You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The original paper decayed the lr at half-way and two third iterations of training. I did the same thing. It is possible that earlier / later decay is the optimal way
why do you choose 40000 as the first step to change lr? it seems that smaller step of changing lr works better.
The text was updated successfully, but these errors were encountered: