Skip to content
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

Auto wrap modules in ZeRO 3 based on _no_split_modules as FSDP #3323

Open
wants to merge 1 commit into
base: main
Choose a base branch
from

Conversation

oraluben
Copy link
Contributor

@oraluben oraluben commented Jan 6, 2025

What does this PR do?

When using FSDP with auto_wrap, transformers will try to wrap modules defined in _no_split_modules. This can reduce overhead of small modules.

DeepSpeed have similar mechanism for ZeRO3, but need to be explicit called by user (set_z3_leaf_modules).
This PR automatically set leaf module based on _no_split_modules.

Some point that might need discussion:

  1. Should we only enable this when enabling torch.compile or not;
  2. how to disable this feature as there's no fsdp_config='auto_wrap' for this case

Related: microsoft/DeepSpeed#6683

Fixes # (issue)

Before submitting

  • This PR fixes a typo or improves the docs (you can dismiss the other checks if that's the case).
  • Did you read the contributor guideline,
    Pull Request section?
  • Was this discussed/approved via a Github issue or the forum? Please add a link
    to it if that's the case.
  • Did you make sure to update the documentation with your changes? Here are the
    documentation guidelines, and
    here are tips on formatting docstrings.
  • Did you write any new necessary tests?

Who can review?

Anyone in the community is free to review the PR once the tests have passed. Feel free to tag
members/contributors who may be interested in your PR.

@muellerzr

@oraluben oraluben changed the title Auto wrap modules based on _no_split_modules as FSDP Auto wrap modules in ZeRO 3 based on _no_split_modules as FSDP Jan 6, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

1 participant