How to use rg1683/hindi_bpe_tokenizer with Transformers:
# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("rg1683/hindi_bpe_tokenizer", dtype="auto")