How to use runanywhere/whisper_medium_4bit with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir whisper_medium_4bit runanywhere/whisper_medium_4bit
