beberapa kerugian pytorch

b = nn.MSELoss()
a = nn.CrossEntropyLoss()

loss_a = a(output_x, x_labels)
loss_b = b(output_y, y_labels)

loss = loss_a + loss_b

loss.backward()
coder