-
Notifications
You must be signed in to change notification settings - Fork 2
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
main分支复现 #29
Comments
这个需要加载swin模型,同时在训练初期看训练是否loss下降明显,不行重训练,训练代码我重构过,改了一些参数不过sad 19以下还是能做到。 |
您好,这个有加载swin,在model.py里有加载。loss最后下降到0.05-0.06左右。 |
我训练了3次,都是差不多的结果。loss最后下降到0.05-0.06左右。 |
应该没那么高,我记得最后磨到更低,可能是有些参数重构时候不合适?Loss 0.02717635196633637 AlphaLoss 0.03338925188872963 Alpha2Loss 0.03338925188872963 0.005261471119010821 L_img 0.005261471119010821 L_img2_ 0.005261471119010821 0.015701981075108052 |
应该差不多这个数量级…… |
数量级好像差不多,但是高一些。您的意思是这个分支的复现和论文的结构不一致导致的吗? |
我重构时候的确为了可读性重写了不少…… |
您好,感谢您的工作,我用torch1.13环境进行复现,其他环境都符合Readme要求,并且没有修改代码,在DIM上,测试结果如下,整体结果相比论文报告结果差了很多,请问可能是什么原因?谢谢。
Average:
Whole Image: MSE: 0.0018012286659430776 SAD: 26.623746548828123
Unknown Region: MSE: 0.004980672549931798 SAD: 26.623746548828123
The text was updated successfully, but these errors were encountered: