☆ Yσɠƚԋσʂ ☆@lemmy.ml to Technology@lemmy.mlEnglish · 6 months agoLlama 3-V: Matching GPT4-V with a 100x smaller model and 500 dollarsaksh-garg.medium.comexternal-linkmessage-square2fedilinkarrow-up19arrow-down18cross-posted to: [email protected][email protected][email protected]
arrow-up11arrow-down1external-linkLlama 3-V: Matching GPT4-V with a 100x smaller model and 500 dollarsaksh-garg.medium.com☆ Yσɠƚԋσʂ ☆@lemmy.ml to Technology@lemmy.mlEnglish · 6 months agomessage-square2fedilinkcross-posted to: [email protected][email protected][email protected]
minus-square☆ Yσɠƚԋσʂ ☆@lemmy.mlOPlinkfedilinkarrow-up2arrow-down2·6 months agoGood question, not sure how they account for that. Maybe there’s a higher level layer responsible for dealing with the boundaries?
Good question, not sure how they account for that. Maybe there’s a higher level layer responsible for dealing with the boundaries?