logo Idiap Research Institute        
 [BibTeX] [Marc21]
Optimizer Benchmarking Needs to Account for Hyperparameter Tuning
Type of publication: Conference paper
Citation: Sivaprasad_ICML2020_2020
Publication status: Published
Booktitle: Proceedings of the 37th International Conference on Machine Learning
Year: 2020
Location: Vienna, Austria
Crossref: Sivaprasad_Idiap-RR-19-2019:
URL: https://icml.cc/Conferences/20...
Abstract: The performance of optimizers, particularly in deep learning, depends considerably on their chosen hyperparameter configuration. The efficacy of optimizers is often studied under near-optimal problem-specific hyperparameters, and finding these settings may be prohibitively costly for practitioners. In this work, we argue that a fair assessment of optimizers' performance must take the computational cost of hyperparameter tuning into account, i.e., how easy it is to find good hyperparameter configurations using an automatic hyperparameter search. Evaluating a variety of optimizers on an extensive set of standard datasets and architectures, our results indicate that Adam is the most practical solution, particularly in low-budget scenarios.
Keywords: Benchmarking, Hyperparameter optimization, optimization
Projects Idiap
Authors Sivaprasad, Prabhu Teja
Mai, Florian
Vogels, Thijs
Jaggi, Martin
Fleuret, Francois
Added by: [UNK]
Total mark: 0
Attachments
  • Sivaprasad_ICML2020_2020.pdf
       (Paper + supplementary)
Notes