Hi, I’m trying to train a model. But when I’m calling ClassificationInterpretation.from_learner on my learner it’s returning len(data.valid_ds)
,len(losses)
,len(idxs)
with different length and I’m not sure what I’m doing wrong here.
I’m having a similiar issue with the BIWI dataset
learn_biwi = cnn_learner(data_biwi, models.resnet18, lin_ftrs=[100], ps=0.05)
learn_biwi.fit_one_cycle(5, 5e-2)
interp = ClassificationInterpretation.from_learner(learn_biwi)
losses,idxs = interp.top_losses()
len(data_biwi.valid_ds)==len(losses)==len(idxs)
# False
my validation set has the length of 40, but losses and idxs each have a length of 80