%0 = call @XlaCallModule__resize_0(%arg0) : (tensor<1x2x2x10xf32>) -> tensor<1x4x4x10xf32>
  return %0 : tensor<1x4x4x10xf32>
}
func.func private @XlaCallModule__resize_0(%arg0: tensor<1x2x2x10xf32>) -> (tensor<1x4x4x10xf32>) {
  %0 = mhlo.constant dense<2> : tensor<i32>
  %1 = mhlo.constant dense<0> : tensor<i32>
  %2 = mhlo.constant dense<4.000000e+00> : tensor<f32>

func.func @attribute_cast(%arg0: tensor<1x4x4x1xf32>) -> tensor<1x2x2x1xf32> {
  %0 = "tfr.cast"(%arg0) : (tensor<1x4x4x1xf32>) -> !tfr.tensor
  %stride_i32 = arith.constant 2 : i32
  %1 = tfr.call @tf__my_max_pool(%0, %stride_i32, %stride_i32) : (!tfr.tensor, i32, i32) -> !tfr.tensor
  %2 = "tfr.cast"(%1) : (!tfr.tensor) -> tensor<1x2x2x1xf32>
  func.return %2 : tensor<1x2x2x1xf32>
// CHECK: tf__max_pool
}

    %dimension = "tf.Const"() { value = dense<3> : tensor<1xi64> } : () -> tensor<1xi64>
    %6 = "tf.Sum"(%3, %dimension) { keep_dims = true }: (tensor<1x3x2x2xf32>, tensor<1xi64>) -> tensor<1x3x2x1xf32>
    return %5, %6 : tensor<1x3x2x2xf32>, tensor<1x3x2x1xf32>
  }

// CHECK-LABEL: func @conv_with_multiple_uses
// CHECK: %[[div:.*]] = "tf.Div"(%arg0
// CHECK: %[[add:.*]] = "tf.AddV2"(%[[div]]
// CHECK: %[[maximum:.*]] = "tf.Maximum"(%[[add]]

func.func @depth_to_space(%arg0: tensor<1x1x1x4xf32>) -> tensor<1x2x2x1xf32> {
  %0 = "tf.DepthToSpace"(%arg0) {block_size = 2: i64,  data_format = "NHWC"}: (tensor<1x1x1x4xf32>) -> tensor<1x2x2x1xf32>
  func.return %0 : tensor<1x2x2x1xf32>
// CHECK: %[[CUSTOM_0:.*]] = "tfl.custom"(%arg0) <{custom_code = "FlexDepthToSpace", custom_option = #tfl<const_bytes : "{{.*}}">}> : (tensor<1x1x1x4xf32>) -> tensor<1x2x2x1xf32>
// CHECK: return %[[CUSTOM_0]] : tensor<1x2x2x1xf32>
}

      stablehlo.return %6 : tensor<f32>
  }) {
    window_dimensions = array<i64: 1, 1, 2, 2>,
    window_strides = array<i64: 1, 1, 2, 2>
  } : (tensor<1x4x5x5xf32>, tensor<f32>) -> tensor<1x4x2x2xf32>
  return %4 : tensor<1x4x2x2xf32>
}
// CHECK-DAG: %[[WEIGHT_CONST:.+]] = stablehlo.constant {{.*}} : tensor<3x3x2x4xf32>
// CHECK-DAG: %[[BIAS_CONST:.+]] = stablehlo.constant {{.*}} : tensor<1x5x5x4xf32>

func.func @depth_to_space(%arg0: tensor<1x1x1x4xf32>) -> tensor<1x2x2x1xf32> {
  %0 = "tf.DepthToSpace"(%arg0) {block_size = 2: i64,  data_format = "NHWC"}: (tensor<1x1x1x4xf32>) -> tensor<1x2x2x1xf32>
  func.return %0 : tensor<1x2x2x1xf32>
// CHECK: %[[CUSTOM_0:.*]] = "tfl.custom"(%arg0) <{custom_code = "FlexDepthToSpace", custom_option = #tfl<const_bytes : "{{.*}}">}> : (tensor<1x1x1x4xf32>) -> tensor<1x2x2x1xf32>
// CHECK: return %[[CUSTOM_0]] : tensor<1x2x2x1xf32>
}

  %1 = "tfl.reshape"(%0, %shape) : (tensor<40x40xf32>, tensor<4xi32>) -> tensor<1x40x4x10xf32>
  %2 = "tfl.add"(%1, %cst2) {fused_activation_function = "NONE"} : (tensor<1x40x4x10xf32>, tensor<1x1x4x10xf32>) -> tensor<1x40x4x10xf32>

  func.return %2 : tensor<1x40x4x10xf32>

  // CHECK-DAG: %[[cst:.*]] = arith.constant dense<2.000000e+00> : tensor<40xf32>

}

func.func @GroupConv(%arg0: tensor<?x1x26x14xf32>, %arg1: tensor<1x3x2x14xf32>) -> (tensor<?x1x6x14xf32>) {
  %0 = "tf.Conv2D"(%arg0, %arg1) {data_format = "NHWC", device = "", dilations = [1, 1, 1, 1], explicit_paddings = [], padding = "SAME", strides = [1, 1, 5, 1], use_cudnn_on_gpu = true} : (tensor<?x1x26x14xf32>, tensor<1x3x2x14xf32>) -> tensor<?x1x6x14xf32>
  func.return %0 : tensor<?x1x6x14xf32>
  // CHECK-LABEL: GroupConv

// CHECK-NEXT:    return %[[VAL_2]] : tensor<1x2x4x1xf32>
// CHECK-NEXT:  }
}

// -----

module {
// expected-warning @+1 {{Invalid number of results from MaxUnpooling2D}}

  %0:2 = "tf.SplitV"(%arg0, %arg1, %arg2) : (tensor<1x4x3x3xf32>, tensor<2xi32>, tensor<i32>) -> (tensor<1x4x2x3xf32>, tensor<1x4x1x3xf32>)
  func.return %0#0 : tensor<1x4x2x3xf32>

  // CHECK-LABEL: splitv
  // CHECK: "tfl.split_v"(%arg0, %arg1, %arg2) <{num_splits = 2 : i32}> : (tensor<1x4x3x3xf32>, tensor<2xi32>, tensor<i32>) -> (tensor<1x4x2x3xf32>, tensor<1x4x1x3xf32>)
}