English

Measuring the Instability of Fine-Tuning

Computation and Language 2023-10-03 v2

Abstract

Fine-tuning pre-trained language models on downstream tasks with varying random seeds has been shown to be unstable, especially on small datasets. Many previous studies have investigated this instability and proposed methods to mitigate it. However, most studies only used the standard deviation of performance scores (SD) as their measure, which is a narrow characterization of instability. In this paper, we analyze SD and six other measures quantifying instability at different levels of granularity. Moreover, we propose a systematic framework to evaluate the validity of these measures. Finally, we analyze the consistency and difference between different measures by reassessing existing instability mitigation methods. We hope our results will inform the development of better measurements of fine-tuning instability.

Keywords

Cite

@article{arxiv.2302.07778,
  title  = {Measuring the Instability of Fine-Tuning},
  author = {Yupei Du and Dong Nguyen},
  journal= {arXiv preprint arXiv:2302.07778},
  year   = {2023}
}

Comments

20 pages, 26 Figures, accepted to ACL 2023 main conference

R2 v1 2026-06-28T08:40:55.251Z