%8 = stablehlo.convert %7 : (tensor<1x2xi32>) -> tensor<1x2xf32>
    %9 = stablehlo.convert %2 : (tensor<2x3xi8>) -> tensor<2x3xf32>
    %10 = stablehlo.dot_general %8, %9, contracting_dims = [1] x [0] : (tensor<1x2xf32>, tensor<2x3xf32>) -> tensor<1x3xf32>
    %11 = stablehlo.convert %3 : (tensor<1x3xi32>) -> tensor<1x3xf32>

    return %2 : tensor<1x2xf32>
  }

  func.func private @composite_add_fn(%arg0: tensor<1x2xf32>, %arg1: tensor<1x2xf32>) -> tensor<1x2xf32> attributes {_from_xla_call_module} {
    %0 = stablehlo.add %arg0, %arg1 : tensor<1x2xf32>
    %1 = stablehlo.add %0, %arg1 : tensor<1x2xf32>
    return %1 : tensor<1x2xf32>
  }
}

// -----

    %3 = stablehlo.concatenate %2, %1, dim = 0 : (tensor<1x3xf32>, tensor<1x3xf32>) -> tensor<2x3xf32>
    return %3 : tensor<2x3xf32>
  }
  func.func private @composite_dot_general_fn_1(%arg0: tensor<1x2xf32>, %arg1: tensor<2x3xf32>) -> tensor<1x3xf32> attributes {_from_xla_call_module, tf_quant.composite_function} {

// CHECK: return

  func.func private @composite_dot_general_fn(%arg0: tensor<1x4xf32>, %arg1: tensor<4x3xf32>) -> tensor<1x3xf32> attributes {_from_xla_call_module} {
      %0 = stablehlo.dot_general %arg0, %arg1, contracting_dims = [1] x [0] : (tensor<1x4xf32>, tensor<4x3xf32>) -> tensor<1x3xf32>
      return %0 : tensor<1x3xf32>
  }
}

// -----

}

func.func @mean(%arg0: tensor<2x2xf32>, %arg1: tensor<1xi32>) -> tensor<1x2xf32> {
  %0 = "tf.Mean"(%arg0, %arg1) : (tensor<2x2xf32>, tensor<1xi32>) -> tensor<1x2xf32>
  func.return %0 : tensor<1x2xf32>

// CHECK-LABEL: mean
// CHECK:  "tfl.mean"(%arg0, %arg1) <{keep_dims = false}> : (tensor<2x2xf32>, tensor<1xi32>) -> tensor<1x2xf32>
}

func.func @mean_true(%arg0: tensor<2x2xf32>, %arg1: tensor<1xi32>) -> tensor<1x2xf32> {

  %4 = "tfl.minimum"(%0, %cst) : (tensor<32xf32>, tensor<32xf32>) -> tensor<32xf32>
  %5 = "tfl.minimum"(%1, %cst) : (tensor<32xf32>, tensor<32xf32>) -> tensor<32xf32>
  %6 = "tfl.minimum"(%2, %cst) : (tensor<32xf32>, tensor<32xf32>) -> tensor<32xf32>
  %7 = "tfl.minimum"(%3, %cst) : (tensor<32xf32>, tensor<32xf32>) -> tensor<32xf32>

func.func @lift_float_gather(%arg0: tensor<6xi64>) -> (tensor<6x32xf32>) {
  %cst = "tf.Const"() {device = "", value = dense<0> : tensor<i32>} : () -> tensor<i32>
  %cst_0 = "tf.Const"() {device = "", value = dense<1.0> : tensor<128x32xf32>} : () -> tensor<128x32xf32>
  %0 = "tf.GatherV2"(%cst_0, %arg0, %cst) {batch_dims = 0 : i64, device = ""} : (tensor<128x32xf32>, tensor<6xi64>, tensor<i32>) -> tensor<6x32xf32>
  return %0 : tensor<6x32xf32>

  %1 = "tf.ReadVariableOp"(%0) {device = "/CPU:0"} : (tensor<!tf_type.resource<tensor<1x1xf32>>>) -> tensor<1x1xf32>
  %2:2 = "tf.Split"(%cst, %arg0) {device = "/CPU:0"} : (tensor<i32>,  tensor<1x4xf32>) -> (tensor<1x2xf32>, tensor<1x2xf32>)
  %3:2 = "tf.Split"(%cst, %2#0) {device = "/CPU:0"} : (tensor<i32>,  tensor<1x2xf32>) -> (tensor<1x1xf32>, tensor<1x1xf32>)

  func.return %0 : tensor<1x40xf32>
}

func.func @testAdd(%arg0: tensor<10x10x10xf32>, %arg1: tensor<10x10x10xf32>) -> tensor<10x10x10xf32> {
  // CHECK: _arithmetic_count = 1000 : i64
  %0 = "tfl.add"(%arg0, %arg1) {fused_activation_function = "NONE"} : (tensor<10x10x10xf32>, tensor<10x10x10xf32>) -> tensor<10x10x10xf32>

// CHECK:           }) : (tensor<1x?xf32>, tensor<1x0xf32>, tensor<1x0xf32>, tensor<1x0xf32>, tensor<1x0xf32>, tensor<1x3xf32>, tensor<1x3xf32>, tensor<1x3xf32>, tensor<1x3xf32>, none, none, none, tensor<1xf32>, tensor<1xf32>, tensor<1xf32>, tensor<1xf32>, tensor<3x1xf32>, tensor<3xf32>, tensor<1x3xf32>, tensor<1x1xf32>, none, none, none, none) -> tensor<1x3xf32>
// CHECK:           [[VAL_50:%.*]] = tensor.cast [[VAL_51:%.*]] : tensor<1x3xf32> to tensor<1x?xf32>