| """CLI: compare style detectors on the realism corpus (debug).""" |
|
|
| from __future__ import annotations |
|
|
| import json |
| import logging |
| import sys |
| from datetime import datetime, timezone |
| from pathlib import Path |
|
|
| ROOT = Path(__file__).resolve().parents[1] |
| sys.path.insert(0, str(ROOT)) |
|
|
| from app.style_eval import run_style_detector_eval |
|
|
| logging.basicConfig( |
| level=logging.INFO, |
| format="%(asctime)s %(levelname)s [%(name)s] %(message)s", |
| ) |
| logger = logging.getLogger("run_style_debug_eval") |
|
|
|
|
| def main() -> int: |
| count = int(sys.argv[1]) if len(sys.argv) > 1 else 12 |
| out_dir = Path(__file__).resolve().parent / "out" |
| out_dir.mkdir(parents=True, exist_ok=True) |
| from app.api import _settings_from_db |
|
|
| settings = _settings_from_db() |
| logger.info("starting style detector compare count_per_class=%s", count) |
| result = run_style_detector_eval(count_per_class=count, settings=settings) |
| stamp = datetime.now(timezone.utc).strftime("%Y%m%dT%H%M%SZ") |
| latest = out_dir / "style_debug_eval_latest.json" |
| stamped = out_dir / f"style_debug_eval_{stamp}.json" |
| text = json.dumps(result, indent=2) |
| latest.write_text(text, encoding="utf-8") |
| stamped.write_text(text, encoding="utf-8") |
| oc = result.get("overall_conclusion") or {} |
| print( |
| f"decision={oc.get('decision')} best={oc.get('best_detector')} " |
| f"wrote={latest}" |
| ) |
| for report in result.get("reports") or []: |
| t = (report.get("conclusion") or {}).get("typical_metrics") or {} |
| print( |
| f" {report.get('detector_id')}: {report.get('conclusion', {}).get('decision')} " |
| f"P={t.get('precision')} R={t.get('recall')} F1={t.get('f1')} " |
| f"animeFP={t.get('anime_false_positive_rate')} " |
| f"uncertain={report.get('conclusion', {}).get('uncertain_count')}" |
| ) |
| return 0 |
|
|
|
|
| if __name__ == "__main__": |
| raise SystemExit(main()) |
|
|
