Harry Potter personified dataset (HP dataset)

该数据集包含了《哈利·波特》小说中的6个虚构角色，这些角色代表了不同的年龄、性别、身份和信息丰富度，旨在评估个性化训练的语言模型。在GPT-4-turbo的帮助下进行了标注，并根据剧情信息将数据分为早期和晚期阶段。该数据集规模涵盖6个角色，拥有大量的对话和上下文数据，其任务是进行个性化语言模型的训练和数值评估。

This dataset contains six fictional characters from the *Harry Potter* novel series. These characters span a variety of ages, genders, identities and information richness levels, and are designed for the training and quantitative evaluation of personalized large language models. Annotated with the assistance of GPT-4-turbo, the dataset is split into early and late stages based on the plot information of the novels. Featuring a large volume of dialogue and contextual data centered on these six characters, this dataset supports the training and quantitative assessment of personalized large language models.