Benchmarking the tradeoff between neural network accuracy and training time is computationally expensive. Here we show how a multiplicative cyclic learning rate schedule can be used to construct a tradeoff curve in a single training run. We generate cyclic tradeoff curves for combinations of training methods such as Blurpool, Channels Last, Label Smoothing and MixUp, and highlight how these cyclic tradeoff curves can be used to evaluate the effects of algorithmic choices on network training efficiency.
@article{arxiv.2206.00832,
title = {Fast Benchmarking of Accuracy vs. Training Time with Cyclic Learning Rates},
author = {Jacob Portes and Davis Blalock and Cory Stephenson and Jonathan Frankle},
journal= {arXiv preprint arXiv:2206.00832},
year = {2022}
}
Comments
9 pages, 5 figures, "Has it Trained Yet?" Workshop at NeurIPS 2022