A distilled version of the MLX model, improving efficiency and performance.
Discovered on HuggingFace via HuggingFace:unknown