A variety of Tensorflow and Numpy implementations of the gradient descent backpropogation algorithm for simple one and two layer networks.
The documentation for these programs is rather poor at the moment, and I will be continually updating them with inline documentation, docstrings, and command line argument parsers. For the moment, just download the code, insert the correct path names, and run in a Python interpreter. Examples and demos will also be added shortly.