CARS: Continuous Evolution for Efficient Neural Architecture Search
release_r53hs6hgjfct3gmglq4j4l66vy
by
Zhaohui Yang, Yunhe Wang, Xinghao Chen, Boxin Shi, Chao Xu, Chunjing
Xu, Qi Tian, Chang Xu
2020
Abstract
Searching techniques in most of existing neural architecture search (NAS)
algorithms are mainly dominated by differentiable methods for the efficiency
reason. In contrast, we develop an efficient continuous evolutionary approach
for searching neural networks. Architectures in the population that share
parameters within one SuperNet in the latest generation will be tuned over the
training dataset with a few epochs. The searching in the next evolution
generation will directly inherit both the SuperNet and the population, which
accelerates the optimal network generation. The non-dominated sorting strategy
is further applied to preserve only results on the Pareto front for accurately
updating the SuperNet. Several neural networks with different model sizes and
performances will be produced after the continuous search with only 0.4 GPU
days. As a result, our framework provides a series of networks with the number
of parameters ranging from 3.7M to 5.1M under mobile settings. These networks
surpass those produced by the state-of-the-art methods on the benchmark
ImageNet dataset.
In text/plain
format
Archived Files and Locations
application/pdf 549.7 kB
file_h7oqu22fgvggzoelu6gmhqxoti
|
arxiv.org (repository) web.archive.org (webarchive) |
1909.04977v6
access all versions, variants, and formats of this works (eg, pre-prints)