Edit model card


  • source group: Hebrew

  • target group: Arabic

  • OPUS readme: heb-ara

  • model: transformer

  • source language(s): heb

  • target language(s): apc apc_Latn ara arq arz

  • model: transformer

  • pre-processing: normalization + SentencePiece (spm32k,spm32k)

  • a sentence initial language token is required in the form of >>id<< (id = valid target language ID)

  • download original weights: opus-2020-07-03.zip

  • test set translations: opus-2020-07-03.test.txt

  • test set scores: opus-2020-07-03.eval.txt


testset BLEU chr-F
Tatoeba-test.heb.ara 23.6 0.532

System Info:

Downloads last month
Hosted inference API
This model can be loaded on the Inference API on-demand.