When should I stop learning?
StepHaze opened this issue · comments
If (after self-play and learning) the network has stopped replacing itself (<55% points ?) during 5-10 iterations, should I stop learning?
There is still a small possibility the network is going to improve later in this case but I would probably stop the experiment and investigate better hyperparameters in such a case.
Thanks Jonathan.
What hyperparameters should I pay attention first of all?