RANK-NOSH: Efficient Predictor-Based Architecture Search via
            Non-Uniform Successive Halving
    Ruochen Wang1 , Xiangning Chen1 , Minhao Cheng1 , Xiaocheng Tang2 , Cho-Jui Hsieh1
           1
            Department of Computer Science, UCLA, 2 DiDi AI Labs

             Abstract                  cess in improving the search efficiency of NAS [2, 7, 23,
                                  29, 44]. Those methods train a supernet that encompasses
  Predictor-based algorithms have achieved ...