How to use Mediform/gemma4-e4b-v13-assistant-rollout-mlx-bf16 with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir gemma4-e4b-v13-assistant-rollout-mlx-bf16 Mediform/gemma4-e4b-v13-assistant-rollout-mlx-bf16
