episodes.jsonl

{"episode_index": 0, "episode_path": "data/chunk-000/episode_000000.parquet", "length": 960, "tasks": [0]}