[negvet]

Description

Please include a brief summary of the changes, relevant motivation and context.

Fixes # (issue)

Type of change

• Documentation change (change only to the documentation, either a fix or a new content)
• Bug fix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• Infra/Build change
• Code refactoring

Changes

Please list the changes introduced in this PR:

• Change A
• Change B

Checklist:

• I have read and followed the contributing guidelines
• The functionality is complete
• I have commented my code, particularly in hard-to-understand areas
• I have made corresponding changes to the documentation
• My changes generate no new warnings
• I have added tests that prove my fix is effective or that my feature works
• New and existing unit tests pass locally with my changes

[negvet]

/te-ci pytorch

[greptile-apps]

Confidence Score: 5/5

[Low risk] Test file adjusts checkpoint serialization logic.

The PR appears safe to merge based on the changes since the previous review.

Summary

The PR updates the FSDP2 hybrid checkpoint parity test to exclude variable-sized _extra_state entries from the saved and loaded model state, then loads the remaining state non-strictly. The changes since the previous review are a merge from main; they do not change this PR’s test patch.

_{Reviews (2) · Last reviewed commit: "Merge branch 'main' into fix_extra_state..."}

[pggPL]

LGTM