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
It is not an issue. I just want to ask why do you use running_loss = running_loss*0.9 + loss.item()*0.1 for monitoring the loss during training?
Do you have any special reason for this?
Isnt it conventional to monitor the average loss after each epoch (in this case, after each iteration)?
The text was updated successfully, but these errors were encountered:
It is not an issue. I just want to ask why do you use running_loss = running_loss*0.9 + loss.item()*0.1 for monitoring the loss during training?
Do you have any special reason for this?
Isnt it conventional to monitor the average loss after each epoch (in this case, after each iteration)?
The text was updated successfully, but these errors were encountered: