SW-Yoon/AVS_train

该数据集包含5056个训练样本，每个样本包括id、image和conversations三个主要特征。其中，conversations是一个列表，每个列表项包含from和value两个字段，分别表示对话的来源和内容。数据集的总大小为504267093.304字节，下载大小为402223975字节。

This dataset contains 5056 training samples, each with three main features: id, image, and conversations. The conversations feature is a list, with each item containing two fields: from and value, representing the source and content of the conversation, respectively. The total size of the dataset is 504267093.304 bytes, with a download size of 402223975 bytes.