Evidence > Intuition: Transferability Estimation for Encoder Selection

TL;DR

This paper introduces a method to predict the best pre-trained language model for a specific NLP task using transferability estimation, reducing the need for extensive fine-tuning and outperforming intuition-based selection.

Contribution

It applies the Logarithm of Maximum Evidence (LogME) measure from computer vision to NLP, demonstrating its effectiveness in ranking language models across multiple tasks.

Findings

LogME correlates with model performance in 94% of cases

Quantitative transferability measures outperform human intuition in model selection

Evidence-based ranking can identify unexpected suitable models

Abstract

With the increase in availability of large pre-trained language models (LMs) in Natural Language Processing (NLP), it becomes critical to assess their fit for a specific target task a priori - as fine-tuning the entire space of available LMs is computationally prohibitive and unsustainable. However, encoder transferability estimation has received little to no attention in NLP. In this paper, we propose to generate quantitative evidence to predict which LM, out of a pool of models, will perform best on a target task without having to fine-tune all candidates. We provide a comprehensive study on LM ranking for 10 NLP tasks spanning the two fundamental problem types of classification and structured prediction. We adopt the state-of-the-art Logarithm of Maximum Evidence (LogME) measure from Computer Vision (CV) and find that it positively correlates with final LM performance in 94% of the setups. In the first study of its kind, we further compare transferability measures with the de facto standard of human practitioner ranking, finding that evidence from quantitative metrics is more robust than pure intuition and can help identify unexpected LM candidates.