POUF: Prompt-oriented unsupervised fine-tuning for large pre-trained models
POUF: Prompt-oriented unsupervised fine-tuning for large pre-trained models
复制标题
DOI：
10.48550/arxiv.2305.00350
复制
发表时间：
2023-04
期刊：
影响因子：
--
通讯作者：
Korawat Tanwisuth;Shujian Zhang;Huangjie Zheng;Pengcheng He;Mingyuan Zhou
中科院分区：
文献类型：
--
作者：
Korawat Tanwisuth;Shujian Zhang;Huangjie Zheng;Pengcheng He;Mingyuan Zhou
通过提示，大规模的预训练模型变得更具表达力和强大，近年来获得了极大的关注。尽管这些大型模型具有零触发能力，但通常仍需要标记数据来使其适应下游任务。为了克服这一关键限制，我们提出了一个无监督的微调框架来直接微调模型或提示未标记的目标数据。我们演示了如何将我们的方法应用于语言增强视觉和掩蔽语言模型，通过对齐从提示和目标数据中提取的离散分布。为了验证我们的方法的适用性，我们进行了广泛的实验图像分类，情感分析和自然语言推理任务。在13个与图像相关的任务和15个与语言相关的任务中，所提出的方法在基线上实现了一致的改进。
Through prompting, large-scale pre-trained models have become more expressive and powerful, gaining significant attention in recent years. Though these big models have zero-shot capabilities, in general, labeled data are still required to adapt them to downstream tasks. To overcome this critical limitation, we propose an unsupervised fine-tuning framework to directly fine-tune the model or prompt on the unlabeled target data. We demonstrate how to apply our method to both language-augmented vision and masked-language models by aligning the discrete distributions extracted from the prompts and target data. To verify our approach's applicability, we conduct extensive experiments on image classification, sentiment analysis, and natural language inference tasks. Across 13 image-related tasks and 15 language-related ones, the proposed approach achieves consistent improvements over the baselines.