Text Generator based upon next character prediction from an MLP

Text Generator based upon next character prediction from an MLP

Feb, 2024 - Mar, 2024 (Advisor – Prof. Prof. Nipun Batra, IIT Gandhinagar) | Project Link

• This project involves generating text by predicting the next character based on the last ’k’ characters. The model used is a neural network with an embedding layer followed by 2 hidden layers. The model is able to capture the semantics of the English language and generate meaningful words and phrases.
• Observed the difference in results based on the choice of embedding and architecture of the neural network. Performed hyperparameter tuning to get the best results. Learned to deploy an application on Streamlit.