htyjers / StrDiffusion

[CVPR 2024] Structure Matters: Tackling the Semantic Discrepancy in Diffusion Models for Image Inpainting
Apache License 2.0
85 stars 11 forks source link

How to resume trainning? #37

Open Zzsf11 opened 3 months ago

Zzsf11 commented 3 months ago

Hi, it's really awesome work. So I'd like to train on my own dataset, but I accidently interrupt the trainning process. I have read the source code about resume checkpoint part, it is difficult for me to load checkpoint because I found the model seem like didn't save the right format (it only saved weight not optimizers and schedulers). Can you give the method to resume trainning?