Some optimizers skip the first iteration and only yield for i > 0
rueckstiess opened this issue · comments
why?
consistency!
BECAUSE!! :)
Can you point me to where that happens?
https://github.com/BRML/climin/blob/master/climin/gd.py#L30
many minimizers have that "i > 0" condition: gd, rprop, *bfgs family but others don't, e.g. bayesian.
Ah, ok, thanks.