%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>

// CHECK-SAME:                        %[[VAL_1:.*]]: tensor<1x2xi32>) -> tensor<1x2xi32> {
// CHECK:           %[[VAL_2:.*]] = "tf.AddV2"(%[[VAL_0]], %[[VAL_1]]) : (tensor<1xi32>, tensor<1x2xi32>) -> tensor<1x2xi32>
// CHECK:           return %[[VAL_2]] : tensor<1x2xi32>
// CHECK:         }
func.func @broadcast_add_chlo(%arg0: tensor<1xi32>, %arg1: tensor<1x2xi32>) -> tensor<1x2xi32> {

func.func @packNegInputAxis2(%arg0: tensor<1x4xi32>, %arg1: tensor<1x4xi32>) -> tensor<1x2x4xi32> {
  // CHECK: "tfl.pack"(%arg0, %arg1) <{axis = -2 : i32, values_count = 2 : i32}>
  %0 = "tfl.pack"(%arg0, %arg1) {axis = -2 : i32, values_count = 2 : i32} : (tensor<1x4xi32>, tensor<1x4xi32>) -> tensor<1x2x4xi32>
  func.return %0 : tensor<1x2x4xi32>
}

func.func @packNegInputAxis3(%arg0: tensor<1x4xi32>, %arg1: tensor<1x4xi32>) -> tensor<2x1x4xi32> {

tensor<1024xf32>, tensor<32x1024xf32>, tensor<1024x1xf32>, tensor<1024xf32>, tensor<1024x1xf32>, tensor<i64>, tensor<i32>, tensor<32x1024xf32>, tensor<i64>, tensor<i64>, tensor<1024xf32>, tensor<32x1024xf32>, tensor<1024x1xf32>, tensor<1024xf32>, tensor<1024x1xf32>, tensor<i64>, tensor<i32>, tensor<32x1024xf32>, tensor<i64>, tensor<i64>, tensor<1024xf32>, tensor<32x1024xf32>, tensor<1024x1xf32>, tensor<1024xf32>, tensor<1024x1xf32>, tensor<i64>, tensor<i32>, tensor<32x1024xf32>, tensor<i64>, tensor<i64>,...

  // CHECK-NEXT:     %[[GROUP:.*]] = "tf.Const"()
  // CHECK-SAME:       [0, 1, 2, 3]
  // CHECK-NEXT:     %[[REDUCED:.*]] = "tf.XlaAllReduce"(%[[ID]], %[[GROUP]]) <{mode = "CrossReplica", reduce_op = "Add"}> : (tensor<f32>, tensor<1x4xi32>) -> tensor<f32>
  // CHECK-NEXT:     "tf.OpA"(%[[REDUCED]]) : (tensor<f32>) -> ()
  tf_device.replicate {n = 4 : i32} {
    "tf_device.cluster"() ({
      "tf.OpA"(%arg0) : (tensor<f32>) -> ()
      tf_device.return

  %values, %indices = "tf.ApproxTopK"(%arg0) {aggregate_to_topk = true, is_max_k = true, k = 4 : i64, recall_target = 8.500000e-01 : f32, reduction_dimension = 1 : i64, reduction_input_size_override = -1 : i64} : (tensor<1x4xf32>) -> (tensor<1x4xf32>, tensor<1x4xi32>)
  func.return %values, %indices: tensor<1x4xf32>, tensor<1x4xi32>

  %cst = arith.constant dense<[[1,2]]> : tensor<1x2xi32>
  %cst1 = arith.constant dense<[[3.0, 4.0]]> : tensor<1x2xf32>
  %0 = "tfl.sqrt"(%arg0) : (tensor<1x1xf32>) -> tensor<1x1xf32>
  %1 = "tfl.tile"(%0, %cst) : (tensor<1x1xf32>, tensor<1x2xi32>) -> tensor<1x2xf32>
  %2 = "tfl.add"(%cst1, %1) {fused_activation_function = "NONE"} : (tensor<1x2xf32>, tensor<1x2xf32>) -> tensor<1x2xf32>

      %%arg1 = "tf.Const"() {value = dense<-3.0> : tensor<1x2x4xf32>} : () -> tensor<1x2x4xf32>

      %%1 = "tf.%s"(%%arg0, %%arg1) {T = f32, adj_x = false, adj_y = false, grad_x = false, grad_y = false, device = ""} : (tensor<1x4x2xf32>, tensor<1x2x4xf32>) -> tensor<1x4x4xf32>

      func.return %%1 : tensor<1x4x4xf32>
    }
  })";
  std::string mat_mul_method =

}

// CHECK-LABEL: QuantizeConcat
func.func @QuantizeConcat(tensor<1x2xf32>, tensor<1x2xf32>) -> tensor<2x2x!quant.uniform<u8:f32, 1.000000e-01:128>> {
^bb0(%arg0: tensor<1x2xf32>, %arg1: tensor<1x2xf32>):
  %0 = "tfl.concatenation"(%arg0, %arg1) {axis = 0 : i32, fused_activation_function = "NONE"} : (tensor<1x2xf32>, tensor<1x2xf32>) -> tensor<2x2xf32>