Repository navigation
fix: validate classifier answers and score partial batches - #187
Merged
Merged
Conversation
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Stacked on #185 so this PR includes the shared CI prerequisites. Review/merge #185 first, then retarget this PR to
main.A classifier batch containing fewer items than
n_examples_shownis currently rejected even when the model returns the correct number of predictions. Parse against the actual batch length, including log probabilities and failure outputs.Validate that JSON answers contain only binary integers or booleans, instead of silently casting values such as
2or"0"to True. Match exact binary tokens when extracting probabilities so alternatives such as10do not contribute to P(1). Reject nonpositive batch sizes.Validation: seven focused classifier cases pass independently, covering five malformed answers, a partial final batch, and an exact-token probability calculation. The combined fixes pass 49 focused tests; Ruff passes. The CPU launcher bypasses only the eager vLLM initializer; no live model requests were made.
The pre-existing Actions collection failure is fixed separately by #185.