nnunet_run_training script

Run nnUNetv2_train command to start nnUNet training for the specified fold.

usage: nnunet_run_training [-h] --config-file CONFIG_FILE [--run-fold [-1-4]]
                           [--run-validation-only RUN_VALIDATION_ONLY]
                           [--post-processing-folds POST_PROCESSING_FOLDS [POST_PROCESSING_FOLDS ...]]
                           [--output-model-file OUTPUT_MODEL_FILE]
                           [--resume-training RESUME_TRAINING]
                           [--n-workers N_WORKERS] [-v | -q]

Named Arguments

--config-file

File path for the configuration dictionary, used to retrieve experiments variables (Task_ID)

--run-fold

Possible choices: -1, 0, 1, 2, 3, 4

int value indicating which fold (in the range 0-4) to run

Default: 0

--run-validation-only

Flag to run only the Validation step ( after the Training step is completed). Default no.

Default: no

--post-processing-folds

Trained Folds to include in the post-processing and model export. Default -1 (All Folds are used).

Default: '-1'

--output-model-file

File Path where to save the zipped Model File.

--resume-training

Flag to indicate training resume after stopping it. Default no.

Default: no

--n-workers

Number of parallel processes used when pre-processing and unpacking the image data (Default: N_THREADS)

-v, --verbose

Enable verbose output in terminal. Add multiple times to increase verbosity.

-q, --silent

Suppress most log outputs in terminal.

Example call:

nnunet_run_training.py --config-file ../CONFIG_FILE.json --run-fold 0
nnunet_run_training.py --config-file ../CONFIG_FILE.json --run-fold 0 --resume-training y