Shoutout to nostr:npub1ex7mdykw786qxvmtuls208uyxmn0hse95rfwsarvfde5yg6wy7jq6qvyt9 and nostr:npub1cn4t4cd78nm900qc2hhqte5aa8c9njm6qkfzw95tszufwcwtcnsq7g3vle these Ross hoodies are sick 😮💨
Is it just me or is PoE magic? #asknostr
There’s no way, that’s like 100x more efficient than an S21- I think that link is a scam
Wait there’s a 70 TH machine that draws 15W? That seems impossible https://avalonasic.org/product/avalon-nano-3-4th-s/
Dante crushing it per usual- well done!
Here's some #freeross jams 🎶
Have you looked at the training methods used by deepseek for their recent model? It splits the training up via horizontally per node with this DualPipe. From link (https://adasci.org/deepseek-v3-explained-optimizing-efficiency-and-scale/), might work at more distributed scale as it needs close to zero all-to-all communication
Interesting, I’ll take a look at those- I was thinking it might be good to keep the training deliberately slow- ie train over a week or two to enable the LLM to see globally distributed data and train on that- similar to how Bitcoin keeps blocks at 10 min- what do you think?
Wtf did this just happen? Minutes before the ceremony?
#asknostr
I had an idea last night about distributed data preparation for AI training. AI models are trained by RLHF with biased individuals, fuck that. Let's distribute this shit, here's an idea:
- Plebs host data relays that scrape and process data from Nostr to a standard (read: in-code) Nostr note type
- distribute some H100 instances to 'slowly' train a public LLM (leveraging DeepSeekV3's approach to distributed training)
- training slowly to agree on model updates: 2 weeks collect new processed data notes, 1 week to train on the new data, grab timestamp from the best timestamp server available (#Bitcoin blockheight), 1 week to distribute new LLM weights. Repeat.
This let's plebs focus on data they are knowledgeable about - hopefully get enough coverage for a general LLM - and should give a public AI advocate a shot... No more censored BS LLMs gaslighting us.. let's take this shit back
#GM nostr, have a good day and God Bless
Let's hope nostr:npub1qtvl2em0llpnnllffhat8zltugwwz97x79gfmxfz4qk52n6zpk3qq87dze hits a block during this inaugural fee spike - they seem to keep missin' them.. 🤔
nostr:npub1py0w8ymhdsll4tksvk69u0ethqdp32k4cdcmwhlaz8sfv6p8as8sc77ryh
Cerra un poco los ojos
#negr0art
Very cool!
I think Fauci will go down as one of the most demonic actors of our time, so will the 11th hour pardon from Biden. smh.
#lableak
I enjoyed the Hyperion series- big world building guy, so this and the Dune series get me goin
Hear me out... a DVM that filters my phone calls so I don't get 1,000 spam calls per day... 🤔 #Nostrthoughts
Do we have a way to efficiently index relays for DVMs to use a RAG like LLM?
