Building Domain-Specific Small Language Models via Guided Data Generation
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909927953399808 |
|---|---|
| author | Kumar, Aman Amin, Ekant Muljibhai Lee, Xian Yeow Vidyaratne, Lasitha Farahat, Ahmed K. Ghosh, Dipanjan D. Koreeda, Yuta Gupta, Chetan |
| author_facet | Kumar, Aman Amin, Ekant Muljibhai Lee, Xian Yeow Vidyaratne, Lasitha Farahat, Ahmed K. Ghosh, Dipanjan D. Koreeda, Yuta Gupta, Chetan |
| contents | Large Language Models (LLMs) have shown remarkable success in supporting a wide range of knowledge-intensive tasks. In specialized domains, there is growing interest in leveraging LLMs to assist subject matter experts with domain-specific challenges. However, deploying LLMs as SaaS solutions raises data privacy concerns, while many open-source models demand significant computational resources for effective domain adaptation and deployment. A promising alternative is to develop smaller, domain-specialized LLMs, though this approach is often constrained by the lack of high-quality domain-specific training data. In this work, we address these limitations by presenting a cost-efficient and scalable training pipeline that combines guided synthetic data generation from a small seed corpus with bottom-up domain data curation. Our pipeline integrates Domain-Adaptive Pretraining (DAPT), Domain-specific Supervised Fine-tuning (DSFT), and Direct Preference Optimization (DPO) to train effective small-scale models for specialized use cases. We demonstrate this approach through DiagnosticSLM, a 3B-parameter domain-specific model tailored for fault diagnosis, root cause analysis, and repair recommendation in industrial settings. To evaluate model performance, we introduce four domain-specific benchmarks: multiple-choice questions (DiagnosticMCQ), question answering (DiagnosticQA), sentence completion (DiagnosticComp), and summarization (DiagnosticSum). DiagnosticSLM achieves up to 25% accuracy improvement over open-source models of comparable or larger size (2B-9B) on the MCQ task, while also outperforming or matching them in other tasks, demonstrating effective domain-specific reasoning and generalization capabilities. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_21748 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Building Domain-Specific Small Language Models via Guided Data Generation Kumar, Aman Amin, Ekant Muljibhai Lee, Xian Yeow Vidyaratne, Lasitha Farahat, Ahmed K. Ghosh, Dipanjan D. Koreeda, Yuta Gupta, Chetan Computation and Language Artificial Intelligence Large Language Models (LLMs) have shown remarkable success in supporting a wide range of knowledge-intensive tasks. In specialized domains, there is growing interest in leveraging LLMs to assist subject matter experts with domain-specific challenges. However, deploying LLMs as SaaS solutions raises data privacy concerns, while many open-source models demand significant computational resources for effective domain adaptation and deployment. A promising alternative is to develop smaller, domain-specialized LLMs, though this approach is often constrained by the lack of high-quality domain-specific training data. In this work, we address these limitations by presenting a cost-efficient and scalable training pipeline that combines guided synthetic data generation from a small seed corpus with bottom-up domain data curation. Our pipeline integrates Domain-Adaptive Pretraining (DAPT), Domain-specific Supervised Fine-tuning (DSFT), and Direct Preference Optimization (DPO) to train effective small-scale models for specialized use cases. We demonstrate this approach through DiagnosticSLM, a 3B-parameter domain-specific model tailored for fault diagnosis, root cause analysis, and repair recommendation in industrial settings. To evaluate model performance, we introduce four domain-specific benchmarks: multiple-choice questions (DiagnosticMCQ), question answering (DiagnosticQA), sentence completion (DiagnosticComp), and summarization (DiagnosticSum). DiagnosticSLM achieves up to 25% accuracy improvement over open-source models of comparable or larger size (2B-9B) on the MCQ task, while also outperforming or matching them in other tasks, demonstrating effective domain-specific reasoning and generalization capabilities. |
| title | Building Domain-Specific Small Language Models via Guided Data Generation |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2511.21748 |