KittenML / KittenTTS

Open-source text-to-speech library with a 1.7B 1-bit speech model and lightweight ONNX models that run on CPU.

facebookresearch / audiocraft

PyTorch library for audio generation research, providing inference and training code for MusicGen, AudioGen, and EnCodec.