Skip to content

Latest commit

 

History

History
835 lines (811 loc) · 43.2 KB

README.md

File metadata and controls

835 lines (811 loc) · 43.2 KB
Raw

opus-2020-06-28.zip

Benchmarks

testset BLEU chr-F
Tatoeba-test.hye-eng.hye.eng 35.0 0.521
Tatoeba-test.multi.eng 48.1 0.629
Tatoeba-test.sqi-eng.sqi.eng 53.4 0.695

opus-2020-07-14.zip

  • dataset: opus
  • model: transformer
  • source language(s): afr aln ang_Latn arg asm ast awa bel bel_Latn ben bho bos_Latn bre bul bul_Latn cat ces cor cos csb_Latn cym dan deu dsb egl ell enm_Latn ext fao fra frm_Latn frr fry gcf_Latn gla gle glg glv gom gos got_Goth grc_Grek gsw guj hat hif_Latn hin hrv hsb hye ind isl ita jdt_Cyrl ksh kur_Arab kur_Latn lad lad_Latn lat_Latn lav lij lit lld_Latn lmo ltg ltz mai mar max_Latn mfe min mkd mwl nds nld nno nob nob_Hebr non_Latn npi oci ori orv_Cyrl oss pan_Guru pap pdc pes pes_Latn pes_Thaa pms pnb pol por prg_Latn pus roh rom ron rue rus san_Deva scn sco sgs sin slv snd_Arab spa sqi srp_Cyrl srp_Latn stq swe swg tgk_Cyrl tly_Latn tmw_Latn ukr urd vec wln yid zlm_Latn zsm_Latn zza
  • target language(s): eng
  • model: transformer
  • pre-processing: normalization + SentencePiece (spm32k,spm32k)
  • download: opus-2020-07-14.zip
  • test set translations: opus-2020-07-14.test.txt
  • test set scores: opus-2020-07-14.eval.txt

