关于cnn网络结构疑惑

[snowlord]

网络中第二层conv2d的输入应该是outputs_1[0]吧，而不是每次都是input_x1。代码是theano_cnn版本

[white127]

为什么是outputs呢？
卷积的输入应该是input，对词向量组成的矩阵进行卷积

[white127]

github已给出答复

2016-09-23 17:52 GMT+08:00 snowlord notifications@github.com:

[image: image]
https://cloud.githubusercontent.com/assets/8564700/18781837/119403dc-81b6-11e6-9d2c-f229f52e5138.png

网络中第二层conv2d的输入应该是outputs_1[0]吧，而不是每次都是input_x1。

—
You are receiving this because you are subscribed to this thread.
Reply to this email directly, view it on GitHub
#1, or mute the
thread
https://github.com/notifications/unsubscribe-auth/ABffUBW7AE5KFKUJxqp8e_qnSo2fT8xTks5qs6FzgaJpZM4KEzXV
.

[snowlord]

@white127 谢谢！