# tf.train.Saver类提供了保存和恢复模型的方法 # tf.train.Saver的构造函数 提供了save和恢复的参数选项 # Saver对象提供了方法来运行这些计算节点，制定了写和读的路径 # Saver会恢复所有在你模型当中已经定义的变量 # 如果你加载一个模型没有通知如果构建该模型的计算图 # TensorFlow 在二进制检查点文件中保存变量, 粗略地说, 将变量名映射到张量值 # Saving variable # 创建一个Saver使用tf.train.Saver()来管理所有的变量在模型中 # 列如下面的代码块解释了如何调用tf.train.Saver.save方法来保存变量到checkpoint文件中 import tensorflow as tf # Create some variables v1 = tf.get_variable("v1", shape=[3], initializer=tf.zeros_initializer) v2 = tf.get_variable("v2", shape=[5], initializer=tf.zeros_initializer) inc_v1 = v1.assign(v1 + 1) dec_v2 = v2.assign(v2 - 1) # Add an op to initialize the variables init_op = tf.global_variables_initializer() # Add ops to save and restore all the variables saver = tf.train.Saver() # Later,launch the model,initialize the variables, # do some work, and save the variables to disk with tf.Session() as sess: sess.run(init_op) # Do some work with the model inc_v1.op.run() dec_v2.op.run() # Save the variables to disk save_path = saver.save(sess, "tmp/model.ckpt") print("Model saved in path : %s " % save_path)
保存的结果如下：
2018-02-17 11:21:22.621424: I C: f_jenkinsworkspace el-winMwindowsPY35 ensorflowcoreplatformcpu_feature_guard.cc:137] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX AVX2 Model saved in path : tmp/model.ckpt