segment parquet files by lang

#2
by mbanon - opened

Hi Max,
is it possible to have the parquet files segmented by language? Something like https://huggingface.co/datasets/openeurollm/nemotron-cc-10K-sample-translated/tree/EuroLLM-9B-Instruct
I am extracting samples for the human evaluation of translations, and that would really ease automating the work.
thanks!

Sign up or log in to comment