{
  "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": 0.7134594283320687
    },
    {
      "epoch": 2,
      "train_loss": 0.3401663822477514
    },
    {
      "epoch": 3,
      "train_loss": 0.19433609214696018
    },
    {
      "epoch": 4,
      "train_loss": 0.32963184771212667
    },
    {
      "epoch": 5,
      "train_loss": 0.08546664945103906
    },
    {
      "epoch": 6,
      "train_loss": 0.2630213478343053
    },
    {
      "epoch": 7,
      "train_loss": 0.11469778050075878
    },
    {
      "epoch": 8,
      "train_loss": 0.05664703453128988
    },
    {
      "epoch": 9,
      "train_loss": 0.04731606664982709
    },
    {
      "epoch": 10,
      "train_loss": 0.041082520572976636
    },
    {
      "epoch": 11,
      "train_loss": 0.20956628300926902
    },
    {
      "epoch": 12,
      "train_loss": 0.05752169869840145
    },
    {
      "epoch": 13,
      "train_loss": 0.03557884735817259
    },
    {
      "epoch": 14,
      "train_loss": 0.03279060132124207
    },
    {
      "epoch": 15,
      "train_loss": 0.031631635776297615
    },
    {
      "epoch": 16,
      "train_loss": 0.03048184691166336
    },
    {
      "epoch": 17,
      "train_loss": 0.3688752563975074
    },
    {
      "epoch": 18,
      "train_loss": 0.07081738247112794
    },
    {
      "epoch": 19,
      "train_loss": 0.054225066940892824
    },
    {
      "epoch": 20,
      "train_loss": 0.046492697589776734
    },
    {
      "epoch": 21,
      "train_loss": 0.047163276597857476
    },
    {
      "epoch": 22,
      "train_loss": 0.7845130688493902
    },
    {
      "epoch": 23,
      "train_loss": 0.3934751184691082
    },
    {
      "epoch": 24,
      "train_loss": 0.14599689253351905
    },
    {
      "epoch": 25,
      "train_loss": 0.3236034153672782
    },
    {
      "epoch": 26,
      "train_loss": 0.16323385662653228
    },
    {
      "epoch": 27,
      "train_loss": 0.09238848522305489
    },
    {
      "epoch": 28,
      "train_loss": 0.11827930971980095
    },
    {
      "epoch": 29,
      "train_loss": 0.15449962057850578
    },
    {
      "epoch": 30,
      "train_loss": 0.14310111538930373
    },
    {
      "epoch": 31,
      "train_loss": 0.12045336569574747
    },
    {
      "epoch": 32,
      "train_loss": 0.14461230769753455
    },
    {
      "epoch": 33,
      "train_loss": 0.08665146978064017
    },
    {
      "epoch": 34,
      "train_loss": 0.12435456068678336
    },
    {
      "epoch": 35,
      "train_loss": 0.15446093495596538
    },
    {
      "epoch": 36,
      "train_loss": 0.09183386780998924
    },
    {
      "epoch": 37,
      "train_loss": 0.06461878389120101
    },
    {
      "epoch": 38,
      "train_loss": 0.05828312648290938
    },
    {
      "epoch": 39,
      "train_loss": 0.15417793202129276
    },
    {
      "epoch": 40,
      "train_loss": 0.6200590939007022
    }
  ],
  "epochs_completed": 40,
  "finite": true,
  "first_nonfinite_epoch": null,
  "mission_lesion_rollout": {
    "episodes": 500,
    "mean_length": 8.284,
    "mean_return": 0.7789,
    "success": 0.86
  },
  "mission_lesion_validation": {
    "accuracy": 0.8928622310870494,
    "loss": 0.36016965703859616
  },
  "predictor": [
    {
      "action_observations": 0,
      "context_rms": 0.2660917639732361,
      "mean_per_cell_r2": 0.47728049755096436,
      "neutral_observations": 4096,
      "residual_context_rms_ratio": 0.3765350580215454,
      "teaching_observations": 0
    },
    {
      "action_observations": 0,
      "context_rms": 0.2768581211566925,
      "mean_per_cell_r2": 0.6187853217124939,
      "neutral_observations": 4096,
      "residual_context_rms_ratio": 0.3115047812461853,
      "teaching_observations": 0
    }
  ],
  "provenance": {
    "cuda_device_name": "NVIDIA GeForce GTX 1080",
    "cuda_peak_allocated_bytes": 114644992,
    "cuda_visible_devices": "7",
    "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": 5.956,
    "mean_return": 0.8441000000000002,
    "success": 0.93
  },
  "stage": "babyai_shared_b1",
  "training": {
    "batch_size": 256,
    "model_seed": 4102,
    "neutral_examples_per_epoch": 4096,
    "parameter_count_including_reciprocal_context_predictor": 394759,
    "shuffle_seed": 4102,
    "total_wall_seconds_including_rollouts": 186.63246417045593,
    "training_wall_seconds": 183.2198555469513
  },
  "validation": {
    "accuracy": 0.9477133494799829,
    "loss": 0.29150108231966915
  }
}