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
train_ppon.py import errors #28
Comments
Thanks for trying it out! The PPON codes have not been updated to the latest changes in the repo, it's on the TODO list |
@sean-horton PPON training is now fully working in my local test repo, will commit it soon, after I do some more testing to check everything works fine |
Awesome, I look forward to trying it out. Thanks! |
@sean-horton the reworked code for PPON is now up. Now you can train PPON using the normal train.py file, only need to change these options in the option file: model: ppon Note that you may need to adjust your scheduler and coordinate the ppon_stages to match, the original paper used "StepLR_Restart". The last part is used to select which losses will be used during each phase, out of those configured and the GAN is at the moment only enabled on the last phase, following the paper. You can configure any of the losses in any of the phases, I recommend testing "contextual" (cx) if possible, specially on phases 2 and 3. |
I tried train_ppon.py and got an import error on
I tried commenting it out and the usage of lpips and there seemed to be other import errors.
I've not tried test_ppon.py, but maybe that has errors too.
The text was updated successfully, but these errors were encountered: