arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2501.06471cs.AI
收藏

大语言模型互联网:面向通用人工智能的LLM训练与知识交换编排框架

The Internet of Large Language Models: An Orchestration Framework for LLM Training and Knowledge Exchange Toward Artificial General Intelligence

EureXa Labs(优瑞科萨实验室);Peking University(北京大学);University of Waterloo(滑铁卢大学)

Wilson Wei, Nicholas Chen, Yuxuan Li

AI总结:

针对大语言模型开发中规模大、配置复杂、功能单一、算力成本高等挑战,提出含共享协议、通用环境框架、Agent最优路径模块的编排框架及联合挖矿机制,实现算力与设计方价值共享,降低研究成本,推动LLM发展。

摘要:

本文探讨了大语言模型(Large Language Models, LLMs)开发过程中面临的多维度挑战,包括模型参数与文件规模庞大、开发环境配置复杂、模型功能单一以及计算资源成本高昂。为应对这些挑战,本文提出三项核心技术方案:LLM共享协议、LLM通用环境框架以及Agent最优路径模块。为解决研究初期的计算资源约束,我们进一步创新性地提出联合挖矿机制,实现算力提供方与模型设计者的双边价值共享,包括最优模型路径的突破奖励与长期收益分配,从而为研究者提供成本优化的计算资源支持,推动LLM研究与应用的持续发展。

英文摘要:

This paper explores the multi-dimensional challenges faced during the development of Large Language Models (LLMs), including the massive scale of model parameters and file sizes, the complexity of development environment configuration, the singularity of model functionality, and the high costs of computational resources. To address these challenges, this paper proposes three core technical solutions: LLM sharing protocol, LLM universal environment framework, and Agent optimal path module. To solve the computational resource constraints in the early stages of research, we further innovatively propose a joint mining mechanism, achieving bilateral value sharing between computing power providers and model designers, including breakthrough rewards for optimal model paths and long-term profit distribution, thereby providing researchers with cost-optimized computational resource support and promoting the continuous development of LLM research and applications.

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。

↑