g-ronimo/imagenet-1k-vl-enriched_moondream2

这是一个包含图像ID和标题的视觉数据集。数据集由经过 vikhyatk/moondream2 重新配字的 visual-layer/imagenet-1k-vl-enriched 数据集构成，包含简短的标题。数据集仅包含图像ID和标题信息，不包括图像本身。标题是基于提示生成的，提示内容为：这是一张 {class_name} 的图片。当前图片的标题是 {caption}。请根据图片内容和当前标题写一个简短的标题。保持简短和精确。

This is a visual dataset containing image IDs and captions. The dataset is composed of the visual-layer/imagenet-1k-vl-enriched dataset that has been recaptioned using vikhyatk/moondream2, featuring short captions. The dataset includes only the image IDs and captions, without the actual images. The captions are generated based on a prompt which reads: This is an image of a {class_name}. The current image caption is {caption}. Please write a short caption based on the image content and the current caption. Keep it short and precise.