Update README.md

README.md

@@ -23,4 +23,23 @@ configs:
     path: data/train-*
   - split: validation
     path: data/validation-*
+task_categories:
+- image-to-text
+- visual-question-answering
+language:
+- dv
+tags:
+- thaana
+- dhivehi
+- ocr
+pretty_name: ocr_thaa
+license: apache-2.0
 ---
+
+# Dhivehi Single-Line Text-Image Dataset
+
+A collection of synthetic Dhivehi text images for training and evaluating text-image / vision models etc. 
+
+Each image contains a single line of Dhivehi text with various visual styles and augmentations (Check the config field for more info on the row).
+
+***Note: This dataset is a subset from alakxender/dhivehi-vrd-images.***