opus-2020-06-28.yml

release: eng-gmq/opus-2020-06-28.zip
release-date: 2020-06-28
dataset-name: opus
modeltype: transformer
pre-processing: normalization + SentencePiece (spm32k,spm32k)
subwords:
   source: spm32k
   target: spm32k
subword-models:
   source: source.spm
   target: target.spm
source-languages:
   - eng
target-languages:
   - dan
   - fao
   - isl
   - non
   - swe
use-target-labels:
   - ">>dan<<"
   - ">>fao<<"
   - ">>isl<<"
   - ">>non_Latn<<"
   - ">>swe<<"
test-data:
   Tatoeba-test.eng-dan: 10000/61469
   Tatoeba-test.eng-fao: 294/1564
   Tatoeba-test.eng-isl: 2500/16288
   Tatoeba-test.eng-multi: 0/0
   Tatoeba-test.eng-non: 15/109
   Tatoeba-test.eng-swe: 10000/54832
BLEU-scores:
   Tatoeba-test.eng-dan: 57.2
   Tatoeba-test.eng-fao: 8.2
   Tatoeba-test.eng-isl: 23.1
   Tatoeba-test.eng-multi: 52.0
   Tatoeba-test.eng-non: 0.7
   Tatoeba-test.eng-swe: 57.4
chr-F-scores:
   Tatoeba-test.eng-dan: 0.720
   Tatoeba-test.eng-fao: 0.314
   Tatoeba-test.eng-isl: 0.500
   Tatoeba-test.eng-multi: 0.681
   Tatoeba-test.eng-non: 0.193
   Tatoeba-test.eng-swe: 0.713