Upload 20260304_1431/checkpoint_eval_iter3000.json with huggingface_hub

20260304_1431/checkpoint_eval_iter3000.json

@@ -0,0 +1,72 @@
+{
+  "iteration": 3000,
+  "date": "2026-03-05",
+  "timestamp": "2026-03-05T02:05:42.341868",
+  "training": {
+    "lr": 3e-05,
+    "batch_size": 256,
+    "aux_loss_weight": 0.5,
+    "buffer_size": 10000,
+    "buffer_capacity": 10000,
+    "ema_decay": 0.999,
+    "recent_loss_total": 0.0067021773407689765,
+    "recent_loss_diffusion": 0.0063638939912838395,
+    "recent_loss_aux": 0.0006765666980734387,
+    "grad_steps_per_iteration": 50,
+    "episodes_per_iteration": 10
+  },
+  "eval_episodes_per_env": 20,
+  "id_winrate": 0.6,
+  "ood_winrate": 0.06666666666666667,
+  "per_env": {
+    "MiniHack-Room-Random-5x5-v0": {
+      "win_rate": 0.85,
+      "avg_reward": 18.601499999999994,
+      "avg_steps": 17.75,
+      "wins": 17,
+      "episodes": 20
+    },
+    "MiniHack-Room-Random-15x15-v0": {
+      "win_rate": 0.8,
+      "avg_reward": 20.3895,
+      "avg_steps": 47.55,
+      "wins": 16,
+      "episodes": 20
+    },
+    "MiniHack-Corridor-R2-v0": {
+      "win_rate": 0.3,
+      "avg_reward": 4.555999999999997,
+      "avg_steps": 140.9,
+      "wins": 6,
+      "episodes": 20
+    },
+    "MiniHack-MazeWalk-9x9-v0": {
+      "win_rate": 0.45,
+      "avg_reward": 8.184499999999996,
+      "avg_steps": 111.5,
+      "wins": 9,
+      "episodes": 20
+    },
+    "MiniHack-Room-Dark-15x15-v0": {
+      "win_rate": 0.2,
+      "avg_reward": 2.9609999999999985,
+      "avg_steps": 161.05,
+      "wins": 4,
+      "episodes": 20
+    },
+    "MiniHack-Corridor-R5-v0": {
+      "win_rate": 0.0,
+      "avg_reward": -2.7510000000000017,
+      "avg_steps": 200.0,
+      "wins": 0,
+      "episodes": 20
+    },
+    "MiniHack-MazeWalk-45x19-v0": {
+      "win_rate": 0.0,
+      "avg_reward": -2.3645000000000023,
+      "avg_steps": 200.0,
+      "wins": 0,
+      "episodes": 20
+    }
+  }
+}