Closed xiechengmude closed 2 months ago
Hi! Could you specify the precision and hardware you are using?
8A100 for 70b model.
Is there any examples for evaluating a max model for open_llm_leaderboard_tasks in a fast way?
Hi!
It's not going to need the same model parallelism/data parallelism ratio depending on it.
Hi, closing this issue for inactivity, but feel free to re-open if you need help.
CUDA Memory OOM when I launch an evaluation for 30b model using lighteval.
Whats the correct config for it?