opus-2020-07-27.yml

release: iir-iir/opus-2020-07-27.zip
release-date: 2020-07-27
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:
   - asm
   - hin
   - mar
   - urd
   - zza
target-languages:
   - asm
   - hin
   - mar
   - urd
   - zza
use-target-labels:
   - ">>asm<<"
   - ">>hin<<"
   - ">>mar<<"
   - ">>urd<<"
   - ">>zza<<"
test-data:
   Tatoeba-test.asm-hin: 4/15
   Tatoeba-test.asm-zza: 1/4
   Tatoeba-test.hin-asm: 4/13
   Tatoeba-test.hin-mar: 158/692
   Tatoeba-test.hin-urd: 239/1591
   Tatoeba-test.mar-hin: 158/860
   Tatoeba-test.multi-multi: 804/4689
   Tatoeba-test.urd-hin: 239/1512
   Tatoeba-test.zza-asm: 1/2
BLEU-scores:
   Tatoeba-test.asm-hin: 3.5
   Tatoeba-test.asm-zza: 12.4
   Tatoeba-test.hin-asm: 6.2
   Tatoeba-test.hin-mar: 27.0
   Tatoeba-test.hin-urd: 21.4
   Tatoeba-test.mar-hin: 13.4
   Tatoeba-test.multi-multi: 17.7
   Tatoeba-test.urd-hin: 13.4
   Tatoeba-test.zza-asm: 5.3
chr-F-scores:
   Tatoeba-test.asm-hin: 0.202
   Tatoeba-test.asm-zza: 0.014
   Tatoeba-test.hin-asm: 0.238
   Tatoeba-test.hin-mar: 0.560
   Tatoeba-test.hin-urd: 0.507
   Tatoeba-test.mar-hin: 0.463
   Tatoeba-test.multi-multi: 0.460
   Tatoeba-test.urd-hin: 0.363
   Tatoeba-test.zza-asm: 0.000