Repository navigation
Re-grade workflow: 4 evaluator workers by default, log memory - #252
Merged
Merged
Conversation
The first full re-grade run lost all ten of its first jobs: each runner got a shutdown signal about two minutes into grading with 16 worker threads, while the PR evaluation of the same files succeeds. The worker count is now an input (default 4), progress is logged as in the PR evaluation, and memory and swap use are printed every 30 seconds so a failure shows its cause. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Fixes the re-grade workflow from #244. Its first full run (run 40967618215) lost all 10 of its first jobs.
What happened: each runner received a shutdown signal (exit 143) about 2 minutes into grading, with 16 worker threads. No Python error was logged, and every job died at the same stage regardless of router. The PR evaluation workflow grades the same files on the same runner type and succeeds. Running out of memory is the likely cause, but the logs couldn't confirm it, because
--save-interval 0also silenced the progress lines.Change:
num_workers, defaulting to 4.I'll test it on one router before re-running all 25.
🤖 Generated with Claude Code