AIDE: Attribute-Guided MultI-Hop Data Expansion for Data Scarcity in Task-Specific Fine-tuning
This addresses the problem of insufficient training data for researchers and practitioners fine-tuning LLMs, offering a novel solution that is not incremental but demonstrates strong gains.
The paper tackles the challenge of data scarcity for fine-tuning large language models on specific tasks by proposing AIDE, a multi-hop data synthesis framework that expands from very few seed data points, achieving over 30% improvement over state-of-the-art methods and enabling fine-tuning from just 10 seeds.
Fine-tuning large language models (LLMs) for specific tasks requires diverse, high-quality training data. However, obtaining sufficient relevant data remains a significant challenge. Existing data synthesis methods either depend on extensive seed datasets or struggle to balance task relevance and data diversity. To address these challenges, we propose Attribute-guided multI-hop Data Expansion (AIDE), a novel data synthesis framework that uses a multi-hop process to expand very few seed data points while ensuring data diversity and task relevance. AIDE extracts the main topic and key knowledge attributes from the seeds to guide the synthesis steps. The process repeats for K hops, using the generated data as seeds. To prevent irrelevant data generation as the hop depth increases, AIDE incorporates a residual connection mechanism. Our empirical results show that AIDE enables fine-tuning of Mistral-7B, Llama-3.1-8B and Llama-3.2-3B from 10 seeds, surpassing the models fine-tuned on human curated data. Furthermore, AIDE outperforms state-of-the-art data synthesis methods, such as Evol-Instruct, by over 30% in task-specific fine-tuning. Code is available at https://github.com/Code4Graph/AIDE.