The core code is as follows:
[ for tensor in tf.get_default_graph().as_graph_def().node]
Example code: (The model of Inceptino_v3 is loaded and the names of all nodes of the model are obtained)
# -*- coding: utf-8 -*- import tensorflow as tf import os model_dir = 'C:/Inception_v3' model_name = 'output_graph.pb' # Read and create a graph to store the trained Inception_v3 model (function)def create_graph(): with (( model_dir, model_name), 'rb') as f: # Use() to define an empty Graph graph_def = () graph_def.ParseFromString(()) # Imports the graph from graph_def into the current default Graph. tf.import_graph_def(graph_def, name='') # Create graphcreate_graph() tensor_name_list = [ for tensor in tf.get_default_graph().as_graph_def().node] for tensor_name in tensor_name_list: print(tensor_name,'\n')
Output result:
mixed_8/tower/conv_1/batchnorm/moving_variance mixed_8/tower/conv_1/batchnorm r_1/mixed/conv_1/batchnorm . . . mixed_10/tower_1/mixed/conv_1/CheckNumerics mixed_10/tower_1/mixed/conv_1/control_dependency mixed_10/tower_1/mixed/conv_1 pool_3 pool_3/_reshape/shape pool_3/_reshape input/BottleneckInputPlaceholder . . . . final_training_ops/weights/final_weights final_training_ops/weights/final_weights/read final_training_ops/biases/final_biases final_training_ops/biases/final_biases/read final_training_ops/Wx_plus_b/MatMul final_training_ops/Wx_plus_b/add final_result
Because the result was too long, some were omitted.
If you don't want to print the output like this, you can also write it to txt to view it.
Write the txt code as follows:
tensor_name_list = [ for tensor in tf.get_default_graph().as_graph_def().node] txt_path = './txt/node name' full_path = txt_path+ '.txt' for tensor_name in tensor_name_list: name = tensor_name + '\n' file = open(full_path,'a+') (name) ()
The above article TensorFlow obtains all the tensor name codes in the loading model are all the content I share with you. I hope you can give you a reference and I hope you can support me more.