search

mjpost / sacrebleu

Reference BLEU implementation that auto-downloads test sets and reports a version string to facilitate cross-lab comparisons
Apache License 2.0
1.07k stars 164 forks source link

Incorrect sample size? #206

Closed kocmitom closed 2 years ago

kocmitom commented 2 years ago

https://github.com/mjpost/sacrebleu/blob/a73315b839dee92c9237c30f64380aa36dda4795/sacrebleu/significance.py#L117

Hi @mjpost , shouldn't this be one zero shorter? It could make everything statistically significant.

kocmitom commented 2 years ago

My bad, it is a number of samples for bootstrap.