{"id":46643,"date":"2023-10-29T08:40:22","date_gmt":"2023-01-28T07:45:08","guid":{"rendered":"https:\/\/www.silicloud.com\/zh\/blog\/46643-2\/"},"modified":"2024-05-04T06:31:34","modified_gmt":"2024-05-03T22:31:34","slug":"46643-2","status":"publish","type":"post","link":"https:\/\/www.silicloud.com\/zh\/blog\/46643-2\/","title":{"rendered":""},"content":{"rendered":"

\u306f\u3058\u3081\u306b<\/h1>\n

Google colaboratory\u304b\u3089ABCL\u306b\u74b0\u5883\u3092\u5909\u3048\u305f\u304c\uff0cABCL\u306e\u74b0\u5883\u69cb\u7bc9\u3068Jupyter notebook\u4e0a\u3067GPU\u3092\u4f7f\u3046\u65b9\u6cd5\u304c\u5206\u304b\u308a\u306b\u304f\u304b\u3063\u305f\u306e\u3067\u3053\u3053\u306b\u30e1\u30e2\u3057\u3066\u304a\u304f<\/p>\n

\u3053\u3093\u306a\u4eba\u306b\u898b\u3066\u307b\u3057\u3044 :
\nABCL\u3092\u4f7f\u3046\u969b\u306b
\n\u30fbABCL\u306e\u958b\u59cb\u65b9\u6cd5\u304c\u308f\u304b\u3089\u306a\u3044\/\u74b0\u5883\u69cb\u7bc9\u306e\u65b9\u6cd5\u304c\u308f\u304b\u3089\u306a\u3044<\/p>\n

Jupyter notebook\u3067TensorFlow\u3092\u4f7f\u3046\u969b\u306b\uff0c
\n\u30fbGPU\u3092\u4f7f\u3063\u3066\u5b66\u7fd2\u3092\u3057\u305f\u3044
\n\u30fbGPU\u304c\u8a8d\u8b58\u3055\u308c\u306a\u3044
\n\u30fbGPU\u3092\u4f7f\u3063\u3066\u308b\u306f\u305a\u306a\u306e\u306b\u9045\u3044
\n\u30fbGPU\u3092\u4f7f\u3063\u305f\u969b\u306b\u6bce\u56de\u5b66\u7fd2\u7d50\u679c\u304c\u7570\u306a\u308b<\/p>\n

\u76ee\u6b21<\/h1>\n
    \n
  1. \n
      ABCL\u306e\u5229\u7528\u65b9\u6cd5<\/ol>\n<\/li>\n<\/ol>\n

       <\/p>\n

        \n
      1. \n
          ABCI\u3067Jupyter Notebook\u3092\u4f7f\u7528\u3059\u308b<\/ol>\n<\/li>\n<\/ol>\n

           <\/p>\n

            \n
          1. \n
              GPU\u306e\u518d\u73fe\u6027\u3092\u78ba\u4fdd\u3059\u308b<\/ol>\n<\/li>\n<\/ol>\n

               <\/p>\n

                GPU\u3092\u8907\u6570\u4f7f\u3063\u3066\u4e26\u5217\u51e6\u7406\u3092\u3059\u308b<\/ol>\n

                1. ABCL\u306e\u5229\u7528\u65b9\u6cd5<\/h1>\n

                ABCI\u5229\u2f64\u30dd\u30fc\u30bf\u30eb\u3078\u306e\u521d\u56de\u30ed\u30b0\u30a4\u30f3\u3068\u30d1\u30b9\u30ef\u30fc\u30c9\u767b\u9332, \u516c\u958b\u9375\u306e\u767b\u9332\u306f[ABCI\u306e\u30de\u30cb\u30e5\u30a2\u30eb\u3011(https:\/\/abci.ai\/ja\/how_to_use\/data\/ABCI_handson_2020.pdf)\u3092\u898b\u308c\u3070\u3067\u304d\u308b\u306e\u3067\u98db\u3070\u3059.<\/p>\n

                \u5229\u7528\u30b0\u30eb\u30fc\u30d7:gccXXX
                \nABCI\u306e\u30a2\u30ab\u30a6\u30f3\u30c8:aXXX
                \nABCI\u306e\u30d1\u30b9\u30ef\u30fc\u30c9:pass
                \n\u3068\u3059\u308b\uff0e<\/p>\n

