arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2402.08594cs.CL
收藏

用于软提示调优的贝叶斯多任务迁移学习

Bayesian Multi-Task Transfer Learning for Soft Prompt Tuning

Kim Jaechul Graduate School of AI, KAIST(KAIST金在哲人工智能研究生院)

Haeju Lee, Minchan Jeong, Se-Young Yun, Kee-Eung Kim

AI总结:

本文提出一种贝叶斯多任务迁移学习方法,利用Stein变分梯度下降获取源任务提示的后验分布样本并聚合以初始化目标提示,有效建模源任务间相关性,在标准NLP基准上超越现有方法且无需辅助模型。

摘要:

提示调优(Prompt tuning)通过优化提示来使大规模预训练语言模型适应下游任务,而非微调全部模型参数,已被证明在多任务迁移学习环境中训练提示时特别有效。这些方法通常涉及为每个源任务单独训练提示,然后将其聚合以提供目标任务提示的初始化。然而,这种方法严重忽略了一个事实:某些源任务之间可能存在负面或正面的相互干扰。我们认为,在通过训练源提示从源任务中提取知识时,需要考虑源任务之间的这种相关性,以便更好地迁移到目标任务。为此,我们提出了一种贝叶斯方法,在其中我们处理跨源任务的提示的后验分布。我们利用Stein变分梯度下降(Stein Variational Gradient Descent)获得对应于后验样本的代表性源提示,然后将其聚合以构成初始目标提示。我们在标准基准NLP任务上展示了广泛的实验结果,其中我们的贝叶斯多任务迁移学习方法在许多环境中优于最先进的方法。此外,我们的方法除了提示本身外不需要辅助模型,实现了高度的参数效率。

英文摘要:

Prompt tuning, in which prompts are optimized to adapt large-scale pre-trained language models to downstream tasks instead of fine-tuning the full model parameters, has been shown to be particularly effective when the prompts are trained in a multi-task transfer learning setting. These methods generally involve individually training prompts for each source task and then aggregating them to provide the initialization of the prompt for the target task. However, this approach critically ignores the fact that some of the source tasks could be negatively or positively interfering with each other. We argue that when we extract knowledge from source tasks via training source prompts, we need to consider this correlation among source tasks for better transfer to target tasks. To this end, we propose a Bayesian approach where we work with the posterior distribution of prompts across source tasks. We obtain representative source prompts corresponding to the samples from the posterior utilizing Stein Variational Gradient Descent, which are then aggregated to constitute the initial target prompt. We show extensive experimental results on the standard benchmark NLP tasks, where our Bayesian multi-task transfer learning approach outperforms the state-of-the-art methods in many settings. Furthermore, our approach requires no auxiliary models other than the prompt itself, achieving a high degree of parameter efficiency.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