Get the latest tech news
Intellect-2 Release: The First 32B Model Trained Through Globally Distributed RL
We're excited to release INTELLECT-2, the first 32B parameter model trained via globally distributed reinforcement learning. Unlike traditional centralized training efforts, INTELLECT-2 trains a reasoning language model using fully asynchronous RL across a dynamic, heterogeneous swarm of permissionless compute contributors.
SHARDCAST: A library for distributing large files via a HTTP-based tree-topology network that efficiently propagates updated model weights to the decentralized inference workers. Now, we’re focusing on tool-assisted reasoning, crowdsourcing higher-quality data, and optimizing our infrastructure and training recipe to build frontier open models. Prime Intellect Research Team: Sami Jaghouar, Justus Mattern, Jack Min Ong, Jannik Straube, Manveer Basra, Aaron Pazdera, Matthew Di Ferrante, Kushal Thaman, Felix Gabriel, Fares Obeid, Kemal Erdem, Michael Keiblinger, Johannes Hagemann
Or read this on Hacker News
