thchi102/cgl_finetune_train_CH

该数据集包含消息和图像两种类型的特征。消息特征由角色(role)和内容(content)两部分组成，角色和内容均为文本形式。图像特征包含了一系列图像数据。数据集划分为训练集，包含24909个样本，总大小约为3.28GB。数据集还提供了一个默认配置，用于指定训练集的数据文件。

The dataset includes two types of features: messages and images. The messages feature consists of two sub-features: role and content, both of which are strings. The images feature is a list containing image data. The dataset is split into a training set with 24,909 samples and a total size of approximately 3.28GB. The dataset also provides a default configuration for specifying the data files of the training set.