arXiv 2407.030402024-07-03cs.CLcs.AI

原始文本即所需：面向大语言模型的知识密集型多轮指令微调

Raw Text is All you Need: Knowledge-intensive Multi-turn Instruction Tuning for Large Language Model

Computer School, Beijing Information Science & Technology University（北京信息科技大学计算机学院）；Beihang University（北京航空航天大学）

Xia Hou, Qifeng Li, Jian Yang, Tongliang Li, Linzheng Chai, Xianjie Wu, Hangyuan Ji, Zhoujun Li, Jixuan Nie, Jingbo Dun, Wenfeng Song

备注：: 11 pages, 3 figures

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文提出 R2S 框架，通过对话逻辑链引导大语言模型从原始文档生成知识密集型多轮对话，构建 G I NSTRUCT 数据集并微调 GLLM，以增强跨领域指令微调效果。

摘要：

指令微调作为一种有效技术，可使大语言模型（LLMs）的输出与人类偏好对齐。但如何从原始文档中生成合乎情境的多轮对话以用于指令微调，仍有待进一步探索。本文提出一个名为 R2S 的新框架，利用“对话逻辑链（CoD-Chain of Dialogue）”引导大语言模型生成知识密集型多轮对话以用于指令微调。通过将来自开源数据集和领域特定网络爬取文档的原始文本整合到基准 K-BENCH 中，我们覆盖了多个领域，如维基百科（英文）、科学（中文）和文物（中文）。我们的方法首先确定当前对话的逻辑流程，然后提示大语言模型生成关键短语以获取相关回复内容。该方法能够构建 G I NSTRUCT 指令数据集，在对话式交互中保留原始文档知识。利用该数据集，我们对 GLLM 进行微调，该模型旨在将原始文档转换为结构化多轮对话，从而将全面的领域知识注入 SFT 模型以增强指令微调。这项工作标志着在提升大语言模型处理与生成跨领域更准确、更具上下文细微差别回复的适应性和有效性方面迈出了一步。

英文摘要：

Instruction tuning as an effective technique aligns the outputs of large language models (LLMs) with human preference. But how to generate the seasonal multi-turn dialogues from raw documents for instruction tuning still requires further exploration. In this paper, we present a novel framework named R2S that leverages the CoD-Chain of Dialogue logic to guide large language models (LLMs) in generating knowledge-intensive multi-turn dialogues for instruction tuning. By integrating raw documents from both open-source datasets and domain-specific web-crawled documents into a benchmark K-BENCH, we cover diverse areas such as Wikipedia (English), Science (Chinese), and Artifacts (Chinese). Our approach first decides the logic flow of the current dialogue and then prompts LLMs to produce key phrases for sourcing relevant response content. This methodology enables the creation of the G I NSTRUCT instruction dataset, retaining raw document knowledge within dialoguestyle interactions. Utilizing this dataset, we fine-tune GLLM, a model designed to transform raw documents into structured multi-turn dialogues, thereby injecting comprehensive domain knowledge into the SFT model for enhanced instruction tuning. This work signifies a stride towards refining the adaptability and effectiveness of LLMs in processing and generating more accurate, contextually nuanced responses across various fields.

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。