Skip to content

Improve normalization #580

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
Merged
merged 22 commits into from
Feb 15, 2023
Merged

Improve normalization #580

merged 22 commits into from
Feb 15, 2023

Conversation

@rocking5566
Copy link
Collaborator

@rocking5566 rocking5566 commented Feb 10, 2023 •
edited
Loading

This PR improve performance of normalization kernel
Improve fp32 at least 15%

What I have done in this PR

  1. Add more instances (tune kernel parameter)
  2. Separate the pipeline of sweep once from normal pipeline, so that I could move the buffer_load(gamma) in advance to hide the latency of welford
  3. Support naive variance for normalization

@rocking5566 rocking5566 removed the WIP label Feb 15, 2023
@rocking5566 rocking5566 changed the title Improve layernorm Improve normalization Feb 15, 2023
@asroy asroy merged commit 6a6163a into develop Feb 15, 2023
@illsilin illsilin deleted the improve_layernorm branch December 7, 2023 18:44
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

@qianfengz qianfengz Awaiting requested review from qianfengz qianfengz is a code owner

@carlushuang carlushuang Awaiting requested review from carlushuang carlushuang is a code owner

1 more reviewer

@asroy asroy asroy approved these changes

Reviewers whose approvals may not affect merge requirements

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

3 participants

@rocking5566 @asroy