将170亿参数大模型压缩至24亿，实现手机端快速高质量文生图
NanoFLUX: Distillation-Driven Compression of Large Text-to-Image Generation Models for Mobile Devices
- 通过渐进式压缩删除扩散模型冗余组件，参数从120亿减至20亿
- 采用基于ResNet的令牌下采样机制，使中间层处理低分辨率特征，降低延迟
- 利用去噪器早期视觉信号优化文本编码器蒸馏，提升生成质量
尽管大规模文生图扩散模型在视觉质量上持续进步，其不断增长的规模仍拉大了顶尖模型与设备端方案之间的差距。为弥合这一差距，我们提出NanoFLUX，一个由170亿参数的FLUX.1-Schnell蒸馏而来的24亿参数文本到图像流匹配模型，采用渐进式压缩流程以保持生成质量。主要贡献包括：（1）一种由剪枝驱动的模型压缩策略，通过移除扩散Transformer中的冗余组件，将模型大小从120亿降至20亿；（2）一种基于ResNet的令牌下采样机制，允许中间模块在低分辨率令牌上操作，同时保留其他部分的高分辨率处理，从而降低延迟；（3）一种新颖的文本编码器蒸馏方法，利用采样过程中去噪器早期层的视觉信号。实验证明，NanoFLUX在移动端可约2.5秒生成512×512图像，证明了高质量文生图在设备端的可行性。
原文摘要 · Abstract (English)
While large-scale text-to-image diffusion models continue to improve in visual quality, their increasing scale has widened the gap between state-of-the-art models and on-device solutions. To address this gap, we introduce NanoFLUX, a 2.4B text-to-image flow-matching model distilled from 17B FLUX.1-Schnell using a progressive compression pipeline designed to preserve generation quality. Our contributions include: (1) A model compression strategy driven by pruning redundant components in the diffusion transformer, reducing its size from 12B to 2B; (2) A ResNet-based token downsampling mechanism that reduces latency by allowing intermediate blocks to operate on lower-resolution tokens while preserving high-resolution processing elsewhere; (3) A novel text encoder distillation approach that leverages visual signals from early layers of the denoiser during sampling. Empirically, NanoFLUX generates 512 x 512 images in approximately 2.5 seconds on mobile devices, demonstrating the feasibility of high-quality on-device text-to-image generation.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。