DS$^2$-Instruct: Domain-Specific Data Synthesis for Large Language Models Instruction Tuning

TL;DR

DS$^2$-Instruct is a zero-shot framework that synthesizes high-quality, domain-specific instruction datasets for LLMs, improving domain adaptation without human annotation by leveraging keyword generation, cognitive pairing, and self-validation.

Contribution

The paper introduces DS$^2$-Instruct, a novel zero-shot data synthesis method for domain-specific instruction tuning, addressing limitations of general-purpose data generation.

Findings

Models fine-tuned on DS$^2$-Instruct data outperform existing methods.

Effective across diverse domains like mathematics, finance, and reasoning.

Self-consistency validation enhances data quality.

Abstract

Adapting Large Language Models (LLMs) to specialized domains requires high-quality instruction tuning datasets, which are expensive to create through human annotation. Existing data synthesis methods focus on general-purpose tasks and fail to capture domain-specific terminology and reasoning patterns. To address this, we introduce DS$^2$-Instruct, a zero-shot framework that generates domain-specific instruction datasets without human supervision. Our approach first generates task-informed keywords to ensure comprehensive domain coverage. It then creates diverse instructions by pairing these keywords with different cognitive levels from Bloom's Taxonomy. Finally, it uses self-consistency validation to ensure data quality. We apply this framework to generate datasets across seven challenging domains, such as mathematics, finance, and logical reasoning. Comprehensive evaluation demonstrates that models fine-tuned on our generated data achieve substantial improvements over existing data generation methods.