Synthetic datasets in low resource languages created using the Kakugo pipeline (https://github.com/Peter-Devine/kakugo)