Small Data? No Problem! Exploring the Viability of Pretrained Multilingual Language Models for Low-resourced Languages