HAT-CL: A Hard-Attention-to-the-Task PyTorch Library for Continual Learning

TL;DR

HAT-CL is a user-friendly PyTorch library that simplifies implementing the HAT mechanism for continual learning, improving usability, compatibility, and performance with novel mask techniques.

Contribution

We developed HAT-CL, a PyTorch-compatible toolkit that streamlines HAT integration and introduces new mask manipulation methods for better continual learning.

Findings

HAT-CL improves ease of use and integration in existing architectures.

Novel mask techniques enhance continual learning performance.

HAT-CL demonstrates consistent improvements across experiments.

Abstract

Catastrophic forgetting, the phenomenon in which a neural network loses previously obtained knowledge during the learning of new tasks, poses a significant challenge in continual learning. The Hard-Attention-to-the-Task (HAT) mechanism has shown potential in mitigating this problem, but its practical implementation has been complicated by issues of usability and compatibility, and a lack of support for existing network reuse. In this paper, we introduce HAT-CL, a user-friendly, PyTorch-compatible redesign of the HAT mechanism. HAT-CL not only automates gradient manipulation but also streamlines the transformation of PyTorch modules into HAT modules. It achieves this by providing a comprehensive suite of modules that can be seamlessly integrated into existing architectures. Additionally, HAT-CL offers ready-to-use HAT networks that are smoothly integrated with the TIMM library. Beyond the redesign and reimplementation of HAT, we also introduce novel mask manipulation techniques for HAT, which have consistently shown improvements across various experiments. Our work paves the way for a broader application of the HAT mechanism, opening up new possibilities in continual learning across diverse models and applications.