{
  "level": "high",
  "problem": 3,
  "stage": 1,
  "stage_name": "s1-single",
  "algorithm": "maskable_ppo",
  "feature_version": 4,
  "obs_dim": 1248,
  "act_dim": 120,
  "seed": 320001,
  "requested_timesteps": 2048,
  "actual_timesteps": 2048,
  "n_envs": 1,
  "vec_env": "dummy",
  "config_profile": "production",
  "fidelity": "production",
  "local_selector": "geometry",
  "local_model": null,
  "device": "mps",
  "wall_seconds": 250.817,
  "max_wall_s": 800.0,
  "created_unix_s": 1789127731.636131,
  "python": "3.13.9",
  "platform": "macOS-27.0-arm64-arm-64bit-Mach-O",
  "torch": "2.11.0",
  "numpy": "2.5.3",
  "stable_baselines3": "2.9.0",
  "sb3_contrib": "2.9.0",
  "mp_actor": true,
  "model_sha256": "5884cf4a249f8f8c93c4fefbd0d514d3eacceb90a2ffacf173c35c89f790b604",
  "monitor_files": [
    "output/hrl/p1-smoke/high_curve_q3_env0.monitor.csv"
  ],
  "reward_note": "HierarchicalEnv 当前使用 HRLReward 的时序奖励；P0.4 纯时间口径在消融/训练配置中显式记录，不静默替换。",
  "evaluation": {
    "episodes": 2,
    "successes": 2,
    "success_rate": 1.0,
    "rows": [
      {
        "episode": 0,
        "seed": 1320001,
        "steps": 9,
        "cleared": 1,
        "accounted": 20,
        "virtual_time_s": 2655.664582,
        "stop_reason": null,
        "anomalies": [],
        "success": true
      },
      {
        "episode": 1,
        "seed": 1320002,
        "steps": 9,
        "cleared": 1,
        "accounted": 20,
        "virtual_time_s": 3140.481152,
        "stop_reason": null,
        "anomalies": [],
        "success": true
      }
    ]
  },
  "eval_manifest": null
}