Abstract
We analyze the performance of large language models (LLMs) on Text Style Transfer (TST), specifically focusing on sentiment transfer and text detoxification across three languages: English, Hindi, and Bengali. Text Style Transfer involves modifying the linguistic style of a text while preserving its core content. We evaluate the capabilities of pre-trained LLMs using zero-shot and few-shot prompting as well as parameter-efficient finetuning on publicly available datasets. Our evaluation using automatic metrics, GPT-4 and human evaluations reveals that while some prompted LLMs perform well in English, their performance in on other languages (Hindi, Bengali) remains average. However, finetuning significantly improves results compared to zero-shot and few-shot prompting, making them comparable to previous state-of-the-art. This underscores the necessity of dedicated datasets and specialized models for effective TST.- Anthology ID:
- 2024.inlg-main.42
- Volume:
- Proceedings of the 17th International Natural Language Generation Conference
- Month:
- September
- Year:
- 2024
- Address:
- Tokyo, Japan
- Editors:
- Saad Mahamood, Nguyen Le Minh, Daphne Ippolito
- Venue:
- INLG
- SIG:
- SIGGEN
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 523–539
- Language:
- URL:
- https://preview.aclanthology.org/add_missing_videos/2024.inlg-main.42/
- DOI:
- Cite (ACL):
- Sourabrata Mukherjee, Atul Kr. Ojha, and Ondrej Dusek. 2024. Are Large Language Models Actually Good at Text Style Transfer?. In Proceedings of the 17th International Natural Language Generation Conference, pages 523–539, Tokyo, Japan. Association for Computational Linguistics.
- Cite (Informal):
- Are Large Language Models Actually Good at Text Style Transfer? (Mukherjee et al., INLG 2024)
- PDF:
- https://preview.aclanthology.org/add_missing_videos/2024.inlg-main.42.pdf