results: retain GoToObj as diagnostic only

1 files changed, 251 insertions, 0 deletions

diff --git a/results/babyai_shared/b1/seed4101_raw_shared.json b/results/babyai_shared/b1/seed4101_raw_shared.json
new file mode 100644
index 0000000..91d5cb3
--- /dev/null
+++ b/results/babyai_shared/b1/seed4101_raw_shared.json
@@ -0,0 +1,251 @@
+{
+  "condition": "raw_shared",
+  "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": 0.912005416913466
+    },
+    {
+      "epoch": 2,
+      "train_loss": 0.4067533041672273
+    },
+    {
+      "epoch": 3,
+      "train_loss": 0.17175627085295592
+    },
+    {
+      "epoch": 4,
+      "train_loss": 0.101405300850218
+    },
+    {
+      "epoch": 5,
+      "train_loss": 0.15046165440570225
+    },
+    {
+      "epoch": 6,
+      "train_loss": 0.06658889784054323
+    },
+    {
+      "epoch": 7,
+      "train_loss": 0.06327508542348038
+    },
+    {
+      "epoch": 8,
+      "train_loss": 0.06550759131258184
+    },
+    {
+      "epoch": 9,
+      "train_loss": 0.17024030198597095
+    },
+    {
+      "epoch": 10,
+      "train_loss": 0.07377955204383893
+    },
+    {
+      "epoch": 11,
+      "train_loss": 0.053011001592332664
+    },
+    {
+      "epoch": 12,
+      "train_loss": 0.06881083751943978
+    },
+    {
+      "epoch": 13,
+      "train_loss": 0.10856242247603157
+    },
+    {
+      "epoch": 14,
+      "train_loss": 0.06692509006370198
+    },
+    {
+      "epoch": 15,
+      "train_loss": 0.0633675326068293
+    },
+    {
+      "epoch": 16,
+      "train_loss": 0.11487563056024638
+    },
+    {
+      "epoch": 17,
+      "train_loss": 0.0888773624836044
+    },
+    {
+      "epoch": 18,
+      "train_loss": 0.06327656554227526
+    },
+    {
+      "epoch": 19,
+      "train_loss": 0.057851375937461855
+    },
+    {
+      "epoch": 20,
+      "train_loss": 0.07349031143229116
+    },
+    {
+      "epoch": 21,
+      "train_loss": 0.04631090391088616
+    },
+    {
+      "epoch": 22,
+      "train_loss": 0.07982008521868424
+    },
+    {
+      "epoch": 23,
+      "train_loss": 0.058804434524341066
+    },
+    {
+      "epoch": 24,
+      "train_loss": 0.09953950692645529
+    },
+    {
+      "epoch": 25,
+      "train_loss": 0.06437903018498962
+    },
+    {
+      "epoch": 26,
+      "train_loss": 0.057973786971785804
+    },
+    {
+      "epoch": 27,
+      "train_loss": 0.09784971290014007
+    },
+    {
+      "epoch": 28,
+      "train_loss": 0.04299957272003997
+    },
+    {
+      "epoch": 29,
+      "train_loss": 0.0515958441590721
+    },
+    {
+      "epoch": 30,
+      "train_loss": 0.06496320383453912
+    },
+    {
+      "epoch": 31,
+      "train_loss": 0.1328263165361502
+    },
+    {
+      "epoch": 32,
+      "train_loss": 0.04388667389750481
+    },
+    {
+      "epoch": 33,
+      "train_loss": 0.07855626332150265
+    },
+    {
+      "epoch": 34,
+      "train_loss": 0.050679853653365914
+    },
+    {
+      "epoch": 35,
+      "train_loss": 0.049945260245691646
+    },
+    {
+      "epoch": 36,
+      "train_loss": 0.051781387623738157
+    },
+    {
+      "epoch": 37,
+      "train_loss": 0.04179245889525522
+    },
+    {
+      "epoch": 38,
+      "train_loss": 0.04486359935592521
+    },
+    {
+      "epoch": 39,
+      "train_loss": 0.04852475961162286
+    },
+    {
+      "epoch": 40,
+      "train_loss": 0.05743650953878056
+    }
+  ],
+  "epochs_completed": 40,
+  "finite": true,
+  "first_nonfinite_epoch": null,
+  "mission_lesion_rollout": {
+    "episodes": 500,
+    "mean_length": 13.522,
+    "mean_return": 0.63055,
+    "success": 0.686
+  },
+  "mission_lesion_validation": {
+    "accuracy": 0.870921783729876,
+    "loss": 0.24419261227477426
+  },
+  "predictor": [],
+  "provenance": {
+    "cuda_device_name": "NVIDIA GeForce GTX 1080",
+    "cuda_peak_allocated_bytes": 39770112,
+    "cuda_visible_devices": "2",
+    "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": 14.164,
+    "mean_return": 0.6130999999999999,
+    "success": 0.672
+  },
+  "stage": "babyai_shared_b1",
+  "training": {
+    "batch_size": 256,
+    "model_seed": 4101,
+    "neutral_examples_per_epoch": 0,
+    "parameter_count_including_reciprocal_context_predictor": 394759,
+    "shuffle_seed": 4101,
+    "total_wall_seconds_including_rollouts": 184.85894656181335,
+    "training_wall_seconds": 178.62242436408997
+  },
+  "validation": {
+    "accuracy": 0.8724889585411028,
+    "loss": 0.3364062811306254
+  }
+}