int dilation_i = mlir::cast<IntegerAttr>(dilations[i]).getInt();
      int out_size = tflite::ComputeOutSize(kTfLitePaddingSame, input_size,
                                            filter_size, stride_i, dilation_i);

      int offset = 0;
      int padding_before = tflite::ComputePaddingWithOffset(
          stride_i, dilation_i, input_size, filter_size, out_size, &offset);

  IntegerAttr dilation_depth_factor, dilation_height_factor,
      dilation_width_factor;
  if (!TFIntListIs1XYZ1(op, "dilations", &dilation_depth_factor,
                        &dilation_height_factor, &dilation_width_factor)) {
    // If the 'dilations' attribute is missing, we use the default value (1)
    // for all dilation depth, height and width factor.
    dilation_depth_factor = rewriter.getI32IntegerAttr(1);

        offset = [0.5] * self.out_channel_size
        mean, variance = scale, offset
        out = nn_ops.conv2d(
            input_tensor,
            self.filters,
            strides=strides,
            dilations=dilations,
            padding=padding,
            data_format='NHWC',
            name='sample/conv',
        )
        if bias_fn is not None:
          out = nn_ops.bias_add(out, self.bias)

     IsIntList1XY1:$strides,
     BoolAttr:$use_cudnn_on_gpu,
     IsSameOrValid:$padding,
     I64ArrayAttr:$explicit_paddings,
     IsDataFormatNHWC:$data_format,
     IsAllOnes:$dilations),
  (TFL_TransposeConvOp $input_sizes,
     (TFL_TransposeOp $filter,
       (Arith_ConstantOp ConstantAttr<I32VectorElementsAttr<4>, "{2, 0, 1, 3}">)),
     $out_backprop,
     /*bias=*/ (CreateNoneValue $input_sizes),