paper-with-me

홈 › Papers

Convergence guarantee for consistency models

2023-08-22 · Junlong Lyu, Zhitang Chen, Shoubo Feng

We provide the first convergence guarantees for the Consistency Models (CMs), a newly emerging type of one-step generative models that can generate comparable samples to those generated by Diffusion Models. Our main result is that, under the basic assumptions on score-matching errors, consistency errors and smoothness of the data distribution, CMs can efficiently sample from any realistic data distribution in one step with small $W_2$ error. Our results (1) hold for $L^2$-accurate score and consistency assumption (rather than $L^\infty$-accurate); (2) do note require strong assumptions on the data distribution such as log-Sobelev inequality; (3) scale polynomially in all parameters; and (4) match the state-of-the-art convergence guarantee for score-based generative models (SGMs). We also provide the result that the Multistep Consistency Sampling procedure can further reduce the error comparing to one step sampling, which support the original statement of "Consistency Models, Yang Song 2023". Our result further imply a TV error guarantee when take some Langevin-based modifications to the output distributions.

📄 PDF Abstract BibTeX arXiv:2308.11449

Code (0)

등록된 구현이 없습니다.

Methods 이 논문이 사용한 방법론

Diffusion Diffusion models generate samples by gradually removing noise from a signal, and their training objective can be expressed as a reweighted variational lower-bound…

Similar Papers 제목 키워드 기반

Unified theoretical guarantees for stability, consistency, and convergence in neural PDE solvers from non-IID data to physics-informed networks

2024-09-08 · Ronald Katende

We establish a unified theoretical framework addressing the stability, consistency, and convergence of neural networks under realistic training conditions, specifically, in the presence of non-IID data, geometric constra…

Federated Learning

Regularized Loss Minimizers with Local Data Perturbation: Consistency and Data Irrecoverability

2018-05-19 · Zitao Li, Jean Honorio

We introduce a new concept, data irrecoverability, and show that the well-studied concept of data privacy is sufficient but not necessary for data irrecoverability. We show that there are several regularized loss minimiz…

Multi-Step Consistency Models: Fast Generation with Theoretical Guarantees

2025-05-02 · Nishant Jain, Xunpeng Huang, Yian Ma, Tong Zhang

Consistency models have recently emerged as a compelling alternative to traditional SDE-based diffusion models. They offer a significant acceleration in generation by producing high-quality samples in very few steps. Des…

PCM : Picard Consistency Model for Fast Parallel Sampling of Diffusion Models

2025-03-25 · CVPR 2025 1 · Junhyuk So, Jiwoong Shin, Chaeyeon Jang, Eunhyeok Park

Recently, diffusion models have achieved significant advances in vision, text, and robotics. However, they still face slow generation speeds due to sequential denoising processes. To address this, a parallel sampling met…

DenoisingImage Generation

Memory-efficient model-based deep learning with convergence and robustness guarantees

2022-06-06 · Aniket Pramanik, M. Bridget Zimmerman, Mathews Jacob

Computational imaging has been revolutionized by compressed sensing algorithms, which offer guaranteed uniqueness, convergence, and stability properties. Model-based deep learning methods that combine imaging physics wit…

compressed sensingCompressive SensingOperator learning