-
Notifications
You must be signed in to change notification settings - Fork 1.7k
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
Pseudo and EMA pseudo dice are 0.0 #2318
Comments
I have been able to resolve this issue. The problem was with the default initial learning rate. I set it to a lower value (1e-4) and the model converges now. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
I am using nnUNetv2 for a tumor segmentation task with dual channel input, PT and CT. Using the 2d configuration works well and I obtain realistic dice scores on the dataset. However, when I use the 3d_fullres configuration, I obtain 0 on both pseudo and EMA pseudo dice after about 10 epochs. I did not obtain any errors whatsoever in the preprocessing step. I have attached the learning curves. Is there something I am missing that could be causing this behaviour? I would greatly appreciate help on this, thank you.
![progress](https://private-user-images.githubusercontent.com/70821652/343010342-a2b354c2-8c5c-49bf-80dd-c10c10398c38.png?jwt=eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.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.83FFACq-6cAKtmLs-HmmtW2oUWescb4XtS2czb-kKdo)
The text was updated successfully, but these errors were encountered: