Regularizing Trajectories to Mitigate Catastrophic Forgetting
Regularization-based continual learning approaches generally prevent catastrophic forgetting by augmenting the training loss with an auxiliary objective. However in most practical optimization scenarios with noisy data and/or gradients, it is possible that stochastic gradient descent can inadvertently change critical parameters. In this paper, we argue for the importance of regularizing optimization trajectories directly. We derive a new co-natural gradient update rule for continual learning whereby the new task gradients are preconditioned with the empirical Fisher information of previously learnt tasks. We show that using the co-natural gradient systematically reduces forgetting in continual learning. Moreover, it helps combat overfitting when learning a new task in a low resource scenario.
Code (0)
등록된 구현이 없습니다.
Tasks
Continual LearningSimilar Papers 제목 키워드 기반
P2DT: Mitigating Forgetting in task-incremental Learning with progressive prompt Decision Transformer
Catastrophic forgetting poses a substantial challenge for managing intelligent agents controlled by a large model, causing performance degradation when these agents face new tasks. In our work, we propose a novel solutio…
Incremental Learningreinforcement-learningReinforcement LearningProvable Continual Learning via Sketched Jacobian Approximations
An important problem in machine learning is the ability to learn tasks in a sequential manner. If trained with standard first-order methods most models forget previously learned tasks when trained on a new task, which is…
Continual LearningFew-Shot Regularization to Tackle Catastrophic Forgetting in Multilingual Machine Translation
Increasing the number of tasks supported by a machine learning model without forgetting previously learned tasks is the goal of any lifelong learning system. In this work, we study how to mitigate the effects of the cata…
Lifelong learningMachine TranslationTranslationLifelong Learning with Sketched Structural Regularization
Preventing catastrophic forgetting while continually learning new tasks is an essential problem in lifelong learning. Structural regularization (SR) refers to a family of algorithms that mitigate catastrophic forgetting …
Continual LearningLifelong learningPermuted-MNISTUsing Adapters to Overcome Catastrophic Forgetting in End-to-End Automatic Speech Recognition
Learning a set of tasks in sequence remains a challenge for artificial neural networks, which, in such scenarios, tend to suffer from Catastrophic Forgetting (CF). The same applies to End-to-End (E2E) Automatic Speech Re…
AllAutomatic Speech RecognitionAutomatic Speech Recognition (ASR)speech-recognition+1