Decoupled MeanFlow:将流模型转化为流图以加速采样
Decoupled MeanFlow: Turning Flow Models into Flow Maps for Accelerated Sampling
AI总结:
针对流模型采样步骤多的问题,提出Decoupled MeanFlow策略,无需修改架构即可将预训练流模型转化为流图,结合增强训练实现1-4步高质量生成,在ImageNet上FID表现远超现有技术,4步推理速度提升超100倍。
摘要:
去噪生成模型(如扩散模型和基于流的模型)能生成高质量样本,但由于离散化误差,需要大量去噪步骤。流图(flow maps)可估计时间步之间的平均速度,能缓解该误差并实现更快采样。然而,流图的训练通常需要调整架构,限制了与预训练流模型的兼容性。我们提出Decoupled MeanFlow,这是一种无需修改架构即可将流模型转化为流图模型的简单解码策略。我们的方法让扩散Transformer的最终模块以随后的时间步为条件,使预训练流模型可直接被改造为流图。结合增强的训练技术,该设计仅需1到4步就能实现高质量生成。值得注意的是,我们发现先训练流模型再进行转化,比从头训练流图更高效且效果更好。在ImageNet 256x256和512x512数据集上,我们的模型1步生成的FID分别达到2.16和2.12,大幅超越现有技术。此外,当步数增加到4时,我们取得了1.51和1.68的FID,几乎匹配流模型的性能,同时推理速度提升超过100倍。
英文摘要:
Denoising generative models, such as diffusion and flow-based models, produce high-quality samples but require many denoising steps due to discretization error. Flow maps, which estimate the average velocity between timesteps, mitigate this error and enable faster sampling. However, their training typically demands architectural changes that limit compatibility with pretrained flow models. We introduce Decoupled MeanFlow, a simple decoding strategy that converts flow models into flow map models without architectural modifications. Our method conditions the final blocks of diffusion transformers on the subsequent timestep, allowing pretrained flow models to be directly repurposed as flow maps. Combined with enhanced training techniques, this design enables high-quality generation in as few as 1 to 4 steps. Notably, we find that training flow models and subsequently converting them is more efficient and effective than training flow maps from scratch. On ImageNet 256x256 and 512x512, our models attain 1-step FID of 2.16 and 2.12, respectively, surpassing prior art by a large margin. Furthermore, we achieve FID of 1.51 and 1.68 when increasing the steps to 4, which nearly matches the performance of flow models while delivering over 100x faster inference.
机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。