add a note in the llama3_1 eval_details.md file mentioning that for SQuAD prev Q & A pairs pertaining to a given passage are included in the prompt.
Motivation:
To the best of my knowledge this evaluation detail isn't mentioned publicly except for in the eval details dataset. My hope is that also including it in the eval_details.md will help other researchers catch this important detail without having to dive deep into the eval details dataset.
Change:
Motivation: