Evaluation

Commonsense reasoning benchmarks (PIQA, SIQA, HellaSwag, WinoGrande, ARC, OpenBookQA, CommonsenseQA)

Grouped commonsense reasoning average over PIQA, SIQA, HellaSwag, WinoGrande, ARC easy/challenge, OpenBookQA, CommonsenseQA (7-shot CommonsenseQA, 0-shot others)

Core capabilities Meta

Ask about this evaluation

How it was run

Open a release to see the setup, scoring, and other details that the lab published. Only fields the lab actually disclosed are shown.

Llama 2 Jul 2023