Its interesting this paper & the Orca LLM paper from Microsoft are using GPT3/4 model outputs to train 'powerful' models. Big question is will they allow your a

Its interesting this paper & the Orca LLM paper from Microsoft are using GPT3/4 model outputs to train 'powerful' models. Big question is will they allow your average joes/businesses to do the same on their own data? Doubt it, considering it breaks OAIs terms of use at present. Will opensource lead the way on this? Bring on Llama-v2