metadata
language:
- af
- de
- en
- fy
- gos
- hrx
- lb
- multilingual
- nds
- nl
tags:
- translation
- opus-mt-tc
license: cc-by-4.0
model-index:
- name: opus-mt-tc-big-gmw-gmw
results:
- task:
name: Translation deu-eng
type: translation
args: deu-eng
dataset:
name: news-test2008
type: news-test2008
args: deu-eng
metrics:
- name: BLEU
type: bleu
value: 27.2
- name: chr-F
type: chrf
value: 0.54584
- task:
name: Translation eng-deu
type: translation
args: eng-deu
dataset:
name: news-test2008
type: news-test2008
args: eng-deu
metrics:
- name: BLEU
type: bleu
value: 23.7
- name: chr-F
type: chrf
value: 0.53204
- task:
name: Translation afr-deu
type: translation
args: afr-deu
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: afr-deu
metrics:
- name: BLEU
type: bleu
value: 50.3
- name: chr-F
type: chrf
value: 0.68633
- task:
name: Translation afr-eng
type: translation
args: afr-eng
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: afr-eng
metrics:
- name: BLEU
type: bleu
value: 56.4
- name: chr-F
type: chrf
value: 0.70502
- task:
name: Translation afr-nld
type: translation
args: afr-nld
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: afr-nld
metrics:
- name: BLEU
type: bleu
value: 55.5
- name: chr-F
type: chrf
value: 0.715
- task:
name: Translation bar-bar
type: translation
args: bar-bar
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: bar-bar
metrics:
- name: BLEU
type: bleu
value: 54.1
- name: chr-F
type: chrf
value: 0.57238
- task:
name: Translation deu-afr
type: translation
args: deu-afr
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: deu-afr
metrics:
- name: BLEU
type: bleu
value: 54.2
- name: chr-F
type: chrf
value: 0.70191
- task:
name: Translation deu-deu
type: translation
args: deu-deu
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: deu-deu
metrics:
- name: BLEU
type: bleu
value: 34.6
- name: chr-F
type: chrf
value: 0.57304
- task:
name: Translation deu-eng
type: translation
args: deu-eng
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: deu-eng
metrics:
- name: BLEU
type: bleu
value: 48.4
- name: chr-F
type: chrf
value: 0.65919
- task:
name: Translation deu-nds
type: translation
args: deu-nds
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: deu-nds
metrics:
- name: BLEU
type: bleu
value: 23.2
- name: chr-F
type: chrf
value: 0.48028
- task:
name: Translation deu-nld
type: translation
args: deu-nld
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: deu-nld
metrics:
- name: BLEU
type: bleu
value: 54.4
- name: chr-F
type: chrf
value: 0.71366
- task:
name: Translation deu-tpi
type: translation
args: deu-tpi
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: deu-tpi
metrics:
- name: BLEU
type: bleu
value: 42.9
- name: chr-F
type: chrf
value: 0.6169
- task:
name: Translation drt-deu
type: translation
args: drt-deu
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: drt-deu
metrics:
- name: BLEU
type: bleu
value: 53.4
- name: chr-F
type: chrf
value: 0.71148
- task:
name: Translation drt-eng
type: translation
args: drt-eng
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: drt-eng
metrics:
- name: BLEU
type: bleu
value: 41.9
- name: chr-F
type: chrf
value: 0.60142
- task:
name: Translation drt-fry
type: translation
args: drt-fry
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: drt-fry
metrics:
- name: BLEU
type: bleu
value: 26.3
- name: chr-F
type: chrf
value: 0.43513
- task:
name: Translation drt-nld
type: translation
args: drt-nld
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: drt-nld
metrics:
- name: BLEU
type: bleu
value: 62.9
- name: chr-F
type: chrf
value: 0.73241
- task:
name: Translation eng-afr
type: translation
args: eng-afr
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: eng-afr
metrics:
- name: BLEU
type: bleu
value: 56.4
- name: chr-F
type: chrf
value: 0.7194
- task:
name: Translation eng-deu
type: translation
args: eng-deu
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: eng-deu
metrics:
- name: BLEU
type: bleu
value: 41.8
- name: chr-F
type: chrf
value: 0.62912
- task:
name: Translation eng-eng
type: translation
args: eng-eng
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: eng-eng
metrics:
- name: BLEU
type: bleu
value: 66.3
- name: chr-F
type: chrf
value: 0.80136
- task:
name: Translation eng-nld
type: translation
args: eng-nld
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: eng-nld
