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'm sorry to bother you that, here is the code:
min_level_weight, max_level_weight = int(cur_level+1)-cur_level, cur_level-int(cur_level)
I wonder if the 'cur_level' need to be changed during the iterations ? And when the iteration grows bigger the max_level_weight grows bigger too
But in the code It has be set as follows and never change
I'm sorry to bother you that, here is the code:
min_level_weight, max_level_weight = int(cur_level+1)-cur_level, cur_level-int(cur_level)
I wonder if the 'cur_level' need to be changed during the iterations ? And when the iteration grows bigger the max_level_weight grows bigger too
But in the code It has be set as follows and never change
phases = {'stabilize':[0, train_kimg//batch_size], 'fade_in':[train_kimg//batch_size+1, (transition_kimg+train_kimg)//batch_size]}
The text was updated successfully, but these errors were encountered: