How to use mmm128/qwen3-math-thinking-tokenizer with Transformers:
# Load model directly from transformers import AutoModel model = AutoModel.from_pretrained("mmm128/qwen3-math-thinking-tokenizer", dtype="auto")
