HuggingFace Daily Papers（社区热门论文）· HuggingFace Daily Papers（社区热门论文）·· 2026-07-23精选AI 评分77
SANA-Video 2.0：混合线性注意力与注意力残差实现高效视频生成
AI 导读
SANA-Video 2.0 是一个混合视频扩散 Transformer，提供 5B 和 14B 两种规模，可在单 GPU 上生成最高 720p 视频。其 Hybrid Linear-Softmax Attention 以 3:1 比例混合线性与 softmax 注意力，配合 Block Attention Residuals 将深层有效秩提升约 12%。
推荐理由
SANA-Video 2.0 视频生成模型把单 GPU 的 720p 生成压到 13 秒，比 Wan 2.2 快 120 倍，做视频应用的开发者该重新评估成本模型了。
正文
Authors:Junsong Chen, Jincheng Yu, Yitong Li, Shuchen Xue, Haozhe Liu, Jingyu Xin, Yuyang Zhao, Tian Ye, Zhangjie Wu, Zian Wang, Daquan Zhou, Ping Luo, Song Han, Enze Xie
Abstract:We introduce SANA-Video 2.0, a hybrid video diffusion transformer instantiated at 5B and 14B scales under a unified architecture. Designed to generate high-quality video up to 720p on a single GPU, SANA-Video 2.0 matches full-softmax video DiTs in quality while retaining the favorable long-sequence scaling of linear attention. To avoid quadratic attention throughout, Hybrid Linear-Softmax Attention combines gated linear attention for O(N)-dominated mixing with periodic gated-softmax anchors at a 3:1 ratio, restoring the full-rank token interactions that pure linear attention lacks. To propagate these refreshed representations across depth, Block Attention Residuals (AttnRes) route completed block summaries into later linear layers, enabling anchor-feature reuse and boosting deep-layer effective rank by ~12%. Through from-scratch training, SANA-Video 2.0 learns the complete hybrid directly rather than linearizing pretrained models, with reduced-resolution proxy studies establishing 25% softmax as the optimal quality-efficiency trade-off. With 40-step sampling, SANA-Video 2.0 achieves a VBench score of 84.30 in 13.2s at 480p on a single H100, remaining competitive with far larger softmax video DiTs at a fraction of the latency. Its compiled DiT forward pass is 3.2x faster than a matched full-softmax baseline at 720p/60s, a gap that expands with video duration. Furthermore, full-stack Sol-Engine optimization (kernel fusion, caching, and sparse attention) accelerates this hardware-friendly backbone by a further 3.58x, bringing the 5B pipeline to 13.06s at 720p/5s and making it 120x faster than Wan 2.2-A14B on one H100. Overall, our hybrid design recovers softmax-level expressiveness at substantially reduced cost, unlocking scalable long, high resolution video generation.
| Comments: | 13 pages, 9 figures, 5 tables |
| Subjects: | Computer Vision and Pattern Recognition (cs.CV) |
| Cite as: | arXiv:2607.21553 [cs.CV] |
| (or arXiv:2607.21553v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2607.21553 arXiv-issued DOI via DataCite |
Submission history
From: Junsong Chen [view email]
[v1]
Thu, 23 Jul 2026 17:36:05 UTC (33,292 KB)
来源：HuggingFace Daily Papers（社区热门论文） · arxiv.org