InstOptima: Evolutionary Multi-objective Instruction Optimization via Large Language Model-based Instruction Operators
This addresses the problem of low efficiency in instruction generation for researchers and practitioners in NLP, offering a novel method but with incremental improvements over existing automation techniques.
The paper tackles the inefficiency of instruction engineering in language models by proposing InstOptima, an evolutionary multi-objective optimization approach that uses LLM-based operators to generate instructions, resulting in improved fine-tuning performance and diverse, high-quality instructions.
Instruction-based language modeling has received significant attention in pretrained language models. However, the efficiency of instruction engineering remains low and hinders the development of instruction studies. Recent studies have focused on automating instruction generation, but they primarily aim to improve performance without considering other crucial objectives that impact instruction quality, such as instruction length and perplexity. Therefore, we propose a novel approach (i.e., InstOptima) that treats instruction generation as an evolutionary multi-objective optimization problem. In contrast to text edition-based methods, our approach utilizes a large language model (LLM) to simulate instruction operators, including mutation and crossover. Furthermore, we introduce an objective-guided mechanism for these operators, allowing the LLM to comprehend the objectives and enhance the quality of the generated instructions. Experimental results demonstrate improved fine-tuning performance and the generation of a diverse set of high-quality instructions.