Support splitting physical axis in SPMD mesh#8698
Merged
Conversation
Previously, you cannot create a logical [64, 4] mesh that is backed by a physical [16, 16] torus. That requires splitting one physical axis to support two logical axes. This is now supported by copying a bunch of code from JAX, updating the copies we had before.
Collaborator
Author
|
I couldn't add a test because there's no large enough TPUs in CI to demonstrate breaking a physical axis. But I did cloud build this branch and tested the docker with torchprime. |
qihqi
approved these changes
Feb 14, 2025
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Previously, you cannot create a logical [64, 4] mesh that is backed by a physical [16, 16] torus. That requires splitting one physical axis to support two logical axes.
This is now supported by copying a bunch of code from JAX, updating the copies we had before.
Fixes #8381