INTELLECT-2: A Reasoning Model Trained Through Globally Decentralized Reinforcement Learning

Open in new window