Datasets:
GEM
/

Multilinguality: unknown
Size Categories: unknown
Language Creators: unknown
Annotations Creators: none
Source Datasets: original
License:

sampled test / val set?

#2
by Muennighoff - opened

Why is there no sampled test / val set like for Wiki Lingua? https://huggingface.co/datasets/GEM/wiki_lingua/viewer/tr/sampled_test
The challenge datasets are not randomly sampled I think

GEM benchmark org

We did not create one, but you are welcome to create one and submit a PR.

Sign up or log in to comment