Fetching the paper…

MaskGCT: Zero-Shot Text-to-Speech with Masked Generative Codec Transformer · Around