results/episode_task_suite/timeline_subtask/metrics.json

{
  "accuracy": 0.05813953488372093,
  "balanced_accuracy": 0.05376979652090881,
  "macro_f1": 0.04954121121178666,
  "weighted_f1": 0.06731304264454903,
  "num_eval_windows": 344,
  "num_classes": 14,
  "task": "timeline_subtask",
  "input": "all modalities -> current subtask label",
  "split": "chronological",
  "num_windows": 1147,
  "num_train_windows": 803,
  "num_test_windows": 344,
  "feature_dim": 8378,
  "majority_baseline_accuracy": 0.0,
  "train_final_accuracy": 1.0,
  "train_final_loss": 0.014040183275938034,
  "unseen_test_classes": [
    "Move bottle to coffee equipment",
    "Pour coffee",
    "Pour milk into coffee",
    "Prepare for pouring"
  ]
}