                \u2460\u30a4\u30f3\u30bf\u30e9\u30af\u30c6\u30a3\u30d6\u30ce\u30fc\u30c9(es)\u3078\u306eshh\u30ed\u30b0\u30a4\u30f3<\/h4>\n
                $ ssh -L 10022:es:22 -l aXXX as.abci.ai\r\n<\/code><\/pre>\n
                10022\u304clocalhost\u306e\u30dd\u30fc\u30c8\u756a\u53f7\u3067\uff0c 22\u304cremotehost\u306e\u30dd\u30fc\u30c8\u756a\u53f7\uff0e
                \n\u5165\u529b\u3057\u305f\u3089\u30bb\u30c3\u30b7\u30e7\u30f3\u3092\u5207\u3089\u305a\u306b\u653e\u7f6e\u3057\u3066\u304a\u304f\uff0e<\/p>\n
                <\u521d\u56de\u6642><\/h5>\n
                “Are you sure you want to continue connecting (yes\/no)?”\u3068\u805e\u304b\u308c\u308b\u306e\u3067\uff0c yes\u3068\u5165\u529b\uff0e
                \n“Enter passphrase for key ‘\/home\/username\/.ssh\/id_rsa’:”\u3067\u30d1\u30b9\u30ef\u30fc\u30c9\u3092\u5165\u529b\uff0e<\/p>\n
                <2\u56de\u76ee\u4ee5\u964d><\/h5>\n
                “Enter passphrase for key ‘\/home\/username\/.ssh\/id_rsa’:”\u3067\u521d\u56de\u6642\u306b\u5165\u529b\u3057\u305f\u30d1\u30b9\u30ef\u30fc\u30c9\u3092\u5165\u529b<\/p>\n
                \u2461\u5225\u30bf\u30fc\u30df\u30ca\u30eb\u3092\u958b\u304d\uff0c \u30a4\u30f3\u30bf\u30e9\u30af\u30c6\u30a3\u30d6\u30ce\u30fc\u30c9(es)\u306b\u30dd\u30fc\u30c8\u30d5\u30a9\u30fc\u30ef\u30fc\u30c7\u30a3\u30f3\u30b0<\/h4>\n
                $ ssh -p 10022 -l aXXX localhost\r\n<\/code><\/pre>\n
                \u2462\u30a4\u30f3\u30bf\u30e9\u30af\u30c6\u30a3\u30d6\u30ce\u30fc\u30c9\u304b\u3089\u8a08\u7b97\u30ce\u30fc\u30c9\u3078\u30ed\u30b0\u30a4\u30f3<\/h4>\n
                [aXXX@es1 ~]$ qrsh -g gccXXX -l rt_F=1  h_rt=1:00:00\r\n<\/code><\/pre>\n
                $[\u5b9f\u884c\u30b5\u30fc\u30d3\u30b9] -g gccXXX -l [\u8a08\u7b97\u30ce\u30fc\u30c9\u306e\u30bf\u30a4\u30d7] [\u30b8\u30e7\u30d6\u306e\u5b9f\u884c\u6642\u9593]<\/p>\n
                qrsh\u3067\u30a4\u30f3\u30bf\u30e9\u30af\u30c6\u30a3\u30d6\u30b8\u30e7\u30d6\u306e\u5b9f\u884c\u30b5\u30fc\u30d3\u30b9\u3092\u5229\u7528\u3057\uff0crt_F\u3067\u30ce\u30fc\u30c9\u5360\u6709.h_rt\u306f\u6307\u5b9a\u3057\u306a\u304f\u3066\u3082\u30c7\u30d5\u30a9\u30eb\u30c8\u30671:00:00\u306b\u306a\u308b\uff0e\u8a73\u3057\u304f\u306f\u3053\u3061\u3089\uff0e<\/p>\n
                2. ABCI\u3067Jupyter Notebook\u3092\u4f7f\u7528\u3059\u308b<\/h1>\n
                \u2460Python\u30e2\u30b8\u30e5\u30fc\u30eb\u306e\u30ed\u30fc\u30c9<\/h4>\n
                \u30e2\u30b8\u30e5\u30fc\u30eb\u3092\u30ed\u30fc\u30c9\u3059\u308b\u524d\u306b(\u521d\u56de\u6642or\u5b9a\u671f\u7684\u306b\u78ba\u8a8d)<\/h5>\n
