"""Gets the number of XlaCallModule ops in the output saved model."""
    root = load.load(output_saved_model_path)
    tf_graph_def = root.signatures['serving_default'].graph.as_graph_def()
    count = 0
    for node_def in tf_graph_def.node:
      if node_def.op == 'XlaCallModule':
        count += 1
    for function in tf_graph_def.library.function:
      for node_def in function.node_def:
        if node_def.op == 'XlaCallModule':

    build_graph(out_dir)
    filename = os.path.join(out_dir, 'test_graph_%s.pb' % build_graph.__name__)
    with open(filename, 'wb') as f:
      f.write(
          six.ensure_binary(
              g.as_graph_def().SerializeToString(deterministic=True)
          )
      )


def main(_):
  control_flow_util.enable_control_flow_v2()
  write_graph(tfadd, FLAGS.out_dir)
  write_graph(tfadd_with_ckpt, FLAGS.out_dir)

                          *module_ref, context.get(), quantization_options,
                          *function_aliases, calibration_data_dir));

  // Save and run the calibration model.
  if (calibration_exported_model.has_graph_def()) {
    TF_ASSIGN_OR_RETURN(std::string calibration_saved_model_dir,
                        CreateTmpDir());
    py_function_library.SaveExportedModel(
        calibration_saved_model_dir, calibration_exported_model,

@tf.function
def foo(x, y):
  return mlir_passthrough_op([x, y], mlir_module, Toutputs=[tf.float32])

graph_def = foo.get_concrete_function(tf.TensorSpec([10], tf.float32), tf.TensorSpec([10], tf.float32)).graph.as_graph_def()
```
  }];

  let arguments = (ins
    Variadic<TF_Tensor>:$inputs,

    StrAttr:$mlir_module
  );

  let results = (outs
    Variadic<TF_Tensor>:$outputs