70亿参数视频生成模型仅用66.5万小时显卡训练，性能媲美更大模型。
Seaweed-7B: Cost-Effective Training of Video Generation Foundation Model
- 采用中等规模扩散模型架构，优化训练策略降低资源消耗。
- 在66.5万H100 GPU小时下达到领先大模型的生成效果。
- 适合资源有限的研究者快速复现与轻量微调应用。
本技术报告提出一种低成本训练视频生成基础模型的方法。我们构建了一个约70亿参数（7B）的中等规模研究模型——Seaweed-7B，从头开始训练，共使用66.5万小时的H100 GPU算力。尽管计算资源有限，Seaweed-7B仍表现出与当前更大规模视频生成模型相当甚至更优的性能。在资源受限环境下，设计选择尤为关键。本文重点分析了提升中等规模扩散模型性能的关键决策。实证观察显示：(1) Seaweed-7B性能可媲美甚至超越在更大算力上训练的大型模型；(2) 该模型具备强泛化能力，可通过轻量微调或持续训练有效适配多种下游任务。详见项目页：https://seaweed.video/
原文摘要 · Abstract (English)
This technical report presents a cost-efficient strategy for training a video generation foundation model. We present a mid-sized research model with approximately 7 billion parameters (7B) called Seaweed-7B trained from scratch using 665,000 H100 GPU hours. Despite being trained with moderate computational resources, Seaweed-7B demonstrates highly competitive performance compared to contemporary video generation models of much larger size. Design choices are especially crucial in a resource-constrained setting. This technical report highlights the key design decisions that enhance the performance of the medium-sized diffusion model. Empirically, we make two observations: (1) Seaweed-7B achieves performance comparable to, or even surpasses, larger models trained on substantially greater GPU resources, and (2) our model, which exhibits strong generalization ability, can be effectively adapted across a wide range of downstream applications either by lightweight fine-tuning or continue training. See the project page at https://seaweed.video/
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。