scripts/evaluate_policy_ablations.py

#!/usr/bin/env python3
"""Run policy-stack ablations independently from GRPO training."""

from __future__ import annotations

import argparse
import json
from pathlib import Path

import sys

ROOT = Path(__file__).resolve().parents[1]
if str(ROOT) not in sys.path:
    sys.path.insert(0, str(ROOT))

from app.training.grpo_experiment import run_policy_stack_rollout


def parse_args() -> argparse.Namespace:
    parser = argparse.ArgumentParser(description="Run policy-stack ablations.")
    parser.add_argument("--episodes", type=int, default=6)
    parser.add_argument("--output", default="outputs/reports/grpo_ablation_report.json")
    parser.add_argument("--checkpoint-dir", default="checkpoints")
    return parser.parse_args()


def main() -> None:
    args = parse_args()
    root = Path(__file__).resolve().parents[1]
    checkpoint_dir = root / args.checkpoint_dir
    checkpoint_dir.mkdir(parents=True, exist_ok=True)

    payload = {
        "status": "ok",
        "ablations": {
            "bandit_only": run_policy_stack_rollout(
                "bandit-only",
                episodes=args.episodes,
                checkpoint_dir=checkpoint_dir,
                seed_offset=1_200,
            ),
            "llm_only": run_policy_stack_rollout(
                "llm-only",
                episodes=args.episodes,
                checkpoint_dir=checkpoint_dir,
                seed_offset=2_200,
            ),
            "llm_bandit": run_policy_stack_rollout(
                "llm+bandit",
                episodes=args.episodes,
                checkpoint_dir=checkpoint_dir,
                seed_offset=3_200,
            ),
        },
    }

    output_path = root / args.output
    output_path.parent.mkdir(parents=True, exist_ok=True)
    output_path.write_text(json.dumps(payload, ensure_ascii=True, indent=2), encoding="utf-8")
    print("policy_ablations_done")


if __name__ == "__main__":
    main()