maxshiro/coco_QA_for_visually_impaired

该数据集包含训练分割，共有200个样本，总大小约为10.35 MB。每个样本包括三个特征：instruction（字符串类型，表示指令）、image（图像类型）和output（字符串类型，表示输出）。数据集用于多模态任务，结合文本指令和图像输入生成文本输出。

This dataset includes a train split with 200 examples and a total size of approximately 10.35 MB. Each example consists of three features: instruction (string type, representing the instruction), image (image type), and output (string type, representing the output). The dataset is designed for multimodal tasks, combining text instructions and image inputs to generate text outputs.