LAMDA-CL/PyCIL
A Python toolbox providing reproducible implementations of 20+ class-incremental learning methods in PyTorch.
Not currently ranked — collecting fresh signals.
star history
PyCIL is a research toolbox for class-incremental learning, a machine learning paradigm where models must continuously learn from new classes without forgetting previously acquired knowledge. It implements and reproduces over 20 state-of-the-art methods from the literature, providing standardized training pipelines and evaluation protocols. The toolbox targets researchers working on continual, lifelong, and open-world learning scenarios.
Frequently asked
- What is LAMDA-CL/PyCIL?
- A Python toolbox providing reproducible implementations of 20+ class-incremental learning methods in PyTorch.
- Is PyCIL open source?
- Yes — LAMDA-CL/PyCIL is an open-source project tracked on heatdrop.
- What language is PyCIL written in?
- LAMDA-CL/PyCIL is primarily written in Python.
- How popular is PyCIL?
- LAMDA-CL/PyCIL has 1.1k stars on GitHub.
- Where can I find PyCIL?
- LAMDA-CL/PyCIL is on GitHub at https://github.com/LAMDA-CL/PyCIL.