Implementing Simple Recurrent Neural Network on plain Text files
The original Vanilla RNN file was copied from the Stanford cours CS231n: Deel Learning for Natural Language Processing. Link: http://cs231n.stanford.edu/ Spring 2017. See the slides of RRN lecture for technical background. Vanialla RNN does not incoporate forget gates and creates an artificially "deep" network that might lead to exploding or vanishing gradients.
The 2 files [LSTM.ipynd] and [vanilla_RNN.ipynb] are implemented based on the theories and template for each model. See each file for more detailed citations.