Skip to content

Add integration test for checkpoint resharding#3800

Open
shralex wants to merge 11 commits intomainfrom
add_resharding_integration_test
Open

Add integration test for checkpoint resharding#3800
shralex wants to merge 11 commits intomainfrom
add_resharding_integration_test

Conversation

@shralex
Copy link
Copy Markdown
Collaborator

@shralex shralex commented May 2, 2026

Adds a checkpoint resharding test - trains a model with a specific sharding strategy and saves a checkpoint.
Then train again by restoring this checkpoint using a different sharding strategy. Finally, asserts that the learning metrics are consistent, ensuring that checkpoints can be successfully loaded across different sharding strategies.

If the change fixes a bug or a Github issue, please include a link, e.g.,:
FIXES: b/504742746

Tests

This PR adds a test.

Checklist

Before submitting this PR, please make sure (put X in square brackets):

  • I have performed a self-review of my code. For an optional AI review, add the gemini-review label.
  • I have necessary comments in my code, particularly in hard-to-understand areas.
  • I have run end-to-end tests tests and provided workload links above if applicable.
  • I have made or will make corresponding changes to the doc if needed, including adding new documentation pages to the relevant Table of Contents (toctree directive) as explained in our documentation.

@shralex shralex requested a review from parambole as a code owner May 2, 2026 22:11
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