Stabilizing, Scaling & Enhancing MeanFlow for Large-scale Diffusion Distillation
Researchers have developed a new framework to stabilize and enhance MeanFlow, a technique used for distilling large-scale diffusion models. The method introduces a warm-up phase with a discrete solution before switching to the differential solution for refinement. Additionally, it incorporates trajectory distribution alignment to mitigate "mean-seeking bias" during few-step inference. This approach has demonstrated superior performance when applied to models like FLUX.1-dev and the 80B-parameter HunyuanImage 3.0. AI
IMPACT Enhances distillation efficiency for large diffusion models, potentially speeding up inference and deployment.