A dharma evaluation dataset with the following configuration:
Subject: MMLU, Size: 12
Subject: ARC-Challenge, Size: 12
Subject: ARC-Easy, Size: 12
Subject: BoolQ, Size: 12
Subject: winogrande, Size: 12
Subject: openbookqa, Size: 12
Subject: truthful_qa, Size: 12
Subject: agieval, Size: 12
Made with
https://github.com/pharaouk/dharma 🚀
