How to use eoinf/pile_tokenizer_4096 with Transformers:
# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("eoinf/pile_tokenizer_4096", dtype="auto")
