Yannick Kirchhoff

Hi @avBuffer, the problem seems to be with the matching in [this line](https://github.com/MIC-DKFZ/nnUNet/blob/master/nnunetv2/utilities/utils.py#L47). Did you name your file for the predictions correctly, i.e. `filename_0000.file_ending`? Best, Yannick

Hi @Crystal-Biru, are your predictions saved as png files? nnUNet saves the labels as consecutive integers (0, 1, 2,...) but image viewers usually expect a range [0 - 255], so...

Hi, that should really not happen. There are a few things that might in principle go wrong, but nothing which would explain to me that the predictions are completely blank....

Hi, you can send it to [email protected]. If it is too large to send via mail I can also send you a link to upload the data. Please mention the...

Hi, yes I got it. It will take some time because I am on a conference next week but I will look into it when I am back. Best, Yannick

Hi, the link you sent me unfortunately expired. I just sent you an email with a link where you can upload the data again. Best, Yannick

Hi @Crystal-Biru, I will close this issue for now to clean up open issues a little. Feel free to reopen the issue and send me the data again if your...

Hi @RoloSmajda, that should definitely not happen. What data do you use to predict on? And can you open the test images with ITK-SNAP? Maybe they are already corrupted. Best,...

It is really weird that it works on CPU but fails on GPU. Can one or both of you share the trained models and data with me, so I can...

Hi @RoloSmajda, thanks for the model and the test cases. Unfortunately I am not able to reproduce your issue on my machine so we will need to go back to...