Example of using the new Flatbuffers Python API to modify a TensorFlow Lite file

strip_names.py

def StripTfliteFile(tflite_input, tflite_output):
  """Given 'tflite_input`, write stripped version, 'tflite_output'."""
  if not os.path.exists(tflite_input):
    raise RuntimeError("Invalid filename %r" % tflite_input)
  with open(tflite_input, "rb") as file_handle:
    file_data = bytearray(file_handle.read())
  model_obj = schema_fb.Model.GetRootAsModel(file_data, 0)
  model = schema_fb.ModelT.InitFromObj(model_obj)
  model.description = ""
  for subgraph in model.subgraphs:
    subgraph.name = ""
    for tensor in subgraph.tensors:
      tensor.name = ""
  builder = flatbuffers.Builder(1024)
  model_offset = model.Pack(builder)
  builder.Finish(model_offset)
  model_data = builder.Output()
  with open(tflite_output, "wb") as out_file:
    out_file.write(model_data)

Hmmm, Netron says "RangeError: Offset is outside the bounds of the DataView ". Stripped model doesn't load in TFLite aswell: "ERROR: The model is not a valid Flatbuffer buffer"