arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2402.03766cs.CVcs.AI
收藏

MobileVLM V2:更快速、更强大的视觉语言模型基线

MobileVLM V2: Faster and Stronger Baseline for Vision Language Model

Meituan Inc.(美团公司);Dalian University of Technology(大连理工大学);Zhejiang University(浙江大学)

Xiangxiang Chu, Limeng Qiao, Xinyu Zhang, Shuang Xu, Fei Wei, Yang Yang, Xiaofei Sun, Yiming Hu, Xinyang Lin, Bo Zhang, Chunhua Shen

AI总结:

MobileVLM V2通过架构设计、定制训练方案和高质量数据策展,在1.7B和3B规模上超越更大模型,成为更强的移动端视觉语言模型基线。

摘要:

我们推出MobileVLM V2,这是基于MobileVLM的一系列显著改进的视觉语言模型,证明了新颖架构设计的精细编排、针对移动端VLM定制的改进训练方案以及丰富的高质量数据集策展可以大幅提升VLM的性能。具体而言,MobileVLM V2 1.7B在标准VLM基准上达到或超过3B规模更大VLM的性能。值得注意的是,我们的3B模型在7B+规模的多种VLM中表现更优。我们的模型将在https://github.com/Meituan-AutoML/MobileVLM 发布。

英文摘要:

We introduce MobileVLM V2, a family of significantly improved vision language models upon MobileVLM, which proves that a delicate orchestration of novel architectural design, an improved training scheme tailored for mobile VLMs, and rich high-quality dataset curation can substantially benefit VLMs' performance. Specifically, MobileVLM V2 1.7B achieves better or on-par performance on standard VLM benchmarks compared with much larger VLMs at the 3B scale. Notably, our 3B model outperforms a large variety of VLMs at the 7B+ scale. Our models will be released at https://github.com/Meituan-AutoML/MobileVLM .

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