TwinFlow: Realizing One-step Generation on Large Models with Self-adversarial Flows

TL;DR

TwinFlow introduces a one-step generative model training framework that significantly accelerates large multi-modal model inference, achieving high-quality results with just a single evaluation step, thus reducing computational costs drastically.

Contribution

The paper presents TwinFlow, a novel training method for 1-step generative models that eliminates the need for fixed teachers and adversarial training, enabling efficient large-scale model inference.

Findings

Achieves 0.83 GenEval score with 1-NFE on text-to-image tasks.

Matches 100-NFE model performance with just 1-NFE, reducing computation by 100x.

Outperforms existing baselines like SANA-Sprint and RCGM in efficiency and quality.

Abstract

Recent advances in large multi-modal generative models have demonstrated impressive capabilities in multi-modal generation, including image and video generation. These models are typically built upon multi-step frameworks like diffusion and flow matching, which inherently limits their inference efficiency (requiring 40-100 Number of Function Evaluations (NFEs)). While various few-step methods aim to accelerate the inference, existing solutions have clear limitations. Prominent distillation-based methods, such as progressive and consistency distillation, either require an iterative distillation procedure or show significant degradation at very few steps (< 4-NFE). Meanwhile, integrating adversarial training into distillation (e.g., DMD/DMD2 and SANA-Sprint) to enhance performance introduces training instability, added complexity, and high GPU memory overhead due to the auxiliary trained models. To this end, we propose TwinFlow, a simple yet effective framework for training 1-step generative models that bypasses the need of fixed pretrained teacher models and avoids standard adversarial networks during training, making it ideal for building large-scale, efficient models. On text-to-image tasks, our method achieves a GenEval score of 0.83 in 1-NFE, outperforming strong baselines like SANA-Sprint (a GAN loss-based framework) and RCGM (a consistency-based framework). Notably, we demonstrate the scalability of TwinFlow by full-parameter training on Qwen-Image-20B and transform it into an efficient few-step generator. With just 1-NFE, our approach matches the performance of the original 100-NFE model on both the GenEval and DPG-Bench benchmarks, reducing computational cost by $100\times$ with minor quality degradation. Project page is available at https://zhenglin-cheng.com/twinflow.