Construction of loss

[Fankunkun123]

Hello, Dr. Lu! Your deepXDE library is really powerful.
Now that I read your article on Gradient Boosting (gPINN), there are the following questions in the code reproduction: How does the loss of the gradient boosting part come together with all the losses after the gpde definition? 2. Is there a specific mathematical expression for loss combination? If there is a regularization term in the loss, how should the regularization term be added to the total loss?
I hope you can answer my question, thank you very much!