How to use kcvmk/whisper_small_4bit with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir whisper_small_4bit kcvmk/whisper_small_4bit
