A disciplined approach to neural network hyper-parameters: Part 1 -- learning rate, batch size, momentum, and weight decay

TL;DR

This paper presents practical methods for tuning key neural network hyper-parameters like learning rate, batch size, momentum, and weight decay to reduce training time and improve model performance.

Contribution

It introduces guidelines for analyzing training loss and adjusting hyper-parameters systematically, emphasizing the importance of balancing regularization with other parameters.

Findings

Proper hyper-parameter tuning reduces training time.

Balancing regularization improves model performance.

Optimal weight decay is linked to learning rate and momentum.

Abstract

Although deep learning has produced dazzling successes for applications of image, speech, and video processing in the past few years, most trainings are with suboptimal hyper-parameters, requiring unnecessarily long training times. Setting the hyper-parameters remains a black art that requires years of experience to acquire. This report proposes several efficient ways to set the hyper-parameters that significantly reduce training time and improves performance. Specifically, this report shows how to examine the training validation/test loss function for subtle clues of underfitting and overfitting and suggests guidelines for moving toward the optimal balance point. Then it discusses how to increase/decrease the learning rate/momentum to speed up training. Our experiments show that it is crucial to balance every manner of regularization for each dataset and architecture. Weight decay is used as a sample regularizer to show how its optimal value is tightly coupled with the learning rates and momentums. Files to help replicate the results reported here are available.