La-MAML: Look-ahead Meta Learning for Continual Learning
This addresses the challenge of training models to learn sequentially arriving tasks without forgetting previous ones, which is crucial for real-world AI applications, though it is an incremental improvement over existing meta-learning approaches.
The paper tackles the problem of catastrophic forgetting in continual learning by proposing La-MAML, a fast meta-learning algorithm that uses per-parameter learning rate modulation and a small episodic memory, achieving superior performance on real-world visual classification benchmarks compared to other methods.
The continual learning problem involves training models with limited capacity to perform well on a set of an unknown number of sequentially arriving tasks. While meta-learning shows great potential for reducing interference between old and new tasks, the current training procedures tend to be either slow or offline, and sensitive to many hyper-parameters. In this work, we propose Look-ahead MAML (La-MAML), a fast optimisation-based meta-learning algorithm for online-continual learning, aided by a small episodic memory. Our proposed modulation of per-parameter learning rates in our meta-learning update allows us to draw connections to prior work on hypergradients and meta-descent. This provides a more flexible and efficient way to mitigate catastrophic forgetting compared to conventional prior-based methods. La-MAML achieves performance superior to other replay-based, prior-based and meta-learning based approaches for continual learning on real-world visual classification benchmarks. Source code can be found here: https://github.com/montrealrobotics/La-MAML