How to use voxmenthe/sum-small-unquantized with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir sum-small-unquantized voxmenthe/sum-small-unquantized
