{
  "description": "BDRC Tibetan OCR benchmark — open-access subset (test split).",
  "creation_date": "2026-08-18T09:40:42.257468+00:00",
  "nb_images": 472,
  "parquet_total_size": 122463224,
  "columns": [
    "image",
    "transcription",
    "id",
    "technology",
    "script",
    "script_4",
    "period",
    "rarity",
    "legibility"
  ],
  "data_files": {
    "test": "data/test-*.parquet"
  }
}