opus-2020-07-06.yml

release: eng-afa/opus-2020-07-06.zip
release-date: 2020-07-06
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:
   - acm
   - afb
   - amh
   - apc
   - ara
   - arq
   - ary
   - arz
   - hau
   - heb
   - kab
   - mlt
   - rif
   - shy
   - som
   - tir
use-target-labels:
   - >>acm<<
   - >>afb<<
   - >>amh<<
   - >>apc<<
   - >>ara<<
   - >>arq<<
   - >>ary<<
   - >>arz<<
   - >>hau_Latn<<
   - >>heb<<
   - >>kab<<
   - >>mlt<<
   - >>rif_Latn<<
   - >>shy_Latn<<
   - >>som<<
   - >>tir<<
test-data:
   - Tatoeba-test.eng-amh: 190/554
   - Tatoeba-test.eng-ara: 10000/50193
   - Tatoeba-test.eng-hau: 56/317
   - Tatoeba-test.eng-heb: 10000/49113
   - Tatoeba-test.eng-kab: 10000/43651
   - Tatoeba-test.eng-mlt: 203/673
   - Tatoeba-test.eng-multi: 0/0
   - Tatoeba-test.eng-rif: 53/166
   - Tatoeba-test.eng-shy: 16/55
   - Tatoeba-test.eng-som: 1/2
   - Tatoeba-test.eng-tir: 69/305
BLEU-scores:
   - Tatoeba-test.eng-amh: 9.6
   - Tatoeba-test.eng-ara: 11.5
   - Tatoeba-test.eng-hau: 10.1
   - Tatoeba-test.eng-heb: 31.3
   - Tatoeba-test.eng-kab: 1.2
   - Tatoeba-test.eng-mlt: 15.0
   - Tatoeba-test.eng-multi: 13.8
   - Tatoeba-test.eng-rif: 1.6
   - Tatoeba-test.eng-shy: 0.8
   - Tatoeba-test.eng-som: 0.0
   - Tatoeba-test.eng-tir: 2.4
chr-F-scores:
   - Tatoeba-test.eng-amh: 0.502
   - Tatoeba-test.eng-ara: 0.402
   - Tatoeba-test.eng-hau: 0.450
   - Tatoeba-test.eng-heb: 0.542
   - Tatoeba-test.eng-kab: 0.179
   - Tatoeba-test.eng-mlt: 0.525
   - Tatoeba-test.eng-multi: 0.364
   - Tatoeba-test.eng-rif: 0.072
   - Tatoeba-test.eng-shy: 0.066
   - Tatoeba-test.eng-som: 0.294
   - Tatoeba-test.eng-tir: 0.233