ConvAI2

该数据集旨在为开发对话代理提供支持，专注于理解和生成类人回应。它被用于自动化指标评测，如MaUdE，适用于开放领域对话任务。

This dataset is intended to support the development of dialogue agents, focusing on understanding and generating human-like responses. It is used for automated metric evaluation tasks such as MaUdE, and is applicable to open-domain dialogue scenarios.