gpus and tractors. Neural networks from Scratch book: nnfs.io piped.video/sentdex
Alabama, USA
- Tweets 19,092
- Following 398
- Followers 112,564
- Likes 23,816
Pinned Tweet
dario begged me in dms to not do agent swarms at home but whatever, this is nemotron 120B A12B running 64 concurrents at a bit over 100 tok/sec each for an aggregate of almost 7K tok/sec on a single GB300 (Asus DGX Station)
believe it or not, the agents did NOT magically just start hacking the planet tho. weird.
180
171
3,143
Harrison Kinsley retweeted
We're shipping Asimov 1!
Builders have already received their kits in 🇺🇸 the US, 🇩🇪 Germany, 🇫🇷 France, 🇵🇱 Poland, 🇹🇭 Thailand, and 🇦🇪 the UAE.
We're forming Batch 4 now.
Order now: menlo.ai
9
19
260
Unless it beats fable/astra, being 2.5T+ would be a mistake imo.
We love these models because they are smart, but also bc fast and cheap! Idc anymore about astra/fable...etc bc local ai is so damn fast, which makes it useful.
Maybe they really just want a china AI top charts
1
1
12
Yep. People just wanna skip the work of building a real fanbase, but even that doesnt have to be hard.
@YourLocalAILab is a great example of doing it right. I shill that place because i actually love it, use it, and think people will benefit by knowing it exists. Pretty simple.
4
Im not even against paid promotions. Its just the totally lazy, effortless, soulless crap. I think u can become rich w/o selling out. The prob is its never enough.
I also cannot imagine shilling anything ive not used extensively.
1
2
I get endless of these. They always just say the name and price. Or ask me my price for a paid post and dont even tell me about the product, as if that part is irrelevant.
For everyone promoting things this way on all sides, it *is* irrelevant.
6
27
The moment i see a video like this i move along. Accts that push this stuff with lazy effort get muted.
You might get a big view count that you can incl in ur next investor email, but people are getting fed up wit this slop.
I already forgot the company in the vid.
i find it really sad that X is becoming a pay to win platform and that even "good" releases are going through these fully slop/fake campaigns to get traction
i'm also pretty sure there is very shady stuff where they are boosting each other's posts with bot farms and more. would be great to investigate here @elonmusk and make sure the algorithm doesn't push their posts, you can very easily cluster that if you look at each release and the accounts amplifying 🙏
(not a dunk on recursive, recently a lot of people are doing that unfo and it's definitely working, posts often get over ~1M views)
4
1
70
I dont think anyone is winning doing this to be honest. Everyone's harming their rep imho.
Even just these stupid over produced videos are annoying. I think people are just copying what everyone else is doing thinking it's the way.
23
to clear, its not really about just having some huge # of tokens. Its more about how useful the model is. If you wanna just get silly about optimizing token use, you absolutely can. It's just rare that I hit 1B tokens this fast on a local model.
1
you can host comfy on 4x rtx pro 6k or 1 dgx station + pro 6k sidecar.
it feels really good on the dgx station, check @original_el8 recipe here: github.com/original-el8/dgx-…
21
so is it opus or sonnet? pic suggests opus. 130B tokens of opus 5.5 for $100 sub?
idk man. havent used it in a while, but that doesn't add up
1
no plugins yet for dsh, but I might be using it as a foundation for agent swarm stuff and might use a plugin to do it.
1