LLM4Graph
收藏资源简介:
LLM4Graph数据集是由北京邮电大学和清华大学联合创建的,旨在提升大型语言模型(LLMs)在图分析中的表现。该数据集包含29,260个问答对,内容包括从六个广泛使用的图库中爬取的文档和自动生成的代码。数据集的创建过程结合了文档信息和代码生成技术,通过指令微调来增强模型的性能。LLM4Graph数据集主要应用于增强LLMs在图分析任务中的能力,特别是在处理大规模图数据和复杂图任务方面。
The LLM4Graph dataset was jointly created by Beijing University of Posts and Telecommunications and Tsinghua University, aiming to improve the performance of Large Language Models (LLMs) in graph analysis. This dataset contains 29,260 question-answer pairs, covering documents crawled from six widely used graph libraries and automatically generated code. The construction of the LLM4Graph dataset integrates document information and code generation technologies, and enhances model performance through instruction fine-tuning. The LLM4Graph dataset is primarily applied to strengthen the capabilities of LLMs in graph analysis tasks, especially when dealing with large-scale graph data and complex graph-related tasks.




