tf.split使用

import tensorflow as tf

tensor = [[[1,2,3],
[4,5,6],
[7,8,9]]]

with tf.Session() as sess:

tensor1, tensor2 = tf.split(tensor, [1,2], 2) #将tensor在第2个维度（基1）切为1份和2份
print(tensor2 .eval())

代码运行结果：
[[[2 3]
[5 6]
[8 9]]]

来源：CSDN

作者：不喝牛奶的里昂

链接：https://blog.csdn.net/qq_34895059/article/details/103465782