule:: paddle.trainer_config_helpers.layers diff --git a/doc/_sources/ui/data_provider/index.txt b/doc/_sources/ui/data_provider/index.txt index e702b1a7e9..3db5b57376 100644 --- a/doc/_sources/ui/data_provider/index.txt +++ b/doc/_sources/ui/data_provider/index.txt @@ -1,5 +1,5 @@ -PaddlePaddle DataProvider Introduction -================================ +DataProvider Introduction +========================= DataProvider is a module that loads training or testing data into cpu or gpu memory for the following triaining or testing process. diff --git a/doc/_sources/ui/data_provider/pydataprovider2.txt b/doc/_sources/ui/data_provider/pydataprovider2.txt index 94472ad0d8..152f8a6df6 100644 --- a/doc/_sources/ui/data_provider/pydataprovider2.txt +++ b/doc/_sources/ui/data_provider/pydataprovider2.txt @@ -152,7 +152,6 @@ Please refer to the following section reference for details. Reference --------- -.. _@provider:: @provider +++++++++ @@ -170,31 +169,28 @@ PaddlePaddle from a user defined function. Its parameters are: usefull in sequential model, that defines batch size is counted upon sequence or token. By default, each sample or sequence counts to 1 when calculating batch size. -* cache is a data cache strategy, see `cache`_ +* cache is a data cache strategy, see `cache`_. * Init_hook function is invoked once the data provider is initialized, - see `init_hook`_ + see `init_hook`_. -.. _input_types:: input_types +++++++++++ PaddlePaddle has four data types, and three sequence types. The four data types are: -* dense_vector represents dense float vector. -* sparse_binary_vector sparse binary vector, most of the value is 0, and +* :code:`dense_vector`: dense float vector. +* :code:`sparse_binary_vector`: sparse binary vector, most of the value is 0, and the non zero elements are fixed to 1. -* sparse_float_vector sparse float vector, most of the value is 0, and some - non zero elements that can be any float value. They are given by the user. -* integer represents an integer scalar, that is especially used for label or - word index. +* :code:`sparse_float_vector`: sparse float vector, most of the value is 0, and some + non zero elements can be any float value. They are given by the user. +* :code:`integer`: an integer scalar, that is especially used for label or word index. +The three sequence types are: -The three sequence types are - -* SequenceType.NO_SEQUENCE means the sample is not a sequence -* SequenceType.SEQUENCE means the sample is a sequence -* SequenceType.SUB_SEQUENCE means it is a nested sequence, that each timestep of +* :code:`SequenceType.NO_SEQUENCE` means the sample is not a sequence. +* :code:`SequenceType.SEQUENCE` means the sample is a sequence. +* :code:`SequenceType.SUB_SEQUENCE` means it is a nested sequence, that each timestep of the input sequence is also a sequence. Different input type has a defferenct input format. Their formats are shown @@ -214,36 +210,39 @@ in the above table. where f represents a float value, i represents an integer value. -.. _init_hook:: -.. _settings:: init_hook +++++++++ init_hook is a function that is invoked once the data provoder is initialized. Its parameters lists as follows: -* The first parameter is a settings object, which is the same to :code:'settings' - in :code:`process` method. The object contains several attributes, including: - * settings.input_types the input types. Reference `input_types`_ - * settings.logger a logging object +* The first parameter is a settings object, which is the same to :code:`settings` + in :code:`process` method. The object contains several attributes, including: + + * :code:`settings.input_types`: the input types. Reference `input_types`_. + * :code:`settings.logger`: a logging object. + * The rest parameters are the key word arguments. It is made up of PaddpePaddle pre-defined parameters and user defined parameters. - * PaddlePaddle defines parameters including: - * is_train is a bool parameter that indicates the DataProvider is used in - training or testing - * file_list is the list of all files. + + * PaddlePaddle-defined parameters including: + + * :code:`is_train` is a bool parameter that indicates the DataProvider is used in + training or testing. + * :code:`file_list` is the list of all files. + * User-defined parameters args can be set in training configuration. Note, PaddlePaddle reserves the right to add pre-defined parameter, so please use :code:`**kwargs` in init_hook to ensure compatibility by accepting the parameters which your init_hook does not use. -.. _cache :: cache +++++ -DataProvider provides two simple cache strategy. They are -* CacheType.NO_CACHE means do not cache any data, then data is read at runtime by +DataProvider provides two simple cache strategy. They are: + +* :code:`CacheType.NO_CACHE` means do not cache any data, then data is read at runtime by the user implemented python module every pass. -* CacheType.CACHE_PASS_IN_MEM means the first pass reads data by the user +* :code:`CacheType.CACHE_PASS_IN_MEM` means the first pass reads data by the user implemented python module, and the rest passes will directly read data from memory. diff --git a/doc/_sources/ui/index.txt b/doc/_sources/ui/index.txt index 829994d56b..9c1ba27bdc 100644 --- a/doc/_sources/ui/index.txt +++ b/doc/_sources/ui/index.txt @@ -7,7 +7,7 @@ ## API Reference -* [Trainer Config Helpers](api/trainer_config_helpers/index.md) +* [Model Config Interface](api/trainer_config_helpers/index.md) ## Command Line Argument diff --git a/doc/_sources/ui/predict/swig_py_paddle_en.txt b/doc/_sources/ui/predict/swig_py_paddle_en.txt index e22d0bff33..b743fc4569 100644 --- a/doc/_sources/ui/predict/swig_py_paddle_en.txt +++ b/doc/_sources/ui/predict/swig_py_paddle_en.txt @@ -10,27 +10,35 @@ SWIG. The main steps of predict values in python are: * Predict Here is a sample python script that shows the typical prediction process for the -MNIST classification problem. +MNIST classification problem. A complete sample code could be found at +:code:`src_root/doc/ui/predict/predict_sample.py`. .. literalinclude:: ./predict_sample.py :language: python - :linenos: + :lines: 15-18,90-100,101-104 The module that does the most of the job is py_paddle.swig_paddle, it's generated by SWIG and has complete documents, for more details you can use python's :code:`help()` function. Let's walk through the above python script: -* At the beginning, initialize PaddlePaddle with command line arguments(line 90). -* Parse the configuration file that is used in training(line 93). -* Create a neural network at line 95 according the parsed configuration, then - load the trained parameters from model at line 97. -* A utility class for data transformation is created at line 98. +* At the beginning, use :code:`swig_paddle.initPaddle()` to initialize + PaddlePaddle with command line arguments, for more about command line arguments + see `Command Line Arguments <../cmd_argument/detail_introduction.html>`_. +* Parse the configuration file that is used in training with :code:`parse_config()`. + Because data to predict with always have no label, and output of prediction work + normally is the output layer rather than the cost layer, so you should modify + the configuration file accordingly before using it in the prediction work. +* Create a neural network with + :code:`swig_paddle.GradientMachine.createFromConfigproto()`, which takes the + parsed configuration :code:`conf.model_config` as argument. Then load the + trained parameters from the model with :code:`network.loadParameters()`. +* Create a data converter object of utility class :code:`DataProviderConverter`. - Note: As swig_paddle can only accept C++ matrices, we offer a utility - class DataProviderWraaperConverter that can accept the same input data with - PyDataProviderWrapper, for more information please refer to document - of `PyDataProviderWrapper <../py_data_provider_wrapper_api.html>`_. -* Do the prediction and output the result at line 100, forwardTest is another - utility class that directly takes the activations of the output layer. + class DataProviderConverter that can accept the same input data with + PyDataProvider2, for more information please refer to document + of `PyDataProvider2 <../data_provider/pydataprovider2.html>`_. +* Do the prediction with :code:`forwardTest()`, which takes the converted + input data and outputs the activations of the output layer. Here is a typical output: diff --git a/doc/_static/basic.css b/doc/_static/basic.css index c89fc7e920..2b513f0c96 100644 --- a/doc/_static/basic.css +++ b/doc/_static/basic.css @@ -52,6 +52,8 @@ div.sphinxsidebar { width: 230px; margin-left: -100%; font-size: 90%; + word-wrap: break-word; + overflow-wrap : break-word; } div.sphinxsidebar ul { @@ -83,10 +85,6 @@ div.sphinxsidebar #searchbox input[type="text"] { width: 170px; } -div.sphinxsidebar #searchbox input[type="submit"] { - width: 30px; -} - img { border: 0; max-width: 100%; @@ -187,6 +185,13 @@ div.genindex-jumpbox { /* -- general body styles --------------------------------------------------- */ +div.body p, div.body dd, div.body li, div.body blockquote { + -moz-hyphens: auto; + -ms-hyphens: auto; + -webkit-hyphens: auto; + hyphens: auto; +} + a.headerlink { visibility: hidden; } diff --git a/doc/_static/doctools.js b/doc/_static/doctools.js index e2e70cc287..8163495635 100644 --- a/doc/_static/doctools.js +++ b/doc/_static/doctools.js @@ -124,6 +124,7 @@ var Documentation = { this.fixFirefoxAnchorBug(); this.highlightSearchWords(); this.initIndexTable(); + }, /** @@ -252,6 +253,29 @@ var Documentation = { }); var url = parts.join('/'); return path.substring(url.lastIndexOf('/') + 1, path.length - 1); + }, + + initOnKeyListeners: function() { + $(document).keyup(function(event) { + var activeElementType = document.activeElement.tagName; + // don't navigate when in search box or textarea + if (activeElementType !== 'TEXTAREA' && activeElementType !== 'INPUT' && activeElementType !== 'SELECT') { + switch (event.keyCode) { + case 37: // left + var prevHref = $('link[rel="prev"]').prop('href'); + if (prevHref) { + window.location.href = prevHref; + return false; + } + case 39: // right + var nextHref = $('link[rel="next"]').prop('href'); + if (nextHref) { + window.location.href = nextHref; + return false; + } + } + } + }); } }; @@ -260,4 +284,4 @@ _ = Documentation.gettext; $(document).ready(function() { Documentation.init(); -}); +}); \ No newline at end of file diff --git a/doc/_static/searchtools.js b/doc/_static/searchtools.js index cb7446728a..066857ce21 100644 --- a/doc/_static/searchtools.js +++ b/doc/_static/searchtools.js @@ -2,7 +2,7 @@ * searchtools.js_t * ~~~~~~~~~~~~~~~~ * - * Sphinx JavaScript utilties for the full-text search. + * Sphinx JavaScript utilities for the full-text search. * * :copyright: Copyright 2007-2016 by the Sphinx team, see AUTHORS. * :license: BSD, see LICENSE for details. @@ -623,7 +623,7 @@ var Search = { * helper function to return a node containing the * search summary for a given text. keywords is a list * of stemmed words, hlwords is the list of normal, unstemmed - * words. the first one is used to find the occurance, the + * words. the first one is used to find the occurrence, the * latter for highlighting it. */ makeSearchSummary : function(text, keywords, hlwords) { diff --git a/doc/_static/websupport.js b/doc/_static/websupport.js index ffd9b2bfdc..98e7f40b63 100644 --- a/doc/_static/websupport.js +++ b/doc/_static/websupport.js @@ -2,7 +2,7 @@ * websupport.js * ~~~~~~~~~~~~~ * - * sphinx.websupport utilties for all documentation. + * sphinx.websupport utilities for all documentation. * * :copyright: Copyright 2007-2016 by the Sphinx team, see AUTHORS. * :license: BSD, see LICENSE for details. diff --git a/doc/algorithm/rnn/rnn.html b/doc/algorithm/rnn/rnn.html index 346fd39323..5452b62487 100644 --- a/doc/algorithm/rnn/rnn.html +++ b/doc/algorithm/rnn/rnn.html @@ -6,7 +6,7 @@ - Recurrent Neural Network Configuration — PaddlePaddle documentation + Recurrent Neural Network Configuration — PaddlePaddle documentation @@ -40,7 +40,7 @@
  • previous |
    • -
  • PaddlePaddle documentation »
    • +
  • PaddlePaddle documentation »
    • @@ -71,7 +71,7 @@ 
    yield src_ids, trg_ids, trg_ids_next
    -

    For more details description of how to write a data provider, please refer to Python Data Provider. The full data provider file is located at demo/seqToseq/dataprovider.py.

    +

    For more details description of how to write a data provider, please refer to PyDataProvider2. The full data provider file is located at demo/seqToseq/dataprovider.py.

    Configure Recurrent Neural Network Architecture

    @@ -121,7 +121,7 @@ Its output function simply takes \(x_t\)We will use the sequence to sequence model with attention as an example to demonstrate how you can configure complex recurrent neural network models. An illustration of the sequence to sequence model with attention is shown in the following figure.

    ../../_images/encoder-decoder-attention-model.png

    In this model, the source sequence \(S = \{s_1, \dots, s_T\}\) is encoded with a bidirectional gated recurrent neural networks. The hidden states of the bidirectional gated recurrent neural network \(H_S = \{H_1, \dots, H_T\}\) is called encoder vector The decoder is a gated recurrent neural network. When decoding each token \(y_t\), the gated recurrent neural network generates a set of weights \(W_S^t = \{W_1^t, \dots, W_T^t\}\), which are used to compute a weighted sum of the encoder vector. The weighted sum of the encoder vector is utilized to condition the generation of the token \(y_t\).

    -

    The encoder part of the model is listed below. It calls grumemory to represent gated recurrent neural network. It is the recommended way of using recurrent neural network if the network architecture is simple, because it is faster than recurrent_group. We have implemented most of the commonly used recurrent neural network architectures, you can refer to Layers for more details.

    +

    The encoder part of the model is listed below. It calls grumemory to represent gated recurrent neural network. It is the recommended way of using recurrent neural network if the network architecture is simple, because it is faster than recurrent_group. We have implemented most of the commonly used recurrent neural network architectures, you can refer to Layers for more details.

    We also project the encoder vector to decoder_size dimensional space, get the first instance of the backward recurrent network, and project it to decoder_size dimensional space:

    # Define the data layer of the source sentence.
 src_word_id = data_layer(name='source_language_word', size=source_dict_dim)
@@ -250,7 +250,7 @@ Its output function simply takes \(x_t\)outputs(beam_gen)
    -

    Notice that this generation technique is only useful for decoder like generation process. If you are working on sequence tagging tasks, please refer to Semantic Role Labeling Demo for more details.

    +

    Notice that this generation technique is only useful for decoder like generation process. If you are working on sequence tagging tasks, please refer to Semantic Role Labeling Demo for more details.

    The full configuration file is located at demo/seqToseq/seqToseq_net.py.

    @@ -288,14 +288,11 @@ Its output function simply takes \(x_t\)

    Quick search

    - - +
    +
    -

    - Enter search terms or a module, class or function name. -

    @@ -314,12 +311,12 @@ Its output function simply takes \(x_t\) previous | -
  • PaddlePaddle documentation »
    • +
  • PaddlePaddle documentation »
    • \ No newline at end of file diff --git a/doc/build/build_from_source.html b/doc/build/build_from_source.html index db964de2f4..b5e41d9b08 100644 --- a/doc/build/build_from_source.html +++ b/doc/build/build_from_source.html @@ -6,7 +6,7 @@ - Build and Install — PaddlePaddle documentation + Build and Install — PaddlePaddle documentation @@ -45,8 +45,8 @@
  • previous |
    • -
  • PaddlePaddle documentation »
    • -
  • Build And Install PaddlePaddle »
    • +
  • PaddlePaddle documentation »
    • +
  • Build And Install PaddlePaddle »
    • @@ -98,11 +98,12 @@ 
    # necessary
 sudo apt-get update
-sudo apt-get install -y g++ make cmake build-essential libatlas-base-dev python python-pip libpython-dev m4 libprotobuf-dev protobuf-compiler python-protobuf python-numpy git 
+sudo apt-get install -y g++ make cmake build-essential libatlas-base-dev python python-pip libpython-dev m4 libprotobuf-dev protobuf-compiler python-protobuf python-numpy git
 # optional
 sudo apt-get install libgoogle-glog-dev
 sudo apt-get install libgflags-dev
 sudo apt-get install libgtest-dev
+sudo pip install wheel
 pushd /usr/src/gtest
 cmake .
 make
@@ -171,6 +172,8 @@ mkdir build
 cd build
 # you can add build option here, such as:    
 cmake -DWITH_GPU=ON -DWITH_DOC=OFF -DCMAKE_INSTALL_PREFIX=<path to install> ..
+# please use sudo make install, if you want
+# to install PaddlePaddle into the system
 make -j `nproc` && make install
 # PaddlePaddle installation path
 export PATH=<path to install>/bin:$PATH
@@ -224,14 +227,11 @@ make -j `nproc`
    @@ -253,13 +253,13 @@ make -j `nproc`
  • previous |
    • -
  • PaddlePaddle documentation »
    • -
  • Build And Install PaddlePaddle »
    • +
  • PaddlePaddle documentation »
    • +
  • Build And Install PaddlePaddle »
    • \ No newline at end of file diff --git a/doc/build/contribute_to_paddle.html b/doc/build/contribute_to_paddle.html index 9d03891e15..762171f337 100644 --- a/doc/build/contribute_to_paddle.html +++ b/doc/build/contribute_to_paddle.html @@ -6,7 +6,7 @@ - Contribute to PaddlePaddle — PaddlePaddle documentation + Contribute to PaddlePaddle — PaddlePaddle documentation @@ -45,8 +45,8 @@
  • previous |
    • -
  • PaddlePaddle documentation »
    • -
  • Build And Install PaddlePaddle »
    • +
  • PaddlePaddle documentation »
    • +
  • Build And Install PaddlePaddle »
    • @@ -81,7 +81,7 @@ It’s just that simple.

    Once you’ve created a fork, you can use your favorite git client to clone your repo or just head straight to the command line:

    # Clone your fork to your local machine
-git clone git@github.com:USERNAME/paddle.git
+git clone git@github.com:USERNAME/Paddle.git

    Then you can start to develop.

    @@ -106,7 +106,7 @@ To do this, you’ll need to add a remote at first:

    # see the current configured remote repository
 git remote -v
 # add upstream repository
-git remote add upstream https://github.com/paddle/paddle.git
+git remote add upstream https://github.com/baidu/Paddle.git
 # verify the new upstream
 git remote -v
    @@ -171,14 +171,11 @@ and click the pull request button.

    @@ -200,13 +197,13 @@ and click the pull request button.

  • previous |
    • -
  • PaddlePaddle documentation »
    • -
  • Build And Install PaddlePaddle »
    • +
  • PaddlePaddle documentation »
    • +
  • Build And Install PaddlePaddle »
    • \ No newline at end of file diff --git a/doc/build/docker_install.html b/doc/build/docker_install.html index d7ffc53034..8684281b30 100644 --- a/doc/build/docker_install.html +++ b/doc/build/docker_install.html @@ -6,7 +6,7 @@ - Docker installation guide — PaddlePaddle documentation + Docker installation guide — PaddlePaddle documentation @@ -45,8 +45,8 @@
  • previous |
    • -
  • PaddlePaddle documentation »
    • -
  • Build And Install PaddlePaddle »
    • +
  • PaddlePaddle documentation »
    • +
  • Build And Install PaddlePaddle »
    • @@ -65,12 +65,12 @@

    PaddlePaddle Docker images

    There are six Docker images:

      -
    • paddledev/paddle:latest-cpu: PaddlePaddle CPU binary image.
      • -
    • paddledev/paddle:latest-gpu: PaddlePaddle GPU binary image.
      • -
    • paddledev/paddle:latest-cpu-devel: PaddlePaddle CPU binary image plus source code.
      • -
    • paddledev/paddle:latest-gpu-devel: PaddlePaddle GPU binary image plus source code.
      • -
    • paddledev/paddle:latest-cpu-demo: PaddlePaddle CPU binary image plus source code and demo
      • -
    • paddledev/paddle:latest-gpu-demo: PaddlePaddle GPU binary image plus source code and demo
      • +
    • paddledev/paddle:cpu-latest: PaddlePaddle CPU binary image.
      • +
    • paddledev/paddle:gpu-latest: PaddlePaddle GPU binary image.
      • +
    • paddledev/paddle:cpu-devel-latest: PaddlePaddle CPU binary image plus source code.
      • +
    • paddledev/paddle:gpu-devel-latest: PaddlePaddle GPU binary image plus source code.
      • +
    • paddledev/paddle:cpu-demo-latest: PaddlePaddle CPU binary image plus source code and demo
      • +
    • paddledev/paddle:gpu-demo-latest: PaddlePaddle GPU binary image plus source code and demo

    Tags with latest will be replaced by a released version.

    @@ -78,13 +78,13 @@

    Download and Run Docker images

    You have to install Docker in your machine which has linux kernel version 3.10+ first. You can refer to the official guide https://docs.docker.com/engine/installation/ for further information.

    You can use docker pullto download images first, or just launch a container with docker run:

    -
    docker run -it paddledev/paddle:lastest-cpu
+
    docker run -it paddledev/paddle:cpu-latest
 
    
 
    
 
    If you want to launch container with GPU support, you need to set some environment variables at the same time:
    
 
    export CUDA_SO="$(\ls /usr/lib64/libcuda* | xargs -I{} echo '-v {}:{}') $(\ls /usr/lib64/libnvidia* | xargs -I{} echo '-v {}:{}"
 export DEVICES=$(\ls /dev/nvidia* | xargs -I{} echo '--device {}:{}')
-docker run -it paddledev/paddle:latest-gpu
+docker run -it paddledev/paddle:gpu-latest
    @@ -178,14 +178,11 @@ docker rm paddle_ssh_machine
    @@ -207,13 +204,13 @@ docker rm paddle_ssh_machine
  • previous |
    • -
  • PaddlePaddle documentation »
    • -
  • Build And Install PaddlePaddle »
    • +
  • PaddlePaddle documentation »
    • +
  • Build And Install PaddlePaddle »
    • \ No newline at end of file diff --git a/doc/build/index.html b/doc/build/index.html index a9422691bb..ec0653dbd0 100644 --- a/doc/build/index.html +++ b/doc/build/index.html @@ -6,7 +6,7 @@ - Build And Install PaddlePaddle — PaddlePaddle documentation + Build And Install PaddlePaddle — PaddlePaddle documentation @@ -44,7 +44,7 @@
  • previous |
    • -
  • PaddlePaddle documentation »
    • +
  • PaddlePaddle documentation »
    • @@ -58,8 +58,6 @@

    Install PaddlePaddle

    -
      -
    @@ -100,23 +77,8 @@ state and your experience of installation may not be smooth.

    If you want to pack docker image, the following guide can help you:

    @@ -154,14 +116,11 @@ state and your experience of installation may not be smooth.

    @@ -183,12 +142,12 @@ state and your experience of installation may not be smooth.

  • previous |
    • -
  • PaddlePaddle documentation »
    • +
  • PaddlePaddle documentation »
    • \ No newline at end of file diff --git a/doc/build/ubuntu_install.html b/doc/build/ubuntu_install.html index 6786184240..e8e6c46663 100644 --- a/doc/build/ubuntu_install.html +++ b/doc/build/ubuntu_install.html @@ -6,7 +6,7 @@ - Debian Package installation guide — PaddlePaddle documentation + Debian Package installation guide — PaddlePaddle documentation @@ -45,8 +45,8 @@
  • previous |
    • -
  • PaddlePaddle documentation »
    • -
  • Build And Install PaddlePaddle »
    • +
  • PaddlePaddle documentation »
    • +
  • Build And Install PaddlePaddle »
    • @@ -103,14 +103,11 @@ apt-get install -f @@ -132,13 +129,13 @@ apt-get install -f
  • previous |
    • -
  • PaddlePaddle documentation »
    • -
  • Build And Install PaddlePaddle »
    • +
  • PaddlePaddle documentation »
    • +
  • Build And Install PaddlePaddle »
    • \ No newline at end of file diff --git a/doc/cluster/index.html b/doc/cluster/index.html index c470da5930..a698d0dc69 100644 --- a/doc/cluster/index.html +++ b/doc/cluster/index.html @@ -6,7 +6,7 @@ - Cluster Train — PaddlePaddle documentation + Cluster Train — PaddlePaddle documentation @@ -44,7 +44,7 @@
  • previous |
    • -
  • PaddlePaddle documentation »
    • +
  • PaddlePaddle documentation »
    • @@ -95,14 +95,11 @@ @@ -124,12 +121,12 @@
  • previous |
    • -
  • PaddlePaddle documentation »
    • +
  • PaddlePaddle documentation »
    • \ No newline at end of file diff --git a/doc/cluster/opensource/cluster_train.html b/doc/cluster/opensource/cluster_train.html index 2119e23a0a..63663fae18 100644 --- a/doc/cluster/opensource/cluster_train.html +++ b/doc/cluster/opensource/cluster_train.html @@ -6,7 +6,7 @@ - Cluster Training — PaddlePaddle documentation + Cluster Training — PaddlePaddle documentation @@ -45,8 +45,8 @@
  • previous |
    • -
  • PaddlePaddle documentation »
    • -
  • Cluster Train »
    • +
  • PaddlePaddle documentation »
    • +
  • Cluster Train »
    • @@ -77,7 +77,7 @@

    Following steps are based on demo/recommendation demo in demo directory.

    You just go through demo/recommendation tutorial doc until Train section, and at last you will get train/test data and model configuration file. Finaly, just use demo/recommendation as workspace for cluster training.

    At last your workspace should look like as follow:

    -
    .
+
    .
 |-- common_utils.py
 |-- data
 |   |-- config.json
@@ -158,7 +158,7 @@ all files in data directory are refered by train.list/test.list which are refere
 job_workspace  set it with already deployed workspace directory, paddle.py will skip dispatch stage to directly launch cluster job with all nodes. It could help to reduce heavy
 dispatch latency.
    
 
    cluster_train/run.sh provides command line sample to run demo/recommendation cluster job, just modify job_dispatch_package and job_workspace with your defined directory, then:
    
-
    sh run.sh
+
    sh run.sh
 
    
 
    
 
    The cluster Job will start in several seconds.
    
@@ -225,14 +225,11 @@ It provides stderr and stdout of trainer process. Check error log if training cr
 
 
         
    
@@ -254,13 +251,13 @@ It provides stderr and stdout of trainer process. Check error log if training cr
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Cluster Train »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Cluster Train »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/demo/embedding_model/index.html b/doc/demo/embedding_model/index.html
index 9db2174d60..bc78dc9540 100644
--- a/doc/demo/embedding_model/index.html
+++ b/doc/demo/embedding_model/index.html
@@ -6,7 +6,7 @@
   
     
     
-    Chinese Word Embedding Model Tutorial — PaddlePaddle  documentation
+    Chinese Word Embedding Model Tutorial — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Examples and demos »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Examples and demos »
    •  
       
     
      
 
@@ -80,7 +80,7 @@
 
    
 
    Download and Extract
    
 
    To download and extract our dictionary and pretrained model, run the following commands.
    
-
    cd $PADDLE_ROOT/demo/model_zoo/embedding
+
    cd $PADDLE_ROOT/demo/model_zoo/embedding
 ./pre_DictAndModel.sh
 
    
 
    
@@ -92,12 +92,12 @@
 
    
 
    Data Preparation and Preprocess
    
 
    First, run the following commands to download and extract the in-house dataset. The dataset (using UTF-8 format) has 20 training samples, 5 testing samples and 2 generating samples.
    
-
    cd $PADDLE_ROOT/demo/seqToseq/data
+
    cd $PADDLE_ROOT/demo/seqToseq/data
 ./paraphrase_data.sh
 
    
 
    
 
    Second, preprocess data and build dictionary on train data by running the following commands, and the preprocessed dataset is stored in $PADDLE_SOURCE_ROOT/demo/seqToseq/data/pre-paraphrase:
    
-
    cd $PADDLE_ROOT/demo/seqToseq/
+
    cd $PADDLE_ROOT/demo/seqToseq/
 python preprocess.py -i data/paraphrase [--mergeDict]
 
    
 
    
@@ -108,7 +108,7 @@ python preprocess.py -i data/paraphrase [--mergeDict]
 
    
 
    User Specified Embedding Model
    
 
    The general command of extracting desired parameters from the pretrained embedding model based on user dictionary is:
    
-
    cd $PADDLE_ROOT/demo/model_zoo/embedding
+
    cd $PADDLE_ROOT/demo/model_zoo/embedding
 python extract_para.py --preModel PREMODEL --preDict PREDICT --usrModel USRMODEL--usrDict USRDICT -d DIM
 
    
 
    
@@ -120,22 +120,22 @@ python extract_para.py --preModel PREMODEL --preDict PREDICT --usrModel USRMODEL
 
  • -d DIM: dimension of parameter
    • 
 
 
    Here, you can simply run the command:
    
-
    cd $PADDLE_ROOT/demo/seqToseq/data/
+
    cd $PADDLE_ROOT/demo/seqToseq/data/
 ./paraphrase_model.sh
 
    
 
    
 
    And you will see following embedding model structure:
    
-
    paraphrase_model
-|--- _source_language_embedding
-|--- _target_language_embedding
+
    paraphrase_model
+|--- _source_language_embedding
+|--- _target_language_embedding
 
    
 
    
 
    
 
    
 
    Training Model in PaddlePaddle
    
 
    First, create a model config file, see example demo/seqToseq/paraphrase/train.conf:
    
-
    from seqToseq_net import *
-is_generating = False
+
    from seqToseq_net import *
+is_generating = False
 
 ################## Data Definition #####################
 train_conf = seq_to_seq_data(data_dir = "./data/pre-paraphrase",
@@ -145,7 +145,7 @@ python extract_para.py --preModel PREMODEL --preDict PREDICT --usrModel USRMODEL
 settings(
       learning_method = AdamOptimizer(),
       batch_size = 50,
-      learning_rate = 5e-4)
+      learning_rate = 5e-4)
 
 ################# Network configure #####################
 gru_encoder_decoder(train_conf, is_generating, word_vector_dim = 32)
@@ -153,7 +153,7 @@ python extract_para.py --preModel PREMODEL --preDict PREDICT --usrModel USRMODEL
 
    
 
    This config is almost the same as demo/seqToseq/translation/train.conf.
    
 
    Then, train the model by running the command:
    
-
    cd $PADDLE_SOURCE_ROOT/demo/seqToseq/paraphrase
+
    cd $PADDLE_SOURCE_ROOT/demo/seqToseq/paraphrase
 ./train.sh
 
    
 
    
@@ -162,7 +162,7 @@ python extract_para.py --preModel PREMODEL --preDict PREDICT --usrModel USRMODEL
 
  • --init_model_path: path of the initialization model, here is data/paraphrase_model
    • 
 
  • --load_missing_parameter_strategy: operations when model file is missing, here use a normal distibution to initialize the other parameters except for the embedding layer
    • 
 
-
    For users who want to understand the dataset format, model architecture and training procedure in detail, please refer to Text generation Tutorial.
    
+
    For users who want to understand the dataset format, model architecture and training procedure in detail, please refer to Text generation Tutorial.
    
 
    
 
    
 
    
@@ -170,7 +170,7 @@ python extract_para.py --preModel PREMODEL --preDict PREDICT --usrModel USRMODEL
 
    
 
    Embedding Parameters Observation
    
 
    For users who want to observe the embedding parameters, this function can convert a PaddlePaddle binary embedding model to a text model by running the command:
    
-
    cd $PADDLE_ROOT/demo/model_zoo/embedding
+
    cd $PADDLE_ROOT/demo/model_zoo/embedding
 python paraconvert.py --b2t -i INPUT -o OUTPUT -d DIM
 
    
 
    
@@ -180,7 +180,7 @@ python paraconvert.py --b2t -i INPUT -o OUTPUT -d DIM
 
  • -d DIM: the dimension of parameter
    • 
 
 
    You will see parameters like this in output text model:
    
-
    0,4,32156096
+
    0,4,32156096
 -0.7845433,1.1937413,-0.1704215,0.4154715,0.9566584,-0.5558153,-0.2503305, ......
 0.0000909,0.0009465,-0.0008813,-0.0008428,0.0007879,0.0000183,0.0001984, ......
 ......
@@ -203,7 +203,7 @@ python paraconvert.py --b2t -i INPUT -o OUTPUT -d DIM
 
    
 
    Embedding Parameters Revision
    
 
    For users who want to revise the embedding parameters, this function can convert a revised text embedding model to a PaddlePaddle binary model by running the command:
    
-
    cd $PADDLE_ROOT/demo/model_zoo/embedding
+
    cd $PADDLE_ROOT/demo/model_zoo/embedding
 python paraconvert.py --t2b -i INPUT -o OUTPUT
 
    
 
    
@@ -212,7 +212,7 @@ python paraconvert.py --t2b -i INPUT -o OUTPUT
 
  • -o OUTPUT: the name of output binary embedding model
    • 
 
 
    Note that the format of input text model is as follows:
    
-
    -0.7845433,1.1937413,-0.1704215,0.4154715,0.9566584,-0.5558153,-0.2503305, ......
+
    -0.7845433,1.1937413,-0.1704215,0.4154715,0.9566584,-0.5558153,-0.2503305, ......
 0.0000909,0.0009465,-0.0008813,-0.0008428,0.0007879,0.0000183,0.0001984, ......
 ......
 
    
@@ -271,14 +271,11 @@ python paraconvert.py --t2b -i INPUT -o OUTPUT
 
 
         
    
@@ -300,13 +297,13 @@ python paraconvert.py --t2b -i INPUT -o OUTPUT
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Examples and demos »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Examples and demos »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/demo/image_classification/image_classification.html b/doc/demo/image_classification/image_classification.html
index 92da4c3ac6..5908b631b8 100644
--- a/doc/demo/image_classification/image_classification.html
+++ b/doc/demo/image_classification/image_classification.html
@@ -6,7 +6,7 @@
   
     
     
-    Image Classification Tutorial — PaddlePaddle  documentation
+    Image Classification Tutorial — PaddlePaddle  documentation
     
     
     
@@ -45,9 +45,9 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Examples and demos »
    • 
-          
  • Image Classification Tutorial »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Examples and demos »
    • 
+          
  • Image Classification Tutorial »
    •  
       
     
      
 
@@ -86,7 +86,7 @@ sh download_cifar.sh
 
    Here are the classes in the dataset, as well as 10 random images from each:
 
    Image Classification
    
 
    After downloading and converting, we should find a directory (cifar-out) containing the dataset in the following format:
    
-
    train
+
    train
 ---airplane
 ---automobile
 ---bird
@@ -115,8 +115,8 @@ sh download_cifar.sh
 
    
 
    Preprocess
    
 
    After the data has been downloaded, it needs to be pre-processed into the Paddle format. We can run the following command for preprocessing.
    
-
    cd demo/image_classification/
-sh preprocess.sh
+
    cd demo/image_classification/
+sh preprocess.sh
 
    
 
    
 
    preprocess.sh calls ./demo/image_classification/preprocess.py to preprocess image data.
    
@@ -220,11 +220,11 @@ python -m paddle.utils.plotcurve -i $log > plot.png
 
    Prediction
    
 
    After we train the model, the model file as well as the model parameters are stored in path ./cifar_vgg_model/pass-%05d. For example, the model of the 300-th pass is stored at ./cifar_vgg_model/pass-00299.
    
 
    To make a prediction for an image, one can run predict.sh as follows. The script will output the label of the classfiication.
    
-
    sh predict.sh
+
    sh predict.sh
 
    
 
    
 
    predict.sh:
    
-
    model=cifar_vgg_model/pass-00299/
+
    model=cifar_vgg_model/pass-00299/
 image=data/cifar-out/test/airplane/seaplane_s_000978.png
 use_gpu=1
 python prediction.py $model $image $use_gpu
@@ -292,14 +292,11 @@ python prediction.py $model $image $use_gpu
 
 
         
    
@@ -321,14 +318,14 @@ python prediction.py $model $image $use_gpu
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Examples and demos »
    • 
-          
  • Image Classification Tutorial »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Examples and demos »
    • 
+          
  • Image Classification Tutorial »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/demo/image_classification/index.html b/doc/demo/image_classification/index.html
index f0043780f8..93160d8d9b 100644
--- a/doc/demo/image_classification/index.html
+++ b/doc/demo/image_classification/index.html
@@ -6,7 +6,7 @@
   
     
     
-    Image Classification Tutorial — PaddlePaddle  documentation
+    Image Classification Tutorial — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Examples and demos »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Examples and demos »
    •  
       
     
      
 
@@ -97,14 +97,11 @@
 
 
         
    
@@ -126,13 +123,13 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Examples and demos »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Examples and demos »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/demo/imagenet_model/resnet_model.html b/doc/demo/imagenet_model/resnet_model.html
index a438ba9b55..d6ee31fa9d 100644
--- a/doc/demo/imagenet_model/resnet_model.html
+++ b/doc/demo/imagenet_model/resnet_model.html
@@ -6,7 +6,7 @@
   
     
     
-    Model Zoo - ImageNet — PaddlePaddle  documentation
+    Model Zoo - ImageNet — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Examples and demos »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Examples and demos »
    •  
       
     
      
 
@@ -101,19 +101,19 @@
 
    
 
    Network Visualization
    
 
    You can get a diagram of ResNet network by running the following commands. The script generates dot file and then converts dot file to PNG file, which uses installed draw_dot tool in our server. If you can not access the server, just install graphviz to convert dot file.
    
-
    cd demo/model_zoo/resnet
-./net_diagram.sh
+
    cd demo/model_zoo/resnet
+./net_diagram.sh
 
    
 
    
 
    
 
    
 
    Model Download
    
-
    cd demo/model_zoo/resnet
-./get_model.sh
+
    cd demo/model_zoo/resnet
+./get_model.sh
 
    
 
    
 
    You can run above command to download all models and mean file and save them in demo/model_zoo/resnet/model if downloading successfully.
    
-
    mean_meta_224  resnet_101  resnet_152  resnet_50
+
    mean_meta_224  resnet_101  resnet_152  resnet_50
 
    
 
    
 
      
@@ -183,8 +183,8 @@ shape: (Co, 
 
      Parameter Observation
      
 
      Users who want to observe the parameters can use python to read:
      
-
      import sys
-import numpy as np
+
      import sys
+import numpy as np
 
 def load(file_name):
     with open(file_name, 'rb') as f:
@@ -196,7 +196,7 @@ shape: (Co, 
      
 
      
 
      or simply use following shell command:
      
-
      od -j 16 -f _res2_1_branch1_bn.w0
+
      od -j 16 -f _res2_1_branch1_bn.w0
 
      
 
      
 
      
@@ -206,8 +206,8 @@ shape: (Co, We provide both C++ and Python interfaces to extract features. The following examples use data in demo/model_zoo/resnet/example to show the extracting process in detail.
      
 
      
 
      C++ Interface
      
-
      First, specify image data list in define_py_data_sources in the config, see example demo/model_zoo/resnet/resnet.py.
      
-
          train_list = 'train.list' if not is_test else None
+
      First, specify image data list in define_py_data_sources2 in the config, see example demo/model_zoo/resnet/resnet.py.
      
+
          train_list = 'train.list' if not is_test else None
     # mean.meta is mean file of ImageNet dataset.
     # mean.meta size : 3 x 224 x 224.
     # If you use three mean value, set like:
@@ -215,7 +215,7 @@ shape: (Co, args={
         'mean_meta': "model/mean_meta_224/mean.meta",
         'image_size': 224, 'crop_size': 224,
-        'color': True,'swap_channel:': [2, 1, 0]}
+        'color': True,'swap_channel:': [2, 1, 0]}
     define_py_data_sources2(train_list,
                            'example/test.list',
                            module="example.image_list_provider",
@@ -224,17 +224,17 @@ shape: (Co, 
      
 
      
 
      Second, specify layers to extract features in Outputs() of resnet.py. For example,
      
-
      Outputs("res5_3_branch2c_conv", "res5_3_branch2c_bn")
+
      Outputs("res5_3_branch2c_conv", "res5_3_branch2c_bn")
 
      
 
      
 
      Third, specify model path and output directory in extract_fea_c++.sh, and then run the following commands.
      
-
      cd demo/model_zoo/resnet
-./extract_fea_c++.sh
+
      cd demo/model_zoo/resnet
+./extract_fea_c++.sh
 
      
 
      
 
      If successful, features are saved in fea_output/rank-00000 as follows. And you can use load_feature_c interface in load_feature.py to load such a file.
      
-
      -0.115318 -0.108358 ... -0.087884;-1.27664 ... -1.11516 -2.59123;
--0.126383 -0.116248 ... -0.00534909;-1.42593 ... -1.04501 -1.40769;
+
      -0.115318 -0.108358 ... -0.087884;-1.27664 ... -1.11516 -2.59123;
+-0.126383 -0.116248 ... -0.00534909;-1.42593 ... -1.04501 -1.40769;
 
      
 
      
 
        
@@ -245,20 +245,20 @@ shape: (Co, 
 
        Python Interface
        
 
        demo/model_zoo/resnet/classify.py is an example to show how to use python to extract features. Following example still uses data of ./example/test.list. Command is as follows:
        
-
        cd demo/model_zoo/resnet
-./extract_fea_py.sh
+
        cd demo/model_zoo/resnet
+./extract_fea_py.sh
 
        
 
        
 
        extract_fea_py.sh:
        
-
        python classify.py \
-     --job=extract \
-     --conf=resnet.py\
-     --use_gpu=1 \
-     --mean=model/mean_meta_224/mean.meta \
-     --model=model/resnet_50 \
-     --data=./example/test.list \
-     --output_layer="res5_3_branch2c_conv,res5_3_branch2c_bn" \
-     --output_dir=features
+
        python classify.py \
+     --job=extract \
+     --conf=resnet.py\
+     --use_gpu=1 \
+     --mean=model/mean_meta_224/mean.meta \
+     --model=model/resnet_50 \
+     --data=./example/test.list \
+     --output_layer="res5_3_branch2c_conv,res5_3_branch2c_bn" \
+     --output_dir=features
 
        
 
        
 
          
@@ -272,7 +272,7 @@ shape: (Co, 
 
          Note, since the convolution layer in these ResNet models is suitable for the cudnn implementation which only support GPU. It not support CPU mode because of compatibility issue and we will fix later.
          
 
          If run successfully, you will see features saved in features/batch_0, this file is produced with cPickle. You can use load_feature_py interface in load_feature.py to open the file, and it returns a dictionary as follows:
          
-
          {
+
          {
 'cat.jpg': {'res5_3_branch2c_conv': array([[-0.12638293, -0.116248  , -0.11883899, ..., -0.00895038, 0.01994277, -0.00534909]], dtype=float32), 'res5_3_branch2c_bn': array([[-1.42593431, -1.28918779, -1.32414699, ..., -1.45933616, -1.04501402, -1.40769434]], dtype=float32)},
 'dog.jpg': {'res5_3_branch2c_conv': array([[-0.11531784, -0.10835785, -0.08809858, ...,0.0055237, 0.01505112, -0.08788397]], dtype=float32), 'res5_3_branch2c_bn': array([[-1.27663755, -1.18272924, -0.90937918, ..., -1.25178063, -1.11515927, -2.59122872]], dtype=float32)}
 }
@@ -284,18 +284,18 @@ shape: (Co, 
 
          Prediction
          
 
          classify.py also can be used to predict. We provide an example script predict.sh to predict data in example/test.list using a ResNet model with 50 layers.
          
-
          cd demo/model_zoo/resnet
-./predict.sh
+
          cd demo/model_zoo/resnet
+./predict.sh
 
          
 
          
 
          predict.sh calls the classify.py:
          
-
          python classify.py \
-     --job=predict \
-     --conf=resnet.py\
-     --multi_crop \
-     --model=model/resnet_50 \
-     --use_gpu=1 \
-     --data=./example/test.list
+
          python classify.py \
+     --job=predict \
+     --conf=resnet.py\
+     --multi_crop \
+     --model=model/resnet_50 \
+     --use_gpu=1 \
+     --data=./example/test.list
 
          
 
          
 
            
@@ -307,8 +307,8 @@ shape: (Co, --data=./example/test.list: data list.
 
          
 
          If run successfully, you will see following results, where 156 and 285 are labels of the images.
          
-
          Label of example/dog.jpg is: 156
-Label of example/cat.jpg is: 282
+
          Label of example/dog.jpg is: 156
+Label of example/cat.jpg is: 282
 
          
 
          
 
          
@@ -357,14 +357,11 @@ Label of example/cat.jpg is: 282
 
 
         
          
@@ -386,13 +383,13 @@ Label of example/cat.jpg is: 282
         
        • 
           previous |
          • 
-        
        • PaddlePaddle  documentation »
          • 
-          
        • Examples and demos »
          •  
+        
        • PaddlePaddle  documentation »
          • 
+          
        • Examples and demos »
          •  
       
        
     
        
     
   
 
\ No newline at end of file
diff --git a/doc/demo/index.html b/doc/demo/index.html
index 4068f4db5b..058cc4a91c 100644
--- a/doc/demo/index.html
+++ b/doc/demo/index.html
@@ -6,7 +6,7 @@
   
     
     
-    Examples and demos — PaddlePaddle  documentation
+    Examples and demos — PaddlePaddle  documentation
     
     
     
@@ -44,7 +44,7 @@
         
      • 
           previous |
        • 
-        
      • PaddlePaddle  documentation »
        •  
+        
      • PaddlePaddle  documentation »
        •  
       
      
     
        
 
@@ -70,6 +70,7 @@
 
 
      
 
      
@@ -126,14 +127,11 @@
 
 
         
      
@@ -155,12 +153,12 @@
         
    • 
           previous |
      • 
-        
    • PaddlePaddle  documentation »
      •  
+        
    • PaddlePaddle  documentation »
      •  
       
    
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/demo/quick_start/index_en.html b/doc/demo/quick_start/index_en.html
index 4bda3430a8..a0c6beb71f 100644
--- a/doc/demo/quick_start/index_en.html
+++ b/doc/demo/quick_start/index_en.html
@@ -6,7 +6,7 @@
   
     
     
-    Quick Start Tutorial — PaddlePaddle  documentation
+    Quick Start Tutorial — PaddlePaddle  documentation
     
     
     
@@ -44,7 +44,7 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
      
 
@@ -72,11 +72,11 @@
 
    Overview
    
 
    For the first step, you will use PaddlePaddle to build a text classification system. For example, suppose you run an e-commence  website, and you want to analyze the sentiment of user reviews to evaluate product quality.
    
 
    For example, given the input
    
-
    This monitor is fantastic.
+
    This monitor is fantastic.
 
    
 
    
 
    Your classifier should output “positive”, since this text snippet shows that the user is satisfied with the product. Given this input:
    
-
    The monitor breaks down two months after purchase.
+
    The monitor breaks down two months after purchase.
 
    
 
    
 
    the classifier should output “negative“.
    
@@ -114,7 +114,7 @@
 
    
 
    Preprocess data into standardized format
    
 
    In this example, you are going to use Amazon electronic product review dataset to build a bunch of deep neural network models for text classification. Each text in this dataset is a product review. This dataset has two categories: “positive” and “negative”. Positive means the reviewer likes the product, while negative means the reviewer does not like the product.
    
-
    demo/quick_start provides scripts for downloading data and preprocessing data as shown below. The data process takes several minutes (about 3 minutes in our machine).
    
+
    demo/quick_start in the source code provides scripts for downloading data and preprocessing data as shown below. The data process takes several minutes (about 3 minutes in our machine).
    
 
    cd demo/quick_start
 ./data/get_data.sh
 ./preprocess.sh
@@ -214,8 +214,7 @@
                         args={"dictionary": word_dict})
 
    
 
    
-
    You can refer to the following link for more detailed examples
-: Python Use Case，The detailed documentation on data format is:  PyDataProviderWrapper。
    
+
    You can refer to the following link for more detailed examples and data formats: PyDataProvider2.
    
 
    
 
    
 
    
@@ -443,7 +442,7 @@ paddle train \
 mv rank-00000 result.txt
 
    
 
    
-
    There are several differences between training and inference network configurations.
    
+
    User can choose the best model base on the training log instead of model output/pass-00003. There are several differences between training and inference network configurations.
    
 
      
 
    • You do not need labels during inference.
      • 
 
    • Outputs need to be specified to the classification probability layer (the output of softmax layer), or the id of maximum probability (max_id layer). An example to output the id and probability is given in the code snippet.
      • 
@@ -451,8 +450,8 @@ mv rank-00000 result.txt
 
    • You need to specify the location of test_list in the test data.
      • 
 
    
 
    The results in result.txt is as follows, each line is one sample.
    
-
    predicted_label_id;probability_of_label_0 probability_of_label_1  # the first sample
-predicted_label_id;probability_of_label_0 probability_of_label_1  # the second sample
+
    predicted_label_id;probability_of_label_0 probability_of_label_1  # the first sample
+predicted_label_id;probability_of_label_0 probability_of_label_1  # the second sample
 
    
 
    
 
    is_predict = get_config_arg('is_predict', bool, False)
@@ -520,7 +519,7 @@ predicted_label_id;probability_of_label_0 probability_of_label_1  # the second s
 
    
 
    
 
    Log
    
-
    TrainerInternal.cpp:160]  Batch=20 samples=2560 AvgCost=0.628761 CurrentCost=0.628761 Eval: classification_error_evaluator=0.304297  CurrentEval: classification_error_evaluator=0.304297
+
    TrainerInternal.cpp:160]  Batch=20 samples=2560 AvgCost=0.628761 CurrentCost=0.628761 Eval: classification_error_evaluator=0.304297  CurrentEval: classification_error_evaluator=0.304297
 
    
 
    
 
    During model training, you will see the log like the examples above:
@@ -607,14 +606,11 @@ predicted_label_id;probability_of_label_0 probability_of_label_1  # the second s
 
    
 
         
    
@@ -636,12 +632,12 @@ predicted_label_id;probability_of_label_0 probability_of_label_1  # the second s
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/demo/rec/ml_dataset.html b/doc/demo/rec/ml_dataset.html
index 26fe2104d7..7f0c567559 100644
--- a/doc/demo/rec/ml_dataset.html
+++ b/doc/demo/rec/ml_dataset.html
@@ -6,7 +6,7 @@
   
     
     
-    MovieLens Dataset — PaddlePaddle  documentation
+    MovieLens Dataset — PaddlePaddle  documentation
     
     
     
@@ -27,7 +27,7 @@
     
     
     
-     
+     
   
   
       
 
@@ -183,8 +183,8 @@ entries and/or test entries
 
 
   
    Previous topic
    
-  
    Text generation Tutorial
    
+  
    Semantic Role labeling Tutorial
    
   
    Next topic
    
   
    Regression MovieLens Ratting
    
@@ -198,14 +198,11 @@ entries and/or test entries
 
 
         
    
@@ -225,15 +222,15 @@ entries and/or test entries
           next |
         
  • 
-          previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Examples and demos »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Examples and demos »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/demo/rec/ml_regression.html b/doc/demo/rec/ml_regression.html
index 3661f86aac..246fea5e1c 100644
--- a/doc/demo/rec/ml_regression.html
+++ b/doc/demo/rec/ml_regression.html
@@ -6,7 +6,7 @@
   
     
     
-    Regression MovieLens Ratting — PaddlePaddle  documentation
+    Regression MovieLens Ratting — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Examples and demos »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Examples and demos »
    •  
       
     
      
 
@@ -91,7 +91,7 @@ i.e, specific WHAT type it is in each feature file.
    
 
    The field config file of ml-1m shows in demo/recommendation/data/config.json.
 It specifics the field types and file names: 1) there are four types of field for user file: id, gender, age and occupation;
 2) the filename is “users.dat”, and the delimiter of file is ”::”.
    
-
    {
+
    {
   "user": {
     "file": {
       "name": "users.dat",
@@ -141,7 +141,7 @@ python config_generator.py config.json > meta_config.json
 
    
 
    
 
    The meta config file shows below:
    
-
    {
+
    {
   "meta": {
     "movie": {
       "fields": [
@@ -381,21 +381,7 @@ cp ml-1m/ratings.dat.test .
 
    The network structure shows below.
    
 rec_regression_network
 
    The demo’s neural network config file “trainer_config.py” show as below.
    
-
    # Copyright (c) 2016 Baidu, Inc. All Rights Reserved
-#
-# 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.
-
-from paddle.trainer_config_helpers import *
+
    from paddle.trainer_config_helpers import *
 
 try:
     import cPickle as pickle
@@ -519,34 +505,14 @@ features.
    
 
  • Pooling Layer, pooling_layer
    • 
 
  • Cosine Similarity Layer, cos_sim
    • 
 
  • Text Convolution Pooling Layer, text_conv_pool
    • 
-
  • Declare Python Data Sources, define_py_data_sources
    • 
+
  • Declare Python Data Sources, define_py_data_sources2
    • 
 
 
    
 
    
 
    Data Provider
    
-
    # Copyright (c) 2016 Baidu, Inc. All Rights Reserved
-#
-# 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.
-
-try:
-    import cPickle as pickle
-except ImportError:
-    import pickle
-
-from paddle.trainer.PyDataProvider2 import *
+
    from paddle.trainer.PyDataProvider2 import *
 import common_utils  # parse
 
-
 def hook(settings, meta, **kwargs):
     """
     Init hook is invoked before process data. It will set obj.slots and store
@@ -573,7 +539,6 @@ features.
    
     settings.input_types = headers
     settings.meta = meta
 
-
 @provider(init_hook=hook, cache=CacheType.CACHE_PASS_IN_MEM)
 def process(settings, filename):
     with open(filename, 'r') as f:
@@ -614,42 +579,27 @@ In this dataprovider.pyuse_seq: Whether this dataprovider.py in sequence mode or not.
 
  • process: Return each sample of data to paddle.
    • 
 
-
    The data provider details document see there.
    
+
    The data provider details document see there.
    
 
    
 
    
 
    
 
    Train
    
 
    After prepare data, config network, writting data provider, now we can run paddle training.
    
 
    The run.sh is shown as follow:
    
-
    #!/bin/bash
-# Copyright (c) 2016 Baidu, Inc. All Rights Reserved
-#
-# 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.
-set -e
-paddle train \
-    --config=trainer_config.py \
-    --save_dir=./output \
-    --use_gpu=false \
-    --trainer_count=4\
-    --test_all_data_in_one_period=true \
-    --log_period=100 \
-    --dot_period=1 \
-    --num_passes=50  2>&1 | tee 'log.txt'
+
    paddle train \
+    --config=trainer_config.py \
+    --save_dir=./output \
+    --use_gpu=false \
+    --trainer_count=4\
+    --test_all_data_in_one_period=true \
+    --log_period=100 \
+    --dot_period=1 \
+    --num_passes=50  2>&1 | tee 'log.txt'
 
    
 
    
 
    It just start a paddle training process, write the log to log.txt,
 then print it on screen.
    
-
    Each command line argument in run.sh, please refer to the command line
+
    Each command line argument in run.sh, please refer to the command line
 arguments page. The short description of these arguments is shown as follow.
    
 
      
 
    • config: Tell paddle which file is neural network configuration.
      • 
@@ -759,14 +709,11 @@ Prediction Score is 3.13
 
 
         
    
@@ -788,13 +735,13 @@ Prediction Score is 3.13
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Examples and demos »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Examples and demos »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc_cn/build_and_install/install/index.html b/doc/demo/semantic_role_labeling/index.html
similarity index 50%
rename from doc_cn/build_and_install/install/index.html
rename to doc/demo/semantic_role_labeling/index.html
index 91f89b6b82..820c5f6045 100644
--- a/doc_cn/build_and_install/install/index.html
+++ b/doc/demo/semantic_role_labeling/index.html
@@ -6,7 +6,7 @@
   
     
     
-    安装PaddlePaddle — PaddlePaddle  documentation
+    Semantic Role Labeling Tutorial — PaddlePaddle  documentation
     
     
     
@@ -25,9 +25,9 @@
     
     
     
-    
-    
-     
+    
+    
+     
   
   
       
 
@@ -52,33 +55,23 @@
         
    
           
    
             
-  
    
-
    安装PaddlePaddle
    
-
    PaddlePaddle提供数个预编译的二进制来进行安装。他们包括Docker镜像，ubuntu的deb安装包等
-。欢迎贡献更多的安装包。我们更推荐使用Docker镜像来部署PaddlePaddle环境。
    
-
    Note: The intallation packages are still in pre-release
-state and your experience of installation may not be smooth.
    
-
    注意!目前PaddlePaddle的安装包还处在pre-release的状态，
-使用起来或许会不是很顺畅。
    
+  
    
+
    Semantic Role Labeling Tutorial
    
 
    
 
    
     
   
 
\ No newline at end of file
diff --git a/doc/demo/semantic_role_labeling/semantic_role_labeling.html b/doc/demo/semantic_role_labeling/semantic_role_labeling.html
new file mode 100644
index 0000000000..b79c5dc18c
--- /dev/null
+++ b/doc/demo/semantic_role_labeling/semantic_role_labeling.html
@@ -0,0 +1,317 @@
+
+
+
+
+  
+    
+    
+    Semantic Role labeling Tutorial — PaddlePaddle  documentation
+    
+    
+    
+    
+    
+    
+    
+    
+    
+    
+    
+    
+     
+  
+  
+      
+
+    
    
+      
    
+        
    
+          
    
+            
+  
    
+
    Semantic Role labeling Tutorial
    
+
    Semantic role labeling (SRL) is a form of shallow semantic parsing whose goal is to discover the predicate-argument structure of each predicate in a given input sentence. SRL is useful as an intermediate step in a wide range of natural language processing tasks, such as information extraction. automatic document categorization and question answering.  An instance is as following [1]:
    
+
    [ A0 He ] [ AM-MOD would ][ AM-NEG n’t ] [ V accept] [ A1 anything of value ] from [A2 those he was writing about ].
    
+
      
+
    • V: verb
      • 
+
    • A0: acceptor
      • 
+
    • A1: thing accepted
      • 
+
    • A2: accepted-from
      • 
+
    • A3: Attribute
      • 
+
    • AM-MOD: modal
      • 
+
    • AM-NEG: negation
      • 
+
    
+
    Given the verb “accept”, the chunks in sentence would play certain semantic roles. Here, the label scheme is from Penn Proposition Bank.
    
+
    To this date, most of the successful SRL systems are built on top of some form of parsing results where pre-defined feature templates over the syntactic structure are used. This tutorial will present an end-to-end system using deep bidirectional long short-term memory (DB-LSTM)[2] for solving the SRL task, which largely outperforms the previous state-of-the-art systems. The system regards SRL task as the sequence labelling problem.
    
+
    
+
    Data Description
    
+
    The relevant paper[2] takes the data set in CoNLL-2005&2012 Shared Task for training and testing. Accordingto data license,  the demo adopts the test data set of CoNLL-2005, which can be reached on website.
    
+
    To download and process the original data, user just need to execute the following command:
    
+
    cd data
+./get_data.sh
+
    
+
    
+
    Several new files appear in the datadirectory as follows.
    
+
    conll05st-release：the test data set of CoNll-2005 shared task 
+test.wsj.words：the Wall Street Journal data sentences
+test.wsj.props:  the propositional arguments
+src.dict：the dictionary of words in sentences
+tgt.dict：the labels dictionary
+feature: the extracted features from data set
+
    
+
    
+
    
+
    
+
    Training
    
+
    
+
    DB-LSTM
    
+
    Please refer to the Sentiment Analysis demo to learn more about the long short-term memory unit.
    
+
    Unlike Bidirectional-LSTM that used in Sentiment Analysis demo,  the DB-LSTM adopts another way to stack LSTM layer. First a standard LSTM processes the sequence in forward direction. The input and output of this LSTM layer are taken by the next LSTM layer as input, processed in reversed direction. These two standard LSTM layers compose a pair of LSTM. Then we stack LSTM layers pair after pair to obtain the deep LSTM model.
    
+
    The following figure shows a temporal expanded 2-layer DB-LSTM network.
+
    
+pic
+
    
+
    
+
    
+
    Features
    
+
    Two input features play an essential role in this pipeline: predicate (pred) and argument (argu). Two other features: predicate context (ctx-p) and region mark (mr) are also adopted. Because a single predicate word can not exactly describe the predicate information, especially when the same words appear more than one times in a sentence. With the predicate context, the ambiguity can be largely eliminated. Similarly, we use region mark mr = 1 to denote the argument position if it locates in the predicate context region, or mr = 0 if does not. These four simple features are all we need for our SRL system. Features of one sample with context size set to 1 is showed as following[2]:
+
    
+pic
+
    
+
    In this sample, the coresponding labelled sentence is:
    
+
    [ A1 A record date ] has [ AM-NEG n’t ] been [ V set ] .
    
+
    In the demo, we adopt the feature template as above, consists of :  argument, predicate, ctx-p (p=-1,0,1), mark and use B/I/O scheme to label each argument. These features and labels are stored in feature file, and separated by \t.
    
+
    
+
    
+
    Data Provider
    
+
    dataprovider.py is the python file to wrap data. hook() function is to define the data slots for network. The  Six features and label are all IndexSlots.
    
+
    def hook(settings, word_dict, label_dict, **kwargs):
+    settings.word_dict = word_dict
+    settings.label_dict = label_dict
+    #all inputs are integral and sequential type
+    settings.slots = [
+        integer_value_sequence(len(word_dict)),
+        integer_value_sequence(len(word_dict)),
+        integer_value_sequence(len(word_dict)),
+        integer_value_sequence(len(word_dict)),
+        integer_value_sequence(len(word_dict)),
+        integer_value_sequence(2),
+        integer_value_sequence(len(label_dict))]
+
    
+
    
+
    The corresponding data iterator is as following:
    
+
    @provider(use_seq=True, init_hook=hook)
+def process(obj, file_name):
+    with open(file_name, 'r') as fdata:
+        for line in fdata:
+            sentence, predicate, ctx_n1, ctx_0, ctx_p1, mark, label = line.strip().split('\t')
+            words = sentence.split()
+            sen_len = len(words)
+            word_slot = [obj.word_dict.get(w, UNK_IDX) for w in words]
+
+            predicate_slot = [obj.word_dict.get(predicate, UNK_IDX)] * sen_len
+            ctx_n1_slot = [obj.word_dict.get(ctx_n1, UNK_IDX) ] * sen_len
+            ctx_0_slot = [obj.word_dict.get(ctx_0, UNK_IDX) ] * sen_len
+            ctx_p1_slot = [obj.word_dict.get(ctx_p1, UNK_IDX) ] * sen_len
+
+            marks = mark.split()
+            mark_slot = [int(w) for w in marks]
+
+            label_list = label.split()
+            label_slot = [obj.label_dict.get(w) for w in label_list]
+
+            yield word_slot, predicate_slot, ctx_n1_slot, ctx_0_slot, ctx_p1_slot, mark_slot, label_slot
+
    
+
    
+
    The processfunction yield 7 lists which are six features and labels.
    
+
    
+
    
+
    Neural Network Config
    
+
    db_lstm.py is the neural network config file to load the dictionaries and define the  data provider module and network architecture during the training procedure.
    
+
    Seven data_layer load instances from data provider. Six features are transformed into embedddings respectively, and mixed by mixed_layer .  Deep bidirectional LSTM layers extract features for the softmax layer. The objective function is cross entropy of labels.
    
+
    
+
    
+
    Run Training
    
+
    The script for training is train.sh, user just need to execute:
    
+
      ./train.sh
+
    
+
    
+
    The content in train.sh:
    
+
    paddle train \
+  --config=./db_lstm.py \
+  --save_dir=./output \
+  --trainer_count=4 \
+  --log_period=10 \
+  --num_passes=500 \
+  --use_gpu=false \
+  --show_parameter_stats_period=10 \
+  --test_all_data_in_one_period=1 \
+2>&1 | tee 'train.log'
+
    
+
    
+
      
+
    • --config=./db_lstm.py : network config file.
      • 
+
    • --save_di=./output: output path to save models.
      • 
+
    • --trainer_count=4 : set thread number (or GPU count).
      • 
+
    • --log_period=10 : print log every 20 batches.
      • 
+
    • --num_passes=500: set pass number, one pass in PaddlePaddle means training all samples in dataset one time.
      • 
+
    • --use_gpu=false: use CPU to train, set true, if you install GPU version of PaddlePaddle and want to use GPU to train.
      • 
+
    • --show_parameter_stats_period=10: show parameter statistic every 100 batches.
      • 
+
    • --test_all_data_in_one_period=1: test all data in every testing.
      • 
+
    
+
    After training, the models  will be saved in directory output.
    
+
    
+
    
+
    Run testing
    
+
    The script for testing is test.sh, user just need to execute:
    
+
      ./test.sh
+
    
+
    
+
    The main part in tesh.sh
    
+
    paddle train \
+  --config=./db_lstm.py \
+  --model_list=$model_list \
+  --job=test \
+  --config_args=is_test=1 \
+
    
+
    
+
      
+
    • --config=./db_lstm.py: network config file
      • 
+
    • --model_list=$model_list.list: model list file
      • 
+
    • --job=test: indicate the test job
      • 
+
    • --config_args=is_test=1: flag to indicate test
      • 
+
    
+
    
+
    
+
    Run prediction
    
+
    The script for prediction is predict.sh, user just need to execute:
    
+
      ./predict.sh
+  
+
    
+
    
+
    In predict.sh, user should offer the network config file, model path, label file, word dictionary file, feature file
    
+
    python predict.py 
+     -c $config_file 
+     -w $model_path 
+     -l $label_file 
+     -d $dict_file 
+     -i $input_file
+
    
+
    
+
    predict.py is the main executable python script, which includes functions: load model, load data, data prediction. The network model will output the probability distribution of labels. In the demo, we take the label with maximum probability as result. User can also implement the beam search or viterbi decoding upon the probability distribution matrix.
    
+
    After prediction,  the result is saved in predict.res.
    
+
    
+
    
+
    
+
    Reference
    
+
    [1] Martha Palmer, Dan Gildea, and Paul Kingsbury. The Proposition Bank: An Annotated Corpus of Semantic Roles , Computational Linguistics, 31(1), 2005.
    
+
    [2] Zhou, Jie, and Wei Xu. “End-to-end learning of semantic role labeling using recurrent neural networks.” Proceedings of the Annual Meeting of the Association for Computational Linguistics. 2015.
    
+
    
+
    
+
+
+          
    
+        
    
+      
    
+      
+      
    
+    
    
+    
+    
+  
+
\ No newline at end of file
diff --git a/doc/demo/sentiment_analysis/index.html b/doc/demo/sentiment_analysis/index.html
index e99e4054b8..54f281d70b 100644
--- a/doc/demo/sentiment_analysis/index.html
+++ b/doc/demo/sentiment_analysis/index.html
@@ -6,7 +6,7 @@
   
     
     
-    Sentiment Analasis Tutorial — PaddlePaddle  documentation
+    Sentiment Analasis Tutorial — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Examples and demos »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Examples and demos »
    •  
       
     
      
 
@@ -102,14 +102,11 @@
 
 
         
    
@@ -131,13 +128,13 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Examples and demos »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Examples and demos »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/demo/sentiment_analysis/sentiment_analysis.html b/doc/demo/sentiment_analysis/sentiment_analysis.html
index e5e171f016..b065681ddc 100644
--- a/doc/demo/sentiment_analysis/sentiment_analysis.html
+++ b/doc/demo/sentiment_analysis/sentiment_analysis.html
@@ -6,7 +6,7 @@
   
     
     
-    Sentiment Analysis Tutorial — PaddlePaddle  documentation
+    Sentiment Analysis Tutorial — PaddlePaddle  documentation
     
     
     
@@ -45,9 +45,9 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Examples and demos »
    • 
-          
  • Sentiment Analasis Tutorial »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Examples and demos »
    • 
+          
  • Sentiment Analasis Tutorial »
    •  
       
     
      
 
@@ -67,12 +67,12 @@
 
    
 
    IMDB Data Introduction
    
 
    Before training models, we need to preprocess the data and build a dictionary. First, you can use following script to download IMDB dataset and Moses tool, which is a statistical machine translation system. We provide a data preprocessing script, which is capable of handling not only IMDB data, but also other user-defined data. In order to use the pre-written script, it needs to move labeled train and test samples to another path, which has been done in get_imdb.sh.
    
-
    cd demo/sentiment/data
-./get_imdb.sh
+
    cd demo/sentiment/data
+./get_imdb.sh
 
    
 
    
 
    If the data is obtained successfuly, you will see the following files at ./demo/sentiment/data:
    
-
    aclImdb  get_imdb.sh  imdb  mosesdecoder-master
+
    aclImdb  get_imdb.sh  imdb  mosesdecoder-master
 
    
 
    
 
      
@@ -81,7 +81,7 @@
 
    • mosesdecoder-master: Moses tool.
      • 
 
    
 
    IMDB dataset contains 25,000 highly polar movie reviews for training, and 25,000 for testing. A negative review has a score ≤ 4 out of 10, and a positive review has a score ≥ 7 out of 10. After running ./get_imdb.sh, we can find the dataset has the following structure in aclImdb.
    
-
    imdbEr.txt  imdb.vocab  README  test  train
+
    imdbEr.txt  imdb.vocab  README  test  train
 
    
 
    
 
      
@@ -92,7 +92,7 @@
 
    • README: data documentation.
      • 
 
    
 
    Both train and test set directory contains:
    
-
    labeledBow.feat  neg  pos  unsup  unsupBow.feat  urls_neg.txt  urls_pos.txt  urls_unsup.txt
+
    labeledBow.feat  neg  pos  unsup  unsupBow.feat  urls_neg.txt  urls_pos.txt  urls_unsup.txt
 
    
 
    
 
      
@@ -106,13 +106,13 @@
 
      
 
      IMDB Data Preparation
      
 
      In this demo, we only use labled train and test set and not use imdb.vocab as dictionary. By default, dictionary is builded on train set. Train set is shuffled and test set is not. tokenizer.perl in Moses tool is used to tokenize the words and punctuation. Simply execute the following command to preprcess data.
      
-
      cd demo/sentiment/
-./preprocess.sh
+
      cd demo/sentiment/
+./preprocess.sh
 
      
 
      
 
      preprocess.sh:
      
-
      data_dir="./data/imdb"
-python preprocess.py -i data_dir
+
      data_dir="./data/imdb"
+python preprocess.py -i data_dir
 
      
 
      
 
        
@@ -120,7 +120,7 @@ python preprocess.py -i data_dir
 
      • preprocess.py: preprocess script.
        • 
 
      
 
      If running successfully, you will see demo/sentiment/data/pre-imdb directory as follows:
      
-
      dict.txt  labels.list  test.list  test_part_000  train.list  train_part_000
+
      dict.txt  labels.list  test.list  test_part_000  train.list  train_part_000
 
      
 
      
 
        
@@ -133,19 +133,19 @@ python preprocess.py -i data_dir
 
        
 
        User-defined Data Preparation
        
 
        If you perform other sentiment classifcation task, you can prepare data as follows. We have provided the scripts to build dictionary and preprocess data. So just organize data as follows.
        
-
        dataset
-|----train
-|    |----class1
-|    |    |----text_files
-|    |----class2
-|    |    |----text_files
-|    |    ...
-|----test
-|    |----class1
-|    |    |----text_files
-|    |----class2
-|    |    |----text_files
-|    |    ...
+
        dataset
+|----train
+|    |----class1
+|    |    |----text_files
+|    |----class2
+|    |    |----text_files
+|    |    ...
+|----test
+|    |----class1
+|    |    |----text_files
+|    |----class2
+|    |    |----text_files
+|    |    ...
 
        
 
        
 
          
@@ -227,12 +227,12 @@ python preprocess.py -i data_dir
 
        
 
        Training
        
 
        Install PaddlePaddle first if necessary. Then you can use script train.sh as follows to launch local training.
        
-
        cd demo/sentiment/
-./train.sh
+
        cd demo/sentiment/
+./train.sh
 
        
 
        
 
        train.sh:
        
-
        config=trainer_config.py
+
        config=trainer_config.py
 output=./model_output
 paddle train --config=$config \
              --save_dir=$output \
@@ -259,10 +259,10 @@ paddle train --config=$config \
 
      • --test_all_data_in_one_period=1: test all data every testing.
        • 
 
      
 
      If the run succeeds, the output log is saved in path of demo/sentiment/train.log and model is saved in path of demo/sentiment/model_output/. The output log is explained as follows.
      
-
      Batch=20 samples=2560 AvgCost=0.681644 CurrentCost=0.681644 Eval: classification_error_evaluator=0.36875  CurrentEval: classification_error_evaluator=0.36875
-...
-Pass=0 Batch=196 samples=25000 AvgCost=0.418964 Eval: classification_error_evaluator=0.1922
-Test samples=24999 cost=0.39297 Eval: classification_error_evaluator=0.149406
+
      Batch=20 samples=2560 AvgCost=0.681644 CurrentCost=0.681644 Eval: classification_error_evaluator=0.36875  CurrentEval: classification_error_evaluator=0.36875
+...
+Pass=0 Batch=196 samples=25000 AvgCost=0.418964 Eval: classification_error_evaluator=0.1922
+Test samples=24999 cost=0.39297 Eval: classification_error_evaluator=0.149406
 
      
 
      
 
        
@@ -280,8 +280,8 @@ Test samples=24999 cost=0.39297 Eval: classification_error_evaluator=0.149406
 
        
 
        Testing
        
 
        Testing means evaluating the labeled validation set using trained model.
        
-
        cd demo/sentiment
-./test.sh
+
        cd demo/sentiment
+./test.sh
 
        
 
        
 
        test.sh:
        
@@ -311,19 +311,19 @@ paddle train --config=$net_conf <
 
        
 
        
 
        The function get_best_pass gets the best model by classification error rate for testing. In this example, We use test dataset of IMDB as validation by default. Unlike training, it needs to specify --job=test and model path, namely --model_list=$model_list here. If running successfully, the log is saved in path of demo/sentiment/test.log. For example, in our test, the best model is model_output/pass-00002, the classification error is 0.115645 as follows.
        
-
        Pass=0 samples=24999 AvgCost=0.280471 Eval: classification_error_evaluator=0.115645
+
        Pass=0 samples=24999 AvgCost=0.280471 Eval: classification_error_evaluator=0.115645
 
        
 
        
 
        
 
        
 
        Prediction
        
 
        predict.py provides a predicting interface. You should install python api of PaddlePaddle before using it. One example to predict unlabeled review of IMDB is as follows. Simply running:
        
-
        cd demo/sentiment
-./predict.sh
+
        cd demo/sentiment
+./predict.sh
 
        
 
        
 
        predict.sh:
        
-
        #Note the default model is pass-00002, you shold make sure the model path
+
        #Note the default model is pass-00002, you shold make sure the model path
 #exists or change the mode path.
 model=model_output/pass-00002/
 config=trainer_config.py
@@ -347,8 +347,8 @@ python predict.py \
 
        Note you should make sure the default model path model_output/pass-00002
 exists or change the model path.
        
 
        Predicting result of this example:
        
-
        Loading parameters from model_output/pass-00002/
-./data/aclImdb/test/pos/10014_7.txt: predicting label is pos
+
        Loading parameters from model_output/pass-00002/
+./data/aclImdb/test/pos/10014_7.txt: predicting label is pos
 
        
 
        
 
        We sincerely appreciate your interest and welcome your contributions.
        
@@ -406,14 +406,11 @@ exists or change the model path.
        
 
 
         
        
@@ -435,14 +432,14 @@ exists or change the model path.
        
         
      • 
           previous |
        • 
-        
      • PaddlePaddle  documentation »
        • 
-          
      • Examples and demos »
        • 
-          
      • Sentiment Analasis Tutorial »
        •  
+        
      • PaddlePaddle  documentation »
        • 
+          
      • Examples and demos »
        • 
+          
      • Sentiment Analasis Tutorial »
        •  
       
      
     
      
     
   
 
\ No newline at end of file
diff --git a/doc/demo/text_generation/index.html b/doc/demo/text_generation/index.html
index e26bccfd8b..77322993ed 100644
--- a/doc/demo/text_generation/index.html
+++ b/doc/demo/text_generation/index.html
@@ -6,7 +6,7 @@
   
     
     
-    Text Generation Tutorial — PaddlePaddle  documentation
+    Text Generation Tutorial — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
    • 
           previous |
      • 
-        
    • PaddlePaddle  documentation »
      • 
-          
    • Examples and demos »
      •  
+        
    • PaddlePaddle  documentation »
      • 
+          
    • Examples and demos »
      •  
       
    
     
      
 
@@ -112,14 +112,11 @@
 
 
         
    
@@ -141,13 +138,13 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Examples and demos »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Examples and demos »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/demo/text_generation/text_generation.html b/doc/demo/text_generation/text_generation.html
index df60465e6a..a33d55b76e 100644
--- a/doc/demo/text_generation/text_generation.html
+++ b/doc/demo/text_generation/text_generation.html
@@ -6,7 +6,7 @@
   
     
     
-    Text generation Tutorial — PaddlePaddle  documentation
+    Text generation Tutorial — PaddlePaddle  documentation
     
     
     
@@ -26,7 +26,7 @@
     
     
     
-    
+    
      
   
   
@@ -40,14 +40,14 @@
           modules |
         
  • 
-          next |
    • 
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Examples and demos »
    • 
-          
  • Text Generation Tutorial »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Examples and demos »
    • 
+          
  • Text Generation Tutorial »
    •  
       
     
      
 
@@ -120,19 +120,19 @@
 
    
 
    User Defined Dataset
    
 
    If you need to do other sequence-to-sequence tasks, such as Paraphrasing, you only need to organize the data as follows, and place them in demo/seqToseq/data:
    
-
    dataset
-  train
-    file1.src file1.trg
-    file2.src file2.trg
-    ......
-  test
-    file1.src file1.trg
-    file2.src file2.trg
-    ......
-  gen
-    file1.src file1.trg
-    file2.src file2.trg
-    ......
+
    dataset
+  train
+    file1.src file1.trg
+    file2.src file2.trg
+    ......
+  test
+    file1.src file1.trg
+    file2.src file2.trg
+    ......
+  gen
+    file1.src file1.trg
+    file2.src file2.trg
+    ......
 
    
 
    
 
      
@@ -182,10 +182,10 @@
 
    • -m --mergeDict: merge source and target dictionary, thus, two dictionaries have the same context
      • 
 
    
 
    And you will see messages like this:
    
-
    concat parallel corpora for dataset
-build source dictionary for train data
-build target dictionary for train data
-dictionary size is XXX
+
    concat parallel corpora for dataset
+build source dictionary for train data
+build target dictionary for train data
+dictionary size is XXX
 
    
 
    
 
    Here, you can simply run the command:
    
@@ -193,7 +193,7 @@ dictionary size is XXX
 
    
 
    
 
    It will take several minutes, and store the preprocessed dataset in demo/seqToseq/data/pre-wmt14, the directory has following structure.
    
-
    train test gen train.list test.list gen.list src.dict trg.dict
+
    train test gen train.list test.list gen.list src.dict trg.dict
 
    
 
    
 
      
@@ -274,9 +274,9 @@ dictionary size is XXX
 
    • dot_period: here print ‘.’ every 5 batches
      • 
 
    
 
    The training loss function is printed every 10 batch by default, and you will see messages like this:
    
-
    I0719 19:16:45.952062 15563 TrainerInternal.cpp:160]  Batch=10 samples=500 AvgCost=198.475 CurrentCost=198.475 Eval: classification_error_evaluator=0.737155  CurrentEval: classification_error_evaluator=0.737155
-I0719 19:17:56.707319 15563 TrainerInternal.cpp:160]  Batch=20 samples=1000 AvgCost=157.479 CurrentCost=116.483 Eval: classification_error_evaluator=0.698392  CurrentEval: classification_error_evaluator=0.659065
-.....
+
    I0719 19:16:45.952062 15563 TrainerInternal.cpp:160]  Batch=10 samples=500 AvgCost=198.475 CurrentCost=198.475 Eval: classification_error_evaluator=0.737155  CurrentEval: classification_error_evaluator=0.737155
+I0719 19:17:56.707319 15563 TrainerInternal.cpp:160]  Batch=20 samples=1000 AvgCost=157.479 CurrentCost=116.483 Eval: classification_error_evaluator=0.698392  CurrentEval: classification_error_evaluator=0.659065
+.....
 
    
 
    
 
      
@@ -360,23 +360,23 @@ I0719 19:17:56.707319 15563 TrainerInternal.cpp:160]  Batch=20 samples=1000 AvgC
 
    • num_passes and test_pass: loading model parameters from test_pass to (num_passes - 1), here only loads data/wmt14_model/pass-00012
      • 
 
    
 
    You will see messages like this:
    
-
    I0706 14:48:31.178915 31441 GradientMachine.cpp:143] Loading parameters from data/wmt14_model/pass-00012
-I0706 14:48:40.012039 31441 Tester.cpp:125]  Batch=100 samples=100 AvgCost=0
-I0706 14:48:48.898632 31441 Tester.cpp:125]  Batch=200 samples=200 AvgCost=0
-...
+
    I0706 14:48:31.178915 31441 GradientMachine.cpp:143] Loading parameters from data/wmt14_model/pass-00012
+I0706 14:48:40.012039 31441 Tester.cpp:125]  Batch=100 samples=100 AvgCost=0
+I0706 14:48:48.898632 31441 Tester.cpp:125]  Batch=200 samples=200 AvgCost=0
+...
 
    
 
    
 
    And the generating result in demo/seqToseq/translation/gen_result likes:
    
-
    0
-0       -11.1314         The <unk> <unk> about the width of the seats while large controls are at stake <e>
-1       -11.1519         The <unk> <unk> on the width of the seats while large controls are at stake <e>
-2       -11.5988         The <unk> <unk> about the width of the seats while large controls are at stake . <e>
+
    0
+0       -11.1314         The <unk> <unk> about the width of the seats while large controls are at stake <e>
+1       -11.1519         The <unk> <unk> on the width of the seats while large controls are at stake <e>
+2       -11.5988         The <unk> <unk> about the width of the seats while large controls are at stake . <e>
 
-1
-0       -24.4149         The dispute is between the major aircraft manufacturers about the width of the tourist seats on the <unk> flights , paving the way for a <unk> confrontation during the month of the Dubai <unk> . <e>
-1       -26.9524         The dispute is between the major aircraft manufacturers about the width of the tourist seats on the <unk> flights , paving the way for a <unk> confrontation during the month of Dubai &apos; s <unk> . <e>
-2       -27.9574         The dispute is between the major aircraft manufacturers about the width of the tourist seats on the <unk> flights , paving the way for a <unk> confrontation during the month of Dubai &apos; s Dubai <unk> . <e>
-...
+1
+0       -24.4149         The dispute is between the major aircraft manufacturers about the width of the tourist seats on the <unk> flights , paving the way for a <unk> confrontation during the month of the Dubai <unk> . <e>
+1       -26.9524         The dispute is between the major aircraft manufacturers about the width of the tourist seats on the <unk> flights , paving the way for a <unk> confrontation during the month of Dubai &apos; s <unk> . <e>
+2       -27.9574         The dispute is between the major aircraft manufacturers about the width of the tourist seats on the <unk> flights , paving the way for a <unk> confrontation during the month of Dubai &apos; s Dubai <unk> . <e>
+...
 
    
 
    
 
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/dev/new_layer/index.html b/doc/dev/new_layer/index.html
index a93b0b133a..c2e3f54877 100644
--- a/doc/dev/new_layer/index.html
+++ b/doc/dev/new_layer/index.html
@@ -6,7 +6,7 @@
   
     
     
-    Writing New Layers — PaddlePaddle  documentation
+    Writing New Layers — PaddlePaddle  documentation
     
     
     
@@ -44,7 +44,7 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
      
 
@@ -88,14 +88,11 @@
 
 
         
    
@@ -117,12 +114,12 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/dev/new_layer/new_layer.html b/doc/dev/new_layer/new_layer.html
index 1dddfb7805..c035074962 100644
--- a/doc/dev/new_layer/new_layer.html
+++ b/doc/dev/new_layer/new_layer.html
@@ -6,7 +6,7 @@
   
     
     
-    Writing New Layers — PaddlePaddle  documentation
+    Writing New Layers — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Writing New Layers »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Writing New Layers »
    •  
       
     
      
 
@@ -466,14 +466,11 @@ add_test(NAME test_FCGrad
 
 
         
    
@@ -495,13 +492,13 @@ add_test(NAME test_FCGrad
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Writing New Layers »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Writing New Layers »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/genindex.html b/doc/genindex.html
index dd9ec48cd1..ef15003b24 100644
--- a/doc/genindex.html
+++ b/doc/genindex.html
@@ -7,7 +7,7 @@
   
     
     
-    Index — PaddlePaddle  documentation
+    Index — PaddlePaddle  documentation
     
     
     
@@ -37,7 +37,7 @@
         
  • 
           modules |
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
      
 
@@ -75,6 +75,272 @@
 
+  
-  
    
   
    
       
+  
    __sparse_get_return__ (C macro)
+  
    
+
+      
+  
    __sparse_get_type_return__ (C macro)
+  
    
+
+      
+  
    _cudnn_convolution_descriptor (C++ class)
+  
    
+
+      
+  
    _cudnn_convolution_descriptor::desc (C++ member)
+  
    
+
+      
+  
    _cudnn_convolution_descriptor::filter (C++ member)
+  
    
+
+      
+  
    _cudnn_convolution_descriptor::input_image (C++ member)
+  
    
+
+      
+  
    _cudnn_convolution_descriptor::mode (C++ member)
+  
    
+
+      
+  
    _cudnn_convolution_descriptor::padding_height (C++ member)
+  
    
+
+      
+  
    _cudnn_convolution_descriptor::padding_width (C++ member)
+  
    
+
+      
+  
    _cudnn_convolution_descriptor::stride_height (C++ member)
+  
    
+
+      
+  
    _cudnn_convolution_descriptor::stride_width (C++ member)
+  
    
+
+      
+  
    _cudnn_convolution_descriptor::upscalex (C++ member)
+  
    
+
+      
+  
    _cudnn_convolution_descriptor::upscaley (C++ member)
+  
    
+
+      
+  
    _cudnn_filter_descriptor (C++ class)
+  
    
+
+      
+  
    _cudnn_filter_descriptor::data_type (C++ member)
+  
    
+
+      
+  
    _cudnn_filter_descriptor::desc (C++ member)
+  
    
+
+      
+  
    _cudnn_filter_descriptor::filter_height (C++ member)
+  
    
+
+      
+  
    _cudnn_filter_descriptor::filter_width (C++ member)
+  
    
+
+      
+  
    _cudnn_filter_descriptor::input_feature_maps (C++ member)
+  
    
+
+      
+  
    _cudnn_filter_descriptor::output_feature_maps (C++ member)
+  
    
+
+      
+  
    _cudnn_pooling_descriptor (C++ class)
+  
    
+
+      
+  
    _cudnn_pooling_descriptor::desc (C++ member)
+  
    
+
+      
+  
    _cudnn_pooling_descriptor::mode (C++ member)
+  
    
+
+      
+  
    _cudnn_pooling_descriptor::stride_height (C++ member)
+  
    
+
+      
+  
    _cudnn_pooling_descriptor::stride_width (C++ member)
+  
    
+
+      
+  
    _cudnn_pooling_descriptor::window_height (C++ member)
+  
    
+
+      
+  
    _cudnn_pooling_descriptor::window_width (C++ member)
+  
    
+
+      
+  
    _cudnn_tensor_descriptor (C++ class)
+  
    
+
+      
+  
    _cudnn_tensor_descriptor::batch_size (C++ member)
+  
    
+
+      
+  
    _cudnn_tensor_descriptor::data_type (C++ member)
+  
    
+
+      
+  
    _cudnn_tensor_descriptor::desc (C++ member)
+  
    
+
+      
+  
    _cudnn_tensor_descriptor::feature_maps (C++ member)
+  
    
+
+      
+  
    _cudnn_tensor_descriptor::format (C++ member)
+  
    
+
+      
+  
    _cudnn_tensor_descriptor::height (C++ member)
+  
    
+
+      
+  
    _cudnn_tensor_descriptor::width (C++ member)
+  
    
+
+      
+  
    _global_device_resources (C++ class)
+  
    
+
+      
+  
    _global_device_resources::cudnn_handle (C++ member)
+  
    
+
+      
+  
    _global_device_resources::gen (C++ member)
+  
    
+
+      
+  
    _global_device_resources::gen_mutex (C++ member)
+  
    
+
+      
+  
    _global_device_resources::handle (C++ member)
+  
    
+
+      
+  
    _global_device_resources::stream (C++ member)
+  
    
+
+      
+  
    _hl_csc_matrix (C++ class)
+  
    
+
+      
+  
    _hl_csc_matrix::col_s (C++ member)
+  
    
+
+      
+  
    _hl_csc_matrix::csc_col (C++ member)
+  
    
+
+      
+  
    _hl_csc_matrix::csc_row (C++ member)
+  
    
+
+      
+  
    _hl_csc_matrix::csc_val (C++ member)
+  
    
+
+  
    
+      
+  
    _hl_csc_matrix::nnz_s (C++ member)
+  
    
+
+      
+  
    _hl_csc_matrix::sparsity (C++ member)
+  
    
+
+      
+  
    _hl_csr_matrix (C++ class)
+  
    
+
+      
+  
    _hl_csr_matrix::csr_col (C++ member)
+  
    
+
+      
+  
    _hl_csr_matrix::csr_row (C++ member)
+  
    
+
+      
+  
    _hl_csr_matrix::csr_val (C++ member)
+  
    
+
+      
+  
    _hl_csr_matrix::nnz_s (C++ member)
+  
    
+
+      
+  
    _hl_csr_matrix::row_s (C++ member)
+  
    
+
+      
+  
    _hl_csr_matrix::sparsity (C++ member)
+  
    
+
+      
+  
    _hl_device_prop (C++ class)
+  
    
+
+      
+  
    _hl_device_prop::device (C++ member)
+  
    
+
+      
+  
    _hl_device_prop::device_mem (C++ member)
+  
    
+
+      
+  
    _hl_device_prop::device_name (C++ member)
+  
    
+
+      
+  
    _hl_device_prop::device_resources (C++ member)
+  
    
+
+      
+  
    _hl_device_prop::device_type (C++ member)
+  
    
+
+      
+  
    _hl_device_prop::is_local (C++ member)
+  
    
+
+      
+  
    _hl_device_prop::major (C++ member)
+  
    
+
+      
+  
    _hl_device_prop::minor (C++ member)
+  
    
+
+      
+  
    _hl_event_st (C++ class)
+  
    
+
+      
+  
    _hl_event_st::cu_event (C++ member)
+  
    
+
+      
   
    _hl_sparse_matrix_s (C++ class)
   
    
 
@@ -90,8 +356,6 @@
   
    _hl_sparse_matrix_s::matrix (C++ member)
   
    
 
-  
    
       
   
    _hl_sparse_matrix_s::nnz (C++ member)
   
    
@@ -104,6 +368,78 @@
   
    _hl_sparse_matrix_s::type (C++ member)
   
    
 
+      
+  
    _hl_thread_resource (C++ class)
+  
    
+
+      
+  
    _hl_thread_resource::cpu_mem (C++ member)
+  
    
+
+      
+  
    _hl_thread_resource::cudnn_desc (C++ member)
+  
    
+
+      
+  
    _hl_thread_resource::cudnn_handle (C++ member)
+  
    
+
+      
+  
    _hl_thread_resource::device (C++ member)
+  
    
+
+      
+  
    _hl_thread_resource::event (C++ member)
+  
    
+
+      
+  
    _hl_thread_resource::gen (C++ member)
+  
    
+
+      
+  
    _hl_thread_resource::gen_mutex (C++ member)
+  
    
+
+      
+  
    _hl_thread_resource::gpu_mem (C++ member)
+  
    
+
+      
+  
    _hl_thread_resource::handle (C++ member)
+  
    
+
+      
+  
    _hl_thread_resource::is_init (C++ member)
+  
    
+
+      
+  
    _hl_thread_resource::major (C++ member)
+  
    
+
+      
+  
    _hl_thread_resource::stream (C++ member)
+  
    
+
+      
+  
    _thread_device_resources (C++ class)
+  
    
+
+      
+  
    _thread_device_resources::cpu_mem (C++ member)
+  
    
+
+      
+  
    _thread_device_resources::gpu_mem (C++ member)
+  
    
+
+      
+  
    _thread_device_resources::mem_event (C++ member)
+  
    
+
+      
+  
    _thread_device_resources::stream (C++ member)
+  
    
+
   
    		
 
    
 
@@ -111,6 +447,22 @@
 
+  
-  
    
   
    
       
+  
    aggregate (C++ type)
+  
    
+
+      
+  
    aggregate::max (C++ class)
+  
    
+
+      
+  
    aggregate::min (C++ class)
+  
    
+
+      
+  
    aggregate::sum (C++ class)
+  
    
+
+      
   
    Arguments (C++ class)
   
    
 
@@ -150,6 +502,8 @@
   
    Arguments::getSlotNum (C++ function)
   
    
 
+  
    
       
   
    Arguments::getSlotSequenceDim (C++ function)
   
    
@@ -158,8 +512,6 @@
   
    Arguments::getSlotSequenceStartPositions (C++ function)
   
    
 
-  
    
       
   
    Arguments::getSlotSubSequenceStartPositions (C++ function)
   
    
@@ -211,6 +563,100 @@
 
+  
@@ -241,16 +707,56 @@
 
    
   
    
       
+  
    backward_reset_grad (C++ function)
+  
    
+
+      
+  
    backward_state_grad (C++ function)
+  
    
+
+      
+  
    base (C++ type)
+  
    
+
+      
+  
    base::binary (C++ type)
+  
    
+
+      
+  
    base::binary::add (C++ class)
+  
    
+
+      
+  
    base::binary::add2 (C++ class)
+  
    
+
+      
+  
    base::binary::add2::add2 (C++ function)
+  
    
+
+      
+  
    base::binary::add2::p1 (C++ member)
+  
    
+
+      
+  
    base::binary::add2::p2 (C++ member)
+  
    
+
+      
+  
    base::binary::classificationError (C++ class)
+  
    
+
+      
+  
    base::binary::classificationError::classificationError (C++ function)
+  
    
+
+      
+  
    base::binary::classificationError::p (C++ member)
+  
    
+
+  
    
+      
+  
    base::binary::div (C++ class)
+  
    
+
+      
+  
    base::binary::first (C++ class)
+  
    
+
+      
+  
    base::binary::mul (C++ class)
+  
    
+
+      
+  
    base::binary::second (C++ class)
+  
    
+
+      
+  
    base::binary::squaredDiff (C++ class)
+  
    
+
+      
+  
    base::binary::sub (C++ class)
+  
    
+
+      
+  
    base::unary (C++ type)
+  
    
+
+      
+  
    base::unary::identity (C++ class)
+  
    
+
+      
+  
    BaseOp (C++ class)
+  
    
+
+      
+  
    BaseOp::BaseOp (C++ function), [1], [2]
+  
    
+
+      
+  
    BaseOp::sse (C++ member)
+  
    
+
+      
   
    batchTranspose (C++ function)
   
    
 
@@ -225,13 +671,33 @@
   
 
       
+  
    CBLAS_GEMM (C macro)
+  
    
+
+      
   
    CREATE_MODE_NORMAL (C++ class)
   
    
 
+      
+  
    CREATE_MODE_TESTING (C++ class)
+  
    
+
   
    		
   
    
       
-  
    CREATE_MODE_TESTING (C++ class)
+  
    cudnn_convolution_descriptor (C++ type)
+  
    
+
+      
+  
    cudnn_filter_descriptor (C++ type)
+  
    
+
+      
+  
    cudnn_pooling_descriptor (C++ type)
+  
    
+
+      
+  
    cudnn_tensor_descriptor (C++ type)
   
    
 
   
    
-  
+  
    
   
    
       
-  
    define_py_data_sources() (in module paddle.trainer_config_helpers.data_sources)
+  
    DEFINE_MATRIX_BINARY_OP (C macro)
   
    
 
       
-  
    DISABLE_COPY_AND_ASSIGN (C macro)
+  
    DEFINE_MATRIX_BINARY_PARAMETER_OP (C macro)
+  
    
+
+      
+  
    DEFINE_MATRIX_QUATERNARY_OP (C macro)
+  
    
+
+      
+  
    DEFINE_MATRIX_QUATERNARY_PARAMETER_OP (C macro)
+  
    
+
+      
+  
    DEFINE_MATRIX_TERNARY_OP (C macro)
+  
    
+
+      
+  
    DEFINE_MATRIX_TERNARY_PARAMETER_OP (C macro)
+  
    
+
+      
+  
    DEFINE_MATRIX_UNARY_OP (C macro)
   
    
 
   
    		
   
    
       
+  
    DEFINE_MATRIX_UNARY_PARAMETER_OP (C macro)
+  
    
+
+      
+  
    define_py_data_sources2() (in module paddle.trainer_config_helpers.data_sources)
+  
    
+
+      
+  
    DEVICE_FMAX (C macro)
+  
    
+
+      
+  
    DEVICE_FMIN (C macro)
+  
    
+
+      
+  
    DISABLE_COPY_AND_ASSIGN (C macro)
+  
    
+
+      
   
    DIVUP (C macro)
   
    
 
@@ -288,16 +794,28 @@
   
    FloatArray::FloatArray (C++ function)
   
    
 
-  
    
       
   
    FloatArray::length (C++ member)
   
    
 
+  
    
       
   
    FloatArray::needFree (C++ member)
   
    
 
+      
+  
    forward_final_output (C++ function)
+  
    
+
+      
+  
    forward_reset_output (C++ function)
+  
    
+
+      
+  
    FOUR_PARAMETER (C macro)
+  
    
+
   
    		
 
    
 
@@ -305,6 +823,18 @@
 
+  
-  
+  
-  
    
   
    
       
+  
    GET_CONVOLUTION_DESCRIPTOR (C macro)
+  
    
+
+      
+  
    GET_FILTER_DESCRIPTOR (C macro)
+  
    
+
+      
+  
    GET_TENSOR_DESCRIPTOR (C macro)
+  
    
+
+      
   
    GetCublasDsoHandle (C++ function)
   
    
 
@@ -325,6 +855,10 @@
   
 
       
+  
    global_device_resources (C++ type)
+  
    
+
+      
   
    GradientMachine (C++ class)
   
    
 
@@ -348,6 +882,8 @@
   
    GradientMachine::createByModelConfig (C++ function)
   
    
 
+  
    
       
   
    GradientMachine::createFromPaddleModelPtr (C++ function)
   
    
@@ -356,8 +892,6 @@
   
    GradientMachine::defaultParamTypes (C++ member)
   
    
 
-  
    
       
   
    GradientMachine::DISABLE_COPY_AND_ASSIGN (C++ function)
   
    
@@ -441,6 +975,10 @@
   
 
       
+  
    hl_agg_op (C++ function)
+  
    
+
+      
   
    hl_avgpool_backward (C++ function)
   
    
 
@@ -449,6 +987,22 @@
   
 
       
+  
    hl_avx_gru_backward_reset_grad (C++ function)
+  
    
+
+      
+  
    hl_avx_gru_backward_state_grad (C++ function)
+  
    
+
+      
+  
    hl_avx_gru_forward_final_output (C++ function)
+  
    
+
+      
+  
    hl_avx_gru_forward_reset_output (C++ function)
+  
    
+
+      
   
    hl_batch_norm_backward (C++ function)
   
    
 
@@ -461,6 +1015,10 @@
   
 
       
+  
    hl_check_align (C++ function), [1]
+  
    
+
+      
   
    hl_CMRNorm_backward (C++ function)
   
    
 
@@ -469,79 +1027,159 @@
   
 
       
-  
    hl_construct_sparse_matrix (C++ function), [1]
+  
    hl_construct_sparse_matrix (C++ function), [1]
+  
    
+
+      
+  
    hl_context_projection_backward_data (C++ function)
+  
    
+
+      
+  
    hl_context_projection_backward_weight (C++ function)
+  
    
+
+      
+  
    hl_context_projection_forward (C++ function)
+  
    
+
+      
+  
    hl_conv_workspace (C++ function)
+  
    
+
+      
+  
    hl_convolution_backward_bias (C++ function)
+  
    
+
+      
+  
    hl_convolution_backward_data (C++ function)
+  
    
+
+      
+  
    hl_convolution_backward_filter (C++ function)
+  
    
+
+      
+  
    hl_convolution_descriptor (C++ type)
+  
    
+
+      
+  
    hl_convolution_forward (C++ function)
+  
    
+
+      
+  
    hl_convolution_forward_add_bias (C++ function)
+  
    
+
+      
+  
    hl_cossim (C++ function)
+  
    
+
+      
+  
    hl_cossim_derivative (C++ function)
+  
    
+
+      
+  
    hl_cpu_apply_binary_op (C++ function)
+  
    
+
+      
+  
    hl_cpu_apply_quaternary_op (C++ function)
+  
    
+
+      
+  
    hl_cpu_apply_ternary_op (C++ function)
+  
    
+
+      
+  
    hl_cpu_apply_unary_op (C++ function)
+  
    
+
+      
+  
    hl_cpu_gru_backward (C++ function), [1]
+  
    
+
+      
+  
    HL_CPU_GRU_CUH_ (C macro)
+  
    
+
+      
+  
    hl_cpu_gru_forward (C++ function), [1]
+  
    
+
+      
+  
    hl_cpu_lstm_backward (C++ function)
   
    
 
       
-  
    hl_context_projection_backward_data (C++ function)
+  
    hl_cpu_lstm_forward (C++ function)
   
    
 
       
-  
    hl_context_projection_backward_weight (C++ function)
+  
    hl_cpu_matrix_column_op (C++ function), [1]
   
    
 
       
-  
    hl_context_projection_forward (C++ function)
+  
    hl_cpu_matrix_row_op (C++ function), [1]
   
    
 
       
-  
    hl_conv_workspace (C++ function)
+  
    hl_create_convolution_descriptor (C++ function)
   
    
 
       
-  
    hl_convolution_backward_bias (C++ function)
+  
    hl_create_event (C++ function)
   
    
 
       
-  
    hl_convolution_backward_data (C++ function)
+  
    hl_create_filter_descriptor (C++ function)
   
    
 
       
-  
    hl_convolution_backward_filter (C++ function)
+  
    hl_create_global_resources (C++ function)
   
    
 
       
-  
    hl_convolution_descriptor (C++ type)
+  
    hl_create_pooling_descriptor (C++ function)
   
    
 
       
-  
    hl_convolution_forward (C++ function)
+  
    hl_create_tensor_descriptor (C++ function)
   
    
 
       
-  
    hl_convolution_forward_add_bias (C++ function)
+  
    hl_create_thread_resources (C++ function)
   
    
 
       
-  
    hl_cossim (C++ function)
+  
    hl_csc_matrix (C++ type)
   
    
 
       
-  
    hl_cossim_derivative (C++ function)
+  
    hl_csr_matrix (C++ type)
   
    
 
       
-  
    hl_create_convolution_descriptor (C++ function)
+  
    hl_cublas_init (C++ function)
   
    
 
       
-  
    hl_create_event (C++ function)
+  
    HL_CUDA_CUDNN_PH_ (C macro)
   
    
 
       
-  
    hl_create_filter_descriptor (C++ function)
+  
    hl_cuda_event_is_ready (C++ function)
   
    
 
       
-  
    hl_create_pooling_descriptor (C++ function)
+  
    HL_CUDA_PH_ (C macro)
   
    
 
       
-  
    hl_create_tensor_descriptor (C++ function)
+  
    hl_cudnn_desc_init (C++ function)
   
    
 
       
-  
    hl_cuda_event_query (C++ function)
+  
    hl_cudnn_init (C++ function)
   
    
 
       
@@ -569,6 +1207,10 @@
   
 
       
+  
    HL_DEVICE (C macro)
+  
    
+
+      
   
    hl_device_can_access_peer (C++ function)
   
    
 
@@ -577,6 +1219,14 @@
   
 
       
+  
    HL_DEVICE_FUNCTIONS_CUH_ (C macro)
+  
    
+
+      
+  
    hl_device_prop (C++ type)
+  
    
+
+      
   
    hl_device_synchronize (C++ function)
   
    
 
@@ -657,6 +1307,58 @@
   
 
       
+  
    hl_gpu_apply_binary_op (C++ function)
+  
    
+
+      
+  
    hl_gpu_apply_quaternary_op (C++ function)
+  
    
+
+      
+  
    hl_gpu_apply_ternary_op (C++ function)
+  
    
+
+      
+  
    hl_gpu_apply_unary_op (C++ function)
+  
    
+
+      
+  
    HL_GPU_FUNCTIONS_CUH_ (C macro)
+  
    
+
+      
+  
    hl_gpu_gru_backward (C++ function), [1]
+  
    
+
+      
+  
    HL_GPU_GRU_CUH_ (C macro)
+  
    
+
+      
+  
    hl_gpu_gru_forward (C++ function), [1]
+  
    
+
+      
+  
    hl_gpu_lstm_backward (C++ function), [1]
+  
    
+
+      
+  
    HL_GPU_LSTM_CUH_ (C macro)
+  
    
+
+      
+  
    hl_gpu_lstm_forward (C++ function), [1]
+  
    
+
+      
+  
    hl_gpu_matrix_column_op (C++ function), [1]
+  
    
+
+      
+  
    hl_gpu_matrix_row_op (C++ function), [1]
+  
    
+
+      
   
    hl_gru_grad (C++ class)
   
    
 
@@ -685,6 +1387,10 @@
   
 
       
+  
    HL_GRU_OPS_CUH_ (C macro)
+  
    
+
+      
   
    hl_gru_value (C++ class)
   
    
 
@@ -753,6 +1459,10 @@
   
 
       
+  
    HL_LSTM_OPS_CUH_ (C macro)
+  
    
+
+      
   
    hl_lstm_parallel_backward_data (C++ function)
   
    
 
@@ -821,9 +1531,19 @@
   
 
       
+  
    HL_MATRIX_APPLY_H_ (C macro)
+  
    
+
+      
+  
    HL_MATRIX_BASE_CUH_ (C macro)
+  
    
+
+      
   
    hl_matrix_classification_error (C++ function)
   
    
 
+  
    
       
   
    hl_matrix_column_max (C++ function)
   
    
@@ -833,6 +1553,10 @@
   
 
       
+  
    hl_matrix_column_op (C++ function), [1]
+  
    
+
+      
   
    hl_matrix_column_sum (C++ function)
   
    
 
@@ -856,8 +1580,6 @@
   
    hl_matrix_csr2dense (C++ function)
   
    
 
-  
    
       
   
    hl_matrix_csr_add_bias (C++ function)
   
    
@@ -895,6 +1617,10 @@
   
 
       
+  
    HL_MATRIX_OPS_CUH_ (C macro)
+  
    
+
+      
   
    hl_matrix_row_max (C++ function)
   
    
 
@@ -931,6 +1657,10 @@
   
 
       
+  
    HL_MATRIX_TYPE_CUH_ (C macro)
+  
    
+
+      
   
    hl_matrix_value_t (C++ type)
   
    
 
@@ -999,6 +1729,22 @@
   
 
       
+  
    hl_naive_gru_backward_reset_grad (C++ function)
+  
    
+
+      
+  
    hl_naive_gru_backward_state_grad (C++ function)
+  
    
+
+      
+  
    hl_naive_gru_forward_final_output (C++ function)
+  
    
+
+      
+  
    hl_naive_gru_forward_reset_output (C++ function)
+  
    
+
+      
   
    HL_NO_VALUE (C++ class)
   
    
 
@@ -1051,6 +1797,10 @@
   
 
       
+  
    HL_RECURRENT_APPLY_CUH_ (C macro)
+  
    
+
+      
   
    hl_reset_convolution_descriptor (C++ function)
   
    
 
@@ -1143,6 +1893,10 @@
   
 
       
+  
    HL_SPARSE_PH_ (C macro)
+  
    
+
+      
   
    hl_specify_devices_start (C++ function)
   
    
 
@@ -1151,6 +1905,22 @@
   
 
       
+  
    hl_sse_column_op_with_rem (C++ function), [1]
+  
    
+
+      
+  
    hl_sse_matrix_column_op (C++ function), [1], [2], [3]
+  
    
+
+      
+  
    HL_SSE_MATRIX_KERNEL_CUH_ (C macro)
+  
    
+
+      
+  
    hl_sse_matrix_row_op (C++ function), [1]
+  
    
+
+      
   
    hl_start (C++ function)
   
    
 
@@ -1179,6 +1949,14 @@
   
 
       
+  
    HL_THREAD_PH_ (C macro)
+  
    
+
+      
+  
    hl_thread_resource (C++ type)
+  
    
+
+      
   
    hl_trans_op_t (C++ type)
   
    
 
@@ -1199,7 +1977,7 @@
   
 
       
-  
    hppl (C++ type), [1], [2]
+  
    hppl (C++ type), [1], [2], [3], [4], [5]
   
    
 
       
@@ -1207,11 +1985,39 @@
   
 
       
-  
    hppl::Active<T>::backward (C++ type)
+  
    hppl::Active<T>::backward (C++ type)
+  
    
+
+      
+  
    hppl::Active<T>::forward (C++ type)
+  
    
+
+      
+  
    hppl::backward (C++ type), [1]
+  
    
+
+      
+  
    hppl::backward::gru_resetGrad (C++ class)
+  
    
+
+      
+  
    hppl::backward::gru_resetGrad::avx (C++ member)
   
    
 
       
-  
    hppl::Active<T>::forward (C++ type)
+  
    hppl::backward::gru_stateGrad (C++ class)
+  
    
+
+      
+  
    hppl::backward::gru_stateGrad::avx (C++ member)
+  
    
+
+      
+  
    hppl::backward::lstm (C++ class)
+  
    
+
+      
+  
    hppl::backward::lstm::avx (C++ member)
   
    
 
       
@@ -1227,6 +2033,34 @@
   
 
       
+  
    hppl::forward (C++ type), [1]
+  
    
+
+      
+  
    hppl::forward::gru_finalOutput (C++ class)
+  
    
+
+      
+  
    hppl::forward::gru_finalOutput::avx (C++ member)
+  
    
+
+      
+  
    hppl::forward::gru_resetOutput (C++ class)
+  
    
+
+      
+  
    hppl::forward::gru_resetOutput::avx (C++ member)
+  
    
+
+      
+  
    hppl::forward::lstm (C++ class)
+  
    
+
+      
+  
    hppl::forward::lstm::avx (C++ member)
+  
    
+
+      
   
    hppl::linear (C++ function), [1], [2], [3]
   
    
 
@@ -1309,6 +2143,10 @@
   
 
       
+  
    INLINE (C macro), [1], [2]
+  
    
+
+      
   
    IntArray (C++ class)
   
    
 
@@ -1647,6 +2485,10 @@
 
+  
-  
-  
+  
-  
+  
    
   
    
       
+  
    ONE_PARAMETER (C macro)
+  
    
+
+      
   
    operator<< (C++ function), [1]
   
    
 
@@ -1662,12 +2504,12 @@
   
    OptimizationConfig::DISABLE_COPY_AND_ASSIGN (C++ function)
   
    
 
+  
    
       
   
    OptimizationConfig::getRawPtr (C++ function)
   
    
 
-  
    
       
   
    OptimizationConfig::m (C++ member)
   
    
@@ -1695,7 +2537,7 @@
   
 
       
-  
    paddle (C++ type), [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23]
+  
    paddle (C++ type), [1], [2], [3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22]
   
    
 
       
@@ -2075,7 +2917,7 @@
   
 
       
-  
    paddle::Argument::resizeAndCopyFrom (C++ function), [1]
+  
    paddle::Argument::resizeAndCopyFrom (C++ function), [1], [2], [3]
   
    
 
       
@@ -3583,10 +4425,6 @@
   
 
       
-  
    paddle::CMRProjectionNormLayer::blocked_ (C++ member)
-  
    
-
-      
   
    paddle::CMRProjectionNormLayer::CMRProjectionNormLayer (C++ function)
   
    
 
@@ -4047,10 +4885,6 @@
   
 
       
-  
    paddle::CosSimLayer::kCosSimScale_ (C++ member)
-  
    
-
-      
   
    paddle::CosSimLayer::~CosSimLayer (C++ function)
   
    
 
@@ -4395,11 +5229,11 @@
   
 
       
-  
    paddle::CpuMatrix::crossMapNormalBwd (C++ function)
+  
    paddle::CpuMatrix::crossMapNormalBwd (C++ function)
   
    
 
       
-  
    paddle::CpuMatrix::crossMapNormalFwd (C++ function)
+  
    paddle::CpuMatrix::crossMapNormalFwd (C++ function)
   
    
 
       
@@ -4983,10 +5817,6 @@
   
 
       
-  
    paddle::CRFLayer::backwardImp (C++ function)
-  
    
-
-      
   
    paddle::CRFLayer::coeff_ (C++ member)
   
    
 
@@ -5003,10 +5833,6 @@
   
 
       
-  
    paddle::CRFLayer::forwardImp (C++ function)
-  
    
-
-      
   
    paddle::CRFLayer::init (C++ function)
   
    
 
@@ -5019,10 +5845,6 @@
   
 
       
-  
    paddle::CRFLayer::tmpCpuInput_ (C++ member)
-  
    
-
-      
   
    paddle::CRFLayer::weightLayer_ (C++ member)
   
    
 
@@ -5399,7 +6221,15 @@
   
 
       
-  
    paddle::CustomStackTrace::dump (C++ function)
+  
    paddle::CustomStackTrace::dump (C++ function)
+  
    
+
+      
+  
    paddle::CustomStackTrace::DumpCallback (C++ type)
+  
    
+
+      
+  
    paddle::CustomStackTrace::empty (C++ function)
   
    
 
       
@@ -6151,10 +6981,6 @@
   
 
       
-  
    paddle::findLastSet (C++ function), [1]
-  
    
-
-      
   
    paddle::FLOAT_VALUE (C++ class)
   
    
 
@@ -6507,11 +7333,11 @@
   
 
       
-  
    paddle::GpuMatrix::crossMapNormalBwd (C++ function)
+  
    paddle::GpuMatrix::crossMapNormalBwd (C++ function)
   
    
 
       
-  
    paddle::GpuMatrix::crossMapNormalFwd (C++ function)
+  
    paddle::GpuMatrix::crossMapNormalFwd (C++ function)
   
    
 
       
@@ -7167,11 +7993,11 @@
   
 
       
-  
    paddle::GruCompute::backward (C++ function), [1]
+  
    paddle::GruCompute::backward (C++ function), [1], [2]
   
    
 
       
-  
    paddle::GruCompute::forward (C++ function), [1]
+  
    paddle::GruCompute::forward (C++ function), [1], [2]
   
    
 
       
@@ -8091,19 +8917,19 @@
   
 
       
-  
    paddle::LstmCompute::backwardBatch (C++ function), [1]
+  
    paddle::LstmCompute::backwardBatch (C++ function), [1], [2]
   
    
 
       
-  
    paddle::LstmCompute::backwardOneSequence (C++ function), [1]
+  
    paddle::LstmCompute::backwardOneSequence (C++ function), [1], [2]
   
    
 
       
-  
    paddle::LstmCompute::forwardBatch (C++ function), [1]
+  
    paddle::LstmCompute::forwardBatch (C++ function), [1], [2]
   
    
 
       
-  
    paddle::LstmCompute::forwardOneSequence (C++ function), [1]
+  
    paddle::LstmCompute::forwardOneSequence (C++ function), [1], [2]
   
    
 
       
@@ -8487,11 +9313,11 @@
   
 
       
-  
    paddle::Matrix::crossMapNormalBwd (C++ function)
+  
    paddle::Matrix::crossMapNormalBwd (C++ function)
   
    
 
       
-  
    paddle::Matrix::crossMapNormalFwd (C++ function)
+  
    paddle::Matrix::crossMapNormalFwd (C++ function)
   
    
 
       
@@ -8814,8 +9640,6 @@
   
    paddle::MatrixOffset::bRow_ (C++ member)
   
    
 
-  
    
       
   
    paddle::MatrixOffset::cCol_ (C++ member)
   
    
@@ -8836,6 +9660,8 @@
   
    paddle::MatrixOffset::MatrixOffset (C++ function)
   
    
 
+  
    
       
   
    paddle::MatrixPtr (C++ type)
   
    
@@ -11321,6 +12147,10 @@
   
 
       
+  
    paddle::ParameterServer2::OperatorFunction (C++ type)
+  
    
+
+      
   
    paddle::ParameterServer2::opFuncs (C++ member)
   
    
 
@@ -12277,11 +13107,11 @@
   
 
       
-  
    paddle::ProtoServer::registerServiceFunction (C++ function), [1]
+  
    paddle::ProtoServer::registerServiceFunction (C++ function), [1]
   
    
 
       
-  
    paddle::ProtoServer::registerServiceFunctionEx (C++ function), [1]
+  
    paddle::ProtoServer::registerServiceFunctionEx (C++ function)
   
    
 
       
@@ -12461,7 +13291,7 @@
   
 
       
-  
    paddle::RecurrentGradientMachine::createInFrameInfo (C++ function)
+  
    paddle::RecurrentGradientMachine::createInFrameInfo (C++ function)
   
    
 
       
@@ -12681,6 +13511,10 @@
   
 
       
+  
    paddle::RecurrentGradientMachine::numSeqs_ (C++ member)
+  
    
+
+      
   
    paddle::RecurrentGradientMachine::oneWaySearch (C++ function)
   
    
 
@@ -12829,6 +13663,10 @@
   
 
       
+  
    paddle::RecurrentGradientMachine::targetInfoInlinkId_ (C++ member)
+  
    
+
+      
   
    paddle::RecurrentGradientMachine::useGpu_ (C++ member)
   
    
 
@@ -13953,6 +14791,10 @@
   
 
       
+  
    paddle::SocketClient::SocketClient (C++ function)
+  
    
+
+      
   
    paddle::SocketClient::socketDaemon_ (C++ member)
   
    
 
@@ -14377,7 +15219,7 @@
   
 
       
-  
    paddle::SparseRemoteParameterUpdaterComposite::__anonymous0 (C++ type)
+  
    paddle::SparseRemoteParameterUpdaterComposite::__anonymous4 (C++ type)
   
    
 
       
@@ -15437,6 +16279,10 @@
   
 
       
+  
    paddle::UpdateFunction (C++ type)
+  
    
+
+      
   
    paddle::UserDefinedVectorPtr (C++ type)
   
    
 
@@ -16053,6 +16899,14 @@
 
+  
-  
    
   
    
       
+  
    thread_device_resources (C++ type)
+  
    
+
+      
+  
    THREE_PARAMETER (C macro)
+  
    
+
+      
   
    Trainer (C++ class)
   
    
 
@@ -16100,12 +16954,12 @@
   
    Trainer::setBatchSize (C++ function)
   
    
 
+  
    
       
   
    Trainer::startTrain (C++ function)
   
    
 
-  
    
       
   
    Trainer::startTrainPass (C++ function)
   
    
@@ -16154,6 +17008,10 @@
   
    TrainerConfig::~TrainerConfig (C++ function)
   
    
 
+      
+  
    TWO_PARAMETER (C macro)
+  
    
+
   
    		
 
    
 
@@ -16220,8 +17078,6 @@
   
    Vector::DISABLE_COPY_AND_ASSIGN (C++ function)
   
    
 
-  
    
       
   
    Vector::get (C++ function)
   
    
@@ -16230,6 +17086,8 @@
   
    Vector::getSharedPtr (C++ function)
   
    
 
+  
    
       
   
    Vector::getSize (C++ function)
   
    
@@ -16258,6 +17116,22 @@
   
    Vector::~Vector (C++ function)
   
    
 
+      
+  
    VECTOR_LEN (C macro)
+  
    
+
+      
+  
    VECTOR_SET (C macro)
+  
    
+
+      
+  
    VECTOR_SIZE (C macro)
+  
    
+
+      
+  
    vecType (C++ type)
+  
    
+
   
    		
 
    
 
@@ -16274,14 +17148,11 @@
 
 
         
@@ -16297,12 +17168,12 @@
         
  • 
           modules |
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
     
   
 
\ No newline at end of file
diff --git a/doc/index.html b/doc/index.html
index 552de4b21e..e58482467e 100644
--- a/doc/index.html
+++ b/doc/index.html
@@ -6,7 +6,7 @@
   
     
     
-    PaddlePaddle Documentation — PaddlePaddle  documentation
+    PaddlePaddle Documentation — PaddlePaddle  documentation
     
     
     
@@ -40,7 +40,7 @@
         
  • 
           next |
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
       
 
@@ -114,14 +114,11 @@
 
 
         
@@ -140,12 +137,12 @@
         
  • 
           next |
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
     
   
 
\ No newline at end of file
diff --git a/doc/layer.html b/doc/layer.html
index a4aebf603f..cd2193fd15 100644
--- a/doc/layer.html
+++ b/doc/layer.html
@@ -6,7 +6,7 @@
   
     
     
-    Layer Documents — PaddlePaddle  documentation
+    Layer Documents — PaddlePaddle  documentation
     
     
     
@@ -44,7 +44,7 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
       
 
@@ -85,14 +85,11 @@
 
 
         
@@ -114,12 +111,12 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
     
   
 
\ No newline at end of file
diff --git a/doc/objects.inv b/doc/objects.inv
index 5e9a8ffa269a3ad43ba944e7981484d7b6c916ab..8aa96b25f46d6442fc9434a0bb36ed29d12582c3 100644
GIT binary patch
delta 155290
zcmV)aK&ro{>j<#F36Oe!?7dra+e(%-_&&d)Bj&t)(O(hZUC}cUvIN;0TN0a;?XI4f
zU_fM&LYV}30FbikLBD-6am_mtzy_tNt`k+3NC9iDyzSh1+j({QZFzV3IKTa`<>K*j
z{}QgC^&{I^q0LGKa+n#?|1g&e?P~6``!P!`rm*2+rR%G
zSD(IJE|#C2yF2mC)p70qCG>srfBKzk=jQg_$>~4c-e1oDx_p=^0F=~HnAzq0^3Hj5
z=B)w$^ZO-+0ntW5!Rdn%2e0fqh_3Gds(##FIm?^7%l~Tp>g|2)S8o@WpT0TEtB2WN
zmzVd~%ge>$_Ok?k#yxoC|LnUzzN?aXjJ;iy{!71d^YDFM=2Bg;pPff6^!~@gx9^X)
z5BJM!XK^*ZojpFNkcxkJl$+acq?DQiJUaJ_ho%5uzAdk2-VP&F8w
z=;bADEI-PBhHjqx+xu_07d3zW0e88L`w!f+-96l1CrMp=dH8m%F)R{p$?LZ@l1g4{
zA^!2_L-Ut^I9IrjT7Glx8@{=49-F>-``@*@j^#2sxM75rTNE!t6dyx>xoiazzaCIy
zq|0!7+#^4Z{%v(QT	(6e%1d+^^40_LO<);(~pDe{=L#u^0GSi2G^rQu&*6VsVwu
zTLTumuRwFkXIS=Ch};;pOPe+1P@GZZ{T(f18P%uPnY&)^iidxoRUCxPQajN>Ix{Y#
zaSi(9#;fNA*48YYHVq+9B!1%7k5TBmyW-nf+%SE81)3wKY_SXC+jYb8>y-e?R3(4+
zo7Me);&mDW^(re4PJ
z2~bJju7cHy`Qh%4EgD_{Aty--c8~T?C^W|1J1j_uqNc`Sl)ft$7qxAmk*h5tn}#vZC{apN&BkM1KiLe!-l3;)uzQok`^bg3~%kc
zhwENqv_H8KdN{~@GnHGHB8~niUcR`XKtvgOocmGwa`wadI>5fgu1oAL8BP>z6d|c?
zG(+s;*NK(9!UM}kJW74Ic9!=K^SjG$hJLg)`Gb4?(OEp+-hVcIKXJrwc7YpThN1g^
zYWiA_eeL3cgD|V7RxKw2OZ@ia`!zOst-%=(8r=6=
zV2=e13OdAJwARglWZ06QTgpzAr{%iO_#9?h?4~s+Y{)1&3tIRapKE0+-%)%fJ7ICXKY1?v4
zwn*(!`C6R3ca^+#7pxJ}em%$0=_b$bvgEZgBoaWJffn@}h#RQWsIP9{!UF82h{WlJ%4-%&UbaSmd|>%&q9}4t
z=zNM;w(1gp0~{k{6Hn?+DFG*Ij#}V!2I@<;e14hiX>#tbl`~nw3Vc4I6k*7!`sKVX
zlHC)L=FYe8wG~Z7oNV>Hwj@i0dc3^<(pXZuFX+eRx9?7Ej6sHR|IiR1Pt5G$0hiF{
z<>e1&e)(CAqZsAOQI^hse_eh1eogimTwg0@$<(O*0f(xqiE>>$LbAJ}5kj^l=4x?O
zJB#}~#Jnv;lO~x%B-Nx5f?aH`d6+c96VGdFdc6*3^
zdx$}Mh->HO^82^id1?un+N4$t3y^HXSJMK@`U(-r21DwDAq|>;A>BSWVVF(qRnL|t
z-SBNbi#JCR9wnM?8L~#Qm(vCCRS?$lRz^I`BuAmfCvqz6%#iO|u~`;{lrui5Dbw#D
zDAVsM6IpZcC9sV%ki-WRDc+=2peb4AYTi8^1^oxsCaCjwu(?j>i{o%FlYFrL!O+*-H=6O8PW6+Lf#Dh4qnKxr;bknIiaCU0B
zoceka(en*%>9_Uf=*m)8V5R%yIL52PG5w;*<0g~?6&NWiNjg7s55BfYr+_8YkL0oc
zMq$vTX7tL}8S9v%W9!lYmTdTQsAXjuj?C&eX@%LVY%3&x#_!n;(Yje>N;hQX03vif
z<4j8}O>&OsMYt#%pbe83SuFDjqavlrJW&A<9MVO-|M&lEpve0zE!!?Iqh{4FqN*R$
zSYpZs-uj1~yJ}Lv)>MKcn$)-EeNNud`nC3oBt!B19o*&lwYP~(@^^|~oY`0JD{UG4
zCi_DQxO{VeNC{JOYe<1a?F%WvYVQgu6HIRj!C(IL;^+FCb;f_l-cE`_0%+)ljfJ-*
zUy+SGxTx5Y;$?X2-=FI2>Je#1DTE`(ue;4hg3pvHk6(Y&kI0}e=!=V!yK9?ADFngK
z_8FSho@CUOz5(GUXPX?=ZgRN`GAE_v*5b&1BsR-`2upP`UJh#w@6bL7U$Z?#ama2l
zd3x%Gc9v{xl9pzh9YrO5Qt$!TOlu+n%4Jx?
z^qJ%k0ND~^+36QIL5T4`sK~R+N}E2h)Ihmdb5bJLCYh6vw3m7_6B%|gF(GVFr66n0
zPGqpmgwG%!_^4X}qCgdBbzyj%slsEO0UV>r^{
zV;;d3wE@#q0;3}+6GtigrS$a6c-roDWOL*&5yNUdDklHs9
zaW$1t^0Y#lBDKv`$-YZUM8-~?naB>Ocz#;}wk1UscG$}F;0ix?Uuqr&k!Y9}K+2@o
zXf?82-m8)j2XI+w^%*o}THe5GQPzrQ%WLQA+vVI@J_W(9Q4eViSdhNu?sma{%hL1F
zX5+2M{^vZ<(lI&mL1>wy6@~$mikCnF8x?4pz(%D^hP}zbkP@r>EiD0a+(Z6%*>BZs
z^RL#h_>|Uy#1W(vlnuLfD=mqv5f78j9!hzd44x^fsRT)2L%-1qFiQsn&zU&*-y00b
za*;4j*-};A#EjB+9jjk_l^x!HV)1LYf2uYVuk&?}goCkl1*rV6Hbd{lp1V_T
zylIr?UI{CtHzEgT()J8VY16>YtQ@k5+(~~0p6S*^gclQ^4BFvS-)E42dD>rhUKACd
zZ(MIDb)TB0S_F4=>~q&&cg9Y@wd}HUXA$9@JP6qPizuh7t=?_X(!8gSA+8X_
z>G?Z0)^}(RC}NyJ@{Rnh-He79H=galg8raie$-as6_ZemVfy$
zmHRpkK$3k`H>e4R)gK#Ca3flCc}+&wg2+L?29ZK)qj^7G2si0}Cy>;M*}zu_wp(gF
z*C*ghCi(=nH^3Usn}loz(0Ilzoog3ghnZ^^admU;l0?X*DN@^9mFzo3zBV|nB||Uv
z{3(_gY7>{8Q;@$={$?x(JH-+`PYOXBJhqRXlW^v33&5hVlsTV1LxKynrR$YYN;W#Q
zP8lq!1?gBoRC1GlNt2dz1Aqi$Zy*iaqw*+XjQ#cY(F%avue
zKK*at#+&Z~vzk!Oa+>voHtQerC=eZ@JygQ0`$f%5dcFF83~8@e{U?2STwJ(^!|oME
z#W2BYS35<4=0;E>jkAR!%Uj1Klr@aG>u*t=_!Vi~YklaQ>euE?Ub}y|S4UJ*hReua
zuJvNFMMWOE*YvQV@HU9kJQiRzKC}3+Tvw|Y#UhF1tE8$)nDc`sT;#MyBUX{MZ&%@`
z167HzZTc&JY7%%vk|QG?j|q+`mr%pkk}xLbMS>ck_!DyXb;FxT7%U2ka%&WvLDY_J
zl!8sJOv
z%uKsvqkr|Qi7S^yimnqXCwm1XR8EGa36+xzqU|~fgOBPC*2)w}aFj$fhi3)Jki*~z(%{K$QZeK(Sc1lJ
zlneZ-@X0jQJ&Ip~@KDAe$g8|nLpfkJJcfOV-2jzCQ&O;4G%wXi65{?vee@z
z+HKh8(rrlrWz0234qUSKl016@sxVob`tC=6$HO5A<64srNyuMre`51j>rXQAnNVGKw5
zD2A*Nd24E2v%X}xBDV`JSAl;Dc1QB|a*d)@=p7VYj5;8HdNNB3NN!h(R_bd)rMUO@
zR0}#FzdF^_8*i6ys)HjRrKkqZ;#a3_L&
z^5t_Gc?B>Z1}uCHJugKV4?sD62Hijn8L+HTN14-S(9P9!k~KCBZlL9Oh*s_n{RY-C
z$#Kmkymlnvm6DL!l;(zb@P`Hq+h8|#mvKH*UMDF773j3UboA6s%)gSAxBL=+!tASt
zO-jjZ^%-Cz4b2*oT7pgE>J_B@!v?#b#}F;|?jZuyF%i-AYJy}Yw<|eQ2|T6e5Cl77
zlN&d#;U7A%hE;s11mx?ujIvROuA$^Qa~Wm3u1d|V?xr*)@q}
zoM4d6RKVv14NPf>ksk#>U6g=-MXsQZSI^5RA89tR3<9dvD^VRSd;1Pj0F`!-GU_%R
zq=3g84+$ZN6&CZ_R+wLzf-n_7uxymuUGU`Y@ZV0}
z3N1sF!bXi6r2^~XB9-x+1afYW3I(oZ)3}aS!4~;&#C?obNw?e|;|9Zj4khWdz|>Pi
z=kmxTd(%U?fTZOq-{tdexr$cHy)xo{0dG)y%L;i^%ns756PP9P9Pf+VAWsHvfFNj}u=0<0kiG)jMW)D+7-+P9{$dL@{S%g@74wIZ`9G#wl3ezl?;
z@sK*+CL=`5*)YCMQ6rB*bA{
zvn|?xaPJdmJN-~wpiSdUa+q9T7JV_bCd;VVy}}Zp#23?4cm>|QHI;x32$rKFsv0lW
zqM`?jGBU~w03b_$QIg?a3W%wXp_DYZrvfV}f~6$V8>D(fv^k2O@vvfD7{%t5m;=I<
z_;A#BOnT8A-Z+DQbCh(+$X7q-Q;X5+m5>#zRU^!Lyy75U#$4rH)Z*lP41uh|eAeO`
zGDcLO(#}iNA}S^&6f{-y5w*y;$p|SWY=hWVGzid2IjX3Q*!?JCC;+4_gHzH~I5hR@
zBjeTxR8^|Fzg#mW8zGS8^MGBCo*c_;-WV~3TrTX5QctqCoDpqZzqCY~0Udg{jW
zeF{XXrnuW}h_?CuR`xWWe$1P6&c*Ab*)*7hxt|sYeGK4mkl7)b=8&GxK}N@YJ4g&b
zAjvjwa(atoDvFoeP#K^zfewA2{Gau1Bp8k7U~Ah%77nM4t0J
zyN5rznCp~(94}aswcuoHX@MHWSZ_;ZB?eKpAfMVOi|SXi#0HT!FD{-u-wj`DBis+j
z>!b8-heNx#aM$YwHG;fE!BExTPSi}2QjPG
z-l=VC#1JCx#tSvGJ!Z(m1VqH%A9rmvV+av9LaV@UYn(HxMBM1v4Pi62>05V>pQ7>2kSKYDw#h&|;w
zNd3f}1c#hr(QSXJSCm-))!qC)2=~j}S*`7V{xFiXzZ|K+_ZSTf4SX`PiU3hpF_SsZHWPCRlq!k
zm*tt4{WmK!h0dAmtrjK5LCAS?w%`q)Ij2L`|jXHeKJ;2FtCvEeA??g+4zO)y3cFf0=qFBsI@R
z&)=X>6KqbM=A>A3h+QWzOQteuNvsxi^mI+idV0X_61J`=z|}Ka{aWG*z4aUV;w(sc
zl)D4)qzhkmlvNVqVHE@ipX(jZh_m
zG?JRRM~Tm;M&l$xTZE%qq3@L)Kp2}TJ;h_(PV$q6i{w4JWVpGpaiu$L=E5vN@~-Y
zmEj*EmVseVdxH|750p*0OXIE^w$Q};(;>EPq*3JU8;`})*Chi1&l5g;y!g`FUuTxWh#<+kfo_uM%Y59nw3wiBiC9qqO+m#Iu=GlGKe&$pIdtSZH
zt`I9i!rZbl0LhC_^QlZ2ctX
zPYmtK^|3RJ4Q;eMpa#>Q8Mh28@58dcBHw+&4U+sNd#Y(}E8tkV;vBx|G-nzuq6COu
z{nc|A`05zT8ljed6enBsOPeLyxrYd?aeu#WNFd_StzNVM*-_*+>FUW~R7`Qb{Ll9q
zq2C8=yYr}lb{?yv#+ccXqGbDXbo-}j|Dfk(TW+e%lTU6Xs$DSS?8U5xZ9TnvkJCq*7q|dL
z(GW|#-0a+Kv&P6k_PnbX|CowfueWS#ueX4$VEohQZ?RystzIztAFUT|b!ppZh3kDasHC)A
zV%j!X;W`7oc7-eNjcUhTvi@KJE-toN2iPD~3EeMFavVoS@H5$h3P58GSU7<#J^cW|
z)GMI`mP3e64=^+qg(dKWWseELHRSLFzT|WcTwx7=IXsFlF^{#Yul)yykXElB1z0_?
zjMp(OkQ^z=&x&~EW-L~(#1eVZDhHAr%c8(UrmWV%P}ww^$VJsU7`_pYC~{IYuYFEZ
zU>Q@@nw4_0j=Y!|@xlvh+zbkYi;Lup$<4N5t7V)*%c8&>Inym9)2)u_0g~xK!PKG?
zWN7$*UXY?_)iJSY5qle^qVf2IvZ2wc7Xy|2iASLUEsk_&*}$rQ
zNT=8GFrJO3sRSrC+ogP-560*D)qD+S;gYG``jesW!>y5!$3F$gZVaW%?l6d=CGz7i
zIJ~woG9nt&qCFk&@~OoVu>w%f9u^B|}<^D)^mNn&Pa-2q!T6OYp
zOk&gF+L-Tkl78ta4elxm!Llo?on4`S(IH;__5}%;T7fhJlJ}s>j6&08ghowUJ26nf
zWferCiaE?ny$e~H!|p3{0-n&UvU?)hNM_KadW_RyCOHI^9Cjr+cOYxVs^yRiiEMJsFHfNQFsyUnzXXkhRyQ4l+IB$#67(?o;wa
zL)Tbf()z{8eRvJ6JZ5x<(HcF3U=dCJGP${#X`kF4G~%%kWM)g=?hmCa|=t?gsBw*MW4L2=JX
z?UI;?K-{R1w1Gd5B-~)*-DCrQ6_9G8fqJyz1bttD!$%FH%H((JAhJA$Ck+g&4=I~t
zwAd`kLFhut;^JN%>K_H27O&1uMGzz!ap{uH;LQ%3(+CuNvKb^|PGDwFa_4%RO9_VnR4EVM{UTq_N#_`Z;Yx5+qfjXAd=+kWM>Y0Y^o(34?v>f*&e>@t7dlGmrm8
zIzCaBLWUgzYYn^5O7gsMeX*oCsv?A1UXY+ER~L-(A!>2{X~aX!P9BqZ-^IQaR2Xl&
z3J@xgw}hm5C~vz85Gt5=NdghhyI+aeE}*xF-6o`WzY?)sP;avFVUO{4R0*#0J7Q`#
zMmwdFJ
z<1`H-nsf*pNaFyaCJEMGo}maroFx_yDA$d)h>{l>4SgMrq*>ac)LUKJwn<20aN7#P$mCMwBVlt?eIU8#bc?mgx
zyL0E|`vPEcEIC>}?E?KpMe67za;#HH)bao|1CiH)=g{+ik+m^9i$Kw+mJ2A|#jxOK
zGxBn8*G2ItviZ9jbynRQXsT(!!;B!>EesEpct&f(;;SuIK}Tx9IxEENV{I3S#~MV#
zbz=5!N!W{E2UcExIi5N&i1>W-b%~)8uMU5?KNz5r6o94*D%LRd6Jr9D`EsVbqEyu1
zsOjo43)6h1By}4^uwubPo~&F1T6vp%IxQjt>@aQ?JwysMD~K9-r<456zhx*9Dz~jI
zU%dDk`;_|@Ek%k~y%+EObqI3c&PmJA%(CQvwuXNIDQ(Yxjtfu*N7vtto>kGvuubwe
zb0
z$BuyfLrOiNnRI&ZnH!CUH8IaQY
zx+#~<>(`|mylJ&IdyUy2Q-VheC1KiPaxwnU-E-(dbvU{Ns@_@X_=PC#qFeY@0RXG`X
z+kN1z8$pk=m-+kz#OwsQn>%Ky8Zahwip88v?#pr4>z!(D!##N9
z|Liplh0!1XdiS5;UB%?uKhvAZmPZuh
zmJ%FdZ}0t`{_S5d8I0t&f5E`Q91m+ZM%j<90J6!xWIy#y-e*Y0OnL87?!J!x;`wWT
z#2AqKQ3`5JL{LXdsg$Aa#~LeN7D*fo}{4klY?trwfdnF-!TzUr4J%o)e?2;O&Yu^O~3Bk-}lFV{pD(P
z+#h$i+#;YRrq*jsL{h^N@DW-s$Sz8E_ZX5LO_94hN9Olr#Ezs=8z$qp|DH6GXxV*k
zTNImNHg{mwGyP9JvV<^WyhaB5T4zM$)5TtT{>A74nt?T{P#c*r;>XXv}l
zG^@{$nPgoQqeI84h6#s>7+}+@s_U)nuKT+y%R5lPI|dRe#6e`IQlCb-kqwro2h<8H
z6)NnoT9SPpNw%S+E=-?Nncn!x1$JTn)B^PeP_7Wt1S*xuji6i*uo+Z;SF{#}a1Cqx
zESrXO6jTyofDB@}IVA=LY~AS~hf`NMu32|hkys_IDo-4Ib6@3EC!hA+kq4smhg40k
z#BrqP0+V`c(|mE@~6NQdHA?rulER;lpZNHsaoM8Y@&
zROPEI!$0m9
z?*6a?Fe+truqaejMI_tsl3xLNdnKrfHbU_{e+aS#*fg+{mICg7Og)8+w@%t2Z&k?6
zP;P#flQSy1Ek!x@v_NoVuooA}-BXGL)FT)sP>#gYDA_h4LLg)Vg5=CjavJ?Idp#!k
zntI9g#+FfAI=%wr`wFz~eZ>Z!oTb>M;7AxaE}f@Zm#*Z+)=jok;+zsrQes?puY#i=
zR}4UF)8@GH98WHPOk5nFu;l?yadro9impiYD)4co1E(IPxY<^n^aSs1Spw|$p7{BR
zIq5F#hTl60^(5j7aOvYzW2|^%d=ZZ7dCits(#Et>b5C?p_Hk
zr^em)C2ibcZ7hYZXm|LDWS(fD4qDPefK9`z62K$!OGyZn)|Q
zYY|z33KeC&1QpR|@YNdb@klGP9Z?77iDS^~I6ieKWg?QO)!;|IA%7b`U;68XCVWxL
z8Y|E^GpfUX$G;BaBOV0blDIJr{%_{V1?Mpf;4IrvvuLEP77c)FCHPUU#|im
zI=Li@0);LY@a#s3Kl>+mWtbMwiYn=?p>}E
z_|xlkImiZ9LgOo^MmPsb@`R%()HsB$RRe;lU^_nH58X7mUI|XOHECS7_O3zgc%
zS(@R0P|2jq#{E?g#qh(0Nsg?nXi`1L^#R&X!;o^WGEg`i+b{vu@O9z%TfGumL*1G0
zKVaeI5=bD_4n5UAgRk%o$unJgsRNckq2jHxWfFe%DY;@Ye-mU4${Mhsj*?DF50F@1
zfuP8;9z)mjQ8*8F$fMTaXK|=e>yjf+NfT0kt%$dm+Gj{~d6(P{nmpK*7K@_LDOr)t
zpqb>(%Fw1+G<7(g4G={Npxs-vBILLG~?C=w?RJ*5;>1{P1g?Xt>GC1hfqdrZPy&w^klGvDcd
zwghrflazAEK{#r3Ih~m9nv9|ZiI@CNN=`Zr|4;BlA|OcOq15CZXGq$-pM}8+MbXu>
z>-z{Q!2=f6B$!ZkV9094YX-sx3VBI**$5JzA$G+1LMaNyd3^unn*3HRr2Ba!D
zm8vl~jNVZS=FI2T&GlhEA7Do>(vgRMM<0f{kCF5(B#M={VbtUr5GXy_TPHSwcbAW=
zGBTZJ7jR|7J8G{#1?ecLB*ZPE?a8@0CHcgwQqkHY5qh4ExX(N(&+)g_G+S|)u4T2mOO;`i!askR6P+?n1bkyi0w+v?pGhX!eB
zn8e+p_NSR_8p%;GNcdVLouT9eXw0uv4>}>)?K*9>5ZKDhSj3?lqwOmk|BypgOMp7r
z#qgV+?P8Tc9F}BcNRLZ0oMbogUHlu8bdqDLS%J>6v1!$wWMk+d`T~Z3yH;1gNcok0
zgtI)89MY*IUHo};Lo^S9_#Temno8g`TJ8gXk?nB<;U8y+8cI;U!LJIk;~tdkxL1?y
zgEfL|%EJUu&4*nsVB;~)Al00tdjL&i0;uMD>#W1(Csfiu0SCDl@vx4DQXv8&kb|H@
zqD&t6VrEUmI3iElRJezK!|t^^>B9tHTX+=#q5(vdA)?JQO`mHW#PyLByE>kaLeLd?
z^84YZ7k4RZERY$ABfdt5s(K|$g&d7f10W=IaI!U#nLb>oC@spq`|0`X;N>w0z6Jia
zJF&JcXROG`dBNpsG;j`DatcQ|Xm|bs1wO$a#iPXz(zcmNrrXneoYIG
z3N64v??hj}dEh&Lg;u5`{f7V6q4Q`DRGd}W8lg)v((24~O$!Ez&01iB2!V$>mx5Q*
zYh}SiE|%ge6}KC#3#nVd8yS^!%c97dw<#eB-Gl^JB%xvytq^y0##}ZVQ42yL2eK+{
z^@p>9%&vMyWp*0%kAlp&hwC<6uyrZRC?zqmVfYs?*H|roCm}JB10MWcE~s(u{mG3`
zy0=T(`((4utK^C7hdkz93FJuO1u98)G+|?ypF9Qkt$#Six4|irU8vZEc3L2+lDk0u
znjGz%lhc1ujEe<|+nR{jIu-f$SBSzLXMh!acxad%MxbOWX1m(RuzNBdV{vS?099s<
z1rlRf-Q|~mHR3rcM3J2=P}R0qV(M7`^N5ZJteqB!I?60Fz*1^apeo*M+cRALP76d8
z<;vZy(%;qR*&C(TrjZ>_Wrb5T)jUxX=1C5zO7hw}9S@Lc9VUqCG^Q?TS`-mzCifX=
z#y6t{q#@LZIfU(IOI*jZ?7RmeyI^LM{5PF{ktlnA3|&9EPAWTS)-cfNIg)9z>hAfQ
zKvg&P$6f4^JyVccrUj}q`?{Kc9nZob4sLxqjKd`P?%~L8U8kk_$Nek_L7U37z*LbQ
zLWFw<1YUJ43Yf1u)`0w$&g6O2H4TCaRYzVdyD`
z#f=1vcx;`9tj~hvw`p88_xiX3?P^9Tq*Jyqp{*%hg9!PcZr6YSIyyG+)|P|THsczR
zoiq;y*wM2?`lC6djp$>1+Wq~I=p$@pY-0@PslWcX?=A&IVhuZ-5v%+sJ?
zmCR(4vkU*+rjeE15jk6X@cb>LXIEe<(_Dr&fc5MsU)Ffd=8kN2dZx<;XQa%1m=0JFQ#0GppAm
z26}B;ECsIKZr5;3w%ZVG)tl}b}sC(n1oS5Q+^!a7*KVQYQ|EkY)KXmOVP#FX%pX$fBifPUO0hfHdx
ztZKfOmKrL)S&*ZU6j48U(-~`j4fi>D!v@!2JMY6C&FR@EZ@Wd|F0d=&;H
z_HG;o2SCee_Z1Wko0Jsm(*eGsNJyGV4$fzcub+U3n1g
z#W9*;mc#SvlITT$vA3F|ZHS@>lYI*!z)i8XEoW$C{pC=gQ$+(6MXz5E
z6Y~+IUB_r@yB_#}U1cAmsi_W-n;H5TO;e8WT&86gy8bVJ*sgJsLjtV&tvT1>7>=KeU(E8E!pCK!GiCV*zC|~NP
zLzm&#k%x3a7n_WLGel>mA>FRO>!PVQsw6pP3kTx;bk$PnZTEqs|#jzD*|3YA4PR`iFCzYk=#zI+K9(C%>(>Xb^Umw9xTOEF4acI
zC&1ZBjAok7j(M$jWd0~hj7tK1ouR;VtasuyFM~bsLR*Gt+z!seug!snO}djPf?R5&
zaT`{D%sucLWQJ&1G%@HXc@T@kV;yrhv!=0bIINUIuAKnul13?^flx%-TA>cNs0fAFzCHZRy$1-gix43h+>Nw_GCoya#
z0gxK$QNF5=}%rN6V&EAf~mA2&u67SIbI%6Rzo$>3s6=?+WYRM%p+7
zuF5Cu&fOq4K8Cmf9v158?ya(EEU;A67Od;p^?kJKAZMlpbYK`)vEC9N#&i|k~H
z$p3a0?@8q|x;yXhN$iW(j-%aIfI8K@dO^)|La9kD*1%2RwL7y#tJW>^|PiDGLr_H9L96~HBMMp+NK2qlY{emnVEu+&gpaLu9w}k4{ONLIQ67;CCj2H
zYDs!H#$9IDi07#*t}9_S&0!>J*un&j|Cg&FS;U24hUO|{^B?5c`q=tfH3&O9FChljgh
z^~*{6E5#(oHDUkR5qnjQAb;`vEqse8
zfQ|TncEw+n{&d4GeA(xJxF+mBJ7TX&`&;LJjLc`lh4P01EwtlBQE8mnle)F
zMXAR(i1_^`S9kMS82k+-uw>xTexb>hV>*?rS~V?3RSH2$q3Yp(hDtZo0*^wG;m|{+
z4ruOEAx`TKt0ZB!6<~vvQc?zQP}5?1U1a&k@a>(kO4Bh6Ov!2
z0ncw*w~;z)phIf~}#ELci9ZN-PJ@Flx(Jakn+4fE!N?uHuA9&WG}cBI;MEJsbuetncpws(J47j
zb33^Vx5qusGN+>U9YW8KlU=1<7HzMjNo0-#T++zhSBQ&ZH;g;fHB1Qj769OpWFU?N
zn4@UZp=3CJC7s42DcHe0lAPVpBM}5tw+dqomPLU|bl}c5Zx7V3Qgy3%)w+kxz=jEy
zybeIQF&UuVjp11#bG}z%IyJHx#G%U2YB^Me0o7Q+LEnJzI5XTI50mwRTg#fgK0J~6tyvs8)%m&$-1Rwh
z5*~wW3z7O^B2QRo$q}$e795){eV2+MSMc(*Ww#YA&l&xoFKP>y!U!&jG
zUQ%^`p7eURC~kdSy{KrE!b%p}loDD8*cDYw3uH4P`?isGMG1Y3i3d+Z{6%^lO}3G8
zT0AkpzV|NqM=uW7@Aj%DLNrO^P`J)bgR#cTM5R#6Wmx&|z|V=4t&
zhebg%BA$=_FW^~=0Sl_8>uf@RQ&4v{fD=i7$gHlk4HI-{=FL2>LIdQ`pHoq1+|MZ^
zOMPfd9*a6FKZ_#4S>@3x$JrmWA`ONJjn(Bt25VS+B@UT+JyG(_C$#na&T`GBSsc&P
zR6CH(PC?z?6
z8@fuqgUTGeuxY_(Vz9KzyXk@S^6U0T+5`ai40`2t`U$P-V=||JLuLGk
z=f!rrNlIqKb|9cOI8qas$$+l}8cFS{K%uy^3ZRkRu=s=Baqp8`d0^e-i3qHxRZ4;A
zj3jEL3P~iszRiwAB}zMT>}4uz8dHgXSZL>3qOwk_ktpoIAChIIkxwL%2kRF}l*xP}
ziH@#}iZrtt8WSJeCFJA(Lu}~d&~!|8E2^;or{z-7x7ZsdDv*iNtXhg1p>`08I!qTb
zMNL4v`9w+jWj#>{#C|?eMXmKUQPd#}{)U@k%Uza|^2&joYW|1hY!kQy%c8)44P>Mn
zha-3$HjQ2(opC{xTP5Tyobu_h0P
z*xP3z8`3uE&4~+Z^^U&x=MgP`4{>-)jSTWP>t(|#IJ_=Af4fWmb`u8sYjpBfsLSNr
z0~M^PXdjT9mN(#409SG6(J>-6FRq$1SV-;Q-I9&gfm
z;+XGjdb`CGvLF}jIC|FBKc?q}IJ@Z5yrl*KV<=I;@OTv68E?F*Fc9{INiM;2A@&<(
zd7lPSG+OXj1Zg`e$3p`{0=&;WPxKU>^{p=4)~yM4tGj0Is@=1F?xwkQjBgN
zbI`{V+5Dg-hw)0S<+lH!QJw%1>(0jm!ieN_IN^;6EF?}j1b^LuA>|3XojpyBE3=@h
zKsFHaRxk3}SUd_@{7L#arBNpPZhS4$i@+aC?`{gSU@ahN@0=G@k2)au;)S*vBuktf
z1!&S+Tu0uE8b-KXiW!I2s!+OsyS$C8!@$|xTfC^inFETJfmAQczlGIvo%0iBLz?5>PJ
zj(nuS7Mj-`#2>kfvP@pGO6i6
zVnHL>;f?CA#uXiPyJ|a<$$@_M!aOPDBD%2T(2u&%Xth{rgG(e2(uAC4CCYN>_A62g
z#gvk(Kw>beM(VFGnIsw%x=0V4v-4%}I(bTdbgf=lp{V(p?#5=sdfqJLAb{A_4xK57
znF0zAE~_Dg8B>JXdt-oomiR%B$~jGvFGrTqaDD}DSYL5Hlu=5M-i+^#S;`p>fvM!&
zV`5erbHb8nt|-B1%We8mi3Y^7$jTwq_3+W1FbXX(hcb2rbikPZ_YDv045M0q*5v>IN->x1Byh2%3wI~Bt5Ifw%dg~Shan!9=E`*Kz5=LSF
zSVX7xM;&-l7^76dRE`B`5b&;LF*}kM2
zU=C7})M9Qjl2YJ{>nS#BzPfPuZI9>Tc<+K^_1
z_5CznKmtW%HUED9nd6929M{libK?B{2ds
z){Zhy?;jUlXD-e{B4(Tc7IHSDB*)DN7#fBZ=bQn;XFJML3`Y9nL+HclHzTqu-CBn$
zp4eiEB>}7uYt3qghz5pLCWe~tb4p^O6Wpp~r$=(T58h>em;x%ETylnSMhe;g@X9t}GXD7yj({9g!_2T7&XW
zKA~1Laz{t({p5`%DscHxl+)MEUU+wk5jGRGAUw@!T#nLU3(ojMWr~!9P?6Uddk6#pJF->Woxm2Z3xy
z|1&Lh`)3b!l66B7ptB6^=IIhg>V~shvCX&tgoK(vLPgb7`fq;|6#=0#KOf
zfvjm2?MVR^sv)q7=636GCUDW;l26R4|F~?;h3Y|TyW0Kg+r@r(UoGXQy5Dd)(WiJ9
zesSv^Fl-fO?eZ^;9xw4KeLK5IMf}T^Eol0IS71QZqz=-w{s8
z-|?6p)MTYLTdwYWxkr@(t!^V28rU)v8}rQVPFz5k`n9@OO^y2Iwj-vuA{VzJGL&bf
zWkZ$TE@^$ILjP3uxLSReb3t~SQl$DGSwfg%dx-QU*=SYi_Fg2&Y2Hr~g5`d3%?nl1
zoZr0SxL`~AUE6}nU3aC`;NGAcJy~I2c7D%Q(*67Hn?a62T`*nb7VWpnq$Y1z?l-Ke
zI1GR~N$H3v{-|A{Zta&eLn@>*lBTB0D56PAE)&$S9jq7BEB|3r-C{pHSH3ncfO#yj}ccU5T(7cygJRF_AGYG8>$l@Ki
zQ5fL_s>~vD=`VF$D98T(Y%q&%y8y)DpoP4e{JuFqK8Jq`C3y8=NL6DInp{#?W5|e?VTLiKaTC_rAhDBO%WZ@C|*XfA>e?#
z^r4$q0_IcUbm{s@{`TzgRc{qED2Ek!(Nl=K{tbCT5ib471jfg69%YRcFHK|?Llc&)
z-MqZ9Y`?or^w-Nzu@r@D9hq0H7*0fc#1uI0
z`4%>7`FC3-n5r?yaU8^EJ
zF;$tz06dqxqHU%&yw$%DR;t!I$}VMpJFEO}QLDGZ=j7$&`fFdpA=#Hvy$k@0(nDa;
z+v+#`)f5M?wf}Z-mP*2_GlHmc-l$9v
z!RiK*|5aw8W_cC7U}JCk_5irVq3dJeEgNJRoGpOWr9Ts!lD%np;Er=coJ
z=@L1YSOq}n@KY1VoO=Y1nPHtySA$}#EJ?`kfNaf4$4VTrG|CBzO~szg#&)VE!3k)86caB
zZ{9&hKpG34<^RGq%pxu)9eTuK)BDKFik{NYk5ay9N#Hc6!&F(IjrwA8KUv{Jc#GV;
zBW4ZfNnmzW#uvgUE_zA#fUxHU@CZPxD9XT{K*1SpfGREyv?T$<>XZO`u=!~NDf1|z
zXUP)Fmn)}o3uISfX#skL2s=<`i&Cv4YxWP(89&gVUH6rT7196233N~^gY8+rt&T7e
zhvCJCipXy;MiS4`!X(TBF?#*wMk8CnGRv~OZx80L)D(;&cR$wl(|
z=7W6}oh!^j&UR-&<|h81D3;Nqqqy10ytdUi{f$37NXpCrjNO(Wsnt%{^dxYAi>i*q
z%#eJ|kj7jae*-BMBbVm%gav?C2FcJ#Itmr}qLAD~4v!){HW!s7kwif@Cvi+cVyb#H
zuHc{uY9U*Xt1j_%H8P3pB<((d2OZ>zwi-C$7DO79sIp6=4GrkeA@&_2&a??Ls2uX$YXM?FdJ(B?_^trOe<7U?9jE
zVP9Pc;FC|it2#Rc=`^B>#en$}_bw8)Xo;F2)YT`O4B5C(Qb#ah93L22i~$^n>7>bW
zcc>0veW9bO|3+5zP6Ui3Su<5beHHZ={n0>Nlevx7KZ0umL(wCNg<-+oq}H`!z;Yx#^-j!-X(DG5Z{&A0(m+AIZX
z4UUMYYhj6AO!CQeSM;UJ8d>2eq*~7e$)N>6TY%r*R>1aIXz#9A3^{Vt#*Y)A{SI%t
zHUMD@tLZ{>n2>|Ju+gvk<4W^SsN&}^k%~_CyN6L*$p*%#VuxEVVDI4?&RTd_M)%BKBajg)$T
zqz^mr1$i(GG9#=NTFnE+q&xX4ol~yJxHo3dtzPohr=NkcI5j~B-AQD}*y+VXp4{o4
z-9`~Eh4=Jdsev9bCM@!f$lgK5GNET}#5Jk~OW+C_Qo&O}=n9h&v=0oSkrj14;<2qQ
z$n^0@G2GDFOD}i@N`~b*m!^KusdXtp?5R%9ZeCIFanZX
zbkNc6FkWXX{}~?rQx=)=ioVH(BhkJ43vxFl3-1A-Ljddl({m;}&;YwU_*fwu73%zpj~YrHmiFs{`G{$&
z`H_QQEd1M5pLoM+VNT0^aCFvzPUl4raPHF%x(;LOs72yXyK^mJ0Eib_=@m&jR#
zuD~apA8tGgf?`Z6HA-sfkR_2~1gpBRCf8or6gX{~s7Nh>XePZn;KPspE{iKzHW&a}
ziJXlDS*h6dTeG=w0Dz@#H&U6@_={lSaaX`<3P;q`A^fs#eEf83gZFnHC@sjbg5K%w
z!3224D1p(Lkf5lJb=rbxk7u2&<9l0QZamOO7D#L8h54U-?8p?K4%M
zLIIuy{07V2-7*%VL*HtV|9chNYk$D-ad}W-o25k#m||KpE3b4~a$)TLoKDfNBUcxP
zin4+nX=s5S*AKX@2P%^sASZ(3Gwi)w>{g&saK_dlvI~s)AfDa~How30O%n~WvMSXH
z&%%HQ_XXlj*5;y=-MXMOc`ImmJ^F$uZDHbbux{70K^S>2G{|?c=ZlX(e!vPVB
z0UM3+^`={{xJU-cAMld0=9DcXix1L7jL>++
z;{7vgGJbhjgOpCvRcBz3Nlk9C?S)ZCMp*l*#S;l-W>-kII3umt9rnLep`SER0G?7x
zNdh~seo(TRzX_`!wq4w~yh|#!nYnddfRid7^8`jN+~h>m9EXjxir-_lj0xR7636gm
z)t`5Z^2yd>uuR#&OdJ!}B}T5~R5oo-%@4e9&R{Yyo5=zi4Kq|^{Uj#ArsV2SX=k8r
z%M{Z(7M+oZSJ#~5ZIBafe&HZXK>teo&`R=n9nb9StodMl9E&9WKfo__p>RPd6-oGi
z1ReYqmxL7I+~SC-4uU|gK9?r~uV{CRd3eUBZ#qjUj@Y-SIK)cbSJET?!z02YAeGyv$4^{kIS=uQKluU_46>UtH(jODgn_R3`R+@zNM39f
zJPglNKL1eBOdi>V$NrXEn%H8X=z8L!O2Dam_Rsv9P3J}B9aChP_oc?=&=MEBE&sZe
zFSi~g$EfYRtr0(L>!oD^yqO0(BjGU~wAUM=g;nq)6nwFJ-m2^aHXCXPopf>3h&Q~D
z2|isdxxC0DijWKW(-8|ucq;2=pg&fJe_hxABorB~n(7g^Wi6_5%jHC2ODvN}SKzCo
zWR)suBpDwXU1!K@yOa+XYPu-ToeVXI>`;*^(t7YD-IqJ2Cru9n{D0qT0=c_P(J&dj
z5}hSTu{tj^55JCT;b9?
zOQ5Z8K-07gc?%bANxM@eHOXux#lpH}H*l(HtM_rMGs|+)xE$u|X*>s(2`fOxW{{ab
z{MDi;L3Ni@Hmvjk6w1zQuD=uXl8Vzlqu2aey<&G+9R~XxB7Pc+fm&Fv7WrDz5BzAj
z56!bXmKAt>I8{xcmHQHyh%K7Jj*{LfA~zn=Hq*V<&5||sn=tn)
z6jq@qT=P{65RPybWh~fOHOr_(lCpr16NN(v#&$;syWe$iwdXIaHYa(1xkQ&*WRlR>
zx;{ER2rWFukoOT}(ipr5;f9w0&cK16!+NM#hN!h=WE>|M!{@6^3MH4-NO)@z8!9!h
zP0|C>BZ^cD{zLAy5!30#w{j#3t;#crEm*eWDX2gIl?;JIMKGYvG9qMYwo9B>&vo@8
zvRjf^iflQ_1pP{T=unZBkb&9qni7$#i(9r%fv0XcctciwLvCXMTvh~Jj!~h5&QV57
z-nQVjipRmL1{QK)e1apvX+_&=G~xbh8WXhTO-)P0n00C$!c+jNLcbI{!C$+0*pLck
z^3I?D@yK^FN{)5&Wq-}qo=7vc^@%0_)QtgFx?VILLuwPbR!9#8iI15IC#I4BBI|zU
z-GHU?=42kkuESPc>@(3>e_qMPGC*8xL06XqQ$$Q*szaARe-)Pb*mw_Gk;+2M7KSmJ
zgp%a~vN~i61-N<(5~%f3F%h%3iKfCc6`41{z36};mVo`|aHZ`|;$iqCIBfdc{qw+Y
z?+UT2(m_}bCs{0jmPXp?eVoLtYmA^uqmZ_4th5=+&e^J1w6`mQM8S-G@7^GNw{k-U
z@6TbeC{iH;Odg0vG1sF)6OL0|_gqCxB=T&+Fghcfi9P<)%a(A?e>BEA-CVG5vh
zKNg|W;Y^oHC4S&?zj5;%WzHcg6>RNb>I?B(11BQ!+HWZ*l~^KQ1C#^HFhLHIG>`zr
z?ac{2=wz~4u=J|)iKe7qbQJ?80=iD`tRi{Q5TC#aEV&v71VG`;d2?oiv)v4)bT>Rb
z`8HthE3dn7itVAD6i}{%5?gfymkCIhVr^BS-ye|RD0pKiIu?eY<4@sO$4g)j>!%>o
zH24e#`!#5DukDN6IBf;e(YBKq)(fN&Zy9ioRsVev{%H=*GbW1^rEN_SjtPcJAb(0^
zdb@|VCD653OBHUChWDBs`p5s;ffHo(4|}oBvos52t<@iv!$z{E&F=Qomj?iPdwD2y
zsmg%YD)6};`MED-^|O~c@7*J*uAbj+0b<}4KO|kJ`k2mW5dOdP+E0&}LVrB2Uda^_
zEmlIA-)wHiE%~S$EzpuA%R+)F89;1|g&EfHk>xU>+F_@qV$m$3rV|=r#&4cNr9OpP
zK;w3D;>(r>6VXoyN~Wy`^VF^CF*l*V1bHD(AEnUiLRl4C&gx>u)5P(
zG3%^m2u_Nmt9j`h+)@|S1J)A@j4k7nJh7bhD_S6sO}?-kzI8d$vm0RKY_#p>z%EU$
z+W+FX$(vcUEKsSaNDM-0`vRDk
zA`)B3{NCRQ3f*bwp1*90LS3V!{)i1TnpFGSx`lo&PXesogbt9xv9E3DzZZMYwoh<1h?3v1ONws2XE
z?N|bSJ;t7*v?)`hRZL8x+d?HgaIbwMFj_IEvwJvHPKmAYsYE;B;C@a>u<4APWE--T`ZqXf
z-$6%+({k#aV3J@Wj4G+)5J~o0oU7im6rxzPOh9{i%=|kXf<*jG4BHMDJ3V7@KW|te
zkky=G@Jjqg?32DOp@4EO_kUf6^@r=9Dbn%j7c5}BK+#!4rEBFi!??XJ
zmtIClP-}-Aks>BounHFcAkj>>ut}gfz^V2R!8h{M2q66Fxp(|x=0MJ*Z8(o?EYrrQ
zRG#TOK2G48wJX%<8mOt=?^PQpn!EUC3m*qko>BY-T`|%Im{dydZQZsIJFw+a+kltyBI2&5#HhUeQvElAV
zc(<-0_2{$8efz*aP
zKKz?Rr0D$0zbb-VWC0K?FOFNbf^1`G(@X{*E?YpyPB5C|uo}%jP+z4sj!bfhIJ{pr
zS}v$E+}%Q?Fhyi~)WiYZI5AeL%BQmDy(jS`&sYM1px>j{(UM
zt@s=5h{aq(#gAC&sn%+Wh5Tkdji{HzF71cUyFCwR$3xk2`=RfAxcFyc{TuJkf_?ut
zJPX^ri4=xj$?N^Y;caT}N(%ZLe=;tSmz8sLv6&^hjF~u2ZS%LJHxmnKA*DlM0D&#I
z$Qu${`arpENNb9p$}I&-`CmYGZgl#;
z@%`Y-oknuf5U{ywED2o%3#oyVW=>re^qWa?W8n2U4YKQywV~j%1}EOpXbgTz3KEx;
z)Ppa8Vbja~X(-f`|2k{ArIKQuKRi-{oVfvf%O8i&O`=jBb~bz;HLUu2H5~s{%u;Zo
zmPe6>u7hhTxGSV&QkI!r)u2$0)6159gcqmO=355WtkAvD>?HzwbdhBOKfk(mN#x+V
zfBLw2PgqWZI0#&pL-)|UbyTAK?Zyw?u2hnMYNiJ~yQub-510BPvyS%sv`E9#kyqsg
zVEW0OW^@{IH04G)Rq;sm0eP>rhLTkN5w^)hMrZi~a06Wy;8(TmIgl~EY;NF23LtgL
zKW4<~0-F?B*@+UV)OXC<1TUnlX3oqCEAe(0gGTFJr5^xUVO;W4(0>ZEu+#rxc@)5h
zFQNL^545XX-n*}#m5!?p?0vLzpP#|bW;HV1We{>~>@5g33_LyM0&ca0`Lm?H?h|G6
z8ueJd&hPcx)!Eu(n5`3$?DAeBg<{9=L!N|Rv2~0RXGcctGP*-s7x>u#0~~&Strp|{
z3T~uulby=_5|$=$3k2ZC{~rh^
z%SZ9SKwI-d|DEgyF%yWVuJG5bs-{27F2YO$qCKuM2Gv|i-MWdrs{Hec%aK>N!!<4i
zsJ=vI&Y-d6qDPf=U7unRhCeliwkc{s
zU(k}0%L1H*%^iNi9$wR?2N3P+q93p
z=p8xyO2+Nr{|CmcXv5^>^&Ib6D4e*{R|@m9sxsZntbS`OQ(q?=#QgsMQ90#pAzm)J
zQCtEio~jMo|6n=S(lxnd3#8@$f#nM)&HjM+{no#NC6RX-e&1_GCo*-S3C6IzF|gT%
zMiwPS5zb9jMzOthl^KuaNI*>_B1#JZfz=YeXR{);S|glcKkXLBr|p^&8(i}Pj_K%8
zAlV73VP`Y->&ll#l3$$#-!{U1w{BgLU1K<`(pjNLJmrDjQAUgt{1K`}_5jO|>=l5q
zZKR8~;MoEc61b!=O$4S$Qs70c$wf65jN~*z#&{8$LfRV7Y*@*?5B@gY4Un^2A}O~%
zrG#ZXS0-*F1@#~Ty36cuG|0_4@aisbG9H$5YBZ08`;+p9>dZtICzQOF=8~Ps%LVesF8|*swt?%vAlZbusfG1~
zIX!Pq%B!vArW4X_beYYIdEYk+GM|EULPKmkw%6fd^qE+!ZE$NcK!%0?a)1OVc!qk?
zUX2W6rR=rQzF6T$zT_$hWbk-~96s~tiPJm1ryB0;bKyzTDNc!QXQ!=~DSCANJ_+sd
zS>L{w;AqBpU+Q1)63MYLXZ7iQl=C|m&!Obys*OXEl733}QLn0flG;-`sa9LAj^5OS
zy&+FmZQsx)<=pWUK%d}Kb-SFm-TJyawRg+lX*CR~d9KTdvzD&o@4Y769+{BlQWP^^
z$L~(=edNEr@3FqwOqM*0`j(O>&gD`W4^|kPuATW#uAKa#3|GP*w$B?jGSby)Z{dY6
zbY~P?DeAL-KSQJmcX`EQ_m)<@^jSCFyj23ETPaCb%uIFx0N5?p+lP!!r&zVTi~t2M
zyHSC-JSO_nhe`xJC-xGS(xgmTsO(wndN1;VUYlFz2HcSo
z$0Zi#oLD<*Lz_dgTtscF{L#F^1}Zr}e?`rpbReO3PSR0LcoEVGfxXras?s^;8eOTjUI
zf*?c^-j)aiv12`GSU|nIVV>4ViqzfbpQZ7*xB(r&sbJ9W(dcJAS=_p*Vg+Xr*Fmn3
zf1oQPvCiDQBipj&S|rm`Y&~QVM=81#$IAX(d48iTDaNn+SmE;xS5woT=)k&gG1xVJZp>KT)8SIgmZ(kP3^;hhBMH(Fw!>(<!34S7Cf<
zQkKSX7r=%QKK3_V>_5x)kYZ>LF(ELIx0-32+2G@_d3EaY@Ok0w>(=4^X7g@$zuL4B
z$qTn|LFx0LauMP6(C+98Y1@YU+u4HO0{FQW1@#z4Zu(xAUWQcDHfZ@R;G#i0rNDND
z>W>+EOY@SI=KSm2K(C^R>+mEdj0+i*lT}D*!#RiK;$<^_4Q(L8vX4yrb5t=M5;KG;;&axc3XNGI0?vSE5qR-@Qf&-!ra=b&`$vMI(df`=v*s
zqRuymkIL+4?#>LKRwFwXqytSCWf7h=TKJiQI#I?<*Qd_t-Gn~CJFIhh)f#QvWekgw
z2UU@{Hdafrfr%ZYSy=xs@uO{%M_rHcUTk2tnSv#dHb|+~3P%9gg?HBIz@;R#pA=^|
zbj)=H(mPk1%&&!uJae(~LunZ+Hzcxs)SfRl8!*wdAO3SzrO7g5PJD7>x1Nxz+IGqU
zk7U7*_r|=;jUK6pH?$Z8Mn8hciN+l}$zimMKPNMFaF5>CA>=_jNwR*#HVV54zuIi{
z@N`=L?eZx<7={OEF{bg%)0k~K#Y5^O(V-oFg{n!yml4Pd!Y6_5ekAmH7{9bgE6{D`
zoBvmC^I?K+T*VA?BIQr7p@{Y;+eWj&OJe8!K75Rvt_=oM#n!O!WVc-hAna}~o3~7>
z)ih0qOi=W)*^U13ebJ-!bJ1C5?YqXAk$yHRr5olzM&uO$R!iQy!B)mb*}B7ehl+?<
ztN;##4k)>%bIyK8GCCt78^{N`hI?LbeiEnr+1DwTg^;jiYy923sUmuyyPs^{+OOT-
z-`4R?F&z7BYo(D%5?tWGWR2t?+DfDOHeFUNbcjf!r)H;R_g*i_>WxHyD#8zTh_J9%
zs^#Fl?)5jIVV|&|fmBtDTL)c^#qJGsoKycHe?ndbz+4%}TIH9A39o=ymi}@UmpU6@
ztNus~Pp@!g_^a`*&tmj}-A3Iyxl_X{>rw8<7O?f8p?w4XCV60Q$}ADBY<8{ogx>~1
zZN?+I=LytEeLFsLF@tazSl)7wJmkRaS`jBAu#^i32iha!-=Hh${fXZzYnK;6e4v0WAmO)ja%9Gyc5|7_Z~mrUR1%Ew(OS-C39T1u8Wpk{-ZUtWKRB@
zq8dnkbKLXO_OBwdHOp4az@1{iWqyJd)>VE2s<16m0+%YT=x;DiMy**b_Q&`IIq5lL
zTLmBhG1z%#2b~Emvi>9-cwUOA;=U{z-p#L<;a;YiC^A0~_xcWf2>+o31|24wn<{m<
zko7`z2@_k#@EC@o7Gj7?8~h0mu28&O?yz1Nq&Gqk&Hp=i6(a{(h{2t$0EvA2TOF~T
z_BV8S|5;yXt5aGD6u41Sq<_(JEIKei-HBVpTa)|R9W%a$#!`=9_W8os^`N)wN=$3B
zV9AxpJ}8;puJ~^G^(Vc;_V#waE8$te5nV+=5vmsj=NC?=q^T8e#Nu{e_DF}{OJgO3
zP#|>ug%Qc|2WA3DTt0SL8QhY2?4pUh*>n6SlwA(v?ac~z1Rb1-~(P^a@!-hq@fJxTg#CnLN!@cGi3$TK<
zjA;xT&N?-0K6v|TytuG^?TB~H5O>7Jg>*HOoi^>hLnpKMSe8cbE`TU3DVpZ-YrVXD
zr@XG;iHa{_b1y!v-RcRNj2Q!7?+|*L1KAI_K^M)4>X+CB`0x58rY3g4;;-yS*AyH^
zU-d7dsh_Ks<_@eW-4~iM7gcFUQb<|E&z+r6Y~f&ylW
zB5GL12j0$zPz&1^+rOQlKbOHu
zqDG^4QYWrAXo2kvIF%sg<^N6!Pm0q$f0y!#=;-{iAuYhlL*-cCS5;>b0
z-oG{6y|D~&sK`R@ByuRxwE3)nn8Dah)lS_{>`4@R8w-Ve}US@8c44Ln?L~?gAUZ>WokgS{jP-uf;3WNaHd#C
zY_t!aG`)2Oys((Fd4|t(PHU+tnJLJdCz2Cct(sPlmuvqt*L-^YSzAMr`wz>`$yW#V
zaARY!JnqbSF`}XhlYLOU?i?l}BW=cw7Y+9GQmshJIND>tKEzFa_1n4JUw92nKdIEw
z>#L~4O6djH^0|39jKCw=i`aTAD|Ya?7drPEKMSD)AZr1B5N_EgnKY{syNvvm+y1R9
zjFl-?)G8`%-cRaT&RGteOI)7IL@r=LTke1_|w)?U^STw{tuJ9t7I(
zN|gds@-8;KenC3$BwJVN2Eblv{{qT3q`a#9zd?)psz?i%Ed8+Cg{3poGVXtlAM?@ch4_^B(uX-Hg~&
zbrMa>NfnT!)@-0ZFX=rNyATp^+ZfT7j?Is)6EsB7sH;0i0APGv6?MFSjFN*wX+zrS
z4T#SIH+vLt2)P$#Su*>txB5vZ%{Y9L@;HqoFrUq%UQ#!|e!@WAGEJ>6RAc58@iG84
zpT`~P%>F{@9tw4cwGZk^E44G_E20NoO#$9iJqOKQ-jx;3luZGOz3WRP+O{M6Gh-`p0RdD*!@>n
z8DU&ac8kp!$J!SpJPQ2&3pad`CvC1zh%KQdNQr0i#mCdJ-5qQan-Mk}<$>w+l}-
zYvTJgpuj=Ag-qGQtc$Z>7GGNB_^g8|d4~=)3F|WN_G`3^)S@`nQmN%v@=4>8%6g40
zm~#!{F%D_X@;84xfNcq}KcOLT`
zXi0m)x`3A$M~#2-pSHSF@yze0(VT{NH{WjuXWx5VueHCvke#Nv^P<`7m)lR9j~ZBI
zO7v8%o@<|OJz|6b1dNuaccop^k8Y4yNh<29cFzu<@1NU4=2U=#TUClWK~2EcK(N<3
zn);eu>nUny-zz-DE^0(4aQ{%*PG2du@h1QYYQm)Oaj)R3{p*bfWctqecWT22RAc`I
zXL7t~pXG`;#+wj!3?=}E^w2hETnk(kimXOlcAE@C+*$$XnE87vJy7FyEgru@Xg_87V
zs;Oe%5ykj0&u+#6?tVW|Y*e<6kI7Cp%Cf|UAZyR&f7$zjw9sL%R7)6qDRK{N{xNk%
zm+9y$X2<{PL^Q()!JyRdWu1)qEpK-x`<3+7`65xwW#3Dg^-bn|e(KcuQv8nd)-@@W
zsHgpEq5!~Wsmq1gNt<@zPL_Y}okS=INEPk`QHgcBz^0m4qMovFr9M)XZ4DLc)nR2+eOcyIy5
z5Z)mIt)5|`GZKuo`(K2E(1k`yX2$cRBWftW2_RR}I1jx>aIK0g3c$Ok&?HPwuhsEI
zd1O(7f1kXpN=iEp-o3t$40cSXBD}e&*9dh$7L(f<=!=H!J)hfc<%J6Ii%|`!M$??p
zbFWa~8(Ct#dm2{Km6IT<1y{8ZAl6(Q^O!p}VB}=L7wLailp>gJRi8cI>1DnWeYQEQ
z0N{<+%$``ny``d!@h0p-mc62ozBp4jA{OZUyAp(_d17~q&K3H_@r)UK#G9u47(7Nf
zCYOWcHmNLYd|JkZP^oapnVjEh+>96m_ASk>^z!T@4$~~boHU_AUi(#fRpgP&%p)l(
z&=~#(aJ}Bh0CxUoK3&p+49WEbUDuS@Pv7r
zn&)canOBsM{P6EsI+By}TekFp25kh6#RzO>4K~Za}3Y
zKr0eE#PMUAZzptq{88W|xB?LqUW>lms`}}nnEcx$>|;@3-9q_8g_wHg1l7*O9me4q#Ye21p;$pq5uLs&UJ)=Wd<`{HYO$SZ;`cg%fd7$~Lg4VByTR*t*-F*fZ6ge8Ne6rDU(4d6qVPiHMw
zP!gAyXpc!)+>NeX3C*14Hc{{}{t4bkG!$X=RqGtSE+Z^1QPR-
zBG^tavQ$+Rxkp(7gq<^4YNx|kc@&0~EQ=*HUw;UuLO@x8@TNTb?5lKzS8rd*AcaZj
zfeKtxz9yQV306p6X#?M)31H8aI#O%DwAecy9ChyK^&t$Va6`Qg!fc;E*LX&L=P=<0
z2A4w1)fWm+qk&8k(_7vFa0a2JC2R}@GsE538-n0_&oow8vvO^Ju%ws&_aEXX^m>oq
zBfKk^oWW4wFB8bgYKfJv08h9Y=R|SYw=Ua9eaf^%p6~@k%a6ta6meSCj^~hG5LB2m
zJImqWO}@=$s&EoS38s8j=v0iJ|G{0BT|SrroH2K8Ur@~ufyk*ukr|7Pdqq5?0q@?s
ze89_GGuayG2vqlFrG{=PbTUjCDia{gHHk9+7FSwLZ{t`tOOEJ9mQ}-9W&ny)OECSC
zeNYw#J#@pWpq^cT(1~7ew6EaI_34u2y%xK6lLa~2HAjPheUp(s_2tPx1lzfwFrGeNUD)!bz@BQPcni{_XS!!?Gwz5*rXrc(_MD~gvjDkx
z|KrZjme`p$^DhnPTq{agiuF24&<>!fX0>GZBvDZ14zkIk6PcjnV8(GG!+((`elFmM
z<~KyN5+y9dxqJtqVSoEPXDZO`?IrE)usv(cW2_B;lzITS(*qsCaeAt3ab6Zsw&Acf
zeow|vwND)%Cmom5L6_tpp>+2ShTkmWwu!4Z^i~bCMZ|6G;kNN!tG`?}QbU*_-T)>6*d~qFTp^$M*-QoC`7NKWPcWi8$jeR>a
z^sjYW51wb2Cr2OJ97=G>AujtA@j~e9x4CcbF{K((b$iao_TVsP<7jjaA!R&8Od{o(
z(q}?tG{LqBC8RT%BP#1~?|7;m^;M#TK9CxK{Ae>-q9$}Su{6=abV9eJk|3?Z$Wr<7
zp(Ih297L&Jj+ByC732EhnYPYTEp1v!xMX_M>2G1$v<-PJKIN;I5)ancODR`p3Zs30
zkmaG497l55p0Wt<-=XRqzhKxLH4UxaH)K_ohtwWYOeeK_iC!FWA$UFz{`_GA
ze1J}vy@{MTN}nS-T@g@CR@3k`;8XhvsyE5X*xWAg=Gv_
zP-d~Zh`d>;9u(D`VoFM9G2?WYPZhN*#lNhvESvh`BA1htF9m=yjA>WKAo4)z)lvBM~vRXG0?I%=%7ef>4+@snE0vlk^q6Tg?j8UQ{E_tJB
zlNM3$)UR8uiQ)<4)F_KLq{M^O17oqK-c?D{BQ+&JQcz7v7jOfXaS2q{(k+06E7vmC
zZY5fP3Q?9NkgZE`L!ogEMaKd&nd*P_Q7al_;_lIE0&NtqwuOxiCtSfZfdwmsY#_i2
zq)`}v3MU%`fP!iQ4_j!juHq+h7wPWOv-LHL_u8y&X^tUXy=)PJwo_
zS#Rj^NaTvjEBZGp^5N`vT6V?iBN(oDQL>5S#rHT)zR6GA!KM1B_R>*h^T@V=W|@l)
zC4Rm!`mxmA6H($PAZ$Azcx6Xud
zMP`sUcbmUGtZyH#x1V=Ifi4=(hOnRRwnIkfm^N*PICSEV9ih7cv>X4BywYm{b6MKy
z!Cc?};6{;$BXtc9ncxm(YD5_B$xlCAp5o!^C}+YLYpZtSZ8MPRY`A}`Mu}Vg!Eoun
zXjmg-TKza0G^a9C0v+Nw?MvZ^^nc^wqIXQY>SNa6gbW~Eerjg3LgncZE#H$Dc=7Xa
zg!A0cpun5$7z0GBHk8Q@&uqj$7VbY&1?j`!Rhcv6tUHIXLoqd;qk{}EGga3ytW`s=
z4Pw3t`u>@$RS(*VGRA)fg^KL+;F^a4aaRI+Gw7nDFk}X*U6E&kxDL`RG+^SdDo?X0
zyuF-0W~T&(&Y-oiG3eLu=zy4
zk7rT@edK#tx9282d-yP1E}1a4rjEebyau$Z)D1JR3C{x
z-9=m<##>x}GhlxuuvBp%Pv2sl+#k4}E*h}s$`~7Di4%^F*&7rCPWCyJsXj{tVd6c#
zmOswkyL`d}#_4!0CNSu?1l1UAz{sUckqIF?DjN=T+4MHzEk3FS2-HMnVq|a%#PKuY?CX;gwPSyV;O&CyE8-fq*NRxJ*haGXjPubd_^cJJ~&)8Q90^UIvKcQvP_ISB}tT4
z#yA>$*UW#6GuD&lcP{4b#(PZjj>lpGgPzqT#-GoZ0T=yqLu0bx=1dAdeQN)7z2y3d
zoRlCV!Xf)otg}HJl0W!>pTYzu#>U1)1uJQIEsIR*-U}f{-VIk-9baFaxm
zt)1|O&e7N|1uMy^>Xu@c#k_zbV
zYz{zue>R5zah)~i-}*ut3MV<8Kc)i&*c4fosz~j;i}tX_i$YZ8`)0Mga=rfB8`S@4
zs8R$8ggJ{I54Ds!wVk}GsokO8YpQ?T&1#n>zz5Hbp&*1TD9?A
zfY-tK0%tzwm62x{7eG&2m4eC|w&4nz6&y-b692^~K-1l!MAf$7sXpU`nN3HvJ;Xa=
zrCOU7tR?_@8=y;b|Pe%6l#A(b0M{;ey?0rlQ+UETk9E53W$O7)O=Gwy2KS#GHjF1u~
z!24u?s9F?OD6c~{NmPB|{i*mAd5-#(0y)C^=RhEc=eZMq-Ibh-Tt=3~d2B!LIS%Ef
zh^>R~uhX`LG$GcJMzxuO$ku;7CLE3G0uzoXZSPSkY%+=fiKj7bh@5*R*+4ZdX4&FVU^QM1n@igxvTRjoQ
z=Q;6FD4r;*ZU*a~Oc*~!4kX8BPUP-u(sQtK2)mKUyrqt=sQ-qDHO5cHW$}0pU*EPl
zm&0KiKN|?s|CBr@$U}u;QhYN&l8$LRMO0Zo^V7&{tq=P&i3e-*t8^t3V`hzhe0j`9
zZMSKG8lvbjeoVB5H>rP4iW{Z8$5+6s`)8cA!ZjTUM&TLuYHi+rN-VO0!t8+malvcH
z*}UF8lE+2KTg^&3Kh);>zI4QbtrvI*3)jf=6oO0S$)gXw7s*zY3KB{4F5TyCYa_~1
zu~JqO7){)m+tbrCT5o{K`r5h8Z2evF!f~$!&{6~mggJ{I6Y_thPHjgfUCVEpcfHpH
zP>Y}yK{f6BBJ?wv;59ZA&Eis45Sg~}qVlX!!~!iNRzWQ%dFYooztEFB)z_kV9s6A3
zkKV?<-PXv`o8F0IZZw874O<49g#tq5Gk>V*`)^NDJ8tJ
zfU*Nx0<99N8ti{il1LNvKlQyBt;u9_*$?Nltv#dH|)!&;3Jik&tf0ci0m|!{%tKH4@Ekv
zE!%)9y57);dB?ms61@51%&NT57#nDy1B?&5Gv0fye;>JNxsRobmf#tGBr+%Yy?byf
z9&uF)pH*tZ7S&GLG~9+C@!owMEKApxZL)N>D}K;gVpLmJPYAZ;9k8lcRfcV1HNS*p
zAjEm@hOU1fzO4SpMnY*HL{P?>_X{wxTYS9uZj`k$AZ_?mCXWmdG;k2HD&~07EL8}irpHU}+!2|7INWr>xEf`d*
zy$X5B?A@VY0s-EaP-F&!ZUn;?)!*2s8`ExPd=`H?4vN@o(PwB<4pZJudI}6%`tt$O
zcAndqB`R>gMGghtWkG3G@OpbWWqG97(ciZRGg%qBq4t37LIRK&#(^;pwav~@Bf1{l
z7~cby7~^0%quRPtLhs2Jzn{s+22V`Kq1^2ix2vl|2;L275lG5i8^|u9%X@M+@0pN`
z27`Z5Gt&r)7gxW{3N2_8ZXBGkBji9p>1DHbY92<&&T1owS#RTAbL5;y=$0YDyJ*HC
zXxU?Hki+(tjIrT6PrDR0#xQ-Rk=g_q;~3#W#Jh%A8zP(CpfW0K^dU6ivek7cgJiEw
z)dj3$Q(A&U&eG^u0p@B-Y-435
z4mc&M=$1zj-N`RvIu}1fHrSjoEsv^n)iI^~;@(MjlvyUKiw$zUXAnjw@R(SZ$|$=<
zHv5l;8#TEfF!IjNI%TvLsOH
zZ*RiW`BFaeGhld}Id9Hfs3ifmdC8Cj)UYEe0X}%~nT*yxyjp#6P7=qOg%MK2KGb$$hBvR<}@g$&+^9;+2G0_86~(F`MTYFhzf6nIS3S
zll%OkwtV30%s#q_yD1KHvIi0*=oTNplhjb6*Y6~7_TE2}6l1~Z3JF^F4w>%{-x5t2
zXJN0T8f&X3_61*wpT%hCB9*C(Lp4j*9YsXvI7O-0hUwqbQxD<-L4ohXAc;c%ceAOz
z9?%oU$*irWInB07D5$(?c7YP++lEHuUQ=TNaUVVrac_%>wc
zQL;|axTKO@GidER9*LBd=|j4zcAUvAMx67B+#y%>JOOO;sAgmvo?zpXm2_BAWGTh+
zPY*Nc`cL8dTbkR!LpXoWRh5>c<^3mysbCt;Q=+Z0&ubui46DjO2ZE!R)ERa#7KOY7
z;_>nnKR{d0GF6G=-p6iKo)LwrHCObYl{RS9oHPY&KZas(D0wA8)>Y0wcd)>*XqZT8
z6!$#!AQ0X7M7bV{33PDZRi!L(#)13j<8@HK6%L#W83W`@s$YMR@0_C~cjA8+cE
zCCl1w0$(8%x{_jf#L0)t6MQ_(b74LTR1h&d#iCG_`XJMWJk#
z2fbjQCzOSHq0*Km1KC9XTNwU;_gS$#>tOg~^@E;Cb=IL76;!a~(sX*X8uYI7@H^h_5ir>J6!UC~(972SSu
zM(b*>M+#=STf{mj(#CSZoN*Sik{Q>q@ZF>_tu8Ag7B*O{Wcw;Ml32*krDmM^8+tnA
z3RWMfAnf#(oKdZgl#u{G*}g1=mEn#EnvJeHBQJkL$ac(H7%b+((im6l4DK^NkZb_X
zPgwh8erGIP(k*uS?z4{+{|Hyc=e(TCN@6aE=;Vn;vLW*1?=?@A~`dryu-q|3fw6>P|Qwi=Eih
zBkX^sVx;OF)gj0>C+AC27Gicx>u0x4PR2$=c^AqaQLZnp5-XGgj*}>Sg{N2)$~ylr
ze>_4lx7#Bo97-N-o#iINXVOusei%P8R~9$1DTA3sA*XuksR!+{>C%VU
z3&_;X#X^&|ETsuvp6Z2_n#Z$HK*~Ez6fKpODnFrRqwRMnc}-F9V$wUvdoVVN0
zA#Yb^rPo?JIpZv(57Rf{P8UD%$Yxb|tW~9rHedGJFN;kSCgD0@PRnLLySpTse5HSX
z%ehbd$)9c(k(9{}&*4zw8j3JV!2M^Ls>V3)c`#-M*jD-2a=vHOB8Hwm0YizXpXbH9b0O=LAWn|$Ot
zIBfAAdoS2x9v@F6Rz{VLx+-bWr~DTQZ{NzbWsHO=#r
zujzo$AN@Z#3T>~ZnQhHVTx;8$WiTOxgOZFZwgaLjW
zwGx8X;s6EiI!J8PO^b$!2D^V_C~m^Dg|!uDRI6*a4H!R$3Au{*vbT{p=DbIL-re5y
ztrT!d9(7Fb?2j$w*%@ciyP%Pq7Shg)twuM$%Aqo7)$)?Rv^o<#WgRqKlk&qm#m#-a
zW2!RHz`WVtw+6QonCbY^U$HY9Hq(2~yO|s^p<&Nw|J5epEOQXo13X9eoIKFu
zE5g`vC>vcel6^$ksWmS+ZUG-%5ho+>R!y}m6f*)eZFo9|bvBJH({gU4iV+nR4YY*^
zyv4~bd%9Z=W#Q!_A)0^YaVR+;+&qA2t0MZF}l^LR~=Kk=pypW6e^ogGtNp+7Te^RW%_?QhzQ3?7+I})7lntm
zLW^;78)m&eyDetzqemhmbo)J@QSE~ady=*uZWQ+-6v1?V
zg;B)&Htd9iYzheIs!?uxtjKAL{Y(E>D3?t=wVRZn$AfnQ2qBpKWl9kz)a#T2k`@mG
z@l-&sR0^P!FI9hvP#|8b6mYb8u~LMU7K|kAQ|&INhkh?Pin5hMsp~TOPwX9^6L^S4
zA*}O%8>FPkYO;JWLU}-w`A0K#UdSPZ)zwvYtZf}Ug}|<>O4!i9ylGUqD%}fYo*NkR
zkgx2~Ovx9K
zIb&Mf9D^hdIZPNv|Ih!cS%~XRwI{>xDysB|lI6O-$s}8|W6hZh6u}6b2_vrIxI8^c
zRuH=r%yU%}XVOKBF?5+N0H#FSj6#0Hdw&Y)4eu3&>O%bct^E$r&>PAN#-9NiQCT1Uc*)(lQy}v|vSr_HOKJBPBYe^a4U&)86-w0cd
zc+6F0rn+=O&Klt~fLT&p`>DzR#LkTKR1Sm0$Z3B-$>d(9@tdc(4jGVpJQ5j~v2GIi
z=`YCb1RwCx3*a;Ggpo%kBnHv|`hLD?Y-4!Bw?jb3Baw@StJQ5ko$@Vz+Yi~UwW~LY7x6!KvcuYC46A0*
zSonX>QI@bb;pyZhH_v$g<2pEyy_4`^nT*tq(#)j~#aH)u#8obhZ)KBSeX_??=7dc7
z$Qgt9IrI^D$VVGi;$y}c7|@j4FD-OYCX4~v*3-R$W$1F7LD9qJR;_P)
z>;bRByRUbc!bkm>R@ZQ3&%yH8wd{_7j+2R#N2EpCpQgY`%=)p~{*8hf2
z@Xf-iQbt+;a@D9EfJ|8fV0;N?r^b=F1T!6@eyqRmk!<7*wY=@Oz&xtevmGexy|6}H
zcl8=BEZdZI=)h-s2}Gx2%G@^jPJ@3un!)J)kWK@Yri*f`Hy!M9O>Rsx%<9@;F+JY_Dt}B+o>cs@7{IhU(28M
zsFqJ8o9wZL(L|-Jj=K|%j@aU?!h|tE8*WFGn(gda8R#raVRh65ms7odm#^o=7Me_<
zjh1ZZp3Xjw(;Szkaf~gj1dV@0E^4muOg0w9;4?6VuFm}7=5F&h_&mHQl+!}Ls%oT-
zGoyy3F}bQr1GhyxbZ=patd#DFXJ!3vo(uOxQ(hj?FXr-FBYNUp;Mr!GZf~%M#}`ZS
zMx-I|9`-+NOd?V8Jt@f)terZADEEXrc+tZFN8XE#sTo?oCue@g!^3~s&eoPG>(HC^
zR~GKrI|epr=e!31Z+OB-*BbG%>C%&#bf@hUDx%yI?&SW!4{?I=IqlS^*+1Tsw1UUY
zDtWP4&k#iEkS;vwFBa`uEmZnFImz|)cCrCc?h8lGX4|T=CX7K((o6DUr|Xsq_dz?C
zcd(PQ59y%U-9KB}|2%)5%Dhl@@9gBT*Y;kOh$Qje>
z8ji>dlIwgQs7q}GB{SvO8wggG(v|P2a7swqHCCBcMFx@P)mPK&gc?&>30w3BwE)|
zkI;5J5l#@IUz*UFKfcp3x6qI8kkzy;{gWQEn{6KF>4RQP%2w#Tw|PYQpIX`aQT
zpD7YpbxFJOcE5j4!L`#=EMr9jp=uwxH4kkb_~}%~vJd>oHe1*T82Q4&KJ-;Sgh#5UG0XR&_aGET{N77jx%OJTkJ
z1%9P;QRr&STfFhpt-oaO#t%bd_Qp@AekA?I58QE1O0|C&id$Ir9*JDkk(uDr*>{sP
zXk;X|yPRzFx*8@!4fB!^3ajl~J>1i2X_9-mC+#>n
z-jAB5Gs)RW(ZTl}N(D+H0kOk%IP
z47US;faygE)%ppUPEcK*j;Rri9**h2rFt-?a^JVOpSnxjID6fd)D=pbM@%}Q(2ti$
zUt{BNj4sITLPNg_qK9e!)I}F2(U})
zX7A2cSCeG>#t?Bf9*LaT%jcA!K^Tw;0&h=#wf-nS`P?sM1vYA~cF#q0VKRCqqRC^0
z#i8WY9AwMNFd0kaq^T%OAF=4bqkg=iA|R_k!$gnBsYSVK_jYH`wHe+=yrO#O7zxUUK$sw@2em*
z>I73vc6!6gDD2fOrqM0+k5%w}I?6OUV2@c@)3)n6^Mux$0>jQU&1JIa4HJ1r>)%a6
z$lkUxZCiJ{8;}4q6{^IoOblc0?S1;e{4gOnyIzD|T>s*h8&M^1+~M*=G4-WzY%+i2
zRFT|JOdx-Gt2;!K>P3ka;e)rQ4#ZfAnRFP(`JgsX5=<-6*B}+JXZ>5xYjkt<;qcu@
z>?>b@gG8TYb}geiyd;AGPExa$c>lQI=+!$6hPWsUTVm#K@(o5g&;68;d)MDvD>qZ3
z&JJ9f<>#1<
zkD`vkC)GBP{%G@DU)`Xb^F+UnBKP&dR{ozbE=bLjc9Dymm}kS-Ufx;>+;q{E{sCDc
zvq&|nO|o3I2-5<9-IFrjV#TJM{!Ft>UL7^D{P7Kd`iiJ?uiR{!Ro3=E{?~t?%r17T
zuMq3{TJoG%LkcH}caPUw8}doG#l8u8-Zu0!sOXB%){$C--gpolqeUV)=
z^c@^|fX%A51Mf6WEe3#Z6?F7#Y=?its)~7{dj3O61Vy
zMH^o7x%68&=S88&7CzSWh?86Q1Q^^!p{O!F$uD5&;5a`uuUlaKf|Li$ZkDMj+|KE@
z-|#7nUgb9%cFDXmPFRBg7vOVa5|DGtu1kQ-xyw>gm5WGu`s$s$1h9X8s5KSgeX~?y
z^Yd}(@9G!mJNBPel?oC`^Df<+-P(wB$g8j1_1lW(&BcT@lm7pZZgq2(c(6(UQz3@&^hj
zYl!Jp%C}v2tj0)A=Wofce%YMDnCE)3-1lZB`|+dE2=Bc(oAf8ITEcD<8Bb;M$N)hD
z2e33E=Tnfs7@>b>OHSndQ_B&#s*F%OuB)q0w=ipzCUE0Vx9#~hL-K7}^RY>6H9OXz
zRC>R(1->6L!SCBA`2COxes7rIcJ^k-5!-Ikkb(p6){qmo*|s62h2IRIh+?B%0a8@Z
z?E!L1(LRBycigk*Tj?O7dc&z#*Q1hRjLB7P^|yKVmpy;*@hD+~>kU_;roCZ`LiJ>I
z7X;y3)wCbPTrX`%&GvRo2(SS`9SP-tVDqJ_K&UT;!-QdgUL=<4h1O1%!j)FmGQTPc
zVkU+zo)bn{m#bm`Syd-c&Q=%qwsy6$>`K1k8B`UYgix8aa
zs*;p?M)-dT%MS;2$}y!TT~$czBYY%XDi<`SYf&V%jp0iQF^+YG#e(m$GIb`|JVN#j
zx4())xWxGL`2yHH7J{V{#;Nt!_ZstjLG+>sNH_26KkJ
zam72i1ULhgk)vT5dCn>y%QTC^MZ@BGsyD9SXbK-)Qjr_3hTvJ2M)C`k{ycEyYn(MF
zG9>-v4bXOtX}5Oqx&L9g6wERY4M!0CGh3szT|ZPlI-bkv54_Ksj0f92IND2t6)At-
z%Y{p@sQEf>6D7_X(63>5ww}93wHF%Sokh2S$@p$BC+i95P*$4si;M2jAkj9a-C9Ik
z@mJa7AscedRLW;N63KIuLR(#gK8HWE?wrAPlKwF|VeGUenzb@raN=R%_eA9Rai}+>}+dXY1&uq1$
zLBHU5WX~L`iKx`*^V}Cr!ln*0M)5ou4EfjF<4+|6RJ4%CA{k|AbUeF3`on+X5jEHd
znK0T&^<1ShS3TS&<^6zUzbtivqj?YI>
z={TN9T-Z#Wez`0y*PmXtU-C}CI*?Tv=zl_$`Tc?o*G%eT7_v@}sJ9Qdd_#Z7fY0Sd
zY{(mxBTfFwbWH%)24|c!lDmJ3F~EA~JV_Ic>qM$Z?Jpwmblv)k0&e&P1Rhnqe88hv
znDyd8L~~v|hL!PrBSwUyc^>6$(EQJJ*Z=E+FT#h1^1G7$L4GaLon?QBcqLk}83ZHE
zLiwdBeZc?dE50Z+MDqA_PF{PFX9_~;pg(z`L8_~J%MM$SLoBaAjO2f(F<&a@+>Iq^bg;sfP=SpTN*HbD`BMtpDif4HkzoPdyg(xfFAie`iW`zuJuC6{{_q4-qw12*Fg9CXH2glckzhB}+
zv6ICT@nou}F$>klbR0Z9zcZ$N%}7!D#V*0`K|ANY4T;0}6R9n~4x62wF>Tx)l{_L!
zDrV(IP^*8;^Blko{WA{oVflbjysAuFGL|=*4H>sJ>^wJb&FmfG_W_=y&aG*~OOkdU
z-zkQ3b2H|3@GjMZL;1D_y}ELb>92;Z9BATs?%x6Y*^PIX#~A
z-!`P)Av;GuIO?J>Zo`_^C*{$t_kniiJVHx+!SS=f12GuuVDQRC5P%jHO(Ou0o*NI}#`9rVRTy1^5r)a1As8h_$6!2>FW)&kQ0Eq+?jJ34y`8kc~W5VRQ|~dvt{PP_*m}Hv;hv9pXm8XwxZf1RzRl
zhHKK&yl-ZzFnR>zPf*WBJ?q(c?mSY_!UV|P&)w~67xZhaV>c2!w-;{5~f2S;|sw0#@w^4P$?ze7Iu6V$$TIH&pO
zH!n_+Ux$xohjMQd%Msp((SgPuxWaVHwu$2;JP*IXM68#kvl*>&
zPvKzMLb2g+`$hiC|G$Kgy47UkOm2ViC35{+ocs)@9qpi`tIE|35>uPh2aZz4+K^U%P8qthjL6~p=W)Q|%Gn!2d-bOFc;Lp9
zu||fpPvM}Hv&_*HPJ7fFx@9j8gFDW*@PPgN8V6^QWrKD4tFpdjYHojLQe_iTJ0oPa
z&~|cvl@-lb92}G9VROr}>2H=QCw90ra3)_g?9Hfy%P9@*u&~S0)C@5U@OSt1U1rtB
zw0Em{HQjAdmWZ?4rFzaThtwvNfphVc^)|~NXR0w8xMg9i4QV6%gp;9j^0Krwx`0Ru
zUtVNR-9HcpNfi2J12BJZXJj(f6Qv;Y4Xch$79%=;p&Hb+g{eHSOZrpR-LOVv|7h5Fo^j@8k({Q-5hrBTaUGDuX#=m2J1YTz8Tlo~pU;hS^3fGBO63>QFJbFBm($TNYmMKtFPj3X3Ni<1=tBSQMgc6dW(U8|j}T6MQAP
zHM>@n7-fHDUi+R-7elh00jU>(x9XRsXiCF}f4YF0KJ_CyflCp-@yQ3x*%&#U!ZhcW
zf1da1g&8M_+;rd-mH!IIYh$q-(i$p3no-b$Man3CoX*>B`A34O=TpkRH3^`5KKrlm
z;hAN+25DOKrRBSKdq5PudBOHv6-ioWUSfEbF|B`YLX+QO3=Xm=bPW;6qLFVS$2+51
z-6)^bAHUBo;fGJ6MZ>}@lFqNyPvcKfc>0LR+twB;iK)yBam`O_?EBCf^jPfF1^Ilb
zb>NeK%Ice>J^po08&!NDL)V+|5+v|c%ch43ng0FHK2XeapIP6*7>NYKMr~578v`i@
zWj%jTig6j;H>`(cY46LZS@z;jUKA%JI2M{BTHS=^$8|akaE}<8WLXLuL7=qfmw0!Kc%NkvHK9E~l(aVhg#(p~OvC&g^S`@IZ6k!HmjUO+7lds-YuH
zu28Vh;w}nhqh3{vNi32x55eI&IHUvl7w~_W8IMFR$cQfal+QPo>ja8z1HC#aVkfk`
z2p&6@!RlUQ?v=ST4=}5kZfthrXPkvuuFEcHw9*q;RoX@*`TWTGboOl|TCI*YtLoP#
zSTVjPBj|Oo$tU3~bS#@5CPngXs40=gln_~G=)aQgP(nzo-!$v>x?iounyZ%1DHeaY
z*FtMm<3)3d1MFGOT2!<4N14OC`dBNrTRe%+VGv`oYr22_TQua5%1dV7BOaOrA}D3<
zhBx^_Mer
z$r`1wCL$o}j;4FB8YDY>biO424Ih6grLZPa&f!mR?Pkr5g1uDS5)>djgv;pP)|)FUVWqq>G1QhCYj`}9y7iXtdw
z(cl@uMs*Es^FqQiW!F6R9#elcgcpUbCe@(ER84d1f@DvUu
zuEF5niiL++G%U18$+nC|3(d`#)~8R_vKNy+nnPX(hkP$NJnh&K8qt5)RUe)`8_evI
z6$fA^e;f@q5HqP+cDn_C0jv)FZGc`0-wngZQ~5?=-#y`rLJT$a$&KYlFZk%4gvGG4
z)v}b8dzKdi1ACG?e?^`+IDm(8o{O@6zg*0Z??nmEi&Ei>rU>OuOZ3QqPV1_C(Z}i^9S@H#~X1Uwz`Eo48h^(OES7^1~v|
zD1LS$e8@KJQ>|k9<@D@&(Rx#!rbF5PnTHQ#T%w6}`0JRz*S|#T=1es
z1L)GLsao~!m8ga{s`ab_steR!jKRo>yVM$_tEMSg6UM8&sgQsD$Cx`mre?LOT9zK>
zW%=#hnyyCI?Bt%G%PoprdwxkgKmKcy9g`3_t*XC9B1vAo86pGCt$I*zzZY-IS=v;_MAB3p8?WJ|qH~
z++8;HkkU;`&|`ljC4dlusXCZVPN*qf0f}*EvPlIr%_@M>bYhw)Z-b&!!4C_64
z`|Ri^ilQ*8-<(k7=<3ObnEF!0GMSrLp_XV@R|k*0iS~FQXYjSJa{$viQ`VilgtK8!
z3XEF^h=A!=je$e5C|sGtMB=M#%|elHN6uD^aeR9@4V-^4bKc+NlHN)=3{uhlmHA2_
zsPU6S^r!_y9AY?qjn1$_Rp}y@mUTMI1u!olgc45&2nc|h6(NI3yTZUgtR;y^+Ko2_
zU~dCuc1fG}*-f6ImgUD&nJO}H)-d%Lg-Nj#!1`LC$qzMYD5YvyA2bPpCa$D_M&+`J
z^)`6QOG)&-H_z%9a?WedlVK{RecELFS4G5|GcVhWSijKzUy$b>9i
z+AtO^%D*c%ZjmdSEhL_a*%h{{^9HBJt#ud(cH
zhF5=ZwXC}&o8yBtmzG-rq1tM}-qVJaVOD_dvbWqxF9-96iMXlPCDz{Fp#({)mV5jw
z*+gEDV3;MSdTx+XMrva@4ZvcC5J*hOQP5e(HYom=#m_h>d$4BV~W}nlO(ln)K(;1e5f~XgmLw9Rx4`RYH^(
z>{7end#*O|OyiNr18JckUJ7ZXknE|D1_HT%6*_gYn`ql55BZA^X;-G4=z+|w4kcbu
z>0rq9xz=C`aWk0nCH>hr`)+;s1)Wjmygt&ypx}^xavfwp1TRG*4+AI*aWkrwl>mQ7
zoZr2d6*ur4GhyVbJi{K>c~_)A5z32(66iN#~
z`9xkvsB9W9qu^g<8OXbA*3B3m{c!)|Yx@6^sl%>G<50p#3y_X7LG4$Ss@^+o^euQ3
zRi!ssBlti*37Z90l`eZWRAhgj{}8xLj%sB!LH5?mhs)D8{0W@*CX6&Pfq=~cm8#PF
zbn$_Uf2xQ!Sxd4@yB0Cdf4o^$pD?Qv-mKQz{4%Q_p3{DR+F9A-GstnLrXW)L$su~3
z0BLH9;S{?wM67Vfoto;VJ2FIEV26#G;_599A{dy%Mor~bySPMPZv%g2HNKRk4FXPX
z@_YP+_&^$o|dd&IwC_aGT9cXo9qjJPR8nlr+euQ)g+
z&yq{3por3juyQh|3ofU1_+lAtQMj06C1K+Fd0;HZnnPJB_?r1yFuLvm#pdwMGXCW~
z#OcrS9TMcUb#VOC;CX+BR7ad_Fph3r>5F1Lm>f!7UiDmEor}*r0B59CC9H^n=x}oL
zA3@tY4yCOk!0rJ&@=1^GfHp6JiJqo=QQj=%07_kz1IWi-!0l(5EZR^zm}Y4WjQ2>9
z8p-~=O_L2eGb~G8wY8<3II5);{y!rxfzJf9Ocrgv=<{OsNo#+ZaZJmz8{72RBKuqT
zEIQA1+tR$*ZB4{I#q8RUzFoF4Erui}S68H{P=@^i$9OuUS{pH~q$j0mL;xfY8zU@{
z;}j{v4{#W~`~BGt+p1DFjj7^Uu>KVo!?7%7esBQ*$2T3OtksJ7d7hToDZuM&5X|^)7Ip8PfqLu6zNyE6hBX6;w>6?{l_%
zxF!1-;sib}7bbaq{qvT9>XRotl7R%BbqaXIU+9+uPozek>pm?yFd
z<4XV?<&#uX5!JDcXkNs+7><+hd`=4=u<@$Wm0Lr1l#0aFb(HcUzLiq7s+1K8{bsZ4
zA{;b+Pyc@utb>?~A&H5=0Odd$zY-*LxgN)`-SRCRKLd+4Z1hN}p@&Slu42ek8iQ6e
zxu~R~DIIfJ;E*l~Re3<6OM4N+^3q-!zuC-@jq>)CeDDG{dd==1y@|p&MnMvlTZp?}
ze6UamBQt+JT#v)sDn2#5Vv$`HmU2D3{{yN@JaV?NGO9O!Xx9%i364sru)&^vRZ}xh
z!1jiQxOzo}+iLAz1)8`~w%8SY^iDl5PEbbggfU{uU$(vH`jUj7qwwc8{0Xp?cRZJ?
znt5xvy6z!Tp((N1jiAmrJ74h8+b5fZJ?)?fQT1@@fCWURAkgJcK|aAZNx
zZ4fg3Srlz}UH|+n>lMO>X`b7)OOi*N3`aQ2+}Z&i@ZNphx(ObP(izpB?N}|GM>4>b
zpHb2RAKlYwzzmw(wTn|l56uE%-e0TE;WmynwL_18tDoi{9xe+no8IlB{Z#yJFyvL`
z*bX$n_^>2bu#(`?}w{Ra=Ui9
zC2=o<0ZxfoOT1@t6K3ci25jtV&z7v-NjlYLBegTTc8IVCVGSU=ACAqGZ^IzFer2~O
zQ>)5qKm{hBF#%zdA#a?~&JUAff4`QWJQ6zu?vxmz
zdLnAG+`QgB{yohECXA&2wux~1Ydi$3Ghr+#Y(4CA_SU9)3JL^rQ8*c7(jfTf06
zV?c8rE+H4^Mhdpccrn@XrG2LvhP3cPo8r8GbSqBMG5Iu5NnTZMs&cU@UhsvT2NkIz
zBX2EN->ET?uo7Eyy8t^ex^bqv*I<92Jwv()L;uk*)aggaiCb$yToeY#8Lg`({m;g>
zDyAg=d}Nz1f0uQ3#AN3*?fV>zR#D9|antgYR#=yFnn%4owii*(n6?qj-fZzHOmNkI
zf6R}^3VVi0bHsqebe~f0Q?9wi0pCFfxxk{Og=O
zN>5I=--MSSu?4woI+;`IXVm)xGV|wu;$=KF9toJe99Vm_gd@xNbUYHdA?aBFCHR?ro7+}dpYuL#2rbt`
zqjQJ)CrSq|!zH-LtwpmF;W*9R*7z2MehZ@5eZj*;pUACac}aWapYnZXz8m`;w;5;DCevmwPgrczSCvBxsw6FZ(z6V`k(Q;r
zc@fFiI@YJNZ!f3S9%!?wer=8z<7+bTS_d1|&TxP}Zqvy;BH!L)IL&YFR($j(Je|oq
zEV$#W!(?Hkj@sf}UETiv?E~6>KK#DlH4#PrU@dOQxTj^
znKve6I>b&9u@AR*+i%xjEvM#`b?bq<+gzua;bDuu-Q6R*)MH0z+5!II!41OT^%P#>
zsbG65U__5;s}7|7hK7ceUB-wYD?iS%siyY2x~djVU~N~Gt}(#X)zJ@sA6=iGChvHm
zsS?PUFg`8#vTISTV>+^4(|KpO0Ek^u(_u(*apyZ%ML?p=-D+L%Q4@&tSN9<}
z_bih|TW9paE%*B~4?mJ&*<}2XAnvrs`_c-2*A7py1kwjDTl<)PN;ORYzB(Vt_G|de
zIHu*1;?M>QwSUI@A6cG%ghv|H;>awG%0KA6Dk+X4T3$3jXtM0cln_}*f~T3X26`SJ
zyOG7Xq2$e2datnyo1IEp4^h042+pu9mC@!4{rZ81=ZKT4a?h)O3r$YbL>o`(pM2%}
z37}9j&cT?d?IcBN!&|dcXc4Ko&=AkMIo%sE5+yXEjaKQcf#nms{!8z80CwX62hPBLkzJs
z+7SvOoudk*^G;4Niwv>N`)cAKkiP68w!u8lrB&S>7R5Wn*HPrY_S&2)W?5C*2B{in
z??c}`<)y(&+&DANL>fzWPHUBgOnp@A&*ClRfjXxlLK0$ss*i^K>oeOsCnm%-KP2`q
ziD~2TlL)ZQ$)|r2!tAp6&MOx)PKm8vyx?p3Kp|J5T@ufa|C(e~vPgJY{;^1qlWs<*
zv3hN@{Zf5BkmHNc#|QA3`f7b~vu`Xv%QVCZ6P|J8rk(x&=oC3fq5mnC+qV9UtRe!)(+|`~0i$5XNvk3u{pe?V6tpz^vkl6%VE`A|t4dRq)^C=|9AA-K9(Uw_&i-JJ`#-Qc=hcwHN#fn(vCD?N
z07!A!Yl9E!In@S*>07EzBF~Lw)6?B?7KN-jwSXRRZMrL8aBW~fJm1=2Xm$T<11UM3
zKjJunPL^ePyMhos=Y!SP6{f
zf$Zpi&5^uFQ4StCCi8e9hmKs|63mo!K^toCaDHfPECnP+`un9jk}V2#W*2BRnSkUe
zm>*Kuxu>&_IU$o)M~-+T<|qrZ$)kVY0jQyZ~PSv?z2nsRq8+D)yDxNF`^{u+Svf+W_T^??}D|`}(A!@)mZ%SCVlyt%+o2B5&2Tm)-1n8ppYR
z&rjcML(*lT56#Fy-Gsa;Se(wyer}6)*|b9A5&75mvEPRFoBQ`LvdTK=jk=P0!9b*y
zk%ZKfqwc8`(cL23HK#HjNXzoO*>$l!Y08~f%9fAiJL06w6iHgrVr$9D8_8?79I8lv
z`%tuz3hG4DTgp~b)&yy2?B*(m@@U?F55(b+Sp0}zyggnYY{kxub3qy&>xspf=xA&w
zR_&ontaFZRwSHtg6j76}E7*;jF>RuYoH+gVISK>spZSqqi`|`-z)MvF{dB$L+V`_+
z^mF7M$m7JXmf*&Y&`pK+Hbfs$&?T;Kc9n1KAWLRKb?$ChB-OsU`p@t&J80N{+?~?S
zV^IjHyaeHrEySDJHi9bJEBX!?FAD$cl-}LXKz6-KOKXSHQ^h|1h4&zVof0KfCV2U;
z#`!nwpQYueWr#<$Alk?>qK0SVU*aszV2~z^CNh9HW?N>}yK>t?Fq+Cd`1h2rxWzGw
zJfNJoXs99$LKex%@@(3kp@VALVnP5Rf~t1=b}vh#X%fK{THPjN=TyLH0u$+v>$6OUv_3#
zpmII!ZJOg54v5W=G=yj$B;yEFz8Fu{k}_kSjtYXuiZPu?D
zke?YSWQs3KNokV5St@y450yHU-fGTmUYs^>NHLOayVHegDgGjVP(NU@ve-*Kmk_G|
zJ@kS;+EXGRRPH(;Erhdis7SZ___I(J3_DBhMDaR_k@g~KK08v1zxB^2!n|BeO>{XS
z=u8+1B-0jIwLfCTx$)*{<;4EvU1>bI`=fq>$=G)JD41B{{Q8~;o6&Kd05t`PsIgB9}ycel&`ookB
zz;T`{iKf$8M|?Ububn2!Sif9H9;5lu`I|rbTdgW7x!d2v$GE@KER&JDbe857PtvmKBb=gT-L!R|VW1=xUup9OT#osFI88&T6|
zm$*sX@z(b=9F&MmTwvAdDsv_Hx=p;U32^K7c{YIAJvcXwF-fg6L7|S@r?{zyE-8J%7#5
zOHI$N7p<9p%!lM9kK7pumh!VqO%c=C?~ANKegu}lSH-
zPh
z`I6uP^Csh{N@eDQ1H{{uHB?5L{K;ZSB+=`|4V3dMDaZHWhs)E`fR)U6EVfps==pUW
zTt~-vh@3X6?Q~Fgmn2Pj5||;trmUq7Zm&!Y0+?1Rc6_T$tOSJAReT@F=Z|3X#@IF
zsLdC}JD~^}geboOd59p*I8WPQ%HLTXVo>V{o$Z8=Y@dsVXdWmx6%&~f9IkDcFn$Vu
zt{_ajr`H>9vSE-J4zabW2F(AWtci*a%pe(8=&VP*cjF(b)5Ng#2Ir}&sPm6zUHgw-
ze0KGFwLuV90$+VBW*;9c8t1&J3NQ@FR$UYfF|b%wW~!78^YLv)zQaRwkmtN5($d+d
zT9!@rz_b()j
zXp=Bc5}pSD5$uR+a>%ztfmV^?EmJH`QZ|soRF4zJPsOD?mkk7rOf}+`QgO;H%2^f{)ve#_>pGtBcGmd)|b>5ViV@b2MTbYW{J~Tk7b42AhT)
z({>6fvXk?U>W~f6W#&y)#o}$~4Li-vra|2QsOG5FK|=M0Q;*YPlo?~JIHgH(KpoPg
zoK4SYQWQW3CMgK@c}dC|;?yGL+v?CF`KxmnK&3%P-KKw76x%G19!o^fIdAF*+tODZ_MCXBq9Zm@w}Bf=>gR!0pUsB1)jIALYe-GCqu
zUk6X2ozxoB>P!VF;`r)5V&4wPgpt>hnfbVyWmaa1qp_FCySOu5-Wi*$O}ufDo64nMZuu2
zW6|1E^5N{b`Ypahu78Vv6LQm?k_7ZYELy(BgIFw;y$7+F?1%T#H+Z$Oi$Xyl%BoAA|1>tJjEc*Oo!K{h3ZvKG+#vl_8Z5yj#)E)t
zoO_`s**yO#3Qr)W%Tf((zBu=V^ZU3u&a?lJpROAeVf&^0+1e3*XSYF&je>t4f-uZ;
zEwph)H|L8%=;3`o>wVLxNHCCZrll+5!%{pBWVfkGn8LAq|AuN|`udMYBIzphjl04{
zaEfcv+ri}5q~DrclwX-YdA?u1x(Wi}jA==>k5-x^7onK8HBEr
zVZu1$ROQxQ02f5=iapwA6M8`Dw=C7rm*u6n4nktOOA{9?gAy@m@Sw`~Jl7KYJbQ5x
z9_7!&wX8}iIre%7Ph{=I=pPr?2kH#1Dmktvn2wFf
zwZ&h;RGTU;z-gJ2^4o7@MKk>^R4qHtwUp_Vec~>IzM()6l@Kt4dhC`=R8hG=W}mEt7dv3yaENY&ny0`bb+4^UYTz
zxqu_Wp_Gx-7}iv@l`Q|wSFgA#wC^^nQLRV2JUqBbTIoMt68xYGQ@%i2{3G7n%*fF`
z?yqxyzY*({D4;K1+UFtto(>!S4IV>=Y%b`%pk6q6Pd
z)ZAH-yUx7wonCXGS(Rx
zknVO5pKpMG7KNOdZ3@0!Ls>fxrP;=R{~>!d7>Z)iP;Sri@O^#%Tbg`myA?kL-6gAx
z)HY_6_h<*zRIVyPLu<%X8r70|_i{QTJJC{0jLQ5a12BA>m;za{@tNMxUXZgy$k
zL+<(zlLgYwAs>&$f^;mDFUtJ;zxRp@pL@h)`(>UVnHj=LvdcR5PKVBMHzj|~t7We~
z_1vVB3C}Vew3(9pzWI{<(_9#_V)?ZB@+7l*B-fc-!}vL)TCE7^^+OR-Pb8BL90D)=+Ba
z#o0bh99{d~F#sqqQ&tgeUK*Z(>%HK9mFg_hKw@dQ>MoOOQ_mLukX29)(WxdZXrSte
zoW!(tD$K$1bMD^03K@{YfaV47ANJ2!`btX+kX5B6*o{~8*q;2*O&+p;1+|ABYk}jD
zNDpcJvYgjrT~O>gNk#R}eb*$#-p!@S;!{C71nC{i&a$bd_UaqMB1b-cljzcO@U*HV
zh;E#uBewmG8z;)sT<+N;|4-{VbN+8P)S2Y*{TQ*Ql*ji@Hl3AUqWF-%qVy2q+#wc(
zXMZ`k^?zvZ@#KG{N~y4aWUQL~ku)Y2|Er47e=FX0d?+3@d&Yh`q>2BHY!!(Aa(=x!
zg@=oe|1$c^C5wQ+ia-C??DM%t{ulq-(Bt3Guk_c~>6gT<+M$pJ#n`ZwQ+5}^b!C%!
zmW2b{FPnK7BxKbb$AAH2?EWprNt&ad+>4())Vq~}^xrHK;p91gJiwwH+{Aq-%Z_^56RPE80g+RFeF^s(A`SjN)2+05cB~A`khj9Ow|Ly#x
z@yqP*{<6jU{KeP5A`SaWT=JodtHxV53^5usX@mwDxi{8Ed_F)viZuGHF
zJ+%AFAAj=AyVu|R_VSa5KR*~iJAcIZy?C(tb<0T-oLPq
zZN7MA{^zfox_tZd$Nk4okNa5H-pl(}@7_P%wJ(4Ep=&NL)+@B8e_Y@BV_npj@83pT
z*H@kV;s+l_@8MhDT(VdG_?xeP@K+zb!Na${y-dT$8hh%qkF@{d)p!43{{8D;goiha
zoXOuu72B_Wd;0XJJ~Vqt#OdGs@7Hhan_qwY+&7n+xZnK#JzVCC_42{ii|;W^-(K+S
z#eSiefqL|tU)le=fAgVFzlopVfxl?;@9E2<5C8N>f1b5h0N01%16%mfdYQ}=JiYe_GKjD!q1)<&A;~H2ZR6i
z;S=9~{0^V^=AS=&;wJ<1L^F?n`|s_<mx2k-pP@4k(7_U5lG{YO9bVBz=p(Kr9!
zk3aew&Hv=zf78%U{{5NB?;bVzr(eAC|NQoUl1A*YNBQc<4-NiraFCb;|HJ?Dn+v0U
z`;(@A{OalEe*EgmCehy{M&y3jy}?vAO7;+KK|fGj13z4(5(lK^H0E{hm)gk`48XbrMhK^=DMUDF=RTIoys9+JyqbboIjb(Cf0VABiZ
zv)1CSS;uUg98c3}IE^kw)K1#=zE|sig_>z?DQ&8gYZvw1mOAkjw#jDJO0uMvl2uzD
zcxAm<9lbKi!KPQ;S!d-ksEsz3koX#{rqOq^S=LG09n(MW#?w;Hy8BLEO0uD|O&T_@
zusu57(q6k`yk<+}JiBAsxvO{b
zwf9NeJE@m3$DS3o`+#|>O42<6$}{~O+aF3BzW&e8T}3LnW+JM2&Q>cXmu}1qsk)=K
z!+xf#xJZ(ZNfY1xoL+wUo6*C6TmVMhqzOWL@u
z@htF3Rr*?(pA%@;X-^>Eu6PZ@uWG!el{+v(HEr3f1aD=X%`s55w(DMhXamp)C+XZl
zNfOpZ%>fTpM01E3aJJbuSd;OyPGi4zpLxmP|zfnIZ)0S#4d{xXr|v
z&y>VF26=%#d4rdqEmh2am#Z3|o!G}pX${@vJGX30fy%wq45oHn<3pyIPSs9mE@2iL
zP@wM|?{$~!8h6rKbQLu>Eg&TBAl_32V>0Eso-fm_gi#PN12Nb=tzPm>1HSJm{qT1;
zzLh&ho&khha|0?)P8f7ROJceXazQ^z%b;RIY!M)Mv#ml?RuII0Dbi%qsM{s}(G
zu7g|yp_~DVSn^)8V%>m5u#kaJ=6u}?-9wePWbzb_ItL=IQAUB|)^X(ug>-G<9w3hd
zwAd5vMutZa40x@7S2pg{joTQqmt(1xI>*Q$armm%oPT!Xq4NLrhev9d?ahZTFDgv*de-i^_=`BM(X0q%|T11Z)+}1madK|n5^Md-o!{#_}(?)IIX9>
zwOWCDn5$$6;N1I7jI7q+h4CT_)#SW`6Adb5cPC(h-NcB0qzyfZQBp}umake?5lR@m
z;hb;tPE584NPbnz8{}oB0Zl2=s4d;jeD{gdAh8p_U<8axjGckCcT
z+jdih!vL6<8<;57!PBL!)CW+vNCztBnlvS7Gi|pp(K$mSBA_vJFcO>+l99s7_j&20PRe1Y>L?5t$bUctWn;!UXNAx`DzA
zOtp8O1L|fA9?bSyx0q36>cJchkVl}vWkDhsaj=q=oNi%a_l*Gp3$|47exXFz>YjL#
z1&yOOFwtk#rU7xb@uqDmsC}>n%_%VH7A9PU;SaihfgEp{3mgKP6)batVeMNyg`j!P
z)*9ZeH32*6WelrrTfx%1#f;LfD`T)?$3|yT$1BF9ja{!gd%FdISRdMIfifPcH%hzY
z+Ar%)dbWIvrDq%Rk_t3ACp5lT2aVFe&r@vX-QsnF|17-gS;Ynze=}-Cm`_k}NuNUy
zzc#Ud`&w(FUAzeIEvRQ<-la_dFJ99vOqj@SIhPewhP2nUM&7fQltxNFAIyLAufKlt
z=JlJO|9c;^ivP~vzJK|Q4X~d!5PQNvz^Vf)QK7RzLol;}S&|cwKlj$O3AAjn)_*OWM4N4WS0tG=W{RU&fEo30!mYc`qF1S$A|dPXZ~JnIMU90^h_P(WiB6^FvhX%
zr~_Z<6-{Rl3y}E?(cZXBsD2JK>jMEiO7j%5!HGxnMqq;Y(ToT5ytSW0>)9W~Y
zbJJZW*9_&Jb$V@cLMw2G;CaS&z^A6$Y`i*;v~uIgoHiOKJO;QV9RD><6X@Z@It+)~
zM$*h8b58-ii)}BzqUjDgqDu$XgO${M0_{yrYndap!B;dr+JL~@35*@%+Z4`e(HNsk
zwytSk(e#2VvI?>4JaX(YYKr9@f`C?keFjPYtm(1)OkLu@qeBJH0`Y~Z0$B_c$yfCn
z$YTSHUovPdm>{Ki5Vo=KAU3XPdVz3cyt}qzOkaM$N)wJh^xjZojYqzNx
zGBK@cb%a2gS2QhF3)YIe^Q<8{%dF5psqr9B&)CZS)U;A@>w*!{kqHA1whb|VQDeUj
zf}&j0^r!_cYP8IZ8MvKxY(-nb*h^{GjJ1LS1O8-!$QPA9L48x~Y=PZLFV}#oRft(o
zEKscuc|caSL;cUvwe@mc(+%&h4FdjA&jRpp@Z<`L17^PSoWn<)i7ev*
zab7^$<(k*HL6c-+0OCBWfgmP-kf=euf`}H#ke}W3hpXDp`?bsIpjVmZ5!IrYTo5O9
z7q1pHY&>K0@l(qijvsCxyay1ZaWYjiR__M3QCs8rEoXQ;$N|hvo0mg2XvC&&TQgCc
zc16o&L-R{V_lCf07``w9#{1rBKEr!_YPn+uwGG9=4uMl#hfC@YE~7
z2G}s=`7PJFO2sSVoLl8KFdP~KK1f<@;XJ?Pw$W3;bcXEefp}LAgHJI1|3>tKEGw?m4kgM~O+Ncz0Ov
zRAr(O*CtTPY6k08-3W;W+k!yrev@c!PLktsOjzMg>@ZpudSb6rTZ-=2eIU
zF7GrG=#W77_6k}Bmp`kd36bBfn9q_xFvUUdx5sg()R|lJ5^E=Oy#hA`4r+#EkIgPD
z!xF(EvrdT3
zlAYWko}fimvQEv$H7$=u-Qo4xtT6z=kOyblo=DdfesBvY3?}dvEm6|C}kW%IYkp)Z=(+*b;cB&#hr@XiTr8m`JalM`Lr
zd{t@@qvzlr;>~wwD7C%1Lr6U%+TbV6!^;|g4b*ffl<#6podJotzz~eCZQk0%Ec7~h
zf%cQ2{k*b&Xi(F-{O5bbfBoUzi|=3j-^(S8XD*$un+RzIrbuvtLwrjEo1O;DSPd^T
z&&qNeD=je`Y@-38%HE-f@nwPERZM+(hBSQC3SMXg8Wf6nS9h>?P_rey&jr4tqVcaVy5H(Ht{&_~N)EkL}&ornnpMjC0FJt-$^Fco&dU8H|hI
z#C0nSRfw(UoJjnb8Q}T9`Aj_c$?pdk2ve=8JIL9v8q2k|o>jrGb}4)V6Q##|;{^#K
z#v_k^gw;ARA3(b57Qcas3O%}k0;AOISo=V;qItFmRI{Gt7AAUiO4Q0C5*@oHaY!KQ
z1Swc&y@d$_Y_?X?>KlSJ#c8<;ctIVrQ@zDN^#~R+j)RBRF+vICjw@Y4N$PVv_}3<|
z6d|vA=5D!q)x1l>?Amp%A+*XZOlWe(TGWYuKcE39rCNm4@!At6@eNFzTl7DgaA&T7
z-45v-;MgO2N3^BbxWKf!fr&ll6>Ih$t_i1yo9$+`*yC0Jp4-FEOWPn*Bv2w(3}&r6v72&3rLqpzKxY8GwwF_fWIx=AYD2<(Z=Y-8OSti(~(s#_)PxXC&27>d;ol<7t{
zdM~E1VufpIr439kx5hDoeYQ5z6@dHuH|tOS=Iy7-dN*t7A_*Y^0yxxC2Y6L~XagF3
zOaWK3X16uQp1=%xrewORd+Vt#Qh*I_-sPsoGN?g|9rXgpheNDTJOJX=AsU0K+|(F&
z+k)25y@+-NQ7{nEB?HT-?QzXae#F>0hVqE9h@RauLt{(0vYujV44h&|U|R3{`Nl36
zSSi7;#Lb89(;&8f@75md@IwxskT-UgxwYU>La@jcolK{E3mRWjW
zyj|7!NCWG-7n50YS!-&kqTEN?McC!4#uug_9MOWj1D3lLsQ|<>nix
zUo<`{)Le=J7@184
z8~_|0sv|Fsn;a~>QXC>w88W+LBfHY7~PbEbp&oeaCPYsl=mlRxyBHGps0+hvo&1;pJ
zXq<;f|Vyw^z2mhnqJcal{HL2
zvG6!^`E*QwLx4o}G^t(Fv<|6IvI9;Ggma9tyTyeE3`db?+#~tabk9(;v7t=`d?iEm
zC}trrfP*|eBZ}m+ri(17hY8h&8J;!m(^;*dSEy=G|TS=KcReO1i7
z8d3YP``NHKJ%?gWKwYc5U(qyX9sY<&v$Q1Q<~N*C2PH|8LIqvZG+y5(j<~Rb5x^Wr
z=RMEVgML`LqG^!v$RG#*mb|7^FgP0bcNziB?W!5FdumI$fRJa>l>r;viT7GfweZiM
z*Z=;1>h1d-zyDzC-Sc-_zx?$J+F)u*-0d~mSO79AoCbxUYUMehi(grA
zmN`pNw^H#=Fjtu);Um6Cp#*=ng=gO7{B)Jm1wvO70YNZB;31(=QryLYAnwn&%KYgj
zCnk7?7uH6RjsTC#3fS7#R)yN7D_Wl5YBRQfsyes=<~B67;k_|jj#>BTw_NfB5!`Lz
zZFH`q#KC(W2?}jcHM^o^2arVc*;s93;VkpPaO;XWw);xYZyAa;$BDp*D`d$uF69*M
zyQPRwrz^&}XK@8RD{x>_(fc^N_X!}3uX&Lkhy{%rMIT|Q}Q&MZ-e{<7+Gj^6fwS4QSmrqv}Gv@ICW~Rtdn7un3q)Y6U3
zZNc<6WBBy}ss~Bw(t4!SA#bV3&CI=h|NSl7x}=thoPm*k9BSSP>MWHFZx0Po_^EV8
z&o|d2Eq-5Rt_+t4=#<_d@eB2Tpo#ga?0T}^UM$IrSL@AR;;VIClDkJ6EVpkl=)knd
z9n#{Enz>krw>ou-#tquMj~HDQ%AqrFookj6IV#I8E3z?Hq9=~wt~KhB^*=YEoO^6C
zCg?qmD`PX|*?gTby3>w$rf_+YEUzx@vzE3EBmY#|lWUAV1W;5heS-9VR?PCbQq4kB
zvA#rk+}P%q)s&tVz_&CI;;B)j*xN=4FO(x2i&o#5L)74_$36N0XH6L9KzOfX4$|Cv
z*0^^y3pnIm`T$;8pi%b*{h8Gj3zy~?78!PZ^xO8qUi|6i4PG{NIU<=r+9KDsG1#mO
zsBT&KDnD_#;@LL$i#Kn7-yLZ3aybCPp$^6LY_S_XCnm-g9_kF?Qg3VZ`}Z$z-e6$X
z5KfJ2D;hUANBq9WQh{bN_8Xi1
zkN4}1+pk{ieXQY!3>|w5(X2JeC6Pp3&bjpka$GCj*!Fw5d4rdKyh$J%cr6_&kZ0Z+
z+c9ytkfFVAgq`m{ntWkVMC{#Pjx>77#^M3MQ6{Lxo#hH2Fnf*Yby+iC#oYez!w)aR
z+&}8O7k~OI6ZZ4+qkp}POTnC#4U=VHRSyF_h%6zX39YrrFR^cB2Ciu*SK?`$oT)%J
zy7Wj5y-O?CnTfrB=>J@AEWZBk&HHbobhE#UG?%B^y&UL!bx*v&>A*P%jH9;#S9M8K
zUi_eGt}_a06pjGAwpeLwe9+=xt@qSGEcn*;;4dO#-Xn^O8(WF{#$DnrWNeB((=|VQ`6CCoDAoV|
zvHJRBEq?db-~7qn{9^t1@z2Z8{+#A7mzs-l*+h+h6?AMiJYf{!iVQ1rKQZ}k+Jnds
zPVr*!q>fe7FA23)4S*{MavEfBf91ubyoxAZ+Japh}lEtxyu+{WSpC-DgQp
zNV;RaAy9Mxo;k@}WkVNU!rGp=ymqQH=s|6xu9o4vjB|$YiLI-))pahqYiB%ZTw{)MGfgyq
z;=F_-p)mJZB0qxPh;jFj)!6OWs`JfnPGvnVREqn=ITg#JY~*!%Tlx415s
z75Xgg2-nQ9tD7~)MZIokU~B#PPQdk;q02#7iknEKX2CNWf%RD`+9LCs(#Usz?F$Ge
z+&VrACkEH*2>1FC~O1Ma*Z7(&$ouG!{Gl$F)cEsq^wBz1pDj4CYcT}o^H4sKNHKj98y0Y0N<2P8OC!bh~
zAefw*!AV!qGe+T@e3k}JQGu@o=A%GvDwSt!UL7$iElrQYzi>r=O_njy_8#}?t+)L|
zhQw=)8lWt$kz%%LTvnMtoID3;YQz6kZf5l5%h&p35~T0fyFdKm{ks=0-#+~L_cnGp
z6YM3b^R->AMszTs8ge|%f^{(Vy0yKwqGAE35>?ZncS{4%j1D)%86!AuZLcT7LwS@b
z8Q#;B+M~c>*IpEV9KPM!o~4b)NHL@Y@!lf-h*?^toeS3BP3*n>%Palsnae-1u1CK5
zSOnEnNJqmDOFm@xEWuNcClm_21BL2`Lb1J6RsM!dA6r79CJG=!jH&qNeN9BMO
z^?M_a7A434iE7K|dmy_e9DzZ;3}%3
z3PkgMz6U0u6oNU-(VLYbF5Pn(#o+(k&i7!`s*xK%3h525l`#o(`He2lYR|CxG~mKd-UNd763|qeRp#WRmQ`Sk$f4B|Y2|MY684x1$V0)T(+hadrk;i{~wpM@r!IH-EmlD`Wvw
zI>){moT+3m!`UnNSrWSW^F`_jb8oxX9x1+Nlz-c8TG_et6Y019etYlr*MC|s-#^LL
zKhl4H|M;o5pMCi=8o%V{13obXGhrHZ%dAr11|^e}JbF8AKZ8O;ypImQQew4(fVkACc4H~7PT5{*CvEw({oQw~
zbuGi&UTSfm+`53JJ5;yRIAQ;Pt(Pxs{mqDf)t@77KW{WNY2;KBU1?m>=9I`n4C!G*
zCon#>V|E4DnrfPLN1SYwXy2h12aPlr^;34)GSng-b$W^0Q37JUW}?z6cwWz^>1nJomTxkrhtpfmD*R~s9AXV7L&N~g>!I6lO0
zsfE0E2DKd8;d#=eI4b@xG8@|i5GAsD1}l=NLK`JxdldPBL|gm6s$WYua}*od+*DVl
z#+B|MEHqfOaUz4ZNG)i1wbQH!8~0@tgTD)-Wv=h7te6oO+#
z#Aa2*_ZNtBU{hRrlSsf{iuO~8w7hyvYS4IAq1O8ZE`ZLN?YyJ)Ecfg9_kQeg7klNbodj$=L$(!2CQnWNBGZjp$ux6wkx)JS80Hk0gjk3JAE
z6zjgK0$D^kQBI`F68p(@+~r+#h`A;h|46UlDMIkEIwUGR50K8gtc$qSr!?^v01h7e
zcadg!V`Or-HRvvVz=PXvphnde`PS?LbVyrkWLmw;+T`9l)<7bzk%1p05C=pa`_z%u
z)9%s-sepl*;OZSroUm4ZM=7~u7$6LW-=hza$1sDMM}Y4Q+!Edov)=SKrYpL)W13s%sRGGUAR(tV0_nMbz
zSdaXZDWVxtDfnWw`}AQZXLu*bAz&u^9?M$P-DoQgGw!lBL8PVh;#9!rD3SnXK$*YQ
z_IbIKjio}Z-DOVhp$Q5ye|AsM4dcp?h7~JBV%tdXvNks})lzB{qa}z$Ap@<8gHPI#
z()&~&{ta!sJujtxV&|kK-UTd$>mZ2~
z_q&>i&1{h(AB%D4C}ihN;66CktlY0B10Ks)-3G;FCi#+#ytGDfFaOBhr>AhZ)`KYZBN(#uCa{PlM?`B35wBg<;=(W^&o0SL@m
zb`*3VeddaD+hN}&_s1`_DNwn&jA5x8t1
z%7>mccBUg6R2i{(vkIta)I~D*Th{gz7rol5#*RrTyQb7?e^2m$Nou58hakvLlUygq*C}lz5$RL<*5ox$tS1FEc%lUO
z4XhFnDC5cMXw|S-X6e+mtO$dAqHf^zRyobc^#_m6SVflh;#4XvCyY7C8LI;@6meDD
zp^1BgLInNeou+n>jHj+tUu%^sl{EpoSff$MS=`hHe@1xLw(}`Z$1Tpn+G-p8))P2X
z+8gf}Z`mP8_>fhOm=%NGjYGlYL07i#I52}3Ooz~)qsl{Wt|=}aO{L8_ty(Tu54W6p
zud(?DVYR_wJfc8)%EYrdkP0N`5F$=0mg8D#R2*cEo5z^ZZI8sQG9U7EbLfOU=y(|d
zAJ?fye^NPUIO^rey|t^Y&I-n0X<*hatX$wj+L%h<->Q9|8sT7PB{uJ7rBGoy9d5pX
zPDr5g8LLr;Hv;w6WCkh4n=L)}(zd&;+US&3>2Ys0#url<^t7^SlvyhZs(gqe#i7W}KN3XDaz`W&?!e@BNm0B73Z;-JM5(s9VD%^r7|Y{$)v
z9SKGz?c
zT4opc6iYdep~dBl%{^yS{fP*ts;ZdDbwo0_IEUzT#OBOxk$sK_#J+Hb1BLX*
zGCE_^q5WD?uN6SvvnBvg7Y)x*)0yX#(0{p!s0op8hH&xLAs>>CCyEf?z$Pu~u{2r&
z6hwB?2KkwTO+Qi0?OK}-@fk7se|1D6hZ&Xm6#tC)y~U{=vAO#K6x6sR(LFMA!oGWr=h?O=>ex5sEN}?s8-fmF%|D!ppSoQ8BgV%~0H_70Rkd@D{`o
z{@Uk=6#6Zg-Uu2CJxXLxmGZdSCNMV2Uxg
z%^4-x#7p%Qnbkvh5ls~mu=aLleZuOXv`ZAwU4zXk7ghjZPo+>lW3#uh5`)w#bRXaz
zQ@CM)F3I_+y`gJuDrPG1e;jCFkr9(8KtZ(i!uaMs;y+6iJM&a*l0+sF@CFV+l-hyME$kC<-2oWRU_<577#EEF+YRs`+8<$AHcz2r-MUwj&
zn=NZ8QZv<b@V0I60TqVO
zjVy#?yNrUD3bf^LEGKMI>=iE3)ijQIfrB7OQ*2Ed)cDAAe^RR!fIBSGAIF~k`FSnCn&Be2X4a24(FnyKuZb7Xt2&}T{Yh|LZQ
zky)?9hfInpZBcOR%4?{V9xD?`^pptfMK1PNEK4Wu3326<6LUD%?gnEb>Wt~*#
zne}M}e^Ydfj8__)VbDcIpwQ;Xs-vEHPOGB_Z>QKFj}jAAZA^(ud=+Z&h)pjeV*KW+
z!Hh^Q)#G`0m#C(8X5PYW=D0_%WlLmkEAZ6pgnxh;o!P?xo;N^39vNOuN~SeI0-<+z
zcym4CvqyHDQMo-*XqO?X=z^Si&I)G2
zYdGOWiU9^HW^IH82()QG!uv)uu;s($3h_
zf0PrH;7+Z^gAp8BrKk=U8GFxgFf2_;sZ_5t%37F3#N^pAlU$Pe8Jo~w6JQrHI*~g9
z4rSHc^UR)PAHhDrIK*y?de>_M-#4{RtPX7^K;R=ho@;lw=CL%c*z(#XF9={J1`Zyx
zIiu!h>}AX{qhJ#rbzrzVn07sLhR(p)e;moIDYiB73Tqj)0s=vLBfRH`&6(iG@G2?b
z4l5N}qFnOO)C|cchiq=ny~*s6V5CP3DxjdrhGWGtI$~2MMP)VLfCuI{fm|8^W@MoD
zht^6Y#w?a9gmsIok~=#^&T#JN57`-;YjGRCMr3p+y~GXn*ps*ARgdfsaYXile-TZf
z!c5?Q8K|2(TwG=#)RDCUO=2o7EodSZ=v`@zm#6F{YJVT$CbRSvdy5WK$LK~nZ}zwV
z7a8^IvNJX(_~nWyjNUUItm=-U^{$wfIb-$BD1S)a4Zf5viw$IcNmNj);Z;olDR0u^y)-a^;z1hxPmV))G6
zd`A_C<(ReY%wEj|-sptiywoD)UU!H2d^v-yx%D$PXDummwNSZ3SErBbgl!6Bl$l0v<5Dp0kf96Q@WP-OE
zY40`J_Jk?T>&)0HG2SH}jWb2&x?K9O1ZTSUEc?v7^&GKTn7c4ieQX^Q6t{bk$ij3)
z3$EO!$4UAeuAY@IF?O0Jyo-U=AF{cc3!oi%*PHu@cr6Iv4KX;Zp0SA^jHr0nb}d#r
z-fCNkOgv&hj;xi{A{YTue>-bjsxEcUmL#o`;vQ-`^FC)}E8|g%dt~H^&H1>6>Vjdm
zGrVs_R3Q{69B7T2Y=B>oqvW~4FedI9n?@`39r4o(Tm)(buXI4Ft+?rRgvSH))+K#;
z8c=8Ws?JrV1|-GxoHOrJibh2fquL=wM9wI8f9wcfjmmV1D;V`=a^GeI)4u
zN0!+%GNr~9k6emMcFWBDOzek%hom86lB&pbZBR_rYGx3>HjnTGnL7_qqDX@ZAKjqS
z%I>?k!LE)ob0~HbOlK?+jpJm#HRPQ;~YH-PK>hSQ8z%IN1UjyGwXA=IUVyAY(8Tq
z5o020;wBjEL)=7@E{tW$l6ZI4W@$#=mC!7?ow?VcB@=WYt7%E1ap%DnZF1O!ml^7e
zf_W_r_B6Ig@OS1p+X#WwDPv3tgcIC0#~Wp#p?>Clj?5riA<3dnSVY%Q2M+@SwU&tv
z*#sY<-gkn&kC#ZsaajnDIyWD|cx8%f9gw#vlweXYf9nSP9iV=roQZMKnMxD(=CNtA
zt#geUxvO`I!`~x3L08lp0NgtGNUIB6s8?w0PLU#VoOz#`;60`g91I+r<(4WY?~IZ!
z+vLpHYQ$cY9=pVmsv2Mf5Q?%wi7KYi5u08o;Cv^@K|B}ebb(3XU5PE9*-M44S16NA
zEO^+We>ODgqDF!^4gqt-rp7%7b<9|oOUbhcv-9mXc6hlXHob#Qhfq-=W{M~#O^K~X
zy61dkt#~ruv+o*Lb#&&)Fr7h;C{CQ`nR6tEKLOvW5q+^DnIV-pRG$jvoxrh=*aYxF
zsEma%PEm(BL6-&Vl+jw{BRboAH(op%9@jw{f8ydPWExm5R?IvThvCpi9NASNjVQ|q
zz*A~s5eq!!%p9@=k%Yz}wk!E$n+DETeAgNU_72&MXnbjP5QrDnO|O{1-ZfuJyqvL_
zy~CS1YgDj;;JxfLt~F-H|BkE`C_w__3x~-&B}i`0sm67)I9@&zn*}gf7}ZiKoXPM3
ze~DB)@^S?2J9Ce2$7Q`{E04L-oE3aNGTgUa#?*D@ZqOc7GL|9$o+oHg;kGg~c8df2
zT=E&4R)YFy_e7EHV>lOM1~9>oo;gR_9O<}YJIEMPPzuyUlu+kk4zG3uhxQzJLdy{$
z1=w@EN08kzM`V4|Gk6=>xMl;9wz(rWe@0QQ%c#-pvzIfN5V?QVQ^~1i06o>nBUo46
zAoinnZ=P_Yc`8$`&(we`O#$Gl6u6U
zvzc}VyNeJHI4GKHmD?Toz0*JZkJoR$58F2_QT%nwmyBE81mbFVTP@d!Hj1jXsX_eg
zGWo3WSt}TEuS00;xK0i{v6eNef0F4FX_Ibcyg&=cD0pU0Sn(b5Qik^lny0&ueAYNO
za>!=x-WplvS@0Ke%Zo;Tf;8XCIKff*jP&ZM0orop8gG%sByt-+g{-$S9v&I`vX)Jt
zcd=VbTd;&J60?F#pEW+cfIXLTxn;E^(-=}J6vLp_#y)C%P~;K@%d|Lte-t<|O!3^*
z5_|P04kT}7oV#v}bnKgqT@!>1W;!HVq_h#B^;zSS@X~wk4J>!D$2}fM35s}|;q&ZL
zn_Z)F4W?LI6UuvoCMq0j1G?p5XO~*s0Ev{T@Qjgg52PC|KJwm267S}8)Hq~gZ)UEb
zr|xb<(?`_u;S@*V{itzme-=Bn$v9(6-XPOt)w
z6tcUa4`+=_tsrn49T){@2b5V6A6YJ$5a;1x*GO?K9n-u`NxLwM<*IWJPfTZ@+%LOq
zlX95`Sv0|7wX7BO47tvHd*kcni~qacy?FiV3{m>B`Abez_L4Wye;cwV0v*D6#%{DM
zKuOBydjOgc#kYcVLWb?RsR6?;LE^mn=;wP7xkn?$r+LuOtTiyE(J?Kei
z1QW8sd0Jti4euRg3falz4n6S55HvEp_MF@eJc1UG#UjIuR65^-2Ev{o|8+xdNo$%W
zSQKpot6I6EY|0?=S{&({aM(bS%dbsU>@fw1^$Fy76vzSPPt6
z=E$6`6p?*B?Eu9!gXb0P^&UOYv`90SKY$6VQsb~g;elGte{I?)&igq{%5BOO%?U&N`|%sOQ_k)X1xO*sCe2D
z=mlKzxE~0Le^S9bGZly)y~A7BIWko7oC_T&k=Aku7AF|-j5N`A=>f;hoD~Th!cAAI
zf~U8dTj;NKhaNOMa>)g#oLLy~fK|yN7wXpO4tM~S-#esM!cxSSapv3;prAv8deS@a
zgALndSL@VgDsjbefF>1T8}6&!4v}8I)<6FG&p)hJe|B3H@RN2Q=)X#_O1u
zhDpsK?vvz$rfV4!V`x}N6g-0fEjrRbKMe18A2i(xuZY*G!=f@_otG+Aal_iDm3-24
zlRS|s5>9oCt5@?%br52?J84=#
zl~=9lQZa(Y1aEXm>kto-wf>-KPBb*`a+I>he}qeHc;yz^=1QA&oCH$1W69gNk1{dQ
zU3~WyM?Wh9(Vi%NdK=S|qE@#P7~$C22e7KW>w|>dDU!xMwil|zF#&0FTf0d|wDUk#Ecn2c6
zzu?V{Qd^siEJBC@YO7;`44&sZAQi8Z>fAw2lYu+*OYw@0_q({C??6P^V8KLEf2lGe
z?XFL7MY!%9cd+Dm2S8S^ZWS!qVcsW0O=`m*m>IRnI^KcBQHq%(b$*&KT)YzX#CZL^
z8r^OV{O-U1%JsEjRo`}X{iNNf$rR;~3iQc;k%f2Ve$ezx-WceVA~9Gb
zx`Ts`?15Do1la0H(-xVWYh>+;fAA;RmIx^mD$smjr3XzfvAEcw8u$8SgMDtr!h&Hc
zR`WP$TC_z-<64H{g%{st^klA;f*WPJq3K&s+aC`>SM0)Unk2T$BrT$4;6K9SY)L0f
zN~v?W*Am6KAsISeU#*dX7Snb0112j33Flnrg{TM*NfiXCv=&w3ZbzIye>2%Gm!hN6
zJ6xL#15;1nU=#-{IhS(4Bpf*&%NDgw4Bx7|2sew1yZ{23XG}t9M`6hf7Nd1nt%VxU%z?t`pvKZv|henPsKEUY}h}3>g{J={#*;O
zI4rU)f@SPEG$MX4DG*DE&?Y>h+@}*eC5qk1sCB@3;H^TdXvM1WR;Jyh6SKySppD@c
zM~uV!N>y@q=Utk&yL4g|vl7+Ufs|ZE)Go2yBcnnd4O;aso$yi9e{@;X6(H%D=KL%)|z?OsnCci{}$2
zqpp1EwUX@&jE;=4e^dqd3=zGG9WXf}CAYRTfmADj;?CS+j%I=<^dbjLF6rQ~$s@|T
zj>+8N1|$D)0&g1X2TW?@?HVJBYP&;Ccwx3nz0MvvkM)2_!xK@IQA-%$+NpZ*m$fLW
zBi!)n!e=JE`ba~7{c$DTEy_TwLXb?%RD@2L^gi$&)g|(*f9wgeHm6~^ZWB?tSun*f
zm;S;3^_Bk?-~G+2fAQ~Mzxm5wZQItn^~0Oje|lj*`N!L5Sb8`YXw%?5cTJ^i2g#4j
z84$m*RX5swI|JP7ApIUmth;o8ah4tl;p6C{+_y8}Hf!gaWtql)bb;U^r5Um9-sZe(
zXPh9BSH?0}e>I5ydD;0&=D0oaLIwiXvWf2aJc`*OHS=1HsNm
zpBbu?C@KOE%Os$zsg@=YI|MS@ZH)AG6vy%K|w%
z_e>_R1Wy@xDAVFzOLr*pea^DYsA!WTqcLe643C}qE-9+D>AY1YqF?^u{i{D7RHa`u
z9|1{R<0^mNpgg{J6w;xx@EQxKa~giW2S5g2F0-V-h`O7Bo!Hz_yHeVgJM_T44oqT;
z3Q;*Ke;BPkY6Hu)#{F+xI=M*?S|QA5v+xK&OL*=%4CBzE1Q`E~Si76_U_=EP&Q_uK
zdP`e&84HNgXH?+fdcFri11BJKN*&uGwm@gkyIWamMepzyz(=Fd?(o&cYLr65JlM3?
zyu>8mVJ3nSP)@T)y>Y1jVZgTvqGFo@UHNBDj({p3*H7$iGn_&4M)o}$E
z93vTZxp#j&YK|q-*yM;)(?K4!Eq4?%$WIlCKG~lh1ziB1ZB27x0TsjV=5oma@Hc{_
ze{a;E-@bYA>Ycy2qp8czr3gfp7Jx+4%e^rZOp~DAQ3>Y0y@9q$;MqCH`J;IxgP;~E
zkPJKm6ZZDKdHwSJ_pko#{r3lv;wuwD6c~$+7sQ6fwZ-n5M`Y2eqHL7*oq7ScDC%i&
zq#ua)om(R;#nvD#Pi$L1){7R&c`dpUfAuz2js3jf`9Gyf;;mSYXav3!-wX}A@NqRpWGa}!1^l0Vm+`bpcANp&xFct;TL_E`_;iO(c?pIIVQU_>o6!kF}Sm2@6rBfdS8&wcv}s
zb0a%xduEV%^SJMt)}SbEvkXEHUaQPK^xg(i{m~Di1$(3k$zjq
z;h-V%dGI!!ykjqXN43m$(za8SfAosf>N4kOJ{qktGN!%^DgD^{4oYu{dm52YPg{h*
zEL&xFQn9$@IS!i6-ZS)T@#G@hOR5Wu0d%b=$Jon3(@@Ye(tf(nn$1cs7mBi6Dfn?e
zXgbo6dn&{n2gwixU`i`R&KPTHx03_DM&qp|UBa=`*xKsO;3U0WGA-Qnf2rTUdE>9%
z{m#FC{my^8-mLE(T=?IAu=SJ{c=u+#cm+Ej{=U-0C4)rkI(LhEeTkJjUTh?>vz#mv
z%igOOGGl`XA|MixNf*d5P~$k>OLMOK^rFF&k?p7>nEOazVO_z;X57q4^hBi6&3i%H
zBXQt@_yO=Niezw>#D-%Lf9exq05|VN-*sjdypN(%N9x%}S`~w$J3UpI?&iIi3+lA@
z>QQu|ZnzzRcTKfNRqyI|8Vhk7k#$tV4d8E}L#YbYmaMk+M0&uR_af?iNXHn=mZx^L
z!P#?`H9PgrcAs82)<$dFxen75XYo*Nlj4H8!av-p7n_%kFmi#KswAv;wjAJt~r<8&QqG3qVmPR&
zo7iAkLT_iOXG_Lllb!PFRWg2HcJ-ZW+8EQJSv8lj>R{a+hTE3YrVhV}JsXSuxZWpA
z@X1%H2sOut?75>~o!@at7uI~WaRSF&;10E#6;Zk)J&FCt$K8_u?4qyYm@|yTu(yTu
zhgm}(leO4xf6;Oow{VLL%+@BclYn?|mw(5013$(c?Q;#cH;mS%vo2*R0drASJZwM!
z=hLp?s+mlla2zZZ!LZqS7^(vQ3OaZV_ngBsSudWcyBw*s@8S^+HbRVBxRp(RM7g}t
zD`ZBGAwBD;6xhg;*-e|Te9zlQ6z4@jFr8odj<8E?uQ
z%&EGpT>HMgnD;5RV$?l=*~fzLr9Eps-q$r;1w~e6leyWA#7fx1%Eqg6&MLQXbu95R
zpz6;u9ReXs0g1U(vV~oHUY#jwPp9|IZJb>onN$Tonr!rbEpF(ZeOY)xmaE-P;)(1y
ziio*ef9lubvn{-8l32q)P7uwy*%lzLVBhsTZ{f~9s#QIy;r$L)e;{iIuroPs&z0Op~qcoe0YvXz0b5fg`v(OAo
zZqaD6(?XryDzC+7*6?i%Utx9Th{6JYZgW`Le~|nA+PK=DG88o?(rM9sFYsn4!zeP%
zlU{qz$h2){M5un)MSG@Z(-Z9j3qpt6EnEyL^Vt?VNlU7<78tP-!!}fDz4m=i)-;W-
z;BnS$Ajky~8tF*(Y3ClVJCbv*NVj;umRSAJI<(O2Au&PEpTzx#@8aXH)_43TuSv>Q
ze^#w&gMX)vT*?;C6*P66fVPZjmtYa-E%ti3-y!B=ia^hROSq?Gpskl+Aw34W7m6k1
zxPebsFnls7ap0d
zVfr1&sG&8VVryeEXv1sG*MOlVnkM$Xe~7)Sw{u>DhF+hP)TP>HB}4LIsT!b%l+S|&}w?JT8NP}x9x
zbQANB@h^Y-;hXi2n#q}F#Y{kCDaJETbh3&hivx(zc
zz3gX2`!C+WVA96J8q$JjrcY>C!I*HbP(A7X4a~CInFyqi0{v8}FAMsrh7>h&y?z5W
z5;%Sf8iNIlM<5FR&&qWOAdj~&!IHmc1ihQ
z*+sbpm5bG}<2I9si<>vWo+QM-o*NbvAF@8MFVk8J5w
z(8a2X6gePkbDfcE+{5)aC-&?c5+g^Od(i9_l*A-Gc&7PhxW&7-$PSS+f2H>dqC9%b
zy9K=!=e6%ETzA4dPNUe!t%{r51}9kB^GWF%?(2YXX>&y8&X0ABW?!Ys?z-(=h6YogP^`T=D#gT8v$+Rr9rRWm0p~omX{`EWvzcQDo1fYVuOA&7+r@=F$a8
z?ckO?B!e5F|5HQ2|Hi%!kihTxo@ZLn;UE!XpY
z_*wr~?BC5#7l!F)e;4N7dS}*XFTuVsM!DH5%UT)@9ATcD*tIrRokl|r8DvN_s;iw#
zGXt5$u3~qCX2NF#^*lKtugug+kAa}G!I^GiV^#tkY@v=h)g-8-Jw<0J&a7A8d8bK$
z_6SzFSu(^ZNRB!!?U61&_gJ8!Qjkx}LG~`R6x!7zN?V{%f9WRni!N0*Dch%HjZ&8M
zRTfupd6MNS_TJouwZZ;er-i0Yx6hHfc+N>~Vy{ITf&l;S);8iLH{!Fg>9wBi>n64@
z4SwBBvP7*@Eza&`owUX&SD!nI&Xr_fU?UaM?6;N*K_I1BYvakW_IQx~^izj;1V~44
z5(H%-ov0h%e~rC_H#=Tl!88_5AO{RKYBmhn1`?Q$k)YluBZ1h{pptCPX@T~cq
zjX4T^;gI1EZq|SI!{2=U^@siD(&x*e>BqJeuV4R)>ydY!yt2n&*Urs;4$D^84K9k%
z6M8N{e@@MD+R_A(S)p-T#mE35Ib2o1{J#LX6EdzN_4PAyp9TT~;jr7#Dpg9k0O?~J
z;%I5P)kc$rvhXia=>|IH8;$b*ddR=~?gu1XK>pK1_FbP}@|A}6^gI#++B(byeFe|#fTtcn6
z%^3*Q5MajEr9#11sHpc_j^SU93gvK_#l{@F*Vv%rD=wj8(jYI4R&t?Xt3m03iKkSH
zf3#809aQPj4PQ`oLSuqt>l&NR5P0M3v`gPrXvLnRrj6&+ZHD(6n>`H6TCHB1ui)X4
z+3XisGV*C2_@+J}E|%At@1TM)$~wlj?_kPm;5s08gU8m8M~~h@W!QC)+cToGl6Q;g;Udy*O>Sar{)1T-(XMy^2)Y(_2^E!LQ
zPjPRVbT4H#j9+nS=I}60+h$Eib%rAF5{sg%f4IvW
zmm;FBDoVnKG;=UrC#yBmnFvurr~NyqChgGp9g`#yILiry{*@7xdz-!ALamhs==9HB
z!)H#zeZ{aeUTdp<36-8lQOMtU@Hs8%vu(U{wVOY?PlS-2sL)hca%Tt;QjWm;3O+9T!^Ft$t49|7rX;~D7S78>_^T(r=6
zqp~0HCDem_PJ0n&Hv^Re`T`=3Bp14Jzw}+f%OK;@;&F2e@ZfV66jshse|l|FFQKX(
zt;QvAyxQaw++|rm-l{-jUiz-SYXN~dY~Tk1)iZ=uk)+7$G-~H!z^5aBb%jY7B%2r*rBSWFKF%_6>f_W;WI?
zG1HTfR>NoNdo?^Ip0tgAfAFbWOKXhZ26ZO7zir3Q=zopu4_*h((-67Xw&FK>8LUw*
zQFhePEtQ0P+1M*0e05K0w%_QsHqSVPfe|7O6!zkoZ0jNgax$7cC#4>3R`HpUKSHNO
z0?QM&6mRs>r3lTrR-U067Z6JyVPy;J>>&Ze(*`N%HK#IXiW`V4e>uecY>%mEtBMk$Ovmv4E$xV<@k}EmRc?s5UVQ>u!r%4kL*5}gHd;W2rjJ|V{Cau&IH@@Rw7nD*UkwN*9Ec`4w#Xu>avTwxIthlUq;Ef6D
zizMZxKuCgoBaM_BBgc_|CkwG>(^6a3<=k#({T}A!g|}LerBwQhWaCW=q&bKkhU_V4
z%yJ>FmP)I?f8Nt1SIeSRcvjoRY&zJ$4a9kR;koqG_O{wFHyux1f`=#c0_!NP(CEBu
zJG%@I_Do@0%nF_C(QhCQ^}STo+G&#pA=9L+oT8hIb>pqF;O@j=$ror%jnr1kn;!4R
zM!FXexpv+|K^G?`NFYtxnQp0q=g%v%!)tJ-LLCEDe@0Z$kRq|I_o7)Z#2mrU>4<9C
zvaf|-SkKUsjzTxRRBs?6e1n#QMaAh_Udb651gvtf@caVestfwfDzq(FIamiiB1R2l
zWpvdGJO_etdlbm%8EF}QMVbsrXO4B+x-cKbp>cNAw0J~nGY0J1i>y+ry=71Lg@_(U
z5!~u1f5ukHnZ2}W=$7{iUC$%dyfWe{)KzA!0c3cEl;cTNcA?uB=3_0k_gauuUif${
z^ag!}I^J!+$~kscFte^pX%3z7p8;1~#GhM-%0mUID8@Ms7!
zpfu;l=9IMR6i|@Of}jgywY7D&fd02&aY4&md_batiru^W4aDWNb3Uyv451|plDTLf
zo2Ti9UEp$ULnpFy8rnSen9W-OEzewDX+AG7T1Sym^CIwE?f>l|)Y@04ov>4f4Fl59)06^|8
zAoi*(h|~146fZCFm~*ie2)SBsAO^Tie|RyPxaT@#X3WS;s{@aAA$JT+JbDu1^ZCH1
zTQ4)c;ga*8|!@}EYGg=d_=2cK*e3>{zg
zMt)R`m4c@!)BwuHv&hf~EI>E6&u<_~PhfKa;x@}VY-rp!yEgddlKF-Cs4X+|f3~ub
z{K4HCCTH*Lt!oe?>k4AW>{$4G+IKy59mtQt-Qx}hkZ+8fwpa>fRm&8XSF_x&>&n#}
zy63{|(77i!#kV6NgS_e40Vcw=?7n>57&+c~+-0|7IvjuXdB&)Do<%uw<2#0)EW7pe
zk~ayHzD8o3P1Vxp1#M;NoS}IKe^_H?4G9=4B2Ej1RNB-3uzJQYpkA)yC7LFtXktk
zwl+!{E77PQfrhph+A4-^_B?MO+F(-=S!*Hv*He~qk!Po3>B%z>KOwHLe=-CU9|LC!
z-sFZVugE0yhq)0~X%#lNL+5xlOTB{*fOdd4_B!{CUERr6uB9TBD)jhrZMIHR>v(Kh
zH9U{FK#B`y#stJb&fxZ=L0!pQ4zWCLAkL0l3@y&pv58oEkR*4|9u3l<-$0C=>S;xi
zfFw5w>&d_&i?fpMLd?Nrf8QGc4XlQz}ps
zo%?YK0TyihfMB_RA+t?}jO^RmHWp{#DKl5JLktKI)zL(6
z>>_Q}#O}u7dD_Ic9G(u;oYQ^AZ*aMNvI-An9Jer*@k687(bqsr
zIv2*OSC)^PARn8IgitDHr0juF%Z^MJ5W85x2Za16NLD8w&>PE;%nkGN#=aG|biu2%
z%_JD6d)3#FIRzsLnfbU+0_WbTWoeIm#tuu_)985;xweLtE{1HsX0`Lh9(4
z9Ed``po08De<*ny-`ls>8*38$9rg`I#<^
zoFkZdkl+wfXU)UXmBli)M9;!{g998{t|W`CVD
zMha_1x_BfeMgCk$Wb@2TPPJz!WpfwfhKX{~M<*@!Bcl$YR
zKw6Z6OaVa^Y#K9=wAkN#Z6xP$3pXbOYfK~Cf2`Bpc;D<1h5(yR8~Wg`_0MtngrHNW
zo~j3<87zIC83maVf$`cr%CaCJ*{9NVLF%wgct8n=7RzH@8&~v8t#v?V3B*ptQ*vXu
zxwcls(zWkf6pUG&&gvk|WpJ@oN9Bq=B<6$s?&rABT!YKJg`2LscJ{9v)z}lB2O{s`
zfASa)_<$dW@)LCrYdnbsT2iLn!j+`Vqz3`g)^JYYJ$w%QgC`31JzVIoWvvT)!PN}j
z&Xy-9sp7vp=*skST_H@PD*#_taw}`2p-bChMQZ90b!V<)f#8D4)L#T
zngpKu9xe}wKnJA5)&z*oXD;^ZocJF!BCmZ6?T8BO_jhi@SK#$`Gd8Z|}e{fsc
zb@t5yYkK->e+80eTG-h!U3<^SLuu$~UgdNZB*5gHy~lV`_g$B`>Arc=t1%eA<{;v*1NbDGtTa}vn=u(TQ;-lkoP
z&!eW5ckTm;o+TH60*?lzmfY~ue{17;&?%$$1!cVhDru5b702VULb_d>zhjgdkiOd4
zXF!*-?hERH8?oK@wfsA*#9a82;5a%ll3vRg3TiiaF8o}>o%oFSlo3b;!zOM7Zy*~x
zN5|gR^a4ANM;?6NFBgx!!0C$hiXS0FUwhA*X!^kAW+b3t2GIa^1Lm{Gf6%?J&0o#Z
z4Ow`wLv^BexEZT75j!Da;u>#Pt1XqbXX^*}gWDO+8U(|{0Poj$yMi!>XYX1MhOz;3
z^4lUf`#ydxZs2w9eds@#Seg5g
z@MWahx+7VFJSgT6psnSyfBLZI*W!lNDk}!WJuNIYLgk1ki{9;t?(x@f+jMSKC)QNe
zvV@K1xiuz*RS?(WmGxP`ImXe7+<(T2(ZK9LFDli?_PQ1)CvAHDRNX738XgUAoA$Lie<3@TYrKx)o~>wP
z7f1zNTQP%A1JBt$o;<(%bKGI4iXb#+8rUcNe(upSPc9%GU7JVvtd?!{vjgG?idVtt
zlvUE=DfMglXsw(z_r%l6V9k|v%CGXamY3VOHm)|Ze50Xw9;wIjHC0(CCC(#orq^({
zY3Fhwd!Q)3tHy>9e+PrElH(qKc4w6oAS;y7sCZQ8_Ol)?UG%%2L
zZ~cn)G|al7lo`B6wJdHLuWYPtmTUVdho((nG#Qyc`-(DTe>DMl@2eso-h1Ac9T!w!
z*X?q0YBp2a(1v6{`dplRx|_mhO%1ef?Ps==c{*&B%Ja59)G%Z6gQe)7L9QaukRak9
zY3zWJ5pWTT=)Z>?TP{Gd$iI%0mnKg7guv?s%ve;@s1Fi{&meuYRuOJ!@=OcJA03rO
z_wHIAbd3GXf7~BYJ!%r5SRkej{L&f^8w^a4b>mwWHk@flpb*NA01wV^2sV0K>8-w;
zUj`|S9p)jA%V<3zxSmv#+GL+TDE7DkiTnh@*&-p5wzCF%12%}}MFYZo<6BD9bwbxL
zsilD2`r2}$XAs?_VHY4PC{{m9n#Ae>K`!FWcpha8e@xYN1G0l|83Ip=mAQ-6=r`1P
zRCVJ~Zw#M0SfJ@(JxEnH=!|1AxrWCxO?u-g+s=xZS}QBw5(GmGW&7v6NkeR2fE;+N
z5bEzLJ?9k}JOpZSp2Y}pWB8~bsK(wSAcJ?_QHu$fwGO=D`Jk!s=g#urkUq00#QkRV
zZ%-MTf1y?5Y2b4Ka)R89p87<(#0%Aw8QP%`Nt~Z-$nY8DN`eUMjc0H>Ex=~3LWgcP
zdp>@+Ut1ZxDkC2Sfzp|szP8^f3Q`_C5}V;&Cd34Hy{l$J&HN4z#@4P
z9+cC1+TfO;Ux3^x*~ZYUJRgg6$wi?e__HYq?Q*X`&ewq(s7VYV*(zE`6bDN{;?6g|
z<+kuToPITvXVZJ|^K!8W&%5@GOtL{>vs+(a%1XyPN0JN~c)y!X1usD25xGCmjUm7a
ze^_PnHBM)#JVwurfamKV&h`y5>ly=(2o|di1PJ1CwR8hAT18^?co{BF8{i}Ct$KI_
zFxMLaFOao%p>rh9;j;6A2sntBRCp-(##0{W5%SH_H*>a8<$7KN!^`{Fb28`$pMU@1
z`%lkyi2V6QDLzum-2$UhRs>sJih#hfe$Ot`^Cn#ICemefmkhn(9
zh&AOFE}N%`BGO>maV4!gccH;1PYKYFc?;JeS`&}Lu@A0*fRt3Bz53S0LzYiH@3-In
z*ZFDx;s5#Y@tbFxCjaH7Up1-Stb$<7QwpoPM+ePOXsrohSN3Ie$M?eM=@Fgse?aOk
zY8XnO>~^BuMMs9DeZ%sA5ISQ^%i7oQv_atCI_>xzx{`V=Q<{s%X
zdSj<^kNV>YZ~ShQlr)1eB#%{{+_OnNv-w;PHU<7=bQ*`$WuTvX
zYzVWJf*U#xO-cityXaOCs>eiVe}-ZzKy8hZM(}r*voF8--EGI2(ozKtE*aAf&o68T
zHj)3@U33R)VhMV%SK?^eh6bXkl@dE5^~>m(Rz1lz$)sDSOP^vhLvo*M!n(mh~S#EKN8
z2&Ci$LOv@!$8-sD$B1jdwU!2%aeCb|oy8UVEIQ(pA(*&!5m;N-c=grm9jva6!UIAt
zH}+)rMm0T`7npP^awewGDLh4z#(yk5rzU<6c+eqWRGxQhtSV_j2i?|3?X(NO2Hc{9
zHXlI_j;hKvSug-yiwkPW;v(=F=In=sVa=wAQxK9duLhfb>b?d%av7@KSpB_w^}*JO
zsm>1Ngm`-I8xLK|6O$onZys3*D9|;Bi`CiAy#RtI#kh{Jide;BVT(S;Ie+-af`-c1
zfGHt5Pm4gyi=4dnFoV#iYtz~n1BmuI>Y?XmWAzLaa3EH{&Tnp0-xq-qv)sX}1u;A4
zH1YI>TRBG`w
zegv46D}Aw4xCzp5o~JdWrhjpuVaFR4_fG&@rISNw0Sk~tTP~yxa1R!~tIT^(jFH94
z!mf}EhM-Zsuz_&XffzQ|#V0=W>D=XH7J5i2#k-G4pd?T$a&K_4KuQXbF_EPDlmW@y
zhA2|ijEgLDv2h3HY;SXAgGvRnhB~8b*x2XX`^Kwp>EqQP0MRSgw0~;BtdgmmT=`yh
z+TxY%3Bc*WaOW(MqSGTyMgd8f_keAI;v#OPIZRVa%Hylhr*#;z)kWalwKeeR%`Wl@
zfux-SQZg}->*)8UQ}k+7+8RX?8-Vl7J8ganK6AM^xMI|`*ensc8#xIXxl~X+w%&q_
zd!s4q6TqjpTo5Sb)PG$EVwFf;#|?R62xj!7!2kH+!`IKN?tj1PtNEq0y)a5MTP7PV
zFKHrZx{tnT2mc7L=H7RL{wkYurnBHXpC*g&59hqkfy+3{=fdZM5F2Or7It){!h;3t
z^lQL~1T>w?BP|+Ga%(vAWKv>dEcJ8^xYpES7WPlD()a417JpuHwyaaSm3zRDoKQs@
zgX-)IilMK(KRQigW$b&v%r6iLE=z#cD|Bzv4ozySdy{*>1KKyIh{Ya#_^q(qA&O_V
z&5U*r_++)>bT9~+ol8sYb)X+1+RnK5jr*?dh<)Q2Gwl)sy-IYI2c_~kt^p53kT%w|
z9PMTGID4xiqkkRNd|dp-1H?SLd$EFnHdu>=w2M!?7u{{{y=r6|$W5qhwzDsIX3K0k
z4e4M*v)l`r)ELhwf5PN#S8a_?2!
z?6v55_0#K1^J2|Ew7+6%A077wm+ghD>jR7q^XL>j=6^`1o~cfsn|Td5*iEP`ylKuK
z$FoXcKu2Bl9P9VyaV(DJ(PC8y{T|9*Z7~#brAfAX`S|q7g(Geu?cxb)zq_}&2UYR8
zNPXB9n{HIesD49^mzn!k*|hc9dJp)-GRKEbgKbCXqdxH4-4s7s`N6^U{ihG#{QlG2
z|NEal&VT*OE3ZFR5$vD+(ZBM_uQtP-d)<AstI^8jmfm8u%?8zpjFpG`U<;i{ueV
zz<9HuTU-vF9f|4glaD>6wyK##G9wM(|04{e80p}=VRZHs`0k6Av8lIxxYgYeCf6Bt
zyiWM9a}^x1aORQ6Zrn86SwKEEW-;~5u{W$TpML^BWh+lUfmHgYCop2=Dj0Lp>X4+*
zf&aJpX+A3d{{4clcEAk+cGA`4cnDN;Eu~ljk+PN6zJV!Q_OyzOMvXT&n^n=yHDSWA
z^T`d&+EObJ1YSzyiSUYNi^7(KF@2OPn5^Ln#DXMp0=)lajpwb^R3;*SyMZ};B9}R{
z^?&P#y6oH|Nomt~)b`xKR4#a(nOT7dElXEm4b4#1kRJYACi@(7kuhKrR{SGZHki#2
z(2JfX;U@7U=Hpdse|FK=;jUc65IVR^dcx`V-_Dt!)|OkcTeu>UXz@aVI9e2gOKHO
zgtg_`!OLd5Hm+tlelU;=%PhIWu~@%5H-ZG4lwEt?l}p}uqjQx!KuKs7Mw_sqU4K1N
z{;hEFAsVzJ270BKZMI%M1c&xk>$v`QH?|gr11Law^*1U;o8D+=@4QpaY5!(77Oijv
zZ6R|aBH4=F*2Ea9rYV6d+`6&7P8f39-qjrU*Ks#1@*1=S^XIm^e~#>f=>|-&Mc?_IkfN!9W?D=G+smoY=_Ef48^;cKXAH9=xAiKM95$9EZJFQqX
z9&@bhc4F^$Ctmsv2oBbl9e>OlQ`($+a;Y`sSBk!$y%Ux#7@Z60-dBntGxiOkq_Oy4
z-|tRNqs@$BuzMFo?u01^>VBs*
zp;$w==Q2Hy#M8(+Fmc}JCbG-SyHx>u-O{2!Wu%tN-d;vZ=bdh%dVgs%o4Blo)Q5nI
z+YlR7WX*V&n-G|jHFW_b-bBoS*xe2S=`1zfr
zXB+p$L9J(EJ?FAtsee&^>?YKW#)_*KojeF9txlbTr@w8E^*%QdZEiLQjY%M1+
z9>9Ek?K*3hovMb*``pAjV@jz_**Pcgve%)dZH`FDEOYxuTZ*os7+V(oooRAhH22{d-7!KNUiQ%+e?N|h72Op`fj?~jeGj``}qBy-+p^-
z5&JWQSATO%h}kx1pN3!z=~8zQ5ckoCN!s=l?FTgSK@gE!YEE;(cyn
z(s1!CV|&g#b`mcK$=JZ&Me~Svxrt3LAQpkLtrSV3TP+r|d-C~fCldejH}T=q?|%6D
zAHIu^ZyHSfXEtR1?5eL#`Gd)`>6w+BCYn9Xmw#z-J2ldChADpuynKu<$W?8FE5ke0}kZrh&&L`Y_PDo$y(LzaPzZ_hk-#c|G(IuzWVU(t?RzbpZlbDTpx2B
z5!;K$J&Get?uOWX5t_$0Cv}6C4sHfrof_mE+oQB4GA}~wHtVjf(?`RQo)$A_HMmwxW?_gFsDMXH`4l=Z=g*TZpS=w9ud1m(CKYaf`{crye`+q-L-1f7}z9uQg
z3x;SokKdkBN_$x(XG;a?ZQeuXhU46bvVXIrp3nBh)3B|~Qcbnyd&tNu2YIrBKI{y@
zJ)oKFbZ;B+ww`;)(Q9F6@zrE#HI{!uo`KUij~PC96zjkLaen>7uV(1@w?Fm$XTSZw
zz4EI~`$1kldXT6@d9-S-
zA_7>?`v^~H+LCOml?7_tFN2pBn%M0bNwh~)$P4h`*0{QkcYp4Juaz)!h77N%3Qx^<
z-I&9ft)V=wf=iYX+Gl@Ro-g$!Y8BFWsBv58F1QB8oPg)&bJC)ZBVNC<3paADT?MbT
zw#@YjV}UvdF{s31KI)r&MA}vGHh;!Iu$pQ6s|1jHdZ%OB86Lg=%I49Rf_Jfs$tGh&
z1`F7Sy%E$pBu^P#`*`1!wUefdJp~>V-@eV~q;{b!A_7!F$FuYY6T>9a=wFDQj!H){YcC?tg+u1z~1i
z_sSywe4Y}>086^wSWvRF$leDi<#Zhq^JyJr6EvZP7K6+R0RWA*
zGDu;KvFmA*ZK6$`o7$Mc%f@qu=6(SgSwWE?$`B-MmScHZh~3cSh6k5>$gl34V1zsp
zUR_%q2w%gEwMU~pXFKo<$g|PPI_m*B8a&3pqtXy?*>};KBTRyx3V)hf)YzvhJ*C;-
zSIp_Q-0vaRHtIl3V$X|pWGkBQre5hBts+4DtYTaf4M#8ny#aN7h2
zZU;O(P4F+hhfMq5vws`thpoP2FBX}01Tzr-oZ7{)J*#Ba8w*dRCT=buo>u8gYSO=D*uvoI680#D#*3TYi%zMZ_aQ$U;mC$6h>JbRqTwkqv-s+|I2r}&}OKR-p
zT;@{twX8MT?m2_@u8UbRWHUE%d7LX?DVQ@by4cRMYfl%E*?+YMsu{uCyOlni26bs3
zP~B4F=G*2P{AfG(-XePJRu{rpYjbzxEWe1n21tSTIklZxLwMG&@Ce458g1M|K8P6}
zov{+xV<0oro3`O+h~hSHBI7-S^l+1O%Xk^pY$&)rCdeU%*+t~QP-dH?#9lw+EIq3s
zYPNk8JnZ2td4EAX{6wyFbQ%nSS2h-rXSQ&zDHXn2jXS87hK;2_@I_1DR5}w~3JyXc
zzMb7djmcA;t50M&muRk3|AMMk1?l;o-T4@##`wl81c*19wc
zd@DvzX}Y3J=49_<4GRVLo}htuP=jW$WkIOZQf;Q$*C}hW@RtPx=C8+yXXQ!Gn3Pgv9m?mI!zN~PC&1f{G89+8|a6gF>7ZXt%JsQT5KkS
zOzSUm{JZFN&S4_8gcf?7u4p?q%(C3r>VL!TqK`80iFrgt2M^DK;UQ5YLCqHaU{A9*
z&}Wk-Ihumq_I^&`5#hpdaH2og^y-(c3oXHlvOT5gfn*7%+N&X7-zncL4
ze`Ejf@nd}Z>Cr&;3zudf6W)ATWc^dR9QK&~5Q}d3hchLpWdt-cej`5o4
z%3XWd2%0wT$d}PeE41}JAfNQQ5Px>fQHS!OC&_b7D&9apECegyr1?au3aK-XSH-+f
zG`zoyUaLnrQ`@T}iCXTBJ0*cRkek@KjE)gD#|}I?4arY%gPq>SN)vV&chQ@LSd~|6
zs)MhvgSIEl9k&E3Jny3ShM2kra@X0&V1o2Hz|0#xSH^5PxjJ6Qkp2
zwO527#GT9Nky(46dED1I_@yj|;P*Uc*SI|9K@>8-F`af+X0Ob&z#AmF(!29=teb7`
zwxmaD{wgy;q+Q&mMd(YTU4Hg;`q>78YS!!8lN~&xs4jL*mv%YBZe7cTrKf$d2N8W5
zvY+CvY0JL+=4pPgW=<_YT7S)!JMLrv8v~!5<1V^OJ5|a!2raA{V~sh-e=~J(_RDXc
z`|%bS+^s^PwMB!F>I+&)vOT!;{swx0QSxLxBfe8E9BmM{TlF1qR;45y3Gt089gYMcI72%JUF
z%!4M&8$7+;d}>H}$FOR9=GZUeO-Xxokh5+zpPR#MBS0J^isLyqL~44lE$;}J^6E`=
z5jtpa4;fUzi!k01AAd2uQjHCbGETbrtxpCiLq0ACD#V)B2}!VZvP0iUW8Z`Z_f|Z3
zW52wfz{i|LCKv;KrJcu3=pCi3nXIt#RCkHfFw9tO(xT?)5_5d|haWzD_rt^C`OmIO
zW-mP5(A&nu^~weqs~m&If|i#D4Y~gkcnzN^ufAR3_qGDJ&3I^%^i+do*c0
z&4z)CPq;*NJ4THi?sr@R-V5>t`q5t$*MPmTH<-*xH`UrlQ#Lly
zWzbJW*K^#slTH_u>xgX#3IEoWZ8{uS5bqSK(0qmS0g+Ry+7VV2mn*TdOG
z;R0mE-!h}MeSaz>FgHvf`n@5@Rkmrk0678)OELMlzBdYO&@=10W-Es)o);kZ90kEU
zc))ukJgww|F^Yk@pgr}er~Gk`?d^+w^ZM`q`j7wH%e{KVvS=mtp)ww{gv}W#0i&K>
z)BIdOlw4xhlzOd;itL0ymTk?!6S;l?5#I*dR}dak+kY4wrC=zIGKY+O-q^GIGQ@`D
zR(aa8;du)PJdHdZgw`;G^Bes>z7WwQX$jg(A&^#ITviJ)8@m?E7Z6W|M5?XN44!$w(WGHQ
zHg+iVCgXWTk%XvPb3ukucG{?izNLv(cQ^dD)M3)U^uKB5Kydv=?}d$9fMfjaCZMfMAHFMg$J9q
zdTri2Px)CKGtU=hhlH`N8?zyJjxkz35EjB3LYgmJ;Bw%N4CMPPHoDuAW%XHCJJQRW
zOEzDInA!Kdy;g>%r^fCphdT00(=C@Q*8*Og0zv-@|D*}MfAda?8v|)H2N6BR65TE73g!w
zkyTf}kgYHofKsL4FV0Boh~OD|D}NX6)L)1I{lwV_H>v|9qiM5)>G&e88DdLsAcB+A
zi6wC-!$heg;~>O#9j9I3R=jqp=KW%rEf
zh+&hpdwL#G#Yan{mYzz)cJ`k=?1wfD+;!eS1d%G%k(4V)2RqN!;+^NKW`E@|gA0hm
z*g3NXo6PPpyRDS^!GkY0x4nU=ZQw<|dW!B-cg|-IXb;A|`Idjv7a`&ac|3vXP}4w?
z?2wTS^CE#&uDLHDg1N1(<&`BaII3SWgAtX>0~|
z0TIfW#V+a_tPW{eUgg>?(tn}P>y2j|AjT~#9w?f`7WRavcaaw1d8QkPT?)kM(k>LM
z`|JXB*UjoV78ejFcZzMy0g?ei6(L?yU3;gIBgqAxGY5}xx4t!MLUi?4-y7Ymq>DZ;
ztXi}Gn!7;fP`!CiZHzMu0>k!n)&*@fef6^WHqQb1u*%lub?oF2dVhXHKhCqs+SEW;
zMFt|Qgjobm11C0YN+!D{Xtv>tMlyeI~Zr;$yJ4?wB
z85K!!U}l<*K|{+7R)4T;U-*vDZX3igt5si(PQU`H1*SC@0$;lD9e3r{QSYPn)fRNX
z$fh~jK8s}q7iPz9&}_%+ppW!!_^hpfQbs_(%F+w@arXtzCbJB7Fk3N#oS-!|`nBFz
z2iPN8aPPYxE^&IV4a~|hpmx;NZagDp&1^kd-58vqiUC?HcYh+e33@@X?yVl@By%Xa
zF;YtvqsRx_e46bGYHmBO%j0)p^q95qy@nwCq{Mdg7v`hC3i3+rliTIc|EthgxXU=y
z1s9&tD=yz^>zNI?+X5|i9+b&e@NRD~E4I1ZOJ-`;ARO^_Qzwf6
zh^-PqFU!RR`qTu098_+Na)Xl}9_5&oZ63M*fTx?bhJVrfHaxW(h@b%YbbU1GbU8)<
zQlp)FpSjj^Ap#te0ttRvKwxPpAUmw4>VuSdZM}hLP;N<9fDtPj%a;YJoHo#6-dNE+
zo>TMgYz|lx3u5DRM7K8ULU6e8jHLU*T>#AZfsj78(!#Dtrt9p3-!Z=V@bOQ7KflF~
zo~k|j?0=H4&cGSuLNG$Q1#Lsv?;fdtX!XN`~0FWduOXf5u9tT#;$!SAX%1s!}!>`ui@r&_Gt&_BmPy@u;oOm?Xj%m)v7ev
zbSoco^=r6C9LzNHtOo74xbIN&v7B2#SzVh)Ib#;}wGL*lAFBg#Fq8`f_z!1Xd(T;d
zZGT&knN6b?wNrS8g6&ahO>wT_g1rju40BiZc1}|ItdlD(TIaoal(o`gWydJHr&W4Q
z(FSxLQcwlkyM{YkOiW$8VnmEol=gjRF4=5o(bqLxkR7&}#{*+PoAf|_c$I-##w))!
zp3ekzpX`uS4wRS3;UWxl(a5^{wQ*HgkAHRYh_Ml#FqtFE=)S??OJ3#LdoDg9Nw!ut
zU2gcs2w3Orx{go?;$Bm?H5b
z8{mmk)RfI{;R44&XbM~Jb!cH87d8RyS_k^T@8Q1Is6EF{98H_)UkbvdZYfLR9&XJE
zymd9&aW^hqFQg)wQQ5`e-t+q65r1>!+T0*iIj5leJi9vd-o3anGQGxiZ8se$qFa?J
zwt7#goOz8;vTQCxi)r!AD&)>}Y-hX2
zz6g#W7z2?qlf5_E5ou=m&IIt#d$=39-&l1n3N@*x+=YI&V+La4E
z?8kyxTvszHQMraYU!6y4?mYH=3i#HVbb(2%?~8ly`6ym*%Ipl{G;u)0lu)FJ9a`mD
zUYJ@=?MU*eE5^1omfX)^%kge@kJn8d!{#bwEf!EW^;Tz2CXlC7*0pi9TEfGYOn-Cl
zE7{z$Nt#>w0^zu(Q{>ormXhSgrBYU@$&ypcW?Lmb
zcDnX`7u%kshUCZM+W_rIJDXyRl@{k*TUU9N!#4rR$xsYt5{O$pO*$p(y!X7j#9p=W
z#8Y^HxFe2akm(IEn%ccMxz`9VjB1;vt)Q(ei^H6FXY*MG%T9em}Xg;fA~
zk5hfrgYSDyA9+91xr&`qE)!D+I!%^gR8{1rd-E616?rW@g*vmYo~>evMzTp~&K{
zH=+kv^llyL?;szu+IK$4+y7VWPai)1)6>`E68bn0&jVtz+kdG%HyliBXGqRgO%-{K}x}lPneMSK9G6D&ush7CGDNs4M3`R(pw{P%7zuu(Z*D~&iX1=Vq4mSH*f(G
zFV$bInamW5DR=He#5zbH%TD(jkX2HZeFnu^iQ@;u2@kxWA$@sVf#jk348ljsQ$P(k
zLmOhk9=HX)0f_`SvLhrc5G_$8E3d5O;TG4?8^ebV(|?#LV9wz1Gw0(_Ps4d^U_kxf
z_?Ae3m=cxD-BK%HSZ<)S2hSKAD&httYcUIw;h--iHz?IBn?iwqAReU)keZ2r4BU1&
zwMm^_FmzlYRgT;kv4*5U?K?L+RGWcTg0*AD$kAAKBY7F#AgFTPz3q`pwNemTS3ase
z?mXpONq_4Bl`D0DM6r0I(jF06r4#l7EQSZQJN#_&OqtF5o7<0LCN);y`N
zRWO7hO6HB>$T@6j8eIF^2fsNOIScAeY}tc{mp+3mxgq#uJZo>w
zXwQ0PZ4HFL=`3^s5()}B`6!5p)5I*Zf~nK23A+8H8<_-3P6kj1Px!XHkU4i3FQc}d
zo_}=R`V10kCyzA7k_M?>I)y8^J5t$!qI^oEf1LhReEaElbN}J1Z~tTK@qhl;Kl{y(
zfA;&o_`zR?^-`DYbbopb1=}hzcU_aqTy?E+?A>qYAa6F!xg<=QhLnTc(lpA+P`g&&
z`F3*G0?ME#QZu$7In#Dp&~eMw8Snjeihl$m_CA?Qfjn3@Kvlb^mZ3p@%?$a%?X>DH
zv=TEd4L36TDtXLNz0>97y>CakZ&;Gi%Ra^Humi-3ms)k#oZkC(c&elK;sFv-D-?~T
zS82dWwaxSH9}7t-IpSDWI=psBs!_JhbC7`7ukb`fr9br?`oK`~;YpgJ+JdMj9&8_v&@!q$iQ|93EXobWFb*gkHEqpteSM|MzvRE0~
z>UKxCOYwfS!
zh)YQ(MYYD{RNbq?5vRM=yoss;G9yrlh{>&{E;Jsurm38Lp8c@E)swRie}A1_VuUx8
z_BAF5O98=sb{aJ5tn_RnnstgYF95Q2YVKHcU~JDGj;m&B-SyB@QU;OgZT0|pw4Tpj
z+i<2%y01vxAhoq+;eOtu@np@$4#>Z@@ySDsN)5G_4J$n7pa0s%OGSwu
zxkr-IJlSfE+r{0>v6}YyYkwQ>iBa?n*G`dI3^^a2mCfvvDrfD<h
zo>y~YND*miee!Td-p32DpA46!6t|&fN3=O@Zh1Dj8n`XX#n(7IU3(_b4hTOU8)}8<
z(}%-T2dkHT=&sx55grHtso;_EPnpAAKO9V8IVmBpvC+5p=9?CMDSxdaSbxt(q&%d%
zeXo{IDb`ls-F-QxQl<#xv!^mJ4y8lMW)7ZEqImj|EPaC!yFDA^tU6}l*+M^UEzDhd
z(9AHba*TC8`(Xjw1ckXR6&<8Az*_V4;viSc?Ac|}7f;S>PUkr2gU1}Fu3B1KRkA+&
zVVzQuZBRgbH_Aft>3@NNKDIeqwr3ZKgY-K`)#@WfQ=b1Sy&pU|phBO|!34kh@b2r6
zpKyhWf1h7|M~9bhfAHnc`}_wz@!s*=Etk>ozezSwX+YdaD(wkFn7|4Y%ES||I;lLC
zmK4xQC6&LoN$j+^g6$_AXu6u%J`Ufk>|5$8>
zJ`4DBQRdNNrY5%ptzBh}o*lS6QGV6(ILU&ND}9$;?GU3O
z+{jXNY9yvI+~zu)SCHk!vzBT>ixTYX*zoA^B;ab
zv>q)!R)4FVDwA`9xVW<)|K#dV7=nNggU$iP$C0s81^
zfsY(rY33&kq0PF-M(P|rWTil@T60*{-s{rwjDI1FEQ#+O2g!%N%rNo>S8k
zhOjnAG4UWk5o|62XQ5iso~qnKe!>u39RtAs77E;=8>Yzeb3m92DBF3$5a@#4h;jj~
zQPMJ)(lo<*40-rtT{eX8R($x^|MUI}{`i&;eVp>3CX+U!Q!(w3V;vU|h?8fmzHW=#
zL4Pp=H#2Q|L$s88bUK_awyuqNTU>m89pDWSfPL&$;j3C)Z8NNJxh<}lmat5D>YnnZ
z8eg_5k1Z!E7u*$hm2vcO#JqIlk*6gff$VX&VSCrR;s!k_0LP5yVNV}GM1Ap9N$s<*
zxa>av!{5D&FZZu~e*N_)aLJw@=QkhT|9|_t{U_gi`*T7NnzW(alyB6#Ti~{`;nBTU
z+}U0HIimodPnb9-FJC%$q|<#)!=S+R%O~2vE*nM7!1rA`ozR>oP||8=<C`9P
z3*`N>)v7nTgtkuBGpA^!!=CGS&M3MR1HZJA!0-0(?kNg;cUHomI-fL(T6Wcvk$>of
zmh%xa2c8iPT-%cVoKfIr@RWdnX4BXtkZo0v1wfJn-=F7xw4LpwyJqK~yno^Nr}z8bH@SZ<;6vC|@Cc9#uowzeH&g4&0DrQP
zPwZEWNlIy5DA_oM+E{yNfYvaj=ABl3#h7#!u|ho<$xBk1AR0&e?zmH4YkS3*>Y%W9
zF&&zE*S*`s>aBGB04=_`vG}bq9lCxrU0nyaMs_O`1bqNbdm5#`VoZ(O@PAYUFQb$!
zF0=|BhvrrVjN&D8+zutnwc=^zI726;OE!`Z2>d#tuNV_w?OJQcK%8TuLL8)kG%T;_
ztAxE`OhYi8CeQHn2$flJb5>PM^T%rydC8c3t*KGcqJe4ivsOj<;4xiIC?>BM)8Vf6
z1nh1)x~)0Rjx$$gHcuKi=YPy!%@6ad_!uAlzi&_kT{{F|r!^|j(x5AF#g@W#mJgKJI))Pu)lN&zxx|NyV1Jl5;vj5KNIOu_}x!V*>hJlPzmE}
z^wTnO7-IvCq)uL`&fpw?-2c)0th6wYsS9D3?tM&~mhspeGxVNvNZ0Y}l~itloM
z>R^0fNrL*LQ1wH%S%13c(1krxx3ix+m?I;(=OjF$7Z1fy+z~bd4s#!O4Q8W?y!-gO
z|9Q^0@eyPC-{BwMz5n{My|y<#|6yssV|K&y;Te-02Ifse!?s7p>N(FC!Wp=;y947<
zdqxi6!KA>>M%~b^PZ+|=>|fX|XN==)KS!`QPNpHiUYf5S!hd~__(zu|JgP`R!_>A)
zh}hQNTk%#8R{p*<_`_nb=9+SD$SuM4iCVJzY9(gat-)i5(b_<)XAz;kFlKi`C78(m
z?Xnl`|M2yW|G0M(`v(!YCzX~t4FpnPd0?6@;Ctlqb{^-P^HT;9^|05LT7%x(m>oc$
ztvHKr#9QSlgMXk^*vQc*0NwG{5*Y_T9!dxXm(FtrF~J3;iEIk-%qn$Q!VJgLDX#Oj
z#{n?*&HFDWXZxj%9Y6hK?&_4exbX>)SPHc_68A#)QZT@}3bnQA`;e0uET$)(~;Gu)C<02OWxK7R}|TBm|Kw^0{}N?lsp>9s8N
z3*W^3x!_X)dX|+cV{Bxb1>!gZ%&mgIoqBKZRl*%Bo3EB}p|mkMc63Em+sV(Z!F`*i
zQ?j&JH0Mq}dx<7Sec;+Id$InH`5((ae)yO8&39kU5AXi(8*BJKx!0fh>gPsy>||pC
zpJm6Y0e|&l6*Vkk+-orLH!Y4ncMx}94<*E5bl1C
zfASP%mS&oztP#ys8{Jp3Zu}<$Yc$_5gnKXW$Cp<}!Zl$Df{$gFH8wyFyG!^gF5WnD
z^jQPAdnNv49gf5<8e2PtZgH<6YgS69RcQPs?l$hS
z+6pS8#gZE)===(N@r(E0e!MI{|M3$!BWZinKwJ2trlK;iB*vNviF5a%w*+kh9{#|X
zyPef5^7hOLrEnZJL6v)gqRci-!aQd5l=}Aq=7ztfXD@M2&o@vXtB`A+bc+Q9`UVY-vI&0KI7sv6bbECgFD1Z&k&B2m~
z;P1|}w6wKtJcAt!8p?=0SOYv!L4boeFlKwHBXwY!UK
zw-}qv^OzZiJC^nGc({y7pAYc1S#ux9@{U|wcuJsOe|^1UI^>wS+L~o5d8B*uh^f7m
z>*FB+k3w3jrK18Sjc%(7$~tXKFON4`%fWZ5L2W7NZZvG?o7PlxU4MPsaD|y~lz)rp
z!4;t@c~lC!Y8!ERyztg}v=&U?K*};jtE7*%kgjcVW$s_=6>Ppt;FWAhjC7sE)r|Ss781m*4
zNhhDn=Svf07l4UWdHlFEtVVrQVMUr*WnLc-^J7w1fetj?FNH48*I(ZdL(wfG*xA=Bj|T3P
zkjub0CHwX9cC8zzrSeTpYoM1Mq<%RTNjJY7FSS_)>d#t}^|qyXa%44kfbJ#Z`s)J@
z1C<1AP86XJvi8)IpW{R^m(LeNIYJ6+h`hKvg88DLnmKr+`qXuGynmhonYlFF)RvXb
zEy|5-U~aY->*e?wn5C@g#lX;0E2k`*nF!-haA9A6eT(95`K)bKQF0B4CW1%*v!uQ*
z&(}O+(R8eu61{(e^&LPrP@_ii>-kFGo7(!O^dXeo20bQik+9@3u3ry5kh}V!?R)_8
z4y?n9g^)n2c@Xyccz=h^dqQ_#stZQR=2>bNMFE`iTpw@c4nno$8
z_)r2pv#z-pz>M|F*Wb|GYR?||)RuUH<4pQ;cBQmSy8eEGrGIEc#@wSFq?#TjCXbI`
zwpbc=eLNr3Z>NH0fLQO9n=E$v<=rpT_V~dIbD0#8AE|)2H=374UykhUrJc+3*_c2L
zjgv8(BT_M;dr{_z1*bxYl~1z|SA!
z?Y78iz0d&;g_n*!YviRH-|WlR-#lP?Ros9I+3O3)bar=cy#OQTdVM@VY)hy}BLHB4
zb-gBvVei=~iGOvxrKKD9Aj^tYO-fpZIWrsBV61Za{eN7wP}$GofF5l+txP)lB5DQ_
z*30+nKpZ;NW2hk;3+18K4pzMy+mf*iua4&#_tDt(JftoxL$A7!{BgY6
zYgRp(x}nEU>FI`iedcnIuVPyYDBrYGqIS&KI`%#n?q9$AI_8HjemmcN{l&lh-94Ym
zR)0t!ufuwiXlSzNZnDjmW5P7EF1H0nJ&^rqXl3i=YOyEF&%v!Mb)NeIi*D4u9`s_U
zrL3mNp=`8E#Y!{3EwHvTGqQJp`MI*A7GIx1C*;#h8TSQVyA{ARIUak6oKZn#=q2gM
z3EA!oOhw1B9g7jnCFw(MHVu@;Qd6J(wtv8NQmhp7c-aDKqZt)gV%fuVoYUrQft#}G
z7r+hBZX`Z6Bz6T{Zu{Xpbz9*6kTdj?*>ABbm5sJ>e;u50l?&VCeiAsNq9r7dnN%id
zach~U+^4GNa$&yzN#FxO-Ua{>ZGDKrCKy)^wm7Ge^tQkx1MI|g{6XSMz?EQoAb(%9
zSUfl0>S@iJoa)sI&x2G!p{3V%iLU9~mq6y8YiOi7u##)!L_U@TDa=vq?Y>*RBLD<@
zLW!L!b+`4f6IL~>vX#zlfkPHR6Ou-HNM)jEfIduAhuU`Oquv&n2GwgNkqwrH$z6jY
z-Q35MW>;>H{3P%>RQRK+h|iLdE`KqUuWM-kTlN9rSJO!+e(?Oc!K21R&6mrI#m41<8g9pV1Fq+cclj@
zrO>cy6!0;FAKTgYJ!8>^z}xtSvf88jBE={XnlcK*eb4xDqP5MsdM^vN7Z;pmA#6=C
zPrdI}SMQSz8RL{`t7ck}QMGXt3%v=gKLVqyrs9WhQjQkI^#flOP^DzOMbX;=6DzNs
z_ga`veWxy{yVRK-Y&Ny~0)Ou=;9>x87O6Alj1?#02cB{5?Y6+UoBMzc3|wTcWo(I+
zN?m&t6s2|F6(1$KkR3V#OE>B$bc#iAnc&_omLmcOTz<`R#xE_}dRL_s{z{GCWlp6(9`;ZW;)@;D6~!8bS@#+|Kn8
z!J@2GP>an)g2gh$fD1LR(|0ND`iK@z*XeSSrxpYCmFj{T!;^vQ$@LLwb2zp4NDFhK
zw46G{<3VV*kCe;x5s_xrnJpcjpB$Q|;87-L_YHu|8*c@EPQ-6Nes!@`(GNb7C7Ji7
z4$-Uv%tF_
zOwpi_@VQQn`^QlmU%~kTh_%@YJ8AhATn2X!=ekZL=f_cL43bw~3M7Br@~FUD#Wp+l
z({4B2Vd{=kLo}+llUa;_1GamppiW(#t8h;Nh2`+!1%CiFd!dGQK
zjtUTMJA=!Mc5kkA;Mw}g;0(L3ab?tNPmKNH14;LIQmn{oWnyX^r)QCQ)9_o*>1VBkp3hj=V#jGX1PDNg&IAl~@`eH6`
z(fm5Oq#wmZI_Wvpoih6efo8z!Hdz+RZ<4qrX3!4@*OWnAO^jT*>|}haYsfx##FW&v
zBab>DMUjED#y%C&%9=2p$~`gH?kRjM8pu-P!(EYKv424KvUP5WxpI!~RNXqYqeR+F
zl(3o9sFn*)oBSXqB`kyf*EHDfn2CH0JQ)GYQAL-wZ#*PnOqKF>;ZDeI1yCcs(X`H2QLt1DIt_ulCs+fi|D$wQ_yv@FnmkVW08I`X{H
zFdWEs_$@i!LVZxpl7junvT1g<$d}gFYM5Df#2lEV+|ww1r-QjWqsT-~SPZ@TE!boX
zM9bz>D?@#uMYS!d1xxdswr*J~BW6AU`ak6CI2;`@rM9<=%QOW%h)Z*A>p)2@uRmvDZOKf#9J(
z)yTKxArMyQ0|C{r`W(8q+<5qu@=j6%A?mNcGu#!
z#N3B3gitA~9G>sTOd8od->vYQO<
zQCRPidp(7Rpzwf0Z^^eOIp$|tKJXy*C>t;GPP4W)o7_`Z4j-CGaxs9;MbZttk~H;Y
zNcNYmmCJpyfApab2@w@6h>R!2R-lEi8htlW$7Eco2lOX#5tRGnAkjO;YE{^wv;YRW
z+OmJV?}}T~I5jzKWbgP>7KX6Z2+?!sWc(XvXM*Dib@$RQH(d;6&GBZ
zJ@sSlHguv^lZOqpBXy5??V|i)TppLxLkbx5bRN6PI@}_XaX=vQZE=SZ*zPs)FoVJ>
zh^MrZbsYd^cf|!0jDi5R&S;woE}boft5ko?vzl(3&r2)jvjEAbx{=&LMh8u}YUA6m
zQf`YIwIJ)&RQjxfi8Tqa3t%oiuF7rKi~CvuJ#C}RqzwuI@1LATi0y=)CTdK%Nf38K2EuVf)++cRn0d=o%47Hkq4q3ScYpyZGfNJctHHR
zov=TNx;hvN2>v=6m@tj%Qr;%GJ+gn$+V?~)nK$+z@iwnVo^lLq6BVf1$FJF=PKQ#_%`?I%~unhWAZF8Pu0}xknW#G1%2tb$yocc
zy&~ZqTGX@6SgYO?wE>YjprER@VSv~06}Kd%
z%_s_hWw|FR2=VTH(xrH#)dfVO{gu0j{i4yJ4d
zyzjkW?c9VVPI#y5Zr;QzRfg1lEOApnhsbs7b$2b>4;<~|J@!Q#1?0tx752oH|S=?8%P?+nqvem
z-COHf*F2m|_srhi7bSBN(0V0!whBBsvS=a5}j4YHw^PenvQv6I}e+>U8QW(VkrdFv`LF;_7f3
zV|FQNY6z`|QTcRCNx5n)iSa(##%VjAp
zHnKL*_mLfW{px>keS)5^*futm4MPE0>;w=KSgYyP;n=|Sz3Odlfq-7cS0_k}gbZ|b
zI7$eUR_N}-WNonkS+;}-l2^SvTwQYdt{s_#Cb+YR%E0^qEuMQ`9S%g)6AFGa9|1_M
zz+ad>8*tRyx;)$g!k{D;QfI+R-;q}lpP3zVI_%X@_j>YSOCB)coZy1#jvhN0^1Xr1QfRQ@^CX-M;-6L57{X9&f_8!S{gtV>Edt&{9zX#
zNY5iv{CsMSRWr+7ujlHVS?J|cbYymG51T<9sfEgYRuNMYUCc)!du>iqP}uWkyhTiSoAN_#Kr)3i3({#7MnoU6l~EYwPC
z2PLQXCI&i6N{H(_=Uk1%BzJMl6k;!SB6N_4lz6d1&-bfSG^(7E8_*PZHwLE-6jf=W
zY28;BAn8b5dy~*c=W+O2vC0jMbGL$ob9uNnc~+L&=Z12mT>%0?ye18LbzdEB6iqU*
z_{e|1LcB?1FOn4y&VD%q0&C`Z49YYj!pcW2+zwVp)?EAL;lkR%L_B?2;szWWz5yWE
zW7YGxd?8RSXSO8`$Qik1da2tmTeBKn{Vor8z(mp*b!Ne>ObLaqj_t&pD__0tfQZH_
z4KP95tm*J=R*bgSY$y!rYL39xBH!%N4L^TEz_O9Qu;PJWHdg=D;aapb)E8i0#oRo@
zcLylKDx^^ArgRpi%J|R8OnL!ywnrDjC%frRPE*y&uX^iJ2`%qCEC=ko%)hL<*WT-*Mhv9#Y
zf^d-~)K>VG)ikcgnR@E*$eySkk+JjyJbhPcc900Q%fk)K!4BxB5An0Wjg+Jcgv11J
z`OG|z!yE%RR!^HHFFy&DeS(poDqTG@tF~Te5Acz8M&Z62Am47(wJF-Jo|%3RfHHC7Cs__3?YYb*39k`r0#0Pg`iXUFA0W#ysvtJhuA+$>L?6^O-h=P6hqwyKVk
zuFfir2PYSJIi(|X+M!>|q+{RQzTXf^{ETp2X_x1QYZIhuNwX8QbMqUB{8p#LDi2hii^I~{y>4W7m+U4S#QougOJ%uaXV)=&i!Km<5Y
z#eB?5vFHKO%GD{l1MiKJJNH)jCg#)vT8Bqk$K;E>tHbpa)Ye~FHZtX;98B^v@f0_Y
zK3yKJHLU98tF8#g57+`uU&p@qTJPd;gHjf2CzWD0RC|fB67YW+CD_uw`eqLCvbWW;
zS_2y&4VOoww|P1c_PLsy4^6{8Rj=XX3DU6unuaE2&+u73_eXT~#9OA#GhBj+W>&cXKQ@&}Pp(
z6aca%)U&)gtDtmaa-@fE2*C5?>B%DN6xnMJzkFt%(pO%yM;xn;(xX`^N4?VIY4MB0
zO~9d8-8=ijYoY@G2CoBG$O*ZvT^{aqq2UpGN37Op3Z{Q`fHu!uaxT9*tgo}3TeJvW
z&VGpVfSF`#R-<1Y&an=p2VgqcYpFEgopto;L&w!K6YM8)RcqolvGVBaL|vZ7gW{5}
zUU!qPBV*epf8@K+J`NHKxjM*+>gC~PIS2PMwU`4CUKC^<6X?XCa^}_HyeoMsZFjS%
zDV9MOEI5C_3ap8Fc{tE4IwWK5WtMVamu-A$X$AT0@|j6Ge(Fx0uk8f9)i_yAz*5TT
zSA+I>5;$Zoe1{!Ie;ySS1Qef>=Jhvo2?5yBF@hGc<4DIZp*G;VQZAe)cJyI_K!+GvTKf>Ia-4sC50~G}IcwMIEo4(O{w(gXuR`fs
z2|JgE+nSExf~2rFr?NLEEeBqSR<^3wo_c0aAJC{O3L#kUX
zs$zfUvcfOt=H95KZygwO2pgRY@W>CyY|dKj>YM3(cx$8cB?A3>bSpQax!W?*?ECU?
zJC8MFBXg~Xi!Z%ne9Muka(NPSFrOrnNfXMd^YQw;Rv`91o2AR29f&xa(ONgIfST#A
zWdIL=(str%vb?kL==bOiz_xI=q3-IN=B0lDtae6h_7e4^6U9S=yK7oA(?z~hhAbiPn#uTpW
z%3jQ5{nXlVd9!VxYb2Pb38Q5fy$Og
z?Y*(H_RPz}9UkMTxm4v*OKybm2vC2gVTPxxlNgf7-j73YK%5Og$x=vab9-L3UH+B?
zPiRvyk^LOfYk*kpQ~OB(DqHK-HxoFFovnewL)UKP>0T&5g)3B~nyZo6DS)*yHu&lS
zg`_9ViOC)+q_uf@=P;P$Dm~6IYhz0pG#{zeT8n!6Dl`iW8nbQSGQw}X$`sJ{eS`6+K
znp|BX_p_=ma4=;DxLvme_aA@vVSW|A`ta`SkDpL+ihrM9em6h9|M2Y(zWjNg|2S7T
zMn5^>_Spb}#f6;$s)a{~y!H#Iz}Zr`jD0Y3rpyZ%jvj{P2g4DQ;WU8s@Uc
z$$jW<5+T)x8B)(X8fxY(0|JJ85<$PX=4eDj)#ieqss<6
zrRsV?NIXW1nZ&>Ezg~$t%XCVi?nKpcUNNRU1gT^WF_4Kv69r_C*6Br54u1KDF%@w{
zNUcZ~9z=;kw@(_2tL1i`=KT*sf}L&I4NSf#w&~7>rJ)HG}*q;
z!0ijY-R95+_&9C|N%PcV6ZqS>!)$EGnlwfplA>OxoBM;1bf-nUO70m}H(2*PT|Bpy
zmaRLU<)*wo$|~QfB68cPh0=G>GVkly~10>>{7
z87#_WBB3Zy!E8#gik4bRc$BA)J3?mPCjnMYRQWP%qXeXY^d_Sz^BvRYJe+mvLuzLp
z2M_czX~lnM+(1d!T;;p}h>JPxA3yPNHjzi#wAo_?$eArh8nvd{%1Y6130e~s-YN}I
zrH+*Aa_aHdLObYsoxRJCgI1#uK#aD4*7``)SLo~!X$>{$o-6H*8vNipry#SHzH07?
z;;-AL%g|ebj?t~;(x)e*P7`fEJpqja8Jp#Mf=++r>2m7&6JwZ=ebFfR3`X6A`
zJkb@jNo%R5bZ!Y6ZAVC=xj>BAEm8*_(41Nr9kTBU8qIgje3jM!Jb3K9vqxZ_Rrk42
zllgy#K~Ha(K%{MfB
zx!<11+tBpMm$kb?)!;!P=x=YbG-z>rfYV(bAwFe8G0r{+4{`=Z6K+Z@5T2)1GL(8E4ptMO{%(
z;|+7Y&ks8*U)j(fqk;F^0T)U1znT=6dQ0Z%^TRHxfSmvdmYz1f8rj5iHp)w`R_=d#
zwo8;R%YacpU2SM`kiF4?XnR8)+c$;Xy-m*tJX&W*(H`|^?2UYX>UgW2?(@S2=CRwT
z25!sOfXzHfpGk&O6=Pf(_PZ}XzWeX*zMda`_mBL*FFw5g_S?Vx`s0W9-~8^fr2ZFT
zJfwwyF6#oY1z$%cU0x(3bs$@sUg>`^@Rx_tBD9MvWak~)+-g%>-}j>0NaIPvFpVx}
z7e8WU0VxLfFDlhElQggM(*49Rwu%9X*IJVMo>sfwv?@VuibxrEE-(M
zo?6Bk=|U>SI~cY(D60;
za|bhkv1?^68Um&mr*w9%w}-1X3hz?~6Il96}Gk
zUOIrknIGq`KD__;cl#52n;XCQ&F_ElN8fz=bHhC7Te1Kii9Q8bKZk!=*9D?l=_SzR
zc_Y!p`nzNyH3OO09%F9<3a9czZuxm5X*SMURF|>RdR6M$c527rp{MkTt?90M1~(DGCs6d*Vm{)HWN5ffLut
zbs`rT3m9@`!BlzPNM?UVtxU*;XMv0FxqWGIb8z(q@OYj#5*K!@o`YKwonB&!kx(o-
z2aTHVyeI$T!`$!U>yQ65{kQMF{>yfFzxn;|e)awDe*2Rj{DOcFD+|+okeqp>{9#N~
z<9Nx2JE8W8uBw~Hl!e4*P_>%Y3^snd>>V7_x3FBy}j
zz1UsbQL?R!!kx!Nt$u(+|Af)!U1O?~V5pbYgu<1xBKW5F1@Ge
z>VV3^Goy4Zi<5usCG$zMCg9jy4g8aa12nwJqw~oX`$YTuZDZPWuINaQu8Zw*eR3B&
zY_3ydy<|*;w!rRl6uEeGHo1NnChBl=4+s4P9=Y29yWf-x~sCcgHz;iz~^jK
z3hOpIfFk_>Ub610f;OPXw(OpKeg+$q64CzsL>My-!-8B>tunpXi
zp6P*CHr9V!(M;*rrLW1Et#wQY&eMg;%_J!*Q3c@O;Za{~U{EjNoE=H2WdlVRpavBTil7tsEuT5dYo1+4&Mq^J
zMU{X5T3W6}bZ7A5p)JXa?o$+9^gCdejP-KJr6gfkqc45Um_%`FB2xrWF}P#3mMLaKKuwktnCC6cmyD^l
zYM{c%Oa)9&)&_1uZm6fUWd4%;gkmxcXhnZdXw$aA613!ULXxLlU$Xx+$To$0JwAN^
z3wSS!KJf+g?&Oz@X_kr#;sgU`+gtW(Je^1PhdKIu&6o}_=Go3(aCX+R(23F4V`mvJ
z$>URaq)!6SL_LiI(t_1@rUD%RZoXzrT0E?I@WqaAo|tKEH*+ryb@(OkHl`T}XQh8z
z3EOJJ_6}t61>RX#+poD#)9c)>>tJwT=W&W4XXcHEr^{c$&6j|bV=j-J(;Q$_rtHUU
zV8*|vYb%0Yw6krd3c>4{uY&4Y$PI&GQRc*#D>YCt1P
zo8*dYZ6>>e;iZ|X<&*YHFx(ow>KviCiLn>3ca?H-o!WS^v%h3aUQ=%qD7W#gJQeX1
zpYe_~=BzKt+YTxoP6~ka+oSl38l8q2>XOP!_AH!0caB~9PNmhV3&aIeX&rw$OSP9Q
zOdj|@hpxy$caMgrHcz9#u7JPvm+Uw3S(Q?}ud-Ah@*M&WRY5R8hF`%_%3bs&?=pChR3{zKlJB7QM;R?=rYRq5z2w<#l$w*|v?OBA#Kar4XmNrx
zo?lYtgXXKtdS1ZaktibVr;dN&S&^~IOLFGHN~U(}ZmzmJsIR={K85gHU9R_~6U;{%=3`
z07#Vi{}8
zV{0|Ffq*t+=|uG#QUMss4H2i7-ct87k0u>uj5cbHwvKqb;U4o@5f$|T<*LoxkJ53W
z{$m>-vb*SFL}WP0kGvS?cjX1y8+1qr7UnJ|YN%Q5?-NV`<(T*$3^vDiUkl
zGIN@1KWzcORVsM&lQ)IWc)$fu_YFsj&x&Z{C?@Do5IZD7ZX_jQk;HwC!wjfdfJ8Xn%juGJkxl6vvbZWH2+WSgg6!F((UtuAh)(mNgUTfN
zfH+Chp>y`Vb7z0g{tc_+%r<7KYZvg4?3wJEG7vm4R&c2oMEv^wSMTPRzn))x^W}va
zNI&^x7JEB%TVAy6owNrAf7<;7&nYH+b!%`#!AFj1Wg{E_2X13dy7ycs0rKWsgD=lL
z98YtgPMS}Xu@7a3H)veyHwWM0s2PRzKc#|>GXpK~;D${QW*}&y|9jrmNiSj%c1Or)fux#?Zp!Ql5Z5|JD}oKm
zcU_tOLCCTC!lZ8{hjZC)AfZBKB;4^GdE5}v4fhihChfol24!eFRz2SjCW>p{5OP1n
z3rByK{W$It2gEsV^|jsFyiQE}has^Ri`AUiOUF}y!0g(SQ-TE~{*I8r&cWSVUb3=K
z5Ty1($Jvo7K)Y`Ud0GX8wU8qVfYv%F&`4J3q`l|4WBQy0{IRrz9y(Weph&h>^AVno
zk<04xKgV|;=9mBcy&%-BJF7p6#FU=eSi6696hSNqq;?f2Z^7!swcmXF_a}qN0Pf;v
zQMROaOzTYj5*6uFB(bQ4AWo)`u|@FCNG?qKNZjBLg(2Puc|pZIoF
zGmRl_jwcUhl}c-=WqzXU-jQKTBo`o>d~*xaPX;rK*Ilx0*NqB;VHMQuljyUh*`9wo
zm}14w0!m^tWj{RE=QDF4tw^us$%CO|b?xB)`Qh#C2@hveY>|@W&z-=MihxG#HITo3
z&5D&>#5JQ{(Jbt_gTZ3vTkb|2(ojQWfRK1z+D8m~Zqk!F)W_!W;Wak0%+9&Wp|9o|
zV?1|hQ;h>~K4Cf64sPZ(H{|1r%5Hyajs4WYG}1J>!}{j2g^H}Jc{JTU6_x(v!K?(P
z2!1Yfh%TvhtaBvaNwVzqxsM5CgxeBl&Idzjd)F4dH{YJQhw;?G^t}7YLq=|?&dDQi
zg}t{}=rn9CPae$arOvYtP+9}40HEMGfSf6{bDo=LdNIdj1>Zj^S_3qaD%5||Qwj`|XP7}TJiX+k1AOAEwV
zw%guhaA|>WK6Nk$m_e%UTn^j7YXjU%?a{T;>fJ}Keli%Owu&W;@u+`14iUCiD2KiQ
zvwFXoH@*J?zZk#%_+fsG-~8?#g|$BkLfEvAvWw=NScyFY1CNB(vRzU-pV+^?a2!K;
zMluVnYrRab&GuLGW6l{8TM-E$ejpM^t*S?|n!v{WI
zQDFEjp>A<1bE7VJgnoa79~xoQJ7OxJcg-2;&Vav1bn3z748AJ?RGjaKIZp95MnO?7
z%WS0s%+hx8WUb_M+0Nwu{3brotMuy6j6kom64k4;CqNz
z;2qV3N?bRD)DtNUB+|&ac<=Pss2ZNTmaKlqv$Q&Z{)}oA4Tvzt8QGRbL+*-s%Z)Xw
zX8~Pdng;SpKH)3P+!l9aKcz+Y3q$sh?0AM}G=K%V8VYb8N_8$Xu9JfQVaRjVNlpoG
zLRk&I`^dQ2m_mOi&pV!F7sKMnK&{Q?2$u(aT1l|
zkAB5@$FrQH?M_Yh0oLnTStz8;eCT~V(jDv4B~CxtXRg&FP$4V<70(iU=i_=u$hvz=
zQ)WH7nt4W0k&2-#&lb0Y+!G}1RJNX$RcexioOgg>dM|%^o%z)dLhdqYi#HGuH8*7y
z_p!lw;$+*U-x0EEmxpL+D-Y+W)VK9Eaix9LsMm%3`0lHB|2IBR)c+3$L%#U(-8bL7
z`}#M3-q~h;m|w+5gn9pm|Mow>`kZKwcY+hu8ZZHuTf7dy1Q3_hFqvz|lz#evvSK7V8<%ujXs`R2V}^Ss16wrcb_8D6Ya#1(xmVG#(VKQ2Sm*kA*G4{CW@bU=$Yc8=~G(EMK2YdAY)H!A!*
zyMTmMskxjtaOj-_@&-V&Yp2i=4a<&Wh(OAqsThA{m8TDAD{5rPNOA@D2@g*P-8g{^
z^d!vh=MRYb-#pX6P8&QcGQ7R?te(z}HxTZf0}9NR;tMp_Mo+usktjbrJ@VMT{N4fm
zOZ?`$FXsn-`?NrRi$eChfBx>P|2{wZZ{yon#bwGJVF&4>H4kp$Fw2#Q6^2ZVS@FW5
zNwI%PQ7%xG$B|O;jL89N-I5}^)E5q|bvj_QR!&)hqOJsq+nNG(k3C*_*V>sX2V|0`
z6=WX;sq?m@?1F-NUO6=5`LKne@Q|L!XBEsC
zH(#Hz`=z{cXtsnF>U+>Zw&q2|yy(g`mW+Qlj82|2v{4-B40oxg?D*2;2U1p__%
zl`C{kjf7MYDVfZnlXcEo&&{wh>MI}Ui3H3&*T{`3gXGMh@Roq>f-)#i8``h&q2I=n
z!+n2zeDoc3vSzb3ckjDPNLenuJOECZq`W^{HbFBs82JT!lZ
zoT)V>U!eo3T_uHe-Ikv}B9M$VL`oYukJ@_$ex2QW?xaZN1tV%j`&i(KI(C(b$&-EI
z3pUMg8}9id+AZeu(96s2K4C;TspE{Pu8GWB9^iGr4L0>m&7xCxqiS>a%7pxE~
z>Flf{24?O=x$l6LUFmtho;en-q^+IL>PTB@1BRbIEygZq#Z#XVeXytGC=}kA($qyP
zB`LG9WBaN11&_gzX-ZwapVQcnj3wSsv6?p3_SCTyu;re^263~Sr-E7$7FK`fjOVWD
zg#v|gS%*;}Ees23>x`5uo6zQSpAm{dRc|%C3tuDmjx=K+7_$>LKXWXO)KZ46O6z*x
zC%c}~v^xc5`>f9#%U)$@O6@%Kj{<40A=~a5Q~=iW)HMwv+BF&GhSH(H3IJEJrMZ*?
zz-8BukvS5ABrOPYIqumyGTa%H8tZ{@^i8NlN7)O5?63x8s@ZL`X
zqw3pid+Ulm5Dy%nJWdW|rDM*0&-i5UpX$5v4b=iJqn}tUt?|V9r@9?~M&Mr|vwZnw
zeEG#U-~IpLAK!oY*FXEoU(Jud<-b0>`;XuM^|L?saad<`iaHx5q7JtufPIxaV(o-$
zx-oQf+|0>?ZfC7s=g5EF7a|3fsd}{=L-%EX)YO<=-bxv>i@RXCB4GkajT=MLM|v0*
z$keoO`A8YZ23D?YiLu(Pp)pMh8>SpM4H-kIk~+rpgf#ia+39OSyH?@(?3|DogVq{-
zmyp&d?YZ92s=Oxj4$;aSr?Rug8}l$6gv4%zv}(CmR5Wu<=!)J4EU<}msqHt&up
zer@RI9drJ(5lLX{#$!Erwp*PVNtj(dR}Z9WU9G%iP+(@QcofR(1b
zh13Wvj_(d?e5-$1`{NDm%?HKb8Wg>_oV_55sdW0noW~$Rz4tjxZLb*=yQ6&^@Jnv_
zJ&b0OS;+~H^;6A^cMWPts<=%HB?h<0Mm3~@(sFU2#ByG<`W81H*TC{Wq(r63nKQKv
zo{*dc}YI8R+$ytW+i}xRg3lYltHv+-7$>Zq;t24`#+Xe
z4ol;tPwpaKGl((Mtl1o>4YLxtV#x^&V3JOwu=#(KK}-Q)q}2^*K<(4cNaoy&Yu!-Z
zp0h-KHkE>BlCUx6!6FaDe5}L0q3?h(sKq8vq>u!
zSRP58*2kxy^Ov-`9A8G4dOn`_V~Roz9g1>V@>@Yg8Jike*f3c{NTqS
z+@_uL)1IEJtd;bOGd=y8
zBmlc6vPxYlr91QMXYZ@nkz+kB%XN95){^T@zURyw;dj0Rrk>HrMbVtUhd%~ShK
zT6LM`a#(MzrSYt)*}UspYUQeal>d6mV>#fO*W9R7ZX#Fb9PfJo3n^*CaqM8qG{DFR
z%si(4s_g|pzOw<9vIwAKn`48^Qa@PhF|OKnVQyVZKVHKjj*eO9>1!5Z!>fP$yH!hT
zv1M`-9v{?_N#QHdUpjBy06xn$sF?B^Y^E-!F0H$4c2>yD9%xs6OFJ_AIi=+EV{>-k
z1%`>6TDuy*S$Q-Q-dP&_iLJ)IT9PXDQu*Um+Z`WzAMUT1
zT{Jp8EOj=q^Z;kB`nAEMxa{yJ?1(H<%2aRl$8Lk-S7QlJI%g(T
zWv@y3a`pgQJcqP|n0D2lQwyRk@V}EBoKPOXrP>~DaCZHwZ-E2t6!Vy!P|0|r(j(?>
zAS=9{y=oirt17(t1QiB6hZV)%8e)k_X;-;{9?F??8{nW?rPOK)m*Qnf({&dkEBq*C
z{A@5!Fa(L*Gxk>xCO3aezv|bP7HrfjvfXu0ISM;iHS-WOR=Ya<1+*5(4
zG3J$bSy}U4%(T0nwsdSn@B(a_9Na|VP=R2ypR2rN1OF$Wv(1!8u;9`U6aIyI=i0~B
zbEa(*u&nJ;nbVLOVp_*8_W_CuCVX{|oixq_UO4970NLQxO7wrJW!6I3$*Zd_Eg=DfPr5V*0{WhXlc;Vaspv^;!Tk(0LHeeWTovX@=2NoNCysKBQH
zY*Ei?Z*(MkmTf69bH_Y;*C~tD{yT?{6q_ntS8Z!zTP`{`z?NCGfOfAk5yC1PWL&ik
zU}jrRgBv{t?k|6BS_fV)!IGz|xWyXv6~F{)9kpfbwg6X?>SDY-k8LafQc8bJTQqt(nIL=T6jKVY@%GTvx%)!wMWD
zUyUW60a(~&MdheXm(-=VAt7VzboYGCfmx*FCdshQjDmlVR+5QrYBlz&>)gW;EwRar
zWmvteaW&a#-HlzdR@1w-_n3PeaL&q1WI#n}@aX_;
zt?#uj__aUOSUtUB>Q7BPE@iSCuTf9)yqRGj>a~IH&eQTqFB-*hS=ExU5j>BK5yMVz
zwk(gF^hJN8!0PH%0@!Glvsh}?d9-Upin=UbGYV=0g9!wDIk>;FK{PxIm^f>yYt+|_
zqS`5ejn2XfUl|{yWxBWq=NniZuNg&aZD1aXfvctg)|P;
z>*}Kd*F8EA%5I893!?s_FVa`73y>-!zSiEF0ONmnXXweneuXa@1-lQ$OoyWNtN=)?
znjVU%KE6qtHu)+_Mt)U%g@+?9g(7;J%V
zuNj5R2J(*ACk6tijNML(6;sosp6hEyG1UMD&*~Fwn++ZV%m<6Jix=4;FB(N)Z=3`;
zXcK>QpK6!*7^efKWj0)U%_vqcZF`}tlT<(ywhzvmbtxED9xvJ>d%;Vo?$%F1^a5BG
zgjU-nOMnHxXcVWwQJvY@{si%i3u5kl@F!#ITTL7;j3QED8-s%FnNk>{sagh8+k37w
zUNj0&l#&1?8s?w|Pno6Z?g!MJ9Q{S3Sm1xfZD7aFVy&;;5!P&-mVh(*tD+f#dZES!
z<#c=}cK7X_!W)q@du+aD6s!`V=!knxpirePG1)1}1idO3)R=yL6cZMv*bJ3P^)+k~
zGrF#H5K78xM&SUSU9vk*&SnDowOSi>2y*1N8m&Kn6n(PmaR{8Cz>8kMzNOD$|oB`Fw!P4?aCJ#O~>v``QFFYfLi}Osi{C!w=LWc*h
zlcbN!#(A9B*5P%#ltb4pcG(>18P*MV2;(~4y-Pvb{iLmvl5}MmexwPB@9a;
z!WMK>P5~>pYP*6ZvjcBpb|w6v
zRWIdG!DvLJx4NJ`%eJ>Vn3G@_02X>It?)qLev~9HSNB*>WvnvTfX#+08I~3Y*DVJT
z>lWdcVoqyK_tT@xH2rlcc?HFQJ
z;a0=g2YF8c9bpw5Fx1_3-UdQeZhaA~$tr#Dh%ORS4&deVs%^lNy2Vv{q{Hmijy3GWntu;AW;~K*-L>rnK^&YjGrwq*?66*_@S2j>tC*=*CN1MtOJ-f`w6ppE
zc)?l&SQpfViYXBc6Y5EEmPUz!Ao*@^K5{~RDa(_jKPFlwQU<7-@C
zo@v*%5XMs^xoSI(=t!*B?|L$Qn^TZfF$bmi)jc)}W*PD8xPzg97S5^WmbF!u5A3_P
zqgi3&=>|Zab{p`l={Q&##e1&KIbw#xook&*XUt{V;zWc?Wg^|(IbUthIR@)?K|MDG
zw6pdBMt_Ddqj%>#aulC!7QtedUaXb5S$o$DulDcSwl?6xw^gB_ArHIG?3q_=ectsg
zpdv!SSi*7u2mxvcYIWxfR|k}Q*EZX-m>q->giiI)QO0Dq+|KI#tF}9fcDsX|V10L~
zKn#0=o9b>r%p2lyWV;?jL~QNk6Q(~5zG0d)t-gb|oOZ-{cn+m_MC
zs(&^U_E=oSioi>Rh-!`f!|;Lqk-w2`vGb7JU28iFrRK$WdR$7;M!CUR!3rJJ@TzU9
z)Okv!pStnMS)!jCz(m*3mah7>=}Bg`#=t!Qj6wPFXCZ+|GS=O5MWuS#0Z`6Ghq54M
zjnlN2lrlzg*|v2MUPSB0ChF@jociW6$A5I(<=@(y4a)Pat)(+1rypNH&4A~5*wt7<
zn1a`>Zs3HI+6+B5I8luey{@bAThFr9oxq`Y;AQ4vrrsx@8%0R2`j(Z!`K?TEwOQDs
zheVEJ(PABV*}JwWoHkQXL@Ex*CnbhT#eSPC^1N$%vt{P7bn2t6veq@Dwi**?s()Se
zsjKM#LKg*8I7|iyNEJ*4C`#}3YRb~WI-5Nqeq_mNsm|PXSQflJocrB##U-n?Oj&1@
z31&DNra-DS%E!9jUFU*@wbd6k7I0E7vVCky$jue2UyWmUYL-1R0vXuvE>n1_J-NoU
zeMuPcD{b$bVZ}<*hxVyGpkK&)x
z3)_;(v4-y)>?GOkT=l7RB{Ksofr|91J&zEuf3^i`bn!QBgM?5!I!VWO2STeYsV>mM
z6=Q!%WcRCW$Eppl7>~?xmw(YtNnrnzt)$0Qyk?Y3$XyQz*jh~9UmcLSf)kFN@8&yL
zONf+MWIVkW!3GEaRY17Fp7(Os_Sirt6GV*3xdn3%(2`@PK^>B~dfsBQ;f?pBJqv}>
zu+wDaWQJESO|R~6kdg>1h-J+tQBW`jSeqo
z@7lIHq)_*>_VBz{I^8puEuXfvF_DRB0tF3X9*)I%+H2j
zOLuV#2+4G7Q}}U5O@EOO7Tpr+BxbC>yS8-~2j^t|RE0Z(mx
zd9s@B&iQ5~o=~%!wT_-MGivfw_Y9PN_1qy0liRq_;^fIjy>+czz=12b{@po`eI_fK
z68kZE>A+85z<+Gnw1FEhy!U_m&eoUP68-S_TmSmgFMRZ`KK|~zkKcdz?60uF8l4!~
z(yC^SNV~XTzwE&PctD50CSVTO`>J8&(RxRuc$5V@@L-b)ijNUR?r-X$KYth={pdN3
zSK1iV0Scx>`_fiS&wSM|;-m?hKik@Y5MApdy}vF?YWBsRc+W-}Mjzv}3CCSi3oz1A
zlGFkOvGA*t>x+g#o&G$81CiZR@;=rDrRJk}PKPJW_kTqkq>OHM#Aa9ml
zHH?MWLX%438-I@@2!Yr@Ey_+sUo?!3ok5v_6Qve$8#{W+=~$f(FJR44_T0x;!x(&W71otLsI>IFlldz}s8dem5{6UhVrV
zK5WfbeUt;V9{@(WbCSf$3Ih@@L7heH*jG7o%L&nS0e^W%O^Hq32bE8G+JQ~eU-eP2
zRr6eH9TL%fYDbaNPt?_Fak>teG{)*X0(SJ&PU=#FSywyptgI$GwPWCAzoy_Hl4yG5AS
zbJ@E4Y=2;fA-uPN{TK~@pAxJwZTGv@OST52X>85V+G-K2D-%x=Li-B4YrUP6=U~-C
z3SV3B3F@l-41gKZrn}ZVK6$dODYwQr5mZQ|23>C(o3PxrJ}vLl124_Pi2!qKtfhrZ
zX5Wmv))A1wk>><{PPxo}MBIT7GTH0A8X0MwO@Be1$D4(1jCVc2aaJo(OPM1sdmFLv
z#Cm+t8eI(mvFkVhX|=L#;$`b|0wXpDTI>^CbLQDK%?gITK|bqk>k2L_rNg`N26f+pPy?Y$~N)DE^;6c57qI}m*7A3+cU2SCm;nyJ?nAATZG=20^r_x?tat*ZSQcGZq;mV2}+t3>J+0AeF>^$C7S*1$0;Hhr
z%hp#D+lgp$?g~bfn9&L@)H?$GxE$jEV;pkr%}l|^lLFv?`z3H69~JTT@~UIqbUE2I
z1*VhyydBeWo%2XB=>XjK!Ig1`w0eTz1uXWkEGUcm-@-K&)qA
z`JW6F#|}m7Zd8LyWe16%$Plx0&JZ}cgrKtE@a67mudDUiSF70M6z@YIO+y+R;-pi%
z-Zg#_Jj`ji&An`9R?nvef1{Kt9e1}{%W$fg;S7#wYzut$hJUw#@8J$uceh$=@n|-0
zE7z17OKjPCXlGVvrG0n70isHA_?YPes@SJ%r>5G5+tsRf154B5Ubg_pKo!iE(PwQN76_Bd=P#9$x$9&`E@^4>@lZ;SR4jVxYk&2+4PL+9HQq|@ZT2yB^9r&h
zmY(JvDOrqc3eu2-OZ>F0Emiwe9%*aCzU+0mx;MRd^D@3#JYD>~MF=!;*ITJCAPF
zx?_|oRW$)~W3c_vVla*jPiXHux*igqnO>Q5muTGv$dCdHR!son?>oAkj18d?J#!uu
zJ<81xygI&-LqM8X@~M%cNZMIUe66bw^Kx8%sHcA
z|HkvpIWBXIDSct3fz%wA0O)UPfm!baax+fcg+y}!
zPLD|7*0YXH~
z?(TDV@`_bf=Q;?WQ?SZ=;eoZZEmo<;$p*@?>u<
z1JDE0U^D7!m~^WX@7?6G@q(ZRMtFF1HZhVUQpd~ycLc(BjTf-waFFFXA7$fHKmaMh
z$=b8a3vM;vIiKSA#EWf(TRCR{&9zK*>APEvg0_gEGoqNrD*b@vPE$%*0YzSZa~ER^
zf1h4va*CG$%Oapo5VZn?b@%XUg&wTAX=AJR1Tqac%l=W|TR<{*x7wayE4iPE__BF!
z=X$2q4F))>jk|nmZ*~nkyld3V%!1t%dyNZj!QI^@4cim8l`h0{8BEMZEF0Rv?gW0O
z-aWxMxn_@AwF5W9*&@E~;iIOdgMw0be~sG#cJrEYl4i_bV;WI%OoCCi?XGc0JVhw)
z*k)&Bpm$H%3Rv=z(#l;Y!@XQG_1W6$RPcgxUpvonPRgU+bu!lQxeNPIfF|g+(=r=d
z9W3|B&f#~Bt01z_u_Ly0t7Bk^TRJ5*Hn+OF;E13GDPn*#g1H53t~Nuxh}V8Mf1Z0{
zGo5w@c6o>{AH$hXmFW5AWv=F(T?pt_6E<9DBX~)GlVwFoUrv+#O5;Gu++AI(>=h^H
zQo$kWY*U8SyHU-t(FSW!03R2qQc3VMwzg*Qb-jG!?W|TTQ9;o)!sL=WLs8WJdrZ^8j6EX1XSaFd%3&Xql}yr$Vu(1
zpHlG7wpXLb6QbMY8*dOLO^y$qPMZhFqUYePfBg~E-Tmf4nU@9c5@}JNky%Ax#(3vE
z-7a5rg~CSC&I|4X7^fqFq6TPluCwF|`~2JQ)`w63@t@Z>KdjFmKmFN8XW#z&zxDl3
ze)-Fu&xEaB@f|xHXJN5EH9rU5?tm*`qrGwF_1qz}lFx!kL%akd-&kbF8$Non#eVm%
z7y{UMT?3XAJ}+Bfx@W5NInTQ1D~2FBu&kYzrOa16HrqHIK1V3@y@FRaE5n^aUat5cLLxg9dG@RH*gnb6v415m+%bCRWV!
zrN`RA?oYyvYhcm6Q{4$7U^F!r<1=sjlo>*Wpg3yJcYM
zD|>|!sO&8*J;89hgUGNuGp*@e-r5yacT0rSK!E$068t{%eS~E!|qBM6o={y`HD>-Mjqr@M&y?PFMih2(#Wjxv@7d&e;PBvQ);X7fTdDK#k!`R
z4!}nlH*9K(*(SGR+ouA;9fZoY^Jwb2)0Ow!U8|ZO)l$;P@Pfyjr^T~k2>Z%=9=y&0
z8Q6?U3tWmFFOIWv1C&zahD{r~fOHV-6(k_cCzf(d2kHPISy$F=u)RT*J!f|y8>Vkf
zw?}AeJ9&Gif6f2$>Er+FhxGC3q50oE`VlYz;bd93nlN9W2|$&GNn-h~UA8W%@MudF
z!Aq@oS-rA5tuK;~Y<2t?6@mxuT3c6_j|Wue^qI4{Dqv;iPf-gB+>6H*Y1}coIzkvA6jDN(7S{WPPW{CxNx8}ZkE);Hf=yzhI%
z&;yQTe+UH#s9Dn8YTE_uRw`S&-ni^}dn{^EH8E`pAjz7>f*)jYH%R0x$6q-X(AE}{WX_eADD@=AyTxghJ!Gx1l@^DCJdfO-fAPKy3
zrv?I_M^YI~o=||1ed$fjs7#j=a0iz1a(vS%xh9Iuq&BV?9Sa*3F9x@K98+B>gUqYp
zm{Dx?J4Vk)J$@9E4fB^+_=20c1d?I5Cg64O6cGG^ICIBnJF1}1WnkS7${6nL9+Q22
zTe-B5uyFG@K&gohD?s*GBdlkE7M)SjGdwOiSolYR6p1Yf6OGLx?Ke?fFk0A~agDGQnmNfxT9(`beA>ZsZaAhPWsVS%nGtTEYBO~${
z036Hnlg<3Xk!S}GUWj{@tz30smXiYZD}T%d$h5B!`a<*=aIt#8>H(N^*8|3r!r1Ps
zOwhucU*{kFfBfLz9)A(}{W)`wUUzl51Uy|1AM`W_L!M|IHuz`QisHvAy<~$<9u8xQgd0waNe|sj69qLXaAAi{y
zoYEVRI$}Fh$f~Qpsax{ZX1l~hcp9UD7l7x3Ob#9%CN^owF)x`N4H(Af)-eDs{g}Z2
z7Lk*8^S8k66U?sO3XfqMJPa5!i!EnJWwX%*P<6?ynr-vE1GIt-)q>qfFvTpjE9Wf{
zq^~x+6mL~$R_1CPj<@1D`NTjqD*%P
z$XTvho0fHJNd<${8E{CYXr5;*+XW0HBo(aZC+^#S`KjN(b?ExXMtFdg=6^`=5WuR~
zF97nnvNt-Gm8v0pY0Ovtp-PGj@BJZfDJtxVu|*PiOD#t}Jf
zmq(pI!d4u8=iSA|qmI4$rGJBw|N4Xf@bg>X$I7Gs!TGPI
z0()^+>_2b|vf$qdg7J#s0rH6rRsDXG&GzEmyGdK=4~9o~5SbRG#Zxk-SWFeeE7
zBz>isU;o2D!U25u2eFv{{N3jdzx(k2y~RiVAO!RA_kZu-e)ZQ{B0~klYTK&&Pu60RNUDLxat=hJwJCaHV-DJ1ch1m$$vs8E)|=pAFS&yHdlJ?
zU|fZe&VVF}j1o)tURXeV&dlaRFC7dsE|xhWLp}#6v>`eS;)5}z+ZS~$|BAu<_rL#m
z{^`&7=+D3X2*BKQ1vJa3%C{*-#4|g6SEen
zXdhwqX@(H15r3trnp|5>D!nJ*-~t4i9B$8^XK3r32rMN5gUCwlj7W_=$~QZM%hL=|
zBO#Q_eAd=FxYp8wfPExuwe&ng7QJ@(I<|AC_y#Ztt6;Iq4WH|cRw93mAzzD~vm)4)
z34|OL*gcEkVBnjjrx^--l$tX;RHVfgyIN~WvO8~4^nZ<(LBGV%`u3mx=7)2yOy%Dj
zc!JDZLfby$umE4!x8R$eg%Wikd8Qf74ahVQ3V1kJH6ow=>OVhm>wRxnTY
z-snLh(@#R6Okgw{T=x}YaQ2?F-7OZD^4e?~{?Jv4;VVVGGKLS|{^nO7zyJKKWXhjB
z@*$#hYJV%d@7!Ptt?r$zo>D*+)l;-fR`JE~9T7o~;v-6136uwBhvbr%+_6fYEz#L7
z5*TX?%9v*Mo%dC8^?Ju@4>nLqqe;p$_Z7pE)~KTm?Ba6Cs%*BO?WR;CF0-vcMagMV
zMoWF}J60!~*Y`MFN1z#DTUr82V@nY>DZOOXsDH3c!Q)XCpLI=r9IWW>Gi$6nR-HQP
zy90=_25e~@!RI`P+icDFyfp7Wwi+oY?hFXRjO0e}(^A>htO`QJ(`syfjbO;2`nNx<
zPhXFre)=$i!(zP`P{HN+NOQ^_1?cKv{JW~nXWLU3JJc;<1qtRaYD>f7B&pbsbv@gj
zJAV&y02cw_ntH`v7j4~2?+Xz43HHtppFa7wpa1sLvJd$AAJ+Nphj0J@Q4T0ieME<~7I_DE;&Q=bIHG7ao;P#vkh
z+qhPrJ`nt7*S=W?DQ!|sx^;$&@?PW0yMKw#8U|*t00YLJWgIZ6BC6-?cN>?@R-LU|
zIyL~5ZvLxbw!9@xEU*H5)WVz^dW!~J5%O;1
znn|nyLZ^%Ym8FLbLN=mgk9yX8w{aP`k)N1@ch3GG_jQ&AT2@xA@3!uHMa=06M3Di+
zsMiWdPxVIKA!)xGj_yhk;GP=hCIkq=q>c?&-w(XSJnuNJTprR6weO{r3x7YJ+5UI3
z^U}Nd`yQ(~Ts~9jnw@6aQvoDvL>Wuo%}*60ocsO8H9^hTkN0tf
z#@o9cXL^$A89p}NZCpC$0p>E+p)_X(Rm1_pQ2g8U-NwaC1PYW&p?{o7+SHLru4&MB
zj4xb|-5i(ZGX~&St6r%IkcMoauu;u;CHrpU3UI{?Dr8`ea|{GaI12guA6hUq}1=y=Zdbjh5lQNdQ
zWr+@YR$`d2(0fv4MaFrzbGbA1z+>%GXOs0vjS6pqxnqTwzkl2L7lSjEop+xY-e5%n
zb9JrmnFbu|yNwG3TWVR>w-SN@R%5xru~r=`-p^0)6I!uesk{wu+8fvg!qC(-
z>bYzleG%>efBEBvO~UK8enwmGHm+8)3N@FywPjNm8w{->wfRZ&-Ed)~Dq~DaU<51c
z;TkBdu0*VR+kd;QugHyLRth_%d16-Plom5ts->iN^Aj;vM+lOpLopJayG|zkCTBO2
z`ELH6tuwNvUUn%Kb{an3*_rnq3&`T#a89gQN86)f-P<{}Rjl?=d*Z?7e!n;r8)3Wf
zFv68>GuvSk14VFR{zH9{|9Yy-{vSW}Aq)tV1jb&$`hP06cpke@F1aULMW5HifsvegdcS8_NMx5p)?Rb+
z0w9scjFrk}oW5qbrem2-(YgrUzjUy$)Ku9(&(!Z3&JDC36MO`Wr8UV>v7dm*%fy)E
zn&GZ_?ti&DsNEvg7`?<)@=d|S$Gm5l0{6f;b#=`8EShVN(C!&tSH``?H9%X1Efv`J
z6=A032J>~9@`m<#&C>qF@N5I@&vm7Wc^=|i<(aZbnoDThGrVRwOzx#fD~mK!mfb1T
zWy_P>J;U)3x?nTw^vK_isCCycUN99c*9?P33V$$t1-cK4J3Hc2slpBmEaSMhk5$1^
zdA)epzPo~l$w>?sz4KUg-7{=>L9FGnT&ilyh{0X3ofiJTx-`5xh65Mm<k253*V8ik*GW
zFn^`)Tbi}hQMSbqpn>Q)0bq$)yEoMV9$3y)#E0^T4U5Sdg>{_`i~Wf~;x|8hvp)R=
z#l%1JtH1o&L>K_^vRG+F&7T&@H$$F-fERG3bJ;w+K`GSJ^r}+>7m`17X)%>U{4KmSz^
zeq;iVZL)I7sn*rHW0&N*kI$91=eTNHg#CorG`BS#Mhn8c`oRurhz0Gc?S)UMtBg~@
z^F=&w77|vrSwUFXH*M#!_B4fOfCCQj=R@f{3dXm_oVcsDX9t*Ek(SO}bif_66MtO=ZTJF(QhvgQH9F&{wo=cern
zGTc?m$(<74eW#w-id&cO;%}TrJjphY&K%l2v-kGSz;=}0Yq4wwyMJD_ExoT3t7234
zb{IWcD_HvZh`)7;{3P4#vCMY7V1EqbjG1ae;Ab124&>)=oX9@O_DF;04+&&^0>!5?
z{n1@$G6QSxt9xwp>`bQObxVKSThi6MQ
zwj!WV;JJ^(blg&u3XNCS8t;M-JKH`@4}oWAvcXl`
z3|U~WyBJ(c$0A#lQCo2CTYt~FYy0pB*D-`xBlZBR1$J3R#Yj%VRom==E(oIz;!6Ca
zV6RmFZ!BN-tA6eDs5@aZ1E)^-lEpLLu^;!H0Pa`U`8kw}G)s%U9Niia)7I%WbVh$e
zdgocT186b8xi%KOD;+yRc&r6hzm{B$-^Qj~TYPj;FMGt1zI7xJ6Mt%ET*WOeM{8Kg
z_(!NdBfR%3Eby?33At(;-h~itT7z}&>8N1;)2DWu`8jj`J-GFgFQESL$j8kD1+XFP01W14cdU{?pEPsJBOhh@0WZ^G
zFo5>^8F#F18P;cHEq@w7JChQFc*QbVw)mwc62HvqcOSp~-v86bPrrNig8AbIlNPU&
z(&|kR#`XkXNeS$HjDVjzwfarl)d3z0!rAho$JzZ45oWfM*y=v)rtQ|Ln$U_EcDTEO
zs%=xL36U;N-ZyOz>YL0~Q{7;L@bhe8z?Jow3bu$0?}+0R8R5s
zxN+O;ww{MB{?ziWWq2SGsetw|0K|e3ZoU`%7l!43_{^ZmXYd6-;u~<+}G6
z2>Wu$?Cw2byv(xwfxt9E#3y7plAIwo87J`AU6-B+0Z4!FePIp@ufxL>!0%Rb9^fM3H*IeKxL%X&F|pIZ2DXUDVq@$&n_vCj
zOgisnv0O66k~%90Ku-hMuCzU``WApk5U-qSpWwIfXOtT(zK7Mk>+_O1f7bRIR3#Xv
zmC5(6!sP#yXKVx&*;j2xn$4oJ-(BjPy5eUgHj`de$9{hv`f~iy=MVqv|JM)SeDg2A
z`uOd)_^m&gh5Z>D4B6uKR93_zT1clVOO|yg&0GLG^+m%dy7VC;+dzzJV86|}x^H05
zQq%L2@4hgMp$&|bDOId@SS|0iniv3u95}pJ&4xIe6JYB*sH%!$D`4%rHSfcZziKvS
zjZ*_SJe_|EXD5Tg!b~3&_@Wi^s$uYIob9N-5GbfU#Tl0;f#v
z-kzrPSIx$G0MEL1bw7L1bLhzi2)=3pIC|AE@E3p2V0~H#Is+Cz%q1>ZGci-oPy8sq
zTi^Zm<0t=Gt=&I7=n(==!%8kE&s7lQdI7c~HBK`0ocv5%73^?utQVu6g-{AYOvjaF`vj-f6llzdS$=~
z)*p-aT8QVtqT-mX=hbuktZ@puY=?6xh!20s8A36jz^5Q_micVszv3T%h!24FA3lEj
zA3uHi_~~AAlfT61Llvu7bS9O~VN8-=V0G*|?J@zV(tC|dPgSgiOqN@9V>L8@t&Wv)
z1h8Vi*SOA>Zhf-Pt!7PSP+uu@yB!rv|92R}qUKXAsVorWGp8-zyxt8f$J?p-e~Uq`m34wN&QrkY
zG;am+SaVk<9F3^27-TE0{)#coGuXHckR5~pWMv^Z^Wd#^
z-YeTnJ{B-?jZ$NjG*xx*wh@2e;7D%Mqv}h>U|aUd6YhJ*#7fBsC_T1fEn!gdiZPsS
zHKoIL%eye2XjENwHTQ_>{WW6%{SSC<
z*d7$yhu0rP5xDfS`Oz{!SIj4W#yWbm5Hmxe!6qJmk{=r
z0dDW0%on1lf#a7Y&ekgIk+XO7wEPug$OSNKPxwp2^p@UK*|e1P<)TyPE5=Y$88sKU
zwA0f8yeh4YC9Ng+nknfOW58nDsilFvzVc}2q?le2)yEh>!uEfPF{qEZ+O#UWMZnAK
zh0PW%IXh~cm#l{^RGzf1*l?L`YPK=M1~EQ~sGVnz;s5sk_4&Vi`ta>%zjqG&6B`UG
zBMlLJ+?7T?QPP=f;VnxuiS`A@`n+MV3yBTH7gnHdOUy%;#%5g!zs>1I!j4b
z&Xij=1)8lC#F&eg=K|)BG3{h;dCm&x2x7275!un>qGf-t_{8LeMNiY5N@w%L+Uhcj
zAKtV4I4{`;U$^uv)V>{C`vz3TJ{tR6w9Nc4t2r9Cs-&qKGhmo_Ry*c>(eegFTnZOu
zur(2*$6(_t9nulF*FDQHkl>n@Tbj#L;424M5sS=cv*+Jb{KPT~p?hIsQNU|c7MLe?
z4vv7A@8y4<SqO!0AqK=m6{+E29Ksx*AUGn*D=mC+XM
zEMwk1XF#vp_St^c+geTAX;u&H3jDRFw5nL~+xC4@LKRVI77(6YS=CFogpEIy@uR_EFBB;H#kGYrat@_yoIUw}F_xy^je$G=f_Sh$DL*7*yP{
zJKB(19#=QP+ffhJ&m-_-QB@$UYjzh5*q5LZ0Y_QkOT#OsmJDpXfTpcCsx6*i7rabr
z2OBC)C6&Uie(V}TEsK@)+Il7{^Vwu(3gO2T;ADK=SrFmpOzoCkc9f&0bJvcSO(V(}
z&4yJk)w*h_sT=|WBYUj{L!}Tt_@yNy#Tn@yViLsxqqwgv8@bt
zzhrMtWe&qC_#DqWy!v?l$v^+VgJSFvvVYE*$pBN2C!WkvItt+R4UwfMnZ~#6nCTItOFl_+1BNqm
znLKQko2GRcXdcij03@s#BH#vtpFD*<%deW=o1G}}zrl8%dk#1-DG0flO_na2Zn=w!
zhxoSCZLbE}KRe#|YbH+<*u`^eu
zICL;iJo?QEM+oYAG$!8hQFWIcAfvaG4dJQuo2Cz@wRh@{&w^l7rIb+^aLAufQIa&a>TbEXU~ZNO4Jec?MBO8yUPZki4+CP
z2n;BEAp~G5<$sM+$7?KWE2+sSx!4&6Ysak};EA($&qHstyrTMG!)jQSoDf;raA5>*
zW!_S>y~J{EGvTk@WAa9o-R>GbKzFQLsbk-0xz`c}pZRF+g)+JepLsB$5uxI?EkoFP
z^eoEVPzenx;e(fpV5p>zH7}S8Ody;3Va{@4H
zP5nO0N)~o}Os0se+UwS}%fcUS#kbG*SzfjFz1f4QR>^zppd(2QqnAR=0}8VizI|Duvu<`3#ovfRFVx9e+b`m9!FED*ha+?0U=p6aV22@b1%N
zALmsGyA@xJjS#j`B(Q*CSGILeTkCz6t0ho(I|>KcB@TG?XuL!}VD;$+c=xkqN}vkD
z{;Z+`U$I!876Ovc8bXBMXPL?g1{j&G?(m}pWWKOz&O%H|(eASx+@NC%klj{c@&G8=
z*ng&3W1crs2d=Sf-4+60!Io(a5L&2)%7((&!P%}2f#>6Y{`TGa@P~&O&_8^5|KpGU
z`0nF}_dopMx;uk6O%T2}&+IEY`(Rx&*h|4z9%(&#`PA{6Q4CqUB{E9Oic)<(sNwD#
z*ze4`UNZ`iZlDiu(kDx@Cg1?vWXrl*oqxHf*NkGH+67?&>~)R0u>Lxw7q<2@JYOgh
z@!3(F1n^UK!lwYnx9H6!4@650#QTK_!auAJKgR#@)5rJ!=ZE+I{@>sK`_-vXZ$6(q
zj{TDyeiWX?O81%#i(Vp+nZ@#W!4LrX*ySy@B{_iJ7BMEkXTaM_DfWUPXiHdN*?*bk
z#zt0^4aMi02_8Sn8+p;!4#DYX8<-9-uZV$fL?8W841gH3&Xq&>f8Ku|?>_$Y{m&f)
zZ%pp|8wMXo2_Sa>zgp@
zF#&Oz&G1ICi_OcFYBT4??Bh}D*MAsOVO@?6IBd)m+(^cVKyARE&W!HkYmBLm+1Tkb
zHh5g>XTpyqa|gr92rTn!jH$~g`sjhwj@mvOZU&$#>4-hMeU&xNK5{K4bF0kp69Q}-
zlq}}8&RY$)ZywW}%vo{3>@nIyMcRbIe7i18=NpZRZ9(Di5f7Nu;izkvuEE|`A)}#hu;8(OsbELh%(ru
z4WFy>jN+y-{nz*J{>S?6C(38Md-pdoIDLIwKYV!qZ{PY~{N~3mG4+V?g-x_FyFiQy
zilBY0U7UJ7XLWt?XsVxF4?7!{!n>~o;!q3bgk2EoeDP>F_g&MfWq(+G|HMtGa-^)D#LHy;SRnp$bN$XZ?+?$d`|1
zg^gLO$j*;WMk(uD?2?AiK0@njM`Lji1*){Q)auT&H+C0F%~|-|;hJRwdX(Nffs^rFJ+b-ESM>A%#VXHjGDW=
zU04c$J;U
z*=Fl=gWj@%olu{LiTpaHH%Zz2g+sw0RM^H$*uAc#*@l&H1m#>?yrFq_&rl|}eX9w}
zoegWsu?mC@GJn~Rn!eG$@#_qQ$=XZ|&@&DwTQ!Ts?l!GBvp`Dv-)XE(XIV^{oNKWQ
zr>{^&5XCM|i1hV;r?D>fpdnlY2WCNMB%6Y<@4)YU-@Qp68H!*Ot5ETDv?wAPeaYzQ
zu&dhE7mNUmt%q%dxkD}E)<>Ij_1vi#^2QO$wIlFi<9{&Sw9EkgUYsD0*atAo-^dZa
zbOaeeXHThhZ%YPf18`q9t!S5JP4NWKWbHZ|&huY7P>l;_TmyV#N(s}{@GYV2garnW687R6-
ztr{;GL3eCID%pq~kok!ai1nFL8zo;@uqkC(9_0K_|60F=Ux%lD4}o@gVy
zN3V#P>u3W6n9fiZ^}LlOcIgPTj#7I^I2uja8GnfxV@;?7ms%SySO0v!^}I)U9|*ro)w@Kkn9C_!7xRK66?c
zuNXm#rHT2Nz4gvU?}_V^2jxp&5gYAL>|u@xcARUXWI#1HA<}mh36b?+o;Bqr|IpK^LOmhbEIgRxhj}l(!FAc-+ySj
zWYtYg)x8~Hx`9h%pa-{^1YkDqG(FY=b7MC}XT4GxyR>sa@r~F1cP_J{Qgihx)+RYD
zTih|LZ-Qg({)V{yHKwazeUIb=AyU2V6ix>)gh=vPL%bA^vmb}xWpB!-frDHoh9%y&1*ch2?H94WVr5P9H2Gvddg
zBDoYSx;)aIrdc(FNCje?fXD%Su--$+_^`2o+^oLD(3Dtbdm2ovzRrP
zuBG-m*iI3|y4eYK6R|gszArM3eT`6!$W?ifHdEqWTan?>zzpu(*N2EuyzNy7I5Y@A
zl8KO#C{l-!+-bUOb{D|sZt!V!V|95+O}9Km3~qhvS!C7*Pu_cBgMatIQdoqrTmjG0
z)!yv?@*nRavA=xugU9Y3ELa1%Dgic=gj&ODiy5V*-)kM5b<0d=mzo>1&A=;lI%rNS
z*^b_89h2x&Rw+y?Xbqt_A!O{bGbL#y-D_Q$d$?K=0VrtT8)J?3sFCaJ@o*WT{r27e
z{oeldSD*IN$8W#;@qa&k{QU>ey1(-|k9x}%huSQw7+Zp?HEKj-$!I6xbZib^_Q9IOqDb^E<}g34nhM0$Jl#BKUv-=5BxGy(|?jX!oL@#YKW*yc
z`w#1P@sIC*_wBpSK(tR!eEj%Z9{7;Wm(G)BuYIMltQ9sxhsT{DCAq9SY=Yr6-oNId
zM-xD?IWHaxe19B;GWH!dsaKU+_h5ILx&?78lhbF51-~wL*jz?k@m`J1XnKY#+9I?dQQH@=7|CJ=4>-M55BZc$p8z>6dV3?H1f-B
zT12Vhz;$_ca>mmTF*6`OxO4d&y8d#TM;l<|mW;)VcYhtM9e>mopHAV)af!{J_J6b{
zzf#TnFP{5QRLjk~G!yN80IX^m(biBKiMoI{nx2{l-2dsSus#BYZSyo@v3p0(H=5Sr
zIgNhO@m|LA>}zIIW@*`Zt5owE)4TC*#EZ8IAXF5{VKTr2>w&24J3qKl<%{Wp@P7Gm
zy_}AYd4H&R7SZKK(}=h&gN~^lXXD|4zl>S=ArY(G`N3V;g{)8wHiHvpBW;1_SeqS(
z*0|Ah;1yM7NLc`kWDpfz-$7#zdq!Q?Li4BZ|M5>9YOil;`-_JYn3{syQ~VIYWSh2Q
z*slWAIKj)r?zWHFWJ_(}hCWYan{nF}es@sNY=3XP+ddGHpE|`XrVF!}PU)-QKc7iX
zZFk$ZwJS^9AJSSPYfqB^rr8D$YCZ0-|KH9(|F56chxqXy-v4-Ue~Djv_#<`rfZ;c0
z=nN3uR09F&qyh@Ni*G0DBso^3r+n~FJ
zE%<32t5raOg5kuSwn1~KG_Z-$9vK{KbB?Ar&BA5{Yb>(>WkFo+Ahp?}CZ
z8}ND5`v%k2Ze$4|opT+Pl-><~YBPs}cdp5oa316Ye$
zq5`Z*CF(S)fW^3&U@2HG_YZ)yVT+YgH-O2Wr$}85xVNny0Bt~$zn*R%KvkPq1a)YR
zkueNSr6?$zB3_^Nf&r|Zg-wIb29=9k^A3NnfZt`S0BsZpG^y1nQrfk7+B{H8
zykc!{^q#zO0KF$~5z?nJ-7ByT3-D6je&7&hzkL8e4W3E9Y_oQQI_Z`VHSsm@tNRBq
z^DxKa;|8<|Op2}Gpg}q1z20nIb>#r2)yh7vQhOFDvm@5GNSF;?HQhgeg=ny4TLyoP
zcM1=y%GfOQSL^2M{sH&`mIyy=BM6_|#74nXP`hQWbH5+3n%3lXL17hr5=w`T%3CUg
zi7EF7BQAb`OJ<6vhc?!cHcqxY;S6%V-w#m55Ud|TxOxJjpIcHheEZ?l{Jtj`{mbLR
z=%3!d|L!xlO5eS=fBj9*dbBVeSv!A|SrJK_ZZ^k*rysU5;=Pxv&26lm&b&7Y@*ixYeyu_4Cm3yxY!()0ieOyjhW{|J&rd=;C8Iou(2;T
zHzG})2ugF6vs+7wm{8rV!%43d7n@T!d@71Gs=TMeqfKU>o9G@1VCZUdEYp8#pKf>p
zuXv`9GO$!?D=COuDdv9{n*-zM(Y5j*C-7DG>T62=pUH%f-lH4loZcYvm#@5Fy}0?f*jKJ#DJ
z{x1IM!~6d)?BloZ-(7XU^B0eOoB-4Wh!AjDa|tnE!GxuTXv$I;yU%hP2m*p_CGh{w
zb2ZRyv6x5Ut4G{u*$e@-wOM^2qSib%7@%%vgRIl^KFhWMPOM@lOzEowA?XSk=fI^(
zy6&3#uPsNZh<)Nk0vmsY&AvCNwk;8$7<;1))|Xh0cR#-WaOuhaFCIyp=^{Gt^%B-h
z-imO`ZfcX4D(pUWkJVCp%PjpYaG)g?g>c#cVU#(Bmv)a;ElSKX4Q>U?)!7CRu(Nl`
z2d`ajvI=8>W&lkH3p|_3aI4kV9t+?8mZd_gF=}oFk)Gh9B&>hZ6emB#Joe}}S$&9A
zq}30-cgFU#CL9NKhReAwTjtNLg6Nh2j+qArnXXcF7TcKLNg|qylqL+^jN~otFFWjMX$DZdO2DZk*C+X
zq|MTR*dsn(q$4Lj_4~w83mSK&a
zsR6a`1m
z1!S^;*A9Q!MbF$jEwHP$BG%ip2a^!vj7Qk*Gz}Xq>(TJUFjwN#On{8RerDnjv
zQez{U05Y<+J8}`Z_RXxe-b+fdwfBu^0up7mthg)o_Po}Zp4AQO-p8C#i?uUXorYJY
z)YH1c7$b7s>#&Cu&&?bma_jJIU6cp^Yg}Ze)LefMJV3VUIVT^RY|k!f%3AZS2JOGy
z%)pk7IoaZ;QNje6wseGgN~CFsUuEWqr-nVHQ*BnBTiK3UYsaJC5@G)BW+-UK>VsOS
zaE5Tp1}dFpQ%!rmX1(#Wh=HlhQ$`Ux&Q>i8K?x{w>
z?B<4DIzY_i9Pig#tbpVh9uHHPmNDCg)zyE}6DXl1yI~%iZe!VDt2^6h&8lJT_@F}L
z^0u$HsD*8_*ya7e7e-(B4?ks?q?PixVIJ870NAsNl`sR$hz9(0VuRnQr5omPbyqR=
zloH0o>~%$vhS|*1&-Uvr#@M=dXx$S8ix=i5$FWoTQPI(^v-lYj|ED8?-+fsAZM=W`
z`2Diq$ceF7JfhWdQ0v2uEFFay~TkK8O5x-i?!;M;AZk-dGTPl
zxhJ?-R=4>$5k6~Vs#6Qdn!a{0JeGeRHNnq5a>#%v*m&-+q1oQ6zjiQA8J5O2#0d~C
zG6?fQ039iTj{A7&V1&5~7=%j?1*S-?)sCfGOjCX98uzxrEa0fJ5@ux0XoY!RfS=fG*TuqtSSd{14+co#;wZA3Ye%oLIWVfm3^YFRs
zv4xd0{Bf}XPWsxxoRi0+n7*f?n4*(ehm_&;hPM|Zy_RQ292li-PrZLt(*egf_gE7y
zPhX4ldO9&Q#!P?^Tqq78x@ObLvBYcFOb5L`1wdw0D%Y_2Ij9tng@H$n{930sFn|UQ
zjQ9xH5Q0^kD0M~UxMcmccxFV+v-wF(ax14Eq55gVwraV*B`JK{V9rRho#P472wwur
z*tHa+nm~#+UObp7CWe2h8XfH5oDh7SIhc+ZJ-v3CX3O{h8^DvPSy;vb_OhxC#7|rH
zTH-&udZ%PkRq;s4DRiz_&j&o=Esgiv2D8h;+|1~jS)+s1!X>1gnNQ`sc56=nOz)N^
zGk**BPpxx0=y&UEV*A>`1e4ag1gi`xR!}x8R?XS
zkyb(D4u!{FI~ad#;_PS7T+&9QMuc1j{{o)bd9eP{!GIpHd5Pte*i6Y=qwFyFTaJ!@
z#)}6-4G-#EoHKOju4S}DU0mjB*4HlJ1r9L5WtWU0kDkbW&ZFl}8)L?P?O-OG!&*p=
zTv}}l>j!*4b)c?_t@hHv?9nuz4Uz(b$FX+L7^lpHg~Wdn!`BWbysx8?K%4-=Yx0_^
zZcyF9rt`1mO@wrma;{pJ)@E4Eerjy29q{e`+9MGI>px5)oR0>-@Vx2Kunay=fRJCi
zo4hR+K!`dszpPZY@KS)A3@7aEwc)^Lii)Vz(_S(uuY<6yJUgN!uf29_VX*`)|ZKM3yveG3v9GGvweckI#%?zE|!=7)inW)-6V30r>KYCjsj%v-NPoGO8Eq
z*R+32jsEcd$MFC1(|6x}((%8cF7vZicxQd2gL|yBPa2_zB}9X%vOLBc-Cj130&=E_
z?a!H$k}azHqn(m?-FwE%2GZEhbWc;BEwtGzIyn3qopst%*UJWCryk(O=@q2R6JX_N
zyZ9(Fqoj4&M*kn{yKjBb=%r5UFCP1l>QsMhps^G766O^pzDB`J_a*O}!$%S{@Pi)lGl6OU&4IV_z?Mrq-6qvfvThg9TD17k2Ue4ff1
z(4?%<@3TCPvpw2_7qVhO0G5X0&$bo-rMS=Xs+J!G=GL_UiCA@T^`T7IWP$X3mVJNk
zMx8w}(XJ6JP|C4WU$Ov(_I;KE$kR6y0E*6tnbIgR)~O3_d|sy?`74(F;NwzqrAJ-o
zQ4O|g$r%ut3S$>q{we3mYk%at1c#Zu=8k
zwgedIQ4wBDPwmNcfiFwlHQjALz=eNhCA_OICb?ii&)pLC4j@AR8)u{!*rz;S$!AaY
zm!Ag%*@MM2bteRjbAx@v+xH*-<(mFTe|#8Yom9jY1Z|d%j^^sh{vJGODE=0Tc*Y)L
zXhBe!ZQ2mv^}3lhUf>LY>SkPS55!7A&DrGWQ{KkNeb+rsIcu!T?d4XdFHe6=69yzq
z(t!!7C)tzs7X5O2TjngiA#gWP#s!Dw=ag)$S<`fby*}nvp6AM`wF3S1aVUWlaI03X
zwkN#6^G+E+Ej671gfMuY7ma-g{BnD6B8U}>O%6{15Sw^6U_MGQg!Rk!Ly?qDZivWp
zk2pYKOoL^LPxKigm)mRD-m-tC0{!d`Dl^8OQnJ=iO3+?zj}1VzU`K#{<3+Pz|KS&-
z@0@14++M}xg?iJzC|5ElDrr4M0oL(a#^v^43t#jhJT8TD17paU_^J(|xm>;Bxp%uu?sRiXp}EZN*`;UTzNoVYGiWVpdy{YjtU(
z_kN^~k?mYy}?_kF`Z@&kc$UH-wD?yqnyYfr9k0>*WrCG}YNnukgELj!&i3$wC}CsrSn_%K95BxMP(o9<@#0Q(TIx+zeEQBJ&dYx{=xn(J^IFEK15njQ
zpMY25UIh#K@&L(}l|0S~%KT?v1L$9S3Fe=U(Cgn%*)8v(q174~j3_WUwkQ*&b^GOj
z3N&U;KvU)t01&eh-L>SB>RPMk%N>GZx9K@-jpS=4Z7$iQ6rdie!CVgb6_~B;jIGUX
zCzHJh*-rEHoJ)U;dO09g#4FY@A`r7Efpyg4(Rf_7;;n=JC+x8W?#vn^>8yi{22rtM
zTb{(LtIHki>b10)i?+bNvm#Nk5eC&kx8&;q06cmT-WFNwRVi;DtbOkjRD8F7IiSJ^
z?KV?G4Dotun&|-2%`+%emB;n=%(*;HufqOeHBq!S&1`?5r6Bm1$D@TPczP4hU`>6t
zPPV~j^pse(k#DfKXX*e?g2)SReJo=_OP6X)6TjLX+k{bOX-xVvwh13$Y}K9ETF@@v
z@4JueXMxz~zID`+=8R6cI4$r3m-|l>_VfdVX>BI#gjv0eHbqVCwCe#X8{1%8TU!b)
zV+zY_By4~FEVab-@8`140(Vgl`0Wua87-$t1S#+%tirbVgb*dy9kJu%WH+Zq)LnC2N>gE0{EK@)o
zt8+3brKgeXam=sS^rv3_eh7BE1n42=9R?c3R?(u8(GY=0z4FVwZzK4@WcevH#!MReEj&F3EsUr
zSR%6$G*##A`cj4epbRl)@fBkz^(bItVfRdv6s%p2Dqh&sujQAFVSpohN;xc*y@LA0
z2-=il-rmX^m#M#T45ij`BwGq?C$$c%a4+zk($;xHYv5WFOG@Ll)kFLlcwK*4nO(91
zuao4J`ie0;#0;2GH65Ndx0X6Knv~c9iks{eV`w{7;LZx_N&OV+xSYBZOIdk^F<&u;
zS!+Ly5r-8gvu4Z)h%aWyIz>xfF@{HN&|>@P8l!q$ojuiaslvwCB`z4l%%{PFo&vCs
zpbDCAV86X6Xet=i6<%&)*Cv0rp+K2IDcaIYS6~DHY;5U%fr$mr&)2|5iV>B)Crt4X
zF;!c3j~AF&SpvV3hn8$wPPtz&`!svDF6#|duZzd9C)bHn3UyG<7(i1JHt}v1$TQ7X
zj3HVSb~A4ldkfnaoWu@b0|7&HT(NOBuzdAF*{K!zu+ef>u@3y?pv`~%1tus}#snM8
zuVsQM2f()Z=pdJqy&+h7@fhkLiC}8;sbMZ~)q60Qj^ZVJy?Lu&ZXOxnNg#ZB^iTRzVEI|8pqz)Cx8pIk1B`C1Mp0vOlv;Nu9p0X;|hEC?vsrKpjTGb
zs3Wszgq<^D4g?<-LGz?!!;tO#A7v?E6UUUk30!0W%5-
zCXBG<>TZ9d`!zG?tB}PW+&h(#jme*0zI(Q!g#Z7;Wg@@OCE=2R*Jj0XLhOGhYaMA*N
zow>63=bATBWs>$u$ubiw+Q1%z%s*!YY5`k_^x0N%Y7D{WKJu&GE}XW4r{;9x@Jzo5J6_e0!Y~AmJd;PYD4Pd
zMDrpecrx&k!gZ|(fp-~|J@2{?OiTL0&4_>hM*x+1Cs(}hnKx_Ky#hReKTd3seLjP1
zfBxZb80hmn^m>=TF?tAm{6X-|zK563&BTA-Yk>L?@UYk!6*DbBRfat>n#n%l7Z?Du
zkFkKxtdxdVD3jA5kuvLD{AmN3XEp^tba-4H;*GiGV_{NO>^D+Oe~STNd`&ue!_s+7
zTM+pGe;jleJFm6p4QvJ#B#9
zF?&#gFJtsfP_7f^4h0JlBcA_6o$c;byA1en7VFCaVa4`<=i8?~KgsZHr{Fad91%=Q
z@i!*iH~eW$j@cAf{pQ>cNIexdmQ59oU|jCG7Aeu$DRvu
z^nz71+M>1=@BKl)@WJXJ#pY64E!*=ud2|Y0ote@uF(Os*UX1DB0)HgW`=X&0AAL{v
z?HZD_Vl&uoR@~I~^UHJ~Vp5$gmcheG^*v87)#h=a^Yi>1uv@ef>7|k=V2gjKuzA$n
zfyK|x&(FaGEA24Rs5cF-S3q70lS;;@p38dP0I*yyMcfe>=0KdBI*KY-b{8M{X#-m4
z1Udq7m(mfGQ?B^E*2b8>*Ylr9TjubAfSEIz8hdHx+*e&EcbDgPvUboO1Okw&+Gb5M
zG-KWcQ{4i4xuakm9plN@VO(gf&h;D+A
zk59Cv&kGD_{nYydTpE@N+kXr@Y3zWrd){4B?}zt?4!W`6;A#U;R~z`}^j)6+MDRly
z5r1xW+uhb@B+zroHJ6jydj1Vg#?wGz)8R$I1ahneTP@`{DpsDKgH!NnI=nRS%7j-P
zrU0gZEI=2kX3rZq$FP6BK+~dRtf1|IZxy*sfR|;@VbdQz`Rhj^I{?I#3ZU0(h=qKhgV{HrCi9;6BF6+Ml-o
zzl=p0W>vQj%&%j_P{uOy=sH~dz4g!BfF@I11%Ro`7T{lOV9cBox*uaD_EwQOXbY^QE{G0rwA^-Wlak5c6
zmP1voMWs#l7Q)*NMBt_EGCL2vwuzk)5f@b{tDTVn=W?)urW?HwJ5FA<4r&4ySsT%z
zc^-#?p4-^v2l%Y?(XO+z8~kY5PDbL`rG}T;MYISte6F(7nw2$jt+`R9JEYo?6at63
zR*H3%9k_pNgDaE8jyu?&g=K7+qF`&Z_Oyfh4?6&6t>f!;1U+hhw2aM~fE8*qDfKEl
zgCc(eXjW&YXazeNW2GDH&$8!b`)5$Ya7BYR102~JU0NO!9KP^6C|zY|);=UHpIS75
zx+q*72@XYpS5SYz%lzA1Rvuj{42~fMfXS$Uin)Sz
zN6fg&4pmaRu(5#x=L!~IU6d(emK|{EW${!2p|KV~1{&qi#MxdAxOxL1FpK*#J6O+b
z^UOfUi~|7
zYWJ+7TxVNf2&$EIbakZ^*Q@Lx26f=M#zv}{s^DZv+-(w#{T}Aa1~MoZvf;Bz@7RMS
z1gOaip+;h>iI)uoF`xpjE654nzcNv`4^w}N5*J&mzic3`a@fzhxp#^+4r}-%V3gR=
zK(W^iWS>11k#z$N$({%Pe3U9XVd<~yVyzG6ZK>JBMymZRmq&x_IVeXfWgPsSJ^qEri{Ny^D@kV6@h<8
zEO;b%r93WqpU06&Dp-c2WGNdPFvWPd^TF$?CHWdK^us4M!L(=(#N22L;E`%U%rKCi
z@E!MchBO{Z&Z(O)FR=?&VWRNx>}{X8%nC~|RnF7{_y8Vr;T$z>zAffP
z6i>ojaB@EuYED_0x62HT)Ihykbb^0MY@%Hj;C{wt67Zyc-C`KK1sR;%%UCMPqOzf#gZQX_N~RA$xZw*V@j>zh)S!SXC`2L%3zI(qsKK${dp&v!qNCK){W%gZ79*v>qbXfXQFZ6#`cE(~d
zI9w`&l)(XkGjFx5JQ$wGud+jV^#U5^CDNlhw*VHm$Z7FhXs^rc1m+*ek>}~EutLVU
z?-
zw=cxyeB1c-bl
zaKaGgmS@eE7{O+g#6D98L=Q2@V?JVE0pnY;u%v
zM!x&;TPsz_=251&4`3Q4Z&S3M%YpkeKNOu5!V88E!tqFgN1y(xT+-`o2
z71yJq1FdOL<1c&^<38&
zIb#`!#6y-mV?ubcEolZRYdK#w_n$u118@urP!iB!XTqtc5hlIlmbR|_g#}0d@%Yh-
z_OX*&@@{7^v+HE5c6NWJa^;hcKdp9~K$>a5@kSkIj8(7~C)}bW`SQo#KF@z;g`E)-
z9-X#>3beBy7iB<$V0Z;)zsSm`lCwWXcqw%iK;yM+70{W6X%6KnQdw<2BUbcFGct1x
zCY*M_$i-H#Nx%qwf(y)<;wm#sDlZWq$+@7JC$cnQOKs-$`Bi^!!-z4nu($)%_7q;7
zZex{_%yvFEE5FH%=Dh$MM#IF!YE*8|p>484bacJyqbVC-9HD4tYwM0tJm*I^s4)CS(uYre|N?%S*x-pEpF&FqBkHvEW!=$>}5?bDb;O
z46C23^!`m|f{F#BX3F#J+}R>T;H8%tFl>CT19F@n1D6#}>V*;3;#wC4(UX>1PLcdZSfh^qXxVfk!{
zdc8fLfN*nkUlT7@m@Z#)g^!_1*%j;alfK8Z%L6CL1wp1jTg5!Ay=s0-loFX?Pf
z);nVk{B^L-r0v~cy49@bvsvU{VogB#2YK=YbZAT;nZmY^Ls;6hFMhYuI$#S>XAi+%
zAGz$_T^cwQKCxbH4STvXFA6@|&tzHZ23LQ~_+OXY`nA@ovIlgYns^zsS1$+#@L=0V
z>&q{Gw+6g`S5o{`!g(f)&$t?y2=DMOez&0puMYypgH&tq%#jm#<2=>ituKBzP}DA;zvo{WE(
zxJn)*0@`=+yP2~Fg$M2Io7?+56Fk9BuFN`_SHIg~^MF#UC0UeOLoiUG{GVc>s9$T%
zbDpfUsjNb_0?@qj2J4?xsyJ7_o9i^ZU*5fzgN=|_rmrCg=@m@x3a|X}4?p}IavFl!
z@KD~u9IOmDO7K>S740^qTy1VlK`eh=p^-I(_0bk~;570Q2a$fl+=sBAK75FGAOG?|
z;EqeCw~P_b7M8$k-kw%XiCr!MX+fCJDf3cex9I=8W$c!_r?-sVB5?DTv0HAU-ZI8g
zid_`1NR1Nayq~kCf`4FnfTLe)jO76=dv6WUd<%O=%4Q{{8XIa|Z0v?B&>w%}fBh8i
z>`zTMl!71d#uJPff*F)_ci7MDEAdi01y2{Ks`{X&AIL-8j@>H(9D8*);LI{7qG$SArU5gj6
zC+XU-3c4x(XcP_a7nvyu8>C{|@j?^SFk4S?j-Nig$y2B2e{KeoJ~#F|oMU5gm{i^%
z*^#MP=2d0_Uub2g1W(yKpPRex=$d0tUf?1#;0!Tk@>XNa2ABhSO*IdzeYQNWdYd-!
z3bpA;>~J#rm{?yqQ;>fXC{C|hX${-cs{vil;g2e}UEm;XQ6;i!y~xZ%1fVN?m@?6^
zuL|zm8=?)c+R0aK&%W4&epyk#P0AgPwYwgP|NE#(SWag~_>&^vVk
zv{bOu0Tfnq9=x*_Y#zVJ46q}fIOm=zf;mrfrn(}r!l_%k$_oHaP^gvEEn8z1*h4j_
zQF6W4tXG*i830BaFjJ;-6GG1}SY#lhQAFBhW}KZl3LAl!$(HB4F|T&83ekOEWQJ0g
z>R1Xu0G%~;yHS5W2sjN?VChwJvn|rVH|hXQC;57$LYx}FI7!A%7nv~%Uoe|TgdM&|
z|D-<;Hf_>j{US3CFmre|w=3JL!~$Rz!^abZh|;d|0>KOz9(D(UbnoMY0kv&YQnFL*
zDl@&CC&U@v+pCX}POS~tAjPb@uh=(_B>UcFFYmU8690d}|0g-~(H)M$q3tp|?4wZk
ziJ1VGEwk@HI#dbjNsrl8b_Or%YN{1`6TU1+{RA%q05^GNbd?=FTN5Y~+c(aM7>YgP
z#T{H^^OW%_J1d8g{qTh{T^9)*#gkN5sHKD>+Xem-r}UgiQ2nd2-t
z&~zF$5qp2yIzFS+o2>F2WhC}=Jy`hc<(4FaF3QSv7P-f2ezeHbmI;5^P<{cmDk((Y
ztYo~&>Z6JZM2wn1dv@M2!8!8qUFEY1)6ZCySyyXd{N4L$ShFE452AQSU*oAP%g+2)N*fLArM?QxIQa@uUvePcDXs55^zheCgCn(Ak3AD*!~;CBF^M$c#Q
zE;dVz6HCgVAXd6*H5vBKT(P(B(bwt)AOYSLue@0(+HSInx7#ZZWBbna}Ccv
zr{A=Oo!Ws9S(~0~gTsZN00U??e5}t3;y!1UrEUrCkcnmKp{U8+sIV4psyA6>CRCDE
zcrkyyI#2+j0Y#Vq(3?JWuk#tJA#vFAW48e_G1+^%9daC{B-%~$8YRvIj8Aa?Sst54
z5zfIK<227V?Hs&mod9lHVaAx-48TaCsj)H$<6|I_F|8g
z=lb$Svwvn~X3%h8ec8pP1~b4|tpK!_%Kq%7(r>XsLGg(&F}&~$lj;$PXLB>B#6Ev;
z*I3y-0uahD{#r{^_>WRE7QA4h+Lmjqn9V&XeVT(nfaye-_vZ@>Q&6R&lw!8`$7V@2Y=K!xQO^fYsiqN4Dx&
zixrE}tb+rl!wJt%8|XX}k<@<{$TTY@B+CF)two27BIN=DV05RoC~S=j?2R82ub?AS
z?C1GO?%t<3w;JYq+G3NDX6#K2KG*#DIT!?G)ik-9oEYr7i`@Vubpj5fK5bwYUsy#}
zav6zQdd!Rf(1pFePk-K3Oce+H5q@^zIdXs$!CXHAK9--~$tsUn3}S!fW@{~P`=sP~
z*uvb{c-{b(>A<%hiWs~n=HqxT15TV@&YypST_+x%c4lV82I#ChPHV3)A#XV
zQJ>?RZ~pxQXiR+g)#rbHf^_`PzwoJ_^oA!LQk__(9qhTEln0wR4@t@9S2IjC6>;hG
zY54bN?F4UWL3MHXYsp`4WEj-rkh@ldw{{S@D||cF!@}b3Z}gpg$;fJYv|E$qVoO=m
z$i4vxKjPio(pMZ=HcDs1O&uy{f&AB`Nrp|TvevITGK69wE<%57J{dSEau-8*+hw0D
z@fAk~%v_Qm-Jq#(kANtx&nQ@G09$X=0={HqJyCLPEE9kj^3MRWL`=md)QF{m|QIl9(63?QNwa!lCUvXrrO9PSED?!F+
z$Ho-=wBXU!8_j>mUox^$>Y67h(+g|Tz6WT{tMlXfRz~GZMyA#@Ko2*iXj11zABM{ot&!_nDdVph
z8GF&farIi+$5Pm?wkwuhT(=Ju5QaODfsBfW0JJ8zrihI-5d61Eqil{PK&D`qj
zyDiH6&##Cz4poh^m9#g@BFXo4via#Q*aPjpZ^pJwsbdKhaH|9sRe71`@LGX|pW~|;
zUZqIWmP~&-k5+7?nRKP)wHG+DH_CipGP25B^K*(9?Qjolopy}fCrXEtkNWjS24a+?
zk#{rseY0aV{kOR3v7ah%>WGEq3M7Fqf+Ome8>|4avT82CjTzGk4xp;7F6x1=2wuYW^e)R9YF!386QpdOHvqE6IMo!7tN+CY9a7KKs)
zun?JNFN0mnn6CMTxc4&#sR%Nii*n)RfR|F#%pOc^&*$ax5V?3N%slELxd1$tLydn1
zI99R37A@BsJi8z2Oz)KoyZ}q>b7vz}z=`wL-O?8r#AX|NY_BAHO1U-w9Qc=QLY>X)
zT-qXi>yIXb*_8+Foh2A#o3q4vBhm8?L(7>E{MrsWg
zpJ$Tp{`wpJus+~HALGNX{p#-9S*+;|#)Q%?(RJ%AEVUodWDcc&LB7_T3Q&
zhqB>OV2{%zelC#X78n7&EaJ)*ZN|c
z7HMpwy$$GL9q;qx*%WYo-12{N{V>-U{C(HfsZ@&agWqzmF>(I7mjj$0_>j)
zY8e@;Z#PANd31OlV`&FWpBFB`nGuQ+e5sJX6Wffzn_ydKt!1`1bX%VMigO}jHERgN
zGlhY1#W~VjdK5O+RB~s!hDY2WI
z-;BJ;&2Cho8>m_fC{(4~I6y^1F$E^AwQ5@1mk{ciQ@DWUWZh{_~40ZVaC_>o+9I@B$h6N@j;AW@(pCX0!s=U?0Unm
zMP+ePGi!;Z*88eM!pOJMiCfginA})qUafCn0eD+1=(@%0dQ4^m`EGx#g*Abn2A(%7yKAh=xhk!D
zQrvE>6u=x)Et<}fCk{9tln}tb85iu7au}t!U|PM8)&Mno7RLeZOU5&yfP`reqEs`i
zY83~eL&fufaRXEgMe*dATlXMJ^~4aIF2W;|z&Hj!ROY2FGGO5AL6lMp^c-dY)CKZs
zBe)b`TLgcq!Y8O~q113<(P5AA+ap75m0J
zv22H8@Y^IFpcUK$;>50mk7sf4A0rmSBqL*B7A}X@SVwGcjjC_zPw_0cJ!5l5jN<7*
zVBtP5cvX*&`G~8t(Jtrp#}lNDBHfGus>E~eOhp;M+j6d@1x@QQ@u_(Gjcjm6jC0Eq
z)R=!`#%nhK%V+@?|%uU>ttM>5XB}oC?sU|+R`Nn@$^H6HR8?7DI3$L{dCL9kV#XWn;Pj4Qu
zID`E*4`0qOx}ZSLaBDGe5y8sWH~-Vgx4yNTJD0GF;y{_pY_*$UJy+Jdh5+KG%`oGe^%r0D9wUw4>L1kH(@421D;7G1IF#y?0jnDndyJt
z8G(`+uhALO5FvrdIT(Q0zp3&6e{}iFb0*stO)Nnr8A#Su=Odm(Wrc`I*gvcBc*Juh
zDa|^=-I0=t!HcB7oun8!F6Xz`k_LlrU6Yu{Ie>#L>oVPU9B)#QIV|Hp(6EC(E6uKV!cFx}t1ldExZf{)XaUr&#(
zb`gzZUdZ{;+2!f^H`I22TEk2FH(-9)@i2R4EkEX`%#rzU)nY89EoFN+(r(HGon_wg
zs(s=y2IruR3lAklTtPOA*L0W@K#IgFRo+G|;(P{F?-DkZU1}1Zj7~Fpc5k6EoD!}rq_S$yazFuu=z6Uu%$4;1<2|O&s>j=0q57Xm0z~W_^S@H
zz_ebONdsC=gVEv!q!@766}ypGs{x+g4HJqJ;Njvuw;yVa243``)ZWh$8?!MH0IUGL
z=3abPNs&uyxq|g=^WwO14KM8_HNY@+EqRtD&&xD%26P~7B$l;F8%uwDk6piwnk5Zi
z=ce+5+2EfgR#+N~YjOrCC>WoBZ#zbq9jrH9SLp58>FM#&$@_Ju$32a5pgIwoVOQYQ
ziWGSE0cVPrhK@`xVJo@bY7&->C`Je*0$!jSF2Lr2
zxjm%>d@Z4buC@*}-1Kk0)!@Kao-g-uo?M)+>_Wb5bEzHLCmtM2uZl~t0fmSkgUcK+
zbPr_@dzM%{xYYaNNqXve6s-g#Q$37KekiN-v&5#53~N5eRMmeAgGr_^?7*qDL)XB1
z_lm8iW;+T`#s;`AFz4V$gNg8EdH}zT#I|fBNFgk`ngKrBU
zhXn|jlv$M}(pV+gazS=(BzBUdWah)QtnS#Q&AcGUbrY#w7Y|zv;63bGzJxn{e?vhw4XgQJ_mQ}CzQ2M&(
ziET5xNRI{A)wLy9b*u1O>gNb$Y$SG;92KjlJh@*96+uF5^FE$I!F4jzAiP)ZNkx|$5^H3np7oY-7l
zppVY$?1aC7Ph1r-BlDb9(wcjd)diwcCb`Ir&%-dDn1PlN*f%IQ`r`TmncXCS3e#Zb
zVULUsLA?|wDKHyn8r&~o=MW|f*5?d
z*xR&z6@h=}tB(2e+e6R+9%L2kV$l-9gf|OyWS-N8Tayh0GVL?qtMxGX2IfB6oU_&t
z`Hfj!V6oQOsX)?^2aa8n76_zsw+et(*B2PjDKD%(<4j`NAr);g1-NVsL%X`b0az|U
zfjI-qOj}rxnR2n}{Ibiiy1*$DcrpX?F`+!q28e(7FiOd_l={t8E^Z6FaV=-(r)SS}
z8Mm8S%FM)UEo{GIezE(K6stpDAtrx>R_+xLm6pt`_zHP&o)AFv4{D
z>Lxx+q2gAbiz=_HUwtf`aq&&L?;mV&$!={}yI$dc9YUWx;Sfvd#(}tU+0|_#)X!pz
zW6ldEzl)~vAJC>SY2a{e6QRMgMrlvX+oPrs7AXKOArfAkDVqqbRT~p|s84W2xrvEf
zElfbt77sdYBDC>j8LNP{S%aDmM2$HGFbZ8EwmS>eCn{VPq>l*(drI_tBA
zUYDYO_(<$lZ;*btmEv9QPG_`zs1x5uH{6@HK!}^jvweb4k0ii%0gJ3K!jy1QOL51OxD}6C1=b#
z$q`U6c*X3uDZyk{yM@fm;!VIi)8W0EUYqZKJsJV)8_CZZwvbsgz)veiZeSj~Twz{}
z12*tJhhqB{GHYx}V96Op7K-m#8f(FbWwo_5Hj&x2*u-Z#LL_sUal+;50NtH+mbQgV
z2|)mRm_NUoH
z(V6hVcvzqbsz5`YpC$7~-kPhqd)uTTx3d?Z4CW28U>ux@27ufOX+hSl#I`c>a>63D
zXlhL?>>J=m80b+`*+^{UaW|&_%pD-gu+j>K$((qTYC$YF5?gA|v2w+HxafTGxwbe)
z&0ezW{@b?#IYQr`U-3WfIb;1!J4=3lmt2Nrg=du!OU)Xp0nKq%nC&jEC@QyR<(jDfMAzc$C|w=l?lfL%p>*8Ijd8d0%aIx2g_zfQ8l>P(PVE-!mmthF|?VH
znBG|Inh%g!#cB>oSnEZrfWEUbj>6s((7F{x4dabAo16jN#7xa0FIi87Rfdg!uJQnX
zZO2!0wT700GG>Ttz6O&$d~tP_xlDp{cu6j{HDwJ5;|Z$HbAD0AY|*>H9kM9Z2VGOv
zOdy{O5KaRyrU0kPDA`Zd%+T+;s;m|x3oUt)C50itM6$DirZLgTOzX-jCR@&kn|{^zrLc2l_C6hd2Xx@77`JJG*MfoZd+yvhzPS9bif#VX5RLN;KvmsX
zm@Z9{thdfoGb>jkinVLX3T22X)lLI1z>>1o+HS@iy@8hWx~44Vw@O*nw>g>KA-5sS
z-aT>vYJIkKxg!gdDrb!ie%W9ocf9q%vxw7|Qin-v%4#vN2uny7b0`OYjkZ>p=5E2W
zX7swIEHK9;(Q`+}1(Pry68B{+Fi-2H%{65W;C#R~29&X=M&x@7+ATP<3)8GB3(tue
zkd+5HvWE|Ez>9-*17cv3=~ZN{;d*Yh#3c*ZkyMq-SiZA@W^PVc3M{t@dd$coi@Nl3eWrog4`W5py==RfT2WJ06=tT7{Vl
zbq;U@V2_w;d~mD%8-!KwM`|v!Bzs}?wyu$p_l`xEL()}+HCmxy6KoJB>?{})aH^ik
zRJF^qf%aWt-G`uRBu_sKXrIM%WstB_xRIRPs=_1;P-C8t30$pz%QSw2VamO8v-3Ig
z>{CwB@6X>~eV8Yg&p*$)uVr#oR`s$))G;QH**(gDJ8Bq?G42awUr%^ys#XUkoUu!-
zu@Vd>Lk*p8aa-ZFC@`gCp<$l?8xHb|0|OkA$DW#`YYDfxx(w#&!IpX2iKmWyiOGU$
z6ytPT;RUF@%|1haO0jkj#|r!j59g(THOjWad-0wpMSFEx@W3mK?Evt0AX=uUZH0>t
zmbI-eCMM@$B74x~jt4(ElQ$Laq60LO(A!*ggt7pfgIF3SdfmF`4o_Vq#u{_Vxm(~3
z$9ZSquQA`prosWDvUG0qgC|DeLDte8U}|@X#jds$4oWe9dtg4+3HTOy1}bYRsmub3
zvF$wf!KErC!?kxX4I9~CfaU3DHKuPn;C#%vifY673KHf3eP0p=G2hHg+s<>JOe=Y0
z1$s6?ERV^~5qQ8@fE>%V!bhu#1*EFl5fWTnuq7n0Fj$#=yS%FjSBnlZOBAMdY%G80
z)OoSR0vyzTZM|CXuJe!t@QCGs8#^-Icw*kVQqyeP3dggZ$yXU%d+*WuqaOYE0
zdat8^s@TzO&4q#GIp>XqdmTh%!dk&Q0{xd*@m(BRi(REN$DV
zmZQ`Rn+gM8fS*Iu9DOx20$F7M*=>bS%{**Y#K5!~Yr4c9G3@c;8>r>B!gWgZ;(F=8
zaJB}2!-Sbkt*UVHbK3z|08%@P*jZq>I!on3pDn5T3Z5(5K6bELKYRy?k5hZafU7+)
z={Kk_Y~8jl6+qFmD1(fiILYhqIb_l4@EE|?wiPacg`vt|GPho0@5}(~$Luw>=(oji
zb!I7E6SFEZU{3w0aqRGvH)|yCen2=+c>6Ga9A+8dv*E7pu!-3(g^3fk?Ro1EV||Ot
z(px}pfY3a)puu%#$F}nv-?2zA!>^Txp)6DHXi&=_=IQIE!n=bjWl~(GY=oJ^o9`@6
z$K{!{ue$SU!hZJ$Z&~z{Plp
zxrS}c>$FD1u%LWXR&K&j9|-`G=RduF=*GeYF4T9M86w1kCQAm`mInx2?i*XUsYY6_
z#dlYM-0|4q&Em2=Fr}Ammp@^~L+FO7Xn3f=Ms%K`(LnP>)7W;Z!D}Wn)d9PM8%LlS
zcvvnzGoUbRvIux3|J{y{#;!0bzWDikjw9i3f1eju$M^3qyuX;|ze3D$`+c5&U+g}f
zEos@uv%ZYx0p1w^%V<)L3RujK_EO?K~O(?{v@V$aU+oTz&oVP0e%d
z9RuFO)-iU=fVXQXt$G&5EO-xp`xb}R)FAjctOZfCEJN%C(|=}H)w#&f^R&GN-FPdm
zwdt9TGq9Y5%??PjdgkWL>uK=Ev#Ynu`Ha6@AB}ux|8P9_ZYMr9gX~iTkR~(iB=_hm
z3oh1C{V1)WmnlYtiGK+ln0%)Ls#nZqsV(WKp;ywTR-2gm5<19IooFO~COQL_0i|)x
zFUfjFV8xD3PCp#A5W0}~oud4|Z_~)O0c{RWJ;q>hHpHB4j1w<@9Jv-mqs3n3ZnJHYIWWj(q)cHMo
zi>qr9Tul*V84wGz48S54?^*Q4m#6^PFQWyIO>&VAAR;tZ#()LH>xZl9S<;@(pVhUf
zV9(Qg39dwAJSen*&2t!it|z-ELi<~^`1uBfzFqEhi}O1^m4EDi9ep?6?9Rut#PHpN
zGt?d?y{Ko+-)Mcxv19frNzTh@lH8CqJQoa8a;m^@!}+L!v1r4dn9@x&X<-RgjLoIt
zg7~@|y!SycKUo|6tsgzSn!b5ohSPNnK3cVO0fo4m-(
zTWqr5uSdzfq!PTbfHXAz
zWhQ~jV|SMJuqTsUmQiIMO0S4Pga^brhpxjr=HhZ436LG=G4DCP@B(@z$4H76h#fG%
zc|4KzHM|FZeap_oL+LfZ7+Z10r)t`4E*mKhPpzjRd*j)jr`P$Eql-lcX~)NR6nuNx
zE-xPCIopY46fq`T?n_0rSa>u=oU=u97v?u|=75^f^@f@ZfnUlg4bXp?prg9f2|t`>
zHuYk`$}8(DMA#@gFc3x7_HO%f(7%XgA#qWJwVh3Wh=sMdU010gHAXF(d^pXblTrgP
zwDsEXl&15xIx%^a9`{8qKU1^Q%hMm*Wy24@9*uea@yoZzrw=>0fxFJG@4EBLGledh
zio(F8gQb>SjcX+i1H)M1OrkBmA55=hn>@-Kb5t&Xc7Z`O6C=FEP}`yOirhwT#Y;||
zksG#u#>6sBu~;)SnM3K783U}EsvDM}4x&d?t#7*3$u18$(iz0CEaTK4F}=p*=NtrpDBqH&PAQ>FWG^o?QO%ybbPu
zj{c>|FGnZK-~MoV@o(hyf5z_*ihqB7?}y#J7him_mKMv7NsGxaXJ_PDJl}jCBjyC>
z*Nmt;oT7i6C-c1Rt9tRWDlA*}KzP$6;~jZI8Q!(bYRB13@Ank%esL`;oXw2zcqaD|
zBmE$tLL7B6CVo#m{EKVZ#9KQ|_cklUFcBZz)0cJZ!0c
zsB$N7WNBr3jZ9CHn}%*)i=L4I1|UU#0*bi+w(92qJy1sDDS3zC
zW@Bcus-%*KORq&wHPhJu=UU)exeO=8hm?E>!lP`jM^EuGfw+xVI1F2IpDvkyieZbN
zS~%!req=pvbLmU{vBn*eIhqceI*6ueX2Bw?`mFZg3W8{Y>Z*skHx&^*|AWfiqJi_FOsO)MpSa0OW#HujcmZ*`aA}_@Lbq0y03?&9Aum>crk#O#lb%h~Y2B;Rm4_jEXw&%;Y6+k`04){{Lt0Ib
zgcZH8uJ9Q-i~yfz4G)=jmL|_N^w2f#t_)wB$kVF$|>1TC&62Ne6N#(<0J
zi>P+Pqs@VbTT^hhichUK`?{Vh&&2}Wu#8v9Hu3PcVi0)ih8iQUO4lBL9WZTYnppBs
zX2PBs&ooXSMeJ+RHCPXAYhh7hc6=~e0K%3v`CpEcMCSeC%+TLc`5
zKx*=;Y!-7X)(kIC56~s^Y$;0_yD-YMzP~12vr&~=XVE+aGs`jgwpvZLsH?@iCS5Cx
zXe!R=Xp@mBMu5oiScd|C?}@9gN!L*Z7gq{vRj}(ED-qM3i?{rHRDaO(Ev-t2!l7{2qVUJP{wnuAfpsZ7#Z)g
z7$Ji20kkJXW@|2~9bCbjEqGV;E{`}nV9X3sa-kV%E%tJF1zAnb*5O0wHj50_d8?{}
zhc1lHgDY4P7m?L}%wCzeBLmNBbyiJG#oH1Ou3+s6hNUKGY5@To+=jt01Q_t8&ejjE
zV4H>oJ4!7vyXFeB!fsdx2Ec5s*_u3+wUP!C4FL{xE~OlQ-qnrjqb>z3mlH;gNus3y
ztQ4Mo(8G^GwSg;8&8C^%C_o8-rp5au`(N;fpT*iriNQn-WpI-!@{H2Ra3Ud&5jNhe
zdOP25lTP`%VWJ@>F@cYa1BSlNflk3bxe`Y=b6>@#m1Ng8qN`vhvBy~x0EHHi`;cqw
zHm0(r?8)eV;tZ0JZ5k4lJ5z;|gL%Jw*`nX4@{f7>!!PaRWFGIh?ZriVKOZ`ab3Kqw
z=H+jv=l{6*e4eBw31I{pG75*(9o~T}1Kg{su_DIT@S@739II=oG^cvt_5Ajys&XD`Cp^Hm?GUf5Hm-t$wXp07)_J9OF^xqJ^9nVUXgx!
z$|~HaCl{A(wd4|a8|6rH0!~yUy2grl%{8ylw*bh+9E52lxi=+E_4u@1-lY3~ce!or
z%ZUhoS7t8><|F1>hsjAz@1>Yn#aG!}fzhS|J;~Zx3BY9LHO$pc2B;?8lL7486*xIP
z+3e}wX`AG$cMw6=ag7$>H%7dwrxy1SP&-P-rbNPH}X!&lr=FQvP-+zWB{O-EB
zc6P*n%RS>%ZV9^n%kIwafBxyadkf^;D|=y!4{S)uABI;AHToT
z#PvgXu#wx>?3&q`r287$eLT|Atw(-eQ|Biq4=eQXf?X4T|Lb4BxF7!>x*Pv?_x^QW
zoi~0gUg+YxFN=9RsL8EMcJ2KazxZxXNsxzI`BXJzY31nhXnt(z1|>JSJoGqqoT
zvAx$`kHSx1fA{1N-3P=x64(2NF~x6gf{Gs&`B;y>y!bv5Z+<+LB_G=t!tOmN^?v6RF{M|!Y`0i_RZ??WSyNrN)8r=H~?z!xCbN6)rGXnis
z%XhB30X@rCkGFmO`GnSQ)A*bAzq>yL%VP@PzU&5a^PBzZ-uZaEcid!
z`Nv%&{D}udn3nXn^LBRjxspt2w6`7p@w$Ucf$+Px{?FCC+TW?=ue+Q7gFoKAcUN$H
z@6EeEU%PH7@!i$JqHZ2KR}S>({kvc0@frQ`Euee5$tU0R$(4V?#!nZ2`RX15nR
y1wR?zvt3%Pf4TT04P)r+_jlJz*z@z3wD-E}5%&A#FL<$^l5{Q0$NvX3u=D5&`{cp^

