Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Yang, Yushi, Bean, Andrew M., McCraith, Robert, Mahdi, Adam
Format: Preprint
Veröffentlicht: 2024
Schlagworte:
Online-Zugang:https://arxiv.org/abs/2408.07888
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866929575997472768
author Yang, Yushi
Bean, Andrew M.
McCraith, Robert
Mahdi, Adam
author_facet Yang, Yushi
Bean, Andrew M.
McCraith, Robert
Mahdi, Adam
contents Fine-tuning Large Language Models (LLMs) incurs considerable training costs, driving the need for data-efficient training with optimised data ordering. Human-inspired strategies offer a solution by organising data based on human learning practices. This study evaluates the fine-tuning efficiency of five human-inspired strategies across four language models, three datasets, and both human- and LLM-labelled data in the context of medical question answering. These strategies achieve the best accuracy gain of 1.81% and an average gain of 1.02% across datasets, with interleaved strategies delivering the best average results. However, the best strategy varies across model-dataset combinations, limiting the generalisability of the effects of any single strategy. Additionally, LLM-defined question difficulty outperforms human-defined labels in curriculum-based learning, showing the potential of model-generated data as a cost-effective alternative for optimising fine-tuning.
format Preprint
id arxiv_https___arxiv_org_abs_2408_07888
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Evaluating Fine-Tuning Efficiency of Human-Inspired Learning Strategies in Medical Question Answering
Yang, Yushi
Bean, Andrew M.
McCraith, Robert
Mahdi, Adam
Computation and Language
Fine-tuning Large Language Models (LLMs) incurs considerable training costs, driving the need for data-efficient training with optimised data ordering. Human-inspired strategies offer a solution by organising data based on human learning practices. This study evaluates the fine-tuning efficiency of five human-inspired strategies across four language models, three datasets, and both human- and LLM-labelled data in the context of medical question answering. These strategies achieve the best accuracy gain of 1.81% and an average gain of 1.02% across datasets, with interleaved strategies delivering the best average results. However, the best strategy varies across model-dataset combinations, limiting the generalisability of the effects of any single strategy. Additionally, LLM-defined question difficulty outperforms human-defined labels in curriculum-based learning, showing the potential of model-generated data as a cost-effective alternative for optimising fine-tuning.
title Evaluating Fine-Tuning Efficiency of Human-Inspired Learning Strategies in Medical Question Answering
topic Computation and Language
url https://arxiv.org/abs/2408.07888