Tensorflow reads tensor operations from ckpt files

in usingpre-train modellength of time，We needrestore variables from checkpoint files.

Often "Tensor name not found" is not found in checkpoint.

This is a good time to check what variables are actually in ckpt

import os
from  import pywrap_tensorflow

checkpoint_path = (model_dir, "")
# 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))

You can display the tensor names and values in ckpt, and of course you can debug with pycharm.

ADDITIONAL: Reading the values saved in the model in tensorflow, the

Use (model_dir)

A standard model file has the following files, model_dir is MyModel (no suffix)

checkpoint

-00000-of-00001

import tensorflow as tf
import pprint # Improve print readability with pprint
NewCheck =("model")

Print all variables in the model

print("debug_string:\n")
(NewCheck.debug_string().decode("utf-8"))

There are three fields, name, data type and shape.

Getting the value in a variable

print("get_tensor:\n")
(NewCheck.get_tensor("D/conv2d/bias"))

print("get_variable_to_dtype_map\n")
(NewCheck.get_variable_to_dtype_map())
print("get_variable_to_shape_map\n")
(NewCheck.get_variable_to_shape_map())

The above is a personal experience, I hope it can give you a reference, and I hope you can support me more. If there is any mistake or something that has not been fully considered, please do not hesitate to give me advice.