TE-NAS (Chen et al., 2021)

Reported on 6 benchmarks across 2 tasks · 1 paper

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Methodology6 results

Neural Architecture SearchonNATS-Bench Topology, CIFAR-10
Test Accuracy· 2021-02-23
93.9
best: 94.13 (RE (Real et al., 2019))
Neural Architecture Search on ImageNet in Four GPU Hours: A Theoretically Inspired Perspective arXiv:2102.11535
Neural Architecture SearchonNATS-Bench Topology, CIFAR-100
Test Accuracy· 2021-02-23
71.24
best: 71.68 (PPO (Schulman et al., 2017))
Neural Architecture Search on ImageNet in Four GPU Hours: A Theoretically Inspired Perspective arXiv:2102.11535
Neural Architecture SearchonNATS-Bench Topology, ImageNet16-120
Test Accuracy· 2021-02-23
42.38
best: 45.54 (EigenNas (Zhu et al., 2022))
Neural Architecture Search on ImageNet in Four GPU Hours: A Theoretically Inspired Perspective arXiv:2102.11535
AutoMLonNATS-Bench Topology, CIFAR-10
Test Accuracy· 2021-02-23
93.9
best: 94.13 (RE (Real et al., 2019))
Neural Architecture Search on ImageNet in Four GPU Hours: A Theoretically Inspired Perspective arXiv:2102.11535
AutoMLonNATS-Bench Topology, CIFAR-100
Test Accuracy· 2021-02-23
71.24
best: 71.68 (PPO (Schulman et al., 2017))
Neural Architecture Search on ImageNet in Four GPU Hours: A Theoretically Inspired Perspective arXiv:2102.11535
AutoMLonNATS-Bench Topology, ImageNet16-120
Test Accuracy· 2021-02-23
42.38
best: 45.54 (EigenNas (Zhu et al., 2022))
Neural Architecture Search on ImageNet in Four GPU Hours: A Theoretically Inspired Perspective arXiv:2102.11535