luandro/mapeo-docs

该数据集包含两种配置：ingested和summarized。ingested配置包含文档ID、文本内容、文件名和文件大小。summarized配置在ingested的基础上增加了原始文档摘要、生成的摘要和摘要模型名称。数据集仅包含训练集，每个样本都包含了上述特征信息。

The dataset consists of two configurations: ingested and summarized. The ingested configuration includes document ID, text content, filename, and file size. The summarized configuration adds the original document summary, generated summary, and the name of the summarization model to the ingested features. The dataset contains only the training set, with each sample having the aforementioned feature information.