get_or_create_global_step

import tensorflow.compat.v1 as tf
tf.disable_v2_behavior()
# tf.enable_eager_execution()
import numpy as np

x = tf.placeholder(tf.float32, shape=[None, 1], name='x')
y = tf.placeholder(tf.float32, shape=[None, 1], name='y')
w = tf.Variable(tf.constant(0.0))

# 手动设置
# global_steps = tf.Variable(0, trainable=False)
# 接口设置
global_steps = tf.train.get_or_create_global_step()

learning_rate = tf.train.exponential_decay(0.1, global_steps, 10, 2, staircase=False)
loss = tf.pow(w * x - y, 2)

train_step = tf.train.GradientDescentOptimizer(learning_rate).minimize(loss, global_step=global_steps)

with tf.Session() as sess:
    sess.run(tf.global_variables_initializer())
    for i in range(10):
        sess.run(train_step, feed_dict={
     x: np.linspace(1, 2, 10).reshape([10, 1]),
                                        y: np.linspace(1, 2, 10).reshape([10, 1])})
        print(sess.run(learning_rate))
        print(sess.run(global_steps))

由global_step =tf.train.get_or_create_global_step（）产生的疑问