Economic Hyperparameter Optimization with Blended Search Strategy
Abstract
We study the problem of using low cost to search for hyperparameter configurations in a large search space with heterogeneous evaluation cost and model quality. We propose a blended search strategy to combine the strengths of global and local search, and prioritize them on the fly with the goal of minimizing the total cost spent in finding good configurations. Our approach demonstrates robust performance for tuning both tree-based models and deep neural networks on a large AutoML benchmark, as well as superior performance in model quality, time, and resource consumption for a production transformer-based NLP model fine-tuning task.
Cite
Text
Wang et al. "Economic Hyperparameter Optimization with Blended Search Strategy." International Conference on Learning Representations, 2021.Markdown
[Wang et al. "Economic Hyperparameter Optimization with Blended Search Strategy." International Conference on Learning Representations, 2021.](https://mlanthology.org/iclr/2021/wang2021iclr-economic/)BibTeX
@inproceedings{wang2021iclr-economic,
title = {{Economic Hyperparameter Optimization with Blended Search Strategy}},
author = {Wang, Chi and Wu, Qingyun and Huang, Silu and Saied, Amin},
booktitle = {International Conference on Learning Representations},
year = {2021},
url = {https://mlanthology.org/iclr/2021/wang2021iclr-economic/}
}