Collections

This collection provides detailed information on how we derived the reported benchmark metrics for the Llama 3.1 models, including the configurations,

NGL, don't look in here.