delta 144387
zcmV(_K-9mmzX_)62#|Vz#JyXyow;!(_&&cv5i_qI;g&$=yt&ot4n>T{8p-yn1IUZS
zF^jAwS!%g|{jA(os+)K$)!F;oa=418C=!Vemz8S)iOk>oyEpI5|MatupX~p7yN|#9
z*+0~u{r;QZ{XhTk>5I4j`_F!rKmTuEytmK%(J%kwAO8G5eDkM&fBoLRdh6f(r7ySH
zTmLWnqyKt;`tasY|Hgm#x4Kp1tfRH>>Y=0aveRv7q1_gK?}sKnzW?fj`p>`EpZ<{l
zLMz|=^0mSL<7fZyjhSD)`{Wx#KrH3`&dz?CS`&vQL%4C*Ui=*DrbT6^CjW)N=Sf{ql=zfH+zS_2S
zVbdGqv-jqyRhJAdj<1=toYp)>G_JJmGxnH`hG`$A1a;v)k9v%)PJD-LGO60>>@!=d
zs$(up*=*L=>`dpvrgx93(zz`fYpi|f_+F!zHDj2xeWmSxiRE8ouvn>8V@C1TiY=8P
zv`p@>y(VTUc1-MC!)HmQ(?zFa$@Eg@9Zf5B)?Q;aHyYK4MO~9O%bwfEJsUt*@Tp#F
zw^c1Rvr+B6F}Zc~dj`bJNai||NJMyr0TZIdDP-JcyRhl4jA6lCV4AIvvrNabb620_
zJK{>)QPkUik`>ZndoGx#u5_jsKzXH~V*7Vr?hhaRx9|Vw_gzJ4wO1nQRb=#z#ifCj
zak|E~+TxY2Lf`I>AOG>g@4x!O-+lV{Q+oNsznML(4RBR#7plIe4DI->;bkHh$3XQy
z?wim7pgCO5cIACfw3^%&3@wur5&AC9PXzs~sBEjq+CN1Qy#;FWAaPrZ`
zS-}W@22f@ZId)+g<-VR%CtK15j%oMZC$_l`Oe-tz+Q)qp3Ua3KyBG(V9jm2MQ$Xf9
z1nSvs-`BWFVXS8=$0r7PgFg9ymzPmG*2`Uu%Ov)-OW8|n6{tL0uV8BTHNK9R
zsi;mua|Nr=fC4jW&TF*$8h6s(bT>6OZ6G9n?jYWx2V)}VzMgN>P{JsjV+LXnv+UWb
zlm*}Sl79G`8xQ4)k!JuQ_d0-z!Ucm4XemthK`!V=8x>S+9oqy54z}G$&Q5}OC*+)V19w3hnXt5WHL55E_81UZjY}{!YkF`#0$5w4!)~XH-?g!f1wgoJWzWD(rAjoM4Eq1U`
z5b~7iMCQ%0I3WifVS=Ko23Gk76PS}_fj!2-he56FN319^_4LektP&_j+n`{7Mx3a>
zi_;@aKxt!uz%r~f%rBT8TMxmPY;Z_@fQczxhqelO7no@i1$9a^)>>jdsf%l6BOU#g2p%NqSXfYPsziQM@%>T
z%f?(wKeoWAhS4CxdICc(`WgrPmnH&YZ427PMED%(+8cAUApl-X(<4mSIWa2q8k_?N
z=WmHwX50OU3voC#x-1@HRA%pki
zRd&Uefp!iwPWTLPNjUy{nkLXg5x_hg?pQ@DpCcg$^e>{Enu7C+LW7T@5{KWx4?E
z4WQDStTGng(e#3VJv#~*JH~e?oYSVY);twTw8R}vZ@41s6Wo(m%`+S=1I#l80qy<@
zlK#D>7i4XjCl8co?MqtO2TO&uZ8`O>Uc-&r0^>gwv=&T|N`462+V~J#_cXmhIH2o!
zMD9sT_YJA9HHv$G$h@m*uLEmYFa|q@XjPN4yY-a0SDWLnCN*P0zuL4X?pc`c3q<~V+BGsM;`TA!PuWR?ip(Z
z1qS>b0g-QdCVV-d1HuaI?#y-%s9J}Z1;ql@nve%%EqvR5kv4VAc3;y2^EVa&|7g|&
z@NnfI!m=TfdJjCgf*OFCM-^GV^2`~2c0rsskoI%WGzRF!5ez`y7Yq=@0`jqFb`a49
z8S;~x{`!{cm4EGa`{MnGc}5K=78k@xkLKO##>cJWYZ!ho0>*oQvAx23eAn{C
z3K{{$!484bT$i6RZSg$2w|iQ)L0f??@TlWu$AgEH&uXVdwt8`$-)p-_$9xdV+k+JCb-rB%eZ3(=PaiuJ81##
zGk50#K`lB5bSQx6^(`-kb}D0(T`I>|3&f&PlM2e;t<*bO26K?c)Mf(%fK1<5$XIh=
zM&K}iukab)wv3%ObgYKW(Q?r#<_;U4`dOvIo4med^cT>^)|vx6fD|%197k-iLsjmW
zL430I37$o(11!+r!n5@-K$6^rNZ|5LQb31JtT8)i6Y~9oH8Z;eVZ%EGpNQS()Qg|VM
z-FNuGIG`|Cz;U!yJvGnifYV9`%aQ#Ri1gn4?JwTHi#NZ&=kKPAh6lpRWhmSUcWofm
z+>Dq(UshRqXY&*$xqz9-0Yzi7pPlL&J8uiyoy{`yVA%QCfK7`5Gxoq_R_S~mV`UVEgB`RWRM{sKF@9`+@Vkno
ze_kOCKWhaOTCf@$Wjvuf*gL4%K6Aqk;UesLKFFF_5>iip7JhU>lrOrN9kA<|UypEmCiWsE)^2dH7!YycOdyNS5RY
z-+?o&D}34x%Gm2y+*^Lu3f%vH%(;M^s$g6MC$6C|)FHNBb0hJa{b7IMpZvq?|K=<4
z;75Ny!Qe3UUZ#VbE$g{m``BuK$7{ShV%usKG5D|-e6YsuSk
zJ>Ug>$)x^>ftndCWSkcdV`78~$Q^gOg_88wc<{e8fvpI6H7if6-MiLk1*~(0j
zN0`v!jJ>FWe?bFKDZL$1$FvtL;s=2tU28aD_E2LL)L`U^dIRLwucMs+#Ct+CCRKT;G4Qs34Xs~iJ319a!9Ya!S=dG+
z);%lvGsdoCD4#Kw(X$~H8vCRxnN@#%>P;slcy9XrC@71gAEel*KJy
zr}JHnS3VtWa0XSlFZ9=YdNF2c>0<6`oX(bxH9R#LuXU#4bA>dh9Bn~&H9iYh2y|DF
zp}d@KZRh!aG@7pEjJ57-+|gRTxm~qI&A=o;ZtqRnOpLd?8ee5$U(aSodTo1)wtkf7
zD$x%6xvTMwB?w2fA@6|YZVl>xZD@=Df7E*4usr+#^#_fw4mDRYj`s}gqH5GKf&)?j
z{3q^fTub9Y8{=N#JyjYu(JidLRWOlPoVR_y@xa1=-hE-cD^>?#%<+p>2>S_&KFfWL
zA3M$3mMMN#gR86<8gSY0yvzORJM_2v-S7Q%LC=2<{u}>|38EF3yBQo@W1AwDLb=1xIl7kQLl9vem
ztcMtRxF7kP5d{gvGS&(go^okqlvt?H8H}cX`f!2xy?OVFBb{&l-YXBKfAF`H4r(SN
zN-nFB`2YZEpq^e@$w;18Xs90>Xj5+~jtXV8ts4X=9|Jb;-K5gK+rY4h0*A)f{h9fH
z;#5d0P9WZ+XgPNqn4oPXLxc=IvbK!;J{dr@twr9DKfr)PFEf~Og%1%SC8N1!MpMHF
z)VmEt29?#Z4cafpvc68O@QfMH(dH}O2mFNr00-4ZTyTYwboI89Ml=QrNJj1f+F?>UkH9z_@1WQAeT~R)1K=c
zbOhYoVqtC%K{ejd^kD6hPP)v~eTd}UegAZH7vkm>2YMcIF=zmqI(e<
z_cX2RbST*cCkDc~R*Px*-~q$Y&MV%LeAo1>P_ucUO$~gdLiH$CAuxcGJiQ`+isXAu
zx3i%hq~(*jS^|UM1e$zyk&NuWr|DcE7>KdK=4xP1g7s@z6%K!S-}f{fjcP!DZ!|X7
z=o$S26f;OoEJnJcX^w@axsSQ$fE9=AIpMuRD@)t=Z1ml+@_I(?=jrFb=JcA1xd3&o
z$GoFytUCNFBhAV_88;trMiZ2Oq@0o}=$@uAeLTnvAf@-wagf9Mt1?MBun!2@rjyb_vWzK|p
z_@Sf{{K*zx`Ihs$r<{HubhQu=1S7_l`o)w6>BW!F0XGHinZpAz{)3N#WWw~G@UU@MyS&r<2>te1wCtU
zV4~=K-rc(Z2s12O_>Pu?p&&Y#Q_VZX9_;Y)69huBLHv$BV^A>?2SoLDV4ZIbYKPZc
z_@ykpW1K;L!JoiwN|$n4U;(BlAS~vL%j&OhnMYy6KXtNz0{6Us1o2F?DE7M7Xy3IQ
z4lSULbx||&1{5Kq@oH>twa6WP*1XYNMGhnsln9?yFczS}b-3wn_p}W9gx+@V&>RC4
z!KtVX@IMK9#mUllEk8Qy<-4bf73+8bGfU=xt3DkLQhWgG8nFGImLGkC^xYfCJaL|4
z)EX8_o_zv8CH6djfa%G1AU^~a`W$)Mbu!xB@mtW0G|pPF*FAlHbU4l*zWnm-ztw-M
z|EQnd{PFXSL%*{9^AGRd?L{1yX|*}_{dkL`i&|LSi)?Y}W}@5?{`v7f*H
z^Pl_fr&pUwW(^c+02PpS5cZKF^`fCuIKiKFyBT4bk66@%YX@bp1K6Tlh2=?
z+nZS(OA`J(uOqthoOzN40>lT3GC2re^4uoxwF;$Q*WOZ+`z_+jI8I=Pjjp20<0t;CVV`zGN2Az}c6-FFpjllY>G#&W{P-)et_639!ZXJIVP7JQyTe1%;qRw4be`a5(^bw#HRe)jG
zD3Hlwng3?rsPR+#vH|P7E9b4Xu6fE5gU@N;WYEz2xqUgOm7Ya)O2ENH+mT1VC50>>
zrf2rW(37@D@i1QCsNof8R%lz^uI_k$W?x*n#917Q9hB)f@=DWZ_c3JnzCu;nm4IPC
z^yQ70<>4!VPGBvPa}@Xrfc2=YUTr@MyY18b>m)Ez4j(O^ON{Oby+}@WMRil|X`b!1n2Q-(mckI>sI1pnmIEj0W
z8VIzP1%;s-1QOVAON*L^tSa??Jnl<>@9#c6c0YXE;w_maVYGPnEP`^H=7tADA3Rp%
z8?I}No}Q_wnKJHdIaOk9x1?WWHaspd+K3^9EHdwRg*@E~h0cw7-wUkUE-@-vA|r@O
z!w6^Q&DIsHfiSXVC|&cVyPDl{&wxdGaluvu!Q?UwPP!kx#weVVOB?Wi6dm~5U_MJV
zl-61D@ahtyXB6=)K>=4Z&Ne0rpBdR{b3|E}7&Smy1~g%{YF?WP5U0ulO&$1uJr6Vb
z_U(Ip@j1?|RN_~k-n{+z^}qkt#%_0lvt^#>5#9RsL4az=@gfcTV8wp4y|FWQ7MGU!
zd!cvR0MM)nH^dbqxE^hPZx+Hwc_z$TK8u__Gu8Ig*%TapJldX>z-N>kQi6CNncS55
zj80J-_TWS8ef*Di`k&{}`Hppc=BJ<3{H+Lc-zU>CEjhJ-0A}X%shQQU@P1*DKePv2
zSKCv7US*;SquK;hidAU$!{_xqu#-w5nS-o3thM83dTpy2{GZ$PJqX%8
zvovK&=f%A<79op&)qJTP-Y?n<|NMs!5NW@8|NiY~+y5Fq0mT3Q(;BB+UGPk2<>=Zer)5{_|5D_CUmC6e>;P0>}1T~wL
z$)c~kI|Hr7_qJ2B{MLB<`jqIB1>_dzxf-15sbGdOJNQ{S^ziS?AZT+RG5gHNW|E1B
zV~Dk=lV5m$2==$z`$~WL$Nl!J7uous>A(N^PksE}k3XUD+e=a46GJc)mcg{jdK%o|
zsdQR}cevVqmc)d3UlV?(7avDzCCZ|S9XoPdZl7S$L9aPEhy}&oUecpR#^`--!)$N0yigv0^MIv0RJYT*!v4SRw{L9!)ri%f
zB5uFmXll}Y-FI}Cb$h_kG9Lz{hbNP7J;2t|i}c7i*(??msTUWmBAez_cG)V_
zB0g2Tac1cPvEEXsjUHLh>MFa3w0F?fbL9^LyTKt;!{`;>!q9xRVv+g)cJO2u?Fr
z%lGM$8ECq#0N+^@%th%cvkHz6@!NVMpOZl?vj87dm{497|DDX{@c=~0*;yj8pR9+F
z#S5|!twW;i_%LT}3|vD2xoGqZT-cA0FxFETrtqj{Un*AmSqH{5Hdd=(<#|9X|#Q*CD@oRdX^
zvtE3b2do(Qy=kATzIOSbfZ!)Ga1Di@2iw77J9)7;M!U%DZWD5GWRU5sz5{wv$1MSW
zR+j1+`gbzB+um6!wYA6_q;{hUdlpQIo&#m^_v#mjL}Wph%zB$B;bpb6#;yFPw;6x0
ze)$cHMg|QuxL&>0TRM|jww5DTAvktK9#&<1e}gy&HjzL@XDX=HbY6u>+q?Io290MO
zYJCcD0d&r+*1X}k$ZS3@#V%E7XUq40EFo4h7&&V}qOPIm(4Qvks~J6&X~I7DSbH9HOr^(miI>m{$jOVGQV)&R-dvrGh1KYAM6dP
zs?u8)FV1%Z8kFwfjU&+AEA49Rfe
zw(?4G3gw++Ebv@fy*#fc6B4(wWwIu0#;BQ6*~fwV;M}w4c|BS1SiXB06qnhlSsT8!
zF(}{VU-I_prJih^CF?>IA<($y3iam%r4#k&kovry=#<5!nC%0PT7P7QPY_xen^kC4xCPTL7%
zPUjk{3ow*%RXm}I=YT>4{o^R26C~rMC)M{_|5x
zq_nX|fZuunhbj@6W6ZKcknlxTIb&7~dN(fxizi(PkGwF07c3W{Ka0Y-4A+tmk3=cB
zmfhR!>0z9!<~0xhAglr$#xn|JmP#zcfmHEQ<>b4q3N@?wEH!gqGv~`=tmqM|@aW7J
zc{*8tDzFC~f7XGIo76L@95fuw_Tt&v-BzW8F<4nxwHrGZ_)r2%3H)1i##JMn?5ySC
zopeeSrs8n(19ajfDqmwYD?((T-WFz%Qp{|db+!@iw)>!~te%#gWt
zQ&8oLIAR`)8DxvC<1S-xOqD)iSuHLKk6EwMKN_zh@mq+Ng#`=SwsrWz2V!M7^!oiN@r
zCM~ZrfQ|Jcs~BKfdr``}kPXUIF~_#*1Vimtjj(u)7Qi9d06+}%FxD);T;UexMV?N7
zqZ+`8_Ra*+V8CV6tb?~^-u>}KIBaWgY0z`fSe8{6vletWY#p0`#fzA!X#-*Uzuuody!q2Vzx;ami`yUg
z=9j;{!CRh9uv$<~2bw3-9Qg_cKL&_@XCCQYVsnM%TlsX++-#}I*_xYVU^l){=k8vc
zd)?}hM?N*Dnt9Ef@sFP@mD8^IPRl~45`JpW5OH8lX6>h&jN&wu5)AvtnF=o#98uZ
zAGyS4x{8rSilC0psGOyEA((1iOPil-Y=%hnb&GK7s=BihZ6<@udx%Mw*pxa}&3Cb$
z+2?pc>>sXhppgF9M%UPMXunZ3%TuqC1>XQb^JsdGUatAhllre_Qhz%{9t`2)V?sU@
zT`!cwcz{jXG;?c=0w~Drq$A6Jx+a^he2Gnm_{^C6zA}-61f;LxKQn%Ba~YS|jJW{?
zJs-(T;aiY)wA*#!Y}CpXHd}!R+W@|LGM&#jeHf?+-753>xnvEs>K0{VGHtr-M!g0@
zahI&crg{n9f;hroyMXV=9?A4p(pczGB51;TjZI)|*$wgtq>@W;@e)gaPTmAATbJ0J
ztGc)_`%A<)7e5XTjcR(%8S*TuPGxG`dBW=t~^MS@S
z>*nAqTic@j5}ShUSdO>uB{Qq1@FEgDbimrjHTx4*7nNu#NB3g0&W$z8sF&7gUSo5P
zwX*SYbc3JurHIULIqA}XK=8*cWIyrCsXUfm(@%B%IGa`L^D@omQm<+I_-g0O}&KN(4d{_L0R
zl|jM_%&-Nz~mPJ?=o)soJ>RJv-Q%)
zp?MA7jsPFf!w|Zeg>Xey1jeKSZ8;sw6*jX3sc+IfG_QGofrB7Oa%>AN8hpuj(&(0e
zyDZZoN@N->+mp&N2MUgBY_b6F+aVeXk;R$eHfc&|--
zlTTV>JK+6)G&>FCsN?ssZ_hQlj4+#4GPzd3vm8AE!C(w=1_0K63HA|K<^;GNo$#8a
zMXfcnJ$LA{qI!wV2@9E7Z^DNNWtFz9z#zoz!%QyWx3hx;SqAvz9m9ZD$tP2M@@5bV
z_YxjI$7qw)AavKr3IZ0#bH8Lh4W5^rVNA*@>OvuZS<@Ap1Zsj>s@?^c-RK&dd6x@Z
zhvYO8@S)ZKz@{Z1q}w$%*OGvd*~jfLvll=DxKO~6`u0mOeuBbhM!_CuH!kDU#r3c&
z_yVT%n(sWNo@o{!*1#U{5pLE(A%jKAx`!Z0qH9%5X�_7qtV0CabEhe$97UUo&|-<@tD)nCNOt
zPFCXUP=lA)^tLj_Pj*jcL{KeRbR(vptfqF&x`o@U`Ho&&oXp(T;HlXO{{d!n%^3#p
zd;k*as_<&3RoVg)2)!HW&GizWy{en6$aklI)Uu>Dh)`Fz;#vEsUa~(|Ynjbo-C04f
zK@qG$yR4&%E}Uz=vxAxNUQU?EV}OBx^q`YN-WEFv{rpVFZWPDXYU}#@=f<7`EmqrFHML+TK`2#N^en
zlKd3)YivS)3BWF6bTW4Y9Lnx>R;gZ2T!MXoafscK)uwj<--E^^R);n_K;TPwJoo8v
z&1-8uvE?1DZU|r|1`b|kQ?lk~o@GdXlPuVTPhA-938vkzxkG1QY|UiWl*gKw!roTz
zfIyJm4DY$brWE)wOeH1UVW&b%v|ApUULo1eMK*)$97kp*7?~M^3Md#S;8?MZF0pxp
zva*^7;DI@BAh%|K85yYkMSCR^V>T-t!g^#@$*4}5Gn^;-Lw1eLy?HF(D>J%(liuP6
zXP(L1_O6$l4{=8Jf)NQ&K??X^1?tub7grewb;(|VCNVvuY-l1r(7R_0-d?J=to?ln
zH<31Xo-I019iyA+yxH>sTxQho+pe)G;Fmk2Fy^fIu&$AZMWt9jS}r-0l5Q(Ye1Q-R
ztyslF=AfQfXfxI|HmUk36|#PR``*npMypLZGv7hOT(Uo9p1^j2$F3PONi#FQ;lCZ~
z1uEj2bqie^1#AI8#PFFhJhBSJcC6ZQ%~_2AZ&ctnZ?#By_LwkVZ+Ec0j(Lqu>8)gL
zoD|zo`5?O!uv;;{StsEVn+M-?us26Ok@Uh|1Yf)fo}5d+X5AiYnKbo()Ugj-DV35u
zh(OMy*NbrIaAM9GUhH62Gwr<=3olsGyssHsCB|Ih)3{P*uB&Y>D{!U}(&CzR>osGu
zuy$dj`wCqGiaWiX%))eu7Tk4;=S{jSSFg&q7&|Qj?_yx}7unpy51<`*H;4Pmcr6Iv
zfEZlXud#{$VP(a`5xrS|pP1FL3)%6E0l8$a?2*9;klI!AQFZGfqnxr^$#
zR>rG0_sqzXhx7Rg)enZ*uHk(EQw)3UeB|ILWw{Pjo%Yr(?PhECDJs~Nd=Ug+Nr)gF+G3pb4Qgn4vOsj;(0YB;2
z%qQTyQIww9GSgM-R)ybgG@K#4ud&G)AFkryT{Fow5C}6#a{w*;k~0hl1U@~lSm?|#
zHPZ*Be%Kj3RIcH-rzRFP%bE$3@n4#+iQ!Gz#^P)4=niBFzGr3;@DV0Q*1^8q15}~$
zHG1XLL$`YXaIZao6??@2v^pU6ivl7pvDrq3g=&b6)2s2CUJurBQtNe1j7z>Vt;vD1
zmbn9F7X$OBXWAF7*StqM^T3f+HqA__`NX5Prl-19=6Ox*#{my1>ntJFbHsH(F?H*e
zLHveX!V``}U7$pn1{FSfK&Q2s(cEBH*EMS>PZLZfw}|F{buwfcmhH}Gm&|?%?>ol_
z6yW3rY75x2dE4e)KFr-K(6{M2wP)(ri7~jurq2<1kDdi5R*U@9
z0Lb&o8})t7{){2wSg&C76)TAt3rP!)gTcOtn`qIEu`Klzj>$T#NakHh%~HoT?>e+q
zf(~RgZRKcxzIm`sK@Pj{GV2fi{Jb1ij-NI|BAA`2uks+~QUK@ig#+C{GuKCWfQXoyL7*hh_1h=g@qkL$oUo)Sp
zGRRg+vaA!9(KR%|!@xjoRHBP)f{)OgQK0X063IA!9}D4ACHNAISEhW{0eM?e2^Iyj
z9>CuP>bKf8F)k{ljl(`X4^6@*_pFh-=cK&+y@V&|&UyoYTL&K*eS-_l4vjr2Q$((7
z=2Hv2ha7@~fn&2eO2^`rEcp_4t{Gd+*o!vvlsHpW1B?JdSyre}$1=Largs6HM}Zu~
zcY#iSH<&yeo!II%XQ|Nj3T0A>4?ec3EzP>9nIMi+z+7Te^PPh_R;-^}$+HZz^Ke^{
zUhWc`KEb9_sOTYP$|xo+g>7cK=X%Lr@nRk_qvul{m6{o*E65S$jk8>HkL2(t;9EVT
zFLov~q?VWJqEOxi9QzWR06qwnwK2ve>o6C8=(1#;XN}SM5}oaH8gE_$pX(qE`S27n
z4J;QsMy`p&aOfk>?5dDPRL@GlQyRxI7FfdBWbYJr}
zC>FJftq6c;0WE4gc7?_sd4XSBy~d`sq&_;mP-gpB&drzsOz@-E+#>~NI_^9UGKLg9
z4Qe7wsPi(1SGxpX}Ed?;arbtFLRm(;;v3EiX99n4^zfv<_@o=%W?m5`G)eF^^-$
ztSRpWLUe-bat}Bhy5@}{v>3acnyDyC-j&IK3_eH4#<;hOzVqx1!Emf5sE5pdDsL-g
zFv#^8^?`fkUtM3SJ$^@f`NZ77G&%?lgx0}4W!bk5-s*LBpsZeCa+HiDtImvlo*^=u
zGrg7!%V_&YX21E?SX}hYUUar
znZ+b?8^46Ck20Pf8TzufAke!!t))2F!j_3yL8h-YF5bYNTf4oo+NaWgTBmg=hDBqn
zxYYQf%q0w#X?gv4;KZ=R>rgA~-Cwwne3WsX8W`z3H(AjOgbY?XBwD7lIY8@cjSFGY
zXPpBqceCd^9!Lqwc$?*N?N*biS-A#FY;1w@4$wq}V;w-Zy6oDmmM=guWhy*lCfoz*
zhKtX<_nE|dI9+NSvT+W7Ggr`4cQ+#Om9>1hrPf#>!
zT`I*NJ#-J;7lPK1AVyG_~{)4F3^G1fObHco$-=_o^Re1hVtHY@_RIZt$|h8s7-W%v#cJC->lt#F`gr5}?VJC9wLiUi|Lz*1^uy+F
zIZ@eL4WKtI2w1^Y5dT@*<>;Y=4>)SZat_)VA=WyTOztcT_*Bj_Rw8eXvnFeqZNM4rwbZGB?NtU7d%_r;vS#G7
zsaCltI+Ixj1&SJuVfqU(Dv#~K;WOIIq!uG06B$+Vjz!t}>YJai7XXFuKT$d~Ro16#
z71|Okp;vEz!TJO|(BmaC&#f(3`fER-2LqqnY6B`)J`8xkuH%af1nJw4?$3h
zjWEyS*gOiC3>C5~O&?I@-FvxJjG!gpjSgvlo#G+0)?a9v3oXsJ9IfpoaESv`ZkcVa
z4e9GjAcZ?VHH_zKf`#tKWA42A*%^rTLh;kbm=?-f-A-VHE2=MG`P4#NNAXijxzMz2
z(>>sbyZ2iSY8&-T{5!aSxL-Nf7I2V|=kOwYOO!!z{TRH3iO&h?vjzyE)~@~=KWlwW;%^Y-J<
z|407sKc%f3BByHFP*RXmT4`kR+g=Mqmd*J>gw*TJG3ZsMTgzuCKZ#s4wq8Ru?!}I;-_DWRPWl3^P!5
zPwddea(xF*$9keJk?gn_$jiJ{&j>8A&GY&WWS9?jY-Y7;EA#%kfUJT$)O-YXF7E)?
z5`aTN(;enr7;;tz{)Ls%$Jv*6U~|@j=FHt+1csY;qFETzKfBT6*1&K7+kbU^Z-AA@
zZtowpn?<3r9#ezBi*(vuPk7dUM)t@wz`M>1O-u2?fX9-V$uiR(7(%8m?8;!^)~_^e
znM%56+OUkjg6)$rZHF2}Cus47ruVUYS)!T`3r~RlhGJvEFg5FSU1<8~$e7B#4Z{l>
z5pB(4uAP#cReGT5M=#rN{m{;1pdg`Ry9sGK+7|vHd`_Hlg~`*pEcf1jvN||4%f$5c
zo;h%_T=%@dWQQW+TD#ns(&2SV2SaOPWSP3hG57DA-0*FgCjyE^XPvu3{nrV}6dAxl
z7nmHkhu~S=9^UjUgE}eTos^eJwYGMFNpJ^@q-`2O40o>44iC%6%fOx^*O(k(S$#be
z#IR+S%~rbuntlJ()!6a8}S@&!Q-uuGl0#vqD+fYgh7Vb5wB>`HJ_{AdJ%nmHm|FEAOHJwSG8nNsAa)=VN*!bc$>IG0OI
zuE{i4%@YIQkmZhPDVZ?6YLDYQ|NM(ze)#bI!!Q51-+r}UidO%BbHo1UKlSl@KmJq;
z2?3UQG(q1&fIb^eFq6}?mO*Ry#Pgg^L@AW#%vr^P>%u|;=xGNw_-JBJ>4fw=>$SDq
z^7453+@+s7-8tIu@sv*NW(KWn4u8OHWmTHo%A51S3k`otCw%oHezv&66*v~TXSvi$
zt>}^{Pw7Np_GwFhUOZtbzzI)`!Khf0GP7Xf6`go2XZohSbnSf0s0HE1%|P+B8P^D`
zEHIn%xQM0inH-{$k*5;YrV*8tN$|E{zPuM*VUnOhQy|t%lntAe*npO04}5g(%4F+0iU6s0fp(BOa*bAhg1`56E-<;zgg7Xk@$-EN
z^MvQk%+dvfbg5rpQZpmlT3IgL9Wu=uv)wAO&dg@57nn4B5oLiqxbx_fx+hE9o3if3
z10SV)&!l%>WgTFDCL3_e0w22(v?w#RLsyvexiF9FCo|}TfRc(BwrkiqdL9-g^NX#2
z@PB&ezrpW+{_5SY{fqY>{^LJw+xDmZ<%jowd}BZQkB_gg^!1I;5QF!O4z?B!^^`eL
zhV?vzUE{f(0q#vm8_)FQ^GtwoJu_4N=cQD8ZfC%4cGO^MlCKtgh&;73
zPMzT0u?_ahlNe^sU@1u~Z>!LSo%!wi|M`nAzxw)rHRxj=5dKB8dAtS@J(i6Q*$GWj
z%B)gmtKhlDQ}sg2Yb~-&J9L2Fgk63skD0}0K(4f1XSvt4vpkzb!g%_9ZM7UR5bS#C
zH$z(!Wv%CBl@2It>1~{h3Imy)kIIhtIm=lisX*1kLoujxp|bJcx2uV{q}z3t4-~M3
z6g#kg1W%b6M$6{e+w`zUWpVzTWs|H`R5MpKX0cZ)&Ez2+tBa6H7Ux}i8NDo%l7UF6h
zdT)*rC)(OTlqp$TiTm|ENE#@BP&s{tWz>kutTC-^wWd#)1@O@6ILQ90o5WiE6GFsF9W`H5ETnbbmi-N5N&f-?Y0gs>6{k<(|gviUhpdI0zAln
z|9lf~^cK{6^`YEch>lMZf~Hi3V3#{d9iDrX+CIV
zaw`Lmz=C~zZ{ENC>Wg>3`|67ek>W2*BvD{&I^GZ)nh#Q^Yh9TitsiBh5>M&{+~QF$
zi!-rA&UYPyuocIEw7lZj`g6S)nReN-D_Kcp*W7!VG^s^DeRO(GFP5CFSjbs_O`)MD
z8D^U)xz5g{z9vuVg|3pcxp@$;b3k9`IGc)kE7RI@dSRVnl~(c=orwpDdSph*e5A;K
zCDh_|y?6&a@OSvn8D#*NfzGNG`+_kloseDx;0f`O*l)
z`%>GPSpAU0K||#8;xHACJPVJkgF3FX?UcpZGHL!1S;N<$y;jB2KkKx4*?cFZf49Q3
ztW1@tBV&uU?G}^tSiav~7n-g(6w0{V}C9*4FxTk(A8afHEXrq
zD9U!FBD~W4tBoz^-dGF+#qAB#K;$$g_S2J
zHq#SYEtV;RpVf;ac|ZgakSXTQJdk6c#(BM0MD}xfG2qF_A}d7Jxe8d=RPZs$H&=yT
z2#$JqFDN`yLT-p30MDjO^JnGQa%@6c3mIgK80I|tsi^qKYAyFY0xj@!z7vYKuHzXBavSE#mPeT)|pKR&z{S;6H@
zjDd_QI@K0euk~!1G$-40dg0g`V}x^GrX}yr$|LN0rFDoACeSX;Rjpb-G
z2BWjQyeXJbhc<^K1_^-5Xjj_KC-eD~fp-(lRI77A^!5NCcZCkU(srwnbT(?~ARS_>
z`L34GUzYQ`}*ma^3w*O
zCTw-x>hse5?U8ihf9zXzsM79%=iJ56^@BgGO(`F|j0x;iLoihh38y
zYR<_*no`@k%5LimyOp|Y=^!3?%df5`;K%5Z#$|RVVYL$PaX3Q>(ic_6$2tUX4Z6%O
zi}amOM^}c5VA&`~LY3n`K?g6h8xi@~?Zz`%TlA5M+43Puf0nrPtL%;}=hQimH|BuM
zn7;a~yv7Pz4@`3LW4n|m8U@QYGQme|i`%PVzm-&3GB2}hV=hRWmNZ&h)~VK%IoMO<
zNV{x)XA_&d%kedI!RA~C!Z#D|j<9(yv&*2!%vnBgPf23cvpSiqTTP_XuCmMPbZZN$
zK92C}5V9#Cf3cTZ6_?Atmv`|oOU$WxjiU{ciMivqwi2()aD#QvBW?{|ZcK@iPu)Ej
zh^bWOW%w)yL!Bg6u#gKxGiw%ns0Yqr^}4UJyXP7*I0|8WS4Mw_c%!WFva&M6Rd&IZ
z1=Ix|pd9Doz3*O}c>46w_GRlC!|XHbP-f9uJ7*?K;h&eUdN8E6J3
zcQS?O%;fl4hx9Uh&KN$$^6eR&xx8h7e_nG)QONyy*}76I6^a@gDe`fo9`I%;LvMYH
z6~1ho;c01CAk>VkkG>1Zwvve<8=%AFD!W)zVw+CZ2=8TN&W7cuBy2;Kj?3mZDwi#p
z@7?2Ue*!^nKxnw5u0e4fuZv!59dNf8Unv>=KoeSMPw&1#&mU>`pT6`De>J}2m2Do?;(Bg*4S~7XBG5D75@w|cXzN90Aw5>srR-AR
z;})iFV7XWqyM>d?i_BV&S}30@Psdt4Fbt!4e|P{5V(XzI?stE96Q3Tg;=XJ3_P#Ls
zCc^d$$f(7mMw6r13``i-dYNHp3F4AfOH7l3x8o4x?;5&2dvvb7a;4%$Fj{?dlaa9nPDFX4yz3K>Ao}FyTsv4!qgPj
ze;#^?`A7fLFW!G~-n@H!eTmPvep9Mm37cNI4()Csl=Ty98>eIl*X14P4xOrVtSn5l
z2Hi4;X-s!OlzLp=f#K~I`V)@NHE*gJ?M`rK*ij*Vc?Za{tjAggzSK|4!tC3!Jm)#H
z>0?FvXFIUcY2#xBZlQC{46rT@YrJFwfK*Z@*Uff=KZTXBKG0vEOFxV!_h
z3OG#1IFLd7xYbhrf9Bq`YqQ%p4*f|7K;zaQ9iPO`2ghDVUipqjH*6*(hif==qVwxh
z_0Sea6kjp&Y`nCt)QIb;eF3_wu)C}KYNsQZPaXFc;Tb|J7LwLhbI;pL4uQ_ke=+Iz
z@!SP`_~`%j)6-1qvmfOrs9dy;Ee9zI?0+C3I;*>Ad{a`hYq)b!$Emf-dE#k-M(rLm
zK!S(Ybq|*+ePm6if-crMks=3VZLTYFjeEEr=fqxpLt^A;YY*BOgOaGE2YZ@-hFiRQ
zi|i0NOM0&$%A>cuThKLgU3`Zg2vdJ)e}W;l3J#OIsr{zx-&&
zYIZACcAs0%HQWP=ak8^2GcrQaz+WV-Ye{R_^DW%L&&D;Ux@#+EDgI<&50Mr4s862C
z{2bSMY&r`cBo62vu|PEMf9<0&!QgX+
z`Z=IFe6!#ji-S6kS-ZE7#2`{e1!
z_rLq^
z)Jl(mP`AOe94Fx_sZm
zK!Bwniz8OmRA^+hf2&87W*~0qCiaV@RSvz|r{o!>m~<<~b7S_;T*cm--+py)-rH$0
zQ>T;aNL@VVBsa0mXnqi2{axEeyyWNqtQ?{~&vv_s?KXpNHb*OAYY#R
z^6?WU7-4fBf9;4LYiFKZj6)UM4yf{e1rw^&CX(nuaI0HZKoF;Dj#dCEB>!ty!?T%QFP+xCga&UAlgMPvw!K>zTHufk~jYIrDSR(M<4}bOb*B|zq
z2d*z)s2|6my#Dhq_&xH@lP!A;PJ!JV&tYnPZg5eAf1a4<0_2)GPP3UHZ7Y;@t5_Kz
zB!_1i*#8$GcS6Q>q`rPe?n^;HARKlZvr3gxEkZIWGKlw^SdwE_HLAgc-QMUoDrQ4}fbnZ*2agvI^<~a6~pr7l8
zKH>Mhf6`}}cTn+;*n*Kk&67h8^d^au!iQ>^+sh@?n%i1|Pz?d59EB?se1(d7zvWZ@
z<)}~|c(1zW%&fpadA5Yyi#tC@Prro9QuWkc!x@(h;rF>y)+#AM@V;@9_Qj|LFBT~Y
z{Ga0}N~W>L8oSrnpyL*oP_b!{7e*_&&?Z))fAqk_Q>sPUsOJu<^yrE=IOl@K1j(K@
zHvK5@#O<_8?^S33pQEOY`n<4J;Y?G!MK}9}pMjwdOmh
zV2ol%2Zuu7eJZ#P$lYMG9P;Ryw@?{&9VGemRuB%Imex9b&7RU5I_k?&rIl!Em9q%z
ze_=mf$8+ZBQS9VfaTp6GRfzS3=sRb#wFhfVu${7Zoq7qCm2(0uk^m=7_{nOh4fbH*
z8MiyAec%pLx1P;>&qlqhsdl;|o#%{Ke~!AkW%|9JF|d!^TPEF0S!=2mmv#af+DHA!52lJOuWe2N=`>87?e*!PDD7uQftZ^wKPOCCW_>g7}w(DYzNIDB4
zO6c!@2UVpV+RbB=L;`0yfzWRmQMtF(`z_R3X}eC>UK_r08txWD(|Db&`Xy94Ekz-J
zXS;PW>8ou#b61n%m)5Wj-I-2LbLH<_P+kjXZWnyR=`I7lU4t6?myD%>((AGGZ`f^<3U>p=zbM
zkjto?;Y~wqt36^Z0b{!q{SlDPR(4(oKcVq^kIO9d`>32ReF^m-pVQRDIYVK}0et}x
zN0N(axnFv(;AN0;$=Kws0Umsge?ej8ETz{b^%CmTGpq6l9FI2n0(UX($6FO>%uDap
zcP$_=hw~Ugpn8U|Dv~DhI<28EDajDaz8RA6dIeMh0w8tdfPqvm>C$RK1_wn7Y}YnN
z8fTnW*9aE$5}xf`LXCm&;Ph+Vg6!kAYTw|;9HFD_5;HvsX*GPMzOCUde{rX6^n+L3
zTCy>I8`PEPIkz2uM%Qg*fABhRo`%T9VHv;C!(e4}iL#@PuBjyC%f<;6;nqE+)qbPP
z+U%?f1tUZpD4YPaI7~(g2)3m*A4mH{P6pGo<}quJSq4IvCr}989D6P
zSp}*GFLH41RnEJge+!7XwZ3eo-t&*EcJ!T-G=Wr4apS2DPC_XKf)xZDq1%tqEe5*v
zlzk)0<%ys54!p5IM3FMQ6v#SIVFdLDwcJmLj&!(KjJ?_{wdHBfkKe4O!!Eh-R10#(
zl>VZ;c#;C444Q={d&(KBT!?a|((bSKw8)&Y%qr~Hc2Ql9f82Znah+b+o1WUw3LW)Qy@81E4U!B_69=OIl`4@z!0`slz%L+L
zHK-}AP@3S#-~f1uSTzuN(dWFtIw0S+M}eqakxcQQNRuJytYN3wh5aZF#j)#5#x|+V
z7_e(EVx?4l%cAZJ5k1aCaI2#z$0cXZ)269g-WJoIe@Cp@GU6(97#{`v=rZm<-etRZsA`m(J~e-w4l$rlit9_`y8=t7l^%{gTQpB#>L
z+b%@8-h=1-l9jb$ZJ5qTQ5SPatQ)5*qYMq5oE96b7lq&$XL=U8zom5HJwmWf))MQc
zQxt7Rb{8xeQ4j!-y9WM%BgOuGY*
ze|8~v3~W6565{pwz^k*DA#g8iCM>9Zfgumgky9FmW
z;2m@G^Z5-#=?VNXKowW9!!gQzb5?|Je;%1%hyb%?=K5AP=mywgLpbQ26MYr($F3lD
z>_@}vn|;?)?LhDhem(Bs9C_bBG?qeXon;Bjt6A>Xr{$^+rFCIH&b1dmif;!609~AC
z2e=Q_&BT0p3G$djq%wat%zk=j>;hY>Vvb9mtu*C5I2sAX?n5|;j
zX3y&eq7IHS5iXOf76pa4?iJVXc>ZykA<@YPjZ)$D>4aPX>P<-8kDW=
zP*$E*Q|}-kAUa^Gy{>)ZtM1|e*-{Zo=PuQJHnr2HcHA}%BA!Py(C~ttu>diUGx+&Y
zq32|oL!OTth^r$PL;dqSZXuQ)B*`5#WrG;$HxOf|dRmbrpyf?Mds490f8wm9yAX4*
z+4n|31FK=Td`p>TBBL?t(B9($B0p|!h*p)}N`=(OgPq&-@kZpk5Oc5rhU_{g0uMhS
zk=)^3_6g|&RhMocmect?E+O23jUUh*5apUVD!#hjKm@T`2s*R+bUa0~-2hp``a7rkv8>eKOHDHi*
zK_ywp)9JGP8rAN
zT`XTgWDB6TP9LG1q>tVPBA!5374-{|oVM_moRP8zRxPJ8T|n%j1uqa{pdfXed_Xod
zL%=rd&l}&Z_(>NmXSS6D%XC|PEtylWl8`5FJk_&TYv4Eu0q=k=I^X0xk`D(`UqA$v
z>DmT>mM)NEPwPN7e;!au;P!e0k%b*;f!<_}f-Gs)td`rf;}#ckN7N#9WvOY8Iy7?}
zEd9NFWEwbLh^riE45~QT8_@YxWp3n-Q(t%7+XdDU8}BVxd1zzu^LS;Y6(?v2i)U`U
zN5?kb?9<6uwB<}5T8l=ZCXl%CR2|xCr?ZhkVkRW3j_ruJf9D1jin_Vi{xMT%uEA
zy}>$;SSxhSLePl@zM2Bbr1Zhs(vTa7b80M3c3Az@X^f=Rs0G?Y9yu4zI*xJFWumD_
zZUcA6&#o*Ee}+uPHoc&94uuu?+%Cr1kF!%TS3lDS$ib5Z@$t=vkH7!h^)3GBA^(fd
zfAUps%mS+L;%>8o|5i>_P)W@s|Jmd4FF|b}P&Fzx_-V$ez-d&PMDn^%7PokxQP~CbMlpusqlkHg=C|xLPwk2L~%~_)Uyw#ckz-9q-ofJ!eoh#O+?7G(ZfU
z6O%no=BVPx7jYd5qk>c5^$KWBO-+L>C6+_`gWT=sxB)R!1~LUCTyRv)K+@t=b=yeJ
z;}&jCe+bstM!H!Uyz#s_nG69o9dPCczmxwQmv;y{b?T|}U^PP*th1sZQz9^4+ecXp
zl9kgjeHz3Shm!}ihs>gR*tK;gq=WMRoo>%WjCL#714C<{gw&FJe_{-AkAg)
zuyu~g6MKl-2l?I4aiQ`Ck9i9}y6)OJHFH$ueaSExYU|aF4u(Pj#g_NYaQZ$wxtrd>wCCtOMwnZhXWB1
zozF5(30?RnG~=~(RTszb(OQQ-FO8p#f69O!#~Vujzqy1WO?U4kmy-*e*q})
zXi#d&4S#xVJr6o%^lnhrJD`#lNmX&&E^D&ewf#FrsR8LbJLe2co2>hSdf=zu?)zH)
z9a>@;-Xu8ATv$o3V~h!EH`o_`uHi1cM!d=hq=KOf7lJ2{mGi1&?`wL2^Ufm=KJS-{
z$2M@fqP^me5TdU==b32wz~g2lf1sfX(ExS>=Cj9`dtckXGfP)w;la7siQeI=tkOj6
zgoKG}ynSkIskFUXKfoXSoYAyFFf1(aevP+J5a#ggeb$4etiW7+w+PO@k6()$c$|A5
z`cEd7ro@COt<`!GnX9&I`A8Gz?6C!Sota$S0j8;Bjx-zX+Bp*VGE!~be~~mn9u##5
z(AILXKD7C@xS_R*#ele{g~djw91+Fn-Ja+ke+{=S=g0BHnyOk%INqM0#-yC9ZmiO
zclXJj)QWtMH+;5He}ihB>L8idaO-NDvuyyPW+yGH*i%qwU?6ty{OZi{%DSMG8N6oB
zV*F@qSy|mI*S@D5f0{Od(PU))>=tE+H350=R*?_y-S5ke3o5YB?Q(Kzx>DNEhGanV
zxj6asG=bG+#W{UY>~RAU`3Z!xMM5NP#|CEuHi%}U0b#!JE+wj+m^M^uDIm9QTTb*0qMPP}
zZK9v~>VsnSe`C@ltOo?Sh&N?F${5(Hb_245ZYcs!3d{V8)#z~4*{Zs6t2dTU9W2mv
zupXo;9rWVSSUkgHPt&|{mu=^Xm|80<-Vy{ujLD(Xd6S0dx&S$FTOrinReH`A89W4P
zaUElXxUqbuAgIRPBOrr!-cgGRnDrcZ!s|g(Ve%}q1Dlbb=*r>Ia1Q2T~QJ@<`w2_9lw5n;1~pRQIdxv(ey)!sO6L>~lJso;ro
zLSR;~RmmleL;J1H@o+aF6)`=EI?cx-c@j3t&3fA4Cqcgexl_`OrCHe@i*(6Dp(6Nu
zQxe+cUV&V%79XfdEFsxbW*t!+ECGo--*}gQTf^&cI^-<&ruX3Mz*mRidjer-(TD#Dzl6|{f2;w2R}1>d-T
z%j0Yz-zy7MbE*yJez8nSNT
zIz($?^Bw2l3J6F^722!sX>99!>VCg}{q{etPy3($+lP(3}aawIJ-uzKrg8UsydoqEjA7-DN745-7W#D0k73A!*-W9uPujY-xFR8=f`@
z{9FB%&!JCJ5A(`%*{W=~Lgt1b^0D?vm(d$1oqNvo5zD{$&KC%G6p2RxJx$-I3fJ+kk;?e7T&QXvoXmOCC&GF||$2DMP3QJXr
zZ>WIntn|xO*g*~f5a}K;D`G_oQUp?R0wG_Oo@3JlxnsmN;95(A%s9P&?v;-5#6HU$
zamo;ET)hZ9Tc7cotJgbNT^ofBLNEXKr1w@eU7ih0Iu$t!Tj&&ak)*Lo&u10M7w
z7?u5Qja4O0=-J!)sGSDo*MM7e(B>n^!BOYrnJgHH8uuVwO92wIF8aoEGlB@Ka6}RlT>5k?ryONW>6-q?zJaX&${NM;1%c
zdJ))%vhA#ZPjiF|ecjJKb0m5cZ`+A%ECS&85BXIdf@=!HV0zZvWs_o=+n8&$t?7cQi^vU
zkw8hHR^;B|Vu6%@6d+?GN%bWIlDQ3;NL4E?vP|Qs5bW9BmSu%X1+#`aqfgn`=iGb8
z*0=QW8cBfYm20wEu&ZQEZmxVUJ8kjGfe7IAV7P0QNYUxaCZm8Ptb4#}pty)zX$_sJ
zrOEbH=+inB+3F(j?z1)U>CGwf34x@Y15&cEk?ZL9wo~+fs#MxDiX;vIXU{urehOZ5
zxwyDuR5K1sgziR8LPjnXRFA{HAmiQ`%K8NG=`9xoN;!4cfmkI{?YJUO48e|m6!;%L
zeE9l#)%}NGeGR{qwhgN^b7Zp8@{+cNI`^4x8qhxiJag|mL4TFaJ=0n6olg@Z{KGl#
zbKo+L`7*qJJ_xaKc5k6)S}JTTu+y&rBNEW&E{|liyyU0h%+935QC!p0HQ-uPi&Z$W
zK}+A(K`p%Gta&c!TJ8ZuazYhtEUL3JD8_u{{n2R}D`VdSW`2Q4a4`W|uh6|wJ2a_U
z_a^s%2efZa5yq*2_)B5ALln;(n;GpM@X1q))4?ErWOW`bwO2zwLbRT7?;ZDD-4Xl7
zQD#~b1-(jil?SErHLd{v*$*Y}@3eDk|cYyZ!G{J8cnpS=E9MX-PNxBi8{{A!0qehb8*VFj_d
zNIt1)%7;fbz1{_1BvRo_L%mA{Zs|OnY?CHrV_pS6Wly%?fZh0}IW0D=74a;s4oR1P
z!4Lia>+~Hx)v%Egr;tHQ(0!%pbKw8)`m`PuL4Wwc*BIUg8Mo-|aXbXtv6fQs8u3Og
z*}j1(TlUh5xK5?BHb*wm)HDrs&i}Z9SzBrawAf3Dybxq?Z&5f3FqV&U1(TIbfmoA7
zPJsQs4C1`Cn#w}_Yd0{5FGO@_&U!U}YRk?&l9aZT4X5V@=H#)@S($ZsP;~UhoiQ`C
zB?!1b7tcS(H0Di20PMVzD+jw~$Vj8JN4QEniTQZ#)*t=otHmkTP{b{MD?H)!*>C48
z5KPT2=`CE5-W2J7<<@svmXA(WrT0Rs&bn^l*0m0Y?~SIFhXNU?-y59;wI}O;?G|n?
z6+e0=-8rV!SxSadTTsXYeSFwiAk+ws%#>^>tj0
zMYch6EPw8!@6T~v8s}Dl3LzK?o%Ex+OJndK=_-F
zAAb1monT&0&l1vm6Rm|~-{@I~ppwflyv};w?@F>q!*WS`_7Wk+xqgdAL5+
zGyjO2MLtypsszmQozyY?s-<4pY3~qCN
z8Liyd$j{|8&an$b;rxzQvLqq>t!7zw7LL%2Hm88+=3Z>P>Hgl`}X8gYsjyZ
zK0kXU{CKbQ@1uuZDT>V4H-yp*k-M~K9sNj^2ZE8!b&%?s?D($zprd)<>&UvOa
z{*K-~_arILJ&*eyzr{5bfo2VUzp;`{SE_6mDVgP!_`m9Ja!qY)?BHzf$E1S%#F?v|
zQ42qR(%WxwO|@0zUycT$JrU9!G=HfwK~=4c@tdqCmPm9CcLE_o+yzr&t=PjvLUIC%XH>zvEiju{ykY(Ija{)p+xjFe*aKSPeJ*0rcJP>TNM#;7iI;<9tYGgl^N4r3h)qWv
zmO^s45=o+KEgGwP^7U(ddH?>4`0(lPe)#(Dzl)D=+ARHhKFIvhuf9IYA8ejY=cwd=
zw36&;y{voNsga(u4f#`G^D(+0SG5h2*Za&-wMWZo?%m2Y;6R>^$O8dQ2g`e#m{r%q
zkDp~c3=E3-KVpCS>ch9Ue)na8+!r0x`dH(L=r%Sz6i1lc4YB(oG@BR~b;B$j+zh%p
zHOM&*327#>E<(?(+I_Z8Hwi_0GG@+yt}YFKIOa|0*xz-Rw>S3R2vreeJF%;_2D_E6&0+C})2N|-e;Z2r6
znB`sld1m&XKYagR{jdKy_Wyjexb0^@``V;fFIb}CY(hPyl=d7)4grP+j{OHN3VtRwXMmRRayQC
zc?M47Y_@x@A?ZK=%li6?|3>jWi750wzxv_^c2PAo11(z%(W?%hf{B4ti>a{`|
zkE#5ua~FID#hifW=W~+L%@B`&auzjmtzHGMwYJRj31xvg2r;OHF(37R&1o9#DtH@X
zAXu%m{Z#_UZ+fR=+bM3n|H{XsF9q+S6`M`Qhzu664<`gncC_p=`joriUL`zd==hCv
zWha|?&6Acp*mv_P_&`$V#?Hz{cV6h*pjk$kgVGi0SHW5QM9~IN2N;n|T9yDjAQ^
zhGKJ{d&s$Uygi6*Elt3W=z7~1&&z0-H7?$B$smO_#;&I=4olYQ{HTo`Y&xE6n)?N0
zWQB}&2It$FyqQ*Xb(pi~PZbhA*!~Gs|ZKDpvBu;*?vl4>yE>cTh
z#^-SHX(JeSaNe{95^`Ag%fu}7EAUM!_M)c{p@kZx`*ro
zzi&!!g_*S0c?5zs*Vn+BwtaZ01HDkMjg9
z1#<>g7l&GZ_Sw@#WX`UEYDVz(uB8vBO9QRYMqSZS9Vn
zhvB`=7FpUBl5oi;n*m5oK@
znS+dLN`;Sh#vRm3+rm;H_@X6ns<{$f3JyXc-ksh;jm54`>kApqC7MsFf5fNdyU*VU
z6@LX5>!3bufogJ&1qfIH^BPGx$>2A-ERZo8r>mrdExjx2r}IR75d@Xo<<4dGR%VE%
z)^J6C>>g`@U@RJ%U%PifPlEaqhI!-mEL_gUN547XGqepxi+{Y34|51)ZFKX)$D2Xy^g&)m-f{QvY9Kefr^J{N4KQ
zLYnx+AAe1lf-VAs3q=TFrOf1L$&*@gkiF+zjo-po3Hrfl^OPvaJ2=m7;5}9p)N5bI
zpCEPzs~BuYA<|RtiF_8XxNd#UNa!tmpFBHuDSNdNoG8+0-)&M6zvNuMe-3s)UILMS
z?*VaaWu)12X+3cmrap)N!`EN`IFkPPLZV;#Z62=GTsYW(&ihGthmP3KF`v!wC+#wP
zgHl_HoNJ1}nD^M7t!$N(4Snx@8J<(mCAEBpLYp_>cU%PUU4#8Hm*I`hk7*{*4y#Af
z?x*8=3jLhGhT=*1--F733*>k>Wd7-YPrr2N6E)&45_-nA_8u6S<>@TAiNKs2?Ls1|X>PV>?w0W|#N0i?0q6-MkN{*n+AT7sE
zn+Y-}pw~)%&S&lo^h3`W+j)-GLE}4(CKE!Y_LoKcUG%!vP?1_<7CL=ew4ED&c3Ey5
zc%gUEM;UmjZqiood9uTQd2!I^sCdr!^Hx@^*U`rU?_}w8)I!B#UO_`SzNF0O
zsKGCzXQ-#(P};m97D2vw!*X5LZfU35Y8-dNwAYrH0UQ`a6kf~Ji-@@4eWO5UJcy59B6#
zE~8_Gt+4~QPFwC1{J`FA+N2e^jJxPfL$u0kK-0ln*g@Nq=8j7O6<&AIdqYgMf!uXA
zGMFHJ4lwgZ&s9;m9BJu)1jO2DogoOe;EB=kXSG*^AjF-^=#g1_pV{u~T>MhZA^1Iy
z)n{B@^B@YD@0d}hheb@r+kp+{O?dj@HkhFI-!dK*|+*%3}pSr?%vmT*rg
zvL2q=v4jaY%jl|qu$(SJTSLt3RX6>w5ID;`D-W71Z}9YX^R6M~9m}fim1Dn*HzjTB
zAZJ}`J~x-yMu0d-6vuOJh}868Gw%qP@|w_d5jtpa4;fT{z>Bcn5g)O=QjHCbGETbr
zu1^LjLq0ACDn#3?3zA^Z#W{5&jeQdu+hzg1U}|0GQk??mUbRDp?8#GD`{cn
z>D(nw!!l#FNsF4FOU&`g(h=u$Ewl8?;Y2Gw?V!@Kl*FV
zZnb{0a}4zBdYO&@=10Rx5`mo);kZ8U?{S*x)@9?pE@_8pT32XwUi7UH1?
z=Jo&o>wo><{@iO~DvMTPA1dQPOW2%|60qtyJIv1oM9C#~<$$GXRAeUvvTV;B?8x;C
zh;A!OPAhe3zSL6nwoYVzE
z?`SS`1PSXI={+Mzo^j(n_K{3an_U*9CHDFhdIE4L%gk$CKwO;SUWalvk#oRK;@Q^j
z#`LTG0;08Pvt#+xin#B#cab~|*=Qj7*Bgj`N?%Le5MN-%c(3Ioq`F0eZ)rE~k;5A}
zR~^DttroXp{3Yy)Kx`f$A#UI4I#VIW}WRI+qEDkwOt`wx5OI`NMD41
zSZ1%1+aU`x4gnBX_MimBvst=uk0n&nEJ3p;gth9%L#`0jv1`$M0TG(tIQn^Iu)YJv
zBu&E_hZTC0@jRkPLX4}qAYV+*y{LySqKP$e)5nE-lp0t!0ctC9&SLG~iWo>+(X{x5
z_b81%(vpPJt7#Q7I%A{(=1PovrrI)5MzNpK3ZmXQ!$2f2Fl0pST+5UK
z8b0<0<>_VC7?utKpkCl|eV;TEo@A;GfmDz-ax%AUBNdq!_G3o|9xa}t1^?iV)Q$+Ap?-4Vck>Gopr1G!;bzVOlF??r
zL#8iE;&%m=NmUmIA-2zP+69ITR)e(?1q5f89lL^RKn4i1X2#>bfk=bA*BjgjKP9H@
zp0ORVZ1U`$o=2SGqoq+xPbFeIr@nZCj5ke20DJl!H4x~|^1#{pv8^27~A6Nqw7Xs45Z&@964q8o@^3dE_+
zE)=W#>;iSyRqHt#HxQNeY>}{Dusa5a-W^cLwvn=WwhOsq@k_CeH9%`X;UdIKssf}?~K`7duj$Kf5%dMBq
zr+E(Oj#iE?uVW{FhqUw?igle$o``j`Tx1}oYqm8E@?4+`F6c*o5}pyrSo7qlkmlI;
zGam@Z;+`+uqb-#R!CEx$7Ne7ft{&y3>sR0X7CWxec?SqEpAZNJhi$t
zIspsF9k|^xgxYlBJ?_d+N4<~Qtr?WZ$U1YILuppW!!c&)8~#700d
zi_Hred3OU#lvM^Nm|2V<3u$dC{aSCFJnRuI`0cwN9;$k;NL1CapmtPiH|~)_YPBA%
zZY<826AQF|R(^@(N6-rzdGG0QPEv<<93!<rI)xXU_gU`e@SWa*O~pm3HoZIABdIh>Zg+Iohg*;Bey}N%w`{9kAmE
zLi*a17IsB4U1uMBkMYfikH7!h^)3GBsoJxD&wlbX890Mn2v$hfplt}}-6QpnS^e}s
zcX0b9sJpPjb@9X>!Vm!*uTYa3XA1CUDwq@)JBqn&66d&iNXyOa!aWLYPetl6H=AmP)jlNlLGO
z;ns@eZn;!_@`$k!p0JrC%jmwr;!C!2?Ku};kR*Fn4qa||#|T*G>}sIBh}YNhkpxRJU5`k{
z@;%Xyj$N?1#*2*J!xc#0)Ez;BqsvB{;7voku`eX9J?FZ0b%R<8NZ)-Di++LyRU`K^
zb6mrX5pg(PD~)Pape?X%v7yG{2syolf7|d6`V&dZocjlXI9)cs94Q22FZXbxflg`e
z!#ZgA_ESNUktR$npKJN3c~a;X$W!bi2va0}WCPrBikhF`M(sIv;%eGV|56Yx=aym;_i$@Y;H}nZ$K7~zZAeA3qH>DEz58`ze~USC
zZEg^%+*8nfo_#v>-o3anGQGy9ZPz(cMAs@+)Ot^;oOO*)vQRHWi_PMjbq2Zoqh74hx1sli>o!I5|wMX>os|#=FYb7Q^2=u
z(FG>4qA>0~=QHtmQ|4q4r-=h1ro=>=*fFbI%L{9kQ#+D;>Jw{Q8jJi_aO8Mby~pd;
z97C6tV#bo~HoeuAlM3YN61%pptR>uR$@Dk(Zb{c(P14%Z4TR&GPO&-qf2R7*;I1OO
ztY(JI8_?rJ^jbbLO-XX&Q7P-<#N^bnIaZ0covyuK}h
zs>n_E_Al-$vKj7ibe77@t$gr~bX0TGYtI>?w;erYabR~n*j1(F>6&o!hDdS?mxG7B
zI#^aq=1sIiq9OmUnd!QD?R+Cfr_G^*Q3@yAr=iu`*~^7rE3fU}e?X8xbVSr}MrSe14hvkDdWo?8O7?EJ9r*Z36&iWt9oBYJ>E@79t24)QVUi`Rp^{lCZl^x@+_
zKK*+c|c5de?66b!@;z6hSaV3!6W6LAalv3(*d@uDSJW4b`}s%c4FrFU@iS;
zkW#SY6E-B?4`d$kGsnL3NPDMt19GNg1M1oL*s0(i8?@Cfa`KkiZ#>wNyYg&&*z9{+
z14}5nB@G7K_TcA`pCBR9d1earHbr+|DepFSuY+gkzVrs9f77pdf?A(5V&HK?FQ;-8
zx7)MV3y@%&i6^&fMH_44an@I{635aWJb??4c&Ppw&7?Y^m~!WLh_Hk7VS2jXfUJ_L
z>@z6VN*q5}PPk!%hII3|0?9W148ljsQ$P*4LmOhk9=L>g0}=^vWJgFa5G|QVR$f`l
z!%tji-dH|#f0)Kj0doe2UpXI#dK%8QfdTb@<6R;FVoOvqcTKH;VY!0R9^7MWsE8Yo
zti>!yhJ(J8{6IOqawrt|2jWq>068-ekb&C{r#7jp3zm*gNR=ZuR;(gvQ2VaU3Dv5g
zm0<1IF>+>@-bh}GCkU!scV~M`sags`tK~Cik2`m{e=N;!RN5mVt8~I%
zfHVa1)})sW?9B0=AO+AU2oZagawA1=h28`!W}2)lZmYet)My=S^T|fypCCQ+(`IS|
z*~=(A``n)9ELvLVy0Lr^|2mr>hJe^0t@eFh1&lSdlEq(Q2ePT>mfj#PG_
zD4!DPAD4d>-+ubLwg366Z~wgY_&@#Y-}}vvfA9By@&|tz)=OP-(*5Z%CfHVywW}>2
zb5*l(^xZG#AaAyrb4l1V4Jik?rOhZOL+xsP=gY}`7Ep$HA~od*lCyLtgN`$`Gv52@
ze-sHs?0qti0@+wMKvlb^mZ3p@%?$a%<+SQ9v=Y-<8ZKmYD|xI@z0+6ay)S3-yJ1O6
zFZ&d)W(SBBFSX8Hb9(Q~VOK}*#RDXwRwx=xuhM{%YFp>sUlx*5a>TJb>G0YisYcnl
z&OrjUUztyK;c^gbJYnq)S1&o^6(Msre~4Q-)P3)KIl`f)IaxyzYpgJ+?8fDc=GJ=5
z@!pp+m#o3#(F%zV>Qw1YTKIIZuj+d*WzjNbtJ|^RnH^56Sky?v(>|bcBGUEVT?2iLRFN7=+(-8@0Zon3R$TXP>sc-rYRgeK6*{%_<8S_
z1+AVo>+tP(360&6+D#V(OM$?B?=U#%>@xGv=ws59*+9tbv^=rtpxEB~a(1nm?s_xo
zDJerp&Aw(pJdT;qf3(Y4b#eNNe>8`vZ9NOm^InA~wmK^y|D#>+ntIkoxGJn6Ou41h
zk$YLN!ao1}AMJYSC@~|?NOJm4!+Pg&@!aK%mR$d%UGE7k_2{mXwz#+gKMJRiszSdYSfA1}>6pzKW
zY}3Q;+OxoRAo%gxkSnCV`{l6fz-IBo>AHO#VM72!1zW~{Z8+TXmxBc?Cndx+7W&TF
zw{t{aN_GVM@4Xo*n{<7<^>j)Z-KNjIZDT2A+D^XrQU=GNSt-@b!5d06Pd}1nIu!99
z?@e;7=_>3k+^3si?Q#aqe~N)!j%!`-{j$Jqf>z)om8z68;9BdzK3W@!m<|Ap9^AEqsGn15*Zguo~KVxnh
z+3hja!axa^-2u&Emr7l-M+pJ4B%(pk#B!CT$ObfBx=MeD~}Oz|Vg3W#y6P
z$@1!X$=yp|$ZThiqOJ3c?h-8LN*5`eb>kKLeypg$o(OLV|CCFxg&kWK;C(E6@!oY~
zGh%&m+OG6<30CkROR1|yV$jB8ZmaL6hn=UA;u36kJGqw1@t7GQ0eZZp>aFz#H92lQ
zWboF;)~PgBe~~(S<)$3!=7_z+<|Wt-wqh0>x{G}XhEfvLGEYQODdKt!HZsv5ghTt9
zd#--yvrtlFd+jCI(?o>UlMCy%?X+?arT45S#p{0TA^+`H@o{|wJ>Bc;-+lE@-+cJW
z9}P!u{r+oK5{Imot=8M5pXOSSXnV8s;N-U7;SvPOeCg4hmwo6ScLxXu0xAKZ-)#`g(0u5Yg
zmUl$)Ry-h-cQPoiQ!TCh4wukZb9IDHogtP2xyqJ~R_AP1^*t^@vm|~c;^>9tX|nCh
z%Tmste_pGmcen&|Ud6(TfEXIf1nP{wP`+yAY56-`!mb+^fPWhjoN?}_i0yNhf%G;#
z?{Ep6_ua|oAtI2{GBn{D1N$-LHGAx_OZdU!<3Ikd4}Z-+zVf55eiph<)jl|rnS-Ch
z&K5|elULZZYq%3)I`A-)bM8h@dB)@{S7VW$f39n|`2D8;hq<@yy6ZNMM1RuYrGW;%
zj*{HTT1mz$$-VO#=x*4&mP9Rz&cyTUr|QslB+HK(iF==u6~!WH`f2RFf$l1xQ3bLR
z0occ`93D#JX}e;D*W2Q%N(s#r;vOkGbqBMy`q*5fM#EikH&~!NF4LExyxg@b=q|hY
ze|4&RyDM(c?*nj5dGq&-2}CqD4~5!TW5;Fh`S1VkU3~5TV(Z&)KZ8qreO%vtc>izj
z{Lg;#{m%(OXws()74{Ac)4*$GYp2<0c(Y#YIirw%5+=^c8*FrWq%&+S#h}gx4=>Ld
zMa#hVqYk>v8SmPp-OtXc*8--~p7bt|fA@=vHtY05?SrgmttyS~g(mYkqZqX)_@$Nv
ze&^GMr6}xD+6n)(^`ud>;!R3MqN^QXF=FMwH=jFxaM9*>H_;dDYPBr;cL4%_`XB2*{iDv3dRf}Cw`_#)>!%>KIe?Mt<
zMlx!#OO1m{+pB0@#{iwGfIH_WeUYHh#>_Jc3X}8%^lJ5JTk2U)x@&g+*;&p%zW002
ztp8HLPf~?M@j7=g-q-6JSOAY*JFgDbmyAhPsamKqQ}eE=EP(nW!$%k|!Sq**sU7-u
zblF2DDZlr!Ks+a4qNQ2JE5_8hf2U1N?=nlt)hg|U=b<4M2>JF`j7bkg$E7vTInIIZe%b*)E~47}KnnPK&p2MugPW@N#w&@b||V
zMSaPbZ0{ve(x&NYx3hOcTI5ArNl22f7!!N6;*bvQPKWR)MHThaX*`|#e|*K54qga#
z9%QBEwRg8&H@814Jrl=y$(XpSy#Twr%%OWtysEQy?qh3`xVOdT?|-#EtZ(9DeE5IA
zLlKM*|Kt7p-~Ppj-~I6~{+Z)sI+96nA
zE#@tgZ5R#gGs}4h;%5MgZt;X6grE>^3Yhf)E{y*i)?;Q#pdIH8zujLtgx~)Spx&?k
z0QJ7YB8i(&;$MjLAeUw*r)+tMD5!+8Vc_(PT5^t!rw&Ga`)+{Ce{=hwt>fYDqtGIt
z(_87OgPDWhe5RIxXEV7kCg3cROz^v$pE?-ZSdyUrDCpzR$(KGsbhwW+^!RfJbLvcH
zIaOXpjE6QV-U!X2IHtA?>_$if1dMwe8iak|FDnm-hX@9UfUZ#|6$(13x(72
z;cc263g%7g#_ls?fAySa4B8+gfAm>XeFC%
zkmeP7Yw++;nmAM)8JLqw%9z~=m0%(J_sb62|NdK#f4_GVfBOd!xF?mi8m$siV0mDg
zZtx)F@_rsSy7N;85$&+6n5v34`dkBqh^;uAPR}m&DTAOX+RQN)0Nt_P6PX!69!dxX
zm(FtrvA~&Yuhl6zv|4Rb6J|KRPBB@xeGY)J@7{mCINL96?D*+V=1x;tn;V}1iKS3w
zBylg)@#u4#f4SWnytR~UOuXE)v?ewP<&7`DXsPg)@Y+?u`&g-sQ`$b$k0q-Ozf{F5
znFUbc*5Feyqh(3A(`VZtDs4;aXSA}#EHU7k5bpktfA$sYDqSVlvS)M?>GYY&;`pCTtkHbK5bk}z
zA3t83e+kcoB?vy2t?uanIqWXsQCz%nx$27saQ905$vT{gT{OBqrg$}LQTJ*TiEUuR
zZzzX-*%0o1z#l(e@1u*&J%YQf+->x-`wnWr#jZC@(De=Y;#cp#|9Dw`{?kw7jHJO&
z2W{bjRzqcANsPS`5~q(Tw*>7KJp6$%y`McAfAaRq38iozHbIqpf}+fJOu{@@jFiSF
zsdmM`mSrn(PtXN=CXcfnx*gF`DbM35O^1$q`IewsUDJ8EoHrGAJ@k~tfbt@@aUDVZ
zQBYo1-R#3fo(6)^%q1
zo`Wh{b(qqco+r|&$iDvhT=9gNC-0@o!4sh>Su{%6N}q9gys$oai5N^@LCP}ax$2mG
zBVFq}SH}a-1imaF4D+E52F?puRw=#^GhKIeq&WhG}raDme#{Lfn?Dbrv|*?c_(z$CQ|5W2397jqI(2SNf(C>S&<
z#pbCHZBs9g7i{L1lFs7k={^-;40&_cnod5K&zD}2T>vIhupc#)$v9OWaiTGQhOG9)Cf1Sfw|f8ESKYJV3x9%6$L|2y_~XjWg?72
z!3Dqm`8H+j^4YFUP;wQBCW2T0b4_Dkp09buqRZSZCAu63>pOsMf1pOqV%PJP;Z7F;
z6*?}mEtPJXx=C2_oY$|15y)NR$Udk{4bQk*!$L@))w~FMeY``D!Ac)v>L^CZYPq&i
z1qE=ebA7y>2MG1Xo6K^0b-_YNsDcGErpER04qKf27IE4$MBf3thNn%=J)W!MP43Zw
zilwe4)fG{7=t}e;f0rU@T_5jY%5+YCK*NXT=~-QjOnuWRw72sRnye#R6Vmzu%7woc1b
zp0ka~e}kgm=H=_17yf0`ZVT9&7Lyt5Op(e&2i4XukB9Y;Hb^beS=N+K0p89tK$P6C
z|9&SWWD(WsHKnLg>0;L<6pPTCUXHI13W1DWJF<+;o}2KFBvlfV&$rASza-uSA3M2B
zXD%yrz}N)~}Cepbhr8@@*}PoJ@t(e`BYNbVI&AYa7T{QLhD*Z|Nz~24-xV
z?&re&>v!MA`ta3n*1K=N`scs9=clq263FYY-XvPLZ2Gvk>NZT6t~%;%fl&{tKN?!u
zdbxY}V);3Em8Gq7U*IZ^+Bbr(6Rnip1v!+BcByFTYPSWJepW{I4lq9xcGTk6XVUZe
ze~eP*eSx?40+?2hh0j{gY@jl9%{1hMtoH?`qT_VKVgz$Z^3b?V6J@cqG}gE+a9b2B
z#XR1+f!b)h1(sNRs?2lxx-D=QR{a9F0oskkCyK-_fXm$<9$U8s9uGM~JDL3!yHVN5
zmHX@9jJsUeCij!T85J!dfvhBTffo0kf2jtrL@bvJ^Zic(9{}<`0f=Z|L=3jTxN@+?
zIn5-u1tu9_C$8ZiB(4Np3AP9FMKj5BEQbh_g6nnkzReJ=0;7BO3OKS7JAI8ZdidDALxh-(i4bX(7kseZ+e<&KD
z4-3_ybuVMK+XB;yd+$`MgQa0|x1cgO_wj1iE4N2}68Ib{{84?yR>?@0nBOxsq^|_V
zc*{Ee*RSsRsgEljh6Fh4qB>+9NdD0)5Dp5!!t}M<0_U>w?v&j{!Q(|db(!Ntp=>q4
z)_sA;*vPq6kaV~2NujWCSu=G|f4Hr;1unAsnk|ET>f~O&tl_&NX`G$Q@cRPmPMNKE
z70H*a+EoGRQ1KhD$JJW{*D~@Dx~)4D!SLq1s$`i__kM4^UvqVVB
zC=B;~5bz0B;&;Fy@LK
zC*cn)xBl);orT|heE0SD|Lx;%KE(2$_v6U$
zL}*liG#GekAn+Amucle4p_==-J|b9@mIi9!TqIa5t157zGOz{|*IjJB3^05I*LVymB3EZ-BgCasUF2YQexkF)~@j)f7FFFx04
zfd4pZ=fR(CKxbXOf3Sa-XYX}#&vNbSqY?A
zl4f<3UKVFx*rBRvtVS{Fl>DBk)L5qkbu{ZIv+x4*xLHU=O}TMmyEux+$@#|3e|r
zdt$axe-*`LfM=zq03%}}qSS?)
zuo-%fTd>KTh?do;LWlZ7lXZZ+RHC(oe==j`6JF~YN5sK-RdaEfczTI@>gNpC
zN{whiwx#t-ElnG6bIBe_Z;82D>PS9mXiKM*o%TEdJa&+ZFn{|!F{P;pvY*erK%*xT
zAbp9BmF{v2?pqdV0-?;Qse*cyeCCdpGToSD|hSHVfNs&7R~jPM-Nr_
zkYcyQ^uvQbRLWKlbsI2~CYXDG9j<*3e@BWH+iJ#Jr8n=y*F6SI=DoRtq}{SU<7_S7
z>*UUej#2Yyr|@zcK5)n_`Sv8k{7l;>KBOFBLrwOyTJNjZd+N&JN2@hiRe;V-(ha8FZ1R_B#EPk$B{L3#cG674B=e{YqY
zXB%K(h;Ge(SKO8+?dxeD40=FQNV;Y#yHqeqeccwf1I!Mj(u6QuXD&$^DrNLUk>zw(
zTySaLm6+T6l!aPN9yUc!Z9dnvi}HtYc@0tyDPYj6dwQ$;aEqwU0|Jq6i#vtD*7w54
z3<~BTp3o-PWCEDo6&Fk}3Igapf3rK4c?MfWcWIbsE!{Ssw=~S>3M8MJBe{c&qN)mN
zDo;5}xh-zAf~?n48LL%HtVM`j0CVZ5R&IMe_PgFcqhGJ
zmU%s>=nm*K*KO-J%ARSX8jOkx4h*a#v1x~Ac)T`oRos94o6FA}e()n7JL4rQyIr}F
zJCPI;RX|gVO_5=-d!lA)zF5xXOxtxm)C)jC3!|f|=9{9fb!>ph1JO<_!|l;J;HPw6
z{J(A|?9Za^0fqvCzmEhae@v%pnDqs2k1VwPo~R}B#hxuG=zzGEj)HR2c*vUlx*gRY
zN3CVCBYp-J#cU0WA_f*Ea7!I1{_&9HcaHcWmAj6`+K&
zLntaEQ(rJs>2rACkMW^znoZ->qTbedxXv%I3nVLV1y71+_nW@!42y)6kNhyfYk0(6
zlhD!@1weDXCn^ZBe~+=~zQ!h(dAjVAuxODIWn8yc|Kq5;=Ge&tyyk%YpTj-6{M%7s;7ERIrM#^k_WzS0B2h=tYY1>+!VE(iY&iJOA^S6Wj50e3U?BIZRwts)Dt16P7q(z%Qh$#4D}N+Qw0~FeRC_UbkM?#}DuR{V)EsZ`Q}(@IUb1|KVT%
zsh{=9`tACk_^G9&SImP+jooILet07Ss6{z%iG54ZjtN#8mGfJGa-CIupdh!_0D8_F
z{QzGa)W!fhe=lcL9-hP)G(v7M%htBv*u4DGprf1fekms0^ahf$oes8uweu)|z%4=d
z@XWhRy@HrvVk1#cfVx!tp);ko1hun{7SNJ`NQ&}5^x?s+as0fc%KoK6ofU5)X&`IP
z8EjJQePr3|a5CL9dwpz5Atj*oM)0f+cyi_{xwH&Se|JmJ*k>wYv&UlY-B1|W({6yz
zZu<=p!e1DaX`uvzzsGDWX1r;#L!2IIs
zaQIF}>7%uVpV5KCcN@BiUQ4}p+B|YBxFi+=I5IGq*W=mkK(VdTE;TsWq!FdX-#1Rk8s`xSiZ;>%vy+lu-=V*`|hCe04ZfSJb=g
z-g`OV)?nT2^-i0Yo!)4~|B`S57&BJV=0<X96Fpb+P2C%jMx#)`3blP+_xD=$=(6MXT^)`tMnz_3Kl9WTRJ1ChnRj&Op}O?T!__>hVWtoa@vMF*ulev{E8Xs|
zPSL1xLi9{KWbX`411#$CM3?4Q7oa+jy4*?mVo(U6HLP+4C-YvAa4rwm7jFr3gIQ6I
zWC&0hsL7%=vG~>DW|34E79SZ(RO_{~TSp5h*|?k|fLHRy0tG4&Vdo_QrOUcL}06tw!T9Z(jzWm##in5|iqeukHaJD>(>&bD&J
zt4s-nu1z;$&YiDbcR)mQmkyX9eYJFW-V?L?ULA!YT@BjVd}Gc&*056vSRDBaD;@~O
zvHGtL*Q2K?wgK}R=H{8^1E2`2e~=Qfm%mJ8ucj2&-8v5dl6DcNl;Q!{^s6g^I4@uE
zCSqyS@tqZKc2QASHgMU#JRGmKgF%jpd_sC4D}yr3w9cxvT^=qTM(J2=NMpQ7IHpw5
zi2|{0U5%nCy@nceY%1O;2p3sG>xJi#DseT=l+&g~wnW{GjHOpWoV`)4e}qJ+Umk8^
z4h}#+dC0y6USv%oKuAmgm(R@exb854W99T!^0t#u*%ufIs?yanvq>LqjQ}6%XBO^E
z0rGjLqEC^2^~`L%+*uMYI$*KsV(2w9rOtL5dUXGHh96
zZxeaj-_3&Qyfz&+H1M9^n@q+)5+
zU0X`~a%e>N&J22>s@4Ff+o_nUv19U1j9y({d!aj1J&|jh;5}gHf9y$|sI0v9ef7HQ
zlDp=`Zu;0Pcb0+$!bN1Bbahtgyjr-y%PC(WOAnoD7QG6Wo4+9u`32#I&?wLp&sLGD
zYg&V#ox9OMELX$YKF34NuWzYSiU6mBC?hl&b9u{);|s!B<@JD6XRT6pf4PpH6GD1wO^prKr1
zp#Tio4$y~GwsVSh>>=!Ob+`eH*8)YI{^?_LOF*$6CVO2{xq6xFvP4;9US&3rpSrz|
znrr12$kSJtRvJg-bV^di3d3iY0I?!|0f1Xj(83_=Ar%f*GE4g}^
zD|^elDvG+Lq^Lgt)<|QS$NH;YUGLPHK|rsBmxE9n*1=lPPBAvhV_Y2$Uuu9SfKjG6
zAR)OX_i=#5N?yBP9*#PRInQpW;((Vn*q}2&TqmCV@~i@+jtY>GU1or;jw~d16@};k
z$X^Zo25=>Rf4F4pZPs(->X|uc!nP6B2DsxZbgSY)?{)>J%9z=1zcv)VfQ4^~;>qe~P=dvU>;lpB;}!N4<3h0R6do
zD_}~RW~o^VBQ2=|3!tMaB&>HyS7#L(iELnWBrah0E#52ston4JHtXe2hl&Gyw*~N$
z`yDwokRgd;bppIxUCnv&YL*YEIPbzy9?d`~3la)D8SCZYv>bf4C1WO`IYkCVIFH+~
zHC?_Ze+H5Y#1MeCl7=^s5t^g58p&tv`}%P7o~MhRVX=XwTV|I&xLu`i=+#*Tr5lqY
zJq%ia)fcZbmby==zJ1!|GxLLuALMMy*iu?AuXoAajv3*WiToh9AE|3#JW5j
ze^?!@sIm4kOF6KME6+%^AfH`6Gs*r>^E7x$SV5DOlT|BNN;%_d(7sLrWzB_0=Akq~
zl8SJ_dHsE)qTuw>09D3O48hz0Udr``HqSPVTfWKb7
z5H_#T$ir=dRy=h}8?$Pw$Jl0Z`O93Ze+_BYqdLX;&*ql?E>!ESq37~&F6mS}P%Uf)
zYHS5c+kjW1x9gt{&~iPj6t#jdW2RK@-WW@fzNL@!1>p>78F{)EV0H%+!C%2l!5dwt
zly>!z(>~zDGaFJgifP_}xcF+tPj063^7a_x@lV6@&ZW>fl5$gtTcw`4;HfJe%^~4TycLuIch`2O`eiw?33Bpk*4FDZm4uw4S({EaRb2
zdJd+H#TbAO1sHIe?;*=(Ts>7he=Dzk&(VRS3pY2KH*=cO2B7WLQ^jkvG&3+tOf34$
zP4|vBS_Ns-_tgc+>1tRUbJC;mTJkhx>>3n_kGTA}khUT1q#e0pQPgT+LJxb3^I#}f
zPZh`4YY2#fY8D%hncPlWO(i93S0gdsY3}5pA1PWNBCDndXAc}*46k11e?ZUP(q6e`
zRCCG@MglRn9X{K>yc;K|I5UC3fyZ0j0p^jH~`!ZT&5{%
z>dP(f)-OMl!54BWCbC%o~`2G9eeOTYbuRgr{_Ty(%oZ{cs*S}jI-+%c2
z2fzGzKmT!!an5maf5Pk23BZg8I|a-RpAO5rq$dm^*BKOOZKwAPG8=W$b{-@Nm@|Dw
zHu0Jv%#8FF7{+wq>$L0X;6@3UyXshLKVb;SK&pH8%8S(81rv*S7g
z${z*YRZq5*=nIHRlkur@lla`Wb0&OR@*NKiO;C9QngRrb|;`wzeWAM0EE{%sEP$i+pJ4vEE<`if|@t1TtM5vXJvQ#4KQH-vO8n3tHG!+P~ps_OD%
zjxvGt_Sf`9esRcPQ7#J!MSu!c7fNyT)KkKzJY(JwGMk?SSh-N;%W9nxkQJmiDb2R;
zf0#b!;jGghLPF~pc%YR=DS^I@ABiIO(}B_vu~iaF%#7z23t~kM@_otNk^v!
zKlsim$ZTcomPexatLw5l<(8mx4lTKie`QIjLPgt8OF-j5#;Wt!{B-^
ztJYcoWV+{^_bovUGc?vv;NORBdtHFsMRz3y$)WcI1Gt*4gKxg}_HkEn^}0x`pTqyao2IrUI_klzzDx_Qeye@g2B
z9=udP*pjf$ZhkJ*Wd32$Gdd;^XbhU@|Ei+|dzfXZUgX1ghD?_KN4$U4-8-Ly0q2ftu=(249e_QZ%RMKswW~2^eOO-1<2LAdmdW4KxH?p&bbko*F_ovoZ
zooPO47%DO9@nWZH*+7Z`{)Y00#ozpp`^jgPL2SdY+-Uhp;SL-uN
z8v^=I4>}&BKX)(_7`rv*qFKQd!aIv{?_eGRZfxPKOXjk^fQNDhLZB*9X`x$hAHdjv
zdgXLo1$=u-VJll68^GaswR8IbW}-M=JyIyjM_ZSMFEX%X^0YPQ?E~0#7O#cd@boX6
z$RAlGTavX(jdke&{$_n#zxwe0-`@FW_BI`V@w-3##UK6VfBT;s=0V?*2JlFXRe|+$
zs1Dg6s*PR(qdspWl30It-AK(qChBv#D^NIvCUVcu8%fuB_NI1@mOh$LwRcb%4<9|H
z&uooeI+8(2n9<9A_#o~Mc0Xn9$Hk>%pYiX!bR@IZwkHU$)0);i?F!%w=9{7b(e@{f
z1VGK%NDQ2~e|K&Rxky>SkShzO%JW9DGHPW)E<77t+?U&yVrt;(8{qLgZzNXPwR#R-
zNep_42}VLG=o~a^zVn^@KOdIAi*G;v_;Aip
z@q^^d8|)8bq87(WR=f$(&vaGYG^SihY$i3iX*^Lde<0qor~ev;_L)wNyT&xOck1a5
zV5$SUm8hZE1@i}GZnjs9smLx%c0}~(n{!siaGH}0%-qjp{hP*At2E6PyguowdDgb{
z)Dp~hrt>9ZlH{A+wF4zv%Pib^OmrduNVLxwecm;uwg`qs>Ag~xb2X&j&MjE{!CMAS
zcaLdjf6r9ix3o+QP`UDoDqYLsBz?(zk}L^0Ha7+Tq@6;|^r#s8az#JW{(jq-oW2|b
z>Ct4fU2ZJyVu!AMTI`pMsZbxzfUIXr0Hi>Invx9}^l8`nk};{Vt#a`q8ON%P*;Sf4n8r=h2O+E@@>R19dg(JoB>gOV(Yh
z?gmbg!vLSNQkZP$8UTuPJ9x>uYXb5ZSp9SH?j6spd^DV4+Z56#2}DzMfGXz-FT?L=2P0Pwn~3`Z%{I_QI88nveZd0Y4Tn)kGN
zii+%*I#K|y6po@BaKU9a@t4deb?%#{IBL4Le!6nCa>}ys?mn}pf7h4}%E?GZi5@`B
zNK_5&R;AKNN7F0DGX6n4ks?42qx=fA1}y
zwaRP0-9XMRE6q(0gI0QOrHW_p;we4V7rmz_vgzi)E*Tx=kW1BsVa>7SHDjuRR}-0<
zi-N%&Z9S(*fW*k3`z4r;vd3!!(gl4DSISN3BosFKz?jF4z2NMu
zWudD|*AmZlz9f&Y$}4>}08PZwJRmJteP;>K0pR9q#w5i;y9Ez+%z0y`e|PU{RyykN
zOU^cyDhOwz7Yo~JL)`;ee1mtk-TgK1X<1utvJVCab{?lxA<5FM
zs`DlLD7yoVZ0R*OWNX#x0}L;%)HI*uFTrqojwWk{;3bA{VDDDT$!$sJ&Cce}V3rUWTXATayjq
zf~m9*eY5&Y7A7zJpF_{*pqWU=S6io3&$oiVjF;>;@mrNrZ0xc{9`YR(9IAm}f(*Zg
zi<_H|O!vF8Pz0>SUS89n!`iCpSz`D7K4^ZFwn)eiyx71D5%9+SiNrSAQOVUw?yk<<3A4w}J
z$?9#Tp>%8MT7eNnUxHQ65uWv;&}pf(t7%n{rIIz?C$%Tcf2V)_&4*b2zutfN?HfJ4
zpM(#-es911xj&A5ZwB@-cWO?(7(7%?i7F4l6L))Ky?jx`?EN&hv?a{|5v!irPe)nI
zV8~na`iNz2uaB+O)F%S!iY*J(Z`KCDP;Q90q>P@1opmJ1EOYkRa`b(w#T#ZdUldVL
zA5gCP%Hs$Pe;!KqoLYa#DQDT&M{KrRb2d45K;+tv1uA$NB|7C*`}Gk)7?k2@h8at1
z>c}?fHr9|>$t<2%k;
za~aw}P)cL_rPN{Wnl}=VtQ+3yY-0=3POG(2V85_Chc4v1GzWgWA>s}!4qiH31OEi2
z=`~fQILi>fA#T0qiNI#4Zhvm=~+V{^+MZR6@423}i)D-^VRNKVzJlQBh8-
zeqZ1*df~nrJJJhfQILB^u`@7v-%`%FBJeQNR6K^Eo|KtS27X<^
zYxaRf!haL&VPworc(Fci3yhR7kLO80iDG!a(GoykN*UO)u6kQwIs_K{1Mo$^6^q8^
zhJ5;B8@t~(xlTDvfl*M$6kMe_xEHYd9=!oL=(sO%4?5(4_5s)Ow%a7?Y$;Uxc*{NN
z7X<#Num1MZyzs}LN6!3M4!N_pxc8V;Ts1^!?0*GKmTm}%@>Ob2-FW+Bc(^nJV+WoU
z8z4qEglvsFdzSX6t0vWIEfi>tRH!O5ZV1`I>Ws`P#Fnt+#}B$LYm@zV!|X3JYZ0U
zc7I^i^L#K-T>FNQ{tz#mqx$2zOB@jA>|)zZ`?^j{`-dU17K^r=*h|M-fWYiBl2L*M
zB>s+&!OlS+J@2}6r65Qfg$B1XQ-IcQ2zhz~gw?30ZU9;xpg=Qep>embbI0^K8~9^O
z2|0AGut1UQz2#GRJEmS%m;Wh#_hEhgPk)~Tp>Ew-{ZS;Q^wj6}K2QX)AduRn8hHy=
zC!YQ0>%Tu4Oa^eTb~a^8#=x}B)Gv`*$Er2;rw)dOAt9;b^CNVxg8IEM41nN6rvKc*
z&_o*9j9m`39G*Y%bX78qA$83s4`!D}YpJbvqWBoduqBdJAewx03)4>qvx+q@S%2Nc
zQDHEwf|`92eYPa)QwLMD=-EI?Y^EHC_xfyQ4y2blT6ywdXpr3o_G4qtW6NhxvkUBv~EN>Yzram|6)i%YZ
z^ZM|fjx6Ijw|eNSx#yhEUD{OR0DqiMSk7&Ln_0^p`M9C7>)z9!I+#wHMt4|qUR!9$
zx{_xXAE`(kPae!pV2a@9m4?(M^?`MctP5}=*=bww{V=t4J#{d<
z_LC~=KvJXsE7nj}RBkbz?SIE}3nc(m0+Uw*(Bon%-je7nC|5jsf9{;w6wW$sfOR=e
zjLL~yQLvXI{XMxH&<-1`SjXfW>5r@Rew!)it6$
zcQB|yBh!SW@-8h9XK_<^GPtzCH=jC~1I$pHcrJ%+;H?AhrT*wz>3{9+D_1`m3{qRe
z62^F>J}wdVUMPp|fLUYQ%$weSg})g8^6|s^7{C7gJql}o7KE_LkFbm8oY;vy0~4Qw
z*0Q6fbUw3xec?D}~L|_~Q?J9Dh+@_&uR+aVm48E_j80h8-GVv^!!7pm)m|=}v*aXAJ7Wtr?u5ltyg|N*eSs
z%hY@pjgb2dAqzgt;%;72O{cg&R05jhYQY`z8$wo==6sTXvwu~f$^=qdl*WZ!@RWW>
zNRtDQBvNPU)r02{vB5iO0G?7Rn=VD{fW1~q}9#XQ{9p6&h1o~5|QFI`}
zm}h3)5*@iK)-5kqwVe%gg=redEBS=&v~u6vk^P(&-7gF|qGrQ4JhKBV(ABU4=b=>R
zvf?@^_#cKmXMdk$l<+2$mEhS&#>>VOI(gmkEk{)>jttb=t)9x|L7x__Q|2~VH-wxj
z9nZMhOj!J-6+B%`kgrF-;=1EouGxK1lYN5qS{4?HT2?;vKA!20b!j!uIN4Ti%_2}C
zGyxT_5-?*PM$QRF)F
zs~?2)vS^Dp5fC*uWvOP%Z)5vp%eE;v>Sm|KtDmKYsNk(H>`l3)LDh0gqd(O~3>Y
zm(($t+ke26e)@p2U^;j1GCBZ7FXM6RiyN?Wm-0p*&#ePeWJy$q42bKe)MbfwVn}Iq
zxW92Oe&>Lwxq|#Mk>I-Qy#>oZJ#M}u*WnvJeLy|Spxs|`7eB~Mm_4ZFY108M-q<<1
zb3p5ps@F7j$~r3iI$l7+T5Y+UH*n~k1F{Z4<9}^X=!k}8$1p@7WzbZNvel;#$OSdB
zWF)zQ_k@oplWv?q267VS_wxrt{cqlBV5bef6&apyIeVnz@dm=Zb3lRFQf!0f>Kw^S
zo{93q%OX#=ZTAl7pX1lR`+9xgZ=V+EZ&Ap8|4+aB=D)9x_M7vYq9_-r%Hv9DwT#IDYTc6{yR;V$tq(e2^j=Qc6-8YL5?#9h
zbx)tKd~5wIsRv||w-u@%0jaa@QM{m_o>vY{c|UBUC_JPma@ttvHM)T^V6b;*!oNB+
z+E0&DhiVEGi8`hsMLLhq*!@yoIW*ms7JuqK=^)$kCSu-n<(j+BH;hi6GqhO@=nQXZ
zq-|xl})mPA6Th?KOhDT{SZ`^a6fGTJLY=!pc(z4pwVDud+ApzxZ2
z?Se8WPaE35;D>%6PY(C#`uH3kbh2i%Hh1rb)s#jEj}GedM+B~&>Xx;06Z#otkbkob
zh$fiP*=l*gi2U%U6?%CHh#gT8E(Tpe?;DMP5>;Cz+i>f%M!TjkzO&8-tsQLbwsD4?r%2b
zdD1ds@77z_m3lG*%%~2@4vuYQsRa2^Eb#`n&^}XOT7&1+1dyI2B`;oE4J1SO}F4CVmmVW}aJaXtH
zZZ@@4P%A=1+gkD5HN8=wP%hh4Dx`&CA-Swbxv>dtJ@*@-7&Nio!basW@)$@n3WBi)
zVe>P`Vn{7z>Ta~IH$U0rlrF<4C_7er=2(1}DJk{y&_4>Ky^d__E2sc0>8WcPM07|}
z%nhMKfffL+qFZt)2Y^ejAAig5@W1=jhxdMIqsEVa>SMEFwX$@B6LEjohgKN}Kw#65
zQtiIL-tZ3F79c|;#gwF}Pnpt(tX%tTfjjS(9~-XPQ)_t6aO5NEqnX%+hd6&8cqWu-
zFXJS-rG?hqtZ6kZwtJmoMR6wbL*W}lLhO0ZQ0gP9`&}su7ADGJnNpP(khY0
zs}$o1MWR{PUEcdiU{rlw-CcH!iFn`u<#BT$3mtRr`^G1O|1|T)6RHhdMnADudgqPv
z&viTgjKIG_X8HQ-`1-5we)s={eSH7nU;pAy{$_pr4gcfAyMO<~zkc@5eO%U=gQCt#
ziKxSE31DC0j#xiol7DUt-3>2u@uJ(=+uNGi#zv%|GSz5$W9YFJkeV8^+j=Q;jcQi0
zT#+z=q~?vG=_5TA3uH;!czmRcqZ2E)dX2gJt)VeZ3mc{kI1L%2QYCfHWeI8Wjl0v=
zgf?lF_p@_CVhmbq3@=f8r?lsOL#y(d&>mG9C$&_|UV2NIVSlTpd?s>;=Nm(_2X2;~
z7El)f51GRpo7j99ruem?pLfmqFGeJRZH~`+@NHd;8cCR4BR31AYT3QKWKdvcy;u~=
z>*N94y(bczIR&P9)$ApMlAd_ZFk7jfoo%#Mz#{7kptic&YX+scA9F02B|Lo4aMy^@
zwYC11DB0bEYJa`qm29L&U~&BJpvL!>Wjs!3Z@wu0)}ZLcW$XnJOrh0$+
zwY_Fg?2fj1z%RMw_b{4CswF3U*3UIF-ZiLMQ^RZ8C^2|FI-4RDl%9(LC6@D=)wj9v
zxF?qXAtfrkp0!fX;0bwY#A^m+bbAcY7Vm4t5uGbjJbwu@KHYr*>#afID@8H^cWjtX
z37KH>a5PPd`kwrnlz^lprHo`Ncx8@Whf@I}0K|utF`hE0U;UB)e`a^!FZ}A~Mqp(X
zF}0~-2k3k@i9n*Ep7qW=8-B_lGT>XHnh^!;$Kgfs-73(`0`qOlpEC$5eYsLoO;T>&
zY$2zd(0?csrD8s15CxfIPuZhvOu1PO?L4S@WVOjk<0*rvD^DzeT1wx`JBs84)RBUG
zI}T{+a|U6&h1f?2D|X2sV#DTC-;
zhhZ4G*TLN;?*C}oIW*0aKD&!}%^>DVt7SEyHh;`YxIPr+?Kw--R+n1vO%m27SnuNJDGusO{mW9GG6+v&djq)-P{30UNYRiv
z4>-5>nVvIj&Z55(S*&V9)z`$5Z>*_5
zKu~-9oj?5RXa3;FCEPBvpJJNh^qf}a-hWnZlFFI##^so+Vhi;F!G0Idbo%7wPMX0|
zbmin~x5n<=0s!^LAcV&a{ft_HKzIql-*6PTD)vgelc;S-FS*rGfiaNAcUwJ`x6;=|
zvA5x!Z%{FcUmxovU|y-io$%`A&e%DNmsAqn&xDG;6yFBhN^2T#Fx(njaz-Q+tbff5
zs7Z<;ahC+#zromVo#zzDSJ73DqDZ1r16ZEj@&2=)b?-dS?8O91nJKZgmihrD%@u81
z58b!UwGz3{3=F+yb3keoq;?(#6TVB#d~fWvC2uJJD~&D;UBK&N1^@$=*t|1#&11Ha
zW}VAtF2?)Tvk3R9#K*n4rB=@B$A9(Tkla@Qu0_p4rE*uj^_}B+4X}`sHXO%3Y?%fa
z=?8Bfum0B93xIrQ2P)+yfQDs`1+Hs`VXmKkYwS?=)=Bzw42L?}M;+GID8z=hu5YQH
z)}m#yD7O!4$)xZV=r7q@CxBPQ1{Iq;1{61A-
z@WMp0X{~vd*muTmD5A8W+o)V6d}VLkK>`b*u}!%*wouVsFw+*Io;<+P3%k2hr@iQ{
z^}B8Qb&Rux0z{=EuWS(FzA(?a$gTC8H;+cbGpit<_!9bRX{yvq<$vB9yWvA`+fWY{+Wmv@y%5g|d^k#zusW
z554svJ7p-OjT+QlP+i8jb*`asW3KB?ZVBbxtQDlTSl1$w0o6-5>{!D
zacgV(SnKx_5sqfLXL;(d3wCPC-a7$<;Tt8r!Y&JWilm
zVv#woVYRZlZJI_lI`Ux!-WmJMaZWmw_g?y3;7H)W5r0nCtpCg0+jd=c+g5@!lx7>Igppu6+5
zeA0_XaerJ^wPb7r&m&{Ru+y6@%OfX!(I~LGdX)e+n&m8(T6G@n8j+$di`R^T+Q47}
z0bdU8uWS$v&jKdSn(7+$HKV9@N?@b2u)Y6RK))5D0(ssK01N`l@vSQf0)~+FKJ~9PbQ0
zIoPl8MWbN%p_u7Vw4N0JiB;1>5tW?Ea4#CgqYLdR1#Bw7DIzL2&Tw{fK6$@p6rk^H
znJw6)0~8Yg^!N<09;RrY)Lt|SECA;BD^T~;vzOwdl!7Q7Y{YJ_8HLQ2_rU8DiGkV1
zZht4mim7Q*RrWQbm}-EeXY~or&o+|*{)H{l#f$8a7mXsYLr(%IwYi05{qUqrSxVq$
zk^D8ISiQ9Eh2m0D0hHLpIB!(oF8-oXoC0@w
zW@i%?L{%<`x%a`JjID3Al6d|oKv7Bplz(WjUW2F1(scI&>Q0XSqERgH;x@2XX0g`S
z?g(qPPD{WU{Z-KnLA_98gK|2)6TAENPT`HnnLRdNGYVFTP;|sSCs3%;mYD36WP)Cm
z3u;V1KZ*$pQ*4IHr1~0`rx~3_I&k9UHKT9<&o0@WCucJO{aUS!Is`fLTaDJAKShc@
z*>$Hc-HqueSjn0w~Cr&$j^?f7#lLha;8E>&1fU
z7Gntyk`UBG4YQY4dhu}15zE_C_JAX5QO5{iV=NyZCztZ#;TX>@3b4fiHL6rF$7*s1
zR+mz{jf=xsiBMNdYbDzfhNTc;3%V(%fE8S|UBOS;fj2R`5`NHfp=j;w$wAoLRoe)G
zFpnLA^+@F_1qH}Ge__`KK77Awo8lO(0rt`mdq$~j`CBs!M5;%*YFivZ2wSf{YHbmP
z(f|=1F3pdPxN3V-e1mlXA~D%iy_7=*qY;tb>Vozx+urJ6PJ&?oSm?2|!UKW(QIfn|
z-D5eGvC3crHXE*FSXvxhw;V*UFC>V
zm@7V0^NHUA=T~YQ^>Nj=z^!@G8t~?l_YMjR0T(o2&)&Y2sPNUcy+|^O-X7-IjmF!p
zkWqw!iuI#T7=Bjf?@0PCmwY&g!Bo7aRGa;!;y$qm!e2Xmo?pizA
z7h9+A?hUKEf9D>|8Q?`aHGpo~F6B{nwb}VFOtIp4Cz4KSB`D@~)%Msu+yOh&fg5eJ
zoty>!HG|Sgzq-eY9$U&&8a)F*&JWYZJ=)oDtA5p|@{EdkS9b8(6Z;`VJ-R!hnY3b8
z_gF>X;yt;uU9gwJ_ceZj;(5N=yy{7|fmGF7!?aUHe-pT|i_Vcp*@2CAb&u7hx%65J
zo7@Q;>TcTc#;mLaew}gIU~rMaRaP8^2KI=1jec9a;kzH86x;_vxT)#jMrLNXt|!yCIR!}--+<{{H*JrC
zuZ;M0+`&*l3+GgG%i1c-2lidt(X6oXbORtyyA62ObQ~;=;yqWtIbw#xook&*XUt{V
z;zWc?WeVHfZ@$`|a}3t)f_iQWXlLyMe~b)aM(=*}$WeT@Sp&e+=k#Q5?`_M7{$kycd4svB3&Jiz||5SQ+=ixRHboK_4h
z6o|FURewGwC5*uQc|(*l-nNWBe^#}bu*c#mRs>!mL{w|+ABG9`NB%~(#m+-=cdhL#
zl$sag>2WDV8|4OP1uJw=!>hKbQs*g^e(J`QvqV2PfQhc5EnW3%)050>je&ar7=!ZT
z&q4x`WURaAic0me1E8FX4rM{i8mDP3DP@f0vTf@iyolC~P1M(6IQ7kCe~!;_mw#(-
zHYm@xwwBJ6oPK-(H3OdKVOL`bVG3Tix`7i;YBTiM;6yb>^t!IbZ#~OacLImrftQ(w
znR=gqZWJN8>RVO@=eIJw)n;Lj9uhf@MT>RdW$)UiaN0~k5ve#JpOhFX75i>?02wfCV;V>B-AXV@wKv8j2JCv(r*2;J
zR{%C(lv0|v)o7#tpVce;6~cD2w<_o
zoho1xSkY=%{o3}G*@GPpEyjutAhD8E0CwYdxQdaF45jplw{a3Es8O!Fue6Q*1n-{Z
z9&o{$08!XcVV<$!e@}rSS_6gdKCYhUf7mQjy244u#;rBbJeHPmCOvV1S-<2_B=ws{@E6&
z(Z%1i4H81_=p-HU4un=)QeB{hE5`nk$nIC$j#V38F&>%Ye=eh)lED5aTSlbu(g=Hzd9gu1t%Ok-_3WhmJlhi$as1$f(;J-tAKEUJ@4hN?XiJQCWsi5a|^yb
zKueCD26af{>UoRJhBw}i_AC@i!%mZxlNnyUG`+gNK}sU5AeJ?oOik8=@M2YY3ex4S
zZGUvrl0X!lf2Y(VdPeP&H9EYUy=&X*kV4(h+Qai+>2&Mx=8qPxSJxU|+g%f-IP{2B
za8f_J&vU5N-><&67X)xXRi3PYf@zx+94O-;oMSSs`g5;RB@*^EQQERS%5LDLhBudP
zSK~LLSoi8DksoC2vxJcp=4ZpNrMtKVgk-w4DV*FiFE63#8fZQ;+wbDdx%)G#|f>S@ET&UXws$)uGC
z9`WdKf3`7YJEf*TZdYSTz@sr~Eas|8cP;?sJw@NPE$*Ia#@s~7ohqbtD1IYdw$Pm>
ztLg50FQ)YjL}~?K3ZA5o%u5TbG}fM8wY^!1C)Dg_t)u76jG8>vJp-j*J$DF8SsOQ6
zoIKg6x2}~7IB@0GzlxE0EjU@xl-Q5SO9y@ee*t~KxBh6eG+(tum5YP&
zV{q|3k2Pbp>OHM#Aa9mlHH?M0Mwc4G0ULkG5rkiCpcZAPqAwc8#?GM3z=?uL>o#`u
zl+&>~DS{-g8b$-E1%qj+?vT!;w<^*o1rE2m^Q(rjM%JOrMj1exV0C%oYd9NbW38?i
z4dYCTI0A2PW&7R0fOxggqbmt_Al@$ggT!K1_*s-s2=9Uwp>jHoB
zj+zpiybmg$^0Wh+roU=ZuvPP1YaJ5NeQHaD<0G(W4zTL0+?O48rAsh8I?<_`I0(_A
zkanyyUNj63CPHVOj-9)yTfh1;b}gjDWV$&F#`mr$z;*eKvov!w}xvz}>m^$Q(loYaXl=EK
z)s=}S388(3-L>9M%5$*lA%(9kn1Z@$KLcQfwCS$(jww&JHRaYACxY69)S&BaV-uF!
z)~Drtdf=sbI1ylujkUCJ$?ThP*E#|+IP#pp&ncJLkBB?)K_+{hS0f{>vnhY5^LVqc
zjq$DrIL>MXYAJKXWp5)Eo>-3$TBEBWAa)%GAgxxmO}uPsEi-*#}!~H6X7RwRmfX-C6?w$@Q{zZ>c4)#c*Xsj`2ux
z7lXfJedT5AC!IwpDo~JF8&kflNAO6>*|ajeq4@mfH54o_GYHw<4FN<
z!2J@qkB^FYdwJEdZn~W8ngXAf%)|k60>>hkLB#>ve@+AYuJOsF9g5c7s0NqH-VH&KA!g^CA#idDL1n?=%iYyp
zSL?N}Rsh$_Y5W2OtJVxO*^nra(v
zSF7F)EKQ4h-2xnwhvl(=DGK5{e4a(L%by)#th3^-6$2-?V>24I-w>wk+%|4~TIYJJ
zfK<(g!YDK0EahI~u7~T%bj}XcvZ?jo_GEP&e~XJT8T~S802h;}*xg{mRjt9rQ7|Hp
ztWxc=heH5Sc(Ub<-UeShI|Yc9mjEo6x@&xRZeR-AJp)vjhc7R>X4|^TY6=X{+J~@tMQfA0L(#@=gH)Gu-O;!=I9pZf7
z(PdpcPQnsdBZ-4?t61Noz;_q5q(B5}+JtRCcy)x%6(Yg(g
zAq5t!ngGP#cXT@$8$u&`<~%5Rl$#-Vb8=xnyKbtn;SQBaC$4)o
zOD^qDzU%ISgV*c1;pcXWXpA{$6zo4)eNn&ZuJKAVPYnP>ta`!@YjBwLVl~HQjxnV#
ztTd3C;}QVi%-a
z?5{Mg4O8r^?+ki51|=vv7%f1Eh}qqJ4o_aO%IaJP0dxvhc`rOrm<;KDm!9aZ<9|#5
zXPc+!j$^y5GuZ7VSG#-}wNjq!tz`guU>a-#T@4@I>co3Dxoo^3sDTk49-U2$B#G29
zGr%2z@Ll5tEIAxxxz0z~_!JO8N^r9F?DB$J&3DeHIHq{9t#B*n44}D|sV;qYt5MJv
zF?2>0(^#b+u-s`%DJ!7J%Wv*tY=7a?%S=x3GGJK*)Cr~W;k2K*FAjHv~*BV>VK|rJHT#U
zQ%=&1`D;ugN{)|Ulx@3f+!0R^$~(5%85!u^Q?>$@yri^p*U4}%mrQ-OwmKEO;M~{F
zbDWd%sCS)=HB5J5KMK$U-F8}LW2=MZKG`|^u5lGa7CLssmTq+nEOASxq{ik}cNZKH
z)F4F+a7Hkudxs
z32?HkDCx^-vR`Q&D4DyfYn8p?QiK%&vYP4Ntca-r}gG-%ZTlJ#~3=yz#u56cYpUjD3WZ;>We2*
z246#Q(20O*yl*deS9_F^a{@W3ef3ic-r4qQ6nR2)yL{sfqNK?&;pwz_fGm0r-ulJ{_Y;W!J6^{M$e@OB4W0UPa&Gq2|k
zp_P0VOd8@P82QE`JKiwq$rk(FzhVer<8=*KPMBV{z;w@4>2scS&sPjVa$s3IFH4!P
zcx<+DI(&{$=zG0lBA3m~0VIDu2euw~Q=Rh#L!erM7oGtF;b7-e<2pP02^#2Bju#AJ
z39nhY_r_8|T7bKZO5hee=QA=MNvh`;!K~S^wsE_~WO)@VS4F^yJmyXjjeJ
z&)OoE6N38&iV^A0;QX#dD`Ztabk8EcyYF2?L!=(
z`K=aH&oRxCs&(Mddh1EIYpNo^vJsx^TcP}~G5zu5H-GcZHy_WN@%!8L`8WUDkDot$
z^TTg{{7W9l>X}(HLEwK-O4@>Wq-3!d6OMUdHE!6XqN-V)q67;A{F1pdvCUmnOR!J=7hh;A{iNuUv`Q&WEQ{DCZari
z0o>e~Qhb6=j$Vw_)wykdqw?-+Z3-~8F7^UmYazt5L?v4usqcSytChsp+H_V;I~GhW
zNgEh6Xj`R1tw*2hicN{Yim5TNVx})W)((C*vFjVY@)eu2H1>O9awahJ^j*N4tN>I=
zIT3ek*5V}$ZvC9<%D#0S?o_&42A00ES15tX-qO+&43~`sTqJCrocW=6M!^7?LGgI!R47z=<&dZ
zKwa;rZ=tMHZVD28S{E=b_KvAF-LXk&nzLqaxx4tmQt1dFtc8M$t5!a;roWoG9^aUB5*RLl#yn@m8Wrd1taT<&6TZ=IV>&KX%Qm(99aZ2)G2{lq}xSMs@T`7a&P<k4b><(nZ^sVXk2yJaAZ_j_U`CmVM{D1wBK0ZA(|EEVk0wy4w
zEDKi?<_k0dsM0V=EZ?=u)+H4lZK)!7sr4?aS9YiMMUqhOTCY2;4(AYoD4Ou{@P6Hn
z57xD;bnaR|Y*o79ky0^I60LKlgSgX9^tIMq>#Qfk)@caex~3kW%$^gV`{o^k=DXIx
zf}DT7Dy?>e)FFUrJO$sGC-EsgF|7Zqe|qvR`rQ+Gtn|_
z6{8~q{tN)!1C$a6OE>QY=hp#|+8*VO(OGIlDwSH4HFmGjQqLKj-COB$$7mdU$YzTQ
zSef}#)Pe%{;&DY9cZ{x%5XRS~B6@~1l9_*n*WPt_E`3}vYMW>OWAJ}$X}fptTLDx`
zRH|q{&FCjTAAZ0_{H34u%{LeC``$40fMeMoLIDD5mUOq;b^*JU%GRzoE_>b{i&|7o
zOq&8ovZk@%2U*+=5;@E9SB?d=^~3>T_A#@zp*8iIhk4&~*^P^jx5u&(`3u{&D^@<>
zBZ2OmV${eAF4pwQu~=xUC!ekb8$+ZrWba4&qk~m*;acqX;p%zY?sre*IeKqerS#4U
zlbsV68a}Y#gO?=I0dPp?^R`g{KoWT6P7MS;kEAk~JfQ$1`_h}5QJF3$;0`S1<(Shc
zxh9Iuq&BV?9Sa-vuoDflLUyW@LFUzP%qX_{9iwNY9={cr+|mJ;SeU`hTms3kTNCg)
zcnS!9L7cf`v>jE@=Q6Nv2W1TRc8|%vzO7tZNLaXe9H7+1h7}-ttP$3;K#R^O>6sqs
z^T+>veO~y*zkIMSkY~eR{Qlqn;vf9xhnLIK0eT0RC9B%4Y<`za)B!C7vs3XgmwnU$
ze?jz{0L};~RzCb^Hn`PYm><4uE%`!T!&65x(o7>YYaIkhYJz{kW~PT(b)Wn~eA`bo8(rS6!QJ=Fn2Q8+sbgk*Yu^zoRzdh$6C{!%Pc9qomCXCIcB%nj0D
zEz6n!D33m{z>shD8n`kNlhhQ{g&F5_oRJZE3;>Sh`N?K};YhTD2QS3E%2uwr@RgU*
z)d4Gid<&3iUnBH|=rQ18^?=m_FzK!bj3n#)V$oANGXt>2DFD#@EU+(0>-8ftS=->>twh+s
zzh~6cODS_6kzrRy#D_euQ}%y6lgAEqr;(3;>}bF+KDUkmaOuYc{gY^RNtT
z7o|#}cyv}T`M7gK)k|iN22bL~U|~t_g;+DvjL9A-+HRTk$p*dnvNHe=s16&r8yGZA0RTeDfmOK
z>56
z07s5znb9)#DWJF@FwP6uk)4v8^|ZO`)86K=QG4mqDod){w5=z`nt?|v1xWYxbK_IRPEY!!$U)0uo?qDeEskvw0i_Dz$)j;^t;mg3L
zzNnr3+`+VlKs|tAda#N$T0R5J{4C0(0UEz_FnwFp8q=`lNIcv!=1pu4wln&o9OiQe
zGgxt9U|$3K(pjLjhl)FYI}lg>Vt?u94#wue$0Iduki#?s5I~Wjh
zLOKJIC^AYc-FsmH^$9Aoyu5TUO!Qdhhz$80pwNcsFo+Mvlx|;?)BGz2^WXpRt&7eTY
zVS!D%7!C&JEIrMCP+(GO&TNd47CZZDttH9syu8sjIu-pThSqog{5LXk8neE9CyzxepW=V#?`{^XGl
z?WR*(;Wg_9Q)qSXZ1t1^s;Hi#U9yT9!#pB_9>pX|TM3j0W{2dGm)x;Ro-NVYr4krx
z463JQw!8OLa`k%0Y7h2pNux>1Gxrt4lGdoB4ea7_$*OF2z3rwHK`yhcK_Sa&QASIB
z?mJdz=J{8D9IhkKjIb>&0j05%345bnvT79erq1%HifLU_9|tSC`^*~aj#a08`|bdu
zEInHqM=+fSahn|>pO;!Ph&>3snLKf&Jl@zW>&?(^S&TJ`}y|I<3Z`S9JJUUI+w_rLq?Z~yKO
zfAj}`rBS-`oVR4Gd!Q~)EpsCvB$i+&VC(1o#)Sxz%O2?rPW4GZEYrXq2Su6cyNzr0
z=>x%kZ+7jQ<(JYXMXXzAxG3*7F5u-1NIp8~P4zNs(>rVhV-fK>p3wQi8_3^eR&?=~)s
z1fX1_+7P1GgAdir@`P*cG2U%l^E}EWSD6Ul)-(6O5Wq0tlT_bt9(QLiaDaIohZlKL
zma(_m1i(mN(0{r(E+Fm!mVsGZGwBMJ!31z}TfMmMck^GZSD{+WSs9^`QZ>xNY5C-T
zZ6_?f-?(-MwjQbEF*YcJ$v91g3Q43Rf=q(y>Maa92YbLP<
z2%RzpRF)q0CE19QJ?dHW-NwaA$&fr!O~Afs90a6U99XREE$=sv2F~Uu=HOkpKMo!0
zEDdzAtXkh~KlX}P*cC`810Yhq76FKVVwSo?(tbC5-<6uk9hk755Fi_qIyOR3Kk#Dn
z{k~VM=5YDU;A?i8X-@@^tPy3b_gi0B@7)h(t&w{mVwdb*lGQYe&i9+gQNTKCpUf}l
zb4{5B#wPoKs`(o4<|jrAs(FE~DICAfLy~?nK3P!H_1*Bq=v^9nOG)d3d0P~Jc{EL{
zMfJ4b?|UsR-LRGbAU0+-MO_Gw^c=m-^KNknVX?6f`!?T!#XhhNM{2o(d+V39;BJnq
z0y-Bg`L=-b=U%h<_L1g<+n@XW#?@uCjXA5jWgO6MWt%<~63Oy@ycTP{O|}7kT%qyy
zZpWFPqI;M==ZZpqY5=ZI=9=iU5-S_~e4tuH`n*>%gI
zW?VRX^Y!|E=Q%5HhqI^n$I6-Y7g~MSm34cu{BHPbO{_2W3dDv~#2`ze)@BB9rCQp%
z;R%FnxC*9t<>4cD?8pF+z}o5*rg*n;vA<*9*4X{6Q5&Ax%T1>M6P8Htc0O@Z#tbOWivL2~X;SCrgv9QeF?fi?unS#>0PYiFcB7wQO
zR`*N;j`iKf1tKf8EbChd4FUJD+~8QNjur3cCpd*xtXB$t!<+U77KAV~wT<514@Y}0
zn@3-SJHTK5xM7pTvbabZDwPlBko(waM9RWWJkUjWsI>(Wsb5JEyja*&DSd9*FJt
z^Q*BDh6+i)E4xp&!(IRiXUqHi`l9{#RJHm)eCk7A4^bb?xPrx0EZIDEq3Uo?wu(Nl
z7s?+Q&hsqHajA%Z2$g5DX4t!$VZN-b=NRnm=rD741*R4Flq%_4QeDi;!~-@Mg=TO
zxVMj0!Mb>VEp^xhyMkB9Nemag^O$hmGi-Q4tmU#?s%pxJ!CkPO8YXyBNL>png)_nzabZ4tKRVei@2c(^JE^XdmXs38`#tG3t3
zSkz^l>Zyr%-Ykr(>{f!Xuy5MVW9?}Q&j1G;+0vl`9d*}RV~*EV+p_~qu1HH~E;`_j
z+0eBMmHhXVo3_hJoq@=)=oG?Dg_DF;04+&&{
zd;-O%GX2q{XELK`@2h)k^z2Ly;`K%>V6Fiqg#BHqMpEC%ynmK$@U+qk{LA*o3E#OA;1Q$bZzSs|LcvI-(v!d2UB
zeudnx~P{u;z-{*5{L;k
zAg;6&5U5#Rs`+e;eL~5N%q6b?)h?6tVTG-DZByH~#_L`pFki
zzklT8W`a7>nx+Ey7{HJ*C3x*JL)t#`j#WGwH0A)@u|>g|0}#Ih7gzxp%+2mtC4oL^
z=9Wi3%IkDq4#Qvo?e{b8Slu$L&&XOdfOaM&^YDsgvJCG_y%_#7tKWb8?g#%re?NZu
z_Sp;O4X
zPT31YvsF+%#q4q8w%Khx4_*ARcFv(xwf=)sQ#t9E#3y7pibIG6wD0T^0#tMO8fK99fsob4jAwGww@
zt)9kv=Nfdehn=%_x@wyn20Jz?m4~HKx}L8cGnEw+QHr56oDN1?jr9N48*y7p6Z
zLTHM5Ch_m9hM@?>N@>kBfU#Tl0;fzZ+McHLSAC810G@U2>VEd1=g^Z45Pa1HaP+ES
z;4gok!TPigbOtPbm`hx+W@4tEpExPsuJ3>I@soe8cI@vT^auf`VI`N7=PC$ty#QO0
z8YdZgPJX7X3U;`_RP<0xr>=^U%^m;h;d5
zt?w+@4cfHl!&W0TwO(59sn)@)>$}!@b?xVz=96kI00uJH4$-ACx{bTmt@fN*mEM$Q@r
zZnBF{5g{EYlkK4;ZJjR|LQnAOJHvN@BgG0p+%SCd++>d}uNZ=rm8w2UHVo^GP4YxW
zSkLU3WaR}z2q`V?xyEL#P?|k@)&`=sR@=7h*+clp^~2|X`0j_#>+$Ogi|vmL(2b|8
zwF5_bKbb;fdnx;KIs1HQ9j|{FL)j~_xO2(q>>7J0xvZfxG?SbEuP}y14Wn98Ss=(~
zPJ0xmGDof)Z>Q$}B?h@x)(O%$PXVXXycNh}&0U#rG@`y@kgc?mb2W-~C)Hj%E8pCw
zQ+eF`E5sPvN{8)(cVRx!sJiNE?h)1dYsLWjAMo7DQ-*}^
z4<2{07!#1IwdcpwpB=*~z(`i%ZWoU_a>FLwV_4B}$uIG=ewJekbXJv|(g#91{Ki?V
zQj~L^?rDGNKm2BWeZW<9gTq$bo0g&g?nGW%JqH9@57G+_q|2
zn3wDA0U3Xfhu0rP5xDfS`Oz{!S&TCbh1hLq
zu3lE&yX4E}VR0w*J<0Z61pqNlv2y?z0j2h3^EN$!2s;?*PQAsHm1|N~CaxCy>CWw5
zLfBgdxV?iiUx=axj$f8ITZORa%iht`@>h%@7r=k4J>f46pSSd;$|jtwFBhFUUonQ7
z%BZ=(rJbG*;8kgDENLyl*Gx&T7y}mDPAv`W^_53EC&l!Ns6NI361G>2L4C~Crd8Q3
z0$ye>Y_@30*&*V*WIb#t7G_1I!+|`|)0g#c
zANz3fr&cQxq$g&Ogq_Ip0ffvf*5R2M0TjSXxS^Kn7pv)X_`~%YzA0c
zT}JW4dzK&PCHr7@OW#86+p)E8KxOQsvCl=z%n!4gqj9TBn!51?3?H7=j(K0Sya9g^
zm%>FEY)!=IG1$mShjaw)bHu>lkI!durlD)w2wXZg`fLjlIX1uHMYV#0f_mfhA&
z=i=8^YYl+(oI1-6xNnMJE8GdbeocRtE?N%o%+OYNEU&8_O~7QK1D?Aly*SV7AY2H*
zZ)Wc*I8*jduc<~^g<0PBEbnA1w9I2RQZ0Dr0`6y~3SiQ%TGll;{AI=da!~xThb4S>
zDt(@vFUG2UjOzHC_v)+00Og+Hm9-)WTl3=DJ~;tl_6)F@kJK?!uB|b%7i@nNSpTBs
zLAh8;I)GFRFg(qX(u7U%p7+gn7N636S5d5LcuhF4#Zk<*2X^^A%ih?B!9gSKIIDqJ
zl~SlLVVL+dFIvXatq&d)V{7}?i4~Sc^OOtF%EkCh7IV7iDkE>Eb?uc{ry;3Ty^vY;
z6U&-%LzHHO%)5vK1u;NVdt!f{=bxGW#PSxlquR_RGELHe4d7Gp1J+zTF%S6H_FumL
z{NdX_L$CJZ@BZF@{OT`wI5|XR6Qpx|)HX+nW+_qIB1HqJexpkJiS}EShABL=X;D@g
zZPCs$=G}7!^tx^TI3;}6YT8b-dSF-JuRWzz#fsmy?~@X$h)T1FumXQ~?xVGq12R+6
zx$ymevp%`~;`<-}X?_0m;eY<@^S@o6f5RW}-~aumzwDu8#lX%Tm{kp|v}$Xb(!pX|
z&K?MF)ZjnCZmtKPZ~#DuXT;h*%Gncq6;yo9H|hkRV7Kfx5YxB!G2w?sa7zJkWX}VG
zid%L^8&b>T>Lz$Q>cI~Bc?5nestSa4&F-Sg^_PkA0Y_P3rr{M+O9nPxK-1P6)fP{%
z3tlF*gAJ9Yl1gEhI(7}Amc`0?Z9S8f`D`*Xh45nva583h7DV_tQ@drC9nz@j+_mFn
z(}*%gvtiYXbc#TL)nzu$cP-v5C|
z-p9Xx@;&1pKl;%;YHyK|x+W0%7VwPHz;@U*_+ZoCwO;Z8#q;9A>)Z;@RKl!?ArblGmYKC#)(EjvQ}P1BxY0;(Q%cTv{fvwOc&XEyLgd;
z?GnM$IjXM5q|(6~d92Sd4T`Zx$bUL#CId`8o_I1x=_r8LH$;}6WEykZ@uf$MF8L(Q
z4H(YMW%96DZkpC*pm{*A0Fba|h=3ame)1IdEWc`cZ+3jd{|4Jr?m6JVq#)#GHd(r8
zy5%k^9^%_lx6K-i5j-g>v!6Fy)ssxG(_0^EJo40O6cW_lJ+@ENt8;z2tAD~R=8Tur
zVrQ;Qap+*4c=Vee93iOd(U^G0N7Y?+fQ;T!HiW0rZ<;>%ti4lrd=>ghK)oo
zJ(fYK0G6tazQ9GWa>aIPwSUYp!HV4LJe>KqVD!aDD?5k>dIBJkanJJ1B@e)k#MJHsKfwjO
z&FA8Ef8X~kA7+1<+gXW?=5(;3Rlrx5@N(WLMtqKCSuN6;4c?-8OMeHk7on(+1ddkw
z^JBuV9{<-%kA3t6lPsViJG4ZxZDH&=A)rJ9QU6@DJakyu05qPwVi|z}xh;eMOl5iF
z)bTl%wb#^()Lc(U#oDpc19&3#9$ECF<>k$U4Xa_bQgp9%l$
z;gdJ2fEB0X0}R8uEq_h+Ma!dAulUTNc~r{iE_`NTLL)-OwJk%~8XAjoH&jA}L69=a
zKC-*vnPv5Rmiyxl2tFc#4`Gj`Q`_mK?T$C@a&c{sBrM0a?4SpgsbzVRA;J9eH81XG
z8RdY^d$r8Y)SV?zcwY@efLQ6`*Pg`&f-XB2gBnnEujEFwg$-kRD}-bvv_v7y9UY)aWT1t&7QXZbLJkO%5;9aA3N1P2exYHyf`e(`G`MyRP%
z_lWF@NTu3To@HK}r0Qu#X8T!(6ED8l
zg$RR6UhY{AVHpqj*jUpr1k;jMf=k7pW0gJM@_)x)ya4Zh``E{MRl;t?S7RfDeJB!G
z!LS?Kx~Hw}p5=BDsQU?pgB-;JyjnUh(F3eLU4VDLw|`6t)Pk@-ThRhvu~?qw1tg(&
zgb2H5naT+VFfyOE-9jtKd}Y&|g_xAQ-?Qx8pt1{)-PXe70Z=m8N3;5jH&O?lWBH7;
z5cn##Oz(iug=(m5D2yGP&vQdyeE!=H->pyo_7DU5#ix%y{P1_*eg5?E`+s}hoxz(X
z2;aM9_J0+fLs-`g_EPbcC9OvG1ZY*7JfPXir#R*_h?VMplgt#pjv{9^dO5
zdC|`ug3-@*FdbmtQV#fr4;hbQ0K}LT&m6)(ef-wH`~2g#zwaP;F}d>#1|LTWAa?-3
zd4GC80Q_uXZ*ak7v;h~8-eXLqg(-VGA<)i4oQ@?@4qIrdzKL@#j|qtDoaxpncCmSx
z($37eGy7O-dyg?y*5#jH$`&?DXjyJT8rx@MGzifnj9?
zmiZoI8bzu;dLZ@Cvtz@}08}Lv-*cRIS%2f~BiDU0x9S*vLV)c;$zopXywz~~@|fmi
z&WZtMi}Z(z^eGhP+ihVw-)O{rc}z<2i8_G*fJwk3lceUEwNa_e{YHGcJf>CB&Iw%h
z%tuX0L%hnh`%ZnOy2iVVsq-+nIMs;l3{cTJ*J%znt~I}PY<71{?8MUmUC*>AqJK&M
zf9Bq_i>}*968uSx9f^exL2mTU6n)!>2$s?~Gg?aG@
zZhVh9OY-3dyKhAKBJnPkw1KJWsej`uQ3D1OmMn%Ed-i>i?-W{i_zhslqy2bd)CR&+YAf|vK
ztTFegM!lZcW4(AZ%_6s8XTw@~_mw~#TE(2O3qq3@kA`zUR@$wev#FNlF@IN|y^R3~
zbY44}aZYizcTCfFJ|L;ArkXqLo<}YHwWE=?tbn|<|f}LeEDd$
z>y*|?Ir-MfDCH%aUD6QRr?2+f(WnQaKr4OkrR`)Ho!x~}Hx@p3f9+`OY_-;y{tlN$
zZQyid?JhH7oL}f%@!)8F_kY2|Kj|Nzb?)&~&%en7Zv@{_tJLf)n5iCWVPFiio#hwVbu`PNCUk0}x#&o5>4?uxs>Sms*4O
zo+<)5l)>Ga4p~jl3x>c$?HKFyqlhX3uL-1xAEr$J#yddc+95bisedIPRs2X9yxlv9
zhn3vTd%kiA|I@#I|H=OhZ~pIZuxamLiccT*8@PxcetzKb|M33nZ{B}=^T}Uj=XSQ)
zhaAvbHn1zy=V2oMk}`TpS?z^G!5~&)8?(ahbu~$!THO$obL)Ob^X@rAF~sExz_glp
zYu=MdH}-i*f*<}}4S$8n+Dr^sF$0vXsd{2}n_i7sAT|Aa8f#xtH&Z5MZ+79d?Q0>3
zVizYw`u6W>teZV(2p7SDSAA$
zj9?fxA(d>32FSeVra*6+MYGQ5k6=ih5=S&I9xr1<0EiE^0H~E6m!CfZc%p;w?xP`Q
z?pPfVK){F&!hhV$61#K+tIgWRKscIxu`?1g#(JR+Tz~3)zF-}!NU3S7HWmgk>Z;NQ
z=}5U9K)zrdY~F921~<&S(>&Iy&Yb53&MoI216-k5zJz6)T5~(;>%!gmq;tr>Iq*sBIm(24oqso}q=dJ!(Y6Emwt3>2rr8o~55%>z
zj%laVI#77d-jB{EPc)swFl*I&W>#U5)c|`ZuZ%N`4#yKsm%IS)bsT9?uy*F$tfy?1
zjaN@E&oo_jTL+|~v-En(Y*@F;0d3DL5&M~@k5SIwuuCIKT}g6VV0LMp?X%buP1n5E
zp?{|8IR-G@!6h=#gQ=tfFze4WJ@*E4V>d-(y;7a0_6SgX<+cAamswq<SOHY@zHY&X2g%6
zB6+D;ba|#{nr77yA{B^n0wM?S!Frpy9B?Ffn`fF{CH9PlIge_!S1p60f1pKP!G8;$
zIoFHz+eujmkk%sb@9H3AthU}F{h6jQt!sds%276
zx;`^rGdUN28=<0#R)mdN$@^sjpLRA_h-=YpF$AhbJGSvcdagDSs?NSZ;u4
z>1uEGPxkRyB=*OTzVX;IoCRwjSJlC0QlZxH(#?$0I-YACob?n*acaw**=FFCOAIup
z*Q~LgYaNqlQg*FOE9f1eI3Z-5dNL(xOM0&L#oWW)YbgK)9eiW%*>5#+gFPNDBeZYc
z{nz*UuRr;;Z$G_x`|&?~`hVjG(7IpwoLjwR_nb^>d-AjKz_#
z?_<@TGv>?je4v^EFO*BISDFG8?FVr~SS3vThS}@vGyA;B8eV7$JZk|I2`H|zLu%=+_84W>#e92cTP>_BKh;F#x3ACb$Q^1#nCC7R3-{(r45N*;IDlgF}y
zcdY{KbD=2*su`mV9!Yfe^0m3ali9yZ?Vfb@XPP36>m}_ea1P5IEOpElyBPJZ#>1vQ
zz5lTP;Q#pU4{zT63`G0xiMJnr%L8w+`DO5=+1l7?Zta!L(BW|fq$Jn<44Ytho%gSK
zu2>16SdAAC6+Vta8Gq*)HmO&Y`Y>mAnsy4}SSF{>>KlH&Jj3Qz>WYu1Y(_IOT;T!a
zWqGlfRC~&N9&H)aHDN-hPdW6Cpfdo#8fmB5!?QZm58)`boKb2Rev
zZK`{1#enOw406Wkh?p4=AKbbA9J>B|o3T1zh!+n(wV3W_(apP(uDhuu@n4tczJ%8uv;yuQ7c(??$XzZvaAd0XYZ*Jg^>!+Ii*&SE_vV(*)ix
z->#QatSLdwGk^C{o@g2ow`b6?wZhnVc;bJ`tb8x-yFT-S+hP~8N;TLFPMDRn1)gIc
zb{yLCiKZQ|xQckG8-S4vqQUDMXiTtY)O9U1zy10jf9OzqeM{RPJ)FQ)7q~se4-rg`
zt{#g0T7Vh_ysYT6?PE6CQX9BoNHn$?*VEv4Ck4&US%1&A4@6|qq^P=BVHQ)=u`B*{
zNhMl;wtd~FvBdo*tyRiCqQ!w}j>3cL@MqZnAMyMD{q6qXAOGR~$BX+*{M5s5slx$=
z-^;Qv)ng*IcD91W4Gd0IP>;r%}qY?YA|XI*-i;6SV*r
zSuZo4Eq@htReQF51ZI5g5{}K58>JFQZadXBKtN}?FiL&>&+q@oJOA(RKYV@l!usn6
z-!h>=F>^YgIM^}H#KXaRTUJ!gzSc8s57OR*$S)Q^7*VltH_1GZy57t4y2v&>HC#ky
z8+23G4L`l+?hR0&VmSUx+n_mA8rsd8Es&w&Y=31f!Jk#qQ@^>M;fPjs>quvS(wYHn
z4t5Y9(Pn$5ZL7^Ou>ztCtTEV5WWkvgYQt}E?PuC%UVpYTT1*zc$>@hyP0Qd$$Fb|a
z`MKc{j1LRI`v{OY$n{dS}{`-w7lNT10HZ-$9vK{G|QoTm>U->L>C)~`**U=SC0
z!+%mD4&d`_&mX`!vrjLa&YEXuQ{jnS+>=p1zdU^a`1om>gR5zr#18Sodt!DuwB(`B
zAHd$s5-q@*RH9C^6|fj9TCfx}*XIv_wBZ!3wH*MHBT-A+9k{pe;g+60fVNby2-;lP
zXT~sAs&#?FEyU~SFBrf+xv*(a98fvWEq@=k2KYU9Hqf1?51`L#fF`XeMM}r&mJScp
z60cbMo!*mI4q%LA-MtK1O!soE!wS5Vj|CipjHeF(sKGMHmmbz`P$yk;P!oSSe)aqT
zNSw?NE`cx<#Iva1P
z5GJHNKNxZG16(pwJl(XhqVySAv%neT{Cqz^6+^Io1mR{0h+a-fUGeRwQSE(|NiaI+$z0&umAd+o^@+sJb&{VOlCzS?dz~P9z6Zjoe}S?UTyAR?F{CvIId%c8$|}T8IhFOb49w-hQ#U1&2>XkxrHO6g=8w=DC)2W&#+x+J79&wEA~9
zJdRfku~{dUs*ExA9Sy_JH@699Byt&NuU6}sXNFIoq1MYCDe%uX2iyXF?4T4BM+sxH
zg!n1g47Kcj_1t0GHQnsmi5(Gbtkt=Acq2}Rm9I8u83<*1r3@K)eiV%0Q8gAc#l
z9J?+Y0VmqL`X8*fSvxHdw|`pIE;a|oF?_7XgB-_KJ$sY@N2s~Cc`pQgzB%LVrp%li
zn_=?cosKOEATWHX$kpc9GQy*o*Tm-bc7VB7f*USRf#)0Zab
zMa8zi+GjE$WDKtXAj>Au>^QS|E@$CBD*e``2key!vLC`zJC+b2{2n8^qK#>
z&s+c9hxh-F>!&yG-(7XU^EZ#ZodC21h~RKpvI{X_#e}7fXv$I;eV*k$5d;dhRe}FE
zo~waw7j=mRUp@VamQ@i@dzWSt5w+!czyQ}VHpq&;o@ZG%z=<~O1Sw+=Af&au%m}#D
zlCHa^{(H+_8)9FvT7MiHh0VSXsJ8A>fMWDc8?4W62&k6@xeNzahPY}i;b6V}svAXIaW)X#3!E!aW0R-$E
zqvXJA*C$zpF+ekbCWHl^O=Y;%X8X*IZ~v5~LaRPo?iG=q;D4eftkUF>Z(<&2jwe~Y
ziB+UMg5Eo0dulHn2X%&PT$e5L*H%Gvs{_YKz#!9A>XNDl^D7i`dD3cvn@w!RgZ)Bn
z0%y`($6^ASi;#Ss)mxZ+F#g@1@t^jG4{z=`lHZVS`Bx9SrAWbr3`8Fw}q>F(tq}wG;*Kn2sf>F8}KQVXUUJ}sywsDmi0MjicPsUY-U#a#ntv$T-L;~c>6ACrP~r_UUegOsOxHbF>0)~5P3#x
zyL71?h`r>J!Y$+?dq2~E>^E+kKR1NRscN2!r|W`%e}CBdD)j?+97$I_*BXP{JdL8-
zhx#wf2+s+$ZEaqvJ$fSZH^xd&RWc%RU<;p=a%08vu+vKi#Jbj4%N{8Z{;&vouncRC
zObrP3e*q@A)Yu$AYSPpeg~cSZhqDalffHE8y7tXBeV!f>2?3s>tVCgbFiT`}Th}fS
z7Lb(#ynhy87bA1;bjPkb3$b1wJ&^o6V{y~w+W@|hTmo|_p$jp-ZdMB4IVSDA@osbfzqSt^a^R(ia(Ht^_sBFw+t37HXHm6mks8G*N%
z+kcg%MCqQ?$FDc@oo3PRKkZK+FAz8S@#Aj0WOj*RSrKXX8DM~0Fy<fO6%bfw6oz0o
zH|)|0Vpbw-TyL=fl4p24Okvu}Y#UZrmw%B!2}{x^%wu0iS$5dl$vS)2rdT^RsnEEs
zpX)8I%C=eT@*ePoJ~sZv?=no%PI*3I9@zr`*s~R_t_qmp9r!8yfZr`mPngF&Y!zis
zslu3;t!=NRV>Yu`9J}74&$G_))sBE*@xok77lGOSgPL0SjM>pMVU7pa8tQrUJ6kPT#K`%vdvlQR>-?(hH|_@Z1eI
zR1-i-d+lJnbWi~4x7!WCb=_=q3%~DY-%~q&)?jifqkF)I%&I!fiuPud;AZk-eeqzp
zxdmJ-FH^al2*;k8>eK~fZM}9dJb#v+Ey2&;a>#%v*my43(CnN&UOO103`=Jl;sgli
z8H70yK&dA`i_Mn~rZ9H_gRnAu0aK*j+rZMTx>CD$jr+90Y~ZN66J}(K_sTpkz)#>B
zVKrBJ>0m1MHpM53`VIN!q~OE}VG|VV{6c4}%Lh}eCfU*0R5)sD?5T2mtAF7;-OD9-
z+F*JEizx?g<&eg&sja#n0jc)>;x$tT@L=6Vt(=%Pq#n4LyuiWh?(*8L4fI`0%275x
z$#B5(saZ+WHNy|THa1m&ogs($f==D_Y1;wCs8%@upvy}K!>*BQHNd0*e?ahFNx|?o
zrAt>xeeGZX4Es__QlFz9TYrG|0oP`l0C2}&JD5%PVg=h6J{Zv`O2$yU^Q12qeeqyA
z+{~zb>#Zmc(aG7^)5dD1>$QHSu~Y0wa5Yu1u_(`Tj;%bz+PEj&{w5`Ua)EtrLL$Pmvp=q&rEMwxZcvss^@V}Qut|uiJ7EFSVAAbFoJ9*h(%ihrq^1MDEe3%-s_rX%`Dubn3886V&PcrrB$%ea8O?4}d(Q@6gB
z_{&%}D4En&F;B@TbiS~j13clL#{1I-bLz(2h>w+7ql4ALCA0=JpK82zYYPCT4Nbz#
z-^%^dnuvjZ_rWH%uN{mtX?+xDl|jP_%8Cs&5wh{Ws;`Yr5`X=u7CQ!CQ*zZ5!@Xf(
zeT3K7?xqDa7kk$QYEqZtcd~Sq8*{RLUb|*i2jboPWVqs`j=5+H1qr
zoCf-qqVz$jT4^O~Uaa(y_Ts_B2F(u2V2r+!ltSK;*wdHQudf{pH6i9H+hc5Iq*D^6
z^adIi6drr+V1KZQv&E5I(?O(0gxm`M0-iZ|u>R7)fF7`UiRP5pOv!qu>@fIS_JP0W
ziw8pu59(Y*%r)0(b@oJET;^)F*Dl}<4j|yNyUbZ`J&`Tq)^n%NDSlo%7-4f*^^!f;
z-ulM+0pCvpsHu{V_Lbv+q`+Vp)^5xhRT35wOMeVsI~cdIV|fB`0tj!(T59Wn
z>P|MDe=Tn!q@$E`TWzn_6|32z`N7%&-yW|$5>c@JU=pF+8vMfZrdz`@_`Cvy{My}Q
zeX{_fEMw-E*2WfI3*aWh30r?{I50`x5S2zcy9niV5VjZ3jws2iuie@hNylmh8qd;B
z=~0!E&41YBl-WzKohGnsri|wl!n<=lzK=+0&6Dr)V%P~*@Ld`Ri1lHQKdzwOn5%el
z>aV4@E<7l@iM}Q#qa1eXUD(!nF-h~aXL=4m6VCEuMC-5$NMQpJd*&&-zxW(D(kfGX
zuWQaRwp0Y`+EZ%q%DJ4^dSD(D^-gua%r3$`fPZ3IsO2If*B8Ui;MEBzwMS{=Oy*3~
zGpO!=5a$qC(tL_hlU`$%z*~k%Mle4+k6IW>!Tk19jQR%ffv}Ns=4Md{_aHCiQDYFDvIs2(L
zYZ0$$U$)WzWxsu6n?^5nT7UD{n^fn*27j6-VJ~4`vD5Y#4kX&CfZ^?VmWv-f%SK%T
zY!-{FQ-mcgAd9gim*-jTr31O>s-Z3MXl6vPbS5dKh0_x)k41h+Mb0!a1|`Gisk{PB
zmOaPwEYC2uN4xPtRxA{NrMd8DXH@_tf1c&tG~Wu$b*u^`qRqh7=VHPp3#6ZC*?-RI
z)Y&5w9m|~sO4&~uyKI1=^E}HAQKiHj*3V3{=Xq?F+VS
zbuiLcrSM`})=0WG_;P8drDxlBaDU;l6W-M}lU%T%<@AKT1Bftw;f(YG`;_OaIgYTu
zyaWbvCW~p>NeCG61pA1$??3$WHT{vke;8w3X^1Td+AJNN-Igu(_uxrGwRSq7>QiGUny>8Ya(|^))u-Q6@z!btJFwFxKk4+}=E;ViN=r
zdJQJyR!_ptkTnp2uC~`ZJJ8czS8yQn(>uUo9-Y`)Uj98tIW-xggmteq0}@a8x1bB(ybQnGUhlQn
zvRWj=CcuMvG;XY8UAtZH&juWIz>f$%CKk^Gu}W~;g^C6Fo|f?=_TVq4cFJZIaG1kV
zSxn|yy`+=!<@!p(CbSm8TM2cL*u=Z%SKDLlR6m07
zQQ`KmQGw@^*;Y_2gpYQ)10oGKb_=Zh?pRaNDCM%tq{P*@e1C&Z8?c$R`R4jTsKQ@{
z_htUVjLQM4^jw{JG&9-+RMj~I@T!=#f(3SYfMknI9v?Z|3S`?8Yy(@xneRG%UG9KD
zV?+QInM(kG%1U%wHP_Vk-XmYWVHQ^7iePIb+mcqdlC{(d6hbwa%K`Yxn81^<^x5r1
z*o%zd@Ba0GPM_tT2kE^!dJLrGF9$Vl_UJghO8iq5&!fAdkePa8#bx+f(Co
z#b|~7yjr4&Y?7>?Z6Lsx`_B!p9itb^U=>69AR1?Lcz;SB>&#EEcO(sfi-Jf2Zx)&{
zp>>p|OxwQN9@}oqv?^j>PdJ_8P0O0lJq5;`tbGGrN-9TMW;a#|PU~lwwjzO35j#R+=B_DpJqt
z*V}V%q64#~ncadsFSA*&<$+kw^#B|TV^AegdBk2kS%=rGHkT*(fOWb5bj=h{$8HcN
zrHnL_-j4YfoJ_RK-w(m=R0mz79JWR_0e0FWrhh2l#hruv-)!%H`I|rd@zdq~Bffj=
z?HGINJh$G0J@V)cUksB~x-DU{&+{y2;UQ?t2O9vmdQ)I&-44?_dkWR(S?<%>Imi&#
zRS?qQ+CUTGNOvE+6M3HHVRSPJrPWoG86+Jz`+sZH*&!t-WkoOWL1knKxQ7!2Ge!EG?Zt`*D+uoA_Zo
z&$7a!@@ND5YXd66EC7e6vw0!`@9;d!jGDlx9cvo|9URA;#5~uKg23s|vmAIksz4@S
z^3mohCcTX<%LKH%U$CG5`1Y%>wtjl^r+?ce^;ciTx9@JLG#|f`Ki*3Ctt?iDpA;G<
zt4%`Z66EUbPSw1AYx_TZc=wKf|D?S?{NaQDVSn=PZNN#Sexq?M^%Wcb>SQl=u$#u6
zTcewepLO~_VQDZK%>Z@MM&KSG##04;qPqIFaiyj1U~Fm629!gY%P!$JFR?Ok$bY)d
zlANJBsFsG^p`dMS6v{*SH
z7OPX3W(99Y_e(jNNEksKUj9b(zcV~X0Y_4U1uq57Xu;HxpSs!9va(!e)oc-~8BRmK0m>IdWo`d*9IU(pP+_=+*q7K>L>?4GG5Ict}_6{~FO*YiupFu}qf3Lv5`_$(nriNPWO6+zWhX={xRd4P0wtm)d!abQ6CDUSF)tu77y}uai{P
z_KGpw#0;2GB?eEMdrt!!tyDh_N}co-W9TO};LZlBOZ^n;xI{aNrK~*SmsgBIS{uD{PEi{em$_jt&nR6<{AhH3s$`
zHeg+#sbE-FcsaqYO{Q}JWq&%QXzQqB0V4olV@r<*6sS;W{x&p-##f18A0nO?-L-@=WrIF?jdN
zZsuM6oWk}6k=OxjB4Cs?uh=*TSiae$?9}!gY_yzL(E}YsL@&
z5E>4KopOSI%{`A+Tz^JbZ7=x_r>#si7wigay%l(#T@l0ZpYGc2C1dd1szgRWZFN;r
z(peB_oNZ%r-!Iu0f%o?)JKD&%7G4WLVo-IrHU`6&jG4tcw-Y;S+5A*4d@Xj0Ir~E>^4@eSaQaf>YdHfDEt^h
zPOlThTMpLOB}9xpj(OmX+2q=u;2r9vR(7O;o4*2)Mgim|vs~Cjt6HixPUVlmeBehG
z%qRWhS6_Ylp?|XOJ99N0xL}RhN>F(Ko@rLmZc9-b?4=6W)_XG
zbB4=A@L^F-*Uvg!{PWW@xF1C%S_1j*?E1T@>0f|&&iPqq_G_Nr>IBVg!wSC4VIo{+
zm+iWm!%6)dl=?MK@9ms{5Ih@j+U#YMPGXkX_^|BK9)Ei}J3U(g{3**fC9$!j5Px}%
z(V~hy_VkHwe7s~h?o_&R9qo2N2pNDe!XBTfV)>y|yR6_;md9?-zXL8EUn0Bp=oqTEk*TyqA5(iJ$RAfBxYY
z3=D~zzJKB_dw}c|*(30=vYlylug49rp4Zk;A8jWq*abSqoL$Qaw+jpaiTT{vO}b4W0+>Z
zm)vIWj~j?X^~!9c&(Vi8PC6h8n&jhw6
zFc%b-xLbAImE8kq3)c#Vaaj>x$17gm&=(o#$H
z0)JmLF~{2yD_qm)ECn4ayiC89i1f!@#VngWsqk?k9Z+eICQHlakqtO`!77@4QwxRn
zGzgdYU~Li-aVc$J=EryP926!AN$CKHJE(w-s!SIw@JHowUo=<4M?YeAYBdpn-oY3E5U^NlP7k4lY
zHsVgX;rFX`#{9D%|3s^24qp(3K>|`|?=vvC-F9RE;Kz6J8fgatfma|eX{pSWF@Ns`
zQ{5eV%da27wiLTLV<_^KgXtDXC6eif{>3-bjpwm
z46OE@R{L5BmMZ1ir$(9wh%7(uuBp$!^Iikp*l=*I=hc@^{2k-e$3GGLP)5X;!|sSj
zhj>~&N^Th-vhT;=ATpljCsOcWM}Qn#W#{nhv5MBm=O8LR&48B%UJ-cJsecQ=6p#hz
zVlCCj4MZrm7igN-j1_du@vS^}0eIQ^5H`h+=k5MJcPnFw#vqVH^k)_=Jhy<{A~56s
zu9q1qC!fZ!hDn)V59Zl#?Rg}%dV0Cc&;ZD0^Y2a>;=K!)VfO$;Zn}HP=Q2Zz2z{t?
zWbDkyqQgSE(vHSj`!OEOZ~bz!t5{$2NV}^+(i?G
zlYQ_;H1JRIgNE$SXL`bBs8|kdtyWj*Lj8Wc-GtZNt6yg47Gn?b4!qa~)CtfT>>zm1
zD6y`xlTWQRv%3>KWl-7z`>GzMsIJEy~s)?-AGhFz+7nN!QEg>~y!
z+3Dbbpy=R(ol2Nqntw(~Uf@u{s(oK&2QFLT%0Sf@*cv_}Q^j*Glk~bj?BM=l2f?%t
ze7%97$1?n$u~}2NHt(gjtL#h)+e{#kC7Fex>=uZXuCOol$YuLyQnzP&hc_E*p5>$T
zJO#X~I
ze*>|rHzjC=V@RIaUH%w#yDhtmUuB1t$fHm^c!g-e;%l3w^1R6gR=g~pt_hsJK<%|t
z0;CvQm?%6E00J|mF0+I6%vOpFbWE@t)L~DF{blibcG>*w)@~D?5B0sOt5ws%b{DO5
zJ>xn%pg96d>3_JD?fhDeV3DN3^)mmqrvh1F-}DxT-P!W(ItM;(jvi@VX6MX0x+zGf
z8>1h31_g?yc6<^rO6+K$*y{#z;>bc|JwQX{NWh=Z+R8~-`un7qP0J{XTzZXKto$U~;(J6|VyHeL%&rr~zZ+viG4lVFjSG>gkqCUlCxt
z=d{T%&%VqKm}5f>V!h~EutMy*OC&%_wl)1sJAYFmICFnrm=;&?6=g%F1n1p+rH$Zg
z`SWrn{>FwFi-#)aL?tm-E-GSQK_Iscukd*}C2v1|`ucm)H5K$B5W0B8NhVu
ztbIN+5$P|FpS`XzPclt5G$wBlBP$r!$yCsI@bS^xP^lTyiz)UtW6oVQLNEa(mCt(o
zcj;T-?>~R=r(MfPdn-HHwT@5Yi)fMUwjU|{`>~lKxjB;oMg@{$+muw``AraXMUr1-
zW|zjp!V%FL!bx~8y|6tM?_BMww-H1T$>rGS2w_oqM0!9<@FAtgJQtZ+`D|?K*)cJ`
zJ5|E6uP)5CoMT<}(H17*fEzeq&k<$sD*QzG+VFql=PEOBz}hxGoUgKO+Od?xPJQA%
zt>vrCoB=S|6>p>|fxH7juhXZo>oq)PTxA9%5uk0H#5%b?B`sj6GC@S~qh7Vr6t0}=
zf_;Iq4a`u;(G*F{J>ZV7G7}0o!PXa);?SNAz_`g~M?h^qmo@#H%s8|EXUi;g2bQG#
zWr2TZ+p35#{&VSrzsU>`g{_e_TJ4r9CdZhL*9&_i%YI~ge)RwSt-sSBs@oCjqh(0~
z(1)4VmFh$|sA1b<*3a_Je{G_xt&Jc6G_hx`hc-c1njYgB#ks%)o_#>20gO|nWNz6C
z1eQ`kh#7nA3rtKk;HrDVq00GxNjP)s*g3xTdelRtCu6cn8z@Mi3
z6nK!FN{#6@0lY)HgZk=cPWLza!w>1@fS@3f>W!^pI^b!`@L)DZJFDdSh@=0p8RmbM
zo8-Q&4HO%)(9w($1JypW5`G)I5rE#g=RQ~EC0IF)Hw+6K6=T=h0x@CvQ3NR$$o
zV+z(+&dhw@ZtJR**0DXUX=M|7yIpl1@FPGLEW*)NyU5Hfdt)s4u(d{WDU)
zOXRDz=h&2A-ddORSynhtYHQUjvDtg!A~WpP;jgK`7hy#A}XmPe|#vn*p4lN_-xvWo2XPbW8W_tZf`E
zudB=iSlBEbm?<;m7NO@e@Rf?YVBk9Cu-`TZ;IxXZyRMp>qwJa%Islp?*}n0lXby~%$~@^JGYDT6r1}N1Y))|STcM|H
z(=uTFA~Oas6IFpN8+8nP127lG$14aCwO{20ob;Fp*p47Q2H3s=wH<$5g(-MduQD@+
z8b~F!=LQ9sGev71*dV#IbspI_pZ0m{UwyT|`R47veGj8r2b!7r00n@{GCQ8amd7g4
z8o>3X#;CqXT-&2Au-ROJsJiNCwOaGkE;Sb5P0UBhh1EwaGR7L82Y?qqEA2=8!lyT1
z|FDl*n|sC#p{R^4YxsYXn-U&cv2j1YQ2o>UZ@&8KU%y9fpysv3Sz`gVsbZ6Ys_W=I
z0lTk9*4>ZIuyAnHUDjy|W~;|RaUIbIfDu$Tf2Nst`)kMYR{taBhJt?j&9}dqGX2y;
zZmYKfl9#Ydn!FbLzvDOTP0stkb(J0AA-e&^5Nz-mruPcIT(W;O%Sj%#vy->o2OF%^
zGfok@=g4)Dk(Ue-+gX3|IuYJWy}-aZyjS%~;Svy5{Fr!d;hEROk5BTjA;q{8FyB;r12^tU
zYyTDltK33BmqDG$O(DHV$?|_>qiNImxB)EFIoUD?uKx(;
z0}QTC_}G1GYVojv(*zGsC$o-J};43M@ai*>wDxJN*gJ_L*=
zcYv)sVpuO3*+}fYFt-+f7}6dGvP4YHfNFE!DM5P8$TB!3wKm&nN8A9nG;Bu^vAyz_
z{AP6=^)|eKvGn5mltg-(w6i}}{emN-X1*g5TO-K0Nj^%!Pa7U>-;suT$;f7HTOuu<
zRV{x|Yk=Ir94KwM7cKe)M;1Il1JD3d0}69p;&U6kE!(Ks7xZTW7dukfL(Jh^$@c)Q
zS19(ux_1rtl94U#%Rvthlbg~wM<#~5of>#RYV!+@EO=M3^JtB#HHAt_OP6WJI@@>Z
zW?nKfY{=8tNyQc0bH<7U)J$1gjBc+lII@4E4WNCCZYeAoJ}5)
zVePBx?lSQ!#L0eavgh;GG{4-)fNg39?#h(C%?*TPoH8rc*=TTO^5sTGX@=cZ0QYT|
z1$dtjCE3I|QA+QEZ}#!c*Pq*-@y>tK`18B{=1=~l$9|W#)uwMOXF$-Y!)Bb?8z={J
z>Y_c*a;ZL)J&Uw1nmC)`zA$N$!_e-jyvVg?MOm9hx}RoC
zu)*CSlCSK>(iF11TF*T}n`uUvJQ;k_As=ZdB#c93m@f8E3b9I--vlYyh!kx6J0c8Pf?4u-57|PXlL3
za=pPpN$p;J*-YUSK~Bdqr6%C-(!ShaSI(G9P%<^fAZDnlWvS+KuLS=QgCf|OtJv{X
z&GDwfX0-cY>74L={U=?3kraQi&hSJTPzd;_fuM|@)_}$9-w+se=GwM%!IVl+4@eqA
zlvO!${TuEBRrr{@5v54Vi2y?#JDIoUJiIEdCRQcr?89F%R}Vm
zsW7FdXJPXNptbFbj9}{d^cjB!vDxOC#~R7r
zQusFj9JUwJ!pe>7TpDIZM(k6;$ailujWZC2=h~L>K6fM&e^Xb&zqd)sIBt?S*%lA?
z2XlqUaD|PM*JUsFONKVWAixfqY`pzV?}5U%-FOQd$B}PB-ZozuN%wX6R>9etyb~Gjnpy;)_Z3JiLwsQjU)d
z*fXv&V}Sk*tf)$>EMd1wk27*fo0l_}eKt1Y#YA-Yw28eAch)98AE+0iR6lZ^p&r>;
zk~M5}2Sh8axfPQTGWqn34awW_UVCMqgMB+*F~N$!$}3gmT(^HA0Sk$I*J<=K;&V|2
zz;jN+w%UbX@rcx9%SIC`i(O$B_f&n>;Mn}>LAlv^kdIh7riT%98ljF!Wxy~jX{Zm1kN1X%U-UJV3wo1oYujT
z+z{pPTdw=<+G$
zxh}E{1dS|gEDTICcTZr_vUO&lSgYx}1vuEgxQJJ)6yCoby=r7OBd3mPf5rj?P}Q8_
z)aNK|q$6RI4Xzfu3++1vm!I=;(|W5jWA&ZB;OuV_gmZuHHQ0{(!UZT9VUew6QudQ(
z%@o4G7R#$u=`%m$D~cd`bnOUwZcc!oDi)+;pBD?&@HvkR|R42$AY|ip?`5^0Bno
z2Dx=CnB1@`t*;D5P_K)9M>K@(g_%IU=jkbSO16ivWud2w_Vdy2RK~CzPA;S5b4pPU
z&<%g3()eZXqp7k-62b}mU+Y+jqm}@kWO6Px*mZ_7!uaS43J7|;5JFb2n@x|fw@-S?
z*~u^2+~C`LD>);Zl$%lEH2UhRf{m;-zQ~|}8L@X-I+O${0SCcDCOD_kva5WVLGk!^
z)!K6cIIk3s0lXZkZnl_jp1;VTb_#IRZbW}53*QGm6o|A*;LDA)vE)q&{)
zluG!;%9IUMTlX3}K6_A+C)*A;AZh@tZLN$3-=Y{QLQm*324zJvUbtfYp3cM})&j4#
ztICPluJg&DBe}GOPkgXGTPnuh&P*O;H{0?Qwr}7c?y87$4ZHJD3@8^LY`8(G$m@Sz
z4zR8dWnVQOYUK)`Ucrn&>|~xhdOZ)moTr`xP^j|08Q?fXcq|FRS4#nUxa{SC4=9m;
zZergu__nUe^i*>_=Q?EP)d7K;&2+K_7JnXEfsH|NU{3QzVcVr3epr_TY%tBR3vVTO
z)Y-bL;q+PWBbCCF-r1`$1->2(3U7a{LEldQr5N$;H?}|d_dop#YbJV<85kK0d#T34
zc3N{%@HfVFgx>?Rzezg1W1Z*9b~Z3w<0RmIf5u`Do;87#fd=Aj$0h){(1sLj`bWrT*ySHxJ62
zuiT&G@f>JUiBaWbcYt&ZNK=K+fz{(0Epn^|wF%pWc1$to`s`
z-jYb9fSHT$byCZfhv_F;tBrr{6ox(a@Usx@7{&2?gl*?RZ8~j7KXUF=vi{nKBckjK
zPH)L4Gg09t+olq;O!k}_4?TRAh4|dz^_hKgn4r;Nf$9THr#zqa@c;GZ(;pw%Y=8HN
zTZ+sClBMEu$Cp@a{O*MPzEZ^}_KU3SToNxt*k}b4$$0Ce7!!Qw)O>%9m0iZchh9hU
zBIlgW+Kt6F#ZtA3Tx5j`(x6Geuch_PT%rL-)v?j3#BJADIhNt5WW^|l|EFpKC1yL@
ztXchx?f%tQKT-hsL6`BXkDnzS@v~37Rp}?sD+T*f9_JBi$D&VgU*Py+WC;eu|Ig<*s9J+$#w;?ccx)7rnnH~-~9bk2X?>TD-vxULvFG9~<1@WIESjiyj(|9Q9fY{&=<4Rk*H)>0?k+Fwni=G#ItDijk^j_miTZ*u$z0QDdYYE_6
z2B>w)hS-01fLyP%WmZDV*frm@0y^id1#ncsR`PVIEqo^yb_E&Ttcc;@NntF9?v1r`
zm)N=@^!(^ax9*&n1n?iAT51Th6ID?v9L>FeJnE&!O0MzIxGr~*cybWeXI_Mb*ZscD;tf^
zI_40HQl02)s+HUhyF6WLY|{(1Jm7<{VO_8$5L7E`MB?bfK>M$ao!VMN(&^QbAIbQx
zNk%K+xL7XbQez4_Rd0Z9zWQ=_hgnW
zP^^C_i$-BW8L#=p5RNvMWA8=slZ}rZk;AazQoczjNlVxas~yhbbB!+vV9G+-P96kW
zfqh{k64Pm`Jk@xft4uJ@TIUMehs{!}^pQa|B;D!!d6{tmYTtgmjnUUx>F=z4}V&wC5V{b1se8SYJ(>8yNI0d55_lKAwNO
z)xc?i(U`Y{14FCzx)BAa@aYmFgqXxF*x+O;M`?Zl~-eD4`)F`AkO-87o31i9x7C2
zz^LW2c-)m`oVvjF^tqoMaH+BYS?vsN@?~|lso@19aeHn&r_upin;Fv!kMwg``clo$
z>buwHBAMk=rN&^MFE4VV+UkGSG#_w>bVb$L&lqQWx7w^=p#wywZU)9On2)Wt?rBV1
zX8gbmylD?g+ekw+1(L}%S~(@#bGI7)j3S(+dyS}WnaLp00&@TcQP)$Avr~Yw8iYy2
z@HCkDL-8k`6O4Q6sh1gN_q&0mo*eXe)pe^|GB$ndyz19f8eRb+ijM-~sV+!%hQ`n!da8u)WvN=VCbJ
z(I6b!W>EoSn^|>|2RNR)6*QL_Hv)7AlyIG3X8{JI%m+)Ou-)@hjazCkTe_V#`XGj^
z-d689d(qsV>)~vNAh>_v>o>|wol%*%fcRFm&()sm+dU`MM78K0lx4wEw*dgj=7Ck?
z$;K7&o2|T#uxP&YTS3I02KG45lijOF(tX(Kg&kNAK088bcWZ2daOAm%fQTWj9k$Gz
zENaRfFVB1-pn1>+d+t_4$Y%S@!WKr2atI6Ioky@Kj0f7&1Y
zAKt#Xt;etO@BM$B{h#me9L;?HbiMuC-~aSkkLU3>7JP-TV+K#@y+dH!bTBgQmRcwD}1JF{y+|Qj}e3y>kq?)zt?UmW-jYZsf7Qoz9nW{do8^Qna
z&9}e5B`yB9?@|`=NB{WN?_R3?+JcH-V}LZ7$rtnVx~PAqwro|CeQR6A?w+o%7y~{w
zz(i%~f7Z1}*lK5yvjNg-TCW&`-ZVjMs8d3{gHM$W;yN1GGf-W7#TeXx9H}Bqm{bs-Y{r>u!w-na>D*w|vM5sUQx0gAupEvV;|EziN
zeZ#<13ow6GO#^_opMm+m1Eilal2vjpF&8$&QqRFY;Dg;SQa(-Z6k2@tNHBo^@4xcj
z=AZn-`)~g3pWem$7nr!k^jYfJwxok7C0CwKvVOK(lzQiS@A6T!vPK6}2gJfI1F#rU
zE2S2ybn3UiW)!R?;2|AAMB7Eo-~loH2sLNwxO0E=bonS2*t6HFtw>L)fB%_!Pt30h7qeK9|S*bk~YcaPwnUzl&$!@oly}rw^(zF6%*aX>7>`w~+zHTI*
z2OEFHIy^B|g~ql>Q|k~J2Q+uSU?l&S|J#4rAOH9|OFUb^w^l05nu1;2&TvO%X$wd>e76UEaku7y&ec$6=l=mX|h5@v#P64l~)^*NcWR)1p2}$hO85(&k$uAE)M<
zj`oa&#^3xG-*)d)Kf>I9!pL{BsJzkUt&@K&3VE-}BF;Rjs?TTK@c;BUWF8sPYRZ5(
zwr7LC?V2O6=*hgiY#h5I+D!^0@Jg`Z6G$HYG=bk+;|b@7zkB!1e)HiUnG(4_z5no=
z&s6+}|MKIHetCgWlI)Som?gMAHgj;YM$PPz6sG|0f2nZ+c8sPJcLOT#r?7V!{w#mZ
zGS!y)ml{{cBEkovx>z}gj%@d~Yyh$YJLSwc0b;@irbUh2l(8hWn;9No`xhg4KZ6bGPZ|p@XURtCdf+E0w
zJjW5r+Ar@8slI`W@LQ+)E#|{a(sKL072!&UJ$1P8R%YPZ(F!|84>u;Z;jVYTOjkKa
zZ-Vef@o;{AS4_5+3FfZboE
z3wfTm+PfCH#ogV6YjQ^JdgGC4YI1J+FVnT(tvzg2<-NGJW3JcH3Ri>(7oojh?Tokg
ztA7O9FKPd3-g02i_MCr!tcCMJXSKqe7uz5mzLgz@R;zu7D#4aZLBEJI20DGMSpJNh*UpW>wa%}?`>q+kC9f7a4J*6$vxcXQzj
zC>7<5Nnksxo#Wx&UF2YgwBAIKy>W2k7p%g)`Rjy!LJ7GbGtGbg0ust8dmQ1Rm9tu#
zq@UnMYctr8!#_7Gv|cODvVC)v9&T)qja$6cW*6BuX$NtrU`HMzYvL8_;on?k5Uvd)
zTJ6?`Ro=cq?k%w`Yfsj{*!ytz>r`GPe}O7Wi6}2MGhA;S?e}CRFWQ8Gi`ILa9MZ7~
z_SSCoSJ7#c1qOe(Q?y1n6f-T{TV10Fcln(Nhx$0k_^ap)*E?gs#!*H<_KKU>+zfj|
zj%;%F;U+SjzlzTNid+tk8&?-Mtz;N3bk~;StSu`f`{@2UI>U8{aKyb@?{R}+xSJg^
zot1~p7XKAkzqB&4d*r;4UNwI+BCQaqk@W%+1I~Zio;$sEC}b`pX5PV2zvF
zz%h1#7df7{BPZr!-|UeqdA$8Etf^noG0k570!f@_phZj!oh*H_Ey0LHyK9cItH@#uIV`@j$;2(HelNOwHxmDyA|12ZIR8b
zj%6BNxOReFVEnc6mF$bf(G4;p+uaxr*?GRlMzDW2%xJ$_g?}lnJKMxD?Jiu|E7A{j
z9I_yl+z^b`w(8{nU!>JC)>y4)TOPT{0^v~7jV#<*_AJXXjKKLzX*J!5oG`rJ>@5S4
z^Wh_Mp%e~BY#gbDe<`0uP8jqyxRoa@z%Ja7jjeV-u1SnsWUP08C9Rv)ejj=sA~)29
zxMqKuQGm19*gJ;1H$8XtSJLWXC+fK7Y99rNtVCIbI3mZ{8trks7nJ$hC|U)n-kycqcd~N?x0V}^Tv0*p5?GK720{^)Q}C1XZN0I>{qtqt_|1z8Tdak`O1JfmEq`N_|3k_UXcGj=A!4PZo
z$PG)l6NP(H-p*S^F8F1gxt
z?hT8KKm~?_HfH4FC;^A2?s{DwJ!q|eeJ$Pxu4`K>dvCcBnWVDG9;Il58xhNnB5UEv
zsyphcC=d|*NK+o_JYVZJ={C7cQ*1I98m0C
z+_|?lp1m`0L$f3|BCCtZ-SK}L!-+zBPiCfEQaQ-!*K2o)bf@8WR9rp<@Q+WcSjiaH}9ydrk6PJPMM!4_g?N>$^uGXIQred6p?U{k&
z2Id}z9IMsdayH})E3$CbjjZrZ#q4p{4#(JJDJ(evca5DA^7d;oCWe24J#s+BlpEl0
z(;t_obGHrtDv8j~4{-=}>Ol`n+27ccHEPjnXYS42-o;7A^}?`GdHAfoe={MwwTU}s
zCW$r4DQ5eKtfp~xrep1m&`})THeSdsO%D$bVVdMc0Cu3~`^x4Q(?((Vkqs
zwBFn6z``LHZaXGl$j*QBHurq($$fXtq4y8PvgqZG2-8J7*TL??-gsWL+Bo~wZs)oy
z+Y8rTxVV5QKql^Bka{q5C1AMWli
zZtov1K76?K@9zEU!jx+wGT!i_80~Km3P}zhC1&Zd3?a
zl92zW-^RJ|_2z%-!^8b=ID3D)`SL*{ec0<)BiJi5{Ldl_a3q=$&$W*qAD+^ijR+M_
zX#^{UEVq8BY
zhs*8u+4Pu~-o0brr^KisuD$l9?l1qD#Z|iV2`uJXf#!G5ur+^A1qS+U$RW)r$F}ynJtR#cLttBvrv4(Q${e
z!MI1C1<6o8)j2rTyK-^)+zOft6?D6~s$Hj)X<|*y`I@?Qsv!j_%lvAUE31`0eW3L#
zA5y=1^Z-}1d|*Y>tC~~ea!HMo8^i0o=;6bvblR7<{>8oj@b%_{|8RYC`}N|>z&Oot
z|IUBCe(&GiUtND1`hFSceI(-`7g-j(41I0IzV_}N2Qe%kj8*p<{kOmUy#L(!5}Ids
z-VgS5G~^n;=yO;Hlspzsli-}{T^=z(DEgQoGzrsf-X^G@P;q(Q;ez;(r@sXW4)7HD
zTK2@-|FE_pf^5sUu>3y4Rcbak144r*$qaw&jetQxXZVZRx*3ofoSm*4S`+NQi|<18
zYYdwMRd;WMM`em+$(yJ|_`-jdF$Q_#zPrt|b67nIpnv=M`mS)E%Vyw;{F!4Pojm$}
z+L+ft$W`1zTF>mj*v!0!I#g!19doshp90ZdA{PqVpo
zb5mnrd+Xw_ZTF|U#lD^>(D1(p@DYE8&4Kj_8shkzBw1Cb1_Sy#bU8FRKeX2
z-_Bw-X(V6ct?B1Rj%kUk4v!{p&~T&{wU6cVa+t;H&*9^YI(C#Ik`M?gH7u4q_wwKw
z_3{C=d4|QZ_2=&^V0=_}>eNoQ1TToI(E*!IJ&#z<1zwP8zz6^OgDfzWPbYsi(XA6}ci#J?}VLk|9FfUtIq-SyDA;`^UwX-~G`TR*i9eGZCOk
z%=YF6m(Zt&i$DC^i%)7CMGr@g^5FmT^2_fZ{DR#k-GYz>AQvAd7f{_+h)5KSYidU%#OeL#B%`0jl!B9)AUV#{@Y^Gt<
zBb3xcP0M>VdhtVp()>#&{oT7TJ-v|Oj_dRB^fGoPqJcDA5ys!fF;8@#&ZBX75P!Rk
zw=9irY)}Wz4)}FLIXiz5y;(d`eLHT2COYK~8gBRzxC?}xV<%=xV@Zp&@rRAth3y^=FstJw%{eLDXihoMW
zHlJoHIb|x0VHcFYx=K!mAe>Sdh9|)h-QVl<_MvGMlTmZpWw*I`2RBd2=;>^?d8hru
z-+qoi)3rF6GLPNDHIvvK%*ZM17Kxg_?!ZRRUUvzmr>^mr-{RtF{Gx)*-(_!?MIiwk
z7&2{yw3B>e=J&QH>G3XIhr
z&k5-?>G$OeLUy~If=^qWhuN;-ihyz%hFv{Vt~exfz7km@dlZcZNrysGw&S`LHFnvn
z5LVVukRyM)7Bwu>v!Kv^L-FI&1esp@E{`~+tI`^m&{Ls-%hLRqX2(gM)#f1N@MPC>
zddR1=4LpV~Ek5QEY?>38JPC}BtgQeOxLgs1&$w@KREJojpxx$G25)Q!vIe92j~&uf
zALll&GI+zgqfOo&ed_fLda5)6w4^KZSjt>*o`QURFT
zHyt&wTA8XPFihGLF#7XJPCFNz_MXNp$u8^IaY|IkA_&8f+RuBRw!{3q11Ju!h3Q)p
z+VB=_OgnyLCt(mE(X;b~*1$y$7;ZnOP30UZ+ey#;cgPV3USq
zKJb6}Q!g27mqN~?GFLJ*k#oedEhZ(EJEoCd0aVvkbA8jSU8on_4A56CIJh|#@iG2)
zKzPTQohyGG#QIzjS7U^drxnr^shzG$_FY*bGInYKZ@p4FPUZ?Q4JoRyJWi$u8~knX
zW8}JIi3YC)QYL*wt&!#OQI&)uh097~&7glN)A9;#MRhISK78;mzg*n<59E}fNiC#d
zu^?+BI2_U+4{?I_`#2HHNj{g=-v*W*$$*@mZrQBh4
ze^n*W;da$(bRFqH4Tp!R#+jr%Q@XAcl_xtFg=%aBbP4_aA-_k*6E3A3lG|dTdEz(;t9=uIoL&nm-{Zy5$3+|*)N|I8
z(!A$eI5I4kVK2I6L2kiJ3hSxdpUf~vXd<|kgXHJ6Cz=XZ^R)RT{4aw2-zu<
z9${UA*MjoKNp`_NW)Z>DWl@d_^;F^%MyEpLoA#KcExIP)0nP$V8kS3F6%K!snd@ek
z52&{j;e89oYX?@uzj2-TijIKHn;Me0qt|(HbJJsKK93tZu>UT;3$w$X1J&)Y=e0`|
z$&xowiSUIi|C-(a_jR@9xBhDEHewEIy%|KobsUDvo6?C9L|W?+L<*^s1~+&i+{lJN
zGG-4aZxC!dGJ8Nnz?Y0>2yB0^fELcXgnF1`@UTKPpeVi$GoUEq>IM`giI7WEq;|S0
z*>{S3ov@%>hTbg4*(%0SWyE{jW67V&vjgX**PyVJ39X(XyXWW9tvV*0yvM^Ozx8Ng7#qM&Xk6g?A?u;tuY8hNFMYEFZP(Kw5QI
zGEmZcADGshoIVC(zN-$(m-M1}P5=}3S|FtnoprtS!G7V~jEPXQxm%~+s3Xg4efnR)
zeI?)G)RxeV-3)7ZOg0?lQ6SnM(^kT(n<33hdcFD#nXg#=lfFFOy$epK!wZZGyTqDj
zoua_uPEaF_^MxYI>%xB}R9B4KAbCV%HndiU0P6SR_rG>K<+baZYxNNe%KfwT%cEW_
zr@iIpm<`?0AiNC{qizIPO^7N!EY~&bz&4RY@>OYS66R*1HFuQCqzE
z#qT>Gkfyq>uC1F+=I|hFYDVn5ZV-=*##4r{^Qbav>{j?zBM(A3(3U8lb_GXEV&?J-
zI64B-)t_9qR^97t4tL*AHF%4KVBw3ob#~1l$TH@2y1##{+ot-{xQVOv-}yH|BL6i5
zsfqg82SqJnq9!TeK8pE8FI!`N_!uA1jFfLQuqvNKP54Y)P|YRYVv%I2$4|^#1j?n`
zkOIn>NAw)HWE~}W_6lS$b)EV>$ZJLYL0b2JlAd2%#|
zJngp($=E45M{~r}zB=Lw?t~|2&;d_?=Q}z5$2;v~F^!)9IX_7`1LO*`Ak
zkss|e^CTumc@Wd|bC?{-F-%iWVRD2+n8u&Ml
z$m>+^TDK(+PlslgN3})dA)j{ms04`k(S!Vc03F`7;Yq+Og5|4-_T6Ut`BYAYrX)Ld
z%jf~fsy~!uN%C(XrZt9A(q!omu#zHJN}{3}p#vZyScGUt`hGadO!8BjYp(^;?1o}N
zwx55?HJ97+zj1jk$zyqUP==D?K0czmB38z7`r0DFAvF=F@AB*IH8#D6p4(`|`0wq$@ECEOxVZk`wo^|hp?+&d)^dLka^8xqH2Q~RsL@GJW;abakh#L&7bwZo
z>>)yVn8ha&I+pB9{6_ybqCW^ptuy8!PWC9%%Y;zA
znv9C1KtGFzQu4~=G%RIAa~qT-`1u5oF2|)VyK!EH(T*!e|1S}
z-fT-`7YF%xAMfhmrq+9n%D$>GQYC+Zo8`HLqN#t=6wotujIK&o1v|}2YXJX3N$s9e
z#bWZJwEr$*4R8$V2t@+)7s{rlnGK?eHMERKbck~+X`U&lrc8I&r7(x@qi>vYCEs(j2o$gRe7#xxpK+gj0~XZ03EO@n502wQ)H(3O9-
z4NUd3AYptB-DuFZ1Jg-5Jvc(0rNh<1ZB0_i#P7?tzxlfAcrqbQg(no>tXvXntK3Fv
z2^gs^xQM8qKvTkC-{~U+d%?HUt5QZqvA5LgOlvi`h
zuo~A=|B8}ecR=MY^;1)GTLH(?g4Fkd80{AjRilz2r
zkEp2H{6oBZAlrrr+31R~Rvr7zANGgfaa!GAAR5^0{6MaUo^}N)1@P|OBUu5`
zK#iN7G$zZluq4?W
zp1_wEpoL{~cobhbVKD6qU>cz^uRmI_abisx%4?B)Dap5PreJ>ti?QmkM4rmzK$2q^
z6qv|VTOACQNu!Bel-0rTIXt4sNm<_foTR{NCe0}n`k1Ud$*{sWyT9XDd3<}JbGhIP4UFn!^Aen9yOfwomhK6qiDT-MQ6Pp&XWukaURlF;J
z(qpU{IFYTa85e(K{Rs@8%>gPHA-KF;mMn5y+~I%@^QNm;dblB@@}vLznM~@I{c&ME
z(3o$vc`nu>dk)fYGwG0!_L;Z0xeI`s*NyrY%=qGZ?a|Y<^*B_w+-N
zKD6t-Ck-NsST$?&Ps}H@bm#Ft4lBpsJIJKuFCz*KXq^e|N;6hPdN?0(vlINBzOq1o
zVzXV!7bkyUd~wpxm!FSNd>Bf{m!*0BfRZ9hPcL(7nTW=ibgbij*qx@Y0aXC%+4Inh
zG`FwzR9#PkBwcbU&G9^ZIG)K2+Ejif$7wXFTO$t_WK+aLdTQo$q+iBLj}zt@SaxNj
zvnw04q)MGzrW{PIK$-#Ndr)OYq3JS0qh_w3uvC9=wSh=fF-xnccOi{ATi%#U@RA1C
zo!OEZG^w7!=@=$C1C^XDCAo>9MsDA!&V}w^u*zwZeck&SBk4~Q-*Avzd)DpO>{(tQ
zz-WYsUVvAF!?wwh5FOgqIGQk_d}3k@Brs_MAo9diuf{X!L!W%CZ5pg?=IDTn5GD?%
zu=;=E@I2iY5QSdj=rX8!V)klYR|hHU_%JF8TqC|uORdeJw0X+!s%ah?vH?}S5`ahH
z_o*;kb$4>C4Q^N)Bo}^=#U|78#}$EUwG9ce%<8zmsT(QlNJ9do>gQ39P8=UyD{%N7
zL{yploV{6;$B3mnb@f|E8;luSM>z=H)=__4+~H~bLSnCdb^F|Ih@cT40%nVLk6_xouXu+boDHb|+Z{FpwLAcY%7KcKPG$Iv2{
zj=0|SF!`{Ava#n;`uD~UctW=+L$N1h@B-|n@6zB*-H2IMBo1WqCd?rT)}I2R*vMTc
zHft`_9J{iKyvTUg8jmd10^1j5^dNtNZzgQQVY50MqG#qxwf8_o;fLCHQo68bfqjC?JRe;(ID~?8vc2!;=A_+z;cy#{;)fw`g4-hTc6}u
z&mGas_2djhV+BuHoO~xE+Gf2!W|EZ)D9i1z;d_BixzB5&cof+X%LtoQw+4S2Gi`a8
z$ZOtoj;+KSG@~QFI%Dp_Q3H;d*I^%CpcL4X=QBO*7H9{bIw&iSw~)5m?Qut5ZChkxbCYDAzg
z-Z#S#`9pd3{j|+gQmAc9G%0^Nb(=T;RijLj=QeLPO>xx=&QO@*e%rKN_EoNnKukAV
z+)Z2mGJ8X^h$HFWx1T1?{M7IN2;yRDF0EuRJ$WSGi9dStZ4HC_98bY%;`vxChP)rr
z^*B0DGB(GX1;N!U=|ix~=gkVTI82Pf5Y-cLWRz|CfP{#(HSXi>=kk9)F-JHxRci)A
zNYUwZQ02KK<}J_GC-1y_S1njouHcrud~K08o$5uO1NyJ~|L@&9CXyYdhylGa?tQlb
zkuuN`>_mc1jnUzU_0iU20N3~>27ybxbcwGPO_J~umQM0`R57)OI6*-+va?j@V)d^7
zxeW?RL^32d3C|*L@@Ri~{CG^`XcFM2ewoKCg5+tMLh@#ZMz+FL$LkM2$J{FuB*JzqI$Qk!{zGVkBv%+>uL0MJ7m3ce;<$2z8%}Pt;S8#
z*lw)GR%2^}#@AdodegeC%oyF`NbIv*6foq^o=9ga`>~Bp_
z2xH*0j%*>4>x6;^h5Ex9WwiyJ;h&O;!L`5Rt|xTC)nqBm^>s_iY-?+T$RGw2@uQ6xakC|TwE#RsK8!MMq~0XuZjZ<0PD>G3Bnp^mDUj;yTz61ia}h6k;xwmU2y8TMZTxWzfSucpo>`
zkvXa4$(MSB!6p`&ghq7vmx&{|DoX62CYf{u6FJ8bZ5b+(OJfhoP^&i?_CDhTR~_tJ
zUT~iYI>KTBrh8S1Kb)qn>(b3D@axF6Q*n036
zt=uKW*KG5p2Vs(-a-hIkgE;1k1a8`#$(YFyo^h7p?^ii<+{%~$RT>jDrFyE|r_6PA
zh8;3ZWEA=Mno@A}79m>{Q!EpX^KHA@@wG%xLEgp#qG|dzobF!y^OW2U7}~5ubDP7*
zdeH&klmcdG(uP03z`sP1`G2P|FDwc+4&#~+p8?m*)|twi4_5&<-(W7WC%NXli;XD
zi4*dF>O;z|BJ~}y&J6n*2=m(hTz?iR9s%|x_kSb3mss$4QiiEY9aJ+%OGBR)0j4-2
z*G-d71aF)#JTy;PUW0cdDbu=q8(szD?({%@WTsr40B8LRyL*-;pF(@8@+b13o)OufXbm7UY{1O_{8dJ3PZbMQFT@d
z_Pk&MtIHO9XCx|aE59{*s(4@It>vxH^F}RUMs3bDk&l%QVEqYs{XvBNeQK>7MMp83
zun2L*xrdJxK~MSLCFVZdP_|d(0DuHegIulQ5YA}P5=k?@`FV1rBhZ&
znZkRixwyeI7>{q2bLs91il2YGhF-%Q1T8vlFw^y8%gk9{QW_{-Yw_z*O=p|eFUUt?
zTfQ>SPwfhw??kn_Yns}}#C|kI?+2=*jYq!t0n?*q8N#%<
znz#gdEL-$LqlQ=C9ZW8DEQ4hR&|QWa2I8(w7holOl+zt4EM#XWHadN*o&3NbdfDil
zn!L*Rg8*`Jl<#>;Gq&JQKsPm%yNe|q^&}cTPFK|tyFj~-+P}igcvQlGAPos~tAzux
zmwz#vNDS6?G3#-Cxc)GYUQIYNu*<-!n=bCkBoa
zWwL
zWs)e#j|=zi^~H5O?&!LTpSIb@fu0#gB2N_~LBgSNuYY8!K%0|W
zHyS5FURO3^741kJFqFdFMbH#%QG?fJimZx@5sCSD^Ibxy91a$*e>|}MVK5`D$@&%u
ztLW+nmowdyj~vlwugIl@aDuW-->EmVCjg&r>HP3r%c((OHfYRObGCNxNQAy{T;$Q(
zg*Q9xB0ehrmqm8ShtPWb+d@JRuH?0yHPz3ECT)Uz{!SkRAOT8*ddXPXj95?~_PryE
zyeq?*44Wo;SaNGXYO~Yq#X~>j`X(@)k?c%ed=r$#c1dk86^zjYs|l{nO3ekZ`tv8PRkc
zqO)L{H_`S)&j_xsDEb+a-#q8=;}4lXZ@?)W>;)>Yn5P7jY81#5zIH$Kw>uxTyDCPE
z=8Ne0tW_$}?>cq?+)-;C6PRd<+h;H$Ow^-YFooNoV@jl~m+G>EUTn&(AiIs`QEB-5|!2sF`qjCfQ&90?Oj
zP7s{Hg=QJ70OfLgnsX!C(#BXbBl}vdpHIMU6R&9_R7o}=(#KRgVFgCmxKb3-u^U}T
zVnykz&g(BKS3-F%#wkH;?#h*H1|$LTA6*Wy)(Cj?o2Ri3IfFmR-t=c$g(gKp>}DI9
zLw3LN!?#I}x@N4+Gx`cYe&xy!!AP2X-VoCOV%f#wvtzC6uDQv&%4}2ptp$mkz`w;R
zAeSCcjYC=z$O&_8o*E-_3xQPLpxVOMw~c&}3q_&&B!7y8il_USyev5MJ43jws6iFa
zZkzEQyM%^$kbLFpyX0z$r#k)#g8oULAmk~BshtG&`Z3M`iC>=cS~Oen>)g!ta0>Xbu+pzGq&AbmZ}`L%`ZgQ
zi^s57Fkj^UY!b`m5WZ=3E?BiB`P!ync!sTQj(SDz8_z*L34L*=RsnD`s_
z7j0v>q<|1cg1cQ2UPkL+$Up+a%WE3JG1We8wAX7cP-!TpLU#4>qJqr1G(|Z*J7`5p
z#JslW53+k&T~dFkbjTy?E9~b7Ek!i}DIU_Vxn6f_Z{gG3`M34Y)P}oKsqEhC@MqGe
zbRFWjFTuDaRLg9^(MB8n2p8phfD232jK%VNbE0KCj&kzaw#9ij){S@40El@ME
zX5T`oWlE+r8a7>@+Y+lrO(_z1b=T-s1mv0S!OEnD0m
zSzw9(5}TO3Jh9f?-xv8C_XaxzxP^Uy>zVr}yAvF$(^_WEq3YTOYF!CVxDgXJ3moUy
zrg@=$T4y25{IgC|T1?6liNTnv8Kv;=h(T`q!v*@i-YmICF&_kee|HvlWeHjOTQ{G3
zs@d9Z=7|xFz8P^w2tF5_D~%DqLq|?>BCTB`N0ya&24t8#W*X-l{>l{r
z?|)b|=wW?*X@9;ClxYkM+%J&DnY;|#7z8)mUz!UO%w;BkPCedcK|+D!kt0|dP?qPU
zK~hx+0UzP8THAJg64UV!Al-8DU@qFUm~AiX@Oy*g3DQ|BwF5!3ioCHR^M79#*1Vy@b-kA?d*CFi&;(KlhIA0&_me6JoYH?q#Y)vfx^;|(^2
zXus(}v-0@jdfK?{7gs@Rl+wykZVt-Cwfk#KqAUyo*RTs<*E41aOR}xVrs=ViYCdda
zyMGa;7j=}MqTna2@9WE;9U$5kvo=A@MGVUUb)w6*#_J%axzI
z8=94)Bp<0b9eTc|3BwRjgQ|9Q1#5&mr;OvJ+>qU`b}2_Las#32(D>$Lj0_+6T+)nq
z`2Q3EL!XGuK_+B!$CpUsFw6|!^wDfZHi5i$2kPA4Z`Eav3GJja&ZM2CJ}HY_-SilK
z>U)|EbicTFvT?gmH(|st(~LbN;t7ZAGmJ^n?)IU}(s2<~EG!b^uKYT!uufHEHIcTX<#`9(85=X!s&b@kePc8#eaXYgp
z%P8S%l32h!z>Gae)jhaHYbaveP|5(X%tG${jUt#P9gE$yHS`PQNfJ^7%^HxS9EK_a
z!8}PHK3WAI;=3rR5F3Cs8<&V!AejJzq^kEEC%7%s0-l0`Yhc!cwh5;Kin+)&@gFM(
zLTTWgQwtJ$83(T%1zVH?VOKRJq#In#II@g8@gki%U)!0w_hHZ+!4}qH91M7w#ntYK
zG5A3lQjJU4{)Yxx-{LqHaR|Cg5t_#y0@kEW=%0~%oR|A{#F@s%Qmwi_ib33U<{wxlSGJQmM3s53c=A3Qb(v=
zIM>9H{K*UGBL@~GUp=FY(X0jC04)+^a~dJT%nob*)+aJxpYVyAx$iXaHe@yck2m`AEZNxGL{CxdEbD*b~MNKg@aD%
zC{VtT$f*lE4bp*u$Lvk$@MspkX5E=VxN)-b(*e(cD~CNmXs5&@aC8DearoXo34Eeb
z2TJj?&AAAzQJBeT=MDS?Hmz?LB~Y;9h}wn6Tjaly7|rKBs@MitIH}Mvgrx9ch~$BM
z3}fb1KFh!&+0#>qaO{XmxZJAQGwGZ(K&Xmw=ShdVFc2U2Qbt@lxYT|!`53CDMV&&-
z4JeNy4gF54tdzx}3;x~x#M#u`Y*y0n_;Z2jDZMm?ix!#1=l-+5U+rBUF301A^}4?H
zsO^|shWY*IQnNM$;69oPA3Kq^fVBZ3+*){(`2f^7$Ku<0XYpoar)mg)xiIQilPZdo
zVZNeC$7_$P*GD!)S#89mM8<^2AX8t7ZVWFZ1mCncp+H;sm-DU{)4gtvh^mA`vtMbP
zuVnOy$zx@Z|A0ghFRV}(T%wBzkXBFY5|LOgh%`gJte*cIb=gTh{^L#|pH_VgkLbiZ
zz#KN_KnMKf;pPNdVG_QOrf~_}2%>s@b7Q=Yq%e(3k|fEymF)GMawxyBZuBRd^u5-6
z9}aJJrjLL)DT++EbF*}Qt1(;rKt^7-N_$zZToo)H9Z>hvkad8x+xWo31EY={yFquG
zlnPyONoA!iC1ALF)iGT=Aa3AT6-OL(bPl!m!1C2`CMCg43F|l2dR8QL{phymK*SH(
zG#O_kjK;^#RuUHXK07jry_GOs`l(IV!FpiFy7gC~2?6BhoT34}XE@@IQ`hQ0Hief5186
ziOy`PV+Scz!Wy}2CBuyTs>IAgn#z7f;2W2)`1JzyYQ0yC!0%y{+w=WbVDIlI0vhOt
zsEHckFX#owxA4#Y&pP=LdD%?L=ue-!$s5Tm0M*{-x3$fb?e~RVWX_-DHHPhtt
z0>1rvTNOR?wF30`W8(X5o0a}%8npPI>-L4dk^%|2g-|vy-I^k!X?>>Al)Ej|t;<^Q
z=+Br@KB_^b&Nu7dT==oSklDJy7w*e~&1wY2$6f0yq?a$|NA1=-`Lz%4FRuPfNjQyS
zWjgj4%e7?RPxJdyt}z~^pDY@uQu4Po512vb>i(_7S#z4FCM6Fxe%{O2IDaHIf@_JT
zK`udS1t+3RhLX*57P!a~eO~KOaP;lb6wgWZmojzhGx6
z1;8f3%{-EckS@i+q6*CYgNx{~75ZAWE5aOD`h~UxMJTr7fiVy%E~Y5%l?}jtyx%?J
z+uu%g@jEa?2x4%C%j#EKOOK`kE=_f6v7LomMp^qy3PsM?UL
z{h_0+2u;%jp-b&dM$+1sFxxeEno?HHS6>k)(RRY>>*?t=j}!Pa-W8CU2j@xGSj5`A
zzR+_LOo2V3q{I9?sEQqYF(8SBwO=Nd#DOz7yMOGHoOMuDb;P4UpgN&bslX)UUpteA
zboyrIYTxVZzylkxE`dyHO)^p{RLHy}*)^X~=29~59fOtr*P^|Rjs{Y$)j41&RnEFz
zivh6V`Ds>Bowpz-F^bSIL|724WANJuVPDI^deH+Pz>-!S0%Ks_
z@^??O2DtX}9q_&{C`eIq%`H!$s(<0oZDR9WwpF_ig|xqt#30r}<|FzQp2FNC*+*iq
zp@Ug*TXtCCOEHsg4CO`8TyQ2Mz8`6)Wb(>0*)RG%!H1LRy4zcrXQrNAdq;2ERA&Xv
zkXqlC<#=?NwYTe+o4Yue=^sZ_{E&6bVkaEvWOqx{@c`Ws_ges?MB|?0fHE{8_}C&_
z=@6~`G47LD!b!{2Gih&|EXD1Om~O^|V)X+$cKw$XENouv3eyA)dZYXJ;(UZ^2@;w;
zo`*A4j0^G~Ae9B5^DcLv6`-H-8{mHtl|MtT`@j#nMLGr}^8q<=^)XRTa
z^&lJ+ARudf&x&As^Hao6&ytAUtM#e9=AgsKQfI4lQvCp7^MWE-we@DrSkLYmYe&aU
zhez!0!GcC$z^v9Z12Fy=BfBwnks1HsT1LgtoR3hVXMpTM8XLuO@;He;`6EXut
zd>fBX=w1U0i-jDqm+W-${hkrXV*np`sf9NEC&cc&}
zLif*L9yR$2Wis;ZJ}JpI{GYI}>y$DkQX;rL6oJf*JV?DQ4cxNeJJb8EgCNab55^Mf
zGJqX$*C0!e79B4htwzq<@A$J$KskIS7{WV!Bz{7T0*9|TrG!Y;h6({UEYwt+myRZ3
zC@4CTYQ(fQUVKnRQ6fVdGg~I`4MXT7ESFMgLLUNOTPMbGnwfAIefRgKqE76>{RcU(p*Do$Om|W&!8)(S9LgODGE9*VF;WumV1-{{
zqTZLYh>?>J;utbRH_La}+HoN*myAX*du(LOw;OtA
zjuha(#^Q`09y(v^ouW!;!)YnzbbQkb1XOqLYas9c@xIgC`z;`nsUAZEU0hfgzb!#8
zR7r9hI2a-<29vNST8hD|vaAbca|V7}DHJkL@x}5UvlWhS;AdNViqNaQ$WEf|RQm~#
zCs$&Vtc!4!_;l2vQBEO^wH0k}=Uq^h#|-rEbcI+{%{h#{6b>0gU@a+Og-PqueYGLo
zez6@z=n&N08pKsGL&+Sj
zc>>QkFfG3zh=r+LoTneh;6~Xoi#yM?A^dWav|D@zCe{f^uH~2phx@pE;d@(V%#D%nh|i
zqqcr-GTf<2J5FORK6gKINyt|KfgP(oG}bzb3$ij%DQ6lk^Uw}*WfbvrfjTwHxk8-`
z)piA$ZqnOXRC`mjTAw9RQiV8cCr4Z`n
zWBue`fU7@nKsEoxfJ0LH{jf7>xSjyK+
zETo*&5uE~l=|CwZc1fw}CLXg}Vtzgr?5*?=nwI7_ZfB?aKRKYPCTtunEae_v-fxTs
z7G4q5EXb|aD_==JAXA=Q2zs7LKfp)5_)oOzp>_Wz-aX&83Dat+L<)1$HrE8PJn8X&
z)WP+F{C#*kq`4_QIZ5TJPHU8OKz#3<4_j66g^)nW==ag1<6la9XxY9Qp1STq7}xF3
zF1RqEExmuE%qs8=3ynHhOw&cD)PUDkm#18gO{TH?w7?d5ZoL&MZ4{o
z`UC-{^8?Yd1G4)&(X&&M`^B#gg~qFFo#U4}SOzfT_441pONlF66GCjkfPNESJRn|T
zW`Dz8i=R+*rP=uD-sSKXt&85JK$eZQtWHLZEpM_|d)Md)eXBj8mctRs8$0H-Cia1G
zHU<8N8_xztQ@>a4+8URewY49j{0Mv}D#isc^-vZz$QGeC@O2+J_HpDX$g^f`<-e`_
zn_v4{Z9J1Ls+{+EYvY7mLt|rhYUOyAmn)Et1;yZ^W+^P+@3n^|LpB_)0@3ZPg$)#*
z7_i8p*fwEuCvr15^r%l4U`e8_XIMKRa?WPc#tfaM84_Obi}|YY$*6;(9Io?PnP`D8
zBHVxbWyuNsh8N7FvF#uu^wx^rFGiXlkav$Sm5BBF7UlG3d?Px4&F%e%FY6t@xT55T
zeUA)#d8h~qQgsXX>G+4}cD_-~DL|oEt1uHl$gNgweyE_#4%Hv+%Kl$1MqU
zhS_PNf=Jsp%>rK;Q3su=q&YFHmh6b7}7jTJeDL?G+loH9)mxF@))_
zaoU*Hk%JpI%W}o(^C7AIj2#dXkx8I>h@+2-nA+blHounhSuS1`PSty{*}#NSrg<|)
z({@IM{QU*9cby(HpsL4JpLhdntFA7eA8xc+>ibob7L$~Z=?-^MBfLOq6!$@hbE#D=
z6IE^Ies^t!G5Z=KV&rDhqEJO=wpQG9j7o>DkI17F;HBlA0N(_9uQb|793uIU9$0v}
zy#=#vT@ut<0gzbk1U
z$<+8(cPvy4Ma?JQZm?GVsXi1#7uzo>q(Hk?Q#I)o%OGQ2r8VrOC}tJ#K?e137mJ?4
zCk%TnFAN2KM)bXz@~C-8gI$TM=%qX|X{UKoo=117D3~an
zfi|uBW%S(j~$j;ukk|+cfn%`f9$!3xN893;wZLGoLkk!
    XYrpQ*nHOXUSA`OFra6+~r
zhMZch>6HWf`hnMupKca_APxU({S0m%5-Zs__Kpn{LN)4@f}KQ_{2)QmSC_n
zG_YQ>KGrhn1XEiAh(tLYaP|p%Q0lI=lAeBZ3_5}s_*&ydDf&Xm1s;lyg37Ok+H#3y
zoz(^fs2XIA)NWA0RiR4Rq1v)(!>me
zJR1AoL1-_%H$bB3s#yvo@$y=d|3yupt-1OGpg7fJ?}v>UiByZ!YAxkWE8XF2@nO|D
zGz1WRfpDN7$q*rK!J)49Uc;S?oqF>`*O|$Uy{o&J@pXi{Q4bm4mW3+46-KE)7-O}0
zo4mj2{FCe22+zsL&6^6Tqw-i<5be*;sz~|IjOTBVwG072fXD2k*i7M)n{tEV=MxSC
z{Dc+oV(5Qr{uT3EXVCvx{PlhO{bJJDpORf#nfXZRTcJMqY0J()Yyfddp+6#2MPLlr
zoXh21pRv&kC)oE*4`#AP2XVGegj%bB>$!11XbGVMY_I6HW{Et7%eM0cW~5Dg$-zwr
z+ADufBwxaTpD0oV()Fn)i^EcgOGI%YfS4kmBM1K^BSQScSA2nLL<9`xg9|KE-i%g(
z0;>ExAFjwOx9ck69PK2E?0Jr}lXRiLUi|pWRo3z^f(aKCooP|?AKf1wS1+kApQE^f
z0zR$J4j-#oz8Zn=(O}i%@~>|M+_20L$M)ZZX2pcO1qwliLw|5J244H?1H;H|083u<
z(i4RGVGNb`Yxa1|ohX=`G<{)&#Vr%%8s{?{@!u3!%KZ4hW@EAjy#l
zzL<9917v!){Dpc`QftjYjnGmkge5651LP1@LLr~^WYY=PbynHTOMv-E(ZJln0AT2zaN*th-G;
zl)EkLM1b0wHQ2tNnIwM1ZN|1Cm)POa_Se$Iqen9VpzSaApJxd?LF3!SxQ^o37Y^u%
zd=v=2%{lC9AuJ?`10+>u;Kf>ssJF{U#0^Gt3rz@p4}-ITzBHPqQxkDO#I>-b+gE
zl5H5v;Olc0KYZG^NW7?((#xUakkql3SOl-;UXSB8i|;$W8V*AF5K#$h98;=e^Bc#@
z+oTdFjd6cDA3<75Yjx=;zs%Tzb%d&jmr{L?G&?oE#lVmz*0b_{71niNdBvRnzp
zKkB^)K6p}CEXbJw88(;s8G(M!ctLY;(D|P@+uW}jbYae3AbWqijhzjxY%O$(zU4Of|NP
z+^Po|-gZy>BNyoP9FMi>tf3IVaWOtnKS(I$8XJ^a2Dr%sesGPV-T21-(fONm{%?h-
zGw{FD3*V}pv0g(M*I!5dxgnCOcfp%+q)pW}k*Q`T+sUL&Q`a%5t|4@>sZ>V)Az}<0
z?C&EN1^qL~{Y~XbinWco+{5(&q4gc&==5oh!hB7SZk8OkxCLCPX#Nvbw{9uZ7)L96
zJHww4=%o7tdAJy|uGIVER-{TKf?XETMl+A?=}<0Ts^r7p7lhOC!de~8WF_?sXt7gR
z)O0clTAH#P3U09IiX9KFN@|x$BIU&$`Qx!`a_h)+`kX!!ihr6tBK?)lnLc*4{a&Qm
zI2td5ryrtetyV{Z#0Vc{I#DfIL2+K5%RFO!@~J;iBzvu6sBqBB#Yyy%#vP&&1P$rI
zD9EUz)fxI?-Z{9AuI1JXife>W)b`Tp!>Rd&^1~iQh@vj#X6G^u;xJG~?Bxxkqi^t2
zucZPa(h6h5xE*@JBXv-170%IOdL+Y_+mCEXP(Gpx-nV+#N^T6Ecl;-pcX<{fLT&@y
z)t3Vh`2tHaa*ycUs|Bl|XRm|)vf}>0Q`5e|E%Y~-(aus4Z_iHV?iYng#Hs!*OQ{DI
zJl*5eRnR@h7vn%1m{x7FmG>
zQP2$Zx40>YSbS;FoPWq6vmM}}DC5EKHiYv*KpTtJn8#Si6;xS|(}mVD3at)=jotYD
z2z%1o2iI~g$D29u8x6)HXpyc
zn;mTpx_@GdRakOE5lMOb!<|n^_r%Jw
zQ?sXuhatzmoFupcQy8b!z}gNV%4X=Y38*FQ=wKwuhyCjF!{o-rqa=Da&qX0vasxFV
zh=*A&$n&lIMI{R`8B=q*bpE2Q_$9gk@1e)}3p?<(O*+;c8U^=XJy{Y-fCh5~d<$)%
zDN5|2P4f`FES(ay14_9bgbn8Ng`SH>PQ$m*Rf^^gM&uq(_+l4$Xd5JektqwSW%6kj
zW0pk1W-5EGWK_Mk!!
zElfNY<7T*+lpuT$v7Qt^=j)})A8%E3$RlnTG1eFuaD|lwvb-Nsm$H+3mByBUnkD@=
z)F9v(@sN}!z|ce6ENTy+scUE0OloDs%@D-h3V*2^bZ&XhFbmNhZ!rh?V1HsA$U!%$d1086|yOKTXqS2EosA+E8)Mj6elKfkp&5w3bh@nMhXdR}WrfXa7eQyHnc6
z*ag0(c~w~TFditLS3|7&8<>8{a9u1u`A(nDhtmE!@kC>hCIp}wfSlR8kTm>I?KoqL
z{>x}{E{<>J!xgs(AA-7)jH^Hjiwd-0pBlzDfR{2w;m_CcWa8(8@I)>Y!-kk3)sW=u
z=;0C_*A|wN@L72gpwHiu)4hBq+?o!SB|2vT>+4l+x!hf8QinK)OEsZEpSXYeY*xAk
z-LyuUw9YYLiU5k+5#zN4cF0-MNzb*r@AhPTV=25(c3>M^!p=n;&XbcUQY7^Z+z1+q
z3Jw!Pu`!;mq_Wor@zl$WUu+#ThvoQX$xShTS0j>L>L^7uwerHAH7(Jzc9Pr!ZlyM7
zd}pcqlKm*jcL_mp^5t%w{qsN_1yw~}?K|~A$Q~*lCZN=lnv$s8ql;>RNk~K0axA`Y
z4);k8!%rqC*tdP$(Mrwih;730pS{dv?TNHcG}@iH3mo>{y^E{jmchS{!Qxh0T}l+W
z6@-@in4%dg9o;dSf4I{!dQ;N!8InI~4*9#LIq(~_lp;lHj|xEFi*EvJl@&l5m(P;j
zziA9vaDh1p$#FvPwMDk_KL7oo5-A!-tr87JJ3+h13IQZ7!IG~A<>T>By#>rwqXp?+
z;^(`N&w;C`QyDbY*E%^Q0;7XM`OV8BO2x#27#REaA!3bLV)!{sx;2>@{r_e${0y|h
z^_*^xEQM3KA5w~?CLsdSsVr`%wVBQ09i)v}eE~LQ=EJ-3Rgw;JNu$fa{75>FvA_~y
zP$w@o8?WvUz6@9EoM)*O{)IQ{#9Hnb%UARSm=Bl@LO=rgt^Pr0`j{_G5|v
z1pELb!s7WMr_>_5AYho=cBzs^h#
z0wjh8(nS?n{5);sv|KBGYH`TA);>c9+ox12N_=H8xgsrd^Ft<>u5<5&~~QAxQQ#Zh(5a${9hZzX;L);4-e#fi@v|2TY8i8ruhMsh+@
zUrE)|6$GrZY%Kp>P0d{IblA4SLjTeUj8Dnwy;?V#qxRGOhVF3-bp$7Wx8*Sioi|&uy)?`MifDo|rxCIM9s1{b5+eF`U>48yc%Q-em%stRlZn?93NBEHm
zc{9)Nw1bZwe2CLuvbch<0$PKwHUfP`zXjvP)Q@pfFC~L8)}yaB92-%|U89=AMaVnI
z%X(J*3)UWdV@AZE%{Pt-;L%@L@IJ-4YG8($jh0(3kjtghOsO!Yo!!gT;<$-Y)bW#T
zWM6%oY)_=egu=%e_5Y@#cu0@RhBBw7+Jh%v=to9KM13`rsZj7WuT{<~FhDMKVC^LdQ+KMEU=bPqUX!hKvZ
zCG=x#dkzEicFIoV;&Dui#WV7t5pp3k3tA$++aM|%qzFm{_d3scABXbQEkghBvCE(M
zqzDOc2Vs`~1rYd*p!POVByGer)Os8D>)X0&T;8>sg@c(5jJ_s8VZvsj4=*#f6i5?t
z@pDB?sGB5|z+TCIWDVFPZnZ=ai+!|rVOJI%GZuz(flG0*=0qU3tuxe2>x4J7ALF3I
z)_gkW^Av{oJ}xOB{&o1wcPBx1vTHU2ChRYTl+rkn2!P2EK5+jcu|{=dQq<9D@fR(!
zzxcX{gmUHpZOI=Bv4S*iO9JUf8@K*1+u96a5>nD(u}*NpuuuH{x%7>>@4YgV*pw6DzL&1{F%p#w@xs$nfT@`Jgf%m$F4+FBJdmHYRydzKN?_vwDgF4D2sKZ?!
zEj8CY7YK+PVPShCS*S*=@jEwb*Z05EnT+u6kck=JN4-i5Hw8)E$H;U;>j)@_-sWR)o=J|a#vC?%eeBtQh9L%>6x9^0kNJLrl(7OZAZbvl#r!xM=G0>g4SBarr+tDBi;jWmcC6!F&*p1soxg
z(yS?g%8_M1y`poBnVxQE@cNH&|}l+UD2k3
z*PeVQ7p=RF<$nkqW$%t9^UsN;1wY$XjZAayFNacPQJgIqY6#TxViqAa>d0paLy3}4
zs-UD)2Xmcmln?sDn;7mX-rsZq=4=S?>PJpw{rQ3o;k5q;%2O?q0TagMq+%_u5Ozw|
zOij=VzSouTNPpDVmlC!Q8YkD`jt2HVyB>)e+W%iN*t3^lV;HYYs^Lg~sc*HKjX$
    NFsYnHb>qs#RIUXSs~HoJe0a&^
zw{~$$T!>9u)mG{UZu7(S=$(ulZTCur8U}jv3p6YgVpxsbn!Z~Js~AKTGKY(CO1T-$
zIP3T?n_m((nd9;Ds?TAauQ0X8GIqOIZ9?KB7~JnXCFk=d;}f>drl#C8f%=>At-q*R=KYVYsk
zG^{meZrmGx43QN(oMO}uhseNQ(G_`ujq2A&yTE9cnzG-!g
zyxggjeePyIRPQSOHA_jHxf+2ruz+k~0rs+8Dzj)!uBQambmNg
z?Ej?_UgY3l%wLv(4yhvU(lRU4rINvDX#gKnNdNo`og&1L!Zy=UiO$Hdr{0{yW0@nL
z0TGH-2_G{^S6fvtLHk|P&k{3whjD5YjS?Aq7=O|v)uE#~A)R{N0xrN&2Wos%ka2lP
zyZ1zZZWsP57pS_(2Ri~SP>VB>4=adOqM$#S#O^SClm1;2{#>;^y|g0FMIY!R78piQwhL9zr_S9d9ePD0nc{ZeMo2SE|w~R8h{HZxPD8k<^+FASW%l;#w|#-Um|eYvRm1UMHq3-?$}GAb#^x7F_*T7nx_SqX-@Wt4X-R-;
zToU)V2{5}l9;wNHEJ22lWNgyXYImsgYdKMJu6@(0b<@%_|DM}7b+k@@6pk4A;HWhL
zde(-qc+G*zT)C8fkWgQlfeJ7~2H*X|RC-M%z`3X~LF>lQKr;DhkxUy67XKGnU+5e>
z6&y6ng@oW7c8EOtpVNfh_>Wx*&PC!_Y%?z%0B3~GUg+yZ{Bk+Jc(vQrj8?zR?@*~c
z-h~$e63hk7jFbrBzh*fL>zd?=ht`c>l9H?Vh8CoEF92(N_y2@Zu*u3IBc`{F9;SW&
z-Vp6JMyDBTaAJdr33%h0hVaAT=^)}iM4SKwX
zwl8kHx|1U|;OcsOul|7We)_6XM_JHqDAb=v$1_n-nQ8C=s-X|C0eRpAy)=rb+3;ym
zLci!;hyRjvV5m4yAfny}+dx~4A$A9h+wgFU&OKI`Tsp`5(TA?8uD2Kz->KTj9{DjK
zB5r)|zZ`F)a3S}=IIG&W{N0!H7z6=6p`-aLhYYpiqJk~yqb2fpsnCl&s(Sz+7b!P0
z^du9%4Ts1SD}Iz0@cZ%L1ye27lo~RZW<2dE10%fGyg&2li5W{*i7sdkcS{;uc8AwXmtx21-$+i
zcB{rVc!5JOf)pa^jIn_P7~w}kq5B^Uj`;!0%Tw(vKHzYAL9(<|)8B*gGqk)3thnDH
zr6I0YLJfSY|Ax}VgI-jwXH@IUDa$&pRLzRf5>b_SoGhwowUC9^+c``!FfSH}01wix
z9d03-N(g-T&bwyBur)~j72>+9#J??dTMMFW-tzL*sNs@#I|7*U)tMN~d|S^wBoi}4
z87mM8#eu#<(9#5Zae;9Ir9f2q)+78iQl0J24D9c@w*n*ibE6|{?WVx?{%W68Xn?Lp
zR$scUMt;J0i$LQDuH7W!cunyoF=X2t_vn!@<^xr_cd7|$;=n;Xg44UyFS|`^=D?qm
z$Y~~g^mDXIISgoH>^}0;Q;M@>b8r5U`MV|ZL6(AGM60v1>J;H2|OLdn#vrW`DCb>F3o?sk*$541nyShLe;i|eB_&9ZN)=aeWU>Ec5
znsSQ~Rm34tuilD|N9m#;PxGV|exF2Gqd|h~?dRgy?G!MnPfd-p;g4Nbfn<9avCfT2
zy+oZ~efYVim>kXh^c$9q@jeC1Kb8Q6MG)xiM{ma-0PyjviN*Oi7pZu(X7#MLARM|)
zNDx*jm)=l1ke2#vk;^289$x>#sANeJ2XX&I7p|1~CkH^ve)hu~+hX^$K0YyUGm9s4
z75_y-IN!w{dMU@hj$>-H&8BC0k
z&FXAh2Zy!KNw=w*$1r7js9vNcLsCik0LNi;r3;1f%U90-AGlpKg=?FiL2|PUcqz&7
zG)f-NmB*%R-{4i(fanqjeuG2=H+d@Dr>e2}xo>sW0wO{!+w`P&nSTvDsa7PgS6k?;
z2CG2!_`aez2o;OZ5%}o8fy>w6P^83x
zg|8U5iI?r)GlNKf@4IO;H+eDlvtP}|St097`{)|_{1Ez8N2gkLzWT70D(z@7$yuhDSklp%%O=rqqO`XJF}Vj(T!N=#WzcCBiZ4|3cvKlM;4Adxiy(Qf
z=JDBp)%#DIP?sPKvE4
zfw-)i3Kp_Vr?U*V3g%b0M7ol_M0Aa(yu9X&V@Zu8qN)LYZNl!p2&1}6C~m+tvErB*
zjpJZuupU#%c4IxvYe_a&VN%^%YU^tw)Zd{p4RNl;(qNLo&kJe0l6U--%MdsYa_O-&
zxB6g2%*O)``xeQ3A}|ZSG2Pbs-D(MBAR5ql88QV`(*O?_7#30ipK_fUT|DI9kWz!p
z=^C?LH)2L=$O7s~xZxRGKjaV_;1WCUeApw<9K4x4)L(eQZ;HQ8@2lzZt&(^>GDLyy
zr8N%FPVYfBo0}t@DLGqZ(nBvzoa6U=DR#dH`;vfYj-O6cmKkwB&~|DBRJeges?S*g
zU3)XC4B$uhK*si2^4Gm<1TtV@1JT(g?am$ByR+w20sG+n=Q7@=9hCs9{hbEiR$W9VtKPj5$QdPDT_!^vEJlaeai@%);AJY)9pUj4`#o+ho{&+Pn$oxun7jHR>z26%T4(?#6Tc^K5IXhc$zjt#Pxqtm(Xt4Ko1#MTwjK!_XCbzje
z#IVWI>}{4l$B}Sx6fd_RZ(ij$$=mqCmQHV-ud-pBrN!}c>GiK$GOVk=b_O+#cuxJP
z=d4kEHKAOAIuoXg5LG?vWDOXbnXaW8k0CaFjm}wcxJJc~H2dDC=jwP`Owy(%C*z;w2#2y*Vx>ae_;mFbJTgef?jpP)UVNj@g
z;_Eb8fVf0e5kuY6MDcc`sfPJ4i`$jVVFfVioN
zqOWghQeK>uo$m^Cnty4SNhA3R1_|GcR5FyD0DsL@VT>Lq17z3-59l*W=HG?Rh>%=@
z&hPV{Il8m7TxE7mD?xV!PQG5{9iXj|z{TU+#?xJU|1}JZMX6QjK$YqONb~7%+
z`MOxW=K|8-uI}z{KHXk?eenOey}9Hq
zlYh%GOcYGRqRAp%Bcqn}h>9=44jq=Q4&|&|$HV$2r8a=ZNiAtiL#39OoR62pr7dUr
zhCy)*GqqEOAlgxlrd3BVOj6YpE|y7dgMZJesh3x_d#0ydFUjCA1ZSL#`yXjpI2u(Z1YHFH}{(+RSiVO<6JF2j|Azj>>s+M+@KxY>NK1p8fvf(ps
z)gjLm2A*vujXg;mM4yEfHKr}NdLJNe`_<*?Jb!nTv1V|T1hckDIe$D9
zL-R34hv+U0l0+Z)<2W-o)zb~khg^{`8d%Br?%jv0yNmZ<{D;e%?SJ>!$}9Y|Qgb
zbXtV2&DLIV2|`i17UkPhdO#$BSIJF+-Ukbr79#u}^Z4zM06|@C$
zOO_%St8)_nc}5SXBD+#Jn0^nFKc(kG^d5bE!}J-ejVKrfh7{HBqt&Sq^JcyvmComWq1vEnzA`FCvfJOQ
zsuA~#(+qsgmsX8|ZM|;!eXpL1W4X+#-WaFp^=(Z}sC6&&0@_PlJ%170@E5N55`4dh
zf04l)xs=Adz48+`2PxMUW$Eb+lv#wQ+>L2RZuP(=oj>=H2q}-@)dCA(0Oy}sV`#A8ZBR4$;Zhf+=$1;%gug|vl%Ch7$nK7
zS!-ofs6&{0G|icbSAPL(1mkwn={{BsB{Mg}oi?dX71#!}KYFrDLl}z7AgMv3DnJBptB@y}wKiOx}xW0lJjEZ5CR-B^6CWqMU8Z1Dm05t43R5rpk?Dl=>AXMZ|>GmLZb@Y!oSL6QNm^2ixx{}+J*&mfya_0@#nnZt9>%9?DEtkGCb
zwsh7&Sq*r`tQkySMFyTJEbojgF&D8Gjh``>wiYlSV=iDVO8;UmYb`*&B{=S);4;mN
zc{{s19AAnqFwNQ8+d(nJ>S&hM0k+?|105K>(kjB>5PuFhQe~8i2Gg%E@M-?3N?jx5
z`orZXXdwiX6QYs;wc)(O)JEI3m@PFHTx8}BWEnW1ltfy0Z
z*V$DqQg>JHQ{^~JraG~$mYkVh`^7~|u*?~X)a1p9*MooiA$SEq+(-^E{jgV8&p3+w
zeII}UV1M8?#{LjS2wcFFLKgCaytQutyR2168M?8yx@ZwZ)Ed0TpSM=^<=O(G4GZu&
zcYXfUrgRSn6vzAYsS1Ds!EP;B*y8rhssi48dfhRRR~*OTsY|@qd{s1T)mO^02s|F!
z@mTb@diM#WGL|h|$K#T3BlHSFxh%n)q8HJjnSWq6W9D79Ul{7^x;0=bdesJQciiIj
zaXq4d*wu`N0tS4-n(ANk9%CjCV6v??}>o(4BMH!sS^``K6CV
zm4Co6&uCB{ZRYBHAiFToEXcwNs6Uq02E%F*hyub*Z~>~^RfWd6SI+POtX8TXLqhWA
z((2#yAObL)QwlR1j*8F4+_LQ7eYuO&-~0mKWhMqQMwDOz+bO015Tys=i~M=#qez1L4x!^G-#tMYCWR>~F!Z8$MuR$AW(CYVf1tu@
zTGP6c@-}jwc$$eNCd|5qhI9m+`Oh(oGfV<<_|9(heUF4%SJ--YM1EOodXopYmCsz%YU4F7iQaS(!e=9t{e?;y<$x%%*xRe`kO>7QJ|$g
zfvCs6zo3Kn{01pv)dnW6NYbjFo>6yAzyQl1p**93u69e!IpPdxKRMO5UZ`sn&uzmf8n03#4q8`LaR{@6AjoL8mzo6sA2Ga8gf
zi*wF`m3LL4@tEG>PMz6Xt6qTI=;p=sf}7o^LHC1ZG?pV|;j#|64!{t%gm_!xj*LW=
zJMh{v!)v>bvN~}6%L_2$qJO}Yu|5Hp@k!1GbPtPRZH}_Zmfg~q-dHIp0j}+N4k;`C
z5twm$RPa4P^U0g{hz{zI7@du{Ky`E)_4pIIz(1u0M!zoKfsH7KRB-+mJm-H&Ie~8L
zF}(H3wocQY-Cxkf&LL&Rh6?`Onvp~yzgYoc08Ks5A!VG|+WhbXcz+%$FlDU39oH83
z+lL!Cujb)K%)(`b>O2oOEg{~MM2eA#_u;1N#7j!~(~UrFJFAi@g)M8EC*KWzz_Kq2
zOc|?*Vf5jK8Ad&M^_O2ZQ2RCsEvp6`1M2}rXJ`SF23SV##hc(YtEVG0eFH3`_rufm
zAptJ?szOyTZ=>)Do_{$+g%-fUb=h4eqF9;D&?lhYc`AA0B?7<4qVGJy95xH1YiS-;$SX
z7(EJ}rO_Y~{LEub6ZA{a(w&*+IU2oH!zJWbW@o-d0VXYA5Pv4#^V_2<41v*^)tWSz
zj(va8U=M`>G5fP;m@0O>8wkalZ!J3OZG?{AfleX)HBQbDSy(Ls$l>Dy=3yn-go#)=
z)Bu@SdDN{`tQ_k6U;vWaCUJ%2mBXqNp0*qRl!A7w>5Kp9Xa8Pig|(}{w$j3K_1y=k
zIywSXx#kckw10}Emfk0PP1+5f)({btlBX<<0-!~9L#z8lt1SVd7c@H9!Ne<1&bME@brA}NIZ|y4%4`(hhV=cx
zx3URYYYuu(aS$DpP_|&9yUEHbdH9(yA!pJ+Iug$OWPcNk)FLoNtb>z3H+=2H&jiuo
z1s#5TI)PTIO@hjKp*U9+o{Vl|qL>?XF*m@)+{lZ0Z=#s@>SEpl7xP|NOqlJ$hEM*I
zu*C~JhtO_RhvBgNJtH&7stw*x8u0uvXOO#y8lGaF7*tQZmrslX>m3IYDX4r@fKm&*7J?IEQe{m}HM8PpCg!F+=H1Ed^~2-M_01p7
z@DP?%l{(_zT&?svD>UtyRg4IrQ9asS8l&iM!w0ltOHlJ2+HyFFk{&S+3E-Q)GQ
zhx^CRP{j631TT@W`M8HE^{Ndlq22DBPk+wC%^xrYp9;`51U!EF_BAt%&Hr&gq)ewu
z9#+(!*I&OmpeDayVJXUwAP3BrMF0u#+hjvd&T)YIJv^0@XA@1yuYUT{Iqm}@2
zfriwuV6lZFpcXb=UB3E-y81y7*-HtL;0_nrSiC!2`r7b>`6-M}D1jdS7c90T6#-2@X;9_FML;a*xsQR?kN|OS
zRUn~tceO2{jD=vt1_*2&5s(6G5`R}tzKsm<&lzaB1aSAqB4CBoAQw3a#4g)1l#n1j
zfAGV@574=6(x5sL;*Z`Ll6F-=Qk-rRw3E&c-FY#yEU_5)Dpen|deKWiy=W)oD?Z7S
zMWpA{$L?$B@6)ruTsraOxgsc3XIK6cPr>mxgyCK0QfCc7_H=JMSaX;s(1`9Z!a>ZHk@@Q^v$q|!1|NzVb1z
    <
z%^6TfQ8$i{@@Huvg5l
zc-&9(P9^MF1^y@LYNeaz8Gnr_mn>Dz(VHJ0LCBovNY3tVpDe>T$*QLrV4rxzUUa}|
z;cmj<6=nz`5gP*&cP{e#A5oO=+q{AdhqlefZmAOe$y1BxSnnA>w!@D&qX=j`T62Jj
z3Co&n>O_5ds%}$1@jz;ECPHu3qo_9?eB?$j){T+q)1R6{{G78Z2Y-IV$wao~F^2xF
z`wN0{Q9#K&A^Z&-y$*tIRe>o8{mkN^d?L#t0F{o>Aq*1qE7?ZjU&W3Ts7Epv!V^8N
z`Kj?yJWpy40??0c;)3X2ODP<_;x0JH>J|9=w**uO+x@B>ajfyWadb;xi!u@7&V)42(b=ogGbMq_F@v-np&
zMd2A}7%jl0YllM^@4Yk7EP_yjmCBS
zvVYAM@-Q1WC6Y9?QnSR$TsZvbY{ZtxuCGPL>b4G{g5f)o|_6SnjO%2*0C5!WAJ
zPR_bU6LCSU4+;+)O^8yh|Ho<>LpP?G8Ky##h0BWag
zSz?Jww$FZo*?;+6*Jwg6n8A2aK#5o$A~%RHVGOjLO(KdY^C$t5v}oNaYeW&{ckh_I
zfVhwq7?;V5Z?aX1cS^$)W({tWhyu#GzK>9p4I-D)N6*P^>Vx#R5{uqn5ILS__ZR%e
zHUp~g+?apuvC&=ffo^oun6Mk&8FqDF1+?PV9N@;|Z+}jiBx^$g0kS$*ZQQQHh^ZLD
zd!JH(JpcU+(;uf9j`OA2Hz9tM8+GafpNo=+l57oYG)mX5oT1p}hX0KVB7dhN%u)5K0uwQb)1P-+^cPU-{lf@5
z06>emRRyNEnv24pP`hmfl1{i6a(
zMiw7j9~~du;00y_yus76q!{6Q3ZNJMW;CXjpHln^)K+9x4KvL?61bC~63f|&tuXRU22TB7Fkh`XLePy$Mb@
zb-emn1%!ecQSC!z--aV!*JxVAJM4;p7Js}()ZssaH~m@M^s0L)*)QKe@BdES%=rbA
zQzL2(0mS!V*7}(vLSgEtNPrx%SwebX0wUnSXGUX6+0U2kM}Y11Wr@UF`zDabqw5-t
zxTN^RW%bTrm)}4L6`=wsSP&sX9xI5Th6ry#@PBrSZh?Ob
zbU3w1NJl`$A%^II6gQ*fRVW2?@T5q!MovZv71yP1$jjFBSZ_3xhrS8-Jv`EK4wc4gY|mnf(xQoU`!*Kc@sTnRSiFr^DS3h-#i=
z@oC?+axNoJLmrujNRm$8v#0pORz=MiH;%X?DtR_vLPot?D2QROWtn$
zlLpd})8fewm^OzL(>|roP=qs#C1yIJSMPu}M<}{=4JhWs&nv6C2Y<9D5?0t(a(2b*
zu1>s*0#8l|TGC|+D4~o2JIXwkcae7^!-HhFlWa+;DCP_fRVwcn0$H=yIDGUL)1U=o
z4Mh=4oKivu+?u@x)pZw16#xz+DS_)V5`18L1P!CqaU`gLk}Xmd&P}&SlEQ3E)PS8_
zRzgO2DMADTwzM^AKz~K8eTmRB8lD@Ls`3g`Sml>Gs7u~7#WuLudFdm;sycoIV*!k%o7f|`xOPJ|e20ziXm
zut{VI`vXqxX{bF2Y>|Q}`_n+_Ck-yWk-U9%9I=wR(LlvaDQIUasET(Z10*}zt6s16
zT)x;V8M33PKz~DD|2a@DQBeTN!bn=4YFVQ`*cm3@vIwT;vSfX3!n;o>6<#h)jO(5>
zL=dt87AYazV)ru;fQcSJBc!PU9*f?PGG1aV6H3M2xAV?@w
z(bGA)0sFW>RNvms5ewI^MOypyXpU5fzAO;gxfiz(T7OURdv=)VOpWh!QlqW7f3@m}
z)+$)n&PJCLuH)&51?z-#A;3DMRvCagCtU=9j;bRbw$ol+#m_ia;XWNW`|8zuL)Nu7
z$LV#`4Pu4``)Ge9`%=4km8pGmokgQh;44L9-hR^<|g=q
zV!(QmD}S!`0o>&9vcdP90_A11!J_R6kJVXvL9&=VCfhasa{g~h?UK_+FuD9g$+jFX
zen4?@D}HeY*XomoOQkB`M>Z{Jp0lW=#4qXdvv0n=N*<-0^ojgHac}_8Ra=CL(d2#f
z;e@vp^Eq;lUN-aIx`!7pK(gWWRDShbd^K);z<+gsCp;QQAwwcgTt9&N1gf#OhX?&C;xYfULuM^EI<{q~=ao4d#B-RJ#SLYFOOO|hTucVkwl
zO_Fw#ICkUP7MEQ^v>X2rBwsWjb5+r)g}J`}!HpshMe-3GvcWBaR7+uaz@L6dpX2ey
zD1U285i5sw?Q1iT?W}pLMk=@Zh4I>d*`ju|)vDLgpq5H)3G}Gr%$GtD_J8B?s@Eo2
z)itZ}LIzM>@oHwfcB;KeWFDO?nU8XaEm#u%Wq`a+rL@ywS0G57wNtVkb5
zAIhvbL*2WKJ-Vs(9vx(Xsja%6WvwTAWq%QiOVAH5_^f)=S(GBy=v4Tg2Qxj4D0j`F
z*Arbj3S$~j!|HfC5;s9+LL)Z*y5`iY!n^eRF@Gg6wt+Up+CsmEC-30;qw6iU|5$!)
z*}rE1Xo&TM-lxg9mYVmal6CkbiVx$f?~drJ60~~Ny!YK#R3EU6S>3BPyPjlf9e+Ju
zz1Hh#J(0Psk2}s5oI3MBxu6|wJ+iJD9W8PHnW7Ypw_2JsoO=3*55`luV42lgU)=aE
z4C2I1yu-~U&3oFo-Q1puG-&iA(XWnH=k{DEL0$P??)Kb-7Y`lBvyv%AtE+M}ean(h
zLs*PwGW_wFK0w2&VV}*HN8C6iNPpFp=#wtu=4kIw^Oq3^ft5^$(sXS3N-wdpkAYmSh)nLqoa^s
zm5nEK)#ldHTdvB+tDu2goPVe0{q&e2lvUNg^aYaXj
z$6|k_-I*~xDVYu(Pb$tgxh`okvS&S|i2JugTBQ%*A5f_#NB4w#`?d;aR0I
z`g}=8JoGOJwUv#xXHxR1Tl=T$CD+gRr379Pj_FI$%o=rwf8hgq4kI+S);UA1DfUws
z-GrC7u@6wKWsAHmxaeKq!{`V9`9=}@3O`Mk++!v|92vva)qjLfpMSE#b8a$rW?Gh5
z3Pyf0+bz(UpI3Nso{6@w9g4FLw@DfYeae9y1M!+e7Zg!z2L+j{d?4SgXk!!V<6Ufj-}uWQr|Dk;ZF2^
zu6uE3G^#>Elz$Pb)PJiAPQ{uAP=+N
zp+uQdcm?wuw#A+O)9@4*){Mpyvs_&j3(4mYl5bUk>JT#$1{j-2MI*;+Z;5ji9q>=b
zU&9(*BxhNo>wi^_-$d#ISF$L;^@QX*%El^XRl&=|aC!)wxwJZsB2*2b3o1M|TR9yb
z^AT#WdF>A`=$QIL!w9ogG^Ma*O$8J?n-L(7pUp^s2xrakcMeEH$VtwZkJ$nNwnbJI
zsv~FKM+aEqWeFc9L3XuDeksGJtO)#nT_b?K;~hsYyVp>>(T8U;YP5t^}wn;3!OySP!691yI&XEZKN
ze*gIA^?%oIPX9dWs|tdeJ*Z5O&@oAvOp`)lo(|yiN=6E+4BzjGvziM|IBdTtKPDsE
zWc&<9gzjue3q0i0wSl95M%}U`a->oMbVx?1s%1$fYg6K%B$6YfPPYM}cu_F1;iSaM
zaImj@$cW9itz{8xlG-}@{w5o{@eplX(xf4iD1R~)CxlTW3lPF6%H~ri1J?1Zktj+PC8nKG<;bt`VWJ>W#ED%3SIpf&U*mE%Q$hpLlBzQPrlw_UogdS*6^}OO
z*X@cop7bXDIDN`jZFgCR8l&q{d{niCH-8yUCO69Tp56dMUrT@NYLqt$YPNEtGS{pa|^yK!8k{e(%ykYLL8-8DA5FT{ESb;bkV1{DAhJSa3
z)6n5f&FU}hSn7=es6)nDLA|Z|%H`A9K-DJW?UE8y5Jp>b5jLw;u|UUkby2HL9x~JA
z7jl!Q{+YM0)9A{U;AL+EQns!U1x*f=+PpC_ZeH_-&l{S
zuQAtB4o6pvj#f?{U|_kb(qxm`s{DqFl@#Oo)>_rBQT_Y+1uufA7kQ}0DSxJA5uEx4
zQ%S}1=2%K64P_Uh5@#4%V&N*6k6^c7ta)t7V$x97N%K1D9dW2w
z*MwbawFtX0lEg*9=3GC1S%3eTuT3%^gww~`FX}1S9Xh4H8|AYM$Q(Ws!8;lf6yhLi
zRj5y*yu(zzwJcJ#PLZW+X21dxt7%SYIHvRTq@~}4&mk;?rn|ur2s(U-^=Ua7lsr)X
zLMW`SYe7jh>Q#t)=3s|{Dg?MMAu$FGH-gd@HGeU$&L-LP^cip*B!9Kn!Dncz9H+nQ
zcnXxZ?DG-Yc2UsiH7c;*A{GVivN+jL@OGD;a~mn|un+ZMrfW?$r@O#zB>`v*?ZW7n
z+N3kos;(zDMh}2JM!T5Sq@nKB&X#AHS#1?D_cng3MJ`wg
z{R|S^Mbj=pM~|(g9M)UXmJRPb?K814g7KLqauP>dj+R`AdRHh`mdKhwE{#enz+O!y+t`Gms_uUCL?t_!WPfG{yVU@E#gh8d`#t
z01~{12JLj&laZ60FSp6B$-odgOnsD~j+}QF{)FenplV*r1g|CIeK*=cnQKG1?R_8l
zH+yB_)ne}5>PD;mv_a$Q|HztBMtIK4mVszOz<3rgRv?fzQ0l~VZ*)@=V49Lh`(
zVdyL5qO4fp0%sW2xnrg9qR{g+UCPfK6WT1Bgm>z8RcXrf^SzV)Qsz0Rsy1*y%OH%-
z;5D(TkexhkYW^REmNf95O-Nam!1_uuS1G(j29%V*+7#!Xl$!rDf`%@+3d4|5{uH$3N+^X
z93!^KJbx#|U3~gbq$VpCzSjJso4DKZc_sZIK_J~=x;McH#mo055N9sqm!KPSURMaD
z<(ap9fBczfN-=Z2l6tKz-`MB;#7_j#!gXY%A|0|_vi?$d>MX156}vF|oo4HS-ysO$
zhcHN@(Ep~lwKqWYlw#7Gt4W=D*G2matnP_QYJbv`OT;mo!h<`K4Yb(g{UVc>^}szk
ztG1W>Yg$zdioW94vdw9+EP_R$=j{F!M&>(!MesPH!)a!
zjm2*!mCa}>m3?Pp3o>U=^Bi7bNpZHy^O5;Bvq>VIMvHy-bp%}&!J{SXEjXB0HEC$*
z(SKc2;Y$<--oH@(so_z5&EZqoJ-+k$J#&!oQyJrZa@k^Y6tg0Z;)aaQJL{T4(LH~T
zRZCNfN!5VwV>*w5bBb2jNczm6GqXM6Axql_yQ*fGaaN|Y
zeX^Xp&W&QK%Swzxzhb1uGYsP-leI$Lkc
zthP7@C8a8SRO}}3ovg$anEDANAJS*|dRP>|Y!%2v#PBMXC9=p8R1wXF3^M%KW;D6t
zh;q;@L%>TEe9EoEOz{hB33)XS?|;z=pRyS+0Bg;mkto`Hgx4_7ab=`9dj@g?Lsn41
zD-`S#H#oY{5gSeZ6u`hZJri+BwV4^A)-|^FAoVr#mx7%cuh1s(CDn0jqPH#zVC{;)
z(T>fG))-k*`obx>2~#7hgBi^r(@`xA!7CtN6N-rQ^O#}~19udxDFrX^7k^&d*o@<}
zERpqW&rwI&T1CA>=P*I0b_MJSk4K^@$>z<2nd-=)#nibpyv8x%uMqHJ1%@|9W;BCH
z$B)9l{nN}2oi(HJ1syZ{f`5T6x*S!&7+32G14D0(-8zJ7$c?z@Ij$q!d5I0WqOr&o
z-F>kpn|ixND9o^1L^DX*#&E}+GZdnp88>3#-6Wet7xfVX_XgIY>jGOrEfmi!bDiqI
zGd-+=#&={Q%y>)Iq(OJ2l>onoysChu<&Lm46RtW1FOra71U0Z&EPsH7H7>dEI%Ity
zxYt@d9QN`4j?F!(EHAJg@~@PCa1VSJES!u=K+dV?{`?$8Po5_gK%wUi@@lPG+HSFtRS
zRs3O*6G5`KJD?^k0^ZqKt8IiX*iouKjGw4Tyls8u6P^6|R~Q{})H?7%%%1RAwi;SQ5Ya_DWOv|Fy}XOciDi-nTSh6qUs=cNOvUh*?+~9#?B&Fr+n*afZA1?
zWr$`kz*{$|fhlcOL1TGEtCv)AJ)VIMQibM_%v4HM#ls^LbH7EvE4qRgWAC8wU`%uc
z5?>(&CrQkEud#V@&QJ(5ptr@HQ$O+W9ZYyx>l)kHemU&EEVogZgqwhR1e$+#pC$^s
zQYQ@_Qh(wc!GAm@RkOqEun2I4iZDvR`R6&Rg1B%z_9!^wr=%u~nW1p3W}RAW%HW)^
zt*tzA#9B7veswFM4i26_V|ZdG0xOT8SPZ+{C8qw5Q0`Af=G`d!p|&u8+stYl^9d1iaI;z4Ko-WV3UI
zVsJ;puh27R#?+)+JaexJI?cTJH;ry*&S(aOYf^oAFNtb4xTb0Xg_w8qR|Pmx03cug
z)3L9)lUI}F8zK`2FaLy`RwxA^vOP}?S0R60q|jYv*Fjp*3x+6OuuM)zWq+cBF>z}S
znSV-O_D}RSfXVh=uxvUn6-cp{i~nX);XG&Hj{~fZJk>T(VhV0;Spw
zI7ciHGQAg9nJCvXUs4n&lAAG5f+`kO-+!%MEsoK3=M0&$vdHTdc!W0FjLR0CQXphD
zCZ47a8JKwTe4?`VhZl5AeXta!Jh`^bXLLH~G}$kv6p>1m$vlqS+Ut|DktaN3zR`;K
zGH8rxSX32xp*odGQFDe$tt>XFXP)D$L_{b~!pP{*yDT}B16qudyD<0q%+8n%kAGSc
zc&>MWBx#m+p|Y{uA9*4>p*+*hk06aHPY-OiI-Z#R8R~d?-^bmM5N!bgg<)jwSn`;5
z$WMo7=v%e%>MzZ2CLi);&`a`jG
zUcxyNs;jH~ZPX@s4uM(MHLxas6@^iHDm_TZq9D-LgMTuoW`=*z))GEHXD*?^x+w_s
z^-wdu5lQP%dN4w(`hwngPC$Phs|r?+{13N06ol(RmQ8}|%fO9;U!;Z6|9|+Oii5bp
zst$PhU00P_qTsl0Xfuh<>}YZ-3LlsPYf6DDbfnMEf)m943Kj*b4l{PqVg%i1O93;C
zt*1g9{d%|xVf5=-g)G?k08^9(8?VW^vpsAVZMy)}%~k#r`&uAQ0q{NY5@f_!tvQ45
zbP;oPb?tkn!0{38Y)w2EKz}o83AC0i&W@Tl@xxa)I-%&{#SL;W^P*R6(vGSRX>^}E
z1&`*{P8za~Dg*Q@`A~amu>FXYxvt4%u>!bn*gt1D;awj*b$I*YtgqSbqe3Hq6_TJ^5H;
zhyG4coa~6#7?ro+Y;BAAZhQ%Y7;Sl9mY2}?i)&h2&4-yi3TjVyEDBwng0%FqTQ*1w
zz96-4u4R|?Z(I2r+p30T`&P90&r#m6H{tp0B{whV@Z&l-;@gVwWtmRoj6&m5m*T5?
zI-%N1<2%td(-8BRiGP?JU48ggfBX{qM|g1-mqtU`{_{upmR?oRziqt8A5m0Eaj~Ic
z1B0*jsenjd-A6AzxY5y!yqSneU*A4Kml%F}TeUefG!>ijM(WC1unZma7^$LVi$bpG
z2_H;u;K!C#1+1XYch!!{&<(v@5CU9BMerQP##H3s8Y+s_hJTUzm@@<#YN`v523(XW
zMd0k34OY&<
z+CnQ3;$WR+AjIF-E?Vzv`;2upHg1H)=jv}mMwoIdCx6dN{lSeQPk2Fp$o>qMMqZB^
z!5@1_DQ@tT-`?4NRF_8l|$md%pN6&MYo#;}{t@Ynkv^<~`1ES1-Y_(TyvM^&6WcTd{6Lj>lhe$Yf{_P`J9^RM5k=msiOmlZf1u7H
zZo%Wl%;k|8%`jN&ue{u`cM43XU9cPgdc!k1xyBT)+AIS#V|Ut2r^3hs#hu+B=rK+Z
zx_@L&eP;gg0k;*r+Zfze<6;X
ztVtSp+20;7NwjII-f?#Jl)FLb{8E%oKPyR3Yaz3e;M24n`N@FY%`RC>hDfhBWq)hw
zgD+A<#jhHby7-WN$>jO`vPHr(had)eYU*&BG0
zyQ+ZI^eOyl;(
zIZoDh>Z1Vzb(j1uy&96v?}DAa@8l-~q`#Ywa3cz!S3Jj5WN0a3pq76q_)0)WLMEV>
z1wrkhfTQVEDynnpkY1wtq<;))Bx)rL>7h%OD5PdTbcjpbr*5o+Dim=orAzXPUQ)=U
z6~ZiBj2x}N$bHn1*&x(nx~Dkkqe(Zl0asSGkUWQpnI5+05IIjzQf*pE7kY>(qs)?-
zed(CV@?jERaTue{>h
z7|~)8@NymaTE{qs#bZ(o6uQX^dgzfSEvOaXlag_&hsV^R-*wBGJxFS=tl5oF{frB}
zCP>Xp3$z|n^R+60L9IeFeL)WiRoa4WlX+f4$>x0V&9qO0uPy0)j75TCS#*=9TJy>_
z>c2j+=R#yhhhC*@P=BpR#L0BO>Iny0vPhI@tX&8_po*O=p_arE};FQ+ZnF
z-))7E-nO)DJGsFd
znKP)4n2@hPoN=kYh_o+ruPA^>A{YEP9=wV0kDaEe0Fvl`L-OXqsXG`oc6(K%snM_QVP0z;kElUEbE?3vl
zDaH3uXhPbin7(XXS(eS0)z|5J*^k$>yST;$Rc3p0$(+18rFrmC*&
z7w{38K~sr^)94jUz
zyw*Rfr+6VH!}|8jLqDHunD
z&!cOXz6T>KV7sn=E<3Yu?}^suN|IFd&^6ETc$u>}wD?E<6xpXnCJm;fyw4FraK)Wr
zIA{B+EHD0E_E$LOWr;|4KAQA|k~{YdSlnfaD1Re8i-Q#OU|gKrAqi-{AkzazH_y=|
zY?th>x9A*3Z{p~GeT;X`04pTG75LKH1Z3FqkOGjFyQ&~%R)qW0SMTg4fa61hDJk!}
zp#huSkIVk8$vocEe_Gcl4iB1T=|Al@Lw>75p$}gFkd{@P{!Q{C`2S!Oh&w5KHWONkb?c__l^vahGcwLbdRh0VGwd
zbt^z974&+5SS9bCKIHCiKh3~IHlMBj;fQ7!O~h2+yockpWqLQIsEBKL-O7a6QlLEu4x=iIoXm61u?24e2WYgP};g?>>$!R
zk0o!Kap?>5f+PAhYVFmGokZd+uwzMFmVEjG<2ks9BI6fX>qfoX8Se@y>mRS!i-R%@bH2TKk!F?SP>8=G$1brhmrJv#N;7Xhv0r#^!OE
z(iljq4M`YjjEh*Dom0(J6b#OZxn5e(+zG
zJeqQjwHJIe-l!8Er>aFq7;_h~&R`p-Py6u(xhX|Qc2|#;S=FYQ6CisoByn_scZ$s&J{96mw$msZ**%$x5W
z52O{OmFlL(`30v(l-5tKh-i%=>8(g5bn8%S6wech!GC%O^r_MSRV_FH&A}Je9gZvU5aKEpXPW6JV@5*@6%@B)
zitK8v9YM9DJr#ej&_OM0{_p=x|4%Rp
zlt27W&hC!FqRJr~%%U3$sLiZ9^Z?-$TKdB?b>3oiEATAmup%7djZOyu`r6)kB*h7I$wg
z_*u;M+!B8nXKHqa*37y_;ZeK~wi(8#cene@-Y@kdq5DhXowM>OQZt9u*P5UBnSz-DeYC`7?(l`YFD6
zqsf1;EHQX(od&~OV^M%`E9o7=m{`Bc4Srmw2jFB)3l`RqP#Vid(o)ee(nVV+bUu=n
z!VZxw+Uj7lNVK$Z!b7w*6r*o(q8{I+XM>HoV7cf^gW_8Ag40bHCxaqdSOky&zfCsa
zRVv9xlZykoQCgs4SdiSBK`gjB0P-`?bwYmtJG(`ZnPFEOLG_797EojfQj#>*9FH8j
zN+u0%j~Y3l-1?^J&ETzoZdeZUVXC8BaPl<8e^xySf`&H+dOz_XqRoDifWfA<&?o0k*`A3IBjC|IV%hNw*DmBJGqs{Qa(oAj8q
zyDW*j)No0+IgYM&T~mMe()%Ymj4s)-N%}6bs~heA{u}JrPn_r)pKn+k
zHzQZJHKoX@hajGlnayNVKWp>!3Uf)FP!Zb^xDd;d)6o^UIgYAtQBZ2j!#>ui0#_BO
zro?=CP+o1&NDvKev&)tMe7RkgL{n83Vw-LJ^4JzvMArFQ8~#0twnu+&!+7KtvPo|G
zp;SJqEH*cn^z(lut;ZIv0^wLCSt}RUOGoV>){N$MI_bqRZN`;5rKr299aZVNX4Mmk
zh|9$zS1CoUsaE~I6xx$T`9srn^TJkFXa4PRt8~)v>eZ*-vpHz*qIy^-dhfaAf81@b
zE=fN87MI(aGxWX7+S6q(cK6418cR++%R3eK!qJ|*hcSO%`z%;~&87dTy3Rgw$8LCa
zt2g=N1+I_R;bZ$MQ;J->Vw3s-Wp!6=#O}(_Yl`-o%gFJ_lP5*(XsUeVx%8=*rup8)kf$UC1YFis-o43$|4kB#Y83^Wrsl$g?=@>AT#OR@>tiqDCJnQ`)ND@v5QfFHG4?uxp?ACvqVJ=dd=vOnb9HdME%gac
zgSPxkQH;htBU;z2nln|_CgoY~xITeRGV5uLk)KV+yYOdxAD{@(G;Tf58FIZ!PGj`>
zl8(E%S#s(LeE;upmqip$s<}GfH+l{BT$9j}fz?rA8>26)Xq`p&x691edvtDQ)cDMbsGK)bTCip9N8J}7n8GiiVw
z75$&CpCVv>i$InN@Z1=ddY(hdIDtZbGQ9GrpL_{MI;~BDYnoyCD+asls!bv9vE}zo
z>6tV5x)p_&FGT=R;2vX>&@%Hq>8+SM&p#*WdblKq2#CrMcZ!M1$hl-|H6yKI5y*eC
z#?aot=stq0b|`8EI&0*v^om7r$!)0YtGeLUH2)y~i{NfrDdi{fPA|Y#uq;98DmbOS
z8>ur8aefj6)d`8nlIo?M1hshR7ek5Jw09Vi
zgvLbMe5O8NQeFtqx
zxELlzlLlQ|kYrIdKuWnTBYneWS(bTUTGO%@fAXR@!NoB!6_MyVYJOa2%K-O;pp~pD
zU@Z|&=J_Sw!H^bf&JgHm^22|F>z99nZ>K2*ucIeO&v~Cj2CT*+z;(1-=tN=gf)*@;
zo+@J__2klNgpLq?0&if(U6#mNUR8{6E#hnW!SOmcW()aO@S3qFJQhbwUGk~eZ>-J<
zB!};VoD@Ejvr0MnO+{i=NmvINmm=q?Ow)}?H-64g=viGJ`^|_au&#fxwUiY5Bb)QZ
zH_>Pfy0cx^|Ewb`M%Q=+y$QC(CY%Asvdy8>CEv%|5@~G-;d6%JC+QXei{A)lJXMY4aOhUN^vD@KEd%MtJDK)$$KovvFhSL(U(P)|fIb6Ha8Tf?*G`}H?Ix`}JF
z8m(oEuXrqCP2v|fLdX2VJ~1n1Kh7?$7j3rHYdU`?|DS&OK-MLST9+{VMBesPykgScvno;9Vos(63&AKJQ$
zXKF@=s#V3I-R!*y%s>&8P!D#t>9i;+@fV*Bf4lRfXhi^B?02I`AofLeR}Him1-s
z;Y^@qxy6zQXmfYf#&e`>LInd>QXCOnVKN7^trKdCm!m}6nQW^9+Ggd5QaCYfDet4S
zq`-b}u)%uI-XVYb=|oW&HE9tf7F}m0h>3q6d0M6m9V(}zw5zM5hqHD)rT7iLGJyoJ
zT5CqL<76#c`lW!~xjB+HVOdQ8L*2&-3$;yQAEVvORu^gMFIELeV>P`5R5
z7-XQs2lF)tP{Ah_F<=zn>EMe~kLZ+Es3D!FrQ?`R>jKyp;ELjJ2XGVsO)FdtI`e-D
zB?jUwiKnF7dXppUeI#QS%z2;PX&U<@1wW8pf`MV<75-WiJSjCv^&g&u#OJ5O4fjS5i664R`&^68PF^8
zIX7JaG+R@OM6+_CdaGDh$L`A02(f>;FBT}9$~kI)Mh}S|o#BHc3}RPVtg6ox&k_0K
zE)24NHZ-CvOL)zu{9#VX?1#RL4JNuPOq4J^ScjuD_M|x@xixU`)$AN7q#H
z{u^c@!4$ni@r!!_n4Kp)L`T6W(J`PBxKacH)srx-7iya0s_8hxG0Gq&H?Dsy*}O(;
z0GiVHDs*NuS`NfS#?{hi4x^<-#ox+{TR3F1frhhIZ5nkWzK_-?1!EHE<@RPslwWce
zeMuz?VzB01jT4Jjk67j|!%K8^9J>UU%2Y+5)QnSGE>J@{1zT`I`t4XfuH(6J*L4VJbdCo&q!&!wwNcGn(T9E
z1=H+fXScY`AA(nZN<~xwc8Oc?15cY+P4V7zwzxE{@kb`Kie7&TY(eW9S86{^S`^z%i!E4&PL)IF8U+_Oqw$^YlFU@vnnOY+
zGTAtUg772+q=ifpMCA_ds7P9(GdY_0X3?W`<(i~-9JNgw$?oRV4n^1F)BXZ
zuIq2;jS25Y$4vHg#}6-=50*_F2Xp~>Wzhz7ZRAduM|i^CM=~|Gmt_+IN^Xkq;oLPuNfw~wXBhpc{^asa
zDKtk!b9Hr#;@CZ*KahJAtXz7tT2l&KS43tR;p*+f_~>8lIU^Co;X
zENxk`=*xdf!o>B9!WgbKi=dL|Yuhttr87VjZQ)*V9*f_01}BNC6H2xSMR%?+Q_uiS
z7J)AA+gDeY@;i^fEw!$JB{dKo&u;M%G`(dJ*fIskJ%U$0cA1XO_LZ}$ZL^mKZNLHu
zbeRR&N{Y0|*d>t{}ClfJChf<+hWXcZmA*h(FA
zUC_PG6H3ZHjKEsG0G;QU;#xT5=j`BpuB?VB1+Q8!RGN+`am6P_eZ?#aSWe%scr%sR
zr#pP`7$@*`LBTtD@X0jrvcKiQC)c;}X;*)&*`2l-3pJnJUt1I$v-&I}$B8dl_Lgqj
zd=_yFUcxAWHsl;rOo>-PH=<+b-mI$%R!(ea6qz*G60?O|ThjdUr6tM#E%Hp%JcgAJ
zrvY@8PcuwORmUc(c^Ru}I8MULCF^{^#OoSYObzXel#Q$MMT%vrkzTc~ktGfNZqk1{
z;4XZ=XTJ*OgB4(ofP_N{Qn~_<(XiXGGab8&=xpig5vqn>l@z#=RY^q*I?xoLf`O)D
zl%<&N&AH?xm8NN-dwW^K;@(~nzniqkR)2d=K6rr}z2&j-ZlW-b9h@w^yz|%hQSMR#
z?C9~U#SM5^DAOkwFFA>om8Ci~9iD&OPbv?;yQztq8gmTqXdcwh?
zA;JR4YYk9wumJEe%I{+HhYxlWVta}@sEg+BrA-q{N=svfse
zSKo^!GHFuT^o>2s&MyRi^bW~3p|2-P%9OQ0_JJ||ivr)_?D;E2eMp~;=~{o+6dk1v
zz565+)%O}y{@qVul(BO5>NIaV*(6N6N8E;Z&`Zzthyr`Ddupp_x=7L#4RFU;ty`mH
z>dWKG^1A-zMdTI2muXSZb!(C*l#Hivp0jlmcti*HZRaLvG?mt*;q0c>zyksex#C9!
zO2H@hd>+w23%YJ~D)G<^NGyN+wff;)>m5Qj>2dvO@#XPMc-7|aR_*8V-$s+Xt{J-t
z4G=o+FX-U8{zK$u4sa}t({DhvJ>l^?yMdMi{{Tp5*i$?TQmRfSf1<;9qe(&6%~}#S
z9Zld2lXaB$f}^yB{&B>~uHo#c)jJEf+Dx=|PS;H$?8`L`k^Nt@<;Z_`VUPz-*dOrL
zs`TMP#^YY1@YOpqf8hoSvMnxwqf#IfCf+dt(I%3-wI+K%Ov+2(O8q#2;TUpfn8;~1
zqPDB;+x^pDnGxt{#QxhhLfJo~F~nL^ijt767WSkjqP}1-lqtjBkHcDRYRzC@s01t+&m$N;>Jpj>)Hydh)tvll6;jnOzcgAC!mc
zXmKlb4c!{!0ZV1)J7%}&{6~I&Cz}Xm8>3rmwtou_4|$5bZ5aAbnx)RZf}dp?E8?;w
zaMtCvUlM65yJV#p@(xs{YMz7ZT1M9O(o0s`-T|4bD9a{kIQxGPJ9G{cRC_|jYe24#
zlSHnUb;RmVLHqO8h)W*KXb_m?7@++`cea
zkTq!-^xOTASE?aN%M!hloFoc;-<;7|lZK;1>ovbr49%t*aoHl$uRQeYlBfQ~k9lvx
zG)PQ|T(z0>RjPmYhKCEJc7pi_QgV5?VCF2a;piw1A4pdp#>JP6sa$i&oG1O!SF&F{
zSX_CVfOw5WTqejo6Gz`1{V?;hMygYL!eeuUx7ww`FLVY}Q+<8G@^n#XcpREsdW=6&
z=7x{g;5@bt(@uoqM4V20_yk?zcTg0&uV}pL+k8G}
zm5NRjgu+#4AOIan=fp#*@5a32HfM;s7;PTk+u)qOt{FP0O0v!;yp=y#(yD^j?;^#n
z{pNh}%~@K*;cVCSKkG}3(KTLpZGx@rjx&Id+h)?Q$oJo|oMx{nOTKy&o-a7sJ?gpZ
zFkNb)J6(U}TwUFLyZzwo9=~lrUOzayy;*(MoT2L>SihQkHMwRhzAw{w`r6V63T
zz>2m>svb&*Ee$nQ_E{r>oMBk2HkzF4>Z(2hfTe$3*SOjOuC7jg_~iQJHhItZjZ#6@
zl;Y{2FOLFgT%JwRXsg4gBE9c8a|F?cLn;{pEAAa)k|&5~+?`Iio)m@1{^=nE*FDc+
zot-uL;8r)+sq|WSSvIvF6Nh=Xpi(P2V;G_~KU|o7$}n{Ve6>E}o4W9=ZIkehhOPp4nOinhnF;Iz#Y9ds(xYcu^+#RNO+zH-qy0mdoK8(242mKCQ$3~ksBGT8v{q_B6EUc$Im4i>
zXy}76t{RI#ceGp1gcQ}AH4|^uxP40fMvZ@wLi&4;z@D2mqv2Jql1H06xSRz`pcW)t
z6ZMxtz2eYor61pdG)BkG5xhz!4YaBNf{hXo+DwR%`dvjrP?H7!eErLC0TEAF_!8}5|C?=}LykFbp9@70GtH^&&_7OlkvbX1Ho
zA-%6k^0kO0Dz*}y)J4F)>@hZ>c~L;i+#OnqcZ{#2$bB1h`<1U{U1MvsYMkf(ynilw
zgOR$i<_w9jmdrftsXCebRS)k;kFkIEJUmHUnevxBV?B}K%GG}#WBx$}PhgkV6#4Jv
zNltq2(9Rho+QPixYn6A3vFSAN{P^FJyl3!$SJf}`02%0Zb!wwGG_PD9AAn=#%klYR
zp*H+H$KY2}ctMeyc|iYrXc^o+j#O0FI*$u~BD}ky-yjtXzd&gK(ISfiM$msI?@-3Y
zx-?ls;0X4?$qSIsl_}8k{|X|$lQ515pGPN{;JPug0$#-m4wv-W$4krXW66Zym7ju@
zQ~-`nPJ9)x3g!&7tP--Xem19qgr=jMMRO!f0`37B#ai?nL6TsJEqzcq>sl&w6V|mAd2TG)
zv+FNsS;ER&3n)`wn_ZeFuO$N_M_x-qr}&92Ny+*05yc5~v#csyN5#I64&XLX!pgbd
zjd$VJaCLTGv-Lqe{RxIDRDCHXeiVehBOo0C6^B?{$e!Gcl2@UWqlbSl=Rc+Rse|hq
z5;LP&vQ>0*W3hOuVhaUgB6WP$p30UbI(7G~UN}4A7A&qt@7?poNAX>HHdJv7mWR&k
zBvg9;y#G6OBNX3+@$iJQw#9{1a>}F(h+WH9BN8c_qq&*^LmBfW^$B9kmkJlYSfI3f
z;9|Z|h>*p6iQOjK^Fx1lAE3K1IuG|xmn}F)#wDU3Fu%1G@I*;F0ILzoOKt1Wc2irL
zj$yJ-2L_Pl2wIl7idF;iR5|=|SG6Ss!dGpp7}S68nfAfEMCD~06Rl(|TNH{0dmo{n
zv5w?haA@8~NuwBZeiE#+NlZsX#^aWSjq69xsRV7_O(=s!%=Ga
zrWbB;pqlDBigf0q^_@^s8AX!ywAfg)vP9yMtq@hVzj;w4kxbN?RBy?0di4%yb2L>ubB4uH=rP_{6p2cW>Bed}RE>2hBpV$cDGo{16z2+N
z>&7NYRFxBFe|?Fgqqjr~LI*`>segj%9xdB;(?gOLP#gOWP{ytAsy)
zGj#tm5JgkyNNo{#vSOoO=m1jK8Afudny&s-F#m!4i_ma8mUz+tI$MziDD-UoG|uY`
zM$(kRbTmYq@(XDZJ8&gl
zu%I#lAvS-F;seopkFKK*4JjiRC&lc>nP5&_3ib4-0xsE5Lazr3GM#D0pfpMbCAWvn>$10#lS+^{pRlTSI0zo
z3uH0z*XutW6Xnf;DO&f*N82dG*Q>qq!dc&wt~moH;o%bx*n4NW+hTTKtVQ$J;*g6v
z#9#thv=l>~d9lhgPrO=sfKHrPIrVwdtBm|ohDisnj0)JCn9oi@7`
z<=fX56b0GUoi0%;@K+A{5#b}fLFfep7yW;)&}Ar%UB8|n1z3{?Q&MdFP&kbu6oB(DThMYDcjg$ESJ=m;ia&H&3L
zfLh+F0(72)oRtsw|Nd_%*2~x8Y0&KAdeMe5K1ba1aLh2UmY?TnlA6x{y)=If{1I4d
zFIZ%Wi1-CZCwnYAfKg|s~SpE#mAxUE%uvJQ!{v#|brTCP(M(}_2OCqV1LwUOgXLMHHseoxhmQ*&#o>6o{H>;!*sK8BIxU8<6pjwqGQP_#8*0L`CS`F^V*2I8B#HnnW9PQEPFT&6E#c
zSc}J)9tbv>CcGs$-rO*y_$0c5F!7$>ZkWx6QFb`^)Upw<_>F&RBq|-4Q4E*Mtd{!V
z#y?~aXJN+;hEq0B7hla?`;T6Han-k@K_sp@yy{xazdl+uE?82Tz%a;@t_Nc*EY>xd
ztYyPueVZ%aqcI%h1xx8D)tOhds+c5VE_CkRF&yJ%i&ID6DDxP>Cf9^REsU{0X-yh3
znIgT{9Uw+kKlp!;bU@jCTSzgyYq_kF@G^oB&W)(H4)K*J&>@n4Wr~)QR1@Sl!{e0V
zlj)Maga)mqg!e+4k&}|^(lHJQX+nvzSJHhLZ(ncC;gy*^=j-OH(Vp;VRh5~W=S>)l
zF{;lQMlEe)?LRJ9N|n0NwjphjPNE9mX5CkjgbJW1ii8OH3q^t&;+-JD+v%Ag{;f+GqvNL(
z$G6H8t%5vZpHAf1jl6Tvby_-p0f$;
zSmgG_p>hE@?FkQ=Wd76jy-ZbB@5ATEIRhr);S&!D)WG1i#fT@)mNu|Z8Vs`L0MXI|
z6-t8vmQJQE5jYiX@En?HEt^EAh9sVjukI7_%?N*)Qt(P8#ic_tQf4&1TnmK1!z$?F
z^t1%&Zj*ObCD;>h33#nP7@k;!VOVKv+f&@V&X(AyV0TqWB
zB)t!b$|!jiN))XSM|+bh0nNrl!gt8VM5{8FjfrZmUfGxm
zo9=%dv2IcPejAbR=;(p2`5sOF;Y0si_RA8114^U!O=w;b)I??Ca%+Bz&SCV%W}<)M
z3Bgy(KSkjgM9Hc`%S@sAY-90jAivWdg3vAsES!$~=8~#-BP5RD-IX`Ta-{W}f2GO{
zL74wga(lvquC=s+`SQd`;uv=mQ^Ya8(+PheU6Vh1zF*}d1QB7`BqTk7RHP^21J7<;
z*uSglTEKJt6yLhPK-ktboUl~cKkw5sBwL<(Pr!yV*TalmC76)O@lhTd55zzw4WOd&
zpjxiTCn9b(a-@bSvsmb;G
z$CN_k&P)ONS@a;7BG+)TG3OFN^^$oGMYpVq_&z|O1MQqa)m9U^!j*nUj|3y}I}`qw
zAdG-q-U_Wq*uus)X-FlpB|5fm5mXpFocOqQI&c
zC6-Zq5whxQK&I-|TGGIKhU)Ok5=_zG|AnG38}(r*OECU$_71t1z}$Esg7be44bi?R
zL-MxJDHecCOeImpM_Y$5J!hqbd{0|#D#3}$y>se0$}3;PcVJ?C2=HAZHh@rd=IbdzaFcsHCNxoy4Y@_g)9)|f|ILJr^
zlvUogtUZDV1?d^bvdh39e@5=5%G1URl#9R1&!NIWapWXdS9jlTAGd$kyX#vg^F{Nk
zD0^`X;7-)mnuD#>(|rGWyZ!o4C@pgaw}XzG`>$YCGNWNTsQBo7eYiGI8#!(V9iMmC
zA2-h31C*3i8@h*@$NhhP+d2F4YSwUB_EN$zM4tsEa3v$@r4X+2Y(9JkRbQR!-QDKy
z^W*ON!GV%B72tQ!b$fq*n-vDErbz3rpnIr!{07w>ZHtTEN7J81HT~H|)9(Jei4h#b
zc2Kc*Ztm|sLYt617cz9w1~)R1HPmq|%D!%!QDJveF*T!N+CfEiH+XaF80bd?dIwE^
zUVr`O81O-I+zvXvJDbmcJUAa8H+P>lcbkWQKHlB$pgPplRuO+`DBbTLKi>ceT9$Bn
zw3+bT28!CU2+S`0k9k@}C>6^Va(69{-!~6`WX
zNe$UblLk`FUTtUOx4uDEy4I8e)sRr$IR{~{udtK6&#oV%$AnL!yLHfjP|#j*{OZQZ
zUVb=OlqCNzKW%>nf#Xm+5+`&Xb_g-U2+kL4e)Hk$b#51T`|bUn_lcMPU$H4tY>yOw
z#eWi#Ui`;6590J#us+84RIhi{+BP4MtmZ1iz`y_dpt|sxp!T~ji{ipIK^1%FWAC!e
z@NYW|TAr0-;Qfmnc%I`p+Zi_S?U($I_NiGx`GntqU&DX!T9XE?WWRna`8|<y^+CVda!UOfE&B%ZZke|Jkfxv?K?
Shkk@(KUn#K{Qn1FF;#C?C|KqI

diff --git a/doc/py-modindex.html b/doc/py-modindex.html
index 12f6ffbabb..543f0da954 100644
--- a/doc/py-modindex.html
+++ b/doc/py-modindex.html
@@ -6,7 +6,7 @@
   
     
     
-    Python Module Index — PaddlePaddle  documentation
+    Python Module Index — PaddlePaddle  documentation
     
     
     
@@ -39,7 +39,7 @@
         
  • 
           modules |
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
       
 
@@ -56,7 +56,7 @@
    
 
    
-     
+     
@@ -67,12 +67,12 @@
        
-       
-       
     
     
    
      
    
        p
    
      
    
      
    
           
+          
        paddle.trainer_config_helpers.attrs
        
    
      
    
           
+          
        paddle.trainer_config_helpers.data_sources
        
    
    
    
@@ -86,14 +86,11 @@
 
 
         
@@ -109,12 +106,12 @@
         
  • 
           modules |
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
     
   
 
\ No newline at end of file
diff --git a/doc/search.html b/doc/search.html
index e8848a3460..2d1f5f908e 100644
--- a/doc/search.html
+++ b/doc/search.html
@@ -6,7 +6,7 @@
   
     
     
-    Search — PaddlePaddle  documentation
+    Search — PaddlePaddle  documentation
     
     
     
@@ -44,7 +44,7 @@
         
  • 
           modules |
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
       
 
@@ -95,12 +95,12 @@
         
  • 
           modules |
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
     
   
 
\ No newline at end of file
diff --git a/doc/searchindex.js b/doc/searchindex.js
index 9dbe69e48d..2e97cc18f1 100644
--- a/doc/searchindex.js
+++ b/doc/searchindex.js
@@ -1 +1 @@
-Search.setIndex({envversion:46,filenames:["algorithm/rnn/rnn","build/build_from_source","build/contribute_to_paddle","build/docker_install","build/index","build/ubuntu_install","cluster/index","cluster/opensource/cluster_train","demo/embedding_model/index","demo/image_classification/image_classification","demo/image_classification/index","demo/imagenet_model/resnet_model","demo/index","demo/quick_start/index_en","demo/rec/ml_dataset","demo/rec/ml_regression","demo/sentiment_analysis/index","demo/sentiment_analysis/sentiment_analysis","demo/text_generation/index","demo/text_generation/text_generation","dev/new_layer/index","dev/new_layer/new_layer","index","layer","source/api/api","source/cuda/cuda/cuda","source/cuda/cuda/index","source/cuda/matrix/index","source/cuda/matrix/matrix","source/cuda/rnn/index","source/cuda/rnn/rnn","source/cuda/utils/index","source/cuda/utils/utils","source/gserver/activations/index","source/gserver/dataprovider/dataproviders","source/gserver/dataprovider/index","source/gserver/evaluators/evaluators","source/gserver/evaluators/index","source/gserver/gradientmachines/gradientmachines","source/gserver/gradientmachines/index","source/gserver/layers/index","source/gserver/layers/layer","source/index","source/math/matrix/index","source/math/matrix/matrix","source/math/utils/index","source/math/utils/utils","source/parameter/optimizer/index","source/parameter/optimizer/optimizer","source/parameter/parameter/index","source/parameter/parameter/parameter","source/parameter/update/index","source/parameter/update/update","source/pserver/client/client","source/pserver/client/index","source/pserver/network/index","source/pserver/network/network","source/pserver/server/index","source/pserver/server/server","source/trainer/trainer","source/utils/customStackTrace","source/utils/enum","source/utils/lock","source/utils/queue","source/utils/thread","ui/api/trainer_config_helpers/activations","ui/api/trainer_config_helpers/activations_index","ui/api/trainer_config_helpers/attrs","ui/api/trainer_config_helpers/data_sources","ui/api/trainer_config_helpers/evaluators","ui/api/trainer_config_helpers/evaluators_index","ui/api/trainer_config_helpers/index","ui/api/trainer_config_helpers/layers","ui/api/trainer_config_helpers/layers_index","ui/api/trainer_config_helpers/networks","ui/api/trainer_config_helpers/networks_index","ui/api/trainer_config_helpers/optimizers","ui/api/trainer_config_helpers/optimizers_index","ui/api/trainer_config_helpers/poolings","ui/api/trainer_config_helpers/poolings_index","ui/cmd_argument/argument_outline","ui/cmd_argument/detail_introduction","ui/cmd_argument/use_case","ui/data_provider/index","ui/data_provider/pydataprovider2","ui/index","ui/predict/swig_py_paddle_en"],objects:{"":{"Arguments::Arguments":[24,6,1,"_CPPv2N9Arguments9ArgumentsEv"],"Arguments::DISABLE_COPY_AND_ASSIGN":[24,6,1,"_CPPv2N9Arguments23DISABLE_COPY_AND_ASSIGNE9Arguments"],"Arguments::createArguments":[24,6,1,"_CPPv2N9Arguments15createArgumentsE6size_t"],"Arguments::createByPaddleArgumentVector":[24,6,1,"_CPPv2N9Arguments28createByPaddleArgumentVectorEPv"],"Arguments::getBatchSize":[24,6,1,"_CPPv2N9Arguments12getBatchSizeE6size_t"],"Arguments::getInternalArgumentsPtr":[24,6,1,"_CPPv2N9Arguments23getInternalArgumentsPtrEv"],"Arguments::getSlotIds":[24,6,1,"_CPPv2N9Arguments10getSlotIdsE6size_t"],"Arguments::getSlotIn":[24,6,1,"_CPPv2N9Arguments9getSlotInE6size_t"],"Arguments::getSlotNum":[24,6,1,"_CPPv2N9Arguments10getSlotNumEv"],"Arguments::getSlotSequenceDim":[24,6,1,"_CPPv2N9Arguments18getSlotSequenceDimE6size_t"],"Arguments::getSlotSequenceStartPositions":[24,6,1,"_CPPv2N9Arguments29getSlotSequenceStartPositionsE6size_t"],"Arguments::getSlotSubSequenceStartPositions":[24,6,1,"_CPPv2N9Arguments32getSlotSubSequenceStartPositionsE6size_t"],"Arguments::getSlotValue":[24,6,1,"_CPPv2N9Arguments12getSlotValueE6size_t"],"Arguments::m":[24,7,1,"_CPPv2N9Arguments1mE"],"Arguments::resize":[24,6,1,"_CPPv2N9Arguments6resizeE6size_t"],"Arguments::setSlotIds":[24,6,1,"_CPPv2N9Arguments10setSlotIdsE6size_tP7IVector"],"Arguments::setSlotIn":[24,6,1,"_CPPv2N9Arguments9setSlotInE6size_tP6Matrix"],"Arguments::setSlotSequenceDim":[24,6,1,"_CPPv2N9Arguments18setSlotSequenceDimE6size_tP7IVector"],"Arguments::setSlotSequenceStartPositions":[24,6,1,"_CPPv2N9Arguments29setSlotSequenceStartPositionsE6size_tP7IVector"],"Arguments::setSlotSubSequenceStartPositions":[24,6,1,"_CPPv2N9Arguments32setSlotSubSequenceStartPositionsE6size_tP7IVector"],"Arguments::setSlotValue":[24,6,1,"_CPPv2N9Arguments12setSlotValueE6size_tP6Matrix"],"Arguments::~Arguments":[24,6,1,"_CPPv2N9ArgumentsD0Ev"],"FloatArray::FloatArray":[24,6,1,"_CPPv2N10FloatArray10FloatArrayEPKfK6size_t"],"FloatArray::buf":[24,7,1,"_CPPv2N10FloatArray3bufE"],"FloatArray::length":[24,7,1,"_CPPv2N10FloatArray6lengthE"],"FloatArray::needFree":[24,7,1,"_CPPv2N10FloatArray8needFreeE"],"GradientMachine::DISABLE_COPY_AND_ASSIGN":[24,6,1,"_CPPv2N15GradientMachine23DISABLE_COPY_AND_ASSIGNE15GradientMachine"],"GradientMachine::GradientMachine":[24,6,1,"_CPPv2N15GradientMachine15GradientMachineEv"],"GradientMachine::asSequenceGenerator":[24,6,1,"_CPPv2N15GradientMachine19asSequenceGeneratorERKNSt6vectorINSt6stringEEE6size_t6size_t6size_t6size_t"],"GradientMachine::backward":[24,6,1,"_CPPv2N15GradientMachine8backwardERK14UpdateCallback"],"GradientMachine::create":[38,6,1,"_CPPv2N15GradientMachine6createERKNSt6stringEP10DataConfig"],"GradientMachine::createByConfigProtoStr":[24,6,1,"_CPPv2N15GradientMachine22createByConfigProtoStrERKNSt6stringE26GradientMatchineCreateModeRKNSt6vectorIiEE"],"GradientMachine::createByModelConfig":[24,6,1,"_CPPv2N15GradientMachine19createByModelConfigEP11ModelConfig26GradientMatchineCreateModeRKNSt6vectorIiEE"],"GradientMachine::createFromPaddleModelPtr":[24,6,1,"_CPPv2N15GradientMachine24createFromPaddleModelPtrEPv26GradientMatchineCreateModeRKNSt6vectorIiEE"],"GradientMachine::defaultParamTypes":[24,7,1,"_CPPv2N15GradientMachine17defaultParamTypesE"],"GradientMachine::forward":[24,6,1,"_CPPv2N15GradientMachine7forwardERK9ArgumentsP9Arguments8PassType"],"GradientMachine::forwardBackward":[24,6,1,"_CPPv2N15GradientMachine15forwardBackwardERK9ArgumentsP9Arguments8PassTypeRK14UpdateCallback"],"GradientMachine::getLayerOutput":[24,6,1,"_CPPv2N15GradientMachine14getLayerOutputERKNSt6stringE"],"GradientMachine::getParameter":[24,6,1,"_CPPv2N15GradientMachine12getParameterE6size_t"],"GradientMachine::getParameterSize":[24,6,1,"_CPPv2N15GradientMachine16getParameterSizeEv"],"GradientMachine::loadParameters":[24,6,1,"_CPPv2N15GradientMachine14loadParametersERKNSt6stringE"],"GradientMachine::m":[24,7,1,"_CPPv2N15GradientMachine1mE"],"GradientMachine::randParameters":[24,6,1,"_CPPv2N15GradientMachine14randParametersEv"],"GradientMachine::saveParameters":[38,6,1,"_CPPv2N15GradientMachine14saveParametersERKNSt6stringE"],"GradientMachine::~GradientMachine":[24,6,1,"_CPPv2N15GradientMachineD0Ev"],"ISequenceResults::getScore":[24,6,1,"_CPPv2N16ISequenceResults8getScoreE6size_t"],"ISequenceResults::getSentence":[24,6,1,"_CPPv2N16ISequenceResults11getSentenceE6size_tb"],"ISequenceResults::getSequence":[24,6,1,"_CPPv2N16ISequenceResults11getSequenceE6size_t"],"ISequenceResults::getSize":[24,6,1,"_CPPv2N16ISequenceResults7getSizeEv"],"ISequenceResults::~ISequenceResults":[24,6,1,"_CPPv2N16ISequenceResultsD0Ev"],"IVector::DISABLE_COPY_AND_ASSIGN":[24,6,1,"_CPPv2N7IVector23DISABLE_COPY_AND_ASSIGNE7IVector"],"IVector::IVector":[24,6,1,"_CPPv2N7IVector7IVectorEv"],"IVector::copyFromNumpyArray":[24,6,1,"_CPPv2N7IVector18copyFromNumpyArrayEPii"],"IVector::copyToNumpyArray":[24,6,1,"_CPPv2N7IVector16copyToNumpyArrayEPPiPi"],"IVector::create":[24,6,1,"_CPPv2N7IVector6createERKNSt6vectorIiEEb"],"IVector::createByPaddleVectorPtr":[24,6,1,"_CPPv2N7IVector23createByPaddleVectorPtrEPv"],"IVector::createCpuVectorFromNumpy":[24,6,1,"_CPPv2N7IVector24createCpuVectorFromNumpyEPiib"],"IVector::createGpuVectorFromNumy":[24,6,1,"_CPPv2N7IVector23createGpuVectorFromNumyEPii"],"IVector::createZero":[24,6,1,"_CPPv2N7IVector10createZeroE6size_tb"],"IVector::get":[24,6,1,"_CPPv2N7IVector3getEK6size_t"],"IVector::getData":[24,6,1,"_CPPv2N7IVector7getDataEv"],"IVector::getSharedPtr":[24,6,1,"_CPPv2N7IVector12getSharedPtrEv"],"IVector::getSize":[24,6,1,"_CPPv2N7IVector7getSizeEv"],"IVector::isGpu":[24,6,1,"_CPPv2N7IVector5isGpuEv"],"IVector::m":[24,7,1,"_CPPv2N7IVector1mE"],"IVector::operator[]":[24,6,1,"_CPPv2N7IVectorixEK6size_t"],"IVector::set":[24,6,1,"_CPPv2N7IVector3setEK6size_ti"],"IVector::toNumpyArrayInplace":[24,6,1,"_CPPv2N7IVector19toNumpyArrayInplaceEPPiPi"],"IVector::~IVector":[24,6,1,"_CPPv2N7IVectorD0Ev"],"IntArray::IntArray":[24,6,1,"_CPPv2N8IntArray8IntArrayEPKiK6size_tb"],"IntArray::buf":[24,7,1,"_CPPv2N8IntArray3bufE"],"IntArray::length":[24,7,1,"_CPPv2N8IntArray6lengthE"],"IntArray::needFree":[24,7,1,"_CPPv2N8IntArray8needFreeE"],"IntWithFloatArray::IntWithFloatArray":[24,6,1,"_CPPv2N17IntWithFloatArray17IntWithFloatArrayEPKfPKi6size_tb"],"IntWithFloatArray::idxBuf":[24,7,1,"_CPPv2N17IntWithFloatArray6idxBufE"],"IntWithFloatArray::length":[24,7,1,"_CPPv2N17IntWithFloatArray6lengthE"],"IntWithFloatArray::needFree":[24,7,1,"_CPPv2N17IntWithFloatArray8needFreeE"],"IntWithFloatArray::valBuf":[24,7,1,"_CPPv2N17IntWithFloatArray6valBufE"],"Matrix::DISABLE_COPY_AND_ASSIGN":[24,6,1,"_CPPv2N6Matrix23DISABLE_COPY_AND_ASSIGNE6Matrix"],"Matrix::Matrix":[24,6,1,"_CPPv2N6Matrix6MatrixEv"],"Matrix::copyFromNumpyMat":[24,6,1,"_CPPv2N6Matrix16copyFromNumpyMatEPfii"],"Matrix::copyToNumpyMat":[24,6,1,"_CPPv2N6Matrix14copyToNumpyMatEPPfPiPi"],"Matrix::create":[44,6,1,"_CPPv2N6Matrix6createE15MemoryHandlePtr6size_t6size_tb"],"Matrix::createByPaddleMatrixPtr":[24,6,1,"_CPPv2N6Matrix23createByPaddleMatrixPtrEPv"],"Matrix::createCpuDenseFromNumpy":[24,6,1,"_CPPv2N6Matrix23createCpuDenseFromNumpyEPfiib"],"Matrix::createDense":[24,6,1,"_CPPv2N6Matrix11createDenseERKNSt6vectorIfEE6size_t6size_tb"],"Matrix::createGpuDenseFromNumpy":[24,6,1,"_CPPv2N6Matrix23createGpuDenseFromNumpyEPfii"],"Matrix::createSparse":[24,6,1,"_CPPv2N6Matrix12createSparseE6size_t6size_t6size_tbbb"],"Matrix::createSparseMatrix":[44,6,1,"_CPPv2N6Matrix18createSparseMatrixE6size_t6size_t6size_t15SparseValueTypebb"],"Matrix::createZero":[24,6,1,"_CPPv2N6Matrix10createZeroE6size_t6size_tb"],"Matrix::get":[24,6,1,"_CPPv2N6Matrix3getE6size_t6size_t"],"Matrix::getData":[24,6,1,"_CPPv2N6Matrix7getDataEv"],"Matrix::getHeight":[24,6,1,"_CPPv2N6Matrix9getHeightEv"],"Matrix::getSharedPtr":[24,6,1,"_CPPv2N6Matrix12getSharedPtrEv"],"Matrix::getSparseFormat":[24,6,1,"_CPPv2N6Matrix15getSparseFormatEv"],"Matrix::getSparseRowCols":[24,6,1,"_CPPv2N6Matrix16getSparseRowColsE6size_t"],"Matrix::getSparseRowColsVal":[24,6,1,"_CPPv2N6Matrix19getSparseRowColsValE6size_t"],"Matrix::getSparseValueType":[24,6,1,"_CPPv2N6Matrix18getSparseValueTypeEv"],"Matrix::getWidth":[24,6,1,"_CPPv2N6Matrix8getWidthEv"],"Matrix::isGpu":[24,6,1,"_CPPv2N6Matrix5isGpuEv"],"Matrix::isSparse":[24,6,1,"_CPPv2N6Matrix8isSparseEv"],"Matrix::m":[24,7,1,"_CPPv2N6Matrix1mE"],"Matrix::reshape":[44,6,1,"_CPPv2N6Matrix7reshapeE6size_t6size_t"],"Matrix::resizeOrCreate":[44,6,1,"_CPPv2N6Matrix14resizeOrCreateER9MatrixPtr6size_t6size_tbb"],"Matrix::resizeOrCreateSparseMatrix":[44,6,1,"_CPPv2N6Matrix26resizeOrCreateSparseMatrixER9MatrixPtr6size_t6size_t6size_t15SparseValueType12SparseFormatbb"],"Matrix::set":[24,6,1,"_CPPv2N6Matrix3setE6size_t6size_tf"],"Matrix::sparseCopyFrom":[24,6,1,"_CPPv2N6Matrix14sparseCopyFromERKNSt6vectorIiEERKNSt6vectorIiEERKNSt6vectorIfEE"],"Matrix::subMatrix":[44,6,1,"_CPPv2N6Matrix9subMatrixE6size_t6size_t6size_t6size_t"],"Matrix::toNumpyMatInplace":[24,6,1,"_CPPv2N6Matrix17toNumpyMatInplaceEPPfPiPi"],"Matrix::~Matrix":[24,6,1,"_CPPv2N6MatrixD0Ev"],"ModelConfig::DISABLE_COPY_AND_ASSIGN":[24,6,1,"_CPPv2N11ModelConfig23DISABLE_COPY_AND_ASSIGNE11ModelConfig"],"ModelConfig::ModelConfig":[24,6,1,"_CPPv2N11ModelConfig11ModelConfigEv"],"ModelConfig::getPaddleModelConfig":[24,6,1,"_CPPv2N11ModelConfig20getPaddleModelConfigEv"],"ModelConfig::m":[24,7,1,"_CPPv2N11ModelConfig1mE"],"ModelConfig::~ModelConfig":[24,6,1,"_CPPv2N11ModelConfigD0Ev"],"OptimizationConfig::DISABLE_COPY_AND_ASSIGN":[24,6,1,"_CPPv2N18OptimizationConfig23DISABLE_COPY_AND_ASSIGNE18OptimizationConfig"],"OptimizationConfig::OptimizationConfig":[24,6,1,"_CPPv2N18OptimizationConfig18OptimizationConfigEv"],"OptimizationConfig::createFromProtoString":[24,6,1,"_CPPv2N18OptimizationConfig21createFromProtoStringERKNSt6stringE"],"OptimizationConfig::getRawPtr":[24,6,1,"_CPPv2N18OptimizationConfig9getRawPtrEv"],"OptimizationConfig::m":[24,7,1,"_CPPv2N18OptimizationConfig1mE"],"OptimizationConfig::toProtoString":[24,6,1,"_CPPv2N18OptimizationConfig13toProtoStringEv"],"OptimizationConfig::~OptimizationConfig":[24,6,1,"_CPPv2N18OptimizationConfigD0Ev"],"Parameter::DISABLE_COPY_AND_ASSIGN":[24,6,1,"_CPPv2N9Parameter23DISABLE_COPY_AND_ASSIGNE9Parameter"],"Parameter::Parameter":[24,6,1,"_CPPv2N9Parameter9ParameterEv"],"Parameter::createFromRawPtr":[24,6,1,"_CPPv2N9Parameter16createFromRawPtrEPv"],"Parameter::createFromSharedPtr":[24,6,1,"_CPPv2N9Parameter19createFromSharedPtrEPv"],"Parameter::exec":[50,6,1,"_CPPv2N9Parameter4execE8ExecFunc"],"Parameter::getBuf":[24,6,1,"_CPPv2N9Parameter6getBufE13ParameterType"],"Parameter::getConfig":[24,6,1,"_CPPv2N9Parameter9getConfigEv"],"Parameter::getID":[24,6,1,"_CPPv2N9Parameter5getIDEv"],"Parameter::getName":[24,6,1,"_CPPv2N9Parameter7getNameEv"],"Parameter::getPrefetchMatrix":[50,6,1,"_CPPv2N9Parameter17getPrefetchMatrixEv"],"Parameter::getTlsTempBufs":[50,6,1,"_CPPv2N9Parameter14getTlsTempBufsEv"],"Parameter::incUpdate":[50,6,1,"_CPPv2N9Parameter9incUpdateERK14UpdateCallback"],"Parameter::initialize":[50,6,1,"_CPPv2N9Parameter10initializeEv"],"Parameter::isGradShared":[50,6,1,"_CPPv2N9Parameter12isGradSharedEP6size_t"],"Parameter::isGradSparseUpdate":[50,6,1,"_CPPv2N9Parameter18isGradSparseUpdateEv"],"Parameter::isValueShared":[50,6,1,"_CPPv2N9Parameter13isValueSharedEv"],"Parameter::kMissParameterFail":[50,7,1,"_CPPv2N9Parameter18kMissParameterFailE"],"Parameter::kMissParameterRand":[50,7,1,"_CPPv2N9Parameter18kMissParameterRandE"],"Parameter::kMissParameterZero":[50,7,1,"_CPPv2N9Parameter18kMissParameterZeroE"],"Parameter::load":[50,6,1,"_CPPv2N9Parameter4loadERKNSt6stringE"],"Parameter::m":[24,7,1,"_CPPv2N9Parameter1mE"],"Parameter::randomize":[50,6,1,"_CPPv2N9Parameter9randomizeEv"],"Parameter::save":[50,6,1,"_CPPv2N9Parameter4saveERKNSt6stringE"],"Parameter::setMat":[50,6,1,"_CPPv2N9Parameter6setMatE13ParameterTypei"],"Parameter::tlsTempBufs_":[50,7,1,"_CPPv2N9Parameter12tlsTempBufs_E"],"Parameter::updateWithGradient":[50,6,1,"_CPPv2N9Parameter18updateWithGradientE4real"],"Parameter::zeroMem":[50,6,1,"_CPPv2N9Parameter7zeroMemEv"],"Parameter::~Parameter":[24,6,1,"_CPPv2N9ParameterD0Ev"],"ParameterConfig::DISABLE_COPY_AND_ASSIGN":[24,6,1,"_CPPv2N15ParameterConfig23DISABLE_COPY_AND_ASSIGNE15ParameterConfig"],"ParameterConfig::ParameterConfig":[24,6,1,"_CPPv2N15ParameterConfig15ParameterConfigEv"],"ParameterConfig::createParameterConfigFromParameterPtr":[24,6,1,"_CPPv2N15ParameterConfig37createParameterConfigFromParameterPtrEPv"],"ParameterConfig::createParameterConfigFromParameterSharedPtr":[24,6,1,"_CPPv2N15ParameterConfig43createParameterConfigFromParameterSharedPtrEPv"],"ParameterConfig::getRawPtr":[24,6,1,"_CPPv2N15ParameterConfig9getRawPtrEv"],"ParameterConfig::m":[24,7,1,"_CPPv2N15ParameterConfig1mE"],"ParameterConfig::toProtoString":[24,6,1,"_CPPv2N15ParameterConfig13toProtoStringEv"],"ParameterConfig::~ParameterConfig":[24,6,1,"_CPPv2N15ParameterConfigD0Ev"],"ParameterOptimizer::DISABLE_COPY_AND_ASSIGN":[24,6,1,"_CPPv2N18ParameterOptimizer23DISABLE_COPY_AND_ASSIGNE18ParameterOptimizer"],"ParameterOptimizer::ParameterOptimizer":[24,6,1,"_CPPv2N18ParameterOptimizer18ParameterOptimizerEv"],"ParameterOptimizer::create":[24,6,1,"_CPPv2N18ParameterOptimizer6createEP18OptimizationConfig"],"ParameterOptimizer::finishBatch":[24,6,1,"_CPPv2N18ParameterOptimizer11finishBatchEv"],"ParameterOptimizer::finishPass":[24,6,1,"_CPPv2N18ParameterOptimizer10finishPassEv"],"ParameterOptimizer::getParameterTypes":[24,6,1,"_CPPv2N18ParameterOptimizer17getParameterTypesEv"],"ParameterOptimizer::init":[24,6,1,"_CPPv2N18ParameterOptimizer4initE6size_tPK15ParameterConfig"],"ParameterOptimizer::m":[24,7,1,"_CPPv2N18ParameterOptimizer1mE"],"ParameterOptimizer::needSpecialTraversal":[24,6,1,"_CPPv2N18ParameterOptimizer20needSpecialTraversalERK15ParameterConfig"],"ParameterOptimizer::startBatch":[24,6,1,"_CPPv2N18ParameterOptimizer10startBatchE6size_t"],"ParameterOptimizer::startPass":[24,6,1,"_CPPv2N18ParameterOptimizer9startPassEv"],"ParameterOptimizer::update":[24,6,1,"_CPPv2N18ParameterOptimizer6updateERKNSt6vectorIP6VectorEERK15ParameterConfig6size_t"],"ParameterOptimizer::~ParameterOptimizer":[24,6,1,"_CPPv2N18ParameterOptimizerD0Ev"],"ParameterTraverseCallback::DISABLE_COPY_AND_ASSIGN":[24,6,1,"_CPPv2N25ParameterTraverseCallback23DISABLE_COPY_AND_ASSIGNE25ParameterTraverseCallback"],"ParameterTraverseCallback::ParameterTraverseCallback":[24,6,1,"_CPPv2N25ParameterTraverseCallback25ParameterTraverseCallbackEv"],"ParameterTraverseCallback::apply":[24,6,1,"_CPPv2N25ParameterTraverseCallback5applyERKNSt6vectorIP6VectorEERK15ParameterConfig6size_t"],"ParameterTraverseCallback::m":[24,7,1,"_CPPv2N25ParameterTraverseCallback1mE"],"ParameterTraverseCallback::~ParameterTraverseCallback":[24,6,1,"_CPPv2N25ParameterTraverseCallbackD0Ev"],"SequenceGenerator::DISABLE_COPY_AND_ASSIGN":[24,6,1,"_CPPv2N17SequenceGenerator23DISABLE_COPY_AND_ASSIGNE17SequenceGenerator"],"SequenceGenerator::SequenceGenerator":[24,6,1,"_CPPv2N17SequenceGenerator17SequenceGeneratorEv"],"SequenceGenerator::createByGradientMachineSharedPtr":[24,6,1,"_CPPv2N17SequenceGenerator32createByGradientMachineSharedPtrEPv"],"SequenceGenerator::generateSequence":[24,6,1,"_CPPv2N17SequenceGenerator16generateSequenceERK9Arguments"],"SequenceGenerator::m":[24,7,1,"_CPPv2N17SequenceGenerator1mE"],"SequenceGenerator::setBeamSize":[24,6,1,"_CPPv2N17SequenceGenerator11setBeamSizeE6size_t"],"SequenceGenerator::setBos":[24,6,1,"_CPPv2N17SequenceGenerator6setBosE6size_t"],"SequenceGenerator::setDict":[24,6,1,"_CPPv2N17SequenceGenerator7setDictERKNSt6vectorINSt6stringEEE"],"SequenceGenerator::setEos":[24,6,1,"_CPPv2N17SequenceGenerator6setEosE6size_t"],"SequenceGenerator::setMaxLength":[24,6,1,"_CPPv2N17SequenceGenerator12setMaxLengthE6size_t"],"SequenceGenerator::~SequenceGenerator":[24,6,1,"_CPPv2N17SequenceGeneratorD0Ev"],"Trainer::DISABLE_COPY_AND_ASSIGN":[24,6,1,"_CPPv2N7Trainer23DISABLE_COPY_AND_ASSIGNE7Trainer"],"Trainer::Trainer":[24,6,1,"_CPPv2N7Trainer7TrainerEv"],"Trainer::createByCommandLine":[24,6,1,"_CPPv2N7Trainer19createByCommandLineEv"],"Trainer::finishTrain":[24,6,1,"_CPPv2N7Trainer11finishTrainEv"],"Trainer::finishTrainOneBatch":[24,6,1,"_CPPv2N7Trainer19finishTrainOneBatchEv"],"Trainer::finishTrainPass":[24,6,1,"_CPPv2N7Trainer15finishTrainPassEv"],"Trainer::forwardOneBatch":[24,6,1,"_CPPv2N7Trainer15forwardOneBatchEv"],"Trainer::getLayerOutput":[24,6,1,"_CPPv2N7Trainer14getLayerOutputERKNSt6stringE"],"Trainer::getNetworkOutput":[24,6,1,"_CPPv2N7Trainer16getNetworkOutputEv"],"Trainer::m":[24,7,1,"_CPPv2N7Trainer1mE"],"Trainer::prepareBatchData":[24,6,1,"_CPPv2N7Trainer16prepareBatchDataE6size_t"],"Trainer::setBatchSize":[24,6,1,"_CPPv2N7Trainer12setBatchSizeE6size_t"],"Trainer::startTrain":[24,6,1,"_CPPv2N7Trainer10startTrainEv"],"Trainer::startTrainPass":[24,6,1,"_CPPv2N7Trainer14startTrainPassEv"],"Trainer::trainOneBatch":[24,6,1,"_CPPv2N7Trainer13trainOneBatchE6size_t"],"Trainer::~Trainer":[24,6,1,"_CPPv2N7TrainerD0Ev"],"TrainerConfig::DISABLE_COPY_AND_ASSIGN":[24,6,1,"_CPPv2N13TrainerConfig23DISABLE_COPY_AND_ASSIGNE13TrainerConfig"],"TrainerConfig::TrainerConfig":[24,6,1,"_CPPv2N13TrainerConfig13TrainerConfigEv"],"TrainerConfig::createFromTrainerConfigFile":[24,6,1,"_CPPv2N13TrainerConfig27createFromTrainerConfigFileERKNSt6stringE"],"TrainerConfig::getModelConfig":[24,6,1,"_CPPv2N13TrainerConfig14getModelConfigEv"],"TrainerConfig::getOptimizationConfig":[24,6,1,"_CPPv2N13TrainerConfig21getOptimizationConfigEv"],"TrainerConfig::m":[24,7,1,"_CPPv2N13TrainerConfig1mE"],"TrainerConfig::~TrainerConfig":[24,6,1,"_CPPv2N13TrainerConfigD0Ev"],"UpdateCallback::apply":[24,6,1,"_CPPv2N14UpdateCallback5applyEP9Parameter"],"UpdateCallback::~UpdateCallback":[24,6,1,"_CPPv2N14UpdateCallbackD0Ev"],"Vector::DISABLE_COPY_AND_ASSIGN":[24,6,1,"_CPPv2N6Vector23DISABLE_COPY_AND_ASSIGNE6Vector"],"Vector::Vector":[24,6,1,"_CPPv2N6Vector6VectorEv"],"Vector::copyFromNumpyArray":[24,6,1,"_CPPv2N6Vector18copyFromNumpyArrayEPfi"],"Vector::copyToNumpyArray":[24,6,1,"_CPPv2N6Vector16copyToNumpyArrayEPPfPi"],"Vector::create":[24,6,1,"_CPPv2N6Vector6createERKNSt6vectorIfEEb"],"Vector::createByPaddleVectorPtr":[24,6,1,"_CPPv2N6Vector23createByPaddleVectorPtrEPv"],"Vector::createCpuVectorFromNumpy":[24,6,1,"_CPPv2N6Vector24createCpuVectorFromNumpyEPfib"],"Vector::createGpuVectorFromNumpy":[24,6,1,"_CPPv2N6Vector24createGpuVectorFromNumpyEPfi"],"Vector::createZero":[24,6,1,"_CPPv2N6Vector10createZeroE6size_tb"],"Vector::get":[24,6,1,"_CPPv2N6Vector3getEK6size_t"],"Vector::getSharedPtr":[24,6,1,"_CPPv2N6Vector12getSharedPtrEv"],"Vector::getSize":[24,6,1,"_CPPv2N6Vector7getSizeEv"],"Vector::isGpu":[24,6,1,"_CPPv2N6Vector5isGpuEv"],"Vector::m":[24,7,1,"_CPPv2N6Vector1mE"],"Vector::set":[24,6,1,"_CPPv2N6Vector3setEK6size_tf"],"Vector::toNumpyArrayInplace":[24,6,1,"_CPPv2N6Vector19toNumpyArrayInplaceEPPfPi"],"Vector::~Vector":[24,6,1,"_CPPv2N6VectorD0Ev"],"_hl_sparse_matrix_s":[32,8,1,"_CPPv219_hl_sparse_matrix_s"],"_hl_sparse_matrix_s::cols":[32,7,1,"_CPPv2N19_hl_sparse_matrix_s4colsE"],"_hl_sparse_matrix_s::format":[32,7,1,"_CPPv2N19_hl_sparse_matrix_s6formatE"],"_hl_sparse_matrix_s::matrix":[32,7,1,"_CPPv2N19_hl_sparse_matrix_s6matrixE"],"_hl_sparse_matrix_s::nnz":[32,7,1,"_CPPv2N19_hl_sparse_matrix_s3nnzE"],"_hl_sparse_matrix_s::rows":[32,7,1,"_CPPv2N19_hl_sparse_matrix_s4rowsE"],"_hl_sparse_matrix_s::type":[32,7,1,"_CPPv2N19_hl_sparse_matrix_s4typeE"],"hl_gru_grad::gateGrad":[32,7,1,"_CPPv2N11hl_gru_grad8gateGradE"],"hl_gru_grad::gateWeightGrad":[32,7,1,"_CPPv2N11hl_gru_grad14gateWeightGradE"],"hl_gru_grad::outputGrad":[32,7,1,"_CPPv2N11hl_gru_grad10outputGradE"],"hl_gru_grad::prevOutGrad":[32,7,1,"_CPPv2N11hl_gru_grad11prevOutGradE"],"hl_gru_grad::resetOutputGrad":[32,7,1,"_CPPv2N11hl_gru_grad15resetOutputGradE"],"hl_gru_grad::stateWeightGrad":[32,7,1,"_CPPv2N11hl_gru_grad15stateWeightGradE"],"hl_gru_value::gateValue":[32,7,1,"_CPPv2N12hl_gru_value9gateValueE"],"hl_gru_value::gateWeight":[32,7,1,"_CPPv2N12hl_gru_value10gateWeightE"],"hl_gru_value::outputValue":[32,7,1,"_CPPv2N12hl_gru_value11outputValueE"],"hl_gru_value::prevOutValue":[32,7,1,"_CPPv2N12hl_gru_value12prevOutValueE"],"hl_gru_value::resetOutputValue":[32,7,1,"_CPPv2N12hl_gru_value16resetOutputValueE"],"hl_gru_value::stateWeight":[32,7,1,"_CPPv2N12hl_gru_value11stateWeightE"],"hl_lstm_grad::checkFgGrad":[32,7,1,"_CPPv2N12hl_lstm_grad11checkFgGradE"],"hl_lstm_grad::checkIgGrad":[32,7,1,"_CPPv2N12hl_lstm_grad11checkIgGradE"],"hl_lstm_grad::checkOgGrad":[32,7,1,"_CPPv2N12hl_lstm_grad11checkOgGradE"],"hl_lstm_grad::gateGrad":[32,7,1,"_CPPv2N12hl_lstm_grad8gateGradE"],"hl_lstm_grad::outputGrad":[32,7,1,"_CPPv2N12hl_lstm_grad10outputGradE"],"hl_lstm_grad::prevStateGrad":[32,7,1,"_CPPv2N12hl_lstm_grad13prevStateGradE"],"hl_lstm_grad::stateActiveGrad":[32,7,1,"_CPPv2N12hl_lstm_grad15stateActiveGradE"],"hl_lstm_grad::stateGrad":[32,7,1,"_CPPv2N12hl_lstm_grad9stateGradE"],"hl_lstm_value::checkFg":[32,7,1,"_CPPv2N13hl_lstm_value7checkFgE"],"hl_lstm_value::checkIg":[32,7,1,"_CPPv2N13hl_lstm_value7checkIgE"],"hl_lstm_value::checkOg":[32,7,1,"_CPPv2N13hl_lstm_value7checkOgE"],"hl_lstm_value::gateValue":[32,7,1,"_CPPv2N13hl_lstm_value9gateValueE"],"hl_lstm_value::outputValue":[32,7,1,"_CPPv2N13hl_lstm_value11outputValueE"],"hl_lstm_value::prevStateValue":[32,7,1,"_CPPv2N13hl_lstm_value14prevStateValueE"],"hl_lstm_value::stateActiveValue":[32,7,1,"_CPPv2N13hl_lstm_value16stateActiveValueE"],"hl_lstm_value::stateValue":[32,7,1,"_CPPv2N13hl_lstm_value10stateValueE"],"hppl::Active":[30,8,1,"_CPPv2N4hppl6ActiveE"],"hppl::Active::backward":[30,9,1,"_CPPv2N4hppl6ActiveI1TE8backwardE"],"hppl::Active::forward":[30,9,1,"_CPPv2N4hppl6ActiveI1TE7forwardE"],"hppl::cpu":[30,9,1,"_CPPv2N4hppl3cpuE"],"hppl::cpu::backward":[30,7,1,"_CPPv2N4hppl3cpu8backwardE"],"hppl::cpu::forward":[30,7,1,"_CPPv2N4hppl3cpu7forwardE"],"hppl::linear":[30,6,1,"_CPPv2N4hppl6linearEK4real"],"hppl::relu":[30,6,1,"_CPPv2N4hppl4reluEK4real"],"hppl::sigmoid":[30,6,1,"_CPPv2N4hppl7sigmoidEK4real"],"hppl::tanh":[30,6,1,"_CPPv2N4hppl4tanhEK4real"],"operator<<":[36,6,1,"_CPPv2lsRNSt7ostreamER9Evaluator"],"paddle::ActivationFunction":[33,8,1,"_CPPv2N6paddle18ActivationFunctionE"],"paddle::ActivationFunction::ActivationFunction":[33,6,1,"_CPPv2N6paddle18ActivationFunction18ActivationFunctionEv"],"paddle::ActivationFunction::backward":[33,6,1,"_CPPv2N6paddle18ActivationFunction8backwardER8Argument"],"paddle::ActivationFunction::create":[33,6,1,"_CPPv2N6paddle18ActivationFunction6createERKNSt6stringE"],"paddle::ActivationFunction::forward":[33,6,1,"_CPPv2N6paddle18ActivationFunction7forwardER8Argument"],"paddle::ActivationFunction::getName":[33,6,1,"_CPPv2N6paddle18ActivationFunction7getNameEv"],"paddle::ActivationFunction::~ActivationFunction":[33,6,1,"_CPPv2N6paddle18ActivationFunctionD0Ev"],"paddle::AdaDeltaParameterOptimizer":[48,8,1,"_CPPv2N6paddle26AdaDeltaParameterOptimizerE"],"paddle::AdaDeltaParameterOptimizer::AdaDeltaParameterOptimizer":[48,6,1,"_CPPv2N6paddle26AdaDeltaParameterOptimizer26AdaDeltaParameterOptimizerERK18OptimizationConfig"],"paddle::AdaDeltaParameterOptimizer::epsilon_":[48,7,1,"_CPPv2N6paddle26AdaDeltaParameterOptimizer8epsilon_E"],"paddle::AdaDeltaParameterOptimizer::rou_":[48,7,1,"_CPPv2N6paddle26AdaDeltaParameterOptimizer4rou_E"],"paddle::AdaDeltaParameterOptimizer::startBatch":[48,6,1,"_CPPv2N6paddle26AdaDeltaParameterOptimizer10startBatchE7int64_t"],"paddle::AdaDeltaParameterOptimizer::update":[48,6,1,"_CPPv2N6paddle26AdaDeltaParameterOptimizer6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::AdagradParameterOptimizer":[48,8,1,"_CPPv2N6paddle25AdagradParameterOptimizerE"],"paddle::AdagradParameterOptimizer::AdagradParameterOptimizer":[48,6,1,"_CPPv2N6paddle25AdagradParameterOptimizer25AdagradParameterOptimizerERK18OptimizationConfig"],"paddle::AdagradParameterOptimizer::kMaxNumAccumulates":[48,7,1,"_CPPv2N6paddle25AdagradParameterOptimizer18kMaxNumAccumulatesE"],"paddle::AdagradParameterOptimizer::needSpecialTraversal":[48,6,1,"_CPPv2N6paddle25AdagradParameterOptimizer20needSpecialTraversalERK15ParameterConfig"],"paddle::AdagradParameterOptimizer::numUpdates_":[48,7,1,"_CPPv2N6paddle25AdagradParameterOptimizer11numUpdates_E"],"paddle::AdagradParameterOptimizer::startBatch":[48,6,1,"_CPPv2N6paddle25AdagradParameterOptimizer10startBatchE7int64_t"],"paddle::AdagradParameterOptimizer::update":[48,6,1,"_CPPv2N6paddle25AdagradParameterOptimizer6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::AdamParameterOptimizer":[48,8,1,"_CPPv2N6paddle22AdamParameterOptimizerE"],"paddle::AdamParameterOptimizer::AdamParameterOptimizer":[48,6,1,"_CPPv2N6paddle22AdamParameterOptimizer22AdamParameterOptimizerERK18OptimizationConfig"],"paddle::AdamParameterOptimizer::beta1_":[48,7,1,"_CPPv2N6paddle22AdamParameterOptimizer6beta1_E"],"paddle::AdamParameterOptimizer::beta2_":[48,7,1,"_CPPv2N6paddle22AdamParameterOptimizer6beta2_E"],"paddle::AdamParameterOptimizer::epsilon_":[48,7,1,"_CPPv2N6paddle22AdamParameterOptimizer8epsilon_E"],"paddle::AdamParameterOptimizer::finishBatch":[48,6,1,"_CPPv2N6paddle22AdamParameterOptimizer11finishBatchEv"],"paddle::AdamParameterOptimizer::learningRate_":[48,7,1,"_CPPv2N6paddle22AdamParameterOptimizer13learningRate_E"],"paddle::AdamParameterOptimizer::step_":[48,7,1,"_CPPv2N6paddle22AdamParameterOptimizer5step_E"],"paddle::AdamParameterOptimizer::update":[48,6,1,"_CPPv2N6paddle22AdamParameterOptimizer6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::AdamaxParameterOptimizer":[48,8,1,"_CPPv2N6paddle24AdamaxParameterOptimizerE"],"paddle::AdamaxParameterOptimizer::AdamaxParameterOptimizer":[48,6,1,"_CPPv2N6paddle24AdamaxParameterOptimizer24AdamaxParameterOptimizerERK18OptimizationConfig"],"paddle::AdamaxParameterOptimizer::beta1_":[48,7,1,"_CPPv2N6paddle24AdamaxParameterOptimizer6beta1_E"],"paddle::AdamaxParameterOptimizer::beta2_":[48,7,1,"_CPPv2N6paddle24AdamaxParameterOptimizer6beta2_E"],"paddle::AdamaxParameterOptimizer::finishBatch":[48,6,1,"_CPPv2N6paddle24AdamaxParameterOptimizer11finishBatchEv"],"paddle::AdamaxParameterOptimizer::learningRate_":[48,7,1,"_CPPv2N6paddle24AdamaxParameterOptimizer13learningRate_E"],"paddle::AdamaxParameterOptimizer::step_":[48,7,1,"_CPPv2N6paddle24AdamaxParameterOptimizer5step_E"],"paddle::AdamaxParameterOptimizer::update":[48,6,1,"_CPPv2N6paddle24AdamaxParameterOptimizer6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::AddOptimizer":[48,8,1,"_CPPv2N6paddle12AddOptimizerE"],"paddle::AddOptimizer::AddOptimizer":[48,6,1,"_CPPv2N6paddle12AddOptimizer12AddOptimizerERK18OptimizationConfig"],"paddle::AddOptimizer::startBatch":[48,6,1,"_CPPv2N6paddle12AddOptimizer10startBatchE7int64_t"],"paddle::AddOptimizer::update":[48,6,1,"_CPPv2N6paddle12AddOptimizer6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::AddtoLayer":[41,8,1,"_CPPv2N6paddle10AddtoLayerE"],"paddle::AddtoLayer::AddtoLayer":[41,6,1,"_CPPv2N6paddle10AddtoLayer10AddtoLayerERK11LayerConfig"],"paddle::AddtoLayer::backward":[41,6,1,"_CPPv2N6paddle10AddtoLayer8backwardERK14UpdateCallback"],"paddle::AddtoLayer::biases_":[41,7,1,"_CPPv2N6paddle10AddtoLayer7biases_E"],"paddle::AddtoLayer::forward":[41,6,1,"_CPPv2N6paddle10AddtoLayer7forwardE8PassType"],"paddle::AddtoLayer::init":[41,6,1,"_CPPv2N6paddle10AddtoLayer4initERK8LayerMapRK12ParameterMap"],"paddle::AddtoLayer::~AddtoLayer":[41,6,1,"_CPPv2N6paddle10AddtoLayerD0Ev"],"paddle::AgentLayer":[41,8,1,"_CPPv2N6paddle10AgentLayerE"],"paddle::AgentLayer::AgentLayer":[41,6,1,"_CPPv2N6paddle10AgentLayer10AgentLayerERK11LayerConfig"],"paddle::AgentLayer::backward":[41,6,1,"_CPPv2N6paddle10AgentLayer8backwardERK14UpdateCallback"],"paddle::AgentLayer::forward":[41,6,1,"_CPPv2N6paddle10AgentLayer7forwardE8PassType"],"paddle::AgentLayer::init":[41,6,1,"_CPPv2N6paddle10AgentLayer4initERK8LayerMapRK12ParameterMap"],"paddle::AgentLayer::numSamples_":[41,7,1,"_CPPv2N6paddle10AgentLayer11numSamples_E"],"paddle::AgentLayer::realLayer_":[41,7,1,"_CPPv2N6paddle10AgentLayer10realLayer_E"],"paddle::AgentLayer::setRealLayer":[41,6,1,"_CPPv2N6paddle10AgentLayer12setRealLayerE8LayerPtri"],"paddle::AgentLayer::~AgentLayer":[41,6,1,"_CPPv2N6paddle10AgentLayerD0Ev"],"paddle::Allocator":[46,8,1,"_CPPv2N6paddle9AllocatorE"],"paddle::Allocator::alloc":[46,6,1,"_CPPv2N6paddle9Allocator5allocE6size_t"],"paddle::Allocator::free":[46,6,1,"_CPPv2N6paddle9Allocator4freeEPv"],"paddle::Allocator::getName":[46,6,1,"_CPPv2N6paddle9Allocator7getNameEv"],"paddle::Allocator::~Allocator":[46,6,1,"_CPPv2N6paddle9AllocatorD0Ev"],"paddle::Argument":[50,8,1,"_CPPv2N6paddle8ArgumentE"],"paddle::Argument::Argument":[50,6,1,"_CPPv2N6paddle8Argument8ArgumentEv"],"paddle::Argument::allCount":[50,7,1,"_CPPv2N6paddle8Argument8allCountE"],"paddle::Argument::checkSubset":[50,6,1,"_CPPv2N6paddle8Argument11checkSubsetEv"],"paddle::Argument::concat":[50,6,1,"_CPPv2N6paddle8Argument6concatERKNSt6vectorI8ArgumentEERKNSt6vectorIiEERKNSt6vectorIiEEb11hl_stream_t8PassType"],"paddle::Argument::countIncrement":[50,6,1,"_CPPv2N6paddle8Argument14countIncrementEv"],"paddle::Argument::cpuSequenceDims":[50,7,1,"_CPPv2N6paddle8Argument15cpuSequenceDimsE"],"paddle::Argument::dataId":[50,7,1,"_CPPv2N6paddle8Argument6dataIdE"],"paddle::Argument::degradeSequence":[50,6,1,"_CPPv2N6paddle8Argument15degradeSequenceERK8Argumentb"],"paddle::Argument::deviceId":[50,7,1,"_CPPv2N6paddle8Argument8deviceIdE"],"paddle::Argument::frameHeight":[50,7,1,"_CPPv2N6paddle8Argument11frameHeightE"],"paddle::Argument::frameWidth":[50,7,1,"_CPPv2N6paddle8Argument10frameWidthE"],"paddle::Argument::getAllCount":[50,6,1,"_CPPv2N6paddle8Argument11getAllCountEv"],"paddle::Argument::getBatchSize":[50,6,1,"_CPPv2N6paddle8Argument12getBatchSizeEv"],"paddle::Argument::getCpuStartPositions":[50,6,1,"_CPPv2N6paddle8Argument20getCpuStartPositionsEv"],"paddle::Argument::getFrameHeight":[50,6,1,"_CPPv2N6paddle8Argument14getFrameHeightEv"],"paddle::Argument::getFrameWidth":[50,6,1,"_CPPv2N6paddle8Argument13getFrameWidthEv"],"paddle::Argument::getNumSequences":[50,6,1,"_CPPv2N6paddle8Argument15getNumSequencesEv"],"paddle::Argument::getNumSubSequences":[50,6,1,"_CPPv2N6paddle8Argument18getNumSubSequencesEv"],"paddle::Argument::getSeqLengthAndStart":[50,6,1,"_CPPv2N6paddle8Argument20getSeqLengthAndStartEPNSt6vectorINSt5tupleIiiiiEEEEPi"],"paddle::Argument::grad":[50,7,1,"_CPPv2N6paddle8Argument4gradE"],"paddle::Argument::gradCount":[50,7,1,"_CPPv2N6paddle8Argument9gradCountE"],"paddle::Argument::gradReadyCond":[50,7,1,"_CPPv2N6paddle8Argument13gradReadyCondE"],"paddle::Argument::hasSubseq":[50,6,1,"_CPPv2N6paddle8Argument9hasSubseqEv"],"paddle::Argument::ids":[50,7,1,"_CPPv2N6paddle8Argument3idsE"],"paddle::Argument::in":[50,7,1,"_CPPv2N6paddle8Argument2inE"],"paddle::Argument::notifyGradReady":[50,6,1,"_CPPv2N6paddle8Argument15notifyGradReadyEv"],"paddle::Argument::notifyValueReady":[50,6,1,"_CPPv2N6paddle8Argument16notifyValueReadyEv"],"paddle::Argument::operator=":[50,6,1,"_CPPv2N6paddle8ArgumentaSERK8Argument"],"paddle::Argument::resizeAndCopyFrom":[50,6,1,"_CPPv2N6paddle8Argument17resizeAndCopyFromERK8Argument7int32_t7int32_tb11hl_stream_t"],"paddle::Argument::sequenceStartPositions":[50,7,1,"_CPPv2N6paddle8Argument22sequenceStartPositionsE"],"paddle::Argument::setFrameHeight":[50,6,1,"_CPPv2N6paddle8Argument14setFrameHeightE6size_t"],"paddle::Argument::setFrameWidth":[50,6,1,"_CPPv2N6paddle8Argument13setFrameWidthE6size_t"],"paddle::Argument::splitByDataId":[50,6,1,"_CPPv2N6paddle8Argument13splitByDataIdERKNSt6vectorI8ArgumentEEPNSt6vectorINSt6vectorI8ArgumentEEEE"],"paddle::Argument::strs":[50,7,1,"_CPPv2N6paddle8Argument4strsE"],"paddle::Argument::subArgFrom":[50,6,1,"_CPPv2N6paddle8Argument10subArgFromERK8Argument6size_t6size_t6size_tbbb6size_t6size_t"],"paddle::Argument::subSequenceStartPositions":[50,7,1,"_CPPv2N6paddle8Argument25subSequenceStartPositionsE"],"paddle::Argument::sumCosts":[50,6,1,"_CPPv2N6paddle8Argument8sumCostsERKNSt6vectorI8ArgumentEE"],"paddle::Argument::udp":[50,7,1,"_CPPv2N6paddle8Argument3udpE"],"paddle::Argument::value":[50,7,1,"_CPPv2N6paddle8Argument5valueE"],"paddle::Argument::valueCount":[50,7,1,"_CPPv2N6paddle8Argument10valueCountE"],"paddle::Argument::valueReadyCond":[50,7,1,"_CPPv2N6paddle8Argument14valueReadyCondE"],"paddle::Argument::waitGradReady":[50,6,1,"_CPPv2N6paddle8Argument13waitGradReadyEv"],"paddle::Argument::waitValueReady":[50,6,1,"_CPPv2N6paddle8Argument14waitValueReadyEv"],"paddle::Argument::~Argument":[50,6,1,"_CPPv2N6paddle8ArgumentD0Ev"],"paddle::AsyncParameter":[50,8,1,"_CPPv2N6paddle14AsyncParameterE"],"paddle::AsyncParameter::AsyncParameter":[50,6,1,"_CPPv2N6paddle14AsyncParameter14AsyncParameterE11TrainerRolei12ParameterPtr"],"paddle::AsyncParameter::accumCounter_":[50,7,1,"_CPPv2N6paddle14AsyncParameter13accumCounter_E"],"paddle::AsyncParameter::asyncCount_":[50,7,1,"_CPPv2N6paddle14AsyncParameter11asyncCount_E"],"paddle::AsyncParameter::clearCounter":[50,6,1,"_CPPv2N6paddle14AsyncParameter12clearCounterEv"],"paddle::AsyncParameter::getAccum":[50,6,1,"_CPPv2N6paddle14AsyncParameter8getAccumEv"],"paddle::AsyncParameter::gradientAccum_":[50,7,1,"_CPPv2N6paddle14AsyncParameter14gradientAccum_E"],"paddle::AsyncParameter::masterUpdate":[50,6,1,"_CPPv2N6paddle14AsyncParameter12masterUpdateE20ParallelParameterPtrRK14UpdateCallback"],"paddle::AsyncParameter::slaveUpdate":[50,6,1,"_CPPv2N6paddle14AsyncParameter11slaveUpdateE4real"],"paddle::AsyncParameter::synchronizeParamter":[50,6,1,"_CPPv2N6paddle14AsyncParameter19synchronizeParamterEv"],"paddle::AsyncThreadPool":[62,8,1,"_CPPv2N6paddle15AsyncThreadPoolE"],"paddle::AsyncThreadPool::AsyncThreadPool":[62,6,1,"_CPPv2N6paddle15AsyncThreadPool15AsyncThreadPoolEv"],"paddle::AsyncThreadPool::JobFunc":[62,9,1,"_CPPv2N6paddle15AsyncThreadPool7JobFuncE"],"paddle::AsyncThreadPool::ProcessChannel":[62,6,1,"_CPPv2N6paddle15AsyncThreadPool14ProcessChannelE6size_t"],"paddle::AsyncThreadPool::addBatchJobs":[62,6,1,"_CPPv2N6paddle15AsyncThreadPool12addBatchJobsERKNSt6vectorI1FEERNSt6vectorINSt9result_ofIF1FvEE4typeEEE"],"paddle::AsyncThreadPool::addJob":[62,6,1,"_CPPv2N6paddle15AsyncThreadPool6addJobERR1FDpRR4Args"],"paddle::AsyncThreadPool::run":[62,6,1,"_CPPv2N6paddle15AsyncThreadPool3runEv"],"paddle::AsyncThreadPool::stop":[62,6,1,"_CPPv2N6paddle15AsyncThreadPool4stopEv"],"paddle::AsyncThreadPool::~AsyncThreadPool":[62,6,1,"_CPPv2N6paddle15AsyncThreadPoolD0Ev"],"paddle::AucEvaluator":[36,8,1,"_CPPv2N6paddle12AucEvaluatorE"],"paddle::AucEvaluator::AucEvaluator":[36,6,1,"_CPPv2N6paddle12AucEvaluator12AucEvaluatorE7int32_t"],"paddle::AucEvaluator::distributeEval":[36,6,1,"_CPPv2N6paddle12AucEvaluator14distributeEvalEP16ParameterClient2"],"paddle::AucEvaluator::evalImp":[36,6,1,"_CPPv2N6paddle12AucEvaluator7evalImpERNSt6vectorI8ArgumentEE"],"paddle::AucEvaluator::printStats":[36,6,1,"_CPPv2N6paddle12AucEvaluator10printStatsERNSt7ostreamE"],"paddle::AucEvaluator::start":[36,6,1,"_CPPv2N6paddle12AucEvaluator5startEv"],"paddle::AucValidation":[41,8,1,"_CPPv2N6paddle13AucValidationE"],"paddle::AucValidation::AucValidation":[41,6,1,"_CPPv2N6paddle13AucValidation13AucValidationERK11LayerConfig"],"paddle::AucValidation::PredictionResult":[41,8,1,"_CPPv2N6paddle13AucValidation16PredictionResultE"],"paddle::AucValidation::PredictionResult::PredictionResult":[41,6,1,"_CPPv2N6paddle13AucValidation16PredictionResult16PredictionResultE4reali"],"paddle::AucValidation::PredictionResult::label":[41,7,1,"_CPPv2N6paddle13AucValidation16PredictionResult5labelE"],"paddle::AucValidation::PredictionResult::out":[41,7,1,"_CPPv2N6paddle13AucValidation16PredictionResult3outE"],"paddle::AucValidation::init":[41,6,1,"_CPPv2N6paddle13AucValidation4initERK8LayerMapRK12ParameterMap"],"paddle::AucValidation::onPassEnd":[41,6,1,"_CPPv2N6paddle13AucValidation9onPassEndEv"],"paddle::AucValidation::predictArray_":[41,7,1,"_CPPv2N6paddle13AucValidation13predictArray_E"],"paddle::AucValidation::validationImp":[41,6,1,"_CPPv2N6paddle13AucValidation13validationImpE9MatrixPtr10IVectorPtr"],"paddle::AverageLayer":[41,8,1,"_CPPv2N6paddle12AverageLayerE"],"paddle::AverageLayer::AverageLayer":[41,6,1,"_CPPv2N6paddle12AverageLayer12AverageLayerERK11LayerConfig"],"paddle::AverageLayer::AverageLevel":[41,9,1,"_CPPv2N6paddle12AverageLayer12AverageLevelE"],"paddle::AverageLayer::AverageStrategy":[41,9,1,"_CPPv2N6paddle12AverageLayer15AverageStrategyE"],"paddle::AverageLayer::backward":[41,6,1,"_CPPv2N6paddle12AverageLayer8backwardERK14UpdateCallback"],"paddle::AverageLayer::biases_":[41,7,1,"_CPPv2N6paddle12AverageLayer7biases_E"],"paddle::AverageLayer::dataMtx_":[41,7,1,"_CPPv2N6paddle12AverageLayer8dataMtx_E"],"paddle::AverageLayer::forward":[41,6,1,"_CPPv2N6paddle12AverageLayer7forwardE8PassType"],"paddle::AverageLayer::init":[41,6,1,"_CPPv2N6paddle12AverageLayer4initERK8LayerMapRK12ParameterMap"],"paddle::AverageLayer::kAverage":[41,8,1,"_CPPv2N6paddle12AverageLayer8kAverageE"],"paddle::AverageLayer::kAverageSquareRootN":[41,8,1,"_CPPv2N6paddle12AverageLayer19kAverageSquareRootNE"],"paddle::AverageLayer::kNonSeq":[41,8,1,"_CPPv2N6paddle12AverageLayer7kNonSeqE"],"paddle::AverageLayer::kSeq":[41,8,1,"_CPPv2N6paddle12AverageLayer4kSeqE"],"paddle::AverageLayer::kSum":[41,8,1,"_CPPv2N6paddle12AverageLayer4kSumE"],"paddle::AverageLayer::mode_":[41,7,1,"_CPPv2N6paddle12AverageLayer5mode_E"],"paddle::AverageLayer::outMtx_":[41,7,1,"_CPPv2N6paddle12AverageLayer7outMtx_E"],"paddle::AverageLayer::type_":[41,7,1,"_CPPv2N6paddle12AverageLayer5type_E"],"paddle::AverageLayer::~AverageLayer":[41,6,1,"_CPPv2N6paddle12AverageLayerD0Ev"],"paddle::AverageOptimizer":[48,8,1,"_CPPv2N6paddle16AverageOptimizerE"],"paddle::AverageOptimizer::AverageOptimizer":[48,6,1,"_CPPv2N6paddle16AverageOptimizer16AverageOptimizerERK18OptimizationConfigP18ParameterOptimizerb"],"paddle::AverageOptimizer::apply":[48,6,1,"_CPPv2N6paddle16AverageOptimizer5applyEv"],"paddle::AverageOptimizer::create":[48,6,1,"_CPPv2N6paddle16AverageOptimizer6createERK18OptimizationConfigP18ParameterOptimizerbb"],"paddle::AverageOptimizer::finishBatch":[48,6,1,"_CPPv2N6paddle16AverageOptimizer11finishBatchEv"],"paddle::AverageOptimizer::finishCatchUpWith":[48,6,1,"_CPPv2N6paddle16AverageOptimizer17finishCatchUpWithEv"],"paddle::AverageOptimizer::finishPass":[48,6,1,"_CPPv2N6paddle16AverageOptimizer10finishPassEv"],"paddle::AverageOptimizer::init":[48,6,1,"_CPPv2N6paddle16AverageOptimizer4initE6size_tPK15ParameterConfig"],"paddle::AverageOptimizer::isAverageWindowTooLong":[48,6,1,"_CPPv2N6paddle16AverageOptimizer22isAverageWindowTooLongEv"],"paddle::AverageOptimizer::kMaxNumAccumulates":[48,7,1,"_CPPv2N6paddle16AverageOptimizer18kMaxNumAccumulatesE"],"paddle::AverageOptimizer::maxAverageWindow_":[48,7,1,"_CPPv2N6paddle16AverageOptimizer17maxAverageWindow_E"],"paddle::AverageOptimizer::minAverageWindow_":[48,7,1,"_CPPv2N6paddle16AverageOptimizer17minAverageWindow_E"],"paddle::AverageOptimizer::needSpecialTraversal":[48,6,1,"_CPPv2N6paddle16AverageOptimizer20needSpecialTraversalERK15ParameterConfig"],"paddle::AverageOptimizer::numAccumulates_":[48,7,1,"_CPPv2N6paddle16AverageOptimizer15numAccumulates_E"],"paddle::AverageOptimizer::numUpdates_":[48,7,1,"_CPPv2N6paddle16AverageOptimizer11numUpdates_E"],"paddle::AverageOptimizer::oldNumAccumulates_":[48,7,1,"_CPPv2N6paddle16AverageOptimizer18oldNumAccumulates_E"],"paddle::AverageOptimizer::optimizer_":[48,7,1,"_CPPv2N6paddle16AverageOptimizer10optimizer_E"],"paddle::AverageOptimizer::prevNumUpdates_":[48,7,1,"_CPPv2N6paddle16AverageOptimizer15prevNumUpdates_E"],"paddle::AverageOptimizer::restore":[48,6,1,"_CPPv2N6paddle16AverageOptimizer7restoreEv"],"paddle::AverageOptimizer::setNoDecay":[48,6,1,"_CPPv2N6paddle16AverageOptimizer10setNoDecayEv"],"paddle::AverageOptimizer::startBatch":[48,6,1,"_CPPv2N6paddle16AverageOptimizer10startBatchE7int64_t"],"paddle::AverageOptimizer::startCatchUpWith":[48,6,1,"_CPPv2N6paddle16AverageOptimizer16startCatchUpWithEv"],"paddle::AverageOptimizer::startPass":[48,6,1,"_CPPv2N6paddle16AverageOptimizer9startPassEv"],"paddle::AverageOptimizer::update":[48,6,1,"_CPPv2N6paddle16AverageOptimizer6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::AverageOptimizer::updateAverageWindowLimit":[48,6,1,"_CPPv2N6paddle16AverageOptimizer24updateAverageWindowLimitEv"],"paddle::AverageOptimizer::useApply_":[48,7,1,"_CPPv2N6paddle16AverageOptimizer9useApply_E"],"paddle::AverageSparseOptimizer":[48,8,1,"_CPPv2N6paddle22AverageSparseOptimizerE"],"paddle::AverageSparseOptimizer::AverageSparseOptimizer":[48,6,1,"_CPPv2N6paddle22AverageSparseOptimizer22AverageSparseOptimizerERK18OptimizationConfigP18ParameterOptimizerb"],"paddle::AverageSparseOptimizer::catchUpWith":[48,6,1,"_CPPv2N6paddle22AverageSparseOptimizer11catchUpWithEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::AverageSparseOptimizer::finishBatch":[48,6,1,"_CPPv2N6paddle22AverageSparseOptimizer11finishBatchEv"],"paddle::AverageSparseOptimizer::finishCatchUpWith":[48,6,1,"_CPPv2N6paddle22AverageSparseOptimizer17finishCatchUpWithEv"],"paddle::AverageSparseOptimizer::init":[48,6,1,"_CPPv2N6paddle22AverageSparseOptimizer4initE6size_tPK15ParameterConfig"],"paddle::AverageSparseOptimizer::startCatchUpWith":[48,6,1,"_CPPv2N6paddle22AverageSparseOptimizer16startCatchUpWithEv"],"paddle::AverageSparseOptimizer::t0Vec_":[48,7,1,"_CPPv2N6paddle22AverageSparseOptimizer6t0Vec_E"],"paddle::AverageSparseOptimizer::timer_":[48,7,1,"_CPPv2N6paddle22AverageSparseOptimizer6timer_E"],"paddle::AverageSparseOptimizer::update":[48,6,1,"_CPPv2N6paddle22AverageSparseOptimizer6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::BaseClient":[53,8,1,"_CPPv2N6paddle10BaseClientE"],"paddle::BaseClient::BaseClient":[53,6,1,"_CPPv2N6paddle10BaseClient10BaseClientEbi"],"paddle::BaseClient::InputIovs":[53,9,1,"_CPPv2N6paddle10BaseClient9InputIovsE"],"paddle::BaseClient::SendDataRequestVec":[53,9,1,"_CPPv2N6paddle10BaseClient18SendDataRequestVecE"],"paddle::BaseClient::SendJob":[53,8,1,"_CPPv2N6paddle10BaseClient7SendJobE"],"paddle::BaseClient::SendJob::parallelDataRequests":[53,7,1,"_CPPv2N6paddle10BaseClient7SendJob20parallelDataRequestsE"],"paddle::BaseClient::SendJob::parallelInputIovs":[53,7,1,"_CPPv2N6paddle10BaseClient7SendJob17parallelInputIovsE"],"paddle::BaseClient::SendJob::parallelRequests":[53,7,1,"_CPPv2N6paddle10BaseClient7SendJob16parallelRequestsE"],"paddle::BaseClient::SendJobPtr":[53,9,1,"_CPPv2N6paddle10BaseClient10SendJobPtrE"],"paddle::BaseClient::SendQueue":[53,9,1,"_CPPv2N6paddle10BaseClient9SendQueueE"],"paddle::BaseClient::SendRequest":[53,9,1,"_CPPv2N6paddle10BaseClient11SendRequestE"],"paddle::BaseClient::ThreadPtr":[53,9,1,"_CPPv2N6paddle10BaseClient9ThreadPtrE"],"paddle::BaseClient::calcClientId":[53,6,1,"_CPPv2N6paddle10BaseClient12calcClientIdEii"],"paddle::BaseClient::clients_":[53,7,1,"_CPPv2N6paddle10BaseClient8clients_E"],"paddle::BaseClient::divup":[53,6,1,"_CPPv2N6paddle10BaseClient5divupEii"],"paddle::BaseClient::finishThreads":[53,6,1,"_CPPv2N6paddle10BaseClient13finishThreadsEv"],"paddle::BaseClient::getAllData":[53,6,1,"_CPPv2N6paddle10BaseClient10getAllDataEi12SendDataTypeP8DataType6size_t"],"paddle::BaseClient::getTransDtype":[53,6,1,"_CPPv2N6paddle10BaseClient13getTransDtypeERKNSt9type_infoE"],"paddle::BaseClient::multiCall":[53,6,1,"_CPPv2N6paddle10BaseClient9multiCallEPKcRK7ProtoInPNSt6vectorI8ProtoOutEE"],"paddle::BaseClient::numPorts_":[53,7,1,"_CPPv2N6paddle10BaseClient9numPorts_E"],"paddle::BaseClient::prepareData":[53,6,1,"_CPPv2N6paddle10BaseClient11prepareDataEi12SendDataType14DataUpdateModeP8DataType6size_tP7SendJob"],"paddle::BaseClient::putData":[53,6,1,"_CPPv2N6paddle10BaseClient7putDataEi12SendDataTypeP8DataType6size_t14DataUpdateMode"],"paddle::BaseClient::putOwnData":[53,6,1,"_CPPv2N6paddle10BaseClient10putOwnDataEi12SendDataTypeP8DataType6size_t"],"paddle::BaseClient::recv":[53,6,1,"_CPPv2N6paddle10BaseClient4recvEi"],"paddle::BaseClient::recvData":[53,6,1,"_CPPv2N6paddle10BaseClient8recvDataEv"],"paddle::BaseClient::recvDataMems_":[53,7,1,"_CPPv2N6paddle10BaseClient13recvDataMems_E"],"paddle::BaseClient::recvJobQueue_":[53,7,1,"_CPPv2N6paddle10BaseClient13recvJobQueue_E"],"paddle::BaseClient::recvSyncBarrier_":[53,7,1,"_CPPv2N6paddle10BaseClient16recvSyncBarrier_E"],"paddle::BaseClient::recvThreads_":[53,7,1,"_CPPv2N6paddle10BaseClient12recvThreads_E"],"paddle::BaseClient::reduce":[53,6,1,"_CPPv2N6paddle10BaseClient6reduceEP8DataTypeP8DataType6size_tii"],"paddle::BaseClient::send":[53,6,1,"_CPPv2N6paddle10BaseClient4sendEi"],"paddle::BaseClient::sendData":[53,6,1,"_CPPv2N6paddle10BaseClient8sendDataEi12SendDataType14DataUpdateModeP8DataType6size_t"],"paddle::BaseClient::sendJobQueue_":[53,7,1,"_CPPv2N6paddle10BaseClient13sendJobQueue_E"],"paddle::BaseClient::sendJob_":[53,7,1,"_CPPv2N6paddle10BaseClient8sendJob_E"],"paddle::BaseClient::sendThreads_":[53,7,1,"_CPPv2N6paddle10BaseClient12sendThreads_E"],"paddle::BaseClient::separateSendAndRecv_":[53,7,1,"_CPPv2N6paddle10BaseClient20separateSendAndRecv_E"],"paddle::BaseClient::serviceNum_":[53,7,1,"_CPPv2N6paddle10BaseClient11serviceNum_E"],"paddle::BaseClient::startThreads":[53,6,1,"_CPPv2N6paddle10BaseClient12startThreadsEv"],"paddle::BaseClient::stopping_":[53,7,1,"_CPPv2N6paddle10BaseClient9stopping_E"],"paddle::BaseClient::synchronize":[53,6,1,"_CPPv2N6paddle10BaseClient11synchronizeE10SyncObject"],"paddle::BaseClient::threadNum_":[53,7,1,"_CPPv2N6paddle10BaseClient10threadNum_E"],"paddle::BaseClient::~BaseClient":[53,6,1,"_CPPv2N6paddle10BaseClientD0Ev"],"paddle::BaseMatrix":[44,9,1,"_CPPv2N6paddle10BaseMatrixE"],"paddle::BaseMatrixT":[44,8,1,"_CPPv2N6paddle11BaseMatrixTE"],"paddle::BaseMatrixT::BaseMatrixT":[44,6,1,"_CPPv2N6paddle11BaseMatrixT11BaseMatrixTE6size_t6size_tP1Tbb"],"paddle::BaseMatrixT::abs":[44,6,1,"_CPPv2N6paddle11BaseMatrixT3absEv"],"paddle::BaseMatrixT::absDerivative":[44,6,1,"_CPPv2N6paddle11BaseMatrixT13absDerivativeER11BaseMatrixT"],"paddle::BaseMatrixT::add":[44,6,1,"_CPPv2N6paddle11BaseMatrixT3addE1T"],"paddle::BaseMatrixT::add2":[44,6,1,"_CPPv2N6paddle11BaseMatrixT4add2ER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::add3":[44,6,1,"_CPPv2N6paddle11BaseMatrixT4add3ER11BaseMatrixTR11BaseMatrixTR11BaseMatrixT1T1T1T"],"paddle::BaseMatrixT::addAtOffset":[44,6,1,"_CPPv2N6paddle11BaseMatrixT11addAtOffsetER11BaseMatrixT7int64_t"],"paddle::BaseMatrixT::addBias":[44,6,1,"_CPPv2N6paddle11BaseMatrixT7addBiasER11BaseMatrixT1T"],"paddle::BaseMatrixT::addColScale":[44,6,1,"_CPPv2N6paddle11BaseMatrixT11addColScaleE6size_tR11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::addColVector":[44,6,1,"_CPPv2N6paddle11BaseMatrixT12addColVectorER11BaseMatrixT"],"paddle::BaseMatrixT::addDotMul":[44,6,1,"_CPPv2N6paddle11BaseMatrixT9addDotMulER11BaseMatrixTR11BaseMatrixT1T1T"],"paddle::BaseMatrixT::addDotMulMMV":[44,6,1,"_CPPv2N6paddle11BaseMatrixT12addDotMulMMVER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::addDotMulMMV2":[44,6,1,"_CPPv2N6paddle11BaseMatrixT13addDotMulMMV2ER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::addDotMulVMM":[44,6,1,"_CPPv2N6paddle11BaseMatrixT12addDotMulVMMER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::addDotMulVMM2":[44,6,1,"_CPPv2N6paddle11BaseMatrixT13addDotMulVMM2ER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::addP2P":[44,6,1,"_CPPv2N6paddle11BaseMatrixT6addP2PER11BaseMatrixT"],"paddle::BaseMatrixT::addRowScale":[44,6,1,"_CPPv2N6paddle11BaseMatrixT11addRowScaleE6size_tR11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::addRowVector":[44,6,1,"_CPPv2N6paddle11BaseMatrixT12addRowVectorER11BaseMatrixT"],"paddle::BaseMatrixT::addScalar":[44,6,1,"_CPPv2N6paddle11BaseMatrixT9addScalarER11BaseMatrixT1T"],"paddle::BaseMatrixT::addSquare":[44,6,1,"_CPPv2N6paddle11BaseMatrixT9addSquareER11BaseMatrixT1T"],"paddle::BaseMatrixT::addSquareSum":[44,6,1,"_CPPv2N6paddle11BaseMatrixT12addSquareSumER11BaseMatrixTR11BaseMatrixT11BaseMatrixT1T1T1T"],"paddle::BaseMatrixT::aggregate":[44,6,1,"_CPPv2N6paddle11BaseMatrixT9aggregateE3Agg2Op5SaverR11BaseMatrixTiiR12MatrixOffset12aAsRowVector12aAsColVector"],"paddle::BaseMatrixT::applyBinary":[44,6,1,"_CPPv2N6paddle11BaseMatrixT11applyBinaryE2OpR11BaseMatrixT"],"paddle::BaseMatrixT::applyCol":[44,6,1,"_CPPv2N6paddle11BaseMatrixT8applyColE3AggR11BaseMatrixT"],"paddle::BaseMatrixT::applyL1":[44,6,1,"_CPPv2N6paddle11BaseMatrixT7applyL1E1T1T"],"paddle::BaseMatrixT::applyL2":[44,6,1,"_CPPv2N6paddle11BaseMatrixT7applyL2E1T1T"],"paddle::BaseMatrixT::applyQuaternary":[44,6,1,"_CPPv2N6paddle11BaseMatrixT15applyQuaternaryE2OpR11BaseMatrixTR11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::applyRow":[44,6,1,"_CPPv2N6paddle11BaseMatrixT8applyRowE3AggR11BaseMatrixT"],"paddle::BaseMatrixT::applyTernary":[44,6,1,"_CPPv2N6paddle11BaseMatrixT12applyTernaryE2OpR11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::applyUnary":[44,6,1,"_CPPv2N6paddle11BaseMatrixT10applyUnaryE2Op"],"paddle::BaseMatrixT::assign":[44,6,1,"_CPPv2N6paddle11BaseMatrixT6assignE1T"],"paddle::BaseMatrixT::assignAtOffset":[44,6,1,"_CPPv2N6paddle11BaseMatrixT14assignAtOffsetER11BaseMatrixT7int64_t"],"paddle::BaseMatrixT::biggerThan":[44,6,1,"_CPPv2N6paddle11BaseMatrixT10biggerThanER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::biggerThanScalar":[44,6,1,"_CPPv2N6paddle11BaseMatrixT16biggerThanScalarE1T"],"paddle::BaseMatrixT::binaryClassificationError":[44,6,1,"_CPPv2N6paddle11BaseMatrixT25binaryClassificationErrorE6size_tR11BaseMatrixTR11BaseMatrixT1T"],"paddle::BaseMatrixT::binaryClassificationError2":[44,6,1,"_CPPv2N6paddle11BaseMatrixT26binaryClassificationError2E6size_tR11BaseMatrixTR11BaseMatrixT1T"],"paddle::BaseMatrixT::binaryLabelCrossEntropy":[44,6,1,"_CPPv2N6paddle11BaseMatrixT23binaryLabelCrossEntropyER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::binaryLabelCrossEntropyBp":[44,6,1,"_CPPv2N6paddle11BaseMatrixT25binaryLabelCrossEntropyBpER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::brelu":[44,6,1,"_CPPv2N6paddle11BaseMatrixT5breluER11BaseMatrixT"],"paddle::BaseMatrixT::breluDerivative":[44,6,1,"_CPPv2N6paddle11BaseMatrixT15breluDerivativeER11BaseMatrixT"],"paddle::BaseMatrixT::clip":[44,6,1,"_CPPv2N6paddle11BaseMatrixT4clipE1T1T"],"paddle::BaseMatrixT::colScale":[44,6,1,"_CPPv2N6paddle11BaseMatrixT8colScaleE6size_tR11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::copyAndClear":[44,6,1,"_CPPv2N6paddle11BaseMatrixT12copyAndClearER11BaseMatrixT"],"paddle::BaseMatrixT::data_":[44,7,1,"_CPPv2N6paddle11BaseMatrixT5data_E"],"paddle::BaseMatrixT::decayAddSquare":[44,6,1,"_CPPv2N6paddle11BaseMatrixT14decayAddSquareER11BaseMatrixT1T1T"],"paddle::BaseMatrixT::decayAddSquareMul":[44,6,1,"_CPPv2N6paddle11BaseMatrixT17decayAddSquareMulER11BaseMatrixTR11BaseMatrixT1T1T"],"paddle::BaseMatrixT::divRowVector":[44,6,1,"_CPPv2N6paddle11BaseMatrixT12divRowVectorER11BaseMatrixT"],"paddle::BaseMatrixT::divScalar":[44,6,1,"_CPPv2N6paddle11BaseMatrixT9divScalarE1T"],"paddle::BaseMatrixT::dotDiv":[44,6,1,"_CPPv2N6paddle11BaseMatrixT6dotDivER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::dotMul":[44,6,1,"_CPPv2N6paddle11BaseMatrixT6dotMulER11BaseMatrixT"],"paddle::BaseMatrixT::dotMulSquare":[44,6,1,"_CPPv2N6paddle11BaseMatrixT12dotMulSquareER11BaseMatrixT"],"paddle::BaseMatrixT::dotMulSquareSum":[44,6,1,"_CPPv2N6paddle11BaseMatrixT15dotMulSquareSumER11BaseMatrixTR11BaseMatrixT1T1T"],"paddle::BaseMatrixT::dotMulSum":[44,6,1,"_CPPv2N6paddle11BaseMatrixT9dotMulSumER11BaseMatrixTR11BaseMatrixT1T1T"],"paddle::BaseMatrixT::dotSquareMul":[44,6,1,"_CPPv2N6paddle11BaseMatrixT12dotSquareMulER11BaseMatrixT"],"paddle::BaseMatrixT::dotSquareSquare":[44,6,1,"_CPPv2N6paddle11BaseMatrixT15dotSquareSquareER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::dotSquareSum":[44,6,1,"_CPPv2N6paddle11BaseMatrixT12dotSquareSumER11BaseMatrixTR11BaseMatrixT1T1T"],"paddle::BaseMatrixT::downClip":[44,6,1,"_CPPv2N6paddle11BaseMatrixT8downClipE1T"],"paddle::BaseMatrixT::exp":[44,6,1,"_CPPv2N6paddle11BaseMatrixT3expEv"],"paddle::BaseMatrixT::expDerivative":[44,6,1,"_CPPv2N6paddle11BaseMatrixT13expDerivativeER11BaseMatrixT"],"paddle::BaseMatrixT::height_":[44,7,1,"_CPPv2N6paddle11BaseMatrixT7height_E"],"paddle::BaseMatrixT::invSqrt":[44,6,1,"_CPPv2N6paddle11BaseMatrixT7invSqrtER11BaseMatrixT"],"paddle::BaseMatrixT::isEqualTo":[44,6,1,"_CPPv2N6paddle11BaseMatrixT9isEqualToER11BaseMatrixT1T"],"paddle::BaseMatrixT::isSparse":[44,6,1,"_CPPv2N6paddle11BaseMatrixT8isSparseEv"],"paddle::BaseMatrixT::log":[44,6,1,"_CPPv2N6paddle11BaseMatrixT3logEv"],"paddle::BaseMatrixT::logisticRegressionLoss":[44,6,1,"_CPPv2N6paddle11BaseMatrixT22logisticRegressionLossER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::logisticRegressionLossBp":[44,6,1,"_CPPv2N6paddle11BaseMatrixT24logisticRegressionLossBpER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::max":[44,6,1,"_CPPv2N6paddle11BaseMatrixT3maxER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::maxCols":[44,6,1,"_CPPv2N6paddle11BaseMatrixT7maxColsER11BaseMatrixT"],"paddle::BaseMatrixT::maxRows":[44,6,1,"_CPPv2N6paddle11BaseMatrixT7maxRowsER11BaseMatrixT"],"paddle::BaseMatrixT::minCols":[44,6,1,"_CPPv2N6paddle11BaseMatrixT7minColsER11BaseMatrixT"],"paddle::BaseMatrixT::minRows":[44,6,1,"_CPPv2N6paddle11BaseMatrixT7minRowsER11BaseMatrixT"],"paddle::BaseMatrixT::mulRowVector":[44,6,1,"_CPPv2N6paddle11BaseMatrixT12mulRowVectorER11BaseMatrixT"],"paddle::BaseMatrixT::mulScalar":[44,6,1,"_CPPv2N6paddle11BaseMatrixT9mulScalarE1T"],"paddle::BaseMatrixT::neg":[44,6,1,"_CPPv2N6paddle11BaseMatrixT3negEv"],"paddle::BaseMatrixT::one":[44,6,1,"_CPPv2N6paddle11BaseMatrixT3oneEv"],"paddle::BaseMatrixT::pow":[44,6,1,"_CPPv2N6paddle11BaseMatrixT3powE1T"],"paddle::BaseMatrixT::rankLoss":[44,6,1,"_CPPv2N6paddle11BaseMatrixT8rankLossER11BaseMatrixTR11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::rankLossBp":[44,6,1,"_CPPv2N6paddle11BaseMatrixT10rankLossBpER11BaseMatrixTR11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::reciprocal":[44,6,1,"_CPPv2N6paddle11BaseMatrixT10reciprocalEv"],"paddle::BaseMatrixT::reciprocalDerivative":[44,6,1,"_CPPv2N6paddle11BaseMatrixT20reciprocalDerivativeER11BaseMatrixT"],"paddle::BaseMatrixT::reciprocalSum":[44,6,1,"_CPPv2N6paddle11BaseMatrixT13reciprocalSumER11BaseMatrixTR11BaseMatrixT1T1T1T"],"paddle::BaseMatrixT::relu":[44,6,1,"_CPPv2N6paddle11BaseMatrixT4reluER11BaseMatrixT"],"paddle::BaseMatrixT::reluDerivative":[44,6,1,"_CPPv2N6paddle11BaseMatrixT14reluDerivativeER11BaseMatrixT"],"paddle::BaseMatrixT::rowAdd":[44,6,1,"_CPPv2N6paddle11BaseMatrixT6rowAddE6size_tR11BaseMatrixTR11BaseMatrixT1T"],"paddle::BaseMatrixT::rowBuf":[44,6,1,"_CPPv2N6paddle11BaseMatrixT6rowBufE6size_t"],"paddle::BaseMatrixT::rowDotMul":[44,6,1,"_CPPv2N6paddle11BaseMatrixT9rowDotMulE6size_tR11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::rowDotMul2":[44,6,1,"_CPPv2N6paddle11BaseMatrixT10rowDotMul2E6size_tR11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::rowPow":[44,6,1,"_CPPv2N6paddle11BaseMatrixT6rowPowE6size_tR11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::rowScale":[44,6,1,"_CPPv2N6paddle11BaseMatrixT8rowScaleE6size_tR11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::rowScale2":[44,6,1,"_CPPv2N6paddle11BaseMatrixT9rowScale2E6size_tR11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::scalarDiv":[44,6,1,"_CPPv2N6paddle11BaseMatrixT9scalarDivER11BaseMatrixT1T"],"paddle::BaseMatrixT::scaledTanh":[44,6,1,"_CPPv2N6paddle11BaseMatrixT10scaledTanhER11BaseMatrixT1T1T"],"paddle::BaseMatrixT::scaledTanhDerivative":[44,6,1,"_CPPv2N6paddle11BaseMatrixT20scaledTanhDerivativeER11BaseMatrixT1T1T"],"paddle::BaseMatrixT::setData":[44,6,1,"_CPPv2N6paddle11BaseMatrixT7setDataEP1T"],"paddle::BaseMatrixT::sgdUpdate":[44,6,1,"_CPPv2N6paddle11BaseMatrixT9sgdUpdateER11BaseMatrixTR11BaseMatrixT1T1T1T"],"paddle::BaseMatrixT::sigmoid":[44,6,1,"_CPPv2N6paddle11BaseMatrixT7sigmoidER11BaseMatrixT"],"paddle::BaseMatrixT::sigmoidDerivative":[44,6,1,"_CPPv2N6paddle11BaseMatrixT17sigmoidDerivativeER11BaseMatrixT"],"paddle::BaseMatrixT::sign":[44,6,1,"_CPPv2N6paddle11BaseMatrixT4signEv"],"paddle::BaseMatrixT::softCrossEntropy":[44,6,1,"_CPPv2N6paddle11BaseMatrixT16softCrossEntropyER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::softCrossEntropyBp":[44,6,1,"_CPPv2N6paddle11BaseMatrixT18softCrossEntropyBpER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::softrelu":[44,6,1,"_CPPv2N6paddle11BaseMatrixT8softreluER11BaseMatrixT"],"paddle::BaseMatrixT::softreluDerivative":[44,6,1,"_CPPv2N6paddle11BaseMatrixT18softreluDerivativeER11BaseMatrixT"],"paddle::BaseMatrixT::sqrt":[44,6,1,"_CPPv2N6paddle11BaseMatrixT4sqrtEv"],"paddle::BaseMatrixT::square":[44,6,1,"_CPPv2N6paddle11BaseMatrixT6squareEv"],"paddle::BaseMatrixT::squareDerivative":[44,6,1,"_CPPv2N6paddle11BaseMatrixT16squareDerivativeER11BaseMatrixT"],"paddle::BaseMatrixT::stride_":[44,7,1,"_CPPv2N6paddle11BaseMatrixT7stride_E"],"paddle::BaseMatrixT::sub":[44,6,1,"_CPPv2N6paddle11BaseMatrixT3subER11BaseMatrixT"],"paddle::BaseMatrixT::subScalar":[44,6,1,"_CPPv2N6paddle11BaseMatrixT9subScalarE1T"],"paddle::BaseMatrixT::sumCols":[44,6,1,"_CPPv2N6paddle11BaseMatrixT7sumColsER11BaseMatrixT"],"paddle::BaseMatrixT::sumOfSquares":[44,6,1,"_CPPv2N6paddle11BaseMatrixT12sumOfSquaresER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::sumRows":[44,6,1,"_CPPv2N6paddle11BaseMatrixT7sumRowsER11BaseMatrixT"],"paddle::BaseMatrixT::tanh":[44,6,1,"_CPPv2N6paddle11BaseMatrixT4tanhER11BaseMatrixT"],"paddle::BaseMatrixT::tanhDerivative":[44,6,1,"_CPPv2N6paddle11BaseMatrixT14tanhDerivativeER11BaseMatrixT"],"paddle::BaseMatrixT::trans_":[44,7,1,"_CPPv2N6paddle11BaseMatrixT6trans_E"],"paddle::BaseMatrixT::useGpu":[44,6,1,"_CPPv2N6paddle11BaseMatrixT6useGpuEv"],"paddle::BaseMatrixT::useGpu_":[44,7,1,"_CPPv2N6paddle11BaseMatrixT7useGpu_E"],"paddle::BaseMatrixT::width_":[44,7,1,"_CPPv2N6paddle11BaseMatrixT6width_E"],"paddle::BaseMatrixT::zero":[44,6,1,"_CPPv2N6paddle11BaseMatrixT4zeroEv"],"paddle::BaseMatrixT::zeroAtOffset":[44,6,1,"_CPPv2N6paddle11BaseMatrixT12zeroAtOffsetE7int64_t7int64_t"],"paddle::BaseMatrixT::~BaseMatrixT":[44,6,1,"_CPPv2N6paddle11BaseMatrixTD0Ev"],"paddle::BaseVector":[44,8,1,"_CPPv2N6paddle10BaseVectorE"],"paddle::BaseVector::BaseVector":[44,6,1,"_CPPv2N6paddle10BaseVector10BaseVectorE6size_tP1Tb"],"paddle::BaseVector::size_":[44,7,1,"_CPPv2N6paddle10BaseVector5size_E"],"paddle::BaseVector::~BaseVector":[44,6,1,"_CPPv2N6paddle10BaseVectorD0Ev"],"paddle::BatchNormBaseLayer":[41,8,1,"_CPPv2N6paddle18BatchNormBaseLayerE"],"paddle::BatchNormBaseLayer::BatchNormBaseLayer":[41,6,1,"_CPPv2N6paddle18BatchNormBaseLayer18BatchNormBaseLayerERK11LayerConfig"],"paddle::BatchNormBaseLayer::biases_":[41,7,1,"_CPPv2N6paddle18BatchNormBaseLayer7biases_E"],"paddle::BatchNormBaseLayer::calFeatureMapSize":[41,6,1,"_CPPv2N6paddle18BatchNormBaseLayer17calFeatureMapSizeEv"],"paddle::BatchNormBaseLayer::channels_":[41,7,1,"_CPPv2N6paddle18BatchNormBaseLayer9channels_E"],"paddle::BatchNormBaseLayer::create":[41,6,1,"_CPPv2N6paddle18BatchNormBaseLayer6createERK11LayerConfig"],"paddle::BatchNormBaseLayer::imageH_":[41,7,1,"_CPPv2N6paddle18BatchNormBaseLayer7imageH_E"],"paddle::BatchNormBaseLayer::imageW_":[41,7,1,"_CPPv2N6paddle18BatchNormBaseLayer7imageW_E"],"paddle::BatchNormBaseLayer::imgPixels_":[41,7,1,"_CPPv2N6paddle18BatchNormBaseLayer10imgPixels_E"],"paddle::BatchNormBaseLayer::imgSize_":[41,7,1,"_CPPv2N6paddle18BatchNormBaseLayer8imgSize_E"],"paddle::BatchNormBaseLayer::init":[41,6,1,"_CPPv2N6paddle18BatchNormBaseLayer4initERK8LayerMapRK12ParameterMap"],"paddle::BatchNormBaseLayer::movingAvgFraction_":[41,7,1,"_CPPv2N6paddle18BatchNormBaseLayer18movingAvgFraction_E"],"paddle::BatchNormBaseLayer::movingMean_":[41,7,1,"_CPPv2N6paddle18BatchNormBaseLayer11movingMean_E"],"paddle::BatchNormBaseLayer::movingVar_":[41,7,1,"_CPPv2N6paddle18BatchNormBaseLayer10movingVar_E"],"paddle::BatchNormBaseLayer::savedInvVar_":[41,7,1,"_CPPv2N6paddle18BatchNormBaseLayer12savedInvVar_E"],"paddle::BatchNormBaseLayer::savedMean_":[41,7,1,"_CPPv2N6paddle18BatchNormBaseLayer10savedMean_E"],"paddle::BatchNormBaseLayer::useGlobalStats_":[41,7,1,"_CPPv2N6paddle18BatchNormBaseLayer15useGlobalStats_E"],"paddle::BatchNormBaseLayer::weight_":[41,7,1,"_CPPv2N6paddle18BatchNormBaseLayer7weight_E"],"paddle::BatchNormBaseLayer::~BatchNormBaseLayer":[41,6,1,"_CPPv2N6paddle18BatchNormBaseLayerD0Ev"],"paddle::BatchNormalizationLayer":[41,8,1,"_CPPv2N6paddle23BatchNormalizationLayerE"],"paddle::BatchNormalizationLayer::BatchNormalizationLayer":[41,6,1,"_CPPv2N6paddle23BatchNormalizationLayer23BatchNormalizationLayerERK11LayerConfig"],"paddle::BatchNormalizationLayer::EPS":[41,7,1,"_CPPv2N6paddle23BatchNormalizationLayer3EPSE"],"paddle::BatchNormalizationLayer::backward":[41,6,1,"_CPPv2N6paddle23BatchNormalizationLayer8backwardERK14UpdateCallback"],"paddle::BatchNormalizationLayer::calMeanAndStd":[41,6,1,"_CPPv2N6paddle23BatchNormalizationLayer13calMeanAndStdERK9MatrixPtr"],"paddle::BatchNormalizationLayer::calMovingMeanAndVar":[41,6,1,"_CPPv2N6paddle23BatchNormalizationLayer19calMovingMeanAndVarEv"],"paddle::BatchNormalizationLayer::expandMat":[41,6,1,"_CPPv2N6paddle23BatchNormalizationLayer9expandMatERK9MatrixPtrR9MatrixPtr"],"paddle::BatchNormalizationLayer::expandedInGrad_":[41,7,1,"_CPPv2N6paddle23BatchNormalizationLayer15expandedInGrad_E"],"paddle::BatchNormalizationLayer::expandedIn_":[41,7,1,"_CPPv2N6paddle23BatchNormalizationLayer11expandedIn_E"],"paddle::BatchNormalizationLayer::expandedOutGrad_":[41,7,1,"_CPPv2N6paddle23BatchNormalizationLayer16expandedOutGrad_E"],"paddle::BatchNormalizationLayer::expandedOut_":[41,7,1,"_CPPv2N6paddle23BatchNormalizationLayer12expandedOut_E"],"paddle::BatchNormalizationLayer::firstTest_":[41,7,1,"_CPPv2N6paddle23BatchNormalizationLayer10firstTest_E"],"paddle::BatchNormalizationLayer::forward":[41,6,1,"_CPPv2N6paddle23BatchNormalizationLayer7forwardE8PassType"],"paddle::BatchNormalizationLayer::inGrad_":[41,7,1,"_CPPv2N6paddle23BatchNormalizationLayer7inGrad_E"],"paddle::BatchNormalizationLayer::init":[41,6,1,"_CPPv2N6paddle23BatchNormalizationLayer4initERK8LayerMapRK12ParameterMap"],"paddle::BatchNormalizationLayer::meanGrad_":[41,7,1,"_CPPv2N6paddle23BatchNormalizationLayer9meanGrad_E"],"paddle::BatchNormalizationLayer::normInGrad_":[41,7,1,"_CPPv2N6paddle23BatchNormalizationLayer11normInGrad_E"],"paddle::BatchNormalizationLayer::normIn_":[41,7,1,"_CPPv2N6paddle23BatchNormalizationLayer7normIn_E"],"paddle::BatchNormalizationLayer::setMeanAndStd":[41,6,1,"_CPPv2N6paddle23BatchNormalizationLayer13setMeanAndStdEv"],"paddle::BatchNormalizationLayer::shrinkMat":[41,6,1,"_CPPv2N6paddle23BatchNormalizationLayer9shrinkMatERK9MatrixPtrR9MatrixPtr"],"paddle::BatchNormalizationLayer::stdGrad_":[41,7,1,"_CPPv2N6paddle23BatchNormalizationLayer8stdGrad_E"],"paddle::BatchNormalizationLayer::tmpGrad_":[41,7,1,"_CPPv2N6paddle23BatchNormalizationLayer8tmpGrad_E"],"paddle::BatchNormalizationLayer::tmpMat_":[41,7,1,"_CPPv2N6paddle23BatchNormalizationLayer7tmpMat_E"],"paddle::BatchNormalizationLayer::~BatchNormalizationLayer":[41,6,1,"_CPPv2N6paddle23BatchNormalizationLayerD0Ev"],"paddle::BlockExpandLayer":[41,8,1,"_CPPv2N6paddle16BlockExpandLayerE"],"paddle::BlockExpandLayer::BlockExpandLayer":[41,6,1,"_CPPv2N6paddle16BlockExpandLayer16BlockExpandLayerERK11LayerConfig"],"paddle::BlockExpandLayer::backward":[41,6,1,"_CPPv2N6paddle16BlockExpandLayer8backwardERK14UpdateCallback"],"paddle::BlockExpandLayer::blockH_":[41,7,1,"_CPPv2N6paddle16BlockExpandLayer7blockH_E"],"paddle::BlockExpandLayer::blockW_":[41,7,1,"_CPPv2N6paddle16BlockExpandLayer7blockW_E"],"paddle::BlockExpandLayer::channels_":[41,7,1,"_CPPv2N6paddle16BlockExpandLayer9channels_E"],"paddle::BlockExpandLayer::forward":[41,6,1,"_CPPv2N6paddle16BlockExpandLayer7forwardE8PassType"],"paddle::BlockExpandLayer::getBlockNum":[41,6,1,"_CPPv2N6paddle16BlockExpandLayer11getBlockNumEv"],"paddle::BlockExpandLayer::imgSizeH_":[41,7,1,"_CPPv2N6paddle16BlockExpandLayer9imgSizeH_E"],"paddle::BlockExpandLayer::imgSizeW_":[41,7,1,"_CPPv2N6paddle16BlockExpandLayer9imgSizeW_E"],"paddle::BlockExpandLayer::init":[41,6,1,"_CPPv2N6paddle16BlockExpandLayer4initERK8LayerMapRK12ParameterMap"],"paddle::BlockExpandLayer::outVTrans_":[41,7,1,"_CPPv2N6paddle16BlockExpandLayer10outVTrans_E"],"paddle::BlockExpandLayer::outputH_":[41,7,1,"_CPPv2N6paddle16BlockExpandLayer8outputH_E"],"paddle::BlockExpandLayer::outputW_":[41,7,1,"_CPPv2N6paddle16BlockExpandLayer8outputW_E"],"paddle::BlockExpandLayer::paddingH_":[41,7,1,"_CPPv2N6paddle16BlockExpandLayer9paddingH_E"],"paddle::BlockExpandLayer::paddingW_":[41,7,1,"_CPPv2N6paddle16BlockExpandLayer9paddingW_E"],"paddle::BlockExpandLayer::strideH_":[41,7,1,"_CPPv2N6paddle16BlockExpandLayer8strideH_E"],"paddle::BlockExpandLayer::strideW_":[41,7,1,"_CPPv2N6paddle16BlockExpandLayer8strideW_E"],"paddle::BlockExpandLayer::~BlockExpandLayer":[41,6,1,"_CPPv2N6paddle16BlockExpandLayerD0Ev"],"paddle::BlockingQueue":[63,8,1,"_CPPv2N6paddle13BlockingQueueE"],"paddle::BlockingQueue::BlockingQueue":[63,6,1,"_CPPv2N6paddle13BlockingQueue13BlockingQueueE6size_t"],"paddle::BlockingQueue::dequeue":[63,6,1,"_CPPv2N6paddle13BlockingQueue7dequeueEv"],"paddle::BlockingQueue::empty":[63,6,1,"_CPPv2N6paddle13BlockingQueue5emptyEv"],"paddle::BlockingQueue::enqueue":[63,6,1,"_CPPv2N6paddle13BlockingQueue7enqueueERK1T"],"paddle::BlockingQueue::size":[63,6,1,"_CPPv2N6paddle13BlockingQueue4sizeEv"],"paddle::CMRProjectionNormLayer":[41,8,1,"_CPPv2N6paddle22CMRProjectionNormLayerE"],"paddle::CMRProjectionNormLayer::CMRProjectionNormLayer":[41,6,1,"_CPPv2N6paddle22CMRProjectionNormLayer22CMRProjectionNormLayerERK11LayerConfig"],"paddle::CMRProjectionNormLayer::backward":[41,6,1,"_CPPv2N6paddle22CMRProjectionNormLayer8backwardERK14UpdateCallback"],"paddle::CMRProjectionNormLayer::blocked_":[41,7,1,"_CPPv2N6paddle22CMRProjectionNormLayer8blocked_E"],"paddle::CMRProjectionNormLayer::forward":[41,6,1,"_CPPv2N6paddle22CMRProjectionNormLayer7forwardE8PassType"],"paddle::CMRProjectionNormLayer::getSize":[41,6,1,"_CPPv2N6paddle22CMRProjectionNormLayer7getSizeEv"],"paddle::CMRProjectionNormLayer::init":[41,6,1,"_CPPv2N6paddle22CMRProjectionNormLayer4initERK8LayerMapRK12ParameterMap"],"paddle::CMRProjectionNormLayer::~CMRProjectionNormLayer":[41,6,1,"_CPPv2N6paddle22CMRProjectionNormLayerD0Ev"],"paddle::CRFDecodingLayer":[41,8,1,"_CPPv2N6paddle16CRFDecodingLayerE"],"paddle::CRFDecodingLayer::CRFDecodingLayer":[41,6,1,"_CPPv2N6paddle16CRFDecodingLayer16CRFDecodingLayerERK11LayerConfig"],"paddle::CRFDecodingLayer::backward":[41,6,1,"_CPPv2N6paddle16CRFDecodingLayer8backwardERK14UpdateCallback"],"paddle::CRFDecodingLayer::crf_":[41,7,1,"_CPPv2N6paddle16CRFDecodingLayer4crf_E"],"paddle::CRFDecodingLayer::forward":[41,6,1,"_CPPv2N6paddle16CRFDecodingLayer7forwardE8PassType"],"paddle::CRFDecodingLayer::init":[41,6,1,"_CPPv2N6paddle16CRFDecodingLayer4initERK8LayerMapRK12ParameterMap"],"paddle::CRFLayer":[41,8,1,"_CPPv2N6paddle8CRFLayerE"],"paddle::CRFLayer::CRFLayer":[41,6,1,"_CPPv2N6paddle8CRFLayer8CRFLayerERK11LayerConfig"],"paddle::CRFLayer::backward":[41,6,1,"_CPPv2N6paddle8CRFLayer8backwardERK14UpdateCallback"],"paddle::CRFLayer::backwardImp":[41,6,1,"_CPPv2N6paddle8CRFLayer11backwardImpERK14UpdateCallbackRK8ArgumentRK8Argument"],"paddle::CRFLayer::coeff_":[41,7,1,"_CPPv2N6paddle8CRFLayer6coeff_E"],"paddle::CRFLayer::crfs_":[41,7,1,"_CPPv2N6paddle8CRFLayer5crfs_E"],"paddle::CRFLayer::forward":[41,6,1,"_CPPv2N6paddle8CRFLayer7forwardE8PassType"],"paddle::CRFLayer::forwardImp":[41,6,1,"_CPPv2N6paddle8CRFLayer10forwardImpERK8ArgumentRK8Argument9VectorPtr9VectorPtr"],"paddle::CRFLayer::init":[41,6,1,"_CPPv2N6paddle8CRFLayer4initERK8LayerMapRK12ParameterMap"],"paddle::CRFLayer::numClasses_":[41,7,1,"_CPPv2N6paddle8CRFLayer11numClasses_E"],"paddle::CRFLayer::parameter_":[41,7,1,"_CPPv2N6paddle8CRFLayer10parameter_E"],"paddle::CRFLayer::tmpCpuInput_":[41,7,1,"_CPPv2N6paddle8CRFLayer12tmpCpuInput_E"],"paddle::CRFLayer::weightLayer_":[41,7,1,"_CPPv2N6paddle8CRFLayer12weightLayer_E"],"paddle::CTCErrorEvaluator":[36,8,1,"_CPPv2N6paddle17CTCErrorEvaluatorE"],"paddle::CTCErrorEvaluator::CTCErrorEvaluator":[36,6,1,"_CPPv2N6paddle17CTCErrorEvaluator17CTCErrorEvaluatorEv"],"paddle::CTCErrorEvaluator::distributeEval":[36,6,1,"_CPPv2N6paddle17CTCErrorEvaluator14distributeEvalEP16ParameterClient2"],"paddle::CTCErrorEvaluator::eval":[36,6,1,"_CPPv2N6paddle17CTCErrorEvaluator4evalERK13NeuralNetwork"],"paddle::CTCErrorEvaluator::evalImp":[36,6,1,"_CPPv2N6paddle17CTCErrorEvaluator7evalImpERNSt6vectorI8ArgumentEE"],"paddle::CTCErrorEvaluator::printStats":[36,6,1,"_CPPv2N6paddle17CTCErrorEvaluator10printStatsERNSt7ostreamE"],"paddle::CTCErrorEvaluator::start":[36,6,1,"_CPPv2N6paddle17CTCErrorEvaluator5startEv"],"paddle::CTCLayer":[41,8,1,"_CPPv2N6paddle8CTCLayerE"],"paddle::CTCLayer::CTCLayer":[41,6,1,"_CPPv2N6paddle8CTCLayer8CTCLayerERK11LayerConfig"],"paddle::CTCLayer::backward":[41,6,1,"_CPPv2N6paddle8CTCLayer8backwardERK14UpdateCallback"],"paddle::CTCLayer::backwardImp":[41,6,1,"_CPPv2N6paddle8CTCLayer11backwardImpERK14UpdateCallbackRK8ArgumentRK8Argument"],"paddle::CTCLayer::ctcs_":[41,7,1,"_CPPv2N6paddle8CTCLayer5ctcs_E"],"paddle::CTCLayer::forward":[41,6,1,"_CPPv2N6paddle8CTCLayer7forwardE8PassType"],"paddle::CTCLayer::forwardImp":[41,6,1,"_CPPv2N6paddle8CTCLayer10forwardImpERK8ArgumentRK8Argument"],"paddle::CTCLayer::init":[41,6,1,"_CPPv2N6paddle8CTCLayer4initERK8LayerMapRK12ParameterMap"],"paddle::CTCLayer::normByTimes_":[41,7,1,"_CPPv2N6paddle8CTCLayer12normByTimes_E"],"paddle::CTCLayer::numClasses_":[41,7,1,"_CPPv2N6paddle8CTCLayer11numClasses_E"],"paddle::CTCLayer::tmpCpuInput_":[41,7,1,"_CPPv2N6paddle8CTCLayer12tmpCpuInput_E"],"paddle::CacheOnePassInMemory":[34,8,1,"_CPPv2N6paddle20CacheOnePassInMemoryE"],"paddle::CacheOnePassInMemory::CacheOnePassInMemory":[34,6,1,"_CPPv2N6paddle20CacheOnePassInMemory20CacheOnePassInMemoryEv"],"paddle::CacheOnePassInMemory::drop":[34,6,1,"_CPPv2N6paddle20CacheOnePassInMemory4dropEPNSt5dequeI11PyObjectPtrEE"],"paddle::CacheOnePassInMemory::load":[34,6,1,"_CPPv2N6paddle20CacheOnePassInMemory4loadEv"],"paddle::CacheOnePassInMemory::reset":[34,6,1,"_CPPv2N6paddle20CacheOnePassInMemory5resetEv"],"paddle::CacheRowCpuMatrix":[44,8,1,"_CPPv2N6paddle17CacheRowCpuMatrixE"],"paddle::CacheRowCpuMatrix::CacheRowCpuMatrix":[44,6,1,"_CPPv2N6paddle17CacheRowCpuMatrix17CacheRowCpuMatrixE6size_t6size_t12IndexDictPtrb"],"paddle::CacheRowCpuMatrix::getRow":[44,6,1,"_CPPv2N6paddle17CacheRowCpuMatrix6getRowE6size_t"],"paddle::CacheRowCpuMatrix::getRowBuf":[44,6,1,"_CPPv2N6paddle17CacheRowCpuMatrix9getRowBufE6size_t"],"paddle::CacheRowCpuMatrix::mul":[44,6,1,"_CPPv2N6paddle17CacheRowCpuMatrix3mulEP15CpuSparseMatrixP9CpuMatrix4real4real"],"paddle::CacheRowCpuMatrix::setSourceData":[44,6,1,"_CPPv2N6paddle17CacheRowCpuMatrix13setSourceDataE12CpuVectorPtr"],"paddle::CacheRowCpuMatrix::sourceDataVec_":[44,7,1,"_CPPv2N6paddle17CacheRowCpuMatrix14sourceDataVec_E"],"paddle::CacheRowCpuMatrix::sourceData_":[44,7,1,"_CPPv2N6paddle17CacheRowCpuMatrix11sourceData_E"],"paddle::ChunkEvaluator":[36,8,1,"_CPPv2N6paddle14ChunkEvaluatorE"],"paddle::ChunkEvaluator::distributeEval":[36,6,1,"_CPPv2N6paddle14ChunkEvaluator14distributeEvalEP16ParameterClient2"],"paddle::ChunkEvaluator::eval1":[36,6,1,"_CPPv2N6paddle14ChunkEvaluator5eval1EPiPii"],"paddle::ChunkEvaluator::evalImp":[36,6,1,"_CPPv2N6paddle14ChunkEvaluator7evalImpERNSt6vectorI8ArgumentEE"],"paddle::ChunkEvaluator::getSegments":[36,6,1,"_CPPv2N6paddle14ChunkEvaluator11getSegmentsEPiiRNSt6vectorI7SegmentEE"],"paddle::ChunkEvaluator::init":[36,6,1,"_CPPv2N6paddle14ChunkEvaluator4initERK15EvaluatorConfig"],"paddle::ChunkEvaluator::isChunkBegin":[36,6,1,"_CPPv2N6paddle14ChunkEvaluator12isChunkBeginEiiii"],"paddle::ChunkEvaluator::isChunkEnd":[36,6,1,"_CPPv2N6paddle14ChunkEvaluator10isChunkEndEiiii"],"paddle::ChunkEvaluator::printStats":[36,6,1,"_CPPv2N6paddle14ChunkEvaluator10printStatsERNSt7ostreamE"],"paddle::ChunkEvaluator::start":[36,6,1,"_CPPv2N6paddle14ChunkEvaluator5startEv"],"paddle::ClassificationErrorEvaluator":[36,8,1,"_CPPv2N6paddle28ClassificationErrorEvaluatorE"],"paddle::ClassificationErrorEvaluator::calcError":[36,6,1,"_CPPv2N6paddle28ClassificationErrorEvaluator9calcErrorERNSt6vectorI8ArgumentEE"],"paddle::ClassificationErrorEvaluator::distributeEval":[36,6,1,"_CPPv2N6paddle28ClassificationErrorEvaluator14distributeEvalEP16ParameterClient2"],"paddle::ClassificationErrorEvaluator::evalImp":[36,6,1,"_CPPv2N6paddle28ClassificationErrorEvaluator7evalImpERNSt6vectorI8ArgumentEE"],"paddle::ClassificationErrorEvaluator::updateSamplesNum":[36,6,1,"_CPPv2N6paddle28ClassificationErrorEvaluator16updateSamplesNumERKNSt6vectorI8ArgumentEE"],"paddle::ClassificationErrorPrinter":[36,8,1,"_CPPv2N6paddle26ClassificationErrorPrinterE"],"paddle::ClassificationErrorPrinter::evalImp":[36,6,1,"_CPPv2N6paddle26ClassificationErrorPrinter7evalImpERNSt6vectorI8ArgumentEE"],"paddle::ClassificationErrorPrinter::updateSamplesNum":[36,6,1,"_CPPv2N6paddle26ClassificationErrorPrinter16updateSamplesNumERKNSt6vectorI8ArgumentEE"],"paddle::ColumnSumEvaluator":[36,8,1,"_CPPv2N6paddle18ColumnSumEvaluatorE"],"paddle::ColumnSumEvaluator::ColumnSumEvaluator":[36,6,1,"_CPPv2N6paddle18ColumnSumEvaluator18ColumnSumEvaluatorE7int32_t"],"paddle::ColumnSumEvaluator::distributeEval":[36,6,1,"_CPPv2N6paddle18ColumnSumEvaluator14distributeEvalEP16ParameterClient2"],"paddle::ColumnSumEvaluator::evalImp":[36,6,1,"_CPPv2N6paddle18ColumnSumEvaluator7evalImpERNSt6vectorI8ArgumentEE"],"paddle::ColumnSumEvaluator::printStats":[36,6,1,"_CPPv2N6paddle18ColumnSumEvaluator10printStatsERNSt7ostreamE"],"paddle::ColumnSumEvaluator::start":[36,6,1,"_CPPv2N6paddle18ColumnSumEvaluator5startEv"],"paddle::ColumnSumEvaluator::updateSamplesNum":[36,6,1,"_CPPv2N6paddle18ColumnSumEvaluator16updateSamplesNumERKNSt6vectorI8ArgumentEE"],"paddle::ConcatenateLayer":[41,8,1,"_CPPv2N6paddle16ConcatenateLayerE"],"paddle::ConcatenateLayer2":[41,8,1,"_CPPv2N6paddle17ConcatenateLayer2E"],"paddle::ConcatenateLayer2::ConcatenateLayer2":[41,6,1,"_CPPv2N6paddle17ConcatenateLayer217ConcatenateLayer2ERK11LayerConfig"],"paddle::ConcatenateLayer2::backward":[41,6,1,"_CPPv2N6paddle17ConcatenateLayer28backwardERK14UpdateCallback"],"paddle::ConcatenateLayer2::forward":[41,6,1,"_CPPv2N6paddle17ConcatenateLayer27forwardE8PassType"],"paddle::ConcatenateLayer2::init":[41,6,1,"_CPPv2N6paddle17ConcatenateLayer24initERK8LayerMapRK12ParameterMap"],"paddle::ConcatenateLayer2::projCol_":[41,7,1,"_CPPv2N6paddle17ConcatenateLayer28projCol_E"],"paddle::ConcatenateLayer2::projOutput_":[41,7,1,"_CPPv2N6paddle17ConcatenateLayer211projOutput_E"],"paddle::ConcatenateLayer2::projections_":[41,7,1,"_CPPv2N6paddle17ConcatenateLayer212projections_E"],"paddle::ConcatenateLayer2::~ConcatenateLayer2":[41,6,1,"_CPPv2N6paddle17ConcatenateLayer2D0Ev"],"paddle::ConcatenateLayer::ConcatenateLayer":[41,6,1,"_CPPv2N6paddle16ConcatenateLayer16ConcatenateLayerERK11LayerConfig"],"paddle::ConcatenateLayer::backward":[41,6,1,"_CPPv2N6paddle16ConcatenateLayer8backwardERK14UpdateCallback"],"paddle::ConcatenateLayer::forward":[41,6,1,"_CPPv2N6paddle16ConcatenateLayer7forwardE8PassType"],"paddle::ConcatenateLayer::init":[41,6,1,"_CPPv2N6paddle16ConcatenateLayer4initERK8LayerMapRK12ParameterMap"],"paddle::ConcatenateLayer::~ConcatenateLayer":[41,6,1,"_CPPv2N6paddle16ConcatenateLayerD0Ev"],"paddle::ConcurrentRemoteParameterUpdater":[59,8,1,"_CPPv2N6paddle32ConcurrentRemoteParameterUpdaterE"],"paddle::ConcurrentRemoteParameterUpdater::ConcurrentRemoteParameterUpdater":[59,6,1,"_CPPv2N6paddle32ConcurrentRemoteParameterUpdater32ConcurrentRemoteParameterUpdaterE18OptimizationConfigiRRNSt10unique_ptrI16ParameterUpdaterEE"],"paddle::ConcurrentRemoteParameterUpdater::copySingleParaFromDevice":[59,6,1,"_CPPv2N6paddle32ConcurrentRemoteParameterUpdater24copySingleParaFromDeviceEP9Parameter13ParameterType"],"paddle::ConcurrentRemoteParameterUpdater::copySingleParaToDevice":[59,6,1,"_CPPv2N6paddle32ConcurrentRemoteParameterUpdater22copySingleParaToDeviceEP9Parameter13ParameterType"],"paddle::ConcurrentRemoteParameterUpdater::finishBatch":[59,6,1,"_CPPv2N6paddle32ConcurrentRemoteParameterUpdater11finishBatchE4real"],"paddle::ConcurrentRemoteParameterUpdater::needToUpdateRemotely":[59,6,1,"_CPPv2N6paddle32ConcurrentRemoteParameterUpdater20needToUpdateRemotelyEv"],"paddle::ConcurrentRemoteParameterUpdater::recv":[59,6,1,"_CPPv2N6paddle32ConcurrentRemoteParameterUpdater4recvEP9Parameter"],"paddle::ConcurrentRemoteParameterUpdater::send":[59,6,1,"_CPPv2N6paddle32ConcurrentRemoteParameterUpdater4sendEP9Parameter"],"paddle::ConcurrentRemoteParameterUpdater::updateImpl":[59,6,1,"_CPPv2N6paddle32ConcurrentRemoteParameterUpdater10updateImplEP9Parameter"],"paddle::ConcurrentRemoteParameterUpdater::~ConcurrentRemoteParameterUpdater":[59,6,1,"_CPPv2N6paddle32ConcurrentRemoteParameterUpdaterD0Ev"],"paddle::ContextProjection":[41,8,1,"_CPPv2N6paddle17ContextProjectionE"],"paddle::ContextProjection::ContextProjection":[41,6,1,"_CPPv2N6paddle17ContextProjection17ContextProjectionERK16ProjectionConfig12ParameterPtrb"],"paddle::ContextProjection::backward":[41,6,1,"_CPPv2N6paddle17ContextProjection8backwardERK14UpdateCallback"],"paddle::ContextProjection::beginPad_":[41,7,1,"_CPPv2N6paddle17ContextProjection9beginPad_E"],"paddle::ContextProjection::endPad_":[41,7,1,"_CPPv2N6paddle17ContextProjection7endPad_E"],"paddle::ContextProjection::forward":[41,6,1,"_CPPv2N6paddle17ContextProjection7forwardEv"],"paddle::ContextProjection::getState":[41,6,1,"_CPPv2N6paddle17ContextProjection8getStateEv"],"paddle::ContextProjection::resetState":[41,6,1,"_CPPv2N6paddle17ContextProjection10resetStateEv"],"paddle::ContextProjection::setState":[41,6,1,"_CPPv2N6paddle17ContextProjection8setStateE13LayerStatePtr"],"paddle::ContextProjection::state2_":[41,7,1,"_CPPv2N6paddle17ContextProjection7state2_E"],"paddle::ContextProjection::state_":[41,7,1,"_CPPv2N6paddle17ContextProjection6state_E"],"paddle::ContextProjection::weight_":[41,7,1,"_CPPv2N6paddle17ContextProjection7weight_E"],"paddle::ConvBaseLayer":[41,8,1,"_CPPv2N6paddle13ConvBaseLayerE"],"paddle::ConvBaseLayer::ConvBaseLayer":[41,6,1,"_CPPv2N6paddle13ConvBaseLayer13ConvBaseLayerERK11LayerConfig"],"paddle::ConvBaseLayer::IntV":[41,9,1,"_CPPv2N6paddle13ConvBaseLayer4IntVE"],"paddle::ConvBaseLayer::biases_":[41,7,1,"_CPPv2N6paddle13ConvBaseLayer7biases_E"],"paddle::ConvBaseLayer::caffeMode_":[41,7,1,"_CPPv2N6paddle13ConvBaseLayer10caffeMode_E"],"paddle::ConvBaseLayer::channels_":[41,7,1,"_CPPv2N6paddle13ConvBaseLayer9channels_E"],"paddle::ConvBaseLayer::filterChannels_":[41,7,1,"_CPPv2N6paddle13ConvBaseLayer15filterChannels_E"],"paddle::ConvBaseLayer::filterPixels_":[41,7,1,"_CPPv2N6paddle13ConvBaseLayer13filterPixels_E"],"paddle::ConvBaseLayer::filterSizeY_":[41,7,1,"_CPPv2N6paddle13ConvBaseLayer12filterSizeY_E"],"paddle::ConvBaseLayer::filterSize_":[41,7,1,"_CPPv2N6paddle13ConvBaseLayer11filterSize_E"],"paddle::ConvBaseLayer::getWeight":[41,6,1,"_CPPv2N6paddle13ConvBaseLayer9getWeightEi"],"paddle::ConvBaseLayer::groups_":[41,7,1,"_CPPv2N6paddle13ConvBaseLayer7groups_E"],"paddle::ConvBaseLayer::imgPixels_":[41,7,1,"_CPPv2N6paddle13ConvBaseLayer10imgPixels_E"],"paddle::ConvBaseLayer::imgSize_":[41,7,1,"_CPPv2N6paddle13ConvBaseLayer8imgSize_E"],"paddle::ConvBaseLayer::init":[41,6,1,"_CPPv2N6paddle13ConvBaseLayer4initERK8LayerMapRK12ParameterMap"],"paddle::ConvBaseLayer::numFilters_":[41,7,1,"_CPPv2N6paddle13ConvBaseLayer11numFilters_E"],"paddle::ConvBaseLayer::outputSize":[41,6,1,"_CPPv2N6paddle13ConvBaseLayer10outputSizeEiiii"],"paddle::ConvBaseLayer::outputX_":[41,7,1,"_CPPv2N6paddle13ConvBaseLayer8outputX_E"],"paddle::ConvBaseLayer::outputs_":[41,7,1,"_CPPv2N6paddle13ConvBaseLayer8outputs_E"],"paddle::ConvBaseLayer::paddingY_":[41,7,1,"_CPPv2N6paddle13ConvBaseLayer9paddingY_E"],"paddle::ConvBaseLayer::padding_":[41,7,1,"_CPPv2N6paddle13ConvBaseLayer8padding_E"],"paddle::ConvBaseLayer::sharedBiases_":[41,7,1,"_CPPv2N6paddle13ConvBaseLayer13sharedBiases_E"],"paddle::ConvBaseLayer::strideY_":[41,7,1,"_CPPv2N6paddle13ConvBaseLayer8strideY_E"],"paddle::ConvBaseLayer::stride_":[41,7,1,"_CPPv2N6paddle13ConvBaseLayer7stride_E"],"paddle::ConvBaseLayer::weights_":[41,7,1,"_CPPv2N6paddle13ConvBaseLayer8weights_E"],"paddle::ConvOperator":[41,8,1,"_CPPv2N6paddle12ConvOperatorE"],"paddle::ConvOperator::ConvOperator":[41,6,1,"_CPPv2N6paddle12ConvOperator12ConvOperatorERK14OperatorConfigb"],"paddle::ConvOperator::backward":[41,6,1,"_CPPv2N6paddle12ConvOperator8backwardEv"],"paddle::ConvOperator::forward":[41,6,1,"_CPPv2N6paddle12ConvOperator7forwardEv"],"paddle::ConvOperator::~ConvOperator":[41,6,1,"_CPPv2N6paddle12ConvOperatorD0Ev"],"paddle::ConvShiftLayer":[41,8,1,"_CPPv2N6paddle14ConvShiftLayerE"],"paddle::ConvShiftLayer::ConvShiftLayer":[41,6,1,"_CPPv2N6paddle14ConvShiftLayer14ConvShiftLayerERK11LayerConfig"],"paddle::ConvShiftLayer::backward":[41,6,1,"_CPPv2N6paddle14ConvShiftLayer8backwardERK14UpdateCallback"],"paddle::ConvShiftLayer::forward":[41,6,1,"_CPPv2N6paddle14ConvShiftLayer7forwardE8PassType"],"paddle::ConvShiftLayer::init":[41,6,1,"_CPPv2N6paddle14ConvShiftLayer4initERK8LayerMapRK12ParameterMap"],"paddle::ConvShiftLayer::~ConvShiftLayer":[41,6,1,"_CPPv2N6paddle14ConvShiftLayerD0Ev"],"paddle::ConvexCombinationLayer":[41,8,1,"_CPPv2N6paddle22ConvexCombinationLayerE"],"paddle::ConvexCombinationLayer::ConvexCombinationLayer":[41,6,1,"_CPPv2N6paddle22ConvexCombinationLayer22ConvexCombinationLayerERK11LayerConfig"],"paddle::ConvexCombinationLayer::backward":[41,6,1,"_CPPv2N6paddle22ConvexCombinationLayer8backwardERK14UpdateCallback"],"paddle::ConvexCombinationLayer::forward":[41,6,1,"_CPPv2N6paddle22ConvexCombinationLayer7forwardE8PassType"],"paddle::ConvexCombinationLayer::init":[41,6,1,"_CPPv2N6paddle22ConvexCombinationLayer4initERK8LayerMapRK12ParameterMap"],"paddle::ConvexCombinationLayer::tmpMtx0":[41,7,1,"_CPPv2N6paddle22ConvexCombinationLayer7tmpMtx0E"],"paddle::ConvexCombinationLayer::tmpRow0":[41,7,1,"_CPPv2N6paddle22ConvexCombinationLayer7tmpRow0E"],"paddle::ConvexCombinationLayer::tmpRow1":[41,7,1,"_CPPv2N6paddle22ConvexCombinationLayer7tmpRow1E"],"paddle::ConvexCombinationLayer::~ConvexCombinationLayer":[41,6,1,"_CPPv2N6paddle22ConvexCombinationLayerD0Ev"],"paddle::CoordIterator":[41,8,1,"_CPPv2N6paddle13CoordIteratorE"],"paddle::CoordIterator::CoordIterator":[41,6,1,"_CPPv2N6paddle13CoordIterator13CoordIteratorENSt6vectorIiEENSt6vectorIbEE"],"paddle::CoordIterator::begin":[41,6,1,"_CPPv2N6paddle13CoordIterator5beginEv"],"paddle::CoordIterator::curPos":[41,6,1,"_CPPv2N6paddle13CoordIterator6curPosEv"],"paddle::CoordIterator::curPos_":[41,7,1,"_CPPv2N6paddle13CoordIterator7curPos_E"],"paddle::CoordIterator::dims_":[41,7,1,"_CPPv2N6paddle13CoordIterator5dims_E"],"paddle::CoordIterator::directions_":[41,7,1,"_CPPv2N6paddle13CoordIterator11directions_E"],"paddle::CoordIterator::end":[41,6,1,"_CPPv2N6paddle13CoordIterator3endEv"],"paddle::CoordIterator::end_":[41,7,1,"_CPPv2N6paddle13CoordIterator4end_E"],"paddle::CoordIterator::getNextPos":[41,6,1,"_CPPv2N6paddle13CoordIterator10getNextPosERKNSt6vectorIiEEiRNSt6vectorIiEE"],"paddle::CoordIterator::getPrePos":[41,6,1,"_CPPv2N6paddle13CoordIterator9getPrePosERKNSt6vectorIiEEiRNSt6vectorIiEE"],"paddle::CoordIterator::offset":[41,6,1,"_CPPv2N6paddle13CoordIterator6offsetEv"],"paddle::CoordIterator::operator++":[41,6,1,"_CPPv2N6paddle13CoordIteratorppEv"],"paddle::CoordIterator::operator--":[41,6,1,"_CPPv2N6paddle13CoordIteratormmEv"],"paddle::CoordIterator::rbegin":[41,6,1,"_CPPv2N6paddle13CoordIterator6rbeginEv"],"paddle::CoordIterator::step":[41,6,1,"_CPPv2N6paddle13CoordIterator4stepE6size_tb"],"paddle::CosSimLayer":[41,8,1,"_CPPv2N6paddle11CosSimLayerE"],"paddle::CosSimLayer::CosSimLayer":[41,6,1,"_CPPv2N6paddle11CosSimLayer11CosSimLayerERK11LayerConfig"],"paddle::CosSimLayer::backward":[41,6,1,"_CPPv2N6paddle11CosSimLayer8backwardERK14UpdateCallback"],"paddle::CosSimLayer::forward":[41,6,1,"_CPPv2N6paddle11CosSimLayer7forwardE8PassType"],"paddle::CosSimLayer::init":[41,6,1,"_CPPv2N6paddle11CosSimLayer4initERK8LayerMapRK12ParameterMap"],"paddle::CosSimLayer::kCosSimScale_":[41,7,1,"_CPPv2N6paddle11CosSimLayer13kCosSimScale_E"],"paddle::CosSimLayer::~CosSimLayer":[41,6,1,"_CPPv2N6paddle11CosSimLayerD0Ev"],"paddle::CosSimVecMatLayer":[41,8,1,"_CPPv2N6paddle17CosSimVecMatLayerE"],"paddle::CosSimVecMatLayer::CosSimVecMatLayer":[41,6,1,"_CPPv2N6paddle17CosSimVecMatLayer17CosSimVecMatLayerERK11LayerConfig"],"paddle::CosSimVecMatLayer::backward":[41,6,1,"_CPPv2N6paddle17CosSimVecMatLayer8backwardERK14UpdateCallback"],"paddle::CosSimVecMatLayer::forward":[41,6,1,"_CPPv2N6paddle17CosSimVecMatLayer7forwardE8PassType"],"paddle::CosSimVecMatLayer::init":[41,6,1,"_CPPv2N6paddle17CosSimVecMatLayer4initERK8LayerMapRK12ParameterMap"],"paddle::CosSimVecMatLayer::tmpMtx0":[41,7,1,"_CPPv2N6paddle17CosSimVecMatLayer7tmpMtx0E"],"paddle::CosSimVecMatLayer::tmpMtx1":[41,7,1,"_CPPv2N6paddle17CosSimVecMatLayer7tmpMtx1E"],"paddle::CosSimVecMatLayer::tmpRow0":[41,7,1,"_CPPv2N6paddle17CosSimVecMatLayer7tmpRow0E"],"paddle::CosSimVecMatLayer::tmpRow1":[41,7,1,"_CPPv2N6paddle17CosSimVecMatLayer7tmpRow1E"],"paddle::CosSimVecMatLayer::tmpRow2":[41,7,1,"_CPPv2N6paddle17CosSimVecMatLayer7tmpRow2E"],"paddle::CosSimVecMatLayer::tmpRow3":[41,7,1,"_CPPv2N6paddle17CosSimVecMatLayer7tmpRow3E"],"paddle::CosSimVecMatLayer::~CosSimVecMatLayer":[41,6,1,"_CPPv2N6paddle17CosSimVecMatLayerD0Ev"],"paddle::CostLayer":[41,8,1,"_CPPv2N6paddle9CostLayerE"],"paddle::CostLayer::CostLayer":[41,6,1,"_CPPv2N6paddle9CostLayer9CostLayerERK11LayerConfig"],"paddle::CostLayer::backward":[41,6,1,"_CPPv2N6paddle9CostLayer8backwardERK14UpdateCallback"],"paddle::CostLayer::backwardImp":[41,6,1,"_CPPv2N6paddle9CostLayer11backwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::CostLayer::coeff_":[41,7,1,"_CPPv2N6paddle9CostLayer6coeff_E"],"paddle::CostLayer::forward":[41,6,1,"_CPPv2N6paddle9CostLayer7forwardE8PassType"],"paddle::CostLayer::forwardImp":[41,6,1,"_CPPv2N6paddle9CostLayer10forwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::CostLayer::getLabelLayer":[41,6,1,"_CPPv2N6paddle9CostLayer13getLabelLayerEv"],"paddle::CostLayer::getOutputLayer":[41,6,1,"_CPPv2N6paddle9CostLayer14getOutputLayerEv"],"paddle::CostLayer::init":[41,6,1,"_CPPv2N6paddle9CostLayer4initERK8LayerMapRK12ParameterMap"],"paddle::CostLayer::weightLayer_":[41,7,1,"_CPPv2N6paddle9CostLayer12weightLayer_E"],"paddle::CpuAllocator":[46,8,1,"_CPPv2N6paddle12CpuAllocatorE"],"paddle::CpuAllocator::alloc":[46,6,1,"_CPPv2N6paddle12CpuAllocator5allocE6size_t"],"paddle::CpuAllocator::free":[46,6,1,"_CPPv2N6paddle12CpuAllocator4freeEPv"],"paddle::CpuAllocator::getName":[46,6,1,"_CPPv2N6paddle12CpuAllocator7getNameEv"],"paddle::CpuAllocator::~CpuAllocator":[46,6,1,"_CPPv2N6paddle12CpuAllocatorD0Ev"],"paddle::CpuGpuVector":[44,9,1,"_CPPv2N6paddle12CpuGpuVectorE"],"paddle::CpuGpuVectorPtr":[44,9,1,"_CPPv2N6paddle15CpuGpuVectorPtrE"],"paddle::CpuGpuVectorT":[44,8,1,"_CPPv2N6paddle13CpuGpuVectorTE"],"paddle::CpuGpuVectorT::CpuGpuVectorT":[44,6,1,"_CPPv2N6paddle13CpuGpuVectorT13CpuGpuVectorTE6size_tb"],"paddle::CpuGpuVectorT::copyFrom":[44,6,1,"_CPPv2N6paddle13CpuGpuVectorT8copyFromERK7VectorTI1TE11hl_stream_t"],"paddle::CpuGpuVectorT::copyToCpu":[44,6,1,"_CPPv2N6paddle13CpuGpuVectorT9copyToCpuEPK1T6size_t"],"paddle::CpuGpuVectorT::copyToGpu":[44,6,1,"_CPPv2N6paddle13CpuGpuVectorT9copyToGpuEPK1T6size_t"],"paddle::CpuGpuVectorT::cpuVectorT_":[44,7,1,"_CPPv2N6paddle13CpuGpuVectorT11cpuVectorT_E"],"paddle::CpuGpuVectorT::create":[44,6,1,"_CPPv2N6paddle13CpuGpuVectorT6createE6size_tb"],"paddle::CpuGpuVectorT::fillSequence":[44,6,1,"_CPPv2N6paddle13CpuGpuVectorT12fillSequenceEb"],"paddle::CpuGpuVectorT::getData":[44,6,1,"_CPPv2N6paddle13CpuGpuVectorT7getDataEb"],"paddle::CpuGpuVectorT::getElement":[44,6,1,"_CPPv2N6paddle13CpuGpuVectorT10getElementE6size_t"],"paddle::CpuGpuVectorT::getMutableData":[44,6,1,"_CPPv2N6paddle13CpuGpuVectorT14getMutableDataEb"],"paddle::CpuGpuVectorT::getMutableVector":[44,6,1,"_CPPv2N6paddle13CpuGpuVectorT16getMutableVectorEb"],"paddle::CpuGpuVectorT::getSize":[44,6,1,"_CPPv2N6paddle13CpuGpuVectorT7getSizeEv"],"paddle::CpuGpuVectorT::getSync":[44,6,1,"_CPPv2N6paddle13CpuGpuVectorT7getSyncEv"],"paddle::CpuGpuVectorT::getVector":[44,6,1,"_CPPv2N6paddle13CpuGpuVectorT9getVectorEb"],"paddle::CpuGpuVectorT::gpuVectorT_":[44,7,1,"_CPPv2N6paddle13CpuGpuVectorT11gpuVectorT_E"],"paddle::CpuGpuVectorT::resize":[44,6,1,"_CPPv2N6paddle13CpuGpuVectorT6resizeE6size_tb"],"paddle::CpuGpuVectorT::resizeOrCreate":[44,6,1,"_CPPv2N6paddle13CpuGpuVectorT14resizeOrCreateERNSt10shared_ptrI13CpuGpuVectorTI1TEEE6size_tb"],"paddle::CpuGpuVectorT::setElement":[44,6,1,"_CPPv2N6paddle13CpuGpuVectorT10setElementE6size_tRK1Tb"],"paddle::CpuGpuVectorT::setSync":[44,6,1,"_CPPv2N6paddle13CpuGpuVectorT7setSyncEP10SyncedFlag"],"paddle::CpuGpuVectorT::syncFlag_":[44,7,1,"_CPPv2N6paddle13CpuGpuVectorT9syncFlag_E"],"paddle::CpuGpuVectorT::sync_":[44,7,1,"_CPPv2N6paddle13CpuGpuVectorT5sync_E"],"paddle::CpuGpuVectorT::zeroMem":[44,6,1,"_CPPv2N6paddle13CpuGpuVectorT7zeroMemEb"],"paddle::CpuGpuVectorT::~CpuGpuVectorT":[44,6,1,"_CPPv2N6paddle13CpuGpuVectorTD0Ev"],"paddle::CpuIVector":[44,9,1,"_CPPv2N6paddle10CpuIVectorE"],"paddle::CpuIVectorPtr":[44,9,1,"_CPPv2N6paddle13CpuIVectorPtrE"],"paddle::CpuMatrix":[44,8,1,"_CPPv2N6paddle9CpuMatrixE"],"paddle::CpuMatrix::CpuMatrix":[44,6,1,"_CPPv2N6paddle9CpuMatrix9CpuMatrixE6size_t6size_tb"],"paddle::CpuMatrix::accumulateColSum":[44,6,1,"_CPPv2N6paddle9CpuMatrix16accumulateColSumER6Matrix"],"paddle::CpuMatrix::addBias":[44,6,1,"_CPPv2N6paddle9CpuMatrix7addBiasER6Matrix4real"],"paddle::CpuMatrix::addByBitCode":[44,6,1,"_CPPv2N6paddle9CpuMatrix12addByBitCodeE6size_tRK7IVectorRK6Matrix"],"paddle::CpuMatrix::addByBitCodeBackward":[44,6,1,"_CPPv2N6paddle9CpuMatrix20addByBitCodeBackwardE6size_tRK7IVectorR6Matrix"],"paddle::CpuMatrix::addColumnVector":[44,6,1,"_CPPv2N6paddle9CpuMatrix15addColumnVectorERK6Matrix"],"paddle::CpuMatrix::addElements":[44,6,1,"_CPPv2N6paddle9CpuMatrix11addElementsER6MatrixR7IVector"],"paddle::CpuMatrix::addToRows":[44,6,1,"_CPPv2N6paddle9CpuMatrix9addToRowsER6MatrixR7IVector"],"paddle::CpuMatrix::addToRowsImp":[44,6,1,"_CPPv2N6paddle9CpuMatrix12addToRowsImpER12TableMatTypeR7IVector"],"paddle::CpuMatrix::avgPoolBackward":[44,6,1,"_CPPv2N6paddle9CpuMatrix15avgPoolBackwardER6Matrix6size_t6size_t6size_ti6size_t6size_t6size_t4real4real"],"paddle::CpuMatrix::avgPoolForward":[44,6,1,"_CPPv2N6paddle9CpuMatrix14avgPoolForwardER6Matrix6size_t6size_t6size_t6size_ti6size_t6size_t6size_t"],"paddle::CpuMatrix::check":[44,6,1,"_CPPv2N6paddle9CpuMatrix5checkERNSt7ostreamER6Matrixb"],"paddle::CpuMatrix::circularConv":[44,6,1,"_CPPv2N6paddle9CpuMatrix12circularConvER6MatrixR6Matrix"],"paddle::CpuMatrix::circularConvDerivative":[44,6,1,"_CPPv2N6paddle9CpuMatrix22circularConvDerivativeER6MatrixR6MatrixR6MatrixR6MatrixR6Matrix"],"paddle::CpuMatrix::classificationError":[44,6,1,"_CPPv2N6paddle9CpuMatrix19classificationErrorE9MatrixPtr10IVectorPtr"],"paddle::CpuMatrix::classificationErrorMulti":[44,6,1,"_CPPv2N6paddle9CpuMatrix24classificationErrorMultiER6MatrixR6Matrix4real"],"paddle::CpuMatrix::clone":[44,6,1,"_CPPv2N6paddle9CpuMatrix5cloneE6size_t6size_tb"],"paddle::CpuMatrix::colMax":[44,6,1,"_CPPv2N6paddle9CpuMatrix6colMaxER6Matrix"],"paddle::CpuMatrix::colMerge":[44,6,1,"_CPPv2N6paddle9CpuMatrix8colMergeER6Matrix"],"paddle::CpuMatrix::collectBias":[44,6,1,"_CPPv2N6paddle9CpuMatrix11collectBiasER6Matrix4real"],"paddle::CpuMatrix::contextProjectionBackward":[44,6,1,"_CPPv2N6paddle9CpuMatrix25contextProjectionBackwardE9MatrixPtr9MatrixPtrRK7IVectorii6size_tb"],"paddle::CpuMatrix::contextProjectionForward":[44,6,1,"_CPPv2N6paddle9CpuMatrix24contextProjectionForwardE9MatrixPtr9MatrixPtrRK7IVectorii6size_tb"],"paddle::CpuMatrix::convExpand":[44,6,1,"_CPPv2N6paddle9CpuMatrix10convExpandER6Matrixiiiiiiiiiii"],"paddle::CpuMatrix::convShrink":[44,6,1,"_CPPv2N6paddle9CpuMatrix10convShrinkER6Matrixiiiiiiiiiii4real4real"],"paddle::CpuMatrix::copyByRowIndex":[44,6,1,"_CPPv2N6paddle9CpuMatrix14copyByRowIndexER6MatrixR7IVector"],"paddle::CpuMatrix::copyFrom":[44,6,1,"_CPPv2N6paddle9CpuMatrix8copyFromERK6Matrix"],"paddle::CpuMatrix::cosSim":[44,6,1,"_CPPv2N6paddle9CpuMatrix6cosSimER6MatrixR6Matrix4real"],"paddle::CpuMatrix::cosSimDerivative":[44,6,1,"_CPPv2N6paddle9CpuMatrix16cosSimDerivativeER6MatrixR6MatrixR6MatrixR6MatrixR6Matrix4real"],"paddle::CpuMatrix::crossMapNormalBwd":[44,6,1,"_CPPv2N6paddle9CpuMatrix17crossMapNormalBwdER6MatrixR6MatrixR6MatrixR6Matrix6size_t6size_t6size_t6size_tffb"],"paddle::CpuMatrix::crossMapNormalFwd":[44,6,1,"_CPPv2N6paddle9CpuMatrix17crossMapNormalFwdER6Matrix6size_t6size_tR6Matrix6size_t6size_tffb"],"paddle::CpuMatrix::getAbsSum":[44,6,1,"_CPPv2N6paddle9CpuMatrix9getAbsSumEv"],"paddle::CpuMatrix::getElement":[44,6,1,"_CPPv2N6paddle9CpuMatrix10getElementE6size_t6size_t"],"paddle::CpuMatrix::getMax":[44,6,1,"_CPPv2N6paddle9CpuMatrix6getMaxEv"],"paddle::CpuMatrix::getMin":[44,6,1,"_CPPv2N6paddle9CpuMatrix6getMinEv"],"paddle::CpuMatrix::getRow":[44,6,1,"_CPPv2N6paddle9CpuMatrix6getRowE6size_t"],"paddle::CpuMatrix::getRowBuf":[44,6,1,"_CPPv2N6paddle9CpuMatrix9getRowBufE6size_t"],"paddle::CpuMatrix::getSum":[44,6,1,"_CPPv2N6paddle9CpuMatrix6getSumEv"],"paddle::CpuMatrix::getTranspose":[44,6,1,"_CPPv2N6paddle9CpuMatrix12getTransposeEv"],"paddle::CpuMatrix::leftMul":[44,6,1,"_CPPv2N6paddle9CpuMatrix7leftMulER6Matrix4real4real"],"paddle::CpuMatrix::maxPoolBackward":[44,6,1,"_CPPv2N6paddle9CpuMatrix15maxPoolBackwardER6Matrix6size_t6size_tR6MatrixR6Matrix6size_ti6size_t6size_t6size_t4real4real"],"paddle::CpuMatrix::maxPoolForward":[44,6,1,"_CPPv2N6paddle9CpuMatrix14maxPoolForwardER6Matrix6size_t6size_t6size_t6size_ti6size_t6size_t6size_t"],"paddle::CpuMatrix::maxSequenceBackward":[44,6,1,"_CPPv2N6paddle9CpuMatrix19maxSequenceBackwardER6MatrixRK7IVectorR7IVector"],"paddle::CpuMatrix::maxSequenceForward":[44,6,1,"_CPPv2N6paddle9CpuMatrix18maxSequenceForwardER6MatrixRK7IVectorR7IVector"],"paddle::CpuMatrix::mul":[44,6,1,"_CPPv2N6paddle9CpuMatrix3mulEK9MatrixPtrK9MatrixPtr4real4real"],"paddle::CpuMatrix::mulByBitCode":[44,6,1,"_CPPv2N6paddle9CpuMatrix12mulByBitCodeE6size_tRK7IVectorRK6MatrixRK6Matrix"],"paddle::CpuMatrix::mulByBitCodeBackwardError":[44,6,1,"_CPPv2N6paddle9CpuMatrix25mulByBitCodeBackwardErrorE6size_tRK7IVectorRK6MatrixR6Matrix"],"paddle::CpuMatrix::mulByBitCodeBackwardWeight":[44,6,1,"_CPPv2N6paddle9CpuMatrix26mulByBitCodeBackwardWeightE6size_tRK7IVectorR6MatrixRK6Matrix"],"paddle::CpuMatrix::multiBinaryLabelCrossEntropy":[44,6,1,"_CPPv2N6paddle9CpuMatrix28multiBinaryLabelCrossEntropyER6MatrixR6Matrix"],"paddle::CpuMatrix::multiBinaryLabelCrossEntropyBp":[44,6,1,"_CPPv2N6paddle9CpuMatrix30multiBinaryLabelCrossEntropyBpER6MatrixR6Matrix"],"paddle::CpuMatrix::oneHotCrossEntropy":[44,6,1,"_CPPv2N6paddle9CpuMatrix18oneHotCrossEntropyER6MatrixR7IVector"],"paddle::CpuMatrix::oneHotCrossEntropyBp":[44,6,1,"_CPPv2N6paddle9CpuMatrix20oneHotCrossEntropyBpER6MatrixR7IVector"],"paddle::CpuMatrix::oneHotCrossEntropyWithSelfNorm":[44,6,1,"_CPPv2N6paddle9CpuMatrix30oneHotCrossEntropyWithSelfNormER6MatrixR7IVector4real"],"paddle::CpuMatrix::oneHotCrossEntropyWithSelfNormBp":[44,6,1,"_CPPv2N6paddle9CpuMatrix32oneHotCrossEntropyWithSelfNormBpER6MatrixR7IVector4real"],"paddle::CpuMatrix::paramReluBackwardDiff":[44,6,1,"_CPPv2N6paddle9CpuMatrix21paramReluBackwardDiffER6MatrixR6MatrixR6Matrix"],"paddle::CpuMatrix::paramReluBackwardW":[44,6,1,"_CPPv2N6paddle9CpuMatrix18paramReluBackwardWER6MatrixR6Matrix"],"paddle::CpuMatrix::paramReluForward":[44,6,1,"_CPPv2N6paddle9CpuMatrix16paramReluForwardER6MatrixR6Matrix"],"paddle::CpuMatrix::print":[44,6,1,"_CPPv2N6paddle9CpuMatrix5printERNSt7ostreamE"],"paddle::CpuMatrix::printOneRow":[44,6,1,"_CPPv2N6paddle9CpuMatrix11printOneRowERNSt7ostreamE6size_t"],"paddle::CpuMatrix::randomizeUniform":[44,6,1,"_CPPv2N6paddle9CpuMatrix16randomizeUniformEv"],"paddle::CpuMatrix::resetOne":[44,6,1,"_CPPv2N6paddle9CpuMatrix8resetOneEv"],"paddle::CpuMatrix::resize":[44,6,1,"_CPPv2N6paddle9CpuMatrix6resizeE6size_t6size_t"],"paddle::CpuMatrix::rightMul":[44,6,1,"_CPPv2N6paddle9CpuMatrix8rightMulER6Matrix4real4real"],"paddle::CpuMatrix::rowMax":[44,6,1,"_CPPv2N6paddle9CpuMatrix6rowMaxER6Matrix"],"paddle::CpuMatrix::rowMaxId":[44,6,1,"_CPPv2N6paddle9CpuMatrix8rowMaxIdER7IVector"],"paddle::CpuMatrix::rowNormalizeL1":[44,6,1,"_CPPv2N6paddle9CpuMatrix14rowNormalizeL1ER6Matrix"],"paddle::CpuMatrix::rowSum":[44,6,1,"_CPPv2N6paddle9CpuMatrix6rowSumER6Matrix"],"paddle::CpuMatrix::scaledTanh":[44,6,1,"_CPPv2N6paddle9CpuMatrix10scaledTanhER6Matrix4real4real"],"paddle::CpuMatrix::selectElements":[44,6,1,"_CPPv2N6paddle9CpuMatrix14selectElementsER6MatrixR7IVector"],"paddle::CpuMatrix::selectRows":[44,6,1,"_CPPv2N6paddle9CpuMatrix10selectRowsER6MatrixR7IVector"],"paddle::CpuMatrix::selectRowsImp":[44,6,1,"_CPPv2N6paddle9CpuMatrix13selectRowsImpER12TableMatTypeR7IVector"],"paddle::CpuMatrix::sequenceAvgForward":[44,6,1,"_CPPv2N6paddle9CpuMatrix18sequenceAvgForwardER6MatrixRK7IVectori"],"paddle::CpuMatrix::sequenceSoftmax":[44,6,1,"_CPPv2N6paddle9CpuMatrix15sequenceSoftmaxER6MatrixRK7IVector"],"paddle::CpuMatrix::setRow":[44,6,1,"_CPPv2N6paddle9CpuMatrix6setRowE6size_t6size_tPKjPK4real"],"paddle::CpuMatrix::softmax":[44,6,1,"_CPPv2N6paddle9CpuMatrix7softmaxER6Matrix"],"paddle::CpuMatrix::softmaxDerivative":[44,6,1,"_CPPv2N6paddle9CpuMatrix17softmaxDerivativeER6MatrixR6Matrix"],"paddle::CpuMatrix::softrelu":[44,6,1,"_CPPv2N6paddle9CpuMatrix8softreluER6Matrix"],"paddle::CpuMatrix::softreluDerivative":[44,6,1,"_CPPv2N6paddle9CpuMatrix18softreluDerivativeER6Matrix"],"paddle::CpuMatrix::subByBitCode":[44,6,1,"_CPPv2N6paddle9CpuMatrix12subByBitCodeE6size_tR7IVector"],"paddle::CpuMatrix::sumByBitCode":[44,6,1,"_CPPv2N6paddle9CpuMatrix12sumByBitCodeE6size_tR7IVectorR6Matrix4real"],"paddle::CpuMatrix::sumOfSquares":[44,6,1,"_CPPv2N6paddle9CpuMatrix12sumOfSquaresER6MatrixR6Matrix"],"paddle::CpuMatrix::sumOfSquaresBp":[44,6,1,"_CPPv2N6paddle9CpuMatrix14sumOfSquaresBpER6MatrixR6Matrix"],"paddle::CpuMatrix::tanh":[44,6,1,"_CPPv2N6paddle9CpuMatrix4tanhER6Matrix"],"paddle::CpuMatrix::tanhDerivative":[44,6,1,"_CPPv2N6paddle9CpuMatrix14tanhDerivativeER6Matrix"],"paddle::CpuMatrix::transpose":[44,6,1,"_CPPv2N6paddle9CpuMatrix9transposeE9MatrixPtrb"],"paddle::CpuMatrix::zeroMem":[44,6,1,"_CPPv2N6paddle9CpuMatrix7zeroMemEv"],"paddle::CpuMatrix::~CpuMatrix":[44,6,1,"_CPPv2N6paddle9CpuMatrixD0Ev"],"paddle::CpuMatrixPtr":[44,9,1,"_CPPv2N6paddle12CpuMatrixPtrE"],"paddle::CpuMemHandlePtr":[46,9,1,"_CPPv2N6paddle15CpuMemHandlePtrE"],"paddle::CpuMemoryHandle":[46,8,1,"_CPPv2N6paddle15CpuMemoryHandleE"],"paddle::CpuMemoryHandle::CpuMemoryHandle":[46,6,1,"_CPPv2N6paddle15CpuMemoryHandle15CpuMemoryHandleE6size_t"],"paddle::CpuMemoryHandle::~CpuMemoryHandle":[46,6,1,"_CPPv2N6paddle15CpuMemoryHandleD0Ev"],"paddle::CpuSparseMatrix":[44,8,1,"_CPPv2N6paddle15CpuSparseMatrixE"],"paddle::CpuSparseMatrix::CpuSparseMatrix":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix15CpuSparseMatrixE6size_t6size_t6size_t15SparseValueType12SparseFormatb"],"paddle::CpuSparseMatrix::DEFAULT_AVG_WIDTH":[44,7,1,"_CPPv2N6paddle15CpuSparseMatrix17DEFAULT_AVG_WIDTHE"],"paddle::CpuSparseMatrix::add3":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix4add3EP9CpuMatrix"],"paddle::CpuSparseMatrix::addBias":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix7addBiasER6Matrix4real"],"paddle::CpuSparseMatrix::clone":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix5cloneE6size_t6size_tb"],"paddle::CpuSparseMatrix::cols_":[44,7,1,"_CPPv2N6paddle15CpuSparseMatrix5cols_E"],"paddle::CpuSparseMatrix::copyFrom":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix8copyFromERK15GpuSparseMatrix11hl_stream_t"],"paddle::CpuSparseMatrix::copyRow":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix7copyRowEi6size_tPK18sparse_non_value_t"],"paddle::CpuSparseMatrix::cpuLocalMats_":[44,7,1,"_CPPv2N6paddle15CpuSparseMatrix13cpuLocalMats_E"],"paddle::CpuSparseMatrix::fillRowIndices":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix14fillRowIndicesER10IVectorPtr"],"paddle::CpuSparseMatrix::format_":[44,7,1,"_CPPv2N6paddle15CpuSparseMatrix7format_E"],"paddle::CpuSparseMatrix::getColNum":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix9getColNumE6size_t"],"paddle::CpuSparseMatrix::getColStartIdx":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix14getColStartIdxE6size_t"],"paddle::CpuSparseMatrix::getCols":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix7getColsEv"],"paddle::CpuSparseMatrix::getColumn":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix9getColumnE6size_t"],"paddle::CpuSparseMatrix::getData":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix7getDataEv"],"paddle::CpuSparseMatrix::getFormat":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix9getFormatEv"],"paddle::CpuSparseMatrix::getMax":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix6getMaxEv"],"paddle::CpuSparseMatrix::getMin":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix6getMinEv"],"paddle::CpuSparseMatrix::getRowCols":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix10getRowColsE6size_t"],"paddle::CpuSparseMatrix::getRowNum":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix9getRowNumE6size_t"],"paddle::CpuSparseMatrix::getRowStartIdx":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix14getRowStartIdxE6size_t"],"paddle::CpuSparseMatrix::getRowValues":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix12getRowValuesE6size_t"],"paddle::CpuSparseMatrix::getRows":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix7getRowsEv"],"paddle::CpuSparseMatrix::getSum":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix6getSumEv"],"paddle::CpuSparseMatrix::getTmpSparseMatrix":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix18getTmpSparseMatrixE6size_t6size_t"],"paddle::CpuSparseMatrix::getTranspose":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix12getTransposeEv"],"paddle::CpuSparseMatrix::getValue":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix8getValueEv"],"paddle::CpuSparseMatrix::getValueType":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix12getValueTypeEv"],"paddle::CpuSparseMatrix::isSparse":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix8isSparseEv"],"paddle::CpuSparseMatrix::mul":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix3mulE9MatrixPtr9MatrixPtr4real4real"],"paddle::CpuSparseMatrix::print":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix5printERNSt7ostreamE"],"paddle::CpuSparseMatrix::printOneRow":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix11printOneRowERNSt7ostreamE6size_t"],"paddle::CpuSparseMatrix::randomizeUniform":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix16randomizeUniformEv"],"paddle::CpuSparseMatrix::resize":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix6resizeE6size_t6size_t6size_t15SparseValueType12SparseFormat"],"paddle::CpuSparseMatrix::rowMax":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix6rowMaxER7IVectorR6Matrix"],"paddle::CpuSparseMatrix::rows_":[44,7,1,"_CPPv2N6paddle15CpuSparseMatrix5rows_E"],"paddle::CpuSparseMatrix::setRow":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix6setRowE6size_t6size_tPKjPK4real"],"paddle::CpuSparseMatrix::sparseResize":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix12sparseResizeEv"],"paddle::CpuSparseMatrix::square":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix6squareEv"],"paddle::CpuSparseMatrix::subMatrix":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix9subMatrixE6size_t6size_t"],"paddle::CpuSparseMatrix::transpose":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix9transposeE9MatrixPtrb"],"paddle::CpuSparseMatrix::trimFrom":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix8trimFromERK15CpuSparseMatrix"],"paddle::CpuSparseMatrix::valueType_":[44,7,1,"_CPPv2N6paddle15CpuSparseMatrix10valueType_E"],"paddle::CpuSparseMatrix::value_":[44,7,1,"_CPPv2N6paddle15CpuSparseMatrix6value_E"],"paddle::CpuSparseMatrix::zeroMem":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrix7zeroMemEv"],"paddle::CpuSparseMatrix::~CpuSparseMatrix":[44,6,1,"_CPPv2N6paddle15CpuSparseMatrixD0Ev"],"paddle::CpuSparseMatrixPtr":[44,9,1,"_CPPv2N6paddle18CpuSparseMatrixPtrE"],"paddle::CpuVector":[44,9,1,"_CPPv2N6paddle9CpuVectorE"],"paddle::CpuVectorPtr":[44,9,1,"_CPPv2N6paddle12CpuVectorPtrE"],"paddle::CpuVectorT":[44,8,1,"_CPPv2N6paddle10CpuVectorTE"],"paddle::CpuVectorT::CpuVectorT":[44,6,1,"_CPPv2N6paddle10CpuVectorT10CpuVectorTE6size_t"],"paddle::CpuVectorT::copyFrom":[44,6,1,"_CPPv2N6paddle10CpuVectorT8copyFromEPK1T6size_t"],"paddle::CpuVectorT::copyTo":[44,6,1,"_CPPv2N6paddle10CpuVectorT6copyToEP10CpuVectorTI1TE"],"paddle::CpuVectorT::fillSequence":[44,6,1,"_CPPv2N6paddle10CpuVectorT12fillSequenceEv"],"paddle::CpuVectorT::get":[44,6,1,"_CPPv2N6paddle10CpuVectorT3getE6size_t"],"paddle::CpuVectorT::getAbsMax":[44,6,1,"_CPPv2N6paddle10CpuVectorT9getAbsMaxEv"],"paddle::CpuVectorT::getAbsSum":[44,6,1,"_CPPv2N6paddle10CpuVectorT9getAbsSumEv"],"paddle::CpuVectorT::getElement":[44,6,1,"_CPPv2N6paddle10CpuVectorT10getElementE6size_t"],"paddle::CpuVectorT::getMax":[44,6,1,"_CPPv2N6paddle10CpuVectorT6getMaxEv"],"paddle::CpuVectorT::getMin":[44,6,1,"_CPPv2N6paddle10CpuVectorT6getMinEv"],"paddle::CpuVectorT::getPoint":[44,6,1,"_CPPv2N6paddle10CpuVectorT8getPointEK8uint64_t"],"paddle::CpuVectorT::getSum":[44,6,1,"_CPPv2N6paddle10CpuVectorT6getSumEv"],"paddle::CpuVectorT::histogram":[44,6,1,"_CPPv2N6paddle10CpuVectorT9histogramERNSt7ostreamEi"],"paddle::CpuVectorT::isEqualTo":[44,6,1,"_CPPv2N6paddle10CpuVectorT9isEqualToERK7VectorTI1TERK1T"],"paddle::CpuVectorT::newMemory":[44,6,1,"_CPPv2N6paddle10CpuVectorT9newMemoryE6size_t"],"paddle::CpuVectorT::print":[44,6,1,"_CPPv2N6paddle10CpuVectorT5printERNSt7ostreamE6size_t"],"paddle::CpuVectorT::printOneElement":[44,6,1,"_CPPv2N6paddle10CpuVectorT15printOneElementERNSt7ostreamE6size_t"],"paddle::CpuVectorT::rand":[44,6,1,"_CPPv2N6paddle10CpuVectorT4randEv"],"paddle::CpuVectorT::randnorm":[44,6,1,"_CPPv2N6paddle10CpuVectorT8randnormE4real4real"],"paddle::CpuVectorT::reset":[44,6,1,"_CPPv2N6paddle10CpuVectorT5resetERK1T"],"paddle::CpuVectorT::selectFrom":[44,6,1,"_CPPv2N6paddle10CpuVectorT10selectFromERK7VectorTI1TERK7VectorTIiE"],"paddle::CpuVectorT::setElement":[44,6,1,"_CPPv2N6paddle10CpuVectorT10setElementE6size_tRK1T"],"paddle::CpuVectorT::uniform":[44,6,1,"_CPPv2N6paddle10CpuVectorT7uniformE4real4real"],"paddle::CpuVectorT::zeroMem":[44,6,1,"_CPPv2N6paddle10CpuVectorT7zeroMemEv"],"paddle::CudaHostAllocator":[46,8,1,"_CPPv2N6paddle17CudaHostAllocatorE"],"paddle::CudaHostAllocator::alloc":[46,6,1,"_CPPv2N6paddle17CudaHostAllocator5allocE6size_t"],"paddle::CudaHostAllocator::free":[46,6,1,"_CPPv2N6paddle17CudaHostAllocator4freeEPv"],"paddle::CudaHostAllocator::getName":[46,6,1,"_CPPv2N6paddle17CudaHostAllocator7getNameEv"],"paddle::CudaHostAllocator::~CudaHostAllocator":[46,6,1,"_CPPv2N6paddle17CudaHostAllocatorD0Ev"],"paddle::CudnnBatchNormLayer":[41,8,1,"_CPPv2N6paddle19CudnnBatchNormLayerE"],"paddle::CudnnBatchNormLayer::CudnnBatchNormLayer":[41,6,1,"_CPPv2N6paddle19CudnnBatchNormLayer19CudnnBatchNormLayerERK11LayerConfig"],"paddle::CudnnBatchNormLayer::EPS":[41,7,1,"_CPPv2N6paddle19CudnnBatchNormLayer3EPSE"],"paddle::CudnnBatchNormLayer::backward":[41,6,1,"_CPPv2N6paddle19CudnnBatchNormLayer8backwardERK14UpdateCallback"],"paddle::CudnnBatchNormLayer::bnParamDesc_":[41,7,1,"_CPPv2N6paddle19CudnnBatchNormLayer12bnParamDesc_E"],"paddle::CudnnBatchNormLayer::forward":[41,6,1,"_CPPv2N6paddle19CudnnBatchNormLayer7forwardE8PassType"],"paddle::CudnnBatchNormLayer::init":[41,6,1,"_CPPv2N6paddle19CudnnBatchNormLayer4initERK8LayerMapRK12ParameterMap"],"paddle::CudnnBatchNormLayer::ioDesc_":[41,7,1,"_CPPv2N6paddle19CudnnBatchNormLayer7ioDesc_E"],"paddle::CudnnBatchNormLayer::reshape":[41,6,1,"_CPPv2N6paddle19CudnnBatchNormLayer7reshapeEi"],"paddle::CudnnBatchNormLayer::tmpBiasGrad_":[41,7,1,"_CPPv2N6paddle19CudnnBatchNormLayer12tmpBiasGrad_E"],"paddle::CudnnBatchNormLayer::tmpWGrad_":[41,7,1,"_CPPv2N6paddle19CudnnBatchNormLayer9tmpWGrad_E"],"paddle::CudnnBatchNormLayer::~CudnnBatchNormLayer":[41,6,1,"_CPPv2N6paddle19CudnnBatchNormLayerD0Ev"],"paddle::CudnnConvLayer":[41,8,1,"_CPPv2N6paddle14CudnnConvLayerE"],"paddle::CudnnConvLayer::CudnnConvLayer":[41,6,1,"_CPPv2N6paddle14CudnnConvLayer14CudnnConvLayerERK11LayerConfig"],"paddle::CudnnConvLayer::addBiases":[41,6,1,"_CPPv2N6paddle14CudnnConvLayer9addBiasesEv"],"paddle::CudnnConvLayer::backward":[41,6,1,"_CPPv2N6paddle14CudnnConvLayer8backwardERK14UpdateCallback"],"paddle::CudnnConvLayer::biasDesc_":[41,7,1,"_CPPv2N6paddle14CudnnConvLayer9biasDesc_E"],"paddle::CudnnConvLayer::biasOffset_":[41,7,1,"_CPPv2N6paddle14CudnnConvLayer11biasOffset_E"],"paddle::CudnnConvLayer::bpropBiases":[41,6,1,"_CPPv2N6paddle14CudnnConvLayer11bpropBiasesEv"],"paddle::CudnnConvLayer::bwdDataAlgo_":[41,7,1,"_CPPv2N6paddle14CudnnConvLayer12bwdDataAlgo_E"],"paddle::CudnnConvLayer::bwdDataLimitBytes_":[41,7,1,"_CPPv2N6paddle14CudnnConvLayer18bwdDataLimitBytes_E"],"paddle::CudnnConvLayer::bwdFilterAlgo_":[41,7,1,"_CPPv2N6paddle14CudnnConvLayer14bwdFilterAlgo_E"],"paddle::CudnnConvLayer::bwdFilterLimitBytes_":[41,7,1,"_CPPv2N6paddle14CudnnConvLayer20bwdFilterLimitBytes_E"],"paddle::CudnnConvLayer::convDesc_":[41,7,1,"_CPPv2N6paddle14CudnnConvLayer9convDesc_E"],"paddle::CudnnConvLayer::filterDesc_":[41,7,1,"_CPPv2N6paddle14CudnnConvLayer11filterDesc_E"],"paddle::CudnnConvLayer::forward":[41,6,1,"_CPPv2N6paddle14CudnnConvLayer7forwardE8PassType"],"paddle::CudnnConvLayer::fwdAlgo_":[41,7,1,"_CPPv2N6paddle14CudnnConvLayer8fwdAlgo_E"],"paddle::CudnnConvLayer::fwdLimitBytes_":[41,7,1,"_CPPv2N6paddle14CudnnConvLayer14fwdLimitBytes_E"],"paddle::CudnnConvLayer::imageH_":[41,7,1,"_CPPv2N6paddle14CudnnConvLayer7imageH_E"],"paddle::CudnnConvLayer::imageW_":[41,7,1,"_CPPv2N6paddle14CudnnConvLayer7imageW_E"],"paddle::CudnnConvLayer::init":[41,6,1,"_CPPv2N6paddle14CudnnConvLayer4initERK8LayerMapRK12ParameterMap"],"paddle::CudnnConvLayer::inputDesc_":[41,7,1,"_CPPv2N6paddle14CudnnConvLayer10inputDesc_E"],"paddle::CudnnConvLayer::inputOffset_":[41,7,1,"_CPPv2N6paddle14CudnnConvLayer12inputOffset_E"],"paddle::CudnnConvLayer::isSelectAlgo_":[41,7,1,"_CPPv2N6paddle14CudnnConvLayer13isSelectAlgo_E"],"paddle::CudnnConvLayer::maxGroups_":[41,7,1,"_CPPv2N6paddle14CudnnConvLayer10maxGroups_E"],"paddle::CudnnConvLayer::outputDesc_":[41,7,1,"_CPPv2N6paddle14CudnnConvLayer11outputDesc_E"],"paddle::CudnnConvLayer::outputH_":[41,7,1,"_CPPv2N6paddle14CudnnConvLayer8outputH_E"],"paddle::CudnnConvLayer::outputOffset_":[41,7,1,"_CPPv2N6paddle14CudnnConvLayer13outputOffset_E"],"paddle::CudnnConvLayer::outputW_":[41,7,1,"_CPPv2N6paddle14CudnnConvLayer8outputW_E"],"paddle::CudnnConvLayer::reshape":[41,6,1,"_CPPv2N6paddle14CudnnConvLayer7reshapeEi"],"paddle::CudnnConvLayer::weightOffset_":[41,7,1,"_CPPv2N6paddle14CudnnConvLayer13weightOffset_E"],"paddle::CudnnConvLayer::workSpaceData_":[41,7,1,"_CPPv2N6paddle14CudnnConvLayer14workSpaceData_E"],"paddle::CudnnConvLayer::workSpaceInBytes_":[41,7,1,"_CPPv2N6paddle14CudnnConvLayer17workSpaceInBytes_E"],"paddle::CudnnConvLayer::workSpace_":[41,7,1,"_CPPv2N6paddle14CudnnConvLayer10workSpace_E"],"paddle::CudnnConvLayer::~CudnnConvLayer":[41,6,1,"_CPPv2N6paddle14CudnnConvLayerD0Ev"],"paddle::CudnnPoolLayer":[41,8,1,"_CPPv2N6paddle14CudnnPoolLayerE"],"paddle::CudnnPoolLayer::CudnnPoolLayer":[41,6,1,"_CPPv2N6paddle14CudnnPoolLayer14CudnnPoolLayerERK11LayerConfig"],"paddle::CudnnPoolLayer::backward":[41,6,1,"_CPPv2N6paddle14CudnnPoolLayer8backwardERK14UpdateCallback"],"paddle::CudnnPoolLayer::forward":[41,6,1,"_CPPv2N6paddle14CudnnPoolLayer7forwardE8PassType"],"paddle::CudnnPoolLayer::heightPadding":[41,7,1,"_CPPv2N6paddle14CudnnPoolLayer13heightPaddingE"],"paddle::CudnnPoolLayer::imageH_":[41,7,1,"_CPPv2N6paddle14CudnnPoolLayer7imageH_E"],"paddle::CudnnPoolLayer::imageW_":[41,7,1,"_CPPv2N6paddle14CudnnPoolLayer7imageW_E"],"paddle::CudnnPoolLayer::init":[41,6,1,"_CPPv2N6paddle14CudnnPoolLayer4initERK8LayerMapRK12ParameterMap"],"paddle::CudnnPoolLayer::inputDesc_":[41,7,1,"_CPPv2N6paddle14CudnnPoolLayer10inputDesc_E"],"paddle::CudnnPoolLayer::mode_":[41,7,1,"_CPPv2N6paddle14CudnnPoolLayer5mode_E"],"paddle::CudnnPoolLayer::outputDesc_":[41,7,1,"_CPPv2N6paddle14CudnnPoolLayer11outputDesc_E"],"paddle::CudnnPoolLayer::outputH_":[41,7,1,"_CPPv2N6paddle14CudnnPoolLayer8outputH_E"],"paddle::CudnnPoolLayer::outputSize":[41,6,1,"_CPPv2N6paddle14CudnnPoolLayer10outputSizeEiiii"],"paddle::CudnnPoolLayer::outputW_":[41,7,1,"_CPPv2N6paddle14CudnnPoolLayer8outputW_E"],"paddle::CudnnPoolLayer::poolingDesc_":[41,7,1,"_CPPv2N6paddle14CudnnPoolLayer12poolingDesc_E"],"paddle::CudnnPoolLayer::reshape":[41,6,1,"_CPPv2N6paddle14CudnnPoolLayer7reshapeEi"],"paddle::CudnnPoolLayer::strideHeight":[41,7,1,"_CPPv2N6paddle14CudnnPoolLayer12strideHeightE"],"paddle::CudnnPoolLayer::strideWidth":[41,7,1,"_CPPv2N6paddle14CudnnPoolLayer11strideWidthE"],"paddle::CudnnPoolLayer::typeCheck":[41,6,1,"_CPPv2N6paddle14CudnnPoolLayer9typeCheckERKNSt6stringEP17hl_pooling_mode_t"],"paddle::CudnnPoolLayer::widthPadding":[41,7,1,"_CPPv2N6paddle14CudnnPoolLayer12widthPaddingE"],"paddle::CudnnPoolLayer::windowHeight":[41,7,1,"_CPPv2N6paddle14CudnnPoolLayer12windowHeightE"],"paddle::CudnnPoolLayer::windowWidth":[41,7,1,"_CPPv2N6paddle14CudnnPoolLayer11windowWidthE"],"paddle::CudnnPoolLayer::~CudnnPoolLayer":[41,6,1,"_CPPv2N6paddle14CudnnPoolLayerD0Ev"],"paddle::CustomStackTrace":[60,8,1,"_CPPv2N6paddle16CustomStackTraceE"],"paddle::CustomStackTrace::clear":[60,6,1,"_CPPv2N6paddle16CustomStackTrace5clearEv"],"paddle::CustomStackTrace::dump":[60,6,1,"_CPPv2N6paddle16CustomStackTrace4dumpE8Callback"],"paddle::CustomStackTrace::pop":[60,6,1,"_CPPv2N6paddle16CustomStackTrace3popERK1T"],"paddle::CustomStackTrace::push":[60,6,1,"_CPPv2N6paddle16CustomStackTrace4pushERK1T"],"paddle::DATA_AT_CPU":[44,8,1,"_CPPv2N6paddle11DATA_AT_CPUE"],"paddle::DATA_AT_GPU":[44,8,1,"_CPPv2N6paddle11DATA_AT_GPUE"],"paddle::DataLayer":[41,8,1,"_CPPv2N6paddle9DataLayerE"],"paddle::DataLayer::DataLayer":[41,6,1,"_CPPv2N6paddle9DataLayer9DataLayerERK11LayerConfig"],"paddle::DataLayer::backward":[41,6,1,"_CPPv2N6paddle9DataLayer8backwardERK14UpdateCallback"],"paddle::DataLayer::copyOutputToOtherDevice":[41,6,1,"_CPPv2N6paddle9DataLayer23copyOutputToOtherDeviceEv"],"paddle::DataLayer::data_":[41,7,1,"_CPPv2N6paddle9DataLayer5data_E"],"paddle::DataLayer::forward":[41,6,1,"_CPPv2N6paddle9DataLayer7forwardE8PassType"],"paddle::DataLayer::prefetch":[41,6,1,"_CPPv2N6paddle9DataLayer8prefetchEv"],"paddle::DataLayer::setData":[41,6,1,"_CPPv2N6paddle9DataLayer7setDataERK8Argument"],"paddle::DataNormLayer":[41,8,1,"_CPPv2N6paddle13DataNormLayerE"],"paddle::DataNormLayer::DataNormLayer":[41,6,1,"_CPPv2N6paddle13DataNormLayer13DataNormLayerERK11LayerConfig"],"paddle::DataNormLayer::NormalizationStrategy":[41,9,1,"_CPPv2N6paddle13DataNormLayer21NormalizationStrategyE"],"paddle::DataNormLayer::backward":[41,6,1,"_CPPv2N6paddle13DataNormLayer8backwardERK14UpdateCallback"],"paddle::DataNormLayer::decimalReciprocal_":[41,7,1,"_CPPv2N6paddle13DataNormLayer18decimalReciprocal_E"],"paddle::DataNormLayer::forward":[41,6,1,"_CPPv2N6paddle13DataNormLayer7forwardE8PassType"],"paddle::DataNormLayer::init":[41,6,1,"_CPPv2N6paddle13DataNormLayer4initERK8LayerMapRK12ParameterMap"],"paddle::DataNormLayer::kDecimalScaling":[41,8,1,"_CPPv2N6paddle13DataNormLayer15kDecimalScalingE"],"paddle::DataNormLayer::kMinMax":[41,8,1,"_CPPv2N6paddle13DataNormLayer7kMinMaxE"],"paddle::DataNormLayer::kZScore":[41,8,1,"_CPPv2N6paddle13DataNormLayer7kZScoreE"],"paddle::DataNormLayer::mean_":[41,7,1,"_CPPv2N6paddle13DataNormLayer5mean_E"],"paddle::DataNormLayer::min_":[41,7,1,"_CPPv2N6paddle13DataNormLayer4min_E"],"paddle::DataNormLayer::mode_":[41,7,1,"_CPPv2N6paddle13DataNormLayer5mode_E"],"paddle::DataNormLayer::rangeReciprocal_":[41,7,1,"_CPPv2N6paddle13DataNormLayer16rangeReciprocal_E"],"paddle::DataNormLayer::stdReciprocal_":[41,7,1,"_CPPv2N6paddle13DataNormLayer14stdReciprocal_E"],"paddle::DataNormLayer::weight_":[41,7,1,"_CPPv2N6paddle13DataNormLayer7weight_E"],"paddle::DataNormLayer::~DataNormLayer":[41,6,1,"_CPPv2N6paddle13DataNormLayerD0Ev"],"paddle::DataProvider":[34,8,1,"_CPPv2N6paddle12DataProviderE"],"paddle::DataProvider::DataProvider":[34,6,1,"_CPPv2N6paddle12DataProvider12DataProviderERK10DataConfigb"],"paddle::DataProvider::config_":[34,7,1,"_CPPv2N6paddle12DataProvider7config_E"],"paddle::DataProvider::constantSlots_":[34,7,1,"_CPPv2N6paddle12DataProvider14constantSlots_E"],"paddle::DataProvider::create":[34,6,1,"_CPPv2N6paddle12DataProvider6createERK10DataConfigb"],"paddle::DataProvider::doubleBuffer_":[34,7,1,"_CPPv2N6paddle12DataProvider13doubleBuffer_E"],"paddle::DataProvider::getConfig":[34,6,1,"_CPPv2N6paddle12DataProvider9getConfigEv"],"paddle::DataProvider::getNextBatch":[34,6,1,"_CPPv2N6paddle12DataProvider12getNextBatchE7int64_tP9DataBatch"],"paddle::DataProvider::getNextBatchFromBuffer":[34,6,1,"_CPPv2N6paddle12DataProvider22getNextBatchFromBufferE7int64_tP9DataBatch"],"paddle::DataProvider::getNextBatchInternal":[34,6,1,"_CPPv2N6paddle12DataProvider20getNextBatchInternalE7int64_tP9DataBatch"],"paddle::DataProvider::getSize":[34,6,1,"_CPPv2N6paddle12DataProvider7getSizeEv"],"paddle::DataProvider::initAsyncLoader":[34,6,1,"_CPPv2N6paddle12DataProvider15initAsyncLoaderEv"],"paddle::DataProvider::registrar_":[34,7,1,"_CPPv2N6paddle12DataProvider10registrar_E"],"paddle::DataProvider::reset":[34,6,1,"_CPPv2N6paddle12DataProvider5resetEv"],"paddle::DataProvider::setSkipShuffle":[34,6,1,"_CPPv2N6paddle12DataProvider14setSkipShuffleEv"],"paddle::DataProvider::shuffle":[34,6,1,"_CPPv2N6paddle12DataProvider7shuffleEv"],"paddle::DataProvider::skipShuffle_":[34,7,1,"_CPPv2N6paddle12DataProvider12skipShuffle_E"],"paddle::DataProvider::usageRatio_":[34,7,1,"_CPPv2N6paddle12DataProvider11usageRatio_E"],"paddle::DataProvider::useGpu_":[34,7,1,"_CPPv2N6paddle12DataProvider7useGpu_E"],"paddle::DataProvider::~DataProvider":[34,6,1,"_CPPv2N6paddle12DataProviderD0Ev"],"paddle::DataProviderGroup":[34,8,1,"_CPPv2N6paddle17DataProviderGroupE"],"paddle::DataProviderGroup::DataProviderGroup":[34,6,1,"_CPPv2N6paddle17DataProviderGroup17DataProviderGroupERK10DataConfigb"],"paddle::DataProviderGroup::ProviderPtrType":[34,9,1,"_CPPv2N6paddle17DataProviderGroup15ProviderPtrTypeE"],"paddle::DataProviderGroup::ProviderType":[34,9,1,"_CPPv2N6paddle17DataProviderGroup12ProviderTypeE"],"paddle::DataProviderGroup::fileList_":[34,7,1,"_CPPv2N6paddle17DataProviderGroup9fileList_E"],"paddle::DataProviderGroup::getNextBatchInternal":[34,6,1,"_CPPv2N6paddle17DataProviderGroup20getNextBatchInternalE7int64_tP9DataBatch"],"paddle::DataProviderGroup::getSize":[34,6,1,"_CPPv2N6paddle17DataProviderGroup7getSizeEv"],"paddle::DataProviderGroup::loader_":[34,7,1,"_CPPv2N6paddle17DataProviderGroup7loader_E"],"paddle::DataProviderGroup::lock_":[34,7,1,"_CPPv2N6paddle17DataProviderGroup5lock_E"],"paddle::DataProviderGroup::provider_":[34,7,1,"_CPPv2N6paddle17DataProviderGroup9provider_E"],"paddle::DataProviderGroup::reset":[34,6,1,"_CPPv2N6paddle17DataProviderGroup5resetEv"],"paddle::DataProviderGroup::shuffle":[34,6,1,"_CPPv2N6paddle17DataProviderGroup7shuffleEv"],"paddle::DataProviderGroup::~DataProviderGroup":[34,6,1,"_CPPv2N6paddle17DataProviderGroupD0Ev"],"paddle::DecayedAdagradParameterOptimizer":[48,8,1,"_CPPv2N6paddle32DecayedAdagradParameterOptimizerE"],"paddle::DecayedAdagradParameterOptimizer::DecayedAdagradParameterOptimizer":[48,6,1,"_CPPv2N6paddle32DecayedAdagradParameterOptimizer32DecayedAdagradParameterOptimizerERK18OptimizationConfig"],"paddle::DecayedAdagradParameterOptimizer::epsilon_":[48,7,1,"_CPPv2N6paddle32DecayedAdagradParameterOptimizer8epsilon_E"],"paddle::DecayedAdagradParameterOptimizer::finishBatch":[48,6,1,"_CPPv2N6paddle32DecayedAdagradParameterOptimizer11finishBatchEv"],"paddle::DecayedAdagradParameterOptimizer::init":[48,6,1,"_CPPv2N6paddle32DecayedAdagradParameterOptimizer4initE6size_tPK15ParameterConfig"],"paddle::DecayedAdagradParameterOptimizer::rou_":[48,7,1,"_CPPv2N6paddle32DecayedAdagradParameterOptimizer4rou_E"],"paddle::DecayedAdagradParameterOptimizer::startBatch":[48,6,1,"_CPPv2N6paddle32DecayedAdagradParameterOptimizer10startBatchE7int64_t"],"paddle::DecayedAdagradParameterOptimizer::t0Vec_":[48,7,1,"_CPPv2N6paddle32DecayedAdagradParameterOptimizer6t0Vec_E"],"paddle::DecayedAdagradParameterOptimizer::timer_":[48,7,1,"_CPPv2N6paddle32DecayedAdagradParameterOptimizer6timer_E"],"paddle::DecayedAdagradParameterOptimizer::update":[48,6,1,"_CPPv2N6paddle32DecayedAdagradParameterOptimizer6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::DenseScanner":[34,8,1,"_CPPv2N6paddle12DenseScannerE"],"paddle::DenseScanner::DenseScanner":[34,6,1,"_CPPv2N6paddle12DenseScanner12DenseScannerEP10SlotHeader"],"paddle::DenseScanner::fill":[34,6,1,"_CPPv2N6paddle12DenseScanner4fillER8ArgumentP8PyObject"],"paddle::DenseScanner::finishPrepare":[34,6,1,"_CPPv2N6paddle12DenseScanner13finishPrepareER8Argument"],"paddle::DenseScanner::prepare":[34,6,1,"_CPPv2N6paddle12DenseScanner7prepareER8ArgumentP8PyObject"],"paddle::DotMulOperator":[41,8,1,"_CPPv2N6paddle14DotMulOperatorE"],"paddle::DotMulOperator::DotMulOperator":[41,6,1,"_CPPv2N6paddle14DotMulOperator14DotMulOperatorERK14OperatorConfigb"],"paddle::DotMulOperator::backward":[41,6,1,"_CPPv2N6paddle14DotMulOperator8backwardEv"],"paddle::DotMulOperator::forward":[41,6,1,"_CPPv2N6paddle14DotMulOperator7forwardEv"],"paddle::DotMulProjection":[41,8,1,"_CPPv2N6paddle16DotMulProjectionE"],"paddle::DotMulProjection::DotMulProjection":[41,6,1,"_CPPv2N6paddle16DotMulProjection16DotMulProjectionERK16ProjectionConfigRK12ParameterPtrb"],"paddle::DotMulProjection::backward":[41,6,1,"_CPPv2N6paddle16DotMulProjection8backwardERK14UpdateCallback"],"paddle::DotMulProjection::forward":[41,6,1,"_CPPv2N6paddle16DotMulProjection7forwardEv"],"paddle::DotMulProjection::weight_":[41,7,1,"_CPPv2N6paddle16DotMulProjection7weight_E"],"paddle::DummyOptimizer":[48,8,1,"_CPPv2N6paddle14DummyOptimizerE"],"paddle::DummyOptimizer::DummyOptimizer":[48,6,1,"_CPPv2N6paddle14DummyOptimizer14DummyOptimizerERK18OptimizationConfig"],"paddle::DummyOptimizer::update":[48,6,1,"_CPPv2N6paddle14DummyOptimizer6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::EosIdCheckLayer":[41,8,1,"_CPPv2N6paddle15EosIdCheckLayerE"],"paddle::EosIdCheckLayer::EosIdCheckLayer":[41,6,1,"_CPPv2N6paddle15EosIdCheckLayer15EosIdCheckLayerERK11LayerConfig"],"paddle::EosIdCheckLayer::backward":[41,6,1,"_CPPv2N6paddle15EosIdCheckLayer8backwardERK14UpdateCallback"],"paddle::EosIdCheckLayer::forward":[41,6,1,"_CPPv2N6paddle15EosIdCheckLayer7forwardE8PassType"],"paddle::EosIdCheckLayer::init":[41,6,1,"_CPPv2N6paddle15EosIdCheckLayer4initERK8LayerMapRK12ParameterMap"],"paddle::Evaluator":[36,8,1,"_CPPv2N6paddle9EvaluatorE"],"paddle::Evaluator::Evaluator":[36,6,1,"_CPPv2N6paddle9Evaluator9EvaluatorEv"],"paddle::Evaluator::config_":[36,7,1,"_CPPv2N6paddle9Evaluator7config_E"],"paddle::Evaluator::create":[36,6,1,"_CPPv2N6paddle9Evaluator6createERK15EvaluatorConfig"],"paddle::Evaluator::distributeEval":[36,6,1,"_CPPv2N6paddle9Evaluator14distributeEvalEP16ParameterClient2"],"paddle::Evaluator::eval":[36,6,1,"_CPPv2N6paddle9Evaluator4evalERK13NeuralNetwork"],"paddle::Evaluator::evalImp":[36,6,1,"_CPPv2N6paddle9Evaluator7evalImpERNSt6vectorI8ArgumentEE"],"paddle::Evaluator::finish":[36,6,1,"_CPPv2N6paddle9Evaluator6finishEv"],"paddle::Evaluator::init":[36,6,1,"_CPPv2N6paddle9Evaluator4initERK15EvaluatorConfig"],"paddle::Evaluator::mergeResultsOfAllClients":[36,6,1,"_CPPv2N6paddle9Evaluator24mergeResultsOfAllClientsEP16ParameterClient2"],"paddle::Evaluator::numSamples_":[36,7,1,"_CPPv2N6paddle9Evaluator11numSamples_E"],"paddle::Evaluator::printStats":[36,6,1,"_CPPv2N6paddle9Evaluator10printStatsERNSt7ostreamE"],"paddle::Evaluator::registrar_":[36,7,1,"_CPPv2N6paddle9Evaluator10registrar_E"],"paddle::Evaluator::start":[36,6,1,"_CPPv2N6paddle9Evaluator5startEv"],"paddle::Evaluator::totalScore_":[36,7,1,"_CPPv2N6paddle9Evaluator11totalScore_E"],"paddle::Evaluator::updateSamplesNum":[36,6,1,"_CPPv2N6paddle9Evaluator16updateSamplesNumERKNSt6vectorI8ArgumentEE"],"paddle::Evaluator::~Evaluator":[36,6,1,"_CPPv2N6paddle9EvaluatorD0Ev"],"paddle::ExpandConvLayer":[41,8,1,"_CPPv2N6paddle15ExpandConvLayerE"],"paddle::ExpandConvLayer::ExpandConvLayer":[41,6,1,"_CPPv2N6paddle15ExpandConvLayer15ExpandConvLayerERK11LayerConfig"],"paddle::ExpandConvLayer::addSharedBias":[41,6,1,"_CPPv2N6paddle15ExpandConvLayer13addSharedBiasEv"],"paddle::ExpandConvLayer::addUnsharedBias":[41,6,1,"_CPPv2N6paddle15ExpandConvLayer15addUnsharedBiasEv"],"paddle::ExpandConvLayer::backward":[41,6,1,"_CPPv2N6paddle15ExpandConvLayer8backwardERK14UpdateCallback"],"paddle::ExpandConvLayer::bpropActs":[41,6,1,"_CPPv2N6paddle15ExpandConvLayer9bpropActsE9MatrixPtri"],"paddle::ExpandConvLayer::bpropBiases":[41,6,1,"_CPPv2N6paddle15ExpandConvLayer11bpropBiasesE9MatrixPtr"],"paddle::ExpandConvLayer::bpropSharedBias":[41,6,1,"_CPPv2N6paddle15ExpandConvLayer15bpropSharedBiasE9MatrixPtr9MatrixPtr"],"paddle::ExpandConvLayer::bpropWeights":[41,6,1,"_CPPv2N6paddle15ExpandConvLayer12bpropWeightsE9MatrixPtri"],"paddle::ExpandConvLayer::expandFwdOnce":[41,6,1,"_CPPv2N6paddle15ExpandConvLayer13expandFwdOnceE9MatrixPtrii"],"paddle::ExpandConvLayer::expandInput_":[41,7,1,"_CPPv2N6paddle15ExpandConvLayer12expandInput_E"],"paddle::ExpandConvLayer::expandOneFrame":[41,6,1,"_CPPv2N6paddle15ExpandConvLayer14expandOneFrameE9MatrixPtr6size_ti"],"paddle::ExpandConvLayer::forward":[41,6,1,"_CPPv2N6paddle15ExpandConvLayer7forwardE8PassType"],"paddle::ExpandConvLayer::getSize":[41,6,1,"_CPPv2N6paddle15ExpandConvLayer7getSizeEv"],"paddle::ExpandConvLayer::imgSizeH_":[41,7,1,"_CPPv2N6paddle15ExpandConvLayer9imgSizeH_E"],"paddle::ExpandConvLayer::imgSizeW_":[41,7,1,"_CPPv2N6paddle15ExpandConvLayer9imgSizeW_E"],"paddle::ExpandConvLayer::init":[41,6,1,"_CPPv2N6paddle15ExpandConvLayer4initERK8LayerMapRK12ParameterMap"],"paddle::ExpandConvLayer::outputH_":[41,7,1,"_CPPv2N6paddle15ExpandConvLayer8outputH_E"],"paddle::ExpandConvLayer::outputW_":[41,7,1,"_CPPv2N6paddle15ExpandConvLayer8outputW_E"],"paddle::ExpandConvLayer::resetConvOutput":[41,6,1,"_CPPv2N6paddle15ExpandConvLayer15resetConvOutputE6size_ti"],"paddle::ExpandConvLayer::resetExpandInput":[41,6,1,"_CPPv2N6paddle15ExpandConvLayer16resetExpandInputE6size_t6size_t"],"paddle::ExpandConvLayer::subK_":[41,7,1,"_CPPv2N6paddle15ExpandConvLayer5subK_E"],"paddle::ExpandConvLayer::subM_":[41,7,1,"_CPPv2N6paddle15ExpandConvLayer5subM_E"],"paddle::ExpandConvLayer::subN_":[41,7,1,"_CPPv2N6paddle15ExpandConvLayer5subN_E"],"paddle::ExpandConvLayer::transOutValue_":[41,7,1,"_CPPv2N6paddle15ExpandConvLayer14transOutValue_E"],"paddle::ExpandConvLayer::~ExpandConvLayer":[41,6,1,"_CPPv2N6paddle15ExpandConvLayerD0Ev"],"paddle::ExpandLayer":[41,8,1,"_CPPv2N6paddle11ExpandLayerE"],"paddle::ExpandLayer::ExpandLayer":[41,6,1,"_CPPv2N6paddle11ExpandLayer11ExpandLayerERK11LayerConfig"],"paddle::ExpandLayer::ExpandLevel":[41,9,1,"_CPPv2N6paddle11ExpandLayer11ExpandLevelE"],"paddle::ExpandLayer::backward":[41,6,1,"_CPPv2N6paddle11ExpandLayer8backwardERK14UpdateCallback"],"paddle::ExpandLayer::biases_":[41,7,1,"_CPPv2N6paddle11ExpandLayer7biases_E"],"paddle::ExpandLayer::cpuExpandStartsPos_":[41,7,1,"_CPPv2N6paddle11ExpandLayer19cpuExpandStartsPos_E"],"paddle::ExpandLayer::expandStartsPos_":[41,7,1,"_CPPv2N6paddle11ExpandLayer16expandStartsPos_E"],"paddle::ExpandLayer::forward":[41,6,1,"_CPPv2N6paddle11ExpandLayer7forwardE8PassType"],"paddle::ExpandLayer::init":[41,6,1,"_CPPv2N6paddle11ExpandLayer4initERK8LayerMapRK12ParameterMap"],"paddle::ExpandLayer::kNonSeq":[41,8,1,"_CPPv2N6paddle11ExpandLayer7kNonSeqE"],"paddle::ExpandLayer::kSeq":[41,8,1,"_CPPv2N6paddle11ExpandLayer4kSeqE"],"paddle::ExpandLayer::type_":[41,7,1,"_CPPv2N6paddle11ExpandLayer5type_E"],"paddle::ExpandLayer::~ExpandLayer":[41,6,1,"_CPPv2N6paddle11ExpandLayerD0Ev"],"paddle::FLOAT_VALUE":[44,8,1,"_CPPv2N6paddle11FLOAT_VALUEE"],"paddle::FeatureMapExpandLayer":[41,8,1,"_CPPv2N6paddle21FeatureMapExpandLayerE"],"paddle::FeatureMapExpandLayer::FeatureMapExpandLayer":[41,6,1,"_CPPv2N6paddle21FeatureMapExpandLayer21FeatureMapExpandLayerERK11LayerConfig"],"paddle::FeatureMapExpandLayer::backward":[41,6,1,"_CPPv2N6paddle21FeatureMapExpandLayer8backwardERK14UpdateCallback"],"paddle::FeatureMapExpandLayer::forward":[41,6,1,"_CPPv2N6paddle21FeatureMapExpandLayer7forwardE8PassType"],"paddle::FeatureMapExpandLayer::init":[41,6,1,"_CPPv2N6paddle21FeatureMapExpandLayer4initERK8LayerMapRK12ParameterMap"],"paddle::FeatureMapExpandLayer::~FeatureMapExpandLayer":[41,6,1,"_CPPv2N6paddle21FeatureMapExpandLayerD0Ev"],"paddle::FullMatrixProjection":[41,8,1,"_CPPv2N6paddle20FullMatrixProjectionE"],"paddle::FullMatrixProjection::FullMatrixProjection":[41,6,1,"_CPPv2N6paddle20FullMatrixProjection20FullMatrixProjectionERK16ProjectionConfigRK12ParameterPtrb"],"paddle::FullMatrixProjection::backward":[41,6,1,"_CPPv2N6paddle20FullMatrixProjection8backwardERK14UpdateCallback"],"paddle::FullMatrixProjection::forward":[41,6,1,"_CPPv2N6paddle20FullMatrixProjection7forwardEv"],"paddle::FullMatrixProjection::weight_":[41,7,1,"_CPPv2N6paddle20FullMatrixProjection7weight_E"],"paddle::FullyConnectedLayer":[41,8,1,"_CPPv2N6paddle19FullyConnectedLayerE"],"paddle::FullyConnectedLayer::FullyConnectedLayer":[41,6,1,"_CPPv2N6paddle19FullyConnectedLayer19FullyConnectedLayerERK11LayerConfig"],"paddle::FullyConnectedLayer::backward":[41,6,1,"_CPPv2N6paddle19FullyConnectedLayer8backwardERK14UpdateCallback"],"paddle::FullyConnectedLayer::biases_":[41,7,1,"_CPPv2N6paddle19FullyConnectedLayer7biases_E"],"paddle::FullyConnectedLayer::forward":[41,6,1,"_CPPv2N6paddle19FullyConnectedLayer7forwardE8PassType"],"paddle::FullyConnectedLayer::getWeight":[41,6,1,"_CPPv2N6paddle19FullyConnectedLayer9getWeightEi"],"paddle::FullyConnectedLayer::init":[41,6,1,"_CPPv2N6paddle19FullyConnectedLayer4initERK8LayerMapRK12ParameterMap"],"paddle::FullyConnectedLayer::prefetch":[41,6,1,"_CPPv2N6paddle19FullyConnectedLayer8prefetchEv"],"paddle::FullyConnectedLayer::weights_":[41,7,1,"_CPPv2N6paddle19FullyConnectedLayer8weights_E"],"paddle::FullyConnectedLayer::~FullyConnectedLayer":[41,6,1,"_CPPv2N6paddle19FullyConnectedLayerD0Ev"],"paddle::GatedRecurrentLayer":[41,8,1,"_CPPv2N6paddle19GatedRecurrentLayerE"],"paddle::GatedRecurrentLayer::GatedRecurrentLayer":[41,6,1,"_CPPv2N6paddle19GatedRecurrentLayer19GatedRecurrentLayerERK11LayerConfig"],"paddle::GatedRecurrentLayer::activationGate_":[41,7,1,"_CPPv2N6paddle19GatedRecurrentLayer15activationGate_E"],"paddle::GatedRecurrentLayer::backward":[41,6,1,"_CPPv2N6paddle19GatedRecurrentLayer8backwardERK14UpdateCallback"],"paddle::GatedRecurrentLayer::backwardBatch":[41,6,1,"_CPPv2N6paddle19GatedRecurrentLayer13backwardBatchEi9MatrixPtr"],"paddle::GatedRecurrentLayer::backwardSequence":[41,6,1,"_CPPv2N6paddle19GatedRecurrentLayer16backwardSequenceEi6size_tPKi9MatrixPtr"],"paddle::GatedRecurrentLayer::batchGrad_":[41,7,1,"_CPPv2N6paddle19GatedRecurrentLayer10batchGrad_E"],"paddle::GatedRecurrentLayer::batchValue_":[41,7,1,"_CPPv2N6paddle19GatedRecurrentLayer11batchValue_E"],"paddle::GatedRecurrentLayer::bias_":[41,7,1,"_CPPv2N6paddle19GatedRecurrentLayer5bias_E"],"paddle::GatedRecurrentLayer::forward":[41,6,1,"_CPPv2N6paddle19GatedRecurrentLayer7forwardE8PassType"],"paddle::GatedRecurrentLayer::forwardBatch":[41,6,1,"_CPPv2N6paddle19GatedRecurrentLayer12forwardBatchEi6size_tPKi9MatrixPtr"],"paddle::GatedRecurrentLayer::forwardSequence":[41,6,1,"_CPPv2N6paddle19GatedRecurrentLayer15forwardSequenceEi6size_tPKi9MatrixPtr"],"paddle::GatedRecurrentLayer::gateWeight_":[41,7,1,"_CPPv2N6paddle19GatedRecurrentLayer11gateWeight_E"],"paddle::GatedRecurrentLayer::gate_":[41,7,1,"_CPPv2N6paddle19GatedRecurrentLayer5gate_E"],"paddle::GatedRecurrentLayer::getState":[41,6,1,"_CPPv2N6paddle19GatedRecurrentLayer8getStateEv"],"paddle::GatedRecurrentLayer::init":[41,6,1,"_CPPv2N6paddle19GatedRecurrentLayer4initERK8LayerMapRK12ParameterMap"],"paddle::GatedRecurrentLayer::prevOutput_":[41,7,1,"_CPPv2N6paddle19GatedRecurrentLayer11prevOutput_E"],"paddle::GatedRecurrentLayer::resetOutput_":[41,7,1,"_CPPv2N6paddle19GatedRecurrentLayer12resetOutput_E"],"paddle::GatedRecurrentLayer::resetState":[41,6,1,"_CPPv2N6paddle19GatedRecurrentLayer10resetStateEv"],"paddle::GatedRecurrentLayer::reversed_":[41,7,1,"_CPPv2N6paddle19GatedRecurrentLayer9reversed_E"],"paddle::GatedRecurrentLayer::setState":[41,6,1,"_CPPv2N6paddle19GatedRecurrentLayer8setStateE13LayerStatePtr"],"paddle::GatedRecurrentLayer::stateWeight_":[41,7,1,"_CPPv2N6paddle19GatedRecurrentLayer12stateWeight_E"],"paddle::GatedRecurrentLayer::useBatch_":[41,7,1,"_CPPv2N6paddle19GatedRecurrentLayer9useBatch_E"],"paddle::GatedRecurrentLayer::weight_":[41,7,1,"_CPPv2N6paddle19GatedRecurrentLayer7weight_E"],"paddle::GatherAgentLayer":[41,8,1,"_CPPv2N6paddle16GatherAgentLayerE"],"paddle::GatherAgentLayer::GatherAgentLayer":[41,6,1,"_CPPv2N6paddle16GatherAgentLayer16GatherAgentLayerERK11LayerConfig"],"paddle::GatherAgentLayer::addRealLayer":[41,6,1,"_CPPv2N6paddle16GatherAgentLayer12addRealLayerE8LayerPtr"],"paddle::GatherAgentLayer::allIds_":[41,7,1,"_CPPv2N6paddle16GatherAgentLayer7allIds_E"],"paddle::GatherAgentLayer::backward":[41,6,1,"_CPPv2N6paddle16GatherAgentLayer8backwardERK14UpdateCallback"],"paddle::GatherAgentLayer::copyIdAndSequenceInfo":[41,6,1,"_CPPv2N6paddle16GatherAgentLayer21copyIdAndSequenceInfoERK8ArgumentRK10IVectorPtrRKNSt6vectorIiEE"],"paddle::GatherAgentLayer::forward":[41,6,1,"_CPPv2N6paddle16GatherAgentLayer7forwardE8PassType"],"paddle::GatherAgentLayer::idIndex_":[41,7,1,"_CPPv2N6paddle16GatherAgentLayer8idIndex_E"],"paddle::GatherAgentLayer::idsVec_":[41,7,1,"_CPPv2N6paddle16GatherAgentLayer7idsVec_E"],"paddle::GatherAgentLayer::init":[41,6,1,"_CPPv2N6paddle16GatherAgentLayer4initERK8LayerMapRK12ParameterMap"],"paddle::GatherAgentLayer::realLayers_":[41,7,1,"_CPPv2N6paddle16GatherAgentLayer11realLayers_E"],"paddle::GatherAgentLayer::~GatherAgentLayer":[41,6,1,"_CPPv2N6paddle16GatherAgentLayerD0Ev"],"paddle::GetOutputLayer":[41,8,1,"_CPPv2N6paddle14GetOutputLayerE"],"paddle::GetOutputLayer::GetOutputLayer":[41,6,1,"_CPPv2N6paddle14GetOutputLayer14GetOutputLayerERK11LayerConfig"],"paddle::GetOutputLayer::backward":[41,6,1,"_CPPv2N6paddle14GetOutputLayer8backwardERK14UpdateCallback"],"paddle::GetOutputLayer::forward":[41,6,1,"_CPPv2N6paddle14GetOutputLayer7forwardE8PassType"],"paddle::GetOutputLayer::init":[41,6,1,"_CPPv2N6paddle14GetOutputLayer4initERK8LayerMapRK12ParameterMap"],"paddle::GetOutputLayer::~GetOutputLayer":[41,6,1,"_CPPv2N6paddle14GetOutputLayerD0Ev"],"paddle::GpuAllocator":[46,8,1,"_CPPv2N6paddle12GpuAllocatorE"],"paddle::GpuAllocator::alloc":[46,6,1,"_CPPv2N6paddle12GpuAllocator5allocE6size_t"],"paddle::GpuAllocator::free":[46,6,1,"_CPPv2N6paddle12GpuAllocator4freeEPv"],"paddle::GpuAllocator::getName":[46,6,1,"_CPPv2N6paddle12GpuAllocator7getNameEv"],"paddle::GpuAllocator::~GpuAllocator":[46,6,1,"_CPPv2N6paddle12GpuAllocatorD0Ev"],"paddle::GpuIVector":[44,9,1,"_CPPv2N6paddle10GpuIVectorE"],"paddle::GpuIVectorPtr":[44,9,1,"_CPPv2N6paddle13GpuIVectorPtrE"],"paddle::GpuMatrix":[44,8,1,"_CPPv2N6paddle9GpuMatrixE"],"paddle::GpuMatrix::GpuMatrix":[44,6,1,"_CPPv2N6paddle9GpuMatrix9GpuMatrixEv"],"paddle::GpuMatrix::accumulateColSum":[44,6,1,"_CPPv2N6paddle9GpuMatrix16accumulateColSumER6Matrix"],"paddle::GpuMatrix::addBias":[44,6,1,"_CPPv2N6paddle9GpuMatrix7addBiasER6Matrix4real"],"paddle::GpuMatrix::addColumnVector":[44,6,1,"_CPPv2N6paddle9GpuMatrix15addColumnVectorERK6Matrix"],"paddle::GpuMatrix::addToRows":[44,6,1,"_CPPv2N6paddle9GpuMatrix9addToRowsER6MatrixR7IVector"],"paddle::GpuMatrix::avgPoolBackward":[44,6,1,"_CPPv2N6paddle9GpuMatrix15avgPoolBackwardER6Matrix6size_t6size_t6size_ti6size_t6size_t6size_t4real4real"],"paddle::GpuMatrix::avgPoolForward":[44,6,1,"_CPPv2N6paddle9GpuMatrix14avgPoolForwardER6Matrix6size_t6size_t6size_t6size_ti6size_t6size_t6size_t"],"paddle::GpuMatrix::check":[44,6,1,"_CPPv2N6paddle9GpuMatrix5checkERNSt7ostreamER6Matrixb"],"paddle::GpuMatrix::classificationError":[44,6,1,"_CPPv2N6paddle9GpuMatrix19classificationErrorE9MatrixPtr10IVectorPtr"],"paddle::GpuMatrix::clone":[44,6,1,"_CPPv2N6paddle9GpuMatrix5cloneE6size_t6size_tb"],"paddle::GpuMatrix::colMax":[44,6,1,"_CPPv2N6paddle9GpuMatrix6colMaxER6Matrix"],"paddle::GpuMatrix::colMerge":[44,6,1,"_CPPv2N6paddle9GpuMatrix8colMergeER6Matrix"],"paddle::GpuMatrix::collectBias":[44,6,1,"_CPPv2N6paddle9GpuMatrix11collectBiasER6Matrix4real"],"paddle::GpuMatrix::contextProjectionBackwardData":[44,6,1,"_CPPv2N6paddle9GpuMatrix29contextProjectionBackwardDataE9MatrixPtrRK7IVectorii"],"paddle::GpuMatrix::contextProjectionBackwardWeight":[44,6,1,"_CPPv2N6paddle9GpuMatrix31contextProjectionBackwardWeightE9MatrixPtrRK7IVectoriii6size_t"],"paddle::GpuMatrix::contextProjectionForward":[44,6,1,"_CPPv2N6paddle9GpuMatrix24contextProjectionForwardE9MatrixPtr9MatrixPtrRK7IVectorii6size_tb"],"paddle::GpuMatrix::convExpand":[44,6,1,"_CPPv2N6paddle9GpuMatrix10convExpandER6Matrixiiiiiiiiiii"],"paddle::GpuMatrix::convShrink":[44,6,1,"_CPPv2N6paddle9GpuMatrix10convShrinkER6Matrixiiiiiiiiiii4real4real"],"paddle::GpuMatrix::copyByRowIndex":[44,6,1,"_CPPv2N6paddle9GpuMatrix14copyByRowIndexER6MatrixR7IVector"],"paddle::GpuMatrix::copyFrom":[44,6,1,"_CPPv2N6paddle9GpuMatrix8copyFromEPK4real6size_t"],"paddle::GpuMatrix::cosSim":[44,6,1,"_CPPv2N6paddle9GpuMatrix6cosSimER6MatrixR6Matrix4real"],"paddle::GpuMatrix::cosSimDerivative":[44,6,1,"_CPPv2N6paddle9GpuMatrix16cosSimDerivativeER6MatrixR6MatrixR6MatrixR6MatrixR6Matrix4real"],"paddle::GpuMatrix::crossMapNormalBwd":[44,6,1,"_CPPv2N6paddle9GpuMatrix17crossMapNormalBwdER6MatrixR6MatrixR6MatrixR6Matrix6size_t6size_t6size_t6size_tffb"],"paddle::GpuMatrix::crossMapNormalFwd":[44,6,1,"_CPPv2N6paddle9GpuMatrix17crossMapNormalFwdER6Matrix6size_t6size_tR6Matrix6size_t6size_tffb"],"paddle::GpuMatrix::getAbsSum":[44,6,1,"_CPPv2N6paddle9GpuMatrix9getAbsSumEv"],"paddle::GpuMatrix::getElement":[44,6,1,"_CPPv2N6paddle9GpuMatrix10getElementE6size_t6size_t"],"paddle::GpuMatrix::getRow":[44,6,1,"_CPPv2N6paddle9GpuMatrix6getRowE6size_t"],"paddle::GpuMatrix::getRowBuf":[44,6,1,"_CPPv2N6paddle9GpuMatrix9getRowBufE6size_t"],"paddle::GpuMatrix::getSum":[44,6,1,"_CPPv2N6paddle9GpuMatrix6getSumEv"],"paddle::GpuMatrix::getTranspose":[44,6,1,"_CPPv2N6paddle9GpuMatrix12getTransposeEv"],"paddle::GpuMatrix::leftMul":[44,6,1,"_CPPv2N6paddle9GpuMatrix7leftMulER6Matrix4real4real"],"paddle::GpuMatrix::maxPoolBackward":[44,6,1,"_CPPv2N6paddle9GpuMatrix15maxPoolBackwardER6Matrix6size_t6size_tR6MatrixR6Matrix6size_ti6size_t6size_t6size_t4real4real"],"paddle::GpuMatrix::maxPoolForward":[44,6,1,"_CPPv2N6paddle9GpuMatrix14maxPoolForwardER6Matrix6size_t6size_t6size_t6size_ti6size_t6size_t6size_t"],"paddle::GpuMatrix::maxSequenceBackward":[44,6,1,"_CPPv2N6paddle9GpuMatrix19maxSequenceBackwardER6MatrixRK7IVectorR7IVector"],"paddle::GpuMatrix::maxSequenceForward":[44,6,1,"_CPPv2N6paddle9GpuMatrix18maxSequenceForwardER6MatrixRK7IVectorR7IVector"],"paddle::GpuMatrix::mul":[44,6,1,"_CPPv2N6paddle9GpuMatrix3mulEK9MatrixPtrK9MatrixPtr4real4real"],"paddle::GpuMatrix::oneHotCrossEntropy":[44,6,1,"_CPPv2N6paddle9GpuMatrix18oneHotCrossEntropyER6MatrixR7IVector"],"paddle::GpuMatrix::oneHotCrossEntropyBp":[44,6,1,"_CPPv2N6paddle9GpuMatrix20oneHotCrossEntropyBpER6MatrixR7IVector"],"paddle::GpuMatrix::oneHotCrossEntropyWithSelfNorm":[44,6,1,"_CPPv2N6paddle9GpuMatrix30oneHotCrossEntropyWithSelfNormER6MatrixR7IVector4real"],"paddle::GpuMatrix::oneHotCrossEntropyWithSelfNormBp":[44,6,1,"_CPPv2N6paddle9GpuMatrix32oneHotCrossEntropyWithSelfNormBpER6MatrixR7IVector4real"],"paddle::GpuMatrix::paramReluBackwardDiff":[44,6,1,"_CPPv2N6paddle9GpuMatrix21paramReluBackwardDiffER6MatrixR6MatrixR6Matrix"],"paddle::GpuMatrix::paramReluBackwardW":[44,6,1,"_CPPv2N6paddle9GpuMatrix18paramReluBackwardWER6MatrixR6Matrix"],"paddle::GpuMatrix::paramReluForward":[44,6,1,"_CPPv2N6paddle9GpuMatrix16paramReluForwardER6MatrixR6Matrix"],"paddle::GpuMatrix::print":[44,6,1,"_CPPv2N6paddle9GpuMatrix5printERNSt7ostreamE"],"paddle::GpuMatrix::randomizeUniform":[44,6,1,"_CPPv2N6paddle9GpuMatrix16randomizeUniformEv"],"paddle::GpuMatrix::resetOne":[44,6,1,"_CPPv2N6paddle9GpuMatrix8resetOneEv"],"paddle::GpuMatrix::resize":[44,6,1,"_CPPv2N6paddle9GpuMatrix6resizeE6size_t6size_t"],"paddle::GpuMatrix::rightMul":[44,6,1,"_CPPv2N6paddle9GpuMatrix8rightMulER6Matrix4real4real"],"paddle::GpuMatrix::rowMax":[44,6,1,"_CPPv2N6paddle9GpuMatrix6rowMaxER6Matrix"],"paddle::GpuMatrix::rowSum":[44,6,1,"_CPPv2N6paddle9GpuMatrix6rowSumER6Matrix"],"paddle::GpuMatrix::scaledTanh":[44,6,1,"_CPPv2N6paddle9GpuMatrix10scaledTanhER6Matrix4real4real"],"paddle::GpuMatrix::selectElements":[44,6,1,"_CPPv2N6paddle9GpuMatrix14selectElementsER6MatrixR7IVector"],"paddle::GpuMatrix::selectRows":[44,6,1,"_CPPv2N6paddle9GpuMatrix10selectRowsER6MatrixR7IVector"],"paddle::GpuMatrix::sequenceAvgForward":[44,6,1,"_CPPv2N6paddle9GpuMatrix18sequenceAvgForwardER6MatrixRK7IVectori"],"paddle::GpuMatrix::sequenceSoftmax":[44,6,1,"_CPPv2N6paddle9GpuMatrix15sequenceSoftmaxER6MatrixRK7IVector"],"paddle::GpuMatrix::setRow":[44,6,1,"_CPPv2N6paddle9GpuMatrix6setRowE6size_t6size_tPKjPK4real"],"paddle::GpuMatrix::softmax":[44,6,1,"_CPPv2N6paddle9GpuMatrix7softmaxER6Matrix"],"paddle::GpuMatrix::softmaxBackward":[44,6,1,"_CPPv2N6paddle9GpuMatrix15softmaxBackwardER6Matrix"],"paddle::GpuMatrix::softmaxDerivative":[44,6,1,"_CPPv2N6paddle9GpuMatrix17softmaxDerivativeER6MatrixR6Matrix"],"paddle::GpuMatrix::softrelu":[44,6,1,"_CPPv2N6paddle9GpuMatrix8softreluER6Matrix"],"paddle::GpuMatrix::softreluDerivative":[44,6,1,"_CPPv2N6paddle9GpuMatrix18softreluDerivativeER6Matrix"],"paddle::GpuMatrix::sumOfSquares":[44,6,1,"_CPPv2N6paddle9GpuMatrix12sumOfSquaresER6MatrixR6Matrix"],"paddle::GpuMatrix::sumOfSquaresBp":[44,6,1,"_CPPv2N6paddle9GpuMatrix14sumOfSquaresBpER6MatrixR6Matrix"],"paddle::GpuMatrix::tanh":[44,6,1,"_CPPv2N6paddle9GpuMatrix4tanhER6Matrix"],"paddle::GpuMatrix::tanhDerivative":[44,6,1,"_CPPv2N6paddle9GpuMatrix14tanhDerivativeER6Matrix"],"paddle::GpuMatrix::transpose":[44,6,1,"_CPPv2N6paddle9GpuMatrix9transposeE9MatrixPtrb"],"paddle::GpuMatrix::zeroMem":[44,6,1,"_CPPv2N6paddle9GpuMatrix7zeroMemEv"],"paddle::GpuMatrix::~GpuMatrix":[44,6,1,"_CPPv2N6paddle9GpuMatrixD0Ev"],"paddle::GpuMatrixPtr":[44,9,1,"_CPPv2N6paddle12GpuMatrixPtrE"],"paddle::GpuMemHandlePtr":[46,9,1,"_CPPv2N6paddle15GpuMemHandlePtrE"],"paddle::GpuMemoryHandle":[46,8,1,"_CPPv2N6paddle15GpuMemoryHandleE"],"paddle::GpuMemoryHandle::GpuMemoryHandle":[46,6,1,"_CPPv2N6paddle15GpuMemoryHandle15GpuMemoryHandleE6size_t"],"paddle::GpuMemoryHandle::~GpuMemoryHandle":[46,6,1,"_CPPv2N6paddle15GpuMemoryHandleD0Ev"],"paddle::GpuSparseMatrix":[44,8,1,"_CPPv2N6paddle15GpuSparseMatrixE"],"paddle::GpuSparseMatrix::Element":[44,8,1,"_CPPv2N6paddle15GpuSparseMatrix7ElementE"],"paddle::GpuSparseMatrix::Element::Element":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix7Element7ElementEii4real"],"paddle::GpuSparseMatrix::Element::col":[44,7,1,"_CPPv2N6paddle15GpuSparseMatrix7Element3colE"],"paddle::GpuSparseMatrix::Element::row":[44,7,1,"_CPPv2N6paddle15GpuSparseMatrix7Element3rowE"],"paddle::GpuSparseMatrix::Element::val":[44,7,1,"_CPPv2N6paddle15GpuSparseMatrix7Element3valE"],"paddle::GpuSparseMatrix::GpuSparseMatrix":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix15GpuSparseMatrixE6size_t6size_t6size_t15SparseValueType12SparseFormatb"],"paddle::GpuSparseMatrix::add3":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix4add3EP9GpuMatrix"],"paddle::GpuSparseMatrix::addBias":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix7addBiasER6Matrix4real"],"paddle::GpuSparseMatrix::cols_":[44,7,1,"_CPPv2N6paddle15GpuSparseMatrix5cols_E"],"paddle::GpuSparseMatrix::copyFrom":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix8copyFromERK6Matrix"],"paddle::GpuSparseMatrix::copyFromCSC":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix11copyFromCSCER15CpuSparseMatrix11hl_stream_t"],"paddle::GpuSparseMatrix::copyFromCSR":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix11copyFromCSRER15CpuSparseMatrix11hl_stream_t"],"paddle::GpuSparseMatrix::copyRow":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix7copyRowEi6size_tPK18sparse_non_value_t"],"paddle::GpuSparseMatrix::end_":[44,7,1,"_CPPv2N6paddle15GpuSparseMatrix4end_E"],"paddle::GpuSparseMatrix::format_":[44,7,1,"_CPPv2N6paddle15GpuSparseMatrix7format_E"],"paddle::GpuSparseMatrix::getColNum":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix9getColNumE6size_t"],"paddle::GpuSparseMatrix::getCols":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix7getColsEv"],"paddle::GpuSparseMatrix::getData":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix7getDataEv"],"paddle::GpuSparseMatrix::getFormat":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix9getFormatEv"],"paddle::GpuSparseMatrix::getRowCols":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix10getRowColsE6size_t"],"paddle::GpuSparseMatrix::getRowValues":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix12getRowValuesE6size_t"],"paddle::GpuSparseMatrix::getRows":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix7getRowsEv"],"paddle::GpuSparseMatrix::getTranspose":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix12getTransposeEv"],"paddle::GpuSparseMatrix::getValue":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix8getValueEv"],"paddle::GpuSparseMatrix::getValueType":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix12getValueTypeEv"],"paddle::GpuSparseMatrix::isSparse":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix8isSparseEv"],"paddle::GpuSparseMatrix::mul":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix3mulEK12GpuMatrixPtrK12GpuMatrixPtr4real4real"],"paddle::GpuSparseMatrix::print":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix5printERNSt7ostreamE"],"paddle::GpuSparseMatrix::resize":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix6resizeE6size_t6size_t6size_t15SparseValueType12SparseFormat"],"paddle::GpuSparseMatrix::resizeCSC":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix9resizeCSCE6size_t6size_t6size_t15SparseValueType"],"paddle::GpuSparseMatrix::resizeCSR":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix9resizeCSRE6size_t6size_t6size_t15SparseValueType"],"paddle::GpuSparseMatrix::rowMax":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix6rowMaxER7IVectorR6Matrix"],"paddle::GpuSparseMatrix::rows_":[44,7,1,"_CPPv2N6paddle15GpuSparseMatrix5rows_E"],"paddle::GpuSparseMatrix::sMatrix_":[44,7,1,"_CPPv2N6paddle15GpuSparseMatrix8sMatrix_E"],"paddle::GpuSparseMatrix::sMemoryHandle_":[44,7,1,"_CPPv2N6paddle15GpuSparseMatrix14sMemoryHandle_E"],"paddle::GpuSparseMatrix::setRow":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix6setRowE6size_t6size_tPKjPK4real"],"paddle::GpuSparseMatrix::sparseResize":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix12sparseResizeEv"],"paddle::GpuSparseMatrix::sparseResizeCSC":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix15sparseResizeCSCEv"],"paddle::GpuSparseMatrix::sparseResizeCSR":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix15sparseResizeCSREv"],"paddle::GpuSparseMatrix::transpose":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix9transposeE9MatrixPtrb"],"paddle::GpuSparseMatrix::trimFrom":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix8trimFromERK15CpuSparseMatrix"],"paddle::GpuSparseMatrix::trimFromCSC":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix11trimFromCSCERK15CpuSparseMatrix"],"paddle::GpuSparseMatrix::trimFromCSR":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix11trimFromCSRERK15CpuSparseMatrix"],"paddle::GpuSparseMatrix::valueType_":[44,7,1,"_CPPv2N6paddle15GpuSparseMatrix10valueType_E"],"paddle::GpuSparseMatrix::value_":[44,7,1,"_CPPv2N6paddle15GpuSparseMatrix6value_E"],"paddle::GpuSparseMatrix::zeroMem":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrix7zeroMemEv"],"paddle::GpuSparseMatrix::~GpuSparseMatrix":[44,6,1,"_CPPv2N6paddle15GpuSparseMatrixD0Ev"],"paddle::GpuSparseMatrixPtr":[44,9,1,"_CPPv2N6paddle18GpuSparseMatrixPtrE"],"paddle::GpuVector":[44,9,1,"_CPPv2N6paddle9GpuVectorE"],"paddle::GpuVectorPtr":[44,9,1,"_CPPv2N6paddle12GpuVectorPtrE"],"paddle::GpuVectorT":[44,8,1,"_CPPv2N6paddle10GpuVectorTE"],"paddle::GpuVectorT::GpuVectorT":[44,6,1,"_CPPv2N6paddle10GpuVectorT10GpuVectorTE6size_t"],"paddle::GpuVectorT::copyFrom":[44,6,1,"_CPPv2N6paddle10GpuVectorT8copyFromEPK1T6size_t"],"paddle::GpuVectorT::copyTo":[44,6,1,"_CPPv2N6paddle10GpuVectorT6copyToEP10CpuVectorTI1TE"],"paddle::GpuVectorT::fillSequence":[44,6,1,"_CPPv2N6paddle10GpuVectorT12fillSequenceEv"],"paddle::GpuVectorT::get":[44,6,1,"_CPPv2N6paddle10GpuVectorT3getE6size_t"],"paddle::GpuVectorT::getAbsMax":[44,6,1,"_CPPv2N6paddle10GpuVectorT9getAbsMaxEv"],"paddle::GpuVectorT::getAbsSum":[44,6,1,"_CPPv2N6paddle10GpuVectorT9getAbsSumEv"],"paddle::GpuVectorT::getElement":[44,6,1,"_CPPv2N6paddle10GpuVectorT10getElementE6size_t"],"paddle::GpuVectorT::getMax":[44,6,1,"_CPPv2N6paddle10GpuVectorT6getMaxEv"],"paddle::GpuVectorT::getMin":[44,6,1,"_CPPv2N6paddle10GpuVectorT6getMinEv"],"paddle::GpuVectorT::getPoint":[44,6,1,"_CPPv2N6paddle10GpuVectorT8getPointEK8uint64_t"],"paddle::GpuVectorT::getSum":[44,6,1,"_CPPv2N6paddle10GpuVectorT6getSumEv"],"paddle::GpuVectorT::histogram":[44,6,1,"_CPPv2N6paddle10GpuVectorT9histogramERNSt7ostreamEi"],"paddle::GpuVectorT::isEqualTo":[44,6,1,"_CPPv2N6paddle10GpuVectorT9isEqualToERK7VectorTI1TERK1T"],"paddle::GpuVectorT::newMemory":[44,6,1,"_CPPv2N6paddle10GpuVectorT9newMemoryE6size_t"],"paddle::GpuVectorT::print":[44,6,1,"_CPPv2N6paddle10GpuVectorT5printERNSt7ostreamE6size_t"],"paddle::GpuVectorT::printOneElement":[44,6,1,"_CPPv2N6paddle10GpuVectorT15printOneElementERNSt7ostreamE6size_t"],"paddle::GpuVectorT::rand":[44,6,1,"_CPPv2N6paddle10GpuVectorT4randEv"],"paddle::GpuVectorT::randnorm":[44,6,1,"_CPPv2N6paddle10GpuVectorT8randnormE4real4real"],"paddle::GpuVectorT::reset":[44,6,1,"_CPPv2N6paddle10GpuVectorT5resetERK1T"],"paddle::GpuVectorT::selectFrom":[44,6,1,"_CPPv2N6paddle10GpuVectorT10selectFromERK7VectorTI1TERK7VectorTIiE"],"paddle::GpuVectorT::setElement":[44,6,1,"_CPPv2N6paddle10GpuVectorT10setElementE6size_tRK1T"],"paddle::GpuVectorT::uniform":[44,6,1,"_CPPv2N6paddle10GpuVectorT7uniformE4real4real"],"paddle::GpuVectorT::zeroMem":[44,6,1,"_CPPv2N6paddle10GpuVectorT7zeroMemEv"],"paddle::GradientMachine":[38,8,1,"_CPPv2N6paddle15GradientMachineE"],"paddle::GradientMachine::CreateMode":[38,9,1,"_CPPv2N6paddle15GradientMachine10CreateModeE"],"paddle::GradientMachine::backward":[38,6,1,"_CPPv2N6paddle15GradientMachine8backwardERK14UpdateCallback"],"paddle::GradientMachine::eval":[38,6,1,"_CPPv2N6paddle15GradientMachine4evalEP9Evaluator"],"paddle::GradientMachine::finish":[38,6,1,"_CPPv2N6paddle15GradientMachine6finishEv"],"paddle::GradientMachine::forward":[38,6,1,"_CPPv2N6paddle15GradientMachine7forwardERKNSt6vectorI8ArgumentEEPNSt6vectorI8ArgumentEE8PassType"],"paddle::GradientMachine::forwardBackward":[38,6,1,"_CPPv2N6paddle15GradientMachine15forwardBackwardERKNSt6vectorI8ArgumentEEPNSt6vectorI8ArgumentEE8PassTypeRK14UpdateCallback"],"paddle::GradientMachine::getNonStaticParameters":[38,6,1,"_CPPv2N6paddle15GradientMachine22getNonStaticParametersEv"],"paddle::GradientMachine::getParameters":[38,6,1,"_CPPv2N6paddle15GradientMachine13getParametersEv"],"paddle::GradientMachine::getState":[38,6,1,"_CPPv2N6paddle15GradientMachine8getStateER12MachineState"],"paddle::GradientMachine::getStats":[38,6,1,"_CPPv2N6paddle15GradientMachine8getStatsER4realR7int64_t"],"paddle::GradientMachine::hasStaticParameters":[38,6,1,"_CPPv2N6paddle15GradientMachine19hasStaticParametersEv"],"paddle::GradientMachine::kCustom":[38,8,1,"_CPPv2N6paddle15GradientMachine7kCustomE"],"paddle::GradientMachine::kNormal":[38,8,1,"_CPPv2N6paddle15GradientMachine7kNormalE"],"paddle::GradientMachine::kSgdSparseCpuTraining":[38,8,1,"_CPPv2N6paddle15GradientMachine21kSgdSparseCpuTrainingE"],"paddle::GradientMachine::kTesting":[38,8,1,"_CPPv2N6paddle15GradientMachine8kTestingE"],"paddle::GradientMachine::makeEvaluator":[38,6,1,"_CPPv2N6paddle15GradientMachine13makeEvaluatorEv"],"paddle::GradientMachine::nonStaticParameters_":[38,7,1,"_CPPv2N6paddle15GradientMachine20nonStaticParameters_E"],"paddle::GradientMachine::onLoadParameter":[38,6,1,"_CPPv2N6paddle15GradientMachine15onLoadParameterEv"],"paddle::GradientMachine::onPassEnd":[38,6,1,"_CPPv2N6paddle15GradientMachine9onPassEndEv"],"paddle::GradientMachine::parameters_":[38,7,1,"_CPPv2N6paddle15GradientMachine11parameters_E"],"paddle::GradientMachine::prefetch":[38,6,1,"_CPPv2N6paddle15GradientMachine8prefetchERKNSt6vectorI8ArgumentEE"],"paddle::GradientMachine::resetState":[38,6,1,"_CPPv2N6paddle15GradientMachine10resetStateEv"],"paddle::GradientMachine::restart":[38,6,1,"_CPPv2N6paddle15GradientMachine7restartEv"],"paddle::GradientMachine::setOutputGrad":[38,6,1,"_CPPv2N6paddle15GradientMachine13setOutputGradERKNSt6vectorI8ArgumentEE"],"paddle::GradientMachine::setState":[38,6,1,"_CPPv2N6paddle15GradientMachine8setStateERK12MachineState"],"paddle::GradientMachine::start":[38,6,1,"_CPPv2N6paddle15GradientMachine5startERK13TrainerConfig15DataProviderPtr"],"paddle::GradientMachine::trainIsOn":[38,6,1,"_CPPv2N6paddle15GradientMachine9trainIsOnEv"],"paddle::GradientMachine::~GradientMachine":[38,6,1,"_CPPv2N6paddle15GradientMachineD0Ev"],"paddle::GradientPrinter":[36,8,1,"_CPPv2N6paddle15GradientPrinterE"],"paddle::GradientPrinter::GradientPrinter":[36,6,1,"_CPPv2N6paddle15GradientPrinter15GradientPrinterEv"],"paddle::GradientPrinter::eval":[36,6,1,"_CPPv2N6paddle15GradientPrinter4evalERK13NeuralNetwork"],"paddle::GradientPrinter::evalImp":[36,6,1,"_CPPv2N6paddle15GradientPrinter7evalImpERNSt6vectorI8ArgumentEE"],"paddle::GradientPrinter::updateSamplesNum":[36,6,1,"_CPPv2N6paddle15GradientPrinter16updateSamplesNumERKNSt6vectorI8ArgumentEE"],"paddle::GruCompute":[41,8,1,"_CPPv2N6paddle10GruComputeE"],"paddle::GruCompute::activeGate_":[41,7,1,"_CPPv2N6paddle10GruCompute11activeGate_E"],"paddle::GruCompute::activeNode_":[41,7,1,"_CPPv2N6paddle10GruCompute11activeNode_E"],"paddle::GruCompute::backward":[41,6,1,"_CPPv2N6paddle10GruCompute8backwardE12hl_gru_value11hl_gru_gradii"],"paddle::GruCompute::forward":[41,6,1,"_CPPv2N6paddle10GruCompute7forwardE12hl_gru_valueii"],"paddle::GruCompute::init":[41,6,1,"_CPPv2N6paddle10GruCompute4initER11LayerConfig"],"paddle::GruStepLayer":[41,8,1,"_CPPv2N6paddle12GruStepLayerE"],"paddle::GruStepLayer::GruStepLayer":[41,6,1,"_CPPv2N6paddle12GruStepLayer12GruStepLayerERK11LayerConfig"],"paddle::GruStepLayer::backward":[41,6,1,"_CPPv2N6paddle12GruStepLayer8backwardERK14UpdateCallback"],"paddle::GruStepLayer::bias_":[41,7,1,"_CPPv2N6paddle12GruStepLayer5bias_E"],"paddle::GruStepLayer::forward":[41,6,1,"_CPPv2N6paddle12GruStepLayer7forwardE8PassType"],"paddle::GruStepLayer::gate_":[41,7,1,"_CPPv2N6paddle12GruStepLayer5gate_E"],"paddle::GruStepLayer::init":[41,6,1,"_CPPv2N6paddle12GruStepLayer4initERK8LayerMapRK12ParameterMap"],"paddle::GruStepLayer::resetOutput_":[41,7,1,"_CPPv2N6paddle12GruStepLayer12resetOutput_E"],"paddle::GruStepLayer::weight_":[41,7,1,"_CPPv2N6paddle12GruStepLayer7weight_E"],"paddle::GruStepLayer::~GruStepLayer":[41,6,1,"_CPPv2N6paddle12GruStepLayerD0Ev"],"paddle::HISTOGRAM_EXPONENT":[44,8,1,"_CPPv2N6paddle18HISTOGRAM_EXPONENTE"],"paddle::HierarchicalSigmoidLayer":[41,8,1,"_CPPv2N6paddle24HierarchicalSigmoidLayerE"],"paddle::HierarchicalSigmoidLayer::HierarchicalSigmoidLayer":[41,6,1,"_CPPv2N6paddle24HierarchicalSigmoidLayer24HierarchicalSigmoidLayerERK11LayerConfig"],"paddle::HierarchicalSigmoidLayer::backward":[41,6,1,"_CPPv2N6paddle24HierarchicalSigmoidLayer8backwardERK14UpdateCallback"],"paddle::HierarchicalSigmoidLayer::biases_":[41,7,1,"_CPPv2N6paddle24HierarchicalSigmoidLayer7biases_E"],"paddle::HierarchicalSigmoidLayer::codeLength_":[41,7,1,"_CPPv2N6paddle24HierarchicalSigmoidLayer11codeLength_E"],"paddle::HierarchicalSigmoidLayer::forward":[41,6,1,"_CPPv2N6paddle24HierarchicalSigmoidLayer7forwardE8PassType"],"paddle::HierarchicalSigmoidLayer::getLabelLayer":[41,6,1,"_CPPv2N6paddle24HierarchicalSigmoidLayer13getLabelLayerEv"],"paddle::HierarchicalSigmoidLayer::init":[41,6,1,"_CPPv2N6paddle24HierarchicalSigmoidLayer4initERK8LayerMapRK12ParameterMap"],"paddle::HierarchicalSigmoidLayer::numClasses_":[41,7,1,"_CPPv2N6paddle24HierarchicalSigmoidLayer11numClasses_E"],"paddle::HierarchicalSigmoidLayer::preOutput_":[41,7,1,"_CPPv2N6paddle24HierarchicalSigmoidLayer10preOutput_E"],"paddle::HierarchicalSigmoidLayer::weights_":[41,7,1,"_CPPv2N6paddle24HierarchicalSigmoidLayer8weights_E"],"paddle::HistogramType":[44,9,1,"_CPPv2N6paddle13HistogramTypeE"],"paddle::HuberTwoClass":[41,8,1,"_CPPv2N6paddle13HuberTwoClassE"],"paddle::HuberTwoClass::HuberTwoClass":[41,6,1,"_CPPv2N6paddle13HuberTwoClass13HuberTwoClassERK11LayerConfig"],"paddle::HuberTwoClass::backwardImp":[41,6,1,"_CPPv2N6paddle13HuberTwoClass11backwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::HuberTwoClass::backwardImpIn":[41,6,1,"_CPPv2N6paddle13HuberTwoClass13backwardImpInER6MatrixR8ArgumentR6Matrix"],"paddle::HuberTwoClass::forwardImp":[41,6,1,"_CPPv2N6paddle13HuberTwoClass10forwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::HuberTwoClass::forwardImpIn":[41,6,1,"_CPPv2N6paddle13HuberTwoClass12forwardImpInER6MatrixR8ArgumentR6Matrix"],"paddle::HuberTwoClass::init":[41,6,1,"_CPPv2N6paddle13HuberTwoClass4initERK8LayerMapRK12ParameterMap"],"paddle::IBaseMatrix":[44,9,1,"_CPPv2N6paddle11IBaseMatrixE"],"paddle::ICpuGpuVector":[44,9,1,"_CPPv2N6paddle13ICpuGpuVectorE"],"paddle::ICpuGpuVectorPtr":[44,9,1,"_CPPv2N6paddle16ICpuGpuVectorPtrE"],"paddle::IFieldScanner":[34,8,1,"_CPPv2N6paddle13IFieldScannerE"],"paddle::IFieldScanner::DISABLE_COPY":[34,6,1,"_CPPv2N6paddle13IFieldScanner12DISABLE_COPYE13IFieldScanner"],"paddle::IFieldScanner::IFieldScanner":[34,6,1,"_CPPv2N6paddle13IFieldScanner13IFieldScannerEP10SlotHeader"],"paddle::IFieldScanner::create":[34,6,1,"_CPPv2N6paddle13IFieldScanner6createEP10SlotHeader"],"paddle::IFieldScanner::fill":[34,6,1,"_CPPv2N6paddle13IFieldScanner4fillER8ArgumentP8PyObject"],"paddle::IFieldScanner::finishFill":[34,6,1,"_CPPv2N6paddle13IFieldScanner10finishFillER8Argument"],"paddle::IFieldScanner::finishPrepare":[34,6,1,"_CPPv2N6paddle13IFieldScanner13finishPrepareER8Argument"],"paddle::IFieldScanner::headerPtr_":[34,7,1,"_CPPv2N6paddle13IFieldScanner10headerPtr_E"],"paddle::IFieldScanner::prepare":[34,6,1,"_CPPv2N6paddle13IFieldScanner7prepareER8ArgumentP8PyObject"],"paddle::IFieldScanner::startFill":[34,6,1,"_CPPv2N6paddle13IFieldScanner9startFillER8Argument"],"paddle::IFieldScanner::startPrepare":[34,6,1,"_CPPv2N6paddle13IFieldScanner12startPrepareER8Argument"],"paddle::IFieldScanner::~IFieldScanner":[34,6,1,"_CPPv2N6paddle13IFieldScannerD0Ev"],"paddle::IGradientMachineMode":[38,8,1,"_CPPv2N6paddle20IGradientMachineModeE"],"paddle::IGradientMachineMode::create":[38,6,1,"_CPPv2N6paddle20IGradientMachineMode6createERK11ModelConfig"],"paddle::IGradientMachineMode::dataMustInCpu":[38,6,1,"_CPPv2N6paddle20IGradientMachineMode13dataMustInCpuE7int32_t6size_t"],"paddle::IGradientMachineMode::isDataMustInCpu":[38,6,1,"_CPPv2N6paddle20IGradientMachineMode15isDataMustInCpuE6size_t"],"paddle::IGradientMachineMode::mode":[38,6,1,"_CPPv2N6paddle20IGradientMachineMode4modeE7int32_t"],"paddle::IGradientMachineMode::needTrainWholeDataInOneBatch":[38,6,1,"_CPPv2N6paddle20IGradientMachineMode28needTrainWholeDataInOneBatchEv"],"paddle::IGradientMachineMode::regGradientMachineMode":[38,6,1,"_CPPv2N6paddle20IGradientMachineMode22regGradientMachineModeE7int32_tRRNSt10unique_ptrI20IGradientMachineModeEE"],"paddle::IGradientMachineMode::shouldBeMe":[38,6,1,"_CPPv2N6paddle20IGradientMachineMode10shouldBeMeERKNSt6stringE6size_tbb"],"paddle::IGradientMachineMode::trainWholeDataInOneBatch":[38,6,1,"_CPPv2N6paddle20IGradientMachineMode24trainWholeDataInOneBatchE7int32_t"],"paddle::IGradientMachineMode::tryCreateGradientMachine":[38,6,1,"_CPPv2N6paddle20IGradientMachineMode24tryCreateGradientMachineE7int32_tRK11ModelConfig"],"paddle::IGradientMachineMode::tryGetMode":[38,6,1,"_CPPv2N6paddle20IGradientMachineMode10tryGetModeEPiRKNSt6stringE7int32_tbb"],"paddle::IGradientMachineMode::~IGradientMachineMode":[38,6,1,"_CPPv2N6paddle20IGradientMachineModeD0Ev"],"paddle::IParameterUpdaterHook":[52,8,1,"_CPPv2N6paddle21IParameterUpdaterHookE"],"paddle::IParameterUpdaterHook::IParameterUpdaterHook":[52,6,1,"_CPPv2N6paddle21IParameterUpdaterHook21IParameterUpdaterHookEv"],"paddle::IParameterUpdaterHook::create":[52,6,1,"_CPPv2N6paddle21IParameterUpdaterHook6createERK15ParameterConfigi"],"paddle::IParameterUpdaterHook::init":[52,6,1,"_CPPv2N6paddle21IParameterUpdaterHook4initEP9Parameter"],"paddle::IParameterUpdaterHook::update":[52,6,1,"_CPPv2N6paddle21IParameterUpdaterHook6updateEP9Parameter"],"paddle::IParameterUpdaterHook::~IParameterUpdaterHook":[52,6,1,"_CPPv2N6paddle21IParameterUpdaterHookD0Ev"],"paddle::IPyDataProviderCache":[34,8,1,"_CPPv2N6paddle20IPyDataProviderCacheE"],"paddle::IPyDataProviderCache::create":[34,6,1,"_CPPv2N6paddle20IPyDataProviderCache6createE9CacheType"],"paddle::IPyDataProviderCache::drop":[34,6,1,"_CPPv2N6paddle20IPyDataProviderCache4dropEPNSt5dequeI11PyObjectPtrEE"],"paddle::IPyDataProviderCache::load":[34,6,1,"_CPPv2N6paddle20IPyDataProviderCache4loadEv"],"paddle::IPyDataProviderCache::reset":[34,6,1,"_CPPv2N6paddle20IPyDataProviderCache5resetEv"],"paddle::IPyDataProviderCache::~IPyDataProviderCache":[34,6,1,"_CPPv2N6paddle20IPyDataProviderCacheD0Ev"],"paddle::IVector":[44,9,1,"_CPPv2N6paddle7IVectorE"],"paddle::IVectorPtr":[44,9,1,"_CPPv2N6paddle10IVectorPtrE"],"paddle::IdentityOffsetProjection":[41,8,1,"_CPPv2N6paddle24IdentityOffsetProjectionE"],"paddle::IdentityOffsetProjection::IdentityOffsetProjection":[41,6,1,"_CPPv2N6paddle24IdentityOffsetProjection24IdentityOffsetProjectionERK16ProjectionConfigRK12ParameterPtrb"],"paddle::IdentityOffsetProjection::backward":[41,6,1,"_CPPv2N6paddle24IdentityOffsetProjection8backwardERK14UpdateCallback"],"paddle::IdentityOffsetProjection::forward":[41,6,1,"_CPPv2N6paddle24IdentityOffsetProjection7forwardEv"],"paddle::IdentityProjection":[41,8,1,"_CPPv2N6paddle18IdentityProjectionE"],"paddle::IdentityProjection::IdentityProjection":[41,6,1,"_CPPv2N6paddle18IdentityProjection18IdentityProjectionERK16ProjectionConfigRK12ParameterPtrb"],"paddle::IdentityProjection::backward":[41,6,1,"_CPPv2N6paddle18IdentityProjection8backwardERK14UpdateCallback"],"paddle::IdentityProjection::forward":[41,6,1,"_CPPv2N6paddle18IdentityProjection7forwardEv"],"paddle::IndexScanner":[34,8,1,"_CPPv2N6paddle12IndexScannerE"],"paddle::IndexScanner::IndexScanner":[34,6,1,"_CPPv2N6paddle12IndexScanner12IndexScannerEP10SlotHeader"],"paddle::IndexScanner::fill":[34,6,1,"_CPPv2N6paddle12IndexScanner4fillER8ArgumentP8PyObject"],"paddle::IndexScanner::finishPrepare":[34,6,1,"_CPPv2N6paddle12IndexScanner13finishPrepareER8Argument"],"paddle::IndexScanner::prepare":[34,6,1,"_CPPv2N6paddle12IndexScanner7prepareER8ArgumentP8PyObject"],"paddle::InterpolationLayer":[41,8,1,"_CPPv2N6paddle18InterpolationLayerE"],"paddle::InterpolationLayer::InterpolationLayer":[41,6,1,"_CPPv2N6paddle18InterpolationLayer18InterpolationLayerERK11LayerConfig"],"paddle::InterpolationLayer::backward":[41,6,1,"_CPPv2N6paddle18InterpolationLayer8backwardERK14UpdateCallback"],"paddle::InterpolationLayer::forward":[41,6,1,"_CPPv2N6paddle18InterpolationLayer7forwardE8PassType"],"paddle::InterpolationLayer::init":[41,6,1,"_CPPv2N6paddle18InterpolationLayer4initERK8LayerMapRK12ParameterMap"],"paddle::InterpolationLayer::tmpMatrix":[41,7,1,"_CPPv2N6paddle18InterpolationLayer9tmpMatrixE"],"paddle::InterpolationLayer::weightLast_":[41,7,1,"_CPPv2N6paddle18InterpolationLayer11weightLast_E"],"paddle::InterpolationLayer::~InterpolationLayer":[41,6,1,"_CPPv2N6paddle18InterpolationLayerD0Ev"],"paddle::L1L2LrRegularizer":[50,8,1,"_CPPv2N6paddle17L1L2LrRegularizerE"],"paddle::L1L2LrRegularizer::update":[50,6,1,"_CPPv2N6paddle17L1L2LrRegularizer6updateEA_K9VectorPtrRK15ParameterConfig4realii"],"paddle::L1L2Regularizer":[50,8,1,"_CPPv2N6paddle15L1L2RegularizerE"],"paddle::L1L2Regularizer::update":[50,6,1,"_CPPv2N6paddle15L1L2Regularizer6updateEA_K9VectorPtrRK15ParameterConfig4realii"],"paddle::L1LrRegularizer":[50,8,1,"_CPPv2N6paddle15L1LrRegularizerE"],"paddle::L1LrRegularizer::update":[50,6,1,"_CPPv2N6paddle15L1LrRegularizer6updateEA_K9VectorPtrRK15ParameterConfig4realii"],"paddle::L1Regularizer":[50,8,1,"_CPPv2N6paddle13L1RegularizerE"],"paddle::L1Regularizer::update":[50,6,1,"_CPPv2N6paddle13L1Regularizer6updateEA_K9VectorPtrRK15ParameterConfig4realii"],"paddle::L2LrRegularizer":[50,8,1,"_CPPv2N6paddle15L2LrRegularizerE"],"paddle::L2LrRegularizer::update":[50,6,1,"_CPPv2N6paddle15L2LrRegularizer6updateEA_K9VectorPtrRK15ParameterConfig4realii"],"paddle::L2Regularizer":[50,8,1,"_CPPv2N6paddle13L2RegularizerE"],"paddle::L2Regularizer::update":[50,6,1,"_CPPv2N6paddle13L2Regularizer6updateEA_K9VectorPtrRK15ParameterConfig4realii"],"paddle::LambdaCost":[41,8,1,"_CPPv2N6paddle10LambdaCostE"],"paddle::LambdaCost::LambdaCost":[41,6,1,"_CPPv2N6paddle10LambdaCost10LambdaCostERK11LayerConfig"],"paddle::LambdaCost::backward":[41,6,1,"_CPPv2N6paddle10LambdaCost8backwardERK14UpdateCallback"],"paddle::LambdaCost::calcGrad":[41,6,1,"_CPPv2N6paddle10LambdaCost8calcGradEPK4realPK4realP4reali"],"paddle::LambdaCost::calcNDCG":[41,6,1,"_CPPv2N6paddle10LambdaCost8calcNDCGEPK4realPK4reali"],"paddle::LambdaCost::forward":[41,6,1,"_CPPv2N6paddle10LambdaCost7forwardE8PassType"],"paddle::LambdaCost::getOutputLayer":[41,6,1,"_CPPv2N6paddle10LambdaCost14getOutputLayerEv"],"paddle::LambdaCost::getScoreLayer":[41,6,1,"_CPPv2N6paddle10LambdaCost13getScoreLayerEv"],"paddle::LambdaCost::init":[41,6,1,"_CPPv2N6paddle10LambdaCost4initERK8LayerMapRK12ParameterMap"],"paddle::LambdaCost::onPassEnd":[41,6,1,"_CPPv2N6paddle10LambdaCost9onPassEndEv"],"paddle::Layer":[41,8,1,"_CPPv2N6paddle5LayerE"],"paddle::Layer::Layer":[41,6,1,"_CPPv2N6paddle5Layer5LayerERK11LayerConfigb"],"paddle::Layer::accessSubNetwork":[41,6,1,"_CPPv2N6paddle5Layer16accessSubNetworkERKNSt8functionIFvR13NeuralNetworkEEE"],"paddle::Layer::activation_":[41,7,1,"_CPPv2N6paddle5Layer11activation_E"],"paddle::Layer::addOutputArgument":[41,6,1,"_CPPv2N6paddle5Layer17addOutputArgumentEi"],"paddle::Layer::addPrev":[41,6,1,"_CPPv2N6paddle5Layer7addPrevE8LayerPtr"],"paddle::Layer::backward":[41,6,1,"_CPPv2N6paddle5Layer8backwardERK14UpdateCallback"],"paddle::Layer::backwardActivation":[41,6,1,"_CPPv2N6paddle5Layer18backwardActivationEv"],"paddle::Layer::biasParameter_":[41,7,1,"_CPPv2N6paddle5Layer14biasParameter_E"],"paddle::Layer::config_":[41,7,1,"_CPPv2N6paddle5Layer7config_E"],"paddle::Layer::copyOutputToOtherDevice":[41,6,1,"_CPPv2N6paddle5Layer23copyOutputToOtherDeviceEv"],"paddle::Layer::create":[41,6,1,"_CPPv2N6paddle5Layer6createERK11LayerConfig"],"paddle::Layer::deviceId_":[41,7,1,"_CPPv2N6paddle5Layer9deviceId_E"],"paddle::Layer::dropOutMask_":[41,7,1,"_CPPv2N6paddle5Layer12dropOutMask_E"],"paddle::Layer::forward":[41,6,1,"_CPPv2N6paddle5Layer7forwardE8PassType"],"paddle::Layer::forwardActivation":[41,6,1,"_CPPv2N6paddle5Layer17forwardActivationEv"],"paddle::Layer::forwardDropOut":[41,6,1,"_CPPv2N6paddle5Layer14forwardDropOutEv"],"paddle::Layer::getBiasParameter":[41,6,1,"_CPPv2N6paddle5Layer16getBiasParameterEv"],"paddle::Layer::getDeviceId":[41,6,1,"_CPPv2N6paddle5Layer11getDeviceIdEv"],"paddle::Layer::getInput":[41,6,1,"_CPPv2N6paddle5Layer8getInputE6size_t"],"paddle::Layer::getInputGrad":[41,6,1,"_CPPv2N6paddle5Layer12getInputGradEi"],"paddle::Layer::getInputLabel":[41,6,1,"_CPPv2N6paddle5Layer13getInputLabelERK5Layer"],"paddle::Layer::getInputValue":[41,6,1,"_CPPv2N6paddle5Layer13getInputValueEi"],"paddle::Layer::getName":[41,6,1,"_CPPv2N6paddle5Layer7getNameEv"],"paddle::Layer::getOutput":[41,6,1,"_CPPv2N6paddle5Layer9getOutputERKNSt6stringE"],"paddle::Layer::getOutputGrad":[41,6,1,"_CPPv2N6paddle5Layer13getOutputGradEv"],"paddle::Layer::getOutputLabel":[41,6,1,"_CPPv2N6paddle5Layer14getOutputLabelEv"],"paddle::Layer::getOutputValue":[41,6,1,"_CPPv2N6paddle5Layer14getOutputValueEv"],"paddle::Layer::getParameters":[41,6,1,"_CPPv2N6paddle5Layer13getParametersEv"],"paddle::Layer::getPrev":[41,6,1,"_CPPv2N6paddle5Layer7getPrevE6size_t"],"paddle::Layer::getSize":[41,6,1,"_CPPv2N6paddle5Layer7getSizeEv"],"paddle::Layer::getState":[41,6,1,"_CPPv2N6paddle5Layer8getStateEv"],"paddle::Layer::getType":[41,6,1,"_CPPv2N6paddle5Layer7getTypeEv"],"paddle::Layer::init":[41,6,1,"_CPPv2N6paddle5Layer4initERK8LayerMapRK12ParameterMap"],"paddle::Layer::initNeedFlags":[41,6,1,"_CPPv2N6paddle5Layer13initNeedFlagsEv"],"paddle::Layer::initSubNetwork":[41,6,1,"_CPPv2N6paddle5Layer14initSubNetworkEP13NeuralNetworkRK11ModelConfigRKNSt6vectorI13ParameterTypeEEb"],"paddle::Layer::inputArgument_":[41,7,1,"_CPPv2N6paddle5Layer14inputArgument_E"],"paddle::Layer::inputLayers_":[41,7,1,"_CPPv2N6paddle5Layer12inputLayers_E"],"paddle::Layer::markAllInputGrad":[41,6,1,"_CPPv2N6paddle5Layer16markAllInputGradEv"],"paddle::Layer::markInBackward_":[41,7,1,"_CPPv2N6paddle5Layer15markInBackward_E"],"paddle::Layer::markInputGrad":[41,6,1,"_CPPv2N6paddle5Layer13markInputGradEi"],"paddle::Layer::needGradient":[41,6,1,"_CPPv2N6paddle5Layer12needGradientEv"],"paddle::Layer::needGradient_":[41,7,1,"_CPPv2N6paddle5Layer13needGradient_E"],"paddle::Layer::needSequenceInfo_":[41,7,1,"_CPPv2N6paddle5Layer17needSequenceInfo_E"],"paddle::Layer::onPassEnd":[41,6,1,"_CPPv2N6paddle5Layer9onPassEndEv"],"paddle::Layer::outputMap_":[41,7,1,"_CPPv2N6paddle5Layer10outputMap_E"],"paddle::Layer::outputOtherDevice_":[41,7,1,"_CPPv2N6paddle5Layer18outputOtherDevice_E"],"paddle::Layer::output_":[41,7,1,"_CPPv2N6paddle5Layer7output_E"],"paddle::Layer::parameters_":[41,7,1,"_CPPv2N6paddle5Layer11parameters_E"],"paddle::Layer::passType_":[41,7,1,"_CPPv2N6paddle5Layer9passType_E"],"paddle::Layer::prefetch":[41,6,1,"_CPPv2N6paddle5Layer8prefetchEv"],"paddle::Layer::registrar_":[41,7,1,"_CPPv2N6paddle5Layer10registrar_E"],"paddle::Layer::reserveOutput":[41,6,1,"_CPPv2N6paddle5Layer13reserveOutputE6size_t6size_t"],"paddle::Layer::resetOutput":[41,6,1,"_CPPv2N6paddle5Layer11resetOutputE6size_t6size_t"],"paddle::Layer::resetSpecifyOutput":[41,6,1,"_CPPv2N6paddle5Layer18resetSpecifyOutputER8Argument6size_t6size_tbb"],"paddle::Layer::resetState":[41,6,1,"_CPPv2N6paddle5Layer10resetStateEv"],"paddle::Layer::resizeOutput":[41,6,1,"_CPPv2N6paddle5Layer12resizeOutputE6size_t6size_t"],"paddle::Layer::setNeedGradient":[41,6,1,"_CPPv2N6paddle5Layer15setNeedGradientEb"],"paddle::Layer::setNeedSequenceInfo":[41,6,1,"_CPPv2N6paddle5Layer19setNeedSequenceInfoEb"],"paddle::Layer::setOutput":[41,6,1,"_CPPv2N6paddle5Layer9setOutputERKNSt6stringEP8Argument"],"paddle::Layer::setState":[41,6,1,"_CPPv2N6paddle5Layer8setStateE13LayerStatePtr"],"paddle::Layer::showOutputStats":[41,6,1,"_CPPv2N6paddle5Layer15showOutputStatsEv"],"paddle::Layer::tmpGrad_":[41,7,1,"_CPPv2N6paddle5Layer8tmpGrad_E"],"paddle::Layer::useGpu_":[41,7,1,"_CPPv2N6paddle5Layer7useGpu_E"],"paddle::Layer::waitAndMergeOutputGrad":[41,6,1,"_CPPv2N6paddle5Layer22waitAndMergeOutputGradEv"],"paddle::Layer::waitInputValue":[41,6,1,"_CPPv2N6paddle5Layer14waitInputValueEv"],"paddle::Layer::zeroGrad":[41,6,1,"_CPPv2N6paddle5Layer8zeroGradEv"],"paddle::Layer::~Layer":[41,6,1,"_CPPv2N6paddle5LayerD0Ev"],"paddle::LinearChainCRF":[41,8,1,"_CPPv2N6paddle14LinearChainCRFE"],"paddle::LinearChainCRF::LinearChainCRF":[41,6,1,"_CPPv2N6paddle14LinearChainCRF14LinearChainCRFEiP4realP4real"],"paddle::LinearChainCRF::a_":[41,7,1,"_CPPv2N6paddle14LinearChainCRF2a_E"],"paddle::LinearChainCRF::alpha_":[41,7,1,"_CPPv2N6paddle14LinearChainCRF6alpha_E"],"paddle::LinearChainCRF::b_":[41,7,1,"_CPPv2N6paddle14LinearChainCRF2b_E"],"paddle::LinearChainCRF::backward":[41,6,1,"_CPPv2N6paddle14LinearChainCRF8backwardEP4realP4realPii"],"paddle::LinearChainCRF::beta_":[41,7,1,"_CPPv2N6paddle14LinearChainCRF5beta_E"],"paddle::LinearChainCRF::da_":[41,7,1,"_CPPv2N6paddle14LinearChainCRF3da_E"],"paddle::LinearChainCRF::db_":[41,7,1,"_CPPv2N6paddle14LinearChainCRF3db_E"],"paddle::LinearChainCRF::decode":[41,6,1,"_CPPv2N6paddle14LinearChainCRF6decodeEP4realPii"],"paddle::LinearChainCRF::dw_":[41,7,1,"_CPPv2N6paddle14LinearChainCRF3dw_E"],"paddle::LinearChainCRF::expW_":[41,7,1,"_CPPv2N6paddle14LinearChainCRF5expW_E"],"paddle::LinearChainCRF::expX_":[41,7,1,"_CPPv2N6paddle14LinearChainCRF5expX_E"],"paddle::LinearChainCRF::forward":[41,6,1,"_CPPv2N6paddle14LinearChainCRF7forwardEP4realPii"],"paddle::LinearChainCRF::maxX_":[41,7,1,"_CPPv2N6paddle14LinearChainCRF5maxX_E"],"paddle::LinearChainCRF::numClasses_":[41,7,1,"_CPPv2N6paddle14LinearChainCRF11numClasses_E"],"paddle::LinearChainCRF::ones_":[41,7,1,"_CPPv2N6paddle14LinearChainCRF5ones_E"],"paddle::LinearChainCRF::track_":[41,7,1,"_CPPv2N6paddle14LinearChainCRF6track_E"],"paddle::LinearChainCRF::w_":[41,7,1,"_CPPv2N6paddle14LinearChainCRF2w_E"],"paddle::LinearChainCTC":[41,8,1,"_CPPv2N6paddle14LinearChainCTCE"],"paddle::LinearChainCTC::LinearChainCTC":[41,6,1,"_CPPv2N6paddle14LinearChainCTC14LinearChainCTCEib"],"paddle::LinearChainCTC::backward":[41,6,1,"_CPPv2N6paddle14LinearChainCTC8backwardEP4realP4realPii"],"paddle::LinearChainCTC::backwardVars_":[41,7,1,"_CPPv2N6paddle14LinearChainCTC13backwardVars_E"],"paddle::LinearChainCTC::blank_":[41,7,1,"_CPPv2N6paddle14LinearChainCTC6blank_E"],"paddle::LinearChainCTC::forward":[41,6,1,"_CPPv2N6paddle14LinearChainCTC7forwardEP4realiPii"],"paddle::LinearChainCTC::forwardVars_":[41,7,1,"_CPPv2N6paddle14LinearChainCTC12forwardVars_E"],"paddle::LinearChainCTC::gradTerms_":[41,7,1,"_CPPv2N6paddle14LinearChainCTC10gradTerms_E"],"paddle::LinearChainCTC::isInvalid_":[41,7,1,"_CPPv2N6paddle14LinearChainCTC10isInvalid_E"],"paddle::LinearChainCTC::logActs_":[41,7,1,"_CPPv2N6paddle14LinearChainCTC8logActs_E"],"paddle::LinearChainCTC::logProb_":[41,7,1,"_CPPv2N6paddle14LinearChainCTC8logProb_E"],"paddle::LinearChainCTC::normByTimes_":[41,7,1,"_CPPv2N6paddle14LinearChainCTC12normByTimes_E"],"paddle::LinearChainCTC::numClasses_":[41,7,1,"_CPPv2N6paddle14LinearChainCTC11numClasses_E"],"paddle::LinearChainCTC::segmentRange":[41,6,1,"_CPPv2N6paddle14LinearChainCTC12segmentRangeERiRii"],"paddle::LinearChainCTC::totalSegments_":[41,7,1,"_CPPv2N6paddle14LinearChainCTC14totalSegments_E"],"paddle::LinearChainCTC::totalTime_":[41,7,1,"_CPPv2N6paddle14LinearChainCTC10totalTime_E"],"paddle::LockedCondition":[64,8,1,"_CPPv2N6paddle15LockedConditionE"],"paddle::LockedCondition::mutex":[64,6,1,"_CPPv2N6paddle15LockedCondition5mutexEv"],"paddle::LockedCondition::mutex_":[64,7,1,"_CPPv2N6paddle15LockedCondition6mutex_E"],"paddle::LockedCondition::notify_all":[64,6,1,"_CPPv2N6paddle15LockedCondition10notify_allE2Op"],"paddle::LockedCondition::notify_one":[64,6,1,"_CPPv2N6paddle15LockedCondition10notify_oneE2Op"],"paddle::LockedCondition::wait":[64,6,1,"_CPPv2N6paddle15LockedCondition4waitE9Predicate"],"paddle::LstmCompute":[41,8,1,"_CPPv2N6paddle11LstmComputeE"],"paddle::LstmCompute::activeGate_":[41,7,1,"_CPPv2N6paddle11LstmCompute11activeGate_E"],"paddle::LstmCompute::activeNode_":[41,7,1,"_CPPv2N6paddle11LstmCompute11activeNode_E"],"paddle::LstmCompute::activeState_":[41,7,1,"_CPPv2N6paddle11LstmCompute12activeState_E"],"paddle::LstmCompute::backwardBatch":[41,6,1,"_CPPv2N6paddle11LstmCompute13backwardBatchE13hl_lstm_value12hl_lstm_gradii"],"paddle::LstmCompute::backwardOneSequence":[41,6,1,"_CPPv2N6paddle11LstmCompute19backwardOneSequenceE13hl_lstm_value12hl_lstm_gradi"],"paddle::LstmCompute::forwardBatch":[41,6,1,"_CPPv2N6paddle11LstmCompute12forwardBatchE13hl_lstm_valueii"],"paddle::LstmCompute::forwardOneSequence":[41,6,1,"_CPPv2N6paddle11LstmCompute18forwardOneSequenceE13hl_lstm_valuei"],"paddle::LstmCompute::init":[41,6,1,"_CPPv2N6paddle11LstmCompute4initER11LayerConfig"],"paddle::LstmLayer":[41,8,1,"_CPPv2N6paddle9LstmLayerE"],"paddle::LstmLayer::LstmLayer":[41,6,1,"_CPPv2N6paddle9LstmLayer9LstmLayerERK11LayerConfig"],"paddle::LstmLayer::backward":[41,6,1,"_CPPv2N6paddle9LstmLayer8backwardERK14UpdateCallback"],"paddle::LstmLayer::backwardBatch":[41,6,1,"_CPPv2N6paddle9LstmLayer13backwardBatchEi6size_tPKi9MatrixPtr"],"paddle::LstmLayer::backwardSeqParallel":[41,6,1,"_CPPv2N6paddle9LstmLayer19backwardSeqParallelEi6size_tPKi9MatrixPtr"],"paddle::LstmLayer::backwardSequence":[41,6,1,"_CPPv2N6paddle9LstmLayer16backwardSequenceEi6size_tPKi9MatrixPtr"],"paddle::LstmLayer::batchGrad_":[41,7,1,"_CPPv2N6paddle9LstmLayer10batchGrad_E"],"paddle::LstmLayer::batchValue_":[41,7,1,"_CPPv2N6paddle9LstmLayer11batchValue_E"],"paddle::LstmLayer::bias_":[41,7,1,"_CPPv2N6paddle9LstmLayer5bias_E"],"paddle::LstmLayer::checkFgGrad_":[41,7,1,"_CPPv2N6paddle9LstmLayer12checkFgGrad_E"],"paddle::LstmLayer::checkFg_":[41,7,1,"_CPPv2N6paddle9LstmLayer8checkFg_E"],"paddle::LstmLayer::checkIgGrad_":[41,7,1,"_CPPv2N6paddle9LstmLayer12checkIgGrad_E"],"paddle::LstmLayer::checkIg_":[41,7,1,"_CPPv2N6paddle9LstmLayer8checkIg_E"],"paddle::LstmLayer::checkOgGrad_":[41,7,1,"_CPPv2N6paddle9LstmLayer12checkOgGrad_E"],"paddle::LstmLayer::checkOg_":[41,7,1,"_CPPv2N6paddle9LstmLayer8checkOg_E"],"paddle::LstmLayer::forward":[41,6,1,"_CPPv2N6paddle9LstmLayer7forwardE8PassType"],"paddle::LstmLayer::forwardBatch":[41,6,1,"_CPPv2N6paddle9LstmLayer12forwardBatchEi6size_tPKi9MatrixPtr"],"paddle::LstmLayer::forwardSeqParallel":[41,6,1,"_CPPv2N6paddle9LstmLayer18forwardSeqParallelEi6size_tPKi9MatrixPtr"],"paddle::LstmLayer::forwardSequence":[41,6,1,"_CPPv2N6paddle9LstmLayer15forwardSequenceEi6size_tPKi9MatrixPtr"],"paddle::LstmLayer::gate_":[41,7,1,"_CPPv2N6paddle9LstmLayer5gate_E"],"paddle::LstmLayer::getPrevBatchOutput":[41,6,1,"_CPPv2N6paddle9LstmLayer18getPrevBatchOutputE6size_t"],"paddle::LstmLayer::getPrevBatchState":[41,6,1,"_CPPv2N6paddle9LstmLayer17getPrevBatchStateE6size_t"],"paddle::LstmLayer::getState":[41,6,1,"_CPPv2N6paddle9LstmLayer8getStateEv"],"paddle::LstmLayer::init":[41,6,1,"_CPPv2N6paddle9LstmLayer4initERK8LayerMapRK12ParameterMap"],"paddle::LstmLayer::localBiasGrad_":[41,7,1,"_CPPv2N6paddle9LstmLayer14localBiasGrad_E"],"paddle::LstmLayer::localBias_":[41,7,1,"_CPPv2N6paddle9LstmLayer10localBias_E"],"paddle::LstmLayer::preOutput_":[41,7,1,"_CPPv2N6paddle9LstmLayer10preOutput_E"],"paddle::LstmLayer::prevBatchOutput2_":[41,7,1,"_CPPv2N6paddle9LstmLayer17prevBatchOutput2_E"],"paddle::LstmLayer::prevOutput_":[41,7,1,"_CPPv2N6paddle9LstmLayer11prevOutput_E"],"paddle::LstmLayer::prevState_":[41,7,1,"_CPPv2N6paddle9LstmLayer10prevState_E"],"paddle::LstmLayer::resetState":[41,6,1,"_CPPv2N6paddle9LstmLayer10resetStateEv"],"paddle::LstmLayer::reversed_":[41,7,1,"_CPPv2N6paddle9LstmLayer9reversed_E"],"paddle::LstmLayer::setState":[41,6,1,"_CPPv2N6paddle9LstmLayer8setStateE13LayerStatePtr"],"paddle::LstmLayer::state_":[41,7,1,"_CPPv2N6paddle9LstmLayer6state_E"],"paddle::LstmLayer::totalState_":[41,7,1,"_CPPv2N6paddle9LstmLayer11totalState_E"],"paddle::LstmLayer::useBatch_":[41,7,1,"_CPPv2N6paddle9LstmLayer9useBatch_E"],"paddle::LstmLayer::useSeqParallel_":[41,7,1,"_CPPv2N6paddle9LstmLayer15useSeqParallel_E"],"paddle::LstmLayer::weight_":[41,7,1,"_CPPv2N6paddle9LstmLayer7weight_E"],"paddle::LstmStepLayer":[41,8,1,"_CPPv2N6paddle13LstmStepLayerE"],"paddle::LstmStepLayer::LstmStepLayer":[41,6,1,"_CPPv2N6paddle13LstmStepLayer13LstmStepLayerERK11LayerConfig"],"paddle::LstmStepLayer::backward":[41,6,1,"_CPPv2N6paddle13LstmStepLayer8backwardERK14UpdateCallback"],"paddle::LstmStepLayer::checkFgGrad_":[41,7,1,"_CPPv2N6paddle13LstmStepLayer12checkFgGrad_E"],"paddle::LstmStepLayer::checkFg_":[41,7,1,"_CPPv2N6paddle13LstmStepLayer8checkFg_E"],"paddle::LstmStepLayer::checkIgGrad_":[41,7,1,"_CPPv2N6paddle13LstmStepLayer12checkIgGrad_E"],"paddle::LstmStepLayer::checkIg_":[41,7,1,"_CPPv2N6paddle13LstmStepLayer8checkIg_E"],"paddle::LstmStepLayer::checkOgGrad_":[41,7,1,"_CPPv2N6paddle13LstmStepLayer12checkOgGrad_E"],"paddle::LstmStepLayer::checkOg_":[41,7,1,"_CPPv2N6paddle13LstmStepLayer8checkOg_E"],"paddle::LstmStepLayer::forward":[41,6,1,"_CPPv2N6paddle13LstmStepLayer7forwardE8PassType"],"paddle::LstmStepLayer::gate_":[41,7,1,"_CPPv2N6paddle13LstmStepLayer5gate_E"],"paddle::LstmStepLayer::init":[41,6,1,"_CPPv2N6paddle13LstmStepLayer4initERK8LayerMapRK12ParameterMap"],"paddle::LstmStepLayer::stateActive_":[41,7,1,"_CPPv2N6paddle13LstmStepLayer12stateActive_E"],"paddle::LstmStepLayer::state_":[41,7,1,"_CPPv2N6paddle13LstmStepLayer6state_E"],"paddle::LstmStepLayer::weight_":[41,7,1,"_CPPv2N6paddle13LstmStepLayer7weight_E"],"paddle::LstmStepLayer::~LstmStepLayer":[41,6,1,"_CPPv2N6paddle13LstmStepLayerD0Ev"],"paddle::MAT_CACHE_ROW":[50,8,1,"_CPPv2N6paddle13MAT_CACHE_ROWE"],"paddle::MAT_NORMAL":[50,8,1,"_CPPv2N6paddle10MAT_NORMALE"],"paddle::MAT_NORMAL_SHARED":[50,8,1,"_CPPv2N6paddle17MAT_NORMAL_SHAREDE"],"paddle::MAT_SPARSE_ROW":[50,8,1,"_CPPv2N6paddle14MAT_SPARSE_ROWE"],"paddle::MAT_SPARSE_ROW_AUTO_GROW":[50,8,1,"_CPPv2N6paddle24MAT_SPARSE_ROW_AUTO_GROWE"],"paddle::MAT_SPARSE_ROW_IDS":[50,8,1,"_CPPv2N6paddle18MAT_SPARSE_ROW_IDSE"],"paddle::MAT_SPARSE_ROW_PREFETCH":[50,8,1,"_CPPv2N6paddle23MAT_SPARSE_ROW_PREFETCHE"],"paddle::MAT_SPARSE_ROW_PREFETCH_FULL_SIZE":[50,8,1,"_CPPv2N6paddle33MAT_SPARSE_ROW_PREFETCH_FULL_SIZEE"],"paddle::MAT_VALUE_SHARED":[50,8,1,"_CPPv2N6paddle16MAT_VALUE_SHAREDE"],"paddle::MDLstmLayer":[41,8,1,"_CPPv2N6paddle11MDLstmLayerE"],"paddle::MDLstmLayer::MDLstmLayer":[41,6,1,"_CPPv2N6paddle11MDLstmLayer11MDLstmLayerERK11LayerConfig"],"paddle::MDLstmLayer::activationGate_":[41,7,1,"_CPPv2N6paddle11MDLstmLayer15activationGate_E"],"paddle::MDLstmLayer::activationState_":[41,7,1,"_CPPv2N6paddle11MDLstmLayer16activationState_E"],"paddle::MDLstmLayer::backward":[41,6,1,"_CPPv2N6paddle11MDLstmLayer8backwardERK14UpdateCallback"],"paddle::MDLstmLayer::backwardGate2OutputSequence":[41,6,1,"_CPPv2N6paddle11MDLstmLayer27backwardGate2OutputSequenceEiR13CoordIterator"],"paddle::MDLstmLayer::backwardOneSequence":[41,6,1,"_CPPv2N6paddle11MDLstmLayer19backwardOneSequenceEiR13CoordIterator"],"paddle::MDLstmLayer::delays_":[41,7,1,"_CPPv2N6paddle11MDLstmLayer7delays_E"],"paddle::MDLstmLayer::dimsV_":[41,7,1,"_CPPv2N6paddle11MDLstmLayer6dimsV_E"],"paddle::MDLstmLayer::directions_":[41,7,1,"_CPPv2N6paddle11MDLstmLayer11directions_E"],"paddle::MDLstmLayer::forward":[41,6,1,"_CPPv2N6paddle11MDLstmLayer7forwardE8PassType"],"paddle::MDLstmLayer::forwardGate2OutputSequence":[41,6,1,"_CPPv2N6paddle11MDLstmLayer26forwardGate2OutputSequenceEiR13CoordIterator"],"paddle::MDLstmLayer::forwardOneSequence":[41,6,1,"_CPPv2N6paddle11MDLstmLayer18forwardOneSequenceEiR13CoordIterator"],"paddle::MDLstmLayer::frameForgetGate_":[41,7,1,"_CPPv2N6paddle11MDLstmLayer16frameForgetGate_E"],"paddle::MDLstmLayer::frameGate_":[41,7,1,"_CPPv2N6paddle11MDLstmLayer10frameGate_E"],"paddle::MDLstmLayer::frameInputGate_":[41,7,1,"_CPPv2N6paddle11MDLstmLayer15frameInputGate_E"],"paddle::MDLstmLayer::frameInputNode_":[41,7,1,"_CPPv2N6paddle11MDLstmLayer15frameInputNode_E"],"paddle::MDLstmLayer::frameOutputGate_":[41,7,1,"_CPPv2N6paddle11MDLstmLayer16frameOutputGate_E"],"paddle::MDLstmLayer::frameOutput_":[41,7,1,"_CPPv2N6paddle11MDLstmLayer12frameOutput_E"],"paddle::MDLstmLayer::framePreOutput_":[41,7,1,"_CPPv2N6paddle11MDLstmLayer15framePreOutput_E"],"paddle::MDLstmLayer::frameState_":[41,7,1,"_CPPv2N6paddle11MDLstmLayer11frameState_E"],"paddle::MDLstmLayer::init":[41,6,1,"_CPPv2N6paddle11MDLstmLayer4initERK8LayerMapRK12ParameterMap"],"paddle::MDLstmLayer::numBlocks_":[41,7,1,"_CPPv2N6paddle11MDLstmLayer10numBlocks_E"],"paddle::MDLstmLayer::numDims_":[41,7,1,"_CPPv2N6paddle11MDLstmLayer8numDims_E"],"paddle::MatType":[50,9,1,"_CPPv2N6paddle7MatTypeE"],"paddle::Matrix":[44,8,1,"_CPPv2N6paddle6MatrixE"],"paddle::Matrix::accumulateColSum":[44,6,1,"_CPPv2N6paddle6Matrix16accumulateColSumER6Matrix"],"paddle::Matrix::add3":[44,6,1,"_CPPv2N6paddle6Matrix4add3E9MatrixPtr"],"paddle::Matrix::addBias":[44,6,1,"_CPPv2N6paddle6Matrix7addBiasER6Matrix4real"],"paddle::Matrix::addByBitCode":[44,6,1,"_CPPv2N6paddle6Matrix12addByBitCodeE6size_tRK7IVectorRK6Matrix"],"paddle::Matrix::addByBitCodeBackward":[44,6,1,"_CPPv2N6paddle6Matrix20addByBitCodeBackwardE6size_tRK7IVectorR6Matrix"],"paddle::Matrix::addColumnVector":[44,6,1,"_CPPv2N6paddle6Matrix15addColumnVectorERK6Matrix"],"paddle::Matrix::addElements":[44,6,1,"_CPPv2N6paddle6Matrix11addElementsER6MatrixR7IVector"],"paddle::Matrix::addToRows":[44,6,1,"_CPPv2N6paddle6Matrix9addToRowsER6MatrixR7IVector"],"paddle::Matrix::avgPoolBackward":[44,6,1,"_CPPv2N6paddle6Matrix15avgPoolBackwardER6Matrix6size_t6size_t6size_ti6size_t6size_t6size_t4real4real"],"paddle::Matrix::avgPoolForward":[44,6,1,"_CPPv2N6paddle6Matrix14avgPoolForwardER6Matrix6size_t6size_t6size_t6size_ti6size_t6size_t6size_t"],"paddle::Matrix::check":[44,6,1,"_CPPv2N6paddle6Matrix5checkERNSt7ostreamER6Matrixb"],"paddle::Matrix::circularConv":[44,6,1,"_CPPv2N6paddle6Matrix12circularConvER6MatrixR6Matrix"],"paddle::Matrix::circularConvDerivative":[44,6,1,"_CPPv2N6paddle6Matrix22circularConvDerivativeER6MatrixR6MatrixR6MatrixR6MatrixR6Matrix"],"paddle::Matrix::classificationError":[44,6,1,"_CPPv2N6paddle6Matrix19classificationErrorE9MatrixPtr10IVectorPtr"],"paddle::Matrix::classificationErrorMulti":[44,6,1,"_CPPv2N6paddle6Matrix24classificationErrorMultiER6MatrixR6Matrix4real"],"paddle::Matrix::clear":[44,6,1,"_CPPv2N6paddle6Matrix5clearEv"],"paddle::Matrix::clone":[44,6,1,"_CPPv2N6paddle6Matrix5cloneE6size_t6size_tb"],"paddle::Matrix::colMax":[44,6,1,"_CPPv2N6paddle6Matrix6colMaxER6Matrix"],"paddle::Matrix::colMerge":[44,6,1,"_CPPv2N6paddle6Matrix8colMergeER6Matrix"],"paddle::Matrix::collectBias":[44,6,1,"_CPPv2N6paddle6Matrix11collectBiasER6Matrix4real"],"paddle::Matrix::contextProjectionBackward":[44,6,1,"_CPPv2N6paddle6Matrix25contextProjectionBackwardE9MatrixPtr9MatrixPtrRK7IVectorii6size_tb"],"paddle::Matrix::contextProjectionBackwardData":[44,6,1,"_CPPv2N6paddle6Matrix29contextProjectionBackwardDataE9MatrixPtrRK7IVectorii"],"paddle::Matrix::contextProjectionBackwardWeight":[44,6,1,"_CPPv2N6paddle6Matrix31contextProjectionBackwardWeightE9MatrixPtrRK7IVectoriii6size_t"],"paddle::Matrix::contextProjectionForward":[44,6,1,"_CPPv2N6paddle6Matrix24contextProjectionForwardE9MatrixPtr9MatrixPtrRK7IVectorii6size_tb"],"paddle::Matrix::convExpand":[44,6,1,"_CPPv2N6paddle6Matrix10convExpandER6Matrixiiiiiiiiiii"],"paddle::Matrix::convShrink":[44,6,1,"_CPPv2N6paddle6Matrix10convShrinkER6Matrixiiiiiiiiiii4real4real"],"paddle::Matrix::copyByRowIndex":[44,6,1,"_CPPv2N6paddle6Matrix14copyByRowIndexER6MatrixR7IVector"],"paddle::Matrix::copyFrom":[44,6,1,"_CPPv2N6paddle6Matrix8copyFromERK6Matrix"],"paddle::Matrix::cosSim":[44,6,1,"_CPPv2N6paddle6Matrix6cosSimER6MatrixR6Matrix4real"],"paddle::Matrix::cosSimDerivative":[44,6,1,"_CPPv2N6paddle6Matrix16cosSimDerivativeER6MatrixR6MatrixR6MatrixR6MatrixR6Matrix4real"],"paddle::Matrix::crossMapNormalBwd":[44,6,1,"_CPPv2N6paddle6Matrix17crossMapNormalBwdER6MatrixR6MatrixR6MatrixR6Matrix6size_t6size_t6size_t6size_tffb"],"paddle::Matrix::crossMapNormalFwd":[44,6,1,"_CPPv2N6paddle6Matrix17crossMapNormalFwdER6Matrix6size_t6size_tR6Matrix6size_t6size_tffb"],"paddle::Matrix::elementCnt_":[44,7,1,"_CPPv2N6paddle6Matrix11elementCnt_E"],"paddle::Matrix::getAbsSum":[44,6,1,"_CPPv2N6paddle6Matrix9getAbsSumEv"],"paddle::Matrix::getCols":[44,6,1,"_CPPv2N6paddle6Matrix7getColsEv"],"paddle::Matrix::getData":[44,6,1,"_CPPv2N6paddle6Matrix7getDataEv"],"paddle::Matrix::getElement":[44,6,1,"_CPPv2N6paddle6Matrix10getElementE6size_t6size_t"],"paddle::Matrix::getElementCnt":[44,6,1,"_CPPv2N6paddle6Matrix13getElementCntEv"],"paddle::Matrix::getFormat":[44,6,1,"_CPPv2N6paddle6Matrix9getFormatEv"],"paddle::Matrix::getHeight":[44,6,1,"_CPPv2N6paddle6Matrix9getHeightEv"],"paddle::Matrix::getMax":[44,6,1,"_CPPv2N6paddle6Matrix6getMaxEv"],"paddle::Matrix::getMemoryHandle":[44,6,1,"_CPPv2N6paddle6Matrix15getMemoryHandleEv"],"paddle::Matrix::getMin":[44,6,1,"_CPPv2N6paddle6Matrix6getMinEv"],"paddle::Matrix::getRowBuf":[44,6,1,"_CPPv2N6paddle6Matrix9getRowBufE6size_t"],"paddle::Matrix::getRows":[44,6,1,"_CPPv2N6paddle6Matrix7getRowsEv"],"paddle::Matrix::getStride":[44,6,1,"_CPPv2N6paddle6Matrix9getStrideEv"],"paddle::Matrix::getSum":[44,6,1,"_CPPv2N6paddle6Matrix6getSumEv"],"paddle::Matrix::getTranspose":[44,6,1,"_CPPv2N6paddle6Matrix12getTransposeEv"],"paddle::Matrix::getValueType":[44,6,1,"_CPPv2N6paddle6Matrix12getValueTypeEv"],"paddle::Matrix::getWidth":[44,6,1,"_CPPv2N6paddle6Matrix8getWidthEv"],"paddle::Matrix::isContiguous":[44,6,1,"_CPPv2N6paddle6Matrix12isContiguousEv"],"paddle::Matrix::isTransposed":[44,6,1,"_CPPv2N6paddle6Matrix12isTransposedEv"],"paddle::Matrix::leftMul":[44,6,1,"_CPPv2N6paddle6Matrix7leftMulER6Matrix4real4real"],"paddle::Matrix::maxPoolBackward":[44,6,1,"_CPPv2N6paddle6Matrix15maxPoolBackwardER6Matrix6size_t6size_tR6MatrixR6Matrix6size_ti6size_t6size_t6size_t4real4real"],"paddle::Matrix::maxPoolForward":[44,6,1,"_CPPv2N6paddle6Matrix14maxPoolForwardER6Matrix6size_t6size_t6size_t6size_ti6size_t6size_t6size_t"],"paddle::Matrix::maxSequenceBackward":[44,6,1,"_CPPv2N6paddle6Matrix19maxSequenceBackwardER6MatrixRK7IVectorR7IVector"],"paddle::Matrix::maxSequenceForward":[44,6,1,"_CPPv2N6paddle6Matrix18maxSequenceForwardER6MatrixRK7IVectorR7IVector"],"paddle::Matrix::memoryHandle_":[44,7,1,"_CPPv2N6paddle6Matrix13memoryHandle_E"],"paddle::Matrix::mul":[44,6,1,"_CPPv2N6paddle6Matrix3mulEK9MatrixPtrK9MatrixPtr4real4real"],"paddle::Matrix::mulByBitCode":[44,6,1,"_CPPv2N6paddle6Matrix12mulByBitCodeE6size_tRK7IVectorRK6MatrixRK6Matrix"],"paddle::Matrix::mulByBitCodeBackwardError":[44,6,1,"_CPPv2N6paddle6Matrix25mulByBitCodeBackwardErrorE6size_tRK7IVectorRK6MatrixR6Matrix"],"paddle::Matrix::mulByBitCodeBackwardWeight":[44,6,1,"_CPPv2N6paddle6Matrix26mulByBitCodeBackwardWeightE6size_tRK7IVectorR6MatrixRK6Matrix"],"paddle::Matrix::multiBinaryLabelCrossEntropy":[44,6,1,"_CPPv2N6paddle6Matrix28multiBinaryLabelCrossEntropyER6MatrixR6Matrix"],"paddle::Matrix::multiBinaryLabelCrossEntropyBp":[44,6,1,"_CPPv2N6paddle6Matrix30multiBinaryLabelCrossEntropyBpER6MatrixR6Matrix"],"paddle::Matrix::oneHotCrossEntropy":[44,6,1,"_CPPv2N6paddle6Matrix18oneHotCrossEntropyER6MatrixR7IVector"],"paddle::Matrix::oneHotCrossEntropyBp":[44,6,1,"_CPPv2N6paddle6Matrix20oneHotCrossEntropyBpER6MatrixR7IVector"],"paddle::Matrix::oneHotCrossEntropyWithSelfNorm":[44,6,1,"_CPPv2N6paddle6Matrix30oneHotCrossEntropyWithSelfNormER6MatrixR7IVector4real"],"paddle::Matrix::oneHotCrossEntropyWithSelfNormBp":[44,6,1,"_CPPv2N6paddle6Matrix32oneHotCrossEntropyWithSelfNormBpER6MatrixR7IVector4real"],"paddle::Matrix::paramReluBackwardDiff":[44,6,1,"_CPPv2N6paddle6Matrix21paramReluBackwardDiffER6MatrixR6MatrixR6Matrix"],"paddle::Matrix::paramReluBackwardW":[44,6,1,"_CPPv2N6paddle6Matrix18paramReluBackwardWER6MatrixR6Matrix"],"paddle::Matrix::paramReluForward":[44,6,1,"_CPPv2N6paddle6Matrix16paramReluForwardER6MatrixR6Matrix"],"paddle::Matrix::print":[44,6,1,"_CPPv2N6paddle6Matrix5printERNSt7ostreamE"],"paddle::Matrix::printOneRow":[44,6,1,"_CPPv2N6paddle6Matrix11printOneRowERNSt7ostreamE6size_t"],"paddle::Matrix::randomizeUniform":[44,6,1,"_CPPv2N6paddle6Matrix16randomizeUniformEv"],"paddle::Matrix::resetOne":[44,6,1,"_CPPv2N6paddle6Matrix8resetOneEv"],"paddle::Matrix::resize":[44,6,1,"_CPPv2N6paddle6Matrix6resizeE6size_t6size_t"],"paddle::Matrix::rightMul":[44,6,1,"_CPPv2N6paddle6Matrix8rightMulER6Matrix4real4real"],"paddle::Matrix::rowMax":[44,6,1,"_CPPv2N6paddle6Matrix6rowMaxER6Matrix"],"paddle::Matrix::rowMaxId":[44,6,1,"_CPPv2N6paddle6Matrix8rowMaxIdER7IVector"],"paddle::Matrix::rowNormalizeL1":[44,6,1,"_CPPv2N6paddle6Matrix14rowNormalizeL1ER6Matrix"],"paddle::Matrix::rowSum":[44,6,1,"_CPPv2N6paddle6Matrix6rowSumER6Matrix"],"paddle::Matrix::scaledTanh":[44,6,1,"_CPPv2N6paddle6Matrix10scaledTanhER6Matrix4real4real"],"paddle::Matrix::selectElements":[44,6,1,"_CPPv2N6paddle6Matrix14selectElementsER6MatrixR7IVector"],"paddle::Matrix::selectRows":[44,6,1,"_CPPv2N6paddle6Matrix10selectRowsER6MatrixR7IVector"],"paddle::Matrix::sequenceAvgForward":[44,6,1,"_CPPv2N6paddle6Matrix18sequenceAvgForwardER6MatrixRK7IVectori"],"paddle::Matrix::sequenceSoftmax":[44,6,1,"_CPPv2N6paddle6Matrix15sequenceSoftmaxER6MatrixRK7IVector"],"paddle::Matrix::setData":[44,6,1,"_CPPv2N6paddle6Matrix7setDataEP4real"],"paddle::Matrix::setRow":[44,6,1,"_CPPv2N6paddle6Matrix6setRowE6size_t6size_tPKjPK4real"],"paddle::Matrix::softmax":[44,6,1,"_CPPv2N6paddle6Matrix7softmaxER6Matrix"],"paddle::Matrix::softmaxBackward":[44,6,1,"_CPPv2N6paddle6Matrix15softmaxBackwardER6Matrix"],"paddle::Matrix::softmaxDerivative":[44,6,1,"_CPPv2N6paddle6Matrix17softmaxDerivativeER6MatrixR6Matrix"],"paddle::Matrix::softrelu":[44,6,1,"_CPPv2N6paddle6Matrix8softreluER6Matrix"],"paddle::Matrix::softreluDerivative":[44,6,1,"_CPPv2N6paddle6Matrix18softreluDerivativeER6Matrix"],"paddle::Matrix::subByBitCode":[44,6,1,"_CPPv2N6paddle6Matrix12subByBitCodeE6size_tR7IVector"],"paddle::Matrix::subColMatrix":[44,6,1,"_CPPv2N6paddle6Matrix12subColMatrixE6size_t6size_t"],"paddle::Matrix::subMatrix":[44,6,1,"_CPPv2N6paddle6Matrix9subMatrixE6size_t6size_t"],"paddle::Matrix::subRowMatrix":[44,6,1,"_CPPv2N6paddle6Matrix12subRowMatrixE6size_t6size_t"],"paddle::Matrix::sumByBitCode":[44,6,1,"_CPPv2N6paddle6Matrix12sumByBitCodeE6size_tR7IVectorR6Matrix4real"],"paddle::Matrix::sumOfSquares":[44,6,1,"_CPPv2N6paddle6Matrix12sumOfSquaresER6MatrixR6Matrix"],"paddle::Matrix::sumOfSquaresBp":[44,6,1,"_CPPv2N6paddle6Matrix14sumOfSquaresBpER6MatrixR6Matrix"],"paddle::Matrix::tanh":[44,6,1,"_CPPv2N6paddle6Matrix4tanhER6Matrix"],"paddle::Matrix::tanhDerivative":[44,6,1,"_CPPv2N6paddle6Matrix14tanhDerivativeER6Matrix"],"paddle::Matrix::tmpMat_":[44,7,1,"_CPPv2N6paddle6Matrix7tmpMat_E"],"paddle::Matrix::transpose":[44,6,1,"_CPPv2N6paddle6Matrix9transposeE9MatrixPtrb"],"paddle::Matrix::trimFrom":[44,6,1,"_CPPv2N6paddle6Matrix8trimFromERK15CpuSparseMatrix"],"paddle::Matrix::zeroMem":[44,6,1,"_CPPv2N6paddle6Matrix7zeroMemEv"],"paddle::Matrix::~Matrix":[44,6,1,"_CPPv2N6paddle6MatrixD0Ev"],"paddle::MatrixOffset":[44,8,1,"_CPPv2N6paddle12MatrixOffsetE"],"paddle::MatrixOffset::MatrixOffset":[44,6,1,"_CPPv2N6paddle12MatrixOffset12MatrixOffsetE6size_t6size_t6size_t6size_t6size_t6size_t6size_t6size_t"],"paddle::MatrixOffset::aCol_":[44,7,1,"_CPPv2N6paddle12MatrixOffset5aCol_E"],"paddle::MatrixOffset::aRow_":[44,7,1,"_CPPv2N6paddle12MatrixOffset5aRow_E"],"paddle::MatrixOffset::bCol_":[44,7,1,"_CPPv2N6paddle12MatrixOffset5bCol_E"],"paddle::MatrixOffset::bRow_":[44,7,1,"_CPPv2N6paddle12MatrixOffset5bRow_E"],"paddle::MatrixOffset::cCol_":[44,7,1,"_CPPv2N6paddle12MatrixOffset5cCol_E"],"paddle::MatrixOffset::cRow_":[44,7,1,"_CPPv2N6paddle12MatrixOffset5cRow_E"],"paddle::MatrixOffset::dCol_":[44,7,1,"_CPPv2N6paddle12MatrixOffset5dCol_E"],"paddle::MatrixOffset::dRow_":[44,7,1,"_CPPv2N6paddle12MatrixOffset5dRow_E"],"paddle::MatrixPtr":[44,9,1,"_CPPv2N6paddle9MatrixPtrE"],"paddle::MaxFramePrinter":[36,8,1,"_CPPv2N6paddle15MaxFramePrinterE"],"paddle::MaxFramePrinter::MaxFramePrinter":[36,6,1,"_CPPv2N6paddle15MaxFramePrinter15MaxFramePrinterEv"],"paddle::MaxFramePrinter::eval":[36,6,1,"_CPPv2N6paddle15MaxFramePrinter4evalERK13NeuralNetwork"],"paddle::MaxFramePrinter::evalImp":[36,6,1,"_CPPv2N6paddle15MaxFramePrinter7evalImpERNSt6vectorI8ArgumentEE"],"paddle::MaxFramePrinter::updateSamplesNum":[36,6,1,"_CPPv2N6paddle15MaxFramePrinter16updateSamplesNumERKNSt6vectorI8ArgumentEE"],"paddle::MaxIdLayer":[41,8,1,"_CPPv2N6paddle10MaxIdLayerE"],"paddle::MaxIdLayer::MaxIdLayer":[41,6,1,"_CPPv2N6paddle10MaxIdLayer10MaxIdLayerERK11LayerConfig"],"paddle::MaxIdLayer::backward":[41,6,1,"_CPPv2N6paddle10MaxIdLayer8backwardERK14UpdateCallback"],"paddle::MaxIdLayer::forward":[41,6,1,"_CPPv2N6paddle10MaxIdLayer7forwardE8PassType"],"paddle::MaxIdLayer::init":[41,6,1,"_CPPv2N6paddle10MaxIdLayer4initERK8LayerMapRK12ParameterMap"],"paddle::MaxIdPrinter":[36,8,1,"_CPPv2N6paddle12MaxIdPrinterE"],"paddle::MaxIdPrinter::MaxIdPrinter":[36,6,1,"_CPPv2N6paddle12MaxIdPrinter12MaxIdPrinterEv"],"paddle::MaxIdPrinter::eval":[36,6,1,"_CPPv2N6paddle12MaxIdPrinter4evalERK13NeuralNetwork"],"paddle::MaxIdPrinter::evalImp":[36,6,1,"_CPPv2N6paddle12MaxIdPrinter7evalImpERNSt6vectorI8ArgumentEE"],"paddle::MaxIdPrinter::updateSamplesNum":[36,6,1,"_CPPv2N6paddle12MaxIdPrinter16updateSamplesNumERKNSt6vectorI8ArgumentEE"],"paddle::MaxLayer":[41,8,1,"_CPPv2N6paddle8MaxLayerE"],"paddle::MaxLayer::MaxLayer":[41,6,1,"_CPPv2N6paddle8MaxLayer8MaxLayerERK11LayerConfig"],"paddle::MaxLayer::MaxLevel":[41,9,1,"_CPPv2N6paddle8MaxLayer8MaxLevelE"],"paddle::MaxLayer::backward":[41,6,1,"_CPPv2N6paddle8MaxLayer8backwardERK14UpdateCallback"],"paddle::MaxLayer::biases_":[41,7,1,"_CPPv2N6paddle8MaxLayer7biases_E"],"paddle::MaxLayer::forward":[41,6,1,"_CPPv2N6paddle8MaxLayer7forwardE8PassType"],"paddle::MaxLayer::init":[41,6,1,"_CPPv2N6paddle8MaxLayer4initERK8LayerMapRK12ParameterMap"],"paddle::MaxLayer::kNonSeq":[41,8,1,"_CPPv2N6paddle8MaxLayer7kNonSeqE"],"paddle::MaxLayer::kSeq":[41,8,1,"_CPPv2N6paddle8MaxLayer4kSeqE"],"paddle::MaxLayer::maxIndex_":[41,7,1,"_CPPv2N6paddle8MaxLayer9maxIndex_E"],"paddle::MaxLayer::type_":[41,7,1,"_CPPv2N6paddle8MaxLayer5type_E"],"paddle::MaxLayer::~MaxLayer":[41,6,1,"_CPPv2N6paddle8MaxLayerD0Ev"],"paddle::MemoryHandle":[46,8,1,"_CPPv2N6paddle12MemoryHandleE"],"paddle::MemoryHandle::MemoryHandle":[46,6,1,"_CPPv2N6paddle12MemoryHandle12MemoryHandleE6size_t"],"paddle::MemoryHandle::allocSize_":[46,7,1,"_CPPv2N6paddle12MemoryHandle10allocSize_E"],"paddle::MemoryHandle::allocator_":[46,7,1,"_CPPv2N6paddle12MemoryHandle10allocator_E"],"paddle::MemoryHandle::buf_":[46,7,1,"_CPPv2N6paddle12MemoryHandle4buf_E"],"paddle::MemoryHandle::deviceId_":[46,7,1,"_CPPv2N6paddle12MemoryHandle9deviceId_E"],"paddle::MemoryHandle::getAllocSize":[46,6,1,"_CPPv2N6paddle12MemoryHandle12getAllocSizeEv"],"paddle::MemoryHandle::getBuf":[46,6,1,"_CPPv2N6paddle12MemoryHandle6getBufEv"],"paddle::MemoryHandle::getSize":[46,6,1,"_CPPv2N6paddle12MemoryHandle7getSizeEv"],"paddle::MemoryHandle::size_":[46,7,1,"_CPPv2N6paddle12MemoryHandle5size_E"],"paddle::MemoryHandle::~MemoryHandle":[46,6,1,"_CPPv2N6paddle12MemoryHandleD0Ev"],"paddle::MemoryHandlePtr":[46,9,1,"_CPPv2N6paddle15MemoryHandlePtrE"],"paddle::MixedLayer":[41,8,1,"_CPPv2N6paddle10MixedLayerE"],"paddle::MixedLayer::MixedLayer":[41,6,1,"_CPPv2N6paddle10MixedLayer10MixedLayerERK11LayerConfig"],"paddle::MixedLayer::backward":[41,6,1,"_CPPv2N6paddle10MixedLayer8backwardERK14UpdateCallback"],"paddle::MixedLayer::biases_":[41,7,1,"_CPPv2N6paddle10MixedLayer7biases_E"],"paddle::MixedLayer::forward":[41,6,1,"_CPPv2N6paddle10MixedLayer7forwardE8PassType"],"paddle::MixedLayer::getState":[41,6,1,"_CPPv2N6paddle10MixedLayer8getStateEv"],"paddle::MixedLayer::init":[41,6,1,"_CPPv2N6paddle10MixedLayer4initERK8LayerMapRK12ParameterMap"],"paddle::MixedLayer::operators_":[41,7,1,"_CPPv2N6paddle10MixedLayer10operators_E"],"paddle::MixedLayer::prefetch":[41,6,1,"_CPPv2N6paddle10MixedLayer8prefetchEv"],"paddle::MixedLayer::projectionStateMatrixSize_":[41,7,1,"_CPPv2N6paddle10MixedLayer26projectionStateMatrixSize_E"],"paddle::MixedLayer::projections_":[41,7,1,"_CPPv2N6paddle10MixedLayer12projections_E"],"paddle::MixedLayer::resetState":[41,6,1,"_CPPv2N6paddle10MixedLayer10resetStateEv"],"paddle::MixedLayer::setState":[41,6,1,"_CPPv2N6paddle10MixedLayer8setStateE13LayerStatePtr"],"paddle::MixedLayer::~MixedLayer":[41,6,1,"_CPPv2N6paddle10MixedLayerD0Ev"],"paddle::MsgReader":[56,8,1,"_CPPv2N6paddle9MsgReaderE"],"paddle::MsgReader::MsgReader":[56,6,1,"_CPPv2N6paddle9MsgReader9MsgReaderEP13SocketChannel6size_t"],"paddle::MsgReader::blockLengths_":[56,7,1,"_CPPv2N6paddle9MsgReader13blockLengths_E"],"paddle::MsgReader::channel_":[56,7,1,"_CPPv2N6paddle9MsgReader8channel_E"],"paddle::MsgReader::currentBlockIndex_":[56,7,1,"_CPPv2N6paddle9MsgReader18currentBlockIndex_E"],"paddle::MsgReader::getBlockLength":[56,6,1,"_CPPv2N6paddle9MsgReader14getBlockLengthE6size_t"],"paddle::MsgReader::getNextBlockLength":[56,6,1,"_CPPv2N6paddle9MsgReader18getNextBlockLengthEv"],"paddle::MsgReader::getNumBlocks":[56,6,1,"_CPPv2N6paddle9MsgReader12getNumBlocksEv"],"paddle::MsgReader::getTotalLength":[56,6,1,"_CPPv2N6paddle9MsgReader14getTotalLengthEv"],"paddle::MsgReader::readBlocks":[56,6,1,"_CPPv2N6paddle9MsgReader10readBlocksERKNSt6vectorIPvEE"],"paddle::MsgReader::readNextBlock":[56,6,1,"_CPPv2N6paddle9MsgReader13readNextBlockEPv"],"paddle::MsgReader::~MsgReader":[56,6,1,"_CPPv2N6paddle9MsgReaderD0Ev"],"paddle::MultiBinaryLabelCrossEntropy":[41,8,1,"_CPPv2N6paddle28MultiBinaryLabelCrossEntropyE"],"paddle::MultiBinaryLabelCrossEntropy::MultiBinaryLabelCrossEntropy":[41,6,1,"_CPPv2N6paddle28MultiBinaryLabelCrossEntropy28MultiBinaryLabelCrossEntropyERK11LayerConfig"],"paddle::MultiBinaryLabelCrossEntropy::backwardImp":[41,6,1,"_CPPv2N6paddle28MultiBinaryLabelCrossEntropy11backwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::MultiBinaryLabelCrossEntropy::forwardImp":[41,6,1,"_CPPv2N6paddle28MultiBinaryLabelCrossEntropy10forwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::MultiBinaryLabelCrossEntropy::init":[41,6,1,"_CPPv2N6paddle28MultiBinaryLabelCrossEntropy4initERK8LayerMapRK12ParameterMap"],"paddle::MultiBinaryLabelCrossEntropy::targetPerDim_":[41,7,1,"_CPPv2N6paddle28MultiBinaryLabelCrossEntropy13targetPerDim_E"],"paddle::MultiClassCrossEntropy":[41,8,1,"_CPPv2N6paddle22MultiClassCrossEntropyE"],"paddle::MultiClassCrossEntropy::MultiClassCrossEntropy":[41,6,1,"_CPPv2N6paddle22MultiClassCrossEntropy22MultiClassCrossEntropyERK11LayerConfig"],"paddle::MultiClassCrossEntropy::backwardImp":[41,6,1,"_CPPv2N6paddle22MultiClassCrossEntropy11backwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::MultiClassCrossEntropy::forwardImp":[41,6,1,"_CPPv2N6paddle22MultiClassCrossEntropy10forwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::MultiClassCrossEntropy::init":[41,6,1,"_CPPv2N6paddle22MultiClassCrossEntropy4initERK8LayerMapRK12ParameterMap"],"paddle::MultiClassCrossEntropyWithSelfNorm":[41,8,1,"_CPPv2N6paddle34MultiClassCrossEntropyWithSelfNormE"],"paddle::MultiClassCrossEntropyWithSelfNorm::MultiClassCrossEntropyWithSelfNorm":[41,6,1,"_CPPv2N6paddle34MultiClassCrossEntropyWithSelfNorm34MultiClassCrossEntropyWithSelfNormERK11LayerConfig"],"paddle::MultiClassCrossEntropyWithSelfNorm::backwardImp":[41,6,1,"_CPPv2N6paddle34MultiClassCrossEntropyWithSelfNorm11backwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::MultiClassCrossEntropyWithSelfNorm::forwardImp":[41,6,1,"_CPPv2N6paddle34MultiClassCrossEntropyWithSelfNorm10forwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::MultiClassCrossEntropyWithSelfNorm::init":[41,6,1,"_CPPv2N6paddle34MultiClassCrossEntropyWithSelfNorm4initERK8LayerMapRK12ParameterMap"],"paddle::MultiClassCrossEntropyWithSelfNorm::sftMaxSum_":[41,7,1,"_CPPv2N6paddle34MultiClassCrossEntropyWithSelfNorm10sftMaxSum_E"],"paddle::MultiClassCrossEntropyWithSelfNorm::sumInv_":[41,7,1,"_CPPv2N6paddle34MultiClassCrossEntropyWithSelfNorm7sumInv_E"],"paddle::MultiDataProvider":[34,8,1,"_CPPv2N6paddle17MultiDataProviderE"],"paddle::MultiDataProvider::MultiDataProvider":[34,6,1,"_CPPv2N6paddle17MultiDataProvider17MultiDataProviderERK10DataConfigb"],"paddle::MultiDataProvider::getNextBatchInternal":[34,6,1,"_CPPv2N6paddle17MultiDataProvider20getNextBatchInternalE7int64_tP9DataBatch"],"paddle::MultiDataProvider::getSize":[34,6,1,"_CPPv2N6paddle17MultiDataProvider7getSizeEv"],"paddle::MultiDataProvider::isTestMode":[34,6,1,"_CPPv2N6paddle17MultiDataProvider10isTestModeEv"],"paddle::MultiDataProvider::reset":[34,6,1,"_CPPv2N6paddle17MultiDataProvider5resetEv"],"paddle::MultiDataProvider::shuffle":[34,6,1,"_CPPv2N6paddle17MultiDataProvider7shuffleEv"],"paddle::MultiDataProvider::subDataProviders_":[34,7,1,"_CPPv2N6paddle17MultiDataProvider17subDataProviders_E"],"paddle::MultiDataProvider::~MultiDataProvider":[34,6,1,"_CPPv2N6paddle17MultiDataProviderD0Ev"],"paddle::MultiGradientMachine":[38,8,1,"_CPPv2N6paddle20MultiGradientMachineE"],"paddle::MultiGradientMachine::MultiGradientMachine":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine20MultiGradientMachineERK11ModelConfigb"],"paddle::MultiGradientMachine::TASK_BACKWARD":[38,8,1,"_CPPv2N6paddle20MultiGradientMachine13TASK_BACKWARDE"],"paddle::MultiGradientMachine::TASK_COPY_IN_ARGS":[38,8,1,"_CPPv2N6paddle20MultiGradientMachine17TASK_COPY_IN_ARGSE"],"paddle::MultiGradientMachine::TASK_FORWARD":[38,8,1,"_CPPv2N6paddle20MultiGradientMachine12TASK_FORWARDE"],"paddle::MultiGradientMachine::TASK_FORWARD_BACKWARD":[38,8,1,"_CPPv2N6paddle20MultiGradientMachine21TASK_FORWARD_BACKWARDE"],"paddle::MultiGradientMachine::TaskType":[38,9,1,"_CPPv2N6paddle20MultiGradientMachine8TaskTypeE"],"paddle::MultiGradientMachine::allBarrier_":[38,7,1,"_CPPv2N6paddle20MultiGradientMachine11allBarrier_E"],"paddle::MultiGradientMachine::allocGradBufs":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine13allocGradBufsEv"],"paddle::MultiGradientMachine::backward":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine8backwardERK14UpdateCallback"],"paddle::MultiGradientMachine::backwardCallback_":[38,7,1,"_CPPv2N6paddle20MultiGradientMachine17backwardCallback_E"],"paddle::MultiGradientMachine::backwardImp":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine11backwardImpERK14UpdateCallback"],"paddle::MultiGradientMachine::bufferSizes_":[38,7,1,"_CPPv2N6paddle20MultiGradientMachine12bufferSizes_E"],"paddle::MultiGradientMachine::eval":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine4evalEP9Evaluator"],"paddle::MultiGradientMachine::finish":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine6finishEv"],"paddle::MultiGradientMachine::forward":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine7forwardERKNSt6vectorI8ArgumentEEPNSt6vectorI8ArgumentEE8PassType"],"paddle::MultiGradientMachine::forwardBackward":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine15forwardBackwardERKNSt6vectorI8ArgumentEEPNSt6vectorI8ArgumentEE8PassTypeRK14UpdateCallback"],"paddle::MultiGradientMachine::forwardImp":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine10forwardImpERKNSt6vectorI8ArgumentEEPNSt6vectorI8ArgumentEE8PassType8TaskType"],"paddle::MultiGradientMachine::getAllThreads":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine13getAllThreadsEv"],"paddle::MultiGradientMachine::getBackwardCallback":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine19getBackwardCallbackEv"],"paddle::MultiGradientMachine::getGradBuf":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine10getGradBufEi"],"paddle::MultiGradientMachine::getInArgs":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine9getInArgsEv"],"paddle::MultiGradientMachine::getNumDevices":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine13getNumDevicesEv"],"paddle::MultiGradientMachine::getNumLogicalDevices":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine20getNumLogicalDevicesEv"],"paddle::MultiGradientMachine::getNumThreads":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine13getNumThreadsEv"],"paddle::MultiGradientMachine::getOutArgs":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine10getOutArgsEPNSt6vectorI8ArgumentEE8PassType"],"paddle::MultiGradientMachine::getPassType":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine11getPassTypeEv"],"paddle::MultiGradientMachine::getSlaveParameters":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine18getSlaveParametersEv"],"paddle::MultiGradientMachine::getTaskType":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine11getTaskTypeEv"],"paddle::MultiGradientMachine::getThread":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine9getThreadEi"],"paddle::MultiGradientMachine::gradBufs_":[38,7,1,"_CPPv2N6paddle20MultiGradientMachine9gradBufs_E"],"paddle::MultiGradientMachine::gradQueue_":[38,7,1,"_CPPv2N6paddle20MultiGradientMachine10gradQueue_E"],"paddle::MultiGradientMachine::gradientMachine_":[38,7,1,"_CPPv2N6paddle20MultiGradientMachine16gradientMachine_E"],"paddle::MultiGradientMachine::hasNonstaticCpuParamters":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine24hasNonstaticCpuParamtersEv"],"paddle::MultiGradientMachine::hasNonstaticCpuParamters_":[38,7,1,"_CPPv2N6paddle20MultiGradientMachine25hasNonstaticCpuParamters_E"],"paddle::MultiGradientMachine::inArgsCopied_":[38,7,1,"_CPPv2N6paddle20MultiGradientMachine13inArgsCopied_E"],"paddle::MultiGradientMachine::inArgs_":[38,7,1,"_CPPv2N6paddle20MultiGradientMachine7inArgs_E"],"paddle::MultiGradientMachine::isPassGrad":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine10isPassGradEv"],"paddle::MultiGradientMachine::isPassGrad_":[38,7,1,"_CPPv2N6paddle20MultiGradientMachine11isPassGrad_E"],"paddle::MultiGradientMachine::logicalDeviceId2RealDeviceId":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine28logicalDeviceId2RealDeviceIdEii"],"paddle::MultiGradientMachine::makeEvaluator":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine13makeEvaluatorEv"],"paddle::MultiGradientMachine::mergeTypes_":[38,7,1,"_CPPv2N6paddle20MultiGradientMachine11mergeTypes_E"],"paddle::MultiGradientMachine::notifyGradientTransfer":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine22notifyGradientTransferEi"],"paddle::MultiGradientMachine::numDevices_":[38,7,1,"_CPPv2N6paddle20MultiGradientMachine11numDevices_E"],"paddle::MultiGradientMachine::numLogicalDevices_":[38,7,1,"_CPPv2N6paddle20MultiGradientMachine18numLogicalDevices_E"],"paddle::MultiGradientMachine::numThreads_":[38,7,1,"_CPPv2N6paddle20MultiGradientMachine11numThreads_E"],"paddle::MultiGradientMachine::onPassEnd":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine9onPassEndEv"],"paddle::MultiGradientMachine::outArgStream_":[38,7,1,"_CPPv2N6paddle20MultiGradientMachine13outArgStream_E"],"paddle::MultiGradientMachine::outArgs_":[38,7,1,"_CPPv2N6paddle20MultiGradientMachine8outArgs_E"],"paddle::MultiGradientMachine::paraMainThread":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine14paraMainThreadEi"],"paddle::MultiGradientMachine::paraMainThread_":[38,7,1,"_CPPv2N6paddle20MultiGradientMachine15paraMainThread_E"],"paddle::MultiGradientMachine::passType_":[38,7,1,"_CPPv2N6paddle20MultiGradientMachine9passType_E"],"paddle::MultiGradientMachine::prefetch":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine8prefetchERKNSt6vectorI8ArgumentEE"],"paddle::MultiGradientMachine::realDeviceId2LogicalDeviceId":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine28realDeviceId2LogicalDeviceIdEii"],"paddle::MultiGradientMachine::setOutputGrad":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine13setOutputGradERKNSt6vectorI8ArgumentEE"],"paddle::MultiGradientMachine::setPassGrad":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine11setPassGradEb"],"paddle::MultiGradientMachine::startTask":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine9startTaskE8TaskType"],"paddle::MultiGradientMachine::taskType_":[38,7,1,"_CPPv2N6paddle20MultiGradientMachine9taskType_E"],"paddle::MultiGradientMachine::threads_":[38,7,1,"_CPPv2N6paddle20MultiGradientMachine8threads_E"],"paddle::MultiGradientMachine::trainerBarrier_":[38,7,1,"_CPPv2N6paddle20MultiGradientMachine15trainerBarrier_E"],"paddle::MultiGradientMachine::updateThreadParameters":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine22updateThreadParametersEv"],"paddle::MultiGradientMachine::useGpu":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine6useGpuEv"],"paddle::MultiGradientMachine::useGpu_":[38,7,1,"_CPPv2N6paddle20MultiGradientMachine7useGpu_E"],"paddle::MultiGradientMachine::waitAfterMerge":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine14waitAfterMergeEv"],"paddle::MultiGradientMachine::waitBeforeMerge":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine15waitBeforeMergeEv"],"paddle::MultiGradientMachine::waitForCopyInArgs":[38,6,1,"_CPPv2N6paddle20MultiGradientMachine17waitForCopyInArgsEv"],"paddle::MultiThreadWorker":[62,8,1,"_CPPv2N6paddle17MultiThreadWorkerE"],"paddle::MultiThreadWorker::JobFunc":[62,9,1,"_CPPv2N6paddle17MultiThreadWorker7JobFuncE"],"paddle::MultiThreadWorker::MultiThreadWorker":[62,6,1,"_CPPv2N6paddle17MultiThreadWorker17MultiThreadWorkerE6size_t6size_t"],"paddle::MultiThreadWorker::ResultPtrType":[62,9,1,"_CPPv2N6paddle17MultiThreadWorker13ResultPtrTypeE"],"paddle::MultiThreadWorker::ResultType":[62,9,1,"_CPPv2N6paddle17MultiThreadWorker10ResultTypeE"],"paddle::MultiThreadWorker::addJob":[62,6,1,"_CPPv2N6paddle17MultiThreadWorker6addJobE7JobFunc"],"paddle::MultiThreadWorker::forceStop":[62,6,1,"_CPPv2N6paddle17MultiThreadWorker9forceStopEv"],"paddle::MultiThreadWorker::jobAdding_":[62,7,1,"_CPPv2N6paddle17MultiThreadWorker10jobAdding_E"],"paddle::MultiThreadWorker::jobs_":[62,7,1,"_CPPv2N6paddle17MultiThreadWorker5jobs_E"],"paddle::MultiThreadWorker::nullResultNum_":[62,7,1,"_CPPv2N6paddle17MultiThreadWorker14nullResultNum_E"],"paddle::MultiThreadWorker::results_":[62,7,1,"_CPPv2N6paddle17MultiThreadWorker8results_E"],"paddle::MultiThreadWorker::run":[62,6,1,"_CPPv2N6paddle17MultiThreadWorker3runEv"],"paddle::MultiThreadWorker::stop":[62,6,1,"_CPPv2N6paddle17MultiThreadWorker4stopEv"],"paddle::MultiThreadWorker::stopAddJob":[62,6,1,"_CPPv2N6paddle17MultiThreadWorker10stopAddJobEv"],"paddle::MultiThreadWorker::stopping_":[62,7,1,"_CPPv2N6paddle17MultiThreadWorker9stopping_E"],"paddle::MultiThreadWorker::testResult":[62,6,1,"_CPPv2N6paddle17MultiThreadWorker10testResultEv"],"paddle::MultiThreadWorker::waitResult":[62,6,1,"_CPPv2N6paddle17MultiThreadWorker10waitResultEv"],"paddle::MultiThreadWorker::workers_":[62,7,1,"_CPPv2N6paddle17MultiThreadWorker8workers_E"],"paddle::MultiThreadWorker::~MultiThreadWorker":[62,6,1,"_CPPv2N6paddle17MultiThreadWorkerD0Ev"],"paddle::MultinomialSampler":[41,8,1,"_CPPv2N6paddle18MultinomialSamplerE"],"paddle::MultinomialSampler::Interval":[41,8,1,"_CPPv2N6paddle18MultinomialSampler8IntervalE"],"paddle::MultinomialSampler::Interval::otherId":[41,7,1,"_CPPv2N6paddle18MultinomialSampler8Interval7otherIdE"],"paddle::MultinomialSampler::Interval::thresh":[41,7,1,"_CPPv2N6paddle18MultinomialSampler8Interval6threshE"],"paddle::MultinomialSampler::MultinomialSampler":[41,6,1,"_CPPv2N6paddle18MultinomialSampler18MultinomialSamplerEPK4reali"],"paddle::MultinomialSampler::gen":[41,6,1,"_CPPv2N6paddle18MultinomialSampler3genER4URNG"],"paddle::MultinomialSampler::gen1":[41,6,1,"_CPPv2N6paddle18MultinomialSampler4gen1E4Rand"],"paddle::MultinomialSampler::intervals_":[41,7,1,"_CPPv2N6paddle18MultinomialSampler10intervals_E"],"paddle::MultinomialSampler::rand_":[41,7,1,"_CPPv2N6paddle18MultinomialSampler5rand_E"],"paddle::MultiplexLayer":[41,8,1,"_CPPv2N6paddle14MultiplexLayerE"],"paddle::MultiplexLayer::CopyInfo":[41,8,1,"_CPPv2N6paddle14MultiplexLayer8CopyInfoE"],"paddle::MultiplexLayer::CopyInfo::CopyInfo":[41,6,1,"_CPPv2N6paddle14MultiplexLayer8CopyInfo8CopyInfoEiii"],"paddle::MultiplexLayer::CopyInfo::copyIdx":[41,7,1,"_CPPv2N6paddle14MultiplexLayer8CopyInfo7copyIdxE"],"paddle::MultiplexLayer::CopyInfo::length":[41,7,1,"_CPPv2N6paddle14MultiplexLayer8CopyInfo6lengthE"],"paddle::MultiplexLayer::CopyInfo::startIdx":[41,7,1,"_CPPv2N6paddle14MultiplexLayer8CopyInfo8startIdxE"],"paddle::MultiplexLayer::MultiplexLayer":[41,6,1,"_CPPv2N6paddle14MultiplexLayer14MultiplexLayerERK11LayerConfig"],"paddle::MultiplexLayer::backward":[41,6,1,"_CPPv2N6paddle14MultiplexLayer8backwardERK14UpdateCallback"],"paddle::MultiplexLayer::copySchedule_":[41,7,1,"_CPPv2N6paddle14MultiplexLayer13copySchedule_E"],"paddle::MultiplexLayer::forward":[41,6,1,"_CPPv2N6paddle14MultiplexLayer7forwardE8PassType"],"paddle::MultiplexLayer::init":[41,6,1,"_CPPv2N6paddle14MultiplexLayer4initERK8LayerMapRK12ParameterMap"],"paddle::MultiplexLayer::tmpDest_":[41,7,1,"_CPPv2N6paddle14MultiplexLayer8tmpDest_E"],"paddle::MultiplexLayer::tmpSrc_":[41,7,1,"_CPPv2N6paddle14MultiplexLayer7tmpSrc_E"],"paddle::MultiplexLayer::~MultiplexLayer":[41,6,1,"_CPPv2N6paddle14MultiplexLayerD0Ev"],"paddle::NCELayer":[41,8,1,"_CPPv2N6paddle8NCELayerE"],"paddle::NCELayer::NCELayer":[41,6,1,"_CPPv2N6paddle8NCELayer8NCELayerERK11LayerConfig"],"paddle::NCELayer::backward":[41,6,1,"_CPPv2N6paddle8NCELayer8backwardERK14UpdateCallback"],"paddle::NCELayer::backwardBias":[41,6,1,"_CPPv2N6paddle8NCELayer12backwardBiasERK14UpdateCallback"],"paddle::NCELayer::backwardCost":[41,6,1,"_CPPv2N6paddle8NCELayer12backwardCostEv"],"paddle::NCELayer::backwardOneInput":[41,6,1,"_CPPv2N6paddle8NCELayer16backwardOneInputEiRK14UpdateCallback"],"paddle::NCELayer::forward":[41,6,1,"_CPPv2N6paddle8NCELayer7forwardE8PassType"],"paddle::NCELayer::forwardBias":[41,6,1,"_CPPv2N6paddle8NCELayer11forwardBiasEv"],"paddle::NCELayer::forwardCost":[41,6,1,"_CPPv2N6paddle8NCELayer11forwardCostEv"],"paddle::NCELayer::forwardOneInput":[41,6,1,"_CPPv2N6paddle8NCELayer15forwardOneInputEi"],"paddle::NCELayer::init":[41,6,1,"_CPPv2N6paddle8NCELayer4initERK8LayerMapRK12ParameterMap"],"paddle::NCELayer::prefetch":[41,6,1,"_CPPv2N6paddle8NCELayer8prefetchEv"],"paddle::NCELayer::prepareSamples":[41,6,1,"_CPPv2N6paddle8NCELayer14prepareSamplesEv"],"paddle::NO_VALUE":[44,8,1,"_CPPv2N6paddle8NO_VALUEE"],"paddle::NeuralNetwork":[38,8,1,"_CPPv2N6paddle13NeuralNetworkE"],"paddle::NeuralNetwork::NeuralNetwork":[38,6,1,"_CPPv2N6paddle13NeuralNetwork13NeuralNetworkENSt6stringEP13NeuralNetwork"],"paddle::NeuralNetwork::backward":[38,6,1,"_CPPv2N6paddle13NeuralNetwork8backwardERK14UpdateCallback"],"paddle::NeuralNetwork::config_":[38,7,1,"_CPPv2N6paddle13NeuralNetwork7config_E"],"paddle::NeuralNetwork::connect":[38,6,1,"_CPPv2N6paddle13NeuralNetwork7connectENSt6stringEP13NeuralNetworkNSt6stringE"],"paddle::NeuralNetwork::create":[38,6,1,"_CPPv2N6paddle13NeuralNetwork6createERK11ModelConfig"],"paddle::NeuralNetwork::dataLayers_":[38,7,1,"_CPPv2N6paddle13NeuralNetwork11dataLayers_E"],"paddle::NeuralNetwork::dllInitMap":[38,7,1,"_CPPv2N6paddle13NeuralNetwork10dllInitMapE"],"paddle::NeuralNetwork::eval":[38,6,1,"_CPPv2N6paddle13NeuralNetwork4evalEP9Evaluator"],"paddle::NeuralNetwork::forEachLayer":[38,6,1,"_CPPv2N6paddle13NeuralNetwork12forEachLayerE1T"],"paddle::NeuralNetwork::forward":[38,6,1,"_CPPv2N6paddle13NeuralNetwork7forwardERKNSt6vectorI8ArgumentEEPNSt6vectorI8ArgumentEE8PassType"],"paddle::NeuralNetwork::getLayer":[38,6,1,"_CPPv2N6paddle13NeuralNetwork8getLayerERKNSt6stringE"],"paddle::NeuralNetwork::getLayerOutput":[38,6,1,"_CPPv2N6paddle13NeuralNetwork14getLayerOutputERKNSt6stringE"],"paddle::NeuralNetwork::getParameterMap":[38,6,1,"_CPPv2N6paddle13NeuralNetwork15getParameterMapEv"],"paddle::NeuralNetwork::getState":[38,6,1,"_CPPv2N6paddle13NeuralNetwork8getStateER12MachineState"],"paddle::NeuralNetwork::layerMap_":[38,7,1,"_CPPv2N6paddle13NeuralNetwork9layerMap_E"],"paddle::NeuralNetwork::layers_":[38,7,1,"_CPPv2N6paddle13NeuralNetwork7layers_E"],"paddle::NeuralNetwork::makeEvaluator":[38,6,1,"_CPPv2N6paddle13NeuralNetwork13makeEvaluatorEv"],"paddle::NeuralNetwork::newNeuralNetwork":[38,6,1,"_CPPv2N6paddle13NeuralNetwork16newNeuralNetworkERKNSt6stringEP13NeuralNetwork"],"paddle::NeuralNetwork::onPassEnd":[38,6,1,"_CPPv2N6paddle13NeuralNetwork9onPassEndEv"],"paddle::NeuralNetwork::outputLayers_":[38,7,1,"_CPPv2N6paddle13NeuralNetwork13outputLayers_E"],"paddle::NeuralNetwork::paramSelfInited_":[38,7,1,"_CPPv2N6paddle13NeuralNetwork16paramSelfInited_E"],"paddle::NeuralNetwork::parameterMap_":[38,7,1,"_CPPv2N6paddle13NeuralNetwork13parameterMap_E"],"paddle::NeuralNetwork::prefetch":[38,6,1,"_CPPv2N6paddle13NeuralNetwork8prefetchERKNSt6vectorI8ArgumentEE"],"paddle::NeuralNetwork::resetState":[38,6,1,"_CPPv2N6paddle13NeuralNetwork10resetStateEv"],"paddle::NeuralNetwork::rootNetwork_":[38,7,1,"_CPPv2N6paddle13NeuralNetwork12rootNetwork_E"],"paddle::NeuralNetwork::setOutputGrad":[38,6,1,"_CPPv2N6paddle13NeuralNetwork13setOutputGradERKNSt6vectorI8ArgumentEE"],"paddle::NeuralNetwork::setState":[38,6,1,"_CPPv2N6paddle13NeuralNetwork8setStateERK12MachineState"],"paddle::NeuralNetwork::subModelName_":[38,7,1,"_CPPv2N6paddle13NeuralNetwork13subModelName_E"],"paddle::NoCacheStrategy":[34,8,1,"_CPPv2N6paddle15NoCacheStrategyE"],"paddle::NoCacheStrategy::drop":[34,6,1,"_CPPv2N6paddle15NoCacheStrategy4dropEPNSt5dequeI11PyObjectPtrEE"],"paddle::NoCacheStrategy::load":[34,6,1,"_CPPv2N6paddle15NoCacheStrategy4loadEv"],"paddle::NoCacheStrategy::reset":[34,6,1,"_CPPv2N6paddle15NoCacheStrategy5resetEv"],"paddle::NormLayer":[41,8,1,"_CPPv2N6paddle9NormLayerE"],"paddle::NormLayer::NormLayer":[41,6,1,"_CPPv2N6paddle9NormLayer9NormLayerERK11LayerConfig"],"paddle::NormLayer::create":[41,6,1,"_CPPv2N6paddle9NormLayer6createERK11LayerConfig"],"paddle::NormLayer::init":[41,6,1,"_CPPv2N6paddle9NormLayer4initERK8LayerMapRK12ParameterMap"],"paddle::Operator":[41,8,1,"_CPPv2N6paddle8OperatorE"],"paddle::Operator::Operator":[41,6,1,"_CPPv2N6paddle8Operator8OperatorERK14OperatorConfigb"],"paddle::Operator::backward":[41,6,1,"_CPPv2N6paddle8Operator8backwardEv"],"paddle::Operator::config_":[41,7,1,"_CPPv2N6paddle8Operator7config_E"],"paddle::Operator::create":[41,6,1,"_CPPv2N6paddle8Operator6createERK14OperatorConfigb"],"paddle::Operator::forward":[41,6,1,"_CPPv2N6paddle8Operator7forwardENSt6vectorIPK8ArgumentEEP8Argument8PassType"],"paddle::Operator::getConfig":[41,6,1,"_CPPv2N6paddle8Operator9getConfigEv"],"paddle::Operator::getState":[41,6,1,"_CPPv2N6paddle8Operator8getStateEv"],"paddle::Operator::ins_":[41,7,1,"_CPPv2N6paddle8Operator4ins_E"],"paddle::Operator::out_":[41,7,1,"_CPPv2N6paddle8Operator4out_E"],"paddle::Operator::passType_":[41,7,1,"_CPPv2N6paddle8Operator9passType_E"],"paddle::Operator::prefetch":[41,6,1,"_CPPv2N6paddle8Operator8prefetchEPK8Argument"],"paddle::Operator::registrar_":[41,7,1,"_CPPv2N6paddle8Operator10registrar_E"],"paddle::Operator::resetState":[41,6,1,"_CPPv2N6paddle8Operator10resetStateEv"],"paddle::Operator::setState":[41,6,1,"_CPPv2N6paddle8Operator8setStateE13LayerStatePtr"],"paddle::Operator::useGpu_":[41,7,1,"_CPPv2N6paddle8Operator7useGpu_E"],"paddle::Operator::~Operator":[41,6,1,"_CPPv2N6paddle8OperatorD0Ev"],"paddle::OptimizerWithGradientClipping":[48,8,1,"_CPPv2N6paddle29OptimizerWithGradientClippingE"],"paddle::OptimizerWithGradientClipping::OptimizerWithGradientClipping":[48,6,1,"_CPPv2N6paddle29OptimizerWithGradientClipping29OptimizerWithGradientClippingERK18OptimizationConfigP18ParameterOptimizer"],"paddle::OptimizerWithGradientClipping::finishBatch":[48,6,1,"_CPPv2N6paddle29OptimizerWithGradientClipping11finishBatchEv"],"paddle::OptimizerWithGradientClipping::finishPass":[48,6,1,"_CPPv2N6paddle29OptimizerWithGradientClipping10finishPassEv"],"paddle::OptimizerWithGradientClipping::init":[48,6,1,"_CPPv2N6paddle29OptimizerWithGradientClipping4initE6size_tPK15ParameterConfig"],"paddle::OptimizerWithGradientClipping::needSpecialTraversal":[48,6,1,"_CPPv2N6paddle29OptimizerWithGradientClipping20needSpecialTraversalERK15ParameterConfig"],"paddle::OptimizerWithGradientClipping::optimizer_":[48,7,1,"_CPPv2N6paddle29OptimizerWithGradientClipping10optimizer_E"],"paddle::OptimizerWithGradientClipping::setNoDecay":[48,6,1,"_CPPv2N6paddle29OptimizerWithGradientClipping10setNoDecayEv"],"paddle::OptimizerWithGradientClipping::startBatch":[48,6,1,"_CPPv2N6paddle29OptimizerWithGradientClipping10startBatchE7int64_t"],"paddle::OptimizerWithGradientClipping::startPass":[48,6,1,"_CPPv2N6paddle29OptimizerWithGradientClipping9startPassEv"],"paddle::OptimizerWithGradientClipping::update":[48,6,1,"_CPPv2N6paddle29OptimizerWithGradientClipping6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::OptimizerWithRegularizer":[48,8,1,"_CPPv2N6paddle24OptimizerWithRegularizerE"],"paddle::OptimizerWithRegularizer::OptimizerWithRegularizer":[48,6,1,"_CPPv2N6paddle24OptimizerWithRegularizer24OptimizerWithRegularizerERK18OptimizationConfigP18ParameterOptimizerP11Regularizer"],"paddle::OptimizerWithRegularizer::create":[48,6,1,"_CPPv2N6paddle24OptimizerWithRegularizer6createERK18OptimizationConfigRK15ParameterConfigbb"],"paddle::OptimizerWithRegularizer::finishBatch":[48,6,1,"_CPPv2N6paddle24OptimizerWithRegularizer11finishBatchEv"],"paddle::OptimizerWithRegularizer::finishPass":[48,6,1,"_CPPv2N6paddle24OptimizerWithRegularizer10finishPassEv"],"paddle::OptimizerWithRegularizer::init":[48,6,1,"_CPPv2N6paddle24OptimizerWithRegularizer4initE6size_tPK15ParameterConfig"],"paddle::OptimizerWithRegularizer::needSpecialTraversal":[48,6,1,"_CPPv2N6paddle24OptimizerWithRegularizer20needSpecialTraversalERK15ParameterConfig"],"paddle::OptimizerWithRegularizer::optimizer_":[48,7,1,"_CPPv2N6paddle24OptimizerWithRegularizer10optimizer_E"],"paddle::OptimizerWithRegularizer::regularizer_":[48,7,1,"_CPPv2N6paddle24OptimizerWithRegularizer12regularizer_E"],"paddle::OptimizerWithRegularizer::startBatch":[48,6,1,"_CPPv2N6paddle24OptimizerWithRegularizer10startBatchE7int64_t"],"paddle::OptimizerWithRegularizer::startPass":[48,6,1,"_CPPv2N6paddle24OptimizerWithRegularizer9startPassEv"],"paddle::OptimizerWithRegularizer::timer_":[48,7,1,"_CPPv2N6paddle24OptimizerWithRegularizer6timer_E"],"paddle::OptimizerWithRegularizer::update":[48,6,1,"_CPPv2N6paddle24OptimizerWithRegularizer6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::OptimizerWithRegularizerEveryNumBatches":[48,8,1,"_CPPv2N6paddle39OptimizerWithRegularizerEveryNumBatchesE"],"paddle::OptimizerWithRegularizerEveryNumBatches::OptimizerWithRegularizerEveryNumBatches":[48,6,1,"_CPPv2N6paddle39OptimizerWithRegularizerEveryNumBatches39OptimizerWithRegularizerEveryNumBatchesERK18OptimizationConfigP18ParameterOptimizerP11Regularizer"],"paddle::OptimizerWithRegularizerEveryNumBatches::baseTimer_":[48,7,1,"_CPPv2N6paddle39OptimizerWithRegularizerEveryNumBatches10baseTimer_E"],"paddle::OptimizerWithRegularizerEveryNumBatches::catchUpWith":[48,6,1,"_CPPv2N6paddle39OptimizerWithRegularizerEveryNumBatches11catchUpWithEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::OptimizerWithRegularizerEveryNumBatches::doTraversal":[48,6,1,"_CPPv2N6paddle39OptimizerWithRegularizerEveryNumBatches11doTraversalEA_K9VectorPtrRK15ParameterConfig"],"paddle::OptimizerWithRegularizerEveryNumBatches::finishCatchUpWith":[48,6,1,"_CPPv2N6paddle39OptimizerWithRegularizerEveryNumBatches17finishCatchUpWithEv"],"paddle::OptimizerWithRegularizerEveryNumBatches::isRegularizationBatch":[48,6,1,"_CPPv2N6paddle39OptimizerWithRegularizerEveryNumBatches21isRegularizationBatchERK15ParameterConfig"],"paddle::OptimizerWithRegularizerEveryNumBatches::needSpecialTraversal":[48,6,1,"_CPPv2N6paddle39OptimizerWithRegularizerEveryNumBatches20needSpecialTraversalERK15ParameterConfig"],"paddle::OptimizerWithRegularizerEveryNumBatches::startCatchUpWith":[48,6,1,"_CPPv2N6paddle39OptimizerWithRegularizerEveryNumBatches16startCatchUpWithEv"],"paddle::OptimizerWithRegularizerEveryNumBatches::startPass":[48,6,1,"_CPPv2N6paddle39OptimizerWithRegularizerEveryNumBatches9startPassEv"],"paddle::OptimizerWithRegularizerEveryNumBatches::update":[48,6,1,"_CPPv2N6paddle39OptimizerWithRegularizerEveryNumBatches6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::OptimizerWithRegularizerSparse":[48,8,1,"_CPPv2N6paddle30OptimizerWithRegularizerSparseE"],"paddle::OptimizerWithRegularizerSparse::OptimizerWithRegularizerSparse":[48,6,1,"_CPPv2N6paddle30OptimizerWithRegularizerSparse30OptimizerWithRegularizerSparseERK18OptimizationConfigP18ParameterOptimizerP11Regularizer"],"paddle::OptimizerWithRegularizerSparse::catchUpWith":[48,6,1,"_CPPv2N6paddle30OptimizerWithRegularizerSparse11catchUpWithEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::OptimizerWithRegularizerSparse::finishCatchUpWith":[48,6,1,"_CPPv2N6paddle30OptimizerWithRegularizerSparse17finishCatchUpWithEv"],"paddle::OptimizerWithRegularizerSparse::init":[48,6,1,"_CPPv2N6paddle30OptimizerWithRegularizerSparse4initE6size_tPK15ParameterConfig"],"paddle::OptimizerWithRegularizerSparse::startCatchUpWith":[48,6,1,"_CPPv2N6paddle30OptimizerWithRegularizerSparse16startCatchUpWithEv"],"paddle::OptimizerWithRegularizerSparse::t0Vec_":[48,7,1,"_CPPv2N6paddle30OptimizerWithRegularizerSparse6t0Vec_E"],"paddle::OptimizerWithRegularizerSparse::update":[48,6,1,"_CPPv2N6paddle30OptimizerWithRegularizerSparse6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::OuterProdLayer":[41,8,1,"_CPPv2N6paddle14OuterProdLayerE"],"paddle::OuterProdLayer::OuterProdLayer":[41,6,1,"_CPPv2N6paddle14OuterProdLayer14OuterProdLayerERK11LayerConfig"],"paddle::OuterProdLayer::backward":[41,6,1,"_CPPv2N6paddle14OuterProdLayer8backwardERK14UpdateCallback"],"paddle::OuterProdLayer::forward":[41,6,1,"_CPPv2N6paddle14OuterProdLayer7forwardE8PassType"],"paddle::OuterProdLayer::init":[41,6,1,"_CPPv2N6paddle14OuterProdLayer4initERK8LayerMapRK12ParameterMap"],"paddle::OuterProdLayer::tmpMtx0":[41,7,1,"_CPPv2N6paddle14OuterProdLayer7tmpMtx0E"],"paddle::OuterProdLayer::tmpRow0":[41,7,1,"_CPPv2N6paddle14OuterProdLayer7tmpRow0E"],"paddle::OuterProdLayer::tmpRow1":[41,7,1,"_CPPv2N6paddle14OuterProdLayer7tmpRow1E"],"paddle::OuterProdLayer::~OuterProdLayer":[41,6,1,"_CPPv2N6paddle14OuterProdLayerD0Ev"],"paddle::ParallelCpuVectorT":[44,8,1,"_CPPv2N6paddle18ParallelCpuVectorTE"],"paddle::ParallelCpuVectorT::ExecFunc":[44,9,1,"_CPPv2N6paddle18ParallelCpuVectorT8ExecFuncE"],"paddle::ParallelCpuVectorT::ParallelCpuVectorT":[44,6,1,"_CPPv2N6paddle18ParallelCpuVectorT18ParallelCpuVectorTE6size_tP14SyncThreadPool"],"paddle::ParallelCpuVectorT::exec":[44,6,1,"_CPPv2N6paddle18ParallelCpuVectorT4execEN14SyncThreadPool7JobFuncE"],"paddle::ParallelCpuVectorT::parallelExec":[44,6,1,"_CPPv2N6paddle18ParallelCpuVectorT12parallelExecE8ExecFunc"],"paddle::ParallelCpuVectorT::pool_":[44,7,1,"_CPPv2N6paddle18ParallelCpuVectorT5pool_E"],"paddle::ParallelCpuVectorT::randnorm":[44,6,1,"_CPPv2N6paddle18ParallelCpuVectorT8randnormE4real4real"],"paddle::ParallelCpuVectorT::uniform":[44,6,1,"_CPPv2N6paddle18ParallelCpuVectorT7uniformE4real4real"],"paddle::ParallelCpuVectorT::zeroMem":[44,6,1,"_CPPv2N6paddle18ParallelCpuVectorT7zeroMemEv"],"paddle::ParallelNeuralNetwork":[38,8,1,"_CPPv2N6paddle21ParallelNeuralNetworkE"],"paddle::ParallelNeuralNetwork::ParallelNeuralNetwork":[38,6,1,"_CPPv2N6paddle21ParallelNeuralNetwork21ParallelNeuralNetworkENSt6stringEP13NeuralNetwork"],"paddle::ParallelNeuralNetwork::addComputeThread":[38,6,1,"_CPPv2N6paddle21ParallelNeuralNetwork16addComputeThreadEi"],"paddle::ParallelNeuralNetwork::backward":[38,6,1,"_CPPv2N6paddle21ParallelNeuralNetwork8backwardERK14UpdateCallback"],"paddle::ParallelNeuralNetwork::dispatchByDeviceId":[38,6,1,"_CPPv2N6paddle21ParallelNeuralNetwork18dispatchByDeviceIdEi8LayerPtr8TaskType"],"paddle::ParallelNeuralNetwork::forward":[38,6,1,"_CPPv2N6paddle21ParallelNeuralNetwork7forwardERKNSt6vectorI8ArgumentEEPNSt6vectorI8ArgumentEE8PassType"],"paddle::ParallelNeuralNetwork::forwardBackward":[38,6,1,"_CPPv2N6paddle21ParallelNeuralNetwork15forwardBackwardERKNSt6vectorI8ArgumentEEPNSt6vectorI8ArgumentEE8PassTypeRK14UpdateCallback"],"paddle::ParallelNeuralNetwork::numDevices_":[38,7,1,"_CPPv2N6paddle21ParallelNeuralNetwork11numDevices_E"],"paddle::ParallelNeuralNetwork::start":[38,6,1,"_CPPv2N6paddle21ParallelNeuralNetwork5startERK13TrainerConfig15DataProviderPtr"],"paddle::ParallelNeuralNetwork::threads_":[38,7,1,"_CPPv2N6paddle21ParallelNeuralNetwork8threads_E"],"paddle::ParallelNeuralNetwork::useGpu_":[38,7,1,"_CPPv2N6paddle21ParallelNeuralNetwork7useGpu_E"],"paddle::ParallelNeuralNetwork::waitAllThread":[38,6,1,"_CPPv2N6paddle21ParallelNeuralNetwork13waitAllThreadEv"],"paddle::ParallelParameter":[50,8,1,"_CPPv2N6paddle17ParallelParameterE"],"paddle::ParallelParameter::ParallelParameter":[50,6,1,"_CPPv2N6paddle17ParallelParameter17ParallelParameterE11TrainerRole12ParameterPtr"],"paddle::ParallelParameter::controlUpdate":[50,6,1,"_CPPv2N6paddle17ParallelParameter13controlUpdateERK14UpdateCallback"],"paddle::ParallelParameter::create":[50,6,1,"_CPPv2N6paddle17ParallelParameter6createE11TrainerRole12ParameterPtri"],"paddle::ParallelParameter::getLocalParameter":[50,6,1,"_CPPv2N6paddle17ParallelParameter17getLocalParameterEv"],"paddle::ParallelParameter::gradSem_":[50,7,1,"_CPPv2N6paddle17ParallelParameter8gradSem_E"],"paddle::ParallelParameter::localParam_":[50,7,1,"_CPPv2N6paddle17ParallelParameter11localParam_E"],"paddle::ParallelParameter::majorUpdate":[50,6,1,"_CPPv2N6paddle17ParallelParameter11majorUpdateE4real"],"paddle::ParallelParameter::minorUpdate":[50,6,1,"_CPPv2N6paddle17ParallelParameter11minorUpdateE4real"],"paddle::ParallelParameter::postValueReady":[50,6,1,"_CPPv2N6paddle17ParallelParameter14postValueReadyEv"],"paddle::ParallelParameter::role_":[50,7,1,"_CPPv2N6paddle17ParallelParameter5role_E"],"paddle::ParallelParameter::singleUpdate":[50,6,1,"_CPPv2N6paddle17ParallelParameter12singleUpdateE4real"],"paddle::ParallelParameter::slaveUpdate":[50,6,1,"_CPPv2N6paddle17ParallelParameter11slaveUpdateE4real"],"paddle::ParallelParameter::syncUpdate":[50,6,1,"_CPPv2N6paddle17ParallelParameter10syncUpdateE11TrainerRole4real"],"paddle::ParallelParameter::synchronizeParamter":[50,6,1,"_CPPv2N6paddle17ParallelParameter19synchronizeParamterEv"],"paddle::ParallelParameter::timeWaitGradReady":[50,6,1,"_CPPv2N6paddle17ParallelParameter17timeWaitGradReadyEi"],"paddle::ParallelParameter::valueSem_":[50,7,1,"_CPPv2N6paddle17ParallelParameter9valueSem_E"],"paddle::ParallelParameter::waitGradReady":[50,6,1,"_CPPv2N6paddle17ParallelParameter13waitGradReadyEv"],"paddle::ParallelParameter::~ParallelParameter":[50,6,1,"_CPPv2N6paddle17ParallelParameterD0Ev"],"paddle::ParallelParameterMap":[50,9,1,"_CPPv2N6paddle20ParallelParameterMapE"],"paddle::ParallelParameterPtr":[50,9,1,"_CPPv2N6paddle20ParallelParameterPtrE"],"paddle::ParamInitCallback":[50,9,1,"_CPPv2N6paddle17ParamInitCallbackE"],"paddle::Parameter":[50,8,1,"_CPPv2N6paddle9ParameterE"],"paddle::Parameter::ExecFunc":[50,9,1,"_CPPv2N6paddle9Parameter8ExecFuncE"],"paddle::Parameter::Header":[50,8,1,"_CPPv2N6paddle9Parameter6HeaderE"],"paddle::Parameter::Header::size":[50,7,1,"_CPPv2N6paddle9Parameter6Header4sizeE"],"paddle::Parameter::Header::valueSize":[50,7,1,"_CPPv2N6paddle9Parameter6Header9valueSizeE"],"paddle::Parameter::Header::version":[50,7,1,"_CPPv2N6paddle9Parameter6Header7versionE"],"paddle::Parameter::bufs_":[50,7,1,"_CPPv2N6paddle9Parameter5bufs_E"],"paddle::Parameter::clearGradient":[50,6,1,"_CPPv2N6paddle9Parameter13clearGradientEv"],"paddle::Parameter::clearUpdate":[50,6,1,"_CPPv2N6paddle9Parameter11clearUpdateEv"],"paddle::Parameter::clearValueUpdated":[50,6,1,"_CPPv2N6paddle9Parameter17clearValueUpdatedEv"],"paddle::Parameter::config_":[50,7,1,"_CPPv2N6paddle9Parameter7config_E"],"paddle::Parameter::deviceId_":[50,7,1,"_CPPv2N6paddle9Parameter9deviceId_E"],"paddle::Parameter::enableIntType":[50,6,1,"_CPPv2N6paddle9Parameter13enableIntTypeE13ParameterType6size_t"],"paddle::Parameter::enableSharedType":[50,6,1,"_CPPv2N6paddle9Parameter16enableSharedTypeE13ParameterType9VectorPtr9MatrixPtr"],"paddle::Parameter::enableSparseParameter":[50,6,1,"_CPPv2N6paddle9Parameter21enableSparseParameterEv"],"paddle::Parameter::enableType":[50,6,1,"_CPPv2N6paddle9Parameter10enableTypeE13ParameterType7MatType"],"paddle::Parameter::format_":[50,7,1,"_CPPv2N6paddle9Parameter7format_E"],"paddle::Parameter::getBuf":[50,6,1,"_CPPv2N6paddle9Parameter6getBufE13ParameterType"],"paddle::Parameter::getBufs":[50,6,1,"_CPPv2N6paddle9Parameter7getBufsEv"],"paddle::Parameter::getConfig":[50,6,1,"_CPPv2N6paddle9Parameter9getConfigEv"],"paddle::Parameter::getDeviceId":[50,6,1,"_CPPv2N6paddle9Parameter11getDeviceIdEv"],"paddle::Parameter::getFormat":[50,6,1,"_CPPv2N6paddle9Parameter9getFormatEv"],"paddle::Parameter::getGradientSegments":[50,6,1,"_CPPv2N6paddle9Parameter19getGradientSegmentsEv"],"paddle::Parameter::getID":[50,6,1,"_CPPv2N6paddle9Parameter5getIDEv"],"paddle::Parameter::getInitMean":[50,6,1,"_CPPv2N6paddle9Parameter11getInitMeanEv"],"paddle::Parameter::getInitStandardDeviation":[50,6,1,"_CPPv2N6paddle9Parameter24getInitStandardDeviationEv"],"paddle::Parameter::getIntBuf":[50,6,1,"_CPPv2N6paddle9Parameter9getIntBufE13ParameterType"],"paddle::Parameter::getLearnRate":[50,6,1,"_CPPv2N6paddle9Parameter12getLearnRateEv"],"paddle::Parameter::getMat":[50,6,1,"_CPPv2N6paddle9Parameter6getMatE13ParameterType"],"paddle::Parameter::getName":[50,6,1,"_CPPv2N6paddle9Parameter7getNameEv"],"paddle::Parameter::getSharedCount":[50,6,1,"_CPPv2N6paddle9Parameter14getSharedCountEv"],"paddle::Parameter::getSize":[50,6,1,"_CPPv2N6paddle9Parameter7getSizeEv"],"paddle::Parameter::gradSegments_":[50,7,1,"_CPPv2N6paddle9Parameter13gradSegments_E"],"paddle::Parameter::hasType":[50,6,1,"_CPPv2N6paddle9Parameter7hasTypeE13ParameterType"],"paddle::Parameter::incShared":[50,6,1,"_CPPv2N6paddle9Parameter9incSharedEv"],"paddle::Parameter::initHook":[50,6,1,"_CPPv2N6paddle9Parameter8initHookEv"],"paddle::Parameter::intBufs_":[50,7,1,"_CPPv2N6paddle9Parameter8intBufs_E"],"paddle::Parameter::isFullSize":[50,6,1,"_CPPv2N6paddle9Parameter10isFullSizeEv"],"paddle::Parameter::isSparse":[50,6,1,"_CPPv2N6paddle9Parameter8isSparseEv"],"paddle::Parameter::isSparseRemoteUpdate":[50,6,1,"_CPPv2N6paddle9Parameter20isSparseRemoteUpdateEv"],"paddle::Parameter::isStatic":[50,6,1,"_CPPv2N6paddle9Parameter8isStaticEv"],"paddle::Parameter::isUpdatable":[50,6,1,"_CPPv2N6paddle9Parameter11isUpdatableEv"],"paddle::Parameter::isValueUpdated":[50,6,1,"_CPPv2N6paddle9Parameter14isValueUpdatedEv"],"paddle::Parameter::kFormatVersion":[50,7,1,"_CPPv2N6paddle9Parameter14kFormatVersionE"],"paddle::Parameter::mats_":[50,7,1,"_CPPv2N6paddle9Parameter5mats_E"],"paddle::Parameter::setDevice":[50,6,1,"_CPPv2N6paddle9Parameter9setDeviceEi"],"paddle::Parameter::setID":[50,6,1,"_CPPv2N6paddle9Parameter5setIDE6size_t"],"paddle::Parameter::setIntBuf":[50,6,1,"_CPPv2N6paddle9Parameter9setIntBufE13ParameterTypeRK10IVectorPtr"],"paddle::Parameter::setSharedCount":[50,6,1,"_CPPv2N6paddle9Parameter14setSharedCountEi"],"paddle::Parameter::setValueUpdated":[50,6,1,"_CPPv2N6paddle9Parameter15setValueUpdatedEv"],"paddle::Parameter::sharedCount_":[50,7,1,"_CPPv2N6paddle9Parameter12sharedCount_E"],"paddle::Parameter::singleUpdate":[50,6,1,"_CPPv2N6paddle9Parameter12singleUpdateEPv"],"paddle::Parameter::updateCounter_":[50,7,1,"_CPPv2N6paddle9Parameter14updateCounter_E"],"paddle::Parameter::updateHook":[50,6,1,"_CPPv2N6paddle9Parameter10updateHookEv"],"paddle::Parameter::updated_":[50,7,1,"_CPPv2N6paddle9Parameter8updated_E"],"paddle::Parameter::updaterHooks_":[50,7,1,"_CPPv2N6paddle9Parameter13updaterHooks_E"],"paddle::Parameter::useGpu":[50,6,1,"_CPPv2N6paddle9Parameter6useGpuEv"],"paddle::Parameter::useGpu_":[50,7,1,"_CPPv2N6paddle9Parameter7useGpu_E"],"paddle::ParameterClient2":[53,8,1,"_CPPv2N6paddle16ParameterClient2E"],"paddle::ParameterClient2::ParameterClient2":[53,6,1,"_CPPv2N6paddle16ParameterClient216ParameterClient2Ebii"],"paddle::ParameterClient2::allSegments_":[53,7,1,"_CPPv2N6paddle16ParameterClient212allSegments_E"],"paddle::ParameterClient2::asyncFinishPass":[53,6,1,"_CPPv2N6paddle16ParameterClient215asyncFinishPassE10SyncObject"],"paddle::ParameterClient2::asyncStartPass":[53,6,1,"_CPPv2N6paddle16ParameterClient214asyncStartPassE10SyncObject"],"paddle::ParameterClient2::calcParameterBlockSize":[53,6,1,"_CPPv2N6paddle16ParameterClient222calcParameterBlockSizeERKNSt6vectorI12ParameterPtrEE6size_t"],"paddle::ParameterClient2::createMatrix":[53,6,1,"_CPPv2N6paddle16ParameterClient212createMatrixE7int32_t"],"paddle::ParameterClient2::createVector":[53,6,1,"_CPPv2N6paddle16ParameterClient212createVectorEv"],"paddle::ParameterClient2::destroy":[53,6,1,"_CPPv2N6paddle16ParameterClient27destroyEv"],"paddle::ParameterClient2::doOperation":[53,6,1,"_CPPv2N6paddle16ParameterClient211doOperationER18PreparedOperationsbbb"],"paddle::ParameterClient2::forwardbackwordTime_":[53,7,1,"_CPPv2N6paddle16ParameterClient220forwardbackwordTime_E"],"paddle::ParameterClient2::getPServerParameterGradient":[53,6,1,"_CPPv2N6paddle16ParameterClient227getPServerParameterGradientEv"],"paddle::ParameterClient2::getPServerParameterValue":[53,6,1,"_CPPv2N6paddle16ParameterClient224getPServerParameterValueEv"],"paddle::ParameterClient2::getParameter":[53,6,1,"_CPPv2N6paddle16ParameterClient212getParameterE13ParameterType13ParameterType"],"paddle::ParameterClient2::getParameterSparse":[53,6,1,"_CPPv2N6paddle16ParameterClient218getParameterSparseE13ParameterType13ParameterType"],"paddle::ParameterClient2::inStatus":[53,6,1,"_CPPv2N6paddle16ParameterClient28inStatusE13PServerStatus"],"paddle::ParameterClient2::init":[53,6,1,"_CPPv2N6paddle16ParameterClient24initERKNSt6vectorI12ParameterPtrEE"],"paddle::ParameterClient2::initThreads":[53,6,1,"_CPPv2N6paddle16ParameterClient211initThreadsEv"],"paddle::ParameterClient2::isPassFinish":[53,6,1,"_CPPv2N6paddle16ParameterClient212isPassFinishEv"],"paddle::ParameterClient2::loadValueVector":[53,6,1,"_CPPv2N6paddle16ParameterClient215loadValueVectorERKNSt6stringE"],"paddle::ParameterClient2::multiCall":[53,6,1,"_CPPv2N6paddle16ParameterClient29multiCallEPKcRK7ProtoInPNSt6vectorI8ProtoOutEE"],"paddle::ParameterClient2::parameterMap_":[53,7,1,"_CPPv2N6paddle16ParameterClient213parameterMap_E"],"paddle::ParameterClient2::passFinish_":[53,7,1,"_CPPv2N6paddle16ParameterClient211passFinish_E"],"paddle::ParameterClient2::port_":[53,7,1,"_CPPv2N6paddle16ParameterClient25port_E"],"paddle::ParameterClient2::prepareSendData":[53,6,1,"_CPPv2N6paddle16ParameterClient215prepareSendDataE19ParameterUpdateMode13ParameterTypeRKNSt6vectorI17ParameterSegmentsEE7int64_t4realb13ParameterType11BatchStatusP7SendJob"],"paddle::ParameterClient2::recv":[53,6,1,"_CPPv2N6paddle16ParameterClient24recvEi"],"paddle::ParameterClient2::recvParameter":[53,6,1,"_CPPv2N6paddle16ParameterClient213recvParameterEv"],"paddle::ParameterClient2::releaseMatrix":[53,6,1,"_CPPv2N6paddle16ParameterClient213releaseMatrixE13PServerMatrix"],"paddle::ParameterClient2::releaseVector":[53,6,1,"_CPPv2N6paddle16ParameterClient213releaseVectorE13PServerVector"],"paddle::ParameterClient2::saveValueVector":[53,6,1,"_CPPv2N6paddle16ParameterClient215saveValueVectorERKNSt6stringE"],"paddle::ParameterClient2::send":[53,6,1,"_CPPv2N6paddle16ParameterClient24sendEi"],"paddle::ParameterClient2::sendAndReceiveParameter":[53,6,1,"_CPPv2N6paddle16ParameterClient223sendAndReceiveParameterE19ParameterUpdateMode13ParameterTypeRKNSt6vectorI17ParameterSegmentsEE7int64_t4realb13ParameterType13ParameterType"],"paddle::ParameterClient2::sendParallel":[53,6,1,"_CPPv2N6paddle16ParameterClient212sendParallelEi6size_t13ParameterType"],"paddle::ParameterClient2::sendParameter":[53,6,1,"_CPPv2N6paddle16ParameterClient213sendParameterE19ParameterUpdateMode13ParameterTypeRKNSt6vectorI17ParameterSegmentsEE7int64_t4realb11BatchStatus"],"paddle::ParameterClient2::setConfig":[53,6,1,"_CPPv2N6paddle16ParameterClient29setConfigERK18OptimizationConfigRKNSt6stringEb"],"paddle::ParameterClient2::setForwardbackwardTime":[53,6,1,"_CPPv2N6paddle16ParameterClient222setForwardbackwardTimeE8uint64_t"],"paddle::ParameterClient2::setParameter":[53,6,1,"_CPPv2N6paddle16ParameterClient212setParameterEv"],"paddle::ParameterClient2::setParameterZero":[53,6,1,"_CPPv2N6paddle16ParameterClient216setParameterZeroEv"],"paddle::ParameterClient2::setStatus":[53,6,1,"_CPPv2N6paddle16ParameterClient29setStatusE13PServerStatus"],"paddle::ParameterClient2::setTrainerId":[53,6,1,"_CPPv2N6paddle16ParameterClient212setTrainerIdEi"],"paddle::ParameterClient2::sparseDistribution_":[53,7,1,"_CPPv2N6paddle16ParameterClient219sparseDistribution_E"],"paddle::ParameterClient2::syncThreadPool_":[53,7,1,"_CPPv2N6paddle16ParameterClient215syncThreadPool_E"],"paddle::ParameterClient2::synchronize":[53,6,1,"_CPPv2N6paddle16ParameterClient211synchronizeE10SyncObject"],"paddle::ParameterClient2::trainerId_":[53,7,1,"_CPPv2N6paddle16ParameterClient210trainerId_E"],"paddle::ParameterClient2::vectorAddMult":[53,6,1,"_CPPv2N6paddle16ParameterClient213vectorAddMultE13PServerVector13PServerVector4real"],"paddle::ParameterClient2::vectorAddMultInto":[53,6,1,"_CPPv2N6paddle16ParameterClient217vectorAddMultIntoE13PServerVector13PServerVector13PServerVector4real"],"paddle::ParameterClient2::vectorCopy":[53,6,1,"_CPPv2N6paddle16ParameterClient210vectorCopyE13PServerVector13PServerVector"],"paddle::ParameterClient2::vectorDotProduct":[53,6,1,"_CPPv2N6paddle16ParameterClient216vectorDotProductE13PServerVector13PServerVector"],"paddle::ParameterClient2::vectorScale":[53,6,1,"_CPPv2N6paddle16ParameterClient211vectorScaleE13PServerVector4real"],"paddle::ParameterClient2::vectorScaleInto":[53,6,1,"_CPPv2N6paddle16ParameterClient215vectorScaleIntoE13PServerVector13PServerVector4real"],"paddle::ParameterClient2::waitForStatus":[53,6,1,"_CPPv2N6paddle16ParameterClient213waitForStatusE13PServerStatus"],"paddle::ParameterClient2::waitPassFinish":[53,6,1,"_CPPv2N6paddle16ParameterClient214waitPassFinishEv"],"paddle::ParameterClient2::waitPassStart":[53,6,1,"_CPPv2N6paddle16ParameterClient213waitPassStartEv"],"paddle::ParameterClient2::~ParameterClient2":[53,6,1,"_CPPv2N6paddle16ParameterClient2D0Ev"],"paddle::ParameterMap":[50,9,1,"_CPPv2N6paddle12ParameterMapE"],"paddle::ParameterOptimizer":[48,8,1,"_CPPv2N6paddle18ParameterOptimizerE"],"paddle::ParameterOptimizer::ParameterOptimizer":[48,6,1,"_CPPv2N6paddle18ParameterOptimizer18ParameterOptimizerERK18OptimizationConfig"],"paddle::ParameterOptimizer::TraverseCallback":[48,9,1,"_CPPv2N6paddle18ParameterOptimizer16TraverseCallbackE"],"paddle::ParameterOptimizer::TraverseCallbackVec":[48,9,1,"_CPPv2N6paddle18ParameterOptimizer19TraverseCallbackVecE"],"paddle::ParameterOptimizer::addParameterType":[48,6,1,"_CPPv2N6paddle18ParameterOptimizer16addParameterTypeE13ParameterType"],"paddle::ParameterOptimizer::apply":[48,6,1,"_CPPv2N6paddle18ParameterOptimizer5applyEv"],"paddle::ParameterOptimizer::applyDecay_":[48,7,1,"_CPPv2N6paddle18ParameterOptimizer11applyDecay_E"],"paddle::ParameterOptimizer::calcLearningRate":[48,6,1,"_CPPv2N6paddle18ParameterOptimizer16calcLearningRateE7int64_t7int64_t"],"paddle::ParameterOptimizer::composeCallbacks":[48,6,1,"_CPPv2N6paddle18ParameterOptimizer16composeCallbacksERK19TraverseCallbackVec"],"paddle::ParameterOptimizer::finishBatch":[48,6,1,"_CPPv2N6paddle18ParameterOptimizer11finishBatchEv"],"paddle::ParameterOptimizer::finishCatchUpWith":[48,6,1,"_CPPv2N6paddle18ParameterOptimizer17finishCatchUpWithEv"],"paddle::ParameterOptimizer::finishPass":[48,6,1,"_CPPv2N6paddle18ParameterOptimizer10finishPassEv"],"paddle::ParameterOptimizer::firstTime_":[48,7,1,"_CPPv2N6paddle18ParameterOptimizer10firstTime_E"],"paddle::ParameterOptimizer::getLearningRate":[48,6,1,"_CPPv2N6paddle18ParameterOptimizer15getLearningRateEv"],"paddle::ParameterOptimizer::getParameterTypes":[48,6,1,"_CPPv2N6paddle18ParameterOptimizer17getParameterTypesEv"],"paddle::ParameterOptimizer::init":[48,6,1,"_CPPv2N6paddle18ParameterOptimizer4initE6size_tPK15ParameterConfig"],"paddle::ParameterOptimizer::learningRateScheduler_":[48,7,1,"_CPPv2N6paddle18ParameterOptimizer22learningRateScheduler_E"],"paddle::ParameterOptimizer::learningRate_":[48,7,1,"_CPPv2N6paddle18ParameterOptimizer13learningRate_E"],"paddle::ParameterOptimizer::needSpecialTraversal":[48,6,1,"_CPPv2N6paddle18ParameterOptimizer20needSpecialTraversalERK15ParameterConfig"],"paddle::ParameterOptimizer::optConfig_":[48,7,1,"_CPPv2N6paddle18ParameterOptimizer10optConfig_E"],"paddle::ParameterOptimizer::parameterTypes_":[48,7,1,"_CPPv2N6paddle18ParameterOptimizer15parameterTypes_E"],"paddle::ParameterOptimizer::pass_":[48,7,1,"_CPPv2N6paddle18ParameterOptimizer5pass_E"],"paddle::ParameterOptimizer::restore":[48,6,1,"_CPPv2N6paddle18ParameterOptimizer7restoreEv"],"paddle::ParameterOptimizer::setNoDecay":[48,6,1,"_CPPv2N6paddle18ParameterOptimizer10setNoDecayEv"],"paddle::ParameterOptimizer::startBatch":[48,6,1,"_CPPv2N6paddle18ParameterOptimizer10startBatchE7int64_t"],"paddle::ParameterOptimizer::startCatchUpWith":[48,6,1,"_CPPv2N6paddle18ParameterOptimizer16startCatchUpWithEv"],"paddle::ParameterOptimizer::startPass":[48,6,1,"_CPPv2N6paddle18ParameterOptimizer9startPassEv"],"paddle::ParameterOptimizer::update":[48,6,1,"_CPPv2N6paddle18ParameterOptimizer6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::ParameterOptimizer::~ParameterOptimizer":[48,6,1,"_CPPv2N6paddle18ParameterOptimizerD0Ev"],"paddle::ParameterPtr":[50,9,1,"_CPPv2N6paddle12ParameterPtrE"],"paddle::ParameterReluLayer":[41,8,1,"_CPPv2N6paddle18ParameterReluLayerE"],"paddle::ParameterReluLayer::ParameterReluLayer":[41,6,1,"_CPPv2N6paddle18ParameterReluLayer18ParameterReluLayerERK11LayerConfig"],"paddle::ParameterReluLayer::backward":[41,6,1,"_CPPv2N6paddle18ParameterReluLayer8backwardERK14UpdateCallback"],"paddle::ParameterReluLayer::forward":[41,6,1,"_CPPv2N6paddle18ParameterReluLayer7forwardE8PassType"],"paddle::ParameterReluLayer::init":[41,6,1,"_CPPv2N6paddle18ParameterReluLayer4initERK8LayerMapRK12ParameterMap"],"paddle::ParameterReluLayer::partialSum_":[41,7,1,"_CPPv2N6paddle18ParameterReluLayer11partialSum_E"],"paddle::ParameterReluLayer::weight_":[41,7,1,"_CPPv2N6paddle18ParameterReluLayer7weight_E"],"paddle::ParameterReluLayer::~ParameterReluLayer":[41,6,1,"_CPPv2N6paddle18ParameterReluLayerD0Ev"],"paddle::ParameterServer2":[58,8,1,"_CPPv2N6paddle16ParameterServer2E"],"paddle::ParameterServer2::BlockInfo":[58,8,1,"_CPPv2N6paddle16ParameterServer29BlockInfoE"],"paddle::ParameterServer2::BlockInfo::config":[58,7,1,"_CPPv2N6paddle16ParameterServer29BlockInfo6configE"],"paddle::ParameterServer2::BlockInfo::lock":[58,7,1,"_CPPv2N6paddle16ParameterServer29BlockInfo4lockE"],"paddle::ParameterServer2::BlockInfo::offset":[58,7,1,"_CPPv2N6paddle16ParameterServer29BlockInfo6offsetE"],"paddle::ParameterServer2::BlockInfo::optimizer":[58,7,1,"_CPPv2N6paddle16ParameterServer29BlockInfo9optimizerE"],"paddle::ParameterServer2::BlockKey":[58,9,1,"_CPPv2N6paddle16ParameterServer28BlockKeyE"],"paddle::ParameterServer2::BlockKeyHash":[58,8,1,"_CPPv2N6paddle16ParameterServer212BlockKeyHashE"],"paddle::ParameterServer2::BlockKeyHash::operator()":[58,6,1,"_CPPv2N6paddle16ParameterServer212BlockKeyHashclERK8BlockKey"],"paddle::ParameterServer2::BlockMap":[58,9,1,"_CPPv2N6paddle16ParameterServer28BlockMapE"],"paddle::ParameterServer2::BlockSegments":[58,9,1,"_CPPv2N6paddle16ParameterServer213BlockSegmentsE"],"paddle::ParameterServer2::Buffer":[58,8,1,"_CPPv2N6paddle16ParameterServer26BufferE"],"paddle::ParameterServer2::Buffer::base":[58,7,1,"_CPPv2N6paddle16ParameterServer26Buffer4baseE"],"paddle::ParameterServer2::Buffer::size":[58,7,1,"_CPPv2N6paddle16ParameterServer26Buffer4sizeE"],"paddle::ParameterServer2::ExecFunc":[58,9,1,"_CPPv2N6paddle16ParameterServer28ExecFuncE"],"paddle::ParameterServer2::ParameterServer2":[58,6,1,"_CPPv2N6paddle16ParameterServer216ParameterServer2ERKNSt6stringEii"],"paddle::ParameterServer2::ReadWriteBuffer":[58,8,1,"_CPPv2N6paddle16ParameterServer215ReadWriteBufferE"],"paddle::ParameterServer2::ReadWriteBuffer::AlignElementCount":[58,7,1,"_CPPv2N6paddle16ParameterServer215ReadWriteBuffer17AlignElementCountE"],"paddle::ParameterServer2::ReadWriteBuffer::IsTLargerThanAlign":[58,7,1,"_CPPv2N6paddle16ParameterServer215ReadWriteBuffer18IsTLargerThanAlignE"],"paddle::ParameterServer2::ReadWriteBuffer::curOffset_":[58,7,1,"_CPPv2N6paddle16ParameterServer215ReadWriteBuffer10curOffset_E"],"paddle::ParameterServer2::ReadWriteBuffer::nextBlock":[58,6,1,"_CPPv2N6paddle16ParameterServer215ReadWriteBuffer9nextBlockE6size_t"],"paddle::ParameterServer2::ReadWriteBuffer::resetAlignAlloc":[58,6,1,"_CPPv2N6paddle16ParameterServer215ReadWriteBuffer15resetAlignAllocEv"],"paddle::ParameterServer2::ReadWriteBuffer::resizeWithAlignHints":[58,6,1,"_CPPv2N6paddle16ParameterServer215ReadWriteBuffer20resizeWithAlignHintsE6size_t6size_t"],"paddle::ParameterServer2::addGradBegin_":[58,7,1,"_CPPv2N6paddle16ParameterServer213addGradBegin_E"],"paddle::ParameterServer2::addGradient":[58,6,1,"_CPPv2N6paddle16ParameterServer211addGradientERK20SendParameterRequestRNSt6vectorI6BufferEEP21SendParameterResponsePNSt6vectorI6BufferEE"],"paddle::ParameterServer2::allClientPassFinish_":[58,7,1,"_CPPv2N6paddle16ParameterServer220allClientPassFinish_E"],"paddle::ParameterServer2::asyncFinishPass":[58,6,1,"_CPPv2N6paddle16ParameterServer215asyncFinishPassERK18SynchronizeRequest21ProtoResponseCallback"],"paddle::ParameterServer2::asyncGrdientCommitCheckAndStat":[58,6,1,"_CPPv2N6paddle16ParameterServer230asyncGrdientCommitCheckAndStatERK20SendParameterRequest"],"paddle::ParameterServer2::asyncLaggedGradientsNum_":[58,7,1,"_CPPv2N6paddle16ParameterServer224asyncLaggedGradientsNum_E"],"paddle::ParameterServer2::asyncLaggedThreshold_":[58,7,1,"_CPPv2N6paddle16ParameterServer221asyncLaggedThreshold_E"],"paddle::ParameterServer2::asyncSGD":[58,6,1,"_CPPv2N6paddle16ParameterServer28asyncSGDERK20SendParameterRequestRNSt6vectorI6BufferEEP21SendParameterResponsePNSt6vectorI6BufferEE"],"paddle::ParameterServer2::asyncTrainerCommitStat_":[58,7,1,"_CPPv2N6paddle16ParameterServer223asyncTrainerCommitStat_E"],"paddle::ParameterServer2::asyncTrainerDiscardStat_":[58,7,1,"_CPPv2N6paddle16ParameterServer224asyncTrainerDiscardStat_E"],"paddle::ParameterServer2::asyncTrainerSteps_":[58,7,1,"_CPPv2N6paddle16ParameterServer218asyncTrainerSteps_E"],"paddle::ParameterServer2::asyncUpdateStat_":[58,7,1,"_CPPv2N6paddle16ParameterServer216asyncUpdateStat_E"],"paddle::ParameterServer2::asyncUpdateSteps_":[58,7,1,"_CPPv2N6paddle16ParameterServer217asyncUpdateSteps_E"],"paddle::ParameterServer2::batchId_":[58,7,1,"_CPPv2N6paddle16ParameterServer28batchId_E"],"paddle::ParameterServer2::blockIdMap_":[58,7,1,"_CPPv2N6paddle16ParameterServer211blockIdMap_E"],"paddle::ParameterServer2::blockInfos_":[58,7,1,"_CPPv2N6paddle16ParameterServer211blockInfos_E"],"paddle::ParameterServer2::blockOffsetMap_":[58,7,1,"_CPPv2N6paddle16ParameterServer215blockOffsetMap_E"],"paddle::ParameterServer2::blockTraverse":[58,6,1,"_CPPv2N6paddle16ParameterServer213blockTraverseER9BlockInfoRK15ParameterConfig7int64_t6size_tA_K9VectorPtrRKN18ParameterOptimizer16TraverseCallbackE"],"paddle::ParameterServer2::callbackVec_":[58,7,1,"_CPPv2N6paddle16ParameterServer212callbackVec_E"],"paddle::ParameterServer2::clearUnusedSegments":[58,6,1,"_CPPv2N6paddle16ParameterServer219clearUnusedSegmentsEP9CpuVector"],"paddle::ParameterServer2::configMap_":[58,7,1,"_CPPv2N6paddle16ParameterServer210configMap_E"],"paddle::ParameterServer2::config_":[58,7,1,"_CPPv2N6paddle16ParameterServer27config_E"],"paddle::ParameterServer2::cost_":[58,7,1,"_CPPv2N6paddle16ParameterServer25cost_E"],"paddle::ParameterServer2::createMatrix":[58,6,1,"_CPPv2N6paddle16ParameterServer212createMatrixERK19CreateMatrixRequest21ProtoResponseCallback"],"paddle::ParameterServer2::createVector":[58,6,1,"_CPPv2N6paddle16ParameterServer212createVectorERK19CreateVectorRequest21ProtoResponseCallback"],"paddle::ParameterServer2::dataMems_":[58,7,1,"_CPPv2N6paddle16ParameterServer29dataMems_E"],"paddle::ParameterServer2::dataSize_":[58,7,1,"_CPPv2N6paddle16ParameterServer29dataSize_E"],"paddle::ParameterServer2::doOperation":[58,6,1,"_CPPv2N6paddle16ParameterServer211doOperationERK18DoOperationRequest21ProtoResponseCallback"],"paddle::ParameterServer2::getBlockId":[58,6,1,"_CPPv2N6paddle16ParameterServer210getBlockIdERK14ParameterBlock"],"paddle::ParameterServer2::getBlockOffset":[58,6,1,"_CPPv2N6paddle16ParameterServer214getBlockOffsetERK14ParameterBlock"],"paddle::ParameterServer2::getParameter":[58,6,1,"_CPPv2N6paddle16ParameterServer212getParameterERK20SendParameterRequestRNSt6vectorI6BufferEEP21SendParameterResponsePNSt6vectorI6BufferEE"],"paddle::ParameterServer2::getParameterConfig":[58,6,1,"_CPPv2N6paddle16ParameterServer218getParameterConfigERK14ParameterBlock"],"paddle::ParameterServer2::getParameterSparse":[58,6,1,"_CPPv2N6paddle16ParameterServer218getParameterSparseERK20SendParameterRequestRNSt6vectorI6BufferEEP21SendParameterResponsePNSt6vectorI6BufferEE"],"paddle::ParameterServer2::getStatus":[58,6,1,"_CPPv2N6paddle16ParameterServer29getStatusERK16GetStatusRequest21ProtoResponseCallback"],"paddle::ParameterServer2::gradientReadyBarrier_":[58,7,1,"_CPPv2N6paddle16ParameterServer221gradientReadyBarrier_E"],"paddle::ParameterServer2::init":[58,6,1,"_CPPv2N6paddle16ParameterServer24initEv"],"paddle::ParameterServer2::isSparseServer_":[58,7,1,"_CPPv2N6paddle16ParameterServer215isSparseServer_E"],"paddle::ParameterServer2::isValidMatrixHandle":[58,6,1,"_CPPv2N6paddle16ParameterServer219isValidMatrixHandleE7int64_tP8Response"],"paddle::ParameterServer2::isValidVectorHandle":[58,6,1,"_CPPv2N6paddle16ParameterServer219isValidVectorHandleE7int64_tP8Response"],"paddle::ParameterServer2::kRetMsgInvalidMatrixHandle":[58,7,1,"_CPPv2N6paddle16ParameterServer226kRetMsgInvalidMatrixHandleE"],"paddle::ParameterServer2::kRetMsgInvalidVectorHandle":[58,7,1,"_CPPv2N6paddle16ParameterServer226kRetMsgInvalidVectorHandleE"],"paddle::ParameterServer2::kRetMsgUnknownOperation":[58,7,1,"_CPPv2N6paddle16ParameterServer223kRetMsgUnknownOperationE"],"paddle::ParameterServer2::loadValueVector":[58,6,1,"_CPPv2N6paddle16ParameterServer215loadValueVectorERK16LoadValueRequest21ProtoResponseCallback"],"paddle::ParameterServer2::matrices_":[58,7,1,"_CPPv2N6paddle16ParameterServer29matrices_E"],"paddle::ParameterServer2::mergeSegments":[58,6,1,"_CPPv2N6paddle16ParameterServer213mergeSegmentsEP13BlockSegments"],"paddle::ParameterServer2::mpiSize_":[58,7,1,"_CPPv2N6paddle16ParameterServer28mpiSize_E"],"paddle::ParameterServer2::numPassFinishClients_":[58,7,1,"_CPPv2N6paddle16ParameterServer221numPassFinishClients_E"],"paddle::ParameterServer2::numSamplesProcessed_":[58,7,1,"_CPPv2N6paddle16ParameterServer220numSamplesProcessed_E"],"paddle::ParameterServer2::opFuncs":[58,7,1,"_CPPv2N6paddle16ParameterServer27opFuncsE"],"paddle::ParameterServer2::op_COPY":[58,6,1,"_CPPv2N6paddle16ParameterServer27op_COPYERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_RESET":[58,6,1,"_CPPv2N6paddle16ParameterServer28op_RESETERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_SGD":[58,6,1,"_CPPv2N6paddle16ParameterServer26op_SGDERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_apply":[58,6,1,"_CPPv2N6paddle16ParameterServer28op_applyERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_au":[58,6,1,"_CPPv2N6paddle16ParameterServer25op_auERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_au_bv":[58,6,1,"_CPPv2N6paddle16ParameterServer28op_au_bvERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_au_bv_cw":[58,6,1,"_CPPv2N6paddle16ParameterServer211op_au_bv_cwERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_cost":[58,6,1,"_CPPv2N6paddle16ParameterServer27op_costERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_dir_deriv":[58,6,1,"_CPPv2N6paddle16ParameterServer212op_dir_derivERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_finish_pass":[58,6,1,"_CPPv2N6paddle16ParameterServer214op_finish_passERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_fix_dir_signs":[58,6,1,"_CPPv2N6paddle16ParameterServer216op_fix_dir_signsERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_fix_omega_signs":[58,6,1,"_CPPv2N6paddle16ParameterServer218op_fix_omega_signsERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_load":[58,6,1,"_CPPv2N6paddle16ParameterServer27op_loadERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_make_steepest_desc_dir":[58,6,1,"_CPPv2N6paddle16ParameterServer225op_make_steepest_desc_dirERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_randomize":[58,6,1,"_CPPv2N6paddle16ParameterServer212op_randomizeERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_save":[58,6,1,"_CPPv2N6paddle16ParameterServer27op_saveERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_start_pass":[58,6,1,"_CPPv2N6paddle16ParameterServer213op_start_passERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_utv":[58,6,1,"_CPPv2N6paddle16ParameterServer26op_utvERK9OperationP15OperationResult"],"paddle::ParameterServer2::parallelExecForEachBlock":[58,6,1,"_CPPv2N6paddle16ParameterServer224parallelExecForEachBlockE8ExecFunc"],"paddle::ParameterServer2::parameterMutex_":[58,7,1,"_CPPv2N6paddle16ParameterServer215parameterMutex_E"],"paddle::ParameterServer2::parameterReadyBarrier_":[58,7,1,"_CPPv2N6paddle16ParameterServer222parameterReadyBarrier_E"],"paddle::ParameterServer2::passBarrier_":[58,7,1,"_CPPv2N6paddle16ParameterServer212passBarrier_E"],"paddle::ParameterServer2::printAsyncGradientCommitStatAndReset":[58,6,1,"_CPPv2N6paddle16ParameterServer236printAsyncGradientCommitStatAndResetEv"],"paddle::ParameterServer2::readAllBlocks":[58,6,1,"_CPPv2N6paddle16ParameterServer213readAllBlocksEP9MsgReaderPNSt6vectorIN16ParameterServer26BufferEEE"],"paddle::ParameterServer2::readWriteBuffer_":[58,7,1,"_CPPv2N6paddle16ParameterServer216readWriteBuffer_E"],"paddle::ParameterServer2::reduceAndSendData":[58,6,1,"_CPPv2N6paddle16ParameterServer217reduceAndSendDataERK15SendDataRequestRNSt10unique_ptrI9MsgReaderEER23ProtoResponseCallbackEx"],"paddle::ParameterServer2::releaseMatrix":[58,6,1,"_CPPv2N6paddle16ParameterServer213releaseMatrixERK20ReleaseMatrixRequest21ProtoResponseCallback"],"paddle::ParameterServer2::releaseVector":[58,6,1,"_CPPv2N6paddle16ParameterServer213releaseVectorERK20ReleaseVectorRequest21ProtoResponseCallback"],"paddle::ParameterServer2::requestVec_":[58,7,1,"_CPPv2N6paddle16ParameterServer211requestVec_E"],"paddle::ParameterServer2::saveValueVector":[58,6,1,"_CPPv2N6paddle16ParameterServer215saveValueVectorERK16SaveValueRequest21ProtoResponseCallback"],"paddle::ParameterServer2::sendBackParameter":[58,6,1,"_CPPv2N6paddle16ParameterServer217sendBackParameterERK14ParameterBlockiP21SendParameterResponsePNSt6vectorI6BufferEE"],"paddle::ParameterServer2::sendBackParameterSparse":[58,6,1,"_CPPv2N6paddle16ParameterServer223sendBackParameterSparseERK14ParameterBlockiP21SendParameterResponseP6Buffer6size_tPNSt6vectorI6BufferEE"],"paddle::ParameterServer2::sendData":[58,6,1,"_CPPv2N6paddle16ParameterServer28sendDataERK15SendDataRequestNSt10unique_ptrI9MsgReaderEE23ProtoResponseCallbackEx"],"paddle::ParameterServer2::sendParameter":[58,6,1,"_CPPv2N6paddle16ParameterServer213sendParameterERK20SendParameterRequestNSt10unique_ptrI9MsgReaderEE23ProtoResponseCallbackEx"],"paddle::ParameterServer2::serverId_":[58,7,1,"_CPPv2N6paddle16ParameterServer29serverId_E"],"paddle::ParameterServer2::setConfig":[58,6,1,"_CPPv2N6paddle16ParameterServer29setConfigERK16SetConfigRequest21ProtoResponseCallback"],"paddle::ParameterServer2::setParameter":[58,6,1,"_CPPv2N6paddle16ParameterServer212setParameterERK20SendParameterRequestRNSt6vectorI6BufferEEP21SendParameterResponsePNSt6vectorI6BufferEE"],"paddle::ParameterServer2::setStatus":[58,6,1,"_CPPv2N6paddle16ParameterServer29setStatusERK16SetStatusRequest21ProtoResponseCallback"],"paddle::ParameterServer2::size_":[58,7,1,"_CPPv2N6paddle16ParameterServer25size_E"],"paddle::ParameterServer2::statSet_":[58,7,1,"_CPPv2N6paddle16ParameterServer28statSet_E"],"paddle::ParameterServer2::status_":[58,7,1,"_CPPv2N6paddle16ParameterServer27status_E"],"paddle::ParameterServer2::syncThreadPool_":[58,7,1,"_CPPv2N6paddle16ParameterServer215syncThreadPool_E"],"paddle::ParameterServer2::synchronize":[58,6,1,"_CPPv2N6paddle16ParameterServer211synchronizeERK18SynchronizeRequest21ProtoResponseCallback"],"paddle::ParameterServer2::synchronizeBarriers_":[58,7,1,"_CPPv2N6paddle16ParameterServer220synchronizeBarriers_E"],"paddle::ParameterServer2::templateReduceSum":[58,6,1,"_CPPv2N6paddle16ParameterServer217templateReduceSumERK15SendDataRequestRNSt10unique_ptrI9MsgReaderEER23ProtoResponseCallbackEx"],"paddle::ParameterServer2::tuningAsyncsgdFinished":[58,6,1,"_CPPv2N6paddle16ParameterServer222tuningAsyncsgdFinishedEv"],"paddle::ParameterServer2::tuningAsyncsgdMidOutput":[58,6,1,"_CPPv2N6paddle16ParameterServer223tuningAsyncsgdMidOutputEv"],"paddle::ParameterServer2::tuningSgdFinished":[58,6,1,"_CPPv2N6paddle16ParameterServer217tuningSgdFinishedEv"],"paddle::ParameterServer2::tuningSgdMidOutput":[58,6,1,"_CPPv2N6paddle16ParameterServer218tuningSgdMidOutputEv"],"paddle::ParameterServer2::usedSegments_":[58,7,1,"_CPPv2N6paddle16ParameterServer213usedSegments_E"],"paddle::ParameterServer2::vectors_":[58,7,1,"_CPPv2N6paddle16ParameterServer28vectors_E"],"paddle::ParameterServer2::waitPassFinish":[58,6,1,"_CPPv2N6paddle16ParameterServer214waitPassFinishERK21WaitPassFinishRequest21ProtoResponseCallback"],"paddle::ParameterServer2::waitPassStart":[58,6,1,"_CPPv2N6paddle16ParameterServer213waitPassStartERK20WaitPassStartRequest21ProtoResponseCallback"],"paddle::ParameterServer2::~ParameterServer2":[58,6,1,"_CPPv2N6paddle16ParameterServer2D0Ev"],"paddle::ParameterUpdater":[52,8,1,"_CPPv2N6paddle16ParameterUpdaterE"],"paddle::ParameterUpdater::ParameterUpdater":[52,6,1,"_CPPv2N6paddle16ParameterUpdater16ParameterUpdaterEv"],"paddle::ParameterUpdater::addParameterType":[52,6,1,"_CPPv2N6paddle16ParameterUpdater16addParameterTypeE13ParameterType"],"paddle::ParameterUpdater::apply":[52,6,1,"_CPPv2N6paddle16ParameterUpdater5applyEv"],"paddle::ParameterUpdater::catchUpWith":[52,6,1,"_CPPv2N6paddle16ParameterUpdater11catchUpWithEv"],"paddle::ParameterUpdater::finishBatch":[52,6,1,"_CPPv2N6paddle16ParameterUpdater11finishBatchE4real"],"paddle::ParameterUpdater::finishPass":[52,6,1,"_CPPv2N6paddle16ParameterUpdater10finishPassE4real"],"paddle::ParameterUpdater::getParameterTypes":[52,6,1,"_CPPv2N6paddle16ParameterUpdater17getParameterTypesEv"],"paddle::ParameterUpdater::getParametersRemote":[52,6,1,"_CPPv2N6paddle16ParameterUpdater19getParametersRemoteEbb"],"paddle::ParameterUpdater::init":[52,6,1,"_CPPv2N6paddle16ParameterUpdater4initERNSt6vectorI12ParameterPtrEE"],"paddle::ParameterUpdater::loadParametersRemote":[52,6,1,"_CPPv2N6paddle16ParameterUpdater20loadParametersRemoteERKNSt6stringE"],"paddle::ParameterUpdater::nonStaticParaIDMap_":[52,7,1,"_CPPv2N6paddle16ParameterUpdater19nonStaticParaIDMap_E"],"paddle::ParameterUpdater::parameterTypes_":[52,7,1,"_CPPv2N6paddle16ParameterUpdater15parameterTypes_E"],"paddle::ParameterUpdater::parameters_":[52,7,1,"_CPPv2N6paddle16ParameterUpdater11parameters_E"],"paddle::ParameterUpdater::randParametersRemote":[52,6,1,"_CPPv2N6paddle16ParameterUpdater20randParametersRemoteEv"],"paddle::ParameterUpdater::restore":[52,6,1,"_CPPv2N6paddle16ParameterUpdater7restoreEv"],"paddle::ParameterUpdater::saveParametersRemote":[52,6,1,"_CPPv2N6paddle16ParameterUpdater20saveParametersRemoteERKNSt6stringE"],"paddle::ParameterUpdater::setForwardbackwardTime":[52,6,1,"_CPPv2N6paddle16ParameterUpdater22setForwardbackwardTimeE8uint64_t"],"paddle::ParameterUpdater::startBatch":[52,6,1,"_CPPv2N6paddle16ParameterUpdater10startBatchE7int64_t"],"paddle::ParameterUpdater::startPass":[52,6,1,"_CPPv2N6paddle16ParameterUpdater9startPassEv"],"paddle::ParameterUpdater::update":[52,6,1,"_CPPv2N6paddle16ParameterUpdater6updateEP9Parameter"],"paddle::ParameterUpdater::updateImpl":[52,6,1,"_CPPv2N6paddle16ParameterUpdater10updateImplEP9Parameter"],"paddle::ParameterUpdater::~ParameterUpdater":[52,6,1,"_CPPv2N6paddle16ParameterUpdaterD0Ev"],"paddle::ParameterUpdaterComposite":[52,8,1,"_CPPv2N6paddle25ParameterUpdaterCompositeE"],"paddle::ParameterUpdaterComposite::ParameterUpdaterComposite":[52,6,1,"_CPPv2N6paddle25ParameterUpdaterComposite25ParameterUpdaterCompositeEv"],"paddle::ParameterUpdaterComposite::apply":[52,6,1,"_CPPv2N6paddle25ParameterUpdaterComposite5applyEv"],"paddle::ParameterUpdaterComposite::catchUpWith":[52,6,1,"_CPPv2N6paddle25ParameterUpdaterComposite11catchUpWithEv"],"paddle::ParameterUpdaterComposite::finishBatch":[52,6,1,"_CPPv2N6paddle25ParameterUpdaterComposite11finishBatchE4real"],"paddle::ParameterUpdaterComposite::finishPass":[52,6,1,"_CPPv2N6paddle25ParameterUpdaterComposite10finishPassE4real"],"paddle::ParameterUpdaterComposite::getParametersRemote":[52,6,1,"_CPPv2N6paddle25ParameterUpdaterComposite19getParametersRemoteEbb"],"paddle::ParameterUpdaterComposite::init":[52,6,1,"_CPPv2N6paddle25ParameterUpdaterComposite4initERNSt6vectorI12ParameterPtrEE"],"paddle::ParameterUpdaterComposite::loadParametersRemote":[52,6,1,"_CPPv2N6paddle25ParameterUpdaterComposite20loadParametersRemoteERKNSt6stringE"],"paddle::ParameterUpdaterComposite::randParametersRemote":[52,6,1,"_CPPv2N6paddle25ParameterUpdaterComposite20randParametersRemoteEv"],"paddle::ParameterUpdaterComposite::restore":[52,6,1,"_CPPv2N6paddle25ParameterUpdaterComposite7restoreEv"],"paddle::ParameterUpdaterComposite::saveParametersRemote":[52,6,1,"_CPPv2N6paddle25ParameterUpdaterComposite20saveParametersRemoteERKNSt6stringE"],"paddle::ParameterUpdaterComposite::setForwardbackwardTime":[52,6,1,"_CPPv2N6paddle25ParameterUpdaterComposite22setForwardbackwardTimeE8uint64_t"],"paddle::ParameterUpdaterComposite::startBatch":[52,6,1,"_CPPv2N6paddle25ParameterUpdaterComposite10startBatchE7int64_t"],"paddle::ParameterUpdaterComposite::startPass":[52,6,1,"_CPPv2N6paddle25ParameterUpdaterComposite9startPassEv"],"paddle::ParameterUpdaterComposite::syncThreadPool_":[52,7,1,"_CPPv2N6paddle25ParameterUpdaterComposite15syncThreadPool_E"],"paddle::ParameterUpdaterComposite::updateImpl":[52,6,1,"_CPPv2N6paddle25ParameterUpdaterComposite10updateImplEP9Parameter"],"paddle::ParameterUpdaterComposite::updaters_":[52,7,1,"_CPPv2N6paddle25ParameterUpdaterComposite9updaters_E"],"paddle::ParameterUpdaterComposite::~ParameterUpdaterComposite":[52,6,1,"_CPPv2N6paddle25ParameterUpdaterCompositeD0Ev"],"paddle::PnpairEvaluator":[36,8,1,"_CPPv2N6paddle15PnpairEvaluatorE"],"paddle::PnpairEvaluator::PnpairEvaluator":[36,6,1,"_CPPv2N6paddle15PnpairEvaluator15PnpairEvaluatorEv"],"paddle::PnpairEvaluator::PredictionResult":[36,8,1,"_CPPv2N6paddle15PnpairEvaluator16PredictionResultE"],"paddle::PnpairEvaluator::PredictionResult::PredictionResult":[36,6,1,"_CPPv2N6paddle15PnpairEvaluator16PredictionResult16PredictionResultE4realii4real"],"paddle::PnpairEvaluator::PredictionResult::label":[36,7,1,"_CPPv2N6paddle15PnpairEvaluator16PredictionResult5labelE"],"paddle::PnpairEvaluator::PredictionResult::out":[36,7,1,"_CPPv2N6paddle15PnpairEvaluator16PredictionResult3outE"],"paddle::PnpairEvaluator::PredictionResult::queryid":[36,7,1,"_CPPv2N6paddle15PnpairEvaluator16PredictionResult7queryidE"],"paddle::PnpairEvaluator::PredictionResult::weight":[36,7,1,"_CPPv2N6paddle15PnpairEvaluator16PredictionResult6weightE"],"paddle::PnpairEvaluator::calc":[36,6,1,"_CPPv2N6paddle15PnpairEvaluator4calcERNSt6vectorI16PredictionResultEE"],"paddle::PnpairEvaluator::distributeEval":[36,6,1,"_CPPv2N6paddle15PnpairEvaluator14distributeEvalEP16ParameterClient2"],"paddle::PnpairEvaluator::evalImp":[36,6,1,"_CPPv2N6paddle15PnpairEvaluator7evalImpERNSt6vectorI8ArgumentEE"],"paddle::PnpairEvaluator::finish":[36,6,1,"_CPPv2N6paddle15PnpairEvaluator6finishEv"],"paddle::PnpairEvaluator::predictArray_":[36,7,1,"_CPPv2N6paddle15PnpairEvaluator13predictArray_E"],"paddle::PnpairEvaluator::printPredictResults":[36,6,1,"_CPPv2N6paddle15PnpairEvaluator19printPredictResultsEv"],"paddle::PnpairEvaluator::printStats":[36,6,1,"_CPPv2N6paddle15PnpairEvaluator10printStatsERNSt7ostreamE"],"paddle::PnpairEvaluator::start":[36,6,1,"_CPPv2N6paddle15PnpairEvaluator5startEv"],"paddle::PnpairEvaluator::stat":[36,6,1,"_CPPv2N6paddle15PnpairEvaluator4statE6size_t6size_tP16PredictionResultRdRdRd"],"paddle::PnpairValidation":[41,8,1,"_CPPv2N6paddle16PnpairValidationE"],"paddle::PnpairValidation::PnpairValidation":[41,6,1,"_CPPv2N6paddle16PnpairValidation16PnpairValidationERK11LayerConfig"],"paddle::PnpairValidation::init":[41,6,1,"_CPPv2N6paddle16PnpairValidation4initERK8LayerMapRK12ParameterMap"],"paddle::PnpairValidation::onPassEnd":[41,6,1,"_CPPv2N6paddle16PnpairValidation9onPassEndEv"],"paddle::PnpairValidation::validationImp":[41,6,1,"_CPPv2N6paddle16PnpairValidation13validationImpE9MatrixPtr10IVectorPtr"],"paddle::PoolAllocator":[46,8,1,"_CPPv2N6paddle13PoolAllocatorE"],"paddle::PoolAllocator::PoolAllocator":[46,6,1,"_CPPv2N6paddle13PoolAllocator13PoolAllocatorEP9Allocator6size_tRKNSt6stringE"],"paddle::PoolAllocator::alloc":[46,6,1,"_CPPv2N6paddle13PoolAllocator5allocE6size_t"],"paddle::PoolAllocator::allocator_":[46,7,1,"_CPPv2N6paddle13PoolAllocator10allocator_E"],"paddle::PoolAllocator::free":[46,6,1,"_CPPv2N6paddle13PoolAllocator4freeEPv6size_t"],"paddle::PoolAllocator::freeAll":[46,6,1,"_CPPv2N6paddle13PoolAllocator7freeAllEv"],"paddle::PoolAllocator::getName":[46,6,1,"_CPPv2N6paddle13PoolAllocator7getNameEv"],"paddle::PoolAllocator::mutex_":[46,7,1,"_CPPv2N6paddle13PoolAllocator6mutex_E"],"paddle::PoolAllocator::name_":[46,7,1,"_CPPv2N6paddle13PoolAllocator5name_E"],"paddle::PoolAllocator::poolMemorySize_":[46,7,1,"_CPPv2N6paddle13PoolAllocator15poolMemorySize_E"],"paddle::PoolAllocator::pool_":[46,7,1,"_CPPv2N6paddle13PoolAllocator5pool_E"],"paddle::PoolAllocator::printAll":[46,6,1,"_CPPv2N6paddle13PoolAllocator8printAllEv"],"paddle::PoolAllocator::sizeLimit_":[46,7,1,"_CPPv2N6paddle13PoolAllocator10sizeLimit_E"],"paddle::PoolAllocator::~PoolAllocator":[46,6,1,"_CPPv2N6paddle13PoolAllocatorD0Ev"],"paddle::PoolLayer":[41,8,1,"_CPPv2N6paddle9PoolLayerE"],"paddle::PoolLayer::PoolLayer":[41,6,1,"_CPPv2N6paddle9PoolLayer9PoolLayerERK11LayerConfig"],"paddle::PoolLayer::channels_":[41,7,1,"_CPPv2N6paddle9PoolLayer9channels_E"],"paddle::PoolLayer::confPaddingY_":[41,7,1,"_CPPv2N6paddle9PoolLayer13confPaddingY_E"],"paddle::PoolLayer::confPadding_":[41,7,1,"_CPPv2N6paddle9PoolLayer12confPadding_E"],"paddle::PoolLayer::create":[41,6,1,"_CPPv2N6paddle9PoolLayer6createERK11LayerConfig"],"paddle::PoolLayer::imgSizeY_":[41,7,1,"_CPPv2N6paddle9PoolLayer9imgSizeY_E"],"paddle::PoolLayer::imgSize_":[41,7,1,"_CPPv2N6paddle9PoolLayer8imgSize_E"],"paddle::PoolLayer::init":[41,6,1,"_CPPv2N6paddle9PoolLayer4initERK8LayerMapRK12ParameterMap"],"paddle::PoolLayer::outputX_":[41,7,1,"_CPPv2N6paddle9PoolLayer8outputX_E"],"paddle::PoolLayer::outputY_":[41,7,1,"_CPPv2N6paddle9PoolLayer8outputY_E"],"paddle::PoolLayer::poolType_":[41,7,1,"_CPPv2N6paddle9PoolLayer9poolType_E"],"paddle::PoolLayer::sizeX_":[41,7,1,"_CPPv2N6paddle9PoolLayer6sizeX_E"],"paddle::PoolLayer::sizeY_":[41,7,1,"_CPPv2N6paddle9PoolLayer6sizeY_E"],"paddle::PoolLayer::start_":[41,7,1,"_CPPv2N6paddle9PoolLayer6start_E"],"paddle::PoolLayer::strideY_":[41,7,1,"_CPPv2N6paddle9PoolLayer8strideY_E"],"paddle::PoolLayer::stride_":[41,7,1,"_CPPv2N6paddle9PoolLayer7stride_E"],"paddle::PoolProjectionLayer":[41,8,1,"_CPPv2N6paddle19PoolProjectionLayerE"],"paddle::PoolProjectionLayer::PoolProjectionLayer":[41,6,1,"_CPPv2N6paddle19PoolProjectionLayer19PoolProjectionLayerERK11LayerConfig"],"paddle::PoolProjectionLayer::getSize":[41,6,1,"_CPPv2N6paddle19PoolProjectionLayer7getSizeEv"],"paddle::PoolProjectionLayer::imgSizeH_":[41,7,1,"_CPPv2N6paddle19PoolProjectionLayer9imgSizeH_E"],"paddle::PoolProjectionLayer::imgSizeW_":[41,7,1,"_CPPv2N6paddle19PoolProjectionLayer9imgSizeW_E"],"paddle::PoolProjectionLayer::outputH_":[41,7,1,"_CPPv2N6paddle19PoolProjectionLayer8outputH_E"],"paddle::PoolProjectionLayer::outputW_":[41,7,1,"_CPPv2N6paddle19PoolProjectionLayer8outputW_E"],"paddle::PowerLayer":[41,8,1,"_CPPv2N6paddle10PowerLayerE"],"paddle::PowerLayer::PowerLayer":[41,6,1,"_CPPv2N6paddle10PowerLayer10PowerLayerERK11LayerConfig"],"paddle::PowerLayer::backward":[41,6,1,"_CPPv2N6paddle10PowerLayer8backwardERK14UpdateCallback"],"paddle::PowerLayer::forward":[41,6,1,"_CPPv2N6paddle10PowerLayer7forwardE8PassType"],"paddle::PowerLayer::init":[41,6,1,"_CPPv2N6paddle10PowerLayer4initERK8LayerMapRK12ParameterMap"],"paddle::PowerLayer::tmpMtx":[41,7,1,"_CPPv2N6paddle10PowerLayer6tmpMtxE"],"paddle::PowerLayer::~PowerLayer":[41,6,1,"_CPPv2N6paddle10PowerLayerD0Ev"],"paddle::PrecisionRecallEvaluator":[36,8,1,"_CPPv2N6paddle24PrecisionRecallEvaluatorE"],"paddle::PrecisionRecallEvaluator::PrecisionRecallEvaluator":[36,6,1,"_CPPv2N6paddle24PrecisionRecallEvaluator24PrecisionRecallEvaluatorEv"],"paddle::PrecisionRecallEvaluator::StatsInfo":[36,8,1,"_CPPv2N6paddle24PrecisionRecallEvaluator9StatsInfoE"],"paddle::PrecisionRecallEvaluator::StatsInfo::FN":[36,7,1,"_CPPv2N6paddle24PrecisionRecallEvaluator9StatsInfo2FNE"],"paddle::PrecisionRecallEvaluator::StatsInfo::FP":[36,7,1,"_CPPv2N6paddle24PrecisionRecallEvaluator9StatsInfo2FPE"],"paddle::PrecisionRecallEvaluator::StatsInfo::StatsInfo":[36,6,1,"_CPPv2N6paddle24PrecisionRecallEvaluator9StatsInfo9StatsInfoEv"],"paddle::PrecisionRecallEvaluator::StatsInfo::TN":[36,7,1,"_CPPv2N6paddle24PrecisionRecallEvaluator9StatsInfo2TNE"],"paddle::PrecisionRecallEvaluator::StatsInfo::TP":[36,7,1,"_CPPv2N6paddle24PrecisionRecallEvaluator9StatsInfo2TPE"],"paddle::PrecisionRecallEvaluator::distributeEval":[36,6,1,"_CPPv2N6paddle24PrecisionRecallEvaluator14distributeEvalEP16ParameterClient2"],"paddle::PrecisionRecallEvaluator::evalImp":[36,6,1,"_CPPv2N6paddle24PrecisionRecallEvaluator7evalImpERNSt6vectorI8ArgumentEE"],"paddle::PrecisionRecallEvaluator::printStats":[36,6,1,"_CPPv2N6paddle24PrecisionRecallEvaluator10printStatsERNSt7ostreamE"],"paddle::PrecisionRecallEvaluator::start":[36,6,1,"_CPPv2N6paddle24PrecisionRecallEvaluator5startEv"],"paddle::Projection":[41,8,1,"_CPPv2N6paddle10ProjectionE"],"paddle::Projection::Projection":[41,6,1,"_CPPv2N6paddle10Projection10ProjectionERK16ProjectionConfig12ParameterPtrb"],"paddle::Projection::backward":[41,6,1,"_CPPv2N6paddle10Projection8backwardERK14UpdateCallback"],"paddle::Projection::config_":[41,7,1,"_CPPv2N6paddle10Projection7config_E"],"paddle::Projection::create":[41,6,1,"_CPPv2N6paddle10Projection6createERK16ProjectionConfig12ParameterPtrb"],"paddle::Projection::forward":[41,6,1,"_CPPv2N6paddle10Projection7forwardEPK8ArgumentPK8Argument8PassType"],"paddle::Projection::getName":[41,6,1,"_CPPv2N6paddle10Projection7getNameEv"],"paddle::Projection::getOutputSize":[41,6,1,"_CPPv2N6paddle10Projection13getOutputSizeEv"],"paddle::Projection::getState":[41,6,1,"_CPPv2N6paddle10Projection8getStateEv"],"paddle::Projection::in_":[41,7,1,"_CPPv2N6paddle10Projection3in_E"],"paddle::Projection::out_":[41,7,1,"_CPPv2N6paddle10Projection4out_E"],"paddle::Projection::parameter_":[41,7,1,"_CPPv2N6paddle10Projection10parameter_E"],"paddle::Projection::passType_":[41,7,1,"_CPPv2N6paddle10Projection9passType_E"],"paddle::Projection::prefetch":[41,6,1,"_CPPv2N6paddle10Projection8prefetchEPK8Argument"],"paddle::Projection::registrar_":[41,7,1,"_CPPv2N6paddle10Projection10registrar_E"],"paddle::Projection::resetState":[41,6,1,"_CPPv2N6paddle10Projection10resetStateEv"],"paddle::Projection::setState":[41,6,1,"_CPPv2N6paddle10Projection8setStateE13LayerStatePtr"],"paddle::Projection::useGpu_":[41,7,1,"_CPPv2N6paddle10Projection7useGpu_E"],"paddle::Projection::~Projection":[41,6,1,"_CPPv2N6paddle10ProjectionD0Ev"],"paddle::ProtoDataProvider":[34,8,1,"_CPPv2N6paddle17ProtoDataProviderE"],"paddle::ProtoDataProvider::ProtoDataProvider":[34,6,1,"_CPPv2N6paddle17ProtoDataProvider17ProtoDataProviderERK10DataConfigbb"],"paddle::ProtoDataProvider::ProtoSlot":[34,8,1,"_CPPv2N6paddle17ProtoDataProvider9ProtoSlotE"],"paddle::ProtoDataProvider::ProtoSlot::denseData":[34,7,1,"_CPPv2N6paddle17ProtoDataProvider9ProtoSlot9denseDataE"],"paddle::ProtoDataProvider::ProtoSlot::dim":[34,7,1,"_CPPv2N6paddle17ProtoDataProvider9ProtoSlot3dimE"],"paddle::ProtoDataProvider::ProtoSlot::indexData":[34,7,1,"_CPPv2N6paddle17ProtoDataProvider9ProtoSlot9indexDataE"],"paddle::ProtoDataProvider::ProtoSlot::indices":[34,7,1,"_CPPv2N6paddle17ProtoDataProvider9ProtoSlot7indicesE"],"paddle::ProtoDataProvider::ProtoSlot::sparseFloatValueData":[34,7,1,"_CPPv2N6paddle17ProtoDataProvider9ProtoSlot20sparseFloatValueDataE"],"paddle::ProtoDataProvider::ProtoSlot::sparseNonValueData":[34,7,1,"_CPPv2N6paddle17ProtoDataProvider9ProtoSlot18sparseNonValueDataE"],"paddle::ProtoDataProvider::ProtoSlot::strData":[34,7,1,"_CPPv2N6paddle17ProtoDataProvider9ProtoSlot7strDataE"],"paddle::ProtoDataProvider::ProtoSlot::subIndices":[34,7,1,"_CPPv2N6paddle17ProtoDataProvider9ProtoSlot10subIndicesE"],"paddle::ProtoDataProvider::ProtoSlot::type":[34,7,1,"_CPPv2N6paddle17ProtoDataProvider9ProtoSlot4typeE"],"paddle::ProtoDataProvider::ProtoSlot::varDenseData":[34,7,1,"_CPPv2N6paddle17ProtoDataProvider9ProtoSlot12varDenseDataE"],"paddle::ProtoDataProvider::ProtoSlot::varIndices":[34,7,1,"_CPPv2N6paddle17ProtoDataProvider9ProtoSlot10varIndicesE"],"paddle::ProtoDataProvider::ProtoVarSlot":[34,8,1,"_CPPv2N6paddle17ProtoDataProvider12ProtoVarSlotE"],"paddle::ProtoDataProvider::ProtoVarSlot::data":[34,7,1,"_CPPv2N6paddle17ProtoDataProvider12ProtoVarSlot4dataE"],"paddle::ProtoDataProvider::ProtoVarSlot::dims":[34,7,1,"_CPPv2N6paddle17ProtoDataProvider12ProtoVarSlot4dimsE"],"paddle::ProtoDataProvider::checkDataHeader":[34,6,1,"_CPPv2N6paddle17ProtoDataProvider15checkDataHeaderERK10DataHeader"],"paddle::ProtoDataProvider::checkSample":[34,6,1,"_CPPv2N6paddle17ProtoDataProvider11checkSampleERK10DataSample"],"paddle::ProtoDataProvider::cpuBatch_":[34,7,1,"_CPPv2N6paddle17ProtoDataProvider9cpuBatch_E"],"paddle::ProtoDataProvider::currentSequenceIndex_":[34,7,1,"_CPPv2N6paddle17ProtoDataProvider21currentSequenceIndex_E"],"paddle::ProtoDataProvider::fillSlots":[34,6,1,"_CPPv2N6paddle17ProtoDataProvider9fillSlotsERK10DataSample"],"paddle::ProtoDataProvider::getNextBatchInternal":[34,6,1,"_CPPv2N6paddle17ProtoDataProvider20getNextBatchInternalE7int64_tP9DataBatch"],"paddle::ProtoDataProvider::getSize":[34,6,1,"_CPPv2N6paddle17ProtoDataProvider7getSizeEv"],"paddle::ProtoDataProvider::gpuBatch_":[34,7,1,"_CPPv2N6paddle17ProtoDataProvider9gpuBatch_E"],"paddle::ProtoDataProvider::header_":[34,7,1,"_CPPv2N6paddle17ProtoDataProvider7header_E"],"paddle::ProtoDataProvider::iidData":[34,6,1,"_CPPv2N6paddle17ProtoDataProvider7iidDataEv"],"paddle::ProtoDataProvider::loadData":[34,6,1,"_CPPv2N6paddle17ProtoDataProvider8loadDataERKNSt6vectorINSt6stringEEE"],"paddle::ProtoDataProvider::loadDataFile":[34,6,1,"_CPPv2N6paddle17ProtoDataProvider12loadDataFileERKNSt6stringE"],"paddle::ProtoDataProvider::lock_":[34,7,1,"_CPPv2N6paddle17ProtoDataProvider5lock_E"],"paddle::ProtoDataProvider::nnzStats_":[34,7,1,"_CPPv2N6paddle17ProtoDataProvider9nnzStats_E"],"paddle::ProtoDataProvider::numVecSlots_":[34,7,1,"_CPPv2N6paddle17ProtoDataProvider12numVecSlots_E"],"paddle::ProtoDataProvider::reset":[34,6,1,"_CPPv2N6paddle17ProtoDataProvider5resetEv"],"paddle::ProtoDataProvider::sampleLoop":[34,6,1,"_CPPv2N6paddle17ProtoDataProvider10sampleLoopE2Op7int64_t"],"paddle::ProtoDataProvider::sampleNums_":[34,7,1,"_CPPv2N6paddle17ProtoDataProvider11sampleNums_E"],"paddle::ProtoDataProvider::sequenceLoop":[34,6,1,"_CPPv2N6paddle17ProtoDataProvider12sequenceLoopE2Op7int64_t"],"paddle::ProtoDataProvider::sequenceStartPositions_":[34,7,1,"_CPPv2N6paddle17ProtoDataProvider23sequenceStartPositions_E"],"paddle::ProtoDataProvider::showDataStats":[34,6,1,"_CPPv2N6paddle17ProtoDataProvider13showDataStatsEv"],"paddle::ProtoDataProvider::shuffle":[34,6,1,"_CPPv2N6paddle17ProtoDataProvider7shuffleEv"],"paddle::ProtoDataProvider::shuffledSequenceIds_":[34,7,1,"_CPPv2N6paddle17ProtoDataProvider20shuffledSequenceIds_E"],"paddle::ProtoDataProvider::slots_":[34,7,1,"_CPPv2N6paddle17ProtoDataProvider6slots_E"],"paddle::ProtoDataProvider::subSampleLoop":[34,6,1,"_CPPv2N6paddle17ProtoDataProvider13subSampleLoopE2Op7int64_ti"],"paddle::ProtoSequenceDataProvider":[34,8,1,"_CPPv2N6paddle25ProtoSequenceDataProviderE"],"paddle::ProtoSequenceDataProvider::ProtoSequenceDataProvider":[34,6,1,"_CPPv2N6paddle25ProtoSequenceDataProvider25ProtoSequenceDataProviderERK10DataConfigbb"],"paddle::ProtoSequenceDataProvider::getNextBatchInternal":[34,6,1,"_CPPv2N6paddle25ProtoSequenceDataProvider20getNextBatchInternalE7int64_tP9DataBatch"],"paddle::ProtoSequenceDataProvider::~ProtoSequenceDataProvider":[34,6,1,"_CPPv2N6paddle25ProtoSequenceDataProviderD0Ev"],"paddle::ProtoServer":[58,8,1,"_CPPv2N6paddle11ProtoServerE"],"paddle::ProtoServer::ProtoResponseCallback":[58,9,1,"_CPPv2N6paddle11ProtoServer21ProtoResponseCallbackE"],"paddle::ProtoServer::ProtoResponseCallbackEx":[58,9,1,"_CPPv2N6paddle11ProtoServer23ProtoResponseCallbackExE"],"paddle::ProtoServer::ProtoServer":[58,6,1,"_CPPv2N6paddle11ProtoServer11ProtoServerERKNSt6stringEii"],"paddle::ProtoServer::ServiceFunction":[58,9,1,"_CPPv2N6paddle11ProtoServer15ServiceFunctionE"],"paddle::ProtoServer::handleRequest":[58,6,1,"_CPPv2N6paddle11ProtoServer13handleRequestENSt10unique_ptrI9MsgReaderEE16ResponseCallback"],"paddle::ProtoServer::handleRequestBegin_":[58,7,1,"_CPPv2N6paddle11ProtoServer19handleRequestBegin_E"],"paddle::ProtoServer::nameToFuncMap_":[58,7,1,"_CPPv2N6paddle11ProtoServer14nameToFuncMap_E"],"paddle::ProtoServer::registerServiceFunction":[58,6,1,"_CPPv2N6paddle11ProtoServer23registerServiceFunctionERKNSt6stringENSt8functionIXvoid(const ProtoIn &request,                                                                                                                                                                                                                                                                                                                                                                   ProtoResponseCallback callback)EEE"],"paddle::ProtoServer::registerServiceFunctionEx":[58,6,1,"_CPPv2N6paddle11ProtoServer25registerServiceFunctionExERKNSt6stringENSt8functionIXvoid(const ProtoIn &, std::unique_ptr<  MsgReader  > msgReader,                                                                                                                                                                                                                                                                                                                                                                   ProtoResponseCallbackEx callback)EEE"],"paddle::ProtoServer::registerServiceFunctionImp":[58,6,1,"_CPPv2N6paddle11ProtoServer26registerServiceFunctionImpERKNSt6stringE15ServiceFunction"],"paddle::PyDataProvider2":[34,8,1,"_CPPv2N6paddle15PyDataProvider2E"],"paddle::PyDataProvider2::PyDataProvider2":[34,6,1,"_CPPv2N6paddle15PyDataProvider215PyDataProvider2ERK10DataConfigb"],"paddle::PyDataProvider2::getNextBatchInternal":[34,6,1,"_CPPv2N6paddle15PyDataProvider220getNextBatchInternalE7int64_tP9DataBatch"],"paddle::PyDataProvider2::getSize":[34,6,1,"_CPPv2N6paddle15PyDataProvider27getSizeEv"],"paddle::PyDataProvider2::reset":[34,6,1,"_CPPv2N6paddle15PyDataProvider25resetEv"],"paddle::PyDataProvider2::shuffle":[34,6,1,"_CPPv2N6paddle15PyDataProvider27shuffleEv"],"paddle::PyDataProvider2::~PyDataProvider2":[34,6,1,"_CPPv2N6paddle15PyDataProvider2D0Ev"],"paddle::Queue":[63,8,1,"_CPPv2N6paddle5QueueE"],"paddle::Queue::Queue":[63,6,1,"_CPPv2N6paddle5Queue5QueueEv"],"paddle::Queue::dequeue":[63,6,1,"_CPPv2N6paddle5Queue7dequeueEv"],"paddle::Queue::empty":[63,6,1,"_CPPv2N6paddle5Queue5emptyEv"],"paddle::Queue::enqueue":[63,6,1,"_CPPv2N6paddle5Queue7enqueueERK1T"],"paddle::Queue::size":[63,6,1,"_CPPv2N6paddle5Queue4sizeEv"],"paddle::Queue::waitEmpty":[63,6,1,"_CPPv2N6paddle5Queue9waitEmptyEv"],"paddle::Queue::~Queue":[63,6,1,"_CPPv2N6paddle5QueueD0Ev"],"paddle::RMSPropParameterOptimizer":[48,8,1,"_CPPv2N6paddle25RMSPropParameterOptimizerE"],"paddle::RMSPropParameterOptimizer::RMSPropParameterOptimizer":[48,6,1,"_CPPv2N6paddle25RMSPropParameterOptimizer25RMSPropParameterOptimizerERK18OptimizationConfig"],"paddle::RMSPropParameterOptimizer::epsilon_":[48,7,1,"_CPPv2N6paddle25RMSPropParameterOptimizer8epsilon_E"],"paddle::RMSPropParameterOptimizer::finishBatch":[48,6,1,"_CPPv2N6paddle25RMSPropParameterOptimizer11finishBatchEv"],"paddle::RMSPropParameterOptimizer::init":[48,6,1,"_CPPv2N6paddle25RMSPropParameterOptimizer4initE6size_tPK15ParameterConfig"],"paddle::RMSPropParameterOptimizer::rou_":[48,7,1,"_CPPv2N6paddle25RMSPropParameterOptimizer4rou_E"],"paddle::RMSPropParameterOptimizer::startBatch":[48,6,1,"_CPPv2N6paddle25RMSPropParameterOptimizer10startBatchE7int64_t"],"paddle::RMSPropParameterOptimizer::t0Vec_":[48,7,1,"_CPPv2N6paddle25RMSPropParameterOptimizer6t0Vec_E"],"paddle::RMSPropParameterOptimizer::timer_":[48,7,1,"_CPPv2N6paddle25RMSPropParameterOptimizer6timer_E"],"paddle::RMSPropParameterOptimizer::update":[48,6,1,"_CPPv2N6paddle25RMSPropParameterOptimizer6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::RWLock":[64,8,1,"_CPPv2N6paddle6RWLockE"],"paddle::RWLock::RWLock":[64,6,1,"_CPPv2N6paddle6RWLock6RWLockEv"],"paddle::RWLock::lock":[64,6,1,"_CPPv2N6paddle6RWLock4lockEv"],"paddle::RWLock::lock_shared":[64,6,1,"_CPPv2N6paddle6RWLock11lock_sharedEv"],"paddle::RWLock::operator=":[64,6,1,"_CPPv2N6paddle6RWLockaSERK6RWLock"],"paddle::RWLock::rwlock_":[64,7,1,"_CPPv2N6paddle6RWLock7rwlock_E"],"paddle::RWLock::unlock":[64,6,1,"_CPPv2N6paddle6RWLock6unlockEv"],"paddle::RWLock::~RWLock":[64,6,1,"_CPPv2N6paddle6RWLockD0Ev"],"paddle::RankAucEvaluator":[36,8,1,"_CPPv2N6paddle16RankAucEvaluatorE"],"paddle::RankAucEvaluator::distributeEval":[36,6,1,"_CPPv2N6paddle16RankAucEvaluator14distributeEvalEP16ParameterClient2"],"paddle::RankAucEvaluator::evalImp":[36,6,1,"_CPPv2N6paddle16RankAucEvaluator7evalImpERNSt6vectorI8ArgumentEE"],"paddle::RankAucEvaluator::start":[36,6,1,"_CPPv2N6paddle16RankAucEvaluator5startEv"],"paddle::RankAucEvaluator::updateSamplesNum":[36,6,1,"_CPPv2N6paddle16RankAucEvaluator16updateSamplesNumERKNSt6vectorI8ArgumentEE"],"paddle::RankingCost":[41,8,1,"_CPPv2N6paddle11RankingCostE"],"paddle::RankingCost::RankingCost":[41,6,1,"_CPPv2N6paddle11RankingCost11RankingCostERK11LayerConfig"],"paddle::RankingCost::backward":[41,6,1,"_CPPv2N6paddle11RankingCost8backwardERK14UpdateCallback"],"paddle::RankingCost::backwardImp":[41,6,1,"_CPPv2N6paddle11RankingCost11backwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::RankingCost::forward":[41,6,1,"_CPPv2N6paddle11RankingCost7forwardE8PassType"],"paddle::RankingCost::forwardImp":[41,6,1,"_CPPv2N6paddle11RankingCost10forwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::RankingCost::getLabelLayer":[41,6,1,"_CPPv2N6paddle11RankingCost13getLabelLayerEv"],"paddle::RankingCost::getOutputLayer":[41,6,1,"_CPPv2N6paddle11RankingCost14getOutputLayerE6size_t"],"paddle::RankingCost::init":[41,6,1,"_CPPv2N6paddle11RankingCost4initERK8LayerMapRK12ParameterMap"],"paddle::RankingCost::onPassEnd":[41,6,1,"_CPPv2N6paddle11RankingCost9onPassEndEv"],"paddle::ReadLockGuard":[64,8,1,"_CPPv2N6paddle13ReadLockGuardE"],"paddle::ReadLockGuard::ReadLockGuard":[64,6,1,"_CPPv2N6paddle13ReadLockGuard13ReadLockGuardER6RWLock"],"paddle::ReadLockGuard::rwlock_":[64,7,1,"_CPPv2N6paddle13ReadLockGuard7rwlock_E"],"paddle::ReadLockGuard::~ReadLockGuard":[64,6,1,"_CPPv2N6paddle13ReadLockGuardD0Ev"],"paddle::RecurrentGradientMachine":[38,8,1,"_CPPv2N6paddle24RecurrentGradientMachineE"],"paddle::RecurrentGradientMachine::BeamSearchCandidatesAdjustCallback":[38,9,1,"_CPPv2N6paddle24RecurrentGradientMachine34BeamSearchCandidatesAdjustCallbackE"],"paddle::RecurrentGradientMachine::DropCallback":[38,9,1,"_CPPv2N6paddle24RecurrentGradientMachine12DropCallbackE"],"paddle::RecurrentGradientMachine::EachStepCallback":[38,9,1,"_CPPv2N6paddle24RecurrentGradientMachine16EachStepCallbackE"],"paddle::RecurrentGradientMachine::EosFrameLine":[38,8,1,"_CPPv2N6paddle24RecurrentGradientMachine12EosFrameLineE"],"paddle::RecurrentGradientMachine::EosFrameLine::layers":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine12EosFrameLine6layersE"],"paddle::RecurrentGradientMachine::Generator":[38,8,1,"_CPPv2N6paddle24RecurrentGradientMachine9GeneratorE"],"paddle::RecurrentGradientMachine::Generator::config":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine9Generator6configE"],"paddle::RecurrentGradientMachine::Generator::ids":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine9Generator3idsE"],"paddle::RecurrentGradientMachine::Generator::outArg":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine9Generator6outArgE"],"paddle::RecurrentGradientMachine::InFrameLine":[38,8,1,"_CPPv2N6paddle24RecurrentGradientMachine11InFrameLineE"],"paddle::RecurrentGradientMachine::InFrameLine::agents":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine11InFrameLine6agentsE"],"paddle::RecurrentGradientMachine::InFrameLine::hasSubseq":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine11InFrameLine9hasSubseqE"],"paddle::RecurrentGradientMachine::InFrameLine::inLayer":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine11InFrameLine7inLayerE"],"paddle::RecurrentGradientMachine::InFrameLine::linkName":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine11InFrameLine8linkNameE"],"paddle::RecurrentGradientMachine::InFrameLine::outArg":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine11InFrameLine6outArgE"],"paddle::RecurrentGradientMachine::Info":[38,8,1,"_CPPv2N6paddle24RecurrentGradientMachine4InfoE"],"paddle::RecurrentGradientMachine::Info::allIds":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine4Info6allIdsE"],"paddle::RecurrentGradientMachine::Info::idIndex":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine4Info7idIndexE"],"paddle::RecurrentGradientMachine::Info::seqStartPosIndex":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine4Info16seqStartPosIndexE"],"paddle::RecurrentGradientMachine::Info::sequenceStartPositions":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine4Info22sequenceStartPositionsE"],"paddle::RecurrentGradientMachine::MemoryFrameLine":[38,8,1,"_CPPv2N6paddle24RecurrentGradientMachine15MemoryFrameLineE"],"paddle::RecurrentGradientMachine::MemoryFrameLine::agents":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine15MemoryFrameLine6agentsE"],"paddle::RecurrentGradientMachine::MemoryFrameLine::allIds":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine15MemoryFrameLine6allIdsE"],"paddle::RecurrentGradientMachine::MemoryFrameLine::biasLayer":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine15MemoryFrameLine9biasLayerE"],"paddle::RecurrentGradientMachine::MemoryFrameLine::bootLayer":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine15MemoryFrameLine9bootLayerE"],"paddle::RecurrentGradientMachine::MemoryFrameLine::frames":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine15MemoryFrameLine6framesE"],"paddle::RecurrentGradientMachine::MemoryFrameLine::is_sequence":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine15MemoryFrameLine11is_sequenceE"],"paddle::RecurrentGradientMachine::MemoryFrameLine::layerName":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine15MemoryFrameLine9layerNameE"],"paddle::RecurrentGradientMachine::MemoryFrameLine::linkName":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine15MemoryFrameLine8linkNameE"],"paddle::RecurrentGradientMachine::MemoryFrameLine::outArg":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine15MemoryFrameLine6outArgE"],"paddle::RecurrentGradientMachine::MemoryFrameLine::rootAgent":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine15MemoryFrameLine9rootAgentE"],"paddle::RecurrentGradientMachine::MemoryFrameLine::rootLayer":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine15MemoryFrameLine9rootLayerE"],"paddle::RecurrentGradientMachine::MemoryFrameLine::scatterAgents":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine15MemoryFrameLine13scatterAgentsE"],"paddle::RecurrentGradientMachine::MemoryFrameLine::sequenceStartPositions":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine15MemoryFrameLine22sequenceStartPositionsE"],"paddle::RecurrentGradientMachine::NormOrDropNodeCallback":[38,9,1,"_CPPv2N6paddle24RecurrentGradientMachine22NormOrDropNodeCallbackE"],"paddle::RecurrentGradientMachine::OutFrameLine":[38,8,1,"_CPPv2N6paddle24RecurrentGradientMachine12OutFrameLineE"],"paddle::RecurrentGradientMachine::OutFrameLine::agentLayer":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine12OutFrameLine10agentLayerE"],"paddle::RecurrentGradientMachine::OutFrameLine::frames":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine12OutFrameLine6framesE"],"paddle::RecurrentGradientMachine::OutFrameLine::layerName":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine12OutFrameLine9layerNameE"],"paddle::RecurrentGradientMachine::Path":[38,8,1,"_CPPv2N6paddle24RecurrentGradientMachine4PathE"],"paddle::RecurrentGradientMachine::Path::Path":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine4Path4PathEv"],"paddle::RecurrentGradientMachine::Path::adjustProb":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine4Path10adjustProbEib"],"paddle::RecurrentGradientMachine::Path::greaterPath":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine4Path11greaterPathERK4PathRK4Path"],"paddle::RecurrentGradientMachine::Path::ids":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine4Path3idsE"],"paddle::RecurrentGradientMachine::Path::isDropable":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine4Path10isDropableEv"],"paddle::RecurrentGradientMachine::Path::logProb":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine4Path7logProbE"],"paddle::RecurrentGradientMachine::Path::machineId":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine4Path9machineIdE"],"paddle::RecurrentGradientMachine::Path::machineIdVec":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine4Path12machineIdVecE"],"paddle::RecurrentGradientMachine::Path::operator<":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine4PathltERK4Path"],"paddle::RecurrentGradientMachine::Path::probHistory":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine4Path11probHistoryE"],"paddle::RecurrentGradientMachine::Path::recordHistory":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine4Path13recordHistoryEv"],"paddle::RecurrentGradientMachine::Path::seqId":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine4Path5seqIdE"],"paddle::RecurrentGradientMachine::Path::topIndex":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine4Path8topIndexE"],"paddle::RecurrentGradientMachine::RecurrentGradientMachine":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine24RecurrentGradientMachineERKNSt6stringEP13NeuralNetwork"],"paddle::RecurrentGradientMachine::backward":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine8backwardERK14UpdateCallback"],"paddle::RecurrentGradientMachine::beamSearch":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine10beamSearchE6size_t"],"paddle::RecurrentGradientMachine::copyScattedId":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine13copyScattedIdERNSt6vectorIiEEP10IVectorPtri"],"paddle::RecurrentGradientMachine::cpuEos_":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine7cpuEos_E"],"paddle::RecurrentGradientMachine::cpuId_":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine6cpuId_E"],"paddle::RecurrentGradientMachine::cpuProb_":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine8cpuProb_E"],"paddle::RecurrentGradientMachine::createInFrameInfo":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine17createInFrameInfoERK8Argument8PassType"],"paddle::RecurrentGradientMachine::createMemoryFrameInfo":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine21createMemoryFrameInfoEP15MemoryFrameLine8PassType"],"paddle::RecurrentGradientMachine::createSeqPos":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine12createSeqPosERKNSt6vectorIiEEP16ICpuGpuVectorPtr"],"paddle::RecurrentGradientMachine::dataArgsFrame_":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine14dataArgsFrame_E"],"paddle::RecurrentGradientMachine::dataArgsSize_":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine13dataArgsSize_E"],"paddle::RecurrentGradientMachine::dataArgs_":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine9dataArgs_E"],"paddle::RecurrentGradientMachine::eosFrameLine_":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine13eosFrameLine_E"],"paddle::RecurrentGradientMachine::eval":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine4evalEP9Evaluator"],"paddle::RecurrentGradientMachine::evaluator_":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine10evaluator_E"],"paddle::RecurrentGradientMachine::forward":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine7forwardERKNSt6vectorI8ArgumentEEPNSt6vectorI8ArgumentEE8PassType"],"paddle::RecurrentGradientMachine::forwardBackward":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine15forwardBackwardERKNSt6vectorI8ArgumentEEPNSt6vectorI8ArgumentEE8PassTypeRK14UpdateCallback"],"paddle::RecurrentGradientMachine::frames_":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine7frames_E"],"paddle::RecurrentGradientMachine::generateSequence":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine16generateSequenceEv"],"paddle::RecurrentGradientMachine::generator_":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine10generator_E"],"paddle::RecurrentGradientMachine::getFinalPaths":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine13getFinalPathsEv"],"paddle::RecurrentGradientMachine::getParameterIds":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine15getParameterIdsEv"],"paddle::RecurrentGradientMachine::inFrameLines_":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine13inFrameLines_E"],"paddle::RecurrentGradientMachine::info_":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine5info_E"],"paddle::RecurrentGradientMachine::init":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine4initERK11ModelConfig17ParamInitCallbackRKNSt6vectorI13ParameterTypeEEb"],"paddle::RecurrentGradientMachine::maxSequenceLength_":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine18maxSequenceLength_E"],"paddle::RecurrentGradientMachine::memoryFrameLines_":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine17memoryFrameLines_E"],"paddle::RecurrentGradientMachine::oneWaySearch":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine12oneWaySearchE6size_t"],"paddle::RecurrentGradientMachine::operator=":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachineaSERK24RecurrentGradientMachine"],"paddle::RecurrentGradientMachine::outFrameLines_":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine14outFrameLines_E"],"paddle::RecurrentGradientMachine::parameterIds_":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine13parameterIds_E"],"paddle::RecurrentGradientMachine::prefetch":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine8prefetchERKNSt6vectorI8ArgumentEE"],"paddle::RecurrentGradientMachine::registerBeamSearchControlCallbacks":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine34registerBeamSearchControlCallbacksERK34BeamSearchCandidatesAdjustCallbackRK22NormOrDropNodeCallbackRK12DropCallback"],"paddle::RecurrentGradientMachine::registerBeamSearchStatisticsCallbacks":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine37registerBeamSearchStatisticsCallbacksERK16EachStepCallbackRK16EachStepCallback"],"paddle::RecurrentGradientMachine::removeBeamSearchControlCallbacks":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine32removeBeamSearchControlCallbacksEv"],"paddle::RecurrentGradientMachine::removeBeamSearchStatisticsCallbacks":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine35removeBeamSearchStatisticsCallbacksEv"],"paddle::RecurrentGradientMachine::resetState":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine10resetStateEv"],"paddle::RecurrentGradientMachine::resizeBootFrame":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine15resizeBootFrameEi"],"paddle::RecurrentGradientMachine::resizeOrCreateFrames":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine20resizeOrCreateFramesEi"],"paddle::RecurrentGradientMachine::reversed_":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine9reversed_E"],"paddle::RecurrentGradientMachine::rootNetwork_":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine12rootNetwork_E"],"paddle::RecurrentGradientMachine::selectRowsOneTime":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine17selectRowsOneTimeE8LayerPtrRK10IVectorPtrP8Argument8PassType"],"paddle::RecurrentGradientMachine::seqLengthAndStart_":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine18seqLengthAndStart_E"],"paddle::RecurrentGradientMachine::stopBeamSearch":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachine14stopBeamSearchEv"],"paddle::RecurrentGradientMachine::stopBeamSearch_":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine15stopBeamSearch_E"],"paddle::RecurrentGradientMachine::useGpu_":[38,7,1,"_CPPv2N6paddle24RecurrentGradientMachine7useGpu_E"],"paddle::RecurrentGradientMachine::~RecurrentGradientMachine":[38,6,1,"_CPPv2N6paddle24RecurrentGradientMachineD0Ev"],"paddle::RecurrentLayer":[41,8,1,"_CPPv2N6paddle14RecurrentLayerE"],"paddle::RecurrentLayer::RecurrentLayer":[41,6,1,"_CPPv2N6paddle14RecurrentLayer14RecurrentLayerERK11LayerConfig"],"paddle::RecurrentLayer::backward":[41,6,1,"_CPPv2N6paddle14RecurrentLayer8backwardERK14UpdateCallback"],"paddle::RecurrentLayer::backwardBatch":[41,6,1,"_CPPv2N6paddle14RecurrentLayer13backwardBatchEi6size_tPKi"],"paddle::RecurrentLayer::backwardOneSequence":[41,6,1,"_CPPv2N6paddle14RecurrentLayer19backwardOneSequenceEii"],"paddle::RecurrentLayer::backwardSequence":[41,6,1,"_CPPv2N6paddle14RecurrentLayer16backwardSequenceEi6size_tPKi"],"paddle::RecurrentLayer::batchGrad_":[41,7,1,"_CPPv2N6paddle14RecurrentLayer10batchGrad_E"],"paddle::RecurrentLayer::batchValue_":[41,7,1,"_CPPv2N6paddle14RecurrentLayer11batchValue_E"],"paddle::RecurrentLayer::bias_":[41,7,1,"_CPPv2N6paddle14RecurrentLayer5bias_E"],"paddle::RecurrentLayer::forward":[41,6,1,"_CPPv2N6paddle14RecurrentLayer7forwardE8PassType"],"paddle::RecurrentLayer::forwardBatch":[41,6,1,"_CPPv2N6paddle14RecurrentLayer12forwardBatchEi6size_tPKi"],"paddle::RecurrentLayer::forwardOneSequence":[41,6,1,"_CPPv2N6paddle14RecurrentLayer18forwardOneSequenceEii"],"paddle::RecurrentLayer::forwardSequence":[41,6,1,"_CPPv2N6paddle14RecurrentLayer15forwardSequenceEi6size_tPKi"],"paddle::RecurrentLayer::frameOutput_":[41,7,1,"_CPPv2N6paddle14RecurrentLayer12frameOutput_E"],"paddle::RecurrentLayer::getState":[41,6,1,"_CPPv2N6paddle14RecurrentLayer8getStateEv"],"paddle::RecurrentLayer::init":[41,6,1,"_CPPv2N6paddle14RecurrentLayer4initERK8LayerMapRK12ParameterMap"],"paddle::RecurrentLayer::prevOutput_":[41,7,1,"_CPPv2N6paddle14RecurrentLayer11prevOutput_E"],"paddle::RecurrentLayer::resetState":[41,6,1,"_CPPv2N6paddle14RecurrentLayer10resetStateEv"],"paddle::RecurrentLayer::reversed_":[41,7,1,"_CPPv2N6paddle14RecurrentLayer9reversed_E"],"paddle::RecurrentLayer::setState":[41,6,1,"_CPPv2N6paddle14RecurrentLayer8setStateE13LayerStatePtr"],"paddle::RecurrentLayer::weight_":[41,7,1,"_CPPv2N6paddle14RecurrentLayer7weight_E"],"paddle::Regularizer":[50,8,1,"_CPPv2N6paddle11RegularizerE"],"paddle::Regularizer::get":[50,6,1,"_CPPv2N6paddle11Regularizer3getERKNSt6vectorI13ParameterTypeEERK15ParameterConfig"],"paddle::Regularizer::update":[50,6,1,"_CPPv2N6paddle11Regularizer6updateEA_K9VectorPtrRK15ParameterConfig4realii"],"paddle::Regularizer::~Regularizer":[50,6,1,"_CPPv2N6paddle11RegularizerD0Ev"],"paddle::RemoteParameterUpdater":[59,8,1,"_CPPv2N6paddle22RemoteParameterUpdaterE"],"paddle::RemoteParameterUpdater::RemoteParameterUpdater":[59,6,1,"_CPPv2N6paddle22RemoteParameterUpdater22RemoteParameterUpdaterERK18OptimizationConfigiRRNSt10unique_ptrI16ParameterUpdaterEE"],"paddle::RemoteParameterUpdater::apply":[59,6,1,"_CPPv2N6paddle22RemoteParameterUpdater5applyEv"],"paddle::RemoteParameterUpdater::batchSize_":[59,7,1,"_CPPv2N6paddle22RemoteParameterUpdater10batchSize_E"],"paddle::RemoteParameterUpdater::batchStatus_":[59,7,1,"_CPPv2N6paddle22RemoteParameterUpdater12batchStatus_E"],"paddle::RemoteParameterUpdater::config_":[59,7,1,"_CPPv2N6paddle22RemoteParameterUpdater7config_E"],"paddle::RemoteParameterUpdater::controller":[59,6,1,"_CPPv2N6paddle22RemoteParameterUpdater10controllerEv"],"paddle::RemoteParameterUpdater::controllerThread_":[59,7,1,"_CPPv2N6paddle22RemoteParameterUpdater17controllerThread_E"],"paddle::RemoteParameterUpdater::copyParametersFromDevice":[59,6,1,"_CPPv2N6paddle22RemoteParameterUpdater24copyParametersFromDeviceE13ParameterType"],"paddle::RemoteParameterUpdater::copyParametersToDevice":[59,6,1,"_CPPv2N6paddle22RemoteParameterUpdater22copyParametersToDeviceE13ParameterType"],"paddle::RemoteParameterUpdater::cpuParameters_":[59,7,1,"_CPPv2N6paddle22RemoteParameterUpdater14cpuParameters_E"],"paddle::RemoteParameterUpdater::expectedPassCount_":[59,7,1,"_CPPv2N6paddle22RemoteParameterUpdater18expectedPassCount_E"],"paddle::RemoteParameterUpdater::finishBatch":[59,6,1,"_CPPv2N6paddle22RemoteParameterUpdater11finishBatchE4real"],"paddle::RemoteParameterUpdater::finishPass":[59,6,1,"_CPPv2N6paddle22RemoteParameterUpdater10finishPassE4real"],"paddle::RemoteParameterUpdater::init":[59,6,1,"_CPPv2N6paddle22RemoteParameterUpdater4initERNSt6vectorI12ParameterPtrEE"],"paddle::RemoteParameterUpdater::isFirstPass_":[59,7,1,"_CPPv2N6paddle22RemoteParameterUpdater12isFirstPass_E"],"paddle::RemoteParameterUpdater::kAverage":[59,7,1,"_CPPv2N6paddle22RemoteParameterUpdater8kAverageE"],"paddle::RemoteParameterUpdater::kElasticAverage":[59,7,1,"_CPPv2N6paddle22RemoteParameterUpdater15kElasticAverageE"],"paddle::RemoteParameterUpdater::localUpdater_":[59,7,1,"_CPPv2N6paddle22RemoteParameterUpdater13localUpdater_E"],"paddle::RemoteParameterUpdater::numBatches_":[59,7,1,"_CPPv2N6paddle22RemoteParameterUpdater11numBatches_E"],"paddle::RemoteParameterUpdater::parameterClient_":[59,7,1,"_CPPv2N6paddle22RemoteParameterUpdater16parameterClient_E"],"paddle::RemoteParameterUpdater::passCount_":[59,7,1,"_CPPv2N6paddle22RemoteParameterUpdater10passCount_E"],"paddle::RemoteParameterUpdater::restore":[59,6,1,"_CPPv2N6paddle22RemoteParameterUpdater7restoreEv"],"paddle::RemoteParameterUpdater::separateSendAndRecv_":[59,7,1,"_CPPv2N6paddle22RemoteParameterUpdater20separateSendAndRecv_E"],"paddle::RemoteParameterUpdater::setForwardbackwardTime":[59,6,1,"_CPPv2N6paddle22RemoteParameterUpdater22setForwardbackwardTimeE8uint64_t"],"paddle::RemoteParameterUpdater::startBatch":[59,6,1,"_CPPv2N6paddle22RemoteParameterUpdater10startBatchE7int64_t"],"paddle::RemoteParameterUpdater::startController":[59,6,1,"_CPPv2N6paddle22RemoteParameterUpdater15startControllerEv"],"paddle::RemoteParameterUpdater::startPass":[59,6,1,"_CPPv2N6paddle22RemoteParameterUpdater9startPassEv"],"paddle::RemoteParameterUpdater::updateImpl":[59,6,1,"_CPPv2N6paddle22RemoteParameterUpdater10updateImplEP9Parameter"],"paddle::RemoteParameterUpdater::useApplyInPserver_":[59,7,1,"_CPPv2N6paddle22RemoteParameterUpdater18useApplyInPserver_E"],"paddle::RemoteParameterUpdater::~RemoteParameterUpdater":[59,6,1,"_CPPv2N6paddle22RemoteParameterUpdaterD0Ev"],"paddle::ResizeLayer":[41,8,1,"_CPPv2N6paddle11ResizeLayerE"],"paddle::ResizeLayer::ResizeLayer":[41,6,1,"_CPPv2N6paddle11ResizeLayer11ResizeLayerERK11LayerConfig"],"paddle::ResizeLayer::backward":[41,6,1,"_CPPv2N6paddle11ResizeLayer8backwardERK14UpdateCallback"],"paddle::ResizeLayer::forward":[41,6,1,"_CPPv2N6paddle11ResizeLayer7forwardE8PassType"],"paddle::ResizeLayer::init":[41,6,1,"_CPPv2N6paddle11ResizeLayer4initERK8LayerMapRK12ParameterMap"],"paddle::ResponseNormLayer":[41,8,1,"_CPPv2N6paddle17ResponseNormLayerE"],"paddle::ResponseNormLayer::ResponseNormLayer":[41,6,1,"_CPPv2N6paddle17ResponseNormLayer17ResponseNormLayerERK11LayerConfig"],"paddle::ResponseNormLayer::backward":[41,6,1,"_CPPv2N6paddle17ResponseNormLayer8backwardERK14UpdateCallback"],"paddle::ResponseNormLayer::channels_":[41,7,1,"_CPPv2N6paddle17ResponseNormLayer9channels_E"],"paddle::ResponseNormLayer::denoms_":[41,7,1,"_CPPv2N6paddle17ResponseNormLayer7denoms_E"],"paddle::ResponseNormLayer::forward":[41,6,1,"_CPPv2N6paddle17ResponseNormLayer7forwardE8PassType"],"paddle::ResponseNormLayer::imgSize_":[41,7,1,"_CPPv2N6paddle17ResponseNormLayer8imgSize_E"],"paddle::ResponseNormLayer::init":[41,6,1,"_CPPv2N6paddle17ResponseNormLayer4initERK8LayerMapRK12ParameterMap"],"paddle::ResponseNormLayer::outputX_":[41,7,1,"_CPPv2N6paddle17ResponseNormLayer8outputX_E"],"paddle::ResponseNormLayer::pow_":[41,7,1,"_CPPv2N6paddle17ResponseNormLayer4pow_E"],"paddle::ResponseNormLayer::scale_":[41,7,1,"_CPPv2N6paddle17ResponseNormLayer6scale_E"],"paddle::ResponseNormLayer::size_":[41,7,1,"_CPPv2N6paddle17ResponseNormLayer5size_E"],"paddle::SPARSE_CSC":[44,8,1,"_CPPv2N6paddle10SPARSE_CSCE"],"paddle::SPARSE_CSR":[44,8,1,"_CPPv2N6paddle10SPARSE_CSRE"],"paddle::SVectorPtr":[50,9,1,"_CPPv2N6paddle10SVectorPtrE"],"paddle::SYNCED":[44,8,1,"_CPPv2N6paddle6SYNCEDE"],"paddle::SamplingIdLayer":[41,8,1,"_CPPv2N6paddle15SamplingIdLayerE"],"paddle::SamplingIdLayer::SamplingIdLayer":[41,6,1,"_CPPv2N6paddle15SamplingIdLayer15SamplingIdLayerERK11LayerConfig"],"paddle::SamplingIdLayer::backward":[41,6,1,"_CPPv2N6paddle15SamplingIdLayer8backwardERK14UpdateCallback"],"paddle::SamplingIdLayer::forward":[41,6,1,"_CPPv2N6paddle15SamplingIdLayer7forwardE8PassType"],"paddle::SamplingIdLayer::forwardImp":[41,6,1,"_CPPv2N6paddle15SamplingIdLayer10forwardImpERK8Argument"],"paddle::SamplingIdLayer::init":[41,6,1,"_CPPv2N6paddle15SamplingIdLayer4initERK8LayerMapRK12ParameterMap"],"paddle::ScalingLayer":[41,8,1,"_CPPv2N6paddle12ScalingLayerE"],"paddle::ScalingLayer::ScalingLayer":[41,6,1,"_CPPv2N6paddle12ScalingLayer12ScalingLayerERK11LayerConfig"],"paddle::ScalingLayer::backward":[41,6,1,"_CPPv2N6paddle12ScalingLayer8backwardERK14UpdateCallback"],"paddle::ScalingLayer::forward":[41,6,1,"_CPPv2N6paddle12ScalingLayer7forwardE8PassType"],"paddle::ScalingLayer::init":[41,6,1,"_CPPv2N6paddle12ScalingLayer4initERK8LayerMapRK12ParameterMap"],"paddle::ScalingLayer::~ScalingLayer":[41,6,1,"_CPPv2N6paddle12ScalingLayerD0Ev"],"paddle::ScatterAgentLayer":[41,8,1,"_CPPv2N6paddle17ScatterAgentLayerE"],"paddle::ScatterAgentLayer::ScatterAgentLayer":[41,6,1,"_CPPv2N6paddle17ScatterAgentLayer17ScatterAgentLayerERK11LayerConfig"],"paddle::ScatterAgentLayer::backward":[41,6,1,"_CPPv2N6paddle17ScatterAgentLayer8backwardERK14UpdateCallback"],"paddle::ScatterAgentLayer::cpuIds_":[41,7,1,"_CPPv2N6paddle17ScatterAgentLayer7cpuIds_E"],"paddle::ScatterAgentLayer::forward":[41,6,1,"_CPPv2N6paddle17ScatterAgentLayer7forwardE8PassType"],"paddle::ScatterAgentLayer::idIndex_":[41,7,1,"_CPPv2N6paddle17ScatterAgentLayer8idIndex_E"],"paddle::ScatterAgentLayer::idSize_":[41,7,1,"_CPPv2N6paddle17ScatterAgentLayer7idSize_E"],"paddle::ScatterAgentLayer::ids_":[41,7,1,"_CPPv2N6paddle17ScatterAgentLayer4ids_E"],"paddle::ScatterAgentLayer::init":[41,6,1,"_CPPv2N6paddle17ScatterAgentLayer4initERK8LayerMapRK12ParameterMap"],"paddle::ScatterAgentLayer::numSequences_":[41,7,1,"_CPPv2N6paddle17ScatterAgentLayer13numSequences_E"],"paddle::ScatterAgentLayer::realLayer_":[41,7,1,"_CPPv2N6paddle17ScatterAgentLayer10realLayer_E"],"paddle::ScatterAgentLayer::realOutArg_":[41,7,1,"_CPPv2N6paddle17ScatterAgentLayer11realOutArg_E"],"paddle::ScatterAgentLayer::seqStartPosIndex_":[41,7,1,"_CPPv2N6paddle17ScatterAgentLayer17seqStartPosIndex_E"],"paddle::ScatterAgentLayer::setRealLayer":[41,6,1,"_CPPv2N6paddle17ScatterAgentLayer12setRealLayerE8LayerPtrRKNSt6vectorIiEEb"],"paddle::ScatterAgentLayer::setRealLayerAndOutput":[41,6,1,"_CPPv2N6paddle17ScatterAgentLayer21setRealLayerAndOutputE8LayerPtrRK8ArgumentRK10IVectorPtrii"],"paddle::ScatterAgentLayer::setSequenceStartPositions":[41,6,1,"_CPPv2N6paddle17ScatterAgentLayer25setSequenceStartPositionsERK16ICpuGpuVectorPtrii"],"paddle::ScatterAgentLayer::~ScatterAgentLayer":[41,6,1,"_CPPv2N6paddle17ScatterAgentLayerD0Ev"],"paddle::Segment":[50,8,1,"_CPPv2N6paddle7SegmentE"],"paddle::Segment::beginDim":[50,7,1,"_CPPv2N6paddle7Segment8beginDimE"],"paddle::Segment::beginPos":[50,7,1,"_CPPv2N6paddle7Segment8beginPosE"],"paddle::Segment::endDim":[50,7,1,"_CPPv2N6paddle7Segment6endDimE"],"paddle::SelectiveFullyConnectedLayer":[41,8,1,"_CPPv2N6paddle28SelectiveFullyConnectedLayerE"],"paddle::SelectiveFullyConnectedLayer::SelectiveFullyConnectedLayer":[41,6,1,"_CPPv2N6paddle28SelectiveFullyConnectedLayer28SelectiveFullyConnectedLayerERK11LayerConfig"],"paddle::SelectiveFullyConnectedLayer::backward":[41,6,1,"_CPPv2N6paddle28SelectiveFullyConnectedLayer8backwardERK14UpdateCallback"],"paddle::SelectiveFullyConnectedLayer::biases_":[41,7,1,"_CPPv2N6paddle28SelectiveFullyConnectedLayer7biases_E"],"paddle::SelectiveFullyConnectedLayer::fillSelectiveData":[41,6,1,"_CPPv2N6paddle28SelectiveFullyConnectedLayer17fillSelectiveDataERKNSt10shared_ptrINSt6vectorINSt4pairIPi6size_tEEEEEE"],"paddle::SelectiveFullyConnectedLayer::forward":[41,6,1,"_CPPv2N6paddle28SelectiveFullyConnectedLayer7forwardE8PassType"],"paddle::SelectiveFullyConnectedLayer::getWeight":[41,6,1,"_CPPv2N6paddle28SelectiveFullyConnectedLayer9getWeightEi"],"paddle::SelectiveFullyConnectedLayer::init":[41,6,1,"_CPPv2N6paddle28SelectiveFullyConnectedLayer4initERK8LayerMapRK12ParameterMap"],"paddle::SelectiveFullyConnectedLayer::prefetch":[41,6,1,"_CPPv2N6paddle28SelectiveFullyConnectedLayer8prefetchEv"],"paddle::SelectiveFullyConnectedLayer::reserveOutput":[41,6,1,"_CPPv2N6paddle28SelectiveFullyConnectedLayer13reserveOutputE6size_t6size_t6size_t"],"paddle::SelectiveFullyConnectedLayer::weights_":[41,7,1,"_CPPv2N6paddle28SelectiveFullyConnectedLayer8weights_E"],"paddle::SelectiveFullyConnectedLayer::~SelectiveFullyConnectedLayer":[41,6,1,"_CPPv2N6paddle28SelectiveFullyConnectedLayerD0Ev"],"paddle::Semaphore":[64,8,1,"_CPPv2N6paddle9SemaphoreE"],"paddle::Semaphore::Semaphore":[64,6,1,"_CPPv2N6paddle9Semaphore9SemaphoreEi"],"paddle::Semaphore::post":[64,6,1,"_CPPv2N6paddle9Semaphore4postEv"],"paddle::Semaphore::sem_":[64,7,1,"_CPPv2N6paddle9Semaphore4sem_E"],"paddle::Semaphore::timeWait":[64,6,1,"_CPPv2N6paddle9Semaphore8timeWaitEP8timespec"],"paddle::Semaphore::wait":[64,6,1,"_CPPv2N6paddle9Semaphore4waitEv"],"paddle::Semaphore::~Semaphore":[64,6,1,"_CPPv2N6paddle9SemaphoreD0Ev"],"paddle::SequenceAgentLayer":[41,8,1,"_CPPv2N6paddle18SequenceAgentLayerE"],"paddle::SequenceAgentLayer::SequenceAgentLayer":[41,6,1,"_CPPv2N6paddle18SequenceAgentLayer18SequenceAgentLayerERK11LayerConfig"],"paddle::SequenceAgentLayer::backward":[41,6,1,"_CPPv2N6paddle18SequenceAgentLayer8backwardERK14UpdateCallback"],"paddle::SequenceAgentLayer::forward":[41,6,1,"_CPPv2N6paddle18SequenceAgentLayer7forwardE8PassType"],"paddle::SequenceAgentLayer::~SequenceAgentLayer":[41,6,1,"_CPPv2N6paddle18SequenceAgentLayerD0Ev"],"paddle::SequenceClassificationErrorEvaluator":[36,8,1,"_CPPv2N6paddle36SequenceClassificationErrorEvaluatorE"],"paddle::SequenceClassificationErrorEvaluator::distributeEval":[36,6,1,"_CPPv2N6paddle36SequenceClassificationErrorEvaluator14distributeEvalEP16ParameterClient2"],"paddle::SequenceClassificationErrorEvaluator::evalImp":[36,6,1,"_CPPv2N6paddle36SequenceClassificationErrorEvaluator7evalImpERNSt6vectorI8ArgumentEE"],"paddle::SequenceClassificationErrorEvaluator::updateSamplesNum":[36,6,1,"_CPPv2N6paddle36SequenceClassificationErrorEvaluator16updateSamplesNumERKNSt6vectorI8ArgumentEE"],"paddle::SequenceConcatLayer":[41,8,1,"_CPPv2N6paddle19SequenceConcatLayerE"],"paddle::SequenceConcatLayer::SequenceConcatLayer":[41,6,1,"_CPPv2N6paddle19SequenceConcatLayer19SequenceConcatLayerERK11LayerConfig"],"paddle::SequenceConcatLayer::backward":[41,6,1,"_CPPv2N6paddle19SequenceConcatLayer8backwardERK14UpdateCallback"],"paddle::SequenceConcatLayer::biases_":[41,7,1,"_CPPv2N6paddle19SequenceConcatLayer7biases_E"],"paddle::SequenceConcatLayer::forward":[41,6,1,"_CPPv2N6paddle19SequenceConcatLayer7forwardE8PassType"],"paddle::SequenceConcatLayer::init":[41,6,1,"_CPPv2N6paddle19SequenceConcatLayer4initERK8LayerMapRK12ParameterMap"],"paddle::SequenceConcatLayer::~SequenceConcatLayer":[41,6,1,"_CPPv2N6paddle19SequenceConcatLayerD0Ev"],"paddle::SequenceGatherAgentLayer":[41,8,1,"_CPPv2N6paddle24SequenceGatherAgentLayerE"],"paddle::SequenceGatherAgentLayer::SequenceGatherAgentLayer":[41,6,1,"_CPPv2N6paddle24SequenceGatherAgentLayer24SequenceGatherAgentLayerERK11LayerConfig"],"paddle::SequenceGatherAgentLayer::backward":[41,6,1,"_CPPv2N6paddle24SequenceGatherAgentLayer8backwardERK14UpdateCallback"],"paddle::SequenceGatherAgentLayer::forward":[41,6,1,"_CPPv2N6paddle24SequenceGatherAgentLayer7forwardE8PassType"],"paddle::SequenceGatherAgentLayer::~SequenceGatherAgentLayer":[41,6,1,"_CPPv2N6paddle24SequenceGatherAgentLayerD0Ev"],"paddle::SequenceLastInstanceLayer":[41,8,1,"_CPPv2N6paddle25SequenceLastInstanceLayerE"],"paddle::SequenceLastInstanceLayer::SequenceLastInstanceLayer":[41,6,1,"_CPPv2N6paddle25SequenceLastInstanceLayer25SequenceLastInstanceLayerERK11LayerConfig"],"paddle::SequenceLastInstanceLayer::SequenceLevel":[41,9,1,"_CPPv2N6paddle25SequenceLastInstanceLayer13SequenceLevelE"],"paddle::SequenceLastInstanceLayer::backward":[41,6,1,"_CPPv2N6paddle25SequenceLastInstanceLayer8backwardERK14UpdateCallback"],"paddle::SequenceLastInstanceLayer::biases_":[41,7,1,"_CPPv2N6paddle25SequenceLastInstanceLayer7biases_E"],"paddle::SequenceLastInstanceLayer::forward":[41,6,1,"_CPPv2N6paddle25SequenceLastInstanceLayer7forwardE8PassType"],"paddle::SequenceLastInstanceLayer::init":[41,6,1,"_CPPv2N6paddle25SequenceLastInstanceLayer4initERK8LayerMapRK12ParameterMap"],"paddle::SequenceLastInstanceLayer::kNonSeq":[41,8,1,"_CPPv2N6paddle25SequenceLastInstanceLayer7kNonSeqE"],"paddle::SequenceLastInstanceLayer::kSeq":[41,8,1,"_CPPv2N6paddle25SequenceLastInstanceLayer4kSeqE"],"paddle::SequenceLastInstanceLayer::tmpDest_":[41,7,1,"_CPPv2N6paddle25SequenceLastInstanceLayer8tmpDest_E"],"paddle::SequenceLastInstanceLayer::tmpSrc_":[41,7,1,"_CPPv2N6paddle25SequenceLastInstanceLayer7tmpSrc_E"],"paddle::SequenceLastInstanceLayer::type_":[41,7,1,"_CPPv2N6paddle25SequenceLastInstanceLayer5type_E"],"paddle::SequenceLastInstanceLayer::~SequenceLastInstanceLayer":[41,6,1,"_CPPv2N6paddle25SequenceLastInstanceLayerD0Ev"],"paddle::SequenceReshapeLayer":[41,8,1,"_CPPv2N6paddle20SequenceReshapeLayerE"],"paddle::SequenceReshapeLayer::SequenceReshapeLayer":[41,6,1,"_CPPv2N6paddle20SequenceReshapeLayer20SequenceReshapeLayerERK11LayerConfig"],"paddle::SequenceReshapeLayer::backward":[41,6,1,"_CPPv2N6paddle20SequenceReshapeLayer8backwardERK14UpdateCallback"],"paddle::SequenceReshapeLayer::biases_":[41,7,1,"_CPPv2N6paddle20SequenceReshapeLayer7biases_E"],"paddle::SequenceReshapeLayer::forward":[41,6,1,"_CPPv2N6paddle20SequenceReshapeLayer7forwardE8PassType"],"paddle::SequenceReshapeLayer::init":[41,6,1,"_CPPv2N6paddle20SequenceReshapeLayer4initERK8LayerMapRK12ParameterMap"],"paddle::SequenceReshapeLayer::reshapedOutputGrad":[41,7,1,"_CPPv2N6paddle20SequenceReshapeLayer18reshapedOutputGradE"],"paddle::SequenceReshapeLayer::~SequenceReshapeLayer":[41,6,1,"_CPPv2N6paddle20SequenceReshapeLayerD0Ev"],"paddle::SequenceScatterAgentLayer":[41,8,1,"_CPPv2N6paddle25SequenceScatterAgentLayerE"],"paddle::SequenceScatterAgentLayer::SequenceScatterAgentLayer":[41,6,1,"_CPPv2N6paddle25SequenceScatterAgentLayer25SequenceScatterAgentLayerERK11LayerConfig"],"paddle::SequenceScatterAgentLayer::backward":[41,6,1,"_CPPv2N6paddle25SequenceScatterAgentLayer8backwardERK14UpdateCallback"],"paddle::SequenceScatterAgentLayer::cpuInputStartPos_":[41,7,1,"_CPPv2N6paddle25SequenceScatterAgentLayer17cpuInputStartPos_E"],"paddle::SequenceScatterAgentLayer::forward":[41,6,1,"_CPPv2N6paddle25SequenceScatterAgentLayer7forwardE8PassType"],"paddle::SequenceScatterAgentLayer::inputStartPos_":[41,7,1,"_CPPv2N6paddle25SequenceScatterAgentLayer14inputStartPos_E"],"paddle::SequenceScatterAgentLayer::~SequenceScatterAgentLayer":[41,6,1,"_CPPv2N6paddle25SequenceScatterAgentLayerD0Ev"],"paddle::SequenceTextPrinter":[36,8,1,"_CPPv2N6paddle19SequenceTextPrinterE"],"paddle::SequenceTextPrinter::SequenceTextPrinter":[36,6,1,"_CPPv2N6paddle19SequenceTextPrinter19SequenceTextPrinterEv"],"paddle::SequenceTextPrinter::evalImp":[36,6,1,"_CPPv2N6paddle19SequenceTextPrinter7evalImpERNSt6vectorI8ArgumentEE"],"paddle::SequenceTextPrinter::init":[36,6,1,"_CPPv2N6paddle19SequenceTextPrinter4initERK15EvaluatorConfig"],"paddle::SequenceTextPrinter::updateSamplesNum":[36,6,1,"_CPPv2N6paddle19SequenceTextPrinter16updateSamplesNumERKNSt6vectorI8ArgumentEE"],"paddle::SequenceToBatch":[41,8,1,"_CPPv2N6paddle15SequenceToBatchE"],"paddle::SequenceToBatch::SequenceToBatch":[41,6,1,"_CPPv2N6paddle15SequenceToBatch15SequenceToBatchEb"],"paddle::SequenceToBatch::add":[41,6,1,"_CPPv2N6paddle15SequenceToBatch3addER6MatrixR6Matrixb"],"paddle::SequenceToBatch::batchStartPositions_":[41,7,1,"_CPPv2N6paddle15SequenceToBatch20batchStartPositions_E"],"paddle::SequenceToBatch::batchValue_":[41,7,1,"_CPPv2N6paddle15SequenceToBatch11batchValue_E"],"paddle::SequenceToBatch::copy":[41,6,1,"_CPPv2N6paddle15SequenceToBatch4copyER6MatrixR6Matrixb"],"paddle::SequenceToBatch::copyBackSeq":[41,6,1,"_CPPv2N6paddle15SequenceToBatch11copyBackSeqER6Matrix"],"paddle::SequenceToBatch::copyFromSeq":[41,6,1,"_CPPv2N6paddle15SequenceToBatch11copyFromSeqER6Matrix"],"paddle::SequenceToBatch::cpuSeq2BatchIdx_":[41,7,1,"_CPPv2N6paddle15SequenceToBatch16cpuSeq2BatchIdx_E"],"paddle::SequenceToBatch::cpuSeqEndIdxInBatch_":[41,7,1,"_CPPv2N6paddle15SequenceToBatch20cpuSeqEndIdxInBatch_E"],"paddle::SequenceToBatch::cpuSeqIdx_":[41,7,1,"_CPPv2N6paddle15SequenceToBatch10cpuSeqIdx_E"],"paddle::SequenceToBatch::getBatchValue":[41,6,1,"_CPPv2N6paddle15SequenceToBatch13getBatchValueER6Matrixii"],"paddle::SequenceToBatch::getNumBatch":[41,6,1,"_CPPv2N6paddle15SequenceToBatch11getNumBatchEv"],"paddle::SequenceToBatch::getSeqOutputFromBatch":[41,6,1,"_CPPv2N6paddle15SequenceToBatch21getSeqOutputFromBatchER6MatrixR6Matrix"],"paddle::SequenceToBatch::numBatch_":[41,7,1,"_CPPv2N6paddle15SequenceToBatch9numBatch_E"],"paddle::SequenceToBatch::prevOutput2Batch":[41,6,1,"_CPPv2N6paddle15SequenceToBatch16prevOutput2BatchER6MatrixR6Matrix"],"paddle::SequenceToBatch::resizeOrCreate":[41,6,1,"_CPPv2N6paddle15SequenceToBatch14resizeOrCreateER6Matrix"],"paddle::SequenceToBatch::resizeOrCreateBatch":[41,6,1,"_CPPv2N6paddle15SequenceToBatch19resizeOrCreateBatchEi6size_tPKibb"],"paddle::SequenceToBatch::seq2BatchIdx_":[41,7,1,"_CPPv2N6paddle15SequenceToBatch13seq2BatchIdx_E"],"paddle::SequenceToBatch::seqEndIdxInBatch_":[41,7,1,"_CPPv2N6paddle15SequenceToBatch17seqEndIdxInBatch_E"],"paddle::SequenceToBatch::seqIdx_":[41,7,1,"_CPPv2N6paddle15SequenceToBatch7seqIdx_E"],"paddle::SequenceToBatch::sequence2BatchAdd":[41,6,1,"_CPPv2N6paddle15SequenceToBatch17sequence2BatchAddER6MatrixR6MatrixR7IVectorb"],"paddle::SequenceToBatch::sequence2BatchCopy":[41,6,1,"_CPPv2N6paddle15SequenceToBatch18sequence2BatchCopyER6MatrixR6MatrixR7IVectorb"],"paddle::SequenceToBatch::shareIndexWith":[41,6,1,"_CPPv2N6paddle15SequenceToBatch14shareIndexWithERK15SequenceToBatch"],"paddle::SequenceToBatch::useGpu_":[41,7,1,"_CPPv2N6paddle15SequenceToBatch7useGpu_E"],"paddle::SgdOptimizer":[48,8,1,"_CPPv2N6paddle12SgdOptimizerE"],"paddle::SgdOptimizer::SgdOptimizer":[48,6,1,"_CPPv2N6paddle12SgdOptimizer12SgdOptimizerERK18OptimizationConfig"],"paddle::SgdOptimizer::finishBatch":[48,6,1,"_CPPv2N6paddle12SgdOptimizer11finishBatchEv"],"paddle::SgdOptimizer::startBatch":[48,6,1,"_CPPv2N6paddle12SgdOptimizer10startBatchE7int64_t"],"paddle::SgdOptimizer::update":[48,6,1,"_CPPv2N6paddle12SgdOptimizer6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::SharedCpuMatrix":[44,8,1,"_CPPv2N6paddle15SharedCpuMatrixE"],"paddle::SharedCpuMatrix::SharedCpuMatrix":[44,6,1,"_CPPv2N6paddle15SharedCpuMatrix15SharedCpuMatrixEi6size_t6size_tb"],"paddle::SharedCpuMatrix::add":[44,6,1,"_CPPv2N6paddle15SharedCpuMatrix3addER6Matrix4real4real"],"paddle::SharedCpuMatrix::blockLocks_":[44,7,1,"_CPPv2N6paddle15SharedCpuMatrix11blockLocks_E"],"paddle::SharedCpuMatrix::blockNum_":[44,7,1,"_CPPv2N6paddle15SharedCpuMatrix9blockNum_E"],"paddle::SharedCpuMatrix::blockSeq_":[44,7,1,"_CPPv2N6paddle15SharedCpuMatrix9blockSeq_E"],"paddle::SharedCpuMatrix::initBlock":[44,6,1,"_CPPv2N6paddle15SharedCpuMatrix9initBlockEi"],"paddle::SharedCpuMatrix::initShared":[44,6,1,"_CPPv2N6paddle15SharedCpuMatrix10initSharedEi"],"paddle::SharedCpuMatrix::localBufRows_":[44,7,1,"_CPPv2N6paddle15SharedCpuMatrix13localBufRows_E"],"paddle::SharedCpuMatrix::localBuf_":[44,7,1,"_CPPv2N6paddle15SharedCpuMatrix9localBuf_E"],"paddle::SharedCpuMatrix::mul":[44,6,1,"_CPPv2N6paddle15SharedCpuMatrix3mulEP15CpuSparseMatrixP9CpuMatrix4real4real"],"paddle::SharedCpuMatrix::~SharedCpuMatrix":[44,6,1,"_CPPv2N6paddle15SharedCpuMatrixD0Ev"],"paddle::SlopeInterceptLayer":[41,8,1,"_CPPv2N6paddle19SlopeInterceptLayerE"],"paddle::SlopeInterceptLayer::SlopeInterceptLayer":[41,6,1,"_CPPv2N6paddle19SlopeInterceptLayer19SlopeInterceptLayerERK11LayerConfig"],"paddle::SlopeInterceptLayer::backward":[41,6,1,"_CPPv2N6paddle19SlopeInterceptLayer8backwardERK14UpdateCallback"],"paddle::SlopeInterceptLayer::forward":[41,6,1,"_CPPv2N6paddle19SlopeInterceptLayer7forwardE8PassType"],"paddle::SlopeInterceptLayer::init":[41,6,1,"_CPPv2N6paddle19SlopeInterceptLayer4initERK8LayerMapRK12ParameterMap"],"paddle::SlopeInterceptLayer::~SlopeInterceptLayer":[41,6,1,"_CPPv2N6paddle19SlopeInterceptLayerD0Ev"],"paddle::SocketChannel":[56,8,1,"_CPPv2N6paddle13SocketChannelE"],"paddle::SocketChannel::MessageHeader":[56,8,1,"_CPPv2N6paddle13SocketChannel13MessageHeaderE"],"paddle::SocketChannel::MessageHeader::iovLengths":[56,7,1,"_CPPv2N6paddle13SocketChannel13MessageHeader10iovLengthsE"],"paddle::SocketChannel::MessageHeader::numIovs":[56,7,1,"_CPPv2N6paddle13SocketChannel13MessageHeader7numIovsE"],"paddle::SocketChannel::MessageHeader::totalLength":[56,7,1,"_CPPv2N6paddle13SocketChannel13MessageHeader11totalLengthE"],"paddle::SocketChannel::SocketChannel":[56,6,1,"_CPPv2N6paddle13SocketChannel13SocketChannelEiRKNSt6stringE"],"paddle::SocketChannel::getPeerName":[56,6,1,"_CPPv2N6paddle13SocketChannel11getPeerNameEv"],"paddle::SocketChannel::peerName_":[56,7,1,"_CPPv2N6paddle13SocketChannel9peerName_E"],"paddle::SocketChannel::rdmaSocket_":[56,7,1,"_CPPv2N6paddle13SocketChannel11rdmaSocket_E"],"paddle::SocketChannel::read":[56,6,1,"_CPPv2N6paddle13SocketChannel4readEPv6size_t"],"paddle::SocketChannel::readMessage":[56,6,1,"_CPPv2N6paddle13SocketChannel11readMessageEv"],"paddle::SocketChannel::readv":[56,6,1,"_CPPv2N6paddle13SocketChannel5readvEPNSt6vectorI5iovecEE"],"paddle::SocketChannel::tcpRdma_":[56,7,1,"_CPPv2N6paddle13SocketChannel8tcpRdma_E"],"paddle::SocketChannel::tcpSocket_":[56,7,1,"_CPPv2N6paddle13SocketChannel10tcpSocket_E"],"paddle::SocketChannel::write":[56,6,1,"_CPPv2N6paddle13SocketChannel5writeEPKv6size_t"],"paddle::SocketChannel::writeMessage":[56,6,1,"_CPPv2N6paddle13SocketChannel12writeMessageERKNSt6vectorI5iovecEE"],"paddle::SocketChannel::writev":[56,6,1,"_CPPv2N6paddle13SocketChannel6writevERKNSt6vectorI5iovecEE"],"paddle::SocketChannel::~SocketChannel":[56,6,1,"_CPPv2N6paddle13SocketChannelD0Ev"],"paddle::SocketClient":[56,8,1,"_CPPv2N6paddle12SocketClientE"],"paddle::SocketClient::RdmaClient":[56,6,1,"_CPPv2N6paddle12SocketClient10RdmaClientERKNSt6stringEi"],"paddle::SocketClient::TcpClient":[56,6,1,"_CPPv2N6paddle12SocketClient9TcpClientERKNSt6stringEi"],"paddle::SocketClient::channel_":[56,7,1,"_CPPv2N6paddle12SocketClient8channel_E"],"paddle::SocketClient::getChannel":[56,6,1,"_CPPv2N6paddle12SocketClient10getChannelEv"],"paddle::SocketClient::socketDaemon_":[56,7,1,"_CPPv2N6paddle12SocketClient13socketDaemon_E"],"paddle::SocketClient::tcpRdma_":[56,7,1,"_CPPv2N6paddle12SocketClient8tcpRdma_E"],"paddle::SocketServer":[56,8,1,"_CPPv2N6paddle12SocketServerE"],"paddle::SocketServer::ResponseCallback":[56,9,1,"_CPPv2N6paddle12SocketServer16ResponseCallbackE"],"paddle::SocketServer::SocketServer":[56,6,1,"_CPPv2N6paddle12SocketServer12SocketServerERKNSt6stringEii"],"paddle::SocketServer::addr_":[56,7,1,"_CPPv2N6paddle12SocketServer5addr_E"],"paddle::SocketServer::createChannel":[56,6,1,"_CPPv2N6paddle12SocketServer13createChannelEiRKNSt6stringE"],"paddle::SocketServer::detach":[56,6,1,"_CPPv2N6paddle12SocketServer6detachEv"],"paddle::SocketServer::handleRequest":[56,6,1,"_CPPv2N6paddle12SocketServer13handleRequestENSt10unique_ptrI9MsgReaderEE16ResponseCallback"],"paddle::SocketServer::maxPendingConnections_":[56,7,1,"_CPPv2N6paddle12SocketServer22maxPendingConnections_E"],"paddle::SocketServer::port_":[56,7,1,"_CPPv2N6paddle12SocketServer5port_E"],"paddle::SocketServer::rdmaCpu_":[56,7,1,"_CPPv2N6paddle12SocketServer8rdmaCpu_E"],"paddle::SocketServer::rdmaServer":[56,6,1,"_CPPv2N6paddle12SocketServer10rdmaServerEv"],"paddle::SocketServer::rdmaSocket_":[56,7,1,"_CPPv2N6paddle12SocketServer11rdmaSocket_E"],"paddle::SocketServer::rdmaUri_":[56,7,1,"_CPPv2N6paddle12SocketServer8rdmaUri_E"],"paddle::SocketServer::run":[56,6,1,"_CPPv2N6paddle12SocketServer3runEv"],"paddle::SocketServer::socket_":[56,7,1,"_CPPv2N6paddle12SocketServer7socket_E"],"paddle::SocketServer::stopping_":[56,7,1,"_CPPv2N6paddle12SocketServer9stopping_E"],"paddle::SocketServer::tcpRdma_":[56,7,1,"_CPPv2N6paddle12SocketServer8tcpRdma_E"],"paddle::SocketServer::tcpServer":[56,6,1,"_CPPv2N6paddle12SocketServer9tcpServerEv"],"paddle::SocketServer::~SocketServer":[56,6,1,"_CPPv2N6paddle12SocketServerD0Ev"],"paddle::SocketWorker":[56,8,1,"_CPPv2N6paddle12SocketWorkerE"],"paddle::SocketWorker::SocketWorker":[56,6,1,"_CPPv2N6paddle12SocketWorker12SocketWorkerERRNSt10unique_ptrI13SocketChannelEEP12SocketServer"],"paddle::SocketWorker::channel_":[56,7,1,"_CPPv2N6paddle12SocketWorker8channel_E"],"paddle::SocketWorker::run":[56,6,1,"_CPPv2N6paddle12SocketWorker3runEv"],"paddle::SocketWorker::server_":[56,7,1,"_CPPv2N6paddle12SocketWorker7server_E"],"paddle::SocketWorker::tcpRdma_":[56,7,1,"_CPPv2N6paddle12SocketWorker8tcpRdma_E"],"paddle::SocketWorker::~SocketWorker":[56,6,1,"_CPPv2N6paddle12SocketWorkerD0Ev"],"paddle::SoftBinaryClassCrossEntropy":[41,8,1,"_CPPv2N6paddle27SoftBinaryClassCrossEntropyE"],"paddle::SoftBinaryClassCrossEntropy::SoftBinaryClassCrossEntropy":[41,6,1,"_CPPv2N6paddle27SoftBinaryClassCrossEntropy27SoftBinaryClassCrossEntropyERK11LayerConfig"],"paddle::SoftBinaryClassCrossEntropy::backwardImp":[41,6,1,"_CPPv2N6paddle27SoftBinaryClassCrossEntropy11backwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::SoftBinaryClassCrossEntropy::forwardImp":[41,6,1,"_CPPv2N6paddle27SoftBinaryClassCrossEntropy10forwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::SoftBinaryClassCrossEntropy::init":[41,6,1,"_CPPv2N6paddle27SoftBinaryClassCrossEntropy4initERK8LayerMapRK12ParameterMap"],"paddle::SoftBinaryClassCrossEntropy::targetPerDim_":[41,7,1,"_CPPv2N6paddle27SoftBinaryClassCrossEntropy13targetPerDim_E"],"paddle::SparseAutoGrowRowCpuMatrix":[44,8,1,"_CPPv2N6paddle26SparseAutoGrowRowCpuMatrixE"],"paddle::SparseAutoGrowRowCpuMatrix::SparseAutoGrowRowCpuMatrix":[44,6,1,"_CPPv2N6paddle26SparseAutoGrowRowCpuMatrix26SparseAutoGrowRowCpuMatrixE6size_t6size_t12IndexDictPtrb"],"paddle::SparseAutoGrowRowCpuMatrix::getRow":[44,6,1,"_CPPv2N6paddle26SparseAutoGrowRowCpuMatrix6getRowE6size_t"],"paddle::SparseAutoGrowRowCpuMatrix::getRowBuf":[44,6,1,"_CPPv2N6paddle26SparseAutoGrowRowCpuMatrix9getRowBufE6size_t"],"paddle::SparseAutoGrowRowCpuMatrix::mul":[44,6,1,"_CPPv2N6paddle26SparseAutoGrowRowCpuMatrix3mulEP15CpuSparseMatrixP9CpuMatrix4real4real"],"paddle::SparseFormat":[44,9,1,"_CPPv2N6paddle12SparseFormatE"],"paddle::SparseMomentumParameterOptimizer":[48,8,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizerE"],"paddle::SparseMomentumParameterOptimizer::SparseMomentumParameterOptimizer":[48,6,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer32SparseMomentumParameterOptimizerERK18OptimizationConfig"],"paddle::SparseMomentumParameterOptimizer::alpha_":[48,7,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer6alpha_E"],"paddle::SparseMomentumParameterOptimizer::beta_":[48,7,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer5beta_E"],"paddle::SparseMomentumParameterOptimizer::decayRate_":[48,7,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer10decayRate_E"],"paddle::SparseMomentumParameterOptimizer::finishBatch":[48,6,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer11finishBatchEv"],"paddle::SparseMomentumParameterOptimizer::gamma_":[48,7,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer6gamma_E"],"paddle::SparseMomentumParameterOptimizer::init":[48,6,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer4initE6size_tPK15ParameterConfig"],"paddle::SparseMomentumParameterOptimizer::isParameterSparse_":[48,7,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer18isParameterSparse_E"],"paddle::SparseMomentumParameterOptimizer::momentum_":[48,7,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer9momentum_E"],"paddle::SparseMomentumParameterOptimizer::needSpecialTraversal":[48,6,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer20needSpecialTraversalERK15ParameterConfig"],"paddle::SparseMomentumParameterOptimizer::startBatch":[48,6,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer10startBatchE7int64_t"],"paddle::SparseMomentumParameterOptimizer::t0Vec_":[48,7,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer6t0Vec_E"],"paddle::SparseMomentumParameterOptimizer::tau_":[48,7,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer4tau_E"],"paddle::SparseMomentumParameterOptimizer::threshold_":[48,7,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer10threshold_E"],"paddle::SparseMomentumParameterOptimizer::timer_":[48,7,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer6timer_E"],"paddle::SparseMomentumParameterOptimizer::update":[48,6,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::SparseNonValueScanner":[34,8,1,"_CPPv2N6paddle21SparseNonValueScannerE"],"paddle::SparseNonValueScanner::SparseNonValueScanner":[34,6,1,"_CPPv2N6paddle21SparseNonValueScanner21SparseNonValueScannerEP10SlotHeader"],"paddle::SparseNonValueScanner::fill":[34,6,1,"_CPPv2N6paddle21SparseNonValueScanner4fillER8ArgumentP8PyObject"],"paddle::SparseNonValueScanner::finishPrepare":[34,6,1,"_CPPv2N6paddle21SparseNonValueScanner13finishPrepareER8Argument"],"paddle::SparseNonValueScanner::height_":[34,7,1,"_CPPv2N6paddle21SparseNonValueScanner7height_E"],"paddle::SparseNonValueScanner::nnz_":[34,7,1,"_CPPv2N6paddle21SparseNonValueScanner4nnz_E"],"paddle::SparseNonValueScanner::prepare":[34,6,1,"_CPPv2N6paddle21SparseNonValueScanner7prepareER8ArgumentP8PyObject"],"paddle::SparseNonValueScanner::setData":[34,6,1,"_CPPv2N6paddle21SparseNonValueScanner7setDataEPiP4realP8PyObject"],"paddle::SparseNonValueScanner::startFill":[34,6,1,"_CPPv2N6paddle21SparseNonValueScanner9startFillER8Argument"],"paddle::SparsePrefetchRowCpuMatrix":[44,8,1,"_CPPv2N6paddle26SparsePrefetchRowCpuMatrixE"],"paddle::SparsePrefetchRowCpuMatrix::SparsePrefetchRowCpuMatrix":[44,6,1,"_CPPv2N6paddle26SparsePrefetchRowCpuMatrix26SparsePrefetchRowCpuMatrixE15CpuMemHandlePtr6size_t6size_t12IndexDictPtrP14SyncThreadPoolb"],"paddle::SparsePrefetchRowCpuMatrix::addRows":[44,6,1,"_CPPv2N6paddle26SparsePrefetchRowCpuMatrix7addRowsE9MatrixPtr"],"paddle::SparsePrefetchRowCpuMatrix::pool_":[44,7,1,"_CPPv2N6paddle26SparsePrefetchRowCpuMatrix5pool_E"],"paddle::SparsePrefetchRowCpuMatrix::setupIndices":[44,6,1,"_CPPv2N6paddle26SparsePrefetchRowCpuMatrix12setupIndicesEv"],"paddle::SparseRemoteParameterUpdater":[59,8,1,"_CPPv2N6paddle28SparseRemoteParameterUpdaterE"],"paddle::SparseRemoteParameterUpdater::SparseRemoteParameterUpdater":[59,6,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater28SparseRemoteParameterUpdaterERK18OptimizationConfigib"],"paddle::SparseRemoteParameterUpdater::apply":[59,6,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater5applyEv"],"paddle::SparseRemoteParameterUpdater::batchSize_":[59,7,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater10batchSize_E"],"paddle::SparseRemoteParameterUpdater::config_":[59,7,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater7config_E"],"paddle::SparseRemoteParameterUpdater::controller":[59,6,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater10controllerEv"],"paddle::SparseRemoteParameterUpdater::controllerThread_":[59,7,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater17controllerThread_E"],"paddle::SparseRemoteParameterUpdater::expectedPassCount_":[59,7,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater18expectedPassCount_E"],"paddle::SparseRemoteParameterUpdater::finishBatch":[59,6,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater11finishBatchE4real"],"paddle::SparseRemoteParameterUpdater::finishPass":[59,6,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater10finishPassE4real"],"paddle::SparseRemoteParameterUpdater::getParametersRemote":[59,6,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater19getParametersRemoteEbb"],"paddle::SparseRemoteParameterUpdater::init":[59,6,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater4initERNSt6vectorI12ParameterPtrEE"],"paddle::SparseRemoteParameterUpdater::loadParametersRemote":[59,6,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater20loadParametersRemoteERKNSt6stringE"],"paddle::SparseRemoteParameterUpdater::parameterClient_":[59,7,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater16parameterClient_E"],"paddle::SparseRemoteParameterUpdater::passCount_":[59,7,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater10passCount_E"],"paddle::SparseRemoteParameterUpdater::randParametersRemote":[59,6,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater20randParametersRemoteEv"],"paddle::SparseRemoteParameterUpdater::restore":[59,6,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater7restoreEv"],"paddle::SparseRemoteParameterUpdater::saveParametersRemote":[59,6,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater20saveParametersRemoteERKNSt6stringE"],"paddle::SparseRemoteParameterUpdater::setForwardbackwardTime":[59,6,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater22setForwardbackwardTimeE8uint64_t"],"paddle::SparseRemoteParameterUpdater::startBatch":[59,6,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater10startBatchE7int64_t"],"paddle::SparseRemoteParameterUpdater::startController":[59,6,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater15startControllerEv"],"paddle::SparseRemoteParameterUpdater::startPass":[59,6,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater9startPassEv"],"paddle::SparseRemoteParameterUpdater::testing_":[59,7,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater8testing_E"],"paddle::SparseRemoteParameterUpdater::updateImpl":[59,6,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater10updateImplEP9Parameter"],"paddle::SparseRemoteParameterUpdater::useApplyInPserver_":[59,7,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater18useApplyInPserver_E"],"paddle::SparseRemoteParameterUpdater::~SparseRemoteParameterUpdater":[59,6,1,"_CPPv2N6paddle28SparseRemoteParameterUpdaterD0Ev"],"paddle::SparseRemoteParameterUpdaterComposite":[59,8,1,"_CPPv2N6paddle37SparseRemoteParameterUpdaterCompositeE"],"paddle::SparseRemoteParameterUpdaterComposite::NUMBER_UPDATERS":[59,8,1,"_CPPv2N6paddle37SparseRemoteParameterUpdaterComposite15NUMBER_UPDATERSE"],"paddle::SparseRemoteParameterUpdaterComposite::SparseRemoteParameterUpdaterComposite":[59,6,1,"_CPPv2N6paddle37SparseRemoteParameterUpdaterComposite37SparseRemoteParameterUpdaterCompositeERK18OptimizationConfigibRRNSt10unique_ptrI16ParameterUpdaterEE"],"paddle::SparseRemoteParameterUpdaterComposite::UPDATER_NORMAL":[59,8,1,"_CPPv2N6paddle37SparseRemoteParameterUpdaterComposite14UPDATER_NORMALE"],"paddle::SparseRemoteParameterUpdaterComposite::UPDATER_SPARSE_REMOTE":[59,8,1,"_CPPv2N6paddle37SparseRemoteParameterUpdaterComposite21UPDATER_SPARSE_REMOTEE"],"paddle::SparseRemoteParameterUpdaterComposite::__anonymous0":[59,9,1,"_CPPv2N6paddle37SparseRemoteParameterUpdaterComposite12__anonymous0E"],"paddle::SparseRemoteParameterUpdaterComposite::init":[59,6,1,"_CPPv2N6paddle37SparseRemoteParameterUpdaterComposite4initERNSt6vectorI12ParameterPtrEE"],"paddle::SparseRowCpuMatrix":[44,8,1,"_CPPv2N6paddle18SparseRowCpuMatrixE"],"paddle::SparseRowCpuMatrix::IndexDict":[44,8,1,"_CPPv2N6paddle18SparseRowCpuMatrix9IndexDictE"],"paddle::SparseRowCpuMatrix::IndexDict::globalIndices":[44,7,1,"_CPPv2N6paddle18SparseRowCpuMatrix9IndexDict13globalIndicesE"],"paddle::SparseRowCpuMatrix::IndexDict::localIndices":[44,7,1,"_CPPv2N6paddle18SparseRowCpuMatrix9IndexDict12localIndicesE"],"paddle::SparseRowCpuMatrix::IndexDictPtr":[44,9,1,"_CPPv2N6paddle18SparseRowCpuMatrix12IndexDictPtrE"],"paddle::SparseRowCpuMatrix::SparseRowCpuMatrix":[44,6,1,"_CPPv2N6paddle18SparseRowCpuMatrix18SparseRowCpuMatrixE15CpuMemHandlePtr6size_t6size_t12IndexDictPtrb"],"paddle::SparseRowCpuMatrix::addTo":[44,6,1,"_CPPv2N6paddle18SparseRowCpuMatrix5addToER10BaseMatrixRNSt6vectorI8uint32_tEE6size_t6size_t"],"paddle::SparseRowCpuMatrix::apply":[44,6,1,"_CPPv2N6paddle18SparseRowCpuMatrix5applyE4Func"],"paddle::SparseRowCpuMatrix::applyL1Decay":[44,6,1,"_CPPv2N6paddle18SparseRowCpuMatrix12applyL1DecayE4real4real"],"paddle::SparseRowCpuMatrix::checkIndex":[44,6,1,"_CPPv2N6paddle18SparseRowCpuMatrix10checkIndexE6size_t"],"paddle::SparseRowCpuMatrix::checkIndices":[44,6,1,"_CPPv2N6paddle18SparseRowCpuMatrix12checkIndicesEv"],"paddle::SparseRowCpuMatrix::checkStoreSize":[44,6,1,"_CPPv2N6paddle18SparseRowCpuMatrix14checkStoreSizeEv"],"paddle::SparseRowCpuMatrix::clearIndices":[44,6,1,"_CPPv2N6paddle18SparseRowCpuMatrix12clearIndicesEv"],"paddle::SparseRowCpuMatrix::clearRows":[44,6,1,"_CPPv2N6paddle18SparseRowCpuMatrix9clearRowsEv"],"paddle::SparseRowCpuMatrix::copyFrom":[44,6,1,"_CPPv2N6paddle18SparseRowCpuMatrix8copyFromEPK4real6size_t"],"paddle::SparseRowCpuMatrix::getIndexDictHandle":[44,6,1,"_CPPv2N6paddle18SparseRowCpuMatrix18getIndexDictHandleEv"],"paddle::SparseRowCpuMatrix::getLocalIndices":[44,6,1,"_CPPv2N6paddle18SparseRowCpuMatrix15getLocalIndicesEv"],"paddle::SparseRowCpuMatrix::getLocalRow":[44,6,1,"_CPPv2N6paddle18SparseRowCpuMatrix11getLocalRowE6size_t"],"paddle::SparseRowCpuMatrix::getRow":[44,6,1,"_CPPv2N6paddle18SparseRowCpuMatrix6getRowE6size_t"],"paddle::SparseRowCpuMatrix::getRowBuf":[44,6,1,"_CPPv2N6paddle18SparseRowCpuMatrix9getRowBufE6size_t"],"paddle::SparseRowCpuMatrix::globalIndices_":[44,7,1,"_CPPv2N6paddle18SparseRowCpuMatrix14globalIndices_E"],"paddle::SparseRowCpuMatrix::indexDictHandle_":[44,7,1,"_CPPv2N6paddle18SparseRowCpuMatrix16indexDictHandle_E"],"paddle::SparseRowCpuMatrix::init":[44,6,1,"_CPPv2N6paddle18SparseRowCpuMatrix4initE6size_t6size_t"],"paddle::SparseRowCpuMatrix::kUnusedId_":[44,7,1,"_CPPv2N6paddle18SparseRowCpuMatrix10kUnusedId_E"],"paddle::SparseRowCpuMatrix::localIndices_":[44,7,1,"_CPPv2N6paddle18SparseRowCpuMatrix13localIndices_E"],"paddle::SparseRowCpuMatrix::mul":[44,6,1,"_CPPv2N6paddle18SparseRowCpuMatrix3mulEP15CpuSparseMatrixP9CpuMatrix4real4real"],"paddle::SparseRowCpuMatrix::reserveStore":[44,6,1,"_CPPv2N6paddle18SparseRowCpuMatrix12reserveStoreEv"],"paddle::SparseRowCpuMatrix::rowStore_":[44,7,1,"_CPPv2N6paddle18SparseRowCpuMatrix9rowStore_E"],"paddle::SparseRowCpuMatrix::sgdUpdate":[44,6,1,"_CPPv2N6paddle18SparseRowCpuMatrix9sgdUpdateER10BaseMatrixR7IVector4reali4realbb"],"paddle::SparseRowCpuMatrix::storeMat_":[44,7,1,"_CPPv2N6paddle18SparseRowCpuMatrix9storeMat_E"],"paddle::SparseRowCpuMatrix::zeroMem":[44,6,1,"_CPPv2N6paddle18SparseRowCpuMatrix7zeroMemEv"],"paddle::SparseRowCpuMatrix::zeroMemThread":[44,6,1,"_CPPv2N6paddle18SparseRowCpuMatrix13zeroMemThreadE6size_t6size_t"],"paddle::SparseRowCpuMatrix::~SparseRowCpuMatrix":[44,6,1,"_CPPv2N6paddle18SparseRowCpuMatrixD0Ev"],"paddle::SparseRowIdsCpuMatrix":[44,8,1,"_CPPv2N6paddle21SparseRowIdsCpuMatrixE"],"paddle::SparseRowIdsCpuMatrix::SparseRowIdsCpuMatrix":[44,6,1,"_CPPv2N6paddle21SparseRowIdsCpuMatrix21SparseRowIdsCpuMatrixE15CpuMemHandlePtr6size_t6size_tb"],"paddle::SparseRowIdsCpuMatrix::getIds":[44,6,1,"_CPPv2N6paddle21SparseRowIdsCpuMatrix6getIdsE6size_t"],"paddle::SparseRowIdsCpuMatrix::idsArray_":[44,7,1,"_CPPv2N6paddle21SparseRowIdsCpuMatrix9idsArray_E"],"paddle::SparseRowIdsCpuMatrix::setNumOfThreads":[44,6,1,"_CPPv2N6paddle21SparseRowIdsCpuMatrix15setNumOfThreadsE6size_t"],"paddle::SparseValueScanner":[34,8,1,"_CPPv2N6paddle18SparseValueScannerE"],"paddle::SparseValueScanner::SparseValueScanner":[34,6,1,"_CPPv2N6paddle18SparseValueScanner18SparseValueScannerEP10SlotHeader"],"paddle::SparseValueScanner::finishPrepare":[34,6,1,"_CPPv2N6paddle18SparseValueScanner13finishPrepareER8Argument"],"paddle::SparseValueScanner::setData":[34,6,1,"_CPPv2N6paddle18SparseValueScanner7setDataEPiP4realP8PyObject"],"paddle::SparseValueType":[44,9,1,"_CPPv2N6paddle15SparseValueTypeE"],"paddle::SpinLock":[64,8,1,"_CPPv2N6paddle8SpinLockE"],"paddle::SpinLock::SpinLock":[64,6,1,"_CPPv2N6paddle8SpinLock8SpinLockEv"],"paddle::SpinLock::lock":[64,6,1,"_CPPv2N6paddle8SpinLock4lockEv"],"paddle::SpinLock::lock_":[64,7,1,"_CPPv2N6paddle8SpinLock5lock_E"],"paddle::SpinLock::operator=":[64,6,1,"_CPPv2N6paddle8SpinLockaSERK8SpinLock"],"paddle::SpinLock::padding_":[64,7,1,"_CPPv2N6paddle8SpinLock8padding_E"],"paddle::SpinLock::unlock":[64,6,1,"_CPPv2N6paddle8SpinLock6unlockEv"],"paddle::SpinLock::~SpinLock":[64,6,1,"_CPPv2N6paddle8SpinLockD0Ev"],"paddle::StorageEngine":[46,8,1,"_CPPv2N6paddle13StorageEngineE"],"paddle::StorageEngine::StorageEngine":[46,6,1,"_CPPv2N6paddle13StorageEngine13StorageEngineEv"],"paddle::StorageEngine::cpuAllocator_":[46,7,1,"_CPPv2N6paddle13StorageEngine13cpuAllocator_E"],"paddle::StorageEngine::getCpuAllocator":[46,6,1,"_CPPv2N6paddle13StorageEngine15getCpuAllocatorEv"],"paddle::StorageEngine::getGpuAllocator":[46,6,1,"_CPPv2N6paddle13StorageEngine15getGpuAllocatorEi"],"paddle::StorageEngine::gpuAllocator_":[46,7,1,"_CPPv2N6paddle13StorageEngine13gpuAllocator_E"],"paddle::StorageEngine::lock_":[46,7,1,"_CPPv2N6paddle13StorageEngine5lock_E"],"paddle::StorageEngine::singleton":[46,6,1,"_CPPv2N6paddle13StorageEngine9singletonEv"],"paddle::StorageEngine::~StorageEngine":[46,6,1,"_CPPv2N6paddle13StorageEngineD0Ev"],"paddle::SubSequenceLayer":[41,8,1,"_CPPv2N6paddle16SubSequenceLayerE"],"paddle::SubSequenceLayer::SubSequenceLayer":[41,6,1,"_CPPv2N6paddle16SubSequenceLayer16SubSequenceLayerERK11LayerConfig"],"paddle::SubSequenceLayer::backward":[41,6,1,"_CPPv2N6paddle16SubSequenceLayer8backwardERK14UpdateCallback"],"paddle::SubSequenceLayer::biases_":[41,7,1,"_CPPv2N6paddle16SubSequenceLayer7biases_E"],"paddle::SubSequenceLayer::forward":[41,6,1,"_CPPv2N6paddle16SubSequenceLayer7forwardE8PassType"],"paddle::SubSequenceLayer::init":[41,6,1,"_CPPv2N6paddle16SubSequenceLayer4initERK8LayerMapRK12ParameterMap"],"paddle::SubSequenceLayer::tmpDest_":[41,7,1,"_CPPv2N6paddle16SubSequenceLayer8tmpDest_E"],"paddle::SubSequenceLayer::tmpSrc_":[41,7,1,"_CPPv2N6paddle16SubSequenceLayer7tmpSrc_E"],"paddle::SubSequenceLayer::~SubSequenceLayer":[41,6,1,"_CPPv2N6paddle16SubSequenceLayerD0Ev"],"paddle::SumEvaluator":[36,8,1,"_CPPv2N6paddle12SumEvaluatorE"],"paddle::SumEvaluator::SumEvaluator":[36,6,1,"_CPPv2N6paddle12SumEvaluator12SumEvaluatorEv"],"paddle::SumEvaluator::distributeEval":[36,6,1,"_CPPv2N6paddle12SumEvaluator14distributeEvalEP16ParameterClient2"],"paddle::SumEvaluator::evalImp":[36,6,1,"_CPPv2N6paddle12SumEvaluator7evalImpERNSt6vectorI8ArgumentEE"],"paddle::SumEvaluator::updateSamplesNum":[36,6,1,"_CPPv2N6paddle12SumEvaluator16updateSamplesNumERKNSt6vectorI8ArgumentEE"],"paddle::SumOfSquaresCostLayer":[41,8,1,"_CPPv2N6paddle21SumOfSquaresCostLayerE"],"paddle::SumOfSquaresCostLayer::SumOfSquaresCostLayer":[41,6,1,"_CPPv2N6paddle21SumOfSquaresCostLayer21SumOfSquaresCostLayerERK11LayerConfig"],"paddle::SumOfSquaresCostLayer::backwardImp":[41,6,1,"_CPPv2N6paddle21SumOfSquaresCostLayer11backwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::SumOfSquaresCostLayer::forwardImp":[41,6,1,"_CPPv2N6paddle21SumOfSquaresCostLayer10forwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::SumOfSquaresCostLayer::init":[41,6,1,"_CPPv2N6paddle21SumOfSquaresCostLayer4initERK8LayerMapRK12ParameterMap"],"paddle::SumToOneNormLayer":[41,8,1,"_CPPv2N6paddle17SumToOneNormLayerE"],"paddle::SumToOneNormLayer::SumToOneNormLayer":[41,6,1,"_CPPv2N6paddle17SumToOneNormLayer17SumToOneNormLayerERK11LayerConfig"],"paddle::SumToOneNormLayer::backward":[41,6,1,"_CPPv2N6paddle17SumToOneNormLayer8backwardERK14UpdateCallback"],"paddle::SumToOneNormLayer::dotSum_":[41,7,1,"_CPPv2N6paddle17SumToOneNormLayer7dotSum_E"],"paddle::SumToOneNormLayer::forward":[41,6,1,"_CPPv2N6paddle17SumToOneNormLayer7forwardE8PassType"],"paddle::SumToOneNormLayer::init":[41,6,1,"_CPPv2N6paddle17SumToOneNormLayer4initERK8LayerMapRK12ParameterMap"],"paddle::SumToOneNormLayer::reciprocalRowSum_":[41,7,1,"_CPPv2N6paddle17SumToOneNormLayer17reciprocalRowSum_E"],"paddle::SumToOneNormLayer::~SumToOneNormLayer":[41,6,1,"_CPPv2N6paddle17SumToOneNormLayerD0Ev"],"paddle::SyncParameter":[50,8,1,"_CPPv2N6paddle13SyncParameterE"],"paddle::SyncParameter::SyncParameter":[50,6,1,"_CPPv2N6paddle13SyncParameter13SyncParameterE11TrainerRole12ParameterPtr"],"paddle::SyncParameter::attachControlParam":[50,6,1,"_CPPv2N6paddle13SyncParameter18attachControlParamE20ParallelParameterPtr"],"paddle::SyncParameter::attachMajorParam":[50,6,1,"_CPPv2N6paddle13SyncParameter16attachMajorParamE20ParallelParameterPtr"],"paddle::SyncParameter::attachMinorParam":[50,6,1,"_CPPv2N6paddle13SyncParameter16attachMinorParamE20ParallelParameterPtri"],"paddle::SyncParameter::controlParam_":[50,7,1,"_CPPv2N6paddle13SyncParameter13controlParam_E"],"paddle::SyncParameter::controlUpdate":[50,6,1,"_CPPv2N6paddle13SyncParameter13controlUpdateERK14UpdateCallback"],"paddle::SyncParameter::getMajorPartners":[50,6,1,"_CPPv2N6paddle13SyncParameter16getMajorPartnersEv"],"paddle::SyncParameter::getMinorPartners":[50,6,1,"_CPPv2N6paddle13SyncParameter16getMinorPartnersEv"],"paddle::SyncParameter::majorPartners_":[50,7,1,"_CPPv2N6paddle13SyncParameter14majorPartners_E"],"paddle::SyncParameter::majorUpdate":[50,6,1,"_CPPv2N6paddle13SyncParameter11majorUpdateE4real"],"paddle::SyncParameter::minorDeviceIds_":[50,7,1,"_CPPv2N6paddle13SyncParameter15minorDeviceIds_E"],"paddle::SyncParameter::minorPartners_":[50,7,1,"_CPPv2N6paddle13SyncParameter14minorPartners_E"],"paddle::SyncParameter::minorUpdate":[50,6,1,"_CPPv2N6paddle13SyncParameter11minorUpdateE4real"],"paddle::SyncParameter::partnerParam_":[50,7,1,"_CPPv2N6paddle13SyncParameter13partnerParam_E"],"paddle::SyncParameter::singleUpdate":[50,6,1,"_CPPv2N6paddle13SyncParameter12singleUpdateE4real"],"paddle::SyncParameter::synchronizeParamter":[50,6,1,"_CPPv2N6paddle13SyncParameter19synchronizeParamterEv"],"paddle::SyncParameter::waitAllMajorGradReady":[50,6,1,"_CPPv2N6paddle13SyncParameter21waitAllMajorGradReadyEv"],"paddle::SyncParameter::~SyncParameter":[50,6,1,"_CPPv2N6paddle13SyncParameterD0Ev"],"paddle::SyncThreadPool":[62,8,1,"_CPPv2N6paddle14SyncThreadPoolE"],"paddle::SyncThreadPool::JobFunc":[62,9,1,"_CPPv2N6paddle14SyncThreadPool7JobFuncE"],"paddle::SyncThreadPool::SyncThreadPool":[62,6,1,"_CPPv2N6paddle14SyncThreadPool14SyncThreadPoolEv"],"paddle::SyncThreadPool::checkOwner_":[62,7,1,"_CPPv2N6paddle14SyncThreadPool11checkOwner_E"],"paddle::SyncThreadPool::exec":[62,6,1,"_CPPv2N6paddle14SyncThreadPool4execE7JobFunc7JobFunc"],"paddle::SyncThreadPool::execHelper":[62,6,1,"_CPPv2N6paddle14SyncThreadPool10execHelperEP14SyncThreadPool7JobFunc"],"paddle::SyncThreadPool::execPlusOwner":[62,6,1,"_CPPv2N6paddle14SyncThreadPool13execPlusOwnerE7JobFunc"],"paddle::SyncThreadPool::getNumThreads":[62,6,1,"_CPPv2N6paddle14SyncThreadPool13getNumThreadsEv"],"paddle::SyncThreadPool::jobFinishBarrier_":[62,7,1,"_CPPv2N6paddle14SyncThreadPool17jobFinishBarrier_E"],"paddle::SyncThreadPool::jobFunc_":[62,7,1,"_CPPv2N6paddle14SyncThreadPool8jobFunc_E"],"paddle::SyncThreadPool::jobStartBarrier_":[62,7,1,"_CPPv2N6paddle14SyncThreadPool16jobStartBarrier_E"],"paddle::SyncThreadPool::ownerThreadId_":[62,7,1,"_CPPv2N6paddle14SyncThreadPool14ownerThreadId_E"],"paddle::SyncThreadPool::run":[62,6,1,"_CPPv2N6paddle14SyncThreadPool3runEi"],"paddle::SyncThreadPool::start":[62,6,1,"_CPPv2N6paddle14SyncThreadPool5startEv"],"paddle::SyncThreadPool::stop":[62,6,1,"_CPPv2N6paddle14SyncThreadPool4stopEv"],"paddle::SyncThreadPool::stopping_":[62,7,1,"_CPPv2N6paddle14SyncThreadPool9stopping_E"],"paddle::SyncThreadPool::workers_":[62,7,1,"_CPPv2N6paddle14SyncThreadPool8workers_E"],"paddle::SyncThreadPool::~SyncThreadPool":[62,6,1,"_CPPv2N6paddle14SyncThreadPoolD0Ev"],"paddle::SyncedFlag":[44,9,1,"_CPPv2N6paddle10SyncedFlagE"],"paddle::TRAINER_ROLE_CONTROL":[50,8,1,"_CPPv2N6paddle20TRAINER_ROLE_CONTROLE"],"paddle::TRAINER_ROLE_MAJOR":[50,8,1,"_CPPv2N6paddle18TRAINER_ROLE_MAJORE"],"paddle::TRAINER_ROLE_MASTER":[50,8,1,"_CPPv2N6paddle19TRAINER_ROLE_MASTERE"],"paddle::TRAINER_ROLE_MINOR":[50,8,1,"_CPPv2N6paddle18TRAINER_ROLE_MINORE"],"paddle::TRAINER_ROLE_SINGLE":[50,8,1,"_CPPv2N6paddle19TRAINER_ROLE_SINGLEE"],"paddle::TRAINER_ROLE_SLAVE":[50,8,1,"_CPPv2N6paddle18TRAINER_ROLE_SLAVEE"],"paddle::TableProjection":[41,8,1,"_CPPv2N6paddle15TableProjectionE"],"paddle::TableProjection::TableProjection":[41,6,1,"_CPPv2N6paddle15TableProjection15TableProjectionERK16ProjectionConfigRK12ParameterPtrb"],"paddle::TableProjection::backward":[41,6,1,"_CPPv2N6paddle15TableProjection8backwardERK14UpdateCallback"],"paddle::TableProjection::forward":[41,6,1,"_CPPv2N6paddle15TableProjection7forwardEv"],"paddle::TableProjection::prefetch":[41,6,1,"_CPPv2N6paddle15TableProjection8prefetchEPK8Argument"],"paddle::TableProjection::table_":[41,7,1,"_CPPv2N6paddle15TableProjection6table_E"],"paddle::TensorLayer":[41,8,1,"_CPPv2N6paddle11TensorLayerE"],"paddle::TensorLayer::TensorLayer":[41,6,1,"_CPPv2N6paddle11TensorLayer11TensorLayerERK11LayerConfig"],"paddle::TensorLayer::backward":[41,6,1,"_CPPv2N6paddle11TensorLayer8backwardERK14UpdateCallback"],"paddle::TensorLayer::biases_":[41,7,1,"_CPPv2N6paddle11TensorLayer7biases_E"],"paddle::TensorLayer::forward":[41,6,1,"_CPPv2N6paddle11TensorLayer7forwardE8PassType"],"paddle::TensorLayer::getWeight":[41,6,1,"_CPPv2N6paddle11TensorLayer9getWeightEi"],"paddle::TensorLayer::init":[41,6,1,"_CPPv2N6paddle11TensorLayer4initERK8LayerMapRK12ParameterMap"],"paddle::TensorLayer::weights_":[41,7,1,"_CPPv2N6paddle11TensorLayer8weights_E"],"paddle::TensorLayer::~TensorLayer":[41,6,1,"_CPPv2N6paddle11TensorLayerD0Ev"],"paddle::Thread":[62,8,1,"_CPPv2N6paddle6ThreadE"],"paddle::Thread::Thread":[62,6,1,"_CPPv2N6paddle6Thread6ThreadEv"],"paddle::Thread::detach":[62,6,1,"_CPPv2N6paddle6Thread6detachEv"],"paddle::Thread::join":[62,6,1,"_CPPv2N6paddle6Thread4joinEv"],"paddle::Thread::run":[62,6,1,"_CPPv2N6paddle6Thread3runEv"],"paddle::Thread::start":[62,6,1,"_CPPv2N6paddle6Thread5startEv"],"paddle::Thread::thread_":[62,7,1,"_CPPv2N6paddle6Thread7thread_E"],"paddle::Thread::~Thread":[62,6,1,"_CPPv2N6paddle6ThreadD0Ev"],"paddle::ThreadBarrier":[64,8,1,"_CPPv2N6paddle13ThreadBarrierE"],"paddle::ThreadBarrier::ThreadBarrier":[64,6,1,"_CPPv2N6paddle13ThreadBarrier13ThreadBarrierEi"],"paddle::ThreadBarrier::barrier_":[64,7,1,"_CPPv2N6paddle13ThreadBarrier8barrier_E"],"paddle::ThreadBarrier::operator=":[64,6,1,"_CPPv2N6paddle13ThreadBarrieraSERK13ThreadBarrier"],"paddle::ThreadBarrier::wait":[64,6,1,"_CPPv2N6paddle13ThreadBarrier4waitEv"],"paddle::ThreadBarrier::~ThreadBarrier":[64,6,1,"_CPPv2N6paddle13ThreadBarrierD0Ev"],"paddle::ThreadWorker":[62,8,1,"_CPPv2N6paddle12ThreadWorkerE"],"paddle::ThreadWorker::JobFunc":[62,9,1,"_CPPv2N6paddle12ThreadWorker7JobFuncE"],"paddle::ThreadWorker::ThreadWorker":[62,6,1,"_CPPv2N6paddle12ThreadWorker12ThreadWorkerEv"],"paddle::ThreadWorker::addJob":[62,6,1,"_CPPv2N6paddle12ThreadWorker6addJobE7JobFunc"],"paddle::ThreadWorker::empty_":[62,7,1,"_CPPv2N6paddle12ThreadWorker6empty_E"],"paddle::ThreadWorker::finishCV_":[62,7,1,"_CPPv2N6paddle12ThreadWorker9finishCV_E"],"paddle::ThreadWorker::jobs_":[62,7,1,"_CPPv2N6paddle12ThreadWorker5jobs_E"],"paddle::ThreadWorker::run":[62,6,1,"_CPPv2N6paddle12ThreadWorker3runEv"],"paddle::ThreadWorker::stop":[62,6,1,"_CPPv2N6paddle12ThreadWorker4stopEv"],"paddle::ThreadWorker::stopping_":[62,7,1,"_CPPv2N6paddle12ThreadWorker9stopping_E"],"paddle::ThreadWorker::wait":[62,6,1,"_CPPv2N6paddle12ThreadWorker4waitEv"],"paddle::ThreadWorker::~ThreadWorker":[62,6,1,"_CPPv2N6paddle12ThreadWorkerD0Ev"],"paddle::TrainerRole":[50,9,1,"_CPPv2N6paddle11TrainerRoleE"],"paddle::TrainerStats":[59,8,1,"_CPPv2N6paddle12TrainerStatsE"],"paddle::TrainerStats::TrainerStats":[59,6,1,"_CPPv2N6paddle12TrainerStats12TrainerStatsEv"],"paddle::TrainerStats::addCost":[59,6,1,"_CPPv2N6paddle12TrainerStats7addCostE7int64_t4real"],"paddle::TrainerStats::getAvgCost":[59,6,1,"_CPPv2N6paddle12TrainerStats10getAvgCostEv"],"paddle::TrainerStats::getCurrentAvgCost":[59,6,1,"_CPPv2N6paddle12TrainerStats17getCurrentAvgCostEv"],"paddle::TrainerStats::getNumProcessed":[59,6,1,"_CPPv2N6paddle12TrainerStats15getNumProcessedEv"],"paddle::TrainerStats::getStats":[59,6,1,"_CPPv2N6paddle12TrainerStats8getStatsEb"],"paddle::TrainerStats::operator+=":[59,6,1,"_CPPv2N6paddle12TrainerStatspLERKNSt4pairI7int64_t4realEE"],"paddle::TrainerStats::reset":[59,6,1,"_CPPv2N6paddle12TrainerStats5resetEv"],"paddle::TrainerStats::resetCurrentStat":[59,6,1,"_CPPv2N6paddle12TrainerStats16resetCurrentStatEv"],"paddle::TrainerStats::showStats":[59,6,1,"_CPPv2N6paddle12TrainerStats9showStatsERNSt7ostreamEb"],"paddle::TrainerThread":[38,8,1,"_CPPv2N6paddle13TrainerThreadE"],"paddle::TrainerThread::TrainerThread":[38,6,1,"_CPPv2N6paddle13TrainerThread13TrainerThreadERK11ModelConfigiP20MultiGradientMachine"],"paddle::TrainerThread::backward":[38,6,1,"_CPPv2N6paddle13TrainerThread8backwardEv"],"paddle::TrainerThread::backwardCallback":[38,6,1,"_CPPv2N6paddle13TrainerThread16backwardCallbackEP9Parameter"],"paddle::TrainerThread::backwardCallback_":[38,7,1,"_CPPv2N6paddle13TrainerThread17backwardCallback_E"],"paddle::TrainerThread::computeThread":[38,6,1,"_CPPv2N6paddle13TrainerThread13computeThreadEv"],"paddle::TrainerThread::computeThread_":[38,7,1,"_CPPv2N6paddle13TrainerThread14computeThread_E"],"paddle::TrainerThread::config_":[38,7,1,"_CPPv2N6paddle13TrainerThread7config_E"],"paddle::TrainerThread::copyGradToBufferThread":[38,6,1,"_CPPv2N6paddle13TrainerThread22copyGradToBufferThreadEv"],"paddle::TrainerThread::copyInArgs":[38,6,1,"_CPPv2N6paddle13TrainerThread10copyInArgsEv"],"paddle::TrainerThread::copyOutputGrad":[38,6,1,"_CPPv2N6paddle13TrainerThread14copyOutputGradEv"],"paddle::TrainerThread::copyThread_":[38,7,1,"_CPPv2N6paddle13TrainerThread11copyThread_E"],"paddle::TrainerThread::deviceId_":[38,7,1,"_CPPv2N6paddle13TrainerThread9deviceId_E"],"paddle::TrainerThread::doCallback":[38,6,1,"_CPPv2N6paddle13TrainerThread10doCallbackEi"],"paddle::TrainerThread::forward":[38,6,1,"_CPPv2N6paddle13TrainerThread7forwardEv"],"paddle::TrainerThread::getDeviceId":[38,6,1,"_CPPv2N6paddle13TrainerThread11getDeviceIdEv"],"paddle::TrainerThread::getGradientMachine":[38,6,1,"_CPPv2N6paddle13TrainerThread18getGradientMachineEv"],"paddle::TrainerThread::getOutArgs":[38,6,1,"_CPPv2N6paddle13TrainerThread10getOutArgsEv"],"paddle::TrainerThread::getParameters":[38,6,1,"_CPPv2N6paddle13TrainerThread13getParametersEv"],"paddle::TrainerThread::getValueBuf":[38,6,1,"_CPPv2N6paddle13TrainerThread11getValueBufEi"],"paddle::TrainerThread::gradBufQueue_":[38,7,1,"_CPPv2N6paddle13TrainerThread13gradBufQueue_E"],"paddle::TrainerThread::gradCollectThread":[38,6,1,"_CPPv2N6paddle13TrainerThread17gradCollectThreadEv"],"paddle::TrainerThread::gradCollectThread_":[38,7,1,"_CPPv2N6paddle13TrainerThread18gradCollectThread_E"],"paddle::TrainerThread::gradQueue_":[38,7,1,"_CPPv2N6paddle13TrainerThread10gradQueue_E"],"paddle::TrainerThread::gradStream_":[38,7,1,"_CPPv2N6paddle13TrainerThread11gradStream_E"],"paddle::TrainerThread::gradientMachine_":[38,7,1,"_CPPv2N6paddle13TrainerThread16gradientMachine_E"],"paddle::TrainerThread::inArgsCopied_":[38,7,1,"_CPPv2N6paddle13TrainerThread13inArgsCopied_E"],"paddle::TrainerThread::inArgs_":[38,7,1,"_CPPv2N6paddle13TrainerThread7inArgs_E"],"paddle::TrainerThread::incUpdateCounter":[38,6,1,"_CPPv2N6paddle13TrainerThread16incUpdateCounterEi"],"paddle::TrainerThread::mergeCpuGradients":[38,6,1,"_CPPv2N6paddle13TrainerThread17mergeCpuGradientsEv"],"paddle::TrainerThread::mergeGradDense":[38,6,1,"_CPPv2N6paddle13TrainerThread14mergeGradDenseEP9ParameterRNSt6vectorIPKNSt6vectorI12ParameterPtrEEEE"],"paddle::TrainerThread::mergeGradSparse":[38,6,1,"_CPPv2N6paddle13TrainerThread15mergeGradSparseEP9ParameterRNSt6vectorIPKNSt6vectorI12ParameterPtrEEEE"],"paddle::TrainerThread::mergeGradSparseRemote":[38,6,1,"_CPPv2N6paddle13TrainerThread21mergeGradSparseRemoteEP9ParameterRNSt6vectorIPKNSt6vectorI12ParameterPtrEEEE"],"paddle::TrainerThread::mergeTypes_":[38,7,1,"_CPPv2N6paddle13TrainerThread11mergeTypes_E"],"paddle::TrainerThread::multiMachine_":[38,7,1,"_CPPv2N6paddle13TrainerThread13multiMachine_E"],"paddle::TrainerThread::notifyCopyGradToBuffer":[38,6,1,"_CPPv2N6paddle13TrainerThread22notifyCopyGradToBufferEi"],"paddle::TrainerThread::notifyGradientCollect":[38,6,1,"_CPPv2N6paddle13TrainerThread21notifyGradientCollectEi"],"paddle::TrainerThread::notifyTaskReady":[38,6,1,"_CPPv2N6paddle13TrainerThread15notifyTaskReadyEv"],"paddle::TrainerThread::notifyValueDispatch":[38,6,1,"_CPPv2N6paddle13TrainerThread19notifyValueDispatchEi"],"paddle::TrainerThread::notifyValueReady":[38,6,1,"_CPPv2N6paddle13TrainerThread16notifyValueReadyEi"],"paddle::TrainerThread::onPassEnd":[38,6,1,"_CPPv2N6paddle13TrainerThread9onPassEndEv"],"paddle::TrainerThread::outArgsReadySem_":[38,7,1,"_CPPv2N6paddle13TrainerThread16outArgsReadySem_E"],"paddle::TrainerThread::outArgs_":[38,7,1,"_CPPv2N6paddle13TrainerThread8outArgs_E"],"paddle::TrainerThread::parameterUpdated_":[38,7,1,"_CPPv2N6paddle13TrainerThread17parameterUpdated_E"],"paddle::TrainerThread::parameters_":[38,7,1,"_CPPv2N6paddle13TrainerThread11parameters_E"],"paddle::TrainerThread::partnerId_":[38,7,1,"_CPPv2N6paddle13TrainerThread10partnerId_E"],"paddle::TrainerThread::prefetch":[38,6,1,"_CPPv2N6paddle13TrainerThread8prefetchEv"],"paddle::TrainerThread::start":[38,6,1,"_CPPv2N6paddle13TrainerThread5startEv"],"paddle::TrainerThread::stop":[38,6,1,"_CPPv2N6paddle13TrainerThread4stopEv"],"paddle::TrainerThread::stopping_":[38,7,1,"_CPPv2N6paddle13TrainerThread9stopping_E"],"paddle::TrainerThread::taskReadySem_":[38,7,1,"_CPPv2N6paddle13TrainerThread13taskReadySem_E"],"paddle::TrainerThread::threadId_":[38,7,1,"_CPPv2N6paddle13TrainerThread9threadId_E"],"paddle::TrainerThread::updateCounter_":[38,7,1,"_CPPv2N6paddle13TrainerThread14updateCounter_E"],"paddle::TrainerThread::valueDispatchThread":[38,6,1,"_CPPv2N6paddle13TrainerThread19valueDispatchThreadEv"],"paddle::TrainerThread::valueDispatchThread_":[38,7,1,"_CPPv2N6paddle13TrainerThread20valueDispatchThread_E"],"paddle::TrainerThread::valueReadyCond_":[38,7,1,"_CPPv2N6paddle13TrainerThread15valueReadyCond_E"],"paddle::TrainerThread::valueReadyQueue_":[38,7,1,"_CPPv2N6paddle13TrainerThread16valueReadyQueue_E"],"paddle::TrainerThread::valueStream_":[38,7,1,"_CPPv2N6paddle13TrainerThread12valueStream_E"],"paddle::TrainerThread::waitOutArgsReady":[38,6,1,"_CPPv2N6paddle13TrainerThread16waitOutArgsReadyEv"],"paddle::TrainerThread::~TrainerThread":[38,6,1,"_CPPv2N6paddle13TrainerThreadD0Ev"],"paddle::TransLayer":[41,8,1,"_CPPv2N6paddle10TransLayerE"],"paddle::TransLayer::TransLayer":[41,6,1,"_CPPv2N6paddle10TransLayer10TransLayerERK11LayerConfig"],"paddle::TransLayer::backward":[41,6,1,"_CPPv2N6paddle10TransLayer8backwardERK14UpdateCallback"],"paddle::TransLayer::forward":[41,6,1,"_CPPv2N6paddle10TransLayer7forwardE8PassType"],"paddle::TransLayer::init":[41,6,1,"_CPPv2N6paddle10TransLayer4initERK8LayerMapRK12ParameterMap"],"paddle::TransposedFullMatrixProjection":[41,8,1,"_CPPv2N6paddle30TransposedFullMatrixProjectionE"],"paddle::TransposedFullMatrixProjection::TransposedFullMatrixProjection":[41,6,1,"_CPPv2N6paddle30TransposedFullMatrixProjection30TransposedFullMatrixProjectionERK16ProjectionConfig12ParameterPtrb"],"paddle::TransposedFullMatrixProjection::backward":[41,6,1,"_CPPv2N6paddle30TransposedFullMatrixProjection8backwardERK14UpdateCallback"],"paddle::TransposedFullMatrixProjection::forward":[41,6,1,"_CPPv2N6paddle30TransposedFullMatrixProjection7forwardEv"],"paddle::TransposedFullMatrixProjection::weight_":[41,7,1,"_CPPv2N6paddle30TransposedFullMatrixProjection7weight_E"],"paddle::UPDATE_TYPE_NUM":[50,7,1,"_CPPv2N6paddle15UPDATE_TYPE_NUME"],"paddle::UpdateCallback":[50,9,1,"_CPPv2N6paddle14UpdateCallbackE"],"paddle::UserDefinedVectorPtr":[50,9,1,"_CPPv2N6paddle20UserDefinedVectorPtrE"],"paddle::ValidationLayer":[41,8,1,"_CPPv2N6paddle15ValidationLayerE"],"paddle::ValidationLayer::ValidationLayer":[41,6,1,"_CPPv2N6paddle15ValidationLayer15ValidationLayerERK11LayerConfig"],"paddle::ValidationLayer::backward":[41,6,1,"_CPPv2N6paddle15ValidationLayer8backwardERK14UpdateCallback"],"paddle::ValidationLayer::forward":[41,6,1,"_CPPv2N6paddle15ValidationLayer7forwardE8PassType"],"paddle::ValidationLayer::getInfoLayer":[41,6,1,"_CPPv2N6paddle15ValidationLayer12getInfoLayerEv"],"paddle::ValidationLayer::getLabelLayer":[41,6,1,"_CPPv2N6paddle15ValidationLayer13getLabelLayerEv"],"paddle::ValidationLayer::getOutputLayer":[41,6,1,"_CPPv2N6paddle15ValidationLayer14getOutputLayerEv"],"paddle::ValidationLayer::init":[41,6,1,"_CPPv2N6paddle15ValidationLayer4initERK8LayerMapRK12ParameterMap"],"paddle::ValidationLayer::onPassEnd":[41,6,1,"_CPPv2N6paddle15ValidationLayer9onPassEndEv"],"paddle::ValidationLayer::validationImp":[41,6,1,"_CPPv2N6paddle15ValidationLayer13validationImpE9MatrixPtr10IVectorPtr"],"paddle::ValuePrinter":[36,8,1,"_CPPv2N6paddle12ValuePrinterE"],"paddle::ValuePrinter::ValuePrinter":[36,6,1,"_CPPv2N6paddle12ValuePrinter12ValuePrinterEv"],"paddle::ValuePrinter::eval":[36,6,1,"_CPPv2N6paddle12ValuePrinter4evalERK13NeuralNetwork"],"paddle::ValuePrinter::evalImp":[36,6,1,"_CPPv2N6paddle12ValuePrinter7evalImpERNSt6vectorI8ArgumentEE"],"paddle::ValuePrinter::updateSamplesNum":[36,6,1,"_CPPv2N6paddle12ValuePrinter16updateSamplesNumERKNSt6vectorI8ArgumentEE"],"paddle::Vector":[44,9,1,"_CPPv2N6paddle6VectorE"],"paddle::VectorPtr":[44,9,1,"_CPPv2N6paddle9VectorPtrE"],"paddle::VectorT":[44,8,1,"_CPPv2N6paddle7VectorTE"],"paddle::VectorT::VectorT":[44,6,1,"_CPPv2N6paddle7VectorT7VectorTE6size_t15MemoryHandlePtr6size_tb"],"paddle::VectorT::copyFrom":[44,6,1,"_CPPv2N6paddle7VectorT8copyFromERK7VectorTI1TE"],"paddle::VectorT::copyTo":[44,6,1,"_CPPv2N6paddle7VectorT6copyToEP10CpuVectorTI1TE"],"paddle::VectorT::create":[44,6,1,"_CPPv2N6paddle7VectorT6createE6size_tb"],"paddle::VectorT::createParallelVector":[44,6,1,"_CPPv2N6paddle7VectorT20createParallelVectorE6size_tbP14SyncThreadPool"],"paddle::VectorT::exec":[44,6,1,"_CPPv2N6paddle7VectorT4execEN14SyncThreadPool7JobFuncE"],"paddle::VectorT::fillSequence":[44,6,1,"_CPPv2N6paddle7VectorT12fillSequenceEv"],"paddle::VectorT::get":[44,6,1,"_CPPv2N6paddle7VectorT3getE6size_t"],"paddle::VectorT::getAbsMax":[44,6,1,"_CPPv2N6paddle7VectorT9getAbsMaxEv"],"paddle::VectorT::getAbsSum":[44,6,1,"_CPPv2N6paddle7VectorT9getAbsSumEv"],"paddle::VectorT::getData":[44,6,1,"_CPPv2N6paddle7VectorT7getDataEv"],"paddle::VectorT::getElement":[44,6,1,"_CPPv2N6paddle7VectorT10getElementE6size_t"],"paddle::VectorT::getMax":[44,6,1,"_CPPv2N6paddle7VectorT6getMaxEv"],"paddle::VectorT::getMemoryHandle":[44,6,1,"_CPPv2N6paddle7VectorT15getMemoryHandleEv"],"paddle::VectorT::getMin":[44,6,1,"_CPPv2N6paddle7VectorT6getMinEv"],"paddle::VectorT::getPoint":[44,6,1,"_CPPv2N6paddle7VectorT8getPointEK8uint64_t"],"paddle::VectorT::getSize":[44,6,1,"_CPPv2N6paddle7VectorT7getSizeEv"],"paddle::VectorT::getSum":[44,6,1,"_CPPv2N6paddle7VectorT6getSumEv"],"paddle::VectorT::histogram":[44,6,1,"_CPPv2N6paddle7VectorT9histogramERNSt7ostreamEi"],"paddle::VectorT::isEqualTo":[44,6,1,"_CPPv2N6paddle7VectorT9isEqualToERK7VectorTI1TERK1T"],"paddle::VectorT::memoryHandle_":[44,7,1,"_CPPv2N6paddle7VectorT13memoryHandle_E"],"paddle::VectorT::newMemory":[44,6,1,"_CPPv2N6paddle7VectorT9newMemoryE6size_t"],"paddle::VectorT::print":[44,6,1,"_CPPv2N6paddle7VectorT5printERNSt7ostreamE6size_t"],"paddle::VectorT::printOneElement":[44,6,1,"_CPPv2N6paddle7VectorT15printOneElementERNSt7ostreamE6size_t"],"paddle::VectorT::rand":[44,6,1,"_CPPv2N6paddle7VectorT4randEv"],"paddle::VectorT::randnorm":[44,6,1,"_CPPv2N6paddle7VectorT8randnormE4real4real"],"paddle::VectorT::reset":[44,6,1,"_CPPv2N6paddle7VectorT5resetERK1T"],"paddle::VectorT::resize":[44,6,1,"_CPPv2N6paddle7VectorT6resizeE6size_t"],"paddle::VectorT::resizeOrCreate":[44,6,1,"_CPPv2N6paddle7VectorT14resizeOrCreateERNSt10shared_ptrI7VectorTI1TEEE6size_tb"],"paddle::VectorT::selectFrom":[44,6,1,"_CPPv2N6paddle7VectorT10selectFromERK7VectorTI1TERK7VectorTIiE"],"paddle::VectorT::setElement":[44,6,1,"_CPPv2N6paddle7VectorT10setElementE6size_tRK1T"],"paddle::VectorT::subVec":[44,6,1,"_CPPv2N6paddle7VectorT6subVecE6size_t6size_t"],"paddle::VectorT::subVecFrom":[44,6,1,"_CPPv2N6paddle7VectorT10subVecFromERK7VectorTI1TE6size_t6size_t"],"paddle::VectorT::uniform":[44,6,1,"_CPPv2N6paddle7VectorT7uniformE4real4real"],"paddle::VectorT::zeroMem":[44,6,1,"_CPPv2N6paddle7VectorT7zeroMemEv"],"paddle::VectorT::~VectorT":[44,6,1,"_CPPv2N6paddle7VectorTD0Ev"],"paddle::Weight":[50,8,1,"_CPPv2N6paddle6WeightE"],"paddle::Weight::Weight":[50,6,1,"_CPPv2N6paddle6Weight6WeightE6size_t6size_t12ParameterPtr"],"paddle::Weight::getParameterPtr":[50,6,1,"_CPPv2N6paddle6Weight15getParameterPtrEv"],"paddle::Weight::getW":[50,6,1,"_CPPv2N6paddle6Weight4getWEv"],"paddle::Weight::getWGrad":[50,6,1,"_CPPv2N6paddle6Weight8getWGradEv"],"paddle::Weight::incUpdate":[50,6,1,"_CPPv2N6paddle6Weight9incUpdateERK14UpdateCallback"],"paddle::Weight::parameter_":[50,7,1,"_CPPv2N6paddle6Weight10parameter_E"],"paddle::Weight::setParameterPtr":[50,6,1,"_CPPv2N6paddle6Weight15setParameterPtrE12ParameterPtr"],"paddle::Weight::weightGrad_":[50,7,1,"_CPPv2N6paddle6Weight11weightGrad_E"],"paddle::Weight::weight_":[50,7,1,"_CPPv2N6paddle6Weight7weight_E"],"paddle::WeightList":[50,9,1,"_CPPv2N6paddle10WeightListE"],"paddle::__attribute__":[52,7,1,"_CPPv2N6paddle13__attribute__E"],"paddle::axpy":[44,6,1,"_CPPv2N6paddle4axpyEKiK1TPK1TP1T"],"paddle::bool_constant":[44,8,1,"_CPPv2N6paddle13bool_constantE"],"paddle::bool_constant::value":[44,7,1,"_CPPv2N6paddle13bool_constant5valueE"],"paddle::dotProduct":[44,6,1,"_CPPv2N6paddle10dotProductEKiPK1TPK1T"],"paddle::enumeration_wrapper":[61,9,1,"_CPPv2N6paddle19enumeration_wrapperE"],"paddle::enumeration_wrapper::NUM_PARAMETER_TYPES":[61,8,1,"_CPPv2N6paddle19enumeration_wrapper19NUM_PARAMETER_TYPESE"],"paddle::enumeration_wrapper::PARAMETER_APPLY":[61,8,1,"_CPPv2N6paddle19enumeration_wrapper15PARAMETER_APPLYE"],"paddle::enumeration_wrapper::PARAMETER_COLS":[61,8,1,"_CPPv2N6paddle19enumeration_wrapper14PARAMETER_COLSE"],"paddle::enumeration_wrapper::PARAMETER_DELTA":[61,8,1,"_CPPv2N6paddle19enumeration_wrapper15PARAMETER_DELTAE"],"paddle::enumeration_wrapper::PARAMETER_GRADIENT":[61,8,1,"_CPPv2N6paddle19enumeration_wrapper18PARAMETER_GRADIENTE"],"paddle::enumeration_wrapper::PARAMETER_GRADIENT_SQURESUM":[61,8,1,"_CPPv2N6paddle19enumeration_wrapper27PARAMETER_GRADIENT_SQURESUME"],"paddle::enumeration_wrapper::PARAMETER_GRADIENT_SQURESUM1":[61,8,1,"_CPPv2N6paddle19enumeration_wrapper28PARAMETER_GRADIENT_SQURESUM1E"],"paddle::enumeration_wrapper::PARAMETER_GRADIENT_SUM":[61,8,1,"_CPPv2N6paddle19enumeration_wrapper22PARAMETER_GRADIENT_SUME"],"paddle::enumeration_wrapper::PARAMETER_LEARNING_RATE":[61,8,1,"_CPPv2N6paddle19enumeration_wrapper23PARAMETER_LEARNING_RATEE"],"paddle::enumeration_wrapper::PARAMETER_MOMENTUM":[61,8,1,"_CPPv2N6paddle19enumeration_wrapper18PARAMETER_MOMENTUME"],"paddle::enumeration_wrapper::PARAMETER_MOMENTUM_UT":[61,8,1,"_CPPv2N6paddle19enumeration_wrapper21PARAMETER_MOMENTUM_UTE"],"paddle::enumeration_wrapper::PARAMETER_MOMENTUM_VT":[61,8,1,"_CPPv2N6paddle19enumeration_wrapper21PARAMETER_MOMENTUM_VTE"],"paddle::enumeration_wrapper::PARAMETER_ROWS":[61,8,1,"_CPPv2N6paddle19enumeration_wrapper14PARAMETER_ROWSE"],"paddle::enumeration_wrapper::PARAMETER_SECOND_MOMENTUM":[61,8,1,"_CPPv2N6paddle19enumeration_wrapper25PARAMETER_SECOND_MOMENTUME"],"paddle::enumeration_wrapper::PARAMETER_SUM1":[61,8,1,"_CPPv2N6paddle19enumeration_wrapper14PARAMETER_SUM1E"],"paddle::enumeration_wrapper::PARAMETER_SUM2":[61,8,1,"_CPPv2N6paddle19enumeration_wrapper14PARAMETER_SUM2E"],"paddle::enumeration_wrapper::PARAMETER_SUM3":[61,8,1,"_CPPv2N6paddle19enumeration_wrapper14PARAMETER_SUM3E"],"paddle::enumeration_wrapper::PARAMETER_UPDATE_TIME":[61,8,1,"_CPPv2N6paddle19enumeration_wrapper21PARAMETER_UPDATE_TIMEE"],"paddle::enumeration_wrapper::PARAMETER_VALUE":[61,8,1,"_CPPv2N6paddle19enumeration_wrapper15PARAMETER_VALUEE"],"paddle::enumeration_wrapper::PARAMETER_WEIGHTED_INFINITY_NORM":[61,8,1,"_CPPv2N6paddle19enumeration_wrapper32PARAMETER_WEIGHTED_INFINITY_NORME"],"paddle::enumeration_wrapper::PASS_GC":[61,8,1,"_CPPv2N6paddle19enumeration_wrapper7PASS_GCE"],"paddle::enumeration_wrapper::PASS_METRIC":[61,8,1,"_CPPv2N6paddle19enumeration_wrapper11PASS_METRICE"],"paddle::enumeration_wrapper::PASS_METRIC_TRAIN":[61,8,1,"_CPPv2N6paddle19enumeration_wrapper17PASS_METRIC_TRAINE"],"paddle::enumeration_wrapper::PASS_METRIC_TRAIN_WITH_NOERROR":[61,8,1,"_CPPv2N6paddle19enumeration_wrapper30PASS_METRIC_TRAIN_WITH_NOERRORE"],"paddle::enumeration_wrapper::PASS_TEST":[61,8,1,"_CPPv2N6paddle19enumeration_wrapper9PASS_TESTE"],"paddle::enumeration_wrapper::PASS_TRAIN":[61,8,1,"_CPPv2N6paddle19enumeration_wrapper10PASS_TRAINE"],"paddle::enumeration_wrapper::ParameterType":[61,9,1,"_CPPv2N6paddle19enumeration_wrapper13ParameterTypeE"],"paddle::enumeration_wrapper::PassType":[61,9,1,"_CPPv2N6paddle19enumeration_wrapper8PassTypeE"],"paddle::false_type":[44,9,1,"_CPPv2N6paddle10false_typeE"],"paddle::findLastSet":[46,6,1,"_CPPv2N6paddle11findLastSetE1T"],"paddle::gemm":[44,6,1,"_CPPv2N6paddle4gemmEK15CBLAS_TRANSPOSEK15CBLAS_TRANSPOSEKiKiKiK1TPK1TKiPK1TKiK1TP1TKi"],"paddle::hl_sparse_matrix_s_ptr":[44,9,1,"_CPPv2N6paddle22hl_sparse_matrix_s_ptrE"],"paddle::operator<<":[44,6,1,"_CPPv2N6paddlelsERNSt7ostreamERK6Matrix"],"paddle::simd":[44,9,1,"_CPPv2N6paddle4simdE"],"paddle::simd::addTo":[44,6,1,"_CPPv2N6paddle4simd5addToEP4TypePK4Type6size_t"],"paddle::simd::batchAddTo":[44,6,1,"_CPPv2N6paddle4simd10batchAddToEP4TypeA_PK4Typei6size_t"],"paddle::simd::colMax":[44,6,1,"_CPPv2N6paddle4simd6colMaxEP4TypePK4Typeii"],"paddle::simd::decayL1":[44,6,1,"_CPPv2N6paddle4simd7decayL1EP4TypeP4TypeP4Type4Type6size_t"],"paddle::simd::isPointerAlign":[44,6,1,"_CPPv2N6paddle4simd14isPointerAlignEPv"],"paddle::simd::naive":[44,9,1,"_CPPv2N6paddle4simd5naiveE"],"paddle::simd::naive::addTo":[44,6,1,"_CPPv2N6paddle4simd5naive5addToEP4TypePK4Type6size_t"],"paddle::simd::naive::batchAddTo":[44,6,1,"_CPPv2N6paddle4simd5naive10batchAddToEP4TypeA_PK4Typei6size_t"],"paddle::simd::naive::colMax":[44,6,1,"_CPPv2N6paddle4simd5naive6colMaxEP4TypePK4Typeii"],"paddle::simd::naive::decayL1":[44,6,1,"_CPPv2N6paddle4simd5naive7decayL1EP4TypeP4TypeP4Type4Type6size_t"],"paddle::simd::vec_check":[44,6,1,"_CPPv2N6paddle4simd9vec_checkE6size_t"],"paddle::sparse_float_value_t":[44,8,1,"_CPPv2N6paddle20sparse_float_value_tE"],"paddle::sparse_float_value_t::col":[44,7,1,"_CPPv2N6paddle20sparse_float_value_t3colE"],"paddle::sparse_float_value_t::value":[44,7,1,"_CPPv2N6paddle20sparse_float_value_t5valueE"],"paddle::sparse_non_value_t":[44,8,1,"_CPPv2N6paddle18sparse_non_value_tE"],"paddle::sparse_non_value_t::col":[44,7,1,"_CPPv2N6paddle18sparse_non_value_t3colE"],"paddle::true_type":[44,9,1,"_CPPv2N6paddle9true_typeE"],Arguments:[24,8,1,"_CPPv29Arguments"],CAL_MATRIX_START_ADDRESS:[44,0,1,"c.CAL_MATRIX_START_ADDRESS"],CREATE_MODE_NORMAL:[24,8,1,"_CPPv218CREATE_MODE_NORMAL"],CREATE_MODE_TESTING:[24,8,1,"_CPPv219CREATE_MODE_TESTING"],DISABLE_COPY_AND_ASSIGN:[24,0,1,"c.DISABLE_COPY_AND_ASSIGN"],DIVUP:[32,0,1,"c.DIVUP"],FloatArray:[24,8,1,"_CPPv210FloatArray"],GetCublasDsoHandle:[25,6,1,"_CPPv218GetCublasDsoHandlePPv"],GetCudartDsoHandle:[25,6,1,"_CPPv218GetCudartDsoHandlePPv"],GetCudnnDsoHandle:[25,6,1,"_CPPv217GetCudnnDsoHandlePPv"],GetCurandDsoHandle:[25,6,1,"_CPPv218GetCurandDsoHandlePPv"],GradientMachine:[24,8,1,"_CPPv215GradientMachine"],GradientMatchineCreateMode:[24,9,1,"_CPPv226GradientMatchineCreateMode"],HL_ACTIVATION_END:[32,8,1,"_CPPv217HL_ACTIVATION_END"],HL_ACTIVATION_LINEAR:[32,8,1,"_CPPv220HL_ACTIVATION_LINEAR"],HL_ACTIVATION_RELU:[32,8,1,"_CPPv218HL_ACTIVATION_RELU"],HL_ACTIVATION_SIGMOID:[32,8,1,"_CPPv221HL_ACTIVATION_SIGMOID"],HL_ACTIVATION_TANH:[32,8,1,"_CPPv218HL_ACTIVATION_TANH"],HL_FLOAT_MAX:[32,0,1,"c.HL_FLOAT_MAX"],HL_FLOAT_MIN:[32,0,1,"c.HL_FLOAT_MIN"],HL_FLOAT_VALUE:[32,8,1,"_CPPv214HL_FLOAT_VALUE"],HL_NO_VALUE:[32,8,1,"_CPPv211HL_NO_VALUE"],HL_POOLING_AVERAGE:[25,8,1,"_CPPv218HL_POOLING_AVERAGE"],HL_POOLING_AVERAGE_EXCLUDE_PADDING:[25,8,1,"_CPPv234HL_POOLING_AVERAGE_EXCLUDE_PADDING"],HL_POOLING_END:[25,8,1,"_CPPv214HL_POOLING_END"],HL_POOLING_MAX:[25,8,1,"_CPPv214HL_POOLING_MAX"],HL_SPARSE_CSC:[32,8,1,"_CPPv213HL_SPARSE_CSC"],HL_SPARSE_CSR:[32,8,1,"_CPPv213HL_SPARSE_CSR"],HL_SPARSE_END:[32,8,1,"_CPPv213HL_SPARSE_END"],HL_VALUE_END:[32,8,1,"_CPPv212HL_VALUE_END"],HPPL_ACTIVE_FUNCTION:[30,0,1,"c.HPPL_ACTIVE_FUNCTION"],HPPL_OP_END:[32,8,1,"_CPPv211HPPL_OP_END"],HPPL_OP_N:[32,8,1,"_CPPv29HPPL_OP_N"],HPPL_OP_T:[32,8,1,"_CPPv29HPPL_OP_T"],HPPL_STREAM_1:[32,8,1,"_CPPv213HPPL_STREAM_1"],HPPL_STREAM_2:[32,8,1,"_CPPv213HPPL_STREAM_2"],HPPL_STREAM_3:[32,8,1,"_CPPv213HPPL_STREAM_3"],HPPL_STREAM_4:[32,8,1,"_CPPv213HPPL_STREAM_4"],HPPL_STREAM_DEFAULT:[32,8,1,"_CPPv219HPPL_STREAM_DEFAULT"],HPPL_STREAM_END:[32,8,1,"_CPPv215HPPL_STREAM_END"],HPPL_THREAD_STREAM_1:[32,8,1,"_CPPv220HPPL_THREAD_STREAM_1"],HPPL_THREAD_STREAM_2:[32,8,1,"_CPPv220HPPL_THREAD_STREAM_2"],HPPL_THREAD_STREAM_3:[32,8,1,"_CPPv220HPPL_THREAD_STREAM_3"],HPPL_THREAD_STREAM_4:[32,8,1,"_CPPv220HPPL_THREAD_STREAM_4"],IOError:[24,8,1,"_CPPv27IOError"],ISequenceResults:[24,8,1,"_CPPv216ISequenceResults"],IVector:[24,8,1,"_CPPv27IVector"],IntArray:[24,8,1,"_CPPv28IntArray"],IntWithFloatArray:[24,8,1,"_CPPv217IntWithFloatArray"],Matrix:[24,8,1,"_CPPv26Matrix"],ModelConfig:[24,8,1,"_CPPv211ModelConfig"],NO_SPARSE_ID:[24,7,1,"_CPPv212NO_SPARSE_ID"],OptimizationConfig:[24,8,1,"_CPPv218OptimizationConfig"],P_DECLARE_bool:[44,6,1,"_CPPv214P_DECLARE_bool31allow_inefficient_sparse_update"],Parameter:[24,8,1,"_CPPv29Parameter"],ParameterConfig:[24,8,1,"_CPPv215ParameterConfig"],ParameterOptimizer:[24,8,1,"_CPPv218ParameterOptimizer"],ParameterTraverseCallback:[24,8,1,"_CPPv225ParameterTraverseCallback"],RangeError:[24,8,1,"_CPPv210RangeError"],SIGMOID_THRESHOLD_MAX:[30,0,1,"c.SIGMOID_THRESHOLD_MAX"],SIGMOID_THRESHOLD_MIN:[30,0,1,"c.SIGMOID_THRESHOLD_MIN"],SPARSE_CSC:[24,8,1,"_CPPv210SPARSE_CSC"],SPARSE_CSR:[24,8,1,"_CPPv210SPARSE_CSR"],SPARSE_NON_VALUE:[24,8,1,"_CPPv216SPARSE_NON_VALUE"],SPARSE_VALUE:[24,8,1,"_CPPv212SPARSE_VALUE"],SequenceGenerator:[24,8,1,"_CPPv217SequenceGenerator"],SparseFormatType:[24,9,1,"_CPPv216SparseFormatType"],SparseValueType:[24,9,1,"_CPPv215SparseValueType"],Trainer:[24,8,1,"_CPPv27Trainer"],TrainerConfig:[24,8,1,"_CPPv213TrainerConfig"],UnsupportError:[24,8,1,"_CPPv214UnsupportError"],UpdateCallback:[24,8,1,"_CPPv214UpdateCallback"],Vector:[24,8,1,"_CPPv26Vector"],batchTranspose:[28,6,1,"_CPPv214batchTransposePK4realP4realiii"],getCurrentTimeStick:[32,6,1,"_CPPv219getCurrentTimeStickv"],hl_CMRNorm_backward:[30,6,1,"_CPPv219hl_CMRNorm_backward6size_tPK4realPK4realPK4realPK4realP4real6size_t6size_t6size_t6size_t4real4real"],hl_CMRNorm_forward:[30,6,1,"_CPPv218hl_CMRNorm_forward6size_tPK4realP4realP4real6size_t6size_t6size_t6size_t4real4real"],hl_activation_mode_t:[32,9,1,"_CPPv220hl_activation_mode_t"],hl_avgpool_backward:[30,6,1,"_CPPv219hl_avgpool_backwardiPK4realiiiiiiiiP4real4real4real"],hl_avgpool_forward:[30,6,1,"_CPPv218hl_avgpool_forwardiPK4realiiiiiiiiP4real"],hl_batch_norm_backward:[25,6,1,"_CPPv222hl_batch_norm_backward20hl_tensor_descriptorP4real20hl_tensor_descriptorP4real20hl_tensor_descriptorP4real20hl_tensor_descriptorP4realP4realP4realdP4realP4real"],hl_batch_norm_forward_inference:[25,6,1,"_CPPv231hl_batch_norm_forward_inference20hl_tensor_descriptorP4real20hl_tensor_descriptorP4real20hl_tensor_descriptorP4realP4realP4realP4reald"],hl_batch_norm_forward_training:[25,6,1,"_CPPv230hl_batch_norm_forward_training20hl_tensor_descriptorP4real20hl_tensor_descriptorP4real20hl_tensor_descriptorP4realP4realdP4realP4realdP4realP4real"],hl_construct_sparse_matrix:[28,6,1,"_CPPv226hl_construct_sparse_matrixP18hl_sparse_matrix_sPv6size_t18hl_matrix_format_t17hl_matrix_value_tiii"],hl_context_projection_backward_data:[30,6,1,"_CPPv235hl_context_projection_backward_dataP4realPKiP4realiiii"],hl_context_projection_backward_weight:[30,6,1,"_CPPv237hl_context_projection_backward_weightP4realPKiP4realiiiiii"],hl_context_projection_forward:[30,6,1,"_CPPv229hl_context_projection_forwardP4realPKiP4realP4realiiiiib"],hl_conv_workspace:[25,6,1,"_CPPv217hl_conv_workspace20hl_tensor_descriptor20hl_tensor_descriptor20hl_filter_descriptor25hl_convolution_descriptorPiP6size_tPiP6size_tPiP6size_t"],hl_convolution_backward_bias:[25,6,1,"_CPPv228hl_convolution_backward_bias20hl_tensor_descriptorP4real20hl_tensor_descriptorP4real"],hl_convolution_backward_data:[25,6,1,"_CPPv228hl_convolution_backward_data20hl_tensor_descriptorP4real20hl_tensor_descriptorP4real20hl_filter_descriptorP4real25hl_convolution_descriptorPv6size_ti"],hl_convolution_backward_filter:[25,6,1,"_CPPv230hl_convolution_backward_filter20hl_tensor_descriptorP4real20hl_tensor_descriptorP4real20hl_filter_descriptorP4real25hl_convolution_descriptorPv6size_ti"],hl_convolution_descriptor:[25,9,1,"_CPPv225hl_convolution_descriptor"],hl_convolution_forward:[25,6,1,"_CPPv222hl_convolution_forward20hl_tensor_descriptorP4real20hl_tensor_descriptorP4real20hl_filter_descriptorP4real25hl_convolution_descriptorPv6size_ti"],hl_convolution_forward_add_bias:[25,6,1,"_CPPv231hl_convolution_forward_add_bias20hl_tensor_descriptorP4real20hl_tensor_descriptorP4real"],hl_cossim:[28,6,1,"_CPPv29hl_cossimP4realP4realP4realiii4real"],hl_cossim_derivative:[28,6,1,"_CPPv220hl_cossim_derivativeP4realP4realP4realP4realP4realP4realiii4real"],hl_create_convolution_descriptor:[25,6,1,"_CPPv232hl_create_convolution_descriptorP25hl_convolution_descriptor20hl_tensor_descriptor20hl_filter_descriptoriiii"],hl_create_event:[25,6,1,"_CPPv215hl_create_eventP10hl_event_t"],hl_create_filter_descriptor:[25,6,1,"_CPPv227hl_create_filter_descriptorP20hl_filter_descriptoriiii"],hl_create_pooling_descriptor:[25,6,1,"_CPPv228hl_create_pooling_descriptorP21hl_pooling_descriptor17hl_pooling_mode_tiiiiii"],hl_create_tensor_descriptor:[25,6,1,"_CPPv227hl_create_tensor_descriptorP20hl_tensor_descriptor"],hl_cuda_event_query:[25,6,1,"_CPPv219hl_cuda_event_query10hl_event_tRb"],hl_destroy_convolution_descriptor:[25,6,1,"_CPPv233hl_destroy_convolution_descriptor25hl_convolution_descriptor"],hl_destroy_event:[25,6,1,"_CPPv216hl_destroy_event10hl_event_t"],hl_destroy_filter_descriptor:[25,6,1,"_CPPv228hl_destroy_filter_descriptor20hl_filter_descriptor"],hl_destroy_pooling_descriptor:[25,6,1,"_CPPv229hl_destroy_pooling_descriptor21hl_pooling_descriptor"],hl_destroy_tensor_descriptor:[25,6,1,"_CPPv228hl_destroy_tensor_descriptor20hl_tensor_descriptor"],hl_destruct_sparse_matrix:[28,6,1,"_CPPv225hl_destruct_sparse_matrix18hl_sparse_matrix_s"],hl_device_can_access_peer:[25,6,1,"_CPPv225hl_device_can_access_peerii"],hl_device_enable_peer_access:[25,6,1,"_CPPv228hl_device_enable_peer_accessi"],hl_device_synchronize:[25,6,1,"_CPPv221hl_device_synchronizev"],hl_event_elapsed_time:[25,6,1,"_CPPv221hl_event_elapsed_time10hl_event_t10hl_event_t"],hl_event_synchronize:[25,6,1,"_CPPv220hl_event_synchronize10hl_event_t"],hl_event_t:[25,9,1,"_CPPv210hl_event_t"],hl_expand_feature2col:[30,6,1,"_CPPv221hl_expand_feature2colPK4real6size_t6size_t6size_t6size_t6size_t6size_t6size_t6size_t6size_t6size_t6size_tP4real"],hl_filter_descriptor:[25,9,1,"_CPPv220hl_filter_descriptor"],hl_fini:[25,6,1,"_CPPv27hl_finiv"],hl_free_mem_device:[25,6,1,"_CPPv218hl_free_mem_devicePv"],hl_free_mem_host:[25,6,1,"_CPPv216hl_free_mem_hostPv"],hl_free_sparse_matrix:[28,6,1,"_CPPv221hl_free_sparse_matrix18hl_sparse_matrix_s"],hl_get_cuda_lib_version:[25,6,1,"_CPPv223hl_get_cuda_lib_versionv"],hl_get_cudnn_lib_version:[25,6,1,"_CPPv224hl_get_cudnn_lib_versionv"],hl_get_device:[25,6,1,"_CPPv213hl_get_devicev"],hl_get_device_count:[25,6,1,"_CPPv219hl_get_device_countv"],hl_get_device_error_string:[25,6,1,"_CPPv226hl_get_device_error_stringv"],hl_get_device_last_error:[25,6,1,"_CPPv224hl_get_device_last_errorv"],hl_get_sync_flag:[25,6,1,"_CPPv216hl_get_sync_flagv"],hl_gru_grad:[32,8,1,"_CPPv211hl_gru_grad"],hl_gru_value:[32,8,1,"_CPPv212hl_gru_value"],hl_init:[25,6,1,"_CPPv27hl_initi"],hl_lstm_grad:[32,8,1,"_CPPv212hl_lstm_grad"],hl_lstm_parallel_backward_data:[30,6,1,"_CPPv230hl_lstm_parallel_backward_dataP4realP4realP4realP4realP4realP4realP4realP4realP4realP4realP4realP4realP4realP4realPKiiib20hl_activation_mode_t20hl_activation_mode_t20hl_activation_mode_t"],hl_lstm_parallel_backward_weight:[30,6,1,"_CPPv232hl_lstm_parallel_backward_weightP4realP4realP4realPKiiiib"],hl_lstm_parallel_forward:[30,6,1,"_CPPv224hl_lstm_parallel_forwardP4realP4realP4realP4realP4realP4realP4realP4realPKiiib20hl_activation_mode_t20hl_activation_mode_t20hl_activation_mode_t"],hl_lstm_value:[32,8,1,"_CPPv213hl_lstm_value"],hl_malloc_device:[25,6,1,"_CPPv216hl_malloc_device6size_t"],hl_malloc_host:[25,6,1,"_CPPv214hl_malloc_host6size_t"],hl_malloc_sparse_matrix:[28,6,1,"_CPPv223hl_malloc_sparse_matrixP18hl_sparse_matrix_s18hl_matrix_format_t17hl_matrix_value_tiii"],hl_matrix_add:[28,6,1,"_CPPv213hl_matrix_addP4realP4realP4realii4real4real"],hl_matrix_add_to_rows:[28,6,1,"_CPPv221hl_matrix_add_to_rowsP4realiP4realiPiiii"],hl_matrix_classification_error:[28,6,1,"_CPPv230hl_matrix_classification_errorP4realPiP4realii"],hl_matrix_column_max:[28,6,1,"_CPPv220hl_matrix_column_maxP4realP4realii"],hl_matrix_column_min:[28,6,1,"_CPPv220hl_matrix_column_minP4realP4realii"],hl_matrix_column_sum:[28,6,1,"_CPPv220hl_matrix_column_sumP4realP4realii"],hl_matrix_cross_entropy:[28,6,1,"_CPPv223hl_matrix_cross_entropyP4realP4realPiii"],hl_matrix_cross_entropy_bp:[28,6,1,"_CPPv226hl_matrix_cross_entropy_bpP4realP4realPiii"],hl_matrix_csc2dense:[28,6,1,"_CPPv219hl_matrix_csc2dense18hl_sparse_matrix_sP4realii"],hl_matrix_csc_mul_dense:[28,6,1,"_CPPv223hl_matrix_csc_mul_dense18hl_sparse_matrix_s13hl_trans_op_tP4real13hl_trans_op_tP4realiii4real4real"],hl_matrix_csr2dense:[28,6,1,"_CPPv219hl_matrix_csr2dense18hl_sparse_matrix_sP4realii"],hl_matrix_csr_add_bias:[28,6,1,"_CPPv222hl_matrix_csr_add_bias18hl_sparse_matrix_sP4real4real"],hl_matrix_csr_add_dense:[28,6,1,"_CPPv223hl_matrix_csr_add_dense18hl_sparse_matrix_sP4realii4real4real"],hl_matrix_csr_column_sum:[28,6,1,"_CPPv224hl_matrix_csr_column_sumP4real18hl_sparse_matrix_sii4real"],hl_matrix_csr_mul_dense:[28,6,1,"_CPPv223hl_matrix_csr_mul_dense18hl_sparse_matrix_s13hl_trans_op_tP4real13hl_trans_op_tP4realiii4real4real"],hl_matrix_dense_mul_csc:[28,6,1,"_CPPv223hl_matrix_dense_mul_cscP4real13hl_trans_op_t18hl_sparse_matrix_s13hl_trans_op_tP4realiii4real4real"],hl_matrix_dense_mul_csr:[28,6,1,"_CPPv223hl_matrix_dense_mul_csrP4real13hl_trans_op_t18hl_sparse_matrix_s13hl_trans_op_tP4realiii4real4real"],hl_matrix_format_t:[32,9,1,"_CPPv218hl_matrix_format_t"],hl_matrix_mul:[25,6,1,"_CPPv213hl_matrix_mulP4real13hl_trans_op_tP4real13hl_trans_op_tP4realiii4real4realiii"],hl_matrix_mul_vector:[25,6,1,"_CPPv220hl_matrix_mul_vectorP4real13hl_trans_op_tP4realP4realii4real4realiii"],hl_matrix_row_max:[28,6,1,"_CPPv217hl_matrix_row_maxP4realP4realii"],hl_matrix_row_min:[28,6,1,"_CPPv217hl_matrix_row_minP4realP4realii"],hl_matrix_row_sum:[28,6,1,"_CPPv217hl_matrix_row_sumP4realP4realii"],hl_matrix_s:[32,9,1,"_CPPv211hl_matrix_s"],hl_matrix_select_rows:[28,6,1,"_CPPv221hl_matrix_select_rowsP4realiP4realiPiiii"],hl_matrix_softmax:[28,6,1,"_CPPv217hl_matrix_softmaxP4realP4realii"],hl_matrix_softmax_derivative:[28,6,1,"_CPPv228hl_matrix_softmax_derivativeP4realP4realP4realii"],hl_matrix_top_k:[28,6,1,"_CPPv215hl_matrix_top_kP4realiPiP4realiiii"],hl_matrix_transpose:[25,6,1,"_CPPv219hl_matrix_transposeP4realP4realiiii"],hl_matrix_value_t:[32,9,1,"_CPPv217hl_matrix_value_t"],hl_matrix_zero_mem:[28,6,1,"_CPPv218hl_matrix_zero_memP4reali"],hl_max_sequence_backward:[30,6,1,"_CPPv224hl_max_sequence_backwardP4realPiP4realii"],hl_max_sequence_forward:[30,6,1,"_CPPv223hl_max_sequence_forwardP4realPKiP4realPiii"],hl_maxpool_backward:[30,6,1,"_CPPv219hl_maxpool_backwardiPK4realPK4realPK4realiiiiiiiiP4real4real4real"],hl_maxpool_forward:[30,6,1,"_CPPv218hl_maxpool_forwardiPK4realiiiiiiiiP4real"],hl_memcpy:[25,6,1,"_CPPv29hl_memcpyPvPv6size_t"],hl_memcpy_async:[25,6,1,"_CPPv215hl_memcpy_asyncPvPv6size_t11hl_stream_t"],hl_memcpy_csc_matrix:[28,6,1,"_CPPv220hl_memcpy_csc_matrix18hl_sparse_matrix_sP4realPiPi11hl_stream_t"],hl_memcpy_csr_matrix:[28,6,1,"_CPPv220hl_memcpy_csr_matrix18hl_sparse_matrix_sP4realPiPi11hl_stream_t"],hl_memcpy_device2device:[25,6,1,"_CPPv223hl_memcpy_device2devicePvPv6size_t"],hl_memcpy_device2host:[25,6,1,"_CPPv221hl_memcpy_device2hostPvPv6size_t"],hl_memcpy_from_csc_matrix:[28,6,1,"_CPPv225hl_memcpy_from_csc_matrixP4real6size_tPi6size_tPi6size_t18hl_sparse_matrix_s11hl_stream_t"],hl_memcpy_from_csr_matrix:[28,6,1,"_CPPv225hl_memcpy_from_csr_matrixP4real6size_tPi6size_tPi6size_t18hl_sparse_matrix_s11hl_stream_t"],hl_memcpy_host2device:[25,6,1,"_CPPv221hl_memcpy_host2devicePvPv6size_t"],hl_memcpy_sparse_matrix:[28,6,1,"_CPPv223hl_memcpy_sparse_matrix18hl_sparse_matrix_s18hl_sparse_matrix_s11hl_stream_t"],hl_memset_device:[25,6,1,"_CPPv216hl_memset_devicePvi6size_t"],hl_param_relu_backward_diff:[28,6,1,"_CPPv227hl_param_relu_backward_diffP4realP4realP4realP4realiii"],hl_param_relu_backward_w:[28,6,1,"_CPPv224hl_param_relu_backward_wP4realP4realP4realiii"],hl_param_relu_forward:[28,6,1,"_CPPv221hl_param_relu_forwardP4realP4realP4realiii"],hl_pooling_backward:[25,6,1,"_CPPv219hl_pooling_backward20hl_tensor_descriptorP4realP4real20hl_tensor_descriptorP4realP4real21hl_pooling_descriptor"],hl_pooling_descriptor:[25,9,1,"_CPPv221hl_pooling_descriptor"],hl_pooling_forward:[25,6,1,"_CPPv218hl_pooling_forward20hl_tensor_descriptorP4real20hl_tensor_descriptorP4real21hl_pooling_descriptor"],hl_pooling_mode_t:[25,9,1,"_CPPv217hl_pooling_mode_t"],hl_rand:[25,6,1,"_CPPv27hl_randP4real6size_t"],hl_reset_convolution_descriptor:[25,6,1,"_CPPv231hl_reset_convolution_descriptor25hl_convolution_descriptor20hl_tensor_descriptor20hl_filter_descriptoriiii"],hl_sequence2batch_add:[30,6,1,"_CPPv221hl_sequence2batch_addP4realP4realPiiib"],hl_sequence2batch_copy:[30,6,1,"_CPPv222hl_sequence2batch_copyP4realP4realPiiib"],hl_sequence_avg_forward:[30,6,1,"_CPPv223hl_sequence_avg_forwardP4realP4realPKiiiKi"],hl_sequence_softmax_forward:[28,6,1,"_CPPv227hl_sequence_softmax_forwardP4realP4realPKii"],hl_set_device:[25,6,1,"_CPPv213hl_set_devicei"],hl_set_device_flags_block:[25,6,1,"_CPPv225hl_set_device_flags_blockv"],hl_set_sync_flag:[25,6,1,"_CPPv216hl_set_sync_flagb"],hl_shrink_col2feature:[30,6,1,"_CPPv221hl_shrink_col2featurePK4real6size_t6size_t6size_t6size_t6size_t6size_t6size_t6size_t6size_t6size_t6size_tP4real4real4real"],hl_softmax_backward:[25,6,1,"_CPPv219hl_softmax_backwardP4realP4realii"],hl_softmax_forward:[25,6,1,"_CPPv218hl_softmax_forwardP4realP4realii"],hl_sparse_matrix_add_bias:[28,6,1,"_CPPv225hl_sparse_matrix_add_bias18hl_sparse_matrix_sP4real4real"],hl_sparse_matrix_add_dense:[28,6,1,"_CPPv226hl_sparse_matrix_add_dense18hl_sparse_matrix_sP4realii4real4real"],hl_sparse_matrix_column_sum:[28,6,1,"_CPPv227hl_sparse_matrix_column_sumP4real18hl_sparse_matrix_sii4real"],hl_sparse_matrix_get_cols:[28,6,1,"_CPPv225hl_sparse_matrix_get_cols18hl_sparse_matrix_s"],hl_sparse_matrix_get_rows:[28,6,1,"_CPPv225hl_sparse_matrix_get_rows18hl_sparse_matrix_s"],hl_sparse_matrix_get_value:[28,6,1,"_CPPv226hl_sparse_matrix_get_value18hl_sparse_matrix_s"],hl_sparse_matrix_mul:[28,6,1,"_CPPv220hl_sparse_matrix_mulP4real13hl_trans_op_tP4real13hl_trans_op_t18hl_sparse_matrix_siii4real4real"],hl_sparse_matrix_s:[32,9,1,"_CPPv218hl_sparse_matrix_s"],hl_sparse_matrix_top_k:[28,6,1,"_CPPv222hl_sparse_matrix_top_kP4realiPi18hl_sparse_matrix_sii"],hl_specify_devices_start:[25,6,1,"_CPPv224hl_specify_devices_startPii"],hl_srand:[25,6,1,"_CPPv28hl_srandj"],hl_start:[25,6,1,"_CPPv28hl_startv"],hl_stream_record_event:[25,6,1,"_CPPv222hl_stream_record_event11hl_stream_t10hl_event_t"],hl_stream_synchronize:[25,6,1,"_CPPv221hl_stream_synchronize11hl_stream_t"],hl_stream_t:[32,9,1,"_CPPv211hl_stream_t"],hl_stream_wait_event:[25,6,1,"_CPPv220hl_stream_wait_event11hl_stream_t10hl_event_t"],hl_tensor_descriptor:[25,9,1,"_CPPv220hl_tensor_descriptor"],hl_tensor_reshape:[25,6,1,"_CPPv217hl_tensor_reshape20hl_tensor_descriptoriiii"],hl_trans_op_t:[32,9,1,"_CPPv213hl_trans_op_t"],hl_vector_abs_sum:[28,6,1,"_CPPv217hl_vector_abs_sumP4realP4reali"],hl_vector_select_from:[28,6,1,"_CPPv221hl_vector_select_fromP1TiPK1TiPKii"],hl_vector_sum:[28,6,1,"_CPPv213hl_vector_sumP4realP4reali"],hppl:[30,9,1,"_CPPv24hppl"],initPaddle:[24,6,1,"_CPPv210initPaddleiPPc"],isGpuVersion:[24,6,1,"_CPPv212isGpuVersionv"],paddle:[44,9,1,"_CPPv26paddle"],staticCastVector:[24,6,1,"_CPPv216staticCastVectorPNSt6vectorI2T2EERKNSt6vectorI2T1EE"]},"paddle.trainer_config_helpers":{attrs:[67,1,0,"-"],data_sources:[68,1,0,"-"]},"paddle.trainer_config_helpers.attrs":{ExtraAttr:[67,2,1,""],ExtraLayerAttribute:[67,3,1,""],ParamAttr:[67,2,1,""],ParameterAttribute:[67,3,1,""]},"paddle.trainer_config_helpers.attrs.ParameterAttribute":{set_default_parameter_name:[67,4,1,""]},"paddle.trainer_config_helpers.data_sources":{define_py_data_sources:[68,5,1,""]}},objnames:{"0":["c","macro","C