English

DEFT: Data Efficient Fine-Tuning for Pre-Trained Language Models via Unsupervised Core-Set Selection

Computation and Language 2024-06-14 v5 Artificial Intelligence

Abstract

Recent advances have led to the availability of many pre-trained language models (PLMs); however, a question that remains is how much data is truly needed to fine-tune PLMs for downstream tasks? In this work, we introduce DEFT-UCS, a data-efficient fine-tuning framework that leverages unsupervised core-set selection to identify a smaller, representative dataset that reduces the amount of data needed to fine-tune PLMs for downstream tasks. We examine the efficacy of DEFT-UCS in the context of text-editing LMs, and compare to the state-of-the art text-editing model, CoEDIT. Our results demonstrate that DEFT-UCS models are just as accurate as CoEDIT, across eight different datasets consisting of six different editing tasks, while finetuned on 70% less data.

Keywords

Cite

@article{arxiv.2310.16776,
  title  = {DEFT: Data Efficient Fine-Tuning for Pre-Trained Language Models via Unsupervised Core-Set Selection},
  author = {Devleena Das and Vivek Khetan},
  journal= {arXiv preprint arXiv:2310.16776},
  year   = {2024}
}
R2 v1 2026-06-28T13:01:48.678Z