andreaparker/small-wiki-ss-corpus-1k-rev

该数据集包含1000个样本，每个样本包含图像、文档ID、文本和标题四个特征。数据集仅包含一个训练集，总大小为315303810字节，下载大小为314304991字节。数据文件路径为data/train-*。

This dataset contains 1000 samples, each with four features: image, docid, text, and title. The dataset includes only a training set, with a total size of 315303810 bytes and a download size of 314304991 bytes. The data files are located at data/train-*.