In this notebook I will be experimenting with an LSTM to generate text. I will be using a model text (Sherlock Holmes story) that I will train my network on to finally generate new text one charachter at a time. To do so, I will process my text first by using character embedding, then I will create the network and train it, finally I will generate a new charachter based on the last charachter generated by the model. So let's get started !
Generate new text one charachter at a time using an LSTM
- Numpy
- Torch