macro"],"1":["py","module","Python module"],"2":["py","attribute","Python attribute"],"3":["py","class","Python class"],"4":["py","method","Python method"],"5":["py","function","Python function"],"6":["cpp","function","C++ function"],"7":["cpp","member","C++ member"],"8":["cpp","class","C++ class"],"9":["cpp","type","C++ type"]},objtypes:{"0":"c:macro","1":"py:module","2":"py:attribute","3":"py:class","4":"py:method","5":"py:function","6":"cpp:function","7":"cpp:member","8":"cpp:class","9":"cpp:type"},terms:{"0000x":13,"00186201e":86,"0473v3":74,"05d":9,"08823112e":86,"0ab":72,"0b0":5,"0th":19,"0ul":24,"10007_10":17,"10014_7":17,"100ul":24,"10gbe":3,"12194102e":86,"15501715e":86,"15mb":13,"1636k":19,"16mb":13,"173m":11,"17549435e":32,"1lu":48,"1st":[8,11,17,19,53,72],"1ul":24,"202mb":19,"211839e770f7b538e2d8":74,"2250738585072014e":32,"234m":11,"252kb":13,"25639710e":86,"25k":13,"27787406e":86,"28x28":84,"2nd":[17,19,53,72],"32777140e":86,"32x32":9,"36540484e":86,"38f":32,"3rd":[15,17,19],"40282347e":32,"40gbe":3,"43630644e":86,"48565123e":86,"48684503e":86,"49316648e":86,"51111044e":86,"53018653e":86,"55g":19,"56gbe":3,"6node":7,"6th":19,"70634608e":86,"72296313e":86,"7976931348623157e":32,"85625684e":86,"864k":19,"93137714e":86,"96644767e":86,"99982715e":86,"99m":11,"\u7ea2\u697c\u68a6":8,"\ufb01xed":19,"__attribute__":52,"__init__":[21,24],"__label":[36,41],"__m256":30,"__main__":[11,86],"__meta__":15,"__name__":[11,86],"__out":[36,41],"__queryid":36,"__repr__":72,"__rnn_step__":0,"__str__":72,"__weight":36,"_full_siz":38,"_hl_convolution_descriptor":25,"_hl_event_st":25,"_hl_filter_descriptor":25,"_hl_matrix_":32,"_hl_pooling_descriptor":25,"_hl_sparse_matrix_":[32,44],"_hl_tensor_descriptor":25,"_link":74,"_proj":72,"_res2_1_branch1_bn":11,"_source_language_embed":[0,8],"_target_language_embed":[0,8],"abstract":[21,44,53,81],"break":[13,63],"byte":[25,56,58],"case":[0,2,13,21,41,58,72],"catch":[44,48],"char":[15,24,25,44,53,64],"class":[9,17,20],"const":[21,24,25,28,30,33,34,36,38,41,44,46,48,50,52,53,56,58,59,60,62,63,64],"default":[1,7,13,15,17,19,24,25,32,38,41,53,58,62,63,64,67,69,72,74,76,81,82,84],"enum":[24,25,32,38,41,44,50,56,59,61],"export":[1,3,9],"final":[1,15,17,21,34,48,74],"float":[8,11,15,21,24,25,32,34,41,44,50,67,69,72,76,82,84],"function":[0,7],"import":[0,8,9,11,13,15,17,19,24,34,36,69,72,84,86],"int":[13,15,21,24,25,28,30,32,34,36,38,41,44,46,48,50,52,53,56,58,59,62,63,64,69,72,74,76,82,84],"long":[17,46,72,74,83],"new":[2,13,17],"null":[21,25,28,38,41,44,48,50,56,62,81],"public":[17,21,24,30,32,33,34,36,38,41,44,46,48,50,52,53,56,58,59,60,62,63,64],"return":[0,11,13,14,15,19,21,24,25,28,32,33,34,36,38,41,44,46,48,52,53,56,58,59,62,63,64,68,69,72,74,78,84],"short":[15,17,72,74],"static":[24,33,34,36,38,41,44,46,48,50,52,53,58,59,62,72],"super":21,"switch":17,"true":[0,11,15,17,19,21,24,25,30,34,36,38,41,44,50,53,59,62,63,65,67,69,72,74,81,82,84],"try":[15,38,64,76],"var":[3,25,41],"void":[21,24,25,28,30,32,33,34,36,38,41,44,46,48,50,52,53,56,58,59,60,62,63,64],"while":[0,13,17,19,44,48,58,59,60,63,67,69,81,83],a_d:[25,28],aascolvector:44,aasrowvector:44,abc:72,abl:[17,41],about:[13,17,19,41,72,74,80,81,83],abov:[11,13,33,44,50,58,59,72,84,86],absderiv:44,absolut:[7,64,83],academ:14,acceler:[41,58,59,82],accept:[13,41,56,84,86],access:0,accesssubnetwork:41,accident:14,accomplish:53,accord:[7,28,36,41,44,46,50,53,62,69,80,81,82,83,84,86],accordingli:[9,21],account:7,accumcounter_:50,accumul:[25,38,50],accumulatecolsum:44,accuraci:[13,14,17,21,36,69],achiev:[9,58,59],ack:81,acl:[17,41],aclimdb:17,acol:44,acol_:44,across:[36,41,72],act:[0,13,33,38,41,72,74],act_typ:13,actgat:41,actinput:41,action:[14,53,58,59],activ:[13,17,21,29],activation_:41,activationfunct:[33,41],activationgate_:41,activationstate_:41,active_g:30,active_nod:30,active_st:30,active_state_typ:41,active_typ:41,activegate_:41,activenode_:41,activestate_:41,activi:74,actnod:41,actstat:41,actual:[24,34,38,41,44,46,48,63],actvi:30,actvie_gate_typ:41,adadelta:[13,76],adadeltaparameteroptim:48,adagrad:[13,76],adagradparameteroptim:48,adam:[13,17,19,48,76],adamax:[13,48,76],adamaxparameteroptim:48,adamoptim:[8,13,17,19],adamparameteroptim:48,adapt:[17,19,69,76],add2:44,add3:44,add:[1,2,13,15,21,25,28,30,41,44,58,59,62,72,74,82,84],add_input:21,add_test:21,add_to:72,add_unittest_without_exec:21,addatoffset:44,addbatchjob:62,addbia:[21,44],addbias:41,addbybitcod:44,addbybitcodebackward:44,addcolscal:44,addcolumnvector:44,addcolvector:44,addcomputethread:38,addcost:59,adddotmul:44,adddotmulmmv2:44,adddotmulmmv:44,adddotmulvmm2:44,adddotmulvmm:44,addel:44,addgradbegin_:58,addgradi:58,addit:[13,28,41,50,53,58,59,72,74],addjob:62,addoptim:48,addoutputargu:41,addp2p:44,addparametertyp:[48,52],addprev:41,addr:[56,58],addr_:56,addreallay:41,address:[5,25,28,41,44,56,58,81],addrow:[21,44],addrowscal:44,addrowvector:44,addscalar:44,addsharedbia:41,addsquar:44,addsquaresum:44,addtion:7,addto:[44,72],addto_lay:41,addtolay:40,addtorow:44,addtorowsimp:44,addunsharedbia:41,adjac:41,adject:17,adjust:38,adjustbeamsearch:38,adjustprob:38,admin:14,adn:28,adopt:50,advanc:[0,81],advantag:[17,56],adventur:14,adverb:17,affect:[3,72],affin:58,afi:84,after:[0,1,5,7,9,11,13,15,17,19,21,24,32,33,38,41,44,48,50,53,58,59,62,72,81,82],afterward:53,again:50,agent:[38,41],agentlay:[38,40],agentlayernam:38,agg:44,agg_level:72,aggreg:40,aggregatelevel:72,agre:[15,86],agument:24,aim:[17,19],aircraft:19,airplan:9,aistat:[41,72],alex:[17,41,72],alexnet_pass1:82,alexnet_pass2:82,algo:[38,41],algorihtm:41,algorithm:[0,8,9],alia:[65,67],align:[19,41,46,58,72,74],align_hint:58,alignblockcount:58,alignbyt:58,alignedalloc:[44,58],alignelementcount:58,aligns:44,all:[0,2,3,7,8,11,13,14,15,17,19,21,24,25,34,38,41,44,46,48,50,53,56,58,59,62,64,67,68,69,72,76,80,81,82,84,86],all_rows_in_block:48,allbarrier_:38,allclientpassfinish_:58,allcount:50,allid:[38,41],allids_:41,alloc:[21,24,25,34,41,44,46,50,58,82],allocator_:46,allocgradbuf:38,allocsize_:46,allow:[13,21,38,41,58,59,64,81],allow_inefficient_sparse_upd:[44,80,81],allow_only_one_model_on_one_gpu:[80,81,82],allsegments_:53,almost:[7,8,53,58,74],along:17,alpha:[25,28,30,44],alpha_:[41,48],alread:59,alreadi:[7,17,38,44,50,81],alreali:[19,80],also:[0,1,3,7,9,11,13,17,21,28,38,41,53,56,58,59,69,72,74,83,84],alwai:[19,34,38,72,74,81],amaz:9,amazon:[13,17],american:9,among:17,amount:[17,41,44],analysi:12,analyz:[13,17],ancestor:41,ani:[0,2,13,15,19,24,34,36,41,64,72,74,83,84,86],anim:14,annot:58,anonym:59,anoth:[17,41,63,64,72,86],answer:36,apach:[15,86],api:[1,13,15,17,21,23],apo:19,appar:19,append:[0,7,15,21,44,84],appl:72,appli:[0,9,13,21,24,41,44,48,52,59,64,72,74],applic:[3,15,17,86],applybinari:44,applycol:44,applydecay_:48,applyl1:44,applyl1decai:44,applyl2:44,applyquaternari:44,applyrow:44,applyternari:44,applyunari:44,apport:50,appreci:[2,17],approach:72,apt:[1,3,5,9],arbitrari:[41,72],architecur:17,area:36,arg:[9,11,13,15,17,38,50,62,68,69,72,74,76,80,84],arg_nam:72,argc:24,argu:50,arguemnt:24,argugroup:50,argument:[0,8,9,11],argumentspriv:24,argv:[11,24],arithmet:44,arm:53,around:[72,84],arow:44,arow_:44,arrai:[11,24,28,34,38,52,72,86],arrang:28,arriv:58,articl:83,artist:14,arxiv:[17,41,48,74],aspect:17,assequencegener:24,assert:[24,86],assign:[38,41,44,48,59,81],assignatoffset:44,assignemnt:44,associ:[17,19],assum:[0,7,8,9,38,41,44,62,82],assur:83,async:[53,58,59,76,80],async_count:81,async_lagged_grad_discard_ratio:81,async_lagged_ratio_default:[80,81],async_lagged_ratio_min:[80,81],asynccount:50,asynccount_:50,asyncfinishpass:[53,58],asyncgrdientcommitcheckandstat:58,asynchron:[25,28,34,50,53,59,62,68,81],asynchroniz:58,asynclaggedgradientsnum_:58,asynclaggedthreshold_:58,asyncparamet:50,asyncsgd:58,asyncstartpass:53,asynctrainercommitstat_:58,asynctrainerdiscardstat_:58,asynctrainersteps_:58,asyncupdaterstep:58,asyncupdatersteps_:58,asyncupdatestat_:58,asyncupdatesteps_:58,asynsgd:58,atal:44,ateo:38,atla:1,atom:[38,58],attachcontrolparam:50,attachmajorparam:50,attachminorparam:50,attitud:17,attr:[67,74],attribut:[8,21,24,34,36,38,41,44,46,48,50,52,53,56,58,59,62,64],auc:[36,69,80],auc_evalu:36,aucvalid:40,aucvalidationlay:81,author:11,auto:[15,21,48,60,62,63],autom:19,automat:[0,1,7,15,19,21,41,44,63,72,74,80,81],automaticli:72,automobil:9,auxiliari:41,avaiabl:58,averag:[11,13,15,17,19,25,30,36,41,44,48,59,69,72,76,78,81],average_:41,average_test_period:[80,81],averagelay:40,averagelevel:41,averageoptim:48,averagepool:72,averagesparseoptim:48,averagestrategi:41,averg:30,avg:[13,41],avgcost:[13,15,17,19,59,69],avgpool:[13,72],avgpoolbackward:44,avgpoolforward:44,avgpoolprojectionlay:41,avoid:[48,84],avreag:30,avx:[5,58],axpi:44,aynchron:59,aysnc:50,b2t:8,b_col:44,b_d:[25,28],b_row:44,b_row_i:44,back:[28,38,53,58,84],backgrad:30,background:3,backward:[0,21,24,25,28,30,33,38,41,44,48,59,60,65,72,74,81,82],backward_first:0,backwardactiv:[21,41],backwardbatch:41,backwardbia:41,backwardcallback:38,backwardcallback_:38,backwardcost:41,backwarddata:41,backwardfilt:41,backwardgate2outputsequ:41,backwardimp:[38,41],backwardimpin:41,backwardoneinput:41,backwardonesequ:41,backwardseqparallel:41,backwardsequ:41,backwardvars_:41,backword:24,bad:58,bag:[13,17],baidu:[1,3,5,8,15,86],baik:8,baisparamet:41,balanc:81,balasubramanyan:17,bandwidth:58,bardward:74,bare:58,barrier:[38,58,64,81],barrier_:64,barrrier:38,bascolvector:44,base:[0,1,3,7,8,15,17,19,21,27],basecli:53,baselay:21,basematrix:[21,44],basematrixt:44,basenam:69,basepoolingtyp:[72,74],baseregular:76,basestr:[15,65,67,68,69,72,74,78],basetimer_:48,basevector:44,bash:15,bashrc:1,basi:[15,86],basic:[13,14,17,21,38,41,53,72,84],basrowvector:44,batch:[7,9,11,13,15,17,19,21,24,25,28,30,34,36,38,41,44,48,50,53,59,62,69,72,74,76,81,84],batch_0:11,batch_norm:[41,72],batch_norm_lay:41,batch_norm_typ:72,batch_siz:[7,8,9,13,15,17,19,25,41,76],batchaddto:44,batchcount:30,batchgrad_:41,batchgradientmachin:50,batchid:[41,44],batchid_:58,batchindex:30,batchnorm:[41,50],batchnormalizationlay:40,batchnormbaselay:40,batchsiz:[21,24,28,30,38,41,52,59,72],batchsize_:59,batchstartpositions_:41,batchstatu:[53,59],batchstatus_:59,batchtranspos:28,batchvalu:41,batchvalue_:41,bcd:72,bcol:44,bcol_:44,beam:[0,19,28,38,41,72,81],beam_gen:[0,72],beam_search:0,beam_siz:[0,24,72,80,81,82],beamsearch:38,beamsearchcandidatesadjustcallback:38,beamsiz:[19,24,28],becaus:[0,9,11,13,21,34,38,41,44,48,53,58,63,72,82],becom:38,been:[2,9,13,17,19,25,36,38,41,44,53,62,81,84],befor:[2,7,9,15,17,19,24,25,33,34,36,38,41,44,48,50,52,53,59,62,63,64,72,74],begin:[21,30,36,41,48,58,69,72,86],begin_dim:58,begin_id:24,begindim:50,beginn:0,beginpad:[30,44],beginpad_:41,beginpo:[44,50],behaivor:38,belong:[19,34,41,72],below:[0,7,9,13,15,21,58,72,84],benefit:74,bengio:[41,72],benifit:[53,58],bertolami:17,besid:[19,72,83],best:[1,15,17,19,24,41,68,72,81],beta1:76,beta1_:48,beta2:76,beta2_:48,beta:[11,25,28,30,41,44],beta_:[41,48],better:[7,15,41,56,58,72,74,84],between:[7,13,14,17,19,25,41,44,48,53,58,59,72,76],beyond:50,bgr:11,bia:[0,11,21,25,30,41,44,72,74,76],bias:[21,41,72],bias_:41,bias_attr:[0,72,74],bias_data:25,bias_fg:41,bias_grad_data:25,bias_ig:41,bias_o:41,bias_og:41,bias_param_attr:74,bias_r:41,bias_z:41,biasdesc_:41,biases_:[21,41],biasgrad:25,biaslay:38,biasoffset_:41,biasparamet:41,biasparameter_:[21,41],biassiz:21,bidirect:[0,16],bidirectional_lstm_net:17,big:[44,58],biggerthan:44,biggerthanscalar:44,biggest:17,bin:[1,7,15],binari:[3,8,13,17,36,41,44,69,72,84],binaryclassificationerror2:44,binaryclassificationerror:44,binarylabelcrossentropi:44,binarylabelcrossentropybp:44,bind:[56,58],bird:9,bit:[13,44,45],bitext:19,bla:1,blank_:41,bleu:18,blob:46,block:[11,17,21,25,41,44,48,53,56,58,62,63,64,72,81],block_expand:72,block_expand_lay:41,block_i:72,block_x:72,blockdim:32,blocked_:41,blockexpandlay:40,blockh:[30,44],blockh_:41,blockid:58,blockidmap_:58,blockinfo:58,blockinfos_:58,blockingqueu:62,blockkei:58,blockkeyhash:58,blocklengths_:56,blocklocks_:44,blockmap:58,blockmap_:58,blocknum:[44,50],blocknum_:44,blockoffsetmap_:58,blockseg:58,blockseq_:44,blocksiz:58,blocktravers:58,blockvec:48,blockw:[30,44],blockw_:41,blog:17,bn_bias_attr:74,bn_layer_attr:74,bn_param_attr:74,bnbia:[25,41],bnparamdesc:25,bnparamdesc_:41,bnscale:[25,41],bollen:17,bool:[13,15,17,21,24,25,30,34,36,38,41,44,48,50,52,53,56,58,59,62,63,64,65,67,68,69,72,74,76,81,82,84],bool_const:44,boot:0,boot_lay:0,bootbiaslay:41,bootlay:38,bos_id:[0,72],both:[0,11,13,17,21,28,38,41,44,50,56,65,67,72,74],bottleneck:11,bottom:17,bow:[13,17],bpropact:41,bpropbias:41,bpropsharedbia:41,bpropweight:41,branch:[2,41,72],breadth:[19,81],breath:1,brelu:[33,44,65],breluderiv:44,brendan:17,bridg:[53,59],brief:34,broadcast:53,brow:44,brow_:44,bryan:17,buf:[24,44,56],buf_:46,buffer:[30,34,38,44,50,53,56,58,59,81,84],buffersizes_:38,bufs_:50,bui:17,builder:3,built:50,bunch:13,bunk:17,burg:41,busi:64,button:[2,60],bwddataalgo_:41,bwddatalimitbyt:25,bwddatalimitbytes_:41,bwdfilteralgo_:41,bwdfilterlimitbyt:25,bwdfilterlimitbytes_:41,c_col:44,c_d:[25,28],c_h:28,c_row:44,c_row_i:44,cach:[13,15,25,34,83],cache_pass_in_mem:[13,15,84],cacherowcpumatrix:44,cachetyp:[13,15,34,84],cacul:[38,44],caffemode_:41,cal_matrix_start_address:44,calc:[36,44],calc_batch_s:84,calc_id:38,calcclientid:53,calcerror:36,calcgrad:41,calclearningr:48,calcndcg:41,calcparameterblocks:53,calcuat:58,calcul:[0,15,21,25,28,36,38,41,44,46,50,53,58,69,72,74,76,81,82,84],calfeaturemaps:41,call:[0,9,11,13,17,19,21,25,28,32,33,34,36,38,41,44,48,50,53,56,58,59,62,64,72,74,81,84],callabl:72,callback:[21,24,38,41,48,50,56,58,60],callbackvec_:58,caller:[38,44,48,50,62],calmeanandstd:41,calmovingmeanandvar:41,caltech:9,calul:44,can:[0,1,2,3,4,5,7,8,9,11,13,15,17,19,21,24,28,32,34,36,38,41,44,46,48,52,53,56,58,59,60,62,63,64,65,67,68,69,72,74,80,81,82,83,84,86],can_over_batch_s:84,candid:[38,41,50,72],cannot:[21,24,38,58],caoi:19,capabl:[17,25,38],capac:63,capapc:62,caption:19,captur:7,card:[7,50],care:[14,74,80,81],carefulli:[7,11,62,81],cascolvector:44,casrowvector:44,cast:24,cat:[9,11,17],catchupwith:[48,52],categori:[13,17,72],cblas_transpos:44,ccb2_pc30:19,ccol:44,ccol_:44,cde:72,ceil:32,cell:[17,41,72,74],center:84,cento:3,certain:[80,83],chain:21,chanc:[13,21],chang:[0,2,9,13,17,21,24,38,41,44,48,53,59,63,72,81],channel:[7,11,30,41,44,55],channel_:56,channels_:41,channeltyp:56,channl:[7,11],char_bas:15,charact:[13,15],characterist:[9,36,82],check:[1,2,6],check_eq:21,check_l:21,check_sparse_distribution_batch:[80,81],check_sparse_distribution_in_pserv:[80,81],check_sparse_distribution_ratio:[80,81],check_sparse_distribution_unbalance_degre:[80,81],checkdatahead:34,checkfg:[30,32],checkfg_:41,checkfggrad:[30,32],checkfggrad_:41,checkgrad:81,checkgrad_ep:81,checkig:[30,32],checkig_:41,checkiggrad:[30,32],checkiggrad_:41,checkindex:44,checkindic:44,checkog:[30,32],checkog_:41,checkoggrad:[30,32],checkoggrad_:41,checkown:62,checkowner_:62,checksampl:34,checkstores:44,checksubset:50,child:[41,53,56],children:14,chmod:1,choic:14,choos:[15,81],chosen:[14,19,83],chpasswd:3,chri:41,christoph:41,chung:41,chunk:[36,69],chunk_schem:69,chunktyp:[36,69],cifar:9,cifar_vgg_model:9,circular:41,circularconv:44,circularconvderiv:44,class1:17,class2:17,class_dim:17,classfic:[11,17],classfiic:9,classic:72,classifc:17,classifi:[9,11,13,17,69],classification_cost:[9,13],classification_error_evalu:[13,17,19,36],classification_error_printer_evalu:36,classification_threshold:69,classificationerror:44,classificationerrormulti:44,classnam:24,classregistrar:[34,36,41],clean:[15,86],clear:[34,41,44,48,60,62],clearcount:50,cleargradi:50,clearindic:44,clearrow:44,clearunusedseg:58,clearupd:50,clearvalueupd:50,cleric:14,click:[2,36],client:[2,36,42],clientid:53,clients_:53,clip:[13,44,67,76,81],clone:1,close:[56,84],cluster_train:7,cmake:[1,21],cmakelist:21,cmd:[3,58],cmrprojectionnormlay:40,cnn:[11,13],cnt:50,code:[0,1],codelength:44,codelength_:41,coeff:72,coeff_:41,coeffici:[41,72],col:[24,28,32,34,44],col_siz:28,colidx:36,colin:44,collect:[14,38,41],collectbia:[21,44],colleg:14,colmax:44,colmerg:44,colnum:44,color:[9,11],cols_:44,cols_d:28,colscal:44,column:[8,19,21,25,28,30,32,36,41,44,53,58,69,72],column_sum_evalu:36,columnoffset:44,columoffset:44,colunm:19,com:[1,2,3,5,11,46,74],combin:[15,17,24,34,38,72,74,75],combinedevalu:36,come:17,comedi:14,comma:[8,81],command:[1,2,7,8,9,11],commandlin:17,commenc:13,comment:[13,17,33,41,60,74],commnun:7,common:[0,21,33,36,56,80],common_util:[7,15],commonli:[0,82],commun:[7,21,38,53,56,58,59],compact:25,compani:17,compar:[13,21,44,58],compat:[11,84],compet:17,compil:[1,2,3,21,24,58],complet:[13,21,25,41,62,72,74,86],complex:[0,13,41,83,84],complianc:[15,86],complic:72,compon:21,composecallback:48,comput:[0,3,13,15,17,21,25,32,36,38,41,44,53,56,58,59,72,74,82],computation:0,computethread:38,computethread_:38,computint:38,comupt:41,concat2:41,concat:[19,40],concat_lay:0,concaten:[41,74],concatenatelay:40,concatenatelayer2:40,concept:[0,64,84],concurr:[58,59],concurrentremoteparameterupdat:58,condit:[0,7,15,19,38,41,53,58,64,72,86],condition_vari:64,conf:[7,8,11,19,24,41,72,86],confer:41,confid:17,config:[3,7,8,9,11,13],config_:[34,36,38,41,50,58,59,81],config_arg:[11,13,17,80,81,82],config_gener:[7,15],config_lay:21,config_pars:[21,86],configmap_:58,configpath:24,conflict:[2,48],confpadding_:41,confpaddingy_:41,confptr:24,confront:19,congest:81,connect:[7,9,11,13,15,17,21,38,40],connectionist:[17,72],connnect:58,connor:17,consecut:[25,41],consequ:[72,74],consid:[9,41,44,76,82],consider:[74,84],consist:[9,11,13,19,28,34,41,44,50],constant:[21,58,64],constantslots_:34,constexpr:[46,58],construct:[0,15,28,41,44,59,62,63,64,84,86],construct_featur:15,constructor:[21,38,41,44,46,53,56,59],consum:17,contain:[0,3,7,9,11,13,14,17,19,24,30,34,36,38,41,44,53,56,59,68,69,72,75,78,84],contemporan:17,content:[17,58],context:[0,8,13,15,17,19,30,41,56,58,72,74],context_attr:74,context_len:[13,15,72,74],context_length:41,context_lenth:41,context_proj_layer_nam:74,context_proj_param_attr:74,context_project:[15,41],context_start:[13,41,72,74],contextlay:41,contextlength:[30,44],contextproject:40,contextprojectionbackward:44,contextprojectionbackwarddata:44,contextprojectionbackwardweight:44,contextprojectionforward:44,contextstart:[30,44],contibut:2,contigu:44,continu:[5,62,64,81],contrast:[19,41],control:[19,38,50,53,56,58,59,67,81],controllerthread_:59,controlparam_:50,controlupd:50,conv:[25,40],conv_act:74,conv_batchnorm_drop_r:74,conv_bias_attr:74,conv_filter_s:74,conv_layer_attr:74,conv_num_filt:74,conv_op:72,conv_oper:41,conv_pad:74,conv_param_attr:74,conv_shif_lay:72,conv_shift:72,conv_shift_lay:41,conv_strid:74,conv_with_batchnorm:74,convbaselay:40,convbwddataalgo:25,convbwdfilteralgo:25,convdesc_:41,conveni:44,converg:[7,17,58],convers:44,convert:[0,8,9,11,13,15,34,41,44,84,86],convex:[41,72],convex_comb:72,convex_comb_lay:41,convexcombinationlay:40,convexpand:44,convfwdalgo:25,convluat:41,convolv:41,convoper:40,convshiftlay:40,convshrink:44,cool:84,cooper:[53,58,59],coopi:38,coordit:41,coorditer:40,copi:[1,15,24,25,28,30,38,41,44,50,53,58,59,64,86],copyandclear:44,copybackseq:41,copybyrowindex:44,copyfrom:44,copyfromcsc:44,copyfromcsr:44,copyfromnumpyarrai:24,copyfromnumpymat:24,copyfromseq:41,copygradtobufferthread:38,copyid:41,copyidandsequenceinfo:41,copyidx:41,copyinarg:38,copyinfo:41,copyoutputgrad:38,copyoutputtootherdevic:41,copyparametersfromdevic:59,copyparameterstodevic:59,copyright:[15,86],copyrow:44,copys:50,copyscattedid:38,copyschedule_:41,copysingleparafromdevic:59,copysingleparatodevic:59,copythread_:38,copyto:44,copytocpu:44,copytogpu:44,copytonumpyarrai:24,copytonumpymat:24,core:[19,53,56,58,67,81,84],corpora:19,correct:[21,36,38,41,44,69,72],correctli:[21,36,69],correl:[9,17],correspond:[0,1,9,14,17,19,21,28,38,41,50,84,86],cos_scal:41,cos_sim:[15,41],cosin:[15,41,44,72],cossim:44,cossimderiv:44,cossimlay:40,cossimvecmatlay:40,cost:[13,15,17,19,33,38,40],cost_:58,costlay:40,costum:38,could:[7,13,15,24,34,38,50,58,59,69,72],count:[8,15,17,19,32,38,48,50,58,64,81,82,84],counter:50,countincr:50,covari:41,cpickl:[11,15],cpp:[13,15,19,21,50],cpu:[1,3,5,11,17,19,24,30,38,41,44,46,50,53,56,58,59,72,81],cpu_memori:44,cpualloc:46,cpuallocator_:46,cpubatch_:34,cpudensematrix:24,cpueos_:38,cpuexpandstartspos_:41,cpugpuvector:44,cpugpuvectorptr:44,cpugpuvectort:44,cpuid_:38,cpuids_:41,cpuinputstartpos_:41,cpuivector:44,cpuivectorptr:44,cpulocalmats_:44,cpumatrix:44,cpumatrixptr:[44,58],cpumemhandleptr:[44,46,53,58],cpumemoryhandl:46,cpuparameters_:59,cpuprob_:38,cpuseq2batchidx_:41,cpuseqendidxinbatch_:41,cpuseqidx_:41,cpusequencedim:[41,50],cpusparsematrix:44,cpusparsematrixptr:44,cpusrc:44,cpuvector:[44,58],cpuvectorptr:[44,58],cpuvectort:44,cpuvectort_:44,craftsman:14,crash:[7,44,81],crazi:7,create_bias_paramet:21,create_input_paramet:21,create_mode_norm:24,create_mode_test:24,createargu:24,createbycommandlin:24,createbyconfigprotostr:24,createbygradientmachinesharedptr:24,createbymodelconfig:24,createbypaddleargumentvector:24,createbypaddlematrixptr:24,createbypaddlevectorptr:24,createchannel:56,createcpudensefromnumpi:24,createcpuvectorfromnumpi:24,createdens:24,createfromconfigproto:86,createfrompaddlemodelptr:24,createfromprotostr:24,createfromrawptr:24,createfromsharedptr:24,createfromtrainerconfigfil:24,creategpudensefromnumpi:24,creategpuvectorfromnumi:24,creategpuvectorfromnumpi:24,createinframeinfo:38,creatematrix:[53,58],creatematrixrequest:58,creatememoryframeinfo:38,createmod:38,createparallelvector:44,createparameterconfigfromparameterptr:24,createparameterconfigfromparametersharedptr:24,createseqpo:38,createspars:24,createsparsematrix:44,createvector:[53,58],createvectorrequest:58,createzero:24,creation:53,crf:[41,72],crf_:41,crfdecodinglay:40,crflayer:40,crfs_:41,crime:14,critic:17,crop:11,crop_siz:11,cross:[13,28,30,41,44,72],crossmapnormalbwd:44,crossmapnormalfwd:44,crow:44,crow_:44,csc:[21,28],csc_col:28,csc_matrix:28,csc_row:28,csc_val:28,csc_valu:28,cslm:19,csr:[21,28,44],csr_col:28,csr_matrix:28,csr_row:28,csr_val:28,csr_valu:28,cstride:25,csv:14,ctc:72,ctcerrorevalu:36,ctclayer:40,ctcs_:41,ctor:[34,38,52],ctrl:[7,15],cub:9,cubla:25,cuda:[1,5,7],cuda_dir:[80,81],cuda_hom:1,cuda_so:3,cudahostalloc:46,cudnn:[1,5,11,25,41,72,81],cudnn_batch_norm:[41,72],cudnn_bn_min_epsilon:41,cudnn_conv:41,cudnn_dir:[80,81],cudnnbatchnormalizationforwardtrain:25,cudnnbatchnormlay:40,cudnnconvlay:40,cudnnpoollay:40,cuh:27,cumul:72,curand:25,curoffset_:58,curpo:41,curpos_:41,current:[0,1,2,5,7,13,17,19,21,24,25,38,41,44,48,50,53,56,58,59,62,63,65,72,74,76,81,84],current_word:0,currentavgcost:59,currentblockindex:56,currentblockindex_:56,currentcost:[13,15,17,19,69],currentev:[13,15,17,19,69],currentsequenceindex_:34,currenttim:[44,50],curv:[9,36],custom:[14,17,21,38,83],customstacktrac:42,cyclic:72,da_:41,daemon:58,dai:19,daili:17,dalla:84,dat:7,data_:[41,44],data_at_cpu:44,data_at_gpu:44,data_cl:68,data_config:86,data_dir:[8,9,17,19],data_initialz:13,data_lay:[0,9,13,15,41],data_provid:68,data_server_port:[80,81],data_sourc:68,dataargs_:38,dataargsframe_:38,dataargssize_:38,databas:17,databatch:34,datacol:30,dataconfig:[34,38],datadim:[41,72],dataformat:34,datahandl:44,datahead:34,dataid:50,dataim:30,datalay:[41,72],datalayerptr:38,datalayers_:38,datamems_:58,datamtx_:41,datamustincpu:38,datanormlay:40,datapool:34,dataprovid:[0,7,15],dataprovider_bow:13,dataprovider_emb:13,dataprovider_pow:13,dataproviderptr:38,dataproviderwraaperconvert:86,dataproviderwrapperconvert:86,datasampl:34,dataset:[8,9,11,12,13],datasize_:58,datatyp:53,dataupdatemod:53,db_:41,dbnparamdesc:25,dcmake_install_prefix:1,dcol:44,dcol_:44,deadlock:84,deal:[2,60],deamon:56,deb:[3,5],debian:3,debug:[44,72],decai:[9,44,76],decay_r:58,decayaddsquar:44,decayaddsquaremul:44,decayedadagradparameteroptim:48,decayl1:44,decayr:44,decayrate_:48,decid:38,decim:41,decimalreciprocal_:41,declar:[15,72,74],decod:[0,19,41,53,72,74],decoder_boot:0,decoder_group_nam:0,decoder_input:0,decoder_mem:0,decoder_prev:74,decoder_s:0,decoder_st:[0,74],decor:[21,84],decrement:64,decriptor:41,deduc:53,deep:[9,11,13,41],deeper:[11,53],deer:9,def:[0,11,13,15,21,24,72,84,86],defalut:[81,82],default_avg_width:44,default_devic:82,default_valu:82,defaultparamtyp:24,defferenct:84,defin:[0,7,8,9],define_py_data_sourc:[11,15,68],define_py_data_sources2:[9,11,13,15,84],defini:19,definit:[8,13,17,84],degradesequ:50,del:15,delai:[41,81],delar:13,delays_:41,delet:[3,34,41],delimit:[14,15,69],delta:[52,53,58,59],demo:[0,3,7,8,9,11],demograph:14,demonstr:[0,15],denom:44,denoms_:41,denot:[13,14,41,50,82],dens:[13,15,21,24,28,34,36,38,41,44,48,53,58,59,72,84],dense_vector:[15,84],densedata:34,densematrix:[28,44],denseslot:86,deploi:[3,7,82],deploy:7,dequ:34,dequeu:[62,63],dereferenc:64,deriv:20,desc:[25,41],descent:[41,72,76],describ:[13,21,64],descript:[0,9],descriptor:[25,41,56],design:[3,17,44,50,72,84],desir:8,dest:[24,28,44,53],dest_d:[25,28],dest_h:25,destcol:44,destin:[28,30],destroi:[25,41,53,64],destruct:[28,34,62,64],destructor:[21,28,46],detach:[56,62],detail:[0,2,7,8],detect:[36,69],determin:[21,72],dev:[1,3,9,19],devel:3,develop:[1,2,19],deverlop:81,deviat:67,devic:[3,19,25,38,41,44,46,59,81],deviceid:[38,41,46,50,82],deviceid_:[38,41,46,50],devid:[38,81],devlin:41,dez:17,diagnos:7,diagram:11,dict:[13,15,17,19,24,68,84],dict_dim:17,dict_fil:[0,13,69,72],dict_nam:68,dict_test:68,dict_train:68,dictionai:13,dictionari:0,dictsiz:19,diff:[28,44],differ:[0,1,3,7,8,9,11,13,17,19,21,36,38,41,44,50,53,58,68,69,72,74,81],digit:84,dii:38,dim1:[24,41],dim2:[24,41],dim:[8,11,13,17,21,24,28,30,34,41,44],dimens:[8,13,15,17,21,24,25,28,30,41,53,58,65,72,78,82],dimension:[0,13,21,41,84],dimenst:8,dimes:72,dimk:[25,28],dimm:[25,28],dimn:[25,28],dims_:41,dimsv_:41,din:15,dir:[7,11,13,15,17,19,38],direct:[11,25,30,41,59,72,74],directions_:41,directli:[7,17,24,25,58,59,74,83,84,86],directori:[1,3,7,9,11,13,15,17,19,53,81,83],diretcoti:11,dirnam:[52,53,59],disabl:[53,58,64,72],disable_copi:34,disable_copy_and_assign:24,discard:[58,81],discount:72,dispatch:[7,38,50,81],dispatchbydeviceid:38,disput:19,distanc:[36,69],distibut:8,distinguish:[7,19],distribut:[3,7,13,15,25,41,44,53,58,72],distribute_test:[80,81],distributedli:21,distributeev:36,dived:41,divid:[9,19,30,62,76,80],divrowvector:44,divscalar:44,divup:[32,53],diy_beam_search_prob_so:[80,81],dllinitmap:38,doc:[1,3,7,72,74],docallback:38,dockerfil:3,doctor:14,document:[1,2,9,13,15,17],documentari:[14,84],doe:[0,13,15,21,41,59,74,84,86],doesn:[19,67,72],dog:[9,11],doinit:50,don:[17,62,74],done:[0,17,38,41,62,72,74],donot:48,dooper:[53,58],dooperationrequest:58,dot:[11,19,41,53,81],dot_period:[9,15,17,19,81,82],dotdiv:44,dotmul:44,dotmul_oper:41,dotmul_project:41,dotmuloper:40,dotmulproject:40,dotmulsquar:44,dotmulsquaresum:44,dotmulsum:44,dotprod:44,dotproduct:44,dotravers:48,dotsquaremul:44,dotsquaresquar:44,dotsquaresum:44,dotsum:41,dotsum_:41,doubl:[1,25,32,36,41,58,81,84],doublebuff:34,doublebuffer_:34,down:[13,38],downclip:44,download:1,download_cifar:9,downsampl:9,doxygen:[1,2],dpkg:5,drama:14,draw:11,driver:3,drop:[34,38,50],drop_rat:67,dropcallback:38,droponenod:38,dropout:[13,21,41,67,72],dropout_lay:72,dropout_r:74,dropoutmask_:41,drow:44,drow_:44,dserver:[53,81],dso:25,dso_handl:25,dst:[25,28,30,41,44,53],dstid:38,dtor:34,dtype:[11,24,58,86],dubai:19,due:[14,15],dummydataprovid:34,dummyevalu:36,dummyoptim:48,dump:60,duplic:[14,58],dure:[13,15,19,21,25,34,38,41,44,72,80,81,83,84],durn:84,dw_:41,dwith:1,dwith_doc:1,dwith_gpu:1,dwith_swig_pi:1,dynamic_cast:[21,44],each:[0,7,8,9,11,13,14,15,17,19,21,24,25,28,30,32,34,36,38,41,44,48,50,53,56,58,60,62,65,69,72,78,81,82,83,84,86],each_feature_vector:65,each_meta:15,each_pixel_str:84,each_sequ:72,each_time_step_output:65,each_timestep:72,each_word:84,eachstepcallback:38,earli:38,eas:11,easi:[7,13,21,41],easier:[3,21],echo:[3,15,17],edit:[36,69],edu:9,educ:14,effect:[81,84],effici:[0,21,38,58,83,84],efg:72,either:[13,15,86],elaps:25,elec:13,electron:13,elem_dim:72,element:[13,17,19,25,28,34,41,44,52,58,63,69,72,74,84,86],elementcnt_:44,elif:15,elment:44,els:[11,13,15,21,28,44,48,58,60,62,64],emb:13,embed:0,embedding_lay:[0,13,15],embedding_nam:0,embedding_s:0,empir:[41,72],emplace_back:21,emploi:[0,14],empti:[34,38,41,60,62,63,72],empty_:62,emul:19,enabl:[3,7,13,25,50,58,67,81],enable_grad_shar:[80,81],enable_if:46,enable_parallel_vector:81,enableinttyp:50,enablesharedtyp:50,enablesparseparamet:50,enabletyp:50,enc_proj:[0,74],enc_seq:74,enc_vec:0,encapsul:59,encod:[0,19,53,74],encoded_proj:[0,74],encoded_sequ:[0,74],encoded_vector:0,encoder_last:72,encoder_proj:0,encoder_s:0,end:[0,8,17,19,25,34,36,38,41,48,53,58,59,69,72,74,81],end_:[41,44],end_id:24,end_of_job:63,endcol:44,enddim:50,endpad_:41,endrow:44,engin:[3,14,41],english:[19,84],enough:59,enqueu:[62,63],ensenti:72,ensur:[21,28,41,48,53,84],enter:[7,14],entir:[17,72,74],entiti:58,entri:[14,21],entropi:[13,28,41,44,72],enumer:[13,42,65,72],environ:[1,3,5,7,9,15,38,80,81],eos_id:[0,41,72],eosframelin:38,eosframeline_:38,eosidchecklay:40,epoch:[14,64],epsilon:[25,41,76],epsilon_:48,equal:[41,56,58,60,72,74,76],equat:20,equenc:30,equip:[0,3],erin:41,err:25,error:[5,7,9,11,13,14,15,17,19,21,24,25,28,33,36,38,41,44,67,69,72,76,81],error_clipping_threshold:67,errorr:69,especi:[58,74,84],essenti:[1,19],estim:41,estimatedmean:25,estimatedvar:25,etc:[3,17,19,24,32,44,50,52,58,76,80,82],eth0:7,ethernet:3,ethternet:7,euclidean:41,eval1:36,eval:[13,15,17,19,36,38,69],eval_bleu:19,evalimp:36,evalu:[7,13],evaluate_pass:17,evaluator_:38,evaluator_bas:69,evaluatorconfig:36,evalut:19,evalutaion:18,even:[17,38,58,62,81],event:25,everi:[0,13,17,19,21,34,41,50,72,74,81,83,84],everyth:2,exactli:[36,41,44,69,72,74,84],exampl:[0,1],exceed:72,except:[8,15,38,64,82,84,86],exchang:[53,59],excl:41,exconv:41,excut:62,exec:[44,50,62,81],execfunc:[44,50,58],exechelp:62,execplusown:62,exectur:62,execut:[14,17,21,25,41,53,58,62,64],exhibit:[56,58,59],exist:[14,17,21,41,44,50,53,59,81],exit:[38,62,81],exp:44,expand:[17,19,21,30,38,41,44,72],expand_a:72,expand_lay:41,expand_level:72,expandcolmat:44,expandconvlay:40,expandedin_:41,expandedingrad_:41,expandedout_:41,expandedoutgrad_:41,expandfwdonc:41,expandinput_:41,expandlay:40,expandlevel:[41,72],expandmat:41,expandonefram:41,expandstartspos_:41,expderiv:44,expect:[17,38,59],expectedpasscount:59,expectedpasscount_:59,expectedppasscount:59,expens:19,experi:[4,82],explain:[17,69,84],explan:[13,19],explicit:21,explicitli:38,exploit:9,explor:72,expon:44,exponenti:33,expos:3,express:[15,17,44,86],expw_:41,expx_:41,extend:[15,58],extens:[14,15,19,76],extern:[38,81],extra:[30,41,58],extraattr:[67,82],extract_fea_c:11,extract_fea_pi:11,extract_para:8,extralayerattribut:[67,72,74],extralayeroutput:74,extrem:72,extremli:83,f_rdma:56,f_tcp:56,fabric:7,facebook:46,facilit:58,facotr:72,fact:[11,59],factor:[25,28,67,72,76],factori:34,fail:[9,38,64,81,82],fals:[0,8,13,15,17,19,21,24,25,30,36,38,41,44,48,50,52,53,59,64,67,68,69,72,74,76,81,82,84,86],false_typ:44,famili:19,familiar:84,fanscin:84,fantasi:14,fantast:13,farmer:14,fascinatingli:83,fast:[2,41],faster:[0,17,38,41,72,74],fastest:41,fat:56,fatal:[34,44],fatur:41,favorit:2,fc1:[21,82],fc2:82,fc3:82,fc4:82,fc_act:74,fc_attr:74,fc_bias_attr:74,fc_layer:[13,15,21,41],fc_layer_nam:74,fc_param_attr:74,fclayer:21,fea:11,fea_output:11,feaimgheight:44,feaimgwidth:44,feat:17,featur:9,feature_map:[15,25],featuremapexpandlay:40,feed:[17,74],feedforward:9,femal:14,fernan:17,festiv:84,fetch:[0,2,21,50,53,59],few:[41,84],fewer:72,fg0:72,fieldscann:34,figur:[0,8,9,11,17,19,21],file1:19,file2:19,file:[0,1,2,7,8,9,11],file_list:84,file_nam:[11,13],filelist:34,filelist_:34,filenam:[15,34,50,84],filer:72,fill:[13,24,34,38,41,44,72],fillrowindic:44,fillselectivedata:41,fillsequ:44,fillslot:34,film:14,filter:[11,25,30,41,44,72],filter_data:25,filter_grad_data:25,filter_s:[72,74],filter_size_i:72,filterchannels_:41,filterdesc_:41,filterpixels_:41,filters:41,filtersize_:41,filtersizex_:41,filtersizey_:41,finali:7,find:[1,9,17,19,28,41,58,60,72,76],findlastset:46,fine:[15,58,67],fini:[44,50],finish:[7,9,24,25,34,36,38,41,44,48,53,58,59,62,84],finishbatch:[24,48,52,58,59],finishcatchupwith:48,finishcv_:62,finishfil:34,finishpass:[24,48,52,59],finishprepar:34,finishthread:53,finishtrain:24,finishtrainonebatch:24,finishtrainpass:24,finit:21,first:[0,1,2,3,5,8,9,11,13,15,17,19,21,25,28,34,36,38,41,44,53,59,62,69,72,81,82,84],first_seq:0,firstli:[7,53],firstorderoptim:48,firsttest_:41,firsttime_:48,fit:[38,83],five:13,fix:[11,19,67,84],flag:[1,9,25,41,44,81],flags_port:53,flags_ports_num:53,flags_use_gpu:[34,38,41,50],flexibl:[0,72,83],flight:19,float32:[11,24,86],float_valu:44,floatarrai:24,fluctuat:58,flush:58,fly:13,focu:[58,84],folder:[9,17,19],folli:46,follow:[0,1,2,3,4,7,8,9,11,13,14,15,17,19,21,36,41,48,58,69,72,74,76,82,83,84],foramt:44,forc:62,forcestop:62,forcibl:62,foreachlay:38,forecast:17,forget:[17,41,76],forgetcheck:41,forgetg:41,forgetgate_:41,form:[38,41,44,50,59,72,74,76,83,84],formal:38,format:[8,9],format_:[44,50],former:19,formul:41,formula:[25,41,72],formular:72,forward:[0,2,17,21,24,25,28,30,33,38,41,44,48,59,60,65,74,82],forwardactiv:[21,41],forwardbackward:[24,38,53,58,59],forwardbackwordtime_:53,forwardbatch:41,forwardbia:41,forwardcost:41,forwarddropout:41,forwardgate2outputsequ:41,forwardimp:[38,41],forwardimpin:41,forwardonebatch:24,forwardoneinput:41,forwardonesequ:41,forwardseqparallel:41,forwardsequ:41,forwardtest:86,forwardvars_:41,found:[0,1,9,13,17,24,58,72,84],four:[8,11,13,15,17,41,84],fourth:38,foward:[33,41],fpga:32,frame:[30,32,36,38,69],framecnt:30,frameforgetgate_:41,framegate_:41,frameheight:[41,50],frameinputgate_:41,frameinputnode_:41,frameoutput_:41,frameoutputgate_:41,framepreoutput_:41,frames:[30,41],frames_:38,framestate_:41,framewidth:[41,50],framework:[11,13,17,21,38,56,58],free:[19,25,28,41,44,46],freeall:46,french:19,frequenc:[8,13,17],frequent:[7,19,53],friend:[24,36,38,44,56],frog:9,from:[0,1,2,3],from_timestep:72,fromfil:11,fucntion:62,fuction:62,full:[0,21,41,50,53,56,59,63,72],full_matrix_project:[0,41],fulli:[2,9,11,13,15,17,21,40],fullmatrixproject:40,fullsiz:[52,59],fully_matrix_project:74,fullyconnect:8,fullyconnectedlay:[21,40],func:[44,50,58,62],funcnam:[53,58],further:[3,53,58,72],fusion:15,futru:41,futur:[53,62],fwdalgo_:41,fwdlimitbyt:25,fwdlimitbytes_:41,gain:[58,72],gamma:[11,25,41],gamma_:48,gap:81,gate_:41,gate_act:[72,74],gate_recurr:72,gatedrecurrentlay:40,gategrad:[30,32],gatevalu:[30,32],gateweight:32,gateweight_:41,gateweightgrad:32,gather:[15,21,38,41,72],gatheragentlay:40,gauss:67,gaussian:44,gemm:44,gen1:41,gen:[19,41,72],gen_conf:19,gen_data:19,gen_input:0,gen_result:19,gen_trans_fil:0,gender:[14,15],generated_sequ:72,generatedinput:[0,72],generatesequ:[24,38],generator_:38,generatorconfig:38,genert:84,genr:[14,15],gentoo:3,gereat:72,get:[0,1,3,5,7,9,11,13,15,17,21,24,25,28,34,38,41,44,48,50,53,56,58,59,62,64,72,74,84],get_best_pass:17,get_config_arg:[13,15,17,82],get_data:13,get_imdb:17,get_input_lay:21,get_model:11,get_output_layer_attr:74,getabsmax:44,getabssum:44,getaccum:50,getallcount:50,getalldata:53,getallocs:46,getallthread:38,getavgcost:59,getbackwardcallback:38,getbatchs:[21,24,50],getbatchvalu:41,getbiasparamet:41,getblockid:58,getblocklength:56,getblocknum:41,getblockoffset:58,getbuf:[24,46,50],getchannel:56,getcol:44,getcolnum:44,getcolstartidx:44,getcolumn:44,getconfig:[24,34,41,50],getcpualloc:46,getcpustartposit:50,getcublasdsohandl:25,getcudartdsohandl:25,getcudnndsohandl:25,getcuranddsohandl:25,getcurrentavgcost:59,getcurrenttimestick:32,getdata:[24,44],getdeviceid:[38,41,50],getel:44,getelementcnt:44,getfinalpath:38,getformat:[44,50],getframeheight:50,getframewidth:50,getgpualloc:46,getgradbuf:38,getgradientmachin:38,getgradientseg:50,getheight:[24,44],getid:[24,44,50],getinarg:38,getindexdicthandl:44,getinfolay:41,getinitmean:50,getinitstandarddevi:50,getinput:[21,41],getinputgrad:[21,41],getinputlabel:41,getinputvalu:[21,41],getintbuf:50,getinternalargumentsptr:24,getitem:24,getjob:63,getlabellay:41,getlay:38,getlayeroutput:[24,38],getlearningr:48,getlearnr:50,getlocalindic:44,getlocalparamet:50,getlocalrow:44,getmajorpartn:50,getmat:50,getmax:44,getmemoryhandl:44,getmin:44,getminorpartn:50,getmodelconfig:24,getmutabledata:44,getmutablevector:44,getnam:[24,33,41,46,50,60],getnetworkoutput:24,getnextbatch:34,getnextbatchfrombuff:34,getnextbatchintern:34,getnextblocklength:56,getnextpo:41,getnonstaticparamet:38,getnumbatch:41,getnumblock:56,getnumdevic:38,getnumlogicaldevic:38,getnumprocess:59,getnumsequ:50,getnumsubsequ:50,getnumthread:[38,62],getoptimizationconfig:24,getoutarg:38,getoutput:41,getoutputgrad:[21,41],getoutputlabel:41,getoutputlay:40,getoutputs:41,getoutputvalu:[21,41],getpaddlemodelconfig:24,getparamet:[24,38,41,53,58],getparameterconfig:58,getparameterid:38,getparametermap:38,getparameterptr:[21,50],getparameters:24,getparameterspars:[53,58],getparametersremot:[52,59],getparametertyp:[24,48,52],getpasstyp:38,getpeernam:56,getpoint:44,getprefetchmatrix:50,getprepo:41,getprev:41,getprevbatchoutput:41,getprevbatchst:41,getpserverparametergradi:53,getpserverparametervalu:53,getrawptr:24,getrow:44,getrowbuf:44,getrowcol:44,getrownum:44,getrowstartidx:44,getrowvalu:44,getscor:24,getscorelay:41,getseg:36,getsent:24,getseqlengthandstart:50,getseqoutputfrombatch:41,getsequ:24,getsharedcount:50,getsharedptr:24,getsiz:[21,24,34,41,44,46,50],getslaveparamet:38,getslotid:24,getslotin:24,getslotnum:24,getslotsequencedim:24,getslotsequencestartposit:24,getslotsubsequencestartposit:24,getslotvalu:24,getsparseformat:24,getsparserowcol:24,getsparserowcolsv:24,getsparsevaluetyp:24,getstat:[38,41,59],getstatu:58,getstatusrequest:58,getstrid:44,getsum:44,getsync:44,gettasktyp:38,getthread:38,gettlstempbuf:50,gettmpsparsematrix:44,gettotallength:56,gettransdtyp:53,gettranspos:[21,44],gettyp:41,getvalu:44,getvaluebuf:38,getvaluetyp:44,getvector:44,getw:[21,50],getweight:[21,41],getwgrad:[21,50],getwidth:[24,44],gflag:1,gist:74,git:[1,2],github:1,give:[13,21,38,41,50,84],given:[13,17,19,21,38,41,44,53,81,84],global:[15,17,41,44,48,53,58,62,67,76,81],global_learning_r:67,globalindic:44,globalindices_:44,globe:84,glog:1,goe:[72,74],good:[17,19,72],googl:58,govern:[15,86],gpu:[1,3,5,9,11,15,17,19,24],gpu_id:[81,82],gpualloc:46,gpuallocator_:46,gpubatch_:34,gpugpu_id:80,gpuivector:44,gpuivectorptr:44,gpumatrix:44,gpumatrixptr:44,gpumemhandleptr:[44,46],gpumemoryhandl:46,gpusparsematrix:[28,44],gpusparsematrixptr:44,gpuvector:44,gpuvectorptr:44,gpuvectort:44,gpuvectort_:44,gpuworkspac:25,grab:17,grad:[14,25,28,30,33,38,41,44,50,81],grad_d:28,grad_o:28,grad_share_block_num:[80,81],grad_w:28,gradbuff:38,gradbufqueue_:38,gradbufs_:38,gradcollectthread:38,gradcollectthread_:38,gradcount:50,graddata:41,gradient:[13,17,20],gradient_clipping_threshold:[13,17,76],gradient_printer_evalu:36,gradientaccum_:50,gradientmachin:[15,19,24],gradientmachine_:38,gradientmachinepriv:24,gradientmatchinecreatemod:24,gradientreadybarrier_:58,gradietn:38,gradmat:50,gradqueue_:38,gradreadycond:50,gradsegments_:50,gradsem_:50,gradstream_:38,gradterms_:41,grai:9,graident:38,grain:58,gram:[8,17],graph:[8,72],graphviz:11,grave:17,grayscal:84,greater:[62,64,72],greaterpath:38,grep:17,groudtruth:0,ground:[13,19,69,72],group:[17,40],group_id:15,grouplen:14,groups_:41,grow:63,gru:[0,13,19,29],gru_bias_attr:74,gru_decod:0,gru_decoder_with_attent:0,gru_encoder_decod:[8,19],gru_layer_attr:74,gru_siz:13,gru_step:0,gru_step_lay:[0,41],grucomput:40,grumemori:[0,41],grusteplay:40,gserver:21,gtest:1,guarante:21,guess:17,guid:[0,2],gur_group:74,hack:[4,7],half:41,hand:[14,15,17],handl:[7,15,17,38,41,45],handler:25,handlerequest:[56,58],handlerequestbegin_:58,handwrit:[17,84],happen:38,har:[7,56,58],hard:[13,24,44],hardward:58,hash:[44,48],hasnonstaticcpuparamt:38,hasnonstaticcpuparamters_:38,hasstaticparamet:38,hassubseq:[38,50],hastyp:50,have:[0,1,2,3,5,7,8,9,13,14,15,17,19,21,34,36,38,41,44,50,53,60,62,63,69,72,81,82,83,84],hdf:83,head:[2,8,17],header:[8,11,15,21,34,50,56],header_:34,headerptr:34,headerptr_:34,health:14,heavi:7,height:[9,21,24,25,28,30,38,41,44,50,72],height_:[34,44],height_pad:25,heightpad:41,heightstor:44,help:[4,7,58,59,84,86],helper:[21,38,58,68,72,74,85],helpful:50,here:[0,1,7,8,9,11,12,13,14,15,17,19,25,34,41,48,50,58,67,68,72,74,80,82,84,86],heterogen:32,heurist:[19,72,81],hidden:[0,13,15,17,41,58,59,72,74],hidden_s:[15,74],hide:59,hierarch:[0,41,72],hierarchicalsigmoidlay:40,high:[3,21,32,44,58,67],higher:83,highest:[19,46],highli:[0,15,17,82,83,84],himself:53,hint:86,histogram:44,histogram_expon:44,histogramtyp:44,histor:17,histori:38,histroi:58,hl_activation_end:32,hl_activation_funct:30,hl_activation_linear:32,hl_activation_mode_t:[30,32,41],hl_activation_relu:32,hl_activation_sigmoid:32,hl_activation_tanh:32,hl_aggreg:27,hl_avgpool_backward:30,hl_avgpool_forward:30,hl_base:31,hl_batch_norm_backward:25,hl_batch_norm_forward_infer:25,hl_batch_norm_forward_train:25,hl_batch_transpos:27,hl_cmrnorm_backward:30,hl_cmrnorm_forward:30,hl_construct_sparse_matrix:28,hl_context_projection_backward_data:30,hl_context_projection_backward_weight:30,hl_context_projection_forward:30,hl_conv_workspac:25,hl_convolution_backward_bia:25,hl_convolution_backward_data:25,hl_convolution_backward_filt:25,hl_convolution_descriptor:[25,41],hl_convolution_forward:25,hl_convolution_forward_add_bia:25,hl_cossim:28,hl_cossim_deriv:28,hl_create_convolution_descriptor:25,hl_create_ev:25,hl_create_filter_descriptor:25,hl_create_pooling_descriptor:25,hl_create_tensor_descriptor:25,hl_cuda_event_queri:25,hl_destroy_convolution_descriptor:25,hl_destroy_ev:25,hl_destroy_filter_descriptor:25,hl_destroy_pooling_descriptor:25,hl_destroy_tensor_descriptor:25,hl_destruct_sparse_matrix:28,hl_device_can_access_p:25,hl_device_enable_peer_access:25,hl_device_synchron:25,hl_event_elapsed_tim:25,hl_event_synchron:25,hl_event_t:25,hl_expand_feature2col:30,hl_filter_descriptor:[25,41],hl_fini:25,hl_float_max:32,hl_float_min:32,hl_float_valu:[28,32],hl_free_mem_devic:25,hl_free_mem_host:25,hl_free_sparse_matrix:28,hl_get_cuda_lib_vers:25,hl_get_cudnn_lib_vers:25,hl_get_devic:25,hl_get_device_count:25,hl_get_device_error_str:25,hl_get_device_last_error:25,hl_get_sync_flag:[21,25],hl_gru_grad:[32,41],hl_gru_valu:[32,41],hl_init:[25,32],hl_lstm_grad:[32,41],hl_lstm_parallel_backward_data:30,hl_lstm_parallel_backward_weight:30,hl_lstm_parallel_forward:30,hl_lstm_valu:[32,41],hl_malloc_devic:25,hl_malloc_host:25,hl_malloc_sparse_matrix:28,hl_matrix:27,hl_matrix_:32,hl_matrix_add:28,hl_matrix_add_to_row:28,hl_matrix_appli:27,hl_matrix_bas:27,hl_matrix_classification_error:28,hl_matrix_column_max:28,hl_matrix_column_min:28,hl_matrix_column_sum:28,hl_matrix_cross_entropi:28,hl_matrix_cross_entropy_bp:28,hl_matrix_csc2dens:28,hl_matrix_csc_mul_dens:28,hl_matrix_csr2dens:28,hl_matrix_csr_add_bia:28,hl_matrix_csr_add_dens:28,hl_matrix_csr_column_sum:28,hl_matrix_csr_mul_dens:28,hl_matrix_dense_mul_csc:28,hl_matrix_dense_mul_csr:28,hl_matrix_format_t:[28,32],hl_matrix_mul:25,hl_matrix_mul_vector:25,hl_matrix_op:27,hl_matrix_row_max:28,hl_matrix_row_min:28,hl_matrix_row_sum:28,hl_matrix_select_row:28,hl_matrix_softmax:28,hl_matrix_softmax_deriv:28,hl_matrix_top_k:28,hl_matrix_transpos:25,hl_matrix_typ:27,hl_matrix_value_t:[28,32],hl_matrix_zero_mem:28,hl_max_sequence_backward:30,hl_max_sequence_forward:30,hl_maxpool_backward:30,hl_maxpool_forward:30,hl_memcpi:25,hl_memcpy_async:25,hl_memcpy_csc_matrix:28,hl_memcpy_csr_matrix:28,hl_memcpy_device2devic:25,hl_memcpy_device2host:25,hl_memcpy_from_csc_matrix:28,hl_memcpy_from_csr_matrix:28,hl_memcpy_host2devic:25,hl_memcpy_sparse_matrix:28,hl_memset_devic:25,hl_no_valu:[28,32],hl_param_relu_backward_diff:28,hl_param_relu_backward_w:28,hl_param_relu_forward:28,hl_pooling_averag:25,hl_pooling_average_exclude_pad:25,hl_pooling_backward:25,hl_pooling_descriptor:[25,41],hl_pooling_end:25,hl_pooling_forward:25,hl_pooling_max:25,hl_pooling_mode_t:[25,41],hl_rand:25,hl_reset_convolution_descriptor:25,hl_sequence2batch_add:30,hl_sequence2batch_copi:30,hl_sequence_avg_forward:30,hl_sequence_softmax_forward:28,hl_set_devic:25,hl_set_device_flags_block:25,hl_set_sync_flag:25,hl_shrink_col2featur:30,hl_softmax_backward:25,hl_softmax_forward:25,hl_spars:27,hl_sparse_csc:[28,32],hl_sparse_csr:[28,32],hl_sparse_end:32,hl_sparse_matirx_column_sum:28,hl_sparse_matrix_:[28,32],hl_sparse_matrix_add_bia:28,hl_sparse_matrix_add_dens:28,hl_sparse_matrix_column_sum:28,hl_sparse_matrix_get_col:28,hl_sparse_matrix_get_row:28,hl_sparse_matrix_get_valu:28,hl_sparse_matrix_mul:28,hl_sparse_matrix_s_ptr:44,hl_sparse_matrix_top_k:28,hl_specify_devices_start:25,hl_srand:25,hl_sse_matrix_kernel:27,hl_start:25,hl_stream_record_ev:25,hl_stream_synchron:25,hl_stream_t:[25,28,32,38,44,50],hl_stream_wait_ev:25,hl_table_appli:27,hl_tensor_descriptor:[25,41],hl_tensor_reshap:25,hl_thread:31,hl_time:31,hl_top_k:27,hl_trans_op_t:[25,28,32],hl_value_end:32,hl_vector_abs_sum:28,hl_vector_select_from:28,hl_vector_sum:28,hold:[38,41,44,56],home:[1,7],homemak:14,hook:[15,48,50,52],horizont:[11,25,72],horror:14,hors:9,horst:17,host:[3,7,25,28,44,53,56,58,59],hostnam:7,hostsrc:44,hot:15,hour:19,hous:[8,84],how:[0,9,11,13,15,36,38,53,58,67,72,81,83],howev:[0,2,17,19,38,74,80,81,84],hppl:[25,30,31],hppl_active_funct:30,hppl_op_end:32,hppl_op_n:[25,32],hppl_op_t:[25,28,32],hppl_stream_1:32,hppl_stream_2:32,hppl_stream_3:32,hppl_stream_4:32,hppl_stream_:32,hppl_stream_default:[32,44,50],hppl_stream_end:32,hppl_thread_stream_1:32,hppl_thread_stream_2:32,hppl_thread_stream_3:32,hppl_thread_stream_4:32,hsigmod_lay:41,hstride:25,html:9,http:[1,2,3,5,9,11,15,19,46,48,74,86],huang:41,huber:[41,72],hubertwoclass:40,huge:[59,72],huina:17,human:19,hundr:53,hyper:[21,72],i0601:15,i0706:19,i0719:19,ib0:7,ibasematrix:44,icpugpuvector:44,icpugpuvectorptr:[38,41,44,50],icwsm:17,id_input:[0,72],idea:[41,53,72],ident:[14,65],identifi:[0,21,53],identity_project:41,identityactiv:33,identityoffsetproject:40,identityproject:40,idindex:[38,41],idindex_:41,ids_:41,idsarray_:44,idsiz:41,idsize_:41,idsvec_:41,idx:[21,24,41,44,52,58],idxbuf:24,ieee:17,ignor:[28,36,41,44,69,81,84],igradientmachinemod:38,iiddata:34,ijcnlp:17,illustr:[0,13,21,84],ilsvrc:11,image_classif:9,image_desc:25,image_list_provid:11,image_provid:9,image_s:11,imageh_:41,imagepixel:41,images:41,imagew_:41,imdb:[14,16],imdber:17,img:9,img_conv_lay:41,img_norm_typ:72,img_pool_lay:41,img_siz:9,imgpixels_:41,imgsiz:41,imgsize_:41,imgsizeh:44,imgsizeh_:41,imgsizew:44,imgsizew_:41,imgsizex_:41,imgsizey_:41,immedi:[34,44,53,58],implement:[0,1,7,11,13,20],implemnt:41,implent:41,impli:[15,86],implicitli:34,implict:50,implictli:[58,59],impliment:59,importerror:15,improv:[17,19],in1:41,in2:41,in_:41,inarg:[24,38,86],inargs_:38,inargscopied_:38,inc:[15,86],incb:25,incc:25,includ:[0,1,3,5,8,13,14,19,21,24,30,32,33,34,38,41,44,46,48,50,52,53,56,72,74,81,83,84],inconsist:14,incopyidx:41,incorrect:[41,72],increas:[19,21,25,81],increment:[64,81],incshar:50,incupd:[21,50],incupdatecount:38,indac:38,inde:58,indeedli:59,indeped:41,independ:[13,34,41,72],index:[7,15,24,28,30,34,36,38,41,44,46,52,58,69,72,84],indexdata:34,indexdict:44,indexdicthandl:44,indexdicthandle_:[38,44],indexdictptr:44,indic:[7,28,34,36,38,41,44,53,56,58,69,72,84],indicationg:58,indiff:30,indirectli:58,individu:59,ineffici:[44,81],inf:38,infiniband:7,info:[2,7],info_:38,infom:2,infomatrion:81,inform:[3,14,15,17,19,21,41,50,69,81,83,86],inframelin:38,inframelines_:38,ingor:81,ingrad:25,ingrad_:41,ingraddesc:25,inherit:[24,34,36,38,41,44,46,48,50,52,53,56,58,59,62,64,65],inidx:41,init:[13,15,21,24,25,36,38,41,44,48,50,52,53,58,59,67,82],init_hook:[13,15,83],init_hook_wrapp:68,init_model_path:[8,13,80,81,82],initasyncload:34,initblock:44,inithook:50,initi:[0,8,13,21,24,25,38,41,50,58,59,62,64,67,72,81],initial_max:67,initial_mean:[50,67,72],initial_min:67,initial_std:[50,67,72],initial_strategi:50,initil:41,initneedflag:41,initpaddl:[24,86],initshar:44,initsubnetwork:41,initthread:53,initthreadnum:62,initvalu:64,inlay:38,inlcud:41,inlength:41,inner:[21,41],inner_param_attr:74,inpidx:41,inplac:24,inpserv:48,inpt:28,input1:[28,41,74],input1_height:28,input2:[28,41],input2_height:28,input:[0,8,9,11,13,15,17,19,21,24,25,28,30,32,33,34,36,38,41,44,50,53,58,59,62,65,69,72,74,78,79,82,84,86],input_:41,input_data:[21,25],input_data_grad:25,input_data_target:21,input_featur:65,input_feature_map:25,input_fg:41,input_fg_:41,input_hassub_sequence_data:21,input_id:[41,72],input_ig:41,input_ig_:41,input_imag:[9,25,74],input_image_grad:25,input_index:21,input_label:21,input_lay:21,input_og:41,input_og_:41,input_s_:41,input_sequence_data:21,input_sequence_label:21,input_sparse_float_value_data:21,input_sparse_non_value_data:21,input_t:21,input_typ:[13,15,83],inputargument_:41,inputbuff:58,inputcheck:41,inputdata:30,inputdef:21,inputdesc:25,inputdesc_:41,inputdim:30,inputg:41,inputgate_:41,inputgrad:[30,41,44],inputindex:41,inputiov:53,inputlay:41,inputlayers_:[21,41],inputmat:44,inputoffset_:41,inputstartpos_:41,inputvalu:41,ins_:41,insert:25,insid:[36,44,69,72],inspir:8,instanc:[0,21,30,34,41,44,76,81],instartidx:41,instatu:53,instead:[7,13,19,41,44,50,53,56,58,59],instruct:[5,13,38],int32:[24,81],int32_t:[36,38,48,50,53],int64_t:[24,32,34,38,44,48,50,52,53,56,58,59],intarrai:24,intbufs_:50,integ:[0,13,17,21,38,41,84],integer_valu:[13,84],integer_value_sequ:[0,13,84],inter:7,intercept:[41,72],interest:17,interfac:[7,9],interg:13,intermedi:[25,41],intern:[1,24,32,34,38,41,44,59,65,72,74],internet:17,interpol:[41,72],interpolation_lay:41,interpolationlay:40,interpret:[36,69,84],interv:[17,41,44],intervals_:41,intial:41,intput:28,introduc:[15,17,33,84],intstores:50,intv:41,intwithfloatarrai:24,inv:30,invari:9,invok:[15,24,34,38,41,50,52,53,59,72,84],invsqrt:44,iob:[36,69],iodesc_:41,ioe:[36,69],ioerror:24,ioff:41,iov:56,iovec:[53,56,58],iovlength:56,iparameterupdaterhook:[50,52],ipt:0,is_async:76,is_gener:[8,19],is_integr:46,is_kei:15,is_layer_typ:72,is_predict:[13,15,17],is_seq:[0,15],is_sequ:[15,38],is_stat:[41,67],is_test:[11,17],is_train:84,is_unsign:46,isaveragewindowtoolong:48,ischunkbegin:36,ischunkend:36,iscontigu:44,isdatamustincpu:38,isdrop:38,isequalto:44,isequenceresult:24,isfirstpass_:59,isfulls:50,isgpu:[24,38],isgpuvers:24,isgradclean:41,isgradshar:50,isgradsparseupd:50,isinst:[24,86],isinvalid_:41,isloc:38,isnonv:24,isnotreadi:25,ispad:[30,44],isparameterspars:48,isparametersparse_:48,ispass:38,ispassfinish:53,ispassgrad:38,ispassgrad_:38,ispointeralign:44,isregularizationbatch:48,isselectalgo_:41,isspars:[21,24,44,50],issparseremoteupd:50,issparseserv:53,issparseserver_:58,isstat:50,issu:11,istestmod:34,istlargerthanalign:58,istranspos:44,istream:[38,50],isupdat:50,isvalidmatrixhandl:58,isvalidvectorhandl:58,isvalueclean:41,isvalueshar:50,isvalueupd:50,item:60,iter:[9,17,38,64,72,74,76],itself:[59,74],ivec:50,ivector:[24,41,44],ivectorpriv:24,ivectorptr:[38,41,44,50],jacob:41,jie:17,job:6,job_dispatch_packag:7,job_mod:8,job_workspac:7,jobadding_:62,jobfinishbarrier_:62,jobfunc:[44,62],jobfunc_:62,jobqueu:53,jobs_:62,jobstartbarrier_:62,johan:17,john:41,join:62,joint:[8,19,41],jointli:[19,41,74],journal:17,jpeg:9,jpg:11,json:[7,15],jth:74,judg:19,junyoung:41,just:[2,3,7,9,11,15,17,24,38,41,44,53,59,64,65,69,72,74,82,84],kaimingh:11,kasasbeh:41,kaverag:[41,59],kaveragesquarerootn:41,kcossimscale_:41,kcustom:38,kdecimalsc:41,kei:[15,17,58,59,84],kelasticaverag:59,kept:41,kernel:[3,13,41,72],key1:81,key2:81,keyword:84,kformatvers:50,kill:6,kind:[13,15,41,50,53,83,84,86],kmaxnumaccumul:48,kminmax:41,kmissparameterfail:50,kmissparameterrand:50,kmissparameterzero:50,knonseq:41,knormal:38,know:[15,21,74,84],knowledg:17,known:[17,19],kretmsginvalidmatrixhandl:58,kretmsginvalidvectorhandl:58,kretmsgunknownoper:58,kriz:9,krizhevski:41,kseq:41,ksgdsparsecputrain:38,ksimonyan:74,ksum:41,ktest:38,kunusedid_:44,kwarg:[13,15,69,72,74,76,84],kzscore:41,l1_rate:67,l1l2lrregular:50,l1l2regular:50,l1lrregular:50,l1regular:50,l2_rate:67,l2lrregular:50,l2regular:[9,13,17,50],label:[0,9,11,13,15,17,34,36,41,44,69,72,76,84],label_d:28,label_dim:[13,72],labeledbow:17,labelseq:41,labelseqlen:41,labl:17,lag:[58,81],lake:84,lamar:41,lambda:44,lambdacost:40,lambdarank:[41,72],languag:[8,15,17,19,41,72,82,86],larg:[17,19,44,48,78],larger:[7,58,59,69,76],largest:44,last:[0,7,13,17,19,25,34,36,38,41,44,48,69,72,74,81],last_time_step_output:72,lastest:3,late:17,latenc:[7,58,59],later:[2,11,13,34],latest:[2,3,17,41,50,58,59,72],latter:19,launch:[3,6],law:[15,86],lawyer:14,layer1:[72,74],layer2:72,layer3:72,layer:[0,8,9,11,13,15,17],layer_0:21,layer_attr:[0,72,82],layer_num:[11,82],layer_s:[41,72],layer_typ:72,layerbas:21,layerconfig:[21,41],layergradutil:21,layerid:41,layermap:[21,38,41],layermap_:38,layernam:[24,38,60],layeroutput:[15,69],layerptr:[38,41],layers_:38,layerstateptr:41,lbl:[9,69],ld_library_path:[1,7],lda:[25,44],ldb:[25,44],ldc:[25,44],ldd:44,ldi:28,ldo:28,ldt:28,ldv:28,lead:28,leaf:41,learn:[0,9,11,13,17,19,21,41,48,53,67,69,72,74,76],learnabl:[41,72],learning_method:[8,9,13,15,17,19,76],learning_r:[8,9,13,15,17,19,48,67,76],learningextern:80,learningr:[44,50],learningrate_:48,learningrateschedul:48,learningratescheduler_:48,learnrat:50,least:[14,44,69,72],leat:41,leav:84,left:[11,41,44,72],leftmul:44,leman:19,len:[0,13,15,21,24,44,72,84],lenght:[50,56],length:[0,17,19,24,28,30,34,36,41,44,56,72,74,81],less:[7,19,41,53,72],let:[15,41,44,86],level:[7,15,17,19,32,36,41,53,56,58,67,72,81],lib64:[1,3,7,81],lib:1,libatla:1,libcuda:3,libcudnn:1,libgflag:1,libgoogl:1,libgtest:1,libjpeg:9,libnvidia:3,libprotobuf:1,libpython:1,librari:[1,5,7,15,32,56,65,81],licens:[15,86],liearilti:33,life:44,lightweight:3,like:[0,7,8,11,13,15,17,19,24,38,41,48,56,69,72,80,82,84],limit:[15,25,34,38,41,72,84,86],line:[2,7,8,9,11],linear:[30,33,41,72],linearchaincrf:40,linearchainctc:40,link:[13,17],linknam:38,linux:[1,3,19],lipeng:8,lipton:17,list:[0,7,9,11,13,15,17,19,21,24,34,36,41,68,69,72,74,81,82,83,84],listen:[56,81],listwis:41,literatur:17,littl:[13,17,81,83,84],lium:19,liwicki:17,load:[11,15,17,19,25,34,41,50,53,59,68,72,81,83,84,86],load_data_arg:86,load_featur:11,load_feature_c:11,load_feature_pi:11,load_missing_parameter_strategi:[8,80,81,82],loaddata:34,loaddataal:34,loaddatafil:34,loader_:34,loadparamet:[24,38,86],loadparametersremot:[52,59],loadsave_parameters_in_pserv:[80,81],loadvaluerequest:58,loadvaluevector:[53,58],local:[1,2,7,9,10,16,17,18,38,41,44,50,53,59,67,72,80,81],localbias_:41,localbiasgrad_:41,localbuf_:44,localbufrows_:44,localgrad:44,localindic:44,localindices_:44,localoutv:44,localparam:50,localparam_:50,localupdat:59,localupdater_:59,locat:[0,13,21,41,44],lock:[25,41,42,58],lock_:[34,46,64],lock_shar:64,lockedcondit:[38,50,62],log:[1,7,9],log_barrier_abstract:81,log_barrier_lowest_nod:[80,81],log_barrier_show_log:[80,81],log_clip:[80,81],log_error_clip:[80,81],log_period:[9,13,15,17,19,59,81,82],log_period_serv:[80,81],logacts_:41,logger:84,logic:[7,38,84],logicaldeviceid2realdeviceid:38,logicalid:38,login:7,logisticregressionloss:44,logisticregressionlossbp:44,logprob:38,logprob_:41,longer:[19,34],look:[7,13,63,69,80,84],lookup:[13,58],loop:[38,41],loss:[13,17,19,21,41,72],lot:[58,80],low:[44,56,72],lower:7,lowest:81,lstm:[0,13,16],lstm_bias_attr:74,lstm_cell_attr:74,lstm_group:74,lstm_later:41,lstm_layer_attr:74,lstm_size:13,lstm_step:74,lstmcomput:40,lstmemori:0,lstmemory_group:72,lstmlayer:40,lstmsteplay:40,ltr:[41,72],machan:74,machin:[1,2,3,13,15,17,19,21],machineid:38,machineidvec:38,machinest:38,macro:58,made:[0,14,38,84],mai:[2,4,14,15,25,34,38,41,44,48,62,68,69,72,86],main:[9,17,38,48,53,56,62,84,86],mainli:81,maintain:[3,44,48,53,62,72],major:[2,11,17,19,44,50,53,58],majorpartners_:50,majorupd:50,make:[1,2,3,7,9,13,15,17,21,25,34,36,38,41,44,72,84],makeevalu:38,makhoul:41,male:14,malloc:[21,28],manag:[46,53,56,64],manageri:14,mani:[13,14,15,17,32,34,38,41,44,48,58,81],mannal:7,manner:28,manual:1,manufactur:19,mao:17,map:[3,9,11,15,24,25,30,38,41,50,52,53,58,72,81,84],marcu:17,mark:[0,41,44,65,84],markallinputgrad:41,market:[14,17],markinbackward_:41,markinputgrad:41,mask:67,master:[2,17,46,50,81],masterupd:50,mat:[24,41,44,50],mat_cache_row:50,mat_norm:[38,50],mat_normal_shar:50,mat_param_attr:74,mat_sparse_row:[38,50],mat_sparse_row_auto_grow:[38,50],mat_sparse_row_id:[38,50],mat_sparse_row_prefetch:[38,50],mat_sparse_row_prefetch_full_s:[38,50],mat_value_shar:50,matbtyp:44,match:44,matctyp:44,math:[21,40],matirx:[28,41,44,72],matplotlib:9,matric:[0,21,41,86],matrices_:58,matrix:[11,21,24,25],matrixoffset:44,matrixpriv:24,matrixptr:[21,34,36,38,41,44,50],mats_:50,matter:84,mattran:44,mattyp:50,max:[9,13,15,30,36,41,44,63,67,72,78,82],max_:[41,44],max_id:13,max_input:44,max_length:[0,24],max_sequence_length:41,max_sort_s:72,maxaveragewindow_:48,maxcol:44,maxframe_printer_evalu:36,maxgroups_:41,maxid:[13,36,44,69,72],maxid_lay:[13,41],maxid_printer_evalu:36,maxidlay:[38,40],maxim:[19,41,72],maximum:[0,13,17,28,30,44,46,50,69,81],maxindex_:41,maxlay:40,maxlength:24,maxlevel:41,maxpendingconnections_:56,maxpool:72,maxpoolbackward:44,maxpoolforward:44,maxpoolprojectionlay:41,maxrow:44,maxsequencebackward:44,maxsequenceforward:44,maxsequencelength:50,maxsequencelength_:38,maxtrix:28,maxval:44,maxx_:41,mayb:[9,41,44,48,72],mdlstm:40,mdlstmlayer:40,mean:[0,7,8,9,11,13,15,17,19,25,36,38,41,44,50,53,58,59,64,67,69,72,74,76,81,82,84],mean_:41,mean_img_s:9,mean_meta:11,mean_meta_224:11,mean_valu:11,meangrad_:41,measur:41,mechan:[0,17,59,64,72,74],media:17,mem:44,memalloc:44,member:[24,32,34,36,38,41,44,46,48,50,53,56,58],memcpi:28,memeori:41,memhandl:44,memor:17,memori:[0,13,17,19,21,24,25,28,30,34,38,41,44,45],memory_threshold_on_load_data:81,memoryframelin:38,memoryframelines_:38,memoryhandl:[44,46],memoryhandle_:44,memoryhandleptr:[44,46],mere:74,merg:[2,8,19,38,41,44,50,58,59,81],merge_model:38,mergecpugradi:38,mergedict:[8,19],mergegraddens:38,mergegradspars:38,mergegradsparseremot:38,mergeresultsofallcli:36,mergeseg:58,mergetypes_:38,messag:[15,17,19,34,55],messagehead:56,messagelit:58,meta:[7,9,11,13],meta_config:[7,15],meta_fil:15,meta_gener:[7,15],meta_path:9,meta_to_head:15,metho:41,method:[13,15,17,19,21,24,34,38,41,44,50,52,58,63,64,68,72,76,81,82,84],metric:[36,53,80],middl:58,might:[21,38,58,72],million:[14,82],min:[15,41,44,67,82],min_:41,minaveragewindow_:48,mincol:44,mind:7,mine:41,mini:[38,50,59,72],minibatch:[41,72],minim:[76,81],minimum:[28,30,41,44,72],minimun:81,minior:50,minor:50,minordeviceids_:50,minorpartners_:50,minorupd:50,minrow:44,minst:84,minut:[13,19],miss:[7,8,81],mix:[0,40],mixed_bias_attr:74,mixed_lay:[0,41],mixed_layer_attr:74,mixedlay:41,mixedlayertyp:72,mkdir:[1,3],mkl:1,ml_data:[7,15],mnist:[84,86],mnist_model:86,mnist_provid:84,mnist_train:84,mode:[1,7,9,11,13,15,17,19,24,25,30,32,38,41,44,50,53,64,72,81],mode_:41,model_averag:76,model_config:86,model_list:[17,81,82],model_output:17,model_path:82,model_zoo:[8,11],modelaverag:76,modelconfig:[24,38,41],modelconfigpriv:24,modelfil:38,modifi:[0,2,7,21,38,50,52,58],modul:[9,11,13,15,34,53,59,68,75,83,84,86],modulo:[41,44,72],mom:44,momentum:[13,50,52,67,76],momentum_:48,momentumoptim:9,monitor:[13,17],mono:72,month:[13,19],mood:17,more:[0,3,7,9,13,17,19,21,36,38,41,44,50,53,56,58,60,69,72,74,82,83,84,86],morin:[41,72],mose:[17,19],moses_bleu:19,mosesdecod:17,most:[0,15,17,19,21,58,59,64,72,80,84,86],mostli:[9,14,44],move:[15,17,25,41,72],movement:17,movie_featur:15,movie_id:15,movie_meta:15,movieid:14,movielen:12,moving_average_fract:72,movingavgfraction_:41,movingmean_:41,movingvar_:41,mpi:7,mpisize_:58,mse:72,msgreader:[56,58],much:72,mul:[21,44],mulbybitcod:44,mulbybitcodebackwarderror:44,mulbybitcodebackwardweight:44,mulit:7,mulrowvector:44,mulscalar:44,multi:[11,19,21,38,41,44,48,50,53,58,59,62,72,80,81],multi_crop:11,multibinarylabelcrossentropi:40,multibinarylabelcrossentropybp:44,multical:53,multiclasscrossentropi:40,multiclasscrossentropywithselfnorm:40,multicombinedevalu:36,multicor:58,multimachin:38,multimachine_:38,multinetwork:38,multinomi:[41,72],multinomialsampl:40,multipl:[0,13,15,17,21,25,28,34,36,38,41,46,48,50,69,72,74,75,81,82],multiplex:41,multiplexlay:40,multipli:[9,21,41,69,72],multithread:[38,53,58,62,84],multithreadwork:34,mush:2,music:14,must:[0,2,7,19,21,28,34,36,38,41,44,48,53,60,62,65,69,72,74,81,82,84],mutex:[34,44,46,58,64],mutex_:[46,64],mutipl:50,mutual:7,mysteri:14,naiv:[7,44,56],name:[0,3,7,8,9,11,13,15,17,19,21,24,34,38,41,44,46,58,65,67,68,69,72,74,78,81,82],name_:46,namespac:[21,30,44,46,48,50,52],nametofuncmap_:58,nanosecond:64,natur:[17,82],nblocksinallparamet:58,ncelay:40,ndcg:72,ndcg_num:72,nearest:13,neccessi:58,necessari:[1,7,13,17,21,41,53,56,58,59,69,72,84],necessarili:[21,32],need:[0,1,2,3,5,7,9,13,15,17,19,21,25,28,34,38,41,48,50,53,58,59,62,72,74,80,81,82,84],needfre:24,needgradi:41,needgradient_:41,needsequenceinfo_:41,needspecialtravers:[24,48],needtoupdateremot:59,needtrainwholedatainonebatch:38,neg:[13,17,36,44,69,84],neighbor:13,nest:84,net:[3,74],net_conf:17,net_diagram:11,network_config:82,neuralnetwork:36,neuron:[13,17,21,41,86],never:63,newheight:44,newid:38,newmean:25,newmemori:44,newneuralnetwork:38,newnnz:44,newsiz:44,newtork:17,newwidth:44,next:[0,17,19,21,34,38,56,58,59,72,81],nextblock:58,nextpo:41,nic:[3,7,58,80,81],nmt:19,nnz:[21,24,28,32,41,44],nnz_:34,nnzstats_:34,no_cach:84,no_sequ:[15,84],no_sparse_id:24,no_valu:44,noah:17,node:[7,17,19,21,38,41,53,58,59,72,81],nodefil:7,noir:14,nois:41,non:[21,24,25,28,33,34,38,53,64,83],none:[0,11,13,28,67,68,69,72,74,83,84,86],nonlinear:21,nonsmooth:41,nonstaticparaidmap_:52,nonstaticparameters_:38,nonvaluesparsematrix:44,nonzero:[32,38,44],norm:[25,40],norm_by_tim:72,norm_typ:41,normal:[0,5,7,8,11,21,25,30,38,41,44,53,59,62,72,74,81,84],normalizationstrategi:41,normalupd:50,normalupdat:59,normbytim:41,normbytimes_:41,normin_:41,normingrad_:41,normlay:40,normordropnod:38,normordropnodecallback:38,normzal:11,north:9,notat:72,note:[1,4,5,8,9,11,15,17,24,25,28,30,32,33,34,36,38,41,44,50,53,56,58,59,62,63,64,67,72,74,76,78,81,82,84,86],noth:[34,38,41,44,65,81],notic:0,notifi:[38,41,50,62,64],notify_al:64,notify_on:64,notifycopygradtobuff:38,notifygradientcollect:38,notifygradienttransf:38,notifygradreadi:50,notifytaskreadi:38,notifyvaluedispatch:38,notifyvaluereadi:[38,50],novel:17,now:[2,3,15,41,50,53,72,81,84],nproc:1,nstride:25,ntst1213:19,ntst14:19,nullptr:[21,38,41,44,48,50,59,62],nullresultnum_:62,num:[7,13,17,19,25,28,34,44,53,62,72,81],num_channel:[9,72,74],num_chunk_typ:69,num_class:[9,72,74],num_col:58,num_filt:[41,72,74],num_gradient_serv:[80,81],num_group:72,num_parameter_typ:[50,61],num_pass:[13,15,17,19,80,81,82],num_result:69,num_results_per_sampl:72,numaccumulates_:48,numbatch:41,numbatch_:41,numbatches_:59,number:[7,8,9,11,13,17,19,21,24,25,28,30,34,36,38,41,44,50,53,56,58,59,62,63,64,69,72,81,84],number_updat:59,numblocks_:41,numchannel:41,numchunktyp:[36,69],numclass:[41,44],numclasses_:[41,44],numcol:[44,53],numcolumn:44,numdevices_:[38,82],numdims_:41,numer:63,numfilt:41,numfilters_:41,numfram:38,numiov:56,numlogicaldevices_:[38,82],numofallsampl:69,numofthread:44,numofwrongpredict:69,numpassfinishclients_:58,numpi:[1,11,24],numport:53,numports_:53,numprocess:[38,59],numpy_mat:24,numrow:[24,41,44,48],numsampl:[28,41,44,53],numsamples_:[36,41],numsamplesprocess:[24,48],numsamplesprocessed_:58,numsequ:[28,30,38,41],numsequences_:41,numtagtyp:[36,69],numthread:[44,53,62],numthreads_:38,numupdates_:48,numvecslots_:34,numwork:62,nvidia:[1,3,81],obj:[9,11,13,15,34,68,84],object:[9,11,13],obtain:[13,15,17,34,38,41,86],obvious:63,occup:[14,15],occupi:53,occur:[44,48],odd:[41,44,72],off:1,offer:86,offici:[3,9],offload:56,offset:[15,41,44,50,58,72],often:[13,19,59,69],ograd:[21,44],old:[34,38,44,81],oldnumaccumulates_:48,omit:13,on_init:84,onc:[2,13,21,38,41,62,72,84],one_host_dens:15,one_hot_dens:15,oneachstepstart:38,oneachstepstop:38,onehotcrossentropi:44,onehotcrossentropybp:44,onehotcrossentropywithselfnorm:44,onehotcrossentropywithselfnormbp:44,ones_:41,onesequ:41,onewaysearch:38,onli:[0,1,5,8,11,13,14,17,19,21,24,25,28,30,34,36,38,41,44,50,53,58,60,62,64,65,69,72,74,80,81,82,83,84,86],onlin:76,onloadparamet:38,onpassend:[38,41],op_appli:58,op_au:58,op_au_bv:58,op_au_bv_cw:58,op_copi:58,op_cost:58,op_dir_deriv:58,op_finish_pass:58,op_fix_dir_sign:58,op_fix_omega_sign:58,op_load:58,op_make_steepest_desc_dir:58,op_random:58,op_reset:58,op_sav:58,op_sgd:58,op_start_pass:58,op_utv:58,open:[11,13,15,84],openbla:1,openssh:3,oper:[0,8,9,15,21,24,25,28,36,38,40],operationresult:58,operatorconfig:41,operatorfunct:58,operators_:41,opfunc:58,opinion:17,oppos:44,opt:1,opt_config:48,optconfig:[48,53],optconfig_:48,optimiz:[53,58],optimizat:58,optimizationconfig:[24,48,53,58,59],optimizationconfigpriv:24,optimizer_:48,optimizerwithgradientclip:48,optimizerwithregular:48,optimizerwithregularizereverynumbatch:48,optimizerwithregularizerspars:48,optmiz:58,order:[11,13,17,19,21,28,41,44,50,62,72,74,81,84],ordinari:17,org:[15,48,74,86],organ:[9,17,19,36,41,72],orgin:48,origin:[2,19,25,41,44,59,72,83,84],orin:44,ostream:[36,44,50,59],other:[0,1,7,8,9,11,13,14,15,17,19,24,27],otherchunktyp:[36,69],otherid:41,otherwis:[0,1,7,15,19,24,25,34,36,41,44,53,59,68,72,81,82,83],ouputg:41,our:[0,8,9,11,13,17,19,21],out:[0,9,17,24,30,36,41,44,60,72,81],out_:41,out_left:72,out_mem:0,out_right:72,outarg:[24,38,41],outargs_:38,outargsreadysem_:38,outargstream_:38,outdata:30,outdiff:30,outer:41,outerprodlay:40,outframelin:38,outframelines_:38,outgrad:[25,30,44],outgraddesc:25,outmtx_:41,outouth_:41,output1:44,output2:44,output:[0,6],output_:[21,33,41,72],output_d:28,output_data:25,output_dir:11,output_feature_map:25,output_grad:25,output_grad_data:25,output_id:[41,72],output_imag:25,output_image_grad:25,output_lay:11,output_mem:[0,72],output_valu:25,outputbuff:58,outputcheck:41,outputdesc:25,outputdesc_:41,outputgate_:41,outputgrad:[30,32,41,44],outputh:[30,44,72],outputh_:41,outputi:41,outputiov:[56,58],outputlayers_:38,outputmap_:41,outputoffset_:41,outputotherdevice_:41,outputs:41,outputs_:41,outputscor:41,outputsizeh:41,outputsizew:41,outputv:44,outputvalu:[30,32,41],outputw:[30,44,72],outputw_:41,outputx:41,outputx_:41,outputy_:41,outsid:[38,44,72,74,84],outv:[21,30,44],outvec:44,outvtrans_:41,over:[2,13,17,21,72,74,83],overcom:17,overhead:58,overlap:[21,53,59],overload:58,overrid:[21,36,62],owerfunc:62,owlqn:58,own:[7,38,44,53,58],owner:62,ownerfunc:62,ownerthreadid_:62,p_declare_bool:44,pack:[4,59],packag:3,pad:[0,13,25,30,41,72],paddepaddl:83,padding_:[41,64],padding_attr:72,padding_height:25,padding_i:72,padding_width:25,padding_x:72,paddingh:[30,44],paddingh_:41,paddingw:[30,44],paddingw_:41,paddingy_:41,paddl:[1,2,3,5,7,9,13,15,17,19,21,24,33,34,36,38,41,44,46,48,50,52,53,56,58,59,60],paddle_gen_sequ:24,paddle_n:7,paddle_port:7,paddle_ports_num:7,paddle_ports_num_for_spars:7,paddle_pserver2:7,paddle_root:8,paddle_source_root:8,paddle_ssh:3,paddle_ssh_machin:3,paddle_train:7,paddleapi:24,paddledev:3,paddlepaddl:[0,1],paddlepadl:84,paddpepaddl:84,page:[2,15,25],pair:[41,44,53,58,59,69],paper:[8,11,17,19,25,41,48,72],para:[38,41,52,58,59],paraconfig:[48,50],paraconvert:8,paraemet:59,paragraph:17,parallel:[19,30,32,38,41,50,53,56,58,81,82],parallel_nn:[41,80,81],parallelcpuvectort:44,paralleldatarequest:53,parallelexec:44,parallelexecforeachblock:58,parallelinputiov:53,parallelli:[41,48],parallelparamet:50,parallelparametermap:50,parallelparameterptr:50,parallelrequest:53,parallelthread:38,param:[15,24,50,67,72],param_attr:[0,72,74],param_config:52,paramainthread:38,paramainthread_:38,paramattr:[0,67,72],paramconfig:52,paramemt:50,parameter_:[41,50,58],parameter_appli:[48,61],parameter_block_s:[80,81],parameter_block_size_for_spars:[80,81],parameter_col:61,parameter_delta:61,parameter_gradi:[38,48,50,61],parameter_gradient_squresum1:61,parameter_gradient_squresum:61,parameter_gradient_sum:61,parameter_learning_r:[61,67],parameter_momentum:[38,61],parameter_momentum_ut:61,parameter_momentum_vt:61,parameter_row:61,parameter_second_momentum:61,parameter_sum1:61,parameter_sum2:61,parameter_sum3:61,parameter_update_tim:61,parameter_valu:[38,48,50,53,61],parameter_weighted_infinity_norm:61,parameterattribut:[67,72,74],parameterblock:58,parametercli:59,parameterclient2:[36,53,59],parameterclient_:59,parameterconfig:[24,48,50,52,58],parameterconfigpriv:24,parametergradi:41,parameterids_:38,parametermap:[21,38,41,50],parametermap_:[38,53],parametermutex_:58,parameteroptim:[24,48,58],parameteroptimizerpriv:24,parameterpriv:24,parameterptr:[38,41,50,52,53,59],parameterreadybarrier_:58,parameterrelulay:40,parameters_:[21,38,41,52,59],parameterseg:53,parameterserv:58,parameterserver2:58,parameterservic:58,parametertraversecallback:24,parametertraversecallbackpriv:24,parametertyp:[24,38,41,48,50,52,53,58,59,61],parametertypes_:[48,52],parameterupdat:[24,50,52,59],parameterupdated_:38,parameterupdatemod:53,parameterupdatercomposit:[52,59],parameterupdaterhook:52,parametervalu:41,parametris:76,paramid:[38,50],paraminitcallback:[38,50],paramrelubackwarddiff:44,paramrelubackwardw:44,paramreluforward:44,paramselfinited_:38,paramt:8,paramutil:15,paraphrase_data:8,paraphrase_model:8,paraspars:21,parent:[21,41,72],parmet:59,pars:[15,82,86],parse_config:86,parser:15,part:[0,13,15,17,19,21,41,44,56,58,59,84],parti:15,partial:72,partial_sum:28,partialsum_:41,particip:50,participl:8,particular:[38,41],partit:50,partner:[38,50],partnerid_:38,partnerparam_:50,pass:[2,7,9,13,15,17,19,21,24,25,34,38,41,44,48,53,56,58,59,63,68,72,81],pass_:48,pass_gc:61,pass_metr:61,pass_metric_train:61,pass_metric_train_with_noerror:61,pass_test:[38,41,50,60,61],pass_train:[41,61],passbarrier_:58,passcount_:59,passfinish_:53,passtyp:[21,24,38,41,50,52,59,60,61],passtype_:[38,41],password:[3,7],path:[0,1,5,7,8,9,11,13,17,19,24,38,68,72,81,82,83,84],pathsequenceresult:24,paticip:50,pattern:[15,17,38,59],pave:19,pdf:74,peephol:41,peer:25,peerdevic:25,peernam:56,peername_:56,pend:38,per:[9,13,48,58,72,81],perceiv:58,perfer:58,perfom:[81,82],perform:[0,2],period:[13,15,17,19,32,81,83],perl:[17,19],permiss:[15,86],permitrootlogin:3,peroid:[41,72,81],perturb:[21,32],perver:58,phase:25,photo:9,pick:44,pickl:15,picklabl:68,pictur:13,pid:38,pid_t:62,pidqueu:38,piec:[72,75],pillow:9,pin:46,pionter:28,pip:[1,7,9,15],pipe:[14,59],pipelin:59,pixel:[41,72,84],pixels_float:84,pixels_str:84,place:[7,11,19,21,83,84],plai:17,plain:[36,69,72,83],plan:21,platform:[7,13],pleas:[0,3,8,9,13,15,21,28,34,41,48,67,72,74,76,83,84,86],plot:9,plotcurv:9,plu:3,png:[9,11],pnpairvalid:40,pnpairvalidationlay:81,pnpairvalidationpredict_fil:80,point:[28,41,44,59,64],pointer:[25,28,38,41,44,46,58,62],polar:17,polit:17,poll:[17,50],pollut:58,poo:9,pool3:21,pool:[9,13,15,25,30,40],pool_:[44,46],pool_attr:74,pool_bias_attr:74,pool_layer_attr:74,pool_pad:74,pool_siz:[72,74,84],pool_start:74,pool_strid:74,pool_typ:[41,72,74],poolalloc:46,pooledh:30,pooledw:30,pooling_desc:25,pooling_lay:[13,15],pooling_typ:[13,72],poolingdesc_:41,poolingtyp:78,poolint:41,poollay:40,poolmemorysize_:46,poolprojectionlay:40,pooltyp:41,pooltype_:41,pop:60,popd:1,pope:60,popular:11,port:[3,7,53,56,58,80,81],port_:[53,56],port_num:80,ports_num:[56,81],ports_num_for_spars:[80,81,82],posit:[13,17,19,28,30,34,36,41,69,72,84],positive_label:69,possibil:38,post:64,postion:[28,44],postvaluereadi:50,pow:44,pow_:41,power:[13,19,41,72],power_lay:41,powerlay:40,practic:[0,21,68,72],pre:[3,4,6],pre_batch_st:41,pre_dictandmodel:8,precis:[1,36,69],precision_recall_evalu:36,pred:[13,64],predefin:17,predetermin:[19,72,81],predic:[48,64],predict:[0,1,7,8],predict_fil:81,predict_output_dir:[13,80,81],predictarrai:36,predictarray_:[36,41],predicted_label_id:13,predictionresult:[36,41],predictor:15,predin:9,prefer:17,prefetch:[21,38,41,44,50,58,59],prefix:[34,38],pregrad:21,premodel:8,preoutput_:41,preoutputgrad:30,preoutputvalu:30,preoutv:44,preparebatchdata:24,preparedata:53,preparedoper:53,preparesampl:41,preparesenddata:53,prepo:41,preprcess:17,preprint:41,preprocess:[0,7],prerequisit:1,present:[11,19],preserv:44,prev:30,prev_batch_st:[80,81],prev_out:41,prevbatchoutput2_:41,prevbatchst:41,prevent:[58,76,83],prevgrad1:44,prevgrad2:44,prevgradi:28,prevgradx:28,previou:[19,21,32,41,72,74,81],previous:11,prevnumupdates_:48,prevout1:44,prevout2:44,prevoutgrad:32,prevouti:28,prevoutput2batch:41,prevoutput_:41,prevoutvalu:32,prevoutx:28,prevstate_:41,prevstategrad:32,prevstatevalu:32,prevtag:36,prevtyp:36,primarili:17,print:[8,13,15,17,19,36,44,59,67],printal:46,printasyncgradientcommitstatandreset:58,printdiff:44,printoneel:44,printonerow:44,printpredictresult:36,printstat:36,prite:72,privat:[24,44,46,48,50,53,56,58],prob:[36,41,69],probabilist:[8,41,72],probabilit:38,probability_of_label_0:13,probability_of_label_1:13,probabl:[0,2,11,13,36,38,41,69,72],probal:38,probhistori:38,problem:[13,17,36,72,76,86],proce:64,procedur:[8,19],proceed:41,process:[0,7,8,9,11,13,15,17,19,34,36,38,41,50,56,58,59,60,62,64,67,68,72,74,76,81,82,83,84,86],process_pr:13,process_test:68,process_train:68,processchannel:62,processdata:[9,11],processjob:63,produc:[11,13,74],product:[13,17,21,41,53],profil:38,program:[7,81,83],programm:14,progress:81,proivid:84,proj:72,projcol_:41,project:[0,15,21,30,40],projectionconfig:41,projections_:41,projectionstatematrixsize_:41,projoutput_:41,promis:[72,74],propag:[28,33,38,41,76,81,82],propagaion:33,properli:13,properti:[81,84],propos:19,protbuf:34,protect:[21,34,36,38,41,44,46,48,50,52,53,56,58,59,62,64],proto:24,protobuf:[1,24,34,53,58,68],protocli:[53,56],protocol:81,protoin:[53,58],protoout:[53,58],protoread:34,protoresponsecallback:58,protoresponsecallbackex:58,protoserv:[56,58],protoslot:34,protostr:24,protovarslot:34,prove:13,proven:19,provid:[0,3,5,7,8,9,11],provider_:34,providermemory_threshold_on_load_data:80,providerptrtyp:34,providertyp:[34,68],provied:33,provod:84,prune:72,pserver:[7,38],pserver_num_thread:[80,81],pservermatrix:53,pserverstart_pserv:80,pserverstatu:[53,58],pservervector:53,psize:21,pthread_barrier_t:64,pthread_rwlock_t:64,pthread_spinlock_t:64,ptr:[24,34,38,44,46,48],ptype:50,punctuat:17,purchas:13,purpos:33,push_back:21,pushd:1,put:[13,21,62],putdata:53,putowndata:53,py_paddl:[24,86],pydata:68,pydataprovid:[13,17],pydataprovider2:[13,15,34,83],pydataproviderwrapp:[13,17,68,86],pyguard:34,pyint:34,pylong:34,pyobject:34,pyobjectptr:34,python:[0,1,8,9],pythonpath:9,pzo:17,qualiti:[13,58],quaternari:44,queri:[19,25,36,72],querybeamsearch:38,queryid:36,queue:[38,42,53,62],queuecapac:62,quick_start:13,quit:62,quoc:41,quot:14,rabih:41,ragno:41,raii:64,ram:59,ramnath:17,rand:[41,44,81,82],rand_:41,randnorm:44,random:[9,25,32,34,41,44,50,67,72,81],random_int_numb:41,randomizeuniform:44,randomli:17,randomnumberse:80,randparamet:[24,38],randparametersremot:[52,59],rang:[9,14,24,28,41,44,50,62,81,82,84],rangeerror:24,rangereciprocal_:41,rank:[11,13],rankaucevalu:36,rankingcost:40,rankloss:44,ranklossbp:44,rare:84,rate:[7,9,13],rather:17,ratio:81,raw:[13,17,46,72],raw_meta:15,rbegin:41,rdma:[3,56,58,81],rdma_tcp:[80,81],rdmaclient:56,rdmacpu:[56,58],rdmacpu_:56,rdmaserv:56,rdmasocket_:56,rdmauri_:56,read:[0,7,11,13,15,34,53,56,58,59,64,83,84],readallblock:58,readblock:56,reader:[19,55],readi:[9,38,41,44,50,53,58],readm:[14,15,17],readmessag:56,readnextblock:56,readv:56,readwritebuff:58,readwritebuffer_:58,real:[25,28,30,32,34,36,38,41,44,48,50,52,53,58,59,84],realdeviceid2logicaldeviceid:38,realid:38,realiti:58,reallay:[38,41],reallayer_:41,reallayernam:38,reallayers_:41,realoutarg_:41,reason:[48,58,72],recal:[36,69],receiv:[36,53,58,68],recent:[1,19,59],reciev:[53,81],reciproc:44,reciprocalderiv:44,reciprocalrowsum_:41,reciprocalsum:44,recogn:9,recognit:[11,17,84],recommand:[74,84],recommend:[0,3,7],recommonmark:1,recor:48,record:[15,25,38,41,58],recordhistori:38,rectangular:72,recurrent_group:0,recurrentgradientmachin:38,recurrentgroup:[36,69],recurrentlay:40,recurrentlayergroup:41,recurrfgw:41,recurrigw:41,recurriw:41,recurrogw:41,recv:[53,59],recvbuf:53,recvdata:53,recvdatamems_:53,recvjobqueue_:53,recvparamet:53,recvparametertyp:53,recvsyncbarrier_:53,recvthreads_:53,reduc:[7,41,53,58,76,81,82],reduceandsenddata:58,redund:58,reeal:41,refactor:41,refer:[0,3,7,8,9,13,15,16],referenc:72,refmat:44,regard:53,regardless:19,regex:15,reggradientmachinemod:38,region:[41,72],regist:[21,38,41,58],register_lay:21,register_service_funct:58,register_service_function_ex:58,registerbeamsearchcontrolcallback:38,registerbeamsearchstatisticscallback:38,registerservicefunct:58,registerservicefunctionex:58,registerservicefunctionimp:58,registrar_:[34,36,41],regress:12,regression_cost:15,reguardless:62,regular:[9,13,17,21,48,49],regularizer_:48,rel:[7,53,74,83],relai:59,relat:[1,5,7,15,17,29],relationship:7,releas:[3,4,5,14,44,46,53,72],releasematrix:[53,58],releasematrixrequest:58,releasepass:53,releasevector:[53,58],releasevectorrequest:58,relev:19,relu:[21,28,30,33,44,65,72],reluderiv:44,remain:[13,56,59,62],remot:2,remoteparameterupdat:52,remov:[7,17,38,81],removebeamsearchcontrolcallback:38,removebeamsearchstatisticscallback:38,renam:19,renshaw:41,reorder:41,reorgan:[41,72],replac:[3,17],replic:41,repo:2,repons:[50,58],report:7,repositori:2,repres:[0,9,13,14,21,41,44,76,84,86],represent:[13,17,32,41],reproduc:19,requestvec_:58,res5_3_branch2c_bn:11,res5_3_branch2c_conv:11,research:[9,14,17],resembl:17,reserv:[15,44,84,86],reserveoutput:[21,41],reservestor:44,reset:[25,34,41,44,58,59,72],resetalignalloc:58,resetconvoutput:41,resetcurrentstat:59,resetexpandinput:41,resetg:32,reseton:44,resetoutput:[32,41],resetoutput_:41,resetoutputgrad:32,resetoutputvalu:32,resetspecifyoutput:41,resetst:[38,41],reshap:[25,40],reshapedoutputgrad:41,resid:25,residu:11,resiz:[24,41,44,58],resizeandcopyfrom:50,resizebootfram:38,resizecsc:44,resizecsr:44,resizelay:40,resizeorcr:[41,44],resizeorcreatebatch:41,resizeorcreatefram:38,resizeorcreatesparsematrix:44,resizeoutput:41,resizewithalignhint:58,resnet_101:11,resnet_152:11,resnet_50:11,resolut:32,respect:[0,9,11,19,21,41,44,81,84],respons:[38,41,44,50,53,56,58,72],responsecallback:[56,58],responsenormlay:40,respos:30,rest:[34,72,84],restart:38,restor:[44,48,52,59],restrict:[38,81],result:[0,6],result_fil:[0,69,72],result_of:62,resultptrtyp:62,resultrunningmean:25,resultrunningvari:25,results_:62,resulttyp:62,retir:14,retriev:[21,38,41,53,58],return_seq:74,reus:[21,41],revers:[0,17,30,41,44,72,74],reversed_:[38,41],review:[13,17],rewrit:19,rgb:72,rgen:17,rho:76,richard:41,right:[11,15,36,41,44,72,84,86],rightmul:44,ring:38,rmsprop:[13,76],rmspropoptim:15,rmspropparameteroptim:48,rnn:[0,13,17,22],rnn_bias_attr:0,rnn_layer_attr:0,rnn_out:0,rnn_step:72,rnn_use_batch:[41,80,81],robert:41,robot:9,robust:[7,41],roc:36,roce:3,role:[0,17,50],role_:50,roman:17,romanc:14,root:[1,3,7,30,41,76,78],root_dir:7,rootag:38,rootid:53,rootlay:38,rootnetwork:[38,41],rootnetwork_:38,rou_:48,rout:53,routin:[15,32,53,58,59],routledg:17,row:[11,21,24,25,28,32,36,38,41,44,48,50,53,58,69,72,86],row_id:[41,72],row_siz:28,rowadd:44,rowbuf:44,rowdotmul2:44,rowdotmul:44,rowin:44,rowindex:44,rowmax:44,rowmaxid:44,rownormalizel1:44,rowpow:44,rows_:44,rows_d:28,rows_in_block:48,rowscal:44,rowscale2:44,rowstore_:44,rowsum:44,rowvec:48,rpc:58,rtype:[15,72],rule:[21,53],run:1,runninginvvar:25,runningmean:25,runtim:[7,25,83,84],rvalu:63,rwlock:[34,46,58],rwlock_:64,s_fusion:15,s_id:15,s_recurrent_group:0,sacrif:83,safe:[48,63],safeti:60,sai:[81,82],sake:[21,44],sale:14,same:[0,1,3,7,8,13,15,17,19,28,36,38,41,44,48,50,52,53,59,62,64,68,69,72,74,82,84,86],samping_id:72,sampl:[7,8,11,13,15,17,19,34,36,38,40],sample1:34,sample2:34,sample_num:69,sampleloop:34,samplen:34,samplenums_:34,sampler:41,sampling_id_lay:41,samplingidlay:40,santiago:17,sate:74,satisfi:[7,13,53],save:[9,11,13,15,17,19,25,41,50,53,59,69,72,81,82,84],save_dir:[9,13,15,17,19,81,82],save_mean:[25,41],save_only_on:[80,81],savedinvvar:25,savedinvvar_:41,savedir:53,savedmean:25,savedmean_:41,savedvar:25,saveparamet:38,saveparametersremot:[52,59],saver:44,savevaluerequest:58,savevaluevector:[53,58],saving_period:[80,81],saving_period_by_batch:[13,80,81,82],saw:84,sbin:3,scalabl:58,scalar:[25,28,41,72,84],scalardiv:44,scale:[11,14,15,25,28,30,41,44,50,53,65,72],scale_:41,scalea:30,scaleab:44,scaleb:30,scaledtanh:44,scaledtanhderiv:44,scalegrad:25,scaleoutput:44,scalesum:44,scalet:44,scaletarget:44,scaling_lay:41,scalinglay:40,scanner:34,scatter:[38,41,72],scatterag:38,scatteragentlay:40,scenario:80,scene:80,schedul:7,scheduler_factor:67,schema:8,scheme:[36,69],schmidhub:17,schwartz:41,schwenk:19,sci:14,scienc:17,scientist:14,score:[15,17,19,24,36,41,69,72],screen:15,scrip:13,script:[7,9,11,13,17,19,86],seaplane_s_000978:9,search:[0,19,38,41,72,81],seat:19,sec:50,second:[2,7,8,11,13,14,15,17,32,38,41,44,53,59,64,72,84],secondli:[7,53],section:[0,7,13,21,84],sed:[3,17,58],see:[2,8,11,13,15,17,19,25,41,50,72,74,84,86],seed:[25,81],segment:[36,50,53,58,69],segmentor:8,segmentrang:41,sel_fc:72,select:[2,14,19,28,34,41,44,72],selectel:44,selectfrom:44,selectiv:72,selective_fc_lay:41,selectivefullyconnect:41,selectivefullyconnectedlay:40,selectrow:[44,50],selectrowsimp:44,selectrowsonetim:38,self:[14,17,21,24,41,50,56],selfnorm:72,sem_:64,sem_t:64,semant:[0,17],semaphor:[38,50],send:[38,53,58,59,81],sendandreceiveparamet:53,sendbackparamet:[53,58],sendbackparameterspars:58,sendbackparametertyp:53,sendbuf:53,senddata:[53,58],senddatarequest:[53,58],senddatarequestvec:53,senddatatyp:53,sendjob:53,sendjob_:53,sendjobptr:53,sendjobqueue_:53,sendparallel:53,sendparamet:[53,58],sendparameterrequest:[53,58],sendparameterrespons:58,sendqueu:53,sendrequest:53,sendthreads_:53,sens:[36,53,72],sent:[53,58,59],sent_id:0,sentenc:[0,13,17,19,24,84],sentiment:[12,13],sentiment_data:17,sentiment_net:17,sentimental_provid:84,separ:[8,13,14,15,19,38,53,58,59,62,69,81,84],separatesendandrecv_:[53,59],seq2batch:[30,41],seq2batchidx:41,seq2batchidx_:41,seq:[15,44],seq_pool:72,seq_text_print:72,seq_to_seq_data:[8,19],seq_typ:[15,84],seqendidxinbatch_:41,seqflag:50,seqid:[38,41],seqidx_:41,seqlength:41,seqlengthandstart:50,seqlengthandstart_:38,seqsiz:50,seqstart:[41,50],seqstartpo:50,seqstartposindex:[38,41],seqstartposindex_:41,seqtext_printer_evalu:36,seqtoseq:[0,8,19,72],seqtoseq_net:[0,8,19,72],sequel:84,sequence2batchadd:41,sequence2batchcopi:41,sequence_conv_pool:13,sequence_layer_group:72,sequence_max:33,sequence_nest_layer_group:72,sequenceagentlay:40,sequenceavgforward:44,sequenceconcatlay:40,sequencegatheragentlay:40,sequencegener:24,sequencegeneratorpriv:24,sequencelastinstancelay:40,sequencelength:30,sequencelevel:41,sequenceloop:34,sequencereshapelay:40,sequencescatteragentlay:40,sequencesoftmax:44,sequencestartposit:[38,41,50,72],sequencestartpositions_:34,sequencetobatch:40,sequencetyp:84,sequenti:[0,13,41,62,72,83],sequentianlli:62,seqvalu:41,seqwidth:30,seri:17,serial:84,serious:3,server:[3,7,11,12,21,38,42,44,53,55],server_:56,serveraddr:56,serverid_:58,serverport:56,servic:[14,53,58],servicefunct:58,servicenum:53,servicenum_:53,set:[0,1,3,5,7,8,9,11,13,14,15,17,19,21,24,25,33,34,36,38,41,44,46,48,53,56,58,59,67,69,72,74],set_active_typ:21,set_default_parameter_nam:67,set_drop_r:21,set_siz:21,set_typ:21,setbatchs:24,setbeams:24,setbo:24,setconfig:[53,58],setconfigrequest:58,setdata:[34,41,44],setdevic:50,setdict:24,setel:44,seteo:24,setforwardbackwardtim:[52,53,59],setframeheight:50,setframewidth:50,setid:50,setintbuf:50,setitem:24,setmat:50,setmaxlength:24,setmeanandstd:41,setneedgradi:41,setneedsequenceinfo:41,setnodecai:48,setnumofthread:44,setoutput:41,setoutputgrad:38,setparamet:[53,58],setparameterptr:50,setparameterzero:53,setpassgrad:38,setreallay:41,setreallayerandoutput:41,setrow:44,setsequencestartposit:41,setsharedcount:50,setskipshuffl:34,setslotid:24,setslotin:24,setslotsequencedim:24,setslotsequencestartposit:24,setslotsubsequencestartposit:24,setslotvalu:24,setsourcedata:44,setstat:[38,41],setstatu:[53,58],setstatusrequest:58,setsync:44,settrainerid:53,settup:21,setup:[13,21,44],setupindic:44,setvalueupd:50,sever:[7,13,15,17,19,38,41,53,58,68,72,82,84],sftmaxsum:44,sftmaxsum_:41,sftmaxsum_d:28,sgd:[7,17,19,50,53,58,59,76],sgdasync_count:80,sgdlocalupdat:52,sgdoptim:48,sgdthreadupdat:52,sgdupdat:[44,50],shaaban:41,shadow:59,shake:41,shalabi:41,shall:64,shallow:44,shape:[11,41,72],share:[1,38,41,44,50,52,53,56,58,64,72,81],sharecpumatrix:41,shared_bia:74,shared_bias:[41,72],shared_ptr:[34,41,44,46,50,52,53,62],sharedbiases_:41,sharedcount_:50,sharedcpumatrix:44,sharedptr:24,shareindexwith:41,shell:11,shift:[11,41],shift_start:41,shiftdim:41,ship:9,shold:[2,17],shop:17,shorter:11,should:[0,7,9,13,15,17,19,24,34,36,38,41,44,48,50,53,60,62,64,69,72,76,84],should_shuffl:84,shouldbem:38,shouldn:2,show:[2,8,11,13,15,17,19,41,59,76,81,86],show_check_sparse_distribution_log:[80,81],show_layer_stat:[80,81],show_parameter_stats_period:[13,17,19,80,81,82],showdatastat:34,shown:[0,9,11,13,15,17,19,21,36,69,72,84],showoutputstat:41,showstat:59,shrink:[21,30,41,63],shrinkmat:41,shuf:15,shuffl:[15,17,34,84],shuffledsequenceids_:34,side:[11,24,53],sigint:7,sigmoid:[21,30,33,41,44,65,72],sigmoid_threshold_max:30,sigmoid_threshold_min:30,sigmoidderiv:44,sign:44,signal:[7,33,53,59],signatur:58,sim:28,simd:44,similar:[13,15,32,41,44,72],similarli:72,simple_attent:0,simple_gru:13,simple_lstm:[13,72],simple_rnn:[0,72],simpledataproviderbas:34,simpli:[0,2,8,11,15,17,19,41,72,83],simplifi:[21,58],simultan:[44,62],simutan:58,sinc:[3,11,13,14,19,58,64,72],sincer:[2,17],singl:[1,11,13,19,21,34,36,38,44,50,53,56,58,59,64,68,69,74,76,84],singleton:[46,62],singleupd:50,situat:38,six:[3,8,19],size:[0,7,9,11,13,14,15,17,19,21,24,25,28,30,32,34,38,41,44,46,50,53,56,58,59,62,63,69,72,74,76,81,84],size_:[34,41,44,46,58],size_a:72,size_b:72,size_t:[21,24,25,28,30,32,34,36,38,41,44,46,48,50,52,53,56,58,62,63],sizei:28,sizeinbyt:25,sizelimit:46,sizelimit_:46,sizeof:[8,41,44,46,58,64],sizex:[30,44],sizex_:41,sizey_:41,skill:19,skip:[7,11,34],skipshuffle_:34,slave:[38,50],slaveparam:50,slaveparamet:38,slaveupd:50,sleep:53,slide:76,slightli:9,slope:[41,72],slope_intercept_lay:41,slopeinterceptlay:40,slot:[0,15,24,34],slot_:34,slot_dim:15,slot_nam:15,slotdef:34,slothead:34,slotnum:24,slots_:34,slottyp:[15,34],slow:58,small:[7,9,19,21,38,41,58,81,84],small_messag:[80,81],small_vgg:9,smaller:[38,41,72],smallest:[41,53],smat:28,smatrix:44,smatrix_:44,smemoryhandl:44,smemoryhandle_:44,smith:17,smooth:4,snippet:[0,13,21],social:17,sock:56,sock_recv_buf_s:[80,81],sock_send_buf_s:[80,81],socket:55,socket_:56,socketchannel:56,socketcli:56,socketdaemon_:56,socketserv:[56,58,62],socketwork:[56,62],socr:72,soft:41,softbinaryclasscrossentropi:40,softcrossentropi:44,softcrossentropybp:44,softmax:[0,8,13,17,21,25,28,33,41,44,65,72,74],softmax_param_attr:74,softmax_selfnorm_alpha:72,softmaxactiv:[0,13],softmaxbackward:44,softmaxderiv:44,softmaxseq:41,softmaxseqgrad:41,softmaxseqlen:41,softrelu:[33,44,65],softreluderiv:44,softwar:[15,86],solut:53,solver:19,some:[1,3,7,13,14,15,17,19,21,30,36,38,41,44,48,50,53,56,58,67,72,76,80,81,82,84],someth:[62,64,72,84],sometim:[17,76],sophist:21,sort:[15,17,19,24,41,44,72,81],sourc:[0,1,3],source_dict_dim:0,source_language_word:0,sourcedata_:44,sourcedatavec_:44,sourcevec:44,space:[0,34,41,46,69],space_seperated_tokens_from_dictionary_according_to_seq:[36,69],space_seperated_tokens_from_dictionary_according_to_sub_seq:[36,69],spars:[7,13,21,24,27],sparse_binary_vector:[13,84],sparse_csc:[24,44],sparse_csr:[24,44],sparse_float_value_t:[34,44],sparse_float_vector:84,sparse_format:44,sparse_non_valu:[24,34],sparse_non_value_t:[34,44],sparse_nonvalu:24,sparse_upd:67,sparse_valu:[24,34],sparseautogrowrowcpumatrix:[38,44],sparsecopyfrom:24,sparsedistribution_:53,sparsefloatvaluedata:34,sparseformat:[44,50],sparseformattyp:24,sparseid:[24,48],sparsematrix:[28,44],sparsemomentumparameteroptim:48,sparsenonvaluedata:34,sparseparam:21,sparseparameterdistribut:53,sparseprefetchrowcpumatrix:[21,38,44,50],sparseremoteparameterupdat:52,sparseremoteparameterupdatercomposit:52,sparseres:44,sparseresizecsc:44,sparseresizecsr:44,sparserowcpumatrix:[38,44,50],sparserowcpumatrixto:38,sparserowidscpumatrix:[38,44],sparserowmatrix:44,sparsevaluetyp:[24,44],sparsiti:38,spatial:[9,41],spawn:56,spe:36,speak:[0,19],specfii:[41,81],speci:9,special:[8,13,19,34,58,72],specif:[9,13,15,25,44,53,58,82,83,86],specifi:0,speed:[41,48,74],spefici:11,sphinx:1,sphinx_rtd_them:1,spin:64,split:[7,8,11,13],splitbydataid:50,splite:24,sprare:28,sql:83,sqr:44,sqrt:[30,44],squar:[30,33,44,65,72,76,78],squarederiv:44,squarerootnpool:72,squash:19,srand:81,src:[1,19,24,25,28,30,41,44,50,53],src_backward:0,src_d:25,src_dict:0,src_embed:0,src_forward:0,src_h:25,src_id:0,src_word_id:0,srcid:38,srcnn:38,sse:58,ssh:[3,7],sshd:3,sshd_config:3,stack:[13,16],stacked_lstm_net:17,stacked_num:17,stage:[7,24,58,59],stake:19,stale:58,standard:[8,9],standarddevi:44,stanh:33,star:14,start:[0,2,7,8],start_:[41,44],start_pass:[80,81],start_pserv:81,startbatch:[24,48,52,58,59],startcatchupwith:48,startcol:44,startcontrol:59,startfil:34,startidx:41,startpass:[24,48,52,59],startprepar:34,startrow:44,startseq:50,startspo:44,starttask:38,startthread:53,starttrain:24,starttrainpass:24,stat:[17,19,36,58,59,81],state2_:41,state:[0,4,19,30,32,38,41,48,58,59,72,74,81],state_:41,state_act:[72,74],stateactive_:41,stateactivegrad:32,stateactivevalu:32,stategrad:[30,32],statement:21,statevalu:[30,32],stateweight:32,stateweight_:41,stateweightgrad:32,staticcastvector:24,staticinput:[0,72],staticpruninghook:52,statist:[13,17,19,36,38,41,59,72,81],statptr:34,statset:58,statset_:58,statsinfo:36,statu:[2,38,53,58,69],status_:58,std:[21,24,33,34,36,38,41,44,46,48,50,52,53,56,58,59,60,62,64,81],stderr:7,stdgrad_:41,stdout:7,stdreciprocal_:41,step:[0,7,13,15,17,19,21,34,38,41,72,74,78,86],step_:48,still:[4,11,41],stochast:76,stock:17,stop:[3,7,15,34,38,62,72,81],stopaddjob:62,stopbeamsearch:38,stopbeamsearch_:38,stopping_:[38,53,56,62],storag:[9,44,46],storageengin:46,store:[7,8,9,11,13,15,17,19,21,25,34,38,41,44,50,53,56,58,59,62,72,81,84],storemat_:44,str:[24,41,50,82],straight:2,strat:25,strategi:[34,50,78,81,84],strdata:34,stream:[25,28,32,34,38,44,50,56,59],strictli:58,stride:[25,30,41,44,72],stride_:[41,44],stride_height:25,stride_i:72,stride_width:25,stride_x:72,strideh:[30,44],strideh_:41,strideheight:41,stridew:[30,44],stridew_:41,stridewidth:41,stridey_:41,string:[17,21,24,25,33,34,38,41,46,50,52,53,56,58,59,60,68,72,81,83,84],strip:[13,15],struct:[24,25,32,34,36,38,41,44,50,53,56,58,64],structur:[7,8,9,13,15,17,19,38,41,50,58],student:14,stun:84,style:[1,2,38,56,72],sub:[0,9,13,19,21,36,38,41,44,69,72],sub_sequ:84,sub_work_thread:38,subargfrom:50,subbybitcod:44,subclass:[24,33,34,36,38,41,44,46,48,50,52,53,56,58,59,62],subcolmatrix:44,subdataproviders_:34,subgradi:76,subindic:34,subk_:41,subm_:41,submatrix:44,submit:[2,80,81],submodel:[38,41],submodelnam:38,submodelname_:38,subn_:41,subnetwork:41,subnetwrk:41,subrowmatrix:44,subsampleloop:34,subscalar:44,subsequ:41,subsequenceinput:72,subsequencelay:40,subsequencestartposit:[41,50],subset:[19,21,38,40],substanti:11,substitut:19,subvec:44,subvecfrom:44,succe:17,success:[11,36],successfuli:17,successfulli:[11,15,17,44,64],successor:[19,81],sucessfulli:19,sudo:[1,9],suffici:81,suffix:19,suggest:[25,72],suit:41,suitabl:[9,11,53,81],sum:[0,21,28,30,36,41,44,50,53,69,72,76,78],sum_evalu:36,sum_j:44,sum_to_one_norm:72,sum_to_one_norm_lay:41,sumbybitcod:44,sumcol:44,sumcost:50,suminv_:41,summar:[13,17],sumofsquar:44,sumofsquaresbp:44,sumofsquarescostlay:40,sumpool:72,sumrow:44,sumtoonenormlay:40,supervis:34,suppli:[34,38],support:[0,1,3,5,11,21,24,28,30,32,34,36,38,41,44,50,53,59,65,67,69,72,76,81],support_hppl:65,suppos:[13,21,33],sure:[2,7,9,17,21,34,44],survei:17,suspend:38,svectorptr:50,swap_channel:11,swig:[1,24,86],swig_paddl:86,sxi_sock:56,sxi_socket:56,symbol:72,sync:[2,44,53,58,59,62,81],sync_:44,sync_default:53,syncedflag:44,syncflag:[21,44],syncflag_:44,synchron:[7,25,28,38,50,53,58,59,76,81],synchronizebarriers_:58,synchronizeparamt:50,synchronizerequest:58,syncobject:53,syncobjectid:53,syncparamet:50,syncthreadpool:[44,52,53,58,59],syncthreadpool_:[52,53,58],synctrainersteps_:58,syncupd:50,syncupdaterstep:58,syntax:15,system:[1,13,17,19],szegedi:41,t0vec_:48,t2b:8,tab:13,tabl:[11,13,19,28,41,44,58,72,84],table_:41,table_project:41,tablemattyp:44,tableproject:40,tables:28,tag:[0,3,5,36,69],tagtyp:[36,69],take:[0,13,19,21,24,36,41,56,58,64,69,72,74,84,86],taken:84,tal:41,tanh:[21,30,33,44,65,72,74],tanhactiv:0,tanhderiv:44,taobao:17,tar:1,target:[0,8,13,19,34,72],target_dict_dim:0,target_language_word:0,targetgrad:30,targetperdim_:41,task:[0,8,11,17,19,25,38,41,44,69,72,82,84],task_backward:38,task_copy_in_arg:38,task_forward:38,task_forward_backward:38,taskreadysem_:38,tasktyp:38,tasktype_:38,tau_:48,tcp:[56,58,81],tcpclient:56,tcprdma_:56,tcpserver:56,tcpsocket_:56,teach:13,team:3,technician:14,techniqu:[0,21],tee:[9,15,17,19],tell:[15,53],tellig:17,templat:[24,28,30,34,38,41,44,46,53,58,60,62,63,64],templatereducesum:58,tempor:[13,72],temporari:[41,44],temporarili:44,tenor:41,tenro:41,tensor:[25,41,72],tensor_lay:41,tensorlay:40,term:[17,72,74],ternari:44,tese:83,test:[1,2,7,8,9,11,13,14],test_all_data_in_one_period:[9,15,17,81,82],test_data:[19,86],test_fcgrad:21,test_layergrad:21,test_list:[9,13,68,84],test_part_000:17,test_pass:[19,80,81,82],test_period:[80,81,82],test_ratio:15,test_wait:[80,81],testabl:64,testconfig:21,testdata:17,tester:[15,19],testfcgrad:21,testfclay:21,testing_:59,testlayergrad:21,testmodel_list:80,testresult:62,testsave_dir:80,testutil:21,text:[8,12,13,15,17],text_conv:13,text_conv_pool:15,text_fil:17,tgtdata:30,tgz:1,than:[0,7,11,17,19,21,34,36,38,41,53,58,59,62,64,69,72,74,76,84],thank:[8,19],the_total_number_of_paramet:50,thead:62,thei:[0,11,17,21,34,38,41,44,80,84],them:[0,9,11,13,15,19,34,36,41,44,53,59,74,80,81,83,84],therun:11,thi:[0,2,7,8,9,11,13,14,15,17,19,21,24,25,30,34,36,38,41,44,46,48,50,53,56,58,59,62,63,64,67,68,69,72,74,76,81,82,83,84,86],thing:[0,1,15,84],third:[11,17,38,72],this_col:44,this_row:44,this_row_i:44,thisimgheight:44,thisimgwidth:44,thoma:41,those:[11,38],though:62,thread1:63,thread2:63,thread:[15,17,19,21,25,31],thread_:62,thread_local_rand_use_global_se:[80,81],threadbarri:[38,53,58,62],threadid:[38,44,53,82],threadid_:38,threadloc:[34,44,50,58,60],threadlocald:34,threadnum:62,threadnum_:53,threadpool:53,threadptr:53,threads_:38,threadsaf:44,threds_:38,three:[0,11,17,19,28,41,50,69,72,76,81,84],thresh:41,threshold:[30,44,58,67,69,76,81],threshold_:48,thriller:14,through:[0,7,8,9,17,19,21,38,41,59,62,64,72,86],throughout:13,thu:[19,21,38,72,84],tid:[44,53,62],time:[0,1,3,13,14,17,19,25,41,44,48,53,58,62,63,64,72,74,78,81,82,84],timelin:[41,72],timeout:64,timer:31,timer_:48,timespec:64,timestamp:[14,41,72],timestep:[30,34,41,72,84],timev:58,timewait:64,timewaitgradreadi:50,titil:15,titl:[2,14,15,36],tlstempbufs_:50,tmall:17,tmp:44,tmpbiasgrad_:41,tmpcpuinput_:41,tmpdest_:41,tmpgrad_:41,tmpmat_:[41,44],tmpmatrix:41,tmpmtx0:41,tmpmtx1:41,tmpmtx:41,tmprow0:41,tmprow1:41,tmprow2:41,tmprow3:41,tmpsrc_:41,tmpwgrad_:41,todo:[53,69,72,74],togeth:[0,41,64,72,74],token:[0,8,17,19,69,72,84],tonumpyarrayinplac:24,tonumpymat:24,tonumpymatinplac:24,too:58,tool:[0,3,11,17,38],toolkit:[1,5],top:[11,28,44,60,69],topid:28,topindex:38,toprotostr:24,topval:28,toronto:9,total:[7,8,19,34,36,41,53,56,59,69],totallength:56,totalpad:[30,44],totalscore_:36,totalsegments_:41,totalstate_:41,totaltime_:41,touch:17,tourism:17,tourist:19,tra:19,trace:60,track:72,track_:41,tractabl:72,tradesman:14,tradit:72,traffic:53,train:0,train_async:68,train_conf:[8,19],train_data:19,train_list:[9,11,13,68,84],train_part_000:17,trainabl:[30,72],trainable_pad:41,traindata:17,traindot_period:80,trainer:[7,13],trainer_config:[7,13,15,17,83,84,86],trainer_config_help:[9,13,15,21,65,67,68,69,72,74,76,78,84],trainer_count:[15,17,19,80,81,82],trainer_id:[58,81],trainer_role_control:50,trainer_role_major:50,trainer_role_mast:50,trainer_role_minor:50,trainer_role_singl:50,trainer_role_slav:50,trainerbarrier_:38,trainerconfig:[24,38],trainerconfigpriv:24,trainercount:38,trainerid:53,trainerid_:53,trainerintern:[13,15,19],trainerpriv:24,trainerrol:50,trainerthreadptr:38,trainingtest_all_data_in_one_period:80,trainison:38,trainonebatch:24,trainrol:50,trainwholedatainonebatch:38,tran:[21,24,25,44,50,53,72,81],trane:[81,84],trans_:44,trans_full_matrix_project:41,trans_lay:41,transa:[25,28,44],transact:17,transb:[25,28,44],transdatatyp:53,transform:[0,9,13,21,33,50,86],transform_param_attr:74,translat:[8,15,17,19,41,72,74],translay:40,transoutvalue_:41,transpar:7,transport:[58,81],transpos:[21,25,28,32,41,44,72],transposedfullmatrixproject:40,transposit:[41,72],tranvers:38,travel:[38,84],travers:48,traversecallback:[48,58],traversecallbackvec:48,treat:[0,34,38,72],tree:[2,19,41,72,81],trg:19,trg_dict:0,trg_dict_path:0,trg_embed:0,trg_id:0,trg_ids_next:0,triain:83,trimfrom:44,trimfromcsc:44,trimfromcsr:44,trivial:84,trn:13,truck:9,true_typ:44,trust:7,truth:[13,19,69,72],trycreategradientmachin:38,trygetmod:38,tst:13,tune:[13,15,19,58,59,67],tuningasyncsgdfinish:58,tuningasyncsgdmidoutput:58,tuninglog_barrier_abstract:80,tuningsgdfinish:58,tuningsgdmidoutput:58,tupl:[34,38,50,68,72,74],ture:[38,41,64,72],turn:72,tutori:[0,2,7],tweet:17,twelv:19,twitter:17,two:[0,5,7,8,9,11,13,15,17,19,25,34,38,41,44,48,50,53,58,59,65,72,74,82,83,84],txt:[7,13,15,17,21,72,84],type:[0,9,11,13,15,21,24,28,30,32,33,34,36,38,41,44,46,48,50,52,53,56,58,59,62,68,69,72,74,76,78,79,81,82,84],type_:41,type_info:53,type_nam:[15,72],typecheck:41,typedef:[25,30,32,34,38,41,44,46,48,50,53,56,58,62],typenam:[24,38,41,44,53,58,60,62,63],typic:[17,34,36,69,86],u_r:41,u_z:41,ubuntu:3,udp:50,udpat:59,ufldl:72,uint32_t:[44,50],uint64_t:[44,50,52,53,58,59],unari:44,unbalanc:81,unbound:0,unconstrain:17,under:[14,15,17,25,36,53,58,86],understand:[8,9,17,58],understudi:19,unemploi:14,uniform:[44,67,81],uniform_real_distribut:41,uniformli:25,uniqu:[2,81],unique_ptr:[21,34,38,41,44,46,48,50,52,53,56,58,59,62],unit:[0,2,20],unittest:1,unittestcheckgrad_ep:80,univ:19,univari:44,unk:[8,19],unk_idx:13,unknown:72,unlabel:17,unless:[15,86],unlik:[17,19,62],unlimit:34,unlock:64,unnorm:41,unordered_map:[46,53,58],unshar:41,unsign:[25,44,46],unsup:17,unsupbow:17,unsupporterror:24,until:[7,41,44,50,53,56,62,63,64],unus:[15,44,58],unzip:15,updat:[1,2,3,7,17,21,24,28,36,38,42,44,48,50,51],update_type_num:50,updateaveragewindowlimit:48,updatecallback:[21,24,38,41,50],updatecallbackinpython:24,updatecallbackwrapp:24,updatecounter_:[38,50],updated_:50,updatefunct:50,updateg:32,updatehook:50,updateimpl:[50,52,59],updatemod:53,updater_hook:52,updater_norm:59,updater_sparse_remot:59,updaterhooks_:50,updaters_:52,updatesamplesnum:36,updatethreadparamet:38,updatewithgradi:50,upon:84,upstream:2,url:17,urls_neg:17,urls_po:17,urls_unsup:17,urng:41,usag:[8,34,58,59,69,72,74,83,84],usageratio_:34,use_global_stat:72,use_gpu:[9,11,13,15,17,19,44,80,81,82,86],use_jpeg:9,use_old_updat:[80,81],use_seq:15,use_seq_or_not:15,useapply_:48,useapplyinpserver_:59,usebatch_:41,usedsegments_:58,useful:[24,72,84],useglobalstats_:41,usegpu:[21,24,34,38,41,44,50],usegpu_:[34,38,41,44,50],usel1:44,useless:7,usepam:3,useparameterappli:48,user:7,user_featur:15,user_id:15,user_meta:15,userdefinedvectorptr:50,userid:14,usernam:[2,3],useseqparallel_:41,usr:[1,3,7,81],usrdict:8,usrmodel:8,usual:[17,33,38,81,82],utc:64,utf:8,util:[0,1,9,15,17,21],val:[24,44],val_siz:28,valbuf:24,vale:38,valid:[11,17,25,40],validationimp:41,validationlay:40,valin:44,valu:[0,9,11,17,21,24,25,28,30,32,33,34,36,38,41,44,46,48,50,53,58,59,61,62,64,67,69,72,76,78,81,82,84,86],valuabl:58,value1:81,value2:81,value_:44,value_d:28,value_printer_evalu:36,value_typ:28,valuecount:50,valuedispatchthread:38,valuedispatchthread_:38,valuereadycond:50,valuereadycond_:38,valuereadyqueue_:38,values:50,valuesem_:50,valuestream_:38,valuetyp:[28,44],valuetype_:44,vanilla:0,vanish:17,vardensedata:34,variabl:[1,3,5,17,21,24,30,41,44,50,52,64,72,84],varianc:[11,25,41,72],varindic:34,vast:2,vctor:36,vec:[24,44,48,50,58],vec_check:44,vector:[0,8,13,15,17,19,21,24,25,28,30,34,36,38,41,44,46,48,50,52,53,56,58,59,62,72,74],vectoraddmult:53,vectoraddmultinto:53,vectorcopi:53,vectordotproduct:53,vectorenable_parallel_vector:80,vectorpriv:24,vectorptr:[38,41,44,48,50,58],vectors_:58,vectorscal:53,vectorscaleinto:53,vectort:44,veri:[0,9,13,17,24,36,44,58,72,78,84],verifi:[2,21],versa:[1,25],version:[1,3,5,8,9,14,15,17,19,21,24,25,41,44,50,58,72,74,80,81,86],vertic:[11,25,72],vgg:[9,74],vgg_16_cifar:9,via:[13,41,50],vice:[1,25],viet:41,view_data:24,view_m_data:24,virtual:[3,24,33,34,36,38,41,44,46,48,50,52,53,56,58,59,62],virtualenv:15,vision:9,visipedia:9,voc_dim:13,vocab:17,voluntarili:14,wai:[0,15,19,21,41,48,72,74,82],wait:[25,38,41,50,53,59,62,63,64,76,81],waitaftermerg:38,waitallmajorgradreadi:50,waitallthread:38,waitandmergeoutputgrad:41,waitbeforemerg:38,waitempti:63,waitforcopyinarg:38,waitforgradi:53,waitforstatu:53,waitgradreadi:50,waitinputvalu:41,waitoutargsreadi:38,waitpassfinish:[53,58],waitpassfinishrequest:58,waitpassstart:[53,58],waitpassstartrequest:58,waitresult:62,waitvaluereadi:50,wake:[38,63,64],walk:86,want:[1,3,4,8,11,13,15,17,21,34,38,72,74,81,82,84],wapper:64,war:14,warranti:[15,86],wbia:11,websit:[1,9,13,17],wei:17,weight:[0,9,11,21,30,32,36,38,41,44,49],weight_:[41,50],weight_act:74,weightdata:30,weightdim:[30,41,72],weightgrad:[30,44],weightgrad_:50,weightlast:41,weightlast_:41,weightlayer_:41,weightlist:[21,41,50],weightoffset_:41,weights_:[21,41],weights_t:21,welcom:[3,15,17],well:[3,7,9,13,21,81],were:[41,64],western:14,wether:[50,72],what:[7,13,15,34,62,67,74,76],wheel:1,when:[0,5,8,9,17,19,21,34,36,38,41,44,48,50,53,58,59,62,67,69,72,81,82,83,84],whenev:15,where:[0,8,11,19,21,33,38,41,44,72,74,76,81,82,84],whether:[9,15,17,19,21,38,41,44,50,69,72,74,81,84],which:[0,3,7,9,11,13,14,15,17,19,21,24,25,28,32,34,38,41,44,50,56,58,59,60,64,65,69,72,74,76,81,82,83,84],whl:1,who:[8,11,14,63],whoe:[28,44],whole:[13,14,15,19,34,36,38,69,84],whole_cont:15,whose:[0,15,38,58,62,84],wide:62,width:[9,19,21,24,25,28,30,41,44,50,58,69,72],width_:44,width_pad:25,widthpad:41,wiil:24,wilder:84,window:[17,25,30,41],windowheight:41,windows:41,windowwidth:41,wise:[41,44,72],witdth:41,with_doubl:21,withcurrentcost:59,within:[38,41,53,58,59,72],without:[15,17,36,38,41,58,69,86],wmt14:19,wmt14_data:19,wmt14_model:19,wmt:19,woker:56,won:[11,63,64],wonder:84,word:[0,7],word_dict:13,word_dim:13,word_id:84,word_vector:13,word_vector_dim:[0,8],work:[0,1,2,13,15,21,25,38,41,48,59,81,84],worker:[25,44,48,55],workernum:62,workers_:62,workflow:[2,18],workspac:6,workspace_:41,workspacedata_:41,workspaceinbytes_:41,workthread:38,would:[3,13,38],wrap:[24,58],wrapper:20,writ:15,write:[0,7,9],writemessag:[56,58],writer:[14,64],writev:56,written:[15,17,56],wstride:25,www:[9,15,19,86],x64:1,x86:5,xarg:[3,21],xgbe0:81,xgbe1:81,xi_t:41,xiaojun:17,xr_t:41,xrang:21,xxbow:17,xxx:[11,19],xz_t:41,xzf:1,yaml:15,yanfei:53,year:14,yeild:9,yet:25,yield:[0,13,15,17,84],you:[0,1,2,3,4,5,7,8,9,11,13,15,17,19,21,24,34,41,44,50,58,63,67,72,74,76,81,82,83,84,86],your:[2,3,4,7,13,17,21,82,84],your_host_machin:3,yourself:[3,7],yuyang18:[72,74],zachari:17,zbib:41,zeng:17,zere:81,zero:[13,21,24,28,38,41,44,53,58,59,63,67,72,76,81,84],zeroatoffset:44,zerograd:41,zeromem:[44,50],zeromemthread:44,zhidao:8,zhongqiang:41,zhou:17,zip:14},titles:["Recurrent Neural Network Configuration","Build and Install","Contribute to PaddlePaddle","Docker installation guide","Build And Install PaddlePaddle","Debian Package installation guide","Cluster Train","Cluster Training","Chinese Word Embedding Model Tutorial","Image Classification Tutorial","Image Classification Tutorial","Model Zoo - ImageNet","Examples and demos","Quick Start Tutorial","MovieLens Dataset","Regression MovieLens Ratting","Sentiment Analasis Tutorial","Sentiment Analysis Tutorial","Text Generation Tutorial","Text generation Tutorial","Writing New Layers","Writing New Layers","PaddlePaddle Documentation","Layer Documents","API","Cuda","CUDA","Matrix","Matrix","RNN","Neural Networks","Utils","Utilities","Activations","Data Providers","Data Providers Documents","Base Evaluator","Evaluators","Gradient Machines","Gradient Machines Documents","Layers Documents","Base","Source Code Documents","Matrix Documents","Matrix","Utils Documents","Utils","Parameter Documents","Optimizer","Parameter Documents","Parameter","Parameter Documents","Update","Client","Client Documents","Network Documents","Network","Server Documents","Server","Trainer","CustomStackTrace","enumeration_wrapper","Thread","Queue","Lock","BaseActivation","Activations","Parameter and Extra Layer Attribute","DataSources","Base","Evaluators","Model Config Interface","Base","Layers","NLP","Networks","BaseSGDOptimizer","Optimizers","BasePoolingType","Poolings","Argument Outline","Detail Description","Use Case","PaddlePaddle DataProvider Introduction","How to use PyDataProvider2","User Interface","Python Prediction API"],titleterms:{"case":82,"class":[21,60,62,63,64],"function":[8,30],"new":[20,21],absactiv:65,access:3,activ:[30,33,41,66],adadeltaoptim:76,adagradoptim:76,adamaxoptim:76,adamoptim:76,addto_lay:72,addtolay:41,agentlay:41,aggreg:[41,72],algorithm:[13,22],analasi:16,analysi:17,api:[24,30,42,85,86],appendix:13,architectur:[0,13],argument:[13,80,82,85],async:81,asyncthreadpool:62,attent:0,attribut:67,auc_evalu:69,aucevalu:36,aucvalid:41,averagelay:41,avgpool:78,base:[28,30,32,34,36,41,44,69,72],baseactiv:65,basepoolingtyp:78,basesgdoptim:76,batch_norm_lay:72,batchnormalizationlay:41,batchnormbaselay:41,beam_search:72,bidirect:17,bidirectional_lstm:74,bit:46,bleu:19,block_expand_lay:72,blockexpandlay:41,blockingqueu:63,breluactiv:65,build:[1,4],cach:84,cacheonepassinmemori:34,channel:56,check:[7,21,41,72],chines:8,chunk_evalu:69,chunkevalu:36,classif:[9,10,36,69],classification_error_evalu:69,classification_error_printer_evalu:69,classificationerrorevalu:36,classificationerrorprint:36,client:[53,54,56],clone:2,cluster:[6,7,82],cmrprojectionnormlay:41,code:[2,42],column_sum_evalu:69,columnsumevalu:36,command:[13,19,82,85],commit:2,common:81,commun:81,concat:41,concat_lay:72,concatenatelay:41,concatenatelayer2:41,concurrentremoteparameterupdat:59,config:[15,71,82],configur:[0,7,13,15],connect:[41,72],context_project:72,contextproject:41,contribut:2,conv:[41,72],conv_oper:72,conv_shift_lay:72,convbaselay:41,convex_comb_lay:72,convexcombinationlay:41,convolut:[9,13],convoper:41,convshiftlay:41,coorditer:41,cos_sim:72,cossimlay:41,cossimvecmatlay:41,cost:[41,72],costlay:41,cpu:82,creat:2,crf_decoding_lay:72,crf_layer:72,crfdecodinglay:41,crflayer:41,cross_entropi:72,cross_entropy_with_selfnorm:72,ctc_error_evalu:69,ctc_layer:72,ctcevalu:36,ctclayer:41,cuda:[25,26,42],cudnnbatchnormlay:41,cudnnconvlay:41,cudnnpoollay:41,cuh:28,customstacktrac:60,dat:14,data:[0,8,9,13,15,17,19,34,35,41,72,85],data_lay:72,datanormlay:41,dataprovid:[34,81,83,84],dataprovidergroup:34,dataset:[14,15,19],datasourc:68,date:2,debian:[4,5],decayedadagradoptim:76,defin:[13,17,19],delv:9,demo:12,densescann:34,depend:1,deriv:21,descript:[14,81],detail:[9,81],develop:22,devic:82,dictionari:8,differ:82,distribut:81,docker:[3,4],document:[22,23,35,39,40,42,43,45,47,49,51,54,55,57],dotmul_project:72,dotmuloper:41,dotmulproject:41,download:[3,8,11,15,19],dropout_lay:74,dynam:25,embed:[8,13],embedding_lay:72,enumeration_wrapp:61,eos_lay:72,eosidchecklay:41,equat:21,evalu:[15,36,37,70],evalutaion:19,exampl:[8,12],exercis:9,expand_lay:72,expandconvlay:41,expandlay:41,extra:67,extract:[8,11,15,19],fc_layer:72,featur:[11,14,15],featuremapexpandlay:41,field:15,file:[13,14,15],first_seq:72,fork:2,format:13,from:4,full_matrix_project:72,fulli:[41,72],fullmatrixproject:41,fullyconnectedlay:41,gate:0,gatedrecurrentlay:41,gatheragentlay:41,gener:[0,18,19],get_output_lay:72,getoutputlay:41,github:2,gpu:[25,81,82],gradient:[21,38,39],gradient_printer_evalu:69,gradientmachin:38,gradientmachinemodel:38,gradientprint:36,group:[41,72],gru:[30,41,74,81],gru_group:74,gru_step_lay:72,gru_unit:74,grucomput:41,grumemori:72,grusteplay:41,gserver:42,guid:[3,5,22],handl:46,hierarchicalsigmoidlay:41,hl_aggreg:28,hl_base:32,hl_batch_transpos:28,hl_cuda:25,hl_cuda_cubla:25,hl_cuda_cudnn:25,hl_dso_load:25,hl_matrix:28,hl_matrix_appli:28,hl_matrix_bas:28,hl_matrix_op:28,hl_matrix_typ:28,hl_spars:28,hl_sse_matrix_kernel:28,hl_table_appli:28,hl_thread:32,hl_time:32,hl_top_k:28,how:84,hppl:32,hsigmoid:72,huber_cost:72,hubertwoclass:41,identity_project:72,identityactiv:65,identityoffsetproject:41,identityproject:41,ifieldscann:34,imag:[3,9,10,12,72,74],imagenet:11,imdb:17,img_cmrnorm_lay:72,img_conv_bn_pool:74,img_conv_group:74,img_conv_lay:72,img_pool_lay:72,img_rnorm_lay:72,implement:21,indexscann:34,infer:13,info:11,init_hook:84,initi:82,input_typ:84,instal:[1,3,4,5,13],interfac:[11,71,85],interpolation_lay:72,interpolationlay:41,introduct:[8,11,17,19,83],ipydataprovid:34,ipydataprovidercach:34,job:7,keep:2,kill:7,lambda_cost:72,lambdacost:41,last_seq:72,launch:7,layer:[20,21,23,40,41,67,72,73,82],layeroutput:72,layertyp:72,learn:81,lib:25,line:[13,85],linearactiv:65,linearchaincrf:41,linearchainctc:41,link:25,local:82,lock:64,lockedcondit:64,log:13,logist:13,lstm:[17,30,41,74,81],lstm_step_lay:72,lstmcomput:41,lstmemori:72,lstmemory_group:74,lstmemory_unit:74,lstmlayer:41,lstmsteplay:41,machin:[38,39],math:[41,42,72],matrix:[27,28,43,44,81],maxframe_printer_evalu:69,maxframeprint:36,maxid_lay:72,maxid_printer_evalu:69,maxidlay:41,maxidprint:36,maxlay:41,maxpool:78,mdlstm:41,mdlstmlayer:41,memori:46,messag:56,meta:15,metric:81,misc:74,mix:[41,72,82],mixed_lay:72,model:[0,7,8,9,11,12,13,19,30,71,82,84],movi:[14,15],movielen:[14,15],multi_binary_label_cross_entropi:72,multibinarylabelcrossentropi:41,multiclasscrossentropi:41,multiclasscrossentropywithselfnorm:41,multidataprovid:34,multigradientmachin:38,multinomialsampl:41,multiplexlay:41,multithreadwork:62,namespac:61,ncelay:41,network:[0,9,11,13,15,30,38,55,56,75,82],neural:[0,9,13,15,30],neuralnetwork:38,nlp:[12,74,81],nocachestrategi:34,non:84,norm:[41,72],normlay:41,notic:3,object:15,observ:[8,11],oper:41,optim:[13,48,77],option:[1,8],other:[28,44],outerprodlay:41,outlin:80,output:[7,74],overview:13,packag:[4,5],paddl:61,paddlepaddl:[2,3,4,8,19,22,83],parallel_nn:82,parallelneuralnetwork:38,paramet:[8,11,42,47,49,50,51,67,81],parameterrelulay:41,paraphras:8,pass:82,perform:[3,81],pnpair_evalu:69,pnpairevalu:36,pnpairvalid:41,pool:[41,72,79],pooling_lay:72,poollay:41,poolprojectionlay:41,power_lay:72,powerlay:41,pre:7,precision_recall_evalu:69,precisionrecallevalu:36,predict:[9,11,15,17,85,86],prepar:[0,7,8,9,15,17,19],preprocess:[8,9,13,15,19],pretrain:[8,19],print:69,printer:36,project:41,proto:34,protodataprovid:34,protosequencedataprovid:34,provid:[13,15,34,35,84,85],pserver:42,pull:2,push:2,pydataprovid:34,pydataprovider2:84,python:[11,13,15,21,86],queue:63,quick:13,randomnumb:81,rank:[36,69],rank_cost:72,rankingcost:41,rat:15,rate:14,reader:56,readlockguard:64,recommend:12,recurr:[0,13,38,41,72,74],recurrent_group:72,recurrent_lay:72,recurrentlay:41,refer:[17,84,85],regress:[13,15],regular:50,relat:30,reluactiv:65,remot:3,remoteparameterupdat:59,request:2,requir:[1,2,7],reshap:[41,72],resizelay:41,resnet:11,resourc:[25,32],responsenormlay:41,result:[7,19],revis:8,rmspropoptim:76,rnn:[29,30,81],run:3,rwlock:64,sampl:[41,72],sampling_id_lay:72,samplingidlay:41,scaling_lay:72,scalinglay:41,scatteragentlay:41,selective_fc_lay:72,selectivefullyconnectedlay:41,semaphor:64,sentiment:[16,17],seqtext_printer_evalu:69,sequenc:0,sequence_conv_pool:74,sequenceagentlay:41,sequenceclassificationerrorevalu:36,sequenceconcatlay:41,sequencegatheragentlay:41,sequencelastinstancelay:41,sequencereshapelay:41,sequencescann:34,sequencescatteragentlay:41,sequencesoftmaxactiv:65,sequencetextprint:36,sequencetobatch:41,sequenti:84,server:[56,57,58,81],set:76,sgd:81,sigmoidactiv:65,simpl:0,simple_attent:74,simple_gru:74,simple_img_conv_pool:74,simple_lstm:74,slope_intercept_lay:72,slopeinterceptlay:41,socket:56,softbinaryclasscrossentropi:41,softmaxactiv:65,softreluactiv:65,sourc:[4,42],spars:[28,44,82],sparsenonvaluescann:34,sparseremoteparameterupdat:59,sparseremoteparameterupdatercomposit:59,sparsevaluescann:34,specifi:[8,82],spinlock:64,split:15,squareactiv:65,squarerootnpool:78,stack:17,standard:13,stanhactiv:65,start:13,subsequencelay:41,subset:41,sum_evalu:69,sum_to_one_norm_lay:72,sumevalu:36,summari:13,sumofsquarescostlay:41,sumpool:78,sumtoonenormlay:41,syncthreadpool:62,table_project:72,tableproject:41,tanhactiv:65,tensor_lay:72,tensorlay:41,test:[15,17,21,81,82],text:[18,19],text_conv_pool:74,thread:[32,62],threadbarri:64,threadwork:62,timer:32,train:[6,7,8,9,13,15,17,19,81,82],trainer:[15,42,59],trainerstat:59,trainerthread:38,trans_full_matrix_project:72,trans_lay:72,transfer:13,translay:41,transposedfullmatrixproject:41,tune:81,tutori:[8,9,10,13,16,17,18,19,22],ubuntu14:1,unit:[21,81],updat:52,user:[8,14,15,17,19,22,85],util:[31,32,36,42,45,46,69],valid:41,validationlay:41,value_printer_evalu:69,valueprint:36,vector:81,vgg_16_network:74,visual:11,weight:50,word:[8,13],worker:56,workflow:19,workspac:7,wrapper:[21,25],write:[13,20,21],zoo:[11,12]}})
\ No newline at end of file
+Search.setIndex({envversion:49,filenames:["algorithm/rnn/rnn","build/build_from_source","build/contribute_to_paddle","build/docker_install","build/index","build/ubuntu_install","cluster/index","cluster/opensource/cluster_train","demo/embedding_model/index","demo/image_classification/image_classification","demo/image_classification/index","demo/imagenet_model/resnet_model","demo/index","demo/quick_start/index_en","demo/rec/ml_dataset","demo/rec/ml_regression","demo/semantic_role_labeling/index","demo/semantic_role_labeling/semantic_role_labeling","demo/sentiment_analysis/index","demo/sentiment_analysis/sentiment_analysis","demo/text_generation/index","demo/text_generation/text_generation","dev/new_layer/index","dev/new_layer/new_layer","index","layer","source/api/api","source/cuda/cuda/cuda","source/cuda/cuda/index","source/cuda/matrix/index","source/cuda/matrix/matrix","source/cuda/rnn/index","source/cuda/rnn/rnn","source/cuda/utils/index","source/cuda/utils/utils","source/gserver/activations/index","source/gserver/dataprovider/dataproviders","source/gserver/dataprovider/index","source/gserver/evaluators/evaluators","source/gserver/evaluators/index","source/gserver/gradientmachines/gradientmachines","source/gserver/gradientmachines/index","source/gserver/layers/index","source/gserver/layers/layer","source/index","source/math/matrix/index","source/math/matrix/matrix","source/math/utils/index","source/math/utils/utils","source/parameter/optimizer/index","source/parameter/optimizer/optimizer","source/parameter/parameter/index","source/parameter/parameter/parameter","source/parameter/update/index","source/parameter/update/update","source/pserver/client/client","source/pserver/client/index","source/pserver/network/index","source/pserver/network/network","source/pserver/server/index","source/pserver/server/server","source/trainer/trainer","source/utils/customStackTrace","source/utils/enum","source/utils/lock","source/utils/queue","source/utils/thread","ui/api/trainer_config_helpers/activations","ui/api/trainer_config_helpers/activations_index","ui/api/trainer_config_helpers/attrs","ui/api/trainer_config_helpers/data_sources","ui/api/trainer_config_helpers/evaluators","ui/api/trainer_config_helpers/evaluators_index","ui/api/trainer_config_helpers/index","ui/api/trainer_config_helpers/layers","ui/api/trainer_config_helpers/layers_index","ui/api/trainer_config_helpers/networks","ui/api/trainer_config_helpers/networks_index","ui/api/trainer_config_helpers/optimizers","ui/api/trainer_config_helpers/optimizers_index","ui/api/trainer_config_helpers/poolings","ui/api/trainer_config_helpers/poolings_index","ui/cmd_argument/argument_outline","ui/cmd_argument/detail_introduction","ui/cmd_argument/use_case","ui/data_provider/index","ui/data_provider/pydataprovider2","ui/index","ui/predict/swig_py_paddle_en"],objects:{"":{"Arguments::Arguments":[26,2,1,"_CPPv2N9Arguments9ArgumentsEv"],"Arguments::DISABLE_COPY_AND_ASSIGN":[26,2,1,"_CPPv2N9Arguments23DISABLE_COPY_AND_ASSIGNE9Arguments"],"Arguments::createArguments":[26,2,1,"_CPPv2N9Arguments15createArgumentsE6size_t"],"Arguments::createByPaddleArgumentVector":[26,2,1,"_CPPv2N9Arguments28createByPaddleArgumentVectorEPv"],"Arguments::getBatchSize":[26,2,1,"_CPPv2N9Arguments12getBatchSizeE6size_t"],"Arguments::getInternalArgumentsPtr":[26,2,1,"_CPPv2N9Arguments23getInternalArgumentsPtrEv"],"Arguments::getSlotIds":[26,2,1,"_CPPv2N9Arguments10getSlotIdsE6size_t"],"Arguments::getSlotIn":[26,2,1,"_CPPv2N9Arguments9getSlotInE6size_t"],"Arguments::getSlotNum":[26,2,1,"_CPPv2N9Arguments10getSlotNumEv"],"Arguments::getSlotSequenceDim":[26,2,1,"_CPPv2N9Arguments18getSlotSequenceDimE6size_t"],"Arguments::getSlotSequenceStartPositions":[26,2,1,"_CPPv2N9Arguments29getSlotSequenceStartPositionsE6size_t"],"Arguments::getSlotSubSequenceStartPositions":[26,2,1,"_CPPv2N9Arguments32getSlotSubSequenceStartPositionsE6size_t"],"Arguments::getSlotValue":[26,2,1,"_CPPv2N9Arguments12getSlotValueE6size_t"],"Arguments::m":[26,3,1,"_CPPv2N9Arguments1mE"],"Arguments::resize":[26,2,1,"_CPPv2N9Arguments6resizeE6size_t"],"Arguments::setSlotIds":[26,2,1,"_CPPv2N9Arguments10setSlotIdsE6size_tP7IVector"],"Arguments::setSlotIn":[26,2,1,"_CPPv2N9Arguments9setSlotInE6size_tP6Matrix"],"Arguments::setSlotSequenceDim":[26,2,1,"_CPPv2N9Arguments18setSlotSequenceDimE6size_tP7IVector"],"Arguments::setSlotSequenceStartPositions":[26,2,1,"_CPPv2N9Arguments29setSlotSequenceStartPositionsE6size_tP7IVector"],"Arguments::setSlotSubSequenceStartPositions":[26,2,1,"_CPPv2N9Arguments32setSlotSubSequenceStartPositionsE6size_tP7IVector"],"Arguments::setSlotValue":[26,2,1,"_CPPv2N9Arguments12setSlotValueE6size_tP6Matrix"],"Arguments::~Arguments":[26,2,1,"_CPPv2N9ArgumentsD0Ev"],"BaseOp::BaseOp":[30,2,1,"_CPPv2N6BaseOp6BaseOpEv"],"BaseOp::sse":[30,3,1,"_CPPv2N6BaseOp3sseE"],"FloatArray::FloatArray":[26,2,1,"_CPPv2N10FloatArray10FloatArrayEPKfK6size_t"],"FloatArray::buf":[26,3,1,"_CPPv2N10FloatArray3bufE"],"FloatArray::length":[26,3,1,"_CPPv2N10FloatArray6lengthE"],"FloatArray::needFree":[26,3,1,"_CPPv2N10FloatArray8needFreeE"],"GradientMachine::DISABLE_COPY_AND_ASSIGN":[26,2,1,"_CPPv2N15GradientMachine23DISABLE_COPY_AND_ASSIGNE15GradientMachine"],"GradientMachine::GradientMachine":[26,2,1,"_CPPv2N15GradientMachine15GradientMachineEv"],"GradientMachine::asSequenceGenerator":[26,2,1,"_CPPv2N15GradientMachine19asSequenceGeneratorERKNSt6vectorINSt6stringEEE6size_t6size_t6size_t6size_t"],"GradientMachine::backward":[26,2,1,"_CPPv2N15GradientMachine8backwardERK14UpdateCallback"],"GradientMachine::create":[40,2,1,"_CPPv2N15GradientMachine6createERNSt7istreamEP13TrainerConfig"],"GradientMachine::createByConfigProtoStr":[26,2,1,"_CPPv2N15GradientMachine22createByConfigProtoStrERKNSt6stringE26GradientMatchineCreateModeRKNSt6vectorIiEE"],"GradientMachine::createByModelConfig":[26,2,1,"_CPPv2N15GradientMachine19createByModelConfigEP11ModelConfig26GradientMatchineCreateModeRKNSt6vectorIiEE"],"GradientMachine::createFromPaddleModelPtr":[26,2,1,"_CPPv2N15GradientMachine24createFromPaddleModelPtrEPv26GradientMatchineCreateModeRKNSt6vectorIiEE"],"GradientMachine::defaultParamTypes":[26,3,1,"_CPPv2N15GradientMachine17defaultParamTypesE"],"GradientMachine::forward":[26,2,1,"_CPPv2N15GradientMachine7forwardERK9ArgumentsP9Arguments8PassType"],"GradientMachine::forwardBackward":[26,2,1,"_CPPv2N15GradientMachine15forwardBackwardERK9ArgumentsP9Arguments8PassTypeRK14UpdateCallback"],"GradientMachine::getLayerOutput":[26,2,1,"_CPPv2N15GradientMachine14getLayerOutputERKNSt6stringE"],"GradientMachine::getParameter":[26,2,1,"_CPPv2N15GradientMachine12getParameterE6size_t"],"GradientMachine::getParameterSize":[26,2,1,"_CPPv2N15GradientMachine16getParameterSizeEv"],"GradientMachine::loadParameters":[40,2,1,"_CPPv2N15GradientMachine14loadParametersERKNSt6stringE"],"GradientMachine::m":[26,3,1,"_CPPv2N15GradientMachine1mE"],"GradientMachine::randParameters":[40,2,1,"_CPPv2N15GradientMachine14randParametersEv"],"GradientMachine::saveParameters":[40,2,1,"_CPPv2N15GradientMachine14saveParametersERKNSt6stringE"],"GradientMachine::~GradientMachine":[26,2,1,"_CPPv2N15GradientMachineD0Ev"],"ISequenceResults::getScore":[26,2,1,"_CPPv2N16ISequenceResults8getScoreE6size_t"],"ISequenceResults::getSentence":[26,2,1,"_CPPv2N16ISequenceResults11getSentenceE6size_tb"],"ISequenceResults::getSequence":[26,2,1,"_CPPv2N16ISequenceResults11getSequenceE6size_t"],"ISequenceResults::getSize":[26,2,1,"_CPPv2N16ISequenceResults7getSizeEv"],"ISequenceResults::~ISequenceResults":[26,2,1,"_CPPv2N16ISequenceResultsD0Ev"],"IVector::DISABLE_COPY_AND_ASSIGN":[26,2,1,"_CPPv2N7IVector23DISABLE_COPY_AND_ASSIGNE7IVector"],"IVector::IVector":[26,2,1,"_CPPv2N7IVector7IVectorEv"],"IVector::copyFromNumpyArray":[26,2,1,"_CPPv2N7IVector18copyFromNumpyArrayEPii"],"IVector::copyToNumpyArray":[26,2,1,"_CPPv2N7IVector16copyToNumpyArrayEPPiPi"],"IVector::create":[26,2,1,"_CPPv2N7IVector6createERKNSt6vectorIiEEb"],"IVector::createByPaddleVectorPtr":[26,2,1,"_CPPv2N7IVector23createByPaddleVectorPtrEPv"],"IVector::createCpuVectorFromNumpy":[26,2,1,"_CPPv2N7IVector24createCpuVectorFromNumpyEPiib"],"IVector::createGpuVectorFromNumy":[26,2,1,"_CPPv2N7IVector23createGpuVectorFromNumyEPii"],"IVector::createZero":[26,2,1,"_CPPv2N7IVector10createZeroE6size_tb"],"IVector::get":[26,2,1,"_CPPv2N7IVector3getEK6size_t"],"IVector::getData":[26,2,1,"_CPPv2N7IVector7getDataEv"],"IVector::getSharedPtr":[26,2,1,"_CPPv2N7IVector12getSharedPtrEv"],"IVector::getSize":[26,2,1,"_CPPv2N7IVector7getSizeEv"],"IVector::isGpu":[26,2,1,"_CPPv2N7IVector5isGpuEv"],"IVector::m":[26,3,1,"_CPPv2N7IVector1mE"],"IVector::operator[]":[26,2,1,"_CPPv2N7IVectorixEK6size_t"],"IVector::set":[26,2,1,"_CPPv2N7IVector3setEK6size_ti"],"IVector::toNumpyArrayInplace":[26,2,1,"_CPPv2N7IVector19toNumpyArrayInplaceEPPiPi"],"IVector::~IVector":[26,2,1,"_CPPv2N7IVectorD0Ev"],"IntArray::IntArray":[26,2,1,"_CPPv2N8IntArray8IntArrayEPKiK6size_tb"],"IntArray::buf":[26,3,1,"_CPPv2N8IntArray3bufE"],"IntArray::length":[26,3,1,"_CPPv2N8IntArray6lengthE"],"IntArray::needFree":[26,3,1,"_CPPv2N8IntArray8needFreeE"],"IntWithFloatArray::IntWithFloatArray":[26,2,1,"_CPPv2N17IntWithFloatArray17IntWithFloatArrayEPKfPKi6size_tb"],"IntWithFloatArray::idxBuf":[26,3,1,"_CPPv2N17IntWithFloatArray6idxBufE"],"IntWithFloatArray::length":[26,3,1,"_CPPv2N17IntWithFloatArray6lengthE"],"IntWithFloatArray::needFree":[26,3,1,"_CPPv2N17IntWithFloatArray8needFreeE"],"IntWithFloatArray::valBuf":[26,3,1,"_CPPv2N17IntWithFloatArray6valBufE"],"Matrix::DISABLE_COPY_AND_ASSIGN":[26,2,1,"_CPPv2N6Matrix23DISABLE_COPY_AND_ASSIGNE6Matrix"],"Matrix::Matrix":[46,2,1,"_CPPv2N6Matrix6MatrixEP4real6size_t6size_tbb"],"Matrix::copyFromNumpyMat":[26,2,1,"_CPPv2N6Matrix16copyFromNumpyMatEPfii"],"Matrix::copyToNumpyMat":[26,2,1,"_CPPv2N6Matrix14copyToNumpyMatEPPfPiPi"],"Matrix::create":[46,2,1,"_CPPv2N6Matrix6createEP4real6size_t6size_tbb"],"Matrix::createByPaddleMatrixPtr":[26,2,1,"_CPPv2N6Matrix23createByPaddleMatrixPtrEPv"],"Matrix::createCpuDenseFromNumpy":[26,2,1,"_CPPv2N6Matrix23createCpuDenseFromNumpyEPfiib"],"Matrix::createDense":[26,2,1,"_CPPv2N6Matrix11createDenseERKNSt6vectorIfEE6size_t6size_tb"],"Matrix::createGpuDenseFromNumpy":[26,2,1,"_CPPv2N6Matrix23createGpuDenseFromNumpyEPfii"],"Matrix::createSparse":[26,2,1,"_CPPv2N6Matrix12createSparseE6size_t6size_t6size_tbbb"],"Matrix::createSparseMatrix":[46,2,1,"_CPPv2N6Matrix18createSparseMatrixEP4realPiPi6size_t6size_t6size_t15SparseValueType12SparseFormatbb"],"Matrix::createZero":[26,2,1,"_CPPv2N6Matrix10createZeroE6size_t6size_tb"],"Matrix::get":[26,2,1,"_CPPv2N6Matrix3getE6size_t6size_t"],"Matrix::getData":[26,2,1,"_CPPv2N6Matrix7getDataEv"],"Matrix::getHeight":[26,2,1,"_CPPv2N6Matrix9getHeightEv"],"Matrix::getSharedPtr":[26,2,1,"_CPPv2N6Matrix12getSharedPtrEv"],"Matrix::getSparseFormat":[26,2,1,"_CPPv2N6Matrix15getSparseFormatEv"],"Matrix::getSparseRowCols":[26,2,1,"_CPPv2N6Matrix16getSparseRowColsE6size_t"],"Matrix::getSparseRowColsVal":[26,2,1,"_CPPv2N6Matrix19getSparseRowColsValE6size_t"],"Matrix::getSparseValueType":[26,2,1,"_CPPv2N6Matrix18getSparseValueTypeEv"],"Matrix::getWidth":[26,2,1,"_CPPv2N6Matrix8getWidthEv"],"Matrix::isGpu":[26,2,1,"_CPPv2N6Matrix5isGpuEv"],"Matrix::isSparse":[26,2,1,"_CPPv2N6Matrix8isSparseEv"],"Matrix::m":[26,3,1,"_CPPv2N6Matrix1mE"],"Matrix::reshape":[46,2,1,"_CPPv2N6Matrix7reshapeE6size_t6size_t"],"Matrix::resizeOrCreate":[46,2,1,"_CPPv2N6Matrix14resizeOrCreateER9MatrixPtr6size_t6size_tbb"],"Matrix::resizeOrCreateSparseMatrix":[46,2,1,"_CPPv2N6Matrix26resizeOrCreateSparseMatrixER9MatrixPtr6size_t6size_t6size_t15SparseValueType12SparseFormatbb"],"Matrix::set":[26,2,1,"_CPPv2N6Matrix3setE6size_t6size_tf"],"Matrix::sparseCopyFrom":[26,2,1,"_CPPv2N6Matrix14sparseCopyFromERKNSt6vectorIiEERKNSt6vectorIiEERKNSt6vectorIfEE"],"Matrix::subMatrix":[46,2,1,"_CPPv2N6Matrix9subMatrixE6size_t6size_t6size_t6size_t"],"Matrix::toNumpyMatInplace":[26,2,1,"_CPPv2N6Matrix17toNumpyMatInplaceEPPfPiPi"],"Matrix::~Matrix":[26,2,1,"_CPPv2N6MatrixD0Ev"],"ModelConfig::DISABLE_COPY_AND_ASSIGN":[26,2,1,"_CPPv2N11ModelConfig23DISABLE_COPY_AND_ASSIGNE11ModelConfig"],"ModelConfig::ModelConfig":[26,2,1,"_CPPv2N11ModelConfig11ModelConfigEv"],"ModelConfig::getPaddleModelConfig":[26,2,1,"_CPPv2N11ModelConfig20getPaddleModelConfigEv"],"ModelConfig::m":[26,3,1,"_CPPv2N11ModelConfig1mE"],"ModelConfig::~ModelConfig":[26,2,1,"_CPPv2N11ModelConfigD0Ev"],"OptimizationConfig::DISABLE_COPY_AND_ASSIGN":[26,2,1,"_CPPv2N18OptimizationConfig23DISABLE_COPY_AND_ASSIGNE18OptimizationConfig"],"OptimizationConfig::OptimizationConfig":[26,2,1,"_CPPv2N18OptimizationConfig18OptimizationConfigEv"],"OptimizationConfig::createFromProtoString":[26,2,1,"_CPPv2N18OptimizationConfig21createFromProtoStringERKNSt6stringE"],"OptimizationConfig::getRawPtr":[26,2,1,"_CPPv2N18OptimizationConfig9getRawPtrEv"],"OptimizationConfig::m":[26,3,1,"_CPPv2N18OptimizationConfig1mE"],"OptimizationConfig::toProtoString":[26,2,1,"_CPPv2N18OptimizationConfig13toProtoStringEv"],"OptimizationConfig::~OptimizationConfig":[26,2,1,"_CPPv2N18OptimizationConfigD0Ev"],"Parameter::DISABLE_COPY_AND_ASSIGN":[26,2,1,"_CPPv2N9Parameter23DISABLE_COPY_AND_ASSIGNE9Parameter"],"Parameter::Parameter":[52,2,1,"_CPPv2N9Parameter9ParameterERK15ParameterConfigbb"],"Parameter::createFromRawPtr":[26,2,1,"_CPPv2N9Parameter16createFromRawPtrEPv"],"Parameter::createFromSharedPtr":[26,2,1,"_CPPv2N9Parameter19createFromSharedPtrEPv"],"Parameter::exec":[52,2,1,"_CPPv2N9Parameter4execE8ExecFunc"],"Parameter::getBuf":[26,2,1,"_CPPv2N9Parameter6getBufE13ParameterType"],"Parameter::getConfig":[26,2,1,"_CPPv2N9Parameter9getConfigEv"],"Parameter::getID":[26,2,1,"_CPPv2N9Parameter5getIDEv"],"Parameter::getName":[26,2,1,"_CPPv2N9Parameter7getNameEv"],"Parameter::getPrefetchMatrix":[52,2,1,"_CPPv2N9Parameter17getPrefetchMatrixEv"],"Parameter::getTlsTempBufs":[52,2,1,"_CPPv2N9Parameter14getTlsTempBufsEv"],"Parameter::incUpdate":[52,2,1,"_CPPv2N9Parameter9incUpdateERK14UpdateCallback"],"Parameter::initialize":[52,2,1,"_CPPv2N9Parameter10initializeEv"],"Parameter::isGradShared":[52,2,1,"_CPPv2N9Parameter12isGradSharedEP6size_t"],"Parameter::isGradSparseUpdate":[52,2,1,"_CPPv2N9Parameter18isGradSparseUpdateEv"],"Parameter::isValueShared":[52,2,1,"_CPPv2N9Parameter13isValueSharedEv"],"Parameter::kMissParameterFail":[52,3,1,"_CPPv2N9Parameter18kMissParameterFailE"],"Parameter::kMissParameterRand":[52,3,1,"_CPPv2N9Parameter18kMissParameterRandE"],"Parameter::kMissParameterZero":[52,3,1,"_CPPv2N9Parameter18kMissParameterZeroE"],"Parameter::load":[52,2,1,"_CPPv2N9Parameter4loadERNSt7istreamE"],"Parameter::m":[26,3,1,"_CPPv2N9Parameter1mE"],"Parameter::randomize":[52,2,1,"_CPPv2N9Parameter9randomizeEv"],"Parameter::save":[52,2,1,"_CPPv2N9Parameter4saveERNSt7ostreamE"],"Parameter::setMat":[52,2,1,"_CPPv2N9Parameter6setMatE13ParameterTypei"],"Parameter::tlsTempBufs_":[52,3,1,"_CPPv2N9Parameter12tlsTempBufs_E"],"Parameter::updateWithGradient":[52,2,1,"_CPPv2N9Parameter18updateWithGradientE4real9VectorPtrb"],"Parameter::zeroMem":[52,2,1,"_CPPv2N9Parameter7zeroMemEv"],"Parameter::~Parameter":[26,2,1,"_CPPv2N9ParameterD0Ev"],"ParameterConfig::DISABLE_COPY_AND_ASSIGN":[26,2,1,"_CPPv2N15ParameterConfig23DISABLE_COPY_AND_ASSIGNE15ParameterConfig"],"ParameterConfig::ParameterConfig":[26,2,1,"_CPPv2N15ParameterConfig15ParameterConfigEv"],"ParameterConfig::createParameterConfigFromParameterPtr":[26,2,1,"_CPPv2N15ParameterConfig37createParameterConfigFromParameterPtrEPv"],"ParameterConfig::createParameterConfigFromParameterSharedPtr":[26,2,1,"_CPPv2N15ParameterConfig43createParameterConfigFromParameterSharedPtrEPv"],"ParameterConfig::getRawPtr":[26,2,1,"_CPPv2N15ParameterConfig9getRawPtrEv"],"ParameterConfig::m":[26,3,1,"_CPPv2N15ParameterConfig1mE"],"ParameterConfig::toProtoString":[26,2,1,"_CPPv2N15ParameterConfig13toProtoStringEv"],"ParameterConfig::~ParameterConfig":[26,2,1,"_CPPv2N15ParameterConfigD0Ev"],"ParameterOptimizer::DISABLE_COPY_AND_ASSIGN":[26,2,1,"_CPPv2N18ParameterOptimizer23DISABLE_COPY_AND_ASSIGNE18ParameterOptimizer"],"ParameterOptimizer::ParameterOptimizer":[26,2,1,"_CPPv2N18ParameterOptimizer18ParameterOptimizerEv"],"ParameterOptimizer::create":[50,2,1,"_CPPv2N18ParameterOptimizer6createERK18OptimizationConfigb"],"ParameterOptimizer::finishBatch":[26,2,1,"_CPPv2N18ParameterOptimizer11finishBatchEv"],"ParameterOptimizer::finishPass":[26,2,1,"_CPPv2N18ParameterOptimizer10finishPassEv"],"ParameterOptimizer::getParameterTypes":[26,2,1,"_CPPv2N18ParameterOptimizer17getParameterTypesEv"],"ParameterOptimizer::init":[26,2,1,"_CPPv2N18ParameterOptimizer4initE6size_tPK15ParameterConfig"],"ParameterOptimizer::m":[26,3,1,"_CPPv2N18ParameterOptimizer1mE"],"ParameterOptimizer::needSpecialTraversal":[26,2,1,"_CPPv2N18ParameterOptimizer20needSpecialTraversalERK15ParameterConfig"],"ParameterOptimizer::startBatch":[26,2,1,"_CPPv2N18ParameterOptimizer10startBatchE6size_t"],"ParameterOptimizer::startPass":[26,2,1,"_CPPv2N18ParameterOptimizer9startPassEv"],"ParameterOptimizer::update":[26,2,1,"_CPPv2N18ParameterOptimizer6updateERKNSt6vectorIP6VectorEERK15ParameterConfig6size_t"],"ParameterOptimizer::~ParameterOptimizer":[26,2,1,"_CPPv2N18ParameterOptimizerD0Ev"],"ParameterTraverseCallback::DISABLE_COPY_AND_ASSIGN":[26,2,1,"_CPPv2N25ParameterTraverseCallback23DISABLE_COPY_AND_ASSIGNE25ParameterTraverseCallback"],"ParameterTraverseCallback::ParameterTraverseCallback":[26,2,1,"_CPPv2N25ParameterTraverseCallback25ParameterTraverseCallbackEv"],"ParameterTraverseCallback::apply":[26,2,1,"_CPPv2N25ParameterTraverseCallback5applyERKNSt6vectorIP6VectorEERK15ParameterConfig6size_t"],"ParameterTraverseCallback::m":[26,3,1,"_CPPv2N25ParameterTraverseCallback1mE"],"ParameterTraverseCallback::~ParameterTraverseCallback":[26,2,1,"_CPPv2N25ParameterTraverseCallbackD0Ev"],"SequenceGenerator::DISABLE_COPY_AND_ASSIGN":[26,2,1,"_CPPv2N17SequenceGenerator23DISABLE_COPY_AND_ASSIGNE17SequenceGenerator"],"SequenceGenerator::SequenceGenerator":[26,2,1,"_CPPv2N17SequenceGenerator17SequenceGeneratorEv"],"SequenceGenerator::createByGradientMachineSharedPtr":[26,2,1,"_CPPv2N17SequenceGenerator32createByGradientMachineSharedPtrEPv"],"SequenceGenerator::generateSequence":[26,2,1,"_CPPv2N17SequenceGenerator16generateSequenceERK9Arguments"],"SequenceGenerator::m":[26,3,1,"_CPPv2N17SequenceGenerator1mE"],"SequenceGenerator::setBeamSize":[26,2,1,"_CPPv2N17SequenceGenerator11setBeamSizeE6size_t"],"SequenceGenerator::setBos":[26,2,1,"_CPPv2N17SequenceGenerator6setBosE6size_t"],"SequenceGenerator::setDict":[26,2,1,"_CPPv2N17SequenceGenerator7setDictERKNSt6vectorINSt6stringEEE"],"SequenceGenerator::setEos":[26,2,1,"_CPPv2N17SequenceGenerator6setEosE6size_t"],"SequenceGenerator::setMaxLength":[26,2,1,"_CPPv2N17SequenceGenerator12setMaxLengthE6size_t"],"SequenceGenerator::~SequenceGenerator":[26,2,1,"_CPPv2N17SequenceGeneratorD0Ev"],"Trainer::DISABLE_COPY_AND_ASSIGN":[26,2,1,"_CPPv2N7Trainer23DISABLE_COPY_AND_ASSIGNE7Trainer"],"Trainer::Trainer":[26,2,1,"_CPPv2N7Trainer7TrainerEv"],"Trainer::createByCommandLine":[26,2,1,"_CPPv2N7Trainer19createByCommandLineEv"],"Trainer::finishTrain":[26,2,1,"_CPPv2N7Trainer11finishTrainEv"],"Trainer::finishTrainOneBatch":[26,2,1,"_CPPv2N7Trainer19finishTrainOneBatchEv"],"Trainer::finishTrainPass":[26,2,1,"_CPPv2N7Trainer15finishTrainPassEv"],"Trainer::forwardOneBatch":[26,2,1,"_CPPv2N7Trainer15forwardOneBatchEv"],"Trainer::getLayerOutput":[26,2,1,"_CPPv2N7Trainer14getLayerOutputERKNSt6stringE"],"Trainer::getNetworkOutput":[26,2,1,"_CPPv2N7Trainer16getNetworkOutputEv"],"Trainer::m":[26,3,1,"_CPPv2N7Trainer1mE"],"Trainer::prepareBatchData":[26,2,1,"_CPPv2N7Trainer16prepareBatchDataE6size_t"],"Trainer::setBatchSize":[26,2,1,"_CPPv2N7Trainer12setBatchSizeE6size_t"],"Trainer::startTrain":[26,2,1,"_CPPv2N7Trainer10startTrainEv"],"Trainer::startTrainPass":[26,2,1,"_CPPv2N7Trainer14startTrainPassEv"],"Trainer::trainOneBatch":[26,2,1,"_CPPv2N7Trainer13trainOneBatchE6size_t"],"Trainer::~Trainer":[26,2,1,"_CPPv2N7TrainerD0Ev"],"TrainerConfig::DISABLE_COPY_AND_ASSIGN":[26,2,1,"_CPPv2N13TrainerConfig23DISABLE_COPY_AND_ASSIGNE13TrainerConfig"],"TrainerConfig::TrainerConfig":[26,2,1,"_CPPv2N13TrainerConfig13TrainerConfigEv"],"TrainerConfig::createFromTrainerConfigFile":[26,2,1,"_CPPv2N13TrainerConfig27createFromTrainerConfigFileERKNSt6stringE"],"TrainerConfig::getModelConfig":[26,2,1,"_CPPv2N13TrainerConfig14getModelConfigEv"],"TrainerConfig::getOptimizationConfig":[26,2,1,"_CPPv2N13TrainerConfig21getOptimizationConfigEv"],"TrainerConfig::m":[26,3,1,"_CPPv2N13TrainerConfig1mE"],"TrainerConfig::~TrainerConfig":[26,2,1,"_CPPv2N13TrainerConfigD0Ev"],"UpdateCallback::apply":[26,2,1,"_CPPv2N14UpdateCallback5applyEP9Parameter"],"UpdateCallback::~UpdateCallback":[26,2,1,"_CPPv2N14UpdateCallbackD0Ev"],"Vector::DISABLE_COPY_AND_ASSIGN":[26,2,1,"_CPPv2N6Vector23DISABLE_COPY_AND_ASSIGNE6Vector"],"Vector::Vector":[26,2,1,"_CPPv2N6Vector6VectorEv"],"Vector::copyFromNumpyArray":[26,2,1,"_CPPv2N6Vector18copyFromNumpyArrayEPfi"],"Vector::copyToNumpyArray":[26,2,1,"_CPPv2N6Vector16copyToNumpyArrayEPPfPi"],"Vector::create":[26,2,1,"_CPPv2N6Vector6createERKNSt6vectorIfEEb"],"Vector::createByPaddleVectorPtr":[26,2,1,"_CPPv2N6Vector23createByPaddleVectorPtrEPv"],"Vector::createCpuVectorFromNumpy":[26,2,1,"_CPPv2N6Vector24createCpuVectorFromNumpyEPfib"],"Vector::createGpuVectorFromNumpy":[26,2,1,"_CPPv2N6Vector24createGpuVectorFromNumpyEPfi"],"Vector::createZero":[26,2,1,"_CPPv2N6Vector10createZeroE6size_tb"],"Vector::get":[26,2,1,"_CPPv2N6Vector3getEK6size_t"],"Vector::getSharedPtr":[26,2,1,"_CPPv2N6Vector12getSharedPtrEv"],"Vector::getSize":[26,2,1,"_CPPv2N6Vector7getSizeEv"],"Vector::isGpu":[26,2,1,"_CPPv2N6Vector5isGpuEv"],"Vector::m":[26,3,1,"_CPPv2N6Vector1mE"],"Vector::set":[26,2,1,"_CPPv2N6Vector3setEK6size_tf"],"Vector::toNumpyArrayInplace":[26,2,1,"_CPPv2N6Vector19toNumpyArrayInplaceEPPfPi"],"Vector::~Vector":[26,2,1,"_CPPv2N6VectorD0Ev"],"_cudnn_convolution_descriptor::desc":[27,3,1,"_CPPv2N29_cudnn_convolution_descriptor4descE"],"_cudnn_convolution_descriptor::filter":[27,3,1,"_CPPv2N29_cudnn_convolution_descriptor6filterE"],"_cudnn_convolution_descriptor::input_image":[27,3,1,"_CPPv2N29_cudnn_convolution_descriptor11input_imageE"],"_cudnn_convolution_descriptor::mode":[27,3,1,"_CPPv2N29_cudnn_convolution_descriptor4modeE"],"_cudnn_convolution_descriptor::padding_height":[27,3,1,"_CPPv2N29_cudnn_convolution_descriptor14padding_heightE"],"_cudnn_convolution_descriptor::padding_width":[27,3,1,"_CPPv2N29_cudnn_convolution_descriptor13padding_widthE"],"_cudnn_convolution_descriptor::stride_height":[27,3,1,"_CPPv2N29_cudnn_convolution_descriptor13stride_heightE"],"_cudnn_convolution_descriptor::stride_width":[27,3,1,"_CPPv2N29_cudnn_convolution_descriptor12stride_widthE"],"_cudnn_convolution_descriptor::upscalex":[27,3,1,"_CPPv2N29_cudnn_convolution_descriptor8upscalexE"],"_cudnn_convolution_descriptor::upscaley":[27,3,1,"_CPPv2N29_cudnn_convolution_descriptor8upscaleyE"],"_cudnn_filter_descriptor::data_type":[27,3,1,"_CPPv2N24_cudnn_filter_descriptor9data_typeE"],"_cudnn_filter_descriptor::desc":[27,3,1,"_CPPv2N24_cudnn_filter_descriptor4descE"],"_cudnn_filter_descriptor::filter_height":[27,3,1,"_CPPv2N24_cudnn_filter_descriptor13filter_heightE"],"_cudnn_filter_descriptor::filter_width":[27,3,1,"_CPPv2N24_cudnn_filter_descriptor12filter_widthE"],"_cudnn_filter_descriptor::input_feature_maps":[27,3,1,"_CPPv2N24_cudnn_filter_descriptor18input_feature_mapsE"],"_cudnn_filter_descriptor::output_feature_maps":[27,3,1,"_CPPv2N24_cudnn_filter_descriptor19output_feature_mapsE"],"_cudnn_pooling_descriptor::desc":[27,3,1,"_CPPv2N25_cudnn_pooling_descriptor4descE"],"_cudnn_pooling_descriptor::mode":[27,3,1,"_CPPv2N25_cudnn_pooling_descriptor4modeE"],"_cudnn_pooling_descriptor::stride_height":[27,3,1,"_CPPv2N25_cudnn_pooling_descriptor13stride_heightE"],"_cudnn_pooling_descriptor::stride_width":[27,3,1,"_CPPv2N25_cudnn_pooling_descriptor12stride_widthE"],"_cudnn_pooling_descriptor::window_height":[27,3,1,"_CPPv2N25_cudnn_pooling_descriptor13window_heightE"],"_cudnn_pooling_descriptor::window_width":[27,3,1,"_CPPv2N25_cudnn_pooling_descriptor12window_widthE"],"_cudnn_tensor_descriptor::batch_size":[27,3,1,"_CPPv2N24_cudnn_tensor_descriptor10batch_sizeE"],"_cudnn_tensor_descriptor::data_type":[27,3,1,"_CPPv2N24_cudnn_tensor_descriptor9data_typeE"],"_cudnn_tensor_descriptor::desc":[27,3,1,"_CPPv2N24_cudnn_tensor_descriptor4descE"],"_cudnn_tensor_descriptor::feature_maps":[27,3,1,"_CPPv2N24_cudnn_tensor_descriptor12feature_mapsE"],"_cudnn_tensor_descriptor::format":[27,3,1,"_CPPv2N24_cudnn_tensor_descriptor6formatE"],"_cudnn_tensor_descriptor::height":[27,3,1,"_CPPv2N24_cudnn_tensor_descriptor6heightE"],"_cudnn_tensor_descriptor::width":[27,3,1,"_CPPv2N24_cudnn_tensor_descriptor5widthE"],"_global_device_resources::cudnn_handle":[27,3,1,"_CPPv2N24_global_device_resources12cudnn_handleE"],"_global_device_resources::gen":[27,3,1,"_CPPv2N24_global_device_resources3genE"],"_global_device_resources::gen_mutex":[27,3,1,"_CPPv2N24_global_device_resources9gen_mutexE"],"_global_device_resources::handle":[27,3,1,"_CPPv2N24_global_device_resources6handleE"],"_global_device_resources::stream":[27,3,1,"_CPPv2N24_global_device_resources6streamE"],"_hl_csc_matrix::col_s":[30,3,1,"_CPPv2N14_hl_csc_matrix5col_sE"],"_hl_csc_matrix::csc_col":[30,3,1,"_CPPv2N14_hl_csc_matrix7csc_colE"],"_hl_csc_matrix::csc_row":[30,3,1,"_CPPv2N14_hl_csc_matrix7csc_rowE"],"_hl_csc_matrix::csc_val":[30,3,1,"_CPPv2N14_hl_csc_matrix7csc_valE"],"_hl_csc_matrix::nnz_s":[30,3,1,"_CPPv2N14_hl_csc_matrix5nnz_sE"],"_hl_csc_matrix::sparsity":[30,3,1,"_CPPv2N14_hl_csc_matrix8sparsityE"],"_hl_csr_matrix::csr_col":[30,3,1,"_CPPv2N14_hl_csr_matrix7csr_colE"],"_hl_csr_matrix::csr_row":[30,3,1,"_CPPv2N14_hl_csr_matrix7csr_rowE"],"_hl_csr_matrix::csr_val":[30,3,1,"_CPPv2N14_hl_csr_matrix7csr_valE"],"_hl_csr_matrix::nnz_s":[30,3,1,"_CPPv2N14_hl_csr_matrix5nnz_sE"],"_hl_csr_matrix::row_s":[30,3,1,"_CPPv2N14_hl_csr_matrix5row_sE"],"_hl_csr_matrix::sparsity":[30,3,1,"_CPPv2N14_hl_csr_matrix8sparsityE"],"_hl_device_prop::device":[27,3,1,"_CPPv2N15_hl_device_prop6deviceE"],"_hl_device_prop::device_mem":[27,3,1,"_CPPv2N15_hl_device_prop10device_memE"],"_hl_device_prop::device_name":[27,3,1,"_CPPv2N15_hl_device_prop11device_nameE"],"_hl_device_prop::device_resources":[27,3,1,"_CPPv2N15_hl_device_prop16device_resourcesE"],"_hl_device_prop::device_type":[27,3,1,"_CPPv2N15_hl_device_prop11device_typeE"],"_hl_device_prop::is_local":[27,3,1,"_CPPv2N15_hl_device_prop8is_localE"],"_hl_device_prop::major":[27,3,1,"_CPPv2N15_hl_device_prop5majorE"],"_hl_device_prop::minor":[27,3,1,"_CPPv2N15_hl_device_prop5minorE"],"_hl_event_st::cu_event":[27,3,1,"_CPPv2N12_hl_event_st8cu_eventE"],"_hl_sparse_matrix_s::cols":[34,3,1,"_CPPv2N19_hl_sparse_matrix_s4colsE"],"_hl_sparse_matrix_s::format":[34,3,1,"_CPPv2N19_hl_sparse_matrix_s6formatE"],"_hl_sparse_matrix_s::matrix":[34,3,1,"_CPPv2N19_hl_sparse_matrix_s6matrixE"],"_hl_sparse_matrix_s::nnz":[34,3,1,"_CPPv2N19_hl_sparse_matrix_s3nnzE"],"_hl_sparse_matrix_s::rows":[34,3,1,"_CPPv2N19_hl_sparse_matrix_s4rowsE"],"_hl_sparse_matrix_s::type":[34,3,1,"_CPPv2N19_hl_sparse_matrix_s4typeE"],"_hl_thread_resource::cpu_mem":[34,3,1,"_CPPv2N19_hl_thread_resource7cpu_memE"],"_hl_thread_resource::cudnn_desc":[34,3,1,"_CPPv2N19_hl_thread_resource10cudnn_descE"],"_hl_thread_resource::cudnn_handle":[34,3,1,"_CPPv2N19_hl_thread_resource12cudnn_handleE"],"_hl_thread_resource::device":[34,3,1,"_CPPv2N19_hl_thread_resource6deviceE"],"_hl_thread_resource::event":[34,3,1,"_CPPv2N19_hl_thread_resource5eventE"],"_hl_thread_resource::gen":[34,3,1,"_CPPv2N19_hl_thread_resource3genE"],"_hl_thread_resource::gen_mutex":[34,3,1,"_CPPv2N19_hl_thread_resource9gen_mutexE"],"_hl_thread_resource::gpu_mem":[34,3,1,"_CPPv2N19_hl_thread_resource7gpu_memE"],"_hl_thread_resource::handle":[34,3,1,"_CPPv2N19_hl_thread_resource6handleE"],"_hl_thread_resource::is_init":[34,3,1,"_CPPv2N19_hl_thread_resource7is_initE"],"_hl_thread_resource::major":[34,3,1,"_CPPv2N19_hl_thread_resource5majorE"],"_hl_thread_resource::stream":[34,3,1,"_CPPv2N19_hl_thread_resource6streamE"],"_thread_device_resources::cpu_mem":[27,3,1,"_CPPv2N24_thread_device_resources7cpu_memE"],"_thread_device_resources::gpu_mem":[27,3,1,"_CPPv2N24_thread_device_resources7gpu_memE"],"_thread_device_resources::mem_event":[27,3,1,"_CPPv2N24_thread_device_resources9mem_eventE"],"_thread_device_resources::stream":[27,3,1,"_CPPv2N24_thread_device_resources6streamE"],"aggregate::max":[30,1,1,"_CPPv2N9aggregate3maxE"],"aggregate::min":[30,1,1,"_CPPv2N9aggregate3minE"],"aggregate::sum":[30,1,1,"_CPPv2N9aggregate3sumE"],"base::binary":[30,4,1,"_CPPv2N4base6binaryE"],"base::binary::add":[30,1,1,"_CPPv2N4base6binary3addE"],"base::binary::add2":[30,1,1,"_CPPv2N4base6binary4add2E"],"base::binary::add2::add2":[30,2,1,"_CPPv2N4base6binary4add24add2EK4realK4real"],"base::binary::add2::p1":[30,3,1,"_CPPv2N4base6binary4add22p1E"],"base::binary::add2::p2":[30,3,1,"_CPPv2N4base6binary4add22p2E"],"base::binary::classificationError":[30,1,1,"_CPPv2N4base6binary19classificationErrorE"],"base::binary::classificationError::classificationError":[30,2,1,"_CPPv2N4base6binary19classificationError19classificationErrorEK4real"],"base::binary::classificationError::p":[30,3,1,"_CPPv2N4base6binary19classificationError1pE"],"base::binary::div":[30,1,1,"_CPPv2N4base6binary3divE"],"base::binary::first":[30,1,1,"_CPPv2N4base6binary5firstE"],"base::binary::mul":[30,1,1,"_CPPv2N4base6binary3mulE"],"base::binary::second":[30,1,1,"_CPPv2N4base6binary6secondE"],"base::binary::squaredDiff":[30,1,1,"_CPPv2N4base6binary11squaredDiffE"],"base::binary::sub":[30,1,1,"_CPPv2N4base6binary3subE"],"base::unary":[30,4,1,"_CPPv2N4base5unaryE"],"base::unary::identity":[30,1,1,"_CPPv2N4base5unary8identityE"],"hl_gru_grad::gateGrad":[34,3,1,"_CPPv2N11hl_gru_grad8gateGradE"],"hl_gru_grad::gateWeightGrad":[34,3,1,"_CPPv2N11hl_gru_grad14gateWeightGradE"],"hl_gru_grad::outputGrad":[34,3,1,"_CPPv2N11hl_gru_grad10outputGradE"],"hl_gru_grad::prevOutGrad":[34,3,1,"_CPPv2N11hl_gru_grad11prevOutGradE"],"hl_gru_grad::resetOutputGrad":[34,3,1,"_CPPv2N11hl_gru_grad15resetOutputGradE"],"hl_gru_grad::stateWeightGrad":[34,3,1,"_CPPv2N11hl_gru_grad15stateWeightGradE"],"hl_gru_value::gateValue":[34,3,1,"_CPPv2N12hl_gru_value9gateValueE"],"hl_gru_value::gateWeight":[34,3,1,"_CPPv2N12hl_gru_value10gateWeightE"],"hl_gru_value::outputValue":[34,3,1,"_CPPv2N12hl_gru_value11outputValueE"],"hl_gru_value::prevOutValue":[34,3,1,"_CPPv2N12hl_gru_value12prevOutValueE"],"hl_gru_value::resetOutputValue":[34,3,1,"_CPPv2N12hl_gru_value16resetOutputValueE"],"hl_gru_value::stateWeight":[34,3,1,"_CPPv2N12hl_gru_value11stateWeightE"],"hl_lstm_grad::checkFgGrad":[34,3,1,"_CPPv2N12hl_lstm_grad11checkFgGradE"],"hl_lstm_grad::checkIgGrad":[34,3,1,"_CPPv2N12hl_lstm_grad11checkIgGradE"],"hl_lstm_grad::checkOgGrad":[34,3,1,"_CPPv2N12hl_lstm_grad11checkOgGradE"],"hl_lstm_grad::gateGrad":[34,3,1,"_CPPv2N12hl_lstm_grad8gateGradE"],"hl_lstm_grad::outputGrad":[34,3,1,"_CPPv2N12hl_lstm_grad10outputGradE"],"hl_lstm_grad::prevStateGrad":[34,3,1,"_CPPv2N12hl_lstm_grad13prevStateGradE"],"hl_lstm_grad::stateActiveGrad":[34,3,1,"_CPPv2N12hl_lstm_grad15stateActiveGradE"],"hl_lstm_grad::stateGrad":[34,3,1,"_CPPv2N12hl_lstm_grad9stateGradE"],"hl_lstm_value::checkFg":[34,3,1,"_CPPv2N13hl_lstm_value7checkFgE"],"hl_lstm_value::checkIg":[34,3,1,"_CPPv2N13hl_lstm_value7checkIgE"],"hl_lstm_value::checkOg":[34,3,1,"_CPPv2N13hl_lstm_value7checkOgE"],"hl_lstm_value::gateValue":[34,3,1,"_CPPv2N13hl_lstm_value9gateValueE"],"hl_lstm_value::outputValue":[34,3,1,"_CPPv2N13hl_lstm_value11outputValueE"],"hl_lstm_value::prevStateValue":[34,3,1,"_CPPv2N13hl_lstm_value14prevStateValueE"],"hl_lstm_value::stateActiveValue":[34,3,1,"_CPPv2N13hl_lstm_value16stateActiveValueE"],"hl_lstm_value::stateValue":[34,3,1,"_CPPv2N13hl_lstm_value10stateValueE"],"hppl::Active":[32,1,1,"_CPPv2N4hppl6ActiveE"],"hppl::Active<T>::backward":[32,4,1,"_CPPv2IEN4hppl6ActiveI1TE8backwardE"],"hppl::Active<T>::forward":[32,4,1,"_CPPv2IEN4hppl6ActiveI1TE7forwardE"],"hppl::backward":[32,4,1,"_CPPv2N4hppl8backwardE"],"hppl::backward::gru_resetGrad":[32,1,1,"_CPPv2N4hppl8backward13gru_resetGradE"],"hppl::backward::gru_resetGrad::avx":[32,3,1,"_CPPv2N4hppl8backward13gru_resetGrad3avxE"],"hppl::backward::gru_stateGrad":[32,1,1,"_CPPv2N4hppl8backward13gru_stateGradE"],"hppl::backward::gru_stateGrad::avx":[32,3,1,"_CPPv2N4hppl8backward13gru_stateGrad3avxE"],"hppl::backward::lstm":[32,1,1,"_CPPv2N4hppl8backward4lstmE"],"hppl::backward::lstm::avx":[32,3,1,"_CPPv2N4hppl8backward4lstm3avxE"],"hppl::cpu":[32,4,1,"_CPPv2N4hppl3cpuE"],"hppl::cpu::backward":[32,3,1,"_CPPv2N4hppl3cpu8backwardE"],"hppl::cpu::forward":[32,3,1,"_CPPv2N4hppl3cpu7forwardE"],"hppl::forward":[32,4,1,"_CPPv2N4hppl7forwardE"],"hppl::forward::gru_finalOutput":[32,1,1,"_CPPv2N4hppl7forward15gru_finalOutputE"],"hppl::forward::gru_finalOutput::avx":[32,3,1,"_CPPv2N4hppl7forward15gru_finalOutput3avxE"],"hppl::forward::gru_resetOutput":[32,1,1,"_CPPv2N4hppl7forward15gru_resetOutputE"],"hppl::forward::gru_resetOutput::avx":[32,3,1,"_CPPv2N4hppl7forward15gru_resetOutput3avxE"],"hppl::forward::lstm":[32,1,1,"_CPPv2N4hppl7forward4lstmE"],"hppl::forward::lstm::avx":[32,3,1,"_CPPv2N4hppl7forward4lstm3avxE"],"hppl::linear":[32,2,1,"_CPPv2N4hppl6linearEK6__m256K6__m256"],"hppl::relu":[32,2,1,"_CPPv2N4hppl4reluEK6__m256K6__m256"],"hppl::sigmoid":[32,2,1,"_CPPv2N4hppl7sigmoidEK6__m256K6__m256"],"hppl::tanh":[32,2,1,"_CPPv2N4hppl4tanhEK6__m256K6__m256"],"operator<<":[38,2,1,"_CPPv2lsRRNSt7ostreamER9Evaluator"],"paddle::ActivationFunction":[35,1,1,"_CPPv2N6paddle18ActivationFunctionE"],"paddle::ActivationFunction::ActivationFunction":[35,2,1,"_CPPv2N6paddle18ActivationFunction18ActivationFunctionEv"],"paddle::ActivationFunction::backward":[35,2,1,"_CPPv2N6paddle18ActivationFunction8backwardER8Argument"],"paddle::ActivationFunction::create":[35,2,1,"_CPPv2N6paddle18ActivationFunction6createERKNSt6stringE"],"paddle::ActivationFunction::forward":[35,2,1,"_CPPv2N6paddle18ActivationFunction7forwardER8Argument"],"paddle::ActivationFunction::getName":[35,2,1,"_CPPv2N6paddle18ActivationFunction7getNameEv"],"paddle::ActivationFunction::~ActivationFunction":[35,2,1,"_CPPv2N6paddle18ActivationFunctionD0Ev"],"paddle::AdaDeltaParameterOptimizer":[50,1,1,"_CPPv2N6paddle26AdaDeltaParameterOptimizerE"],"paddle::AdaDeltaParameterOptimizer::AdaDeltaParameterOptimizer":[50,2,1,"_CPPv2N6paddle26AdaDeltaParameterOptimizer26AdaDeltaParameterOptimizerERK18OptimizationConfig"],"paddle::AdaDeltaParameterOptimizer::epsilon_":[50,3,1,"_CPPv2N6paddle26AdaDeltaParameterOptimizer8epsilon_E"],"paddle::AdaDeltaParameterOptimizer::rou_":[50,3,1,"_CPPv2N6paddle26AdaDeltaParameterOptimizer4rou_E"],"paddle::AdaDeltaParameterOptimizer::startBatch":[50,2,1,"_CPPv2N6paddle26AdaDeltaParameterOptimizer10startBatchE7int64_t"],"paddle::AdaDeltaParameterOptimizer::update":[50,2,1,"_CPPv2N6paddle26AdaDeltaParameterOptimizer6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::AdagradParameterOptimizer":[50,1,1,"_CPPv2N6paddle25AdagradParameterOptimizerE"],"paddle::AdagradParameterOptimizer::AdagradParameterOptimizer":[50,2,1,"_CPPv2N6paddle25AdagradParameterOptimizer25AdagradParameterOptimizerERK18OptimizationConfig"],"paddle::AdagradParameterOptimizer::kMaxNumAccumulates":[50,3,1,"_CPPv2N6paddle25AdagradParameterOptimizer18kMaxNumAccumulatesE"],"paddle::AdagradParameterOptimizer::needSpecialTraversal":[50,2,1,"_CPPv2N6paddle25AdagradParameterOptimizer20needSpecialTraversalERK15ParameterConfig"],"paddle::AdagradParameterOptimizer::numUpdates_":[50,3,1,"_CPPv2N6paddle25AdagradParameterOptimizer11numUpdates_E"],"paddle::AdagradParameterOptimizer::startBatch":[50,2,1,"_CPPv2N6paddle25AdagradParameterOptimizer10startBatchE7int64_t"],"paddle::AdagradParameterOptimizer::update":[50,2,1,"_CPPv2N6paddle25AdagradParameterOptimizer6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::AdamParameterOptimizer":[50,1,1,"_CPPv2N6paddle22AdamParameterOptimizerE"],"paddle::AdamParameterOptimizer::AdamParameterOptimizer":[50,2,1,"_CPPv2N6paddle22AdamParameterOptimizer22AdamParameterOptimizerERK18OptimizationConfig"],"paddle::AdamParameterOptimizer::beta1_":[50,3,1,"_CPPv2N6paddle22AdamParameterOptimizer6beta1_E"],"paddle::AdamParameterOptimizer::beta2_":[50,3,1,"_CPPv2N6paddle22AdamParameterOptimizer6beta2_E"],"paddle::AdamParameterOptimizer::epsilon_":[50,3,1,"_CPPv2N6paddle22AdamParameterOptimizer8epsilon_E"],"paddle::AdamParameterOptimizer::finishBatch":[50,2,1,"_CPPv2N6paddle22AdamParameterOptimizer11finishBatchEv"],"paddle::AdamParameterOptimizer::learningRate_":[50,3,1,"_CPPv2N6paddle22AdamParameterOptimizer13learningRate_E"],"paddle::AdamParameterOptimizer::step_":[50,3,1,"_CPPv2N6paddle22AdamParameterOptimizer5step_E"],"paddle::AdamParameterOptimizer::update":[50,2,1,"_CPPv2N6paddle22AdamParameterOptimizer6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::AdamaxParameterOptimizer":[50,1,1,"_CPPv2N6paddle24AdamaxParameterOptimizerE"],"paddle::AdamaxParameterOptimizer::AdamaxParameterOptimizer":[50,2,1,"_CPPv2N6paddle24AdamaxParameterOptimizer24AdamaxParameterOptimizerERK18OptimizationConfig"],"paddle::AdamaxParameterOptimizer::beta1_":[50,3,1,"_CPPv2N6paddle24AdamaxParameterOptimizer6beta1_E"],"paddle::AdamaxParameterOptimizer::beta2_":[50,3,1,"_CPPv2N6paddle24AdamaxParameterOptimizer6beta2_E"],"paddle::AdamaxParameterOptimizer::finishBatch":[50,2,1,"_CPPv2N6paddle24AdamaxParameterOptimizer11finishBatchEv"],"paddle::AdamaxParameterOptimizer::learningRate_":[50,3,1,"_CPPv2N6paddle24AdamaxParameterOptimizer13learningRate_E"],"paddle::AdamaxParameterOptimizer::step_":[50,3,1,"_CPPv2N6paddle24AdamaxParameterOptimizer5step_E"],"paddle::AdamaxParameterOptimizer::update":[50,2,1,"_CPPv2N6paddle24AdamaxParameterOptimizer6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::AddOptimizer":[50,1,1,"_CPPv2N6paddle12AddOptimizerE"],"paddle::AddOptimizer::AddOptimizer":[50,2,1,"_CPPv2N6paddle12AddOptimizer12AddOptimizerERK18OptimizationConfig"],"paddle::AddOptimizer::startBatch":[50,2,1,"_CPPv2N6paddle12AddOptimizer10startBatchE7int64_t"],"paddle::AddOptimizer::update":[50,2,1,"_CPPv2N6paddle12AddOptimizer6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::AddtoLayer":[43,1,1,"_CPPv2N6paddle10AddtoLayerE"],"paddle::AddtoLayer::AddtoLayer":[43,2,1,"_CPPv2N6paddle10AddtoLayer10AddtoLayerERK11LayerConfig"],"paddle::AddtoLayer::backward":[43,2,1,"_CPPv2N6paddle10AddtoLayer8backwardERK14UpdateCallback"],"paddle::AddtoLayer::biases_":[43,3,1,"_CPPv2N6paddle10AddtoLayer7biases_E"],"paddle::AddtoLayer::forward":[43,2,1,"_CPPv2N6paddle10AddtoLayer7forwardE8PassType"],"paddle::AddtoLayer::init":[43,2,1,"_CPPv2N6paddle10AddtoLayer4initERK8LayerMapRK12ParameterMap"],"paddle::AddtoLayer::~AddtoLayer":[43,2,1,"_CPPv2N6paddle10AddtoLayerD0Ev"],"paddle::AgentLayer":[43,1,1,"_CPPv2N6paddle10AgentLayerE"],"paddle::AgentLayer::AgentLayer":[43,2,1,"_CPPv2N6paddle10AgentLayer10AgentLayerERK11LayerConfig"],"paddle::AgentLayer::backward":[43,2,1,"_CPPv2N6paddle10AgentLayer8backwardERK14UpdateCallback"],"paddle::AgentLayer::forward":[43,2,1,"_CPPv2N6paddle10AgentLayer7forwardE8PassType"],"paddle::AgentLayer::init":[43,2,1,"_CPPv2N6paddle10AgentLayer4initERK8LayerMapRK12ParameterMap"],"paddle::AgentLayer::numSamples_":[43,3,1,"_CPPv2N6paddle10AgentLayer11numSamples_E"],"paddle::AgentLayer::realLayer_":[43,3,1,"_CPPv2N6paddle10AgentLayer10realLayer_E"],"paddle::AgentLayer::setRealLayer":[43,2,1,"_CPPv2N6paddle10AgentLayer12setRealLayerE8LayerPtri"],"paddle::AgentLayer::~AgentLayer":[43,2,1,"_CPPv2N6paddle10AgentLayerD0Ev"],"paddle::Allocator":[48,1,1,"_CPPv2N6paddle9AllocatorE"],"paddle::Allocator::alloc":[48,2,1,"_CPPv2N6paddle9Allocator5allocE6size_t"],"paddle::Allocator::free":[48,2,1,"_CPPv2N6paddle9Allocator4freeEPv"],"paddle::Allocator::getName":[48,2,1,"_CPPv2N6paddle9Allocator7getNameEv"],"paddle::Allocator::~Allocator":[48,2,1,"_CPPv2N6paddle9AllocatorD0Ev"],"paddle::Argument":[52,1,1,"_CPPv2N6paddle8ArgumentE"],"paddle::Argument::Argument":[52,2,1,"_CPPv2N6paddle8Argument8ArgumentEv"],"paddle::Argument::allCount":[52,3,1,"_CPPv2N6paddle8Argument8allCountE"],"paddle::Argument::checkSubset":[52,2,1,"_CPPv2N6paddle8Argument11checkSubsetEv"],"paddle::Argument::concat":[52,2,1,"_CPPv2N6paddle8Argument6concatERKNSt6vectorI8ArgumentEEb11hl_stream_t8PassType"],"paddle::Argument::countIncrement":[52,2,1,"_CPPv2N6paddle8Argument14countIncrementEv"],"paddle::Argument::cpuSequenceDims":[52,3,1,"_CPPv2N6paddle8Argument15cpuSequenceDimsE"],"paddle::Argument::dataId":[52,3,1,"_CPPv2N6paddle8Argument6dataIdE"],"paddle::Argument::degradeSequence":[52,2,1,"_CPPv2N6paddle8Argument15degradeSequenceERK8Argumentb"],"paddle::Argument::deviceId":[52,3,1,"_CPPv2N6paddle8Argument8deviceIdE"],"paddle::Argument::frameHeight":[52,3,1,"_CPPv2N6paddle8Argument11frameHeightE"],"paddle::Argument::frameWidth":[52,3,1,"_CPPv2N6paddle8Argument10frameWidthE"],"paddle::Argument::getAllCount":[52,2,1,"_CPPv2N6paddle8Argument11getAllCountEv"],"paddle::Argument::getBatchSize":[52,2,1,"_CPPv2N6paddle8Argument12getBatchSizeEv"],"paddle::Argument::getCpuStartPositions":[52,2,1,"_CPPv2N6paddle8Argument20getCpuStartPositionsEv"],"paddle::Argument::getFrameHeight":[52,2,1,"_CPPv2N6paddle8Argument14getFrameHeightEv"],"paddle::Argument::getFrameWidth":[52,2,1,"_CPPv2N6paddle8Argument13getFrameWidthEv"],"paddle::Argument::getNumSequences":[52,2,1,"_CPPv2N6paddle8Argument15getNumSequencesEv"],"paddle::Argument::getNumSubSequences":[52,2,1,"_CPPv2N6paddle8Argument18getNumSubSequencesEv"],"paddle::Argument::getSeqLengthAndStart":[52,2,1,"_CPPv2N6paddle8Argument20getSeqLengthAndStartEPNSt6vectorINSt5tupleIiiiiEEEEPi"],"paddle::Argument::grad":[52,3,1,"_CPPv2N6paddle8Argument4gradE"],"paddle::Argument::gradCount":[52,3,1,"_CPPv2N6paddle8Argument9gradCountE"],"paddle::Argument::gradReadyCond":[52,3,1,"_CPPv2N6paddle8Argument13gradReadyCondE"],"paddle::Argument::hasSubseq":[52,2,1,"_CPPv2N6paddle8Argument9hasSubseqEv"],"paddle::Argument::ids":[52,3,1,"_CPPv2N6paddle8Argument3idsE"],"paddle::Argument::in":[52,3,1,"_CPPv2N6paddle8Argument2inE"],"paddle::Argument::notifyGradReady":[52,2,1,"_CPPv2N6paddle8Argument15notifyGradReadyEv"],"paddle::Argument::notifyValueReady":[52,2,1,"_CPPv2N6paddle8Argument16notifyValueReadyEv"],"paddle::Argument::operator=":[52,2,1,"_CPPv2N6paddle8ArgumentaSERK8Argument"],"paddle::Argument::resizeAndCopyFrom":[52,2,1,"_CPPv2N6paddle8Argument17resizeAndCopyFromERK8Argumentb11hl_stream_t"],"paddle::Argument::sequenceStartPositions":[52,3,1,"_CPPv2N6paddle8Argument22sequenceStartPositionsE"],"paddle::Argument::setFrameHeight":[52,2,1,"_CPPv2N6paddle8Argument14setFrameHeightE6size_t"],"paddle::Argument::setFrameWidth":[52,2,1,"_CPPv2N6paddle8Argument13setFrameWidthE6size_t"],"paddle::Argument::splitByDataId":[52,2,1,"_CPPv2N6paddle8Argument13splitByDataIdERKNSt6vectorI8ArgumentEEPNSt6vectorINSt6vectorI8ArgumentEEEE"],"paddle::Argument::strs":[52,3,1,"_CPPv2N6paddle8Argument4strsE"],"paddle::Argument::subArgFrom":[52,2,1,"_CPPv2N6paddle8Argument10subArgFromERK8Argument6size_t6size_t6size_tbbb6size_t6size_t"],"paddle::Argument::subSequenceStartPositions":[52,3,1,"_CPPv2N6paddle8Argument25subSequenceStartPositionsE"],"paddle::Argument::sumCosts":[52,2,1,"_CPPv2N6paddle8Argument8sumCostsERKNSt6vectorI8ArgumentEE"],"paddle::Argument::udp":[52,3,1,"_CPPv2N6paddle8Argument3udpE"],"paddle::Argument::value":[52,3,1,"_CPPv2N6paddle8Argument5valueE"],"paddle::Argument::valueCount":[52,3,1,"_CPPv2N6paddle8Argument10valueCountE"],"paddle::Argument::valueReadyCond":[52,3,1,"_CPPv2N6paddle8Argument14valueReadyCondE"],"paddle::Argument::waitGradReady":[52,2,1,"_CPPv2N6paddle8Argument13waitGradReadyEv"],"paddle::Argument::waitValueReady":[52,2,1,"_CPPv2N6paddle8Argument14waitValueReadyEv"],"paddle::Argument::~Argument":[52,2,1,"_CPPv2N6paddle8ArgumentD0Ev"],"paddle::AsyncParameter":[52,1,1,"_CPPv2N6paddle14AsyncParameterE"],"paddle::AsyncParameter::AsyncParameter":[52,2,1,"_CPPv2N6paddle14AsyncParameter14AsyncParameterE11TrainerRolei12ParameterPtr"],"paddle::AsyncParameter::accumCounter_":[52,3,1,"_CPPv2N6paddle14AsyncParameter13accumCounter_E"],"paddle::AsyncParameter::asyncCount_":[52,3,1,"_CPPv2N6paddle14AsyncParameter11asyncCount_E"],"paddle::AsyncParameter::clearCounter":[52,2,1,"_CPPv2N6paddle14AsyncParameter12clearCounterEv"],"paddle::AsyncParameter::getAccum":[52,2,1,"_CPPv2N6paddle14AsyncParameter8getAccumEv"],"paddle::AsyncParameter::gradientAccum_":[52,3,1,"_CPPv2N6paddle14AsyncParameter14gradientAccum_E"],"paddle::AsyncParameter::masterUpdate":[52,2,1,"_CPPv2N6paddle14AsyncParameter12masterUpdateE20ParallelParameterPtrRK14UpdateCallback"],"paddle::AsyncParameter::slaveUpdate":[52,2,1,"_CPPv2N6paddle14AsyncParameter11slaveUpdateE4real"],"paddle::AsyncParameter::synchronizeParamter":[52,2,1,"_CPPv2N6paddle14AsyncParameter19synchronizeParamterEv"],"paddle::AsyncThreadPool":[64,1,1,"_CPPv2N6paddle15AsyncThreadPoolE"],"paddle::AsyncThreadPool::AsyncThreadPool":[64,2,1,"_CPPv2N6paddle15AsyncThreadPool15AsyncThreadPoolEv"],"paddle::AsyncThreadPool::JobFunc":[64,4,1,"_CPPv2N6paddle15AsyncThreadPool7JobFuncE"],"paddle::AsyncThreadPool::ProcessChannel":[64,2,1,"_CPPv2N6paddle15AsyncThreadPool14ProcessChannelE6size_t"],"paddle::AsyncThreadPool::addBatchJobs":[64,2,1,"_CPPv2N6paddle15AsyncThreadPool12addBatchJobsERKNSt6vectorI1FEERNSt6vectorINSt9result_ofIF1FvEE4typeEEE"],"paddle::AsyncThreadPool::addJob":[64,2,1,"_CPPv2N6paddle15AsyncThreadPool6addJobERR1FDpRR4Args"],"paddle::AsyncThreadPool::run":[64,2,1,"_CPPv2N6paddle15AsyncThreadPool3runEv"],"paddle::AsyncThreadPool::stop":[64,2,1,"_CPPv2N6paddle15AsyncThreadPool4stopEv"],"paddle::AsyncThreadPool::~AsyncThreadPool":[64,2,1,"_CPPv2N6paddle15AsyncThreadPoolD0Ev"],"paddle::AucEvaluator":[38,1,1,"_CPPv2N6paddle12AucEvaluatorE"],"paddle::AucEvaluator::AucEvaluator":[38,2,1,"_CPPv2N6paddle12AucEvaluator12AucEvaluatorE7int32_t"],"paddle::AucEvaluator::distributeEval":[38,2,1,"_CPPv2N6paddle12AucEvaluator14distributeEvalEP16ParameterClient2"],"paddle::AucEvaluator::evalImp":[38,2,1,"_CPPv2N6paddle12AucEvaluator7evalImpERNSt6vectorI8ArgumentEE"],"paddle::AucEvaluator::printStats":[38,2,1,"_CPPv2N6paddle12AucEvaluator10printStatsERNSt7ostreamE"],"paddle::AucEvaluator::start":[38,2,1,"_CPPv2N6paddle12AucEvaluator5startEv"],"paddle::AucValidation":[43,1,1,"_CPPv2N6paddle13AucValidationE"],"paddle::AucValidation::AucValidation":[43,2,1,"_CPPv2N6paddle13AucValidation13AucValidationERK11LayerConfig"],"paddle::AucValidation::PredictionResult":[43,1,1,"_CPPv2N6paddle13AucValidation16PredictionResultE"],"paddle::AucValidation::PredictionResult::PredictionResult":[43,2,1,"_CPPv2N6paddle13AucValidation16PredictionResult16PredictionResultE4reali"],"paddle::AucValidation::PredictionResult::label":[43,3,1,"_CPPv2N6paddle13AucValidation16PredictionResult5labelE"],"paddle::AucValidation::PredictionResult::out":[43,3,1,"_CPPv2N6paddle13AucValidation16PredictionResult3outE"],"paddle::AucValidation::init":[43,2,1,"_CPPv2N6paddle13AucValidation4initERK8LayerMapRK12ParameterMap"],"paddle::AucValidation::onPassEnd":[43,2,1,"_CPPv2N6paddle13AucValidation9onPassEndEv"],"paddle::AucValidation::predictArray_":[43,3,1,"_CPPv2N6paddle13AucValidation13predictArray_E"],"paddle::AucValidation::validationImp":[43,2,1,"_CPPv2N6paddle13AucValidation13validationImpE9MatrixPtr10IVectorPtr"],"paddle::AverageLayer":[43,1,1,"_CPPv2N6paddle12AverageLayerE"],"paddle::AverageLayer::AverageLayer":[43,2,1,"_CPPv2N6paddle12AverageLayer12AverageLayerERK11LayerConfig"],"paddle::AverageLayer::AverageLevel":[43,4,1,"_CPPv2N6paddle12AverageLayer12AverageLevelE"],"paddle::AverageLayer::AverageStrategy":[43,4,1,"_CPPv2N6paddle12AverageLayer15AverageStrategyE"],"paddle::AverageLayer::backward":[43,2,1,"_CPPv2N6paddle12AverageLayer8backwardERK14UpdateCallback"],"paddle::AverageLayer::biases_":[43,3,1,"_CPPv2N6paddle12AverageLayer7biases_E"],"paddle::AverageLayer::dataMtx_":[43,3,1,"_CPPv2N6paddle12AverageLayer8dataMtx_E"],"paddle::AverageLayer::forward":[43,2,1,"_CPPv2N6paddle12AverageLayer7forwardE8PassType"],"paddle::AverageLayer::init":[43,2,1,"_CPPv2N6paddle12AverageLayer4initERK8LayerMapRK12ParameterMap"],"paddle::AverageLayer::kAverage":[43,1,1,"_CPPv2N6paddle12AverageLayer8kAverageE"],"paddle::AverageLayer::kAverageSquareRootN":[43,1,1,"_CPPv2N6paddle12AverageLayer19kAverageSquareRootNE"],"paddle::AverageLayer::kNonSeq":[43,1,1,"_CPPv2N6paddle12AverageLayer7kNonSeqE"],"paddle::AverageLayer::kSeq":[43,1,1,"_CPPv2N6paddle12AverageLayer4kSeqE"],"paddle::AverageLayer::kSum":[43,1,1,"_CPPv2N6paddle12AverageLayer4kSumE"],"paddle::AverageLayer::mode_":[43,3,1,"_CPPv2N6paddle12AverageLayer5mode_E"],"paddle::AverageLayer::outMtx_":[43,3,1,"_CPPv2N6paddle12AverageLayer7outMtx_E"],"paddle::AverageLayer::type_":[43,3,1,"_CPPv2N6paddle12AverageLayer5type_E"],"paddle::AverageLayer::~AverageLayer":[43,2,1,"_CPPv2N6paddle12AverageLayerD0Ev"],"paddle::AverageOptimizer":[50,1,1,"_CPPv2N6paddle16AverageOptimizerE"],"paddle::AverageOptimizer::AverageOptimizer":[50,2,1,"_CPPv2N6paddle16AverageOptimizer16AverageOptimizerERK18OptimizationConfigP18ParameterOptimizerb"],"paddle::AverageOptimizer::apply":[50,2,1,"_CPPv2N6paddle16AverageOptimizer5applyEv"],"paddle::AverageOptimizer::create":[50,2,1,"_CPPv2N6paddle16AverageOptimizer6createERK18OptimizationConfigP18ParameterOptimizerbb"],"paddle::AverageOptimizer::finishBatch":[50,2,1,"_CPPv2N6paddle16AverageOptimizer11finishBatchEv"],"paddle::AverageOptimizer::finishCatchUpWith":[50,2,1,"_CPPv2N6paddle16AverageOptimizer17finishCatchUpWithEv"],"paddle::AverageOptimizer::finishPass":[50,2,1,"_CPPv2N6paddle16AverageOptimizer10finishPassEv"],"paddle::AverageOptimizer::init":[50,2,1,"_CPPv2N6paddle16AverageOptimizer4initE6size_tPK15ParameterConfig"],"paddle::AverageOptimizer::isAverageWindowTooLong":[50,2,1,"_CPPv2N6paddle16AverageOptimizer22isAverageWindowTooLongEv"],"paddle::AverageOptimizer::kMaxNumAccumulates":[50,3,1,"_CPPv2N6paddle16AverageOptimizer18kMaxNumAccumulatesE"],"paddle::AverageOptimizer::maxAverageWindow_":[50,3,1,"_CPPv2N6paddle16AverageOptimizer17maxAverageWindow_E"],"paddle::AverageOptimizer::minAverageWindow_":[50,3,1,"_CPPv2N6paddle16AverageOptimizer17minAverageWindow_E"],"paddle::AverageOptimizer::needSpecialTraversal":[50,2,1,"_CPPv2N6paddle16AverageOptimizer20needSpecialTraversalERK15ParameterConfig"],"paddle::AverageOptimizer::numAccumulates_":[50,3,1,"_CPPv2N6paddle16AverageOptimizer15numAccumulates_E"],"paddle::AverageOptimizer::numUpdates_":[50,3,1,"_CPPv2N6paddle16AverageOptimizer11numUpdates_E"],"paddle::AverageOptimizer::oldNumAccumulates_":[50,3,1,"_CPPv2N6paddle16AverageOptimizer18oldNumAccumulates_E"],"paddle::AverageOptimizer::optimizer_":[50,3,1,"_CPPv2N6paddle16AverageOptimizer10optimizer_E"],"paddle::AverageOptimizer::prevNumUpdates_":[50,3,1,"_CPPv2N6paddle16AverageOptimizer15prevNumUpdates_E"],"paddle::AverageOptimizer::restore":[50,2,1,"_CPPv2N6paddle16AverageOptimizer7restoreEv"],"paddle::AverageOptimizer::setNoDecay":[50,2,1,"_CPPv2N6paddle16AverageOptimizer10setNoDecayEv"],"paddle::AverageOptimizer::startBatch":[50,2,1,"_CPPv2N6paddle16AverageOptimizer10startBatchE7int64_t"],"paddle::AverageOptimizer::startCatchUpWith":[50,2,1,"_CPPv2N6paddle16AverageOptimizer16startCatchUpWithEv"],"paddle::AverageOptimizer::startPass":[50,2,1,"_CPPv2N6paddle16AverageOptimizer9startPassEv"],"paddle::AverageOptimizer::update":[50,2,1,"_CPPv2N6paddle16AverageOptimizer6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::AverageOptimizer::updateAverageWindowLimit":[50,2,1,"_CPPv2N6paddle16AverageOptimizer24updateAverageWindowLimitEv"],"paddle::AverageOptimizer::useApply_":[50,3,1,"_CPPv2N6paddle16AverageOptimizer9useApply_E"],"paddle::AverageSparseOptimizer":[50,1,1,"_CPPv2N6paddle22AverageSparseOptimizerE"],"paddle::AverageSparseOptimizer::AverageSparseOptimizer":[50,2,1,"_CPPv2N6paddle22AverageSparseOptimizer22AverageSparseOptimizerERK18OptimizationConfigP18ParameterOptimizerb"],"paddle::AverageSparseOptimizer::catchUpWith":[50,2,1,"_CPPv2N6paddle22AverageSparseOptimizer11catchUpWithEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::AverageSparseOptimizer::finishBatch":[50,2,1,"_CPPv2N6paddle22AverageSparseOptimizer11finishBatchEv"],"paddle::AverageSparseOptimizer::finishCatchUpWith":[50,2,1,"_CPPv2N6paddle22AverageSparseOptimizer17finishCatchUpWithEv"],"paddle::AverageSparseOptimizer::init":[50,2,1,"_CPPv2N6paddle22AverageSparseOptimizer4initE6size_tPK15ParameterConfig"],"paddle::AverageSparseOptimizer::startCatchUpWith":[50,2,1,"_CPPv2N6paddle22AverageSparseOptimizer16startCatchUpWithEv"],"paddle::AverageSparseOptimizer::t0Vec_":[50,3,1,"_CPPv2N6paddle22AverageSparseOptimizer6t0Vec_E"],"paddle::AverageSparseOptimizer::timer_":[50,3,1,"_CPPv2N6paddle22AverageSparseOptimizer6timer_E"],"paddle::AverageSparseOptimizer::update":[50,2,1,"_CPPv2N6paddle22AverageSparseOptimizer6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::BaseClient":[55,1,1,"_CPPv2N6paddle10BaseClientE"],"paddle::BaseClient::BaseClient":[55,2,1,"_CPPv2N6paddle10BaseClient10BaseClientEbi"],"paddle::BaseClient::InputIovs":[55,4,1,"_CPPv2N6paddle10BaseClient9InputIovsE"],"paddle::BaseClient::SendDataRequestVec":[55,4,1,"_CPPv2N6paddle10BaseClient18SendDataRequestVecE"],"paddle::BaseClient::SendJob":[55,1,1,"_CPPv2N6paddle10BaseClient7SendJobE"],"paddle::BaseClient::SendJob::parallelDataRequests":[55,3,1,"_CPPv2N6paddle10BaseClient7SendJob20parallelDataRequestsE"],"paddle::BaseClient::SendJob::parallelInputIovs":[55,3,1,"_CPPv2N6paddle10BaseClient7SendJob17parallelInputIovsE"],"paddle::BaseClient::SendJob::parallelRequests":[55,3,1,"_CPPv2N6paddle10BaseClient7SendJob16parallelRequestsE"],"paddle::BaseClient::SendJobPtr":[55,4,1,"_CPPv2N6paddle10BaseClient10SendJobPtrE"],"paddle::BaseClient::SendQueue":[55,4,1,"_CPPv2N6paddle10BaseClient9SendQueueE"],"paddle::BaseClient::SendRequest":[55,4,1,"_CPPv2N6paddle10BaseClient11SendRequestE"],"paddle::BaseClient::ThreadPtr":[55,4,1,"_CPPv2N6paddle10BaseClient9ThreadPtrE"],"paddle::BaseClient::calcClientId":[55,2,1,"_CPPv2N6paddle10BaseClient12calcClientIdEii"],"paddle::BaseClient::clients_":[55,3,1,"_CPPv2N6paddle10BaseClient8clients_E"],"paddle::BaseClient::divup":[55,2,1,"_CPPv2N6paddle10BaseClient5divupEii"],"paddle::BaseClient::finishThreads":[55,2,1,"_CPPv2N6paddle10BaseClient13finishThreadsEv"],"paddle::BaseClient::getAllData":[55,2,1,"_CPPv2N6paddle10BaseClient10getAllDataEi12SendDataTypeP8DataType6size_t"],"paddle::BaseClient::getTransDtype":[55,2,1,"_CPPv2N6paddle10BaseClient13getTransDtypeERKNSt9type_infoE"],"paddle::BaseClient::multiCall":[55,2,1,"_CPPv2N6paddle10BaseClient9multiCallEPKcRK7ProtoInPNSt6vectorI8ProtoOutEE"],"paddle::BaseClient::numPorts_":[55,3,1,"_CPPv2N6paddle10BaseClient9numPorts_E"],"paddle::BaseClient::prepareData":[55,2,1,"_CPPv2N6paddle10BaseClient11prepareDataEi12SendDataType14DataUpdateModeP8DataType6size_tP7SendJob"],"paddle::BaseClient::putData":[55,2,1,"_CPPv2N6paddle10BaseClient7putDataEi12SendDataTypeP8DataType6size_t14DataUpdateMode"],"paddle::BaseClient::putOwnData":[55,2,1,"_CPPv2N6paddle10BaseClient10putOwnDataEi12SendDataTypeP8DataType6size_t"],"paddle::BaseClient::recv":[55,2,1,"_CPPv2N6paddle10BaseClient4recvEi"],"paddle::BaseClient::recvData":[55,2,1,"_CPPv2N6paddle10BaseClient8recvDataEv"],"paddle::BaseClient::recvDataMems_":[55,3,1,"_CPPv2N6paddle10BaseClient13recvDataMems_E"],"paddle::BaseClient::recvJobQueue_":[55,3,1,"_CPPv2N6paddle10BaseClient13recvJobQueue_E"],"paddle::BaseClient::recvSyncBarrier_":[55,3,1,"_CPPv2N6paddle10BaseClient16recvSyncBarrier_E"],"paddle::BaseClient::recvThreads_":[55,3,1,"_CPPv2N6paddle10BaseClient12recvThreads_E"],"paddle::BaseClient::reduce":[55,2,1,"_CPPv2N6paddle10BaseClient6reduceEP8DataTypeP8DataType6size_tii"],"paddle::BaseClient::send":[55,2,1,"_CPPv2N6paddle10BaseClient4sendEi"],"paddle::BaseClient::sendData":[55,2,1,"_CPPv2N6paddle10BaseClient8sendDataEi12SendDataType14DataUpdateModeP8DataType6size_t"],"paddle::BaseClient::sendJobQueue_":[55,3,1,"_CPPv2N6paddle10BaseClient13sendJobQueue_E"],"paddle::BaseClient::sendJob_":[55,3,1,"_CPPv2N6paddle10BaseClient8sendJob_E"],"paddle::BaseClient::sendThreads_":[55,3,1,"_CPPv2N6paddle10BaseClient12sendThreads_E"],"paddle::BaseClient::separateSendAndRecv_":[55,3,1,"_CPPv2N6paddle10BaseClient20separateSendAndRecv_E"],"paddle::BaseClient::serviceNum_":[55,3,1,"_CPPv2N6paddle10BaseClient11serviceNum_E"],"paddle::BaseClient::startThreads":[55,2,1,"_CPPv2N6paddle10BaseClient12startThreadsEv"],"paddle::BaseClient::stopping_":[55,3,1,"_CPPv2N6paddle10BaseClient9stopping_E"],"paddle::BaseClient::synchronize":[55,2,1,"_CPPv2N6paddle10BaseClient11synchronizeE10SyncObject"],"paddle::BaseClient::threadNum_":[55,3,1,"_CPPv2N6paddle10BaseClient10threadNum_E"],"paddle::BaseClient::~BaseClient":[55,2,1,"_CPPv2N6paddle10BaseClientD0Ev"],"paddle::BaseMatrix":[46,4,1,"_CPPv2N6paddle10BaseMatrixE"],"paddle::BaseMatrixT":[46,1,1,"_CPPv2N6paddle11BaseMatrixTE"],"paddle::BaseMatrixT::BaseMatrixT":[46,2,1,"_CPPv2N6paddle11BaseMatrixT11BaseMatrixTER11BaseMatrixTb"],"paddle::BaseMatrixT::abs":[46,2,1,"_CPPv2N6paddle11BaseMatrixT3absEv"],"paddle::BaseMatrixT::absDerivative":[46,2,1,"_CPPv2N6paddle11BaseMatrixT13absDerivativeER11BaseMatrixT"],"paddle::BaseMatrixT::add":[46,2,1,"_CPPv2N6paddle11BaseMatrixT3addER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::add2":[46,2,1,"_CPPv2N6paddle11BaseMatrixT4add2ER11BaseMatrixTR11BaseMatrixT1T1T1T"],"paddle::BaseMatrixT::add3":[46,2,1,"_CPPv2N6paddle11BaseMatrixT4add3ER11BaseMatrixTR11BaseMatrixTR11BaseMatrixT1T1T1T"],"paddle::BaseMatrixT::addAtOffset":[46,2,1,"_CPPv2N6paddle11BaseMatrixT11addAtOffsetER11BaseMatrixT7int64_t"],"paddle::BaseMatrixT::addBias":[46,2,1,"_CPPv2N6paddle11BaseMatrixT7addBiasER11BaseMatrixT1T"],"paddle::BaseMatrixT::addColScale":[46,2,1,"_CPPv2N6paddle11BaseMatrixT11addColScaleE6size_tR11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::addColVector":[46,2,1,"_CPPv2N6paddle11BaseMatrixT12addColVectorER11BaseMatrixT"],"paddle::BaseMatrixT::addDotMul":[46,2,1,"_CPPv2N6paddle11BaseMatrixT9addDotMulER11BaseMatrixTR11BaseMatrixT1T1T"],"paddle::BaseMatrixT::addDotMulMMV":[46,2,1,"_CPPv2N6paddle11BaseMatrixT12addDotMulMMVER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::addDotMulMMV2":[46,2,1,"_CPPv2N6paddle11BaseMatrixT13addDotMulMMV2ER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::addDotMulVMM":[46,2,1,"_CPPv2N6paddle11BaseMatrixT12addDotMulVMMER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::addDotMulVMM2":[46,2,1,"_CPPv2N6paddle11BaseMatrixT13addDotMulVMM2ER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::addP2P":[46,2,1,"_CPPv2N6paddle11BaseMatrixT6addP2PER11BaseMatrixT"],"paddle::BaseMatrixT::addRowScale":[46,2,1,"_CPPv2N6paddle11BaseMatrixT11addRowScaleE6size_tR11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::addRowVector":[46,2,1,"_CPPv2N6paddle11BaseMatrixT12addRowVectorER11BaseMatrixT"],"paddle::BaseMatrixT::addScalar":[46,2,1,"_CPPv2N6paddle11BaseMatrixT9addScalarER11BaseMatrixT1T"],"paddle::BaseMatrixT::addSquare":[46,2,1,"_CPPv2N6paddle11BaseMatrixT9addSquareER11BaseMatrixT1T"],"paddle::BaseMatrixT::addSquareSum":[46,2,1,"_CPPv2N6paddle11BaseMatrixT12addSquareSumER11BaseMatrixTR11BaseMatrixT11BaseMatrixT1T1T1T"],"paddle::BaseMatrixT::aggregate":[46,2,1,"_CPPv2N6paddle11BaseMatrixT9aggregateE3Agg2Op5SaverR11BaseMatrixTiiR12MatrixOffset12aAsRowVector12aAsColVector"],"paddle::BaseMatrixT::applyBinary":[46,2,1,"_CPPv2N6paddle11BaseMatrixT11applyBinaryE2OpR11BaseMatrixTiiR12MatrixOffset12bAsRowVector12bAsColVector"],"paddle::BaseMatrixT::applyCol":[46,2,1,"_CPPv2N6paddle11BaseMatrixT8applyColE3AggR11BaseMatrixT"],"paddle::BaseMatrixT::applyL1":[46,2,1,"_CPPv2N6paddle11BaseMatrixT7applyL1ER11BaseMatrixT1T1T"],"paddle::BaseMatrixT::applyL2":[46,2,1,"_CPPv2N6paddle11BaseMatrixT7applyL2ER11BaseMatrixT1T1T"],"paddle::BaseMatrixT::applyQuaternary":[46,2,1,"_CPPv2N6paddle11BaseMatrixT15applyQuaternaryE2OpR11BaseMatrixTR11BaseMatrixTR11BaseMatrixTiiR12MatrixOffset"],"paddle::BaseMatrixT::applyRow":[46,2,1,"_CPPv2N6paddle11BaseMatrixT8applyRowE3AggR11BaseMatrixT"],"paddle::BaseMatrixT::applyTernary":[46,2,1,"_CPPv2N6paddle11BaseMatrixT12applyTernaryE2OpR11BaseMatrixTR11BaseMatrixTiiR12MatrixOffset12cAsRowVector12cAsColVector"],"paddle::BaseMatrixT::applyUnary":[46,2,1,"_CPPv2N6paddle11BaseMatrixT10applyUnaryE2OpiiR12MatrixOffset"],"paddle::BaseMatrixT::assign":[46,2,1,"_CPPv2N6paddle11BaseMatrixT6assignER11BaseMatrixT"],"paddle::BaseMatrixT::assignAtOffset":[46,2,1,"_CPPv2N6paddle11BaseMatrixT14assignAtOffsetER11BaseMatrixT7int64_t"],"paddle::BaseMatrixT::biggerThan":[46,2,1,"_CPPv2N6paddle11BaseMatrixT10biggerThanER11BaseMatrixTR11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::biggerThanScalar":[46,2,1,"_CPPv2N6paddle11BaseMatrixT16biggerThanScalarE1T"],"paddle::BaseMatrixT::binaryClassificationError":[46,2,1,"_CPPv2N6paddle11BaseMatrixT25binaryClassificationErrorE6size_tR11BaseMatrixTR11BaseMatrixT1T"],"paddle::BaseMatrixT::binaryClassificationError2":[46,2,1,"_CPPv2N6paddle11BaseMatrixT26binaryClassificationError2E6size_tR11BaseMatrixTR11BaseMatrixT1T"],"paddle::BaseMatrixT::binaryLabelCrossEntropy":[46,2,1,"_CPPv2N6paddle11BaseMatrixT23binaryLabelCrossEntropyER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::binaryLabelCrossEntropyBp":[46,2,1,"_CPPv2N6paddle11BaseMatrixT25binaryLabelCrossEntropyBpER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::brelu":[46,2,1,"_CPPv2N6paddle11BaseMatrixT5breluER11BaseMatrixT"],"paddle::BaseMatrixT::breluDerivative":[46,2,1,"_CPPv2N6paddle11BaseMatrixT15breluDerivativeER11BaseMatrixT"],"paddle::BaseMatrixT::clip":[46,2,1,"_CPPv2N6paddle11BaseMatrixT4clipE1T1T"],"paddle::BaseMatrixT::colScale":[46,2,1,"_CPPv2N6paddle11BaseMatrixT8colScaleE6size_tR11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::copyAndClear":[46,2,1,"_CPPv2N6paddle11BaseMatrixT12copyAndClearER11BaseMatrixT"],"paddle::BaseMatrixT::data_":[46,3,1,"_CPPv2N6paddle11BaseMatrixT5data_E"],"paddle::BaseMatrixT::decayAddSquare":[46,2,1,"_CPPv2N6paddle11BaseMatrixT14decayAddSquareER11BaseMatrixT1T1T"],"paddle::BaseMatrixT::decayAddSquareMul":[46,2,1,"_CPPv2N6paddle11BaseMatrixT17decayAddSquareMulER11BaseMatrixTR11BaseMatrixT1T1T"],"paddle::BaseMatrixT::divRowVector":[46,2,1,"_CPPv2N6paddle11BaseMatrixT12divRowVectorER11BaseMatrixT"],"paddle::BaseMatrixT::divScalar":[46,2,1,"_CPPv2N6paddle11BaseMatrixT9divScalarER11BaseMatrixT1T"],"paddle::BaseMatrixT::dotDiv":[46,2,1,"_CPPv2N6paddle11BaseMatrixT6dotDivER11BaseMatrixTR11BaseMatrixT1T1T"],"paddle::BaseMatrixT::dotMul":[46,2,1,"_CPPv2N6paddle11BaseMatrixT6dotMulER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::dotMulSquare":[46,2,1,"_CPPv2N6paddle11BaseMatrixT12dotMulSquareER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::dotMulSquareSum":[46,2,1,"_CPPv2N6paddle11BaseMatrixT15dotMulSquareSumER11BaseMatrixTR11BaseMatrixT1T1T"],"paddle::BaseMatrixT::dotMulSum":[46,2,1,"_CPPv2N6paddle11BaseMatrixT9dotMulSumER11BaseMatrixTR11BaseMatrixT1T1T"],"paddle::BaseMatrixT::dotSquareMul":[46,2,1,"_CPPv2N6paddle11BaseMatrixT12dotSquareMulER11BaseMatrixT"],"paddle::BaseMatrixT::dotSquareSquare":[46,2,1,"_CPPv2N6paddle11BaseMatrixT15dotSquareSquareER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::dotSquareSum":[46,2,1,"_CPPv2N6paddle11BaseMatrixT12dotSquareSumER11BaseMatrixTR11BaseMatrixT1T1T"],"paddle::BaseMatrixT::downClip":[46,2,1,"_CPPv2N6paddle11BaseMatrixT8downClipE1T"],"paddle::BaseMatrixT::exp":[46,2,1,"_CPPv2N6paddle11BaseMatrixT3expEv"],"paddle::BaseMatrixT::expDerivative":[46,2,1,"_CPPv2N6paddle11BaseMatrixT13expDerivativeER11BaseMatrixT"],"paddle::BaseMatrixT::height_":[46,3,1,"_CPPv2N6paddle11BaseMatrixT7height_E"],"paddle::BaseMatrixT::invSqrt":[46,2,1,"_CPPv2N6paddle11BaseMatrixT7invSqrtER11BaseMatrixT"],"paddle::BaseMatrixT::isEqualTo":[46,2,1,"_CPPv2N6paddle11BaseMatrixT9isEqualToER11BaseMatrixT1T"],"paddle::BaseMatrixT::isSparse":[46,2,1,"_CPPv2N6paddle11BaseMatrixT8isSparseEv"],"paddle::BaseMatrixT::log":[46,2,1,"_CPPv2N6paddle11BaseMatrixT3logEv"],"paddle::BaseMatrixT::logisticRegressionLoss":[46,2,1,"_CPPv2N6paddle11BaseMatrixT22logisticRegressionLossER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::logisticRegressionLossBp":[46,2,1,"_CPPv2N6paddle11BaseMatrixT24logisticRegressionLossBpER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::max":[46,2,1,"_CPPv2N6paddle11BaseMatrixT3maxER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::maxCols":[46,2,1,"_CPPv2N6paddle11BaseMatrixT7maxColsER11BaseMatrixT"],"paddle::BaseMatrixT::maxRows":[46,2,1,"_CPPv2N6paddle11BaseMatrixT7maxRowsER11BaseMatrixT"],"paddle::BaseMatrixT::minCols":[46,2,1,"_CPPv2N6paddle11BaseMatrixT7minColsER11BaseMatrixT"],"paddle::BaseMatrixT::minRows":[46,2,1,"_CPPv2N6paddle11BaseMatrixT7minRowsER11BaseMatrixT"],"paddle::BaseMatrixT::mulRowVector":[46,2,1,"_CPPv2N6paddle11BaseMatrixT12mulRowVectorER11BaseMatrixT"],"paddle::BaseMatrixT::mulScalar":[46,2,1,"_CPPv2N6paddle11BaseMatrixT9mulScalarER11BaseMatrixT1T"],"paddle::BaseMatrixT::neg":[46,2,1,"_CPPv2N6paddle11BaseMatrixT3negEv"],"paddle::BaseMatrixT::one":[46,2,1,"_CPPv2N6paddle11BaseMatrixT3oneEv"],"paddle::BaseMatrixT::pow":[46,2,1,"_CPPv2N6paddle11BaseMatrixT3powER11BaseMatrixT1T"],"paddle::BaseMatrixT::rankLoss":[46,2,1,"_CPPv2N6paddle11BaseMatrixT8rankLossER11BaseMatrixTR11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::rankLossBp":[46,2,1,"_CPPv2N6paddle11BaseMatrixT10rankLossBpER11BaseMatrixTR11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::reciprocal":[46,2,1,"_CPPv2N6paddle11BaseMatrixT10reciprocalEv"],"paddle::BaseMatrixT::reciprocalDerivative":[46,2,1,"_CPPv2N6paddle11BaseMatrixT20reciprocalDerivativeER11BaseMatrixT"],"paddle::BaseMatrixT::reciprocalSum":[46,2,1,"_CPPv2N6paddle11BaseMatrixT13reciprocalSumER11BaseMatrixTR11BaseMatrixT1T1T1T"],"paddle::BaseMatrixT::relu":[46,2,1,"_CPPv2N6paddle11BaseMatrixT4reluER11BaseMatrixT"],"paddle::BaseMatrixT::reluDerivative":[46,2,1,"_CPPv2N6paddle11BaseMatrixT14reluDerivativeER11BaseMatrixT"],"paddle::BaseMatrixT::rowAdd":[46,2,1,"_CPPv2N6paddle11BaseMatrixT6rowAddE6size_tR11BaseMatrixTR11BaseMatrixT1T"],"paddle::BaseMatrixT::rowBuf":[46,2,1,"_CPPv2N6paddle11BaseMatrixT6rowBufE6size_t"],"paddle::BaseMatrixT::rowDotMul":[46,2,1,"_CPPv2N6paddle11BaseMatrixT9rowDotMulE6size_tR11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::rowDotMul2":[46,2,1,"_CPPv2N6paddle11BaseMatrixT10rowDotMul2E6size_tR11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::rowPow":[46,2,1,"_CPPv2N6paddle11BaseMatrixT6rowPowE6size_tR11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::rowScale":[46,2,1,"_CPPv2N6paddle11BaseMatrixT8rowScaleE6size_tR11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::rowScale2":[46,2,1,"_CPPv2N6paddle11BaseMatrixT9rowScale2E6size_tR11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::scalarDiv":[46,2,1,"_CPPv2N6paddle11BaseMatrixT9scalarDivER11BaseMatrixT1T"],"paddle::BaseMatrixT::scaledTanh":[46,2,1,"_CPPv2N6paddle11BaseMatrixT10scaledTanhER11BaseMatrixT1T1T"],"paddle::BaseMatrixT::scaledTanhDerivative":[46,2,1,"_CPPv2N6paddle11BaseMatrixT20scaledTanhDerivativeER11BaseMatrixT1T1T"],"paddle::BaseMatrixT::setData":[46,2,1,"_CPPv2N6paddle11BaseMatrixT7setDataEP1T"],"paddle::BaseMatrixT::sgdUpdate":[46,2,1,"_CPPv2N6paddle11BaseMatrixT9sgdUpdateER11BaseMatrixTR11BaseMatrixTR11BaseMatrixT1T1T1T"],"paddle::BaseMatrixT::sigmoid":[46,2,1,"_CPPv2N6paddle11BaseMatrixT7sigmoidER11BaseMatrixT"],"paddle::BaseMatrixT::sigmoidDerivative":[46,2,1,"_CPPv2N6paddle11BaseMatrixT17sigmoidDerivativeER11BaseMatrixT"],"paddle::BaseMatrixT::sign":[46,2,1,"_CPPv2N6paddle11BaseMatrixT4signEv"],"paddle::BaseMatrixT::softCrossEntropy":[46,2,1,"_CPPv2N6paddle11BaseMatrixT16softCrossEntropyER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::softCrossEntropyBp":[46,2,1,"_CPPv2N6paddle11BaseMatrixT18softCrossEntropyBpER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::softrelu":[46,2,1,"_CPPv2N6paddle11BaseMatrixT8softreluER11BaseMatrixT"],"paddle::BaseMatrixT::softreluDerivative":[46,2,1,"_CPPv2N6paddle11BaseMatrixT18softreluDerivativeER11BaseMatrixT"],"paddle::BaseMatrixT::sqrt":[46,2,1,"_CPPv2N6paddle11BaseMatrixT4sqrtEv"],"paddle::BaseMatrixT::square":[46,2,1,"_CPPv2N6paddle11BaseMatrixT6squareEv"],"paddle::BaseMatrixT::squareDerivative":[46,2,1,"_CPPv2N6paddle11BaseMatrixT16squareDerivativeER11BaseMatrixT"],"paddle::BaseMatrixT::stride_":[46,3,1,"_CPPv2N6paddle11BaseMatrixT7stride_E"],"paddle::BaseMatrixT::sub":[46,2,1,"_CPPv2N6paddle11BaseMatrixT3subER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::subScalar":[46,2,1,"_CPPv2N6paddle11BaseMatrixT9subScalarER11BaseMatrixT1T"],"paddle::BaseMatrixT::sumCols":[46,2,1,"_CPPv2N6paddle11BaseMatrixT7sumColsER11BaseMatrixT1T"],"paddle::BaseMatrixT::sumOfSquares":[46,2,1,"_CPPv2N6paddle11BaseMatrixT12sumOfSquaresER11BaseMatrixTR11BaseMatrixT"],"paddle::BaseMatrixT::sumRows":[46,2,1,"_CPPv2N6paddle11BaseMatrixT7sumRowsER11BaseMatrixT"],"paddle::BaseMatrixT::tanh":[46,2,1,"_CPPv2N6paddle11BaseMatrixT4tanhER11BaseMatrixT"],"paddle::BaseMatrixT::tanhDerivative":[46,2,1,"_CPPv2N6paddle11BaseMatrixT14tanhDerivativeER11BaseMatrixT"],"paddle::BaseMatrixT::trans_":[46,3,1,"_CPPv2N6paddle11BaseMatrixT6trans_E"],"paddle::BaseMatrixT::useGpu":[46,2,1,"_CPPv2N6paddle11BaseMatrixT6useGpuEv"],"paddle::BaseMatrixT::useGpu_":[46,3,1,"_CPPv2N6paddle11BaseMatrixT7useGpu_E"],"paddle::BaseMatrixT::width_":[46,3,1,"_CPPv2N6paddle11BaseMatrixT6width_E"],"paddle::BaseMatrixT::zero":[46,2,1,"_CPPv2N6paddle11BaseMatrixT4zeroEv"],"paddle::BaseMatrixT::zeroAtOffset":[46,2,1,"_CPPv2N6paddle11BaseMatrixT12zeroAtOffsetE7int64_t7int64_t"],"paddle::BaseMatrixT::~BaseMatrixT":[46,2,1,"_CPPv2N6paddle11BaseMatrixTD0Ev"],"paddle::BaseVector":[46,1,1,"_CPPv2N6paddle10BaseVectorE"],"paddle::BaseVector::BaseVector":[46,2,1,"_CPPv2N6paddle10BaseVector10BaseVectorE6size_tP1Tb"],"paddle::BaseVector::size_":[46,3,1,"_CPPv2N6paddle10BaseVector5size_E"],"paddle::BaseVector::~BaseVector":[46,2,1,"_CPPv2N6paddle10BaseVectorD0Ev"],"paddle::BatchNormBaseLayer":[43,1,1,"_CPPv2N6paddle18BatchNormBaseLayerE"],"paddle::BatchNormBaseLayer::BatchNormBaseLayer":[43,2,1,"_CPPv2N6paddle18BatchNormBaseLayer18BatchNormBaseLayerERK11LayerConfig"],"paddle::BatchNormBaseLayer::biases_":[43,3,1,"_CPPv2N6paddle18BatchNormBaseLayer7biases_E"],"paddle::BatchNormBaseLayer::calFeatureMapSize":[43,2,1,"_CPPv2N6paddle18BatchNormBaseLayer17calFeatureMapSizeEv"],"paddle::BatchNormBaseLayer::channels_":[43,3,1,"_CPPv2N6paddle18BatchNormBaseLayer9channels_E"],"paddle::BatchNormBaseLayer::create":[43,2,1,"_CPPv2N6paddle18BatchNormBaseLayer6createERK11LayerConfig"],"paddle::BatchNormBaseLayer::imageH_":[43,3,1,"_CPPv2N6paddle18BatchNormBaseLayer7imageH_E"],"paddle::BatchNormBaseLayer::imageW_":[43,3,1,"_CPPv2N6paddle18BatchNormBaseLayer7imageW_E"],"paddle::BatchNormBaseLayer::imgPixels_":[43,3,1,"_CPPv2N6paddle18BatchNormBaseLayer10imgPixels_E"],"paddle::BatchNormBaseLayer::imgSize_":[43,3,1,"_CPPv2N6paddle18BatchNormBaseLayer8imgSize_E"],"paddle::BatchNormBaseLayer::init":[43,2,1,"_CPPv2N6paddle18BatchNormBaseLayer4initERK8LayerMapRK12ParameterMap"],"paddle::BatchNormBaseLayer::movingAvgFraction_":[43,3,1,"_CPPv2N6paddle18BatchNormBaseLayer18movingAvgFraction_E"],"paddle::BatchNormBaseLayer::movingMean_":[43,3,1,"_CPPv2N6paddle18BatchNormBaseLayer11movingMean_E"],"paddle::BatchNormBaseLayer::movingVar_":[43,3,1,"_CPPv2N6paddle18BatchNormBaseLayer10movingVar_E"],"paddle::BatchNormBaseLayer::savedInvVar_":[43,3,1,"_CPPv2N6paddle18BatchNormBaseLayer12savedInvVar_E"],"paddle::BatchNormBaseLayer::savedMean_":[43,3,1,"_CPPv2N6paddle18BatchNormBaseLayer10savedMean_E"],"paddle::BatchNormBaseLayer::useGlobalStats_":[43,3,1,"_CPPv2N6paddle18BatchNormBaseLayer15useGlobalStats_E"],"paddle::BatchNormBaseLayer::weight_":[43,3,1,"_CPPv2N6paddle18BatchNormBaseLayer7weight_E"],"paddle::BatchNormBaseLayer::~BatchNormBaseLayer":[43,2,1,"_CPPv2N6paddle18BatchNormBaseLayerD0Ev"],"paddle::BatchNormalizationLayer":[43,1,1,"_CPPv2N6paddle23BatchNormalizationLayerE"],"paddle::BatchNormalizationLayer::BatchNormalizationLayer":[43,2,1,"_CPPv2N6paddle23BatchNormalizationLayer23BatchNormalizationLayerERK11LayerConfig"],"paddle::BatchNormalizationLayer::EPS":[43,3,1,"_CPPv2N6paddle23BatchNormalizationLayer3EPSE"],"paddle::BatchNormalizationLayer::backward":[43,2,1,"_CPPv2N6paddle23BatchNormalizationLayer8backwardERK14UpdateCallback"],"paddle::BatchNormalizationLayer::calMeanAndStd":[43,2,1,"_CPPv2N6paddle23BatchNormalizationLayer13calMeanAndStdERK9MatrixPtr"],"paddle::BatchNormalizationLayer::calMovingMeanAndVar":[43,2,1,"_CPPv2N6paddle23BatchNormalizationLayer19calMovingMeanAndVarEv"],"paddle::BatchNormalizationLayer::expandMat":[43,2,1,"_CPPv2N6paddle23BatchNormalizationLayer9expandMatERK9MatrixPtrR9MatrixPtr"],"paddle::BatchNormalizationLayer::expandedInGrad_":[43,3,1,"_CPPv2N6paddle23BatchNormalizationLayer15expandedInGrad_E"],"paddle::BatchNormalizationLayer::expandedIn_":[43,3,1,"_CPPv2N6paddle23BatchNormalizationLayer11expandedIn_E"],"paddle::BatchNormalizationLayer::expandedOutGrad_":[43,3,1,"_CPPv2N6paddle23BatchNormalizationLayer16expandedOutGrad_E"],"paddle::BatchNormalizationLayer::expandedOut_":[43,3,1,"_CPPv2N6paddle23BatchNormalizationLayer12expandedOut_E"],"paddle::BatchNormalizationLayer::firstTest_":[43,3,1,"_CPPv2N6paddle23BatchNormalizationLayer10firstTest_E"],"paddle::BatchNormalizationLayer::forward":[43,2,1,"_CPPv2N6paddle23BatchNormalizationLayer7forwardE8PassType"],"paddle::BatchNormalizationLayer::inGrad_":[43,3,1,"_CPPv2N6paddle23BatchNormalizationLayer7inGrad_E"],"paddle::BatchNormalizationLayer::init":[43,2,1,"_CPPv2N6paddle23BatchNormalizationLayer4initERK8LayerMapRK12ParameterMap"],"paddle::BatchNormalizationLayer::meanGrad_":[43,3,1,"_CPPv2N6paddle23BatchNormalizationLayer9meanGrad_E"],"paddle::BatchNormalizationLayer::normInGrad_":[43,3,1,"_CPPv2N6paddle23BatchNormalizationLayer11normInGrad_E"],"paddle::BatchNormalizationLayer::normIn_":[43,3,1,"_CPPv2N6paddle23BatchNormalizationLayer7normIn_E"],"paddle::BatchNormalizationLayer::setMeanAndStd":[43,2,1,"_CPPv2N6paddle23BatchNormalizationLayer13setMeanAndStdEv"],"paddle::BatchNormalizationLayer::shrinkMat":[43,2,1,"_CPPv2N6paddle23BatchNormalizationLayer9shrinkMatERK9MatrixPtrR9MatrixPtr"],"paddle::BatchNormalizationLayer::stdGrad_":[43,3,1,"_CPPv2N6paddle23BatchNormalizationLayer8stdGrad_E"],"paddle::BatchNormalizationLayer::tmpGrad_":[43,3,1,"_CPPv2N6paddle23BatchNormalizationLayer8tmpGrad_E"],"paddle::BatchNormalizationLayer::tmpMat_":[43,3,1,"_CPPv2N6paddle23BatchNormalizationLayer7tmpMat_E"],"paddle::BatchNormalizationLayer::~BatchNormalizationLayer":[43,2,1,"_CPPv2N6paddle23BatchNormalizationLayerD0Ev"],"paddle::BlockExpandLayer":[43,1,1,"_CPPv2N6paddle16BlockExpandLayerE"],"paddle::BlockExpandLayer::BlockExpandLayer":[43,2,1,"_CPPv2N6paddle16BlockExpandLayer16BlockExpandLayerERK11LayerConfig"],"paddle::BlockExpandLayer::backward":[43,2,1,"_CPPv2N6paddle16BlockExpandLayer8backwardERK14UpdateCallback"],"paddle::BlockExpandLayer::blockH_":[43,3,1,"_CPPv2N6paddle16BlockExpandLayer7blockH_E"],"paddle::BlockExpandLayer::blockW_":[43,3,1,"_CPPv2N6paddle16BlockExpandLayer7blockW_E"],"paddle::BlockExpandLayer::channels_":[43,3,1,"_CPPv2N6paddle16BlockExpandLayer9channels_E"],"paddle::BlockExpandLayer::forward":[43,2,1,"_CPPv2N6paddle16BlockExpandLayer7forwardE8PassType"],"paddle::BlockExpandLayer::getBlockNum":[43,2,1,"_CPPv2N6paddle16BlockExpandLayer11getBlockNumEv"],"paddle::BlockExpandLayer::imgSizeH_":[43,3,1,"_CPPv2N6paddle16BlockExpandLayer9imgSizeH_E"],"paddle::BlockExpandLayer::imgSizeW_":[43,3,1,"_CPPv2N6paddle16BlockExpandLayer9imgSizeW_E"],"paddle::BlockExpandLayer::init":[43,2,1,"_CPPv2N6paddle16BlockExpandLayer4initERK8LayerMapRK12ParameterMap"],"paddle::BlockExpandLayer::outVTrans_":[43,3,1,"_CPPv2N6paddle16BlockExpandLayer10outVTrans_E"],"paddle::BlockExpandLayer::outputH_":[43,3,1,"_CPPv2N6paddle16BlockExpandLayer8outputH_E"],"paddle::BlockExpandLayer::outputW_":[43,3,1,"_CPPv2N6paddle16BlockExpandLayer8outputW_E"],"paddle::BlockExpandLayer::paddingH_":[43,3,1,"_CPPv2N6paddle16BlockExpandLayer9paddingH_E"],"paddle::BlockExpandLayer::paddingW_":[43,3,1,"_CPPv2N6paddle16BlockExpandLayer9paddingW_E"],"paddle::BlockExpandLayer::strideH_":[43,3,1,"_CPPv2N6paddle16BlockExpandLayer8strideH_E"],"paddle::BlockExpandLayer::strideW_":[43,3,1,"_CPPv2N6paddle16BlockExpandLayer8strideW_E"],"paddle::BlockExpandLayer::~BlockExpandLayer":[43,2,1,"_CPPv2N6paddle16BlockExpandLayerD0Ev"],"paddle::BlockingQueue":[65,1,1,"_CPPv2N6paddle13BlockingQueueE"],"paddle::BlockingQueue::BlockingQueue":[65,2,1,"_CPPv2N6paddle13BlockingQueue13BlockingQueueE6size_t"],"paddle::BlockingQueue::dequeue":[65,2,1,"_CPPv2N6paddle13BlockingQueue7dequeueEv"],"paddle::BlockingQueue::empty":[65,2,1,"_CPPv2N6paddle13BlockingQueue5emptyEv"],"paddle::BlockingQueue::enqueue":[65,2,1,"_CPPv2N6paddle13BlockingQueue7enqueueERK1T"],"paddle::BlockingQueue::size":[65,2,1,"_CPPv2N6paddle13BlockingQueue4sizeEv"],"paddle::CMRProjectionNormLayer":[43,1,1,"_CPPv2N6paddle22CMRProjectionNormLayerE"],"paddle::CMRProjectionNormLayer::CMRProjectionNormLayer":[43,2,1,"_CPPv2N6paddle22CMRProjectionNormLayer22CMRProjectionNormLayerERK11LayerConfig"],"paddle::CMRProjectionNormLayer::backward":[43,2,1,"_CPPv2N6paddle22CMRProjectionNormLayer8backwardERK14UpdateCallback"],"paddle::CMRProjectionNormLayer::forward":[43,2,1,"_CPPv2N6paddle22CMRProjectionNormLayer7forwardE8PassType"],"paddle::CMRProjectionNormLayer::getSize":[43,2,1,"_CPPv2N6paddle22CMRProjectionNormLayer7getSizeEv"],"paddle::CMRProjectionNormLayer::init":[43,2,1,"_CPPv2N6paddle22CMRProjectionNormLayer4initERK8LayerMapRK12ParameterMap"],"paddle::CMRProjectionNormLayer::~CMRProjectionNormLayer":[43,2,1,"_CPPv2N6paddle22CMRProjectionNormLayerD0Ev"],"paddle::CRFDecodingLayer":[43,1,1,"_CPPv2N6paddle16CRFDecodingLayerE"],"paddle::CRFDecodingLayer::CRFDecodingLayer":[43,2,1,"_CPPv2N6paddle16CRFDecodingLayer16CRFDecodingLayerERK11LayerConfig"],"paddle::CRFDecodingLayer::backward":[43,2,1,"_CPPv2N6paddle16CRFDecodingLayer8backwardERK14UpdateCallback"],"paddle::CRFDecodingLayer::crf_":[43,3,1,"_CPPv2N6paddle16CRFDecodingLayer4crf_E"],"paddle::CRFDecodingLayer::forward":[43,2,1,"_CPPv2N6paddle16CRFDecodingLayer7forwardE8PassType"],"paddle::CRFDecodingLayer::init":[43,2,1,"_CPPv2N6paddle16CRFDecodingLayer4initERK8LayerMapRK12ParameterMap"],"paddle::CRFLayer":[43,1,1,"_CPPv2N6paddle8CRFLayerE"],"paddle::CRFLayer::CRFLayer":[43,2,1,"_CPPv2N6paddle8CRFLayer8CRFLayerERK11LayerConfig"],"paddle::CRFLayer::backward":[43,2,1,"_CPPv2N6paddle8CRFLayer8backwardERK14UpdateCallback"],"paddle::CRFLayer::coeff_":[43,3,1,"_CPPv2N6paddle8CRFLayer6coeff_E"],"paddle::CRFLayer::crfs_":[43,3,1,"_CPPv2N6paddle8CRFLayer5crfs_E"],"paddle::CRFLayer::forward":[43,2,1,"_CPPv2N6paddle8CRFLayer7forwardE8PassType"],"paddle::CRFLayer::init":[43,2,1,"_CPPv2N6paddle8CRFLayer4initERK8LayerMapRK12ParameterMap"],"paddle::CRFLayer::numClasses_":[43,3,1,"_CPPv2N6paddle8CRFLayer11numClasses_E"],"paddle::CRFLayer::parameter_":[43,3,1,"_CPPv2N6paddle8CRFLayer10parameter_E"],"paddle::CRFLayer::weightLayer_":[43,3,1,"_CPPv2N6paddle8CRFLayer12weightLayer_E"],"paddle::CTCErrorEvaluator":[38,1,1,"_CPPv2N6paddle17CTCErrorEvaluatorE"],"paddle::CTCErrorEvaluator::CTCErrorEvaluator":[38,2,1,"_CPPv2N6paddle17CTCErrorEvaluator17CTCErrorEvaluatorEv"],"paddle::CTCErrorEvaluator::distributeEval":[38,2,1,"_CPPv2N6paddle17CTCErrorEvaluator14distributeEvalEP16ParameterClient2"],"paddle::CTCErrorEvaluator::eval":[38,2,1,"_CPPv2N6paddle17CTCErrorEvaluator4evalERK13NeuralNetwork"],"paddle::CTCErrorEvaluator::evalImp":[38,2,1,"_CPPv2N6paddle17CTCErrorEvaluator7evalImpERNSt6vectorI8ArgumentEE"],"paddle::CTCErrorEvaluator::printStats":[38,2,1,"_CPPv2N6paddle17CTCErrorEvaluator10printStatsERNSt7ostreamE"],"paddle::CTCErrorEvaluator::start":[38,2,1,"_CPPv2N6paddle17CTCErrorEvaluator5startEv"],"paddle::CTCLayer":[43,1,1,"_CPPv2N6paddle8CTCLayerE"],"paddle::CTCLayer::CTCLayer":[43,2,1,"_CPPv2N6paddle8CTCLayer8CTCLayerERK11LayerConfig"],"paddle::CTCLayer::backward":[43,2,1,"_CPPv2N6paddle8CTCLayer8backwardERK14UpdateCallback"],"paddle::CTCLayer::backwardImp":[43,2,1,"_CPPv2N6paddle8CTCLayer11backwardImpERK14UpdateCallbackRK8ArgumentRK8Argument"],"paddle::CTCLayer::ctcs_":[43,3,1,"_CPPv2N6paddle8CTCLayer5ctcs_E"],"paddle::CTCLayer::forward":[43,2,1,"_CPPv2N6paddle8CTCLayer7forwardE8PassType"],"paddle::CTCLayer::forwardImp":[43,2,1,"_CPPv2N6paddle8CTCLayer10forwardImpERK8ArgumentRK8Argument"],"paddle::CTCLayer::init":[43,2,1,"_CPPv2N6paddle8CTCLayer4initERK8LayerMapRK12ParameterMap"],"paddle::CTCLayer::normByTimes_":[43,3,1,"_CPPv2N6paddle8CTCLayer12normByTimes_E"],"paddle::CTCLayer::numClasses_":[43,3,1,"_CPPv2N6paddle8CTCLayer11numClasses_E"],"paddle::CTCLayer::tmpCpuInput_":[43,3,1,"_CPPv2N6paddle8CTCLayer12tmpCpuInput_E"],"paddle::CacheOnePassInMemory":[36,1,1,"_CPPv2N6paddle20CacheOnePassInMemoryE"],"paddle::CacheOnePassInMemory::CacheOnePassInMemory":[36,2,1,"_CPPv2N6paddle20CacheOnePassInMemory20CacheOnePassInMemoryEv"],"paddle::CacheOnePassInMemory::drop":[36,2,1,"_CPPv2N6paddle20CacheOnePassInMemory4dropEPNSt5dequeI11PyObjectPtrEE"],"paddle::CacheOnePassInMemory::load":[36,2,1,"_CPPv2N6paddle20CacheOnePassInMemory4loadEv"],"paddle::CacheOnePassInMemory::reset":[36,2,1,"_CPPv2N6paddle20CacheOnePassInMemory5resetEv"],"paddle::CacheRowCpuMatrix":[46,1,1,"_CPPv2N6paddle17CacheRowCpuMatrixE"],"paddle::CacheRowCpuMatrix::CacheRowCpuMatrix":[46,2,1,"_CPPv2N6paddle17CacheRowCpuMatrix17CacheRowCpuMatrixE6size_t6size_t12IndexDictPtrb"],"paddle::CacheRowCpuMatrix::getRow":[46,2,1,"_CPPv2N6paddle17CacheRowCpuMatrix6getRowE6size_t"],"paddle::CacheRowCpuMatrix::getRowBuf":[46,2,1,"_CPPv2N6paddle17CacheRowCpuMatrix9getRowBufE6size_t"],"paddle::CacheRowCpuMatrix::mul":[46,2,1,"_CPPv2N6paddle17CacheRowCpuMatrix3mulEP15CpuSparseMatrixP9CpuMatrix4real4real"],"paddle::CacheRowCpuMatrix::setSourceData":[46,2,1,"_CPPv2N6paddle17CacheRowCpuMatrix13setSourceDataE12CpuVectorPtr"],"paddle::CacheRowCpuMatrix::sourceDataVec_":[46,3,1,"_CPPv2N6paddle17CacheRowCpuMatrix14sourceDataVec_E"],"paddle::CacheRowCpuMatrix::sourceData_":[46,3,1,"_CPPv2N6paddle17CacheRowCpuMatrix11sourceData_E"],"paddle::ChunkEvaluator":[38,1,1,"_CPPv2N6paddle14ChunkEvaluatorE"],"paddle::ChunkEvaluator::distributeEval":[38,2,1,"_CPPv2N6paddle14ChunkEvaluator14distributeEvalEP16ParameterClient2"],"paddle::ChunkEvaluator::eval1":[38,2,1,"_CPPv2N6paddle14ChunkEvaluator5eval1EPiPii"],"paddle::ChunkEvaluator::evalImp":[38,2,1,"_CPPv2N6paddle14ChunkEvaluator7evalImpERNSt6vectorI8ArgumentEE"],"paddle::ChunkEvaluator::getSegments":[38,2,1,"_CPPv2N6paddle14ChunkEvaluator11getSegmentsEPiiRNSt6vectorI7SegmentEE"],"paddle::ChunkEvaluator::init":[38,2,1,"_CPPv2N6paddle14ChunkEvaluator4initERK15EvaluatorConfig"],"paddle::ChunkEvaluator::isChunkBegin":[38,2,1,"_CPPv2N6paddle14ChunkEvaluator12isChunkBeginEiiii"],"paddle::ChunkEvaluator::isChunkEnd":[38,2,1,"_CPPv2N6paddle14ChunkEvaluator10isChunkEndEiiii"],"paddle::ChunkEvaluator::printStats":[38,2,1,"_CPPv2N6paddle14ChunkEvaluator10printStatsERNSt7ostreamE"],"paddle::ChunkEvaluator::start":[38,2,1,"_CPPv2N6paddle14ChunkEvaluator5startEv"],"paddle::ClassificationErrorEvaluator":[38,1,1,"_CPPv2N6paddle28ClassificationErrorEvaluatorE"],"paddle::ClassificationErrorEvaluator::calcError":[38,2,1,"_CPPv2N6paddle28ClassificationErrorEvaluator9calcErrorERNSt6vectorI8ArgumentEE"],"paddle::ClassificationErrorEvaluator::distributeEval":[38,2,1,"_CPPv2N6paddle28ClassificationErrorEvaluator14distributeEvalEP16ParameterClient2"],"paddle::ClassificationErrorEvaluator::evalImp":[38,2,1,"_CPPv2N6paddle28ClassificationErrorEvaluator7evalImpERNSt6vectorI8ArgumentEE"],"paddle::ClassificationErrorEvaluator::updateSamplesNum":[38,2,1,"_CPPv2N6paddle28ClassificationErrorEvaluator16updateSamplesNumERKNSt6vectorI8ArgumentEE"],"paddle::ClassificationErrorPrinter":[38,1,1,"_CPPv2N6paddle26ClassificationErrorPrinterE"],"paddle::ClassificationErrorPrinter::evalImp":[38,2,1,"_CPPv2N6paddle26ClassificationErrorPrinter7evalImpERNSt6vectorI8ArgumentEE"],"paddle::ClassificationErrorPrinter::updateSamplesNum":[38,2,1,"_CPPv2N6paddle26ClassificationErrorPrinter16updateSamplesNumERKNSt6vectorI8ArgumentEE"],"paddle::ColumnSumEvaluator":[38,1,1,"_CPPv2N6paddle18ColumnSumEvaluatorE"],"paddle::ColumnSumEvaluator::ColumnSumEvaluator":[38,2,1,"_CPPv2N6paddle18ColumnSumEvaluator18ColumnSumEvaluatorE7int32_t"],"paddle::ColumnSumEvaluator::distributeEval":[38,2,1,"_CPPv2N6paddle18ColumnSumEvaluator14distributeEvalEP16ParameterClient2"],"paddle::ColumnSumEvaluator::evalImp":[38,2,1,"_CPPv2N6paddle18ColumnSumEvaluator7evalImpERNSt6vectorI8ArgumentEE"],"paddle::ColumnSumEvaluator::printStats":[38,2,1,"_CPPv2N6paddle18ColumnSumEvaluator10printStatsERNSt7ostreamE"],"paddle::ColumnSumEvaluator::start":[38,2,1,"_CPPv2N6paddle18ColumnSumEvaluator5startEv"],"paddle::ColumnSumEvaluator::updateSamplesNum":[38,2,1,"_CPPv2N6paddle18ColumnSumEvaluator16updateSamplesNumERKNSt6vectorI8ArgumentEE"],"paddle::ConcatenateLayer":[43,1,1,"_CPPv2N6paddle16ConcatenateLayerE"],"paddle::ConcatenateLayer2":[43,1,1,"_CPPv2N6paddle17ConcatenateLayer2E"],"paddle::ConcatenateLayer2::ConcatenateLayer2":[43,2,1,"_CPPv2N6paddle17ConcatenateLayer217ConcatenateLayer2ERK11LayerConfig"],"paddle::ConcatenateLayer2::backward":[43,2,1,"_CPPv2N6paddle17ConcatenateLayer28backwardERK14UpdateCallback"],"paddle::ConcatenateLayer2::forward":[43,2,1,"_CPPv2N6paddle17ConcatenateLayer27forwardE8PassType"],"paddle::ConcatenateLayer2::init":[43,2,1,"_CPPv2N6paddle17ConcatenateLayer24initERK8LayerMapRK12ParameterMap"],"paddle::ConcatenateLayer2::projCol_":[43,3,1,"_CPPv2N6paddle17ConcatenateLayer28projCol_E"],"paddle::ConcatenateLayer2::projOutput_":[43,3,1,"_CPPv2N6paddle17ConcatenateLayer211projOutput_E"],"paddle::ConcatenateLayer2::projections_":[43,3,1,"_CPPv2N6paddle17ConcatenateLayer212projections_E"],"paddle::ConcatenateLayer2::~ConcatenateLayer2":[43,2,1,"_CPPv2N6paddle17ConcatenateLayer2D0Ev"],"paddle::ConcatenateLayer::ConcatenateLayer":[43,2,1,"_CPPv2N6paddle16ConcatenateLayer16ConcatenateLayerERK11LayerConfig"],"paddle::ConcatenateLayer::backward":[43,2,1,"_CPPv2N6paddle16ConcatenateLayer8backwardERK14UpdateCallback"],"paddle::ConcatenateLayer::forward":[43,2,1,"_CPPv2N6paddle16ConcatenateLayer7forwardE8PassType"],"paddle::ConcatenateLayer::init":[43,2,1,"_CPPv2N6paddle16ConcatenateLayer4initERK8LayerMapRK12ParameterMap"],"paddle::ConcatenateLayer::~ConcatenateLayer":[43,2,1,"_CPPv2N6paddle16ConcatenateLayerD0Ev"],"paddle::ConcurrentRemoteParameterUpdater":[61,1,1,"_CPPv2N6paddle32ConcurrentRemoteParameterUpdaterE"],"paddle::ConcurrentRemoteParameterUpdater::ConcurrentRemoteParameterUpdater":[61,2,1,"_CPPv2N6paddle32ConcurrentRemoteParameterUpdater32ConcurrentRemoteParameterUpdaterE18OptimizationConfigiRRNSt10unique_ptrI16ParameterUpdaterEE"],"paddle::ConcurrentRemoteParameterUpdater::copySingleParaFromDevice":[61,2,1,"_CPPv2N6paddle32ConcurrentRemoteParameterUpdater24copySingleParaFromDeviceEP9Parameter13ParameterType"],"paddle::ConcurrentRemoteParameterUpdater::copySingleParaToDevice":[61,2,1,"_CPPv2N6paddle32ConcurrentRemoteParameterUpdater22copySingleParaToDeviceEP9Parameter13ParameterType"],"paddle::ConcurrentRemoteParameterUpdater::finishBatch":[61,2,1,"_CPPv2N6paddle32ConcurrentRemoteParameterUpdater11finishBatchE4real"],"paddle::ConcurrentRemoteParameterUpdater::needToUpdateRemotely":[61,2,1,"_CPPv2N6paddle32ConcurrentRemoteParameterUpdater20needToUpdateRemotelyEv"],"paddle::ConcurrentRemoteParameterUpdater::recv":[61,2,1,"_CPPv2N6paddle32ConcurrentRemoteParameterUpdater4recvEv"],"paddle::ConcurrentRemoteParameterUpdater::send":[61,2,1,"_CPPv2N6paddle32ConcurrentRemoteParameterUpdater4sendEv"],"paddle::ConcurrentRemoteParameterUpdater::updateImpl":[61,2,1,"_CPPv2N6paddle32ConcurrentRemoteParameterUpdater10updateImplEP9Parameter"],"paddle::ConcurrentRemoteParameterUpdater::~ConcurrentRemoteParameterUpdater":[61,2,1,"_CPPv2N6paddle32ConcurrentRemoteParameterUpdaterD0Ev"],"paddle::ContextProjection":[43,1,1,"_CPPv2N6paddle17ContextProjectionE"],"paddle::ContextProjection::ContextProjection":[43,2,1,"_CPPv2N6paddle17ContextProjection17ContextProjectionERK16ProjectionConfig12ParameterPtrb"],"paddle::ContextProjection::backward":[43,2,1,"_CPPv2N6paddle17ContextProjection8backwardERK14UpdateCallback"],"paddle::ContextProjection::beginPad_":[43,3,1,"_CPPv2N6paddle17ContextProjection9beginPad_E"],"paddle::ContextProjection::endPad_":[43,3,1,"_CPPv2N6paddle17ContextProjection7endPad_E"],"paddle::ContextProjection::forward":[43,2,1,"_CPPv2N6paddle17ContextProjection7forwardEv"],"paddle::ContextProjection::getState":[43,2,1,"_CPPv2N6paddle17ContextProjection8getStateEv"],"paddle::ContextProjection::resetState":[43,2,1,"_CPPv2N6paddle17ContextProjection10resetStateEv"],"paddle::ContextProjection::setState":[43,2,1,"_CPPv2N6paddle17ContextProjection8setStateE13LayerStatePtr"],"paddle::ContextProjection::state2_":[43,3,1,"_CPPv2N6paddle17ContextProjection7state2_E"],"paddle::ContextProjection::state_":[43,3,1,"_CPPv2N6paddle17ContextProjection6state_E"],"paddle::ContextProjection::weight_":[43,3,1,"_CPPv2N6paddle17ContextProjection7weight_E"],"paddle::ConvBaseLayer":[43,1,1,"_CPPv2N6paddle13ConvBaseLayerE"],"paddle::ConvBaseLayer::ConvBaseLayer":[43,2,1,"_CPPv2N6paddle13ConvBaseLayer13ConvBaseLayerERK11LayerConfig"],"paddle::ConvBaseLayer::IntV":[43,4,1,"_CPPv2N6paddle13ConvBaseLayer4IntVE"],"paddle::ConvBaseLayer::biases_":[43,3,1,"_CPPv2N6paddle13ConvBaseLayer7biases_E"],"paddle::ConvBaseLayer::caffeMode_":[43,3,1,"_CPPv2N6paddle13ConvBaseLayer10caffeMode_E"],"paddle::ConvBaseLayer::channels_":[43,3,1,"_CPPv2N6paddle13ConvBaseLayer9channels_E"],"paddle::ConvBaseLayer::filterChannels_":[43,3,1,"_CPPv2N6paddle13ConvBaseLayer15filterChannels_E"],"paddle::ConvBaseLayer::filterPixels_":[43,3,1,"_CPPv2N6paddle13ConvBaseLayer13filterPixels_E"],"paddle::ConvBaseLayer::filterSizeY_":[43,3,1,"_CPPv2N6paddle13ConvBaseLayer12filterSizeY_E"],"paddle::ConvBaseLayer::filterSize_":[43,3,1,"_CPPv2N6paddle13ConvBaseLayer11filterSize_E"],"paddle::ConvBaseLayer::getWeight":[43,2,1,"_CPPv2N6paddle13ConvBaseLayer9getWeightEi"],"paddle::ConvBaseLayer::groups_":[43,3,1,"_CPPv2N6paddle13ConvBaseLayer7groups_E"],"paddle::ConvBaseLayer::imgPixels_":[43,3,1,"_CPPv2N6paddle13ConvBaseLayer10imgPixels_E"],"paddle::ConvBaseLayer::imgSize_":[43,3,1,"_CPPv2N6paddle13ConvBaseLayer8imgSize_E"],"paddle::ConvBaseLayer::init":[43,2,1,"_CPPv2N6paddle13ConvBaseLayer4initERK8LayerMapRK12ParameterMap"],"paddle::ConvBaseLayer::numFilters_":[43,3,1,"_CPPv2N6paddle13ConvBaseLayer11numFilters_E"],"paddle::ConvBaseLayer::outputSize":[43,2,1,"_CPPv2N6paddle13ConvBaseLayer10outputSizeEiiii"],"paddle::ConvBaseLayer::outputX_":[43,3,1,"_CPPv2N6paddle13ConvBaseLayer8outputX_E"],"paddle::ConvBaseLayer::outputs_":[43,3,1,"_CPPv2N6paddle13ConvBaseLayer8outputs_E"],"paddle::ConvBaseLayer::paddingY_":[43,3,1,"_CPPv2N6paddle13ConvBaseLayer9paddingY_E"],"paddle::ConvBaseLayer::padding_":[43,3,1,"_CPPv2N6paddle13ConvBaseLayer8padding_E"],"paddle::ConvBaseLayer::sharedBiases_":[43,3,1,"_CPPv2N6paddle13ConvBaseLayer13sharedBiases_E"],"paddle::ConvBaseLayer::strideY_":[43,3,1,"_CPPv2N6paddle13ConvBaseLayer8strideY_E"],"paddle::ConvBaseLayer::stride_":[43,3,1,"_CPPv2N6paddle13ConvBaseLayer7stride_E"],"paddle::ConvBaseLayer::weights_":[43,3,1,"_CPPv2N6paddle13ConvBaseLayer8weights_E"],"paddle::ConvOperator":[43,1,1,"_CPPv2N6paddle12ConvOperatorE"],"paddle::ConvOperator::ConvOperator":[43,2,1,"_CPPv2N6paddle12ConvOperator12ConvOperatorERK14OperatorConfigb"],"paddle::ConvOperator::backward":[43,2,1,"_CPPv2N6paddle12ConvOperator8backwardEv"],"paddle::ConvOperator::forward":[43,2,1,"_CPPv2N6paddle12ConvOperator7forwardEv"],"paddle::ConvOperator::~ConvOperator":[43,2,1,"_CPPv2N6paddle12ConvOperatorD0Ev"],"paddle::ConvShiftLayer":[43,1,1,"_CPPv2N6paddle14ConvShiftLayerE"],"paddle::ConvShiftLayer::ConvShiftLayer":[43,2,1,"_CPPv2N6paddle14ConvShiftLayer14ConvShiftLayerERK11LayerConfig"],"paddle::ConvShiftLayer::backward":[43,2,1,"_CPPv2N6paddle14ConvShiftLayer8backwardERK14UpdateCallback"],"paddle::ConvShiftLayer::forward":[43,2,1,"_CPPv2N6paddle14ConvShiftLayer7forwardE8PassType"],"paddle::ConvShiftLayer::init":[43,2,1,"_CPPv2N6paddle14ConvShiftLayer4initERK8LayerMapRK12ParameterMap"],"paddle::ConvShiftLayer::~ConvShiftLayer":[43,2,1,"_CPPv2N6paddle14ConvShiftLayerD0Ev"],"paddle::ConvexCombinationLayer":[43,1,1,"_CPPv2N6paddle22ConvexCombinationLayerE"],"paddle::ConvexCombinationLayer::ConvexCombinationLayer":[43,2,1,"_CPPv2N6paddle22ConvexCombinationLayer22ConvexCombinationLayerERK11LayerConfig"],"paddle::ConvexCombinationLayer::backward":[43,2,1,"_CPPv2N6paddle22ConvexCombinationLayer8backwardERK14UpdateCallback"],"paddle::ConvexCombinationLayer::forward":[43,2,1,"_CPPv2N6paddle22ConvexCombinationLayer7forwardE8PassType"],"paddle::ConvexCombinationLayer::init":[43,2,1,"_CPPv2N6paddle22ConvexCombinationLayer4initERK8LayerMapRK12ParameterMap"],"paddle::ConvexCombinationLayer::tmpMtx0":[43,3,1,"_CPPv2N6paddle22ConvexCombinationLayer7tmpMtx0E"],"paddle::ConvexCombinationLayer::tmpRow0":[43,3,1,"_CPPv2N6paddle22ConvexCombinationLayer7tmpRow0E"],"paddle::ConvexCombinationLayer::tmpRow1":[43,3,1,"_CPPv2N6paddle22ConvexCombinationLayer7tmpRow1E"],"paddle::ConvexCombinationLayer::~ConvexCombinationLayer":[43,2,1,"_CPPv2N6paddle22ConvexCombinationLayerD0Ev"],"paddle::CoordIterator":[43,1,1,"_CPPv2N6paddle13CoordIteratorE"],"paddle::CoordIterator::CoordIterator":[43,2,1,"_CPPv2N6paddle13CoordIterator13CoordIteratorENSt6vectorIiEENSt6vectorIbEE"],"paddle::CoordIterator::begin":[43,2,1,"_CPPv2N6paddle13CoordIterator5beginEv"],"paddle::CoordIterator::curPos":[43,2,1,"_CPPv2N6paddle13CoordIterator6curPosEv"],"paddle::CoordIterator::curPos_":[43,3,1,"_CPPv2N6paddle13CoordIterator7curPos_E"],"paddle::CoordIterator::dims_":[43,3,1,"_CPPv2N6paddle13CoordIterator5dims_E"],"paddle::CoordIterator::directions_":[43,3,1,"_CPPv2N6paddle13CoordIterator11directions_E"],"paddle::CoordIterator::end":[43,2,1,"_CPPv2N6paddle13CoordIterator3endEv"],"paddle::CoordIterator::end_":[43,3,1,"_CPPv2N6paddle13CoordIterator4end_E"],"paddle::CoordIterator::getNextPos":[43,2,1,"_CPPv2N6paddle13CoordIterator10getNextPosERKNSt6vectorIiEEiRNSt6vectorIiEE"],"paddle::CoordIterator::getPrePos":[43,2,1,"_CPPv2N6paddle13CoordIterator9getPrePosERKNSt6vectorIiEEiRNSt6vectorIiEE"],"paddle::CoordIterator::offset":[43,2,1,"_CPPv2N6paddle13CoordIterator6offsetEv"],"paddle::CoordIterator::operator++":[43,2,1,"_CPPv2N6paddle13CoordIteratorppEv"],"paddle::CoordIterator::operator--":[43,2,1,"_CPPv2N6paddle13CoordIteratormmEv"],"paddle::CoordIterator::rbegin":[43,2,1,"_CPPv2N6paddle13CoordIterator6rbeginEv"],"paddle::CoordIterator::step":[43,2,1,"_CPPv2N6paddle13CoordIterator4stepE6size_tb"],"paddle::CosSimLayer":[43,1,1,"_CPPv2N6paddle11CosSimLayerE"],"paddle::CosSimLayer::CosSimLayer":[43,2,1,"_CPPv2N6paddle11CosSimLayer11CosSimLayerERK11LayerConfig"],"paddle::CosSimLayer::backward":[43,2,1,"_CPPv2N6paddle11CosSimLayer8backwardERK14UpdateCallback"],"paddle::CosSimLayer::forward":[43,2,1,"_CPPv2N6paddle11CosSimLayer7forwardE8PassType"],"paddle::CosSimLayer::init":[43,2,1,"_CPPv2N6paddle11CosSimLayer4initERK8LayerMapRK12ParameterMap"],"paddle::CosSimLayer::~CosSimLayer":[43,2,1,"_CPPv2N6paddle11CosSimLayerD0Ev"],"paddle::CosSimVecMatLayer":[43,1,1,"_CPPv2N6paddle17CosSimVecMatLayerE"],"paddle::CosSimVecMatLayer::CosSimVecMatLayer":[43,2,1,"_CPPv2N6paddle17CosSimVecMatLayer17CosSimVecMatLayerERK11LayerConfig"],"paddle::CosSimVecMatLayer::backward":[43,2,1,"_CPPv2N6paddle17CosSimVecMatLayer8backwardERK14UpdateCallback"],"paddle::CosSimVecMatLayer::forward":[43,2,1,"_CPPv2N6paddle17CosSimVecMatLayer7forwardE8PassType"],"paddle::CosSimVecMatLayer::init":[43,2,1,"_CPPv2N6paddle17CosSimVecMatLayer4initERK8LayerMapRK12ParameterMap"],"paddle::CosSimVecMatLayer::tmpMtx0":[43,3,1,"_CPPv2N6paddle17CosSimVecMatLayer7tmpMtx0E"],"paddle::CosSimVecMatLayer::tmpMtx1":[43,3,1,"_CPPv2N6paddle17CosSimVecMatLayer7tmpMtx1E"],"paddle::CosSimVecMatLayer::tmpRow0":[43,3,1,"_CPPv2N6paddle17CosSimVecMatLayer7tmpRow0E"],"paddle::CosSimVecMatLayer::tmpRow1":[43,3,1,"_CPPv2N6paddle17CosSimVecMatLayer7tmpRow1E"],"paddle::CosSimVecMatLayer::tmpRow2":[43,3,1,"_CPPv2N6paddle17CosSimVecMatLayer7tmpRow2E"],"paddle::CosSimVecMatLayer::tmpRow3":[43,3,1,"_CPPv2N6paddle17CosSimVecMatLayer7tmpRow3E"],"paddle::CosSimVecMatLayer::~CosSimVecMatLayer":[43,2,1,"_CPPv2N6paddle17CosSimVecMatLayerD0Ev"],"paddle::CostLayer":[43,1,1,"_CPPv2N6paddle9CostLayerE"],"paddle::CostLayer::CostLayer":[43,2,1,"_CPPv2N6paddle9CostLayer9CostLayerERK11LayerConfig"],"paddle::CostLayer::backward":[43,2,1,"_CPPv2N6paddle9CostLayer8backwardERK14UpdateCallback"],"paddle::CostLayer::backwardImp":[43,2,1,"_CPPv2N6paddle9CostLayer11backwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::CostLayer::coeff_":[43,3,1,"_CPPv2N6paddle9CostLayer6coeff_E"],"paddle::CostLayer::forward":[43,2,1,"_CPPv2N6paddle9CostLayer7forwardE8PassType"],"paddle::CostLayer::forwardImp":[43,2,1,"_CPPv2N6paddle9CostLayer10forwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::CostLayer::getLabelLayer":[43,2,1,"_CPPv2N6paddle9CostLayer13getLabelLayerEv"],"paddle::CostLayer::getOutputLayer":[43,2,1,"_CPPv2N6paddle9CostLayer14getOutputLayerEv"],"paddle::CostLayer::init":[43,2,1,"_CPPv2N6paddle9CostLayer4initERK8LayerMapRK12ParameterMap"],"paddle::CostLayer::weightLayer_":[43,3,1,"_CPPv2N6paddle9CostLayer12weightLayer_E"],"paddle::CpuAllocator":[48,1,1,"_CPPv2N6paddle12CpuAllocatorE"],"paddle::CpuAllocator::alloc":[48,2,1,"_CPPv2N6paddle12CpuAllocator5allocE6size_t"],"paddle::CpuAllocator::free":[48,2,1,"_CPPv2N6paddle12CpuAllocator4freeEPv"],"paddle::CpuAllocator::getName":[48,2,1,"_CPPv2N6paddle12CpuAllocator7getNameEv"],"paddle::CpuAllocator::~CpuAllocator":[48,2,1,"_CPPv2N6paddle12CpuAllocatorD0Ev"],"paddle::CpuGpuVector":[46,4,1,"_CPPv2N6paddle12CpuGpuVectorE"],"paddle::CpuGpuVectorPtr":[46,4,1,"_CPPv2N6paddle15CpuGpuVectorPtrE"],"paddle::CpuGpuVectorT":[46,1,1,"_CPPv2N6paddle13CpuGpuVectorTE"],"paddle::CpuGpuVectorT::CpuGpuVectorT":[46,2,1,"_CPPv2N6paddle13CpuGpuVectorT13CpuGpuVectorTERKNSt10shared_ptrI7VectorTI1TEEE"],"paddle::CpuGpuVectorT::copyFrom":[46,2,1,"_CPPv2N6paddle13CpuGpuVectorT8copyFromERK7VectorTI1TE11hl_stream_t"],"paddle::CpuGpuVectorT::copyToCpu":[46,2,1,"_CPPv2N6paddle13CpuGpuVectorT9copyToCpuEv"],"paddle::CpuGpuVectorT::copyToGpu":[46,2,1,"_CPPv2N6paddle13CpuGpuVectorT9copyToGpuEv"],"paddle::CpuGpuVectorT::cpuVectorT_":[46,3,1,"_CPPv2N6paddle13CpuGpuVectorT11cpuVectorT_E"],"paddle::CpuGpuVectorT::create":[46,2,1,"_CPPv2N6paddle13CpuGpuVectorT6createE6size_tb"],"paddle::CpuGpuVectorT::fillSequence":[46,2,1,"_CPPv2N6paddle13CpuGpuVectorT12fillSequenceEb"],"paddle::CpuGpuVectorT::getData":[46,2,1,"_CPPv2N6paddle13CpuGpuVectorT7getDataEb"],"paddle::CpuGpuVectorT::getElement":[46,2,1,"_CPPv2N6paddle13CpuGpuVectorT10getElementE6size_t"],"paddle::CpuGpuVectorT::getMutableData":[46,2,1,"_CPPv2N6paddle13CpuGpuVectorT14getMutableDataEb"],"paddle::CpuGpuVectorT::getMutableVector":[46,2,1,"_CPPv2N6paddle13CpuGpuVectorT16getMutableVectorEb"],"paddle::CpuGpuVectorT::getSize":[46,2,1,"_CPPv2N6paddle13CpuGpuVectorT7getSizeEv"],"paddle::CpuGpuVectorT::getSync":[46,2,1,"_CPPv2N6paddle13CpuGpuVectorT7getSyncEv"],"paddle::CpuGpuVectorT::getVector":[46,2,1,"_CPPv2N6paddle13CpuGpuVectorT9getVectorEb"],"paddle::CpuGpuVectorT::gpuVectorT_":[46,3,1,"_CPPv2N6paddle13CpuGpuVectorT11gpuVectorT_E"],"paddle::CpuGpuVectorT::resize":[46,2,1,"_CPPv2N6paddle13CpuGpuVectorT6resizeE6size_tb"],"paddle::CpuGpuVectorT::resizeOrCreate":[46,2,1,"_CPPv2N6paddle13CpuGpuVectorT14resizeOrCreateERNSt10shared_ptrI13CpuGpuVectorTI1TEEE6size_tb"],"paddle::CpuGpuVectorT::setElement":[46,2,1,"_CPPv2N6paddle13CpuGpuVectorT10setElementE6size_tRK1Tb"],"paddle::CpuGpuVectorT::setSync":[46,2,1,"_CPPv2N6paddle13CpuGpuVectorT7setSyncEb"],"paddle::CpuGpuVectorT::syncFlag_":[46,3,1,"_CPPv2N6paddle13CpuGpuVectorT9syncFlag_E"],"paddle::CpuGpuVectorT::sync_":[46,3,1,"_CPPv2N6paddle13CpuGpuVectorT5sync_E"],"paddle::CpuGpuVectorT::zeroMem":[46,2,1,"_CPPv2N6paddle13CpuGpuVectorT7zeroMemEb"],"paddle::CpuGpuVectorT::~CpuGpuVectorT":[46,2,1,"_CPPv2N6paddle13CpuGpuVectorTD0Ev"],"paddle::CpuIVector":[46,4,1,"_CPPv2N6paddle10CpuIVectorE"],"paddle::CpuIVectorPtr":[46,4,1,"_CPPv2N6paddle13CpuIVectorPtrE"],"paddle::CpuMatrix":[46,1,1,"_CPPv2N6paddle9CpuMatrixE"],"paddle::CpuMatrix::CpuMatrix":[46,2,1,"_CPPv2N6paddle9CpuMatrix9CpuMatrixEP4real6size_t6size_tb"],"paddle::CpuMatrix::accumulateColSum":[46,2,1,"_CPPv2N6paddle9CpuMatrix16accumulateColSumER6Matrix"],"paddle::CpuMatrix::addBias":[46,2,1,"_CPPv2N6paddle9CpuMatrix7addBiasER6Matrix4real"],"paddle::CpuMatrix::addByBitCode":[46,2,1,"_CPPv2N6paddle9CpuMatrix12addByBitCodeE6size_tRK7IVectorRK6Matrix"],"paddle::CpuMatrix::addByBitCodeBackward":[46,2,1,"_CPPv2N6paddle9CpuMatrix20addByBitCodeBackwardE6size_tRK7IVectorR6Matrix"],"paddle::CpuMatrix::addColumnVector":[46,2,1,"_CPPv2N6paddle9CpuMatrix15addColumnVectorERK6Matrix"],"paddle::CpuMatrix::addElements":[46,2,1,"_CPPv2N6paddle9CpuMatrix11addElementsER6MatrixR7IVector"],"paddle::CpuMatrix::addToRows":[46,2,1,"_CPPv2N6paddle9CpuMatrix9addToRowsER6MatrixR7IVector"],"paddle::CpuMatrix::addToRowsImp":[46,2,1,"_CPPv2N6paddle9CpuMatrix12addToRowsImpER12TableMatTypeR7IVector"],"paddle::CpuMatrix::avgPoolBackward":[46,2,1,"_CPPv2N6paddle9CpuMatrix15avgPoolBackwardER6Matrix6size_t6size_t6size_ti6size_t6size_t6size_t4real4real"],"paddle::CpuMatrix::avgPoolForward":[46,2,1,"_CPPv2N6paddle9CpuMatrix14avgPoolForwardER6Matrix6size_t6size_t6size_t6size_ti6size_t6size_t6size_t"],"paddle::CpuMatrix::check":[46,2,1,"_CPPv2N6paddle9CpuMatrix5checkERNSt7ostreamER6Matrixb"],"paddle::CpuMatrix::circularConv":[46,2,1,"_CPPv2N6paddle9CpuMatrix12circularConvER6MatrixR6Matrix"],"paddle::CpuMatrix::circularConvDerivative":[46,2,1,"_CPPv2N6paddle9CpuMatrix22circularConvDerivativeER6MatrixR6MatrixR6MatrixR6MatrixR6Matrix"],"paddle::CpuMatrix::classificationError":[46,2,1,"_CPPv2N6paddle9CpuMatrix19classificationErrorE9MatrixPtr10IVectorPtr"],"paddle::CpuMatrix::classificationErrorMulti":[46,2,1,"_CPPv2N6paddle9CpuMatrix24classificationErrorMultiER6MatrixR6Matrix4real"],"paddle::CpuMatrix::clone":[46,2,1,"_CPPv2N6paddle9CpuMatrix5cloneE6size_t6size_tb"],"paddle::CpuMatrix::colMax":[46,2,1,"_CPPv2N6paddle9CpuMatrix6colMaxER6Matrix"],"paddle::CpuMatrix::colMerge":[46,2,1,"_CPPv2N6paddle9CpuMatrix8colMergeER6Matrix"],"paddle::CpuMatrix::collectBias":[46,2,1,"_CPPv2N6paddle9CpuMatrix11collectBiasER6Matrix4real"],"paddle::CpuMatrix::contextProjectionBackward":[46,2,1,"_CPPv2N6paddle9CpuMatrix25contextProjectionBackwardE9MatrixPtr9MatrixPtrRK7IVectorii6size_tb"],"paddle::CpuMatrix::contextProjectionForward":[46,2,1,"_CPPv2N6paddle9CpuMatrix24contextProjectionForwardE9MatrixPtr9MatrixPtrRK7IVectorii6size_tb"],"paddle::CpuMatrix::convExpand":[46,2,1,"_CPPv2N6paddle9CpuMatrix10convExpandER6Matrixiiiiiiiiiii"],"paddle::CpuMatrix::convShrink":[46,2,1,"_CPPv2N6paddle9CpuMatrix10convShrinkER6Matrixiiiiiiiiiii4real4real"],"paddle::CpuMatrix::copyByRowIndex":[46,2,1,"_CPPv2N6paddle9CpuMatrix14copyByRowIndexER6MatrixR7IVector"],"paddle::CpuMatrix::copyFrom":[46,2,1,"_CPPv2N6paddle9CpuMatrix8copyFromERK7IVector"],"paddle::CpuMatrix::cosSim":[46,2,1,"_CPPv2N6paddle9CpuMatrix6cosSimER6MatrixR6Matrix4real"],"paddle::CpuMatrix::cosSimDerivative":[46,2,1,"_CPPv2N6paddle9CpuMatrix16cosSimDerivativeER6MatrixR6MatrixR6MatrixR6MatrixR6Matrix4real"],"paddle::CpuMatrix::crossMapNormalBwd":[46,2,1,"_CPPv2N6paddle9CpuMatrix17crossMapNormalBwdER6MatrixR6MatrixR6MatrixR6Matrix6size_t6size_t6size_t6size_tff"],"paddle::CpuMatrix::crossMapNormalFwd":[46,2,1,"_CPPv2N6paddle9CpuMatrix17crossMapNormalFwdER6Matrix6size_t6size_tR6Matrix6size_t6size_tff"],"paddle::CpuMatrix::getAbsSum":[46,2,1,"_CPPv2N6paddle9CpuMatrix9getAbsSumEv"],"paddle::CpuMatrix::getElement":[46,2,1,"_CPPv2N6paddle9CpuMatrix10getElementE6size_t6size_t"],"paddle::CpuMatrix::getMax":[46,2,1,"_CPPv2N6paddle9CpuMatrix6getMaxEv"],"paddle::CpuMatrix::getMin":[46,2,1,"_CPPv2N6paddle9CpuMatrix6getMinEv"],"paddle::CpuMatrix::getRow":[46,2,1,"_CPPv2N6paddle9CpuMatrix6getRowE6size_t"],"paddle::CpuMatrix::getRowBuf":[46,2,1,"_CPPv2N6paddle9CpuMatrix9getRowBufE6size_t"],"paddle::CpuMatrix::getSum":[46,2,1,"_CPPv2N6paddle9CpuMatrix6getSumEv"],"paddle::CpuMatrix::getTranspose":[46,2,1,"_CPPv2N6paddle9CpuMatrix12getTransposeEv"],"paddle::CpuMatrix::leftMul":[46,2,1,"_CPPv2N6paddle9CpuMatrix7leftMulER6Matrix4real4real"],"paddle::CpuMatrix::maxPoolBackward":[46,2,1,"_CPPv2N6paddle9CpuMatrix15maxPoolBackwardER6Matrix6size_t6size_tR6MatrixR6Matrix6size_ti6size_t6size_t6size_t4real4real"],"paddle::CpuMatrix::maxPoolForward":[46,2,1,"_CPPv2N6paddle9CpuMatrix14maxPoolForwardER6Matrix6size_t6size_t6size_t6size_ti6size_t6size_t6size_t"],"paddle::CpuMatrix::maxSequenceBackward":[46,2,1,"_CPPv2N6paddle9CpuMatrix19maxSequenceBackwardER6MatrixRK7IVectorR7IVector"],"paddle::CpuMatrix::maxSequenceForward":[46,2,1,"_CPPv2N6paddle9CpuMatrix18maxSequenceForwardER6MatrixRK7IVectorR7IVector"],"paddle::CpuMatrix::mul":[46,2,1,"_CPPv2N6paddle9CpuMatrix3mulEP9CpuMatrixP9CpuMatrixP15CpuSparseMatrix4real4real"],"paddle::CpuMatrix::mulByBitCode":[46,2,1,"_CPPv2N6paddle9CpuMatrix12mulByBitCodeE6size_tRK7IVectorRK6MatrixRK6Matrix"],"paddle::CpuMatrix::mulByBitCodeBackwardError":[46,2,1,"_CPPv2N6paddle9CpuMatrix25mulByBitCodeBackwardErrorE6size_tRK7IVectorRK6MatrixR6Matrix"],"paddle::CpuMatrix::mulByBitCodeBackwardWeight":[46,2,1,"_CPPv2N6paddle9CpuMatrix26mulByBitCodeBackwardWeightE6size_tRK7IVectorR6MatrixRK6Matrix"],"paddle::CpuMatrix::multiBinaryLabelCrossEntropy":[46,2,1,"_CPPv2N6paddle9CpuMatrix28multiBinaryLabelCrossEntropyER6MatrixR6Matrix"],"paddle::CpuMatrix::multiBinaryLabelCrossEntropyBp":[46,2,1,"_CPPv2N6paddle9CpuMatrix30multiBinaryLabelCrossEntropyBpER6MatrixR6Matrix"],"paddle::CpuMatrix::oneHotCrossEntropy":[46,2,1,"_CPPv2N6paddle9CpuMatrix18oneHotCrossEntropyER6MatrixR7IVector"],"paddle::CpuMatrix::oneHotCrossEntropyBp":[46,2,1,"_CPPv2N6paddle9CpuMatrix20oneHotCrossEntropyBpER6MatrixR7IVector"],"paddle::CpuMatrix::oneHotCrossEntropyWithSelfNorm":[46,2,1,"_CPPv2N6paddle9CpuMatrix30oneHotCrossEntropyWithSelfNormER6MatrixR7IVector4real"],"paddle::CpuMatrix::oneHotCrossEntropyWithSelfNormBp":[46,2,1,"_CPPv2N6paddle9CpuMatrix32oneHotCrossEntropyWithSelfNormBpER6MatrixR7IVector4real"],"paddle::CpuMatrix::paramReluBackwardDiff":[46,2,1,"_CPPv2N6paddle9CpuMatrix21paramReluBackwardDiffER6MatrixR6MatrixR6Matrix"],"paddle::CpuMatrix::paramReluBackwardW":[46,2,1,"_CPPv2N6paddle9CpuMatrix18paramReluBackwardWER6MatrixR6Matrix"],"paddle::CpuMatrix::paramReluForward":[46,2,1,"_CPPv2N6paddle9CpuMatrix16paramReluForwardER6MatrixR6Matrix"],"paddle::CpuMatrix::print":[46,2,1,"_CPPv2N6paddle9CpuMatrix5printERNSt7ostreamE6size_t6size_t"],"paddle::CpuMatrix::printOneRow":[46,2,1,"_CPPv2N6paddle9CpuMatrix11printOneRowERNSt7ostreamE6size_t"],"paddle::CpuMatrix::randomizeUniform":[46,2,1,"_CPPv2N6paddle9CpuMatrix16randomizeUniformEv"],"paddle::CpuMatrix::resetOne":[46,2,1,"_CPPv2N6paddle9CpuMatrix8resetOneEv"],"paddle::CpuMatrix::resize":[46,2,1,"_CPPv2N6paddle9CpuMatrix6resizeE6size_t6size_t6size_t15SparseValueType12SparseFormat"],"paddle::CpuMatrix::rightMul":[46,2,1,"_CPPv2N6paddle9CpuMatrix8rightMulER6Matrix4real4real"],"paddle::CpuMatrix::rowMax":[46,2,1,"_CPPv2N6paddle9CpuMatrix6rowMaxER7IVectorR6Matrix"],"paddle::CpuMatrix::rowMaxId":[46,2,1,"_CPPv2N6paddle9CpuMatrix8rowMaxIdER7IVector"],"paddle::CpuMatrix::rowNormalizeL1":[46,2,1,"_CPPv2N6paddle9CpuMatrix14rowNormalizeL1ER6Matrix"],"paddle::CpuMatrix::rowSum":[46,2,1,"_CPPv2N6paddle9CpuMatrix6rowSumER6Matrix"],"paddle::CpuMatrix::scaledTanh":[46,2,1,"_CPPv2N6paddle9CpuMatrix10scaledTanhER6Matrix4real4real"],"paddle::CpuMatrix::selectElements":[46,2,1,"_CPPv2N6paddle9CpuMatrix14selectElementsER6MatrixR7IVector"],"paddle::CpuMatrix::selectRows":[46,2,1,"_CPPv2N6paddle9CpuMatrix10selectRowsER6MatrixR7IVector"],"paddle::CpuMatrix::selectRowsImp":[46,2,1,"_CPPv2N6paddle9CpuMatrix13selectRowsImpER12TableMatTypeR7IVector"],"paddle::CpuMatrix::sequenceAvgForward":[46,2,1,"_CPPv2N6paddle9CpuMatrix18sequenceAvgForwardER6MatrixRK7IVectori"],"paddle::CpuMatrix::sequenceSoftmax":[46,2,1,"_CPPv2N6paddle9CpuMatrix15sequenceSoftmaxER6MatrixRK7IVector"],"paddle::CpuMatrix::setRow":[46,2,1,"_CPPv2N6paddle9CpuMatrix6setRowE6size_t6size_tPKjPK4real"],"paddle::CpuMatrix::softmax":[46,2,1,"_CPPv2N6paddle9CpuMatrix7softmaxER6Matrix"],"paddle::CpuMatrix::softmaxDerivative":[46,2,1,"_CPPv2N6paddle9CpuMatrix17softmaxDerivativeER6MatrixR6Matrix"],"paddle::CpuMatrix::softrelu":[46,2,1,"_CPPv2N6paddle9CpuMatrix8softreluER6Matrix"],"paddle::CpuMatrix::softreluDerivative":[46,2,1,"_CPPv2N6paddle9CpuMatrix18softreluDerivativeER6Matrix"],"paddle::CpuMatrix::subByBitCode":[46,2,1,"_CPPv2N6paddle9CpuMatrix12subByBitCodeE6size_tR7IVector"],"paddle::CpuMatrix::sumByBitCode":[46,2,1,"_CPPv2N6paddle9CpuMatrix12sumByBitCodeE6size_tR7IVectorR6Matrix4real"],"paddle::CpuMatrix::sumOfSquares":[46,2,1,"_CPPv2N6paddle9CpuMatrix12sumOfSquaresER6MatrixR6Matrix"],"paddle::CpuMatrix::sumOfSquaresBp":[46,2,1,"_CPPv2N6paddle9CpuMatrix14sumOfSquaresBpER6MatrixR6Matrix"],"paddle::CpuMatrix::tanh":[46,2,1,"_CPPv2N6paddle9CpuMatrix4tanhER6Matrix"],"paddle::CpuMatrix::tanhDerivative":[46,2,1,"_CPPv2N6paddle9CpuMatrix14tanhDerivativeER6Matrix"],"paddle::CpuMatrix::transpose":[46,2,1,"_CPPv2N6paddle9CpuMatrix9transposeE9MatrixPtrb"],"paddle::CpuMatrix::zeroMem":[46,2,1,"_CPPv2N6paddle9CpuMatrix7zeroMemEv"],"paddle::CpuMatrix::~CpuMatrix":[46,2,1,"_CPPv2N6paddle9CpuMatrixD0Ev"],"paddle::CpuMatrixPtr":[46,4,1,"_CPPv2N6paddle12CpuMatrixPtrE"],"paddle::CpuMemHandlePtr":[48,4,1,"_CPPv2N6paddle15CpuMemHandlePtrE"],"paddle::CpuMemoryHandle":[48,1,1,"_CPPv2N6paddle15CpuMemoryHandleE"],"paddle::CpuMemoryHandle::CpuMemoryHandle":[48,2,1,"_CPPv2N6paddle15CpuMemoryHandle15CpuMemoryHandleE6size_t"],"paddle::CpuMemoryHandle::~CpuMemoryHandle":[48,2,1,"_CPPv2N6paddle15CpuMemoryHandleD0Ev"],"paddle::CpuSparseMatrix":[46,1,1,"_CPPv2N6paddle15CpuSparseMatrixE"],"paddle::CpuSparseMatrix::CpuSparseMatrix":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix15CpuSparseMatrixEP4realPiPi6size_t6size_t6size_t15SparseValueType12SparseFormatb"],"paddle::CpuSparseMatrix::DEFAULT_AVG_WIDTH":[46,3,1,"_CPPv2N6paddle15CpuSparseMatrix17DEFAULT_AVG_WIDTHE"],"paddle::CpuSparseMatrix::add3":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix4add3EP9CpuMatrix"],"paddle::CpuSparseMatrix::addBias":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix7addBiasER6Matrix4real"],"paddle::CpuSparseMatrix::clone":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix5cloneE6size_t6size_tb"],"paddle::CpuSparseMatrix::cols_":[46,3,1,"_CPPv2N6paddle15CpuSparseMatrix5cols_E"],"paddle::CpuSparseMatrix::copyFrom":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix8copyFromERNSt6vectorIiEERNSt6vectorIiEERNSt6vectorI4realEE"],"paddle::CpuSparseMatrix::copyRow":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix7copyRowEi6size_tPK20sparse_float_value_t"],"paddle::CpuSparseMatrix::cpuLocalMats_":[46,3,1,"_CPPv2N6paddle15CpuSparseMatrix13cpuLocalMats_E"],"paddle::CpuSparseMatrix::fillRowIndices":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix14fillRowIndicesER10IVectorPtr"],"paddle::CpuSparseMatrix::format_":[46,3,1,"_CPPv2N6paddle15CpuSparseMatrix7format_E"],"paddle::CpuSparseMatrix::getColNum":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix9getColNumE6size_t"],"paddle::CpuSparseMatrix::getColStartIdx":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix14getColStartIdxE6size_t"],"paddle::CpuSparseMatrix::getCols":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix7getColsEv"],"paddle::CpuSparseMatrix::getColumn":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix9getColumnE6size_t"],"paddle::CpuSparseMatrix::getData":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix7getDataEv"],"paddle::CpuSparseMatrix::getFormat":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix9getFormatEv"],"paddle::CpuSparseMatrix::getMax":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix6getMaxEv"],"paddle::CpuSparseMatrix::getMin":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix6getMinEv"],"paddle::CpuSparseMatrix::getRowCols":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix10getRowColsE6size_t"],"paddle::CpuSparseMatrix::getRowNum":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix9getRowNumE6size_t"],"paddle::CpuSparseMatrix::getRowStartIdx":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix14getRowStartIdxE6size_t"],"paddle::CpuSparseMatrix::getRowValues":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix12getRowValuesE6size_t"],"paddle::CpuSparseMatrix::getRows":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix7getRowsEv"],"paddle::CpuSparseMatrix::getSum":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix6getSumEv"],"paddle::CpuSparseMatrix::getTmpSparseMatrix":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix18getTmpSparseMatrixE6size_t6size_t"],"paddle::CpuSparseMatrix::getTranspose":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix12getTransposeEv"],"paddle::CpuSparseMatrix::getValue":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix8getValueEv"],"paddle::CpuSparseMatrix::getValueType":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix12getValueTypeEv"],"paddle::CpuSparseMatrix::isSparse":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix8isSparseEv"],"paddle::CpuSparseMatrix::mul":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix3mulE9MatrixPtr9MatrixPtr4real4real"],"paddle::CpuSparseMatrix::print":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix5printERNSt7ostreamE"],"paddle::CpuSparseMatrix::printOneRow":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix11printOneRowERNSt7ostreamE6size_t"],"paddle::CpuSparseMatrix::randomizeUniform":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix16randomizeUniformEv"],"paddle::CpuSparseMatrix::resize":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix6resizeE6size_t6size_t6size_t15SparseValueType12SparseFormat"],"paddle::CpuSparseMatrix::rowMax":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix6rowMaxER7IVectorR6Matrix"],"paddle::CpuSparseMatrix::rows_":[46,3,1,"_CPPv2N6paddle15CpuSparseMatrix5rows_E"],"paddle::CpuSparseMatrix::setRow":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix6setRowE6size_t6size_tPKjPK4real"],"paddle::CpuSparseMatrix::sparseResize":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix12sparseResizeEv"],"paddle::CpuSparseMatrix::square":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix6squareEv"],"paddle::CpuSparseMatrix::subMatrix":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix9subMatrixE6size_t6size_t"],"paddle::CpuSparseMatrix::transpose":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix9transposeE9MatrixPtrb"],"paddle::CpuSparseMatrix::trimFrom":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix8trimFromERK15CpuSparseMatrix"],"paddle::CpuSparseMatrix::valueType_":[46,3,1,"_CPPv2N6paddle15CpuSparseMatrix10valueType_E"],"paddle::CpuSparseMatrix::value_":[46,3,1,"_CPPv2N6paddle15CpuSparseMatrix6value_E"],"paddle::CpuSparseMatrix::zeroMem":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrix7zeroMemEv"],"paddle::CpuSparseMatrix::~CpuSparseMatrix":[46,2,1,"_CPPv2N6paddle15CpuSparseMatrixD0Ev"],"paddle::CpuSparseMatrixPtr":[46,4,1,"_CPPv2N6paddle18CpuSparseMatrixPtrE"],"paddle::CpuVector":[46,4,1,"_CPPv2N6paddle9CpuVectorE"],"paddle::CpuVectorPtr":[46,4,1,"_CPPv2N6paddle12CpuVectorPtrE"],"paddle::CpuVectorT":[46,1,1,"_CPPv2N6paddle10CpuVectorTE"],"paddle::CpuVectorT::CpuVectorT":[46,2,1,"_CPPv2N6paddle10CpuVectorT10CpuVectorTERK7VectorTI1TE"],"paddle::CpuVectorT::copyFrom":[46,2,1,"_CPPv2N6paddle10CpuVectorT8copyFromERK7VectorTI1TE11hl_stream_t"],"paddle::CpuVectorT::copyTo":[46,2,1,"_CPPv2N6paddle10CpuVectorT6copyToEP10GpuVectorTI1TE"],"paddle::CpuVectorT::fillSequence":[46,2,1,"_CPPv2N6paddle10CpuVectorT12fillSequenceEv"],"paddle::CpuVectorT::get":[46,2,1,"_CPPv2N6paddle10CpuVectorT3getE6size_t"],"paddle::CpuVectorT::getAbsMax":[46,2,1,"_CPPv2N6paddle10CpuVectorT9getAbsMaxEv"],"paddle::CpuVectorT::getAbsSum":[46,2,1,"_CPPv2N6paddle10CpuVectorT9getAbsSumEv"],"paddle::CpuVectorT::getElement":[46,2,1,"_CPPv2N6paddle10CpuVectorT10getElementE6size_t"],"paddle::CpuVectorT::getMax":[46,2,1,"_CPPv2N6paddle10CpuVectorT6getMaxEv"],"paddle::CpuVectorT::getMin":[46,2,1,"_CPPv2N6paddle10CpuVectorT6getMinEv"],"paddle::CpuVectorT::getPoint":[46,2,1,"_CPPv2N6paddle10CpuVectorT8getPointEK8uint64_t"],"paddle::CpuVectorT::getSum":[46,2,1,"_CPPv2N6paddle10CpuVectorT6getSumEv"],"paddle::CpuVectorT::histogram":[46,2,1,"_CPPv2N6paddle10CpuVectorT9histogramERNSt7ostreamEi"],"paddle::CpuVectorT::isEqualTo":[46,2,1,"_CPPv2N6paddle10CpuVectorT9isEqualToERK7VectorTI1TERK1T"],"paddle::CpuVectorT::newMemory":[46,2,1,"_CPPv2N6paddle10CpuVectorT9newMemoryE6size_t"],"paddle::CpuVectorT::print":[46,2,1,"_CPPv2N6paddle10CpuVectorT5printERNSt7ostreamE6size_t"],"paddle::CpuVectorT::printOneElement":[46,2,1,"_CPPv2N6paddle10CpuVectorT15printOneElementERNSt7ostreamE6size_t"],"paddle::CpuVectorT::rand":[46,2,1,"_CPPv2N6paddle10CpuVectorT4randEv"],"paddle::CpuVectorT::randnorm":[46,2,1,"_CPPv2N6paddle10CpuVectorT8randnormE4real4real"],"paddle::CpuVectorT::reset":[46,2,1,"_CPPv2N6paddle10CpuVectorT5resetERK1T"],"paddle::CpuVectorT::selectFrom":[46,2,1,"_CPPv2N6paddle10CpuVectorT10selectFromERK7VectorTI1TERK7VectorTIiE"],"paddle::CpuVectorT::setElement":[46,2,1,"_CPPv2N6paddle10CpuVectorT10setElementE6size_tRK1T"],"paddle::CpuVectorT::uniform":[46,2,1,"_CPPv2N6paddle10CpuVectorT7uniformE4real4real"],"paddle::CpuVectorT::zeroMem":[46,2,1,"_CPPv2N6paddle10CpuVectorT7zeroMemEv"],"paddle::CudaHostAllocator":[48,1,1,"_CPPv2N6paddle17CudaHostAllocatorE"],"paddle::CudaHostAllocator::alloc":[48,2,1,"_CPPv2N6paddle17CudaHostAllocator5allocE6size_t"],"paddle::CudaHostAllocator::free":[48,2,1,"_CPPv2N6paddle17CudaHostAllocator4freeEPv"],"paddle::CudaHostAllocator::getName":[48,2,1,"_CPPv2N6paddle17CudaHostAllocator7getNameEv"],"paddle::CudaHostAllocator::~CudaHostAllocator":[48,2,1,"_CPPv2N6paddle17CudaHostAllocatorD0Ev"],"paddle::CudnnBatchNormLayer":[43,1,1,"_CPPv2N6paddle19CudnnBatchNormLayerE"],"paddle::CudnnBatchNormLayer::CudnnBatchNormLayer":[43,2,1,"_CPPv2N6paddle19CudnnBatchNormLayer19CudnnBatchNormLayerERK11LayerConfig"],"paddle::CudnnBatchNormLayer::EPS":[43,3,1,"_CPPv2N6paddle19CudnnBatchNormLayer3EPSE"],"paddle::CudnnBatchNormLayer::backward":[43,2,1,"_CPPv2N6paddle19CudnnBatchNormLayer8backwardERK14UpdateCallback"],"paddle::CudnnBatchNormLayer::bnParamDesc_":[43,3,1,"_CPPv2N6paddle19CudnnBatchNormLayer12bnParamDesc_E"],"paddle::CudnnBatchNormLayer::forward":[43,2,1,"_CPPv2N6paddle19CudnnBatchNormLayer7forwardE8PassType"],"paddle::CudnnBatchNormLayer::init":[43,2,1,"_CPPv2N6paddle19CudnnBatchNormLayer4initERK8LayerMapRK12ParameterMap"],"paddle::CudnnBatchNormLayer::ioDesc_":[43,3,1,"_CPPv2N6paddle19CudnnBatchNormLayer7ioDesc_E"],"paddle::CudnnBatchNormLayer::reshape":[43,2,1,"_CPPv2N6paddle19CudnnBatchNormLayer7reshapeEi"],"paddle::CudnnBatchNormLayer::tmpBiasGrad_":[43,3,1,"_CPPv2N6paddle19CudnnBatchNormLayer12tmpBiasGrad_E"],"paddle::CudnnBatchNormLayer::tmpWGrad_":[43,3,1,"_CPPv2N6paddle19CudnnBatchNormLayer9tmpWGrad_E"],"paddle::CudnnBatchNormLayer::~CudnnBatchNormLayer":[43,2,1,"_CPPv2N6paddle19CudnnBatchNormLayerD0Ev"],"paddle::CudnnConvLayer":[43,1,1,"_CPPv2N6paddle14CudnnConvLayerE"],"paddle::CudnnConvLayer::CudnnConvLayer":[43,2,1,"_CPPv2N6paddle14CudnnConvLayer14CudnnConvLayerERK11LayerConfig"],"paddle::CudnnConvLayer::addBiases":[43,2,1,"_CPPv2N6paddle14CudnnConvLayer9addBiasesEv"],"paddle::CudnnConvLayer::backward":[43,2,1,"_CPPv2N6paddle14CudnnConvLayer8backwardERK14UpdateCallback"],"paddle::CudnnConvLayer::biasDesc_":[43,3,1,"_CPPv2N6paddle14CudnnConvLayer9biasDesc_E"],"paddle::CudnnConvLayer::biasOffset_":[43,3,1,"_CPPv2N6paddle14CudnnConvLayer11biasOffset_E"],"paddle::CudnnConvLayer::bpropBiases":[43,2,1,"_CPPv2N6paddle14CudnnConvLayer11bpropBiasesEv"],"paddle::CudnnConvLayer::bwdDataAlgo_":[43,3,1,"_CPPv2N6paddle14CudnnConvLayer12bwdDataAlgo_E"],"paddle::CudnnConvLayer::bwdDataLimitBytes_":[43,3,1,"_CPPv2N6paddle14CudnnConvLayer18bwdDataLimitBytes_E"],"paddle::CudnnConvLayer::bwdFilterAlgo_":[43,3,1,"_CPPv2N6paddle14CudnnConvLayer14bwdFilterAlgo_E"],"paddle::CudnnConvLayer::bwdFilterLimitBytes_":[43,3,1,"_CPPv2N6paddle14CudnnConvLayer20bwdFilterLimitBytes_E"],"paddle::CudnnConvLayer::convDesc_":[43,3,1,"_CPPv2N6paddle14CudnnConvLayer9convDesc_E"],"paddle::CudnnConvLayer::filterDesc_":[43,3,1,"_CPPv2N6paddle14CudnnConvLayer11filterDesc_E"],"paddle::CudnnConvLayer::forward":[43,2,1,"_CPPv2N6paddle14CudnnConvLayer7forwardE8PassType"],"paddle::CudnnConvLayer::fwdAlgo_":[43,3,1,"_CPPv2N6paddle14CudnnConvLayer8fwdAlgo_E"],"paddle::CudnnConvLayer::fwdLimitBytes_":[43,3,1,"_CPPv2N6paddle14CudnnConvLayer14fwdLimitBytes_E"],"paddle::CudnnConvLayer::imageH_":[43,3,1,"_CPPv2N6paddle14CudnnConvLayer7imageH_E"],"paddle::CudnnConvLayer::imageW_":[43,3,1,"_CPPv2N6paddle14CudnnConvLayer7imageW_E"],"paddle::CudnnConvLayer::init":[43,2,1,"_CPPv2N6paddle14CudnnConvLayer4initERK8LayerMapRK12ParameterMap"],"paddle::CudnnConvLayer::inputDesc_":[43,3,1,"_CPPv2N6paddle14CudnnConvLayer10inputDesc_E"],"paddle::CudnnConvLayer::inputOffset_":[43,3,1,"_CPPv2N6paddle14CudnnConvLayer12inputOffset_E"],"paddle::CudnnConvLayer::isSelectAlgo_":[43,3,1,"_CPPv2N6paddle14CudnnConvLayer13isSelectAlgo_E"],"paddle::CudnnConvLayer::maxGroups_":[43,3,1,"_CPPv2N6paddle14CudnnConvLayer10maxGroups_E"],"paddle::CudnnConvLayer::outputDesc_":[43,3,1,"_CPPv2N6paddle14CudnnConvLayer11outputDesc_E"],"paddle::CudnnConvLayer::outputH_":[43,3,1,"_CPPv2N6paddle14CudnnConvLayer8outputH_E"],"paddle::CudnnConvLayer::outputOffset_":[43,3,1,"_CPPv2N6paddle14CudnnConvLayer13outputOffset_E"],"paddle::CudnnConvLayer::outputW_":[43,3,1,"_CPPv2N6paddle14CudnnConvLayer8outputW_E"],"paddle::CudnnConvLayer::reshape":[43,2,1,"_CPPv2N6paddle14CudnnConvLayer7reshapeEi"],"paddle::CudnnConvLayer::weightOffset_":[43,3,1,"_CPPv2N6paddle14CudnnConvLayer13weightOffset_E"],"paddle::CudnnConvLayer::workSpaceData_":[43,3,1,"_CPPv2N6paddle14CudnnConvLayer14workSpaceData_E"],"paddle::CudnnConvLayer::workSpaceInBytes_":[43,3,1,"_CPPv2N6paddle14CudnnConvLayer17workSpaceInBytes_E"],"paddle::CudnnConvLayer::workSpace_":[43,3,1,"_CPPv2N6paddle14CudnnConvLayer10workSpace_E"],"paddle::CudnnConvLayer::~CudnnConvLayer":[43,2,1,"_CPPv2N6paddle14CudnnConvLayerD0Ev"],"paddle::CudnnPoolLayer":[43,1,1,"_CPPv2N6paddle14CudnnPoolLayerE"],"paddle::CudnnPoolLayer::CudnnPoolLayer":[43,2,1,"_CPPv2N6paddle14CudnnPoolLayer14CudnnPoolLayerERK11LayerConfig"],"paddle::CudnnPoolLayer::backward":[43,2,1,"_CPPv2N6paddle14CudnnPoolLayer8backwardERK14UpdateCallback"],"paddle::CudnnPoolLayer::forward":[43,2,1,"_CPPv2N6paddle14CudnnPoolLayer7forwardE8PassType"],"paddle::CudnnPoolLayer::heightPadding":[43,3,1,"_CPPv2N6paddle14CudnnPoolLayer13heightPaddingE"],"paddle::CudnnPoolLayer::imageH_":[43,3,1,"_CPPv2N6paddle14CudnnPoolLayer7imageH_E"],"paddle::CudnnPoolLayer::imageW_":[43,3,1,"_CPPv2N6paddle14CudnnPoolLayer7imageW_E"],"paddle::CudnnPoolLayer::init":[43,2,1,"_CPPv2N6paddle14CudnnPoolLayer4initERK8LayerMapRK12ParameterMap"],"paddle::CudnnPoolLayer::inputDesc_":[43,3,1,"_CPPv2N6paddle14CudnnPoolLayer10inputDesc_E"],"paddle::CudnnPoolLayer::mode_":[43,3,1,"_CPPv2N6paddle14CudnnPoolLayer5mode_E"],"paddle::CudnnPoolLayer::outputDesc_":[43,3,1,"_CPPv2N6paddle14CudnnPoolLayer11outputDesc_E"],"paddle::CudnnPoolLayer::outputH_":[43,3,1,"_CPPv2N6paddle14CudnnPoolLayer8outputH_E"],"paddle::CudnnPoolLayer::outputSize":[43,2,1,"_CPPv2N6paddle14CudnnPoolLayer10outputSizeEiiii"],"paddle::CudnnPoolLayer::outputW_":[43,3,1,"_CPPv2N6paddle14CudnnPoolLayer8outputW_E"],"paddle::CudnnPoolLayer::poolingDesc_":[43,3,1,"_CPPv2N6paddle14CudnnPoolLayer12poolingDesc_E"],"paddle::CudnnPoolLayer::reshape":[43,2,1,"_CPPv2N6paddle14CudnnPoolLayer7reshapeEi"],"paddle::CudnnPoolLayer::strideHeight":[43,3,1,"_CPPv2N6paddle14CudnnPoolLayer12strideHeightE"],"paddle::CudnnPoolLayer::strideWidth":[43,3,1,"_CPPv2N6paddle14CudnnPoolLayer11strideWidthE"],"paddle::CudnnPoolLayer::typeCheck":[43,2,1,"_CPPv2N6paddle14CudnnPoolLayer9typeCheckERKNSt6stringEP17hl_pooling_mode_t"],"paddle::CudnnPoolLayer::widthPadding":[43,3,1,"_CPPv2N6paddle14CudnnPoolLayer12widthPaddingE"],"paddle::CudnnPoolLayer::windowHeight":[43,3,1,"_CPPv2N6paddle14CudnnPoolLayer12windowHeightE"],"paddle::CudnnPoolLayer::windowWidth":[43,3,1,"_CPPv2N6paddle14CudnnPoolLayer11windowWidthE"],"paddle::CudnnPoolLayer::~CudnnPoolLayer":[43,2,1,"_CPPv2N6paddle14CudnnPoolLayerD0Ev"],"paddle::CustomStackTrace":[62,1,1,"_CPPv2N6paddle16CustomStackTraceE"],"paddle::CustomStackTrace::DumpCallback":[62,4,1,"_CPPv2N6paddle16CustomStackTrace12DumpCallbackE"],"paddle::CustomStackTrace::clear":[62,2,1,"_CPPv2N6paddle16CustomStackTrace5clearEv"],"paddle::CustomStackTrace::dump":[62,2,1,"_CPPv2N6paddle16CustomStackTrace4dumpERK12DumpCallbackb"],"paddle::CustomStackTrace::empty":[62,2,1,"_CPPv2N6paddle16CustomStackTrace5emptyEv"],"paddle::CustomStackTrace::pop":[62,2,1,"_CPPv2N6paddle16CustomStackTrace3popERK1T"],"paddle::CustomStackTrace::push":[62,2,1,"_CPPv2N6paddle16CustomStackTrace4pushERK1T"],"paddle::DATA_AT_CPU":[46,1,1,"_CPPv2N6paddle11DATA_AT_CPUE"],"paddle::DATA_AT_GPU":[46,1,1,"_CPPv2N6paddle11DATA_AT_GPUE"],"paddle::DataLayer":[43,1,1,"_CPPv2N6paddle9DataLayerE"],"paddle::DataLayer::DataLayer":[43,2,1,"_CPPv2N6paddle9DataLayer9DataLayerERK11LayerConfig"],"paddle::DataLayer::backward":[43,2,1,"_CPPv2N6paddle9DataLayer8backwardERK14UpdateCallback"],"paddle::DataLayer::copyOutputToOtherDevice":[43,2,1,"_CPPv2N6paddle9DataLayer23copyOutputToOtherDeviceEv"],"paddle::DataLayer::data_":[43,3,1,"_CPPv2N6paddle9DataLayer5data_E"],"paddle::DataLayer::forward":[43,2,1,"_CPPv2N6paddle9DataLayer7forwardE8PassType"],"paddle::DataLayer::prefetch":[43,2,1,"_CPPv2N6paddle9DataLayer8prefetchEv"],"paddle::DataLayer::setData":[43,2,1,"_CPPv2N6paddle9DataLayer7setDataERK8Argument"],"paddle::DataNormLayer":[43,1,1,"_CPPv2N6paddle13DataNormLayerE"],"paddle::DataNormLayer::DataNormLayer":[43,2,1,"_CPPv2N6paddle13DataNormLayer13DataNormLayerERK11LayerConfig"],"paddle::DataNormLayer::NormalizationStrategy":[43,4,1,"_CPPv2N6paddle13DataNormLayer21NormalizationStrategyE"],"paddle::DataNormLayer::backward":[43,2,1,"_CPPv2N6paddle13DataNormLayer8backwardERK14UpdateCallback"],"paddle::DataNormLayer::decimalReciprocal_":[43,3,1,"_CPPv2N6paddle13DataNormLayer18decimalReciprocal_E"],"paddle::DataNormLayer::forward":[43,2,1,"_CPPv2N6paddle13DataNormLayer7forwardE8PassType"],"paddle::DataNormLayer::init":[43,2,1,"_CPPv2N6paddle13DataNormLayer4initERK8LayerMapRK12ParameterMap"],"paddle::DataNormLayer::kDecimalScaling":[43,1,1,"_CPPv2N6paddle13DataNormLayer15kDecimalScalingE"],"paddle::DataNormLayer::kMinMax":[43,1,1,"_CPPv2N6paddle13DataNormLayer7kMinMaxE"],"paddle::DataNormLayer::kZScore":[43,1,1,"_CPPv2N6paddle13DataNormLayer7kZScoreE"],"paddle::DataNormLayer::mean_":[43,3,1,"_CPPv2N6paddle13DataNormLayer5mean_E"],"paddle::DataNormLayer::min_":[43,3,1,"_CPPv2N6paddle13DataNormLayer4min_E"],"paddle::DataNormLayer::mode_":[43,3,1,"_CPPv2N6paddle13DataNormLayer5mode_E"],"paddle::DataNormLayer::rangeReciprocal_":[43,3,1,"_CPPv2N6paddle13DataNormLayer16rangeReciprocal_E"],"paddle::DataNormLayer::stdReciprocal_":[43,3,1,"_CPPv2N6paddle13DataNormLayer14stdReciprocal_E"],"paddle::DataNormLayer::weight_":[43,3,1,"_CPPv2N6paddle13DataNormLayer7weight_E"],"paddle::DataNormLayer::~DataNormLayer":[43,2,1,"_CPPv2N6paddle13DataNormLayerD0Ev"],"paddle::DataProvider":[36,1,1,"_CPPv2N6paddle12DataProviderE"],"paddle::DataProvider::DataProvider":[36,2,1,"_CPPv2N6paddle12DataProvider12DataProviderERK10DataConfigb"],"paddle::DataProvider::config_":[36,3,1,"_CPPv2N6paddle12DataProvider7config_E"],"paddle::DataProvider::constantSlots_":[36,3,1,"_CPPv2N6paddle12DataProvider14constantSlots_E"],"paddle::DataProvider::create":[36,2,1,"_CPPv2N6paddle12DataProvider6createERK10DataConfigb"],"paddle::DataProvider::doubleBuffer_":[36,3,1,"_CPPv2N6paddle12DataProvider13doubleBuffer_E"],"paddle::DataProvider::getConfig":[36,2,1,"_CPPv2N6paddle12DataProvider9getConfigEv"],"paddle::DataProvider::getNextBatch":[36,2,1,"_CPPv2N6paddle12DataProvider12getNextBatchE7int64_tP9DataBatch"],"paddle::DataProvider::getNextBatchFromBuffer":[36,2,1,"_CPPv2N6paddle12DataProvider22getNextBatchFromBufferE7int64_tP9DataBatch"],"paddle::DataProvider::getNextBatchInternal":[36,2,1,"_CPPv2N6paddle12DataProvider20getNextBatchInternalE7int64_tP9DataBatch"],"paddle::DataProvider::getSize":[36,2,1,"_CPPv2N6paddle12DataProvider7getSizeEv"],"paddle::DataProvider::initAsyncLoader":[36,2,1,"_CPPv2N6paddle12DataProvider15initAsyncLoaderEv"],"paddle::DataProvider::registrar_":[36,3,1,"_CPPv2N6paddle12DataProvider10registrar_E"],"paddle::DataProvider::reset":[36,2,1,"_CPPv2N6paddle12DataProvider5resetEv"],"paddle::DataProvider::setSkipShuffle":[36,2,1,"_CPPv2N6paddle12DataProvider14setSkipShuffleEv"],"paddle::DataProvider::shuffle":[36,2,1,"_CPPv2N6paddle12DataProvider7shuffleEv"],"paddle::DataProvider::skipShuffle_":[36,3,1,"_CPPv2N6paddle12DataProvider12skipShuffle_E"],"paddle::DataProvider::usageRatio_":[36,3,1,"_CPPv2N6paddle12DataProvider11usageRatio_E"],"paddle::DataProvider::useGpu_":[36,3,1,"_CPPv2N6paddle12DataProvider7useGpu_E"],"paddle::DataProvider::~DataProvider":[36,2,1,"_CPPv2N6paddle12DataProviderD0Ev"],"paddle::DataProviderGroup":[36,1,1,"_CPPv2N6paddle17DataProviderGroupE"],"paddle::DataProviderGroup::DataProviderGroup":[36,2,1,"_CPPv2N6paddle17DataProviderGroup17DataProviderGroupERK10DataConfigb"],"paddle::DataProviderGroup::ProviderPtrType":[36,4,1,"_CPPv2N6paddle17DataProviderGroup15ProviderPtrTypeE"],"paddle::DataProviderGroup::ProviderType":[36,4,1,"_CPPv2N6paddle17DataProviderGroup12ProviderTypeE"],"paddle::DataProviderGroup::fileList_":[36,3,1,"_CPPv2N6paddle17DataProviderGroup9fileList_E"],"paddle::DataProviderGroup::getNextBatchInternal":[36,2,1,"_CPPv2N6paddle17DataProviderGroup20getNextBatchInternalE7int64_tP9DataBatch"],"paddle::DataProviderGroup::getSize":[36,2,1,"_CPPv2N6paddle17DataProviderGroup7getSizeEv"],"paddle::DataProviderGroup::loader_":[36,3,1,"_CPPv2N6paddle17DataProviderGroup7loader_E"],"paddle::DataProviderGroup::lock_":[36,3,1,"_CPPv2N6paddle17DataProviderGroup5lock_E"],"paddle::DataProviderGroup::provider_":[36,3,1,"_CPPv2N6paddle17DataProviderGroup9provider_E"],"paddle::DataProviderGroup::reset":[36,2,1,"_CPPv2N6paddle17DataProviderGroup5resetEv"],"paddle::DataProviderGroup::shuffle":[36,2,1,"_CPPv2N6paddle17DataProviderGroup7shuffleEv"],"paddle::DataProviderGroup::~DataProviderGroup":[36,2,1,"_CPPv2N6paddle17DataProviderGroupD0Ev"],"paddle::DecayedAdagradParameterOptimizer":[50,1,1,"_CPPv2N6paddle32DecayedAdagradParameterOptimizerE"],"paddle::DecayedAdagradParameterOptimizer::DecayedAdagradParameterOptimizer":[50,2,1,"_CPPv2N6paddle32DecayedAdagradParameterOptimizer32DecayedAdagradParameterOptimizerERK18OptimizationConfig"],"paddle::DecayedAdagradParameterOptimizer::epsilon_":[50,3,1,"_CPPv2N6paddle32DecayedAdagradParameterOptimizer8epsilon_E"],"paddle::DecayedAdagradParameterOptimizer::finishBatch":[50,2,1,"_CPPv2N6paddle32DecayedAdagradParameterOptimizer11finishBatchEv"],"paddle::DecayedAdagradParameterOptimizer::init":[50,2,1,"_CPPv2N6paddle32DecayedAdagradParameterOptimizer4initE6size_tPK15ParameterConfig"],"paddle::DecayedAdagradParameterOptimizer::rou_":[50,3,1,"_CPPv2N6paddle32DecayedAdagradParameterOptimizer4rou_E"],"paddle::DecayedAdagradParameterOptimizer::startBatch":[50,2,1,"_CPPv2N6paddle32DecayedAdagradParameterOptimizer10startBatchE7int64_t"],"paddle::DecayedAdagradParameterOptimizer::t0Vec_":[50,3,1,"_CPPv2N6paddle32DecayedAdagradParameterOptimizer6t0Vec_E"],"paddle::DecayedAdagradParameterOptimizer::timer_":[50,3,1,"_CPPv2N6paddle32DecayedAdagradParameterOptimizer6timer_E"],"paddle::DecayedAdagradParameterOptimizer::update":[50,2,1,"_CPPv2N6paddle32DecayedAdagradParameterOptimizer6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::DenseScanner":[36,1,1,"_CPPv2N6paddle12DenseScannerE"],"paddle::DenseScanner::DenseScanner":[36,2,1,"_CPPv2N6paddle12DenseScanner12DenseScannerEP10SlotHeader"],"paddle::DenseScanner::fill":[36,2,1,"_CPPv2N6paddle12DenseScanner4fillER8ArgumentP8PyObject"],"paddle::DenseScanner::finishPrepare":[36,2,1,"_CPPv2N6paddle12DenseScanner13finishPrepareER8Argument"],"paddle::DenseScanner::prepare":[36,2,1,"_CPPv2N6paddle12DenseScanner7prepareER8ArgumentP8PyObject"],"paddle::DotMulOperator":[43,1,1,"_CPPv2N6paddle14DotMulOperatorE"],"paddle::DotMulOperator::DotMulOperator":[43,2,1,"_CPPv2N6paddle14DotMulOperator14DotMulOperatorERK14OperatorConfigb"],"paddle::DotMulOperator::backward":[43,2,1,"_CPPv2N6paddle14DotMulOperator8backwardEv"],"paddle::DotMulOperator::forward":[43,2,1,"_CPPv2N6paddle14DotMulOperator7forwardEv"],"paddle::DotMulProjection":[43,1,1,"_CPPv2N6paddle16DotMulProjectionE"],"paddle::DotMulProjection::DotMulProjection":[43,2,1,"_CPPv2N6paddle16DotMulProjection16DotMulProjectionERK16ProjectionConfigRK12ParameterPtrb"],"paddle::DotMulProjection::backward":[43,2,1,"_CPPv2N6paddle16DotMulProjection8backwardERK14UpdateCallback"],"paddle::DotMulProjection::forward":[43,2,1,"_CPPv2N6paddle16DotMulProjection7forwardEv"],"paddle::DotMulProjection::weight_":[43,3,1,"_CPPv2N6paddle16DotMulProjection7weight_E"],"paddle::DummyOptimizer":[50,1,1,"_CPPv2N6paddle14DummyOptimizerE"],"paddle::DummyOptimizer::DummyOptimizer":[50,2,1,"_CPPv2N6paddle14DummyOptimizer14DummyOptimizerERK18OptimizationConfig"],"paddle::DummyOptimizer::update":[50,2,1,"_CPPv2N6paddle14DummyOptimizer6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::EosIdCheckLayer":[43,1,1,"_CPPv2N6paddle15EosIdCheckLayerE"],"paddle::EosIdCheckLayer::EosIdCheckLayer":[43,2,1,"_CPPv2N6paddle15EosIdCheckLayer15EosIdCheckLayerERK11LayerConfig"],"paddle::EosIdCheckLayer::backward":[43,2,1,"_CPPv2N6paddle15EosIdCheckLayer8backwardERK14UpdateCallback"],"paddle::EosIdCheckLayer::forward":[43,2,1,"_CPPv2N6paddle15EosIdCheckLayer7forwardE8PassType"],"paddle::EosIdCheckLayer::init":[43,2,1,"_CPPv2N6paddle15EosIdCheckLayer4initERK8LayerMapRK12ParameterMap"],"paddle::Evaluator":[38,1,1,"_CPPv2N6paddle9EvaluatorE"],"paddle::Evaluator::Evaluator":[38,2,1,"_CPPv2N6paddle9Evaluator9EvaluatorEv"],"paddle::Evaluator::config_":[38,3,1,"_CPPv2N6paddle9Evaluator7config_E"],"paddle::Evaluator::create":[38,2,1,"_CPPv2N6paddle9Evaluator6createERK15EvaluatorConfig"],"paddle::Evaluator::distributeEval":[38,2,1,"_CPPv2N6paddle9Evaluator14distributeEvalEP16ParameterClient2"],"paddle::Evaluator::eval":[38,2,1,"_CPPv2N6paddle9Evaluator4evalERK13NeuralNetwork"],"paddle::Evaluator::evalImp":[38,2,1,"_CPPv2N6paddle9Evaluator7evalImpERNSt6vectorI8ArgumentEE"],"paddle::Evaluator::finish":[38,2,1,"_CPPv2N6paddle9Evaluator6finishEv"],"paddle::Evaluator::init":[38,2,1,"_CPPv2N6paddle9Evaluator4initERK15EvaluatorConfig"],"paddle::Evaluator::mergeResultsOfAllClients":[38,2,1,"_CPPv2N6paddle9Evaluator24mergeResultsOfAllClientsEP16ParameterClient2"],"paddle::Evaluator::numSamples_":[38,3,1,"_CPPv2N6paddle9Evaluator11numSamples_E"],"paddle::Evaluator::printStats":[38,2,1,"_CPPv2N6paddle9Evaluator10printStatsERNSt7ostreamE"],"paddle::Evaluator::registrar_":[38,3,1,"_CPPv2N6paddle9Evaluator10registrar_E"],"paddle::Evaluator::start":[38,2,1,"_CPPv2N6paddle9Evaluator5startEv"],"paddle::Evaluator::totalScore_":[38,3,1,"_CPPv2N6paddle9Evaluator11totalScore_E"],"paddle::Evaluator::updateSamplesNum":[38,2,1,"_CPPv2N6paddle9Evaluator16updateSamplesNumERKNSt6vectorI8ArgumentEE"],"paddle::Evaluator::~Evaluator":[38,2,1,"_CPPv2N6paddle9EvaluatorD0Ev"],"paddle::ExpandConvLayer":[43,1,1,"_CPPv2N6paddle15ExpandConvLayerE"],"paddle::ExpandConvLayer::ExpandConvLayer":[43,2,1,"_CPPv2N6paddle15ExpandConvLayer15ExpandConvLayerERK11LayerConfig"],"paddle::ExpandConvLayer::addSharedBias":[43,2,1,"_CPPv2N6paddle15ExpandConvLayer13addSharedBiasEv"],"paddle::ExpandConvLayer::addUnsharedBias":[43,2,1,"_CPPv2N6paddle15ExpandConvLayer15addUnsharedBiasEv"],"paddle::ExpandConvLayer::backward":[43,2,1,"_CPPv2N6paddle15ExpandConvLayer8backwardERK14UpdateCallback"],"paddle::ExpandConvLayer::bpropActs":[43,2,1,"_CPPv2N6paddle15ExpandConvLayer9bpropActsE9MatrixPtri"],"paddle::ExpandConvLayer::bpropBiases":[43,2,1,"_CPPv2N6paddle15ExpandConvLayer11bpropBiasesE9MatrixPtr"],"paddle::ExpandConvLayer::bpropSharedBias":[43,2,1,"_CPPv2N6paddle15ExpandConvLayer15bpropSharedBiasE9MatrixPtr9MatrixPtr"],"paddle::ExpandConvLayer::bpropWeights":[43,2,1,"_CPPv2N6paddle15ExpandConvLayer12bpropWeightsE9MatrixPtri"],"paddle::ExpandConvLayer::expandFwdOnce":[43,2,1,"_CPPv2N6paddle15ExpandConvLayer13expandFwdOnceE9MatrixPtrii"],"paddle::ExpandConvLayer::expandInput_":[43,3,1,"_CPPv2N6paddle15ExpandConvLayer12expandInput_E"],"paddle::ExpandConvLayer::expandOneFrame":[43,2,1,"_CPPv2N6paddle15ExpandConvLayer14expandOneFrameE9MatrixPtr6size_ti"],"paddle::ExpandConvLayer::forward":[43,2,1,"_CPPv2N6paddle15ExpandConvLayer7forwardE8PassType"],"paddle::ExpandConvLayer::getSize":[43,2,1,"_CPPv2N6paddle15ExpandConvLayer7getSizeEv"],"paddle::ExpandConvLayer::imgSizeH_":[43,3,1,"_CPPv2N6paddle15ExpandConvLayer9imgSizeH_E"],"paddle::ExpandConvLayer::imgSizeW_":[43,3,1,"_CPPv2N6paddle15ExpandConvLayer9imgSizeW_E"],"paddle::ExpandConvLayer::init":[43,2,1,"_CPPv2N6paddle15ExpandConvLayer4initERK8LayerMapRK12ParameterMap"],"paddle::ExpandConvLayer::outputH_":[43,3,1,"_CPPv2N6paddle15ExpandConvLayer8outputH_E"],"paddle::ExpandConvLayer::outputW_":[43,3,1,"_CPPv2N6paddle15ExpandConvLayer8outputW_E"],"paddle::ExpandConvLayer::resetConvOutput":[43,2,1,"_CPPv2N6paddle15ExpandConvLayer15resetConvOutputE6size_ti"],"paddle::ExpandConvLayer::resetExpandInput":[43,2,1,"_CPPv2N6paddle15ExpandConvLayer16resetExpandInputE6size_t6size_t"],"paddle::ExpandConvLayer::subK_":[43,3,1,"_CPPv2N6paddle15ExpandConvLayer5subK_E"],"paddle::ExpandConvLayer::subM_":[43,3,1,"_CPPv2N6paddle15ExpandConvLayer5subM_E"],"paddle::ExpandConvLayer::subN_":[43,3,1,"_CPPv2N6paddle15ExpandConvLayer5subN_E"],"paddle::ExpandConvLayer::transOutValue_":[43,3,1,"_CPPv2N6paddle15ExpandConvLayer14transOutValue_E"],"paddle::ExpandConvLayer::~ExpandConvLayer":[43,2,1,"_CPPv2N6paddle15ExpandConvLayerD0Ev"],"paddle::ExpandLayer":[43,1,1,"_CPPv2N6paddle11ExpandLayerE"],"paddle::ExpandLayer::ExpandLayer":[43,2,1,"_CPPv2N6paddle11ExpandLayer11ExpandLayerERK11LayerConfig"],"paddle::ExpandLayer::ExpandLevel":[43,4,1,"_CPPv2N6paddle11ExpandLayer11ExpandLevelE"],"paddle::ExpandLayer::backward":[43,2,1,"_CPPv2N6paddle11ExpandLayer8backwardERK14UpdateCallback"],"paddle::ExpandLayer::biases_":[43,3,1,"_CPPv2N6paddle11ExpandLayer7biases_E"],"paddle::ExpandLayer::cpuExpandStartsPos_":[43,3,1,"_CPPv2N6paddle11ExpandLayer19cpuExpandStartsPos_E"],"paddle::ExpandLayer::expandStartsPos_":[43,3,1,"_CPPv2N6paddle11ExpandLayer16expandStartsPos_E"],"paddle::ExpandLayer::forward":[43,2,1,"_CPPv2N6paddle11ExpandLayer7forwardE8PassType"],"paddle::ExpandLayer::init":[43,2,1,"_CPPv2N6paddle11ExpandLayer4initERK8LayerMapRK12ParameterMap"],"paddle::ExpandLayer::kNonSeq":[43,1,1,"_CPPv2N6paddle11ExpandLayer7kNonSeqE"],"paddle::ExpandLayer::kSeq":[43,1,1,"_CPPv2N6paddle11ExpandLayer4kSeqE"],"paddle::ExpandLayer::type_":[43,3,1,"_CPPv2N6paddle11ExpandLayer5type_E"],"paddle::ExpandLayer::~ExpandLayer":[43,2,1,"_CPPv2N6paddle11ExpandLayerD0Ev"],"paddle::FLOAT_VALUE":[46,1,1,"_CPPv2N6paddle11FLOAT_VALUEE"],"paddle::FeatureMapExpandLayer":[43,1,1,"_CPPv2N6paddle21FeatureMapExpandLayerE"],"paddle::FeatureMapExpandLayer::FeatureMapExpandLayer":[43,2,1,"_CPPv2N6paddle21FeatureMapExpandLayer21FeatureMapExpandLayerERK11LayerConfig"],"paddle::FeatureMapExpandLayer::backward":[43,2,1,"_CPPv2N6paddle21FeatureMapExpandLayer8backwardERK14UpdateCallback"],"paddle::FeatureMapExpandLayer::forward":[43,2,1,"_CPPv2N6paddle21FeatureMapExpandLayer7forwardE8PassType"],"paddle::FeatureMapExpandLayer::init":[43,2,1,"_CPPv2N6paddle21FeatureMapExpandLayer4initERK8LayerMapRK12ParameterMap"],"paddle::FeatureMapExpandLayer::~FeatureMapExpandLayer":[43,2,1,"_CPPv2N6paddle21FeatureMapExpandLayerD0Ev"],"paddle::FullMatrixProjection":[43,1,1,"_CPPv2N6paddle20FullMatrixProjectionE"],"paddle::FullMatrixProjection::FullMatrixProjection":[43,2,1,"_CPPv2N6paddle20FullMatrixProjection20FullMatrixProjectionERK16ProjectionConfigRK12ParameterPtrb"],"paddle::FullMatrixProjection::backward":[43,2,1,"_CPPv2N6paddle20FullMatrixProjection8backwardERK14UpdateCallback"],"paddle::FullMatrixProjection::forward":[43,2,1,"_CPPv2N6paddle20FullMatrixProjection7forwardEv"],"paddle::FullMatrixProjection::weight_":[43,3,1,"_CPPv2N6paddle20FullMatrixProjection7weight_E"],"paddle::FullyConnectedLayer":[43,1,1,"_CPPv2N6paddle19FullyConnectedLayerE"],"paddle::FullyConnectedLayer::FullyConnectedLayer":[43,2,1,"_CPPv2N6paddle19FullyConnectedLayer19FullyConnectedLayerERK11LayerConfig"],"paddle::FullyConnectedLayer::backward":[43,2,1,"_CPPv2N6paddle19FullyConnectedLayer8backwardERK14UpdateCallback"],"paddle::FullyConnectedLayer::biases_":[43,3,1,"_CPPv2N6paddle19FullyConnectedLayer7biases_E"],"paddle::FullyConnectedLayer::forward":[43,2,1,"_CPPv2N6paddle19FullyConnectedLayer7forwardE8PassType"],"paddle::FullyConnectedLayer::getWeight":[43,2,1,"_CPPv2N6paddle19FullyConnectedLayer9getWeightEi"],"paddle::FullyConnectedLayer::init":[43,2,1,"_CPPv2N6paddle19FullyConnectedLayer4initERK8LayerMapRK12ParameterMap"],"paddle::FullyConnectedLayer::prefetch":[43,2,1,"_CPPv2N6paddle19FullyConnectedLayer8prefetchEv"],"paddle::FullyConnectedLayer::weights_":[43,3,1,"_CPPv2N6paddle19FullyConnectedLayer8weights_E"],"paddle::FullyConnectedLayer::~FullyConnectedLayer":[43,2,1,"_CPPv2N6paddle19FullyConnectedLayerD0Ev"],"paddle::GatedRecurrentLayer":[43,1,1,"_CPPv2N6paddle19GatedRecurrentLayerE"],"paddle::GatedRecurrentLayer::GatedRecurrentLayer":[43,2,1,"_CPPv2N6paddle19GatedRecurrentLayer19GatedRecurrentLayerERK11LayerConfig"],"paddle::GatedRecurrentLayer::activationGate_":[43,3,1,"_CPPv2N6paddle19GatedRecurrentLayer15activationGate_E"],"paddle::GatedRecurrentLayer::backward":[43,2,1,"_CPPv2N6paddle19GatedRecurrentLayer8backwardERK14UpdateCallback"],"paddle::GatedRecurrentLayer::backwardBatch":[43,2,1,"_CPPv2N6paddle19GatedRecurrentLayer13backwardBatchEi9MatrixPtr"],"paddle::GatedRecurrentLayer::backwardSequence":[43,2,1,"_CPPv2N6paddle19GatedRecurrentLayer16backwardSequenceEi6size_tPKi9MatrixPtr"],"paddle::GatedRecurrentLayer::batchGrad_":[43,3,1,"_CPPv2N6paddle19GatedRecurrentLayer10batchGrad_E"],"paddle::GatedRecurrentLayer::batchValue_":[43,3,1,"_CPPv2N6paddle19GatedRecurrentLayer11batchValue_E"],"paddle::GatedRecurrentLayer::bias_":[43,3,1,"_CPPv2N6paddle19GatedRecurrentLayer5bias_E"],"paddle::GatedRecurrentLayer::forward":[43,2,1,"_CPPv2N6paddle19GatedRecurrentLayer7forwardE8PassType"],"paddle::GatedRecurrentLayer::forwardBatch":[43,2,1,"_CPPv2N6paddle19GatedRecurrentLayer12forwardBatchEi6size_tPKi9MatrixPtr"],"paddle::GatedRecurrentLayer::forwardSequence":[43,2,1,"_CPPv2N6paddle19GatedRecurrentLayer15forwardSequenceEi6size_tPKi9MatrixPtr"],"paddle::GatedRecurrentLayer::gateWeight_":[43,3,1,"_CPPv2N6paddle19GatedRecurrentLayer11gateWeight_E"],"paddle::GatedRecurrentLayer::gate_":[43,3,1,"_CPPv2N6paddle19GatedRecurrentLayer5gate_E"],"paddle::GatedRecurrentLayer::getState":[43,2,1,"_CPPv2N6paddle19GatedRecurrentLayer8getStateEv"],"paddle::GatedRecurrentLayer::init":[43,2,1,"_CPPv2N6paddle19GatedRecurrentLayer4initERK8LayerMapRK12ParameterMap"],"paddle::GatedRecurrentLayer::prevOutput_":[43,3,1,"_CPPv2N6paddle19GatedRecurrentLayer11prevOutput_E"],"paddle::GatedRecurrentLayer::resetOutput_":[43,3,1,"_CPPv2N6paddle19GatedRecurrentLayer12resetOutput_E"],"paddle::GatedRecurrentLayer::resetState":[43,2,1,"_CPPv2N6paddle19GatedRecurrentLayer10resetStateEv"],"paddle::GatedRecurrentLayer::reversed_":[43,3,1,"_CPPv2N6paddle19GatedRecurrentLayer9reversed_E"],"paddle::GatedRecurrentLayer::setState":[43,2,1,"_CPPv2N6paddle19GatedRecurrentLayer8setStateE13LayerStatePtr"],"paddle::GatedRecurrentLayer::stateWeight_":[43,3,1,"_CPPv2N6paddle19GatedRecurrentLayer12stateWeight_E"],"paddle::GatedRecurrentLayer::useBatch_":[43,3,1,"_CPPv2N6paddle19GatedRecurrentLayer9useBatch_E"],"paddle::GatedRecurrentLayer::weight_":[43,3,1,"_CPPv2N6paddle19GatedRecurrentLayer7weight_E"],"paddle::GatherAgentLayer":[43,1,1,"_CPPv2N6paddle16GatherAgentLayerE"],"paddle::GatherAgentLayer::GatherAgentLayer":[43,2,1,"_CPPv2N6paddle16GatherAgentLayer16GatherAgentLayerERK11LayerConfig"],"paddle::GatherAgentLayer::addRealLayer":[43,2,1,"_CPPv2N6paddle16GatherAgentLayer12addRealLayerE8LayerPtr"],"paddle::GatherAgentLayer::allIds_":[43,3,1,"_CPPv2N6paddle16GatherAgentLayer7allIds_E"],"paddle::GatherAgentLayer::backward":[43,2,1,"_CPPv2N6paddle16GatherAgentLayer8backwardERK14UpdateCallback"],"paddle::GatherAgentLayer::copyIdAndSequenceInfo":[43,2,1,"_CPPv2N6paddle16GatherAgentLayer21copyIdAndSequenceInfoERK8ArgumentRK10IVectorPtrRKNSt6vectorIiEE"],"paddle::GatherAgentLayer::forward":[43,2,1,"_CPPv2N6paddle16GatherAgentLayer7forwardE8PassType"],"paddle::GatherAgentLayer::idIndex_":[43,3,1,"_CPPv2N6paddle16GatherAgentLayer8idIndex_E"],"paddle::GatherAgentLayer::idsVec_":[43,3,1,"_CPPv2N6paddle16GatherAgentLayer7idsVec_E"],"paddle::GatherAgentLayer::init":[43,2,1,"_CPPv2N6paddle16GatherAgentLayer4initERK8LayerMapRK12ParameterMap"],"paddle::GatherAgentLayer::realLayers_":[43,3,1,"_CPPv2N6paddle16GatherAgentLayer11realLayers_E"],"paddle::GatherAgentLayer::~GatherAgentLayer":[43,2,1,"_CPPv2N6paddle16GatherAgentLayerD0Ev"],"paddle::GetOutputLayer":[43,1,1,"_CPPv2N6paddle14GetOutputLayerE"],"paddle::GetOutputLayer::GetOutputLayer":[43,2,1,"_CPPv2N6paddle14GetOutputLayer14GetOutputLayerERK11LayerConfig"],"paddle::GetOutputLayer::backward":[43,2,1,"_CPPv2N6paddle14GetOutputLayer8backwardERK14UpdateCallback"],"paddle::GetOutputLayer::forward":[43,2,1,"_CPPv2N6paddle14GetOutputLayer7forwardE8PassType"],"paddle::GetOutputLayer::init":[43,2,1,"_CPPv2N6paddle14GetOutputLayer4initERK8LayerMapRK12ParameterMap"],"paddle::GetOutputLayer::~GetOutputLayer":[43,2,1,"_CPPv2N6paddle14GetOutputLayerD0Ev"],"paddle::GpuAllocator":[48,1,1,"_CPPv2N6paddle12GpuAllocatorE"],"paddle::GpuAllocator::alloc":[48,2,1,"_CPPv2N6paddle12GpuAllocator5allocE6size_t"],"paddle::GpuAllocator::free":[48,2,1,"_CPPv2N6paddle12GpuAllocator4freeEPv"],"paddle::GpuAllocator::getName":[48,2,1,"_CPPv2N6paddle12GpuAllocator7getNameEv"],"paddle::GpuAllocator::~GpuAllocator":[48,2,1,"_CPPv2N6paddle12GpuAllocatorD0Ev"],"paddle::GpuIVector":[46,4,1,"_CPPv2N6paddle10GpuIVectorE"],"paddle::GpuIVectorPtr":[46,4,1,"_CPPv2N6paddle13GpuIVectorPtrE"],"paddle::GpuMatrix":[46,1,1,"_CPPv2N6paddle9GpuMatrixE"],"paddle::GpuMatrix::GpuMatrix":[46,2,1,"_CPPv2N6paddle9GpuMatrix9GpuMatrixEv"],"paddle::GpuMatrix::accumulateColSum":[46,2,1,"_CPPv2N6paddle9GpuMatrix16accumulateColSumER6Matrix"],"paddle::GpuMatrix::addBias":[46,2,1,"_CPPv2N6paddle9GpuMatrix7addBiasER6Matrix4real"],"paddle::GpuMatrix::addColumnVector":[46,2,1,"_CPPv2N6paddle9GpuMatrix15addColumnVectorERK6Matrix"],"paddle::GpuMatrix::addToRows":[46,2,1,"_CPPv2N6paddle9GpuMatrix9addToRowsER6MatrixR7IVector"],"paddle::GpuMatrix::avgPoolBackward":[46,2,1,"_CPPv2N6paddle9GpuMatrix15avgPoolBackwardER6Matrix6size_t6size_t6size_ti6size_t6size_t6size_t4real4real"],"paddle::GpuMatrix::avgPoolForward":[46,2,1,"_CPPv2N6paddle9GpuMatrix14avgPoolForwardER6Matrix6size_t6size_t6size_t6size_ti6size_t6size_t6size_t"],"paddle::GpuMatrix::check":[46,2,1,"_CPPv2N6paddle9GpuMatrix5checkERNSt7ostreamER6Matrixb"],"paddle::GpuMatrix::classificationError":[46,2,1,"_CPPv2N6paddle9GpuMatrix19classificationErrorE9MatrixPtr10IVectorPtr"],"paddle::GpuMatrix::clone":[46,2,1,"_CPPv2N6paddle9GpuMatrix5cloneE6size_t6size_tb"],"paddle::GpuMatrix::colMax":[46,2,1,"_CPPv2N6paddle9GpuMatrix6colMaxER6Matrix"],"paddle::GpuMatrix::colMerge":[46,2,1,"_CPPv2N6paddle9GpuMatrix8colMergeER6Matrix"],"paddle::GpuMatrix::collectBias":[46,2,1,"_CPPv2N6paddle9GpuMatrix11collectBiasER6Matrix4real"],"paddle::GpuMatrix::contextProjectionBackwardData":[46,2,1,"_CPPv2N6paddle9GpuMatrix29contextProjectionBackwardDataE9MatrixPtrRK7IVectorii"],"paddle::GpuMatrix::contextProjectionBackwardWeight":[46,2,1,"_CPPv2N6paddle9GpuMatrix31contextProjectionBackwardWeightE9MatrixPtrRK7IVectoriii6size_t"],"paddle::GpuMatrix::contextProjectionForward":[46,2,1,"_CPPv2N6paddle9GpuMatrix24contextProjectionForwardE9MatrixPtr9MatrixPtrRK7IVectorii6size_tb"],"paddle::GpuMatrix::convExpand":[46,2,1,"_CPPv2N6paddle9GpuMatrix10convExpandER6Matrixiiiiiiiiiii"],"paddle::GpuMatrix::convShrink":[46,2,1,"_CPPv2N6paddle9GpuMatrix10convShrinkER6Matrixiiiiiiiiiii4real4real"],"paddle::GpuMatrix::copyByRowIndex":[46,2,1,"_CPPv2N6paddle9GpuMatrix14copyByRowIndexER6MatrixR7IVector"],"paddle::GpuMatrix::copyFrom":[46,2,1,"_CPPv2N6paddle9GpuMatrix8copyFromERK7IVector"],"paddle::GpuMatrix::cosSim":[46,2,1,"_CPPv2N6paddle9GpuMatrix6cosSimER6MatrixR6Matrix4real"],"paddle::GpuMatrix::cosSimDerivative":[46,2,1,"_CPPv2N6paddle9GpuMatrix16cosSimDerivativeER6MatrixR6MatrixR6MatrixR6MatrixR6Matrix4real"],"paddle::GpuMatrix::crossMapNormalBwd":[46,2,1,"_CPPv2N6paddle9GpuMatrix17crossMapNormalBwdER6MatrixR6MatrixR6MatrixR6Matrix6size_t6size_t6size_t6size_tff"],"paddle::GpuMatrix::crossMapNormalFwd":[46,2,1,"_CPPv2N6paddle9GpuMatrix17crossMapNormalFwdER6Matrix6size_t6size_tR6Matrix6size_t6size_tff"],"paddle::GpuMatrix::getAbsSum":[46,2,1,"_CPPv2N6paddle9GpuMatrix9getAbsSumEv"],"paddle::GpuMatrix::getElement":[46,2,1,"_CPPv2N6paddle9GpuMatrix10getElementE6size_t6size_t"],"paddle::GpuMatrix::getRow":[46,2,1,"_CPPv2N6paddle9GpuMatrix6getRowE6size_t"],"paddle::GpuMatrix::getRowBuf":[46,2,1,"_CPPv2N6paddle9GpuMatrix9getRowBufE6size_t"],"paddle::GpuMatrix::getSum":[46,2,1,"_CPPv2N6paddle9GpuMatrix6getSumEv"],"paddle::GpuMatrix::getTranspose":[46,2,1,"_CPPv2N6paddle9GpuMatrix12getTransposeEv"],"paddle::GpuMatrix::leftMul":[46,2,1,"_CPPv2N6paddle9GpuMatrix7leftMulER6Matrix4real4real"],"paddle::GpuMatrix::maxPoolBackward":[46,2,1,"_CPPv2N6paddle9GpuMatrix15maxPoolBackwardER6Matrix6size_t6size_tR6MatrixR6Matrix6size_ti6size_t6size_t6size_t4real4real"],"paddle::GpuMatrix::maxPoolForward":[46,2,1,"_CPPv2N6paddle9GpuMatrix14maxPoolForwardER6Matrix6size_t6size_t6size_t6size_ti6size_t6size_t6size_t"],"paddle::GpuMatrix::maxSequenceBackward":[46,2,1,"_CPPv2N6paddle9GpuMatrix19maxSequenceBackwardER6MatrixRK7IVectorR7IVector"],"paddle::GpuMatrix::maxSequenceForward":[46,2,1,"_CPPv2N6paddle9GpuMatrix18maxSequenceForwardER6MatrixRK7IVectorR7IVector"],"paddle::GpuMatrix::mul":[46,2,1,"_CPPv2N6paddle9GpuMatrix3mulERK9GpuMatrixRK9GpuMatrix4real4real"],"paddle::GpuMatrix::oneHotCrossEntropy":[46,2,1,"_CPPv2N6paddle9GpuMatrix18oneHotCrossEntropyER6MatrixR7IVector"],"paddle::GpuMatrix::oneHotCrossEntropyBp":[46,2,1,"_CPPv2N6paddle9GpuMatrix20oneHotCrossEntropyBpER6MatrixR7IVector"],"paddle::GpuMatrix::oneHotCrossEntropyWithSelfNorm":[46,2,1,"_CPPv2N6paddle9GpuMatrix30oneHotCrossEntropyWithSelfNormER6MatrixR7IVector4real"],"paddle::GpuMatrix::oneHotCrossEntropyWithSelfNormBp":[46,2,1,"_CPPv2N6paddle9GpuMatrix32oneHotCrossEntropyWithSelfNormBpER6MatrixR7IVector4real"],"paddle::GpuMatrix::paramReluBackwardDiff":[46,2,1,"_CPPv2N6paddle9GpuMatrix21paramReluBackwardDiffER6MatrixR6MatrixR6Matrix"],"paddle::GpuMatrix::paramReluBackwardW":[46,2,1,"_CPPv2N6paddle9GpuMatrix18paramReluBackwardWER6MatrixR6Matrix"],"paddle::GpuMatrix::paramReluForward":[46,2,1,"_CPPv2N6paddle9GpuMatrix16paramReluForwardER6MatrixR6Matrix"],"paddle::GpuMatrix::print":[46,2,1,"_CPPv2N6paddle9GpuMatrix5printERNSt7ostreamE6size_t6size_t"],"paddle::GpuMatrix::randomizeUniform":[46,2,1,"_CPPv2N6paddle9GpuMatrix16randomizeUniformEv"],"paddle::GpuMatrix::resetOne":[46,2,1,"_CPPv2N6paddle9GpuMatrix8resetOneEv"],"paddle::GpuMatrix::resize":[46,2,1,"_CPPv2N6paddle9GpuMatrix6resizeE6size_t6size_t6size_t15SparseValueType12SparseFormat"],"paddle::GpuMatrix::rightMul":[46,2,1,"_CPPv2N6paddle9GpuMatrix8rightMulER6Matrix4real4real"],"paddle::GpuMatrix::rowMax":[46,2,1,"_CPPv2N6paddle9GpuMatrix6rowMaxER7IVectorR6Matrix"],"paddle::GpuMatrix::rowSum":[46,2,1,"_CPPv2N6paddle9GpuMatrix6rowSumER6Matrix"],"paddle::GpuMatrix::scaledTanh":[46,2,1,"_CPPv2N6paddle9GpuMatrix10scaledTanhER6Matrix4real4real"],"paddle::GpuMatrix::selectElements":[46,2,1,"_CPPv2N6paddle9GpuMatrix14selectElementsER6MatrixR7IVector"],"paddle::GpuMatrix::selectRows":[46,2,1,"_CPPv2N6paddle9GpuMatrix10selectRowsER6MatrixR7IVector"],"paddle::GpuMatrix::sequenceAvgForward":[46,2,1,"_CPPv2N6paddle9GpuMatrix18sequenceAvgForwardER6MatrixRK7IVectori"],"paddle::GpuMatrix::sequenceSoftmax":[46,2,1,"_CPPv2N6paddle9GpuMatrix15sequenceSoftmaxER6MatrixRK7IVector"],"paddle::GpuMatrix::setRow":[46,2,1,"_CPPv2N6paddle9GpuMatrix6setRowE6size_t6size_tPKjPK4real"],"paddle::GpuMatrix::softmax":[46,2,1,"_CPPv2N6paddle9GpuMatrix7softmaxER6Matrix"],"paddle::GpuMatrix::softmaxBackward":[46,2,1,"_CPPv2N6paddle9GpuMatrix15softmaxBackwardER6Matrix"],"paddle::GpuMatrix::softmaxDerivative":[46,2,1,"_CPPv2N6paddle9GpuMatrix17softmaxDerivativeER6MatrixR6Matrix"],"paddle::GpuMatrix::softrelu":[46,2,1,"_CPPv2N6paddle9GpuMatrix8softreluER6Matrix"],"paddle::GpuMatrix::softreluDerivative":[46,2,1,"_CPPv2N6paddle9GpuMatrix18softreluDerivativeER6Matrix"],"paddle::GpuMatrix::sumOfSquares":[46,2,1,"_CPPv2N6paddle9GpuMatrix12sumOfSquaresER6MatrixR6Matrix"],"paddle::GpuMatrix::sumOfSquaresBp":[46,2,1,"_CPPv2N6paddle9GpuMatrix14sumOfSquaresBpER6MatrixR6Matrix"],"paddle::GpuMatrix::tanh":[46,2,1,"_CPPv2N6paddle9GpuMatrix4tanhER6Matrix"],"paddle::GpuMatrix::tanhDerivative":[46,2,1,"_CPPv2N6paddle9GpuMatrix14tanhDerivativeER6Matrix"],"paddle::GpuMatrix::transpose":[46,2,1,"_CPPv2N6paddle9GpuMatrix9transposeE9MatrixPtrb"],"paddle::GpuMatrix::zeroMem":[46,2,1,"_CPPv2N6paddle9GpuMatrix7zeroMemEv"],"paddle::GpuMatrix::~GpuMatrix":[46,2,1,"_CPPv2N6paddle9GpuMatrixD0Ev"],"paddle::GpuMatrixPtr":[46,4,1,"_CPPv2N6paddle12GpuMatrixPtrE"],"paddle::GpuMemHandlePtr":[48,4,1,"_CPPv2N6paddle15GpuMemHandlePtrE"],"paddle::GpuMemoryHandle":[48,1,1,"_CPPv2N6paddle15GpuMemoryHandleE"],"paddle::GpuMemoryHandle::GpuMemoryHandle":[48,2,1,"_CPPv2N6paddle15GpuMemoryHandle15GpuMemoryHandleE6size_t"],"paddle::GpuMemoryHandle::~GpuMemoryHandle":[48,2,1,"_CPPv2N6paddle15GpuMemoryHandleD0Ev"],"paddle::GpuSparseMatrix":[46,1,1,"_CPPv2N6paddle15GpuSparseMatrixE"],"paddle::GpuSparseMatrix::Element":[46,1,1,"_CPPv2N6paddle15GpuSparseMatrix7ElementE"],"paddle::GpuSparseMatrix::Element::Element":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix7Element7ElementEii4real"],"paddle::GpuSparseMatrix::Element::col":[46,3,1,"_CPPv2N6paddle15GpuSparseMatrix7Element3colE"],"paddle::GpuSparseMatrix::Element::row":[46,3,1,"_CPPv2N6paddle15GpuSparseMatrix7Element3rowE"],"paddle::GpuSparseMatrix::Element::val":[46,3,1,"_CPPv2N6paddle15GpuSparseMatrix7Element3valE"],"paddle::GpuSparseMatrix::GpuSparseMatrix":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix15GpuSparseMatrixEP4realPiPi6size_t6size_t6size_t15SparseValueType12SparseFormatb"],"paddle::GpuSparseMatrix::add3":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix4add3EP9GpuMatrix"],"paddle::GpuSparseMatrix::addBias":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix7addBiasER6Matrix4real"],"paddle::GpuSparseMatrix::cols_":[46,3,1,"_CPPv2N6paddle15GpuSparseMatrix5cols_E"],"paddle::GpuSparseMatrix::copyFrom":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix8copyFromERK7IVector11hl_stream_t"],"paddle::GpuSparseMatrix::copyFromCSC":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix11copyFromCSCER15CpuSparseMatrix11hl_stream_t"],"paddle::GpuSparseMatrix::copyFromCSR":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix11copyFromCSRER15CpuSparseMatrix11hl_stream_t"],"paddle::GpuSparseMatrix::copyRow":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix7copyRowEi6size_tPK20sparse_float_value_t"],"paddle::GpuSparseMatrix::end_":[46,3,1,"_CPPv2N6paddle15GpuSparseMatrix4end_E"],"paddle::GpuSparseMatrix::format_":[46,3,1,"_CPPv2N6paddle15GpuSparseMatrix7format_E"],"paddle::GpuSparseMatrix::getColNum":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix9getColNumE6size_t"],"paddle::GpuSparseMatrix::getCols":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix7getColsEv"],"paddle::GpuSparseMatrix::getData":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix7getDataEv"],"paddle::GpuSparseMatrix::getFormat":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix9getFormatEv"],"paddle::GpuSparseMatrix::getRowCols":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix10getRowColsE6size_t"],"paddle::GpuSparseMatrix::getRowValues":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix12getRowValuesE6size_t"],"paddle::GpuSparseMatrix::getRows":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix7getRowsEv"],"paddle::GpuSparseMatrix::getTranspose":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix12getTransposeEv"],"paddle::GpuSparseMatrix::getValue":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix8getValueEv"],"paddle::GpuSparseMatrix::getValueType":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix12getValueTypeEv"],"paddle::GpuSparseMatrix::isSparse":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix8isSparseEv"],"paddle::GpuSparseMatrix::mul":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix3mulEK9MatrixPtrK9MatrixPtr4real4real"],"paddle::GpuSparseMatrix::print":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix5printERNSt7ostreamE"],"paddle::GpuSparseMatrix::resize":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix6resizeE6size_t6size_t6size_t15SparseValueType12SparseFormat"],"paddle::GpuSparseMatrix::resizeCSC":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix9resizeCSCE6size_t6size_t6size_t15SparseValueType"],"paddle::GpuSparseMatrix::resizeCSR":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix9resizeCSRE6size_t6size_t6size_t15SparseValueType"],"paddle::GpuSparseMatrix::rowMax":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix6rowMaxER7IVectorR6Matrix"],"paddle::GpuSparseMatrix::rows_":[46,3,1,"_CPPv2N6paddle15GpuSparseMatrix5rows_E"],"paddle::GpuSparseMatrix::sMatrix_":[46,3,1,"_CPPv2N6paddle15GpuSparseMatrix8sMatrix_E"],"paddle::GpuSparseMatrix::sMemoryHandle_":[46,3,1,"_CPPv2N6paddle15GpuSparseMatrix14sMemoryHandle_E"],"paddle::GpuSparseMatrix::setRow":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix6setRowE6size_t6size_tPKjPK4real"],"paddle::GpuSparseMatrix::sparseResize":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix12sparseResizeEv"],"paddle::GpuSparseMatrix::sparseResizeCSC":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix15sparseResizeCSCEv"],"paddle::GpuSparseMatrix::sparseResizeCSR":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix15sparseResizeCSREv"],"paddle::GpuSparseMatrix::transpose":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix9transposeE9MatrixPtrb"],"paddle::GpuSparseMatrix::trimFrom":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix8trimFromERK15CpuSparseMatrix"],"paddle::GpuSparseMatrix::trimFromCSC":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix11trimFromCSCERK15CpuSparseMatrix"],"paddle::GpuSparseMatrix::trimFromCSR":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix11trimFromCSRERK15CpuSparseMatrix"],"paddle::GpuSparseMatrix::valueType_":[46,3,1,"_CPPv2N6paddle15GpuSparseMatrix10valueType_E"],"paddle::GpuSparseMatrix::value_":[46,3,1,"_CPPv2N6paddle15GpuSparseMatrix6value_E"],"paddle::GpuSparseMatrix::zeroMem":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrix7zeroMemEv"],"paddle::GpuSparseMatrix::~GpuSparseMatrix":[46,2,1,"_CPPv2N6paddle15GpuSparseMatrixD0Ev"],"paddle::GpuSparseMatrixPtr":[46,4,1,"_CPPv2N6paddle18GpuSparseMatrixPtrE"],"paddle::GpuVector":[46,4,1,"_CPPv2N6paddle9GpuVectorE"],"paddle::GpuVectorPtr":[46,4,1,"_CPPv2N6paddle12GpuVectorPtrE"],"paddle::GpuVectorT":[46,1,1,"_CPPv2N6paddle10GpuVectorTE"],"paddle::GpuVectorT::GpuVectorT":[46,2,1,"_CPPv2N6paddle10GpuVectorT10GpuVectorTE6size_tP1T"],"paddle::GpuVectorT::copyFrom":[46,2,1,"_CPPv2N6paddle10GpuVectorT8copyFromERK7VectorTI1TE11hl_stream_t"],"paddle::GpuVectorT::copyTo":[46,2,1,"_CPPv2N6paddle10GpuVectorT6copyToEP10GpuVectorTI1TE"],"paddle::GpuVectorT::fillSequence":[46,2,1,"_CPPv2N6paddle10GpuVectorT12fillSequenceEv"],"paddle::GpuVectorT::get":[46,2,1,"_CPPv2N6paddle10GpuVectorT3getE6size_t"],"paddle::GpuVectorT::getAbsMax":[46,2,1,"_CPPv2N6paddle10GpuVectorT9getAbsMaxEv"],"paddle::GpuVectorT::getAbsSum":[46,2,1,"_CPPv2N6paddle10GpuVectorT9getAbsSumEv"],"paddle::GpuVectorT::getElement":[46,2,1,"_CPPv2N6paddle10GpuVectorT10getElementE6size_t"],"paddle::GpuVectorT::getMax":[46,2,1,"_CPPv2N6paddle10GpuVectorT6getMaxEv"],"paddle::GpuVectorT::getMin":[46,2,1,"_CPPv2N6paddle10GpuVectorT6getMinEv"],"paddle::GpuVectorT::getPoint":[46,2,1,"_CPPv2N6paddle10GpuVectorT8getPointEK8uint64_t"],"paddle::GpuVectorT::getSum":[46,2,1,"_CPPv2N6paddle10GpuVectorT6getSumEv"],"paddle::GpuVectorT::histogram":[46,2,1,"_CPPv2N6paddle10GpuVectorT9histogramERNSt7ostreamEi"],"paddle::GpuVectorT::isEqualTo":[46,2,1,"_CPPv2N6paddle10GpuVectorT9isEqualToERK7VectorTI1TERK1T"],"paddle::GpuVectorT::newMemory":[46,2,1,"_CPPv2N6paddle10GpuVectorT9newMemoryE6size_t"],"paddle::GpuVectorT::print":[46,2,1,"_CPPv2N6paddle10GpuVectorT5printERNSt7ostreamE6size_t"],"paddle::GpuVectorT::printOneElement":[46,2,1,"_CPPv2N6paddle10GpuVectorT15printOneElementERNSt7ostreamE6size_t"],"paddle::GpuVectorT::rand":[46,2,1,"_CPPv2N6paddle10GpuVectorT4randEv"],"paddle::GpuVectorT::randnorm":[46,2,1,"_CPPv2N6paddle10GpuVectorT8randnormE4real4real"],"paddle::GpuVectorT::reset":[46,2,1,"_CPPv2N6paddle10GpuVectorT5resetERK1T"],"paddle::GpuVectorT::selectFrom":[46,2,1,"_CPPv2N6paddle10GpuVectorT10selectFromERK7VectorTI1TERK7VectorTIiE"],"paddle::GpuVectorT::setElement":[46,2,1,"_CPPv2N6paddle10GpuVectorT10setElementE6size_tRK1T"],"paddle::GpuVectorT::uniform":[46,2,1,"_CPPv2N6paddle10GpuVectorT7uniformE4real4real"],"paddle::GpuVectorT::zeroMem":[46,2,1,"_CPPv2N6paddle10GpuVectorT7zeroMemEv"],"paddle::GradientMachine":[40,1,1,"_CPPv2N6paddle15GradientMachineE"],"paddle::GradientMachine::CreateMode":[40,4,1,"_CPPv2N6paddle15GradientMachine10CreateModeE"],"paddle::GradientMachine::backward":[40,2,1,"_CPPv2N6paddle15GradientMachine8backwardERK14UpdateCallback"],"paddle::GradientMachine::eval":[40,2,1,"_CPPv2N6paddle15GradientMachine4evalEP9Evaluator"],"paddle::GradientMachine::finish":[40,2,1,"_CPPv2N6paddle15GradientMachine6finishEv"],"paddle::GradientMachine::forward":[40,2,1,"_CPPv2N6paddle15GradientMachine7forwardERKNSt6vectorI8ArgumentEEPNSt6vectorI8ArgumentEE8PassType"],"paddle::GradientMachine::forwardBackward":[40,2,1,"_CPPv2N6paddle15GradientMachine15forwardBackwardERKNSt6vectorI8ArgumentEEPNSt6vectorI8ArgumentEE8PassTypeRK14UpdateCallback"],"paddle::GradientMachine::getNonStaticParameters":[40,2,1,"_CPPv2N6paddle15GradientMachine22getNonStaticParametersEv"],"paddle::GradientMachine::getParameters":[40,2,1,"_CPPv2N6paddle15GradientMachine13getParametersEv"],"paddle::GradientMachine::getState":[40,2,1,"_CPPv2N6paddle15GradientMachine8getStateER12MachineState"],"paddle::GradientMachine::getStats":[40,2,1,"_CPPv2N6paddle15GradientMachine8getStatsER4realR7int64_t"],"paddle::GradientMachine::hasStaticParameters":[40,2,1,"_CPPv2N6paddle15GradientMachine19hasStaticParametersEv"],"paddle::GradientMachine::kCustom":[40,1,1,"_CPPv2N6paddle15GradientMachine7kCustomE"],"paddle::GradientMachine::kNormal":[40,1,1,"_CPPv2N6paddle15GradientMachine7kNormalE"],"paddle::GradientMachine::kSgdSparseCpuTraining":[40,1,1,"_CPPv2N6paddle15GradientMachine21kSgdSparseCpuTrainingE"],"paddle::GradientMachine::kTesting":[40,1,1,"_CPPv2N6paddle15GradientMachine8kTestingE"],"paddle::GradientMachine::makeEvaluator":[40,2,1,"_CPPv2N6paddle15GradientMachine13makeEvaluatorEv"],"paddle::GradientMachine::nonStaticParameters_":[40,3,1,"_CPPv2N6paddle15GradientMachine20nonStaticParameters_E"],"paddle::GradientMachine::onLoadParameter":[40,2,1,"_CPPv2N6paddle15GradientMachine15onLoadParameterEv"],"paddle::GradientMachine::onPassEnd":[40,2,1,"_CPPv2N6paddle15GradientMachine9onPassEndEv"],"paddle::GradientMachine::parameters_":[40,3,1,"_CPPv2N6paddle15GradientMachine11parameters_E"],"paddle::GradientMachine::prefetch":[40,2,1,"_CPPv2N6paddle15GradientMachine8prefetchERKNSt6vectorI8ArgumentEE"],"paddle::GradientMachine::resetState":[40,2,1,"_CPPv2N6paddle15GradientMachine10resetStateEv"],"paddle::GradientMachine::restart":[40,2,1,"_CPPv2N6paddle15GradientMachine7restartEv"],"paddle::GradientMachine::setOutputGrad":[40,2,1,"_CPPv2N6paddle15GradientMachine13setOutputGradERKNSt6vectorI8ArgumentEE"],"paddle::GradientMachine::setState":[40,2,1,"_CPPv2N6paddle15GradientMachine8setStateERK12MachineState"],"paddle::GradientMachine::start":[40,2,1,"_CPPv2N6paddle15GradientMachine5startERK13TrainerConfig15DataProviderPtr"],"paddle::GradientMachine::trainIsOn":[40,2,1,"_CPPv2N6paddle15GradientMachine9trainIsOnEv"],"paddle::GradientMachine::~GradientMachine":[40,2,1,"_CPPv2N6paddle15GradientMachineD0Ev"],"paddle::GradientPrinter":[38,1,1,"_CPPv2N6paddle15GradientPrinterE"],"paddle::GradientPrinter::GradientPrinter":[38,2,1,"_CPPv2N6paddle15GradientPrinter15GradientPrinterEv"],"paddle::GradientPrinter::eval":[38,2,1,"_CPPv2N6paddle15GradientPrinter4evalERK13NeuralNetwork"],"paddle::GradientPrinter::evalImp":[38,2,1,"_CPPv2N6paddle15GradientPrinter7evalImpERNSt6vectorI8ArgumentEE"],"paddle::GradientPrinter::updateSamplesNum":[38,2,1,"_CPPv2N6paddle15GradientPrinter16updateSamplesNumERKNSt6vectorI8ArgumentEE"],"paddle::GruCompute":[43,1,1,"_CPPv2N6paddle10GruComputeE"],"paddle::GruCompute::activeGate_":[43,3,1,"_CPPv2N6paddle10GruCompute11activeGate_E"],"paddle::GruCompute::activeNode_":[43,3,1,"_CPPv2N6paddle10GruCompute11activeNode_E"],"paddle::GruCompute::backward":[43,2,1,"_CPPv2N6paddle10GruCompute8backwardE12hl_gru_value11hl_gru_gradii"],"paddle::GruCompute::forward":[43,2,1,"_CPPv2N6paddle10GruCompute7forwardE12hl_gru_valueii"],"paddle::GruCompute::init":[43,2,1,"_CPPv2N6paddle10GruCompute4initER11LayerConfig"],"paddle::GruStepLayer":[43,1,1,"_CPPv2N6paddle12GruStepLayerE"],"paddle::GruStepLayer::GruStepLayer":[43,2,1,"_CPPv2N6paddle12GruStepLayer12GruStepLayerERK11LayerConfig"],"paddle::GruStepLayer::backward":[43,2,1,"_CPPv2N6paddle12GruStepLayer8backwardERK14UpdateCallback"],"paddle::GruStepLayer::bias_":[43,3,1,"_CPPv2N6paddle12GruStepLayer5bias_E"],"paddle::GruStepLayer::forward":[43,2,1,"_CPPv2N6paddle12GruStepLayer7forwardE8PassType"],"paddle::GruStepLayer::gate_":[43,3,1,"_CPPv2N6paddle12GruStepLayer5gate_E"],"paddle::GruStepLayer::init":[43,2,1,"_CPPv2N6paddle12GruStepLayer4initERK8LayerMapRK12ParameterMap"],"paddle::GruStepLayer::resetOutput_":[43,3,1,"_CPPv2N6paddle12GruStepLayer12resetOutput_E"],"paddle::GruStepLayer::weight_":[43,3,1,"_CPPv2N6paddle12GruStepLayer7weight_E"],"paddle::GruStepLayer::~GruStepLayer":[43,2,1,"_CPPv2N6paddle12GruStepLayerD0Ev"],"paddle::HISTOGRAM_EXPONENT":[46,1,1,"_CPPv2N6paddle18HISTOGRAM_EXPONENTE"],"paddle::HierarchicalSigmoidLayer":[43,1,1,"_CPPv2N6paddle24HierarchicalSigmoidLayerE"],"paddle::HierarchicalSigmoidLayer::HierarchicalSigmoidLayer":[43,2,1,"_CPPv2N6paddle24HierarchicalSigmoidLayer24HierarchicalSigmoidLayerERK11LayerConfig"],"paddle::HierarchicalSigmoidLayer::backward":[43,2,1,"_CPPv2N6paddle24HierarchicalSigmoidLayer8backwardERK14UpdateCallback"],"paddle::HierarchicalSigmoidLayer::biases_":[43,3,1,"_CPPv2N6paddle24HierarchicalSigmoidLayer7biases_E"],"paddle::HierarchicalSigmoidLayer::codeLength_":[43,3,1,"_CPPv2N6paddle24HierarchicalSigmoidLayer11codeLength_E"],"paddle::HierarchicalSigmoidLayer::forward":[43,2,1,"_CPPv2N6paddle24HierarchicalSigmoidLayer7forwardE8PassType"],"paddle::HierarchicalSigmoidLayer::getLabelLayer":[43,2,1,"_CPPv2N6paddle24HierarchicalSigmoidLayer13getLabelLayerEv"],"paddle::HierarchicalSigmoidLayer::init":[43,2,1,"_CPPv2N6paddle24HierarchicalSigmoidLayer4initERK8LayerMapRK12ParameterMap"],"paddle::HierarchicalSigmoidLayer::numClasses_":[43,3,1,"_CPPv2N6paddle24HierarchicalSigmoidLayer11numClasses_E"],"paddle::HierarchicalSigmoidLayer::preOutput_":[43,3,1,"_CPPv2N6paddle24HierarchicalSigmoidLayer10preOutput_E"],"paddle::HierarchicalSigmoidLayer::weights_":[43,3,1,"_CPPv2N6paddle24HierarchicalSigmoidLayer8weights_E"],"paddle::HistogramType":[46,4,1,"_CPPv2N6paddle13HistogramTypeE"],"paddle::HuberTwoClass":[43,1,1,"_CPPv2N6paddle13HuberTwoClassE"],"paddle::HuberTwoClass::HuberTwoClass":[43,2,1,"_CPPv2N6paddle13HuberTwoClass13HuberTwoClassERK11LayerConfig"],"paddle::HuberTwoClass::backwardImp":[43,2,1,"_CPPv2N6paddle13HuberTwoClass11backwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::HuberTwoClass::backwardImpIn":[43,2,1,"_CPPv2N6paddle13HuberTwoClass13backwardImpInER6MatrixR8ArgumentR6Matrix"],"paddle::HuberTwoClass::forwardImp":[43,2,1,"_CPPv2N6paddle13HuberTwoClass10forwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::HuberTwoClass::forwardImpIn":[43,2,1,"_CPPv2N6paddle13HuberTwoClass12forwardImpInER6MatrixR8ArgumentR6Matrix"],"paddle::HuberTwoClass::init":[43,2,1,"_CPPv2N6paddle13HuberTwoClass4initERK8LayerMapRK12ParameterMap"],"paddle::IBaseMatrix":[46,4,1,"_CPPv2N6paddle11IBaseMatrixE"],"paddle::ICpuGpuVector":[46,4,1,"_CPPv2N6paddle13ICpuGpuVectorE"],"paddle::ICpuGpuVectorPtr":[46,4,1,"_CPPv2N6paddle16ICpuGpuVectorPtrE"],"paddle::IFieldScanner":[36,1,1,"_CPPv2N6paddle13IFieldScannerE"],"paddle::IFieldScanner::DISABLE_COPY":[36,2,1,"_CPPv2N6paddle13IFieldScanner12DISABLE_COPYE13IFieldScanner"],"paddle::IFieldScanner::IFieldScanner":[36,2,1,"_CPPv2N6paddle13IFieldScanner13IFieldScannerEP10SlotHeader"],"paddle::IFieldScanner::create":[36,2,1,"_CPPv2N6paddle13IFieldScanner6createEP10SlotHeader"],"paddle::IFieldScanner::fill":[36,2,1,"_CPPv2N6paddle13IFieldScanner4fillER8ArgumentP8PyObject"],"paddle::IFieldScanner::finishFill":[36,2,1,"_CPPv2N6paddle13IFieldScanner10finishFillER8Argument"],"paddle::IFieldScanner::finishPrepare":[36,2,1,"_CPPv2N6paddle13IFieldScanner13finishPrepareER8Argument"],"paddle::IFieldScanner::headerPtr_":[36,3,1,"_CPPv2N6paddle13IFieldScanner10headerPtr_E"],"paddle::IFieldScanner::prepare":[36,2,1,"_CPPv2N6paddle13IFieldScanner7prepareER8ArgumentP8PyObject"],"paddle::IFieldScanner::startFill":[36,2,1,"_CPPv2N6paddle13IFieldScanner9startFillER8Argument"],"paddle::IFieldScanner::startPrepare":[36,2,1,"_CPPv2N6paddle13IFieldScanner12startPrepareER8Argument"],"paddle::IFieldScanner::~IFieldScanner":[36,2,1,"_CPPv2N6paddle13IFieldScannerD0Ev"],"paddle::IGradientMachineMode":[40,1,1,"_CPPv2N6paddle20IGradientMachineModeE"],"paddle::IGradientMachineMode::create":[40,2,1,"_CPPv2N6paddle20IGradientMachineMode6createERK11ModelConfig"],"paddle::IGradientMachineMode::dataMustInCpu":[40,2,1,"_CPPv2N6paddle20IGradientMachineMode13dataMustInCpuE7int32_t6size_t"],"paddle::IGradientMachineMode::isDataMustInCpu":[40,2,1,"_CPPv2N6paddle20IGradientMachineMode15isDataMustInCpuE6size_t"],"paddle::IGradientMachineMode::mode":[40,2,1,"_CPPv2N6paddle20IGradientMachineMode4modeE7int32_t"],"paddle::IGradientMachineMode::needTrainWholeDataInOneBatch":[40,2,1,"_CPPv2N6paddle20IGradientMachineMode28needTrainWholeDataInOneBatchEv"],"paddle::IGradientMachineMode::regGradientMachineMode":[40,2,1,"_CPPv2N6paddle20IGradientMachineMode22regGradientMachineModeE7int32_tRRNSt10unique_ptrI20IGradientMachineModeEE"],"paddle::IGradientMachineMode::shouldBeMe":[40,2,1,"_CPPv2N6paddle20IGradientMachineMode10shouldBeMeERKNSt6stringE6size_tbb"],"paddle::IGradientMachineMode::trainWholeDataInOneBatch":[40,2,1,"_CPPv2N6paddle20IGradientMachineMode24trainWholeDataInOneBatchE7int32_t"],"paddle::IGradientMachineMode::tryCreateGradientMachine":[40,2,1,"_CPPv2N6paddle20IGradientMachineMode24tryCreateGradientMachineE7int32_tRK11ModelConfig"],"paddle::IGradientMachineMode::tryGetMode":[40,2,1,"_CPPv2N6paddle20IGradientMachineMode10tryGetModeEPiRKNSt6stringE7int32_tbb"],"paddle::IGradientMachineMode::~IGradientMachineMode":[40,2,1,"_CPPv2N6paddle20IGradientMachineModeD0Ev"],"paddle::IParameterUpdaterHook":[54,1,1,"_CPPv2N6paddle21IParameterUpdaterHookE"],"paddle::IParameterUpdaterHook::IParameterUpdaterHook":[54,2,1,"_CPPv2N6paddle21IParameterUpdaterHook21IParameterUpdaterHookEv"],"paddle::IParameterUpdaterHook::create":[54,2,1,"_CPPv2N6paddle21IParameterUpdaterHook6createERK15ParameterConfigi"],"paddle::IParameterUpdaterHook::init":[54,2,1,"_CPPv2N6paddle21IParameterUpdaterHook4initEP9Parameter"],"paddle::IParameterUpdaterHook::update":[54,2,1,"_CPPv2N6paddle21IParameterUpdaterHook6updateEP9Parameter"],"paddle::IParameterUpdaterHook::~IParameterUpdaterHook":[54,2,1,"_CPPv2N6paddle21IParameterUpdaterHookD0Ev"],"paddle::IPyDataProviderCache":[36,1,1,"_CPPv2N6paddle20IPyDataProviderCacheE"],"paddle::IPyDataProviderCache::create":[36,2,1,"_CPPv2N6paddle20IPyDataProviderCache6createE9CacheType"],"paddle::IPyDataProviderCache::drop":[36,2,1,"_CPPv2N6paddle20IPyDataProviderCache4dropEPNSt5dequeI11PyObjectPtrEE"],"paddle::IPyDataProviderCache::load":[36,2,1,"_CPPv2N6paddle20IPyDataProviderCache4loadEv"],"paddle::IPyDataProviderCache::reset":[36,2,1,"_CPPv2N6paddle20IPyDataProviderCache5resetEv"],"paddle::IPyDataProviderCache::~IPyDataProviderCache":[36,2,1,"_CPPv2N6paddle20IPyDataProviderCacheD0Ev"],"paddle::IVector":[46,4,1,"_CPPv2N6paddle7IVectorE"],"paddle::IVectorPtr":[46,4,1,"_CPPv2N6paddle10IVectorPtrE"],"paddle::IdentityOffsetProjection":[43,1,1,"_CPPv2N6paddle24IdentityOffsetProjectionE"],"paddle::IdentityOffsetProjection::IdentityOffsetProjection":[43,2,1,"_CPPv2N6paddle24IdentityOffsetProjection24IdentityOffsetProjectionERK16ProjectionConfigRK12ParameterPtrb"],"paddle::IdentityOffsetProjection::backward":[43,2,1,"_CPPv2N6paddle24IdentityOffsetProjection8backwardERK14UpdateCallback"],"paddle::IdentityOffsetProjection::forward":[43,2,1,"_CPPv2N6paddle24IdentityOffsetProjection7forwardEv"],"paddle::IdentityProjection":[43,1,1,"_CPPv2N6paddle18IdentityProjectionE"],"paddle::IdentityProjection::IdentityProjection":[43,2,1,"_CPPv2N6paddle18IdentityProjection18IdentityProjectionERK16ProjectionConfigRK12ParameterPtrb"],"paddle::IdentityProjection::backward":[43,2,1,"_CPPv2N6paddle18IdentityProjection8backwardERK14UpdateCallback"],"paddle::IdentityProjection::forward":[43,2,1,"_CPPv2N6paddle18IdentityProjection7forwardEv"],"paddle::IndexScanner":[36,1,1,"_CPPv2N6paddle12IndexScannerE"],"paddle::IndexScanner::IndexScanner":[36,2,1,"_CPPv2N6paddle12IndexScanner12IndexScannerEP10SlotHeader"],"paddle::IndexScanner::fill":[36,2,1,"_CPPv2N6paddle12IndexScanner4fillER8ArgumentP8PyObject"],"paddle::IndexScanner::finishPrepare":[36,2,1,"_CPPv2N6paddle12IndexScanner13finishPrepareER8Argument"],"paddle::IndexScanner::prepare":[36,2,1,"_CPPv2N6paddle12IndexScanner7prepareER8ArgumentP8PyObject"],"paddle::InterpolationLayer":[43,1,1,"_CPPv2N6paddle18InterpolationLayerE"],"paddle::InterpolationLayer::InterpolationLayer":[43,2,1,"_CPPv2N6paddle18InterpolationLayer18InterpolationLayerERK11LayerConfig"],"paddle::InterpolationLayer::backward":[43,2,1,"_CPPv2N6paddle18InterpolationLayer8backwardERK14UpdateCallback"],"paddle::InterpolationLayer::forward":[43,2,1,"_CPPv2N6paddle18InterpolationLayer7forwardE8PassType"],"paddle::InterpolationLayer::init":[43,2,1,"_CPPv2N6paddle18InterpolationLayer4initERK8LayerMapRK12ParameterMap"],"paddle::InterpolationLayer::tmpMatrix":[43,3,1,"_CPPv2N6paddle18InterpolationLayer9tmpMatrixE"],"paddle::InterpolationLayer::weightLast_":[43,3,1,"_CPPv2N6paddle18InterpolationLayer11weightLast_E"],"paddle::InterpolationLayer::~InterpolationLayer":[43,2,1,"_CPPv2N6paddle18InterpolationLayerD0Ev"],"paddle::L1L2LrRegularizer":[52,1,1,"_CPPv2N6paddle17L1L2LrRegularizerE"],"paddle::L1L2LrRegularizer::update":[52,2,1,"_CPPv2N6paddle17L1L2LrRegularizer6updateEA_K9VectorPtrRK15ParameterConfig4realii"],"paddle::L1L2Regularizer":[52,1,1,"_CPPv2N6paddle15L1L2RegularizerE"],"paddle::L1L2Regularizer::update":[52,2,1,"_CPPv2N6paddle15L1L2Regularizer6updateEA_K9VectorPtrRK15ParameterConfig4realii"],"paddle::L1LrRegularizer":[52,1,1,"_CPPv2N6paddle15L1LrRegularizerE"],"paddle::L1LrRegularizer::update":[52,2,1,"_CPPv2N6paddle15L1LrRegularizer6updateEA_K9VectorPtrRK15ParameterConfig4realii"],"paddle::L1Regularizer":[52,1,1,"_CPPv2N6paddle13L1RegularizerE"],"paddle::L1Regularizer::update":[52,2,1,"_CPPv2N6paddle13L1Regularizer6updateEA_K9VectorPtrRK15ParameterConfig4realii"],"paddle::L2LrRegularizer":[52,1,1,"_CPPv2N6paddle15L2LrRegularizerE"],"paddle::L2LrRegularizer::update":[52,2,1,"_CPPv2N6paddle15L2LrRegularizer6updateEA_K9VectorPtrRK15ParameterConfig4realii"],"paddle::L2Regularizer":[52,1,1,"_CPPv2N6paddle13L2RegularizerE"],"paddle::L2Regularizer::update":[52,2,1,"_CPPv2N6paddle13L2Regularizer6updateEA_K9VectorPtrRK15ParameterConfig4realii"],"paddle::LambdaCost":[43,1,1,"_CPPv2N6paddle10LambdaCostE"],"paddle::LambdaCost::LambdaCost":[43,2,1,"_CPPv2N6paddle10LambdaCost10LambdaCostERK11LayerConfig"],"paddle::LambdaCost::backward":[43,2,1,"_CPPv2N6paddle10LambdaCost8backwardERK14UpdateCallback"],"paddle::LambdaCost::calcGrad":[43,2,1,"_CPPv2N6paddle10LambdaCost8calcGradEPK4realPK4realP4reali"],"paddle::LambdaCost::calcNDCG":[43,2,1,"_CPPv2N6paddle10LambdaCost8calcNDCGEPK4realPK4reali"],"paddle::LambdaCost::forward":[43,2,1,"_CPPv2N6paddle10LambdaCost7forwardE8PassType"],"paddle::LambdaCost::getOutputLayer":[43,2,1,"_CPPv2N6paddle10LambdaCost14getOutputLayerEv"],"paddle::LambdaCost::getScoreLayer":[43,2,1,"_CPPv2N6paddle10LambdaCost13getScoreLayerEv"],"paddle::LambdaCost::init":[43,2,1,"_CPPv2N6paddle10LambdaCost4initERK8LayerMapRK12ParameterMap"],"paddle::LambdaCost::onPassEnd":[43,2,1,"_CPPv2N6paddle10LambdaCost9onPassEndEv"],"paddle::Layer":[43,1,1,"_CPPv2N6paddle5LayerE"],"paddle::Layer::Layer":[43,2,1,"_CPPv2N6paddle5Layer5LayerERK11LayerConfigb"],"paddle::Layer::accessSubNetwork":[43,2,1,"_CPPv2N6paddle5Layer16accessSubNetworkERKNSt8functionIFvR13NeuralNetworkEEE"],"paddle::Layer::activation_":[43,3,1,"_CPPv2N6paddle5Layer11activation_E"],"paddle::Layer::addOutputArgument":[43,2,1,"_CPPv2N6paddle5Layer17addOutputArgumentEi"],"paddle::Layer::addPrev":[43,2,1,"_CPPv2N6paddle5Layer7addPrevE8LayerPtr"],"paddle::Layer::backward":[43,2,1,"_CPPv2N6paddle5Layer8backwardERK14UpdateCallback"],"paddle::Layer::backwardActivation":[43,2,1,"_CPPv2N6paddle5Layer18backwardActivationEv"],"paddle::Layer::biasParameter_":[43,3,1,"_CPPv2N6paddle5Layer14biasParameter_E"],"paddle::Layer::config_":[43,3,1,"_CPPv2N6paddle5Layer7config_E"],"paddle::Layer::copyOutputToOtherDevice":[43,2,1,"_CPPv2N6paddle5Layer23copyOutputToOtherDeviceEv"],"paddle::Layer::create":[43,2,1,"_CPPv2N6paddle5Layer6createERK11LayerConfig"],"paddle::Layer::deviceId_":[43,3,1,"_CPPv2N6paddle5Layer9deviceId_E"],"paddle::Layer::dropOutMask_":[43,3,1,"_CPPv2N6paddle5Layer12dropOutMask_E"],"paddle::Layer::forward":[43,2,1,"_CPPv2N6paddle5Layer7forwardE8PassType"],"paddle::Layer::forwardActivation":[43,2,1,"_CPPv2N6paddle5Layer17forwardActivationEv"],"paddle::Layer::forwardDropOut":[43,2,1,"_CPPv2N6paddle5Layer14forwardDropOutEv"],"paddle::Layer::getBiasParameter":[43,2,1,"_CPPv2N6paddle5Layer16getBiasParameterEv"],"paddle::Layer::getDeviceId":[43,2,1,"_CPPv2N6paddle5Layer11getDeviceIdEv"],"paddle::Layer::getInput":[43,2,1,"_CPPv2N6paddle5Layer8getInputERK5Layer"],"paddle::Layer::getInputGrad":[43,2,1,"_CPPv2N6paddle5Layer12getInputGradEi"],"paddle::Layer::getInputLabel":[43,2,1,"_CPPv2N6paddle5Layer13getInputLabelERK5Layer"],"paddle::Layer::getInputValue":[43,2,1,"_CPPv2N6paddle5Layer13getInputValueEi"],"paddle::Layer::getName":[43,2,1,"_CPPv2N6paddle5Layer7getNameEv"],"paddle::Layer::getOutput":[43,2,1,"_CPPv2N6paddle5Layer9getOutputEi"],"paddle::Layer::getOutputGrad":[43,2,1,"_CPPv2N6paddle5Layer13getOutputGradEv"],"paddle::Layer::getOutputLabel":[43,2,1,"_CPPv2N6paddle5Layer14getOutputLabelEv"],"paddle::Layer::getOutputValue":[43,2,1,"_CPPv2N6paddle5Layer14getOutputValueEv"],"paddle::Layer::getParameters":[43,2,1,"_CPPv2N6paddle5Layer13getParametersEv"],"paddle::Layer::getPrev":[43,2,1,"_CPPv2N6paddle5Layer7getPrevE6size_t"],"paddle::Layer::getSize":[43,2,1,"_CPPv2N6paddle5Layer7getSizeEv"],"paddle::Layer::getState":[43,2,1,"_CPPv2N6paddle5Layer8getStateEv"],"paddle::Layer::getType":[43,2,1,"_CPPv2N6paddle5Layer7getTypeEv"],"paddle::Layer::init":[43,2,1,"_CPPv2N6paddle5Layer4initERK8LayerMapRK12ParameterMap"],"paddle::Layer::initNeedFlags":[43,2,1,"_CPPv2N6paddle5Layer13initNeedFlagsEv"],"paddle::Layer::initSubNetwork":[43,2,1,"_CPPv2N6paddle5Layer14initSubNetworkEP13NeuralNetworkRK11ModelConfigRKNSt6vectorI13ParameterTypeEEb"],"paddle::Layer::inputArgument_":[43,3,1,"_CPPv2N6paddle5Layer14inputArgument_E"],"paddle::Layer::inputLayers_":[43,3,1,"_CPPv2N6paddle5Layer12inputLayers_E"],"paddle::Layer::markAllInputGrad":[43,2,1,"_CPPv2N6paddle5Layer16markAllInputGradEv"],"paddle::Layer::markInBackward_":[43,3,1,"_CPPv2N6paddle5Layer15markInBackward_E"],"paddle::Layer::markInputGrad":[43,2,1,"_CPPv2N6paddle5Layer13markInputGradEi"],"paddle::Layer::needGradient":[43,2,1,"_CPPv2N6paddle5Layer12needGradientEv"],"paddle::Layer::needGradient_":[43,3,1,"_CPPv2N6paddle5Layer13needGradient_E"],"paddle::Layer::needSequenceInfo_":[43,3,1,"_CPPv2N6paddle5Layer17needSequenceInfo_E"],"paddle::Layer::onPassEnd":[43,2,1,"_CPPv2N6paddle5Layer9onPassEndEv"],"paddle::Layer::outputMap_":[43,3,1,"_CPPv2N6paddle5Layer10outputMap_E"],"paddle::Layer::outputOtherDevice_":[43,3,1,"_CPPv2N6paddle5Layer18outputOtherDevice_E"],"paddle::Layer::output_":[43,3,1,"_CPPv2N6paddle5Layer7output_E"],"paddle::Layer::parameters_":[43,3,1,"_CPPv2N6paddle5Layer11parameters_E"],"paddle::Layer::passType_":[43,3,1,"_CPPv2N6paddle5Layer9passType_E"],"paddle::Layer::prefetch":[43,2,1,"_CPPv2N6paddle5Layer8prefetchEv"],"paddle::Layer::registrar_":[43,3,1,"_CPPv2N6paddle5Layer10registrar_E"],"paddle::Layer::reserveOutput":[43,2,1,"_CPPv2N6paddle5Layer13reserveOutputE6size_t6size_t"],"paddle::Layer::resetOutput":[43,2,1,"_CPPv2N6paddle5Layer11resetOutputE6size_t6size_t"],"paddle::Layer::resetSpecifyOutput":[43,2,1,"_CPPv2N6paddle5Layer18resetSpecifyOutputER8Argument6size_t6size_tbb"],"paddle::Layer::resetState":[43,2,1,"_CPPv2N6paddle5Layer10resetStateEv"],"paddle::Layer::resizeOutput":[43,2,1,"_CPPv2N6paddle5Layer12resizeOutputE6size_t6size_t"],"paddle::Layer::setNeedGradient":[43,2,1,"_CPPv2N6paddle5Layer15setNeedGradientEb"],"paddle::Layer::setNeedSequenceInfo":[43,2,1,"_CPPv2N6paddle5Layer19setNeedSequenceInfoEb"],"paddle::Layer::setOutput":[43,2,1,"_CPPv2N6paddle5Layer9setOutputERKNSt6stringEP8Argument"],"paddle::Layer::setState":[43,2,1,"_CPPv2N6paddle5Layer8setStateE13LayerStatePtr"],"paddle::Layer::showOutputStats":[43,2,1,"_CPPv2N6paddle5Layer15showOutputStatsEv"],"paddle::Layer::tmpGrad_":[43,3,1,"_CPPv2N6paddle5Layer8tmpGrad_E"],"paddle::Layer::useGpu_":[43,3,1,"_CPPv2N6paddle5Layer7useGpu_E"],"paddle::Layer::waitAndMergeOutputGrad":[43,2,1,"_CPPv2N6paddle5Layer22waitAndMergeOutputGradEv"],"paddle::Layer::waitInputValue":[43,2,1,"_CPPv2N6paddle5Layer14waitInputValueEv"],"paddle::Layer::zeroGrad":[43,2,1,"_CPPv2N6paddle5Layer8zeroGradEv"],"paddle::Layer::~Layer":[43,2,1,"_CPPv2N6paddle5LayerD0Ev"],"paddle::LinearChainCRF":[43,1,1,"_CPPv2N6paddle14LinearChainCRFE"],"paddle::LinearChainCRF::LinearChainCRF":[43,2,1,"_CPPv2N6paddle14LinearChainCRF14LinearChainCRFEiP4realP4real"],"paddle::LinearChainCRF::a_":[43,3,1,"_CPPv2N6paddle14LinearChainCRF2a_E"],"paddle::LinearChainCRF::alpha_":[43,3,1,"_CPPv2N6paddle14LinearChainCRF6alpha_E"],"paddle::LinearChainCRF::b_":[43,3,1,"_CPPv2N6paddle14LinearChainCRF2b_E"],"paddle::LinearChainCRF::backward":[43,2,1,"_CPPv2N6paddle14LinearChainCRF8backwardEP4realP4realPii"],"paddle::LinearChainCRF::beta_":[43,3,1,"_CPPv2N6paddle14LinearChainCRF5beta_E"],"paddle::LinearChainCRF::da_":[43,3,1,"_CPPv2N6paddle14LinearChainCRF3da_E"],"paddle::LinearChainCRF::db_":[43,3,1,"_CPPv2N6paddle14LinearChainCRF3db_E"],"paddle::LinearChainCRF::decode":[43,2,1,"_CPPv2N6paddle14LinearChainCRF6decodeEP4realPii"],"paddle::LinearChainCRF::dw_":[43,3,1,"_CPPv2N6paddle14LinearChainCRF3dw_E"],"paddle::LinearChainCRF::expW_":[43,3,1,"_CPPv2N6paddle14LinearChainCRF5expW_E"],"paddle::LinearChainCRF::expX_":[43,3,1,"_CPPv2N6paddle14LinearChainCRF5expX_E"],"paddle::LinearChainCRF::forward":[43,2,1,"_CPPv2N6paddle14LinearChainCRF7forwardEP4realPii"],"paddle::LinearChainCRF::maxX_":[43,3,1,"_CPPv2N6paddle14LinearChainCRF5maxX_E"],"paddle::LinearChainCRF::numClasses_":[43,3,1,"_CPPv2N6paddle14LinearChainCRF11numClasses_E"],"paddle::LinearChainCRF::ones_":[43,3,1,"_CPPv2N6paddle14LinearChainCRF5ones_E"],"paddle::LinearChainCRF::track_":[43,3,1,"_CPPv2N6paddle14LinearChainCRF6track_E"],"paddle::LinearChainCRF::w_":[43,3,1,"_CPPv2N6paddle14LinearChainCRF2w_E"],"paddle::LinearChainCTC":[43,1,1,"_CPPv2N6paddle14LinearChainCTCE"],"paddle::LinearChainCTC::LinearChainCTC":[43,2,1,"_CPPv2N6paddle14LinearChainCTC14LinearChainCTCEib"],"paddle::LinearChainCTC::backward":[43,2,1,"_CPPv2N6paddle14LinearChainCTC8backwardEP4realP4realPii"],"paddle::LinearChainCTC::backwardVars_":[43,3,1,"_CPPv2N6paddle14LinearChainCTC13backwardVars_E"],"paddle::LinearChainCTC::blank_":[43,3,1,"_CPPv2N6paddle14LinearChainCTC6blank_E"],"paddle::LinearChainCTC::forward":[43,2,1,"_CPPv2N6paddle14LinearChainCTC7forwardEP4realiPii"],"paddle::LinearChainCTC::forwardVars_":[43,3,1,"_CPPv2N6paddle14LinearChainCTC12forwardVars_E"],"paddle::LinearChainCTC::gradTerms_":[43,3,1,"_CPPv2N6paddle14LinearChainCTC10gradTerms_E"],"paddle::LinearChainCTC::isInvalid_":[43,3,1,"_CPPv2N6paddle14LinearChainCTC10isInvalid_E"],"paddle::LinearChainCTC::logActs_":[43,3,1,"_CPPv2N6paddle14LinearChainCTC8logActs_E"],"paddle::LinearChainCTC::logProb_":[43,3,1,"_CPPv2N6paddle14LinearChainCTC8logProb_E"],"paddle::LinearChainCTC::normByTimes_":[43,3,1,"_CPPv2N6paddle14LinearChainCTC12normByTimes_E"],"paddle::LinearChainCTC::numClasses_":[43,3,1,"_CPPv2N6paddle14LinearChainCTC11numClasses_E"],"paddle::LinearChainCTC::segmentRange":[43,2,1,"_CPPv2N6paddle14LinearChainCTC12segmentRangeERiRii"],"paddle::LinearChainCTC::totalSegments_":[43,3,1,"_CPPv2N6paddle14LinearChainCTC14totalSegments_E"],"paddle::LinearChainCTC::totalTime_":[43,3,1,"_CPPv2N6paddle14LinearChainCTC10totalTime_E"],"paddle::LockedCondition":[66,1,1,"_CPPv2N6paddle15LockedConditionE"],"paddle::LockedCondition::mutex":[66,2,1,"_CPPv2N6paddle15LockedCondition5mutexEv"],"paddle::LockedCondition::mutex_":[66,3,1,"_CPPv2N6paddle15LockedCondition6mutex_E"],"paddle::LockedCondition::notify_all":[66,2,1,"_CPPv2N6paddle15LockedCondition10notify_allE2Op"],"paddle::LockedCondition::notify_one":[66,2,1,"_CPPv2N6paddle15LockedCondition10notify_oneE2Op"],"paddle::LockedCondition::wait":[66,2,1,"_CPPv2N6paddle15LockedCondition4waitE9Predicate"],"paddle::LstmCompute":[43,1,1,"_CPPv2N6paddle11LstmComputeE"],"paddle::LstmCompute::activeGate_":[43,3,1,"_CPPv2N6paddle11LstmCompute11activeGate_E"],"paddle::LstmCompute::activeNode_":[43,3,1,"_CPPv2N6paddle11LstmCompute11activeNode_E"],"paddle::LstmCompute::activeState_":[43,3,1,"_CPPv2N6paddle11LstmCompute12activeState_E"],"paddle::LstmCompute::backwardBatch":[43,2,1,"_CPPv2N6paddle11LstmCompute13backwardBatchE13hl_lstm_value12hl_lstm_gradii"],"paddle::LstmCompute::backwardOneSequence":[43,2,1,"_CPPv2N6paddle11LstmCompute19backwardOneSequenceE13hl_lstm_value12hl_lstm_gradi"],"