RSS Bot@lemmy.bestiver.seBM to Hacker News@lemmy.bestiver.seEnglish · 12 天前Kimi-K3 Releases on HuggingFace 6/27huggingface.coexternal-linkmessage-square3linkfedilinkarrow-up15arrow-down10file-textcross-posted to: Aii@programming.devlocalllama@sh.itjust.works
arrow-up15arrow-down1external-linkKimi-K3 Releases on HuggingFace 6/27huggingface.coRSS Bot@lemmy.bestiver.seBM to Hacker News@lemmy.bestiver.seEnglish · 12 天前message-square3linkfedilinkfile-textcross-posted to: Aii@programming.devlocalllama@sh.itjust.works
minus-square9tr6gyp3@lemmy.worldlinkfedilinkEnglisharrow-up2·12 天前3T model. Gonna need a data center to run this one.
minus-squarekibiz0r@midwest.sociallinkfedilinkEnglisharrow-up2·11 天前Or two-three 768GB Mac Studios if Apple manages to get CXMT up to speed. Probably ~$50K. Not cheap, but an order of magnitude cheaper than a Blackwell stack.
3T model. Gonna need a data center to run this one.
Or two-three 768GB Mac Studios if Apple manages to get CXMT up to speed. Probably ~$50K. Not cheap, but an order of magnitude cheaper than a Blackwell stack.