ToolQA

ToolQA数据集由佐治亚理工学院计算机学院创建，旨在评估大型语言模型（LLMs）利用外部工具解答问题的能力。该数据集通过自动化三阶段流程构建，包括参考数据收集、模板引导的问题生成和程序化答案生成。ToolQA涵盖8个不同领域，每个问题都需要使用特定的工具从外部知识库中获取信息来解答，确保LLMs不能仅依赖内部知识。数据集的应用领域包括增强LLMs在复杂问题解答中的工具使用和逻辑推理能力，特别是在需要多步骤推理的场景中。

The ToolQA dataset was developed by the School of Computer Science at Georgia Institute of Technology, with the goal of evaluating the ability of Large Language Models (LLMs) to leverage external tools for question answering. This dataset is constructed through an automated three-stage pipeline, which consists of reference data collection, template-guided question generation, and programmatic answer generation. ToolQA covers 8 distinct domains, and each question necessitates retrieving information from external knowledge bases using specific tools, thus ensuring that LLMs cannot rely solely on their internal knowledge. The application scenarios of ToolQA focus on enhancing the tool utilization and logical reasoning capabilities of LLMs in complex question answering, particularly in scenarios requiring multi-step reasoning.