{
  "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": 0.7408094794641842
    },
    {
      "epoch": 2,
      "train_loss": 0.4141200262849981
    },
    {
      "epoch": 3,
      "train_loss": 0.20642850984226574
    },
    {
      "epoch": 4,
      "train_loss": 0.2396153619072654
    },
    {
      "epoch": 5,
      "train_loss": 0.1747013689035719
    },
    {
      "epoch": 6,
      "train_loss": 0.09160391047596932
    },
    {
      "epoch": 7,
      "train_loss": 0.20036431182514539
    },
    {
      "epoch": 8,
      "train_loss": 0.05948227807879448
    },
    {
      "epoch": 9,
      "train_loss": 0.21365087948062203
    },
    {
      "epoch": 10,
      "train_loss": 0.0611378302899274
    },
    {
      "epoch": 11,
      "train_loss": 0.06920530608431859
    },
    {
      "epoch": 12,
      "train_loss": 0.047904286282983696
    },
    {
      "epoch": 13,
      "train_loss": 0.21460180244662544
    },
    {
      "epoch": 14,
      "train_loss": 0.05637756698510864
    },
    {
      "epoch": 15,
      "train_loss": 0.05161521524190903
    },
    {
      "epoch": 16,
      "train_loss": 0.3368245812234553
    },
    {
      "epoch": 17,
      "train_loss": 0.19509815630587665
    },
    {
      "epoch": 18,
      "train_loss": 0.2183473968234929
    },
    {
      "epoch": 19,
      "train_loss": 0.09419559308073737
    },
    {
      "epoch": 20,
      "train_loss": 0.07092469604177909
    },
    {
      "epoch": 21,
      "train_loss": 0.059987902383912696
    },
    {
      "epoch": 22,
      "train_loss": 0.05138562641360543
    },
    {
      "epoch": 23,
      "train_loss": 0.05321425412866202
    },
    {
      "epoch": 24,
      "train_loss": 0.2473546962236816
    },
    {
      "epoch": 25,
      "train_loss": 0.05551467421379956
    },
    {
      "epoch": 26,
      "train_loss": 0.11517106813463297
    },
    {
      "epoch": 27,
      "train_loss": 0.1695636088502678
    },
    {
      "epoch": 28,
      "train_loss": 0.04232350076464089
    },
    {
      "epoch": 29,
      "train_loss": 0.040572751760482785
    },
    {
      "epoch": 30,
      "train_loss": 0.041964036503976045
    },
    {
      "epoch": 31,
      "train_loss": 0.03777888340706175
    },
    {
      "epoch": 32,
      "train_loss": 0.043543690524318
    },
    {
      "epoch": 33,
      "train_loss": 0.041410166783766314
    },
    {
      "epoch": 34,
      "train_loss": 0.23046876887706194
    },
    {
      "epoch": 35,
      "train_loss": 0.04775922474536029
    },
    {
      "epoch": 36,
      "train_loss": 0.04496883080086925
    },
    {
      "epoch": 37,
      "train_loss": 0.038768304579637265
    },
    {
      "epoch": 38,
      "train_loss": 0.03775290479375557
    },
    {
      "epoch": 39,
      "train_loss": 0.1857465661791238
    },
    {
      "epoch": 40,
      "train_loss": 0.20626876744357023
    }
  ],
  "epochs_completed": 40,
  "finite": true,
  "first_nonfinite_epoch": null,
  "mission_lesion_rollout": {
    "episodes": 500,
    "mean_length": 5.742,
    "mean_return": 0.8496500000000001,
    "success": 0.932
  },
  "mission_lesion_validation": {
    "accuracy": 0.9762074369568314,
    "loss": 0.11846651761242086
  },
  "predictor": [
    {
      "action_observations": 0,
      "context_rms": 0.2629075348377228,
      "mean_per_cell_r2": 0.49027836322784424,
      "neutral_observations": 4096,
      "residual_context_rms_ratio": 0.3944878876209259,
      "teaching_observations": 0
    },
    {
      "action_observations": 0,
      "context_rms": 0.2731991410255432,
      "mean_per_cell_r2": 0.6234094500541687,
      "neutral_observations": 4096,
      "residual_context_rms_ratio": 0.32784295082092285,
      "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": 4.59,
    "mean_return": 0.8820500000000002,
    "success": 0.968
  },
  "stage": "babyai_shared_b1",
  "training": {
    "batch_size": 256,
    "model_seed": 4103,
    "neutral_examples_per_epoch": 4096,
    "parameter_count_including_reciprocal_context_predictor": 394759,
    "shuffle_seed": 4103,
    "total_wall_seconds_including_rollouts": 87.60878133773804,
    "training_wall_seconds": 85.27321362495422
  },
  "validation": {
    "accuracy": 0.9844707223251176,
    "loss": 0.11933256204455374
  }
}