DisTrO

This is the repository for DisTrO (Distributed Training Over-The-Internet), a family of low latency distributed optimizers that reduce inter-GPU communication requirements by three to four orders of magnitude.

• Aug. 26th, 2024: DisTrO (Preliminary Report)
• Dec. 2nd, 2024: DeMo Optimization (Paper) (Code), original seed research/idea for DisTrO
• Dec. 2nd, 2024: Nous trains a 15b model using DisTrO
• May 14th, 2025: Psyche Network
• May 14th, 2025: Nous Consilience 40b LLM, Huggingface
• Oct. 14th, 2025: DeMo Optimization (Paper v2), (Production Code)

Join us on Discord if you're interested in helping research and build the future of distributed training.