Q-Tuning: Queue-based Prompt Tuning for Lifelong Few-shot Language Learning

arXiv:2404.14607v1 Announce Type: new
Abstract: This paper introduces \textbf{Q-tuning}, a novel approach for continual prompt tuning that enables the lifelong learning of a pre-trained language model. When learning a new task, Q-tuning trains a task-specific prompt by adding it to a prompt queue consisting of the prompts from older tasks. To better transfer the knowledge of old tasks, we design an adaptive knowledge aggregation technique that reweighs previous prompts in the queue with a learnable low-rank matrix. Once the prompt …

abstract arxiv continual cs.cl few-shot language language model lifelong learning novel paper prompt prompts prompt tuning tasks trains type