通过动态合并核心专家,让LoRA专家模型更省参数且适应更细粒度任务。
CoMoL: Efficient Mixture of LoRA Experts via Dynamic Core Space Merging

- 用核心空间专家和路由机制,实现低开销的专家动态选择与合并。
- 在多个任务上性能超越现有方法,参数量接近标准LoRA。
- 适合追求高效微调、需细粒度适配的LLM应用开发者。
大语言模型通过参数高效微调(PEFT)在多种下游任务中表现卓越。然而,现有PEFT方法特别是MoE-LoRA架构存在参数效率低、适应粒度粗的问题,源于LoRA专家数量多及实例级路由。为此,我们提出核心空间混合LoRA(CoMoL),一种新型MoE-LoRA框架,融合专家多样性、参数效率与细粒度适应性。CoMoL引入两个关键组件:核心空间专家与核心空间路由。核心空间专家以紧凑的核心矩阵存储每个专家,保持多样性的同时控制参数增长;核心空间路由为每个令牌动态选择并激活合适的专家,实现细粒度输入自适应路由。激活的核心专家通过软合并策略合并为单一核心专家,并与共享的LoRA结合形成专用的LoRA模块。此外,路由网络被投影至与LoRA矩阵相同的低秩空间,进一步降低参数开销而不损失表达能力。大量实验表明,CoMoL在保留MoE-LoRA可适应性的同时,达到与标准LoRA相当的参数效率,且在多个任务上持续优于现有方法。
原文摘要 · Abstract (English)
Large language models (LLMs) achieve remarkable performance on diverse downstream and domain-specific tasks via parameter-efficient fine-tuning (PEFT). However, existing PEFT methods, particularly MoE-LoRA architectures, suffer from limited parameter efficiency and coarse-grained adaptation due to the proliferation of LoRA experts and instance-level routing. To address these issues, we propose Core Space Mixture of LoRA (\textbf{CoMoL}), a novel MoE-LoRA framework that incorporates expert diversity, parameter efficiency, and fine-grained adaptation. Specifically, CoMoL introduces two key components: core space experts and core space routing. Core space experts store each expert in a compact core matrix, preserving diversity while controlling parameter growth. Core space routing dynamically selects and activates the appropriate core experts for each token, enabling fine-grained, input-adaptive routing. Activated core experts are then merged via a soft-merging strategy into a single core expert, which is combined with a shared LoRA to form a specialized LoRA module. Besides, the routing network is projected into the same low-rank space as the LoRA matrices, further reducing parameter overhead without compromising expressiveness. Extensive experiments demonstrate that CoMoL retains the adaptability of MoE-LoRA architectures while achieving parameter efficiency comparable to standard LoRA, consistently outperforming existing methods across multiple tasks.
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。