func.func @main() -> (tensor<32x64xf32> {mhlo.sharding = "\08\01\1A\01\01\22\01\00"}) {
        %cst = "tf.Const"() {value = dense<[524170, 523952]> : tensor<2xi32>} : () -> tensor<2xi32>
        %cst_0 = "tf.Const"() {value = dense<[32, 64]> : tensor<2xi32>} : () -> tensor<2xi32>
        %0 = "tf.StatelessRandomNormal"(%cst_0, %cst) : (tensor<2xi32>, tensor<2xi32>) -> tensor<32x64xf32>
        return %0 : tensor<32x64xf32>
    }

// -----

// CHECK-LABEL: @duplicate_const_for_shape_determining_operand_at_idx_2
// CHECK-SAME: (%[[ARG_0:.*]]: tensor<16x4xf32>, %[[ARG_1:.*]]: tensor<16xi32>)
func.func private @duplicate_const_for_shape_determining_operand_at_idx_2(%arg0: tensor<16x4xf32>, %arg1: tensor<16xi32>) -> tensor<16xf32> {
  %cst = "tf.Const"() {device = "", value = dense<[1]> : tensor<1xi32>} : () -> tensor<1xi32>

  // CHECK-NEXT: %8 = "mhlo.set_dimension_size"(%arg0, %7) <{dimension = 0 : i64}> : (tensor<3x4xi32>, tensor<i32>) -> tensor<3x4xi32>
  // CHECK-NEXT: %9 = mhlo.constant dense<4> : tensor<i32>
  // CHECK-NEXT: %10 = "mhlo.set_dimension_size"(%8, %9) <{dimension = 1 : i64}> : (tensor<3x4xi32>, tensor<i32>) -> tensor<3x4xi32>
  // CHECK-NEXT: %11 = mhlo.constant dense<0> : tensor<i32>

    %5 = "quantfork.qcast"(%4) {volatile} : (tensor<3x4xf32>) -> tensor<3x4x!quant.uniform<i8:f32, 0.13170163023705575:-1>>
    %6 = "quantfork.dcast"(%5) : (tensor<3x4x!quant.uniform<i8:f32, 0.13170163023705575:-1>>) -> tensor<3x4xf32>
    %7 = stablehlo.slice %6 [1:3, 2:4] : (tensor<3x4xf32>) -> tensor<2x2xf32>

//CHECK-NEXT:}

func.func @iota() -> tensor<3x4xf32> {
 %0 = "vhlo.iota_v1" () <{iota_dimension = #vhlo.integer_v1<0 : i64>}> : () -> tensor<3x4xf32>
 return %0 : tensor<3x4xf32>
}

//CHECK:func.func private @iota() -> tensor<3x4xf32> {
//CHECK-NEXT: %0 = "vhlo.iota_v1"() <{iota_dimension = #vhlo.integer_v1<0 : i64>}> : () -> tensor<3x4xf32>
//CHECK-NEXT: return %0 : tensor<3x4xf32>
//CHECK-NEXT:}

  %1 = "tf.ReadVariableOp"(%0) {device = "/CPU:0"} : (tensor<!tf_type.resource<tensor<3x1xf32>>>) -> tensor<3x1xf32>
  %2 = "tf.TPUExecute"(%arg0, %1, %program) {_producer_name = "UNKNOWN", device = "/TPU:0"} : (tensor<1x3xf32>, tensor<3x1xf32>, tensor<3x!tf_type.string>) -> tensor<1x1xf32>
  return %2 : tensor<1x1xf32>
}

}

    %16 = "tf.Mean"(%15, %8) {keep_dims = false} : (tensor<2x112x112x64xf32>, tensor<2xi32>) -> tensor<2x64xf32>
    %17 = "tf.MatMul"(%16, %arg3) {transpose_a = false, transpose_b = false} : (tensor<2x64xf32>, tensor<64x1001xf32>) -> tensor<2x1001xf32>
    %18 = "tf.BiasAdd"(%17, %arg4) {data_format = "NHWC"} : (tensor<2x1001xf32>, tensor<1001xf32>) -> tensor<2x1001xf32>

    %2 = "tf.ConcatV2"(%0, %1, %cst_1) : (tensor<2x2xf32>, tensor<2x2xf32>, tensor<i32>) -> tensor<2x4xf32>
    return %2 : tensor<2x4xf32>
  }
  func.func private @composite_matmul_fn_2(%arg0: tensor<2x2xf32>, %arg1: tensor<2x2xf32>) -> tensor<2x2xf32> attributes {tf_quant.composite_function} {

// Tests ops that variable shapes are correctly evaluated on static types.
// CHECK-LABEL: func @testVariableShape
func.func @testVariableShape(%arg0: tensor<!tf_type.resource<tensor<2x4xf32>>>) -> tensor<2xi32> {
  %0 = "tf.VariableShape"(%arg0) : (tensor<!tf_type.resource<tensor<2x4xf32>>>) -> tensor<2xi32>
  // CHECK:         [[cst:%.*]] = "tf.Const{{.*}} dense<{{\[}}2, 4]> : tensor<2xi32>
  // CHECK-NEXT:    return [[cst]] : tensor<2xi32>

  %value0 = "tfl.pseudo_const"() {value = dense<[[1.0, 2.0], [3.0, 4.0], [5.0, 6.0]]> : tensor<3x2xf32>} : () -> tensor<3x2xf32>
  %value1 = "tfl.pseudo_const"() {value = dense<[[[1.0, 2.0], [3.0, 4.0]], [[5.0, 6.0], [7.0, 8.0]]]> : tensor<2x2x2xf32>} : () -> tensor<2x2x2xf32>
  %lookup0 = "tfl.embedding_lookup"(%index0, %value0) : (tensor<5xi32>, tensor<3x2xf32>) -> tensor<5x2xf32>