Benchmarks

testset BLEU chr-F
Tatoeba-test.afr-eng.afr.eng 58.1 0.719
Tatoeba-test.ang-eng.ang.eng 9.5 0.257
Tatoeba-test.arg-eng.arg.eng 36.0 0.467
Tatoeba-test.asm-eng.asm.eng 21.4 0.384
Tatoeba-test.ast-eng.ast.eng 35.5 0.511
Tatoeba-test.awa-eng.awa.eng 11.7 0.280
Tatoeba-test.bel-eng.bel.eng 40.4 0.581
Tatoeba-test.ben-eng.ben.eng 44.8 0.588
Tatoeba-test.bho-eng.bho.eng 32.7 0.515
Tatoeba-test.bre-eng.bre.eng 22.0 0.386
Tatoeba-test.bul-eng.bul.eng 50.2 0.656
Tatoeba-test.cat-eng.cat.eng 51.1 0.675
Tatoeba-test.ces-eng.ces.eng 46.3 0.631
Tatoeba-test.cor-eng.cor.eng 5.0 0.202
Tatoeba-test.cos-eng.cos.eng 57.5 0.624
Tatoeba-test.csb-eng.csb.eng 22.6 0.405
Tatoeba-test.cym-eng.cym.eng 33.8 0.513
Tatoeba-test.dan-eng.dan.eng 55.5 0.702
Tatoeba-test.deu-eng.deu.eng 43.7 0.614
Tatoeba-test.dsb-eng.dsb.eng 16.4 0.358
Tatoeba-test.egl-eng.egl.eng 4.4 0.189
Tatoeba-test.ell-eng.ell.eng 61.0 0.733
Tatoeba-test.enm-eng.enm.eng 24.3 0.479
Tatoeba-test.ext-eng.ext.eng 27.9 0.429
Tatoeba-test.fao-eng.fao.eng 29.0 0.467
Tatoeba-test.fas-eng.fas.eng 9.4 0.302
Tatoeba-test.fra-eng.fra.eng 49.3 0.653
Tatoeba-test.frm-eng.frm.eng 28.0 0.439
Tatoeba-test.frr-eng.frr.eng 4.8 0.120
Tatoeba-test.fry-eng.fry.eng 26.9 0.476
Tatoeba-test.gcf-eng.gcf.eng 17.4 0.294
Tatoeba-test.gla-eng.gla.eng 21.7 0.377
Tatoeba-test.gle-eng.gle.eng 51.0 0.662
Tatoeba-test.glg-eng.glg.eng 49.1 0.657
Tatoeba-test.glv-eng.glv.eng 3.2 0.179
Tatoeba-test.gos-eng.gos.eng 17.5 0.359
Tatoeba-test.got-eng.got.eng 0.0 0.020
Tatoeba-test.grc-eng.grc.eng 1.4 0.131
Tatoeba-test.gsw-eng.gsw.eng 23.9 0.393
Tatoeba-test.guj-eng.guj.eng 26.6 0.407
Tatoeba-test.hat-eng.hat.eng 42.8 0.566
Tatoeba-test.hif-eng.hif.eng 23.4 0.456
Tatoeba-test.hin-eng.hin.eng 40.5 0.574
Tatoeba-test.hsb-eng.hsb.eng 46.7 0.596
Tatoeba-test.hye-eng.hye.eng 36.7 0.541
Tatoeba-test.isl-eng.isl.eng 42.8 0.593
Tatoeba-test.ita-eng.ita.eng 59.9 0.728
Tatoeba-test.jdt-eng.jdt.eng 3.0 0.087
Tatoeba-test.kok-eng.kok.eng 4.0 0.229
Tatoeba-test.ksh-eng.ksh.eng 11.7 0.283
Tatoeba-test.kur-eng.kur.eng 9.1 0.211
Tatoeba-test.lad-eng.lad.eng 30.2 0.475
Tatoeba-test.lah-eng.lah.eng 13.8 0.309
Tatoeba-test.lat-eng.lat.eng 23.7 0.424
Tatoeba-test.lav-eng.lav.eng 46.3 0.635
Tatoeba-test.lij-eng.lij.eng 9.1 0.282
Tatoeba-test.lit-eng.lit.eng 50.2 0.656
Tatoeba-test.lld-eng.lld.eng 21.0 0.361
Tatoeba-test.lmo-eng.lmo.eng 16.4 0.304
Tatoeba-test.ltz-eng.ltz.eng 36.3 0.523
Tatoeba-test.mai-eng.mai.eng 37.0 0.614
Tatoeba-test.mar-eng.mar.eng 30.1 0.538
Tatoeba-test.mfe-eng.mfe.eng 63.1 0.810
Tatoeba-test.mkd-eng.mkd.eng 48.6 0.629
Tatoeba-test.multi.eng 45.1 0.609
Tatoeba-test.mwl-eng.mwl.eng 28.1 0.624
Tatoeba-test.nds-eng.nds.eng 33.1 0.522
Tatoeba-test.nep-eng.nep.eng 2.4 0.174
Tatoeba-test.nld-eng.nld.eng 53.4 0.688
Tatoeba-test.non-eng.non.eng 39.5 0.565
Tatoeba-test.oci-eng.oci.eng 19.4 0.389
Tatoeba-test.ori-eng.ori.eng 15.5 0.326
Tatoeba-test.orv-eng.orv.eng 11.5 0.262
Tatoeba-test.oss-eng.oss.eng 4.4 0.268
Tatoeba-test.pan-eng.pan.eng 17.9 0.386
Tatoeba-test.pap-eng.pap.eng 58.1 0.697
Tatoeba-test.pdc-eng.pdc.eng 42.2 0.533
Tatoeba-test.pms-eng.pms.eng 11.2 0.322
Tatoeba-test.pol-eng.pol.eng 45.7 0.626
Tatoeba-test.por-eng.por.eng 56.1 0.707
Tatoeba-test.prg-eng.prg.eng 3.9 0.203
Tatoeba-test.pus-eng.pus.eng 2.3 0.199
Tatoeba-test.roh-eng.roh.eng 20.2 0.420
Tatoeba-test.rom-eng.rom.eng 4.7 0.230
Tatoeba-test.ron-eng.ron.eng 50.7 0.666
Tatoeba-test.rue-eng.rue.eng 26.0 0.447
Tatoeba-test.rus-eng.rus.eng 46.9 0.626
Tatoeba-test.san-eng.san.eng 4.3 0.202
Tatoeba-test.scn-eng.scn.eng 35.9 0.389
Tatoeba-test.sco-eng.sco.eng 41.9 0.584
Tatoeba-test.sgs-eng.sgs.eng 15.6 0.310
Tatoeba-test.sin-eng.sin.eng 31.0 0.505
Tatoeba-test.slv-eng.slv.eng 37.6 0.558
Tatoeba-test.snd-eng.snd.eng 15.9 0.409
Tatoeba-test.spa-eng.spa.eng 52.3 0.682
Tatoeba-test.sqi-eng.sqi.eng 51.4 0.675
Tatoeba-test.stq-eng.stq.eng 8.3 0.344
Tatoeba-test.swe-eng.swe.eng 56.6 0.699
Tatoeba-test.swg-eng.swg.eng 15.9 0.326
Tatoeba-test.tgk-eng.tgk.eng 27.8 0.467
Tatoeba-test.tly-eng.tly.eng 1.1 0.088
Tatoeba-test.ukr-eng.ukr.eng 47.6 0.635
Tatoeba-test.urd-eng.urd.eng 26.8 0.460
Tatoeba-test.vec-eng.vec.eng 36.0 0.434
Tatoeba-test.wln-eng.wln.eng 12.0 0.284
Tatoeba-test.yid-eng.yid.eng 22.4 0.420
Tatoeba-test.zza-eng.zza.eng 1.4 0.124

