Re:PMC-LLaMA: Further Finetuning LLaMA on Medical Papers
by opisthenar84 on Tuesday May 09, 2023 @01:30AM
(#35870511)
With 75B tokens, would it be better to train from scratch, then apply instruction fine-tuning? Training from llama seems like it could introduce a lot of unexpected behavior.
[ Parent
| on HN ]
Re:PMC-LLaMA: Further Finetuning LLaMA on Medical Papers
by Taek on Tuesday May 09, 2023 @02:52AM
(#35870984)
75B tokens is not really enough data to make an intelligent model. Llama was trained on over 1 trillion tokens.
And yes, training on top of LLaMA could introduce a lot of unexpected behavior, but that's just where the State-of-the-Art is today