ValueError: Shapes (40, 4, 4, 1) and (?, 8, 8, 1) are not compatible
使用conv_transpose中出现的问题,即是解卷积,是卷积的一个逆过程。
sess = tf.Session()
batch_size = 3
output_shape = [batch_size, 8, 8, 128]
strides = [1, 2, 2, 1]
l = tf.constant(0.1, shape=[batch_si
原创
2017-10-30 12:12:16 ·
4960 阅读 ·
0 评论