Key Features:

• Increase scalability
  • #112 Increase scalability of replay-based methods by supporting buffers that do not fit in memory
• Increase interoperability with other open-source libraries
  • #113 Support for Avalanche training strategies (at least the most relevant ones)
• Improve library usability
  • #116 Renate non-intuitive arguments and methods (need to break this down further)
  • #115 Simplify scripts for benchmarking
  • #109 Reduce the amount of dependencies in the default installation
  • Support more ML tasks by making data loaders more generic. (need to break this down further)
• Improve documentation
  • #107 Add example for text classification
  • #59 Create an overview page for HPO
• Improve testing
  • #114 Create a suite of integration tests

Due date is an approximation and subject to change.