Rickaym/

Burmese-MMLU-Lite

Human annotated translations of the Global-MMLU-Lite into the Burmese language. This dataset was duly prepared for use in the Aya Expedition.

To evaluate models on this benchmark, you can use any standard MMLU toolkits. To get started quickly, you can also use ayamytk (Aya Myanmar Toolkit) which was originally developed for running this benchmark.

Install the toolkit directly.

!pip install git+https://github.com/Rickaym/aya-my-tk… See the full description on the dataset page: https://huggingface.co/datasets/Rickaym/Burmese-MMLU-Lite.