Fast Diffusion Probabilistic Model Sampling through the lens of Backward Error Analysis
This addresses the computational bottleneck in DDPMs for researchers and practitioners, offering a significant speedup without retraining, though it is incremental as it builds on existing diffusion ODE frameworks.
The paper tackles the slow sampling problem in denoising diffusion probabilistic models (DDPMs) by proposing a Restricting Backward Error (RBE) schedule, achieving high-quality samples with only 8 to 20 function evaluations, such as 12.01 FID on ImageNet 128x128 and a 20x speedup.
Denoising diffusion probabilistic models (DDPMs) are a class of powerful generative models. The past few years have witnessed the great success of DDPMs in generating high-fidelity samples. A significant limitation of the DDPMs is the slow sampling procedure. DDPMs generally need hundreds or thousands of sequential function evaluations (steps) of neural networks to generate a sample. This paper aims to develop a fast sampling method for DDPMs requiring much fewer steps while retaining high sample quality. The inference process of DDPMs approximates solving the corresponding diffusion ordinary differential equations (diffusion ODEs) in the continuous limit. This work analyzes how the backward error affects the diffusion ODEs and the sample quality in DDPMs. We propose fast sampling through the \textbf{Restricting Backward Error schedule (RBE schedule)} based on dynamically moderating the long-time backward error. Our method accelerates DDPMs without any further training. Our experiments show that sampling with an RBE schedule generates high-quality samples within only 8 to 20 function evaluations on various benchmark datasets. We achieved 12.01 FID in 8 function evaluations on the ImageNet $128\times128$, and a $20\times$ speedup compared with previous baseline samplers.