func.func @QuantizeUnidirectionalLstm(%arg0: tensor<1x2x3xf32>) -> (tensor<1x2x3xf32>) {
  %1 = "tfl.pseudo_const"() {value = dense<[[0.1]]> : tensor<1x1xf32>} : () -> tensor<1x1xf32>
  %2 = "tfl.pseudo_const"() {value = dense<[[0.2]]> : tensor<1x1xf32>} : () -> tensor<1x1xf32>
  %3 = "tfl.pseudo_const"() {value = dense<[[0.3]]> : tensor<1x1xf32>} : () -> tensor<1x1xf32>
  %4 = "tfl.pseudo_const"() {value = dense<[[0.4]]> : tensor<1x1xf32>} : () -> tensor<1x1xf32>

module {

func.func @main(%arg0: tensor<4x384x32xf32>) -> tensor<1x384x32xf32> {
  %0 = "tfl.pseudo_const"() {value = dense<0> : tensor<3xi32>} : () -> tensor<3xi32>
  %1 = "tfl.pseudo_const"() {value = dense<[1, 384, 32]> : tensor<3xi32>} : () -> tensor<3xi32>

// CHECK-DAG:       [[VAL_2:%.*]] = "tfl.pseudo_const"{{.*}}dense<0> : tensor<4xi32>
// CHECK-DAG:       [[VAL_3:%.*]] = "tfl.pseudo_const"{{.*}}dense<[0, 0, 0, 256]> : tensor<4xi32>
// CHECK-DAG:       [[VAL_4:%.*]] = "tfl.pseudo_const"{{.*}}dense<[0, 0, 0, 768]> : tensor<4xi32>
// CHECK-DAG:       [[VAL_5:%.*]] = "tfl.pseudo_const"{{.*}}dense<[-1, -1, -1, 256]> : tensor<4xi32>

  %0 = "tfl.pseudo_const"() { value = dense<[1.0, 2.0, 3.0, 4.0]> : tensor<4xf16> } : () -> tensor<4xf16>
  func.return %0 : tensor<4xf16>
}

func.func @f32() -> tensor<4xf32> {
  // CHECK-LABEL: @f32
  // CHECK: value = dense<[1.000000e+00, 2.000000e+00, 3.000000e+00, 4.000000e+00]> : tensor<4xf32>

  %0 = "tfl.pseudo_const"() { value = dense<[1.0, 2.0, 3.0, 4.0]> : tensor<4xf16> } : () -> tensor<4xf16>
  func.return %0 : tensor<4xf16>
}

func.func @f32() -> tensor<4xf32> {
  // CHECK-LABEL: @f32
  // CHECK: value = dense<[1.000000e+00, 2.000000e+00, 3.000000e+00, 4.000000e+00]> : tensor<4xf32>

  %10 = "tfl.pseudo_const"() {value = dense<[-1.59288621, 0.904723584]> : tensor<2xf32>} : () -> tensor<2xf32>
  %11 = "tfl.pseudo_const"() {value = dense<[-0.323118627, 1.77580559]> : tensor<2xf32>} : () -> tensor<2xf32>
  %12 = "tfl.pseudo_const"() {value = dense<[-1.0347594, -1.09994471]> : tensor<2xf32>} : () -> tensor<2xf32>
  %13 = "tfl.pseudo_const"() {value = dense<[-2.03072214, -1.63648951]> : tensor<2xf32>} : () -> tensor<2xf32>

  %1 = "tfl.pseudo_const"() {value = dense<[[0.1]]> : tensor<1x1xf32>} : () -> tensor<1x1xf32>
  %2 = "tfl.pseudo_const"() {value = dense<[[0.2]]> : tensor<1x1xf32>} : () -> tensor<1x1xf32>
  %3 = "tfl.pseudo_const"() {value = dense<[[0.3]]> : tensor<1x1xf32>} : () -> tensor<1x1xf32>
  %4 = "tfl.pseudo_const"() {value = dense<[[0.4]]> : tensor<1x1xf32>} : () -> tensor<1x1xf32>

// RUN: tf-opt %s -tfl-partitioned-topological-sort | FileCheck %s

// CHECK-LABEL: @tf_ops_will_be_partitioned
func.func @tf_ops_will_be_partitioned() -> tensor<1xf32> {
  %const = "tfl.pseudo_const"() {value = dense<[1.0]> : tensor<1xf32>} : () -> tensor<1xf32>
  %tmp1 = "tfl.add"(%const, %const) { fused_activation_function = "NONE" } : (tensor<1xf32>,tensor<1xf32>) -> (tensor<1xf32>)

func.func @test_transpose_conv2d(%arg0: tensor<1x32x32x8xf32>) -> tensor<1x32x32x16xf32> {
  // CHECK-DAG: %[[SHAPE:.+]] = "tfl.pseudo_const"() <{value = dense<[1, 32, 32, 16]>
  // CHECK-DAG: %[[VAL0:.+]] = "tfl.pseudo_qconst"() <{qtype = tensor<16x{{.+}}>, value = dense<1> : tensor<16xi32>}>
  // CHECK-DAG: %[[VAL1:.+]] = "tfl.pseudo_qconst"() <{qtype = tensor<16x{{.+}}>, value = dense<2> : tensor<16xi32>}>
  // CHECK-DAG: %[[VAL2:.+]] = "tfl.dequantize"(%[[VAL0]])

// CHECK-NEXT:      %4 = "tfl.shape"(%arg1) : (tensor<4x?x256xf32>) -> tensor<3xi32>
// CHECK-NEXT:      %5 = "tfl.pseudo_const"() <{value = dense<[-1, 0, -1]> : tensor<3xi32>}> : () -> tensor<3xi32>
// CHECK-NEXT:      %6 = "tfl.pseudo_const"() <{value = dense<[-1, -1, 0]> : tensor<3xi32>}> : () -> tensor<3xi32>
// CHECK-NEXT:      %7 = "tfl.pseudo_const"() <{value = dense<1> : tensor<i32>}> : () -> tensor<i32>