Tensorflow -( retrain.py ) OSError：[Errno 2]没有这样的文件或目录：'‘

Question

我在Ubuntu中运行tensorflow (在VirtualEnv中)。

我想创建训练模型...(使用retrain.py训练模型)..

Github代码链接==> https://github.com/tensorflow/tensorflow/blob/master/tensorflow/examples/image_retraining/retrain.py

我试着用命令在终端上运行

python image_retraining/retrain.py \

   --bottleneck_dir= /tf_files/bottlenecks \

   --how_many_training_steps 100 \

   --model_dir= /tf_files/inception \

   --output_graph= /tf_files/retrained_graph.pb \

   --output_labels= /tf_files/retrained_labels.txt \

   --image_dir = /Image_Processing/theory/Green

我得到了错误。如何读取错误以及如何解决它..

Traceback (most recent call last):
  File "image_retraining/retrain.py", line 1061, in <module>
    tf.app.run(main=main, argv=[sys.argv[0]] + unparsed)
  File "/Image_Processing/tensorflow/local/lib/python2.7/site-packages/tensorflow/python/platform/app.py", line 43, in run
    sys.exit(main(sys.argv[:1] + flags_passthrough))
  File "image_retraining/retrain.py", line 777, in main
    maybe_download_and_extract()
  File "image_retraining/retrain.py", line 288, in maybe_download_and_extract
    os.makedirs(dest_directory)
  File "/Image_Processing/tensorflow/lib/python2.7/os.py", line 157, in makedirs
    mkdir(name, mode)
OSError: [Errno 2] No such file or directory: ''

注意:-我是tensorflow的新手，刚刚开始使用它。

Answer 1

在运行retrain.py时，您不必指定所有参数。只需指定

python /path_to_file/retrain.py --img_dir /dir_containing_subfolders_of_images

在运行代码时，它将使用剩余参数的默认路径。例如：/tmp/some_dir_or_file用于大多数情况。

如果要指定自定义位置，请创建所需的目录并指定这些路径。例如：

python /path_to_file/retrain.py 
--img_dir /dir_containing_subfolders_of_images
--how_many_training_steps number_of_steps
--output_graph /dirPath_to_store_output_graph/file_name.pb
--output_labels /dirPath_to_store_output_labels/file_name.txt