用一个模型搞定多种论辩分析任务，效果好还省算力。
AMELIA: A Family of Multi-task End-to-end Language Models for Argumentation
- 构建统一格式的19个论辩数据集，支持多任务训练。
- 单任务微调显著提升各任务表现，多任务微调无性能下降。
- 模型合并策略兼顾效果与效率，适合资源受限场景。
论辩挖掘是论辩研究的一个分支，旨在从自然语言文本中自动提取论辩结构及其关系。本文探讨如何利用单一大型语言模型完成一项或多项论辩挖掘任务。主要贡献有二：其一，通过调研并转换文献中的19个知名论辩挖掘数据集，构建了一个统一格式的多任务数据集；其二，基于Meta AI的Llama-3.1-8B-Instruct模型，探索了三种训练策略：(1) 单任务微调，(2) 多任务联合微调，(3) 分别微调后合并模型。实验表明，任务特定微调显著提升了所有任务的性能。同时，多任务微调在不降低性能的前提下实现了跨任务的有效知识迁移。最后，模型合并策略在保持竞争力表现的同时，有效缓解了全量多任务微调带来的计算开销，是一种可行的折衷方案。
原文摘要 · Abstract (English)
Argument mining is a subfield of argumentation that aims to automatically extract argumentative structures and their relations from natural language texts. This paper investigates how a single large language model can be leveraged to perform one or several argument mining tasks. Our contributions are two-fold. First, we construct a multi-task dataset by surveying and converting 19 well-known argument mining datasets from the literature into a unified format. Second, we explore various training strategies using Meta AI's Llama-3.1-8B-Instruct model: (1) fine-tuning on individual tasks, (2) fine-tuning jointly on multiple tasks, and (3) merging models fine-tuned separately on individual tasks. Our experiments show that task-specific fine-tuning significantly improves individual performance across all tasks. Moreover, multi-task fine-tuning maintains strong performance without degradation, suggesting effective transfer learning across related tasks. Finally, we demonstrate that model merging offers a viable compromise: it yields competitive performance while mitigating the computational costs associated with full multi-task fine-tuning.
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。