If I have two different neural networks (parametrized by model1 and model2) and corresponding two optimizers, would the below operation using model2.parameters without detach()
