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

propagate vllm_kwargs to cot-configs #5

Open ggbetz opened 5 months ago

ggbetz commented 5 months ago

cot-configs have vllm_kwargs sub-dict, but this is static and does not reflect the vllm arguments given in config.env