Avalanche: and End-to-End Library for Continual Learning based on PyTorch
Antonio Carta, Andrea Cossu, Hamed Hemati, Lorenzo Pellegrini, Gabriele Graffieti, Vincenzo Lomonaco
Why Avalanche? ❄️
Avalanche is an End-to-End Continual Learning Library (now part of the PyTorch Ecosystem!) powered by ContinualAI with the unique goal of providing a shared and collaborative open-source (MIT licensed) codebase for fast prototyping, training and reproducible evaluation of continual learning algorithms.
Learning continually from a non-stationary stream of experiences is a challenging task, especially for deep neural networks, where simply fine-tuning a pre-trained model on the new available data often incurs in catastrophic forgetting of previously learned knowledge.
Avalanche can help Continual Learning researchers and practitioners in several ways:
- Write less code, prototype faster & reduce errors
- Improve reproducibility
- Improve modularity and reusability
- Increase code efficiency, scalability & portability
- Augment impact and usability of their research products
Avalanche in a Nutshell 🌰
The library is organized in five main modules, in order to provide the basic building blocks of any continual learning experiment:
Benchmarks: This module maintains a uniform API for data handling, by generating a stream of data from one or more datasets. It contains all the major CL benchmarks (similar to what has been done for torchvision).
Training: This module provides all the necessary utilities concerning model training. This includes simple and efficient ways to implement new continual learning strategies as well as a set of CL baselines and state-of-the-art algorithms to use for comparison!
Evaluation: This module provides all the utilities and metrics that can help evaluate a CL algorithm with respect to all the factors we believe to be important for a continually learning system like accuracy, forgetting, forward transfer and so on.
Models: This module contains several model architectures and pre-trained models that can be used for your continual learning experiment (similar to what has been done in torchvision.models).
Logging: It includes advanced logging and plotting features, including native stdout, file and TensorBoard support (How cool is it to have a complete, interactive dashboard, tracking your experiment metrics in real-time with a single line of code?)
Check out how your code changes when you start using Avalanche! 👇
Avalanche is the first experiment of a End-to-end Library for reproducible continual learning research & development where you can find benchmarks, algorithms, evaluation metrics and much more, in the same place.
Do you want to start using Avalanche right now? Check out the complete “From Zero to Hero” tutorial runnable on google colab!
Maintained by ContinualAI 🤗
Avalanche is the flagship open-source collaborative project of ContinualAI: a non profit research organization and the largest open community on Continual Learning for AI.
Do you have a question, do you want to report an issue or simply ask for a new feature? Check out the Questions & Issues center. Do you want to improve Avalanche yourself? Follow these simple rules on How to Contribute.
The Avalanche project is maintained by the collaborative research team ContinualAI Lab and used extensively by the Units of the ContinualAI Research (CLAIR) consortium, a research network of the major continual learning stakeholders around the world.
Learn more about the Avalanche team and all the people who made it great!
Let’s make it together 👫 a wonderful ride! 🎈
Cite Avalanche 📑
If you used Avalanche in your research project, please remember to cite our reference paper “Avalanche: an End-to-End Library for Continual Learning”. This will help us make Avalanche better known in the machine learning community, ultimately making a better tool for everyone: