Rapidata/Ideogram-V2_t2i_human_preference
收藏官方服务：
资源简介：
这是一个文本到图像的数据集，包含了195k个来自42k个 annotators 的人类响应，使用Rapidata Python API在一天内收集。该数据集用于评估Ideogram-v2模型在偏好、连贯性和对齐性三个类别上的表现，并与11个其他模型进行了1v1的比较。
This is a text-to-image dataset containing 195k human responses from 42k annotators collected in just one day using the Rapidata Python API. The dataset is used to evaluate the performance of the Ideogram-v2 model across three categories: preference, coherence, and alignment, and includes 1v1 comparisons with 11 other models.
提供机构：
Rapidata
