Apply loss function if RNN is the output layer

Author: castelao

src/nf/nf_network_submodule.f90

@@ -317,6 +317,11 @@ pure module subroutine backward(self, output)
               self % layers(n - 1), &
               quadratic_derivative(output, this_layer % output) &
             )
+          type is(rnn_layer)
+            call self % layers(n) % backward( &
+              self % layers(n - 1), &
+              quadratic_derivative(output, this_layer % output) &
+            )
         end select
       else
         ! Hidden layer; take the gradient from the next layer