RuntimeError: "upsample_bilinear2d_channels_last" not implemented for 'Long' after converting np array to list

Question

I am trying to send image data across modules.

The following works fine:

def process_image(pdf_path, page_dimensions):
  pdf_path = get_pdf(pdf_path, False)
  pdf_name = os.path.basename(pdf_path)
  with tempfile.TemporaryDirectory() as path:
    conversion_time = time.time()
    chart_images = convert_from_path(
          pdf_path=pdf_path,
          dpi=300,
          fmt="jpg",
          output_file=os.path.basename(str(pdf_path)).split(".")[0],
          output_folder=path,
          use_pdftocairo=False,
          paths_only=True,
          thread_count=8,
      )
    pg_dim, pg_image_path = page_dimensions[0], chart_images[0]
    pg_image = cv2.imread(pg_image_path, cv2.IMREAD_UNCHANGED)
    pg_image = cv2.resize(pg_image, (pg_dim[1], pg_dim[2]))
  return pg_image

The result of this function (pg_image) works fine when taken as input to my Detectron2 model.

However, when I send pg_image.tolist() and convert back to np array on receiving (np.array(pg_image)) and send it to my Detectron2 model, I keep getting the following error:

RuntimeError: "upsample_bilinear2d_channels_last" not implemented for 'Long'

So, if converting the np array to a list and converting back changing the data?