HN user
decentralized model training
Got any examples? Would be interesting to see what's out there...