How to use ToPo-ToPo/Qwen3.8-27B-MTP-4bit with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir Qwen3.8-27B-MTP-4bit ToPo-ToPo/Qwen3.8-27B-MTP-4bit
ToPo-ToPo commited on 