after_training_plots#
- radionets.evaluation.train_inspection.after_training_plots(conf, num_images=3, rand=False, diff=True)[source]#
Create quickly inspection plots right after the training finished. Note, that these images are taken from the validation dataset and are therefore known by the network.
- Parameters:
- confdict
contains configurations
- num_imagesint, optional
number of images to plot, by default 3
- randbool, optional
take images randomly or from the beginning of the dataset, by default False
- diffbool, optional
show the difference or the input, by default True