metrics:
- name: BLEU
type: bleu
value: 54.3
- name: chr-F
type: chrf
value: 0.70929
- task:
name: Translation eng-sco
type: translation
args: eng-sco
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: eng-sco
metrics:
- name: BLEU
type: bleu
value: 21.4
- name: chr-F
type: chrf
value: 0.50135
- task:
name: Translation eng-srn
type: translation
args: eng-srn
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: eng-srn
metrics:
- name: BLEU
type: bleu
value: 37
- name: chr-F
type: chrf
value: 0.6592
- task:
name: Translation eng-tpi
type: translation
args: eng-tpi
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: eng-tpi
metrics:
- name: BLEU
type: bleu
value: 34.5
- name: chr-F
type: chrf
value: 0.53886
- task:
name: Translation enm-deu
type: translation
args: enm-deu
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: enm-deu
metrics:
- name: BLEU
type: bleu
value: 36.4
- name: chr-F
type: chrf
value: 0.52972
- task:
name: Translation enm-eng
type: translation
args: enm-eng
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: enm-eng
metrics:
- name: BLEU
type: bleu
value: 22.1
- name: chr-F
type: chrf
value: 0.44876
- task:
name: Translation enm-fry
type: translation
args: enm-fry
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: enm-fry
metrics:
- name: BLEU
type: bleu
value: 58.7
- name: chr-F
type: chrf
value: 0.6386
- task:
name: Translation enm-ltz
type: translation
args: enm-ltz
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: enm-ltz
metrics:
- name: BLEU
type: bleu
value: 20.1
- name: chr-F
type: chrf
value: 0.49027
- task:
name: Translation enm-nld
type: translation
args: enm-nld
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: enm-nld
metrics:
- name: BLEU
type: bleu
value: 33
- name: chr-F
type: chrf
value: 0.53731
- task:
name: Translation fry-deu
type: translation
args: fry-deu
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: fry-deu
metrics:
- name: BLEU
type: bleu
value: 28.7
- name: chr-F
type: chrf
value: 0.4882
- task:
name: Translation fry-eng
type: translation
args: fry-eng
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: fry-eng
metrics:
- name: BLEU
type: bleu
value: 24.5
- name: chr-F
type: chrf
value: 0.40304
- task:
name: Translation fry-ltz
type: translation
args: fry-ltz
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: fry-ltz
metrics:
- name: BLEU
type: bleu
value: 22.2
- name: chr-F
type: chrf
value: 0.4158
- task:
name: Translation fry-nld
type: translation
args: fry-nld
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: fry-nld
metrics:
- name: BLEU
type: bleu
value: 40.5
- name: chr-F
type: chrf
value: 0.54939
- task:
name: Translation gos-afr
type: translation
args: gos-afr
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: gos-afr
metrics:
- name: BLEU
type: bleu
value: 43.2
- name: chr-F
type: chrf
value: 0.59703
- task:
name: Translation gos-deu
type: translation
args: gos-deu
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: gos-deu
metrics:
- name: BLEU
type: bleu
value: 25.4
- name: chr-F
type: chrf
value: 0.45302
- task:
name: Translation gos-eng
type: translation
args: gos-eng
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: gos-eng
metrics:
- name: BLEU
type: bleu
value: 23.9
- name: chr-F
type: chrf
value: 0.37587
- task:
name: Translation gos-fry
type: translation
args: gos-fry
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: gos-fry
metrics:
- name: BLEU
type: bleu
value: 22.2
- name: chr-F
type: chrf
value: 0.44738
- task:
name: Translation gos-nld
type: translation
args: gos-nld
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: gos-nld
metrics:
- name: BLEU
type: bleu
value: 26.1
- name: chr-F
type: chrf
value: 0.45701
- task:
name: Translation gsw-deu
type: translation
args: gsw-deu
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: gsw-deu
metrics:
- name: BLEU
type: bleu
value: 27.5
- name: chr-F
type: chrf
value: 0.44989
- task:
name: Translation gsw-nld
type: translation
args: gsw-nld
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: gsw-nld
metrics:
- name: BLEU
type: bleu
value: 23.6
- name: chr-F
type: chrf
value: 0.43814
- task:
name: Translation hrx-deu
type: translation
args: hrx-deu
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: hrx-deu
metrics:
- name: BLEU
type: bleu
value: 30
