MDIT: A Model-free Data Interpolation Method for Diverse Instruction Tuning
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866913793016070144 |
|---|---|
| author | Li, Yangning Lan, Zihua Qingsong, Lv Li, Yinghui Zheng, Hai-Tao |
| author_facet | Li, Yangning Lan, Zihua Qingsong, Lv Li, Yinghui Zheng, Hai-Tao |
| contents | As Large Language Models (LLMs) are increasingly applied across various tasks, instruction tuning has emerged as a critical method for enhancing model performance. However, current data management strategies face substantial challenges in generating diverse and comprehensive data, restricting further improvements in model performance. To address this gap, we propose MDIT, a novel model-free data interpolation method for diverse instruction tuning, which generates varied and high-quality instruction data by performing task interpolation. Moreover, it contains diversity-based clustering strategies to ensure the diversity of the training data. Extensive experiments show that our method achieves superior performance in multiple benchmark tasks. The LLMs finetuned with MDIT show significant improvements in numerous tasks such as general question answering, math reasoning, and code generation. MDIT offers an efficient and automatic data synthetic method, generating diverse instruction data without depending on external resources while expanding the application potential of LLMs in complex environments. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_07288 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | MDIT: A Model-free Data Interpolation Method for Diverse Instruction Tuning Li, Yangning Lan, Zihua Qingsong, Lv Li, Yinghui Zheng, Hai-Tao Computation and Language As Large Language Models (LLMs) are increasingly applied across various tasks, instruction tuning has emerged as a critical method for enhancing model performance. However, current data management strategies face substantial challenges in generating diverse and comprehensive data, restricting further improvements in model performance. To address this gap, we propose MDIT, a novel model-free data interpolation method for diverse instruction tuning, which generates varied and high-quality instruction data by performing task interpolation. Moreover, it contains diversity-based clustering strategies to ensure the diversity of the training data. Extensive experiments show that our method achieves superior performance in multiple benchmark tasks. The LLMs finetuned with MDIT show significant improvements in numerous tasks such as general question answering, math reasoning, and code generation. MDIT offers an efficient and automatic data synthetic method, generating diverse instruction data without depending on external resources while expanding the application potential of LLMs in complex environments. |
| title | MDIT: A Model-free Data Interpolation Method for Diverse Instruction Tuning |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2504.07288 |