OP_REQUIRES_OK(context, context->GetAttr("ids", &ids_));
    OP_REQUIRES_OK(context, context->GetAttr("calibration_methods",
                                             &calibration_methods_));
    OP_REQUIRES(
        context, ids_.size() == calibration_methods_.size(),
        absl::AbortedError(
            "The `ids` and `calibration_methods` must have the same size."));

    // Check the number and type of inputs.

                  .Input(inputs)
                  .Attr("ids", ids)
                  .Attr("calibration_methods", calibration_methods)
                  .Finalize(node_def()));
  ASSERT_THAT(InitOp(),
              StatusIs(tsl::error::INVALID_ARGUMENT,
                       HasSubstr("NodeDef missing attr 'output_file_path'")));
}

TEST_F(CalibrationStatisticsSaverTest, WrongNumInputs) {

    SmallVector<Value>& statistics_outputs, SmallVector<StringRef>& ids,
    SmallVector<int32_t>& calibration_methods) {
  for (auto op : region.getOps<TF::CustomAggregatorOp>()) {
    if (aggregator_ops_to_ignore.count(op.getId().str())) continue;

    ids.push_back(op.getId());
    calibration_methods.push_back(op.getCalibrationMethod());
    statistics_outputs.push_back(op.getMin());

// CKECK-SAME: <{calibration_method = 5 : i32, id = "skipping_id", num_bins = 32 : i32, max_percentile = 0.000000e+00 : f32, min_percentile = 0.000000e+00 : f32}> : (tensor<1x3x4x3xf32>) -> (tensor<1x3x4x3xf32>, tensor<f32>, tensor<f32>, tensor<512xi64>)
// CHECK: %[[CUSTOM_AGGREGATOR_1:.*]], %[[MIN_1:.*]], %[[MAX_1:.*]], %[[HISTOGRAM_1:.*]] = "tf.CustomAggregator"

  let arguments = (ins
    TensorOf<[TF_Float32]>:$input,

    // The unique id of this `CustomAggregator` op.
    StrAttr:$id,
    // The integer value of the enforcing `CalibrationMethod`.
    I32Attr:$calibration_method,
    // The number of histogram bins.
    I32Attr:$num_bins,
    // Min percentile to be included in the selected range, only used in the
    // `HISTOGRAM_PERCENTILE` method.

// CKECK-SAME: <{calibration_method = 5 : i32, id = "0", num_bins = 32 : i32, max_percentile = 0.000000e+00 : f32, min_percentile = 0.000000e+00 : f32}> : (tensor<1x3x4x3xf32>) -> (tensor<1x3x4x3xf32>, tensor<f32>, tensor<f32>, tensor<512xi64>)
// CHECK: %[[CUSTOM_AGGREGATOR_1:.*]], %[[MIN_1:.*]], %[[MAX_1:.*]], %[[HISTOGRAM_1:.*]] = "tf.CustomAggregator"

from tensorflow.python.platform import test

_CalibrationMethod = (
    stablehlo_quant_config_pb2.CalibrationOptions.CalibrationMethod
)


class CalibrationAlgorithmTest(test.TestCase, parameterized.TestCase):

  def test_min_max_max(self):
    calib_opts = stablehlo_quant_config_pb2.CalibrationOptions(
        calibration_method=_CalibrationMethod.CALIBRATION_METHOD_MIN_MAX
    )
    statistics = calib_stats_pb2.CalibrationStatistics()

    auto calibration_method =
        static_cast<CalibrationMethod>(calibration_method_value);
    OP_REQUIRES(
        context,
        calibration_method !=
            CalibrationOptions::CALIBRATION_METHOD_UNSPECIFIED,
        absl::AbortedError("The calibration method must be specified."));

    calib_opts_.set_calibration_method(calibration_method);

          [1.0, 2.0, 3.0, 4.0, 5.0], dtypes.float32
      )

      aggregator = custom_aggregator_op_wrapper.custom_aggregator(
          input_tensor,
          id='1',
          calibration_method=_CalibrationMethod.CALIBRATION_METHOD_MIN_MAX,
      )
      aggregator_output = self.evaluate(aggregator)
      self.assertAllEqual(aggregator_output.output, [1.0, 2.0, 3.0, 4.0, 5.0])

      SmallVector<NamedAttribute, 5> attributes{
          rewriter.getNamedAttr("id", rewriter.getStringAttr(aggregator_id)),
          rewriter.getNamedAttr(
              "calibration_method",
              rewriter.getI32IntegerAttr(calib_opts_.calibration_method())),
          rewriter.getNamedAttr("num_bins",
                                rewriter.getI32IntegerAttr(effective_num_bins)),
          rewriter.getNamedAttr(