Provable Statistical Rates for Consistency Diffusion Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Dou, Zehao, Chen, Minshuo, Wang, Mengdi, Yang, Zhuoran
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910499603480576
author Dou, Zehao
Chen, Minshuo
Wang, Mengdi
Yang, Zhuoran
author_facet Dou, Zehao
Chen, Minshuo
Wang, Mengdi
Yang, Zhuoran
contents Diffusion models have revolutionized various application domains, including computer vision and audio generation. Despite the state-of-the-art performance, diffusion models are known for their slow sample generation due to the extensive number of steps involved. In response, consistency models have been developed to merge multiple steps in the sampling process, thereby significantly boosting the speed of sample generation without compromising quality. This paper contributes towards the first statistical theory for consistency models, formulating their training as a distribution discrepancy minimization problem. Our analysis yields statistical estimation rates based on the Wasserstein distance for consistency models, matching those of vanilla diffusion models. Additionally, our results encompass the training of consistency models through both distillation and isolation methods, demystifying their underlying advantage.
format Preprint
id arxiv_https___arxiv_org_abs_2406_16213
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Provable Statistical Rates for Consistency Diffusion Models
Dou, Zehao
Chen, Minshuo
Wang, Mengdi
Yang, Zhuoran
Machine Learning
Diffusion models have revolutionized various application domains, including computer vision and audio generation. Despite the state-of-the-art performance, diffusion models are known for their slow sample generation due to the extensive number of steps involved. In response, consistency models have been developed to merge multiple steps in the sampling process, thereby significantly boosting the speed of sample generation without compromising quality. This paper contributes towards the first statistical theory for consistency models, formulating their training as a distribution discrepancy minimization problem. Our analysis yields statistical estimation rates based on the Wasserstein distance for consistency models, matching those of vanilla diffusion models. Additionally, our results encompass the training of consistency models through both distillation and isolation methods, demystifying their underlying advantage.
title Provable Statistical Rates for Consistency Diffusion Models
topic Machine Learning
url https://arxiv.org/abs/2406.16213