PyCIL: a Python toolbox for class-incremental learning

Traditional machine learning systems are deployed under the closed-world setting, which requires entire training data before offline process. However, real-world applications often face incoming new classes, and a model should incorporate them continually. The paradigm is called Class-Incremental Learning (CIL). We propose Python toolbox that implements several key algorithms for class-incremental to ease burden of researchers in community. contains implementations number founding works CIL such as EWC iCaRL, but also provides current state-of-the-art can be used conducting novel fundamental research. This toolbox, named PyCIL Learning, available at https://github.com/G-U-N/PyCIL