由于出现以下错误
所以怀疑是没有运行fully_connected_feed.py文件的问题。
错误：
不能运行fully_connected_feed.py
解决：
把克隆好的mnist数据集放入C:\Users\yyuqi\Anaconda3\Lib\site-packages\tensorflow\tensorflow\examples\tutorials\mnist中
运行python fully_connected_feed.py
又出现错误：
ModuleNotFoundError: No module named 'tensorflow.examples'
解决：
需要在创建的虚拟环境（Anaconda Prompt）中运行。
又出现错误：
C:\Users\yyuqi\Anaconda3\envs\tensorflow\lib\site-packages\h5py\__init__.py:34: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`. from ._conv import register_converters as _register_convertersTraceback (most recent call last): File "fully_connected_feed.py", line 279, in tf.app.run(main=main, argv=[sys.argv[0]] + unparsed)
File "C:\Users\yyuqi\Anaconda3\envs\tensorflow\lib\site-packages\tensorflow\python\platform\app.py", line 48, in run
_sys.exit(main(_sys.argv[:1] + flags_passthrough))
File "fully_connected_feed.py", line 221, in main
tf.gfile.MakeDirs(FLAGS.log_dir)
File "C:\Users\yyuqi\Anaconda3\envs\tensorflow\lib\site-packages\tensorflow\python\lib\io\file_io.py", line 367, in recursive_create_dir
pywrap_tensorflow.RecursivelyCreateDir(compat.as_bytes(dirname), status)
File "C:\Users\yyuqi\Anaconda3\envs\tensorflow\lib\contextlib.py", line 66, in __exit__
next(self.gen)
File "C:\Users\yyuqi\Anaconda3\envs\tensorflow\lib\site-packages\tensorflow\python\framework\errors_impl.py", line 466, in raise_exception_on_not_ok_status
pywrap_tensorflow.TF_GetCode(status))
tensorflow.python.framework.errors_impl.NotFoundError: Failed to create a directory: /tmp\tensorflow
其中错误：
File "fully_connected_feed.py", line 279, in tf.app.run(main=main, argv=[sys.argv[0]] + unparsed)
以上报错原因是tensorflow版本太低。
查看tensorflow版本办法：
<<< import tensorflow
<<< tensorflow.__version__
升级tensorflow办法：
GPU版：pip3 install --upgrade tensorflow-gpu
CPU版：pip3 install --upgrade tensorflow
更新pip：python -m pip install --upgrade pip
