Skip to content

v2.3.0

Latest
Compare
Choose a tag to compare
@mjpost mjpost released this 18 Oct 20:50
· 16 commits to master since this release
5166cf7

Features:

  • (#203) Added -tok flores101 and -tok flores200, a.k.a. spbleu.
    These are multilingual tokenizations that make use of the
    multilingual SPM models released by Facebook and described in the
    following papers:
  • (#213) Added JSON formatting for multi-system output (thanks to Manikanta Inugurthi @me-manikanta)
  • (#211) You can now list all test sets for a language pair with --list SRC-TRG.
    Thanks to Jaume Zaragoza (@ZJaume) for adding this feature.
  • Added WMT22 test sets (test set wmt22)
  • System outputs: include with wmt22. Also added wmt21/systems which will produce WMT21 submitted systems.
    To see available systems, give a dummy system to --echo, e.g., sacrebleu -t wmt22 -l en-de --echo ?