search

logikon-ai / cot-eval

A framework for evaluating the effectiveness of chain-of-thought reasoning in language models.
https://huggingface.co/spaces/logikon/open_cot_leaderboard
MIT License
7 stars 1 forks source link

add pipeline-parallel-size #6

Open ggbetz opened 5 months ago

ggbetz commented 5 months ago

add --pipeline-parallel-size for vllm, to efficiently use multi-gpu resources