Jiazheng Xu
Jiazheng Xu
Thanks for pointing this out, we do have a small number of image files in our dataset that don't exist, we'll be fixing this in the next version, you can...
This is a very interesting discovery, and I believe it may be related to the learning rate schedule and warm-up settings, although there could be other factors worth exploring.
You can refer https://github.com/xujz18/icl-selective-annotation, which is based on original codebase: https://github.com/xlang-ai/icl-selective-annotation
Your understanding is correct and I appreciate the discussion.
Thank you very much for your attention. The timing of ImageReward-v2.0 depends on the progress of the research, and communication is welcome.
你好，多谢关注！ 1. 可以的，有一些3D的Reward使用了ImageReward，例如https://github.com/THU-LYJ-Lab/T3Bench 和 https://github.com/liuff19/DreamReward 2. 只提供一个prompt和image完全可行，就是这么用的 3. 这些图片类型没有验证过，应该需要专门微调一下 > Hi, thanks for your attention! 1. it's possible, there are some 3D Rewards that use ImageReward, for example https://github.com/THU-LYJ-Lab/T3Bench https://github.com/liuff19/DreamReward...
https://github.com/THUDM/ImageReward/blob/main/train/scripts/train_one_node.sh
Thanks for your interest. We have released the test set for now, other data is still being compiled and studied.
Thank you for your interest, our dataset is now fully public.
Hello, thank you very much for your interest and welcome to the conversation. I think ImageReward's distribution of scores is approximately normal but not strictly so. It can be analogized...