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
I think we usually tune the batch_size to be small as possible while making good estimates of the gradient and that shouldn't depend on the dataset size.
Maybe we could accept both integers and fractions and just interpret (0-1] as fractions and >1 as batch_size?
This would be much easier to tune
The text was updated successfully, but these errors were encountered: