mm-eval/PMC-VQA

该数据集是一个多模态数据集，包含2000个测试示例，每个示例由id（唯一标识符）、media（图像列表）和messages（文本消息）组成。数据以test划分提供，总大小约253MB，下载大小约249MB。数据集主要用于处理图像和文本的交互任务，可能适用于视觉语言模型训练或评估。

This dataset is a multimodal dataset comprising 2000 test examples, each consisting of an id (unique identifier), media (a list of images), and messages (text messages). It is provided with a test split, with a total size of approximately 253 MB and a download size of about 249 MB. The dataset is primarily designed for tasks involving interaction between images and text, potentially suitable for training or evaluating vision-language models.