Upload dataset

README.md

@@ -39,11 +39,17 @@ dataset_info:
   - name: Drutsa
     num_bytes: 272793
     num_examples: 595
+  - name: KhyentseWangpo
+    num_bytes: 88706
+    num_examples: 143
   - name: Derge_Tenjur
     num_bytes: 50626920
     num_examples: 61656
-  download_size: 57826524
-  dataset_size: 211899367
+  - name: NorbukatekaNumbers
+    num_bytes: 97483
+    num_examples: 317
+  download_size: 57878208
+  dataset_size: 212085556
 configs:
 - config_name: default
   data_files:
@@ -61,8 +67,12 @@ configs:
     path: data/Karmapa8-*
   - split: Drutsa
     path: data/Drutsa-*
+  - split: KhyentseWangpo
+    path: data/KhyentseWangpo-*
   - split: Derge_Tenjur
     path: data/Derge_Tenjur-*
+  - split: NorbukatekaNumbers
+    path: data/NorbukatekaNumbers-*
 ---
 # Tibetan OCR-line-to-text Benchmark Dataset
 This repository hosts a line-to-text benchmark dataset to evaluate and compare Tibetan OCR models. The dataset includes diverse scripts, writing styles, and print methods, enabling comprehensive testing across multiple domains.

data/KhyentseWangpo-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:5a175a4eb3d6646bbaa3edf1d12614416215a834719bec6a69d04fdfcb4679ed
+size 28215

data/NorbukatekaNumbers-00000-of-00001.parquet

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:afbcd28a067797e35e281a62e9307c332297f5dd8bee20d06658d649c6641844
+size 23469