opus-2020-07-19.zip

  • dataset: opus
  • model: transformer
  • source language(s): afr aln ang_Latn arg asm ast awa bel bel_Latn ben bho bos_Latn bre bul bul_Latn cat ces cor cos csb_Latn cym dan deu dsb egl ell enm_Latn ext fao fra frm_Latn frr fry gcf_Latn gla gle glg glv gom gos got_Goth grc_Grek gsw guj hat hif_Latn hin hrv hsb hye ind isl ita jdt_Cyrl ksh kur_Arab kur_Latn lad lad_Latn lat_Latn lav lij lit lld_Latn lmo ltg ltz mai mar max_Latn mfe min mkd mwl nds nld nno nob nob_Hebr non_Latn npi oci ori orv_Cyrl oss pan_Guru pap pdc pes pes_Latn pes_Thaa pms pnb pol por prg_Latn pus roh rom ron rue rus san_Deva scn sco sgs sin slv snd_Arab spa sqi srp_Cyrl srp_Latn stq swe swg tgk_Cyrl tly_Latn tmw_Latn ukr urd vec wln yid zlm_Latn zsm_Latn zza
  • target language(s): eng
  • model: transformer
  • pre-processing: normalization + SentencePiece (spm32k,spm32k)
  • download: opus-2020-07-19.zip
  • test set translations: opus-2020-07-19.test.txt
  • test set scores: opus-2020-07-19.eval.txt

