Role Prompting Guided Domain Adaptation with General Capability Preserve for Large Language Models

TL;DR

This paper introduces REGA, a novel multi-domain adaptation strategy for large language models that preserves general capabilities while improving domain-specific performance by managing catastrophic forgetting and inter-domain confusion.

Contribution

REGA combines self-distillation, role prompting, and role integration to effectively adapt LLMs to multiple domains without sacrificing their general capabilities.

Findings

REGA reduces catastrophic forgetting in multi-domain LLM adaptation.

REGA improves domain-specific performance over standard fine-tuning.

REGA maintains strong general capabilities while adapting to specific domains.

Abstract

The growing interest in Large Language Models (LLMs) for specialized applications has revealed a significant challenge: when tailored to specific domains, LLMs tend to experience catastrophic forgetting, compromising their general capabilities and leading to a suboptimal user experience. Additionally, crafting a versatile model for multiple domains simultaneously often results in a decline in overall performance due to confusion between domains. In response to these issues, we present the RolE Prompting Guided Multi-Domain Adaptation (REGA) strategy. This novel approach effectively manages multi-domain LLM adaptation through three key components: 1) Self-Distillation constructs and replays general-domain exemplars to alleviate catastrophic forgetting. 2) Role Prompting assigns a central prompt to the general domain and a unique role prompt to each specific domain to minimize inter-domain confusion during training. 3) Role Integration reuses and integrates a small portion of domain-specific data to the general-domain data, which are trained under the guidance of the central prompt. The central prompt is used for a streamlined inference process, removing the necessity to switch prompts for different domains. Empirical results demonstrate that REGA effectively alleviates catastrophic forgetting and inter-domain confusion. This leads to improved domain-specific performance compared to standard fine-tuned models, while still preserving robust general capabilities.