Open-access subset (472) + documentable metadata (private draft)

README.md

@@ -48,6 +48,7 @@ published here.
 | `id` | stable per-page id (the anonymized page number, e.g. `0002`) |
 | `technology` | how the page was produced: `manuscripts` / `blockprints` / `modern` / `metal_type` / `typewriter` / … |
 | `script` | coarse writing system: `Uchen` / `Ume` |
+| `script_4` | writing-style family: `Uchen` / `Pedri` / `Gyuyig` / `Druma` / `Tsugdri` / `Danyig` / `Non-applicable` |
 | `period` | palaeographic period of the writing style (e.g. *Medieval (15)-present*) |
 | `rarity` | how common the writing style is (e.g. *4 - very common*) |
 | `legibility` | annotator legibility grade of the page (e.g. *2 - Fair*, *4 - Excellent*) |

data/test-00000-of-00001.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:6311f30ab2d0a55abb8a5f211d409d1b5895ebe101ebcb904575b0694fbd84dd
-size 122462298
+oid sha256:39b426ac868bb24f4d2c4d5d63b2c51bae2e79b374ee240b3eb9e016cafbb035
+size 122463224

subset_info.json

@@ -1,14 +1,15 @@
 {
   "description": "BDRC Tibetan OCR benchmark — open-access subset (test split).",
-  "creation_date": "2026-08-17T14:58:27.201216+00:00",
+  "creation_date": "2026-08-17T15:05:35.681823+00:00",
   "nb_images": 472,
-  "parquet_total_size": 122462298,
+  "parquet_total_size": 122463224,
   "columns": [
     "image",
     "transcription",
     "id",
     "technology",
     "script",
+    "script_4",
     "period",
     "rarity",
     "legibility"