opus4m-2020-08-12.yml

release: tut-eng/opus4m-2020-08-12.zip
release-date: 2020-08-12
dataset-name: opus4m
modeltype: transformer
pre-processing: normalization + SentencePiece (spm32k,spm32k)
subwords:
   source: spm32k
   target: spm32k
subword-models:
   source: source.spm
   target: target.spm
source-languages:
   - aze
   - bak
   - chv
   - crh
   - kaz
   - kir
   - kjh
   - kum
   - mon
   - nog
   - ota
   - sah
   - tat
   - tuk
   - tur
   - tyv
   - uig
   - uzb
   - xal
target-languages:
   - eng
test-data:
   Tatoeba-test.aze-eng: 2659/13295
   Tatoeba-test.bak-eng: 39/173
   Tatoeba-test.chv-eng: 333/1876
   Tatoeba-test.crh-eng: 22/92
   Tatoeba-test.kaz-eng: 397/2176
   Tatoeba-test.kir-eng: 118/583
   Tatoeba-test.kjh-eng: 17/60
   Tatoeba-test.kum-eng: 8/41
   Tatoeba-test.mon-eng: 337/2256
   Tatoeba-test.multi-eng: 3070/15499
   Tatoeba-test.nog-eng: 83/373
   Tatoeba-test.ota-eng: 678/3612
   Tatoeba-test.sah-eng: 39/183
   Tatoeba-test.tat-eng: 1451/9447
   Tatoeba-test.tuk-eng: 2500/16260
   Tatoeba-test.tur-eng: 10000/66526
   Tatoeba-test.tyv-eng: 5/20
   Tatoeba-test.uig-eng: 3024/19231
   Tatoeba-test.uzb-eng: 457/2184
   Tatoeba-test.xal-eng: 281/1726
   newsdev2016-entr.tur-eng: 1001/19474
   newstest2016-entr.tur-eng: 3000/58396
   newstest2017-entr.tur-eng: 3007/59741
   newstest2018-entr.tur-eng: 3000/60232
BLEU-scores:
   Tatoeba-test.aze-eng: 13.2
   Tatoeba-test.bak-eng: 2.9
   Tatoeba-test.chv-eng: 0.8
   Tatoeba-test.crh-eng: 14.6
   Tatoeba-test.kaz-eng: 11.8
   Tatoeba-test.kir-eng: 15.1
   Tatoeba-test.kjh-eng: 1.7
   Tatoeba-test.kum-eng: 14.6
   Tatoeba-test.mon-eng: 6.4
   Tatoeba-test.multi-eng: 10.6
   Tatoeba-test.nog-eng: 9.0
   Tatoeba-test.ota-eng: 1.4
   Tatoeba-test.sah-eng: 0.7
   Tatoeba-test.tat-eng: 4.6
   Tatoeba-test.tuk-eng: 5.5
   Tatoeba-test.tur-eng: 28.6
   Tatoeba-test.tyv-eng: 5.5
   Tatoeba-test.uig-eng: 0.1
   Tatoeba-test.uzb-eng: 5.1
   Tatoeba-test.xal-eng: 1.4
   newsdev2016-entr.tur-eng: 5.9
   newstest2016-entr.tur-eng: 4.6
   newstest2017-entr.tur-eng: 4.6
   newstest2018-entr.tur-eng: 4.9
chr-F-scores:
   Tatoeba-test.aze-eng: 0.322
   Tatoeba-test.bak-eng: 0.190
   Tatoeba-test.chv-eng: 0.162
   Tatoeba-test.crh-eng: 0.322
   Tatoeba-test.kaz-eng: 0.302
   Tatoeba-test.kir-eng: 0.337
   Tatoeba-test.kjh-eng: 0.099
   Tatoeba-test.kum-eng: 0.314
   Tatoeba-test.mon-eng: 0.251
   Tatoeba-test.multi-eng: 0.296
   Tatoeba-test.nog-eng: 0.263
   Tatoeba-test.ota-eng: 0.150
   Tatoeba-test.sah-eng: 0.118
   Tatoeba-test.tat-eng: 0.228
   Tatoeba-test.tuk-eng: 0.256
   Tatoeba-test.tur-eng: 0.459
   Tatoeba-test.tyv-eng: 0.190
   Tatoeba-test.uig-eng: 0.113
   Tatoeba-test.uzb-eng: 0.208
   Tatoeba-test.xal-eng: 0.150
   newsdev2016-entr.tur-eng: 0.259
   newstest2016-entr.tur-eng: 0.246
   newstest2017-entr.tur-eng: 0.246
   newstest2018-entr.tur-eng: 0.247