Catastrophic forgetting has remained a critical challenge for deep neural networks in Continual Learning (CL) as it undermines consolidated knowledge when learning new tasks. Parameter efficient fine tuning CL techniques are gaining traction for their effectiveness in addressing catastrophic forgetting with a lightweight training schedule while avoiding degradation of consolidated knowledge in pre-trained models. However, low rank adapters (LoRA) in these approaches are highly sensitive to rank selection which can lead to sub-optimal resource allocation and performance. To this end, we introduce PEARL, a rehearsal-free CL framework that entails dynamic rank allocation for LoRA components during CL training. Specifically, PEARL leverages reference task weights and adaptively determines the rank of task-specific LoRA components based on the current tasks' proximity to reference task weights in parameter space. To demonstrate the versatility of PEARL, we evaluate it across three vision architectures (ResNet, Separable Convolutional Network and Vision Transformer) and a multitude of CL scenarios, and show that PEARL outperforms all considered baselines by a large margin.

翻译：灾难性遗忘一直是深度神经网络在持续学习（CL）中面临的关键挑战，因为它会在学习新任务时破坏已巩固的知识。参数高效微调CL技术因其能以轻量级训练方案有效解决灾难性遗忘问题，同时避免预训练模型中已巩固知识的退化，正日益受到关注。然而，这些方法中的低秩适配器（LoRA）对秩的选择高度敏感，可能导致次优的资源分配和性能表现。为此，我们提出了PEARL——一种无需回放的CL框架，该框架在CL训练过程中为LoRA组件实现动态秩分配。具体而言，PEARL利用参考任务权重，根据当前任务在参数空间中与参考任务权重的接近程度，自适应地确定任务特定LoRA组件的秩。为展示PEARL的通用性，我们在三种视觉架构（ResNet、可分离卷积网络和Vision Transformer）及多种CL场景下进行评估，结果表明PEARL以显著优势超越所有对比基线方法。