eval_healthbench: skip when --output already exists - #246
Merged
Conversation
stage 4 runs 4 healthbench configs in sequence; on resume after preempt the wrapper short-circuits but a direct invocation would re-run a config whose JSON already landed (~10-15 min wasted per config). bake the idempotency check into the script so it protects direct invocations and future wrapper changes. empty (size 0) files are still overwritten in case a previous run crashed mid-write.
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
run_eval()wrapper intpu/run_pipeline.shalready skips configs whose JSON exists, but the underlyingeval_healthbench.pydid not. direct invocations or future wrapper changes would re-do ~10-15 min of inference + grading.--outputis already a non-empty file, right afterargs.outputis expanded. empty (size 0) files are treated as partial writes from a crashed run and are allowed to be overwritten.tests/test_eval_healthbench_skip.pycovering both the non-empty skip path (rc==0, mtime unchanged) and the empty-file no-skip path.Test plan
python3 -m py_compile scripts/eval_healthbench.py/tmp/preflight-py311/bin/python -m pytest tests/ -x -q— 187 passed, 1 skipped/tmp/preflight-py311/bin/python -m pytest tests/test_eval_healthbench_skip.py -v— both new tests pass