Generative Adversarial Neural Architecture Search

Authors: Seyed Saeed Changiz Rezaei, Fred X. Han, Di Niu, Mohammad Salameh, Keith Mills, Shuo Lian, Wei Lu, Shangling Jui

IJCAI 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments show that GA-NAS beats the best published results under several cases on three public NAS benchmarks. In the meantime, GA-NAS can handle ad-hoc search constraints and search spaces. We show that GA-NAS can be used to improve already optimized baselines found by other NAS methods, including Efficient Net and Proxyless NAS, in terms of Image Net accuracy or the number of parameters, in their original search space.
Researcher Affiliation Collaboration 1Huawei Technologies Canada Co., Ltd. 2Department of Electrical and Computer Engineering, University of Alberta 3Huawei Kirin Solution, Shanghai, China
Pseudocode Yes Algorithm 1 GA-NAS Algorithm; Algorithm 2 Training Discriminator D and the Generator G
Open Source Code No The paper does not provide an explicit statement or a link indicating the availability of its source code.
Open Datasets Yes To evaluate search algorithm and decouple it from the impact of search spaces, we query three NAS benchmarks: NAS-Bench-101 [Ying et al., 2019], NAS-Bench-201 [Dong and Yang, 2020], and NAS-Bench-301 [Siems et al., 2020]. ... We test GA-NAS on NAS-Bench-201 by conducting 20 runs for CIFAR-10, CIFAR-100, and Image Net-16-120 using the true test accuracy.
Dataset Splits Yes We use the supernet to evaluate the accuracy of a cell on a validation set of 10k instances of CIFAR10 (see Appendix).
Hardware Specification Yes the search cost for a run is 8 GPU hours on Ge Force GTX 1080 Ti. ... Total search time including supernet training is around 680 GPU hours on Tesla V100 GPUs. ... GA-NAS for 38 hours on 8 Tesla V100 GPUs,
Software Dependencies No The paper mentions various algorithms and model architectures like RNN, GNN, MLP, and training methods such as SGD and PPO, but does not provide specific software library names with version numbers (e.g., TensorFlow, PyTorch, CUDA versions).
Experiment Setup Yes In the first setup, we set |X0| = 50, |Xt| = |Xt 1| + 50, t ≥ 1, and k = 25, In the second setup, we set |X0| = 100, |Xt| = |Xt 1| + 100, t ≥ 1, and k = 50. For both setups, the initial set X0 is picked to be a random set, and the number of iterations T is 10.