How to get the Jacobian and Hessian value after training the neural network?

[DOUBLEYANG-Github]

I know that there is a method in deepXDE: https://deepxde.readthedocs.io/en/latest/modules/deepxde.html#module-deepxde.gradients that calculates the Jacobian matrix when training the neural network. Suppose I have the trained network, how do I get the corresponding Jacobian given an input value?

[Dkupetsk]

I have the same question

[lululxvi]

See https://deepxde.readthedocs.io/en/latest/demos/pinn_forward/poisson.1d.dirichlet.html

model.predict(x, operator=pde)