Track_Shovel@slrpnk.net to Microblog Memes@lemmy.worldEnglish · 5 days agoSubscription based KPIsslrpnk.netimagemessage-square141linkfedilinkarrow-up11.11Karrow-down127
arrow-up11.08Karrow-down1imageSubscription based KPIsslrpnk.netTrack_Shovel@slrpnk.net to Microblog Memes@lemmy.worldEnglish · 5 days agomessage-square141linkfedilink
minus-squaretyler@programming.devlinkfedilinkEnglisharrow-up4·5 days agoThey literally use them as base models for compression.
minus-squarejonathan@piefed.sociallinkfedilinkEnglisharrow-up3·5 days agoHow do they do that exactly?
minus-squarecrunchy@lemmy.dbzer0.comlinkfedilinkEnglisharrow-up4·5 days agoIt’s a process called distillation. One way of doing it is creating thousands of accounts to prompt a big flagship model, then use the output to pre-train their own model.
They literally use them as base models for compression.
How do they do that exactly?
It’s a process called distillation. One way of doing it is creating thousands of accounts to prompt a big flagship model, then use the output to pre-train their own model.