Models

models.MLP_models.MLP_surrogate([channels, ...])
models.Spline_models.CubicSpline([x, y, n_knot])
models.Spline_models.MultiDim_CubicSpline(y)
models._pde_informed_params.pde_params(channels)	Default model : PINN prediction + NeuralODE simulation to estimate parameters
models._pde_informed_params.log_pde_params(...)
models._density_transport.Density_Transfer([...])	Wrap up the density transfer function into a nn Module, for calling odeint_adjoint