Turkish Data
#80
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
{{ message }}
-
I've been emailing Baran about Turkish data and apparently some of the data used from CCAligned is Azerbaijani_language not Turkish. He recommends excluding it and training the next Turkish model on just OpenSubtitles.
Beta Was this translation helpful? Give feedback.
All reactions