Saved in:
Bibliographic Details
Main Authors: Hu, Zhengyu, Zhang, Jieyu, Yu, Yue, Zhuang, Yuchen, Xiong, Hui
Format: Preprint
Published: 2023
Subjects:
Online Access:https://arxiv.org/abs/2312.01619
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913236113162240
author Hu, Zhengyu
Zhang, Jieyu
Yu, Yue
Zhuang, Yuchen
Xiong, Hui
author_facet Hu, Zhengyu
Zhang, Jieyu
Yu, Yue
Zhuang, Yuchen
Xiong, Hui
contents This paper presents LEMR (Label-Efficient Model Ranking) and introduces the MoraBench Benchmark. LEMR is a novel framework that minimizes the need for costly annotations in model selection by strategically annotating instances from an unlabeled validation set. To evaluate LEMR, we leverage the MoraBench Benchmark, a comprehensive collection of model outputs across diverse scenarios. Our extensive evaluation across 23 different NLP tasks in semi-supervised learning, weak supervision, and prompt selection tasks demonstrates LEMR's effectiveness in significantly reducing labeling costs. Key findings highlight the impact of suitable ensemble methods, uncertainty sampling strategies, and model committee selection in enhancing model ranking accuracy. LEMR, supported by the insights from MoraBench, provides a cost-effective and accurate solution for model selection, especially valuable in resource-constrained environments.
format Preprint
id arxiv_https___arxiv_org_abs_2312_01619
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle How Many Validation Labels Do You Need? Exploring the Design Space of Label-Efficient Model Ranking
Hu, Zhengyu
Zhang, Jieyu
Yu, Yue
Zhuang, Yuchen
Xiong, Hui
Machine Learning
This paper presents LEMR (Label-Efficient Model Ranking) and introduces the MoraBench Benchmark. LEMR is a novel framework that minimizes the need for costly annotations in model selection by strategically annotating instances from an unlabeled validation set. To evaluate LEMR, we leverage the MoraBench Benchmark, a comprehensive collection of model outputs across diverse scenarios. Our extensive evaluation across 23 different NLP tasks in semi-supervised learning, weak supervision, and prompt selection tasks demonstrates LEMR's effectiveness in significantly reducing labeling costs. Key findings highlight the impact of suitable ensemble methods, uncertainty sampling strategies, and model committee selection in enhancing model ranking accuracy. LEMR, supported by the insights from MoraBench, provides a cost-effective and accurate solution for model selection, especially valuable in resource-constrained environments.
title How Many Validation Labels Do You Need? Exploring the Design Space of Label-Efficient Model Ranking
topic Machine Learning
url https://arxiv.org/abs/2312.01619