huggingface / lighteval

Lighteval is your all-in-one toolkit for evaluating LLMs across multiple backends
MIT License
845 stars 100 forks source link

Allow kwargs for BERTScore compute function and remove unused var #311

Closed chuandudx closed 2 months ago

chuandudx commented 2 months ago

This PR addresses items 1 and 3 in this issue: https://github.com/huggingface/lighteval/issues/310. Thank you in advance for your feedback.

I also am curious if there is an easier way to set rescale_with_baseline if we don't do it directly within this file.

Thanks!

chuandudx commented 2 months ago

I checked the failure and saw user authentication related error - is this something related to any configurations I have?

FAILED tests/logging/test_evaluation_tracker.py::test_push_to_hub_works - huggingface_hub.utils._errors.HfHubHTTPError: 401 Client Error: Unauthorized for url: https://huggingface.co/api/repos/create (Request ID: Root=1-66e99f24-188495d77a850f1131fd8f50;312a7ad8-514c-4932-92fb-ea98d5d420a6)
chuandudx commented 2 months ago

Updated branch and doubled checked for style issues :)