research
∙
07/31/2023
Lookbehind Optimizer: k steps back, 1 step forward
The Lookahead optimizer improves the training stability of deep neural n...
share
research
∙
11/18/2022
