dddraxxx/refchartqa-charxiv-val-sample

该数据集是一个包含图像和文本信息的综合数据集，适用于视觉问答和对话系统。它包含图像特征、数据源信息、奖励模型、额外信息（包括答案、定位框、ID、问题和问题类型）、提示信息（包括内容和角色）、能力和环境名称等字段。数据集分为训练集，共有20个示例。额外信息中的定位框提供了问题的具体视觉定位信息。

This dataset is a comprehensive dataset containing both image and text information, suitable for visual question answering and dialogue systems. It includes features such as image features, data source information, reward model, extra information (including answer, bounding boxes, ID, question, and question type), prompt information (including content and role), ability, and environment name. The dataset is split into a training set with a total of 20 examples. The bounding boxes in the extra information provide specific visual定位 information for the questions.