English

Fine-tuning Large Language Models for Adaptive Machine Translation

Computation and Language 2023-12-21 v1 Information Retrieval

Abstract

This paper presents the outcomes of fine-tuning Mistral 7B, a general-purpose large language model (LLM), for adaptive machine translation (MT). The fine-tuning process involves utilising a combination of zero-shot and one-shot translation prompts within the medical domain. The primary objective is to enhance real-time adaptive MT capabilities of Mistral 7B, enabling it to adapt translations to the required domain at inference time. The results, particularly for Spanish-to-English MT, showcase the efficacy of the fine-tuned model, demonstrating quality improvements in both zero-shot and one-shot translation scenarios, surpassing Mistral 7B's baseline performance. Notably, the fine-tuned Mistral outperforms ChatGPT "gpt-3.5-turbo" in zero-shot translation while achieving comparable one-shot translation quality. Moreover, the zero-shot translation of the fine-tuned Mistral matches NLLB 3.3B's performance, and its one-shot translation quality surpasses that of NLLB 3.3B. These findings emphasise the significance of fine-tuning efficient LLMs like Mistral 7B to yield high-quality zero-shot translations comparable to task-oriented models like NLLB 3.3B. Additionally, the adaptive gains achieved in one-shot translation are comparable to those of commercial LLMs such as ChatGPT. Our experiments demonstrate that, with a relatively small dataset of 20,000 segments that incorporate a mix of zero-shot and one-shot prompts, fine-tuning significantly enhances Mistral's in-context learning ability, especially for real-time adaptive MT.

Keywords

Cite

@article{arxiv.2312.12740,
  title  = {Fine-tuning Large Language Models for Adaptive Machine Translation},
  author = {Yasmin Moslem and Rejwanul Haque and Andy Way},
  journal= {arXiv preprint arXiv:2312.12740},
  year   = {2023}
}
R2 v1 2026-06-28T13:57:08.167Z