tsfile_conversion.json

{
  "source_dataset": "jackvial/koch_test",
  "converted_data_files": [
    "data/jackvial_koch_test_chunk_v2.tsfile",
    "data/jackvial_koch_test_train_legacy.tsfile"
  ],
  "time_precision": "ms",
  "time_mapping": {
    "source": "timestamp",
    "formula": "round(timestamp * 1000)",
    "unit": "milliseconds",
    "restarts_per_episode": true
  },
  "layouts": [
    {
      "layout": "chunk_v2",
      "table_name": "jackvial_koch_test_chunk_v2",
      "source_files": [
        "data/chunk-000/episode_000000.parquet",
        "data/chunk-000/episode_000001.parquet"
      ],
      "source_rows": 1192,
      "converted_rows": 1192,
      "episodes": {
        "0": 596,
        "1": 596
      },
      "tag_columns": [
        "episode_index",
        "task_index"
      ],
      "video_columns_omitted": [],
      "staged_parquet": "jackvial_koch_test_chunk_v2.parquet",
      "tsfile": "data/jackvial_koch_test_chunk_v2.tsfile"
    },
    {
      "layout": "train_legacy",
      "table_name": "jackvial_koch_test_train_legacy",
      "source_files": [
        "data/train-00000-of-00001.parquet"
      ],
      "source_rows": 880,
      "converted_rows": 880,
      "episodes": {
        "0": 289,
        "1": 294,
        "2": 297
      },
      "tag_columns": [
        "episode_index"
      ],
      "video_columns_omitted": [
        "observation.images.laptop"
      ],
      "staged_parquet": "jackvial_koch_test_train_legacy.parquet",
      "tsfile": "data/jackvial_koch_test_train_legacy.tsfile"
    }
  ],
  "dropped_columns": {
    "timestamp": "Converted to Time and not retained as a separate field."
  },
  "omitted_video_columns": {
    "observation.images.laptop": "Legacy train layout video path struct; videos remain in the original dataset."
  },
  "video_policy": "Videos are not uploaded; use the original Hugging Face dataset videos."
}