Giters

MIC-DKFZ / nnDetection

nnDetection is a self-configuring framework for 3D (volumetric) medical object detection which can be applied to new data sets without manual intervention. It includes guides for 12 data sets that were used to develop and evaluate the performance of the proposed method.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

When is plan_inference.pkl generated?

dzeego opened this issue · comments

commented

Hello @mibaumgartner,

After finishing the training (multiprocessing error at the last epoch persists) I am trying to run nndet_predict on my test set which is located in the raw_splitted folder in imagesTs.
However, when i launch nndet_predict 0003 RetinaUNetV001_D3V001_3d --fold 0 I get:
FileNotFoundError: [Errno 2] No such file or directory: '/datadrive/nndetection_data/nndetection_models/Task0003_LesionDetection/RetinaUNetV001_D3V001_3d/fold0/plan_inference.pkl'

At what stage does the nndetection generate this file? and how can I mitigate this issue moving forward?

Also for some reason the val_analysis, val_analysis_preprocessed, val_predictions, val_predictions_preprocessed, val_results, and val_results_preprocessed folders were not generated. Is that normal?

Thank you in advance,
Best regards !

commented

Dear @dzeego ,

did you run the training command with the --sweep option? - that should generate the plan and results
nndet_sweep can be used to perform this manually -> it will predict the validaiton cases and derive the best set of postprocessing paramters for the validation set (those are the inference parameters which are saved into the plan_inference.pkl.

commented

This issue is stale because it has been open for 30 days with no activity.

commented

This issue was closed because it has been inactive for 14 days since being marked as stale.