paddle::LstmCompute::forwardBatch":[43,2,1,"_CPPv2N6paddle11LstmCompute12forwardBatchE13hl_lstm_valueii"],"paddle::LstmCompute::forwardOneSequence":[43,2,1,"_CPPv2N6paddle11LstmCompute18forwardOneSequenceE13hl_lstm_valuei"],"paddle::LstmCompute::init":[43,2,1,"_CPPv2N6paddle11LstmCompute4initER11LayerConfig"],"paddle::LstmLayer":[43,1,1,"_CPPv2N6paddle9LstmLayerE"],"paddle::LstmLayer::LstmLayer":[43,2,1,"_CPPv2N6paddle9LstmLayer9LstmLayerERK11LayerConfig"],"paddle::LstmLayer::backward":[43,2,1,"_CPPv2N6paddle9LstmLayer8backwardERK14UpdateCallback"],"paddle::LstmLayer::backwardBatch":[43,2,1,"_CPPv2N6paddle9LstmLayer13backwardBatchEi6size_tPKi9MatrixPtr"],"paddle::LstmLayer::backwardSeqParallel":[43,2,1,"_CPPv2N6paddle9LstmLayer19backwardSeqParallelEi6size_tPKi9MatrixPtr"],"paddle::LstmLayer::backwardSequence":[43,2,1,"_CPPv2N6paddle9LstmLayer16backwardSequenceEi6size_tPKi9MatrixPtr"],"paddle::LstmLayer::batchGrad_":[43,3,1,"_CPPv2N6paddle9LstmLayer10batchGrad_E"],"paddle::LstmLayer::batchValue_":[43,3,1,"_CPPv2N6paddle9LstmLayer11batchValue_E"],"paddle::LstmLayer::bias_":[43,3,1,"_CPPv2N6paddle9LstmLayer5bias_E"],"paddle::LstmLayer::checkFgGrad_":[43,3,1,"_CPPv2N6paddle9LstmLayer12checkFgGrad_E"],"paddle::LstmLayer::checkFg_":[43,3,1,"_CPPv2N6paddle9LstmLayer8checkFg_E"],"paddle::LstmLayer::checkIgGrad_":[43,3,1,"_CPPv2N6paddle9LstmLayer12checkIgGrad_E"],"paddle::LstmLayer::checkIg_":[43,3,1,"_CPPv2N6paddle9LstmLayer8checkIg_E"],"paddle::LstmLayer::checkOgGrad_":[43,3,1,"_CPPv2N6paddle9LstmLayer12checkOgGrad_E"],"paddle::LstmLayer::checkOg_":[43,3,1,"_CPPv2N6paddle9LstmLayer8checkOg_E"],"paddle::LstmLayer::forward":[43,2,1,"_CPPv2N6paddle9LstmLayer7forwardE8PassType"],"paddle::LstmLayer::forwardBatch":[43,2,1,"_CPPv2N6paddle9LstmLayer12forwardBatchEi6size_tPKi9MatrixPtr"],"paddle::LstmLayer::forwardSeqParallel":[43,2,1,"_CPPv2N6paddle9LstmLayer18forwardSeqParallelEi6size_tPKi9MatrixPtr"],"paddle::LstmLayer::forwardSequence":[43,2,1,"_CPPv2N6paddle9LstmLayer15forwardSequenceEi6size_tPKi9MatrixPtr"],"paddle::LstmLayer::gate_":[43,3,1,"_CPPv2N6paddle9LstmLayer5gate_E"],"paddle::LstmLayer::getPrevBatchOutput":[43,2,1,"_CPPv2N6paddle9LstmLayer18getPrevBatchOutputE6size_t"],"paddle::LstmLayer::getPrevBatchState":[43,2,1,"_CPPv2N6paddle9LstmLayer17getPrevBatchStateE6size_t"],"paddle::LstmLayer::getState":[43,2,1,"_CPPv2N6paddle9LstmLayer8getStateEv"],"paddle::LstmLayer::init":[43,2,1,"_CPPv2N6paddle9LstmLayer4initERK8LayerMapRK12ParameterMap"],"paddle::LstmLayer::localBiasGrad_":[43,3,1,"_CPPv2N6paddle9LstmLayer14localBiasGrad_E"],"paddle::LstmLayer::localBias_":[43,3,1,"_CPPv2N6paddle9LstmLayer10localBias_E"],"paddle::LstmLayer::preOutput_":[43,3,1,"_CPPv2N6paddle9LstmLayer10preOutput_E"],"paddle::LstmLayer::prevBatchOutput2_":[43,3,1,"_CPPv2N6paddle9LstmLayer17prevBatchOutput2_E"],"paddle::LstmLayer::prevOutput_":[43,3,1,"_CPPv2N6paddle9LstmLayer11prevOutput_E"],"paddle::LstmLayer::prevState_":[43,3,1,"_CPPv2N6paddle9LstmLayer10prevState_E"],"paddle::LstmLayer::resetState":[43,2,1,"_CPPv2N6paddle9LstmLayer10resetStateEv"],"paddle::LstmLayer::reversed_":[43,3,1,"_CPPv2N6paddle9LstmLayer9reversed_E"],"paddle::LstmLayer::setState":[43,2,1,"_CPPv2N6paddle9LstmLayer8setStateE13LayerStatePtr"],"paddle::LstmLayer::state_":[43,3,1,"_CPPv2N6paddle9LstmLayer6state_E"],"paddle::LstmLayer::totalState_":[43,3,1,"_CPPv2N6paddle9LstmLayer11totalState_E"],"paddle::LstmLayer::useBatch_":[43,3,1,"_CPPv2N6paddle9LstmLayer9useBatch_E"],"paddle::LstmLayer::useSeqParallel_":[43,3,1,"_CPPv2N6paddle9LstmLayer15useSeqParallel_E"],"paddle::LstmLayer::weight_":[43,3,1,"_CPPv2N6paddle9LstmLayer7weight_E"],"paddle::LstmStepLayer":[43,1,1,"_CPPv2N6paddle13LstmStepLayerE"],"paddle::LstmStepLayer::LstmStepLayer":[43,2,1,"_CPPv2N6paddle13LstmStepLayer13LstmStepLayerERK11LayerConfig"],"paddle::LstmStepLayer::backward":[43,2,1,"_CPPv2N6paddle13LstmStepLayer8backwardERK14UpdateCallback"],"paddle::LstmStepLayer::checkFgGrad_":[43,3,1,"_CPPv2N6paddle13LstmStepLayer12checkFgGrad_E"],"paddle::LstmStepLayer::checkFg_":[43,3,1,"_CPPv2N6paddle13LstmStepLayer8checkFg_E"],"paddle::LstmStepLayer::checkIgGrad_":[43,3,1,"_CPPv2N6paddle13LstmStepLayer12checkIgGrad_E"],"paddle::LstmStepLayer::checkIg_":[43,3,1,"_CPPv2N6paddle13LstmStepLayer8checkIg_E"],"paddle::LstmStepLayer::checkOgGrad_":[43,3,1,"_CPPv2N6paddle13LstmStepLayer12checkOgGrad_E"],"paddle::LstmStepLayer::checkOg_":[43,3,1,"_CPPv2N6paddle13LstmStepLayer8checkOg_E"],"paddle::LstmStepLayer::forward":[43,2,1,"_CPPv2N6paddle13LstmStepLayer7forwardE8PassType"],"paddle::LstmStepLayer::gate_":[43,3,1,"_CPPv2N6paddle13LstmStepLayer5gate_E"],"paddle::LstmStepLayer::init":[43,2,1,"_CPPv2N6paddle13LstmStepLayer4initERK8LayerMapRK12ParameterMap"],"paddle::LstmStepLayer::stateActive_":[43,3,1,"_CPPv2N6paddle13LstmStepLayer12stateActive_E"],"paddle::LstmStepLayer::state_":[43,3,1,"_CPPv2N6paddle13LstmStepLayer6state_E"],"paddle::LstmStepLayer::weight_":[43,3,1,"_CPPv2N6paddle13LstmStepLayer7weight_E"],"paddle::LstmStepLayer::~LstmStepLayer":[43,2,1,"_CPPv2N6paddle13LstmStepLayerD0Ev"],"paddle::MAT_CACHE_ROW":[52,1,1,"_CPPv2N6paddle13MAT_CACHE_ROWE"],"paddle::MAT_NORMAL":[52,1,1,"_CPPv2N6paddle10MAT_NORMALE"],"paddle::MAT_NORMAL_SHARED":[52,1,1,"_CPPv2N6paddle17MAT_NORMAL_SHAREDE"],"paddle::MAT_SPARSE_ROW":[52,1,1,"_CPPv2N6paddle14MAT_SPARSE_ROWE"],"paddle::MAT_SPARSE_ROW_AUTO_GROW":[52,1,1,"_CPPv2N6paddle24MAT_SPARSE_ROW_AUTO_GROWE"],"paddle::MAT_SPARSE_ROW_IDS":[52,1,1,"_CPPv2N6paddle18MAT_SPARSE_ROW_IDSE"],"paddle::MAT_SPARSE_ROW_PREFETCH":[52,1,1,"_CPPv2N6paddle23MAT_SPARSE_ROW_PREFETCHE"],"paddle::MAT_SPARSE_ROW_PREFETCH_FULL_SIZE":[52,1,1,"_CPPv2N6paddle33MAT_SPARSE_ROW_PREFETCH_FULL_SIZEE"],"paddle::MAT_VALUE_SHARED":[52,1,1,"_CPPv2N6paddle16MAT_VALUE_SHAREDE"],"paddle::MDLstmLayer":[43,1,1,"_CPPv2N6paddle11MDLstmLayerE"],"paddle::MDLstmLayer::MDLstmLayer":[43,2,1,"_CPPv2N6paddle11MDLstmLayer11MDLstmLayerERK11LayerConfig"],"paddle::MDLstmLayer::activationGate_":[43,3,1,"_CPPv2N6paddle11MDLstmLayer15activationGate_E"],"paddle::MDLstmLayer::activationState_":[43,3,1,"_CPPv2N6paddle11MDLstmLayer16activationState_E"],"paddle::MDLstmLayer::backward":[43,2,1,"_CPPv2N6paddle11MDLstmLayer8backwardERK14UpdateCallback"],"paddle::MDLstmLayer::backwardGate2OutputSequence":[43,2,1,"_CPPv2N6paddle11MDLstmLayer27backwardGate2OutputSequenceEiR13CoordIterator"],"paddle::MDLstmLayer::backwardOneSequence":[43,2,1,"_CPPv2N6paddle11MDLstmLayer19backwardOneSequenceEiR13CoordIterator"],"paddle::MDLstmLayer::delays_":[43,3,1,"_CPPv2N6paddle11MDLstmLayer7delays_E"],"paddle::MDLstmLayer::dimsV_":[43,3,1,"_CPPv2N6paddle11MDLstmLayer6dimsV_E"],"paddle::MDLstmLayer::directions_":[43,3,1,"_CPPv2N6paddle11MDLstmLayer11directions_E"],"paddle::MDLstmLayer::forward":[43,2,1,"_CPPv2N6paddle11MDLstmLayer7forwardE8PassType"],"paddle::MDLstmLayer::forwardGate2OutputSequence":[43,2,1,"_CPPv2N6paddle11MDLstmLayer26forwardGate2OutputSequenceEiR13CoordIterator"],"paddle::MDLstmLayer::forwardOneSequence":[43,2,1,"_CPPv2N6paddle11MDLstmLayer18forwardOneSequenceEiR13CoordIterator"],"paddle::MDLstmLayer::frameForgetGate_":[43,3,1,"_CPPv2N6paddle11MDLstmLayer16frameForgetGate_E"],"paddle::MDLstmLayer::frameGate_":[43,3,1,"_CPPv2N6paddle11MDLstmLayer10frameGate_E"],"paddle::MDLstmLayer::frameInputGate_":[43,3,1,"_CPPv2N6paddle11MDLstmLayer15frameInputGate_E"],"paddle::MDLstmLayer::frameInputNode_":[43,3,1,"_CPPv2N6paddle11MDLstmLayer15frameInputNode_E"],"paddle::MDLstmLayer::frameOutputGate_":[43,3,1,"_CPPv2N6paddle11MDLstmLayer16frameOutputGate_E"],"paddle::MDLstmLayer::frameOutput_":[43,3,1,"_CPPv2N6paddle11MDLstmLayer12frameOutput_E"],"paddle::MDLstmLayer::framePreOutput_":[43,3,1,"_CPPv2N6paddle11MDLstmLayer15framePreOutput_E"],"paddle::MDLstmLayer::frameState_":[43,3,1,"_CPPv2N6paddle11MDLstmLayer11frameState_E"],"paddle::MDLstmLayer::init":[43,2,1,"_CPPv2N6paddle11MDLstmLayer4initERK8LayerMapRK12ParameterMap"],"paddle::MDLstmLayer::numBlocks_":[43,3,1,"_CPPv2N6paddle11MDLstmLayer10numBlocks_E"],"paddle::MDLstmLayer::numDims_":[43,3,1,"_CPPv2N6paddle11MDLstmLayer8numDims_E"],"paddle::MatType":[52,4,1,"_CPPv2N6paddle7MatTypeE"],"paddle::Matrix":[46,1,1,"_CPPv2N6paddle6MatrixE"],"paddle::Matrix::accumulateColSum":[46,2,1,"_CPPv2N6paddle6Matrix16accumulateColSumER6Matrix"],"paddle::Matrix::add3":[46,2,1,"_CPPv2N6paddle6Matrix4add3E9MatrixPtr"],"paddle::Matrix::addBias":[46,2,1,"_CPPv2N6paddle6Matrix7addBiasER6Matrix4real"],"paddle::Matrix::addByBitCode":[46,2,1,"_CPPv2N6paddle6Matrix12addByBitCodeE6size_tRK7IVectorRK6Matrix"],"paddle::Matrix::addByBitCodeBackward":[46,2,1,"_CPPv2N6paddle6Matrix20addByBitCodeBackwardE6size_tRK7IVectorR6Matrix"],"paddle::Matrix::addColumnVector":[46,2,1,"_CPPv2N6paddle6Matrix15addColumnVectorERK6Matrix"],"paddle::Matrix::addElements":[46,2,1,"_CPPv2N6paddle6Matrix11addElementsER6MatrixR7IVector"],"paddle::Matrix::addToRows":[46,2,1,"_CPPv2N6paddle6Matrix9addToRowsER6MatrixR7IVector"],"paddle::Matrix::avgPoolBackward":[46,2,1,"_CPPv2N6paddle6Matrix15avgPoolBackwardER6Matrix6size_t6size_t6size_ti6size_t6size_t6size_t4real4real"],"paddle::Matrix::avgPoolForward":[46,2,1,"_CPPv2N6paddle6Matrix14avgPoolForwardER6Matrix6size_t6size_t6size_t6size_ti6size_t6size_t6size_t"],"paddle::Matrix::check":[46,2,1,"_CPPv2N6paddle6Matrix5checkERNSt7ostreamER6Matrixb"],"paddle::Matrix::circularConv":[46,2,1,"_CPPv2N6paddle6Matrix12circularConvER6MatrixR6Matrix"],"paddle::Matrix::circularConvDerivative":[46,2,1,"_CPPv2N6paddle6Matrix22circularConvDerivativeER6MatrixR6MatrixR6MatrixR6MatrixR6Matrix"],"paddle::Matrix::classificationError":[46,2,1,"_CPPv2N6paddle6Matrix19classificationErrorE9MatrixPtr10IVectorPtr"],"paddle::Matrix::classificationErrorMulti":[46,2,1,"_CPPv2N6paddle6Matrix24classificationErrorMultiER6MatrixR6Matrix4real"],"paddle::Matrix::clear":[46,2,1,"_CPPv2N6paddle6Matrix5clearEv"],"paddle::Matrix::clone":[46,2,1,"_CPPv2N6paddle6Matrix5cloneE6size_t6size_tb"],"paddle::Matrix::colMax":[46,2,1,"_CPPv2N6paddle6Matrix6colMaxER6Matrix"],"paddle::Matrix::colMerge":[46,2,1,"_CPPv2N6paddle6Matrix8colMergeER6Matrix"],"paddle::Matrix::collectBias":[46,2,1,"_CPPv2N6paddle6Matrix11collectBiasER6Matrix4real"],"paddle::Matrix::contextProjectionBackward":[46,2,1,"_CPPv2N6paddle6Matrix25contextProjectionBackwardE9MatrixPtr9MatrixPtrRK7IVectorii6size_tb"],"paddle::Matrix::contextProjectionBackwardData":[46,2,1,"_CPPv2N6paddle6Matrix29contextProjectionBackwardDataE9MatrixPtrRK7IVectorii"],"paddle::Matrix::contextProjectionBackwardWeight":[46,2,1,"_CPPv2N6paddle6Matrix31contextProjectionBackwardWeightE9MatrixPtrRK7IVectoriii6size_t"],"paddle::Matrix::contextProjectionForward":[46,2,1,"_CPPv2N6paddle6Matrix24contextProjectionForwardE9MatrixPtr9MatrixPtrRK7IVectorii6size_tb"],"paddle::Matrix::convExpand":[46,2,1,"_CPPv2N6paddle6Matrix10convExpandER6Matrixiiiiiiiiiii"],"paddle::Matrix::convShrink":[46,2,1,"_CPPv2N6paddle6Matrix10convShrinkER6Matrixiiiiiiiiiii4real4real"],"paddle::Matrix::copyByRowIndex":[46,2,1,"_CPPv2N6paddle6Matrix14copyByRowIndexER6MatrixR7IVector"],"paddle::Matrix::copyFrom":[46,2,1,"_CPPv2N6paddle6Matrix8copyFromERK7IVector"],"paddle::Matrix::cosSim":[46,2,1,"_CPPv2N6paddle6Matrix6cosSimER6MatrixR6Matrix4real"],"paddle::Matrix::cosSimDerivative":[46,2,1,"_CPPv2N6paddle6Matrix16cosSimDerivativeER6MatrixR6MatrixR6MatrixR6MatrixR6Matrix4real"],"paddle::Matrix::crossMapNormalBwd":[46,2,1,"_CPPv2N6paddle6Matrix17crossMapNormalBwdER6MatrixR6MatrixR6MatrixR6Matrix6size_t6size_t6size_t6size_tff"],"paddle::Matrix::crossMapNormalFwd":[46,2,1,"_CPPv2N6paddle6Matrix17crossMapNormalFwdER6Matrix6size_t6size_tR6Matrix6size_t6size_tff"],"paddle::Matrix::elementCnt_":[46,3,1,"_CPPv2N6paddle6Matrix11elementCnt_E"],"paddle::Matrix::getAbsSum":[46,2,1,"_CPPv2N6paddle6Matrix9getAbsSumEv"],"paddle::Matrix::getCols":[46,2,1,"_CPPv2N6paddle6Matrix7getColsEv"],"paddle::Matrix::getData":[46,2,1,"_CPPv2N6paddle6Matrix7getDataEv"],"paddle::Matrix::getElement":[46,2,1,"_CPPv2N6paddle6Matrix10getElementE6size_t6size_t"],"paddle::Matrix::getElementCnt":[46,2,1,"_CPPv2N6paddle6Matrix13getElementCntEv"],"paddle::Matrix::getFormat":[46,2,1,"_CPPv2N6paddle6Matrix9getFormatEv"],"paddle::Matrix::getHeight":[46,2,1,"_CPPv2N6paddle6Matrix9getHeightEv"],"paddle::Matrix::getMax":[46,2,1,"_CPPv2N6paddle6Matrix6getMaxEv"],"paddle::Matrix::getMemoryHandle":[46,2,1,"_CPPv2N6paddle6Matrix15getMemoryHandleEv"],"paddle::Matrix::getMin":[46,2,1,"_CPPv2N6paddle6Matrix6getMinEv"],"paddle::Matrix::getRowBuf":[46,2,1,"_CPPv2N6paddle6Matrix9getRowBufE6size_t"],"paddle::Matrix::getRows":[46,2,1,"_CPPv2N6paddle6Matrix7getRowsEv"],"paddle::Matrix::getStride":[46,2,1,"_CPPv2N6paddle6Matrix9getStrideEv"],"paddle::Matrix::getSum":[46,2,1,"_CPPv2N6paddle6Matrix6getSumEv"],"paddle::Matrix::getTranspose":[46,2,1,"_CPPv2N6paddle6Matrix12getTransposeEv"],"paddle::Matrix::getValueType":[46,2,1,"_CPPv2N6paddle6Matrix12getValueTypeEv"],"paddle::Matrix::getWidth":[46,2,1,"_CPPv2N6paddle6Matrix8getWidthEv"],"paddle::Matrix::isContiguous":[46,2,1,"_CPPv2N6paddle6Matrix12isContiguousEv"],"paddle::Matrix::isTransposed":[46,2,1,"_CPPv2N6paddle6Matrix12isTransposedEv"],"paddle::Matrix::leftMul":[46,2,1,"_CPPv2N6paddle6Matrix7leftMulER6Matrix4real4real"],"paddle::Matrix::maxPoolBackward":[46,2,1,"_CPPv2N6paddle6Matrix15maxPoolBackwardER6Matrix6size_t6size_tR6MatrixR6Matrix6size_ti6size_t6size_t6size_t4real4real"],"paddle::Matrix::maxPoolForward":[46,2,1,"_CPPv2N6paddle6Matrix14maxPoolForwardER6Matrix6size_t6size_t6size_t6size_ti6size_t6size_t6size_t"],"paddle::Matrix::maxSequenceBackward":[46,2,1,"_CPPv2N6paddle6Matrix19maxSequenceBackwardER6MatrixRK7IVectorR7IVector"],"paddle::Matrix::maxSequenceForward":[46,2,1,"_CPPv2N6paddle6Matrix18maxSequenceForwardER6MatrixRK7IVectorR7IVector"],"paddle::Matrix::memoryHandle_":[46,3,1,"_CPPv2N6paddle6Matrix13memoryHandle_E"],"paddle::Matrix::mul":[46,2,1,"_CPPv2N6paddle6Matrix3mulEK9MatrixPtrK9MatrixPtr4real4real"],"paddle::Matrix::mulByBitCode":[46,2,1,"_CPPv2N6paddle6Matrix12mulByBitCodeE6size_tRK7IVectorRK6MatrixRK6Matrix"],"paddle::Matrix::mulByBitCodeBackwardError":[46,2,1,"_CPPv2N6paddle6Matrix25mulByBitCodeBackwardErrorE6size_tRK7IVectorRK6MatrixR6Matrix"],"paddle::Matrix::mulByBitCodeBackwardWeight":[46,2,1,"_CPPv2N6paddle6Matrix26mulByBitCodeBackwardWeightE6size_tRK7IVectorR6MatrixRK6Matrix"],"paddle::Matrix::multiBinaryLabelCrossEntropy":[46,2,1,"_CPPv2N6paddle6Matrix28multiBinaryLabelCrossEntropyER6MatrixR6Matrix"],"paddle::Matrix::multiBinaryLabelCrossEntropyBp":[46,2,1,"_CPPv2N6paddle6Matrix30multiBinaryLabelCrossEntropyBpER6MatrixR6Matrix"],"paddle::Matrix::oneHotCrossEntropy":[46,2,1,"_CPPv2N6paddle6Matrix18oneHotCrossEntropyER6MatrixR7IVector"],"paddle::Matrix::oneHotCrossEntropyBp":[46,2,1,"_CPPv2N6paddle6Matrix20oneHotCrossEntropyBpER6MatrixR7IVector"],"paddle::Matrix::oneHotCrossEntropyWithSelfNorm":[46,2,1,"_CPPv2N6paddle6Matrix30oneHotCrossEntropyWithSelfNormER6MatrixR7IVector4real"],"paddle::Matrix::oneHotCrossEntropyWithSelfNormBp":[46,2,1,"_CPPv2N6paddle6Matrix32oneHotCrossEntropyWithSelfNormBpER6MatrixR7IVector4real"],"paddle::Matrix::paramReluBackwardDiff":[46,2,1,"_CPPv2N6paddle6Matrix21paramReluBackwardDiffER6MatrixR6MatrixR6Matrix"],"paddle::Matrix::paramReluBackwardW":[46,2,1,"_CPPv2N6paddle6Matrix18paramReluBackwardWER6MatrixR6Matrix"],"paddle::Matrix::paramReluForward":[46,2,1,"_CPPv2N6paddle6Matrix16paramReluForwardER6MatrixR6Matrix"],"paddle::Matrix::print":[46,2,1,"_CPPv2N6paddle6Matrix5printERNSt7ostreamE6size_t6size_t"],"paddle::Matrix::printOneRow":[46,2,1,"_CPPv2N6paddle6Matrix11printOneRowERNSt7ostreamE6size_t"],"paddle::Matrix::randomizeUniform":[46,2,1,"_CPPv2N6paddle6Matrix16randomizeUniformEv"],"paddle::Matrix::resetOne":[46,2,1,"_CPPv2N6paddle6Matrix8resetOneEv"],"paddle::Matrix::resize":[46,2,1,"_CPPv2N6paddle6Matrix6resizeE6size_t6size_t6size_t15SparseValueType12SparseFormat"],"paddle::Matrix::rightMul":[46,2,1,"_CPPv2N6paddle6Matrix8rightMulER6Matrix4real4real"],"paddle::Matrix::rowMax":[46,2,1,"_CPPv2N6paddle6Matrix6rowMaxER7IVectorR6Matrix"],"paddle::Matrix::rowMaxId":[46,2,1,"_CPPv2N6paddle6Matrix8rowMaxIdER7IVector"],"paddle::Matrix::rowNormalizeL1":[46,2,1,"_CPPv2N6paddle6Matrix14rowNormalizeL1ER6Matrix"],"paddle::Matrix::rowSum":[46,2,1,"_CPPv2N6paddle6Matrix6rowSumER6Matrix"],"paddle::Matrix::scaledTanh":[46,2,1,"_CPPv2N6paddle6Matrix10scaledTanhER6Matrix4real4real"],"paddle::Matrix::selectElements":[46,2,1,"_CPPv2N6paddle6Matrix14selectElementsER6MatrixR7IVector"],"paddle::Matrix::selectRows":[46,2,1,"_CPPv2N6paddle6Matrix10selectRowsER6MatrixR7IVector"],"paddle::Matrix::sequenceAvgForward":[46,2,1,"_CPPv2N6paddle6Matrix18sequenceAvgForwardER6MatrixRK7IVectori"],"paddle::Matrix::sequenceSoftmax":[46,2,1,"_CPPv2N6paddle6Matrix15sequenceSoftmaxER6MatrixRK7IVector"],"paddle::Matrix::setData":[46,2,1,"_CPPv2N6paddle6Matrix7setDataEP4real6size_t6size_t"],"paddle::Matrix::setRow":[46,2,1,"_CPPv2N6paddle6Matrix6setRowE6size_t6size_tPKjPK4real"],"paddle::Matrix::softmax":[46,2,1,"_CPPv2N6paddle6Matrix7softmaxER6Matrix"],"paddle::Matrix::softmaxBackward":[46,2,1,"_CPPv2N6paddle6Matrix15softmaxBackwardER6Matrix"],"paddle::Matrix::softmaxDerivative":[46,2,1,"_CPPv2N6paddle6Matrix17softmaxDerivativeER6MatrixR6Matrix"],"paddle::Matrix::softrelu":[46,2,1,"_CPPv2N6paddle6Matrix8softreluER6Matrix"],"paddle::Matrix::softreluDerivative":[46,2,1,"_CPPv2N6paddle6Matrix18softreluDerivativeER6Matrix"],"paddle::Matrix::subByBitCode":[46,2,1,"_CPPv2N6paddle6Matrix12subByBitCodeE6size_tR7IVector"],"paddle::Matrix::subColMatrix":[46,2,1,"_CPPv2N6paddle6Matrix12subColMatrixE6size_t6size_t"],"paddle::Matrix::subMatrix":[46,2,1,"_CPPv2N6paddle6Matrix9subMatrixE6size_t6size_t9MatrixPtr"],"paddle::Matrix::subRowMatrix":[46,2,1,"_CPPv2N6paddle6Matrix12subRowMatrixE6size_t6size_t"],"paddle::Matrix::sumByBitCode":[46,2,1,"_CPPv2N6paddle6Matrix12sumByBitCodeE6size_tR7IVectorR6Matrix4real"],"paddle::Matrix::sumOfSquares":[46,2,1,"_CPPv2N6paddle6Matrix12sumOfSquaresER6MatrixR6Matrix"],"paddle::Matrix::sumOfSquaresBp":[46,2,1,"_CPPv2N6paddle6Matrix14sumOfSquaresBpER6MatrixR6Matrix"],"paddle::Matrix::tanh":[46,2,1,"_CPPv2N6paddle6Matrix4tanhER6Matrix"],"paddle::Matrix::tanhDerivative":[46,2,1,"_CPPv2N6paddle6Matrix14tanhDerivativeER6Matrix"],"paddle::Matrix::tmpMat_":[46,3,1,"_CPPv2N6paddle6Matrix7tmpMat_E"],"paddle::Matrix::transpose":[46,2,1,"_CPPv2N6paddle6Matrix9transposeE9MatrixPtrb"],"paddle::Matrix::trimFrom":[46,2,1,"_CPPv2N6paddle6Matrix8trimFromERK15CpuSparseMatrix"],"paddle::Matrix::zeroMem":[46,2,1,"_CPPv2N6paddle6Matrix7zeroMemEv"],"paddle::Matrix::~Matrix":[46,2,1,"_CPPv2N6paddle6MatrixD0Ev"],"paddle::MatrixOffset":[46,1,1,"_CPPv2N6paddle12MatrixOffsetE"],"paddle::MatrixOffset::MatrixOffset":[46,2,1,"_CPPv2N6paddle12MatrixOffset12MatrixOffsetE6size_t6size_t6size_t6size_t6size_t6size_t6size_t6size_t"],"paddle::MatrixOffset::aCol_":[46,3,1,"_CPPv2N6paddle12MatrixOffset5aCol_E"],"paddle::MatrixOffset::aRow_":[46,3,1,"_CPPv2N6paddle12MatrixOffset5aRow_E"],"paddle::MatrixOffset::bCol_":[46,3,1,"_CPPv2N6paddle12MatrixOffset5bCol_E"],"paddle::MatrixOffset::bRow_":[46,3,1,"_CPPv2N6paddle12MatrixOffset5bRow_E"],"paddle::MatrixOffset::cCol_":[46,3,1,"_CPPv2N6paddle12MatrixOffset5cCol_E"],"paddle::MatrixOffset::cRow_":[46,3,1,"_CPPv2N6paddle12MatrixOffset5cRow_E"],"paddle::MatrixOffset::dCol_":[46,3,1,"_CPPv2N6paddle12MatrixOffset5dCol_E"],"paddle::MatrixOffset::dRow_":[46,3,1,"_CPPv2N6paddle12MatrixOffset5dRow_E"],"paddle::MatrixPtr":[46,4,1,"_CPPv2N6paddle9MatrixPtrE"],"paddle::MaxFramePrinter":[38,1,1,"_CPPv2N6paddle15MaxFramePrinterE"],"paddle::MaxFramePrinter::MaxFramePrinter":[38,2,1,"_CPPv2N6paddle15MaxFramePrinter15MaxFramePrinterEv"],"paddle::MaxFramePrinter::eval":[38,2,1,"_CPPv2N6paddle15MaxFramePrinter4evalERK13NeuralNetwork"],"paddle::MaxFramePrinter::evalImp":[38,2,1,"_CPPv2N6paddle15MaxFramePrinter7evalImpERNSt6vectorI8ArgumentEE"],"paddle::MaxFramePrinter::updateSamplesNum":[38,2,1,"_CPPv2N6paddle15MaxFramePrinter16updateSamplesNumERKNSt6vectorI8ArgumentEE"],"paddle::MaxIdLayer":[43,1,1,"_CPPv2N6paddle10MaxIdLayerE"],"paddle::MaxIdLayer::MaxIdLayer":[43,2,1,"_CPPv2N6paddle10MaxIdLayer10MaxIdLayerERK11LayerConfig"],"paddle::MaxIdLayer::backward":[43,2,1,"_CPPv2N6paddle10MaxIdLayer8backwardERK14UpdateCallback"],"paddle::MaxIdLayer::forward":[43,2,1,"_CPPv2N6paddle10MaxIdLayer7forwardE8PassType"],"paddle::MaxIdLayer::init":[43,2,1,"_CPPv2N6paddle10MaxIdLayer4initERK8LayerMapRK12ParameterMap"],"paddle::MaxIdPrinter":[38,1,1,"_CPPv2N6paddle12MaxIdPrinterE"],"paddle::MaxIdPrinter::MaxIdPrinter":[38,2,1,"_CPPv2N6paddle12MaxIdPrinter12MaxIdPrinterEv"],"paddle::MaxIdPrinter::eval":[38,2,1,"_CPPv2N6paddle12MaxIdPrinter4evalERK13NeuralNetwork"],"paddle::MaxIdPrinter::evalImp":[38,2,1,"_CPPv2N6paddle12MaxIdPrinter7evalImpERNSt6vectorI8ArgumentEE"],"paddle::MaxIdPrinter::updateSamplesNum":[38,2,1,"_CPPv2N6paddle12MaxIdPrinter16updateSamplesNumERKNSt6vectorI8ArgumentEE"],"paddle::MaxLayer":[43,1,1,"_CPPv2N6paddle8MaxLayerE"],"paddle::MaxLayer::MaxLayer":[43,2,1,"_CPPv2N6paddle8MaxLayer8MaxLayerERK11LayerConfig"],"paddle::MaxLayer::MaxLevel":[43,4,1,"_CPPv2N6paddle8MaxLayer8MaxLevelE"],"paddle::MaxLayer::backward":[43,2,1,"_CPPv2N6paddle8MaxLayer8backwardERK14UpdateCallback"],"paddle::MaxLayer::biases_":[43,3,1,"_CPPv2N6paddle8MaxLayer7biases_E"],"paddle::MaxLayer::forward":[43,2,1,"_CPPv2N6paddle8MaxLayer7forwardE8PassType"],"paddle::MaxLayer::init":[43,2,1,"_CPPv2N6paddle8MaxLayer4initERK8LayerMapRK12ParameterMap"],"paddle::MaxLayer::kNonSeq":[43,1,1,"_CPPv2N6paddle8MaxLayer7kNonSeqE"],"paddle::MaxLayer::kSeq":[43,1,1,"_CPPv2N6paddle8MaxLayer4kSeqE"],"paddle::MaxLayer::maxIndex_":[43,3,1,"_CPPv2N6paddle8MaxLayer9maxIndex_E"],"paddle::MaxLayer::type_":[43,3,1,"_CPPv2N6paddle8MaxLayer5type_E"],"paddle::MaxLayer::~MaxLayer":[43,2,1,"_CPPv2N6paddle8MaxLayerD0Ev"],"paddle::MemoryHandle":[48,1,1,"_CPPv2N6paddle12MemoryHandleE"],"paddle::MemoryHandle::MemoryHandle":[48,2,1,"_CPPv2N6paddle12MemoryHandle12MemoryHandleE6size_t"],"paddle::MemoryHandle::allocSize_":[48,3,1,"_CPPv2N6paddle12MemoryHandle10allocSize_E"],"paddle::MemoryHandle::allocator_":[48,3,1,"_CPPv2N6paddle12MemoryHandle10allocator_E"],"paddle::MemoryHandle::buf_":[48,3,1,"_CPPv2N6paddle12MemoryHandle4buf_E"],"paddle::MemoryHandle::deviceId_":[48,3,1,"_CPPv2N6paddle12MemoryHandle9deviceId_E"],"paddle::MemoryHandle::getAllocSize":[48,2,1,"_CPPv2N6paddle12MemoryHandle12getAllocSizeEv"],"paddle::MemoryHandle::getBuf":[48,2,1,"_CPPv2N6paddle12MemoryHandle6getBufEv"],"paddle::MemoryHandle::getSize":[48,2,1,"_CPPv2N6paddle12MemoryHandle7getSizeEv"],"paddle::MemoryHandle::size_":[48,3,1,"_CPPv2N6paddle12MemoryHandle5size_E"],"paddle::MemoryHandle::~MemoryHandle":[48,2,1,"_CPPv2N6paddle12MemoryHandleD0Ev"],"paddle::MemoryHandlePtr":[48,4,1,"_CPPv2N6paddle15MemoryHandlePtrE"],"paddle::MixedLayer":[43,1,1,"_CPPv2N6paddle10MixedLayerE"],"paddle::MixedLayer::MixedLayer":[43,2,1,"_CPPv2N6paddle10MixedLayer10MixedLayerERK11LayerConfig"],"paddle::MixedLayer::backward":[43,2,1,"_CPPv2N6paddle10MixedLayer8backwardERK14UpdateCallback"],"paddle::MixedLayer::biases_":[43,3,1,"_CPPv2N6paddle10MixedLayer7biases_E"],"paddle::MixedLayer::forward":[43,2,1,"_CPPv2N6paddle10MixedLayer7forwardE8PassType"],"paddle::MixedLayer::getState":[43,2,1,"_CPPv2N6paddle10MixedLayer8getStateEv"],"paddle::MixedLayer::init":[43,2,1,"_CPPv2N6paddle10MixedLayer4initERK8LayerMapRK12ParameterMap"],"paddle::MixedLayer::operators_":[43,3,1,"_CPPv2N6paddle10MixedLayer10operators_E"],"paddle::MixedLayer::prefetch":[43,2,1,"_CPPv2N6paddle10MixedLayer8prefetchEv"],"paddle::MixedLayer::projectionStateMatrixSize_":[43,3,1,"_CPPv2N6paddle10MixedLayer26projectionStateMatrixSize_E"],"paddle::MixedLayer::projections_":[43,3,1,"_CPPv2N6paddle10MixedLayer12projections_E"],"paddle::MixedLayer::resetState":[43,2,1,"_CPPv2N6paddle10MixedLayer10resetStateEv"],"paddle::MixedLayer::setState":[43,2,1,"_CPPv2N6paddle10MixedLayer8setStateE13LayerStatePtr"],"paddle::MixedLayer::~MixedLayer":[43,2,1,"_CPPv2N6paddle10MixedLayerD0Ev"],"paddle::MsgReader":[58,1,1,"_CPPv2N6paddle9MsgReaderE"],"paddle::MsgReader::MsgReader":[58,2,1,"_CPPv2N6paddle9MsgReader9MsgReaderEP13SocketChannel6size_t"],"paddle::MsgReader::blockLengths_":[58,3,1,"_CPPv2N6paddle9MsgReader13blockLengths_E"],"paddle::MsgReader::channel_":[58,3,1,"_CPPv2N6paddle9MsgReader8channel_E"],"paddle::MsgReader::currentBlockIndex_":[58,3,1,"_CPPv2N6paddle9MsgReader18currentBlockIndex_E"],"paddle::MsgReader::getBlockLength":[58,2,1,"_CPPv2N6paddle9MsgReader14getBlockLengthE6size_t"],"paddle::MsgReader::getNextBlockLength":[58,2,1,"_CPPv2N6paddle9MsgReader18getNextBlockLengthEv"],"paddle::MsgReader::getNumBlocks":[58,2,1,"_CPPv2N6paddle9MsgReader12getNumBlocksEv"],"paddle::MsgReader::getTotalLength":[58,2,1,"_CPPv2N6paddle9MsgReader14getTotalLengthEv"],"paddle::MsgReader::readBlocks":[58,2,1,"_CPPv2N6paddle9MsgReader10readBlocksERKNSt6vectorIPvEE"],"paddle::MsgReader::readNextBlock":[58,2,1,"_CPPv2N6paddle9MsgReader13readNextBlockEPv"],"paddle::MsgReader::~MsgReader":[58,2,1,"_CPPv2N6paddle9MsgReaderD0Ev"],"paddle::MultiBinaryLabelCrossEntropy":[43,1,1,"_CPPv2N6paddle28MultiBinaryLabelCrossEntropyE"],"paddle::MultiBinaryLabelCrossEntropy::MultiBinaryLabelCrossEntropy":[43,2,1,"_CPPv2N6paddle28MultiBinaryLabelCrossEntropy28MultiBinaryLabelCrossEntropyERK11LayerConfig"],"paddle::MultiBinaryLabelCrossEntropy::backwardImp":[43,2,1,"_CPPv2N6paddle28MultiBinaryLabelCrossEntropy11backwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::MultiBinaryLabelCrossEntropy::forwardImp":[43,2,1,"_CPPv2N6paddle28MultiBinaryLabelCrossEntropy10forwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::MultiBinaryLabelCrossEntropy::init":[43,2,1,"_CPPv2N6paddle28MultiBinaryLabelCrossEntropy4initERK8LayerMapRK12ParameterMap"],"paddle::MultiBinaryLabelCrossEntropy::targetPerDim_":[43,3,1,"_CPPv2N6paddle28MultiBinaryLabelCrossEntropy13targetPerDim_E"],"paddle::MultiClassCrossEntropy":[43,1,1,"_CPPv2N6paddle22MultiClassCrossEntropyE"],"paddle::MultiClassCrossEntropy::MultiClassCrossEntropy":[43,2,1,"_CPPv2N6paddle22MultiClassCrossEntropy22MultiClassCrossEntropyERK11LayerConfig"],"paddle::MultiClassCrossEntropy::backwardImp":[43,2,1,"_CPPv2N6paddle22MultiClassCrossEntropy11backwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::MultiClassCrossEntropy::forwardImp":[43,2,1,"_CPPv2N6paddle22MultiClassCrossEntropy10forwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::MultiClassCrossEntropy::init":[43,2,1,"_CPPv2N6paddle22MultiClassCrossEntropy4initERK8LayerMapRK12ParameterMap"],"paddle::MultiClassCrossEntropyWithSelfNorm":[43,1,1,"_CPPv2N6paddle34MultiClassCrossEntropyWithSelfNormE"],"paddle::MultiClassCrossEntropyWithSelfNorm::MultiClassCrossEntropyWithSelfNorm":[43,2,1,"_CPPv2N6paddle34MultiClassCrossEntropyWithSelfNorm34MultiClassCrossEntropyWithSelfNormERK11LayerConfig"],"paddle::MultiClassCrossEntropyWithSelfNorm::backwardImp":[43,2,1,"_CPPv2N6paddle34MultiClassCrossEntropyWithSelfNorm11backwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::MultiClassCrossEntropyWithSelfNorm::forwardImp":[43,2,1,"_CPPv2N6paddle34MultiClassCrossEntropyWithSelfNorm10forwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::MultiClassCrossEntropyWithSelfNorm::init":[43,2,1,"_CPPv2N6paddle34MultiClassCrossEntropyWithSelfNorm4initERK8LayerMapRK12ParameterMap"],"paddle::MultiClassCrossEntropyWithSelfNorm::sftMaxSum_":[43,3,1,"_CPPv2N6paddle34MultiClassCrossEntropyWithSelfNorm10sftMaxSum_E"],"paddle::MultiClassCrossEntropyWithSelfNorm::sumInv_":[43,3,1,"_CPPv2N6paddle34MultiClassCrossEntropyWithSelfNorm7sumInv_E"],"paddle::MultiDataProvider":[36,1,1,"_CPPv2N6paddle17MultiDataProviderE"],"paddle::MultiDataProvider::MultiDataProvider":[36,2,1,"_CPPv2N6paddle17MultiDataProvider17MultiDataProviderERK10DataConfigb"],"paddle::MultiDataProvider::getNextBatchInternal":[36,2,1,"_CPPv2N6paddle17MultiDataProvider20getNextBatchInternalE7int64_tP9DataBatch"],"paddle::MultiDataProvider::getSize":[36,2,1,"_CPPv2N6paddle17MultiDataProvider7getSizeEv"],"paddle::MultiDataProvider::isTestMode":[36,2,1,"_CPPv2N6paddle17MultiDataProvider10isTestModeEv"],"paddle::MultiDataProvider::reset":[36,2,1,"_CPPv2N6paddle17MultiDataProvider5resetEv"],"paddle::MultiDataProvider::shuffle":[36,2,1,"_CPPv2N6paddle17MultiDataProvider7shuffleEv"],"paddle::MultiDataProvider::subDataProviders_":[36,3,1,"_CPPv2N6paddle17MultiDataProvider17subDataProviders_E"],"paddle::MultiDataProvider::~MultiDataProvider":[36,2,1,"_CPPv2N6paddle17MultiDataProviderD0Ev"],"paddle::MultiGradientMachine":[40,1,1,"_CPPv2N6paddle20MultiGradientMachineE"],"paddle::MultiGradientMachine::MultiGradientMachine":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine20MultiGradientMachineERK11ModelConfigb"],"paddle::MultiGradientMachine::TASK_BACKWARD":[40,1,1,"_CPPv2N6paddle20MultiGradientMachine13TASK_BACKWARDE"],"paddle::MultiGradientMachine::TASK_COPY_IN_ARGS":[40,1,1,"_CPPv2N6paddle20MultiGradientMachine17TASK_COPY_IN_ARGSE"],"paddle::MultiGradientMachine::TASK_FORWARD":[40,1,1,"_CPPv2N6paddle20MultiGradientMachine12TASK_FORWARDE"],"paddle::MultiGradientMachine::TASK_FORWARD_BACKWARD":[40,1,1,"_CPPv2N6paddle20MultiGradientMachine21TASK_FORWARD_BACKWARDE"],"paddle::MultiGradientMachine::TaskType":[40,4,1,"_CPPv2N6paddle20MultiGradientMachine8TaskTypeE"],"paddle::MultiGradientMachine::allBarrier_":[40,3,1,"_CPPv2N6paddle20MultiGradientMachine11allBarrier_E"],"paddle::MultiGradientMachine::allocGradBufs":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine13allocGradBufsEv"],"paddle::MultiGradientMachine::backward":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine8backwardERK14UpdateCallback"],"paddle::MultiGradientMachine::backwardCallback_":[40,3,1,"_CPPv2N6paddle20MultiGradientMachine17backwardCallback_E"],"paddle::MultiGradientMachine::backwardImp":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine11backwardImpERK14UpdateCallback"],"paddle::MultiGradientMachine::bufferSizes_":[40,3,1,"_CPPv2N6paddle20MultiGradientMachine12bufferSizes_E"],"paddle::MultiGradientMachine::eval":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine4evalEP9Evaluator"],"paddle::MultiGradientMachine::finish":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine6finishEv"],"paddle::MultiGradientMachine::forward":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine7forwardERKNSt6vectorI8ArgumentEEPNSt6vectorI8ArgumentEE8PassType"],"paddle::MultiGradientMachine::forwardBackward":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine15forwardBackwardERKNSt6vectorI8ArgumentEEPNSt6vectorI8ArgumentEE8PassTypeRK14UpdateCallback"],"paddle::MultiGradientMachine::forwardImp":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine10forwardImpERKNSt6vectorI8ArgumentEEPNSt6vectorI8ArgumentEE8PassType8TaskType"],"paddle::MultiGradientMachine::getAllThreads":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine13getAllThreadsEv"],"paddle::MultiGradientMachine::getBackwardCallback":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine19getBackwardCallbackEv"],"paddle::MultiGradientMachine::getGradBuf":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine10getGradBufEi"],"paddle::MultiGradientMachine::getInArgs":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine9getInArgsEv"],"paddle::MultiGradientMachine::getNumDevices":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine13getNumDevicesEv"],"paddle::MultiGradientMachine::getNumLogicalDevices":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine20getNumLogicalDevicesEv"],"paddle::MultiGradientMachine::getNumThreads":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine13getNumThreadsEv"],"paddle::MultiGradientMachine::getOutArgs":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine10getOutArgsEPNSt6vectorI8ArgumentEE8PassType"],"paddle::MultiGradientMachine::getPassType":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine11getPassTypeEv"],"paddle::MultiGradientMachine::getSlaveParameters":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine18getSlaveParametersEv"],"paddle::MultiGradientMachine::getTaskType":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine11getTaskTypeEv"],"paddle::MultiGradientMachine::getThread":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine9getThreadEi"],"paddle::MultiGradientMachine::gradBufs_":[40,3,1,"_CPPv2N6paddle20MultiGradientMachine9gradBufs_E"],"paddle::MultiGradientMachine::gradQueue_":[40,3,1,"_CPPv2N6paddle20MultiGradientMachine10gradQueue_E"],"paddle::MultiGradientMachine::gradientMachine_":[40,3,1,"_CPPv2N6paddle20MultiGradientMachine16gradientMachine_E"],"paddle::MultiGradientMachine::hasNonstaticCpuParamters":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine24hasNonstaticCpuParamtersEv"],"paddle::MultiGradientMachine::hasNonstaticCpuParamters_":[40,3,1,"_CPPv2N6paddle20MultiGradientMachine25hasNonstaticCpuParamters_E"],"paddle::MultiGradientMachine::inArgsCopied_":[40,3,1,"_CPPv2N6paddle20MultiGradientMachine13inArgsCopied_E"],"paddle::MultiGradientMachine::inArgs_":[40,3,1,"_CPPv2N6paddle20MultiGradientMachine7inArgs_E"],"paddle::MultiGradientMachine::isPassGrad":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine10isPassGradEv"],"paddle::MultiGradientMachine::isPassGrad_":[40,3,1,"_CPPv2N6paddle20MultiGradientMachine11isPassGrad_E"],"paddle::MultiGradientMachine::logicalDeviceId2RealDeviceId":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine28logicalDeviceId2RealDeviceIdEii"],"paddle::MultiGradientMachine::makeEvaluator":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine13makeEvaluatorEv"],"paddle::MultiGradientMachine::mergeTypes_":[40,3,1,"_CPPv2N6paddle20MultiGradientMachine11mergeTypes_E"],"paddle::MultiGradientMachine::notifyGradientTransfer":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine22notifyGradientTransferEi"],"paddle::MultiGradientMachine::numDevices_":[40,3,1,"_CPPv2N6paddle20MultiGradientMachine11numDevices_E"],"paddle::MultiGradientMachine::numLogicalDevices_":[40,3,1,"_CPPv2N6paddle20MultiGradientMachine18numLogicalDevices_E"],"paddle::MultiGradientMachine::numThreads_":[40,3,1,"_CPPv2N6paddle20MultiGradientMachine11numThreads_E"],"paddle::MultiGradientMachine::onPassEnd":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine9onPassEndEv"],"paddle::MultiGradientMachine::outArgStream_":[40,3,1,"_CPPv2N6paddle20MultiGradientMachine13outArgStream_E"],"paddle::MultiGradientMachine::outArgs_":[40,3,1,"_CPPv2N6paddle20MultiGradientMachine8outArgs_E"],"paddle::MultiGradientMachine::paraMainThread":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine14paraMainThreadEi"],"paddle::MultiGradientMachine::paraMainThread_":[40,3,1,"_CPPv2N6paddle20MultiGradientMachine15paraMainThread_E"],"paddle::MultiGradientMachine::passType_":[40,3,1,"_CPPv2N6paddle20MultiGradientMachine9passType_E"],"paddle::MultiGradientMachine::prefetch":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine8prefetchERKNSt6vectorI8ArgumentEE"],"paddle::MultiGradientMachine::realDeviceId2LogicalDeviceId":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine28realDeviceId2LogicalDeviceIdEii"],"paddle::MultiGradientMachine::setOutputGrad":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine13setOutputGradERKNSt6vectorI8ArgumentEE"],"paddle::MultiGradientMachine::setPassGrad":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine11setPassGradEb"],"paddle::MultiGradientMachine::startTask":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine9startTaskE8TaskType"],"paddle::MultiGradientMachine::taskType_":[40,3,1,"_CPPv2N6paddle20MultiGradientMachine9taskType_E"],"paddle::MultiGradientMachine::threads_":[40,3,1,"_CPPv2N6paddle20MultiGradientMachine8threads_E"],"paddle::MultiGradientMachine::trainerBarrier_":[40,3,1,"_CPPv2N6paddle20MultiGradientMachine15trainerBarrier_E"],"paddle::MultiGradientMachine::updateThreadParameters":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine22updateThreadParametersEv"],"paddle::MultiGradientMachine::useGpu":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine6useGpuEv"],"paddle::MultiGradientMachine::useGpu_":[40,3,1,"_CPPv2N6paddle20MultiGradientMachine7useGpu_E"],"paddle::MultiGradientMachine::waitAfterMerge":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine14waitAfterMergeEv"],"paddle::MultiGradientMachine::waitBeforeMerge":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine15waitBeforeMergeEv"],"paddle::MultiGradientMachine::waitForCopyInArgs":[40,2,1,"_CPPv2N6paddle20MultiGradientMachine17waitForCopyInArgsEv"],"paddle::MultiThreadWorker":[64,1,1,"_CPPv2N6paddle17MultiThreadWorkerE"],"paddle::MultiThreadWorker::JobFunc":[64,4,1,"_CPPv2N6paddle17MultiThreadWorker7JobFuncE"],"paddle::MultiThreadWorker::MultiThreadWorker":[64,2,1,"_CPPv2N6paddle17MultiThreadWorker17MultiThreadWorkerE6size_t6size_t"],"paddle::MultiThreadWorker::ResultPtrType":[64,4,1,"_CPPv2N6paddle17MultiThreadWorker13ResultPtrTypeE"],"paddle::MultiThreadWorker::ResultType":[64,4,1,"_CPPv2N6paddle17MultiThreadWorker10ResultTypeE"],"paddle::MultiThreadWorker::addJob":[64,2,1,"_CPPv2N6paddle17MultiThreadWorker6addJobE7JobFunc"],"paddle::MultiThreadWorker::forceStop":[64,2,1,"_CPPv2N6paddle17MultiThreadWorker9forceStopEv"],"paddle::MultiThreadWorker::jobAdding_":[64,3,1,"_CPPv2N6paddle17MultiThreadWorker10jobAdding_E"],"paddle::MultiThreadWorker::jobs_":[64,3,1,"_CPPv2N6paddle17MultiThreadWorker5jobs_E"],"paddle::MultiThreadWorker::nullResultNum_":[64,3,1,"_CPPv2N6paddle17MultiThreadWorker14nullResultNum_E"],"paddle::MultiThreadWorker::results_":[64,3,1,"_CPPv2N6paddle17MultiThreadWorker8results_E"],"paddle::MultiThreadWorker::run":[64,2,1,"_CPPv2N6paddle17MultiThreadWorker3runEv"],"paddle::MultiThreadWorker::stop":[64,2,1,"_CPPv2N6paddle17MultiThreadWorker4stopEv"],"paddle::MultiThreadWorker::stopAddJob":[64,2,1,"_CPPv2N6paddle17MultiThreadWorker10stopAddJobEv"],"paddle::MultiThreadWorker::stopping_":[64,3,1,"_CPPv2N6paddle17MultiThreadWorker9stopping_E"],"paddle::MultiThreadWorker::testResult":[64,2,1,"_CPPv2N6paddle17MultiThreadWorker10testResultEv"],"paddle::MultiThreadWorker::waitResult":[64,2,1,"_CPPv2N6paddle17MultiThreadWorker10waitResultEv"],"paddle::MultiThreadWorker::workers_":[64,3,1,"_CPPv2N6paddle17MultiThreadWorker8workers_E"],"paddle::MultiThreadWorker::~MultiThreadWorker":[64,2,1,"_CPPv2N6paddle17MultiThreadWorkerD0Ev"],"paddle::MultinomialSampler":[43,1,1,"_CPPv2N6paddle18MultinomialSamplerE"],"paddle::MultinomialSampler::Interval":[43,1,1,"_CPPv2N6paddle18MultinomialSampler8IntervalE"],"paddle::MultinomialSampler::Interval::otherId":[43,3,1,"_CPPv2N6paddle18MultinomialSampler8Interval7otherIdE"],"paddle::MultinomialSampler::Interval::thresh":[43,3,1,"_CPPv2N6paddle18MultinomialSampler8Interval6threshE"],"paddle::MultinomialSampler::MultinomialSampler":[43,2,1,"_CPPv2N6paddle18MultinomialSampler18MultinomialSamplerEPK4reali"],"paddle::MultinomialSampler::gen":[43,2,1,"_CPPv2N6paddle18MultinomialSampler3genER4URNG"],"paddle::MultinomialSampler::gen1":[43,2,1,"_CPPv2N6paddle18MultinomialSampler4gen1E4Rand"],"paddle::MultinomialSampler::intervals_":[43,3,1,"_CPPv2N6paddle18MultinomialSampler10intervals_E"],"paddle::MultinomialSampler::rand_":[43,3,1,"_CPPv2N6paddle18MultinomialSampler5rand_E"],"paddle::MultiplexLayer":[43,1,1,"_CPPv2N6paddle14MultiplexLayerE"],"paddle::MultiplexLayer::CopyInfo":[43,1,1,"_CPPv2N6paddle14MultiplexLayer8CopyInfoE"],"paddle::MultiplexLayer::CopyInfo::CopyInfo":[43,2,1,"_CPPv2N6paddle14MultiplexLayer8CopyInfo8CopyInfoEiii"],"paddle::MultiplexLayer::CopyInfo::copyIdx":[43,3,1,"_CPPv2N6paddle14MultiplexLayer8CopyInfo7copyIdxE"],"paddle::MultiplexLayer::CopyInfo::length":[43,3,1,"_CPPv2N6paddle14MultiplexLayer8CopyInfo6lengthE"],"paddle::MultiplexLayer::CopyInfo::startIdx":[43,3,1,"_CPPv2N6paddle14MultiplexLayer8CopyInfo8startIdxE"],"paddle::MultiplexLayer::MultiplexLayer":[43,2,1,"_CPPv2N6paddle14MultiplexLayer14MultiplexLayerERK11LayerConfig"],"paddle::MultiplexLayer::backward":[43,2,1,"_CPPv2N6paddle14MultiplexLayer8backwardERK14UpdateCallback"],"paddle::MultiplexLayer::copySchedule_":[43,3,1,"_CPPv2N6paddle14MultiplexLayer13copySchedule_E"],"paddle::MultiplexLayer::forward":[43,2,1,"_CPPv2N6paddle14MultiplexLayer7forwardE8PassType"],"paddle::MultiplexLayer::init":[43,2,1,"_CPPv2N6paddle14MultiplexLayer4initERK8LayerMapRK12ParameterMap"],"paddle::MultiplexLayer::tmpDest_":[43,3,1,"_CPPv2N6paddle14MultiplexLayer8tmpDest_E"],"paddle::MultiplexLayer::tmpSrc_":[43,3,1,"_CPPv2N6paddle14MultiplexLayer7tmpSrc_E"],"paddle::MultiplexLayer::~MultiplexLayer":[43,2,1,"_CPPv2N6paddle14MultiplexLayerD0Ev"],"paddle::NCELayer":[43,1,1,"_CPPv2N6paddle8NCELayerE"],"paddle::NCELayer::NCELayer":[43,2,1,"_CPPv2N6paddle8NCELayer8NCELayerERK11LayerConfig"],"paddle::NCELayer::backward":[43,2,1,"_CPPv2N6paddle8NCELayer8backwardERK14UpdateCallback"],"paddle::NCELayer::backwardBias":[43,2,1,"_CPPv2N6paddle8NCELayer12backwardBiasERK14UpdateCallback"],"paddle::NCELayer::backwardCost":[43,2,1,"_CPPv2N6paddle8NCELayer12backwardCostEv"],"paddle::NCELayer::backwardOneInput":[43,2,1,"_CPPv2N6paddle8NCELayer16backwardOneInputEiRK14UpdateCallback"],"paddle::NCELayer::forward":[43,2,1,"_CPPv2N6paddle8NCELayer7forwardE8PassType"],"paddle::NCELayer::forwardBias":[43,2,1,"_CPPv2N6paddle8NCELayer11forwardBiasEv"],"paddle::NCELayer::forwardCost":[43,2,1,"_CPPv2N6paddle8NCELayer11forwardCostEv"],"paddle::NCELayer::forwardOneInput":[43,2,1,"_CPPv2N6paddle8NCELayer15forwardOneInputEi"],"paddle::NCELayer::init":[43,2,1,"_CPPv2N6paddle8NCELayer4initERK8LayerMapRK12ParameterMap"],"paddle::NCELayer::prefetch":[43,2,1,"_CPPv2N6paddle8NCELayer8prefetchEv"],"paddle::NCELayer::prepareSamples":[43,2,1,"_CPPv2N6paddle8NCELayer14prepareSamplesEv"],"paddle::NO_VALUE":[46,1,1,"_CPPv2N6paddle8NO_VALUEE"],"paddle::NeuralNetwork":[40,1,1,"_CPPv2N6paddle13NeuralNetworkE"],"paddle::NeuralNetwork::NeuralNetwork":[40,2,1,"_CPPv2N6paddle13NeuralNetwork13NeuralNetworkENSt6stringEP13NeuralNetwork"],"paddle::NeuralNetwork::backward":[40,2,1,"_CPPv2N6paddle13NeuralNetwork8backwardERK14UpdateCallback"],"paddle::NeuralNetwork::config_":[40,3,1,"_CPPv2N6paddle13NeuralNetwork7config_E"],"paddle::NeuralNetwork::connect":[40,2,1,"_CPPv2N6paddle13NeuralNetwork7connectENSt6stringEP13NeuralNetworkNSt6stringE"],"paddle::NeuralNetwork::create":[40,2,1,"_CPPv2N6paddle13NeuralNetwork6createERK11ModelConfig"],"paddle::NeuralNetwork::dataLayers_":[40,3,1,"_CPPv2N6paddle13NeuralNetwork11dataLayers_E"],"paddle::NeuralNetwork::dllInitMap":[40,3,1,"_CPPv2N6paddle13NeuralNetwork10dllInitMapE"],"paddle::NeuralNetwork::eval":[40,2,1,"_CPPv2N6paddle13NeuralNetwork4evalEP9Evaluator"],"paddle::NeuralNetwork::forEachLayer":[40,2,1,"_CPPv2N6paddle13NeuralNetwork12forEachLayerE1T"],"paddle::NeuralNetwork::forward":[40,2,1,"_CPPv2N6paddle13NeuralNetwork7forwardERKNSt6vectorI8ArgumentEEPNSt6vectorI8ArgumentEE8PassType"],"paddle::NeuralNetwork::getLayer":[40,2,1,"_CPPv2N6paddle13NeuralNetwork8getLayerERKNSt6stringE"],"paddle::NeuralNetwork::getLayerOutput":[40,2,1,"_CPPv2N6paddle13NeuralNetwork14getLayerOutputERKNSt6stringE"],"paddle::NeuralNetwork::getParameterMap":[40,2,1,"_CPPv2N6paddle13NeuralNetwork15getParameterMapEv"],"paddle::NeuralNetwork::getState":[40,2,1,"_CPPv2N6paddle13NeuralNetwork8getStateER12MachineState"],"paddle::NeuralNetwork::layerMap_":[40,3,1,"_CPPv2N6paddle13NeuralNetwork9layerMap_E"],"paddle::NeuralNetwork::layers_":[40,3,1,"_CPPv2N6paddle13NeuralNetwork7layers_E"],"paddle::NeuralNetwork::makeEvaluator":[40,2,1,"_CPPv2N6paddle13NeuralNetwork13makeEvaluatorEv"],"paddle::NeuralNetwork::newNeuralNetwork":[40,2,1,"_CPPv2N6paddle13NeuralNetwork16newNeuralNetworkERKNSt6stringEP13NeuralNetwork"],"paddle::NeuralNetwork::onPassEnd":[40,2,1,"_CPPv2N6paddle13NeuralNetwork9onPassEndEv"],"paddle::NeuralNetwork::outputLayers_":[40,3,1,"_CPPv2N6paddle13NeuralNetwork13outputLayers_E"],"paddle::NeuralNetwork::paramSelfInited_":[40,3,1,"_CPPv2N6paddle13NeuralNetwork16paramSelfInited_E"],"paddle::NeuralNetwork::parameterMap_":[40,3,1,"_CPPv2N6paddle13NeuralNetwork13parameterMap_E"],"paddle::NeuralNetwork::prefetch":[40,2,1,"_CPPv2N6paddle13NeuralNetwork8prefetchERKNSt6vectorI8ArgumentEE"],"paddle::NeuralNetwork::resetState":[40,2,1,"_CPPv2N6paddle13NeuralNetwork10resetStateEv"],"paddle::NeuralNetwork::rootNetwork_":[40,3,1,"_CPPv2N6paddle13NeuralNetwork12rootNetwork_E"],"paddle::NeuralNetwork::setOutputGrad":[40,2,1,"_CPPv2N6paddle13NeuralNetwork13setOutputGradERKNSt6vectorI8ArgumentEE"],"paddle::NeuralNetwork::setState":[40,2,1,"_CPPv2N6paddle13NeuralNetwork8setStateERK12MachineState"],"paddle::NeuralNetwork::subModelName_":[40,3,1,"_CPPv2N6paddle13NeuralNetwork13subModelName_E"],"paddle::NoCacheStrategy":[36,1,1,"_CPPv2N6paddle15NoCacheStrategyE"],"paddle::NoCacheStrategy::drop":[36,2,1,"_CPPv2N6paddle15NoCacheStrategy4dropEPNSt5dequeI11PyObjectPtrEE"],"paddle::NoCacheStrategy::load":[36,2,1,"_CPPv2N6paddle15NoCacheStrategy4loadEv"],"paddle::NoCacheStrategy::reset":[36,2,1,"_CPPv2N6paddle15NoCacheStrategy5resetEv"],"paddle::NormLayer":[43,1,1,"_CPPv2N6paddle9NormLayerE"],"paddle::NormLayer::NormLayer":[43,2,1,"_CPPv2N6paddle9NormLayer9NormLayerERK11LayerConfig"],"paddle::NormLayer::create":[43,2,1,"_CPPv2N6paddle9NormLayer6createERK11LayerConfig"],"paddle::NormLayer::init":[43,2,1,"_CPPv2N6paddle9NormLayer4initERK8LayerMapRK12ParameterMap"],"paddle::Operator":[43,1,1,"_CPPv2N6paddle8OperatorE"],"paddle::Operator::Operator":[43,2,1,"_CPPv2N6paddle8Operator8OperatorERK14OperatorConfigb"],"paddle::Operator::backward":[43,2,1,"_CPPv2N6paddle8Operator8backwardEv"],"paddle::Operator::config_":[43,3,1,"_CPPv2N6paddle8Operator7config_E"],"paddle::Operator::create":[43,2,1,"_CPPv2N6paddle8Operator6createERK14OperatorConfigb"],"paddle::Operator::forward":[43,2,1,"_CPPv2N6paddle8Operator7forwardEv"],"paddle::Operator::getConfig":[43,2,1,"_CPPv2N6paddle8Operator9getConfigEv"],"paddle::Operator::getState":[43,2,1,"_CPPv2N6paddle8Operator8getStateEv"],"paddle::Operator::ins_":[43,3,1,"_CPPv2N6paddle8Operator4ins_E"],"paddle::Operator::out_":[43,3,1,"_CPPv2N6paddle8Operator4out_E"],"paddle::Operator::passType_":[43,3,1,"_CPPv2N6paddle8Operator9passType_E"],"paddle::Operator::prefetch":[43,2,1,"_CPPv2N6paddle8Operator8prefetchEPK8Argument"],"paddle::Operator::registrar_":[43,3,1,"_CPPv2N6paddle8Operator10registrar_E"],"paddle::Operator::resetState":[43,2,1,"_CPPv2N6paddle8Operator10resetStateEv"],"paddle::Operator::setState":[43,2,1,"_CPPv2N6paddle8Operator8setStateE13LayerStatePtr"],"paddle::Operator::useGpu_":[43,3,1,"_CPPv2N6paddle8Operator7useGpu_E"],"paddle::Operator::~Operator":[43,2,1,"_CPPv2N6paddle8OperatorD0Ev"],"paddle::OptimizerWithGradientClipping":[50,1,1,"_CPPv2N6paddle29OptimizerWithGradientClippingE"],"paddle::OptimizerWithGradientClipping::OptimizerWithGradientClipping":[50,2,1,"_CPPv2N6paddle29OptimizerWithGradientClipping29OptimizerWithGradientClippingERK18OptimizationConfigP18ParameterOptimizer"],"paddle::OptimizerWithGradientClipping::finishBatch":[50,2,1,"_CPPv2N6paddle29OptimizerWithGradientClipping11finishBatchEv"],"paddle::OptimizerWithGradientClipping::finishPass":[50,2,1,"_CPPv2N6paddle29OptimizerWithGradientClipping10finishPassEv"],"paddle::OptimizerWithGradientClipping::init":[50,2,1,"_CPPv2N6paddle29OptimizerWithGradientClipping4initE6size_tPK15ParameterConfig"],"paddle::OptimizerWithGradientClipping::needSpecialTraversal":[50,2,1,"_CPPv2N6paddle29OptimizerWithGradientClipping20needSpecialTraversalERK15ParameterConfig"],"paddle::OptimizerWithGradientClipping::optimizer_":[50,3,1,"_CPPv2N6paddle29OptimizerWithGradientClipping10optimizer_E"],"paddle::OptimizerWithGradientClipping::setNoDecay":[50,2,1,"_CPPv2N6paddle29OptimizerWithGradientClipping10setNoDecayEv"],"paddle::OptimizerWithGradientClipping::startBatch":[50,2,1,"_CPPv2N6paddle29OptimizerWithGradientClipping10startBatchE7int64_t"],"paddle::OptimizerWithGradientClipping::startPass":[50,2,1,"_CPPv2N6paddle29OptimizerWithGradientClipping9startPassEv"],"paddle::OptimizerWithGradientClipping::update":[50,2,1,"_CPPv2N6paddle29OptimizerWithGradientClipping6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::OptimizerWithRegularizer":[50,1,1,"_CPPv2N6paddle24OptimizerWithRegularizerE"],"paddle::OptimizerWithRegularizer::OptimizerWithRegularizer":[50,2,1,"_CPPv2N6paddle24OptimizerWithRegularizer24OptimizerWithRegularizerERK18OptimizationConfigP18ParameterOptimizerP11Regularizer"],"paddle::OptimizerWithRegularizer::create":[50,2,1,"_CPPv2N6paddle24OptimizerWithRegularizer6createERK18OptimizationConfigRK15ParameterConfigbb"],"paddle::OptimizerWithRegularizer::finishBatch":[50,2,1,"_CPPv2N6paddle24OptimizerWithRegularizer11finishBatchEv"],"paddle::OptimizerWithRegularizer::finishPass":[50,2,1,"_CPPv2N6paddle24OptimizerWithRegularizer10finishPassEv"],"paddle::OptimizerWithRegularizer::init":[50,2,1,"_CPPv2N6paddle24OptimizerWithRegularizer4initE6size_tPK15ParameterConfig"],"paddle::OptimizerWithRegularizer::needSpecialTraversal":[50,2,1,"_CPPv2N6paddle24OptimizerWithRegularizer20needSpecialTraversalERK15ParameterConfig"],"paddle::OptimizerWithRegularizer::optimizer_":[50,3,1,"_CPPv2N6paddle24OptimizerWithRegularizer10optimizer_E"],"paddle::OptimizerWithRegularizer::regularizer_":[50,3,1,"_CPPv2N6paddle24OptimizerWithRegularizer12regularizer_E"],"paddle::OptimizerWithRegularizer::startBatch":[50,2,1,"_CPPv2N6paddle24OptimizerWithRegularizer10startBatchE7int64_t"],"paddle::OptimizerWithRegularizer::startPass":[50,2,1,"_CPPv2N6paddle24OptimizerWithRegularizer9startPassEv"],"paddle::OptimizerWithRegularizer::timer_":[50,3,1,"_CPPv2N6paddle24OptimizerWithRegularizer6timer_E"],"paddle::OptimizerWithRegularizer::update":[50,2,1,"_CPPv2N6paddle24OptimizerWithRegularizer6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::OptimizerWithRegularizerEveryNumBatches":[50,1,1,"_CPPv2N6paddle39OptimizerWithRegularizerEveryNumBatchesE"],"paddle::OptimizerWithRegularizerEveryNumBatches::OptimizerWithRegularizerEveryNumBatches":[50,2,1,"_CPPv2N6paddle39OptimizerWithRegularizerEveryNumBatches39OptimizerWithRegularizerEveryNumBatchesERK18OptimizationConfigP18ParameterOptimizerP11Regularizer"],"paddle::OptimizerWithRegularizerEveryNumBatches::baseTimer_":[50,3,1,"_CPPv2N6paddle39OptimizerWithRegularizerEveryNumBatches10baseTimer_E"],"paddle::OptimizerWithRegularizerEveryNumBatches::catchUpWith":[50,2,1,"_CPPv2N6paddle39OptimizerWithRegularizerEveryNumBatches11catchUpWithEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::OptimizerWithRegularizerEveryNumBatches::doTraversal":[50,2,1,"_CPPv2N6paddle39OptimizerWithRegularizerEveryNumBatches11doTraversalEA_K9VectorPtrRK15ParameterConfig"],"paddle::OptimizerWithRegularizerEveryNumBatches::finishCatchUpWith":[50,2,1,"_CPPv2N6paddle39OptimizerWithRegularizerEveryNumBatches17finishCatchUpWithEv"],"paddle::OptimizerWithRegularizerEveryNumBatches::isRegularizationBatch":[50,2,1,"_CPPv2N6paddle39OptimizerWithRegularizerEveryNumBatches21isRegularizationBatchERK15ParameterConfig"],"paddle::OptimizerWithRegularizerEveryNumBatches::needSpecialTraversal":[50,2,1,"_CPPv2N6paddle39OptimizerWithRegularizerEveryNumBatches20needSpecialTraversalERK15ParameterConfig"],"paddle::OptimizerWithRegularizerEveryNumBatches::startCatchUpWith":[50,2,1,"_CPPv2N6paddle39OptimizerWithRegularizerEveryNumBatches16startCatchUpWithEv"],"paddle::OptimizerWithRegularizerEveryNumBatches::startPass":[50,2,1,"_CPPv2N6paddle39OptimizerWithRegularizerEveryNumBatches9startPassEv"],"paddle::OptimizerWithRegularizerEveryNumBatches::update":[50,2,1,"_CPPv2N6paddle39OptimizerWithRegularizerEveryNumBatches6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::OptimizerWithRegularizerSparse":[50,1,1,"_CPPv2N6paddle30OptimizerWithRegularizerSparseE"],"paddle::OptimizerWithRegularizerSparse::OptimizerWithRegularizerSparse":[50,2,1,"_CPPv2N6paddle30OptimizerWithRegularizerSparse30OptimizerWithRegularizerSparseERK18OptimizationConfigP18ParameterOptimizerP11Regularizer"],"paddle::OptimizerWithRegularizerSparse::catchUpWith":[50,2,1,"_CPPv2N6paddle30OptimizerWithRegularizerSparse11catchUpWithEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::OptimizerWithRegularizerSparse::finishCatchUpWith":[50,2,1,"_CPPv2N6paddle30OptimizerWithRegularizerSparse17finishCatchUpWithEv"],"paddle::OptimizerWithRegularizerSparse::init":[50,2,1,"_CPPv2N6paddle30OptimizerWithRegularizerSparse4initE6size_tPK15ParameterConfig"],"paddle::OptimizerWithRegularizerSparse::startCatchUpWith":[50,2,1,"_CPPv2N6paddle30OptimizerWithRegularizerSparse16startCatchUpWithEv"],"paddle::OptimizerWithRegularizerSparse::t0Vec_":[50,3,1,"_CPPv2N6paddle30OptimizerWithRegularizerSparse6t0Vec_E"],"paddle::OptimizerWithRegularizerSparse::update":[50,2,1,"_CPPv2N6paddle30OptimizerWithRegularizerSparse6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::OuterProdLayer":[43,1,1,"_CPPv2N6paddle14OuterProdLayerE"],"paddle::OuterProdLayer::OuterProdLayer":[43,2,1,"_CPPv2N6paddle14OuterProdLayer14OuterProdLayerERK11LayerConfig"],"paddle::OuterProdLayer::backward":[43,2,1,"_CPPv2N6paddle14OuterProdLayer8backwardERK14UpdateCallback"],"paddle::OuterProdLayer::forward":[43,2,1,"_CPPv2N6paddle14OuterProdLayer7forwardE8PassType"],"paddle::OuterProdLayer::init":[43,2,1,"_CPPv2N6paddle14OuterProdLayer4initERK8LayerMapRK12ParameterMap"],"paddle::OuterProdLayer::tmpMtx0":[43,3,1,"_CPPv2N6paddle14OuterProdLayer7tmpMtx0E"],"paddle::OuterProdLayer::tmpRow0":[43,3,1,"_CPPv2N6paddle14OuterProdLayer7tmpRow0E"],"paddle::OuterProdLayer::tmpRow1":[43,3,1,"_CPPv2N6paddle14OuterProdLayer7tmpRow1E"],"paddle::OuterProdLayer::~OuterProdLayer":[43,2,1,"_CPPv2N6paddle14OuterProdLayerD0Ev"],"paddle::ParallelCpuVectorT":[46,1,1,"_CPPv2N6paddle18ParallelCpuVectorTE"],"paddle::ParallelCpuVectorT::ExecFunc":[46,4,1,"_CPPv2N6paddle18ParallelCpuVectorT8ExecFuncE"],"paddle::ParallelCpuVectorT::ParallelCpuVectorT":[46,2,1,"_CPPv2N6paddle18ParallelCpuVectorT18ParallelCpuVectorTE6size_tP14SyncThreadPool"],"paddle::ParallelCpuVectorT::exec":[46,2,1,"_CPPv2N6paddle18ParallelCpuVectorT4execEN14SyncThreadPool7JobFuncE"],"paddle::ParallelCpuVectorT::parallelExec":[46,2,1,"_CPPv2N6paddle18ParallelCpuVectorT12parallelExecE8ExecFunc"],"paddle::ParallelCpuVectorT::pool_":[46,3,1,"_CPPv2N6paddle18ParallelCpuVectorT5pool_E"],"paddle::ParallelCpuVectorT::randnorm":[46,2,1,"_CPPv2N6paddle18ParallelCpuVectorT8randnormE4real4real"],"paddle::ParallelCpuVectorT::uniform":[46,2,1,"_CPPv2N6paddle18ParallelCpuVectorT7uniformE4real4real"],"paddle::ParallelCpuVectorT::zeroMem":[46,2,1,"_CPPv2N6paddle18ParallelCpuVectorT7zeroMemEv"],"paddle::ParallelNeuralNetwork":[40,1,1,"_CPPv2N6paddle21ParallelNeuralNetworkE"],"paddle::ParallelNeuralNetwork::ParallelNeuralNetwork":[40,2,1,"_CPPv2N6paddle21ParallelNeuralNetwork21ParallelNeuralNetworkENSt6stringEP13NeuralNetwork"],"paddle::ParallelNeuralNetwork::addComputeThread":[40,2,1,"_CPPv2N6paddle21ParallelNeuralNetwork16addComputeThreadEi"],"paddle::ParallelNeuralNetwork::backward":[40,2,1,"_CPPv2N6paddle21ParallelNeuralNetwork8backwardERK14UpdateCallback"],"paddle::ParallelNeuralNetwork::dispatchByDeviceId":[40,2,1,"_CPPv2N6paddle21ParallelNeuralNetwork18dispatchByDeviceIdEi8LayerPtr8TaskType"],"paddle::ParallelNeuralNetwork::forward":[40,2,1,"_CPPv2N6paddle21ParallelNeuralNetwork7forwardERKNSt6vectorI8ArgumentEEPNSt6vectorI8ArgumentEE8PassType"],"paddle::ParallelNeuralNetwork::forwardBackward":[40,2,1,"_CPPv2N6paddle21ParallelNeuralNetwork15forwardBackwardERKNSt6vectorI8ArgumentEEPNSt6vectorI8ArgumentEE8PassTypeRK14UpdateCallback"],"paddle::ParallelNeuralNetwork::numDevices_":[40,3,1,"_CPPv2N6paddle21ParallelNeuralNetwork11numDevices_E"],"paddle::ParallelNeuralNetwork::start":[40,2,1,"_CPPv2N6paddle21ParallelNeuralNetwork5startERK13TrainerConfig15DataProviderPtr"],"paddle::ParallelNeuralNetwork::threads_":[40,3,1,"_CPPv2N6paddle21ParallelNeuralNetwork8threads_E"],"paddle::ParallelNeuralNetwork::useGpu_":[40,3,1,"_CPPv2N6paddle21ParallelNeuralNetwork7useGpu_E"],"paddle::ParallelNeuralNetwork::waitAllThread":[40,2,1,"_CPPv2N6paddle21ParallelNeuralNetwork13waitAllThreadEv"],"paddle::ParallelParameter":[52,1,1,"_CPPv2N6paddle17ParallelParameterE"],"paddle::ParallelParameter::ParallelParameter":[52,2,1,"_CPPv2N6paddle17ParallelParameter17ParallelParameterE11TrainerRole12ParameterPtr"],"paddle::ParallelParameter::controlUpdate":[52,2,1,"_CPPv2N6paddle17ParallelParameter13controlUpdateERK14UpdateCallback"],"paddle::ParallelParameter::create":[52,2,1,"_CPPv2N6paddle17ParallelParameter6createE11TrainerRole12ParameterPtri"],"paddle::ParallelParameter::getLocalParameter":[52,2,1,"_CPPv2N6paddle17ParallelParameter17getLocalParameterEv"],"paddle::ParallelParameter::gradSem_":[52,3,1,"_CPPv2N6paddle17ParallelParameter8gradSem_E"],"paddle::ParallelParameter::localParam_":[52,3,1,"_CPPv2N6paddle17ParallelParameter11localParam_E"],"paddle::ParallelParameter::majorUpdate":[52,2,1,"_CPPv2N6paddle17ParallelParameter11majorUpdateE4real"],"paddle::ParallelParameter::minorUpdate":[52,2,1,"_CPPv2N6paddle17ParallelParameter11minorUpdateE4real"],"paddle::ParallelParameter::postValueReady":[52,2,1,"_CPPv2N6paddle17ParallelParameter14postValueReadyEv"],"paddle::ParallelParameter::role_":[52,3,1,"_CPPv2N6paddle17ParallelParameter5role_E"],"paddle::ParallelParameter::singleUpdate":[52,2,1,"_CPPv2N6paddle17ParallelParameter12singleUpdateE4real"],"paddle::ParallelParameter::slaveUpdate":[52,2,1,"_CPPv2N6paddle17ParallelParameter11slaveUpdateE4real"],"paddle::ParallelParameter::syncUpdate":[52,2,1,"_CPPv2N6paddle17ParallelParameter10syncUpdateE11TrainerRole4real"],"paddle::ParallelParameter::synchronizeParamter":[52,2,1,"_CPPv2N6paddle17ParallelParameter19synchronizeParamterEv"],"paddle::ParallelParameter::timeWaitGradReady":[52,2,1,"_CPPv2N6paddle17ParallelParameter17timeWaitGradReadyEi"],"paddle::ParallelParameter::valueSem_":[52,3,1,"_CPPv2N6paddle17ParallelParameter9valueSem_E"],"paddle::ParallelParameter::waitGradReady":[52,2,1,"_CPPv2N6paddle17ParallelParameter13waitGradReadyEv"],"paddle::ParallelParameter::~ParallelParameter":[52,2,1,"_CPPv2N6paddle17ParallelParameterD0Ev"],"paddle::ParallelParameterMap":[52,4,1,"_CPPv2N6paddle20ParallelParameterMapE"],"paddle::ParallelParameterPtr":[52,4,1,"_CPPv2N6paddle20ParallelParameterPtrE"],"paddle::ParamInitCallback":[52,4,1,"_CPPv2N6paddle17ParamInitCallbackE"],"paddle::Parameter":[52,1,1,"_CPPv2N6paddle9ParameterE"],"paddle::Parameter::ExecFunc":[52,4,1,"_CPPv2N6paddle9Parameter8ExecFuncE"],"paddle::Parameter::Header":[52,1,1,"_CPPv2N6paddle9Parameter6HeaderE"],"paddle::Parameter::Header::size":[52,3,1,"_CPPv2N6paddle9Parameter6Header4sizeE"],"paddle::Parameter::Header::valueSize":[52,3,1,"_CPPv2N6paddle9Parameter6Header9valueSizeE"],"paddle::Parameter::Header::version":[52,3,1,"_CPPv2N6paddle9Parameter6Header7versionE"],"paddle::Parameter::bufs_":[52,3,1,"_CPPv2N6paddle9Parameter5bufs_E"],"paddle::Parameter::clearGradient":[52,2,1,"_CPPv2N6paddle9Parameter13clearGradientEv"],"paddle::Parameter::clearUpdate":[52,2,1,"_CPPv2N6paddle9Parameter11clearUpdateEv"],"paddle::Parameter::clearValueUpdated":[52,2,1,"_CPPv2N6paddle9Parameter17clearValueUpdatedEv"],"paddle::Parameter::config_":[52,3,1,"_CPPv2N6paddle9Parameter7config_E"],"paddle::Parameter::deviceId_":[52,3,1,"_CPPv2N6paddle9Parameter9deviceId_E"],"paddle::Parameter::enableIntType":[52,2,1,"_CPPv2N6paddle9Parameter13enableIntTypeE13ParameterType6size_t"],"paddle::Parameter::enableSharedType":[52,2,1,"_CPPv2N6paddle9Parameter16enableSharedTypeE13ParameterType9VectorPtr9MatrixPtr"],"paddle::Parameter::enableSparseParameter":[52,2,1,"_CPPv2N6paddle9Parameter21enableSparseParameterEv"],"paddle::Parameter::enableType":[52,2,1,"_CPPv2N6paddle9Parameter10enableTypeE13ParameterType7MatType"],"paddle::Parameter::format_":[52,3,1,"_CPPv2N6paddle9Parameter7format_E"],"paddle::Parameter::getBuf":[52,2,1,"_CPPv2N6paddle9Parameter6getBufE13ParameterType"],"paddle::Parameter::getBufs":[52,2,1,"_CPPv2N6paddle9Parameter7getBufsEv"],"paddle::Parameter::getConfig":[52,2,1,"_CPPv2N6paddle9Parameter9getConfigEv"],"paddle::Parameter::getDeviceId":[52,2,1,"_CPPv2N6paddle9Parameter11getDeviceIdEv"],"paddle::Parameter::getFormat":[52,2,1,"_CPPv2N6paddle9Parameter9getFormatEv"],"paddle::Parameter::getGradientSegments":[52,2,1,"_CPPv2N6paddle9Parameter19getGradientSegmentsEv"],"paddle::Parameter::getID":[52,2,1,"_CPPv2N6paddle9Parameter5getIDEv"],"paddle::Parameter::getInitMean":[52,2,1,"_CPPv2N6paddle9Parameter11getInitMeanEv"],"paddle::Parameter::getInitStandardDeviation":[52,2,1,"_CPPv2N6paddle9Parameter24getInitStandardDeviationEv"],"paddle::Parameter::getIntBuf":[52,2,1,"_CPPv2N6paddle9Parameter9getIntBufE13ParameterType"],"paddle::Parameter::getLearnRate":[52,2,1,"_CPPv2N6paddle9Parameter12getLearnRateEv"],"paddle::Parameter::getMat":[52,2,1,"_CPPv2N6paddle9Parameter6getMatE13ParameterType"],"paddle::Parameter::getName":[52,2,1,"_CPPv2N6paddle9Parameter7getNameEv"],"paddle::Parameter::getSharedCount":[52,2,1,"_CPPv2N6paddle9Parameter14getSharedCountEv"],"paddle::Parameter::getSize":[52,2,1,"_CPPv2N6paddle9Parameter7getSizeEv"],"paddle::Parameter::gradSegments_":[52,3,1,"_CPPv2N6paddle9Parameter13gradSegments_E"],"paddle::Parameter::hasType":[52,2,1,"_CPPv2N6paddle9Parameter7hasTypeE13ParameterType"],"paddle::Parameter::incShared":[52,2,1,"_CPPv2N6paddle9Parameter9incSharedEv"],"paddle::Parameter::initHook":[52,2,1,"_CPPv2N6paddle9Parameter8initHookEv"],"paddle::Parameter::intBufs_":[52,3,1,"_CPPv2N6paddle9Parameter8intBufs_E"],"paddle::Parameter::isFullSize":[52,2,1,"_CPPv2N6paddle9Parameter10isFullSizeEv"],"paddle::Parameter::isSparse":[52,2,1,"_CPPv2N6paddle9Parameter8isSparseEv"],"paddle::Parameter::isSparseRemoteUpdate":[52,2,1,"_CPPv2N6paddle9Parameter20isSparseRemoteUpdateEv"],"paddle::Parameter::isStatic":[52,2,1,"_CPPv2N6paddle9Parameter8isStaticEv"],"paddle::Parameter::isUpdatable":[52,2,1,"_CPPv2N6paddle9Parameter11isUpdatableEv"],"paddle::Parameter::isValueUpdated":[52,2,1,"_CPPv2N6paddle9Parameter14isValueUpdatedEv"],"paddle::Parameter::kFormatVersion":[52,3,1,"_CPPv2N6paddle9Parameter14kFormatVersionE"],"paddle::Parameter::mats_":[52,3,1,"_CPPv2N6paddle9Parameter5mats_E"],"paddle::Parameter::setDevice":[52,2,1,"_CPPv2N6paddle9Parameter9setDeviceEi"],"paddle::Parameter::setID":[52,2,1,"_CPPv2N6paddle9Parameter5setIDE6size_t"],"paddle::Parameter::setIntBuf":[52,2,1,"_CPPv2N6paddle9Parameter9setIntBufE13ParameterTypeRK10IVectorPtr"],"paddle::Parameter::setSharedCount":[52,2,1,"_CPPv2N6paddle9Parameter14setSharedCountEi"],"paddle::Parameter::setValueUpdated":[52,2,1,"_CPPv2N6paddle9Parameter15setValueUpdatedEv"],"paddle::Parameter::sharedCount_":[52,3,1,"_CPPv2N6paddle9Parameter12sharedCount_E"],"paddle::Parameter::singleUpdate":[52,2,1,"_CPPv2N6paddle9Parameter12singleUpdateEPv"],"paddle::Parameter::updateCounter_":[52,3,1,"_CPPv2N6paddle9Parameter14updateCounter_E"],"paddle::Parameter::updateHook":[52,2,1,"_CPPv2N6paddle9Parameter10updateHookEv"],"paddle::Parameter::updated_":[52,3,1,"_CPPv2N6paddle9Parameter8updated_E"],"paddle::Parameter::updaterHooks_":[52,3,1,"_CPPv2N6paddle9Parameter13updaterHooks_E"],"paddle::Parameter::useGpu":[52,2,1,"_CPPv2N6paddle9Parameter6useGpuEv"],"paddle::Parameter::useGpu_":[52,3,1,"_CPPv2N6paddle9Parameter7useGpu_E"],"paddle::ParameterClient2":[55,1,1,"_CPPv2N6paddle16ParameterClient2E"],"paddle::ParameterClient2::ParameterClient2":[55,2,1,"_CPPv2N6paddle16ParameterClient216ParameterClient2Ebii"],"paddle::ParameterClient2::allSegments_":[55,3,1,"_CPPv2N6paddle16ParameterClient212allSegments_E"],"paddle::ParameterClient2::asyncFinishPass":[55,2,1,"_CPPv2N6paddle16ParameterClient215asyncFinishPassE10SyncObject"],"paddle::ParameterClient2::asyncStartPass":[55,2,1,"_CPPv2N6paddle16ParameterClient214asyncStartPassE10SyncObject"],"paddle::ParameterClient2::calcParameterBlockSize":[55,2,1,"_CPPv2N6paddle16ParameterClient222calcParameterBlockSizeERKNSt6vectorI12ParameterPtrEE6size_t"],"paddle::ParameterClient2::createMatrix":[55,2,1,"_CPPv2N6paddle16ParameterClient212createMatrixE7int32_t"],"paddle::ParameterClient2::createVector":[55,2,1,"_CPPv2N6paddle16ParameterClient212createVectorEv"],"paddle::ParameterClient2::destroy":[55,2,1,"_CPPv2N6paddle16ParameterClient27destroyEv"],"paddle::ParameterClient2::doOperation":[55,2,1,"_CPPv2N6paddle16ParameterClient211doOperationER18PreparedOperationsbbb"],"paddle::ParameterClient2::forwardbackwordTime_":[55,3,1,"_CPPv2N6paddle16ParameterClient220forwardbackwordTime_E"],"paddle::ParameterClient2::getPServerParameterGradient":[55,2,1,"_CPPv2N6paddle16ParameterClient227getPServerParameterGradientEv"],"paddle::ParameterClient2::getPServerParameterValue":[55,2,1,"_CPPv2N6paddle16ParameterClient224getPServerParameterValueEv"],"paddle::ParameterClient2::getParameter":[55,2,1,"_CPPv2N6paddle16ParameterClient212getParameterE13ParameterType13ParameterType"],"paddle::ParameterClient2::getParameterSparse":[55,2,1,"_CPPv2N6paddle16ParameterClient218getParameterSparseE13ParameterType13ParameterType"],"paddle::ParameterClient2::inStatus":[55,2,1,"_CPPv2N6paddle16ParameterClient28inStatusE13PServerStatus"],"paddle::ParameterClient2::init":[55,2,1,"_CPPv2N6paddle16ParameterClient24initERKNSt6vectorI12ParameterPtrEE"],"paddle::ParameterClient2::initThreads":[55,2,1,"_CPPv2N6paddle16ParameterClient211initThreadsEv"],"paddle::ParameterClient2::isPassFinish":[55,2,1,"_CPPv2N6paddle16ParameterClient212isPassFinishEv"],"paddle::ParameterClient2::loadValueVector":[55,2,1,"_CPPv2N6paddle16ParameterClient215loadValueVectorERKNSt6stringE"],"paddle::ParameterClient2::multiCall":[55,2,1,"_CPPv2N6paddle16ParameterClient29multiCallEPKcRK7ProtoInPNSt6vectorI8ProtoOutEE"],"paddle::ParameterClient2::parameterMap_":[55,3,1,"_CPPv2N6paddle16ParameterClient213parameterMap_E"],"paddle::ParameterClient2::passFinish_":[55,3,1,"_CPPv2N6paddle16ParameterClient211passFinish_E"],"paddle::ParameterClient2::port_":[55,3,1,"_CPPv2N6paddle16ParameterClient25port_E"],"paddle::ParameterClient2::prepareSendData":[55,2,1,"_CPPv2N6paddle16ParameterClient215prepareSendDataE19ParameterUpdateMode13ParameterTypeRKNSt6vectorI17ParameterSegmentsEE7int64_t4realb13ParameterType11BatchStatusP7SendJob"],"paddle::ParameterClient2::recv":[55,2,1,"_CPPv2N6paddle16ParameterClient24recvEi"],"paddle::ParameterClient2::recvParameter":[55,2,1,"_CPPv2N6paddle16ParameterClient213recvParameterEv"],"paddle::ParameterClient2::releaseMatrix":[55,2,1,"_CPPv2N6paddle16ParameterClient213releaseMatrixE13PServerMatrix"],"paddle::ParameterClient2::releaseVector":[55,2,1,"_CPPv2N6paddle16ParameterClient213releaseVectorE13PServerVector"],"paddle::ParameterClient2::saveValueVector":[55,2,1,"_CPPv2N6paddle16ParameterClient215saveValueVectorERKNSt6stringE"],"paddle::ParameterClient2::send":[55,2,1,"_CPPv2N6paddle16ParameterClient24sendEi"],"paddle::ParameterClient2::sendAndReceiveParameter":[55,2,1,"_CPPv2N6paddle16ParameterClient223sendAndReceiveParameterE19ParameterUpdateMode13ParameterTypeRKNSt6vectorI17ParameterSegmentsEE7int64_t4realb13ParameterType13ParameterType"],"paddle::ParameterClient2::sendParallel":[55,2,1,"_CPPv2N6paddle16ParameterClient212sendParallelEi6size_t13ParameterType"],"paddle::ParameterClient2::sendParameter":[55,2,1,"_CPPv2N6paddle16ParameterClient213sendParameterE19ParameterUpdateMode13ParameterTypeRKNSt6vectorI17ParameterSegmentsEE7int64_t4realb11BatchStatus"],"paddle::ParameterClient2::setConfig":[55,2,1,"_CPPv2N6paddle16ParameterClient29setConfigERK18OptimizationConfigRKNSt6stringEb"],"paddle::ParameterClient2::setForwardbackwardTime":[55,2,1,"_CPPv2N6paddle16ParameterClient222setForwardbackwardTimeE8uint64_t"],"paddle::ParameterClient2::setParameter":[55,2,1,"_CPPv2N6paddle16ParameterClient212setParameterEv"],"paddle::ParameterClient2::setParameterZero":[55,2,1,"_CPPv2N6paddle16ParameterClient216setParameterZeroEv"],"paddle::ParameterClient2::setStatus":[55,2,1,"_CPPv2N6paddle16ParameterClient29setStatusE13PServerStatus"],"paddle::ParameterClient2::setTrainerId":[55,2,1,"_CPPv2N6paddle16ParameterClient212setTrainerIdEi"],"paddle::ParameterClient2::sparseDistribution_":[55,3,1,"_CPPv2N6paddle16ParameterClient219sparseDistribution_E"],"paddle::ParameterClient2::syncThreadPool_":[55,3,1,"_CPPv2N6paddle16ParameterClient215syncThreadPool_E"],"paddle::ParameterClient2::synchronize":[55,2,1,"_CPPv2N6paddle16ParameterClient211synchronizeE10SyncObject"],"paddle::ParameterClient2::trainerId_":[55,3,1,"_CPPv2N6paddle16ParameterClient210trainerId_E"],"paddle::ParameterClient2::vectorAddMult":[55,2,1,"_CPPv2N6paddle16ParameterClient213vectorAddMultE13PServerVector13PServerVector4real"],"paddle::ParameterClient2::vectorAddMultInto":[55,2,1,"_CPPv2N6paddle16ParameterClient217vectorAddMultIntoE13PServerVector13PServerVector13PServerVector4real"],"paddle::ParameterClient2::vectorCopy":[55,2,1,"_CPPv2N6paddle16ParameterClient210vectorCopyE13PServerVector13PServerVector"],"paddle::ParameterClient2::vectorDotProduct":[55,2,1,"_CPPv2N6paddle16ParameterClient216vectorDotProductE13PServerVector13PServerVector"],"paddle::ParameterClient2::vectorScale":[55,2,1,"_CPPv2N6paddle16ParameterClient211vectorScaleE13PServerVector4real"],"paddle::ParameterClient2::vectorScaleInto":[55,2,1,"_CPPv2N6paddle16ParameterClient215vectorScaleIntoE13PServerVector13PServerVector4real"],"paddle::ParameterClient2::waitForStatus":[55,2,1,"_CPPv2N6paddle16ParameterClient213waitForStatusE13PServerStatus"],"paddle::ParameterClient2::waitPassFinish":[55,2,1,"_CPPv2N6paddle16ParameterClient214waitPassFinishEv"],"paddle::ParameterClient2::waitPassStart":[55,2,1,"_CPPv2N6paddle16ParameterClient213waitPassStartEv"],"paddle::ParameterClient2::~ParameterClient2":[55,2,1,"_CPPv2N6paddle16ParameterClient2D0Ev"],"paddle::ParameterMap":[52,4,1,"_CPPv2N6paddle12ParameterMapE"],"paddle::ParameterOptimizer":[50,1,1,"_CPPv2N6paddle18ParameterOptimizerE"],"paddle::ParameterOptimizer::ParameterOptimizer":[50,2,1,"_CPPv2N6paddle18ParameterOptimizer18ParameterOptimizerERK18OptimizationConfig"],"paddle::ParameterOptimizer::TraverseCallback":[50,4,1,"_CPPv2N6paddle18ParameterOptimizer16TraverseCallbackE"],"paddle::ParameterOptimizer::TraverseCallbackVec":[50,4,1,"_CPPv2N6paddle18ParameterOptimizer19TraverseCallbackVecE"],"paddle::ParameterOptimizer::addParameterType":[50,2,1,"_CPPv2N6paddle18ParameterOptimizer16addParameterTypeE13ParameterType"],"paddle::ParameterOptimizer::apply":[50,2,1,"_CPPv2N6paddle18ParameterOptimizer5applyEv"],"paddle::ParameterOptimizer::applyDecay_":[50,3,1,"_CPPv2N6paddle18ParameterOptimizer11applyDecay_E"],"paddle::ParameterOptimizer::calcLearningRate":[50,2,1,"_CPPv2N6paddle18ParameterOptimizer16calcLearningRateE7int64_t7int64_t"],"paddle::ParameterOptimizer::composeCallbacks":[50,2,1,"_CPPv2N6paddle18ParameterOptimizer16composeCallbacksERK19TraverseCallbackVec"],"paddle::ParameterOptimizer::finishBatch":[50,2,1,"_CPPv2N6paddle18ParameterOptimizer11finishBatchEv"],"paddle::ParameterOptimizer::finishCatchUpWith":[50,2,1,"_CPPv2N6paddle18ParameterOptimizer17finishCatchUpWithEv"],"paddle::ParameterOptimizer::finishPass":[50,2,1,"_CPPv2N6paddle18ParameterOptimizer10finishPassEv"],"paddle::ParameterOptimizer::firstTime_":[50,3,1,"_CPPv2N6paddle18ParameterOptimizer10firstTime_E"],"paddle::ParameterOptimizer::getLearningRate":[50,2,1,"_CPPv2N6paddle18ParameterOptimizer15getLearningRateEv"],"paddle::ParameterOptimizer::getParameterTypes":[50,2,1,"_CPPv2N6paddle18ParameterOptimizer17getParameterTypesEv"],"paddle::ParameterOptimizer::init":[50,2,1,"_CPPv2N6paddle18ParameterOptimizer4initE6size_tPK15ParameterConfig"],"paddle::ParameterOptimizer::learningRateScheduler_":[50,3,1,"_CPPv2N6paddle18ParameterOptimizer22learningRateScheduler_E"],"paddle::ParameterOptimizer::learningRate_":[50,3,1,"_CPPv2N6paddle18ParameterOptimizer13learningRate_E"],"paddle::ParameterOptimizer::needSpecialTraversal":[50,2,1,"_CPPv2N6paddle18ParameterOptimizer20needSpecialTraversalERK15ParameterConfig"],"paddle::ParameterOptimizer::optConfig_":[50,3,1,"_CPPv2N6paddle18ParameterOptimizer10optConfig_E"],"paddle::ParameterOptimizer::parameterTypes_":[50,3,1,"_CPPv2N6paddle18ParameterOptimizer15parameterTypes_E"],"paddle::ParameterOptimizer::pass_":[50,3,1,"_CPPv2N6paddle18ParameterOptimizer5pass_E"],"paddle::ParameterOptimizer::restore":[50,2,1,"_CPPv2N6paddle18ParameterOptimizer7restoreEv"],"paddle::ParameterOptimizer::setNoDecay":[50,2,1,"_CPPv2N6paddle18ParameterOptimizer10setNoDecayEv"],"paddle::ParameterOptimizer::startBatch":[50,2,1,"_CPPv2N6paddle18ParameterOptimizer10startBatchE7int64_t"],"paddle::ParameterOptimizer::startCatchUpWith":[50,2,1,"_CPPv2N6paddle18ParameterOptimizer16startCatchUpWithEv"],"paddle::ParameterOptimizer::startPass":[50,2,1,"_CPPv2N6paddle18ParameterOptimizer9startPassEv"],"paddle::ParameterOptimizer::update":[50,2,1,"_CPPv2N6paddle18ParameterOptimizer6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::ParameterOptimizer::~ParameterOptimizer":[50,2,1,"_CPPv2N6paddle18ParameterOptimizerD0Ev"],"paddle::ParameterPtr":[52,4,1,"_CPPv2N6paddle12ParameterPtrE"],"paddle::ParameterReluLayer":[43,1,1,"_CPPv2N6paddle18ParameterReluLayerE"],"paddle::ParameterReluLayer::ParameterReluLayer":[43,2,1,"_CPPv2N6paddle18ParameterReluLayer18ParameterReluLayerERK11LayerConfig"],"paddle::ParameterReluLayer::backward":[43,2,1,"_CPPv2N6paddle18ParameterReluLayer8backwardERK14UpdateCallback"],"paddle::ParameterReluLayer::forward":[43,2,1,"_CPPv2N6paddle18ParameterReluLayer7forwardE8PassType"],"paddle::ParameterReluLayer::init":[43,2,1,"_CPPv2N6paddle18ParameterReluLayer4initERK8LayerMapRK12ParameterMap"],"paddle::ParameterReluLayer::partialSum_":[43,3,1,"_CPPv2N6paddle18ParameterReluLayer11partialSum_E"],"paddle::ParameterReluLayer::weight_":[43,3,1,"_CPPv2N6paddle18ParameterReluLayer7weight_E"],"paddle::ParameterReluLayer::~ParameterReluLayer":[43,2,1,"_CPPv2N6paddle18ParameterReluLayerD0Ev"],"paddle::ParameterServer2":[60,1,1,"_CPPv2N6paddle16ParameterServer2E"],"paddle::ParameterServer2::BlockInfo":[60,1,1,"_CPPv2N6paddle16ParameterServer29BlockInfoE"],"paddle::ParameterServer2::BlockInfo::config":[60,3,1,"_CPPv2N6paddle16ParameterServer29BlockInfo6configE"],"paddle::ParameterServer2::BlockInfo::lock":[60,3,1,"_CPPv2N6paddle16ParameterServer29BlockInfo4lockE"],"paddle::ParameterServer2::BlockInfo::offset":[60,3,1,"_CPPv2N6paddle16ParameterServer29BlockInfo6offsetE"],"paddle::ParameterServer2::BlockInfo::optimizer":[60,3,1,"_CPPv2N6paddle16ParameterServer29BlockInfo9optimizerE"],"paddle::ParameterServer2::BlockKey":[60,4,1,"_CPPv2N6paddle16ParameterServer28BlockKeyE"],"paddle::ParameterServer2::BlockKeyHash":[60,1,1,"_CPPv2N6paddle16ParameterServer212BlockKeyHashE"],"paddle::ParameterServer2::BlockKeyHash::operator()":[60,2,1,"_CPPv2N6paddle16ParameterServer212BlockKeyHashclERK8BlockKey"],"paddle::ParameterServer2::BlockMap":[60,4,1,"_CPPv2N6paddle16ParameterServer28BlockMapE"],"paddle::ParameterServer2::BlockSegments":[60,4,1,"_CPPv2N6paddle16ParameterServer213BlockSegmentsE"],"paddle::ParameterServer2::Buffer":[60,1,1,"_CPPv2N6paddle16ParameterServer26BufferE"],"paddle::ParameterServer2::Buffer::base":[60,3,1,"_CPPv2N6paddle16ParameterServer26Buffer4baseE"],"paddle::ParameterServer2::Buffer::size":[60,3,1,"_CPPv2N6paddle16ParameterServer26Buffer4sizeE"],"paddle::ParameterServer2::ExecFunc":[60,4,1,"_CPPv2N6paddle16ParameterServer28ExecFuncE"],"paddle::ParameterServer2::OperatorFunction":[60,4,1,"_CPPv2N6paddle16ParameterServer216OperatorFunctionE"],"paddle::ParameterServer2::ParameterServer2":[60,2,1,"_CPPv2N6paddle16ParameterServer216ParameterServer2ERKNSt6stringEii"],"paddle::ParameterServer2::ReadWriteBuffer":[60,1,1,"_CPPv2N6paddle16ParameterServer215ReadWriteBufferE"],"paddle::ParameterServer2::ReadWriteBuffer::AlignElementCount":[60,3,1,"_CPPv2N6paddle16ParameterServer215ReadWriteBuffer17AlignElementCountE"],"paddle::ParameterServer2::ReadWriteBuffer::IsTLargerThanAlign":[60,3,1,"_CPPv2N6paddle16ParameterServer215ReadWriteBuffer18IsTLargerThanAlignE"],"paddle::ParameterServer2::ReadWriteBuffer::curOffset_":[60,3,1,"_CPPv2N6paddle16ParameterServer215ReadWriteBuffer10curOffset_E"],"paddle::ParameterServer2::ReadWriteBuffer::nextBlock":[60,2,1,"_CPPv2N6paddle16ParameterServer215ReadWriteBuffer9nextBlockE6size_t"],"paddle::ParameterServer2::ReadWriteBuffer::resetAlignAlloc":[60,2,1,"_CPPv2N6paddle16ParameterServer215ReadWriteBuffer15resetAlignAllocEv"],"paddle::ParameterServer2::ReadWriteBuffer::resizeWithAlignHints":[60,2,1,"_CPPv2N6paddle16ParameterServer215ReadWriteBuffer20resizeWithAlignHintsE6size_t6size_t"],"paddle::ParameterServer2::addGradBegin_":[60,3,1,"_CPPv2N6paddle16ParameterServer213addGradBegin_E"],"paddle::ParameterServer2::addGradient":[60,2,1,"_CPPv2N6paddle16ParameterServer211addGradientERK20SendParameterRequestRNSt6vectorI6BufferEEP21SendParameterResponsePNSt6vectorI6BufferEE"],"paddle::ParameterServer2::allClientPassFinish_":[60,3,1,"_CPPv2N6paddle16ParameterServer220allClientPassFinish_E"],"paddle::ParameterServer2::asyncFinishPass":[60,2,1,"_CPPv2N6paddle16ParameterServer215asyncFinishPassERK18SynchronizeRequest21ProtoResponseCallback"],"paddle::ParameterServer2::asyncGrdientCommitCheckAndStat":[60,2,1,"_CPPv2N6paddle16ParameterServer230asyncGrdientCommitCheckAndStatERK20SendParameterRequest"],"paddle::ParameterServer2::asyncLaggedGradientsNum_":[60,3,1,"_CPPv2N6paddle16ParameterServer224asyncLaggedGradientsNum_E"],"paddle::ParameterServer2::asyncLaggedThreshold_":[60,3,1,"_CPPv2N6paddle16ParameterServer221asyncLaggedThreshold_E"],"paddle::ParameterServer2::asyncSGD":[60,2,1,"_CPPv2N6paddle16ParameterServer28asyncSGDERK20SendParameterRequestRNSt6vectorI6BufferEEP21SendParameterResponsePNSt6vectorI6BufferEE"],"paddle::ParameterServer2::asyncTrainerCommitStat_":[60,3,1,"_CPPv2N6paddle16ParameterServer223asyncTrainerCommitStat_E"],"paddle::ParameterServer2::asyncTrainerDiscardStat_":[60,3,1,"_CPPv2N6paddle16ParameterServer224asyncTrainerDiscardStat_E"],"paddle::ParameterServer2::asyncTrainerSteps_":[60,3,1,"_CPPv2N6paddle16ParameterServer218asyncTrainerSteps_E"],"paddle::ParameterServer2::asyncUpdateStat_":[60,3,1,"_CPPv2N6paddle16ParameterServer216asyncUpdateStat_E"],"paddle::ParameterServer2::asyncUpdateSteps_":[60,3,1,"_CPPv2N6paddle16ParameterServer217asyncUpdateSteps_E"],"paddle::ParameterServer2::batchId_":[60,3,1,"_CPPv2N6paddle16ParameterServer28batchId_E"],"paddle::ParameterServer2::blockIdMap_":[60,3,1,"_CPPv2N6paddle16ParameterServer211blockIdMap_E"],"paddle::ParameterServer2::blockInfos_":[60,3,1,"_CPPv2N6paddle16ParameterServer211blockInfos_E"],"paddle::ParameterServer2::blockOffsetMap_":[60,3,1,"_CPPv2N6paddle16ParameterServer215blockOffsetMap_E"],"paddle::ParameterServer2::blockTraverse":[60,2,1,"_CPPv2N6paddle16ParameterServer213blockTraverseER9BlockInfoRK15ParameterConfig7int64_t6size_tA_K9VectorPtrRKN18ParameterOptimizer16TraverseCallbackE"],"paddle::ParameterServer2::callbackVec_":[60,3,1,"_CPPv2N6paddle16ParameterServer212callbackVec_E"],"paddle::ParameterServer2::clearUnusedSegments":[60,2,1,"_CPPv2N6paddle16ParameterServer219clearUnusedSegmentsEP9CpuVector"],"paddle::ParameterServer2::configMap_":[60,3,1,"_CPPv2N6paddle16ParameterServer210configMap_E"],"paddle::ParameterServer2::config_":[60,3,1,"_CPPv2N6paddle16ParameterServer27config_E"],"paddle::ParameterServer2::cost_":[60,3,1,"_CPPv2N6paddle16ParameterServer25cost_E"],"paddle::ParameterServer2::createMatrix":[60,2,1,"_CPPv2N6paddle16ParameterServer212createMatrixERK19CreateMatrixRequest21ProtoResponseCallback"],"paddle::ParameterServer2::createVector":[60,2,1,"_CPPv2N6paddle16ParameterServer212createVectorERK19CreateVectorRequest21ProtoResponseCallback"],"paddle::ParameterServer2::dataMems_":[60,3,1,"_CPPv2N6paddle16ParameterServer29dataMems_E"],"paddle::ParameterServer2::dataSize_":[60,3,1,"_CPPv2N6paddle16ParameterServer29dataSize_E"],"paddle::ParameterServer2::doOperation":[60,2,1,"_CPPv2N6paddle16ParameterServer211doOperationERK18DoOperationRequest21ProtoResponseCallback"],"paddle::ParameterServer2::getBlockId":[60,2,1,"_CPPv2N6paddle16ParameterServer210getBlockIdERK14ParameterBlock"],"paddle::ParameterServer2::getBlockOffset":[60,2,1,"_CPPv2N6paddle16ParameterServer214getBlockOffsetERK14ParameterBlock"],"paddle::ParameterServer2::getParameter":[60,2,1,"_CPPv2N6paddle16ParameterServer212getParameterERK20SendParameterRequestRNSt6vectorI6BufferEEP21SendParameterResponsePNSt6vectorI6BufferEE"],"paddle::ParameterServer2::getParameterConfig":[60,2,1,"_CPPv2N6paddle16ParameterServer218getParameterConfigERK14ParameterBlock"],"paddle::ParameterServer2::getParameterSparse":[60,2,1,"_CPPv2N6paddle16ParameterServer218getParameterSparseERK20SendParameterRequestRNSt6vectorI6BufferEEP21SendParameterResponsePNSt6vectorI6BufferEE"],"paddle::ParameterServer2::getStatus":[60,2,1,"_CPPv2N6paddle16ParameterServer29getStatusERK16GetStatusRequest21ProtoResponseCallback"],"paddle::ParameterServer2::gradientReadyBarrier_":[60,3,1,"_CPPv2N6paddle16ParameterServer221gradientReadyBarrier_E"],"paddle::ParameterServer2::init":[60,2,1,"_CPPv2N6paddle16ParameterServer24initEv"],"paddle::ParameterServer2::isSparseServer_":[60,3,1,"_CPPv2N6paddle16ParameterServer215isSparseServer_E"],"paddle::ParameterServer2::isValidMatrixHandle":[60,2,1,"_CPPv2N6paddle16ParameterServer219isValidMatrixHandleE7int64_tP8Response"],"paddle::ParameterServer2::isValidVectorHandle":[60,2,1,"_CPPv2N6paddle16ParameterServer219isValidVectorHandleE7int64_tP8Response"],"paddle::ParameterServer2::kRetMsgInvalidMatrixHandle":[60,3,1,"_CPPv2N6paddle16ParameterServer226kRetMsgInvalidMatrixHandleE"],"paddle::ParameterServer2::kRetMsgInvalidVectorHandle":[60,3,1,"_CPPv2N6paddle16ParameterServer226kRetMsgInvalidVectorHandleE"],"paddle::ParameterServer2::kRetMsgUnknownOperation":[60,3,1,"_CPPv2N6paddle16ParameterServer223kRetMsgUnknownOperationE"],"paddle::ParameterServer2::loadValueVector":[60,2,1,"_CPPv2N6paddle16ParameterServer215loadValueVectorERK16LoadValueRequest21ProtoResponseCallback"],"paddle::ParameterServer2::matrices_":[60,3,1,"_CPPv2N6paddle16ParameterServer29matrices_E"],"paddle::ParameterServer2::mergeSegments":[60,2,1,"_CPPv2N6paddle16ParameterServer213mergeSegmentsEP13BlockSegments"],"paddle::ParameterServer2::mpiSize_":[60,3,1,"_CPPv2N6paddle16ParameterServer28mpiSize_E"],"paddle::ParameterServer2::numPassFinishClients_":[60,3,1,"_CPPv2N6paddle16ParameterServer221numPassFinishClients_E"],"paddle::ParameterServer2::numSamplesProcessed_":[60,3,1,"_CPPv2N6paddle16ParameterServer220numSamplesProcessed_E"],"paddle::ParameterServer2::opFuncs":[60,3,1,"_CPPv2N6paddle16ParameterServer27opFuncsE"],"paddle::ParameterServer2::op_COPY":[60,2,1,"_CPPv2N6paddle16ParameterServer27op_COPYERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_RESET":[60,2,1,"_CPPv2N6paddle16ParameterServer28op_RESETERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_SGD":[60,2,1,"_CPPv2N6paddle16ParameterServer26op_SGDERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_apply":[60,2,1,"_CPPv2N6paddle16ParameterServer28op_applyERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_au":[60,2,1,"_CPPv2N6paddle16ParameterServer25op_auERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_au_bv":[60,2,1,"_CPPv2N6paddle16ParameterServer28op_au_bvERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_au_bv_cw":[60,2,1,"_CPPv2N6paddle16ParameterServer211op_au_bv_cwERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_cost":[60,2,1,"_CPPv2N6paddle16ParameterServer27op_costERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_dir_deriv":[60,2,1,"_CPPv2N6paddle16ParameterServer212op_dir_derivERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_finish_pass":[60,2,1,"_CPPv2N6paddle16ParameterServer214op_finish_passERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_fix_dir_signs":[60,2,1,"_CPPv2N6paddle16ParameterServer216op_fix_dir_signsERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_fix_omega_signs":[60,2,1,"_CPPv2N6paddle16ParameterServer218op_fix_omega_signsERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_load":[60,2,1,"_CPPv2N6paddle16ParameterServer27op_loadERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_make_steepest_desc_dir":[60,2,1,"_CPPv2N6paddle16ParameterServer225op_make_steepest_desc_dirERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_randomize":[60,2,1,"_CPPv2N6paddle16ParameterServer212op_randomizeERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_save":[60,2,1,"_CPPv2N6paddle16ParameterServer27op_saveERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_start_pass":[60,2,1,"_CPPv2N6paddle16ParameterServer213op_start_passERK9OperationP15OperationResult"],"paddle::ParameterServer2::op_utv":[60,2,1,"_CPPv2N6paddle16ParameterServer26op_utvERK9OperationP15OperationResult"],"paddle::ParameterServer2::parallelExecForEachBlock":[60,2,1,"_CPPv2N6paddle16ParameterServer224parallelExecForEachBlockE8ExecFunc"],"paddle::ParameterServer2::parameterMutex_":[60,3,1,"_CPPv2N6paddle16ParameterServer215parameterMutex_E"],"paddle::ParameterServer2::parameterReadyBarrier_":[60,3,1,"_CPPv2N6paddle16ParameterServer222parameterReadyBarrier_E"],"paddle::ParameterServer2::passBarrier_":[60,3,1,"_CPPv2N6paddle16ParameterServer212passBarrier_E"],"paddle::ParameterServer2::printAsyncGradientCommitStatAndReset":[60,2,1,"_CPPv2N6paddle16ParameterServer236printAsyncGradientCommitStatAndResetEv"],"paddle::ParameterServer2::readAllBlocks":[60,2,1,"_CPPv2N6paddle16ParameterServer213readAllBlocksEP9MsgReaderPNSt6vectorIN16ParameterServer26BufferEEE"],"paddle::ParameterServer2::readWriteBuffer_":[60,3,1,"_CPPv2N6paddle16ParameterServer216readWriteBuffer_E"],"paddle::ParameterServer2::reduceAndSendData":[60,2,1,"_CPPv2N6paddle16ParameterServer217reduceAndSendDataERK15SendDataRequestRNSt10unique_ptrI9MsgReaderEER23ProtoResponseCallbackEx"],"paddle::ParameterServer2::releaseMatrix":[60,2,1,"_CPPv2N6paddle16ParameterServer213releaseMatrixERK20ReleaseMatrixRequest21ProtoResponseCallback"],"paddle::ParameterServer2::releaseVector":[60,2,1,"_CPPv2N6paddle16ParameterServer213releaseVectorERK20ReleaseVectorRequest21ProtoResponseCallback"],"paddle::ParameterServer2::requestVec_":[60,3,1,"_CPPv2N6paddle16ParameterServer211requestVec_E"],"paddle::ParameterServer2::saveValueVector":[60,2,1,"_CPPv2N6paddle16ParameterServer215saveValueVectorERK16SaveValueRequest21ProtoResponseCallback"],"paddle::ParameterServer2::sendBackParameter":[60,2,1,"_CPPv2N6paddle16ParameterServer217sendBackParameterERK14ParameterBlockiP21SendParameterResponsePNSt6vectorI6BufferEE"],"paddle::ParameterServer2::sendBackParameterSparse":[60,2,1,"_CPPv2N6paddle16ParameterServer223sendBackParameterSparseERK14ParameterBlockiP21SendParameterResponseP6Buffer6size_tPNSt6vectorI6BufferEE"],"paddle::ParameterServer2::sendData":[60,2,1,"_CPPv2N6paddle16ParameterServer28sendDataERK15SendDataRequestNSt10unique_ptrI9MsgReaderEE23ProtoResponseCallbackEx"],"paddle::ParameterServer2::sendParameter":[60,2,1,"_CPPv2N6paddle16ParameterServer213sendParameterERK20SendParameterRequestNSt10unique_ptrI9MsgReaderEE23ProtoResponseCallbackEx"],"paddle::ParameterServer2::serverId_":[60,3,1,"_CPPv2N6paddle16ParameterServer29serverId_E"],"paddle::ParameterServer2::setConfig":[60,2,1,"_CPPv2N6paddle16ParameterServer29setConfigERK16SetConfigRequest21ProtoResponseCallback"],"paddle::ParameterServer2::setParameter":[60,2,1,"_CPPv2N6paddle16ParameterServer212setParameterERK20SendParameterRequestRNSt6vectorI6BufferEEP21SendParameterResponsePNSt6vectorI6BufferEE"],"paddle::ParameterServer2::setStatus":[60,2,1,"_CPPv2N6paddle16ParameterServer29setStatusERK16SetStatusRequest21ProtoResponseCallback"],"paddle::ParameterServer2::size_":[60,3,1,"_CPPv2N6paddle16ParameterServer25size_E"],"paddle::ParameterServer2::statSet_":[60,3,1,"_CPPv2N6paddle16ParameterServer28statSet_E"],"paddle::ParameterServer2::status_":[60,3,1,"_CPPv2N6paddle16ParameterServer27status_E"],"paddle::ParameterServer2::syncThreadPool_":[60,3,1,"_CPPv2N6paddle16ParameterServer215syncThreadPool_E"],"paddle::ParameterServer2::synchronize":[60,2,1,"_CPPv2N6paddle16ParameterServer211synchronizeERK18SynchronizeRequest21ProtoResponseCallback"],"paddle::ParameterServer2::synchronizeBarriers_":[60,3,1,"_CPPv2N6paddle16ParameterServer220synchronizeBarriers_E"],"paddle::ParameterServer2::templateReduceSum":[60,2,1,"_CPPv2N6paddle16ParameterServer217templateReduceSumERK15SendDataRequestRNSt10unique_ptrI9MsgReaderEER23ProtoResponseCallbackEx"],"paddle::ParameterServer2::tuningAsyncsgdFinished":[60,2,1,"_CPPv2N6paddle16ParameterServer222tuningAsyncsgdFinishedEv"],"paddle::ParameterServer2::tuningAsyncsgdMidOutput":[60,2,1,"_CPPv2N6paddle16ParameterServer223tuningAsyncsgdMidOutputEv"],"paddle::ParameterServer2::tuningSgdFinished":[60,2,1,"_CPPv2N6paddle16ParameterServer217tuningSgdFinishedEv"],"paddle::ParameterServer2::tuningSgdMidOutput":[60,2,1,"_CPPv2N6paddle16ParameterServer218tuningSgdMidOutputEv"],"paddle::ParameterServer2::usedSegments_":[60,3,1,"_CPPv2N6paddle16ParameterServer213usedSegments_E"],"paddle::ParameterServer2::vectors_":[60,3,1,"_CPPv2N6paddle16ParameterServer28vectors_E"],"paddle::ParameterServer2::waitPassFinish":[60,2,1,"_CPPv2N6paddle16ParameterServer214waitPassFinishERK21WaitPassFinishRequest21ProtoResponseCallback"],"paddle::ParameterServer2::waitPassStart":[60,2,1,"_CPPv2N6paddle16ParameterServer213waitPassStartERK20WaitPassStartRequest21ProtoResponseCallback"],"paddle::ParameterServer2::~ParameterServer2":[60,2,1,"_CPPv2N6paddle16ParameterServer2D0Ev"],"paddle::ParameterUpdater":[54,1,1,"_CPPv2N6paddle16ParameterUpdaterE"],"paddle::ParameterUpdater::ParameterUpdater":[54,2,1,"_CPPv2N6paddle16ParameterUpdater16ParameterUpdaterEv"],"paddle::ParameterUpdater::addParameterType":[54,2,1,"_CPPv2N6paddle16ParameterUpdater16addParameterTypeE13ParameterType"],"paddle::ParameterUpdater::apply":[54,2,1,"_CPPv2N6paddle16ParameterUpdater5applyEv"],"paddle::ParameterUpdater::catchUpWith":[54,2,1,"_CPPv2N6paddle16ParameterUpdater11catchUpWithEv"],"paddle::ParameterUpdater::finishBatch":[54,2,1,"_CPPv2N6paddle16ParameterUpdater11finishBatchE4real"],"paddle::ParameterUpdater::finishPass":[54,2,1,"_CPPv2N6paddle16ParameterUpdater10finishPassE4real"],"paddle::ParameterUpdater::getParameterTypes":[54,2,1,"_CPPv2N6paddle16ParameterUpdater17getParameterTypesEv"],"paddle::ParameterUpdater::getParametersRemote":[54,2,1,"_CPPv2N6paddle16ParameterUpdater19getParametersRemoteEbb"],"paddle::ParameterUpdater::init":[54,2,1,"_CPPv2N6paddle16ParameterUpdater4initERNSt6vectorI12ParameterPtrEE"],"paddle::ParameterUpdater::loadParametersRemote":[54,2,1,"_CPPv2N6paddle16ParameterUpdater20loadParametersRemoteERKNSt6stringE"],"paddle::ParameterUpdater::nonStaticParaIDMap_":[54,3,1,"_CPPv2N6paddle16ParameterUpdater19nonStaticParaIDMap_E"],"paddle::ParameterUpdater::parameterTypes_":[54,3,1,"_CPPv2N6paddle16ParameterUpdater15parameterTypes_E"],"paddle::ParameterUpdater::parameters_":[54,3,1,"_CPPv2N6paddle16ParameterUpdater11parameters_E"],"paddle::ParameterUpdater::randParametersRemote":[54,2,1,"_CPPv2N6paddle16ParameterUpdater20randParametersRemoteEv"],"paddle::ParameterUpdater::restore":[54,2,1,"_CPPv2N6paddle16ParameterUpdater7restoreEv"],"paddle::ParameterUpdater::saveParametersRemote":[54,2,1,"_CPPv2N6paddle16ParameterUpdater20saveParametersRemoteERKNSt6stringE"],"paddle::ParameterUpdater::setForwardbackwardTime":[54,2,1,"_CPPv2N6paddle16ParameterUpdater22setForwardbackwardTimeE8uint64_t"],"paddle::ParameterUpdater::startBatch":[54,2,1,"_CPPv2N6paddle16ParameterUpdater10startBatchE7int64_t"],"paddle::ParameterUpdater::startPass":[54,2,1,"_CPPv2N6paddle16ParameterUpdater9startPassEv"],"paddle::ParameterUpdater::update":[54,2,1,"_CPPv2N6paddle16ParameterUpdater6updateEP9Parameter"],"paddle::ParameterUpdater::updateImpl":[54,2,1,"_CPPv2N6paddle16ParameterUpdater10updateImplEP9Parameter"],"paddle::ParameterUpdater::~ParameterUpdater":[54,2,1,"_CPPv2N6paddle16ParameterUpdaterD0Ev"],"paddle::ParameterUpdaterComposite":[54,1,1,"_CPPv2N6paddle25ParameterUpdaterCompositeE"],"paddle::ParameterUpdaterComposite::ParameterUpdaterComposite":[54,2,1,"_CPPv2N6paddle25ParameterUpdaterComposite25ParameterUpdaterCompositeEv"],"paddle::ParameterUpdaterComposite::apply":[54,2,1,"_CPPv2N6paddle25ParameterUpdaterComposite5applyEv"],"paddle::ParameterUpdaterComposite::catchUpWith":[54,2,1,"_CPPv2N6paddle25ParameterUpdaterComposite11catchUpWithEv"],"paddle::ParameterUpdaterComposite::finishBatch":[54,2,1,"_CPPv2N6paddle25ParameterUpdaterComposite11finishBatchE4real"],"paddle::ParameterUpdaterComposite::finishPass":[54,2,1,"_CPPv2N6paddle25ParameterUpdaterComposite10finishPassE4real"],"paddle::ParameterUpdaterComposite::getParametersRemote":[54,2,1,"_CPPv2N6paddle25ParameterUpdaterComposite19getParametersRemoteEbb"],"paddle::ParameterUpdaterComposite::init":[54,2,1,"_CPPv2N6paddle25ParameterUpdaterComposite4initERNSt6vectorI12ParameterPtrEE"],"paddle::ParameterUpdaterComposite::loadParametersRemote":[54,2,1,"_CPPv2N6paddle25ParameterUpdaterComposite20loadParametersRemoteERKNSt6stringE"],"paddle::ParameterUpdaterComposite::randParametersRemote":[54,2,1,"_CPPv2N6paddle25ParameterUpdaterComposite20randParametersRemoteEv"],"paddle::ParameterUpdaterComposite::restore":[54,2,1,"_CPPv2N6paddle25ParameterUpdaterComposite7restoreEv"],"paddle::ParameterUpdaterComposite::saveParametersRemote":[54,2,1,"_CPPv2N6paddle25ParameterUpdaterComposite20saveParametersRemoteERKNSt6stringE"],"paddle::ParameterUpdaterComposite::setForwardbackwardTime":[54,2,1,"_CPPv2N6paddle25ParameterUpdaterComposite22setForwardbackwardTimeE8uint64_t"],"paddle::ParameterUpdaterComposite::startBatch":[54,2,1,"_CPPv2N6paddle25ParameterUpdaterComposite10startBatchE7int64_t"],"paddle::ParameterUpdaterComposite::startPass":[54,2,1,"_CPPv2N6paddle25ParameterUpdaterComposite9startPassEv"],"paddle::ParameterUpdaterComposite::syncThreadPool_":[54,3,1,"_CPPv2N6paddle25ParameterUpdaterComposite15syncThreadPool_E"],"paddle::ParameterUpdaterComposite::updateImpl":[54,2,1,"_CPPv2N6paddle25ParameterUpdaterComposite10updateImplEP9Parameter"],"paddle::ParameterUpdaterComposite::updaters_":[54,3,1,"_CPPv2N6paddle25ParameterUpdaterComposite9updaters_E"],"paddle::ParameterUpdaterComposite::~ParameterUpdaterComposite":[54,2,1,"_CPPv2N6paddle25ParameterUpdaterCompositeD0Ev"],"paddle::PnpairEvaluator":[38,1,1,"_CPPv2N6paddle15PnpairEvaluatorE"],"paddle::PnpairEvaluator::PnpairEvaluator":[38,2,1,"_CPPv2N6paddle15PnpairEvaluator15PnpairEvaluatorEv"],"paddle::PnpairEvaluator::PredictionResult":[38,1,1,"_CPPv2N6paddle15PnpairEvaluator16PredictionResultE"],"paddle::PnpairEvaluator::PredictionResult::PredictionResult":[38,2,1,"_CPPv2N6paddle15PnpairEvaluator16PredictionResult16PredictionResultE4realii4real"],"paddle::PnpairEvaluator::PredictionResult::label":[38,3,1,"_CPPv2N6paddle15PnpairEvaluator16PredictionResult5labelE"],"paddle::PnpairEvaluator::PredictionResult::out":[38,3,1,"_CPPv2N6paddle15PnpairEvaluator16PredictionResult3outE"],"paddle::PnpairEvaluator::PredictionResult::queryid":[38,3,1,"_CPPv2N6paddle15PnpairEvaluator16PredictionResult7queryidE"],"paddle::PnpairEvaluator::PredictionResult::weight":[38,3,1,"_CPPv2N6paddle15PnpairEvaluator16PredictionResult6weightE"],"paddle::PnpairEvaluator::calc":[38,2,1,"_CPPv2N6paddle15PnpairEvaluator4calcERNSt6vectorI16PredictionResultEE"],"paddle::PnpairEvaluator::distributeEval":[38,2,1,"_CPPv2N6paddle15PnpairEvaluator14distributeEvalEP16ParameterClient2"],"paddle::PnpairEvaluator::evalImp":[38,2,1,"_CPPv2N6paddle15PnpairEvaluator7evalImpERNSt6vectorI8ArgumentEE"],"paddle::PnpairEvaluator::finish":[38,2,1,"_CPPv2N6paddle15PnpairEvaluator6finishEv"],"paddle::PnpairEvaluator::predictArray_":[38,3,1,"_CPPv2N6paddle15PnpairEvaluator13predictArray_E"],"paddle::PnpairEvaluator::printPredictResults":[38,2,1,"_CPPv2N6paddle15PnpairEvaluator19printPredictResultsEv"],"paddle::PnpairEvaluator::printStats":[38,2,1,"_CPPv2N6paddle15PnpairEvaluator10printStatsERNSt7ostreamE"],"paddle::PnpairEvaluator::start":[38,2,1,"_CPPv2N6paddle15PnpairEvaluator5startEv"],"paddle::PnpairEvaluator::stat":[38,2,1,"_CPPv2N6paddle15PnpairEvaluator4statE6size_t6size_tP16PredictionResultRdRdRd"],"paddle::PnpairValidation":[43,1,1,"_CPPv2N6paddle16PnpairValidationE"],"paddle::PnpairValidation::PnpairValidation":[43,2,1,"_CPPv2N6paddle16PnpairValidation16PnpairValidationERK11LayerConfig"],"paddle::PnpairValidation::init":[43,2,1,"_CPPv2N6paddle16PnpairValidation4initERK8LayerMapRK12ParameterMap"],"paddle::PnpairValidation::onPassEnd":[43,2,1,"_CPPv2N6paddle16PnpairValidation9onPassEndEv"],"paddle::PnpairValidation::validationImp":[43,2,1,"_CPPv2N6paddle16PnpairValidation13validationImpE9MatrixPtr10IVectorPtr"],"paddle::PoolAllocator":[48,1,1,"_CPPv2N6paddle13PoolAllocatorE"],"paddle::PoolAllocator::PoolAllocator":[48,2,1,"_CPPv2N6paddle13PoolAllocator13PoolAllocatorEP9Allocator6size_tRKNSt6stringE"],"paddle::PoolAllocator::alloc":[48,2,1,"_CPPv2N6paddle13PoolAllocator5allocE6size_t"],"paddle::PoolAllocator::allocator_":[48,3,1,"_CPPv2N6paddle13PoolAllocator10allocator_E"],"paddle::PoolAllocator::free":[48,2,1,"_CPPv2N6paddle13PoolAllocator4freeEPv6size_t"],"paddle::PoolAllocator::freeAll":[48,2,1,"_CPPv2N6paddle13PoolAllocator7freeAllEv"],"paddle::PoolAllocator::getName":[48,2,1,"_CPPv2N6paddle13PoolAllocator7getNameEv"],"paddle::PoolAllocator::mutex_":[48,3,1,"_CPPv2N6paddle13PoolAllocator6mutex_E"],"paddle::PoolAllocator::name_":[48,3,1,"_CPPv2N6paddle13PoolAllocator5name_E"],"paddle::PoolAllocator::poolMemorySize_":[48,3,1,"_CPPv2N6paddle13PoolAllocator15poolMemorySize_E"],"paddle::PoolAllocator::pool_":[48,3,1,"_CPPv2N6paddle13PoolAllocator5pool_E"],"paddle::PoolAllocator::printAll":[48,2,1,"_CPPv2N6paddle13PoolAllocator8printAllEv"],"paddle::PoolAllocator::sizeLimit_":[48,3,1,"_CPPv2N6paddle13PoolAllocator10sizeLimit_E"],"paddle::PoolAllocator::~PoolAllocator":[48,2,1,"_CPPv2N6paddle13PoolAllocatorD0Ev"],"paddle::PoolLayer":[43,1,1,"_CPPv2N6paddle9PoolLayerE"],"paddle::PoolLayer::PoolLayer":[43,2,1,"_CPPv2N6paddle9PoolLayer9PoolLayerERK11LayerConfig"],"paddle::PoolLayer::channels_":[43,3,1,"_CPPv2N6paddle9PoolLayer9channels_E"],"paddle::PoolLayer::confPaddingY_":[43,3,1,"_CPPv2N6paddle9PoolLayer13confPaddingY_E"],"paddle::PoolLayer::confPadding_":[43,3,1,"_CPPv2N6paddle9PoolLayer12confPadding_E"],"paddle::PoolLayer::create":[43,2,1,"_CPPv2N6paddle9PoolLayer6createERK11LayerConfig"],"paddle::PoolLayer::imgSizeY_":[43,3,1,"_CPPv2N6paddle9PoolLayer9imgSizeY_E"],"paddle::PoolLayer::imgSize_":[43,3,1,"_CPPv2N6paddle9PoolLayer8imgSize_E"],"paddle::PoolLayer::init":[43,2,1,"_CPPv2N6paddle9PoolLayer4initERK8LayerMapRK12ParameterMap"],"paddle::PoolLayer::outputX_":[43,3,1,"_CPPv2N6paddle9PoolLayer8outputX_E"],"paddle::PoolLayer::outputY_":[43,3,1,"_CPPv2N6paddle9PoolLayer8outputY_E"],"paddle::PoolLayer::poolType_":[43,3,1,"_CPPv2N6paddle9PoolLayer9poolType_E"],"paddle::PoolLayer::sizeX_":[43,3,1,"_CPPv2N6paddle9PoolLayer6sizeX_E"],"paddle::PoolLayer::sizeY_":[43,3,1,"_CPPv2N6paddle9PoolLayer6sizeY_E"],"paddle::PoolLayer::start_":[43,3,1,"_CPPv2N6paddle9PoolLayer6start_E"],"paddle::PoolLayer::strideY_":[43,3,1,"_CPPv2N6paddle9PoolLayer8strideY_E"],"paddle::PoolLayer::stride_":[43,3,1,"_CPPv2N6paddle9PoolLayer7stride_E"],"paddle::PoolProjectionLayer":[43,1,1,"_CPPv2N6paddle19PoolProjectionLayerE"],"paddle::PoolProjectionLayer::PoolProjectionLayer":[43,2,1,"_CPPv2N6paddle19PoolProjectionLayer19PoolProjectionLayerERK11LayerConfig"],"paddle::PoolProjectionLayer::getSize":[43,2,1,"_CPPv2N6paddle19PoolProjectionLayer7getSizeEv"],"paddle::PoolProjectionLayer::imgSizeH_":[43,3,1,"_CPPv2N6paddle19PoolProjectionLayer9imgSizeH_E"],"paddle::PoolProjectionLayer::imgSizeW_":[43,3,1,"_CPPv2N6paddle19PoolProjectionLayer9imgSizeW_E"],"paddle::PoolProjectionLayer::outputH_":[43,3,1,"_CPPv2N6paddle19PoolProjectionLayer8outputH_E"],"paddle::PoolProjectionLayer::outputW_":[43,3,1,"_CPPv2N6paddle19PoolProjectionLayer8outputW_E"],"paddle::PowerLayer":[43,1,1,"_CPPv2N6paddle10PowerLayerE"],"paddle::PowerLayer::PowerLayer":[43,2,1,"_CPPv2N6paddle10PowerLayer10PowerLayerERK11LayerConfig"],"paddle::PowerLayer::backward":[43,2,1,"_CPPv2N6paddle10PowerLayer8backwardERK14UpdateCallback"],"paddle::PowerLayer::forward":[43,2,1,"_CPPv2N6paddle10PowerLayer7forwardE8PassType"],"paddle::PowerLayer::init":[43,2,1,"_CPPv2N6paddle10PowerLayer4initERK8LayerMapRK12ParameterMap"],"paddle::PowerLayer::tmpMtx":[43,3,1,"_CPPv2N6paddle10PowerLayer6tmpMtxE"],"paddle::PowerLayer::~PowerLayer":[43,2,1,"_CPPv2N6paddle10PowerLayerD0Ev"],"paddle::PrecisionRecallEvaluator":[38,1,1,"_CPPv2N6paddle24PrecisionRecallEvaluatorE"],"paddle::PrecisionRecallEvaluator::PrecisionRecallEvaluator":[38,2,1,"_CPPv2N6paddle24PrecisionRecallEvaluator24PrecisionRecallEvaluatorEv"],"paddle::PrecisionRecallEvaluator::StatsInfo":[38,1,1,"_CPPv2N6paddle24PrecisionRecallEvaluator9StatsInfoE"],"paddle::PrecisionRecallEvaluator::StatsInfo::FN":[38,3,1,"_CPPv2N6paddle24PrecisionRecallEvaluator9StatsInfo2FNE"],"paddle::PrecisionRecallEvaluator::StatsInfo::FP":[38,3,1,"_CPPv2N6paddle24PrecisionRecallEvaluator9StatsInfo2FPE"],"paddle::PrecisionRecallEvaluator::StatsInfo::StatsInfo":[38,2,1,"_CPPv2N6paddle24PrecisionRecallEvaluator9StatsInfo9StatsInfoEv"],"paddle::PrecisionRecallEvaluator::StatsInfo::TN":[38,3,1,"_CPPv2N6paddle24PrecisionRecallEvaluator9StatsInfo2TNE"],"paddle::PrecisionRecallEvaluator::StatsInfo::TP":[38,3,1,"_CPPv2N6paddle24PrecisionRecallEvaluator9StatsInfo2TPE"],"paddle::PrecisionRecallEvaluator::distributeEval":[38,2,1,"_CPPv2N6paddle24PrecisionRecallEvaluator14distributeEvalEP16ParameterClient2"],"paddle::PrecisionRecallEvaluator::evalImp":[38,2,1,"_CPPv2N6paddle24PrecisionRecallEvaluator7evalImpERNSt6vectorI8ArgumentEE"],"paddle::PrecisionRecallEvaluator::printStats":[38,2,1,"_CPPv2N6paddle24PrecisionRecallEvaluator10printStatsERNSt7ostreamE"],"paddle::PrecisionRecallEvaluator::start":[38,2,1,"_CPPv2N6paddle24PrecisionRecallEvaluator5startEv"],"paddle::Projection":[43,1,1,"_CPPv2N6paddle10ProjectionE"],"paddle::Projection::Projection":[43,2,1,"_CPPv2N6paddle10Projection10ProjectionERK16ProjectionConfig12ParameterPtrb"],"paddle::Projection::backward":[43,2,1,"_CPPv2N6paddle10Projection8backwardERK14UpdateCallback"],"paddle::Projection::config_":[43,3,1,"_CPPv2N6paddle10Projection7config_E"],"paddle::Projection::create":[43,2,1,"_CPPv2N6paddle10Projection6createERK16ProjectionConfig12ParameterPtrb"],"paddle::Projection::forward":[43,2,1,"_CPPv2N6paddle10Projection7forwardEv"],"paddle::Projection::getName":[43,2,1,"_CPPv2N6paddle10Projection7getNameEv"],"paddle::Projection::getOutputSize":[43,2,1,"_CPPv2N6paddle10Projection13getOutputSizeEv"],"paddle::Projection::getState":[43,2,1,"_CPPv2N6paddle10Projection8getStateEv"],"paddle::Projection::in_":[43,3,1,"_CPPv2N6paddle10Projection3in_E"],"paddle::Projection::out_":[43,3,1,"_CPPv2N6paddle10Projection4out_E"],"paddle::Projection::parameter_":[43,3,1,"_CPPv2N6paddle10Projection10parameter_E"],"paddle::Projection::passType_":[43,3,1,"_CPPv2N6paddle10Projection9passType_E"],"paddle::Projection::prefetch":[43,2,1,"_CPPv2N6paddle10Projection8prefetchEPK8Argument"],"paddle::Projection::registrar_":[43,3,1,"_CPPv2N6paddle10Projection10registrar_E"],"paddle::Projection::resetState":[43,2,1,"_CPPv2N6paddle10Projection10resetStateEv"],"paddle::Projection::setState":[43,2,1,"_CPPv2N6paddle10Projection8setStateE13LayerStatePtr"],"paddle::Projection::useGpu_":[43,3,1,"_CPPv2N6paddle10Projection7useGpu_E"],"paddle::Projection::~Projection":[43,2,1,"_CPPv2N6paddle10ProjectionD0Ev"],"paddle::ProtoDataProvider":[36,1,1,"_CPPv2N6paddle17ProtoDataProviderE"],"paddle::ProtoDataProvider::ProtoDataProvider":[36,2,1,"_CPPv2N6paddle17ProtoDataProvider17ProtoDataProviderERK10DataConfigbb"],"paddle::ProtoDataProvider::ProtoSlot":[36,1,1,"_CPPv2N6paddle17ProtoDataProvider9ProtoSlotE"],"paddle::ProtoDataProvider::ProtoSlot::denseData":[36,3,1,"_CPPv2N6paddle17ProtoDataProvider9ProtoSlot9denseDataE"],"paddle::ProtoDataProvider::ProtoSlot::dim":[36,3,1,"_CPPv2N6paddle17ProtoDataProvider9ProtoSlot3dimE"],"paddle::ProtoDataProvider::ProtoSlot::indexData":[36,3,1,"_CPPv2N6paddle17ProtoDataProvider9ProtoSlot9indexDataE"],"paddle::ProtoDataProvider::ProtoSlot::indices":[36,3,1,"_CPPv2N6paddle17ProtoDataProvider9ProtoSlot7indicesE"],"paddle::ProtoDataProvider::ProtoSlot::sparseFloatValueData":[36,3,1,"_CPPv2N6paddle17ProtoDataProvider9ProtoSlot20sparseFloatValueDataE"],"paddle::ProtoDataProvider::ProtoSlot::sparseNonValueData":[36,3,1,"_CPPv2N6paddle17ProtoDataProvider9ProtoSlot18sparseNonValueDataE"],"paddle::ProtoDataProvider::ProtoSlot::strData":[36,3,1,"_CPPv2N6paddle17ProtoDataProvider9ProtoSlot7strDataE"],"paddle::ProtoDataProvider::ProtoSlot::subIndices":[36,3,1,"_CPPv2N6paddle17ProtoDataProvider9ProtoSlot10subIndicesE"],"paddle::ProtoDataProvider::ProtoSlot::type":[36,3,1,"_CPPv2N6paddle17ProtoDataProvider9ProtoSlot4typeE"],"paddle::ProtoDataProvider::ProtoSlot::varDenseData":[36,3,1,"_CPPv2N6paddle17ProtoDataProvider9ProtoSlot12varDenseDataE"],"paddle::ProtoDataProvider::ProtoSlot::varIndices":[36,3,1,"_CPPv2N6paddle17ProtoDataProvider9ProtoSlot10varIndicesE"],"paddle::ProtoDataProvider::ProtoVarSlot":[36,1,1,"_CPPv2N6paddle17ProtoDataProvider12ProtoVarSlotE"],"paddle::ProtoDataProvider::ProtoVarSlot::data":[36,3,1,"_CPPv2N6paddle17ProtoDataProvider12ProtoVarSlot4dataE"],"paddle::ProtoDataProvider::ProtoVarSlot::dims":[36,3,1,"_CPPv2N6paddle17ProtoDataProvider12ProtoVarSlot4dimsE"],"paddle::ProtoDataProvider::checkDataHeader":[36,2,1,"_CPPv2N6paddle17ProtoDataProvider15checkDataHeaderERK10DataHeader"],"paddle::ProtoDataProvider::checkSample":[36,2,1,"_CPPv2N6paddle17ProtoDataProvider11checkSampleERK10DataSample"],"paddle::ProtoDataProvider::cpuBatch_":[36,3,1,"_CPPv2N6paddle17ProtoDataProvider9cpuBatch_E"],"paddle::ProtoDataProvider::currentSequenceIndex_":[36,3,1,"_CPPv2N6paddle17ProtoDataProvider21currentSequenceIndex_E"],"paddle::ProtoDataProvider::fillSlots":[36,2,1,"_CPPv2N6paddle17ProtoDataProvider9fillSlotsERK10DataSample"],"paddle::ProtoDataProvider::getNextBatchInternal":[36,2,1,"_CPPv2N6paddle17ProtoDataProvider20getNextBatchInternalE7int64_tP9DataBatch"],"paddle::ProtoDataProvider::getSize":[36,2,1,"_CPPv2N6paddle17ProtoDataProvider7getSizeEv"],"paddle::ProtoDataProvider::gpuBatch_":[36,3,1,"_CPPv2N6paddle17ProtoDataProvider9gpuBatch_E"],"paddle::ProtoDataProvider::header_":[36,3,1,"_CPPv2N6paddle17ProtoDataProvider7header_E"],"paddle::ProtoDataProvider::iidData":[36,2,1,"_CPPv2N6paddle17ProtoDataProvider7iidDataEv"],"paddle::ProtoDataProvider::loadData":[36,2,1,"_CPPv2N6paddle17ProtoDataProvider8loadDataERKNSt6vectorINSt6stringEEE"],"paddle::ProtoDataProvider::loadDataFile":[36,2,1,"_CPPv2N6paddle17ProtoDataProvider12loadDataFileERKNSt6stringE"],"paddle::ProtoDataProvider::lock_":[36,3,1,"_CPPv2N6paddle17ProtoDataProvider5lock_E"],"paddle::ProtoDataProvider::nnzStats_":[36,3,1,"_CPPv2N6paddle17ProtoDataProvider9nnzStats_E"],"paddle::ProtoDataProvider::numVecSlots_":[36,3,1,"_CPPv2N6paddle17ProtoDataProvider12numVecSlots_E"],"paddle::ProtoDataProvider::reset":[36,2,1,"_CPPv2N6paddle17ProtoDataProvider5resetEv"],"paddle::ProtoDataProvider::sampleLoop":[36,2,1,"_CPPv2N6paddle17ProtoDataProvider10sampleLoopE2Op7int64_t"],"paddle::ProtoDataProvider::sampleNums_":[36,3,1,"_CPPv2N6paddle17ProtoDataProvider11sampleNums_E"],"paddle::ProtoDataProvider::sequenceLoop":[36,2,1,"_CPPv2N6paddle17ProtoDataProvider12sequenceLoopE2Op7int64_t"],"paddle::ProtoDataProvider::sequenceStartPositions_":[36,3,1,"_CPPv2N6paddle17ProtoDataProvider23sequenceStartPositions_E"],"paddle::ProtoDataProvider::showDataStats":[36,2,1,"_CPPv2N6paddle17ProtoDataProvider13showDataStatsEv"],"paddle::ProtoDataProvider::shuffle":[36,2,1,"_CPPv2N6paddle17ProtoDataProvider7shuffleEv"],"paddle::ProtoDataProvider::shuffledSequenceIds_":[36,3,1,"_CPPv2N6paddle17ProtoDataProvider20shuffledSequenceIds_E"],"paddle::ProtoDataProvider::slots_":[36,3,1,"_CPPv2N6paddle17ProtoDataProvider6slots_E"],"paddle::ProtoDataProvider::subSampleLoop":[36,2,1,"_CPPv2N6paddle17ProtoDataProvider13subSampleLoopE2Op7int64_ti"],"paddle::ProtoSequenceDataProvider":[36,1,1,"_CPPv2N6paddle25ProtoSequenceDataProviderE"],"paddle::ProtoSequenceDataProvider::ProtoSequenceDataProvider":[36,2,1,"_CPPv2N6paddle25ProtoSequenceDataProvider25ProtoSequenceDataProviderERK10DataConfigbb"],"paddle::ProtoSequenceDataProvider::getNextBatchInternal":[36,2,1,"_CPPv2N6paddle25ProtoSequenceDataProvider20getNextBatchInternalE7int64_tP9DataBatch"],"paddle::ProtoSequenceDataProvider::~ProtoSequenceDataProvider":[36,2,1,"_CPPv2N6paddle25ProtoSequenceDataProviderD0Ev"],"paddle::ProtoServer":[60,1,1,"_CPPv2N6paddle11ProtoServerE"],"paddle::ProtoServer::ProtoResponseCallback":[60,4,1,"_CPPv2N6paddle11ProtoServer21ProtoResponseCallbackE"],"paddle::ProtoServer::ProtoResponseCallbackEx":[60,4,1,"_CPPv2N6paddle11ProtoServer23ProtoResponseCallbackExE"],"paddle::ProtoServer::ProtoServer":[60,2,1,"_CPPv2N6paddle11ProtoServer11ProtoServerERKNSt6stringEii"],"paddle::ProtoServer::ServiceFunction":[60,4,1,"_CPPv2N6paddle11ProtoServer15ServiceFunctionE"],"paddle::ProtoServer::handleRequest":[60,2,1,"_CPPv2N6paddle11ProtoServer13handleRequestENSt10unique_ptrI9MsgReaderEE16ResponseCallback"],"paddle::ProtoServer::handleRequestBegin_":[60,3,1,"_CPPv2N6paddle11ProtoServer19handleRequestBegin_E"],"paddle::ProtoServer::nameToFuncMap_":[60,3,1,"_CPPv2N6paddle11ProtoServer14nameToFuncMap_E"],"paddle::ProtoServer::registerServiceFunction":[60,2,1,"_CPPv2N6paddle11ProtoServer23registerServiceFunctionERKNSt6stringENSt8functionIFvRK7ProtoIn21ProtoResponseCallbackEEE"],"paddle::ProtoServer::registerServiceFunctionEx":[60,2,1,"_CPPv2N6paddle11ProtoServer25registerServiceFunctionExERKNSt6stringENSt8functionIFvRK7ProtoInNSt10unique_ptrI9MsgReaderEE23ProtoResponseCallbackExEEE"],"paddle::ProtoServer::registerServiceFunctionImp":[60,2,1,"_CPPv2N6paddle11ProtoServer26registerServiceFunctionImpERKNSt6stringE15ServiceFunction"],"paddle::PyDataProvider2":[36,1,1,"_CPPv2N6paddle15PyDataProvider2E"],"paddle::PyDataProvider2::PyDataProvider2":[36,2,1,"_CPPv2N6paddle15PyDataProvider215PyDataProvider2ERK10DataConfigb"],"paddle::PyDataProvider2::getNextBatchInternal":[36,2,1,"_CPPv2N6paddle15PyDataProvider220getNextBatchInternalE7int64_tP9DataBatch"],"paddle::PyDataProvider2::getSize":[36,2,1,"_CPPv2N6paddle15PyDataProvider27getSizeEv"],"paddle::PyDataProvider2::reset":[36,2,1,"_CPPv2N6paddle15PyDataProvider25resetEv"],"paddle::PyDataProvider2::shuffle":[36,2,1,"_CPPv2N6paddle15PyDataProvider27shuffleEv"],"paddle::PyDataProvider2::~PyDataProvider2":[36,2,1,"_CPPv2N6paddle15PyDataProvider2D0Ev"],"paddle::Queue":[65,1,1,"_CPPv2N6paddle5QueueE"],"paddle::Queue::Queue":[65,2,1,"_CPPv2N6paddle5Queue5QueueEv"],"paddle::Queue::dequeue":[65,2,1,"_CPPv2N6paddle5Queue7dequeueEv"],"paddle::Queue::empty":[65,2,1,"_CPPv2N6paddle5Queue5emptyEv"],"paddle::Queue::enqueue":[65,2,1,"_CPPv2N6paddle5Queue7enqueueERR1T"],"paddle::Queue::size":[65,2,1,"_CPPv2N6paddle5Queue4sizeEv"],"paddle::Queue::waitEmpty":[65,2,1,"_CPPv2N6paddle5Queue9waitEmptyEv"],"paddle::Queue::~Queue":[65,2,1,"_CPPv2N6paddle5QueueD0Ev"],"paddle::RMSPropParameterOptimizer":[50,1,1,"_CPPv2N6paddle25RMSPropParameterOptimizerE"],"paddle::RMSPropParameterOptimizer::RMSPropParameterOptimizer":[50,2,1,"_CPPv2N6paddle25RMSPropParameterOptimizer25RMSPropParameterOptimizerERK18OptimizationConfig"],"paddle::RMSPropParameterOptimizer::epsilon_":[50,3,1,"_CPPv2N6paddle25RMSPropParameterOptimizer8epsilon_E"],"paddle::RMSPropParameterOptimizer::finishBatch":[50,2,1,"_CPPv2N6paddle25RMSPropParameterOptimizer11finishBatchEv"],"paddle::RMSPropParameterOptimizer::init":[50,2,1,"_CPPv2N6paddle25RMSPropParameterOptimizer4initE6size_tPK15ParameterConfig"],"paddle::RMSPropParameterOptimizer::rou_":[50,3,1,"_CPPv2N6paddle25RMSPropParameterOptimizer4rou_E"],"paddle::RMSPropParameterOptimizer::startBatch":[50,2,1,"_CPPv2N6paddle25RMSPropParameterOptimizer10startBatchE7int64_t"],"paddle::RMSPropParameterOptimizer::t0Vec_":[50,3,1,"_CPPv2N6paddle25RMSPropParameterOptimizer6t0Vec_E"],"paddle::RMSPropParameterOptimizer::timer_":[50,3,1,"_CPPv2N6paddle25RMSPropParameterOptimizer6timer_E"],"paddle::RMSPropParameterOptimizer::update":[50,2,1,"_CPPv2N6paddle25RMSPropParameterOptimizer6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::RWLock":[66,1,1,"_CPPv2N6paddle6RWLockE"],"paddle::RWLock::RWLock":[66,2,1,"_CPPv2N6paddle6RWLock6RWLockEv"],"paddle::RWLock::lock":[66,2,1,"_CPPv2N6paddle6RWLock4lockEv"],"paddle::RWLock::lock_shared":[66,2,1,"_CPPv2N6paddle6RWLock11lock_sharedEv"],"paddle::RWLock::operator=":[66,2,1,"_CPPv2N6paddle6RWLockaSERK6RWLock"],"paddle::RWLock::rwlock_":[66,3,1,"_CPPv2N6paddle6RWLock7rwlock_E"],"paddle::RWLock::unlock":[66,2,1,"_CPPv2N6paddle6RWLock6unlockEv"],"paddle::RWLock::~RWLock":[66,2,1,"_CPPv2N6paddle6RWLockD0Ev"],"paddle::RankAucEvaluator":[38,1,1,"_CPPv2N6paddle16RankAucEvaluatorE"],"paddle::RankAucEvaluator::distributeEval":[38,2,1,"_CPPv2N6paddle16RankAucEvaluator14distributeEvalEP16ParameterClient2"],"paddle::RankAucEvaluator::evalImp":[38,2,1,"_CPPv2N6paddle16RankAucEvaluator7evalImpERNSt6vectorI8ArgumentEE"],"paddle::RankAucEvaluator::start":[38,2,1,"_CPPv2N6paddle16RankAucEvaluator5startEv"],"paddle::RankAucEvaluator::updateSamplesNum":[38,2,1,"_CPPv2N6paddle16RankAucEvaluator16updateSamplesNumERKNSt6vectorI8ArgumentEE"],"paddle::RankingCost":[43,1,1,"_CPPv2N6paddle11RankingCostE"],"paddle::RankingCost::RankingCost":[43,2,1,"_CPPv2N6paddle11RankingCost11RankingCostERK11LayerConfig"],"paddle::RankingCost::backward":[43,2,1,"_CPPv2N6paddle11RankingCost8backwardERK14UpdateCallback"],"paddle::RankingCost::backwardImp":[43,2,1,"_CPPv2N6paddle11RankingCost11backwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::RankingCost::forward":[43,2,1,"_CPPv2N6paddle11RankingCost7forwardE8PassType"],"paddle::RankingCost::forwardImp":[43,2,1,"_CPPv2N6paddle11RankingCost10forwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::RankingCost::getLabelLayer":[43,2,1,"_CPPv2N6paddle11RankingCost13getLabelLayerEv"],"paddle::RankingCost::getOutputLayer":[43,2,1,"_CPPv2N6paddle11RankingCost14getOutputLayerE6size_t"],"paddle::RankingCost::init":[43,2,1,"_CPPv2N6paddle11RankingCost4initERK8LayerMapRK12ParameterMap"],"paddle::RankingCost::onPassEnd":[43,2,1,"_CPPv2N6paddle11RankingCost9onPassEndEv"],"paddle::ReadLockGuard":[66,1,1,"_CPPv2N6paddle13ReadLockGuardE"],"paddle::ReadLockGuard::ReadLockGuard":[66,2,1,"_CPPv2N6paddle13ReadLockGuard13ReadLockGuardER6RWLock"],"paddle::ReadLockGuard::rwlock_":[66,3,1,"_CPPv2N6paddle13ReadLockGuard7rwlock_E"],"paddle::ReadLockGuard::~ReadLockGuard":[66,2,1,"_CPPv2N6paddle13ReadLockGuardD0Ev"],"paddle::RecurrentGradientMachine":[40,1,1,"_CPPv2N6paddle24RecurrentGradientMachineE"],"paddle::RecurrentGradientMachine::BeamSearchCandidatesAdjustCallback":[40,4,1,"_CPPv2N6paddle24RecurrentGradientMachine34BeamSearchCandidatesAdjustCallbackE"],"paddle::RecurrentGradientMachine::DropCallback":[40,4,1,"_CPPv2N6paddle24RecurrentGradientMachine12DropCallbackE"],"paddle::RecurrentGradientMachine::EachStepCallback":[40,4,1,"_CPPv2N6paddle24RecurrentGradientMachine16EachStepCallbackE"],"paddle::RecurrentGradientMachine::EosFrameLine":[40,1,1,"_CPPv2N6paddle24RecurrentGradientMachine12EosFrameLineE"],"paddle::RecurrentGradientMachine::EosFrameLine::layers":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine12EosFrameLine6layersE"],"paddle::RecurrentGradientMachine::Generator":[40,1,1,"_CPPv2N6paddle24RecurrentGradientMachine9GeneratorE"],"paddle::RecurrentGradientMachine::Generator::config":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine9Generator6configE"],"paddle::RecurrentGradientMachine::Generator::ids":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine9Generator3idsE"],"paddle::RecurrentGradientMachine::Generator::outArg":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine9Generator6outArgE"],"paddle::RecurrentGradientMachine::InFrameLine":[40,1,1,"_CPPv2N6paddle24RecurrentGradientMachine11InFrameLineE"],"paddle::RecurrentGradientMachine::InFrameLine::agents":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine11InFrameLine6agentsE"],"paddle::RecurrentGradientMachine::InFrameLine::hasSubseq":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine11InFrameLine9hasSubseqE"],"paddle::RecurrentGradientMachine::InFrameLine::inLayer":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine11InFrameLine7inLayerE"],"paddle::RecurrentGradientMachine::InFrameLine::linkName":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine11InFrameLine8linkNameE"],"paddle::RecurrentGradientMachine::InFrameLine::outArg":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine11InFrameLine6outArgE"],"paddle::RecurrentGradientMachine::Info":[40,1,1,"_CPPv2N6paddle24RecurrentGradientMachine4InfoE"],"paddle::RecurrentGradientMachine::Info::allIds":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine4Info6allIdsE"],"paddle::RecurrentGradientMachine::Info::idIndex":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine4Info7idIndexE"],"paddle::RecurrentGradientMachine::Info::seqStartPosIndex":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine4Info16seqStartPosIndexE"],"paddle::RecurrentGradientMachine::Info::sequenceStartPositions":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine4Info22sequenceStartPositionsE"],"paddle::RecurrentGradientMachine::MemoryFrameLine":[40,1,1,"_CPPv2N6paddle24RecurrentGradientMachine15MemoryFrameLineE"],"paddle::RecurrentGradientMachine::MemoryFrameLine::agents":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine15MemoryFrameLine6agentsE"],"paddle::RecurrentGradientMachine::MemoryFrameLine::allIds":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine15MemoryFrameLine6allIdsE"],"paddle::RecurrentGradientMachine::MemoryFrameLine::biasLayer":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine15MemoryFrameLine9biasLayerE"],"paddle::RecurrentGradientMachine::MemoryFrameLine::bootLayer":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine15MemoryFrameLine9bootLayerE"],"paddle::RecurrentGradientMachine::MemoryFrameLine::frames":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine15MemoryFrameLine6framesE"],"paddle::RecurrentGradientMachine::MemoryFrameLine::is_sequence":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine15MemoryFrameLine11is_sequenceE"],"paddle::RecurrentGradientMachine::MemoryFrameLine::layerName":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine15MemoryFrameLine9layerNameE"],"paddle::RecurrentGradientMachine::MemoryFrameLine::linkName":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine15MemoryFrameLine8linkNameE"],"paddle::RecurrentGradientMachine::MemoryFrameLine::outArg":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine15MemoryFrameLine6outArgE"],"paddle::RecurrentGradientMachine::MemoryFrameLine::rootAgent":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine15MemoryFrameLine9rootAgentE"],"paddle::RecurrentGradientMachine::MemoryFrameLine::rootLayer":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine15MemoryFrameLine9rootLayerE"],"paddle::RecurrentGradientMachine::MemoryFrameLine::scatterAgents":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine15MemoryFrameLine13scatterAgentsE"],"paddle::RecurrentGradientMachine::MemoryFrameLine::sequenceStartPositions":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine15MemoryFrameLine22sequenceStartPositionsE"],"paddle::RecurrentGradientMachine::NormOrDropNodeCallback":[40,4,1,"_CPPv2N6paddle24RecurrentGradientMachine22NormOrDropNodeCallbackE"],"paddle::RecurrentGradientMachine::OutFrameLine":[40,1,1,"_CPPv2N6paddle24RecurrentGradientMachine12OutFrameLineE"],"paddle::RecurrentGradientMachine::OutFrameLine::agentLayer":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine12OutFrameLine10agentLayerE"],"paddle::RecurrentGradientMachine::OutFrameLine::frames":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine12OutFrameLine6framesE"],"paddle::RecurrentGradientMachine::OutFrameLine::layerName":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine12OutFrameLine9layerNameE"],"paddle::RecurrentGradientMachine::Path":[40,1,1,"_CPPv2N6paddle24RecurrentGradientMachine4PathE"],"paddle::RecurrentGradientMachine::Path::Path":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine4Path4PathEv"],"paddle::RecurrentGradientMachine::Path::adjustProb":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine4Path10adjustProbEib"],"paddle::RecurrentGradientMachine::Path::greaterPath":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine4Path11greaterPathERK4PathRK4Path"],"paddle::RecurrentGradientMachine::Path::ids":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine4Path3idsE"],"paddle::RecurrentGradientMachine::Path::isDropable":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine4Path10isDropableEv"],"paddle::RecurrentGradientMachine::Path::logProb":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine4Path7logProbE"],"paddle::RecurrentGradientMachine::Path::machineId":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine4Path9machineIdE"],"paddle::RecurrentGradientMachine::Path::machineIdVec":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine4Path12machineIdVecE"],"paddle::RecurrentGradientMachine::Path::operator<":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine4PathltERK4Path"],"paddle::RecurrentGradientMachine::Path::probHistory":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine4Path11probHistoryE"],"paddle::RecurrentGradientMachine::Path::recordHistory":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine4Path13recordHistoryEv"],"paddle::RecurrentGradientMachine::Path::seqId":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine4Path5seqIdE"],"paddle::RecurrentGradientMachine::Path::topIndex":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine4Path8topIndexE"],"paddle::RecurrentGradientMachine::RecurrentGradientMachine":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine24RecurrentGradientMachineERKNSt6stringEP13NeuralNetwork"],"paddle::RecurrentGradientMachine::backward":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine8backwardERK14UpdateCallback"],"paddle::RecurrentGradientMachine::beamSearch":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine10beamSearchE6size_t"],"paddle::RecurrentGradientMachine::copyScattedId":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine13copyScattedIdERNSt6vectorIiEEP10IVectorPtri"],"paddle::RecurrentGradientMachine::cpuEos_":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine7cpuEos_E"],"paddle::RecurrentGradientMachine::cpuId_":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine6cpuId_E"],"paddle::RecurrentGradientMachine::cpuProb_":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine8cpuProb_E"],"paddle::RecurrentGradientMachine::createInFrameInfo":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine17createInFrameInfoEiRK8Argument8PassType"],"paddle::RecurrentGradientMachine::createMemoryFrameInfo":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine21createMemoryFrameInfoEP15MemoryFrameLine8PassType"],"paddle::RecurrentGradientMachine::createSeqPos":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine12createSeqPosERKNSt6vectorIiEEP16ICpuGpuVectorPtr"],"paddle::RecurrentGradientMachine::dataArgsFrame_":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine14dataArgsFrame_E"],"paddle::RecurrentGradientMachine::dataArgsSize_":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine13dataArgsSize_E"],"paddle::RecurrentGradientMachine::dataArgs_":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine9dataArgs_E"],"paddle::RecurrentGradientMachine::eosFrameLine_":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine13eosFrameLine_E"],"paddle::RecurrentGradientMachine::eval":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine4evalEP9Evaluator"],"paddle::RecurrentGradientMachine::evaluator_":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine10evaluator_E"],"paddle::RecurrentGradientMachine::forward":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine7forwardERKNSt6vectorI8ArgumentEEPNSt6vectorI8ArgumentEE8PassType"],"paddle::RecurrentGradientMachine::forwardBackward":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine15forwardBackwardERKNSt6vectorI8ArgumentEEPNSt6vectorI8ArgumentEE8PassTypeRK14UpdateCallback"],"paddle::RecurrentGradientMachine::frames_":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine7frames_E"],"paddle::RecurrentGradientMachine::generateSequence":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine16generateSequenceEv"],"paddle::RecurrentGradientMachine::generator_":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine10generator_E"],"paddle::RecurrentGradientMachine::getFinalPaths":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine13getFinalPathsEv"],"paddle::RecurrentGradientMachine::getParameterIds":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine15getParameterIdsEv"],"paddle::RecurrentGradientMachine::inFrameLines_":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine13inFrameLines_E"],"paddle::RecurrentGradientMachine::info_":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine5info_E"],"paddle::RecurrentGradientMachine::init":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine4initERK11ModelConfig17ParamInitCallbackRKNSt6vectorI13ParameterTypeEEb"],"paddle::RecurrentGradientMachine::maxSequenceLength_":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine18maxSequenceLength_E"],"paddle::RecurrentGradientMachine::memoryFrameLines_":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine17memoryFrameLines_E"],"paddle::RecurrentGradientMachine::numSeqs_":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine8numSeqs_E"],"paddle::RecurrentGradientMachine::oneWaySearch":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine12oneWaySearchE6size_t"],"paddle::RecurrentGradientMachine::operator=":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachineaSERK24RecurrentGradientMachine"],"paddle::RecurrentGradientMachine::outFrameLines_":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine14outFrameLines_E"],"paddle::RecurrentGradientMachine::parameterIds_":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine13parameterIds_E"],"paddle::RecurrentGradientMachine::prefetch":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine8prefetchERKNSt6vectorI8ArgumentEE"],"paddle::RecurrentGradientMachine::registerBeamSearchControlCallbacks":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine34registerBeamSearchControlCallbacksERK34BeamSearchCandidatesAdjustCallbackRK22NormOrDropNodeCallbackRK12DropCallback"],"paddle::RecurrentGradientMachine::registerBeamSearchStatisticsCallbacks":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine37registerBeamSearchStatisticsCallbacksERK16EachStepCallbackRK16EachStepCallback"],"paddle::RecurrentGradientMachine::removeBeamSearchControlCallbacks":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine32removeBeamSearchControlCallbacksEv"],"paddle::RecurrentGradientMachine::removeBeamSearchStatisticsCallbacks":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine35removeBeamSearchStatisticsCallbacksEv"],"paddle::RecurrentGradientMachine::resetState":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine10resetStateEv"],"paddle::RecurrentGradientMachine::resizeBootFrame":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine15resizeBootFrameEi"],"paddle::RecurrentGradientMachine::resizeOrCreateFrames":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine20resizeOrCreateFramesEi"],"paddle::RecurrentGradientMachine::reversed_":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine9reversed_E"],"paddle::RecurrentGradientMachine::rootNetwork_":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine12rootNetwork_E"],"paddle::RecurrentGradientMachine::selectRowsOneTime":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine17selectRowsOneTimeE8LayerPtrRK10IVectorPtrP8Argument8PassType"],"paddle::RecurrentGradientMachine::seqLengthAndStart_":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine18seqLengthAndStart_E"],"paddle::RecurrentGradientMachine::stopBeamSearch":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachine14stopBeamSearchEv"],"paddle::RecurrentGradientMachine::stopBeamSearch_":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine15stopBeamSearch_E"],"paddle::RecurrentGradientMachine::targetInfoInlinkId_":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine19targetInfoInlinkId_E"],"paddle::RecurrentGradientMachine::useGpu_":[40,3,1,"_CPPv2N6paddle24RecurrentGradientMachine7useGpu_E"],"paddle::RecurrentGradientMachine::~RecurrentGradientMachine":[40,2,1,"_CPPv2N6paddle24RecurrentGradientMachineD0Ev"],"paddle::RecurrentLayer":[43,1,1,"_CPPv2N6paddle14RecurrentLayerE"],"paddle::RecurrentLayer::RecurrentLayer":[43,2,1,"_CPPv2N6paddle14RecurrentLayer14RecurrentLayerERK11LayerConfig"],"paddle::RecurrentLayer::backward":[43,2,1,"_CPPv2N6paddle14RecurrentLayer8backwardERK14UpdateCallback"],"paddle::RecurrentLayer::backwardBatch":[43,2,1,"_CPPv2N6paddle14RecurrentLayer13backwardBatchEi6size_tPKi"],"paddle::RecurrentLayer::backwardOneSequence":[43,2,1,"_CPPv2N6paddle14RecurrentLayer19backwardOneSequenceEii"],"paddle::RecurrentLayer::backwardSequence":[43,2,1,"_CPPv2N6paddle14RecurrentLayer16backwardSequenceEi6size_tPKi"],"paddle::RecurrentLayer::batchGrad_":[43,3,1,"_CPPv2N6paddle14RecurrentLayer10batchGrad_E"],"paddle::RecurrentLayer::batchValue_":[43,3,1,"_CPPv2N6paddle14RecurrentLayer11batchValue_E"],"paddle::RecurrentLayer::bias_":[43,3,1,"_CPPv2N6paddle14RecurrentLayer5bias_E"],"paddle::RecurrentLayer::forward":[43,2,1,"_CPPv2N6paddle14RecurrentLayer7forwardE8PassType"],"paddle::RecurrentLayer::forwardBatch":[43,2,1,"_CPPv2N6paddle14RecurrentLayer12forwardBatchEi6size_tPKi"],"paddle::RecurrentLayer::forwardOneSequence":[43,2,1,"_CPPv2N6paddle14RecurrentLayer18forwardOneSequenceEii"],"paddle::RecurrentLayer::forwardSequence":[43,2,1,"_CPPv2N6paddle14RecurrentLayer15forwardSequenceEi6size_tPKi"],"paddle::RecurrentLayer::frameOutput_":[43,3,1,"_CPPv2N6paddle14RecurrentLayer12frameOutput_E"],"paddle::RecurrentLayer::getState":[43,2,1,"_CPPv2N6paddle14RecurrentLayer8getStateEv"],"paddle::RecurrentLayer::init":[43,2,1,"_CPPv2N6paddle14RecurrentLayer4initERK8LayerMapRK12ParameterMap"],"paddle::RecurrentLayer::prevOutput_":[43,3,1,"_CPPv2N6paddle14RecurrentLayer11prevOutput_E"],"paddle::RecurrentLayer::resetState":[43,2,1,"_CPPv2N6paddle14RecurrentLayer10resetStateEv"],"paddle::RecurrentLayer::reversed_":[43,3,1,"_CPPv2N6paddle14RecurrentLayer9reversed_E"],"paddle::RecurrentLayer::setState":[43,2,1,"_CPPv2N6paddle14RecurrentLayer8setStateE13LayerStatePtr"],"paddle::RecurrentLayer::weight_":[43,3,1,"_CPPv2N6paddle14RecurrentLayer7weight_E"],"paddle::Regularizer":[52,1,1,"_CPPv2N6paddle11RegularizerE"],"paddle::Regularizer::get":[52,2,1,"_CPPv2N6paddle11Regularizer3getERKNSt6vectorI13ParameterTypeEERK15ParameterConfig"],"paddle::Regularizer::update":[52,2,1,"_CPPv2N6paddle11Regularizer6updateEA_K9VectorPtrRK15ParameterConfig4realii"],"paddle::Regularizer::~Regularizer":[52,2,1,"_CPPv2N6paddle11RegularizerD0Ev"],"paddle::RemoteParameterUpdater":[61,1,1,"_CPPv2N6paddle22RemoteParameterUpdaterE"],"paddle::RemoteParameterUpdater::RemoteParameterUpdater":[61,2,1,"_CPPv2N6paddle22RemoteParameterUpdater22RemoteParameterUpdaterERK18OptimizationConfigiRRNSt10unique_ptrI16ParameterUpdaterEE"],"paddle::RemoteParameterUpdater::apply":[61,2,1,"_CPPv2N6paddle22RemoteParameterUpdater5applyEv"],"paddle::RemoteParameterUpdater::batchSize_":[61,3,1,"_CPPv2N6paddle22RemoteParameterUpdater10batchSize_E"],"paddle::RemoteParameterUpdater::batchStatus_":[61,3,1,"_CPPv2N6paddle22RemoteParameterUpdater12batchStatus_E"],"paddle::RemoteParameterUpdater::config_":[61,3,1,"_CPPv2N6paddle22RemoteParameterUpdater7config_E"],"paddle::RemoteParameterUpdater::controller":[61,2,1,"_CPPv2N6paddle22RemoteParameterUpdater10controllerEv"],"paddle::RemoteParameterUpdater::controllerThread_":[61,3,1,"_CPPv2N6paddle22RemoteParameterUpdater17controllerThread_E"],"paddle::RemoteParameterUpdater::copyParametersFromDevice":[61,2,1,"_CPPv2N6paddle22RemoteParameterUpdater24copyParametersFromDeviceE13ParameterType"],"paddle::RemoteParameterUpdater::copyParametersToDevice":[61,2,1,"_CPPv2N6paddle22RemoteParameterUpdater22copyParametersToDeviceE13ParameterType"],"paddle::RemoteParameterUpdater::cpuParameters_":[61,3,1,"_CPPv2N6paddle22RemoteParameterUpdater14cpuParameters_E"],"paddle::RemoteParameterUpdater::expectedPassCount_":[61,3,1,"_CPPv2N6paddle22RemoteParameterUpdater18expectedPassCount_E"],"paddle::RemoteParameterUpdater::finishBatch":[61,2,1,"_CPPv2N6paddle22RemoteParameterUpdater11finishBatchE4real"],"paddle::RemoteParameterUpdater::finishPass":[61,2,1,"_CPPv2N6paddle22RemoteParameterUpdater10finishPassE4real"],"paddle::RemoteParameterUpdater::init":[61,2,1,"_CPPv2N6paddle22RemoteParameterUpdater4initERNSt6vectorI12ParameterPtrEE"],"paddle::RemoteParameterUpdater::isFirstPass_":[61,3,1,"_CPPv2N6paddle22RemoteParameterUpdater12isFirstPass_E"],"paddle::RemoteParameterUpdater::kAverage":[61,3,1,"_CPPv2N6paddle22RemoteParameterUpdater8kAverageE"],"paddle::RemoteParameterUpdater::kElasticAverage":[61,3,1,"_CPPv2N6paddle22RemoteParameterUpdater15kElasticAverageE"],"paddle::RemoteParameterUpdater::localUpdater_":[61,3,1,"_CPPv2N6paddle22RemoteParameterUpdater13localUpdater_E"],"paddle::RemoteParameterUpdater::numBatches_":[61,3,1,"_CPPv2N6paddle22RemoteParameterUpdater11numBatches_E"],"paddle::RemoteParameterUpdater::parameterClient_":[61,3,1,"_CPPv2N6paddle22RemoteParameterUpdater16parameterClient_E"],"paddle::RemoteParameterUpdater::passCount_":[61,3,1,"_CPPv2N6paddle22RemoteParameterUpdater10passCount_E"],"paddle::RemoteParameterUpdater::restore":[61,2,1,"_CPPv2N6paddle22RemoteParameterUpdater7restoreEv"],"paddle::RemoteParameterUpdater::separateSendAndRecv_":[61,3,1,"_CPPv2N6paddle22RemoteParameterUpdater20separateSendAndRecv_E"],"paddle::RemoteParameterUpdater::setForwardbackwardTime":[61,2,1,"_CPPv2N6paddle22RemoteParameterUpdater22setForwardbackwardTimeE8uint64_t"],"paddle::RemoteParameterUpdater::startBatch":[61,2,1,"_CPPv2N6paddle22RemoteParameterUpdater10startBatchE7int64_t"],"paddle::RemoteParameterUpdater::startController":[61,2,1,"_CPPv2N6paddle22RemoteParameterUpdater15startControllerEv"],"paddle::RemoteParameterUpdater::startPass":[61,2,1,"_CPPv2N6paddle22RemoteParameterUpdater9startPassEv"],"paddle::RemoteParameterUpdater::updateImpl":[61,2,1,"_CPPv2N6paddle22RemoteParameterUpdater10updateImplEP9Parameter"],"paddle::RemoteParameterUpdater::useApplyInPserver_":[61,3,1,"_CPPv2N6paddle22RemoteParameterUpdater18useApplyInPserver_E"],"paddle::RemoteParameterUpdater::~RemoteParameterUpdater":[61,2,1,"_CPPv2N6paddle22RemoteParameterUpdaterD0Ev"],"paddle::ResizeLayer":[43,1,1,"_CPPv2N6paddle11ResizeLayerE"],"paddle::ResizeLayer::ResizeLayer":[43,2,1,"_CPPv2N6paddle11ResizeLayer11ResizeLayerERK11LayerConfig"],"paddle::ResizeLayer::backward":[43,2,1,"_CPPv2N6paddle11ResizeLayer8backwardERK14UpdateCallback"],"paddle::ResizeLayer::forward":[43,2,1,"_CPPv2N6paddle11ResizeLayer7forwardE8PassType"],"paddle::ResizeLayer::init":[43,2,1,"_CPPv2N6paddle11ResizeLayer4initERK8LayerMapRK12ParameterMap"],"paddle::ResponseNormLayer":[43,1,1,"_CPPv2N6paddle17ResponseNormLayerE"],"paddle::ResponseNormLayer::ResponseNormLayer":[43,2,1,"_CPPv2N6paddle17ResponseNormLayer17ResponseNormLayerERK11LayerConfig"],"paddle::ResponseNormLayer::backward":[43,2,1,"_CPPv2N6paddle17ResponseNormLayer8backwardERK14UpdateCallback"],"paddle::ResponseNormLayer::channels_":[43,3,1,"_CPPv2N6paddle17ResponseNormLayer9channels_E"],"paddle::ResponseNormLayer::denoms_":[43,3,1,"_CPPv2N6paddle17ResponseNormLayer7denoms_E"],"paddle::ResponseNormLayer::forward":[43,2,1,"_CPPv2N6paddle17ResponseNormLayer7forwardE8PassType"],"paddle::ResponseNormLayer::imgSize_":[43,3,1,"_CPPv2N6paddle17ResponseNormLayer8imgSize_E"],"paddle::ResponseNormLayer::init":[43,2,1,"_CPPv2N6paddle17ResponseNormLayer4initERK8LayerMapRK12ParameterMap"],"paddle::ResponseNormLayer::outputX_":[43,3,1,"_CPPv2N6paddle17ResponseNormLayer8outputX_E"],"paddle::ResponseNormLayer::pow_":[43,3,1,"_CPPv2N6paddle17ResponseNormLayer4pow_E"],"paddle::ResponseNormLayer::scale_":[43,3,1,"_CPPv2N6paddle17ResponseNormLayer6scale_E"],"paddle::ResponseNormLayer::size_":[43,3,1,"_CPPv2N6paddle17ResponseNormLayer5size_E"],"paddle::SPARSE_CSC":[46,1,1,"_CPPv2N6paddle10SPARSE_CSCE"],"paddle::SPARSE_CSR":[46,1,1,"_CPPv2N6paddle10SPARSE_CSRE"],"paddle::SVectorPtr":[52,4,1,"_CPPv2N6paddle10SVectorPtrE"],"paddle::SYNCED":[46,1,1,"_CPPv2N6paddle6SYNCEDE"],"paddle::SamplingIdLayer":[43,1,1,"_CPPv2N6paddle15SamplingIdLayerE"],"paddle::SamplingIdLayer::SamplingIdLayer":[43,2,1,"_CPPv2N6paddle15SamplingIdLayer15SamplingIdLayerERK11LayerConfig"],"paddle::SamplingIdLayer::backward":[43,2,1,"_CPPv2N6paddle15SamplingIdLayer8backwardERK14UpdateCallback"],"paddle::SamplingIdLayer::forward":[43,2,1,"_CPPv2N6paddle15SamplingIdLayer7forwardE8PassType"],"paddle::SamplingIdLayer::forwardImp":[43,2,1,"_CPPv2N6paddle15SamplingIdLayer10forwardImpERK8Argument"],"paddle::SamplingIdLayer::init":[43,2,1,"_CPPv2N6paddle15SamplingIdLayer4initERK8LayerMapRK12ParameterMap"],"paddle::ScalingLayer":[43,1,1,"_CPPv2N6paddle12ScalingLayerE"],"paddle::ScalingLayer::ScalingLayer":[43,2,1,"_CPPv2N6paddle12ScalingLayer12ScalingLayerERK11LayerConfig"],"paddle::ScalingLayer::backward":[43,2,1,"_CPPv2N6paddle12ScalingLayer8backwardERK14UpdateCallback"],"paddle::ScalingLayer::forward":[43,2,1,"_CPPv2N6paddle12ScalingLayer7forwardE8PassType"],"paddle::ScalingLayer::init":[43,2,1,"_CPPv2N6paddle12ScalingLayer4initERK8LayerMapRK12ParameterMap"],"paddle::ScalingLayer::~ScalingLayer":[43,2,1,"_CPPv2N6paddle12ScalingLayerD0Ev"],"paddle::ScatterAgentLayer":[43,1,1,"_CPPv2N6paddle17ScatterAgentLayerE"],"paddle::ScatterAgentLayer::ScatterAgentLayer":[43,2,1,"_CPPv2N6paddle17ScatterAgentLayer17ScatterAgentLayerERK11LayerConfig"],"paddle::ScatterAgentLayer::backward":[43,2,1,"_CPPv2N6paddle17ScatterAgentLayer8backwardERK14UpdateCallback"],"paddle::ScatterAgentLayer::cpuIds_":[43,3,1,"_CPPv2N6paddle17ScatterAgentLayer7cpuIds_E"],"paddle::ScatterAgentLayer::forward":[43,2,1,"_CPPv2N6paddle17ScatterAgentLayer7forwardE8PassType"],"paddle::ScatterAgentLayer::idIndex_":[43,3,1,"_CPPv2N6paddle17ScatterAgentLayer8idIndex_E"],"paddle::ScatterAgentLayer::idSize_":[43,3,1,"_CPPv2N6paddle17ScatterAgentLayer7idSize_E"],"paddle::ScatterAgentLayer::ids_":[43,3,1,"_CPPv2N6paddle17ScatterAgentLayer4ids_E"],"paddle::ScatterAgentLayer::init":[43,2,1,"_CPPv2N6paddle17ScatterAgentLayer4initERK8LayerMapRK12ParameterMap"],"paddle::ScatterAgentLayer::numSequences_":[43,3,1,"_CPPv2N6paddle17ScatterAgentLayer13numSequences_E"],"paddle::ScatterAgentLayer::realLayer_":[43,3,1,"_CPPv2N6paddle17ScatterAgentLayer10realLayer_E"],"paddle::ScatterAgentLayer::realOutArg_":[43,3,1,"_CPPv2N6paddle17ScatterAgentLayer11realOutArg_E"],"paddle::ScatterAgentLayer::seqStartPosIndex_":[43,3,1,"_CPPv2N6paddle17ScatterAgentLayer17seqStartPosIndex_E"],"paddle::ScatterAgentLayer::setRealLayer":[43,2,1,"_CPPv2N6paddle17ScatterAgentLayer12setRealLayerE8LayerPtrRKNSt6vectorIiEEb"],"paddle::ScatterAgentLayer::setRealLayerAndOutput":[43,2,1,"_CPPv2N6paddle17ScatterAgentLayer21setRealLayerAndOutputE8LayerPtrRK8ArgumentRK10IVectorPtrii"],"paddle::ScatterAgentLayer::setSequenceStartPositions":[43,2,1,"_CPPv2N6paddle17ScatterAgentLayer25setSequenceStartPositionsERK16ICpuGpuVectorPtrii"],"paddle::ScatterAgentLayer::~ScatterAgentLayer":[43,2,1,"_CPPv2N6paddle17ScatterAgentLayerD0Ev"],"paddle::Segment":[52,1,1,"_CPPv2N6paddle7SegmentE"],"paddle::Segment::beginDim":[52,3,1,"_CPPv2N6paddle7Segment8beginDimE"],"paddle::Segment::beginPos":[52,3,1,"_CPPv2N6paddle7Segment8beginPosE"],"paddle::Segment::endDim":[52,3,1,"_CPPv2N6paddle7Segment6endDimE"],"paddle::SelectiveFullyConnectedLayer":[43,1,1,"_CPPv2N6paddle28SelectiveFullyConnectedLayerE"],"paddle::SelectiveFullyConnectedLayer::SelectiveFullyConnectedLayer":[43,2,1,"_CPPv2N6paddle28SelectiveFullyConnectedLayer28SelectiveFullyConnectedLayerERK11LayerConfig"],"paddle::SelectiveFullyConnectedLayer::backward":[43,2,1,"_CPPv2N6paddle28SelectiveFullyConnectedLayer8backwardERK14UpdateCallback"],"paddle::SelectiveFullyConnectedLayer::biases_":[43,3,1,"_CPPv2N6paddle28SelectiveFullyConnectedLayer7biases_E"],"paddle::SelectiveFullyConnectedLayer::fillSelectiveData":[43,2,1,"_CPPv2N6paddle28SelectiveFullyConnectedLayer17fillSelectiveDataERKNSt10shared_ptrINSt6vectorINSt4pairIPi6size_tEEEEEE"],"paddle::SelectiveFullyConnectedLayer::forward":[43,2,1,"_CPPv2N6paddle28SelectiveFullyConnectedLayer7forwardE8PassType"],"paddle::SelectiveFullyConnectedLayer::getWeight":[43,2,1,"_CPPv2N6paddle28SelectiveFullyConnectedLayer9getWeightEi"],"paddle::SelectiveFullyConnectedLayer::init":[43,2,1,"_CPPv2N6paddle28SelectiveFullyConnectedLayer4initERK8LayerMapRK12ParameterMap"],"paddle::SelectiveFullyConnectedLayer::prefetch":[43,2,1,"_CPPv2N6paddle28SelectiveFullyConnectedLayer8prefetchEv"],"paddle::SelectiveFullyConnectedLayer::reserveOutput":[43,2,1,"_CPPv2N6paddle28SelectiveFullyConnectedLayer13reserveOutputE6size_t6size_t6size_t"],"paddle::SelectiveFullyConnectedLayer::weights_":[43,3,1,"_CPPv2N6paddle28SelectiveFullyConnectedLayer8weights_E"],"paddle::SelectiveFullyConnectedLayer::~SelectiveFullyConnectedLayer":[43,2,1,"_CPPv2N6paddle28SelectiveFullyConnectedLayerD0Ev"],"paddle::Semaphore":[66,1,1,"_CPPv2N6paddle9SemaphoreE"],"paddle::Semaphore::Semaphore":[66,2,1,"_CPPv2N6paddle9Semaphore9SemaphoreEi"],"paddle::Semaphore::post":[66,2,1,"_CPPv2N6paddle9Semaphore4postEv"],"paddle::Semaphore::sem_":[66,3,1,"_CPPv2N6paddle9Semaphore4sem_E"],"paddle::Semaphore::timeWait":[66,2,1,"_CPPv2N6paddle9Semaphore8timeWaitEP8timespec"],"paddle::Semaphore::wait":[66,2,1,"_CPPv2N6paddle9Semaphore4waitEv"],"paddle::Semaphore::~Semaphore":[66,2,1,"_CPPv2N6paddle9SemaphoreD0Ev"],"paddle::SequenceAgentLayer":[43,1,1,"_CPPv2N6paddle18SequenceAgentLayerE"],"paddle::SequenceAgentLayer::SequenceAgentLayer":[43,2,1,"_CPPv2N6paddle18SequenceAgentLayer18SequenceAgentLayerERK11LayerConfig"],"paddle::SequenceAgentLayer::backward":[43,2,1,"_CPPv2N6paddle18SequenceAgentLayer8backwardERK14UpdateCallback"],"paddle::SequenceAgentLayer::forward":[43,2,1,"_CPPv2N6paddle18SequenceAgentLayer7forwardE8PassType"],"paddle::SequenceAgentLayer::~SequenceAgentLayer":[43,2,1,"_CPPv2N6paddle18SequenceAgentLayerD0Ev"],"paddle::SequenceClassificationErrorEvaluator":[38,1,1,"_CPPv2N6paddle36SequenceClassificationErrorEvaluatorE"],"paddle::SequenceClassificationErrorEvaluator::distributeEval":[38,2,1,"_CPPv2N6paddle36SequenceClassificationErrorEvaluator14distributeEvalEP16ParameterClient2"],"paddle::SequenceClassificationErrorEvaluator::evalImp":[38,2,1,"_CPPv2N6paddle36SequenceClassificationErrorEvaluator7evalImpERNSt6vectorI8ArgumentEE"],"paddle::SequenceClassificationErrorEvaluator::updateSamplesNum":[38,2,1,"_CPPv2N6paddle36SequenceClassificationErrorEvaluator16updateSamplesNumERKNSt6vectorI8ArgumentEE"],"paddle::SequenceConcatLayer":[43,1,1,"_CPPv2N6paddle19SequenceConcatLayerE"],"paddle::SequenceConcatLayer::SequenceConcatLayer":[43,2,1,"_CPPv2N6paddle19SequenceConcatLayer19SequenceConcatLayerERK11LayerConfig"],"paddle::SequenceConcatLayer::backward":[43,2,1,"_CPPv2N6paddle19SequenceConcatLayer8backwardERK14UpdateCallback"],"paddle::SequenceConcatLayer::biases_":[43,3,1,"_CPPv2N6paddle19SequenceConcatLayer7biases_E"],"paddle::SequenceConcatLayer::forward":[43,2,1,"_CPPv2N6paddle19SequenceConcatLayer7forwardE8PassType"],"paddle::SequenceConcatLayer::init":[43,2,1,"_CPPv2N6paddle19SequenceConcatLayer4initERK8LayerMapRK12ParameterMap"],"paddle::SequenceConcatLayer::~SequenceConcatLayer":[43,2,1,"_CPPv2N6paddle19SequenceConcatLayerD0Ev"],"paddle::SequenceGatherAgentLayer":[43,1,1,"_CPPv2N6paddle24SequenceGatherAgentLayerE"],"paddle::SequenceGatherAgentLayer::SequenceGatherAgentLayer":[43,2,1,"_CPPv2N6paddle24SequenceGatherAgentLayer24SequenceGatherAgentLayerERK11LayerConfig"],"paddle::SequenceGatherAgentLayer::backward":[43,2,1,"_CPPv2N6paddle24SequenceGatherAgentLayer8backwardERK14UpdateCallback"],"paddle::SequenceGatherAgentLayer::forward":[43,2,1,"_CPPv2N6paddle24SequenceGatherAgentLayer7forwardE8PassType"],"paddle::SequenceGatherAgentLayer::~SequenceGatherAgentLayer":[43,2,1,"_CPPv2N6paddle24SequenceGatherAgentLayerD0Ev"],"paddle::SequenceLastInstanceLayer":[43,1,1,"_CPPv2N6paddle25SequenceLastInstanceLayerE"],"paddle::SequenceLastInstanceLayer::SequenceLastInstanceLayer":[43,2,1,"_CPPv2N6paddle25SequenceLastInstanceLayer25SequenceLastInstanceLayerERK11LayerConfig"],"paddle::SequenceLastInstanceLayer::SequenceLevel":[43,4,1,"_CPPv2N6paddle25SequenceLastInstanceLayer13SequenceLevelE"],"paddle::SequenceLastInstanceLayer::backward":[43,2,1,"_CPPv2N6paddle25SequenceLastInstanceLayer8backwardERK14UpdateCallback"],"paddle::SequenceLastInstanceLayer::biases_":[43,3,1,"_CPPv2N6paddle25SequenceLastInstanceLayer7biases_E"],"paddle::SequenceLastInstanceLayer::forward":[43,2,1,"_CPPv2N6paddle25SequenceLastInstanceLayer7forwardE8PassType"],"paddle::SequenceLastInstanceLayer::init":[43,2,1,"_CPPv2N6paddle25SequenceLastInstanceLayer4initERK8LayerMapRK12ParameterMap"],"paddle::SequenceLastInstanceLayer::kNonSeq":[43,1,1,"_CPPv2N6paddle25SequenceLastInstanceLayer7kNonSeqE"],"paddle::SequenceLastInstanceLayer::kSeq":[43,1,1,"_CPPv2N6paddle25SequenceLastInstanceLayer4kSeqE"],"paddle::SequenceLastInstanceLayer::tmpDest_":[43,3,1,"_CPPv2N6paddle25SequenceLastInstanceLayer8tmpDest_E"],"paddle::SequenceLastInstanceLayer::tmpSrc_":[43,3,1,"_CPPv2N6paddle25SequenceLastInstanceLayer7tmpSrc_E"],"paddle::SequenceLastInstanceLayer::type_":[43,3,1,"_CPPv2N6paddle25SequenceLastInstanceLayer5type_E"],"paddle::SequenceLastInstanceLayer::~SequenceLastInstanceLayer":[43,2,1,"_CPPv2N6paddle25SequenceLastInstanceLayerD0Ev"],"paddle::SequenceReshapeLayer":[43,1,1,"_CPPv2N6paddle20SequenceReshapeLayerE"],"paddle::SequenceReshapeLayer::SequenceReshapeLayer":[43,2,1,"_CPPv2N6paddle20SequenceReshapeLayer20SequenceReshapeLayerERK11LayerConfig"],"paddle::SequenceReshapeLayer::backward":[43,2,1,"_CPPv2N6paddle20SequenceReshapeLayer8backwardERK14UpdateCallback"],"paddle::SequenceReshapeLayer::biases_":[43,3,1,"_CPPv2N6paddle20SequenceReshapeLayer7biases_E"],"paddle::SequenceReshapeLayer::forward":[43,2,1,"_CPPv2N6paddle20SequenceReshapeLayer7forwardE8PassType"],"paddle::SequenceReshapeLayer::init":[43,2,1,"_CPPv2N6paddle20SequenceReshapeLayer4initERK8LayerMapRK12ParameterMap"],"paddle::SequenceReshapeLayer::reshapedOutputGrad":[43,3,1,"_CPPv2N6paddle20SequenceReshapeLayer18reshapedOutputGradE"],"paddle::SequenceReshapeLayer::~SequenceReshapeLayer":[43,2,1,"_CPPv2N6paddle20SequenceReshapeLayerD0Ev"],"paddle::SequenceScatterAgentLayer":[43,1,1,"_CPPv2N6paddle25SequenceScatterAgentLayerE"],"paddle::SequenceScatterAgentLayer::SequenceScatterAgentLayer":[43,2,1,"_CPPv2N6paddle25SequenceScatterAgentLayer25SequenceScatterAgentLayerERK11LayerConfig"],"paddle::SequenceScatterAgentLayer::backward":[43,2,1,"_CPPv2N6paddle25SequenceScatterAgentLayer8backwardERK14UpdateCallback"],"paddle::SequenceScatterAgentLayer::cpuInputStartPos_":[43,3,1,"_CPPv2N6paddle25SequenceScatterAgentLayer17cpuInputStartPos_E"],"paddle::SequenceScatterAgentLayer::forward":[43,2,1,"_CPPv2N6paddle25SequenceScatterAgentLayer7forwardE8PassType"],"paddle::SequenceScatterAgentLayer::inputStartPos_":[43,3,1,"_CPPv2N6paddle25SequenceScatterAgentLayer14inputStartPos_E"],"paddle::SequenceScatterAgentLayer::~SequenceScatterAgentLayer":[43,2,1,"_CPPv2N6paddle25SequenceScatterAgentLayerD0Ev"],"paddle::SequenceTextPrinter":[38,1,1,"_CPPv2N6paddle19SequenceTextPrinterE"],"paddle::SequenceTextPrinter::SequenceTextPrinter":[38,2,1,"_CPPv2N6paddle19SequenceTextPrinter19SequenceTextPrinterEv"],"paddle::SequenceTextPrinter::evalImp":[38,2,1,"_CPPv2N6paddle19SequenceTextPrinter7evalImpERNSt6vectorI8ArgumentEE"],"paddle::SequenceTextPrinter::init":[38,2,1,"_CPPv2N6paddle19SequenceTextPrinter4initERK15EvaluatorConfig"],"paddle::SequenceTextPrinter::updateSamplesNum":[38,2,1,"_CPPv2N6paddle19SequenceTextPrinter16updateSamplesNumERKNSt6vectorI8ArgumentEE"],"paddle::SequenceToBatch":[43,1,1,"_CPPv2N6paddle15SequenceToBatchE"],"paddle::SequenceToBatch::SequenceToBatch":[43,2,1,"_CPPv2N6paddle15SequenceToBatch15SequenceToBatchEb"],"paddle::SequenceToBatch::add":[43,2,1,"_CPPv2N6paddle15SequenceToBatch3addER6MatrixR6Matrixb"],"paddle::SequenceToBatch::batchStartPositions_":[43,3,1,"_CPPv2N6paddle15SequenceToBatch20batchStartPositions_E"],"paddle::SequenceToBatch::batchValue_":[43,3,1,"_CPPv2N6paddle15SequenceToBatch11batchValue_E"],"paddle::SequenceToBatch::copy":[43,2,1,"_CPPv2N6paddle15SequenceToBatch4copyER6MatrixR6Matrixb"],"paddle::SequenceToBatch::copyBackSeq":[43,2,1,"_CPPv2N6paddle15SequenceToBatch11copyBackSeqER6Matrix"],"paddle::SequenceToBatch::copyFromSeq":[43,2,1,"_CPPv2N6paddle15SequenceToBatch11copyFromSeqER6Matrix"],"paddle::SequenceToBatch::cpuSeq2BatchIdx_":[43,3,1,"_CPPv2N6paddle15SequenceToBatch16cpuSeq2BatchIdx_E"],"paddle::SequenceToBatch::cpuSeqEndIdxInBatch_":[43,3,1,"_CPPv2N6paddle15SequenceToBatch20cpuSeqEndIdxInBatch_E"],"paddle::SequenceToBatch::cpuSeqIdx_":[43,3,1,"_CPPv2N6paddle15SequenceToBatch10cpuSeqIdx_E"],"paddle::SequenceToBatch::getBatchValue":[43,2,1,"_CPPv2N6paddle15SequenceToBatch13getBatchValueEv"],"paddle::SequenceToBatch::getNumBatch":[43,2,1,"_CPPv2N6paddle15SequenceToBatch11getNumBatchEv"],"paddle::SequenceToBatch::getSeqOutputFromBatch":[43,2,1,"_CPPv2N6paddle15SequenceToBatch21getSeqOutputFromBatchER6MatrixR6Matrix"],"paddle::SequenceToBatch::numBatch_":[43,3,1,"_CPPv2N6paddle15SequenceToBatch9numBatch_E"],"paddle::SequenceToBatch::prevOutput2Batch":[43,2,1,"_CPPv2N6paddle15SequenceToBatch16prevOutput2BatchER6MatrixR6Matrix"],"paddle::SequenceToBatch::resizeOrCreate":[43,2,1,"_CPPv2N6paddle15SequenceToBatch14resizeOrCreateER6Matrix"],"paddle::SequenceToBatch::resizeOrCreateBatch":[43,2,1,"_CPPv2N6paddle15SequenceToBatch19resizeOrCreateBatchEi6size_tPKibb"],"paddle::SequenceToBatch::seq2BatchIdx_":[43,3,1,"_CPPv2N6paddle15SequenceToBatch13seq2BatchIdx_E"],"paddle::SequenceToBatch::seqEndIdxInBatch_":[43,3,1,"_CPPv2N6paddle15SequenceToBatch17seqEndIdxInBatch_E"],"paddle::SequenceToBatch::seqIdx_":[43,3,1,"_CPPv2N6paddle15SequenceToBatch7seqIdx_E"],"paddle::SequenceToBatch::sequence2BatchAdd":[43,2,1,"_CPPv2N6paddle15SequenceToBatch17sequence2BatchAddER6MatrixR6MatrixR7IVectorb"],"paddle::SequenceToBatch::sequence2BatchCopy":[43,2,1,"_CPPv2N6paddle15SequenceToBatch18sequence2BatchCopyER6MatrixR6MatrixR7IVectorb"],"paddle::SequenceToBatch::shareIndexWith":[43,2,1,"_CPPv2N6paddle15SequenceToBatch14shareIndexWithERK15SequenceToBatch"],"paddle::SequenceToBatch::useGpu_":[43,3,1,"_CPPv2N6paddle15SequenceToBatch7useGpu_E"],"paddle::SgdOptimizer":[50,1,1,"_CPPv2N6paddle12SgdOptimizerE"],"paddle::SgdOptimizer::SgdOptimizer":[50,2,1,"_CPPv2N6paddle12SgdOptimizer12SgdOptimizerERK18OptimizationConfig"],"paddle::SgdOptimizer::finishBatch":[50,2,1,"_CPPv2N6paddle12SgdOptimizer11finishBatchEv"],"paddle::SgdOptimizer::startBatch":[50,2,1,"_CPPv2N6paddle12SgdOptimizer10startBatchE7int64_t"],"paddle::SgdOptimizer::update":[50,2,1,"_CPPv2N6paddle12SgdOptimizer6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::SharedCpuMatrix":[46,1,1,"_CPPv2N6paddle15SharedCpuMatrixE"],"paddle::SharedCpuMatrix::SharedCpuMatrix":[46,2,1,"_CPPv2N6paddle15SharedCpuMatrix15SharedCpuMatrixEiP4real6size_t6size_tb"],"paddle::SharedCpuMatrix::add":[46,2,1,"_CPPv2N6paddle15SharedCpuMatrix3addER6Matrix4real4real"],"paddle::SharedCpuMatrix::blockLocks_":[46,3,1,"_CPPv2N6paddle15SharedCpuMatrix11blockLocks_E"],"paddle::SharedCpuMatrix::blockNum_":[46,3,1,"_CPPv2N6paddle15SharedCpuMatrix9blockNum_E"],"paddle::SharedCpuMatrix::blockSeq_":[46,3,1,"_CPPv2N6paddle15SharedCpuMatrix9blockSeq_E"],"paddle::SharedCpuMatrix::initBlock":[46,2,1,"_CPPv2N6paddle15SharedCpuMatrix9initBlockEi"],"paddle::SharedCpuMatrix::initShared":[46,2,1,"_CPPv2N6paddle15SharedCpuMatrix10initSharedEi"],"paddle::SharedCpuMatrix::localBufRows_":[46,3,1,"_CPPv2N6paddle15SharedCpuMatrix13localBufRows_E"],"paddle::SharedCpuMatrix::localBuf_":[46,3,1,"_CPPv2N6paddle15SharedCpuMatrix9localBuf_E"],"paddle::SharedCpuMatrix::mul":[46,2,1,"_CPPv2N6paddle15SharedCpuMatrix3mulEP15CpuSparseMatrixP9CpuMatrix4real4real"],"paddle::SharedCpuMatrix::~SharedCpuMatrix":[46,2,1,"_CPPv2N6paddle15SharedCpuMatrixD0Ev"],"paddle::SlopeInterceptLayer":[43,1,1,"_CPPv2N6paddle19SlopeInterceptLayerE"],"paddle::SlopeInterceptLayer::SlopeInterceptLayer":[43,2,1,"_CPPv2N6paddle19SlopeInterceptLayer19SlopeInterceptLayerERK11LayerConfig"],"paddle::SlopeInterceptLayer::backward":[43,2,1,"_CPPv2N6paddle19SlopeInterceptLayer8backwardERK14UpdateCallback"],"paddle::SlopeInterceptLayer::forward":[43,2,1,"_CPPv2N6paddle19SlopeInterceptLayer7forwardE8PassType"],"paddle::SlopeInterceptLayer::init":[43,2,1,"_CPPv2N6paddle19SlopeInterceptLayer4initERK8LayerMapRK12ParameterMap"],"paddle::SlopeInterceptLayer::~SlopeInterceptLayer":[43,2,1,"_CPPv2N6paddle19SlopeInterceptLayerD0Ev"],"paddle::SocketChannel":[58,1,1,"_CPPv2N6paddle13SocketChannelE"],"paddle::SocketChannel::MessageHeader":[58,1,1,"_CPPv2N6paddle13SocketChannel13MessageHeaderE"],"paddle::SocketChannel::MessageHeader::iovLengths":[58,3,1,"_CPPv2N6paddle13SocketChannel13MessageHeader10iovLengthsE"],"paddle::SocketChannel::MessageHeader::numIovs":[58,3,1,"_CPPv2N6paddle13SocketChannel13MessageHeader7numIovsE"],"paddle::SocketChannel::MessageHeader::totalLength":[58,3,1,"_CPPv2N6paddle13SocketChannel13MessageHeader11totalLengthE"],"paddle::SocketChannel::SocketChannel":[58,2,1,"_CPPv2N6paddle13SocketChannel13SocketChannelEiRKNSt6stringE"],"paddle::SocketChannel::getPeerName":[58,2,1,"_CPPv2N6paddle13SocketChannel11getPeerNameEv"],"paddle::SocketChannel::peerName_":[58,3,1,"_CPPv2N6paddle13SocketChannel9peerName_E"],"paddle::SocketChannel::rdmaSocket_":[58,3,1,"_CPPv2N6paddle13SocketChannel11rdmaSocket_E"],"paddle::SocketChannel::read":[58,2,1,"_CPPv2N6paddle13SocketChannel4readEPv6size_t"],"paddle::SocketChannel::readMessage":[58,2,1,"_CPPv2N6paddle13SocketChannel11readMessageEv"],"paddle::SocketChannel::readv":[58,2,1,"_CPPv2N6paddle13SocketChannel5readvEPNSt6vectorI5iovecEE"],"paddle::SocketChannel::tcpRdma_":[58,3,1,"_CPPv2N6paddle13SocketChannel8tcpRdma_E"],"paddle::SocketChannel::tcpSocket_":[58,3,1,"_CPPv2N6paddle13SocketChannel10tcpSocket_E"],"paddle::SocketChannel::write":[58,2,1,"_CPPv2N6paddle13SocketChannel5writeEPKv6size_t"],"paddle::SocketChannel::writeMessage":[58,2,1,"_CPPv2N6paddle13SocketChannel12writeMessageERKNSt6vectorI5iovecEE"],"paddle::SocketChannel::writev":[58,2,1,"_CPPv2N6paddle13SocketChannel6writevERKNSt6vectorI5iovecEE"],"paddle::SocketChannel::~SocketChannel":[58,2,1,"_CPPv2N6paddle13SocketChannelD0Ev"],"paddle::SocketClient":[58,1,1,"_CPPv2N6paddle12SocketClientE"],"paddle::SocketClient::RdmaClient":[58,2,1,"_CPPv2N6paddle12SocketClient10RdmaClientERKNSt6stringEi"],"paddle::SocketClient::SocketClient":[58,2,1,"_CPPv2N6paddle12SocketClient12SocketClientERKNSt6stringEi11ChannelType"],"paddle::SocketClient::TcpClient":[58,2,1,"_CPPv2N6paddle12SocketClient9TcpClientERKNSt6stringEi"],"paddle::SocketClient::channel_":[58,3,1,"_CPPv2N6paddle12SocketClient8channel_E"],"paddle::SocketClient::getChannel":[58,2,1,"_CPPv2N6paddle12SocketClient10getChannelEv"],"paddle::SocketClient::socketDaemon_":[58,3,1,"_CPPv2N6paddle12SocketClient13socketDaemon_E"],"paddle::SocketClient::tcpRdma_":[58,3,1,"_CPPv2N6paddle12SocketClient8tcpRdma_E"],"paddle::SocketServer":[58,1,1,"_CPPv2N6paddle12SocketServerE"],"paddle::SocketServer::ResponseCallback":[58,4,1,"_CPPv2N6paddle12SocketServer16ResponseCallbackE"],"paddle::SocketServer::SocketServer":[58,2,1,"_CPPv2N6paddle12SocketServer12SocketServerERKNSt6stringEii"],"paddle::SocketServer::addr_":[58,3,1,"_CPPv2N6paddle12SocketServer5addr_E"],"paddle::SocketServer::createChannel":[58,2,1,"_CPPv2N6paddle12SocketServer13createChannelEiRKNSt6stringE"],"paddle::SocketServer::detach":[58,2,1,"_CPPv2N6paddle12SocketServer6detachEv"],"paddle::SocketServer::handleRequest":[58,2,1,"_CPPv2N6paddle12SocketServer13handleRequestENSt10unique_ptrI9MsgReaderEE16ResponseCallback"],"paddle::SocketServer::maxPendingConnections_":[58,3,1,"_CPPv2N6paddle12SocketServer22maxPendingConnections_E"],"paddle::SocketServer::port_":[58,3,1,"_CPPv2N6paddle12SocketServer5port_E"],"paddle::SocketServer::rdmaCpu_":[58,3,1,"_CPPv2N6paddle12SocketServer8rdmaCpu_E"],"paddle::SocketServer::rdmaServer":[58,2,1,"_CPPv2N6paddle12SocketServer10rdmaServerEv"],"paddle::SocketServer::rdmaSocket_":[58,3,1,"_CPPv2N6paddle12SocketServer11rdmaSocket_E"],"paddle::SocketServer::rdmaUri_":[58,3,1,"_CPPv2N6paddle12SocketServer8rdmaUri_E"],"paddle::SocketServer::run":[58,2,1,"_CPPv2N6paddle12SocketServer3runEv"],"paddle::SocketServer::socket_":[58,3,1,"_CPPv2N6paddle12SocketServer7socket_E"],"paddle::SocketServer::stopping_":[58,3,1,"_CPPv2N6paddle12SocketServer9stopping_E"],"paddle::SocketServer::tcpRdma_":[58,3,1,"_CPPv2N6paddle12SocketServer8tcpRdma_E"],"paddle::SocketServer::tcpServer":[58,2,1,"_CPPv2N6paddle12SocketServer9tcpServerEv"],"paddle::SocketServer::~SocketServer":[58,2,1,"_CPPv2N6paddle12SocketServerD0Ev"],"paddle::SocketWorker":[58,1,1,"_CPPv2N6paddle12SocketWorkerE"],"paddle::SocketWorker::SocketWorker":[58,2,1,"_CPPv2N6paddle12SocketWorker12SocketWorkerERRNSt10unique_ptrI13SocketChannelEEP12SocketServer"],"paddle::SocketWorker::channel_":[58,3,1,"_CPPv2N6paddle12SocketWorker8channel_E"],"paddle::SocketWorker::run":[58,2,1,"_CPPv2N6paddle12SocketWorker3runEv"],"paddle::SocketWorker::server_":[58,3,1,"_CPPv2N6paddle12SocketWorker7server_E"],"paddle::SocketWorker::tcpRdma_":[58,3,1,"_CPPv2N6paddle12SocketWorker8tcpRdma_E"],"paddle::SocketWorker::~SocketWorker":[58,2,1,"_CPPv2N6paddle12SocketWorkerD0Ev"],"paddle::SoftBinaryClassCrossEntropy":[43,1,1,"_CPPv2N6paddle27SoftBinaryClassCrossEntropyE"],"paddle::SoftBinaryClassCrossEntropy::SoftBinaryClassCrossEntropy":[43,2,1,"_CPPv2N6paddle27SoftBinaryClassCrossEntropy27SoftBinaryClassCrossEntropyERK11LayerConfig"],"paddle::SoftBinaryClassCrossEntropy::backwardImp":[43,2,1,"_CPPv2N6paddle27SoftBinaryClassCrossEntropy11backwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::SoftBinaryClassCrossEntropy::forwardImp":[43,2,1,"_CPPv2N6paddle27SoftBinaryClassCrossEntropy10forwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::SoftBinaryClassCrossEntropy::init":[43,2,1,"_CPPv2N6paddle27SoftBinaryClassCrossEntropy4initERK8LayerMapRK12ParameterMap"],"paddle::SoftBinaryClassCrossEntropy::targetPerDim_":[43,3,1,"_CPPv2N6paddle27SoftBinaryClassCrossEntropy13targetPerDim_E"],"paddle::SparseAutoGrowRowCpuMatrix":[46,1,1,"_CPPv2N6paddle26SparseAutoGrowRowCpuMatrixE"],"paddle::SparseAutoGrowRowCpuMatrix::SparseAutoGrowRowCpuMatrix":[46,2,1,"_CPPv2N6paddle26SparseAutoGrowRowCpuMatrix26SparseAutoGrowRowCpuMatrixE6size_t6size_t12IndexDictPtrb"],"paddle::SparseAutoGrowRowCpuMatrix::getRow":[46,2,1,"_CPPv2N6paddle26SparseAutoGrowRowCpuMatrix6getRowE6size_t"],"paddle::SparseAutoGrowRowCpuMatrix::getRowBuf":[46,2,1,"_CPPv2N6paddle26SparseAutoGrowRowCpuMatrix9getRowBufE6size_t"],"paddle::SparseAutoGrowRowCpuMatrix::mul":[46,2,1,"_CPPv2N6paddle26SparseAutoGrowRowCpuMatrix3mulEP15CpuSparseMatrixP9CpuMatrix4real4real"],"paddle::SparseFormat":[46,4,1,"_CPPv2N6paddle12SparseFormatE"],"paddle::SparseMomentumParameterOptimizer":[50,1,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizerE"],"paddle::SparseMomentumParameterOptimizer::SparseMomentumParameterOptimizer":[50,2,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer32SparseMomentumParameterOptimizerERK18OptimizationConfig"],"paddle::SparseMomentumParameterOptimizer::alpha_":[50,3,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer6alpha_E"],"paddle::SparseMomentumParameterOptimizer::beta_":[50,3,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer5beta_E"],"paddle::SparseMomentumParameterOptimizer::decayRate_":[50,3,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer10decayRate_E"],"paddle::SparseMomentumParameterOptimizer::finishBatch":[50,2,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer11finishBatchEv"],"paddle::SparseMomentumParameterOptimizer::gamma_":[50,3,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer6gamma_E"],"paddle::SparseMomentumParameterOptimizer::init":[50,2,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer4initE6size_tPK15ParameterConfig"],"paddle::SparseMomentumParameterOptimizer::isParameterSparse_":[50,3,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer18isParameterSparse_E"],"paddle::SparseMomentumParameterOptimizer::momentum_":[50,3,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer9momentum_E"],"paddle::SparseMomentumParameterOptimizer::needSpecialTraversal":[50,2,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer20needSpecialTraversalERK15ParameterConfig"],"paddle::SparseMomentumParameterOptimizer::startBatch":[50,2,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer10startBatchE7int64_t"],"paddle::SparseMomentumParameterOptimizer::t0Vec_":[50,3,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer6t0Vec_E"],"paddle::SparseMomentumParameterOptimizer::tau_":[50,3,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer4tau_E"],"paddle::SparseMomentumParameterOptimizer::threshold_":[50,3,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer10threshold_E"],"paddle::SparseMomentumParameterOptimizer::timer_":[50,3,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer6timer_E"],"paddle::SparseMomentumParameterOptimizer::update":[50,2,1,"_CPPv2N6paddle32SparseMomentumParameterOptimizer6updateEA_K9VectorPtrRK15ParameterConfig6size_t"],"paddle::SparseNonValueScanner":[36,1,1,"_CPPv2N6paddle21SparseNonValueScannerE"],"paddle::SparseNonValueScanner::SparseNonValueScanner":[36,2,1,"_CPPv2N6paddle21SparseNonValueScanner21SparseNonValueScannerEP10SlotHeader"],"paddle::SparseNonValueScanner::fill":[36,2,1,"_CPPv2N6paddle21SparseNonValueScanner4fillER8ArgumentP8PyObject"],"paddle::SparseNonValueScanner::finishPrepare":[36,2,1,"_CPPv2N6paddle21SparseNonValueScanner13finishPrepareER8Argument"],"paddle::SparseNonValueScanner::height_":[36,3,1,"_CPPv2N6paddle21SparseNonValueScanner7height_E"],"paddle::SparseNonValueScanner::nnz_":[36,3,1,"_CPPv2N6paddle21SparseNonValueScanner4nnz_E"],"paddle::SparseNonValueScanner::prepare":[36,2,1,"_CPPv2N6paddle21SparseNonValueScanner7prepareER8ArgumentP8PyObject"],"paddle::SparseNonValueScanner::setData":[36,2,1,"_CPPv2N6paddle21SparseNonValueScanner7setDataEPiP4realP8PyObject"],"paddle::SparseNonValueScanner::startFill":[36,2,1,"_CPPv2N6paddle21SparseNonValueScanner9startFillER8Argument"],"paddle::SparsePrefetchRowCpuMatrix":[46,1,1,"_CPPv2N6paddle26SparsePrefetchRowCpuMatrixE"],"paddle::SparsePrefetchRowCpuMatrix::SparsePrefetchRowCpuMatrix":[46,2,1,"_CPPv2N6paddle26SparsePrefetchRowCpuMatrix26SparsePrefetchRowCpuMatrixE15CpuMemHandlePtr6size_t6size_t12IndexDictPtrP14SyncThreadPoolb"],"paddle::SparsePrefetchRowCpuMatrix::addRows":[46,2,1,"_CPPv2N6paddle26SparsePrefetchRowCpuMatrix7addRowsEPKj6size_t"],"paddle::SparsePrefetchRowCpuMatrix::pool_":[46,3,1,"_CPPv2N6paddle26SparsePrefetchRowCpuMatrix5pool_E"],"paddle::SparsePrefetchRowCpuMatrix::setupIndices":[46,2,1,"_CPPv2N6paddle26SparsePrefetchRowCpuMatrix12setupIndicesEv"],"paddle::SparseRemoteParameterUpdater":[61,1,1,"_CPPv2N6paddle28SparseRemoteParameterUpdaterE"],"paddle::SparseRemoteParameterUpdater::SparseRemoteParameterUpdater":[61,2,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater28SparseRemoteParameterUpdaterERK18OptimizationConfigib"],"paddle::SparseRemoteParameterUpdater::apply":[61,2,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater5applyEv"],"paddle::SparseRemoteParameterUpdater::batchSize_":[61,3,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater10batchSize_E"],"paddle::SparseRemoteParameterUpdater::config_":[61,3,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater7config_E"],"paddle::SparseRemoteParameterUpdater::controller":[61,2,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater10controllerEv"],"paddle::SparseRemoteParameterUpdater::controllerThread_":[61,3,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater17controllerThread_E"],"paddle::SparseRemoteParameterUpdater::expectedPassCount_":[61,3,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater18expectedPassCount_E"],"paddle::SparseRemoteParameterUpdater::finishBatch":[61,2,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater11finishBatchE4real"],"paddle::SparseRemoteParameterUpdater::finishPass":[61,2,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater10finishPassE4real"],"paddle::SparseRemoteParameterUpdater::getParametersRemote":[61,2,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater19getParametersRemoteEbb"],"paddle::SparseRemoteParameterUpdater::init":[61,2,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater4initERNSt6vectorI12ParameterPtrEE"],"paddle::SparseRemoteParameterUpdater::loadParametersRemote":[61,2,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater20loadParametersRemoteERKNSt6stringE"],"paddle::SparseRemoteParameterUpdater::parameterClient_":[61,3,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater16parameterClient_E"],"paddle::SparseRemoteParameterUpdater::passCount_":[61,3,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater10passCount_E"],"paddle::SparseRemoteParameterUpdater::randParametersRemote":[61,2,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater20randParametersRemoteEv"],"paddle::SparseRemoteParameterUpdater::restore":[61,2,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater7restoreEv"],"paddle::SparseRemoteParameterUpdater::saveParametersRemote":[61,2,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater20saveParametersRemoteERKNSt6stringE"],"paddle::SparseRemoteParameterUpdater::setForwardbackwardTime":[61,2,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater22setForwardbackwardTimeE8uint64_t"],"paddle::SparseRemoteParameterUpdater::startBatch":[61,2,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater10startBatchE7int64_t"],"paddle::SparseRemoteParameterUpdater::startController":[61,2,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater15startControllerEv"],"paddle::SparseRemoteParameterUpdater::startPass":[61,2,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater9startPassEv"],"paddle::SparseRemoteParameterUpdater::testing_":[61,3,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater8testing_E"],"paddle::SparseRemoteParameterUpdater::updateImpl":[61,2,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater10updateImplEP9Parameter"],"paddle::SparseRemoteParameterUpdater::useApplyInPserver_":[61,3,1,"_CPPv2N6paddle28SparseRemoteParameterUpdater18useApplyInPserver_E"],"paddle::SparseRemoteParameterUpdater::~SparseRemoteParameterUpdater":[61,2,1,"_CPPv2N6paddle28SparseRemoteParameterUpdaterD0Ev"],"paddle::SparseRemoteParameterUpdaterComposite":[61,1,1,"_CPPv2N6paddle37SparseRemoteParameterUpdaterCompositeE"],"paddle::SparseRemoteParameterUpdaterComposite::NUMBER_UPDATERS":[61,1,1,"_CPPv2N6paddle37SparseRemoteParameterUpdaterComposite15NUMBER_UPDATERSE"],"paddle::SparseRemoteParameterUpdaterComposite::SparseRemoteParameterUpdaterComposite":[61,2,1,"_CPPv2N6paddle37SparseRemoteParameterUpdaterComposite37SparseRemoteParameterUpdaterCompositeERK18OptimizationConfigibRRNSt10unique_ptrI16ParameterUpdaterEE"],"paddle::SparseRemoteParameterUpdaterComposite::UPDATER_NORMAL":[61,1,1,"_CPPv2N6paddle37SparseRemoteParameterUpdaterComposite14UPDATER_NORMALE"],"paddle::SparseRemoteParameterUpdaterComposite::UPDATER_SPARSE_REMOTE":[61,1,1,"_CPPv2N6paddle37SparseRemoteParameterUpdaterComposite21UPDATER_SPARSE_REMOTEE"],"paddle::SparseRemoteParameterUpdaterComposite::__anonymous4":[61,4,1,"_CPPv2N6paddle37SparseRemoteParameterUpdaterComposite12__anonymous4E"],"paddle::SparseRemoteParameterUpdaterComposite::init":[61,2,1,"_CPPv2N6paddle37SparseRemoteParameterUpdaterComposite4initERNSt6vectorI12ParameterPtrEE"],"paddle::SparseRowCpuMatrix":[46,1,1,"_CPPv2N6paddle18SparseRowCpuMatrixE"],"paddle::SparseRowCpuMatrix::IndexDict":[46,1,1,"_CPPv2N6paddle18SparseRowCpuMatrix9IndexDictE"],"paddle::SparseRowCpuMatrix::IndexDict::globalIndices":[46,3,1,"_CPPv2N6paddle18SparseRowCpuMatrix9IndexDict13globalIndicesE"],"paddle::SparseRowCpuMatrix::IndexDict::localIndices":[46,3,1,"_CPPv2N6paddle18SparseRowCpuMatrix9IndexDict12localIndicesE"],"paddle::SparseRowCpuMatrix::IndexDictPtr":[46,4,1,"_CPPv2N6paddle18SparseRowCpuMatrix12IndexDictPtrE"],"paddle::SparseRowCpuMatrix::SparseRowCpuMatrix":[46,2,1,"_CPPv2N6paddle18SparseRowCpuMatrix18SparseRowCpuMatrixE15CpuMemHandlePtr6size_t6size_t12IndexDictPtrb"],"paddle::SparseRowCpuMatrix::addTo":[46,2,1,"_CPPv2N6paddle18SparseRowCpuMatrix5addToER18SparseRowCpuMatrix6size_t6size_t"],"paddle::SparseRowCpuMatrix::apply":[46,2,1,"_CPPv2N6paddle18SparseRowCpuMatrix5applyE4Func"],"paddle::SparseRowCpuMatrix::applyL1Decay":[46,2,1,"_CPPv2N6paddle18SparseRowCpuMatrix12applyL1DecayE4real4real"],"paddle::SparseRowCpuMatrix::checkIndex":[46,2,1,"_CPPv2N6paddle18SparseRowCpuMatrix10checkIndexE6size_t"],"paddle::SparseRowCpuMatrix::checkIndices":[46,2,1,"_CPPv2N6paddle18SparseRowCpuMatrix12checkIndicesEv"],"paddle::SparseRowCpuMatrix::checkStoreSize":[46,2,1,"_CPPv2N6paddle18SparseRowCpuMatrix14checkStoreSizeEv"],"paddle::SparseRowCpuMatrix::clearIndices":[46,2,1,"_CPPv2N6paddle18SparseRowCpuMatrix12clearIndicesEv"],"paddle::SparseRowCpuMatrix::clearRows":[46,2,1,"_CPPv2N6paddle18SparseRowCpuMatrix9clearRowsEv"],"paddle::SparseRowCpuMatrix::copyFrom":[46,2,1,"_CPPv2N6paddle18SparseRowCpuMatrix8copyFromEPK4real6size_t"],"paddle::SparseRowCpuMatrix::getIndexDictHandle":[46,2,1,"_CPPv2N6paddle18SparseRowCpuMatrix18getIndexDictHandleEv"],"paddle::SparseRowCpuMatrix::getLocalIndices":[46,2,1,"_CPPv2N6paddle18SparseRowCpuMatrix15getLocalIndicesEv"],"paddle::SparseRowCpuMatrix::getLocalRow":[46,2,1,"_CPPv2N6paddle18SparseRowCpuMatrix11getLocalRowE6size_t"],"paddle::SparseRowCpuMatrix::getRow":[46,2,1,"_CPPv2N6paddle18SparseRowCpuMatrix6getRowE6size_t"],"paddle::SparseRowCpuMatrix::getRowBuf":[46,2,1,"_CPPv2N6paddle18SparseRowCpuMatrix9getRowBufE6size_t"],"paddle::SparseRowCpuMatrix::globalIndices_":[46,3,1,"_CPPv2N6paddle18SparseRowCpuMatrix14globalIndices_E"],"paddle::SparseRowCpuMatrix::indexDictHandle_":[46,3,1,"_CPPv2N6paddle18SparseRowCpuMatrix16indexDictHandle_E"],"paddle::SparseRowCpuMatrix::init":[46,2,1,"_CPPv2N6paddle18SparseRowCpuMatrix4initE6size_t6size_t"],"paddle::SparseRowCpuMatrix::kUnusedId_":[46,3,1,"_CPPv2N6paddle18SparseRowCpuMatrix10kUnusedId_E"],"paddle::SparseRowCpuMatrix::localIndices_":[46,3,1,"_CPPv2N6paddle18SparseRowCpuMatrix13localIndices_E"],"paddle::SparseRowCpuMatrix::mul":[46,2,1,"_CPPv2N6paddle18SparseRowCpuMatrix3mulEP15CpuSparseMatrixP9CpuMatrix4real4real"],"paddle::SparseRowCpuMatrix::reserveStore":[46,2,1,"_CPPv2N6paddle18SparseRowCpuMatrix12reserveStoreEv"],"paddle::SparseRowCpuMatrix::rowStore_":[46,3,1,"_CPPv2N6paddle18SparseRowCpuMatrix9rowStore_E"],"paddle::SparseRowCpuMatrix::sgdUpdate":[46,2,1,"_CPPv2N6paddle18SparseRowCpuMatrix9sgdUpdateER10BaseMatrixR7IVector4reali4realbb"],"paddle::SparseRowCpuMatrix::storeMat_":[46,3,1,"_CPPv2N6paddle18SparseRowCpuMatrix9storeMat_E"],"paddle::SparseRowCpuMatrix::zeroMem":[46,2,1,"_CPPv2N6paddle18SparseRowCpuMatrix7zeroMemEv"],"paddle::SparseRowCpuMatrix::zeroMemThread":[46,2,1,"_CPPv2N6paddle18SparseRowCpuMatrix13zeroMemThreadE6size_t6size_t"],"paddle::SparseRowCpuMatrix::~SparseRowCpuMatrix":[46,2,1,"_CPPv2N6paddle18SparseRowCpuMatrixD0Ev"],"paddle::SparseRowIdsCpuMatrix":[46,1,1,"_CPPv2N6paddle21SparseRowIdsCpuMatrixE"],"paddle::SparseRowIdsCpuMatrix::SparseRowIdsCpuMatrix":[46,2,1,"_CPPv2N6paddle21SparseRowIdsCpuMatrix21SparseRowIdsCpuMatrixE15CpuMemHandlePtr6size_t6size_tb"],"paddle::SparseRowIdsCpuMatrix::getIds":[46,2,1,"_CPPv2N6paddle21SparseRowIdsCpuMatrix6getIdsE6size_t"],"paddle::SparseRowIdsCpuMatrix::idsArray_":[46,3,1,"_CPPv2N6paddle21SparseRowIdsCpuMatrix9idsArray_E"],"paddle::SparseRowIdsCpuMatrix::setNumOfThreads":[46,2,1,"_CPPv2N6paddle21SparseRowIdsCpuMatrix15setNumOfThreadsE6size_t"],"paddle::SparseValueScanner":[36,1,1,"_CPPv2N6paddle18SparseValueScannerE"],"paddle::SparseValueScanner::SparseValueScanner":[36,2,1,"_CPPv2N6paddle18SparseValueScanner18SparseValueScannerEP10SlotHeader"],"paddle::SparseValueScanner::finishPrepare":[36,2,1,"_CPPv2N6paddle18SparseValueScanner13finishPrepareER8Argument"],"paddle::SparseValueScanner::setData":[36,2,1,"_CPPv2N6paddle18SparseValueScanner7setDataEPiP4realP8PyObject"],"paddle::SparseValueType":[46,4,1,"_CPPv2N6paddle15SparseValueTypeE"],"paddle::SpinLock":[66,1,1,"_CPPv2N6paddle8SpinLockE"],"paddle::SpinLock::SpinLock":[66,2,1,"_CPPv2N6paddle8SpinLock8SpinLockEv"],"paddle::SpinLock::lock":[66,2,1,"_CPPv2N6paddle8SpinLock4lockEv"],"paddle::SpinLock::lock_":[66,3,1,"_CPPv2N6paddle8SpinLock5lock_E"],"paddle::SpinLock::operator=":[66,2,1,"_CPPv2N6paddle8SpinLockaSERK8SpinLock"],"paddle::SpinLock::padding_":[66,3,1,"_CPPv2N6paddle8SpinLock8padding_E"],"paddle::SpinLock::unlock":[66,2,1,"_CPPv2N6paddle8SpinLock6unlockEv"],"paddle::SpinLock::~SpinLock":[66,2,1,"_CPPv2N6paddle8SpinLockD0Ev"],"paddle::StorageEngine":[48,1,1,"_CPPv2N6paddle13StorageEngineE"],"paddle::StorageEngine::StorageEngine":[48,2,1,"_CPPv2N6paddle13StorageEngine13StorageEngineEv"],"paddle::StorageEngine::cpuAllocator_":[48,3,1,"_CPPv2N6paddle13StorageEngine13cpuAllocator_E"],"paddle::StorageEngine::getCpuAllocator":[48,2,1,"_CPPv2N6paddle13StorageEngine15getCpuAllocatorEv"],"paddle::StorageEngine::getGpuAllocator":[48,2,1,"_CPPv2N6paddle13StorageEngine15getGpuAllocatorEi"],"paddle::StorageEngine::gpuAllocator_":[48,3,1,"_CPPv2N6paddle13StorageEngine13gpuAllocator_E"],"paddle::StorageEngine::lock_":[48,3,1,"_CPPv2N6paddle13StorageEngine5lock_E"],"paddle::StorageEngine::singleton":[48,2,1,"_CPPv2N6paddle13StorageEngine9singletonEv"],"paddle::StorageEngine::~StorageEngine":[48,2,1,"_CPPv2N6paddle13StorageEngineD0Ev"],"paddle::SubSequenceLayer":[43,1,1,"_CPPv2N6paddle16SubSequenceLayerE"],"paddle::SubSequenceLayer::SubSequenceLayer":[43,2,1,"_CPPv2N6paddle16SubSequenceLayer16SubSequenceLayerERK11LayerConfig"],"paddle::SubSequenceLayer::backward":[43,2,1,"_CPPv2N6paddle16SubSequenceLayer8backwardERK14UpdateCallback"],"paddle::SubSequenceLayer::biases_":[43,3,1,"_CPPv2N6paddle16SubSequenceLayer7biases_E"],"paddle::SubSequenceLayer::forward":[43,2,1,"_CPPv2N6paddle16SubSequenceLayer7forwardE8PassType"],"paddle::SubSequenceLayer::init":[43,2,1,"_CPPv2N6paddle16SubSequenceLayer4initERK8LayerMapRK12ParameterMap"],"paddle::SubSequenceLayer::tmpDest_":[43,3,1,"_CPPv2N6paddle16SubSequenceLayer8tmpDest_E"],"paddle::SubSequenceLayer::tmpSrc_":[43,3,1,"_CPPv2N6paddle16SubSequenceLayer7tmpSrc_E"],"paddle::SubSequenceLayer::~SubSequenceLayer":[43,2,1,"_CPPv2N6paddle16SubSequenceLayerD0Ev"],"paddle::SumEvaluator":[38,1,1,"_CPPv2N6paddle12SumEvaluatorE"],"paddle::SumEvaluator::SumEvaluator":[38,2,1,"_CPPv2N6paddle12SumEvaluator12SumEvaluatorEv"],"paddle::SumEvaluator::distributeEval":[38,2,1,"_CPPv2N6paddle12SumEvaluator14distributeEvalEP16ParameterClient2"],"paddle::SumEvaluator::evalImp":[38,2,1,"_CPPv2N6paddle12SumEvaluator7evalImpERNSt6vectorI8ArgumentEE"],"paddle::SumEvaluator::updateSamplesNum":[38,2,1,"_CPPv2N6paddle12SumEvaluator16updateSamplesNumERKNSt6vectorI8ArgumentEE"],"paddle::SumOfSquaresCostLayer":[43,1,1,"_CPPv2N6paddle21SumOfSquaresCostLayerE"],"paddle::SumOfSquaresCostLayer::SumOfSquaresCostLayer":[43,2,1,"_CPPv2N6paddle21SumOfSquaresCostLayer21SumOfSquaresCostLayerERK11LayerConfig"],"paddle::SumOfSquaresCostLayer::backwardImp":[43,2,1,"_CPPv2N6paddle21SumOfSquaresCostLayer11backwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::SumOfSquaresCostLayer::forwardImp":[43,2,1,"_CPPv2N6paddle21SumOfSquaresCostLayer10forwardImpER6MatrixR8ArgumentR6Matrix"],"paddle::SumOfSquaresCostLayer::init":[43,2,1,"_CPPv2N6paddle21SumOfSquaresCostLayer4initERK8LayerMapRK12ParameterMap"],"paddle::SumToOneNormLayer":[43,1,1,"_CPPv2N6paddle17SumToOneNormLayerE"],"paddle::SumToOneNormLayer::SumToOneNormLayer":[43,2,1,"_CPPv2N6paddle17SumToOneNormLayer17SumToOneNormLayerERK11LayerConfig"],"paddle::SumToOneNormLayer::backward":[43,2,1,"_CPPv2N6paddle17SumToOneNormLayer8backwardERK14UpdateCallback"],"paddle::SumToOneNormLayer::dotSum_":[43,3,1,"_CPPv2N6paddle17SumToOneNormLayer7dotSum_E"],"paddle::SumToOneNormLayer::forward":[43,2,1,"_CPPv2N6paddle17SumToOneNormLayer7forwardE8PassType"],"paddle::SumToOneNormLayer::init":[43,2,1,"_CPPv2N6paddle17SumToOneNormLayer4initERK8LayerMapRK12ParameterMap"],"paddle::SumToOneNormLayer::reciprocalRowSum_":[43,3,1,"_CPPv2N6paddle17SumToOneNormLayer17reciprocalRowSum_E"],"paddle::SumToOneNormLayer::~SumToOneNormLayer":[43,2,1,"_CPPv2N6paddle17SumToOneNormLayerD0Ev"],"paddle::SyncParameter":[52,1,1,"_CPPv2N6paddle13SyncParameterE"],"paddle::SyncParameter::SyncParameter":[52,2,1,"_CPPv2N6paddle13SyncParameter13SyncParameterE11TrainerRole12ParameterPtr"],"paddle::SyncParameter::attachControlParam":[52,2,1,"_CPPv2N6paddle13SyncParameter18attachControlParamE20ParallelParameterPtr"],"paddle::SyncParameter::attachMajorParam":[52,2,1,"_CPPv2N6paddle13SyncParameter16attachMajorParamE20ParallelParameterPtr"],"paddle::SyncParameter::attachMinorParam":[52,2,1,"_CPPv2N6paddle13SyncParameter16attachMinorParamE20ParallelParameterPtri"],"paddle::SyncParameter::controlParam_":[52,3,1,"_CPPv2N6paddle13SyncParameter13controlParam_E"],"paddle::SyncParameter::controlUpdate":[52,2,1,"_CPPv2N6paddle13SyncParameter13controlUpdateERK14UpdateCallback"],"paddle::SyncParameter::getMajorPartners":[52,2,1,"_CPPv2N6paddle13SyncParameter16getMajorPartnersEv"],"paddle::SyncParameter::getMinorPartners":[52,2,1,"_CPPv2N6paddle13SyncParameter16getMinorPartnersEv"],"paddle::SyncParameter::majorPartners_":[52,3,1,"_CPPv2N6paddle13SyncParameter14majorPartners_E"],"paddle::SyncParameter::majorUpdate":[52,2,1,"_CPPv2N6paddle13SyncParameter11majorUpdateE4real"],"paddle::SyncParameter::minorDeviceIds_":[52,3,1,"_CPPv2N6paddle13SyncParameter15minorDeviceIds_E"],"paddle::SyncParameter::minorPartners_":[52,3,1,"_CPPv2N6paddle13SyncParameter14minorPartners_E"],"paddle::SyncParameter::minorUpdate":[52,2,1,"_CPPv2N6paddle13SyncParameter11minorUpdateE4real"],"paddle::SyncParameter::partnerParam_":[52,3,1,"_CPPv2N6paddle13SyncParameter13partnerParam_E"],"paddle::SyncParameter::singleUpdate":[52,2,1,"_CPPv2N6paddle13SyncParameter12singleUpdateE4real"],"paddle::SyncParameter::synchronizeParamter":[52,2,1,"_CPPv2N6paddle13SyncParameter19synchronizeParamterEv"],"paddle::SyncParameter::waitAllMajorGradReady":[52,2,1,"_CPPv2N6paddle13SyncParameter21waitAllMajorGradReadyEv"],"paddle::SyncParameter::~SyncParameter":[52,2,1,"_CPPv2N6paddle13SyncParameterD0Ev"],"paddle::SyncThreadPool":[64,1,1,"_CPPv2N6paddle14SyncThreadPoolE"],"paddle::SyncThreadPool::JobFunc":[64,4,1,"_CPPv2N6paddle14SyncThreadPool7JobFuncE"],"paddle::SyncThreadPool::SyncThreadPool":[64,2,1,"_CPPv2N6paddle14SyncThreadPool14SyncThreadPoolEv"],"paddle::SyncThreadPool::checkOwner_":[64,3,1,"_CPPv2N6paddle14SyncThreadPool11checkOwner_E"],"paddle::SyncThreadPool::exec":[64,2,1,"_CPPv2N6paddle14SyncThreadPool4execE7JobFunc7JobFunc"],"paddle::SyncThreadPool::execHelper":[64,2,1,"_CPPv2N6paddle14SyncThreadPool10execHelperEP14SyncThreadPool7JobFunc"],"paddle::SyncThreadPool::execPlusOwner":[64,2,1,"_CPPv2N6paddle14SyncThreadPool13execPlusOwnerE7JobFunc"],"paddle::SyncThreadPool::getNumThreads":[64,2,1,"_CPPv2N6paddle14SyncThreadPool13getNumThreadsEv"],"paddle::SyncThreadPool::jobFinishBarrier_":[64,3,1,"_CPPv2N6paddle14SyncThreadPool17jobFinishBarrier_E"],"paddle::SyncThreadPool::jobFunc_":[64,3,1,"_CPPv2N6paddle14SyncThreadPool8jobFunc_E"],"paddle::SyncThreadPool::jobStartBarrier_":[64,3,1,"_CPPv2N6paddle14SyncThreadPool16jobStartBarrier_E"],"paddle::SyncThreadPool::ownerThreadId_":[64,3,1,"_CPPv2N6paddle14SyncThreadPool14ownerThreadId_E"],"paddle::SyncThreadPool::run":[64,2,1,"_CPPv2N6paddle14SyncThreadPool3runEi"],"paddle::SyncThreadPool::start":[64,2,1,"_CPPv2N6paddle14SyncThreadPool5startEv"],"paddle::SyncThreadPool::stop":[64,2,1,"_CPPv2N6paddle14SyncThreadPool4stopEv"],"paddle::SyncThreadPool::stopping_":[64,3,1,"_CPPv2N6paddle14SyncThreadPool9stopping_E"],"paddle::SyncThreadPool::workers_":[64,3,1,"_CPPv2N6paddle14SyncThreadPool8workers_E"],"paddle::SyncThreadPool::~SyncThreadPool":[64,2,1,"_CPPv2N6paddle14SyncThreadPoolD0Ev"],"paddle::SyncedFlag":[46,4,1,"_CPPv2N6paddle10SyncedFlagE"],"paddle::TRAINER_ROLE_CONTROL":[52,1,1,"_CPPv2N6paddle20TRAINER_ROLE_CONTROLE"],"paddle::TRAINER_ROLE_MAJOR":[52,1,1,"_CPPv2N6paddle18TRAINER_ROLE_MAJORE"],"paddle::TRAINER_ROLE_MASTER":[52,1,1,"_CPPv2N6paddle19TRAINER_ROLE_MASTERE"],"paddle::TRAINER_ROLE_MINOR":[52,1,1,"_CPPv2N6paddle18TRAINER_ROLE_MINORE"],"paddle::TRAINER_ROLE_SINGLE":[52,1,1,"_CPPv2N6paddle19TRAINER_ROLE_SINGLEE"],"paddle::TRAINER_ROLE_SLAVE":[52,1,1,"_CPPv2N6paddle18TRAINER_ROLE_SLAVEE"],"paddle::TableProjection":[43,1,1,"_CPPv2N6paddle15TableProjectionE"],"paddle::TableProjection::TableProjection":[43,2,1,"_CPPv2N6paddle15TableProjection15TableProjectionERK16ProjectionConfigRK12ParameterPtrb"],"paddle::TableProjection::backward":[43,2,1,"_CPPv2N6paddle15TableProjection8backwardERK14UpdateCallback"],"paddle::TableProjection::forward":[43,2,1,"_CPPv2N6paddle15TableProjection7forwardEv"],"paddle::TableProjection::prefetch":[43,2,1,"_CPPv2N6paddle15TableProjection8prefetchEPK8Argument"],"paddle::TableProjection::table_":[43,3,1,"_CPPv2N6paddle15TableProjection6table_E"],"paddle::TensorLayer":[43,1,1,"_CPPv2N6paddle11TensorLayerE"],"paddle::TensorLayer::TensorLayer":[43,2,1,"_CPPv2N6paddle11TensorLayer11TensorLayerERK11LayerConfig"],"paddle::TensorLayer::backward":[43,2,1,"_CPPv2N6paddle11TensorLayer8backwardERK14UpdateCallback"],"paddle::TensorLayer::biases_":[43,3,1,"_CPPv2N6paddle11TensorLayer7biases_E"],"paddle::TensorLayer::forward":[43,2,1,"_CPPv2N6paddle11TensorLayer7forwardE8PassType"],"paddle::TensorLayer::getWeight":[43,2,1,"_CPPv2N6paddle11TensorLayer9getWeightEi"],"paddle::TensorLayer::init":[43,2,1,"_CPPv2N6paddle11TensorLayer4initERK8LayerMapRK12ParameterMap"],"paddle::TensorLayer::weights_":[43,3,1,"_CPPv2N6paddle11TensorLayer8weights_E"],"paddle::TensorLayer::~TensorLayer":[43,2,1,"_CPPv2N6paddle11TensorLayerD0Ev"],"paddle::Thread":[64,1,1,"_CPPv2N6paddle6ThreadE"],"paddle::Thread::Thread":[64,2,1,"_CPPv2N6paddle6Thread6ThreadEv"],"paddle::Thread::detach":[64,2,1,"_CPPv2N6paddle6Thread6detachEv"],"paddle::Thread::join":[64,2,1,"_CPPv2N6paddle6Thread4joinEv"],"paddle::Thread::run":[64,2,1,"_CPPv2N6paddle6Thread3runEv"],"paddle::Thread::start":[64,2,1,"_CPPv2N6paddle6Thread5startEv"],"paddle::Thread::thread_":[64,3,1,"_CPPv2N6paddle6Thread7thread_E"],"paddle::Thread::~Thread":[64,2,1,"_CPPv2N6paddle6ThreadD0Ev"],"paddle::ThreadBarrier":[66,1,1,"_CPPv2N6paddle13ThreadBarrierE"],"paddle::ThreadBarrier::ThreadBarrier":[66,2,1,"_CPPv2N6paddle13ThreadBarrier13ThreadBarrierEi"],"paddle::ThreadBarrier::barrier_":[66,3,1,"_CPPv2N6paddle13ThreadBarrier8barrier_E"],"paddle::ThreadBarrier::operator=":[66,2,1,"_CPPv2N6paddle13ThreadBarrieraSERK13ThreadBarrier"],"paddle::ThreadBarrier::wait":[66,2,1,"_CPPv2N6paddle13ThreadBarrier4waitEv"],"paddle::ThreadBarrier::~ThreadBarrier":[66,2,1,"_CPPv2N6paddle13ThreadBarrierD0Ev"],"paddle::ThreadWorker":[64,1,1,"_CPPv2N6paddle12ThreadWorkerE"],"paddle::ThreadWorker::JobFunc":[64,4,1,"_CPPv2N6paddle12ThreadWorker7JobFuncE"],"paddle::ThreadWorker::ThreadWorker":[64,2,1,"_CPPv2N6paddle12ThreadWorker12ThreadWorkerEv"],"paddle::ThreadWorker::addJob":[64,2,1,"_CPPv2N6paddle12ThreadWorker6addJobE7JobFunc"],"paddle::ThreadWorker::empty_":[64,3,1,"_CPPv2N6paddle12ThreadWorker6empty_E"],"paddle::ThreadWorker::finishCV_":[64,3,1,"_CPPv2N6paddle12ThreadWorker9finishCV_E"],"paddle::ThreadWorker::jobs_":[64,3,1,"_CPPv2N6paddle12ThreadWorker5jobs_E"],"paddle::ThreadWorker::run":[64,2,1,"_CPPv2N6paddle12ThreadWorker3runEv"],"paddle::ThreadWorker::stop":[64,2,1,"_CPPv2N6paddle12ThreadWorker4stopEv"],"paddle::ThreadWorker::stopping_":[64,3,1,"_CPPv2N6paddle12ThreadWorker9stopping_E"],"paddle::ThreadWorker::wait":[64,2,1,"_CPPv2N6paddle12ThreadWorker4waitEv"],"paddle::ThreadWorker::~ThreadWorker":[64,2,1,"_CPPv2N6paddle12ThreadWorkerD0Ev"],"paddle::TrainerRole":[52,4,1,"_CPPv2N6paddle11TrainerRoleE"],"paddle::TrainerStats":[61,1,1,"_CPPv2N6paddle12TrainerStatsE"],"paddle::TrainerStats::TrainerStats":[61,2,1,"_CPPv2N6paddle12TrainerStats12TrainerStatsEv"],"paddle::TrainerStats::addCost":[61,2,1,"_CPPv2N6paddle12TrainerStats7addCostE7int64_t4real"],"paddle::TrainerStats::getAvgCost":[61,2,1,"_CPPv2N6paddle12TrainerStats10getAvgCostEv"],"paddle::TrainerStats::getCurrentAvgCost":[61,2,1,"_CPPv2N6paddle12TrainerStats17getCurrentAvgCostEv"],"paddle::TrainerStats::getNumProcessed":[61,2,1,"_CPPv2N6paddle12TrainerStats15getNumProcessedEv"],"paddle::TrainerStats::getStats":[61,2,1,"_CPPv2N6paddle12TrainerStats8getStatsEb"],"paddle::TrainerStats::operator+=":[61,2,1,"_CPPv2N6paddle12TrainerStatspLERKNSt4pairI7int64_t4realEE"],"paddle::TrainerStats::reset":[61,2,1,"_CPPv2N6paddle12TrainerStats5resetEv"],"paddle::TrainerStats::resetCurrentStat":[61,2,1,"_CPPv2N6paddle12TrainerStats16resetCurrentStatEv"],"paddle::TrainerStats::showStats":[61,2,1,"_CPPv2N6paddle12TrainerStats9showStatsERNSt7ostreamEb"],"paddle::TrainerThread":[40,1,1,"_CPPv2N6paddle13TrainerThreadE"],"paddle::TrainerThread::TrainerThread":[40,2,1,"_CPPv2N6paddle13TrainerThread13TrainerThreadERK11ModelConfigiP20MultiGradientMachine"],"paddle::TrainerThread::backward":[40,2,1,"_CPPv2N6paddle13TrainerThread8backwardEv"],"paddle::TrainerThread::backwardCallback":[40,2,1,"_CPPv2N6paddle13TrainerThread16backwardCallbackEP9Parameter"],"paddle::TrainerThread::backwardCallback_":[40,3,1,"_CPPv2N6paddle13TrainerThread17backwardCallback_E"],"paddle::TrainerThread::computeThread":[40,2,1,"_CPPv2N6paddle13TrainerThread13computeThreadEv"],"paddle::TrainerThread::computeThread_":[40,3,1,"_CPPv2N6paddle13TrainerThread14computeThread_E"],"paddle::TrainerThread::config_":[40,3,1,"_CPPv2N6paddle13TrainerThread7config_E"],"paddle::TrainerThread::copyGradToBufferThread":[40,2,1,"_CPPv2N6paddle13TrainerThread22copyGradToBufferThreadEv"],"paddle::TrainerThread::copyInArgs":[40,2,1,"_CPPv2N6paddle13TrainerThread10copyInArgsEv"],"paddle::TrainerThread::copyOutputGrad":[40,2,1,"_CPPv2N6paddle13TrainerThread14copyOutputGradEv"],"paddle::TrainerThread::copyThread_":[40,3,1,"_CPPv2N6paddle13TrainerThread11copyThread_E"],"paddle::TrainerThread::deviceId_":[40,3,1,"_CPPv2N6paddle13TrainerThread9deviceId_E"],"paddle::TrainerThread::doCallback":[40,2,1,"_CPPv2N6paddle13TrainerThread10doCallbackEi"],"paddle::TrainerThread::forward":[40,2,1,"_CPPv2N6paddle13TrainerThread7forwardEv"],"paddle::TrainerThread::getDeviceId":[40,2,1,"_CPPv2N6paddle13TrainerThread11getDeviceIdEv"],"paddle::TrainerThread::getGradientMachine":[40,2,1,"_CPPv2N6paddle13TrainerThread18getGradientMachineEv"],"paddle::TrainerThread::getOutArgs":[40,2,1,"_CPPv2N6paddle13TrainerThread10getOutArgsEv"],"paddle::TrainerThread::getParameters":[40,2,1,"_CPPv2N6paddle13TrainerThread13getParametersEv"],"paddle::TrainerThread::getValueBuf":[40,2,1,"_CPPv2N6paddle13TrainerThread11getValueBufEi"],"paddle::TrainerThread::gradBufQueue_":[40,3,1,"_CPPv2N6paddle13TrainerThread13gradBufQueue_E"],"paddle::TrainerThread::gradCollectThread":[40,2,1,"_CPPv2N6paddle13TrainerThread17gradCollectThreadEv"],"paddle::TrainerThread::gradCollectThread_":[40,3,1,"_CPPv2N6paddle13TrainerThread18gradCollectThread_E"],"paddle::TrainerThread::gradQueue_":[40,3,1,"_CPPv2N6paddle13TrainerThread10gradQueue_E"],"paddle::TrainerThread::gradStream_":[40,3,1,"_CPPv2N6paddle13TrainerThread11gradStream_E"],"paddle::TrainerThread::gradientMachine_":[40,3,1,"_CPPv2N6paddle13TrainerThread16gradientMachine_E"],"paddle::TrainerThread::inArgsCopied_":[40,3,1,"_CPPv2N6paddle13TrainerThread13inArgsCopied_E"],"paddle::TrainerThread::inArgs_":[40,3,1,"_CPPv2N6paddle13TrainerThread7inArgs_E"],"paddle::TrainerThread::incUpdateCounter":[40,2,1,"_CPPv2N6paddle13TrainerThread16incUpdateCounterEi"],"paddle::TrainerThread::mergeCpuGradients":[40,2,1,"_CPPv2N6paddle13TrainerThread17mergeCpuGradientsEv"],"paddle::TrainerThread::mergeGradDense":[40,2,1,"_CPPv2N6paddle13TrainerThread14mergeGradDenseEP9ParameterRNSt6vectorIPKNSt6vectorI12ParameterPtrEEEE"],"paddle::TrainerThread::mergeGradSparse":[40,2,1,"_CPPv2N6paddle13TrainerThread15mergeGradSparseEP9ParameterRNSt6vectorIPKNSt6vectorI12ParameterPtrEEEE"],"paddle::TrainerThread::mergeGradSparseRemote":[40,2,1,"_CPPv2N6paddle13TrainerThread21mergeGradSparseRemoteEP9ParameterRNSt6vectorIPKNSt6vectorI12ParameterPtrEEEE"],"paddle::TrainerThread::mergeTypes_":[40,3,1,"_CPPv2N6paddle13TrainerThread11mergeTypes_E"],"paddle::TrainerThread::multiMachine_":[40,3,1,"_CPPv2N6paddle13TrainerThread13multiMachine_E"],"paddle::TrainerThread::notifyCopyGradToBuffer":[40,2,1,"_CPPv2N6paddle13TrainerThread22notifyCopyGradToBufferEi"],"paddle::TrainerThread::notifyGradientCollect":[40,2,1,"_CPPv2N6paddle13TrainerThread21notifyGradientCollectEi"],"paddle::TrainerThread::notifyTaskReady":[40,2,1,"_CPPv2N6paddle13TrainerThread15notifyTaskReadyEv"],"paddle::TrainerThread::notifyValueDispatch":[40,2,1,"_CPPv2N6paddle13TrainerThread19notifyValueDispatchEi"],"paddle::TrainerThread::notifyValueReady":[40,2,1,"_CPPv2N6paddle13TrainerThread16notifyValueReadyEi"],"paddle::TrainerThread::onPassEnd":[40,2,1,"_CPPv2N6paddle13TrainerThread9onPassEndEv"],"paddle::TrainerThread::outArgsReadySem_":[40,3,1,"_CPPv2N6paddle13TrainerThread16outArgsReadySem_E"],"paddle::TrainerThread::outArgs_":[40,3,1,"_CPPv2N6paddle13TrainerThread8outArgs_E"],"paddle::TrainerThread::parameterUpdated_":[40,3,1,"_CPPv2N6paddle13TrainerThread17parameterUpdated_E"],"paddle::TrainerThread::parameters_":[40,3,1,"_CPPv2N6paddle13TrainerThread11parameters_E"],"paddle::TrainerThread::partnerId_":[40,3,1,"_CPPv2N6paddle13TrainerThread10partnerId_E"],"paddle::TrainerThread::prefetch":[40,2,1,"_CPPv2N6paddle13TrainerThread8prefetchEv"],"paddle::TrainerThread::start":[40,2,1,"_CPPv2N6paddle13TrainerThread5startEv"],"paddle::TrainerThread::stop":[40,2,1,"_CPPv2N6paddle13TrainerThread4stopEv"],"paddle::TrainerThread::stopping_":[40,3,1,"_CPPv2N6paddle13TrainerThread9stopping_E"],"paddle::TrainerThread::taskReadySem_":[40,3,1,"_CPPv2N6paddle13TrainerThread13taskReadySem_E"],"paddle::TrainerThread::threadId_":[40,3,1,"_CPPv2N6paddle13TrainerThread9threadId_E"],"paddle::TrainerThread::updateCounter_":[40,3,1,"_CPPv2N6paddle13TrainerThread14updateCounter_E"],"paddle::TrainerThread::valueDispatchThread":[40,2,1,"_CPPv2N6paddle13TrainerThread19valueDispatchThreadEv"],"paddle::TrainerThread::valueDispatchThread_":[40,3,1,"_CPPv2N6paddle13TrainerThread20valueDispatchThread_E"],"paddle::TrainerThread::valueReadyCond_":[40,3,1,"_CPPv2N6paddle13TrainerThread15valueReadyCond_E"],"paddle::TrainerThread::valueReadyQueue_":[40,3,1,"_CPPv2N6paddle13TrainerThread16valueReadyQueue_E"],"paddle::TrainerThread::valueStream_":[40,3,1,"_CPPv2N6paddle13TrainerThread12valueStream_E"],"paddle::TrainerThread::waitOutArgsReady":[40,2,1,"_CPPv2N6paddle13TrainerThread16waitOutArgsReadyEv"],"paddle::TrainerThread::~TrainerThread":[40,2,1,"_CPPv2N6paddle13TrainerThreadD0Ev"],"paddle::TransLayer":[43,1,1,"_CPPv2N6paddle10TransLayerE"],"paddle::TransLayer::TransLayer":[43,2,1,"_CPPv2N6paddle10TransLayer10TransLayerERK11LayerConfig"],"paddle::TransLayer::backward":[43,2,1,"_CPPv2N6paddle10TransLayer8backwardERK14UpdateCallback"],"paddle::TransLayer::forward":[43,2,1,"_CPPv2N6paddle10TransLayer7forwardE8PassType"],"paddle::TransLayer::init":[43,2,1,"_CPPv2N6paddle10TransLayer4initERK8LayerMapRK12ParameterMap"],"paddle::TransposedFullMatrixProjection":[43,1,1,"_CPPv2N6paddle30TransposedFullMatrixProjectionE"],"paddle::TransposedFullMatrixProjection::TransposedFullMatrixProjection":[43,2,1,"_CPPv2N6paddle30TransposedFullMatrixProjection30TransposedFullMatrixProjectionERK16ProjectionConfig12ParameterPtrb"],"paddle::TransposedFullMatrixProjection::backward":[43,2,1,"_CPPv2N6paddle30TransposedFullMatrixProjection8backwardERK14UpdateCallback"],"paddle::TransposedFullMatrixProjection::forward":[43,2,1,"_CPPv2N6paddle30TransposedFullMatrixProjection7forwardEv"],"paddle::TransposedFullMatrixProjection::weight_":[43,3,1,"_CPPv2N6paddle30TransposedFullMatrixProjection7weight_E"],"paddle::UPDATE_TYPE_NUM":[52,3,1,"_CPPv2N6paddle15UPDATE_TYPE_NUME"],"paddle::UpdateCallback":[52,4,1,"_CPPv2N6paddle14UpdateCallbackE"],"paddle::UpdateFunction":[52,4,1,"_CPPv2N6paddle14UpdateFunctionE"],"paddle::UserDefinedVectorPtr":[52,4,1,"_CPPv2N6paddle20UserDefinedVectorPtrE"],"paddle::ValidationLayer":[43,1,1,"_CPPv2N6paddle15ValidationLayerE"],"paddle::ValidationLayer::ValidationLayer":[43,2,1,"_CPPv2N6paddle15ValidationLayer15ValidationLayerERK11LayerConfig"],"paddle::ValidationLayer::backward":[43,2,1,"_CPPv2N6paddle15ValidationLayer8backwardERK14UpdateCallback"],"paddle::ValidationLayer::forward":[43,2,1,"_CPPv2N6paddle15ValidationLayer7forwardE8PassType"],"paddle::ValidationLayer::getInfoLayer":[43,2,1,"_CPPv2N6paddle15ValidationLayer12getInfoLayerEv"],"paddle::ValidationLayer::getLabelLayer":[43,2,1,"_CPPv2N6paddle15ValidationLayer13getLabelLayerEv"],"paddle::ValidationLayer::getOutputLayer":[43,2,1,"_CPPv2N6paddle15ValidationLayer14getOutputLayerEv"],"paddle::ValidationLayer::init":[43,2,1,"_CPPv2N6paddle15ValidationLayer4initERK8LayerMapRK12ParameterMap"],"paddle::ValidationLayer::onPassEnd":[43,2,1,"_CPPv2N6paddle15ValidationLayer9onPassEndEv"],"paddle::ValidationLayer::validationImp":[43,2,1,"_CPPv2N6paddle15ValidationLayer13validationImpE9MatrixPtr10IVectorPtr"],"paddle::ValuePrinter":[38,1,1,"_CPPv2N6paddle12ValuePrinterE"],"paddle::ValuePrinter::ValuePrinter":[38,2,1,"_CPPv2N6paddle12ValuePrinter12ValuePrinterEv"],"paddle::ValuePrinter::eval":[38,2,1,"_CPPv2N6paddle12ValuePrinter4evalERK13NeuralNetwork"],"paddle::ValuePrinter::evalImp":[38,2,1,"_CPPv2N6paddle12ValuePrinter7evalImpERNSt6vectorI8ArgumentEE"],"paddle::ValuePrinter::updateSamplesNum":[38,2,1,"_CPPv2N6paddle12ValuePrinter16updateSamplesNumERKNSt6vectorI8ArgumentEE"],"paddle::Vector":[46,4,1,"_CPPv2N6paddle6VectorE"],"paddle::VectorPtr":[46,4,1,"_CPPv2N6paddle9VectorPtrE"],"paddle::VectorT":[46,1,1,"_CPPv2N6paddle7VectorTE"],"paddle::VectorT::VectorT":[46,2,1,"_CPPv2N6paddle7VectorT7VectorTE6size_tP1Tb"],"paddle::VectorT::copyFrom":[46,2,1,"_CPPv2N6paddle7VectorT8copyFromERK7VectorTI1TE11hl_stream_t"],"paddle::VectorT::copyTo":[46,2,1,"_CPPv2N6paddle7VectorT6copyToEP10GpuVectorTI1TE"],"paddle::VectorT::create":[46,2,1,"_CPPv2N6paddle7VectorT6createEP1T6size_tb"],"paddle::VectorT::createParallelVector":[46,2,1,"_CPPv2N6paddle7VectorT20createParallelVectorE6size_tbP14SyncThreadPool"],"paddle::VectorT::exec":[46,2,1,"_CPPv2N6paddle7VectorT4execEN14SyncThreadPool7JobFuncE"],"paddle::VectorT::fillSequence":[46,2,1,"_CPPv2N6paddle7VectorT12fillSequenceEv"],"paddle::VectorT::get":[46,2,1,"_CPPv2N6paddle7VectorT3getE6size_t"],"paddle::VectorT::getAbsMax":[46,2,1,"_CPPv2N6paddle7VectorT9getAbsMaxEv"],"paddle::VectorT::getAbsSum":[46,2,1,"_CPPv2N6paddle7VectorT9getAbsSumEv"],"paddle::VectorT::getData":[46,2,1,"_CPPv2N6paddle7VectorT7getDataEv"],"paddle::VectorT::getElement":[46,2,1,"_CPPv2N6paddle7VectorT10getElementE6size_t"],"paddle::VectorT::getMax":[46,2,1,"_CPPv2N6paddle7VectorT6getMaxEv"],"paddle::VectorT::getMemoryHandle":[46,2,1,"_CPPv2N6paddle7VectorT15getMemoryHandleEv"],"paddle::VectorT::getMin":[46,2,1,"_CPPv2N6paddle7VectorT6getMinEv"],"paddle::VectorT::getPoint":[46,2,1,"_CPPv2N6paddle7VectorT8getPointEK8uint64_t"],"paddle::VectorT::getSize":[46,2,1,"_CPPv2N6paddle7VectorT7getSizeEv"],"paddle::VectorT::getSum":[46,2,1,"_CPPv2N6paddle7VectorT6getSumEv"],"paddle::VectorT::histogram":[46,2,1,"_CPPv2N6paddle7VectorT9histogramERNSt7ostreamEi"],"paddle::VectorT::isEqualTo":[46,2,1,"_CPPv2N6paddle7VectorT9isEqualToERK7VectorTI1TERK1T"],"paddle::VectorT::memoryHandle_":[46,3,1,"_CPPv2N6paddle7VectorT13memoryHandle_E"],"paddle::VectorT::newMemory":[46,2,1,"_CPPv2N6paddle7VectorT9newMemoryE6size_t"],"paddle::VectorT::print":[46,2,1,"_CPPv2N6paddle7VectorT5printERNSt7ostreamE6size_t"],"paddle::VectorT::printOneElement":[46,2,1,"_CPPv2N6paddle7VectorT15printOneElementERNSt7ostreamE6size_t"],"paddle::VectorT::rand":[46,2,1,"_CPPv2N6paddle7VectorT4randEv"],"paddle::VectorT::randnorm":[46,2,1,"_CPPv2N6paddle7VectorT8randnormE4real4real"],"paddle::VectorT::reset":[46,2,1,"_CPPv2N6paddle7VectorT5resetERK1T"],"paddle::VectorT::resize":[46,2,1,"_CPPv2N6paddle7VectorT6resizeE6size_t"],"paddle::VectorT::resizeOrCreate":[46,2,1,"_CPPv2N6paddle7VectorT14resizeOrCreateERNSt10shared_ptrI7VectorTI1TEEE6size_tb"],"paddle::VectorT::selectFrom":[46,2,1,"_CPPv2N6paddle7VectorT10selectFromERK7VectorTI1TERK7VectorTIiE"],"paddle::VectorT::setElement":[46,2,1,"_CPPv2N6paddle7VectorT10setElementE6size_tRK1T"],"paddle::VectorT::subVec":[46,2,1,"_CPPv2N6paddle7VectorT6subVecE6size_t6size_t"],"paddle::VectorT::subVecFrom":[46,2,1,"_CPPv2N6paddle7VectorT10subVecFromERK7VectorTI1TENSt4pairI6size_t6size_tEE"],"paddle::VectorT::uniform":[46,2,1,"_CPPv2N6paddle7VectorT7uniformE4real4real"],"paddle::VectorT::zeroMem":[46,2,1,"_CPPv2N6paddle7VectorT7zeroMemEv"],"paddle::VectorT::~VectorT":[46,2,1,"_CPPv2N6paddle7VectorTD0Ev"],"paddle::Weight":[52,1,1,"_CPPv2N6paddle6WeightE"],"paddle::Weight::Weight":[52,2,1,"_CPPv2N6paddle6Weight6WeightE6size_t6size_t12ParameterPtr6size_t"],"paddle::Weight::getParameterPtr":[52,2,1,"_CPPv2N6paddle6Weight15getParameterPtrEv"],"paddle::Weight::getW":[52,2,1,"_CPPv2N6paddle6Weight4getWEv"],"paddle::Weight::getWGrad":[52,2,1,"_CPPv2N6paddle6Weight8getWGradEv"],"paddle::Weight::incUpdate":[52,2,1,"_CPPv2N6paddle6Weight9incUpdateERK14UpdateCallback"],"paddle::Weight::parameter_":[52,3,1,"_CPPv2N6paddle6Weight10parameter_E"],"paddle::Weight::setParameterPtr":[52,2,1,"_CPPv2N6paddle6Weight15setParameterPtrE12ParameterPtr"],"paddle::Weight::weightGrad_":[52,3,1,"_CPPv2N6paddle6Weight11weightGrad_E"],"paddle::Weight::weight_":[52,3,1,"_CPPv2N6paddle6Weight7weight_E"],"paddle::WeightList":[52,4,1,"_CPPv2N6paddle10WeightListE"],"paddle::__attribute__":[54,3,1,"_CPPv2N6paddle13__attribute__E"],"paddle::axpy":[46,2,1,"_CPPv2N6paddle4axpyEKiK1TPK1TP1T"],"paddle::bool_constant":[46,1,1,"_CPPv2N6paddle13bool_constantE"],"paddle::bool_constant::value":[46,3,1,"_CPPv2N6paddle13bool_constant5valueE"],"paddle::dotProduct":[46,2,1,"_CPPv2N6paddle10dotProductEKiPK1TPK1T"],"paddle::enumeration_wrapper":[63,4,1,"_CPPv2N6paddle19enumeration_wrapperE"],"paddle::enumeration_wrapper::NUM_PARAMETER_TYPES":[63,1,1,"_CPPv2N6paddle19enumeration_wrapper19NUM_PARAMETER_TYPESE"],"paddle::enumeration_wrapper::PARAMETER_APPLY":[63,1,1,"_CPPv2N6paddle19enumeration_wrapper15PARAMETER_APPLYE"],"paddle::enumeration_wrapper::PARAMETER_COLS":[63,1,1,"_CPPv2N6paddle19enumeration_wrapper14PARAMETER_COLSE"],"paddle::enumeration_wrapper::PARAMETER_DELTA":[63,1,1,"_CPPv2N6paddle19enumeration_wrapper15PARAMETER_DELTAE"],"paddle::enumeration_wrapper::PARAMETER_GRADIENT":[63,1,1,"_CPPv2N6paddle19enumeration_wrapper18PARAMETER_GRADIENTE"],"paddle::enumeration_wrapper::PARAMETER_GRADIENT_SQURESUM":[63,1,1,"_CPPv2N6paddle19enumeration_wrapper27PARAMETER_GRADIENT_SQURESUME"],"paddle::enumeration_wrapper::PARAMETER_GRADIENT_SQURESUM1":[63,1,1,"_CPPv2N6paddle19enumeration_wrapper28PARAMETER_GRADIENT_SQURESUM1E"],"paddle::enumeration_wrapper::PARAMETER_GRADIENT_SUM":[63,1,1,"_CPPv2N6paddle19enumeration_wrapper22PARAMETER_GRADIENT_SUME"],"paddle::enumeration_wrapper::PARAMETER_LEARNING_RATE":[63,1,1,"_CPPv2N6paddle19enumeration_wrapper23PARAMETER_LEARNING_RATEE"],"paddle::enumeration_wrapper::PARAMETER_MOMENTUM":[63,1,1,"_CPPv2N6paddle19enumeration_wrapper18PARAMETER_MOMENTUME"],"paddle::enumeration_wrapper::PARAMETER_MOMENTUM_UT":[63,1,1,"_CPPv2N6paddle19enumeration_wrapper21PARAMETER_MOMENTUM_UTE"],"paddle::enumeration_wrapper::PARAMETER_MOMENTUM_VT":[63,1,1,"_CPPv2N6paddle19enumeration_wrapper21PARAMETER_MOMENTUM_VTE"],"paddle::enumeration_wrapper::PARAMETER_ROWS":[63,1,1,"_CPPv2N6paddle19enumeration_wrapper14PARAMETER_ROWSE"],"paddle::enumeration_wrapper::PARAMETER_SECOND_MOMENTUM":[63,1,1,"_CPPv2N6paddle19enumeration_wrapper25PARAMETER_SECOND_MOMENTUME"],"paddle::enumeration_wrapper::PARAMETER_SUM1":[63,1,1,"_CPPv2N6paddle19enumeration_wrapper14PARAMETER_SUM1E"],"paddle::enumeration_wrapper::PARAMETER_SUM2":[63,1,1,"_CPPv2N6paddle19enumeration_wrapper14PARAMETER_SUM2E"],"paddle::enumeration_wrapper::PARAMETER_SUM3":[63,1,1,"_CPPv2N6paddle19enumeration_wrapper14PARAMETER_SUM3E"],"paddle::enumeration_wrapper::PARAMETER_UPDATE_TIME":[63,1,1,"_CPPv2N6paddle19enumeration_wrapper21PARAMETER_UPDATE_TIMEE"],"paddle::enumeration_wrapper::PARAMETER_VALUE":[63,1,1,"_CPPv2N6paddle19enumeration_wrapper15PARAMETER_VALUEE"],"paddle::enumeration_wrapper::PARAMETER_WEIGHTED_INFINITY_NORM":[63,1,1,"_CPPv2N6paddle19enumeration_wrapper32PARAMETER_WEIGHTED_INFINITY_NORME"],"paddle::enumeration_wrapper::PASS_GC":[63,1,1,"_CPPv2N6paddle19enumeration_wrapper7PASS_GCE"],"paddle::enumeration_wrapper::PASS_METRIC":[63,1,1,"_CPPv2N6paddle19enumeration_wrapper11PASS_METRICE"],"paddle::enumeration_wrapper::PASS_METRIC_TRAIN":[63,1,1,"_CPPv2N6paddle19enumeration_wrapper17PASS_METRIC_TRAINE"],"paddle::enumeration_wrapper::PASS_METRIC_TRAIN_WITH_NOERROR":[63,1,1,"_CPPv2N6paddle19enumeration_wrapper30PASS_METRIC_TRAIN_WITH_NOERRORE"],"paddle::enumeration_wrapper::PASS_TEST":[63,1,1,"_CPPv2N6paddle19enumeration_wrapper9PASS_TESTE"],"paddle::enumeration_wrapper::PASS_TRAIN":[63,1,1,"_CPPv2N6paddle19enumeration_wrapper10PASS_TRAINE"],"paddle::enumeration_wrapper::ParameterType":[63,4,1,"_CPPv2N6paddle19enumeration_wrapper13ParameterTypeE"],"paddle::enumeration_wrapper::PassType":[63,4,1,"_CPPv2N6paddle19enumeration_wrapper8PassTypeE"],"paddle::false_type":[46,4,1,"_CPPv2N6paddle10false_typeE"],"paddle::gemm":[46,2,1,"_CPPv2N6paddle4gemmEK15CBLAS_TRANSPOSEK15CBLAS_TRANSPOSEKiKiKiK1TPK1TKiPK1TKiK1TP1TKi"],"paddle::hl_sparse_matrix_s_ptr":[46,4,1,"_CPPv2N6paddle22hl_sparse_matrix_s_ptrE"],"paddle::operator<<":[46,2,1,"_CPPv2N6paddlelsERNSt7ostreamERK7VectorTI1TE"],"paddle::simd":[46,4,1,"_CPPv2N6paddle4simdE"],"paddle::simd::addTo":[46,2,1,"_CPPv2N6paddle4simd5addToEPfPKf6size_t"],"paddle::simd::batchAddTo":[46,2,1,"_CPPv2N6paddle4simd10batchAddToEPfA_PKfi6size_t"],"paddle::simd::colMax":[46,2,1,"_CPPv2N6paddle4simd6colMaxEPfPKfii"],"paddle::simd::decayL1":[46,2,1,"_CPPv2N6paddle4simd7decayL1EPfPff6size_t"],"paddle::simd::isPointerAlign":[46,2,1,"_CPPv2N6paddle4simd14isPointerAlignEPv"],"paddle::simd::naive":[46,4,1,"_CPPv2N6paddle4simd5naiveE"],"paddle::simd::naive::addTo":[46,2,1,"_CPPv2N6paddle4simd5naive5addToEP4TypePK4Type6size_t"],"paddle::simd::naive::batchAddTo":[46,2,1,"_CPPv2N6paddle4simd5naive10batchAddToEP4TypeA_PK4Typei6size_t"],"paddle::simd::naive::colMax":[46,2,1,"_CPPv2N6paddle4simd5naive6colMaxEP4TypePK4Typeii"],"paddle::simd::naive::decayL1":[46,2,1,"_CPPv2N6paddle4simd5naive7decayL1EP4TypeP4TypeP4Type4Type6size_t"],"paddle::simd::vec_check":[46,2,1,"_CPPv2N6paddle4simd9vec_checkE6size_t"],"paddle::sparse_float_value_t":[46,1,1,"_CPPv2N6paddle20sparse_float_value_tE"],"paddle::sparse_float_value_t::col":[46,3,1,"_CPPv2N6paddle20sparse_float_value_t3colE"],"paddle::sparse_float_value_t::value":[46,3,1,"_CPPv2N6paddle20sparse_float_value_t5valueE"],"paddle::sparse_non_value_t":[46,1,1,"_CPPv2N6paddle18sparse_non_value_tE"],"paddle::sparse_non_value_t::col":[46,3,1,"_CPPv2N6paddle18sparse_non_value_t3colE"],"paddle::true_type":[46,4,1,"_CPPv2N6paddle9true_typeE"],Arguments:[26,1,1,"_CPPv29Arguments"],BaseOp:[30,1,1,"_CPPv26BaseOp"],CAL_MATRIX_START_ADDRESS:[46,0,1,"c.CAL_MATRIX_START_ADDRESS"],CBLAS_GEMM:[32,0,1,"c.CBLAS_GEMM"],CREATE_MODE_NORMAL:[26,1,1,"_CPPv218CREATE_MODE_NORMAL"],CREATE_MODE_TESTING:[26,1,1,"_CPPv219CREATE_MODE_TESTING"],DEFINE_MATRIX_BINARY_OP:[30,0,1,"c.DEFINE_MATRIX_BINARY_OP"],DEFINE_MATRIX_BINARY_PARAMETER_OP:[30,0,1,"c.DEFINE_MATRIX_BINARY_PARAMETER_OP"],DEFINE_MATRIX_QUATERNARY_OP:[30,0,1,"c.DEFINE_MATRIX_QUATERNARY_OP"],DEFINE_MATRIX_QUATERNARY_PARAMETER_OP:[30,0,1,"c.DEFINE_MATRIX_QUATERNARY_PARAMETER_OP"],DEFINE_MATRIX_TERNARY_OP:[30,0,1,"c.DEFINE_MATRIX_TERNARY_OP"],DEFINE_MATRIX_TERNARY_PARAMETER_OP:[30,0,1,"c.DEFINE_MATRIX_TERNARY_PARAMETER_OP"],DEFINE_MATRIX_UNARY_OP:[30,0,1,"c.DEFINE_MATRIX_UNARY_OP"],DEFINE_MATRIX_UNARY_PARAMETER_OP:[30,0,1,"c.DEFINE_MATRIX_UNARY_PARAMETER_OP"],DEVICE_FMAX:[30,0,1,"c.DEVICE_FMAX"],DEVICE_FMIN:[30,0,1,"c.DEVICE_FMIN"],DISABLE_COPY_AND_ASSIGN:[26,0,1,"c.DISABLE_COPY_AND_ASSIGN"],DIVUP:[34,0,1,"c.DIVUP"],FOUR_PARAMETER:[30,0,1,"c.FOUR_PARAMETER"],FloatArray:[26,1,1,"_CPPv210FloatArray"],GET_CONVOLUTION_DESCRIPTOR:[27,0,1,"c.GET_CONVOLUTION_DESCRIPTOR"],GET_FILTER_DESCRIPTOR:[27,0,1,"c.GET_FILTER_DESCRIPTOR"],GET_TENSOR_DESCRIPTOR:[27,0,1,"c.GET_TENSOR_DESCRIPTOR"],GetCublasDsoHandle:[27,2,1,"_CPPv218GetCublasDsoHandlePPv"],GetCudartDsoHandle:[27,2,1,"_CPPv218GetCudartDsoHandlePPv"],GetCudnnDsoHandle:[27,2,1,"_CPPv217GetCudnnDsoHandlePPv"],GetCurandDsoHandle:[27,2,1,"_CPPv218GetCurandDsoHandlePPv"],GradientMachine:[26,1,1,"_CPPv215GradientMachine"],GradientMatchineCreateMode:[26,4,1,"_CPPv226GradientMatchineCreateMode"],HL_ACTIVATION_END:[34,1,1,"_CPPv217HL_ACTIVATION_END"],HL_ACTIVATION_LINEAR:[34,1,1,"_CPPv220HL_ACTIVATION_LINEAR"],HL_ACTIVATION_RELU:[34,1,1,"_CPPv218HL_ACTIVATION_RELU"],HL_ACTIVATION_SIGMOID:[34,1,1,"_CPPv221HL_ACTIVATION_SIGMOID"],HL_ACTIVATION_TANH:[34,1,1,"_CPPv218HL_ACTIVATION_TANH"],HL_CPU_GRU_CUH_:[32,0,1,"c.HL_CPU_GRU_CUH_"],HL_CUDA_CUDNN_PH_:[27,0,1,"c.HL_CUDA_CUDNN_PH_"],HL_CUDA_PH_:[27,0,1,"c.HL_CUDA_PH_"],HL_DEVICE:[30,0,1,"c.HL_DEVICE"],HL_DEVICE_FUNCTIONS_CUH_:[32,0,1,"c.HL_DEVICE_FUNCTIONS_CUH_"],HL_FLOAT_MAX:[34,0,1,"c.HL_FLOAT_MAX"],HL_FLOAT_MIN:[34,0,1,"c.HL_FLOAT_MIN"],HL_FLOAT_VALUE:[34,1,1,"_CPPv214HL_FLOAT_VALUE"],HL_GPU_FUNCTIONS_CUH_:[32,0,1,"c.HL_GPU_FUNCTIONS_CUH_"],HL_GPU_GRU_CUH_:[32,0,1,"c.HL_GPU_GRU_CUH_"],HL_GPU_LSTM_CUH_:[32,0,1,"c.HL_GPU_LSTM_CUH_"],HL_GRU_OPS_CUH_:[32,0,1,"c.HL_GRU_OPS_CUH_"],HL_LSTM_OPS_CUH_:[32,0,1,"c.HL_LSTM_OPS_CUH_"],HL_MATRIX_APPLY_H_:[30,0,1,"c.HL_MATRIX_APPLY_H_"],HL_MATRIX_BASE_CUH_:[30,0,1,"c.HL_MATRIX_BASE_CUH_"],HL_MATRIX_OPS_CUH_:[30,0,1,"c.HL_MATRIX_OPS_CUH_"],HL_MATRIX_TYPE_CUH_:[30,0,1,"c.HL_MATRIX_TYPE_CUH_"],HL_NO_VALUE:[34,1,1,"_CPPv211HL_NO_VALUE"],HL_POOLING_AVERAGE:[27,1,1,"_CPPv218HL_POOLING_AVERAGE"],HL_POOLING_AVERAGE_EXCLUDE_PADDING:[27,1,1,"_CPPv234HL_POOLING_AVERAGE_EXCLUDE_PADDING"],HL_POOLING_END:[27,1,1,"_CPPv214HL_POOLING_END"],HL_POOLING_MAX:[27,1,1,"_CPPv214HL_POOLING_MAX"],HL_RECURRENT_APPLY_CUH_:[32,0,1,"c.HL_RECURRENT_APPLY_CUH_"],HL_SPARSE_CSC:[34,1,1,"_CPPv213HL_SPARSE_CSC"],HL_SPARSE_CSR:[34,1,1,"_CPPv213HL_SPARSE_CSR"],HL_SPARSE_END:[34,1,1,"_CPPv213HL_SPARSE_END"],HL_SPARSE_PH_:[30,0,1,"c.HL_SPARSE_PH_"],HL_SSE_MATRIX_KERNEL_CUH_:[30,0,1,"c.HL_SSE_MATRIX_KERNEL_CUH_"],HL_THREAD_PH_:[34,0,1,"c.HL_THREAD_PH_"],HL_VALUE_END:[34,1,1,"_CPPv212HL_VALUE_END"],HPPL_ACTIVE_FUNCTION:[32,0,1,"c.HPPL_ACTIVE_FUNCTION"],HPPL_OP_END:[34,1,1,"_CPPv211HPPL_OP_END"],HPPL_OP_N:[34,1,1,"_CPPv29HPPL_OP_N"],HPPL_OP_T:[34,1,1,"_CPPv29HPPL_OP_T"],HPPL_STREAM_1:[34,1,1,"_CPPv213HPPL_STREAM_1"],HPPL_STREAM_2:[34,1,1,"_CPPv213HPPL_STREAM_2"],HPPL_STREAM_3:[34,1,1,"_CPPv213HPPL_STREAM_3"],HPPL_STREAM_4:[34,1,1,"_CPPv213HPPL_STREAM_4"],HPPL_STREAM_DEFAULT:[34,1,1,"_CPPv219HPPL_STREAM_DEFAULT"],HPPL_STREAM_END:[34,1,1,"_CPPv215HPPL_STREAM_END"],HPPL_THREAD_STREAM_1:[34,1,1,"_CPPv220HPPL_THREAD_STREAM_1"],HPPL_THREAD_STREAM_2:[34,1,1,"_CPPv220HPPL_THREAD_STREAM_2"],HPPL_THREAD_STREAM_3:[34,1,1,"_CPPv220HPPL_THREAD_STREAM_3"],HPPL_THREAD_STREAM_4:[34,1,1,"_CPPv220HPPL_THREAD_STREAM_4"],INLINE:[32,0,1,"c.INLINE"],IOError:[26,1,1,"_CPPv27IOError"],ISequenceResults:[26,1,1,"_CPPv216ISequenceResults"],IVector:[26,1,1,"_CPPv27IVector"],IntArray:[26,1,1,"_CPPv28IntArray"],IntWithFloatArray:[26,1,1,"_CPPv217IntWithFloatArray"],Matrix:[26,1,1,"_CPPv26Matrix"],ModelConfig:[26,1,1,"_CPPv211ModelConfig"],NO_SPARSE_ID:[26,3,1,"_CPPv212NO_SPARSE_ID"],ONE_PARAMETER:[30,0,1,"c.ONE_PARAMETER"],OptimizationConfig:[26,1,1,"_CPPv218OptimizationConfig"],P_DECLARE_bool:[46,2,1,"_CPPv214P_DECLARE_bool31allow_inefficient_sparse_update"],Parameter:[26,1,1,"_CPPv29Parameter"],ParameterConfig:[26,1,1,"_CPPv215ParameterConfig"],ParameterOptimizer:[26,1,1,"_CPPv218ParameterOptimizer"],ParameterTraverseCallback:[26,1,1,"_CPPv225ParameterTraverseCallback"],RangeError:[26,1,1,"_CPPv210RangeError"],SIGMOID_THRESHOLD_MAX:[32,0,1,"c.SIGMOID_THRESHOLD_MAX"],SIGMOID_THRESHOLD_MIN:[32,0,1,"c.SIGMOID_THRESHOLD_MIN"],SPARSE_CSC:[26,1,1,"_CPPv210SPARSE_CSC"],SPARSE_CSR:[26,1,1,"_CPPv210SPARSE_CSR"],SPARSE_NON_VALUE:[26,1,1,"_CPPv216SPARSE_NON_VALUE"],SPARSE_VALUE:[26,1,1,"_CPPv212SPARSE_VALUE"],SequenceGenerator:[26,1,1,"_CPPv217SequenceGenerator"],SparseFormatType:[26,4,1,"_CPPv216SparseFormatType"],SparseValueType:[26,4,1,"_CPPv215SparseValueType"],THREE_PARAMETER:[30,0,1,"c.THREE_PARAMETER"],TWO_PARAMETER:[30,0,1,"c.TWO_PARAMETER"],Trainer:[26,1,1,"_CPPv27Trainer"],TrainerConfig:[26,1,1,"_CPPv213TrainerConfig"],UnsupportError:[26,1,1,"_CPPv214UnsupportError"],UpdateCallback:[26,1,1,"_CPPv214UpdateCallback"],VECTOR_LEN:[30,0,1,"c.VECTOR_LEN"],VECTOR_SET:[30,0,1,"c.VECTOR_SET"],VECTOR_SIZE:[30,0,1,"c.VECTOR_SIZE"],Vector:[26,1,1,"_CPPv26Vector"],__sparse_get_return__:[30,0,1,"c.__sparse_get_return__"],__sparse_get_type_return__:[30,0,1,"c.__sparse_get_type_return__"],_cudnn_convolution_descriptor:[27,1,1,"_CPPv229_cudnn_convolution_descriptor"],_cudnn_filter_descriptor:[27,1,1,"_CPPv224_cudnn_filter_descriptor"],_cudnn_pooling_descriptor:[27,1,1,"_CPPv225_cudnn_pooling_descriptor"],_cudnn_tensor_descriptor:[27,1,1,"_CPPv224_cudnn_tensor_descriptor"],_global_device_resources:[27,1,1,"_CPPv224_global_device_resources"],_hl_csc_matrix:[30,1,1,"_CPPv214_hl_csc_matrix"],_hl_csr_matrix:[30,1,1,"_CPPv214_hl_csr_matrix"],_hl_device_prop:[27,1,1,"_CPPv215_hl_device_prop"],_hl_event_st:[27,1,1,"_CPPv212_hl_event_st"],_hl_sparse_matrix_s:[34,1,1,"_CPPv219_hl_sparse_matrix_s"],_hl_thread_resource:[34,1,1,"_CPPv219_hl_thread_resource"],_thread_device_resources:[27,1,1,"_CPPv224_thread_device_resources"],aggregate:[30,4,1,"_CPPv29aggregate"],backward_reset_grad:[32,2,1,"_CPPv219backward_reset_grad11OpResetGrad12hl_gru_value11hl_gru_gradii20hl_activation_mode_t"],backward_state_grad:[32,2,1,"_CPPv219backward_state_grad11OpStateGrad12hl_gru_value11hl_gru_gradii20hl_activation_mode_t"],base:[30,4,1,"_CPPv24base"],batchTranspose:[30,2,1,"_CPPv214batchTransposePK4realP4realiii"],cudnn_convolution_descriptor:[27,4,1,"_CPPv228cudnn_convolution_descriptor"],cudnn_filter_descriptor:[27,4,1,"_CPPv223cudnn_filter_descriptor"],cudnn_pooling_descriptor:[27,4,1,"_CPPv224cudnn_pooling_descriptor"],cudnn_tensor_descriptor:[27,4,1,"_CPPv223cudnn_tensor_descriptor"],forward_final_output:[32,2,1,"_CPPv220forward_final_output13OpFinalOutput12hl_gru_valueii20hl_activation_mode_t"],forward_reset_output:[32,2,1,"_CPPv220forward_reset_output13OpResetOutput12hl_gru_valueii20hl_activation_mode_t"],getCurrentTimeStick:[34,2,1,"_CPPv219getCurrentTimeStickv"],global_device_resources:[27,4,1,"_CPPv223global_device_resources"],hl_CMRNorm_backward:[32,2,1,"_CPPv219hl_CMRNorm_backward6size_tPK4realPK4realPK4realPK4realP4real6size_t6size_t6size_t6size_t4real4real"],hl_CMRNorm_forward:[32,2,1,"_CPPv218hl_CMRNorm_forward6size_tPK4realP4realP4real6size_t6size_t6size_t6size_t4real4real"],hl_activation_mode_t:[34,4,1,"_CPPv220hl_activation_mode_t"],hl_agg_op:[30,2,1,"_CPPv29hl_agg_op3Agg7vecType"],hl_avgpool_backward:[32,2,1,"_CPPv219hl_avgpool_backwardiPK4realiiiiiiiiP4real4real4real"],hl_avgpool_forward:[32,2,1,"_CPPv218hl_avgpool_forwardiPK4realiiiiiiiiP4real"],hl_avx_gru_backward_reset_grad:[32,2,1,"_CPPv230hl_avx_gru_backward_reset_grad11OpResetGradP4realP4realP4realP4realP4reali20hl_activation_mode_t"],hl_avx_gru_backward_state_grad:[32,2,1,"_CPPv230hl_avx_gru_backward_state_grad11OpStateGradP4realP4realP4realP4realP4reali20hl_activation_mode_t"],hl_avx_gru_forward_final_output:[32,2,1,"_CPPv231hl_avx_gru_forward_final_output13OpFinalOutputP4realP4realP4reali20hl_activation_mode_t"],hl_avx_gru_forward_reset_output:[32,2,1,"_CPPv231hl_avx_gru_forward_reset_output13OpResetOutputP4realP4realP4reali20hl_activation_mode_t"],hl_batch_norm_backward:[27,2,1,"_CPPv222hl_batch_norm_backward20hl_tensor_descriptorP4real20hl_tensor_descriptorP4real20hl_tensor_descriptorP4real20hl_tensor_descriptorP4realP4realP4realdP4realP4real"],hl_batch_norm_forward_inference:[27,2,1,"_CPPv231hl_batch_norm_forward_inference20hl_tensor_descriptorP4real20hl_tensor_descriptorP4real20hl_tensor_descriptorP4realP4realP4realP4reald"],hl_batch_norm_forward_training:[27,2,1,"_CPPv230hl_batch_norm_forward_training20hl_tensor_descriptorP4real20hl_tensor_descriptorP4real20hl_tensor_descriptorP4realP4realdP4realP4realdP4realP4real"],hl_check_align:[30,2,1,"_CPPv214hl_check_alignPv"],hl_construct_sparse_matrix:[30,2,1,"_CPPv226hl_construct_sparse_matrixP18hl_sparse_matrix_sPv6size_t18hl_matrix_format_t17hl_matrix_value_tiii"],hl_context_projection_backward_data:[32,2,1,"_CPPv235hl_context_projection_backward_dataP4realPKiP4realiiii"],hl_context_projection_backward_weight:[32,2,1,"_CPPv237hl_context_projection_backward_weightP4realPKiP4realiiiiii"],hl_context_projection_forward:[32,2,1,"_CPPv229hl_context_projection_forwardP4realPKiP4realP4realiiiiib"],hl_conv_workspace:[27,2,1,"_CPPv217hl_conv_workspace20hl_tensor_descriptor20hl_tensor_descriptor20hl_filter_descriptor25hl_convolution_descriptorPiP6size_tPiP6size_tPiP6size_t"],hl_convolution_backward_bias:[27,2,1,"_CPPv228hl_convolution_backward_bias20hl_tensor_descriptorP4real20hl_tensor_descriptorP4real"],hl_convolution_backward_data:[27,2,1,"_CPPv228hl_convolution_backward_data20hl_tensor_descriptorP4real20hl_tensor_descriptorP4real20hl_filter_descriptorP4real25hl_convolution_descriptorPv6size_ti"],hl_convolution_backward_filter:[27,2,1,"_CPPv230hl_convolution_backward_filter20hl_tensor_descriptorP4real20hl_tensor_descriptorP4real20hl_filter_descriptorP4real25hl_convolution_descriptorPv6size_ti"],hl_convolution_descriptor:[27,4,1,"_CPPv225hl_convolution_descriptor"],hl_convolution_forward:[27,2,1,"_CPPv222hl_convolution_forward20hl_tensor_descriptorP4real20hl_tensor_descriptorP4real20hl_filter_descriptorP4real25hl_convolution_descriptorPv6size_ti"],hl_convolution_forward_add_bias:[27,2,1,"_CPPv231hl_convolution_forward_add_bias20hl_tensor_descriptorP4real20hl_tensor_descriptorP4real"],hl_cossim:[30,2,1,"_CPPv29hl_cossimP4realP4realP4realiii4real"],hl_cossim_derivative:[30,2,1,"_CPPv220hl_cossim_derivativeP4realP4realP4realP4realP4realP4realiii4real"],hl_cpu_apply_binary_op:[30,2,1,"_CPPv222hl_cpu_apply_binary_op2OpP1TP1Tiiii"],hl_cpu_apply_quaternary_op:[30,2,1,"_CPPv226hl_cpu_apply_quaternary_op2OpP1TP1TP1TP1Tiiiiii"],hl_cpu_apply_ternary_op:[30,2,1,"_CPPv223hl_cpu_apply_ternary_op2OpP1TP1TP1Tiiiii"],hl_cpu_apply_unary_op:[30,2,1,"_CPPv221hl_cpu_apply_unary_op2OpP1Tiii"],hl_cpu_gru_backward:[32,2,1,"_CPPv219hl_cpu_gru_backward11OpStateGrad11OpResetGrad12hl_gru_value11hl_gru_gradii20hl_activation_mode_t20hl_activation_mode_t"],hl_cpu_gru_forward:[32,2,1,"_CPPv218hl_cpu_gru_forward13OpResetOutput13OpFinalOutput12hl_gru_valueii20hl_activation_mode_t20hl_activation_mode_t"],hl_cpu_lstm_backward:[32,2,1,"_CPPv220hl_cpu_lstm_backward2Op13hl_lstm_value12hl_lstm_gradi20hl_activation_mode_t20hl_activation_mode_t20hl_activation_mode_t"],hl_cpu_lstm_forward:[32,2,1,"_CPPv219hl_cpu_lstm_forward2Op13hl_lstm_valuei20hl_activation_mode_t20hl_activation_mode_t20hl_activation_mode_t"],hl_cpu_matrix_column_op:[30,2,1,"_CPPv223hl_cpu_matrix_column_op3Agg2Op5SaveriiP4realP4realiP4reali"],hl_cpu_matrix_row_op:[30,2,1,"_CPPv220hl_cpu_matrix_row_op3Agg2OpiiP4realiP4realiP4reali"],hl_create_convolution_descriptor:[27,2,1,"_CPPv232hl_create_convolution_descriptorP25hl_convolution_descriptor20hl_tensor_descriptor20hl_filter_descriptoriiii"],hl_create_event:[27,2,1,"_CPPv215hl_create_eventP10hl_event_t"],hl_create_filter_descriptor:[27,2,1,"_CPPv227hl_create_filter_descriptorP20hl_filter_descriptoriiii"],hl_create_global_resources:[27,2,1,"_CPPv226hl_create_global_resources14hl_device_prop"],hl_create_pooling_descriptor:[27,2,1,"_CPPv228hl_create_pooling_descriptorP21hl_pooling_descriptor17hl_pooling_mode_tiiiiii"],hl_create_tensor_descriptor:[27,2,1,"_CPPv227hl_create_tensor_descriptorP20hl_tensor_descriptor"],hl_create_thread_resources:[27,2,1,"_CPPv226hl_create_thread_resourcesi23thread_device_resources"],hl_csc_matrix:[30,4,1,"_CPPv213hl_csc_matrix"],hl_csr_matrix:[30,4,1,"_CPPv213hl_csr_matrix"],hl_cublas_init:[34,2,1,"_CPPv214hl_cublas_initP14cublasHandle_t12cudaStream_t"],hl_cuda_event_is_ready:[27,2,1,"_CPPv222hl_cuda_event_is_ready10hl_event_t"],hl_cudnn_desc_init:[34,2,1,"_CPPv218hl_cudnn_desc_initP23cudnnTensorDescriptor_t"],hl_cudnn_init:[34,2,1,"_CPPv213hl_cudnn_initP13cudnnHandle_t12cudaStream_t"],hl_destroy_convolution_descriptor:[27,2,1,"_CPPv233hl_destroy_convolution_descriptor25hl_convolution_descriptor"],hl_destroy_event:[27,2,1,"_CPPv216hl_destroy_event10hl_event_t"],hl_destroy_filter_descriptor:[27,2,1,"_CPPv228hl_destroy_filter_descriptor20hl_filter_descriptor"],hl_destroy_pooling_descriptor:[27,2,1,"_CPPv229hl_destroy_pooling_descriptor21hl_pooling_descriptor"],hl_destroy_tensor_descriptor:[27,2,1,"_CPPv228hl_destroy_tensor_descriptor20hl_tensor_descriptor"],hl_destruct_sparse_matrix:[30,2,1,"_CPPv225hl_destruct_sparse_matrix18hl_sparse_matrix_s"],hl_device_can_access_peer:[27,2,1,"_CPPv225hl_device_can_access_peerii"],hl_device_enable_peer_access:[27,2,1,"_CPPv228hl_device_enable_peer_accessi"],hl_device_prop:[27,4,1,"_CPPv214hl_device_prop"],hl_device_synchronize:[27,2,1,"_CPPv221hl_device_synchronizev"],hl_event_elapsed_time:[27,2,1,"_CPPv221hl_event_elapsed_time10hl_event_t10hl_event_t"],hl_event_synchronize:[27,2,1,"_CPPv220hl_event_synchronize10hl_event_t"],hl_event_t:[27,4,1,"_CPPv210hl_event_t"],hl_expand_feature2col:[32,2,1,"_CPPv221hl_expand_feature2colPK4real6size_t6size_t6size_t6size_t6size_t6size_t6size_t6size_t6size_t6size_t6size_tP4real"],hl_filter_descriptor:[27,4,1,"_CPPv220hl_filter_descriptor"],hl_fini:[27,2,1,"_CPPv27hl_finiv"],hl_free_mem_device:[27,2,1,"_CPPv218hl_free_mem_devicePv"],hl_free_mem_host:[27,2,1,"_CPPv216hl_free_mem_hostPv"],hl_free_sparse_matrix:[30,2,1,"_CPPv221hl_free_sparse_matrix18hl_sparse_matrix_s"],hl_get_cuda_lib_version:[27,2,1,"_CPPv223hl_get_cuda_lib_versionv"],hl_get_cudnn_lib_version:[27,2,1,"_CPPv224hl_get_cudnn_lib_versionv"],hl_get_device:[27,2,1,"_CPPv213hl_get_devicev"],hl_get_device_count:[27,2,1,"_CPPv219hl_get_device_countv"],hl_get_device_error_string:[27,2,1,"_CPPv226hl_get_device_error_stringv"],hl_get_device_last_error:[27,2,1,"_CPPv224hl_get_device_last_errorv"],hl_get_sync_flag:[27,2,1,"_CPPv216hl_get_sync_flagv"],hl_gpu_apply_binary_op:[30,2,1,"_CPPv222hl_gpu_apply_binary_op2OpP1TP1Tiiii"],hl_gpu_apply_quaternary_op:[30,2,1,"_CPPv226hl_gpu_apply_quaternary_op2OpP1TP1TP1TP1Tiiiiii"],hl_gpu_apply_ternary_op:[30,2,1,"_CPPv223hl_gpu_apply_ternary_op2OpP1TP1TP1Tiiiii"],hl_gpu_apply_unary_op:[30,2,1,"_CPPv221hl_gpu_apply_unary_op2OpP1Tiii"],hl_gpu_gru_backward:[32,2,1,"_CPPv219hl_gpu_gru_backward11OpStateGrad11OpResetGrad12hl_gru_value11hl_gru_gradii20hl_activation_mode_t20hl_activation_mode_t"],hl_gpu_gru_forward:[32,2,1,"_CPPv218hl_gpu_gru_forward13OpResetOutput13OpFinalOutput12hl_gru_valueii20hl_activation_mode_t20hl_activation_mode_t"],hl_gpu_lstm_backward:[32,2,1,"_CPPv220hl_gpu_lstm_backward2Op13hl_lstm_value12hl_lstm_gradii20hl_activation_mode_t20hl_activation_mode_t20hl_activation_mode_t"],hl_gpu_lstm_forward:[32,2,1,"_CPPv219hl_gpu_lstm_forward2Op13hl_lstm_valueii20hl_activation_mode_t20hl_activation_mode_t20hl_activation_mode_t"],hl_gpu_matrix_column_op:[30,2,1,"_CPPv223hl_gpu_matrix_column_op3Agg2Op5SaveriiP4realP4realiP4reali"],hl_gpu_matrix_row_op:[30,2,1,"_CPPv220hl_gpu_matrix_row_op3Agg2OpiiP4realiP4realiP4reali"],hl_gru_grad:[34,1,1,"_CPPv211hl_gru_grad"],hl_gru_value:[34,1,1,"_CPPv212hl_gru_value"],hl_init:[27,2,1,"_CPPv27hl_initi"],hl_lstm_grad:[34,1,1,"_CPPv212hl_lstm_grad"],hl_lstm_parallel_backward_data:[32,2,1,"_CPPv230hl_lstm_parallel_backward_dataP4realP4realP4realP4realP4realP4realP4realP4realP4realP4realP4realP4realP4realP4realPKiiib20hl_activation_mode_t20hl_activation_mode_t20hl_activation_mode_t"],hl_lstm_parallel_backward_weight:[32,2,1,"_CPPv232hl_lstm_parallel_backward_weightP4realP4realP4realPKiiiib"],hl_lstm_parallel_forward:[32,2,1,"_CPPv224hl_lstm_parallel_forwardP4realP4realP4realP4realP4realP4realP4realP4realPKiiib20hl_activation_mode_t20hl_activation_mode_t20hl_activation_mode_t"],hl_lstm_value:[34,1,1,"_CPPv213hl_lstm_value"],hl_malloc_device:[27,2,1,"_CPPv216hl_malloc_device6size_t"],hl_malloc_host:[27,2,1,"_CPPv214hl_malloc_host6size_t"],hl_malloc_sparse_matrix:[30,2,1,"_CPPv223hl_malloc_sparse_matrixP18hl_sparse_matrix_s18hl_matrix_format_t17hl_matrix_value_tiii"],hl_matrix_add:[30,2,1,"_CPPv213hl_matrix_addP4realP4realP4realii4real4real"],hl_matrix_add_to_rows:[30,2,1,"_CPPv221hl_matrix_add_to_rowsP4realiP4realiPiiii"],hl_matrix_classification_error:[30,2,1,"_CPPv230hl_matrix_classification_errorP4realPiP4realii"],hl_matrix_column_max:[30,2,1,"_CPPv220hl_matrix_column_maxP4realP4realii"],hl_matrix_column_min:[30,2,1,"_CPPv220hl_matrix_column_minP4realP4realii"],hl_matrix_column_op:[30,2,1,"_CPPv219hl_matrix_column_op3Agg2Op5SaveriiP4realP4realiP4reali"],hl_matrix_column_sum:[30,2,1,"_CPPv220hl_matrix_column_sumP4realP4realii"],hl_matrix_cross_entropy:[30,2,1,"_CPPv223hl_matrix_cross_entropyP4realP4realPiii"],hl_matrix_cross_entropy_bp:[30,2,1,"_CPPv226hl_matrix_cross_entropy_bpP4realP4realPiii"],hl_matrix_csc2dense:[30,2,1,"_CPPv219hl_matrix_csc2dense18hl_sparse_matrix_sP4realii"],hl_matrix_csc_mul_dense:[30,2,1,"_CPPv223hl_matrix_csc_mul_dense18hl_sparse_matrix_s13hl_trans_op_tP4real13hl_trans_op_tP4realiii4real4real"],hl_matrix_csr2dense:[30,2,1,"_CPPv219hl_matrix_csr2dense18hl_sparse_matrix_sP4realii"],hl_matrix_csr_add_bias:[30,2,1,"_CPPv222hl_matrix_csr_add_bias18hl_sparse_matrix_sP4real4real"],hl_matrix_csr_add_dense:[30,2,1,"_CPPv223hl_matrix_csr_add_dense18hl_sparse_matrix_sP4realii4real4real"],hl_matrix_csr_column_sum:[30,2,1,"_CPPv224hl_matrix_csr_column_sumP4real18hl_sparse_matrix_sii4real"],hl_matrix_csr_mul_dense:[30,2,1,"_CPPv223hl_matrix_csr_mul_dense18hl_sparse_matrix_s13hl_trans_op_tP4real13hl_trans_op_tP4realiii4real4real"],hl_matrix_dense_mul_csc:[30,2,1,"_CPPv223hl_matrix_dense_mul_cscP4real13hl_trans_op_t18hl_sparse_matrix_s13hl_trans_op_tP4realiii4real4real"],hl_matrix_dense_mul_csr:[30,2,1,"_CPPv223hl_matrix_dense_mul_csrP4real13hl_trans_op_t18hl_sparse_matrix_s13hl_trans_op_tP4realiii4real4real"],hl_matrix_format_t:[34,4,1,"_CPPv218hl_matrix_format_t"],hl_matrix_mul:[27,2,1,"_CPPv213hl_matrix_mulP4real13hl_trans_op_tP4real13hl_trans_op_tP4realiii4real4realiii"],hl_matrix_mul_vector:[27,2,1,"_CPPv220hl_matrix_mul_vectorP4real13hl_trans_op_tP4realP4realii4real4realiii"],hl_matrix_row_max:[30,2,1,"_CPPv217hl_matrix_row_maxP4realP4realii"],hl_matrix_row_min:[30,2,1,"_CPPv217hl_matrix_row_minP4realP4realii"],hl_matrix_row_sum:[30,2,1,"_CPPv217hl_matrix_row_sumP4realP4realii"],hl_matrix_s:[34,4,1,"_CPPv211hl_matrix_s"],hl_matrix_select_rows:[30,2,1,"_CPPv221hl_matrix_select_rowsP4realiP4realiPiiii"],hl_matrix_softmax:[30,2,1,"_CPPv217hl_matrix_softmaxP4realP4realii"],hl_matrix_softmax_derivative:[30,2,1,"_CPPv228hl_matrix_softmax_derivativeP4realP4realP4realii"],hl_matrix_top_k:[30,2,1,"_CPPv215hl_matrix_top_kP4realiPiP4realiiii"],hl_matrix_transpose:[27,2,1,"_CPPv219hl_matrix_transposeP4realP4realiiii"],hl_matrix_value_t:[34,4,1,"_CPPv217hl_matrix_value_t"],hl_matrix_zero_mem:[30,2,1,"_CPPv218hl_matrix_zero_memP4reali"],hl_max_sequence_backward:[32,2,1,"_CPPv224hl_max_sequence_backwardP4realPiP4realii"],hl_max_sequence_forward:[32,2,1,"_CPPv223hl_max_sequence_forwardP4realPKiP4realPiii"],hl_maxpool_backward:[32,2,1,"_CPPv219hl_maxpool_backwardiPK4realPK4realPK4realiiiiiiiiP4real4real4real"],hl_maxpool_forward:[32,2,1,"_CPPv218hl_maxpool_forwardiPK4realiiiiiiiiP4real"],hl_memcpy:[27,2,1,"_CPPv29hl_memcpyPvPv6size_t"],hl_memcpy_async:[27,2,1,"_CPPv215hl_memcpy_asyncPvPv6size_t11hl_stream_t"],hl_memcpy_csc_matrix:[30,2,1,"_CPPv220hl_memcpy_csc_matrix18hl_sparse_matrix_sP4realPiPi11hl_stream_t"],hl_memcpy_csr_matrix:[30,2,1,"_CPPv220hl_memcpy_csr_matrix18hl_sparse_matrix_sP4realPiPi11hl_stream_t"],hl_memcpy_device2device:[27,2,1,"_CPPv223hl_memcpy_device2devicePvPv6size_t"],hl_memcpy_device2host:[27,2,1,"_CPPv221hl_memcpy_device2hostPvPv6size_t"],hl_memcpy_from_csc_matrix:[30,2,1,"_CPPv225hl_memcpy_from_csc_matrixP4real6size_tPi6size_tPi6size_t18hl_sparse_matrix_s11hl_stream_t"],hl_memcpy_from_csr_matrix:[30,2,1,"_CPPv225hl_memcpy_from_csr_matrixP4real6size_tPi6size_tPi6size_t18hl_sparse_matrix_s11hl_stream_t"],hl_memcpy_host2device:[27,2,1,"_CPPv221hl_memcpy_host2devicePvPv6size_t"],hl_memcpy_sparse_matrix:[30,2,1,"_CPPv223hl_memcpy_sparse_matrix18hl_sparse_matrix_s18hl_sparse_matrix_s11hl_stream_t"],hl_memset_device:[27,2,1,"_CPPv216hl_memset_devicePvi6size_t"],hl_naive_gru_backward_reset_grad:[32,2,1,"_CPPv232hl_naive_gru_backward_reset_grad11OpResetGradP4realP4realP4realP4realP4reali20hl_activation_mode_t"],hl_naive_gru_backward_state_grad:[32,2,1,"_CPPv232hl_naive_gru_backward_state_grad11OpStateGradP4realP4realP4realP4realP4reali20hl_activation_mode_t"],hl_naive_gru_forward_final_output:[32,2,1,"_CPPv233hl_naive_gru_forward_final_output13OpFinalOutputP4realP4realP4reali20hl_activation_mode_t"],hl_naive_gru_forward_reset_output:[32,2,1,"_CPPv233hl_naive_gru_forward_reset_output13OpResetOutputP4realP4realP4reali20hl_activation_mode_t"],hl_param_relu_backward_diff:[30,2,1,"_CPPv227hl_param_relu_backward_diffP4realP4realP4realP4realiii"],hl_param_relu_backward_w:[30,2,1,"_CPPv224hl_param_relu_backward_wP4realP4realP4realiii"],hl_param_relu_forward:[30,2,1,"_CPPv221hl_param_relu_forwardP4realP4realP4realiii"],hl_pooling_backward:[27,2,1,"_CPPv219hl_pooling_backward20hl_tensor_descriptorP4realP4real20hl_tensor_descriptorP4realP4real21hl_pooling_descriptor"],hl_pooling_descriptor:[27,4,1,"_CPPv221hl_pooling_descriptor"],hl_pooling_forward:[27,2,1,"_CPPv218hl_pooling_forward20hl_tensor_descriptorP4real20hl_tensor_descriptorP4real21hl_pooling_descriptor"],hl_pooling_mode_t:[27,4,1,"_CPPv217hl_pooling_mode_t"],hl_rand:[27,2,1,"_CPPv27hl_randP4real6size_t"],hl_reset_convolution_descriptor:[27,2,1,"_CPPv231hl_reset_convolution_descriptor25hl_convolution_descriptor20hl_tensor_descriptor20hl_filter_descriptoriiii"],hl_sequence2batch_add:[32,2,1,"_CPPv221hl_sequence2batch_addP4realP4realPiiib"],hl_sequence2batch_copy:[32,2,1,"_CPPv222hl_sequence2batch_copyP4realP4realPiiib"],hl_sequence_avg_forward:[32,2,1,"_CPPv223hl_sequence_avg_forwardP4realP4realPKiiiKi"],hl_sequence_softmax_forward:[30,2,1,"_CPPv227hl_sequence_softmax_forwardP4realP4realPKii"],hl_set_device:[27,2,1,"_CPPv213hl_set_devicei"],hl_set_device_flags_block:[27,2,1,"_CPPv225hl_set_device_flags_blockv"],hl_set_sync_flag:[27,2,1,"_CPPv216hl_set_sync_flagb"],hl_shrink_col2feature:[32,2,1,"_CPPv221hl_shrink_col2featurePK4real6size_t6size_t6size_t6size_t6size_t6size_t6size_t6size_t6size_t6size_t6size_tP4real4real4real"],hl_softmax_backward:[27,2,1,"_CPPv219hl_softmax_backwardP4realP4realii"],hl_softmax_forward:[27,2,1,"_CPPv218hl_softmax_forwardP4realP4realii"],hl_sparse_matrix_add_bias:[30,2,1,"_CPPv225hl_sparse_matrix_add_bias18hl_sparse_matrix_sP4real4real"],hl_sparse_matrix_add_dense:[30,2,1,"_CPPv226hl_sparse_matrix_add_dense18hl_sparse_matrix_sP4realii4real4real"],hl_sparse_matrix_column_sum:[30,2,1,"_CPPv227hl_sparse_matrix_column_sumP4real18hl_sparse_matrix_sii4real"],hl_sparse_matrix_get_cols:[30,2,1,"_CPPv225hl_sparse_matrix_get_cols18hl_sparse_matrix_s"],hl_sparse_matrix_get_rows:[30,2,1,"_CPPv225hl_sparse_matrix_get_rows18hl_sparse_matrix_s"],hl_sparse_matrix_get_value:[30,2,1,"_CPPv226hl_sparse_matrix_get_value18hl_sparse_matrix_s"],hl_sparse_matrix_mul:[30,2,1,"_CPPv220hl_sparse_matrix_mulP4real13hl_trans_op_tP4real13hl_trans_op_t18hl_sparse_matrix_siii4real4real"],hl_sparse_matrix_s:[34,4,1,"_CPPv218hl_sparse_matrix_s"],hl_sparse_matrix_top_k:[30,2,1,"_CPPv222hl_sparse_matrix_top_kP4realiPi18hl_sparse_matrix_sii"],hl_specify_devices_start:[27,2,1,"_CPPv224hl_specify_devices_startPii"],hl_srand:[27,2,1,"_CPPv28hl_srandj"],hl_sse_column_op_with_rem:[30,2,1,"_CPPv225hl_sse_column_op_with_rem3Agg2Op5SaveriiP4realP4realiP4reali"],hl_sse_matrix_column_op:[30,2,1,"_CPPv223hl_sse_matrix_column_op3Agg2Op5SaveriiP4realP4realiP4reali"],hl_sse_matrix_row_op:[30,2,1,"_CPPv220hl_sse_matrix_row_op3Agg2Op5SaveriiP4realiP4realiP4reali"],hl_start:[27,2,1,"_CPPv28hl_startv"],hl_stream_record_event:[27,2,1,"_CPPv222hl_stream_record_event11hl_stream_t10hl_event_t"],hl_stream_synchronize:[27,2,1,"_CPPv221hl_stream_synchronize11hl_stream_t"],hl_stream_t:[34,4,1,"_CPPv211hl_stream_t"],hl_stream_wait_event:[27,2,1,"_CPPv220hl_stream_wait_event11hl_stream_t10hl_event_t"],hl_tensor_descriptor:[27,4,1,"_CPPv220hl_tensor_descriptor"],hl_tensor_reshape:[27,2,1,"_CPPv217hl_tensor_reshape20hl_tensor_descriptoriiiiiiii"],hl_thread_resource:[34,4,1,"_CPPv218hl_thread_resource"],hl_trans_op_t:[34,4,1,"_CPPv213hl_trans_op_t"],hl_vector_abs_sum:[30,2,1,"_CPPv217hl_vector_abs_sumP4realP4reali"],hl_vector_select_from:[30,2,1,"_CPPv221hl_vector_select_fromP1TiPK1TiPKii"],hl_vector_sum:[30,2,1,"_CPPv213hl_vector_sumP4realP4reali"],hppl:[32,4,1,"_CPPv24hppl"],initPaddle:[26,2,1,"_CPPv210initPaddleiPPc"],isGpuVersion:[26,2,1,"_CPPv212isGpuVersionv"],paddle:[54,4,1,"_CPPv26paddle"],staticCastVector:[26,2,1,"_CPPv216staticCastVectorPNSt6vectorI2T2EERKNSt6vectorI2T1EE"],thread_device_resources:[27,4,1,"_CPPv223thread_device_resources"],vecType:[30,4,1,"_CPPv27vecType"]},"paddle.trainer_config_helpers":{attrs:[69,5,0,"-"],data_sources:[70,5,0,"-"]},"paddle.trainer_config_helpers.attrs":{ExtraAttr:[69,6,1,""],ExtraLayerAttribute:[69,7,1,""],ParamAttr:[69,6,1,""],ParameterAttribute:[69,7,1,""]},"paddle.trainer_config_helpers.attrs.ParameterAttribute":{set_default_parameter_name:[69,8,1,""]},"paddle.trainer_config_helpers.data_sources":{define_py_data_sources2:[70,9,1,""]}},objnames:{"0":["c","macro","C macro"],"1":["cpp","class","C++ class"],"2":["cpp","function","C++ function"],"3":["cpp","member","C++ member"],"4":["cpp","type","C++ type"],"5":["py","module","Python module"],"6":["py","attribute","Python attribute"],"7":["py","class","Python class"],"8":["py","method","Python method"],"9":["py","function","Python function"]},objtypes:{"0":"c:macro","1":"cpp:class","2":"cpp:function","3":"cpp:member","4":"cpp:type","5":"py:module","6":"py:attribute","7":"py:class","8":"py:method","9":"py:function"},terms:{"0000x":13,"00186201e":88,"0473v3":76,"05d":9,"08823112e":88,"0ab":74,"0b0":5,"0th":21,"0ul":26,"10007_10":19,"10014_7":19,"100ul":26,"10gbe":3,"12194102e":88,"15501715e":88,"15mb":13,"1636k":21,"16mb":13,"173m":11,"17549435e":34,"1lu":50,"1st":[8,11,19,21,55,74],"1ul":26,"202mb":21,"211839e770f7b538e2d8":76,"2250738585072014e":34,"234m":11,"252kb":13,"25639710e":88,"25k":13,"27787406e":88,"28x28":86,"2nd":[19,21,55,74],"32777140e":88,"32x32":9,"36540484e":88,"38f":34,"3rd":[15,19,21],"40282347e":34,"40gbe":3,"43630644e":88,"48565123e":88,"48684503e":88,"49316648e":88,"51111044e":88,"53018653e":88,"55g":21,"56gbe":3,"6node":7,"6th":21,"70634608e":88,"72296313e":88,"7976931348623157e":34,"85625684e":88,"864k":21,"93137714e":88,"96644767e":88,"99982715e":88,"99m":11,"\u7ea2\u697c\u68a6":8,"\ufb01xed":21,"abstract":[23,46,55,83],"break":[13,65],"byte":[27,58,60],"case":[0,2,13,23,43,60,74],"catch":[46,50],"char":[15,26,27,46,55,66],"class":[9,19,22],"const":[23,26,27,30,32,35,36,38,40,43,46,48,50,52,54,55,58,60,61,62,64,65,66],"default":[1,7,13,15,19,21,26,27,34,40,43,55,60,64,65,66,69,71,74,76,78,83,84,86],"double":1,"else":[62,66],"enum":[26,27,34,40,43,46,52,58,61,63],"export":[1,3,9],"final":[1,15,19,23,36,50,76],"float":[8,11,15,23,26,27,30,34,36,43,46,52,69,71,74,78,84,86],"function":[0,7],"import":[0,8,9,11,13,15,19,21,26,38,71,74,86,88],"int":[13,15,17,23,26,27,30,32,34,36,38,40,43,46,48,50,52,54,55,58,60,61,64,65,66,71,74,76,78,84,86],"interface":69,"long":[17,19,74,76,85],"new":[2,13,17,19],"null":[23,27,30,40,43,46,50,52,58,64,83],"public":[19,23,26,27,30,32,34,35,36,38,40,43,46,48,50,52,54,55,58,60,61,62,64,65,66],"return":[0,11,13,14,15,21,23,26,27,30,34,35,36,38,40,43,46,48,50,54,55,58,60,61,62,64,65,66,70,71,74,76,80,86],"short":[15,17,19,74,76],"static":[26,30,32,35,36,38,40,43,46,48,50,52,54,55,60,61,64,74],"super":23,"switch":19,"true":[0,11,15,17,19,21,23,26,27,32,36,38,40,43,46,52,55,61,62,64,65,67,69,71,74,76,83,84,86],"try":[15,40,66,78],"var":[3,27,43],"void":[23,26,27,30,32,34,35,36,38,40,43,46,48,50,52,54,55,58,60,61,62,64,65,66],"while":[0,13,19,21,46,50,60,61,65,69,71,83,85],__attribute__:54,__device__:32,__init__:[23,26],__inline__:32,__label:[38,43],__m128:30,__m256:32,__main__:[11,88],__meta__:15,__name__:[11,88],__out:[38,43],__queryid:38,__rnn_step__:0,__sparse_get_return__:30,__sparse_get_type_return__:30,__thread:34,__weight:38,_cudnn_convolution_descriptor:27,_cudnn_filter_descriptor:27,_cudnn_pooling_descriptor:27,_cudnn_tensor_descriptor:27,_full_size:40,_global_device_resourc:27,_hl_convolution_descriptor:27,_hl_csc_matrix:30,_hl_csr_matrix:30,_hl_device_prop:27,_hl_event_st:27,_hl_filter_descriptor:27,_hl_matrix_:34,_hl_pooling_descriptor:27,_hl_sparse_matrix_:[34,46],_hl_tensor_descriptor:27,_hl_thread_resourc:34,_link:76,_proj:74,_res2_1_branch1_bn:11,_source_language_embed:[0,8],_target_language_embed:[0,8],_thread_device_resourc:27,a_d:[27,30],a_h:30,aascolvector:46,aasrowvector:46,abc:74,abl:[19,43],about:[13,17,19,21,43,74,76,82,83,85,88],abov:[11,13,17,35,43,46,52,60,61,74,86,88],absderiv:46,absolut:[7,66,85],academ:14,acceler:[43,60,61,84],accept:[13,17,43,58,86,88],acceptor:17,access:0,accesssubnetwork:43,accident:14,accomplish:55,accord:[7,30,38,43,46,48,52,55,64,71,82,83,84,85,86],accordingli:[9,23,88],accordingto:17,account:7,accumcounter_:52,accumul:[27,40,52],accumulatecolsum:46,accuraci:[13,14,19,23,38,71],achiev:[9,60,61],ack:83,acl:[19,43],aclimdb:19,acol:46,acol_:46,across:[38,43,74],act:[0,13,35,40,43,74,76],act_typ:13,actgat:[32,43],actinput:[32,43],action:[14,55,60,61,62],activ:[13,19,23,32,34,35,36,43,55,67,74,76,83,88],activat:31,activation_:43,activationfunct:[35,43],activationgate_:43,activationstate_:43,active:32,active_g:32,active_nod:32,active_st:32,active_state_typ:43,active_typ:43,activegate_:43,activenode_:43,activestate_:43,activi:76,actnod:43,actstat:[32,43],actual:[26,36,40,43,46,48,50,65],actvi:32,actvie_gate_typ:43,adadelta:[13,78],adadeltaparameteroptim:50,adagrad:[13,78],adagradparameteroptim:50,adam:[13,19,21,50,78],adamax:[13,50,78],adamaxparameteroptim:50,adamoptim:[8,13,19,21],adamparameteroptim:50,adapt:[19,21,71],adaptiv:78,adaptive:78,add2:[30,46],add3:46,add:[1,2,13,15,23,27,30,32,43,46,60,61,64,74,76,84,86],add_input:23,add_test:23,add_to:74,add_unittest_without_exec:23,addatoffset:46,addbatchjob:64,addbia:[23,46],addbias:43,addbybitcod:46,addbybitcodebackward:46,addcolscal:46,addcolumnvector:46,addcolvector:46,addcomputethread:40,addcost:61,adddotmul:46,adddotmulmmv2:46,adddotmulmmv:46,adddotmulvmm2:46,adddotmulvmm:46,addelement:46,addgradbegin_:60,addgradi:60,addit:[13,30,43,52,55,60,61,74,76],addition:[13,55],addjob:64,addoptimiz:50,addoutputargu:43,addp2p:46,addparametertyp:[50,54],addprev:43,addr:[58,60],addr_:58,addreallay:43,address:[5,27,30,32,43,46,58,60,83],addrow:[23,46],addrowscal:46,addrowvector:46,addscalar:46,addsharedbia:43,addsquar:46,addsquaresum:46,addtion:7,addto:[46,74],addto_lay:43,addtolay:42,addtorow:46,addtorowsimp:46,addunsharedbia:43,adjac:43,adject:19,adjust:40,adjustbeamsearch:40,adjustprob:40,admin:14,adn:30,adopt:[17,52],advanc:[0,83],advantag:[19,58],adventur:14,adverb:19,affect:[3,74],affin:60,afi:86,after:[0,1,5,7,9,11,13,15,17,19,21,23,26,34,35,40,43,46,50,52,55,60,61,64,74,83,84],afterward:55,again:52,age:14,agent:[40,43],agentlay:[40,42],agentlayernam:40,agg:[30,46],agg_level:74,aggreg:[30,46,55,61,74],aggregat:42,aggregatelevel:74,agument:26,aim:[19,21],aircraft:21,airplan:9,aistats:[43,74],alex:[19,43,74],alexnet_pass1:84,alexnet_pass2:84,algo:[40,43],algorihtm:43,algorithm:[0,8,9],alia:[67,69],align:[21,43,48,60,74,76],align_hint:60,alignblockcount:60,alignbyt:60,alignedalloc:[46,60],alignelementcount:60,alignsiz:46,all:[0,2,3,7,8,11,13,14,15,17,19,21,23,26,27,36,40,43,46,48,50,52,55,58,60,61,62,64,66,69,71,74,78,82,83,84,86],all_rows_in_block:50,allbarrier_:40,allclientpassfinish_:60,allcount:52,allid:[40,43],allids_:43,alloc:[23,26,27,36,43,46,48,52,60],allocat:[27,43,48,84],allocator_:48,allocgradbuf:40,allocsize_:48,allow:[13,23,40,43,60,61,66,83],allow_inefficient_sparse_upd:[46,82,83],allow_only_one_model_on_one_gpu:[82,83,84],allsegments_:55,almost:[7,8,55,60,76],along:19,alpha:[27,30,32,46],alpha_:[43,50],alread:61,alreadi:[7,19,40,46,52,83],alreali:[21,82],also:[0,1,3,7,9,11,13,17,19,23,30,40,43,55,58,60,61,71,74,76,85,86],alwai:[21,36,40,74,76,83,88],amaz:9,amazon:[13,19],ambigu:17,american:9,among:19,amount:[19,43,46],analysi:[12,17],analyz:[13,19],ancestor:43,ani:[0,2,13,15,21,26,36,38,43,66,74,76,85,86],animat:14,annot:60,annotat:17,annual:17,anonym:61,anoth:[17,19,43,65,66,74],another:19,answer:[17,38],any:74,anyth:17,api:[1,13,15,19,23,25],apo:21,appar:21,appear:17,append:[0,7,15,23,46,86],appl:74,appli:[9,13,23,26,43,46,50,54,61,66,74,76],applic:[3,19],apply:[0,13,23],applybinari:46,applycol:46,applydecay_:50,applyl1:46,applyl1decai:46,applyl2:46,applyquaternari:46,applyrow:46,applyternari:46,applyunari:46,apport:52,appreci:[2,19],approach:74,apt:[1,3,5,9],arbitrari:[43,74],architecur:19,area:38,arg:[9,11,13,15,17,19,40,52,64,70,71,74,76,78,82,86],arg_nam:74,argc:26,argu:[17,52],arguemnt:26,argugroup:52,argument:[0,8,9,11],argumentspriv:26,argv:[11,26],arithmet:46,arm:55,around:[74,86],arow:46,arow_:46,arrai:[11,26,30,36,40,54,74,88],arrang:30,arriv:60,art:17,articl:85,artist:14,arxiv:[19,43,50,76],aspect:19,assequencegener:26,assert:[26,88],assign:[30,40,43,46,50,61,83],assignatoffset:46,assignemnt:46,associ:[19,21],associat:17,assum:[0,7,8,9,43,46,84],assume:[7,40,43,64],assur:85,async:[55,60,61,78,82],async_count:83,async_lagged_grad_discard_ratio:83,async_lagged_ratio_default:[82,83],async_lagged_ratio_min:[82,83],asynccount:52,asynccount_:52,asyncfinishpass:[55,60],asyncgrdientcommitcheckandstat:60,asynchron:[27,30,36,52,55,61,64,83],asynchroniz:60,asynclaggedgradientsnum_:60,asynclaggedthreshold_:60,asyncparamet:52,asyncsgd:60,asyncstartpass:55,asynctrainercommitstat_:60,asynctrainerdiscardstat_:60,asynctrainersteps_:60,asyncupdaterstep:60,asyncupdatersteps_:60,asyncupdatestat_:60,asyncupdatesteps_:60,asynsgd:60,atal:46,ateo:40,atlas:1,atom:[40,60],atomicadd:32,attachcontrolparam:52,attachmajorparam:52,attachminorparam:52,attent:[0,21,74,76],attitud:19,attr:[69,76],attribut:[8,17,23,26,30,32,36,38,40,43,46,48,50,52,54,55,58,60,61,64,66],auc:[38,71,82],auc_evalu:38,aucvalid:42,aucvalidationlay:83,author:11,auto:[15,23,50,62,64,65],autom:21,automat:[0,1,7,15,17,21,23,43,46,65,74,76,82,83],automaticli:74,automobil:9,auxiliari:43,avaiabl:60,averag:[11,13,15,19,21,27,32,38,43,46,50,61,71,74,78,80,83],average_:43,average_test_period:[82,83],averagelay:42,averagelevel:43,averageoptim:50,averagepool:74,averagesparseoptim:50,averagestrategi:43,averg:32,avg:[13,43],avgcost:[13,15,19,21,61,71],avgpool:[13,74],avgpoolbackward:46,avgpoolforward:46,avgpoolprojectionlay:43,avoid:[50,86],avreag:32,avx:[5,32,60],axpi:46,aynchron:61,aysnc:52,b2t:8,b_col:46,b_d:[27,30],b_h:30,b_row:46,b_row_i:46,back:[30,40,55,60,86],backgrad:32,background:3,backward:[0,23,26,27,30,32,35,40,43,46,50,61,62,67,74,76,83,84],backward_first:0,backward_reset_grad:32,backward_state_grad:32,backwardactiv:[23,43],backwardbatch:43,backwardbia:43,backwardcallback:40,backwardcallback_:40,backwardcost:43,backwarddata:43,backwardfilt:43,backwardgate2outputsequ:43,backwardimp:[40,43],backwardimpin:43,backwardoneinput:43,backwardonesequ:43,backwardseqparallel:43,backwardsequ:43,backwardvars_:43,backword:26,bad:60,bag:[13,19],baidu:[1,2,3,5,8],baik:8,baisparamet:43,balanc:83,balasubramanyan:19,bandwidth:60,bank:17,bardward:76,bare:60,barrier:[40,60,66,83],barrier_:66,barrrier:40,bascolvector:[30,46],base:[0,1,3,7,8,13,15,19,21,23,29],basecli:55,baselay:23,basematrix:[23,46],basematrixt:46,basenam:71,baseop:30,basepoolingtyp:[74,76],baseregular:78,basestr:[15,67,69,70,71,74,76,80],basetimer_:50,basevector:46,bashrc:1,basic:[13,14,19,23,40,43,55,74,86],basrowvector:[30,46],batch:[7,9,11,13,15,17,19,21,23,26,27,30,32,36,38,40,43,46,50,52,55,61,64,71,74,76,78,83,86],batch_0:11,batch_norm:[43,74],batch_norm_lay:43,batch_norm_typ:74,batch_siz:[7,8,9,13,15,19,21,27,43,78],batchaddto:46,batchcount:32,batchgrad_:43,batchgradientmachin:52,batchid:[43,46],batchid_:60,batchindex:32,batchnorm:[43,52],batchnormalizationlay:42,batchnormbaselay:42,batchsiz:[23,26,30,32,40,43,54,61,74],batchsize_:61,batchstartpositions_:43,batchstatu:[55,61],batchstatus_:61,batchtranspos:30,batchvalu:43,batchvalue_:43,bcd:74,bcol:46,bcol_:46,beam:[0,17,21,30,40,43,74,83],beam_gen:[0,74],beam_search:0,beam_siz:[0,26,74,82,83,84],beamsearch:40,beamsearchcandidatesadjustcallback:40,beamsiz:[26,30],beamsize:21,becaus:[0,9,11,13,17,23,36,40,43,46,50,55,60,65,74,84,88],becom:40,been:[2,9,13,17,19,21,38,40,43,46,55,64,83,86],befor:[2,7,9,15,19,21,26,27,35,36,38,40,43,46,50,52,54,55,61,64,65,66,74,76,88],begin:[23,32,38,43,50,60,71,74,88],begin_dim:60,begin_id:26,begindim:52,beginn:0,beginpad:[32,46],beginpad_:43,beginpo:[46,52],behaivor:40,belong:[21,36,43,74],below:[0,7,9,13,15,23,60,74,86],benefit:76,bengio:[43,74],benifit:[55,60],bertolami:19,besid:[21,74,85],best:[1,13,15,19,21,26,43,70,74,83],beta1:78,beta1_:50,beta2:78,beta2_:50,beta:[11,27,30,32,43,46],beta_:[43,50],better:[7,15,43,58,60,74,76,86],between:[7,13,14,19,21,27,43,46,50,55,60,61,74,78],beyond:52,bgr:11,bia:[0,11,23,27,32,43,46,74,76,78],bias:[23,43,74],bias_:43,bias_attr:[0,74,76],bias_data:27,bias_fg:43,bias_grad_data:27,bias_ig:43,bias_o:43,bias_og:43,bias_param_attr:76,bias_r:43,bias_z:43,biasdesc_:43,biases_:[23,43],biasgrad:27,biaslay:40,biasoffset_:43,biasparamet:43,biasparameter_:[23,43],biassiz:23,bidirect:[0,17,18],bidirectional_lstm_net:19,big:[46,60],biggerthan:46,biggerthanscalar:46,biggest:19,bin:[1,7,15],binari:[3,8,13,19,30,38,43,46,71,74,86],binaryclassificationerror2:46,binaryclassificationerror:46,binarylabelcrossentropi:46,binarylabelcrossentropybp:46,bind:[58,60],bird:9,bit:[13,46],bitext:21,blank:74,blank_:43,blas:1,bleu:20,block:[11,19,23,27,43,50,55,58,60,64,65,66,74,83],block_expand:74,block_expand_lay:43,block_i:74,block_x:74,blockdim:34,blockexpandlay:42,blockh:[32,46],blockh_:43,blockid:60,blockidmap_:60,blockinfo:60,blockinfos_:60,blockingqueu:64,blockkei:60,blockkeyhash:60,blocklengths_:58,blocklocks_:46,blockmap:60,blockmap_:60,blocknum:[46,52],blocknum_:46,blockoffsetmap_:60,blockseg:60,blockseq_:46,blocksiz:60,blocktravers:60,blockvec:50,blockw:[32,46],blockw_:43,blog:19,bn_bias_attr:76,bn_layer_attr:76,bn_param_attr:76,bnbia:[27,43],bnparamdesc:27,bnparamdesc_:43,bnscale:[27,43],bollen:19,bool:[13,15,19,23,26,27,30,32,34,36,38,40,43,46,50,52,54,55,58,60,61,62,64,65,66,67,69,71,74,76,78,83,84,86],bool_const:46,boot:0,boot_lay:0,bootbiaslay:43,bootlay:40,bos_id:[0,74],both:[0,11,13,19,23,30,40,43,46,52,58,67,69,74,76],bottleneck:11,bottom:19,bow:[13,19],bpropact:43,bpropbias:43,bpropsharedbia:43,bpropweight:43,branch:[2,43,74],breadth:[21,83],breath:1,brelu:[35,46,67],breluderiv:46,brendan:19,bridg:[55,61],brief:36,broadcast:55,brow:46,brow_:46,bryan:19,buf:[26,46,58],buf_:48,buffer:[32,36,40,46,52,55,58,60,61,83,86],buffersizes_:40,bufs_:52,bui:19,builder:3,built:[17,52],bunch:13,bunk:19,burg:43,busi:66,button:2,bwddataalgo_:43,bwddatalimitbyt:27,bwddatalimitbytes_:43,bwdfilteralgo_:43,bwdfilterlimitbyt:27,bwdfilterlimitbytes_:43,c_col:46,c_d:[27,30],c_h:30,c_row:46,c_row_i:46,cach:[13,15,27,36,85],cache_pass_in_mem:[13,15,86],cacherowcpumatrix:46,cachetyp:[13,15,36,86],cacul:[40,46],caffemode_:43,cal_matrix_start_address:46,calc:[38,46],calc_batch_s:86,calc_id:40,calcclientid:55,calcerror:38,calcgrad:43,calclearningr:50,calcndcg:43,calcparameterblocks:55,calcuat:60,calcul:[0,15,23,27,30,38,40,43,46,48,52,55,60,71,74,76,78,83,84,86],calfeaturemaps:43,call:[0,9,11,13,19,21,23,27,30,34,35,36,38,40,43,46,50,52,55,58,60,61,64,66,74,76,83,86],callabl:74,callback:[23,26,40,43,50,52,58,60,62],callbackvec_:60,caller:[40,46,50,52,64],calmeanandstd:43,calmovingmeanandvar:43,caltech:9,calul:46,can:[0,1,2,3,4,5,7,8,9,11,13,15,17,19,21,23,26,30,34,36,38,40,43,46,48,50,54,55,58,60,61,62,64,65,66,67,69,70,71,74,76,82,83,84,85,86,88],can_over_batch_s:86,candid:[40,43,52,74],cannot:[23,26,40,60],caoi:21,capabl:[19,27,34,40],capac:65,capapc:64,caption:21,captur:7,card:[7,52],care:[14,76,82,83],carefulli:[7,11,64,83],cascolvector:[30,46],casrowvector:[30,46],cast:26,cat:[9,11,19],catchupwith:[50,54],categor:17,categori:[13,19,74],cblas_gemm:32,cblas_transpose:46,ccb2_pc30:21,ccol:46,ccol_:46,cde:74,ceil:34,cell:[19,43,74,76],center:86,cento:3,certain:[17,82,85],chain:23,chanc:[13,23],chang:[0,2,9,13,19,23,26,40,43,46,50,55,61,65,74,83],channel:[7,11,32,43,46,57],channel_:58,channels_:43,channeltyp:58,channl:[7,11],char_bas:15,charact:[13,15],characterist:[9,38,84],check:[1,2,6],check_eq:23,check_le:23,check_sparse_distribution_batch:[82,83],check_sparse_distribution_in_pserv:[82,83],check_sparse_distribution_ratio:[82,83],check_sparse_distribution_unbalance_degre:[82,83],checkdatahead:36,checkf:32,checkfg:[32,34],checkfg_:43,checkfggrad:[32,34],checkfggrad_:43,checkfgrad:32,checkgrad:83,checkgrad_ep:83,checki:32,checkig:[32,34],checkig_:43,checkiggrad:[32,34],checkiggrad_:43,checkigrad:32,checkindex:46,checkindic:46,checko:32,checkog:[32,34],checkog_:43,checkoggrad:[32,34],checkoggrad_:43,checkograd:32,checkowner:64,checkowner_:64,checksampl:36,checkstores:46,checksubset:52,child:[43,55,58],children:14,chmod:1,choic:14,choos:[13,15,83],chosen:[14,21,85],chpasswd:3,chri:43,christoph:43,chung:43,chunk:[17,38,71],chunk_schem:71,chunktyp:[38,71],cifar:9,cifar_vgg_model:9,circular:43,circularconv:46,circularconvderiv:46,class1:19,class2:19,class_dim:19,classfic:[11,19],classfiic:9,classic:74,classifc:19,classifi:[9,11,13,19,71],classification_cost:[9,13],classification_error_evalu:[13,19,21,38],classification_error_printer_evalu:38,classification_threshold:71,classificationerror:[30,46],classificationerrormulti:46,classnam:26,classregistrar:[36,38,43],clean:[15,88],clear:[36,43,46,50,62,64],clearcount:52,cleargradi:52,clearindic:46,clearrow:46,clearunusedseg:60,clearupdat:52,clearvalueupd:52,cleric:14,click:[2,38],client:[2,38,44],clientid:55,clients_:55,clip:[13,46,69,78,83],clone:1,close:[58,86],cluster_train:7,cmake:[1,23],cmakelist:23,cmd:[3,60],cmrprojectionnormlay:42,cnn:[11,13],cnt:52,code:[0,1],codelength:46,codelength_:43,coeff:74,coeff_:43,coeffici:[43,74],col:[26,30,34,36,46],col_:30,col_siz:30,colidx:38,colin:46,collect:[14,40,43],collectbia:[23,46],colleg:14,colmax:46,colmerg:46,colnum:46,color:[9,11],cols_:46,cols_d:30,colscal:46,column:[8,21,23,27,30,32,34,38,43,46,55,60,71,74],column_sum_evalu:38,columnoffset:46,columoffset:46,colunm:21,com:[1,2,3,5,11,76],combin:[15,19,26,36,40,74,76,77],combinedevalu:38,come:19,comedi:14,comma:[8,30,83],command:[1,2,7,8,9,11],commandlin:19,commenc:13,comment:[13,19,35,43,62,76],commnun:7,common:[0,23,35,38,58,82],common_util:[7,15],commonli:[0,84],commun:[7,23,40,55,58,60,61],compact:27,compani:19,compar:[13,23,46,60],compat:[11,86],compet:19,compil:[1,2,3,23,26,60],complet:[13,23,27,43,64,74,76,88],complex:[0,13,43,85,86],complic:74,compon:23,compos:17,composecallback:50,comput:[0,3,13,15,17,19,23,27,34,38,40,43,46,55,58,60,61,74,76,84],computation:0,computethread:40,computethread_:40,computint:40,comupt:43,concat2:43,concat:[21,42],concat_lay:0,concaten:[43,76],concatenatelay:42,concatenatelayer2:42,concept:[0,66,86],concurr:[60,61],concurrentremoteparameterupdat:60,condit:[0,7,21,40,43,55,60,66,74],condition_vari:66,conf:[7,8,11,21,26,43,74,88],confer:43,confid:19,config:[3,7,8,9,11,13],config_:[36,38,40,43,52,60,61,83],config_arg:[11,13,17,19,82,83,84],config_fil:17,config_gener:[7,15],config_lay:23,config_pars:[23,88],configmap_:60,configpath:26,conflict:[2,50],confpadding_:43,confpaddingy_:43,confptr:26,confront:21,congest:83,conll05st:17,conll:17,connect:[7,9,11,13,15,19,23,40,42],connectionist:[19,74],connnect:60,connor:19,consecut:[27,43],consequ:[74,76],consid:[9,43,46,74,78,84],consider:[76,86],consist:[9,11,13,17,21,30,36,43,46,52],constant:[23,60,66],constantslots_:36,constexpr:60,construct:[0,15,30,43,46,61,64,65,66,86,88],construct_featur:15,constructor:[23,40,43,46,48,55,58,61],consum:19,contain:[0,3,7,9,11,13,14,19,21,26,32,36,38,40,43,46,55,58,61,70,71,74,77,80,86],contemporan:19,content:[17,19,60],context:[0,8,13,15,17,19,21,32,34,43,58,60,74,76],context_attr:76,context_len:[13,15,74,76],context_length:43,context_lenth:43,context_proj_layer_nam:76,context_proj_param_attr:76,context_project:[15,43],context_start:[13,43,74,76],contextlay:43,contextlength:[32,46],contextproject:42,contextprojectionbackward:46,contextprojectionbackwarddata:46,contextprojectionbackwardweight:46,contextprojectionforward:46,contextstart:[32,46],contibut:2,contigu:46,continu:[5,64,66,83],contrast:[21,43],control:[21,40,52,55,58,60,61,69,83],controllerthread_:61,controlparam_:52,controlupd:52,conv:[27,42],conv_act:76,conv_batchnorm_drop_r:76,conv_bias_attr:76,conv_filter_s:76,conv_layer_attr:76,conv_num_filt:76,conv_op:74,conv_oper:43,conv_pad:76,conv_param_attr:76,conv_shif_lay:74,conv_shift:74,conv_shift_lay:43,conv_strid:76,conv_with_batchnorm:76,convbaselay:42,convbwddataalgo:27,convbwdfilteralgo:27,convdesc_:43,conveni:46,converg:[7,19,60],convers:46,convert:[0,8,9,11,13,15,36,43,46,86,88],convexcombinationlay:42,convexpand:46,convfwdalgo:27,convluat:43,convolv:43,convoper:42,convshiftlay:42,convshrink:46,cool:86,cooper:[55,60,61],coopi:40,coorditer:42,copi:[1,15,26,27,30,32,40,43,46,52,55,60,61,66],copyandclear:46,copybackseq:43,copybyrowindex:46,copyfrom:46,copyfromcsc:46,copyfromcsr:46,copyfromnumpyarrai:26,copyfromnumpymat:26,copyfromseq:43,copygradtobufferthread:40,copyid:43,copyidandsequenceinfo:43,copyidx:43,copyinarg:40,copyinfo:43,copyoutputgrad:40,copyoutputtootherdevic:43,copyparametersfromdevic:61,copyparameterstodevic:61,copyrow:46,copys:52,copyscattedid:40,copyschedule_:43,copysingleparafromdevic:61,copysingleparatodevic:61,copythread_:40,copyto:46,copytocpu:46,copytogpu:46,copytonumpyarrai:26,copytonumpymat:26,core:[21,55,58,60,69,83,86],corespond:17,corpora:21,corpu:17,correct:[23,38,40,43,46,71,74],correctli:[23,38,71],correl:[9,19],correspond:[0,1,9,14,17,19,21,23,30,40,43,52,86,88],cos_scal:43,cos_sim:[15,43],cosin:[15,43,46,74],cossim:46,cossimderiv:46,cossimlay:42,cossimvecmatlay:42,cost:[13,15,19,21,35,40,42],cost_:60,costlay:42,costum:40,could:[7,13,15,26,36,40,52,60,61,71,74,88],count:[8,15,17,19,21,34,40,50,52,60,66,83,84,86],counter:52,countincrement:52,covari:43,cpickl:[11,15],cpp:[13,15,21,23,30,52],cpu:[1,3,5,11,17,19,21,26,30,32,34,40,43,46,48,52,55,58,60,61,74,83],cpu_mem:[27,34],cpu_memori:46,cpualloc:48,cpuallocator_:48,cpubatch_:36,cpudensematrix:26,cpueos_:40,cpuexpandstartspos_:43,cpugpuvector:46,cpugpuvectorptr:46,cpugpuvectort:46,cpuid_:40,cpuids_:43,cpuinputstartpos_:43,cpuivector:46,cpuivectorptr:46,cpulocalmats_:46,cpumatrix:46,cpumatrixptr:[46,60],cpumemhandleptr:[46,48,55,60],cpumemoryhandl:48,cpuparameters_:61,cpuprob_:40,cpuseq2batchidx_:43,cpuseqendidxinbatch_:43,cpuseqidx_:43,cpusequencedim:[43,52],cpusparsematrix:46,cpusparsematrixptr:46,cpusrc:46,cpuvector:[46,60],cpuvectorptr:[46,60],cpuvectort:46,cpuvectort_:46,craftsman:14,crash:[7,46,83],crazi:7,create_bias_paramet:23,create_input_paramet:23,create_mode_normal:26,create_mode_testing:26,createargu:26,createbycommandlin:26,createbyconfigprotostr:26,createbygradientmachinesharedptr:26,createbymodelconfig:26,createbypaddleargumentvector:26,createbypaddlematrixptr:26,createbypaddlevectorptr:26,createchannel:58,createcpudensefromnumpi:26,createcpuvectorfromnumpi:26,createdens:26,createfromconfigproto:88,createfrompaddlemodelptr:26,createfromprotostr:26,createfromrawptr:26,createfromsharedptr:26,createfromtrainerconfigfil:26,creategpudensefromnumpi:26,creategpuvectorfromnumi:26,creategpuvectorfromnumpi:26,createinframeinfo:40,creatematrix:[55,60],creatematrixrequest:60,creatememoryframeinfo:40,createmod:40,createparallelvector:46,createparameterconfigfromparameterptr:26,createparameterconfigfromparametersharedptr:26,createseqpo:40,createspars:26,createsparsematrix:46,createvector:[55,60],createvectorrequest:60,createzero:26,creation:55,crf:[43,74],crf_:43,crfdecodinglay:42,crflayer:42,crfs_:43,crime:14,critic:19,crop:11,crop_siz:11,cross:[13,17,30,32,43,46,74],crossmapnormalbwd:46,crossmapnormalfwd:46,crow:46,crow_:46,csc:[23,30],csc_col:30,csc_matrix:30,csc_row:30,csc_val:30,csc_valu:30,cslm:21,csr:[23,30,46],csr_col:30,csr_matrix:30,csr_row:30,csr_val:30,csr_valu:30,cstride:27,csv:14,ctc:74,ctc_layer:71,ctcerrorevalu:38,ctclayer:42,ctcs_:43,ctor:[36,40,54],ctrl:[7,15],ctx:17,ctx_0:17,ctx_0_slot:17,ctx_n1:17,ctx_n1_slot:17,ctx_p1:17,ctx_p1_slot:17,cu_ev:27,cub:9,cubla:[27,34],cublas:27,cublas_handl:34,cublashandle_t:[27,34],cuda:[1,5,7],cuda_dir:[82,83],cuda_home:1,cuda_so:3,cudaevent_t:[27,34],cudahostalloc:48,cudastream_t:[27,34],cudnn:[1,5,11,27,34,43,74,83],cudnn_batch_norm:[43,74],cudnn_bn_min_epsilon:43,cudnn_conv:43,cudnn_convolution_descriptor:27,cudnn_desc:34,cudnn_dir:[82,83],cudnn_filter_descriptor:27,cudnn_handl:[27,34],cudnn_pooling_descriptor:27,cudnn_tensor_descriptor:27,cudnnbatchnormalizationforwardtrain:27,cudnnbatchnormlay:42,cudnnconvlay:42,cudnnconvolutiondescriptor_t:27,cudnnconvolutionmode_t:27,cudnndatatype_t:27,cudnnfilterdescriptor_t:27,cudnnhandle_t:[27,34],cudnnpoolingdescriptor_t:27,cudnnpoolingmode_t:27,cudnnpoollay:42,cudnntensordescriptor_t:[27,34],cudnntensorformat_t:27,cuh:29,cumul:74,curand:[27,34],curandgenerator_t:[27,34],curoffset_:60,curpo:43,curpos_:43,current:[0,1,2,5,7,13,19,21,23,26,27,32,40,43,46,50,52,55,58,60,61,62,64,65,67,74,76,78,83,86],current_word:0,currentavgcost:61,currentblockindex:58,currentblockindex_:58,currentcost:[13,15,19,21,71],currentev:[13,15,19,21,71],currently:43,currentsequenceindex_:36,currenttim:[46,52],curv:[9,38],custom:[14,19,23,40,85],customstacktrac:44,cyclic:74,d_d:30,d_h:30,da_:43,daemon:60,dai:21,daili:19,dalla:86,dan:17,dat:7,data_:[43,46],data_at_cpu:46,data_at_gpu:46,data_config:88,data_dir:[8,9,19,21],data_initialz:13,data_lay:[0,9,13,15,17,43,71],data_provid:70,data_server_port:[82,83],data_sourc:70,data_typ:27,dataargs_:40,dataargsframe_:40,dataargssize_:40,databas:19,databatch:36,datacol:32,dataconfig:[36,40],datadim:[43,74],dataformat:36,datahandl:46,datahead:36,dataid:52,dataim:32,datalay:[43,74],datalayerptr:40,datalayers_:40,datamems_:60,datamtx_:43,datamustincpu:40,datanormlay:42,datapool:36,dataprovid:[0,7,15,17],dataprovider_bow:13,dataprovider_emb:13,dataprovider_pow:13,dataproviderconvert:88,dataproviderptr:40,datasampl:36,dataset:[8,9,11,12,13],datasize_:60,datatyp:55,dataupdatemod:55,db_:43,db_lstm:17,dbnparamdesc:27,dcmake_install_prefix:1,dcol:46,dcol_:46,deadlock:86,deal:2,deamon:58,deb:[3,5],debian:3,debug:46,decai:[9,46,78],decay_r:60,decayaddsquar:46,decayaddsquaremul:46,decayedadagradparameteroptim:50,decayl1:46,decayr:46,decayrate_:50,decid:40,decim:43,decimalreciprocal_:43,declar:[15,74,76],decod:[0,17,21,43,55,74,76],decoder_boot:0,decoder_group_nam:0,decoder_input:0,decoder_mem:0,decoder_prev:76,decoder_s:0,decoder_st:[0,76],decor:[23,86],decrement:66,decriptor:43,deduc:55,deep:[9,11,13,17,43],deeper:[11,55],deer:9,def:[0,11,13,15,17,23,26,74,86,88],defalut:[83,84],default_avg_width:46,default_devic:84,default_valu:84,defaultparamtyp:26,defferenct:86,defin:[0,7,8,9],define_matrix_binary_op:30,define_matrix_binary_parameter_op:30,define_matrix_quaternary_op:30,define_matrix_quaternary_parameter_op:30,define_matrix_ternary_op:30,define_matrix_ternary_parameter_op:30,define_matrix_unary_op:30,define_matrix_unary_parameter_op:30,define_py_data_sources2:[9,11,13,15,70,86],defini:21,definit:[8,13,19,86],degradesequ:52,del:15,delai:[43,83],delar:13,delays_:43,delet:[3,36,43],delimit:[14,15,71],delta:[54,55,60,61],demo:[0,3,7,8,9,11],demograph:14,demonstr:[0,15],denom:46,denoms_:43,denot:[13,14,17,43,52,84],dens:[13,15,23,26,30,36,38,40,43,46,50,55,60,61,74,86],dense_vector:[15,86,88],densedata:36,densematrix:[30,46],deploi:[3,7,84],deploy:7,dequ:36,dequeu:[64,65],dereferenc:66,deriv:22,desc:[27,43],descent:[43,74,78],describ:[13,17,23,66],descript:[0,9,14,15,16],descriptor:[27,34,43,58],design:[3,19,46,52,74,86],desir:8,dest:[26,30,46,55],dest_d:[27,30],dest_h:27,destcol:46,destin:[30,32],destroi:[27,43,55,66],destruct:[30,36,64,66],destructor:[23,30,48],detach:[58,64],detail:[0,2,7,8],detect:[38,71],determin:[23,74],dev:[1,3,9,21],devci:27,devel:3,develop:[1,2,21],deverlop:83,deviat:69,devic:[3,21,27,34,40,43,46,48,61,83],device_fmax:30,device_fmin:30,device_mem:27,device_nam:27,device_prop:27,device_r:27,device_resourc:27,device_typ:27,deviceid:[40,43,48,52,84],deviceid_:[40,43,48,52],devices:3,devid:[40,83],devlin:43,dez:19,diagnos:7,diagram:11,dict:[13,15,17,19,21,26,70,86],dict_dim:19,dict_fil:[0,13,17,71,74],dict_nam:70,dictionai:13,dictionari:0,dictsize:21,diff:[30,46],differ:[0,1,3,7,8,9,11,13,19,21,23,38,40,43,46,52,55,60,70,71,74,76,83],digit:86,dim1:[26,43],dim2:[26,43],dim:[8,11,13,19,23,26,30,32,36,43,46],dimens:[8,13,15,19,23,26,27,30,32,43,55,60,67,74,80,84],dimension:[0,13,23,43,86],dimenst:8,dimes:74,dimk:[27,30],dimm:[27,30],dimn:[27,30],dims_:43,dimsv_:43,din:15,dir:[7,11,13,15,19,21,40],direct:[11,17,27,32,43,61,74,76],directions_:43,directli:[7,19,26,27,60,61,76,85,86],directori:[1,3,7,9,11,13,15,17,19,21,55,83,85],diretcoti:11,dirnam:[54,55,61],disabl:[55,60,66],disable_copy:36,disable_copy_and_assign:26,discard:[60,83],discount:74,discov:17,dispatch:[7,40,52,83],dispatchbydeviceid:40,disput:21,distanc:[38,71],distibut:8,distinguish:[7,21],distribut:[3,7,13,17,27,43,46,55,60,74],distribute_test:[82,83],distributedli:23,distributeev:38,div:30,dived:43,divid:[9,21,32,64,78,82],divrowvector:46,divscalar:46,divup:[34,55],diy:40,diy_beam_search_prob_so:[82,83],dllinitmap:40,doc:[1,3,7,74,76,88],docallback:40,dockerfil:3,doctor:14,document:[1,2,9,13,15,17,19],documentari:[14,86],doe:[0,13,15,17,23,43,61,76,86,88],doesn:[21,69,74],dog:[9,11],doinit:52,don:[19,64,76],done:[0,19,40,43,64,74,76],donot:50,dooper:[55,60],dooperationrequest:60,dot:[11,21,43,55,83],dot_period:[9,15,19,21,83,84],dotdiv:46,dotmul:46,dotmul_oper:43,dotmul_project:43,dotmuloper:42,dotmulproject:42,dotmulsquar:46,dotmulsquaresum:46,dotmulsum:46,dotprod:46,dotproduct:46,dotravers:50,dotsquaremul:46,dotsquaresquar:46,dotsquaresum:46,dotsum:43,dotsum_:43,doubl:[1,27,32,34,38,43,60,83,86],doublebuff:36,doublebuffer_:36,down:[13,40],downclip:46,download:1,download_cifar:9,downsampl:9,doxygen:[1,2],dpkg:5,drama:14,draw:11,driver:3,drop:[36,40,52],drop_rat:69,dropcallback:40,droponenod:40,dropout:[13,23,43,69,74],dropout_lay:74,dropout_r:76,dropoutmask_:43,drow:46,drow_:46,dserver:[55,83],dso:27,dso_handl:27,dst:[27,30,32,43,46,55],dstid:40,dtor:36,dtype:[11,26,60,88],dubai:21,due:[14,15],dummydataprovid:36,dummyevalu:38,dummyoptim:50,dump:62,dumpcallback:62,duplic:[14,60],dure:[13,15,17,21,23,27,36,40,43,46,74,82,83,85,86],durn:86,dw_:43,dwith:1,dwith_doc:1,dwith_gpu:1,dwith_swig_py:1,dynamic_cast:[23,46],each:[0,7,8,9,11,13,14,15,17,19,21,23,26,27,30,32,34,36,38,40,43,46,50,52,55,58,60,64,67,71,74,80,83,84,85,86,88],each_feature_vector:67,each_meta:15,each_pixel_str:86,each_sequence:74,each_time_step_output:67,each_timestep:74,each_word:86,eachstepcallback:40,earli:40,eas:11,easi:[7,13,23,43],easier:[3,23],echo:[3,15,19],edit:[38,71],edu:9,educ:14,effect:[83,86],effici:[0,23,40,60,85,86],efg:74,either:[13,15],elaps:27,elec:13,electron:13,elem_dim:74,element:[13,19,21,27,30,36,43,46,54,60,65,71,74,76,86,88],elementcnt_:46,elif:15,elimin:17,elment:46,els:[11,13,15,23,30,46,50,60,64,66],emb:13,embed:0,embedd:17,embedding_lay:[0,13,15],embedding_nam:0,embedding_s:0,empiric:[43,74],emplace_back:23,emploi:[0,14],empti:[36,40,43,62,64,65,74],empty_:64,emul:21,enabl:[3,7,13,27,52,60,69,83],enable_grad_shar:[82,83],enable_parallel_vector:83,enableinttyp:52,enablesharedtyp:52,enablesparseparamet:52,enabletyp:52,enc_proj:[0,76],enc_seq:76,enc_vec:0,encapsul:61,encod:[0,21,55,76],encoded_proj:[0,76],encoded_sequ:[0,76],encoded_vector:0,encoder:21,encoder_last:74,encoder_proj:0,encoder_s:0,end:[0,8,17,19,21,27,36,38,40,43,50,55,60,61,71,74,76,83],end_:[43,46],end_id:26,end_of_job:65,endcol:46,enddim:52,ending:27,endpad_:43,endrow:46,engin:[3,14,43],english:[21,86],enough:61,enqueu:[64,65],enqueue:65,ensenti:74,ensur:[23,30,43,50,55,86],enter:[7,14],entir:[19,74,76],entiti:60,entri:[14,23],entropi:[13,17,30,43,46,74],enumer:[13,44,67,74],environ:[1,3,5,7,9,40,82,83],environment:15,eos:[43,74],eos_id:[0,43,74],eosframelin:40,eosframeline_:40,eosidchecklay:42,epoch:[14,66],eps:43,epsilon:[27,43,78],epsilon_:50,equal:[43,58,60,74,76,78],equat:[23,74,76,78],equation:22,equenc:32,equip:[0,3],erin:43,err:27,error:[5,7,9,11,13,14,15,19,21,23,26,27,30,35,38,40,43,46,69,71,74,78,83],error_clipping_threshold:69,errorrat:71,especi:[17,60,76,86],essenti:[1,17],essential:21,estim:43,estimatedmean:27,estimatedvar:27,etc:[3,19,21,26,34,46,52,54,60,78,82,84],eth0:7,ethernet:3,ethternet:7,euclidean:43,eval1:38,eval:[13,15,19,21,38,40,71],eval_bleu:21,evalimp:38,evalu:[7,13,15,19,21,38,40,71,74,85],evaluate_pass:19,evaluator_:40,evaluator_bas:71,evaluatorconfig:38,evalut:21,evalutaion:20,even:[19,40,60,64,83],event:[27,34],everi:[0,13,17,19,21,23,36,43,52,74,76,83,85,86],everyth:2,exactli:[17,38,43,46,71,74,76,86],exampl:[0,1],exceed:74,except:[8,15,40,66,84,86],exchang:[55,61],excl:43,exconv:43,excut:64,exec:[46,52,64,83],execfunc:[46,52,60],exechelp:64,execplusown:64,exectur:64,execut:[14,17,19,23,27,43,55,60,64,66],exhibit:[58,60,61],exist:[14,19,23,43,46,52,55,61,83],exit:[40,64,83],exp:46,expand:[17,19,21,23,32,40,43,46,74],expand_a:74,expand_lay:43,expand_level:74,expandcolmat:46,expandconvlay:42,expandedin_:43,expandedingrad_:43,expandedout_:43,expandedoutgrad_:43,expandfwdonc:43,expandinput_:43,expandlay:42,expandlevel:[43,74],expandmat:43,expandonefram:43,expandstartspos_:43,expderiv:46,expect:[19,40,61],expectedpasscount:61,expectedpasscount_:61,expectedppasscount:61,expens:21,experi:[4,84],explain:[19,71,86],explan:21,explanat:13,explicit:23,explicitli:40,exploit:9,explor:74,expon:46,exponenti:35,expose:3,express:[19,30,46],expw_:43,expx_:43,extend:[15,60],extens:[14,15,21,78],extern:[40,83],extra:[32,43,60],extraattr:[69,84],extract_fea_c:11,extract_fea_pi:11,extract_para:8,extralayerattribut:[69,74,76],extralayeroutput:76,extrem:74,extremli:85,f_rdma:58,f_tcp:58,fabric:7,facilit:60,facotr:74,fact:[11,61],factor:[27,30,69,74,78],factori:36,fail:[9,40,66,83,84],fals:[0,8,13,15,17,19,21,23,26,27,32,38,40,43,46,50,52,54,55,61,62,66,69,71,74,76,78,83,84,86],false_typ:46,famili:21,familiar:86,fanscin:86,fantasi:14,fantast:13,farmer:14,fascinatingli:85,fast:[2,43],faster:[0,19,40,43,74,76],fastest:43,fat:58,fatal:[36,46],fatur:43,favorit:2,fc1:[23,84],fc2:84,fc3:84,fc4:84,fc_act:76,fc_attr:76,fc_bias_attr:76,fc_layer:[13,15,23,43],fc_layer_nam:76,fc_param_attr:76,fclayer:23,fdata:17,fea:11,fea_output:11,feaimgheight:46,feaimgwidth:46,feat:19,featur:9,feature_map:[15,27],featuremapexpandlay:42,feed:[19,76],feedforward:9,femal:14,fernan:19,festiv:86,fetch:[0,2,23,52,55,61],few:[43,86],fewer:74,fg0:74,fieldscann:36,figur:[0,8,9,11,17,19,21,23],file1:21,file2:21,file:[0,1,2,7,8,9,11],file_list:86,file_nam:[11,13,17],filelist:36,filelist_:36,filenam:[15,36,52,86],filer:74,fill:[13,26,36,40,43,46,74],fillrowindic:46,fillselectivedata:43,fillsequ:46,fillslot:36,film:14,filter:[11,27,32,43,46,74],filter_data:27,filter_grad_data:27,filter_height:27,filter_s:[74,76],filter_size_i:74,filter_width:27,filterchannels_:43,filterdesc_:43,filterpixels_:43,filters:43,filtersize_:43,filtersizex_:43,filtersizey_:43,finali:7,find:[1,9,19,21,30,43,60,62,74,78],fine:[15,60,69],fini:[46,52],finish:[7,9,26,27,36,38,40,43,46,50,55,60,61,64,86],finishbatch:[26,50,54,60,61],finishcatchupwith:50,finishcv_:64,finishfil:36,finishpass:[26,50,54,61],finishprepar:36,finishthread:55,finishtrain:26,finishtrainonebatch:26,finishtrainpass:26,finit:23,first:[0,1,2,3,5,8,9,11,13,15,17,19,21,23,27,30,36,38,40,43,46,55,61,62,64,71,74,83,84,86],first_seq:0,firstli:[7,55],firstorderoptim:50,firsttest_:43,firsttime_:50,fit:[40,85],five:13,fix:[11,21,69,86],flag:[1,9,17,27,43,46,83],flags:83,flags_port:55,flags_ports_num:55,flags_use_gpu:[36,40,43,52],flexibl:[0,74,85],flight:21,float32:[11,26,88],float_value:46,floatarrai:26,fluctuat:60,flush:60,fly:13,focu:[60,86],folder:[9,19,21],follow:[0,1,2,3,4,7,8,9,11,13,14,15,17,19,21,23,38,43,50,60,71,74,76,78,84,85,86],foramt:46,forc:64,forcestop:64,forcibl:64,foreachlay:40,forecast:19,forget:[19,32,43,78],forgetcheck:43,forgetg:43,forgetgate_:43,form:[17,40,43,46,52,61,76,78,85,86],formal:40,format:[8,9],format_:[46,52],former:21,formul:43,formula:[27,43,74],formular:74,forward:[0,2,17,19,23,26,27,30,32,35,40,43,46,50,61,62,67,76,84],forward_final_output:32,forward_reset_output:32,forwardactiv:[23,43],forwardbackward:[26,40,55,60,61],forwardbackwordtime_:55,forwardbatch:43,forwardbia:43,forwardcost:43,forwarddropout:43,forwardgate2outputsequ:43,forwardimp:[40,43],forwardimpin:43,forwardonebatch:26,forwardoneinput:43,forwardonesequ:43,forwardseqparallel:43,forwardsequ:43,forwardtest:88,forwardvars_:43,found:[0,1,9,13,19,26,60,74,86,88],four:[8,11,13,15,17,19,43,86],four_parameter:30,fourth:40,foward:[35,43],fpga:34,frame:[32,34,38,40,71],framecnt:32,frameforgetgate_:43,framegate_:43,frameheight:[43,52],frameinputgate_:43,frameinputnode_:43,frameoutput_:43,frameoutputgate_:43,framepreoutput_:43,frames:[32,43],frames_:40,framestate_:43,framewidth:[43,52],framework:[11,13,19,23,40,58,60],free:[21,27,30,43,46,48],freeall:48,french:21,frequenc:[8,13,19],frequent:[7,21,55],friend:[26,38,40,46,58],frog:9,from:[0,1,2,3],from_timestep:74,fromfil:11,fucntion:64,fuction:64,full:[0,23,43,52,55,58,61,65,74],full_matrix_project:[0,43],fulli:[2,9,11,13,15,19,23,42],fullmatrixproject:42,fullsiz:[54,61],fully_matrix_project:76,fullyconnect:8,fullyconnectedlay:[23,42],func:[46,52,60,64],funcnam:[55,60],further:[3,55,60,74],fusion:15,futru:43,futur:[55,64],fwdalgo_:43,fwdlimitbyt:27,fwdlimitbytes_:43,gain:[60,74],gamma:[11,27,43],gamma_:50,gap:83,gate_:43,gate_act:[74,76],gate_recurr:74,gatedrecurrentlay:42,gategrad:[32,34],gatevalu:[32,34],gateweight:34,gateweight_:43,gateweightgrad:34,gather:[15,23,40,43,74],gatheragentlay:42,gauss:69,gaussian:46,gemm:46,gen1:43,gen:[21,27,34,43,74],gen_conf:21,gen_data:21,gen_input:0,gen_mutex:[27,34],gen_result:21,gen_trans_fil:0,gender:[14,15],generated_sequ:74,generatedinput:[0,74],generatesequ:[26,40],generator_:40,generatorconfig:40,genert:86,genr:[14,15],gentoo:3,gereat:74,get:[0,1,3,5,7,9,11,13,15,17,19,23,26,27,30,36,40,43,46,50,52,55,58,60,61,64,66,74,76,86],get_best_pass:19,get_config_arg:[13,15,19,84],get_convolution_descriptor:27,get_data:[13,17],get_filter_descriptor:27,get_imdb:19,get_input_lay:23,get_model:11,get_output_layer_attr:76,get_tensor_descriptor:27,getabsmax:46,getabssum:46,getaccum:52,getallcount:52,getalldata:55,getallocs:48,getallthread:40,getavgcost:61,getbackwardcallback:40,getbatchs:[23,26,52],getbatchvalu:43,getbiasparamet:43,getblockid:60,getblocklength:58,getblocknum:43,getblockoffset:60,getbuf:[26,48,52],getchannel:58,getcol:46,getcolnum:46,getcolstartidx:46,getcolumn:46,getconfig:[26,36,43,52],getcpualloc:48,getcpustartposit:52,getcublasdsohandl:27,getcudartdsohandl:27,getcudnndsohandl:27,getcuranddsohandl:27,getcurrentavgcost:61,getcurrenttimestick:34,getdata:[26,46],getdeviceid:[40,43,52],getelement:46,getelementcnt:46,getfinalpath:40,getformat:[46,52],getframeheight:52,getframewidth:52,getgpualloc:48,getgradbuf:40,getgradientmachin:40,getgradientseg:52,getheight:[26,46],getid:[26,46,52],getinarg:40,getindexdicthandl:46,getinfolay:43,getinitmean:52,getinitstandarddevi:52,getinput:[23,43],getinputgrad:[23,43],getinputlabel:43,getinputvalu:[23,43],getintbuf:52,getinternalargumentsptr:26,getitem:26,getjob:65,getlabellay:43,getlay:40,getlayeroutput:[26,40],getlearningr:50,getlearnr:52,getlocalindic:46,getlocalparamet:52,getlocalrow:46,getmajorpartn:52,getmat:52,getmax:46,getmemoryhandl:46,getmin:46,getminorpartn:52,getmodelconfig:26,getmutabledata:46,getmutablevector:46,getnam:[26,35,43,48,52,62],getnetworkoutput:26,getnextbatch:36,getnextbatchfrombuff:36,getnextbatchintern:36,getnextblocklength:58,getnextpo:43,getnonstaticparamet:40,getnumbatch:43,getnumblock:58,getnumdevic:40,getnumlogicaldevic:40,getnumprocess:61,getnumsequ:52,getnumsubsequ:52,getnumthread:[40,64],getoptimizationconfig:26,getoutarg:40,getoutput:43,getoutputgrad:[23,43],getoutputlabel:43,getoutputlay:42,getoutputs:43,getoutputvalu:[23,43],getpaddlemodelconfig:26,getparamet:[26,40,43,55,60],getparameterconfig:60,getparameterid:40,getparametermap:40,getparameterptr:[23,52],getparameters:26,getparameterspars:[55,60],getparametersremot:[54,61],getparametertyp:[26,50,54],getpasstyp:40,getpeernam:58,getpoint:46,getprefetchmatrix:52,getprepo:43,getprev:43,getprevbatchoutput:43,getprevbatchst:43,getpserverparametergradi:55,getpserverparametervalu:55,getrawptr:26,getrow:46,getrowbuf:46,getrowcol:46,getrownum:46,getrowstartidx:46,getrowvalu:46,getscor:26,getscorelay:43,getseg:38,getsent:26,getseqlengthandstart:52,getseqoutputfrombatch:43,getsequ:26,getsharedcount:52,getsharedptr:26,getsiz:[23,26,36,43,46,48,52],getslaveparamet:40,getslotid:26,getslotin:26,getslotnum:26,getslotsequencedim:26,getslotsequencestartposit:26,getslotsubsequencestartposit:26,getslotvalu:26,getsparseformat:26,getsparserowcol:26,getsparserowcolsv:26,getsparsevaluetyp:26,getstat:[40,43,61],getstatu:60,getstatusrequest:60,getstrid:46,getsum:46,getsync:46,gettasktyp:40,getthread:40,gettlstempbuf:52,gettmpsparsematrix:46,gettotallength:58,gettransdtyp:55,gettranspos:[23,46],gettyp:43,getvalu:46,getvaluebuf:40,getvaluetyp:46,getvector:46,getw:[23,52],getweight:[23,43],getwgrad:[23,52],getwidth:[26,46],gflag:1,gflags:1,gildea:17,gist:76,git:[1,2],github:1,give:[13,23,40,43,52,86],given:[13,17,19,21,23,40,43,46,55,83,86],global:[15,19,27,43,46,50,55,60,64,69,78,83],global_device_resourc:27,global_learning_rate:69,globalindic:46,globalindices_:46,globe:86,glog:1,goal:17,goe:[74,76],good:[19,21,74],googl:60,gpu:[1,3,5,9,11,15,17,19,21,26],gpu_id:[83,84],gpu_mem:[27,34],gpualloc:48,gpuallocator_:48,gpubatch_:36,gpugpu_id:82,gpuivector:46,gpuivectorptr:46,gpumatrix:46,gpumatrixptr:46,gpumemhandleptr:[46,48],gpumemoryhandl:48,gpusparsematrix:[30,46],gpusparsematrixptr:46,gpuvector:46,gpuvectorptr:46,gpuvectort:46,gpuvectort_:46,gpuworkspac:27,grab:19,grad:[14,27,30,32,35,40,43,46,52,83],grad_d:30,grad_o:30,grad_share_block_num:[82,83],grad_w:30,gradbuff:40,gradbufqueue_:40,gradbufs_:40,gradcollectthread:40,gradcollectthread_:40,gradcount:52,graddata:43,gradfg:32,gradframest:32,gradient:[13,19,22],gradient_clipping_threshold:[13,19,78],gradient_printer_evalu:38,gradientaccum_:52,gradientmachin:[15,21,26],gradientmachine_:40,gradientmachinepriv:26,gradientmatchinecreatemod:26,gradientreadybarrier_:60,gradietn:40,gradig:32,gradin:32,gradmat:52,gradog:32,gradoutput:32,gradprevout:32,gradqueue_:40,gradreadycond:52,gradresetg:32,gradresetoutput:32,gradsegments_:52,gradsem_:52,gradstream_:40,gradterms_:43,gradupdateg:32,grai:9,graident:40,grain:60,gram:[8,19],graph:[8,74],graphviz:11,grave:19,grayscal:86,greater:[64,66,74],greaterpath:40,grep:19,groudtruth:0,ground:[13,21,71,74],group:[19,42],group_id:15,grouplen:14,groups_:43,grow:65,gru:[0,13,21,31],gru_bias_attr:76,gru_decod:0,gru_decoder_with_attent:0,gru_encoder_decod:[8,21],gru_finaloutput:32,gru_layer_attr:76,gru_resetgrad:32,gru_resetoutput:32,gru_siz:13,gru_stategrad:32,gru_step:0,gru_step_lay:[0,43],grucomput:42,grumemori:[0,43],grusteplay:42,gserver:23,gtest:1,guarante:23,guess:19,guid:[0,2],gur_group:76,hack:[4,7],half:43,hand:[14,15,19],handl:[7,15,19,27,34,40,43,47],handler:27,handlerequest:[58,60],handlerequestbegin_:60,handwrit:[19,86],happen:40,har:[7,58,60],hard:[13,26,46],hardward:60,hash:[46,50],hasnonstaticcpuparamt:40,hasnonstaticcpuparamters_:40,hasstaticparamet:40,hassubseq:[40,52],hastyp:52,have:[0,1,2,3,5,7,8,9,13,14,15,19,21,23,36,38,40,43,46,52,55,64,65,71,74,83,84,85,86,88],hdfs:85,head:[2,8,19],header:[8,11,15,23,36,52,58],header_:36,headerptr:36,headerptr_:36,health:14,heavi:7,height:[9,23,26,27,30,32,40,43,46,52,74],height_:[36,46],height_pad:27,heightpad:43,heightstor:46,help:[4,7,60,61,86,88],helper:[23,40,60,70,74,76],helpful:52,here:[0,1,7,8,9,11,12,13,14,15,17,19,21,27,36,43,50,52,60,69,70,74,76,82,84,86,88],heterogen:34,heurist:[21,74,83],hidden:[0,13,15,19,43,60,61,74,76],hidden_s:[15,76],hide:61,hierarch:[0,43,74],hierarchicalsigmoidlay:42,high:[3,23,34,46,60,69],higher:85,highest:21,highli:[0,19,84,85,86],highly:15,himself:55,hint:88,histogram:46,histogram_exponent:46,histogramtyp:46,histor:19,histori:40,histroi:60,hl_activation_end:34,hl_activation_funct:32,hl_activation_linear:34,hl_activation_mode_t:[32,34,43],hl_activation_relu:34,hl_activation_sigmoid:34,hl_activation_tanh:34,hl_agg_op:30,hl_aggreg:29,hl_avgpool_backward:32,hl_avgpool_forward:32,hl_avx_gru_backward_reset_grad:32,hl_avx_gru_backward_state_grad:32,hl_avx_gru_forward_final_output:32,hl_avx_gru_forward_reset_output:32,hl_base:33,hl_batch_norm_backward:27,hl_batch_norm_forward_infer:27,hl_batch_norm_forward_train:27,hl_batch_transpos:29,hl_check_align:30,hl_cmrnorm_backward:32,hl_cmrnorm_forward:32,hl_construct_sparse_matrix:30,hl_context_projection_backward_data:32,hl_context_projection_backward_weight:32,hl_context_projection_forward:32,hl_conv_workspac:27,hl_convolution_backward_bia:27,hl_convolution_backward_data:27,hl_convolution_backward_filt:27,hl_convolution_descriptor:[27,43],hl_convolution_forward:27,hl_convolution_forward_add_bia:27,hl_cossim:30,hl_cossim_deriv:30,hl_cpu_apply_binary_op:30,hl_cpu_apply_quaternary_op:30,hl_cpu_apply_ternary_op:30,hl_cpu_apply_unary_op:30,hl_cpu_gru_backward:32,hl_cpu_gru_cuh_:32,hl_cpu_gru_forward:32,hl_cpu_lstm_backward:32,hl_cpu_lstm_forward:32,hl_cpu_matrix_column_op:30,hl_cpu_matrix_row_op:30,hl_create_convolution_descriptor:27,hl_create_ev:27,hl_create_filter_descriptor:27,hl_create_global_resourc:27,hl_create_pooling_descriptor:27,hl_create_tensor_descriptor:27,hl_create_thread_resourc:27,hl_csc_matrix:30,hl_csr_matrix:30,hl_cublas_init:34,hl_cuda_cudnn_ph_:27,hl_cuda_event_is_readi:27,hl_cuda_ph_:27,hl_cudnn_desc_init:34,hl_cudnn_init:34,hl_destroy_convolution_descriptor:27,hl_destroy_ev:27,hl_destroy_filter_descriptor:27,hl_destroy_pooling_descriptor:27,hl_destroy_tensor_descriptor:27,hl_destruct_sparse_matrix:30,hl_device:30,hl_device_can_access_p:27,hl_device_enable_peer_access:27,hl_device_functions_cuh_:32,hl_device_prop:27,hl_device_synchron:27,hl_event_elapsed_tim:27,hl_event_synchron:27,hl_event_t:27,hl_expand_feature2col:32,hl_filter_descriptor:[27,43],hl_fini:27,hl_float_max:34,hl_float_min:34,hl_float_value:[30,34],hl_free_mem_devic:27,hl_free_mem_host:27,hl_free_sparse_matrix:30,hl_get_cuda_lib_vers:27,hl_get_cudnn_lib_vers:27,hl_get_devic:27,hl_get_device_count:27,hl_get_device_error_str:27,hl_get_device_last_error:27,hl_get_sync_flag:[23,27],hl_gpu_apply_binary_op:30,hl_gpu_apply_quaternary_op:30,hl_gpu_apply_ternary_op:30,hl_gpu_apply_unary_op:30,hl_gpu_functions_cuh_:32,hl_gpu_gru_backward:32,hl_gpu_gru_cuh_:32,hl_gpu_gru_forward:32,hl_gpu_lstm_backward:32,hl_gpu_lstm_cuh_:32,hl_gpu_lstm_forward:32,hl_gpu_matrix_column_op:30,hl_gpu_matrix_row_op:30,hl_gru_grad:[32,34,43],hl_gru_op:32,hl_gru_ops_cuh_:32,hl_gru_valu:[32,34,43],hl_init:[27,34],hl_lstm_grad:[32,34,43],hl_lstm_op:32,hl_lstm_ops_cuh_:32,hl_lstm_parallel_backward_data:32,hl_lstm_parallel_backward_weight:32,hl_lstm_parallel_forward:32,hl_lstm_valu:[32,34,43],hl_malloc_devic:27,hl_malloc_host:27,hl_malloc_sparse_matrix:30,hl_matrix:29,hl_matrix_:34,hl_matrix_add:30,hl_matrix_add_to_row:30,hl_matrix_appli:29,hl_matrix_apply_h_:30,hl_matrix_bas:29,hl_matrix_base_cuh_:30,hl_matrix_classification_error:30,hl_matrix_column_max:30,hl_matrix_column_min:30,hl_matrix_column_op:30,hl_matrix_column_sum:30,hl_matrix_cross_entropi:30,hl_matrix_cross_entropy_bp:30,hl_matrix_csc2dens:30,hl_matrix_csc_mul_dens:30,hl_matrix_csr2dens:30,hl_matrix_csr_add_bia:30,hl_matrix_csr_add_dens:30,hl_matrix_csr_column_sum:30,hl_matrix_csr_mul_dens:30,hl_matrix_dense_mul_csc:30,hl_matrix_dense_mul_csr:30,hl_matrix_format_t:[30,34],hl_matrix_mul:27,hl_matrix_mul_vector:27,hl_matrix_op:29,hl_matrix_ops_cuh_:30,hl_matrix_row_max:30,hl_matrix_row_min:30,hl_matrix_row_sum:30,hl_matrix_select_row:30,hl_matrix_softmax:30,hl_matrix_softmax_deriv:30,hl_matrix_top_k:30,hl_matrix_transpos:27,hl_matrix_typ:29,hl_matrix_type_cuh_:30,hl_matrix_value_t:[30,34],hl_matrix_zero_mem:30,hl_max_sequence_backward:32,hl_max_sequence_forward:32,hl_maxpool_backward:32,hl_maxpool_forward:32,hl_memcpi:27,hl_memcpy_async:27,hl_memcpy_csc_matrix:30,hl_memcpy_csr_matrix:30,hl_memcpy_device2devic:27,hl_memcpy_device2host:27,hl_memcpy_from_csc_matrix:30,hl_memcpy_from_csr_matrix:30,hl_memcpy_host2devic:27,hl_memcpy_sparse_matrix:30,hl_memset_devic:27,hl_naive_gru_backward_reset_grad:32,hl_naive_gru_backward_state_grad:32,hl_naive_gru_forward_final_output:32,hl_naive_gru_forward_reset_output:32,hl_no_value:[30,34],hl_param_relu_backward_diff:30,hl_param_relu_backward_w:30,hl_param_relu_forward:30,hl_pooling_average:27,hl_pooling_average_exclude_padding:27,hl_pooling_backward:27,hl_pooling_descriptor:[27,43],hl_pooling_end:27,hl_pooling_forward:27,hl_pooling_max:27,hl_pooling_mode_t:[27,43],hl_rand:27,hl_recurrent_apply_cuh_:32,hl_reset_convolution_descriptor:27,hl_sequence2batch_add:32,hl_sequence2batch_copi:32,hl_sequence_avg_forward:32,hl_sequence_softmax_forward:30,hl_set_devic:27,hl_set_device_flags_block:27,hl_set_sync_flag:27,hl_shrink_col2featur:32,hl_softmax_backward:27,hl_softmax_forward:27,hl_spars:29,hl_sparse_csc:[30,34],hl_sparse_csr:[30,34],hl_sparse_end:34,hl_sparse_matirx_column_sum:30,hl_sparse_matrix_:[30,34],hl_sparse_matrix_add_bia:30,hl_sparse_matrix_add_dens:30,hl_sparse_matrix_column_sum:30,hl_sparse_matrix_get_col:30,hl_sparse_matrix_get_row:30,hl_sparse_matrix_get_valu:30,hl_sparse_matrix_mul:30,hl_sparse_matrix_s_ptr:46,hl_sparse_matrix_top_k:30,hl_sparse_ph_:30,hl_specify_devices_start:27,hl_srand:27,hl_sse_column_op_with_rem:30,hl_sse_matrix_column_op:30,hl_sse_matrix_kernel:29,hl_sse_matrix_kernel_cuh_:30,hl_sse_matrix_row_op:30,hl_start:27,hl_stream_record_ev:27,hl_stream_synchron:27,hl_stream_t:[27,30,34,40,46,52],hl_stream_wait_ev:27,hl_table_appli:29,hl_tensor_descriptor:[27,43],hl_tensor_reshap:27,hl_thread:33,hl_thread_ph_:34,hl_thread_resourc:34,hl_time:33,hl_top_k:29,hl_trans_op_t:[27,30,34],hl_value_end:34,hl_vector_abs_sum:30,hl_vector_select_from:30,hl_vector_sum:30,hold:[40,43,46,58],home:[1,7],homemak:14,hook:[15,17,50,52,54],horizont:[11,27,74],horror:14,hors:9,horst:19,host:[3,27,30,46,55,58,60,61],hostnam:7,hosts:7,hostsrc:46,hot:15,hour:21,hous:[8,86],how:[0,9,11,13,15,38,40,55,60,69,74,83,85],howev:[0,2,19,21,40,76,82,83,86],hppl:[27,32,33],hppl_active_function:32,hppl_op_end:34,hppl_op_n:[27,34],hppl_op_t:[27,30,34],hppl_stream_1:34,hppl_stream_2:34,hppl_stream_3:34,hppl_stream_4:34,hppl_stream_:34,hppl_stream_default:[34,46,52],hppl_stream_end:34,hppl_thread_stream_1:34,hppl_thread_stream_2:34,hppl_thread_stream_3:34,hppl_thread_stream_4:34,hsigmod_lay:43,hstride:27,html:9,http:[1,2,3,5,9,11,21,50,76],huang:43,huber:[43,74],hubertwoclass:42,huge:[61,74],huina:19,human:21,hundr:55,hyper:[23,74],i0601:15,i0706:21,i0719:21,ib0:7,ibasematrix:46,icpugpuvector:46,icpugpuvectorptr:[40,43,46,52],icwsm:19,id_input:[0,74],idea:[43,55,74],ident:[14,30],identifi:[0,23,55],identiti:67,identity_project:43,identityactiv:35,identityoffsetproject:42,identityproject:42,idindex:[40,43],idindex_:43,ids:40,ids_:43,idsarray_:46,idsiz:43,idsize_:43,idsvec_:43,idx:[23,26,43,46,54,60],idxbuf:26,ieee:19,ignor:[30,38,43,46,71,83,86],igradientmachinemod:40,iiddata:36,ijcnlp:19,illustr:[0,13,23,86],ilsvrc:11,image_classif:9,image_desc:27,image_list_provid:11,image_provid:9,image_s:11,imageh_:43,imagepixel:43,images:43,imagew_:43,imdb:[14,18],imdber:19,img:9,img_conv_lay:43,img_norm_typ:74,img_pool_lay:43,img_siz:9,imgpixels_:43,imgsiz:43,imgsize_:43,imgsizeh:46,imgsizeh_:43,imgsizew:46,imgsizew_:43,imgsizex_:43,imgsizey_:43,immedi:[36,46,55,60],implement:[0,1,7,11,13,17,22],implemnt:43,implent:43,implicitli:36,implict:52,implictli:[60,61],impliment:61,important:36,importerror:15,improv:[19,21],in1:43,in2:43,in_:43,inarg:[26,40,88],inargs_:40,inargscopied_:40,incb:27,incc:27,includ:[0,1,3,5,8,13,14,17,21,23,26,32,34,35,36,40,43,46,48,50,52,54,55,58,74,76,83,85,86],inconsist:14,incopyidx:43,incorrect:[43,74],increas:[21,23,27,83],increment:[66,83],incshar:52,incupdat:[23,52],incupdatecount:40,indac:40,inde:60,indeedli:61,indeped:43,independ:[13,36,43,74],index:[7,15,26,30,32,36,38,40,43,46,54,60,71,74,86],indexdata:36,indexdict:46,indexdicthandl:46,indexdicthandle_:[40,46],indexdictptr:46,indexslot:17,indic:[7,17,30,36,38,40,43,46,58,60,71,74,86],indicat:[7,55],indicationg:60,indiff:32,indirectli:60,individu:61,ineffici:[46,83],inf:40,infer:13,infiniband:7,info:[2,7],info_:40,infom:2,infomatrion:83,inform:[3,14,15,17,19,21,23,43,52,71,83,85,88],inframelin:40,inframelines_:40,ingor:83,ingrad:27,ingrad_:43,ingraddesc:27,inherit:[26,30,36,38,40,43,46,48,50,52,54,55,58,60,61,64,66,67],inidx:43,init:[13,15,23,26,27,30,34,38,40,43,46,50,52,54,55,60,61,69,84],init_hook:[13,15,17,85],init_hook_wrapp:70,init_model_path:[8,13,82,83,84],initasyncload:36,initblock:46,inithook:52,initi:[0,8,13,23,27,40,52,60,61,64,66,69,74,83,86,88],initial:[23,26,27,34,43,52,64,66],initial_max:69,initial_mean:[52,69,74],initial_min:69,initial_std:[52,69,74],initial_strategi:52,initil:43,initneedflag:43,initpaddl:[26,88],initshar:46,initsubnetwork:43,initthread:55,initthreadnum:64,initvalu:66,inlay:40,inlcud:43,inlength:43,inlin:30,inline:[30,32],inlinks_id:40,inner:[23,43],inner_param_attr:76,inpidx:43,inplac:26,inpserv:50,inpt:30,input1:[30,43,76],input1_height:30,input2:[30,43],input2_height:30,input:[0,8,9,11,13,15,17,19,21,23,26,27,30,32,34,35,36,38,40,43,46,52,55,60,61,64,67,71,74,76,80,81,84,86,88],input_:43,input_data:[23,27],input_data_grad:27,input_data_target:23,input_featur:67,input_feature_map:27,input_fg:43,input_fg_:43,input_fil:17,input_hassub_sequence_data:23,input_id:[43,74],input_ig:43,input_ig_:43,input_imag:[9,27,76],input_image_grad:27,input_index:23,input_label:23,input_lay:23,input_og:43,input_og_:43,input_s_:43,input_sequence_data:23,input_sequence_label:23,input_sparse_float_value_data:23,input_sparse_non_value_data:23,input_t:23,input_typ:[13,15,85],inputargument_:43,inputbuff:60,inputcheck:43,inputdata:32,inputdef:23,inputdesc:27,inputdesc_:43,inputdim:32,inputg:43,inputgate_:43,inputgrad:[32,43,46],inputindex:43,inputiov:55,inputlay:43,inputlayers_:[23,43],inputmat:46,inputoffset_:43,inputstartpos_:43,inputvalu:43,ins_:43,insert:27,insid:[46,74],inside:[38,71],inspire:8,instal:1,installat:24,instanc:[0,17,23,32,36,43,46,78,83],instartidx:43,instatu:55,instead:[7,13,21,43,46,52,55,58,60,61],instruct:[5,13,40],int32:[26,83],int32_t:[38,40,50,52,55],int64_t:[26,34,36,40,46,50,52,54,55,58,60,61],intarrai:26,intbufs_:52,integ:[0,13,19,23,40,43,86],integer_valu:[13,86],integer_value_sequ:[0,13,17,86],integr:17,inter:7,intercept:[43,74],interest:19,interfac:[7,9],interg:13,intermedi:[17,27,32,43],intern:[1,34,36,40,43,46,61,67,74,76],internal:[26,43,61],internet:19,interpol:[43,74],interpolation_lay:43,interpolationlay:42,interpret:[38,71,86],interv:[19,43,46],interval:43,intervals_:43,intializ:43,intput:30,introduc:[15,19,35,86],intstores:52,intv:43,intwithfloatarrai:26,inv:32,invari:9,invok:[15,26,36,40,43,52,54,55,61,62,74,86],invoke:[40,54],invsqrt:46,iob:[38,71],iobes:[38,71],iodesc_:43,ioe:[38,71],ioerror:26,ioff:43,iov:58,iovec:[55,58,60],iovlength:58,iparameterupdaterhook:[52,54],ipt:0,is_async:78,is_gener:[8,21],is_init:34,is_kei:15,is_layer_typ:74,is_loc:27,is_predict:[13,15,19],is_seq:[0,15],is_sequ:[15,40],is_stat:[43,69],is_test:[11,17,19],is_train:86,isaveragewindowtoolong:50,ischunkbegin:38,ischunkend:38,iscontigu:46,isdatamustincpu:40,isdrop:40,isequalto:46,isequenceresult:26,isfirstpass_:61,isfulls:52,isgpu:[26,40],isgpuvers:26,isgradclean:43,isgradshar:52,isgradsparseupd:52,isinst:[26,88],isinvalid_:43,isloc:40,isnonv:26,ispad:[32,46],isparameterspars:50,isparametersparse_:50,ispass:40,ispassfinish:55,ispassgrad:40,ispassgrad_:40,ispointeralign:46,isregularizationbatch:50,isselectalgo_:43,isspars:[23,26,46,52],issparseremoteupd:52,issparseserv:55,issparseserver_:60,isstat:52,issu:11,istestmod:36,istlargerthanalign:60,istranspos:46,istream:[40,52],isupdat:52,isvalidmatrixhandl:60,isvalidvectorhandl:60,isvalueclean:43,isvalueshar:52,isvalueupd:52,item:62,iter:[9,17,19,40,66,74,76,78],itself:[61,76],ivec:52,ivector:[26,43,46],ivectorpriv:26,ivectorptr:[40,43,46,52],jacob:43,jie:[17,19],job:6,job_dispatch_packag:7,job_mod:8,job_workspac:7,jobadding_:64,jobfinishbarrier_:64,jobfunc:[46,64],jobfunc_:64,jobqueu:55,jobs_:64,jobstartbarrier_:64,johan:19,john:43,join:64,joint:[8,21,43],jointli:[21,76],jointly:43,journal:[17,19],jpeg:9,jpg:11,json:[7,15],jth:76,judg:21,junyoung:43,just:[2,3,7,9,11,15,17,19,26,40,43,46,55,61,62,66,67,71,74,76,84,86],kaimingh:11,kasasbeh:43,kaverag:[43,61],kaveragesquarerootn:43,kcustom:40,kdecimalsc:43,kei:[15,19,60,61,86],kelasticaverag:61,kept:43,kernel:[3,13,43,74],key1:83,key2:83,keyword:86,kformatvers:52,kill:6,kind:[13,15,43,52,55,85,86],kingsburi:17,kmaxnumaccumul:50,kminmax:43,kmissparameterfail:52,kmissparameterrand:52,kmissparameterzero:52,knonseq:43,knormal:40,know:[15,23,76,86],knowledg:19,known:[19,21],kretmsginvalidmatrixhandl:60,kretmsginvalidvectorhandl:60,kretmsgunknownoper:60,kriz:9,krizhevski:43,kseq:43,ksgdsparsecputrain:40,ksimonyan:76,ksum:43,ktest:40,kunusedid_:46,kwarg:[13,15,17,71,74,76,78,86],kzscore:43,l1_rate:69,l1l2lrregular:52,l1l2regular:52,l1lrregular:52,l1regular:52,l2_rate:69,l2lrregular:52,l2regular:[9,13,19,52],label:[0,9,11,12,13,15],label_d:30,label_dict:17,label_dim:[13,74],label_fil:17,label_list:17,label_slot:17,labeledbow:19,labelseq:43,labelseqlen:43,labl:19,lag:[60,83],lake:86,lamar:43,lambda:46,lambdacost:42,lambdarank:[43,74],languag:[8,17,19,21,43,74,84],larg:[17,19,21,46,50,80],larger:[7,60,61,71,78],largest:46,last:[0,7,13,19,21,27,36,38,40,43,46,50,62,71,74,76,83],last_time_step_output:74,late:19,latenc:[7,60,61],later:[2,11,13,36],latest:[2,3,19,43,52,60,61,74],latter:21,launch:[3,6],lawyer:14,layer1:[74,76],layer2:74,layer3:74,layer:[0,8,9,11,13,15,17,19],layer_0:23,layer_attr:[0,74,84],layer_num:[11,84],layer_s:[43,74],layer_typ:74,layerbas:23,layerconfig:[23,43],layergradutil:23,layerid:43,layermap:[23,40,43],layermap_:40,layernam:[26,40],layeroutput:[15,71],layerptr:[40,43],layers_:40,layerstateptr:43,lbl:[9,71],ld_library_path:[1,7],lda:[27,30,46],ldb:[27,30,46],ldc:[27,30,46],ldd:[30,46],ldi:30,ldo:30,ldt:30,ldv:30,lead:30,leaf:43,learn:[0,9,11,13,17,19,21,23,43,50,55,69,71,74,76,78],learnabl:[43,74],learning:[43,78],learning_method:[8,9,13,15,19,21,78],learning_r:[8,9,13,15,19,21,50,69,78],learning_rate:69,learningextern:82,learningr:[46,52],learningrate_:50,learningrateschedul:50,learningratescheduler_:50,learnrat:52,least:[14,46,71,74],leat:43,leav:86,left:[11,43,46,74],leftmul:46,leman:21,len:[0,13,15,17,23,26,46,74,86],lenght:[52,58],length:[0,19,21,26,30,32,36,38,43,46,58,74,76,83],less:[7,21,43,55,74],let:[15,43,46,88],level:[7,15,19,21,34,38,43,55,58,60,69,74,83],lib64:[1,3,7,83],lib:1,libatla:1,libcuda:3,libcudnn:1,libgflag:1,libgoogl:1,libgtest:1,libjpeg:9,libnvidia:3,libprotobuf:1,libpython:1,librari:[1,7,15,34,58,67,83],library:[1,5,7,83],licens:17,liearilti:35,life:46,lightweight:3,like:[0,7,8,11,13,15,19,21,26,40,43,50,58,71,74,82,84,86],limit:[27,36,40,43,74,86],line:[2,7,8,9,11],linear:[32,35,43,74],linear_comb:74,linear_comb_lay:43,linearchaincrf:42,linearchainctc:42,linguist:17,link:[13,19],linknam:40,linux:[1,3,21],lipeng:8,lipton:19,list:[0,7,9,11,13,15,17,19,21,23,26,36,38,43,70,71,74,76,83,84,85,86],listen:[58,83],listwis:43,literatur:19,littl:[13,19,83,85,86],lium:21,liwicki:19,load:[11,15,17,19,21,27,36,43,52,55,61,74,83,85,86,88],load_data_arg:88,load_featur:11,load_feature_c:11,load_feature_pi:11,load_missing_parameter_strategi:[8,82,83,84],loaddata:36,loaddataal:36,loaddatafil:36,loader_:36,loadparamet:[26,40,88],loadparametersremot:[54,61],loadsave_parameters_in_pserv:[82,83],loadvaluerequest:60,loadvaluevector:[55,60],local:[1,2,7,9,10,18,19,20,40,43,46,52,55,61,69,82,83],localbias_:43,localbiasgrad_:43,localbuf_:46,localbufrows_:46,localgrad:46,localindic:46,localindices_:46,localoutv:46,localparam:52,localparam_:52,localupdat:61,localupdater_:61,locat:[0,13,17,23,43,46],lock:[27,34,43,44,60],lock_:[36,48,66],lock_shar:66,lockedcondit:[40,52,64],log:[1,7,9],log_barrier_abstract:83,log_barrier_lowest_nod:[82,83],log_barrier_show_log:[82,83],log_clip:[82,83],log_error_clip:[82,83],log_period:[9,13,15,17,19,21,61,83,84],log_period_serv:[82,83],logacts_:43,logger:86,logic:[7,40,86],logicaldeviceid2realdeviceid:40,logicalid:40,login:7,logisticregressionloss:46,logisticregressionlossbp:46,logprob:40,logprob_:43,longer:[21,36],look:[7,13,65,71,82,86],lookup:[13,60],loop:[40,43],loss:[13,19,21,23,43,74],lot:[60,82],low:[46,58,74],lower:7,lowest:83,lstm:[0,13,16],lstm_bias_attr:76,lstm_cell_attr:76,lstm_group:76,lstm_later:43,lstm_layer_attr:76,lstm_size:13,lstm_step:76,lstmcomput:42,lstmemori:0,lstmemory_group:74,lstmlayer:42,lstmsteplay:42,ltr:[43,74],machan:76,machin:[1,2,3,13,15,19,21,23],machine:[43,74],machineid:40,machineidvec:40,machinest:40,macro:[30,60],made:[0,14,40,86],mai:[2,4,14,27,36,40,43,46,50,64,70,71,74],main:[9,17,19,40,50,55,58,64,86,88],mainli:83,maintain:[3,46,50,55,64,74],maintainer:3,major:[2,11,19,21,27,34,46,52,55,60],majorpartners_:52,majorupd:52,make:[1,2,3,7,9,13,15,19,23,27,36,38,40,43,46,74,86],makeevalu:40,makhoul:43,male:14,malloc:[23,30],manag:[48,55,58,66],manageri:14,mani:[13,14,15,19,34,36,40,43,46,50,60,62,83],mannal:7,manner:30,manual:1,manufactur:21,mao:19,map:[3,9,11,15,26,27,32,40,43,52,54,55,60,74,83,86],marcu:19,mark:[0,17,43,46,62,67,86],mark_slot:17,markallinputgrad:43,market:[14,19],markinbackward_:43,markinputgrad:43,martha:17,mask:69,master:[2,19,52,83],masterupd:52,mat:[26,30,43,46,52],mat_cache_row:52,mat_normal:[40,52],mat_normal_shared:52,mat_param_attr:76,mat_sparse_row:[40,52],mat_sparse_row_auto_grow:[40,52],mat_sparse_row_ids:[40,52],mat_sparse_row_prefetch:[40,52],mat_sparse_row_prefetch_full_size:[40,52],mat_value_shared:52,matbtyp:46,match:46,matctyp:46,math:[23,42],matirx:[30,43,46,74],matplotlib:9,matric:[0,23,43,88],matrices_:60,matrix:[11,17,23,26,27],matrixoffset:46,matrixpriv:26,matrixptr:[23,36,38,40,43,46,52],mats_:52,matter:86,mattran:46,mattyp:52,max:[9,13,15,30,32,38,43,46,65,69,74,80,84],max_:[43,46],max_id:13,max_input:46,max_length:[0,26],max_sequence_length:43,max_sort_s:74,maxaveragewindow_:50,maxcol:46,maxframe_printer_evalu:38,maxgroups_:43,maxid:[13,38,46,71,74],maxid_lay:[13,43],maxid_printer_evalu:38,maxidlay:[40,42],maxim:[21,43,74],maximum:[0,13,17,19,30,32,46,48,52,71,83],maxindex_:43,maxlay:42,maxlength:26,maxlevel:43,maxpendingconnections_:58,maxpool:74,maxpoolbackward:46,maxpoolforward:46,maxpoolprojectionlay:43,maxrow:[30,46],maxsequencebackward:46,maxsequenceforward:46,maxsequencelength:52,maxsequencelength_:40,maxtrix:30,maxval:46,maxx_:43,mayb:[9,43,46,50,74],mdlstm:42,mdlstmlayer:42,mean:[0,7,8,9,11,13,15,17,19,21,27,38,40,43,46,52,55,60,61,66,69,71,74,76,78,83,84,86],mean_:43,mean_img_s:9,mean_meta:11,mean_meta_224:11,mean_valu:11,meangrad_:43,measur:43,mechan:[0,19,61,66,74,76],media:19,meet:17,mem:46,mem_ev:27,memalloc:46,member:[26,27,30,34,36,38,40,43,46,48,50,52,55,58,60],memcpi:30,memeori:43,memhandl:46,memor:19,memori:[0,13,17,19,21,23,26,27,30,32,34,36,40,43,46,47],memory_threshold_on_load_data:83,memoryframelin:40,memoryframelines_:40,memoryhandl:[46,48],memoryhandle_:46,memoryhandleptr:[46,48],mere:76,merg:[2,8,21,40,43,46,52,60,61,83],merge_model:40,mergecpugradi:40,mergedict:[8,21],mergegraddens:40,mergegradspars:40,mergegradsparseremot:40,mergeresultsofallcli:38,mergeseg:60,mergetypes_:40,messag:[15,19,21,36,57],messagehead:58,messagelit:60,meta:[7,9,11,13],meta_config:[7,15],meta_file:15,meta_gener:[7,15],meta_path:9,meta_to_head:15,metho:43,method:[13,15,19,21,23,26,36,40,43,46,52,54,60,62,65,66,70,74,78,83,84,86],metric:[38,55,82],middl:60,might:[23,40,60,74],million:[14,84],min:[15,30,43,46,69,84],min_:43,minaveragewindow_:50,mincol:46,mind:7,mine:43,mini:[40,52,61,74],minibatch:[43,74],minim:[78,83],minimum:[30,32,43,46,74],minimun:83,minior:52,minor:[27,52],minordeviceids_:52,minorpartners_:52,minorupd:52,minrow:46,minst:86,minut:[13,21],miss:[7,8,83],mix:[0,17,42],mixed_bias_attr:76,mixed_lay:[0,17,43],mixed_layer_attr:76,mixedlay:43,mixedlayertyp:74,mkdir:[1,3],mkl:1,ml_data:[7,15],mnist:[86,88],mnist_model:88,mnist_provid:86,mnist_train:86,mod:17,modal:17,mode:[1,7,9,11,13,15,19,21,26,27,32,34,40,43,46,52,55,66,74,83],mode_:43,model_averag:78,model_config:88,model_list:[17,19,83,84],model_output:19,model_path:[17,84],model_zoo:[8,11],modelaverag:78,modelconfig:[26,40,43],modelconfigpriv:26,modelfil:40,modifi:[0,2,7,23,40,52,54,60,88],modul:[9,11,13,15,17,36,55,61,70,77,85,86,88],modulo:[43,46,74],mom:46,momentum:[13,52,54,69,78],momentum_:50,momentumoptim:9,monitor:[13,19],mono:74,month:[13,21],mood:19,more:[0,3,7,9,13,17,19,21,23,38,40,43,46,52,55,58,60,62,71,74,76,84,85,86,88],morin:[43,74],mose:[19,21],moses_bleu:21,mosesdecod:19,most:[0,15,17,19,21,23,60,61,66,74,82,86,88],mostli:[9,14,46],move:[15,19,27,43,74],movement:19,movie_featur:15,movie_id:15,movie_meta:15,movieid:14,movielen:12,moving_average_fract:74,movingavgfraction_:43,movingmean_:43,movingvar_:43,mpi:7,mpisize_:60,mse:74,msgreader:[58,60],much:74,mul:[23,30,46],mulbybitcod:46,mulbybitcodebackwarderror:46,mulbybitcodebackwardweight:46,mulit:7,mulrowvector:46,mulscalar:46,multi:[11,21,23,40,43,46,50,52,55,60,61,64,74,82,83],multi_crop:11,multibinarylabelcrossentropi:42,multibinarylabelcrossentropybp:46,multical:55,multiclasscrossentropi:42,multiclasscrossentropywithselfnorm:42,multicombinedevalu:38,multicor:60,multimachin:40,multimachine_:40,multinetwork:40,multinomi:[43,74],multinomialsampl:42,multipl:[0,13,15,19,23,27,30,36,38,40,43,48,50,52,71,74,76,77,83,84],multiplex:43,multiplexlay:42,multipli:[9,23,43,71,74],multithread:[40,55,60,64,86],multithreadwork:36,mush:2,music:14,must:[0,2,7,21,23,30,36,38,40,43,46,50,55,64,67,71,74,76,83,84,86],mutex:[36,46,48,60,66],mutex_:[48,66],mutipl:52,mutual:7,mysteri:14,naiv:[7,46,58],name:[0,3,7,8,9,11,13,15,19,21,23,26,30,36,40,43,46,48,60,67,69,70,71,74,76,80,83,84],name_:48,namespac:[23,30,32,46,48,50,52,54],nametofuncmap_:60,nanosecond:66,natur:[17,19,84],nblocksinallparamet:60,ncelayer:42,ndcg:74,ndcg_num:74,nearest:13,neccessi:60,necessari:[1,7,13,19,23,43,55,58,60,61,71,74,86],necessarili:[23,34],need:[0,1,2,3,5,7,9,13,15,17,19,21,23,27,30,36,40,43,50,52,55,60,61,64,74,76,82,83,84,86],needfre:26,needgradi:43,needgradient_:43,needsequenceinfo_:43,needspecialtravers:[26,50],needtoupdateremot:61,needtrainwholedatainonebatch:40,neg:[13,17,19,38,46,71,86],negat:17,neighbor:13,nest:86,net:[3,76],net_conf:19,net_diagram:11,network_config:84,neuralnetwork:38,neuron:[13,19,23,43,88],never:65,newheight:46,newid:40,newmean:27,newmemori:46,newneuralnetwork:40,newnnz:46,newsiz:46,newtork:19,newwidth:46,next:[0,17,19,21,23,36,40,58,60,61,74,83],nextblock:60,nextpo:43,nic:[3,7,60,82,83],nmt:21,nnz:[23,26,30,34,43,46],nnz_:[30,36],nnzstats_:36,no_cache:86,no_sequ:15,no_sequence:86,no_sparse_id:26,no_value:46,noah:19,node:[7,19,21,23,32,40,43,55,60,61,74,83],nodefil:7,noir:14,nois:43,non:[23,26,27,30,35,36,40,55,66,85],none:[0,11,13,30,69,70,71,74,76,85,86,88],nonlinear:23,nonsmooth:43,nonstaticparaidmap_:54,nonstaticparameters_:40,nonvaluesparsematrix:46,nonzero:[30,34,40,46],norm:[27,42],norm_by_tim:74,norm_typ:43,normal:[0,5,7,8,11,23,27,32,40,43,46,55,61,64,74,76,83,86,88],normalizationstrategi:43,normalupd:52,normalupdat:61,normbytim:43,normbytimes_:43,normin_:43,normingrad_:43,normlay:42,normordropnod:40,normordropnodecallback:40,normzal:11,north:9,notat:74,note:[1,4,5,8,9,11,15,19,26,27,30,32,34,35,36,38,40,43,46,52,55,58,60,61,64,65,66,69,74,76,78,80,83,84,86,88],noth:[36,40,43,46,67,83],notic:0,notifi:[40,43,52,64,66],notify_al:66,notify_on:66,notifycopygradtobuff:40,notifygradientcollect:40,notifygradienttransf:40,notifygradreadi:52,notifytaskreadi:40,notifyvaluedispatch:40,notifyvaluereadi:[40,52],novel:19,now:[2,3,15,43,52,55,74,83,86],nproc:1,nstride:27,ntst1213:21,ntst14:21,nullptr:[23,40,43,46,50,52,61,64],nullresultnum_:64,num:[7,13,17,19,21,27,30,36,46,55,64,74,83],num_channel:[9,74,76],num_chunk_typ:71,num_class:[9,74,76],num_col:60,num_filt:[43,74,76],num_gradient_serv:[82,83],num_group:74,num_parameter_types:[52,63],num_pass:[13,15,17,19,21,82,83,84],num_result:71,num_results_per_sampl:74,numaccumulates_:50,numbatch:43,numbatch_:43,numbatches_:61,number:[7,8,9,11,13,17,19,21,23,26,27,30,32,36,38,40,43,46,52,55,58,60,61,64,65,66,71,74,83,86],number_updaters:61,numblocks_:43,numchannel:43,numchunktyp:[38,71],numclass:[43,46],numclasses_:[43,46],numcol:[46,55],numcolumn:46,numdevices_:[40,84],numdims_:43,numer:65,numfilt:43,numfilters_:43,numfram:40,numiov:58,numlogicaldevices_:[40,84],numofallsampl:71,numofthread:46,numofwrongpredict:71,numpassfinishclients_:60,numpi:[1,11,26],numport:55,numports_:55,numprocess:[40,61],numpy_mat:26,numrow:[26,43,46,50],numsampl:[30,43,46,55],numsamples_:[38,43],numsamplesprocess:[26,50],numsamplesprocessed_:60,numseqs_:40,numsequ:[30,32,40,43],numsequences_:43,numtagtyp:[38,71],numthread:[46,55,64],numthreads_:40,numupdates_:50,numvecslots_:36,numwork:64,nvidia:[1,3,83],obj:[9,11,13,15,17,36,70,86],object:[9,11,13],observ:[8,11,21,23,78],obtain:[13,17,19,36,40,43],obviousli:65,occup:15,occupat:14,occupi:55,occur:[46,50],odd:[43,46,74],off:1,offer:[17,88],offici:[3,9],offload:58,offset:[15,43,46,52,60,74],often:[13,21,61,71],ograd:[23,46],old:[36,40,46,83],oldnumaccumulates_:50,omit:13,on_init:86,onc:[13,23,40,43,64,74,86],once:2,one:[0,13,19,21,23,36,43,71,74,83,86],one_host_dens:15,one_hot_dens:15,one_parameter:30,oneachstepstart:40,oneachstepstop:40,onehotcrossentropi:46,onehotcrossentropybp:46,onehotcrossentropywithselfnorm:46,onehotcrossentropywithselfnormbp:46,ones_:43,onesequ:43,onewaysearch:40,onli:[0,1,5,8,11,13,14,19,21,23,26,27,36,38,40,43,46,52,55,60,66,67,71,74,76,82,83,84,85,86,88],online:78,onloadparamet:40,only:[14,26,30,32,43,46,55,64,83,86],onlycurrentthread:62,onpassend:[40,43],op_appli:60,op_au:60,op_au_bv:60,op_au_bv_cw:60,op_copy:60,op_cost:60,op_dir_deriv:60,op_finish_pass:60,op_fix_dir_sign:60,op_fix_omega_sign:60,op_load:60,op_make_steepest_desc_dir:60,op_random:60,op_reset:60,op_sav:60,op_sgd:60,op_start_pass:60,op_utv:60,open:[11,13,15,17,86],openblas:1,openssh:3,oper:[0,8,9,15,23,26,27,30,32,38,40,43,46,52,55,60,61,66,74,78,83],operat:[26,42],operationresult:60,operatorconfig:43,operatorfunct:60,operators_:43,opfinaloutput:32,opfunc:60,opinion:19,oppos:46,opresetgrad:32,opresetoutput:32,opstategrad:32,opt:1,opt_config:50,optconfig:[50,55],optconfig_:50,optim:[13,19,23,50,54,55,60,61,69,78,86],optimizat:60,optimizationconfig:[26,50,55,60,61],optimizationconfigpriv:26,optimizer_:50,optimizerwithgradientclip:50,optimizerwithregular:50,optimizerwithregularizereverynumbatch:50,optimizerwithregularizerspars:50,option:[1,2,7,8,23,27,40,43,74,84],optmiz:60,order:[11,13,19,21,23,30,43,46,52,64,74,76,83,86],ordinari:19,org:[50,76],organ:[9,19,21,38,43],organiz:[43,74],orgin:50,origin:[2,17,21,27,43,46,61,74,85,86],orin:46,ostream:[38,46,52,61],other:[0,1,7,8,9,11,13,14,15,17,19,21,26,29],otherchunktyp:[38,71],otherid:43,otherwis:[0,1,7,15,21,26,27,36,38,43,46,55,61,70,74,83,84,85],ouputg:43,our:[0,8,9,11,13,17,19,21,23],out:[0,9,19,26,32,38,43,46,62,74,83],out_:43,out_left:74,out_mem:0,out_right:74,outarg:[26,40,43],outargs_:40,outargsreadysem_:40,outargstream_:40,outdata:32,outdiff:32,outer:43,outerprodlay:42,outframelin:40,outframelines_:40,outgrad:[27,32,46],outgraddesc:27,outmtx_:43,outouth_:43,outperform:17,output1:46,output2:46,output:[0,6],output_:[23,35,43,74],output_d:30,output_data:27,output_dir:11,output_feature_map:27,output_grad:27,output_grad_data:27,output_id:[43,74],output_imag:27,output_image_grad:27,output_lay:11,output_mem:[0,74],output_valu:27,outputbuff:60,outputcheck:43,outputdesc:27,outputdesc_:43,outputgate_:43,outputgrad:[32,34,43,46],outputh:[32,46,74],outputh_:43,outputiov:[58,60],outputlayers_:40,outputmap_:43,outputoffset_:43,outputotherdevice_:43,outputs:43,outputs_:43,outputscor:43,outputsizeh:43,outputsizew:43,outputv:46,outputvalu:[32,34,43],outputw:[32,46,74],outputw_:43,outputx:43,outputx_:43,outputy:43,outputy_:43,outsid:[40,46,74,76,86],outv:[23,32,46],outvec:46,outvtrans_:43,over:[2,13,17,19,23,74,76,85],overcom:19,overhead:60,overlap:[23,55,61],overload:60,overrid:[23,38,64],owerfunc:64,owlqn:60,own:[7,40,46,55,60],owner:64,ownerfunc:64,ownerthreadid_:64,p_declare_bool:46,pack:[4,61],packag:3,pad:[0,13,27,32,43,74],paddepaddl:85,padding_:[43,66],padding_attr:74,padding_height:27,padding_i:74,padding_width:27,padding_x:74,paddingh:[32,46],paddingh_:43,paddingw:[32,46],paddingw_:43,paddingy_:43,paddl:[1,2,3,5,7,9,13,15,17,19,21,23,26,35,36,38,40,43,46,48,50,52,54,55,58,60,61,62],paddle:7,paddle_gen_sequ:26,paddle_nic:7,paddle_port:7,paddle_ports_num:7,paddle_ports_num_for_sparse:7,paddle_pserver2:7,paddle_root:8,paddle_source_root:8,paddle_ssh:3,paddle_ssh_machin:3,paddle_train:7,paddleapi:26,paddledev:3,paddlepaddl:[0,1],paddlepaddle:15,paddlepadl:86,paddpepaddl:86,page:[2,15,27],pair:[17,43,46,55,60,61,71],palmer:17,paper:[8,11,17,19,21,27,43,50,74],para:[40,43,54,60,61],para_macro:30,paraconfig:[50,52],paraconvert:8,paraemet:61,paragraph:19,parallel:[21,32,34,40,43,52,55,58,60,83,84],parallel_nn:[43,82,83],parallelcpuvectort:46,paralleldatarequest:55,parallelexec:46,parallelexecforeachblock:60,parallelinputiov:55,parallelli:[43,50],parallelparamet:52,parallelparametermap:52,parallelparameterptr:52,parallelrequest:55,parallelthread:40,param:[15,26,52,69,74],param_attr:[0,74,76],param_config:54,paramainthread:40,paramainthread_:40,paramattr:[0,69,74],paramconfig:54,paramemt:52,parameter_:[43,52,60],parameter_apply:[50,63],parameter_block_s:[82,83],parameter_block_size_for_spars:[82,83],parameter_cols:63,parameter_delta:63,parameter_gradient:[40,50,52,63],parameter_gradient_squresum1:63,parameter_gradient_squresum:63,parameter_gradient_sum:63,parameter_learning_rate:[63,69],parameter_momentum:[40,63],parameter_momentum_ut:63,parameter_momentum_vt:63,parameter_rows:63,parameter_second_momentum:63,parameter_sum1:63,parameter_sum2:63,parameter_sum3:63,parameter_update_time:63,parameter_value:[40,50,52,55,63],parameter_weighted_infinity_norm:63,parameterattribut:[69,74,76],parameterblock:60,parametercli:61,parameterclient2:[38,55,61],parameterclient_:61,parameterconfig:[26,50,52,54,60],parameterconfigpriv:26,parameterids_:40,parametermap:[23,40,43,52],parametermap_:[40,55],parametermutex_:60,parameteroptim:[26,50,60],parameteroptimizerpriv:26,parameterpriv:26,parameterptr:[40,43,52,54,55,61],parameterreadybarrier_:60,parameterrelulay:42,parameters_:[23,40,43,54,61],parameterseg:55,parameterserv:60,parameterserver2:60,parameterservic:60,parametertraversecallback:26,parametertraversecallbackpriv:26,parametertyp:[26,40,43,50,52,54,55,60,61,63],parametertypes_:[50,54],parameterupdat:[26,52,54,61],parameterupdated_:40,parameterupdatemod:55,parameterupdatercomposit:[54,61],parameterupdaterhook:54,parametris:78,paramid:[40,52],paraminitcallback:[40,52],paramrelubackwarddiff:46,paramrelubackwardw:46,paramreluforward:46,paramselfinited_:40,paramt:8,paramutil:15,paraphrase_data:8,paraphrase_model:8,paraspars:23,parent:[23,43,74],parmet:61,pars:[15,17,84,88],parse_config:88,parser:15,part:[0,13,15,17,19,21,23,43,46,58,60,61,86],parti:15,partial:74,partial_sum:30,partialsum_:43,particip:52,participl:8,particular:[40,43],partit:52,partner:[40,52],partnerid_:40,partnerparam_:52,pass:[2,7,9,13,15,17,19,21,23,26,27,36,40,43,46,50,55,58,60,61,65,70,74,83],pass_:50,pass_gc:63,pass_metric:63,pass_metric_train:63,pass_metric_train_with_noerror:63,pass_test:[40,43,52,63],pass_train:[43,63],passbarrier_:60,passcount_:61,passfinish_:55,passtyp:[23,26,40,43,52,54,61,62,63],passtype_:[40,43],password:[3,7],path:[0,1,5,7,8,9,11,13,17,19,21,26,40,74,83,84,85,86],pathsequenceresult:26,paticip:52,pattern:[15,19,30,40,61],paul:17,pave:21,pdf:76,peephol:43,peer:27,peerdevic:27,peernam:58,peername_:58,pend:40,penn:17,per:[9,13,50,60,74,83],perceiv:60,perfer:60,perfom:[83,84],perform:[0,2],period:[13,15,17,19,21,34,83,85],perl:[19,21],permitrootlogin:3,peroid:[43,74,83],perturb:[23,34],perver:60,phase:27,photo:9,pick:46,pickl:15,picklabl:70,pictur:13,pid:40,pid_t:64,pidqueu:40,piec:[74,77],pillow:9,pin:48,pionter:30,pip:[1,7,9,15],pipe:[14,61],pipelin:[17,61],pixel:[43,74,86],pixels_float:86,pixels_str:86,place:[7,11,21,23,85,86],plai:[17,19],plain:[38,71,74,85],plan:23,platform:[7,13],pleas:[0,1,3,8,9,13,15,17,23,30,36,43,50,69,74,76,78,85,86,88],plot:9,plotcurv:9,plu:3,png:[9,11],pnpairvalid:42,pnpairvalidationlay:83,pnpairvalidationpredict_fil:82,point:[30,43,46,61,66],pointer:[27,30,40,43,46,48,60,64],polar:19,polit:19,poll:[19,52],pollut:60,poo:9,pool3:23,pool:[9,13,15,27,32,42],pool_:[46,48],pool_attr:76,pool_bias_attr:76,pool_layer_attr:76,pool_pad:76,pool_siz:[74,76,86],pool_start:76,pool_strid:76,pool_typ:[43,74,76],poolalloc:48,pooledh:32,pooledw:32,pooling_desc:27,pooling_lay:[13,15],pooling_typ:[13,74],poolingdesc_:43,poolingtyp:80,poolint:43,poollay:42,poolmemorysize_:48,poolprojectionlay:42,pooltyp:43,pooltype_:43,pop:62,popd:1,popular:11,port:[3,7,55,58,60,82,83],port_:[55,58],port_num:82,ports_num:[58,83],ports_num_for_spars:[82,83,84],posit:[13,17,19,21,30,32,36,38,43,71,74,86],positive_label:71,possibil:40,post:66,postion:[30,46],postvaluereadi:52,pow:46,pow_:43,power:[13,21,43,74],power_lay:43,powerlay:42,practic:[0,23,70,74],pre:[3,4,6],pre_batch_st:43,pre_dictandmodel:8,precis:[1,38,71],precision_recall_evalu:38,pred:[13,17,66],predefin:19,predetermin:[21,74,83],predic:[17,50,66],predicate_slot:17,predict:[0,1,7,8],predict_fil:83,predict_output_dir:[13,82,83],predict_sampl:88,predictarrai:38,predictarray_:[38,43],predicted_label_id:13,predictionresult:[38,43],predictor:15,predin:9,prefer:19,prefetch:[23,40,43,46,52,60,61],prefix:[36,40],pregrad:23,premodel:8,preoutput_:43,preoutputgrad:32,preoutputvalu:32,preoutv:46,preparebatchdata:26,preparedata:55,preparedoper:55,preparesampl:43,preparesenddata:55,prepo:43,preprcess:19,preprint:43,preprocess:[0,7],prerequisit:1,present:[11,17,21],preserv:46,prev:32,prev_batch_st:[82,83],prev_out:43,prevbatchoutput2_:43,prevbatchst:43,prevent:[60,78,85],prevgrad1:46,prevgrad2:46,prevgradx:30,prevgrady:30,previou:[17,21,23,32,34,43,74,76,83],previous:11,prevnumupdates_:50,prevout1:46,prevout2:46,prevout:32,prevoutgrad:[32,34],prevoutput2batch:43,prevoutput_:43,prevoutputvalu:32,prevoutvalu:[32,34],prevoutx:30,prevouty:30,prevstat:32,prevstate_:43,prevstategrad:[32,34],prevstatevalu:34,prevtag:38,prevtyp:38,primarili:19,print:[8,13,15,17,19,21,38,46,61,69],printall:48,printasyncgradientcommitstatandreset:60,printdiff:46,printoneel:46,printonerow:46,printpredictresult:38,printstat:38,prite:74,privat:[26,30,46,48,50,52,55,58,60],prob:[38,43,71],probabilist:[8,43,74],probabilit:40,probability_of_label_0:13,probability_of_label_1:13,probabl:[0,2,11,13,17,38,40,43,71,74],probal:40,probhistori:40,problem:[13,17,19,38,74,78,88],proce:66,procedur:[8,17,21],proceed:[17,43],process:[0,7,8,9,11,13,15,17,19,21,36,38,40,43,52,58,60,61,62,64,66,69,70,74,76,78,83,84,85,86,88],process_pr:13,process_test:70,process_train:70,processchannel:64,processdata:[9,11],processjob:65,produc:[11,13,76],product:[13,19,23,43,55],profil:40,program:[7,83,85],programm:14,progress:83,proivid:86,proj:74,projcol_:43,project:[0,15,23,32,42],projectionconfig:43,projections_:43,projectionstatematrixsize_:43,projoutput_:43,promis:[74,76],prop:17,propag:[30,35,40,43,78,83,84],propagaion:35,properli:13,properti:[27,83,86],propos:21,proposit:17,protbuf:36,protect:[23,36,38,40,43,46,48,50,52,54,55,58,60,61,64,66],proto:26,protobuf:[1,26,36,55,60],protocli:[55,58],protocol:83,protoin:[55,60],protoout:[55,60],protoread:36,protoresponsecallback:60,protoresponsecallbackex:60,protoserv:[58,60],protoslot:36,protostr:26,protovarslot:36,prove:13,proven:21,provid:[0,3,5,7,8,9,11],provider_:36,providermemory_threshold_on_load_data:82,providerptrtyp:36,providertyp:36,provied:35,provod:86,prune:74,pserver:[7,40],pserver_num_thread:[82,83],pservermatrix:55,pserverstart_pserv:82,pserverstatu:[55,60],pservervector:55,psize:23,pthread_barrier_t:66,pthread_mutex_t:[27,34],pthread_rwlock_t:66,pthread_spinlock_t:66,ptr:[26,30,36,40,46,48,50],ptype:52,punctuat:19,purchas:13,purpos:35,push_back:23,pushd:1,put:[13,23,64],putdata:55,putowndata:55,py_paddl:[26,88],pydataprovid:[13,19],pydataprovider2:[0,13,15,36,85],pydataproviderwrapp:[19,70],pyguard:36,pyint:36,pylong:36,pyobject:36,pyobjectptr:36,python:[1,8,9],pythonpath:9,pzo:19,qualiti:[13,60],quaternari:[30,46],queri:[21,27,38,74],querybeamsearch:40,queryid:38,question:17,queue:[40,44,55,64],queuecapac:64,quick_start:13,quit:64,quoc:43,quot:14,rabih:43,ragno:43,raii:66,ram:61,ramnath:19,rand:[43,46,83,84],rand_:43,randnorm:46,random:[9,27,34,36,43,46,52,69,74,83],random_int_numb:43,randomizeuniform:46,randomli:19,randomnumberse:82,randparamet:[26,40],randparametersremot:[54,61],rang:[9,14,17,26,30,43,46,52,64,83,84,86],rangeerror:26,rangereciprocal_:43,rank:[11,13],rankaucevalu:38,rankingcost:42,rankloss:46,ranklossbp:46,rare:86,rate:[7,9,13],rather:[19,88],ratio:83,raw:[13,19,48,74],raw_meta:15,rbegin:[43,62],rdma:[3,58,60,83],rdma_tcp:[82,83],rdmaclient:58,rdmacpu:[58,60],rdmacpu_:58,rdmaserv:58,rdmasocket_:58,rdmauri_:58,reach:17,read:[0,7,11,13,15,36,55,58,60,61,66,85,86],readallblock:60,readblock:58,reader:[21,57],readi:[9,27,40,43,46,52,55,60],readme:[14,15,19],readmessag:58,readnextblock:58,readv:58,readwritebuff:60,readwritebuffer_:60,real:[27,30,32,34,36,38,40,43,46,50,52,54,55,60,61,86],realdeviceid2logicaldeviceid:40,realid:40,realiti:60,reallay:[40,43],reallayer_:43,reallayernam:40,reallayers_:43,realoutarg_:43,reason:[50,60,74],recal:[38,71],receiv:[38,55,60,70],recent:[1,21,61],reciev:[55,83],reciproc:46,reciprocalderiv:46,reciprocalrowsum_:43,reciprocalsum:46,recogn:9,recognit:[11,19,86],recommand:[76,86],recommend:[0,3,7],recommendation:7,recommonmark:1,recor:50,record:[15,17,27,40,43,60],recordhistori:40,rectangular:74,recurrent_group:0,recurrentgradientmachin:40,recurrentgroup:[38,71],recurrentlay:42,recurrentlayergroup:43,recurrfgw:43,recurrigw:43,recurriw:43,recurrogw:43,recv:[55,61],recvbuf:55,recvdata:55,recvdatamems_:55,recvjobqueue_:55,recvparamet:55,recvparametertyp:55,recvsyncbarrier_:55,recvthreads_:55,reduc:[7,43,55,60,78,83,84],reduceandsenddata:60,redund:60,reeal:43,refactor:43,refer:[0,3,7,8,9,13,15,16],referenc:74,refmat:46,regard:[17,55],regardless:21,regex:15,reggradientmachinemod:40,region:[17,43],regist:[23,40,43,60],register_layer:23,register_service_function:60,register_service_function_ex:60,registerbeamsearchcontrolcallback:40,registerbeamsearchstatisticscallback:40,registerservicefunct:60,registerservicefunctionex:60,registerservicefunctionimp:60,registrar_:[36,38,43],regress:12,regression_cost:15,reguardless:64,regular:[9,13,19,23,50,51],regularizer_:50,rel:[7,55,76,85],relai:61,relat:[1,5,7,15,19,31],relationship:7,releas:[3,4,5,14,17,46,48,55],releasematrix:[55,60],releasematrixrequest:60,releasepass:55,releasevector:[55,60],releasevectorrequest:60,relev:[17,21],relu:[23,30,32,35,46,67,74],reluderiv:46,remain:[13,58,61,64],remot:2,remoteparameterupdat:54,remov:[7,19,40,83],removebeamsearchcontrolcallback:40,removebeamsearchstatisticscallback:40,renam:21,rend:62,renshaw:43,reorder:43,reorgan:[43,74],replac:[3,19],replic:43,repo:2,repons:[52,60],report:7,repositori:2,repres:[0,9,13,14,23,43,46,78,86,88],represent:[13,19,34,43],reproduc:21,requestvec_:60,res5_3_branch2c_bn:11,res5_3_branch2c_conv:11,research:[9,14,19],resembl:19,reserv:[46,86],reserveoutput:[23,43],reservestor:46,reset:[27,32,36,43,46,60,61,74],resetalignalloc:60,resetconvoutput:43,resetcurrentstat:61,resetexpandinput:43,resetg:34,reseton:46,resetoutput:[34,43],resetoutput_:43,resetoutputgrad:[32,34],resetoutputvalu:[32,34],resetspecifyoutput:43,resetst:[40,43],reshap:[27,42],reshapedoutputgrad:43,resid:27,residu:11,resiz:[26,43,46,60],resizeandcopyfrom:52,resizebootfram:40,resizecsc:46,resizecsr:46,resizelay:42,resizeorcr:[43,46],resizeorcreatebatch:43,resizeorcreatefram:40,resizeorcreatesparsematrix:46,resizeoutput:43,resizewithalignhint:60,resnet_101:11,resnet_152:11,resnet_50:11,resolut:34,respect:[0,9,11,17,21,23,43,46,83,86],respons:[40,43,46,52,55,58,60,74],responsecallback:[58,60],responsenormlay:42,respos:32,rest:[36,74,86],restart:40,restor:[46,50,54,61],restrict:[40,83],result:[0,6],result_fil:[0,71,74],result_of:64,resultptrtyp:64,resultrunningmean:27,resultrunningvari:27,results_:64,resulttyp:64,retir:14,retriev:[23,40,43,55,60],return_seq:76,reus:[23,43],revers:[0,17,19,32,43,46,74,76],reversed_:[40,43],review:[13,19],rewrit:21,rgb:74,rgen:19,rho:78,richard:43,right:[11,38,43,46,74,86],rightmul:46,ring:40,rmsprop:[13,78],rmspropoptim:15,rmspropparameteroptim:50,rnn:[0,13,19,24],rnn_bias_attr:0,rnn_layer_attr:0,rnn_out:0,rnn_step:74,rnn_use_batch:[43,82,83],robert:43,robot:9,robust:[7,43],roc:38,roce:3,role:[0,12],role_:52,roman:19,romanc:14,root:[1,3,7,32,43,78,80],root_dir:7,rootagent:40,rootid:55,rootlay:40,rootnetwork:[40,43],rootnetwork_:40,rou_:50,rout:55,routin:[15,34,55,60,61],routledg:19,row:[11,23,26,27,30,34,38,40,43,46,50,52,55,60,71,74,88],row_:30,row_id:[43,74],row_siz:30,rowadd:46,rowbuf:46,rowdotmul2:46,rowdotmul:46,rowin:46,rowindex:46,rowmax:46,rowmaxid:46,rownormalizel1:46,rowpow:46,rows_:46,rows_d:30,rows_in_block:50,rowscal:46,rowscale2:46,rowstore_:46,rowsum:46,rowvec:50,rpc:60,rtype:15,rule:[23,55],run:1,runninginvvar:27,runningmean:27,runtim:[7,27,85,86],rvalu:65,rwlock:[36,48,60],rwlock_:66,s_fusion:15,s_id:15,s_recurrent_group:0,sacrif:85,safe:[50,65],sai:[83,84],sake:[23,46],sale:14,same:[0,1,3,7,8,13,15,17,19,21,30,38,40,43,46,50,52,54,55,61,64,66,70,71,74,76,84,86,88],samping_id:74,sampl:[7,8,11,13,15,17,19,21,36,38,40,42],sample1:36,sample2:36,sample_num:71,sampleloop:36,samplen:36,samplenums_:36,sampler:43,sampling_id_lay:43,samplingidlay:42,santiago:19,sate:76,satisfi:[7,13,55],save:[9,11,13,15,17,19,21,27,43,52,55,61,71,74,83,84,86],save_dir:[9,13,15,17,19,21,83,84],save_mean:[27,43],save_only_on:[82,83],savedinvvar:27,savedinvvar_:43,savedir:55,savedmean:27,savedmean_:43,savedvar:27,saveparamet:40,saveparametersremot:[54,61],saver:[30,46],savevaluerequest:60,savevaluevector:[55,60],saving_period:[82,83],saving_period_by_batch:[13,82,83,84],saw:86,sbin:3,scalabl:60,scalar:[27,30,43,74,86],scalardiv:46,scale:[11,14,15,27,30,32,43,46,52,55,67,74],scale_:43,scalea:32,scaleab:46,scaleb:32,scaledtanh:46,scaledtanhderiv:46,scalegrad:27,scaleoutput:46,scalesum:46,scalet:46,scaletarget:46,scaling_lay:43,scalinglay:42,scanner:36,scatter:[40,43,74],scatterag:40,scatteragentlay:42,scenario:82,scene:82,schedul:7,scheduler_factor:69,schema:8,scheme:[17,38,71],schmidhub:19,schwartz:43,schwenk:21,sci:14,scienc:19,scientist:14,score:[15,19,21,26,38,43,71,74],screen:15,scrip:13,script:[7,9,11,13,17,19,21,88],seaplane_s_000978:9,search:[0,17,21,40,43,74,83],seat:21,sec:52,second:[2,7,8,11,13,14,15,19,30,34,40,43,46,55,61,62,66,74,86],secondli:[7,55],section:[0,7,13,23,86],sed:[3,19,60],see:[2,8,11,13,15,19,21,27,30,43,52,74,76,86,88],seed:[27,83],segment:[38,52,55,60,71],segmentor:8,segmentrang:43,sel_fc:74,select:[2,14,21,30,36,43,46,74],selectel:46,selectfrom:46,selectiv:74,selective_fc_lay:43,selectivefullyconnect:43,selectivefullyconnectedlay:42,selectrow:[46,52],selectrowsimp:46,selectrowsonetim:40,self:[14,19,23,26,43,52,58],selfnorm:74,sem_:66,sem_t:66,semant:[0,12],semaphor:[40,52],sen_len:17,send:[40,55,60,61,83],sendandreceiveparamet:55,sendbackparamet:[55,60],sendbackparameterspars:60,sendbackparametertyp:55,sendbuf:55,senddata:[55,60],senddatarequest:[55,60],senddatarequestvec:55,senddatatyp:55,sendjob:55,sendjob_:55,sendjobptr:55,sendjobqueue_:55,sendparallel:55,sendparamet:[55,60],sendparameterrequest:[55,60],sendparameterrespons:60,sendqueu:55,sendrequest:55,sendthreads_:55,sens:[38,55,74],sent:[55,60,61],sent_id:0,sentenc:[0,13,17,19,21,26,86],sentiment:[12,13,17],sentiment_data:19,sentiment_net:19,sentimental_provid:86,separ:[8,13,14,15,17,21,30,40,55,60,61,64,71,83,86],separatesendandrecv_:[55,61],seq2batch:[32,43],seq2batchidx:43,seq2batchidx_:43,seq:[15,46],seq_pool:74,seq_text_print:74,seq_to_seq_data:[8,21],seq_typ:[15,86],seqendidxinbatch_:43,seqflag:52,seqid:[40,43],seqidx_:43,seqlength:43,seqlengthandstart:52,seqlengthandstart_:40,seqsiz:52,seqstart:[43,52],seqstartpo:52,seqstartposindex:[40,43],seqstartposindex_:43,seqtext_printer_evalu:38,seqtoseq:[0,8,21,74],seqtoseq_net:[0,8,21,74],sequel:86,sequence2batchadd:43,sequence2batchcopi:43,sequence:[36,86],sequence_conv_pool:13,sequence_layer_group:74,sequence_max:35,sequence_nest_layer_group:74,sequenceagentlay:42,sequenceavgforward:46,sequenceconcatlay:42,sequencegatheragentlay:42,sequencegener:26,sequencegeneratorpriv:26,sequencelastinstancelay:42,sequencelength:32,sequencelevel:43,sequenceloop:36,sequencereshapelay:42,sequencescatteragentlay:42,sequencesoftmax:46,sequencestartposit:[40,43,52,74],sequencestartpositions_:36,sequencetobatch:42,sequencetyp:86,sequenti:[0,13,17,43,64,74,85],sequentianlli:64,seqvalu:43,seqwidth:32,seri:19,serial:86,serious:3,server:[3,7,11,12,23,40,44,46,55,57],server_:58,serveraddr:58,serverid_:60,serverport:58,servic:[14,55,60],servicefunct:60,servicenum:55,servicenum_:55,set:[0,1,3,5,7,8,9,11,13,14,15,17,19,21,23,26,27,35,36,38,40,43,46,50,55,58,60,61,62,69,71,74,76],set_active_typ:23,set_default_parameter_nam:69,set_drop_r:23,set_siz:23,set_typ:23,setbatchs:26,setbeams:26,setbo:26,setconfig:[55,60],setconfigrequest:60,setdata:[36,43,46],setdevic:52,setdict:26,setelement:46,seteo:26,setforwardbackwardtim:[54,55,61],setframeheight:52,setframewidth:52,setid:52,setintbuf:52,setitem:26,setmat:52,setmaxlength:26,setmeanandstd:43,setneedgradi:43,setneedsequenceinfo:43,setnodecai:50,setnumofthread:46,setoutput:43,setoutputgrad:40,setparamet:[55,60],setparameterptr:52,setparameterzero:55,setpassgrad:40,setreallay:43,setreallayerandoutput:43,setrow:46,setsequencestartposit:43,setsharedcount:52,setskipshuffl:36,setslotid:26,setslotin:26,setslotsequencedim:26,setslotsequencestartposit:26,setslotsubsequencestartposit:26,setslotvalu:26,setsourcedata:46,setstat:[40,43],setstatu:[55,60],setstatusrequest:60,setsync:46,settrainerid:55,settup:23,setup:[13,23,46],setupindic:46,setvalueupd:52,seven:17,sever:[7,13,15,17,19,21,40,43,55,60,74,84,86],sftmaxsum:46,sftmaxsum_:43,sftmaxsum_d:30,sgd:[7,19,21,52,55,60,61,78],sgdasync_count:82,sgdlocalupdat:54,sgdoptimiz:50,sgdthreadupdat:54,sgdupdate:[46,52],shaaban:43,shadow:61,shake:43,shalabi:43,shall:66,shallow:[17,46],shape:[11,43,74],share:[1,17,40,43,46,52,54,55,58,60,66,74,83],sharecpumatrix:43,shared_bia:76,shared_bias:[43,74],shared_ptr:[36,43,46,48,52,54,55,64],sharedbiases_:43,sharedcount_:52,sharedcpumatrix:46,sharedptr:26,shareindexwith:43,shell:11,shift:[11,43],shift_start:43,shiftdim:43,ship:9,shold:[2,19],shop:19,shorter:11,should:[0,7,9,13,15,17,19,21,26,36,38,40,43,46,50,52,55,64,66,71,74,78,86,88],should_shuffl:86,shouldbem:40,shouldn:2,show:[2,8,11,13,15,17,19,21,43,61,78,83,88],show_check_sparse_distribution_log:[82,83],show_layer_stat:[82,83],show_parameter_stats_period:[13,17,19,21,82,83,84],showdatastat:36,shown:[0,9,11,13,15,19,21,23,38,71,74,86],showoutputstat:43,showstat:61,shrink:[23,32,43,65],shrinkmat:43,shuf:15,shuffl:[15,19,36,86],shuffledsequenceids_:36,side:[11,26,55],sigint:7,sigmoid:[23,32,35,43,46,67,74],sigmoid_threshold_max:32,sigmoid_threshold_min:32,sigmoidderiv:46,sign:46,signal:[7,35,55,61],signatur:60,sim:30,simd:46,similar:[13,15,34,43,46,74],similarli:[17,74],simple_attent:0,simple_gru:13,simple_lstm:[13,74],simple_rnn:[0,74],simpledataproviderbas:36,simpli:[0,2,8,11,15,19,21,43,74,85],simplifi:[23,60],simultan:[46,64],simutan:60,sinc:[3,11,13,14,21,60,66,74],sincer:[2,19],singl:[1,11,13,17,21,23,36,38,40,46,52,55,58,60,61,66,71,76,78,86],singleton:[48,64],singleupd:52,situat:40,six:[3,8,17,21],size:[0,7,9,11,13,14,15,17,19,21,23,26,27,30,32,34,36,40,43,46,48,52,55,58,60,61,64,65,71,74,76,78,83,86],size_:[36,43,46,48,60],size_a:74,size_b:74,size_t:[23,26,27,30,32,34,36,38,40,43,46,48,50,52,54,55,58,60,64,65],sizei:30,sizeinbyt:27,sizelimit:48,sizelimit_:48,sizeof:[8,30,43,46,60,66],sizex:[32,46],sizex_:43,sizey_:43,skill:21,skip:[7,11,36],skipshuffle_:36,slave:[40,52],slaveparam:52,slaveparamet:40,slaveupd:52,sleep:55,slide:78,slightli:9,slope:[43,74],slope_intercept_lay:43,slopeinterceptlay:42,slot:[0,15,17,26,36],slot_:36,slot_dim:15,slot_nam:15,slotdef:36,slothead:36,slotnum:26,slots_:36,slottyp:[15,36],slow:60,small:[7,9,21,23,40,43,60,83,86],small_messag:[82,83],small_vgg:9,smaller:[40,43,74],smallest:[43,55],smat:30,smatrix:46,smatrix_:46,smemoryhandl:46,smemoryhandle_:46,smith:19,smooth:4,snippet:[0,13,23],social:19,sock:58,sock_recv_buf_s:[82,83],sock_send_buf_s:[82,83],socket:57,socket_:58,socketchannel:58,socketcli:58,socketdaemon_:58,socketserv:[58,60,64],socketwork:[58,64],socr:74,soft:43,softbinaryclasscrossentropi:42,softcrossentropi:46,softcrossentropybp:46,softmax:[0,8,13,17,19,23,27,30,35,43,46,67,74,76],softmax_param_attr:76,softmax_selfnorm_alpha:74,softmaxactiv:[0,13],softmaxbackward:46,softmaxderiv:46,softmaxseq:43,softmaxseqgrad:43,softmaxseqlen:43,softrelu:[35,46,67],softreluderiv:46,solut:55,solv:17,solver:21,some:[1,3,7,13,14,15,17,19,21,23,32,38,40,43,46,50,52,55,58,60,69,74,78,82,83,84,86],someth:[64,66,74,86],sometim:[19,78],sophist:23,sort:[15,19,21,26,43,46,74,83],sourc:[0,1,3],source_dict_dim:0,source_language_word:0,sourcedata_:46,sourcedatavec_:46,sourcevec:46,space:[0,36,43,48,71],space_seperated_tokens_from_dictionary_according_to_seq:[38,71],space_seperated_tokens_from_dictionary_according_to_sub_seq:[38,71],spars:[7,13,23,26,29],sparse:36,sparse_binary_vector:[13,86],sparse_csc:[26,46],sparse_csr:[26,46],sparse_float_value_t:[36,46],sparse_float_vector:86,sparse_format:46,sparse_non_valu:36,sparse_non_value:26,sparse_non_value_t:[36,46],sparse_nonvalu:26,sparse_upd:69,sparse_valu:36,sparse_value:26,sparseautogrowrowcpumatrix:[40,46],sparsecopyfrom:26,sparsedistribution_:55,sparsefloatvaluedata:36,sparseformat:[46,52],sparseformattyp:26,sparseid:[26,50],sparsematrix:[30,46],sparsemomentumparameteroptim:50,sparsenonvaluedata:36,sparseparam:23,sparseparameterdistribut:55,sparseprefetchrowcpumatrix:[23,40,46,52],sparseremoteparameterupdat:54,sparseremoteparameterupdatercomposit:54,sparseres:46,sparseresizecsc:46,sparseresizecsr:46,sparserowcpumatrix:[40,46,52],sparserowcpumatrixto:40,sparserowidscpumatrix:[40,46],sparserowmatrix:46,sparsevaluetyp:[26,46],sparsiti:[30,40],spatial:[9,43],spawn:58,spe:38,speak:[0,21],specfii:[43,83],speci:9,special:[8,13,21,36,60,74],specif:[9,13,15,27,46,55,60,84,85],specifi:0,speed:[43,50,76],spefici:11,sphinx:1,sphinx_rtd_them:1,spin:66,split:[7,8,11,13],splitbydataid:52,splite:26,sprare:30,sql:85,sqr:46,sqrt:[32,46],squar:[32,35,46,67,74,78,80],squareddiff:30,squarederiv:46,squarerootnpool:74,squash:21,srand:83,src:[1,17,21,26,27,30,32,43,46,52,55],src_backward:0,src_d:27,src_dict:0,src_embed:0,src_forward:0,src_h:27,src_id:0,src_root:88,src_word_id:0,srcid:40,srcnn:40,srl:17,sse:[30,60],sseadd2:30,sseadd:30,sseclassificationerror:30,ssediv:30,ssefirst:30,sseidentiti:30,ssemax:30,ssemin:30,ssemul:30,ssesecond:30,ssesquareddiff:30,ssesub:30,ssesum:30,ssh:[3,7],sshd:3,sshd_config:3,stack:[13,17,18],stacked_lstm_net:19,stacked_num:19,stage:[7,26,60,61,62],stake:21,stale:60,standard:[8,9],standarddevi:46,stanh:35,star:14,start:[0,2,7,8],start_:[43,46],start_pass:[82,83],start_pserv:83,startbatch:[26,50,54,60,61],startcatchupwith:50,startcol:46,startcontrol:61,startfil:36,startidx:43,startpass:[26,50,54,61],startprepar:36,startrow:46,startseq:52,startspo:46,starttask:40,startthread:55,starttrain:26,starttrainpass:26,stat:[17,19,21,38,60,61,83],state2_:43,state:[0,4,17,21,32,34,40,43,50,60,61,74,76,83],state_:43,state_act:[74,76],stateactive_:43,stateactivegrad:34,stateactivevalu:34,stateatv:32,stategrad:[32,34],statement:23,statevalu:[32,34],stateweight:34,stateweight_:43,stateweightgrad:34,staticcastvector:26,staticinput:[0,74],staticpruninghook:54,statist:[13,17,19,21,38,40,43,61,74,83],statptr:36,statset:60,statset_:60,statsinfo:38,statu:[2,40,55,60,62,71],status_:60,std:[23,26,35,36,38,40,43,46,48,50,52,54,55,58,60,61,62,64,66,83],stderr:7,stdgrad_:43,stdout:7,stdreciprocal_:43,step:[0,7,13,15,17,19,21,23,30,36,40,43,74,76,80,88],step_:50,still:[4,11,43],stochast:78,stock:19,stop:[3,7,15,36,40,64,74,83],stopaddjob:64,stopbeamsearch:40,stopbeamsearch_:40,stopping_:[40,55,58,64],storag:[9,46,48],storageengin:48,store:[7,8,9,11,13,15,17,19,21,23,27,36,40,43,46,52,55,58,60,61,64,74,83,86],storemat_:46,str:[26,43,52,84],straight:2,strat:27,strategi:[36,52,80,83,86],strdata:36,stream:[27,30,34,36,40,46,52,58,61],street:17,strictli:60,stride:[27,32,43,46,74],stride_:[43,46],stride_height:27,stride_i:74,stride_width:27,stride_x:74,strideh:[32,46],strideh_:43,strideheight:43,stridew:[32,46],stridew_:43,stridewidth:43,stridey_:43,string:[19,23,26,27,35,36,40,43,48,52,54,55,58,60,61,62,70,74,83,85,86],strip:[13,15,17],struct:[26,27,30,34,36,38,40,43,46,52,55,58,60,66],structur:[7,8,9,13,15,17,19,21,34,40,43,52,60],student:14,stun:86,style:[1,2,40,58,74],sub:[0,9,13,21,23,30,38,40,43,46,71,74],sub_sequence:86,sub_work_thread:40,subargfrom:52,subbybitcod:46,subclass:[26,35,36,38,40,43,46,48,50,52,54,55,58,60,61,64],subcolmatrix:46,subdataproviders_:36,subgradi:78,subindic:36,subk_:43,subm_:43,submatrix:46,submit:[2,82,83],submodel:[40,43],submodelnam:40,submodelname_:40,subn_:43,subnetwork:43,subnetwrk:43,subrowmatrix:46,subsampleloop:36,subscalar:46,subsequ:43,subsequenceinput:74,subsequencelay:42,subsequencestartposit:[43,52],subset:[21,23,40,42],substanti:11,substitut:21,subvec:46,subvecfrom:46,succe:19,success:[11,17,38],successfuli:19,successfulli:[11,15,19,46,66],successor:[21,83],sucessfulli:21,sudo:[1,9],suffici:83,suffix:21,suggest:[27,74],suit:43,suitabl:[9,11,55,83],sum:[0,23,30,32,38,43,46,52,55,71,74,78,80],sum_evalu:38,sum_j:46,sum_to_one_norm:74,sum_to_one_norm_lay:43,sumbybitcod:46,sumcol:46,sumcost:52,suminv_:43,summar:[13,19],sumofsquar:46,sumofsquaresbp:46,sumofsquarescostlay:42,sumpool:74,sumrow:46,sumtoonenormlay:42,supervis:36,suppli:[36,40],support:[0,1,3,5,11,23,26,30,32,34,36,38,40,43,46,52,55,61,67,69,71,74,78,83],support_hppl:67,suppos:[13,23,35],sure:[2,7,9,19,23,36,46],survei:19,suspend:40,svectorptr:52,swap_channel:11,swig:[1,26,88],swig_paddl:88,sxi_sock:58,sxi_socket:58,symbol:74,sync:[2,46,55,60,61,64,83],sync_:46,sync_default:55,synced:46,syncedflag:46,syncflag:[23,46],syncflag_:46,synchron:[7,27,30,40,52,55,60,61,78,83],synchronizebarriers_:60,synchronizeparamt:52,synchronizerequest:60,syncobject:55,syncobjectid:55,syncparamet:52,syncthreadpool:[46,54,55,60,61],syncthreadpool_:[54,55,60],synctrainersteps_:60,syncupdat:52,syncupdaterstep:60,syntact:17,syntax:15,system:[1,13,17,19,21],szegedi:43,t0vec_:50,t2b:8,t_resourc:34,tab:13,tabl:[11,13,21,30,43,46,60,74,86],table_:43,table_project:43,tablemattyp:46,tableproject:42,tables:30,tag:[0,3,5,38,71],tagtyp:[38,71],take:[0,13,17,21,23,26,38,43,58,60,66,71,74,76,86,88],taken:[17,86],tal:43,tanh:[23,32,35,46,67,74,76],tanhactiv:0,tanhderiv:46,taobao:19,tar:1,target:[0,8,13,21,36,74],target_dict_dim:0,target_language_word:0,targetgrad:32,targetinfoinlinkid_:40,targetperdim_:43,task:[0,8,11,17,19,21,27,40,43,46,71,74,84,86],task_backward:40,task_copy_in_args:40,task_forward:40,task_forward_backward:40,taskreadysem_:40,tasktyp:40,tasktype_:40,tau_:50,tcp:[58,60,83],tcpclient:58,tcprdma_:58,tcpserver:58,tcpsocket_:58,teach:13,team:3,technician:14,techniqu:[0,23],tee:[9,15,17,19,21],tell:[15,55],tellig:19,temp:32,templat:[17,26,30,32,36,40,43,46,55,60,62,64,65,66],templatereducesum:60,tempor:[13,17,74],temporari:[43,46],temporarili:46,tenor:43,tenro:43,tensor:[27,34,43,74],tensor_lay:43,tensorlay:42,term:[17,19,74,76],ternari:[30,46],tese:85,tesh:17,test:[2,7,8,9,11,13,14],test_all_data_in_one_period:[9,15,17,19,83,84],test_data:[21,88],test_fcgrad:23,test_layergrad:23,test_list:[9,13,70,86],test_part_000:19,test_pass:[21,82,83,84],test_period:[82,83,84],test_ratio:15,test_wait:[82,83],testabl:66,testconfig:23,testdata:19,tester:[15,21],testfcgrad:23,testfclay:23,testing:1,testing_:61,testlayergrad:23,testmodel_list:82,testresult:64,testsave_dir:82,testutil:23,text:[8,12,13,15,19],text_conv:13,text_conv_pool:15,text_fil:19,tgt:17,tgtdata:32,tgz:1,than:[0,7,11,17,19,21,23,36,38,40,43,55,60,61,64,66,71,74,76,78,86,88],thank:[8,21],the_total_number_of_paramet:52,thead:64,thei:[0,11,19,23,36,40,43,46,82,86],them:[0,9,11,13,15,21,36,38,43,46,55,61,76,82,83,85,86],therun:11,thi:[0,2,7,8,9,11,13,14,15,17,19,21,23,26,27,32,36,38,40,43,46,48,50,52,55,58,60,61,64,65,66,69,70,71,74,76,78,83,84,85,86],thing:[0,1,15,17,86],third:[11,19,40,62,74],this_col:46,this_row:46,this_row_i:46,thisimgheight:46,thisimgwidth:46,thoma:43,those:[11,17,40],though:64,thread1:65,thread2:65,thread:[15,17,19,21,23,27,33],thread_:64,thread_device_resourc:27,thread_local_rand_use_global_se:[82,83],threadbarri:[40,55,60,64],threadid:[40,46,55,84],threadid_:40,threadloc:[36,46,52,60,62],threadlocald:36,threadnum:64,threadnum_:55,threadpool:55,threadptr:55,threads_:40,threadsaf:46,threds_:40,three:[0,11,19,21,30,43,52,71,74,78,83,86],three_parameter:30,thresh:43,threshold:[32,46,60,69,71,78,83],threshold_:50,thriller:14,through:[0,7,8,9,19,21,23,40,43,61,64,66,74,88],throughout:13,thu:[21,23,40,74,86],tid:[46,55,64],time:[0,1,3,13,14,17,19,21,27,43,46,50,55,60,62,64,65,66,74,76,80,83,84,86],timelin:[43,74],timeout:66,timer:33,timer_:50,timespec:66,timestamp:[14,43,74],timestep:[32,36,43,74,86],timev:60,timewait:66,timewaitgradreadi:52,titil:15,titl:[2,14,15,38],tlstempbufs_:52,tmall:19,tmp:46,tmpbiasgrad_:43,tmpcpuinput_:43,tmpdest_:43,tmpgrad_:43,tmpmat_:[43,46],tmpmatrix:43,tmpmtx0:43,tmpmtx1:43,tmpmtx:43,tmprow0:43,tmprow1:43,tmprow2:43,tmprow3:43,tmpsrc_:43,tmpwgrad_:43,todo:[55,71,74,76],togeth:[0,43,66,74,76],token:[0,8,19,21,71,74,86],tonumpyarrayinplac:26,tonumpymat:26,tonumpymatinplac:26,too:60,tool:[0,3,11,19,40],toolkit:[1,5],top:[11,17,30,46,62,71],topid:30,topindex:40,toprotostr:26,topval:30,toronto:9,total:[7,8,21,36,38,43,55,58,61,71],totallength:58,totalpad:[32,46],totalscore_:38,totalsegments_:43,totalstate_:43,totaltime_:43,touch:19,tourism:19,tourist:21,tra:21,trace:62,track:74,track_:43,tractabl:74,tradesman:14,tradit:74,traffic:55,train:0,train_conf:[8,21],train_data:21,train_list:[9,11,13,70,86],train_part_000:19,trainabl:[32,74],trainable_pad:43,traindata:19,traindot_period:82,trained:26,trainer:[7,13],trainer_config:[7,13,15,19,85,86,88],trainer_config_help:[9,13,15,23,67,69,70,71,74,76,78,80,86],trainer_count:[15,17,19,21,82,83,84],trainer_id:[60,83],trainer_role_control:52,trainer_role_major:52,trainer_role_master:52,trainer_role_minor:52,trainer_role_single:52,trainer_role_slave:52,trainerbarrier_:40,trainerconfig:[26,40],trainerconfigpriv:26,trainercount:40,trainerid:55,trainerid_:55,trainerintern:[13,15,21],trainerpriv:26,trainerrol:52,trainerthreadptr:40,trainingtest_all_data_in_one_period:82,trainison:40,trainonebatch:26,trainrol:52,trainwholedatainonebatch:40,tran:[23,26,27,46,52,55,74,83],trane:[83,86],trans_:46,trans_full_matrix_project:43,trans_lay:43,transa:[27,30,46],transact:19,transb:[27,30,46],transdatatyp:55,transform:[0,9,13,17,23,35,52],transform_param_attr:76,translat:[8,15,19,21,43,74,76],translate:43,translation:43,translay:42,transoutvalue_:43,transpar:7,transport:[60,83],transpos:[23,27,30,34,43,46,74],transposedfullmatrixproject:42,transposit:[43,74],tranvers:40,travel:[40,86],travers:50,traversecallback:[50,60],traversecallbackvec:50,treat:[0,36,40,74],tree:[2,21,43,74,83],trg:21,trg_dict:0,trg_dict_path:0,trg_embed:0,trg_id:0,trg_ids_next:0,triain:85,trimfrom:46,trimfromcsc:46,trimfromcsr:46,trivial:86,trn:13,truck:9,true_typ:46,trust:7,truth:[13,21,71,74],trycreategradientmachin:40,trygetmod:40,tst:13,tune:[13,15,21,60,61,69],tuningasyncsgdfinish:60,tuningasyncsgdmidoutput:60,tuninglog_barrier_abstract:82,tuningsgdfinish:60,tuningsgdmidoutput:60,tupl:[36,40,52,70,74,76],ture:[40,43,66],turing:[43,74],turn:74,tutori:[0,2,7],tweet:19,twelv:21,twitter:19,two:[0,5,7,8,9,11,13,15,17,19,21,27,36,40,43,46,50,52,55,60,61,67,74,76,84,85,86],two_parameter:30,txt:[7,13,15,19,23,74,86],type:[0,9,11,13,15,17,23,26,30,32,34,35,36,38,40,43,46,50,52,54,55,58,60,61,62,64,70,71,74,76,78,80,81,83,84,86],type_:43,type_info:55,type_nam:[15,74],typecheck:43,typedef:[27,30,32,34,36,40,43,46,48,50,52,55,58,60,62,64],typenam:[26,40,43,46,55,60,62,64,65],typic:[19,36,38,71,88],u_r:43,u_z:43,ubuntu:3,udp:52,udpat:61,ufldl:74,uint32_t:[46,52],uint64_t:[46,52,54,55,60,61],unari:[30,46],unbalanc:83,unbound:0,unconstrain:19,under:[14,19,38,55,60,62],understand:[8,9,19,60],understudi:21,unemploi:14,uniform:[46,69,83],uniform_real_distribut:43,uniformli:27,uniqu:[2,83],unique_ptr:[23,36,40,43,46,48,50,52,54,55,58,60,61,64],unit:[0,2,17,22],unittest:1,unittestcheckgrad_ep:82,univ:21,univari:46,unk:[8,21],unk_idx:[13,17],unknown:74,unlabel:19,unlik:21,unlike:[17,19,64],unlimit:36,unlock:66,unnorm:43,unordered_map:[48,55,60],unsegment:74,unshar:43,unsign:[27,46],unsup:19,unsupbow:19,unsupporterror:26,until:[7,43,46,52,55,58,64,65,66],unus:[15,46,60],unzip:15,updat:[1,3,7,19,23,26,30,32,40,46,50,52,54,55,60,61,69,74,83,84],update:[2,38,44,52,53],update_type_num:52,updateaveragewindowlimit:50,updatecallback:[23,26,40,43,52],updatecallbackinpython:26,updatecallbackwrapp:26,updatecounter_:[40,52],updated_:52,updatefunct:52,updateg:34,updatehook:52,updateimpl:[52,54,61],updatemod:55,updater:54,updater_hook:54,updater_normal:61,updater_sparse_remote:61,updaterhooks_:52,updaters_:54,updatesamplesnum:38,updatethreadparamet:40,updatewithgradi:52,upon:[17,86],upscalei:27,upscalex:27,upstream:2,url:19,urls_neg:19,urls_po:19,urls_unsup:19,urng:43,usag:[8,36,60,61,71,74,76,85,86],usageratio_:36,use:[15,23,30,50,60,64,66,83],use_global_stat:74,use_gpu:[9,11,13,15,17,19,21,46,82,83,84,88],use_jpeg:9,use_old_updat:[82,83],use_seq:[15,17],use_seq_or_not:15,useapply_:50,useapplyinpserver_:61,usebatch_:43,used:[40,43,64,76],usedsegments_:60,useful:[74,86],usefull:26,useglobalstats_:43,usegpu:[23,26,36,40,43,46,52],usegpu_:[36,40,43,46,52],usel1:46,useless:7,usepam:3,useparameterappli:50,user:7,user_featur:15,user_id:15,user_meta:15,userdefinedvectorptr:52,userid:14,usernam:3,username:2,useseqparallel_:43,using:[15,19],usr:[1,3,7,83],usrdict:8,usrmodel:8,usual:[40,83,84],usualli:[19,35],utc:66,utf:8,util:[0,1,9,15,19,23],utiliti:33,val:[26,32,46],val_siz:30,valbuf:26,vale:40,valid:[11,19,27,42],validationimp:43,validationlay:42,valin:46,valu:[0,9,11,17,19,23,26,27,30,32,34,35,36,38,40,43,46,50,52,55,60,61,63,64,66,69,71,74,78,80,83,84,86,88],valuabl:60,value1:83,value2:83,value_:46,value_d:30,value_printer_evalu:38,value_typ:30,valuecount:52,valuedispatchthread:40,valuedispatchthread_:40,valuefg:32,valueframest:32,valueig:32,valuein:32,valueog:32,valueoutput:32,valueprevout:32,valuereadycond:52,valuereadycond_:40,valuereadyqueue_:40,valueresetg:32,valueresetoutput:32,values:52,valuesem_:52,valuestream_:40,valuetyp:[30,46],valuetype_:46,valueupdateg:32,vanilla:0,vanish:19,vardensedata:36,variabl:[1,3,5,19,23,26,32,34,43,46,52,54,66,74,86],varianc:[11,27,43,74],varindic:36,vast:2,vctor:38,vec:[26,46,50,52,60],vec_check:46,vecop:30,vector:[0,8,13,15,19,21,23,26,27,30,32,36,38,40,43,46,48,50,52,54,55,58,60,61,64,74,76],vector_len:30,vector_set:30,vector_size:30,vectoraddmult:55,vectoraddmultinto:55,vectorcopi:55,vectordotproduct:55,vectorenable_parallel_vector:82,vectorpriv:26,vectorptr:[40,46,50,52,60],vectors_:60,vectorscal:55,vectorscaleinto:55,vectort:46,vectyp:30,verb:17,veri:[0,9,13,19,38,46,60,74,80,86],verifi:[2,23],versa:1,version:[1,3,5,8,9,14,17,19,21,23,26,27,43,46,52,60,74,76,82,83],vertic:[11,27,74],very:26,vgg:[9,76],vgg_16_cifar:9,via:[13,43,52],vice:1,viet:43,view_data:26,view_m_data:26,virtual:[3,26,35,36,38,40,43,46,48,50,52,54,55,58,60,61,64],virtualenv:15,vision:9,visipedia:9,viterbi:17,voc_dim:13,vocab:19,voluntarili:14,wai:[0,15,17,21,23,43,50,74,76,84],wait:[27,40,43,52,55,61,64,65,66,78,83],waitaftermerg:40,waitallmajorgradreadi:52,waitallthread:40,waitandmergeoutputgrad:43,waitbeforemerg:40,waitempti:65,waitforcopyinarg:40,waitforgradi:55,waitforstatu:55,waitgradreadi:52,waitinputvalu:43,waitoutargsreadi:40,waitpassfinish:[55,60],waitpassfinishrequest:60,waitpassstart:[55,60],waitpassstartrequest:60,waitresult:64,waitvaluereadi:52,wake:[40,65,66],walk:88,wall:17,want:[1,3,4,8,11,13,15,17,19,23,36,40,74,76,83,84,86],wapper:66,war:14,wbia:11,websit:[1,9,13,17,19],wei:[17,19],weigh:74,weight:[0,9,11,23,32,34,38,40,43,46,51],weight_:[43,52],weight_act:76,weightdata:32,weightdim:[32,43],weightgrad:[32,46],weightgrad_:52,weightlast:43,weightlast_:43,weightlayer_:43,weightlist:[23,43,52],weightoffset_:43,weights_:[23,43],weights_t:23,welcom:[3,19],welcome:15,well:[3,7,9,13,23,83],were:[43,66],western:14,wether:[52,74],what:[7,13,15,36,64,69,76,78],wheel:1,when:[0,5,8,9,17,19,21,23,36,38,40,43,46,50,52,55,60,61,64,69,71,74,83,84,85,86],whenev:15,where:[0,8,11,17,21,23,35,40,43,46,74,76,78,83,84,86],whether:[9,15,19,21,23,40,43,46,52,71,74,76,83,86],which:[0,3,7,9,11,13,14,15,17,19,21,23,26,27,30,34,36,40,43,46,52,58,60,61,66,67,71,74,76,78,83,84,85,86,88],whl:1,who:[8,11,14,65],whoe:[30,46],whole:[13,14,15,21,36,38,40,71,86],whole_cont:15,whose:[0,15,17,40,60,64,86],wide:[17,64],width:[9,21,23,26,27,30,32,43,46,52,60,71,74],width_:46,width_pad:27,widthpad:43,wiil:26,wilder:86,window:[19,27,32,43],window_height:27,window_width:27,windowheight:43,windows:43,windowwidth:43,wise:[30,43,46,74],witdth:43,with_double:23,withcurrentcost:61,within:[40,43,55,60,61,74],without:[19,38,40,43,60,71],wmt14:21,wmt14_data:21,wmt14_model:21,wmt:21,woker:58,won:[11,65,66],wonder:86,word:[0,7],word_dict:[13,17],word_dim:13,word_id:86,word_slot:17,word_vector:13,word_vector_dim:[0,8],work:[0,1,2,13,15,23,27,40,43,50,61,83,86,88],worker:[27,46,50,57],workernum:64,workers_:64,workflow:[2,20],workspac:6,workspace_:43,workspacedata_:43,workspaceinbytes_:43,workthread:40,would:[3,13,17,40],wrap:[17,26,60],wrapper:22,writ:15,write:[0,7,9],writemessag:[58,60],writer:[14,66],writev:58,written:[15,19,58],wsj:17,wstride:27,www:[9,21],x64:1,x86:5,xarg:[3,23],xgbe0:83,xgbe1:83,xi_t:43,xiaojun:19,xr_t:43,xrang:23,xxbow:19,xxx:[11,21],xz_t:43,xzf:1,yaml:15,yanfei:55,year:14,yeild:9,yield:[0,13,15,17,19,86],you:[0,1,2,3,4,5,7,8,9,11,13,15,17,19,21,23,26,36,43,46,52,60,65,69,74,76,78,83,84,85,86,88],your:[2,3,4,7,13,19,23,84,86],your_host_machine:3,yourself:[3,7],yuyang18:[74,76],zachari:19,zbib:43,zeng:19,zere:83,zero:[13,23,26,30,40,43,46,55,60,61,65,69,74,78,83,86],zeroatoffset:46,zerograd:43,zeromem:[46,52],zeromemthread:46,zhidao:8,zhongqiang:43,zhou:[17,19],zip:14},titles:["Recurrent Neural Network Configuration","Build and Install","Contribute to PaddlePaddle","Docker installation guide","Build And Install PaddlePaddle","Debian Package installation guide","Cluster Train","Cluster Training","Chinese Word Embedding Model Tutorial","Image Classification Tutorial","Image Classification Tutorial","Model Zoo - ImageNet","Examples and demos","Quick Start Tutorial","MovieLens Dataset","Regression MovieLens Ratting","Semantic Role Labeling Tutorial","Semantic Role labeling Tutorial","Sentiment Analasis Tutorial","Sentiment Analysis Tutorial","Text Generation Tutorial","Text generation Tutorial","Writing New Layers","Writing New Layers","PaddlePaddle Documentation","Layer Documents","API","Cuda","CUDA","Matrix","Matrix","RNN","Neural Networks","Utils","Utilities","Activations","Data Providers","Data Providers Documents","Base Evaluator","Evaluators","Gradient Machines","Gradient Machines Documents","Layers Documents","Base","Source Code Documents","Matrix Documents","Matrix","Utils Documents","Utils","Parameter Documents","Optimizer","Parameter Documents","Parameter","Parameter Documents","Update","Client","Client Documents","Network Documents","Network","Server Documents","Server","Trainer","CustomStackTrace","enumeration_wrapper","Thread","Queue","Lock","BaseActivation","Activations","Parameter and Extra Layer Attribute","DataSources","Base","Evaluators","Model Config Interface","Base","Layers","NLP","Networks","BaseSGDOptimizer","Optimizers","BasePoolingType","Poolings","Argument Outline","Detail Description","Use Case","DataProvider Introduction","How to use PyDataProvider2","User Interface","Python Prediction API"],titleterms:{"case":84,"class":[23,62,64,65,66],"function":[8,32],"new":[22,23],absactivat:67,access:3,activat:[32,35,43,68],adadeltaoptim:78,adagradoptim:78,adamaxoptim:78,adamoptim:78,addto_lay:74,addtolay:43,agentlay:43,aggregat:[43,74],algorithm:[13,24],analasi:18,analysi:19,api:[26,32,44,87,88],appendix:13,architectur:[0,13],argument:[13,82,84,87],async:83,asyncthreadpool:64,attention:0,attribut:69,auc_evalu:71,aucevalu:38,aucvalid:43,averagelay:43,avgpool:80,base:[30,32,34,36,38,43,46,71,74],baseactiv:67,basepoolingtyp:80,basesgdoptim:78,batch_norm_lay:74,batchnormalizationlay:43,batchnormbaselay:43,beam_search:74,bidirect:19,bidirectional_lstm:76,bleu:21,block_expand_lay:74,blockexpandlay:43,blockingqueu:65,breluactiv:67,build:[1,4],cach:86,cacheonepassinmemori:36,channel:58,check:[7,23,43,74],chines:8,chunk_evalu:71,chunkevalu:38,classif:[9,10,38,71],classification_error_evalu:71,classification_error_printer_evalu:71,classificationerrorevalu:38,classificationerrorprint:38,client:[55,56,58],clone:2,cluster:[6,7,84],cmrprojectionnormlay:43,code:[2,44],column_sum_evalu:71,columnsumevalu:38,command:[13,21,84,87],commit:2,common:83,commun:83,concat:43,concat_lay:74,concatenatelay:43,concatenatelayer2:43,concurrentremoteparameterupdat:61,config:[15,17,73,84],configur:[0,7,13,15],connect:[43,74],context_project:74,contextproject:43,contribut:2,conv:[43,74],conv_oper:74,conv_shift_lay:74,convbaselay:43,convexcombinationlay:43,convolut:[9,13],convoper:43,convshiftlay:43,coorditer:43,cos_sim:74,cossimlay:43,cossimvecmatlay:43,cost:[43,74],costlay:43,cpu:84,creat:2,crf_decoding_lay:74,crf_layer:74,crfdecodinglay:43,crflayer:43,cross_entropi:74,cross_entropy_with_selfnorm:74,ctc_error_evalu:71,ctc_layer:74,ctcevaluat:38,ctclayer:43,cuda:[27,28,44],cudnnbatchnormlay:43,cudnnconvlay:43,cudnnpoollay:43,cuh:30,customstacktrac:62,dat:14,data:[0,8,9,13,15,17,19,21,36,37,43,74,87],data_lay:74,datanormlay:43,dataprovid:[36,83,85,86],dataprovidergroup:36,dataset:[14,15,21],datasourc:70,date:2,debian:[4,5],decayedadagradoptim:78,defin:[13,19,21],delv:9,demo:12,densescann:36,depend:1,deriv:23,descript:[17,83],description:14,detail:[9,83],develop:24,devic:84,dictionari:8,differ:84,distribut:83,docker:[3,4],document:[24,25,37,41,42,44,45,47,49,51,53,56,57,59],dotmul_project:74,dotmuloper:43,dotmulproject:43,download:[3,8,11,15,21],dropout_lay:76,dynam:27,embed:[8,13],embedding_lay:74,enumeration_wrapp:63,eos_lay:74,eosidchecklay:43,equation:23,evaluat:[15,38,39,72],evalutaion:21,exampl:[8,12],exercis:9,expand_lay:74,expandconvlay:43,expandlay:43,extra:69,extract:[8,15,21],extraction:11,fc_layer:74,featur:[11,14,15,17],featuremapexpandlay:43,field:15,file:[13,14,15],first_seq:74,fork:2,format:13,from:4,full_matrix_project:74,fulli:[43,74],fullmatrixproject:43,fullyconnectedlay:43,gate:0,gatedrecurrentlay:43,gatheragentlay:43,gener:[0,20,21],get_output_lay:74,getoutputlay:43,github:2,gpu:[27,83,84],gradient:[23,40,41],gradient_printer_evalu:71,gradientmachin:40,gradientmachinemodel:40,gradientprint:38,group:[43,74],gru:[32,43,76,83],gru_group:76,gru_step_lay:74,gru_unit:76,grucomput:43,grumemori:74,grusteplay:43,gserver:44,guid:[3,5,24],handl:48,hierarchicalsigmoidlay:43,hl_aggreg:30,hl_base:34,hl_batch_transpos:30,hl_cuda:27,hl_cuda_cubla:27,hl_cuda_cudnn:27,hl_dso_load:27,hl_matrix:30,hl_matrix_appli:30,hl_matrix_bas:30,hl_matrix_op:30,hl_matrix_typ:30,hl_spars:30,hl_sse_matrix_kernel:30,hl_table_appli:30,hl_thread:34,hl_time:34,hl_top_k:30,how:86,hppl:34,hsigmoid:74,huber_cost:74,hubertwoclass:43,identity_project:74,identityactiv:67,identityoffsetproject:43,identityproject:43,ifieldscann:36,imag:3,image:[9,10,12,74,76],imagenet:11,imdb:19,img_cmrnorm_lay:74,img_conv_bn_pool:76,img_conv_group:76,img_conv_lay:74,img_pool_lay:74,implement:23,indexscann:36,inferenc:13,info:11,init_hook:86,initial:84,input_typ:86,instal:[3,4,5],install:[1,4,13],interfac:[11,73,87],interpolation_lay:74,interpolationlay:43,introduct:[8,11,19,21,85],ipydataprovid:36,ipydataprovidercach:36,job:7,keep:2,kill:7,label:[16,17],lambda_cost:74,lambdacost:43,last_seq:74,launch:7,layer:[22,23,25,42,43,69,74,75,84],layeroutput:74,layertyp:74,learn:83,lib:27,line:[13,87],linear_comb_lay:74,linearactiv:67,linearchaincrf:43,linearchainctc:43,link:27,local:84,lock:66,lockedcondit:66,log:13,logist:13,lstm:[17,19,32,43,76,83],lstm_step_lay:74,lstmcomput:43,lstmemori:74,lstmemory_group:76,lstmemory_unit:76,lstmlayer:43,lstmsteplay:43,machin:[40,41],math:[43,44,74],matrix:[29,30,45,46,83],maxframe_printer_evalu:71,maxframeprint:38,maxid_lay:74,maxid_printer_evalu:71,maxidlay:43,maxidprint:38,maxlay:43,maxpool:80,mdlstm:43,mdlstmlayer:43,memori:48,messag:58,meta:15,metric:83,misc:76,mix:[43,74,84],mixed_lay:74,model:[0,7,8,9,11,12,13,21,32,73,84,86],movi:[14,15],movielen:[14,15],movies:14,multi_binary_label_cross_entropi:74,multibinarylabelcrossentropi:43,multiclasscrossentropi:43,multiclasscrossentropywithselfnorm:43,multidataprovid:36,multigradientmachin:40,multinomialsampl:43,multiplexlay:43,multithreadwork:64,namespac:63,ncelayer:43,network:[0,9,11,13,15,17,32,40,57,58,77,84],neural:[0,9,13,15,17,32],neuralnetwork:40,nlp:[12,76,83],nocachestrategi:36,non:86,norm:[43,74],normlay:43,notic:3,object:15,observat:[8,11],operat:43,optimiz:[13,50,79],optional:[1,8],other:[30,46],outerprodlay:43,outlin:82,output:[7,76],overview:13,packag:[4,5],paddl:63,paddlepaddl:[2,3,4,8,21,24],parallel_nn:84,parallelneuralnetwork:40,paramet:[8,11,44,49,51,52,53,69,83],parameterrelulay:43,paraphras:8,pass:84,perform:[3,83],pnpair_evalu:71,pnpairevalu:38,pnpairvalid:43,pool:[43,74,81],pooling_lay:74,poollay:43,poolprojectionlay:43,power_lay:74,powerlay:43,pre:7,precision_recall_evalu:71,precisionrecallevalu:38,predict:[9,11,15,17,19,87,88],prepar:[0,7,8,9,15,19,21],preprocess:[8,9,13,15,21],pretrain:[8,21],print:71,printer:38,project:43,proto:36,protodataprovid:36,protosequencedataprovid:36,provid:[13,15,17,36,37,86,87],pserver:44,pull:2,push:2,pydataprovid:36,pydataprovider2:86,python:[11,13,15,23,88],queue:65,quick:13,randomnumb:83,rank:[38,71],rank_cost:74,rankingcost:43,rat:15,rate:14,ratings:14,reader:58,readlockguard:66,recommend:12,recurr:[0,13,40,43,74,76],recurrent_group:74,recurrent_lay:74,recurrentlay:43,refer:[17,19,86,87],regress:[13,15],regular:52,relat:32,reluactiv:67,remot:3,remoteparameterupdat:61,request:2,requir:[1,2,7],reshap:[43,74],resizelay:43,resnet:11,resourc:[27,34],responsenormlay:43,result:[7,21],revis:8,rmspropoptim:78,rnn:[31,32,83],role:[16,17],run:[3,17],rwlock:66,sampl:[43,74],sampling_id_lay:74,samplingidlay:43,scaling_lay:74,scalinglay:43,scatteragentlay:43,selective_fc_lay:74,selectivefullyconnectedlay:43,semant:[16,17],semaphor:66,sentiment:[18,19],seqtext_printer_evalu:71,sequenc:0,sequence_conv_pool:76,sequenceagentlay:43,sequenceclassificationerrorevalu:38,sequenceconcatlay:43,sequencegatheragentlay:43,sequencelastinstancelay:43,sequencereshapelay:43,sequencescann:36,sequencescatteragentlay:43,sequencesoftmaxactiv:67,sequencetextprint:38,sequencetobatch:43,sequenti:86,server:[58,59,60,83],set:78,sgd:83,sigmoidactiv:67,simpl:0,simple_attent:76,simple_gru:76,simple_img_conv_pool:76,simple_lstm:76,slope_intercept_lay:74,slopeinterceptlay:43,socket:58,softbinaryclasscrossentropi:43,softmaxactiv:67,softreluactiv:67,sourc:[4,44],spars:[30,46,84],sparsenonvaluescann:36,sparseremoteparameterupdat:61,sparseremoteparameterupdatercomposit:61,sparsevaluescann:36,specifi:[8,84],spinlock:66,split:15,squareactiv:67,squarerootnpool:80,stack:19,standard:13,stanhactiv:67,start:13,subsequencelay:43,subset:43,sum_evalu:71,sum_to_one_norm_lay:74,sumevalu:38,summari:13,sumofsquarescostlay:43,sumpool:80,sumtoonenormlay:43,syncthreadpool:64,table_project:74,tableproject:43,tanhactiv:67,tensor_lay:74,tensorlay:43,test:[15,17,19,23,83,84],text:[20,21],text_conv_pool:76,thread:[34,64],threadbarri:66,threadwork:64,timer:34,train:[6,7,8,9,13,15,17,19,21,83,84],trainer:[15,44,61],trainerstat:61,trainerthread:40,trans_full_matrix_project:74,trans_lay:74,transfer:13,translay:43,transposedfullmatrixproject:43,tune:83,tutori:[8,9,10,13,16,17,18,19,20,21,24],ubuntu14:1,unit:[23,83],update:54,use:84,user:[8,14,15,19,21,24,87],users:14,util:[33,38,44,47,48,71],utiliti:34,valid:43,validationlay:43,value_printer_evalu:71,valueprint:38,vector:83,vgg_16_network:76,visual:11,weight:52,word:[8,13],worker:58,workflow:21,workspac:7,wrapper:[23,27],write:[13,22,23],zoo:[11,12]}})
\ No newline at end of file
diff --git a/doc/source/api/api.html b/doc/source/api/api.html
index 969ae7b9ee..3bc68e4eff 100644
--- a/doc/source/api/api.html
+++ b/doc/source/api/api.html
@@ -6,7 +6,7 @@
   
