Open Deep Learning

OpenDL: A ANSI C/C++14/17 implementation of state-of-the-art deep learning models

Author: Shyamal Suhana Chandra
Date Started: June 17, 2017

Purpose: To provide the latest and greatest models of deep learning with clear syntax, documentation, and comments in the code for quick usage.

Time expected to completion: T.B.D.

Phases of the Project:

1. Phase 0 - Define a design specification for the C/C++17 interfaces based on UML and Design Patterns.
2. Phase 1 - Create a diagram of the whole architecture
3. Phase 2 - Implement the skeleton of the architecture along with the pub/sub architecture like ROS.
4. Phase 3 - Agile process with different deep neural networks like Convolutional Networks, Recurrent NN’s, GRU’s, and LSTM’s.
5. Phase 4 - Agile process with different SOTA networks like GANs, VAEs, DQN, DeepRL, AlphaGo (w/ MCTS), and DNC’s.

Please help out and suggest other networks in this README.

Happy and Clear Coding! Best of luck! ~ SSC