研究人员开发了一种树状混合策略剪枝（TMP）框架，旨在减少大规模图像生成模型的参数数量和计算需求。该框架适用于文本到图像和图像到图像任务，并支持混合专家（MoE）和扩散 transformer（DiT）等架构。实验表明，TMP能够将80B参数的HunyuanImage 3.0模型压缩至20B参数，使其能够在单块24GB GPU上进行推理，且生成质量损失极小。该框架还成功将Z-Image turbo模型从6B参数压缩到4B参数。 AI
影响 使得在消费级硬件上更高效地部署和访问大型图像生成模型成为可能。
排序理由 该集群描述了一篇详细介绍新型模型剪枝框架的学术论文。
AI 生成摘要 · Google Gemini · 来自 2 个来源。 我们如何撰写摘要 →
arXiv:2606.27089v1 Announce Type: new Abstract: Modern image generation model rapidly grows their sizes to meet high-fidelity image synthesis. However, they gradually become unaffordable for their enormous parameter consumption and computation budget that lead to massive resource…
Modern image generation model rapidly grows their sizes to meet high-fidelity image synthesis. However, they gradually become unaffordable for their enormous parameter consumption and computation budget that lead to massive resources requirement and gpu memory footprint. In this …