Neural network based solvers for partial differential equations and inverse problems :milky_way:. Implementation of physics-informed neural networks in pytorch.
At the moment the framework is designed such that the model has to be moved by the user. But you are right. maybe its a good idea to move to model by the framework.
https://github.com/ComputationalRadiationPhysics/NeuralSolvers/blob/bf9af0a3d23bc156a05a339473a75d0b65d25d5f/PINNFramework/PINN.py#L97
I feel like there should be a line sending the model to cuda if needed.