Add option to unfuse Wqkv #1367
Merged
+220
−14
Conversation
This file contains 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
irenedea
reviewed
Jul 16, 2024
dakinggg
reviewed
Jul 16, 2024
…h/unfuse_wqkv merged
dakinggg
reviewed
Jul 17, 2024
There was a problem hiding this comment.
lgtm, would like @vchiley or @ShashankMosaicML to take a look before approving
|
Added testing for fuse splits into the unit test @dakinggg |
dakinggg
approved these changes
Jul 17, 2024
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Outdated suggestions cannot be applied.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.
Adding an option to unfuse Wqkv. Gives us flexibility to support various other model architectures that don't fuse Wqkv like we do.
Run with fused Wqkv (main): fused-wqkv-orig-n7eky6
Run with fused Wqkv (this branch): fused-wqkv-new-ggXoVL
Run without fused Wqkv: unfused-wqkv-new-a2Qo2r
Same loss curves:
Same throughput for fused Wqkv before and after, slightly lower throughput for unfused Wqkv (as expected):
