张量流常数为何为可馈tru

Question

我在学张量流

import tensorflow as tf
print(tf.VERSION)

a = tf.placeholder(tf.float32, shape=[3])
b = tf.constant([2, 2, 2], tf.float32)
c = a + b
with tf.Session() as sess:
    print(tf.get_default_graph().is_feedable(b))
    print(sess.run(c, feed_dict={a: [3, 2, 3]}))

结果如下

^{pr2}$

我不明白为什么张量流说常数是可馈的。占位符是feedable true，但为什么是常量？在

Answer 1

因为在TF中，您还可以feed values in constants and variables：

While you can replace any Tensor with feed data, including variables and constants, the best practice is to use a tf.placeholder node

你自己检查一下：

import tensorflow as tf

a = tf.constant([2, 2, 2], tf.float32)
with tf.Session() as sess:
    print(sess.run(a, feed_dict={a: [3, 2, 3]}))

常量a更改了其值。在