Validate canonical checkpoint optimizer counters - #992
Merged
Merged
Conversation
bradhilton
had a problem deploying
to
trainer-rank-gpu-validation
September 26, 2026 17:52 — with 
GitHub Actions
Failure
GitHub Actions
Failure
bradhilton
marked this pull request as draft
September 26, 2026 18:08
bradhilton
deployed
to
trainer-rank-gpu-validation
September 26, 2026 18:08 — with
GitHub Actions
Active
GitHub Actions
Active
bradhilton
marked this pull request as ready for review
September 26, 2026 18:30
bradhilton
deployed
to
trainer-rank-gpu-validation
September 26, 2026 18:30 — with
GitHub Actions
Active
GitHub Actions
Active
This was referenced Sep 26, 2026
This branch was successfully deployed
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
A checkpoint with an authenticated manifest can currently carry a negative, fractional, infinite, or NaN optimizer step into canonical optimizer restore. Validate the counter domain before reading optimizer state: require nonnegative integers or finite integral floats, preserving legacy integer-valued floats and the existing boolean rejection.
Adds public safetensors fixtures covering local and artifact-staged reads. The tests regenerate the manifest digest so they exercise semantic validation; no private checkpoint metadata or tensors are included.
Validation: 14 new cases failed before the fix. All 29 focused cases now pass, including valid counters, legacy format, optimizer mapping, and weights-only behavior. Ruff, formatting, and diff checks pass. Local type checking reports the same six preexisting slot-type diagnostics as the exact baseline; CI will run the complete environment.