The goal of this project is to classify two pairs of headlines as 'agreed', 'disagreed', or 'unrelated'.
The create_to_translate and data_augmentation were used to create the augmented.csv file. To train or deploy the model, please use the agree_disagree notebook.