How to use geodesic-research/nemotron-think-tokenizer with Transformers:
# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("geodesic-research/nemotron-think-tokenizer", dtype="auto")
