Timezone: »

Poster
CSER: Communication-efficient SGD with Error Reset
Cong Xie · Shuai Zheng · Sanmi Koyejo · Indranil Gupta · Mu Li · Haibin Lin

Wed Dec 09 09:00 AM -- 11:00 AM (PST) @ Poster Session 3 #1130
The scalability of Distributed Stochastic Gradient Descent (SGD) is today limited by communication bottlenecks. We propose a novel SGD variant: \underline{C}ommunication-efficient \underline{S}GD with \underline{E}rror \underline{R}eset, or \underline{CSER}. The key idea in CSER is first a new technique called `error reset'' that adapts arbitrary compressors for SGD, producing bifurcated local models with periodic reset of resulting local residual errors. Second we introduce partial synchronization for both the gradients and the models, leveraging advantages from them. We prove the convergence of CSER for smooth non-convex problems. Empirical results show that when combined with highly aggressive compressors, the CSER algorithms accelerate the distributed training by nearly $10\times$ for CIFAR-100, and by $4.5\times$ for ImageNet.