no code implementations • 23 May 2021 • Younghwan Chae, Daniel N. Wilke, Dominic Kafka
The results show that training a model with the recommended learning rate for a class of search directions helps to reduce the model errors in multimodal cases.
1 code implementation • 29 Jun 2020 • Dominic Kafka, Daniel Nicolas Wilke
Gradient-only and probabilistic line searches have recently reintroduced the ability to adaptively determine learning rates in dynamic mini-batch sub-sampled neural network training.
no code implementations • 15 Jan 2020 • Dominic Kafka, Daniel. N. Wilke
This study proposes gradient-only line searches to resolve the learning rate for neural network training algorithms.
no code implementations • 22 Mar 2019 • Dominic Kafka, Daniel Wilke
Line searches are capable of adaptively resolving learning rate schedules.
no code implementations • 20 Mar 2019 • Dominic Kafka, Daniel Wilke
Mini-batch sub-sampling in neural network training is unavoidable, due to growing data demands, memory-limited computational resources such as graphical processing units (GPUs), and the dynamics of on-line learning.