svg-hub/svg-stack-annotated

该数据集包含三个特征：文件名(Filename)、SVG格式图像(Svg)和简短描述(short_captions)。数据集分为验证集、测试集和训练集，分别包含不同数量的示例。验证集包含108,456个示例，测试集包含5,709个示例，训练集包含1,743,717个示例。数据集的总下载大小为4,255,013,228字节，整个数据集的大小为4,372,246,211字节。

The dataset includes three features: filename (Filename), SVG formatted images (Svg), and short descriptions (short_captions). It is divided into three splits: validation set, test set, and training set, each containing a different number of examples. The validation set has 108,456 examples, the test set has 5,709 examples, and the training set has 1,743,717 examples. The total download size of the dataset is 4,255,013,228 bytes, and the entire dataset size is 4,372,246,211 bytes.