Name Generation with RNN

Character level Reccurent Neural Network trained with Nepali names. In order to account for the exploding gradient problem, the model was implemented with Gradient Clipping.

Loss of the RNN Model :

Names at Iteration 0 :

Names at Iteration 316000 :

Character level RNN architecture :

Sources :

techgaun/nepali-names

Character-level Recurrent Neural Networks in Practice: Comparing Training and Sampling Schemes

The Unreasonable Effectiveness of Recurrent Neural Networks

Contributing

Pull requests are welcome.