How to use iris-sfg/Voxtral-Mini-4B-Realtime-2602-4bit-fp16 with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir Voxtral-Mini-4B-Realtime-2602-4bit-fp16 iris-sfg/Voxtral-Mini-4B-Realtime-2602-4bit-fp16
