Decentralized training works — and it's happening on Akash
AI Made in Cosmos - with Greg Osuri, Murthy Vitwit, Valery Litvin & Dean Tribble (Cosmoverse)
“Can we actually create truly decentralized systems that are trained on decentralized compute networks? And the answer is yes, we can. The first example we saw was a billion parameter model by Nous Research which trained truly on distributed GPUs.” — 00:09:27
Context: Cites Nous’s DisTrO framework enabling low-bandwidth internet training; continues that Prime Intellect’s 10B-parameter run was 15% complete — “both these models are being trained on Akash.”