Datasets:
Verifiable downloads
#64
by
jindrahelcl
- opened
We are downloading fineweb 1 and 2 for the HPLT and OpenEuroLLM projects and it would help if we would be able to verify that we have the right dataset.
Could you provide the exact dataset sizes in bytes rather than GiBs, or any other statistic that can help people with this? Something like the fineweb2-language-distribution.csv would serve well.
Thanks!