Benchmarks

testset BLEU chr-F
Tatoeba-test.afr-eng.afr.eng 58.3 0.720
Tatoeba-test.ang-eng.ang.eng 9.0 0.247
Tatoeba-test.arg-eng.arg.eng 31.4 0.461
Tatoeba-test.asm-eng.asm.eng 21.9 0.390
Tatoeba-test.ast-eng.ast.eng 34.6 0.498
Tatoeba-test.awa-eng.awa.eng 10.9 0.280
Tatoeba-test.bel-eng.bel.eng 41.3 0.590
Tatoeba-test.ben-eng.ben.eng 45.2 0.593
Tatoeba-test.bho-eng.bho.eng 30.0 0.504
Tatoeba-test.bre-eng.bre.eng 22.0 0.392
Tatoeba-test.bul-eng.bul.eng 50.4 0.658
Tatoeba-test.cat-eng.cat.eng 51.3 0.677
Tatoeba-test.ces-eng.ces.eng 46.8 0.636
Tatoeba-test.cor-eng.cor.eng 4.9 0.198
Tatoeba-test.cos-eng.cos.eng 53.3 0.616
Tatoeba-test.csb-eng.csb.eng 22.3 0.399
Tatoeba-test.cym-eng.cym.eng 33.8 0.514
Tatoeba-test.dan-eng.dan.eng 55.9 0.706
Tatoeba-test.deu-eng.deu.eng 44.2 0.619
Tatoeba-test.dsb-eng.dsb.eng 23.3 0.445
Tatoeba-test.egl-eng.egl.eng 4.7 0.204
Tatoeba-test.ell-eng.ell.eng 61.6 0.738
Tatoeba-test.enm-eng.enm.eng 22.1 0.477
Tatoeba-test.ext-eng.ext.eng 16.9 0.424
Tatoeba-test.fao-eng.fao.eng 27.7 0.464
Tatoeba-test.fas-eng.fas.eng 8.8 0.301
Tatoeba-test.fra-eng.fra.eng 49.9 0.658
Tatoeba-test.frm-eng.frm.eng 25.5 0.431
Tatoeba-test.frr-eng.frr.eng 3.9 0.068
Tatoeba-test.fry-eng.fry.eng 28.0 0.482
Tatoeba-test.gcf-eng.gcf.eng 15.6 0.295
Tatoeba-test.gla-eng.gla.eng 22.5 0.384
Tatoeba-test.gle-eng.gle.eng 52.1 0.671
Tatoeba-test.glg-eng.glg.eng 49.7 0.661
Tatoeba-test.glv-eng.glv.eng 9.9 0.269
Tatoeba-test.gos-eng.gos.eng 17.7 0.361
Tatoeba-test.got-eng.got.eng 0.1 0.031
Tatoeba-test.grc-eng.grc.eng 1.6 0.133
Tatoeba-test.gsw-eng.gsw.eng 20.5 0.379
Tatoeba-test.guj-eng.guj.eng 25.5 0.406
Tatoeba-test.hat-eng.hat.eng 42.7 0.586
Tatoeba-test.hbs-eng.hbs.eng 51.0 0.663
Tatoeba-test.hif-eng.hif.eng 19.3 0.441
Tatoeba-test.hin-eng.hin.eng 41.0 0.579
Tatoeba-test.hsb-eng.hsb.eng 48.2 0.606
Tatoeba-test.hye-eng.hye.eng 38.0 0.548
Tatoeba-test.isl-eng.isl.eng 43.1 0.599
Tatoeba-test.ita-eng.ita.eng 60.3 0.732
Tatoeba-test.jdt-eng.jdt.eng 4.3 0.114
Tatoeba-test.kok-eng.kok.eng 5.7 0.249
Tatoeba-test.ksh-eng.ksh.eng 11.7 0.286
Tatoeba-test.kur-eng.kur.eng 9.6 0.221
Tatoeba-test.lad-eng.lad.eng 29.0 0.473
Tatoeba-test.lah-eng.lah.eng 11.8 0.331
Tatoeba-test.lat-eng.lat.eng 24.0 0.427
Tatoeba-test.lav-eng.lav.eng 47.1 0.644
Tatoeba-test.lij-eng.lij.eng 7.2 0.293
Tatoeba-test.lit-eng.lit.eng 51.3 0.667
Tatoeba-test.lld-eng.lld.eng 21.4 0.367
Tatoeba-test.lmo-eng.lmo.eng 35.7 0.417
Tatoeba-test.ltz-eng.ltz.eng 35.1 0.512
Tatoeba-test.mai-eng.mai.eng 74.0 0.769
Tatoeba-test.mar-eng.mar.eng 31.2 0.550
Tatoeba-test.mfe-eng.mfe.eng 63.1 0.805
Tatoeba-test.mkd-eng.mkd.eng 49.2 0.635
Tatoeba-test.msa-eng.msa.eng 38.4 0.564
Tatoeba-test.multi.eng 45.3 0.613
Tatoeba-test.mwl-eng.mwl.eng 28.1 0.624
Tatoeba-test.nds-eng.nds.eng 33.8 0.528
Tatoeba-test.nep-eng.nep.eng 4.5 0.193
Tatoeba-test.nld-eng.nld.eng 53.7 0.691
Tatoeba-test.non-eng.non.eng 39.4 0.533
Tatoeba-test.nor-eng.nor.eng 50.5 0.660
Tatoeba-test.oci-eng.oci.eng 19.1 0.386
Tatoeba-test.ori-eng.ori.eng 16.8 0.345
Tatoeba-test.orv-eng.orv.eng 11.6 0.270
Tatoeba-test.oss-eng.oss.eng 7.0 0.267
Tatoeba-test.pan-eng.pan.eng 19.7 0.388
Tatoeba-test.pap-eng.pap.eng 58.0 0.693
Tatoeba-test.pdc-eng.pdc.eng 38.8 0.499
Tatoeba-test.pms-eng.pms.eng 12.8 0.333
Tatoeba-test.pol-eng.pol.eng 46.3 0.631
Tatoeba-test.por-eng.por.eng 56.3 0.709
Tatoeba-test.prg-eng.prg.eng 3.1 0.198
Tatoeba-test.pus-eng.pus.eng 2.5 0.202
Tatoeba-test.roh-eng.roh.eng 19.3 0.422
Tatoeba-test.rom-eng.rom.eng 4.6 0.237
Tatoeba-test.ron-eng.ron.eng 50.9 0.667
Tatoeba-test.rue-eng.rue.eng 26.1 0.450
Tatoeba-test.rus-eng.rus.eng 47.3 0.629
Tatoeba-test.san-eng.san.eng 5.7 0.206
Tatoeba-test.scn-eng.scn.eng 44.4 0.437
Tatoeba-test.sco-eng.sco.eng 45.8 0.592
Tatoeba-test.sgs-eng.sgs.eng 29.2 0.374
Tatoeba-test.sin-eng.sin.eng 25.6 0.481
Tatoeba-test.slv-eng.slv.eng 37.8 0.560
Tatoeba-test.snd-eng.snd.eng 15.9 0.409
Tatoeba-test.spa-eng.spa.eng 52.8 0.684
Tatoeba-test.sqi-eng.sqi.eng 52.5 0.681
Tatoeba-test.stq-eng.stq.eng 18.6 0.445
Tatoeba-test.swe-eng.swe.eng 57.1 0.704
Tatoeba-test.swg-eng.swg.eng 19.0 0.342
Tatoeba-test.tgk-eng.tgk.eng 28.1 0.447
Tatoeba-test.tly-eng.tly.eng 1.3 0.110
Tatoeba-test.ukr-eng.ukr.eng 48.0 0.638
Tatoeba-test.urd-eng.urd.eng 27.5 0.470
Tatoeba-test.vec-eng.vec.eng 42.5 0.458
Tatoeba-test.wln-eng.wln.eng 10.8 0.264
Tatoeba-test.yid-eng.yid.eng 22.9 0.422
Tatoeba-test.zza-eng.zza.eng 1.5 0.129

