784. Don’t be so Monotone: Relaxing Stochastic Line Search in Over-Parameterized Models
Invited abstract in session MB-32: Algorithmic Advances in Large Scale Nonconvex Optimization, stream Advances in large scale nonlinear optimization.
Monday, 10:30-12:00Room: 41 (building: 303A)
Authors (first author is the speaker)
| 1. | Leonardo Galli
|
| Mathematics, LMU Munich |
Abstract
Recent works have shown that line search methods can speed up Stochastic Gradient Descent (SGD) and Adam in modern over-parameterized settings. However,existing line searches may take steps that are smaller than necessary since they require a monotone decrease of the (mini-)batch objective function. We explore nonmonotone line search methods to relax this condition and possibly accept larger step sizes. Despite the lack of a monotonic decrease, we prove the same fast rates of convergence as in the monotone case. Our experiments show that nonmonotone methods improve the speed of convergence and generalization properties of SGD/Adam even beyond the previous monotone line searches. We propose a POlyak NOnmonotone Stochastic (PoNoS) method, obtained by combining a nonmonotone line search with a Polyak initial step size. Furthermore, we develop a new resetting technique that in the majority of the iterations reduces the amount of backtracks to zero while still maintaining a large initial step size. To the best of our knowledge, a first runtime comparison shows that the epoch-wise advantage of line-search-based methods gets reflected in the overall computational time.
Keywords
- Large Scale Optimization
- Stochastic Optimization
- Machine Learning
Status: accepted
Back to the list of papers