LogME:PracticalAssessmentofPre-TrainedModelsforTransferLearningKaichaoYou∗1YongLiu∗1JianminWang1MingshengLong1Abstracttrainedwithlarge-scalesuperviseddata(Dengetal.,2009;Russakovskyetal.,2015)and...
InstabilitiesofOfflineRLwithPre-TrainedNeuralRepresentationRuosongWang1YifanWu1RuslanSalakhutdinov1ShamM.Kakade23Abstract2018;Wangetal.,2018;Yuetal.,2019);itisseeingmuchrecentinterestduetothelargea...
ComposedFine-Tuning:FreezingPre-TrainedDenoisingAutoencodersforImprovedGeneralizationSangMichaelXie1TengyuMa1PercyLiang1AbstractPre-TraineddenoiserWefocusonpredictionproblemswithstructuredoutputsth...