But why they didn't compare it to SOTA finetuned...(like vicuna playtus..)? ... ... | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

transformi on Sept 27, 2023 | parent | context | favorite | on: Mistral 7B

But why they didn't compare it to SOTA finetuned...(like vicuna playtus..)? ... smells a bit strange..

ShrigmaMale on Sept 27, 2023 [–]

Bc that's not as good a comparison? Foundation models are better compared to each other. Can apply vicuna/guanaco/orca/sharegpt/whatever data to it and then do more of an apples-to-apples comparison.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact