论文标题
用指数积分快速采样扩散模型
Fast Sampling of Diffusion Models with Exponential Integrator
论文作者
论文摘要
过去的几年见证了扩散模型〜（DM）在生成建模任务中生成高保真样本方面取得的巨大成功。 DM的主要局限性是其臭名昭著的缓慢采样程序，通常需要数百至数千个学习扩散过程的离散步骤才能达到所需的准确性。我们的目标是为DMS开发一种快速采样方法，该方法的步骤少得多，同时保留了高样本质量。为此，我们系统地分析了DMS中的采样程序，并确定影响样本质量的关键因素，其中离散化方法至关重要。通过仔细检查学习的扩散过程，我们提出了扩散指数积分器〜（DEIS）。它基于设计用于离散的普通微分方程（ODE）的指数积分器，并利用学习扩散过程的半线性结构来减少离散误差。所提出的方法可以应用于任何DMS，并且可以在只有10个步骤中生成高保真样本。在我们的实验中，一个A6000 GPU大约需要3分钟才能从CIFAR10产生$ 50K $的图像。此外，通过直接使用预训练的DMS，当得分函数评估〜（NFE）的数量有限时，我们实现了最先进的采样性能，例如，使用10 NFE，3.37 FID和9.74的4.17 FID，在CIFAR10上只有15个NFE。代码可从https://github.com/qsh-zh/deis获得
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 a 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., 4.17 FID with 10 NFEs, 3.37 FID, and 9.74 IS with only 15 NFEs on CIFAR10. Code is available at https://github.com/qsh-zh/deis
