Fast Sampling of Diffusion Models with Exponential Integrator

04/29/2022
by   Qinsheng Zhang, et al.
0

The past few years have witnessed the great success of Diffusion models (DMs) in generating high-fidelity samples in generative modeling tasks. A major limitation of the DM is its notoriously slow sampling procedure which normally requires hundreds to thousands of time discretization steps of the learned diffusion process to reach the desired accuracy. Our goal is to develop a fast sampling method for DMs with much less number of steps while retaining high sample quality. To this end, we systematically analyze the sampling procedure in DMs and identify key factors that affect the sample quality, among which the method of discretization is most crucial. By carefully examining the learned diffusion process, we propose Diffusion Exponential Integrator Sampler (DEIS). It is based on the Exponential Integrator designed for discretizing ordinary differential equations (ODEs) and leverages a semilinear structure of the learned diffusion process to reduce the discretization error. The proposed method can be applied to any DMs and can generate high-fidelity samples in as few as 10 steps. In our experiments, it takes about 3 minutes on one A6000 GPU to generate 50k images from CIFAR10. Moreover, by directly using pre-trained DMs, we achieve the state-of-art sampling performance when the number of score function evaluation (NFE) is limited, e.g., 3.37 FID and 9.74 Inception score with only 15 NFEs on CIFAR10.

READ FULL TEXT

page 2

page 17

page 18

page 19

page 20

page 21

page 22

research
04/22/2023

Fast Diffusion Probabilistic Model Sampling through the lens of Backward Error Analysis

Denoising diffusion probabilistic models (DDPMs) are a class of powerful...
research
02/11/2022

Learning Fast Samplers for Diffusion Models by Differentiating Through Sample Quality

Diffusion models have emerged as an expressive family of generative mode...
research
06/26/2023

Restart Sampling for Improving Generative Processes

Generative processes that involve solving differential equations, such a...
research
08/15/2023

SciRE-Solver: Accelerating Diffusion Models Sampling by Score-integrand Solver with Recursive Difference

Diffusion models (DMs) have made significant progress in the fields of i...
research
05/23/2023

SEEDS: Exponential SDE Solvers for Fast High-Quality Sampling from Diffusion Models

A potent class of generative models known as Diffusion Probabilistic Mod...
research
11/24/2022

Fast Sampling of Diffusion Models via Operator Learning

Diffusion models have found widespread adoption in various areas. Howeve...
research
07/26/2023

MCMC-Correction of Score-Based Diffusion Models for Model Composition

Diffusion models can be parameterised in terms of either a score or an e...

Please sign up or login with your details

Forgot password? Click here to reset