Tensorflow: 从checkpoint文件中读取tensor

在使用pre-train model时候，我们需要restore variables from checkpoint files.

经常出现在checkpoint 中找不到”Tensor name not found”.

这时候需要查看一下ckpt中到底有哪些变量

import os
from tensorflow.python import pywrap_tensorflow

checkpoint_path = os.path.join(model_dir, "model.ckpt")
# Read data from checkpoint file
reader = pywrap_tensorflow.NewCheckpointReader(checkpoint_path)
var_to_shape_map = reader.get_variable_to_shape_map()
# Print tensor name and values
for key in var_to_shape_map:
print("tensor_name: ", key)
print(reader.get_tensor(key))

可以显示ckpt中的tensor名字和值，当然也可以用pycharm调试。