Better as Generators Than Classifiers: Leveraging LLMs and Synthetic Data for Low-Resource Multilingual Classification
This research explores a novel framework where Large Language Models act as teachers to generate synthetic data, enabling efficient multilingual classificati...