Fetching the paper…

FastDiff: A Fast Conditional Diffusion Model for High-Quality Speech Synthesis · Around