A bfloat16 quantized version of the Kokoro text-to-speech model optimized for Apple MLX framework.
No resource links recorded.