MagMax: Leveraging Model Merging for Seamless Continual Learning

TL;DR

MagMax introduces a novel model merging technique for continual learning, allowing large pre-trained models to learn new tasks sequentially without forgetting, outperforming traditional methods in various scenarios.

Contribution

The paper presents MagMax, a new model-merging strategy that effectively enables continual learning in large pre-trained models, with extensive analysis of simple merging techniques.

Findings

Simple merging methods like weight averaging perform well.

MagMax outperforms traditional continual learning approaches.

Effective in class- and domain-incremental learning scenarios.

Abstract

This paper introduces a continual learning approach named MagMax, which utilizes model merging to enable large pre-trained models to continuously learn from new data without forgetting previously acquired knowledge. Distinct from traditional continual learning methods that aim to reduce forgetting during task training, MagMax combines sequential fine-tuning with a maximum magnitude weight selection for effective knowledge integration across tasks. Our initial contribution is an extensive examination of model merging techniques, revealing that simple approaches like weight averaging and random weight selection surprisingly hold up well in various continual learning contexts. More importantly, we present MagMax, a novel model-merging strategy that enables continual learning of large pre-trained models for successive tasks. Our thorough evaluation demonstrates the superiority of MagMax in various scenarios, including class- and domain-incremental learning settings. The code is available at this URL: https://github.com/danielm1405/magmax.