tensorflow ckpt模型和pb模型获取节点名称,及ckpt转pb模型实例

ckpt

from tensorflow.python import pywrap_tensorflow
checkpoint_path = 'model.ckpt-8000'
reader = pywrap_tensorflow.NewCheckpointReader(checkpoint_path)
var_to_shape_map = reader.get_variable_to_shape_map()
for key in var_to_shape_map:
print("tensor_name: ", key)

pb

import tensorflow as tf
import os

model_name = './mobilenet_v2_140_inf_graph.pb'

def create_graph():
with tf.gfile.FastGFile(model_name, 'rb') as f:
 graph_def = tf.GraphDef()
 graph_def.ParseFromString(f.read())
 tf.import_graph_def(graph_def, name='')

create_graph()
tensor_name_list = [tensor.name for tensor in tf.get_default_graph().as_graph_def().node]
for tensor_name in tensor_name_list:
print(tensor_name,'\n')

ckpt转pb

def freeze_graph(input_checkpoint,output_graph):
'''
:param input_checkpoint:
:param output_graph: PB模型保存路径
:return:
'''
output_node_names = "xxx"
saver = tf.train.import_meta_graph(input_checkpoint + '.meta', clear_devices=True)
graph = tf.get_default_graph()
input_graph_def = graph.as_graph_def()
with tf.Session() as sess:
 saver.restore(sess, input_checkpoint)
 output_graph_def = graph_util.convert_variables_to_constants(
  sess=sess,
  input_graph_def=input_graph_def,# 等于:sess.graph_def
  output_node_names=output_node_names.split(","))
 with tf.gfile.GFile(output_graph, "wb") as f:
  f.write(output_graph_def.SerializeToString())
 print("％d ops in the final graph." ％ len(output_graph_def.node))

for op in graph.get_operations():
  print(op.name, op.values())

来源：https://blog.csdn.net/qq_38109843/article/details/88841306