failed(AddFloatAttr(func, attrs, "nms_iou_threshold", &fbb)) ||
      failed(AddFloatAttr(func, attrs, "y_scale", &fbb)) ||
      failed(AddFloatAttr(func, attrs, "x_scale", &fbb)) ||
      failed(AddFloatAttr(func, attrs, "h_scale", &fbb)) ||
      failed(AddFloatAttr(func, attrs, "w_scale", &fbb)))
    return failure();
  auto use_regular_nms =
      mlir::dyn_cast_or_null<BoolAttr>(attrs.get("use_regular_nms"));

    "'h_scale' type: 'float'} attr : { name: 'max_classes_per_detection' "
    "type: 'int'} attr : { name: 'max_detections' type: 'int'} attr : { "
    "name: 'nms_iou_threshold' type: 'float'} attr : { name: "
    "'nms_score_threshold' type: 'float'} attr : { name: 'num_classes' type: "
    "'int'} attr : { name: 'w_scale' type: 'float'} attr : { name: 'x_scale' "

  %0 = "tf.Const"() {value = dense<0.0> : tensor<f32>} : () -> tensor<f32>

  // CHECK: %[[X_SCALED:.+]] = mhlo.multiply %[[X_CENTER]], %[[SCALE_BCAST]] : tensor<4x256xf32>
  // CHECK: %[[X_NORMED:.+]] = mhlo.divide %[[X_SCALED]], %[[STDDEV_BCAST]] : tensor<4x256xf32>
  // CHECK: %[[RESULT:.+]] = mhlo.add %[[X_NORMED]], %[[OFFSET_BCAST]] : tensor<4x256xf32>
  %0 = "mhlo.batch_norm_inference"(%x, %scale, %offset, %mean, %variance)

apiVersion: apps/v1beta2
kind: Scale
metadata:
  annotations:
    annotationsKey: annotationsValue
  creationTimestamp: "2008-01-01T01:01:01Z"
  deletionGracePeriodSeconds: 10
  deletionTimestamp: "2009-01-01T01:01:01Z"
  finalizers:
  - finalizersValue
  generateName: generateNameValue
  generation: 7
  labels:
    labelsKey: labelsValue
  managedFields:
  - apiVersion: apiVersionValue
    fieldsType: fieldsTypeValue

apiVersion: extensions/v1beta1
kind: Scale
metadata:
  annotations:
    annotationsKey: annotationsValue
  creationTimestamp: "2008-01-01T01:01:01Z"
  deletionGracePeriodSeconds: 10
  deletionTimestamp: "2009-01-01T01:01:01Z"
  finalizers:
  - finalizersValue
  generateName: generateNameValue
  generation: 7
  labels:
    labelsKey: labelsValue
  managedFields:
  - apiVersion: apiVersionValue
    fieldsType: fieldsTypeValue

f32], _support_output_type_float_in_quantized_op = true, detections_per_class = 100 : i64, device = "", h_scale = 5.000000e+00 : f32, max_classes_per_detection = 1 : i64, max_detections = 20 : i64, nms_iou_threshold = 6.000000e-01 : f32, nms_score_threshold = 3.000000e-01 : f32, num_classes = 90 : i64, use_regular_nms = false, w_scale = 5.000000e+00 : f32, x_scale = 1.000000e+01 : f32, y_scale = 1.000000e+01 : f32} : (tensor<1x64x64x32xf32>, tensor<1x64x64x32xf32>, tensor<1x64x64x32xf32>) -> (te...

    for (RequantizeState& rescale : rescales) {
      if (rescale.params == quantized_type) {
        rescale.users.emplace_back(op, operand_index);
        return true;
      }
    }
    RequantizeState& rescale = rescales.emplace_back();
    rescale.pos = RequantizeState::ON_OUTPUT;
    rescale.params = quantized_type;
    rescale.users.emplace_back(op, operand_index);
    return true;
  }

	}
	return q.i.Sign()
}

// AsScale returns the current value, rounded up to the provided scale, and returns
// false if the scale resulted in a loss of precision.
func (q *Quantity) AsScale(scale Scale) (CanonicalValue, bool) {
	if q.d.Dec != nil {
		return q.d.AsScale(scale)
	}
	return q.i.AsScale(scale)
}

// RoundUp updates the quantity to the provided scale, ensuring that the value is at

  auto &state = GetResultQuantState(op, res_index);
  if (state.params == params) {
    return false;
  }
  if (!state.IsEmpty()) {
    auto &rescale = GetResultRequantizeState(op, res_index);
    rescale.params = params;
    rescale.pos = RequantizeState::ON_INPUT;
    return false;
  }
  state.params = params;
  return true;
}