我對(duì) Google Cloud Platform 還很陌生，我正在嘗試使用 TPU 訓(xùn)練模型。我按照本教程使用 Google Colab 設(shè)置 TPU。下面的所有代碼都遵循教程。這是我完成的步驟：import datetimeimport jsonimport osimport pprintimport randomimport stringimport sysimport tensorflow as tfassert 'COLAB_TPU_ADDR' in os.environ, 'ERROR: Not connected to a TPU runtime; please see the first cell in this notebook for instructions!'TPU_ADDRESS = 'grpc://' + os.environ['COLAB_TPU_ADDR']print('TPU address is => ', TPU_ADDRESS)from google.colab import authauth.authenticate_user()with tf.Session(TPU_ADDRESS) as session:  print('TPU devices:')  pprint.pprint(session.list_devices())  # Upload credentials to TPU.  with open('/content/adc.json', 'r') as f:    auth_info = json.load(f)  tf.contrib.cloud.configure_gcs(session, credentials=auth_info)  # Now credentials are set for all future sessions on this TPU.輸出：TPU address is =>  grpc://10.4.89.154:8470提供我的BUCKET名字和OUPUT DIRECTORY姓名：BUCKET = 'my_xlnet' #@param {type:"string"}assert BUCKET, '*** Must specify an existing GCS bucket name ***'output_dir_name = 'xlnet_output' #@param {type:"string"}BUCKET_NAME = 'gs://{}'.format(BUCKET)OUTPUT_DIR = 'gs://{}/{}'.format(BUCKET,output_dir_name)tf.gfile.MakeDirs(OUTPUT_DIR)print('***** Model output directory: {} *****'.format(OUTPUT_DIR))將預(yù)訓(xùn)練模型移至 GCS 存儲(chǔ)桶：!gsutil mv /content/xlnet_extension_tf/model/xlnet_cased_L-24_H-1024_A-16 $BUCKET_NAME輸出：...Operation completed over 5 objects/1.3 GiB.   然后運(yùn)行主要代碼：!python /content/xlnet_extension_tf/run_coqa.py \--use_tpu=True \--tpu_name=grpc://10.4.89.154:8470 \--spiece_model_file=$BUCKET_NAME/xlnet_cased_L-24_H-1024_A-16/spiece.model \--model_config_path=$BUCKET_NAME/xlnet_cased_L-24_H-1024_A-16/xlnet_config.json \--init_checkpoint=$BUCKET_NAME/xlnet_cased_L-24_H-1024_A-16/xlnet_model.ckpt \...然后我得到了這個(gè)錯(cuò)誤：OSError: Not found: "gs://my_xlnet/xlnet_cased_L-24_H-1024_A-16/spiece.model": No such file or directory Error #2這是 GCS 存儲(chǔ)桶屏幕：我不知道為什么會(huì)出現(xiàn)這個(gè)錯(cuò)誤，因?yàn)槲铱梢猿晒Φ貙⑽业念A(yù)訓(xùn)練模型移動(dòng)到桶中。你們知道如何解決這個(gè)問(wèn)題嗎？