- name: chr-F
type: chrf
value: 0.5184
- task:
name: Translation hrx-eng
type: translation
args: hrx-eng
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: hrx-eng
metrics:
- name: BLEU
type: bleu
value: 29.2
- name: chr-F
type: chrf
value: 0.42778
- task:
name: Translation lim-deu
type: translation
args: lim-deu
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: lim-deu
metrics:
- name: BLEU
type: bleu
value: 58.6
- name: chr-F
type: chrf
value: 0.68147
- task:
name: Translation lim-eng
type: translation
args: lim-eng
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: lim-eng
metrics:
- name: BLEU
type: bleu
value: 59.7
- name: chr-F
type: chrf
value: 0.66629
- task:
name: Translation lim-nld
type: translation
args: lim-nld
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: lim-nld
metrics:
- name: BLEU
type: bleu
value: 47.5
- name: chr-F
type: chrf
value: 0.68145
- task:
name: Translation ltz-deu
type: translation
args: ltz-deu
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: ltz-deu
metrics:
- name: BLEU
type: bleu
value: 21
- name: chr-F
type: chrf
value: 0.37005
- task:
name: Translation ltz-eng
type: translation
args: ltz-eng
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: ltz-eng
metrics:
- name: BLEU
type: bleu
value: 30.1
- name: chr-F
type: chrf
value: 0.37764
- task:
name: Translation ltz-fry
type: translation
args: ltz-fry
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: ltz-fry
metrics:
- name: BLEU
type: bleu
value: 27.3
- name: chr-F
type: chrf
value: 0.47498
- task:
name: Translation ltz-nld
type: translation
args: ltz-nld
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: ltz-nld
metrics:
- name: BLEU
type: bleu
value: 26.4
- name: chr-F
type: chrf
value: 0.32392
- task:
name: Translation multi-multi
type: translation
args: multi-multi
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: multi-multi
metrics:
- name: BLEU
type: bleu
value: 40.4
- name: chr-F
type: chrf
value: 0.594
- task:
name: Translation nds-deu
type: translation
args: nds-deu
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: nds-deu
metrics:
- name: BLEU
type: bleu
value: 45.5
- name: chr-F
type: chrf
value: 0.63898
- task:
name: Translation nds-eng
type: translation
args: nds-eng
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: nds-eng
metrics:
- name: BLEU
type: bleu
value: 38.4
- name: chr-F
type: chrf
value: 0.55112
- task:
name: Translation nds-nld
type: translation
args: nds-nld
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: nds-nld
metrics:
- name: BLEU
type: bleu
value: 49.8
- name: chr-F
type: chrf
value: 0.66676
- task:
name: Translation nld-afr
type: translation
args: nld-afr
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: nld-afr
metrics:
- name: BLEU
type: bleu
value: 62.3
- name: chr-F
type: chrf
value: 0.7661
- task:
name: Translation nld-deu
type: translation
args: nld-deu
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: nld-deu
metrics:
- name: BLEU
type: bleu
value: 56.7
- name: chr-F
type: chrf
value: 0.73047
- task:
name: Translation nld-eng
type: translation
args: nld-eng
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: nld-eng
metrics:
- name: BLEU
type: bleu
value: 60.2
- name: chr-F
type: chrf
value: 0.7394
- task:
name: Translation nld-fry
type: translation
args: nld-fry
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: nld-fry
metrics:
- name: BLEU
type: bleu
value: 31
- name: chr-F
type: chrf
value: 0.47959
- task:
name: Translation nld-nds
type: translation
args: nld-nds
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: nld-nds
metrics:
- name: BLEU
type: bleu
value: 20
- name: chr-F
type: chrf
value: 0.43743
- task:
name: Translation nld-nld
type: translation
args: nld-nld
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: nld-nld
metrics:
- name: BLEU
type: bleu
value: 44.9
- name: chr-F
type: chrf
value: 0.63646
- task:
name: Translation nld-sco
type: translation
args: nld-sco
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: nld-sco
metrics:
- name: BLEU
type: bleu
value: 37
- name: chr-F
type: chrf
value: 0.68223
- task:
name: Translation ofs-bar
type: translation
args: ofs-bar
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: ofs-bar
metrics:
- name: BLEU
type: bleu
value: 22.3
- name: chr-F
type: chrf
value: 0.31888
- task:
name: Translation pdc-deu
