scripts/run_evals.py

• TaskExecutionError · class · L48-L55 — class TaskExecutionError(RuntimeError)
• load_manifest · function · L83-L85 — def load_manifest(task_dir: Path) -> dict[str, Any]
• discover_tasks · function · L88-L104 — def discover_tasks(gold_root: Path = GOLD_ROOT) -> list[str]
• _accuracy · function · L110-L111 — def _accuracy(correct: int, total: int) -> float
• measure_citation_extraction · function · L114-L222 — def measure_citation_extraction(task_dir: Path, manifest: dict[str, Any]) -> dict[str, Any]
• measure_rq_framing_patterns · function · L228-L291 — def measure_rq_framing_patterns(task_dir: Path, manifest: dict[str, Any]) -> dict[str, Any]
• _pending_result · function · L297-L309 — def _pending_result(task_dir: Path, manifest: dict[str, Any], notice: str) -> dict[str, Any]
• run_task · function · L320-L365 — def run_task(task_name: str, gold_root: Path = GOLD_ROOT) -> dict[str, Any]
• _new_run_id · function · L371-L373 — def _new_run_id() -> str
• _gold_set_version · function · L376-L378 — def _gold_set_version(per_task: list[dict[str, Any]]) -> str
• _metric_entries · function · L381-L403 — def _metric_entries(task_result: dict[str, Any]) -> list[dict[str, Any]]
• _report_task_index · function · L406-L407 — def _report_task_index(report: dict[str, Any]) -> dict[str, dict[str, Any]]
• build_report · function · L410-L430 — def build_report(task_names: list[str], gold_root: Path = GOLD_ROOT) -> dict[str, Any]
• build_compare_report · function · L433-L476 — def build_compare_report(baseline: dict[str, Any], compare: dict[str, Any]) -> dict[str, Any]
• main · function · L482-L509 — def main(argv: list[str] | None = None) -> int