tamil bert tokenizer 200000

trained using oscar dataset

colab file link