King to Technology@lemmy.worldEnglish · 2 months agoThese Mini AI Models Match OpenAI With 1,000 Times Less Data.singularityhub.comexternal-linkmessage-square8fedilinkarrow-up1128arrow-down116
arrow-up1112arrow-down1external-linkThese Mini AI Models Match OpenAI With 1,000 Times Less Data.singularityhub.comKing to Technology@lemmy.worldEnglish · 2 months agomessage-square8fedilink
minus-squareGeneral_Effort@lemmy.worldlinkfedilinkEnglisharrow-up1·2 months agoAfter a quick skim, seems like the article has lots of errors. Molmo is trained on top of Qwen. The smallest ones are trained on something by the same company as Molmo.
After a quick skim, seems like the article has lots of errors. Molmo is trained on top of Qwen. The smallest ones are trained on something by the same company as Molmo.