2000 character limit reached
Gradient Coding (1612.03301v2)
Published 10 Dec 2016 in stat.ML, cs.DC, cs.IT, cs.LG, math.IT, and stat.CO
Abstract: We propose a novel coding theoretic framework for mitigating stragglers in distributed learning. We show how carefully replicating data blocks and coding across gradients can provide tolerance to failures and stragglers for Synchronous Gradient Descent. We implement our schemes in python (using MPI) to run on Amazon EC2, and show how we compare against baseline approaches in running time and generalization error.