Knowledge Commons of Institute of Automation,CAS
MoDE-CoTD: Chain-of-Thought Distillation for Complex Reasoning Tasks with Mixture of Decoupled LoRA-Experts | |
Xiang Li1,2![]() ![]() ![]() | |
2024-05-20 | |
会议名称 | The 2024 Joint International Conference on Computational Linguistics, Language Resources and Evaluation |
会议日期 | 2024.5.20 - 2024.5.25 |
会议地点 | Torino (Italia) |
出版者 | ELRA and ICCL |
摘要 | Chain-of-thought Distillation (CoTD) aims at distilling Chain-of-thought (CoT) reasoning ability of large language models (LLMs) to much smaller student models. The core of CoTD is using a large teacher model to generate rationales and fine-tune smaller student models. However, current Chain-of-thought Distillation works have the following limitations: 1) Student models are separately distilled from specific reasoning tasks and lack a collaboration mechanism, hindering the enhancement of reasoning performance through collaboration among various reasoning tasks. 2) The parameter update of student models severely harms the CoT reasoning ability on other unseen reasoning tasks not included in the distillation process. In this work, we introduce a novel CoT Distillation method, MoDE-CoTD, which decouples the CoT reasoning abilities out of the student model by distilling multiple LoRA-Experts and freezing the parameters of the student model. Sequentially, LoRA-Experts are combined and adapted to handle both seen and unseen reasoning tasks, enabling collaboration among diverse reasoning tasks to further enhance CoT reasoning performance. Experimental results on 14 datasets (including 4 unseen datasets) demonstrate the strength of MoDE-CoTD, with an average accuracy gain of 6.3% on seen datasets and 7.8% on unseen datasets. |
七大方向——子方向分类 | 自然语言处理 |
国重实验室规划方向分类 | 语音语言处理 |
是否有论文关联数据集需要存交 | 否 |
文献类型 | 会议论文 |
条目标识符 | http://ir.ia.ac.cn/handle/173211/57448 |
专题 | 复杂系统认知与决策实验室 |
通讯作者 | Shizhu He |
作者单位 | 1.Institute of Automation, Chinese Academy of Sciences 2.School of Artificial Intelligence, University of Chinese Academy of Sciences 3.Guangdong OPPO Mobile Telecommunications Corp.,Ltd |
第一作者单位 | 中国科学院自动化研究所 |
通讯作者单位 | 中国科学院自动化研究所 |
推荐引用方式 GB/T 7714 | Xiang Li,Shizhu He,Jiayu Wu,et al. MoDE-CoTD: Chain-of-Thought Distillation for Complex Reasoning Tasks with Mixture of Decoupled LoRA-Experts[C]:ELRA and ICCL,2024. |
条目包含的文件 | 下载所有文件 | |||||
文件名称/大小 | 文献类型 | 版本类型 | 开放类型 | 使用许可 | ||
3178_Paper.pdf(1062KB) | 会议论文 | 开放获取 | CC BY-NC-SA | 浏览 下载 |
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。
修改评论