Introduction. Neural Architecture Search (NAS)1,2,3,4,5 is intended to automatically search for the neural network architecture. It doesn’t require much priori knowledge for designing a good neural network architecture compared to traditional manually design methods6,7,8. However, NAS demands relatively much computing time and resources, for example, the early NAS methods took months to discover a suitable network architecture for a particular task or dataset9,10,11,12,13 and required hund...