     
     
-    API — PaddlePaddle  documentation
+    API — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
       
 
@@ -122,10 +122,10 @@
 
    
 void initPaddle(int argc, char **argv)
    
 
    Initialize paddle. 
    
-
    In python, this method should be invoked as  
    import sys
-import paddle
-paddle.initPaddle(sys.argv)
-or you can change arguments as any list of str.
+
    In python, this method should be invoked as  
    import sys
+import paddle
+paddle.initPaddle(sys.argv)
+or you can change arguments as any list of str.
 
    
 
    
 
    
@@ -281,18 +281,17 @@ or you can change arguments as any list of str.
 void toNumpyMatInplace(float **view_data, int *dim1, int *dim2)
 
    Cast to numpy matrix.
    
 
    
-Example:  
    import paddle
+Example:  
    import paddle
 m = paddle.Matrix.createZero(10,2)
 numpy_mat = m.toNumpyMat()
 
    
 
    
 
    
 
    Note
    
-
    This method take no parameter in python. 
    
-
    Note
    
-
    This method in python will return a numpy matrix, not void. 
    
-
    Note
    
-
    Only CpuDenseMatrix is supported.
    
+
    This method take no parameter in python. 
    
+
    This method in python will return a numpy matrix, not void. 
    
+
    Only CpuDenseMatrix is supported.
    
+
    
 
    
 
    
 
    
@@ -378,14 +377,14 @@ Example:  
    Public Static Functions
    
 
    
 
    
-Matrix *createZero(size_t height, size_t width, bool useGpu = false)
    
-
    Create A Matrix with height,width, which is filled by zero. 
    
+Matrix *createZero(size_t height, size_t width, bool useGpu = false)
+
    Create A Matrix with height,width, which is filled by zero. 
    
 
    
 
 
    
 
    
-Matrix *createSparse(size_t height, size_t width, size_t nnz, bool isNonVal = true, bool trans = false, bool useGpu = false)
    
-
    Create Sparse Matrix.
    
+Matrix *createSparse(size_t height, size_t width, size_t nnz, bool isNonVal = true, bool trans = false, bool useGpu = false)
+
    Create Sparse Matrix.
    
 
    After create sparse, sparseCopyFrom can be used to fill matrix.
    
 
    
 
    Note
    
@@ -402,8 +401,8 @@ Example:  
    
 
    
-Matrix *createDense(const std::vector<float> &data, size_t height, size_t width, bool useGpu = false)
    
-
    Create Dense Matrix.
    
+Matrix *createDense(const std::vector<float> &data, size_t height, size_t width, bool useGpu = false)
+
    Create Dense Matrix.
    
 
    
 
    Note
    
 
    the value will be copy into a new matrix. 
    
@@ -419,8 +418,8 @@ Example:  
    
 
    
-Matrix *createCpuDenseFromNumpy(float *data, int dim1, int dim2, bool copy = false)
    
-
    Create Cpu Dense Matrix from numpy matrix, dtype=float32
    
+Matrix *createCpuDenseFromNumpy(float *data, int dim1, int dim2, bool copy = false)
+
    Create Cpu Dense Matrix from numpy matrix, dtype=float32
    
 
    
 
    Parameters
    
 
      
@@ -440,8 +439,8 @@ Example:  
      
 
      
-Matrix *createGpuDenseFromNumpy(float *data, int dim1, int dim2)
      
-
      Create Gpu Dense Matrix from numpy matrix, dtype=float32. 
      
+Matrix *createGpuDenseFromNumpy(float *data, int dim1, int dim2)
+
      Create Gpu Dense Matrix from numpy matrix, dtype=float32. 
      
 
    
 
 
    
@@ -454,7 +453,7 @@ Example:  
    
 
    
-DISABLE_COPY_AND_ASSIGN(Matrix)
    
+DISABLE_COPY_AND_ASSIGN(Matrix)
 
    
 
 
    
@@ -475,7 +474,7 @@ Example:  
    Private Static Functions
    
 
    
 
    
-Matrix *createByPaddleMatrixPtr(void *sharedPtr)
    
+Matrix *createByPaddleMatrixPtr(void *sharedPtr)
 
    
 
 
    
@@ -556,28 +555,28 @@ Example:  
    Public Static Functions
    
 
    
 
    
-Vector *createZero(size_t sz, bool useGpu = false)
    
-
    Create Vector filled with zero. 
    
+Vector *createZero(size_t sz, bool useGpu = false)
+
    Create Vector filled with zero. 
    
 
    
 
 
    
 
    
-Vector *create(const std::vector<float> &data, bool useGpu = false)
    
-
    Create Vector from list of float.
    
+Vector *create(const std::vector<float> &data, bool useGpu = false)
+
    Create Vector from list of float.
    
 
    It will create a new vector, and copy data into it. 
    
 
    
 
 
    
 
    
-Vector *createCpuVectorFromNumpy(float *data, int dim, bool copy = false)
    
-
    Create Cpu Vector from numpy array, which dtype=float32
    
+Vector *createCpuVectorFromNumpy(float *data, int dim, bool copy = false)
+
    Create Cpu Vector from numpy array, which dtype=float32
    
 
    If copy is false, it will create vector inplace. 
    
 
    
 
 
    
 
    
-Vector *createGpuVectorFromNumpy(float *data, int dim)
    
-
    Create Gpu Vector from numpy array, which dtype=float32. 
    
+Vector *createGpuVectorFromNumpy(float *data, int dim)
+
    Create Gpu Vector from numpy array, which dtype=float32. 
    
 
    
 
 
    
@@ -585,7 +584,7 @@ Example:  
    Private Functions
    
 
    
 
    
-DISABLE_COPY_AND_ASSIGN(Vector)
    
+DISABLE_COPY_AND_ASSIGN(Vector)
 
    
 
 
    
@@ -611,7 +610,7 @@ Example:  
    Private Static Functions
    
 
    
 
    
-Vector *createByPaddleVectorPtr(void *ptr)
    
+Vector *createByPaddleVectorPtr(void *ptr)
 
    
 
 
    
@@ -677,7 +676,7 @@ Example:  
    
 
    
-const int &operator[](const size_t idx) const
    
+const int &operator[](const size_t idx) const
 
    
 
 
    
@@ -707,27 +706,27 @@ Example:  
    Public Static Functions
    
 
    
 
    
-IVector *createZero(size_t sz, bool useGpu = false)
    
-
    Create IVector filled with zero. 
    
+IVector *createZero(size_t sz, bool useGpu = false)
+
    Create IVector filled with zero. 
    
 
    
 
 
    
 
    
-IVector *create(const std::vector<int> &data, bool useGpu = false)
    
-
    Create IVector from list of int. It will create a new vector, and copy data into it. 
    
+IVector *create(const std::vector<int> &data, bool useGpu = false)
+
    Create IVector from list of int. It will create a new vector, and copy data into it. 
    
 
    
 
 
    
 
    
-IVector *createCpuVectorFromNumpy(int *data, int dim, bool copy = false)
    
-
    Create Cpu IVector from numpy array, which dtype=int32
    
+IVector *createCpuVectorFromNumpy(int *data, int dim, bool copy = false)
+
    Create Cpu IVector from numpy array, which dtype=int32
    
 
    If copy is false, it will create vector inplace 
    
 
    
 
 
    
 
    
-IVector *createGpuVectorFromNumy(int *data, int dim)
    
-
    Create Gpu IVector from numpy array, which dtype=int32 
    
+IVector *createGpuVectorFromNumy(int *data, int dim)
+
    Create Gpu IVector from numpy array, which dtype=int32 
    
 
    
 
 
    
@@ -740,7 +739,7 @@ Example:  
    
 
    
-DISABLE_COPY_AND_ASSIGN(IVector)
    
+DISABLE_COPY_AND_ASSIGN(IVector)
 
    
 
 
    
@@ -761,7 +760,7 @@ Example:  
    Private Static Functions
    
 
    
 
    
-IVector *createByPaddleVectorPtr(void *ptr)
    
+IVector *createByPaddleVectorPtr(void *ptr)
 
    
 
 
    
@@ -778,7 +777,7 @@ Example:  
    
 
    
 class Arguments
    
-
    #include <PaddleAPI.h>
    The Arguments is actual a std::vector<paddle::Argument> in paddle. 
    
+
    #include <PaddleAPI.h>
    The Arguments is actual a std::vector<paddle::Argument> in paddle. 
    
 
    
 
    Public Functions
    
 
    
@@ -801,7 +800,7 @@ Example:  
    
 
    
 Matrix *getSlotValue(size_t idx) const
    
-
    The get functions of Arguments
    
+
    The get functions of Arguments
    
 
    the param idx is the slot id 
    
 
    
 
@@ -838,8 +837,8 @@ Example:  
    
 
    
 void setSlotValue(size_t idx, Matrix *mat)
    
-
    The set functions of Arguments.
    
-
    The param idx is the slot id. The other param is the input Matrix or vector. 
    
+
    The set functions of Arguments.
    
+
    The param idx is the slot id. The other param is the input Matrix or vector. 
    
 
    
 
 
    
@@ -872,7 +871,7 @@ Example:  
    Public Static Functions
    
 
    
 
    
-Arguments *createArguments(size_t slotNum)
    
+Arguments *createArguments(size_t slotNum)
 
    Create a arguments with size. Note that it can be zero. 
    
 
    
 
@@ -886,7 +885,7 @@ Example:  
    
 
    
-DISABLE_COPY_AND_ASSIGN(Arguments)
    
+DISABLE_COPY_AND_ASSIGN(Arguments)
 
    
 
 
    
@@ -907,7 +906,7 @@ Example:  
    Private Static Functions
    
 
    
 
    
-Arguments *createByPaddleArgumentVector(void *ptr)
    
+Arguments *createByPaddleArgumentVector(void *ptr)
 
    
 
 
    
@@ -952,7 +951,7 @@ Example:  
    Private Functions
    
 
    
 
    
-DISABLE_COPY_AND_ASSIGN(ParameterConfig)
    
+DISABLE_COPY_AND_ASSIGN(ParameterConfig)
 
    
 
 
    
@@ -978,13 +977,13 @@ Example:  
    Private Static Functions
    
 
    
 
    
-ParameterConfig *createParameterConfigFromParameterSharedPtr(void *ptr)
    
+ParameterConfig *createParameterConfigFromParameterSharedPtr(void *ptr)
 
    Internal methods 
    
 
    
 
 
    
 
    
-ParameterConfig *createParameterConfigFromParameterPtr(void *ptr)
    
+ParameterConfig *createParameterConfigFromParameterPtr(void *ptr)
 
    
 
 
    
@@ -1029,7 +1028,7 @@ Example:  
    Public Static Functions
    
 
    
 
    
-OptimizationConfig *createFromProtoString(const std::string &str)
    
+OptimizationConfig *createFromProtoString(const std::string &str)
 
    
 
 
    
@@ -1037,7 +1036,7 @@ Example:  
    Private Functions
    
 
    
 
    
-DISABLE_COPY_AND_ASSIGN(OptimizationConfig)
    
+DISABLE_COPY_AND_ASSIGN(OptimizationConfig)
 
    
 
 
    
@@ -1093,7 +1092,7 @@ Example:  
    
 
    
 Vector *getBuf(ParameterType type)
    
-
    get buf in Parameter 
    
+
    get buf in Parameter 
    
 
    
 
 
    
@@ -1117,7 +1116,7 @@ Example:  
    
 
    
-DISABLE_COPY_AND_ASSIGN(Parameter)
    
+DISABLE_COPY_AND_ASSIGN(Parameter)
 
    
 
 
    
@@ -1133,12 +1132,12 @@ Example:  
    Private Static Functions
    
 
    
 
    
-Parameter *createFromRawPtr(void *ptr)
    
+Parameter *createFromRawPtr(void *ptr)
 
    
 
 
    
 
    
-Parameter *createFromSharedPtr(void *ptr)
    
+Parameter *createFromSharedPtr(void *ptr)
 
    
 
 
    
@@ -1160,8 +1159,8 @@ Example:  
    
 
    
 class ModelConfig
    
-
    #include <PaddleAPI.h>
    You can only get model config from TrainerConfig.
    
-
    It is used by GradientMachine. 
    
+
    #include <PaddleAPI.h>
    You can only get model config from TrainerConfig.
    
+
    It is used by GradientMachine. 
    
 
    
 
    Public Functions
    
 
    
@@ -1179,7 +1178,7 @@ Example:  
    
 
    
-DISABLE_COPY_AND_ASSIGN(ModelConfig)
    
+DISABLE_COPY_AND_ASSIGN(ModelConfig)
 
    
 
 
    
@@ -1219,8 +1218,8 @@ Example:  
    
 
    
 class TrainerConfig
    
-
    #include <PaddleAPI.h>
    To get TrainerConfig from file.
    
-
    It is used by GradientMachine. 
    
+
    #include <PaddleAPI.h>
    To get TrainerConfig from file.
    
+
    It is used by GradientMachine. 
    
 
    
 
    Public Functions
    
 
    
@@ -1243,7 +1242,7 @@ Example:  
    Public Static Functions
    
 
    
 
    
-TrainerConfig *createFromTrainerConfigFile(const std::string &configPath)
    
+TrainerConfig *createFromTrainerConfigFile(const std::string &configPath)
 
    
 
 
    
@@ -1256,7 +1255,7 @@ Example:  
    
 
    
-DISABLE_COPY_AND_ASSIGN(TrainerConfig)
    
+DISABLE_COPY_AND_ASSIGN(TrainerConfig)
 
    
 
 
    
@@ -1275,7 +1274,7 @@ Example:  
    class UpdateCallback
 
    #include <PaddleAPI.h>
    The callback in backword.
    
 
    You can inherit this class in python.
    
-
     
    class UpdateCallbackInPython(paddle.UpdateCallback):
+
     
    class UpdateCallbackInPython(paddle.UpdateCallback):
   def __init__(self):
     paddle.UpdateCallback.__init__(self)
 
@@ -1319,7 +1318,7 @@ Example:  
    Private Functions
    
 
    
 
    
-DISABLE_COPY_AND_ASSIGN(ParameterTraverseCallback)
    
+DISABLE_COPY_AND_ASSIGN(ParameterTraverseCallback)
 
    
 
 
    
@@ -1401,7 +1400,7 @@ Example:  
    Public Static Functions
    
 
    
 
    
-ParameterOptimizer *create(OptimizationConfig *config)
    
+ParameterOptimizer *create(OptimizationConfig *config)
 
    
 
 
    
@@ -1409,7 +1408,7 @@ Example:  
    Private Functions
    
 
    
 
    
-DISABLE_COPY_AND_ASSIGN(ParameterOptimizer)
    
+DISABLE_COPY_AND_ASSIGN(ParameterOptimizer)
 
    
 
 
    
@@ -1441,12 +1440,12 @@ Example:  
    
 
    
 void forward(const Arguments &inArgs, Arguments *outArgs, PassType passType)
    
-
    The forward stage of GradientMachine.
    
+
    The forward stage of GradientMachine.
    
 
    
 
    Note
    
-
    the outArgs could be zero length arguemnts. 
    
-
    Note
    
-
    THIS METHOD IS VERY USEFULL FOR PREDICT FROM TRAINED MODEL. 
    
+
    the outArgs could be zero length arguemnts. 
    
+
    THIS METHOD IS VERY USEFULL FOR PREDICT FROM TRAINED MODEL. 
    
+
    
 
    
 
    
 
    
@@ -1454,7 +1453,7 @@ Example:  
    
 
    
 void backward(const UpdateCallback &callback = UpdateCallback ())
    
-
    The backward stage of GradientMachine.
    
+
    The backward stage of GradientMachine.
    
 
    
 
    Note
    
 
    Currently the ParameterUpdater is not wrapped in SWIG, so backward cannot actually train a network. But you can write a update callback to change the parameter or implement a ParameterUpdater in python side. 
    
@@ -1509,16 +1508,16 @@ Example:  
    Public Static Functions
    
 
    
 
    
-GradientMachine *createByConfigProtoStr(const std::string &protoStr, GradientMatchineCreateMode mode = CREATE_MODE_NORMAL, const std::vector<int> &parameterTypes = defaultParamTypes)
    
+GradientMachine *createByConfigProtoStr(const std::string &protoStr, GradientMatchineCreateMode mode = CREATE_MODE_NORMAL, const std::vector<int> &parameterTypes = defaultParamTypes)
 
    Create By ProtoStr.
    
 
    The ProtoStr can be generate by python’s protobuf code. 
    
 
    
 
 
    
 
    
-GradientMachine *createByModelConfig(ModelConfig *conf, GradientMatchineCreateMode mode = CREATE_MODE_NORMAL, const std::vector<int> &parameterTypes = defaultParamTypes)
    
-
    Create by ModelConfig object.
    
-
    To get ModelConfig, you can get TrainerConfig from config file, then get model config by TrainerConfig 
    
+GradientMachine *createByModelConfig(ModelConfig *conf, GradientMatchineCreateMode mode = CREATE_MODE_NORMAL, const std::vector<int> &parameterTypes = defaultParamTypes)
+
    Create by ModelConfig object.
    
+
    To get ModelConfig, you can get TrainerConfig from config file, then get model config by TrainerConfig 
    
 
    
 
 
    
@@ -1531,7 +1530,7 @@ Example:  
    
 
    
-DISABLE_COPY_AND_ASSIGN(GradientMachine)
    
+DISABLE_COPY_AND_ASSIGN(GradientMachine)
 
    
 
 
    
@@ -1547,7 +1546,7 @@ Example:  
    Private Static Functions
    
 
    
 
    
-GradientMachine *createFromPaddleModelPtr(void *confPtr, GradientMatchineCreateMode mode, const std::vector<int> &types)
    
+GradientMachine *createFromPaddleModelPtr(void *confPtr, GradientMatchineCreateMode mode, const std::vector<int> &types)
 
    
 
 
    
@@ -1645,8 +1644,8 @@ Example:  
    Public Static Functions
    
 
    
 
    
-Trainer *createByCommandLine()
    
-
    Create A Trainer By TrainerConfig. using paddle command line. 
    
+Trainer *createByCommandLine()
+
    Create A Trainer By TrainerConfig. using paddle command line. 
    
 
    
 
 
    
@@ -1659,7 +1658,7 @@ Example:  
    
 
    
-DISABLE_COPY_AND_ASSIGN(Trainer)
    
+DISABLE_COPY_AND_ASSIGN(Trainer)
 
    
 
 
    
@@ -1687,13 +1686,13 @@ Example:  
    
 
    
-virtual size_t getSize() const = 0
    
+virtual size_t getSize() const = 0
 
    Number of result. 
    
 
    
 
 
    
 
    
-virtual std::string getSentence(size_t id, bool split = false) const = 0
    
+virtual std::string getSentence(size_t id, bool split = false) const = 0
 
    Get sentence from dictionary.
    
 
    
 
    Parameters
    
@@ -1710,12 +1709,12 @@ Example:  
    
 
    
-virtual std::vector<int> getSequence(size_t id) const = 0
    
+virtual std::vector<int> getSequence(size_t id) const = 0
 
    
 
 
    
 
    
-virtual float getScore(size_t id) const = 0
    
+virtual float getScore(size_t id) const = 0
 
    
 
 
    
@@ -1737,9 +1736,9 @@ Example:  
    Generate Sequence by input.
    
 
    
 
    Note
    
-
    The inArgs is just one sequence of data. 
    
-
    Note
    
-
    The return will get a N-best generate result by inArgs. Sort by score. 
    
+
    The inArgs is just one sequence of data. 
    
+
    The return will get a N-best generate result by inArgs. Sort by score. 
    
+
    
 
    
 
    
 
    
@@ -1774,7 +1773,7 @@ Example:  
    Private Functions
    
 
    
 
    
-DISABLE_COPY_AND_ASSIGN(SequenceGenerator)
    
+DISABLE_COPY_AND_ASSIGN(SequenceGenerator)
 
    
 
 
    
@@ -1795,7 +1794,7 @@ Example:  
    Private Static Functions
    
 
    
 
    
-SequenceGenerator *createByGradientMachineSharedPtr(void *ptr)
    
+SequenceGenerator *createByGradientMachineSharedPtr(void *ptr)
 
    
 
 
    
@@ -1843,14 +1842,11 @@ Example:  
    
   
    Quick search
    
     
    
-      
-      
+      
    
+      
    
       
       
     
    
-    
    
-    Enter search terms or a module, class or function name.
-    
    
 
    
 
         
    
@@ -1872,13 +1868,13 @@ Example:  
    
           previous |
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/source/cuda/cuda/cuda.html b/doc/source/cuda/cuda/cuda.html
index 6d5fd216c2..eec5916466 100644
--- a/doc/source/cuda/cuda/cuda.html
+++ b/doc/source/cuda/cuda/cuda.html
@@ -6,7 +6,7 @@
   
     
     
-    Cuda — PaddlePaddle  documentation
+    Cuda — PaddlePaddle  documentation
     
     
     
@@ -45,9 +45,9 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    • 
-          
  • CUDA »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    • 
+          
  • CUDA »
    •  
       
     
      
 
@@ -131,6 +131,219 @@
 
    GPU Resources
    
 
    
 
    hl_cuda.ph
    
+
    
+
    Defines
    
+
    
+
    
+HL_CUDA_PH_
    
+
    
+
+
    
+
    
+
    Typedefs
    
+
    
+
    
+typedef struct _global_device_resources *global_device_resources
    
+
    
+
+
    
+
    
+typedef struct _thread_device_resources *thread_device_resources
    
+
    
+
+
    
+
    
+typedef struct _hl_device_prop *hl_device_prop
    
+
    
+
+
    
+
    
+
    Functions
    
+
    
+
    
+void hl_create_thread_resources(int device, thread_device_resources device_res)
    
+
    thread device resource allocation. 
    
+
    create cuda stream and cuda event, allocate gpu memory and host page-lock memory for threads.
    
+
    
+
    Parameters
    
+
      
+
    • device - 
      device number. 
      
+
      • 
+
    • device_res - 
      device properties. 
      
+
      • 
+
    
+
    
+
    
+
    
+
    
+
+
    
+
    
+void hl_create_global_resources(hl_device_prop device_prop)
    
+
    global device resource allocation. 
    
+
    create cuda stream, initialize cublas, curand and cudnn.
    
+
    
+
    Parameters
    
+
      
+
    • device_prop - 
      device properties. 
      
+
      • 
+
    
+
    
+
    
+
    
+
    
+
+
    
+
    
+
    
+struct _hl_event_st
    
+
    hppl event. 
    
+
    
+
    Parameters
    
+
      
+
    • cuda - 
      event. 
      
+
      • 
+
    
+
    
+
    
+
    
+
    
+
    Public Members
    
+
    
+
    
+cudaEvent_t cu_event
    
+
    
+
+
    
+
    
+
+
    
+
    
+struct _global_device_resources
    
+
    global device resources. 
    
+
    
+
    Parameters
    
+
      
+
    • *stream - 
      device global stream. 
      
+
      • 
+
    • handle - 
      devcie cublas handle. 
      
+
      • 
+
    • gen - 
      device curand generator. 
      
+
      • 
+
    • cudnn_handle - 
      cudnn handle. 
      
+
      • 
+
    • *gen_mutex - 
      gen lock. 
      
+
      • 
+
    
+
    
+
    
+
    
+
    
+
    Public Members
    
+
    
+
    
+cudaStream_t *stream
    
+
    
+
+
    
+
    
+cublasHandle_t handle
    
+
    
+
+
    
+
    
+curandGenerator_t gen
    
+
    
+
+
    
+
    
+cudnnHandle_t cudnn_handle
    
+
    
+
+
    
+
    
+pthread_mutex_t *gen_mutex
    
+
    
+
+
    
+
    
+
+
    
+
    
+struct _thread_device_resources
    
+
    
+
    Public Members
    
+
    
+
    
+cudaStream_t *stream
    
+
    
+
+
    
+
    
+real *gpu_mem
    
+
    
+
+
    
+
    
+real *cpu_mem
    
+
    
+
+
    
+
    
+cudaEvent_t mem_event
    
+
    
+
+
    
+
    
+
+
    
+
    
+struct _hl_device_prop
    
+
    
+
    Public Members
    
+
    
+
    
+int device
    
+
    
+
+
    
+
    
+int device_type
    
+
    
+
+
    
+
    
+char device_name[256]
    
+
    
+
+
    
+
    
+size_t device_mem
    
+
    
+
+
    
+
    
+int major
    
+
    
+
+
    
+
    
+int minor
    
+
    
+
+
    
+
    
+bool is_local
    
+
    
+
+
    
+
    
+global_device_resources device_resources
    
+
    
+
+
    
+
    
+
 
    
 
    
 
    hl_cuda.h
    
@@ -138,7 +351,7 @@
 
    Typedefs
    
 
    
 
    
-typedef struct _hl_event_st *hl_event_t
    
+typedef struct _hl_event_st *hl_event_t
 
    HPPL event. 
    
 
    
 
@@ -662,15 +875,15 @@
 
    
 
 
    
-
    
-void hl_cuda_event_query(hl_event_t event, bool &isNotReady)
    
-
    hppl query event. 
    
+
    
+bool hl_cuda_event_is_ready(hl_event_t event)
    
+
    check cuda event is ready 
    
 
    
+
    Return
    
+
    true cuda event is ready. false cuda event is not ready. 
    
 
    Parameters
    
 
      
-
    • event - 
      cuda event to query. 
      
-
      • 
-
    • isNotReady - 
      this work under device has not yet been completed, vice versa. 
      
+
    • event - 
      cuda event to query.
      
 
      • 
 
    
 
    
@@ -696,7 +909,7 @@
 
    
 
    
 void hl_matrix_transpose(real *A_d, real *C_d, int dimM, int dimN, int lda, int ldc)
    
-
    Matrix transpose: C_d = T(A_d) 
    
+
    Matrix transpose: C_d = T(A_d) 
    
 
    
 
    Parameters
    
 
      
@@ -1546,6 +1759,229 @@
 
    
 
    
 
    hl_cuda_cudnn.h
    
+
    
+
    Defines
    
+
    
+
    
+HL_CUDA_CUDNN_PH_
    
+
    
+
+
    
+
    
+GET_TENSOR_DESCRIPTOR(image)
    
+
    
+
+
    
+
    
+GET_FILTER_DESCRIPTOR(filter)
    
+
    
+
+
    
+
    
+GET_CONVOLUTION_DESCRIPTOR(conv)
    
+
    
+
+
    
+
    
+
    Typedefs
    
+
    
+
    
+typedef struct _cudnn_tensor_descriptor *cudnn_tensor_descriptor
    
+
    
+
+
    
+
    
+typedef struct _cudnn_pooling_descriptor *cudnn_pooling_descriptor
    
+
    
+
+
    
+
    
+typedef struct _cudnn_filter_descriptor *cudnn_filter_descriptor
    
+
    
+
+
    
+
    
+typedef struct _cudnn_convolution_descriptor *cudnn_convolution_descriptor
    
+
    
+
+
    
+
    
+
    
+struct _cudnn_tensor_descriptor
    
+
    
+
    Public Members
    
+
    
+
    
+cudnnTensorDescriptor_t desc
    
+
    
+
+
    
+
    
+cudnnTensorFormat_t format
    
+
    
+
+
    
+
    
+cudnnDataType_t data_type
    
+
    
+
+
    
+
    
+int batch_size
    
+
    
+
+
    
+
    
+int feature_maps
    
+
    
+
+
    
+
    
+int height
    
+
    
+
+
    
+
    
+int width
    
+
    
+
+
    
+
    
+
+
    
+
    
+struct _cudnn_pooling_descriptor
    
+
    
+
    Public Members
    
+
    
+
    
+cudnnPoolingDescriptor_t desc
    
+
    
+
+
    
+
    
+cudnnPoolingMode_t mode
    
+
    
+
+
    
+
    
+int window_height
    
+
    
+
+
    
+
    
+int window_width
    
+
    
+
+
    
+
    
+int stride_height
    
+
    
+
+
    
+
    
+int stride_width
    
+
    
+
+
    
+
    
+
+
    
+
    
+struct _cudnn_filter_descriptor
    
+
    
+
    Public Members
    
+
    
+
    
+cudnnFilterDescriptor_t desc
    
+
    
+
+
    
+
    
+cudnnDataType_t data_type
    
+
    
+
+
    
+
    
+int output_feature_maps
    
+
    
+
+
    
+
    
+int input_feature_maps
    
+
    
+
+
    
+
    
+int filter_height
    
+
    
+
+
    
+
    
+int filter_width
    
+
    
+
+
    
+
    
+
+
    
+
    
+struct _cudnn_convolution_descriptor
    
+
    
+
    Public Members
    
+
    
+
    
+cudnnConvolutionDescriptor_t desc
    
+
    
+
+
    
+
    
+hl_tensor_descriptor input_image
    
+
    
+
+
    
+
    
+hl_filter_descriptor filter
    
+
    
+
+
    
+
    
+int padding_height
    
+
    
+
+
    
+
    
+int padding_width
    
+
    
+
+
    
+
    
+int stride_height
    
+
    
+
+
    
+
    
+int stride_width
    
+
    
+
+
    
+
    
+int upscalex
    
+
    
+
+
    
+
    
+int upscaley
    
+
    
+
+
    
+
    
+cudnnConvolutionMode_t mode
    
+
    
+
+
    
+
    
+
 
    
 
    
 
@@ -1594,14 +2030,11 @@
 
 
         
@@ -1623,14 +2056,14 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    • 
-          
  • CUDA »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    • 
+          
  • CUDA »
    •  
       
     
     
   
 
\ No newline at end of file
diff --git a/doc/source/cuda/cuda/index.html b/doc/source/cuda/cuda/index.html
index a83fda67f9..54a9dcce14 100644
--- a/doc/source/cuda/cuda/index.html
+++ b/doc/source/cuda/cuda/index.html
@@ -6,7 +6,7 @@
   
     
     
-    CUDA — PaddlePaddle  documentation
+    CUDA — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
       
 
@@ -103,14 +103,11 @@
 
 
         
@@ -132,13 +129,13 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
     
   
 
\ No newline at end of file
diff --git a/doc/source/cuda/matrix/index.html b/doc/source/cuda/matrix/index.html
index 734cab01a0..bfb85a57bf 100644
--- a/doc/source/cuda/matrix/index.html
+++ b/doc/source/cuda/matrix/index.html
@@ -6,7 +6,7 @@
   
     
     
-    Matrix — PaddlePaddle  documentation
+    Matrix — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
       
 
@@ -109,14 +109,11 @@
 
 
         
@@ -138,13 +135,13 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
     
   
 
\ No newline at end of file
diff --git a/doc/source/cuda/matrix/matrix.html b/doc/source/cuda/matrix/matrix.html
index 6f402084e4..e9c5d2b6c3 100644
--- a/doc/source/cuda/matrix/matrix.html
+++ b/doc/source/cuda/matrix/matrix.html
@@ -6,7 +6,7 @@
   
     
     
-    Matrix — PaddlePaddle  documentation
+    Matrix — PaddlePaddle  documentation
     
     
     
@@ -45,9 +45,9 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    • 
-          
  • Matrix »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    • 
+          
  • Matrix »
    •  
       
       
 
@@ -67,7 +67,7 @@
 
    
 
    
 void hl_matrix_add(real *A_d, real *B_d, real *C_d, int dimM, int dimN, real alpha, real beta)
    
-
    Matrix addition: C_d[i] = alpha * A_d[i] + beta * B_d[i]. 
    
+
    Matrix addition: C_d[i] = alpha * A_d[i] + beta * B_d[i]. 
    
 
    
 
    Parameters
    
 
      
@@ -94,7 +94,7 @@
 
      
 
      
 void hl_matrix_softmax(real *A_d, real *C_d, int dimM, int dimN)
      
-
      Matrix Softmax. 
      
+
      Matrix Softmax. 
      
 
      
 
      Parameters
      
 
        
@@ -115,7 +115,7 @@
 
        
 
        
 void hl_matrix_softmax_derivative(real *grad_d, real *output_d, real *sftmaxSum_d, int dimM, int dimN)
        
-
        Matrix softmax derivative. 
        
+
        Matrix softmax derivative. 
        
 
        
 
        Parameters
        
 
          
@@ -159,7 +159,7 @@
 
          
 
          
 void hl_matrix_classification_error(real *A_d, int *B_d, real *C_d, int dimM, int dimN)
          
-
          Matrix classification error. 
          
+
          Matrix classification error. 
          
 
          
 
          Parameters
          
 
            
@@ -182,7 +182,7 @@
 
            
 
            
 void hl_matrix_cross_entropy(real *A_d, real *C_d, int *label_d, int dimM, int dimN)
            
-
            Matrix cross entropy. 
            
+
            Matrix cross entropy. 
            
 
            
 
            Parameters
            
 
              
@@ -205,7 +205,7 @@
 
              
 
              
 void hl_matrix_cross_entropy_bp(real *grad_d, real *output_d, int *label_d, int dimM, int dimN)
              
-
              Matrix cross entropy back propagation. 
              
+
              Matrix cross entropy back propagation. 
              
 
              
 
              Parameters
              
 
                
@@ -228,7 +228,7 @@
 
                
 
                
 void hl_matrix_zero_mem(real *data, int num)
                
-
                Matrix zero memory. 
                
+
                Matrix zero memory. 
                
 
                
 
                Parameters
                
 
                  
@@ -383,18 +383,1213 @@
 
 
                  
 
                  hl_matrix_base.h
                  
+
                  
+
                  Defines
                  
+
                  
+
                  
+HL_MATRIX_BASE_CUH_
                  
+
                  
+
+
                  
+
                  
+INLINE
                  
+
                  CPP inline function 
                  
+
                  
+
+
                  
+
                  
+DEVICE_FMAX
                  
+
                  
+
+
                  
+
                  
+DEVICE_FMIN
                  
+
                  
+
+
                  
+
                  
+
                  
+class BaseOp
                  
+
                  
+
                  Public Functions
                  
+
                  
+
                  
+BaseOp()
                  
+
                  
+
+
                  
+
                  
+BaseOp(const real s1)
                  
+
                  
+
+
                  
+
                  
+BaseOp(const real s1, const real s2)
                  
+
                  
+
+
                  
+
                  
+INLINE vecType BaseOp::vecOp(const vecType a) const
                  
+
                  
+
+
                  
+
                  
+INLINE vecType BaseOp::vecOp(const vecType a, const vecType b) const
                  
+
                  
+
+
                  
+
                  
+
                  Public Static Attributes
                  
+
                  
+
                  
+const bool sse
                  
+
                  
+
+
                  
+
                  
+
+
                  
+
                  
+namespace aggregate
                  
+
                  
+
                  
+class sum
                  
+
                  Inherits from aggregate::SSESum
                  
+
                  
+
                  Public Functions
                  
+
                  
+
                  
+INLINE real aggregate::sum::init()
                  
+
                  
+
+
                  
+
                  
+INLINE real aggregate::sum::operator()(const real a, const real b) const
                  
+
                  
+
+
                  
+
                  
+
+
                  
+
                  
+class max
                  
+
                  Inherits from aggregate::SSEMax
                  
+
                  
+
                  Public Functions
                  
+
                  
+
                  
+INLINE real aggregate::max::init()
                  
+
                  
+
+
                  
+
                  
+INLINE real aggregate::max::operator()(const real a, const real b) const
                  
+
                  
+
+
                  
+
                  
+
+
                  
+
                  
+class min
                  
+
                  Inherits from aggregate::SSEMin
                  
+
                  
+
                  Public Functions
                  
+
                  
+
                  
+INLINE real aggregate::min::init()
                  
+
                  
+
+
                  
+
                  
+INLINE real aggregate::min::operator()(const real a, const real b) const
                  
+
                  
+
+
                  
+
                  
+
+
                  
+
+
                  
+
                  
+namespace base
                  
+
                  
+
                  
+namespace binary
                  
+
                  
+
                  
+class add
                  
+
                  Inherits from base::binary::SSEAdd
                  
+
                  
+
                  Public Functions
                  
+
                  
+
                  
+INLINE real base::binary::add::operator()(const real a, const real b) const
                  
+
                  
+
+
                  
+
                  
+
+
                  
+
                  
+class add2
                  
+
                  Inherits from base::binary::SSEAdd2
                  
+
                  
+
                  Public Functions
                  
+
                  
+
                  
+add2(const real s1, const real s2)
                  
+
                  
+
+
                  
+
                  
+INLINE real base::binary::add2::operator()(const real a, const real b) const
                  
+
                  
+
+
                  
+
                  
+
                  Private Members
                  
+
                  
+
                  
+const real p1
                  
+
                  
+
+
                  
+
                  
+const real p2
                  
+
                  
+
+
                  
+
                  
+
+
                  
+
                  
+class sub
                  
+
                  Inherits from base::binary::SSESub
                  
+
                  
+
                  Public Functions
                  
+
                  
+
                  
+INLINE real base::binary::sub::operator()(const real a, const real b) const
                  
+
                  
+
+
                  
+
                  
+
+
                  
+
                  
+class mul
                  
+
                  Inherits from base::binary::SSEMul
                  
+
                  
+
                  Public Functions
                  
+
                  
+
                  
+INLINE real base::binary::mul::operator()(const real a, const real b) const
                  
+
                  
+
+
                  
+
                  
+
+
                  
+
                  
+class div
                  
+
                  Inherits from base::binary::SSEDiv
                  
+
                  
+
                  Public Functions
                  
+
                  
+
                  
+INLINE real base::binary::div::operator()(const real a, const real b) const
                  
+
                  
+
+
                  
+
                  
+
+
                  
+
                  
+class squaredDiff
                  
+
                  Inherits from base::binary::SSESquaredDiff
                  
+
                  
+
                  Public Functions
                  
+
                  
+
                  
+INLINE real base::binary::squaredDiff::operator()(const real a, const real b) const
                  
+
                  
+
+
                  
+
                  
+
+
                  
+
                  
+class first
                  
+
                  Inherits from base::binary::SSEFirst
                  
+
                  
+
                  Public Functions
                  
+
                  
+
                  
+INLINE real base::binary::first::operator()(const real a, const real b) const
                  
+
                  
+
+
                  
+
                  
+
+
                  
+
                  
+class second
                  
+
                  Inherits from base::binary::SSESecond
                  
+
                  
+
                  Public Functions
                  
+
                  
+
                  
+INLINE real base::binary::second::operator()(const real a, const real b) const
                  
+
                  
+
+
                  
+
                  
+
+
                  
+
                  
+class classificationError
                  
+
                  Inherits from base::binary::SSEClassificationError
                  
+
                  
+
                  Public Functions
                  
+
                  
+
                  
+classificationError(const real s)
                  
+
                  
+
+
                  
+
                  
+INLINE real base::binary::classificationError::operator()(const real a, const real b) const
                  
+
                  
+
+
                  
+
                  
+
                  Private Members
                  
+
                  
+
                  
+const real p
                  
+
                  
+
+
                  
+
                  
+
+
                  
+
+
                  
+
                  
+namespace unary
                  
+
                  
+
                  
+class identity
                  
+
                  Inherits from base::unary::SSEIdentity
                  
+
                  
+
                  Public Functions
                  
+
                  
+
                  
+INLINE real base::unary::identity::operator()(const real a) const
                  
+
                  
+
+
                  
+
                  
+
+
                  
+
+
                  
+
+
                  
+
                  
+
                  hl_matrix_apply.cuh
                  
+
                  
+
                  Defines
                  
+
                  
+
                  
+HL_MATRIX_APPLY_H_
                  
+
                  
+
+
                  
+
                  
+
                  Functions
                  
+
                  
+
                  
+template <class T, class Op>
                  
+
                  
+void hl_cpu_apply_unary_op(Op op, T *A_h, int dimM, int dimN, int lda)
                  
+
                  CPU element wise unary operator. 
                  
+
                  element wise op(a) for 0 <= i < dimM & for 0 <= j < dimN.
                  
+
                  
+CPU element wise unary operator. 
                  
+
                  Parameters
                  
+
                    
+
                  • op - 
                    unary op. see namespace unary 
                    
+
                    • 
+
                  • A_h - 
                    matrix. 
                    
+
                    • 
+
                  • dimM - 
                    matrix height. 
                    
+
                    • 
+
                  • dimN - 
                    matrix width. 
                    
+
                    • 
+
                  • lda - 
                    leading dimension of A.
                    
+
                    • 
+
                  
+
                  
+
                  
+
                  
+
                  
+
+
                  
+
                  
+template <class T, class Op, bool BAsRowVector, bool BAsColVector>
                  
+
                  
+void hl_cpu_apply_binary_op(Op op, T *A_h, T *B_h, int dimM, int dimN, int lda, int ldb)
                  
+
                  CPU element wise binary operator. 
                  
+
                  element wise op(a, b) for 0 <= i < dimM & for 0 <= j < dimN.
                  
+
                  if (BAsRowVector == 0 && BAsColVector == 0) op(A[i * lda + j], B[i * ldb + j])
                  
+
                  if (BAsRowVector == 1 && BAsColVector == 0) op(A[i * lda + j], B[j])
                  
+
                  if (BAsRowVector == 0 && BAsColVector == 1) op(A[i * lda + j], B[i * ldb])
                  
+
                  if (BAsRowVector == 1 && BAsColVector == 1) op(A[i * lda + j], B[0])
                  
+
                  
+CPU element wise binary operator. 
                  
+
                  Parameters
                  
+
                    
+
                  • op - 
                    binary op. see namespace binary. 
                    
+
                    • 
+
                  • A_h - 
                    matrix. 
                    
+
                    • 
+
                  • B_h - 
                    matrix. 
                    
+
                    • 
+
                  • dimM - 
                    matrix height. 
                    
+
                    • 
+
                  • dimN - 
                    matrix width. 
                    
+
                    • 
+
                  • lda - 
                    leading dimension of A. 
                    
+
                    • 
+
                  • ldb - 
                    leading dimension of B.
                    
+
                    • 
+
                  
+
                  
+
                  
+
                  
+
                  
+
+
                  
+
                  
+template <class T, class Op, bool CAsRowVector, bool CAsColVector>
                  
+
                  
+void hl_cpu_apply_ternary_op(Op op, T *A_h, T *B_h, T *C_h, int dimM, int dimN, int lda, int ldb, int ldc)
                  
+
                  CPU element wise ternary operator. 
                  
+
                  element wise op(a, b, c) for 0 <= i < dimM & for 0 <= j < dimN.
                  
+
                  if (CAsRowVector == 0 && CAsColVector == 0) op(A[i*lda + j], B[i*ldb + j], C[i*ldc + j])
                  
+
                  if (CAsRowVector == 1 && CAsColVector == 0) op(A[i*lda + j], B[i*ldb + j], C[j])
                  
+
                  if (CAsRowVector == 0 && CAsColVector == 1) op(A[i*lda + j], B[i*ldb + j], C[i*ldc])
                  
+
                  if (CAsRowVector == 1 && CAsColVector == 1) op(A[i*lda + j], B[i*ldb + j], C[0])
                  
+
                  
+CPU element wise ternary operator. 
                  
+
                  Parameters
                  
+
                    
+
                  • op - 
                    ternary op. see namespace ternary. 
                    
+
                    • 
+
                  • A_h - 
                    matrix. 
                    
+
                    • 
+
                  • B_h - 
                    matrix. 
                    
+
                    • 
+
                  • C_h - 
                    matrix. 
                    
+
                    • 
+
                  • dimM - 
                    matrix height. 
                    
+
                    • 
+
                  • dimN - 
                    matrix width. 
                    
+
                    • 
+
                  • lda - 
                    leading dimension of A. 
                    
+
                    • 
+
                  • ldb - 
                    leading dimension of B. 
                    
+
                    • 
+
                  • ldc - 
                    leading dimension of C.
                    
+
                    • 
+
                  
+
                  
+
                  
+
                  
+
                  
+
+
                  
+
                  
+template <class T, class Op>
                  
+
                  
+void hl_cpu_apply_quaternary_op(Op op, T *A_h, T *B_h, T *C_h, T *D_h, int dimM, int dimN, int lda, int ldb, int ldc, int ldd)
                  
+
                  CPU element wise quaternary operator. element wise op(a, b, c, d) for 0 <= i < dimM & for 0 <= j < dimN. 
                  
+
                  
+CPU element wise quaternary operator. element wise op(a, b, c, d) for 0 <= i < dimM & for 0 <= j < dimN. 
                  
+
                  Parameters
                  
+
                    
+
                  • op - 
                    quaternary op. see namespace ternary. 
                    
+
                    • 
+
                  • A_h - 
                    matrix. 
                    
+
                    • 
+
                  • B_h - 
                    matrix. 
                    
+
                    • 
+
                  • C_h - 
                    matrix. 
                    
+
                    • 
+
                  • D_h - 
                    matrix. 
                    
+
                    • 
+
                  • dimM - 
                    matrix height. 
                    
+
                    • 
+
                  • dimN - 
                    matrix width. 
                    
+
                    • 
+
                  • lda - 
                    leading dimension of A. 
                    
+
                    • 
+
                  • ldb - 
                    leading dimension of B. 
                    
+
                    • 
+
                  • ldc - 
                    leading dimension of C. 
                    
+
                    • 
+
                  • ldd - 
                    leading dimension of D.
                    
+
                    • 
+
                  
+
                  
+
                  
+
                  
+
                  
+
+
                  
+
                  
+template <class T, class Op>
                  
+
                  
+void hl_gpu_apply_unary_op(Op op, T *A_d, int dimM, int dimN, int lda)
                  
+
                  GPU element wise unary operator. element wise op(a) for 0 <= i < dimM & for 0 <= j < dimN. 
                  
+
                  
+
                  Parameters
                  
+
                    
+
                  • op - 
                    unary op. see namespace unary. 
                    
+
                    • 
+
                  • A_d - 
                    matrix. 
                    
+
                    • 
+
                  • dimM - 
                    matrix height. 
                    
+
                    • 
+
                  • dimN - 
                    matrix width. 
                    
+
                    • 
+
                  • lda - 
                    leading dimension of A. 
                    
+
                    • 
+
                  
+
                  
+
                  
+
                  
+
                  
+
+
                  
+
                  
+template <class T, class Op, bool BAsRowVector, bool BAsColVector>
                  
+
                  
+void hl_gpu_apply_binary_op(Op op, T *A_d, T *B_d, int dimM, int dimN, int lda, int ldb)
                  
+
                  GPU element wise binary operator. 
                  
+
                  element wise op(a, b) for 0 <= i < dimM & for 0 <= j < dimN
                  
+
                  if (BAsRowVector == 0 && BAsColVector == 0) op(A[i * lda + j], B[i * ldb + j])
                  
+
                  if (BAsRowVector == 1 && BAsColVector == 0) op(A[i * lda + j], B[j])
                  
+
                  if (BAsRowVector == 0 && BAsColVector == 1) op(A[i * lda + j], B[i * ldb])
                  
+
                  if (BAsRowVector == 1 && BAsColVector == 1) op(A[i * lda + j], B[0])
                  
+
                  
+
                  Parameters
                  
+
                    
+
                  • op - 
                    binary op. see namespace binary. 
                    
+
                    • 
+
                  • A_d - 
                    matrix. 
                    
+
                    • 
+
                  • B_d - 
                    matrix. 
                    
+
                    • 
+
                  • dimM - 
                    matrix height. 
                    
+
                    • 
+
                  • dimN - 
                    matrix width. 
                    
+
                    • 
+
                  • lda - 
                    leading dimension of A. 
                    
+
                    • 
+
                  • ldb - 
                    leading dimension of B. 
                    
+
                    • 
+
                  
+
                  
+
                  
+
                  
+
                  
+
+
                  
+
                  
+template <class T, class Op, bool CAsRowVector, bool CAsColVector>
                  
+
                  
+void hl_gpu_apply_ternary_op(Op op, T *A_d, T *B_d, T *C_d, int dimM, int dimN, int lda, int ldb, int ldc)
                  
+
                  GPU element wise ternary operator. 
                  
+
                  element wise op(a, b, c) for 0 <= i < dimM & for 0 <= j < dimN.
                  
+
                  if (CAsRowVector == 0 && CAsColVector == 0) op(A[i*lda + j], B[i*ldb + j], C[i*ldc + j])
                  
+
                  if (CAsRowVector == 1 && CAsColVector == 0) op(A[i*lda + j], B[i*ldb + j], C[j])
                  
+
                  if (CAsRowVector == 0 && CAsColVector == 1) op(A[i*lda + j], B[i*ldb + j], C[i*ldc])
                  
+
                  if (CAsRowVector == 1 && CAsColVector == 1) op(A[i*lda + j], B[i*ldb + j], C[0])
                  
+
                  
+
                  Parameters
                  
+
                    
+
                  • op - 
                    ternary op. see namespace ternary. 
                    
+
                    • 
+
                  • A_d - 
                    matrix. 
                    
+
                    • 
+
                  • B_d - 
                    matrix. 
                    
+
                    • 
+
                  • C_d - 
                    matrix. 
                    
+
                    • 
+
                  • dimM - 
                    matrix height. 
                    
+
                    • 
+
                  • dimN - 
                    matrix width. 
                    
+
                    • 
+
                  • lda - 
                    leading dimension of A. 
                    
+
                    • 
+
                  • ldb - 
                    leading dimension of B. 
                    
+
                    • 
+
                  • ldc - 
                    leading dimension of C. 
                    
+
                    • 
+
                  
+
                  
+
                  
+
                  
+
                  
+
+
                  
+
                  
+template <class T, class Op>
                  
+
                  
+void hl_gpu_apply_quaternary_op(Op op, T *A_d, T *B_d, T *C_d, T *D_d, int dimM, int dimN, int lda, int ldb, int ldc, int ldd)
                  
+
                  GPU element wise quaternary operator. element wise op(a, b, c, d) for 0 <= i < dimM & for 0 <= j < dimN. 
                  
+
                  
+
                  Parameters
                  
+
                    
+
                  • op - 
                    quaternary op. see namespace ternary. 
                    
+
                    • 
+
                  • A_d - 
                    matrix. 
                    
+
                    • 
+
                  • B_d - 
                    matrix. 
                    
+
                    • 
+
                  • C_d - 
                    matrix. 
                    
+
                    • 
+
                  • D_d - 
                    matrix. 
                    
+
                    • 
+
                  • dimM - 
                    matrix height. 
                    
+
                    • 
+
                  • dimN - 
                    matrix width. 
                    
+
                    • 
+
                  • lda - 
                    leading dimension of A. 
                    
+
                    • 
+
                  • ldb - 
                    leading dimension of B. 
                    
+
                    • 
+
                  • ldc - 
                    leading dimension of C. 
                    
+
                    • 
+
                  • ldd - 
                    leading dimension of D. 
                    
+
                    • 
+
                  
+
                  
+
                  
+
                  
+
                  
+
+
                  
+
                  
+template <class Agg, class Op, class Saver>
                  
+
                  
+void hl_cpu_matrix_row_op(Agg agg, Op op, Saver sv, int dimM, int dimN, real *dst, int ld, real *A, int lda)
                  
+
                  CPU matrix row operator. 
                  
+
                  
+
+
                  
+
                  
+template <class Saver, class Agg, class Op>
                  
+
                  
+void hl_cpu_matrix_row_op(Agg agg, Op op, int dimM, int dimN, real *dst, int ld, real *A, int lda, real *B, int ldb)
                  
+
                  CPU matrix row operator. 
                  
+
                  
+
                  Parameters
                  
+
                    
+
                  • agg - 
                    aggregate operator expression. 
                    
+
                    • 
+
                  • op - 
                    operator expression. 
                    
+
                    • 
+
                  • dimM - 
                    matrix height. 
                    
+
                    • 
+
                  • dimN - 
                    matrix width. 
                    
+
                    • 
+
                  • dst - 
                    destination matrix. 
                    
+
                    • 
+
                  • ld - 
                    leading dimension of dst matrix. 
                    
+
                    • 
+
                  • *A - 
                    matrix A. 
                    
+
                    • 
+
                  • lda - 
                    leading dimension of matrix A. 
                    
+
                    • 
+
                  • *B - 
                    matrix B. 
                    
+
                    • 
+
                  • ldb - 
                    leading dimension of matrix B. 
                    
+
                    • 
+
                  
+
                  
+
                  
+
                  
+
                  
+
+
                  
+
                  
+template <class Agg, class Op, class Saver>
                  
+
                  
+void hl_cpu_matrix_column_op(Agg agg, Op op, Saver sv, int dimM, int dimN, real *dst, real *A, int lda)
                  
+
                  CPU matrix column operator. 
                  
+
                  
+
                  Parameters
                  
+
                    
+
                  • agg - 
                    aggregate operator expression. 
                    
+
                    • 
+
                  • op - 
                    operator expression. 
                    
+
                    • 
+
                  • sv - 
                    assignment operator expression. 
                    
+
                    • 
+
                  • dimM - 
                    matrix height. 
                    
+
                    • 
+
                  • dimN - 
                    matrix width. 
                    
+
                    • 
+
                  • dst - 
                    destination matrix. 
                    
+
                    • 
+
                  • *A - 
                    matrix A. 
                    
+
                    • 
+
                  • lda - 
                    leading dimension of matrix A. 
                    
+
                    • 
+
                  
+
                  
+
                  
+
                  
+
                  
+
+
                  
+
                  
+template <class Agg, class Op, class Saver>
                  
+
                  
+void hl_cpu_matrix_column_op(Agg agg, Op op, Saver sv, int dimM, int dimN, real *dst, real *A, int lda, real *B, int ldb)
                  
+
                  CPU matrix column operator. 
                  
+
                  
+
                  Parameters
                  
+
                    
+
                  • agg - 
                    aggregate operator expression. 
                    
+
                    • 
+
                  • op - 
                    operator expression. 
                    
+
                    • 
+
                  • sv - 
                    assignment operator expression. 
                    
+
                    • 
+
                  • dimM - 
                    matrix height. 
                    
+
                    • 
+
                  • dimN - 
                    matrix width. 
                    
+
                    • 
+
                  • dst - 
                    destination matrix. 
                    
+
                    • 
+
                  • *A - 
                    matrix A. 
                    
+
                    • 
+
                  • lda - 
                    leading dimension of matrix A. 
                    
+
                    • 
+
                  • *B - 
                    matrix B. 
                    
+
                    • 
+
                  • ldb - 
                    leading dimension of matrix B. 
                    
+
                    • 
+
                  
+
                  
+
                  
+
                  
+
                  
+
+
                  
+
                  
+template <class Agg, class Op, class Saver>
                  
+
                  
+void hl_gpu_matrix_row_op(Agg agg, Op op, Saver sv, int dimM, int dimN, real *dst, int ld, real *A, int lda)
                  
+
                  GPU matrix row operator. 
                  
+
                  
+
                  Parameters
                  
+
                    
+
                  • agg - 
                    aggregate operator expression. 
                    
+
                    • 
+
                  • op - 
                    operator expression. 
                    
+
                    • 
+
                  • sv - 
                    assignment operator expression. 
                    
+
                    • 
+
                  • dimM - 
                    matrix height. 
                    
+
                    • 
+
                  • dimN - 
                    matrix width. 
                    
+
                    • 
+
                  • dst - 
                    destination matrix. 
                    
+
                    • 
+
                  • ld - 
                    leading dimension of dst. 
                    
+
                    • 
+
                  • *A - 
                    matrix A. 
                    
+
                    • 
+
                  • lda - 
                    leading dimension of matrix A. 
                    
+
                    • 
+
                  
+
                  
+
                  
+
                  
+
                  
+
+
                  
+
                  
+template <class Saver, class Agg, class Op>
                  
+
                  
+void hl_gpu_matrix_row_op(Agg agg, Op op, int dimM, int dimN, real *dst, int ld, real *A, int lda, real *B, int ldb)
                  
+
                  GPU matrix row operator. 
                  
+
                  
+
                  Parameters
                  
+
                    
+
                  • agg - 
                    aggregate operator expression. 
                    
+
                    • 
+
                  • op - 
                    operator expression. 
                    
+
                    • 
+
                  • dimM - 
                    matrix height. 
                    
+
                    • 
+
                  • dimN - 
                    matrix width. 
                    
+
                    • 
+
                  • dst - 
                    destination matrix. 
                    
+
                    • 
+
                  • ld - 
                    leading dimension of dst matrix. 
                    
+
                    • 
+
                  • *A - 
                    matrix A. 
                    
+
                    • 
+
                  • lda - 
                    leading dimension of matrix A. 
                    
+
                    • 
+
                  • *B - 
                    matrix B. 
                    
+
                    • 
+
                  • ldb - 
                    leading dimension of matrix B. 
                    
+
                    • 
+
                  
+
                  
+
                  
+
                  
+
                  
+
+
                  
+
                  
+template <class Agg, class Op, class Saver>
                  
+
                  
+void hl_gpu_matrix_column_op(Agg agg, Op op, Saver sv, int dimM, int dimN, real *dst, real *A, int lda)
                  
+
                  GPU matrix column operator. 
                  
+
                  
+
                  Parameters
                  
+
                    
+
                  • agg - 
                    aggregate operator expression. 
                    
+
                    • 
+
                  • op - 
                    operator expression. 
                    
+
                    • 
+
                  • sv - 
                    assignment operator expression. 
                    
+
                    • 
+
                  • dimM - 
                    matrix height. 
                    
+
                    • 
+
                  • dimN - 
                    matrix width. 
                    
+
                    • 
+
                  • dst - 
                    destination matrix. 
                    
+
                    • 
+
                  • *A - 
                    matrix A. 
                    
+
                    • 
+
                  • lda - 
                    leading dimension of matrix A. 
                    
+
                    • 
+
                  
+
                  
+
                  
+
                  
+
                  
+
+
                  
+
                  
+template <class Agg, class Op, class Saver>
                  
+
                  
+void hl_gpu_matrix_column_op(Agg agg, Op op, Saver sv, int dimM, int dimN, real *dst, real *A, int lda, real *B, int ldb)
                  
+
                  GPU matrix column operator. 
                  
+
                  
+
                  Parameters
                  
+
                    
+
                  • agg - 
                    aggregate operator expression. 
                    
+
                    • 
+
                  • op - 
                    operator expression. 
                    
+
                    • 
+
                  • sv - 
                    assignment operator expression. 
                    
+
                    • 
+
                  • dimM - 
                    matrix height. 
                    
+
                    • 
+
                  • dimN - 
                    matrix width. 
                    
+
                    • 
+
                  • dst - 
                    destination matrix. 
                    
+
                    • 
+
                  • *A - 
                    matrix A. 
                    
+
                    • 
+
                  • lda - 
                    leading dimension of matrix A. 
                    
+
                    • 
+
                  • *B - 
                    matrix B. 
                    
+
                    • 
+
                  • ldb - 
                    leading dimension of matrix B. 
                    
+
                    • 
+
                  
+
                  
+
                  
+
                  
+
                  
+
 
                  
-
                  
-
                  hl_matrix_apply.cuh
                  
 
                  
 
                  
 
                  hl_matrix_ops.cuh
                  
+
                  
+
                  Defines
                  
+
                  
+
                  
+HL_MATRIX_OPS_CUH_
                  
+
                  
+
+
                  
+
                  
+HL_DEVICE
                  
+
                  
+
+
                  
+
                  
+ONE_PARAMETER(name)
                  
+
                  parameter macro. 
                  
+
                  
+
+
                  
+
                  
+TWO_PARAMETER(name)
                  
+
                  
+
+
                  
+
                  
+THREE_PARAMETER(name)
                  
+
                  
+
+
                  
+
                  
+FOUR_PARAMETER(name)
                  
+
                  
+
+
                  
+
                  
+DEFINE_MATRIX_UNARY_OP(name, op)
                  
+
                  unary operator macro. 
                  
+
                  
+
                  Note
                  
+
                  op format: op supports multiple expressions that are separated by a comma. e.g. a, b
                  
+
                  See
                  
+
                  hl_gpu_apply_unary_op 
                  
+
                  hl_cpu_apply_unary_op 
                  
+
                  
+
                  Parameters
                  
+
                    
+
                  • name - 
                    operator name. 
                    
+
                    • 
+
                  • op - 
                    operator expression.
                    
+
                    • 
+
                  
+
                  
+
                  
+
                  
+
                  
+
+
                  
+
                  
+DEFINE_MATRIX_UNARY_PARAMETER_OP(name, PARA_MACRO, op)
                  
+
                  unary operator macro. 
                  
+
                  
+
                  Note
                  
+
                  op format: op supports multiple expressions that are separated by a comma. e.g. a, b
                  
+
                  See
                  
+
                  hl_gpu_apply_unary_op 
                  
+
                  hl_cpu_apply_unary_op 
                  
+
                  
+
                  Parameters
                  
+
                    
+
                  • name - 
                    operator name. 
                    
+
                    • 
+
                  • PARA_MACRO - 
                    parameter macro. 
                    
+
                    • 
+
                  • op - 
                    operator expression.
                    
+
                    • 
+
                  
+
                  
+
                  
+
                  
+
                  
+
+
                  
+
                  
+DEFINE_MATRIX_BINARY_OP(name, op)
                  
+
                  binary operator macro. 
                  
+
                  
+
                  Note
                  
+
                  op format: op supports multiple expressions that are separated by a comma. e.g. a, b
                  
+
                  See
                  
+
                  hl_gpu_apply_unary_op 
                  
+
                  hl_cpu_apply_unary_op 
                  
+
                  
+
                  Parameters
                  
+
                    
+
                  • name - 
                    operator name. 
                    
+
                    • 
+
                  • op - 
                    operator expression.
                    
+
                    • 
+
                  
+
                  
+
                  
+
                  
+
                  
+
+
                  
+
                  
+DEFINE_MATRIX_BINARY_PARAMETER_OP(name, PARA_MACRO, op)
                  
+
                  binary operator macro. 
                  
+
                  
+
                  Note
                  
+
                  op format: op supports multiple expressions that are separated by a comma. e.g. a, b
                  
+
                  See
                  
+
                  hl_gpu_apply_binary_op 
                  
+
                  hl_cpu_apply_binary_op 
                  
+
                  
+
                  Parameters
                  
+
                    
+
                  • name - 
                    operator name. 
                    
+
                    • 
+
                  • PARA_MACRO - 
                    parameter macro. 
                    
+
                    • 
+
                  • op - 
                    operator expression.
                    
+
                    • 
+
                  
+
                  
+
                  
+
                  
+
                  
+
+
                  
+
                  
+DEFINE_MATRIX_TERNARY_OP(name, op)
                  
+
                  ternary operator macro. 
                  
+
                  
+
                  Note
                  
+
                  op format: op supports multiple expressions that are separated by a comma. e.g. a, b, c
                  
+
                  See
                  
+
                  hl_gpu_apply_ternary_op 
                  
+
                  hl_cpu_apply_ternary_op 
                  
+
                  
+
                  Parameters
                  
+
                    
+
                  • name - 
                    operator name. 
                    
+
                    • 
+
                  • op - 
                    operator expression.
                    
+
                    • 
+
                  
+
                  
+
                  
+
                  
+
                  
+
+
                  
+
                  
+DEFINE_MATRIX_TERNARY_PARAMETER_OP(name, PARA_MACRO, op)
                  
+
                  ternary operator macro. 
                  
+
                  
+
                  Note
                  
+
                  op format: op supports multiple expressions that are separated by a comma. e.g. a, b, c
                  
+
                  See
                  
+
                  hl_gpu_apply_ternary_op 
                  
+
                  hl_cpu_apply_ternary_op 
                  
+
                  
+
                  Parameters
                  
+
                    
+
                  • name - 
                    operator name. 
                    
+
                    • 
+
                  • PARA_MACRO - 
                    parameter macro. 
                    
+
                    • 
+
                  • op - 
                    operator expression.
                    
+
                    • 
+
                  
+
                  
+
                  
+
                  
+
                  
+
+
                  
+
                  
+DEFINE_MATRIX_QUATERNARY_OP(name, op)
                  
+
                  quaternary operator macro. 
                  
+
                  
+
                  Note
                  
+
                  op format: op supports multiple expressions that are separated by a comma. e.g. a, b, c, d
                  
+
                  See
                  
+
                  hl_gpu_apply_quaternary_op 
                  
+
                  hl_cpu_apply_quaternary_op 
                  
+
                  
+
                  Parameters
                  
+
                    
+
                  • name - 
                    operator name. 
                    
+
                    • 
+
                  • op - 
                    operator expression.
                    
+
                    • 
+
                  
+
                  
+
                  
+
                  
+
                  
+
+
                  
+
                  
+DEFINE_MATRIX_QUATERNARY_PARAMETER_OP(name, PARA_MACRO, op)
                  
+
                  quaternary operator macro. 
                  
+
                  
+
                  Note
                  
+
                  op format: op supports multiple expressions that are separated by a comma. e.g. a, b, c, d
                  
+
                  See
                  
+
                  hl_gpu_apply_quaternary_op 
                  
+
                  hl_cpu_apply_quaternary_op 
                  
+
                  
+
                  Parameters
                  
+
                    
+
                  • name - 
                    operator name. 
                    
+
                    • 
+
                  • PARA_MACRO - 
                    parameter macro. 
                    
+
                    • 
+
                  • op - 
                    operator expression.
                    
+
                    • 
+
                  
+
                  
+
                  
+
                  
+
                  
+
+
                  
 
                  
 
                  
 
                  hl_matrix_type.cuh
                  
+
                  
+
                  Defines
                  
+
                  
+
                  
+HL_MATRIX_TYPE_CUH_
                  
+
                  
+
+
                  
+
                  
+
                  Typedefs
                  
+
                  
+
                  
+typedef __m128 vecType
                  
+
                  
+
+
                  
 
                  
 
                  
 
                  hl_sse_matrix_kernel.cuh
                  
+
                  
+
                  Defines
                  
+
                  
+
                  
+HL_SSE_MATRIX_KERNEL_CUH_
                  
+
                  
+
+
                  
+
                  
+VECTOR_SIZE
                  
+
                  
+
+
                  
+
                  
+VECTOR_LEN
                  
+
                  
+
+
                  
+
                  
+VECTOR_SET
                  
+
                  
+
+
                  
+
                  
+
                  Functions
                  
+
                  
+
                  
+bool hl_check_align(size_t size)
                  
+
                  
+
+
                  
+
                  
+bool hl_check_align(void *ptr)
                  
+
                  
+
+
                  
+
                  
+template <class Agg>
                  
+
                  
+real hl_agg_op(Agg agg, vecType mm)
                  
+
                  
+
+
                  
+
                  
+template <class Agg, class Op, class Saver>
                  
+
                  
+void hl_sse_matrix_row_op(Agg agg, Op op, Saver sv, int dimM, int dimN, real *dst, int ld, real *A, int lda)
                  
+
                  
+
+
                  
+
                  
+template <class Agg, class Op, class Saver>
                  
+
                  
+void hl_sse_matrix_row_op(Agg agg, Op op, Saver sv, int dimM, int dimN, real *dst, int ld, real *A, int lda, real *B, int ldb)
                  
+
                  
+
+
                  
+
                  
+template <class Agg, class Op, class Saver>
                  
+
                  
+void hl_matrix_column_op(Agg agg, Op op, Saver sv, int dimM, int dimN, real *dst, real *A, int lda)
                  
+
                  
+
+
                  
+
                  
+template <class Agg, class Op, class Saver>
                  
+
                  
+void hl_matrix_column_op(Agg agg, Op op, Saver sv, int dimM, int dimN, real *dst, real *A, int lda, real *B, int ldb)
                  
+
                  
+
+
                  
+
                  
+template <int MaxRow, class Agg, class Op, class Saver>
                  
+
                  
+void hl_sse_column_op_with_rem(Agg agg, Op op, Saver sv, int dimM, int dimN, real *dst, real *A, int lda)
                  
+
                  
+
+
                  
+
                  
+template <int Step, class Agg, class Op, class Saver>
                  
+
                  
+void hl_sse_matrix_column_op(Agg agg, Op op, Saver sv, int dimM, int dimN, real *dst, real *A, int lda)
                  
+
                  
+
+
                  
+
                  
+template <class Agg, class Op, class Saver>
                  
+
                  
+void hl_sse_matrix_column_op(Agg agg, Op op, Saver sv, int dimM, int dimN, real *dst, real *A, int lda)
                  
+
                  
+
+
                  
+
                  
+template <int MaxRow, class Agg, class Op, class Saver>
                  
+
                  
+void hl_sse_column_op_with_rem(Agg agg, Op op, Saver sv, int dimM, int dimN, real *dst, real *A, int lda, real *B, int ldb)
                  
+
                  
+
+
                  
+
                  
+template <int Step, class Agg, class Op, class Saver>
                  
+
                  
+void hl_sse_matrix_column_op(Agg agg, Op op, Saver sv, int dimM, int dimN, real *dst, real *A, int lda, real *B, int ldb)
                  
+
                  
+
+
                  
+
                  
+template <class Agg, class Op, class Saver>
                  
+
                  
+void hl_sse_matrix_column_op(Agg agg, Op op, Saver sv, int dimM, int dimN, real *dst, real *A, int lda, real *B, int ldb)
                  
+
                  
+
+
                  
 
                  
 
                  
 
                  hl_batch_transpose.h
                  
@@ -1079,6 +2274,153 @@
 
                  
 
                  
 
                  hl_sparse.ph
                  
+
                  
+
                  Defines
                  
+
                  
+
                  
+HL_SPARSE_PH_
                  
+
                  
+
+
                  
+
                  
+__sparse_get_type_return__(mat, type, field)
                  
+
                  
+
+
                  
+
                  
+__sparse_get_return__(mat, field)
                  
+
                  
+
+
                  
+
                  
+
                  Typedefs
                  
+
                  
+
                  
+typedef struct _hl_csr_matrix *hl_csr_matrix
                  
+
                  
+
+
                  
+
                  
+typedef struct _hl_csc_matrix *hl_csc_matrix
                  
+
                  
+
+
                  
+
                  
+
                  
+struct _hl_csr_matrix
                  
+
                  sparse matrix csr format. 
                  
+
                  
+
                  Parameters
                  
+
                    
+
                  • *csr_val - 
                    nonzero values of matrix. 
                    
+
                    • 
+
                  • *csr_row - 
                    row indices. 
                    
+
                    • 
+
                  • *csr_col - 
                    column indices. 
                    
+
                    • 
+
                  • nnz_s - 
                    sizeof of csr_val & csr_col. 
                    
+
                    • 
+
                  • row_s - 
                    sizeof of csr_row. 
                    
+
                    • 
+
                  • sparsity - 
                    sparsity pattern. 
                    
+
                    • 
+
                  
+
                  
+
                  
+
                  
+
                  
+
                  Public Members
                  
+
                  
+
                  
+real *csr_val
                  
+
                  
+
+
                  
+
                  
+int *csr_row
                  
+
                  
+
+
                  
+
                  
+int *csr_col
                  
+
                  
+
+
                  
+
                  
+size_t nnz_s
                  
+
                  
+
+
                  
+
                  
+int row_s
                  
+
                  
+
+
                  
+
                  
+float sparsity
                  
+
                  
+
+
                  
+
                  
+
+
                  
+
                  
+struct _hl_csc_matrix
                  
+
                  sparse matrix csc format. 
                  
+
                  
+
                  Parameters
                  
+
                    
+
                  • *csc_val - 
                    nonzero values of matrix. 
                    
+
                    • 
+
                  • *csc_row - 
                    row indices. 
                    
+
                    • 
+
                  • *csc_col - 
                    column indices. 
                    
+
                    • 
+
                  • nnz_s - 
                    sizeof of csc_val & csc_row. 
                    
+
                    • 
+
                  • col_s - 
                    sizeof of csc_col. 
                    
+
                    • 
+
                  • sparsity - 
                    sparsity pattern. 
                    
+
                    • 
+
                  
+
                  
+
                  
+
                  
+
                  
+
                  Public Members
                  
+
                  
+
                  
+real *csc_val
                  
+
                  
+
+
                  
+
                  
+int *csc_row
                  
+
                  
+
+
                  
+
                  
+int *csc_col
                  
+
                  
+
+
                  
+
                  
+size_t nnz_s
                  
+
                  
+
+
                  
+
                  
+int col_s
                  
+
                  
+
+
                  
+
                  
+float sparsity
                  
+
                  
+
+
                  
+
                  
+
 
                  
 
                  
 
                  
@@ -1159,7 +2501,7 @@
 
                    
 
                  • A_d - 
                    input matrix (M x N). 
                    
 
                    • 
-
                  • C_d - 
                    output Matrix (1 x N). 
                    
+
                  • C_d - 
                    output Matrix (1 x N). 
                    
 
                    • 
 
                  • dimM - 
                    matrix height. 
                    
 
                    • 
@@ -1459,14 +2801,11 @@
 
 
         
                  
@@ -1488,14 +2827,14 @@
         
                • 
           previous |
                  • 
-        
                • PaddlePaddle  documentation »
                  • 
-          
                • Source Code Documents »
                  • 
-          
                • Matrix »
                  •  
+        
                • PaddlePaddle  documentation »
                  • 
+          
                • Source Code Documents »
                  • 
+          
                • Matrix »
                  •  
       
                
     
     
   
 
\ No newline at end of file
diff --git a/doc/source/cuda/rnn/index.html b/doc/source/cuda/rnn/index.html
index dbae835725..9edfc6ee0f 100644
--- a/doc/source/cuda/rnn/index.html
+++ b/doc/source/cuda/rnn/index.html
@@ -6,7 +6,7 @@
   
     
     
-    RNN — PaddlePaddle  documentation
+    RNN — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
              • 
           previous |
                • 
-        
              • PaddlePaddle  documentation »
                • 
-          
              • Source Code Documents »
                •  
+        
              • PaddlePaddle  documentation »
                • 
+          
              • Source Code Documents »
                •  
       
              
       
 
@@ -95,14 +95,11 @@
 
 
         
@@ -124,13 +121,13 @@
         
            • 
           previous |
              • 
-        
            • PaddlePaddle  documentation »
              • 
-          
            • Source Code Documents »
              •  
+        
            • PaddlePaddle  documentation »
              • 
+          
            • Source Code Documents »
              •  
       
            
     
     
   
 
\ No newline at end of file
diff --git a/doc/source/cuda/rnn/rnn.html b/doc/source/cuda/rnn/rnn.html
index e52b60f8a5..8017dc848f 100644
--- a/doc/source/cuda/rnn/rnn.html
+++ b/doc/source/cuda/rnn/rnn.html
@@ -6,7 +6,7 @@
   
     
     
-    Neural Networks — PaddlePaddle  documentation
+    Neural Networks — PaddlePaddle  documentation
     
     
     
@@ -45,9 +45,9 @@
         
          • 
           previous |
            • 
-        
          • PaddlePaddle  documentation »
            • 
-          
          • Source Code Documents »
            • 
-          
          • RNN »
            •  
+        
          • PaddlePaddle  documentation »
            • 
+          
          • Source Code Documents »
            • 
+          
          • RNN »
            •  
       
          
       
 
@@ -432,7 +432,7 @@
 
 
          
 
          
-namespace hppl
          
+namespace hppl
 
          
 
          Functions
          
 
          
@@ -478,6 +478,35 @@
 
          
 
          
 
+
          
+
          Defines
          
+
          
+
          
+HL_DEVICE_FUNCTIONS_CUH_
          
+
          
+
+
          
+
          
+
          
+namespace hppl
          
+
          
+
          Functions
          
+
          
+
          
+static __inline__ __device__ double hppl::atomicAdd(double * address, double val)
          
+
          
+
+
          
+
          
+
+
          
+
          Defines
          
+
          
+
          
+HL_GPU_FUNCTIONS_CUH_
          
+
          
+
+
          
 
 
          
 
          Activation Functions
          
@@ -491,8 +520,8 @@
 
 
          
 
          
-
          
-namespace hppl
          
+
          
+namespace hppl
          
 
          
 
          
 template <class T>
          
@@ -502,13 +531,17 @@
 
          
 
          Public Types
          
 
          
-
          
-typedef T (*forward)(T)
          
+
          
+typedef >
          
+
          
+typedef T (*hppl::Active<T>::forward)(T)
          
 
          
 
 
          
-
          
-typedef T (*backward)(T, T)
          
+
          
+typedef >
          
+
          
+typedef T (*hppl::Active<T>::backward)(T, T)
          
 
          
 
 
          
@@ -538,6 +571,253 @@
 
          
 
 
+
          
+
          Defines
          
+
          
+
          
+HL_GPU_LSTM_CUH_
          
+
          
+
+
          
+
          
+
          Functions
          
+
          
+
          
+template <class Op>
          
+
          
+void hl_gpu_lstm_forward(Op op, hl_lstm_value value, int frameSize, int batchSize, hl_activation_mode_t active_node, hl_activation_mode_t active_gate, hl_activation_mode_t active_state)
          
+
          Gpu lstm batch forward. 
          
+
          
+
          Parameters
          
+
            
+
          • op - 
            hl_lstm_ops.cuh 
            
+
            • 
+
          • value - 
            lstm value. 
            
+
            • 
+
          • frameSize - 
            frame size. 
            
+
            • 
+
          • batchSize - 
            size of current batch. 
            
+
            • 
+
          • active_node - 
            active input type. 
            
+
            • 
+
          • active_gate - 
            active state type. 
            
+
            • 
+
          • active_state - 
            actvie gate type. 
            
+
            • 
+
          
+
          
+
          
+
          
+
          
+
+
          
+
          
+template <class Op>
          
+
          
+void hl_gpu_lstm_backward(Op op, hl_lstm_value value, hl_lstm_grad grad, int frameSize, int batchSize, hl_activation_mode_t active_node, hl_activation_mode_t active_gate, hl_activation_mode_t active_state)
          
+
          Gpu lstm batch backward. 
          
+
          
+
          Parameters
          
+
            
+
          • op - 
            hl_lstm_ops.cuh 
            
+
            • 
+
          • value - 
            lstm value. 
            
+
            • 
+
          • grad - 
            lstm gradient. 
            
+
            • 
+
          • frameSize - 
            frame size. 
            
+
            • 
+
          • batchSize - 
            size of current batch. 
            
+
            • 
+
          • active_node - 
            active input type. 
            
+
            • 
+
          • active_gate - 
            active state type. 
            
+
            • 
+
          • active_state - 
            actvie gate type. 
            
+
            • 
+
          
+
          
+
          
+
          
+
          
+
+
          
+
          
+
          Defines
          
+
          
+
          
+HL_LSTM_OPS_CUH_
          
+
          
+
+
          
+
          
+INLINE
          
+
          
+
+
          
+
          
+
          
+namespace hppl
          
+
          
+
          
+namespace backward
          
+
          
+
          
+class lstm
          
+
          
+
          Public Functions
          
+
          
+
          
+INLINE void hppl::backward::lstm::operator()(real & valueIn, real & valueIg, real & valueFg, real & valueOg, real & gradIn, real & gradIg, real & gradFg, real & gradOg, real & prevState, real & prevStateGrad, real & state, real & stateGrad, real & stateAtv, real & outputGrad, real & checkI, real & checkF, real & checkO, real & checkIGrad, real & checkFGrad, real & checkOGrad, Active < real >::backward actInput, Active < real >::backward actGate, Active < real >::backward actState)
          
+
          
+
          Parameters
          
+
            
+
          • valueIn - 
            input 
            
+
            • 
+
          • valueIg - 
            input gate 
            
+
            • 
+
          • valueFg - 
            forget gate 
            
+
            • 
+
          • valueOg - 
            output gate 
            
+
            • 
+
          • gradIn - 
            input grad 
            
+
            • 
+
          • gradIg - 
            input gate grad 
            
+
            • 
+
          • gradFg - 
            forget gate grad 
            
+
            • 
+
          • gradOg - 
            output gate grad 
            
+
            • 
+
          • prevState - 
            previous state value 
            
+
            • 
+
          • prevStateGrad - 
            previous state grad 
            
+
            • 
+
          • state - 
            current state value 
            
+
            • 
+
          • stateGrad - 
            current state grad 
            
+
            • 
+
          • stateAtv - 
            state active 
            
+
            • 
+
          • outputGrad - 
            output grad 
            
+
            • 
+
          • checkI - 
            check input gate 
            
+
            • 
+
          • checkF - 
            check forget gate 
            
+
            • 
+
          • checkO - 
            check output gate 
            
+
            • 
+
          • checkIGrad - 
            check input gate grad 
            
+
            • 
+
          • checkFGrad - 
            check forget gate grad 
            
+
            • 
+
          • checkOGrad - 
            check output gate grad 
            
+
            • 
+
          • actInput - 
            backward function of input 
            
+
            • 
+
          • actGate - 
            backward function of gate 
            
+
            • 
+
          • actState - 
            backward function of state 
            
+
            • 
+
          
+
          
+
          
+
          
+
          
+
+
          
+
          
+
          Public Static Attributes
          
+
          
+
          
+const bool avx
          
+
          
+
+
          
+
          
+
+
          
+
+
          
+
          
+namespace forward
          
+
          
+
          
+class lstm
          
+
          
+
          Public Functions
          
+
          
+
          
+INLINE void hppl::forward::lstm::operator()(real & valueIn, real & valueIg, real & valueFg, real & valueOg, real & prevState, real & state, real & stateAtv, real & output, real & checkI, real & checkF, real & checkO, Active < real >::forward actInput, Active < real >::forward actGate, Active < real >::forward actState)
          
+
          
+
          Parameters
          
+
            
+
          • valueIn - 
            input 
            
+
            • 
+
          • valueIg - 
            input gate 
            
+
            • 
+
          • valueFg - 
            forget gate 
            
+
            • 
+
          • valueOg - 
            output gate 
            
+
            • 
+
          • prevState - 
            previous state 
            
+
            • 
+
          • state - 
            current state 
            
+
            • 
+
          • stateAtv - 
            state active 
            
+
            • 
+
          • output - 
            output 
            
+
            • 
+
          • checkI - 
            check input gate 
            
+
            • 
+
          • checkF - 
            check forget gate 
            
+
            • 
+
          • checkO - 
            check output gate 
            
+
            • 
+
          • actInput - 
            forward function of input 
            
+
            • 
+
          • actGate - 
            forward function of gate 
            
+
            • 
+
          • actState - 
            forward function of state 
            
+
            • 
+
          
+
          
+
          
+
          
+
          
+
+
          
+
          
+
          Public Static Attributes
          
+
          
+
          
+const bool avx
          
+
          
+
+
          
+
          
+
+
          
+
+
          
+
+
+
          
+
          GRU Model
          
+
          
+
          Defines
          
+
          
+
          
+HL_GRU_OPS_CUH_
          
+
          
+
+
          
+
          
+INLINE
          
+
          
+
+
          
+
          
+
          
+namespace hppl
          
+
          
+
          
+namespace backward
          
+
          
+
          
+class gru_stateGrad
          
+
          
+
          Public Functions
          
+
          
+
          
+INLINE void hppl::backward::gru_stateGrad::operator()(real & valueUpdateGate, real & gradUpdateGate, real & valueFrameState, real & gradFrameState, real & valuePrevOut, real & gradPrevOut, real & gradOutput, Active < real >::backward actInput)
          
+
          
+
          Parameters
          
+
            
+
          • valueUpdateGate - 
            update gate value 
            
+
            • 
+
          • gradUpdateGate - 
            update gate grad 
            
+
            • 
+
          • valueFrameState - 
            frame state value 
            
+
            • 
+
          • gradFrameState - 
            frame state grad 
            
+
            • 
+
          • valuePrevOut - 
            previous output value 
            
+
            • 
+
          • gradPrevOut - 
            previous output grad 
            
+
            • 
+
          • gradOutput - 
            output grad 
            
+
            • 
+
          • actInput - 
            backward function of frame state 
            
+
            • 
+
          
+
          
+
          
+
          
+
          
+
+
          
+
          
+
          Public Static Attributes
          
+
          
+
          
+const bool avx
          
+
          
+
+
          
+
          
+
+
          
+
          
+class gru_resetGrad
          
+
          
+
          Public Functions
          
+
          
+
          
+INLINE void hppl::backward::gru_resetGrad::operator()(real & valueUpdateGate, real & gradUpdateGate, real & valueResetGate, real & gradResetGate, real & valuePrevOut, real & gradPrevOut, real & gradResetOutput, Active < real >::backward actGate)
          
+
          
+
          Parameters
          
+
            
+
          • valueUpdateGate - 
            update gate value 
            
+
            • 
+
          • gradUpdateGate - 
            update gate grad 
            
+
            • 
+
          • valueResetGate - 
            reset gate value 
            
+
            • 
+
          • gradResetGate - 
            reset gate grad 
            
+
            • 
+
          • valuePrevOut - 
            previous output value 
            
+
            • 
+
          • gradPrevOut - 
            previous output grad 
            
+
            • 
+
          • gradResetOutput - 
            reset output grad (temp val) 
            
+
            • 
+
          • actGate - 
            backward function of gate 
            
+
            • 
+
          
+
          
+
          
+
          
+
          
+
+
          
+
          
+
          Public Static Attributes
          
+
          
+
          
+const bool avx
          
+
          
+
+
          
+
          
+
+
          
+
+
          
+
          
+namespace forward
          
+
          
+
          
+class gru_resetOutput
          
+
          
+
          Public Functions
          
+
          
+
          
+INLINE void hppl::forward::gru_resetOutput::operator()(real & valueUpdateGate, real & valueResetGate, real & prevOut, real & valueResetOutput, Active < real >::forward actGate)
          
+
          
+
          Parameters
          
+
            
+
          • valueUpdateGate - 
            update gate 
            
+
            • 
+
          • valueResetGate - 
            reset gate 
            
+
            • 
+
          • prevOut - 
            previous output 
            
+
            • 
+
          • valueResetOutput - 
            intermediate value for frame state 
            
+
            • 
+
          • actGate - 
            forward function of gate 
            
+
            • 
+
          
+
          
+
          
+
          
+
          
+
+
          
+
          
+
          Public Static Attributes
          
+
          
+
          
+const bool avx
          
+
          
+
+
          
+
          
+
+
          
+
          
+class gru_finalOutput
          
+
          
+
          Public Functions
          
+
          
+
          
+INLINE void hppl::forward::gru_finalOutput::operator()(real & valueUpdateGate, real & valueFrameState, real & prevOut, real & valueOutput, Active < real >::forward actInput)
          
+
          
+
          Parameters
          
+
            
+
          • valueUpdateGate - 
            update gate 
            
+
            • 
+
          • valueFrameState - 
            frame state ({{h}_t}) 
            
+
            • 
+
          • prevOut - 
            previous output 
            
+
            • 
+
          • valueOutput - 
            output 
            
+
            • 
+
          • actInput - 
            forward function of node 
            
+
            • 
+
          
+
          
+
          
+
          
+
          
+
+
          
+
          
+
          Public Static Attributes
          
+
          
+
          
+const bool avx
          
+
          
+
+
          
+
          
+
+
          
+
+
          
+
+
          
+
          Defines
          
+
          
+
          
+HL_CPU_GRU_CUH_
          
+
          
+
+
          
+
          
+CBLAS_GEMM
          
+
          
+
+
          
+
          
+
          Functions
          
+
          
+
          
+template <class OpResetOutput>
          
+
          
+void hl_naive_gru_forward_reset_output(OpResetOutput opResetOutput, real *gateValue, real *resetOutputValue, real *prevOutputValue, int frameSize, hl_activation_mode_t active_gate)
          
+
          
+
+
          
+
          
+template <class OpFinalOutput>
          
+
          
+void hl_naive_gru_forward_final_output(OpFinalOutput opFinalOutput, real *gateValue, real *prevOutputValue, real *outputValue, int frameSize, hl_activation_mode_t active_node)
          
+
          
+
+
          
+
          
+template <class OpResetOutput>
          
+
          
+void hl_avx_gru_forward_reset_output(OpResetOutput opResetOutput, real *gateValue, real *resetOutputValue, real *prevOutputValue, int frameSize, hl_activation_mode_t active_gate)
          
+
          
+
+
          
+
          
+template <class OpFinalOutput>
          
+
          
+void hl_avx_gru_forward_final_output(OpFinalOutput opFinalOutput, real *gateValue, real *prevOutputValue, real *outputValue, int frameSize, hl_activation_mode_t active_node)
          
+
          
+
+
          
+
          
+template <class OpResetOutput>
          
+
          
+void forward_reset_output(OpResetOutput opResetOutput, hl_gru_value value, int frameSize, int batchSize, hl_activation_mode_t active_gate)
          
+
          
+
+
          
+
          
+template <class OpFinalOutput>
          
+
          
+void forward_final_output(OpFinalOutput opFinalOutput, hl_gru_value value, int frameSize, int batchSize, hl_activation_mode_t active_node)
          
+
          
+
+
          
+
          
+template <class OpResetOutput, class OpFinalOutput>
          
+
          
+void hl_cpu_gru_forward(OpResetOutput opResetOutput, OpFinalOutput opFinalOutput, hl_gru_value value, int frameSize, int batchSize, hl_activation_mode_t active_node, hl_activation_mode_t active_gate)
          
+
          Cpu gru forward. 
          
+
          
+
          Parameters
          
+
            
+
          • opResetOutput - 
            hl_gru_ops.cuh 
            
+
            • 
+
          • opFinalOutput - 
            hl_gru_ops.cuh 
            
+
            • 
+
          • value - 
            gru value. 
            
+
            • 
+
          • frameSize - 
            frame length/size. 
            
+
            • 
+
          • batchSize - 
            size of current batch. 
            
+
            • 
+
          • active_node - 
            active input type. 
            
+
            • 
+
          • active_gate - 
            active state type. 
            
+
            • 
+
          
+
          
+
          
+
          
+
          
+
+
          
+
          
+template <class OpStateGrad>
          
+
          
+void hl_naive_gru_backward_state_grad(OpStateGrad opStateGrad, real *gateValue, real *gateGrad, real *prevOutValue, real *prevOutGrad, real *outputGrad, int frameSize, hl_activation_mode_t active_node)
          
+
          
+
+
          
+
          
+template <class OpResetGrad>
          
+
          
+void hl_naive_gru_backward_reset_grad(OpResetGrad opResetGrad, real *gateValue, real *gateGrad, real *prevOutValue, real *prevOutGrad, real *resetOutputGrad, int frameSize, hl_activation_mode_t active_gate)
          
+
          
+
+
          
+
          
+template <class OpStateGrad>
          
+
          
+void hl_avx_gru_backward_state_grad(OpStateGrad opStateGrad, real *gateValue, real *gateGrad, real *prevOutValue, real *prevOutGrad, real *outputGrad, int frameSize, hl_activation_mode_t active_node)
          
+
          
+
+
          
+
          
+template <class OpResetGrad>
          
+
          
+void hl_avx_gru_backward_reset_grad(OpResetGrad opResetGrad, real *gateValue, real *gateGrad, real *prevOutValue, real *prevOutGrad, real *resetOutputGrad, int frameSize, hl_activation_mode_t active_gate)
          
+
          
+
+
          
+
          
+template <class OpStateGrad>
          
+
          
+void backward_state_grad(OpStateGrad opStateGrad, hl_gru_value value, hl_gru_grad grad, int frameSize, int batchSize, hl_activation_mode_t active_node)
          
+
          
+
+
          
+
          
+template <class OpResetGrad>
          
+
          
+void backward_reset_grad(OpResetGrad opResetGrad, hl_gru_value value, hl_gru_grad grad, int frameSize, int batchSize, hl_activation_mode_t active_gate)
          
+
          
+
+
          
+
          
+template <class OpStateGrad, class OpResetGrad>
          
+
          
+void hl_cpu_gru_backward(OpStateGrad opStateGrad, OpResetGrad opResetGrad, hl_gru_value value, hl_gru_grad grad, int frameSize, int batchSize, hl_activation_mode_t active_node, hl_activation_mode_t active_gate)
          
+
          Cpu gru forward. 
          
+
          
+
          Parameters
          
+
            
+
          • opStateGrad - 
            hl_gru_ops.cuh 
            
+
            • 
+
          • opResetGrad - 
            hl_gru_ops.cuh 
            
+
            • 
+
          • value - 
            gru value. 
            
+
            • 
+
          • grad - 
            gru gradient. 
            
+
            • 
+
          • frameSize - 
            frame length/size. 
            
+
            • 
+
          • batchSize - 
            size of current batch. 
            
+
            • 
+
          • active_node - 
            active input type. 
            
+
            • 
+
          • active_gate - 
            active state type. 
            
+
            • 
+
          
+
          
+
          
+
          
+
          
+
+
          
+
          
+
          Defines
          
+
          
+
          
+HL_GPU_GRU_CUH_
          
+
          
+
+
          
+
          
+
          Functions
          
+
          
+
          
+template <class OpResetOutput, class OpFinalOutput>
          
+
          
+void hl_gpu_gru_forward(OpResetOutput opResetOutput, OpFinalOutput opFinalOutput, hl_gru_value value, int frameSize, int batchSize, hl_activation_mode_t active_node, hl_activation_mode_t active_gate)
          
+
          Gpu gru forward. 
          
+
          
+
          Parameters
          
+
            
+
          • opResetOutput - 
            hl_gru_ops.cuh 
            
+
            • 
+
          • opFinalOutput - 
            hl_gru_ops.cuh 
            
+
            • 
+
          • value - 
            gru value. 
            
+
            • 
+
          • frameSize - 
            frame length/size. 
            
+
            • 
+
          • batchSize - 
            size of current batch. 
            
+
            • 
+
          • active_node - 
            active input type. 
            
+
            • 
+
          • active_gate - 
            active state type. 
            
+
            • 
+
          
+
          
+
          
+
          
+
          
+
+
          
+
          
+template <class OpStateGrad, class OpResetGrad>
          
+
          
+void hl_gpu_gru_backward(OpStateGrad opStateGrad, OpResetGrad opResetGrad, hl_gru_value value, hl_gru_grad grad, int frameSize, int batchSize, hl_activation_mode_t active_node, hl_activation_mode_t active_gate)
          
+
          Gpu gru forward. 
          
+
          
+
          Parameters
          
+
            
+
          • opStateGrad - 
            hl_gru_ops.cuh 
            
+
            • 
+
          • opResetGrad - 
            hl_gru_ops.cuh 
            
+
            • 
+
          • value - 
            gru value. 
            
+
            • 
+
          • grad - 
            gru gradient. 
            
+
            • 
+
          • frameSize - 
            frame length/size. 
            
+
            • 
+
          • batchSize - 
            size of current batch. 
            
+
            • 
+
          • active_node - 
            active input type. 
            
+
            • 
+
          • active_gate - 
            active state type. 
            
+
            • 
+
          
+
          
+
          
+
          
+
          
+
 
          
-
          
-
          GRU Model
          
 
          
 
          
 
@@ -944,14 +1876,11 @@
 
 
         
@@ -973,14 +1902,14 @@
         
        • 
           previous |
          • 
-        
        • PaddlePaddle  documentation »
          • 
-          
        • Source Code Documents »
          • 
-          
        • RNN »
          •  
+        
        • PaddlePaddle  documentation »
          • 
+          
        • Source Code Documents »
          • 
+          
        • RNN »
          •  
       
        
     
     
   
 
\ No newline at end of file
diff --git a/doc/source/cuda/utils/index.html b/doc/source/cuda/utils/index.html
index aec0ffa710..1311c8bd27 100644
--- a/doc/source/cuda/utils/index.html
+++ b/doc/source/cuda/utils/index.html
@@ -6,7 +6,7 @@
   
     
     
-    Utils — PaddlePaddle  documentation
+    Utils — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
      • 
           previous |
        • 
-        
      • PaddlePaddle  documentation »
        • 
-          
      • Source Code Documents »
        •  
+        
      • PaddlePaddle  documentation »
        • 
+          
      • Source Code Documents »
        •  
       
      
       
 
@@ -100,14 +100,11 @@
 
 
         
@@ -129,13 +126,13 @@
         
    • 
           previous |
      • 
-        
    • PaddlePaddle  documentation »
      • 
-          
    • Source Code Documents »
      •  
+        
    • PaddlePaddle  documentation »
      • 
+          
    • Source Code Documents »
      •  
       
    
     
     
   
 
\ No newline at end of file
diff --git a/doc/source/cuda/utils/utils.html b/doc/source/cuda/utils/utils.html
index cf760cc92c..c1bfd92865 100644
--- a/doc/source/cuda/utils/utils.html
+++ b/doc/source/cuda/utils/utils.html
@@ -6,7 +6,7 @@
   
     
     
-    Utilities — PaddlePaddle  documentation
+    Utilities — PaddlePaddle  documentation
     
     
     
@@ -45,9 +45,9 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    • 
-          
  • Utils »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    • 
+          
  • Utils »
    •  
       
       
 
@@ -603,6 +603,183 @@
 
    Thread Resource
    
 
    
 
    hl_thread.ph
    
+
    
+
    Defines
    
+
    
+
    
+HL_THREAD_PH_
    
+
    
+
+
    
+
    
+
    Typedefs
    
+
    
+
    
+typedef struct _hl_thread_resource *hl_thread_resource
    
+
    
+
+
    
+
    
+
    Functions
    
+
    
+
    
+void hl_cudnn_init(cudnnHandle_t *cudnn_handle, cudaStream_t stream)
    
+
    Initialize cudnn. 
    
+
    
+
    Parameters
    
+
      
+
    • cudnn_handle - 
      Cudnn handle. 
      
+
      • 
+
    • stream - 
      Cudnn stream. 
      
+
      • 
+
    
+
    
+
    
+
    
+
    
+
+
    
+
    
+void hl_cublas_init(cublasHandle_t *cublas_handle, cudaStream_t stream)
    
+
    Initialize cublas. 
    
+
    
+
    Parameters
    
+
      
+
    • cublas_handle - 
      Cublas handle. 
      
+
      • 
+
    • stream - 
      Cuda stream. 
      
+
      • 
+
    
+
    
+
    
+
    
+
    
+
+
    
+
    
+void hl_cudnn_desc_init(cudnnTensorDescriptor_t *cudnn_desc)
    
+
    Initialize cudnn tensor descriptor. 
    
+
    
+
    Parameters
    
+
      
+
    • cudnn_desc - 
      Cudnn tensor descriptor. 
      
+
      • 
+
    
+
    
+
    
+
    
+
    
+
+
    
+
    
+
    Variables
    
+
    
+
    
+__thread _hl_thread_resource t_resource
    
+
    thread resource. 
    
+
    
+
+
    
+
    
+
    
+struct _hl_thread_resource
    
+
    Thread resource structure. 
    
+
    
+
    Parameters
    
+
      
+
    • stream[HPPL_STREAM_END] - 
      Stream for thread. 
      
+
      • 
+
    • handle - 
      Cublas Handle. 
      
+
      • 
+
    • gen - 
      Curand Generator. 
      
+
      • 
+
    • cudnn_handle - 
      Cudnn handle. 
      
+
      • 
+
    • cudnn_desc - 
      Cudnn image descriptor. 
      
+
      • 
+
    • *gen_mutex - 
      Gen lock. 
      
+
      • 
+
    • *gpu_mem - 
      HPPL GPU Memory. 
      
+
      • 
+
    • *cpu_mem - 
      HPPL CPU Memory. 
      
+
      • 
+
    • event - 
      gpu_mem event. 
      
+
      • 
+
    • device - 
      Thread device context. 
      
+
      • 
+
    • major - 
      Compute capability. 
      
+
      • 
+
    • is_init - 
      Thread init or not. 
      
+
      • 
+
    
+
    
+
    
+
    
+
    
+
    Public Members
    
+
    
+
    
+cudaStream_t stream[HPPL_STREAM_END]
    
+
    
+
+
    
+
    
+cublasHandle_t handle
    
+
    
+
+
    
+
    
+curandGenerator_t gen
    
+
    
+
+
    
+
    
+cudnnHandle_t cudnn_handle
    
+
    
+
+
    
+
    
+cudnnTensorDescriptor_t cudnn_desc
    
+
    
+
+
    
+
    
+pthread_mutex_t *gen_mutex
    
+
    
+
+
    
+
    
+real *gpu_mem
    
+
    
+
+
    
+
    
+real *cpu_mem
    
+
    
+
+
    
+
    
+cudaEvent_t event
    
+
    
+
+
    
+
    
+int device
    
+
    
+
+
    
+
    
+int major
    
+
    
+
+
    
+
    
+bool is_init
    
+
    
+
+
    
+
    
+
 
    
 
 
@@ -648,14 +825,11 @@
 
 
         
@@ -677,14 +851,14 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    • 
-          
  • Utils »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    • 
+          
  • Utils »
    •  
       
     
     
   
 
\ No newline at end of file
diff --git a/doc/source/gserver/activations/index.html b/doc/source/gserver/activations/index.html
index adea6c71e7..ea1d25bb85 100644
--- a/doc/source/gserver/activations/index.html
+++ b/doc/source/gserver/activations/index.html
@@ -6,7 +6,7 @@
   
     
     
-    Activations — PaddlePaddle  documentation
+    Activations — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
       
 
@@ -83,24 +83,24 @@
 
    
 virtual void forward(Argument &act) = 0
    
 
    Foward propagation. 
    
-
    act.value <- f(act.value), where f is the activation function. Suppose that before calling forward(), act.value is x and after forward() is called, act.value is y, then y = f(x).
    
-
    Usually, act is Layer::output_ 
    
+
    act.value <- f(act.value), where f is the activation function. Suppose that before calling forward(), act.value is x and after forward() is called, act.value is y, then y = f(x).
    
+
    Usually, act is Layer::output_ 
    
 
    
 
 
    
 
    
 virtual void backward(Argument &act) = 0
    
 
    Backward propagaion. 
    
-
    x and y are defined in the above comment for forward().
      
-
    • Before calling backward(), act.grad = dE / dy, where E is the error/cost
      • 
-
    • After backward() returns, act.grad = dE / dx = (dE/dy) * (dy/dx) 
      • 
+
      x and y are defined in the above comment for forward().
        
+
      • Before calling backward(), act.grad = dE / dy, where E is the error/cost
        • 
+
      • After backward() returns, act.grad = dE / dx = (dE/dy) * (dy/dx) 
        • 
 
      
 
      
 
    
 
 
    
 
    
-virtual const std::string &getName() const = 0
    
+virtual const std::string &getName() const = 0
 
    
 
 
@@ -108,7 +108,7 @@
 
    Public Static Functions
    
 
    
 
    
-ActivationFunction *create(const std::string &type)
    
+ActivationFunction *create(const std::string &type)
 
    
 
 
@@ -138,14 +138,11 @@
 
 
         
@@ -167,13 +164,13 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
     
   
 
\ No newline at end of file
diff --git a/doc/source/gserver/dataprovider/dataproviders.html b/doc/source/gserver/dataprovider/dataproviders.html
index 7f76f66e18..f1fb1da744 100644
--- a/doc/source/gserver/dataprovider/dataproviders.html
+++ b/doc/source/gserver/dataprovider/dataproviders.html
@@ -6,7 +6,7 @@
   
     
     
-    Data Providers — PaddlePaddle  documentation
+    Data Providers — PaddlePaddle  documentation
     
     
     
@@ -45,9 +45,9 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    • 
-          
  • Data Providers Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    • 
+          
  • Data Providers Documents »
    •  
       
       
 
@@ -63,13 +63,13 @@
 
    
 
    
 class paddle::DataProvider
    
-
    Base class for DataProvider, which supplies data for training. 
    
+
    Base class for DataProvider, which supplies data for training. 
    
 
    
 
    Note
    
 
    It can supplies multiple streams of data. For typical supervised training, there are two streams: one is for input, one is for label. 
    
 
    
 
    
-
    Subclassed by paddle::DataProviderGroup< T >, paddle::DummyDataProvider, paddle::MultiDataProvider, paddle::ProtoDataProvider, paddle::PyDataProvider, paddle::PyDataProvider2, paddle::SimpleDataProviderBase
    
+
    Subclassed by paddle::DataProviderGroup< T >, paddle::DummyDataProvider, paddle::MultiDataProvider, paddle::ProtoDataProvider, paddle::PyDataProvider, paddle::PyDataProvider2, paddle::SimpleDataProviderBase
    
 
    
 
    Public Functions
    
 
    
@@ -123,7 +123,7 @@
 
    reset all the value of index 
    
 
    
 
    Note
    
-
    reset() must be called before any calls to getNextBatch() IMPORTANT: subclass reset() should always call the base class reset() at the end of the function 
    
+
    reset() must be called before any calls to getNextBatch() IMPORTANT: subclass reset() should always call the base class reset() at the end of the function 
    
 
    
 
    
 
    
@@ -165,7 +165,7 @@
 
    Public Static Functions
    
 
    
 
    
-DataProvider *create(const DataConfig &config, bool useGpu = FLAGS_use_gpu)
    
+DataProvider *create(const DataConfig &config, bool useGpu = FLAGS_use_gpu)
 
    
 
 
    
@@ -173,7 +173,7 @@
 
    Public Static Attributes
    
 
    
 
    
-ClassRegistrar<DataProvider, DataConfig, bool> registrar_
    
+ClassRegistrar<DataProvider, DataConfig, bool> registrar_
 
    
 
 
@@ -246,7 +246,7 @@
 template <class T>
 
    
 class paddle::DataProviderGroup
    
-
    Inherits from paddle::DataProvider
    
+
    Inherits from paddle::DataProvider
    
 
    
 
    Public Functions
    
 
    
@@ -265,7 +265,7 @@
 
    reset all the value of index 
    
 
    
 
    Note
    
-
    reset() must be called before any calls to getNextBatch() IMPORTANT: subclass reset() should always call the base class reset() at the end of the function 
    
+
    reset() must be called before any calls to getNextBatch() IMPORTANT: subclass reset() should always call the base class reset() at the end of the function 
    
 
    
 
    
 
    
@@ -353,7 +353,7 @@
 
    
 
    
 class paddle::MultiDataProvider
    
-
    Inherits from paddle::DataProvider
    
+
    Inherits from paddle::DataProvider
    
 
    
 
    Public Functions
    
 
    
@@ -372,7 +372,7 @@
 
    reset all the value of index 
    
 
    
 
    Note
    
-
    reset() must be called before any calls to getNextBatch() IMPORTANT: subclass reset() should always call the base class reset() at the end of the function 
    
+
    reset() must be called before any calls to getNextBatch() IMPORTANT: subclass reset() should always call the base class reset() at the end of the function 
    
 
    
 
    
 
    
@@ -442,12 +442,12 @@
 class paddle::IFieldScanner
 
    FieldScanner Interface.
    
 
    It will read python object, and fill to argument’s each slot. There are two steps, prepare and fill. Scanner will alloc memory during prepare step, fill data into argument during fill step. 
    
-
    Subclassed by paddle::DenseScanner, paddle::IndexScanner, paddle::SequenceScanner, paddle::SparseNonValueScanner
    
+
    Subclassed by paddle::DenseScanner, paddle::IndexScanner, paddle::SequenceScanner, paddle::SparseNonValueScanner
    
 
    
 
    Public Functions
    
 
    
 
    
-DISABLE_COPY(IFieldScanner)
    
+DISABLE_COPY(IFieldScanner)
 
    
 
 
    
@@ -520,7 +520,7 @@
 
    Public Static Functions
    
 
    
 
    
-IFieldScanner *create(SlotHeader *header)
    
+IFieldScanner *create(SlotHeader *header)
 
    Factory method. Create a scanner by header. The final scanner may be combine many scanners.
    
 
    
 
    Note
    
@@ -547,7 +547,7 @@
 
    
 class paddle::DenseScanner
    
 
    Scanner for dense slot. 
    
-
    Inherits from paddle::IFieldScanner
    
+
    Inherits from paddle::IFieldScanner
    
 
    
 
    Public Functions
    
 
    
@@ -603,7 +603,7 @@
 
    
 class paddle::IndexScanner
    
 
    Scanner for index slot 
    
-
    Inherits from paddle::IFieldScanner
    
+
    Inherits from paddle::IFieldScanner
    
 
    
 
    Public Functions
    
 
    
@@ -643,7 +643,7 @@
 
    
 
    
 class paddle::SparseNonValueScanner
    
-
    Inherits from paddle::IFieldScanner
    
+
    Inherits from paddle::IFieldScanner
    
 
    Subclassed by paddle::SparseValueScanner
    
 
    
 
    Public Functions
    
@@ -727,7 +727,7 @@
 
    
 
    
 class paddle::SparseValueScanner
    
-
    Inherits from paddle::SparseNonValueScanner
    
+
    Inherits from paddle::SparseNonValueScanner
    
 
    
 
    Public Functions
    
 
    
@@ -770,18 +770,18 @@
 
    SequenceScanner
    
 
    
 
    
-class paddle::SparseValueScanner
    
-
    Inherits from paddle::SparseNonValueScanner
    
+class paddle::SparseValueScanner
+
    Inherits from paddle::SparseNonValueScanner
    
 
    
 
    Public Functions
    
 
    
 
    
-SparseValueScanner(SlotHeader *ptr)
    
+SparseValueScanner(SlotHeader *ptr)
 
    
 
 
    
 
    
-virtual void finishPrepare(Argument &argument)
    
+virtual void finishPrepare(Argument &argument)
 
    Finish Prepare step. 
    
 
    
 
@@ -790,7 +790,7 @@
 
    Protected Functions
    
 
    
 
    
-virtual void setData(int *col, real *dat, PyObject *obj)
    
+virtual void setData(int *col, real *dat, PyObject *obj)
 
    Set a single sparse index and value. 
    
 
    Parameters
    
 
    
     
      
 
@@ -109,14 +109,11 @@
 
 
         
    
@@ -138,13 +135,13 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/source/gserver/evaluators/evaluators.html b/doc/source/gserver/evaluators/evaluators.html
index 4ae25889f8..8941a097e8 100644
--- a/doc/source/gserver/evaluators/evaluators.html
+++ b/doc/source/gserver/evaluators/evaluators.html
@@ -6,7 +6,7 @@
   
     
     
-    Base Evaluator — PaddlePaddle  documentation
+    Base Evaluator — PaddlePaddle  documentation
     
     
     
@@ -45,9 +45,9 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    • 
-          
  • Evaluators »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    • 
+          
  • Evaluators »
    •  
       
     
      
 
@@ -63,8 +63,8 @@
 
    
 
    
 class paddle::Evaluator
    
-
    Base class for Evaluator Evaluating the performance of a model is very important. It indicates how successful the scores(predictions) of a datasets has been by a trained model. 
    
-
    Subclassed by paddle::AucEvaluator, paddle::ChunkEvaluator, paddle::ClassificationErrorEvaluator, paddle::ColumnSumEvaluator, paddle::CombinedEvaluator, paddle::CTCErrorEvaluator, paddle::DummyEvaluator, paddle::GradientPrinter, paddle::MaxFramePrinter, paddle::MaxIdPrinter, paddle::MultiCombinedEvaluator, paddle::PnpairEvaluator, paddle::PrecisionRecallEvaluator, paddle::RankAucEvaluator, paddle::SequenceTextPrinter, paddle::SumEvaluator, paddle::ValuePrinter
    
+
    Base class for Evaluator Evaluating the performance of a model is very important. It indicates how successful the scores(predictions) of a datasets has been by a trained model. 
    
+
    Subclassed by paddle::AucEvaluator, paddle::ChunkEvaluator, paddle::ClassificationErrorEvaluator, paddle::ColumnSumEvaluator, paddle::CombinedEvaluator, paddle::CTCErrorEvaluator, paddle::DummyEvaluator, paddle::GradientPrinter, paddle::MaxFramePrinter, paddle::MaxIdPrinter, paddle::MultiCombinedEvaluator, paddle::PnpairEvaluator, paddle::PrecisionRecallEvaluator, paddle::RankAucEvaluator, paddle::SequenceTextPrinter, paddle::SumEvaluator, paddle::ValuePrinter
    
 
    
 
    Public Functions
    
 
    
@@ -102,7 +102,7 @@
 
    Return
    
 
    the score for the batch if it make sense to sum the score across batches. 
    
 
    Note
    
-
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
+
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
 
    
 
    
 
    
@@ -116,7 +116,7 @@
 
    
 
    
 virtual void distributeEval(ParameterClient2 *client)
    
-
    finish() should be called before distributeEval 
    
+
    finish() should be called before distributeEval 
    
 
    
 
 
    
@@ -136,7 +136,7 @@
 
    print the statistics of evaluate result 
    
 
    
 
    Note
    
-
    finish() should be called before printStats 
    
+
    finish() should be called before printStats 
    
 
    
 
    
 
    
@@ -146,7 +146,7 @@
 
    Public Static Functions
    
 
    
 
    
-Evaluator *create(const EvaluatorConfig &config)
    
+Evaluator *create(const EvaluatorConfig &config)
 
    
 
 
    
@@ -154,7 +154,7 @@
 
    Public Static Attributes
    
 
    
 
    
-ClassRegistrar<Evaluator> registrar_
    
+ClassRegistrar<Evaluator> registrar_
 
    
 
 
    
@@ -200,9 +200,9 @@
 
    
 
    
 class paddle::SumEvaluator
    
-
    sum Evaluator Calculate the sum of output or label 
    
+
    sum Evaluator Calculate the sum of output or label 
    
 
    The config file api is sum_evaluator. 
    
-
    Inherits from paddle::Evaluator
    
+
    Inherits from paddle::Evaluator
    
 
    
 
    Public Functions
    
 
    
@@ -224,7 +224,7 @@
 
    Return
    
 
    the score for the batch if it make sense to sum the score across batches. 
    
 
    Note
    
-
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
+
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
 
    
 
    
 
    
@@ -232,7 +232,7 @@
 
    
 
    
 virtual void distributeEval(ParameterClient2 *client)
    
-
    finish() should be called before distributeEval 
    
+
    finish() should be called before distributeEval 
    
 
    
 
 
    
@@ -244,7 +244,7 @@
 
    
 
    
 class paddle::ColumnSumEvaluator
    
-
    column sum Evaluator 
    
+
    column sum Evaluator 
    
 
    
 The config file api is column_sum_evaluator. 
    
 
    Note
    
@@ -256,7 +256,7 @@ The config file api is column_sum_evaluator. 
    
 
    
 
    
 
    
-
    Inherits from paddle::Evaluator
    
+
    Inherits from paddle::Evaluator
    
 
    
 
    Public Functions
    
 
    
@@ -284,7 +284,7 @@ The config file api is column_sum_evaluator. 
    
 
    Return
    
 
    the score for the batch if it make sense to sum the score across batches. 
    
 
    Note
    
-
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
+
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
 
    
 
    
 
    
@@ -295,7 +295,7 @@ The config file api is column_sum_evaluator. 
    
 
    print the statistics of evaluate result 
    
 
    
 
    Note
    
-
    finish() should be called before printStats 
    
+
    finish() should be called before printStats 
    
 
    
 
    
 
    
@@ -303,7 +303,7 @@ The config file api is column_sum_evaluator. 
    
 
    
 
    
 virtual void distributeEval(ParameterClient2 *client)
    
-
    finish() should be called before distributeEval 
    
+
    finish() should be called before distributeEval 
    
 
    
 
 
@@ -318,10 +318,10 @@ The config file api is column_sum_evaluator. 
    
 
    
 
    
 class paddle::ClassificationErrorEvaluator
    
-
    classification error Evaluator 
    
+
    classification error Evaluator 
    
 
    The config file api is classification_error_evaluator. 
    
-
    Inherits from paddle::Evaluator
    
-
    Subclassed by paddle::ClassificationErrorPrinter, paddle::SequenceClassificationErrorEvaluator
    
+
    Inherits from paddle::Evaluator
    
+
    Subclassed by paddle::ClassificationErrorPrinter, paddle::SequenceClassificationErrorEvaluator
    
 
    
 
    Public Functions
    
 
    
@@ -343,7 +343,7 @@ The config file api is column_sum_evaluator. 
    
 
    Return
    
 
    the score for the batch if it make sense to sum the score across batches. 
    
 
    Note
    
-
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
+
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
 
    
 
    
 
    
@@ -351,7 +351,7 @@ The config file api is column_sum_evaluator. 
    
 
    
 
    
 virtual void distributeEval(ParameterClient2 *client)
    
-
    finish() should be called before distributeEval 
    
+
    finish() should be called before distributeEval 
    
 
    
 
 
@@ -363,13 +363,13 @@ The config file api is column_sum_evaluator. 
    
 
    
 
    
 class paddle::SequenceClassificationErrorEvaluator
    
-
    sequence classification error Evaluator 
    
+
    sequence classification error Evaluator 
    
 
    
 
    Note
    
 
    sequence level classification error stats, if any frame in one sequence has error, the sequence is error 
    
 
    
 
    
-
    Inherits from paddle::ClassificationErrorEvaluator
    
+
    Inherits from paddle::ClassificationErrorEvaluator
    
 
    
 
    Public Functions
    
 
    
@@ -386,7 +386,7 @@ The config file api is column_sum_evaluator. 
    
 
    Return
    
 
    the score for the batch if it make sense to sum the score across batches. 
    
 
    Note
    
-
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
+
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
 
    
 
    
 
    
@@ -394,7 +394,7 @@ The config file api is column_sum_evaluator. 
    
 
    
 
    
 virtual void distributeEval(ParameterClient2 *client)
    
-
    finish() should be called before distributeEval 
    
+
    finish() should be called before distributeEval 
    
 
    
 
 
@@ -418,7 +418,7 @@ The config file api is column_sum_evaluator. 
    
 
    
 
    
 
    The config file api is auc_evaluator. 
    
-
    Inherits from paddle::Evaluator
    
+
    Inherits from paddle::Evaluator
    
 
    
 
    Public Functions
    
 
    
@@ -440,7 +440,7 @@ The config file api is column_sum_evaluator. 
    
 
    Return
    
 
    the score for the batch if it make sense to sum the score across batches. 
    
 
    Note
    
-
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
+
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
 
    
 
    
 
    
@@ -451,7 +451,7 @@ The config file api is column_sum_evaluator. 
    
 
    print the statistics of evaluate result 
    
 
    
 
    Note
    
-
    finish() should be called before printStats 
    
+
    finish() should be called before printStats 
    
 
    
 
    
 
    
@@ -459,7 +459,7 @@ The config file api is column_sum_evaluator. 
    
 
    
 
    
 virtual void distributeEval(ParameterClient2 *client)
    
-
    finish() should be called before distributeEval 
    
+
    finish() should be called before distributeEval 
    
 
    
 
 
@@ -471,11 +471,11 @@ The config file api is column_sum_evaluator. 
    
 
    
 
    
 class paddle::PrecisionRecallEvaluator
    
-
    precision, recall and f1 score Evaluator 
    
+
    precision, recall and f1 score Evaluator 
    
 \[\begin{split} precision = \frac{tp}{tp+tn} \\ recall=\frac{tp}{tp+fn} \\ f1=2*\frac{precsion*recall}{precision+recall} \end{split}\]
    
  
    
 
    The config file api is precision_recall_evaluator. 
    
-
    Inherits from paddle::Evaluator
    
+
    Inherits from paddle::Evaluator
    
 
    
 
    Public Functions
    
 
    
@@ -497,7 +497,7 @@ The config file api is column_sum_evaluator. 
    
 
    Return
    
 
    the score for the batch if it make sense to sum the score across batches. 
    
 
    Note
    
-
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
+
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
 
    
 
    
 
    
@@ -508,7 +508,7 @@ The config file api is column_sum_evaluator. 
    
 
    print the statistics of evaluate result 
    
 
    
 
    Note
    
-
    finish() should be called before printStats 
    
+
    finish() should be called before printStats 
    
 
    
 
    
 
    
@@ -516,7 +516,7 @@ The config file api is column_sum_evaluator. 
    
