@timctho Thanks for sharing your code. I don't see the use of Levenberg Marquardt optimization as indicated by the authors.