Papers › Evolving Neural Architecture Using One Shot Model

Evolving Neural Architecture Using One Shot Model

23 Dec 2020arXiv:2012.12540archive 2025-07-28

Nilotpal Sinha, Kuan-Wen Chen

Neural Architecture Search (NAS) is emerging as a new research direction which has the potential to replace the hand-crafted neural architectures designed for specific tasks. Previous evolution based architecture search requires high computational resources resulting in high search time. In this work, we propose a novel way of applying a simple genetic algorithm to the NAS problem called EvNAS (Evolving Neural Architecture using One Shot Model) which reduces the search time significantly while still achieving better result than previous evolution based methods. The architectures are represented by using the architecture parameter of the one shot model which results in the weight sharing among the architectures for a given population of architectures and also weight inheritance from one generation to the next generation of architectures. We propose a decoding technique for the architecture parameter which is used to divert majority of the gradient information towards the given architecture and is also used for improving the performance prediction of the given architecture from the one shot model during the search process. Furthermore, we use the accuracy of the partially trained architecture on the validation data as a prediction of its fitness in order to reduce the search time. EvNAS searches for the architecture on the proxy dataset i.e. CIFAR-10 for 4.4 GPU day on a single GPU and achieves top-1 test error of 2.47% with 3.63M parameters which is then transferred to CIFAR-100 and ImageNet achieving top-1 error of 16.37% and top-5 error of 7.4% respectively. All of these results show the potential of evolutionary methods in solving the architecture search problem.

PaperPDFCode

Code

nightstorm0909/EvNAS officialpytorch report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

Not run by Syntology. Nothing on this page verifies that the listed code works.

Tasks

Neural Architecture Searchmodel

1 archive task tag without a task page not shown.

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Neural Architecture Search ImageNet EvNAS-B Accuracy 75.6 #104 of 135 Archive leaderboard report
Neural Architecture Search ImageNet EvNAS-B MACs 599 #104 of 135 Archive leaderboard report
Neural Architecture Search ImageNet EvNAS-B Params 5.3 #104 of 135 Archive leaderboard report
Neural Architecture Search ImageNet EvNAS-B Top-1 Error Rate 24.4 #104 of 135 Archive leaderboard report
Neural Architecture Search ImageNet EvNAS-A Accuracy 75.6 #105 of 135 Archive leaderboard report
Neural Architecture Search ImageNet EvNAS-A MACs 570 #105 of 135 Archive leaderboard report
Neural Architecture Search ImageNet EvNAS-A Params 5.1 #105 of 135 Archive leaderboard report
Neural Architecture Search ImageNet EvNAS-A Top-1 Error Rate 24.4 #105 of 135 Archive leaderboard report
Neural Architecture Search ImageNet EvNAS-C Accuracy 74.9 #115 of 135 Archive leaderboard report
Neural Architecture Search ImageNet EvNAS-C MACs 547 #115 of 135 Archive leaderboard report
Neural Architecture Search ImageNet EvNAS-C Params 4.9 #115 of 135 Archive leaderboard report
Neural Architecture Search ImageNet EvNAS-C Top-1 Error Rate 25.1 #115 of 135 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections