Zenkai - A framework to beyond backpropagation
Zenkai is a framework to help researchers to test techniques beyond simple backpropagation for deep learning machines and also define more flexible architectures and training mechanics.
Get Developing With Zenkai
- Overview
Understand what Zenkai is and how it can help you
- Getting Started
Get started using Zenkai
Modules
- Core (Kaku)
Develop a LearningMachine and implement all its components
- Tansaku
Implement classes for adding exploration in learning
- Ensemble
Modules for implementing ensembles
- TargetProp
Modules for implementing target propagation
- Feedback
Modules for implementing feedback alignment
- Scikit
Modules for adapting scikit learn classes.
- Utils
Global utilities in Zenkai
Other
- Contribute
Contribute to Zenkai
Note
This project is under active development.