opus2m-2020-08-01.yml

release: eng-cel/opus2m-2020-08-01.zip
release-date: 2020-08-01
dataset-name: opus2m
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:
   - bre
   - cor
   - cym
   - gla
   - gle
   - glv
use-target-labels:
   - >>bre<<
   - >>cor<<
   - >>cym<<
   - >>gla<<
   - >>gle<<
   - >>glv<<
test-data:
   - Tatoeba-test.eng-bre: 383/1864
   - Tatoeba-test.eng-cor: 3198/14464
   - Tatoeba-test.eng-cym: 818/5073
   - Tatoeba-test.eng-gla: 917/5898
   - Tatoeba-test.eng-gle: 1924/10141
   - Tatoeba-test.eng-glv: 5/20
   - Tatoeba-test.eng-multi: 7245/37460
BLEU-scores:
   - Tatoeba-test.eng-bre: 11.5
   - Tatoeba-test.eng-cor: 0.3
   - Tatoeba-test.eng-cym: 31.0
   - Tatoeba-test.eng-gla: 7.6
   - Tatoeba-test.eng-gle: 35.9
   - Tatoeba-test.eng-glv: 9.9
   - Tatoeba-test.eng-multi: 18.0
chr-F-scores:
   - Tatoeba-test.eng-bre: 0.338
   - Tatoeba-test.eng-cor: 0.095
   - Tatoeba-test.eng-cym: 0.549
   - Tatoeba-test.eng-gla: 0.317
   - Tatoeba-test.eng-gle: 0.582
   - Tatoeba-test.eng-glv: 0.454
   - Tatoeba-test.eng-multi: 0.342