type: translation
args: pdc-deu
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: pdc-deu
metrics:
- name: BLEU
type: bleu
value: 39.5
- name: chr-F
type: chrf
value: 0.49683
- task:
name: Translation pdc-eng
type: translation
args: pdc-eng
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: pdc-eng
metrics:
- name: BLEU
type: bleu
value: 34.6
- name: chr-F
type: chrf
value: 0.47919
- task:
name: Translation sco-eng
type: translation
args: sco-eng
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: sco-eng
metrics:
- name: BLEU
type: bleu
value: 25.9
- name: chr-F
type: chrf
value: 0.43718
- task:
name: Translation sco-nld
type: translation
args: sco-nld
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: sco-nld
metrics:
- name: BLEU
type: bleu
value: 43.7
- name: chr-F
type: chrf
value: 0.55897
- task:
name: Translation srn-eng
type: translation
args: srn-eng
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: srn-eng
metrics:
- name: BLEU
type: bleu
value: 42.2
- name: chr-F
type: chrf
value: 0.43374
- task:
name: Translation stq-deu
type: translation
args: stq-deu
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: stq-deu
metrics:
- name: BLEU
type: bleu
value: 35.5
- name: chr-F
type: chrf
value: 0.5622
- task:
name: Translation stq-eng
type: translation
args: stq-eng
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: stq-eng
metrics:
- name: BLEU
type: bleu
value: 28.9
- name: chr-F
type: chrf
value: 0.43818
- task:
name: Translation stq-nld
type: translation
args: stq-nld
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: stq-nld
metrics:
- name: BLEU
type: bleu
value: 27
- name: chr-F
type: chrf
value: 0.49046
- task:
name: Translation swg-eng
type: translation
args: swg-eng
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: swg-eng
metrics:
- name: BLEU
type: bleu
value: 22.2
- name: chr-F
type: chrf
value: 0.40521
- task:
name: Translation swg-nld
type: translation
args: swg-nld
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: swg-nld
metrics:
- name: BLEU
type: bleu
value: 32.8
- name: chr-F
type: chrf
value: 0.38592
- task:
name: Translation tpi-deu
type: translation
args: tpi-deu
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: tpi-deu
metrics:
- name: BLEU
type: bleu
value: 26.8
- name: chr-F
type: chrf
value: 0.5189
- task:
name: Translation tpi-eng
type: translation
args: tpi-eng
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: tpi-eng
metrics:
- name: BLEU
type: bleu
value: 27.1
- name: chr-F
type: chrf
value: 0.43439
- task:
name: Translation zea-deu
type: translation
args: zea-deu
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: zea-deu
metrics:
- name: BLEU
type: bleu
value: 60.4
- name: chr-F
type: chrf
value: 0.72725
- task:
name: Translation zea-eng
type: translation
args: zea-eng
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: zea-eng
metrics:
- name: BLEU
type: bleu
value: 50.7
- name: chr-F
type: chrf
value: 0.56884
- task:
name: Translation zea-fry
type: translation
args: zea-fry
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: zea-fry
metrics:
- name: BLEU
type: bleu
value: 28.6
- name: chr-F
type: chrf
value: 0.44983
- task:
name: Translation zea-nds
type: translation
args: zea-nds
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: zea-nds
metrics:
- name: BLEU
type: bleu
value: 34.6
- name: chr-F
type: chrf
value: 0.51589
- task:
name: Translation zea-nld
type: translation
args: zea-nld
dataset:
name: tatoeba-test-v2020-07-28-v2021-08-07
type: tatoeba_mt
args: zea-nld
metrics:
- name: BLEU
type: bleu
value: 47.5
- name: chr-F
type: chrf
value: 0.71248
- task:
name: Translation deu-eng
type: translation
args: deu-eng
dataset:
name: newstest2009
type: wmt-2009-news
args: deu-eng
metrics:
- name: BLEU
type: bleu
value: 25.9
- name: chr-F
type: chrf
value: 0.53749
- task:
name: Translation eng-deu
type: translation
args: eng-deu
dataset:
name: newstest2009
type: wmt-2009-news
args: eng-deu
metrics:
- name: BLEU
type: bleu
value: 22.9
- name: chr-F
type: chrf
value: 0.53283
- task:
name: Translation deu-eng
type: translation
args: deu-eng
dataset:
name: newstest2010
type: wmt-2010-news
args: deu-eng
metrics:
- name: BLEU
type: bleu
value: 30.6
- name: chr-F
type: chrf
value: 0.58356
- task:
name: Translation eng-deu
type: translation
args: eng-deu
dataset:
name: newstest2010
