Salvato in:
Dettagli Bibliografici
Autori principali: Kallusky, Dominik, Rao, Vinay, Nandavanam, Vishal, Shi, Hao-Jun Michael
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:https://arxiv.org/abs/2510.15830
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866909852983361536
author Kallusky, Dominik
Rao, Vinay
Nandavanam, Vishal
Shi, Hao-Jun Michael
author_facet Kallusky, Dominik
Rao, Vinay
Nandavanam, Vishal
Shi, Hao-Jun Michael
contents The rapid development of large language models (LLMs) has driven the demand for more efficient optimization techniques. Among these, the Lookahead family of optimizers employs a two-loop framework, maintaining fast and slow sets of model weights. Multiple inner optimizer steps on the fast weights produce a trajectory - the pseudo-gradient - that is used to update the slow weights. DiLoCo, a notable example originally designed for distributed training, applies Nesterov momentum to the averaged pseudo-gradient from multiple workers, claiming to even outperform AdamW in a non-distributed setup. In this paper, we empirically show that DiLoCo's surprising effectiveness stems primarily from applying Nesterov momentum to the pseudo-gradient, which improves training in a non-distributed setting. We call this Lookahead variant the Step-$K$ Nesterov Outer Optimizer (SNOO). We demonstrate that SNOO achieves compute factor gains of 1.5 - 2.5$\times$ in a non-distributed setting up to a scale of 1e23 training FLOPs, with improvements that increase with model size. Because of its minimal compute and memory overhead and compatibility with model sharding, SNOO is a practical enhancement for a variety of inner optimizers, including AdamW and Muon.
format Preprint
id arxiv_https___arxiv_org_abs_2510_15830
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle SNOO: Step-K Nesterov Outer Optimizer - The Surprising Effectiveness of Nesterov Momentum Applied to Pseudo-Gradients
Kallusky, Dominik
Rao, Vinay
Nandavanam, Vishal
Shi, Hao-Jun Michael
Machine Learning
Artificial Intelligence
The rapid development of large language models (LLMs) has driven the demand for more efficient optimization techniques. Among these, the Lookahead family of optimizers employs a two-loop framework, maintaining fast and slow sets of model weights. Multiple inner optimizer steps on the fast weights produce a trajectory - the pseudo-gradient - that is used to update the slow weights. DiLoCo, a notable example originally designed for distributed training, applies Nesterov momentum to the averaged pseudo-gradient from multiple workers, claiming to even outperform AdamW in a non-distributed setup. In this paper, we empirically show that DiLoCo's surprising effectiveness stems primarily from applying Nesterov momentum to the pseudo-gradient, which improves training in a non-distributed setting. We call this Lookahead variant the Step-$K$ Nesterov Outer Optimizer (SNOO). We demonstrate that SNOO achieves compute factor gains of 1.5 - 2.5$\times$ in a non-distributed setting up to a scale of 1e23 training FLOPs, with improvements that increase with model size. Because of its minimal compute and memory overhead and compatibility with model sharding, SNOO is a practical enhancement for a variety of inner optimizers, including AdamW and Muon.
title SNOO: Step-K Nesterov Outer Optimizer - The Surprising Effectiveness of Nesterov Momentum Applied to Pseudo-Gradients
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2510.15830