{"id":327,"date":"2014-08-22T22:26:47","date_gmt":"2014-08-22T13:26:47","guid":{"rendered":"https:\/\/www.dogrow.net\/nnet\/?p=327"},"modified":"2019-10-19T04:39:24","modified_gmt":"2019-10-18T19:39:24","slug":"blog22","status":"publish","type":"post","link":"https:\/\/www.dogrow.net\/nnet\/blog22\/","title":{"rendered":"(22) cuda-convnet2\u306f\u3084\u3063\u3066\u307f\u308c\u306a\u3044 (T_T)"},"content":{"rendered":"

2014\u5e747\u670817\u65e5\u3001cuda-convnet <\/span>\u306e\u4f5c\u8005\u304c cuda-convnet2 <\/span>\u306a\u308b\u30d7\u30ed\u30b0\u30e9\u30e0\u3092GIT\u4e0a\u306b\u516c\u958b\u3057\u305f\u3002
\n\u73fe\u5728\u3082\u958b\u767a\u304c\u9032\u884c\u4e2d\u3067\u3001\u305d\u306e\u69d8\u5b50\u306f\u3053\u3061\u3089\uff08\u2193\uff09\u3067\u773a\u3081\u3089\u308c\u308b\u3002
\nhttps:\/\/github.com\/akrizhevsky\/cuda-convnet2
\n\"20140822_01\"<\/a><\/p>\n

\u3053\u308c\u306f\u305c\u3072\u3084\u3063\u3066\u307f\u3088\u3046\uff01<\/span><\/p>\n

\u3068\u610f\u6c17\u63da\u3005\u3068\u30d7\u30ed\u30b0\u30e9\u30e0\u3092\u30c0\u30a6\u30f3\u30ed\u30fc\u30c9\u3057\u3001
\n  \u5fc5\u8981\u306a\u30e9\u30a4\u30d6\u30e9\u30ea\u3092\u30a4\u30f3\u30b9\u30c8\u30fc\u30eb\u3057\u3001
\n    \u30d3\u30eb\u30c9\u304c\u6210\u529f\u3057\u3001<\/p>\n

\u3044\u3056\u5b9f\u884c\uff01<\/span><\/p>\n

=========================\r\nRunning on CUDA device(s) 0\r\nCurrent time: Fri Aug 22 21:51:07 2014\r\nSaving checkpoints to \/home\/user\/cuda\/cuda-convnet2\/save\/cifar-10\/ConvNet__2014-08-22_21.51.06\r\n=========================\r\nsrc\/nvmatrix.cu(394) : getLastCudaError() CUDA error : kSetupCurand: Kernel execution failed : (8) invalid device function .\r\n<\/pre>\n
\u30a8\u30e9\u30fc\u3067\u6b62\u307e\u3063\u305f (TT)<\/span><\/p>\n
\u3044\u308d\u3044\u308d\u8abf\u3079\u308b\u3068\u3001\u30b3\u30f3\u30d1\u30a4\u30eb\u6642\u306b\u6307\u5b9a\u3057\u305f Compte Capability version<\/span> \u304c\u3001
\n\u30de\u30b7\u30f3\u306b\u642d\u8f09\u3057\u3066\u3044\u308b GPU\u3068\u5408\u3063\u3066\u3044\u306a\u3044\u3053\u3068\u304c\u539f\u56e0\u306e\u3088\u3046\u3060\u3002<\/p>\n
\u305d\u3053\u3067…
\n\u3046\u3061\u306e\u65b0\u3057\u3044 GTX-760 <\/span>\u306e Compute Capability\u3092\u8abf\u3079\u3066\u307f\u308b\u3002
\n\u3053\u308c\u306b\u306f NVIDIA_CUDA-6.0_Samples<\/span> \u306b\u4ed8\u5c5e\u306e deviceQuery<\/span> \u30b3\u30de\u30f3\u30c9\u304c\u4f7f\u3048\u308b\u3002<\/p>\n
[user@]$ .\/deviceQuery\r\n.\/deviceQuery Starting...\r\n\r\n CUDA Device Query (Runtime API) version (CUDART static linking)\r\n\r\nDetected 1 CUDA Capable device(s)\r\n\r\nDevice 0: \"GeForce GTX 760\"\r\n  CUDA Driver Version \/ Runtime Version          6.0 \/ 6.0\r\n  CUDA Capability Major\/Minor version number:    3.0<\/span>\r\n  Total amount of global memory:                 2047 MBytes (2146762752 bytes)\r\n  ( 6) Multiprocessors, (192) CUDA Cores\/MP:     1152 CUDA Cores\r\n  GPU Clock rate:                                1058 MHz (1.06 GHz)\r\n  Memory Clock rate:                             3004 Mhz\r\n  Memory Bus Width:                              256-bit\r\n  L2 Cache Size:                                 524288 bytes\r\n :\r\n<\/pre>\n
\u3046\u3061\u306e\u306f Compute Capability version 3.0<\/span> \u306a\u306e\u3060\u305d\u3046\u3060\u3002<\/p>\n
\u3067\u306f\u3001cuda-convnet2 \u3092 Compute Capability ver.3.0 \u6307\u5b9a\u3067\u30d3\u30eb\u30c9\u3057\u3066\u307f\u308b\u3002
\nMakeFile\u4e2d\u3067\u4ee5\u4e0b\u306e\u3088\u3046\u306b\u66f8\u304b\u308c\u3066\u3044\u308b\u3068\u3053\u308d\u3092\u63a2\u3057\u3001\u3059\u3079\u30663.0\u6307\u5b9a\u306b\u66f8\u304d\u63db\u3048\u305f\u3002<\/p>\n
GENCODE_SM35    := -gencode arch=compute_35,code=sm_35\r\nGENCODE_FLAGS   := $(GENCODE_SM35)\r\n<\/pre>\n
\u518d\u5ea6\u30d3\u30eb\u30c9\u3092\u5b9f\u884c\u3059\u308b\u3002<\/p>\n
[user@]$ .\/build.sh clean\r\n[user@]$ .\/build.sh\r\n<\/pre>\n
\u4eca\u5ea6\u3053\u305d\u306f\uff01
\n\u3068 cuda-convnet2 \u3067 Cifar-10 \u81ea\u52d5\u8a8d\u8b58\u3092\u8d77\u52d5\uff01<\/p>\n
\u3059\u308b\u3068…<\/p>\n
\u3055\u3063\u304d\u3088\u308a\u3082\u30e4\u30d0\u30b2\u306a\u30a8\u30e9\u30fc\u304c\u51fa\u305f (T^T)<\/span><\/p>\n
=========================\r\nRunning on CUDA device(s) 0\r\nCurrent time: Fri Aug 22 22:00:59 2014\r\nSaving checkpoints to \/home\/user\/cuda\/cuda-convnet2\/save\/cifar-10\/ConvNet__2014-08-22_22.00.58\r\n=========================\r\n1.1 (0.00%)...python: src\/nvmatrix.cu:1473: virtual cudaTextureObject_t NVMatrix::getTextureObject(): Assertion `_texObj != 0' failed.\r\nError signal 6:\r\n\/home\/user\/cuda\/cuda-convnet2\/cuda-convnet2\/cudaconvnet\/_ConvNet.so(_Z13signalHandleri+0x26)[0x7fb5b9f2a3e6]\r\n\/lib64\/libc.so.6[0x35e22329a0]\r\n\/lib64\/libc.so.6(gsignal+0x35)[0x35e2232925]\r\n\/lib64\/libc.so.6(abort+0x175)[0x35e2234105]\r\n\/lib64\/libc.so.6[0x35e222ba4e]\r\n\/lib64\/libc.so.6(__assert_perror_fail+0x0)[0x35e222bb10]\r\n.\/nvmatrix\/libnvmatrix.so(_ZN8NVMatrix16getTextureObjectEv+0x147)[0x7fb5b8c15f67]\r\n.\/cudaconv3\/libcudaconv.so(_Z11_filterActsR8NVMatrixS0_S0_iiiiiiiffb+0x5c25)[0x7fb5b7f984e5]\r\n.\/cudaconv3\/libcudaconv.so(_Z14convFilterActsR8NVMatrixS0_S0_iiiiiiiff+0x30)[0x7fb5b7f9dd30]\r\n\/home\/user\/cuda\/cuda-convnet2\/cuda-convnet2\/cudaconvnet\/_ConvNet.so(_ZN9ConvLayer9fpropActsEifji+0x18e)[0x7fb5b9ee5d8e]\r\n\/home\/user\/cuda\/cuda-convnet2\/cuda-convnet2\/cudaconvnet\/_ConvNet.so(_ZN5Layer5fpropERSt3mapIiP8NVMatrixSt4lessIiESaISt4pairIKiS2_EEEji+0x240)[0x7fb5b9ef2ab0]\r\n\/home\/user\/cuda\/cuda-convnet2\/cuda-convnet2\/cudaconvnet\/_ConvNet.so(_ZN5Layer5fpropEji+0x23f)[0x7fb5b9ef2d4f]\r\n\/home\/user\/cuda\/cuda-convnet2\/cuda-convnet2\/cudaconvnet\/_ConvNet.so(_ZN13ConvNetThread3runEv+0x17d)[0x7fb5b9f44d6d]\r\n\/home\/user\/cuda\/cuda-convnet2\/cuda-convnet2\/cudaconvnet\/_ConvNet.so(_ZN6Thread18start_pthread_funcEPv+0x9)[0x7fb5b9ef7b69]\r\n\/lib64\/libpthread.so.0[0x35e26079d1]\r\n\/lib64\/libc.so.6(clone+0x6d)[0x35e22e8b5d]\r\nCUDA error at src\/nvmatrix.cu:522 code=29(cudaErrorCudartUnloading) \"cudaSetDevice(d)\"\r\nError signal 11:\r\n\/home\/user\/cuda\/cuda-convnet2\/cuda-convnet2\/cudaconvnet\/_ConvNet.so(_Z13signalHandleri+0x26)[0x7fb5b9f2a3e6]\r\n\/lib64\/libc.so.6[0x35e22329a0]\r\n\/lib64\/libc.so.6(exit+0x35)[0x35e2235d75]\r\n.\/nvmatrix\/libnvmatrix.so(+0x30886)[0x7fb5b8c15886]\r\n\/home\/user\/cuda\/cuda-convnet2\/cuda-convnet2\/cudaconvnet\/_ConvNet.so(_ZN14DataCopyThread3runEv+0x4ac)[0x7fb5b9ee3b9c]\r\n\/home\/user\/cuda\/cuda-convnet2\/cuda-convnet2\/cudaconvnet\/_ConvNet.so(_ZN6Thread18start_pthread_funcEPv+0x9)[0x7fb5b9ef7b69]\r\n\/lib64\/libpthread.so.0[0x35e26079d1]\r\n\/lib64\/libc.so.6(clone+0x6d)[0x35e22e8b5d]\r\n<\/pre>\n
\u30a8\u30e9\u30fc\u306e\u5185\u5bb9\u3092\u898b\u308b\u3068\u3001\u3069\u3046\u3084\u3089\u3046\u3061\u306eGTX760\u3067\u306f\u672a\u30b5\u30dd\u30fc\u30c8\u306e\u6a5f\u80fd\u3092\u4f7f\u304a\u3046\u3068\u3057\u3066\u3044\u308b\u69d8\u5b50…<\/p>\n
\u305d\u3053\u3067 cuda-convnet2 \u306e\u30db\u30fc\u30e0\u30da\u30fc\u30b8\u3092\u6539\u3081\u3066\u898b\u3066\u307f\u308b\u3068…
\n\"20140822_02\"<\/a><\/p>\n
Compute Capability 3.5\u4ee5\u4e0a\u3058\u3083\u306a\u3044\u3068\u30c0\u30e1<\/span><\/p>\n
\u3068\u66f8\u3044\u3066\u3042\u3063\u305f…<\/p>\n
2014\u5e748\u6708\u73fe\u5728\u3001GTX780\u3092\u624b\u306b\u5165\u308c\u308b\u306b\u306f 6\u4e07\u5186<\/span>\u524d\u5f8c\u306e\u304a\u91d1\u304c\u5fc5\u8981\u3060…
