CATIE-AQ/retriever-vidore-tabfquad_test_subsampled-clean

该数据集名为vidore/tabfquad_test_subsampled，包含图像、问题、答案和数据集名称等特征。数据集适用于图像特征提取任务，语言为法语。它被设计用于与VQA数据集结合使用，以便训练类似Colpali模型或其衍生模型。测试集包含280个示例，总共占用37496998字节。尽管答案列在数据集中是空的，但它是为了方便与只有问题和图像列的VQA数据集结合而设置的。

The dataset named vidore/tabfquad_test_subsampled includes features such as images, questions, answers, and dataset names. It is designed for image feature extraction tasks and is in French. The dataset is intended to be used in conjunction with VQA datasets for training Colpali-like models or their derivatives. The test set contains 280 examples, totaling 37496998 bytes. Although the answer column is empty in the dataset, it is set up to facilitate concatenation with VQA datasets that only use the question and image columns.