Tot_epochs in fit_one_cycle


#1

learn.fit_one_cycle(cyc_len=1, tot_epochs=2)

This completes after just 1 epoch though the arguments suggest it should run for a total of 2 epochs. I don’t see any documentation on tot_epochs. What does it do?

Incidentally if you run learn.fit_one_cycle(cyc_len=2, tot_epochs=1) then it completes after 1 epoch but with the top progress bar showing only 50% which feels like an bug.