func.func @testBroadcastGradientArgs5() -> (tensor<1xi32>, tensor<1xi32>) {
  %s4 = "tf.Const"() {value = dense<[]> : tensor<0xi32>} : () -> tensor<0xi32>
  %s5 = "tf.Const"() {value = dense<[1]> : tensor<1xi32>} : () -> tensor<1xi32>
  %r4, %r5 = "tf.BroadcastGradientArgs"(%s4, %s5) {} : (tensor<0xi32>, tensor<1xi32>) -> (tensor<1xi32>, tensor<1xi32>)
  // CHECK: %[[R0:.*]] = "tf.Const"() <{value = dense<0> : tensor<1xi32>}> : () -> tensor<1xi32>

  %b = arith.constant dense<0.0> : tensor<512xf32>
  %fc = "tfl.fully_connected"(%arg0, %w, %b) {fused_activation_function = "NONE", keep_num_dims = false, weights_format = "DEFAULT"} : (tensor<1x224x224x3xf32>, tensor<512x12xf32>, tensor<512xf32>) -> tensor<1x112x112x512xf32>
  func.return %fc : tensor<1x112x112x512xf32>

// CHECK: %[[b:.*]] = arith.constant dense<0.000000e+00> : tensor<512xf32>

  %dense_default_0 = "tf.Const"() {device = "/device:CPU:0", dtype = f32, value = dense<[]> : tensor<0xf32>} : () -> tensor<0xf32>
  %dense_default_1 = "tf.Const"() {device = "/device:CPU:0", dtype = f32, value = dense<[]> : tensor<0xf32>} : () -> tensor<0xf32>
  %dense_keys = "tf.Const"() {device = "/device:CPU:0", dtype = !tf_type.string, value = dense<""> : tensor<2x!tf_type.string>} : () -> tensor<2x!tf_type.string>

  ^bb0(%arg2: tensor<*xi32>, %arg3: tensor<*xf32>):
    %1:2 = func.call @WhileOp_body(%arg2, %arg3) : (tensor<*xi32>, tensor<*xf32>) -> (tensor<*xi32>, tensor<*xf32>)
    "tfl.yield"(%1#0, %1#1) : (tensor<*xi32>, tensor<*xf32>) -> ()
  }) : (tensor<i32>, tensor<1xf32>) -> (tensor<i32>, tensor<1xf32>) loc("WhileOp")
  func.return %0#1 : tensor<1xf32>

  %1 = "tfl.pseudo_const"() { value = dense<1.0> : tensor<16xf32>} : () -> tensor<16xf32>
  %2 = "tfl.conv_3d"(%arg0, %0, %1) {dilation_h_factor = 1 : i32, dilation_w_factor = 1 : i32, dilation_d_factor = 1 : i32, fused_activation_function = "NONE", padding = "SAME", stride_h = 1 : i32, stride_w = 1 : i32, stride_d = 1 : i32} : (tensor<1x32x32x32x8xf32>, tensor<1x1x1x8x16x!quant.uniform<i8:f32, 1.0>>, tensor<16xf32>) -> tensor<1x32x32x32x16xf32>

          [%u0#0, %u0#1] as %u : tensor<32xf32>)
          {n = 2 : i32, devices = {CORE_0 = ["/CPU:0", "/GPU:1"]}} {
        %read0 = "tf.ReadVariableOp"(%r0) : (tensor<*x!tf_type.resource<tensor<32xf32>>>) -> tensor<32xf32>
        // expected-remark@above {{ID: 1}}
        "tf.AssignVariableOp"(%r1, %u) : (tensor<*x!tf_type.resource<tensor<32xf32>>>, tensor<32xf32>) -> ()
        // expected-remark@above {{ID: 2}}

    return %1 : tensor<1x3xf32>
  }

  func.func private @composite_matmul_with_bias_fn_2(%arg0: tensor<1x4xf32>, %arg1: tensor<4x3xf32>, %arg2: tensor<3xf32>) -> tensor<1x3xf32> attributes {tf_quant.composite_function} {

  // Just to introduce an extra use for %cst.
  %1 = "tf.AddV2"(%cst, %cst) {device = ""} : (tensor<1xi32>, tensor<1xi32>) -> tensor<1xi32>

  return %0 : tensor<16xf32>
}
// Check that the constant is cloned with same value.
// CHECK-DAG: %[[CST_0:.*]] = "tf.Const"()
// CHECK-SAME: value = dense<1> : tensor<1xi32>
// CHECK-DAG: %[[CST_1:.*]] = "tf.Const"()

  %0 = "tfl.unidirectional_sequence_lstm"(%arg0,...

  func.func @f(%arg0: tensor<!tf_type.resource<tensor<?xf32>>> {tf_saved_model.bound_input = @v})
  -> (tensor<!tf_type.resource<tensor<?xf32>>> {tf_saved_model.index_path = []})
  attributes {tf_saved_model.exported_names = ["f"]} {
    // CHECK: return
    return %arg0 : tensor<!tf_type.resource<tensor<?xf32>>>
  }