CARDIAX-NNFE is a framework that implements the Neural Network Finite Element method. It learns a family of solutions for parameterized PDEs. It depends on CARDIAX for finite element machinery, Equinox for network creation, and Optax for optimization.

Learning pressure-volume loops for a prolate spheroidal geometry over a range of pressures and active stresses. The wireframe mesh is the FE solution and red points are NNFE.