%0 = "tfl.unidirectional_sequence_lstm"(%arg0,...

}

func.func @div(%arg0: tensor<1xf32>, %arg1: tensor<1xf32>) -> tensor<1xf32> {
  %0 = "tf.Div"(%arg0, %arg1) : (tensor<1xf32>, tensor<1xf32>) -> tensor<1xf32>
  func.return %0: tensor<1xf32>

// CHECK-LABEL: div
// CHECK:  tfl.div %arg0, %arg1 {fused_activation_function = "NONE"} : tensor<1xf32>
// CHECK:  return
}

func.func @squaredDifferenceRelu(tensor<1xf32>, tensor<1xf32>) -> tensor<1xf32> {

  %12 = mhlo.clamp %9, %11, %1 : tensor<2xf32>
  %13 = mhlo.multiply %arg0, %12 : tensor<2xf32>
  %14 = mhlo.divide %13, %1 : tensor<2xf32>
  return %14 : tensor<2xf32>
}

// CHECK-LABEL:   func.func @hardswish(
// CHECK-SAME:                         %[[VAL_0:.*]]: tensor<2xf32>) -> tensor<*xf32> {
// CHECK:           %[[VAL_1:.*]] = "tfl.hard_swish"(%[[VAL_0]]) : (tensor<2xf32>) -> tensor<2xf32>

  %2 = "tfl.add"(%arg0, %arg3) {fused_activation_function = "RELU6", per_device_costs = {CPU = 5.0 : f32, GPU = 1.0 : f32}, tac.device = "GPU"} : (tensor<1xf32>, tensor<1xf32>) -> tensor<1xf32>
  %3 = "tfl.pack"(%1, %2) {axis = 0 : i32, per_device_costs = {CPU = 2.0 : f32, GPU = -1.0 : f32}, values_count = 2 : i32, tac.device = "CPU"} : (tensor<1xf32>, tensor<1xf32>) -> tensor<2x1xf32>

    } else {
      // Recurse on the subtypes in the variant/resource. Basically if the input
      // were:
      //   tensor<!tf_type.variant<tensor<?x8xf32>>>
      // and:
      //   tensor<!tf_type.variant<tensor<10x8xf32>>>
      // we'll try here to refine tensor<?x8xf32> with tensor<10x8xf32>.
      auto refined_subtype = mlir::cast<TensorType>(
          TypeMeet(lhs_element_type_with_subtype.GetSubtypes().front(),

///      : (tensor<f32>, tensor<f32>, tensor<f32>) -> tensor<5xf32>
///
/// Output would be:
///   %iota = "mhlo.iota"() {iota_dimension = 0 : i64} : () -> tensor<5xf32>
///   %scaled = "mhlo.multiply"(%iota, %delta)
///       {broadcast_dimensions = dense<[]> : tensor<0xi64>} :
///       (tensor<5xf32>, tensor<f32>) -> tensor<5xf32>
///   %result = "mhlo.add"(%scaled, %offset)

          %1 = call @then_branch_func(%arg1) : (tensor<*xf32>) -> tensor<*xf32>
          "tf.Yield"(%1) : (tensor<*xf32>) -> ()
        },  {
          %1 = call @else_branch_func(%arg1) : (tensor<*xf32>) -> tensor<*xf32>
          "tf.Yield"(%1) : (tensor<*xf32>) -> ()
        }) {is_stateless = false} : (tensor<i1>) -> tensor<*xf32>
    ```

    will be transformed into this functional operation