- New “distilled diffusion models” research can create high quality images 256x faster with step counts as low as 4 https://arxiv.org/abs/2210.03142 47 comments machinelearning
Linking pages
- SD RESOURCE GOLDMINE https://rentry.co/sdupdates 23 comments
- GitHub - heejkoo/Awesome-Diffusion-Models: A collection of resources and papers on Diffusion Models https://github.com/heejkoo/Awesome-Diffusion-Models 0 comments
- Google and Stanford Researchers Propose a Novel Approach for Distilling Classifier-Free Guided Diffusion Models with High Sampling Efficiency - MarkTechPost https://www.marktechpost.com/2022/10/22/google-and-stanford-researchers-propose-a-novel-approach-for-distilling-classifier-free-guided-diffusion-models-with-high-sampling-efficiency/ 0 comments
- Perspectives on diffusion – Sander Dieleman https://sander.ai/2023/07/20/perspectives.html 0 comments
- The paradox of diffusion distillation – Sander Dieleman https://sander.ai/2024/02/28/paradox.html 0 comments
Would you like to stay up to date with Computer science? Checkout Computer science
Weekly.
Related searches:
Search whole site: site:arxiv.org
Search title: [2210.03142] On Distillation of Guided Diffusion Models
See how to search.