[JAX] Wrap triton_call custom call using the FFI. #345
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
![copybara-service[bot]](https://avatars.githubusercontent.com/in/44061?s=60&v=4){kind=small}
copybara-service
bot
force-pushed
the
test_766642128
branch
3 times, most recently
from
ad63665 to
e2b8b1a
Compare
We've been putting off porting the `triton_call` kernel to the FFI because it's possible that we could make better use of the FFI features to improve the design. However, XLA is planning to remove support for legacy custom calls in a few months, so we need to update to the newer API. As a stopgap solution, this change wraps the existing kernel so that it can be directly called using the new API version without any major refactoring. This will support XLA's deprecation plan without holding them up with our design discussions! PiperOrigin-RevId: 766642128
copybara-service
bot
force-pushed
the
test_766642128
branch
from
e2b8b1a to
66c5ab6
Compare
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Outdated suggestions cannot be applied.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.
[JAX] Wrap triton_call custom call using the FFI.
We've been putting off porting the
triton_callkernel to the FFI because it's possible that we could make better use of the FFI features to improve the design. However, XLA is planning to remove support for legacy custom calls in a few months, so we need to update to the newer API. As a stopgap solution, this change wraps the existing kernel so that it can be directly called using the new API version without any major refactoring. This will support XLA's deprecation plan without holding them up with our design discussions!