Scaling Optimal LR Across Token Horizons

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Bjorck, Johan, Benhaim, Alon, Chaudhary, Vishrav, Wei, Furu, Song, Xia
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866909507275194368
author Bjorck, Johan
Benhaim, Alon
Chaudhary, Vishrav
Wei, Furu
Song, Xia
author_facet Bjorck, Johan
Benhaim, Alon
Chaudhary, Vishrav
Wei, Furu
Song, Xia
contents State-of-the-art LLMs are powered by scaling -- scaling model size, dataset size and cluster size. It is economically infeasible to extensively tune hyperparameter for the largest runs. Instead, approximately optimal hyperparameters must be inferred or \textit{transferred} from smaller experiments. Hyperparameter transfer across model sizes has been studied in Yang et al. However, hyperparameter transfer across dataset size -- or token horizon -- has not been studied yet. To remedy this we conduct a large scale empirical study on how optimal learning rate (LR) depends on token horizon in LLM training. We first demonstrate that the optimal LR changes significantly with token horizon -- longer training necessitates smaller LR. Secondly we demonstrate the the optimal LR follows a scaling law, and that the optimal LR for longer horizons can be accurately estimated from shorter horizons via such scaling laws. We also provide a rule-of-thumb for transferring LR across token horizons with zero overhead over current practices. Lastly we provide evidence that LLama-1 used too high LR, and estimate the performance hit from this. We thus argue that hyperparameter transfer across data size is an important and overlooked component of LLM training.
format Preprint
id arxiv_https___arxiv_org_abs_2409_19913
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Scaling Optimal LR Across Token Horizons
Bjorck, Johan
Benhaim, Alon
Chaudhary, Vishrav
Wei, Furu
Song, Xia
Machine Learning
Artificial Intelligence
Computation and Language
State-of-the-art LLMs are powered by scaling -- scaling model size, dataset size and cluster size. It is economically infeasible to extensively tune hyperparameter for the largest runs. Instead, approximately optimal hyperparameters must be inferred or \textit{transferred} from smaller experiments. Hyperparameter transfer across model sizes has been studied in Yang et al. However, hyperparameter transfer across dataset size -- or token horizon -- has not been studied yet. To remedy this we conduct a large scale empirical study on how optimal learning rate (LR) depends on token horizon in LLM training. We first demonstrate that the optimal LR changes significantly with token horizon -- longer training necessitates smaller LR. Secondly we demonstrate the the optimal LR follows a scaling law, and that the optimal LR for longer horizons can be accurately estimated from shorter horizons via such scaling laws. We also provide a rule-of-thumb for transferring LR across token horizons with zero overhead over current practices. Lastly we provide evidence that LLama-1 used too high LR, and estimate the performance hit from this. We thus argue that hyperparameter transfer across data size is an important and overlooked component of LLM training.
title Scaling Optimal LR Across Token Horizons
topic Machine Learning
Artificial Intelligence
Computation and Language
url https://arxiv.org/abs/2409.19913