TIGER-Lab/EditReward-Bench
收藏Hugging Face2025-12-01 更新2025-10-18 收录
下载链接：
https://hf-mirror.com/datasets/TIGER-Lab/EditReward-Bench
下载链接
链接失效反馈官方服务：
资源简介：
EditReward-Bench是一个针对指令引导的图像编辑任务的人类对齐奖励模型基准，包含了超过200K偏好对的大型人类偏好数据集EditReward-Data。该基准旨在评估模型在指令引导的图像编辑任务中对人类偏好的对齐程度。
EditReward-Bench is a human-aligned reward model benchmark for instruction-guided image editing tasks, which includes a large-scale human preference dataset called EditReward-Data with over 200K preference pairs. The benchmark is designed to evaluate the alignment of models with human preferences in instruction-guided image editing tasks.
提供机构：
TIGER-Lab
