data-is-better-together/open-image-preferences-v1-binarized

Open Image Preferences项目旨在创建10K文本到图像的偏好对，用于评估图像生成模型在不同图像类别中的表现。项目通过使用来自fal/imgsys-results的提示，并根据复杂性和质量进行演化，然后让社区为每个提示的两个生成图像进行偏好标注。最终，项目成功标注了10K偏好对。

The dataset aims to create 10,000 text-to-image preference pairs for evaluating the performance of image generation models across a wide variety of common image categories. The dataset uses prompts from a specific source, evolves them based on complexity and quality, and then annotates the preference between two generated images by the community. Ultimately, the dataset contains 10,000 preference pairs.