@JOMusic@lemmy.ml to Technology@lemmy.worldEnglish • 2 months agoUS Bill proposed to jail people who download Deepseekwww.404media.coexternal-linkmessage-square128fedilinkarrow-up1813arrow-down10cross-posted to: technology@beehaw.orgnottheonion@lemmy.worldusa@lemmy.mlopensource@lemmy.mlpolitics@lemmy.world
arrow-up1813arrow-down1external-linkUS Bill proposed to jail people who download Deepseekwww.404media.co@JOMusic@lemmy.ml to Technology@lemmy.worldEnglish • 2 months agomessage-square128fedilinkcross-posted to: technology@beehaw.orgnottheonion@lemmy.worldusa@lemmy.mlopensource@lemmy.mlpolitics@lemmy.world
minus-square@FrederikNJS@lemm.eelinkfedilinkEnglish5•2 months ago32b is still distilled. The full one is 671b.
minus-square@DogWater@lemmy.worldlinkfedilinkEnglish2•2 months agoI know, but the fall off in performance isn’t supposed to be severe
minus-square@FrederikNJS@lemm.eelinkfedilinkEnglish1•2 months agoYou are correct. And yes that is kinda the whole point of the distilled models.
32b is still distilled. The full one is 671b.
I know, but the fall off in performance isn’t supposed to be severe
You are correct. And yes that is kinda the whole point of the distilled models.
I know. Lmao