Removing duplicate lines in dictionary file

I tried using
sort filename | uniq -u

but of course if a line was duplicated it was removing all the same lines and not keeping a unique line.