Quxiaolong2024/entity_extend_triples-small

该数据集旨在让大型语言模型（llms）学习如何根据给定的实体和文本扩展三元组。每个数据单元包含三部分：指令、输入和输出。指令是固定的，要求从文本中提取与给定实体相关的三元组。输入格式为文本加实体，实体是从文本中随机采样的。输出是与实体相关的三元组，实体作为三元组的头节点或尾节点。

This dataset is designed to enable large language models (llms) to learn how to extend triples based on a given entity and text. Each data unit consists of three parts: instruction, input, and output. The instruction is fixed, requiring the extraction of triples associated with the given entity from the text. The input format is text plus entity, where the entity is randomly sampled from the text. The output is the triples associated with the entity, with the entity serving as the head or tail node of the triple.