\n\u5148\u65e5Xeon x5570 x2 \u30de\u30b7\u30f3\u3092 6\u4e07\u5186\u3067\u8cfc\u5165\u3057\u305f\u306e\u3067\u3001\u624b\u5143\u306b\u304a\u91d1\u306f\u6b8b\u3063\u3066\u3044\u306a\u3044…<\/p>\n
cuda-convnet2<\/span> \u306f\u30e4\u30d5\u30aa\u30af\u3067 GTX780\u304c2\u4e07\u5186\u53f0\u306b\u5024\u4e0b\u304c\u308a\u3059\u308b\u307e\u3067\u3057\u3070\u3089\u304f\u3042\u304d\u3089\u3081\u3088\u3046(T^T)<\/p>\n
\u53c2\u8003\u60c5\u5831<\/h1>\n
CUDA ZONE\u306e\u30da\u30fc\u30b8\u306b GPU\u578b\u756a\u5225\u306e Compute Capavility Version\u4e00\u89a7\u304c\u8f09\u3063\u3066\u3044\u308b\u3002
\nhttps:\/\/developer.nvidia.com\/cuda-gpus
\n\"20140822_03\"<\/a><\/p>\n
\n","protected":false},"excerpt":{"rendered":"
2014\u5e747\u670817\u65e5\u3001cuda-convnet \u306e\u4f5c\u8005\u304c cuda-convnet2 \u306a\u308b\u30d7\u30ed\u30b0\u30e9\u30e0\u3092GIT\u4e0a\u306b\u516c\u958b\u3057\u305f\u3002 \u73fe\u5728\u3082\u958b\u767a\u304c\u9032\u884c\u4e2d\u3067\u3001\u305d\u306e\u69d8\u5b50\u306f\u3053\u3061\u3089\uff08\u2193\uff09\u3067\u773a\u3081\u3089\u308c\u308b\u3002 https:\/\/github.com\u2026 \u7d9a\u304d\u3092\u8aad\u3080 »<\/a><\/span><\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[8,10],"tags":[],"class_list":["post-327","post","type-post","status-publish","format-standard","hentry","category-cuda","category-cuda-convnet"],"views":3827,"amp_enabled":true,"_links":{"self":[{"href":"https:\/\/www.dogrow.net\/nnet\/wp-json\/wp\/v2\/posts\/327","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.dogrow.net\/nnet\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.dogrow.net\/nnet\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.dogrow.net\/nnet\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.dogrow.net\/nnet\/wp-json\/wp\/v2\/comments?post=327"}],"version-history":[{"count":47,"href":"https:\/\/www.dogrow.net\/nnet\/wp-json\/wp\/v2\/posts\/327\/revisions"}],"predecessor-version":[{"id":767,"href":"https:\/\/www.dogrow.net\/nnet\/wp-json\/wp\/v2\/posts\/327\/revisions\/767"}],"wp:attachment":[{"href":"https:\/\/www.dogrow.net\/nnet\/wp-json\/wp\/v2\/media?parent=327"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.dogrow.net\/nnet\/wp-json\/wp\/v2\/categories?post=327"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.dogrow.net\/nnet\/wp-json\/wp\/v2\/tags?post=327"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}