114514

2026-09-11 21:00:39 +08:00
parent 2a3ead5d3d
commit 1725cd0891
168 changed files with 170687 additions and 874 deletions
@@ -0,0 +1,4 @@
+#{"t_start": 1789127784.936411, "env_id": "None"}
+r,l,t
+0.132297,9,6.497756
+-0.015601,9,15.666472
@@ -0,0 +1,45 @@
+{
+  "level": "high",
+  "problem": 3,
+  "stage": 1,
+  "stage_name": "s1-single",
+  "algorithm": "maskable_ppo",
+  "feature_version": 4,
+  "action_layout_version": 1,
+  "reward_version": 1,
+  "obs_dim": 1248,
+  "act_dim": 120,
+  "seed": 2,
+  "requested_timesteps": 32,
+  "actual_timesteps": 32,
+  "n_envs": 1,
+  "vec_env": "dummy",
+  "config_profile": "production",
+  "fidelity": "production",
+  "reward_profile": "time",
+  "local_selector": "geometry",
+  "local_model": null,
+  "device": "mps",
+  "wall_seconds": 15.971,
+  "max_wall_s": 120.0,
+  "created_unix_s": 1789127801.8324401,
+  "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": "ccd89a54a9735ff550fd5692e1375441dc668f9c5c2bb496193f3bf616f0263e",
+  "monitor_files": [
+    "output/hrl/p1-tiny-time/high_curve_q3_env0.monitor.csv"
+  ],
+  "reward_note": "reward_profile=time 使用 HRLReward.time_only（P0.4 纯时间口径）；reward_profile=prompt 使用原始 §5.2 塑形。模型元数据已固化口径。",
+  "evaluation": {
+    "episodes": 0,
+    "successes": 0,
+    "success_rate": null,
+    "rows": []
+  },
+  "eval_manifest": null
+}