Is it possible to see in results of fit_one_cycle() function also error_rate for Training set in addition to error_rate for validation set?
No it’s not, but you can always ask for data.validate(ds_type=DatasetType.train)
when the training is finished.
Thanks for the answer. But I do not find a function validate() for a class vision.data?
I found that I can call it in this way:
learn.validate(learn.data.train_dl)
Ah yes, sorry. It’s Learner.validate
and it takes a dataloader, not a DatasetType
.