English

How to Train BERT with an Academic Budget

Computation and Language 2021-09-10 v2 Artificial Intelligence Machine Learning

Abstract

While large language models a la BERT are used ubiquitously in NLP, pretraining them is considered a luxury that only a few well-funded industry labs can afford. How can one train such models with a more modest budget? We present a recipe for pretraining a masked language model in 24 hours using a single low-end deep learning server. We demonstrate that through a combination of software optimizations, design choices, and hyperparameter tuning, it is possible to produce models that are competitive with BERT-base on GLUE tasks at a fraction of the original pretraining cost.

Keywords

Cite

@article{arxiv.2104.07705,
  title  = {How to Train BERT with an Academic Budget},
  author = {Peter Izsak and Moshe Berchansky and Omer Levy},
  journal= {arXiv preprint arXiv:2104.07705},
  year   = {2021}
}
R2 v1 2026-06-24T01:13:02.860Z