Stochastic Gradient Methods with Compressed Communication for Decentralized Saddle Point Problems

Open in new window