Teaching According to Talents! Instruction Tuning LLMs with Competence-Aware Curriculum Learning

Fuente: arXiv
Guardado en:
Detalles Bibliográficos
Autores principales: Li, Yangning, Lu, Tingwei, Li, Yinghui, Chen, Yankai, Huang, Wei-Chieh, Jiang, Wenhao, Wang, Hui, Zheng, Hai-Tao, Yu, Philip S.
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866908624147709952
author Li, Yangning
Lu, Tingwei
Li, Yinghui
Chen, Yankai
Huang, Wei-Chieh
Jiang, Wenhao
Wang, Hui
Zheng, Hai-Tao
Yu, Philip S.
author_facet Li, Yangning
Lu, Tingwei
Li, Yinghui
Chen, Yankai
Huang, Wei-Chieh
Jiang, Wenhao
Wang, Hui
Zheng, Hai-Tao
Yu, Philip S.
contents Efficient instruction tuning aims to enhance the ultimate performance of large language models (LLMs) trained on a given instruction dataset. Curriculum learning as a typical data organization strategy has shown preliminary effectiveness in instruction tuning. However, current curriculum tuning methods suffer from the curriculum rigidity, since they rely solely on static heuristic difficulty metrics. These methods fail to adapt to the evolving capabilities of models during training, resulting in a fixed and potentially sub-optimal learning trajectory. To address the issue, Competence-Aware Multi-Perspective cUrriculum inStruction tuning framework termed CAMPUS is proposed. CAMPUS offers several advantages: (1) Dynamic selection for sub-curriculum. (2) Competency-aware adjustment to the curriculum schedule. (3) Multiple difficulty-based scheduling. Extensive experiments prove the superior performance of CAMPUS, compared to other state-of-the-art baselines for efficient instruction tuning.
format Preprint
id arxiv_https___arxiv_org_abs_2509_13790
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Teaching According to Talents! Instruction Tuning LLMs with Competence-Aware Curriculum Learning
Li, Yangning
Lu, Tingwei
Li, Yinghui
Chen, Yankai
Huang, Wei-Chieh
Jiang, Wenhao
Wang, Hui
Zheng, Hai-Tao
Yu, Philip S.
Computation and Language
Artificial Intelligence
Efficient instruction tuning aims to enhance the ultimate performance of large language models (LLMs) trained on a given instruction dataset. Curriculum learning as a typical data organization strategy has shown preliminary effectiveness in instruction tuning. However, current curriculum tuning methods suffer from the curriculum rigidity, since they rely solely on static heuristic difficulty metrics. These methods fail to adapt to the evolving capabilities of models during training, resulting in a fixed and potentially sub-optimal learning trajectory. To address the issue, Competence-Aware Multi-Perspective cUrriculum inStruction tuning framework termed CAMPUS is proposed. CAMPUS offers several advantages: (1) Dynamic selection for sub-curriculum. (2) Competency-aware adjustment to the curriculum schedule. (3) Multiple difficulty-based scheduling. Extensive experiments prove the superior performance of CAMPUS, compared to other state-of-the-art baselines for efficient instruction tuning.
title Teaching According to Talents! Instruction Tuning LLMs with Competence-Aware Curriculum Learning
topic Computation and Language
Artificial Intelligence
url https://arxiv.org/abs/2509.13790