Getting training accuracy in addition to validation accuracy in learn.fit()

Is there a way to ouput training accuracy in addition to training loss, validation loss and validation accuracy from learn.fit(1e-2,5) (the learner object).It is really necessary that i plot validation accuracy along with training accuracy for my project. I tried to look inside the fastai libraries but could not find any information.

2 Likes

The learn.sched object contains lots of useful information, including, e.g., learn.sched.val_losses.

I saw that there is no such object anymore in Learner. Do you know the current way to get training set accuracy after the learning?

1 Like