StreamFlow: Theory, Algorithm, and Implementation for High-Efficiency Rectified Flow Generation

TL;DR

This paper introduces StreamFlow, a comprehensive acceleration pipeline for Rectified Flow models, significantly boosting image generation speed through innovative theory, design, and implementation strategies.

Contribution

The paper presents a novel acceleration pipeline for Rectified Flow models, incorporating new methods like batch processing, vectorization, and dynamic compilation to vastly improve efficiency.

Findings

Achieves up to 611% speedup in 512x512 image generation

Outperforms existing acceleration methods by a large margin

Provides a comprehensive framework for high-efficiency flow generation

Abstract

New technologies such as Rectified Flow and Flow Matching have significantly improved the performance of generative models in the past two years, especially in terms of control accuracy, generation quality, and generation efficiency. However, due to some differences in its theory, design, and existing diffusion models, the existing acceleration methods cannot be directly applied to the Rectified Flow model. In this article, we have comprehensively implemented an overall acceleration pipeline from the aspects of theory, design, and reasoning strategies. This pipeline uses new methods such as batch processing with a new velocity field, vectorization of heterogeneous time-step batch processing, and dynamic TensorRT compilation for the new methods to comprehensively accelerate related models based on flow models. Currently, the existing public methods usually achieve an acceleration of 18%, while experiments have proved that our new method can accelerate the 512*512 image generation speed to up to 611%, which is far beyond the current non-generalized acceleration methods.