Demystifying Instruction Mixing for Fine-tuning Large Language Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wang, Renxi, Li, Haonan, Wu, Minghao, Wang, Yuxia, Han, Xudong, Zhang, Chiyu, Baldwin, Timothy
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929247351734272
author Wang, Renxi
Li, Haonan
Wu, Minghao
Wang, Yuxia
Han, Xudong
Zhang, Chiyu
Baldwin, Timothy
author_facet Wang, Renxi
Li, Haonan
Wu, Minghao
Wang, Yuxia
Han, Xudong
Zhang, Chiyu
Baldwin, Timothy
contents Instruction tuning significantly enhances the performance of large language models (LLMs) across various tasks. However, the procedure to optimizing the mixing of instruction datasets for LLM fine-tuning is still poorly understood. This study categorizes instructions into three primary types: NLP downstream tasks, coding, and general chat. We explore the effects of instruction tuning on different combinations of datasets on LLM performance, and find that certain instruction types are more advantageous for specific applications but can negatively impact other areas. This work provides insights into instruction mixtures, laying the foundations for future research.
format Preprint
id arxiv_https___arxiv_org_abs_2312_10793
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Demystifying Instruction Mixing for Fine-tuning Large Language Models
Wang, Renxi
Li, Haonan
Wu, Minghao
Wang, Yuxia
Han, Xudong
Zhang, Chiyu
Baldwin, Timothy
Computation and Language
Artificial Intelligence
I.2.7
Instruction tuning significantly enhances the performance of large language models (LLMs) across various tasks. However, the procedure to optimizing the mixing of instruction datasets for LLM fine-tuning is still poorly understood. This study categorizes instructions into three primary types: NLP downstream tasks, coding, and general chat. We explore the effects of instruction tuning on different combinations of datasets on LLM performance, and find that certain instruction types are more advantageous for specific applications but can negatively impact other areas. This work provides insights into instruction mixtures, laying the foundations for future research.
title Demystifying Instruction Mixing for Fine-tuning Large Language Models
topic Computation and Language
Artificial Intelligence
I.2.7
url https://arxiv.org/abs/2312.10793