Computing Loss in PyTorch

How it works:

How the implementation in PyTorch looks like:

 

Rule of thumb: The more accurate the network, the smaller the loss.

reference: Datacamp's Introduction to Deep Learning with PyTorch