                \u4f7f\u7528\u3067\u304d\u308b\u30e2\u30b8\u30e5\u30fc\u30eb\u306e\u30d0\u30fc\u30b8\u30e7\u30f3\u3092\u78ba\u8a8d\u3057\uff0c\u3000\u4f7f\u7528\u3059\u308btensorflow(\u2462\u3067\u30a4\u30f3\u30b9\u30c8\u30fc\u30eb\u3059\u308b)\u306b\u5408\u308f\u305b\u3066\uff0ccuda\u3068cudnn\u306e\u30d0\u30fc\u30b8\u30e7\u30f3\u3092\u5408\u308f\u305b\u308b\u5fc5\u8981\u304c\u3042\u308b. (\u3057\u3063\u304b\u308a\u30d0\u30fc\u30b8\u30e7\u30f3\u304c\u3042\u3063\u3066\u306a\u3044\u3068GPU\u304c\u4f7f\u308f\u308c\u306a\u3044)<\/p>\n
                tensorflow\u3068cuda, cudnn\u306e\u30d0\u30fc\u30b8\u30e7\u30f3\u306e\u7d71\u5408\u6027\u306e\u78ba\u8a8d\u306f\u3053\u3061\u3089\u3092\u53c2\u7167. 2022\u5e74\u6700\u65b0\u306eTensorFlow 2.9.1 \u306f\u3053\u3061\u3089\u306b\u8f09\u3063\u3066\u308b. \u3068\u308a\u3042\u3048\u305a, \u6700\u65b0\u306ecuda\u3068cudnn\u3092\u4f7f\u7528\u3059\u308c\u3070GPU\u306e\u51e6\u7406\u306f\u901f\u304f\u306a\u308b\uff0e<\/p>\n
                \u30e2\u30b8\u30e5\u30fc\u30eb\u306e\u30d0\u30fc\u30b8\u30e7\u30f3\u78ba\u8a8d\u306f\uff0c\u4e0b\u8a18\u306e\u30b3\u30de\u30f3\u30c9\u3067\u3067\u304d\u308b\u305f\u3081\uff0c\u521d\u56de\u6642\u306b\u78ba\u8a8d\u3057\u3066\u304a\u304f\uff0e<\/p>\n
                [aXXX@gXXX ~]$ module avail gcc\r\n[aXXX@gXXX ~]$ module avail python\r\n[aXXX@gXXX ~]$ module avail cuda\r\n[aXXX@gXXX ~]$ module avail cudnn\r\n<\/code><\/pre>\n
                \u30e2\u30b8\u30e5\u30fc\u30eb\u306e\u30ed\u30fc\u30c9\u306f\u4ee5\u4e0b\u306e\u30b3\u30de\u30f3\u30c9<\/h5>\n
                [aXXX@gXXX ~]$ module load gcc\/9.3.0 python\/3.8\/3.8.13 cuda\/11.7\/11.7.0 cudnn\/8.4\/8.4.1\r\n<\/code><\/pre>\n
                \u2461Python\u4eee\u60f3\u74b0\u5883\u3092\u4f5c\u6210 \/ \u4eee\u60f3\u74b0\u5883\u306b\u5165\u308b<\/h4>\n
                \u521d\u56de\u6642<\/h5>\n
                (1) \u4eee\u60f3\u74b0\u5883\u3092\u4f5c\u6210\u3057\uff0c\u4eee\u60f3\u74b0\u5883\u306b\u5165\u308b\uff0e<\/p>\n
                [aXXX@gXXX ~]$ python3 -m venv ~\/jupyter_env\r\n[aXXX@gXXX ~]$ source ~\/jupyter_env\/bin\/activate\r\n<\/code><\/pre>\n
                \u203b\u3061\u306a\u307f\u306b\u51fa\u308b\u5834\u5408\u306f\uff0c deactivate\u3068\u5165\u529b<\/p>\n
                (2)\u30e9\u30a4\u30d6\u30e9\u30ea\u306e\u30a4\u30f3\u30b9\u30c8\u30fc\u30eb<\/p>\n
                [aXXX@gXXX ~]$ pip3 install --upgrade pip setuptools\r\n[aXXX@gXXX ~]$ pip3 install jupyter numpy tensorflow==2.9.1\r\n[aXXX@gXXX ~]$ pip3 install tensorflow-gpu tensorflow-determinism #GPU\u4f7f\u3046\u4eba\u7528\r\n[aXXX@gXXX ~]$ pip3 install ipython-sql #sql\u3092\u4f7f\u3046\u4eba\u7528\r\n[aXXX@gXXX ~]$ pip3 install jupyter-resource-usage #\u30e1\u30e2\u30ea\u306e\u4f7f\u7528\u91cf\u3092\u78ba\u8a8d\u3057\u305f\u3044\u4eba\u7528\r\n<\/code><\/pre>\n
                pip3 list\u3067\u30a4\u30f3\u30b9\u30c8\u30fc\u30eb\u3057\u305f\u30e9\u30a4\u30d6\u30e9\u30ea\u3092\u78ba\u8a8d\u3067\u304d\u308b.<\/p>\n
                \uff12\u56de\u76ee\u4ee5\u964d<\/h5>\n
