shijianS01/6k-templates-mm-vsft-300k

该数据集包含消息和图像两种类型的数据。消息数据包括内容和角色两个部分，其中内容部分又细分为索引、文本和类型三个子部分。图像数据以图像格式存储。数据集分为训练集和测试集，训练集包含270,000个样本，测试集包含30,000个样本。训练集的数据大小为4,502,291,850字节，测试集的数据大小为500,254,650字节。总下载大小为2,058,480,414字节，数据集总大小为5,002,546,500字节。数据文件的路径在配置信息中指定。

The dataset contains two types of data: messages and images. The message data includes content and role, with content further divided into index, text, and type. The image data is stored in image format. The dataset is divided into a training set and a test set, with the training set containing 270,000 samples and the test set containing 30,000 samples. The training set has a data size of 4,502,291,850 bytes, and the test set has a data size of 500,254,650 bytes. The total download size is 2,058,480,414 bytes, and the total dataset size is 5,002,546,500 bytes. The paths to the data files are specified in the configuration information.