Hi. I would like to solve an Inverse Problem with NVIDIA PhysicsNeMo Sym. In the documentation, this is done by creating additional neural networks depending on space for the unknown parameters. However my unknown parameters are constant in my inverse problem and using NN depending on space doesn’t work.
In the literature for such an inverse problem, the unknown constant parameters are just added with the weights in the optimization process. I would like to know how to do this with NVIDIA PhysicsNeMo Sym.
Thanks for your help.