                [aXXX@gXXX ~]$ source ~\/jupyter_env\/bin\/activate\r\n<\/code><\/pre>\n
                \u306e\u307f\u3067OK<\/p>\n
                \u2462\u30b8\u30e5\u30d4\u30bf\u30fc\u30ce\u30fc\u30c8\u30d6\u30c3\u30af\u3092port8888\u3067\u8d77\u52d5\u3057\u3066\uff0c\u30c8\u30fc\u30af\u30f3\u3092\u78ba\u8a8d<\/h4>\n
                [aXXX@gXXX ~]$ jupyter notebook --ip=`hostname`  --port=8888 --no-browser\r\n<\/code><\/pre>\n
                \u2463\u5225\u30bf\u30fc\u30df\u30ca\u30eb(3\u3064\u76ee\u306e\u30bf\u30fc\u30df\u30ca\u30eb)\u3067\uff0c \u30ed\u30fc\u30ab\u30ebPC\u306e8888\u756a\u30dd\u30fc\u30c8\u3092\u8a08\u7b97\u30ce\u30fc\u30c9\u306e8888\u756a\u30dd\u30fc\u30c8\u306b\u8ee2\u9001\u3059\u308bSSH\u30c8\u30f3\u30cd\u30eb\u3092\u4f5c\u6210<\/h4>\n
                \u203b\uff11\u3064\u76ee\u3068\uff12\u3064\u76ee\u306e\u30bf\u30fc\u30df\u30ca\u30eb\u306f\u958b\u3044\u305f\u307e\u307e\u306b\u3057\u3066\u304a\u304f
                \n8888:gXXX:8888\u306eXXX\u306f,\uff12\u3064\u76ee\u306e\u30bf\u30fc\u30df\u30ca\u30eb\u306e[aXXX@gXXX ~]$\u306e\u90e8\u5206\u3092\u78ba\u8a8d.<\/p>\n
                $ ssh -L 8888:gXXX:8888 -l aXXX -p 10022 localhost\r\n<\/code><\/pre>\n
                \u2463Jupyter Notebook\u3078\u306e\u63a5\u7d9a<\/h4>\n
                \u30d6\u30e9\u30a6\u30b6\u3067
                \nhttp:\/\/127.0.0.1:8888\/
                \n\u3092\u958b\u304f\uff0e\u958b\u3044\u305f\u3089password\u304btoken\u3092\u6c42\u3081\u3089\u308c\u308b\u306e\u3067\uff0c\uff12\u3064\u76ee\u306e\u30bf\u30fc\u30df\u30ca\u30eb\u306b\u8868\u793a\u3055\u308c\u305f’http:\/\/localhost:8888\/?token=’\u306e\u5f8c\u306e\u6587\u5b57\u5217\u3092\u5165\u529b<\/p>\n
                Jupyter Notebook\u306e\u7d42\u4e86\u306f3\u3064\u76ee\u306e\u30bf\u30fc\u30df\u30ca\u30eb\u306bquit\u3068\u5165\u529b\u3059\u308b\u3053\u3068\u3067\u3067\u304d\u308b<\/p>\n
                3.GPU\u306e\u518d\u73fe\u6027\u3092\u78ba\u4fdd\u3059\u308b<\/h1>\n
                tensorflow\u3067GPU\u3092\u4f7f\u3063\u305f\u969b\u306b\u5b66\u7fd2\u7d50\u679c\u304c\u6bce\u56de\u7570\u306a\u308b\u4eba\u3078<\/p>\n
                (1) 2. (2) \u3067tensorflow-determinism\u3092\u30a4\u30f3\u30b9\u30c8\u30fc\u30eb<\/h4>\n
                [aXXX@gXXX ~]$ pip3 install tensorflow-determinism \r\n<\/code><\/pre>\n
                (2)\u30ce\u30fc\u30c8\u30d6\u30c3\u30af\u4e0a\u306b\u8a18\u5165<\/h4>\n
                os.environ['TF_DETERMINISTIC_OPS'] = '1' # GPU\u5185\u306e\u8a08\u7b97\u9806\u3092\u56fa\u5b9a\u3059\u308b\u304b\u3069\u3046\u304b\u306e\u30d5\u30e9\u30b0\u3092\u6307\u5b9a(1'\u307e\u305f\u306f'True'\u3067\u56fa\u5b9a)\r\nos.environ['TF_XLA_FLAGS'] = '--tf_xla_enable_xla_devices'\r\n\r\n# \u30b0\u30ed\u30fc\u30d0\u30eb\u30b7\u30fc\u30c9\u306e\u6307\u5b9a\r\nseed=1234\r\nnp.random.seed(seed)\r\ntf.random.set_seed(seed)\r\nos.environ[\"PYTHONHASHSEED\"] = str(seed)\r\n<\/code><\/pre>\n
