Avalanche: A PyTorch Library for Deep Continual Learning
This addresses the need for sustainable and efficient training of deep neural networks over time for researchers and practitioners in machine learning, though it is incremental as it builds on existing PyTorch primitives.
The authors tackled the problem of deep continual learning by introducing Avalanche, a PyTorch library that provides first-class support for dynamic architectures, data streams, and incremental training methods, resulting in an open-source tool with extensive benchmarks and modular design.
Continual learning is the problem of learning from a nonstationary stream of data, a fundamental issue for sustainable and efficient training of deep neural networks over time. Unfortunately, deep learning libraries only provide primitives for offline training, assuming that model's architecture and data are fixed. Avalanche is an open source library maintained by the ContinualAI non-profit organization that extends PyTorch by providing first-class support for dynamic architectures, streams of datasets, and incremental training and evaluation methods. Avalanche provides a large set of predefined benchmarks and training algorithms and it is easy to extend and modular while supporting a wide range of continual learning scenarios. Documentation is available at \url{https://avalanche.continualai.org}.