opus-2021-02-23.yml

release: nor-spa/opus-2021-02-23.zip
release-date: 2021-02-23
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:
   - nno
   - nob
target-languages:
   - spa
training-data:
   - nno-spa: Tatoeba-train (279713) 
   - nob-spa: Tatoeba-train (9149277) 
validation-data:
   - nno-spa: Tatoeba-dev, 24
   - nob-spa: Tatoeba-dev, 976
   - total size of shuffled dev data: 999
   - devset = top 999  lines of Tatoeba-dev.src.shuffled!
test-data:
   - Tatoeba-test.nno-spa: 75/445
   - Tatoeba-test.nob-spa: 885/6866
   - Tatoeba-test.nor-spa: 960/7311
BLEU-scores:
   - Tatoeba-test.nno-spa: 48.2
   - Tatoeba-test.nob-spa: 53.1
   - Tatoeba-test.nor-spa: 53.2
chr-F-scores:
   - Tatoeba-test.nno-spa: 0.638
   - Tatoeba-test.nob-spa: 0.709
   - Tatoeba-test.nor-spa: 0.708