As suggested by https://huggingface.co/datasets/openlanguagedata/flores_plus/discussions/39#6a33e65093b77d390c6d05fd, we could also report the tokens/sentence ratio, which is more comparable across languages.
As suggested by https://huggingface.co/datasets/openlanguagedata/flores_plus/discussions/39#6a33e65093b77d390c6d05fd, we could also report the tokens/sentence ratio, which is more comparable across languages.