Fetching the paper…

ConsistencyTTA: Accelerating Diffusion-Based Text-to-Audio Generation with Consistency Distillation · Around