3 ms·
My understanding is that the models are pretty comparable, but nobody's reinforcement training set is not nearly as good as OpenAI's, so they're able to fine-tu
by adastra22 3y ago
My understanding is that the models are pretty comparable, but nobody's reinforcement training set is not nearly as good as OpenAI's, so they're able to fine-tune their model to give more accurate results.