Progressive Distillation for Fast Sampling of Diffusion Models

02/01/2022
by   Tim Salimans, et al.
0

Diffusion models have recently shown great promise for generative modeling, outperforming GANs on perceptual quality and autoregressive models at density estimation. A remaining downside is their slow sampling time: generating high quality samples takes many hundreds or thousands of model evaluations. Here we make two contributions to help eliminate this downside: First, we present new parameterizations of diffusion models that provide increased stability when using few sampling steps. Second, we present a method to distill a trained deterministic diffusion sampler, using many steps, into a new diffusion model that takes half as many sampling steps. We then keep progressively applying this distillation procedure to our model, halving the number of required sampling steps each time. On standard image generation benchmarks like CIFAR-10, ImageNet, and LSUN, we start out with state-of-the-art samplers taking as many as 8192 steps, and are able to distill down to models taking as few as 4 steps without losing much perceptual quality; achieving, for example, a FID of 3.0 on CIFAR-10 in 4 steps. Finally, we show that the full progressive distillation procedure does not take more time than it takes to train the original model, thus representing an efficient solution for generative modeling using diffusion at both train and test time.

READ FULL TEXT

page 7

page 18

page 19

research
11/22/2022

Accelerating Diffusion Sampling with Classifier-based Feature Distillation

Although diffusion model has shown great potential for generating higher...
research
09/12/2023

InstaFlow: One Step is Enough for High-Quality Diffusion-Based Text-to-Image Generation

Diffusion models have revolutionized text-to-image generation with its e...
research
08/12/2023

Accelerating Diffusion-based Combinatorial Optimization Solvers by Progressive Distillation

Graph-based diffusion models have shown promising results in terms of ge...
research
05/18/2023

Catch-Up Distillation: You Only Need to Train Once for Accelerating Sampling

Diffusion Probability Models (DPMs) have made impressive advancements in...
research
02/18/2023

Modelos Generativos basados en Mecanismos de Difusión

Diffusion-based generative models are a design framework that allows gen...
research
10/10/2022

f-DM: A Multi-stage Diffusion Model via Progressive Signal Transformation

Diffusion models (DMs) have recently emerged as SoTA tools for generativ...
research
01/13/2023

Neural Image Compression with a Diffusion-Based Decoder

Diffusion probabilistic models have recently achieved remarkable success...

Please sign up or login with your details

Forgot password? Click here to reset