TrainLarge,ThenCompress:RethinkingModelSizeforEfficientTrainingandInferenceofTransformersZhuohanLi1EricWallace1ShengShen1KevinLin1KurtKeutzer1DanKlein1JosephE.Gonzalez1AbstractCommonTrainSmallStopT...
AutoGAN-Distiller:SearchingtoCompressGenerativeAdversarialNetworksYongganFu1WuyangChen2HaotaoWang2HaoranLi1YingyanLin1ZhangyangWang2Abstractandenhancement.Thoseapplicationshavedriventhegrow-ingdema...
Progress&Compress:AscalableframeworkforcontinuallearningJonathanSchwarz1JelenaLuketina2WojciechM.Czarnecki1AgnieszkaGrabska-Barwinska1YeeWhyeTeh1RazvanPascanu1RaiaHadsell1Abstractpreviouslytrainedp...