{
  "condition": "sdil",
  "config": {
    "action_dim": 7,
    "context_gain": 1.0,
    "hidden_layers": 2,
    "input_dim": 984,
    "learning_rate": 0.01,
    "mission_dim": 13,
    "momentum": 0.9,
    "reciprocal_learning_rate": 0.01,
    "weight_decay": 0.0001,
    "width": 256
  },
  "data": {
    "color_cardinality": 6,
    "elapsed_seconds": 37.730633020401,
    "env_id": "BabyAI-GoToObjS6-v1",
    "minigrid_version": "3.1.0",
    "object_cardinality": 11,
    "protocol": "babyai_shared_feedback_b0",
    "rollout_episodes": 500,
    "rollout_seed_start": 200000,
    "state_cardinality": 3,
    "train_episodes": 20000,
    "train_steps": 70159,
    "validation_episodes": 2000,
    "validation_steps": 7019,
    "vocabulary": [
      "<unk>",
      "ball",
      "blue",
      "box",
      "go",
      "green",
      "grey",
      "key",
      "purple",
      "red",
      "the",
      "to",
      "yellow"
    ]
  },
  "epoch_history": [
    {
      "epoch": 1,
      "train_loss": 0.7169609726017172
    },
    {
      "epoch": 2,
      "train_loss": 0.37284291326999663
    },
    {
      "epoch": 3,
      "train_loss": 0.17600450354543598
    },
    {
      "epoch": 4,
      "train_loss": 0.29167314236814323
    },
    {
      "epoch": 5,
      "train_loss": 0.0823398195952177
    },
    {
      "epoch": 6,
      "train_loss": 0.05786021038889885
    },
    {
      "epoch": 7,
      "train_loss": 0.04764524250545285
    },
    {
      "epoch": 8,
      "train_loss": 0.47599790914492174
    },
    {
      "epoch": 9,
      "train_loss": 0.27363136577335273
    },
    {
      "epoch": 10,
      "train_loss": 0.18569429102269086
    },
    {
      "epoch": 11,
      "train_loss": 0.1496148065545342
    },
    {
      "epoch": 12,
      "train_loss": 0.1059368799355897
    },
    {
      "epoch": 13,
      "train_loss": 0.11592144590887156
    },
    {
      "epoch": 14,
      "train_loss": 0.04884156241335652
    },
    {
      "epoch": 15,
      "train_loss": 0.044375065808946436
    },
    {
      "epoch": 16,
      "train_loss": 0.5301400216601112
    },
    {
      "epoch": 17,
      "train_loss": 0.2342631328918717
    },
    {
      "epoch": 18,
      "train_loss": 0.1805937432294542
    },
    {
      "epoch": 19,
      "train_loss": 0.15011224646459925
    },
    {
      "epoch": 20,
      "train_loss": 0.06355005253445019
    },
    {
      "epoch": 21,
      "train_loss": 0.16213352036747067
    },
    {
      "epoch": 22,
      "train_loss": 0.07662442246621305
    },
    {
      "epoch": 23,
      "train_loss": 0.06566207943992182
    },
    {
      "epoch": 24,
      "train_loss": 0.09292673818767071
    },
    {
      "epoch": 25,
      "train_loss": 0.27312749338420955
    },
    {
      "epoch": 26,
      "train_loss": 0.3949113512581045
    },
    {
      "epoch": 27,
      "train_loss": 0.25578115195035933
    },
    {
      "epoch": 28,
      "train_loss": 0.20474397618662227
    },
    {
      "epoch": 29,
      "train_loss": 0.1151967611096122
    },
    {
      "epoch": 30,
      "train_loss": 0.22775728605010293
    },
    {
      "epoch": 31,
      "train_loss": 0.11137387694282966
    },
    {
      "epoch": 32,
      "train_loss": 0.32518788596445863
    },
    {
      "epoch": 33,
      "train_loss": 0.3679282115806233
    },
    {
      "epoch": 34,
      "train_loss": 0.11811869431625713
    },
    {
      "epoch": 35,
      "train_loss": 0.25989035964012147
    },
    {
      "epoch": 36,
      "train_loss": 0.1275611624392596
    },
    {
      "epoch": 37,
      "train_loss": 0.24255421275442296
    },
    {
      "epoch": 38,
      "train_loss": 0.24983196166428653
    },
    {
      "epoch": 39,
      "train_loss": 0.10486641916361722
    },
    {
      "epoch": 40,
      "train_loss": 0.2554104011031714
    }
  ],
  "epochs_completed": 40,
  "finite": true,
  "first_nonfinite_epoch": null,
  "mission_lesion_rollout": {
    "episodes": 500,
    "mean_length": 4.29,
    "mean_return": 0.8905500000000001,
    "success": 0.978
  },
  "mission_lesion_validation": {
    "accuracy": 0.997008120814931,
    "loss": 0.09852080094472926
  },
  "predictor": [
    {
      "action_observations": 0,
      "context_rms": 0.29070043563842773,
      "mean_per_cell_r2": 0.45361873507499695,
      "neutral_observations": 4096,
      "residual_context_rms_ratio": 0.3701525032520294,
      "teaching_observations": 0
    },
    {
      "action_observations": 0,
      "context_rms": 0.2747928202152252,
      "mean_per_cell_r2": 0.5599287748336792,
      "neutral_observations": 4096,
      "residual_context_rms_ratio": 0.33980196714401245,
      "teaching_observations": 0
    }
  ],
  "provenance": {
    "cuda_device_name": "NVIDIA GeForce GTX 1080",
    "cuda_peak_allocated_bytes": 114644992,
    "cuda_visible_devices": "3",
    "device": "cuda:0",
    "git_commit": "e3011f144a018f6415b3c2a636c0646e412b3ced",
    "git_dirty_tracked": false,
    "minigrid_version": "3.1.0",
    "torch_version": "2.3.1+cu118"
  },
  "rollout": {
    "episodes": 500,
    "mean_length": 4.298,
    "mean_return": 0.8903500000000001,
    "success": 0.978
  },
  "stage": "babyai_shared_b1",
  "training": {
    "batch_size": 256,
    "model_seed": 4101,
    "neutral_examples_per_epoch": 4096,
    "parameter_count_including_reciprocal_context_predictor": 394759,
    "shuffle_seed": 4101,
    "total_wall_seconds_including_rollouts": 185.46793723106384,
    "training_wall_seconds": 183.09582328796387
  },
  "validation": {
    "accuracy": 0.994728593816783,
    "loss": 0.1036817075384531
  }
}