mthandazo/mainland_xview_conv_instruct

该数据集包含图像和消息两个主要特征。图像特征的数据类型为图像，消息特征是一个列表，包含内容和角色两个子特征。内容子特征进一步包含索引、文本和类型三个子特征，数据类型分别为整数、字符串和字符串。角色子特征的数据类型为字符串。数据集分为训练集和测试集两个分割，训练集包含40744个样本，测试集包含10186个样本。数据集的下载大小为30021136151字节，总大小为30105840033.0字节。

The dataset contains two main features: image and messages. The image feature is of type image, while the messages feature is a list containing two sub-features: content and role. The content sub-feature further contains three sub-features: index, text, and type, with data types int64, string, and string respectively. The role sub-feature is of type string. The dataset is divided into two splits: a training set with 40,744 samples and a test set with 10,186 samples. The download size of the dataset is 30,021,136,151 bytes, and the total size is 30,105,840,033.0 bytes.