liuhuohuo2/pick-a-pic-v2

pickapic_v2数据集包含图片对及其相关信息，每个样本包括两个图片的字节数据（jpg_0和jpg_1），图片的唯一标识符（image_0_uid和image_1_uid），图片的URLs，是否为不同图片的标识（are_different），图片的描述（caption），创建时间（created_at），是否具有标签的标识（has_label），标签值（label_0和label_1），模型标识（model_0和model_1），排名ID（ranking_id），用户ID（user_id），每个提示的示例数量（num_example_per_prompt）等。数据集分为训练集、验证集、测试集，以及具有唯一样本的验证集和测试集。

The pickapic_v2 dataset contains pairs of images and their associated information. Each sample includes the byte data of two images (jpg_0 and jpg_1), unique identifiers for the images (image_0_uid and image_1_uid), URLs for the images, a boolean indicating whether the images are different (are_different), captions for the images (caption), creation timestamps (created_at), a boolean indicating whether the images have labels (has_label), label values (label_0 and label_1), model identifiers (model_0 and model_1), ranking IDs (ranking_id), user IDs (user_id), and the number of examples per prompt (num_example_per_prompt). The dataset is split into training, validation, and test sets, as well as validation and test sets with unique samples.