fix llm as judge warnings (#173)
* commit
* fixes
* fix style
* fixes
* make style
* Fix import error detection for open ai package (llm as a judge metric)
---------
Co-authored-by: Nathan Habib <nathan.habib@huggingface.com>
Co-authored-by: Clémentine Fourrier <22726840+clefourrier@users.noreply.github.com>