type: wmt-2010-news
args: eng-deu
metrics:
- name: BLEU
type: bleu
value: 25.8
- name: chr-F
type: chrf
value: 0.54886
- task:
name: Translation deu-eng
type: translation
args: deu-eng
dataset:
name: newstest2011
type: wmt-2011-news
args: deu-eng
metrics:
- name: BLEU
type: bleu
value: 26.3
- name: chr-F
type: chrf
value: 0.54883
- task:
name: Translation eng-deu
type: translation
args: eng-deu
dataset:
name: newstest2011
type: wmt-2011-news
args: eng-deu
metrics:
- name: BLEU
type: bleu
value: 23.1
- name: chr-F
type: chrf
value: 0.52712
- task:
name: Translation deu-eng
type: translation
args: deu-eng
dataset:
name: newstest2012
type: wmt-2012-news
args: deu-eng
metrics:
- name: BLEU
type: bleu
value: 28.5
- name: chr-F
type: chrf
value: 0.5616
- task:
name: Translation eng-deu
type: translation
args: eng-deu
dataset:
name: newstest2012
type: wmt-2012-news
args: eng-deu
metrics:
- name: BLEU
type: bleu
value: 23.3
- name: chr-F
type: chrf
value: 0.52662
- task:
name: Translation deu-eng
type: translation
args: deu-eng
dataset:
name: newstest2013
type: wmt-2013-news
args: deu-eng
metrics:
- name: BLEU
type: bleu
value: 31.4
- name: chr-F
type: chrf
value: 0.5777
- task:
name: Translation eng-deu
type: translation
args: eng-deu
dataset:
name: newstest2013
type: wmt-2013-news
args: eng-deu
metrics:
- name: BLEU
type: bleu
value: 27.8
- name: chr-F
type: chrf
value: 0.55774
- task:
name: Translation deu-eng
type: translation
args: deu-eng
dataset:
name: newstest2014-deen
type: wmt-2014-news
args: deu-eng
metrics:
- name: BLEU
type: bleu
value: 33.2
- name: chr-F
type: chrf
value: 0.59826
- task:
name: Translation eng-deu
type: translation
args: eng-deu
dataset:
name: newstest2014-deen
type: wmt-2014-news
args: eng-deu
metrics:
- name: BLEU
type: bleu
value: 29.6
- name: chr-F
type: chrf
value: 0.59441
- task:
name: Translation deu-eng
type: translation
args: deu-eng
dataset:
name: newstest2015-ende
type: wmt-2015-news
args: deu-eng
metrics:
- name: BLEU
type: bleu
value: 33.4
- name: chr-F
type: chrf
value: 0.5966
- task:
name: Translation eng-deu
type: translation
args: eng-deu
dataset:
name: newstest2015-ende
type: wmt-2015-news
args: eng-deu
metrics:
- name: BLEU
type: bleu
value: 32.3
- name: chr-F
type: chrf
value: 0.59889
- task:
name: Translation deu-eng
type: translation
args: deu-eng
dataset:
name: newstest2016-ende
type: wmt-2016-news
args: deu-eng
metrics:
- name: BLEU
type: bleu
value: 39.8
- name: chr-F
type: chrf
value: 0.64736
- task:
name: Translation eng-deu
type: translation
args: eng-deu
dataset:
name: newstest2016-ende
type: wmt-2016-news
args: eng-deu
metrics:
- name: BLEU
type: bleu
value: 38.3
- name: chr-F
type: chrf
value: 0.64429
- task:
name: Translation deu-eng
type: translation
args: deu-eng
dataset:
name: newstest2017-ende
type: wmt-2017-news
args: deu-eng
metrics:
- name: BLEU
type: bleu
value: 35.2
- name: chr-F
type: chrf
value: 0.60933
- task:
name: Translation eng-deu
type: translation
args: eng-deu
dataset:
name: newstest2017-ende
type: wmt-2017-news
args: eng-deu
metrics:
- name: BLEU
type: bleu
value: 30.7
- name: chr-F
type: chrf
value: 0.59258
- task:
name: Translation deu-eng
type: translation
args: deu-eng
dataset:
name: newstest2018-ende
type: wmt-2018-news
args: deu-eng
metrics:
- name: BLEU
type: bleu
value: 42.6
- name: chr-F
type: chrf
value: 0.66796
- task:
name: Translation eng-deu
type: translation
args: eng-deu
dataset:
name: newstest2018-ende
type: wmt-2018-news
args: eng-deu
metrics:
- name: BLEU
type: bleu
value: 46.5
- name: chr-F
type: chrf
value: 0.69605
- task:
name: Translation deu-eng
type: translation
args: deu-eng
dataset:
name: newstest2019-deen
type: wmt-2019-news
args: deu-eng
metrics:
- name: BLEU
type: bleu
value: 39.8
- name: chr-F
type: chrf
value: 0.63766
- task:
name: Translation eng-deu
type: translation
args: eng-deu
dataset:
name: newstest2019-ende
type: wmt-2019-news
args: eng-deu
metrics:
- name: BLEU
type: bleu
value: 43.3
- name: chr-F
type: chrf
value: 0.6688
opus-mt-tc-big-gmw-gmw
Table of Contents
- Model Details
- Uses
- Risks, Limitations and Biases
- How to Get Started With the Model
- Training
- Evaluation
- Citation Information
- Acknowledgements
Model Details
Neural machine translation model for translating from West Germanic languages (gmw) to West Germanic languages (gmw).
