I\'m trying to extract the jacobian matrix from an autoencoder using pytorch automatic differentiation package (autograd.functional.jacobian).
My code is as follow:
