Original Address:
https://blog.csdn.net/uestc_c2_403/article/details/73350457
Since tensorflow updated version problem usage slightly modified
----------------------------------------------------------------------------------
tf.split(input, num_split, dimension):
dimension input which means that one dimension tensor, if it is 0 0 means that the first cutting dimensions. num_split is cutting the number of, if it means 2 is input tensor is cut into 2 parts, each one is a list.
E.g:
import tensorflow as tf; import numpy as np; A = [[1,2,3],[4,5,6]] x = tf.split(A, 3, 1) with tf.Session() as sess: c = sess.run(x) for ele in c: print( ele )
Output:
[[1]
[4]]
[[2]
[5]]
[[3]
[6]]
Note: This program is installed tf version is 0.12.0, subject to change in a different version, which is a function of usage will be different, all of a sudden attention.
---------------------------------------------------------------------------------------------------------------------
import tensorflow as tf; import numpy as np; A = [[1,2,3],[4,5,6]] x = tf.split(A, 2, 0) with tf.Session() as sess: c = sess.run(x) for ele in c: print( ele )
Output:
[[1 2 3]]
[[4 5 6]]