%w = arith.constant dense<2.0> : tensor<3x3x3x3xf32>
  %q = "tfl.quantize"(%w) {qtype = tensor<3x3x3x3x!quant.uniform<i8<-127:127>:f32:0,{1.0,2.0,3.0}>>} : (tensor<3x3x3x3xf32>) -> tensor<3x3x3x3x!quant.uniform<i8<-127:127>:f32:0,{1.0,2.0,3.0}>>
  %dq = "tfl.dequantize"(%q) : (tensor<3x3x3x3x!quant.uniform<i8<-127:127>:f32:0,{1.0,2.0,3.0}>>) -> tensor<3x3x3x3xf32>

    strides = [1, 1, 1, 1],
    use_cudnn_on_gpu = true
  } : (tensor<4xi32>, tensor<2x2x5x21xf32>, tensor<5x2x2x21xf32>) -> tensor<5x3x3x15xf32>
  func.return %result : tensor<5x3x3x15xf32>
}


// CHECK-LABEL: @conv3d_backprop_input
func.func @conv3d_backprop_input(%filter: tensor<3x3x3x1x6xf32>, %out_backprop: tensor<2x8x8x8x6xf32>) -> tensor<2x8x8x8x1xf32> {

  %0 = "mhlo.broadcast_in_dim"(%arg0) <{broadcast_dimensions = dense<[1, 2, 3]> : tensor<3xi64>, name = "broadcast.0"}> : (tensor<8x1x16xf32>) -> tensor<3x8x8x16xf32>
  func.return %0 : tensor<3x8x8x16xf32>
}

// CHECK-LABEL:   func @broadcast_in_dim_general_case(
// CHECK-SAME:                                        %[[VAL_0:.*]]: tensor<3x1x16xf32>) -> tensor<3x8x8x16xf32> {