Fix deprecations that weren't caught by CI (#6879)

**Context:**

PR #6822 deprecated the arguments `config` and `inner_transform_program`
to `qml.execute`. However, because of some bug in our CI, the
`qml.PennyLaneDeprecationWarning` are not being treated as errors. I
just happened to find this warning in our CI logs.

**Description of the Change:**

Promoted `qml.PennyLaneDeprecationWarning` to a `ValueError` to fish out
any deprecated code. Updated said code to use `qml.workflow.run` instead
of `qml.execute`.

[sc-81529]

doc/releases/changelog-dev.md

@@ -100,6 +100,7 @@
 * The ``inner_transform_program`` and ``config`` keyword arguments in ``qml.execute`` have been deprecated.
   If more detailed control over the execution is required, use ``qml.workflow.run`` with these arguments instead.
   [(#6822)](https://github.com/PennyLaneAI/pennylane/pull/6822)
+  [(#6879)](https://github.com/PennyLaneAI/pennylane/pull/6879)
 
 <h3>Internal changes ⚙️</h3>
 

pennylane/optimize/qnspsa.py

@@ -221,20 +221,7 @@ def _step_core(self, cost, args, kwargs):
             all_grad_dirs.append(grad_dirs)
             all_tensor_dirs.append(tensor_dirs)
 
-        if isinstance(cost.device, qml.devices.Device):
-            program, config = cost.device.preprocess()
-
-            raw_results = qml.execute(
-                all_grad_tapes + all_metric_tapes,
-                cost.device,
-                None,
-                transform_program=program,
-                config=config,
-            )
-        else:
-            raw_results = qml.execute(
-                all_grad_tapes + all_metric_tapes, cost.device, None
-            )  # pragma: no cover
+        raw_results = qml.execute(all_grad_tapes + all_metric_tapes, cost.device)
         grads = [
             self._post_process_grad(raw_results[2 * i : 2 * i + 2], all_grad_dirs[i])
             for i in range(self.resamplings)

pennylane/optimize/riemannian_gradient.py

@@ -392,31 +392,16 @@ def get_omegas(self):
             self.nqubits,
         )
 
-        if isinstance(self.circuit.device, qml.devices.Device):
-            program, config = self.circuit.device.preprocess()
-
-            circuits = qml.execute(
-                circuits,
-                self.circuit.device,
-                transform_program=program,
-                config=config,
-                diff_method=None,
-            )
-        else:
-            circuits = qml.execute(
-                circuits, self.circuit.device, diff_method=None
-            )  # pragma: no cover
-
-        program = self.circuit.device.preprocess_transforms()
+        raw_results = qml.execute(circuits, self.circuit.device)
 
-        circuits_plus = np.array(circuits[: len(circuits) // 2]).reshape(
+        results_plus = np.array(raw_results[: len(raw_results) // 2]).reshape(
             len(self.coeffs), len(self.lie_algebra_basis_names)
         )
-        circuits_min = np.array(circuits[len(circuits) // 2 :]).reshape(
+        results_min = np.array(raw_results[len(raw_results) // 2 :]).reshape(
             len(self.coeffs), len(self.lie_algebra_basis_names)
         )
 
         # For each observable O_i in the Hamiltonian, we have to calculate all Lie coefficients
-        omegas = circuits_plus - circuits_min
+        omegas = results_plus - results_min
 
         return np.dot(self.coeffs, omegas)