Closed jamesbraza closed 1 month ago
should probably add a better unit test to call query, then add a few more, then call query again to get coverage of mixed
to_embed != self.texts
But can wait until future for that
@whitead this was actually somewhat hard to test (which is a good sign) because our API has so many checks in place for this. That being said, I have written a test test_partly_embedded_texts
that fails on main
and passes here. Cheers 🥳
https://github.com/Future-House/paper-qa/pull/289 changed the logic of
add_texts
such that the retrying logic of OpenAI rate limits is no longer relevant. Other comments were stale too, so this PR just cleans them up