 
    
 
    
 virtual void distributeEval(ParameterClient2 *client)
    
-
    finish() should be called before distributeEval 
    
+
    finish() should be called before distributeEval 
    
 
    
 
 
@@ -570,23 +570,23 @@ The config file api is column_sum_evaluator. 
    
 class paddle::ChunkEvaluator
 
    Chunk evaluator is used to evaluate segment labelling accuracy for a sequence. It calculates the chunk detection F1 score.
    
 
    A chunk is correctly detected if its beginning, end and type are correct. Other chunk type is ignored. For each label in the label sequence, we have
    
-
    tagType = label % numTagType
+
    tagType = label % numTagType
 chunkType = label / numTagType
 otherChunkType = numChunkTypes
 
    
 
    
 
    
 
    The total number of different labels is numTagType*numChunkTypes+1 We support 4 labelling scheme The tag type for each of the scheme is shown as follows:
    
-
    Scheme Begin Inside End   Single
- plain  0     -      -     -
- IOB    0     1      -     -
- IOE    -     0      1     -
- IOBES  0     1      2     3
+
    Scheme Begin Inside End   Single
+ plain  0     -      -     -
+ IOB    0     1      -     -
+ IOE    -     0      1     -
+ IOBES  0     1      2     3
 
    
 
    
 
    
 
    ‘plain’ means the whole chunk must contain exactly the same chunk label. 
    
-
    Inherits from paddle::Evaluator
    
+
    Inherits from paddle::Evaluator
    
 
    
 
    Public Functions
    
 
    
@@ -606,7 +606,7 @@ The config file api is column_sum_evaluator. 
    
 
    print the statistics of evaluate result 
    
 
    
 
    Note
    
-
    finish() should be called before printStats 
    
+
    finish() should be called before printStats 
    
 
    
 
    
 
    
@@ -614,7 +614,7 @@ The config file api is column_sum_evaluator. 
    
 
    
 
    
 virtual void distributeEval(ParameterClient2 *client)
    
-
    finish() should be called before distributeEval 
    
+
    finish() should be called before distributeEval 
    
 
    
 
 
    
@@ -625,7 +625,7 @@ The config file api is column_sum_evaluator. 
    
 
    Return
    
 
    the score for the batch if it make sense to sum the score across batches. 
    
 
    Note
    
-
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
+
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
 
    
 
    
 
    
@@ -660,7 +660,7 @@ The config file api is column_sum_evaluator. 
    
 
    
 class paddle::CTCErrorEvaluator
    
 
    calculate sequence-to-sequence edit distance 
    
-
    Inherits from paddle::Evaluator
    
+
    Inherits from paddle::Evaluator
    
 
    
 
    Public Functions
    
 
    
@@ -676,7 +676,7 @@ The config file api is column_sum_evaluator. 
    
 
    Return
    
 
    the score for the batch if it make sense to sum the score across batches. 
    
 
    Note
    
-
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
+
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
 
    
 
    
 
    
@@ -699,7 +699,7 @@ The config file api is column_sum_evaluator. 
    
 
    print the statistics of evaluate result 
    
 
    
 
    Note
    
-
    finish() should be called before printStats 
    
+
    finish() should be called before printStats 
    
 
    
 
    
 
    
@@ -707,7 +707,7 @@ The config file api is column_sum_evaluator. 
    
 
    
 
    
 virtual void distributeEval(ParameterClient2 *client)
    
-
    finish() should be called before distributeEval 
    
+
    finish() should be called before distributeEval 
    
 
    
 
 
@@ -722,7 +722,7 @@ The config file api is column_sum_evaluator. 
    
 
    
 
    
 class paddle::PnpairEvaluator
    
-
    Inherits from paddle::Evaluator
    
+
    Inherits from paddle::Evaluator
    
 
    
 
    Public Functions
    
 
    
@@ -744,7 +744,7 @@ The config file api is column_sum_evaluator. 
    
 
    Return
    
 
    the score for the batch if it make sense to sum the score across batches. 
    
 
    Note
    
-
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
+
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
 
    
 
    
 
    
@@ -776,7 +776,7 @@ The config file api is column_sum_evaluator. 
    
 
    print the statistics of evaluate result 
    
 
    
 
    Note
    
-
    finish() should be called before printStats 
    
+
    finish() should be called before printStats 
    
 
    
 
    
 
    
@@ -784,7 +784,7 @@ The config file api is column_sum_evaluator. 
    
 
    
 
    
 virtual void distributeEval(ParameterClient2 *client)
    
-
    finish() should be called before distributeEval 
    
+
    finish() should be called before distributeEval 
    
 
    
 
 
@@ -840,8 +840,8 @@ The config file api is column_sum_evaluator. 
    
 
    
 
    
 class paddle::RankAucEvaluator
    
-
    RankAucEvaluator calculates the AUC of each list (i.e., titles under the same query), and averages them. Each list should be organized as a sequence. The inputs of this evaluator is [output, click, pv]. If pv is not provided, it will be set to 1. The types of click and pv are dense value. 
    
-
    Inherits from paddle::Evaluator
    
+
    RankAucEvaluator calculates the AUC of each list (i.e., titles under the same query), and averages them. Each list should be organized as a sequence. The inputs of this evaluator is [output, click, pv]. If pv is not provided, it will be set to 1. The types of click and pv are dense value. 
    
+
    Inherits from paddle::Evaluator
    
 
    
 
    Public Functions
    
 
    
@@ -864,7 +864,7 @@ The config file api is column_sum_evaluator. 
    
 
    Return
    
 
    the score for the batch if it make sense to sum the score across batches. 
    
 
    Note
    
-
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
+
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
 
    
 
    
 
    
@@ -872,7 +872,7 @@ The config file api is column_sum_evaluator. 
    
 
    
 
    
 virtual void distributeEval(ParameterClient2 *client)
    
-
    finish() should be called before distributeEval 
    
+
    finish() should be called before distributeEval 
    
 
    
 
 
@@ -889,7 +889,7 @@ The config file api is column_sum_evaluator. 
    
 class paddle::ValuePrinter
 
    print value of each layer. 
    
 
    The config file api is value_printer_evaluator. 
    
-
    Inherits from paddle::Evaluator
    
+
    Inherits from paddle::Evaluator
    
 
    
 
    Public Functions
    
 
    
@@ -917,7 +917,7 @@ The config file api is column_sum_evaluator. 
    
 
    Return
    
 
    the score for the batch if it make sense to sum the score across batches. 
    
 
    Note
    
-
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
+
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
 
    
 
    
 
    
@@ -933,7 +933,7 @@ The config file api is column_sum_evaluator. 
    
 class paddle::GradientPrinter
 
    print gradient of each layer. 
    
 
    The config file api is gradient_printer_evaluator. 
    
-
    Inherits from paddle::Evaluator
    
+
    Inherits from paddle::Evaluator
    
 
    
 
    Public Functions
    
 
    
@@ -961,7 +961,7 @@ The config file api is column_sum_evaluator. 
    
 
    Return
    
 
    the score for the batch if it make sense to sum the score across batches. 
    
 
    Note
    
-
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
+
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
 
    
 
    
 
    
@@ -977,7 +977,7 @@ The config file api is column_sum_evaluator. 
    
 class paddle::MaxIdPrinter
 
    print row max id vctor of each layer 
    
 
    The config file api is maxid_printer_evaluator. 
    
-
    Inherits from paddle::Evaluator
    
+
    Inherits from paddle::Evaluator
    
 
    
 
    Public Functions
    
 
    
@@ -1005,7 +1005,7 @@ The config file api is column_sum_evaluator. 
    
 
    Return
    
 
    the score for the batch if it make sense to sum the score across batches. 
    
 
    Note
    
-
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
+
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
 
    
 
    
 
    
@@ -1021,7 +1021,7 @@ The config file api is column_sum_evaluator. 
    
 class paddle::MaxFramePrinter
 
    print sequence max frames of each layer 
    
 
    The config file api is maxframe_printer_evaluator. 
    
-
    Inherits from paddle::Evaluator
    
+
    Inherits from paddle::Evaluator
    
 
    
 
    Public Functions
    
 
    
@@ -1049,7 +1049,7 @@ The config file api is column_sum_evaluator. 
    
 
    Return
    
 
    the score for the batch if it make sense to sum the score across batches. 
    
 
    Note
    
-
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
+
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
 
    
 
    
 
    
@@ -1074,7 +1074,7 @@ The config file api is column_sum_evaluator. 
    
 
  • sequence without sub-sequence, and there is probability.
    • 
 
 
    
-
    id \t prob space_seperated_tokens_from_dictionary_according_to_seq
+
    id \t prob space_seperated_tokens_from_dictionary_according_to_seq
 
    
 
    
 
    
@@ -1082,7 +1082,7 @@ The config file api is column_sum_evaluator. 
    
 
  • sequence without sub-sequence, and there is not probability.
    • 
 
 
    
-
    id \t space_seperated_tokens_from_dictionary_according_to_seq
+
    id \t space_seperated_tokens_from_dictionary_according_to_seq
 
    
 
    
 
    
@@ -1090,15 +1090,15 @@ The config file api is column_sum_evaluator. 
    
 
  • sequence with sub-sequence, and there is not probability.
    • 
 
 
    
-
    id \t space_seperated_tokens_from_dictionary_according_to_sub_seq
-\t \t space_seperated_tokens_from_dictionary_according_to_sub_seq
-...
+
    id \t space_seperated_tokens_from_dictionary_according_to_sub_seq
+\t \t space_seperated_tokens_from_dictionary_according_to_sub_seq
+...
 
    
 
    
 
    
-
    Typically SequenceTextPrinter layer takes output of maxid or RecurrentGroup with maxid (when generating) as an input.
    
+
    Typically SequenceTextPrinter layer takes output of maxid or RecurrentGroup with maxid (when generating) as an input.
    
 
    The config file api is seqtext_printer_evaluator. 
    
-
    Inherits from paddle::Evaluator
    
+
    Inherits from paddle::Evaluator
    
 
    
 
    Public Functions
    
 
    
@@ -1125,7 +1125,7 @@ The config file api is column_sum_evaluator. 
    
 
    Return
    
 
    the score for the batch if it make sense to sum the score across batches. 
    
 
    Note
    
-
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
+
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
 
    
 
    
 
    
@@ -1141,7 +1141,7 @@ The config file api is column_sum_evaluator. 
    
 class paddle::ClassificationErrorPrinter
 
    print classification error. 
    
 
    The config file api is classification_error_printer_evaluator. 
    
-
    Inherits from paddle::ClassificationErrorEvaluator
    
+
    Inherits from paddle::ClassificationErrorEvaluator
    
 
    
 
    Public Functions
    
 
    
@@ -1158,7 +1158,7 @@ The config file api is column_sum_evaluator. 
    
 
    Return
    
 
    the score for the batch if it make sense to sum the score across batches. 
    
 
    Note
    
-
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
+
    Otherwise evaluator should return 0 and override finish() and printStats() to do the right calculation. 
    
 
    
 
    
 
    
@@ -1227,14 +1227,11 @@ The config file api is column_sum_evaluator. 
    
 
 
         
    
@@ -1256,14 +1253,14 @@ The config file api is column_sum_evaluator. 
    
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    • 
-          
  • Evaluators »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    • 
+          
  • Evaluators »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/source/gserver/evaluators/index.html b/doc/source/gserver/evaluators/index.html
index 80890914d8..3644250675 100644
--- a/doc/source/gserver/evaluators/index.html
+++ b/doc/source/gserver/evaluators/index.html
@@ -6,7 +6,7 @@
   
     
     
-    Evaluators — PaddlePaddle  documentation
+    Evaluators — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
      
 
@@ -117,14 +117,11 @@
 
 
         
    
@@ -146,13 +143,13 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/source/gserver/gradientmachines/gradientmachines.html b/doc/source/gserver/gradientmachines/gradientmachines.html
index 57e97ef395..75b1a8e3a0 100644
--- a/doc/source/gserver/gradientmachines/gradientmachines.html
+++ b/doc/source/gserver/gradientmachines/gradientmachines.html
@@ -6,7 +6,7 @@
   
     
     
-    Gradient Machines — PaddlePaddle  documentation
+    Gradient Machines — PaddlePaddle  documentation
     
     
     
@@ -45,9 +45,9 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    • 
-          
  • Gradient Machines Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    • 
+          
  • Gradient Machines Documents »
    •  
       
     
      
 
@@ -63,7 +63,7 @@
 
    
 
    
 class paddle::GradientMachine
    
-
    Subclassed by paddle::MultiGradientMachine, paddle::NeuralNetwork
    
+
    Subclassed by paddle::MultiGradientMachine, paddle::NeuralNetwork
    
 
    
 
    Public Types
    
 
    
@@ -113,7 +113,7 @@
 
    Calculate outputs (outArgs) based the inputs (inArgs)
    
 
    
 
    Note
    
-
    : if passType==PASS_TEST, then backward() should not be called 
    
+
    : if passType==PASS_TEST, then backward() should not be called 
    
 
    
 
    
 
    
@@ -123,17 +123,17 @@
 virtual void backward(const UpdateCallback &callback = nullptr) = 0
 
    Backward propagation. 
    
 
    Calculate the gradient of inArgs and parameter.
    
-
    This function should only be called after a corresponding forward() call. The caller is responsible for filling the correct grad for the outArgs obtained using forward().
    
-
    It may also change the grad field for the inArgs supplied at forward() 
    
+
    This function should only be called after a corresponding forward() call. The caller is responsible for filling the correct grad for the outArgs obtained using forward().
    
+
    It may also change the grad field for the inArgs supplied at forward() 
    
 
    
 
 
    
 
    
 virtual void forwardBackward(const std::vector<Argument> &inArgs, std::vector<Argument> *outArgs, PassType passType, const UpdateCallback &callback = nullptr)
    
-
    Combine forward() and backward(). For multithread training, this may be faster.
    
+
    Combine forward() and backward(). For multithread training, this may be faster.
    
 
    
 
    Note
    
-
    : passType PASS_TEST is not allowed for forwardBackward(). 
    
+
    : passType PASS_TEST is not allowed for forwardBackward(). 
    
 
    
 
    
 
    
@@ -161,7 +161,7 @@
 
    
 
    
 virtual Evaluator *makeEvaluator() = 0
    
-
    Create an evaluator which can be used for eval() 
    
+
    Create an evaluator which can be used for eval() 
    
 
    
 
 
    
@@ -261,31 +261,31 @@
 
    
 
    
 GradientMachine * GradientMachine::create(const  ModelConfig  & config, int mode = kNormal, const std::vector< ParameterType > & parameterTypes = std::vector< ParameterType >{PARAMETER_VALUE, PARAMETER_GRADIENT, PARAMETER_MOMENTUM})
    
-
    Create a gradient machine from ModelConfig Parameter will have parameterTypes 
    
+
    Create a gradient machine from ModelConfig Parameter will have parameterTypes 
    
 
    
 
 
    
 
    
-GradientMachine *create(const std::string &modelFile, DataConfig *dataConfig)
    
-
    Create a gradient machine from the merged model file. The merged model file can be generated using tools/merge_model If dataConfig is not null, it will be filled with the DataConfig from the TrainerConfig 
    
+GradientMachine *create(const std::string &modelFile, DataConfig *dataConfig)
+
    Create a gradient machine from the merged model file. The merged model file can be generated using tools/merge_model If dataConfig is not null, it will be filled with the DataConfig from the TrainerConfig 
    
 
    
 
 
    
 
    
-GradientMachine *create(std::istream &is, DataConfig *dataConfig)
    
-
    Create a gradient machine from a stream which contains the merged model file. The merged model file can be generated using tools/merge_model If dataConfig is not null, it will be filled with the DataConfig from the TrainerConfig 
    
+GradientMachine *create(std::istream &is, DataConfig *dataConfig)
+
    Create a gradient machine from a stream which contains the merged model file. The merged model file can be generated using tools/merge_model If dataConfig is not null, it will be filled with the DataConfig from the TrainerConfig 
    
 
    
 
 
    
 
    
-GradientMachine *create(const std::string &modelFile, TrainerConfig *trainerConfig)
    
-
    Create a gradient machine from the merged model file. The merged model file can be generated using tools/merge_model If trainerConfig is not null, it will be filled with the TrainerConfig 
    
+GradientMachine *create(const std::string &modelFile, TrainerConfig *trainerConfig)
+
    Create a gradient machine from the merged model file. The merged model file can be generated using tools/merge_model If trainerConfig is not null, it will be filled with the TrainerConfig 
    
 
    
 
 
    
 
    
-GradientMachine *create(std::istream &is, TrainerConfig *trainerConfig)
    
-
    Create a gradient machine from a stream which contains the merged model file. The merged model file can be generated using tools/merge_model If trainerConfig is not null, it will be filled with the TrainerConfig 
    
+GradientMachine *create(std::istream &is, TrainerConfig *trainerConfig)
+
    Create a gradient machine from a stream which contains the merged model file. The merged model file can be generated using tools/merge_model If trainerConfig is not null, it will be filled with the TrainerConfig 
    
 
    
 
 
@@ -342,8 +342,8 @@
 
 
    
 
    
-virtual bool shouldBeMe(const std::string &algo, size_t trainerCount, bool isLocal, bool isGpu) const = 0
    
-
    shouldBeMe the current mode of GradientMachine should be this mode. 
    
+virtual bool shouldBeMe(const std::string &algo, size_t trainerCount, bool isLocal, bool isGpu) const = 0
+
    shouldBeMe the current mode of GradientMachine should be this mode. 
    
 
    
 
    Return
    
 
    true if mode should be this mode. 
    
@@ -365,7 +365,7 @@
 
 
    
 
    
-virtual bool isDataMustInCpu(size_t trainerCount) const = 0
    
+virtual bool isDataMustInCpu(size_t trainerCount) const = 0
 
    Is data must be in cpu even if using gpu mode. 
    
 
    
 
    Return
    
@@ -382,7 +382,7 @@
 
 
    
 
    
-virtual bool needTrainWholeDataInOneBatch() const = 0
    
+virtual bool needTrainWholeDataInOneBatch() const = 0
 
    Need not to use mini-batch method, and should train all data in one batch in one pass. 
    
 
    
 
@@ -482,18 +482,18 @@
 
    
 
    
 class paddle::MultiGradientMachine
    
-
    A MultiGradientMachine is a synchronous GradientMachine which devides one data batch into several smaller batches and assign each one small batch to one computint thread for computation. After each thread finishes computation, it merges result (including output Argument and gradient during backward()). It basically is the same as single thread gradient machine, except that it uses multi-thread to do the computation.
    
+
    A MultiGradientMachine is a synchronous GradientMachine which devides one data batch into several smaller batches and assign each one small batch to one computint thread for computation. After each thread finishes computation, it merges result (including output Argument and gradient during backward()). It basically is the same as single thread gradient machine, except that it uses multi-thread to do the computation.
    
 
    It handles GPU and Cpu parameters differently. In GPU, one computing thread generally corresponds to one GPU device. Thus, each thread keeps a separate copy of the parameter in its own device’s memory. In CPU, we only need to keep one copy of the parameters in the main memory. After, each computing thread computes its own parameter gradient, the update process needs to accumulate the parameter gradients from all the computing threads, and update the accumulated parameter gradient to the corresponding parameter value.
    
 
    Each GPU parameter is assigned to a thread called its main thread. For each parameter, the accumulation of its gradients and the update of its value happens in its main thread. The main thread first gather the parameter gradients from all the computing thread. Then, it performs parameter update. After a gradient is updated by the main thread, it is scattered to all the computing thread so that the parameters in all the computing threads are synchronized. The scatter and gather process are implemented by ring-style communication. Assume we have N computing threads, its thread ids will be 0, 1, ..., N-1. For each parameter, the id of the main thread is specified in paraMainThread_[pid], where pid is the id of the parameter. Each thread i only sends data to its partner thread (i - 1) % N. For example, for a parameter gradient that is computed in thread 4, and its main thread is 2. Its traveling process would be 4, 5,..., N-1, 0, 1, 2. In each step, the gradient buffer is added to the local gradient, and the local gradient is then copied to the gradient buffer of the next thread. At last, its main thread 2 will get the accumulated parameter gradient. For the same parameter, after its value is updated, the value’s traveling process would be 2, 1, 0, N-1, ... 3. At the end, all the computing threads would have the updated parameter value.
    
-
    A computing thread (TrainerThread) uses 4 threads to do different jobs:
    
+
    A computing thread (TrainerThread) uses 4 threads to do different jobs:
    
 
      
-
    1. computeThread(): performing forward(), backward(), prefetch().
      2. 
+
    2. computeThread(): performing forward(), backward(), prefetch().
      3. 
 
    3. valueDispatchThread(): copying parameter values to partner thread.
      4. 
 
    4. copyGradToBufferThread(): copying parameter gradient to partner thread.
      5. 
 
    5. gradCollectThread(): merging the gradient from step 3 with local gradient and call the callback supplied by the user to update parameter value.
      6. 
 
    
 
    
-
    CPU parameter value has only one copy. And their gradients are merged at the end of backward().
    
+
    CPU parameter value has only one copy. And their gradients are merged at the end of backward().
    
 
      
 
    • Handling of sparse update Currently, sparse update is only supported for CPU parameters.
      • 
 
    
@@ -504,21 +504,21 @@
 
      
 
    1. Local sparse update
      
 
      Main parameter value type is MAT_NORMAL. It is a dense matrix.
      
-
      Main parameter grad type is MAT_SPARSE_ROW_IDS (SparseRowIdsCpuMatrix) It is also a dense matrix, but the updated values are specified by IDS.
      
+
      Main parameter grad type is MAT_SPARSE_ROW_IDS (SparseRowIdsCpuMatrix) It is also a dense matrix, but the updated values are specified by IDS.
      
 
      Slave parameter value shares with main parameter value.
      
-
      Slave parameter grad type is MAT_SPARSE_ROW_AUTO_GROW (SparseAutoGrowRowCpuMatrix). It is a sparse row matrix.
      
-
      During backward() of each TrainerThread, SparseAutoGrowRowCpuMatrix will gather all the non-zero gradient. And After backward(), they will be merged into main parameter grad (SparseRowIdsCpuMatrix), with indices indicating which rows have nonzero gradient.
      
+
      Slave parameter grad type is MAT_SPARSE_ROW_AUTO_GROW (SparseAutoGrowRowCpuMatrix). It is a sparse row matrix.
      
+
      During backward() of each TrainerThread, SparseAutoGrowRowCpuMatrix will gather all the non-zero gradient. And After backward(), they will be merged into main parameter grad (SparseRowIdsCpuMatrix), with indices indicating which rows have nonzero gradient.
      
 
      2. 
 
    2. Remote sparse update
      
-
      Main parameter value type is MAT_SPARSE_ROW_PREFETCH(_FULL_SIZE) (SparsePrefetchRowCpuMatrix). MAT_SPARSE_ROW_PREFETCH is a sparse matrix. MAT_SPARSE_ROW_PREFETCH_FULL_SIZE is a dense matrix. However, only the parameter values that are prefetched is up-to-date.
      
-
      Main parameter grad type is MAT_SPARSE_ROW (SparseRowCpuMatrix). And it shares sparse pattern with value by sharing indexDictHandle_, which is an internal data structure used by SparseRowCpuMatrixto specify the sparsity pattern of Slave parameter value shares with main parameter value.
      
-
      Slave parameter grad type is MAT_SPARSE_ROW_AUTO_GROW (SparsePrefetchRowCpuMatrix). It is a sparse row matrix
      
-
      During prefetch(), all the layers will indicates which rows of each parameter are needed. Then the framework will retrieve those rows from parameter server.
      
-
      During backward() of each TrainerThread, SparseAutoGrowRowCpuMatrix will gather all the non-zero gradient. And After backward(), they will be merged into main parameter grad (SparseRowCpuMatrix). And the framework will send the merged gradient to parameter server. 
      
+
      Main parameter value type is MAT_SPARSE_ROW_PREFETCH(_FULL_SIZE) (SparsePrefetchRowCpuMatrix). MAT_SPARSE_ROW_PREFETCH is a sparse matrix. MAT_SPARSE_ROW_PREFETCH_FULL_SIZE is a dense matrix. However, only the parameter values that are prefetched is up-to-date.
      
+
      Main parameter grad type is MAT_SPARSE_ROW (SparseRowCpuMatrix). And it shares sparse pattern with value by sharing indexDictHandle_, which is an internal data structure used by SparseRowCpuMatrixto specify the sparsity pattern of Slave parameter value shares with main parameter value.
      
+
      Slave parameter grad type is MAT_SPARSE_ROW_AUTO_GROW (SparsePrefetchRowCpuMatrix). It is a sparse row matrix
      
+
      During prefetch(), all the layers will indicates which rows of each parameter are needed. Then the framework will retrieve those rows from parameter server.
      
+
      During backward() of each TrainerThread, SparseAutoGrowRowCpuMatrix will gather all the non-zero gradient. And After backward(), they will be merged into main parameter grad (SparseRowCpuMatrix). And the framework will send the merged gradient to parameter server. 
      
 
      3. 
 
    
 
    
-
    Inherits from paddle::GradientMachine
    
+
    Inherits from paddle::GradientMachine
    
 
    
 
    Public Types
    
 
    
@@ -568,7 +568,7 @@
 
    Calculate outputs (outArgs) based the inputs (inArgs)
    
 
    
 
    Note
    
-
    : if passType==PASS_TEST, then backward() should not be called 
    
+
    : if passType==PASS_TEST, then backward() should not be called 
    
 
    
 
    
 
    
@@ -578,17 +578,17 @@
 virtual void backward(const UpdateCallback &callback = nullptr)
 
    Backward propagation. 
    
 
    Calculate the gradient of inArgs and parameter.
    
-
    This function should only be called after a corresponding forward() call. The caller is responsible for filling the correct grad for the outArgs obtained using forward().
    
-
    It may also change the grad field for the inArgs supplied at forward() 
    
+
    This function should only be called after a corresponding forward() call. The caller is responsible for filling the correct grad for the outArgs obtained using forward().
    
+
    It may also change the grad field for the inArgs supplied at forward() 
    
 
    
 
 
    
 
    
 virtual void forwardBackward(const std::vector<Argument> &inArgs, std::vector<Argument> *outArgs, PassType passType, const UpdateCallback &callback)
    
-
    Combine forward() and backward(). For multithread training, this may be faster.
    
+
    Combine forward() and backward(). For multithread training, this may be faster.
    
 
    
 
    Note
    
-
    : passType PASS_TEST is not allowed for forwardBackward(). 
    
+
    : passType PASS_TEST is not allowed for forwardBackward(). 
    
 
    
 
    
 
    
@@ -612,7 +612,7 @@
 
    
 
    
 virtual Evaluator *makeEvaluator()
    
-
    Create an evaluator which can be used for eval() 
    
+
    Create an evaluator which can be used for eval() 
    
 
    
 
 
    
@@ -681,19 +681,19 @@
 
    
 
    
 void waitBeforeMerge()
    
-
    Called TrainerThread to wait before merging CPU parameter gradients. 
    
+
    Called TrainerThread to wait before merging CPU parameter gradients. 
    
 
    
 
 
    
 
    
 void waitAfterMerge()
    
-
    called by MultiGradientMachine and TrainerThread to wait after merging CPU parameter graidents. 
    
+
    called by MultiGradientMachine and TrainerThread to wait after merging CPU parameter graidents. 
    
 
    
 
 
    
 
    
 void waitForCopyInArgs()
    
-
    called by MultiGradientMachine and TrainerThread to wait for copyInArgs() finishing 
    
+
    called by MultiGradientMachine and TrainerThread to wait for copyInArgs() finishing 
    
 
    
 
 
    
@@ -714,7 +714,7 @@
 
    
 
    
 void notifyGradientTransfer(int paramId)
    
-
    Called by TrainerThread to notify MultiGradientMachine that the gradient for paramId is ready 
    
+
    Called by TrainerThread to notify MultiGradientMachine that the gradient for paramId is ready 
    
 
    
 
 
    
@@ -887,13 +887,13 @@
 
    
 
    
 ThreadBarrier allBarrier_
    
-
    barrier for both MultiGradientMachine and threds_ 
    
+
    barrier for both MultiGradientMachine and threds_ 
    
 
    
 
 
    
 
    
 bool inArgsCopied_
    
-
    indicate whether inArgs is copied before forward() 
    
+
    indicate whether inArgs is copied before forward() 
    
 
    
 
 
    
@@ -1013,7 +1013,7 @@
 
    
 
    
 void copyOutputGrad()
    
-
    copy the output gradient from the main GradientMachine. 
    
+
    copy the output gradient from the main GradientMachine. 
    
 
    
 
 
@@ -1082,7 +1082,7 @@
 
    
 
    
 void doCallback(int pid)
    
-
    call the actuall callback supplied by the caller of GradientMachine::backward 
    
+
    call the actuall callback supplied by the caller of GradientMachine::backward 
    
 
    
 
 
@@ -1247,7 +1247,7 @@
 
    
 
    
 class paddle::RecurrentGradientMachine
    
-
    Inherits from paddle::NeuralNetwork
    
+
    Inherits from paddle::NeuralNetwork
    
 
    
 
    Public Types
    
 
    
@@ -1262,7 +1262,7 @@
 
 
    
 
    
-typedef std::function<bool(int seqId, const std::vector<int>&, const std::vector<real>&)> DropCallback
    
+typedef std::function<bool(int seqId, const std::vector<int>&, const std::vector<real>&)> DropCallback
 
    DropCallback. 
    
 
    Drop a whole prefix or one candidate in beam search or not.
    
 
    The first parameter is sequence index in a batch
    
@@ -1273,7 +1273,7 @@
 
 
    
 
    
-typedef std::function<void(int seqId, const std::vector<int>&, std::vector<real>&, real*)> NormOrDropNodeCallback
    
+typedef std::function<void(int seqId, const std::vector<int>&, std::vector<real>&, real *)> NormOrDropNodeCallback
 
    NormOrDropNodeCallback. 
    
 
    Normalize a path’s probabilities or just drop it by modifying path.logProb
    
 
    The first parameter is sequence index in a batch
    
@@ -1304,7 +1304,7 @@
 
 
    
 
    
-RecurrentGradientMachine &operator=(const RecurrentGradientMachine &other)
    
+RecurrentGradientMachine &operator=(const RecurrentGradientMachine &other)
 
    
 
 
    
@@ -1330,7 +1330,7 @@
 
    Calculate outputs (outArgs) based the inputs (inArgs)
    
 
    
 
    Note
    
-
    : if passType==PASS_TEST, then backward() should not be called 
    
+
    : if passType==PASS_TEST, then backward() should not be called 
    
 
    
 
    
 
    
@@ -1340,17 +1340,17 @@
 virtual void backward(const UpdateCallback &callback = nullptr)
 
    Backward propagation. 
    
 
    Calculate the gradient of inArgs and parameter.
    
-
    This function should only be called after a corresponding forward() call. The caller is responsible for filling the correct grad for the outArgs obtained using forward().
    
-
    It may also change the grad field for the inArgs supplied at forward() 
    
+
    This function should only be called after a corresponding forward() call. The caller is responsible for filling the correct grad for the outArgs obtained using forward().
    
+
    It may also change the grad field for the inArgs supplied at forward() 
    
 
    
 
 
    
 
    
 virtual void forwardBackward(const std::vector<Argument> &inArgs, std::vector<Argument> *outArgs, PassType passType, const UpdateCallback &callback)
    
-
    Combine forward() and backward(). For multithread training, this may be faster.
    
+
    Combine forward() and backward(). For multithread training, this may be faster.
    
 
    
 
    Note
    
-
    : passType PASS_TEST is not allowed for forwardBackward(). 
    
+
    : passType PASS_TEST is not allowed for forwardBackward(). 
    
 
    
 
    
 
    
@@ -1467,8 +1467,8 @@
 
    
 
 
    
-
    
-void createInFrameInfo(const Argument &input, PassType passType)
    
+
    
+void createInFrameInfo(int inlinks_id, const Argument &input, PassType passType)
    
 
    
 
 
    
@@ -1511,12 +1511,22 @@
 
 
    
 
    
-Info info_
    
+std::vector<Info> info_
+
    
+
+
    
+
    
+std::vector<int> numSeqs_
    
 
    
 
 
    
 
    
-std::vector<std::tuple<int, int, int, int>> seqLengthAndStart_
    
+std::vector<std::vector<std::tuple<int, int, int, int>>> seqLengthAndStart_
+
    
+
+
    
+
    
+int targetInfoInlinkId_
    
 
    
 
 
    
@@ -1801,7 +1811,7 @@
 
    
 
    
 Path()
    
-
    Path default ctor, first logProb is 0. 
    
+
    Path default ctor, first logProb is 0. 
    
 
    
 
 
    
@@ -1824,7 +1834,7 @@
 
 
  • machineId - 
    sample index of a frame in RNN 
    
 
    • 
-
  • topIndex - 
    index of MaxIdLayer output in one sample 
    
+
  • topIndex - 
    index of MaxIdLayer output in one sample 
    
 
    • 
 
 
    
@@ -1834,9 +1844,9 @@
 
 
    
 
    
-bool operator<(const Path &other) const
    
+bool operator<(const Path &other) const
 
    operator < 
    
-
    Path a < Path b means log probability of a is smaller than that of b 
    
+
    Path a < Path b means log probability of a is smaller than that of b 
    
 
    
 
 
    
@@ -1916,7 +1926,7 @@
 
    A record of each node’s probality in a formed path in beam search. 
    
 
    
 
    Note
    
-
    It could be empty when history is not recorded. If the history is wanted to be recorded, recordHistory() MUST be invoked first. 
    
+
    It could be empty when history is not recorded. If the history is wanted to be recorded, recordHistory() MUST be invoked first. 
    
 
    
 
    
 
    
@@ -1926,7 +1936,7 @@
 
    Public Static Functions
    
 
    
 
    
-static bool greaterPath(const Path &a, const Path &b)
    
+static bool greaterPath(const Path &a, const Path &b)
 
    
 
 
@@ -1943,8 +1953,8 @@
 
    
 
    
 class paddle::NeuralNetwork
    
-
    Inherits from paddle::GradientMachine
    
-
    Subclassed by paddle::MultiNetwork, paddle::ParallelNeuralNetwork, paddle::RecurrentGradientMachine
    
+
    Inherits from paddle::GradientMachine
    
+
    Subclassed by paddle::MultiNetwork, paddle::ParallelNeuralNetwork, paddle::RecurrentGradientMachine
    
 
    
 
    Public Functions
    
 
    
@@ -1954,7 +1964,7 @@ virtual 
 
    
-void connect(std::string agentLayerName, NeuralNetwork *srcNN, std::string realLayerName)
    
+void connect(std::string agentLayerName, NeuralNetwork *srcNN, std::string realLayerName)
 
    
 
 
    
@@ -1970,7 +1980,7 @@ virtual 
 
    Note
    
-
    : if passType==PASS_TEST, then backward() should not be called 
    
+
    : if passType==PASS_TEST, then backward() should not be called 
    
 
    
 
    
 
    
@@ -1980,8 +1990,8 @@ virtual virtual void backward(const UpdateCallback &callback = nullptr)
 
    Backward propagation. 
    
 
    Calculate the gradient of inArgs and parameter.
    
-
    This function should only be called after a corresponding forward() call. The caller is responsible for filling the correct grad for the outArgs obtained using forward().
    
-
    It may also change the grad field for the inArgs supplied at forward() 
    
+
    This function should only be called after a corresponding forward() call. The caller is responsible for filling the correct grad for the outArgs obtained using forward().
    
+
    It may also change the grad field for the inArgs supplied at forward() 
    
 
    
 
 
    
@@ -2002,7 +2012,7 @@ virtual 
 
    
 virtual Evaluator *makeEvaluator()
    
-
    Create an evaluator which can be used for eval() 
    
+
    Create an evaluator which can be used for eval() 
    
 
    
 
 
    
@@ -2078,12 +2088,12 @@ virtual 
 
    
-NeuralNetwork *create(const ModelConfig &config)
    
+NeuralNetwork *create(const ModelConfig &config)
 
    
 
 
    
 
    
-NeuralNetwork *newNeuralNetwork(const std::string &name = "", NeuralNetwork *rootNetwork = nullptr)
    
+NeuralNetwork *newNeuralNetwork(const std::string &name = "", NeuralNetwork *rootNetwork = nullptr)
 
    
 
 
@@ -2092,7 +2102,7 @@ virtual 
 
    
 NeuralNetwork(std::string subModelName = "", NeuralNetwork *rootNetwork = nullptr)
    
-
    The constructor of NeuralNetwork. The sub networks can get parameters_ and parameterMap_ from base NeuralNetwork.
    
+
    The constructor of NeuralNetwork. The sub networks can get parameters_ and parameterMap_ from base NeuralNetwork.
    
 
    
 
    Parameters
    
 
    
 
 
    
@@ -2172,8 +2182,8 @@ virtual 
 
    
 class paddle::ParallelNeuralNetwork
    
-
    A ParallelNeuralNetwork is capable of calculating a neural network through multiple threads in parallel. 
    
-
    Inherits from paddle::NeuralNetwork
    
+
    A ParallelNeuralNetwork is capable of calculating a neural network through multiple threads in parallel. 
    
+
    Inherits from paddle::NeuralNetwork
    
 
    
 
    Public Functions
    
 
    
@@ -2193,7 +2203,7 @@ virtual 
 
    Note
    
-
    : if passType==PASS_TEST, then backward() should not be called 
    
+
    : if passType==PASS_TEST, then backward() should not be called 
    
 
    
 
    
 
    
@@ -2203,17 +2213,17 @@ virtual virtual void backward(const UpdateCallback &callback = nullptr)
 
    Backward propagation. 
    
 
    Calculate the gradient of inArgs and parameter.
    
-
    This function should only be called after a corresponding forward() call. The caller is responsible for filling the correct grad for the outArgs obtained using forward().
    
-
    It may also change the grad field for the inArgs supplied at forward() 
    
+
    This function should only be called after a corresponding forward() call. The caller is responsible for filling the correct grad for the outArgs obtained using forward().
    
+
    It may also change the grad field for the inArgs supplied at forward() 
    
 
    
 
 
    
 
    
 virtual void forwardBackward(const std::vector<Argument> &inArgs, std::vector<Argument> *outArgs, PassType passType, const UpdateCallback &callback = NULL)
    
-
    Combine forward() and backward(). For multithread training, this may be faster.
    
+
    Combine forward() and backward(). For multithread training, this may be faster.
    
 
    
 
    Note
    
-
    : passType PASS_TEST is not allowed for forwardBackward(). 
    
+
    : passType PASS_TEST is not allowed for forwardBackward(). 
    
 
    
 
    
 
    
@@ -2310,14 +2320,11 @@ virtual 
    
 
@@ -368,7 +368,7 @@ virtual 
 
    
-const MatrixPtr &getInputValue(const Layer &inputLayer)
    
+const MatrixPtr &getInputValue(const Layer &inputLayer)
 
    Get the forward-input value. 
    
 
    
 
@@ -380,13 +380,13 @@ virtual 
 
    
-const MatrixPtr &getInputGrad(const Layer &inputLayer)
    
+const MatrixPtr &getInputGrad(const Layer &inputLayer)
 
    Get the forward-input grad. 
    
 
    
 
 
    
 
    
-const IVectorPtr &getInputLabel(const Layer &inputLayer)
    
+const IVectorPtr &getInputLabel(const Layer &inputLayer)
 
    Get the forward-input label. 
    
 
    
 
@@ -432,7 +432,7 @@ virtual 
 
    
 LayerConfig config_
    
-
    Layer config. 
    
+
    Layer config. 
    
 
    
 
 
    
@@ -456,13 +456,13 @@ virtual 
 
    
 std::vector<std::string> inputArgument_
    
-
    Argument of input layers. 
    
+
    Argument of input layers. 
    
 
    
 
 
    
 
    
 std::vector<ParameterPtr> parameters_
    
-
    Parameter for each input layer. Parameters_[i] is nullptr if inputLayers_[i] does not need parameter. 
    
+
    Parameter for each input layer. Parameters_[i] is nullptr if inputLayers_[i] does not need parameter. 
    
 
    
 
 
    
@@ -539,8 +539,8 @@ virtual 
 
    
 class paddle::Projection
    
-
    A projection takes one Argument as input, calculate the result and add it to output Argument. 
    
-
    Subclassed by paddle::ContextProjection, paddle::DotMulProjection, paddle::FullMatrixProjection, paddle::IdentityOffsetProjection, paddle::IdentityProjection, paddle::TableProjection, paddle::TransposedFullMatrixProjection
    
+
    A projection takes one Argument as input, calculate the result and add it to output Argument. 
    
+
    Subclassed by paddle::ContextProjection, paddle::DotMulProjection, paddle::FullMatrixProjection, paddle::IdentityOffsetProjection, paddle::IdentityProjection, paddle::TableProjection, paddle::TransposedFullMatrixProjection
    
 
    
 
    Public Functions
    
 
    
@@ -620,7 +620,7 @@ virtual Public Static Functions
    
 
    
 
    
-Projection *create(const ProjectionConfig &config, ParameterPtr parameter, bool useGpu)
    
+Projection *create(const ProjectionConfig &config, ParameterPtr parameter, bool useGpu)
 
    
 
 
    
@@ -628,7 +628,7 @@ virtual Public Static Attributes
    
 
    
 
    
-ClassRegistrar<Projection, ProjectionConfig, ParameterPtr, bool> registrar_
    
+ClassRegistrar<Projection, ProjectionConfig, ParameterPtr, bool> registrar_
 
    Register a projection. 
    
 
    
 
@@ -644,7 +644,7 @@ virtual 
 
    
 ParameterPtr parameter_
    
-
    Parameter of projection. 
    
+
    Parameter of projection. 
    
 
    
 
 
    
@@ -679,12 +679,12 @@ virtual 
 
    
 class paddle::Operator
    
-
    Operator like Projection, but takes more than one Arguments as input. 
    
+
    Operator like Projection, but takes more than one Arguments as input. 
    
 
    Note
    
-
    : Operator can’t have parameters. 
    
+
    : Operator can’t have parameters. 
    
 
    
 
    
-
    Subclassed by paddle::ConvOperator, paddle::DotMulOperator
    
+
    Subclassed by paddle::ConvOperator, paddle::DotMulOperator
    
 
    
 
    Public Functions
    
 
    
@@ -758,7 +758,7 @@ virtual Public Static Functions
    
 
    
 
    
-Operator *create(const OperatorConfig &config, bool useGpu)
    
+Operator *create(const OperatorConfig &config, bool useGpu)
 
    
 
 
    
@@ -766,7 +766,7 @@ virtual Public Static Attributes
    
 
    
 
    
-ClassRegistrar<Operator, OperatorConfig, bool> registrar_
    
+ClassRegistrar<Operator, OperatorConfig, bool> registrar_
 
    
 
 
@@ -813,7 +813,7 @@ virtual class paddle::DataLayer
 
    This layer just copy data to output, and has no backward propagation.
    
 
    The config file api is data_layer. 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -847,7 +847,7 @@ virtual 
 
    
 virtual void copyOutputToOtherDevice()
    
-
    Copy layer’s output_ to other device. If output layer is in other device, called after Layer::forward() function. 
    
+
    Copy layer’s output_ to other device. If output layer is in other device, called after Layer::forward() function. 
    
 
    
 
 
    
@@ -871,7 +871,7 @@ virtual class paddle::FullyConnectedLayer
 
    A layer has full connections to all neurons in the previous layer. It computes an inner product with a set of learned weights, and (optionally) adds biases.
    
 
    The config file api is fc_layer. 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -910,7 +910,7 @@ virtual 
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -935,10 +935,10 @@ virtual 
 
    
 class paddle::SelectiveFullyConnectedLayer
    
-
    The SelectiveFullyConnectedLayer class. 
    
-
    SelectiveFullyConnectedLayer differs from FullyConnectedLayer by that it requires an additional input to indicate several selected columns, and only compute the multiplications between the input matrices and the selected columns of the parameter matrices of this layer. If the selected columns is not specified, SelectiveFullyConnected layer acts exactly like FullyConnectedLayer.
    
+
    The SelectiveFullyConnectedLayer class. 
    
+
    SelectiveFullyConnectedLayer differs from FullyConnectedLayer by that it requires an additional input to indicate several selected columns, and only compute the multiplications between the input matrices and the selected columns of the parameter matrices of this layer. If the selected columns is not specified, SelectiveFullyConnected layer acts exactly like FullyConnectedLayer.
    
 
    The config file api is selective_fc_layer. 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -1000,7 +1000,7 @@ virtual 
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -1028,9 +1028,9 @@ virtual 
 
    
 class paddle::ConvBaseLayer
    
-
    A Base Convolution Layer, which convolves the input image with learned filters and (optionally) adds biases. 
    
-
    Inherits from paddle::Layer
    
-
    Subclassed by paddle::CudnnConvLayer, paddle::ExpandConvLayer
    
+
    A Base Convolution Layer, which convolves the input image with learned filters and (optionally) adds biases. 
    
+
    Inherits from paddle::Layer
    
+
    Subclassed by paddle::CudnnConvLayer, paddle::ExpandConvLayer
    
 
    
 
    Public Functions
    
 
    
@@ -1205,9 +1205,9 @@ virtual 
 
    
 class paddle::ConvOperator
    
-
    ConvOperator takes two inputs to perform the convolution. The first input is the image, and the second input is the convolution kernel. The height of data for two inputs are the same. Each data of the first input is convolved with each data of the second input indepedently. 
    
+
    ConvOperator takes two inputs to perform the convolution. The first input is the image, and the second input is the convolution kernel. The height of data for two inputs are the same. Each data of the first input is convolved with each data of the second input indepedently. 
    
 
    The config file api is conv_operator. 
    
-
    Inherits from paddle::Operator
    
+
    Inherits from paddle::Operator
    
 
    
 
    Public Functions
    
 
    
@@ -1263,7 +1263,7 @@ virtual paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -1291,7 +1291,7 @@ virtual 
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -1303,9 +1303,9 @@ virtual 
 
    
 class paddle::CudnnConvLayer
    
-
    A subclass of ConvBaseLayer by cuDNN implementation. It only supports GPU mode. We automatic select CudnnConvLayer for GPU mode and ExpandConvLayer for CPU mode if you set type of “conv”. User also can specfiy type of “exconv” or “cudnn_conv” for particular type. 
    
+
    A subclass of ConvBaseLayer by cuDNN implementation. It only supports GPU mode. We automatic select CudnnConvLayer for GPU mode and ExpandConvLayer for CPU mode if you set type of “conv”. User also can specfiy type of “exconv” or “cudnn_conv” for particular type. 
    
 
    The config file api is img_conv_layer. 
    
-
    Inherits from paddle::ConvBaseLayer
    
+
    Inherits from paddle::ConvBaseLayer
    
 
    
 
    Public Functions
    
 
    
@@ -1339,7 +1339,7 @@ virtual 
 
    
 virtual void backward(const UpdateCallback &callback)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -1506,7 +1506,7 @@ virtual class paddle::ExpandConvLayer
 
    A subclass of convolution layer. This layer expands input and use matrix multiplication to calculate convolution operation. 
    
 
    The config file api is img_conv_layer. 
    
-
    Inherits from paddle::ConvBaseLayer
    
+
    Inherits from paddle::ConvBaseLayer
    
 
    
 
    Public Functions
    
 
    
@@ -1585,7 +1585,7 @@ virtual 
 
    
 virtual void backward(const UpdateCallback &callback)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -1666,19 +1666,19 @@ virtual class paddle::ContextProjection
 
    Context projection concatenate features in adjacent time steps in a sequence. The i-th row of the output is the concatenation of context_length rows of the input. The context_length rows are the consecutive rows from the i+shift_start row. 
    
 
    For example, assumed input (x) has 4 words and the dimension of each word representation is 2. If we use zero to pad instead of learned weight to pad, and the context_lenth is 3, the output (y) is:
    
-
    x = [a1, a2;
-     b1, b2;
-     c1, c2;
-     d1, d2]
-y = [0,  0,  a1, a2, b1, b2;
-     a1, a2, b1, b2, c1, c2;
-     b1, b2, c1, c2, d1, d2;
-     c1, c2, d1, d2, 0,  0]
+
    x = [a1, a2;
+     b1, b2;
+     c1, c2;
+     d1, d2]
+y = [0,  0,  a1, a2, b1, b2;
+     a1, a2, b1, b2, c1, c2;
+     b1, b2, c1, c2, d1, d2;
+     c1, c2, d1, d2, 0,  0]
 
    
 
    
 
    
 
    The config file api is context_projection. 
    
-
    Inherits from paddle::Projection
    
+
    Inherits from paddle::Projection
    
 
    
 
    Public Functions
    
 
    
@@ -1759,8 +1759,8 @@ y = [0,  0,  a1, a2, b1, b2;
 
    
 class paddle::PoolLayer
    
 
    Basic parent layer of pooling Pools the input within regions. 
    
-
    Inherits from paddle::Layer
    
-
    Subclassed by paddle::CudnnPoolLayer, paddle::PoolProjectionLayer
    
+
    Inherits from paddle::Layer
    
+
    Subclassed by paddle::CudnnPoolLayer, paddle::PoolProjectionLayer
    
 
    
 
    Public Functions
    
 
    
@@ -1861,7 +1861,7 @@ y = [0,  0,  a1, a2, b1, b2;
 
    
 class paddle::PoolProjectionLayer
    
 
    Basic parent layer of different kinds of pooling. 
    
-
    Inherits from paddle::PoolLayer
    
+
    Inherits from paddle::PoolLayer
    
 
    Subclassed by paddle::AvgPoolProjectionLayer, paddle::MaxPoolProjectionLayer
    
 
    
 
    Public Functions
    
@@ -1907,9 +1907,9 @@ y = [0,  0,  a1, a2, b1, b2;
 
    
 
    
 class paddle::CudnnPoolLayer
    
-
    CudnnPoolLayer is subclass of PoolLayer, which is implemented by cudnn api and only supports GPU. 
    
+
    CudnnPoolLayer is subclass of PoolLayer, which is implemented by cudnn api and only supports GPU. 
    
 
    The config file api is img_pool_layer. 
    
-
    Inherits from paddle::PoolLayer
    
+
    Inherits from paddle::PoolLayer
    
 
    
 
    Public Functions
    
 
    
@@ -1943,7 +1943,7 @@ y = [0,  0,  a1, a2, b1, b2;
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -2055,8 +2055,8 @@ y = [0,  0,  a1, a2, b1, b2;
 
    Normalize the input in local region 
    
 
    
 
    
-
    Inherits from paddle::Layer
    
-
    Subclassed by paddle::ResponseNormLayer
    
+
    Inherits from paddle::Layer
    
+
    Subclassed by paddle::ResponseNormLayer
    
 
    
 
    Public Functions
    
 
    
@@ -2089,7 +2089,7 @@ y = [0,  0,  a1, a2, b1, b2;
 
    
 class paddle::CMRProjectionNormLayer
    
 
    response normalization across feature maps namely normalize in number of size_ channels 
    
-
    Inherits from paddle::ResponseNormLayer
    
+
    Inherits from paddle::ResponseNormLayer
    
 
    
 
    Public Functions
    
 
    
@@ -2122,17 +2122,9 @@ y = [0,  0,  a1, a2, b1, b2;
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
-
    
-
    
-
    Protected Attributes
    
-
    
-
    
-bool blocked_
    
-
    
-
 
    
 
    
 
@@ -2144,7 +2136,7 @@ y = [0,  0,  a1, a2, b1, b2;
 class paddle::DataNormLayer
 
    A layer for data normalization. 
    
 
      
-
    • Input: One and only one input layer is accepted. The input layer must be DataLayer with dense data type.
      • 
+
    • Input: One and only one input layer is accepted. The input layer must be DataLayer with dense data type.
      • 
 
    • Output: The normalization of the input data
      • 
 
    
 
    
@@ -2155,7 +2147,7 @@ y = [0,  0,  a1, a2, b1, b2;
 
  • decimal-scaling: y = x/10^j, where j is the smallest integer such that max(|y|)<1 
    • 
 
 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Types
    
 
    
@@ -2207,7 +2199,7 @@ y = [0,  0,  a1, a2, b1, b2;
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -2258,8 +2250,8 @@ y = [0,  0,  a1, a2, b1, b2;
 
    
 class paddle::ResponseNormLayer
    
 
    response normalization within feature maps namely normalize in independent channel When code refactoring, we delete the original implementation. Need to implement in the futrue. 
    
-
    Inherits from paddle::NormLayer
    
-
    Subclassed by paddle::CMRProjectionNormLayer
    
+
    Inherits from paddle::NormLayer
    
+
    Subclassed by paddle::CMRProjectionNormLayer
    
 
    
 
    Public Functions
    
 
    
@@ -2282,7 +2274,7 @@ y = [0,  0,  a1, a2, b1, b2;
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -2334,11 +2326,11 @@ y = [0,  0,  a1, a2, b1, b2;
 class paddle::BatchNormBaseLayer
 
    Batch normalization layer use to normalizes the input to across the batch. 
    
 
    By default, calculating global mean and variance statistics via a running average in the training peroid. Then the pre-calculated global mean and variance are used for testing.
    
-
    Moving mean and variance are located in Parameter object when constructing and the calculation will change them. Now we only save global mean and variance of one thread in first node for GPU. But the calculation in CPU is different, because parameters are shared by multiple threads. Here using ShareCpuMatrix with lock to calculate. We still save global mean and variance in first node in CPU when multi machine.
    
+
    Moving mean and variance are located in Parameter object when constructing and the calculation will change them. Now we only save global mean and variance of one thread in first node for GPU. But the calculation in CPU is different, because parameters are shared by multiple threads. Here using ShareCpuMatrix with lock to calculate. We still save global mean and variance in first node in CPU when multi machine.
    
 
    [1] S. Ioffe and C. Szegedy, “Batch Normalization: Accelerating Deep Network
     Training by Reducing Internal Covariate Shift.” arXiv preprint arXiv:1502.03167 (2015). 
    
-
    Inherits from paddle::Layer
    
-
    Subclassed by paddle::BatchNormalizationLayer, paddle::CudnnBatchNormLayer
    
+
    Inherits from paddle::Layer
    
+
    Subclassed by paddle::BatchNormalizationLayer, paddle::CudnnBatchNormLayer
    
 
    
 
    Public Functions
    
 
    
@@ -2459,7 +2451,7 @@ y = [0,  0,  a1, a2, b1, b2;
 class paddle::BatchNormalizationLayer
 
    A Inheritance class of Batch normalization layer. It supports both CPU and GPU. 
    
 
    The config file api is batch_norm_layer. 
    
-
    Inherits from paddle::BatchNormBaseLayer
    
+
    Inherits from paddle::BatchNormBaseLayer
    
 
    
 
    Public Functions
    
 
    
@@ -2487,7 +2479,7 @@ y = [0,  0,  a1, a2, b1, b2;
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -2514,13 +2506,13 @@ y = [0,  0,  a1, a2, b1, b2;
 
    
 
    
 void expandMat(const MatrixPtr &in, MatrixPtr &out)
    
-
    expand a Matrix from batch, channels* imagePixels to batch * ImagePixels * channels. 
    
+
    expand a Matrix from batch, channels* imagePixels to batch * ImagePixels * channels. 
    
 
    
 
 
    
 
    
 void shrinkMat(const MatrixPtr &in, MatrixPtr &out)
    
-
    Shrink a Matrix from from batch * ImagePixels * channels to batch, channels* imagePixels. 
    
+
    Shrink a Matrix from from batch * ImagePixels * channels to batch, channels* imagePixels. 
    
 
    
 
 
    
@@ -2612,7 +2604,7 @@ The config file api is batch_norm_layer. 
    
 
    Cudnn version must >= v4.0, and better to use the latest version (v5.1).
    
 
    
 
    
-
    Inherits from paddle::BatchNormBaseLayer
    
+
    Inherits from paddle::BatchNormBaseLayer
    
 
    
 
    Public Functions
    
 
    
@@ -2646,7 +2638,7 @@ The config file api is batch_norm_layer. 
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -2697,7 +2689,7 @@ The config file api is batch_norm_layer. 
    
 \[ out[i] = \frac {in[i]} {\sum_{k=1}^N in[k]} \]
    
  where \(in\) is a (batchSize x dataDim) input vector, and \(out\) is a (batchSize x dataDim) output vector.
    
 
    The config file api is sum_to_one_norm_layer. 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -2725,7 +2717,7 @@ The config file api is batch_norm_layer. 
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -2755,12 +2747,12 @@ The config file api is batch_norm_layer. 
    
 
    
 
    
 class paddle::ParameterReluLayer
    
-
    ParameterReluLayer active inputs with learnable parameter weight_. forward: 
    
-\[\begin{split} y = x > 0 ? x : w .* x \end{split}\]
    
+
    ParameterReluLayer active inputs with learnable parameter weight_. forward: 
    
+\[ y = x > 0 ? x : w .* x \]
    
  backward: 
    
 \[\begin{split} dx = x > 0 ? dy : w .* dy \\ dw = x > 0 ? 0 : dy.*x \end{split}\]
    
  Here, x is the input, w is the weight, y is the output. dx, dw, dy is the gradient. 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -2788,7 +2780,7 @@ The config file api is batch_norm_layer. 
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -2823,12 +2815,12 @@ The config file api is batch_norm_layer. 
    
 
    
 
    
 class paddle::RecurrentLayer
    
-
    RecurrentLayer takes 1 input layer. The output size is the same with input layer. For each sequence [start, end] it performs the following computation: 
    
+
    RecurrentLayer takes 1 input layer. The output size is the same with input layer. For each sequence [start, end] it performs the following computation: 
    
 \[\begin{split} out_{i} = act(in_{i}) \ \ \text{for} \ i = start \\ out_{i} = act(in_{i} + out_{i-1} * W) \ \ \text{for} \ start < i <= end \end{split}\]
    
  If reversed is true, the order is reversed: 
    
 \[\begin{split} out_{i} = act(in_{i}) \ \ \text{for} \ i = end \\ out_{i} = act(in_{i} + out_{i+1} * W) \ \ \text{for} \ start <= i < end \end{split}\]
    
  There are two methods to calculate rnn. One way is to compute rnn one sequence by one sequence. The other way is to reorganize the input into batches, then compute rnn one batch by one batch. Users can select them by rnn_use_batch flag. 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -2851,14 +2843,14 @@ The config file api is batch_norm_layer. 
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
 
    
 virtual void resetState()
    
-
    Reset the internal state variables. Allocate them if they have not been allocated. This function need to called before Layer::forward() for generating sequence.
    
-
    This is used for sequence generation. When generating sequence, the calculation at current timestamp depends on the state from previous timestamp. The model needs to keep the information about the previous timestamp in the state variables. Layers such as RecurrentLayer, LstmLayer and ContextLayer have state variables. 
    
+
    Reset the internal state variables. Allocate them if they have not been allocated. This function need to called before Layer::forward() for generating sequence.
    
+
    This is used for sequence generation. When generating sequence, the calculation at current timestamp depends on the state from previous timestamp. The model needs to keep the information about the previous timestamp in the state variables. Layers such as RecurrentLayer, LstmLayer and ContextLayer have state variables. 
    
 
    
 
 
    
@@ -2955,7 +2947,7 @@ The config file api is batch_norm_layer. 
    
 
    
 
    
 void forwardBatch(int batchSize, size_t numSequences, const int *starts)
    
-
    Reorganize input into batches and compute rnn forward batch by batch. It will convert batch shape to sequence after finishing forward. The batch info can refer to SequenceToBatch class. 
    
+
    Reorganize input into batches and compute rnn forward batch by batch. It will convert batch shape to sequence after finishing forward. The batch info can refer to SequenceToBatch class. 
    
 
    
 
    Parameters
    
 
      
@@ -3110,7 +3102,7 @@ The config file api is batch_norm_layer. 
      
 
 
      
 
      
-void shareIndexWith(const SequenceToBatch &seq2batch)
      
+void shareIndexWith(const SequenceToBatch &seq2batch)
 
      
 
 
    
@@ -3190,17 +3182,17 @@ The config file api is batch_norm_layer. 
    
 
    
 
    
 class paddle::LstmLayer
    
-
    LstmLayer takes 1 input layer with size * 4. Input layer is diveded into 4 equal parts: (input_s, input_ig, input_fg, input_og) 
    
-
    For each sequence [start, end] it performs the following computation: 
    output_{i} = actState(state_{i}) * actGate(outputGate_{i})
-state_{i} = actInput(input_s_{i} + bias_s +
-            output_{i-1} * recurrIW) * actGate(inputGate_{i}) +
-            actGate(forgetGate_{i}) * state_{i-1}
-inputGate = input_ig_{i} + bias_ig + output_{i-1} * recurrIGW +
-            state_{i-1} * inputCheck
-ouputGate = input_og_{i} + bias_og + output_{i-1} * recurrOGW +
-            state_{i} * outputCheck
-forgetGate = input_fg_{i} + bias_fg + output_{i-1} * recurrFGW +
-             state_{i-1} * forgetCheck
+
    LstmLayer takes 1 input layer with size * 4. Input layer is diveded into 4 equal parts: (input_s, input_ig, input_fg, input_og) 
    
+
    For each sequence [start, end] it performs the following computation: 
    output_{i} = actState(state_{i}) * actGate(outputGate_{i})
+state_{i} = actInput(input_s_{i} + bias_s +
+            output_{i-1} * recurrIW) * actGate(inputGate_{i}) +
+            actGate(forgetGate_{i}) * state_{i-1}
+inputGate = input_ig_{i} + bias_ig + output_{i-1} * recurrIGW +
+            state_{i-1} * inputCheck
+ouputGate = input_og_{i} + bias_og + output_{i-1} * recurrOGW +
+            state_{i} * outputCheck
+forgetGate = input_fg_{i} + bias_fg + output_{i-1} * recurrFGW +
+             state_{i-1} * forgetCheck
 
    
 
    
 
    
@@ -3219,11 +3211,11 @@ forgetGate = input_fg_{i} + bias_fg + output_{i-1} * recurrFGW +
 
    
 The weight ([size, 4*size]) contains \(W_{hi}, W_{hf}, W_{hc}, W_{ho}\). The bias contains \(b_i, b_f, b_c, b_o\) and \(W_{ci}, W_{cf}, W_{co}\). 
    
 
    Note
    
-
    These \(W_{xi}x_{t}, W_{xf}x_{t}, W_{xc}x_{t}, W_{xo}x_{t}\) operations on the input sequence were NOT included in LstmLayer. So users should use fc_layer or mixed_layer before lstm_later.
    
+
    These \(W_{xi}x_{t}, W_{xf}x_{t}, W_{xc}x_{t}, W_{xo}x_{t}\) operations on the input sequence were NOT included in LstmLayer. So users should use fc_layer or mixed_layer before lstm_later.
    
 
    
 
    
-
    Inherits from paddle::Layer, paddle::LstmCompute
    
-
    Subclassed by paddle::MDLstmLayer
    
+
    Inherits from paddle::Layer, paddle::LstmCompute
    
+
    Subclassed by paddle::MDLstmLayer
    
 
    
 
    Public Functions
    
 
    
@@ -3246,14 +3238,14 @@ The weight ([size, 4*size]) contains \(W_{hi}, W_{hf}, W_{hc}
 
    
 
    
 virtual void backward(const UpdateCallback &callback)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
 
    
 virtual void resetState()
    
-
    Reset the internal state variables. Allocate them if they have not been allocated. This function need to called before Layer::forward() for generating sequence.
    
-
    This is used for sequence generation. When generating sequence, the calculation at current timestamp depends on the state from previous timestamp. The model needs to keep the information about the previous timestamp in the state variables. Layers such as RecurrentLayer, LstmLayer and ContextLayer have state variables. 
    
+
    Reset the internal state variables. Allocate them if they have not been allocated. This function need to called before Layer::forward() for generating sequence.
    
+
    This is used for sequence generation. When generating sequence, the calculation at current timestamp depends on the state from previous timestamp. The model needs to keep the information about the previous timestamp in the state variables. Layers such as RecurrentLayer, LstmLayer and ContextLayer have state variables. 
    
 
    
 
 
    
@@ -3305,10 +3297,10 @@ The weight ([size, 4*size]) contains \(W_{hi}, W_{hf}, W_{hc}
 
    
 
    
 void forwardBatch(int batchSize, size_t numSequences, const int *starts, MatrixPtr inputValue)
    
-
    Compute lstm forward one batch by one batch. The batch value is reorganized by SequenceToBatch class. The batch output value will be convert into sequence value after finishing forward. Here, one batch contains one word of each sample. If the length of each sample is not equality, the batch will not pads zero and contains less words. The total batch numbers are the max length of the sequence. The details can refer to SequenceToBatch class. On GPU mode, it will launch GPU kernel for loop.
    
-
     
    for (int i = 0; i < numBatch(max_sequence_length); ++i) {
-  compute one batch.
-}
+
    Compute lstm forward one batch by one batch. The batch value is reorganized by SequenceToBatch class. The batch output value will be convert into sequence value after finishing forward. Here, one batch contains one word of each sample. If the length of each sample is not equality, the batch will not pads zero and contains less words. The total batch numbers are the max length of the sequence. The details can refer to SequenceToBatch class. On GPU mode, it will launch GPU kernel for loop.
    
+
     
    for (int i = 0; i < numBatch(max_sequence_length); ++i) {
+  compute one batch.
+}
 
    
 
    
 
    
@@ -3487,7 +3479,7 @@ The weight ([size, 4*size]) contains \(W_{hi}, W_{hf}, W_{hc}
 
    
 
    
 class paddle::LstmStepLayer
    
-
    Inherits from paddle::Layer, paddle::LstmCompute
    
+
    Inherits from paddle::Layer, paddle::LstmCompute
    
 
    
 
    Public Functions
    
 
    
@@ -3515,7 +3507,7 @@ The weight ([size, 4*size]) contains \(W_{hi}, W_{hf}, W_{hc}
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -3580,7 +3572,7 @@ The weight ([size, 4*size]) contains \(W_{hi}, W_{hf}, W_{hc}
 
    
 
    
 class paddle::LstmCompute
    
-
    Subclassed by paddle::LstmLayer, paddle::LstmStepLayer
    
+
    Subclassed by paddle::LstmLayer, paddle::LstmStepLayer
    
 
    
 
    Public Functions
    
 
    
@@ -3593,7 +3585,7 @@ The weight ([size, 4*size]) contains \(W_{hi}, W_{hf}, W_{hc}
 template <bool useGpu>
 
    
 void forwardBatch(hl_lstm_value value, int frameSize, int batchSize)
    
-
    LstmLayer batch compute API (forwardBatch, backwardBatch). If use batch compute api, lstm value(and grad) need to be batch structure. Compute order: forwardBatch: for 0 <= id < numBatch backwardBatch: for numBatch > id >= 0 
    
+
    LstmLayer batch compute API (forwardBatch, backwardBatch). If use batch compute api, lstm value(and grad) need to be batch structure. Compute order: forwardBatch: for 0 <= id < numBatch backwardBatch: for numBatch > id >= 0 
    
 
    
 
 
    
@@ -3608,7 +3600,7 @@ The weight ([size, 4*size]) contains \(W_{hi}, W_{hf}, W_{hc}
 template <bool useGpu>
 
    
 void forwardOneSequence(hl_lstm_value value, int frameSize)
    
-
    LstmLayer sequence compute API (forwardOneSequence, backwardOneSequence). Compute order(for each sequence): forwardOneSequence: if (!reversed) for 0 <= seqId < seqLength if (reversed) for seqLength > seqId >= 0 backwardOneSequence: if (!reversed) for seqLength > seqId >= 0 if (reversed) for 0 <= seqId < seqLength 
    
+
    LstmLayer sequence compute API (forwardOneSequence, backwardOneSequence). Compute order(for each sequence): forwardOneSequence: if (!reversed) for 0 <= seqId < seqLength if (reversed) for seqLength > seqId >= 0 backwardOneSequence: if (!reversed) for seqLength > seqId >= 0 if (reversed) for 0 <= seqId < seqLength 
    
 
    
 
 
    
@@ -3622,28 +3614,56 @@ The weight ([size, 4*size]) contains \(W_{hi}, W_{hf}, W_{hc}
 
    
 template <>
    
 
    
-void forwardOneSequence(hl_lstm_value value, int frameSize)
    
+void forwardOneSequence(hl_lstm_value value, int frameSize)
 
    
 
 
    
 
    
 template <>
    
 
    
-void backwardOneSequence(hl_lstm_value value, hl_lstm_grad grad, int frameSize)
    
+void backwardOneSequence(hl_lstm_value value, hl_lstm_grad grad, int frameSize)
 
    
 
 
    
 
    
 template <>
    
 
    
-void forwardBatch(hl_lstm_value value, int frameSize, int batchSize)
    
+void forwardBatch(hl_lstm_value value, int frameSize, int batchSize)
 
    
 
 
    
 
    
 template <>
    
 
    
-void backwardBatch(hl_lstm_value value, hl_lstm_grad grad, int frameSize, int batchSize)
    
+void backwardBatch(hl_lstm_value value, hl_lstm_grad grad, int frameSize, int batchSize)
+
    
+
+
    
+
    
+template <>
    
+
    
+void forwardBatch(hl_lstm_value value, int frameSize, int batchSize)
    
+
    
+
+
    
+
    
+template <>
    
+
    
+void backwardBatch(hl_lstm_value value, hl_lstm_grad grad, int frameSize, int batchSize)
    
+
    
+
+
    
+
    
+template <>
    
+
    
+void forwardOneSequence(hl_lstm_value value, int frameSize)
    
+
    
+
+
    
+
    
+template <>
    
+
    
+void backwardOneSequence(hl_lstm_value value, hl_lstm_grad grad, int frameSize)
    
 
    
 
 
    
@@ -3676,7 +3696,7 @@ The weight ([size, 4*size]) contains \(W_{hi}, W_{hf}, W_{hc}
 
    
 
    
 class paddle::MDLstmLayer
    
-
    Inherits from paddle::LstmLayer
    
+
    Inherits from paddle::LstmLayer
    
 
    
 
    Public Functions
    
 
    
@@ -3699,7 +3719,7 @@ The weight ([size, 4*size]) contains \(W_{hi}, W_{hf}, W_{hc}
 
    
 
    
 virtual void backward(const UpdateCallback &callback)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -3826,12 +3846,12 @@ The weight ([size, 4*size]) contains \(W_{hi}, W_{hf}, W_{hc}
 
 
    
 
    
-CoordIterator &operator++()
    
+CoordIterator &operator++()
 
    
 
 
    
 
    
-CoordIterator &operator--()
    
+CoordIterator &operator--()
 
    
 
 
    
@@ -3911,7 +3931,7 @@ The weight ([size, 4*size]) contains \(W_{hi}, W_{hf}, W_{hc}
 class paddle::GatedRecurrentLayer
 
    Please refer to “Junyoung Chung, Empirical Evaluation
 of Gated Recurrent Neural Networks on Sequence Modeling”. 
    
-
    GatedRecurrentLayer takes 1 input layer with size * 3. Input layer is diveded into 3 equal parts: (xz_t, xr_t, xi_t). parameter and biasParameter is also diveded into 3 equal parts:
      
+
      GatedRecurrentLayer takes 1 input layer with size * 3. Input layer is diveded into 3 equal parts: (xz_t, xr_t, xi_t). parameter and biasParameter is also diveded into 3 equal parts:
        
 
      • parameter consists of (U_z, U_r, U)
        • 
 
      • baisParameter consists of (bias_z, bias_r, bias_o)
        • 
 
      
@@ -3930,7 +3950,7 @@ The config file is grumemory. 
      
 
    
 
    
 
    
-
    Inherits from paddle::Layer, paddle::GruCompute
    
+
    Inherits from paddle::Layer, paddle::GruCompute
    
 
    
 
    Public Functions
    
 
    
@@ -3953,14 +3973,14 @@ The config file is grumemory. 
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
 
    
 virtual void resetState()
    
-
    Reset the internal state variables. Allocate them if they have not been allocated. This function need to called before Layer::forward() for generating sequence.
    
-
    This is used for sequence generation. When generating sequence, the calculation at current timestamp depends on the state from previous timestamp. The model needs to keep the information about the previous timestamp in the state variables. Layers such as RecurrentLayer, LstmLayer and ContextLayer have state variables. 
    
+
    Reset the internal state variables. Allocate them if they have not been allocated. This function need to called before Layer::forward() for generating sequence.
    
+
    This is used for sequence generation. When generating sequence, the calculation at current timestamp depends on the state from previous timestamp. The model needs to keep the information about the previous timestamp in the state variables. Layers such as RecurrentLayer, LstmLayer and ContextLayer have state variables. 
    
 
    
 
 
    
@@ -4074,7 +4094,7 @@ The config file is grumemory. 
    
 
    
 
    
 class paddle::GruStepLayer
    
-
    GruStepLayer is like GatedRecurrentLayer, but used in recurrent layer group. GruStepLayer takes 2 input layer. 
    
+
    GruStepLayer is like GatedRecurrentLayer, but used in recurrent layer group. GruStepLayer takes 2 input layer. 
    
 
      
 
    • input[0] with size * 3 and diveded into 3 equal parts: (xz_t, xr_t, xi_t).
      • 
 
    • input[1] with size: {prev_out}.
      • 
@@ -4101,7 +4121,7 @@ The config file api if gru_step_layer. 
      
 
    
 
    
 
    
-
    Inherits from paddle::Layer, paddle::GruCompute
    
+
    Inherits from paddle::Layer, paddle::GruCompute
    
 
    
 
    Public Functions
    
 
    
@@ -4129,7 +4149,7 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -4164,7 +4184,7 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 class paddle::GruCompute
    
-
    Subclassed by paddle::GatedRecurrentLayer, paddle::GruStepLayer
    
+
    Subclassed by paddle::GatedRecurrentLayer, paddle::GruStepLayer
    
 
    
 
    Public Functions
    
 
    
@@ -4190,14 +4210,28 @@ The config file api if gru_step_layer. 
    
 
    
 template <>
    
 
    
-void forward(hl_gru_value value, int frameSize, int batchSize)
    
+void forward(hl_gru_value value, int frameSize, int batchSize)
 
    
 
 
    
 
    
 template <>
    
 
    
-void backward(hl_gru_value value, hl_gru_grad grad, int frameSize, int batchSize)
    
+void backward(hl_gru_value value, hl_gru_grad grad, int frameSize, int batchSize)
+
    
+
+
    
+
    
+template <>
    
+
    
+void forward(hl_gru_value value, int frameSize, int batchSize)
    
+
    
+
+
    
+
    
+template <>
    
+
    
+void backward(hl_gru_value value, hl_gru_grad grad, int frameSize, int batchSize)
    
 
    
 
 
    
@@ -4226,9 +4260,9 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 class paddle::AgentLayer
    
-
    AgentLayer use as a virtual input of another layer in config, before execute forward/backward, setRealLayer() should be called to set one and only one real layer 
    
-
    Inherits from paddle::Layer
    
-
    Subclassed by paddle::SequenceAgentLayer
    
+
    AgentLayer use as a virtual input of another layer in config, before execute forward/backward, setRealLayer() should be called to set one and only one real layer 
    
+
    Inherits from paddle::Layer
    
+
    Subclassed by paddle::SequenceAgentLayer
    
 
    
 
    Public Functions
    
 
    
@@ -4261,7 +4295,7 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -4286,8 +4320,8 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 class paddle::SequenceAgentLayer
    
-
    like AgentLayer, but use first numSamples sequences 
    
-
    Inherits from paddle::AgentLayer
    
+
    like AgentLayer, but use first numSamples sequences 
    
+
    Inherits from paddle::AgentLayer
    
 
    
 
    Public Functions
    
 
    
@@ -4309,7 +4343,7 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -4321,9 +4355,9 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 class paddle::GatherAgentLayer
    
-
    Like AgentLayer, but it can gather many real layers. Each real layer give a few rows of a sequence, after gather all real layers, GatherAgentLayer collect a complete sequence. 
    
-
    Inherits from paddle::Layer
    
-
    Subclassed by paddle::SequenceGatherAgentLayer
    
+
    Like AgentLayer, but it can gather many real layers. Each real layer give a few rows of a sequence, after gather all real layers, GatherAgentLayer collect a complete sequence. 
    
+
    Inherits from paddle::Layer
    
+
    Subclassed by paddle::SequenceGatherAgentLayer
    
 
    
 
    Public Functions
    
 
    
@@ -4361,7 +4395,7 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -4396,8 +4430,8 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 class paddle::SequenceGatherAgentLayer
    
-
    Like GatherAgentLayer, but select a few sequence in real layer. ids in addRealLayer() are the ids of selected sequence. It’s used to reorder sequence output. 
    
-
    Inherits from paddle::GatherAgentLayer
    
+
    Like GatherAgentLayer, but select a few sequence in real layer. ids in addRealLayer() are the ids of selected sequence. It’s used to reorder sequence output. 
    
+
    Inherits from paddle::GatherAgentLayer
    
 
    
 
    Public Functions
    
 
    
@@ -4419,7 +4453,7 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -4431,9 +4465,9 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 class paddle::ScatterAgentLayer
    
-
    Like AgentLayer, but only select a few rows in real layer. [idIndex, idIndex + idSize) of ids in setRealLayerAndOutput() are the selected row ids. It’s used to scatter one layer’s output to many small submodels. ScatterAgentLayer can support ids real layer, if it is, the agent will select a few ids in real layer. 
    
-
    Inherits from paddle::Layer
    
-
    Subclassed by paddle::SequenceScatterAgentLayer
    
+
    Like AgentLayer, but only select a few rows in real layer. [idIndex, idIndex + idSize) of ids in setRealLayerAndOutput() are the selected row ids. It’s used to scatter one layer’s output to many small submodels. ScatterAgentLayer can support ids real layer, if it is, the agent will select a few ids in real layer. 
    
+
    Inherits from paddle::Layer
    
+
    Subclassed by paddle::SequenceScatterAgentLayer
    
 
    
 
    Public Functions
    
 
    
@@ -4463,7 +4497,7 @@ The config file api if gru_step_layer. 
    
 
 
  • ids[input] - 
    row id in real layer 
    
 
    • 
-
  • copyId[input] - 
    whether to copy a cpu version of ids, false(default) in ScatterAgentLayer, and true in SequenceScatterAgentLayer. 
    
+
  • copyId[input] - 
    whether to copy a cpu version of ids, false(default) in ScatterAgentLayer, and true in SequenceScatterAgentLayer. 
    
 
    • 
 
 
    
@@ -4490,7 +4524,7 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -4545,8 +4579,8 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 class paddle::SequenceScatterAgentLayer
    
-
    Like ScatterAgentLayer, but select a few sequence in real layer. ids in setRealLayer() or setRealLayerAndOutput() are the ids of selected sequence. It’s used to reorder sequence input. 
    
-
    Inherits from paddle::ScatterAgentLayer
    
+
    Like ScatterAgentLayer, but select a few sequence in real layer. ids in setRealLayer() or setRealLayerAndOutput() are the ids of selected sequence. It’s used to reorder sequence input. 
    
+
    Inherits from paddle::ScatterAgentLayer
    
 
    
 
    Public Functions
    
 
    
@@ -4568,7 +4602,7 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -4593,7 +4627,7 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 class paddle::GetOutputLayer
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -4621,7 +4655,7 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -4634,9 +4668,9 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 class paddle::MixedLayer
    
-
    A mixed layer has multiple input layers. Each input layer was processed by a Projection or Operator. The results of all projections or Operators are summed together with bias (if configured), and then go through an activation function and dropout (if configured).
    
+
    A mixed layer has multiple input layers. Each input layer was processed by a Projection or Operator. The results of all projections or Operators are summed together with bias (if configured), and then go through an activation function and dropout (if configured).
    
 
    The config file api is mixed_layer. 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -4670,20 +4704,20 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
 
    
 virtual void resetState()
    
-
    Reset the internal state variables. Allocate them if they have not been allocated. This function need to called before Layer::forward() for generating sequence.
    
-
    This is used for sequence generation. When generating sequence, the calculation at current timestamp depends on the state from previous timestamp. The model needs to keep the information about the previous timestamp in the state variables. Layers such as RecurrentLayer, LstmLayer and ContextLayer have state variables. 
    
+
    Reset the internal state variables. Allocate them if they have not been allocated. This function need to called before Layer::forward() for generating sequence.
    
+
    This is used for sequence generation. When generating sequence, the calculation at current timestamp depends on the state from previous timestamp. The model needs to keep the information about the previous timestamp in the state variables. Layers such as RecurrentLayer, LstmLayer and ContextLayer have state variables. 
    
 
    
 
 
    
 
    
 virtual void setState(LayerStatePtr state)
    
-
    setState() should be called after getState(). Argument state consists of all projections states. 
    
+
    setState() should be called after getState(). Argument state consists of all projections states. 
    
 
    
 
 
    
@@ -4724,11 +4758,11 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 class paddle::DotMulProjection
    
-
    DotMulProjection performs element-wise multiplication with weight: 
    
+
    DotMulProjection performs element-wise multiplication with weight: 
    
 \[ out.row[i] += in.row[i] .* weight \]
    
  where \(.*\) means element-wise multiplication.
    
 
    The config file api is dotmul_projection. 
    
-
    Inherits from paddle::Projection
    
+
    Inherits from paddle::Projection
    
 
    
 
    Public Functions
    
 
    
@@ -4764,11 +4798,11 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 class paddle::DotMulOperator
    
-
    DotMulOperator takes two inputs, performs element-wise multiplication: 
    
+
    DotMulOperator takes two inputs, performs element-wise multiplication: 
    
 \[ out.row[i] += scale * (in1.row[i] .* in2.row[i]) \]
    
  where \(.*\) means element-wise multiplication, and scale is a config scalar, its default value is one.
    
 
    The config file api is dotmul_operator. 
    
-
    Inherits from paddle::Operator
    
+
    Inherits from paddle::Operator
    
 
    
 
    Public Functions
    
 
    
@@ -4795,11 +4829,11 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 class paddle::FullMatrixProjection
    
-
    FullMatrixProjection performs full matrix multiplication: 
    
+
    FullMatrixProjection performs full matrix multiplication: 
    
 \[ out.row[i] += in.row[i] * weight \]
    
 
    
 
    The config file api is full_matrix_projection. 
    
-
    Inherits from paddle::Projection
    
+
    Inherits from paddle::Projection
    
 
    
 
    Public Functions
    
 
    
@@ -4834,11 +4868,11 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 class paddle::IdentityProjection
    
-
    IdentityProjection performs addition: 
    
+
    IdentityProjection performs addition: 
    
 \[ out.row[i] += in.row[i] \]
    
 
    
 
    The config file api is identity_projection. 
    
-
    Inherits from paddle::Projection
    
+
    Inherits from paddle::Projection
    
 
    
 
    Public Functions
    
 
    
@@ -4846,7 +4880,7 @@ The config file api if gru_step_layer. 
    
 IdentityProjection(const ProjectionConfig &config, const ParameterPtr &parameter, bool useGpu)
 
    Constructed function. 
    
 
    Note
    
-
    IdentityProjection should not have any parameter. 
    
+
    IdentityProjection should not have any parameter. 
    
 
    
 
    
 
    
@@ -4870,11 +4904,11 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 class paddle::IdentityOffsetProjection
    
-
    IdentityOffsetProjection likes IdentityProjection, but layer size may be smaller than input size. It selects dimensions [offset, offset+layer_size) from input to perform addition: 
    
+
    IdentityOffsetProjection likes IdentityProjection, but layer size may be smaller than input size. It selects dimensions [offset, offset+layer_size) from input to perform addition: 
    
 \[ out.row[i] += in.row[i + \textrm{offset}] \]
    
 
    
 
    The config file api is identity_projection. 
    
-
    Inherits from paddle::Projection
    
+
    Inherits from paddle::Projection
    
 
    
 
    Public Functions
    
 
    
@@ -4882,7 +4916,7 @@ The config file api if gru_step_layer. 
    
 IdentityOffsetProjection(const ProjectionConfig &config, const ParameterPtr &parameter, bool useGpu)
 
    Constructed function. 
    
 
    Note
    
-
    IdentityOffsetProjection should not have any parameter. 
    
+
    IdentityOffsetProjection should not have any parameter. 
    
 
    
 
    
 
    
@@ -4915,7 +4949,7 @@ The config file api if gru_step_layer. 
    
 
    If \(ids[i] = -1\), it will be ignored. 
    
 
    
 
    
-
    Inherits from paddle::Projection
    
+
    Inherits from paddle::Projection
    
 
    
 
    Public Functions
    
 
    
@@ -4956,9 +4990,9 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 class paddle::TransposedFullMatrixProjection
    
-
    TransposedFullMatrixProjection performs full matrix multiplication: out.row[i] += in.row[i] * weight.transpose. 
    
+
    TransposedFullMatrixProjection performs full matrix multiplication: out.row[i] += in.row[i] * weight.transpose. 
    
 
    The config file api is trans_full_matrix_projection. 
    
-
    Inherits from paddle::Projection
    
+
    Inherits from paddle::Projection
    
 
    
 
    Public Functions
    
 
    
@@ -4999,7 +5033,7 @@ The config file api if gru_step_layer. 
    
 
    
 class paddle::AverageLayer
    
 
    A layer for “internal average” for sequence input. Input: one or more sequences. Each sequence contains some instances. If AverageLevel = kNonSeq: Output: output size is the number of input sequences (NOT input instances) output[i] = average_{for each instance in this sequence}{input[i]} If AverageLevel = kSeq: Check input sequence must has sub-sequence Output: output size is the number of input sub-sequences output[i] = average_{for each instance in this sub-sequence}{input[i]} 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Types
    
 
    
@@ -5067,7 +5101,7 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -5108,7 +5142,7 @@ The config file api if gru_step_layer. 
    
 
    
 class paddle::MaxLayer
    
 
    A layer for “internal max” for sequence input. Input: one or more sequences. Each sequence contains some instances. If MaxLevel = kNonSeq: Output: output size is the number of input sequences (NOT input instances) output[i] = max_{for each instance in this sequence}{input[i]} If MaxLevel = kSeq: Check input sequence must has sub-sequence Output: output size is the number of input sub-sequences output[i] = max_{for each instance in this sub-sequence}{input[i]} 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Types
    
 
    
@@ -5155,7 +5189,7 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -5186,7 +5220,7 @@ The config file api if gru_step_layer. 
    
 
    
 class paddle::SequenceLastInstanceLayer
    
 
    A layer for extracting the last instance of the input sequence. Input: a sequence If SequenceLevel = kNonseq: Output: a sequence containing only the last instance of the input sequence If SequenceLevel = kSeq: Check input sequence must has sub-sequence Output: a sequence containing only the last instance of each sub-sequence of the input sequence 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -5214,7 +5248,7 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -5272,7 +5306,7 @@ The config file api if gru_step_layer. 
    
 
    
 class paddle::ConcatenateLayer
    
 
    A concatenate layer has multiple input layers. It concatenates rows of each input as one row for the output of this layer and apply activation. 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -5300,7 +5334,7 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -5312,8 +5346,8 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 class paddle::ConcatenateLayer2
    
-
    concat2 layer is like concat layer, but each input layer was processed by a Projection. 
    
-
    Inherits from paddle::Layer
    
+
    concat2 layer is like concat layer, but each input layer was processed by a Projection. 
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -5341,7 +5375,7 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -5372,7 +5406,7 @@ The config file api if gru_step_layer. 
    
 
    
 class paddle::SequenceConcatLayer
    
 
    A layer for concatenating the first sequence with the second sequence following the first Input: two sequences each containing some instances Output: a concatenated sequence of the two input sequences 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -5400,7 +5434,7 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -5424,7 +5458,7 @@ The config file api if gru_step_layer. 
    
 
    
 class paddle::SubSequenceLayer
    
 
    A layer for taking the subsequence according to given offset and size Input: original sequence, offset, size Output: subsequence 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -5452,7 +5486,7 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -5497,9 +5531,9 @@ The config file api if gru_step_layer. 
    
 
 
 
    
-
    The expand method is the same with ExpandConvLayer, but saved the transposed value. After expanding, output_.sequenceStartPositions will store timeline. The number of time steps are outputH_ * outputW_ and the dimension of each time step is blockH_ * blockW_ * channels_. This layer can be used after convolution neural network, and before recurrent neural network.
    
+
    The expand method is the same with ExpandConvLayer, but saved the transposed value. After expanding, output_.sequenceStartPositions will store timeline. The number of time steps are outputH_ * outputW_ and the dimension of each time step is blockH_ * blockW_ * channels_. This layer can be used after convolution neural network, and before recurrent neural network.
    
 
    The config file api is block_expand_layer. 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -5527,7 +5561,7 @@ The config file api if gru_step_layer. 
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -5633,7 +5667,7 @@ sequence is one) to sequence data.”
    
 
    
 
    And the output size is the batch size(not instances) of second input.
    
 
    The config file api is expand_layer. 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -5661,7 +5695,7 @@ sequence is one) to sequence data.”
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -5724,16 +5758,16 @@ sequence is one) to sequence data.”
    
 
  • Input: Input one should be dense image data.
    • 
 
  • Output: expanded fature maps. 
    
 \[ y.row[i] = x.row[i \mod x.width], i = 0,1,..., (x.width * num\_filters - 1) \]
    
- For example, num_filters = 4:  
    x = [a1,a2;
-     b1,b2]
-y = [a1, a2, a1, a2, a1, a2, a1, a2;
-     b1, b2, b1, b2, b1, b2, b1, b2;]
+ For example, num_filters = 4:  
    x = [a1,a2;
+     b1,b2]
+y = [a1, a2, a1, a2, a1, a2, a1, a2;
+     b1, b2, b1, b2, b1, b2, b1, b2;]
 
    
 
    
 
    • 
 
 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -5761,7 +5795,7 @@ y = [a1, a2, a1, a2, a1, a2, a1, a2;
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -5779,7 +5813,7 @@ y = [a1, a2, a1, a2, a1, a2, a1, a2;
 
    origin matrix height * witdth) resize matrix: (height * width / size) * size 
    
 
    
 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -5802,7 +5836,7 @@ y = [a1, a2, a1, a2, a1, a2, a1, a2;
 
    
 
    
 virtual void backward(const UpdateCallback &callback)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -5815,7 +5849,7 @@ y = [a1, a2, a1, a2, a1, a2, a1, a2;
 
    
 class paddle::SequenceReshapeLayer
    
 
    A layer for reshaping the sequence Input: a sequence Output: a sequence 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -5843,7 +5877,7 @@ y = [a1, a2, a1, a2, a1, a2, a1, a2;
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -5875,7 +5909,7 @@ y = [a1, a2, a1, a2, a1, a2, a1, a2;
 \[ y=f(\sum_{i}x_i + b) \]
    
  where \(y\) is output, \(x\) is input, \(b\) is bias, and \(f\) is activation function.
    
 
    The config file api is addto_layer. 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -5891,7 +5925,7 @@ y = [a1, a2, a1, a2, a1, a2, a1, a2;
 
    
 
    
 virtual bool init(const LayerMap &layerMap, const ParameterMap &parameterMap)
    
-
    Intialization of AddtoLayer. 
    
+
    Intialization of AddtoLayer. 
    
 
    
 
 
    
@@ -5927,16 +5961,16 @@ y = [a1, a2, a1, a2, a1, a2, a1, a2;
 
    
 
    
 class paddle::ConvexCombinationLayer
    
-
    A layer for convex weighted average of vectors, which is used in NEURAL MACHINE TRANSLATION BY JOINTLY LEARNING TO ALIGN AND TRANSLATE. 
    
+
    A layer for weighted sum of vectors, which is used in NEURAL MACHINE TRANSLATION BY JOINTLY LEARNING TO ALIGN AND TRANSLATE. 
    
 
      
-
    • Input: the first input contains the convex weights (batchSize x weightDim), and the shape of second input is (batchSize x (weightdim*dataDim)).
      • 
-
    • Output: the shape of output is (batchSize x dataDim). 
      
-\[ out[i][j] = \sum_{j}(in0(i, j) * in1(i,j + i * dataDim)), i = 0,1,...,(batchSize-1); j = 0, 1,...,(dataDim-1) \]
      
-
      • 
+
    • Input: the the size of the first input is weightDim, and the size of the second input is weightdim * dataDim.
      • 
+
    • Output: the sizeof the output is dataDim 
      
+\[ out(j) = \sum_{i}(in0(i) * in1(i,j + i * dataDim)), i = 0,1,...,(weightDim-1); j = 0, 1,...,(dataDim-1) \]
      
+ Note that the above computation is for one sample. Multiple samples are processed in one batch.
      • 
 
    
 
    
-
    The config file api is convex_comb_layer. 
    
-
    Inherits from paddle::Layer
    
+
    The config file api is linear_comb_layer. 
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -5964,7 +5998,7 @@ y = [a1, a2, a1, a2, a1, a2, a1, a2;
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -6001,7 +6035,7 @@ y = [a1, a2, a1, a2, a1, a2, a1, a2;
 \[ y.row[i] = w[i] * x_1.row[i] + (1 - w[i]) * x_2.row[i] \]
    
  where \(x_1\) and \(x_2\) are two (batchSize x dataDim) inputs, \(w\) is (batchSize x 1) weight vector, and \(y\) is (batchSize x dataDim) output.
    
 
    The config file api is interpolation_layer. 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -6029,7 +6063,7 @@ y = [a1, a2, a1, a2, a1, a2, a1, a2;
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -6065,7 +6099,7 @@ y = [a1, a2, a1, a2, a1, a2, a1, a2;
 
    
 \[ y[i][j] = x_{x_{0}[i] + 1}[i][j], j = 0,1, ... , (x_{1}.width - 1) \]
    
  where, y is output. \(x_{k}\) is the k-th input layer and \(k = x_{0}[i] + 1\). 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -6093,7 +6127,7 @@ y = [a1, a2, a1, a2, a1, a2, a1, a2;
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -6102,7 +6136,7 @@ y = [a1, a2, a1, a2, a1, a2, a1, a2;
 
    
 
    
 std::vector<CopyInfo> copySchedule_
    
-
    A list of CopyInfo used to save copy information. 
    
+
    A list of CopyInfo used to save copy information. 
    
 
    
 
 
    
@@ -6167,7 +6201,7 @@ y = [a1, a2, a1, a2, a1, a2, a1, a2;
 
    used in NEURAL TURING MACHINE Input1: vector (batchSize * dim1) Input2: vector (batchSize * dim2) Output: a matrix: (batchSize * (dim1*dim2)) 
    
 
    
 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -6195,7 +6229,7 @@ y = [a1, a2, a1, a2, a1, a2, a1, a2;
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -6229,7 +6263,7 @@ y = [a1, a2, a1, a2, a1, a2, a1, a2;
 \[ y = x^w \]
    
  where \(x\) is a input vector, \(w\) is scalar weight, and output \(y\) is a vector.
    
 
    The config file api is power_layer. 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -6257,7 +6291,7 @@ y = [a1, a2, a1, a2, a1, a2, a1, a2;
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -6281,7 +6315,7 @@ y = [a1, a2, a1, a2, a1, a2, a1, a2;
 \[ y.row[i] = w[i] * x.row[i] \]
    
  where \(x\) is (batchSize x dataDim) input, \(w\) is (batchSize x 1) weight vector, and \(y\) is (batchSize x dataDim) output.
    
 
    The config file api is scaling_layer. 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -6309,7 +6343,7 @@ y = [a1, a2, a1, a2, a1, a2, a1, a2;
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -6331,7 +6365,7 @@ y = [a1, a2, a1, a2, a1, a2, a1, a2;
 
    
 
    Here, a is scale and b is offset, which are provided as attributes of the layer.
    
 
    The config file api is slope_intercept_layer. 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -6359,7 +6393,7 @@ y = [a1, a2, a1, a2, a1, a2, a1, a2;
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -6371,7 +6405,7 @@ y = [a1, a2, a1, a2, a1, a2, a1, a2;
 
    
 
    
 class paddle::TensorLayer
    
-
    TensorLayer takes two input vectors. 
    
+
    TensorLayer takes two input vectors. 
    
 \[ y_{i} = x_{1} * W_{i} * x_{2}^{\rm T}, i=0, 1, ...,K-1 \]
    
 . 
    
 
      
@@ -6384,7 +6418,7 @@ y = [a1, a2, a1, a2, a1, a2, a1, a2;
 
    
 
    
 
    The config file api is tensor_layer. 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -6417,7 +6451,7 @@ y = [a1, a2, a1, a2, a1, a2, a1, a2;
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -6446,7 +6480,7 @@ y = [a1, a2, a1, a2, a1, a2, a1, a2;
 \[ y = x^\mathrm{T} \]
    
  where \(x\) is (M x N) input, and \(y\) is (N x M) output.
    
 
    The config file api is trans_layer. 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -6469,7 +6503,7 @@ y = [a1, a2, a1, a2, a1, a2, a1, a2;
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -6582,7 +6616,7 @@ The space requirement is O(N)=O(N * sizeof(Interval)). The computational complex
 class paddle::MaxIdLayer
 
    A layer for finding the id which has the maximal value for each sample. The result is stored in output_.ids.
    
 
    The config file api is maxid_layer. 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -6605,7 +6639,7 @@ The space requirement is O(N)=O(N * sizeof(Interval)). The computational complex
 
    
 
    
 virtual void backward(const UpdateCallback &callback)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -6619,7 +6653,7 @@ The space requirement is O(N)=O(N * sizeof(Interval)). The computational complex
 class paddle::SamplingIdLayer
 
    A layer for sampling id from multinomial distribution from the input layer. Sampling one id for one sample. The result is stored in output_.ids. 
    
 
    The config file api is sampling_id_layer. 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -6647,7 +6681,7 @@ The space requirement is O(N)=O(N * sizeof(Interval)). The computational complex
 
    
 
    
 virtual void backward(const UpdateCallback &callback)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -6663,8 +6697,8 @@ The space requirement is O(N)=O(N * sizeof(Interval)). The computational complex
 
    
 class paddle::CostLayer
    
 
    Base class for a particular type of cost layer. This type of cost should have one data layer, one label layer and an optional weight layer as input. The derived class should implemnt forwardImp() and backwardImp() which calculate the cost for data and label. The weight is automatically handled by the base class. 
    
-
    Inherits from paddle::Layer
    
-
    Subclassed by paddle::HuberTwoClass, paddle::MultiBinaryLabelCrossEntropy, paddle::MultiClassCrossEntropy, paddle::MultiClassCrossEntropyWithSelfNorm, paddle::SoftBinaryClassCrossEntropy, paddle::SumOfSquaresCostLayer
    
+
    Inherits from paddle::Layer
    
+
    Subclassed by paddle::HuberTwoClass, paddle::MultiBinaryLabelCrossEntropy, paddle::MultiClassCrossEntropy, paddle::MultiClassCrossEntropyWithSelfNorm, paddle::SoftBinaryClassCrossEntropy, paddle::SumOfSquaresCostLayer
    
 
    
 
    Public Functions
    
 
    
@@ -6697,7 +6731,7 @@ The space requirement is O(N)=O(N * sizeof(Interval)). The computational complex
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -6735,7 +6769,7 @@ The space requirement is O(N)=O(N * sizeof(Interval)). The computational complex
 
    For label={0, 1}, let y=2*label-1. Given output f, the loss is: 
    
 \[\begin{split} Loss = \left\{\begin{matrix} 4 * y * f & \textit{if} \ \ y* f < -1 \\ (1 - y * f)^2 & \textit{if} \ \ -1 < y * f < 1 \\ 0 & \textit{otherwise} \end{matrix}\right. \end{split}\]
    
  
    
-
    Inherits from paddle::CostLayer
    
+
    Inherits from paddle::CostLayer
    
 
    
 
    Public Functions
    
 
    
@@ -6783,7 +6817,7 @@ The space requirement is O(N)=O(N * sizeof(Interval)). The computational complex
 \[ \lambda_{ij} = \frac{1}{1 + e^{o_i - o_j}} \left| \Delta_{NDCG} \right| \]
    
 
    
 
    [1] Christopher J.C. Burges, Robert Ragno, Quoc Viet Le. Learning to Rank with Nonsmooth Cost Functions. 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -6816,7 +6850,7 @@ The space requirement is O(N)=O(N * sizeof(Interval)). The computational complex
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -6847,7 +6881,7 @@ The space requirement is O(N)=O(N * sizeof(Interval)). The computational complex
 
    Cross entropy for multi binary labels. 
    
 \[ cost[i] = -sum(label[i][j]*log(output[i][j]) + (1-label[i][j])*log(1-output[i][j])) \]
    
  
    
-
    Inherits from paddle::CostLayer
    
+
    Inherits from paddle::CostLayer
    
 
    
 
    Public Functions
    
 
    
@@ -6892,7 +6926,7 @@ The space requirement is O(N)=O(N * sizeof(Interval)). The computational complex
 
    
 \[ L = - \sum_{i}{t_{k} * log(P(y=k))} \]
    
  
    
-
    Inherits from paddle::CostLayer
    
+
    Inherits from paddle::CostLayer
    
 
    
 
    Public Functions
    
 
    
@@ -6931,7 +6965,7 @@ The space requirement is O(N)=O(N * sizeof(Interval)). The computational complex
 
    
 
    The \(Z(x)\) is the softmax normalizer.
    
 
    [1] Jacob Devlin, Rabih Zbib, Zhongqiang Huang, Thomas Lamar, Richard Schwartz, and John Makhoul. Fast and robust neural network joint models for statistical machine translation. In Proceedings of the ACL 2014 Conference. 
    
-
    Inherits from paddle::CostLayer
    
+
    Inherits from paddle::CostLayer
    
 
    
 
    Public Functions
    
 
    
@@ -6981,7 +7015,7 @@ The space requirement is O(N)=O(N * sizeof(Interval)). The computational complex
 \[\begin{split} C_{i,j} = -\tilde{P_{ij}} * o_{i,j} + log(1 + e^{o_{i,j}}) \\ o_{i,j} = o_i - o_j \\ \tilde{P_{i,j}} = \left \{0, 0.5, 1 \right \} \ or \ \left \{0, 1 \right \} \end{split}\]
    
 
    
 
    [1]. Chris Burges, Tal Shaked, Erin Renshaw, et al. Learning to Rank useing Gradient Descent. 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -7014,7 +7048,7 @@ The space requirement is O(N)=O(N * sizeof(Interval)). The computational complex
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -7045,7 +7079,7 @@ The space requirement is O(N)=O(N * sizeof(Interval)). The computational complex
 
    The cross-entropy for soft binary class. 
    
 \[ L = \sum_i (\sum_j -y_j(i)*log(x_j(i))-(1-y_j(i))*log(1-x_j(i))) \]
    
  
    
-
    Inherits from paddle::CostLayer
    
+
    Inherits from paddle::CostLayer
    
 
    
 
    Public Functions
    
 
    
@@ -7089,7 +7123,7 @@ The space requirement is O(N)=O(N * sizeof(Interval)). The computational complex
 
    This cost layer compute Euclidean (L2) loss for real-valued regression tasks. 
    
 \[ L = \frac{1}{2N} \sum_{i=1}^N {|| \hat{y}_i - y_i||_2^2} \]
    
  
    
-
    Inherits from paddle::CostLayer
    
+
    Inherits from paddle::CostLayer
    
 
    
 
    Public Functions
    
 
    
@@ -7133,7 +7167,7 @@ The space requirement is O(N)=O(N * sizeof(Interval)). The computational complex
 
 
    
 
    The config file api is cos_sim. 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -7161,17 +7195,9 @@ The space requirement is O(N)=O(N * sizeof(Interval)). The computational complex
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
-
    
-
    
-
    Public Members
    
-
    
-
    
-const real kCosSimScale_
    
-
    
-
 
    
 
    
 
@@ -7190,7 +7216,7 @@ Input1: a vector (batchSize * dataDim)
    
 
    
 
    Input2: a matrix in vector form (batchSize * (weightDim*dataDim))
    
 
    Output: a vector (batchSize * weightDim) 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -7218,7 +7244,7 @@ Input1: a vector (batchSize * dataDim)
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -7264,7 +7290,7 @@ Input1: a vector (batchSize * dataDim)
    
 
    
 class paddle::CRFDecodingLayer
    
 
    A layer for calculating the decoding sequence of sequential conditional random field model. The decoding sequence is stored in output_.ids It also calculate error, output_.value[i] is 1 for incorrect decoding or 0 for correct decoding) See LinearChainCRF.h for the detail of the CRF formulation. 
    
-
    Inherits from paddle::CRFLayer
    
+
    Inherits from paddle::CRFLayer
    
 
    
 
    Public Functions
    
 
    
@@ -7287,7 +7313,7 @@ Input1: a vector (batchSize * dataDim)
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -7308,8 +7334,8 @@ Input1: a vector (batchSize * dataDim)
    
 
    
 class paddle::CRFLayer
    
 
    A layer for calculating the cost of sequential conditional random field model. See LinearChainCRF.h for the detail of the CRF formulation. 
    
-
    Inherits from paddle::Layer
    
-
    Subclassed by paddle::CRFDecodingLayer
    
+
    Inherits from paddle::Layer
    
+
    Subclassed by paddle::CRFDecodingLayer
    
 
    
 
    Public Functions
    
 
    
@@ -7329,22 +7355,12 @@ Input1: a vector (batchSize * dataDim)
    
 
    Forward propagation. All inherited implementation should call Layer::foward() function. 
    
 
    
 
-
    
-
    
-void forwardImp(const Argument &output, const Argument &label, VectorPtr parameterValue, VectorPtr parameterGradient)
    
-
    
-
 
    
 
    
 virtual void backward(const UpdateCallback &callback)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
-
    
-
    
-void backwardImp(const UpdateCallback &callback, const Argument &output, const Argument &label)
    
-
    
-
 
    
 
    
 
    Protected Attributes
    
@@ -7373,11 +7389,6 @@ Input1: a vector (batchSize * dataDim)
    
 real coeff_
 
    
 
-
    
-
    
-std::vector<Argument> tmpCpuInput_
    
-
    
-
 
    
 
    
 
@@ -7387,7 +7398,7 @@ Input1: a vector (batchSize * dataDim)
    
 
    
 
    
 class paddle::CTCLayer
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -7415,7 +7426,7 @@ Input1: a vector (batchSize * dataDim)
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -7458,14 +7469,14 @@ Input1: a vector (batchSize * dataDim)
    
 
    Organize the classes into a binary tree. At each node, a sigmoid function is used to calculate the probability of belonging to the right branch. This idea is from “F. Morin, Y. Bengio (AISTATS 05): 
 Hierarchical Probabilistic Neural Network Language Model.”
    
 
    Here we uses a simple way of making the binary tree. Assuming the number of classes C = 6, The classes are organized as a binary tree in the following way:
    
-
    *-*-*- 2
-| | |- 3
-| |
-| |-*- 4
-|   |- 5
-|
-|-*- 0
-|- 1
+
    *-*-*- 2
+| | |- 3
+| |
+| |-*- 4
+|   |- 5
+|
+|-*- 0
+|- 1
 
    
 
    
 
    
@@ -7488,7 +7499,7 @@ Hierarchical Probabilistic Neural Network Language Model.”
    
 
 
    
 
    The config file api is hsigmod_layer. 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -7511,7 +7522,7 @@ Hierarchical Probabilistic Neural Network Language Model.”
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -7760,7 +7771,7 @@ Hierarchical Probabilistic Neural Network Language Model.”
    
 
    
 class paddle::NCELayer
    
 
    Noise-contrastive estimation Implements the method in the following paper: A fast and simple algorithm for training neural probabilistic language models 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -7794,7 +7805,7 @@ Hierarchical Probabilistic Neural Network Language Model.”
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -7838,8 +7849,8 @@ Hierarchical Probabilistic Neural Network Language Model.”
    
 
    
 
    
 class paddle::ValidationLayer
    
-
    Inherits from paddle::Layer
    
-
    Subclassed by paddle::AucValidation, paddle::PnpairValidation
    
+
    Inherits from paddle::Layer
    
+
    Subclassed by paddle::AucValidation, paddle::PnpairValidation
    
 
    
 
    Public Functions
    
 
    
@@ -7877,7 +7888,7 @@ Hierarchical Probabilistic Neural Network Language Model.”
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback = nullptr)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -7900,7 +7911,7 @@ Hierarchical Probabilistic Neural Network Language Model.”
    
 
    
 
    
 class paddle::AucValidation
    
-
    Inherits from paddle::ValidationLayer
    
+
    Inherits from paddle::ValidationLayer
    
 
    
 
    Public Functions
    
 
    
@@ -7968,7 +7979,7 @@ Hierarchical Probabilistic Neural Network Language Model.”
    
 
    
 
    
 class paddle::PnpairValidation
    
-
    Inherits from paddle::ValidationLayer
    
+
    Inherits from paddle::ValidationLayer
    
 
    
 
    Public Functions
    
 
    
@@ -8011,7 +8022,7 @@ Hierarchical Probabilistic Neural Network Language Model.”
    
 
 
    
 
    The result is stored in output_.ids. It is used by recurrent layer group. 
    
-
    Inherits from paddle::Layer
    
+
    Inherits from paddle::Layer
    
 
    
 
    Public Functions
    
 
    
@@ -8034,7 +8045,7 @@ Hierarchical Probabilistic Neural Network Language Model.”
    
 
    
 
    
 virtual void backward(const UpdateCallback &callback)
    
-
    Backward propagation. Should only be called after Layer::forward() function. 
    
+
    Backward propagation. Should only be called after Layer::forward() function. 
    
 
    
 
 
    
@@ -8232,14 +8243,11 @@ Hierarchical Probabilistic Neural Network Language Model.”
    
 
 
         
    
@@ -8261,14 +8269,14 @@ Hierarchical Probabilistic Neural Network Language Model.”
    
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Layer Documents »
    • 
-          
  • Layers Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Layer Documents »
    • 
+          
  • Layers Documents »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/source/index.html b/doc/source/index.html
index a556b5d219..8c1a75f6a4 100644
--- a/doc/source/index.html
+++ b/doc/source/index.html
@@ -6,7 +6,7 @@
   
     
     
-    Source Code Documents — PaddlePaddle  documentation
+    Source Code Documents — PaddlePaddle  documentation
     
     
     
@@ -44,7 +44,7 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
      
 
@@ -174,14 +174,11 @@
 
 
         
    
@@ -203,12 +200,12 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/source/math/matrix/index.html b/doc/source/math/matrix/index.html
index 413b9b6b20..150659c856 100644
--- a/doc/source/math/matrix/index.html
+++ b/doc/source/math/matrix/index.html
@@ -6,7 +6,7 @@
   
     
     
-    Matrix Documents — PaddlePaddle  documentation
+    Matrix Documents — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
      
 
@@ -91,14 +91,11 @@
 
 
         
    
@@ -120,13 +117,13 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/source/math/matrix/matrix.html b/doc/source/math/matrix/matrix.html
index b6491472af..5ceddde40a 100644
--- a/doc/source/math/matrix/matrix.html
+++ b/doc/source/math/matrix/matrix.html
@@ -6,7 +6,7 @@
   
     
     
-    Matrix — PaddlePaddle  documentation
+    Matrix — PaddlePaddle  documentation
     
     
     
@@ -45,9 +45,9 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    • 
-          
  • Matrix Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    • 
+          
  • Matrix Documents »
    •  
       
     
      
 
@@ -96,12 +96,83 @@
 
    
 
 
    
+
    
+
    
+template <class T, T v>
    
+
    
+struct bool_constant
    
+
    
+
    Public Static Attributes
    
+
    
+
    
+const T value
    
+
    
+
+
    
+
    
+
+
    
+
    
+class MatrixOffset
    
+
    
+
    Public Functions
    
+
    
+
    
+MatrixOffset(size_t aCol = 0, size_t aRow = 0, size_t bCol = 0, size_t bRow = 0, size_t cCol = 0, size_t cRow = 0, size_t dCol = 0, size_t dRow = 0)
    
+
    
+
+
    
+
    
+
    Public Members
    
+
    
+
    
+size_t aCol_
    
+
    
+
+
    
+
    
+size_t aRow_
    
+
    
+
+
    
+
    
+size_t bCol_
    
+
    
+
+
    
+
    
+size_t bRow_
    
+
    
+
+
    
+
    
+size_t cCol_
    
+
    
+
+
    
+
    
+size_t cRow_
    
+
    
+
+
    
+
    
+size_t dCol_
    
+
    
+
+
    
+
    
+size_t dRow_
    
+
    
+
+
    
+
    
+
 
    
 
    
 template <class T>
    
 
    
 class BaseMatrixT
    
-
    Subclassed by paddle::BaseVector< T >, paddle::Matrix
    
+
    Subclassed by paddle::BaseVector< T >, paddle::Matrix
    
 
    
 
    Public Functions
    
 
    
@@ -141,7 +212,7 @@
 
    
 int applyUnary(Op op)
    
 
    unary operator: element wise op(a).
    
-
     
    for 0 <= i < this->height_ & for 0 <= j < this->width_.
+
     
    for 0 <= i < this->height_ & for 0 <= j < this->width_.
 
    
 
    
 
    
@@ -153,9 +224,9 @@
 
    
 int applyUnary(Op op, int numRows, int numCols, MatrixOffset &offset)
    
 
    unary operator: element wise op(a).
    
-
     
    for 0 <= i < numRows & for 0 <= j < numCols.
-While matrix start address is:
- A = this->data_ + offset.aRow_*ld + offset.aCol_;
+
     
    for 0 <= i < numRows & for 0 <= j < numCols.
+While matrix start address is:
+ A = this->data_ + offset.aRow_*ld + offset.aCol_;
 
    
 
    
 
    
@@ -165,10 +236,10 @@ While matrix start address is:
 
    
 template <class Op>
    
 
    
-int applyBinary(Op op, BaseMatrixT &b)
    
+int applyBinary(Op op, BaseMatrixT &b)
 
    binary operator: element wise op(a, b).
    
-
     
    for 0 <= i < this->height_ & for 0 <= j < this->width_.
-While this->height_ == b.height_ && this->width_ == b.width_.
+
     
    for 0 <= i < this->height_ & for 0 <= j < this->width_.
+While this->height_ == b.height_ && this->width_ == b.width_.
 
    
 
    
 
    
@@ -178,24 +249,24 @@ While this->height_ == b.height_ && this->width_ == b.width_.
 
    
 template <class Op, class bAsRowVector, class bAsColVector>
    
 
    
-int applyBinary(Op op, BaseMatrixT &b, int numRows, int numCols, MatrixOffset &offset, bAsRowVector, bAsColVector)
    
+int applyBinary(Op op, BaseMatrixT &b, int numRows, int numCols, MatrixOffset &offset, bAsRowVector, bAsColVector)
 
    binary operator: element wise op(a, b)
    
-
     
    for 0 <= i < numRows & for 0 <= j < numCols.
-While matrix start address is:
-  A = this->data_ + offset.aRow_*lda + offset.aCol_;
-  B = b->data_ + offset.bRow_*ldb + offset.bCol_;
+
     
    for 0 <= i < numRows & for 0 <= j < numCols.
+While matrix start address is:
+  A = this->data_ + offset.aRow_*lda + offset.aCol_;
+  B = b->data_ + offset.bRow_*ldb + offset.bCol_;
 
-if (bAsRowVector == false_type && bAsColVector == false_type)
-  op(A[i * lda + j], B[i * ldb + j])
+if (bAsRowVector == false_type && bAsColVector == false_type)
+  op(A[i * lda + j], B[i * ldb + j])
 
-if (bAsRowVector == true_type && bAsColVector == false_type)
-  op(A[i * lda + j], B[j])
+if (bAsRowVector == true_type && bAsColVector == false_type)
+  op(A[i * lda + j], B[j])
 
-if (bAsRowVector == false_type && bAsColVector == true_type)
-  op(A[i * lda + j], B[i * ldb])
+if (bAsRowVector == false_type && bAsColVector == true_type)
+  op(A[i * lda + j], B[i * ldb])
 
-if (bAsRowVector == true_type && bAsColVector == true_type)
-  op(A[i * lda + j], B[0])
+if (bAsRowVector == true_type && bAsColVector == true_type)
+  op(A[i * lda + j], B[0])
 
    
 
    
 
    
@@ -205,19 +276,19 @@ if (bAsRowVector == true_type && bAsColVector == true_type)
 
    
 template <class Op>
    
 
    
-int applyBinary(Op op, BaseMatrixT &b, int numRows, int numCols, MatrixOffset &offset)
    
+int applyBinary(Op op, BaseMatrixT &b, int numRows, int numCols, MatrixOffset &offset)
 
    
 
 
    
 
    
 template <class Op>
    
 
    
-int applyTernary(Op op, BaseMatrixT &b, BaseMatrixT &c)
    
+int applyTernary(Op op, BaseMatrixT &b, BaseMatrixT &c)
 
    ternary operator: element wise op(a, b, c).
    
-
     
    for 0 <= i < this->height_ & for 0 <= j < this->width_.
+
     
    for 0 <= i < this->height_ & for 0 <= j < this->width_.
 
-While this->height_ == b.height_ && this->width_ == b.width_
-   && this->height_ == c.height_ && this->width_ == c.width_
+While this->height_ == b.height_ && this->width_ == b.width_
+   && this->height_ == c.height_ && this->width_ == c.width_
 
    
 
    
 
    
@@ -227,26 +298,26 @@ While this->height_ == b.height_ && this->width_ == b.width_
 
    
 template <class Op, class cAsRowVector, class cAsColVector>
    
 
    
-int applyTernary(Op op, BaseMatrixT &b, BaseMatrixT &c, int numRows, int numCols, MatrixOffset &offset, cAsRowVector, cAsColVector)
    
+int applyTernary(Op op, BaseMatrixT &b, BaseMatrixT &c, int numRows, int numCols, MatrixOffset &offset, cAsRowVector, cAsColVector)
 
    ternary operator: element wise op(a, b, c).
    
-
     
    for 0 <= i < numRows & for 0 <= j < numCols.
-While matrix start address is:
+
     
    for 0 <= i < numRows & for 0 <= j < numCols.
+While matrix start address is:
 
-  A = this->data_ + offset.aRow_*lda + offset.aCol_;
-  B = b->data_ + offset.bRow_*ldb + offset.bCol_;
-  C = c->data_ + offset.cRow_*ldc + offset.cCol_;
+  A = this->data_ + offset.aRow_*lda + offset.aCol_;
+  B = b->data_ + offset.bRow_*ldb + offset.bCol_;
+  C = c->data_ + offset.cRow_*ldc + offset.cCol_;
 
-  if (cAsRowVector == false_type && cAsColVector == false_type)
-    op(A[i*lda + j], B[i*ldb + j], C[i*ldc + j])
+  if (cAsRowVector == false_type && cAsColVector == false_type)
+    op(A[i*lda + j], B[i*ldb + j], C[i*ldc + j])
 
-  if (cAsRowVector == true_type && cAsColVector == false_type)
-    op(A[i*lda + j], B[i*ldb + j], C[j])
+  if (cAsRowVector == true_type && cAsColVector == false_type)
+    op(A[i*lda + j], B[i*ldb + j], C[j])
 
-  if (cAsRowVector == false_type && cAsColVector == true_type)
-    op(A[i*lda + j], B[i*ldb + j], C[i*ldc])
+  if (cAsRowVector == false_type && cAsColVector == true_type)
+    op(A[i*lda + j], B[i*ldb + j], C[i*ldc])
 
-  if (cAsRowVector == 1 && cAsColVector == 1)
-    op(A[i*lda + j], B[i*ldb + j], C[0])
+  if (cAsRowVector == 1 && cAsColVector == 1)
+    op(A[i*lda + j], B[i*ldb + j], C[0])
 
    
 
    
 
    
@@ -256,20 +327,20 @@ While matrix start address is:
 
    
 template <class Op>
    
 
    
-int applyTernary(Op op, BaseMatrixT &b, BaseMatrixT &c, int numRows, int numCols, MatrixOffset &offset)
    
+int applyTernary(Op op, BaseMatrixT &b, BaseMatrixT &c, int numRows, int numCols, MatrixOffset &offset)
 
    
 
 
    
 
    
 template <class Op>
    
 
    
-int applyQuaternary(Op op, BaseMatrixT &b, BaseMatrixT &c, BaseMatrixT &d)
    
+int applyQuaternary(Op op, BaseMatrixT &b, BaseMatrixT &c, BaseMatrixT &d)
 
    quaternary operator: element wise op(a, b, c, d).
    
-
     
    for 0 <= i < this->height_ & for 0 <= j < this->width_.
+
     
    for 0 <= i < this->height_ & for 0 <= j < this->width_.
 
-While this->height_ == b.height_ && this->width_ == b.width_
-   && this->height_ == c.height_ && this->width_ == c.width_
-   && this->height_ == d.height_ && this->width_ == d.width_
+While this->height_ == b.height_ && this->width_ == b.width_
+   && this->height_ == c.height_ && this->width_ == c.width_
+   && this->height_ == d.height_ && this->width_ == d.width_
 
    
 
    
 
    
@@ -279,14 +350,14 @@ While this->height_ == b.height_ && this->width_ == b.width_
 
    
 template <class Op>
    
 
    
-int applyQuaternary(Op op, BaseMatrixT &b, BaseMatrixT &c, BaseMatrixT &d, int numRows, int numCols, MatrixOffset &offset)
    
+int applyQuaternary(Op op, BaseMatrixT &b, BaseMatrixT &c, BaseMatrixT &d, int numRows, int numCols, MatrixOffset &offset)
 
    quaternary operator: element wise op(a, b, c, d).
    
-
     
    for 0 <= i < numRows & for 0 <= j < numCols.
-While matrix start address is:
-   A = this->data_ + offset.aRow_*lda + offset.aCol_;
-   B = b->data_ + offset.bRow_*ldb + offset.bCol_;
-   C = c->data_ + offset.cRow_*ldc + offset.cCol_;
-   D = d->data_ + offset.dRow_*ldd + offset.dCol_;
+
     
    for 0 <= i < numRows & for 0 <= j < numCols.
+While matrix start address is:
+   A = this->data_ + offset.aRow_*lda + offset.aCol_;
+   B = b->data_ + offset.bRow_*ldb + offset.bCol_;
+   C = c->data_ + offset.cRow_*ldc + offset.cCol_;
+   D = d->data_ + offset.dRow_*ldd + offset.dCol_;
 
    
 
    
 
    
@@ -296,17 +367,17 @@ While matrix start address is:
 
    
 template <class Agg, class Op, class Saver, class aAsRowVector, class aAsColVector>
    
 
    
-int aggregate(Agg agg, Op op, Saver sv, BaseMatrixT &b, int numRows, int numCols, MatrixOffset &offset, aAsRowVector, aAsColVector)
    
+int aggregate(Agg agg, Op op, Saver sv, BaseMatrixT &b, int numRows, int numCols, MatrixOffset &offset, aAsRowVector, aAsColVector)
 
    a aggregate expression that apply each row(or column) of matrix b. op and sv is element wise operator.
    
-
     
    if (aAsRowVector == true_type && aAsColVector == false_type)
- for each column j & 0 <= i < numRows, do:
-   dst = agg(op(b[i*ldb + j]))
-   a[j] = sv(a[j], dst)
+
     
    if (aAsRowVector == true_type && aAsColVector == false_type)
+ for each column j & 0 <= i < numRows, do:
+   dst = agg(op(b[i*ldb + j]))
+   a[j] = sv(a[j], dst)
 
-if (aAsRowVector == false_type && aAsColVector == true_type)
- for each row i & 0 <= j < numCols, do:
-   dst = agg(op(b[i*ldb + j]))
-   a[i] = sv(a[i], dst)
+if (aAsRowVector == false_type && aAsColVector == true_type)
+ for each row i & 0 <= j < numCols, do:
+   dst = agg(op(b[i*ldb + j]))
+   a[i] = sv(a[i], dst)
 
    
 
    
 
    
@@ -316,18 +387,18 @@ if (aAsRowVector == false_type && aAsColVector == true_type)
 
    
 template <class Agg, class Op, class Saver, class aAsRowVector, class aAsColVector>
    
 
    
-int aggregate(Agg agg, Op op, Saver sv, BaseMatrixT &b, BaseMatrixT &c, int numRows, int numCols, MatrixOffset &offset, aAsRowVector, aAsColVector)
    
+int aggregate(Agg agg, Op op, Saver sv, BaseMatrixT &b, BaseMatrixT &c, int numRows, int numCols, MatrixOffset &offset, aAsRowVector, aAsColVector)
 
    a aggregate expression that apply each row(or column) of matrix b and c.
    
 
    op and sv is element wise operator.
    
-
     
    if (aAsRowVector == true_type && aAsColVector == false_type)
-  for each column j & 0 <= i < numRows, do:
-    dst = agg(op(b[i*ldb + j], c[i*ldc + j]))
-    a[j] = sv(a[j], dst)
+
     
    if (aAsRowVector == true_type && aAsColVector == false_type)
+  for each column j & 0 <= i < numRows, do:
+    dst = agg(op(b[i*ldb + j], c[i*ldc + j]))
+    a[j] = sv(a[j], dst)
 
-if (aAsRowVector == false_type && aAsColVector == true_type)
-  for each row i & 0 <= j < numCols, do:
-    dst = agg(op(b[i*ldb + j], c[i*ldc + j]))
-    a[i] = sv(a[i], dst)
+if (aAsRowVector == false_type && aAsColVector == true_type)
+  for each row i & 0 <= j < numCols, do:
+    dst = agg(op(b[i*ldb + j], c[i*ldc + j]))
+    a[i] = sv(a[i], dst)
 
    
 
    
 
    
@@ -337,10 +408,10 @@ if (aAsRowVector == false_type && aAsColVector == true_type)
 
    
 template <class Agg>
    
 
    
-int applyRow(Agg agg, BaseMatrixT &b)
    
+int applyRow(Agg agg, BaseMatrixT &b)
 
    a aggregate expression that apply each row of matrix b.
    
-
     
    for each row i & 0 <= j < b.width_, do:
-  this[i] = agg(b[i*ldb + j])
+
     
    for each row i & 0 <= j < b.width_, do:
+  this[i] = agg(b[i*ldb + j])
 
    
 
    
 
    
@@ -350,11 +421,11 @@ if (aAsRowVector == false_type && aAsColVector == true_type)
 
    
 template <class Agg, class Saver>
    
 
    
-int applyRow(Agg agg, Saver sv, BaseMatrixT &b)
    
+int applyRow(Agg agg, Saver sv, BaseMatrixT &b)
 
    a aggregate expression that apply each row of matrix b.
    
-
     
    for each row i & 0 <= j < b.width_, do:
-  dst = agg(b[i*ldb + j])
-  this[i] = sv(this[i], dst)
+
     
    for each row i & 0 <= j < b.width_, do:
+  dst = agg(b[i*ldb + j])
+  this[i] = sv(this[i], dst)
 
    
 
    
 
    
@@ -364,10 +435,10 @@ if (aAsRowVector == false_type && aAsColVector == true_type)
 
    
 template <class Agg>
    
 
    
-int applyCol(Agg agg, BaseMatrixT &b)
    
+int applyCol(Agg agg, BaseMatrixT &b)
 
    a aggregate expression that apply each column of matrix b.
    
-
     
    for each column j & 0 <= i < b.height_, do:
-  this[j] = agg(b[i*ldb + j])
+
     
    for each column j & 0 <= i < b.height_, do:
+  this[j] = agg(b[i*ldb + j])
 
    
 
    
 
    
@@ -377,11 +448,11 @@ if (aAsRowVector == false_type && aAsColVector == true_type)
 
    
 template <class Agg, class Saver>
    
 
    
-int applyCol(Agg agg, Saver sv, BaseMatrixT &b)
    
+int applyCol(Agg agg, Saver sv, BaseMatrixT &b)
 
    a aggregate expression that apply each column of matrix b.
    
-
     
    for each column j & 0 <= i < b.height_, do:
-  dst = agg(b[i*ldb + j])
-  this[j] = sv(this[j], dst)
+
     
    for each column j & 0 <= i < b.height_, do:
+  dst = agg(b[i*ldb + j])
+  this[j] = sv(this[j], dst)
 
    
 
    
 
    
@@ -399,7 +470,7 @@ if (aAsRowVector == false_type && aAsColVector == true_type)
 
 
    
 
    
-T *rowBuf(size_t row)
    
+T *rowBuf(size_t row)
 
    
 
 
    
@@ -456,7 +527,7 @@ if (aAsRowVector == false_type && aAsColVector == true_type)
 
    
 
    
 void zeroAtOffset(int64_t columnOffset, int64_t numColumns)
    
-
     
    this(row, col + columnOffset) = 0 for 0 <= col < numColumns
+
     
    this(row, col + columnOffset) = 0 for 0 <= col < numColumns
 
    
 
    
 
    
@@ -485,7 +556,7 @@ if (aAsRowVector == false_type && aAsColVector == true_type)
 
    
 
    
 void assign(T p)
    
-
     
    this = p
+
     
    this = p
 
    
 
    
 
    
@@ -494,7 +565,7 @@ if (aAsRowVector == false_type && aAsColVector == true_type)
 
    
 
    
 void add(T p)
    
-
     
    this = this + p
+
     
    this = this + p
 
    
 
    
 
    
@@ -503,7 +574,7 @@ if (aAsRowVector == false_type && aAsColVector == true_type)
 
    
 
    
 void add(T p1, T p2)
    
-
     
    this = this*p1 + p2
+
     
    this = this*p1 + p2
 
    
 
    
 
    
@@ -519,7 +590,7 @@ if (aAsRowVector == false_type && aAsColVector == true_type)
 
    
 
    
 void biggerThanScalar(T p)
    
-
     
    a = a > p ? 1.0f : 0.0f
+
     
    a = a > p ? 1.0f : 0.0f
 
    
 
    
 
    
@@ -528,7 +599,7 @@ if (aAsRowVector == false_type && aAsColVector == true_type)
 
    
 
    
 void downClip(T p)
    
-
     
    a = a > p ? a : p
+
     
    a = a > p ? a : p
 
    
 
    
 
    
@@ -536,8 +607,8 @@ if (aAsRowVector == false_type && aAsColVector == true_type)
 
 
    
 
    
-void assign(BaseMatrixT &b)
    
-
     
    this = b
+void assign(BaseMatrixT &b)
+
     
    this = b
 
    
 
    
 
    
@@ -545,14 +616,14 @@ if (aAsRowVector == false_type && aAsColVector == true_type)
 
 
    
 
    
-void assignAtOffset(BaseMatrixT &b, int64_t columnOffset)
    
-
     
    If b.width + columOffset <= this.width
- this(row, col + columnOffset) = b(row, col) for 0 <= col < b.width
+void assignAtOffset(BaseMatrixT &b, int64_t columnOffset)
+
     
    If b.width + columOffset <= this.width
+ this(row, col + columnOffset) = b(row, col) for 0 <= col < b.width
 
-If this.width + columnOffset <= b.width
- this(row, col) = b(row, col + columnOffset) for 0 <= col < this.width
+If this.width + columnOffset <= b.width
+ this(row, col) = b(row, col + columnOffset) for 0 <= col < this.width
 
-Otherwise, FATAL
+Otherwise, FATAL
 
    
 
    
 
    
@@ -560,20 +631,20 @@ Otherwise, FATAL
 
 
    
 
    
-void add(BaseMatrixT &b)
    
+void add(BaseMatrixT &b)
 
    this = this + b 
    
 
    
 
 
    
 
    
-void addAtOffset(BaseMatrixT &b, int64_t columnOffset)
    
-
     
    If b.width + columOffset <= this.width
- this(row, col + columnOffset) += b(row, col) for 0 <= col < b.width
+void addAtOffset(BaseMatrixT &b, int64_t columnOffset)
+
     
    If b.width + columOffset <= this.width
+ this(row, col + columnOffset) += b(row, col) for 0 <= col < b.width
 
-If this.width + columnOffset <= b.width
- this(row, col) += b(row, col + columnOffset) for 0 <= col < this.width
+If this.width + columnOffset <= b.width
+ this(row, col) += b(row, col + columnOffset) for 0 <= col < this.width
 
-Otherwise, FATAL
+Otherwise, FATAL
 
    
 
    
 
    
@@ -581,38 +652,38 @@ Otherwise, FATAL
 
 
    
 
    
-void addColVector(BaseMatrixT &b)
    
+void addColVector(BaseMatrixT &b)
 
    
 
 
    
 
    
-void addRowVector(BaseMatrixT &b)
    
+void addRowVector(BaseMatrixT &b)
 
    
 
 
    
 
    
-void addBias(BaseMatrixT &b, T scale)
    
+void addBias(BaseMatrixT &b, T scale)
 
    
 
 
    
 
    
-void mulRowVector(BaseMatrixT &b)
    
+void mulRowVector(BaseMatrixT &b)
 
    
 
 
    
 
    
-void divRowVector(BaseMatrixT &b)
    
+void divRowVector(BaseMatrixT &b)
 
    
 
 
    
 
    
-void addP2P(BaseMatrixT &b)
    
+void addP2P(BaseMatrixT &b)
 
    
 
 
    
 
    
-void add(BaseMatrixT &b, T p)
    
-
     
    this = this + b*p
+void add(BaseMatrixT &b, T p)
+
     
    this = this + b*p
 
    
 
    
 
    
@@ -620,8 +691,8 @@ Otherwise, FATAL
 
 
    
 
    
-void add(BaseMatrixT &b, T p1, T p2)
    
-
     
    this = p1*this + p2*b
+void add(BaseMatrixT &b, T p1, T p2)
+
     
    this = p1*this + p2*b
 
    
 
    
 
    
@@ -629,8 +700,8 @@ Otherwise, FATAL
 
 
    
 
    
-void sub(BaseMatrixT &b)
    
-
     
    this = this - b
+void sub(BaseMatrixT &b)
+
     
    this = this - b
 
    
 
    
 
    
@@ -638,8 +709,8 @@ Otherwise, FATAL
 
 
    
 
    
-void sub(BaseMatrixT &b, T p)
    
-
     
    this = this - b*p
+void sub(BaseMatrixT &b, T p)
+
     
    this = this - b*p
 
    
 
    
 
    
@@ -647,8 +718,8 @@ Otherwise, FATAL
 
 
    
 
    
-void relu(BaseMatrixT &b)
    
-
     
    b = max(0, this)
+void relu(BaseMatrixT &b)
+
     
    b = max(0, this)
 
    
 
    
 
    
@@ -656,13 +727,13 @@ Otherwise, FATAL
 
 
    
 
    
-void reluDerivative(BaseMatrixT &b)
    
+void reluDerivative(BaseMatrixT &b)
 
    
 
 
    
 
    
-void softrelu(BaseMatrixT &b)
    
-
     
    b = log(1.0 + exp(this))
+void softrelu(BaseMatrixT &b)
+
     
    b = log(1.0 + exp(this))
 
    
 
    
 
    
@@ -670,13 +741,13 @@ Otherwise, FATAL
 
 
    
 
    
-void softreluDerivative(BaseMatrixT &b)
    
+void softreluDerivative(BaseMatrixT &b)
 
    
 
 
    
 
    
-void brelu(BaseMatrixT &b)
    
-
     
    b = min(max(this, p1), p2)
+void brelu(BaseMatrixT &b)
+
     
    b = min(max(this, p1), p2)
 
    
 
    
 
    
@@ -684,13 +755,13 @@ Otherwise, FATAL
 
 
    
 
    
-void breluDerivative(BaseMatrixT &b)
    
+void breluDerivative(BaseMatrixT &b)
 
    
 
 
    
 
    
-void square(BaseMatrixT &b)
    
-
     
    b = this * this
+void square(BaseMatrixT &b)
+
     
    b = this * this
 
    
 
    
 
    
@@ -698,13 +769,13 @@ Otherwise, FATAL
 
 
    
 
    
-void squareDerivative(BaseMatrixT &b)
    
+void squareDerivative(BaseMatrixT &b)
 
    
 
 
    
 
    
-void tanh(BaseMatrixT &b)
    
-
     
    b = tanh(this)
+void tanh(BaseMatrixT &b)
+
     
    b = tanh(this)
 
    
 
    
 
    
@@ -712,13 +783,13 @@ Otherwise, FATAL
 
 
    
 
    
-void tanhDerivative(BaseMatrixT &b)
    
+void tanhDerivative(BaseMatrixT &b)
 
    
 
 
    
 
    
-void scaledTanh(BaseMatrixT &b, T p1, T p2)
    
-
     
    b = p1 * tanh(p2 * this)
+void scaledTanh(BaseMatrixT &b, T p1, T p2)
+
     
    b = p1 * tanh(p2 * this)
 
    
 
    
 
    
@@ -726,13 +797,13 @@ Otherwise, FATAL
 
 
    
 
    
-void scaledTanhDerivative(BaseMatrixT &b, T p1, T p2)
    
+void scaledTanhDerivative(BaseMatrixT &b, T p1, T p2)
 
    
 
 
    
 
    
-void reciprocal(BaseMatrixT &b)
    
-
     
    b = 1.0f / this
+void reciprocal(BaseMatrixT &b)
+
     
    b = 1.0f / this
 
    
 
    
 
    
@@ -740,13 +811,13 @@ Otherwise, FATAL
 
 
    
 
    
-void reciprocalDerivative(BaseMatrixT &b)
    
+void reciprocalDerivative(BaseMatrixT &b)
 
    
 
 
    
 
    
-void abs(BaseMatrixT &b)
    
-
     
    b = this > 0.0f ? this : -this
+void abs(BaseMatrixT &b)
+
     
    b = this > 0.0f ? this : -this
 
    
 
    
 
    
@@ -754,13 +825,13 @@ Otherwise, FATAL
 
 
    
 
    
-void absDerivative(BaseMatrixT &b)
    
+void absDerivative(BaseMatrixT &b)
 
    
 
 
    
 
    
-void sigmoid(BaseMatrixT &b)
    
-
     
    b = 1.0f / (1.0f + exp(-this))
+void sigmoid(BaseMatrixT &b)
+
     
    b = 1.0f / (1.0f + exp(-this))
 
    
 
    
 
    
@@ -768,13 +839,13 @@ Otherwise, FATAL
 
 
    
 
    
-void sigmoidDerivative(BaseMatrixT &b)
    
+void sigmoidDerivative(BaseMatrixT &b)
 
    
 
 
    
 
    
-void expDerivative(BaseMatrixT &b)
    
-
     
    b = a
+void expDerivative(BaseMatrixT &b)
+
     
    b = a
 
    
 
    
 
    
@@ -782,58 +853,58 @@ Otherwise, FATAL
 
 
    
 
    
-void sign(BaseMatrixT &b)
    
+void sign(BaseMatrixT &b)
 
    
 
 
    
 
    
-void exp(BaseMatrixT &b)
    
+void exp(BaseMatrixT &b)
 
    
 
 
    
 
    
-void pow(BaseMatrixT &b, T p)
    
+void pow(BaseMatrixT &b, T p)
 
    
 
 
    
 
    
-void log(BaseMatrixT &b)
    
+void log(BaseMatrixT &b)
 
    
 
 
    
 
    
-void sqrt(BaseMatrixT &b)
    
+void sqrt(BaseMatrixT &b)
 
    
 
 
    
 
    
-void addScalar(BaseMatrixT &b, T p)
    
+void addScalar(BaseMatrixT &b, T p)
 
    
 
 
    
 
    
-void subScalar(BaseMatrixT &b, T p)
    
+void subScalar(BaseMatrixT &b, T p)
 
    
 
 
    
 
    
-void mulScalar(BaseMatrixT &b, T p)
    
+void mulScalar(BaseMatrixT &b, T p)
 
    
 
 
    
 
    
-void divScalar(BaseMatrixT &b, T p)
    
+void divScalar(BaseMatrixT &b, T p)
 
    
 
 
    
 
    
-void scalarDiv(BaseMatrixT &b, T p)
    
+void scalarDiv(BaseMatrixT &b, T p)
 
    
 
 
    
 
    
-void invSqrt(BaseMatrixT &b)
    
-
     
    this = 1.0f / sqrt(b)
+void invSqrt(BaseMatrixT &b)
+
     
    this = 1.0f / sqrt(b)
 
    
 
    
 
    
@@ -841,35 +912,35 @@ Otherwise, FATAL
 
 
    
 
    
-void isEqualTo(BaseMatrixT &b, T value)
    
+void isEqualTo(BaseMatrixT &b, T value)
 
    this = (b == value) 
    
 
    
 
 
    
 
    
-void softCrossEntropy(BaseMatrixT &b, BaseMatrixT &c)
    
+void softCrossEntropy(BaseMatrixT &b, BaseMatrixT &c)
 
    ternary operator. 
    
 
    
 
 
    
 
    
-void softCrossEntropyBp(BaseMatrixT &b, BaseMatrixT &c)
    
+void softCrossEntropyBp(BaseMatrixT &b, BaseMatrixT &c)
 
    
 
 
    
 
    
-void binaryLabelCrossEntropy(BaseMatrixT &b, BaseMatrixT &c)
    
+void binaryLabelCrossEntropy(BaseMatrixT &b, BaseMatrixT &c)
 
    
 
 
    
 
    
-void binaryLabelCrossEntropyBp(BaseMatrixT &b, BaseMatrixT &c)
    
+void binaryLabelCrossEntropyBp(BaseMatrixT &b, BaseMatrixT &c)
 
    
 
 
    
 
    
-void add(BaseMatrixT &b, BaseMatrixT &c)
    
-
     
    this = b + c
+void add(BaseMatrixT &b, BaseMatrixT &c)
+
     
    this = b + c
 
    
 
    
 
    
@@ -877,8 +948,8 @@ Otherwise, FATAL
 
 
    
 
    
-void add(BaseMatrixT &b, T p1, BaseMatrixT &c, T p2)
    
-
     
    this = b*p1 + c*p2
+void add(BaseMatrixT &b, T p1, BaseMatrixT &c, T p2)
+
     
    this = b*p1 + c*p2
 
    
 
    
 
    
@@ -886,8 +957,8 @@ Otherwise, FATAL
 
 
    
 
    
-void sub(BaseMatrixT &b, BaseMatrixT &c)
    
-
     
    this = b - c
+void sub(BaseMatrixT &b, BaseMatrixT &c)
+
     
    this = b - c
 
    
 
    
 
    
@@ -895,8 +966,8 @@ Otherwise, FATAL
 
 
    
 
    
-void sub(BaseMatrixT &b, T p1, BaseMatrixT &c, T p2)
    
-
     
    this = b*p1 - c*p2
+void sub(BaseMatrixT &b, T p1, BaseMatrixT &c, T p2)
+
     
    this = b*p1 - c*p2
 
    
 
    
 
    
@@ -904,8 +975,8 @@ Otherwise, FATAL
 
 
    
 
    
-void add2(BaseMatrixT &b, BaseMatrixT &c)
    
-
     
    this = this + b + c
+void add2(BaseMatrixT &b, BaseMatrixT &c)
+
     
    this = this + b + c
 
    
 
    
 
    
@@ -913,8 +984,8 @@ Otherwise, FATAL
 
 
    
 
    
-void add2(BaseMatrixT &b, BaseMatrixT &c, T p1, T p2, T p3)
    
-
     
    this = this*p1 + b*p2 + c*p3
+void add2(BaseMatrixT &b, BaseMatrixT &c, T p1, T p2, T p3)
+
     
    this = this*p1 + b*p2 + c*p3
 
    
 
    
 
    
@@ -922,8 +993,8 @@ Otherwise, FATAL
 
 
    
 
    
-void add3(BaseMatrixT &b, BaseMatrixT &c, BaseMatrixT &d, T p1, T p2, T p3)
    
-
     
    this = a*p1 + b*p2 + c*p3
+void add3(BaseMatrixT &b, BaseMatrixT &c, BaseMatrixT &d, T p1, T p2, T p3)
+
     
    this = a*p1 + b*p2 + c*p3
 
    
 
    
 
    
@@ -931,8 +1002,8 @@ Otherwise, FATAL
 
 
    
 
    
-void sgdUpdate(BaseMatrixT &b, BaseMatrixT &c, T p1, T p2, T p3)
    
-
     
    c = p2 * c - p1 *  (b + p3 * this)
+void sgdUpdate(BaseMatrixT &b, BaseMatrixT &c, T p1, T p2, T p3)
+
     
    c = p2 * c - p1 *  (b + p3 * this)
 this += mom
 
    
 
    
@@ -941,8 +1012,8 @@ Otherwise, FATAL
 
 
    
 
    
-void sgdUpdate(BaseMatrixT &b, BaseMatrixT &c, BaseMatrixT &d, T p1, T p2, T p3)
    
-
     
    c = p2 * c - p1 * d * (b + p3 * this)
+void sgdUpdate(BaseMatrixT &b, BaseMatrixT &c, BaseMatrixT &d, T p1, T p2, T p3)
+
     
    c = p2 * c - p1 * d * (b + p3 * this)
 this += mom
 
    
 
    
@@ -957,7 +1028,7 @@ Otherwise, FATAL
 
 
    
 
    
-void applyL1(BaseMatrixT &lr, T learningRate, T decayRate)
    
+void applyL1(BaseMatrixT &lr, T learningRate, T decayRate)
 
    
 
 
    
@@ -967,13 +1038,13 @@ Otherwise, FATAL
 
 
    
 
    
-void applyL2(BaseMatrixT &lr, T learningRate, T decayRate)
    
+void applyL2(BaseMatrixT &lr, T learningRate, T decayRate)
 
    
 
 
    
 
    
-void dotMul(BaseMatrixT &b)
    
-
     
    this *= b
+void dotMul(BaseMatrixT &b)
+
     
    this *= b
 
    
 
    
 
    
@@ -981,8 +1052,8 @@ Otherwise, FATAL
 
 
    
 
    
-void dotMul(BaseMatrixT &b, BaseMatrixT &c)
    
-
     
    this = b * c
+void dotMul(BaseMatrixT &b, BaseMatrixT &c)
+
     
    this = b * c
 
    
 
    
 
    
@@ -990,8 +1061,8 @@ Otherwise, FATAL
 
 
    
 
    
-void dotDiv(BaseMatrixT &b, BaseMatrixT &c)
    
-
     
    this = b / c
+void dotDiv(BaseMatrixT &b, BaseMatrixT &c)
+
     
    this = b / c
 
    
 
    
 
    
@@ -999,8 +1070,8 @@ Otherwise, FATAL
 
 
    
 
    
-void dotDiv(BaseMatrixT &b, BaseMatrixT &c, T p1, T p2)
    
-
     
    this = (b + p1) / (c + p2)
+void dotDiv(BaseMatrixT &b, BaseMatrixT &c, T p1, T p2)
+
     
    this = (b + p1) / (c + p2)
 
    
 
    
 
    
@@ -1008,8 +1079,8 @@ Otherwise, FATAL
 
 
    
 
    
-void rankLoss(BaseMatrixT &b, BaseMatrixT &c, BaseMatrixT &d)
    
-
     
    this = log(1 + exp(b - c)) - d * (b - c)
+void rankLoss(BaseMatrixT &b, BaseMatrixT &c, BaseMatrixT &d)
+
     
    this = log(1 + exp(b - c)) - d * (b - c)
 
    
 
    
 
    
@@ -1017,13 +1088,13 @@ Otherwise, FATAL
 
 
    
 
    
-void rankLossBp(BaseMatrixT &b, BaseMatrixT &c, BaseMatrixT &d)
    
+void rankLossBp(BaseMatrixT &b, BaseMatrixT &c, BaseMatrixT &d)
 
    
 
 
    
 
    
-void logisticRegressionLoss(BaseMatrixT &b, BaseMatrixT &c)
    
-
     
    this = log(1 + exp(b)) - c * b
+void logisticRegressionLoss(BaseMatrixT &b, BaseMatrixT &c)
+
     
    this = log(1 + exp(b)) - c * b
 
    
 
    
 
    
@@ -1031,8 +1102,8 @@ Otherwise, FATAL
 
 
    
 
    
-void logisticRegressionLossBp(BaseMatrixT &b, BaseMatrixT &c)
    
-
     
    this += exp(b)/(1+exp(b)) - c
+void logisticRegressionLossBp(BaseMatrixT &b, BaseMatrixT &c)
+
     
    this += exp(b)/(1+exp(b)) - c
 
    
 
    
 
    
@@ -1040,8 +1111,8 @@ Otherwise, FATAL
 
 
    
 
    
-void biggerThan(BaseMatrixT &b, BaseMatrixT &c)
    
-
     
    this = b > c ? 1.0 : 0.0
+void biggerThan(BaseMatrixT &b, BaseMatrixT &c)
+
     
    this = b > c ? 1.0 : 0.0
 
    
 
    
 
    
@@ -1049,8 +1120,8 @@ Otherwise, FATAL
 
 
    
 
    
-void biggerThan(BaseMatrixT &b, BaseMatrixT &c, BaseMatrixT &d)
    
-
     
    this = ((b>c && d>0.5) || (b<c && d<0.5)) ? 1 : 0)
+void biggerThan(BaseMatrixT &b, BaseMatrixT &c, BaseMatrixT &d)
+
     
    this = ((b>c && d>0.5) || (b<c && d<0.5)) ? 1 : 0)
 
    
 
    
 
    
@@ -1058,8 +1129,8 @@ Otherwise, FATAL
 
 
    
 
    
-void max(BaseMatrixT &b, BaseMatrixT &c)
    
-
     
    this = b>c ? b : c
+void max(BaseMatrixT &b, BaseMatrixT &c)
+
     
    this = b>c ? b : c
 
    
 
    
 
    
@@ -1067,8 +1138,8 @@ Otherwise, FATAL
 
 
    
 
    
-void binaryClassificationError(size_t destCol, BaseMatrixT &b, BaseMatrixT &c, T p)
    
-
     
    this[destCol] += (b>p1 == c>p1) ? 0 : 1)
+void binaryClassificationError(size_t destCol, BaseMatrixT &b, BaseMatrixT &c, T p)
+
     
    this[destCol] += (b>p1 == c>p1) ? 0 : 1)
 
    
 
    
 
    
@@ -1076,13 +1147,13 @@ Otherwise, FATAL
 
 
    
 
    
-void binaryClassificationError2(size_t destCol, BaseMatrixT &b, BaseMatrixT &c, T p)
    
+void binaryClassificationError2(size_t destCol, BaseMatrixT &b, BaseMatrixT &c, T p)
 
    
 
 
    
 
    
-void dotMulSquare(BaseMatrixT &b)
    
-
     
    this = this * b * b
+void dotMulSquare(BaseMatrixT &b)
+
     
    this = this * b * b
 
    
 
    
 
    
@@ -1090,8 +1161,8 @@ Otherwise, FATAL
 
 
    
 
    
-void dotSquareMul(BaseMatrixT &b)
    
-
     
    this = this * this * b
+void dotSquareMul(BaseMatrixT &b)
+
     
    this = this * this * b
 
    
 
    
 
    
@@ -1099,8 +1170,8 @@ Otherwise, FATAL
 
 
    
 
    
-void dotMulSquare(BaseMatrixT &b, BaseMatrixT &c)
    
-
     
    this = b * c * c
+void dotMulSquare(BaseMatrixT &b, BaseMatrixT &c)
+
     
    this = b * c * c
 
    
 
    
 
    
@@ -1108,8 +1179,8 @@ Otherwise, FATAL
 
 
    
 
    
-void dotSquareSquare(BaseMatrixT &b, BaseMatrixT &c)
    
-
     
    this = b * b * c * c
+void dotSquareSquare(BaseMatrixT &b, BaseMatrixT &c)
+
     
    this = b * b * c * c
 
    
 
    
 
    
@@ -1117,8 +1188,8 @@ Otherwise, FATAL
 
 
    
 
    
-void dotMulSquareSum(BaseMatrixT &b, BaseMatrixT &c, T p1, T p2)
    
-
     
    this = this * (p1*b + p2*c)^2
+void dotMulSquareSum(BaseMatrixT &b, BaseMatrixT &c, T p1, T p2)
+
     
    this = this * (p1*b + p2*c)^2
 
    
 
    
 
    
@@ -1126,8 +1197,8 @@ Otherwise, FATAL
 
 
    
 
    
-void dotSquareSum(BaseMatrixT &b, BaseMatrixT &c, T p1, T p2)
    
-
     
    this = (p1*b + p2*c)^2
+void dotSquareSum(BaseMatrixT &b, BaseMatrixT &c, T p1, T p2)
+
     
    this = (p1*b + p2*c)^2
 
    
 
    
 
    
@@ -1135,8 +1206,8 @@ Otherwise, FATAL
 
 
    
 
    
-void dotMulSum(BaseMatrixT &b, BaseMatrixT &c, T p1, T p2)
    
-
     
    this=  this * (p1*b + p2*c)
+void dotMulSum(BaseMatrixT &b, BaseMatrixT &c, T p1, T p2)
+
     
    this=  this * (p1*b + p2*c)
 
    
 
    
 
    
@@ -1144,8 +1215,8 @@ Otherwise, FATAL
 
 
    
 
    
-void addSquareSum(BaseMatrixT &b, BaseMatrixT &c, BaseMatrixT d, T p1, T p2, T p3)
    
-
     
    this += sqr(p1*b + p2*c + p3*d)
+void addSquareSum(BaseMatrixT &b, BaseMatrixT &c, BaseMatrixT d, T p1, T p2, T p3)
+
     
    this += sqr(p1*b + p2*c + p3*d)
 
    
 
    
 
    
@@ -1153,8 +1224,8 @@ Otherwise, FATAL
 
 
    
 
    
-void addSquare(BaseMatrixT &b, T p)
    
-
     
    this += p * sqr(b)
+void addSquare(BaseMatrixT &b, T p)
+
     
    this += p * sqr(b)
 
    
 
    
 
    
@@ -1162,8 +1233,8 @@ Otherwise, FATAL
 
 
    
 
    
-void decayAddSquare(BaseMatrixT &b, T p1, T p2)
    
-
     
    this = p1 * this + p2 * sqr(b)
+void decayAddSquare(BaseMatrixT &b, T p1, T p2)
+
     
    this = p1 * this + p2 * sqr(b)
 
    
 
    
 
    
@@ -1171,8 +1242,8 @@ Otherwise, FATAL
 
 
    
 
    
-void decayAddSquareMul(BaseMatrixT &b, BaseMatrixT &c, T p1, T p2)
    
-
     
    this = p1 * this + p2 * sqr(b * c)
+void decayAddSquareMul(BaseMatrixT &b, BaseMatrixT &c, T p1, T p2)
+
     
    this = p1 * this + p2 * sqr(b * c)
 
    
 
    
 
    
@@ -1180,8 +1251,8 @@ Otherwise, FATAL
 
 
    
 
    
-void reciprocal(BaseMatrixT &b, T p1, T p2)
    
-
     
    this = 1 / (p1 * b + p2)
+void reciprocal(BaseMatrixT &b, T p1, T p2)
+
     
    this = 1 / (p1 * b + p2)
 
    
 
    
 
    
@@ -1189,8 +1260,8 @@ Otherwise, FATAL
 
 
    
 
    
-void reciprocalSum(BaseMatrixT &b, BaseMatrixT &c, T p1, T p2, T p3)
    
-
     
    this = 1 / (p1 * b + p2 * c + p3)
+void reciprocalSum(BaseMatrixT &b, BaseMatrixT &c, T p1, T p2, T p3)
+
     
    this = 1 / (p1 * b + p2 * c + p3)
 
    
 
    
 
    
@@ -1198,8 +1269,8 @@ Otherwise, FATAL
 
 
    
 
    
-void copyAndClear(BaseMatrixT &b)
    
-
     
    b = this; this = 0
+void copyAndClear(BaseMatrixT &b)
+
     
    b = this; this = 0
 
    
 
    
 
    
@@ -1207,8 +1278,8 @@ Otherwise, FATAL
 
 
    
 
    
-void rowDotMul(size_t destCol, BaseMatrixT &b, BaseMatrixT &c)
    
-
     
    this_row[destCol] += dotprod(b_row, c_row)
+void rowDotMul(size_t destCol, BaseMatrixT &b, BaseMatrixT &c)
+
     
    this_row[destCol] += dotprod(b_row, c_row)
 
    
 
    
 
    
@@ -1216,15 +1287,15 @@ Otherwise, FATAL
 
 
    
 
    
-void rowDotMul2(size_t destCol, BaseMatrixT &b, BaseMatrixT &c)
    
+void rowDotMul2(size_t destCol, BaseMatrixT &b, BaseMatrixT &c)
 
    
 
 
    
 
    
-void addDotMulVMM(BaseMatrixT &b, BaseMatrixT &c)
    
+void addDotMulVMM(BaseMatrixT &b, BaseMatrixT &c)
 
    this is vector (one row matrix)
    
-
     
    for each row i, do:
-   this_row += dotmul(b_row_i, c_row_i)
+
     
    for each row i, do:
+   this_row += dotmul(b_row_i, c_row_i)
 
    
 
    
 
    
@@ -1232,15 +1303,15 @@ Otherwise, FATAL
 
 
    
 
    
-void addDotMulVMM2(BaseMatrixT &b, BaseMatrixT &c)
    
+void addDotMulVMM2(BaseMatrixT &b, BaseMatrixT &c)
 
    
 
 
    
 
    
-void addDotMulMMV(BaseMatrixT &b, BaseMatrixT &c)
    
+void addDotMulMMV(BaseMatrixT &b, BaseMatrixT &c)
 
    c is vector (one row matrix)
    
-
     
    for each row i, do:
-   this_row_i += dotmul(b_row_i, c_row)
+
     
    for each row i, do:
+   this_row_i += dotmul(b_row_i, c_row)
 
    
 
    
 
    
@@ -1248,13 +1319,13 @@ Otherwise, FATAL
 
 
    
 
    
-void addDotMulMMV2(BaseMatrixT &b, BaseMatrixT &c)
    
+void addDotMulMMV2(BaseMatrixT &b, BaseMatrixT &c)
 
    
 
 
    
 
    
-void addDotMul(BaseMatrixT &b, BaseMatrixT &c, T p1, T p2)
    
-
     
    this = p1 * this + p2 * b * c
+void addDotMul(BaseMatrixT &b, BaseMatrixT &c, T p1, T p2)
+
     
    this = p1 * this + p2 * b * c
 
    
 
    
 
    
@@ -1262,8 +1333,8 @@ Otherwise, FATAL
 
 
    
 
    
-void rowScale(size_t cCol, BaseMatrixT &b, BaseMatrixT &c)
    
-
     
    this_row = b_row * c_row[cCol]
+void rowScale(size_t cCol, BaseMatrixT &b, BaseMatrixT &c)
+
     
    this_row = b_row * c_row[cCol]
 
    
 
    
 
    
@@ -1271,13 +1342,13 @@ Otherwise, FATAL
 
 
    
 
    
-void rowScale2(size_t cCol, BaseMatrixT &b, BaseMatrixT &c)
    
+void rowScale2(size_t cCol, BaseMatrixT &b, BaseMatrixT &c)
 
    
 
 
    
 
    
-void colScale(size_t cRow, BaseMatrixT &b, BaseMatrixT &c)
    
-
     
    this_col = b_col * c_col[cRow]
+void colScale(size_t cRow, BaseMatrixT &b, BaseMatrixT &c)
+
     
    this_col = b_col * c_col[cRow]
 
    
 
    
 
    
@@ -1285,8 +1356,8 @@ Otherwise, FATAL
 
 
    
 
    
-void addColScale(size_t cRow, BaseMatrixT &b, BaseMatrixT &c)
    
-
     
    this_col += b_col * c_col[cRow]
+void addColScale(size_t cRow, BaseMatrixT &b, BaseMatrixT &c)
+
     
    this_col += b_col * c_col[cRow]
 
    
 
    
 
    
@@ -1294,8 +1365,8 @@ Otherwise, FATAL
 
 
    
 
    
-void addRowScale(size_t cCol, BaseMatrixT &b, BaseMatrixT &c)
    
-
     
    this_row += b_row * c_row[cCol]
+void addRowScale(size_t cCol, BaseMatrixT &b, BaseMatrixT &c)
+
     
    this_row += b_row * c_row[cCol]
 
    
 
    
 
    
@@ -1303,55 +1374,55 @@ Otherwise, FATAL
 
 
    
 
    
-void sumRows(BaseMatrixT &b)
    
+void sumRows(BaseMatrixT &b)
 
    calculate the sum of each row of the matrix b. 
    
 
    
 
 
    
 
    
-void maxRows(BaseMatrixT &b)
    
+void maxRows(BaseMatrixT &b)
 
    calculate the maximum value of each row of the matrix b. 
    
 
    
 
 
    
 
    
-void minRows(BaseMatrixT &b)
    
+void minRows(BaseMatrixT &b)
 
    calculate the minimum value of each row of the matrix b. 
    
 
    
 
 
    
 
    
-void sumCols(BaseMatrixT &b)
    
+void sumCols(BaseMatrixT &b)
 
    calculate the sum of each column of the matrix b. 
    
 
    
 
 
    
 
    
-void maxCols(BaseMatrixT &b)
    
+void maxCols(BaseMatrixT &b)
 
    calculate the maximum value of each column of the matrix b. 
    
 
    
 
 
    
 
    
-void minCols(BaseMatrixT &b)
    
+void minCols(BaseMatrixT &b)
 
    calculate the minimum value of each column of the matrix b. 
    
 
    
 
 
    
 
    
-void sumCols(BaseMatrixT &b, T scale)
    
+void sumCols(BaseMatrixT &b, T scale)
 
    
 
 
    
 
    
-void sumOfSquares(BaseMatrixT &b, BaseMatrixT &c)
    
+void sumOfSquares(BaseMatrixT &b, BaseMatrixT &c)
 
    calculate the sum of each row of (b - c)^2. 
    
 
    
 
 
    
 
    
-void rowAdd(size_t cCol, BaseMatrixT &b, BaseMatrixT &c, T p)
    
-
     
    this_row = b_row + p * ones * c_row[cCol]
+void rowAdd(size_t cCol, BaseMatrixT &b, BaseMatrixT &c, T p)
+
     
    this_row = b_row + p * ones * c_row[cCol]
 
    
 
    
 
    
@@ -1359,8 +1430,8 @@ Otherwise, FATAL
 
 
    
 
    
-void rowPow(size_t cCol, BaseMatrixT &b, BaseMatrixT &c)
    
-
     
    this_row = pow(b_row, c_row[cCol])
+void rowPow(size_t cCol, BaseMatrixT &b, BaseMatrixT &c)
+
     
    this_row = pow(b_row, c_row[cCol])
 
    
 
    
 
    
@@ -1407,77 +1478,6 @@ Otherwise, FATAL
 
    
 
    
 
-
    
-
    
-template <class T, T v>
    
-
    
-struct bool_constant
    
-
    
-
    Public Static Attributes
    
-
    
-
    
-const T value
    
-
    
-
-
    
-
    
-
-
    
-
    
-class MatrixOffset
    
-
    
-
    Public Functions
    
-
    
-
    
-MatrixOffset(size_t aCol = 0, size_t aRow = 0, size_t bCol = 0, size_t bRow = 0, size_t cCol = 0, size_t cRow = 0, size_t dCol = 0, size_t dRow = 0)
    
-
    
-
-
    
-
    
-
    Public Members
    
-
    
-
    
-size_t aCol_
    
-
    
-
-
    
-
    
-size_t aRow_
    
-
    
-
-
    
-
    
-size_t bCol_
    
-
    
-
-
    
-
    
-size_t bRow_
    
-
    
-
-
    
-
    
-size_t cCol_
    
-
    
-
-
    
-
    
-size_t cRow_
    
-
    
-
-
    
-
    
-size_t dCol_
    
-
    
-
-
    
-
    
-size_t dRow_
    
-
    
-
-
    
-
    
-
 
    
 
 
    
@@ -1485,7 +1485,7 @@ Otherwise, FATAL
 
    Sparse Matrix
    
 
    
 
    
-namespace paddle
    
+namespace paddle
 
    
 
    Typedefs
    
 
    
@@ -1537,17 +1537,17 @@ Otherwise, FATAL
 enum SparseFormat
 
    matrix sparse_format . 
    
 
    nnz represents nonzero number in sparse matrix.
    
-
    SPARSE_CSR: row major matrix. length of row is height_ + 1, each element represents row start index in Matrix. length of col and value are nnz.
    
-
    SPARSE_CSC: col major matrix. length of col is width_ + 1, each element represents col start index in Matrix. length of col and value are nnz.
    
-
     
    for example: [0, 1, 0, 2, 0;
-              1, 0, 0, 0, 0;
-              0, 0, 0, 2, 5];
-SPARSE_CSR row   [0, 2, 3, 5];
-           col   [1, 3, 0, 3, 4];
-           value [1, 2, 1, 2, 5]
-SPARSE_CSC col   [0, 1, 2, 2, 4, 5];
-           row   [1, 0, 0, 2, 2];
-           value [1, 1, 2, 2, 5]
+
    SPARSE_CSR: row major matrix. length of row is height_ + 1, each element represents row start index in Matrix. length of col and value are nnz.
    
+
    SPARSE_CSC: col major matrix. length of col is width_ + 1, each element represents col start index in Matrix. length of col and value are nnz.
    
+
     
    for example: [0, 1, 0, 2, 0;
+              1, 0, 0, 0, 0;
+              0, 0, 0, 2, 5];
+SPARSE_CSR row   [0, 2, 3, 5];
+           col   [1, 3, 0, 3, 4];
+           value [1, 2, 1, 2, 5]
+SPARSE_CSC col   [0, 1, 2, 2, 4, 5];
+           row   [1, 0, 0, 2, 2];
+           value [1, 1, 2, 2, 5]
 
    
 
    
 
    
@@ -1574,445 +1574,482 @@ SPARSE_CSC col   [0, 1, 2, 2, 4, 5];
 
 
    
 
    
-
    
-class CpuMatrix
    
-
    Inherits from paddle::Matrix
    
-
    Subclassed by paddle::SharedCpuMatrix, paddle::SparseRowCpuMatrix, paddle::SparseRowIdsCpuMatrix
    
+
    
+class Matrix
    
+
    #include <Matrix.h>
    Copy or assignemnt constructor will share the data as opposed to making a copy of the original data. To make a copy of the orinal data, use copyFrom() instead. 
    
+
    Inherits from paddle::BaseMatrixT< real >
    
+
    Subclassed by paddle::CpuMatrix, paddle::CpuSparseMatrix, paddle::GpuMatrix, paddle::GpuSparseMatrix
    
 
    
 
    Public Functions
    
 
    
-
    
-CpuMatrix(size_t height, size_t width, bool trans = false)
    
-
    CpuMatrix 
    
-
    
+
    
+virtual ~Matrix()
    
+
    
 
 
    
-
    
-CpuMatrix(real *data, size_t height, size_t width, bool trans = false)
    
-
    
+
    
+void setData(real *data)
    
+
    set the data buffer used to hold the matrix data. 
    
+
    caller should make sure that the size of data is at least sizeof(real)*height*width. 
    
+
    
 
 
    
-
    
-CpuMatrix(real *data, size_t height, size_t width, size_t stride, bool trans = false)
    
-
    
+
    
+void setData(real *data, size_t newHeight, size_t newWidth)
    
+
    the data should be contiguous 
    
+
    
 
 
    
-
    
-CpuMatrix(CpuMemHandlePtr dataHandle, size_t height, size_t width, bool trans = false)
    
+
    
+size_t getWidth() const
    
 
    
 
 
    
-
    
-~CpuMatrix()
    
+
    
+size_t getHeight() const
    
 
    
 
 
    
-
    
-virtual void zeroMem()
    
+
    
+size_t getStride() const
    
 
    
 
 
    
-
    
-virtual void resetOne()
    
+
    
+size_t getElementCnt() const
    
 
    
 
 
    
-
    
-virtual void resize(size_t newHeight, size_t newWidth)
    
-
    
-
    Note
    
-
    Original data may not be preserved after resize(). 
    
-
    
-
    
-
    
+
    
+virtual real *getData()
    
+
    
 
 
    
-
    
-virtual void resize(size_t newHeight, size_t newWidth, size_t newNnz, SparseValueType valueType, SparseFormat format)
    
-
    
-
    Note
    
-
    This should only be used for sparse matrix. 
    
-
    
-
    
-
    
+
    
+virtual const real *getData() const
    
+
    
 
 
    
-
    
-virtual void setRow(size_t row, size_t colNum, const unsigned int *cols, const real *values)
    
-
    This should only be used for sparse matrix. 
    
-
    Currently must be called for each row in order. The matrix is not valid until setRow is called for the last row. 
    
-
    
+
    
+bool isTransposed() const
    
+
    
 
 
    
-
    
-virtual real getElement(size_t x, size_t y) const
    
+
    
+bool isContiguous() const
    
 
    
 
 
    
-
    
-virtual real getSum()
    
+
    
+virtual int *getRows() const
    
 
    
 
 
    
-
    
-virtual void accumulateColSum(Matrix &src)
    
+
    
+virtual int *getCols() const
    
 
    
 
 
    
-
    
-virtual real getAbsSum()
    
+
    
+virtual SparseFormat getFormat() const
    
 
    
 
 
    
-
    
-virtual MatrixPtr getTranspose()
    
+
    
+virtual SparseValueType getValueType() const
    
 
    
 
 
    
-
    
-virtual void transpose(MatrixPtr matTrans, bool memAlloc)
    
-
    hard transpose. 
    
-
    allocate matTrans’ memory outside, then set memAlloc as false; else set as true. 
    
+
    
+virtual void add3(MatrixPtr b)
    
+
    matrix elment-wise add 
    
+
    Named add3 just because add/add2 has been used in BaseMatrix.cu and they are not virtual function. 
    
 
    
 
 
    
-
    
-virtual void copyFrom(const Matrix &src)
    
+
    
+MemoryHandlePtr getMemoryHandle() const
    
 
    
 
 
    
-
    
-virtual void copyFrom(const Matrix &src, hl_stream_t stream)
    
+
    
+virtual void zeroMem()
    
 
    
 
 
    
-
    
-virtual void copyFrom(const real *src, size_t size)
    
-
    If this is GpuMatrix, src is assumed to be CPU memory
    
-
    If this is CpuMatrix, src is assumed to be CPU memory 
    
-
    
+
    
+virtual void resetOne()
    
+
    
 
 
    
-
    
-virtual void copyFrom(const real *cpuSrc, const int64_t *seq)
    
+
    
+virtual void copyFrom(const Matrix &src)
    
 
    
 
 
    
-
    
-virtual void copyFrom(const IVector &src)
    
-
    convert a int vector to a real matrix. 
    
-
    (1) source and dest are both in CPU.
    
-
    (2) sizes are exactly match. 
    
-
    
-
-
    
-
    
-void copyFrom(CpuSparseMatrix &src)
    
+
    
+virtual void trimFrom(const CpuSparseMatrix &src)
    
 
    
 
 
    
-
    
-virtual void copyByRowIndex(Matrix &b, IVector &rowIndex)
    
+
    
+virtual void copyFrom(const Matrix &src, hl_stream_t stream)
    
 
    
 
 
    
-
    
-virtual MatrixPtr clone(size_t height, size_t width, bool useGpu = false)
    
-
    Create a matrix with the same type (GpuMatrix, CpuMatrix, NonValueSparseMatrix, etc.) as this. 
    
-
    If height and width is zero, the new matrix will have the same size as this, otherwise the new matrix will have the specified size. 
    
-
    
+
    
+MatrixPtr subMatrix(size_t startRow, size_t endRow, size_t startCol, size_t endCol)
    
+
    
 
 
    
-
    
-virtual void convExpand(Matrix &feature, int feaImgHeight, int feaImgWidth, int channels, int blockH, int blockW, int strideH, int strideW, int paddingH, int paddingW, int outputH, int outputW)
    
-
    This function is used to calculate the convolution:
    
-
    It will expand a feature matrix according to the convolution filters 
    
-
    
+
    
+MatrixPtr subRowMatrix(size_t startRow, size_t endRow)
    
+
    
 
 
    
-
    
-virtual void convShrink(Matrix &expandColMat, int thisImgHeight, int thisImgWidth, int channels, int blockH, int blockW, int strideH, int strideW, int paddingH, int paddingW, int outputH, int outputW, real alpha = 1.0f, real beta = 0.0f)
    
-
    This function is the reverse implementation of convExpand:
    
-
    Its function is to restore a expanded-matrix into a feature matrix 
    
-
    
+
    
+MatrixPtr subColMatrix(size_t startCol, size_t endCol)
    
+
    
 
 
    
-
    
-virtual void maxPoolForward(Matrix &inputMat, size_t imgSizeH, size_t imgSizeW, size_t channels, size_t sizeX, int start_, size_t stride, size_t outputH, size_t outputW)
    
-
    Pooling forward operation, pick out the largest element in the sizeX of value 
    
-
    
+
    
+virtual MatrixPtr subMatrix(size_t startRow, size_t numRows)
    
+
    
 
 
    
-
    
-virtual void maxPoolBackward(Matrix &image, size_t imgSizeH, size_t imgSizeW, Matrix &outGrad, Matrix &outV, size_t sizeX, int start, size_t stride, size_t outputH, size_t outputW, real scaleTargets, real scaleOutput)
    
-
    Pooling backward operation. 
    
-
    
+
    
+virtual MatrixPtr subMatrix(size_t startRow, size_t numRows, MatrixPtr dest)
    
+
    
 
 
    
-
    
-virtual void avgPoolForward(Matrix &input, size_t imgSizeH, size_t imgSizeW, size_t channels, size_t sizeX, int start, size_t stride, size_t outputH, size_t outputW)
    
-
    Pooling forward operation, caculate the average of sizeX elements. 
    
+
    
+virtual void copyFrom(const real *src, size_t size)
    
+
    If this is GpuMatrix, src is assumed to be CPU memory
    
+
    If this is CpuMatrix, src is assumed to be CPU memory 
    
 
    
 
 
    
-
    
-virtual void avgPoolBackward(Matrix &input, size_t imgSizeH, size_t imgSizeW, size_t sizeX, int start, size_t stride, size_t outputH, size_t outputW, real scaleTargets, real scaleOutput)
    
+
    
+virtual void copyFrom(const real *src, const int64_t *seq)
    
 
    
 
 
    
-
    
-virtual void crossMapNormalFwd(Matrix &input, size_t imgSizeH, size_t imgSizeW, Matrix &denoms, size_t channels, size_t sizeX, float scale, float pow, bool blocked)
    
-
    normalize-operation. 
    
+
    
+virtual void copyFrom(const IVector &src)
    
+
    convert a int vector to a real matrix. 
    
+
    (1) source and dest are both in CPU.
    
+
    (2) sizes are exactly match. 
    
 
    
 
 
    
-
    
-virtual void crossMapNormalBwd(Matrix &localGrad, Matrix &denoms, Matrix &preOutV, Matrix &localOutV, size_t channels, size_t imgSizeH, size_t imgSizeW, size_t sizeX, float scale, float pow, bool blocked)
    
+
    
+virtual void copyByRowIndex(Matrix &b, IVector &rowIndex)
    
 
    
 
 
    
-
    
-virtual void maxSequenceForward(Matrix &input, const IVector &sequence, IVector &index)
    
-
    Input: one or more sequences. Each sequence contains some instances. Output: output size is the number of input sequences (NOT input instances). output[i] is set to max_{for each instance in this sequence}{input[i]} 
    
+
    
+virtual MatrixPtr clone(size_t height = 0, size_t width = 0, bool useGpu = false)
    
+
    Create a matrix with the same type (GpuMatrix, CpuMatrix, NonValueSparseMatrix, etc.) as this. 
    
+
    If height and width is zero, the new matrix will have the same size as this, otherwise the new matrix will have the specified size. 
    
 
    
 
 
    
-
    
-virtual void maxSequenceBackward(Matrix &outputGrad, const IVector &sequence, IVector &index)
    
+
    
+virtual real *getRowBuf(size_t row)
    
 
    
 
 
    
-
    
-virtual void contextProjectionForward(MatrixPtr input, MatrixPtr weight, const IVector &sequence, int contextLength, int contextStart, size_t beginPad, bool isPadding)
    
+
    
+virtual real getElement(size_t x, size_t y) const
    
 
    
 
 
    
-
    
-virtual void contextProjectionBackward(MatrixPtr inputGrad, MatrixPtr weightGrad, const IVector &sequence, int contextLength, int contextStart, size_t beginPad, bool isPadding)
    
+
    
+virtual real getSum()
    
 
    
 
 
    
-
    
-real *getRow(size_t row)
    
+
    
+virtual void accumulateColSum(Matrix &src)
    
 
    
 
 
    
-
    
-virtual real *getRowBuf(size_t row)
    
+
    
+virtual real getAbsSum()
    
 
    
 
 
    
-
    
-virtual void addBias(Matrix &b, real scale)
    
-
    add b to each sample of this. 
    
+
    
+virtual void resize(size_t newHeight, size_t newWidth) = 0
    
+
    
+
    Note
    
+
    Original data may not be preserved after resize(). 
    
+
    
+
    
 
    
 
 
    
-
    
-virtual void collectBias(Matrix &a, real scale)
    
-
    add each sample of a to this. 
    
+
    
+virtual void resize(size_t newHeight, size_t newWidth, size_t newNnz, SparseValueType valueType, SparseFormat format) = 0
    
+
    
+
    Note
    
+
    This should only be used for sparse matrix. 
    
+
    
+
    
 
    
 
 
    
-
    
-virtual void sequenceAvgForward(Matrix &a, const IVector &startsPos, int mode)
    
+
    
+virtual void setRow(size_t row, size_t colNum, const unsigned int *cols, const real *values) = 0
    
+
    This should only be used for sparse matrix. 
    
+
    Currently must be called for each row in order. The matrix is not valid until setRow is called for the last row. 
    
+
    
+
+
    
+
    
+virtual MatrixPtr getTranspose() = 0
    
 
    
 
 
    
-
    
-virtual void selectRows(Matrix &table, IVector &ids)
    
-
     
    this.row[i] += table.row[ids[i]]
-
    
-
    
-
    
+
    
+virtual void transpose(MatrixPtr matTrans, bool memAlloc)
    
+
    hard transpose. 
    
+
    allocate matTrans’ memory outside, then set memAlloc as false; else set as true. 
    
 
    
 
 
    
-
    
-virtual void addToRows(Matrix &table, IVector &ids)
    
-
     
    table.row[ids[i]] += this.row[i]
-
    
-
    
-
    
+
    
+virtual void clear()
    
+
    Only set all variables to 0 or NULL but not free them. 
    
 
    
 
 
    
-
    
-virtual void selectElements(Matrix &table, IVector &ids)
    
-
     
    this[i] = table[i, id[i]]
-
    
-
    
-
    
-
    
+
    
+void reshape(size_t height, size_t width)
    
+
    
 
 
    
-
    
-virtual void addElements(Matrix &table, IVector &ids)
    
-
     
    table[i, id[i]] += this[i]
-
    
-
    
-
    
+
    
+virtual void addBias(Matrix &b, real scale)
    
+
    add b to each sample of this. 
    
 
    
 
 
    
-
    
-template <typename TableMatType>
    
-
    
-void selectRowsImp(TableMatType &table, IVector &ids)
    
-
    use abstract getRow() to get row from table.
    
-
    Define table as template instead of virtual class for performance sake. internal used by above two virtual funcs. 
    
+
    
+virtual void collectBias(Matrix &a, real scale)
    
+
    add each sample from a to this. 
    
 
    
 
 
    
-
    
-template <typename TableMatType>
    
-
    
-void addToRowsImp(TableMatType &table, IVector &ids)
    
+
    
+virtual void sequenceAvgForward(Matrix &a, const IVector &startsPos, int mode)
    
 
    
 
 
    
-
    
-virtual void addColumnVector(const Matrix &b)
    
-
    Add a vector (column) b to matrix a, column by column. 
    
-
    
-
-
    
-
    
-virtual void mul(const MatrixPtr a, const MatrixPtr b, real scaleAB, real scaleT)
    
-
     
    this = scaleAB*(a*b) + scaleT*this
+
    
+virtual void mul(const MatrixPtr a, const MatrixPtr b, real scaleAB, real scaleT)
    
+
     
    this = scaleAB*(a*b) + scaleT*this
 
    
 
    
 
    
 
    
 
 
    
-
    
-void mul(CpuMatrix *a, CpuMatrix *b, real scaleAB, real scaleT)
    
-
    
+
    
+virtual void addColumnVector(const Matrix &b)
    
+
    Add a vector (column) b to matrix a, column by column. 
    
+
    
 
 
    
-
    
-void mul(CpuMatrix *a, CpuSparseMatrix *b, real scaleAB, real scaleT)
    
-
    
+
    
+virtual void addByBitCode(size_t numClasses, const IVector &codes, const Matrix &vec)
    
+
     
    For j < codeLength:
+  this(i, j) += vec(index(i, j), 0)
+where index(i, j) = ((codes(i) + numClasses) >> (j + 1)) - 1
+
    
+
    
+
    
+
    
 
 
    
-
    
-virtual void mul(CpuSparseMatrix *a, CpuMatrix *b, real scaleAB, real scaleT)
    
-
    
+
    
+virtual void addByBitCodeBackward(size_t numClasses, const IVector &codes, Matrix &vec)
    
+
     
    For j < codeLength:
+  vec(index(i, j), 0) += this(i, j)
+where index is same as the index for addByBitCode
+
    
+
    
+
    
+
    
 
 
    
-
    
-virtual void mul(const MatrixPtr a, const MatrixPtr b)
    
-
     
    this = a*b
+
    
+virtual void mulByBitCode(size_t numClasses, const IVector &codes, const Matrix &mat, const Matrix &input)
    
+
     
    For j < codeLength:
+  this(i, j) += <mat.row(index(i, j)), input.row(i)>
+where index is same as the index for addByBitCode
 
    
 
    
 
    
 
    
 
 
    
-
    
-virtual void rightMul(Matrix &b, real scaleAB, real scaleT)
    
-
     
    this = scaleAB*(this*b) +  scaleT*this
+
    
+virtual void mulByBitCodeBackwardWeight(size_t numClasses, const IVector &codes, Matrix &mat, const Matrix &input)
    
+
     
    For j < codeLength:
+  mat.row(index(i, j)) += this(i, j) * input.row(i)
+where index is same as the index for addByBitCode
 
    
 
    
 
    
 
    
 
 
    
-
    
-virtual void rightMul(Matrix &b)
    
-
     
    this = this* b
+
    
+virtual void mulByBitCodeBackwardError(size_t numClasses, const IVector &codes, const Matrix &mat, Matrix &input)
    
+
     
    For j < codeLength:
+  input.row(i) += this(i, j) * mat.row(index(i, j))
+where index is same as the index for addByBitCode
 
    
 
    
 
    
 
    
 
 
    
-
    
-virtual void leftMul(Matrix &a, real scaleAB, real scaleT)
    
-
     
    this = scaleAB*(a*this) +  scaleT*this
+
    
+virtual void sumByBitCode(size_t numClasses, IVector &codes, Matrix &sum, real scaleSum)
    
+
     
    For j < codeLength
+  sum(i, 0) = scaleSum * \sum_j  bit(i, j) * this(i, j)
+where bit(i, j) = ((codes(i) + numClasses) & 2^j) ? 1 : 0
 
    
 
    
 
    
 
    
 
 
    
-
    
-virtual void leftMul(Matrix &a)
    
-
     
    this = a*this)
+
    
+virtual void subByBitCode(size_t numClasses_, IVector &codes)
    
+
     
    For j < codeLength
+ this(i, j) -= bit(i, j)
+where bit(i, j) is same as that for sumByBitCode
 
    
 
    
 
    
 
    
 
 
    
-
    
-virtual void colMerge(Matrix &src)
    
-
    merge the element for each col. 
    
+
    
+virtual void rowSum(Matrix &sum)
    
+
    add the sum of each row of this to mat 
    
 
    
 
 
    
-
    
-virtual void rowSum(Matrix &sum)
    
-
    add the sum of each row of this to mat 
    
+
    
+virtual void rowMax(Matrix &max)
    
+
    set the max of each row of this to mat 
    
 
    
 
 
    
-
    
-virtual void rowMaxId(IVector &maxIds)
    
+
    
+virtual void colMax(Matrix &max)
    
 
    
 
 
    
-
    
-virtual void rowMax(Matrix &max)
    
-
    set the max of each row of this to mat 
    
-
    
+
    
+virtual void rowMaxId(IVector &maxIds)
    
+
    
 
 
    
-
    
-virtual void rowMax(IVector &maxIds, Matrix &max)
    
+
    
+virtual void rowMax(IVector &maxIds, Matrix &max)
    
 
    Get the top k elements of each row of this matrix. 
    
 
    The column ids and values of these elements are stored in maxIds and max respectively. Note that the top k elements are not sorted. 
    
 
    
 
 
    
-
    
-virtual void colMax(Matrix &max)
    
-
    
+
    
+virtual void rowNormalizeL1(Matrix &out)
    
+
    normalize each row so that the sum of each row is 1. 
    
+
    
 
 
    
-
    
-virtual void rowNormalizeL1(Matrix &out)
    
-
    normalize each row so that the sum of each row is 1. 
    
+
    
+virtual void mul(const MatrixPtr a, const MatrixPtr b)
    
+
     
    this = a*b
+
    
+
    
+
    
 
    
 
 
    
-
    
-virtual void oneHotCrossEntropy(Matrix &output, IVector &label)
    
+
    
+virtual void rightMul(Matrix &b, real scaleAB, real scaleT)
    
+
     
    this = scaleAB*(this*b) +  scaleT*this
+
    
+
    
+
    
+
    
+
+
    
+
    
+virtual void rightMul(Matrix &b)
    
+
     
    this = this* b
+
    
+
    
+
    
+
    
+
+
    
+
    
+virtual void leftMul(Matrix &a, real scaleAB, real scaleT)
    
+
     
    this = scaleAB*(a*this) +  scaleT*this
+
    
+
    
+
    
+
    
+
+
    
+
    
+virtual void leftMul(Matrix &a)
    
+
     
    this = a*this)
+
    
+
    
+
    
+
    
+
+
    
+
    
+virtual void colMerge(Matrix &src)
    
+
    merge the element for each col. 
    
+
    
+
+
    
+
    
+virtual void oneHotCrossEntropy(Matrix &output, IVector &label)
    
 
    copy -log(output[label]) to this->data[i]. 
    
 
    
 
 
    
-
    
-virtual void oneHotCrossEntropyBp(Matrix &outputV, IVector &label)
    
+
    
+virtual void oneHotCrossEntropyBp(Matrix &outputV, IVector &label)
    
 
    calculate the error of outputV according to label. 
    
 
    
 
 
    
-
    
-virtual void oneHotCrossEntropyWithSelfNorm(Matrix &output, IVector &label, real alpha)
    
+
    
+virtual void oneHotCrossEntropyWithSelfNorm(Matrix &output, IVector &label, real alpha)
    
 
    copy -log(output[label]) to this->data[i]. 
    
 
    
 
 
    
-
    
-virtual void oneHotCrossEntropyWithSelfNormBp(Matrix &outputV, IVector &label, real alpha)
    
+
    
+virtual void oneHotCrossEntropyWithSelfNormBp(Matrix &outputV, IVector &label, real alpha)
    
 
    calculate the error of outputV according to label. 
    
 
    
 
 
    
-
    
-virtual void circularConv(Matrix &b, Matrix &c)
    
+
    
+virtual void circularConv(Matrix &b, Matrix &c)
    
 
    
 \[ a[i] = \sum_{j=-(N-1)/2}^{(N-1)/2} b_{i+j} * c_{j} \]
    
 
    
@@ -2020,262 +2057,379 @@ SPARSE_CSC col   [0, 1, 2, 2, 4, 5];
 
    
 
 
    
-
    
-virtual void circularConvDerivative(Matrix &output, Matrix &prevOut1, Matrix &prevOut2, Matrix &prevGrad1, Matrix &prevGrad2)
    
+
    
+virtual void circularConvDerivative(Matrix &output, Matrix &prevOut1, Matrix &prevOut2, Matrix &prevGrad1, Matrix &prevGrad2)
    
 
    
 
 
    
-
    
-virtual void softmax(Matrix &output)
    
+
    
+virtual void softmax(Matrix &output)
    
 
    
 
 
    
-
    
-virtual void sequenceSoftmax(Matrix &output, const IVector &index)
    
+
    
+virtual void sequenceSoftmax(Matrix &output, const IVector &index)
    
 
    
 
 
    
-
    
-virtual void softmaxDerivative(Matrix &output, Matrix &sftmaxSum)
    
+
    
+virtual void softmaxBackward(Matrix &outputV)
    
 
    
 
 
    
-
    
-virtual void sumOfSquares(Matrix &output, Matrix &label)
    
+
    
+virtual void softmaxDerivative(Matrix &output, Matrix &sftmaxSum)
    
+
    
+
+
    
+
    
+virtual void sumOfSquares(Matrix &output, Matrix &label)
    
 
    calculate the sum of squares diff cost. 
    
 
    
 
 
    
-
    
-virtual void sumOfSquaresBp(Matrix &outputV, Matrix &label)
    
+
    
+virtual void sumOfSquaresBp(Matrix &outputV, Matrix &label)
    
 
    gradient of sumOfSquares. 
    
 
    
 
 
    
-
    
-virtual void tanh(Matrix &output)
    
+
    
+virtual void tanh(Matrix &output)
    
 
    
 
 
    
-
    
-virtual void tanhDerivative(Matrix &output)
    
+
    
+virtual void tanhDerivative(Matrix &output)
    
 
    
 
 
    
-
    
-virtual void softrelu(Matrix &output)
    
+
    
+virtual void softrelu(Matrix &output)
    
 
    
 
 
    
-
    
-virtual void softreluDerivative(Matrix &output)
    
+
    
+virtual void softreluDerivative(Matrix &output)
    
 
    
 
 
    
-
    
-virtual void scaledTanh(Matrix &output, real p1, real p2)
    
+
    
+virtual void scaledTanh(Matrix &output, real p1, real p2)
    
 
    
 
 
    
-
    
-virtual void cosSim(Matrix &output1, Matrix &output2, real scale)
    
+
    
+virtual void cosSim(Matrix &output1, Matrix &output2, real scale = 1.0f)
    
 
    cosine similarity, for each row i, this[i] = cos(output1[i], output2[i])
    
 
    output2 can only have one row, then for each row i, this[i] = cos(output1[i], output2[0]) 
    
 
    
 
 
    
-
    
-virtual void cosSimDerivative(Matrix &output, Matrix &prevOut1, Matrix &prevOut2, Matrix &prevGrad1, Matrix &prevGrad2, real scale)
    
+
    
+virtual void cosSimDerivative(Matrix &output, Matrix &prevOut1, Matrix &prevOut2, Matrix &prevGrad1, Matrix &prevGrad2, real scale = 1.0f)
    
 
    
 
 
    
-
    
-virtual void print(std::ostream &os) const
    
+
    
+virtual void print(std::ostream &os) const
    
 
    print out the values of elements to os 
    
 
    
 
 
    
-
    
-virtual void print(std::ostream &os, size_t height, size_t width) const
    
+
    
+virtual void print(std::ostream &os, size_t height, size_t width) const
    
 
    print a part of the matrix from the (top,left) value to the (height, width) value (not included) 
    
 
    
 
 
    
-
    
-virtual void printOneRow(std::ostream &os, size_t idx) const
    
+
    
+virtual void printOneRow(std::ostream &os, size_t idx) const
    
 
    print one row to os 
    
 
    
 
 
    
-
    
-virtual void paramReluForward(Matrix &data, Matrix &W)
    
-
    
-
-
    
-
    
-virtual void paramReluBackwardW(Matrix &oGrad, Matrix &data)
    
-
    
-
-
    
-
    
-virtual void paramReluBackwardDiff(Matrix &oGrad, Matrix &data, Matrix &W)
    
-
    
-
-
    
-
    
-virtual void check(std::ostream &os, Matrix &refMat, bool printDiff = true)
    
+
    
+virtual void check(std::ostream &os, Matrix &refMat, bool printDiff = true)
    
 
    
 
 
    
-
    
-virtual real getMin()
    
+
    
+virtual real getMin()
    
 
    
 
 
    
-
    
-virtual real getMax()
    
+
    
+virtual real getMax()
    
 
    
 
 
    
-
    
-virtual void randomizeUniform()
    
+
    
+virtual void randomizeUniform()
    
 
    
 
 
    
-
    
-virtual void classificationError(MatrixPtr output, IVectorPtr label)
    
+
    
+virtual void classificationError(MatrixPtr output, IVectorPtr label)
    
 
    calulate the error of classification 
    
 
    output[i] = 1 if row i is an error.
    
 
    output[i] = 0 if row i is correct. 
    
 
    
 
 
    
-
    
-virtual void addByBitCode(size_t numClasses, const IVector &codes, const Matrix &vec)
    
-
     
    For j < codeLength:
-  this(i, j) += vec(index(i, j), 0)
-where index(i, j) = ((codes(i) + numClasses) >> (j + 1)) - 1
-
    
-
    
-
    
+
    
+virtual void convExpand(Matrix &feature, int feaImgHeight, int feaImgWidth, int channels, int blockH, int blockW, int strideH, int strideW, int paddingH, int paddingW, int outputH, int outputW)
    
+
    This function is used to calculate the convolution:
    
+
    It will expand a feature matrix according to the convolution filters 
    
 
    
 
 
    
-
    
-virtual void addByBitCodeBackward(size_t numClasses, const IVector &codes, Matrix &vec)
    
-
     
    For j < codeLength:
-  vec(index(i, j), 0) += this(i, j)
-where index is same as the index for addByBitCode
-
    
-
    
-
    
+
    
+virtual void convShrink(Matrix &expandColMat, int thisImgHeight, int thisImgWidth, int channels, int blockH, int blockW, int strideH, int strideW, int paddingH, int paddingW, int outputH, int outputW, real alpha = 1.0f, real beta = 0.0f)
    
+
    This function is the reverse implementation of convExpand:
    
+
    Its function is to restore a expanded-matrix into a feature matrix 
    
 
    
 
 
    
-
    
-virtual void mulByBitCode(size_t numClasses, const IVector &codes, const Matrix &mat, const Matrix &input)
    
-
     
    For j < codeLength:
-  this(i, j) += <mat.row(index(i, j)), input.row(i)>
-where index is same as the index for addByBitCode
-
    
-
    
-
    
+
    
+virtual void maxPoolForward(Matrix &inputMat, size_t imgSizeH, size_t imgSizeW, size_t channels, size_t sizeX, int start_, size_t stride, size_t outputH, size_t outputW)
    
+
    Pooling forward operation, pick out the largest element in the sizeX of value 
    
 
    
 
 
    
-
    
-virtual void mulByBitCodeBackwardWeight(size_t numClasses, const IVector &codes, Matrix &mat, const Matrix &input)
    
-
     
    For j < codeLength:
-  mat.row(index(i, j)) += this(i, j) * input.row(i)
-where index is same as the index for addByBitCode
-
    
-
    
-
    
+
    
+virtual void maxPoolBackward(Matrix &image, size_t imgSizeH, size_t imgSizeW, Matrix &outGrad, Matrix &outV, size_t sizeX, int start, size_t stride, size_t outputH, size_t outputW, real scaleTargets, real scaleOutput)
    
+
    Pooling backward operation. 
    
 
    
 
 
    
-
    
-virtual void mulByBitCodeBackwardError(size_t numClasses, const IVector &codes, const Matrix &mat, Matrix &input)
    
-
     
    For j < codeLength:
-  input.row(i) += this(i, j) * mat.row(index(i, j))
-where index is same as the index for addByBitCode
-
    
-
    
-
    
+
    
+virtual void avgPoolForward(Matrix &input, size_t imgSizeH, size_t imgSizeW, size_t channels, size_t sizeX, int start, size_t stride, size_t outputH, size_t outputW)
    
+
    Pooling forward operation, caculate the average of sizeX elements. 
    
 
    
 
 
    
-
    
-virtual void sumByBitCode(size_t numClasses, IVector &codes, Matrix &sum, real scaleSum)
    
-
     
    For j < codeLength
-  sum(i, 0) = scaleSum * \sum_j  bit(i, j) * this(i, j)
-where bit(i, j) = ((codes(i) + numClasses) & 2^j) ? 1 : 0
-
    
-
    
-
    
-
    
+
    
+virtual void avgPoolBackward(Matrix &input, size_t imgSizeH, size_t imgSizeW, size_t sizeX, int start, size_t stride, size_t outputH, size_t outputW, real scaleTargets, real scaleOutput)
    
+
    
 
 
    
-
    
-virtual void subByBitCode(size_t numClasses_, IVector &codes)
    
-
     
    For j < codeLength
- this(i, j) -= bit(i, j)
-where bit(i, j) is same as that for sumByBitCode
-
    
-
    
-
    
+
    
+virtual void crossMapNormalFwd(Matrix &input, size_t imgSizeH, size_t imgSizeW, Matrix &denoms, size_t channels, size_t sizeX, float scale, float pow)
    
+
    normalize-operation. 
    
 
    
 
 
    
-
    
-virtual void multiBinaryLabelCrossEntropy(Matrix &output, Matrix &label)
    
-
    cross entropy for multi binary labels 
    
-
     
    this[i] = -sum(label[i][j]*log(output[i][j])
-          + (1-label[i][j])*log(1-output[i][j]))
-
    
-
    
-
    
+
    
+virtual void crossMapNormalBwd(Matrix &localGrad, Matrix &denoms, Matrix &preOutV, Matrix &localOutV, size_t channels, size_t imgSizeH, size_t imgSizeW, size_t size, float scale, float pow)
    
+
    
+
+
    
+
    
+virtual void maxSequenceForward(Matrix &input, const IVector &sequence, IVector &index)
    
+
    Input: one or more sequences. Each sequence contains some instances.
    
+
    Output: output size is the number of input sequences (NOT input instances).
    
+
    output[i] is set to max_input[i]. 
    
 
    
 
 
    
-
    
-virtual void multiBinaryLabelCrossEntropyBp(Matrix &output, Matrix &label)
    
+
    
+virtual void maxSequenceBackward(Matrix &outputGrad, const IVector &sequence, IVector &index)
    
+
    
+
+
    
+
    
+virtual void contextProjectionForward(MatrixPtr input, MatrixPtr weight, const IVector &sequence, int contextLength, int contextStart, size_t beginPad, bool isPadding)
    
+
    
+
+
    
+
    
+virtual void contextProjectionBackward(MatrixPtr inputGrad, MatrixPtr weightGrad, const IVector &sequence, int contextLength, int contextStart, size_t beginPad, bool isPadding)
    
+
    
+
+
    
+
    
+virtual void contextProjectionBackwardData(MatrixPtr inputGrad, const IVector &sequence, int contextLength, int contextStart)
    
+
    
+
+
    
+
    
+virtual void contextProjectionBackwardWeight(MatrixPtr weightGrad, const IVector &sequence, int contextLength, int contextStart, int totalPad, size_t beginPad)
    
+
    
+
+
    
+
    
+virtual void selectRows(Matrix &table, IVector &ids)
    
+
     
    this.row[i] += table.row[ids[i]]
+if ids[i] == -1, it will be ignored
+
    
+
    
+
    
+
    
+
+
    
+
    
+virtual void selectElements(Matrix &table, IVector &ids)
    
+
     
    this[i] = table[i, id[i]]
+
    
+
    
+
    
+
    
+
+
    
+
    
+virtual void addToRows(Matrix &table, IVector &ids)
    
+
     
    table.row[ids[i]] += this.row[i]
+if ids[i] == -1, it will be ignored
+
    
+
    
+
    
+
    
+
+
    
+
    
+virtual void addElements(Matrix &table, IVector &ids)
    
+
     
    table[i, id[i]] += this[i]
+
    
+
    
+
    
+
    
+
+
    
+
    
+virtual void multiBinaryLabelCrossEntropy(Matrix &output, Matrix &label)
    
+
    cross entropy for multi binary labels 
    
+
     
    this[i] = -sum(label[i][j]*log(output[i][j])
+          + (1-label[i][j])*log(1-output[i][j]))
+
    
+
    
+
    
+
    
+
+
    
+
    
+virtual void multiBinaryLabelCrossEntropyBp(Matrix &output, Matrix &label)
    
 
    The gradient of cross entropy for multi binary labels on output. 
    
-
     
    this[i][j] = -label[i][j]/output[i][j]
-             + (1-label[i][j])/(1-output[i][j])
+
     
    this[i][j] = -label[i][j]/output[i][j]
+             + (1-label[i][j])/(1-output[i][j])
 
    
 
    
 
    
 
    
 
 
    
-
    
-virtual void classificationErrorMulti(Matrix &output, Matrix &label, real threshold)
    
+
    
+virtual void classificationErrorMulti(Matrix &output, Matrix &label, real threshold)
    
 
    Calculate the classification error for multi binary labels. 
    
-
     
    this[i] = sum((output[i][j] >= threshold && label[i][j] == 0)
-           || (output[i][j] < threshold && label[i][j] == 1))
-           / output->getWidth()
+
     
    this[i] = sum((output[i][j] >= threshold && label[i][j] == 0)
+           || (output[i][j] < threshold && label[i][j] == 1))
+           / output->getWidth()
 
    
 
    
 
    
 
    
 
+
    
+
    
+virtual void paramReluForward(Matrix &data, Matrix &W)
    
+
    
+
+
    
+
    
+virtual void paramReluBackwardW(Matrix &oGrad, Matrix &data)
    
+
    
+
+
    
+
    
+virtual void paramReluBackwardDiff(Matrix &oGrad, Matrix &data, Matrix &W)
    
+
    
+
+
    
+
    
+
    Public Members
    
+
    
+
    
+size_t elementCnt_
    
+
    
+
+
    
+
    
+MemoryHandlePtr memoryHandle_
    
+
    
+
 
    
 
    
 
    Public Static Functions
    
 
    
-
    
-void mul(CpuMatrix *a, CpuMatrix *b, CpuSparseMatrix *c, real scaleAB, real scaleT)
    
+
    
+MatrixPtr create(MemoryHandlePtr memHandle, size_t height, size_t width, bool trans = false)
    
 
    
 
 
    
-
    
-template <typename MatBType, typename MatCType>
    
-
    
-void mul(CpuSparseMatrix *a, MatBType *b, MatCType *c, real scaleAB, real scaleT)
    
-
    c = a * b
    
-
    use abstract getRow() to get row from B,C. Define B,C as template instead of virtual class for performance sake. 
    
-
    
+
    
+MatrixPtr create(size_t height, size_t width, bool trans = false, bool useGpu = false)
    
+
    
+
+
    
+
    
+MatrixPtr create(real *data, size_t height, size_t width, bool trans = false, bool useGpu = false)
    
+
    
+
+
    
+
    
+MatrixPtr create(real *data, size_t height, size_t width, size_t stride, bool trans = false, bool useGpu = false)
    
+
    
+
+
    
+
    
+MatrixPtr createSparseMatrix(size_t height, size_t width, size_t nnz, SparseValueType valueType = FLOAT_VALUE, bool trans = false, bool useGpu = false)
    
+
    
+
+
    
+
    
+MatrixPtr createSparseMatrix(size_t height, size_t width, size_t nnz, SparseValueType valueType = FLOAT_VALUE, SparseFormat foramt = SPARSE_CSR, bool trans = false, bool useGpu = false)
    
+
    
+
+
    
+
    
+MatrixPtr createSparseMatrix(real *data, int *row, int *col, size_t height, size_t width, size_t nnz, SparseValueType valueType, SparseFormat format, bool trans, bool useGpu)
    
+
    
+
+
    
+
    
+void resizeOrCreateSparseMatrix(MatrixPtr &matrix, size_t height, size_t width, size_t nnz, SparseValueType valueType = FLOAT_VALUE, SparseFormat foramt = SPARSE_CSR, bool trans = false, bool useGpu = false)
    
+
    
+
+
    
+
    
+void resizeOrCreate(MatrixPtr &a, size_t height, size_t width, bool trans = false, bool useGpu = false)
    
+
    
+
+
    
+
    
+
    Protected Functions
    
+
    
+
    
+Matrix(MemoryHandlePtr memHandle, size_t height, size_t width, bool trans, bool use_gpu)
    
+
    
+
+
    
+
    
+Matrix(real *data, size_t height, size_t width, bool trans, bool use_gpu)
    
+
    
+
+
    
+
    
+Matrix(real *data, size_t height, size_t width, size_t stride, bool trans, bool use_gpu)
    
+
    
+
+
    
+
    
+
    Protected Static Attributes
    
+
    
+
    
+ThreadLocal<MatrixPtr> tmpMat_
    
+
    
 
 
    
 
    
@@ -2283,7 +2437,7 @@ where bit(i, j) is same as that for sumByBitCode
 
    
 
    
 class GpuMatrix
    
-
    Inherits from paddle::Matrix
    
+
    Inherits from paddle::Matrix
    
 
    
 
    Public Functions
    
 
    
@@ -2331,7 +2485,7 @@ where bit(i, j) is same as that for sumByBitCode
 virtual void resize(size_t newHeight, size_t newWidth)
 
    
 
    Note
    
-
    Original data may not be preserved after resize(). 
    
+
    Original data may not be preserved after resize(). 
    
 
    
 
    
 
    
@@ -2390,7 +2544,7 @@ where bit(i, j) is same as that for sumByBitCode
 
    
 
    
 virtual MatrixPtr clone(size_t height, size_t width, bool useGpu = false)
    
-
    Create a matrix with the same type (GpuMatrix, CpuMatrix, NonValueSparseMatrix, etc.) as this. 
    
+
    Create a matrix with the same type (GpuMatrix, CpuMatrix, NonValueSparseMatrix, etc.) as this. 
    
 
    If height and width is zero, the new matrix will have the same size as this, otherwise the new matrix will have the specified size. 
    
 
    
 
@@ -2445,7 +2599,7 @@ where bit(i, j) is same as that for sumByBitCode
 
    
 
    
 virtual void collectBias(Matrix &a, real scale)
    
-
     
    add each sample from a to this.
+
     
    add each sample from a to this.
 
    
 
    
 
    
@@ -2459,7 +2613,7 @@ where bit(i, j) is same as that for sumByBitCode
 
    
 
    
 virtual void selectRows(Matrix &table, IVector &ids)
    
-
     
    this.row[i] += table.row[ids[i]]
+
     
    this.row[i] += table.row[ids[i]]
 
    
 
    
 
    
@@ -2468,7 +2622,7 @@ where bit(i, j) is same as that for sumByBitCode
 
    
 
    
 virtual void selectElements(Matrix &table, IVector &ids)
    
-
     
    this[i] = table[i, id[i]]
+
     
    this[i] = table[i, id[i]]
 
    
 
    
 
    
@@ -2477,7 +2631,7 @@ where bit(i, j) is same as that for sumByBitCode
 
    
 
    
 virtual void addToRows(Matrix &table, IVector &ids)
    
-
     
    table.row[ids[i]] += this.row[i]
+
     
    table.row[ids[i]] += this.row[i]
 
    
 
    
 
    
@@ -2492,7 +2646,7 @@ where bit(i, j) is same as that for sumByBitCode
 
    
 
    
 virtual void mul(const MatrixPtr a, const MatrixPtr b, real scaleAB, real scaleT)
    
-
     
    this = scaleAB*(a*b) + scaleT*this
+
     
    this = scaleAB*(a*b) + scaleT*this
 
    
 
    
 
    
@@ -2501,7 +2655,7 @@ where bit(i, j) is same as that for sumByBitCode
 
    
 
    
 virtual void mul(const MatrixPtr a, const MatrixPtr b)
    
-
     
    this = a*b
+
     
    this = a*b
 
    
 
    
 
    
@@ -2509,23 +2663,23 @@ where bit(i, j) is same as that for sumByBitCode
 
 
    
 
    
-void mul(const GpuMatrix &a, const GpuMatrix &b, real scaleAB, real scaleT)
    
+void mul(const GpuMatrix &a, const GpuMatrix &b, real scaleAB, real scaleT)
 
    
 
 
    
 
    
-void mul(const GpuSparseMatrix &a, const GpuMatrix &b, real scaleAB, real scaleT)
    
+void mul(const GpuSparseMatrix &a, const GpuMatrix &b, real scaleAB, real scaleT)
 
    
 
 
    
 
    
-void mul(const GpuMatrix &a, const GpuSparseMatrix &b, real scaleAB, real scaleT)
    
+void mul(const GpuMatrix &a, const GpuSparseMatrix &b, real scaleAB, real scaleT)
 
    
 
 
    
 
    
 virtual void rightMul(Matrix &b, real scaleAB, real scaleT)
    
-
     
    this = scaleAB*(this*b) +  scaleT*this
+
     
    this = scaleAB*(this*b) +  scaleT*this
 
    
 
    
 
    
@@ -2534,7 +2688,7 @@ where bit(i, j) is same as that for sumByBitCode
 
    
 
    
 virtual void rightMul(Matrix &b)
    
-
     
    this = this* b
+
     
    this = this* b
 
    
 
    
 
    
@@ -2543,7 +2697,7 @@ where bit(i, j) is same as that for sumByBitCode
 
    
 
    
 virtual void leftMul(Matrix &a, real scaleAB, real scaleT)
    
-
     
    this = scaleAB*(a*this) +  scaleT*this
+
     
    this = scaleAB*(a*this) +  scaleT*this
 
    
 
    
 
    
@@ -2552,7 +2706,7 @@ where bit(i, j) is same as that for sumByBitCode
 
    
 
    
 virtual void leftMul(Matrix &a)
    
-
     
    this = a*this
+
     
    this = a*this
 
    
 
    
 
    
@@ -2764,14 +2918,14 @@ where bit(i, j) is same as that for sumByBitCode
 
    
 
 
    
-
    
-virtual void crossMapNormalFwd(Matrix &input, size_t imgSizeH, size_t imgSizeW, Matrix &denoms, size_t channels, size_t sizeX, float scale, float pow, bool blocked)
    
+
    
+virtual void crossMapNormalFwd(Matrix &input, size_t imgSizeH, size_t imgSizeW, Matrix &denoms, size_t channels, size_t sizeX, float scale, float pow)
    
 
    normalize-operation. 
    
 
    
 
 
    
-
    
-virtual void crossMapNormalBwd(Matrix &localGrad, Matrix &denoms, Matrix &preOutV, Matrix &localOutV, size_t channels, size_t imgSizeH, size_t imgSizeW, size_t sizeX, float scale, float pow, bool blocked)
    
+
    
+virtual void crossMapNormalBwd(Matrix &localGrad, Matrix &denoms, Matrix &preOutV, Matrix &localOutV, size_t channels, size_t imgSizeH, size_t imgSizeW, size_t sizeX, float scale, float pow)
    
 
    
 
 
    
@@ -2806,482 +2960,445 @@ where bit(i, j) is same as that for sumByBitCode
 
    
 
 
    
-
    
-class Matrix
    
-
    #include <Matrix.h>
    Copy or assignemnt constructor will share the data as opposed to making a copy of the original data. To make a copy of the orinal data, use copyFrom() instead. 
    
-
    Inherits from paddle::BaseMatrixT< real >
    
-
    Subclassed by paddle::CpuMatrix, paddle::CpuSparseMatrix, paddle::GpuMatrix, paddle::GpuSparseMatrix
    
+
    
+class CpuMatrix
    
+
    Inherits from paddle::Matrix
    
+
    Subclassed by paddle::SharedCpuMatrix, paddle::SparseRowCpuMatrix, paddle::SparseRowIdsCpuMatrix
    
 
    
 
    Public Functions
    
 
    
-
    
-virtual ~Matrix()
    
-
    
-
-
    
-
    
-void setData(real *data)
    
-
    set the data buffer used to hold the matrix data. 
    
-
    caller should make sure that the size of data is at least sizeof(real)*height*width. 
    
+
    
+CpuMatrix(size_t height, size_t width, bool trans = false)
    
+
    CpuMatrix 
    
 
    
 
 
    
-
    
-void setData(real *data, size_t newHeight, size_t newWidth)
    
-
    the data should be contiguous 
    
-
    
+
    
+CpuMatrix(real *data, size_t height, size_t width, bool trans = false)
    
+
    
 
 
    
-
    
-size_t getWidth() const
    
+
    
+CpuMatrix(real *data, size_t height, size_t width, size_t stride, bool trans = false)
    
 
    
 
 
    
-
    
-size_t getHeight() const
    
+
    
+CpuMatrix(CpuMemHandlePtr dataHandle, size_t height, size_t width, bool trans = false)
    
 
    
 
 
    
-
    
-size_t getStride() const
    
+
    
+~CpuMatrix()
    
 
    
 
 
    
-
    
-size_t getElementCnt() const
    
+
    
+virtual void zeroMem()
    
 
    
 
 
    
-
    
-virtual real *getData()
    
+
    
+virtual void resetOne()
    
 
    
 
 
    
-
    
-virtual const real *getData() const
    
-
    
+
    
+virtual void resize(size_t newHeight, size_t newWidth)
    
+
    
+
    Note
    
+
    Original data may not be preserved after resize(). 
    
+
    
+
    
+
    
 
 
    
-
    
-bool isTransposed() const
    
-
    
+
    
+virtual void resize(size_t newHeight, size_t newWidth, size_t newNnz, SparseValueType valueType, SparseFormat format)
    
+
    
+
    Note
    
+
    This should only be used for sparse matrix. 
    
+
    
+
    
+
    
 
 
    
-
    
-bool isContiguous() const
    
-
    
+
    
+virtual void setRow(size_t row, size_t colNum, const unsigned int *cols, const real *values)
    
+
    This should only be used for sparse matrix. 
    
+
    Currently must be called for each row in order. The matrix is not valid until setRow is called for the last row. 
    
+
    
 
 
    
-
    
-virtual int *getRows() const
    
+
    
+virtual real getElement(size_t x, size_t y) const
    
 
    
 
 
    
-
    
-virtual int *getCols() const
    
+
    
+virtual real getSum()
    
 
    
 
 
    
-
    
-virtual SparseFormat getFormat() const
    
+
    
+virtual void accumulateColSum(Matrix &src)
    
 
    
 
 
    
-
    
-virtual SparseValueType getValueType() const
    
+
    
+virtual real getAbsSum()
    
 
    
 
 
    
-
    
-virtual void add3(MatrixPtr b)
    
-
    matrix elment-wise add 
    
-
    Named add3 just because add/add2 has been used in BaseMatrix.cu and they are not virtual function. 
    
-
    
+
    
+virtual MatrixPtr getTranspose()
    
+
    
 
 
    
-
    
-MemoryHandlePtr getMemoryHandle() const
    
-
    
+
    
+virtual void transpose(MatrixPtr matTrans, bool memAlloc)
    
+
    hard transpose. 
    
+
    allocate matTrans’ memory outside, then set memAlloc as false; else set as true. 
    
+
    
 
 
    
-
    
-virtual void zeroMem()
    
+
    
+virtual void copyFrom(const Matrix &src)
    
 
    
 
 
    
-
    
-virtual void resetOne()
    
+
    
+virtual void copyFrom(const Matrix &src, hl_stream_t stream)
    
 
    
 
 
    
-
    
-virtual void copyFrom(const Matrix &src)
    
-
    
+
    
+virtual void copyFrom(const real *src, size_t size)
    
+
    If this is GpuMatrix, src is assumed to be CPU memory
    
+
    If this is CpuMatrix, src is assumed to be CPU memory 
    
+
    
 
 
    
-
    
-virtual void trimFrom(const CpuSparseMatrix &src)
    
+
    
+virtual void copyFrom(const real *cpuSrc, const int64_t *seq)
    
 
    
 
 
    
-
    
-virtual void copyFrom(const Matrix &src, hl_stream_t stream)
    
-
    
+
    
+virtual void copyFrom(const IVector &src)
    
+
    convert a int vector to a real matrix. 
    
+
    (1) source and dest are both in CPU.
    
+
    (2) sizes are exactly match. 
    
+
    
 
 
    
-
    
-MatrixPtr subMatrix(size_t startRow, size_t endRow, size_t startCol, size_t endCol)
    
+
    
+void copyFrom(CpuSparseMatrix &src)
    
 
    
 
 
    
-
    
-MatrixPtr subRowMatrix(size_t startRow, size_t endRow)
    
+
    
+virtual void copyByRowIndex(Matrix &b, IVector &rowIndex)
    
 
    
 
 
    
-
    
-MatrixPtr subColMatrix(size_t startCol, size_t endCol)
    
-
    
+
    
+virtual MatrixPtr clone(size_t height, size_t width, bool useGpu = false)
    
+
    Create a matrix with the same type (GpuMatrix, CpuMatrix, NonValueSparseMatrix, etc.) as this. 
    
+
    If height and width is zero, the new matrix will have the same size as this, otherwise the new matrix will have the specified size. 
    
+
    
 
 
    
-
    
-virtual MatrixPtr subMatrix(size_t startRow, size_t numRows)
    
-
    
+
    
+virtual void convExpand(Matrix &feature, int feaImgHeight, int feaImgWidth, int channels, int blockH, int blockW, int strideH, int strideW, int paddingH, int paddingW, int outputH, int outputW)
    
+
    This function is used to calculate the convolution:
    
+
    It will expand a feature matrix according to the convolution filters 
    
+
    
 
 
    
-
    
-virtual MatrixPtr subMatrix(size_t startRow, size_t numRows, MatrixPtr dest)
    
-
    
+
    
+virtual void convShrink(Matrix &expandColMat, int thisImgHeight, int thisImgWidth, int channels, int blockH, int blockW, int strideH, int strideW, int paddingH, int paddingW, int outputH, int outputW, real alpha = 1.0f, real beta = 0.0f)
    
+
    This function is the reverse implementation of convExpand:
    
+
    Its function is to restore a expanded-matrix into a feature matrix 
    
+
    
 
 
    
-
    
-virtual void copyFrom(const real *src, size_t size)
    
-
    If this is GpuMatrix, src is assumed to be CPU memory
    
-
    If this is CpuMatrix, src is assumed to be CPU memory 
    
+
    
+virtual void maxPoolForward(Matrix &inputMat, size_t imgSizeH, size_t imgSizeW, size_t channels, size_t sizeX, int start_, size_t stride, size_t outputH, size_t outputW)
    
+
    Pooling forward operation, pick out the largest element in the sizeX of value 
    
 
    
 
 
    
-
    
-virtual void copyFrom(const real *src, const int64_t *seq)
    
-
    
+
    
+virtual void maxPoolBackward(Matrix &image, size_t imgSizeH, size_t imgSizeW, Matrix &outGrad, Matrix &outV, size_t sizeX, int start, size_t stride, size_t outputH, size_t outputW, real scaleTargets, real scaleOutput)
    
+
    Pooling backward operation. 
    
+
    
 
 
    
-
    
-virtual void copyFrom(const IVector &src)
    
-
    convert a int vector to a real matrix. 
    
-
    (1) source and dest are both in CPU.
    
-
    (2) sizes are exactly match. 
    
+
    
+virtual void avgPoolForward(Matrix &input, size_t imgSizeH, size_t imgSizeW, size_t channels, size_t sizeX, int start, size_t stride, size_t outputH, size_t outputW)
    
+
    Pooling forward operation, caculate the average of sizeX elements. 
    
 
    
 
 
    
-
    
-virtual void copyByRowIndex(Matrix &b, IVector &rowIndex)
    
+
    
+virtual void avgPoolBackward(Matrix &input, size_t imgSizeH, size_t imgSizeW, size_t sizeX, int start, size_t stride, size_t outputH, size_t outputW, real scaleTargets, real scaleOutput)
    
 
    
 
 
    
-
    
-virtual MatrixPtr clone(size_t height = 0, size_t width = 0, bool useGpu = false)
    
-
    Create a matrix with the same type (GpuMatrix, CpuMatrix, NonValueSparseMatrix, etc.) as this. 
    
-
    If height and width is zero, the new matrix will have the same size as this, otherwise the new matrix will have the specified size. 
    
+
    
+virtual void crossMapNormalFwd(Matrix &input, size_t imgSizeH, size_t imgSizeW, Matrix &denoms, size_t channels, size_t sizeX, float scale, float pow)
    
+
    normalize-operation. 
    
 
    
 
 
    
-
    
-virtual real *getRowBuf(size_t row)
    
+
    
+virtual void crossMapNormalBwd(Matrix &localGrad, Matrix &denoms, Matrix &preOutV, Matrix &localOutV, size_t channels, size_t imgSizeH, size_t imgSizeW, size_t sizeX, float scale, float pow)
    
 
    
 
 
    
-
    
-virtual real getElement(size_t x, size_t y) const
    
+
    
+virtual void maxSequenceForward(Matrix &input, const IVector &sequence, IVector &index)
    
+
    Input: one or more sequences. Each sequence contains some instances. Output: output size is the number of input sequences (NOT input instances). output[i] is set to max_{for each instance in this sequence}{input[i]} 
    
+
    
+
+
    
+
    
+virtual void maxSequenceBackward(Matrix &outputGrad, const IVector &sequence, IVector &index)
    
 
    
 
 
    
-
    
-virtual real getSum()
    
+
    
+virtual void contextProjectionForward(MatrixPtr input, MatrixPtr weight, const IVector &sequence, int contextLength, int contextStart, size_t beginPad, bool isPadding)
    
 
    
 
 
    
-
    
-virtual void accumulateColSum(Matrix &src)
    
+
    
+virtual void contextProjectionBackward(MatrixPtr inputGrad, MatrixPtr weightGrad, const IVector &sequence, int contextLength, int contextStart, size_t beginPad, bool isPadding)
    
 
    
 
 
    
-
    
-virtual real getAbsSum()
    
+
    
+real *getRow(size_t row)
    
 
    
 
 
    
-
    
-virtual void resize(size_t newHeight, size_t newWidth) = 0
    
-
    
-
    Note
    
-
    Original data may not be preserved after resize(). 
    
-
    
-
    
-
    
+
    
+virtual real *getRowBuf(size_t row)
    
+
    
 
 
    
-
    
-virtual void resize(size_t newHeight, size_t newWidth, size_t newNnz, SparseValueType valueType, SparseFormat format) = 0
    
-
    
-
    Note
    
-
    This should only be used for sparse matrix. 
    
-
    
-
    
+
    
+virtual void addBias(Matrix &b, real scale)
    
+
    add b to each sample of this. 
    
 
    
 
 
    
-
    
-virtual void setRow(size_t row, size_t colNum, const unsigned int *cols, const real *values) = 0
    
-
    This should only be used for sparse matrix. 
    
-
    Currently must be called for each row in order. The matrix is not valid until setRow is called for the last row. 
    
+
    
+virtual void collectBias(Matrix &a, real scale)
    
+
    add each sample of a to this. 
    
 
    
 
 
    
-
    
-virtual MatrixPtr getTranspose() = 0
    
+
    
+virtual void sequenceAvgForward(Matrix &a, const IVector &startsPos, int mode)
    
 
    
 
 
    
-
    
-virtual void transpose(MatrixPtr matTrans, bool memAlloc)
    
-
    hard transpose. 
    
-
    allocate matTrans’ memory outside, then set memAlloc as false; else set as true. 
    
+
    
+virtual void selectRows(Matrix &table, IVector &ids)
    
+
     
    this.row[i] += table.row[ids[i]]
+
    
+
    
+
    
 
    
 
 
    
-
    
-virtual void clear()
    
-
    Only set all variables to 0 or NULL but not free them. 
    
+
    
+virtual void addToRows(Matrix &table, IVector &ids)
    
+
     
    table.row[ids[i]] += this.row[i]
+
    
+
    
+
    
 
    
 
 
    
-
    
-void reshape(size_t height, size_t width)
    
-
    
+
    
+virtual void selectElements(Matrix &table, IVector &ids)
    
+
     
    this[i] = table[i, id[i]]
+
    
+
    
+
    
+
    
 
 
    
-
    
-virtual void addBias(Matrix &b, real scale)
    
-
    add b to each sample of this. 
    
+
    
+virtual void addElements(Matrix &table, IVector &ids)
    
+
     
    table[i, id[i]] += this[i]
+
    
+
    
+
    
 
    
 
 
    
-
    
-virtual void collectBias(Matrix &a, real scale)
    
-
    add each sample from a to this. 
    
+
    
+template <typename TableMatType>
    
+
    
+void selectRowsImp(TableMatType &table, IVector &ids)
    
+
    use abstract getRow() to get row from table.
    
+
    Define table as template instead of virtual class for performance sake. internal used by above two virtual funcs. 
    
 
    
 
 
    
-
    
-virtual void sequenceAvgForward(Matrix &a, const IVector &startsPos, int mode)
    
+
    
+template <typename TableMatType>
    
+
    
+void addToRowsImp(TableMatType &table, IVector &ids)
    
 
    
 
 
    
-
    
-virtual void mul(const MatrixPtr a, const MatrixPtr b, real scaleAB, real scaleT)
    
-
     
    this = scaleAB*(a*b) + scaleT*this
+
    
+virtual void addColumnVector(const Matrix &b)
    
+
    Add a vector (column) b to matrix a, column by column. 
    
+
    
+
+
    
+
    
+virtual void mul(const MatrixPtr a, const MatrixPtr b, real scaleAB, real scaleT)
    
+
     
    this = scaleAB*(a*b) + scaleT*this
 
    
 
    
 
    
 
    
 
 
    
-
    
-virtual void addColumnVector(const Matrix &b)
    
-
    Add a vector (column) b to matrix a, column by column. 
    
-
    
+
    
+void mul(CpuMatrix *a, CpuMatrix *b, real scaleAB, real scaleT)
    
+
    
 
 
    
-
    
-virtual void addByBitCode(size_t numClasses, const IVector &codes, const Matrix &vec)
    
-
     
    For j < codeLength:
-  this(i, j) += vec(index(i, j), 0)
-where index(i, j) = ((codes(i) + numClasses) >> (j + 1)) - 1
+
    
+void mul(CpuMatrix *a, CpuSparseMatrix *b, real scaleAB, real scaleT)
    
+
    
+
+
    
+
    
+virtual void mul(CpuSparseMatrix *a, CpuMatrix *b, real scaleAB, real scaleT)
    
+
    
+
+
    
+
    
+virtual void mul(const MatrixPtr a, const MatrixPtr b)
    
+
     
    this = a*b
 
    
 
    
 
    
 
    
 
 
    
-
    
-virtual void addByBitCodeBackward(size_t numClasses, const IVector &codes, Matrix &vec)
    
-
     
    For j < codeLength:
-  vec(index(i, j), 0) += this(i, j)
-where index is same as the index for addByBitCode
+
    
+virtual void rightMul(Matrix &b, real scaleAB, real scaleT)
    
+
     
    this = scaleAB*(this*b) +  scaleT*this
 
    
 
    
 
    
 
    
 
 
    
-
    
-virtual void mulByBitCode(size_t numClasses, const IVector &codes, const Matrix &mat, const Matrix &input)
    
-
     
    For j < codeLength:
-  this(i, j) += <mat.row(index(i, j)), input.row(i)>
-where index is same as the index for addByBitCode
+
    
+virtual void rightMul(Matrix &b)
    
+
     
    this = this* b
 
    
 
    
 
    
 
    
 
 
    
-
    
-virtual void mulByBitCodeBackwardWeight(size_t numClasses, const IVector &codes, Matrix &mat, const Matrix &input)
    
-
     
    For j < codeLength:
-  mat.row(index(i, j)) += this(i, j) * input.row(i)
-where index is same as the index for addByBitCode
+
    
+virtual void leftMul(Matrix &a, real scaleAB, real scaleT)
    
+
     
    this = scaleAB*(a*this) +  scaleT*this
 
    
 
    
 
    
 
    
 
 
    
-
    
-virtual void mulByBitCodeBackwardError(size_t numClasses, const IVector &codes, const Matrix &mat, Matrix &input)
    
-
     
    For j < codeLength:
-  input.row(i) += this(i, j) * mat.row(index(i, j))
-where index is same as the index for addByBitCode
+
    
+virtual void leftMul(Matrix &a)
    
+
     
    this = a*this)
 
    
 
    
 
    
 
    
 
 
    
-
    
-virtual void sumByBitCode(size_t numClasses, IVector &codes, Matrix &sum, real scaleSum)
    
-
     
    For j < codeLength
-  sum(i, 0) = scaleSum * \sum_j  bit(i, j) * this(i, j)
-where bit(i, j) = ((codes(i) + numClasses) & 2^j) ? 1 : 0
-
    
-
    
-
    
-
    
-
-
    
-
    
-virtual void subByBitCode(size_t numClasses_, IVector &codes)
    
-
     
    For j < codeLength
- this(i, j) -= bit(i, j)
-where bit(i, j) is same as that for sumByBitCode
-
    
-
    
-
    
+
    
+virtual void colMerge(Matrix &src)
    
+
    merge the element for each col. 
    
 
    
 
 
    
-
    
-virtual void rowSum(Matrix &sum)
    
+
    
+virtual void rowSum(Matrix &sum)
    
 
    add the sum of each row of this to mat 
    
 
    
 
 
    
-
    
-virtual void rowMax(Matrix &max)
    
-
    set the max of each row of this to mat 
    
-
    
-
-
    
-
    
-virtual void colMax(Matrix &max)
    
+
    
+virtual void rowMaxId(IVector &maxIds)
    
 
    
 
 
    
-
    
-virtual void rowMaxId(IVector &maxIds)
    
-
    
+
    
+virtual void rowMax(Matrix &max)
    
+
    set the max of each row of this to mat 
    
+
    
 
 
    
-
    
-virtual void rowMax(IVector &maxIds, Matrix &max)
    
+
    
+virtual void rowMax(IVector &maxIds, Matrix &max)
    
 
    Get the top k elements of each row of this matrix. 
    
 
    The column ids and values of these elements are stored in maxIds and max respectively. Note that the top k elements are not sorted. 
    
 
    
 
 
    
-
    
-virtual void rowNormalizeL1(Matrix &out)
    
-
    normalize each row so that the sum of each row is 1. 
    
-
    
-
-
    
-
    
-virtual void mul(const MatrixPtr a, const MatrixPtr b)
    
-
     
    this = a*b
-
    
-
    
-
    
-
    
-
-
    
-
    
-virtual void rightMul(Matrix &b, real scaleAB, real scaleT)
    
-
     
    this = scaleAB*(this*b) +  scaleT*this
-
    
-
    
-
    
-
    
-
-
    
-
    
-virtual void rightMul(Matrix &b)
    
-
     
    this = this* b
-
    
-
    
-
    
-
    
-
-
    
-
    
-virtual void leftMul(Matrix &a, real scaleAB, real scaleT)
    
-
     
    this = scaleAB*(a*this) +  scaleT*this
-
    
-
    
-
    
-
    
-
-
    
-
    
-virtual void leftMul(Matrix &a)
    
-
     
    this = a*this)
-
    
-
    
-
    
-
    
+
    
+virtual void colMax(Matrix &max)
    
+
    
 
 
    
-
    
-virtual void colMerge(Matrix &src)
    
-
    merge the element for each col. 
    
+
    
+virtual void rowNormalizeL1(Matrix &out)
    
+
    normalize each row so that the sum of each row is 1. 
    
 
    
 
 
    
-
    
-virtual void oneHotCrossEntropy(Matrix &output, IVector &label)
    
+
    
+virtual void oneHotCrossEntropy(Matrix &output, IVector &label)
    
 
    copy -log(output[label]) to this->data[i]. 
    
 
    
 
 
    
-
    
-virtual void oneHotCrossEntropyBp(Matrix &outputV, IVector &label)
    
+
    
+virtual void oneHotCrossEntropyBp(Matrix &outputV, IVector &label)
    
 
    calculate the error of outputV according to label. 
    
 
    
 
 
    
-
    
-virtual void oneHotCrossEntropyWithSelfNorm(Matrix &output, IVector &label, real alpha)
    
+
    
+virtual void oneHotCrossEntropyWithSelfNorm(Matrix &output, IVector &label, real alpha)
    
 
    copy -log(output[label]) to this->data[i]. 
    
 
    
 
 
    
-
    
-virtual void oneHotCrossEntropyWithSelfNormBp(Matrix &outputV, IVector &label, real alpha)
    
+
    
+virtual void oneHotCrossEntropyWithSelfNormBp(Matrix &outputV, IVector &label, real alpha)
    
 
    calculate the error of outputV according to label. 
    
 
    
 
 
    
-
    
-virtual void circularConv(Matrix &b, Matrix &c)
    
+
    
+virtual void circularConv(Matrix &b, Matrix &c)
    
 
    
 \[ a[i] = \sum_{j=-(N-1)/2}^{(N-1)/2} b_{i+j} * c_{j} \]
    
 
    
@@ -3289,249 +3406,217 @@ where bit(i, j) is same as that for sumByBitCode
 
    
 
 
    
-
    
-virtual void circularConvDerivative(Matrix &output, Matrix &prevOut1, Matrix &prevOut2, Matrix &prevGrad1, Matrix &prevGrad2)
    
-
    
-
-
    
-
    
-virtual void softmax(Matrix &output)
    
+
    
+virtual void circularConvDerivative(Matrix &output, Matrix &prevOut1, Matrix &prevOut2, Matrix &prevGrad1, Matrix &prevGrad2)
    
 
    
 
 
    
-
    
-virtual void sequenceSoftmax(Matrix &output, const IVector &index)
    
+
    
+virtual void softmax(Matrix &output)
    
 
    
 
 
    
-
    
-virtual void softmaxBackward(Matrix &outputV)
    
+
    
+virtual void sequenceSoftmax(Matrix &output, const IVector &index)
    
 
    
 
 
    
-
    
-virtual void softmaxDerivative(Matrix &output, Matrix &sftmaxSum)
    
+
    
+virtual void softmaxDerivative(Matrix &output, Matrix &sftmaxSum)
    
 
    
 
 
    
-
    
-virtual void sumOfSquares(Matrix &output, Matrix &label)
    
+
    
+virtual void sumOfSquares(Matrix &output, Matrix &label)
    
 
    calculate the sum of squares diff cost. 
    
 
    
 
 
    
-
    
-virtual void sumOfSquaresBp(Matrix &outputV, Matrix &label)
    
+
    
+virtual void sumOfSquaresBp(Matrix &outputV, Matrix &label)
    
 
    gradient of sumOfSquares. 
    
 
    
 
 
    
-
    
-virtual void tanh(Matrix &output)
    
+
    
+virtual void tanh(Matrix &output)
    
 
    
 
 
    
-
    
-virtual void tanhDerivative(Matrix &output)
    
+
    
+virtual void tanhDerivative(Matrix &output)
    
 
    
 
 
    
-
    
-virtual void softrelu(Matrix &output)
    
+
    
+virtual void softrelu(Matrix &output)
    
 
    
 
 
    
-
    
-virtual void softreluDerivative(Matrix &output)
    
+
    
+virtual void softreluDerivative(Matrix &output)
    
 
    
 
 
    
-
    
-virtual void scaledTanh(Matrix &output, real p1, real p2)
    
+
    
+virtual void scaledTanh(Matrix &output, real p1, real p2)
    
 
    
 
 
    
-
    
-virtual void cosSim(Matrix &output1, Matrix &output2, real scale = 1.0f)
    
+
    
+virtual void cosSim(Matrix &output1, Matrix &output2, real scale)
    
 
    cosine similarity, for each row i, this[i] = cos(output1[i], output2[i])
    
 
    output2 can only have one row, then for each row i, this[i] = cos(output1[i], output2[0]) 
    
 
    
 
 
    
-
    
-virtual void cosSimDerivative(Matrix &output, Matrix &prevOut1, Matrix &prevOut2, Matrix &prevGrad1, Matrix &prevGrad2, real scale = 1.0f)
    
+
    
+virtual void cosSimDerivative(Matrix &output, Matrix &prevOut1, Matrix &prevOut2, Matrix &prevGrad1, Matrix &prevGrad2, real scale)
    
 
    
 
 
    
-
    
-virtual void print(std::ostream &os) const
    
+
    
+virtual void print(std::ostream &os) const
    
 
    print out the values of elements to os 
    
 
    
 
 
    
-
    
-virtual void print(std::ostream &os, size_t height, size_t width) const
    
+
    
+virtual void print(std::ostream &os, size_t height, size_t width) const
    
 
    print a part of the matrix from the (top,left) value to the (height, width) value (not included) 
    
 
    
 
 
    
-
    
-virtual void printOneRow(std::ostream &os, size_t idx) const
    
+
    
+virtual void printOneRow(std::ostream &os, size_t idx) const
    
 
    print one row to os 
    
 
    
 
 
    
-
    
-virtual void check(std::ostream &os, Matrix &refMat, bool printDiff = true)
    
+
    
+virtual void paramReluForward(Matrix &data, Matrix &W)
    
 
    
 
 
    
-
    
-virtual real getMin()
    
+
    
+virtual void paramReluBackwardW(Matrix &oGrad, Matrix &data)
    
 
    
 
 
    
-
    
-virtual real getMax()
    
+
    
+virtual void paramReluBackwardDiff(Matrix &oGrad, Matrix &data, Matrix &W)
    
 
    
 
 
    
-
    
-virtual void randomizeUniform()
    
+
    
+virtual void check(std::ostream &os, Matrix &refMat, bool printDiff = true)
    
 
    
 
 
    
-
    
-virtual void classificationError(MatrixPtr output, IVectorPtr label)
    
+
    
+virtual real getMin()
    
+
    
+
+
    
+
    
+virtual real getMax()
    
+
    
+
+
    
+
    
+virtual void randomizeUniform()
    
+
    
+
+
    
+
    
+virtual void classificationError(MatrixPtr output, IVectorPtr label)
    
 
    calulate the error of classification 
    
 
    output[i] = 1 if row i is an error.
    
 
    output[i] = 0 if row i is correct. 
    
 
    
 
 
    
-
    
-virtual void convExpand(Matrix &feature, int feaImgHeight, int feaImgWidth, int channels, int blockH, int blockW, int strideH, int strideW, int paddingH, int paddingW, int outputH, int outputW)
    
-
    This function is used to calculate the convolution:
    
-
    It will expand a feature matrix according to the convolution filters 
    
+
    
+virtual void addByBitCode(size_t numClasses, const IVector &codes, const Matrix &vec)
    
+
     
    For j < codeLength:
+  this(i, j) += vec(index(i, j), 0)
+where index(i, j) = ((codes(i) + numClasses) >> (j + 1)) - 1
+
    
+
    
+
    
 
    
 
 
    
-
    
-virtual void convShrink(Matrix &expandColMat, int thisImgHeight, int thisImgWidth, int channels, int blockH, int blockW, int strideH, int strideW, int paddingH, int paddingW, int outputH, int outputW, real alpha = 1.0f, real beta = 0.0f)
    
-
    This function is the reverse implementation of convExpand:
    
-
    Its function is to restore a expanded-matrix into a feature matrix 
    
+
    
+virtual void addByBitCodeBackward(size_t numClasses, const IVector &codes, Matrix &vec)
    
+
     
    For j < codeLength:
+  vec(index(i, j), 0) += this(i, j)
+where index is same as the index for addByBitCode
+
    
+
    
+
    
 
    
 
 
    
-
    
-virtual void maxPoolForward(Matrix &inputMat, size_t imgSizeH, size_t imgSizeW, size_t channels, size_t sizeX, int start_, size_t stride, size_t outputH, size_t outputW)
    
-
    Pooling forward operation, pick out the largest element in the sizeX of value 
    
+
    
+virtual void mulByBitCode(size_t numClasses, const IVector &codes, const Matrix &mat, const Matrix &input)
    
+
     
    For j < codeLength:
+  this(i, j) += <mat.row(index(i, j)), input.row(i)>
+where index is same as the index for addByBitCode
+
    
+
    
+
    
 
    
 
 
    
-
    
-virtual void maxPoolBackward(Matrix &image, size_t imgSizeH, size_t imgSizeW, Matrix &outGrad, Matrix &outV, size_t sizeX, int start, size_t stride, size_t outputH, size_t outputW, real scaleTargets, real scaleOutput)
    
-
    Pooling backward operation. 
    
+
    
+virtual void mulByBitCodeBackwardWeight(size_t numClasses, const IVector &codes, Matrix &mat, const Matrix &input)
    
+
     
    For j < codeLength:
+  mat.row(index(i, j)) += this(i, j) * input.row(i)
+where index is same as the index for addByBitCode
+
    
+
    
+
    
 
    
 
 
    
-
    
-virtual void avgPoolForward(Matrix &input, size_t imgSizeH, size_t imgSizeW, size_t channels, size_t sizeX, int start, size_t stride, size_t outputH, size_t outputW)
    
-
    Pooling forward operation, caculate the average of sizeX elements. 
    
-
    
-
-
    
-
    
-virtual void avgPoolBackward(Matrix &input, size_t imgSizeH, size_t imgSizeW, size_t sizeX, int start, size_t stride, size_t outputH, size_t outputW, real scaleTargets, real scaleOutput)
    
-
    
-
-
    
-
    
-virtual void crossMapNormalFwd(Matrix &input, size_t imgSizeH, size_t imgSizeW, Matrix &denoms, size_t channels, size_t sizeX, float scale, float pow, bool blocked)
    
-
    normalize-operation. 
    
-
    
-
-
    
-
    
-virtual void crossMapNormalBwd(Matrix &localGrad, Matrix &denoms, Matrix &preOutV, Matrix &localOutV, size_t channels, size_t imgSizeH, size_t imgSizeW, size_t size, float scale, float pow, bool blocked)
    
-
    
-
-
    
-
    
-virtual void maxSequenceForward(Matrix &input, const IVector &sequence, IVector &index)
    
-
    Input: one or more sequences. Each sequence contains some instances.
    
-
    Output: output size is the number of input sequences (NOT input instances).
    
-
    output[i] is set to max_input[i]. 
    
-
    
-
-
    
-
    
-virtual void maxSequenceBackward(Matrix &outputGrad, const IVector &sequence, IVector &index)
    
-
    
-
-
    
-
    
-virtual void contextProjectionForward(MatrixPtr input, MatrixPtr weight, const IVector &sequence, int contextLength, int contextStart, size_t beginPad, bool isPadding)
    
-
    
-
-
    
-
    
-virtual void contextProjectionBackward(MatrixPtr inputGrad, MatrixPtr weightGrad, const IVector &sequence, int contextLength, int contextStart, size_t beginPad, bool isPadding)
    
-
    
-
-
    
-
    
-virtual void contextProjectionBackwardData(MatrixPtr inputGrad, const IVector &sequence, int contextLength, int contextStart)
    
-
    
-
-
    
-
    
-virtual void contextProjectionBackwardWeight(MatrixPtr weightGrad, const IVector &sequence, int contextLength, int contextStart, int totalPad, size_t beginPad)
    
-
    
-
-
    
-
    
-virtual void selectRows(Matrix &table, IVector &ids)
    
-
     
    this.row[i] += table.row[ids[i]]
-if ids[i] == -1, it will be ignored
-
    
-
    
-
    
-
    
-
-
    
-
    
-virtual void selectElements(Matrix &table, IVector &ids)
    
-
     
    this[i] = table[i, id[i]]
+
    
+virtual void mulByBitCodeBackwardError(size_t numClasses, const IVector &codes, const Matrix &mat, Matrix &input)
    
+
     
    For j < codeLength:
+  input.row(i) += this(i, j) * mat.row(index(i, j))
+where index is same as the index for addByBitCode
 
    
 
    
 
    
 
    
 
 
    
-
    
-virtual void addToRows(Matrix &table, IVector &ids)
    
-
     
    table.row[ids[i]] += this.row[i]
-if ids[i] == -1, it will be ignored
+
    
+virtual void sumByBitCode(size_t numClasses, IVector &codes, Matrix &sum, real scaleSum)
    
+
     
    For j < codeLength
+  sum(i, 0) = scaleSum * \sum_j  bit(i, j) * this(i, j)
+where bit(i, j) = ((codes(i) + numClasses) & 2^j) ? 1 : 0
 
    
 
    
 
    
 
    
 
 
    
-
    
-virtual void addElements(Matrix &table, IVector &ids)
    
-
     
    table[i, id[i]] += this[i]
+
    
+virtual void subByBitCode(size_t numClasses_, IVector &codes)
    
+
     
    For j < codeLength
+ this(i, j) -= bit(i, j)
+where bit(i, j) is same as that for sumByBitCode
 
    
 
    
 
    
 
    
 
 
    
-
    
-virtual void multiBinaryLabelCrossEntropy(Matrix &output, Matrix &label)
    
+
    
+virtual void multiBinaryLabelCrossEntropy(Matrix &output, Matrix &label)
    
 
    cross entropy for multi binary labels 
    
-
     
    this[i] = -sum(label[i][j]*log(output[i][j])
+
     
    this[i] = -sum(label[i][j]*log(output[i][j])
           + (1-label[i][j])*log(1-output[i][j]))
 
    
 
    
@@ -3539,129 +3624,44 @@ if ids[i] == -1, it will be ignored
 
    
 
 
    
-
    
-virtual void multiBinaryLabelCrossEntropyBp(Matrix &output, Matrix &label)
    
+
    
+virtual void multiBinaryLabelCrossEntropyBp(Matrix &output, Matrix &label)
    
 
    The gradient of cross entropy for multi binary labels on output. 
    
-
     
    this[i][j] = -label[i][j]/output[i][j]
-             + (1-label[i][j])/(1-output[i][j])
+
     
    this[i][j] = -label[i][j]/output[i][j]
+             + (1-label[i][j])/(1-output[i][j])
 
    
 
    
 
    
 
    
 
 
    
-
    
-virtual void classificationErrorMulti(Matrix &output, Matrix &label, real threshold)
    
+
    
+virtual void classificationErrorMulti(Matrix &output, Matrix &label, real threshold)
    
 
    Calculate the classification error for multi binary labels. 
    
-
     
    this[i] = sum((output[i][j] >= threshold && label[i][j] == 0)
-           || (output[i][j] < threshold && label[i][j] == 1))
-           / output->getWidth()
+
     
    this[i] = sum((output[i][j] >= threshold && label[i][j] == 0)
+           || (output[i][j] < threshold && label[i][j] == 1))
+           / output->getWidth()
 
    
 
    
 
    
 
    
 
-
    
-
    
-virtual void paramReluForward(Matrix &data, Matrix &W)
    
-
    
-
-
    
-
    
-virtual void paramReluBackwardW(Matrix &oGrad, Matrix &data)
    
-
    
-
-
    
-
    
-virtual void paramReluBackwardDiff(Matrix &oGrad, Matrix &data, Matrix &W)
    
-
    
-
-
    
-
    
-
    Public Members
    
-
    
-
    
-size_t elementCnt_
    
-
    
-
-
    
-
    
-MemoryHandlePtr memoryHandle_
    
-
    
-
 
    
 
    
 
    Public Static Functions
    
 
    
-
    
-MatrixPtr create(MemoryHandlePtr memHandle, size_t height, size_t width, bool trans = false)
    
-
    
-
-
    
-
    
-MatrixPtr create(size_t height, size_t width, bool trans = false, bool useGpu = false)
    
-
    
-
-
    
-
    
-MatrixPtr create(real *data, size_t height, size_t width, bool trans = false, bool useGpu = false)
    
-
    
-
-
    
-
    
-MatrixPtr create(real *data, size_t height, size_t width, size_t stride, bool trans = false, bool useGpu = false)
    
-
    
-
-
    
-
    
-MatrixPtr createSparseMatrix(size_t height, size_t width, size_t nnz, SparseValueType valueType = FLOAT_VALUE, bool trans = false, bool useGpu = false)
    
-
    
-
-
    
-
    
-MatrixPtr createSparseMatrix(size_t height, size_t width, size_t nnz, SparseValueType valueType = FLOAT_VALUE, SparseFormat foramt = SPARSE_CSR, bool trans = false, bool useGpu = false)
    
-
    
-
-
    
-
    
-MatrixPtr createSparseMatrix(real *data, int *row, int *col, size_t height, size_t width, size_t nnz, SparseValueType valueType, SparseFormat format, bool trans, bool useGpu)
    
-
    
-
-
    
-
    
-void resizeOrCreateSparseMatrix(MatrixPtr &matrix, size_t height, size_t width, size_t nnz, SparseValueType valueType = FLOAT_VALUE, SparseFormat foramt = SPARSE_CSR, bool trans = false, bool useGpu = false)
    
-
    
-
-
    
-
    
-void resizeOrCreate(MatrixPtr &a, size_t height, size_t width, bool trans = false, bool useGpu = false)
    
-
    
-
-
    
-
    
-
    Protected Functions
    
-
    
-
    
-Matrix(MemoryHandlePtr memHandle, size_t height, size_t width, bool trans, bool use_gpu)
    
-
    
-
-
    
-
    
-Matrix(real *data, size_t height, size_t width, bool trans, bool use_gpu)
    
+
    
+void mul(CpuMatrix *a, CpuMatrix *b, CpuSparseMatrix *c, real scaleAB, real scaleT)
    
 
    
 
 
    
-
    
-Matrix(real *data, size_t height, size_t width, size_t stride, bool trans, bool use_gpu)
    
-
    
-
-
    
-
    
-
    Protected Static Attributes
    
-
    
-
    
-ThreadLocal<MatrixPtr> tmpMat_
    
-
    
+
    
+template <typename MatBType, typename MatCType>
    
+
    
+void mul(CpuSparseMatrix *a, MatBType *b, MatCType *c, real scaleAB, real scaleT)
    
+
    c = a * b
    
+
    use abstract getRow() to get row from B,C. Define B,C as template instead of virtual class for performance sake. 
    
+
    
 
 
    
 
    
@@ -3669,7 +3669,7 @@ if ids[i] == -1, it will be ignored
 
    
 
    
 class SharedCpuMatrix
    
-
    Inherits from paddle::CpuMatrix
    
+
    Inherits from paddle::CpuMatrix
    
 
    
 
    Public Functions
    
 
    
@@ -3757,31 +3757,31 @@ if ids[i] == -1, it will be ignored
 
    
 
 
    
-
    
-struct sparse_float_value_t
    
+
    
+struct sparse_non_value_t
    
 
    
 
    Public Members
    
 
    
-
    
-unsigned int col
    
-
    
-
-
    
-
    
-float value
    
+
    
+unsigned int col
    
 
    
 
 
    
 
    
 
 
    
-
    
-struct sparse_non_value_t
    
+
    
+struct sparse_float_value_t
    
 
    
 
    Public Members
    
 
    
-
    
-unsigned int col
    
+
    
+unsigned int col
    
+
    
+
+
    
+
    
+float value
    
 
    
 
 
    
@@ -3790,8 +3790,8 @@ if ids[i] == -1, it will be ignored
 
    
 
 
    
-
    
-namespace paddle
    
+
    
+namespace paddle
    
 
    
 
    Typedefs
    
 
    
@@ -3887,339 +3887,194 @@ if ids[i] == -1, it will be ignored
 
    
 
    
 
    
-template <class T>
    
-
    
-class BaseVector
    
-
    Inherits from paddle::BaseMatrixT< T >
    
-
    Subclassed by paddle::VectorT< T >
    
+template <class T>
+
    
+class GpuVectorT
    
+
    Inherits from paddle::VectorT< T >
    
 
    
 
    Public Functions
    
 
    
-
    
-BaseVector(size_t size, T *data, bool useGpu)
    
+
    
+GpuVectorT(size_t size)
    
 
    
 
 
    
-
    
-~BaseVector()
    
+
    
+GpuVectorT(size_t size, GpuMemHandlePtr memHandle, size_t offset)
    
 
    
 
-
    
-
    
-
    Protected Attributes
    
-
    
-
    
-size_t &size_
    
+
    
+
    
+GpuVectorT(size_t size, T *data)
    
 
    
 
-
    
-
    
-
-
    
-
    
-template <class T>
    
-
    
-class CpuGpuVectorT
    
-
    #include <Vector.h>
    A class to do conversion between CpuVector and GpuVector automatically. 
    
-
    
-
    Public Types
    
-
    
-
    
-enum SyncedFlag
    
-
    An enum type of SyncedFlag using to mark data memory is in CPU or GPU. 
    
-
    DATA_AT_CPU: data is located in CPU.
    
-
    DATA_AT_GPU: data is located in GPU.
    
-
    SYNCED: data is located in CPU and GPU simultaneously. 
    
-
    Values:
    
-
    
-
    
-DATA_AT_CPU = 0
    
-
    
-
-
    
-
    
-DATA_AT_GPU = 1
    
-
    
-
-
    
-
    
-SYNCED = 2
    
-
    
-
-
    
-
-
    
-
    
-
    Public Functions
    
-
    
-
    
-CpuGpuVectorT(size_t size, bool useGpu)
    
-
    A constructor, create cpuVectorT_ or gpuVectorT_. 
    
-
    
-
    Parameters
    
-
      
-
    • size - 
      data size. 
      
-
      • 
-
    • useGpu - 
      use gpu or not. 
      
-
      • 
-
    
-
    
-
    
-
    
-
    
+
    
+
    
+virtual MemoryHandlePtr newMemory(size_t size)
    
+
    
 
 
    
-
    
-CpuGpuVectorT(const std::shared_ptr<VectorT<T>> &src)
    
-
    A constructor, create CpuGpuVectorT by VectorT. 
    
-
    If src is CpuVector, cpuVectorT_ is shared data with src.
    
-
    If src is GpuVector, gpuVectorT_ is shared data with src. 
    
-
    
+
    
+virtual void zeroMem()
    
+
    
 
 
    
-
    
-CpuGpuVectorT(size_t size, T *data, bool useGpu)
    
-
    A constructor. 
    
-
    If useGpu is true, data should be located in device and create gpuVectorT_ with data.
    
-
    If useGpu is false, data should be located in host and create cpuVectorT_ with data.
    
-
    
-
    Note
    
-
    Data is owned by the caller and should be valid during the life of this vector. Caller is responsible for release the memory. 
    
-
    
-
    
-
    
+
    
+virtual void reset(const T &value)
    
+
    
 
 
    
-
    
-CpuGpuVectorT(CpuGpuVectorT<T> &src, size_t offset, size_t size)
    
+
    
+virtual void fillSequence()
    
 
    
 
 
    
-
    
-virtual ~CpuGpuVectorT()
    
-
    
+
    
+virtual void copyFrom(const T *src, size_t size)
    
+
    copy size elements from src
    
+
    If this is GpuVector, src can be cpu or gpu memory
    
+
    If this is CpuVector, src is assumed to be cpu memory 
    
+
    
 
 
    
-
    
-void resize(size_t size, bool useGpu)
    
-
    resize vector. 
    
-
    If useGpu is true, resize gpuVectorT_ and set syncFlag_ to DATA_AT_GPU,
    
-
    otherwise resize cpuVectorT_ and set syncFlag_ to DATA_AT_CPU. 
    
+
    
+virtual void copyFrom(const T *src, size_t size, hl_stream_t stream)
    
+
    copy size elements from src
    
+
    If this is GpuVector, src can be cpu or gpu memory
    
+
    If this is CpuVector, src is assumed to be cpu memory, 
    
 
    
 
 
    
-
    
-std::shared_ptr<const VectorT<T>> getVector(bool useGpu) const
    
-
    return a const cpuVectorT_ or gpuVectorT_. 
    
-
    If useGpu is true, return gpuVectorT_.
    
-
    If useGpu is false, return cpuVectorT_.
    
-
    
-
    Note
    
-
    Caller should not change the data. If caller changes const attribute, should set syncFlag_. 
    
-
    
-
    
+
    
+virtual void copyFrom(const VectorT<T> &src)
    
+
    This function will crash if the size of src and dest is different. 
    
 
    
 
 
    
-
    
-std::shared_ptr<VectorT<T>> &getMutableVector(bool useGpu)
    
-
    return a const cpuVectorT_ or gpuVectorT_. 
    
-
    
-
    Note
    
-
    : This interface will change syncFlag_, so if you will not change the data, you should call getVector. 
    
-
    
-
    
+
    
+virtual void copyFrom(const VectorT<T> &src, hl_stream_t stream)
    
+
    If use_gpu, this function will push the copy-task to the specifed-stream and return immediately.
    
+
    If not use GPU, this function is same as the copyFrom(const VectorT<T>& src), which use stream HPPL_STREAM_DEFAULT. 
    
 
    
 
 
    
-
    
-const T *getData(bool useGpu) const
    
-
    return const T* data. 
    
-
    If useGpu is true, return device data.
    
-
    If useGpu is false, return host data. 
    
+
    
+virtual T getElement(size_t i) const
    
+
    Get the value for the i’th element. 
    
 
    
 
 
    
-
    
-T *getMutableData(bool useGpu)
    
+
    
+virtual void setElement(size_t i, const T &value)
    
 
    
 
 
    
-
    
-void zeroMem(bool useGpu)
    
-
    If useGpu is true, gpuVectorT_->Op().
    
-
    If useGpu is false, cpuVectorT_->Op().
    
-
    Op is zeroMem, fillSequence, ... 
    
+
    
+virtual T *getPoint(const uint64_t beginPos)
    
+
    Get the buffer point with beginPos. 
    
 
    
 
 
    
-
    
-void fillSequence(bool useGpu)
    
+
    
+virtual T getAbsSum()
    
 
    
 
 
    
-
    
-void setElement(size_t i, const T &value, bool useGpu)
    
+
    
+virtual T getSum()
    
 
    
 
 
    
-
    
-T getElement(size_t i) const
    
-
    return i-th element. 
    
-
    
-
-
    
-
    
-size_t getSize() const
    
-
    return vector size. 
    
-
    
-
-
    
-
    
-void copyToCpu(const T *data, size_t size)
    
-
    copy data to cpuVectorT_. 
    
-
    
+
    
+virtual T getMax()
    
+
    
 
 
    
-
    
-void copyToCpu(const T *data, size_t size, hl_stream_t stream)
    
-
    copy data to cpuVectorT_ using specifed-stream. 
    
-
    
+
    
+virtual T getAbsMax()
    
+
    
 
 
    
-
    
-void copyToGpu(const T *data, size_t size)
    
-
    copy data to gpuVectorT_. 
    
-
    
+
    
+virtual T getMin()
    
+
    
 
 
    
-
    
-void copyToGpu(const T *data, size_t size, hl_stream_t stream)
    
-
    copy data to gpuVectorT_ using specifed-stream. 
    
+
    
+virtual void isEqualTo(const VectorT<T> &b, const T &value)
    
+
    element-wise calc: this = (b == value) 
    
 
    
 
 
    
-
    
-void copyFrom(const VectorT<T> &src, hl_stream_t stream)
    
-
    copy from src using specifed-stream. 
    
-
    If src is CpuVectorT, copy to cpuVectorT_.
    
-
    If src is GpuVectorT, copy to gpuVectorT_. 
    
+
    
+virtual void selectFrom(const VectorT<T> &src, const VectorT<int> &ids)
    
+
    select elements indexed by ids from vector src 
    
 
    
 
 
    
-
    
-void copyFrom(const T *data, size_t size, bool useGpu)
    
-
    copy data. 
    
-
    If useGpu is false, copy host data to cpuVectorT_.
    
-
    If useGpu is true, copy device data to gpuVectorT_.
    
+
    
+virtual void histogram(std::ostream &os, int type)
    
+
    print histogram of vector values 
    
 
    
 
    Note
    
-
    data address should consistent with useGpu. 
    
+
    only exponent histogram supported currently 
    
 
    
 
    
 
    
 
 
    
-
    
-void copyFrom(const T *data, size_t size, hl_stream_t stream, bool useGpu)
    
-
    
-
-
    
-
    
-void copyFrom(CpuGpuVectorT<T> &src, size_t offset, size_t size, bool useGpu, hl_stream_t stream)
    
-
    copy from (src + offset) using specifed-stream. 
    
+
    
+virtual void rand()
    
+
    generate uniform random value for each element 
    
 
    
 
 
    
-
    
-void copyFrom(CpuGpuVectorT<T> &src, hl_stream_t stream)
    
-
    copy from src using specifed-stream. 
    
+
    
+virtual void rand(size_t classes)
    
+
    generate uniform random value for each element, data range is from 0 to (classes - 1). 
    
 
    
 
 
    
-
    
-SyncedFlag *getSync() const
    
-
    return sync_. 
    
+
    
+virtual void randnorm(real mean, real standardDeviation)
    
+
    generate univariate Gaussian distributed random numbers with given mean and standardDeviation. 
    
 
    
 
 
    
-
    
-void setSync(SyncedFlag *sync)
    
-
    set sync_. 
    
+
    
+virtual void uniform(real left, real right)
    
+
    generate uniform distributed random numbers with given range. 
    
 
    
 
 
    
-
    
-void setSync(SyncedFlag syncFlag)
    
-
    
-
-
    
-
    
-void setSync(bool useGpu)
    
-
    
+
    
+virtual T get(size_t pos)
    
+
    Debug use only. Very inefficient for GPU vector. get the value at pos. 
    
+
    
 
-
    
-
    
-
    Public Static Functions
    
 
    
-
    
-std::shared_ptr<CpuGpuVectorT<T>> create(size_t size, bool useGpu)
    
-
    
+
    
+virtual void print(std::ostream &os, size_t num) const
    
+
    print the first “num” elements of the Vector 
    
+
    
 
 
    
-
    
-void resizeOrCreate(std::shared_ptr<CpuGpuVectorT<T>> &vec, size_t size, bool useGpu)
    
-
    resize or create CpuGpuVectorT. 
    
+
    
+virtual void printOneElement(std::ostream &os, size_t idx) const
    
+
    print the “idx” element of the Vector 
    
 
    
 
 
    
 
    
 
    Protected Functions
    
 
    
-
    
-void resizeOrCreate(size_t size, bool useGpu)
    
+
    
+virtual void copyTo(CpuVectorT<T> *dest) const
    
 
    
 
 
    
-
    
-void copyToCpu()
    
-
    copy between cpuVectorT_ and gpuVectorT_. 
    
-
    If syncFlag_ is DATA_AT_CPU and SYNCED, do nothing.
    
-
    If syncFlag_ is DATA_AT_GPU, copy gpuVectorT_ to cpuVectorT_ and set syncFlag_ to SYNCED. 
    
-
    
-
-
    
-
    
-void copyToGpu()
    
-
    copy between cpuVectorT_ and gpuVectorT_. 
    
-
    If syncFlag_ is DATA_AT_GPU and SYNCED, do nothing.
    
-
    If syncFlag_ is DATA_AT_CPU, copy cpuVectorT_ to gpuVectorT_ and set syncFlag_ to SYNCED. 
    
-
    
-
-
    
-
    
-
    Protected Attributes
    
-
    
-
    
-std::shared_ptr<VectorT<T>> cpuVectorT_
    
-
    host pointer. 
    
-
    
-
-
    
-
    
-std::shared_ptr<VectorT<T>> gpuVectorT_
    
-
    device pointer. 
    
-
    
-
-
    
-
    
-SyncedFlag syncFlag_
    
-
    specify current data address. 
    
-
    
-
-
    
-
    
-SyncedFlag *sync_
    
+
    
+virtual void copyTo(GpuVectorT<T> *dest) const
    
 
    
 
 
    
@@ -4230,8 +4085,8 @@ if ids[i] == -1, it will be ignored
 template <class T>
 
    
 class CpuVectorT
    
-
    Inherits from paddle::VectorT< T >
    
-
    Subclassed by paddle::ParallelCpuVectorT< T >
    
+
    Inherits from paddle::VectorT< T >
    
+
    Subclassed by paddle::ParallelCpuVectorT< T >
    
 
    
 
    Public Functions
    
 
    
@@ -4302,12 +4157,12 @@ if ids[i] == -1, it will be ignored
 
    
 virtual void copyFrom(const VectorT<T> &src, hl_stream_t stream)
    
 
    If use_gpu, this function will push the copy-task to the specifed-stream and return immediately.
    
-
    If not use GPU, this function is same as the copyFrom(const VectorT<T>& src), which use stream HPPL_STREAM_DEFAULT. 
    
+
    If not use GPU, this function is same as the copyFrom(const VectorT<T>& src), which use stream HPPL_STREAM_DEFAULT. 
    
 
    
 
 
    
 
    
-virtual void copyTo(CpuVectorT<T> *dest) const
    
+virtual void copyTo(CpuVectorT<T> *dest) const
 
    
 
 
    
@@ -4413,13 +4268,13 @@ if ids[i] == -1, it will be ignored
 
    
 
    
 virtual void print(std::ostream &os, size_t num) const
    
-
    print the first “num” elements of the Vector 
    
+
    print the first “num” elements of the Vector 
    
 
    
 
 
    
 
    
 virtual void printOneElement(std::ostream &os, size_t idx) const
    
-
    print the “idx” element of the Vector 
    
+
    print the “idx” element of the Vector 
    
 
    
 
 
    
@@ -4427,133 +4282,224 @@ if ids[i] == -1, it will be ignored
 
 
    
 
    
-template <class T>
    
-
    
-class GpuVectorT
    
-
    Inherits from paddle::VectorT< T >
    
+template <class T>
+
    
+class BaseVector
    
+
    Inherits from paddle::BaseMatrixT< T >
    
+
    Subclassed by paddle::VectorT< T >
    
 
    
 
    Public Functions
    
 
    
-
    
-GpuVectorT(size_t size)
    
+
    
+BaseVector(size_t size, T *data, bool useGpu)
    
 
    
 
 
    
-
    
-GpuVectorT(size_t size, GpuMemHandlePtr memHandle, size_t offset)
    
+
    
+~BaseVector()
    
+
    
+
+
    
+
    
+
    Protected Attributes
    
+
    
+
    
+size_t &size_
    
 
    
 
+
    
+
    
+
+
    
+
    
+template <class T>
    
+
    
+class VectorT
    
+
    #include <Vector.h>
    Copy or assignemnt constructor will share the data as opposed to making a copy of the original data. To make a copy of the orinal data, use copyFrom() instead. 
    
+
    Inherits from paddle::BaseVector< T >
    
+
    Subclassed by paddle::CpuVectorT< T >, paddle::GpuVectorT< T >
    
+
    
+
    Public Types
    
+
    
+
    
+enum HistogramType
    
+
    Values:
    
+
    
+
    
+HISTOGRAM_EXPONENT = 0
    
+
    
+
+
    
+
+
    
+
    
+
    Public Functions
    
 
    
-
    
-GpuVectorT(size_t size, T *data)
    
+
    
+virtual ~VectorT()
    
 
    
 
 
    
-
    
-virtual MemoryHandlePtr newMemory(size_t size)
    
+
    
+size_t getSize() const
    
 
    
 
 
    
-
    
-virtual void zeroMem()
    
+
    
+const T *getData() const
    
 
    
 
 
    
-
    
-virtual void reset(const T &value)
    
+
    
+T *getData()
    
 
    
 
 
    
-
    
-virtual void fillSequence()
    
+
    
+virtual void zeroMem() = 0
    
 
    
 
 
    
-
    
-virtual void copyFrom(const T *src, size_t size)
    
+
    
+virtual void reset(const T &value) = 0
    
+
    
+
+
    
+
    
+virtual void fillSequence() = 0
    
+
    
+
+
    
+
    
+MemoryHandlePtr getMemoryHandle() const
    
+
    
+
+
    
+
    
+void resize(size_t newSize)
    
+
    resizing to a big vector will not preserve old values. 
    
+
    
+
+
    
+
    
+virtual MemoryHandlePtr newMemory(size_t size) = 0
    
+
    
+
+
    
+
    
+void subVecFrom(const VectorT<T> &src, size_t start, size_t size)
    
+
    form sub vector from src, shallow copy 
    
+
    
+
+
    
+
    
+std::shared_ptr<VectorT<T>> subVec(size_t start, size_t size)
    
+
    
+
+
    
+
    
+void subVecFrom(const T *src, size_t start, size_t size)
    
+
    form sub vector from src, shallow copy 
    
+
    
+
+
    
+
    
+void subVecFrom(const VectorT<T> &src, std::pair<size_t, size_t> interval)
    
+
    form sub vector from src, shallow copy in interval [interval.first, interval.second) 
    
+
    
+
+
    
+
    
+virtual void copyFrom(const VectorT<T> &src) = 0
    
+
    This function will crash if the size of src and dest is different. 
    
+
    
+
+
    
+
    
+virtual void copyFrom(const VectorT<T> &src, hl_stream_t stream) = 0
    
+
    If use_gpu, this function will push the copy-task to the specifed-stream and return immediately.
    
+
    If not use GPU, this function is same as the copyFrom(const VectorT<T>& src), which use stream HPPL_STREAM_DEFAULT. 
    
+
    
+
+
    
+
    
+virtual void copyFrom(const T *src, size_t size) = 0
    
 
    copy size elements from src
    
 
    If this is GpuVector, src can be cpu or gpu memory
    
 
    If this is CpuVector, src is assumed to be cpu memory 
    
 
    
 
 
    
-
    
-virtual void copyFrom(const T *src, size_t size, hl_stream_t stream)
    
+
    
+virtual void copyFrom(const T *src, size_t size, hl_stream_t stream) = 0
    
 
    copy size elements from src
    
 
    If this is GpuVector, src can be cpu or gpu memory
    
 
    If this is CpuVector, src is assumed to be cpu memory, 
    
 
    
 
 
    
-
    
-virtual void copyFrom(const VectorT<T> &src)
    
-
    This function will crash if the size of src and dest is different. 
    
+
    
+virtual void exec(SyncThreadPool::JobFunc func)
    
+
    exec a func in single/multi thread 
    
 
    
 
 
    
-
    
-virtual void copyFrom(const VectorT<T> &src, hl_stream_t stream)
    
-
    If use_gpu, this function will push the copy-task to the specifed-stream and return immediately.
    
-
    If not use GPU, this function is same as the copyFrom(const VectorT<T>& src), which use stream HPPL_STREAM_DEFAULT. 
    
+
    
+virtual T *getPoint(const uint64_t beginPos) = 0
    
+
    Get the buffer point with beginPos. 
    
 
    
 
 
    
-
    
-virtual T getElement(size_t i) const
    
+
    
+virtual T getElement(size_t i) const = 0
    
 
    Get the value for the i’th element. 
    
 
    
 
 
    
-
    
-virtual void setElement(size_t i, const T &value)
    
+
    
+virtual void setElement(size_t i, const T &value) = 0
    
 
    
 
 
    
-
    
-virtual T *getPoint(const uint64_t beginPos)
    
-
    Get the buffer point with beginPos. 
    
-
    
-
-
    
-
    
-virtual T getAbsSum()
    
+
    
+virtual T getAbsSum() = 0
    
 
    
 
 
    
-
    
-virtual T getSum()
    
+
    
+virtual T getSum() = 0
    
 
    
 
 
    
-
    
-virtual T getMax()
    
+
    
+virtual T getMax() = 0
    
 
    
 
 
    
-
    
-virtual T getAbsMax()
    
+
    
+virtual T getAbsMax() = 0
    
 
    
 
 
    
-
    
-virtual T getMin()
    
+
    
+virtual T getMin() = 0
    
 
    
 
 
    
-
    
-virtual void isEqualTo(const VectorT<T> &b, const T &value)
    
+
    
+virtual void isEqualTo(const VectorT<T> &b, const T &value) = 0
    
 
    element-wise calc: this = (b == value) 
    
 
    
 
 
    
-
    
-virtual void selectFrom(const VectorT<T> &src, const VectorT<int> &ids)
    
+
    
+virtual void selectFrom(const VectorT<T> &src, const VectorT<int> &ids) = 0
    
 
    select elements indexed by ids from vector src 
    
 
    
 
 
    
-
    
-virtual void histogram(std::ostream &os, int type)
    
+
    
+virtual void histogram(std::ostream &os, int type = HISTOGRAM_EXPONENT) = 0
    
 
    print histogram of vector values 
    
 
    
 
    Note
    
@@ -4563,58 +4509,117 @@ if ids[i] == -1, it will be ignored
 
    
 
 
    
-
    
-virtual void rand()
    
+
    
+virtual void rand() = 0
    
 
    generate uniform random value for each element 
    
 
    
 
 
    
-
    
-virtual void rand(size_t classes)
    
+
    
+virtual void rand(size_t classes) = 0
    
 
    generate uniform random value for each element, data range is from 0 to (classes - 1). 
    
 
    
 
 
    
-
    
-virtual void randnorm(real mean, real standardDeviation)
    
+
    
+virtual T get(size_t pos) = 0
    
+
    Debug use only. Very inefficient for GPU vector. get the value at pos. 
    
+
    
+
+
    
+
    
+virtual void randnorm(real mean, real standardDeviation) = 0
    
 
    generate univariate Gaussian distributed random numbers with given mean and standardDeviation. 
    
 
    
 
 
    
-
    
-virtual void uniform(real left, real right)
    
+
    
+virtual void uniform(real left, real right) = 0
    
 
    generate uniform distributed random numbers with given range. 
    
 
    
 
 
    
-
    
-virtual T get(size_t pos)
    
-
    Debug use only. Very inefficient for GPU vector. get the value at pos. 
    
+
    
+virtual void print(std::ostream &os, size_t num) const = 0
    
+
    print the first “num” elements of the Vector 
    
 
    
 
 
    
-
    
-virtual void print(std::ostream &os, size_t num) const
    
-
    print the first “num” elements of the Vector 
    
+
    
+virtual void printOneElement(std::ostream &os, size_t idx) const = 0
    
+
    print the “idx” element of the Vector 
    
 
    
 
+
    
+
    
+
    Public Static Functions
    
+
    
+
    
+std::shared_ptr<VectorT<T>> create(size_t size, bool useGpu)
    
+
    
+
 
    
-
    
-virtual void printOneElement(std::ostream &os, size_t idx) const
    
-
    print the “idx” element of the Vector 
    
-
    
+
    
+std::shared_ptr<VectorT<T>> create(T *data, size_t size, bool useGpu)
    
+
    
+
+
    
+
    
+std::shared_ptr<VectorT<T>> create(size_t size, MemoryHandlePtr memoryHandle, size_t offset = 0)
    
+
    
+
+
    
+
    
+std::shared_ptr<VectorT<T>> createParallelVector(size_t size, bool useGpu, SyncThreadPool *pool = nullptr)
    
+
    
+
+
    
+
    
+static void resizeOrCreate(std::shared_ptr<VectorT<T>> &vec, size_t size, bool useGpu)
    
+
    
 
 
    
 
    
 
    Protected Functions
    
 
    
-
    
-virtual void copyTo(CpuVectorT<T> *dest) const
    
+
    
+VectorT(size_t size, MemoryHandlePtr memoryHandle, size_t offset, bool useGpu)
    
 
    
 
 
    
-
    
-virtual void copyTo(GpuVectorT<T> *dest) const
    
+
    
+VectorT(size_t size, T *data, bool useGpu)
    
+
    
+
+
    
+
    
+virtual void copyTo(CpuVectorT<T> *dest) const = 0
    
+
    
+
+
    
+
    
+virtual void copyTo(GpuVectorT<T> *dest) const = 0
    
+
    
+
+
    
+
    
+
    Protected Attributes
    
+
    
+
    
+MemoryHandlePtr memoryHandle_
    
+
    
+
+
    
+
    
+
    Friends
    
+
    
+
    
+friend paddle::GpuVectorT< T >
    
+
    
+
+
    
+
    
+friend paddle::CpuVectorT< T >
    
 
    
 
 
    
@@ -4625,7 +4630,7 @@ if ids[i] == -1, it will be ignored
 template <class T>
 
    
 class ParallelCpuVectorT
    
-
    Inherits from paddle::CpuVectorT< T >
    
+
    Inherits from paddle::CpuVectorT< T >
    
 
    
 
    Public Functions
    
 
    
@@ -4661,7 +4666,7 @@ if ids[i] == -1, it will be ignored
 
    Private Types
    
 
    
 
    
-typedef std::function<void(CpuVectorT<T>& vec)> ExecFunc
    
+typedef std::function<void(CpuVectorT<T> &vec)> ExecFunc
 
    
 
 
    
@@ -4685,314 +4690,309 @@ if ids[i] == -1, it will be ignored
 
 
    
 
    
-template <class T>
    
-
    
-class VectorT
    
-
    #include <Vector.h>
    Copy or assignemnt constructor will share the data as opposed to making a copy of the original data. To make a copy of the orinal data, use copyFrom() instead. 
    
-
    Inherits from paddle::BaseVector< T >
    
-
    Subclassed by paddle::CpuVectorT< T >, paddle::GpuVectorT< T >
    
+template <class T>
+
    
+class CpuGpuVectorT
    
+
    #include <Vector.h>
    A class to do conversion between CpuVector and GpuVector automatically. 
    
 
    
 
    Public Types
    
 
    
-
    
-enum HistogramType
    
-
    Values:
    
+
    
+enum SyncedFlag
    
+
    An enum type of SyncedFlag using to mark data memory is in CPU or GPU. 
    
+
    DATA_AT_CPU: data is located in CPU.
    
+
    DATA_AT_GPU: data is located in GPU.
    
+
    SYNCED: data is located in CPU and GPU simultaneously. 
    
+
    Values:
    
 
    
-
    
-HISTOGRAM_EXPONENT = 0
    
-
    
-
-
    
-
-
    
-
    
-
    Public Functions
    
-
    
-
    
-virtual ~VectorT()
    
-
    
-
-
    
-
    
-size_t getSize() const
    
-
    
-
-
    
-
    
-const T *getData() const
    
-
    
-
-
    
-
    
-T *getData()
    
-
    
-
-
    
-
    
-virtual void zeroMem() = 0
    
-
    
-
-
    
-
    
-virtual void reset(const T &value) = 0
    
-
    
-
-
    
-
    
-virtual void fillSequence() = 0
    
-
    
-
-
    
-
    
-MemoryHandlePtr getMemoryHandle() const
    
+
    
+DATA_AT_CPU = 0
    
 
    
 
-
    
-
    
-void resize(size_t newSize)
    
-
    resizing to a big vector will not preserve old values. 
    
-
    
-
-
    
-
    
-virtual MemoryHandlePtr newMemory(size_t size) = 0
    
+
    
+
    
+DATA_AT_GPU = 1
    
 
    
 
-
    
-
    
-void subVecFrom(const VectorT<T> &src, size_t start, size_t size)
    
-
    form sub vector from src, shallow copy 
    
-
    
-
-
    
-
    
-std::shared_ptr<VectorT<T>> subVec(size_t start, size_t size)
    
+
    
+
    
+SYNCED = 2
    
 
    
 
-
    
-
    
-void subVecFrom(const T *src, size_t start, size_t size)
    
-
    form sub vector from src, shallow copy 
    
 
    
 
+
    
+
    
+
    Public Functions
    
 
    
-
    
-void subVecFrom(const VectorT<T> &src, std::pair<size_t, size_t> interval)
    
-
    form sub vector from src, shallow copy in interval [interval.first, interval.second) 
    
+
    
+CpuGpuVectorT(size_t size, bool useGpu)
    
+
    A constructor, create cpuVectorT_ or gpuVectorT_. 
    
+
    
+
    Parameters
    
+
      
+
    • size - 
      data size. 
      
+
      • 
+
    • useGpu - 
      use gpu or not. 
      
+
      • 
+
    
+
    
+
    
+
    
 
    
 
 
    
-
    
-virtual void copyFrom(const VectorT<T> &src) = 0
    
-
    This function will crash if the size of src and dest is different. 
    
+
    
+CpuGpuVectorT(const std::shared_ptr<VectorT<T>> &src)
    
+
    A constructor, create CpuGpuVectorT by VectorT. 
    
+
    If src is CpuVector, cpuVectorT_ is shared data with src.
    
+
    If src is GpuVector, gpuVectorT_ is shared data with src. 
    
 
    
 
 
    
-
    
-virtual void copyFrom(const VectorT<T> &src, hl_stream_t stream) = 0
    
-
    If use_gpu, this function will push the copy-task to the specifed-stream and return immediately.
    
-
    If not use GPU, this function is same as the copyFrom(const VectorT<T>& src), which use stream HPPL_STREAM_DEFAULT. 
    
+
    
+CpuGpuVectorT(size_t size, T *data, bool useGpu)
    
+
    A constructor. 
    
+
    If useGpu is true, data should be located in device and create gpuVectorT_ with data.
    
+
    If useGpu is false, data should be located in host and create cpuVectorT_ with data.
    
+
    
+
    Note
    
+
    Data is owned by the caller and should be valid during the life of this vector. Caller is responsible for release the memory. 
    
+
    
+
    
 
    
 
 
    
-
    
-virtual void copyFrom(const T *src, size_t size) = 0
    
-
    copy size elements from src
    
-
    If this is GpuVector, src can be cpu or gpu memory
    
-
    If this is CpuVector, src is assumed to be cpu memory 
    
-
    
+
    
+CpuGpuVectorT(CpuGpuVectorT<T> &src, size_t offset, size_t size)
    
+
    
 
 
    
-
    
-virtual void copyFrom(const T *src, size_t size, hl_stream_t stream) = 0
    
-
    copy size elements from src
    
-
    If this is GpuVector, src can be cpu or gpu memory
    
-
    If this is CpuVector, src is assumed to be cpu memory, 
    
-
    
+
    
+virtual ~CpuGpuVectorT()
    
+
    
 
 
    
-
    
-virtual void exec(SyncThreadPool::JobFunc func)
    
-
    exec a func in single/multi thread 
    
+
    
+void resize(size_t size, bool useGpu)
    
+
    resize vector. 
    
+
    If useGpu is true, resize gpuVectorT_ and set syncFlag_ to DATA_AT_GPU,
    
+
    otherwise resize cpuVectorT_ and set syncFlag_ to DATA_AT_CPU. 
    
 
    
 
 
    
-
    
-virtual T *getPoint(const uint64_t beginPos) = 0
    
-
    Get the buffer point with beginPos. 
    
+
    
+std::shared_ptr<const VectorT<T>> getVector(bool useGpu) const
    
+
    return a const cpuVectorT_ or gpuVectorT_. 
    
+
    If useGpu is true, return gpuVectorT_.
    
+
    If useGpu is false, return cpuVectorT_.
    
+
    
+
    Note
    
+
    Caller should not change the data. If caller changes const attribute, should set syncFlag_. 
    
+
    
+
    
 
    
 
 
    
-
    
-virtual T getElement(size_t i) const = 0
    
-
    Get the value for the i’th element. 
    
+
    
+std::shared_ptr<VectorT<T>> &getMutableVector(bool useGpu)
    
+
    return a const cpuVectorT_ or gpuVectorT_. 
    
+
    
+
    Note
    
+
    : This interface will change syncFlag_, so if you will not change the data, you should call getVector. 
    
+
    
+
    
 
    
 
 
    
-
    
-virtual void setElement(size_t i, const T &value) = 0
    
-
    
+
    
+const T *getData(bool useGpu) const
    
+
    return const T* data. 
    
+
    If useGpu is true, return device data.
    
+
    If useGpu is false, return host data. 
    
+
    
 
 
    
-
    
-virtual T getAbsSum() = 0
    
+
    
+T *getMutableData(bool useGpu)
    
 
    
 
 
    
-
    
-virtual T getSum() = 0
    
-
    
+
    
+void zeroMem(bool useGpu)
    
+
    If useGpu is true, gpuVectorT_->Op().
    
+
    If useGpu is false, cpuVectorT_->Op().
    
+
    Op is zeroMem, fillSequence, ... 
    
+
    
 
 
    
-
    
-virtual T getMax() = 0
    
+
    
+void fillSequence(bool useGpu)
    
 
    
 
 
    
-
    
-virtual T getAbsMax() = 0
    
+
    
+void setElement(size_t i, const T &value, bool useGpu)
    
 
    
 
 
    
-
    
-virtual T getMin() = 0
    
-
    
+
    
+T getElement(size_t i) const
    
+
    return i-th element. 
    
+
    
 
 
    
-
    
-virtual void isEqualTo(const VectorT<T> &b, const T &value) = 0
    
-
    element-wise calc: this = (b == value) 
    
+
    
+size_t getSize() const
    
+
    return vector size. 
    
 
    
 
 
    
-
    
-virtual void selectFrom(const VectorT<T> &src, const VectorT<int> &ids) = 0
    
-
    select elements indexed by ids from vector src 
    
+
    
+void copyToCpu(const T *data, size_t size)
    
+
    copy data to cpuVectorT_. 
    
 
    
 
 
    
-
    
-virtual void histogram(std::ostream &os, int type = HISTOGRAM_EXPONENT) = 0
    
-
    print histogram of vector values 
    
-
    
-
    Note
    
-
    only exponent histogram supported currently 
    
-
    
-
    
+
    
+void copyToCpu(const T *data, size_t size, hl_stream_t stream)
    
+
    copy data to cpuVectorT_ using specifed-stream. 
    
 
    
 
 
    
-
    
-virtual void rand() = 0
    
-
    generate uniform random value for each element 
    
+
    
+void copyToGpu(const T *data, size_t size)
    
+
    copy data to gpuVectorT_. 
    
 
    
 
 
    
-
    
-virtual void rand(size_t classes) = 0
    
-
    generate uniform random value for each element, data range is from 0 to (classes - 1). 
    
+
    
+void copyToGpu(const T *data, size_t size, hl_stream_t stream)
    
+
    copy data to gpuVectorT_ using specifed-stream. 
    
 
    
 
 
    
-
    
-virtual T get(size_t pos) = 0
    
-
    Debug use only. Very inefficient for GPU vector. get the value at pos. 
    
+
    
+void copyFrom(const VectorT<T> &src, hl_stream_t stream)
    
+
    copy from src using specifed-stream. 
    
+
    If src is CpuVectorT, copy to cpuVectorT_.
    
+
    If src is GpuVectorT, copy to gpuVectorT_. 
    
 
    
 
 
    
-
    
-virtual void randnorm(real mean, real standardDeviation) = 0
    
-
    generate univariate Gaussian distributed random numbers with given mean and standardDeviation. 
    
+
    
+void copyFrom(const T *data, size_t size, bool useGpu)
    
+
    copy data. 
    
+
    If useGpu is false, copy host data to cpuVectorT_.
    
+
    If useGpu is true, copy device data to gpuVectorT_.
    
+
    
+
    Note
    
+
    data address should consistent with useGpu. 
    
+
    
+
    
 
    
 
 
    
-
    
-virtual void uniform(real left, real right) = 0
    
-
    generate uniform distributed random numbers with given range. 
    
+
    
+void copyFrom(const T *data, size_t size, hl_stream_t stream, bool useGpu)
    
+
    
+
+
    
+
    
+void copyFrom(CpuGpuVectorT<T> &src, size_t offset, size_t size, bool useGpu, hl_stream_t stream)
    
+
    copy from (src + offset) using specifed-stream. 
    
 
    
 
 
    
-
    
-virtual void print(std::ostream &os, size_t num) const = 0
    
-
    print the first “num” elements of the Vector 
    
+
    
+void copyFrom(CpuGpuVectorT<T> &src, hl_stream_t stream)
    
+
    copy from src using specifed-stream. 
    
 
    
 
 
    
-
    
-virtual void printOneElement(std::ostream &os, size_t idx) const = 0
    
-
    print the “idx” element of the Vector 
    
+
    
+SyncedFlag *getSync() const
    
+
    return sync_. 
    
 
    
 
-
    
-
    
-
    Public Static Functions
    
 
    
-
    
-std::shared_ptr<VectorT<T>> create(size_t size, bool useGpu)
    
-
    
+
    
+void setSync(SyncedFlag *sync)
    
+
    set sync_. 
    
+
    
 
 
    
-
    
-std::shared_ptr<VectorT<T>> create(T *data, size_t size, bool useGpu)
    
+
    
+void setSync(SyncedFlag syncFlag)
    
 
    
 
 
    
-
    
-std::shared_ptr<VectorT<T>> create(size_t size, MemoryHandlePtr memoryHandle, size_t offset = 0)
    
+
    
+void setSync(bool useGpu)
    
 
    
 
+
    
+
    
+
    Public Static Functions
    
 
    
-
    
-std::shared_ptr<VectorT<T>> createParallelVector(size_t size, bool useGpu, SyncThreadPool *pool = nullptr)
    
+
    
+std::shared_ptr<CpuGpuVectorT<T>> create(size_t size, bool useGpu)
    
 
    
 
 
    
-
    
-static void resizeOrCreate(std::shared_ptr<VectorT<T>> &vec, size_t size, bool useGpu)
    
-
    
+
    
+void resizeOrCreate(std::shared_ptr<CpuGpuVectorT<T>> &vec, size_t size, bool useGpu)
    
+
    resize or create CpuGpuVectorT. 
    
+
    
 
 
    
 
    
 
    Protected Functions
    
 
    
-
    
-VectorT(size_t size, MemoryHandlePtr memoryHandle, size_t offset, bool useGpu)
    
-
    
-
-
    
-
    
-VectorT(size_t size, T *data, bool useGpu)
    
+
    
+void resizeOrCreate(size_t size, bool useGpu)
    
 
    
 
 
    
-
    
-virtual void copyTo(CpuVectorT<T> *dest) const = 0
    
-
    
+
    
+void copyToCpu()
    
+
    copy between cpuVectorT_ and gpuVectorT_. 
    
+
    If syncFlag_ is DATA_AT_CPU and SYNCED, do nothing.
    
+
    If syncFlag_ is DATA_AT_GPU, copy gpuVectorT_ to cpuVectorT_ and set syncFlag_ to SYNCED. 
    
+
    
 
 
    
-
    
-virtual void copyTo(GpuVectorT<T> *dest) const = 0
    
-
    
+
    
+void copyToGpu()
    
+
    copy between cpuVectorT_ and gpuVectorT_. 
    
+
    If syncFlag_ is DATA_AT_GPU and SYNCED, do nothing.
    
+
    If syncFlag_ is DATA_AT_CPU, copy cpuVectorT_ to gpuVectorT_ and set syncFlag_ to SYNCED. 
    
+
    
 
 
    
 
    
 
    Protected Attributes
    
 
    
-
    
-MemoryHandlePtr memoryHandle_
    
-
    
+
    
+std::shared_ptr<VectorT<T>> cpuVectorT_
    
+
    host pointer. 
    
+
    
 
-
    
-
    
-
    Friends
    
-
    
-
    
-friend paddle::GpuVectorT< T >
    
-
    
+
    
+
    
+std::shared_ptr<VectorT<T>> gpuVectorT_
    
+
    device pointer. 
    
+
    
 
-
    
-
    
-friend paddle::CpuVectorT< T >
    
+
    
+
    
+SyncedFlag syncFlag_
    
+
    specify current data address. 
    
+
    
+
+
    
+
    
+SyncedFlag *sync_
    
 
    
 
 
    
@@ -5001,8 +5001,8 @@ if ids[i] == -1, it will be ignored
 
    
 
 
    
-
    
-namespace paddle
    
+
    
+namespace paddle
    
 
    
 
    Typedefs
    
 
    
@@ -5014,7 +5014,7 @@ if ids[i] == -1, it will be ignored
 
    
 
    
 class GpuSparseMatrix
    
-
    Inherits from paddle::Matrix
    
+
    Inherits from paddle::Matrix
    
 
    
 
    Public Functions
    
 
    
@@ -5057,7 +5057,7 @@ if ids[i] == -1, it will be ignored
 virtual void resize(size_t newHeight, size_t newWidth)
 
    
 
    Note
    
-
    Original data may not be preserved after resize(). 
    
+
    Original data may not be preserved after resize(). 
    
 
    
 
    
 
    
@@ -5246,12 +5246,12 @@ if ids[i] == -1, it will be ignored
 
    
 virtual real *getData()
    
 
    return value_ of sparse matrix 
    
-
    Some times CpuSparseMatrix maybe Matrix, if getValue, must dynamic_cast to CpuSparseMatrix, getData is convenient to get value 
    
+
    Some times CpuSparseMatrix maybe Matrix, if getValue, must dynamic_cast to CpuSparseMatrix, getData is convenient to get value 
    
 
    
 
 
    
 
    
-virtual const real *getData() const
    
+virtual const real *getData() const
 
    
 
 
    
@@ -5275,7 +5275,7 @@ if ids[i] == -1, it will be ignored
 
    
 
    
 virtual void mul(const MatrixPtr a, const MatrixPtr b, real scaleAB, real scaleT)
    
-
     
    this = scaleAB*(a*b) + scaleT*this
+
     
    this = scaleAB*(a*b) + scaleT*this
 
    
 
    
 
    
@@ -5288,7 +5288,7 @@ if ids[i] == -1, it will be ignored
 
 
    
 
    
-void copyFrom(GpuSparseMatrix &src, hl_stream_t stream)
    
+void copyFrom(GpuSparseMatrix &src, hl_stream_t stream)
 
    
 
 
    
@@ -5417,141 +5417,14 @@ if ids[i] == -1, it will be ignored
 
 
    
 
    
-
    
-namespace paddle
    
+
    
+namespace paddle
    
 
    
-
    
-class CacheRowCpuMatrix
    
-
    Inherits from paddle::SparseAutoGrowRowCpuMatrix
    
-
    
-
    Public Functions
    
-
    
-
    
-CacheRowCpuMatrix(size_t height, size_t width, IndexDictPtr indexDictHandle = nullptr, bool trans = false)
    
-
    
-
-
    
-
    
-void setSourceData(CpuVectorPtr sourceVec)
    
-
    
-
-
    
-
    
-real *getRow(size_t row)
    
-
    
-
-
    
-
    
-virtual real *getRowBuf(size_t row)
    
-
    
-
-
    
-
    
-virtual void mul(CpuSparseMatrix *a, CpuMatrix *b, real scaleAB, real scaleT)
    
-
    
-
-
    
-
    
-
    Public Members
    
-
    
-
    
-CpuVectorPtr sourceDataVec_
    
-
    
-
-
    
-
    
-real *sourceData_
    
-
    
-
-
    
-
    
-
-
    
-
    
-class SparseAutoGrowRowCpuMatrix
    
-
    Inherits from paddle::SparseRowCpuMatrix
    
-
    Subclassed by paddle::CacheRowCpuMatrix
    
-
    
-
    Public Functions
    
-
    
-
    
-SparseAutoGrowRowCpuMatrix(size_t height, size_t width, IndexDictPtr indexDictHandle = nullptr, bool trans = false)
    
-
    
-
-
    
-
    
-real *getRow(size_t row)
    
-
    
-
-
    
-
    
-virtual real *getRowBuf(size_t row)
    
-
    
-
-
    
-
    
-virtual void mul(CpuSparseMatrix *a, CpuMatrix *b, real scaleAB, real scaleT)
    
-
    
-
-
    
-
    
-
-
    
-
    
-class SparsePrefetchRowCpuMatrix
    
-
    #include <SparseRowMatrix.h>
    For prefetching parameters from remote Parameter server. 
    
-
    Inherits from paddle::SparseRowCpuMatrix
    
-
    
-
    Public Functions
    
-
    
-
    
-SparsePrefetchRowCpuMatrix(CpuMemHandlePtr dataHandle, size_t height, size_t width, IndexDictPtr indexDictHandle = nullptr, SyncThreadPool *pool = nullptr, bool trans = false)
    
-
    
-
-
    
-
    
-void addRows(MatrixPtr input)
    
-
    Extract feature ids from input, to fill row indexs.
    
-
    input must be sparse matrix.
    
-
    Can call many times before setup. 
    
-
    
-
-
    
-
    
-void addRows(IVectorPtr ids)
    
-
    
-
-
    
-
    
-void setupIndices()
    
-
    setup global indices of SparseRowMatrix after finish add rows. 
    
-
    
-
-
    
-
    
-
    Protected Functions
    
-
    
-
    
-void addRows(const unsigned int *ids, size_t len)
    
-
    
-
-
    
-
    
-
    Protected Attributes
    
-
    
-
    
-SyncThreadPool *pool_
    
-
    
-
-
    
-
    
-
-
    
 
    
 class SparseRowCpuMatrix
    
 
    #include <SparseRowMatrix.h>
    Sparse Row 
    
-
    Inherits from paddle::CpuMatrix
    
-
    Subclassed by paddle::SparseAutoGrowRowCpuMatrix, paddle::SparsePrefetchRowCpuMatrix
    
+
    Inherits from paddle::CpuMatrix
    
+
    Subclassed by paddle::SparseAutoGrowRowCpuMatrix, paddle::SparsePrefetchRowCpuMatrix
    
 
    
 
    Public Types
    
 
    
@@ -5607,7 +5480,7 @@ if ids[i] == -1, it will be ignored
 
    
 void reserveStore()
    
 
    reserve the storage for rows according to current size of indexDictHandle.
    
-
    This is only used when SparseRowCpuMatrix is constructed with indexDictHandle. 
    
+
    This is only used when SparseRowCpuMatrix is constructed with indexDictHandle. 
    
 
    
 
 
    
@@ -5624,7 +5497,7 @@ if ids[i] == -1, it will be ignored
 
    
 virtual void copyFrom(const real *src, size_t size)
    
 
    Fill data according to row indexs added, setup indices inside.
    
-
    src and size are data and size of normal dense CpuMatrix. 
    
+
    src and size are data and size of normal dense CpuMatrix. 
    
 
    
 
 
    
@@ -5668,9 +5541,9 @@ if ids[i] == -1, it will be ignored
 
 
    
 
    
-void addTo(SparseRowCpuMatrix &dest, size_t tid, size_t numThreads)
    
-
    the second version addTo(), dest is a SparseRowCpuMatrix.
    
-
    The dest’s indices should be setup already, addTo() will check src ids is exist in dest’s indices. 
    
+void addTo(SparseRowCpuMatrix &dest, size_t tid, size_t numThreads)
+
    the second version addTo(), dest is a SparseRowCpuMatrix.
    
+
    The dest’s indices should be setup already, addTo() will check src ids is exist in dest’s indices. 
    
 
    
 
 
    
@@ -5778,12 +5651,139 @@ if ids[i] == -1, it will be ignored
 
 
    
 
+
    
+
    
+class SparsePrefetchRowCpuMatrix
    
+
    #include <SparseRowMatrix.h>
    For prefetching parameters from remote Parameter server. 
    
+
    Inherits from paddle::SparseRowCpuMatrix
    
+
    
+
    Public Functions
    
+
    
+
    
+SparsePrefetchRowCpuMatrix(CpuMemHandlePtr dataHandle, size_t height, size_t width, IndexDictPtr indexDictHandle = nullptr, SyncThreadPool *pool = nullptr, bool trans = false)
    
+
    
+
+
    
+
    
+void addRows(MatrixPtr input)
    
+
    Extract feature ids from input, to fill row indexs.
    
+
    input must be sparse matrix.
    
+
    Can call many times before setup. 
    
+
    
+
+
    
+
    
+void addRows(IVectorPtr ids)
    
+
    
+
+
    
+
    
+void setupIndices()
    
+
    setup global indices of SparseRowMatrix after finish add rows. 
    
+
    
+
+
    
+
    
+
    Protected Functions
    
+
    
+
    
+void addRows(const unsigned int *ids, size_t len)
    
+
    
+
+
    
+
    
+
    Protected Attributes
    
+
    
+
    
+SyncThreadPool *pool_
    
+
    
+
+
    
+
    
+
+
    
+
    
+class SparseAutoGrowRowCpuMatrix
    
+
    Inherits from paddle::SparseRowCpuMatrix
    
+
    Subclassed by paddle::CacheRowCpuMatrix
    
+
    
+
    Public Functions
    
+
    
+
    
+SparseAutoGrowRowCpuMatrix(size_t height, size_t width, IndexDictPtr indexDictHandle = nullptr, bool trans = false)
    
+
    
+
+
    
+
    
+real *getRow(size_t row)
    
+
    
+
+
    
+
    
+virtual real *getRowBuf(size_t row)
    
+
    
+
+
    
+
    
+virtual void mul(CpuSparseMatrix *a, CpuMatrix *b, real scaleAB, real scaleT)
    
+
    
+
+
    
+
    
+
+
    
+
    
+class CacheRowCpuMatrix
    
+
    Inherits from paddle::SparseAutoGrowRowCpuMatrix
    
+
    
+
    Public Functions
    
+
    
+
    
+CacheRowCpuMatrix(size_t height, size_t width, IndexDictPtr indexDictHandle = nullptr, bool trans = false)
    
+
    
+
+
    
+
    
+void setSourceData(CpuVectorPtr sourceVec)
    
+
    
+
+
    
+
    
+real *getRow(size_t row)
    
+
    
+
+
    
+
    
+virtual real *getRowBuf(size_t row)
    
+
    
+
+
    
+
    
+virtual void mul(CpuSparseMatrix *a, CpuMatrix *b, real scaleAB, real scaleT)
    
+
    
+
+
    
+
    
+
    Public Members
    
+
    
+
    
+CpuVectorPtr sourceDataVec_
    
+
    
+
+
    
+
    
+real *sourceData_
    
+
    
+
+
    
+
    
+
 
    
 
    
 class SparseRowIdsCpuMatrix
    
-
    #include <SparseRowMatrix.h>
    Sparse Row Ids Matrix.
    
-
    mostly same as CpuMatrix, but maintain sparse row ids occured, ids are hashed by worker thread id. 
    
-
    Inherits from paddle::CpuMatrix
    
+
    #include <SparseRowMatrix.h>
    Sparse Row Ids Matrix.
    
+
    mostly same as CpuMatrix, but maintain sparse row ids occured, ids are hashed by worker thread id. 
    
+
    Inherits from paddle::CpuMatrix
    
 
    
 
    Public Functions
    
 
    
@@ -5815,12 +5815,12 @@ if ids[i] == -1, it will be ignored
 
    
 
 
    
-
    
-namespace paddle
    
+
    
+namespace paddle
    
 
    
 
    
 class CpuSparseMatrix
    
-
    Inherits from paddle::Matrix
    
+
    Inherits from paddle::Matrix
    
 
    
 
    Public Functions
    
 
    
@@ -5858,7 +5858,7 @@ if ids[i] == -1, it will be ignored
 virtual void resize(size_t newHeight, size_t newWidth)
 
    
 
    Note
    
-
    Original data may not be preserved after resize(). 
    
+
    Original data may not be preserved after resize(). 
    
 
    
 
    
 
    
@@ -5965,19 +5965,19 @@ if ids[i] == -1, it will be ignored
 
 
    
 
    
-virtual SparseValueType getValueType() const
    
+virtual SparseValueType getValueType() const
 
    
 
 
    
 
    
 virtual real *getData()
    
 
    return value_ of sparse matrix 
    
-
    Some times CpuSparseMatrix maybe Matrix, if getValue, must dynamic_cast to CpuSparseMatrix, getData is convenient to get value 
    
+
    Some times CpuSparseMatrix maybe Matrix, if getValue, must dynamic_cast to CpuSparseMatrix, getData is convenient to get value 
    
 
    
 
 
    
 
    
-virtual const real *getData() const
    
+virtual const real *getData() const
 
    
 
 
    
@@ -5995,7 +5995,7 @@ if ids[i] == -1, it will be ignored
 
    
 
    
 virtual void mul(MatrixPtr a, MatrixPtr b, real scaleAB, real scaleT)
    
-
     
    this = scaleAB*(a*b) + scaleT*this
+
     
    this = scaleAB*(a*b) + scaleT*this
 
    
 
    
 
    
@@ -6109,12 +6109,12 @@ if ids[i] == -1, it will be ignored
 
 
    
 
    
-void copyFrom(const CpuSparseMatrix &src)
    
+void copyFrom(const CpuSparseMatrix &src)
 
    
 
 
    
 
    
-virtual void trimFrom(const CpuSparseMatrix &src)
    
+virtual void trimFrom(const CpuSparseMatrix &src)
 
    
 
 
    
@@ -6144,8 +6144,8 @@ if ids[i] == -1, it will be ignored
 
    
 
    
 virtual void copyFrom(const real *src, size_t size)
    
-
    If this is GpuMatrix, src is assumed to be CPU memory
    
-
    If this is CpuMatrix, src is assumed to be CPU memory 
    
+
    If this is GpuMatrix, src is assumed to be CPU memory
    
+
    If this is CpuMatrix, src is assumed to be CPU memory 
    
 
    
 
 
    
@@ -6208,7 +6208,7 @@ if ids[i] == -1, it will be ignored
 
    
 
    
 virtual MatrixPtr clone(size_t height = 0, size_t width = 0, bool useGpu = false)
    
-
    Create a matrix with the same type (GpuMatrix, CpuMatrix, NonValueSparseMatrix, etc.) as this. 
    
+
    Create a matrix with the same type (GpuMatrix, CpuMatrix, NonValueSparseMatrix, etc.) as this. 
    
 
    If height and width is zero, the new matrix will have the same size as this, otherwise the new matrix will have the specified size. 
    
 
    
 
@@ -6221,8 +6221,8 @@ if ids[i] == -1, it will be ignored
 
    
 
    Others
    
 
    
-
    
-namespace paddle
    
+
    
+namespace paddle
    
 
    
 
    Functions
    
 
    
@@ -6250,8 +6250,8 @@ if ids[i] == -1, it will be ignored
 
    
 
 
    
-
    
-namespace paddle
    
+
    
+namespace paddle
    
 
    
 
    
 namespace simd
    
@@ -6427,14 +6427,11 @@ if ids[i] == -1, it will be ignored
 
 
         
    
@@ -6456,14 +6453,14 @@ if ids[i] == -1, it will be ignored
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    • 
-          
  • Matrix Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    • 
+          
  • Matrix Documents »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/source/math/utils/index.html b/doc/source/math/utils/index.html
index 1a7cd038d9..da31ce66b5 100644
--- a/doc/source/math/utils/index.html
+++ b/doc/source/math/utils/index.html
@@ -6,7 +6,7 @@
   
     
     
-    Utils Documents — PaddlePaddle  documentation
+    Utils Documents — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
      
 
@@ -60,7 +60,6 @@
 
    
 
      
 
    • Utils
 
      • 
@@ -90,14 +89,11 @@
 
 
         
    
@@ -119,13 +115,13 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/source/math/utils/utils.html b/doc/source/math/utils/utils.html
index f41363bb43..b45deb6516 100644
--- a/doc/source/math/utils/utils.html
+++ b/doc/source/math/utils/utils.html
@@ -6,7 +6,7 @@
   
     
     
-    Utils — PaddlePaddle  documentation
+    Utils — PaddlePaddle  documentation
     
     
     
@@ -45,9 +45,9 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    • 
-          
  • Utils Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    • 
+          
  • Utils Documents »
    •  
       
     
      
 
@@ -58,36 +58,6 @@
             
   
    
 
    Utils
    
-
    
-
    Bits
    
-
    
-
    
-namespace paddle
    
-
    
-
    Functions
    
-
    
-
    
-template <class T>
    
-
    
-constexpr std::enable_if<(std::is_integral<T>::value && std::is_unsigned<T>::value && sizeof(T) <= sizeof(unsigned int)), unsigned int>::type findLastSet(T x)
    
-
    From Facebook folly: https://github.com/facebook/folly/blob/master/folly/Bits.h
    
-
    findLastSet: return the 1-based index of the highest bit set
    
-
    for x > 0: 
    
-\[ findLastSet(x) = 1 + \floor*{\log_{2}x} \]
    
- 
    
-
    
-
-
    
-
    
-template <class T>
    
-
    
-constexpr std::enable_if<(std::is_integral<T>::value && std::is_unsigned<T>::value && sizeof(T) > sizeof(unsigned int) && sizeof(T) <= sizeof(unsigned long)), unsigned int>::type findLastSet(T x)
    
-
    
-
-
    
-
    
-
-
    
 
    
 
    Memory Handle
    
 
    
@@ -111,52 +81,10 @@
 
    
 
 
    
-
    
-
    
-class CpuMemoryHandle
    
-
    #include <MemoryHandle.h>
    Wrapper class for raw cpu memory handle.
    
-
    The raw handle will be released at destructor 
    
-
    Inherits from paddle::MemoryHandle
    
-
    
-
    Public Functions
    
-
    
-
    
-CpuMemoryHandle(size_t size)
    
-
    
-
-
    
-
    
-virtual ~CpuMemoryHandle()
    
-
    
-
-
    
-
    
-
-
    
-
    
-class GpuMemoryHandle
    
-
    #include <MemoryHandle.h>
    Wrapper class for raw gpu memory handle.
    
-
    The raw handle will be released at destructor 
    
-
    Inherits from paddle::MemoryHandle
    
-
    
-
    Public Functions
    
-
    
-
    
-GpuMemoryHandle(size_t size)
    
-
    
-
-
    
-
    
-virtual ~GpuMemoryHandle()
    
-
    
-
-
    
-
    
-
 
    
 
    
 class MemoryHandle
    
-
    Subclassed by paddle::CpuMemoryHandle, paddle::GpuMemoryHandle
    
+
    Subclassed by paddle::CpuMemoryHandle, paddle::GpuMemoryHandle
    
 
    
 
    Public Functions
    
 
    
@@ -219,17 +147,59 @@
 
    
 
    
 
+
    
+
    
+class GpuMemoryHandle
    
+
    #include <MemoryHandle.h>
    Wrapper class for raw gpu memory handle.
    
+
    The raw handle will be released at destructor 
    
+
    Inherits from paddle::MemoryHandle
    
+
    
+
    Public Functions
    
+
    
+
    
+GpuMemoryHandle(size_t size)
    
+
    
+
+
    
+
    
+virtual ~GpuMemoryHandle()
    
+
    
+
+
    
+
    
+
+
    
+
    
+class CpuMemoryHandle
    
+
    #include <MemoryHandle.h>
    Wrapper class for raw cpu memory handle.
    
+
    The raw handle will be released at destructor 
    
+
    Inherits from paddle::MemoryHandle
    
+
    
+
    Public Functions
    
+
    
+
    
+CpuMemoryHandle(size_t size)
    
+
    
+
+
    
+
    
+virtual ~CpuMemoryHandle()
    
+
    
+
+
    
+
    
+
 
    
 
 
    
 
    
-namespace paddle
    
+namespace paddle
 
    
 
    
 class Allocator
    
-
    #include <Allocator.h>
    Allocator base class. 
    
-
    This is the base class of all Allocator class. 
    
-
    Subclassed by paddle::CpuAllocator, paddle::CudaHostAllocator, paddle::GpuAllocator
    
+
    #include <Allocator.h>
    Allocator base class. 
    
+
    This is the base class of all Allocator class. 
    
+
    Subclassed by paddle::CpuAllocator, paddle::CudaHostAllocator, paddle::GpuAllocator
    
 
    
 
    Public Functions
    
 
    
@@ -259,7 +229,7 @@
 
    
 class CpuAllocator
    
 
    #include <Allocator.h>
    CPU allocator implementation. 
    
-
    Inherits from paddle::Allocator
    
+
    Inherits from paddle::Allocator
    
 
    
 
    Public Functions
    
 
    
@@ -308,21 +278,21 @@
 
    
 
 
    
-
    
-class CudaHostAllocator
    
-
    #include <Allocator.h>
    CPU pinned memory allocator implementation. 
    
-
    Inherits from paddle::Allocator
    
+
    
+class GpuAllocator
    
+
    #include <Allocator.h>
    GPU allocator implementation. 
    
+
    Inherits from paddle::Allocator
    
 
    
 
    Public Functions
    
 
    
-
    
-~CudaHostAllocator()
    
+
    
+~GpuAllocator()
    
 
    
 
 
    
-
    
-virtual void *alloc(size_t size)
    
-
    Allocate pinned memory. 
    
+
    
+virtual void *alloc(size_t size)
    
+
    Allocate GPU memory. 
    
 
    
 
    Return
    
 
    Pointer to the allocated memory 
    
@@ -337,9 +307,9 @@
 
    
 
 
    
-
    
-virtual void free(void *ptr)
    
-
    Free the pinned memory. 
    
+
    
+virtual void free(void *ptr)
    
+
    Free the GPU memory. 
    
 
    
 
    Parameters
    
 
      
@@ -352,29 +322,29 @@
 
    
 
 
    
-
    
-virtual std::string getName()
    
+
    
+virtual std::string getName()
    
 
    
 
 
    
 
    
 
 
    
-
    
-class GpuAllocator
    
-
    #include <Allocator.h>
    GPU allocator implementation. 
    
-
    Inherits from paddle::Allocator
    
+
    
+class CudaHostAllocator
    
+
    #include <Allocator.h>
    CPU pinned memory allocator implementation. 
    
+
    Inherits from paddle::Allocator
    
 
    
 
    Public Functions
    
 
    
-
    
-~GpuAllocator()
    
+
    
+~CudaHostAllocator()
    
 
    
 
 
    
-
    
-virtual void *alloc(size_t size)
    
-
    Allocate GPU memory. 
    
+
    
+virtual void *alloc(size_t size)
    
+
    Allocate pinned memory. 
    
 
    
 
    Return
    
 
    Pointer to the allocated memory 
    
@@ -389,9 +359,9 @@
 
    
 
 
    
-
    
-virtual void free(void *ptr)
    
-
    Free the GPU memory. 
    
+
    
+virtual void free(void *ptr)
    
+
    Free the pinned memory. 
    
 
    
 
    Parameters
    
 
      
@@ -404,8 +374,8 @@
 
    
 
 
    
-
    
-virtual std::string getName()
    
+
    
+virtual std::string getName()
    
 
    
 
 
    
@@ -414,8 +384,8 @@
 
    
 
 
    
-
    
-namespace paddle
    
+
    
+namespace paddle
    
 
    
 
    
 class PoolAllocator
    
@@ -429,7 +399,7 @@
 
    
 
    Parameters
    
 
      
-
    • allocator - 
      a Allocator object. 
      
+
    • allocator - 
      a Allocator object. 
      
 
      • 
 
    • sizeLimit - 
      The maximum size memory can be managed, if sizeLimit == 0, the pool allocator is a simple wrapper of allocator. 
      
 
      • 
@@ -512,8 +482,8 @@
 
    
 
 
    
-
    
-namespace paddle
    
+
    
+namespace paddle
    
 
    
 
    
 class StorageEngine
    
@@ -545,7 +515,7 @@
 
    Public Static Functions
    
 
    
 
    
-StorageEngine *singleton()
    
+StorageEngine *singleton()
 
    
 
    Return
    
 
    Storage singleton 
    
@@ -601,7 +571,6 @@
   
    Table Of Contents
    
   
      
 
    • Utils
 
      • 
@@ -623,14 +592,11 @@
 
 
         
    
@@ -652,14 +618,14 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    • 
-          
  • Utils Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    • 
+          
  • Utils Documents »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/source/parameter/optimizer/index.html b/doc/source/parameter/optimizer/index.html
index 5a5abd1469..2d54860367 100644
--- a/doc/source/parameter/optimizer/index.html
+++ b/doc/source/parameter/optimizer/index.html
@@ -6,7 +6,7 @@
   
     
     
-    Parameter Documents — PaddlePaddle  documentation
+    Parameter Documents — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
      
 
@@ -86,14 +86,11 @@
 
 
         
    
@@ -115,13 +112,13 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/source/parameter/optimizer/optimizer.html b/doc/source/parameter/optimizer/optimizer.html
index f6e4b991e1..5ee7f3ba72 100644
--- a/doc/source/parameter/optimizer/optimizer.html
+++ b/doc/source/parameter/optimizer/optimizer.html
@@ -6,7 +6,7 @@
   
     
     
-    Optimizer — PaddlePaddle  documentation
+    Optimizer — PaddlePaddle  documentation
     
     
     
@@ -45,9 +45,9 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    • 
-          
  • Parameter Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    • 
+          
  • Parameter Documents »
    •  
       
     
      
 
@@ -62,39 +62,156 @@
 
    
 namespace paddle
    
 
    
-
    
-class AdaDeltaParameterOptimizer
    
-
    Inherits from paddle::ParameterOptimizer
    
+
    
+class SgdOptimizer
    
+
    Inherits from paddle::ParameterOptimizer
    
 
    
 
    Public Functions
    
 
    
-
    
-AdaDeltaParameterOptimizer(const OptimizationConfig &optConfig)
    
+
    
+SgdOptimizer(const OptimizationConfig &optConfig)
    
 
    
 
 
    
-
    
-virtual void startBatch(int64_t numSamplesProcessed)
    
+
    
+virtual void startBatch(int64_t numSamplesProcessed)
    
 
    called by Trainer before forward() of a batch. 
    
 
    
 
 
    
-
    
-virtual void update(const VectorPtr vecs[], const ParameterConfig &config, size_t sparseId) const
    
-
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
+
    
+virtual void update(const VectorPtr vecs[], const ParameterConfig &config, size_t sparseId) const
    
+
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
+
    
+
+
    
+
    
+virtual void finishBatch()
    
+
    called by Trainer after backward() of a batch 
    
+
    
+
+
    
+
    
+
+
    
+
    
+class SparseMomentumParameterOptimizer
    
+
    Inherits from paddle::ParameterOptimizer
    
+
    
+
    Public Functions
    
+
    
+
    
+SparseMomentumParameterOptimizer(const OptimizationConfig &optConfig)
    
+
    
+
+
    
+
    
+virtual void init(size_t numRows, const ParameterConfig *config)
    
+
    For sparse update, optimizer can maintain numRows of timer(t0). Some sparse optimizer depends on parameter config in functions such as startBatch(). Optimizer can get it here. But notice that, not all callers can pass config here, so the optimizer should check config passed in is not null ptr. 
    
+
    
+
+
    
+
    
+virtual void startBatch(int64_t numSamplesProcessed)
    
+
    called by Trainer before forward() of a batch. 
    
+
    
+
+
    
+
    
+virtual void update(const VectorPtr vecs[], const ParameterConfig &config, size_t sparseId) const
    
+
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
+
    
+
+
    
+
    
+virtual ParameterOptimizer::TraverseCallback needSpecialTraversal(const ParameterConfig &config) const
    
+
    following hooks useful for sparse update, because the traversal in block costs. called by Trainer after update and before finishBatch e.g. Trainer call like this:
    
+
    startBatch();
+if (dense) {
+  update(blockVec);
+} else {//sparse
+  for (row : rows_in_block) {update(rowVec)}
+}
+auto callback = needSpecialTraversal();
+if (callback) {
+  // do traverse, maybe multi-thread
+  if (dense) {
+    callback();
+  } else {//sparse
+    for (row : all_rows_in_block) {callback();}
+  }
+}
+finishBatch();
+
    
+
    
+
    
+
    
+
    Return
    
+
    callback if need traverse, else return nullptr. It should be no state change. 
    
+
    
+
    
+
    
+
+
    
+
    
+virtual void finishBatch()
    
+
    called by Trainer after backward() of a batch 
    
 
    
 
 
    
 
    
 
    Protected Attributes
    
 
    
-
    
-real rou_
    
+
    
+int64_t timer_
    
 
    
 
 
    
-
    
-real epsilon_
    
+
    
+std::vector<int64_t> t0Vec_
    
+
    
+
+
    
+
    
+bool isParameterSparse_
    
+
    
+
+
    
+
    
+
    Private Members
    
+
    
+
    
+real alpha_
    
+
    
+
+
    
+
    
+real beta_
    
+
    
+
+
    
+
    
+real tau_
    
+
    
+
+
    
+
    
+real gamma_
    
+
    
+
+
    
+
    
+real threshold_
    
+
    
+
+
    
+
    
+real momentum_
    
+
    
+
+
    
+
    
+real decayRate_
    
 
    
 
 
    
@@ -103,7 +220,7 @@
 
    
 
    
 class AdagradParameterOptimizer
    
-
    Inherits from paddle::ParameterOptimizer
    
+
    Inherits from paddle::ParameterOptimizer
    
 
    
 
    Public Functions
    
 
    
@@ -120,29 +237,29 @@
 
    
 
    
 virtual void update(const VectorPtr vecs[], const ParameterConfig &config, size_t sparseId) const
    
-
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
+
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
 
    
 
 
    
 
    
 virtual ParameterOptimizer::TraverseCallback needSpecialTraversal(const ParameterConfig &config) const
    
 
    following hooks useful for sparse update, because the traversal in block costs. called by Trainer after update and before finishBatch e.g. Trainer call like this:
    
-
    startBatch();
-if (dense) {
-  update(blockVec);
-} else {//sparse
-  for (row : rows_in_block) {update(rowVec)}
-}
-auto callback = needSpecialTraversal();
-if (callback) {
-  // do traverse, maybe multi-thread
-  if (dense) {
-    callback();
-  } else {//sparse
-    for (row : all_rows_in_block) {callback();}
-  }
-}
-finishBatch();
+
    startBatch();
+if (dense) {
+  update(blockVec);
+} else {//sparse
+  for (row : rows_in_block) {update(rowVec)}
+}
+auto callback = needSpecialTraversal();
+if (callback) {
+  // do traverse, maybe multi-thread
+  if (dense) {
+    callback();
+  } else {//sparse
+    for (row : all_rows_in_block) {callback();}
+  }
+}
+finishBatch();
 
    
 
    
 
    
@@ -173,140 +290,110 @@ finishBatch();
 
    
 
 
    
-
    
-class AdamaxParameterOptimizer
    
-
    #include <FirstOrderOptimizer.h>
    AdaMax Optimizer. Reference Paper: http://arxiv.org/abs/1412.6980 Algorithm 2 
    
-
    Inherits from paddle::ParameterOptimizer
    
+
    
+class AdaDeltaParameterOptimizer
    
+
    Inherits from paddle::ParameterOptimizer
    
 
    
 
    Public Functions
    
 
    
-
    
-AdamaxParameterOptimizer(const OptimizationConfig &optConfig)
    
+
    
+AdaDeltaParameterOptimizer(const OptimizationConfig &optConfig)
    
 
    
 
 
    
-
    
-virtual void finishBatch()
    
-
    called by Trainer after backward() of a batch 
    
+
    
+virtual void startBatch(int64_t numSamplesProcessed)
    
+
    called by Trainer before forward() of a batch. 
    
 
    
 
 
    
-
    
-virtual void update(const VectorPtr vecs[], const ParameterConfig &config, size_t sparseId) const
    
-
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
+
    
+virtual void update(const VectorPtr vecs[], const ParameterConfig &config, size_t sparseId) const
    
+
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
 
    
 
 
    
 
    
 
    Protected Attributes
    
 
    
-
    
-real beta1_
    
-
    
-
-
    
-
    
-real beta2_
    
-
    
-
-
    
-
    
-int64_t step_
    
+
    
+real rou_
    
 
    
 
 
    
-
    
-real learningRate_
    
+
    
+real epsilon_
    
 
    
 
 
    
 
    
 
 
    
-
    
-class AdamParameterOptimizer
    
-
    #include <FirstOrderOptimizer.h>
    Adam Optimizer. Reference Paper: http://arxiv.org/abs/1412.6980 Algorithm 1 
    
-
    Inherits from paddle::ParameterOptimizer
    
+
    
+class RMSPropParameterOptimizer
    
+
    Inherits from paddle::ParameterOptimizer
    
 
    
 
    Public Functions
    
 
    
-
    
-AdamParameterOptimizer(const OptimizationConfig &optConfig)
    
+
    
+RMSPropParameterOptimizer(const OptimizationConfig &optConfig)
    
 
    
 
 
    
-
    
-virtual void finishBatch()
    
+
    
+virtual void init(size_t numRows, const ParameterConfig *config)
    
+
    For sparse update, optimizer can maintain numRows of timer(t0). Some sparse optimizer depends on parameter config in functions such as startBatch(). Optimizer can get it here. But notice that, not all callers can pass config here, so the optimizer should check config passed in is not null ptr. 
    
+
    
+
+
    
+
    
+virtual void startBatch(int64_t numSamplesProcessed)
    
+
    called by Trainer before forward() of a batch. 
    
+
    
+
+
    
+
    
+virtual void finishBatch()
    
 
    called by Trainer after backward() of a batch 
    
 
    
 
 
    
-
    
-virtual void update(const VectorPtr vecs[], const ParameterConfig &config, size_t sparseId) const
    
-
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
+
    
+virtual void update(const VectorPtr vecs[], const ParameterConfig &config, size_t sparseId) const
    
+
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
 
    
 
 
    
 
    
 
    Protected Attributes
    
 
    
-
    
-real beta1_
    
-
    
-
-
    
-
    
-real beta2_
    
-
    
-
-
    
-
    
-real epsilon_
    
+
    
+real rou_
    
 
    
 
 
    
-
    
-int64_t step_
    
+
    
+real epsilon_
    
 
    
 
 
    
-
    
-real learningRate_
    
-
    
-
-
    
+
    
+int64_t timer_
    
+
    counting batches, donot need catch up with t(timer_) is current time, t0(t0Vec_) are last occur time of i rows. if one block is update by multi threads, caller should hash sparse ids to avoid write conflict in t0Vec_. 
    
 
    
 
-
    
-
    
-class AddOptimizer
    
-
    Inherits from paddle::ParameterOptimizer
    
-
    
-
    Public Functions
    
-
    
-
    
-AddOptimizer(const OptimizationConfig &optConfig)
    
+
    
+
    
+std::vector<int64_t> t0Vec_
    
 
    
 
-
    
-
    
-virtual void startBatch(int64_t numSamplesProcessed)
    
-
    called by Trainer before forward() of a batch. 
    
-
    
-
-
    
-
    
-virtual void update(const VectorPtr vecs[], const ParameterConfig &config, size_t sparseId) const
    
-
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
-
    
-
 
    
 
    
 
 
    
 
    
 class DecayedAdagradParameterOptimizer
    
-
    Inherits from paddle::ParameterOptimizer
    
+
    Inherits from paddle::ParameterOptimizer
    
 
    
 
    Public Functions
    
 
    
@@ -317,7 +404,7 @@ finishBatch();
 
    
 
    
 virtual void init(size_t numRows, const ParameterConfig *config)
    
-
    For sparse update, optimizer can maintain numRows of timer(t0). Some sparse optimizer depends on parameter config in functions such as startBatch(). Optimizer can get it here. But notice that, not all callers can pass config here, so the optimizer should check config passed in is not null ptr. 
    
+
    For sparse update, optimizer can maintain numRows of timer(t0). Some sparse optimizer depends on parameter config in functions such as startBatch(). Optimizer can get it here. But notice that, not all callers can pass config here, so the optimizer should check config passed in is not null ptr. 
    
 
    
 
 
    
@@ -335,7 +422,7 @@ finishBatch();
 
    
 
    
 virtual void update(const VectorPtr vecs[], const ParameterConfig &config, size_t sparseId) const
    
-
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
+
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
 
    
 
 
    
@@ -366,259 +453,215 @@ finishBatch();
 
    
 
 
    
-
    
-class DummyOptimizer
    
-
    Inherits from paddle::ParameterOptimizer
    
+
    
+class AdamParameterOptimizer
    
+
    #include <FirstOrderOptimizer.h>
    Adam Optimizer. Reference Paper: http://arxiv.org/abs/1412.6980 Algorithm 1 
    
+
    Inherits from paddle::ParameterOptimizer
    
 
    
 
    Public Functions
    
 
    
-
    
-DummyOptimizer(const OptimizationConfig &optConfig)
    
+
    
+AdamParameterOptimizer(const OptimizationConfig &optConfig)
    
 
    
 
 
    
-
    
-virtual void update(const VectorPtr vecs[], const ParameterConfig &config, size_t sparseId) const
    
-
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
+
    
+virtual void finishBatch()
    
+
    called by Trainer after backward() of a batch 
    
 
    
 
-
    
+
    
+
    
+virtual void update(const VectorPtr vecs[], const ParameterConfig &config, size_t sparseId) const
    
+
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
 
    
 
-
    
-
    
-class OptimizerWithGradientClipping
    
-
    Inherits from paddle::ParameterOptimizer
    
+
    
 
    
-
    Public Functions
    
-
    
-
    
-OptimizerWithGradientClipping(const OptimizationConfig &optConfig, ParameterOptimizer *optimizer)
    
+
    Protected Attributes
    
+
    
+
    
+real beta1_
    
 
    
 
-
    
-
    
-virtual void init(size_t numRows, const ParameterConfig *config)
    
-
    For sparse update, optimizer can maintain numRows of timer(t0). Some sparse optimizer depends on parameter config in functions such as startBatch(). Optimizer can get it here. But notice that, not all callers can pass config here, so the optimizer should check config passed in is not null ptr. 
    
-
    
-
-
    
-
    
-virtual void startPass()
    
+
    
+
    
+real beta2_
    
 
    
 
-
    
-
    
-virtual void finishPass()
    
+
    
+
    
+real epsilon_
    
 
    
 
-
    
-
    
-virtual void startBatch(int64_t numSamplesProcessed)
    
-
    called by Trainer before forward() of a batch. 
    
-
    
-
-
    
-
    
-virtual void finishBatch()
    
-
    called by Trainer after backward() of a batch 
    
-
    
-
-
    
-
    
-virtual TraverseCallback needSpecialTraversal(const ParameterConfig &config) const
    
-
    following hooks useful for sparse update, because the traversal in block costs. called by Trainer after update and before finishBatch e.g. Trainer call like this:
    
-
    startBatch();
-if (dense) {
-  update(blockVec);
-} else {//sparse
-  for (row : rows_in_block) {update(rowVec)}
-}
-auto callback = needSpecialTraversal();
-if (callback) {
-  // do traverse, maybe multi-thread
-  if (dense) {
-    callback();
-  } else {//sparse
-    for (row : all_rows_in_block) {callback();}
-  }
-}
-finishBatch();
-
    
-
    
-
    
-
    
-
    Return
    
-
    callback if need traverse, else return nullptr. It should be no state change. 
    
-
    
-
    
-
    
-
-
    
-
    
-virtual void update(const VectorPtr vecs[], const ParameterConfig &config, size_t sparseId) const
    
-
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
-
    
-
-
    
-
    
-virtual void setNoDecay()
    
+
    
+
    
+int64_t step_
    
 
    
 
-
    
-
    
-
    Protected Attributes
    
 
    
-
    
-std::unique_ptr<ParameterOptimizer> optimizer_
    
+
    
+real learningRate_
    
 
    
 
 
    
 
    
 
 
    
-
    
-class RMSPropParameterOptimizer
    
-
    Inherits from paddle::ParameterOptimizer
    
+
    
+class AdamaxParameterOptimizer
    
+
    #include <FirstOrderOptimizer.h>
    AdaMax Optimizer. Reference Paper: http://arxiv.org/abs/1412.6980 Algorithm 2 
    
+
    Inherits from paddle::ParameterOptimizer
    
 
    
 
    Public Functions
    
 
    
-
    
-RMSPropParameterOptimizer(const OptimizationConfig &optConfig)
    
+
    
+AdamaxParameterOptimizer(const OptimizationConfig &optConfig)
    
 
    
 
 
    
-
    
-virtual void init(size_t numRows, const ParameterConfig *config)
    
-
    For sparse update, optimizer can maintain numRows of timer(t0). Some sparse optimizer depends on parameter config in functions such as startBatch(). Optimizer can get it here. But notice that, not all callers can pass config here, so the optimizer should check config passed in is not null ptr. 
    
-
    
-
-
    
-
    
-virtual void startBatch(int64_t numSamplesProcessed)
    
-
    called by Trainer before forward() of a batch. 
    
-
    
-
-
    
-
    
-virtual void finishBatch()
    
+
    
+virtual void finishBatch()
    
 
    called by Trainer after backward() of a batch 
    
 
    
 
 
    
-
    
-virtual void update(const VectorPtr vecs[], const ParameterConfig &config, size_t sparseId) const
    
-
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
+
    
+virtual void update(const VectorPtr vecs[], const ParameterConfig &config, size_t sparseId) const
    
+
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
 
    
 
 
    
 
    
 
    Protected Attributes
    
 
    
-
    
-real rou_
    
+
    
+real beta1_
    
 
    
 
 
    
-
    
-real epsilon_
    
+
    
+real beta2_
    
 
    
 
 
    
-
    
-int64_t timer_
    
-
    counting batches, donot need catch up with t(timer_) is current time, t0(t0Vec_) are last occur time of i rows. if one block is update by multi threads, caller should hash sparse ids to avoid write conflict in t0Vec_. 
    
-
    
+
    
+int64_t step_
    
+
    
 
 
    
-
    
-std::vector<int64_t> t0Vec_
    
+
    
+real learningRate_
    
 
    
 
 
    
 
    
 
 
    
-
    
-class SgdOptimizer
    
-
    Inherits from paddle::ParameterOptimizer
    
+
    
+class AddOptimizer
    
+
    Inherits from paddle::ParameterOptimizer
    
 
    
 
    Public Functions
    
 
    
-
    
-SgdOptimizer(const OptimizationConfig &optConfig)
    
+
    
+AddOptimizer(const OptimizationConfig &optConfig)
    
 
    
 
 
    
-
    
-virtual void startBatch(int64_t numSamplesProcessed)
    
+
    
+virtual void startBatch(int64_t numSamplesProcessed)
    
 
    called by Trainer before forward() of a batch. 
    
 
    
 
 
    
-
    
-virtual void update(const VectorPtr vecs[], const ParameterConfig &config, size_t sparseId) const
    
-
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
+
    
+virtual void update(const VectorPtr vecs[], const ParameterConfig &config, size_t sparseId) const
    
+
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
 
    
 
+
    
+
    
+
+
    
+
    
+class DummyOptimizer
    
+
    Inherits from paddle::ParameterOptimizer
    
+
    
+
    Public Functions
    
+
    
+
    
+DummyOptimizer(const OptimizationConfig &optConfig)
    
+
    
+
 
    
-
    
-virtual void finishBatch()
    
-
    called by Trainer after backward() of a batch 
    
+
    
+virtual void update(const VectorPtr vecs[], const ParameterConfig &config, size_t sparseId) const
    
+
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
 
    
 
 
    
 
    
 
 
    
-
    
-class SparseMomentumParameterOptimizer
    
-
    Inherits from paddle::ParameterOptimizer
    
+
    
+class OptimizerWithGradientClipping
    
+
    Inherits from paddle::ParameterOptimizer
    
 
    
 
    Public Functions
    
 
    
-
    
-SparseMomentumParameterOptimizer(const OptimizationConfig &optConfig)
    
+
    
+OptimizerWithGradientClipping(const OptimizationConfig &optConfig, ParameterOptimizer *optimizer)
    
 
    
 
 
    
-
    
-virtual void init(size_t numRows, const ParameterConfig *config)
    
-
    For sparse update, optimizer can maintain numRows of timer(t0). Some sparse optimizer depends on parameter config in functions such as startBatch(). Optimizer can get it here. But notice that, not all callers can pass config here, so the optimizer should check config passed in is not null ptr. 
    
+
    
+virtual void init(size_t numRows, const ParameterConfig *config)
    
+
    For sparse update, optimizer can maintain numRows of timer(t0). Some sparse optimizer depends on parameter config in functions such as startBatch(). Optimizer can get it here. But notice that, not all callers can pass config here, so the optimizer should check config passed in is not null ptr. 
    
 
    
 
 
    
-
    
-virtual void startBatch(int64_t numSamplesProcessed)
    
+
    
+virtual void startPass()
    
+
    
+
+
    
+
    
+virtual void finishPass()
    
+
    
+
+
    
+
    
+virtual void startBatch(int64_t numSamplesProcessed)
    
 
    called by Trainer before forward() of a batch. 
    
 
    
 
 
    
-
    
-virtual void update(const VectorPtr vecs[], const ParameterConfig &config, size_t sparseId) const
    
-
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
+
    
+virtual void finishBatch()
    
+
    called by Trainer after backward() of a batch 
    
 
    
 
 
    
-
    
-virtual ParameterOptimizer::TraverseCallback needSpecialTraversal(const ParameterConfig &config) const
    
+
    
+virtual TraverseCallback needSpecialTraversal(const ParameterConfig &config) const
    
 
    following hooks useful for sparse update, because the traversal in block costs. called by Trainer after update and before finishBatch e.g. Trainer call like this:
    
-
    startBatch();
-if (dense) {
-  update(blockVec);
-} else {//sparse
-  for (row : rows_in_block) {update(rowVec)}
-}
-auto callback = needSpecialTraversal();
-if (callback) {
-  // do traverse, maybe multi-thread
-  if (dense) {
-    callback();
-  } else {//sparse
-    for (row : all_rows_in_block) {callback();}
-  }
-}
-finishBatch();
+
    startBatch();
+if (dense) {
+  update(blockVec);
+} else {//sparse
+  for (row : rows_in_block) {update(rowVec)}
+}
+auto callback = needSpecialTraversal();
+if (callback) {
+  // do traverse, maybe multi-thread
+  if (dense) {
+    callback();
+  } else {//sparse
+    for (row : all_rows_in_block) {callback();}
+  }
+}
+finishBatch();
 
    
 
    
 
    
@@ -630,65 +673,22 @@ finishBatch();
 
    
 
 
    
-
    
-virtual void finishBatch()
    
-
    called by Trainer after backward() of a batch 
    
+
    
+virtual void update(const VectorPtr vecs[], const ParameterConfig &config, size_t sparseId) const
    
+
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
 
    
 
-
    
-
    
-
    Protected Attributes
    
-
    
-
    
-int64_t timer_
    
-
    
-
-
    
-
    
-std::vector<int64_t> t0Vec_
    
-
    
-
-
    
-
    
-bool isParameterSparse_
    
+
    
+
    
+virtual void setNoDecay()
    
 
    
 
 
    
 
    
-
    Private Members
    
-
    
-
    
-real alpha_
    
-
    
-
-
    
-
    
-real beta_
    
-
    
-
-
    
-
    
-real tau_
    
-
    
-
-
    
-
    
-real gamma_
    
-
    
-
-
    
-
    
-real threshold_
    
-
    
-
-
    
-
    
-real momentum_
    
-
    
-
+
    Protected Attributes
    
 
    
-
    
-real decayRate_
    
+
    
+std::unique_ptr<ParameterOptimizer> optimizer_
    
 
    
 
 
    
@@ -698,12 +698,12 @@ finishBatch();
 
 
    
 
    
-namespace paddle
    
+namespace paddle
 
    
 
    
 class AverageOptimizer
    
-
    Inherits from paddle::ParameterOptimizer
    
-
    Subclassed by paddle::AverageSparseOptimizer
    
+
    Inherits from paddle::ParameterOptimizer
    
+
    Subclassed by paddle::AverageSparseOptimizer
    
 
    
 
    Public Functions
    
 
    
@@ -714,7 +714,7 @@ finishBatch();
 
    
 
    
 virtual void init(size_t numRows, const ParameterConfig *config)
    
-
    For sparse update, optimizer can maintain numRows of timer(t0). Some sparse optimizer depends on parameter config in functions such as startBatch(). Optimizer can get it here. But notice that, not all callers can pass config here, so the optimizer should check config passed in is not null ptr. 
    
+
    For sparse update, optimizer can maintain numRows of timer(t0). Some sparse optimizer depends on parameter config in functions such as startBatch(). Optimizer can get it here. But notice that, not all callers can pass config here, so the optimizer should check config passed in is not null ptr. 
    
 
    
 
 
    
@@ -742,29 +742,29 @@ finishBatch();
 
    
 
    
 virtual void update(const VectorPtr vecs[], const ParameterConfig &config, size_t sparseId) const
    
-
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
+
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
 
    
 
 
    
 
    
 virtual ParameterOptimizer::TraverseCallback needSpecialTraversal(const ParameterConfig &config) const
    
 
    following hooks useful for sparse update, because the traversal in block costs. called by Trainer after update and before finishBatch e.g. Trainer call like this:
    
-
    startBatch();
-if (dense) {
-  update(blockVec);
-} else {//sparse
-  for (row : rows_in_block) {update(rowVec)}
-}
-auto callback = needSpecialTraversal();
-if (callback) {
-  // do traverse, maybe multi-thread
-  if (dense) {
-    callback();
-  } else {//sparse
-    for (row : all_rows_in_block) {callback();}
-  }
-}
-finishBatch();
+
    startBatch();
+if (dense) {
+  update(blockVec);
+} else {//sparse
+  for (row : rows_in_block) {update(rowVec)}
+}
+auto callback = needSpecialTraversal();
+if (callback) {
+  // do traverse, maybe multi-thread
+  if (dense) {
+    callback();
+  } else {//sparse
+    for (row : all_rows_in_block) {callback();}
+  }
+}
+finishBatch();
 
    
 
    
 
    
@@ -778,18 +778,18 @@ finishBatch();
 
    
 
    
 virtual TraverseCallback startCatchUpWith() const
    
-
    following hooks catch up with current time for sparse update, In the beginning, call startCatchUpWith() and check return. In the end, call finishCatchUpWith() to finish state. callback do the actual works, can call many times for sparse data. e.g. Trainer call like this:
    
-
    auto callback = startCatchUpWith();
-if (callback) {
-  // do catch up with, maybe multi-thread
-  if (dense) {
-    callback();
-  } else {//sparse
-    for (row : rows_in_block) {callback();}
-  }
-  // finish catch up with, main thread
-  finishCatchUpWith();
-}
+
    following hooks catch up with current time for sparse update, In the beginning, call startCatchUpWith() and check return. In the end, call finishCatchUpWith() to finish state. callback do the actual works, can call many times for sparse data. e.g. Trainer call like this:
    
+
    auto callback = startCatchUpWith();
+if (callback) {
+  // do catch up with, maybe multi-thread
+  if (dense) {
+    callback();
+  } else {//sparse
+    for (row : rows_in_block) {callback();}
+  }
+  // finish catch up with, main thread
+  finishCatchUpWith();
+}
 
    
 
    
 
    
@@ -810,7 +810,7 @@ if (callback) {
 virtual ParameterOptimizer::TraverseCallback apply()
 
    following two hooks used by averager, apply to final parameter value (PARAMETER_VALUE or PARAMETER_APPLY).
    
 
    restore() will restore orginal value if it apply to PARAMETER_VALUE. Caller must ensure it’s catched up with current time before apply.
    
-
    Use returned callback same way as callback returned by ParameterOptimizer::needSpecialTraversal() 
    
+
    Use returned callback same way as callback returned by ParameterOptimizer::needSpecialTraversal() 
    
 
    
 
 
    
@@ -901,7 +901,7 @@ if (callback) {
 
    
 
    
 class AverageSparseOptimizer
    
-
    Inherits from paddle::AverageOptimizer
    
+
    Inherits from paddle::AverageOptimizer
    
 
    
 
    Public Functions
    
 
    
@@ -912,7 +912,7 @@ if (callback) {
 
    
 
    
 virtual void init(size_t numRows, const ParameterConfig *config)
    
-
    For sparse update, optimizer can maintain numRows of timer(t0). Some sparse optimizer depends on parameter config in functions such as startBatch(). Optimizer can get it here. But notice that, not all callers can pass config here, so the optimizer should check config passed in is not null ptr. 
    
+
    For sparse update, optimizer can maintain numRows of timer(t0). Some sparse optimizer depends on parameter config in functions such as startBatch(). Optimizer can get it here. But notice that, not all callers can pass config here, so the optimizer should check config passed in is not null ptr. 
    
 
    
 
 
    
@@ -924,7 +924,7 @@ if (callback) {
 
    
 
    
 virtual void update(const VectorPtr vecs[], const ParameterConfig &config, size_t sparseId) const
    
-
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
+
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
 
    
 
 
    
@@ -935,18 +935,18 @@ if (callback) {
 
    
 
    
 virtual ParameterOptimizer::TraverseCallback startCatchUpWith() const
    
-
    following hooks catch up with current time for sparse update, In the beginning, call startCatchUpWith() and check return. In the end, call finishCatchUpWith() to finish state. callback do the actual works, can call many times for sparse data. e.g. Trainer call like this:
    
-
    auto callback = startCatchUpWith();
-if (callback) {
-  // do catch up with, maybe multi-thread
-  if (dense) {
-    callback();
-  } else {//sparse
-    for (row : rows_in_block) {callback();}
-  }
-  // finish catch up with, main thread
-  finishCatchUpWith();
-}
+
    following hooks catch up with current time for sparse update, In the beginning, call startCatchUpWith() and check return. In the end, call finishCatchUpWith() to finish state. callback do the actual works, can call many times for sparse data. e.g. Trainer call like this:
    
+
    auto callback = startCatchUpWith();
+if (callback) {
+  // do catch up with, maybe multi-thread
+  if (dense) {
+    callback();
+  } else {//sparse
+    for (row : rows_in_block) {callback();}
+  }
+  // finish catch up with, main thread
+  finishCatchUpWith();
+}
 
    
 
    
 
    
@@ -982,23 +982,23 @@ if (callback) {
 
    
 
 
    
-
    
-namespace paddle
    
+
    
+namespace paddle
    
 
    
 
    
 class ParameterOptimizer
    
 
    #include <ParameterOptimizer.h>
    Some member functions are set to const for two reasons:
    
 
      
-
    1. For sparse update thread safe: update(), traverse callback(const this) may be called many times, each time one row, and these function can be called parallelly by multi worker, to speed up large block.
      2. 
-
    2. For predicate functions, needSpecialTraversal(), startCatchUpWith() may be called many times, should be no state change between calls. 
      3. 
+
    3. For sparse update thread safe: update(), traverse callback(const this) may be called many times, each time one row, and these function can be called parallelly by multi worker, to speed up large block.
      4. 
+
    4. For predicate functions, needSpecialTraversal(), startCatchUpWith() may be called many times, should be no state change between calls. 
      5. 
 
    
 
    
-
    Subclassed by paddle::AdaDeltaParameterOptimizer, paddle::AdagradParameterOptimizer, paddle::AdamaxParameterOptimizer, paddle::AdamParameterOptimizer, paddle::AddOptimizer, paddle::AverageOptimizer, paddle::DecayedAdagradParameterOptimizer, paddle::DummyOptimizer, paddle::OptimizerWithGradientClipping, paddle::OptimizerWithRegularizer, paddle::RMSPropParameterOptimizer, paddle::SgdOptimizer, paddle::SparseMomentumParameterOptimizer
    
+
    Subclassed by paddle::AdaDeltaParameterOptimizer, paddle::AdagradParameterOptimizer, paddle::AdamaxParameterOptimizer, paddle::AdamParameterOptimizer, paddle::AddOptimizer, paddle::AverageOptimizer, paddle::DecayedAdagradParameterOptimizer, paddle::DummyOptimizer, paddle::OptimizerWithGradientClipping, paddle::OptimizerWithRegularizer, paddle::RMSPropParameterOptimizer, paddle::SgdOptimizer, paddle::SparseMomentumParameterOptimizer
    
 
    
 
    Public Types
    
 
    
 
    
-typedef std::function<void(const VectorPtr vecs[], const ParameterConfig& config, size_t sparseId)> TraverseCallback
    
+typedef std::function<void(const VectorPtr vecs[], const ParameterConfig &config, size_t sparseId)> TraverseCallback
 
    
 
 
    
@@ -1022,7 +1022,7 @@ if (callback) {
 
    
 
    
 virtual void init(size_t numRows, const ParameterConfig *config)
    
-
    For sparse update, optimizer can maintain numRows of timer(t0). Some sparse optimizer depends on parameter config in functions such as startBatch(). Optimizer can get it here. But notice that, not all callers can pass config here, so the optimizer should check config passed in is not null ptr. 
    
+
    For sparse update, optimizer can maintain numRows of timer(t0). Some sparse optimizer depends on parameter config in functions such as startBatch(). Optimizer can get it here. But notice that, not all callers can pass config here, so the optimizer should check config passed in is not null ptr. 
    
 
    
 
 
    
@@ -1045,22 +1045,22 @@ if (callback) {
 
    
 virtual TraverseCallback needSpecialTraversal(const ParameterConfig &config) const
    
 
    following hooks useful for sparse update, because the traversal in block costs. called by Trainer after update and before finishBatch e.g. Trainer call like this:
    
-
    startBatch();
-if (dense) {
-  update(blockVec);
-} else {//sparse
-  for (row : rows_in_block) {update(rowVec)}
-}
-auto callback = needSpecialTraversal();
-if (callback) {
-  // do traverse, maybe multi-thread
-  if (dense) {
-    callback();
-  } else {//sparse
-    for (row : all_rows_in_block) {callback();}
-  }
-}
-finishBatch();
+
    startBatch();
+if (dense) {
+  update(blockVec);
+} else {//sparse
+  for (row : rows_in_block) {update(rowVec)}
+}
+auto callback = needSpecialTraversal();
+if (callback) {
+  // do traverse, maybe multi-thread
+  if (dense) {
+    callback();
+  } else {//sparse
+    for (row : all_rows_in_block) {callback();}
+  }
+}
+finishBatch();
 
    
 
    
 
    
@@ -1079,25 +1079,25 @@ finishBatch();
 
 
    
 
    
-virtual void update(const VectorPtr vecs[], const ParameterConfig &config, size_t sparseId = -1LU) const = 0
    
-
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
+virtual void update(const VectorPtr vecs[], const ParameterConfig &config, size_t sparseId = -1LU) const = 0
+
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
 
    
 
 
    
 
    
 virtual TraverseCallback startCatchUpWith() const
    
-
    following hooks catch up with current time for sparse update, In the beginning, call startCatchUpWith() and check return. In the end, call finishCatchUpWith() to finish state. callback do the actual works, can call many times for sparse data. e.g. Trainer call like this:
    
-
    auto callback = startCatchUpWith();
-if (callback) {
-  // do catch up with, maybe multi-thread
-  if (dense) {
-    callback();
-  } else {//sparse
-    for (row : rows_in_block) {callback();}
-  }
-  // finish catch up with, main thread
-  finishCatchUpWith();
-}
+
    following hooks catch up with current time for sparse update, In the beginning, call startCatchUpWith() and check return. In the end, call finishCatchUpWith() to finish state. callback do the actual works, can call many times for sparse data. e.g. Trainer call like this:
    
+
    auto callback = startCatchUpWith();
+if (callback) {
+  // do catch up with, maybe multi-thread
+  if (dense) {
+    callback();
+  } else {//sparse
+    for (row : rows_in_block) {callback();}
+  }
+  // finish catch up with, main thread
+  finishCatchUpWith();
+}
 
    
 
    
 
    
@@ -1118,7 +1118,7 @@ if (callback) {
 virtual TraverseCallback apply()
 
    following two hooks used by averager, apply to final parameter value (PARAMETER_VALUE or PARAMETER_APPLY).
    
 
    restore() will restore orginal value if it apply to PARAMETER_VALUE. Caller must ensure it’s catched up with current time before apply.
    
-
    Use returned callback same way as callback returned by ParameterOptimizer::needSpecialTraversal() 
    
+
    Use returned callback same way as callback returned by ParameterOptimizer::needSpecialTraversal() 
    
 
    
 
 
    
@@ -1152,7 +1152,7 @@ if (callback) {
 
    Public Static Functions
    
 
    
 
    
-ParameterOptimizer *create(const OptimizationConfig &optConfig, bool inPserver = false)
    
+ParameterOptimizer *create(const OptimizationConfig &optConfig, bool inPserver = false)
 
    
 
 
    
@@ -1160,7 +1160,7 @@ if (callback) {
 
    Protected Types
    
 
    
 
    
-typedef std::vector<ParameterOptimizer::TraverseCallback> TraverseCallbackVec
    
+typedef std::vector<ParameterOptimizer::TraverseCallback> TraverseCallbackVec
 
    
 
 
    
@@ -1216,13 +1216,13 @@ if (callback) {
 
    
 
 
    
-
    
-namespace paddle
    
+
    
+namespace paddle
    
 
    
 
    
 class OptimizerWithRegularizer
    
-
    Inherits from paddle::ParameterOptimizer
    
-
    Subclassed by paddle::OptimizerWithRegularizerEveryNumBatches, paddle::OptimizerWithRegularizerSparse
    
+
    Inherits from paddle::ParameterOptimizer
    
+
    Subclassed by paddle::OptimizerWithRegularizerEveryNumBatches, paddle::OptimizerWithRegularizerSparse
    
 
    
 
    Public Functions
    
 
    
@@ -1233,7 +1233,7 @@ if (callback) {
 
    
 
    
 virtual void init(size_t numRows, const ParameterConfig *config)
    
-
    For sparse update, optimizer can maintain numRows of timer(t0). Some sparse optimizer depends on parameter config in functions such as startBatch(). Optimizer can get it here. But notice that, not all callers can pass config here, so the optimizer should check config passed in is not null ptr. 
    
+
    For sparse update, optimizer can maintain numRows of timer(t0). Some sparse optimizer depends on parameter config in functions such as startBatch(). Optimizer can get it here. But notice that, not all callers can pass config here, so the optimizer should check config passed in is not null ptr. 
    
 
    
 
 
    
@@ -1262,22 +1262,22 @@ if (callback) {
 
    
 virtual TraverseCallback needSpecialTraversal(const ParameterConfig &config) const
    
 
    following hooks useful for sparse update, because the traversal in block costs. called by Trainer after update and before finishBatch e.g. Trainer call like this:
    
-
    startBatch();
-if (dense) {
-  update(blockVec);
-} else {//sparse
-  for (row : rows_in_block) {update(rowVec)}
-}
-auto callback = needSpecialTraversal();
-if (callback) {
-  // do traverse, maybe multi-thread
-  if (dense) {
-    callback();
-  } else {//sparse
-    for (row : all_rows_in_block) {callback();}
-  }
-}
-finishBatch();
+
    startBatch();
+if (dense) {
+  update(blockVec);
+} else {//sparse
+  for (row : rows_in_block) {update(rowVec)}
+}
+auto callback = needSpecialTraversal();
+if (callback) {
+  // do traverse, maybe multi-thread
+  if (dense) {
+    callback();
+  } else {//sparse
+    for (row : all_rows_in_block) {callback();}
+  }
+}
+finishBatch();
 
    
 
    
 
    
@@ -1291,7 +1291,7 @@ finishBatch();
 
    
 
    
 virtual void update(const VectorPtr vecs[], const ParameterConfig &config, size_t sparseId) const
    
-
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
+
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
 
    
 
 
    
@@ -1327,7 +1327,7 @@ finishBatch();
 
    
 
    
 class OptimizerWithRegularizerEveryNumBatches
    
-
    Inherits from paddle::OptimizerWithRegularizer
    
+
    Inherits from paddle::OptimizerWithRegularizer
    
 
    
 
    Public Functions
    
 
    
@@ -1343,29 +1343,29 @@ finishBatch();
 
    
 
    
 virtual void update(const VectorPtr vecs[], const ParameterConfig &config, size_t sparseId) const
    
-
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
+
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
 
    
 
 
    
 
    
 virtual ParameterOptimizer::TraverseCallback needSpecialTraversal(const ParameterConfig &config) const
    
 
    following hooks useful for sparse update, because the traversal in block costs. called by Trainer after update and before finishBatch e.g. Trainer call like this:
    
-
    startBatch();
-if (dense) {
-  update(blockVec);
-} else {//sparse
-  for (row : rows_in_block) {update(rowVec)}
-}
-auto callback = needSpecialTraversal();
-if (callback) {
-  // do traverse, maybe multi-thread
-  if (dense) {
-    callback();
-  } else {//sparse
-    for (row : all_rows_in_block) {callback();}
-  }
-}
-finishBatch();
+
    startBatch();
+if (dense) {
+  update(blockVec);
+} else {//sparse
+  for (row : rows_in_block) {update(rowVec)}
+}
+auto callback = needSpecialTraversal();
+if (callback) {
+  // do traverse, maybe multi-thread
+  if (dense) {
+    callback();
+  } else {//sparse
+    for (row : all_rows_in_block) {callback();}
+  }
+}
+finishBatch();
 
    
 
    
 
    
@@ -1389,18 +1389,18 @@ finishBatch();
 
    
 
    
 virtual ParameterOptimizer::TraverseCallback startCatchUpWith() const
    
-
    following hooks catch up with current time for sparse update, In the beginning, call startCatchUpWith() and check return. In the end, call finishCatchUpWith() to finish state. callback do the actual works, can call many times for sparse data. e.g. Trainer call like this:
    
-
    auto callback = startCatchUpWith();
-if (callback) {
-  // do catch up with, maybe multi-thread
-  if (dense) {
-    callback();
-  } else {//sparse
-    for (row : rows_in_block) {callback();}
-  }
-  // finish catch up with, main thread
-  finishCatchUpWith();
-}
+
    following hooks catch up with current time for sparse update, In the beginning, call startCatchUpWith() and check return. In the end, call finishCatchUpWith() to finish state. callback do the actual works, can call many times for sparse data. e.g. Trainer call like this:
    
+
    auto callback = startCatchUpWith();
+if (callback) {
+  // do catch up with, maybe multi-thread
+  if (dense) {
+    callback();
+  } else {//sparse
+    for (row : rows_in_block) {callback();}
+  }
+  // finish catch up with, main thread
+  finishCatchUpWith();
+}
 
    
 
    
 
    
@@ -1439,7 +1439,7 @@ if (callback) {
 
    
 
    
 class OptimizerWithRegularizerSparse
    
-
    Inherits from paddle::OptimizerWithRegularizer
    
+
    Inherits from paddle::OptimizerWithRegularizer
    
 
    
 
    Public Functions
    
 
    
@@ -1450,13 +1450,13 @@ if (callback) {
 
    
 
    
 virtual void init(size_t numRows, const ParameterConfig *config)
    
-
    For sparse update, optimizer can maintain numRows of timer(t0). Some sparse optimizer depends on parameter config in functions such as startBatch(). Optimizer can get it here. But notice that, not all callers can pass config here, so the optimizer should check config passed in is not null ptr. 
    
+
    For sparse update, optimizer can maintain numRows of timer(t0). Some sparse optimizer depends on parameter config in functions such as startBatch(). Optimizer can get it here. But notice that, not all callers can pass config here, so the optimizer should check config passed in is not null ptr. 
    
 
    
 
 
    
 
    
 virtual void update(const VectorPtr vecs[], const ParameterConfig &config, size_t sparseId) const
    
-
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
+
    between startBatch() and finishBatch(), update() will be called by the trainer multiple times, each time for updating one Parameter with its gradient in PARAMETER_GRADIENT. sparseId is row id, when sparseId set, update is sparse, each time one row. 
    
 
    
 
 
    
@@ -1467,18 +1467,18 @@ if (callback) {
 
    
 
    
 virtual ParameterOptimizer::TraverseCallback startCatchUpWith() const
    
-
    following hooks catch up with current time for sparse update, In the beginning, call startCatchUpWith() and check return. In the end, call finishCatchUpWith() to finish state. callback do the actual works, can call many times for sparse data. e.g. Trainer call like this:
    
-
    auto callback = startCatchUpWith();
-if (callback) {
-  // do catch up with, maybe multi-thread
-  if (dense) {
-    callback();
-  } else {//sparse
-    for (row : rows_in_block) {callback();}
-  }
-  // finish catch up with, main thread
-  finishCatchUpWith();
-}
+
    following hooks catch up with current time for sparse update, In the beginning, call startCatchUpWith() and check return. In the end, call finishCatchUpWith() to finish state. callback do the actual works, can call many times for sparse data. e.g. Trainer call like this:
    
+
    auto callback = startCatchUpWith();
+if (callback) {
+  // do catch up with, maybe multi-thread
+  if (dense) {
+    callback();
+  } else {//sparse
+    for (row : rows_in_block) {callback();}
+  }
+  // finish catch up with, main thread
+  finishCatchUpWith();
+}
 
    
 
    
 
    
@@ -1532,14 +1532,11 @@ if (callback) {
 
 
         
    
@@ -1561,14 +1558,14 @@ if (callback) {
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    • 
-          
  • Parameter Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    • 
+          
  • Parameter Documents »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/source/parameter/parameter/index.html b/doc/source/parameter/parameter/index.html
index 2e5cc90769..36ceaf5ca5 100644
--- a/doc/source/parameter/parameter/index.html
+++ b/doc/source/parameter/parameter/index.html
@@ -6,7 +6,7 @@
   
     
     
-    Parameter Documents — PaddlePaddle  documentation
+    Parameter Documents — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
      
 
@@ -91,14 +91,11 @@
 
 
         
    
@@ -120,13 +117,13 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/source/parameter/parameter/parameter.html b/doc/source/parameter/parameter/parameter.html
index 24bf4678c9..90dbb049d7 100644
--- a/doc/source/parameter/parameter/parameter.html
+++ b/doc/source/parameter/parameter/parameter.html
@@ -6,7 +6,7 @@
   
     
     
-    Parameter — PaddlePaddle  documentation
+    Parameter — PaddlePaddle  documentation
     
     
     
@@ -45,9 +45,9 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    • 
-          
  • Parameter Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    • 
+          
  • Parameter Documents »
    •  
       
     
      
 
@@ -139,30 +139,43 @@
 
    Regularizer
    
 
    
 
    
-namespace paddle
    
+namespace paddle
 
    
-
    
-class L1L2LrRegularizer
    
-
    Inherits from paddle::Regularizer
    
+
    
+class Regularizer
    
+
    Subclassed by paddle::L1L2LrRegularizer, paddle::L1L2Regularizer, paddle::L1LrRegularizer, paddle::L1Regularizer, paddle::L2LrRegularizer, paddle::L2Regularizer
    
 
    
-
    Private Functions
    
+
    Public Functions
    
 
    
-
    
-virtual void update(const VectorPtr vecs[], const ParameterConfig &paraConfig, real learningRate, int t0, int t) const
    
+
    
+virtual void update(const VectorPtr vecs[], const ParameterConfig &paraConfig, real learningRate, int t0, int t) const = 0
    
+
    
+
+
    
+
    
+virtual ~Regularizer()
    
+
    
+
+
    
+
    
+
    Public Static Functions
    
+
    
+
    
+Regularizer *get(const std::vector<ParameterType> &types, const ParameterConfig &paraConfig)
    
 
    
 
 
    
 
    
 
 
    
-
    
-class L1L2Regularizer
    
-
    Inherits from paddle::Regularizer
    
+
    
+class L1Regularizer
    
+
    Inherits from paddle::Regularizer
    
 
    
 
    Private Functions
    
 
    
-
    
-virtual void update(const VectorPtr vecs[], const ParameterConfig &paraConfig, real learningRate, int t0, int t) const
    
+
    
+virtual void update(const VectorPtr vecs[], const ParameterConfig &paraConfig, real learningRate, int t0, int t) const
    
 
    
 
 
    
@@ -171,7 +184,7 @@
 
    
 
    
 class L1LrRegularizer
    
-
    Inherits from paddle::Regularizer
    
+
    Inherits from paddle::Regularizer
    
 
    
 
    Private Functions
    
 
    
@@ -183,14 +196,14 @@
 
    
 
 
    
-
    
-class L1Regularizer
    
-
    Inherits from paddle::Regularizer
    
+
    
+class L2Regularizer
    
+
    Inherits from paddle::Regularizer
    
 
    
 
    Private Functions
    
 
    
-
    
-virtual void update(const VectorPtr vecs[], const ParameterConfig &paraConfig, real learningRate, int t0, int t) const
    
+
    
+virtual void update(const VectorPtr vecs[], const ParameterConfig &paraConfig, real learningRate, int t0, int t) const
    
 
    
 
 
    
@@ -199,7 +212,7 @@
 
    
 
    
 class L2LrRegularizer
    
-
    Inherits from paddle::Regularizer
    
+
    Inherits from paddle::Regularizer
    
 
    
 
    Private Functions
    
 
    
@@ -211,41 +224,28 @@
 
    
 
 
    
-
    
-class L2Regularizer
    
-
    Inherits from paddle::Regularizer
    
+
    
+class L1L2Regularizer
    
+
    Inherits from paddle::Regularizer
    
 
    
 
    Private Functions
    
 
    
-
    
-virtual void update(const VectorPtr vecs[], const ParameterConfig &paraConfig, real learningRate, int t0, int t) const
    
+
    
+virtual void update(const VectorPtr vecs[], const ParameterConfig &paraConfig, real learningRate, int t0, int t) const
    
 
    
 
 
    
 
    
 
 
    
-
    
-class Regularizer
    
-
    Subclassed by paddle::L1L2LrRegularizer, paddle::L1L2Regularizer, paddle::L1LrRegularizer, paddle::L1Regularizer, paddle::L2LrRegularizer, paddle::L2Regularizer
    
-
    
-
    Public Functions
    
-
    
-
    
-virtual void update(const VectorPtr vecs[], const ParameterConfig &paraConfig, real learningRate, int t0, int t) const = 0
    
-
    
-
-
    
-
    
-virtual ~Regularizer()
    
-
    
-
-
    
+
    
+class L1L2LrRegularizer
    
+
    Inherits from paddle::Regularizer
    
 
    
-
    Public Static Functions
    
+
    Private Functions
    
 
    
-
    
-Regularizer *get(const std::vector<ParameterType> &types, const ParameterConfig &paraConfig)
    
+
    
+virtual void update(const VectorPtr vecs[], const ParameterConfig &paraConfig, real learningRate, int t0, int t) const
    
 
    
 
 
    
@@ -257,8 +257,8 @@
 
    
 
    Parameter
    
 
    
-
    
-namespace paddle
    
+
    
+namespace paddle
    
 
    
 
    Typedefs
    
 
    
@@ -294,7 +294,7 @@
 
 
    
 
    
-void operator=(const Argument &argument)
    
+void operator=(const Argument &argument)
 
    
 
 
    
@@ -374,7 +374,7 @@
 
 
    
 
    
-void subArgFrom(const Argument &input, size_t offset, size_t height, size_t width, bool useGpu, bool trans = false, bool seqFlag = false, size_t seqStart = 0, size_t seqSize = 0)
    
+void subArgFrom(const Argument &input, size_t offset, size_t height, size_t width, bool useGpu, bool trans = false, bool seqFlag = false, size_t seqStart = 0, size_t seqSize = 0)
 
    (value, grad, sequenceStartPositions) of output are subset of input. Note that, output share the same memory of input. 
    
 
    
 
    Parameters
    
@@ -405,22 +405,32 @@
 
 
    
 
    
-int32_t resizeAndCopyFrom(const Argument &src, int32_t startSeq, int32_t copySize, bool useGpu = FLAGS_use_gpu, hl_stream_t stream = HPPL_STREAM_DEFAULT)
    
+int32_t resizeAndCopyFrom(const Argument &src, int32_t startSeq, int32_t copySize, bool useGpu, hl_stream_t stream)
+
    
+
+
    
+
    
+int32_t resizeAndCopyFrom(const Argument &src, int32_t startSeq, int32_t copySize, bool useGpu = FLAGS_use_gpu)
    
 
    
 
 
    
 
    
-void resizeAndCopyFrom(const Argument &src, bool useGpu = FLAGS_use_gpu, hl_stream_t stream = HPPL_STREAM_DEFAULT)
    
+void resizeAndCopyFrom(const Argument &src, bool useGpu, hl_stream_t stream)
+
    
+
+
    
+
    
+void resizeAndCopyFrom(const Argument &src, bool useGpu = FLAGS_use_gpu)
    
 
    
 
 
    
 
    
-void concat(const std::vector<Argument> &args, const std::vector<int> &selectRows, const std::vector<int> &seqStartPos, bool useGpu, hl_stream_t stream, PassType passType)
    
+void concat(const std::vector<Argument> &args, const std::vector<int> &selectRows, const std::vector<int> &seqStartPos, bool useGpu, hl_stream_t stream, PassType passType)
 
    
 
 
    
 
    
-void concat(const std::vector<Argument> &src, bool useGpu = FLAGS_use_gpu, hl_stream_t stream = HPPL_STREAM_DEFAULT, PassType passType = PASS_TEST)
    
+void concat(const std::vector<Argument> &src, bool useGpu = FLAGS_use_gpu, hl_stream_t stream = HPPL_STREAM_DEFAULT, PassType passType = PASS_TEST)
 
    
 
 
    
@@ -435,7 +445,7 @@
 
 
    
 
    
-void degradeSequence(const Argument &input, bool useGpu)
    
+void degradeSequence(const Argument &input, bool useGpu)
 
    
 
 
    
@@ -536,12 +546,12 @@
 
    Public Static Functions
    
 
    
 
    
-static real sumCosts(const std::vector<Argument> &arguments)
    
+static real sumCosts(const std::vector<Argument> &arguments)
 
    
 
 
    
 
    
-void splitByDataId(const std::vector<Argument> &argus, std::vector<std::vector<Argument>> *arguGroups)
    
+void splitByDataId(const std::vector<Argument> &argus, std::vector<std::vector<Argument>> *arguGroups)
 
    
 
 
    
@@ -550,18 +560,18 @@
 
    
 
 
    
-
    
-namespace paddle
    
+
    
+namespace paddle
    
 
    
 
    Typedefs
    
 
    
 
    
-typedef std::function<void(Parameter* param)> UpdateCallback
    
+typedef std::function<void(Parameter *param)> UpdateCallback
 
    
 
 
    
 
    
-typedef std::function<void(int paramId, Parameter* param)> ParamInitCallback
    
+typedef std::function<void(int paramId, Parameter *param)> ParamInitCallback
 
    
 
 
    
@@ -575,6 +585,29 @@
 
    
 
 
    
+
    
+
    
+struct Segment
    
+
    
+
    Public Members
    
+
    
+
    
+int64_t beginDim
    
+
    
+
+
    
+
    
+int64_t endDim
    
+
    
+
+
    
+
    
+int64_t beginPos
    
+
    
+
+
    
+
    
+
 
    
 
    
 class Parameter
    
@@ -639,8 +672,8 @@
 
 
    
 
    
-typedef std::function<void(const VectorPtr vecs[])> ExecFunc
    
-
    exec a func in single/multi thread. vecs is bufs_ of Parameter, as input of ExecFunc. 
    
+typedef std::function<void(const VectorPtr vecs[])> ExecFunc
+
    exec a func in single/multi thread. vecs is bufs_ of Parameter, as input of ExecFunc. 
    
 
    
 
 
    
@@ -752,7 +785,7 @@
 
 
    
 
    
-ParameterConfig &getConfig()
    
+ParameterConfig &getConfig()
 
    
 
 
    
@@ -832,7 +865,7 @@
 
    Update bufs_[PARAMETER_VALUE] using sparse row grad matrix.
    
 
    
 
    See
    
-
    SparseRowCpuMatrix::sgdUpdate for more information. 
    
+
    SparseRowCpuMatrix::sgdUpdate for more information. 
    
 
    
 
    
 
    
@@ -908,7 +941,7 @@
 
    
 
    
 void updateHook()
    
-
    Parameter Update Hook. 
    
+
    Parameter Update Hook. 
    
 
    The parameter’s update hook before ParameterUpdater::updateImpl It could modify gradient/momentum/etc here. Such as drop some gradient, etc. 
    
 
    
 
@@ -1032,13 +1065,13 @@
 
    
 VectorPtr bufs_[NUM_PARAMETER_TYPES]
    
 
    bufs_ stores parameter value and gradient. 
    
-
    Layer should use bufs_[PARAMETER_VALUE] to form weight matrix for calculation and stores gradient to bufs_[PARAMETER_GRADIENT]. 
    
+
    Layer should use bufs_[PARAMETER_VALUE] to form weight matrix for calculation and stores gradient to bufs_[PARAMETER_GRADIENT]. 
    
 
    
 
 
    
 
    
 MatrixPtr mats_[NUM_PARAMETER_TYPES]
    
-
    Weight matrix for bufs_. 
    
+
    Weight matrix for bufs_. 
    
 
    It’s helpfull when parameter shared by multi-layers. Caller should check, if mats exist, do not create it again. 
    
 
    
 
@@ -1113,34 +1146,11 @@
 
 
    
 
-
    
-
    
-struct Segment
    
-
    
-
    Public Members
    
-
    
-
    
-int64_t beginDim
    
-
    
-
-
    
-
    
-int64_t endDim
    
-
    
-
-
    
-
    
-int64_t beginPos
    
-
    
-
-
    
-
    
-
 
    
 
 
    
-
    
-namespace paddle
    
+
    
+namespace paddle
    
 
    
 
    Typedefs
    
 
    
@@ -1149,8 +1159,8 @@
 
    
 
 
    
-
    
-typedef void(ParallelParameter::* paddle::UpdateFunction) (real learnRate)
    
+
    
+typedef void (ParallelParameter::*UpdateFunction)(real learnRate)
    
 
    
 
 
    
@@ -1223,75 +1233,10 @@
 
    
 
 
    
-
    
-
    
-class AsyncParameter
    
-
    Inherits from paddle::ParallelParameter
    
-
    
-
    Public Functions
    
-
    
-
    
-AsyncParameter(TrainerRole role, int asyncCount, ParameterPtr localParam)
    
-
    
-
-
    
-
    
-void clearCounter()
    
-
    
-
-
    
-
    
-VectorPtr getAccum()
    
-
    
-
-
    
-
    
-virtual void synchronizeParamter()
    
-
    
-
-
    
-
    
-virtual void slaveUpdate(real learnRate)
    
-
    When asynchronous training, update strategy including slave and master.
    
-
    slave: If in range asyncCount, adopting self-update method. If beyond asyncCount, waiting for master to update. 
    
-
    
-
-
    
-
    
-bool masterUpdate(ParallelParameterPtr slaveParam, const UpdateCallback &callback)
    
-
    When asynchronous training, update strategy including slave and master.
    
-
    master: it only polls slaves, do not training data. If slave’s gradient is ready, fetch it. Update master’s parameter, then copy it into corresponding slave. 
    
-
    
-
-
    
-
    
-
    Private Members
    
-
    
-
    
-VectorPtr gradientAccum_
    
-
    When asynchronous training, every aysnc trainer needs to accumulate a number of batch gradient.
    
-
    gradientAccum_ is used to save the sum of gradients. 
    
-
    
-
-
    
-
    
-int asyncCount_
    
-
    Asynchronous count. 
    
-
    
-
-
    
-
    
-int accumCounter_
    
-
    Accumulate counter of current gradients. 
    
-
    
-
-
    
-
    
-
 
    
 
    
 class ParallelParameter
    
-
    Subclassed by paddle::AsyncParameter, paddle::SyncParameter
    
+
    Subclassed by paddle::AsyncParameter, paddle::SyncParameter
    
 
    
 
    Public Functions
    
 
    
@@ -1401,7 +1346,7 @@
 class SyncParameter
 
    #include <ParallelParameter.h>
    this class is designed for multi-threading training.
    
 
    “Synchronous” means multiple GPUs calculate 1/4 mini-Batch, but will get only one gradient 
    
-
    Inherits from paddle::ParallelParameter
    
+
    Inherits from paddle::ParallelParameter
    
 
    
 
    Public Functions
    
 
    
@@ -1501,6 +1446,71 @@
 
    
 
    
 
+
    
+
    
+class AsyncParameter
    
+
    Inherits from paddle::ParallelParameter
    
+
    
+
    Public Functions
    
+
    
+
    
+AsyncParameter(TrainerRole role, int asyncCount, ParameterPtr localParam)
    
+
    
+
+
    
+
    
+void clearCounter()
    
+
    
+
+
    
+
    
+VectorPtr getAccum()
    
+
    
+
+
    
+
    
+virtual void synchronizeParamter()
    
+
    
+
+
    
+
    
+virtual void slaveUpdate(real learnRate)
    
+
    When asynchronous training, update strategy including slave and master.
    
+
    slave: If in range asyncCount, adopting self-update method. If beyond asyncCount, waiting for master to update. 
    
+
    
+
+
    
+
    
+bool masterUpdate(ParallelParameterPtr slaveParam, const UpdateCallback &callback)
    
+
    When asynchronous training, update strategy including slave and master.
    
+
    master: it only polls slaves, do not training data. If slave’s gradient is ready, fetch it. Update master’s parameter, then copy it into corresponding slave. 
    
+
    
+
+
    
+
    
+
    Private Members
    
+
    
+
    
+VectorPtr gradientAccum_
    
+
    When asynchronous training, every aysnc trainer needs to accumulate a number of batch gradient.
    
+
    gradientAccum_ is used to save the sum of gradients. 
    
+
    
+
+
    
+
    
+int asyncCount_
    
+
    Asynchronous count. 
    
+
    
+
+
    
+
    
+int accumCounter_
    
+
    Accumulate counter of current gradients. 
    
+
    
+
+
    
+
    
+
 
    
 
 
    
@@ -1538,14 +1548,11 @@
 
 
         
    
@@ -1567,14 +1574,14 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    • 
-          
  • Parameter Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    • 
+          
  • Parameter Documents »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/source/parameter/update/index.html b/doc/source/parameter/update/index.html
index 272a848273..1ebba46a3c 100644
--- a/doc/source/parameter/update/index.html
+++ b/doc/source/parameter/update/index.html
@@ -6,7 +6,7 @@
   
     
     
-    Parameter Documents — PaddlePaddle  documentation
+    Parameter Documents — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
      
 
@@ -86,14 +86,11 @@
 
 
         
    
@@ -115,13 +112,13 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/source/parameter/update/update.html b/doc/source/parameter/update/update.html
index cc35ecd70c..b7c8fa004f 100644
--- a/doc/source/parameter/update/update.html
+++ b/doc/source/parameter/update/update.html
@@ -6,7 +6,7 @@
   
     
     
-    Update — PaddlePaddle  documentation
+    Update — PaddlePaddle  documentation
     
     
     
@@ -45,9 +45,9 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    • 
-          
  • Parameter Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    • 
+          
  • Parameter Documents »
    •  
       
     
      
 
@@ -64,7 +64,7 @@
 
    
 
    
 class ParameterUpdater
    
-
    Subclassed by paddle::ParameterUpdaterComposite, paddle::RemoteParameterUpdater, paddle::SgdLocalUpdater, paddle::SgdThreadUpdater, paddle::SparseRemoteParameterUpdater
    
+
    Subclassed by paddle::ParameterUpdaterComposite, paddle::RemoteParameterUpdater, paddle::SgdLocalUpdater, paddle::SgdThreadUpdater, paddle::SparseRemoteParameterUpdater
    
 
    
 
    Public Functions
    
 
    
@@ -189,8 +189,8 @@
 
    
 
    
 class ParameterUpdaterComposite
    
-
    Inherits from paddle::ParameterUpdater
    
-
    Subclassed by paddle::SparseRemoteParameterUpdaterComposite
    
+
    Inherits from paddle::ParameterUpdater
    
+
    Subclassed by paddle::SparseRemoteParameterUpdaterComposite
    
 
    
 
    Public Functions
    
 
    
@@ -296,7 +296,7 @@
 
 
    
 
    
-namespace paddle
    
+namespace paddle
 
    
 
    Variables
    
 
    
@@ -309,7 +309,7 @@
 
    
 class IParameterUpdaterHook
    
 
    #include <ParameterUpdaterHook.h>
    The parameter updater hook interface.
    
-
    The Parameter Updater hooks is a group of methods invoke before ParameterUpdater::updateImpl. It can modify gradient/momentum/etc before parameter optimization. 
    
+
    The Parameter Updater hooks is a group of methods invoke before ParameterUpdater::updateImpl. It can modify gradient/momentum/etc before parameter optimization. 
    
 
    Subclassed by paddle::StaticPruningHook
    
 
    
 
    Public Functions
    
@@ -335,7 +335,7 @@
 
    Public Static Functions
    
 
    
 
    
-std::shared_ptr<IParameterUpdaterHook> create(const ParameterConfig &paramConfig, int idx)
    
+std::shared_ptr<IParameterUpdaterHook> create(const ParameterConfig &paramConfig, int idx)
 
    Create A ParameterUpdaterHook.
    
 
    The same parameter shared the same hooks. So it returns shared_ptr.
    
 
    
@@ -389,14 +389,11 @@
 
 
         
    
@@ -418,14 +415,14 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    • 
-          
  • Parameter Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    • 
+          
  • Parameter Documents »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/source/pserver/client/client.html b/doc/source/pserver/client/client.html
index d0d76bdb7a..1af097951b 100644
--- a/doc/source/pserver/client/client.html
+++ b/doc/source/pserver/client/client.html
@@ -6,7 +6,7 @@
   
     
     
-    Client — PaddlePaddle  documentation
+    Client — PaddlePaddle  documentation
     
     
     
@@ -45,9 +45,9 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    • 
-          
  • Client Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    • 
+          
  • Client Documents »
    •  
       
     
      
 
@@ -62,7 +62,7 @@
 
    
 class paddle::BaseClient
    
 
    it manages all connections to pservers. it exists two modes to manage connections to all pservers. Firstly, one connection owns two threads that separately manage to send and receive data. Secondly, each thread uses one connection for all activation in it. the first solution arms with sendThreads_/recvThreads_ and sendJobQueue_/ recvJobQueue_. the second solution use some shared thread pool to manage connections. In addition to pserver, metric learning also uses network to exchange features within multi-machines, so this class just abstracts some basic threads and queue buffer creation for them 
    
-
    Subclassed by paddle::ParameterClient2
    
+
    Subclassed by paddle::ParameterClient2
    
 
    
 
    Public Types
    
 
    
@@ -234,7 +234,7 @@
 
    use multithread to separately receive data 
    
 
    
 
    Note
    
-
    almost same as send() 
    
+
    almost same as send() 
    
 
    
 
    
 
    
@@ -358,8 +358,8 @@
 
    
 
    
 class paddle::ParameterClient2
    
-
    The client interface for parameter server. ParameterClient2 supports 2 modes for managing connections to parameter servers, in the 1st mode one connection is shared by 2 threads that are separately responsible for sending and recieving activities, in the 2nd mode one connection is owned by only one thread, and all the sending and recieving activities run in that single thread. TODO(yanfei): Additional core idea to further optimizate pserver performance is to do sync-sgd based parameter level instead of pserver level. full-parallelization based parameter level for sync-sgd also can sense forwardbackward computation layer-by-layer for more deeper layer model. Firstly, pserver can do full-parallelization on all computation based parameter level instead of waiting for all gradients are finished and start to send back parameters value immediately if parameter is ready instead of waiting for all parameters value are ready Secondly, parameter client can write back parameters to GPU instead of waiting until all parameters are received to CPU host end. 
    
-
    Inherits from paddle::BaseClient
    
+
    The client interface for parameter server. ParameterClient2 supports 2 modes for managing connections to parameter servers, in the 1st mode one connection is shared by 2 threads that are separately responsible for sending and recieving activities, in the 2nd mode one connection is owned by only one thread, and all the sending and recieving activities run in that single thread. TODO(yanfei): Additional core idea to further optimizate pserver performance is to do sync-sgd based parameter level instead of pserver level. full-parallelization based parameter level for sync-sgd also can sense forwardbackward computation layer-by-layer for more deeper layer model. Firstly, pserver can do full-parallelization on all computation based parameter level instead of waiting for all gradients are finished and start to send back parameters value immediately if parameter is ready instead of waiting for all parameters value are ready Secondly, parameter client can write back parameters to GPU instead of waiting until all parameters are received to CPU host end. 
    
+
    Inherits from paddle::BaseClient
    
 
    
 
    Public Functions
    
 
    
@@ -430,7 +430,7 @@
 
    
 
    
 void sendParameter(ParameterUpdateMode updateMode, ParameterType parameterType, const std::vector<ParameterSegments> &segments, int64_t numSamples, real cost, bool sendBackParameter, BatchStatus batchStatus)
    
-
    Sends the segments in parameter to parameter servers. Each sendParameter() must be paired with a recvParameter() in the future. Only parameterType will be sent. 
    
+
    Sends the segments in parameter to parameter servers. Each sendParameter() must be paired with a recvParameter() in the future. Only parameterType will be sent. 
    
 
    
 
    Note
    
 
    This function is non-blocking. This means that parameter should not change between this call and recvParameter() 
    
@@ -771,7 +771,7 @@
 
    
 
    
 void sendParallel(int tid, size_t numThreads, ParameterType recvParameterType)
    
-
    management function for parallelizing send/recv all connections to all pservers. it is called under one SyncThreadPool. it supports to use N thread to control M connections. the receiving actions can be started until all sending action to all connections owned by current thread are finished. Different connections controlled by different threads can transfer data asynchronously. 
    
+
    management function for parallelizing send/recv all connections to all pservers. it is called under one SyncThreadPool. it supports to use N thread to control M connections. the receiving actions can be started until all sending action to all connections owned by current thread are finished. Different connections controlled by different threads can transfer data asynchronously. 
    
