GIVEAWAY 5 GTD & 10 FCFS
Want a guaranteed spot?
Requirements:
• Follow @hizuminft and me
• Like + RT the pinned post @hizuminft
⏰ 24 hours deadline
Drop your wallet below 👇
Introducing RL Swarm 72B
@GensynAI introduces RL Swarm 72B, a fully decentralized reinforcement learning system for training a 72B parameter model. There are no whitelists and no barriers. Anyone can contribute compute and participate in large scale RL training.
Introducing SkipPipe
Developed with researchers from the University of Neuchatel and TU Delft, SkipPipe is a new pipeline parallel training method that cuts distributed training time by up to 55 percent and is designed for decentralised environments.
@gensynai