保存/恢复张量流模型的权重

Question

我正在尝试保存和恢复张量流模型的参数。该代码确实将参数保存到给定的路径，但是当我尝试还原最后一个检查点，执行操作然后再次保存时，它不会返回最后一个检查点。我该怎么办？

import tensorflow as tf
import os



v1 = tf.Variable(1.32, name="v1")
v2 = tf.Variable(1.33, name="v2")


saver = tf.train.Saver()

init = tf.initialize_all_variables()


with tf.Session() as sess:


    sess.run(init)


    print (v2.eval(sess)) 

    saver.save(sess, "/tmp/model")
    print("Model restored.")


for i in range(10):

    ckpt = tf.train.get_checkpoint_state(os.path.dirname('/tmp/'))

    print(ckpt)

    sess = tf.Session()

    if(ckpt and ckpt.model_checkpoint_path):


        saver.restore(sess, ckpt.model_checkpoint_path)

        x = sess.run("v2:0")
        x = x + 1

        print(x)


        saver.save(sess, "/tmp/model", global_step = i)

        sess.close()