Align Your Flow: Scaling Continuous-Time Flow Map Distillation
This work addresses the need for efficient and high-quality few-step generative models in image synthesis, representing an incremental improvement over existing consistency and flow matching methods.
The paper tackled the problem of performance degradation in consistency models when increasing sampling steps for generative modeling, introducing continuous-time flow map objectives and training techniques that achieve state-of-the-art few-step generation on ImageNet benchmarks and outperform existing non-adversarial text-to-image samplers.
Diffusion- and flow-based models have emerged as state-of-the-art generative modeling approaches, but they require many sampling steps. Consistency models can distill these models into efficient one-step generators; however, unlike flow- and diffusion-based methods, their performance inevitably degrades when increasing the number of steps, which we show both analytically and empirically. Flow maps generalize these approaches by connecting any two noise levels in a single step and remain effective across all step counts. In this paper, we introduce two new continuous-time objectives for training flow maps, along with additional novel training techniques, generalizing existing consistency and flow matching objectives. We further demonstrate that autoguidance can improve performance, using a low-quality model for guidance during distillation, and an additional boost can be achieved by adversarial finetuning, with minimal loss in sample diversity. We extensively validate our flow map models, called Align Your Flow, on challenging image generation benchmarks and achieve state-of-the-art few-step generation performance on both ImageNet 64x64 and 512x512, using small and efficient neural networks. Finally, we show text-to-image flow map models that outperform all existing non-adversarially trained few-step samplers in text-conditioned synthesis.