CompressingGradientOptimizersviaCount-SketchesRyanSpring1AnastasiosKyrillidis1VijaiMohan2AnshumaliShrivastava12AbstractTraininglarge-scalemodelsefficientlyisachallengingtask.Therearenumerouspublica...
CompressingNeuralNetworksusingtheVariationalInformationBottleneckBinDai1ChenZhu2BainingGuo3DavidWipf3Abstractnetworkpruneddowntoasmalleroneofequivalentsizeoftenproduceshigheraccuracy(Linetal.,2017)...