The code was developed on top of the excellent nnUNet library. Please refer to the original repo for the installation, usages, and common Q&A
Pytorch 1.9.1 Cuda 11.4 RTX3090.
You can train the models that we used for the competition using the command:
nnUNet_train 3d_fullres trainer_name <TASK_ID> <FOLD> --npz
nnUNet_train 3d_fullres trainer_name <TASK_ID> <FOLD> --npz
This repo borrowed heavily from nnUNet library and axial attention