iRoPE varseq flag for pre-calculated kv qparams #4160
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
✅ Deploy Preview for pytorch-fbgemm-docs ready!
To edit notification comments on pull requests, go to your Netlify project configuration. |
|
This pull request was exported from Phabricator. Differential Revision: D73478483 |
Aya-ZIbra
added a commit
to Aya-ZIbra/FBGEMM
that referenced
this pull request
May 20, 2025
Summary: X-link: facebookresearch/FBGEMM#1240 as title. This is needed to handle this case: https://www.internalfb.com/diff/D73833204?dst_version_fbid=9500286030082255&transaction_fbid=676020828512263 This will help avoid amax calc in rope for decode and partial prefill batch lanes. Also, we can rely on it in Kernel2, to return back and avoid unneccessary quantization. Reviewed By: y-sq Differential Revision: D73478483
Aya-ZIbra
added a commit
to Aya-ZIbra/FBGEMM
that referenced
this pull request
May 20, 2025
Summary: X-link: facebookresearch/FBGEMM#1240 as title. This is needed to handle this case: https://www.internalfb.com/diff/D73833204?dst_version_fbid=9500286030082255&transaction_fbid=676020828512263 This will help avoid amax calc in rope for decode and partial prefill batch lanes. Also, we can rely on it in Kernel2, to return back and avoid unneccessary quantization. Reviewed By: y-sq Differential Revision: D73478483
Aya-ZIbra
force-pushed
the
export-D73478483
branch
from
88e480c to
705f27b
Compare
Aya-ZIbra
added a commit
to Aya-ZIbra/FBGEMM
that referenced
this pull request
May 20, 2025
Summary: X-link: facebookresearch/FBGEMM#1240 as title. This is needed to handle this case: https://www.internalfb.com/diff/D73833204?dst_version_fbid=9500286030082255&transaction_fbid=676020828512263 This will help avoid amax calc in rope for decode and partial prefill batch lanes. Also, we can rely on it in Kernel2, to return back and avoid unneccessary quantization. Reviewed By: y-sq Differential Revision: D73478483
Aya-ZIbra
force-pushed
the
export-D73478483
branch
from
705f27b to
3a04a01
Compare
|
This pull request was exported from Phabricator. Differential Revision: D73478483 |
Aya-ZIbra
force-pushed
the
export-D73478483
branch
from
3a04a01 to
ee77a32
Compare
Aya-ZIbra
added a commit
to Aya-ZIbra/FBGEMM
that referenced
this pull request
May 20, 2025
Summary: Pull Request resolved: pytorch#4160 X-link: facebookresearch/FBGEMM#1240 as title. This is needed to handle this case: https://www.internalfb.com/diff/D73833204?dst_version_fbid=9500286030082255&transaction_fbid=676020828512263 This will help avoid amax calc in rope for decode and partial prefill batch lanes. Also, we can rely on it in Kernel2, to return back and avoid unneccessary quantization. Reviewed By: y-sq Differential Revision: D73478483
Summary: Pull Request resolved: pytorch#4160 X-link: facebookresearch/FBGEMM#1240 as title. This is needed to handle this case: https://www.internalfb.com/diff/D73833204?dst_version_fbid=9500286030082255&transaction_fbid=676020828512263 This will help avoid amax calc in rope for decode and partial prefill batch lanes. Also, we can rely on it in Kernel2, to return back and avoid unneccessary quantization. Reviewed By: y-sq Differential Revision: D73478483
|
This pull request was exported from Phabricator. Differential Revision: D73478483 |
Aya-ZIbra
force-pushed
the
export-D73478483
branch
from
ee77a32 to
163169c
Compare
|
This pull request has been merged in c85ea56. |
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/1240
as title.
This is needed to handle this case: https://www.internalfb.com/diff/D73833204?dst_version_fbid=9500286030082255&transaction_fbid=676020828512263
This will help avoid amax calc in rope for decode and partial prefill batch lanes.
Also, we can rely on it in Kernel2, to return back and avoid unneccessary quantization.
Reviewed By: y-sq
Differential Revision: D73478483