How to use kcvmk/qwen3_4B_mlx_4bit with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir qwen3_4B_mlx_4bit kcvmk/qwen3_4B_mlx_4bit
kcvmk commited on 