Re:GPT-4 Trained on 10x More Data, Can Write 60k Word Books
by minimaxir on Wednesday January 11, 2023 @01:56PM
(#34342802)
This misinformed GPT-4 take is new levels of comically bad because it correlates number of model hyperparameters with the quantity of training data.
[ Parent
| on HN ]
There are no comments on this comment yet.