This model is part of the OPUS-MT project, an effort to make neural machine translation models widely available and accessible for many languages in the world. All models are originally trained using the amazing framework of Marian NMT, an efficient NMT implementation written in pure C++. The models have been converted to pyTorch using the transformers library by huggingface. Training data is taken from OPUS and training pipelines use the procedures of OPUS-MT-train. Model Description:
- Developed by: Language Technology Research Group at the University of Helsinki
- Model Type: Translation (transformer-big)
- Release: 2022-08-11
- License: CC-BY-4.0
- Language(s):
- Source Language(s): afr deu eng enm fry gos gsw hrx ksh ltz multi nds nld pdc sco stq swg tpi yid
- Target Language(s): afr ang deu eng enm fry gos ltz multi nds nld sco tpi yid
- Language Pair(s): afr-deu afr-eng afr-nld deu-afr deu-deu deu-eng deu-nds deu-nld eng-afr eng-deu eng-eng eng-nld fry-eng fry-nld gos-deu gos-eng gos-nld hrx-deu hrx-eng ltz-deu ltz-eng ltz-nld multi-multi nds-deu nds-eng nds-nld nld-afr nld-deu nld-eng nld-fry nld-nds nld-nld
- Valid Target Language Labels: >>act<< >>afr<< >>afs<< >>aig<< >>ang<< >>ang_Latn<< >>bah<< >>bar<< >>bis<< >>bjs<< >>brc<< >>bzj<< >>bzj_Latn<< >>bzk<< >>cim<< >>dcr<< >>deu<< >>djk<< >>djk_Latn<< >>drt<< >>drt_Latn<< >>dum<< >>eng<< >>enm<< >>enm_Latn<< >>fpe<< >>frk<< >>frr<< >>fry<< >>gcl<< >>gct<< >>geh<< >>gmh<< >>gml<< >>goh<< >>gos<< >>gpe<< >>gsw<< >>gul<< >>gyn<< >>hrx<< >>hrx_Latn<< >>hwc<< >>icr<< >>jam<< >>jvd<< >>kri<< >>ksh<< >>kww<< >>lim<< >>lng<< >>ltz<< >>mhn<< >>nds<< >>nld<< >>odt<< >>ofs<< >>ofs_Latn<< >>oor<< >>osx<< >>pcm<< >>pdc<< >>pdt<< >>pey<< >>pfl<< >>pih<< >>pih_Latn<< >>pis<< >>pis_Latn<< >>qlm<< >>rop<< >>sco<< >>sdz<< >>skw<< >>sli<< >>srm<< >>srm_Latn<< >>srn<< >>stl<< >>stq<< >>svc<< >>swg<< >>sxu<< >>tch<< >>tcs<< >>tgh<< >>tpi<< >>trf<< >>twd<< >>uln<< >>vel<< >>vic<< >>vls<< >>vmf<< >>wae<< >>wep<< >>wes<< >>wes_Latn<< >>wym<< >>ydd<< >>yec<< >>yid<< >>yih<< >>zea<<
- Original Model: opusTCv20210807_transformer-big_2022-08-11.zip
- Resources for more information:
- OPUS-MT-train GitHub Repo
- More information about released models for this language pair: OPUS-MT gmw-gmw README
- More information about MarianNMT models in the transformers library
- [Tatoeba Translation Challenge](https://github.com/Helsinki-NLP/Tatoeba-Challenge/
This is a multilingual translation model with multiple target languages. A sentence initial language token is required in the form of >>id<< (id = valid target language ID), e.g. >>afr<<
Uses
This model can be used for translation and text-to-text generation.
Risks, Limitations and Biases
CONTENT WARNING: Readers should be aware that the model is trained on various public data sets that may contain content that is disturbing, offensive, and can propagate historical and current stereotypes.
Significant research has explored bias and fairness issues with language models (see, e.g., Sheng et al. (2021) and Bender et al. (2021)).
How to Get Started With the Model
A short example code:
from transformers import MarianMTModel, MarianTokenizer
src_text = [
">>nds<< Red keinen Quatsch.",
">>eng<< Findet ihr das nicht etwas übereilt?"
]
model_name = "pytorch-models/opus-mt-tc-big-gmw-gmw"
tokenizer = MarianTokenizer.from_pretrained(model_name)
model = MarianMTModel.from_pretrained(model_name)
translated = model.generate(**tokenizer(src_text, return_tensors="pt", padding=True))
for t in translated:
print( tokenizer.decode(t, skip_special_tokens=True) )
# expected output:
# Kiek ok bi: Rott.
# Aren't you in a hurry?
You can also use OPUS-MT models with the transformers pipelines, for example:
from transformers import pipeline
pipe = pipeline("translation", model="Helsinki-NLP/opus-mt-tc-big-gmw-gmw")
print(pipe(">>nds<< Red keinen Quatsch."))
