)
    rng = np.random.default_rng(seed=14)

    input_tensor_value = rng.random(size=(2, 2))
    sample = {'input_tensor': input_tensor_value}

    feed_dict = repr_dataset.create_feed_dict_from_input_data(
        sample, signature_def
    )

    self.assertLen(feed_dict, 1)
    self.assertIn('input:0', feed_dict)
    self.assertAllEqual(feed_dict['input:0'], input_tensor_value)

      # Handle this as if the size is unknown.
      logging.info('Cannot determine the size of the dataset (%s).', ex)
      return None
  else:
    return None


def create_feed_dict_from_input_data(
    input_data: RepresentativeSample,
    signature_def: meta_graph_pb2.SignatureDef,
) -> Mapping[str, np.ndarray]:
  """Constructs a feed_dict from input data.

  ):
    # Create a mapping from input tensor name to the input tensor value.
    # ex) "Placeholder:0" -> [0, 1, 2]
    feed_dict = rd.create_feed_dict_from_input_data(sample, signature_def)
    sess.run(output_tensor_names, feed_dict=feed_dict)


def _run_graph_for_calibration_graph_mode(
    model_dir: str,
    tags: Collection[str],