RankNAS: Efficient Neural Architecture Search by Pairwise Ranking
This addresses the problem of slow and resource-intensive architecture search for researchers and practitioners in machine learning, though it appears incremental as it builds on existing NAS methods.
The paper tackles the efficiency challenge in Neural Architecture Search (NAS) by formulating it as a ranking problem, resulting in RankNAS, which designs high-performance architectures while being orders of magnitude faster than state-of-the-art NAS systems.
This paper addresses the efficiency challenge of Neural Architecture Search (NAS) by formulating the task as a ranking problem. Previous methods require numerous training examples to estimate the accurate performance of architectures, although the actual goal is to find the distinction between "good" and "bad" candidates. Here we do not resort to performance predictors. Instead, we propose a performance ranking method (RankNAS) via pairwise ranking. It enables efficient architecture search using much fewer training examples. Moreover, we develop an architecture selection method to prune the search space and concentrate on more promising candidates. Extensive experiments on machine translation and language modeling tasks show that RankNAS can design high-performance architectures while being orders of magnitude faster than state-of-the-art NAS systems.