Repository for the ANLP course of CentraleSupelec 2024
Kaggle link: https://www.kaggle.com/competitions/nlp-cs-2024
This is our report presenting our solution more deeply
Where we instantiate our models and execute our code to create predictions
Contains the models used to create our final prediction and a few others that we also tested.
Contains some utility functions used to import data, load, and save predictions...
Script used to generate sentences with Llama2. We also used other techniques and some other variations of the prompt shown in this file to generate our data.
We used the dataset "datafinaltouttout.txt" to train our models. You can also find other scripts of models or data augmentation techniques in the "old" directory.