opus-2020-07-26.zip

  • dataset: opus
  • model: transformer
  • source language(s): afr aln ang_Latn arg asm ast awa bel bel_Latn ben bho bos_Latn bre bul bul_Latn cat ces cor cos csb_Latn cym dan deu dsb egl ell enm_Latn ext fao fra frm_Latn frr fry gcf_Latn gla gle glg glv gom gos got_Goth grc_Grek gsw guj hat hif_Latn hin hrv hsb hye ind isl ita jdt_Cyrl ksh kur_Arab kur_Latn lad lad_Latn lat_Latn lav lij lit lld_Latn lmo ltg ltz mai mar max_Latn mfe min mkd mwl nds nld nno nob nob_Hebr non_Latn npi oci ori orv_Cyrl oss pan_Guru pap pdc pes pes_Latn pes_Thaa pms pnb pol por prg_Latn pus roh rom ron rue rus san_Deva scn sco sgs sin slv snd_Arab spa sqi srp_Cyrl srp_Latn stq swe swg tgk_Cyrl tly_Latn tmw_Latn ukr urd vec wln yid zlm_Latn zsm_Latn zza
  • target language(s): eng
  • model: transformer
  • pre-processing: normalization + SentencePiece (spm32k,spm32k)
  • download: opus-2020-07-26.zip
  • test set translations: opus-2020-07-26.test.txt
  • test set scores: opus-2020-07-26.eval.txt

