QA-TOOLBOX
收藏资源简介:
QA-TOOLBOX数据集由多个研究机构合作创建,旨在为制造业中的任务指导系统提供数据支持。该数据集包含超过200,000条问答对,这些问答对基于技术人员在高级制造环境中的实际交互,并参考了规范文档和视频演示。数据集的创建过程结合了规范文档的步骤描述和任务执行的叙述,通过大型语言模型(LLM)进行数据增强。该数据集主要应用于制造业中的任务指导和AI辅助系统,旨在帮助技术人员更有效地执行复杂任务。
The QA-TOOLBOX dataset was collaboratively developed by multiple research institutions, designed to provide data support for task guidance systems in the manufacturing industry. This dataset contains over 200,000 question-answer pairs, which are derived from the actual on-site interactions of technicians in advanced manufacturing environments, and compiled with reference to standard technical documents and video demonstrations. The dataset's development workflow integrates step-by-step descriptions from official specifications and narratives of task execution, and leverages Large Language Models (LLMs) for data augmentation. This dataset is primarily utilized for task guidance and AI-assisted systems in the manufacturing sector, with the objective of enabling technicians to perform complex tasks more efficiently.




