FairNAS: Rethinking Evaluation Fairness of Weight Sharing Neural Architecture Search

One of the most critical problems in two-stage weight-sharing neural architecture search is the evaluation of candidate models. A faithful ranking certainly leads to accurate searching results. However, current methods are prone to making misjudgments. In this paper, we prove that they inevitably give biased evaluations due to inherent unfairness in the supernet training. In view of this, we propose two levels of constraints: expectation fairness and strict fairness. Particularly, strict fairness ensures equal optimization opportunities for all choice blocks throughout the training, which neither overestimates nor underestimates their capacity. We demonstrate this is crucial to improving confidence in models' ranking. Incorporating our supernet trained under fairness constraints with a multi-objective evolutionary search algorithm, we obtain various state-of-the-art models on ImageNet. Especially, FairNAS-A attains 77.5% top-1 accuracy. The models and their evaluation codes are made publicly available online http://github.com/fairnas/FairNAS .