Quantized Gradient Projection for Memory-Efficient Continual Learning

Dongjun Kim, Seohyeon Cha, Huancheng Chen, Chianing Wang, Haris Vikalo

Keywords: Continual Learning

TL;DR: We propose QGPM, a memory-efficient and privacy-preserving continual learning framework that compresses task subspaces via quantization.

Abstract: Real-world deployment of machine learning models requires the ability to continually learn from non-stationary data while preserving prior knowledge and user privacy. Therefore, storing knowledge acquired from past data in a resource- and privacy-friendly manner is a crucial consideration in determining their viability. We introduce Quantized Gradient Projection Memory (QGPM), a systematic framework for continual learning that compresses and preserves the previous gradient subspace. QGPM integrates three key components: (i) distribution-aware, basis-wise quantization to minimize storage overhead, (ii) a Quantization Error-Aware (QEA) gradient projection that selectively relaxes orthogonality to mitigate gradient drift caused by accumulated quantization noise, and (iii) an on-the-fly sparse sketching strategy that improves runtime memory and computational efficiency. Experiments across multiple benchmarks demonstrate that QGPM achieves state-of-the-art performance under fixed memory budgets, highlighting its effectiveness in scalable, privacy-preserving continual learning.

Primary Area: transfer learning, meta learning, and lifelong learning

Submission Number: 14203