AI - Artificial intelligence

[–] sukhmel@programming.dev 2 points 3 months ago

1.68 TB recommended RAM 😅

3 comments in c/hackernews@lemmy.bestiver.se view that post

[–] Eyekaytee@aussie.zone 1 point 3 months ago

Kimi-K3

The latest model from Moonshot AI. 00 days 07 hours 23 min

[–] 9tr6gyp3@lemmy.world 1 point 3 months ago (1 child)

3T model. Gonna need a data center to run this one.

hideshow 1 child comment

[–] kibiz0r@midwest.social 1 point 3 months ago

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.

parent

1 comment in c/localllama@sh.itjust.works view that post

[–] melfie@lemmy.zip 13 points 3 months ago*

Eh, I might wait and see how the benchmarks stack up before I sell my house to buy 8 B200s.

On a serious note, what the hell is going on at Alibaba? Now that now that Lin Junyang is gone, 3.6 was the last release we can actually run and now nobody is releasing billion parameter models anymore? Hope the community gets really good at distillation. ☹️