// CHECK:           [[VAL_5:%.*]] = call @func_1_GPU_FLOAT([[VAL_0]], [[VAL_3]]) {tac.device = "GPU", tac.inference_type = "FLOAT", tac.interface_name = "func_1"} : (tensor<100xf32>, tensor<100xf32>) -> tensor<100xf32>

  %2 = func.call @fold_all_test(%arg0, %0, %1) : (tensor<256x32x32x3xf32>, tensor<16x3x3x3xf32>, tensor<16xf32>) -> tensor<256x30x30x16xf32>
  func.return %2 : tensor<256x30x30x16xf32>
}

// ALL-LABEL: @fold_all_test
func.func @fold_all_test(%arg0: tensor<256x32x32x3xf32>, %arg1: tensor<16x3x3x3xf32>, %arg2: tensor<16xf32>) -> tensor<256x30x30x16xf32> {

      Targs = [tensor<32xf32>, tensor<64xf32>, tensor<8xf32>, tensor<2xf32>],
      Tresults = [tensor<32xf32>, tensor<64xf32>, tensor<8xf32>]}
      : (tensor<32xf32>, tensor<64xf32>, tensor<8xf32>, tensor<2xf32>, tensor<2x!tf_type.string>)
        -> (tensor<32xf32>, tensor<64xf32>, tensor<8xf32>)
    tf_device.return %0#0, %0#1, %0#2 : tensor<32xf32>, tensor<64xf32>, tensor<8xf32>

func.func @add(%arg0: tensor<1xf32>, %arg1: tensor<1xf32>) -> tensor<1xf32> {
  %0 = "tf.Add"(%arg0, %arg1) : (tensor<1xf32>, tensor<1xf32>) -> tensor<1xf32>
  func.return %0: tensor<1xf32>
// CHECK: %[[CUSTOM_0:.*]] = "tfl.custom"(%arg0, %arg1) <{custom_code = "FlexAdd", custom_option = #tfl<const_bytes : "0x03416464001412034164641A001A002A070A015412023001320000021B171414042801">}> : (tensor<1xf32>, tensor<1xf32>) -> tensor<1xf32>

  %min = arith.constant dense<0.0> : tensor<16xf32>
  %max = arith.constant dense<255.0> : tensor<16xf32>
  %mini = "tf.Identity"(%min) : (tensor<16xf32>) -> tensor<16xf32>
  %maxi = "tf.Identity"(%max) : (tensor<16xf32>) -> tensor<16xf32>
  %fq = "tf.FakeQuantWithMinMaxVarsPerChannel"(%in, %mini, %maxi) {num_bits = 5, narrow_range = false} : (tensor<3x3x3x16xf32>, tensor<16xf32>, tensor<16xf32>) -> tensor<3x3x3x16xf32>

  %min = arith.constant dense<0.0> : tensor<16xf32>
  %max = arith.constant dense<15.0> : tensor<16xf32>
  %mini = "tf.Identity"(%min) : (tensor<16xf32>) -> tensor<16xf32>
  %maxi = "tf.Identity"(%max) : (tensor<16xf32>) -> tensor<16xf32>
  %fq = "tf.FakeQuantWithMinMaxVarsPerChannel"(%in, %mini, %maxi) {num_bits = 3, narrow_range = false} : (tensor<3x3x3x16xf32>, tensor<16xf32>, tensor<16xf32>) -> tensor<3x3x3x16xf32>

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

  // CHECK: return %[[y]] : tensor<1xf32>
  func.return %0 : tensor<1xf32>
}

// -----

// CHECK-LABEL: func @uniform_quantize_dequantize
func.func @uniform_quantize_dequantize(%arg0: tensor<1xf32>) -> tensor<1xf32>
{
  %scales = "tf.Const"() { value = dense<1.0> : tensor<f32> } : () -> tensor<f32>

  %0 = "tfr.cast"(%arg0) : (tensor<10x!quant.uniform<i8:f32, 0.0039133410900831223:-128>>) -> !tfr.tensor
  %1 = tfr.quant_raw_data(%0) : (!tfr.tensor) -> !tfr.tensor
  %2 = "tfr.cast"(%1) : (!tfr.tensor) -> tensor<10xi8>
  %3 = "tf.Cast"(%2) {Truncate = false} : (tensor<10xi8>) -> tensor<10xi32>
  func.return %3 : tensor<10xi32>

    %4 = "tf.LayerNorm"(%a1, %a2, %a3, %a4) {_tfl_quant_trait = "fully_quantizable", device = ""} : (tensor<128x128xf32>, tensor<1xf32>, tensor<1xf32>, tensor<1xi32>) -> tensor<128x128xf32>
   "tfl.yield"(%4) : (tensor<128x128xf32>) -> ()
  }) {_tfl_quant_trait = "fully_quantizable", device = ""} : (tensor<128x128xf32>, tensor<1xf32>, tensor<1xf32>, tensor<1xi32>) -> tensor<128x128xf32>

func.func @testBroadcastGradientArgs5() -> (tensor<1xi32>, tensor<1xi32>) {
  %s4 = "tf.Const"() {value = dense<[]> : tensor<0xi32>} : () -> tensor<0xi32>
  %s5 = "tf.Const"() {value = dense<[1]> : tensor<1xi32>} : () -> tensor<1xi32>
  %r4, %r5 = "tf.BroadcastGradientArgs"(%s4, %s5) {} : (tensor<0xi32>, tensor<1xi32>) -> (tensor<1xi32>, tensor<1xi32>)
  // CHECK: %[[R0:.*]] = "tf.Const"() <{value = dense<0> : tensor<1xi32>}> : () -> tensor<1xi32>