Fetching the paper…

DiTAR: Diffusion Transformer Autoregressive Modeling for Speech Generation · Around