%cst_1 = arith.constant dense<2.0> : tensor<3xf32>
      %0 = "tf.XlaCallModule"(%arg0, %cst_0, %cst_1) <{Sout = [#tf_type.shape<1x4x4x3>], module = "", version = 9 : i64}> {_entry_function = @composite_fn_1, _original_entry_function = "composite_fn_1", _tfl_quant_trait = "fully_quantizable"} : (tensor<1x4x4x3xf32>, tensor<3x1x1x3xf32>, tensor<3xf32>) -> tensor<1x4x4x3xf32>

  %biasadd = "tf.BiasAdd"(%conv, %dq_bias) {data_format = "NHWC", device = ""} : (tensor<*xf32>, tensor<2xf32>) -> tensor<*xf32>
  %res = "tf.Relu6"(%biasadd) : (tensor<*xf32>) -> tensor<*xf32>
  %q_res = "quantfork.qcast"(%res) : (tensor<*xf32>) -> tensor<*x!quant.uniform<i8:f32, 0.023529411764705882:-128>>

static constexpr char kMlirModuleStr[] = R"(
  module attributes {tf.versions = {bad_consumers = [], min_consumer = 0 : i32, producer = 268 : i32}} {
  func.func @main(%arg0 : tensor<1xf32>) -> tensor<1xf32> {
    %0 = "tf.Acos"(%arg0) : (tensor<1xf32>) -> tensor<1xf32>
   func.return %0 : tensor<1xf32>
  }
})";

static constexpr char kBadMlirModuleStr[] = R"(
  module attributes {tf.versions = {bad_consumers = [], min_consumer = 0 : i32, producer = 268 : i32}} {

      } : (tensor<1xf32>, tensor<f32>, tensor<i32>) -> tensor<1x!tf_type.qint8>
      %1 = "tf.UniformDequantize"(%0, %scales, %zps) {
        quantization_axis = -1 : i64, quantization_min_val = -128 : i64, quantization_max_val = 127 : i64
      } : (tensor<1x!tf_type.qint8>, tensor<f32>, tensor<i32>) -> tensor<1xf32>
      func.return %1 : tensor<1xf32>
    }
  })mlir";

// CHECK: %[[UDQ:.*]] = stablehlo.uniform_dequantize %[[BITCAST_CONVERT_1]] : (tensor<1x!quant.uniform<i8:f32, 1.000000e+00:3>>) -> tensor<1xf32>
// CHECK: return %[[UDQ]] : tensor<1xf32>
module attributes {tf.versions = {bad_consumers = [], min_consumer = 0 : i32, producer = 268 : i32}} {
    func.func @main(%arg0 : tensor<?xf32>) -> tensor<?xf32> {
      %scales = "tf.Const"() { value = dense<1.0> : tensor<f32> } : () -> tensor<f32>

      %cst = "tf.Const"() {value = dense<3.00000000e-1> : tensor<2x3xf32>} : () -> tensor<2x3xf32>
      %0 = "quantfork.stats"(%arg0) {layerStats = dense<[6.00000000e-6, 9.00000000e-1]> : tensor<2xf32>} : (tensor<1x2xf32>) -> tensor<1x2xf32>

      } : (tensor<?x!tf_type.qint8>, tensor<f32>, tensor<i32>) -> tensor<?xf32>
      func.return %1 : tensor<?xf32>
    }
}

// -----

// expected-error@+1 {{Input shapes provided but no `main` function found.}}
module attributes {tf.versions = {bad_consumers = [], min_consumer = 0 : i32, producer = 268 : i32}} {
    func.func @non_main(%arg0 : tensor<?xf32>) -> tensor<?xf32> {

// CHECK-LABEL: @main
// CHECK: %[[RES0:.*]] = "quantfork.stats"(%arg0) <{layerStats = dense<[0.000000e+00, 2.550000e+02]> : tensor<2xf32>}> : (tensor<1x512x672x8xf32>) -> tensor<1x512x672x8xf32>

  %0 = "quantfork.stats"(%arg0) {layerStats = dense<[0.000000e+00, 2.550000e+02]> : tensor<2xf32>} : (tensor<1x512x672x8xf32>) -> tensor<1x512x672x8xf32>
  func.return %0 : tensor<1x512x672x8xf32>

  %handle = "tf.VarHandleOp"() {device = "/device:CPU:0", container = "", shared_name = "x"} : () -> tensor<!tf_type.resource<tensor<3xf32>>>
  %0 = "tf.ReadVariableOp"(%handle) {_output_shapes = ["tfshape$dim { size: 3 }"], device = "/device:CPU:0", dtype = f32} : (tensor<!tf_type.resource<tensor<3xf32>>>) -> tensor<3xf32>

^bb0(%arg0: tensor<40x37xf32>, %arg1: tensor<40x37xf32>):
  %cst = arith.constant dense<1.0> : tensor<40xf32>
  %0:2 = "tfl.fully_connected"(%arg0, %arg1, %cst) {fused_activation_function = "NONE", keep_num_dims = false, weights_format = "DEFAULT"} : (tensor<40x37xf32>, tensor<40x37xf32>, tensor<40xf32>) -> (tensor<40x40xf32>, tensor<40x40xf32>)
  func.return %0 : tensor<40x40xf32>