How to train an SDXL to reproduce your experimental results

[torpedo2003]

How to train an SDXL to reproduce your experimental results

[rt219]

Thank you for your attention. We will release our training scripts soon. And we plan to release our SDXL safety preference dataset firstly

[torpedo2003]

Thank you for your reply. But can you give me some tips for training SD1.5 using the CoProv2 dataset? Because 'train. py' doesn't seem to be compatible with the CoProv2 dataset.