                4. GPU\u3092\u8907\u6570\u4f7f\u3063\u3066\u4e26\u5217\u51e6\u7406\u3092\u3059\u308b<\/h1>\n
                \u8a13\u7df4\u30c7\u30fc\u30bf\u3092X_train, \u8a13\u7df4\u30e9\u30d9\u30eb\u3092y_train, \u691c\u8a3c\u30c7\u30fc\u30bf\u3092X_val, \u691c\u8a3c\u30e9\u30d9\u30eb\u3092y_val\u3068\u3059\u308b<\/p>\n
                import tensorflow as tf\r\n\r\nstrategy = tf.distribute.MirroredStrategy()\r\nwith strategy.scope():\r\n    <scope\u306e\u4e2d\u3067\u30e2\u30c7\u30eb\u3092\u69cb\u7bc9\uff0c\u3000\u30b3\u30f3\u30d1\u30a4\u30eb\u3059\u308b>\r\n\r\n#\u30c7\u30fc\u30bf\u3092\u5909\u5f62\r\ntrain_data=tf.data.Dataset.from_tensor_slices((X_train, y_train)).batch(BATCH_SIZE)\r\nval_data=tf.data.Dataset.from_tensor_slices((X_val, y_val)).batch(BATCH_SIZE)\r\n\r\noptions = tf.data.Options()\r\noptions.experimental_distribute.auto_shard_policy = tf.data.experimental.AutoShardPolicy.OFF\r\ntrain_data, val_data =train_data.with_options(options), val_data.with_options(options)\r\n\r\n#\u901a\u5e38\u901a\u308a\u306b\u30e2\u30c7\u30eb\u3092fit\r\nhistory =model.fit(tr_data, validation_data = (val_data))  \r\n<\/code><\/pre>\n
                \u53c2\u8003\u6587\u732e<\/h1>\n
                ABCI 2.0 User Guide Jupyter Notebook\u306e\u5229\u7528
                \nABCI 2.0 User Guide Python
                \nABCI 2.0 User Guide TensorFlow-Keras
                \nTensorflow\u306b\u3088\u308b\u518d\u73fe\u6027\u306e\u78ba\u4fdd\u306b\u3064\u3044\u3066\uff08GPU\u3092\u4f7f\u7528\u3059\u308b\u5834\u5408\uff09
                \nGPU\u3067\u4e26\u5217\u51e6\u7406\u3092\u3059\u308b\u969b\u306e\u5165\u529b\u30c7\u30fc\u30bf\u306e\u5f62\u5f0f\u306b\u3064\u3044\u3066<\/p>\n","protected":false},"excerpt":{"rendered":"
                \u306f\u3058\u3081\u306b Google colaboratory\u304b\u3089ABCL\u306b\u74b0\u5883\u3092\u5909\u3048\u305f\u304c\uff0cABCL\u306e\u74b0\u5883\u69cb\u7bc9\u3068Jupyte […]<\/p>\n","protected":false},"author":12,"featured_media":0,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[1],"tags":[],"class_list":["post-46643","post","type-post","status-publish","format-standard","hentry","category-uncategorized"],"yoast_head":"\n- Blog - Silicon Cloud<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.silicloud.com\/zh\/blog\/46643-2\/\" \/>\n<meta property=\"og:locale\" content=\"zh_CN\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:description\" content=\"\u306f\u3058\u3081\u306b Google colaboratory\u304b\u3089ABCL\u306b\u74b0\u5883\u3092\u5909\u3048\u305f\u304c\uff0cABCL\u306e\u74b0\u5883\u69cb\u7bc9\u3068Jupyte […]\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.silicloud.com\/zh\/blog\/46643-2\/\" \/>\n<meta property=\"og:site_name\" content=\"Blog - Silicon Cloud\" \/>\n<meta