New text-to-speech model streams naturally from a single GPU

Original: TontaubeV1: Streaming Text-to-Speech with Hierarchical Codec Modeling and Bounded Context

Source: arxiv.org ↗