opus-2020-09-26.yml

release: sem-sem/opus-2020-09-26.zip
release-date: 2020-09-26
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:
   - acm
   - afb
   - amh
   - apc
   - ara
   - arq
   - ary
   - arz
   - eng
   - heb
   - mlt
   - phn
   - syc
   - tir
   - tmr
target-languages:
   - acm
   - afb
   - amh
   - apc
   - ara
   - arq
   - ary
   - arz
   - eng
   - heb
   - mlt
   - phn
   - syc
   - tir
   - tmr
use-target-labels:
   - ">>acm<<"
   - ">>afb<<"
   - ">>amh<<"
   - ">>apc<<"
   - ">>ara<<"
   - ">>arq<<"
   - ">>ary<<"
   - ">>arz<<"
   - ">>eng<<"
   - ">>heb<<"
   - ">>mlt<<"
   - ">>phn_Phnx<<"
   - ">>syc_Syrc<<"
   - ">>tir<<"
   - ">>tmr_Hebr<<"
test-data:
   Tatoeba-test.amh-eng: 190/802
   Tatoeba-test.ara-ara: 16/48
   Tatoeba-test.ara-eng: 10000/62449
   Tatoeba-test.ara-heb: 1208/5440
   Tatoeba-test.ara-mlt: 28/59
   Tatoeba-test.ara-tmr: 8/20
   Tatoeba-test.eng-amh: 190/554
   Tatoeba-test.eng-ara: 10000/50193
   Tatoeba-test.eng-heb: 10000/49113
   Tatoeba-test.eng-mlt: 203/673
   Tatoeba-test.eng-phn: 5/28
   Tatoeba-test.eng-tir: 69/305
   Tatoeba-test.eng-tmr: 19/76
   Tatoeba-test.heb-ara: 1208/5301
   Tatoeba-test.heb-eng: 10000/62332
   Tatoeba-test.heb-phn: 9/38
   Tatoeba-test.heb-syc: 1/5
   Tatoeba-test.heb-tmr: 16/79
   Tatoeba-test.mlt-ara: 28/67
   Tatoeba-test.mlt-eng: 203/935
   Tatoeba-test.multi-multi: 10000/54716
   Tatoeba-test.phn-eng: 5/39
   Tatoeba-test.phn-heb: 9/39
   Tatoeba-test.phn-tmr: 1/2
   Tatoeba-test.syc-heb: 1/5
   Tatoeba-test.tir-eng: 69/426
   Tatoeba-test.tmr-ara: 8/19
   Tatoeba-test.tmr-eng: 19/118
   Tatoeba-test.tmr-heb: 16/80
   Tatoeba-test.tmr-phn: 1/3
BLEU-scores:
   Tatoeba-test.amh-eng: 42.0
   Tatoeba-test.ara-ara: 2.7
   Tatoeba-test.ara-eng: 38.6
   Tatoeba-test.ara-heb: 34.9
   Tatoeba-test.ara-mlt: 24.3
   Tatoeba-test.ara-tmr: 2.7
   Tatoeba-test.eng-amh: 13.7
   Tatoeba-test.eng-ara: 12.2
   Tatoeba-test.eng-heb: 32.1
   Tatoeba-test.eng-mlt: 17.6
   Tatoeba-test.eng-phn: 1.3
   Tatoeba-test.eng-tir: 2.6
   Tatoeba-test.eng-tmr: 1.1
   Tatoeba-test.heb-ara: 19.5
   Tatoeba-test.heb-eng: 43.3
   Tatoeba-test.heb-phn: 2.0
   Tatoeba-test.heb-syc: 3.3
   Tatoeba-test.heb-tmr: 0.2
   Tatoeba-test.mlt-ara: 17.3
   Tatoeba-test.mlt-eng: 48.3
   Tatoeba-test.multi-multi: 33.2
   Tatoeba-test.phn-eng: 2.2
   Tatoeba-test.phn-heb: 0.4
   Tatoeba-test.phn-tmr: 0.5
   Tatoeba-test.syc-heb: 0.0
   Tatoeba-test.tir-eng: 16.1
   Tatoeba-test.tmr-ara: 2.5
   Tatoeba-test.tmr-eng: 2.2
   Tatoeba-test.tmr-heb: 1.0
   Tatoeba-test.tmr-phn: 0.0
chr-F-scores:
   Tatoeba-test.amh-eng: 0.593
   Tatoeba-test.ara-ara: 0.167
   Tatoeba-test.ara-eng: 0.564
   Tatoeba-test.ara-heb: 0.558
   Tatoeba-test.ara-mlt: 0.532
   Tatoeba-test.ara-tmr: 0.014
   Tatoeba-test.eng-amh: 0.510
   Tatoeba-test.eng-ara: 0.412
   Tatoeba-test.eng-heb: 0.550
   Tatoeba-test.eng-mlt: 0.556
   Tatoeba-test.eng-phn: 0.007
   Tatoeba-test.eng-tir: 0.250
   Tatoeba-test.eng-tmr: 0.007
   Tatoeba-test.heb-ara: 0.496
   Tatoeba-test.heb-eng: 0.598
   Tatoeba-test.heb-phn: 0.009
   Tatoeba-test.heb-syc: 0.000
   Tatoeba-test.heb-tmr: 0.005
   Tatoeba-test.mlt-ara: 0.427
   Tatoeba-test.mlt-eng: 0.647
   Tatoeba-test.multi-multi: 0.534
   Tatoeba-test.phn-eng: 0.071
   Tatoeba-test.phn-heb: 0.044
   Tatoeba-test.phn-tmr: 0.000
   Tatoeba-test.syc-heb: 0.000
   Tatoeba-test.tir-eng: 0.344
   Tatoeba-test.tmr-ara: 0.075
   Tatoeba-test.tmr-eng: 0.141
   Tatoeba-test.tmr-heb: 0.142
   Tatoeba-test.tmr-phn: 0.017