takara-ai/multimodal-icon-instruct

该数据集包含了具有四个字段的样本：名称(name)，提示(prompt)，SVG图像描述(svg)和图片(image)。数据集主要包含一个训练集，共有32196个样本，总数据大小约为682MB。数据集支持默认配置，其中定义了训练集的数据文件。

The dataset consists of samples with four fields: name, prompt, SVG image description (svg), and image. It primarily includes a training set with a total of 32,196 samples, with a total dataset size of approximately 682MB. The dataset supports a default configuration, which specifies the data files for the training set.