huggingface / lighteval

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

[BUG] No script named 'run_evals_accelerate.py' #302

Closed mxjmtxrm closed 2 months ago

mxjmtxrm commented 2 months ago

The 'run_evals_accelerate.py' is mentioned in README, but I didn't find it.

NathanHB commented 2 months ago

fixed in #303