Basic character-level word language predictor (English, Turkish, French) with Many-to-One LSTM written in TensorFlow.