auto device_attr = (*this)->getAttrOfType<StringAttr>("device");
  // Treat missing device attribute like unspecified (= empty string) attribute.
  // Note that different op instances with the same string (including empty
  // string) are seen as dependent (same resource instance).
  if (!device_attr) return "";
  return device_attr.str();
}

//===----------------------------------------------------------------------===//
// NotEqualOp

  %0 = "tf.Conv2D"(%arg0, %arg1) {padding = "SAME", strides = [1, 1, 1, 1]} : (tensor<*xf32>, tensor<*xf32>) -> tensor<?x?x?x?xf32>
  func.return %0 : tensor<?x?x?x?xf32>
}

// -----

// CHECK-LABEL: func @testValidConv3D
func.func @testValidConv3D(%arg0: tensor<256x32x32x32x3xf32>, %arg1: tensor<3x3x3x3x16xf32>) -> tensor<256x32x32x32x16xf32> {

      return op.emitOpError("'input' and 'output' should have the same rank.");
    }

    // Check if input and output shapes are same
    for (int i = 0; i < input_type.getRank(); i++) {
      if (input_type.getDimSize(i) != output_type.getDimSize(i)) {
        return op.emitOpError(
            "'input' and 'output' should have the same shape.");
      }
    }
  }
  return success();
}

                should have all the fields, including the ones that have the same value
                as the default. This is different from `response_model_exclude_unset`
                in that if the fields are set but contain the same default values,
                they will be excluded from the response.

                should have all the fields, including the ones that have the same value
                as the default. This is different from `response_model_exclude_unset`
                in that if the fields are set but contain the same default values,
                they will be excluded from the response.

				base.Errorf("%v", p.Error)
			}
		}
	}
	base.ExitIfErrors()

	// Check for duplicate loads of the same package.
	// That should be impossible, but if it does happen then
	// we end up trying to build the same package twice,
	// usually in parallel overwriting the same files,
	// which doesn't work very well.
	seen := map[string]bool{}
	reported := map[string]bool{}

  func.return %0, %1, %2, %3 : tensor<f32>, tensor<f32>, tensor<f32>, tensor<f32>
}

// CHECK-LABEL: func @tflite_custom_nms(
// CHECK-SAME:                          %[[VAL_0:.*]]: tensor<1x100x4xf32>,
// CHECK-SAME:                          %[[VAL_1:.*]]: tensor<1x100x91xf32>,

  let summary = "Form clusters from instructions assigned to same device";
  let constructor = "TFDevice::CreateClusterFormationPass()";
  let dependentDialects = ["tf_device::TensorFlowDeviceDialect"];
  let description = [{
    Clusters operations with the same device assignment id. For each
    cluster, creates a "tf_device.device_launch" op with a Region containing the

 * server address, the canonical representation of the URL, and does not
 * compare authentication information. In essance, two
 * <code>SmbFile</code> objects that refer to
 * the same file should generate the same hashcode provided it is possible
 * to make such a determination.
 *
 * @return  A hashcode for this abstract file
 * @throws SmbException
 */

    public int hashCode() {

			masq:     false,
		},
		{
			name:     "clusterIP with TCP, UDP, and SCTP on same port (TCP)",
			sourceIP: "10.180.0.2",
			protocol: v1.ProtocolTCP,
			destIP:   "172.30.0.42",
			destPort: 53,
			output:   "10.180.0.1:5353, 10.180.2.1:5353",
			masq:     false,
		},
		{
			name:     "clusterIP with TCP, UDP, and SCTP on same port (TCP)",
			sourceIP: "10.180.0.2",
			protocol: v1.ProtocolUDP,