Benchmarks

testset BLEU chr-F
newsdev2014-hineng.hin.eng 10.5 0.374
newsdev2016-enro-roneng.ron.eng 34.8 0.611
newsdev2017-enlv-laveng.lav.eng 24.3 0.536
newsdev2019-engu-gujeng.guj.eng 13.7 0.406
newsdev2019-enlt-liteng.lit.eng 23.6 0.518
newsdiscussdev2015-enfr-fraeng.fra.eng 29.6 0.545
newsdiscusstest2015-enfr-fraeng.fra.eng 32.8 0.568
newssyscomb2009-ceseng.ces.eng 23.8 0.514
newssyscomb2009-deueng.deu.eng 25.5 0.523
newssyscomb2009-fraeng.fra.eng 27.8 0.550
newssyscomb2009-itaeng.ita.eng 30.8 0.570
newssyscomb2009-spaeng.spa.eng 28.4 0.557
news-test2008-deueng.deu.eng 23.6 0.512
news-test2008-fraeng.fra.eng 24.1 0.522
news-test2008-spaeng.spa.eng 25.5 0.532
newstest2009-ceseng.ces.eng 23.2 0.514
newstest2009-deueng.deu.eng 22.9 0.509
newstest2009-fraeng.fra.eng 26.9 0.543
newstest2009-itaeng.ita.eng 30.1 0.565
newstest2009-spaeng.spa.eng 27.8 0.549
newstest2010-ceseng.ces.eng 23.5 0.521
newstest2010-deueng.deu.eng 25.8 0.542
newstest2010-fraeng.fra.eng 28.2 0.558
newstest2010-spaeng.spa.eng 31.4 0.581
newstest2011-ceseng.ces.eng 24.1 0.520
newstest2011-deueng.deu.eng 23.7 0.520
newstest2011-fraeng.fra.eng 29.2 0.568
newstest2011-spaeng.spa.eng 30.1 0.568
newstest2012-ceseng.ces.eng 22.9 0.513
newstest2012-deueng.deu.eng 24.7 0.528
newstest2012-fraeng.fra.eng 29.8 0.567
newstest2012-ruseng.rus.eng 29.5 0.562
newstest2012-spaeng.spa.eng 33.1 0.590
newstest2013-ceseng.ces.eng 25.1 0.527
newstest2013-deueng.deu.eng 27.4 0.542
newstest2013-fraeng.fra.eng 29.9 0.559
newstest2013-ruseng.rus.eng 23.5 0.510
newstest2013-spaeng.spa.eng 30.7 0.576
newstest2014-csen-ceseng.ces.eng 27.4 0.554
newstest2014-deen-deueng.deu.eng 27.5 0.542
newstest2014-fren-fraeng.fra.eng 32.0 0.590
newstest2014-hien-hineng.hin.eng 15.9 0.444
newstest2014-ruen-ruseng.rus.eng 27.1 0.549
newstest2015-encs-ceseng.ces.eng 25.0 0.516
newstest2015-ende-deueng.deu.eng 28.7 0.549
newstest2015-enru-ruseng.rus.eng 25.5 0.523
newstest2016-encs-ceseng.ces.eng 26.6 0.536
newstest2016-ende-deueng.deu.eng 32.9 0.590
newstest2016-enro-roneng.ron.eng 32.6 0.589
newstest2016-enru-ruseng.rus.eng 24.9 0.523
newstest2017-encs-ceseng.ces.eng 23.1 0.506
newstest2017-ende-deueng.deu.eng 29.1 0.554
newstest2017-enlv-laveng.lav.eng 18.5 0.482
newstest2017-enru-ruseng.rus.eng 27.9 0.543
newstest2018-encs-ceseng.ces.eng 24.4 0.519
newstest2018-ende-deueng.deu.eng 35.5 0.601
newstest2018-enru-ruseng.rus.eng 23.5 0.514
newstest2019-deen-deueng.deu.eng 31.2 0.566
newstest2019-guen-gujeng.guj.eng 10.2 0.360
