扩散采样的最优步长

Optimal Stepsize for Diffusion Sampling

March 27, 2025

作者: Jianning Pei, Han Hu, Shuyang Gu

cs.AI

摘要

扩散模型在生成质量上表现出色，但由于次优的步长离散化，其采样过程计算密集。现有研究主要集中于优化去噪方向，而本文则着眼于步长调度的原则性设计。我们提出了最优步长蒸馏法，这是一个动态规划框架，通过从参考轨迹中提炼知识来提取理论上的最优调度方案。通过将步长优化重新表述为递归误差最小化问题，我们的方法利用最优子结构特性，确保了全局离散化边界。关键在于，所提炼的调度方案在架构、ODE求解器和噪声调度方案上均展现出强大的鲁棒性。实验表明，文本到图像生成速度提升了10倍，同时在GenEval基准上保持了99.4%的性能。我们的代码已发布于https://github.com/bebebe666/OptimalSteps。

English

Diffusion models achieve remarkable generation quality but suffer from computational intensive sampling due to suboptimal step discretization. While existing works focus on optimizing denoising directions, we address the principled design of stepsize schedules. This paper proposes Optimal Stepsize Distillation, a dynamic programming framework that extracts theoretically optimal schedules by distilling knowledge from reference trajectories. By reformulating stepsize optimization as recursive error minimization, our method guarantees global discretization bounds through optimal substructure exploitation. Crucially, the distilled schedules demonstrate strong robustness across architectures, ODE solvers, and noise schedules. Experiments show 10x accelerated text-to-image generation while preserving 99.4% performance on GenEval. Our code is available at https://github.com/bebebe666/OptimalSteps.