auto op = cast<ResizeNearestNeighborOp>(src_op);
    auto loc = op.getLoc();
    auto result_ty = mlir::cast<ShapedType>(op.getType());

    auto input = op.getImages();
    auto input_ty = mlir::cast<ShapedType>(input.getType());
    auto input_element_ty = input_ty.getElementType();
    auto out_size = op.getSize();
    auto out_size_ty = mlir::cast<ShapedType>(out_size.getType());

#include "tensorflow/core/common_runtime/eager/tensor_handle.h"
#include "tensorflow/core/distributed_runtime/rpc/grpc_server_lib.h"
#include "tensorflow/core/framework/function.pb.h"
#include "tensorflow/core/platform/casts.h"
#include "tensorflow/core/platform/env.h"
#include "tensorflow/core/platform/logging.h"
#include "tensorflow/core/platform/macros.h"
#include "tensorflow/core/platform/protobuf.h"

  "$0.isa<DenseElementsAttr>() && "
  "$0.cast<DenseElementsAttr>().getNumElements() == 1 && "
  "(($0.cast<DenseElementsAttr>().getValues<APFloat>()[0].isLargest() && "
  "$0.cast<DenseElementsAttr>().getValues<APFloat>()[0].isNegative()) || "
  "$0.cast<DenseElementsAttr>().getValues<APFloat>()[0].isNegInfinity())">>;

def L2NormValidReduceIndex : Constraint<CPred<
  "L2NormalizeReduceAxis($0, $1.cast<DenseElementsAttr>())">>;

// CHECK-DAG:       %[[CST1:.*]] = "tf.Const"() <{value = dense<0> : tensor<1x24xi32>}> : () -> tensor<1x24xi32>
// CHECK:           %[[CAST0:.*]] = "tf.Cast"(%arg0) <{Truncate = false}> : (tensor<1x24xi1>) -> tensor<1x24xi32>
// CHECK:           %[[CAST1:.*]] = "tf.Cast"(%arg2) <{Truncate = false}> : (tensor<1x2xi1>) -> tensor<1x2xi32>

      filter_i8_value_attr = mlir::cast<DenseIntElementsAttr>(
          cast<stablehlo::ConstantOp>(filter_op->getOperand(0).getDefiningOp())
              .getValue());
    }

    // Create Uniform Quantized constant for the filter.
    auto subtract_op =
        cast<stablehlo::SubtractOp>(*op.getResult().user_begin());
    auto other_conv_op = cast<stablehlo::ConvolutionOp>(

								cases = append(cases,
									testCase{
										path:  fmt.Sprintf("/policy-%s-all", ns.Prefix()),
										allow: ns.Name() == to.Config().Namespace.Name(),
									})
							}

							// Make sure the workload-specific paths succeeds.
							cases = append(cases,
								testCase{

      // to
      //     Fill(Concat(num_element, element_shape), 0)
      // because element_shape may contain -1 to represent unknown dimension.
      //
      // In real world use cases (e.g. Keras RNN), `element_shape` is usually
      // a constant, and the first dimension of `element_shape` is usually
      // batch dimension. Currently TFLiteConverter always rewrite unknown

    return errors::InvalidArgument("Cast requires 2 arguments");
  }
  if (grad_inputs.size() != 1) {
    return errors::InvalidArgument("Cast grad requires 1 grad input");
  }

  auto src_type = op.input_type(0);
  auto dst_type = grad_inputs[0].type();
  if (IsFloatingPointDtype(src_type) && IsFloatingPointDtype(dst_type)) {
    grad_outputs->push_back(Cast(scope, grad_inputs[0], src_type));
  } else {

}

func TestCostEstimation(t *testing.T) {
	cases := []struct {
		name            string
		schemaGenerator func(maxLength *int64) *schema.Structural
		setMaxElements  int64

		// calc costs expectations are checked against the generated schema without any max element limits set
		expectedCalcCost           uint64
		expectCalcCostExceedsLimit uint64

					"Diff: %s", cmp.Diff(testCase.expectedAnnotations, existingNode.Annotations))
			}
		})
	}
}

// We can't test failure or autodetection cases here because the relevant code isn't mockable
func TestNodeAddress_NoCloudProvider(t *testing.T) {
	cases := []struct {
		name              string
		nodeIPs           []net.IP
		expectedAddresses []v1.NodeAddress
		shouldError       bool
	}{
		{