Spaces:

mlabonne

/

Yet_Another_LLM_Leaderboard

like 161

Running

It's a combination of the following tasks: hellaswag,openbookqa,winogrande,arc_easy,arc_challenge,boolq,piqa

It's a combination of the following tasks: hellaswag,openbookqa,winogrande,arc_easy,arc_challenge,boolq,piqa

Thanks for your reply. However, this is still too vague for me to figure out its provenance since I have no idea where the GPT4All benchmark originates.

For example, what's its relationship with https://gpt4all.io/index.html?
What's its relationship with https://arxiv.org/pdf/2311.04931.pdf?
What's its relationship with https://huggingface.co/datasets/nomic-ai/gpt4all_prompt_generations?

I think it is better to attach a link accordingly to redirect us to its provenance like other benchmarks. What do you think? @mlabonne