Upload generalization/20260304_1431/checkpoint_eval_iter8000.json with huggingface_hub

generalization/20260304_1431/checkpoint_eval_iter8000.json

@@ -0,0 +1,72 @@
+{
+  "iteration": 8000,
+  "date": "2026-03-05",
+  "timestamp": "2026-03-05T19:56:01.381383",
+  "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": 0.004037455272295629,
+    "recent_loss_diffusion": 0.003938834615365194,
+    "recent_loss_aux": 0.00019724131360162572,
+    "grad_steps_per_iteration": 50,
+    "episodes_per_iteration": 10
+  },
+  "eval_episodes_per_env": 20,
+  "id_winrate": 0.6375000000000001,
+  "ood_winrate": 0.09999999999999999,
+  "per_env": {
+    "MiniHack-Room-Random-5x5-v0": {
+      "win_rate": 1.0,
+      "avg_reward": 22.07,
+      "avg_steps": 3.0,
+      "wins": 20,
+      "episodes": 20
+    },
+    "MiniHack-Room-Random-15x15-v0": {
+      "win_rate": 0.7,
+      "avg_reward": 16.985000000000003,
+      "avg_steps": 65.8,
+      "wins": 14,
+      "episodes": 20
+    },
+    "MiniHack-Corridor-R2-v0": {
+      "win_rate": 0.4,
+      "avg_reward": 6.8424999999999985,
+      "avg_steps": 121.7,
+      "wins": 8,
+      "episodes": 20
+    },
+    "MiniHack-MazeWalk-9x9-v0": {
+      "win_rate": 0.45,
+      "avg_reward": 8.560499999999998,
+      "avg_steps": 111.65,
+      "wins": 9,
+      "episodes": 20
+    },
+    "MiniHack-Room-Dark-15x15-v0": {
+      "win_rate": 0.25,
+      "avg_reward": 4.096999999999999,
+      "avg_steps": 150.95,
+      "wins": 5,
+      "episodes": 20
+    },
+    "MiniHack-Corridor-R5-v0": {
+      "win_rate": 0.05,
+      "avg_reward": -1.7025000000000023,
+      "avg_steps": 191.55,
+      "wins": 1,
+      "episodes": 20
+    },
+    "MiniHack-MazeWalk-45x19-v0": {
+      "win_rate": 0.0,
+      "avg_reward": -2.904000000000002,
+      "avg_steps": 200.0,
+      "wins": 0,
+      "episodes": 20
+    }
+  }
+}