Dear Organizer, I just want to know about docker output file name. I have tested my docker and my model running well and but I got mismatch size. Although my docker is running well locally and produced same size as input. It seems docker evaluate two different files. The file name of input image should be same as prediction file name? for example, input file test_036-US.nii.gz, prediction file test_036-US.nii.gz or input file test_036-US.nii.gz and prediction file test_036-label.nii.gz.? Please can you clarify clearly. I am confused the name of files convention. Most of the competitions, docker input file name should be same as prediction file name. Thanks. I have solved the problem and docker is now running. Thanks

Created by Abdul Qayyum abdulkcl
Here is the log file I got. STDERR: 2023-07-26T11:26:02.037627411Z ['goldstandard/test_001-label.nii.gz', 'goldstandard/test_002-label.nii.gz', 'goldstandard/test_003-label.nii.gz', 'goldstandard/test_004-label.nii.gz', 'goldstandard/test_005-label.nii.gz', 'goldstandard/test_006-label.nii.gz', 'goldstandard/test_007-label.nii.gz', 'goldstandard/test_008-label.nii.gz', 'goldstandard/test_009-label.nii.gz', 'goldstandard. File "/opt/monai/monai/metrics/meandice.py", line 138, in compute_dice STDERR: 2023-07-26T11:26:02.037735111Z raise ValueError(f"y_pred and y should have same shapes, got {y_pred.shape} and {y.shape}.") STDERR: 2023-07-26T11:26:02.037749013Z ValueError: y_pred and y should have same shapes, got torch.Size([1, 2, 379, 399, 197]) and torch.Size([1, 2, 346, 401, 194]). I guess input file name is test_001-US.nii.gz and GT file name is test_001-label.nii.gz? Please correct me is it true? I got this error. I am wondering the shape of input image and prediction is same in my local evaluation.

Docker input and output files names confused? page is loading…