-
Notifications
You must be signed in to change notification settings - Fork 688
Preserve SDPA for CoreML #5258
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
Preserve SDPA for CoreML #5258
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/5258
Note: Links to docs will display an error until the docs builds have been completed. ✅ No FailuresAs of commit b552716 with merge base de30572 ( This comment was automatically generated by Dr. CI and updates every 15 minutes. |
d464b45
to
e017e99
Compare
e017e99
to
cbcc3e9
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thanks! It’s a a lot cleaner to use custom op. Eventually we can use the new api
@cccclai has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator. |
@cccclai has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator. |
1 similar comment
@cccclai has imported this pull request. If you are a Meta employee, you can view this diff on Phabricator. |
Motivation
Starting from iOS18, CoreML has added SDPA op, so there is no longer need to decompose torch SDPA
Solution
Following #3483, add
ops_not_to_decompose
in CoreML partitioner, then useto_edge_transform_and_lower
API in llama export