ContinualAI / avalanche

Avalanche: an End-to-End Library for Continual Learning based on PyTorch.
http://avalanche.continualai.org
MIT License
1.76k stars 289 forks source link

Add a plugin to manage the weight importance computation and penalization to the loss #953

Open AndreaCossu opened 2 years ago

AndreaCossu commented 2 years ago

By using a separate plugin to compute importances and penalize the loss we can reuse the same interface for EWC, MAS, SI and all future strategies.

AntonioCarta commented 2 years ago

I started this PR by implementing the interface for LwF (avalanche.training.regularization). Now it can be used easily outside of Avalanche strategies. We can do the same for the other regularization strategies.