Danescu-Niculescu-Mizil and Lee Dataset

该数据集包含了10,292对之间的220,579次对话交流，这些数据被用于训练谷歌机器人。此外，该数据集被划分为训练集和测试集，以便对模型进行评估。数据规模达到了220,579次交流，任务是对话响应生成。

This dataset contains 220,579 conversational exchanges between 10,292 pairs of interlocutors, which is utilized for training Google bots. Furthermore, the dataset is divided into training and test sets to facilitate model evaluation. With a total of 220,579 exchanges, the core task of this dataset is conversational response generation.