MahmoodAnaam/vqav2-ar-en-validation

该数据集包含图像和文本数据，主要特征包括元数据（metadata）、图像（image）、问题（question）、答案（answers）和多项选择答案（multiple_choice_answer）。元数据包括图像ID、问题ID、问题类型和答案类型；问题部分包含英文和阿拉伯文；答案部分包含英文、阿拉伯文、置信度和ID；多项选择答案也包含英文和阿拉伯文。数据集的分割为验证集，包含214,354个样本，总大小为33,728,130,217.41字节，下载大小为6,929,167,572字节。

This dataset contains image and text data, with main features including metadata, image, question, answers, and multiple choice answer. Metadata includes image ID, question ID, question type, and answer type; the question section contains English and Arabic; the answers section contains English, Arabic, confidence, and ID; the multiple choice answer also contains English and Arabic. The dataset is split into a validation set, containing 214,354 samples, with a total size of 33,728,130,217.41 bytes and a download size of 6,929,167,572 bytes.