Diffusion models have achieved state-of-the-art synthesis quality on both visual and audio tasks, and recent works further adapt them to textual data by diffusing on the embedding space.
BibTex:
Before browse our site, please accept our cookies policy
