Models' forward are incorrect

[sagarwaghmare69]

Hey just realized that the model forward/sequence is wrong.
E.g
Current code:
input is (nc) x 256 x 256

e1 = self.conv1(input)
# input is (ngf) x 128 x 128
e2 = self.batch_norm2(self.conv2(self.leaky_relu(e1)))

This should be
input is (nc) x 256 x 256

e1 = self.conv1(input)
# input is (ngf) x 128 x 128
e2 = self.conv2(self.leaky_relu(self.batch_norm2(e1)))

Hence both the discriminator and generator models are incorrect.

[mrzhu-cool]

Hey, I checked the code carefully today.
The model construction seems to have no wrong.
It's structure is the same as the original lua torch code.
A little mistake is that I forgot a leaky_relu layer after h4 in netD.

[sagarwaghmare69]

Hi, My bad regarding the receptive field. If you pushed the changes to repo I would love to check it out. Thanks

…

On Fri, Mar 17, 2017 at 8:14 AM, mrzhu ***@***.***> wrote: Hey, I checked the code carefully today. The model construction seems to have no wrong. It's structure is the same as the original lua torch code. A little mistake is that I forgot a leaky_relu layer after h4 in netD. — You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub <#2 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ADR7VdVmyK58N_7quK-0FWZ8LfBXQikAks5rmnkYgaJpZM4MfaB2> .

-- Kind regards Sagar M Waghmare

[sagarwaghmare69]

Nice. Thanks. Let me know if the results improve.
Closing the issue.