AI4Industry/OmniScience
收藏官方服务：
资源简介：
该数据集包含了图片、标题、原始标题、上下文、主题、原始主题、来源、数字对象标识符、链接、标题重写模型和子图信息等字段。数据集分为验证集和训练集，每个集合都包含5000个样本。
The dataset includes fields such as images, captions, raw captions, context, titles, raw subjects, sources, digital object identifiers, links, caption rewriting models, and subfigure information. The dataset is split into a validation set and a training set, each containing 5,000 samples.
提供机构：
AI4Industry
