This project consists of five main files. The notebook (.ipynb) was used to create two model files. The primary objective of this project was to explore how model complexity affects training efficiency and convergence time for natural language processing tasks.

My findings indicate that increasing model complexity speeds up problem-solving to a point. For models with an embedding layer larger than 256 and a hidden layer size beyond 2048, performance improvements began to diminish, with gains becoming smaller as complexity increased further.

The RNN.py file represents my effort to modularize, organize, and automate the code, requiring minimal intervention to generate models efficiently.

Finally, input.txt contains a large sample of English and Catalan text used for training.