Datasets:
Download subset_info.json from openpecha/stok: direct link, hf CLI and curl.
- Browser
- Download file 361 Bytes
-
https://huggingface.co/datasets/openpecha/stok/resolve/main/subset_info.json
- Command line
-
hf download hf://datasets/openpecha/stok/subset_info.json
-
curl -L -o subset_info.json https://huggingface.co/datasets/openpecha/stok/resolve/main/subset_info.json
361 Bytes
| { | |
| "description": "OPTranskribus — Tibetan OCR page images + transcriptions (BDRC).", | |
| "creation_date": "2026-08-18T12:16:46.263299+00:00", | |
| "author": "Buddhist Digital Resource Center", | |
| "nb_images": 806, | |
| "parquet_total_size": 265161796, | |
| "image_format": "image_bytes", | |
| "extra_fields": [], | |
| "data_files": { | |
| "train": "data/train-*.parquet" | |
| } | |
| } |