Re:PMC-LLaMA: Further Finetuning LLaMA on Medical Papers
by randomcatuser on Tuesday May 09, 2023 @12:46AM
(#35870249)
can someone explain the PEFT vs Zero-shot results to me? (in the table) How can Llama go from 1.8 to around 68?
[ Parent
| on HN ]
Re:PMC-LLaMA: Further Finetuning LLaMA on Medical Papers
by kristjansson on Tuesday May 09, 2023 @01:14AM
(#35870423)
AIUI they (1) finetuned LLaMA on 75B tokens of medical papers, and called the resulting model PMC-LLaMA, (2) finetuned that model on task-specifc data, then (3) separately finetuned the base model they started with in (1) on the same task specific data for comparison. So, they're reporting 6 separate conditions: {LLaMA, PMC-LLaMA} x {Zero-shot, Finetune, PEFT}.
Could definitely use some more exposition around that though.