Optimizing Millions of Hyperparameters by Implicit Differentiation
Abstract
We propose an algorithm for inexpensive gradient-based hyperparameter optimization that combines the implicit function theorem (IFT) with efficient inverse Hessian approximations. We present results about the relationship between the IFT and differentiating through optimization, motivating our algorithm. We use the proposed approach to train modern network architectures with millions of weights and millions of hyper-parameters. For example, we learn a data-augmentation network—where every weight is a hyperparameter tuned for validation performance—outputting augmented training examples. Jointly tuning weights and hyper-parameters is only a few times more costly in memory and compute than standard training.
Cite
Text
Lorraine et al. "Optimizing Millions of Hyperparameters by Implicit Differentiation." Artificial Intelligence and Statistics, 2020.Markdown
[Lorraine et al. "Optimizing Millions of Hyperparameters by Implicit Differentiation." Artificial Intelligence and Statistics, 2020.](https://mlanthology.org/aistats/2020/lorraine2020aistats-optimizing/)BibTeX
@inproceedings{lorraine2020aistats-optimizing,
title = {{Optimizing Millions of Hyperparameters by Implicit Differentiation}},
author = {Lorraine, Jonathan and Vicol, Paul and Duvenaud, David},
booktitle = {Artificial Intelligence and Statistics},
year = {2020},
pages = {1540-1552},
volume = {108},
url = {https://mlanthology.org/aistats/2020/lorraine2020aistats-optimizing/}
}