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