Multi-class cross entropy loss and softmax in pytorch

What loss function are we supposed to use when we use the F.softmax layer?

1 Like