Optimize kv cache usage for yoco #4030
Closed
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
This pull request was exported from Phabricator. Differential Revision: D73570737 |
✅ Deploy Preview for pytorch-fbgemm-docs ready!
To edit notification comments on pull requests, go to your Netlify site configuration. |
ghjeong12
force-pushed
the
export-D73570737
branch
from
ac05718 to
b51e730
Compare
ghjeong12
pushed a commit
to ghjeong12/FBGEMM-1
that referenced
this pull request
Apr 28, 2025
Summary: X-link: facebookresearch/FBGEMM#1114 This diff aims to remove redundant temporary tensors and KV caches when using yoco. This includes changes for allowing optional xk, xv for rope/nope kernels, adding flag for checking to update kv cache or not, etc. Reviewed By: Aya-ZIbra Differential Revision: D73570737
ghjeong12
pushed a commit
to ghjeong12/FBGEMM-1
that referenced
this pull request
Apr 28, 2025
Summary: X-link: facebookresearch/FBGEMM#1114 This diff aims to remove redundant temporary tensors and KV caches when using yoco. This includes changes for allowing optional xk, xv for rope/nope kernels, adding flag for checking to update kv cache or not, etc. Reviewed By: Aya-ZIbra Differential Revision: D73570737
ghjeong12
force-pushed
the
export-D73570737
branch
from
b51e730 to
14c849d
Compare
|
This pull request was exported from Phabricator. Differential Revision: D73570737 |
ghjeong12
force-pushed
the
export-D73570737
branch
from
14c849d to
d3875fc
Compare
ghjeong12
pushed a commit
to ghjeong12/FBGEMM-1
that referenced
this pull request
Apr 28, 2025
Summary: Pull Request resolved: pytorch#4030 X-link: facebookresearch/FBGEMM#1114 This diff aims to remove redundant temporary tensors and KV caches when using yoco. This includes changes for allowing optional xk, xv for rope/nope kernels, adding flag for checking to update kv cache or not, etc. Reviewed By: Aya-ZIbra Differential Revision: D73570737
Summary: Pull Request resolved: pytorch#4030 X-link: facebookresearch/FBGEMM#1114 This diff aims to remove redundant temporary tensors and KV caches when using yoco. This includes changes for allowing optional xk, xv for rope/nope kernels, adding flag for checking to update kv cache or not, etc. Reviewed By: Aya-ZIbra Differential Revision: D73570737
|
This pull request was exported from Phabricator. Differential Revision: D73570737 |
ghjeong12
force-pushed
the
export-D73570737
branch
from
d3875fc to
93a6361
Compare
|
This pull request has been merged in 517b73d. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Summary:
X-link: https://github.com/facebookresearch/FBGEMM/pull/1114
This diff aims to remove redundant temporary tensors and KV caches when using yoco. This includes changes for allowing optional xk, xv for rope/nope kernels, adding flag for checking to update kv cache or not, etc.
Reviewed By: Aya-ZIbra
Differential Revision: D73570737