Clone a voice and synthesize speech from text
Wavelet-Factorized Autoregressive Mel Model for voice clone
Zero-shot voice cloning with the UniVoice ASR+TTS LLM