opus2m-2020-08-01.yml

release: eng-pqw/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:
   - akl
   - ceb
   - cha
   - dtp
   - hil
   - iba
   - ilo
   - ind
   - jav
   - mad
   - max
   - min
   - mlg
   - pag
   - pau
   - sun
   - tmw
   - war
   - zlm
   - zsm
use-target-labels:
   - ">>akl_Latn<<"
   - ">>ceb<<"
   - ">>cha<<"
   - ">>dtp<<"
   - ">>hil<<"
   - ">>iba<<"
   - ">>ilo<<"
   - ">>ind<<"
   - ">>jav<<"
   - ">>jav_Java<<"
   - ">>mad<<"
   - ">>max_Latn<<"
   - ">>min<<"
   - ">>mlg<<"
   - ">>pag<<"
   - ">>pau<<"
   - ">>sun<<"
   - ">>tmw_Latn<<"
   - ">>war<<"
   - ">>zlm_Latn<<"
   - ">>zsm_Latn<<"
test-data:
   Tatoeba-test.eng-akl: 27/69
   Tatoeba-test.eng-ceb: 378/1678
   Tatoeba-test.eng-cha: 237/796
   Tatoeba-test.eng-dtp: 1929/11440
   Tatoeba-test.eng-hil: 22/104
   Tatoeba-test.eng-iba: 30/235
   Tatoeba-test.eng-ilo: 1093/6040
   Tatoeba-test.eng-jav: 259/1320
   Tatoeba-test.eng-mad: 7/23
   Tatoeba-test.eng-mlg: 51/178
   Tatoeba-test.eng-msa: 5000/27896
   Tatoeba-test.eng-multi: 10000/55384
   Tatoeba-test.eng-pag: 49/268
   Tatoeba-test.eng-pau: 34/106
   Tatoeba-test.eng-sun: 26/87
   Tatoeba-test.eng-war: 1512/9364
BLEU-scores:
   Tatoeba-test.eng-akl: 3.0
   Tatoeba-test.eng-ceb: 11.4
   Tatoeba-test.eng-cha: 1.4
   Tatoeba-test.eng-dtp: 0.6
   Tatoeba-test.eng-hil: 17.7
   Tatoeba-test.eng-iba: 14.6
   Tatoeba-test.eng-ilo: 34.0
   Tatoeba-test.eng-jav: 6.2
   Tatoeba-test.eng-mad: 2.6
   Tatoeba-test.eng-mlg: 34.3
   Tatoeba-test.eng-msa: 31.1
   Tatoeba-test.eng-multi: 17.5
   Tatoeba-test.eng-pag: 19.8
   Tatoeba-test.eng-pau: 1.2
   Tatoeba-test.eng-sun: 30.3
   Tatoeba-test.eng-war: 12.6
chr-F-scores:
   Tatoeba-test.eng-akl: 0.143
   Tatoeba-test.eng-ceb: 0.432
   Tatoeba-test.eng-cha: 0.189
   Tatoeba-test.eng-dtp: 0.139
   Tatoeba-test.eng-hil: 0.525
   Tatoeba-test.eng-iba: 0.365
   Tatoeba-test.eng-ilo: 0.590
   Tatoeba-test.eng-jav: 0.299
   Tatoeba-test.eng-mad: 0.154
   Tatoeba-test.eng-mlg: 0.518
   Tatoeba-test.eng-msa: 0.561
   Tatoeba-test.eng-multi: 0.422
   Tatoeba-test.eng-pag: 0.507
   Tatoeba-test.eng-pau: 0.129
   Tatoeba-test.eng-sun: 0.418
   Tatoeba-test.eng-war: 0.439