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.

Sign up for GitHub

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

Jump to bottom

Some small optimizations for conv2d on cpu #404

Merged

coreylowman merged 1 commit into main from conv2d-optims

Jan 26, 2023

Owner

coreylowman commented Jan 25, 2023

This PR updates CPU conv2d implementation to always iterate the patches buffers in order and with explicit loops. Prior behavior was to:

Use strided indexing in forward, which isn't as fast as just using the loops here, probably thanks to inlining
Change loop order in backward pass. Prior it mirror the loop order of forward to make the logic identical. Now it mirrors the cuda kernel's logic


          Some small optimizations for conv2d

909801c

coreylowman merged commit 1211cd3 into main

coreylowman deleted the conv2d-optims branch

January 26, 2023 15:15

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet