CharLSTM-RNN is a simple plug-and-play script that uses LSTM-based recurrent neural networks to generate text character-by-character. Simply link your source text and go.
CharLSTM-RNN is built on top of Theano, a parallel computing framework that will run on most GPU's.
CharLSTM-RNN was inspired by Andrej Karpathy's char-rnn (https://github.com/karpathy/char-rnn). To learn more about Recurrent Neural Networks, visit http://karpathy.github.io/2015/05/21/rnn-effectiveness/.