方式1：靜態獲取,通過直接解析checkpoint檔案獲取變數名及變數值

通過

reader = tf.train.NewCheckpointReader(model_path)

或者通過：

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

程式碼：

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獲取變數字典,key是變數名,value是變數的shape
var_to_shape_map = reader.get_variable_to_shape_map()
for var_name in var_to_shape_map.keys():
 #用reader獲取變數值
 var_value = reader.get_tensor(var_name)
 
 print("var_name",var_name)
 print("var_value",var_value)

方式2：動態獲取,先載入checkpoint模型,然後用graph.get_tensor_by_name()獲取變數值

程式碼 (注意：要先在指令碼中構建model中對應的變數及scope):

 model_path = "./checkpoints/model.ckpt-75000"
 config = tf.ConfigProto()
 config.gpu_options.allow_growth = True
 with tf.Session(config=config) as sess:
  ## 獲取待載入的變數列表
  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
 
  ## 僅載入目標變數
  loader = tf.train.Saver(var_restore)
  loader.restore(sess,model_path)
 
  ## 顯示載入的變數值
  graph = tf.get_default_graph()
  for var in var_restore:
   tensor = graph.get_tensor_by_name(var.name)
   print("=======變數名=======",tensor)
   print("-------變數值-------",sess.run(tensor))

以上這篇tensorflow 獲取checkpoint中的變數列表例項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，也希望大家多多支援我們。