mlx-lm provides tools for running, fine-tuning, and quantizing large language models using Apple's MLX framework and the Hugging Face Hub. Requires Apple Silicon (M1/M2/M3+).

WWW: https://github.com/ml-explore/mlx-lm
