Progressive Distillation for Fast Sampling of Diffusion Models · HackerPapers discussionshttps://hackerpaper.kaipfister.de/p/2202.00512/Collected HN comments and submissions linking this paper.enMon, 02 Nov 2026 11:16:22 GMTProgressive Distillation for Fast Sampling of Diffusion Models — mentioned by johndoughhttps://news.ycombinator.com/item?id=31086823urn:hackerpapers:mention:2202.00512:31086823Tue, 19 Apr 2022 18:20:51 GMT<p>I was wondering it if would be possible to train a neural network to do multiple iterative steps at once. As it turns out, it has already been done and it requires about 4 to 8 distilled iterations for comparable quality. If this pace keeps up, we will probably see similar running time to GANs in the near future. https://arxiv.org/pdf/2202.00512.pdf</p><p><a href="https://hackerpaper.kaipfister.de/p/2202.00512/">Paper and discussion history</a>. A mention is not an endorsement.</p>cs.LG