כתבה
arXiv cs.AI ·
ART for Diffusion Sampling: גישה ללמידת ריפוי על-פי שגיאות זמן
ART for Diffusion Sampling: A Reinforcement Learning Approach to Timestep Schedule
אורך זמן נבחר עבור דיפוזיה-ב-סקור, כדי לשפר תפוקה ואיכות דגימות.
תקציר מקורי באנגליתarXiv:2601.18681v3 Announce Type: replace-cross Abstract: We consider time discretization for score-based diffusion models to generate samples from a learned reverse-time dynamic on a finite grid. Uniform and hand-crafted grids can be suboptimal given a budget on the number of time steps. We introduce Adaptive Reparameterized Time (ART), which controls the clock speed of a reparameterized time variable to redistribute computation along the sampling trajectory while preserving the terminal time, with the objective of minimizing the aggregate Euler discretization error. We derive a randomized companion ART-RL that recasts ART as a continuous-time reinforcement learning problem with Gaussian policies, and prove a two-directional bridge between the two: the deterministic ART optimum lifts to a
קרא במקור המקורי
arxiv.org
פתח כתבה מקורית