Nayana-cognitivelab/VQA-SFT-test

这是一个包含图像和文本问题的多语言数据集，支持跨语言问答。数据集分为不同的配置，每个配置包含训练集。每种配置都有源语言和目标语言，以及描述性和多项选择题类型的问题。数据集还包括图像的源语言、问题语言和答案语言的信息。

This is a multilingual dataset containing images and text-based questions, supporting cross-lingual question answering. The dataset is divided into different configurations, each with a training split. Each configuration includes source and target languages, as well as descriptive and multiple-choice question types. The dataset also includes information about the image source language, question language, and answer language.