PeRFlow: Piecewise Rectified Flow as Universal Plug-and-Play Accelerator

This demo includes PeRFlow acceleration module for Stable Diffusion v1.5, Generate high-quality images 512x512 with only 4 steps! Set use_refiner to True to generate 1024x1024 images.

More demos are also available at:

Introduction

Rectified Flow is a promising way for accelerating pre-trained diffusion models. However, the generation quality of prior fast flow-based models on Stable Diffusion (such as InstaFlow) is unsatisfactory. In this work, we did several improvements to the original reflow pipeline to significantly boost the performance of flow-based fast SD. Our new model learns a piecewise linear probability flow which can efficiently generate high-quality images in just 4 steps, termed piecewise rectified flow (PeRFlow). Moreover, we found the difference of model weights, ${\Delta}W = W_{\text{PeRFlow}} - W_{\text{SD}}$, can be used as a plug-and-play accelerator module on a wide-range of SD-based models.

Fast image generation via PeRFlow-T2I

Generate high-quality images (512x512) with only 4 steps!

Citation

@article{yan2024perflow,
  title={PeRFlow: Accelerating Diffusion Models with Piecewise Rectified Flows},
  author={Yan, Hanshu and Liu, Xingchao and Pan, Jiachun and Liew, Jun Hao and Liu, Qiang and Feng, Jiashi},
  year={2024},
  url={https://piecewise-rectified-flow.github.io}
}

Related Materials

We provide several related links here:

• The official Rectified Flow github repo (https://github.com/gnobitab/RectifiedFlow)

• The official InstaFlow github repo (https://github.com/gnobitab/InstaFlow)

Acknowledgements

Our training and evaluation scripts are implemented based on the Diffusers and Accelerate libraries. We use several high-quality finetuned versions of Stable Diffusion for model evaluation, including DreamShaper, RealisticVision, LandscapeRealistic, ArchitectureExterior, DisneyCartoon.

Xingchao Liu wishes to express his genuine gratitude to Nat Friedman and the Andromeda cluster for providing free GPU grants during this research.