Gradient-only line searches: An Alternative to Probabilistic Line Searches
This addresses the challenge of adaptive learning rate scheduling for practitioners in machine learning, offering an alternative to probabilistic line searches, though it appears incremental as it builds on existing line search concepts.
The paper tackles the problem of automatically determining learning rates in neural network training by introducing Gradient-Only Line Searches that are Inexact (GOLS-I), which resolves step sizes over 15 orders of magnitude without using surrogates, showing it as a competitive and easy-to-implement strategy.
Step sizes in neural network training are largely determined using predetermined rules such as fixed learning rates and learning rate schedules. These require user input or expensive global optimization strategies to determine their functional form and associated hyperparameters. Line searches are capable of adaptively resolving learning rate schedules. However, due to discontinuities induced by mini-batch sub-sampling, they have largely fallen out of favour. Notwithstanding, probabilistic line searches, which use statistical surrogates over a limited spatial domain, have recently demonstrated viability in resolving learning rates for stochastic loss functions. This paper introduces an alternative paradigm, Gradient-Only Line Searches that are Inexact (GOLS-I), as an alternative strategy to automatically determine learning rates in stochastic loss functions over a range of 15 orders of magnitude without the use of surrogates. We show that GOLS-I is a competitive strategy to reliably determine step sizes, adding high value in terms of performance, while being easy to implement.