Skip to content

Latest commit

 

History

History

aav-aav

README.md

opus-2020-10-04.zip

  • dataset: opus
  • model: transformer
  • source language(s): eng hoc hoc_Latn kha khm khm_Latn mnw vie vie_Hani
  • target language(s): eng hoc hoc_Latn kha khm khm_Latn mnw vie vie_Hani
  • model: transformer
  • pre-processing: normalization + SentencePiece (spm32k,spm32k)
  • a sentence initial language token is required in the form of >>id<< (id = valid target language ID)
  • download: opus-2020-10-04.zip
  • test set translations: opus-2020-10-04.test.txt
  • test set scores: opus-2020-10-04.eval.txt

Benchmarks

testset BLEU chr-F
Tatoeba-test.eng-hoc.eng.hoc 0.1 0.010
Tatoeba-test.eng-kha.eng.kha 0.1 0.013
Tatoeba-test.eng-khm.eng.khm 0.2 0.247
Tatoeba-test.eng-mnw.eng.mnw 0.4 0.003
Tatoeba-test.eng-vie.eng.vie 30.2 0.480
Tatoeba-test.hoc-eng.hoc.eng 0.2 0.099
Tatoeba-test.kha-eng.kha.eng 0.8 0.119
Tatoeba-test.kha-vie.kha.vie 2.0 0.082
Tatoeba-test.khm-eng.khm.eng 12.7 0.302
Tatoeba-test.khm-vie.khm.vie 19.7 0.405
Tatoeba-test.mnw-eng.mnw.eng 1.4 0.059
Tatoeba-test.multi.multi 16.6 0.324
Tatoeba-test.vie-eng.vie.eng 31.6 0.511
Tatoeba-test.vie-kha.vie.kha 0.7 0.015
Tatoeba-test.vie-khm.vie.khm 1.9 0.166

opus4m+btTCv20210807-2021-09-30.zip

Benchmarks

testset BLEU chr-F #sent #words BP
Tatoeba-test-v2021-08-07.kha-vie 2.1 0.078 4 39 0.947
Tatoeba-test-v2021-08-07.khm-vie 37.8 0.581 18 101 1.000
Tatoeba-test-v2021-08-07.multi-multi 28.1 0.412 10000 73768 0.984
Tatoeba-test-v2021-08-07.vie-kha 2.2 0.097 4 37 0.973
Tatoeba-test-v2021-08-07.vie-khm 2.3 0.367 18 37 1.000
Tatoeba-test-v2021-08-07.vie-vie 61.8 0.738 542 6349 0.962
tico19-test.eng-khm 11.0 0.574 2100 20941 0.685
tico19-test.khm-eng 31.1 0.582 2100 56848 1.000