Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

speeding up batch normalization #7

Merged
merged 4 commits into from Jun 15, 2015

Conversation

okuta
Copy link
Member

@okuta okuta commented Jun 8, 2015

This is speedup of batch normalization.

@unnonouno
Copy link
Member

It looks good for me except a comment.

@okuta
Copy link
Member Author

okuta commented Jun 12, 2015

shape[0] is batch size. I expect loop unrolling by compiler, if batch size is small.
In most cases batch size is constant. Therefore, compiling the kernel only once in this code.

@unnonouno
Copy link
Member

!! Cool idea!! So, please write a comment about loop unrolling.

@okuta
Copy link
Member Author

okuta commented Jun 15, 2015

Thank you for your advice.
I added comments.

@unnonouno
Copy link
Member

💯
LGTM

unnonouno added a commit that referenced this pull request Jun 15, 2015
@unnonouno unnonouno merged commit b3d9d8b into chainer:master Jun 15, 2015
@unnonouno
Copy link
Member

Sorry, please hide cuda.cutools here:
https://github.com/pfnet/chainer/pull/7/files#diff-54bc089963209c5c3d69831ae45dd716R27

When CUDA is not enabled, it fails to import batch_normalization.

@okuta okuta modified the milestone: v1.0.1 Jun 23, 2015
@okuta okuta deleted the speeding_up_batch_normalization branch October 2, 2015 05:50
niboshi pushed a commit to niboshi/chainer that referenced this pull request May 31, 2017
hvy pushed a commit that referenced this pull request May 21, 2019
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

3 participants