import tensorflow as tf
import numpy as np
input1 = tf.placeholder(tf.float32)
input2 = tf.placeholder(tf.float32)
output = tf.matmul(input1, input2)
with tf.Session() as sess:
print(sess.run(output, {
input1:[1, 2],
input2:np.array([[1,2]]).T}))
出现报错,In[0] ndims must be >= 2: 1。发现原理是使用matmul时对象必须是秩>2的张量,这里两个张量相乘修改为multiply就好了
output = tf.multiply(input1, input2)