download_document fails on original document format docx

[maharg101]

When downloading a pre-translated document using is_xliff="false" the download_document attempts to decode the api response as UTF-8, which results in a UnicodeDecodeError.

To reproduce, upload a docx file for pretranslation. Once processed, download:

api_instance.download_document(job_id, is_xliff="false")

Observe the following error:

  File "/path/to/src/lilt-pipeline-wrapper/venv/lib/python3.7/site-packages/lilt/api_client.py", line 205, in __call_api
    response_data.data = response_data.data.decode(encoding)
UnicodeDecodeError: 'utf-8' codec can't decode byte 0xce in position 10: invalid continuation byte

I found mention of a bytes response type at

lilt-python/lilt/api_client.py

Line 200 in 364b633

if six.PY3 and response_type not in ["file", "bytes"]:

but it is not clear if / how this may be used.