A discounted Nesterov look-ahead update, implemented with NAdam, is claimed to converge sublinearly under fixed gradient delay and to beat synchronous pipeline baselines in language modeling.
Advances in asynchronous parallel and distributed optimization
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Nesterov Method for Asynchronous Pipeline Parallel Optimization
A discounted Nesterov look-ahead update, implemented with NAdam, is claimed to converge sublinearly under fixed gradient delay and to beat synchronous pipeline baselines in language modeling.