How can I output a gradient in vector format in Op.grad instance?

Questions v5
3

Yes! I started from that page. But, the difference is that the example on the page uses a scalar output as each element in return of grad as

    def grad(self, inputs, output_gradients):
        (
            grad_wrt_emission_obsered,
            grad_wrt_emission_signal,
            grad_wrt_emission_noise,
            grad_wrt_logp_initial_state,
            grad_wrt_logp_transition,
        ) = hmm_logp_grad_op(*inputs)
        # If there are inputs for which the gradients will never be needed or cannot
        # be computed, `aesara.gradient.grad_not_implemented` should  be used as the
        # output gradient for that input.
        output_gradient = output_gradients[0]
        return [
            output_gradient * grad_wrt_emission_obsered,
            output_gradient * grad_wrt_emission_signal,
            output_gradient * grad_wrt_emission_noise,
            output_gradient * grad_wrt_logp_initial_state,
            output_gradient * grad_wrt_logp_transition,
        ]

In my case, I’d like to return [scalar, vector] instead.