Update README.md

"csebuetnlp/banglabert" tokenizer is used to tokenize the dataset. This tokenizer must be used to decode this dataset and train any model.

README.md

@@ -22,4 +22,8 @@ configs:
   data_files:
   - split: train
     path: data/train-*
----
+task_categories:
+- text-generation
+language:
+- bn
+---