meta-llama

/

Llama-3.1-8B-Instruct

like 2.64k

In the model card you say you achieved 51.9 on MATH using Llama 3.1 8B with zero-shots.

I can't reproduce this. Did you use lm-evaluation-harness or some other code?