Fix np.pad issue in numpy 1.16 (#3377)

Cirq minimum requirement for numpy is 1.16.

In 1.16, the mode parameter for np.pad is required.
(It was changed to optional in numpy 1.17).

Add this parameter in with the default value to support numpy 1.16.

Author: dstrain115

cirq/experiments/grid_parallel_two_qubit_xeb.py

@@ -492,7 +492,8 @@ def _get_xeb_result(qubit_pair: GridQubitPair, circuits: List['cirq.Circuit'],
             _, counts = np.unique(measurements, return_counts=True)
             empirical_probs = counts / len(measurements)
             empirical_probs = np.pad(empirical_probs,
-                                     (0, 4 - len(empirical_probs)))
+                                     (0, 4 - len(empirical_probs)),
+                                     mode='constant')
             all_and_observed_probabilities[depth].append(
                 (probabilities, probabilities[measurements]))
             empirical_probabilities[depth].append(empirical_probs)