MRMRbenchmark/theorem
收藏Hugging Face2025-10-07 更新2025-10-25 收录
下载链接：
https://hf-mirror.com/datasets/MRMRbenchmark/theorem
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含三个配置：corpus、qrels和query。corpus配置包含文本和多个图像特征；qrels配置包含查询ID、语料库ID和分数；query配置包含文本、图像、问题、选项、答案等特征。数据集分为测试集，包含相应的字节数和示例数。corpus测试集包含14257个示例，qrels测试集包含1255个示例，query测试集包含579个示例。
The dataset consists of three configurations: corpus, qrels, and query. The corpus configuration includes text and multiple image features; the qrels configuration includes query_id, corpus_id, and score; the query configuration includes text, image, question, options, answer, and other features. The dataset is split into a test set, containing the respective number of bytes and examples. The corpus test set contains 14,257 examples, the qrels test set contains 1,255 examples, and the query test set contains 579 examples.
提供机构：
MRMRbenchmark
