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 2401.13601cs.CL
收藏

MM-LLMs:多模态大语言模型的最新进展

MM-LLMs: Recent Advances in MultiModal Large Language Models

Tencent AI Lab(腾讯AI实验室);Kyoto University(京都大学);Mohamed bin Zayed University of Artificial Intelligence(穆罕默德·本·扎耶德人工智能大学)

Duzhen Zhang, Yahan Yu, Jiahua Dong, Chenxing Li, Dan Su, Chenhui Chu, Dong Yu

AI总结:

本文综述了多模态大语言模型(MM-LLMs)的最新进展,涵盖架构设计、训练流程、126个模型分类、基准性能及关键训练秘诀,并展望未来方向,旨在促进该领域研究。

摘要:

在过去一年中,多模态大语言模型(MM-LLMs)取得了实质性进展,通过经济高效的训练策略增强现成的LLMs以支持多模态输入或输出。由此产生的模型不仅保留了LLMs固有的推理和决策能力,还赋能了多样化的多模态任务。在本文中,我们提供了一项全面的综述,旨在促进MM-LLMs的进一步研究。首先,我们概述了模型架构和训练流程的通用设计公式。随后,我们介绍了一个包含126个MM-LLMs的分类体系,每个模型都有其特定的公式。此外,我们回顾了选定的MM-LLMs在主流基准上的性能,并总结了增强MM-LLMs效能的关键训练秘诀。最后,我们探索了MM-LLMs的有前景方向,同时维护一个实时跟踪网站以获取该领域的最新进展。我们希望这项综述有助于MM-LLMs领域的持续进步。

英文摘要:

In the past year, MultiModal Large Language Models (MM-LLMs) have undergone substantial advancements, augmenting off-the-shelf LLMs to support MM inputs or outputs via cost-effective training strategies. The resulting models not only preserve the inherent reasoning and decision-making capabilities of LLMs but also empower a diverse range of MM tasks. In this paper, we provide a comprehensive survey aimed at facilitating further research of MM-LLMs. Initially, we outline general design formulations for model architecture and training pipeline. Subsequently, we introduce a taxonomy encompassing 126 MM-LLMs, each characterized by its specific formulations. Furthermore, we review the performance of selected MM-LLMs on mainstream benchmarks and summarize key training recipes to enhance the potency of MM-LLMs. Finally, we explore promising directions for MM-LLMs while concurrently maintaining a real-time tracking website for the latest developments in the field. We hope that this survey contributes to the ongoing advancement of the MM-LLMs domain.

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。

↑