CGaLore：曲率引导的GaLore用于ASR基础模型的内存高效持续适应
CGaLore: Curvature-Guided GaLore for Memory-Efficient Continual Adaptation of ASR Foundation Models
浏览论文内容
中文总结 AI 辅助
针对ASR基础模型适应中的灾难性遗忘问题，提出曲率引导的GaLore（CGaLore），利用旧任务曲率信息选择低秩投影基，实现内存高效且缓解遗忘的持续适应，优于现有基线。
中文摘要 AI 辅助
自动语音识别模型在适应新领域、口音或下游任务时，会遭受灾难性遗忘。随着语音基础模型的使用日益增多，这一问题变得愈发重要，因为适应过程应既内存高效又安全，保留预训练期间获得的广泛能力。梯度低秩投影（GaLore）最近被提出作为一种内存高效的微调方法，它保持模型参数为全秩，同时通过低秩梯度投影减少优化器内存。然而，GaLore并未考虑灾难性遗忘。我们提出了曲率引导的GaLore（CGaLore），在选择低秩投影基时纳入旧任务曲率信息。具体而言，CGaLore在计算梯度子空间之前，使用来自先前任务的Kronecker因子近似曲率来过滤当前任务梯度。我们的实验表明，CGaLore能够实现有效适应同时缓解遗忘，优于最先进的持续学习基线。广泛的消融研究证实了CGaLore的实际适用性。
英文摘要
Automatic speech recognition models suffer from catastrophic forgetting when adapted to new domains, accents, or downstream tasks. This problem becomes increasingly important with the growing use of speech foundation models, where adaptation should be both memory-efficient and safe, preserving the broad capabilities learned during pretraining. Gradient Low-Rank Projection (GaLore) has recently been proposed as a memory-efficient fine-tuning method that keeps model parameters full-rank while reducing the optimizer memory through low-rank gradient projection. However, GaLore does not account for catastrophic forgetting. We propose Curvature-Guided GaLore (CGaLore), which incorporates old-task curvature information when selecting the low-rank projection bases. Specifically, CGaLore filters current-task gradients using Kronecker-factored approximate curvature from previous tasks before computing the gradient subspace. Our experiments show that CGaLore enables effective adaptation while alleviating forgetting, outperforming state-of-the-art continual learning baselines. Extensive ablation studies confirm the practical applicability of CGaLore.
发表机构
- KU Leuven（荷语鲁汶大学）
机构由 AI 辅助整理，请以论文原文为准。
