方式1：靜態獲取,通過直接解析checkpoint文件獲取變量名及變量值

通過

reader = tf.train.NewCheckpointReader(model_path)

或者通過：

from tensorflow.python import pywrap_tensorflowreader = pywrap_tensorflow.NewCheckpointReader(model_path)

代碼：

model_path = "./checkpoints/model.ckpt-75000"## 下面兩個reader作用等價#reader = pywrap_tensorflow.NewCheckpointReader(model_path)reader = tf.train.NewCheckpointReader(model_path) ## 用reader獲取變量字典,key是變量名,value是變量的shapevar_to_shape_map = reader.get_variable_to_shape_map()for var_name in var_to_shape_map.keys(): #用reader獲取變量值 var_value = reader.get_tensor(var_name)  print("var_name",var_name) print("var_value",var_value)

方式2：動態獲取,先加載checkpoint模型,然后用graph.get_tensor_by_name()獲取變量值

代碼 (注意：要先在腳本中構建model中對應的變量及scope):

 model_path = "./checkpoints/model.ckpt-75000" config = tf.ConfigProto() config.gpu_options.allow_growth = True with tf.Session(config=config) as sess:  ## 獲取待加載的變量列表  trainable_vars = tf.trainable_variables()  g_vars = tf.get_collection(tf.GraphKeys.GLOBAL_VARIABLES,scope="generator")  d_vars = tf.get_collection(tf.GraphKeys.GLOBAL_VARIABLES,scope='discriminator')  flow_vars = tf.get_collection(tf.GraphKeys.GLOBAL_VARIABLES,scope='flow_net')  var_restore = g_vars + d_vars   ## 僅加載目標變量  loader = tf.train.Saver(var_restore)  loader.restore(sess,model_path)   ## 顯示加載的變量值  graph = tf.get_default_graph()  for var in var_restore:   tensor = graph.get_tensor_by_name(var.name)   print("=======變量名=======",tensor)   print("-------變量值-------",sess.run(tensor))

以上這篇tensorflow 獲取checkpoint中的變量列表實例就是小編分享給大家的全部內容了，希望能給大家一個參考，也希望大家多多支持武林站長站。