MLX How to use zecanard/gemma-4-12b-it-uncensored-heretic-MLX-4bit-mixed_4_6 with MLX:
# Download the model from the Hub
pip install huggingface_hub[hf_xet]
huggingface-cli download --local-dir gemma-4-12b-it-uncensored-heretic-MLX-4bit-mixed_4_6 zecanard/gemma-4-12b-it-uncensored-heretic-MLX-4bit-mixed_4_6
