Deep learning framework written from scratch in c++ as a learning exercise & hopefully ongoing project
Currently in progress, as of now this library contains a basic linear algebra library written from scratch to support matrix manipulation and a basic network class to facilitate back propogation and the training of networks. Eventually I plan on adding modern features to this framework such as convolutions, rnns, transformers, and a simplistic interface. I also plan to implement native reinforcement learning (Q learning, DQNs, Actor-Critic, etc) and evolutionary techniques (NEAT, Spiking networks, bi-directional graphing networks, etc).