How to use Grunzig/gemma-4-12B-it-qat-oQ4e-fp16-mtp with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir gemma-4-12B-it-qat-oQ4e-fp16-mtp Grunzig/gemma-4-12B-it-qat-oQ4e-fp16-mtp
