arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2501.15427cs.CL

OpenCharacter:利用大规模合成角色设定训练可定制的角色扮演大语言模型

OpenCharacter: Training Customizable Role-Playing LLMs with Large-Scale Synthetic Personas

Tencent AI Lab(腾讯AI实验室)

Xiaoyang Wang, Hongming Zhang, Tao Ge, Wenhao Yu, Dian Yu, Dong Yu

AI总结:

本文提出基于Persona Hub大规模合成角色档案,并通过回复重写与生成构建角色对齐指令数据,用LLaMA-3 8B进行SFT,使模型在角色扮演对话上达到接近GPT-4o的表现。

摘要:

大语言模型(LLM)中的可定制角色扮演,也称为角色泛化,因其在开发和部署角色扮演对话智能体方面的多功能性与成本效益,正受到越来越多的关注。本研究探索了一种大规模数据合成方法,以赋予LLM角色泛化能力。我们首先利用来自Persona Hub的角色设定合成大规模角色档案,随后探索两种策略:回复重写和回复生成,以创建与角色对齐的指令回复。为验证我们用于角色泛化的合成指令调优数据的有效性,我们使用LLaMA-3 8B模型进行了监督微调(SFT)。我们性能最佳的模型增强了原始LLaMA-3 8B Instruct模型,并在角色扮演对话上取得了与GPT-4o模型相当的性能。我们公开了合成角色和指令调优对话,以支持公共研究。

英文摘要:

Customizable role-playing in large language models (LLMs), also known as character generalization, is gaining increasing attention for its versatility and cost-efficiency in developing and deploying role-playing dialogue agents. This study explores a large-scale data synthesis approach to equip LLMs with character generalization capabilities. We begin by synthesizing large-scale character profiles using personas from Persona Hub and then explore two strategies: response rewriting and response generation, to create character-aligned instructional responses. To validate the effectiveness of our synthetic instruction tuning data for character generalization, we perform supervised fine-tuning (SFT) using the LLaMA-3 8B model. Our best-performing model strengthens the original LLaMA-3 8B Instruct model and achieves performance comparable to GPT-4o models on role-playing dialogue. We release our synthetic characters and instruction-tuning dialogues to support public research.

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。

↑