Comment by mvieira38

The results are interesting for showing the efficacy of small, fine-tuned models that can be run locally. AI providers as a business need their do-all models to be better than these if they want long-term revenue through the APIs, right?