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.

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

Always scatter in FP32 in TPConv #4260

Merged
merged 1 commit into from
Mar 20, 2024
Merged

Always scatter in FP32 in TPConv #4260

merged 1 commit into from
Mar 20, 2024

Conversation

tingyu66
Copy link
Member

@tingyu66 tingyu66 commented Mar 19, 2024 •
edited
Loading

Always scatter in FP32 in TPConv due to poor performance of torch's native scatter_add in bfloat16.

@tingyu66 tingyu66 requested a review from a team as a code owner March 19, 2024 23:20
@copy-pr-bot copy-pr-bot
Copy link

copy-pr-bot bot commented Mar 19, 2024

This pull request requires additional validation before any workflows can run on NVIDIA's runners.

Pull request vetters can view their responsibilities here.

Contributors can view more details about this message here.

@tingyu66 tingyu66 self-assigned this Mar 19, 2024
@tingyu66 tingyu66 added improvement Improvement / enhancement to an existing function non-breaking Non-breaking change labels Mar 19, 2024
@tingyu66
Copy link
Member Author

/ok to test

alexbarghi-nv
alexbarghi-nv approved these changes Mar 20, 2024
View reviewed changes
Copy link
Member

@alexbarghi-nv alexbarghi-nv left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

👍

@alexbarghi-nv
Copy link
Member

/merge

@rapids-bot rapids-bot bot merged commit 4becbe8 into rapidsai:branch-24.04 Mar 20, 2024
137 checks passed
@tingyu66 tingyu66 deleted the fp32-scatter branch April 5, 2024 02:04
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@alexbarghi-nv alexbarghi-nv alexbarghi-nv approved these changes

Assignees

@tingyu66 tingyu66

Labels
improvement Improvement / enhancement to an existing function non-breaking Non-breaking change python
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
2 participants
@tingyu66 @alexbarghi-nv