Rapidata/OpenGVLab_Lumina_t2i_human_preference
收藏官方服务:
资源简介:
这是一个文本到图像的评估数据集,包含了超过400k的人类反馈,用于评估Lumina模型在偏好、连贯性和一致性三个类别上的表现。数据集包含文本提示、图像、模型名称以及相关的加权结果和详细结果。
This is a text-to-image evaluation dataset containing over 400k human responses, used to assess the Lumina models performance in three categories: preference, coherence, and alignment. The dataset includes text prompts, images, model names, and related weighted results and detailed results.
提供机构:
Rapidata


