StickerConv

该数据集专注于评估多模态对话生成，特别是针对贴纸和文本回应。它被用于为对话生成模型设定基准，同时关注同理心、一致性和多模态交互的具体指标。该数据集的任务是对话生成和评估。

This dataset focuses on evaluating multimodal dialogue generation, with a particular emphasis on sticker and text responses. It is utilized to establish benchmarks for dialogue generation models, while prioritizing specific metrics related to empathy, consistency, and multimodal interaction. The core tasks of this dataset are dialogue generation and evaluation.