The model was not set to evaluation mode in the notebook before evaluating the parameters on test. We included dropout layers. They will randomly set a fraction of the inputs to zero during training, which helps prevent overfitting. However, dropout is also active during evaluation if the model is not set to evaluation mode.
I added weibull_model.eval() and cox_model.eval() before the parameters evaluation on data from dataloader_test.
The model was not set to evaluation mode in the notebook before evaluating the parameters on test. We included dropout layers. They will randomly set a fraction of the inputs to zero during training, which helps prevent overfitting. However, dropout is also active during evaluation if the model is not set to evaluation mode.
I added
weibull_model.eval()andcox_model.eval()before the parameters evaluation on data fromdataloader_test.