-
Notifications
You must be signed in to change notification settings - Fork 528
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
Refactor FP8 grouped GEMM with dynamic and static versions #3561
base: main
Are you sure you want to change the base?
Conversation
✅ Deploy Preview for pytorch-fbgemm-docs ready!
To edit notification comments on pull requests, go to your Netlify site configuration. |
This pull request was exported from Phabricator. Differential Revision: D68004072 |
4208eac
to
27d65f1
Compare
This pull request was exported from Phabricator. Differential Revision: D68004072 |
Summary: Pull Request resolved: pytorch#3560 X-link: facebookresearch/FBGEMM#646 This Diff supports FP8 grouped GEMM with rowwise scaling for MoE, and replaces the existing tensorwise with rowwise scaling to achieve better accuracy with similar performance Differential Revision: D67806685 Reviewed By: jwfromm
27d65f1
to
379880d
Compare
) Summary: Pull Request resolved: pytorch#3561 X-link: facebookresearch/FBGEMM#647 Refactor FP8 grouped GEMM with dynamic and static versions to unify CUTLASS and CK FP8 grouped GEMM in fbgemm Reviewed By: jwfromm Differential Revision: D68004072
This pull request was exported from Phabricator. Differential Revision: D68004072 |
379880d
to
fdf78d4
Compare
) Summary: Pull Request resolved: pytorch#3561 X-link: facebookresearch/FBGEMM#647 Refactor FP8 grouped GEMM with dynamic and static versions to unify CUTLASS and CK FP8 grouped GEMM in fbgemm Reviewed By: jwfromm Differential Revision: D68004072
This pull request was exported from Phabricator. Differential Revision: D68004072 |
) Summary: Pull Request resolved: pytorch#3561 X-link: facebookresearch/FBGEMM#647 Refactor FP8 grouped GEMM with dynamic and static versions to unify CUTLASS and CK FP8 grouped GEMM in fbgemm Reviewed By: jwfromm Differential Revision: D68004072
fdf78d4
to
05a1bdc
Compare
This pull request was exported from Phabricator. Differential Revision: D68004072 |
1 similar comment
This pull request was exported from Phabricator. Differential Revision: D68004072 |
) Summary: Pull Request resolved: pytorch#3561 X-link: facebookresearch/FBGEMM#647 Refactor FP8 grouped GEMM with dynamic and static versions to unify CUTLASS and CK FP8 grouped GEMM in fbgemm Reviewed By: jwfromm Differential Revision: D68004072
05a1bdc
to
66def49
Compare
) Summary: Pull Request resolved: pytorch#3561 X-link: facebookresearch/FBGEMM#647 Refactor FP8 grouped GEMM with dynamic and static versions to unify CUTLASS and CK FP8 grouped GEMM in fbgemm Reviewed By: jwfromm Differential Revision: D68004072
66def49
to
fa9e2e5
Compare
This pull request was exported from Phabricator. Differential Revision: D68004072 |
) Summary: Pull Request resolved: pytorch#3561 X-link: facebookresearch/FBGEMM#647 Refactor FP8 grouped GEMM with dynamic and static versions to unify CUTLASS and CK FP8 grouped GEMM in fbgemm Reviewed By: jwfromm Differential Revision: D68004072
fa9e2e5
to
154c9ad
Compare
This pull request was exported from Phabricator. Differential Revision: D68004072 |
Summary: Refactor FP8 grouped GEMM with dynamic and static versions to unify CUTLASS and CK FP8 grouped GEMM in fbgemm
Differential Revision: D68004072