# expected output: Kiek ok bi: Rott.
Training
- Data: opusTCv20210807 (source)
- Pre-processing: SentencePiece (spm32k,spm32k)
- Model Type: transformer-big
- Original MarianNMT Model: opusTCv20210807_transformer-big_2022-08-11.zip
- Training Scripts: GitHub Repo
Evaluation
- test set translations: opusTCv20210807_transformer-big_2022-08-11.test.txt
- test set scores: opusTCv20210807_transformer-big_2022-08-11.eval.txt
- benchmark results: benchmark_results.txt
- benchmark output: benchmark_translations.zip
| langpair | testset | chr-F | BLEU | #sent | #words |
|---|---|---|---|---|---|
| afr-deu | tatoeba-test-v2020-07-28-v2021-08-07 | 0.68633 | 50.3 | 1583 | 9105 |
| afr-eng | tatoeba-test-v2020-07-28-v2021-08-07 | 0.70502 | 56.4 | 1374 | 9622 |
| afr-nld | tatoeba-test-v2020-07-28-v2021-08-07 | 0.71500 | 55.5 | 1056 | 6710 |
| deu-afr | tatoeba-test-v2020-07-28-v2021-08-07 | 0.70191 | 54.2 | 1583 | 9507 |
| deu-deu | tatoeba-test-v2020-07-28-v2021-08-07 | 0.57304 | 34.6 | 2500 | 20797 |
| deu-eng | tatoeba-test-v2020-07-28-v2021-08-07 | 0.65919 | 48.4 | 17565 | 149415 |
| deu-nds | tatoeba-test-v2020-07-28-v2021-08-07 | 0.48028 | 23.2 | 9999 | 76119 |
| deu-nld | tatoeba-test-v2020-07-28-v2021-08-07 | 0.71366 | 54.4 | 10218 | 75208 |
| deu-yid | tatoeba-test-v2020-07-28-v2021-08-07 | 9.234 | 0.4 | 853 | 5353 |
| eng-afr | tatoeba-test-v2020-07-28-v2021-08-07 | 0.71940 | 56.4 | 1374 | 10314 |
| eng-deu | tatoeba-test-v2020-07-28-v2021-08-07 | 0.62912 | 41.8 | 17565 | 151539 |
| eng-eng | tatoeba-test-v2020-07-28-v2021-08-07 | 0.80136 | 66.3 | 12062 | 115099 |
| eng-nld | tatoeba-test-v2020-07-28-v2021-08-07 | 0.70929 | 54.3 | 12696 | 91769 |
| eng-yid | tatoeba-test-v2020-07-28-v2021-08-07 | 9.648 | 0.4 | 2483 | 16388 |
| fry-eng | tatoeba-test-v2020-07-28-v2021-08-07 | 0.40304 | 24.5 | 220 | 1573 |
| fry-nld | tatoeba-test-v2020-07-28-v2021-08-07 | 0.54939 | 40.5 | 260 | 1854 |
| gos-deu | tatoeba-test-v2020-07-28-v2021-08-07 | 0.45302 | 25.4 | 207 | 1168 |
| gos-eng | tatoeba-test-v2020-07-28-v2021-08-07 | 0.37587 | 23.9 | 1154 | 5634 |
| gos-nld | tatoeba-test-v2020-07-28-v2021-08-07 | 0.45701 | 26.1 | 1852 | 9902 |
| hrx-deu | tatoeba-test-v2020-07-28-v2021-08-07 | 0.51840 | 30.0 | 471 | 2805 |
| hrx-eng | tatoeba-test-v2020-07-28-v2021-08-07 | 0.42778 | 29.2 | 221 | 1235 |
| ltz-deu | tatoeba-test-v2020-07-28-v2021-08-07 | 0.37005 | 21.0 | 347 | 2208 |
| ltz-eng | tatoeba-test-v2020-07-28-v2021-08-07 | 0.37764 | 30.1 | 293 | 1840 |
| ltz-nld | tatoeba-test-v2020-07-28-v2021-08-07 | 0.32392 | 26.4 | 292 | 1685 |
| multi-multi | tatoeba-test-v2020-07-28-v2021-08-07 | 0.59400 | 40.4 | 10000 | 74505 |
| nds-deu | tatoeba-test-v2020-07-28-v2021-08-07 | 0.63898 | 45.5 | 9999 | 74544 |
| nds-eng | tatoeba-test-v2020-07-28-v2021-08-07 | 0.55112 | 38.4 | 2500 | 17584 |
| nds-nld | tatoeba-test-v2020-07-28-v2021-08-07 | 0.66676 | 49.8 | 1657 | 11489 |
| nld-afr | tatoeba-test-v2020-07-28-v2021-08-07 | 0.76610 | 62.3 | 1056 | 6823 |
| nld-deu | tatoeba-test-v2020-07-28-v2021-08-07 | 0.73047 | 56.7 | 10218 | 74121 |
| nld-eng | tatoeba-test-v2020-07-28-v2021-08-07 | 0.73940 | 60.2 | 12696 | 89970 |
| nld-fry | tatoeba-test-v2020-07-28-v2021-08-07 | 0.47959 | 31.0 | 260 | 1857 |
| nld-nds | tatoeba-test-v2020-07-28-v2021-08-07 | 0.43743 | 20.0 | 1657 | 11711 |
| nld-nld | tatoeba-test-v2020-07-28-v2021-08-07 | 0.63646 | 44.9 | 1000 | 7196 |
| swg-deu | tatoeba-test-v2020-07-28-v2021-08-07 | 0.40319 | 16.3 | 1523 | 15630 |
| yid-deu | tatoeba-test-v2020-07-28-v2021-08-07 | 6.304 | 0.1 | 853 | 5172 |
