arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2402.11260cs.CLcs.AI
收藏

MoRAL:面向LLMs终身学习的MoE增强LoRA

MoRAL: MoE Augmented LoRA for LLMs' Lifelong Learning

King Abdullah University of Science and Technology(阿卜杜拉国王科技大学);University of Macau(澳门大学);SDAIA-KAUST AI(SDAIA-KAUST人工智能机构)

Shu Yang, Muhammad Asif Ali, Cheng-Long Wang, Lijie Hu, Di Wang

AI总结:

本文提出MoRAL,结合MoE多任务能力和LoRA微调能力实现LLMs终身学习,引入5L-bench基准,实验证明其在开卷设置下学习更快、对大模型提升更显著且有效抵御灾难性遗忘。

摘要:

将大型语言模型(LLMs)适应到新领域/任务并使其成为高效的终身学习者是一个关键挑战。本文提出MoRAL,即用于终身学习的混合专家增强低秩适应(MoRAL)。MoRAL结合了MoE的多任务处理能力和LoRA的微调能力,以实现LLMs的有效终身学习。与使用事实三元组作为输入的传统方法不同,MoRAL依赖于简单的问答对,这是一种更实用且有效的策略,能够实现稳健且高效的学习。基于新的数据设置,我们引入了一个新的评估基准,即LLM终身学习基准(5L-bench),包含一个新策划的问答对数据集,以及一套评估指标,用于在开卷和闭卷设置下严格评估MoRAL。实验评估表明:(i)在开卷设置下,LLMs学习速度更快,与闭卷相比,使用MoRAL微调的Phi-2-2.7B模型在“RA”上提升了高达30.15%;(ii)MoRAL在参数数量更多的模型上表现出更高的性能提升;(iii)MoRAL对灾难性遗忘具有鲁棒性,与基线相比能更好地保留知识。

英文摘要:

Adapting large language models (LLMs) to new domains/tasks and enabling them to be efficient lifelong learners is a pivotal challenge. In this paper, we propose MoRAL, i.e., Mixture-of-Experts augmented Low-Rank Adaptation for Lifelong Learning. MoRAL combines the multi-tasking abilities of MoE with the fine-tuning abilities of LoRA for effective life-long learning of LLMs. In contrast to the conventional approaches that use factual triplets as inputs MoRAL relies on simple question-answer pairs, which is a more practical and effective strategy for robust and efficient learning. Owing to new data settings, we introduce a new evaluation benchmark namely: Life Long Learning of LLM (5L-bench) encompassing a newly curated dataset of question-answer pairs, and a set of evaluation metrics for rigorous evaluation of MoRAL in open-book and closed-book settings. Experimental evaluation shows (i) LLMs learn fast in open-book settings with up to 30.15% improvement in "RA" for Phi-2-2.7B compared to closed-book (for models fine-tuned with MoRAL); (ii) MoRAL shows higher performance improvement for models with a greater number of parameters; (iii) MoRAL is robust to catastrophic forgetting offering better knowledge retention compared to baselines.

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。

↑