You're right, I've certainly been a bit presumptuous to call this'a benchmark'. It is indeed a flawed test. Yet,It's been giving me the occasion to try some ope

You're right, I've certainly been a bit presumptuous to call this'a benchmark'. It is indeed a flawed test. Yet,It's been giving me the occasion to try some open source models and for my workflow, some of them are incredibly competitive with sota closed source models.