Installation (on cluster):
- Run setup.sh
- Setup python virtual env
- Install Pytorch (cuda 11.2)
- Install nnUnet (from requirements.txt)
Usage:
- Run nnUnet_inference.sh
- Loads virtual env
- Sets enviornment varibles
- DATA - Sample data
- RESULTS_FOLDER - Directory tree with model
- OUTPUT_FOLDER - Where to store results
- nnUNet_preprocessed - Where to store intermediate results
- Runs nnUnet on GPU