Repository navigation
chore: release 0.14.0 - #248
Merged
Merged
Conversation
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
Merged
7 tasks done
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Type
Summary
runner_config.stop_afterand a model'slifecycle, agrader/extractormay name amodels:entry thatsieval runlaunches, a new exit code 6 for an engine lost mid-run, and new--dry-runfieldsbarriers,phasesandlaunches(fix(cli): fail fast when an engine sieval run launched dies mid-run #246, feat: stage barriers and model lifecycles #247).IDENTITY_SCHEMAstays 4 andRESULT_SCHEMA_VERSIONstays 1. Barriers and lifecycles are scheduling only;engine_died,engine_died_unsentandbarrier_iterationare newerror_reasonvalues;runs[].stopped_afterandruns[].serving.lifecyclesare additive. 0.13.x result directories resume.Related Issues
Refs #245
Test Plan
Automated
ruff check .,ruff format --check .)ty check)pytest tests/unit tests/integration --cov --cov-fail-under=95 -n 8: 9971 passed, 1 skipped, coverage 97.15%.test_exec_quotebench_anchor.pyfails on a macOS host, as onv0.13.2(it compares against GNU verdicts). Threetest_wikisql.pytests fail only when the shell has no locale set; withLANG=en_US.UTF-8they pass.pytest tests/acceptance -m "not benchmark" -s: 11 passed, 10 skipped (staged BFCL archives and the code-evaluator service are not available locally)pytest -m benchmark -s: all scenarios PASS once host load fell, four runs in a row (last: single-turn 558.5 SPS / 87.3%, long output 156.7 / 73.4%, multi-iteration 180.2 / 84.5%, high concurrency 1078.9 / 84.3%, framework overhead 1677.9 SPS, resume 0.5 s), matchingv0.13.2run back to back on the same host (1079.6 / 84.3%, 1688.6 SPS). Earlier runs under host load failed single-turn at 61.0% and, once in three, high concurrency at 61.0%.pre-commit run -a: all hooks passcheck_preflight.py --level deep: all pass exceptcheck_version(latest tag 0.13.2 vs CHANGELOG 0.14.0, which resolves whenv0.14.0is tagged after merge) and link warnings for unreachable or not-yet-existing URLsChecklist
Required (all PRs)
type(scope): description)AI-Generated Code - <model> (<provider>)in module docstringcore/🤖 Generated with Claude Code