Paper

ReCache: Learning Budget-Aware Caching Schedules for Diffusion Models via REINFORCE

arXiv:2606.06060v1 Announce Type: new Abstract: Modern diffusion models generate high-quality images and videos, but their iterative denoising process makes inference expensive. Feature caching accelerates sampling by reusing or predicting intermediate activations across neighboring denoising steps, exploiting the redundancy of computations along the reverse trajectory. In this work, we focus on the caching schedule: selecting which denoising steps should be fully recomputed. Existing schedules are either fixed (e.g. uniform) or chosen adaptively from per-step error heuristics; in both cases,…

Authors:

Related discussions