Skip to content

Conversation

@ezyang
Copy link
Contributor

@ezyang ezyang commented Jan 8, 2018

Signed-off-by: Edward Z. Yang ezyang@fb.com

@ezyang ezyang requested a review from soumith January 8, 2018 22:23
@pytorchbot
Copy link
Collaborator

@ezyang, thanks for your PR! We identified @zdevito to be a potential reviewer.

@soumith
Copy link
Contributor

soumith commented Jan 8, 2018

@ezyang isn't it cleaner to just have the CuDNN bindings make weight contiguous always...?

This comment was marked as off-topic.

@ezyang
Copy link
Contributor Author

ezyang commented Jan 9, 2018

@ezyang isn't it cleaner to just have the CuDNN bindings make weight contiguous always...?

Yeah, I was vacillating between the two ways. I'll swap it.

I think we can just return out; here to fix the Windows CI error

It's not just a CI error; it's a legit bug. Thanks for pointing it out!

ezyang added 2 commits January 9, 2018 06:46
Signed-off-by: Edward Z. Yang <ezyang@fb.com>
Signed-off-by: Edward Z. Yang <ezyang@fb.com>
@ezyang ezyang force-pushed the pr/convolution-weight-contiguous branch from ac8c90a to 3224c35 Compare January 9, 2018 14:53
Signed-off-by: Edward Z. Yang <ezyang@fb.com>
@yf225
Copy link
Contributor

yf225 commented Jan 9, 2018

@pytorchbot retest this please

2 similar comments
@yf225
Copy link
Contributor

yf225 commented Jan 9, 2018

@pytorchbot retest this please

@ezyang
Copy link
Contributor Author

ezyang commented Jan 10, 2018

@pytorchbot retest this please

@ezyang ezyang merged commit bc7a41a into pytorch:master Jan 10, 2018
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

5 participants