How to use sgrankin/Qwen3-Reranker-4B-oQ8-fp16 with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir Qwen3-Reranker-4B-oQ8-fp16 sgrankin/Qwen3-Reranker-4B-oQ8-fp16