property=\"article:published_time\" content=\"2023-01-28T07:45:08+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2024-05-03T22:31:34+00:00\" \/>\n<meta name=\"author\" content=\"\u9038, \u79d1\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"\u4f5c\u8005\" \/>\n\t<meta name=\"twitter:data1\" content=\"\u9038, \u79d1\" \/>\n\t<meta name=\"twitter:label2\" content=\"\u9884\u8ba1\u9605\u8bfb\u65f6\u95f4\" \/>\n\t<meta name=\"twitter:data2\" content=\"2 \u5206\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\/\/www.silicloud.com\/zh\/blog\/46643-2\/\",\"url\":\"https:\/\/www.silicloud.com\/zh\/blog\/46643-2\/\",\"name\":\"- Blog - Silicon Cloud\",\"isPartOf\":{\"@id\":\"https:\/\/www.silicloud.com\/zh\/blog\/#website\"},\"datePublished\":\"2023-01-28T07:45:08+00:00\",\"dateModified\":\"2024-05-03T22:31:34+00:00\",\"author\":{\"@id\":\"https:\/\/www.silicloud.com\/zh\/blog\/#\/schema\/person\/85c1dae56e6ea1e695c73d33c684d487\"},\"inLanguage\":\"zh-Hans\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/www.silicloud.com\/zh\/blog\/46643-2\/\"]}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/www.silicloud.com\/zh\/blog\/#website\",\"url\":\"https:\/\/www.silicloud.com\/zh\/blog\/\",\"name\":\"Blog - Silicon Cloud\",\"description\":\"\",\"inLanguage\":\"zh-Hans\"},{\"@type\":\"Person\",\"@id\":\"https:\/\/www.silicloud.com\/zh\/blog\/#\/schema\/person\/85c1dae56e6ea1e695c73d33c684d487\",\"name\":\"\u9038, \u79d1\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"zh-Hans\",\"@id\":\"https:\/\/www.silicloud.com\/zh\/blog\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/secure.gravatar.com\/avatar\/c94f6d9cbbfbca863fab309840bd690c153c95f8490c290ad2ed54dd693dad16?s=96&d=mm&r=g\",\"contentUrl\":\"https:\/\/secure.gravatar.com\/avatar\/c94f6d9cbbfbca863fab309840bd690c153c95f8490c290ad2ed54dd693dad16?s=96&d=mm&r=g\",\"caption\":\"\u9038, \u79d1\"},\"url\":\"https:\/\/www.silicloud.com\/zh\/blog\/author\/keyi\/\"},{\"@type\":\"ImageObject\",\"inLanguage\":\"zh-Hans\",\"@id\":\"https:\/\/www.silicloud.com\/zh\/blog\/46643-2\/#local-main-organization-logo\",\"url\":\"\",\"contentUrl\":\"\",\"caption\":\"Blog - Silicon Cloud\"}]}<\/script>\n<!-- \/ Yoast SEO Premium plugin. -->","yoast_head_json":{"title":"- Blog - Silicon Cloud","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.silicloud.com\/zh\/blog\/46643-2\/","og_locale":"zh_CN","og_type":"article","og_description":"\u306f\u3058\u3081\u306b Google colaboratory\u304b\u3089ABCL\u306b\u74b0\u5883\u3092\u5909\u3048\u305f\u304c\uff0cABCL\u306e\u74b0\u5883\u69cb\u7bc9\u3068Jupyte […]","og_url":"https:\/\/www.silicloud.com\/zh\/blog\/46643-2\/","og_site_name":"Blog - Silicon Cloud","article_published_time":"2023-01-28T07:45:08+00:00","article_modified_time":"2024-05-03T22:31:34+00:00","author":"\u9038, \u79d1","twitter_card":"summary_large_image","twitter_misc":{"\u4f5c\u8005":"\u9038, \u79d1","\u9884\u8ba1\u9605\u8bfb\u65f6\u95f4":"2 \u5206"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/www.silicloud.com\/zh\/blog\/46643-2\/","url":"https:\/\/www.silicloud.com\/zh\/blog\/46643-2\/","name":"- Blog - Silicon Cloud","isPartOf":{"@id":"https:\/\/www.silicloud.com\/zh\/blog\/#website"},"datePublished":"2023-01-28T07:45:08+00:00","dateModified":"2024-05-03T22:31:34+00:00","author":{"@id":"https:\/\/www.silicloud.com\/zh\/blog\/#\/schema\/person\/85c1dae56e6ea1e695c73d33c684d487"},"inLanguage":"zh-Hans","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.silicloud.com\/zh\/blog\/46643-2\/"]}]},{"@type":"WebSite","@id":"https:\/\/www.silicloud.com\/zh\/blog\/#website","url":"https:\/\/www.silicloud.com\/zh\/blog\/","name":"Blog - Silicon Cloud","description":"","inLanguage":"zh-Hans"},{"@type":"Person","@id":"https:\/\/www.silicloud.com\/zh\/blog\/#\/schema\/person\/85c1dae56e6ea1e695c73d33c684d487","name":"\u9038, \u79d1","image":{"@type":"ImageObject","inLanguage":"zh-Hans","@id":"https:\/\/www.silicloud.com\/zh\/blog\/#\/schema\/person\/image\/","url":"https:\/\/secure.gravatar.com\/avatar\/c94f6d9cbbfbca863fab309840bd690c153c95f8490c290ad2ed54dd693dad16?s=96&d=mm&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/c94f6d9cbbfbca863fab309840bd690c153c95f8490c290ad2ed54dd693dad16?s=96&d=mm&r=g","caption":"\u9038, \u79d1"},"url":"https:\/\/www.silicloud.com\/zh\/blog\/author\/keyi\/"},{"@type":"ImageObject","inLanguage":"zh-Hans","@id":"https:\/\/www.silicloud.com\/zh\/blog\/46643-2\/#local-main-organization-logo","url":"","contentUrl":"","caption":"Blog - Silicon Cloud"}]}},"_links":{"self":[{"href":"https:\/\/www.silicloud.com\/zh\/blog\/wp-json\/wp\/v2\/posts\/46643","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.silicloud.com\/zh\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.silicloud.com\/zh\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.silicloud.com\/zh\/blog\/wp-json\/wp\/v2\/users\/12"}],"replies":[{"embeddable":true,"href":"https:\/\/www.silicloud.com\/zh\/blog\/wp-json\/wp\/v2\/comments?post=46643"}],"version-history":[{"count":2,"href":"https:\/\/www.silicloud.com\/zh\/blog\/wp-json\/wp\/v2\/posts\/46643\/revisions"}],"predecessor-version":[{"id":96686,"href":"https:\/\/www.silicloud.com\/zh\/blog\/wp-json\/wp\/v2\/posts\/46643\/revisions\/96686"}],"wp:attachment":[{"href":"https:\/\/www.silicloud.com\/zh\/blog\/wp-json\/wp\/v2\/media?parent=46643"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.silicloud.com\/zh\/blog\/wp-json\/wp\/v2\/categories?post=46643"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.silicloud.com\/zh\/blog\/wp-json\/wp\/v2\/tags?post=46643"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}