 
    
 
 
    
@@ -830,14 +830,11 @@
 
 
         
    
@@ -859,14 +856,14 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    • 
-          
  • Client Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    • 
+          
  • Client Documents »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/source/pserver/client/index.html b/doc/source/pserver/client/index.html
index c362f2afc7..45919f8fe1 100644
--- a/doc/source/pserver/client/index.html
+++ b/doc/source/pserver/client/index.html
@@ -6,7 +6,7 @@
   
     
     
-    Client Documents — PaddlePaddle  documentation
+    Client Documents — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
      
 
@@ -86,14 +86,11 @@
 
 
         
    
@@ -115,13 +112,13 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/source/pserver/network/index.html b/doc/source/pserver/network/index.html
index b728c7fc2e..973cb1d2f8 100644
--- a/doc/source/pserver/network/index.html
+++ b/doc/source/pserver/network/index.html
@@ -6,7 +6,7 @@
   
     
     
-    Network Documents — PaddlePaddle  documentation
+    Network Documents — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
      
 
@@ -93,14 +93,11 @@
 
 
         
    
@@ -122,13 +119,13 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/source/pserver/network/network.html b/doc/source/pserver/network/network.html
index a5fef05f67..b15359e40c 100644
--- a/doc/source/pserver/network/network.html
+++ b/doc/source/pserver/network/network.html
@@ -6,7 +6,7 @@
   
     
     
-    Network — PaddlePaddle  documentation
+    Network — PaddlePaddle  documentation
     
     
     
@@ -45,9 +45,9 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    • 
-          
  • Network Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    • 
+          
  • Network Documents »
    •  
       
     
      
 
@@ -69,13 +69,13 @@
 
    each parameter server inherits from one socket server, each server contains serveral woker threads which are to parallelize the processing of computation, but share some common datas stored in child class of socketserver. 
    
 
    
 
    
-
    Inherits from paddle::Thread
    
-
    Subclassed by paddle::ProtoServer
    
+
    Inherits from paddle::Thread
    
+
    Subclassed by paddle::ProtoServer
    
 
    
 
    Public Types
    
 
    
 
    
-typedef std::function<void(const std::vector<iovec>& outputIovs)> ResponseCallback
    
+typedef std::function<void(const std::vector<iovec> &outputIovs)> ResponseCallback
 
    
 
 
    
@@ -84,7 +84,7 @@
 
    
 
    
 SocketServer(const std::string &addr, int port, int rdmaCpu)
    
-
    class constructor for SocketServer 
    
+
    class constructor for SocketServer 
    
 
    
 
    Note
    
 
    start one socket server which hosts parameter server process. rdmaCpu is passed to rdma deamon for better performance, and start tcp socket instead of rdma socket if rdmaCpu is equal to -1. Each trainer process starts one connection to one socket server, and use ports_num to build more connections to harness fat communication channel if necessary. each connection is controlled by single thread with blocking read and write. 
    
@@ -237,7 +237,7 @@
 
    all parameter processing will run in the context of this worker 
    
 
    
 
    
-
    Inherits from paddle::Thread
    
+
    Inherits from paddle::Thread
    
 
    
 
    Public Functions
    
 
    
@@ -298,8 +298,8 @@
 
    
 
    Public Functions
    
 
    
-
    
-paddle::SocketClient::SocketClient(const std::string & serverAddr, int serverPort, enum ChannelType channelType)
    
+
    
+SocketClient(const std::string &serverAddr, int serverPort, enum ChannelType channelType)
    
 
    class constructor 
    
 
    
 
    Note
    
@@ -532,7 +532,7 @@ rdma::readv and rdma::writev can take advantage of RDMA blocking offload transfe
 
    
 
    
 class paddle::MsgReader
    
-
    reading a set of blocks of data from SocketChannel. 
    
+
    reading a set of blocks of data from SocketChannel. 
    
 
    
 
    Public Functions
    
 
    
@@ -638,14 +638,11 @@ rdma::readv and rdma::writev can take advantage of RDMA blocking offload transfe
 
 
         
    
@@ -667,14 +664,14 @@ rdma::readv and rdma::writev can take advantage of RDMA blocking offload transfe
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    • 
-          
  • Network Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    • 
+          
  • Network Documents »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/source/pserver/server/index.html b/doc/source/pserver/server/index.html
index da92f8f591..d577a0f59d 100644
--- a/doc/source/pserver/server/index.html
+++ b/doc/source/pserver/server/index.html
@@ -6,7 +6,7 @@
   
     
     
-    Server Documents — PaddlePaddle  documentation
+    Server Documents — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
      
 
@@ -86,14 +86,11 @@
 
 
         
    
@@ -115,13 +112,13 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/source/pserver/server/server.html b/doc/source/pserver/server/server.html
index e145284457..2c00b77bcc 100644
--- a/doc/source/pserver/server/server.html
+++ b/doc/source/pserver/server/server.html
@@ -6,7 +6,7 @@
   
     
     
-    Server — PaddlePaddle  documentation
+    Server — PaddlePaddle  documentation
     
     
     
@@ -45,9 +45,9 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    • 
-          
  • Server Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    • 
+          
  • Server Documents »
    •  
       
     
      
 
@@ -61,19 +61,19 @@
 
    
 
    
 class paddle::ProtoServer
    
-
    It implements the rpc framework, which launchs one thread for each connection. Here define one parameter server as single TCP server binding on single port. All connections share single tcp ProtoServer object, each connection handles all requests from specified trainer within single worker thread. to accelerate bandwidth efficiency and harness multicore for pserver optimization to reduce pserver latency, you could launch more port for single NIC hardward with port=N(N>1) for small cluster job. 
    
-
    Inherits from paddle::SocketServer
    
-
    Subclassed by paddle::ParameterServer2
    
+
    It implements the rpc framework, which launchs one thread for each connection. Here define one parameter server as single TCP server binding on single port. All connections share single tcp ProtoServer object, each connection handles all requests from specified trainer within single worker thread. to accelerate bandwidth efficiency and harness multicore for pserver optimization to reduce pserver latency, you could launch more port for single NIC hardward with port=N(N>1) for small cluster job. 
    
+
    Inherits from paddle::SocketServer
    
+
    Subclassed by paddle::ParameterServer2
    
 
    
 
    Public Types
    
 
    
 
    
-typedef std::function<void(const google::protobuf::MessageLite& protoOut, const std::vector<iovec>& outputIovs)> ProtoResponseCallbackEx
    
+typedef std::function<void(const google::protobuf::MessageLite &protoOut, const std::vector<iovec> &outputIovs)> ProtoResponseCallbackEx
 
    
 
 
    
 
    
-typedef std::function<void(const google::protobuf::MessageLite& protoOut)> ProtoResponseCallback
    
+typedef std::function<void(const google::protobuf::MessageLite &protoOut)> ProtoResponseCallback
 
    
 
 
    
@@ -88,8 +88,8 @@
 
    
 
    
 template <class ProtoIn>
    
-
    
-void registerServiceFunction(const std::string &funcName, std::function<void(const ProtoIn &request,                                                                                                                                                                                                                                                                                                                                                                   ProtoResponseCallback callback)> func)
    
+
    
+void registerServiceFunction(const std::string &funcName, std::function<void(const ProtoIn &request, ProtoResponseCallback callback)> func)
    
 
    Register a service function for this server void(const ProtoIn& request, ProtoResponseCallback callback) The service function process the request and call the callback after it finishes the request.
    
 
    Use macro REGISTER_SERVICE_FUNCTION as a helper to simplify the use. 
    
 
    
@@ -97,25 +97,18 @@
 
    
 
    
 template <class ProtoIn>
    
-
    
-void registerServiceFunctionEx(const std::string &funcName, std::function<void(const ProtoIn &, std::unique_ptr<  MsgReader  > msgReader,                                                                                                                                                                                                                                                                                                                                                                   ProtoResponseCallbackEx callback)> func)
    
+
    
+void registerServiceFunctionEx(const std::string &funcName, std::function<void(const ProtoIn&, std::unique_ptr<MsgReader> msgReader, ProtoResponseCallbackEx callback)> func)
    
 
    Register a service function for this server The signature of the service function is void(const ProtoIn&, std::unique_ptr<MsgReader> msgReader, ProtoResponseCallbackEx callback) The service function process the request and call the callback after it finishes the request. The extended service function can take extra input blocks from the communication channel by reading msgReader. It can also send extra blocks to the communication channel by providing outputIovs as the argument for the callback function.
    
-
    Use macro REGISTER_SERVICE_FUNCTION_EX as a helper to simplify the use. 
    
+
    Use macro REGISTER_SERVICE_FUNCTION_EX as a helper to simplify the use.
    
+
    create wrapper function for parameter server high level function and register the wrapper function into function mapping. 
    
 
    
 
 
    
 
    
 template <class ProtoIn>
    
-
    
-void registerServiceFunctionEx(const std::string &funcName, std::function<void(const ProtoIn &, std::unique_ptr<  MsgReader  > msgReader,                                                                                                                                                                                                                                                                                                                                   ProtoResponseCallbackEx callback)> func)
    
-
    create wrapper function for parameter server high level function and register the wrapper function into function mapping. 
    
-
    
-
-
    
 
    
-template <class ProtoIn>
    
-
    
-void registerServiceFunction(const std::string &funcName, std::function<void(const ProtoIn &, ProtoResponseCallback callback)> func)
    
+void registerServiceFunction(const std::string &funcName, std::function<void(const ProtoIn&, ProtoResponseCallback callback)> func)
 
    
 
 
    
@@ -123,7 +116,7 @@
 
    Protected Types
    
 
    
 
    
-typedef std::function<void(std::unique_ptr<MsgReader> msgReader, ResponseCallback callback)> ServiceFunction
    
+typedef std::function<void(std::unique_ptr<MsgReader> msgReader, ResponseCallback callback)> ServiceFunction
 
    
 
 
    
@@ -188,12 +181,12 @@
 class paddle::ParameterServer2
 
    Client interface for the parameter server
    
 
    it implements several rpc API for remote parameter client usage. for sync-sgd, client needs one controller thread to build connections to all pservers, these controller connections do barriers synchronization with these connections used for transfering data. each data connection uses block based fine grained synchronization to gain better scalability. Merging gradients from different trainers are concurrently executed with block units, so that some network overhead will be hidden in merging gradient. for async-sgd, the difference is that pserver will do optimization immediately if the gradients are ready, so that pserver needs to prepare separate buffer to store value for sending back to trainer to prevent from being polluted. 
    
-
    Inherits from paddle::ProtoServer
    
+
    Inherits from paddle::ProtoServer
    
 
    
 
    Public Types
    
 
    
-
    
-typedef void(ParameterServer2::* paddle::ParameterServer2::OperatorFunction) (const Operation &operation, OperationResult *result)
    
+
    
+typedef void (ParameterServer2::*OperatorFunction)(const Operation &operation, OperationResult *result)
    
 
    
 
 
    
@@ -574,7 +567,7 @@
 
 
    
 
    
-ParameterServer2::OperatorFunction opFuncs
    
+ParameterServer2::OperatorFunction opFuncs
 
    doOperation will call following operations indirectly e.g. for sync-sgd control, the controller in remote updater will send op_SGD command to pserver, then send sendParameter request to pserver immediately. the two function at pserver end will do cooperation to achieve the sync-sgd gradient merge and optimization. the most following operations are specified for owlqn, all operations are under the context of doOperation function 
    
 
    
 
@@ -599,7 +592,7 @@
 
 
    
 
    
-typedef std::function<void(int64_t blockId, const VectorPtr vecs[])> ExecFunc
    
+typedef std::function<void(int64_t blockId, const VectorPtr vecs[])> ExecFunc
 
    framework routine for block parallelization e.g. for optimization on all blocks at pserver end, this routine can facilitize the parallelize of do optimization on all blocks with multithreads. 
    
 
    
 
@@ -630,7 +623,7 @@
 
 
    
 
    
-void readAllBlocks(MsgReader *msgReader, std::vector<ParameterServer2::Buffer> *buffers)
    
+void readAllBlocks(MsgReader *msgReader, std::vector<ParameterServer2::Buffer> *buffers)
 
    read all data from connection and store it in static pre-allocated buffer 
    
 
    
 
@@ -820,7 +813,7 @@
 
    
 
    
 ThreadBarrier gradientReadyBarrier_
    
-
    for synchronized training, check details in addGradient() and doOperation() 
    
+
    for synchronized training, check details in addGradient() and doOperation() 
    
 
    
 
 
    
@@ -870,7 +863,7 @@
 
  • initial asyncUpdaterSteps = 0, asyncTrainerSteps_[N] = 0. syncUpdaterSteps means the version of parameter value.
    • 
 
  • when pull arrives, record asyncUpdateSteps_ into syncTrainerSteps_[trainer_id]
    • 
 
  • when push arrives, compare asyncUpdateSteps_ with syncTrainerSteps_[trainer_id] if delta > threshold, discard current gradient, else commit gradient.
    • 
-
  • reset asyncUpdaterSteps_ and asyncTrainerSteps_[N] when pass finished Note: it can not discard all lag-gradient strictly in some special condition. part of gradients could be discarded if ConcurrentRemoteParameterUpdater is sed. this algorithm is implemented in asynSGD() 
    • 
+
  • reset asyncUpdaterSteps_ and asyncTrainerSteps_[N] when pass finished Note: it can not discard all lag-gradient strictly in some special condition. part of gradients could be discarded if ConcurrentRemoteParameterUpdater is sed. this algorithm is implemented in asynSGD() 
    • 
 
 
    
 
    
@@ -1006,7 +999,7 @@
 template <typename T, size_t AlignBytes>
 
    
 class ReadWriteBuffer
    
-
    The ReadWriteBuffer is based on std::vector, but aligned for avx/sse compute. And add some helper method to allocate memory aligned blocks.
    
+
    The ReadWriteBuffer is based on std::vector, but aligned for avx/sse compute. And add some helper method to allocate memory aligned blocks.
    
 
    
 
    Parameters
    
 
      
@@ -1024,7 +1017,7 @@
 
      
 
      
 void resizeWithAlignHints(size_t size, size_t alignBlockCount = 1)
      
-
      Resize Buffer, with block count that will be allocated. Each block will be memory aligned in AlignBytes. 
      
+
      Resize Buffer, with block count that will be allocated. Each block will be memory aligned in AlignBytes. 
      
 
      
 
      Parameters
      
 
        
@@ -1113,14 +1106,11 @@
 
 
         
    
@@ -1142,14 +1132,14 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    • 
-          
  • Server Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    • 
+          
  • Server Documents »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/source/trainer/trainer.html b/doc/source/trainer/trainer.html
index 664d2d421f..6877f82db0 100644
--- a/doc/source/trainer/trainer.html
+++ b/doc/source/trainer/trainer.html
@@ -6,7 +6,7 @@
   
     
     
-    Trainer — PaddlePaddle  documentation
+    Trainer — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
      
 
@@ -62,7 +62,7 @@
 
    
 
    
 class paddle::TrainerStats
    
-
    TrainerStats object will statistics sample processed and total cost. 
    
+
    TrainerStats object will statistics sample processed and total cost. 
    
 
    There are two stats in it, the ‘AvgCost’ and ‘CurrentAvgCost’. ‘AvgCost’ means cost through one pass(all mini-batches). ‘CurrentAvgCost’ means cost through one mini-batch. 
    
 
    
 
    Public Functions
    
@@ -132,11 +132,11 @@
 
 
    
 
    
-TrainerStats &operator+=(const std::pair<int64_t, real> &p)
    
+TrainerStats &operator+=(const std::pair<int64_t, real> &p)
 
    same function as addCost. But it is simple to invoke. For example: 
    
-
    TrainerStats stat;
-cost = neuralNetwork.forward(batchSize);
-stat += {batchSize, cost};
+
    TrainerStats stat;
+cost = neuralNetwork.forward(batchSize);
+stat += {batchSize, cost};
 
    
 
    
 
    
@@ -156,7 +156,7 @@ stat += {batchSize, cost};
 
    
 
    
 TrainerStats()
    
-
    TrainerStats Constructor. 
    
+
    TrainerStats Constructor. 
    
 
    reset stat when constructed. 
    
 
    
 
@@ -207,8 +207,8 @@ stat += {batchSize, cost};
 
    Normal remote parameter updater for dense parameters.
    
 
    It first packs all parameters for all pservers using ParameterClient module, then wait for merged parameters data from all pservers. The synchronization pattern specified by sync-sgd or async-sgd is achieved by all pservers with the help of the controller within this remote parameter updater. This module indeedly bridges the gradient machines and parameter servers. It helps to transfer the parameters from acceleration device to cpu end for network. It contains additional parameters copy buffers for acceleration devices at cpu end, such as gpu, otherwise it will directly use original parameters data to update pservers.
    
 
    This remote parameter updater does not use pipeline mechanism to hide copy latency from gpu to cpu buffer. In addition the overlapped between backward and communication is not supported. 
    
-
    Inherits from paddle::ParameterUpdater
    
-
    Subclassed by paddle::ConcurrentRemoteParameterUpdater
    
+
    Inherits from paddle::ParameterUpdater
    
+
    Subclassed by paddle::ConcurrentRemoteParameterUpdater
    
 
    
 
    Public Functions
    
 
    
@@ -414,8 +414,8 @@ stat += {batchSize, cost};
 
    
 class paddle::ConcurrentRemoteParameterUpdater
    
 
    This updater add additional optimization for overlapping synchronization from pservers with backward computation.
    
-
    Parameter can be sent to pservers when related backward stage is finished. This concurrent udpater does data copy from acceleration device to host memory aynchronously. In addition internal parameter client reads data in host memory and send them to all pservers in next stage. So this class help to pipeline device-to-host copy and host-to-network to hide network latency in backward stage. It contains separate send and recv thread for pipeline usage. 
    
-
    Inherits from paddle::RemoteParameterUpdater
    
+
    Parameter can be sent to pservers when related backward stage is finished. This concurrent udpater does data copy from acceleration device to host memory aynchronously. In addition internal parameter client reads data in host memory and send them to all pservers in next stage. So this class help to pipeline device-to-host copy and host-to-network to hide network latency in backward stage. It contains separate send and recv thread for pipeline usage. 
    
+
    Inherits from paddle::RemoteParameterUpdater
    
 
    
 
    Public Functions
    
 
    
@@ -512,7 +512,7 @@ stat += {batchSize, cost};
 
    It allows part of parameter to be exchanged with all pservers. If sparse input assigned, part gradients of first hidden layer could remained zero which can not need to be exchanged within all pservers. This is the key optimization point for this updater
    
 
    For updating sparse parameters, all latest parameters are stored in pservers instead of keeping full copy at train end, so need to prefetch parameters weight value which can be changed in next-batch before doing next forwardbackward. Also, with above fact that the parameters can be stored in pserver instead of trainer, we can fetch specified parmeters if necessary, and can support huge parameters which is larger enough than the RAM size in single node.
    
 
    Internally, this updater will direct internal parameter client to encapsulate sparse specified message for all pservers. 
    
-
    Inherits from paddle::ParameterUpdater
    
+
    Inherits from paddle::ParameterUpdater
    
 
    
 
    Public Functions
    
 
    
@@ -674,26 +674,26 @@ stat += {batchSize, cost};
 
    Class for supporting normal updater and sparse updater
    
 
    Not all parts of one model are sparse, so it exists dense updater for normal layers while sparse updater is for sparse layers.
    
 
    it directly call internal dense and sparse udpater individually. 
    
-
    Inherits from paddle::ParameterUpdaterComposite
    
+
    Inherits from paddle::ParameterUpdaterComposite
    
 
    
 
    Public Types
    
 
    
-
    
-enum [anonymous]
    
+
    
+enum [anonymous]
    
 
    Values:
    
 
    
 
    
-UPDATER_SPARSE_REMOTE = 0
    
+UPDATER_SPARSE_REMOTE = 0
 
    
 
 
    
 
    
-UPDATER_NORMAL = 1
    
+UPDATER_NORMAL = 1
 
    
 
 
    
 
    
-NUMBER_UPDATERS = 2
    
+NUMBER_UPDATERS = 2
 
    
 
 
    
@@ -758,14 +758,11 @@ stat += {batchSize, cost};
 
 
         
    
@@ -787,13 +784,13 @@ stat += {batchSize, cost};
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/source/utils/customStackTrace.html b/doc/source/utils/customStackTrace.html
index 879f107330..3877eab029 100644
--- a/doc/source/utils/customStackTrace.html
+++ b/doc/source/utils/customStackTrace.html
@@ -6,7 +6,7 @@
   
     
     
-    CustomStackTrace — PaddlePaddle  documentation
+    CustomStackTrace — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
      
 
@@ -65,63 +65,67 @@
 
    
 class paddle::CustomStackTrace
    
 
    A ThreadLocal stack for tracing train/test process. (More details of ThreadLocal can be find in the comments of ThreadLocal class.)
    
-
    For example.  
    paddle::CustomStackTrace<std::string> stack;
-PASS_TEST=0;
-for (auto& layer : layers){
-  stack.push(layer->getName());
-  layer->forward(passType);
-}
-for (auto& layer : layers){
-  layer->backward(passType);
-  stack.pop(layer->getName());
-}
+
    For example.  
    paddle::CustomStackTrace<std::string> stack;
+for (auto& layer : layers){
+  stack.push(layer->getName());
+  layer->forward();
+}
 
-if(passType == PASS_TEST) {
-  stack.clear();
-}
-else {
-  stack.dump([](const std::string& layername){
-    LOG(INFO) << "LayerName: " << layername;
-  })
-}
+stack.pop("");  // mark under pop stage.
+
+for (auto it = layers.rbegin(); it != layers.rend(); ++it){
+  auto& layer = *it;
+  layer->backward(passType);
+  stack.pop(layer->getName());
+}
 
    
 
    
 
    
 
    
+
    Public Types
    
+
    
+
    
+typedef std::function<void(const std::thread::id&, bool *, const T&)> DumpCallback
    
+
    DumpCallback Type. It will be invoked many times by dump method. 
    
+
    The first parameter is stack thread id. The second parameter is the last action of stack is push or not. The third parameter is the item in stack. 
    
+
    
+
+
    
+
    
 
    Public Functions
    
 
    
 
    
-void pop(const T &ip)
    
-
    Pop out an item from the top of the stack. For safety the item will be poped should equal to ip. 
    
+void pop(const T &item)
+
    Pop out an item from the top of the stack if item == top. Else, just set status to popping. 
    
 
    
 
 
    
-
    
-template <typename Callback>
    
-
    
-void dump(Callback callback)
    
-
    Empty the stack by sequence from top to button. 
    
+
    
+void clear()
    
+
    clear current thread stack. 
    
+
    
+
+
    
+
    
+bool empty() const
    
+
    return true if all thread’s stack is empty. 
    
 
    
-
    Parameters
    
-
      
-
    • callback - 
      A function deal with each item while dumping. It must have and only have a in parameter which is the stack item. 
      
-
      • 
-
    
-
    
+
    Return
    
+
    true if empty 
    
 
    
 
    
 
    
 
 
    
-
    
-void clear()
    
-
    Only empty the stack. 
    
+
    
+void dump(const DumpCallback &callback, bool onlyCurrentThread = false)
    
+
    Dump all thread stack, and all stack will be cleared. 
    
 
    
 
 
    
 
    
-void push(const T &ip)
    
-
    Push item ip to the top of the stack. 
    
+void push(const T &item)
+
    Push item to current thread stack. 
    
 
    
 
 
    
@@ -160,14 +164,11 @@ else {
 
 
         
    
@@ -189,13 +190,13 @@ else {
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/source/utils/enum.html b/doc/source/utils/enum.html
index 6879fef842..d7b0f29dd7 100644
--- a/doc/source/utils/enum.html
+++ b/doc/source/utils/enum.html
@@ -6,7 +6,7 @@
   
     
     
-    enumeration_wrapper — PaddlePaddle  documentation
+    enumeration_wrapper — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
      
 
@@ -242,14 +242,11 @@
 
 
         
    
@@ -271,13 +268,13 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/source/utils/lock.html b/doc/source/utils/lock.html
index 4b21b8cd06..bb351fa497 100644
--- a/doc/source/utils/lock.html
+++ b/doc/source/utils/lock.html
@@ -6,7 +6,7 @@
   
     
     
-    Thread — PaddlePaddle  documentation
+    Thread — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
      
 
@@ -63,7 +63,7 @@
 
    
 class paddle::Thread
    
 
    A simple wrapper for std::thread 
    
-
    Subclassed by paddle::SocketServer, paddle::SocketWorker, paddle::ThreadWorker
    
+
    Subclassed by paddle::SocketServer, paddle::SocketWorker, paddle::ThreadWorker
    
 
    
 
    Public Functions
    
 
    
@@ -80,7 +80,7 @@
 
    
 
    
 void start()
    
-
    Creat a new thread and call run() function. 
    
+
    Creat a new thread and call run() function. 
    
 
    
 
 
    
@@ -118,9 +118,9 @@
 
    
 
    
 class paddle::ThreadWorker
    
-
    ThreadWorker maintains a job queue. It executes the jobs in the job queue sequentianlly in a separate thread.
    
-
    Use addJob() to add a new job to the job queue. 
    
-
    Inherits from paddle::Thread
    
+
    ThreadWorker maintains a job queue. It executes the jobs in the job queue sequentianlly in a separate thread.
    
+
    Use addJob() to add a new job to the job queue. 
    
+
    Inherits from paddle::Thread
    
 
    
 
    Public Types
    
 
    
@@ -207,14 +207,14 @@
 
    
 
    
 class paddle::SyncThreadPool
    
-
    SyncThreadPool maintains a pool of threads. It executes the job use all workers in the pool.
    
-
    Use exec() to run a new job, job complete when exec returned. Only one job can exec simultaneously.
    
-
    Each worker has an tid whose range is [0, getNumThreads()). JobFunc can use tid to divide input data. 
    
+
    SyncThreadPool maintains a pool of threads. It executes the job use all workers in the pool.
    
+
    Use exec() to run a new job, job complete when exec returned. Only one job can exec simultaneously.
    
+
    Each worker has an tid whose range is [0, getNumThreads()). JobFunc can use tid to divide input data. 
    
 
    
 
    Public Types
    
 
    
 
    
-typedef std::function<void(int tid, size_t numThreads)> JobFunc
    
+typedef std::function<void(int tid, size_t numThreads)> JobFunc
 
    
 
 
    
@@ -260,7 +260,7 @@
 
    Execute a job using all the theads in the pool. 
    
 
    
 
    Note
    
-
    For the ownerFunc, tid=getNumThreads(). 
    
+
    For the ownerFunc, tid=getNumThreads(). 
    
 
    Parameters
    
 
      
 
    • jobFunc - 
      The function to be executed. 
      
@@ -295,7 +295,7 @@
 
      Public Static Functions
      
 
      
 
      
-static void execHelper(SyncThreadPool *pool, JobFunc jobFunc)
      
+static void execHelper(SyncThreadPool *pool, JobFunc jobFunc)
 
      Execute a job if has pool, else use caller thread as a worker. 
      
 
      
 
      Parameters
      
@@ -316,7 +316,7 @@
 
      
 
      
 void start()
      
-
      Start all the workers in the pool, call their run() function. 
      
+
      Start all the workers in the pool, call their run() function. 
      
 
      
 
 
      
@@ -380,14 +380,14 @@
 template <class T>
 
      
 class paddle::MultiThreadWorker
      
-
      MultiThreadWorker maintains a job queue and a result queue. It executes the jobs in the job queue and puts the results into the result queue sequentially in multi separate threads.
      
+
      MultiThreadWorker maintains a job queue and a result queue. It executes the jobs in the job queue and puts the results into the result queue sequentially in multi separate threads.
      
 
      Add jobs:
      
-
      Use addJob() to add a new job to the job queue (the user added jobs should not return nullptr).
      
-
      Use stopAddJob() to stop adding new jobs to the job queue (addJob() can not be called after stopAddJob()).
      
+
      Use addJob() to add a new job to the job queue (the user added jobs should not return nullptr).
      
+
      Use stopAddJob() to stop adding new jobs to the job queue (addJob() can not be called after stopAddJob()).
      
 
      Normal stop:
      
-
      Use waitResult() to get the results until nullptr is returned. Use stop() to exit normally (stopAddJob() should be called first).
      
+
      Use waitResult() to get the results until nullptr is returned. Use stop() to exit normally (stopAddJob() should be called first).
      
 
      Force stop:
      
-
      Use forceStop() to exit forcibly even though there are remaining jobs in the job queue. 
      
+
      Use forceStop() to exit forcibly even though there are remaining jobs in the job queue. 
      
 
      
 
      Public Types
      
 
      
@@ -437,7 +437,7 @@
 
      Stop all the workers normally. 
      
 
      
 
      Note
      
-
      stopAddJob() should be called before it. 
      
+
      stopAddJob() should be called before it. 
      
 
      
 
      
 
      
@@ -448,7 +448,7 @@
 
      Stop all the workers forcibly. 
      
 
      
 
      Note
      
-
      This function will call stopAddJob() first and empty the result queue. 
      
+
      This function will call stopAddJob() first and empty the result queue. 
      
 
      
 
      
 
      
@@ -459,7 +459,7 @@
 
      Add a job to the job queue. 
      
 
      
 
      Note
      
-
      Job can not be added after calling stopAddJob(). 
      
+
      Job can not be added after calling stopAddJob(). 
      
 
      
 
      
 
      
@@ -553,12 +553,12 @@
 
      
 
      
 class paddle::AsyncThreadPool
      
-
      AsyncThreadPool maintains a job queue and threads pool. It executes the jobs from queue asynchronously.
      
+
      AsyncThreadPool maintains a job queue and threads pool. It executes the jobs from queue asynchronously.
      
 
      Add jobs:
      
-
      Use addJob() to add a new job to the job queue and get a std::future result. The caller’s thread continues running. Call std::future::get() when the result’s value is needed, and the caller’s thread may be blocked until thread-pool finished the job.
      
-
      Use addBatchJobs() to add a batch of jobs. Unlike addJob()‘s asynchronization, addBatchJobs will block caller’s thread until all jobs in the batch are finished.
      
-
      Stop: Use stop() to stop the thread pool. Job can be added once stopped.
      
-
      Process-wide Singleton: Use AsyncThreadPool::ProcessChannel(N) first to create N threads. Then call AsyncThreadPool::ProcessChannel() to get the process-wide global thread pool. 
      
+
      Use addJob() to add a new job to the job queue and get a std::future result. The caller’s thread continues running. Call std::future::get() when the result’s value is needed, and the caller’s thread may be blocked until thread-pool finished the job.
      
+
      Use addBatchJobs() to add a batch of jobs. Unlike addJob()‘s asynchronization, addBatchJobs will block caller’s thread until all jobs in the batch are finished.
      
+
      Stop: Use stop() to stop the thread pool. Job can be added once stopped.
      
+
      Process-wide Singleton: Use AsyncThreadPool::ProcessChannel(N) first to create N threads. Then call AsyncThreadPool::ProcessChannel() to get the process-wide global thread pool. 
      
 
      
 
      Public Types
      
 
      
@@ -621,7 +621,7 @@
 
      Add a batch of jobs to the queue. The main thread will be blocked until these jobs are finished. The results will be stored in results according to jobs order. 
      
 
      
 
      Note
      
-
      results may need to be carefully cleared before addBatchJobs(). 
      
+
      results may need to be carefully cleared before addBatchJobs(). 
      
 
      Template Parameters
      
 
        
 
      • F - 
        should have a return value.
        
@@ -668,8 +668,8 @@
 
        Public Static Functions
        
 
        
 
        
-static AsyncThreadPool &ProcessChannel(size_t initThreadNum = 0)
        
-
        A process-wide singleton. Used as a global thread pool It should be initialized by calling AsyncThreadPool::ProcessChannel(N) first to create N threads, then call AsyncThreadPool::ProcessChannel() will get the thread pool. 
        
+static AsyncThreadPool &ProcessChannel(size_t initThreadNum = 0)
+
        A process-wide singleton. Used as a global thread pool It should be initialized by calling AsyncThreadPool::ProcessChannel(N) first to create N threads, then call AsyncThreadPool::ProcessChannel() will get the thread pool. 
        
 
        
 
 
      
@@ -721,14 +721,11 @@
 
 
         
    
@@ -750,13 +747,13 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/source/utils/queue.html b/doc/source/utils/queue.html
index 5132abffd9..e0e98d1009 100644
--- a/doc/source/utils/queue.html
+++ b/doc/source/utils/queue.html
@@ -6,7 +6,7 @@
   
     
     
-    Queue — PaddlePaddle  documentation
+    Queue — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
      
 
@@ -65,27 +65,27 @@
 
    
 class paddle::Queue
    
 
    A thread-safe queue that automatically grows but never shrinks. Dequeue a empty queue will block current thread. Enqueue an element will wake up another thread that blocked by dequeue method.
    
-
    For example.  
    paddle::Queue<int> q;
-END_OF_JOB=-1
-void thread1() {
-  while (true) {
-    auto job = q.dequeue();
-    if (job == END_OF_JOB) {
-      break;
-    }
-    processJob(job);
-  }
-}
+
    For example.  
    paddle::Queue<int> q;
+END_OF_JOB=-1
+void thread1() {
+  while (true) {
+    auto job = q.dequeue();
+    if (job == END_OF_JOB) {
+      break;
+    }
+    processJob(job);
+  }
+}
 
-void thread2() {
-  while (true) {
-     auto job = getJob();
-     q.enqueue(job);
-     if (job == END_OF_JOB) {
-       break;
-     }
-  }
-}
+void thread2() {
+  while (true) {
+     auto job = getJob();
+     q.enqueue(job);
+     if (job == END_OF_JOB) {
+       break;
+     }
+  }
+}
 
    
 
    
 
    
@@ -94,7 +94,7 @@ void thread2() {
 
    
 
    
 Queue()
    
-
    Construct Function. Default capacity of Queue is zero. 
    
+
    Construct Function. Default capacity of Queue is zero. 
    
 
    
 
 
    
@@ -105,7 +105,7 @@ void thread2() {
 
    
 
    
 void enqueue(const T &el)
    
-
    enqueue an element into Queue. 
    
+
    enqueue an element into Queue. 
    
 
    
 
    Note
    
 
    This method is thread-safe, and will wake up another blocked thread. 
    
@@ -122,7 +122,7 @@ void thread2() {
 
    
 
    
 void enqueue(T &&el)
    
-
    enqueue an element into Queue. 
    
+
    enqueue an element into Queue. 
    
 
    
 
    Note
    
 
    This method is thread-safe, and will wake up another blocked thread. 
    
@@ -207,12 +207,12 @@ void thread2() {
 
    
 
    
 void enqueue(const T &x)
    
-
    enqueue an element into Queue. 
    
+
    enqueue an element into Queue. 
    
 
    
 
    Note
    
-
    This method is thread-safe, and will wake up another thread who was blocked because of the queue is empty. 
    
-
    Note
    
-
    If it’s size() >= capacity before enqueue, this method will block and wait until size() < capacity. 
    
+
    This method is thread-safe, and will wake up another thread who was blocked because of the queue is empty. 
    
+
    If it’s size() >= capacity before enqueue, this method will block and wait until size() < capacity. 
    
+
    
 
    Parameters
    
 
      
 
    • x - 
      The enqueue element, pass by reference . 
      
@@ -228,9 +228,9 @@ void thread2() {
 T dequeue()
 
      Dequeue from a queue and return a element. 
      
 
      Note
      
-
      this method will be blocked until not empty. 
      
-
      Note
      
-
      this method will wake up another thread who was blocked because of the queue is full. 
      
+
      this method will be blocked until not empty. 
      
+
      this method will wake up another thread who was blocked because of the queue is full. 
      
+
      
 
      
 
      
 
    
@@ -296,14 +296,11 @@ void thread2() {
 
 
         
    
@@ -325,13 +322,13 @@ void thread2() {
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/source/utils/thread.html b/doc/source/utils/thread.html
index 3fd5efab64..1ca2b87bf8 100644
--- a/doc/source/utils/thread.html
+++ b/doc/source/utils/thread.html
@@ -6,7 +6,7 @@
   
     
     
-    Lock — PaddlePaddle  documentation
+    Lock — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
      
 
@@ -64,8 +64,8 @@
 class paddle::RWLock
 
    A simple read-write lock. The RWlock allows a number of readers or at most one writer at any point in time. The RWlock disable copy.
    
 
    Lock:
    
-
    Use lock() to lock on write mode, no other thread can get it until unlock.
    
-
    Use lock_shared() to lock on read mode, other thread can get it by using the same method lock_shared().
    
+
    Use lock() to lock on write mode, no other thread can get it until unlock.
    
+
    Use lock_shared() to lock on read mode, other thread can get it by using the same method lock_shared().
    
 
    Unlock:
    
 
    Use unlock() to unlock the lock. 
    
 
    
@@ -87,7 +87,7 @@
 
 
    
 
    
-RWLock &operator=(const RWLock&)
    
+RWLock &operator=(const RWLock&)
 
    
 
 
    
@@ -134,7 +134,7 @@
 
    
 
    
 class paddle::ReadLockGuard
    
-
    The ReadLockGuard is a read mode RWLock using RAII management mechanism. 
    
+
    The ReadLockGuard is a read mode RWLock using RAII management mechanism. 
    
 
    
 
    Public Functions
    
 
    
@@ -171,7 +171,7 @@
 
    
 
    
 class paddle::SpinLock
    
-
    A simple wrapper for spin lock. The lock() method of SpinLock is busy-waiting which means it will keep trying to lock until lock on successfully. The SpinLock disable copy. 
    
+
    A simple wrapper for spin lock. The lock() method of SpinLock is busy-waiting which means it will keep trying to lock until lock on successfully. The SpinLock disable copy. 
    
 
    
 
    Public Functions
    
 
    
@@ -191,7 +191,7 @@
 
 
    
 
    
-SpinLock &operator=(const SpinLock&)
    
+SpinLock &operator=(const SpinLock&)
 
    
 
 
    
@@ -252,7 +252,7 @@
 
    
 
    
 bool timeWait(struct timespec *ts)
    
-
    The same as wait(), except if the decrement can not be performed until ts return false install of blocking. 
    
+
    The same as wait(), except if the decrement can not be performed until ts return false install of blocking. 
    
 
    
 
    Return
    
 
    ture if the decrement proceeds before ts, else return false. 
    
@@ -275,7 +275,7 @@
 
    
 
    
 void post()
    
-
    increment the semaphore. If the semaphore’s value greater than 0, wake up a thread blocked in wait(). 
    
+
    increment the semaphore. If the semaphore’s value greater than 0, wake up a thread blocked in wait(). 
    
 
    
 
 
    
@@ -295,13 +295,13 @@
 
    
 
    
 class paddle::ThreadBarrier
    
-
    A simple wrapper of thread barrier. The ThreadBarrier disable copy. 
    
+
    A simple wrapper of thread barrier. The ThreadBarrier disable copy. 
    
 
    
 
    Public Functions
    
 
    
 
    
 ThreadBarrier(int count)
    
-
    Construct Function. Initialize the barrier should wait for count threads in wait(). 
    
+
    Construct Function. Initialize the barrier should wait for count threads in wait(). 
    
 
    
 
 
    
@@ -316,7 +316,7 @@
 
 
    
 
    
-ThreadBarrier &operator=(const ThreadBarrier&)
    
+ThreadBarrier &operator=(const ThreadBarrier&)
 
    
 
 
    
@@ -454,14 +454,11 @@
 
 
         
    
@@ -483,13 +480,13 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • Source Code Documents »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • Source Code Documents »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/ui/api/trainer_config_helpers/activations.html b/doc/ui/api/trainer_config_helpers/activations.html
index 334093639e..693dedb83f 100644
--- a/doc/ui/api/trainer_config_helpers/activations.html
+++ b/doc/ui/api/trainer_config_helpers/activations.html
@@ -6,7 +6,7 @@
   
     
     
-    BaseActivation — PaddlePaddle  documentation
+    BaseActivation — PaddlePaddle  documentation
     
     
     
@@ -45,10 +45,10 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    • 
-          
  • Model Config Interface »
    • 
-          
  • Activations »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
+          
  • Model Config Interface »
    • 
+          
  • Activations »
    •  
       
     
      
 
@@ -155,9 +155,9 @@ supported by hppl.
 class paddle.trainer_config_helpers.activations.SequenceSoftmaxActivation
 
    Softmax activation for one sequence. The dimension of input feature must be
 1 and a sequence.
    
-
    result = softmax(for each_feature_vector[0] in input_feature)
-for i, each_time_step_output in enumerate(output):
-    each_time_step_output = result[i]
+
    result = softmax(for each_feature_vector[0] in input_feature)
+for i, each_time_step_output in enumerate(output):
+    each_time_step_output = result[i]
 
    
 
    
 
    
@@ -262,14 +262,11 @@ for i, each_time_step_output in enumerate(output):
 
 
         
    
@@ -291,15 +288,15 @@ for i, each_time_step_output in enumerate(output):
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    • 
-          
  • Model Config Interface »
    • 
-          
  • Activations »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
+          
  • Model Config Interface »
    • 
+          
  • Activations »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/ui/api/trainer_config_helpers/activations_index.html b/doc/ui/api/trainer_config_helpers/activations_index.html
index b28b190fc3..924da78c4e 100644
--- a/doc/ui/api/trainer_config_helpers/activations_index.html
+++ b/doc/ui/api/trainer_config_helpers/activations_index.html
@@ -6,7 +6,7 @@
   
     
     
-    Activations — PaddlePaddle  documentation
+    Activations — PaddlePaddle  documentation
     
     
     
@@ -45,9 +45,9 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    • 
-          
  • Model Config Interface »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
+          
  • Model Config Interface »
    •  
       
     
      
 
@@ -99,14 +99,11 @@
 
 
         
    
@@ -128,14 +125,14 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    • 
-          
  • Model Config Interface »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
+          
  • Model Config Interface »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/ui/api/trainer_config_helpers/attrs.html b/doc/ui/api/trainer_config_helpers/attrs.html
index 137ff7455f..9847e4e080 100644
--- a/doc/ui/api/trainer_config_helpers/attrs.html
+++ b/doc/ui/api/trainer_config_helpers/attrs.html
@@ -6,7 +6,7 @@
   
     
     
-    Parameter and Extra Layer Attribute — PaddlePaddle  documentation
+    Parameter and Extra Layer Attribute — PaddlePaddle  documentation
     
     
     
@@ -45,9 +45,9 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    • 
-          
  • Model Config Interface »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
+          
  • Model Config Interface »
    •  
       
     
      
 
@@ -166,14 +166,11 @@ details of what dropout is please refer to 
    
           previous |
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    • 
-          
  • Model Config Interface »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
+          
  • Model Config Interface »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/ui/api/trainer_config_helpers/data_sources.html b/doc/ui/api/trainer_config_helpers/data_sources.html
index 555bac04e7..bdee4d0ce7 100644
--- a/doc/ui/api/trainer_config_helpers/data_sources.html
+++ b/doc/ui/api/trainer_config_helpers/data_sources.html
@@ -6,7 +6,7 @@
   
     
     
-    DataSources — PaddlePaddle  documentation
+    DataSources — PaddlePaddle  documentation
     
     
     
@@ -45,9 +45,9 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    • 
-          
  • Model Config Interface »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
+          
  • Model Config Interface »
    •  
       
     
      
 
@@ -58,36 +58,29 @@
             
   
    
 
    DataSources
    
-
    Data Sources are helpers to define paddle training data or testing data.
-There are several data attributes will be used by paddle:
    
-
      
-
    • Data ProviderType: such as Python, Protobuf
      • 
-
    • Data File list: a single file that contains all data file paths
      • 
-
    
+
    Data Sources are helpers to define paddle training data or testing data.
    
 
    
-
    
-paddle.trainer_config_helpers.data_sources.define_py_data_sources(train_list, test_list, module, obj, args=None, train_async=False, data_cls=<function PyData>)
    
+
    
+paddle.trainer_config_helpers.data_sources.define_py_data_sources2(train_list, test_list, module, obj, args=None)
    
 
    Define python Train/Test data sources in one method. If train/test use
 the same Data Provider configuration, module/obj/args contain one argument,
 otherwise contain a list or tuple of arguments. For example:
    
-
    define_py_data_sources("train.list", "test.list", module="data_provider"
-                       obj="process", args={"dictionary": dict_name})
-
    
-
    
-
    Or.
    
-
    define_py_data_sources("train.list", "test.list", module="data_provider"
-                       obj=["process_train", "process_test"],
-                       args=[{"dictionary": dict_train}, {"dictionary": dict_test}])
+
    define_py_data_sources2(train_list="train.list",
+                        test_list="test.list",
+                        module="data_provider"
+                        # if train/test use different configurations,
+                        # obj=["process_train", "process_test"]
+                        obj="process",
+                        args={"dictionary": dict_name})
 
    
 
    
 
    The related data provider can refer to
-here.
    
+here.
    
 
@@ -137,14 +129,11 @@ or list to this argument.
@@ -166,14 +155,14 @@ or list to this argument.
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    • 
-          
  • Model Config Interface »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
+          
  • Model Config Interface »
    • 
\ No newline at end of file
diff --git a/doc/ui/api/trainer_config_helpers/evaluators.html b/doc/ui/api/trainer_config_helpers/evaluators.html
index 7b4e0ef1ee..850d5004de 100644
--- a/doc/ui/api/trainer_config_helpers/evaluators.html
+++ b/doc/ui/api/trainer_config_helpers/evaluators.html
@@ -6,7 +6,7 @@
   
     
-    Base — PaddlePaddle  documentation
+    Base — PaddlePaddle  documentation
@@ -45,10 +45,10 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    • 
-          
  • Model Config Interface »
    • 
-          
  • Evaluators »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
+          
  • Model Config Interface »
    • 
+          
  • Evaluators »
    •   
 
@@ -171,7 +171,7 @@ important this sample is.paddle.trainer_config_helpers.evaluators.ctc_error_evaluator(*args, **kwargs)
    This evaluator is to calculate sequence-to-sequence edit distance.
    
 
    The simple usage is :
    
-
    eval = ctc_error_evaluator(input)
+
    eval = ctc_error_evaluator(input=input, label=lbl)
 
    
 
    
 
    
 
 
 
 
    Parameters:
      
-
    • data_cls – 
      • 
 
    • train_list (basestring) – Train list name.
      • 
 
    • test_list (basestring) – Test list name.
      • 
 
    • module (basestring or tuple or list) – python module name. If train and test is different, then
@@ -99,7 +92,6 @@ a tuple or list to this argument.
      • 
 DataProvider, and use @init_hook_wrapper to receive
 arguments. If train and test is different, then pass a tuple
 or list to this argument.
-
    • train_async (bool) – Is training data load asynchronously or not.
      • 
 
    
     		
 
    
 
 
         
          
       
     
     
   
 
     
     
     
      
       
     
 
 
    
@@ -180,7 +180,9 @@ important this sample is.
@@ -618,14 +620,11 @@ one or more input layers.
    
@@ -647,15 +646,15 @@ one or more input layers.
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    • 
-          
  • Model Config Interface »
    • 
-          
  • Evaluators »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
+          
  • Model Config Interface »
    • 
+          
  • Evaluators »
    • 
\ No newline at end of file
diff --git a/doc/ui/api/trainer_config_helpers/evaluators_index.html b/doc/ui/api/trainer_config_helpers/evaluators_index.html
index 54b12d7ab3..680b38c130 100644
--- a/doc/ui/api/trainer_config_helpers/evaluators_index.html
+++ b/doc/ui/api/trainer_config_helpers/evaluators_index.html
@@ -6,7 +6,7 @@
   
     
-    Evaluators — PaddlePaddle  documentation
+    Evaluators — PaddlePaddle  documentation
@@ -45,9 +45,9 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    • 
-          
  • Model Config Interface »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
+          
  • Model Config Interface »
    •   
 
@@ -113,14 +113,11 @@
 
@@ -142,14 +139,14 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    • 
-          
  • Model Config Interface »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
+          
  • Model Config Interface »
    • 
\ No newline at end of file
diff --git a/doc/ui/api/trainer_config_helpers/index.html b/doc/ui/api/trainer_config_helpers/index.html
index a82784006d..e753ca1acc 100644
--- a/doc/ui/api/trainer_config_helpers/index.html
+++ b/doc/ui/api/trainer_config_helpers/index.html
@@ -6,7 +6,7 @@
   
     
-    Model Config Interface — PaddlePaddle  documentation
+    Model Config Interface — PaddlePaddle  documentation
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    •   
 
@@ -93,14 +93,11 @@
 
@@ -122,13 +119,13 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
\ No newline at end of file
diff --git a/doc/ui/api/trainer_config_helpers/layers.html b/doc/ui/api/trainer_config_helpers/layers.html
index 7950f1075a..af7e54886b 100644
--- a/doc/ui/api/trainer_config_helpers/layers.html
+++ b/doc/ui/api/trainer_config_helpers/layers.html
@@ -6,7 +6,7 @@
   
     
-    Base — PaddlePaddle  documentation
+    Base — PaddlePaddle  documentation
@@ -45,10 +45,10 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    • 
-          
  • Model Config Interface »
    • 
-          
  • Layers »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
+          
  • Model Config Interface »
    • 
+          
  • Layers »
    •   
 
@@ -121,18 +121,6 @@ reasons.
    
 
 
    Parameters:
      
 
    • name (None|basestring) – Evaluator name.
      • 
-
    • input (LayerOutput) – Input Layer.
      • 
+
    • input (LayerOutput) – Input Layer. Should be the same as the input for ctc_layer.
      • 
+
    • label (LayerOutput) – input label, which is a data_layer. Should be the same as the
+label for ctc_layer
      • 
 
    
     		
 
    
 
 
         
          
       
     
     
   
 
     
     
     
      
       
     
 
          
       
     
     
   
 
     
     
     
      
       
     
 
          
       
     
     
   
 
     
     
     
      
       
     
 
     
 
    
-
    
-
    
-__repr__()
    
-
    Disable __repr__ for debug reason. Will be implemented when release
    
-
    
-
-
    
-
    
-__str__()
    
-
    Disable __str__ for debug reason. Will be implemented when release
    
-
    
-
 
    
 
 
    
@@ -503,7 +491,7 @@ MaxPooling.
 
    
 
    
 paddle.trainer_config_helpers.layers.img_cmrnorm_layer(*args, **kwargs)
    
-
    Convolution cross-map-response-normalize layer.
+
    Response normalization across feature maps.
 The details please refer to
 Alex’s paper.
    
 
@@ -513,7 +501,7 @@ The details please refer to
 
    Parameters:
      
 
    • name (None|basestring) – layer name.
      • 
 
    • input (LayerOutput) – layer’s input.
      • 
-
    • size (int) – cross map response size.
      • 
+
    • size (int) – Normalize in number of \(size\) feature maps.
      • 
 
    • scale (float) – The hyper-parameter.
      • 
 
    • power (float) – The hyper-parameter.
      • 
 
    • num_channels – input layer’s filers number or channels. If
@@ -533,42 +521,6 @@ num_channels is None, it will be set automatically.
      • 
 
    
 
    
 
-
    
-
    
-
    img_rnorm_layer
    
-
    
-
    
-paddle.trainer_config_helpers.layers.img_rnorm_layer(*args, **kwargs)
    
-
    Normalize the input in local region, namely response normalization
-across feature maps.
    
-
---
-
-
-
-
-
-
-
-
-
-
    


    Parameters:
      
-
    • name – The name of this layer.
      • 
-
    • input – The input of this layer.
      • 
-
    • size – 
      • 
-
    • scale – 
      • 
-
    • power – 
      • 
-
    • num_channels – 
      • 
-
    • layer_attr – 
      • 
-
    
-
    Rtype name:
    None|basestring
    
-
    Returns:
    LayerOutput object.
    
-
    Return type:
    LayerOutput
    
-
    
-
    
-
 
    
 
    
 
    batch_norm_layer
    
@@ -715,7 +667,7 @@ and \(out\) is a (batchSize x dataDim) output vector.<
 
    Long Short-term Memory Cell.
    
 
    The memory cell was implemented as follow equations.
    
 
    
-\[\begin{split}i_t & = \sigma(W_{xi}x_{t} + W_{hi}h_{t-1} + W_{ci}c_{t-1} + b_i)\end{split}\]\[\begin{split}f_t & = \sigma(W_{xf}x_{t} + W_{hf}h_{t-1} + W_{cf}c_{t-1} + b_f)\end{split}\]\[\begin{split}c_t & = f_tc_{t-1} + i_t tanh (W_{xc}x_t+W_{hc}h_{t-1} + b_c)\end{split}\]\[\begin{split}o_t & = \sigma(W_{xo}x_{t} + W_{ho}h_{t-1} + W_{co}c_t + b_o)\end{split}\]\[\begin{split}h_t & = o_t tanh(c_t)\end{split}\]
    
+\[ \begin{align}\begin{aligned}i_t & = \sigma(W_{xi}x_{t} + W_{hi}h_{t-1} + W_{ci}c_{t-1} + b_i)\\f_t & = \sigma(W_{xf}x_{t} + W_{hf}h_{t-1} + W_{cf}c_{t-1} + b_f)\\c_t & = f_tc_{t-1} + i_t tanh (W_{xc}x_t+W_{hc}h_{t-1} + b_c)\\o_t & = \sigma(W_{xo}x_{t} + W_{ho}h_{t-1} + W_{co}c_t + b_o)\\h_t & = o_t tanh(c_t)\end{aligned}\end{align} \]
    
 
    NOTE: In PaddlePaddle’s implementation, the multiplications
 \(W_{xi}x_{t}\) , \(W_{xf}x_{t}\),
 \(W_{xc}x_t\), \(W_{xo}x_{t}\) are not done in the lstmemory layer,
@@ -764,13 +716,13 @@ bias.
 
    LSTM Step Layer. It used in recurrent_group. The lstm equations are shown
 as follow.
    
 
    
-\[\begin{split}i_t & = \sigma(W_{xi}x_{t} + W_{hi}h_{t-1} + W_{ci}c_{t-1} + b_i)\end{split}\]\[\begin{split}f_t & = \sigma(W_{xf}x_{t} + W_{hf}h_{t-1} + W_{cf}c_{t-1} + b_f)\end{split}\]\[\begin{split}c_t & = f_tc_{t-1} + i_t tanh (W_{xc}x_t+W_{hc}h_{t-1} + b_c)\end{split}\]\[\begin{split}o_t & = \sigma(W_{xo}x_{t} + W_{ho}h_{t-1} + W_{co}c_t + b_o)\end{split}\]\[\begin{split}h_t & = o_t tanh(c_t)\end{split}\]
    
+\[ \begin{align}\begin{aligned}i_t & = \sigma(W_{xi}x_{t} + W_{hi}h_{t-1} + W_{ci}c_{t-1} + b_i)\\f_t & = \sigma(W_{xf}x_{t} + W_{hf}h_{t-1} + W_{cf}c_{t-1} + b_f)\\c_t & = f_tc_{t-1} + i_t tanh (W_{xc}x_t+W_{hc}h_{t-1} + b_c)\\o_t & = \sigma(W_{xo}x_{t} + W_{ho}h_{t-1} + W_{co}c_t + b_o)\\h_t & = o_t tanh(c_t)\end{aligned}\end{align} \]
    
 
    The input of lstm step is \(Wx_t + Wh_{t-1}\), and user should use
 mixed_layer and full_matrix_projection to calculate these
 input vector.
    
 
    The state of lstm step is \(c_{t-1}\). And lstm step layer will do
    
 
    
-\[i_t = \sigma(input + W_{ci}c_{t-1} + b_i)\]\[...\]
    
+\[ \begin{align}\begin{aligned}i_t = \sigma(input + W_{ci}c_{t-1} + b_i)\\...\end{aligned}\end{align} \]
    
 
    This layer contains two outputs. Default output is \(h_t\). The other
 output is \(o_t\), which name is ‘state’ and can use
 get_output_layer to extract this output.
    
@@ -1549,7 +1501,7 @@ Inputs can be list of LayerOutput or list of projection.
    
 
    
 
    
 
    
-\[outputH = 1 + (2 * padding_y + imgSizeH - block_y + stride_y - 1) / stride_y\]\[outputW = 1 + (2 * padding_x + imgSizeW - block_x + stride_x - 1) / stride_x\]
    
+\[ \begin{align}\begin{aligned}outputH = 1 + (2 * padding_y + imgSizeH - block_y + stride_y - 1) / stride_y\\outputW = 1 + (2 * padding_x + imgSizeW - block_x + stride_x - 1) / stride_x\end{aligned}\end{align} \]
    
 
    The expand method is the same with ExpandConvLayer, but saved the transposed
 value. After expanding, output.sequenceStartPositions will store timeline.
 The number of time steps are outputH * outputW and the dimension of each
@@ -1686,38 +1638,44 @@ bias.
 
    
 
 
    
-
    
-
    convex_comb_layer
    
+
    
+
    linear_comb_layer
    
 
    
 
    
-paddle.trainer_config_helpers.layers.convex_comb_layer(*args, **kwargs)
    
+paddle.trainer_config_helpers.layers.linear_comb_layer(*args, **kwargs)
 
    
-
    A layer for convex weighted average of vectors takes two inputs.
    
+
    A layer for weighted sum of vectors takes two inputs.
    
 
      
 
    • 
-
      Input: a vector containing the convex weights (batchSize x weightdim),
      
-
      and a matrix in a vector form (batchSize x (weightdim * datadim)).
      
+
      Input: size of weights is M
      
+
      size of vectors is M*N
      
 
      
 
      
 
      • 
-
    • Output: a vector (batchSize * datadim).
      
+
    • Output: a vector of size=N
      
 
      • 
 
    
 
    
 
    
 
    
-\[y[i][j] = \sum_{j}(x_{1}(i, j) * x_{2}(i,j + i * dataDim)),\]\[            i = 0,1,...,(batchSize-1); j = 0, 1,...,(dataDim-1)\]
    
+\[z(i) = \sum_{j=0}^{M-1} x(j) y(i+Nj)\]
    
+
    where \(0 \le i \le N-1\)
    
+
    Or in the matrix notation:
    
+
    
+\[z = x^\mathrm{T} Y\]
    
 
    
 
    In this formular:
    
 
      
-
    • \(x_{1}\): the first input.
      • 
-
    • \(x_{2}\): the second input.
      • 
-
    • \(y\): the output.
      • 
+
    • \(x\): weights
      • 
+
    • \(y\): vectors.
      • 
+
    • \(z\): the output.
      • 
 
    
 
    
 
    
+
    Note that the above computation is for one sample. Multiple samples are
+processed in one batch.
    
 
    The simple usage is:
    
-
    convex_comb = convex_comb_layer(input=inputs,
+
    linear_comb = linear_comb_layer(weighs=weight, vectors=vectors,
                                 size=elem_dim)
 
    
 
    
@@ -1825,11 +1783,13 @@ and \(y\) is a output vector.
    
 
    
 
    
 paddle.trainer_config_helpers.layers.scaling_layer(*args, **kwargs)
    
-
    A layer for each row of a matrix, multiplying with a element of a vector.
    
+
    A layer for multiplying input vector by weight scalar.
    
 
    
-\[y.row[i] = w[i] * x.row[i]\]
    
-
    where \(x\) is (batchSize x dataDim) input, \(w\) is
-(batchSize x 1) weight vector, and \(y\) is (batchSize x dataDim) output.
    
+\[y  = w x\]
    
+
    where \(x\) is size=dataDim input, \(w\) is size=1 weight,
+and \(y\) is size=dataDim output.
    
+
    Note that the above computation is for one sample. Multiple samples are
+processed in one batch.
    
 
    The example usage is:
    
 
    scale = scaling_layer(input=layer1, weight=layer2)
 
    
@@ -1944,6 +1904,45 @@ default Bias.
 
    
 
    
 
+
+
    
+
    cos_sim
    
+
    
+
    
+paddle.trainer_config_helpers.layers.cos_sim(*args, **kwargs)
    
+
    Cosine Similarity Layer. The cosine similarity equation is here.
    
+
    
+\[similarity = cos(\theta) = {\mathbf{a} \cdot \mathbf{b}
+\over \|\mathbf{a}\| \|\mathbf{b}\|}\]
    
+
    The size of a is M, size of b is M*N,
+Similarity will be calculated N times by step M. The output size is
+N. The scale will be multiplied to similarity.
    
+
    Note that the above computation is for one sample. Multiple samples are
+processed in one batch.
    
+
+++
+
+
+
+
+
+
+
+
    


    Parameters:
      
+
    • name (basestring) – layer name
      • 
+
    • a (LayerOutput) – input layer a
      • 
+
    • b (LayerOutput) – input layer b
      • 
+
    • scale (float) – scale for cosine value. default is 5.
      • 
+
    • size (int) – layer size. NOTE size_a * size should equal size_b.
      • 
+
    • layer_attr (ExtraLayerAttribute) – Extra Layer Attribute.
      • 
+
    
+
    Returns:
    LayerOutput object.
    
+
    Return type:
    LayerOutput
    
+
    
+
    
+
 
    
 
+\[ \begin{align}\begin{aligned}C_{i,j} & = -\tilde{P_{ij}} * o_{i,j} + log(1 + e^{o_{i,j}})\\o_{i,j} & =  o_i - o_j\\\tilde{P_{i,j}} & = \{0, 0.5, 1\} \ or \ \{0, 1\}\end{aligned}\end{align} \]
 
    
 
    In this formula:
    
 
      
@@ -2285,43 +2284,6 @@ It is an optional argument.
 
    
 
 
-
    
-
    
-
    cos_sim
    
-
    
-
    
-paddle.trainer_config_helpers.layers.cos_sim(*args, **kwargs)
    
-
    Cosine Similarity Layer. The cosine similarity equation is here.
    
-
    
-\[similarity = cos(\theta) = {\mathbf{a} \cdot \mathbf{b}
-\over \|\mathbf{b}\| \|\mathbf{b}\|}\]
    
-
    And the input dimension is \(a \in R^M\), \(b \in R^{MN}\). The
-similarity will be calculated N times by step M. The output dimension is
-\(R^N\). The scale will be multiplied to similarity.
    
-
---
-
-
-
-
-
-
-
-
    


    Parameters:
      
-
    • name (basestring) – layer name
      • 
-
    • a (LayerOutput) – input layer a
      • 
-
    • b (LayerOutput) – input layer b
      • 
-
    • scale (float) – scale for cosine value. default is 5.
      • 
-
    • size (int) – layer size. NOTE size_a * size should equal size_b.
      • 
-
    • layer_attr (ExtraLayerAttribute) – Extra Layer Attribute.
      • 
-
    
-
    Returns:
    LayerOutput object.
    
-
    Return type:
    LayerOutput
    
-
    
-
    
-
 
    
 
    
 
    crf_layer
    
@@ -2404,6 +2366,17 @@ decoding or 0 for correct decoding.
    
 
    Connectionist Temporal Classification (CTC) is designed for temporal
 classication task. That is, for sequence labeling problems where the
 alignment between the inputs and the target labels is unknown.
    
+
    More details can be found by referring to Connectionist Temporal
+Classification: Labelling Unsegmented Sequence Data with Recurrent
+Neural Networks
    
+
    
+
    Note
    
+
    Considering the ‘blank’ label needed by CTC, you need to use
+(num_classes + 1) as the input size. num_classes is the category number.
+And the ‘blank’ is the last category index. So the size of ‘input’ layer, such as
+fc_layer with softmax activation, should be num_classes + 1. The size of ctc_layer
+should also be num_classes + 1.
    
+
    
 
    The simple usage:
    
 
    ctc = ctc_layer(input=input,
                 label=label,
@@ -2418,7 +2391,7 @@ alignment between the inputs and the target labels is unknown.
    
 Parameters:
      
 
    • input (LayerOutput) – The input layers.
      • 
 
    • label (LayerOutput) – The data layer of label with variable length.
      • 
-
    • size (int) – category numbers.
      • 
+
    • size (int) – category numbers + 1.
      • 
 
    • name (string|None) – The name of this layer, which can not specify.
      • 
 
    • norm_by_times (bool) – Whether to normalization by times. False by default.
      • 
 
    
@@ -2553,7 +2526,6 @@ It is used by recurrent layer group.
    
 
 
  • Norm Layer
@@ -2596,12 +2568,13 @@ It is used by recurrent layer group.
    
 
    • 
 
  • Math Layers
 
    • 
@@ -2617,7 +2590,6 @@ It is used by recurrent layer group.
    
 
  • huber_cost
    • 
 
  • lambda_cost
    • 
 
  • rank_cost
    • 
-
  • cos_sim
    • 
 
  • crf_layer
    • 
 
  • crf_decoding_layer
    • 
 
  • ctc_layer
    • 
@@ -2646,14 +2618,11 @@ It is used by recurrent layer group.
    
 
 
         
    
@@ -2675,15 +2644,15 @@ It is used by recurrent layer group.
    
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    • 
-          
  • Model Config Interface »
    • 
-          
  • Layers »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
+          
  • Model Config Interface »
    • 
+          
  • Layers »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/ui/api/trainer_config_helpers/layers_index.html b/doc/ui/api/trainer_config_helpers/layers_index.html
index fd63942bb5..3efe1148b7 100644
--- a/doc/ui/api/trainer_config_helpers/layers_index.html
+++ b/doc/ui/api/trainer_config_helpers/layers_index.html
@@ -6,7 +6,7 @@
   
     
     
-    Layers — PaddlePaddle  documentation
+    Layers — PaddlePaddle  documentation
     
     
     
@@ -45,9 +45,9 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    • 
-          
  • Model Config Interface »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
+          
  • Model Config Interface »
    •  
       
     
      
 
@@ -87,7 +87,6 @@
 
 
  • Norm Layer
@@ -130,12 +129,13 @@
 
    • 
 
  • Math Layers
 
    • 
@@ -151,7 +151,6 @@
 
  • huber_cost
    • 
 
  • lambda_cost
    • 
 
  • rank_cost
    • 
-
  • cos_sim
    • 
 
  • crf_layer
    • 
 
  • crf_decoding_layer
    • 
 
  • ctc_layer
    • 
@@ -188,14 +187,11 @@
 
 
         
    
@@ -217,14 +213,14 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    • 
-          
  • Model Config Interface »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
+          
  • Model Config Interface »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/ui/api/trainer_config_helpers/networks.html b/doc/ui/api/trainer_config_helpers/networks.html
index 9080980e26..ee3da3c9a9 100644
--- a/doc/ui/api/trainer_config_helpers/networks.html
+++ b/doc/ui/api/trainer_config_helpers/networks.html
@@ -6,7 +6,7 @@
   
     
     
-    NLP — PaddlePaddle  documentation
+    NLP — PaddlePaddle  documentation
     
     
     
@@ -45,10 +45,10 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    • 
-          
  • Model Config Interface »
    • 
-          
  • Networks »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
+          
  • Model Config Interface »
    • 
+          
  • Networks »
    •  
       
     
      
 
@@ -328,7 +328,7 @@ mechanism.
    
 for more details about LSTM. The link goes as follows:
 .. _Link: https://arxiv.org/abs/1308.0850
    
 
    
-\[\begin{split}i_t & = \sigma(W_{xi}x_{t} + W_{hi}h_{t-1} + W_{ci}c_{t-1} + b_i)\end{split}\]\[\begin{split}f_t & = \sigma(W_{xf}x_{t} + W_{hf}h_{t-1} + W_{cf}c_{t-1} + b_f)\end{split}\]\[\begin{split}c_t & = f_tc_{t-1} + i_t tanh (W_{xc}x_t+W_{hc}h_{t-1} + b_c)\end{split}\]\[\begin{split}o_t & = \sigma(W_{xo}x_{t} + W_{ho}h_{t-1} + W_{co}c_t + b_o)\end{split}\]\[\begin{split}h_t & = o_t tanh(c_t)\end{split}\]
    
+\[ \begin{align}\begin{aligned}i_t & = \sigma(W_{xi}x_{t} + W_{hi}h_{t-1} + W_{ci}c_{t-1} + b_i)\\f_t & = \sigma(W_{xf}x_{t} + W_{hf}h_{t-1} + W_{cf}c_{t-1} + b_f)\\c_t & = f_tc_{t-1} + i_t tanh (W_{xc}x_t+W_{hc}h_{t-1} + b_c)\\o_t & = \sigma(W_{xo}x_{t} + W_{ho}h_{t-1} + W_{co}c_t + b_o)\\h_t & = o_t tanh(c_t)\end{aligned}\end{align} \]
    
 
    The example usage is:
    
 
    lstm_step = lstmemory_unit(input=[layer1],
                            size=256,
@@ -440,7 +440,7 @@ False means no bias, None means default bias.
 
    It just combine a mixed layer with fully_matrix_projection and a lstmemory
 layer. The simple lstm cell was implemented as follow equations.
    
 
    
-\[\begin{split}i_t & = \sigma(W_{xi}x_{t} + W_{hi}h_{t-1} + W_{ci}c_{t-1} + b_i)\end{split}\]\[\begin{split}f_t & = \sigma(W_{xf}x_{t} + W_{hf}h_{t-1} + W_{cf}c_{t-1} + b_f)\end{split}\]\[\begin{split}c_t & = f_tc_{t-1} + i_t tanh (W_{xc}x_t+W_{hc}h_{t-1} + b_c)\end{split}\]\[\begin{split}o_t & = \sigma(W_{xo}x_{t} + W_{ho}h_{t-1} + W_{co}c_t + b_o)\end{split}\]\[\begin{split}h_t & = o_t tanh(c_t)\end{split}\]
    
+\[ \begin{align}\begin{aligned}i_t & = \sigma(W_{xi}x_{t} + W_{hi}h_{t-1} + W_{ci}c_{t-1} + b_i)\\f_t & = \sigma(W_{xf}x_{t} + W_{hf}h_{t-1} + W_{cf}c_{t-1} + b_f)\\c_t & = f_tc_{t-1} + i_t tanh (W_{xc}x_t+W_{hc}h_{t-1} + b_c)\\o_t & = \sigma(W_{xo}x_{t} + W_{ho}h_{t-1} + W_{co}c_t + b_o)\\h_t & = o_t tanh(c_t)\end{aligned}\end{align} \]
    
 
    Please refer Generating Sequences With Recurrent Neural Networks if you
 want to know what lstm is. Link is here.
    
 
@@ -659,7 +659,7 @@ Please see grumemory in layers.py for more detail about the maths.
    Calculate and then return a context vector by attention machanism.
 Size of the context vector equals to size of the encoded_sequence.
    
 
    
-\[\begin{split}a(s_{i-1},h_{j}) & = v_{a}f(W_{a}s_{t-1} + U_{a}h_{j})\end{split}\]\[\begin{split}e_{i,j} & = a(s_{i-1}, h_{j})\end{split}\]\[\begin{split}a_{i,j} & = \frac{exp(e_{i,j})}{\sum_{k=1}^{T_x}{exp(e_{i,k})}}\end{split}\]\[\begin{split}c_{i} & = \sum_{j=1}^{T_{x}}a_{i,j}h_{j}\end{split}\]
    
+\[ \begin{align}\begin{aligned}a(s_{i-1},h_{j}) & = v_{a}f(W_{a}s_{t-1} + U_{a}h_{j})\\e_{i,j} & = a(s_{i-1}, h_{j})\\a_{i,j} & = \frac{exp(e_{i,j})}{\sum_{k=1}^{T_x}{exp(e_{i,k})}}\\c_{i} & = \sum_{j=1}^{T_{x}}a_{i,j}h_{j}\end{aligned}\end{align} \]
 
    where \(h_{j}\) is the jth element of encoded_sequence,
 \(U_{a}h_{j}\) is the jth element of encoded_proj
 \(s_{i-1}\) is decoder_state
@@ -815,14 +815,11 @@ train network’s output automatically.
    
 
 
         
@@ -844,15 +841,15 @@ train network’s output automatically.
    
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    • 
-          
  • Model Config Interface »
    • 
-          
  • Networks »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
+          
  • Model Config Interface »
    • 
+          
  • Networks »
    •  
       
     
     
   
 
\ No newline at end of file
diff --git a/doc/ui/api/trainer_config_helpers/networks_index.html b/doc/ui/api/trainer_config_helpers/networks_index.html
index dea3c7cfe1..52dae6330e 100644
--- a/doc/ui/api/trainer_config_helpers/networks_index.html
+++ b/doc/ui/api/trainer_config_helpers/networks_index.html
@@ -6,7 +6,7 @@
   
     
     
-    Networks — PaddlePaddle  documentation
+    Networks — PaddlePaddle  documentation
     
     
     
@@ -45,9 +45,9 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    • 
-          
  • Model Config Interface »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
+          
  • Model Config Interface »
    •  
       
       
 
@@ -121,14 +121,11 @@
 
 
         
@@ -150,14 +147,14 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    • 
-          
  • Model Config Interface »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
+          
  • Model Config Interface »
    •  
       
     
     
   
 
\ No newline at end of file
diff --git a/doc/ui/api/trainer_config_helpers/optimizers.html b/doc/ui/api/trainer_config_helpers/optimizers.html
index 9580d8e463..d61f9de5cb 100644
--- a/doc/ui/api/trainer_config_helpers/optimizers.html
+++ b/doc/ui/api/trainer_config_helpers/optimizers.html
@@ -6,7 +6,7 @@
   
     
     
-    BaseSGDOptimizer — PaddlePaddle  documentation
+    BaseSGDOptimizer — PaddlePaddle  documentation
     
     
     
@@ -45,10 +45,10 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    • 
-          
  • Model Config Interface »
    • 
-          
  • Optimizers »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
+          
  • Model Config Interface »
    • 
+          
  • Optimizers »
    •  
       
       
 
@@ -302,14 +302,11 @@ clipped.
 
 
         
@@ -331,15 +328,15 @@ clipped.
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    • 
-          
  • Model Config Interface »
    • 
-          
  • Optimizers »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
+          
  • Model Config Interface »
    • 
+          
  • Optimizers »
    •  
       
     
     
   
 
\ No newline at end of file
diff --git a/doc/ui/api/trainer_config_helpers/optimizers_index.html b/doc/ui/api/trainer_config_helpers/optimizers_index.html
index 35317211aa..0068733f1d 100644
--- a/doc/ui/api/trainer_config_helpers/optimizers_index.html
+++ b/doc/ui/api/trainer_config_helpers/optimizers_index.html
@@ -6,7 +6,7 @@
   
     
     
-    Optimizers — PaddlePaddle  documentation
+    Optimizers — PaddlePaddle  documentation
     
     
     
@@ -45,9 +45,9 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    • 
-          
  • Model Config Interface »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
+          
  • Model Config Interface »
    •  
       
       
 
@@ -94,14 +94,11 @@
 
 
         
@@ -123,14 +120,14 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    • 
-          
  • Model Config Interface »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
+          
  • Model Config Interface »
    •  
       
     
     
   
 
\ No newline at end of file
diff --git a/doc/ui/api/trainer_config_helpers/poolings.html b/doc/ui/api/trainer_config_helpers/poolings.html
index ac55d82b58..9ba59f0e3e 100644
--- a/doc/ui/api/trainer_config_helpers/poolings.html
+++ b/doc/ui/api/trainer_config_helpers/poolings.html
@@ -6,7 +6,7 @@
   
     
     
-    BasePoolingType — PaddlePaddle  documentation
+    BasePoolingType — PaddlePaddle  documentation
     
     
     
@@ -45,10 +45,10 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    • 
-          
  • Model Config Interface »
    • 
-          
  • Poolings »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
+          
  • Model Config Interface »
    • 
+          
  • Poolings »
    •  
       
       
 
@@ -156,14 +156,11 @@ Each PoolingType contains one parameter:
    
 
 
         
@@ -185,15 +182,15 @@ Each PoolingType contains one parameter:
    
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    • 
-          
  • Model Config Interface »
    • 
-          
  • Poolings »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
+          
  • Model Config Interface »
    • 
+          
  • Poolings »
    •  
       
     
     
   
 
\ No newline at end of file
diff --git a/doc/ui/api/trainer_config_helpers/poolings_index.html b/doc/ui/api/trainer_config_helpers/poolings_index.html
index 6c707aa70d..36152ee4da 100644
--- a/doc/ui/api/trainer_config_helpers/poolings_index.html
+++ b/doc/ui/api/trainer_config_helpers/poolings_index.html
@@ -6,7 +6,7 @@
   
     
     
-    Poolings — PaddlePaddle  documentation
+    Poolings — PaddlePaddle  documentation
     
     
     
@@ -45,9 +45,9 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    • 
-          
  • Model Config Interface »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
+          
  • Model Config Interface »
    •  
       
       
 
@@ -92,14 +92,11 @@
 
 
         
@@ -121,14 +118,14 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    • 
-          
  • Model Config Interface »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
+          
  • Model Config Interface »
    •  
       
     
     
   
 
\ No newline at end of file
diff --git a/doc/ui/cmd_argument/argument_outline.html b/doc/ui/cmd_argument/argument_outline.html
index 693419273f..d49687f792 100644
--- a/doc/ui/cmd_argument/argument_outline.html
+++ b/doc/ui/cmd_argument/argument_outline.html
@@ -6,7 +6,7 @@
   
     
     
-    Argument Outline — PaddlePaddle  documentation
+    Argument Outline — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    •  
       
       
 
@@ -325,14 +325,11 @@
 
 
         
@@ -354,13 +351,13 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    •  
       
     
     
   
 
\ No newline at end of file
diff --git a/doc/ui/cmd_argument/detail_introduction.html b/doc/ui/cmd_argument/detail_introduction.html
index 9cf9e9e203..991c295e8e 100644
--- a/doc/ui/cmd_argument/detail_introduction.html
+++ b/doc/ui/cmd_argument/detail_introduction.html
@@ -6,7 +6,7 @@
   
     
     
-    Detail Description — PaddlePaddle  documentation
+    Detail Description — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    •  
       
       
 
@@ -555,14 +555,11 @@
 
 
         
@@ -584,13 +581,13 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    •  
       
     
     
   
 
\ No newline at end of file
diff --git a/doc/ui/cmd_argument/use_case.html b/doc/ui/cmd_argument/use_case.html
index af1e19cc44..31ea5766db 100644
--- a/doc/ui/cmd_argument/use_case.html
+++ b/doc/ui/cmd_argument/use_case.html
@@ -6,7 +6,7 @@
   
     
     
-    Use Case — PaddlePaddle  documentation
+    Use Case — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    •  
       
       
 
@@ -60,35 +60,35 @@
 
    
 
    Local Training
    
 
    These command line arguments are commonly used by local training experiments, such as image classification, natural language processing, et al.
    
-
    paddle train \
-  --use_gpu=1/0 \                        #1:GPU,0:CPU(default:true)
-  --config=network_config \
-  --save_dir=output \
-  --trainer_count=COUNT \                #(default:1)
-  --test_period=M \                      #(default:1000）
-  --test_all_data_in_one_period=true \   #(default:false) 
-  --num_passes=N \                       #(defalut:100）
-  --log_period=K \                       #(default:100)
-  --dot_period=1000 \                    #(default:1)
-  #[--show_parameter_stats_period=100] \ #(default:0)
-  #[--saving_period_by_batches=200] \    #(default:0)
+
    paddle train \
+  --use_gpu=1/0 \                        #1:GPU,0:CPU(default:true)
+  --config=network_config \
+  --save_dir=output \
+  --trainer_count=COUNT \                #(default:1)
+  --test_period=M \                      #(default:1000）
+  --test_all_data_in_one_period=true \   #(default:false) 
+  --num_passes=N \                       #(defalut:100）
+  --log_period=K \                       #(default:100)
+  --dot_period=1000 \                    #(default:1)
+  #[--show_parameter_stats_period=100] \ #(default:0)
+  #[--saving_period_by_batches=200] \    #(default:0)
 
    
 
    
 
    show_parameter_stats_period and saving_period_by_batches are optional according to your task.
    
 
    
 
    1) Pass Command Argument to Network config
    
 
    config_args is a useful parameter to pass arguments to network config.
    
-
    --config_args=generating=1,beam_size=5,layer_num=10 \
+
    --config_args=generating=1,beam_size=5,layer_num=10 \
 
    
 
    
 
    And get_config_arg can be used to parse these arguments in network config as follows:
    
-
    generating = get_config_arg('generating', bool, False)
+
    generating = get_config_arg('generating', bool, False)
 beam_size = get_config_arg('beam_size', int, 3)
 layer_num = get_config_arg('layer_num', int, 8)
 
    
 
    
 
    get_config_arg:
    
-
    get_config_arg(name, type, default_value)
+
    get_config_arg(name, type, default_value)
 
    
 
    
 
      
@@ -100,7 +100,7 @@
 
      
 
      2) Use Model to Initialize Network
      
 
      add argument:
      
-
      --init_model_path=model_path
+
      --init_model_path=model_path
 --load_missing_parameter_strategy=rand
 
      
 
      
@@ -109,11 +109,11 @@
 
      
 
      Local Testing
      
 
      Method 1:
      
-
      paddle train --job=test \
-             --use_gpu=1/0 \ 
-             --config=network_config \
-             --trainer_count=COUNT \ 
-             --init_model_path=model_path \
+
      paddle train --job=test \
+             --use_gpu=1/0 \ 
+             --config=network_config \
+             --trainer_count=COUNT \ 
+             --init_model_path=model_path \
 
      
 
      
 
        
@@ -121,29 +121,29 @@
 
      • only can test one model.
        • 
 
      
 
      Method 2:
      
-
      paddle train --job=test \
-             --use_gpu=1/0 \ 
-             --config=network_config \
-             --trainer_count=COUNT \ 
-             --model_list=model.list \
+
      paddle train --job=test \
+             --use_gpu=1/0 \ 
+             --config=network_config \
+             --trainer_count=COUNT \ 
+             --model_list=model.list \
 
      
 
      
 
        
 
      • use model_list to specify test models
        • 
 
      • can test several models, where model.list likes:
        • 
 
      
-
      ./alexnet_pass1
-./alexnet_pass2
+
      ./alexnet_pass1
+./alexnet_pass2
 
      
 
      
 
      Method 3:
      
-
      paddle train --job=test \
-             --use_gpu=1/0 \
-             --config=network_config \
-             --trainer_count=COUNT \
-             --save_dir=model \
-             --test_pass=M \
-             --num_passes=N \
+
      paddle train --job=test \
+             --use_gpu=1/0 \
+             --config=network_config \
+             --trainer_count=COUNT \
+             --save_dir=model \
+             --test_pass=M \
+             --num_passes=N \
 
      
 
      
 
      This way must use model path saved by Paddle like this: model/pass-%5d. Testing model is from M-th pass to (N-1)-th pass. For example: M=12 and N=14 will test model/pass-00012 and model/pass-00013.
      
@@ -158,7 +158,7 @@
 
      
 
      2) cluster training
      
 
      Add the following argument for cluster training of a sparse model. At the same time you need to set sparse_remote_update=True in network config. Check the network config documentation for more details.
      
-
      --ports_num_for_sparse=1    #(default: 0)
+
      --ports_num_for_sparse=1    #(default: 0)
 
      
 
      
 
      
@@ -167,20 +167,20 @@
 
      parallel_nn
      
 
      parallel_nn can be set to mixed use of GPUs and CPUs to compute layers. That is to say, you can deploy network to use a GPU to compute some layers and use a CPU to compute other layers. The other way is to split layers into different GPUs, which can reduce GPU memory or use parallel computation to accelerate some layers.
      
 
      If you want to use these characteristics, you need to specify device ID in network config (denote it as deviceId) and add command line argument:
      
-
      --parallel_nn=true
+
      --parallel_nn=true
 
      
 
      
 
      
 
      case 1: Mixed Use of GPU and CPU
      
 
      Consider the following example:
      
-
      #command line:
-paddle train --use_gpu=true --parallel_nn=true trainer_count=COUNT
+
      #command line:
+paddle train --use_gpu=true --parallel_nn=true trainer_count=COUNT
 
-default_device(0)
+default_device(0)
 
-fc1=fc_layer(...)
-fc2=fc_layer(...)
-fc3=fc_layer(...,layer_attr=ExtraAttr(device=-1))
+fc1=fc_layer(...)
+fc2=fc_layer(...)
+fc3=fc_layer(...,layer_attr=ExtraAttr(device=-1))
 
      
 
      
 
        
@@ -195,13 +195,13 @@ fc3=fc_layer(...,layer_attr=ExtraAttr(device=-1))
 
      
 
      
 
      Case 2: Specify Layers in Different Devices
      
-
      #command line:
-paddle train --use_gpu=true --parallel_nn=true --trainer_count=COUNT
+
      #command line:
+paddle train --use_gpu=true --parallel_nn=true --trainer_count=COUNT
 
-#network:
-fc2=fc_layer(input=l1, layer_attr=ExtraAttr(device=0), ...)
-fc3=fc_layer(input=l1, layer_attr=ExtraAttr(device=1), ...)
-fc4=fc_layer(input=fc2, layer_attr=ExtraAttr(device=-1), ...)
+#network:
+fc2=fc_layer(input=l1, layer_attr=ExtraAttr(device=0), ...)
+fc3=fc_layer(input=l1, layer_attr=ExtraAttr(device=1), ...)
+fc4=fc_layer(input=fc2, layer_attr=ExtraAttr(device=-1), ...)
 
      
 
      
 
      In this case, we assume that there are 4 GPUs in one machine.
      
@@ -224,13 +224,13 @@ fc4=fc_layer(input=fc2, layer_attr=ExtraAttr(device=-1), ...)
 
 
    
 
    Allocation of device ID when device!=-1:
    
-
    (deviceId + gpu_id + threadId * numLogicalDevices_) % numDevices_
+
    (deviceId + gpu_id + threadId * numLogicalDevices_) % numDevices_
 
-deviceId:             specified in layer.
-gpu_id:               0 by default.
-threadId:             thread ID, range: 0,1,..., trainer_count-1
-numDevices_:          device (GPU) count in machine.
-numLogicalDevices_:   min(max(deviceId + 1), numDevices_)
+deviceId:             specified in layer.
+gpu_id:               0 by default.
+threadId:             thread ID, range: 0,1,..., trainer_count-1
+numDevices_:          device (GPU) count in machine.
+numLogicalDevices_:   min(max(deviceId + 1), numDevices_)
 
    
 
    
 
    
@@ -282,14 +282,11 @@ numLogicalDevices_:   min(max(deviceId + 1), numDevices_)
 
 
         
    
@@ -311,13 +308,13 @@ numLogicalDevices_:   min(max(deviceId + 1), numDevices_)
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/ui/data_provider/index.html b/doc/ui/data_provider/index.html
index ee97ea8626..11faaa60ca 100644
--- a/doc/ui/data_provider/index.html
+++ b/doc/ui/data_provider/index.html
@@ -6,7 +6,7 @@
   
     
     
-    PaddlePaddle DataProvider Introduction — PaddlePaddle  documentation
+    DataProvider Introduction — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    •  
       
     
      
 
@@ -55,8 +55,8 @@
         
    
           
    
             
-  
    
-
    PaddlePaddle DataProvider Introduction
    
+  
    
+
    DataProvider Introduction
    
 
    DataProvider is a module that loads training or testing data into cpu or gpu
 memory for the following triaining or testing process.
    
 
    For simple use, users can use Python PyDataProvider to dynamically reads
@@ -125,14 +125,11 @@ usages of DataProvider and how to implement a new DataProvider,
    
 
 
         
    
@@ -154,13 +151,13 @@ usages of DataProvider and how to implement a new DataProvider,
    
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/ui/data_provider/pydataprovider2.html b/doc/ui/data_provider/pydataprovider2.html
index 93992bf0de..f9c211d04f 100644
--- a/doc/ui/data_provider/pydataprovider2.html
+++ b/doc/ui/data_provider/pydataprovider2.html
@@ -6,7 +6,7 @@
   
     
     
-    How to use PyDataProvider2 — PaddlePaddle  documentation
+    How to use PyDataProvider2 — PaddlePaddle  documentation
     
     
     
@@ -25,9 +25,9 @@
     
     
     
-    
+    
     
-     
+     
   
   
       
 
@@ -74,19 +74,19 @@ grayscale images. Labels of the training sample range from 0 to 9. All the
 images have been size-normalized and centered into images with the same size
 of 28 x 28 pixels.
    
 
    A small part of the original data as an example is shown as below:
    
-
    5;0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.215686 0.533333 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.67451 0.992157 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.070588 0.886275 0.992157 0 0 0 0 0 0 0 0 0 0 0.192157 0.070588 0 0 0 0 0 0 0 0 0 0 0 0 0 0.670588 0.992157 0.992157 0 0 0 0 0 0 0 0 0 0.117647 0.933333 0.858824 0.313725 0 0 0 0 0 0 0 0 0 0 0 0.090196 0.858824 0.992157 0.831373 0 0 0 0 0 0 0 0 0 0.141176 0.992157 0.992157 0.611765 0.054902 0 0 0 0 0 0 0 0 0 0 0.258824 0.992157 0.992157 0.529412 0 0 0 0 0 0 0 0 0 0.368627 0.992157 0.992157 0.419608 0.003922 0 0 0 0 0 0 0 0 0 0.094118 0.835294 0.992157 0.992157 0.517647 0 0 0 0 0 0 0 0 0 0.603922 0.992157 0.992157 0.992157 0.603922 0.545098 0.043137 0 0 0 0 0 0 0 0.447059 0.992157 0.992157 0.956863 0.062745 0 0 0 0 0 0 0 0 0.011765 0.666667 0.992157 0.992157 0.992157 0.992157 0.992157 0.745098 0.137255 0 0 0 0 0 0.152941 0.866667 0.992157 0.992157 0.521569 0 0 0 0 0 0 0 0 0 0.070588 0.992157 0.992157 0.992157 0.803922 0.352941 0.745098 0.992157 0.945098 0.317647 0 0 0 0 0.580392 0.992157 0.992157 0.764706 0.043137 0 0 0 0 0 0 0 0 0 0.070588 0.992157 0.992157 0.776471 0.043137 0 0.007843 0.27451 0.882353 0.941176 0.176471 0 0 0.180392 0.898039 0.992157 0.992157 0.313725 0 0 0 0 0 0 0 0 0 0 0.070588 0.992157 0.992157 0.713725 0 0 0 0 0.627451 0.992157 0.729412 0.062745 0 0.509804 0.992157 0.992157 0.776471 0.035294 0 0 0 0 0 0 0 0 0 0 0.494118 0.992157 0.992157 0.968627 0.168627 0 0 0 0.423529 0.992157 0.992157 0.364706 0 0.717647 0.992157 0.992157 0.317647 0 0 0 0 0 0 0 0 0 0 0 0.533333 0.992157 0.984314 0.945098 0.603922 0 0 0 0.003922 0.466667 0.992157 0.988235 0.976471 0.992157 0.992157 0.788235 0.007843 0 0 0 0 0 0 0 0 0 0 0 0.686275 0.882353 0.364706 0 0 0 0 0 0 0.098039 0.588235 0.992157 0.992157 0.992157 0.980392 0.305882 0 0 0 0 0 0 0 0 0 0 0 0 0.101961 0.67451 0.321569 0 0 0 0 0 0 0 0.105882 0.733333 0.976471 0.811765 0.713725 0 0 0 0 0 0 0 0 0 0 0 0 0 0.65098 0.992157 0.321569 0 0 0 0 0 0 0 0 0 0.25098 0.007843 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0.94902 0.219608 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.968627 0.764706 0.152941 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.498039 0.25098 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0;
-0;0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.298039 0.333333 0.333333 0.333333 0.337255 0.333333 0.333333 0.109804 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.027451 0.223529 0.776471 0.964706 0.988235 0.988235 0.988235 0.992157 0.988235 0.988235 0.780392 0.098039 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.14902 0.698039 0.988235 0.992157 0.988235 0.901961 0.87451 0.568627 0.882353 0.976471 0.988235 0.988235 0.501961 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.188235 0.647059 0.988235 0.988235 0.745098 0.439216 0.098039 0 0 0 0.572549 0.988235 0.988235 0.988235 0 0 0 0 0 0 0 0 0 0 0 0 0 0.2 0.933333 0.992157 0.941176 0.247059 0 0 0 0 0 0 0.188235 0.898039 0.992157 0.992157 0 0 0 0 0 0 0 0 0 0 0 0.039216 0.639216 0.933333 0.988235 0.913725 0.278431 0 0 0 0 0 0 0 0.113725 0.843137 0.988235 0.988235 0 0 0 0 0 0 0 0 0 0 0 0.235294 0.988235 0.992157 0.988235 0.815686 0.07451 0 0 0 0 0 0 0 0.333333 0.988235 0.988235 0.552941 0 0 0 0 0 0 0 0 0 0 0.211765 0.878431 0.988235 0.992157 0.701961 0.329412 0.109804 0 0 0 0 0 0 0 0.698039 0.988235 0.913725 0.145098 0 0 0 0 0 0 0 0 0 0.188235 0.890196 0.988235 0.988235 0.745098 0.047059 0 0 0 0 0 0 0 0 0 0.882353 0.988235 0.568627 0 0 0 0 0 0 0 0 0 0.2 0.933333 0.992157 0.992157 0.992157 0.447059 0.294118 0 0 0 0 0 0 0 0 0.447059 0.992157 0.768627 0 0 0 0 0 0 0 0 0 0 0.623529 0.988235 0.988235 0.988235 0.988235 0.992157 0.47451 0 0 0 0 0 0 0 0.188235 0.933333 0.87451 0.509804 0 0 0 0 0 0 0 0 0 0 0.992157 0.988235 0.937255 0.792157 0.988235 0.894118 0.082353 0 0 0 0 0 0 0.027451 0.647059 0.992157 0.654902 0 0 0 0 0 0 0 0 0 0 0 0.623529 0.988235 0.913725 0.329412 0.376471 0.184314 0 0 0 0 0 0 0.027451 0.513725 0.988235 0.635294 0.219608 0 0 0 0 0 0 0 0 0 0 0 0.196078 0.929412 0.988235 0.988235 0.741176 0.309804 0 0 0 0 0 0 0.529412 0.988235 0.678431 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.223529 0.992157 0.992157 1 0.992157 0.992157 0.992157 0.992157 1 0.992157 0.992157 0.882353 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.023529 0.478431 0.654902 0.658824 0.952941 0.988235 0.988235 0.988235 0.992157 0.988235 0.729412 0.278431 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.196078 0.647059 0.764706 0.764706 0.768627 0.580392 0.047059 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0;
-4;0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.180392 0.470588 0.623529 0.623529 0.623529 0.588235 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.243137 0.494118 0.862745 0.870588 0.960784 0.996078 0.996078 0.996078 0.996078 0.992157 0.466667 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.317647 0.639216 0.639216 0.639216 0.639216 0.639216 0.470588 0.262745 0.333333 0.929412 0.694118 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.811765 0.694118 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.811765 0.694118 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.811765 0.694118 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.184314 0.992157 0.694118 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.192157 0.996078 0.384314 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.454902 0.980392 0.219608 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.564706 0.941176 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.588235 0.776471 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.945098 0.560784 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.054902 0.952941 0.356863 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.337255 0.917647 0.109804 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.698039 0.701961 0.019608 0.4 0.662745 0.662745 0.662745 0.662745 0.662745 0.662745 0.662745 0.376471 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.090196 0.639216 0.972549 0.945098 0.913725 0.996078 0.996078 0.996078 0.996078 1 0.996078 0.996078 1 0.996078 0 0 0 0 0 0 0 0 0 0 0.007843 0.105882 0.717647 0.776471 0.905882 0.996078 0.996078 0.988235 0.980392 0.862745 0.537255 0.223529 0.223529 0.368627 0.376471 0.6 0.6 0.6 0 0 0 0 0 0 0 0 0.262745 0.470588 0.6 0.996078 0.996078 0.996078 0.996078 0.847059 0.356863 0.156863 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.909804 0.705882 0.823529 0.635294 0.490196 0.219608 0.113725 0.062745 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.152941 0.152941 0.156863 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0;
+
    5;0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.215686 0.533333 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.67451 0.992157 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.070588 0.886275 0.992157 0 0 0 0 0 0 0 0 0 0 0.192157 0.070588 0 0 0 0 0 0 0 0 0 0 0 0 0 0.670588 0.992157 0.992157 0 0 0 0 0 0 0 0 0 0.117647 0.933333 0.858824 0.313725 0 0 0 0 0 0 0 0 0 0 0 0.090196 0.858824 0.992157 0.831373 0 0 0 0 0 0 0 0 0 0.141176 0.992157 0.992157 0.611765 0.054902 0 0 0 0 0 0 0 0 0 0 0.258824 0.992157 0.992157 0.529412 0 0 0 0 0 0 0 0 0 0.368627 0.992157 0.992157 0.419608 0.003922 0 0 0 0 0 0 0 0 0 0.094118 0.835294 0.992157 0.992157 0.517647 0 0 0 0 0 0 0 0 0 0.603922 0.992157 0.992157 0.992157 0.603922 0.545098 0.043137 0 0 0 0 0 0 0 0.447059 0.992157 0.992157 0.956863 0.062745 0 0 0 0 0 0 0 0 0.011765 0.666667 0.992157 0.992157 0.992157 0.992157 0.992157 0.745098 0.137255 0 0 0 0 0 0.152941 0.866667 0.992157 0.992157 0.521569 0 0 0 0 0 0 0 0 0 0.070588 0.992157 0.992157 0.992157 0.803922 0.352941 0.745098 0.992157 0.945098 0.317647 0 0 0 0 0.580392 0.992157 0.992157 0.764706 0.043137 0 0 0 0 0 0 0 0 0 0.070588 0.992157 0.992157 0.776471 0.043137 0 0.007843 0.27451 0.882353 0.941176 0.176471 0 0 0.180392 0.898039 0.992157 0.992157 0.313725 0 0 0 0 0 0 0 0 0 0 0.070588 0.992157 0.992157 0.713725 0 0 0 0 0.627451 0.992157 0.729412 0.062745 0 0.509804 0.992157 0.992157 0.776471 0.035294 0 0 0 0 0 0 0 0 0 0 0.494118 0.992157 0.992157 0.968627 0.168627 0 0 0 0.423529 0.992157 0.992157 0.364706 0 0.717647 0.992157 0.992157 0.317647 0 0 0 0 0 0 0 0 0 0 0 0.533333 0.992157 0.984314 0.945098 0.603922 0 0 0 0.003922 0.466667 0.992157 0.988235 0.976471 0.992157 0.992157 0.788235 0.007843 0 0 0 0 0 0 0 0 0 0 0 0.686275 0.882353 0.364706 0 0 0 0 0 0 0.098039 0.588235 0.992157 0.992157 0.992157 0.980392 0.305882 0 0 0 0 0 0 0 0 0 0 0 0 0.101961 0.67451 0.321569 0 0 0 0 0 0 0 0.105882 0.733333 0.976471 0.811765 0.713725 0 0 0 0 0 0 0 0 0 0 0 0 0 0.65098 0.992157 0.321569 0 0 0 0 0 0 0 0 0 0.25098 0.007843 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0.94902 0.219608 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.968627 0.764706 0.152941 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.498039 0.25098 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0;
+0;0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.298039 0.333333 0.333333 0.333333 0.337255 0.333333 0.333333 0.109804 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.027451 0.223529 0.776471 0.964706 0.988235 0.988235 0.988235 0.992157 0.988235 0.988235 0.780392 0.098039 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.14902 0.698039 0.988235 0.992157 0.988235 0.901961 0.87451 0.568627 0.882353 0.976471 0.988235 0.988235 0.501961 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.188235 0.647059 0.988235 0.988235 0.745098 0.439216 0.098039 0 0 0 0.572549 0.988235 0.988235 0.988235 0 0 0 0 0 0 0 0 0 0 0 0 0 0.2 0.933333 0.992157 0.941176 0.247059 0 0 0 0 0 0 0.188235 0.898039 0.992157 0.992157 0 0 0 0 0 0 0 0 0 0 0 0.039216 0.639216 0.933333 0.988235 0.913725 0.278431 0 0 0 0 0 0 0 0.113725 0.843137 0.988235 0.988235 0 0 0 0 0 0 0 0 0 0 0 0.235294 0.988235 0.992157 0.988235 0.815686 0.07451 0 0 0 0 0 0 0 0.333333 0.988235 0.988235 0.552941 0 0 0 0 0 0 0 0 0 0 0.211765 0.878431 0.988235 0.992157 0.701961 0.329412 0.109804 0 0 0 0 0 0 0 0.698039 0.988235 0.913725 0.145098 0 0 0 0 0 0 0 0 0 0.188235 0.890196 0.988235 0.988235 0.745098 0.047059 0 0 0 0 0 0 0 0 0 0.882353 0.988235 0.568627 0 0 0 0 0 0 0 0 0 0.2 0.933333 0.992157 0.992157 0.992157 0.447059 0.294118 0 0 0 0 0 0 0 0 0.447059 0.992157 0.768627 0 0 0 0 0 0 0 0 0 0 0.623529 0.988235 0.988235 0.988235 0.988235 0.992157 0.47451 0 0 0 0 0 0 0 0.188235 0.933333 0.87451 0.509804 0 0 0 0 0 0 0 0 0 0 0.992157 0.988235 0.937255 0.792157 0.988235 0.894118 0.082353 0 0 0 0 0 0 0.027451 0.647059 0.992157 0.654902 0 0 0 0 0 0 0 0 0 0 0 0.623529 0.988235 0.913725 0.329412 0.376471 0.184314 0 0 0 0 0 0 0.027451 0.513725 0.988235 0.635294 0.219608 0 0 0 0 0 0 0 0 0 0 0 0.196078 0.929412 0.988235 0.988235 0.741176 0.309804 0 0 0 0 0 0 0.529412 0.988235 0.678431 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.223529 0.992157 0.992157 1 0.992157 0.992157 0.992157 0.992157 1 0.992157 0.992157 0.882353 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.023529 0.478431 0.654902 0.658824 0.952941 0.988235 0.988235 0.988235 0.992157 0.988235 0.729412 0.278431 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.196078 0.647059 0.764706 0.764706 0.768627 0.580392 0.047059 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0;
+4;0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.180392 0.470588 0.623529 0.623529 0.623529 0.588235 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.243137 0.494118 0.862745 0.870588 0.960784 0.996078 0.996078 0.996078 0.996078 0.992157 0.466667 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.317647 0.639216 0.639216 0.639216 0.639216 0.639216 0.470588 0.262745 0.333333 0.929412 0.694118 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.811765 0.694118 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.811765 0.694118 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.811765 0.694118 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.184314 0.992157 0.694118 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.192157 0.996078 0.384314 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.454902 0.980392 0.219608 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.564706 0.941176 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.588235 0.776471 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.945098 0.560784 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.054902 0.952941 0.356863 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.337255 0.917647 0.109804 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.698039 0.701961 0.019608 0.4 0.662745 0.662745 0.662745 0.662745 0.662745 0.662745 0.662745 0.376471 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.090196 0.639216 0.972549 0.945098 0.913725 0.996078 0.996078 0.996078 0.996078 1 0.996078 0.996078 1 0.996078 0 0 0 0 0 0 0 0 0 0 0.007843 0.105882 0.717647 0.776471 0.905882 0.996078 0.996078 0.988235 0.980392 0.862745 0.537255 0.223529 0.223529 0.368627 0.376471 0.6 0.6 0.6 0 0 0 0 0 0 0 0 0.262745 0.470588 0.6 0.996078 0.996078 0.996078 0.996078 0.847059 0.356863 0.156863 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.909804 0.705882 0.823529 0.635294 0.490196 0.219608 0.113725 0.062745 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0.152941 0.152941 0.156863 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0;
 
    
 
    
 
    Each line of the data contains two parts, separated by ‘;’. The first part is
 label of an image. The second part contains 28x28 pixel float values.
    
 
    Just write path of the above data into train.list. It looks like this:
    
-
    mnist_train.txt
+
    mnist_train.txt
 
    
 
    
 
    The corresponding dataprovider is shown as below:
    
-
    from paddle.trainer.PyDataProvider2 import *
+
    from paddle.trainer.PyDataProvider2 import *
 
 
 # Define a py data provider
@@ -142,10 +142,10 @@ sample by using keywords y
 generator.
    
 
    Only a few lines of codes need to be added into the training configuration file,
 you can take this as an example.
    
-
    from paddle.trainer_config_helpers import *
+
    from paddle.trainer_config_helpers import *
 
 define_py_data_sources2(train_list='train.list',
-                        test_list=None,
+                        test_list=None,
                         module='mnist_provider',
                         obj='process')
 
    
@@ -176,13 +176,13 @@ tasks.
    
 The original input data are simple English text, labeled into positive or
 negative sentiment (marked by 0 and 1 respectively).
    
 
    A small part of the original data as an example can be found in the path below:
    
-
    0       I saw this movie at the AFI Dallas festival . It all takes place at a lake house and it looks wonderful .
-1       This documentary makes you travel all around the globe . It contains rare and stunning sequels from the wilderness .
-...
+
    0       I saw this movie at the AFI Dallas festival . It all takes place at a lake house and it looks wonderful .
+1       This documentary makes you travel all around the globe . It contains rare and stunning sequels from the wilderness .
+...
 
    
 
    
 
    The corresponding data provider can be found in the path below:
    
-
    from paddle.trainer.PyDataProvider2 import *
+
    from paddle.trainer.PyDataProvider2 import *
 
 
 def on_init(settings, dictionary, **kwargs):
@@ -244,12 +244,12 @@ configuration file, and it maps word string to word id.
 
 
    To pass these parameters into DataProvider, the following lines should be added
 into trainer configuration file.
    
-
    from paddle.trainer_config_helpers import *
+
    from paddle.trainer_config_helpers import *
 
 dictionary = dict()
 ...  #  read dictionary from outside
 
-define_py_data_sources2(train_list='train.list', test_list=None,
+define_py_data_sources2(train_list='train.list', test_list=None,
                         module='sentimental_provider', obj='process',
                         # above codes same as mnist sample.
                         args={  # pass to provider.
@@ -292,9 +292,9 @@ samples than pool_size. It is better to set True to avoid some deadlocks.
 usefull in sequential model, that defines batch size is counted upon sequence
 or token. By default, each sample or sequence counts to 1 when calculating
 batch size.
-
  • cache is a data cache strategy, see cache
    • 
+
  • cache is a data cache strategy, see cache.
    • 
 
  • Init_hook function is invoked once the data provider is initialized,
-see init_hook
    • 
+see init_hook.
 
 
    
 
    
@@ -302,19 +302,18 @@ see init_hook
 
    PaddlePaddle has four data types, and three sequence types.
 The four data types are:
    
 
      
-
    • dense_vector represents dense float vector.
      • 
-
    • sparse_binary_vector sparse binary vector, most of the value is 0, and
+
    • dense_vector: dense float vector.
      • 
+
    • sparse_binary_vector: sparse binary vector, most of the value is 0, and
 the non zero elements are fixed to 1.
      • 
-
    • sparse_float_vector sparse float vector, most of the value is 0, and some
-non zero elements that can be any float value. They are given by the user.
      • 
-
    • integer represents an integer scalar, that is especially used for label or
-word index.
      • 
+
    • sparse_float_vector: sparse float vector, most of the value is 0, and some
+non zero elements can be any float value. They are given by the user.
      • 
+
    • integer: an integer scalar, that is especially used for label or word index.
      • 
 
    
-
    The three sequence types are
    
+
    The three sequence types are:
    
 
      
-
    • SequenceType.NO_SEQUENCE means the sample is not a sequence
      • 
-
    • SequenceType.SEQUENCE means the sample is a sequence
      • 
-
    • SequenceType.SUB_SEQUENCE means it is a nested sequence, that each timestep of
+
    • SequenceType.NO_SEQUENCE means the sample is not a sequence.
      • 
+
    • SequenceType.SEQUENCE means the sample is a sequence.
      • 
+
    • SequenceType.SUB_SEQUENCE means it is a nested sequence, that each timestep of
 the input sequence is also a sequence.
      • 
 
    
 
    Different input type has a defferenct input format. Their formats are shown
@@ -362,23 +361,21 @@ in the above table.
    
 
    init_hook
    
 
    init_hook is a function that is invoked once the data provoder is initialized.
 Its parameters lists as follows:
    
-
      
-
    • The first parameter is a settings object, which is the same to :code:’settings’
-in process method.  The object contains several attributes, including:
-* settings.input_types the input types. Reference input_types
-* settings.logger a logging object
      
+
        
+
      • The first parameter is a settings object, which is the same to settings
+in process method. The object contains several attributes, including:
          
+
        • settings.input_types: the input types. Reference input_types.
          • 
+
        • settings.logger: a logging object.
          • 
+
        
 
        • 
-
      • The rest parameters are the key word arguments. It is made up of PaddpePaddle
-pre-defined parameters and user defined parameters.
-* PaddlePaddle defines parameters including:
        
-
        
-
          
-
        • is_train is a bool parameter that indicates the DataProvider is used in
-training or testing
          • 
-
        • file_list is the list of all files.
          • 
+
        • The rest parameters are the key word arguments. It is made up of PaddpePaddle
+pre-defined parameters and user defined parameters.
            
+
          • PaddlePaddle-defined parameters including:
              
+
            • is_train is a bool parameter that indicates the DataProvider is used in
+training or testing.
              • 
+
            • file_list is the list of all files.
              • 
 
            
-
        
-
          
+
 
        • User-defined parameters args can be set in training configuration.
          • 
 
        
 
        • 
@@ -389,12 +386,11 @@ parameters which your init_hook does not use.
        
 
    
 
    
 
    cache
    
-
    DataProvider provides two simple cache strategy. They are
-* CacheType.NO_CACHE means do not cache any data, then data is read at runtime by
    
-
    
-
    the user implemented python module every pass.
    
+
    DataProvider provides two simple cache strategy. They are:
    
 
      
-
    • CacheType.CACHE_PASS_IN_MEM means the first pass reads data by the user
+
    • CacheType.NO_CACHE means do not cache any data, then data is read at runtime by
+the user implemented python module every pass.
      • 
+
    • CacheType.CACHE_PASS_IN_MEM means the first pass reads data by the user
 implemented python module, and the rest passes will directly read data from
 memory.
      • 
 
    
@@ -426,7 +422,7 @@ memory.
 
   
    Previous topic
    
   
    PaddlePaddle DataProvider Introduction
    
+                        title="previous chapter">DataProvider Introduction
    
   
    Next topic
    
   
    Model Config Interface
    
@@ -440,14 +436,11 @@ memory.
 
 
         
    
@@ -467,16 +460,16 @@ memory.
           next |
         
  • 
-          previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    • 
-          
  • PaddlePaddle DataProvider Introduction »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    • 
+          
  • DataProvider Introduction »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/ui/index.html b/doc/ui/index.html
index 73fccf9c46..8a338acbce 100644
--- a/doc/ui/index.html
+++ b/doc/ui/index.html
@@ -6,7 +6,7 @@
   
     
     
-    User Interface — PaddlePaddle  documentation
+    User Interface — PaddlePaddle  documentation
     
     
     
@@ -25,7 +25,7 @@
     
     
     
-    
+    
      
   
   
@@ -39,12 +39,12 @@
           modules |
         
  • 
-          next |
    • 
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
      
 
@@ -68,7 +68,7 @@
 
    API Reference
    
 
    
 
 
    
 
    
@@ -114,7 +114,7 @@
                         title="previous chapter">Debian Package installation guide
    
   
    Next topic
    
   
    PaddlePaddle DataProvider Introduction
    
+                        title="next chapter">DataProvider Introduction
    
   
    
     
    This Page
    
     
      
@@ -125,14 +125,11 @@
 
 
         
    
@@ -149,17 +146,17 @@
           modules |
         
  • 
-          next |
    • 
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc/ui/predict/swig_py_paddle_en.html b/doc/ui/predict/swig_py_paddle_en.html
index 979a3090ca..4f928820f4 100644
--- a/doc/ui/predict/swig_py_paddle_en.html
+++ b/doc/ui/predict/swig_py_paddle_en.html
@@ -6,7 +6,7 @@
   
     
     
-    Python Prediction API — PaddlePaddle  documentation
+    Python Prediction API — PaddlePaddle  documentation
     
     
     
@@ -45,8 +45,8 @@
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    •  
       
     
      
 
@@ -66,207 +66,20 @@ SWIG. The main steps of predict values in python are:
    
 
  • Predict
    • 
 
 
    Here is a sample python script that shows the typical prediction process for the
-MNIST classification problem.
    
-
    
 
      1
-  2
-  3
-  4
-  5
-  6
-  7
-  8
-  9
- 10
- 11
- 12
- 13
- 14
- 15
- 16
- 17
- 18
- 19
- 20
- 21
- 22
- 23
- 24
- 25
- 26
- 27
- 28
- 29
- 30
- 31
- 32
- 33
- 34
- 35
- 36
- 37
- 38
- 39
- 40
- 41
- 42
- 43
- 44
- 45
- 46
- 47
- 48
- 49
- 50
- 51
- 52
- 53
- 54
- 55
- 56
- 57
- 58
- 59
- 60
- 61
- 62
- 63
- 64
- 65
- 66
- 67
- 68
- 69
- 70
- 71
- 72
- 73
- 74
- 75
- 76
- 77
- 78
- 79
- 80
- 81
- 82
- 83
- 84
- 85
- 86
- 87
- 88
- 89
- 90
- 91
- 92
- 93
- 94
- 95
- 96
- 97
- 98
- 99
-100
-101
-102
-103
-104
    # Copyright (c) 2016 Baidu, Inc. All Rights Reserved
-#
-# 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.
-
-from py_paddle import swig_paddle, DataProviderWrapperConverter
-from paddle.trainer.PyDataProviderWrapper import DenseSlot
+MNIST classification problem. A complete sample code could be found at
+src_root/doc/ui/predict/predict_sample.py.
    
+
    from py_paddle import swig_paddle, DataProviderConverter
+from paddle.trainer.PyDataProvider2 import dense_vector
 from paddle.trainer.config_parser import parse_config
 
-TEST_DATA = [[[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
-               0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
-               0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
-               0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.215686,
-               0.533333, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.67451,
-               0.992157, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.070588, 0.886275,
-               0.992157, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.192157, 0.070588, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
-               0.670588, 0.992157, 0.992157, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.117647, 0.933333, 0.858824, 0.313725, 0, 0, 0,
-               0, 0, 0, 0, 0, 0, 0, 0, 0.090196, 0.858824, 0.992157, 0.831373, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.141176,
-               0.992157, 0.992157, 0.611765, 0.054902, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.258824, 0.992157, 0.992157,
-               0.529412, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.368627, 0.992157, 0.992157, 0.419608, 0.003922, 0, 0, 0, 0, 0, 0,
-               0, 0, 0, 0.094118, 0.835294, 0.992157, 0.992157, 0.517647, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.603922, 0.992157,
-               0.992157, 0.992157, 0.603922, 0.545098, 0.043137, 0, 0, 0, 0, 0, 0, 0, 0.447059, 0.992157, 0.992157,
-               0.956863, 0.062745, 0, 0, 0, 0, 0, 0, 0, 0, 0.011765, 0.666667, 0.992157, 0.992157, 0.992157, 0.992157,
-               0.992157, 0.745098, 0.137255, 0, 0, 0, 0, 0, 0.152941, 0.866667, 0.992157, 0.992157, 0.521569, 0, 0, 0,
-               0, 0, 0, 0, 0, 0, 0.070588, 0.992157, 0.992157, 0.992157, 0.803922, 0.352941, 0.745098, 0.992157,
-               0.945098, 0.317647, 0, 0, 0, 0, 0.580392, 0.992157, 0.992157, 0.764706, 0.043137, 0, 0, 0, 0, 0, 0, 0, 0,
-               0, 0.070588, 0.992157, 0.992157, 0.776471, 0.043137, 0, 0.007843, 0.27451, 0.882353, 0.941176, 0.176471,
-               0, 0, 0.180392, 0.898039, 0.992157, 0.992157, 0.313725, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.070588, 0.992157,
-               0.992157, 0.713725, 0, 0, 0, 0, 0.627451, 0.992157, 0.729412, 0.062745, 0, 0.509804, 0.992157, 0.992157,
-               0.776471, 0.035294, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.494118, 0.992157, 0.992157, 0.968627, 0.168627, 0, 0,
-               0, 0.423529, 0.992157, 0.992157, 0.364706, 0, 0.717647, 0.992157, 0.992157, 0.317647, 0, 0, 0, 0, 0, 0,
-               0, 0, 0, 0, 0, 0.533333, 0.992157, 0.984314, 0.945098, 0.603922, 0, 0, 0, 0.003922, 0.466667, 0.992157,
-               0.988235, 0.976471, 0.992157, 0.992157, 0.788235, 0.007843, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.686275,
-               0.882353, 0.364706, 0, 0, 0, 0, 0, 0, 0.098039, 0.588235, 0.992157, 0.992157, 0.992157, 0.980392,
-               0.305882, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.101961, 0.67451, 0.321569, 0, 0, 0, 0, 0, 0, 0, 0.105882,
-               0.733333, 0.976471, 0.811765, 0.713725, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.65098, 0.992157,
-               0.321569, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.25098, 0.007843, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1,
-               0.94902, 0.219608, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.968627,
-               0.764706, 0.152941, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.498039,
-               0.25098, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
-               0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
-               0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
-               0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]], [
-                 [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
-                  0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
-                  0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
-                  0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
-                  0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
-                  0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.298039, 0.333333, 0.333333, 0.333333, 0.337255, 0.333333,
-                  0.333333, 0.109804, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.027451, 0.223529, 0.776471,
-                  0.964706, 0.988235, 0.988235, 0.988235, 0.992157, 0.988235, 0.988235, 0.780392, 0.098039, 0, 0, 0, 0,
-                  0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.14902, 0.698039, 0.988235, 0.992157, 0.988235, 0.901961, 0.87451,
-                  0.568627, 0.882353, 0.976471, 0.988235, 0.988235, 0.501961, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
-                  0.188235, 0.647059, 0.988235, 0.988235, 0.745098, 0.439216, 0.098039, 0, 0, 0, 0.572549, 0.988235,
-                  0.988235, 0.988235, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.2, 0.933333, 0.992157, 0.941176,
-                  0.247059, 0, 0, 0, 0, 0, 0, 0.188235, 0.898039, 0.992157, 0.992157, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
-                  0.039216, 0.639216, 0.933333, 0.988235, 0.913725, 0.278431, 0, 0, 0, 0, 0, 0, 0, 0.113725, 0.843137,
-                  0.988235, 0.988235, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.235294, 0.988235, 0.992157, 0.988235, 0.815686,
-                  0.07451, 0, 0, 0, 0, 0, 0, 0, 0.333333, 0.988235, 0.988235, 0.552941, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
-                  0.211765, 0.878431, 0.988235, 0.992157, 0.701961, 0.329412, 0.109804, 0, 0, 0, 0, 0, 0, 0, 0.698039,
-                  0.988235, 0.913725, 0.145098, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.188235, 0.890196, 0.988235, 0.988235,
-                  0.745098, 0.047059, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.882353, 0.988235, 0.568627, 0, 0, 0, 0, 0, 0, 0, 0,
-                  0, 0.2, 0.933333, 0.992157, 0.992157, 0.992157, 0.447059, 0.294118, 0, 0, 0, 0, 0, 0, 0, 0, 0.447059,
-                  0.992157, 0.768627, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.623529, 0.988235, 0.988235, 0.988235, 0.988235,
-                  0.992157, 0.47451, 0, 0, 0, 0, 0, 0, 0, 0.188235, 0.933333, 0.87451, 0.509804, 0, 0, 0, 0, 0, 0, 0, 0,
-                  0, 0, 0.992157, 0.988235, 0.937255, 0.792157, 0.988235, 0.894118, 0.082353, 0, 0, 0, 0, 0, 0,
-                  0.027451, 0.647059, 0.992157, 0.654902, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.623529, 0.988235, 0.913725,
-                  0.329412, 0.376471, 0.184314, 0, 0, 0, 0, 0, 0, 0.027451, 0.513725, 0.988235, 0.635294, 0.219608, 0,
-                  0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.196078, 0.929412, 0.988235, 0.988235, 0.741176, 0.309804, 0, 0, 0, 0,
-                  0, 0, 0.529412, 0.988235, 0.678431, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.223529, 0.992157,
-                  0.992157, 1, 0.992157, 0.992157, 0.992157, 0.992157, 1, 0.992157, 0.992157, 0.882353, 0, 0, 0, 0, 0,
-                  0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0.023529, 0.478431, 0.654902, 0.658824, 0.952941, 0.988235, 0.988235,
-                  0.988235, 0.992157, 0.988235, 0.729412, 0.278431, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
-                  0, 0, 0, 0.196078, 0.647059, 0.764706, 0.764706, 0.768627, 0.580392, 0.047059, 0, 0, 0, 0, 0, 0, 0, 0,
-                  0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
-                  0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
-                  0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
-                  0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
-                  0, 0, 0, 0, 0, 0, 0]]]
-
 
 def main():
-    conf = parse_config("./mnist_model/trainer_config.conf.norm", "")
+    conf = parse_config("./mnist_model/trainer_config.py", "")
     print conf.data_config.load_data_args
     network = swig_paddle.GradientMachine.createFromConfigProto(conf.model_config)
     assert isinstance(network, swig_paddle.GradientMachine)  # For code hint.
     network.loadParameters("./mnist_model/")
-    converter = DataProviderWrapperConverter(False, [DenseSlot(784)])
+    converter = DataProviderConverter([dense_vector(784)])
     inArg = converter(TEST_DATA)
     print network.forwardTest(inArg)
 
@@ -275,31 +88,38 @@ MNIST classification problem.
    
     swig_paddle.initPaddle("--use_gpu=0")
     main()
 
    
-
    
+
    
 
    The module that does the most of the job is py_paddle.swig_paddle, it’s
 generated by SWIG and has complete documents, for more details you can use
 python’s help() function. Let’s walk through the above python script:
    
 
      
-
    • At the beginning, initialize PaddlePaddle with command line arguments(line 90).
      
+
    • At the beginning, use swig_paddle.initPaddle() to initialize
+PaddlePaddle with command line arguments, for more about command line arguments
+see Command Line Arguments.
      
 
      • 
-
    • Parse the configuration file that is used in training(line 93).
      
+
    • Parse the configuration file that is used in training with parse_config().
+Because data to predict with always have no label, and output of prediction work
+normally is the output layer rather than the cost layer, so you should modify
+the configuration file accordingly before using it in the prediction work.
      
 
      • 
-
    • Create a neural network at line 95 according the parsed configuration, then
-load the trained parameters from model at line 97.
      
+
    • Create a neural network with
+swig_paddle.GradientMachine.createFromConfigproto(), which takes the
+parsed configuration conf.model_config as argument. Then load the
+trained parameters from the model with network.loadParameters().
      
 
      • 
 
    • 
-
      A utility class for data transformation is created at line 98.
      
+
      Create a data converter object of utility class DataProviderConverter.
      
 
        
 
      • Note: As swig_paddle can only accept C++ matrices, we offer a utility
-class DataProviderWraaperConverter that can accept the same input data with
-PyDataProviderWrapper, for more information please refer to document
-of PyDataProviderWrapper.
        • 
+class DataProviderConverter that can accept the same input data with
+PyDataProvider2, for more information please refer to document
+of PyDataProvider2.
 
      
 
      
 
      
 
      • 
-
    • Do the prediction and output the result at line 100, forwardTest is another
-utility class that directly takes the activations of the output layer.
      
+
    • Do the prediction with forwardTest(), which takes the converted
+input data and outputs the activations of the output layer.
      
 
      • 
 
    
 
    Here is a typical output:
    
@@ -340,14 +160,11 @@ the corresponding neuron in the output layer.
    
 
 
         
    
@@ -369,13 +186,13 @@ the corresponding neuron in the output layer.
    
         
  • 
           previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • User Interface »
    •  
+        
  • PaddlePaddle  documentation »
    • 
+          
  • User Interface »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc_cn/.buildinfo b/doc_cn/.buildinfo
index 6b7cc993be..2343be214c 100644
--- a/doc_cn/.buildinfo
+++ b/doc_cn/.buildinfo
@@ -1,4 +1,4 @@
 # Sphinx build info version 1
 # This file hashes the configuration used when building these files. When it is not found, a full rebuild will be done.
-config: 5bb206a2182263ffcb7c4270c50bc7c9
+config: 68487318b6dbd705fd569c9b901c758d
 tags: 645f666f9bcd5a90fca523b33c5a78b7
diff --git a/doc_cn/_sources/algorithm/rnn/rnn.txt b/doc_cn/_sources/algorithm/rnn/rnn.txt
deleted file mode 100644
index f073ac4e20..0000000000
--- a/doc_cn/_sources/algorithm/rnn/rnn.txt
+++ /dev/null
@@ -1,7 +0,0 @@
-RNN 配置
-========
-
-.. toctree::
-  :maxdepth: 3
-
-* `RNN配置 <../../../doc/algorithm/rnn/rnn.html>`_
diff --git a/doc_cn/_sources/build_and_install/index.txt b/doc_cn/_sources/build_and_install/index.txt
index 67d85eca9b..e21fc98c63 100644
--- a/doc_cn/_sources/build_and_install/index.txt
+++ b/doc_cn/_sources/build_and_install/index.txt
@@ -1,8 +1,19 @@
 编译与安装
 ========================
 
-..  toctree::
-    :maxdepth: 1
-    
-    install/index.rst
-    cmake/index.rst
+PaddlePaddle提供数个预编译的二进制来进行安装，包括Docker镜像，ubuntu的deb安装包等。我们推荐使用Docker镜像来部署环境，同时欢迎贡献更多的安装包。
+
+Note: The intallation packages are still in pre-release state and your experience of installation may not be smooth.
+
+注意：目前PaddlePaddle的安装包还处在pre-release的状态，使用起来或许会不是很顺畅。
+
+.. toctree::
+   :maxdepth: 1
+   :glob:
+   
+   源码下载(对内) <../build/internal/download_paddle_source_zh_cn.rst>
+   使用Jumbo安装(对内) <../build/internal/install_from_jumbo.rst>
+   从源码编译安装(对内)  <../build/internal/build_from_source_zh_cn.rst>
+   install/docker_install.rst 
+   install/ubuntu_install.rst
+   cmake/index.rst
diff --git a/doc_cn/_sources/build_and_install/install/index.txt b/doc_cn/_sources/build_and_install/install/index.txt
deleted file mode 100644
index ce463728c7..0000000000
--- a/doc_cn/_sources/build_and_install/install/index.txt
+++ /dev/null
@@ -1,15 +0,0 @@
-安装PaddlePaddle
-==========
-
-PaddlePaddle提供数个预编译的二进制来进行安装。他们包括Docker镜像，ubuntu的deb安装包等
-。欢迎贡献更多的安装包。我们更推荐使用Docker镜像来部署PaddlePaddle环境。
-
-Note: The intallation packages are still in pre-release
-state and your experience of installation may not be smooth.
-
-注意!目前PaddlePaddle的安装包还处在pre-release的状态，
-使用起来或许会不是很顺畅。
-
-..	toctree::
-	docker_install.rst
-	ubuntu_install.rst
diff --git a/doc_cn/_sources/cluster/index.txt b/doc_cn/_sources/cluster/index.txt
index 16c1f0e37b..25313a9635 100644
--- a/doc_cn/_sources/cluster/index.txt
+++ b/doc_cn/_sources/cluster/index.txt
@@ -1,4 +1,11 @@
 集群训练
 ========
 
-参见 `集群训练 <../../doc/cluster/index.html>`_
+* `集群训练 <../../doc/cluster/index.html>`_
+
+.. toctree::
+    :maxdepth: 2
+    :glob:
+
+    集群训练(对内) 
+
diff --git a/doc_cn/_sources/demo/embedding_model/index.txt b/doc_cn/_sources/demo/embedding_model/index.txt
deleted file mode 100644
index 5894a4de5a..0000000000
--- a/doc_cn/_sources/demo/embedding_model/index.txt
+++ /dev/null
@@ -1 +0,0 @@
-# Embedding Demo
diff --git a/doc_cn/_sources/demo/image_classification/index.txt b/doc_cn/_sources/demo/image_classification/index.txt
deleted file mode 100644
index 98cbdc29b9..0000000000
--- a/doc_cn/_sources/demo/image_classification/index.txt
+++ /dev/null
@@ -1,4 +0,0 @@
-图片分类教程
-============
-
-TBD
diff --git a/doc_cn/_sources/demo/imagenet_model/index.txt b/doc_cn/_sources/demo/imagenet_model/index.txt
deleted file mode 100644
index b54b28401c..0000000000
--- a/doc_cn/_sources/demo/imagenet_model/index.txt
+++ /dev/null
@@ -1,2 +0,0 @@
-# Resnet
- TBD
diff --git a/doc_cn/_sources/demo/index.txt b/doc_cn/_sources/demo/index.txt
index 4c948dadae..71f54bc18f 100644
--- a/doc_cn/_sources/demo/index.txt
+++ b/doc_cn/_sources/demo/index.txt
@@ -4,23 +4,23 @@
 图像
 ''''
 
-* `图像分类 `_
+* `图像分类 <../../doc/demo/image_classification/index.html>`_
 
 自然语言处理
 ''''''''''''
 
-* `情感分析 `_
-* `文本生成 `_
-* `词性标注 `_
+* `情感分析 <../../doc/demo/sentiment_analysis/index.html>`_
+* `文本生成 <../../doc/demo/text_generation/index.html>`_
+* `词性标注 <../../doc/demo/semantic_role_labeling/index.html>`_
 
 推荐
 ''''
 
-* `MovieLens数据集 `_
-* `MovieLens评分回归 `_
+* `MovieLens数据集 <../../doc/demo/rec/ml_dataset.html>`_
+* `MovieLens评分回归 <../../doc/demo/rec/ml_regression.html>`_
 
 常用模型
 ''''''''
 
-* `ImageNet: ResNet `_
-* `Embedding: Chinese Word `_
+* `ImageNet: ResNet <../../doc/demo/imagenet_model/resnet_model.html>`_
+* `Embedding: Chinese Word <../../doc/demo/embedding_model/index.html>`_
diff --git a/doc_cn/_sources/demo/quick_start/index.txt b/doc_cn/_sources/demo/quick_start/index.txt
index b1de49068d..34cd4a840e 100644
--- a/doc_cn/_sources/demo/quick_start/index.txt
+++ b/doc_cn/_sources/demo/quick_start/index.txt
@@ -32,7 +32,7 @@
 
 ## 数据格式准备(Data Preparation)
 在本问题中，我们使用[Amazon电子产品评论数据](http://jmcauley.ucsd.edu/data/amazon/)，
-将评论分为好评(正样本)和差评(负样本)两类。`demo/quick_start`里提供了数据下载脚本
+将评论分为好评(正样本)和差评(负样本)两类。[源码](https://github.com/baidu/Paddle)的`demo/quick_start`里提供了数据下载脚本
 和预处理脚本。
 
 ```bash
@@ -144,7 +144,7 @@ PyDataProviderWrapper。
 
 我们将以基本的逻辑回归网络作为起点，并逐渐展示更加深入的功能。更详细的网络配置
 连接请参考Layer文档。
-所有配置在`demo/quick_start`目录，首先列举逻辑回归网络。
+所有配置在[源码](https://github.com/baidu/Paddle)`demo/quick_start`目录，首先列举逻辑回归网络。
 
 ### 逻辑回归模型(Logistic Regression)
 
@@ -407,7 +407,7 @@ paddle train \
 
 mv rank-00000 result.txt
 ```
-与训练网络配置不同的是：无需label相关的层，指定outputs输出概率层(softmax输出)，
+这里以`output/pass-00003`为例进行预测，用户可以根据训练log选择test结果最好的模型来预测。与训练网络配置不同的是：无需label相关的层，指定outputs输出概率层(softmax输出)，
 指定batch_size=1，数据传输无需label数据，预测数据指定test_list的位置。
 
 预测结果以文本的形式保存在`result.txt`中，一行为一个样本，格式如下：
diff --git a/doc_cn/_sources/demo/semantic_role_labeling/index.txt b/doc_cn/_sources/demo/semantic_role_labeling/index.txt
deleted file mode 100644
index a1594577bb..0000000000
--- a/doc_cn/_sources/demo/semantic_role_labeling/index.txt
+++ /dev/null
@@ -1,2 +0,0 @@
-# 语义标注
-TBD
diff --git a/doc_cn/_sources/demo/sentiment_analysis/index.txt b/doc_cn/_sources/demo/sentiment_analysis/index.txt
deleted file mode 100644
index d95f2803a4..0000000000
--- a/doc_cn/_sources/demo/sentiment_analysis/index.txt
+++ /dev/null
@@ -1,2 +0,0 @@
-# 情感分析
-TBD
diff --git a/doc_cn/_sources/demo/text_generation/index.txt b/doc_cn/_sources/demo/text_generation/index.txt
deleted file mode 100644
index 147b776465..0000000000
--- a/doc_cn/_sources/demo/text_generation/index.txt
+++ /dev/null
@@ -1,3 +0,0 @@
-文本生成
-========
-TBD
diff --git a/doc_cn/_sources/dev/new_layer/index.txt b/doc_cn/_sources/dev/new_layer/index.txt
deleted file mode 100644
index aafeceff5b..0000000000
--- a/doc_cn/_sources/dev/new_layer/index.txt
+++ /dev/null
@@ -1,4 +0,0 @@
-新写Layer
-=========
-
-* `新写Layer <../../../doc/dev/new_layer/index.html>`_
diff --git a/doc_cn/_sources/index.txt b/doc_cn/_sources/index.txt
index f21f60e146..6cf5588b5b 100644
--- a/doc_cn/_sources/index.txt
+++ b/doc_cn/_sources/index.txt
@@ -3,17 +3,17 @@ PaddlePaddle文档
 
 使用指南
 --------
-* [快速入门](demo/quick_start/index.md)
-* [编译与安装](build_and_install/index.rst)
-* [用户接口](ui/index.rst)
-* [使用示例](demo/index.rst)
-* [模型配置](ui/model.rst)
-* [集群训练](cluster/index.rst)
+* `快速入门 `_
+* `编译与安装 `_
+* `用户接口 `_
+* `使用示例 `_
+* `模型配置 <../doc/ui/api/trainer_config_helpers/index.html>`_
+* `集群训练 `_
 
 开发指南
 --------
-* [新写Layer](dev/new_layer/index.rst)
+* `新写Layer <../doc/dev/new_layer/index.html>`_
 
 算法教程
 --------
-* [RNN配置](algorithm/rnn/rnn.rst)
+* `RNN配置 <../doc/algorithm/rnn/rnn.html>`_
diff --git a/doc_cn/_sources/ui/data_provider/index.txt b/doc_cn/_sources/ui/data_provider/index.txt
index 681a131b66..ec8f8e5dc5 100644
--- a/doc_cn/_sources/ui/data_provider/index.txt
+++ b/doc_cn/_sources/ui/data_provider/index.txt
@@ -1,24 +1,15 @@
 PaddlePaddle的数据提供(DataProvider)介绍
-==================================
+========================================
 
-数据提供(DataProvider，后用DataProvider代替)是PaddlePaddle负责提供数据的模块。其作用是将训练数据
-传入内存或者显存，让神经网络可以进行训练。简单的使用，用户可以使用Python的
-:code:`PyDataProvider` 来自定义传数据的过程。如果有更复杂的使用，或者需要更高的效率，
-用户也可以在C++端自定义一个 :code:`DataProvider` 。
+数据提供(DataProvider)是PaddlePaddle负责提供数据的模块。其作用是将训练数据传入内存或者显存，让神经网络可以进行训练。简单的使用，用户可以使用Python的 :code:`PyDataProvider` 来自定义传数据的过程。如果有更复杂的使用，或者需要更高的效率，用户也可以在C++端自定义一个 :code:`DataProvider` 。
 
-PaddlePaddle需要用户在网络配置(trainer_config.py)中定义使用什么DataProvider，和DataProvider
-的一些参数，训练文件列表(train.list)和测试文件列表(test.list)。
+PaddlePaddle需要用户在网络配置(trainer_config.py)中定义使用哪种DataProvider及其参数，训练文件列表(train.list)和测试文件列表(test.list)。
 
-其中，train.list和test.list均为本地的两个文件(推荐直接放置到训练目录，以相对路径引用)。如果
-test.list不设置，或者设置为None的话，那么在训练过程中，不会执行测试操作。否则，则会根据命令行
-参数指定的测试方式，在训练过程中进行测试，从而防止过拟合。
+其中，train.list和test.list均为本地的两个文件(推荐直接放置到训练目录，以相对路径引用)。如果test.list不设置，或者设置为None，那么在训练过程中，不会执行测试操作。否则，会根据命令行参数指定的测试方式，在训练过程中进行测试，从而防止过拟合。
 
-一般情况下，train.list和test.list为纯文本文件，其每一行对应这每一个数据文件。数据文件存放在
-本地磁盘中，将文件的绝对路径或相对路径(相对于PaddlePaddle程序运行时的路径)的方式写在train.list和
-test.list中。当然，train.list和test.list也可以放置hdfs文件路径，或者数据库连接地址等等。
-用户在DataProvider中需要实现如何访问其中每一个文件。
+一般情况下，train.list和test.list为纯文本文件，一行对应一个数据文件，数据文件存放在本地磁盘中。将文件的绝对路径或相对路径(相对于PaddlePaddle程序运行时的路径)写在train.list和test.list中。当然，train.list和test.list也可以放置hdfs文件路径，或者数据库连接地址等等。
 
-DataProvider的具体用法和如何实现一个新的DataProvider，请参考下述文章:
+用户在DataProvider中需要实现如何访问其中每一个文件。DataProvider的具体用法和如何实现一个新的DataProvider，请参考下述文章:
 
 ..	toctree::
 
diff --git a/doc_cn/_sources/ui/data_provider/pydataprovider2.txt b/doc_cn/_sources/ui/data_provider/pydataprovider2.txt
index 766f583538..e743e41688 100644
--- a/doc_cn/_sources/ui/data_provider/pydataprovider2.txt
+++ b/doc_cn/_sources/ui/data_provider/pydataprovider2.txt
@@ -116,8 +116,6 @@ DataProvider创建的时候执行。这个初始化函数具有如下参数:
 参考(Reference)
 ---------------
 
-..  _@provider::
-
 @provider
 +++++++++
 
@@ -134,9 +132,6 @@ DataProvider创建的时候执行。这个初始化函数具有如下参数:
 *  cache 是数据缓存的策略，参考 `cache`_
 *  init_hook 是初始化时调用的函数，参考 `init_hook`_
 
-
-..  _input_types::
-
 input_types
 +++++++++++
 
@@ -169,16 +164,11 @@ PaddlePaddle的数据包括四种主要类型，和三种序列模式。其中
 
 其中，f代表一个浮点数，i代表一个整数。
 
-..  _init_hook::
-..  _settings::
-
 init_hook
 +++++++++
 
 init_hook可以传入一个函数。这个函数在初始化的时候会被调用。这个函数的参数是:
 
-
-
 * 第一个参数是 settings 对象。这个对象和process的第一个参数一致。具有的属性有
     * settings.input_types 设置输入类型。参考 `input_types`_
     * settings.logger 一个logging对象
@@ -192,8 +182,6 @@ init_hook可以传入一个函数。这个函数在初始化的时候会被调
 注意，PaddlePaddle保留添加参数的权力，所以init_hook尽量使用 :code:`**kwargs` , 来接受不使用的
 函数来保证兼容性。
 
-..  _cache::
-
 cache
 +++++
 
diff --git a/doc_cn/_sources/ui/model.txt b/doc_cn/_sources/ui/model.txt
deleted file mode 100644
index 7a81236d6f..0000000000
--- a/doc_cn/_sources/ui/model.txt
+++ /dev/null
@@ -1,4 +0,0 @@
-模型配置
-========
-
-* `Model Config Interface <../../doc/ui/api/trainer_config_helpers/index.html>`_
diff --git a/doc_cn/_sources/ui/predict/swig_py_paddle.txt b/doc_cn/_sources/ui/predict/swig_py_paddle.txt
index 284c60686d..012ac4ff6e 100644
--- a/doc_cn/_sources/ui/predict/swig_py_paddle.txt
+++ b/doc_cn/_sources/ui/predict/swig_py_paddle.txt
@@ -9,22 +9,30 @@ PaddlePaddle目前使用Swig对其常用的预测接口进行了封装，使在P
 * 准备数据
 * 预测
 
-典型的预测代码如下，使用mnist手写识别作为样例。
+典型的预测代码如下，使用mnist手写识别作为样例, 完整代码见
+:code:`src_root/doc/ui/predict/predict_sample.py` 。
 
 ..  literalinclude:: ../../../doc/ui/predict/predict_sample.py
     :language: python
-    :linenos:
-
-主要的软件包为py_paddle.swig_paddle，这个软件包文档相对完善。可以使用python的 :code:`help()` 函数查询文档。主要步骤为:
-
-* 在程序开始阶段，使用命令行参数初始化PaddlePaddle
-* 在98行载入PaddlePaddle的训练文件。读取config
-* 在100行创建神经网络，并在83行载入参数。
-* 103行创建一个从工具类，用来转换数据。
+    :lines: 15-18,90-100,101-104
+
+主要的软件包为py_paddle.swig_paddle，这个软件包文档相对完善。可以使用python的
+:code:`help()` 函数查询文档。主要步骤为:
+
+* 在程序开始阶段，使用 :code:`swig_paddle.initPaddle()` 传入命令行参数初始化
+  PaddlePaddle。详细的命令行参数请参考
+  `命令行参数 <../cmd_argument/detail_introduction.html>`_ 。
+* 接下来使用 :code:`parse_config()` 解析训练时的配置文件。这里要注意预测数据通常
+  不包含label, 而且预测网络通常直接输出最后一层的结果而不是像训练时一样以cost
+  layer作为输出，所以用于预测的配置文件要做相应的修改。
+* 使用 :code:`swig_paddle.GradientMachine.createFromConfigproto()` 根据上一步解
+  析好的配置创建神经网络。
+* 创建一个 :code:`DataProviderConverter` 对象converter。
     - swig_paddle接受的原始数据是C++的Matrix，也就是直接写内存的float数组。
-    - 这个接口并不用户友好。所以，我们提供了一个工具类DataProviderWrapperConverter.
-    - 这个工具类接收和PyDataProviderWrapper一样的输入数据，请参考PyDataProviderWrapper的文档。
-* 在第105行执行预测。forwardTest是一个工具类，直接提取出神经网络Output层的输出结果。典型的输出结果为\:
+      这个接口并不用户友好。所以，我们提供了一个工具类DataProviderConverter。
+      这个工具类接收和PyDataProvider2一样的输入数据，详情请参考
+      `PyDataProvider2文档 <../../../doc/ui/data_provider/pydataprovider2.html>`_ 。
+* 最后使用 :code:`forwardTest()` 直接提取出神经网络Output层的输出结果。典型的输出结果为\:
 
 ..  code-block:: text
 
@@ -37,4 +45,4 @@ PaddlePaddle目前使用Swig对其常用的预测接口进行了封装，使在P
           2.70634608e-08,   3.48565123e-08,   5.25639710e-09,
           4.48684503e-08]], dtype=float32)}]
 
-其中，value即为softmax层的输出。由于数据是两个，所以输出的value。
+其中，value即为softmax层的输出。由于数据是两条，所以输出的value包含两个向量 。
diff --git a/doc_cn/_static/basic.css b/doc_cn/_static/basic.css
index c89fc7e920..2b513f0c96 100644
--- a/doc_cn/_static/basic.css
+++ b/doc_cn/_static/basic.css
@@ -52,6 +52,8 @@ div.sphinxsidebar {
     width: 230px;
     margin-left: -100%;
     font-size: 90%;
+    word-wrap: break-word;
+    overflow-wrap : break-word;
 }
 
 div.sphinxsidebar ul {
@@ -83,10 +85,6 @@ div.sphinxsidebar #searchbox input[type="text"] {
     width: 170px;
 }
 
-div.sphinxsidebar #searchbox input[type="submit"] {
-    width: 30px;
-}
-
 img {
     border: 0;
     max-width: 100%;
@@ -187,6 +185,13 @@ div.genindex-jumpbox {
 
 /* -- general body styles --------------------------------------------------- */
 
+div.body p, div.body dd, div.body li, div.body blockquote {
+    -moz-hyphens: auto;
+    -ms-hyphens: auto;
+    -webkit-hyphens: auto;
+    hyphens: auto;
+}
+
 a.headerlink {
     visibility: hidden;
 }
diff --git a/doc_cn/_static/doctools.js b/doc_cn/_static/doctools.js
index e2e70cc287..8163495635 100644
--- a/doc_cn/_static/doctools.js
+++ b/doc_cn/_static/doctools.js
@@ -124,6 +124,7 @@ var Documentation = {
     this.fixFirefoxAnchorBug();
     this.highlightSearchWords();
     this.initIndexTable();
+    
   },
 
   /**
@@ -252,6 +253,29 @@ var Documentation = {
     });
     var url = parts.join('/');
     return path.substring(url.lastIndexOf('/') + 1, path.length - 1);
+  },
+
+  initOnKeyListeners: function() {
+    $(document).keyup(function(event) {
+      var activeElementType = document.activeElement.tagName;
+      // don't navigate when in search box or textarea
+      if (activeElementType !== 'TEXTAREA' && activeElementType !== 'INPUT' && activeElementType !== 'SELECT') {
+        switch (event.keyCode) {
+          case 37: // left
+            var prevHref = $('link[rel="prev"]').prop('href');
+            if (prevHref) {
+              window.location.href = prevHref;
+              return false;
+            }
+          case 39: // right
+            var nextHref = $('link[rel="next"]').prop('href');
+            if (nextHref) {
+              window.location.href = nextHref;
+              return false;
+            }
+        }
+      }
+    });
   }
 };
 
@@ -260,4 +284,4 @@ _ = Documentation.gettext;
 
 $(document).ready(function() {
   Documentation.init();
-});
+});
\ No newline at end of file
diff --git a/doc_cn/_static/searchtools.js b/doc_cn/_static/searchtools.js
index cb7446728a..066857ce21 100644
--- a/doc_cn/_static/searchtools.js
+++ b/doc_cn/_static/searchtools.js
@@ -2,7 +2,7 @@
  * searchtools.js_t
  * ~~~~~~~~~~~~~~~~
  *
- * Sphinx JavaScript utilties for the full-text search.
+ * Sphinx JavaScript utilities for the full-text search.
  *
  * :copyright: Copyright 2007-2016 by the Sphinx team, see AUTHORS.
  * :license: BSD, see LICENSE for details.
@@ -623,7 +623,7 @@ var Search = {
    * helper function to return a node containing the
    * search summary for a given text. keywords is a list
    * of stemmed words, hlwords is the list of normal, unstemmed
-   * words. the first one is used to find the occurance, the
+   * words. the first one is used to find the occurrence, the
    * latter for highlighting it.
    */
   makeSearchSummary : function(text, keywords, hlwords) {
diff --git a/doc_cn/_static/websupport.js b/doc_cn/_static/websupport.js
index ffd9b2bfdc..98e7f40b63 100644
--- a/doc_cn/_static/websupport.js
+++ b/doc_cn/_static/websupport.js
@@ -2,7 +2,7 @@
  * websupport.js
  * ~~~~~~~~~~~~~
  *
- * sphinx.websupport utilties for all documentation.
+ * sphinx.websupport utilities for all documentation.
  *
  * :copyright: Copyright 2007-2016 by the Sphinx team, see AUTHORS.
  * :license: BSD, see LICENSE for details.
diff --git a/doc_cn/algorithm/rnn/rnn.html b/doc_cn/algorithm/rnn/rnn.html
deleted file mode 100644
index 323c5a4ea6..0000000000
--- a/doc_cn/algorithm/rnn/rnn.html
+++ /dev/null
@@ -1,110 +0,0 @@
-
-
-
-
-  
-    
-    
-    RNN 配置 — PaddlePaddle  documentation
-    
-    
-    
-    
-    
-    
-    
-    
-    
-    
-     
-  
-  
-      
-
-    
    
-      
    
-        
    
-          
    
-            
-  
    
-
    RNN 配置
    
-
    
-
      
-
    
-
    
-
-
    
-
-
-          
    
-        
    
-      
    
-      
-      
    
-    
    
-    
-    
-  
-
\ No newline at end of file
diff --git a/doc_cn/build/docker/build_docker_image.html b/doc_cn/build/docker/build_docker_image.html
index cbdd566966..1761207763 100644
--- a/doc_cn/build/docker/build_docker_image.html
+++ b/doc_cn/build/docker/build_docker_image.html
@@ -6,7 +6,7 @@
   
     
     
-    构建PaddlePaddle Docker Image — PaddlePaddle  documentation
+    构建PaddlePaddle Docker Image — PaddlePaddle  documentation
     
     
     
@@ -33,7 +33,7 @@
         
  • 
           index
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
      
 
@@ -103,14 +103,11 @@ docker build --build-arg LOWEST_DL_SPEED
 
 
         
    
@@ -123,12 +120,12 @@ docker build --build-arg LOWEST_DL_SPEED
         
  • 
           index
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc_cn/build_and_install/cmake/compile_options.html b/doc_cn/build_and_install/cmake/compile_options.html
index 0bc28d5124..47678addc2 100644
--- a/doc_cn/build_and_install/cmake/compile_options.html
+++ b/doc_cn/build_and_install/cmake/compile_options.html
@@ -6,7 +6,7 @@
   
     
     
-    设置PaddlePaddle的编译选项 — PaddlePaddle  documentation
+    设置PaddlePaddle的编译选项 — PaddlePaddle  documentation
     
     
     
@@ -24,10 +24,7 @@
     
     
     
-    
-    
-    
-     
+     
   
   
       
 
@@ -212,12 +201,6 @@ cmake
 
 
 
-  
    Previous topic
    
-  
    安装编译PaddlePaddle需要的依赖
    
-  
    Next topic
    
-  
    make和make install
    
   
    
     
    This Page
    
     
      
@@ -228,14 +211,11 @@ cmake
 
 
         
    
@@ -248,20 +228,12 @@ cmake
         
  • 
           index
    • 
-        
  • 
-          next |
    • 
-        
  • 
-          previous |
    • 
-        
  • PaddlePaddle  documentation »
    • 
-          
  • 编译与安装 »
    • 
-          
  • 使用cmake编译PaddlePaddle »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc_cn/build_and_install/cmake/index.html b/doc_cn/build_and_install/cmake/index.html
index 9771c0f55b..83bc138b39 100644
--- a/doc_cn/build_and_install/cmake/index.html
+++ b/doc_cn/build_and_install/cmake/index.html
@@ -6,7 +6,7 @@
   
     
     
-    使用cmake编译PaddlePaddle — PaddlePaddle  documentation
+    使用cmake编译PaddlePaddle — PaddlePaddle  documentation
     
     
     
@@ -24,10 +24,7 @@
     
     
     
-    
-    
-    
-     
+     
   
   
       
 
@@ -77,12 +67,6 @@
       
    
       
       
       
     
   
 
\ No newline at end of file
diff --git a/doc_cn/demo/quick_start/index.html b/doc_cn/demo/quick_start/index.html
index 3a3347078d..3acc83cf78 100644
--- a/doc_cn/demo/quick_start/index.html
+++ b/doc_cn/demo/quick_start/index.html
@@ -6,7 +6,7 @@
   
     
     
-    PaddlePaddle快速入门教程 — PaddlePaddle  documentation
+    PaddlePaddle快速入门教程 — PaddlePaddle  documentation
     
     
     
@@ -24,9 +24,7 @@
     
     
     
-    
-    
-     
+     
   
   
       
 
@@ -91,7 +83,7 @@
 
    
 
    数据格式准备(Data Preparation)
    
 
    在本问题中，我们使用Amazon电子产品评论数据，
-将评论分为好评(正样本)和差评(负样本)两类。demo/quick_start里提供了数据下载脚本
+将评论分为好评(正样本)和差评(负样本)两类。源码的demo/quick_start里提供了数据下载脚本
 和预处理脚本。
    
 
    cd demo/quick_start
 ./data/get_data.sh
@@ -203,7 +195,7 @@ PyDataProviderWrapper。
    
 
      
    
 
    我们将以基本的逻辑回归网络作为起点，并逐渐展示更加深入的功能。更详细的网络配置
 连接请参考Layer文档。
-所有配置在demo/quick_start目录，首先列举逻辑回归网络。
    
+所有配置在源码demo/quick_start目录，首先列举逻辑回归网络。
    
 
    
 
    逻辑回归模型(Logistic Regression)
    
 
    流程如下：
@@ -258,7 +250,7 @@ PyDataProviderWrapper。
    
 
    词向量模型(Word Vector)
    
 
    embedding模型需要稍微改变数据提供的脚本，即dataprovider_emb.py，词向量模型、
 卷积模型、时序模型均使用该脚本。其中文本输入类型定义为整数时序类型integer_value_sequence。
    
-
    def initializer(settings, dictionary, **kwargs):
+
    def initializer(settings, dictionary, **kwargs):
     settings.word_dict = dictionary
     settings.input_types = [
         # Define the type of the first input as sequence of integer.
@@ -425,15 +417,15 @@ paddle train \
 mv rank-00000 result.txt
 
    
 
    
-
    与训练网络配置不同的是：无需label相关的层，指定outputs输出概率层(softmax输出)，
+
    这里以output/pass-00003为例进行预测，用户可以根据训练log选择test结果最好的模型来预测。与训练网络配置不同的是：无需label相关的层，指定outputs输出概率层(softmax输出)，
 指定batch_size=1，数据传输无需label数据，预测数据指定test_list的位置。
    
 
    预测结果以文本的形式保存在result.txt中，一行为一个样本，格式如下：
    
-
    预测ID;ID为0的概率 ID为1的概率
-预测ID;ID为0的概率 ID为1的概率
+
    预测ID;ID为0的概率 ID为1的概率
+预测ID;ID为0的概率 ID为1的概率
 
    
 
    
-
    is_predict = get_config_arg('is_predict', bool, False)
-trn = 'data/train.list' if not is_predict else None
+
    is_predict = get_config_arg('is_predict', bool, False)
+trn = 'data/train.list' if not is_predict else None
 tst = 'data/test.list' if not is_predict else 'data/pred.list'
 obj = 'process' if not is_predict else 'process_pre'
 batch_size = 128 if not is_predict else 1
@@ -501,7 +493,7 @@ mv rank-00000 result.txt
 
    
 
    
 
    输出日志(Log)
    
-
    TrainerInternal.cpp:160]  Batch=20 samples=2560 AvgCost=0.628761 CurrentCost=0.628761 Eval: classification_error_evaluator=0.304297  CurrentEval: classification_error_evaluator=0.304297
+
    TrainerInternal.cpp:160]  Batch=20 samples=2560 AvgCost=0.628761 CurrentCost=0.628761 Eval: classification_error_evaluator=0.304297  CurrentEval: classification_error_evaluator=0.304297
 
    
 
    
 
    模型训练会看到这样的日志，详细的参数解释如下面表格：
@@ -572,12 +564,6 @@ mv rank-00000 result.txt
 
 
 
-  
    Previous topic
    
-  
    PaddlePaddle文档
    
-  
    Next topic
    
-  
    编译与安装
    
   
    
     
    This Page
    
     
      
@@ -588,14 +574,11 @@ mv rank-00000 result.txt
 
 
         
    
@@ -608,18 +591,12 @@ mv rank-00000 result.txt
         
  • 
           index
    • 
-        
  • 
-          next |
    • 
-        
  • 
-          previous |
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc_cn/demo/semantic_role_labeling/index.html b/doc_cn/demo/semantic_role_labeling/index.html
deleted file mode 100644
index 0359806335..0000000000
--- a/doc_cn/demo/semantic_role_labeling/index.html
+++ /dev/null
@@ -1,94 +0,0 @@
-
-
-
-
-  
-    
-    
-    语义标注 — PaddlePaddle  documentation
-    
-    
-    
-    
-    
-    
-    
-    
-    
-     
-  
-  
-      
-
-    
    
-      
    
-        
    
-          
    
-            
-  
    
-
    语义标注
    
-
    TBD
    
-
    
-
-
-          
    
-        
    
-      
    
-      
-      
    
-    
    
-    
-    
-  
-
\ No newline at end of file
diff --git a/doc_cn/demo/sentiment_analysis/index.html b/doc_cn/demo/sentiment_analysis/index.html
deleted file mode 100644
index 02054fa732..0000000000
--- a/doc_cn/demo/sentiment_analysis/index.html
+++ /dev/null
@@ -1,94 +0,0 @@
-
-
-
-
-  
-    
-    
-    情感分析 — PaddlePaddle  documentation
-    
-    
-    
-    
-    
-    
-    
-    
-    
-     
-  
-  
-      
-
-    
    
-      
    
-        
    
-          
    
-            
-  
    
-
    情感分析
    
-
    TBD
    
-
    
-
-
-          
    
-        
    
-      
    
-      
-      
    
-    
    
-    
-    
-  
-
\ No newline at end of file
diff --git a/doc_cn/demo/text_generation/index.html b/doc_cn/demo/text_generation/index.html
deleted file mode 100644
index a294f97dde..0000000000
--- a/doc_cn/demo/text_generation/index.html
+++ /dev/null
@@ -1,94 +0,0 @@
-
-
-
-
-  
-    
-    
-    文本生成 — PaddlePaddle  documentation
-    
-    
-    
-    
-    
-    
-    
-    
-    
-     
-  
-  
-      
-
-    
    
-      
    
-        
    
-          
    
-            
-  
    
-
    文本生成
    
-
    TBD
    
-
    
-
-
-          
    
-        
    
-      
    
-      
-      
    
-    
    
-    
-    
-  
-
\ No newline at end of file
diff --git a/doc_cn/dev/new_layer/index.html b/doc_cn/dev/new_layer/index.html
deleted file mode 100644
index eabbed8445..0000000000
--- a/doc_cn/dev/new_layer/index.html
+++ /dev/null
@@ -1,116 +0,0 @@
-
-
-
-
-  
-    
-    
-    新写Layer — PaddlePaddle  documentation
-    
-    
-    
-    
-    
-    
-    
-    
-    
-    
-    
-     
-  
-  
-      
-
-    
    
-      
    
-        
    
-          
    
-            
-  
    
-
    新写Layer
    
-
-
    
-
-
-          
    
-        
    
-      
    
-      
-      
    
-    
    
-    
-    
-  
-
\ No newline at end of file
diff --git a/doc_cn/genindex.html b/doc_cn/genindex.html
index fcd544b0d0..38ddb02da7 100644
--- a/doc_cn/genindex.html
+++ b/doc_cn/genindex.html
@@ -7,7 +7,7 @@
   
     
     
-    Index — PaddlePaddle  documentation
+    Index — PaddlePaddle  documentation
     
     
     
@@ -34,7 +34,7 @@
         
  • 
           index
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
      
 
@@ -62,14 +62,11 @@
 
 
         
    
@@ -82,12 +79,12 @@
         
  • 
           index
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc_cn/index.html b/doc_cn/index.html
index f2023f27ed..4f8f3162c0 100644
--- a/doc_cn/index.html
+++ b/doc_cn/index.html
@@ -6,7 +6,7 @@
   
     
     
-    PaddlePaddle文档 — PaddlePaddle  documentation
+    PaddlePaddle文档 — PaddlePaddle  documentation
     
     
     
@@ -24,8 +24,7 @@
     
     
     
-    
-     
+     
   
   
       
 
@@ -47,36 +43,30 @@
           
    
             
   
    
-
    PaddlePaddle文档
    
-
    
-
    使用指南
    
-
    
-
    
-
    
-
    开发指南
    
-
    
-
    
-
    
-
    算法教程
    
-
    
-
    
 
    
 
 
@@ -88,16 +78,13 @@
   
    Table Of Contents
    
   
 
-  
    Next topic
    
-  
    PaddlePaddle快速入门教程
    
   
    
     
    This Page
    
     
      
@@ -108,14 +95,11 @@
 
 
         
    
@@ -128,15 +112,12 @@
         
  • 
           index
    • 
-        
  • 
-          next |
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc_cn/objects.inv b/doc_cn/objects.inv
index 05fe95da3c231256c1756e4c5aeb65a0a64ff3ba..7969c18370e0d15b69996dcf3db0c407e6c28fe7 100644
GIT binary patch
delta 741
zcmV`@Qc$k7;2TsW9CF`g&eov}>VFL!6b-l~9z
z1u8}`0%{O24aVw)Sg4dgv!DOM*_~}V%gmN~+1+{H=i{As=cDdn9af>GEh48{PX~Yv
z05;SHQd|Nb(10>?U#U-q-diAsu7qD!gTZA@g-}QAsmx^{Wq&>eRTxh-A;6|fR9&xX
zhPsHVnq}BnN5HZPwoF&>y`94Rhb>3{%Z7hFaEWYv^J8Xi_!|P3DJvbd@~?(
z^RKR>58W)xJb2>?PT2@RWV_QDj`7Sl=xxNuYocXx*owcuq%-!1pX2kr(hAZCAD*U)
zn;ce28V%lWrhhLbPe@#+=q=Y(HZQ-w&88{UWAD-Wn`Rbr4&ToktA?%HWcM&w=}f4>
zt9vH#{;!52hhIw&)q~Bo;IR8NgZTZoVr>;o_gq3Zq0rVsg2^TaucP5f-1`{~Rw@t~
zmdF}M`AVXbOMlpAl;wjyocA9+bhP$7lZoASz=2%Ww
z>qt)pM^Pp*pMF*Z*AkO(Z#_IZt{4`S2;0q>QGziaThOB?6ti%V>Szl^28%K96PB-~
z$C!uduAqmZXV^fq%xAd4*Cu9}!b8(i2qrpuG!hV|#mDotxpUkI{WrGQpfn
z92J}8(pkD7oJDSxrPeOJHOhr^KtVc>$;F{cRA4*S5{9y^DMyG4n2J-uaH|(~-uc6W
zJ9Cs&%Q<)bK|303iEef?9gl{y6Mb)B^9q{0&t7bLgJg$*XtnHMf`EyZ0WUMrQBKM4
z>-Dpsx9?wU`2CjOKjW)Ny=86WZ;B7y@Xeoa`!whrr*PXr5No9BF5?EUn;`#hMZO#)
X{S&n&S+LCH+iG~;&RzZiq|eO^;ZJjl

delta 88
zcmV-e0H^=c2Fn4EdQnSH&CAS7Nv%*QE=jS
     
     
-    Search — PaddlePaddle  documentation
+    Search — PaddlePaddle  documentation
     
     
     
@@ -41,7 +41,7 @@
         
  • 
           index
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
      
 
@@ -89,12 +89,12 @@
         
  • 
           index
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc_cn/searchindex.js b/doc_cn/searchindex.js
index 5fbebb14fd..b9890f04a0 100644
--- a/doc_cn/searchindex.js
+++ b/doc_cn/searchindex.js
@@ -1 +1 @@
-Search.setIndex({envversion:46,filenames:["algorithm/rnn/rnn","build/docker/build_docker_image","build_and_install/cmake/compile_options","build_and_install/cmake/index","build_and_install/cmake/install_deps","build_and_install/cmake/make_and_install","build_and_install/index","build_and_install/install/docker_install","build_and_install/install/index","build_and_install/install/ubuntu_install","cluster/index","demo/embedding_model/index","demo/image_classification/index","demo/imagenet_model/index","demo/index","demo/quick_start/index","demo/semantic_role_labeling/index","demo/sentiment_analysis/index","demo/text_generation/index","dev/new_layer/index","index","ui/cmd/dump_config","ui/cmd/index","ui/cmd/make_diagram","ui/cmd/merge_model","ui/cmd/paddle_pserver","ui/cmd/paddle_train","ui/cmd/paddle_version","ui/data_provider/index","ui/data_provider/pydataprovider2","ui/data_provider/write_new_dataprovider","ui/index","ui/model","ui/predict/swig_py_paddle"],objects:{},objnames:{},objtypes:{},terms:{"0000x":15,"000\u5f20\u7070\u5ea6\u56fe\u7247\u7684\u6570\u5b57\u5206\u7c7b\u6570\u636e\u96c6":29,"00186201e":33,"04\u7248\u672c\u4f7f\u7528deb\u5305\u5b89\u88c5":9,"08823112e":33,"0b0":9,"103\u884c\u521b\u5efa\u4e00\u4e2a\u4ece\u5de5\u5177\u7c7b":33,"10\u4ee5\u4e0a\u7684linux":7,"10\u7ef4\u7684\u6574\u6570\u503c":29,"10gbe":7,"10m":1,"12194102e":33,"12\u7248\u672c\u6d4b\u8bd5\u901a\u8fc7":1,"12\u7248\u672c\u7684\u60c5\u51b5\u4e0b\u5e76\u6ca1\u6709\u6d4b\u8bd5":1,"15501715e":33,"15mb":15,"16mb":15,"1\u7684\u8bdd":29,"252kb":15,"25639710e":33,"25k":15,"27787406e":33,"28\u7684\u50cf\u7d20\u7070\u5ea6\u503c":29,"28\u7684\u7a20\u5bc6\u5411\u91cf\u548c\u4e00\u4e2a":29,"2\u8fdb\u884c\u8fdb\u4e00\u6b65\u6f14\u5316":15,"32777140e":33,"36540484e":33,"40gbe":7,"43630644e":33,"48565123e":33,"48684503e":33,"49316648e":33,"50k":1,"51111044e":33,"53018653e":33,"56gbe":7,"5\u5230\u672c\u5730\u73af\u5883\u4e2d":9,"70634608e":33,"72296313e":33,"85625684e":33,"93137714e":33,"96644767e":33,"99982715e":33,"9\u7684\u6570\u5b57":29,"\u4e00\u4e2a\u6587\u4ef6":29,"\u4e00\u4e2alogging\u5bf9\u8c61":29,"\u4e00\u4e2apass\u8868\u793a\u8fc7\u4e00\u904d\u6240\u6709\u8bad\u7ec3\u6837\u672c":15,"\u4e00\u6761":29,"\u4e00\u81f4\u7684\u7279\u5f81":29,"\u4e00\u822c\u60c5\u51b5\u4e0b":28,"\u4e00\u822c\u63a8\u8350\u8bbe\u7f6e\u6210true":29,"\u4e00\u884c\u4e3a\u4e00\u4e2a\u6837\u672c":15,"\u4e0a\u7684\u6548\u679c":15,"\u4e0b\u8f7d\u8fdb\u7a0b\u4f1a\u91cd\u542f":1,"\u4e0b\u8ff0\u5185\u5bb9\u5c06\u5206\u4e3a\u5982\u4e0b\u51e0\u4e2a\u7c7b\u522b\u63cf\u8ff0":7,"\u4e0b\u9762dataprovid":15,"\u4e0d\u4e00\u5b9a\u548c\u65f6\u95f4\u6709\u5173\u7cfb":29,"\u4e0d\u4f1a\u6267\u884c\u6d4b\u8bd5\u64cd\u4f5c":28,"\u4e0d\u540c\u7684\u6570\u636e\u7c7b\u578b\u548c\u5e8f\u5217\u6a21\u5f0f\u8fd4\u56de\u7684\u683c\u5f0f\u4e0d\u540c":29,"\u4e0d\u540c\u7684\u7a33\u5b9a\u7248\u672c":7,"\u4e0d\u7f13\u5b58\u4efb\u4f55\u6570\u636e":29,"\u4e0e\u8bad\u7ec3\u7f51\u7edc\u914d\u7f6e\u4e0d\u540c\u7684\u662f":15,"\u4e14\u5e8f\u5217\u7684\u6bcf\u4e00\u4e2a\u5143\u7d20\u8fd8\u662f\u4e00\u4e2a\u65f6\u95f4\u5e8f\u5217":29,"\u4e24\u4e2a\u6587\u6863":7,"\u4e24\u7c7b":15,"\u4e2a":15,"\u4e2d":15,"\u4e2d\u5b9a\u4e49\u4f7f\u7528\u4ec0\u4e48dataprovid":28,"\u4e2d\u5b9a\u4e49\u7684\u987a\u5e8f\u4e00\u81f4":29,"\u4e2d\u5bfb\u627e\u8fd9\u4e9bblas\u7684\u5b9e\u73b0":2,"\u4e2d\u7684":29,"\u4e2d\u7684\u4e8c\u8fdb\u5236\u4f7f\u7528\u4e86":7,"\u4e2d\u7684set":29,"\u4e2d\u914d\u7f6e":29,"\u4e3a":29,"\u4e3a\u4e86\u8fd0\u884cpaddlepaddle\u7684docker\u955c\u50cf":7,"\u4e3b\u8981\u51fd\u6570\u662fprocess\u51fd\u6570":29,"\u4e3b\u8981\u5206\u4e3a\u4ee5\u4e0b\u51e0\u4e2a\u6b65\u9aa4":33,"\u4e3b\u8981\u5305\u62ec\u4e24\u90e8\u5206":15,"\u4e3b\u8981\u662f\u589e\u52a0\u4e86\u521d\u59cb\u5316\u673a\u5236":29,"\u4e3b\u8981\u6b65\u9aa4\u4e3a":33,"\u4e3b\u8981\u7531\u4e8e\u65e7\u7248\u672c":1,"\u4e3b\u8981\u7684\u8f6f\u4ef6\u5305\u4e3apy_paddl":33,"\u4e5f\u4f1a\u8bfb\u53d6\u76f8\u5173\u8def\u5f84\u53d8\u91cf\u6765\u8fdb\u884c\u641c\u7d22":2,"\u4e5f\u53ef\u4ee5\u4f7f\u7528":29,"\u4e5f\u53ef\u4ee5\u548cpaddl":22,"\u4e5f\u53ef\u4ee5\u76f4\u63a5\u6267\u884c":7,"\u4e5f\u53ef\u4ee5\u901a\u8fc7\u5982\u4e0b\u65b9\u5f0f\u9884\u6d4b":15,"\u4e5f\u53ef\u4ee5\u901a\u8fc7save":15,"\u4e5f\u53ef\u4ee5\u9884\u6d4b\u6ca1\u6709label\u7684\u6d4b\u8bd5\u96c6":15,"\u4e5f\u5c31\u662f\u5c06\u8bcd\u5411\u91cf\u6a21\u578b\u989d\u6b65":15,"\u4e5f\u5c31\u662f\u76f4\u63a5\u5199\u5185\u5b58\u7684float\u6570\u7ec4":33,"\u4e86\u975e\u5e38\u65b9\u4fbf\u7684\u4e8c\u8fdb\u5236\u5206\u53d1\u624b\u6bb5":7,"\u4e8c\u6b21\u5f00\u53d1\u53ef\u4ee5":7,"\u4eba\u5458\u7b49\u7b49":1,"\u4ec5\u4ec5\u9700\u8981":29,"\u4ecb\u7ecdpaddlepaddle\u4f7f\u7528\u6d41\u7a0b\u548c\u5e38\u7528\u7684\u7f51\u7edc\u57fa\u7840\u5355\u5143\u7684\u914d\u7f6e\u65b9\u6cd5":15,"\u4ece\u6587\u4ef6\u4e2d\u8bfb\u53d6\u6bcf\u4e00\u6761\u6570\u636e":29,"\u4ece\u6587\u672c\u6587\u4ef6\u4e2d\u8bfb\u53d6":29,"\u4ece\u800c\u4e0d\u80fd\u5728\u8fd0\u884c\u7f16\u8bd1\u547d\u4ee4\u7684\u65f6\u5019\u63a5\u53d7\u53c2\u6570":1,"\u4ece\u800c\u9632\u6b62\u8fc7\u62df\u5408":28,"\u4ed6\u4eec\u5305\u62ecdocker\u955c\u50cf":8,"\u4ed6\u4eec\u662f":[22,29],"\u4ed6\u4eec\u7684\u4e0b\u8f7d\u5730\u5740\u662f":9,"\u4ed6\u53ef\u4ee5\u5c06\u67d0\u4e00\u4e2a\u51fd\u6570\u6807\u8bb0\u6210\u4e00\u4e2apydataprovid":29,"\u4ee4\u884c\u53c2\u6570\u6587\u6863":15,"\u4ee5\u592a\u7f51\u5361":7,"\u4ee5\u76f8\u5bf9\u8def\u5f84\u5f15\u7528":28,"\u4efb\u610f\u4e00\u79cdcblas\u5b9e\u73b0":2,"\u4f1a\u5c06latest\u66ff\u6362\u6210\u7a33\u5b9a\u7248\u672c\u7684\u7248\u672c\u53f7":7,"\u4f1a\u62a5\u5bfb\u627e\u4e0d\u5230\u8fd9\u4e9b\u52a8\u6001\u5e93":9,"\u4f20\u5165\u5185\u5b58\u6216\u8005\u663e\u5b58":28,"\u4f20\u5165\u7684\u662f\u4e00\u4e2a\u51fd\u6570":29,"\u4f20\u5165\u7684\u914d\u7f6e\u53c2\u6570\u5305\u62ec":1,"\u4f20\u5165\u8fd9\u4e2a\u53d8\u91cf\u7684\u65b9\u5f0f\u4e3a":29,"\u4f46\u662f":1,"\u4f46\u662f\u5728":29,"\u4f46\u662f\u5982\u679c\u4f7f\u7528\u4e86\u9ad8\u6027\u80fd\u7684\u7f51\u5361":7,"\u4f46\u662f\u65b9\u4fbf\u8c03\u8bd5\u548cbenchmark":2,"\u4f46\u662f\u6709\u65f6\u4e3a\u4e86\u8ba1\u7b97\u5747\u8861\u6027":29,"\u4f46\u9700\u8981\u6ce8\u610f\u7684\u662f\u7f16\u8bd1\u548c":2,"\u4f4e\u4e8edocker":1,"\u4f53\u53ef\u4ee5\u53c2\u8003":29,"\u4f7f\u5728python\u73af\u5883\u4e0b\u7684\u9884\u6d4b\u63a5\u53e3\u66f4\u52a0\u7b80\u5355":33,"\u4f7f\u7528":[2,7],"\u4f7f\u7528\u547d\u4ee4\u884c\u53c2\u6570\u521d\u59cb\u5316paddlepaddl":33,"\u4f7f\u7528\u7684\u547d\u4ee4\u4e5f\u662f":2,"\u4f7f\u7528\u8be5\u63a5\u53e3\u7528\u6237\u53ef\u4ee5\u53ea\u5173\u6ce8\u5982\u4f55":29,"\u4f7f\u7528\u8be5dockerfile\u6784\u5efa\u51fa\u955c\u50cf":7,"\u4f7f\u7528\u8d77\u6765\u6216\u8bb8\u4f1a\u4e0d\u662f\u5f88\u987a\u7545":8,"\u4f7f\u7528\u8fd9\u4e2a\u5173\u952e\u8bcd":29,"\u4f7f\u7528deb\u5305\u5728ubuntu\u4e0a\u5b89\u88c5paddlepaddl":8,"\u4f7f\u7528dockerfile\u6784\u5efa\u4e00\u4e2a\u5168\u65b0\u7684dock":7,"\u4f7f\u7528mnist\u624b\u5199\u8bc6\u522b\u4f5c\u4e3a\u6837\u4f8b":33,"\u4f7f\u7528ssh\u8bbf\u95eepaddlepaddle\u955c\u50cf":7,"\u4f8b\u5982":[2,9,15,29],"\u4f8b\u5982\u6587\u4ef6\u540d\u662f":29,"\u4f8b\u5982rdma\u7f51\u5361":7,"\u4f8b\u5982sigmoid\u53d8\u6362":15,"\u4f9d\u6b21\u8fd4\u56de\u4e86\u6587\u4ef6\u4e2d\u7684\u6bcf\u6761\u6570\u636e":29,"\u4fe1\u606f":7,"\u5173\u95edcontain":7,"\u5176\u4e2d":[1,7,28,29,33],"\u5176\u4e2d\u6587\u672c\u8f93\u5165\u7c7b\u578b\u5b9a\u4e49\u4e3a\u6574\u6570\u65f6\u5e8f\u7c7b\u578binteg":15,"\u5176\u4e2d\u7b2c\u4e00\u884c\u662f\u5f15\u5165paddlepaddle\u7684pydataprovider2\u5305":29,"\u5176\u4ed6\u53c2\u6570\u5747\u4f7f\u7528kei":29,"\u5176\u4ed6\u53c2\u6570\u8bf7\u53c2\u8003":15,"\u5176\u4ed6\u53c2\u6570\u90fd\u4f7f\u7528kei":29,"\u5176\u4f5c\u7528\u662f\u5c06\u8bad\u7ec3\u6570\u636e":28,"\u5176\u5b83\u90e8\u5206\u548c\u903b\u8f91\u56de\u5f52\u7f51\u7edc\u7ed3\u6784\u4e00\u81f4":15,"\u5176\u6570\u636e\u4f7f\u7528":29,"\u5176\u6bcf\u4e00\u884c\u5bf9\u5e94\u8fd9\u6bcf\u4e00\u4e2a\u6570\u636e\u6587\u4ef6":28,"\u5176\u7b2c\u4e00\u884c\u8bf4\u660e\u4e86paddle\u7684\u7248\u672c":27,"\u5177":29,"\u5177\u4f53\u53ef\u4ee5\u8bbe\u7f6e\u6210\u4ec0\u4e48\u5176\u4ed6\u683c":29,"\u5177\u4f53\u6709\u54ea\u4e9b\u683c\u5f0f":29,"\u5177\u4f53\u8bf7\u53c2\u8003\u6ce8\u610f\u4e8b\u9879\u4e2d\u7684":7,"\u5177\u4f53dataprovider\u8fd8\u5177\u6709\u4ec0\u4e48\u529f\u80fd":29,"\u5177\u6709\u4e24\u4e2a\u53c2\u6570":29,"\u5177\u6709\u548c\u5bbf\u4e3b\u673a\u76f8\u8fd1\u7684\u8fd0\u884c\u6548\u7387":7,"\u5177\u6709\u7684\u5c5e\u6027\u6709":29,"\u5178\u578b\u7684\u8f93\u51fa\u7ed3\u679c\u4e3a":33,"\u5178\u578b\u7684\u9884\u6d4b\u4ee3\u7801\u5982\u4e0b":33,"\u518d\u6307\u5b9a":2,"\u5199\u5165train":29,"\u51c6\u5907\u6570\u636e":33,"\u51fd\u6570":29,"\u51fd\u6570\u4e2d":29,"\u51fd\u6570\u4e2d\u4f7f\u7528":29,"\u51fd\u6570\u4e2d\u7684":29,"\u51fd\u6570\u662f\u4f7f\u7528":29,"\u51fd\u6570\u6765\u4fdd\u8bc1\u517c\u5bb9\u6027":29,"\u51fd\u6570\u67e5\u8be2\u6587\u6863":33,"\u5206\u5e03\u5f0f\u8bad\u7ec3":15,"\u5206\u7c7b\u6210\u6b63\u9762\u60c5\u7eea\u548c":29,"\u5217\u8868\u5982\u4e0b":29,"\u5219\u4f1a\u6839\u636e\u547d\u4ee4\u884c":28,"\u5219\u5b57\u4e0e\u5b57\u4e4b\u95f4\u7528\u7a7a\u683c\u5206\u9694":15,"\u5219\u9700\u8981\u5148\u5c06":7,"\u5219\u9700\u8981\u8fdb\u884c\u4e00\u5b9a\u7684\u4e8c\u6b21\u5f00\u53d1":7,"\u5220\u9664contain":7,"\u5229\u7528\u5355\u8bcdid\u67e5\u627e\u5bf9\u5e94\u7684\u8be5\u5355\u8bcd\u7684\u8fde\u7eed\u8868\u793a\u5411\u91cf":15,"\u5229\u7528\u903b\u8f91\u56de\u5f52\u6a21\u578b\u5bf9\u8be5\u5411\u91cf\u8fdb\u884c\u5206\u7c7b":15,"\u522b":15,"\u5269\u4e0b\u7684pass\u4f1a\u76f4\u63a5\u4ece\u5185\u5b58\u91cc":29,"\u52a0\u4e86l2\u6b63\u5219\u548c\u68af\u5ea6\u622a\u65ad":15,"\u52a0\u8f7d\u6570\u636e":15,"\u5305":7,"\u5305\u548c":7,"\u5305\u62ec":15,"\u5305\u62ec\u7b80\u5355\u7684rnn\u6a21\u578b":15,"\u5373":[7,15],"\u5373\u4e0d\u5728\u4e4e\u5185\u5b58\u6682\u5b58\u591a\u5c11\u6761\u6570\u636e":29,"\u5373\u4e0d\u662f\u4e00\u6761\u5e8f\u5217":29,"\u5373\u4ece\u5355\u8bcd\u5b57\u7b26\u4e32\u5230\u5355\u8bcdid\u7684\u5b57\u5178":29,"\u5373\u5305\u542b\u65f6\u95f4\u6b65\u4fe1\u606f":29,"\u5373\u53ef":29,"\u5373\u53ef\u4ee5\u4f7f\u7528ssh\u8bbf\u95ee\u5bbf\u4e3b\u673a\u76848022\u7aef\u53e3":7,"\u5373\u53ef\u542f\u52a8\u548c\u8fdb\u5165paddlepaddle\u7684contain":7,"\u5373\u53ef\u5728\u672c\u5730\u7f16\u8bd1\u51fapaddlepaddle\u7684\u955c\u50cf":1,"\u5373\u53ef\u6253\u5370\u51fapaddlepaddle\u7684\u7248\u672c\u548c\u6784\u5efa":7,"\u5373\u5927\u90e8\u5206\u503c\u4e3a0":29,"\u5373\u5bf9\u7b2c3\u6b65\u8fdb\u884c\u66ff\u6362":15,"\u5373\u662f\u4e00\u6761\u65f6\u95f4\u5e8f\u5217":29,"\u5373cpu\u7248\u672c":9,"\u5373train":29,"\u5377\u79ef\u7f51\u7edc\u662f\u4e00\u79cd\u7279\u6b8a\u7684\u4ece\u8bcd\u5411\u91cf\u8868\u793a\u5230\u53e5\u5b50\u8868\u793a\u7684\u65b9\u6cd5":15,"\u53c2\u6570":1,"\u53c2\u6570\u6307\u5b9a\u7684\u6d4b\u8bd5\u65b9\u5f0f":28,"\u53c2\u6570\u6570\u91cf":15,"\u53c2\u8003":28,"\u53c2\u89c1":[4,5,10],"\u53d6\u51b3\u4e8e\u662f\u5426\u5bfb\u627e\u5230gflag":2,"\u53d6\u51b3\u4e8e\u662f\u5426\u5bfb\u627e\u5230glog":2,"\u53d6\u51b3\u4e8e\u662f\u5426\u5bfb\u627e\u5230gtest":2,"\u53d6\u51b3\u4e8e\u662f\u5426\u627e\u5230swig":2,"\u53d8\u4e3a3\u4e2a\u65b0\u7684\u5b50\u6b65\u9aa4":15,"\u53d8\u4f1a\u62a5\u8fd9\u4e2a\u9519\u8bef":9,"\u53e5\u5b50\u8868\u793a\u7684\u8ba1\u7b97\u66f4\u65b0\u4e3a2\u6b65":15,"\u53ea\u662f\u5c06\u53e5\u5b50\u5229\u7528\u8fde\u7eed\u5411\u91cf\u8868\u793a\u66ff\u6362\u7a00\u758f":15,"\u53ea\u662f\u8bf4\u660e\u6570\u636e\u7684\u987a\u5e8f\u662f\u91cd\u8981\u7684":29,"\u53ea\u9700\u8981\u4f7f\u7528\u4e00\u884c\u4ee3\u7801\u5373\u53ef\u4ee5\u8bbe\u7f6e\u8bad\u7ec3\u5f15\u7528\u8fd9\u4e2adataprovid":29,"\u53ea\u9700\u8981\u5728":29,"\u53ea\u9700\u8981\u77e5\u9053\u8fd9\u53ea\u662f\u4e00\u4e2a\u6807\u8bb0\u5c5e\u6027\u7684\u65b9\u6cd5\u5c31\u53ef\u4ee5\u4e86":29,"\u53ef\u4ee5\u4e3a\u4e00\u4e2a\u6570\u636e\u6587\u4ef6\u8fd4\u56de\u591a\u6761\u8bad\u7ec3\u6837\u672c":29,"\u53ef\u4ee5\u4f20\u516510k":1,"\u53ef\u4ee5\u4f7f\u7528":1,"\u53ef\u4ee5\u4f7f\u7528\u8bad\u7ec3\u597d\u7684\u6a21\u578b\u8bc4\u4f30\u5e26\u6709label\u7684\u9a8c\u8bc1\u96c6":15,"\u53ef\u4ee5\u4f7f\u7528graphviz\u5bf9paddlepaddle\u7684\u7f51\u7edc\u6a21\u578b\u8fdb\u884c\u7ed8\u5236":22,"\u53ef\u4ee5\u4f7f\u7528paddl":22,"\u53ef\u4ee5\u4f7f\u7528python\u7684":33,"\u53ef\u4ee5\u53c2\u8003":15,"\u53ef\u4ee5\u5728\u4e00\u4e2a\u51fd\u6570\u91cc":29,"\u53ef\u4ee5\u5728cmake\u7684\u547d\u4ee4\u884c\u8bbe\u7f6e":2,"\u53ef\u4ee5\u5c06\u4e00\u6761\u6570\u636e\u8bbe\u7f6e\u6210\u591a\u4e2abatch":29,"\u53ef\u4ee5\u5c06paddlepaddle\u7684\u6a21\u578b\u548c\u914d\u7f6e\u6253\u5305\u6210\u4e00\u4e2a\u6587\u4ef6":22,"\u53ef\u4ee5\u5c06paddlepaddle\u7684\u8bad\u7ec3\u6a21\u578b\u4ee5proto":22,"\u53ef\u4ee5\u65b9\u4fbf\u5d4c\u5165\u5f0f\u5de5\u4f5c":2,"\u53ef\u4ee5\u6839\u636e\u4e0d\u540c\u7684\u6570\u636e\u914d\u7f6e\u4e0d\u540c\u7684\u8f93\u5165\u7c7b\u578b":29,"\u53ef\u4ee5\u901a\u8fc7show":15,"\u53ef\u7528\u5728\u6d4b\u8bd5\u6216\u8bad\u7ec3\u65f6\u6307\u5b9a\u521d\u59cb\u5316\u6a21\u578b":15,"\u53ef\u80fd\u9047\u5230\u7684\u95ee\u9898":8,"\u53ef\u9009":29,"\u5404\u79cd\u53c2\u6570\u548c\u7ef4\u62a4":1,"\u540c\u65f6":[1,2,7,29],"\u540c\u65f6\u4f1a\u8ba1\u7b97\u5206\u7c7b\u51c6\u786e\u7387":15,"\u540c\u65f6\u5305\u542bcpu\u5f00\u53d1\u73af\u5883":7,"\u540c\u65f6\u5305\u542bcpu\u5f00\u53d1\u73af\u5883\u548c\u6e90\u7801":7,"\u540c\u65f6\u5305\u542bgpu\u5f00\u53d1\u73af\u5883":7,"\u540c\u65f6\u5305\u542bgpu\u5f00\u53d1\u73af\u5883\u548c\u6e90\u7801":7,"\u540d\u79f0":15,"\u540e\u7528dataprovider\u4ee3\u66ff":28,"\u540e\u9762\u8ddf\u7740\u4e00\u7cfb\u5217\u7f16\u8bd1\u53c2\u6570":27,"\u5411\u91cf\u8868\u793a":15,"\u5426":2,"\u5426\u5219":28,"\u547d\u4ee4":1,"\u547d\u4ee4\u4e3a":7,"\u547d\u4ee4\u6307\u5b9a\u7684\u53c2\u6570\u4f1a\u4f20\u5165\u7f51\u7edc\u914d\u7f6e\u4e2d":15,"\u547d\u4ee4\u8fd0\u884c\u955c\u50cf":7,"\u547d\u4ee4\u9884\u5148\u4e0b\u8f7d\u955c\u50cf":7,"\u548c":[2,7,29],"\u548c\u4e09\u79cd\u5e8f\u5217\u6a21\u5f0f":29,"\u548c\u5dee\u8bc4":15,"\u548c\u6587\u672c\u4fe1\u606f\u7528tab\u95f4\u9694":15,"\u548c\u6d4b\u8bd5\u6587\u4ef6\u5217\u8868":28,"\u548c\u7528\u6237\u4f20\u5165\u7684\u53c2\u6570":29,"\u548c\u9884\u5904\u7406\u811a\u672c":15,"\u548ccudnn":9,"\u548cdataprovid":28,"\u548cgpu\u7248\u672c":9,"\u548cinitalizer\u91cc\u5b9a\u4e49\u987a\u5e8f\u4e00\u81f4":15,"\u5668":15,"\u56db\u79cd\u6570\u636e\u7c7b\u578b\u662f":29,"\u56fe\u50cf\u5206\u7c7b":14,"\u5728":[2,9,29],"\u5728100\u884c\u521b\u5efa\u795e\u7ecf\u7f51\u7edc":33,"\u572898\u884c\u8f7d\u5165paddlepaddle\u7684\u8bad\u7ec3\u6587\u4ef6":33,"\u5728\u58f0\u660edataprovider\u7684\u65f6\u5019\u4f20\u5165\u4e86dictionary\u4f5c\u4e3a\u53c2\u6570":29,"\u5728\u5b8c\u6210\u4e86\u6570\u636e\u548c\u7f51\u7edc\u7ed3\u6784\u642d\u5efa\u4e4b\u540e":15,"\u5728\u672c\u95ee\u9898\u4e2d":15,"\u5728\u6a21\u578b\u914d\u7f6e\u4e2d\u5229\u7528":15,"\u5728\u6b64\u4e3a\u65b9\u4fbf\u5bf9\u6bd4\u4e0d\u540c\u7f51\u7edc\u7ed3\u6784":15,"\u5728\u6bcf\u4e2a\u7ef4\u5ea6\u4e0a\u53d6\u51fa\u5728\u8be5\u53e5\u8bdd\u65b0\u7684\u5411\u91cf\u96c6\u5408\u4e0a\u8be5\u7ef4\u5ea6\u7684\u6700\u5927\u503c\u4f5c\u4e3a\u6700\u540e\u7684\u53e5\u5b50\u8868\u793a\u5411\u91cf":15,"\u5728\u7a0b\u5e8f\u5f00\u59cb\u9636\u6bb5":33,"\u5728\u7b2c105\u884c\u6267\u884c\u9884\u6d4b":33,"\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u8fdb\u884c\u6d4b\u8bd5":28,"\u5728\u8bad\u7ec3\u914d\u7f6e\u91cc":29,"\u5728\u8fd9\u4e2a\u51fd\u6570\u4e2d":29,"\u5728\u914d\u7f6e\u4e2d\u8bfb\u53d6\u4e86\u5b57\u5178":29,"\u5728cmake\u914d\u7f6e\u65f6\u53ef\u4ee5\u4f7f\u7528":2,"\u5728pydataprovider\u4e2d":29,"\u5728python\u73af\u5883\u4e0b\u9884\u6d4b\u7ed3\u679c":33,"\u57fa\u672c\u7684\u5904\u7406\u903b\u8f91\u4e5f\u548cmnist\u903b\u8f91\u4e00\u81f4":29,"\u57fa\u672c\u7684pydataprovider\u4f7f\u7528\u4ecb\u7ecd\u5b8c\u6bd5\u4e86":29,"\u591a\u4e2ainput\u4ee5list\u65b9\u5f0f\u8f93\u5165":15,"\u591a\u6b21\u8fd4\u56de\u53d8\u91cf":29,"\u591a\u7ebf\u7a0b\u4e0b\u8f7d\u8fc7\u7a0b\u4e2d":1,"\u591a\u7ebf\u7a0b\u6570\u636e\u8bfb\u53d6":29,"\u597d\u8bc4":15,"\u5982\u679c":28,"\u5982\u679c\u4e0d\u4e86\u89e3":29,"\u5982\u679c\u4e0d\u4f7f\u7528\u5219\u4f1a\u4f7f\u7528\u4e00\u4e2a\u7b80\u5316\u7248\u7684\u547d\u4ee4\u884c\u53c2\u6570\u89e3\u6790":2,"\u5982\u679c\u4e0d\u4f7f\u7528\u5219\u4f1a\u4f7f\u7528\u4e00\u4e2a\u7b80\u5316\u7248\u7684\u65e5\u5fd7\u5b9e\u73b0":2,"\u5982\u679c\u4e0d\u5207\u8bcd":15,"\u5982\u679c\u4e0d\u8bbe\u7f6e\u7684\u8bdd":29,"\u5982\u679c\u4f7f\u7528gpu\u7248\u672c\u7684paddlepaddl":9,"\u5982\u679c\u5728":9,"\u5982\u679c\u5728\u7b2c\u4e00\u6b21cmake\u4e4b\u540e\u60f3\u8981\u91cd\u65b0\u8bbe":2,"\u5982\u679c\u5c0f\u4e8e\u8fd9\u4e2a\u4e0b\u8f7d\u901f\u5ea6":1,"\u5982\u679c\u60a8\u4f7f\u7528":7,"\u5982\u679c\u60f3\u8981\u5728\u5916\u90e8\u673a\u5668\u8bbf\u95ee\u8fd9\u4e2acontain":7,"\u5982\u679c\u6709\u66f4\u590d\u6742\u7684\u4f7f\u7528":28,"\u5982\u679c\u7cfb\u7edf\u4f7f\u7528\u4e0d\u652f\u6301":9,"\u5982\u679c\u8fd0\u884cgpu\u7248\u672c\u7684paddlepaddl":7,"\u5b83\u5305\u542b\u7684\u53c2\u6570\u6709":29,"\u5b89\u88c5\u597d\u7684paddlepaddle\u811a\u672c\u5305\u62ec\u591a\u6761\u547d\u4ee4":22,"\u5b89\u88c5\u5b8c\u6210\u7684paddlepaddle\u4e3b\u4f53\u5305\u62ec\u4e09\u4e2a\u90e8\u5206":7,"\u5b89\u88c5\u5b8c\u6210paddlepaddle\u540e":9,"\u5b89\u88c5\u6559\u7a0b":15,"\u5b89\u88c5\u65b9\u6cd5\u8bf7\u53c2\u8003":7,"\u5b89\u88c5\u7f16\u8bd1\u4f9d\u8d56":4,"\u5b89\u88c5\u7f16\u8bd1paddlepaddle\u9700\u8981\u7684\u4f9d\u8d56":3,"\u5b89\u88c5docker\u9700\u8981\u60a8\u7684\u673a\u5668":7,"\u5b89\u88c5paddlepaddl":6,"\u5b8c\u6210\u591a\u673a\u8bad\u7ec3":22,"\u5b9a\u4e49\u6587\u672c\u4fe1\u606f":15,"\u5b9e\u73b0\u4e86\u6253\u5f00\u6587\u672c\u6587\u4ef6":29,"\u5bc6\u7801\u4e5f\u662froot":7,"\u5bf9\u4e8e\u7528\u6237\u6765\u8bf4":29,"\u5bf9\u4e8e\u7ed9\u5b9a\u7684\u4e00\u6761\u6587\u672c":15,"\u5bf9\u4e8ecuda\u7684toolkit\u6709\u65ad\u70b9\u7eed\u4f20\u548c\u4f20\u8f93\u901f\u5ea6\u8fc7\u5c0f\u91cd\u542f\u4e0b\u8f7d\u7684":1,"\u5bf9\u4e8emnist\u800c\u8a00":29,"\u5bf9\u8be5\u8868\u793a\u8fdb\u884c\u975e\u7ebf\u6027\u53d8\u6362":15,"\u5bf9\u8c61":29,"\u5c06\u4f1a\u6d4b\u8bd5\u914d\u7f6e\u6587\u4ef6\u4e2dtest":15,"\u5c06\u5b57\u5178\u5b58\u5165\u4e86set":29,"\u5c06\u5bbf\u4e3b\u673a\u76848022\u7aef\u53e3\u6620\u5c04\u5230container\u768422\u7aef\u53e3\u4e0a":7,"\u5c06\u6570\u636e\u7ec4\u5408\u6210batch\u8bad\u7ec3":29,"\u5c06\u6587\u4ef6\u7684\u7edd\u5bf9\u8def\u5f84\u6216\u76f8\u5bf9\u8def\u5f84":28,"\u5c06\u8bc4\u8bba\u5206\u4e3a\u597d\u8bc4":15,"\u5c06\u8be5\u53e5\u8bdd\u5305\u542b\u7684\u6240\u6709\u5355\u8bcd\u5411\u91cf\u6c42\u5e73\u5747\u5f97\u5230\u53e5\u5b50\u7684\u8868\u793a":15,"\u5c06ssh\u88c5\u5165\u7cfb\u7edf\u5185\u5e76\u5f00\u542f\u8fdc\u7a0b\u8bbf\u95ee":7,"\u5c31\u50cf\u8fd9\u4e2a\u6837\u4f8b\u4e00\u6837":29,"\u5c31\u662f\u5c06\u8fd9\u4e9b\u52a8\u6001\u5e93\u52a0\u5230\u73af\u5883\u53d8\u91cf\u91cc\u9762":9,"\u5c5e\u6027":29,"\u5dee\u8bc4":15,"\u5e38\u89c1\u7684\u8f93\u51fa\u683c\u5f0f\u4e3a":27,"\u5e76\u4e14\u4f7f\u7528\u5173\u952e\u8bcd":29,"\u5e76\u4e14\u5220\u9664container\u4e2d\u7684\u6570\u636e":7,"\u5e76\u4e14\u6807\u8bb0process\u51fd\u6570\u662f\u4e00\u4e2adataprovid":29,"\u5e76\u4f7f\u7528\u4e86dropout":15,"\u5e76\u572823\u884c\u8fd4\u56de\u7ed9paddlepaddle\u8fdb\u7a0b":29,"\u5e76\u572883\u884c\u8f7d\u5165\u53c2\u6570":33,"\u5e76\u5c06\u6bcf\u884c\u8f6c\u6362\u6210\u548c":29,"\u5e76\u63d0\u4f9b":7,"\u5e76\u63d0\u4f9b\u4e86\u7b80\u5355\u7684cache\u529f\u80fd":29,"\u5e76\u8bbe\u7f6e\u597d\u5bf9\u5e94\u7684\u73af\u5883\u53d8\u91cf":9,"\u5e76\u9010\u6e10\u5c55\u793a\u66f4\u52a0\u6df1\u5165\u7684\u529f\u80fd":15,"\u5e8f\u5217\u6a21\u578b\u6570\u636e\u63d0\u4f9b":28,"\u5e8f\u5217\u6a21\u578b\u662f\u6307\u6570\u636e\u7684\u67d0\u4e00\u7ef4\u5ea6\u662f\u4e00\u4e2a\u5e8f\u5217\u5f62\u5f0f":29,"\u5e93\u7684\u8bdd":9,"\u5f0f":29,"\u5f15\u7528\u7684dataprovider\u662f":29,"\u5f53\u524dlog_period\u4e2abatch\u6240\u6709\u6837\u672c\u7684\u5e73\u5747\u5206\u7c7b\u9519\u8bef\u7387":15,"\u5f53\u524dlog_period\u4e2abatch\u6240\u6709\u6837\u672c\u7684\u5e73\u5747cost":15,"\u5f53\u7136":28,"\u5f88":15,"\u5f97\u5230\u7ed3\u679c":9,"\u5feb\u901f\u5165\u95e8":20,"\u5ff5\u662f":29,"\u60a8\u53ef\u4ee5\u4f7f\u7528":7,"\u60a8\u5c31\u53ef\u4ee5\u8fdc\u7a0b\u7684\u4f7f\u7528paddlepaddle\u5566":7,"\u60a8\u9700\u8981\u5728\u673a\u5668\u4e2d\u5b89\u88c5\u597ddocker":7,"\u60a8\u9700\u8981\u8fdb\u5165\u955c\u50cf\u8fd0\u884cpaddlepaddl":7,"\u60c5\u611f\u5206\u6790":14,"\u60f3\u8981\u8fd0\u884cpaddlepaddl":7,"\u6210\u4e3a\u7ef4\u5ea6\u4e3ahidden":15,"\u6211\u4eec\u4ece\u63d0\u524d\u7ed9\u5b9a\u7684\u7c7b\u522b\u96c6\u5408\u4e2d\u9009\u62e9\u5176\u6240\u5c5e\u7c7b":15,"\u6211\u4eec\u4ee5\u6587\u672c\u5206\u7c7b\u95ee\u9898\u4f5c\u4e3a\u80cc\u666f":15,"\u6211\u4eec\u4f7f\u7528":15,"\u6211\u4eec\u5728\u6b64\u603b":15,"\u6211\u4eec\u5c06\u4ee5\u57fa\u672c\u7684\u903b\u8f91\u56de\u5f52\u7f51\u7edc\u4f5c\u4e3a\u8d77\u70b9":15,"\u6211\u4eec\u5c06\u5728\u540e\u9762\u4ecb\u7ecd\u8bad\u7ec3\u548c\u9884\u6d4b\u7684\u6d41\u7a0b\u7684\u811a\u672c":15,"\u6211\u4eec\u5c06\u8bad\u7ec3\u7684\u547d\u4ee4\u884c\u4fdd\u5b58\u5728\u4e86":15,"\u6211\u4eec\u63d0\u4f9b\u4e866\u4e2adocker":7,"\u6211\u4eec\u63d0\u4f9b\u4e86\u4e00\u4e2a\u5de5\u5177\u7c7bdataproviderwrapperconvert":33,"\u6211\u4eec\u66f4\u63a8\u8350\u4f7f\u7528docker\u955c\u50cf\u6765\u90e8\u7f72paddlepaddle\u73af\u5883":8,"\u6211\u4eec\u8fdb\u5165\u5230\u8bad\u7ec3\u90e8\u5206":15,"\u6211\u4eec\u91c7\u7528\u5355\u5c42lstm\u6a21\u578b":15,"\u6211\u4eec\u968f\u65f6\u603b\u7ed3\u4e86\u5404\u4e2a\u7f51\u7edc\u7684\u590d\u6742\u5ea6\u548c\u6548\u679c":15,"\u6216\u5176\u4ed6":15,"\u6216\u8005":7,"\u6216\u800510g\u8fd9\u6837\u7684\u5355\u4f4d":1,"\u6216\u8005\u6570\u636e\u5e93\u8fde\u63a5\u5730\u5740\u7b49\u7b49":28,"\u6216\u8005\u8bbe\u7f6e\u4e3anone\u7684\u8bdd":28,"\u6216\u8005\u9700\u8981\u66f4\u9ad8\u7684\u6548\u7387":28,"\u6216\u8005\u9ad8\u6027\u80fd\u7684":7,"\u6240\u4ee5":[29,33],"\u6240\u4ee5\u5728cpu\u7684\u8fd0\u7b97\u6027\u80fd\u4e0a\u5e76\u4e0d\u4f1a\u6709\u4e25\u91cd\u7684\u5f71\u54cd":7,"\u6240\u4ee5\u5982\u679c\u60f3\u8981\u5728\u540e\u53f0\u542f\u7528ssh":7,"\u6240\u4ee5\u5c06":29,"\u6240\u4ee5\u8f93\u51fa\u7684valu":33,"\u6240\u4ee5gpu\u5728\u8fd0\u7b97\u6027\u80fd\u4e0a\u4e5f\u4e0d\u4f1a\u6709\u4e25\u91cd\u7684\u5f71\u54cd":7,"\u6240\u4ee5init_hook\u5c3d\u91cf\u4f7f\u7528":29,"\u6240\u6709\u5b57\u7b26\u90fd\u5c06\u8f6c\u6362\u4e3a\u8fde\u7eed\u6574\u6570\u8868\u793a\u7684id\u4f20\u7ed9\u6a21\u578b":15,"\u6240\u6709\u6587\u4ef6\u5217\u8868":29,"\u6240\u6709\u7684":2,"\u6240\u6709\u7684\u4e0b\u8f7d\u7ebf\u7a0b\u5173\u95ed\u65f6":1,"\u6240\u6709\u914d\u7f6e\u5728":15,"\u6240\u8c13\u65f6\u95f4\u6b65\u4fe1\u606f":29,"\u6253\u5370\u7684\u65e5\u5fd7\u53d8\u591a":2,"\u6267\u884c":1,"\u6267\u884c\u5982\u4e0b\u547d\u4ee4\u5373\u53ef\u4ee5\u5173\u95ed\u8fd9\u4e2acontain":7,"\u6267\u884c\u65b9\u6cd5\u5982\u4e0b":7,"\u62a5\u9519":9,"\u62fc\u63a5\u6210\u4e00\u4e2a\u65b0\u7684\u5411\u91cf\u8868\u793a":15,"\u6307\u4ee4\u96c6\u7684cpu\u8fd0\u884cpaddlepaddl":9,"\u6307\u4ee4\u96c6\u7684x86":9,"\u6307\u5b9a\u521d\u59cb\u5316\u6a21\u578b\u8def\u5f84":15,"\u6307\u5b9a\u751f\u6210\u6570\u636e\u7684\u51fd\u6570":15,"\u6307\u5b9a\u8bad\u7ec3":15,"\u6307\u5b9abatch":15,"\u6307\u5b9aoutputs\u8f93\u51fa\u6982\u7387\u5c42":15,"\u6389\u7f16\u8bd1\u76ee\u5f55\u540e":2,"\u63a5\u53e3\u4f7f\u7528\u591a\u7ebf\u7a0b\u8bfb\u53d6\u6570\u636e":29,"\u63a8\u8350\u4f7f\u7528\u5c06\u672c\u5730\u7f51\u5361":7,"\u63a8\u8350\u4f7f\u7528\u6700\u65b0\u7248\u672c\u7684cudnn":2,"\u63a8\u8350\u6e05\u7406":2,"\u63a8\u8350\u76f4\u63a5\u653e\u7f6e\u5230\u8bad\u7ec3\u76ee\u5f55":28,"\u63cf\u8ff0":2,"\u63cf\u8ff0\u4e86docker":1,"\u6548\u679c\u4e00\u81f4":29,"\u6548\u679c\u603b\u7ed3":15,"\u6559\u7a0b":15,"\u6570\u636e":29,"\u6570\u636e\u4f20\u8f93\u65e0\u9700label\u6570\u636e":15,"\u6570\u636e\u5904\u7406python\u6587\u4ef6\u540d":15,"\u6570\u636e\u5982\u4f55\u5b58\u50a8\u7b49\u7b49":29,"\u6570\u636e\u63d0\u4f9b":28,"\u6570\u636e\u6587\u4ef6\u5b58\u653e\u5728":28,"\u6570\u636e\u662f\u7ed9\u4e00\u6bb5\u82f1\u6587\u6587\u672c":29,"\u6570\u636e\u683c\u5f0f\u548c\u8be6\u7ec6\u6587\u6863\u8bf7\u53c2\u8003":15,"\u6587\u4ef6\u4e2d":15,"\u6587\u672c\u4e2d\u7684\u5355\u8bcd\u7528\u7a7a\u683c\u5206\u9694":15,"\u6587\u672c\u4fe1\u606f\u5c31\u662f\u4e00\u4e2a\u5e8f\u5217":29,"\u6587\u672c\u5206\u7c7b\u95ee\u9898":15,"\u6587\u672c\u5377\u79ef\u5206\u4e3a\u4e09\u4e2a\u6b65\u9aa4":15,"\u6587\u672c\u751f\u6210":14,"\u65b9\u4fbf\u8c03\u8bd5\u4f7f\u7528":22,"\u65b9\u4fbf\u90e8\u7f72\u5206\u53d1":22,"\u65e0\u9700label\u76f8\u5173\u7684\u5c42":15,"\u65f6\u5e8f\u6a21\u578b\u5373\u4e3arnn\u6a21\u578b":15,"\u65f6\u5e8f\u6a21\u578b\u5747\u4f7f\u7528\u8be5\u811a\u672c":15,"\u662f":2,"\u662f\u4e00\u4e2abatch":29,"\u662f\u4e00\u4e2apython\u7684":29,"\u662f\u4e00\u4e2aswig\u5c01\u88c5\u7684paddlepaddle\u5305":7,"\u662f\u4e0d\u662f\u5f88\u7b80\u5355\u5462":29,"\u662f\u4e2adataprovider\u662f\u4e0d\u662f\u8981\u505ashuffl":29,"\u662f\u4ec0\u4e48\u4e5f\u6ca1\u5173\u7cfb":29,"\u662f\u4ece\u8bad\u7ec3\u914d\u7f6e\u4f20\u5165\u7684dict\u5bf9\u8c61":29,"\u662f\u521d\u59cb\u5316\u65f6\u8c03\u7528\u7684\u51fd\u6570":29,"\u662f\u540c\u4e00\u4e2a\u5bf9\u8c61":29,"\u662f\u5426\u4f7f\u7528\u53cc\u7cbe\u5ea6\u6d6e\u70b9\u6570":2,"\u662f\u5426\u4f7f\u7528\u8fd0\u884c\u65f6\u52a8\u6001\u52a0\u8f7dcuda\u52a8\u6001\u5e93":2,"\u662f\u5426\u4f7f\u7528gflag":2,"\u662f\u5426\u4f7f\u7528glog":2,"\u662f\u5426\u5185\u5d4cpython\u89e3\u91ca\u5668":2,"\u662f\u5426\u5bfb\u627e\u5230cuda\u5de5\u5177\u94fe":2,"\u662f\u5426\u5f00\u542f\u5355\u5143\u6d4b\u8bd5":2,"\u662f\u5426\u5f00\u542f\u8ba1\u65f6\u529f\u80fd\u5f00\u542f\u8ba1\u65f6\u529f\u80fd\u4f1a\u5bfc\u81f4\u8fd0\u884c\u7565\u6162":2,"\u662f\u5426\u5f00\u542fgpu\u529f\u80fd":1,"\u662f\u5426\u5f00\u542frdma\u652f\u6301":2,"\u662f\u5426\u7f16\u8bd1\u4e2d\u6587\u6587\u6863":2,"\u662f\u5426\u7f16\u8bd1\u542b\u6709avx\u6307\u4ee4\u96c6\u7684paddlepaddle\u4e8c\u8fdb\u5236":2,"\u662f\u5426\u7f16\u8bd1\u65f6\u8fdb\u884c\u4ee3\u7801\u98ce\u683c\u68c0\u67e5":2,"\u662f\u5426\u7f16\u8bd1\u82f1\u6587\u6587\u6863":2,"\u662f\u5426\u7f16\u8bd1gpu\u652f\u6301":2,"\u662f\u5426\u7f16\u8bd1python\u7684swig\u63a5\u53e3":2,"\u662f\u5728\u8fd0\u884c\u65f6\u6267\u884c\u7684":29,"\u662f\u6570\u636e\u7f13\u5b58\u7684\u7b56\u7565":29,"\u662f\u6570\u636e\u8f93\u5165\u683c\u5f0f":29,"\u662f\u8bbe\u7f6e\u8fd9\u4e2adataprovider\u8fd4\u56de\u4ec0\u4e48\u6837\u7684\u6570\u636e":29,"\u662f\u8bbe\u7f6edataprovider\u5728\u5185\u5b58\u4e2d\u6682\u5b58\u7684\u6570\u636e\u6761\u6570":29,"\u662fpaddlepaddle\u8d1f\u8d23\u63d0\u4f9b\u6570\u636e\u7684\u6a21\u5757":28,"\u662fpython\u7684\u4e00\u4e2a\u5173\u952e\u8bcd":29,"\u663e":15,"\u66f4\u591a\u7684\u5b89\u88c5\u5305paddlepaddle\u4f1a\u5728\u8fd1\u671f\u63d0\u4f9b":9,"\u66f4\u8be6\u7ec6\u7528\u4f8b\u8bf7\u53c2\u8003\u6587\u6863":15,"\u66f4\u8be6\u7ec6\u7684\u4ecb\u7ecd\u8bf7\u53c2\u8003\u5404\u4e2a\u547d\u4ee4\u7684\u547d\u4ee4\u884c\u53c2\u6570\u6587\u6863":22,"\u66f4\u8be6\u7ec6\u7684\u7f51\u7edc\u914d\u7f6e":15,"\u6700\u4f4e\u7ebf\u7a0b\u7684\u4e0b\u8f7d\u901f\u5ea6":1,"\u6709\u503c\u7684\u4f4d\u7f6e\u53ea\u80fd\u53d61":29,"\u6709\u503c\u7684\u90e8\u5206\u53ef\u4ee5\u662f\u4efb\u4f55\u6d6e\u70b9\u6570":29,"\u6709\u90e8\u5206\u53c2\u6570\u662fpaddle\u81ea\u52a8\u751f\u6210\u7684":29,"\u672c\u5730\u78c1\u76d8\u4e2d":28,"\u672c\u8282\u6211\u4eec\u5c06\u4e13\u6ce8\u4e8e\u7f51\u7edc\u7ed3\u6784\u7684\u4ecb\u7ecd":15,"\u6765\u63a5\u53d7\u4e0d\u4f7f\u7528\u7684":29,"\u6765\u81ea\u5b9a\u4e49\u4f20\u6570\u636e\u7684\u8fc7\u7a0b":28,"\u6765\u8bf4\u660e\u7b80\u5355\u7684pydataprovider\u5982\u4f55\u4f7f\u7528":29,"\u6765\u8fdb\u884c\u8bad\u7ec3":7,"\u6765\u914d\u7f6ecudnn\u7684\u5b89\u88c5\u8def\u5f84":2,"\u6784\u9020gradientmachin":33,"\u6807\u7b7e\u662f0":29,"\u6837\u4f8b\u6570\u636e\u4e3a":29,"\u6837\u4f8b\u6570\u636e\u5982\u4e0b":29,"\u6837\u672c":29,"\u683c\u5f0f\u5982\u4e0b":15,"\u68d2":15,"\u6a21\u578b\u5b58\u50a8\u8def\u5f84":15,"\u6a21\u578b\u8bad\u7ec3\u4f1a\u770b\u5230\u8fd9\u6837\u7684\u65e5\u5fd7":15,"\u6a21\u578b\u914d\u7f6e":20,"\u6b22\u8fce\u5927\u5bb6\u8d21\u732e\u5404\u4e2a\u53d1\u884c\u7248\u7684\u5b89\u88c5\u5305":9,"\u6b22\u8fce\u8d21\u732e\u66f4\u591a\u7684\u5b89\u88c5\u5305":8,"\u6b63\u6837\u672c":15,"\u6bcf\u4e00\u4e2a\u4efb\u52a1\u6d41\u7a0b\u90fd\u53ef\u4ee5\u5206\u4e3a\u5982\u4e0b5\u4e2a\u57fa\u7840\u90e8\u5206":15,"\u6bcf\u4e00\u884c":29,"\u6bcf\u4e2a\u5c42\u90fd\u6709\u4e00\u4e2a\u6216\u591a\u4e2ainput":15,"\u6bcf\u4e2apass\u7684\u7b2c0\u4e2abatch\u5230\u5f53\u524dbatch\u6240\u6709\u6837\u672c\u7684\u5e73\u5747\u5206\u7c7b\u9519\u8bef\u7387":15,"\u6bcf\u4e2apass\u7684\u7b2c0\u4e2abatch\u5230\u5f53\u524dbatch\u6240\u6709\u6837\u672c\u7684\u5e73\u5747cost":15,"\u6bcf\u6b21\u90fd\u4f1a\u4ecepython\u7aef\u8bfb\u53d6\u6570\u636e":29,"\u6bcf\u884c\u4fdd\u5b58\u4e00\u6761\u6837\u672c":15,"\u6bcf\u9694\u591a\u5c11batch\u6253\u5370\u4e00\u6b21\u65e5\u5fd7":15,"\u6bd4\u5982\u901a\u8fc7\u7528\u6237\u5bf9\u7535\u5b50\u5546\u52a1\u7f51\u7ad9\u8bc4\u8bba":15,"\u6bd4\u8f83\u53ef\u80fd\u7684\u547d\u4ee4\u5982\u4e0b":9,"\u6ca1\u6709\u8bbe\u7f6e":9,"\u6ce8\u610f":[1,8,29],"\u6d41\u7a0b\u5982\u4e0b":15,"\u6d4b\u8bd5\u6570\u636e":15,"\u6d4b\u8bd5\u7684\u65f6\u5019\u9ed8\u8ba4\u4e0dshuffl":29,"\u6d4b\u8bd5\u811a\u672c\u5982\u4e0b":15,"\u6e90\u7801\u548c\u8fd0\u884cdemo\u7684\u5fc5\u8981\u4f9d\u8d56":7,"\u6e90\u7801\u6839\u76ee\u5f55":1,"\u6fc0\u6d3b\u51fd\u6570\u7c7b\u578b":15,"\u7136\u540e\u6267\u884c\u5982\u4e0b\u547d\u4ee4\u5373\u53ef\u5b8c\u6210\u5b89\u88c5":9,"\u7136\u540e\u8fd0\u884c\u8fd9\u4e2acontainer\u5373\u53ef":7,"\u751f\u6210\u5404\u4e2a\u5e73\u53f0\u7684makefil":2,"\u75280\u548c1\u8868\u793a":29,"\u7528\u4e86\u4e24\u4e2a\u6708\u4e4b\u540e\u8fd9\u4e2a\u663e\u793a\u5668\u5c4f\u5e55\u788e\u4e86":15,"\u7528\u6237\u4e5f\u53ef\u4ee5\u5728c":28,"\u7528\u6237\u53ef\u4ee5\u4f7f\u7528python\u7684":28,"\u7528\u6237\u540d\u4e3a":7,"\u7528\u6237\u5728dataprovider\u4e2d\u9700\u8981\u5b9e\u73b0\u5982\u4f55\u8bbf\u95ee\u5176\u4e2d\u6bcf\u4e00\u4e2a\u6587\u4ef6":28,"\u7528\u6237\u5b9a\u4e49\u7684\u53c2\u6570\u4f7f\u7528args\u5728\u8bad\u7ec3\u914d\u7f6e\u4e2d\u8bbe\u7f6e":29,"\u7528\u6237\u63a5\u53e3":20,"\u7528\u6237\u9700\u8981\u5148\u5c06paddlepaddle\u5b89\u88c5\u5305\u4e0b\u8f7d\u5230\u672c\u5730":9,"\u7528\u6765\u505a\u9884\u6d4b\u548c\u7b80\u5355\u7684\u5b9a\u5236\u5316":7,"\u7528\u6765\u8f6c\u6362\u6570\u636e":33,"\u7531\u4e8e\u6570\u636e\u662f\u4e24\u4e2a":33,"\u7531\u4e8edocker\u662f\u57fa\u4e8e\u5bb9\u5668\u7684\u8f7b\u91cf\u5316\u865a\u62df\u65b9\u6848":7,"\u7531\u4e8epaddlepaddle\u7684docker\u955c\u50cf\u5e76\u4e0d\u5305\u542b\u4efb\u4f55\u9884\u5b9a\u4e49\u7684\u8fd0\u884c\u547d\u4ee4":7,"\u7531\u6613\u5230\u96be\u5c55\u793a4\u79cd\u4e0d\u540c\u7684\u7f51\u7edc\u914d\u7f6e":15,"\u7684\u4e00\u4e9b\u53c2\u6570":28,"\u7684\u5b89\u88c5\u6587\u6863":7,"\u7684\u6587\u6863":29,"\u7684\u65b9\u5f0f\u5199\u5728train":28,"\u7684\u65f6\u5019\u5982\u679c\u62a5\u4e00\u4e9b\u4f9d\u8d56\u672a\u627e\u5230\u7684\u9519\u8bef\u662f\u6b63\u5e38\u7684":9,"\u7684\u662f":29,"\u7684\u673a\u5668\u4e0a\u8fdb\u884c":1,"\u7684\u7f51\u6865\u6765\u8fdb\u884c\u7f51\u7edc\u901a\u4fe1":7,"\u7684demo\u5b66\u4e60\u5982\u4f55\u8fdb\u884c\u591a\u673a\u8bad\u7ec3":15,"\u7684docker\u53ef\u80fd\u7f3a\u4e4f":1,"\u7684matrix":33,"\u7684python\u5305\u662fpaddlepaddle\u7684\u8bad\u7ec3\u4e3b\u8981\u7a0b\u5e8f":7,"\u7684python\u5305\u6765\u505a\u914d\u7f6e\u6587\u4ef6\u89e3\u6790\u7b49\u5de5\u4f5c":7,"\u7684python\u9884\u6d4b\u8fc7\u7a0b":15,"\u76ee\u524dpaddlepaddle\u7684\u5b89\u88c5\u5305\u53ea\u652f\u6301":9,"\u76ee\u524dpaddlepaddle\u7684\u5b89\u88c5\u5305\u8fd8\u5904\u5728pr":8,"\u76ee\u5f55":15,"\u76ee\u5f55\u4e0b":1,"\u76f4\u63a5\u63d0\u53d6\u51fa\u795e\u7ecf\u7f51\u7edcoutput\u5c42\u7684\u8f93\u51fa\u7ed3\u679c":33,"\u76f8\u5173\u547d\u4ee4\u4e3a":7,"\u76f8\u5173\u7684\u6982":29,"\u76f8\u5bf9\u4e8epaddlepaddle\u7a0b\u5e8f\u8fd0\u884c\u65f6\u7684\u8def\u5f84":28,"\u77e5\u9053\u5982\u4f55\u4ece":29,"\u793a":15,"\u7a0b\u5e8f\u6216\u8005\u81ea\u5b9a\u4e49\u4e00\u4e2a\u542b\u6709\u542f\u52a8\u811a\u672c\u7684imag":7,"\u7aef\u81ea\u5b9a\u4e49\u4e00\u4e2a":28,"\u7b2c\u4e00\u4e2a\u53c2\u6570\u662f":29,"\u7b2c\u4e00\u4e2apass\u4f1a\u4ecepython\u7aef\u8bfb\u53d6\u6570\u636e":29,"\u7b2c\u4e00\u6bb5\u6570\u636e\u4e3a\u8fd9\u5f20\u56fe\u7247\u7684label":29,"\u7b2c\u4e8c\u4e2a\u53c2\u6570\u662ffilenam":29,"\u7b2c\u4e8c\u6bb5\u6570\u636e\u4e3a\u8fd9\u4e2a\u56fe\u7247\u7684\u50cf\u7d20\u503c":29,"\u7b80\u5355\u4f18\u5316":1,"\u7b80\u5355\u7684\u4f7f\u7528":28,"\u7b80\u5355\u7684\u4f7f\u7528\u573a\u666f":28,"\u7b80\u5355\u7684\u4f7f\u7528\u6837\u4f8b\u4e3a":1,"\u7b80\u5355\u7684\u542b\u6709ssh\u7684dockerfile\u5982\u4e0b":7,"\u7b80\u5355\u7684pydataprovider\u6837\u4f8b\u5c31\u8bf4\u660e\u5b8c\u6bd5\u4e86":29,"\u7c7b\u522bid":15,"\u7c7b\u522bid\u7684\u6570\u636e\u7c7b\u578b":15,"\u7c7b\u578b\u6765\u8bbe\u7f6e":29,"\u7ec3":22,"\u7ed3\u4e0a\u8ff0\u7f51\u7edc\u7ed3\u6784\u5728amazon":15,"\u7ef4\u5ea6\u4e3aword":15,"\u7ef4\u5ea6\u662f\u7c7b\u522b\u4e2a\u6570":15,"\u7ef4\u5ea6\u662f\u8bcd\u5178\u5927\u5c0f":15,"\u7f13\u5b58\u8bad\u7ec3\u6570\u636e\u5230\u5185\u5b58":29,"\u7f16\u8bd1\u53c2\u6570\u9009\u9879\u6587\u4ef6":27,"\u7f16\u8bd1\u6587\u6863":9,"\u7f16\u8bd1\u9009\u9879":2,"\u7f16\u8bd1\u9009\u9879\u5217\u8868\u5982\u4e0b":2,"\u7f16\u8bd1paddlepaddl":9,"\u7f16\u8bd1paddlepaddle\u7684gpu\u7248\u672c\u5e76\u4e0d\u9700\u8981\u4e00\u5b9a\u5728\u5177\u6709gpu":1,"\u7f51\u7edc\u540d\u79f0":15,"\u7f51\u7edc\u914d\u7f6e":15,"\u7f6e\u8fd9\u4e9b\u53d8\u91cf":2,"\u800c":7,"\u800c\u4e09\u79cd\u5e8f\u5217\u6a21\u5f0f\u4e3a":29,"\u800c\u4e0d\u4f7f\u7528docker":7,"\u800c\u4e0d\u7528\u5173\u5fc3\u6570\u636e\u5982\u4f55\u4f20\u8f93\u7ed9paddlepaddl":29,"\u800c\u5728":[2,29],"\u800c\u6ca1\u6709\u6d4b\u8bd5\u6570\u636e":29,"\u800c\u7279\u5f81\u5373\u4e3a":29,"\u800c\u8fd9\u4e2a\u4e00\u822c\u8bf4\u660epaddlepaddle\u5df2\u7ecf\u5b89\u88c5\u5b8c\u6bd5\u4e86":9,"\u800c\u975e\u9759\u6001\u52a0\u8f7dcuda\u52a8\u6001\u5e93":2,"\u800cgpu\u7684\u9a71\u52a8\u548c\u8bbe\u5907\u5168\u90e8\u6620\u5c04\u5230\u4e86\u5bb9\u5668\u5185":7,"\u800cpaddlepaddle\u8fdb\u7a0b\u5e2e\u52a9\u7528\u6237\u505a\u4e86":29,"\u811a\u672c":7,"\u811a\u672c\u53ef\u4ee5\u542f\u52a8paddlepaddle\u7684\u8bad\u7ec3\u8fdb\u7a0b\u548cpserv":7,"\u811a\u672c\u548c":7,"\u811a\u672c\u7c7b\u4f3c\u4e8e":7,"\u81ea\u5b9a\u4e49\u4e00\u4e2adataprovid":28,"\u81f3\u5c11\u5177\u67093":7,"\u81f3\u6b64":[7,29],"\u83b7\u53d6\u5229\u7528one":15,"\u83b7\u53d6\u6bcf\u4e2a\u5355\u8bcd\u5de6\u53f3\u5404k\u4e2a\u8fd1\u90bb":15,"\u83b7\u53d6\u8be5\u6761\u6837\u672c\u7c7b\u522bid":15,"\u8868\u793a\u6574\u6570\u6807\u7b7e":29,"\u8868\u793a\u662f\u5426\u5141\u8bb8paddle\u6682\u5b58\u7565\u5fae\u591a\u4f59pool_size\u7684\u6570\u636e":29,"\u8868\u793a\u7a00\u758f\u7684\u5411\u91cf":29,"\u8868\u793a\u7a00\u758f\u7684\u96f6\u4e00\u5411\u91cf":29,"\u8868\u793a\u7a20\u5bc6\u7684\u6d6e\u70b9\u6570\u5411\u91cf":29,"\u8868\u793a\u8fc7\u4e8620\u4e2abatch":15,"\u8868\u793a\u8fc7\u4e862560\u4e2a\u6837\u672c":15,"\u8868\u793a\u8fd9\u4e2adataprovider\u662f\u8bad\u7ec3\u7528\u7684dataprovider\u6216\u8005\u6d4b\u8bd5\u7528\u7684":29,"\u89e3\u51b3\u529e\u6cd5\u662f\u5c06cuda":9,"\u89e3\u51b3\u65b9\u6cd5\u5f88\u7b80\u5355":9,"\u89e3\u91ca":15,"\u8ba9\u795e\u7ecf\u7f51\u7edc\u53ef\u4ee5\u8fdb\u884c\u8bad\u7ec3":28,"\u8bad\u7ec3":7,"\u8bad\u7ec3\u6587\u4ef6\u5217\u8868":28,"\u8bad\u7ec3\u65f6\u6240\u9700\u8bbe\u7f6e\u7684\u4e3b\u8981\u53c2\u6570\u5982\u4e0b":15,"\u8bad\u7ec3\u7684\u65f6\u5019\u9ed8\u8ba4shuffl":29,"\u8bad\u7ec3\u811a\u672c":15,"\u8bad\u7ec3\u811a\u672c\u5728":15,"\u8bad\u7ec3\u8f6e\u6b21":15,"\u8bb2\u6570\u636e\u53d1\u9001\u7ed9paddlepaddl":29,"\u8bbe\u7f6e\u4e0b\u5217\u7f16\u8bd1\u9009\u9879\u65f6":2,"\u8bbe\u7f6e\u6210":29,"\u8bbe\u7f6e\u6210\u4e86\u5e8f\u5217":29,"\u8bbe\u7f6e\u8f93\u5165\u7c7b\u578b":29,"\u8bc4\u4f30\u4ea7\u54c1\u7684\u8d28\u91cf":15,"\u8bcd\u6027\u6807\u6ce8":14,"\u8be5\u5c42\u795e\u7ecf\u5143\u4e2a\u6570":15,"\u8be5\u6570\u636e":29,"\u8be5\u6784\u5efa\u811a\u672c\u5145\u5206\u8003\u8651\u4e86\u7f51\u7edc\u4e0d\u7a33\u5b9a\u7684\u60c5\u51b5":1,"\u8be5\u6a21\u578b\u4f9d\u7136\u662f\u4f7f\u7528\u903b\u8f91\u56de\u5f52\u5206\u7c7b\u7f51\u7edc\u7684\u6846\u67b6":15,"\u8be5\u76ee\u5f55\u4e0b\u6709\u4e24\u4e2a\u6587\u4ef6":1,"\u8be5\u811a\u672c\u7684\u4f7f\u7528\u65b9\u6cd5\u662f":1,"\u8be5image\u57fa\u4e8eubuntu":1,"\u8be5image\u7684\u6784\u5efa\u5728dock":1,"\u8be6\u7ec6\u7684\u53c2\u6570\u89e3\u91ca\u5982\u4e0b\u9762\u8868\u683c":15,"\u8be6\u7ec6\u7684cmake\u4f7f\u7528\u65b9\u6cd5\u53ef\u4ee5\u53c2\u8003":2,"\u8bf4\u660e":2,"\u8bf4\u660e\u547d\u4ee4\u884c\u53c2\u6570":9,"\u8bf7\u53c2\u8003":[7,9,29],"\u8bf7\u53c2\u8003\u4e0b\u8282refer":29,"\u8bf7\u53c2\u8003\u4e0b\u8ff0\u6587\u7ae0":28,"\u8bf7\u53c2\u8003pydataproviderwrapper\u7684\u6587\u6863":33,"\u8bf7\u5b89\u88c5cuda":9,"\u8bfb\u5165\u89e3\u6790\u8bad\u7ec3\u914d\u7f6e":33,"\u8bfb\u53d6\u6570\u636e":29,"\u8bfb\u53d6config":33,"\u8c03\u7528":2,"\u8d1f\u6837\u672c":15,"\u8d1f\u8d23\u591a\u673a\u8bad\u7ec3\u4e2d\u7684\u53c2\u6570\u805a\u5408\u5de5\u4f5c":22,"\u8d1f\u9762\u60c5\u7eea\u4e24\u7c7b":29,"\u8def\u5f84\u53d8\u91cf\u4e3a":2,"\u8f93\u5165n\u4e2a\u5355\u8bcd":15,"\u8f93\u51fa\u4e3an\u4e2aword":15,"\u8fd0\u884c":[7,9],"\u8fd0\u884c\u4f7f\u7528\u7684cudnn\u5c3d\u91cf\u662f\u540c\u4e00\u4e2a\u7248\u672c":2,"\u8fd0\u884c\u8fd9\u4e2acontain":7,"\u8fd0\u884cpaddlepaddle\u7684gpu\u7248\u672c\u4e00\u5b9a\u8981\u5728\u5177\u6709cuda\u7684\u673a\u5668\u4e0a\u8fd0\u884c":1,"\u8fd4\u56de\u6570\u636e\u5728paddlepaddle\u4e2d\u662f\u4ec5\u4ec5\u8fd4\u56de\u4e00\u6761\u5b8c\u6574\u7684\u8bad\u7ec3\u6837\u672c":29,"\u8fd4\u56de\u7684\u987a\u5e8f\u9700\u8981\u548c":29,"\u8fd4\u56debatch_size\u7684\u5927\u5c0f":29,"\u8fd9":15,"\u8fd93\u4e2a\u5b50\u6b65\u9aa4\u53ef\u914d\u7f6e\u4e3a":15,"\u8fd9\u4e2a\u51fd\u6570\u4ee5\u4e00\u6761\u6570\u636e\u4e3a\u53c2\u6570":29,"\u8fd9\u4e2a\u51fd\u6570\u4f1a\u5728":29,"\u8fd9\u4e2a\u51fd\u6570\u5728\u521d\u59cb\u5316\u7684\u65f6\u5019\u4f1a\u88ab\u8c03\u7528":29,"\u8fd9\u4e2a\u51fd\u6570\u7684\u53c2\u6570\u662f":29,"\u8fd9\u4e2a\u521d\u59cb\u5316\u51fd\u6570\u5177\u6709\u5982\u4e0b\u53c2\u6570":29,"\u8fd9\u4e2a\u53c2\u6570\u5728\u8fd9\u4e2a\u6837\u4f8b\u91cc\u6ca1\u6709\u4f7f\u7528":29,"\u8fd9\u4e2a\u53c2\u6570\u88abpaddlepaddle\u8fdb\u7a0b\u4f20\u5165":29,"\u8fd9\u4e2a\u548c\u5728":29,"\u8fd9\u4e2a\u58f0\u660e\u57fa\u672c\u4e0a\u548cmnist\u7684\u6837\u4f8b\u4e00\u81f4":29,"\u8fd9\u4e2a\u5b57\u5178\u53ef\u4ee5\u5728":29,"\u8fd9\u4e2a\u5bf9\u8c61\u548cprocess\u7684\u7b2c\u4e00\u4e2a\u53c2\u6570\u4e00\u81f4":29,"\u8fd9\u4e2a\u5de5\u5177\u7c7b\u63a5\u6536\u548cpydataproviderwrapper\u4e00\u6837\u7684\u8f93\u5165\u6570\u636e":33,"\u8fd9\u4e2a\u5e8f\u5217\u6a21\u578b\u6bd4\u8f83\u590d\u6742":29,"\u8fd9\u4e2a\u63a5\u53e3\u5e76\u4e0d\u7528\u6237\u53cb\u597d":33,"\u8fd9\u4e2a\u663e\u793a\u5668\u5f88\u68d2":15,"\u8fd9\u4e2a\u6a21\u5757\u4e2d\u7684":29,"\u8fd9\u4e2a\u8bbe\u7f6e\u4e3a":29,"\u8fd9\u4e2a\u8f6f\u4ef6\u5305\u6587\u6863\u76f8\u5bf9\u5b8c\u5584":33,"\u8fd9\u4e9b\u53c2\u6570\u5305\u62ecpaddle\u5b9a\u4e49\u7684\u53c2\u6570":29,"\u8fd9\u4e9b\u53d8":2,"\u8fd9\u4e9b\u53d8\u91cf\u53ea\u5728\u7b2c\u4e00\u6b21cmake\u7684\u65f6\u5019\u6709\u6548":2,"\u8fd9\u4e9b\u53d8\u91cf\u5747\u53ef\u4ee5\u4f7f\u7528":2,"\u8fd9\u4e9b\u6d41\u7a0b\u4e2d\u7684\u6570\u636e\u4e0b\u8f7d":15,"\u8fd9\u6837\u505a\u53ef\u4ee5\u907f\u514d\u5f88\u591a\u6b7b\u9501\u95ee\u9898":29,"\u8fd9\u884c\u7684\u4f5c\u7528\u662f\u8bbe\u7f6edataprovider\u7684\u4e00\u4e9b\u5c5e\u6027":29,"\u8fd9\u91cc":29,"\u8fd9\u91cc\u4e3e\u4f8b\u7684\u6570\u636e\u662f\u82f1\u6587\u60c5\u611f\u5206\u7c7b\u7684\u6570\u636e":29,"\u8fd9\u91cc\u4ee5mnist\u624b\u5199\u8bc6\u522b\u4e3a\u4f8b":29,"\u8fd9\u91cc\u53ef\u4ee5\u53c2\u8003paddle\u7684":27,"\u8fd9\u91cc\u6211\u4eec\u4f7f\u7528\u7b80\u5355\u7684\u6587\u672c\u6587\u4ef6\u8868\u793amnist\u56fe\u7247":29,"\u8fd9\u91cc\u6307\u5b9a\u8bcd\u5178":15,"\u8fd9\u91cc\u6ca1\u6709\u4ecb\u7ecd\u591a\u673a\u5206\u5e03\u5f0f\u8bad\u7ec3":15,"\u8fd9\u91cc\u7684":29,"\u8fd9\u91cc\u7684\u8f93\u5165\u7279\u5f81\u662f\u8bcdid\u7684\u5e8f\u5217":29,"\u8fd9\u91cc\u8bbe\u7f6e\u7684\u662f\u8fd4\u56de\u4e00\u4e2a":29,"\u8fd9\u91cc\u8bf4\u660e\u4e86\u8bad\u7ec3\u6570\u636e\u662f":29,"\u8fd9\u91cc\u91c7\u7528adam\u4f18\u5316\u65b9\u6cd5":15,"\u8fdb\u5165\u8be5\u6e90\u7801\u76ee\u5f55":1,"\u8fdb\u5165docker":7,"\u8fdc\u7a0b\u8bbf\u95ee":7,"\u8fde\u63a5\u8bf7\u53c2\u8003":15,"\u9009\u9879":2,"\u901a\u8fc7\u7f16\u8bd1\u65f6\u6307\u5b9a\u8def\u5f84\u6765\u5b9e\u73b0\u5f15\u7528\u5404\u79cdbla":2,"\u903b\u8f91\u56de\u5f52":15,"\u90a3\u4e48":29,"\u90a3\u4e48\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d":28,"\u90a3\u4e48\u5bf9\u5e94\u7684dataprovider\u65e2\u4e3a":29,"\u90a3\u4e48\u8fd9\u4e2a\u4e0b\u8f7d\u7ebf\u7a0b\u5c06\u4f1a\u5173\u95ed":1,"\u90a3\u4e48\u9700\u8981\u4ece\u6e90\u7801":9,"\u914d\u7f6e\u4e86":29,"\u914d\u7f6e\u53c2\u6570\u914d\u7f6e\u7ed9dataprovider\u7684":29,"\u914d\u7f6e\u6587\u4ef6":15,"\u91cc\u4f1a\u7ee7\u7eed\u5b89\u88c5":9,"\u91cc\u63d0\u4f9b\u4e86\u6570\u636e\u4e0b\u8f7d\u811a\u672c":15,"\u91cc\u9762\u8bfb\u53d6":29,"\u91cf\u4e5f\u53ef\u4ee5\u901a\u8fc7\u8c03\u7528cmake\u547d\u4ee4\u524d\u901a\u8fc7\u73af\u5883\u53d8\u91cf\u6307\u5b9a":2,"\u9488\u5bf9\u672c\u95ee\u9898":15,"\u94fe\u63a5\u4f55\u79cdblas\u7b49\u7b49":2,"\u9519\u8bef\u7387":15,"\u95f4\u9694":29,"\u9664\u4e86":29,"\u9664\u8fc7data\u5c42":15,"\u9700\u8981\u53c2\u8003":7,"\u9700\u8981\u6ce8\u610f":29,"\u9700\u8981\u6ce8\u610f\u7684\u662f":[2,9],"\u9884\u6d4b\u6570\u636e\u6307\u5b9atest":15,"\u9884\u6d4b\u7ed3\u679c\u4ee5\u6587\u672c\u7684\u5f62\u5f0f\u4fdd\u5b58\u5728":15,"\u9884\u6d4b\u811a\u672c":15,"\u9884\u6d4bid":15,"\u989d\u5916\u7684\u53c2\u6570":15,"\u9996\u5148\u5217\u4e3e\u903b\u8f91\u56de\u5f52\u7f51\u7edc":15,"\u9996\u5148\u6211\u4eec\u5c06\u8fd9\u4e2a\u6570\u636e\u6587\u4ef6":29,"\u9996\u5148\u8bf7\u53c2\u8003":15,"\u9aa43":15,"\u9ed8\u8ba4\u4e00\u4e2apass\u4fdd\u5b58\u4e00\u6b21\u6a21\u578b":15,"\u9ed8\u8ba4\u503c":2,"\u9ed8\u8ba4\u5355\u4f4d\u662fbyte":1,"\u9ed8\u8ba4\u60c5\u51b5\u4e0b\u4e00\u6761\u6570\u636e":29,"__main__":33,"__name__":33,"adamax\u7b49":15,"amazon\u7535\u5b50\u4ea7\u54c1\u8bc4\u8bba\u6570\u636e":15,"api\u9884\u6d4b":15,"argument\u4f20\u5165":29,"argument\u5f62\u5f0f\u4f20\u5165":29,"atlas\u5e93\u7684\u8def\u5f84":2,"batches\u8bbe\u7f6e\u6bcf\u9694\u591a\u5c11batch\u4fdd\u5b58\u4e00\u6b21\u6a21\u578b":15,"bool\u53c2\u6570":29,"case":[15,31],"cd\u5230\u542b\u6709dockerfile\u7684\u8def\u5f84\u4e2d":7,"cmake\u53ef\u4ee5\u5c06cmake\u9879\u76ee\u6587\u4ef6":2,"cmake\u662f\u4e00\u4e2a\u8de8\u5e73\u53f0\u7684\u7f16\u8bd1\u811a\u672c":2,"cmake\u7684\u5b98\u65b9\u6587\u6863":2,"cmake\u7f16\u8bd1\u65f6\u4f1a\u9996\u5148\u5728\u7cfb\u7edf\u8def\u5f84":2,"config\u6587\u4ef6\u627e\u4e0d\u5230":8,"container\u540e":7,"cuda\u76f8\u5173\u7684driver\u548c\u8bbe\u5907\u6620\u5c04\u8fdbcontainer\u4e2d":7,"d\u547d\u4ee4\u5373\u53ef":2,"d\u547d\u4ee4\u6307\u5b9a":2,"dataprovider\u521b\u5efa\u7684\u65f6\u5019\u6267\u884c":29,"dataprovider\u53ef\u4ee5\u662f":29,"dataprovider\u63d0\u4f9b\u4e86\u4e24\u79cd\u7b80\u5355\u7684cache\u7b56\u7565":29,"dataprovider\u7684\u5177\u4f53\u7528\u6cd5\u548c\u5982\u4f55\u5b9e\u73b0\u4e00\u4e2a\u65b0\u7684dataprovid":28,"dim\u7684\u65b0\u7684\u5411\u91cf":15,"dim\u7ef4\u5ea6\u5411\u91cf":15,"docker\u662f\u4e00\u4e2a\u57fa\u4e8e\u5bb9\u5668\u7684\u8f7b\u91cf\u7ea7\u865a\u62df\u73af\u5883":7,"docker\u7684\u5b98\u65b9\u6587\u6863":7,"dockerfile\u548cbuild":1,"dockerfile\u662fdock":1,"dockerfile\u7684\u6587\u6863":7,"dockerfile\u7684\u6700\u4f73\u5b9e\u8df5":7,"driver\u627e\u4e0d\u5230":8,"driver\u6dfb\u52a0\u5230ld_library_path\u4e2d":9,"elec\u6d4b\u8bd5\u96c6":15,"embedding\u6a21\u578b\u9700\u8981\u7a0d\u5fae\u6539\u53d8\u6570\u636e\u63d0\u4f9b\u7684\u811a\u672c":15,"export":[2,7,9],"f\u4ee3\u8868\u4e00\u4e2a\u6d6e\u70b9\u6570":29,"float":29,"forwardtest\u662f\u4e00\u4e2a\u5de5\u5177\u7c7b":33,"gpu\u53cc\u7f13\u5b58":29,"gpu\u7248\u672c\u4e8c\u8fdb\u5236":2,"gru\u6a21\u578b":15,"gru\u6a21\u578b\u914d\u7f6e":15,"i\u4ee3\u8868\u4e00\u4e2a\u6574\u6570":29,"id\u4e3a0\u7684\u6982\u7387":15,"id\u4e3a1\u7684\u6982\u7387":15,"image\u6784\u5efa\u6e90\u7801\u653e\u7f6e\u5728":1,"image\u7684\u4e3b\u8981\u63cf\u8ff0\u6587\u4ef6":1,"image\u7684\u4e3b\u8981\u6784\u5efa\u6b65\u9aa4":1,"image\u7684\u6784\u5efa\u6b65\u9aa4":1,"import":[15,29,33],"include\u4e0b\u9700\u8981\u5305\u542bcbla":2,"include\u4e0b\u9700\u8981\u5305\u542bmkl":2,"init_hook\u53ef\u4ee5\u4f20\u5165\u4e00\u4e2a\u51fd\u6570":29,"int":[15,29],"layer\u6587\u6863":15,"ld_library_path\u7b49\u7b49":9,"ld_library_path\u91cc\u9762\u627e\u4e0d\u5230\u8fd9\u4e9b\u52a8\u6001":9,"lib\u4e0b\u9700\u8981\u5305\u542bcblas\u548catlas\u4e24\u4e2a\u5e93":2,"lib\u4e0b\u9700\u8981\u5305\u542bcblas\u5e93":2,"lib\u4e0b\u9700\u8981\u5305\u542bopenblas\u5e93":2,"lib\u76ee\u5f55\u4e0b\u9700\u8981\u5305\u542b":2,"list\u4e0d\u8bbe\u7f6e":28,"list\u4e2d":28,"list\u4e2d\u7684\u4e00\u884c":29,"list\u4e3a\u7eaf\u6587\u672c\u6587\u4ef6":28,"list\u4e5f\u53ef\u4ee5\u653e\u7f6ehdfs\u6587\u4ef6\u8def\u5f84":28,"list\u5373\u4e3a":29,"list\u548c":28,"list\u548ctest":28,"list\u5747\u4e3a\u672c\u5730\u7684\u4e24\u4e2a\u6587\u4ef6":28,"list\u6307\u5b9a\u7684\u6570\u636e":15,"list\u7684\u4f4d\u7f6e":15,"list\u82e5\u5e72\u6570\u636e\u6587\u4ef6\u8def\u5f84\u7684\u67d0\u4e00\u4e2a\u8def\u5f84":29,"lstm\u6a21\u578b\u7b49":15,"lstm\u6a21\u578b\u914d\u7f6e":15,"make\u548cmak":3,"mkl\u7684\u8def\u5f84":2,"mkl_sequential\u548cmkl_intel_lp64\u4e09\u4e2a\u5e93":2,"mnist\u662f\u4e00\u4e2a\u5305\u542b\u6709":29,"movielens\u6570\u636e\u96c6":14,"movielens\u8bc4\u5206\u56de\u5f52":14,"osx\u6216\u8005\u662fwindows\u673a\u5668":7,"osx\u7684\u5b89\u88c5\u6587\u6863":7,"paddle\u5b9a\u4e49\u7684\u53c2\u6570\u5305\u62ec":29,"paddlepaddle\u4f7f\u7528\u8fd0\u884c\u65f6\u52a8\u6001\u8fde\u63a5cuda\u7684so":9,"paddlepaddle\u4fdd\u7559\u6dfb\u52a0\u53c2\u6570\u7684\u6743\u529b":29,"paddlepaddle\u53ef\u4ee5\u4f7f\u7528":2,"paddlepaddle\u53ef\u4ee5\u8bfb\u53d6python\u5199\u7684\u4f20\u8f93\u6570\u636e\u811a\u672c":15,"paddlepaddle\u5728\u8fd0\u884c\u65f6\u627e\u4e0d\u5230\u5bf9\u5e94\u7684config\u6587\u4ef6":9,"paddlepaddle\u63a8\u8350\u4f7f\u7528docker\u8fdb\u884cpaddlepaddle\u7684\u90e8\u7f72\u548c":7,"paddlepaddle\u63d0\u4f9b\u4e86docker\u7684\u4f7f\u7528\u955c\u50cf":7,"paddlepaddle\u63d0\u4f9b\u6570\u4e2a\u9884\u7f16\u8bd1\u7684\u4e8c\u8fdb\u5236\u6765\u8fdb\u884c\u5b89\u88c5":8,"paddlepaddle\u63d0\u4f9b\u7684\u955c\u50cf\u5e76\u4e0d\u5305\u542b\u4efb\u4f55\u547d\u4ee4\u8fd0\u884c":7,"paddlepaddle\u7684\u6570\u636e\u5305\u62ec\u56db\u79cd\u4e3b\u8981\u7c7b\u578b":29,"paddlepaddle\u7684\u7f16\u8bd1\u9009\u9879\u53ef\u4ee5\u5728\u8c03\u7528cmake\u7684\u65f6\u5019\u8bbe\u7f6e":2,"paddlepaddle\u7684\u7f16\u8bd1\u9009\u9879\u662f\u53ef\u4ee5\u63a7\u5236paddlepaddle\u751f\u6210cpu":2,"paddlepaddle\u7684cpu\u4e8c\u8fdb\u5236":7,"paddlepaddle\u7684dock":1,"paddlepaddle\u7684gpu\u4e8c\u8fdb\u5236":7,"paddlepaddle\u7684python\u9884\u6d4b\u63a5\u53e3":31,"paddlepaddle\u7684ubuntu\u5b89\u88c5\u5305\u5206\u4e3a\u4e24\u4e2a\u7248\u672c":9,"paddlepaddle\u76ee\u524d\u4f7f\u7528swig\u5bf9\u5176\u5e38\u7528\u7684\u9884\u6d4b\u63a5\u53e3\u8fdb\u884c\u4e86\u5c01\u88c5":33,"paddlepaddle\u76ee\u524d\u652f\u6301ubuntu":9,"paddlepaddle\u8fd0\u884c\u65f6\u5982\u679c\u6ca1\u6709\u5bfb\u627e\u5230cuda\u7684driv":9,"paddlepaddle\u9700\u8981\u7528\u6237\u5728\u7f51\u7edc\u914d\u7f6e":28,"period\u8bbe\u7f6e\u6253\u5370\u53c2\u6570\u4fe1\u606f\u7b49":15,"process\u51fd\u6570":29,"process\u51fd\u6570\u662f\u5b9e\u73b0\u6570\u636e\u8f93\u5165\u7684\u4e3b\u51fd\u6570":29,"process\u51fd\u6570\u8c03\u7528\u591a\u6b21":29,"pserver\u4e3apaddlepaddle\u7684paramet":22,"pserver\u7684\u547d\u4ee4\u884c\u53c2\u6570":22,"pserver\u7ec4\u5408\u4f7f\u7528":22,"py\u6587\u4ef6\u7ed9\u51fa\u4e86\u5b8c\u6574\u4f8b\u5b50":15,"pydataprovider2\u7684\u4f7f\u7528":28,"pydataprovider\u662fpaddlepaddle\u4f7f\u7528python\u63d0\u4f9b\u6570\u636e\u7684\u63a8\u8350\u63a5\u53e3":29,"python\u5305":7,"python\u7684":7,"python\u7684swig\u63a5\u53e3\u53ef\u4ee5\u65b9\u4fbf\u8fdb\u884c\u9884\u6d4b\u548c\u5b9a\u5236\u5316\u8bad\u7ec3":2,"release\u7684\u72b6\u6001":8,"return":15,"rnn\u914d\u7f6e":[0,20],"server\u8fdb\u7a0b":22,"sh\u662fdocker":1,"shuffle\u8bad\u7ec3\u6570\u636e":29,"so\u627e\u4e0d\u5230":8,"softmax\u8f93\u51fa":15,"string\u7684\u683c\u5f0f\u6253\u5370\u51fa\u6765":22,"swig_paddle\u63a5\u53d7\u7684\u539f\u59cb\u6570\u636e\u662fc":33,"train\u5373\u4e3apaddlepaddle\u7684\u8bad\u7ec3\u8fdb\u7a0b":22,"train\u5b8c\u6210\u5355\u673a\u591a\u663e\u5361\u591a\u7ebf\u7a0b\u7684\u8bad":22,"train\u7684\u547d\u4ee4\u884c\u53c2\u6570":22,"ubuntu\u7684deb\u5b89\u88c5\u5305\u7b49":8,"v2\u4e4b\u540e\u7684\u4efb\u4f55\u4e00\u4e2acudnn\u7248\u672c\u6765\u7f16\u8bd1\u8fd0\u884c":2,"value\u5373\u4e3asoftmax\u5c42\u7684\u8f93\u51fa":33,"var":7,"vector\u8868\u793a\u7684\u6bcf\u4e2a\u5355\u8bcd":15,"version\u53ef\u4ee5\u6253\u5370\u51fapaddle\u7684\u7248\u672c\u4fe1\u606f\u548c\u7f16\u8bd1\u7684\u9009\u9879":27,"version\u53ef\u4ee5\u6253\u5370\u51fapaddlepaddle\u7684\u7248\u672c\u548c\u7f16\u8bd1\u65f6\u4fe1\u606f":22,"version\u7684\u547d\u4ee4\u884c\u53c2\u6570":22,"yield\u6587\u672c\u4fe1\u606f\u548c\u7c7b\u522bid":15,abov:29,act:15,act_typ:15,activ:15,adadelta:15,adagrad:15,adam:15,adamoptim:15,afi:29,agre:33,all:[29,33],allow:15,alreadi:9,also:15,ani:33,apach:33,append:29,applic:33,apt:[7,9],arg:[1,15,29],around:29,arrai:33,assert:33,atla:2,atlas_root:2,avg:15,avgcost:15,avgpool:15,avx:9,bag:15,baidu:[7,9,33],basi:33,batch:15,batch_siz:15,binari:15,bla:2,bool:15,both:15,bow:15,build:[1,7],cach:[15,28],cache_pass_in_mem:[15,29],cachetyp:[15,29],calc_batch_s:29,call:15,can:15,can_over_batch_s:29,categori:15,cento:9,check:9,chines:14,chpasswd:7,classif:15,classification_cost:15,classification_error_evalu:15,close:29,cmake:2,cmd:7,cnn:15,code:[1,29,33],com:[7,9],comment:15,compil:27,complianc:33,condit:33,conf:33,config:[9,15,32],config_arg:15,config_pars:33,connect:15,contain:[15,29],context_len:15,context_start:15,convert:[15,29,33],copi:33,copyright:33,couldn:9,cpp:[9,15],cpu:[7,9,29],createfromconfigproto:33,cross:15,cuda_so:7,cudastat:9,cudasuccess:9,cudnn:2,cudnn_root:2,cudnnv5:2,current:[15,29],currentcost:15,currentev:15,dalla:29,data:9,data_config:33,data_initialz:15,data_lay:15,dataprovid:15,dataprovider_bow:15,dataprovider_emb:15,dataproviderwrapperconvert:33,dataset:15,deb:9,debian:9,decor:29,def:[15,29,33],defin:[15,29],define_py_data_sources2:[15,29],delar:15,demo:7,dense_vector:29,denseslot:33,describ:15,descript:31,detail:31,dev:7,devel:7,devic:7,dict:[15,29],dict_fil:15,dictionai:15,dictionari:[15,29],dictrionari:15,differ:15,dim:15,dimens:15,dir:15,distribut:33,documentari:29,dpkg:9,driver:9,dso_handl:9,dtype:33,dump_config:22,dure:[15,29],dynam:29,each:[15,29],each_pixel_str:29,each_word:29,echo:7,either:[15,33],els:15,emb:15,embedding_lay:15,entropi:15,enumer:15,error:[9,15],etc:7,eval:15,exampl:15,except:33,experi:8,expos:7,express:33,f0831:9,fail:9,fals:[15,33],fc_layer:15,featur:[15,29],festiv:29,file:[15,29,33],file_list:29,file_nam:15,filenam:29,fill:15,find:9,first:15,float32:33,fly:15,forwardtest:33,framework:15,from:[7,15,29,33],fulli:15,gener:[15,29],gentoo:9,get:[7,9,15,29],get_config_arg:15,get_data:15,github:9,give:29,given:15,globe:29,govern:33,gpu:[7,9],gradient_clipping_threshold:15,gradientmachin:33,gru:15,gru_siz:15,help:33,hint:33,hl_cuda_devic:9,hl_dso_load:9,host:7,hot:15,hous:29,http:[9,33],ignor:29,imagenet:14,impli:33,inarg:33,inc:33,includ:15,init:15,init_hook:[15,28],init_model_path:15,initi:[15,29],initpaddl:33,input:[15,29],input_typ:[15,28],instal:3,insuffici:9,intal:8,integ:[15,29],integer_sequ:29,integer_valu:[15,29],integer_value_sequ:15,interfac:32,invok:29,is_predict:15,is_train:29,isinst:33,job:15,kernel:7,kind:33,kwarg:[15,29],l2regular:15,label:[15,29],label_dim:15,lake:29,languag:33,later:15,latest:[1,7],law:33,layer:15,ld_library_path:9,learning_method:15,learning_r:15,len:[15,29],lib64:[7,9],lib:2,libcuda:7,libcudart:8,libcudnn:8,libnvidia:7,librari:9,licens:33,limit:33,list:[15,28,29],load_data_arg:33,loadparamet:33,local:[2,9],log_period:15,logger:29,look:[15,29],loss:15,lowest_dl_spe:1,lstm:15,lstm_size:15,mac:7,mai:[8,33],main:33,maintain:7,make:[9,29],make_diagram:22,maxid:15,maxid_lay:15,mean:15,memori:15,merge_model:22,method:29,mkdir:7,mkl:2,mkl_core:2,mkl_root:2,mnist:29,mnist_model:33,mnist_provid:29,mnist_train:29,model_config:33,modul:[15,29],momentum:15,movi:29,must:9,name:[7,15],necessari:15,need:15,neg:[15,29],net:7,no_cach:29,no_sequ:29,none:[15,29,33],norm:33,note:[8,9],nullptr:9,num:15,num_pass:15,nvidia:7,obj:[15,29],object:[15,29],obtain:33,off:[1,2,27],omit:15,on_init:29,onli:15,open:[15,29],openbla:2,openblas_root:2,openssh:7,opt:2,org:33,other:15,outlin:31,output:15,outsid:29,packag:8,paddl:[1,7,9,15,22],paddle_gpu:1,paddle_ssh:7,paddle_ssh_machin:7,paddledev:7,paddlepaddl:[7,9,27],paramet:15,parse_config:33,pass:[15,29],path:[9,15],period:15,permiss:33,permitrootlogin:7,pixel:29,pixels_float:29,pixels_str:29,place:29,pleas:9,pool_siz:29,pooling_lay:15,pooling_typ:15,posit:[15,29],pre:8,pred:15,predict_output_dir:15,preprocess:15,print:33,process:[15,29],process_pr:15,properli:15,pull:7,put:15,py_paddl:[7,33],pydataprovid:28,pydataprovider2:[15,29],pydataproviderwrapp:[15,33],python:15,quick_start:15,rang:15,rank:15,rare:29,read:[15,29],refer:28,reference_cblas_root:2,reffer:2,regular:15,releas:[8,9],repres:15,represent:15,requir:33,reserv:33,result:15,right:33,rmsprop:15,roce:7,root:7,run:7,runtim:[9,29],same:[15,29],sampl:[15,29],save:[15,29],save_dir:15,saw:29,sbin:7,script:1,second:15,sed:7,see:[15,33],sentenc:29,sentiment:29,sentimental_provid:29,separ:15,seq_typ:29,sequel:29,sequence_conv_pool:15,sequencetyp:29,server:7,set:[15,29],setup:15,should_shuffl:29,simple_gru:15,simple_lstm:15,size:[15,29],smooth:8,softmax:15,softmaxactiv:15,softwar:33,sourc:15,spars:15,sparse_binary_vector:[15,29],sparse_float_vector:29,specif:33,specifi:[9,15],split:[15,29],ssh:7,sshd:7,sshd_config:7,stat:15,state:8,still:8,stop:7,store:15,string:29,strip:15,structur:15,stun:29,sub_sequ:29,sure:9,swig_paddl:33,tag:[1,9],take:29,tbd:[12,13,16,17,18,30],team:7,test:[15,28],test_data:33,test_list:[15,29],text:[15,29],text_conv:15,them:15,thi:[15,29,33],thing:29,train:9,train_list:[15,29],trainer:[15,29,33],trainer_config:[15,28,29,33],trainer_config_help:[15,29],trainerintern:15,trainermain:9,travel:29,trn:15,tst:15,two:15,txt:[15,29],type:[15,29],ubuntu:9,under:33,unk_idx:15,unless:33,updat:7,use_gpu:[15,33],usepam:7,user:15,usr:[2,7,9],valid:9,valu:[15,29,33],version:[7,9,33],via:9,want:29,warranti:33,what:15,when:29,which:15,whole:29,wilder:29,window:7,with_avx:[2,27],with_doc:2,with_doc_cn:2,with_doubl:[2,27],with_dso:2,with_gflag:[2,27],with_glog:[2,27],with_gpu:[1,2,27],with_metric_learn:27,with_predict_sdk:27,with_python:[2,27],with_rdma:[2,27],with_style_check:2,with_swig_pi:2,with_test:2,with_tim:[2,27],without:33,wonder:29,word:14,word_dict:15,word_dim:15,word_id:29,word_vector:15,write:33,www:33,xarg:7,yield:[15,29],you:[9,29,33],your:8,your_host_machin:7},titles:["RNN \u914d\u7f6e","\u6784\u5efaPaddlePaddle Docker Image","\u8bbe\u7f6ePaddlePaddle\u7684\u7f16\u8bd1\u9009\u9879","\u4f7f\u7528cmake\u7f16\u8bd1PaddlePaddle","\u5b89\u88c5\u7f16\u8bd1PaddlePaddle\u9700\u8981\u7684\u4f9d\u8d56","make\u548cmake install","\u7f16\u8bd1\u4e0e\u5b89\u88c5","\u5b89\u88c5PaddlePaddle\u7684Docker\u955c\u50cf","\u5b89\u88c5PaddlePaddle","\u4f7f\u7528deb\u5305\u5728Ubuntu\u4e0a\u5b89\u88c5PaddlePaddle","\u96c6\u7fa4\u8bad\u7ec3","Embedding Demo","\u56fe\u7247\u5206\u7c7b\u6559\u7a0b","Resnet","\u4f7f\u7528\u793a\u4f8b","PaddlePaddle\u5feb\u901f\u5165\u95e8\u6559\u7a0b","\u8bed\u4e49\u6807\u6ce8","\u60c5\u611f\u5206\u6790","\u6587\u672c\u751f\u6210","\u65b0\u5199Layer","PaddlePaddle\u6587\u6863","<no title>","\u547d\u4ee4\u884c\u53c2\u6570","<no title>","<no title>","paddle pserver\u7684\u547d\u4ee4\u884c\u53c2\u6570","paddle train\u7684\u547d\u4ee4\u884c\u53c2\u6570","paddle version\u7684\u547d\u4ee4\u884c\u53c2\u6570","PaddlePaddle\u7684\u6570\u636e\u63d0\u4f9b(DataProvider)\u4ecb\u7ecd","PyDataProvider2\u7684\u4f7f\u7528","\u81ea\u5b9a\u4e49\u4e00\u4e2aDataProvider","\u7528\u6237\u63a5\u53e3","\u6a21\u578b\u914d\u7f6e","PaddlePaddle\u7684Python\u9884\u6d4b\u63a5\u53e3"],titleterms:{"\u4e0b\u8f7d\u548c\u8fd0\u884cdocker\u955c\u50cf":7,"\u4ecb\u7ecd":28,"\u4f18\u5316\u7b97\u6cd5":15,"\u4f7f\u7528\u6307\u5357":20,"\u4f7f\u7528\u6982\u8ff0":15,"\u4f7f\u7528\u793a\u4f8b":14,"\u4f7f\u7528\u811a\u672c\u6784\u5efapaddlepaddl":1,"\u4f7f\u7528cmake\u7f16\u8bd1paddlepaddl":3,"\u4f7f\u7528deb\u5305\u5728ubuntu\u4e0a\u5b89\u88c5paddlepaddl":9,"\u5377\u79ef\u6a21\u578b":15,"\u53c2\u8003":29,"\u53ef\u80fd\u9047\u5230\u7684\u95ee\u9898":9,"\u547d\u4ee4\u884c\u53c2\u6570":[15,22,31],"\u56fe\u50cf":14,"\u56fe\u7247\u5206\u7c7b\u6559\u7a0b":12,"\u5b89\u88c5":15,"\u5b89\u88c5\u7f16\u8bd1paddlepaddle\u9700\u8981\u7684\u4f9d\u8d56":4,"\u5b89\u88c5paddlepaddl":8,"\u5b89\u88c5paddlepaddle\u7684docker\u955c\u50cf":7,"\u5e38\u7528\u6a21\u578b":14,"\u5e8f\u5217\u6a21\u578b\u6570\u636e\u63d0\u4f9b":29,"\u5f00\u53d1\u6307\u5357":20,"\u6027\u80fd\u95ee\u9898":7,"\u603b\u4f53\u6548\u679c\u603b\u7ed3":15,"\u60c5\u611f\u5206\u6790":17,"\u63a8\u8350":14,"\u6570\u636e\u5411\u6a21\u578b\u4f20\u9001":15,"\u6570\u636e\u63d0\u4f9b":31,"\u6570\u636e\u683c\u5f0f\u51c6\u5907":15,"\u6587\u672c\u751f\u6210":18,"\u65b0\u5199layer":19,"\u65f6\u5e8f\u6a21\u578b":15,"\u6784\u5efapaddlepaddl":1,"\u6a21\u578b\u914d\u7f6e":32,"\u6ce8\u610f\u4e8b\u9879":7,"\u7528\u6237\u63a5\u53e3":31,"\u7b80\u5355\u7684\u4f7f\u7528\u573a\u666f":29,"\u7b97\u6cd5\u6559\u7a0b":20,"\u7f16\u8bd1\u4e0e\u5b89\u88c5":6,"\u7f51\u7edc\u7ed3\u6784":15,"\u81ea\u5b9a\u4e49\u4e00\u4e2adataprovid":30,"\u81ea\u7136\u8bed\u8a00\u5904\u7406":14,"\u8bad\u7ec3\u6a21\u578b":15,"\u8bbe\u7f6epaddlepaddle\u7684\u7f16\u8bd1\u9009\u9879":2,"\u8bcd\u5411\u91cf\u6a21\u578b":15,"\u8bed\u4e49\u6807\u6ce8":16,"\u8f93\u51fa\u65e5\u5fd7":15,"\u8fdc\u7a0b\u8bbf\u95ee\u95ee\u9898\u548c\u4e8c\u6b21\u5f00\u53d1":7,"\u903b\u8f91\u56de\u5f52\u6a21\u578b":15,"\u914d\u7f6e":0,"\u914d\u7f6e\u4e2d\u7684\u6570\u636e\u52a0\u8f7d\u5b9a\u4e49":15,"\u9644\u5f55":15,"\u96c6\u7fa4\u8bad\u7ec3":10,"\u9884\u6d4b":[15,31],"blas\u76f8\u5173\u7684\u7f16\u8bd1\u9009\u9879":2,"bool\u578b\u7684\u7f16\u8bd1\u9009\u9879":2,"config\u6587\u4ef6\u627e\u4e0d\u5230":9,"cudnn\u76f8\u5173\u7684\u7f16\u8bd1\u9009\u9879":2,"driver\u627e\u4e0d\u5230":9,"make\u548cmak":5,"paddlepaddle\u5feb\u901f\u5165\u95e8\u6559\u7a0b":15,"paddlepaddle\u63d0\u4f9b\u7684docker\u955c\u50cf\u7248\u672c":7,"paddlepaddle\u6587\u6863":20,"paddlepaddle\u7684\u6570\u636e\u63d0\u4f9b":28,"paddlepaddle\u7684\u7f16\u8bd1\u9009\u9879":2,"paddlepaddle\u7684bool\u578b\u7f16\u8bd1\u9009\u9879":2,"paddlepaddle\u7684cblas\u7f16\u8bd1\u9009\u9879":2,"paddlepaddle\u7684python\u9884\u6d4b\u63a5\u53e3":33,"pserver\u7684\u547d\u4ee4\u884c\u53c2\u6570":25,"pydataprovider2\u7684\u4f7f\u7528":29,"python\u6570\u636e\u52a0\u8f7d\u811a\u672c":15,"so\u627e\u4e0d\u5230":9,"train\u7684\u547d\u4ee4\u884c\u53c2\u6570":26,"version\u7684\u547d\u4ee4\u884c\u53c2\u6570":27,algorithm:15,appendix:15,architectur:15,argument:15,cach:29,command:15,configur:15,convolut:15,cuda:[2,9],data:15,dataprovid:28,demo:11,docker:1,embed:11,imag:1,init_hook:29,input_typ:29,instal:[5,15],libcudart:9,libcudnn:9,line:15,log:15,logist:15,model:15,network:15,optim:15,overview:15,paddl:[25,26,27],predict:15,prepar:15,provid:[15,29],refer:29,regress:15,resnet:13,rnn:0,script:15,sequenc:15,summari:15,time:15,train:15,transfer:15,vector:15,word:15}})
\ No newline at end of file
+Search.setIndex({envversion:49,filenames:["build/docker/build_docker_image","build_and_install/cmake/compile_options","build_and_install/cmake/index","build_and_install/cmake/install_deps","build_and_install/cmake/make_and_install","build_and_install/index","build_and_install/install/docker_install","build_and_install/install/ubuntu_install","cluster/index","demo/index","demo/quick_start/index","index","ui/cmd/dump_config","ui/cmd/index","ui/cmd/make_diagram","ui/cmd/merge_model","ui/cmd/paddle_pserver","ui/cmd/paddle_train","ui/cmd/paddle_version","ui/data_provider/index","ui/data_provider/pydataprovider2","ui/data_provider/write_new_dataprovider","ui/index","ui/predict/swig_py_paddle"],objects:{},objnames:{},objtypes:{},terms:{"0000x":10,"000\u5f20\u7070\u5ea6\u56fe\u7247\u7684\u6570\u5b57\u5206\u7c7b\u6570\u636e\u96c6":20,"00186201e":23,"04\u7248\u672c\u4f7f\u7528deb\u5305\u5b89\u88c5":7,"08823112e":23,"0b0":7,"10\u4ee5\u4e0a\u7684linux":6,"10\u7ef4\u7684\u6574\u6570\u503c":20,"10gbe":6,"10m":0,"12194102e":23,"12\u7248\u672c\u6d4b\u8bd5\u901a\u8fc7":0,"12\u7248\u672c\u7684\u60c5\u51b5\u4e0b\u5e76\u6ca1\u6709\u6d4b\u8bd5":0,"15501715e":23,"15mb":10,"16mb":10,"1\u7684\u8bdd":20,"252kb":10,"25639710e":23,"25k":10,"27787406e":23,"28\u7684\u50cf\u7d20\u7070\u5ea6\u503c":20,"28\u7684\u7a20\u5bc6\u5411\u91cf\u548c\u4e00\u4e2a":20,"2\u8fdb\u884c\u8fdb\u4e00\u6b65\u6f14\u5316":10,"32777140e":23,"36540484e":23,"40gbe":6,"43630644e":23,"48565123e":23,"48684503e":23,"49316648e":23,"50k":0,"51111044e":23,"53018653e":23,"56gbe":6,"5\u5230\u672c\u5730\u73af\u5883\u4e2d":7,"70634608e":23,"72296313e":23,"85625684e":23,"93137714e":23,"96644767e":23,"99982715e":23,"9\u7684\u6570\u5b57":20,"\u4e00\u4e2a\u6587\u4ef6":20,"\u4e00\u4e2alogging\u5bf9\u8c61":20,"\u4e00\u4e2apass\u8868\u793a\u8fc7\u4e00\u904d\u6240\u6709\u8bad\u7ec3\u6837\u672c":10,"\u4e00\u6761":20,"\u4e00\u81f4\u7684\u7279\u5f81":20,"\u4e00\u822c\u60c5\u51b5\u4e0b":19,"\u4e00\u822c\u63a8\u8350\u8bbe\u7f6e\u6210true":20,"\u4e00\u884c\u4e3a\u4e00\u4e2a\u6837\u672c":10,"\u4e00\u884c\u5bf9\u5e94\u4e00\u4e2a\u6570\u636e\u6587\u4ef6":19,"\u4e0a\u7684\u6548\u679c":10,"\u4e0b\u8f7d\u8fdb\u7a0b\u4f1a\u91cd\u542f":0,"\u4e0b\u8ff0\u5185\u5bb9\u5c06\u5206\u4e3a\u5982\u4e0b\u51e0\u4e2a\u7c7b\u522b\u63cf\u8ff0":6,"\u4e0b\u9762dataprovid":10,"\u4e0d\u4e00\u5b9a\u548c\u65f6\u95f4\u6709\u5173\u7cfb":20,"\u4e0d\u4f1a\u6267\u884c\u6d4b\u8bd5\u64cd\u4f5c":19,"\u4e0d\u5305\u542blabel":23,"\u4e0d\u540c\u7684\u6570\u636e\u7c7b\u578b\u548c\u5e8f\u5217\u6a21\u5f0f\u8fd4\u56de\u7684\u683c\u5f0f\u4e0d\u540c":20,"\u4e0d\u540c\u7684\u7a33\u5b9a\u7248\u672c":6,"\u4e0d\u7f13\u5b58\u4efb\u4f55\u6570\u636e":20,"\u4e0e\u8bad\u7ec3\u7f51\u7edc\u914d\u7f6e\u4e0d\u540c\u7684\u662f":10,"\u4e14\u5e8f\u5217\u7684\u6bcf\u4e00\u4e2a\u5143\u7d20\u8fd8\u662f\u4e00\u4e2a\u65f6\u95f4\u5e8f\u5217":20,"\u4e24\u4e2a\u6587\u6863":6,"\u4e24\u7c7b":10,"\u4e2a":10,"\u4e2d":10,"\u4e2d\u5b9a\u4e49\u4f7f\u7528\u54ea\u79cddataprovider\u53ca\u5176\u53c2\u6570":19,"\u4e2d\u5b9a\u4e49\u7684\u987a\u5e8f\u4e00\u81f4":20,"\u4e2d\u5bfb\u627e\u8fd9\u4e9bblas\u7684\u5b9e\u73b0":1,"\u4e2d\u7684":20,"\u4e2d\u7684\u4e8c\u8fdb\u5236\u4f7f\u7528\u4e86":6,"\u4e2d\u7684set":20,"\u4e2d\u914d\u7f6e":20,"\u4e3a":20,"\u4e3a\u4e86\u8fd0\u884cpaddlepaddle\u7684docker\u955c\u50cf":6,"\u4e3a\u4f8b\u8fdb\u884c\u9884\u6d4b":10,"\u4e3b\u8981\u51fd\u6570\u662fprocess\u51fd\u6570":20,"\u4e3b\u8981\u5206\u4e3a\u4ee5\u4e0b\u51e0\u4e2a\u6b65\u9aa4":23,"\u4e3b\u8981\u5305\u62ec\u4e24\u90e8\u5206":10,"\u4e3b\u8981\u662f\u589e\u52a0\u4e86\u521d\u59cb\u5316\u673a\u5236":20,"\u4e3b\u8981\u6b65\u9aa4\u4e3a":23,"\u4e3b\u8981\u7531\u4e8e\u65e7\u7248\u672c":0,"\u4e3b\u8981\u7684\u8f6f\u4ef6\u5305\u4e3apy_paddl":23,"\u4e5f\u4f1a\u8bfb\u53d6\u76f8\u5173\u8def\u5f84\u53d8\u91cf\u6765\u8fdb\u884c\u641c\u7d22":1,"\u4e5f\u53ef\u4ee5\u4f7f\u7528":20,"\u4e5f\u53ef\u4ee5\u548cpaddl":13,"\u4e5f\u53ef\u4ee5\u76f4\u63a5\u6267\u884c":6,"\u4e5f\u53ef\u4ee5\u901a\u8fc7\u5982\u4e0b\u65b9\u5f0f\u9884\u6d4b":10,"\u4e5f\u53ef\u4ee5\u901a\u8fc7save":10,"\u4e5f\u53ef\u4ee5\u9884\u6d4b\u6ca1\u6709label\u7684\u6d4b\u8bd5\u96c6":10,"\u4e5f\u5c31\u662f\u5c06\u8bcd\u5411\u91cf\u6a21\u578b\u989d\u6b65":10,"\u4e5f\u5c31\u662f\u76f4\u63a5\u5199\u5185\u5b58\u7684float\u6570\u7ec4":23,"\u4e86\u975e\u5e38\u65b9\u4fbf\u7684\u4e8c\u8fdb\u5236\u5206\u53d1\u624b\u6bb5":6,"\u4e8c\u6b21\u5f00\u53d1\u53ef\u4ee5":6,"\u4eba\u5458\u7b49\u7b49":0,"\u4ec5\u4ec5\u9700\u8981":20,"\u4ecb\u7ecdpaddlepaddle\u4f7f\u7528\u6d41\u7a0b\u548c\u5e38\u7528\u7684\u7f51\u7edc\u57fa\u7840\u5355\u5143\u7684\u914d\u7f6e\u65b9\u6cd5":10,"\u4ece\u6587\u4ef6\u4e2d\u8bfb\u53d6\u6bcf\u4e00\u6761\u6570\u636e":20,"\u4ece\u6587\u672c\u6587\u4ef6\u4e2d\u8bfb\u53d6":20,"\u4ece\u800c\u4e0d\u80fd\u5728\u8fd0\u884c\u7f16\u8bd1\u547d\u4ee4\u7684\u65f6\u5019\u63a5\u53d7\u53c2\u6570":0,"\u4ece\u800c\u9632\u6b62\u8fc7\u62df\u5408":19,"\u4ed6\u4eec\u662f":[13,20],"\u4ed6\u4eec\u7684\u4e0b\u8f7d\u5730\u5740\u662f":7,"\u4ed6\u53ef\u4ee5\u5c06\u67d0\u4e00\u4e2a\u51fd\u6570\u6807\u8bb0\u6210\u4e00\u4e2apydataprovid":20,"\u4ee4\u884c\u53c2\u6570\u6587\u6863":10,"\u4ee5\u592a\u7f51\u5361":6,"\u4ee5\u76f8\u5bf9\u8def\u5f84\u5f15\u7528":19,"\u4efb\u610f\u4e00\u79cdcblas\u5b9e\u73b0":1,"\u4f1a\u5c06latest\u66ff\u6362\u6210\u7a33\u5b9a\u7248\u672c\u7684\u7248\u672c\u53f7":6,"\u4f1a\u62a5\u5bfb\u627e\u4e0d\u5230\u8fd9\u4e9b\u52a8\u6001\u5e93":7,"\u4f1a\u6839\u636e\u547d\u4ee4\u884c\u53c2\u6570\u6307\u5b9a\u7684\u6d4b\u8bd5\u65b9\u5f0f":19,"\u4f20\u5165\u547d\u4ee4\u884c\u53c2\u6570\u521d\u59cb\u5316":23,"\u4f20\u5165\u7684\u662f\u4e00\u4e2a\u51fd\u6570":20,"\u4f20\u5165\u7684\u914d\u7f6e\u53c2\u6570\u5305\u62ec":0,"\u4f20\u5165\u8fd9\u4e2a\u53d8\u91cf\u7684\u65b9\u5f0f\u4e3a":20,"\u4f46\u662f":0,"\u4f46\u662f\u5728":20,"\u4f46\u662f\u5982\u679c\u4f7f\u7528\u4e86\u9ad8\u6027\u80fd\u7684\u7f51\u5361":6,"\u4f46\u662f\u65b9\u4fbf\u8c03\u8bd5\u548cbenchmark":1,"\u4f46\u662f\u6709\u65f6\u4e3a\u4e86\u8ba1\u7b97\u5747\u8861\u6027":20,"\u4f46\u9700\u8981\u6ce8\u610f\u7684\u662f\u7f16\u8bd1\u548c":1,"\u4f4e\u4e8edocker":0,"\u4f53\u53ef\u4ee5\u53c2\u8003":20,"\u4f7f\u5728python\u73af\u5883\u4e0b\u7684\u9884\u6d4b\u63a5\u53e3\u66f4\u52a0\u7b80\u5355":23,"\u4f7f\u7528":[1,6,23],"\u4f7f\u7528\u7684\u547d\u4ee4\u4e5f\u662f":1,"\u4f7f\u7528\u8be5\u63a5\u53e3\u7528\u6237\u53ef\u4ee5\u53ea\u5173\u6ce8\u5982\u4f55":20,"\u4f7f\u7528\u8be5dockerfile\u6784\u5efa\u51fa\u955c\u50cf":6,"\u4f7f\u7528\u8d77\u6765\u6216\u8bb8\u4f1a\u4e0d\u662f\u5f88\u987a\u7545":5,"\u4f7f\u7528\u8fd9\u4e2a\u5173\u952e\u8bcd":20,"\u4f7f\u7528deb\u5305\u5728ubuntu\u4e0a\u5b89\u88c5paddlepaddl":5,"\u4f7f\u7528dockerfile\u6784\u5efa\u4e00\u4e2a\u5168\u65b0\u7684dock":6,"\u4f7f\u7528mnist\u624b\u5199\u8bc6\u522b\u4f5c\u4e3a\u6837\u4f8b":23,"\u4f7f\u7528ssh\u8bbf\u95eepaddlepaddle\u955c\u50cf":6,"\u4f8b\u5982":[1,7,10,20],"\u4f8b\u5982\u6587\u4ef6\u540d\u662f":20,"\u4f8b\u5982rdma\u7f51\u5361":6,"\u4f8b\u5982sigmoid\u53d8\u6362":10,"\u4f9d\u6b21\u8fd4\u56de\u4e86\u6587\u4ef6\u4e2d\u7684\u6bcf\u6761\u6570\u636e":20,"\u4fe1\u606f":6,"\u5173\u95edcontain":6,"\u5176\u4e2d":[0,6,19,20,23],"\u5176\u4e2d\u6587\u672c\u8f93\u5165\u7c7b\u578b\u5b9a\u4e49\u4e3a\u6574\u6570\u65f6\u5e8f\u7c7b\u578binteg":10,"\u5176\u4e2d\u7b2c\u4e00\u884c\u662f\u5f15\u5165paddlepaddle\u7684pydataprovider2\u5305":20,"\u5176\u4ed6\u53c2\u6570\u5747\u4f7f\u7528kei":20,"\u5176\u4ed6\u53c2\u6570\u8bf7\u53c2\u8003":10,"\u5176\u4ed6\u53c2\u6570\u90fd\u4f7f\u7528kei":20,"\u5176\u4f5c\u7528\u662f\u5c06\u8bad\u7ec3\u6570\u636e\u4f20\u5165\u5185\u5b58\u6216\u8005\u663e\u5b58":19,"\u5176\u5b83\u90e8\u5206\u548c\u903b\u8f91\u56de\u5f52\u7f51\u7edc\u7ed3\u6784\u4e00\u81f4":10,"\u5176\u6570\u636e\u4f7f\u7528":20,"\u5176\u7b2c\u4e00\u884c\u8bf4\u660e\u4e86paddle\u7684\u7248\u672c":18,"\u5177":20,"\u5177\u4f53\u53ef\u4ee5\u8bbe\u7f6e\u6210\u4ec0\u4e48\u5176\u4ed6\u683c":20,"\u5177\u4f53\u6709\u54ea\u4e9b\u683c\u5f0f":20,"\u5177\u4f53\u8bf7\u53c2\u8003\u6ce8\u610f\u4e8b\u9879\u4e2d\u7684":6,"\u5177\u4f53dataprovider\u8fd8\u5177\u6709\u4ec0\u4e48\u529f\u80fd":20,"\u5177\u6709\u4e24\u4e2a\u53c2\u6570":20,"\u5177\u6709\u548c\u5bbf\u4e3b\u673a\u76f8\u8fd1\u7684\u8fd0\u884c\u6548\u7387":6,"\u5177\u6709\u7684\u5c5e\u6027\u6709":20,"\u5178\u578b\u7684\u8f93\u51fa\u7ed3\u679c\u4e3a":23,"\u5178\u578b\u7684\u9884\u6d4b\u4ee3\u7801\u5982\u4e0b":23,"\u518d\u6307\u5b9a":1,"\u5199\u5165train":20,"\u5199\u5728train":19,"\u51c6\u5907\u6570\u636e":23,"\u51fd\u6570":20,"\u51fd\u6570\u4e2d":20,"\u51fd\u6570\u4e2d\u4f7f\u7528":20,"\u51fd\u6570\u4e2d\u7684":20,"\u51fd\u6570\u662f\u4f7f\u7528":20,"\u51fd\u6570\u6765\u4fdd\u8bc1\u517c\u5bb9\u6027":20,"\u51fd\u6570\u67e5\u8be2\u6587\u6863":23,"\u5206\u5e03\u5f0f\u8bad\u7ec3":10,"\u5206\u7c7b\u6210\u6b63\u9762\u60c5\u7eea\u548c":20,"\u5217\u8868\u5982\u4e0b":20,"\u5219\u5b57\u4e0e\u5b57\u4e4b\u95f4\u7528\u7a7a\u683c\u5206\u9694":10,"\u5219\u9700\u8981\u5148\u5c06":6,"\u5219\u9700\u8981\u8fdb\u884c\u4e00\u5b9a\u7684\u4e8c\u6b21\u5f00\u53d1":6,"\u521b\u5efa\u4e00\u4e2a":23,"\u5220\u9664contain":6,"\u5229\u7528\u5355\u8bcdid\u67e5\u627e\u5bf9\u5e94\u7684\u8be5\u5355\u8bcd\u7684\u8fde\u7eed\u8868\u793a\u5411\u91cf":10,"\u5229\u7528\u903b\u8f91\u56de\u5f52\u6a21\u578b\u5bf9\u8be5\u5411\u91cf\u8fdb\u884c\u5206\u7c7b":10,"\u522b":10,"\u5269\u4e0b\u7684pass\u4f1a\u76f4\u63a5\u4ece\u5185\u5b58\u91cc":20,"\u52a0\u4e86l2\u6b63\u5219\u548c\u68af\u5ea6\u622a\u65ad":10,"\u52a0\u8f7d\u6570\u636e":10,"\u5305":6,"\u5305\u548c":6,"\u5305\u62ec":10,"\u5305\u62ec\u7b80\u5355\u7684rnn\u6a21\u578b":10,"\u5305\u62ecdocker\u955c\u50cf":5,"\u5373":[6,10],"\u5373\u4e0d\u5728\u4e4e\u5185\u5b58\u6682\u5b58\u591a\u5c11\u6761\u6570\u636e":20,"\u5373\u4e0d\u662f\u4e00\u6761\u5e8f\u5217":20,"\u5373\u4ece\u5355\u8bcd\u5b57\u7b26\u4e32\u5230\u5355\u8bcdid\u7684\u5b57\u5178":20,"\u5373\u5305\u542b\u65f6\u95f4\u6b65\u4fe1\u606f":20,"\u5373\u53ef":20,"\u5373\u53ef\u4ee5\u4f7f\u7528ssh\u8bbf\u95ee\u5bbf\u4e3b\u673a\u76848022\u7aef\u53e3":6,"\u5373\u53ef\u542f\u52a8\u548c\u8fdb\u5165paddlepaddle\u7684contain":6,"\u5373\u53ef\u5728\u672c\u5730\u7f16\u8bd1\u51fapaddlepaddle\u7684\u955c\u50cf":0,"\u5373\u53ef\u6253\u5370\u51fapaddlepaddle\u7684\u7248\u672c\u548c\u6784\u5efa":6,"\u5373\u5927\u90e8\u5206\u503c\u4e3a0":20,"\u5373\u5bf9\u7b2c3\u6b65\u8fdb\u884c\u66ff\u6362":10,"\u5373\u662f\u4e00\u6761\u65f6\u95f4\u5e8f\u5217":20,"\u5373cpu\u7248\u672c":7,"\u5373train":20,"\u5377\u79ef\u7f51\u7edc\u662f\u4e00\u79cd\u7279\u6b8a\u7684\u4ece\u8bcd\u5411\u91cf\u8868\u793a\u5230\u53e5\u5b50\u8868\u793a\u7684\u65b9\u6cd5":10,"\u53c2\u6570":0,"\u53c2\u6570\u6570\u91cf":10,"\u53c2\u8003":19,"\u53c2\u89c1":[3,4],"\u53d6\u51b3\u4e8e\u662f\u5426\u5bfb\u627e\u5230gflags":1,"\u53d6\u51b3\u4e8e\u662f\u5426\u5bfb\u627e\u5230glog":1,"\u53d6\u51b3\u4e8e\u662f\u5426\u5bfb\u627e\u5230gtest":1,"\u53d6\u51b3\u4e8e\u662f\u5426\u627e\u5230swig":1,"\u53d8\u4e3a3\u4e2a\u65b0\u7684\u5b50\u6b65\u9aa4":10,"\u53d8\u4f1a\u62a5\u8fd9\u4e2a\u9519\u8bef":7,"\u53e5\u5b50\u8868\u793a\u7684\u8ba1\u7b97\u66f4\u65b0\u4e3a2\u6b65":10,"\u53ea\u662f\u5c06\u53e5\u5b50\u5229\u7528\u8fde\u7eed\u5411\u91cf\u8868\u793a\u66ff\u6362\u7a00\u758f":10,"\u53ea\u662f\u8bf4\u660e\u6570\u636e\u7684\u987a\u5e8f\u662f\u91cd\u8981\u7684":20,"\u53ea\u9700\u8981\u4f7f\u7528\u4e00\u884c\u4ee3\u7801\u5373\u53ef\u4ee5\u8bbe\u7f6e\u8bad\u7ec3\u5f15\u7528\u8fd9\u4e2adataprovid":20,"\u53ea\u9700\u8981\u5728":20,"\u53ea\u9700\u8981\u77e5\u9053\u8fd9\u53ea\u662f\u4e00\u4e2a\u6807\u8bb0\u5c5e\u6027\u7684\u65b9\u6cd5\u5c31\u53ef\u4ee5\u4e86":20,"\u53ef\u4ee5\u4e3a\u4e00\u4e2a\u6570\u636e\u6587\u4ef6\u8fd4\u56de\u591a\u6761\u8bad\u7ec3\u6837\u672c":20,"\u53ef\u4ee5\u4f20\u516510k":0,"\u53ef\u4ee5\u4f7f\u7528":0,"\u53ef\u4ee5\u4f7f\u7528\u8bad\u7ec3\u597d\u7684\u6a21\u578b\u8bc4\u4f30\u5e26\u6709label\u7684\u9a8c\u8bc1\u96c6":10,"\u53ef\u4ee5\u4f7f\u7528graphviz\u5bf9paddlepaddle\u7684\u7f51\u7edc\u6a21\u578b\u8fdb\u884c\u7ed8\u5236":13,"\u53ef\u4ee5\u4f7f\u7528paddl":13,"\u53ef\u4ee5\u4f7f\u7528python\u7684":23,"\u53ef\u4ee5\u53c2\u8003":10,"\u53ef\u4ee5\u5728\u4e00\u4e2a\u51fd\u6570\u91cc":20,"\u53ef\u4ee5\u5728cmake\u7684\u547d\u4ee4\u884c\u8bbe\u7f6e":1,"\u53ef\u4ee5\u5c06\u4e00\u6761\u6570\u636e\u8bbe\u7f6e\u6210\u591a\u4e2abatch":20,"\u53ef\u4ee5\u5c06paddlepaddle\u7684\u6a21\u578b\u548c\u914d\u7f6e\u6253\u5305\u6210\u4e00\u4e2a\u6587\u4ef6":13,"\u53ef\u4ee5\u5c06paddlepaddle\u7684\u8bad\u7ec3\u6a21\u578b\u4ee5proto":13,"\u53ef\u4ee5\u65b9\u4fbf\u5d4c\u5165\u5f0f\u5de5\u4f5c":1,"\u53ef\u4ee5\u6839\u636e\u4e0d\u540c\u7684\u6570\u636e\u914d\u7f6e\u4e0d\u540c\u7684\u8f93\u5165\u7c7b\u578b":20,"\u53ef\u4ee5\u901a\u8fc7show":10,"\u53ef\u7528\u5728\u6d4b\u8bd5\u6216\u8bad\u7ec3\u65f6\u6307\u5b9a\u521d\u59cb\u5316\u6a21\u578b":10,"\u53ef\u9009":20,"\u5404\u79cd\u53c2\u6570\u548c\u7ef4\u62a4":0,"\u540c\u65f6":[0,1,6,20],"\u540c\u65f6\u4f1a\u8ba1\u7b97\u5206\u7c7b\u51c6\u786e\u7387":10,"\u540c\u65f6\u5305\u542bcpu\u5f00\u53d1\u73af\u5883":6,"\u540c\u65f6\u5305\u542bcpu\u5f00\u53d1\u73af\u5883\u548c\u6e90\u7801":6,"\u540c\u65f6\u5305\u542bgpu\u5f00\u53d1\u73af\u5883":6,"\u540c\u65f6\u5305\u542bgpu\u5f00\u53d1\u73af\u5883\u548c\u6e90\u7801":6,"\u540c\u65f6\u6b22\u8fce\u8d21\u732e\u66f4\u591a\u7684\u5b89\u88c5\u5305":5,"\u540d\u79f0":10,"\u540e\u9762\u8ddf\u7740\u4e00\u7cfb\u5217\u7f16\u8bd1\u53c2\u6570":18,"\u5411\u91cf\u8868\u793a":10,"\u5426":1,"\u5426\u5219":19,"\u547d\u4ee4":0,"\u547d\u4ee4\u4e3a":6,"\u547d\u4ee4\u6307\u5b9a\u7684\u53c2\u6570\u4f1a\u4f20\u5165\u7f51\u7edc\u914d\u7f6e\u4e2d":10,"\u547d\u4ee4\u8fd0\u884c\u955c\u50cf":6,"\u547d\u4ee4\u9884\u5148\u4e0b\u8f7d\u955c\u50cf":6,"\u548c":[1,6,20],"\u548c\u4e09\u79cd\u5e8f\u5217\u6a21\u5f0f":20,"\u548c\u5dee\u8bc4":10,"\u548c\u6587\u672c\u4fe1\u606f\u7528tab\u95f4\u9694":10,"\u548c\u6d4b\u8bd5\u6587\u4ef6\u5217\u8868":19,"\u548c\u7528\u6237\u4f20\u5165\u7684\u53c2\u6570":20,"\u548c\u9884\u5904\u7406\u811a\u672c":10,"\u548ccudnn":7,"\u548cgpu\u7248\u672c":7,"\u548cinitalizer\u91cc\u5b9a\u4e49\u987a\u5e8f\u4e00\u81f4":10,"\u5668":10,"\u56db\u79cd\u6570\u636e\u7c7b\u578b\u662f":20,"\u56fe\u50cf\u5206\u7c7b":9,"\u5728":[1,7,20],"\u5728\u58f0\u660edataprovider\u7684\u65f6\u5019\u4f20\u5165\u4e86dictionary\u4f5c\u4e3a\u53c2\u6570":20,"\u5728\u5b8c\u6210\u4e86\u6570\u636e\u548c\u7f51\u7edc\u7ed3\u6784\u642d\u5efa\u4e4b\u540e":10,"\u5728\u672c\u95ee\u9898\u4e2d":10,"\u5728\u6a21\u578b\u914d\u7f6e\u4e2d\u5229\u7528":10,"\u5728\u6b64\u4e3a\u65b9\u4fbf\u5bf9\u6bd4\u4e0d\u540c\u7f51\u7edc\u7ed3\u6784":10,"\u5728\u6bcf\u4e2a\u7ef4\u5ea6\u4e0a\u53d6\u51fa\u5728\u8be5\u53e5\u8bdd\u65b0\u7684\u5411\u91cf\u96c6\u5408\u4e0a\u8be5\u7ef4\u5ea6\u7684\u6700\u5927\u503c\u4f5c\u4e3a\u6700\u540e\u7684\u53e5\u5b50\u8868\u793a\u5411\u91cf":10,"\u5728\u7a0b\u5e8f\u5f00\u59cb\u9636\u6bb5":23,"\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d\u8fdb\u884c\u6d4b\u8bd5":19,"\u5728\u8bad\u7ec3\u914d\u7f6e\u91cc":20,"\u5728\u8fd9\u4e2a\u51fd\u6570\u4e2d":20,"\u5728\u914d\u7f6e\u4e2d\u8bfb\u53d6\u4e86\u5b57\u5178":20,"\u5728cmake\u914d\u7f6e\u65f6\u53ef\u4ee5\u4f7f\u7528":1,"\u5728pydataprovider\u4e2d":20,"\u5728python\u73af\u5883\u4e0b\u9884\u6d4b\u7ed3\u679c":23,"\u57fa\u672c\u7684\u5904\u7406\u903b\u8f91\u4e5f\u548cmnist\u903b\u8f91\u4e00\u81f4":20,"\u57fa\u672c\u7684pydataprovider\u4f7f\u7528\u4ecb\u7ecd\u5b8c\u6bd5\u4e86":20,"\u591a\u4e2ainput\u4ee5list\u65b9\u5f0f\u8f93\u5165":10,"\u591a\u6b21\u8fd4\u56de\u53d8\u91cf":20,"\u591a\u7ebf\u7a0b\u4e0b\u8f7d\u8fc7\u7a0b\u4e2d":0,"\u591a\u7ebf\u7a0b\u6570\u636e\u8bfb\u53d6":20,"\u597d\u8bc4":10,"\u5982\u679c\u4e0d\u4e86\u89e3":20,"\u5982\u679c\u4e0d\u4f7f\u7528\u5219\u4f1a\u4f7f\u7528\u4e00\u4e2a\u7b80\u5316\u7248\u7684\u547d\u4ee4\u884c\u53c2\u6570\u89e3\u6790":1,"\u5982\u679c\u4e0d\u4f7f\u7528\u5219\u4f1a\u4f7f\u7528\u4e00\u4e2a\u7b80\u5316\u7248\u7684\u65e5\u5fd7\u5b9e\u73b0":1,"\u5982\u679c\u4e0d\u5207\u8bcd":10,"\u5982\u679c\u4e0d\u8bbe\u7f6e\u7684\u8bdd":20,"\u5982\u679c\u4f7f\u7528gpu\u7248\u672c\u7684paddlepaddl":7,"\u5982\u679c\u5728":7,"\u5982\u679c\u5728\u7b2c\u4e00\u6b21cmake\u4e4b\u540e\u60f3\u8981\u91cd\u65b0\u8bbe":1,"\u5982\u679c\u5c0f\u4e8e\u8fd9\u4e2a\u4e0b\u8f7d\u901f\u5ea6":0,"\u5982\u679c\u60a8\u4f7f\u7528":6,"\u5982\u679c\u60f3\u8981\u5728\u5916\u90e8\u673a\u5668\u8bbf\u95ee\u8fd9\u4e2acontain":6,"\u5982\u679c\u6709\u66f4\u590d\u6742\u7684\u4f7f\u7528":19,"\u5982\u679c\u7cfb\u7edf\u4f7f\u7528\u4e0d\u652f\u6301":7,"\u5982\u679c\u8fd0\u884cgpu\u7248\u672c\u7684paddlepaddl":6,"\u5982\u679ctest":19,"\u5b83\u5305\u542b\u7684\u53c2\u6570\u6709":20,"\u5b89\u88c5\u597d\u7684paddlepaddle\u811a\u672c\u5305\u62ec\u591a\u6761\u547d\u4ee4":13,"\u5b89\u88c5\u5b8c\u6210\u7684paddlepaddle\u4e3b\u4f53\u5305\u62ec\u4e09\u4e2a\u90e8\u5206":6,"\u5b89\u88c5\u5b8c\u6210paddlepaddle\u540e":7,"\u5b89\u88c5\u6559\u7a0b":10,"\u5b89\u88c5\u65b9\u6cd5\u8bf7\u53c2\u8003":6,"\u5b89\u88c5\u7f16\u8bd1\u4f9d\u8d56":3,"\u5b89\u88c5\u7f16\u8bd1paddlepaddle\u9700\u8981\u7684\u4f9d\u8d56":2,"\u5b89\u88c5docker\u9700\u8981\u60a8\u7684\u673a\u5668":6,"\u5b89\u88c5paddlepaddl":10,"\u5b89\u88c5paddlepaddle\u7684docker\u955c\u50cf":5,"\u5b8c\u6210\u591a\u673a\u8bad\u7ec3":13,"\u5b8c\u6574\u4ee3\u7801\u89c1":23,"\u5b9a\u4e49\u6587\u672c\u4fe1\u606f":10,"\u5b9e\u73b0\u4e86\u6253\u5f00\u6587\u672c\u6587\u4ef6":20,"\u5bc6\u7801\u4e5f\u662froot":6,"\u5bf9\u4e8e\u7528\u6237\u6765\u8bf4":20,"\u5bf9\u4e8e\u7ed9\u5b9a\u7684\u4e00\u6761\u6587\u672c":10,"\u5bf9\u4e8ecuda\u7684toolkit\u6709\u65ad\u70b9\u7eed\u4f20\u548c\u4f20\u8f93\u901f\u5ea6\u8fc7\u5c0f\u91cd\u542f\u4e0b\u8f7d\u7684":0,"\u5bf9\u4e8emnist\u800c\u8a00":20,"\u5bf9\u8be5\u8868\u793a\u8fdb\u884c\u975e\u7ebf\u6027\u53d8\u6362":10,"\u5bf9\u8c61":20,"\u5bf9\u8c61convert":23,"\u5c06\u4f1a\u6d4b\u8bd5\u914d\u7f6e\u6587\u4ef6\u4e2dtest":10,"\u5c06\u5b57\u5178\u5b58\u5165\u4e86set":20,"\u5c06\u5bbf\u4e3b\u673a\u76848022\u7aef\u53e3\u6620\u5c04\u5230container\u768422\u7aef\u53e3\u4e0a":6,"\u5c06\u6570\u636e\u7ec4\u5408\u6210batch\u8bad\u7ec3":20,"\u5c06\u6587\u4ef6\u7684\u7edd\u5bf9\u8def\u5f84\u6216\u76f8\u5bf9\u8def\u5f84":19,"\u5c06\u8bc4\u8bba\u5206\u4e3a\u597d\u8bc4":10,"\u5c06\u8be5\u53e5\u8bdd\u5305\u542b\u7684\u6240\u6709\u5355\u8bcd\u5411\u91cf\u6c42\u5e73\u5747\u5f97\u5230\u53e5\u5b50\u7684\u8868\u793a":10,"\u5c06ssh\u88c5\u5165\u7cfb\u7edf\u5185\u5e76\u5f00\u542f\u8fdc\u7a0b\u8bbf\u95ee":6,"\u5c31\u50cf\u8fd9\u4e2a\u6837\u4f8b\u4e00\u6837":20,"\u5c31\u662f\u5c06\u8fd9\u4e9b\u52a8\u6001\u5e93\u52a0\u5230\u73af\u5883\u53d8\u91cf\u91cc\u9762":7,"\u5c5e\u6027":20,"\u5dee\u8bc4":10,"\u5e38\u89c1\u7684\u8f93\u51fa\u683c\u5f0f\u4e3a":18,"\u5e76\u4e14\u4f7f\u7528\u5173\u952e\u8bcd":20,"\u5e76\u4e14\u5220\u9664container\u4e2d\u7684\u6570\u636e":6,"\u5e76\u4e14\u6807\u8bb0process\u51fd\u6570\u662f\u4e00\u4e2adataprovid":20,"\u5e76\u4f7f\u7528\u4e86dropout":10,"\u5e76\u572823\u884c\u8fd4\u56de\u7ed9paddlepaddle\u8fdb\u7a0b":20,"\u5e76\u5c06\u6bcf\u884c\u8f6c\u6362\u6210\u548c":20,"\u5e76\u63d0\u4f9b":6,"\u5e76\u63d0\u4f9b\u4e86\u7b80\u5355\u7684cache\u529f\u80fd":20,"\u5e76\u8bbe\u7f6e\u597d\u5bf9\u5e94\u7684\u73af\u5883\u53d8\u91cf":7,"\u5e76\u9010\u6e10\u5c55\u793a\u66f4\u52a0\u6df1\u5165\u7684\u529f\u80fd":10,"\u5e8f\u5217\u6a21\u578b\u6570\u636e\u63d0\u4f9b":19,"\u5e8f\u5217\u6a21\u578b\u662f\u6307\u6570\u636e\u7684\u67d0\u4e00\u7ef4\u5ea6\u662f\u4e00\u4e2a\u5e8f\u5217\u5f62\u5f0f":20,"\u5e93\u7684\u8bdd":7,"\u5f0f":20,"\u5f15\u7528\u7684dataprovider\u662f":20,"\u5f53\u524dlog_period\u4e2abatch\u6240\u6709\u6837\u672c\u7684\u5e73\u5747\u5206\u7c7b\u9519\u8bef\u7387":10,"\u5f53\u524dlog_period\u4e2abatch\u6240\u6709\u6837\u672c\u7684\u5e73\u5747cost":10,"\u5f53\u7136":19,"\u5f88":10,"\u5f97\u5230\u7ed3\u679c":7,"\u5feb\u901f\u5165\u95e8":11,"\u5ff5\u662f":20,"\u60a8\u53ef\u4ee5\u4f7f\u7528":6,"\u60a8\u5c31\u53ef\u4ee5\u8fdc\u7a0b\u7684\u4f7f\u7528paddlepaddle\u5566":6,"\u60a8\u9700\u8981\u5728\u673a\u5668\u4e2d\u5b89\u88c5\u597ddocker":6,"\u60a8\u9700\u8981\u8fdb\u5165\u955c\u50cf\u8fd0\u884cpaddlepaddl":6,"\u60c5\u611f\u5206\u6790":9,"\u60f3\u8981\u8fd0\u884cpaddlepaddl":6,"\u6210\u4e3a\u7ef4\u5ea6\u4e3ahidden":10,"\u6211\u4eec\u4ece\u63d0\u524d\u7ed9\u5b9a\u7684\u7c7b\u522b\u96c6\u5408\u4e2d\u9009\u62e9\u5176\u6240\u5c5e\u7c7b":10,"\u6211\u4eec\u4ee5\u6587\u672c\u5206\u7c7b\u95ee\u9898\u4f5c\u4e3a\u80cc\u666f":10,"\u6211\u4eec\u4f7f\u7528":10,"\u6211\u4eec\u5728\u6b64\u603b":10,"\u6211\u4eec\u5c06\u4ee5\u57fa\u672c\u7684\u903b\u8f91\u56de\u5f52\u7f51\u7edc\u4f5c\u4e3a\u8d77\u70b9":10,"\u6211\u4eec\u5c06\u5728\u540e\u9762\u4ecb\u7ecd\u8bad\u7ec3\u548c\u9884\u6d4b\u7684\u6d41\u7a0b\u7684\u811a\u672c":10,"\u6211\u4eec\u5c06\u8bad\u7ec3\u7684\u547d\u4ee4\u884c\u4fdd\u5b58\u5728\u4e86":10,"\u6211\u4eec\u63a8\u8350\u4f7f\u7528docker\u955c\u50cf\u6765\u90e8\u7f72\u73af\u5883":5,"\u6211\u4eec\u63d0\u4f9b\u4e866\u4e2adocker":6,"\u6211\u4eec\u63d0\u4f9b\u4e86\u4e00\u4e2a\u5de5\u5177\u7c7bdataproviderconvert":23,"\u6211\u4eec\u8fdb\u5165\u5230\u8bad\u7ec3\u90e8\u5206":10,"\u6211\u4eec\u91c7\u7528\u5355\u5c42lstm\u6a21\u578b":10,"\u6211\u4eec\u968f\u65f6\u603b\u7ed3\u4e86\u5404\u4e2a\u7f51\u7edc\u7684\u590d\u6742\u5ea6\u548c\u6548\u679c":10,"\u6216\u5176\u4ed6":10,"\u6216\u8005":6,"\u6216\u800510g\u8fd9\u6837\u7684\u5355\u4f4d":0,"\u6216\u8005\u6570\u636e\u5e93\u8fde\u63a5\u5730\u5740\u7b49\u7b49":19,"\u6216\u8005\u8bbe\u7f6e\u4e3anone":19,"\u6216\u8005\u9700\u8981\u66f4\u9ad8\u7684\u6548\u7387":19,"\u6216\u8005\u9ad8\u6027\u80fd\u7684":6,"\u6240\u4ee5":[20,23],"\u6240\u4ee5\u5728cpu\u7684\u8fd0\u7b97\u6027\u80fd\u4e0a\u5e76\u4e0d\u4f1a\u6709\u4e25\u91cd\u7684\u5f71\u54cd":6,"\u6240\u4ee5\u5982\u679c\u60f3\u8981\u5728\u540e\u53f0\u542f\u7528ssh":6,"\u6240\u4ee5\u5c06":20,"\u6240\u4ee5\u7528\u4e8e\u9884\u6d4b\u7684\u914d\u7f6e\u6587\u4ef6\u8981\u505a\u76f8\u5e94\u7684\u4fee\u6539":23,"\u6240\u4ee5\u8f93\u51fa\u7684value\u5305\u542b\u4e24\u4e2a\u5411\u91cf":23,"\u6240\u4ee5gpu\u5728\u8fd0\u7b97\u6027\u80fd\u4e0a\u4e5f\u4e0d\u4f1a\u6709\u4e25\u91cd\u7684\u5f71\u54cd":6,"\u6240\u4ee5init_hook\u5c3d\u91cf\u4f7f\u7528":20,"\u6240\u6709\u5b57\u7b26\u90fd\u5c06\u8f6c\u6362\u4e3a\u8fde\u7eed\u6574\u6570\u8868\u793a\u7684id\u4f20\u7ed9\u6a21\u578b":10,"\u6240\u6709\u6587\u4ef6\u5217\u8868":20,"\u6240\u6709\u7684":1,"\u6240\u6709\u7684\u4e0b\u8f7d\u7ebf\u7a0b\u5173\u95ed\u65f6":0,"\u6240\u6709\u914d\u7f6e\u5728":10,"\u6240\u8c13\u65f6\u95f4\u6b65\u4fe1\u606f":20,"\u6253\u5370\u7684\u65e5\u5fd7\u53d8\u591a":1,"\u6267\u884c":0,"\u6267\u884c\u5982\u4e0b\u547d\u4ee4\u5373\u53ef\u4ee5\u5173\u95ed\u8fd9\u4e2acontain":6,"\u6267\u884c\u65b9\u6cd5\u5982\u4e0b":6,"\u62a5\u9519":7,"\u62fc\u63a5\u6210\u4e00\u4e2a\u65b0\u7684\u5411\u91cf\u8868\u793a":10,"\u6307\u4ee4\u96c6\u7684cpu\u8fd0\u884cpaddlepaddl":7,"\u6307\u4ee4\u96c6\u7684x86":7,"\u6307\u5b9a\u521d\u59cb\u5316\u6a21\u578b\u8def\u5f84":10,"\u6307\u5b9a\u751f\u6210\u6570\u636e\u7684\u51fd\u6570":10,"\u6307\u5b9a\u8bad\u7ec3":10,"\u6307\u5b9abatch":10,"\u6307\u5b9aoutputs\u8f93\u51fa\u6982\u7387\u5c42":10,"\u6389\u7f16\u8bd1\u76ee\u5f55\u540e":1,"\u63a5\u4e0b\u6765\u4f7f\u7528":23,"\u63a5\u53e3\u4f7f\u7528\u591a\u7ebf\u7a0b\u8bfb\u53d6\u6570\u636e":20,"\u63a8\u8350\u4f7f\u7528\u5c06\u672c\u5730\u7f51\u5361":6,"\u63a8\u8350\u4f7f\u7528\u6700\u65b0\u7248\u672c\u7684cudnn":1,"\u63a8\u8350\u6e05\u7406":1,"\u63a8\u8350\u76f4\u63a5\u653e\u7f6e\u5230\u8bad\u7ec3\u76ee\u5f55":19,"\u63cf\u8ff0":1,"\u63cf\u8ff0\u4e86docker":0,"\u6548\u679c\u4e00\u81f4":20,"\u6548\u679c\u603b\u7ed3":10,"\u6559\u7a0b":10,"\u6570\u636e":20,"\u6570\u636e\u4f20\u8f93\u65e0\u9700label\u6570\u636e":10,"\u6570\u636e\u5904\u7406python\u6587\u4ef6\u540d":10,"\u6570\u636e\u5982\u4f55\u5b58\u50a8\u7b49\u7b49":20,"\u6570\u636e\u63d0\u4f9b":19,"\u6570\u636e\u6587\u4ef6\u5b58\u653e\u5728\u672c\u5730\u78c1\u76d8\u4e2d":19,"\u6570\u636e\u662f\u7ed9\u4e00\u6bb5\u82f1\u6587\u6587\u672c":20,"\u6570\u636e\u683c\u5f0f\u548c\u8be6\u7ec6\u6587\u6863\u8bf7\u53c2\u8003":10,"\u6587\u4ef6\u4e2d":10,"\u6587\u672c\u4e2d\u7684\u5355\u8bcd\u7528\u7a7a\u683c\u5206\u9694":10,"\u6587\u672c\u4fe1\u606f\u5c31\u662f\u4e00\u4e2a\u5e8f\u5217":20,"\u6587\u672c\u5206\u7c7b\u95ee\u9898":10,"\u6587\u672c\u5377\u79ef\u5206\u4e3a\u4e09\u4e2a\u6b65\u9aa4":10,"\u6587\u672c\u751f\u6210":9,"\u65b0\u5199layer":11,"\u65b9\u4fbf\u8c03\u8bd5\u4f7f\u7528":13,"\u65b9\u4fbf\u90e8\u7f72\u5206\u53d1":13,"\u65e0\u9700label\u76f8\u5173\u7684\u5c42":10,"\u65f6\u5e8f\u6a21\u578b\u5373\u4e3arnn\u6a21\u578b":10,"\u65f6\u5e8f\u6a21\u578b\u5747\u4f7f\u7528\u8be5\u811a\u672c":10,"\u662f":1,"\u662f\u4e00\u4e2abatch":20,"\u662f\u4e00\u4e2apython\u7684":20,"\u662f\u4e00\u4e2aswig\u5c01\u88c5\u7684paddlepaddle\u5305":6,"\u662f\u4e0d\u662f\u5f88\u7b80\u5355\u5462":20,"\u662f\u4e2adataprovider\u662f\u4e0d\u662f\u8981\u505ashuffl":20,"\u662f\u4ec0\u4e48\u4e5f\u6ca1\u5173\u7cfb":20,"\u662f\u4ece\u8bad\u7ec3\u914d\u7f6e\u4f20\u5165\u7684dict\u5bf9\u8c61":20,"\u662f\u521d\u59cb\u5316\u65f6\u8c03\u7528\u7684\u51fd\u6570":20,"\u662f\u540c\u4e00\u4e2a\u5bf9\u8c61":20,"\u662f\u5426\u4f7f\u7528\u53cc\u7cbe\u5ea6\u6d6e\u70b9\u6570":1,"\u662f\u5426\u4f7f\u7528\u8fd0\u884c\u65f6\u52a8\u6001\u52a0\u8f7dcuda\u52a8\u6001\u5e93":1,"\u662f\u5426\u4f7f\u7528gflags":1,"\u662f\u5426\u4f7f\u7528glog":1,"\u662f\u5426\u5185\u5d4cpython\u89e3\u91ca\u5668":1,"\u662f\u5426\u5bfb\u627e\u5230cuda\u5de5\u5177\u94fe":1,"\u662f\u5426\u5f00\u542f\u5355\u5143\u6d4b\u8bd5":1,"\u662f\u5426\u5f00\u542f\u8ba1\u65f6\u529f\u80fd\u5f00\u542f\u8ba1\u65f6\u529f\u80fd\u4f1a\u5bfc\u81f4\u8fd0\u884c\u7565\u6162":1,"\u662f\u5426\u5f00\u542fgpu\u529f\u80fd":0,"\u662f\u5426\u5f00\u542frdma\u652f\u6301":1,"\u662f\u5426\u7f16\u8bd1\u4e2d\u6587\u6587\u6863":1,"\u662f\u5426\u7f16\u8bd1\u542b\u6709avx\u6307\u4ee4\u96c6\u7684paddlepaddle\u4e8c\u8fdb\u5236":1,"\u662f\u5426\u7f16\u8bd1\u65f6\u8fdb\u884c\u4ee3\u7801\u98ce\u683c\u68c0\u67e5":1,"\u662f\u5426\u7f16\u8bd1\u82f1\u6587\u6587\u6863":1,"\u662f\u5426\u7f16\u8bd1gpu\u652f\u6301":1,"\u662f\u5426\u7f16\u8bd1python\u7684swig\u63a5\u53e3":1,"\u662f\u5728\u8fd0\u884c\u65f6\u6267\u884c\u7684":20,"\u662f\u6570\u636e\u7f13\u5b58\u7684\u7b56\u7565":20,"\u662f\u6570\u636e\u8f93\u5165\u683c\u5f0f":20,"\u662f\u8bbe\u7f6e\u8fd9\u4e2adataprovider\u8fd4\u56de\u4ec0\u4e48\u6837\u7684\u6570\u636e":20,"\u662f\u8bbe\u7f6edataprovider\u5728\u5185\u5b58\u4e2d\u6682\u5b58\u7684\u6570\u636e\u6761\u6570":20,"\u662fpaddlepaddle\u8d1f\u8d23\u63d0\u4f9b\u6570\u636e\u7684\u6a21\u5757":19,"\u662fpython\u7684\u4e00\u4e2a\u5173\u952e\u8bcd":20,"\u663e":10,"\u66f4\u591a\u7684\u5b89\u88c5\u5305paddlepaddle\u4f1a\u5728\u8fd1\u671f\u63d0\u4f9b":7,"\u66f4\u8be6\u7ec6\u7528\u4f8b\u8bf7\u53c2\u8003\u6587\u6863":10,"\u66f4\u8be6\u7ec6\u7684\u4ecb\u7ecd\u8bf7\u53c2\u8003\u5404\u4e2a\u547d\u4ee4\u7684\u547d\u4ee4\u884c\u53c2\u6570\u6587\u6863":13,"\u66f4\u8be6\u7ec6\u7684\u7f51\u7edc\u914d\u7f6e":10,"\u6700\u4f4e\u7ebf\u7a0b\u7684\u4e0b\u8f7d\u901f\u5ea6":0,"\u6700\u540e\u4f7f\u7528":23,"\u6709\u503c\u7684\u4f4d\u7f6e\u53ea\u80fd\u53d61":20,"\u6709\u503c\u7684\u90e8\u5206\u53ef\u4ee5\u662f\u4efb\u4f55\u6d6e\u70b9\u6570":20,"\u6709\u90e8\u5206\u53c2\u6570\u662fpaddle\u81ea\u52a8\u751f\u6210\u7684":20,"\u672c\u8282\u6211\u4eec\u5c06\u4e13\u6ce8\u4e8e\u7f51\u7edc\u7ed3\u6784\u7684\u4ecb\u7ecd":10,"\u6765\u63a5\u53d7\u4e0d\u4f7f\u7528\u7684":20,"\u6765\u81ea\u5b9a\u4e49\u4f20\u6570\u636e\u7684\u8fc7\u7a0b":19,"\u6765\u8bf4\u660e\u7b80\u5355\u7684pydataprovider\u5982\u4f55\u4f7f\u7528":20,"\u6765\u8fdb\u884c\u8bad\u7ec3":6,"\u6765\u914d\u7f6ecudnn\u7684\u5b89\u88c5\u8def\u5f84":1,"\u6784\u9020gradientmachin":23,"\u6790\u597d\u7684\u914d\u7f6e\u521b\u5efa\u795e\u7ecf\u7f51\u7edc":23,"\u6807\u7b7e\u662f0":20,"\u6837\u4f8b\u6570\u636e\u4e3a":20,"\u6837\u4f8b\u6570\u636e\u5982\u4e0b":20,"\u6837\u672c":20,"\u6839\u636e\u4e0a\u4e00\u6b65\u89e3":23,"\u683c\u5f0f\u5982\u4e0b":10,"\u68d2":10,"\u6a21\u578b\u5b58\u50a8\u8def\u5f84":10,"\u6a21\u578b\u8bad\u7ec3\u4f1a\u770b\u5230\u8fd9\u6837\u7684\u65e5\u5fd7":10,"\u6a21\u578b\u914d\u7f6e":11,"\u6b22\u8fce\u5927\u5bb6\u8d21\u732e\u5404\u4e2a\u53d1\u884c\u7248\u7684\u5b89\u88c5\u5305":7,"\u6b63\u6837\u672c":10,"\u6bcf\u4e00\u4e2a\u4efb\u52a1\u6d41\u7a0b\u90fd\u53ef\u4ee5\u5206\u4e3a\u5982\u4e0b5\u4e2a\u57fa\u7840\u90e8\u5206":10,"\u6bcf\u4e00\u884c":20,"\u6bcf\u4e2a\u5c42\u90fd\u6709\u4e00\u4e2a\u6216\u591a\u4e2ainput":10,"\u6bcf\u4e2apass\u7684\u7b2c0\u4e2abatch\u5230\u5f53\u524dbatch\u6240\u6709\u6837\u672c\u7684\u5e73\u5747\u5206\u7c7b\u9519\u8bef\u7387":10,"\u6bcf\u4e2apass\u7684\u7b2c0\u4e2abatch\u5230\u5f53\u524dbatch\u6240\u6709\u6837\u672c\u7684\u5e73\u5747cost":10,"\u6bcf\u6b21\u90fd\u4f1a\u4ecepython\u7aef\u8bfb\u53d6\u6570\u636e":20,"\u6bcf\u884c\u4fdd\u5b58\u4e00\u6761\u6837\u672c":10,"\u6bcf\u9694\u591a\u5c11batch\u6253\u5370\u4e00\u6b21\u65e5\u5fd7":10,"\u6bd4\u5982\u901a\u8fc7\u7528\u6237\u5bf9\u7535\u5b50\u5546\u52a1\u7f51\u7ad9\u8bc4\u8bba":10,"\u6bd4\u8f83\u53ef\u80fd\u7684\u547d\u4ee4\u5982\u4e0b":7,"\u6ca1\u6709\u8bbe\u7f6e":7,"\u6ce8\u610f":[0,5,20],"\u6d41\u7a0b\u5982\u4e0b":10,"\u6d4b\u8bd5\u6570\u636e":10,"\u6d4b\u8bd5\u7684\u65f6\u5019\u9ed8\u8ba4\u4e0dshuffl":20,"\u6d4b\u8bd5\u811a\u672c\u5982\u4e0b":10,"\u6e90\u7801":10,"\u6e90\u7801\u548c\u8fd0\u884cdemo\u7684\u5fc5\u8981\u4f9d\u8d56":6,"\u6e90\u7801\u6839\u76ee\u5f55":0,"\u6fc0\u6d3b\u51fd\u6570\u7c7b\u578b":10,"\u7136\u540e\u6267\u884c\u5982\u4e0b\u547d\u4ee4\u5373\u53ef\u5b8c\u6210\u5b89\u88c5":7,"\u7136\u540e\u8fd0\u884c\u8fd9\u4e2acontainer\u5373\u53ef":6,"\u751f\u6210\u5404\u4e2a\u5e73\u53f0\u7684makefil":1,"\u75280\u548c1\u8868\u793a":20,"\u7528\u4e86\u4e24\u4e2a\u6708\u4e4b\u540e\u8fd9\u4e2a\u663e\u793a\u5668\u5c4f\u5e55\u788e\u4e86":10,"\u7528\u6237\u4e5f\u53ef\u4ee5\u5728c":19,"\u7528\u6237\u53ef\u4ee5\u4f7f\u7528python\u7684":19,"\u7528\u6237\u53ef\u4ee5\u6839\u636e\u8bad\u7ec3log\u9009\u62e9test\u7ed3\u679c\u6700\u597d\u7684\u6a21\u578b\u6765\u9884\u6d4b":10,"\u7528\u6237\u540d\u4e3a":6,"\u7528\u6237\u5728dataprovider\u4e2d\u9700\u8981\u5b9e\u73b0\u5982\u4f55\u8bbf\u95ee\u5176\u4e2d\u6bcf\u4e00\u4e2a\u6587\u4ef6":19,"\u7528\u6237\u5b9a\u4e49\u7684\u53c2\u6570\u4f7f\u7528args\u5728\u8bad\u7ec3\u914d\u7f6e\u4e2d\u8bbe\u7f6e":20,"\u7528\u6237\u63a5\u53e3":11,"\u7528\u6237\u9700\u8981\u5148\u5c06paddlepaddle\u5b89\u88c5\u5305\u4e0b\u8f7d\u5230\u672c\u5730":7,"\u7528\u6765\u505a\u9884\u6d4b\u548c\u7b80\u5355\u7684\u5b9a\u5236\u5316":6,"\u7531\u4e8e\u6570\u636e\u662f\u4e24\u6761":23,"\u7531\u4e8edocker\u662f\u57fa\u4e8e\u5bb9\u5668\u7684\u8f7b\u91cf\u5316\u865a\u62df\u65b9\u6848":6,"\u7531\u4e8epaddlepaddle\u7684docker\u955c\u50cf\u5e76\u4e0d\u5305\u542b\u4efb\u4f55\u9884\u5b9a\u4e49\u7684\u8fd0\u884c\u547d\u4ee4":6,"\u7531\u6613\u5230\u96be\u5c55\u793a4\u79cd\u4e0d\u540c\u7684\u7f51\u7edc\u914d\u7f6e":10,"\u7684":10,"\u7684\u5b89\u88c5\u6587\u6863":6,"\u7684\u6587\u6863":20,"\u7684\u65f6\u5019\u5982\u679c\u62a5\u4e00\u4e9b\u4f9d\u8d56\u672a\u627e\u5230\u7684\u9519\u8bef\u662f\u6b63\u5e38\u7684":7,"\u7684\u662f":20,"\u7684\u673a\u5668\u4e0a\u8fdb\u884c":0,"\u7684\u7f51\u6865\u6765\u8fdb\u884c\u7f51\u7edc\u901a\u4fe1":6,"\u7684demo\u5b66\u4e60\u5982\u4f55\u8fdb\u884c\u591a\u673a\u8bad\u7ec3":10,"\u7684docker\u53ef\u80fd\u7f3a\u4e4f":0,"\u7684matrix":23,"\u7684python\u5305\u662fpaddlepaddle\u7684\u8bad\u7ec3\u4e3b\u8981\u7a0b\u5e8f":6,"\u7684python\u5305\u6765\u505a\u914d\u7f6e\u6587\u4ef6\u89e3\u6790\u7b49\u5de5\u4f5c":6,"\u7684python\u9884\u6d4b\u8fc7\u7a0b":10,"\u76ee\u524dpaddlepaddle\u7684\u5b89\u88c5\u5305\u53ea\u652f\u6301":7,"\u76ee\u524dpaddlepaddle\u7684\u5b89\u88c5\u5305\u8fd8\u5904\u5728pr":5,"\u76ee\u5f55":10,"\u76ee\u5f55\u4e0b":0,"\u76f4\u63a5\u63d0\u53d6\u51fa\u795e\u7ecf\u7f51\u7edcoutput\u5c42\u7684\u8f93\u51fa\u7ed3\u679c":23,"\u76f8\u5173\u547d\u4ee4\u4e3a":6,"\u76f8\u5173\u7684\u6982":20,"\u76f8\u5bf9\u4e8epaddlepaddle\u7a0b\u5e8f\u8fd0\u884c\u65f6\u7684\u8def\u5f84":19,"\u77e5\u9053\u5982\u4f55\u4ece":20,"\u793a":10,"\u7a0b\u5e8f\u6216\u8005\u81ea\u5b9a\u4e49\u4e00\u4e2a\u542b\u6709\u542f\u52a8\u811a\u672c\u7684imag":6,"\u7aef\u81ea\u5b9a\u4e49\u4e00\u4e2a":19,"\u7b2c\u4e00\u4e2a\u53c2\u6570\u662f":20,"\u7b2c\u4e00\u4e2apass\u4f1a\u4ecepython\u7aef\u8bfb\u53d6\u6570\u636e":20,"\u7b2c\u4e00\u6bb5\u6570\u636e\u4e3a\u8fd9\u5f20\u56fe\u7247\u7684label":20,"\u7b2c\u4e8c\u4e2a\u53c2\u6570\u662ffilenam":20,"\u7b2c\u4e8c\u6bb5\u6570\u636e\u4e3a\u8fd9\u4e2a\u56fe\u7247\u7684\u50cf\u7d20\u503c":20,"\u7b80\u5355\u4f18\u5316":0,"\u7b80\u5355\u7684\u4f7f\u7528":19,"\u7b80\u5355\u7684\u4f7f\u7528\u573a\u666f":19,"\u7b80\u5355\u7684\u4f7f\u7528\u6837\u4f8b\u4e3a":0,"\u7b80\u5355\u7684\u542b\u6709ssh\u7684dockerfile\u5982\u4e0b":6,"\u7b80\u5355\u7684pydataprovider\u6837\u4f8b\u5c31\u8bf4\u660e\u5b8c\u6bd5\u4e86":20,"\u7c7b\u522bid":10,"\u7c7b\u522bid\u7684\u6570\u636e\u7c7b\u578b":10,"\u7c7b\u578b\u6765\u8bbe\u7f6e":20,"\u7ec3":13,"\u7ed3\u4e0a\u8ff0\u7f51\u7edc\u7ed3\u6784\u5728amazon":10,"\u7ef4\u5ea6\u4e3aword":10,"\u7ef4\u5ea6\u662f\u7c7b\u522b\u4e2a\u6570":10,"\u7ef4\u5ea6\u662f\u8bcd\u5178\u5927\u5c0f":10,"\u7f13\u5b58\u8bad\u7ec3\u6570\u636e\u5230\u5185\u5b58":20,"\u7f16\u8bd1\u53c2\u6570\u9009\u9879\u6587\u4ef6":18,"\u7f16\u8bd1\u6587\u6863":7,"\u7f16\u8bd1\u9009\u9879":1,"\u7f16\u8bd1\u9009\u9879\u5217\u8868\u5982\u4e0b":1,"\u7f16\u8bd1paddlepaddl":7,"\u7f16\u8bd1paddlepaddle\u7684gpu\u7248\u672c\u5e76\u4e0d\u9700\u8981\u4e00\u5b9a\u5728\u5177\u6709gpu":0,"\u7f51\u7edc\u540d\u79f0":10,"\u7f51\u7edc\u914d\u7f6e":10,"\u7f6e\u8fd9\u4e9b\u53d8\u91cf":1,"\u800c":6,"\u800c\u4e09\u79cd\u5e8f\u5217\u6a21\u5f0f\u4e3a":20,"\u800c\u4e0d\u4f7f\u7528docker":6,"\u800c\u4e0d\u7528\u5173\u5fc3\u6570\u636e\u5982\u4f55\u4f20\u8f93\u7ed9paddlepaddl":20,"\u800c\u4e14\u9884\u6d4b\u7f51\u7edc\u901a\u5e38\u76f4\u63a5\u8f93\u51fa\u6700\u540e\u4e00\u5c42\u7684\u7ed3\u679c\u800c\u4e0d\u662f\u50cf\u8bad\u7ec3\u65f6\u4e00\u6837\u4ee5cost":23,"\u800c\u5728":[1,20],"\u800c\u6ca1\u6709\u6d4b\u8bd5\u6570\u636e":20,"\u800c\u7279\u5f81\u5373\u4e3a":20,"\u800c\u8fd9\u4e2a\u4e00\u822c\u8bf4\u660epaddlepaddle\u5df2\u7ecf\u5b89\u88c5\u5b8c\u6bd5\u4e86":7,"\u800c\u975e\u9759\u6001\u52a0\u8f7dcuda\u52a8\u6001\u5e93":1,"\u800cgpu\u7684\u9a71\u52a8\u548c\u8bbe\u5907\u5168\u90e8\u6620\u5c04\u5230\u4e86\u5bb9\u5668\u5185":6,"\u800cpaddlepaddle\u8fdb\u7a0b\u5e2e\u52a9\u7528\u6237\u505a\u4e86":20,"\u811a\u672c":6,"\u811a\u672c\u53ef\u4ee5\u542f\u52a8paddlepaddle\u7684\u8bad\u7ec3\u8fdb\u7a0b\u548cpserv":6,"\u811a\u672c\u548c":6,"\u811a\u672c\u7c7b\u4f3c\u4e8e":6,"\u81ea\u5b9a\u4e49\u4e00\u4e2adataprovid":19,"\u81f3\u5c11\u5177\u67093":6,"\u81f3\u6b64":[6,20],"\u83b7\u53d6\u5229\u7528one":10,"\u83b7\u53d6\u6bcf\u4e2a\u5355\u8bcd\u5de6\u53f3\u5404k\u4e2a\u8fd1\u90bb":10,"\u83b7\u53d6\u8be5\u6761\u6837\u672c\u7c7b\u522bid":10,"\u8868\u793a\u6574\u6570\u6807\u7b7e":20,"\u8868\u793a\u662f\u5426\u5141\u8bb8paddle\u6682\u5b58\u7565\u5fae\u591a\u4f59pool_size\u7684\u6570\u636e":20,"\u8868\u793a\u7a00\u758f\u7684\u5411\u91cf":20,"\u8868\u793a\u7a00\u758f\u7684\u96f6\u4e00\u5411\u91cf":20,"\u8868\u793a\u7a20\u5bc6\u7684\u6d6e\u70b9\u6570\u5411\u91cf":20,"\u8868\u793a\u8fc7\u4e8620\u4e2abatch":10,"\u8868\u793a\u8fc7\u4e862560\u4e2a\u6837\u672c":10,"\u8868\u793a\u8fd9\u4e2adataprovider\u662f\u8bad\u7ec3\u7528\u7684dataprovider\u6216\u8005\u6d4b\u8bd5\u7528\u7684":20,"\u89e3\u51b3\u529e\u6cd5\u662f\u5c06cuda":7,"\u89e3\u51b3\u65b9\u6cd5\u5f88\u7b80\u5355":7,"\u89e3\u6790\u8bad\u7ec3\u65f6\u7684\u914d\u7f6e\u6587\u4ef6":23,"\u89e3\u91ca":10,"\u8ba9\u795e\u7ecf\u7f51\u7edc\u53ef\u4ee5\u8fdb\u884c\u8bad\u7ec3":19,"\u8bad\u7ec3":6,"\u8bad\u7ec3\u6587\u4ef6\u5217\u8868":19,"\u8bad\u7ec3\u65f6\u6240\u9700\u8bbe\u7f6e\u7684\u4e3b\u8981\u53c2\u6570\u5982\u4e0b":10,"\u8bad\u7ec3\u7684\u65f6\u5019\u9ed8\u8ba4shuffl":20,"\u8bad\u7ec3\u811a\u672c":10,"\u8bad\u7ec3\u811a\u672c\u5728":10,"\u8bad\u7ec3\u8f6e\u6b21":10,"\u8bb2\u6570\u636e\u53d1\u9001\u7ed9paddlepaddl":20,"\u8bbe\u7f6e\u4e0b\u5217\u7f16\u8bd1\u9009\u9879\u65f6":1,"\u8bbe\u7f6e\u6210":20,"\u8bbe\u7f6e\u6210\u4e86\u5e8f\u5217":20,"\u8bbe\u7f6e\u8f93\u5165\u7c7b\u578b":20,"\u8bc4\u4f30\u4ea7\u54c1\u7684\u8d28\u91cf":10,"\u8bcd\u6027\u6807\u6ce8":9,"\u8be5\u5c42\u795e\u7ecf\u5143\u4e2a\u6570":10,"\u8be5\u6570\u636e":20,"\u8be5\u6784\u5efa\u811a\u672c\u5145\u5206\u8003\u8651\u4e86\u7f51\u7edc\u4e0d\u7a33\u5b9a\u7684\u60c5\u51b5":0,"\u8be5\u6a21\u578b\u4f9d\u7136\u662f\u4f7f\u7528\u903b\u8f91\u56de\u5f52\u5206\u7c7b\u7f51\u7edc\u7684\u6846\u67b6":10,"\u8be5\u76ee\u5f55\u4e0b\u6709\u4e24\u4e2a\u6587\u4ef6":0,"\u8be5\u811a\u672c\u7684\u4f7f\u7528\u65b9\u6cd5\u662f":0,"\u8be5image\u57fa\u4e8eubuntu":0,"\u8be5image\u7684\u6784\u5efa\u5728dock":0,"\u8be6\u60c5\u8bf7\u53c2\u8003":23,"\u8be6\u7ec6\u7684\u53c2\u6570\u89e3\u91ca\u5982\u4e0b\u9762\u8868\u683c":10,"\u8be6\u7ec6\u7684\u547d\u4ee4\u884c\u53c2\u6570\u8bf7\u53c2\u8003":23,"\u8be6\u7ec6\u7684cmake\u4f7f\u7528\u65b9\u6cd5\u53ef\u4ee5\u53c2\u8003":1,"\u8bf4\u660e":1,"\u8bf4\u660e\u547d\u4ee4\u884c\u53c2\u6570":7,"\u8bf7\u53c2\u8003":[6,7,20],"\u8bf7\u53c2\u8003\u4e0b\u8282refer":20,"\u8bf7\u53c2\u8003\u4e0b\u8ff0\u6587\u7ae0":19,"\u8bf7\u5b89\u88c5cuda":7,"\u8bfb\u5165\u89e3\u6790\u8bad\u7ec3\u914d\u7f6e":23,"\u8bfb\u53d6\u6570\u636e":20,"\u8c03\u7528":1,"\u8d1f\u6837\u672c":10,"\u8d1f\u8d23\u591a\u673a\u8bad\u7ec3\u4e2d\u7684\u53c2\u6570\u805a\u5408\u5de5\u4f5c":13,"\u8d1f\u9762\u60c5\u7eea\u4e24\u7c7b":20,"\u8def\u5f84\u53d8\u91cf\u4e3a":1,"\u8f93\u5165n\u4e2a\u5355\u8bcd":10,"\u8f93\u51fa\u4e3an\u4e2aword":10,"\u8fd0\u884c":[6,7],"\u8fd0\u884c\u4f7f\u7528\u7684cudnn\u5c3d\u91cf\u662f\u540c\u4e00\u4e2a\u7248\u672c":1,"\u8fd0\u884c\u8fd9\u4e2acontain":6,"\u8fd0\u884cpaddlepaddle\u7684gpu\u7248\u672c\u4e00\u5b9a\u8981\u5728\u5177\u6709cuda\u7684\u673a\u5668\u4e0a\u8fd0\u884c":0,"\u8fd4\u56de\u6570\u636e\u5728paddlepaddle\u4e2d\u662f\u4ec5\u4ec5\u8fd4\u56de\u4e00\u6761\u5b8c\u6574\u7684\u8bad\u7ec3\u6837\u672c":20,"\u8fd4\u56de\u7684\u987a\u5e8f\u9700\u8981\u548c":20,"\u8fd4\u56debatch_size\u7684\u5927\u5c0f":20,"\u8fd9":10,"\u8fd93\u4e2a\u5b50\u6b65\u9aa4\u53ef\u914d\u7f6e\u4e3a":10,"\u8fd9\u4e2a\u51fd\u6570\u4ee5\u4e00\u6761\u6570\u636e\u4e3a\u53c2\u6570":20,"\u8fd9\u4e2a\u51fd\u6570\u4f1a\u5728":20,"\u8fd9\u4e2a\u51fd\u6570\u5728\u521d\u59cb\u5316\u7684\u65f6\u5019\u4f1a\u88ab\u8c03\u7528":20,"\u8fd9\u4e2a\u51fd\u6570\u7684\u53c2\u6570\u662f":20,"\u8fd9\u4e2a\u521d\u59cb\u5316\u51fd\u6570\u5177\u6709\u5982\u4e0b\u53c2\u6570":20,"\u8fd9\u4e2a\u53c2\u6570\u5728\u8fd9\u4e2a\u6837\u4f8b\u91cc\u6ca1\u6709\u4f7f\u7528":20,"\u8fd9\u4e2a\u53c2\u6570\u88abpaddlepaddle\u8fdb\u7a0b\u4f20\u5165":20,"\u8fd9\u4e2a\u548c\u5728":20,"\u8fd9\u4e2a\u58f0\u660e\u57fa\u672c\u4e0a\u548cmnist\u7684\u6837\u4f8b\u4e00\u81f4":20,"\u8fd9\u4e2a\u5b57\u5178\u53ef\u4ee5\u5728":20,"\u8fd9\u4e2a\u5bf9\u8c61\u548cprocess\u7684\u7b2c\u4e00\u4e2a\u53c2\u6570\u4e00\u81f4":20,"\u8fd9\u4e2a\u5de5\u5177\u7c7b\u63a5\u6536\u548cpydataprovider2\u4e00\u6837\u7684\u8f93\u5165\u6570\u636e":23,"\u8fd9\u4e2a\u5e8f\u5217\u6a21\u578b\u6bd4\u8f83\u590d\u6742":20,"\u8fd9\u4e2a\u63a5\u53e3\u5e76\u4e0d\u7528\u6237\u53cb\u597d":23,"\u8fd9\u4e2a\u663e\u793a\u5668\u5f88\u68d2":10,"\u8fd9\u4e2a\u6a21\u5757\u4e2d\u7684":20,"\u8fd9\u4e2a\u8bbe\u7f6e\u4e3a":20,"\u8fd9\u4e2a\u8f6f\u4ef6\u5305\u6587\u6863\u76f8\u5bf9\u5b8c\u5584":23,"\u8fd9\u4e9b\u53c2\u6570\u5305\u62ecpaddle\u5b9a\u4e49\u7684\u53c2\u6570":20,"\u8fd9\u4e9b\u53d8":1,"\u8fd9\u4e9b\u53d8\u91cf\u53ea\u5728\u7b2c\u4e00\u6b21cmake\u7684\u65f6\u5019\u6709\u6548":1,"\u8fd9\u4e9b\u53d8\u91cf\u5747\u53ef\u4ee5\u4f7f\u7528":1,"\u8fd9\u4e9b\u6d41\u7a0b\u4e2d\u7684\u6570\u636e\u4e0b\u8f7d":10,"\u8fd9\u6837\u505a\u53ef\u4ee5\u907f\u514d\u5f88\u591a\u6b7b\u9501\u95ee\u9898":20,"\u8fd9\u884c\u7684\u4f5c\u7528\u662f\u8bbe\u7f6edataprovider\u7684\u4e00\u4e9b\u5c5e\u6027":20,"\u8fd9\u91cc":20,"\u8fd9\u91cc\u4e3e\u4f8b\u7684\u6570\u636e\u662f\u82f1\u6587\u60c5\u611f\u5206\u7c7b\u7684\u6570\u636e":20,"\u8fd9\u91cc\u4ee5":10,"\u8fd9\u91cc\u4ee5mnist\u624b\u5199\u8bc6\u522b\u4e3a\u4f8b":20,"\u8fd9\u91cc\u53ef\u4ee5\u53c2\u8003paddle\u7684":18,"\u8fd9\u91cc\u6211\u4eec\u4f7f\u7528\u7b80\u5355\u7684\u6587\u672c\u6587\u4ef6\u8868\u793amnist\u56fe\u7247":20,"\u8fd9\u91cc\u6307\u5b9a\u8bcd\u5178":10,"\u8fd9\u91cc\u6ca1\u6709\u4ecb\u7ecd\u591a\u673a\u5206\u5e03\u5f0f\u8bad\u7ec3":10,"\u8fd9\u91cc\u7684":20,"\u8fd9\u91cc\u7684\u8f93\u5165\u7279\u5f81\u662f\u8bcdid\u7684\u5e8f\u5217":20,"\u8fd9\u91cc\u8981\u6ce8\u610f\u9884\u6d4b\u6570\u636e\u901a\u5e38":23,"\u8fd9\u91cc\u8bbe\u7f6e\u7684\u662f\u8fd4\u56de\u4e00\u4e2a":20,"\u8fd9\u91cc\u8bf4\u660e\u4e86\u8bad\u7ec3\u6570\u636e\u662f":20,"\u8fd9\u91cc\u91c7\u7528adam\u4f18\u5316\u65b9\u6cd5":10,"\u8fdb\u5165\u8be5\u6e90\u7801\u76ee\u5f55":0,"\u8fdb\u5165docker":6,"\u8fdc\u7a0b\u8bbf\u95ee":6,"\u8fde\u63a5\u8bf7\u53c2\u8003":10,"\u9009\u9879":1,"\u901a\u8fc7\u7f16\u8bd1\u65f6\u6307\u5b9a\u8def\u5f84\u6765\u5b9e\u73b0\u5f15\u7528\u5404\u79cdbla":1,"\u903b\u8f91\u56de\u5f52":10,"\u90a3\u4e48":20,"\u90a3\u4e48\u5728\u8bad\u7ec3\u8fc7\u7a0b\u4e2d":19,"\u90a3\u4e48\u5bf9\u5e94\u7684dataprovider\u65e2\u4e3a":20,"\u90a3\u4e48\u8fd9\u4e2a\u4e0b\u8f7d\u7ebf\u7a0b\u5c06\u4f1a\u5173\u95ed":0,"\u90a3\u4e48\u9700\u8981\u4ece\u6e90\u7801":7,"\u914d\u7f6e\u4e86":20,"\u914d\u7f6e\u53c2\u6570\u914d\u7f6e\u7ed9dataprovider\u7684":20,"\u914d\u7f6e\u6587\u4ef6":10,"\u91cc\u4f1a\u7ee7\u7eed\u5b89\u88c5":7,"\u91cc\u63d0\u4f9b\u4e86\u6570\u636e\u4e0b\u8f7d\u811a\u672c":10,"\u91cc\u9762\u8bfb\u53d6":20,"\u91cf\u4e5f\u53ef\u4ee5\u901a\u8fc7\u8c03\u7528cmake\u547d\u4ee4\u524d\u901a\u8fc7\u73af\u5883\u53d8\u91cf\u6307\u5b9a":1,"\u9488\u5bf9\u672c\u95ee\u9898":10,"\u94fe\u63a5\u4f55\u79cdblas\u7b49\u7b49":1,"\u9519\u8bef\u7387":10,"\u95f4\u9694":20,"\u9664\u4e86":20,"\u9664\u8fc7data\u5c42":10,"\u9700\u8981\u53c2\u8003":6,"\u9700\u8981\u6ce8\u610f":20,"\u9700\u8981\u6ce8\u610f\u7684\u662f":[1,7],"\u9884\u6d4b\u6570\u636e\u6307\u5b9atest":10,"\u9884\u6d4b\u7ed3\u679c\u4ee5\u6587\u672c\u7684\u5f62\u5f0f\u4fdd\u5b58\u5728":10,"\u9884\u6d4b\u811a\u672c":10,"\u9884\u6d4bid":10,"\u989d\u5916\u7684\u53c2\u6570":10,"\u9996\u5148\u5217\u4e3e\u903b\u8f91\u56de\u5f52\u7f51\u7edc":10,"\u9996\u5148\u6211\u4eec\u5c06\u8fd9\u4e2a\u6570\u636e\u6587\u4ef6":20,"\u9996\u5148\u8bf7\u53c2\u8003":10,"\u9aa43":10,"\u9ed8\u8ba4\u4e00\u4e2apass\u4fdd\u5b58\u4e00\u6b21\u6a21\u578b":10,"\u9ed8\u8ba4\u503c":1,"\u9ed8\u8ba4\u5355\u4f4d\u662fbyte":0,"\u9ed8\u8ba4\u60c5\u51b5\u4e0b\u4e00\u6761\u6570\u636e":20,"adamax\u7b49":10,"amazon\u7535\u5b50\u4ea7\u54c1\u8bc4\u8bba\u6570\u636e":10,"api\u9884\u6d4b":10,"argument\u4f20\u5165":20,"argument\u5f62\u5f0f\u4f20\u5165":20,"atlas\u5e93\u7684\u8def\u5f84":1,"batches\u8bbe\u7f6e\u6bcf\u9694\u591a\u5c11batch\u4fdd\u5b58\u4e00\u6b21\u6a21\u578b":10,"bool\u53c2\u6570":20,"case":[10,22],"cd\u5230\u542b\u6709dockerfile\u7684\u8def\u5f84\u4e2d":6,"cmake\u53ef\u4ee5\u5c06cmake\u9879\u76ee\u6587\u4ef6":1,"cmake\u662f\u4e00\u4e2a\u8de8\u5e73\u53f0\u7684\u7f16\u8bd1\u811a\u672c":1,"cmake\u7684\u5b98\u65b9\u6587\u6863":1,"cmake\u7f16\u8bd1\u65f6\u4f1a\u9996\u5148\u5728\u7cfb\u7edf\u8def\u5f84":1,"container\u540e":6,"cuda\u76f8\u5173\u7684driver\u548c\u8bbe\u5907\u6620\u5c04\u8fdbcontainer\u4e2d":6,"d\u547d\u4ee4\u5373\u53ef":1,"d\u547d\u4ee4\u6307\u5b9a":1,"dataprovider\u521b\u5efa\u7684\u65f6\u5019\u6267\u884c":20,"dataprovider\u53ef\u4ee5\u662f":20,"dataprovider\u63d0\u4f9b\u4e86\u4e24\u79cd\u7b80\u5355\u7684cache\u7b56\u7565":20,"dataprovider\u7684\u5177\u4f53\u7528\u6cd5\u548c\u5982\u4f55\u5b9e\u73b0\u4e00\u4e2a\u65b0\u7684dataprovid":19,"dim\u7684\u65b0\u7684\u5411\u91cf":10,"dim\u7ef4\u5ea6\u5411\u91cf":10,"docker\u662f\u4e00\u4e2a\u57fa\u4e8e\u5bb9\u5668\u7684\u8f7b\u91cf\u7ea7\u865a\u62df\u73af\u5883":6,"docker\u7684\u5b98\u65b9\u6587\u6863":6,"dockerfile\u548cbuild":0,"dockerfile\u662fdock":0,"dockerfile\u7684\u6587\u6863":6,"dockerfile\u7684\u6700\u4f73\u5b9e\u8df5":6,"driver\u6dfb\u52a0\u5230ld_library_path\u4e2d":7,"elec\u6d4b\u8bd5\u96c6":10,"embedding\u6a21\u578b\u9700\u8981\u7a0d\u5fae\u6539\u53d8\u6570\u636e\u63d0\u4f9b\u7684\u811a\u672c":10,"export":[1,6,7],"f\u4ee3\u8868\u4e00\u4e2a\u6d6e\u70b9\u6570":20,"float":20,"gpu\u53cc\u7f13\u5b58":20,"gpu\u7248\u672c\u4e8c\u8fdb\u5236":1,"gru\u6a21\u578b":10,"gru\u6a21\u578b\u914d\u7f6e":10,"i\u4ee3\u8868\u4e00\u4e2a\u6574\u6570":20,"id\u4e3a0\u7684\u6982\u7387":10,"id\u4e3a1\u7684\u6982\u7387":10,"image\u6784\u5efa\u6e90\u7801\u653e\u7f6e\u5728":0,"image\u7684\u4e3b\u8981\u63cf\u8ff0\u6587\u4ef6":0,"image\u7684\u4e3b\u8981\u6784\u5efa\u6b65\u9aa4":0,"image\u7684\u6784\u5efa\u6b65\u9aa4":0,"import":[10,20,23],"include\u4e0b\u9700\u8981\u5305\u542bcbla":1,"include\u4e0b\u9700\u8981\u5305\u542bmkl":1,"init_hook\u53ef\u4ee5\u4f20\u5165\u4e00\u4e2a\u51fd\u6570":20,"int":[10,20],"layer\u4f5c\u4e3a\u8f93\u51fa":23,"layer\u6587\u6863":10,"ld_library_path\u7b49\u7b49":7,"ld_library_path\u91cc\u9762\u627e\u4e0d\u5230\u8fd9\u4e9b\u52a8\u6001":7,"lib\u4e0b\u9700\u8981\u5305\u542bcblas\u548catlas\u4e24\u4e2a\u5e93":1,"lib\u4e0b\u9700\u8981\u5305\u542bcblas\u5e93":1,"lib\u4e0b\u9700\u8981\u5305\u542bopenblas\u5e93":1,"lib\u76ee\u5f55\u4e0b\u9700\u8981\u5305\u542b":1,"list\u4e0d\u8bbe\u7f6e":19,"list\u4e2d":19,"list\u4e2d\u7684\u4e00\u884c":20,"list\u4e3a\u7eaf\u6587\u672c\u6587\u4ef6":19,"list\u4e5f\u53ef\u4ee5\u653e\u7f6ehdfs\u6587\u4ef6\u8def\u5f84":19,"list\u5373\u4e3a":20,"list\u548ctest":19,"list\u5747\u4e3a\u672c\u5730\u7684\u4e24\u4e2a\u6587\u4ef6":19,"list\u6307\u5b9a\u7684\u6570\u636e":10,"list\u7684\u4f4d\u7f6e":10,"list\u82e5\u5e72\u6570\u636e\u6587\u4ef6\u8def\u5f84\u7684\u67d0\u4e00\u4e2a\u8def\u5f84":20,"lstm\u6a21\u578b\u7b49":10,"lstm\u6a21\u578b\u914d\u7f6e":10,"make\u548cmak":2,"mkl\u7684\u8def\u5f84":1,"mkl_sequential\u548cmkl_intel_lp64\u4e09\u4e2a\u5e93":1,"mnist\u662f\u4e00\u4e2a\u5305\u542b\u6709":20,"movielens\u6570\u636e\u96c6":9,"movielens\u8bc4\u5206\u56de\u5f52":9,"osx\u6216\u8005\u662fwindows\u673a\u5668":6,"osx\u7684\u5b89\u88c5\u6587\u6863":6,"paddle\u5b9a\u4e49\u7684\u53c2\u6570\u5305\u62ec":20,"paddlepaddle\u4f7f\u7528\u8fd0\u884c\u65f6\u52a8\u6001\u8fde\u63a5cuda\u7684so":7,"paddlepaddle\u4fdd\u7559\u6dfb\u52a0\u53c2\u6570\u7684\u6743\u529b":20,"paddlepaddle\u53ef\u4ee5\u4f7f\u7528":1,"paddlepaddle\u53ef\u4ee5\u8bfb\u53d6python\u5199\u7684\u4f20\u8f93\u6570\u636e\u811a\u672c":10,"paddlepaddle\u5728\u8fd0\u884c\u65f6\u627e\u4e0d\u5230\u5bf9\u5e94\u7684config\u6587\u4ef6":7,"paddlepaddle\u63a8\u8350\u4f7f\u7528docker\u8fdb\u884cpaddlepaddle\u7684\u90e8\u7f72\u548c":6,"paddlepaddle\u63d0\u4f9b\u4e86docker\u7684\u4f7f\u7528\u955c\u50cf":6,"paddlepaddle\u63d0\u4f9b\u6570\u4e2a\u9884\u7f16\u8bd1\u7684\u4e8c\u8fdb\u5236\u6765\u8fdb\u884c\u5b89\u88c5":5,"paddlepaddle\u63d0\u4f9b\u7684\u955c\u50cf\u5e76\u4e0d\u5305\u542b\u4efb\u4f55\u547d\u4ee4\u8fd0\u884c":6,"paddlepaddle\u7684\u6570\u636e\u5305\u62ec\u56db\u79cd\u4e3b\u8981\u7c7b\u578b":20,"paddlepaddle\u7684\u7f16\u8bd1\u9009\u9879\u53ef\u4ee5\u5728\u8c03\u7528cmake\u7684\u65f6\u5019\u8bbe\u7f6e":1,"paddlepaddle\u7684\u7f16\u8bd1\u9009\u9879\u662f\u53ef\u4ee5\u63a7\u5236paddlepaddle\u751f\u6210cpu":1,"paddlepaddle\u7684cpu\u4e8c\u8fdb\u5236":6,"paddlepaddle\u7684dock":0,"paddlepaddle\u7684gpu\u4e8c\u8fdb\u5236":6,"paddlepaddle\u7684python\u9884\u6d4b\u63a5\u53e3":22,"paddlepaddle\u7684ubuntu\u5b89\u88c5\u5305\u5206\u4e3a\u4e24\u4e2a\u7248\u672c":7,"paddlepaddle\u76ee\u524d\u4f7f\u7528swig\u5bf9\u5176\u5e38\u7528\u7684\u9884\u6d4b\u63a5\u53e3\u8fdb\u884c\u4e86\u5c01\u88c5":23,"paddlepaddle\u76ee\u524d\u652f\u6301ubuntu":7,"paddlepaddle\u8fd0\u884c\u65f6\u5982\u679c\u6ca1\u6709\u5bfb\u627e\u5230cuda\u7684driv":7,"paddlepaddle\u9700\u8981\u7528\u6237\u5728\u7f51\u7edc\u914d\u7f6e":19,"period\u8bbe\u7f6e\u6253\u5370\u53c2\u6570\u4fe1\u606f\u7b49":10,"process\u51fd\u6570":20,"process\u51fd\u6570\u662f\u5b9e\u73b0\u6570\u636e\u8f93\u5165\u7684\u4e3b\u51fd\u6570":20,"process\u51fd\u6570\u8c03\u7528\u591a\u6b21":20,"pserver\u4e3apaddlepaddle\u7684paramet":13,"pserver\u7684\u547d\u4ee4\u884c\u53c2\u6570":13,"pserver\u7ec4\u5408\u4f7f\u7528":13,"py\u6587\u4ef6\u7ed9\u51fa\u4e86\u5b8c\u6574\u4f8b\u5b50":10,"pydataprovider2\u6587\u6863":23,"pydataprovider2\u7684\u4f7f\u7528":19,"pydataprovider\u662fpaddlepaddle\u4f7f\u7528python\u63d0\u4f9b\u6570\u636e\u7684\u63a8\u8350\u63a5\u53e3":20,"python\u5305":6,"python\u7684":6,"python\u7684swig\u63a5\u53e3\u53ef\u4ee5\u65b9\u4fbf\u8fdb\u884c\u9884\u6d4b\u548c\u5b9a\u5236\u5316\u8bad\u7ec3":1,"release\u7684\u72b6\u6001":5,"return":10,"rnn\u914d\u7f6e":11,"server\u8fdb\u7a0b":13,"sh\u662fdocker":0,"shuffle\u8bad\u7ec3\u6570\u636e":20,"softmax\u8f93\u51fa":10,"string\u7684\u683c\u5f0f\u6253\u5370\u51fa\u6765":13,"swig_paddle\u63a5\u53d7\u7684\u539f\u59cb\u6570\u636e\u662fc":23,"train\u5373\u4e3apaddlepaddle\u7684\u8bad\u7ec3\u8fdb\u7a0b":13,"train\u5b8c\u6210\u5355\u673a\u591a\u663e\u5361\u591a\u7ebf\u7a0b\u7684\u8bad":13,"train\u7684\u547d\u4ee4\u884c\u53c2\u6570":13,"ubuntu\u7684deb\u5b89\u88c5\u5305\u7b49":5,"v2\u4e4b\u540e\u7684\u4efb\u4f55\u4e00\u4e2acudnn\u7248\u672c\u6765\u7f16\u8bd1\u8fd0\u884c":1,"value\u5373\u4e3asoftmax\u5c42\u7684\u8f93\u51fa":23,"var":6,"vector\u8868\u793a\u7684\u6bcf\u4e2a\u5355\u8bcd":10,"version\u53ef\u4ee5\u6253\u5370\u51fapaddle\u7684\u7248\u672c\u4fe1\u606f\u548c\u7f16\u8bd1\u7684\u9009\u9879":18,"version\u53ef\u4ee5\u6253\u5370\u51fapaddlepaddle\u7684\u7248\u672c\u548c\u7f16\u8bd1\u65f6\u4fe1\u606f":13,"version\u7684\u547d\u4ee4\u884c\u53c2\u6570":13,"yield\u6587\u672c\u4fe1\u606f\u548c\u7c7b\u522bid":10,__main__:23,__name__:23,abov:20,act:10,act_typ:10,activ:10,adadelta:10,adagrad:10,adam:10,adamoptim:10,afi:20,all:20,allow:10,alreadi:7,also:10,append:20,apt:[6,7],arg:[0,10,20],around:20,arrai:23,assert:23,atla:1,atlas_root:1,avg:10,avgcost:10,avgpool:10,avx:7,bag:10,baidu:[6,7],batch:10,batch_siz:10,binari:10,bla:1,bool:10,both:10,bow:10,build:[0,6],cach:[10,19],cache_pass_in_mem:[10,20],cachetyp:[10,20],calc_batch_s:20,call:10,can:10,can_over_batch_s:20,categori:10,cento:7,check:7,chines:9,chpasswd:6,classif:10,classification_cost:10,classification_error_evalu:10,close:20,cmake:1,cmd:6,cnn:10,code:[0,20,23],com:[6,7],comment:10,compil:18,conf:23,config:[7,10],config_arg:10,config_pars:23,connect:10,contain:[10,20],context_len:10,context_start:10,convert:[10,20,23],couldn:7,cpp:[7,10],cpu:[6,7,20],createfromconfigproto:23,cross:10,cuda_so:6,cudastat:7,cudasuccess:7,cudnn:1,cudnn_root:1,cudnnv5:1,current:[10,20],currentcost:10,currentev:10,dalla:20,data:7,data_config:23,data_initialz:10,data_lay:10,dataprovid:10,dataprovider_bow:10,dataprovider_emb:10,dataproviderconvert:23,dataset:10,deb:7,debian:7,decor:20,def:[10,20,23],defin:[10,20],define_py_data_sources2:[10,20],delar:10,demo:[6,10],dense_vector:[20,23],describ:10,descript:22,detail:22,dev:6,devel:6,devic:6,devices:6,dict:[10,20],dict_fil:10,dictionai:10,dictionari:[10,20],dictrionari:10,differ:10,dim:10,dimens:10,dir:10,doc:23,documentari:20,dpkg:7,driver:7,dso_handl:7,dtype:23,dump_config:13,dure:[10,20],dynam:20,each:[10,20],each_pixel_str:20,each_word:20,echo:6,either:10,els:10,emb:10,embed:9,embedding_lay:10,entropi:10,enumer:10,error:[7,10],etc:6,eval:10,exampl:10,experi:5,expose:6,f0831:7,fail:7,fals:10,fc_layer:10,featur:[10,20],festiv:20,file:[10,20],file_list:20,file_nam:10,filenam:20,fill:10,find:7,first:10,float32:23,fly:10,forwardtest:23,framework:10,from:[6,10,20,23],fulli:10,gener:[10,20],gentoo:7,get:[6,7,10,20],get_config_arg:10,get_data:10,github:7,give:20,given:10,globe:20,gpu:[6,7],gradient_clipping_threshold:10,gradientmachin:23,gru:10,gru_siz:10,help:23,hint:23,hl_cuda_devic:7,hl_dso_load:7,host:6,hot:10,hous:20,http:7,ignor:20,imag:6,imagenet:9,inarg:23,includ:10,init:10,init_hook:[10,19],init_model_path:10,initi:[10,20],initpaddl:23,input:[10,20],input_typ:[10,19],instal:2,insuffici:7,intal:5,integ:[10,20],integer_sequ:20,integer_valu:[10,20],integer_value_sequ:10,invok:20,is_predict:10,is_train:20,isinst:23,job:10,kernel:6,kwarg:[10,20],l2regular:10,label:[10,20],label_dim:10,lake:20,later:10,latest:[0,6],layer:10,ld_library_path:7,learning_method:10,learning_r:10,len:[10,20],lib64:[6,7],lib:1,libcuda:6,libnvidia:6,librari:7,list:[10,19,20],load_data_arg:23,loadparamet:23,local:[1,7],log_period:10,logger:20,look:[10,20],loss:10,lowest_dl_speed:0,lstm:10,lstm_size:10,mac:6,mai:5,main:23,maintainer:6,make:[7,20],make_diagram:13,maxid:10,maxid_lay:10,mean:10,memori:10,merge_model:13,method:20,mkdir:6,mkl:1,mkl_core:1,mkl_root:1,mnist:20,mnist_model:23,mnist_provid:20,mnist_train:20,model_config:23,modul:[10,20],momentum:10,movi:20,must:7,name:[6,10],necessari:10,need:10,neg:[10,20],net:6,no_cache:20,no_sequence:20,none:[10,20,23],note:[5,7],nullptr:7,num:10,num_pass:10,nvidia:6,obj:[10,20],object:[10,20],off:[0,1,18],omit:10,on_init:20,onli:10,open:[10,20],openbla:1,openblas_root:1,openssh:6,opt:1,other:10,outlin:22,output:10,outsid:20,packag:5,paddl:[0,6,7,10,13],paddle_gpu:0,paddle_ssh:6,paddle_ssh_machin:6,paddledev:6,paddlepaddl:[6,7,18,23],paramet:10,parse_config:23,pass:[10,20],path:[7,10],period:10,permitrootlogin:6,pixel:20,pixels_float:20,pixels_str:20,place:20,pleas:7,pool_siz:20,pooling_lay:10,pooling_typ:10,posit:[10,20],pre:5,pred:10,predict_output_dir:10,predict_sampl:23,preprocess:10,print:23,process:[10,20],process_pr:10,properli:10,pull:6,put:10,py_paddl:[6,23],pydataprovid:19,pydataprovider2:[10,20,23],pydataproviderwrapp:10,python:10,quick_start:10,rang:10,rank:10,rare:20,read:[10,20],refer:19,reference_cblas_root:1,reffer:1,regular:10,releas:[5,7],repres:10,represent:10,resnet:9,result:10,rmsprop:10,roce:6,root:6,run:6,runtim:[7,20],same:[10,20],sampl:[10,20],save:[10,20],save_dir:10,saw:20,sbin:6,script:0,second:10,sed:6,see:10,sentenc:20,sentiment:20,sentimental_provid:20,separ:10,seq_typ:20,sequel:20,sequence:20,sequence_conv_pool:10,sequencetyp:20,server:6,set:[10,20],setup:10,should_shuffl:20,simple_gru:10,simple_lstm:10,size:[10,20],smooth:5,softmax:10,softmaxactiv:10,sourc:10,spars:10,sparse_binary_vector:[10,20],sparse_float_vector:20,specifi:[7,10],split:[10,20],src_root:23,ssh:6,sshd:6,sshd_config:6,stat:10,state:5,still:5,stop:6,store:10,string:20,strip:10,structur:10,stun:20,sub_sequence:20,sure:7,swig_paddl:23,tag:[0,7],take:20,tbd:21,team:6,test:[10,19],test_data:23,test_list:[10,20],text:[10,20],text_conv:10,them:10,thi:[10,20],thing:20,train:7,train_list:[10,20],trainer:[10,20,23],trainer_config:[10,19,20,23],trainer_config_help:[10,20],trainerintern:10,trainermain:7,travel:20,trn:10,tst:10,two:10,txt:[10,20],type:[10,20],ubuntu:7,unk_idx:10,updat:6,use:[10,22],use_gpu:[10,23],usepam:6,user:10,usr:[1,6,7],valid:7,valu:[10,20,23],version:[6,7],via:7,want:20,what:10,when:20,which:10,whole:20,wilder:20,window:6,with_avx:[1,18],with_doc:1,with_doc_cn:1,with_doubl:18,with_double:1,with_dso:1,with_gflag:18,with_gflags:1,with_glog:[1,18],with_gpu:[0,1,18],with_metric_learn:18,with_predict_sdk:18,with_python:[1,18],with_rdma:[1,18],with_style_check:1,with_swig_py:1,with_testing:1,with_tim:18,with_timer:1,wonder:20,word:9,word_dict:10,word_dim:10,word_id:20,word_vector:10,xarg:6,yield:[10,20],you:[7,20],your:5,your_host_machine:6},titles:["\u6784\u5efaPaddlePaddle Docker Image","\u8bbe\u7f6ePaddlePaddle\u7684\u7f16\u8bd1\u9009\u9879","\u4f7f\u7528cmake\u7f16\u8bd1PaddlePaddle","\u5b89\u88c5\u7f16\u8bd1PaddlePaddle\u9700\u8981\u7684\u4f9d\u8d56","make\u548cmake install","\u7f16\u8bd1\u4e0e\u5b89\u88c5","\u5b89\u88c5PaddlePaddle\u7684Docker\u955c\u50cf","\u4f7f\u7528deb\u5305\u5728Ubuntu\u4e0a\u5b89\u88c5PaddlePaddle","\u96c6\u7fa4\u8bad\u7ec3","\u4f7f\u7528\u793a\u4f8b","PaddlePaddle\u5feb\u901f\u5165\u95e8\u6559\u7a0b","PaddlePaddle\u6587\u6863","<no title>","\u547d\u4ee4\u884c\u53c2\u6570","<no title>","<no title>","paddle pserver\u7684\u547d\u4ee4\u884c\u53c2\u6570","paddle train\u7684\u547d\u4ee4\u884c\u53c2\u6570","paddle version\u7684\u547d\u4ee4\u884c\u53c2\u6570","PaddlePaddle\u7684\u6570\u636e\u63d0\u4f9b(DataProvider)\u4ecb\u7ecd","PyDataProvider2\u7684\u4f7f\u7528","\u81ea\u5b9a\u4e49\u4e00\u4e2aDataProvider","\u7528\u6237\u63a5\u53e3","PaddlePaddle\u7684Python\u9884\u6d4b\u63a5\u53e3"],titleterms:{"\u4e0b\u8f7d\u548c\u8fd0\u884cdocker\u955c\u50cf":6,"\u4ecb\u7ecd":19,"\u4f18\u5316\u7b97\u6cd5":10,"\u4f7f\u7528\u6307\u5357":11,"\u4f7f\u7528\u6982\u8ff0":10,"\u4f7f\u7528\u793a\u4f8b":9,"\u4f7f\u7528\u811a\u672c\u6784\u5efapaddlepaddl":0,"\u4f7f\u7528cmake\u7f16\u8bd1paddlepaddl":2,"\u4f7f\u7528deb\u5305\u5728ubuntu\u4e0a\u5b89\u88c5paddlepaddl":7,"\u5377\u79ef\u6a21\u578b":10,"\u53c2\u8003":20,"\u53ef\u80fd\u9047\u5230\u7684\u95ee\u9898":7,"\u547d\u4ee4\u884c\u53c2\u6570":[10,13,22],"\u56fe\u50cf":9,"\u5b89\u88c5":10,"\u5b89\u88c5\u7f16\u8bd1paddlepaddle\u9700\u8981\u7684\u4f9d\u8d56":3,"\u5b89\u88c5paddlepaddle\u7684docker\u955c\u50cf":6,"\u5e38\u7528\u6a21\u578b":9,"\u5e8f\u5217\u6a21\u578b\u6570\u636e\u63d0\u4f9b":20,"\u5f00\u53d1\u6307\u5357":11,"\u6027\u80fd\u95ee\u9898":6,"\u603b\u4f53\u6548\u679c\u603b\u7ed3":10,"\u63a8\u8350":9,"\u6570\u636e\u5411\u6a21\u578b\u4f20\u9001":10,"\u6570\u636e\u63d0\u4f9b":22,"\u6570\u636e\u683c\u5f0f\u51c6\u5907":10,"\u65f6\u5e8f\u6a21\u578b":10,"\u6784\u5efapaddlepaddl":0,"\u6ce8\u610f\u4e8b\u9879":6,"\u7528\u6237\u63a5\u53e3":22,"\u7b80\u5355\u7684\u4f7f\u7528\u573a\u666f":20,"\u7b97\u6cd5\u6559\u7a0b":11,"\u7f16\u8bd1\u4e0e\u5b89\u88c5":5,"\u7f51\u7edc\u7ed3\u6784":10,"\u81ea\u5b9a\u4e49\u4e00\u4e2adataprovid":21,"\u81ea\u7136\u8bed\u8a00\u5904\u7406":9,"\u8bad\u7ec3\u6a21\u578b":10,"\u8bbe\u7f6epaddlepaddle\u7684\u7f16\u8bd1\u9009\u9879":1,"\u8bcd\u5411\u91cf\u6a21\u578b":10,"\u8f93\u51fa\u65e5\u5fd7":10,"\u8fdc\u7a0b\u8bbf\u95ee\u95ee\u9898\u548c\u4e8c\u6b21\u5f00\u53d1":6,"\u903b\u8f91\u56de\u5f52\u6a21\u578b":10,"\u914d\u7f6e\u4e2d\u7684\u6570\u636e\u52a0\u8f7d\u5b9a\u4e49":10,"\u9644\u5f55":10,"\u96c6\u7fa4\u8bad\u7ec3":8,"\u9884\u6d4b":[10,22],"blas\u76f8\u5173\u7684\u7f16\u8bd1\u9009\u9879":1,"bool\u578b\u7684\u7f16\u8bd1\u9009\u9879":1,"config\u6587\u4ef6\u627e\u4e0d\u5230":7,"cudnn\u76f8\u5173\u7684\u7f16\u8bd1\u9009\u9879":1,"driver\u627e\u4e0d\u5230":7,"make\u548cmak":4,"paddlepaddle\u5feb\u901f\u5165\u95e8\u6559\u7a0b":10,"paddlepaddle\u63d0\u4f9b\u7684docker\u955c\u50cf\u7248\u672c":6,"paddlepaddle\u6587\u6863":11,"paddlepaddle\u7684\u6570\u636e\u63d0\u4f9b":19,"paddlepaddle\u7684\u7f16\u8bd1\u9009\u9879":1,"paddlepaddle\u7684bool\u578b\u7f16\u8bd1\u9009\u9879":1,"paddlepaddle\u7684cblas\u7f16\u8bd1\u9009\u9879":1,"paddlepaddle\u7684python\u9884\u6d4b\u63a5\u53e3":23,"pserver\u7684\u547d\u4ee4\u884c\u53c2\u6570":16,"pydataprovider2\u7684\u4f7f\u7528":20,"python\u6570\u636e\u52a0\u8f7d\u811a\u672c":10,"so\u627e\u4e0d\u5230":7,"train\u7684\u547d\u4ee4\u884c\u53c2\u6570":17,"version\u7684\u547d\u4ee4\u884c\u53c2\u6570":18,algorithm:10,appendix:10,architectur:10,argument:10,cach:20,command:10,configur:10,convolut:10,cuda:[1,7],data:10,dataprovid:19,docker:0,image:0,init_hook:20,input_typ:20,instal:4,install:10,libcudart:7,libcudnn:7,line:10,log:10,logist:10,model:10,network:10,optimiz:10,overview:10,paddl:[16,17,18],predict:10,prepar:10,provid:[10,20],refer:20,regress:10,script:10,sequenc:10,summari:10,time:10,train:10,transfer:10,vector:10,word:10}})
\ No newline at end of file
diff --git a/doc_cn/ui/cmd/dump_config.html b/doc_cn/ui/cmd/dump_config.html
index 151eec68af..d5fc7bf730 100644
--- a/doc_cn/ui/cmd/dump_config.html
+++ b/doc_cn/ui/cmd/dump_config.html
@@ -6,7 +6,7 @@
   
     
     
-    <no title> — PaddlePaddle  documentation
+    <no title> — PaddlePaddle  documentation
     
     
     
@@ -33,7 +33,7 @@
         
  • 
           index
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
      
 
@@ -59,14 +59,11 @@
 
 
         
    
@@ -79,12 +76,12 @@
         
  • 
           index
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc_cn/ui/cmd/index.html b/doc_cn/ui/cmd/index.html
index 0a974571ff..b59961c481 100644
--- a/doc_cn/ui/cmd/index.html
+++ b/doc_cn/ui/cmd/index.html
@@ -6,7 +6,7 @@
   
     
     
-    命令行参数 — PaddlePaddle  documentation
+    命令行参数 — PaddlePaddle  documentation
     
     
     
@@ -33,7 +33,7 @@
         
  • 
           index
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
      
 
@@ -80,14 +80,11 @@
 
 
         
    
@@ -100,12 +97,12 @@
         
  • 
           index
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc_cn/ui/cmd/make_diagram.html b/doc_cn/ui/cmd/make_diagram.html
index af2c1eb0bc..0b242873e3 100644
--- a/doc_cn/ui/cmd/make_diagram.html
+++ b/doc_cn/ui/cmd/make_diagram.html
@@ -6,7 +6,7 @@
   
     
     
-    <no title> — PaddlePaddle  documentation
+    <no title> — PaddlePaddle  documentation
     
     
     
@@ -33,7 +33,7 @@
         
  • 
           index
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
      
 
@@ -59,14 +59,11 @@
 
 
         
    
@@ -79,12 +76,12 @@
         
  • 
           index
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc_cn/ui/cmd/merge_model.html b/doc_cn/ui/cmd/merge_model.html
index 5c3a0c26a2..36b8e5cd35 100644
--- a/doc_cn/ui/cmd/merge_model.html
+++ b/doc_cn/ui/cmd/merge_model.html
@@ -6,7 +6,7 @@
   
     
     
-    <no title> — PaddlePaddle  documentation
+    <no title> — PaddlePaddle  documentation
     
     
     
@@ -33,7 +33,7 @@
         
  • 
           index
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
      
 
@@ -59,14 +59,11 @@
 
 
         
    
@@ -79,12 +76,12 @@
         
  • 
           index
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc_cn/ui/cmd/paddle_pserver.html b/doc_cn/ui/cmd/paddle_pserver.html
index 7c60ea221f..8a18540a04 100644
--- a/doc_cn/ui/cmd/paddle_pserver.html
+++ b/doc_cn/ui/cmd/paddle_pserver.html
@@ -6,7 +6,7 @@
   
     
     
-    paddle pserver的命令行参数 — PaddlePaddle  documentation
+    paddle pserver的命令行参数 — PaddlePaddle  documentation
     
     
     
@@ -33,7 +33,7 @@
         
  • 
           index
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
      
 
@@ -62,14 +62,11 @@
 
 
         
    
@@ -82,12 +79,12 @@
         
  • 
           index
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc_cn/ui/cmd/paddle_train.html b/doc_cn/ui/cmd/paddle_train.html
index dccb900cf7..cc01273522 100644
--- a/doc_cn/ui/cmd/paddle_train.html
+++ b/doc_cn/ui/cmd/paddle_train.html
@@ -6,7 +6,7 @@
   
     
     
-    paddle train的命令行参数 — PaddlePaddle  documentation
+    paddle train的命令行参数 — PaddlePaddle  documentation
     
     
     
@@ -33,7 +33,7 @@
         
  • 
           index
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
      
 
@@ -62,14 +62,11 @@
 
 
         
    
@@ -82,12 +79,12 @@
         
  • 
           index
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc_cn/ui/cmd/paddle_version.html b/doc_cn/ui/cmd/paddle_version.html
index b73a319336..5b6e1f7dab 100644
--- a/doc_cn/ui/cmd/paddle_version.html
+++ b/doc_cn/ui/cmd/paddle_version.html
@@ -6,7 +6,7 @@
   
     
     
-    paddle version的命令行参数 — PaddlePaddle  documentation
+    paddle version的命令行参数 — PaddlePaddle  documentation
     
     
     
@@ -33,7 +33,7 @@
         
  • 
           index
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
      
 
@@ -45,17 +45,17 @@
   
    
 
    paddle version的命令行参数
    
 
    paddle version可以打印出paddle的版本信息和编译的选项。常见的输出格式为
    
-
    PaddlePaddle 0.8.0b, compiled with
-    with_avx: ON
-    with_gpu: ON
-    with_double: OFF
-    with_python: ON
-    with_rdma: OFF
-    with_glog: ON
-    with_gflags: ON
-    with_metric_learning: OFF
-    with_timer: OFF
-    with_predict_sdk: OFF
+
    PaddlePaddle 0.8.0b, compiled with
+    with_avx: ON
+    with_gpu: ON
+    with_double: OFF
+    with_python: ON
+    with_rdma: OFF
+    with_glog: ON
+    with_gflags: ON
+    with_metric_learning: OFF
+    with_timer: OFF
+    with_predict_sdk: OFF
 
    
 
    
 
    其第一行说明了paddle的版本，后面跟着一系列编译参数。这里可以参考paddle的
@@ -78,14 +78,11 @@
 
    
 
         
    
@@ -98,12 +95,12 @@
         
  • 
           index
    • 
-        
  • PaddlePaddle  documentation »
    •  
+        
  • PaddlePaddle  documentation »
    •  
       
     
    
     
   
 
\ No newline at end of file
diff --git a/doc_cn/ui/data_provider/index.html b/doc_cn/ui/data_provider/index.html
index 88f98f1426..e0fd76ee85 100644
--- a/doc_cn/ui/data_provider/index.html
+++ b/doc_cn/ui/data_provider/index.html
@@ -6,7 +6,7 @@
   
     
     
-    PaddlePaddle的数据提供(DataProvider)介绍 — PaddlePaddle  documentation
+    PaddlePaddle的数据提供(DataProvider)介绍 — PaddlePaddle  documentation
     
     
     
@@ -24,10 +24,7 @@
     
     
     
-    
-    
-    
-     
+     
   
   
       
 
@@ -54,20 +44,11 @@
             
   
    
 
    PaddlePaddle的数据提供(DataProvider)介绍
    
-
    数据提供(DataProvider，后用DataProvider代替)是PaddlePaddle负责提供数据的模块。其作用是将训练数据
-传入内存或者显存，让神经网络可以进行训练。简单的使用，用户可以使用Python的
-PyDataProvider 来自定义传数据的过程。如果有更复杂的使用，或者需要更高的效率，
-用户也可以在C++端自定义一个 DataProvider 。
    
-
    PaddlePaddle需要用户在网络配置(trainer_config.py)中定义使用什么DataProvider，和DataProvider
-的一些参数，训练文件列表(train.list)和测试文件列表(test.list)。
    
-
    其中，train.list和test.list均为本地的两个文件(推荐直接放置到训练目录，以相对路径引用)。如果
-test.list不设置，或者设置为None的话，那么在训练过程中，不会执行测试操作。否则，则会根据命令行
-参数指定的测试方式，在训练过程中进行测试，从而防止过拟合。
    
-
    一般情况下，train.list和test.list为纯文本文件，其每一行对应这每一个数据文件。数据文件存放在
-本地磁盘中，将文件的绝对路径或相对路径(相对于PaddlePaddle程序运行时的路径)的方式写在train.list和
-test.list中。当然，train.list和test.list也可以放置hdfs文件路径，或者数据库连接地址等等。
-用户在DataProvider中需要实现如何访问其中每一个文件。
    
-
    DataProvider的具体用法和如何实现一个新的DataProvider，请参考下述文章:
    
+
    数据提供(DataProvider)是PaddlePaddle负责提供数据的模块。其作用是将训练数据传入内存或者显存，让神经网络可以进行训练。简单的使用，用户可以使用Python的 PyDataProvider 来自定义传数据的过程。如果有更复杂的使用，或者需要更高的效率，用户也可以在C++端自定义一个 DataProvider 。
    
+
    PaddlePaddle需要用户在网络配置(trainer_config.py)中定义使用哪种DataProvider及其参数，训练文件列表(train.list)和测试文件列表(test.list)。
    
+
    其中，train.list和test.list均为本地的两个文件(推荐直接放置到训练目录，以相对路径引用)。如果test.list不设置，或者设置为None，那么在训练过程中，不会执行测试操作。否则，会根据命令行参数指定的测试方式，在训练过程中进行测试，从而防止过拟合。
    
+
    一般情况下，train.list和test.list为纯文本文件，一行对应一个数据文件，数据文件存放在本地磁盘中。将文件的绝对路径或相对路径(相对于PaddlePaddle程序运行时的路径)写在train.list和test.list中。当然，train.list和test.list也可以放置hdfs文件路径，或者数据库连接地址等等。
    
+
    用户在DataProvider中需要实现如何访问其中每一个文件。DataProvider的具体用法和如何实现一个新的DataProvider，请参考下述文章:
    
 
    
 
      
 
    • PyDataProvider2的使用
        
@@ -93,12 +74,6 @@ test.list中。当然，train.list和test.list也可以放置hdfs文件路径，