Tuning Less, Prompting More: A Cost-Effective and Flexible Training and Inference Pipeline for Natural Language Transformation

ACL ARR 2025 February Submission976 Authors

Abstract: Natural language transformation (NLT) tasks, such as machine translation (MT) and text style transfer (TST), require models to generate accurate and contextually appropriate outputs. However, existing approaches face significant challenges, including the computational costs of leveraging large pre-trained models and the limited generalization ability of fine-tuned smaller models. In this paper, we propose a novel framework that combines the flexibility of prompting with the cost-effectiveness of fine-tuning. Our method enhances small models by integrating In-Context Examples (ICE) retrieved from training data, enabling the model to better capture contextual information and align with user preferences. We further improve performance through hierarchical contrastive learning and dynamic preference inference mechanisms. Experimental results demonstrate that our approach outperforms existing methods, such as Supervised Instruction Tuning (SIT), Direct Preference Optimization (DPO), and Contrastive Preference Optimization (CPO), across both MT and TST tasks, providing a more efficient solution for resource-constrained environments.

Paper Type: Long

Research Area: Machine Translation

Research Area Keywords: Machine Translation, Text Style Transfer, Preference Alignment, Large Language Model,

Contribution Types: Approaches low compute settings-efficiency

Languages Studied: Chinese, English, German, Russian, Czech, Indonesian

Submission Number: 976