AudioCPP

Work in progress

Just recently started audio.cpp integration, still thinking architecture and expanding abstracted models!

Example using audio.cpp:

from delusion.audio.cpp import AudioCPP, OmniVoice

# No PyTorch or ONNX required
audio = AudioCPP(
    model=OmniVoice(quant="q8_0").download()
)

# Uses NamedTemporaryFile internally
speak = audio.tts(text="お水はもう一杯もらえますか")
Path("output.wav").write_bytes(speak.wav)