tf.pad()的使用

笼统来说,tf.pad()的作用是填充,其函数原型如下:

def pad(tensor, paddings, mode="CONSTANT", name=None, constant_values=0):
# tensor :指待填充张量
# paddings :指用何种方式填充
# comde :填充模式
# constant_value :指定填充指

其中当paddings的维度为2x2时:paddings = [[up,down],[left,right]四个位置,分别表示在四个方向上填充
举例:

import tensorflow as tf

tensor = tf.constant([[1,2,3],[4,5,6]])
paddings = tf.constant([[0, 2,], [0, 0]])#下下方填充2行0
embed = tf.pad(tensor,paddings,'CONSTANT')
with tf.Session() as sess:
    tf.global_variables_initializer().run()
    print(sess.run(embed))

#
[[1 2 3]
 [4 5 6]
 [0 0 0]
 [0 0 0]]
import tensorflow as tf

tensor = tf.constant([[1,2,3],[4,5,6]])
paddings = tf.constant([[0, 2,], [0, 0]])#在下方填充2行0,在右方填充2列0
embed = tf.pad(tensor,paddings,'CONSTANT')
with tf.Session() as sess:
    tf.global_variables_initializer().run()
    print(sess.run(embed))

#
[[1 2 3 0 0 0]
 [4 5 6 0 0 0]
 [0 0 0 0 0 0]
 [0 0 0 0 0 0]]

参考:

TensorFlow中关于pad函数的详细理解

猜你喜欢

转载自blog.csdn.net/The_lastest/article/details/81410912