Yes, we did! The dataset has since been cleaned even more so we're due to update the model.