| yid-eng | tatoeba-test-v2020-07-28-v2021-08-07 | 3.715 | 0.1 | 2483 | 15449 |
| yid-yid | tatoeba-test-v2020-07-28-v2021-08-07 | 6.596 | 0.1 | 292 | 1802 |
| deu-eng | newssyscomb2009 | 0.54992 | 28.2 | 502 | 11821 |
| eng-deu | newssyscomb2009 | 0.53867 | 23.2 | 502 | 11271 |
| deu-eng | news-test2008 | 0.54584 | 27.2 | 2051 | 49380 |
| eng-deu | news-test2008 | 0.53204 | 23.7 | 2051 | 47427 |
| deu-eng | newstest2009 | 0.53749 | 25.9 | 2525 | 65402 |
| eng-deu | newstest2009 | 0.53283 | 22.9 | 2525 | 62816 |
| deu-eng | newstest2010 | 0.58356 | 30.6 | 2489 | 61724 |
| eng-deu | newstest2010 | 0.54886 | 25.8 | 2489 | 61511 |
| deu-eng | newstest2011 | 0.54883 | 26.3 | 3003 | 74681 |
| eng-deu | newstest2011 | 0.52712 | 23.1 | 3003 | 72981 |
| deu-eng | newstest2012 | 0.56160 | 28.5 | 3003 | 72812 |
| eng-deu | newstest2012 | 0.52662 | 23.3 | 3003 | 72886 |
| deu-eng | newstest2013 | 0.57770 | 31.4 | 3000 | 64505 |
| eng-deu | newstest2013 | 0.55774 | 27.8 | 3000 | 63737 |
| deu-eng | newstest2014-deen | 0.59826 | 33.2 | 3003 | 67337 |
| eng-deu | newstest2014-deen | 0.59441 | 29.6 | 3003 | 62964 |
| deu-eng | newstest2015-ende | 0.59660 | 33.4 | 2169 | 46443 |
| eng-deu | newstest2015-ende | 0.59889 | 32.3 | 2169 | 44260 |
| deu-eng | newstest2016-ende | 0.64736 | 39.8 | 2999 | 64126 |
| eng-deu | newstest2016-ende | 0.64429 | 38.3 | 2999 | 62670 |
| deu-eng | newstest2017-ende | 0.60933 | 35.2 | 3004 | 64399 |
| eng-deu | newstest2017-ende | 0.59258 | 30.7 | 3004 | 61291 |
| deu-eng | newstest2018-ende | 0.66796 | 42.6 | 2998 | 67013 |
| eng-deu | newstest2018-ende | 0.69605 | 46.5 | 2998 | 64276 |
| deu-eng | newstest2019-deen | 0.63766 | 39.8 | 2000 | 39282 |
| eng-deu | newstest2019-ende | 0.66880 | 43.3 | 1997 | 48969 |
Citation Information
- Publications: OPUS-MT – Building open translation services for the World and The Tatoeba Translation Challenge – Realistic Data Sets for Low Resource and Multilingual MT (Please, cite if you use this model.)
@inproceedings{tiedemann-thottingal-2020-opus,
title = "{OPUS}-{MT} {--} Building open translation services for the World",
author = {Tiedemann, J{\"o}rg and Thottingal, Santhosh},
booktitle = "Proceedings of the 22nd Annual Conference of the European Association for Machine Translation",
month = nov,
year = "2020",
address = "Lisboa, Portugal",
publisher = "European Association for Machine Translation",
url = "https://aclanthology.org/2020.eamt-1.61",
pages = "479--480",
}
@inproceedings{tiedemann-2020-tatoeba,
title = "The Tatoeba Translation Challenge {--} Realistic Data Sets for Low Resource and Multilingual {MT}",
author = {Tiedemann, J{\"o}rg},
booktitle = "Proceedings of the Fifth Conference on Machine Translation",
month = nov,
year = "2020",
address = "Online",
publisher = "Association for Computational Linguistics",
url = "https://aclanthology.org/2020.wmt-1.139",
pages = "1174--1182",
}
Acknowledgements
The work is supported by the European Language Grid as pilot project 2866, by the FoTran project, funded by the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme (grant agreement No 771113), and the MeMAD project, funded by the European Union’s Horizon 2020 Research and Innovation Programme under grant agreement No 780069. We are also grateful for the generous computational resources and IT infrastructure provided by CSC -- IT Center for Science, Finland.
Model conversion info
- transformers version: 4.16.2
- OPUS-MT git hash: c1980b5
- port time: Sun Oct 8 14:39:59 EEST 2023
- port machine: LM0-400-22516.local