# Decentralized training works — and it's happening on Akash

- **Speaker:** Greg Osuri
- **Attribution:** attributed
- **Said:** 2024-10-24
- **Theme:** Decentralized AI — https://www.gregosuri.com/predictions/decentralized-ai/
- **Source:** AI Made in Cosmos - with Greg Osuri, Murthy Vitwit, Valery Litvin & Dean Tribble (Cosmoverse)
- **Watch at:** 00:09:27 — https://www.youtube.com/watch?v=OHQnZEaMEHU&t=567s
- **Canonical:** https://www.gregosuri.com/predictions/decentralized-ai/2024-10-24-decentralized-training-works-and-it-s-happening-on-6lqv/
- **Cite as:** Greg Osuri, "Decentralized training works — and it's happening on Akash," AI Made in Cosmos - with Greg Osuri, Murthy Vitwit, Valery Litvin & Dean Tribble (Cosmoverse), 24 Oct 2024, 00:09:27. https://www.youtube.com/watch?v=OHQnZEaMEHU&t=567s

## Quote — verbatim

> 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.

## Context — site annotation, not spoken

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."