newstest2019-lten-liteng.lit.eng 26.5 0.555
newstest2019-ruen-ruseng.rus.eng 27.2 0.536
Tatoeba-test.afr-eng.afr.eng 58.1 0.719
Tatoeba-test.ang-eng.ang.eng 9.2 0.250
Tatoeba-test.arg-eng.arg.eng 37.9 0.493
Tatoeba-test.asm-eng.asm.eng 20.5 0.382
Tatoeba-test.ast-eng.ast.eng 35.6 0.507
Tatoeba-test.awa-eng.awa.eng 10.0 0.285
Tatoeba-test.bel-eng.bel.eng 41.0 0.588
Tatoeba-test.ben-eng.ben.eng 45.9 0.599
Tatoeba-test.bho-eng.bho.eng 31.1 0.513
Tatoeba-test.bre-eng.bre.eng 22.3 0.386
Tatoeba-test.bul-eng.bul.eng 50.5 0.658
Tatoeba-test.cat-eng.cat.eng 51.3 0.677
Tatoeba-test.ces-eng.ces.eng 46.7 0.634
Tatoeba-test.cor-eng.cor.eng 5.2 0.204
Tatoeba-test.cos-eng.cos.eng 53.3 0.615
Tatoeba-test.csb-eng.csb.eng 20.0 0.379
Tatoeba-test.cym-eng.cym.eng 34.4 0.520
Tatoeba-test.dan-eng.dan.eng 55.9 0.705
Tatoeba-test.deu-eng.deu.eng 44.3 0.619
Tatoeba-test.dsb-eng.dsb.eng 23.2 0.417
Tatoeba-test.egl-eng.egl.eng 3.7 0.200
Tatoeba-test.ell-eng.ell.eng 61.9 0.740
Tatoeba-test.enm-eng.enm.eng 21.9 0.473
Tatoeba-test.ext-eng.ext.eng 15.5 0.426
Tatoeba-test.fao-eng.fao.eng 28.3 0.469
Tatoeba-test.fas-eng.fas.eng 9.4 0.305
Tatoeba-test.fra-eng.fra.eng 50.1 0.659
Tatoeba-test.frm-eng.frm.eng 29.4 0.448
Tatoeba-test.frr-eng.frr.eng 4.8 0.176
Tatoeba-test.fry-eng.fry.eng 27.4 0.473
Tatoeba-test.gcf-eng.gcf.eng 18.0 0.310
Tatoeba-test.gla-eng.gla.eng 22.8 0.383
Tatoeba-test.gle-eng.gle.eng 51.5 0.666
Tatoeba-test.glg-eng.glg.eng 48.8 0.657
Tatoeba-test.glv-eng.glv.eng 3.0 0.188
Tatoeba-test.gos-eng.gos.eng 18.6 0.369
Tatoeba-test.got-eng.got.eng 0.2 0.039
Tatoeba-test.grc-eng.grc.eng 1.5 0.133
Tatoeba-test.gsw-eng.gsw.eng 19.9 0.370
Tatoeba-test.guj-eng.guj.eng 25.2 0.415
Tatoeba-test.hat-eng.hat.eng 40.7 0.554
Tatoeba-test.hbs-eng.hbs.eng 50.9 0.661
Tatoeba-test.hif-eng.hif.eng 19.3 0.443
Tatoeba-test.hin-eng.hin.eng 41.2 0.581
Tatoeba-test.hsb-eng.hsb.eng 44.8 0.575
Tatoeba-test.hye-eng.hye.eng 38.0 0.554
Tatoeba-test.isl-eng.isl.eng 43.3 0.601
Tatoeba-test.ita-eng.ita.eng 60.5 0.732
Tatoeba-test.jdt-eng.jdt.eng 4.3 0.114
Tatoeba-test.kok-eng.kok.eng 4.8 0.233
Tatoeba-test.ksh-eng.ksh.eng 10.2 0.269
Tatoeba-test.kur-eng.kur.eng 10.1 0.218
Tatoeba-test.lad-eng.lad.eng 30.2 0.480
Tatoeba-test.lah-eng.lah.eng 11.8 0.311
Tatoeba-test.lat-eng.lat.eng 23.8 0.426
Tatoeba-test.lav-eng.lav.eng 46.8 0.641
Tatoeba-test.lij-eng.lij.eng 7.4 0.281
Tatoeba-test.lit-eng.lit.eng 51.0 0.665
Tatoeba-test.lld-eng.lld.eng 20.1 0.351
Tatoeba-test.lmo-eng.lmo.eng 28.0 0.384
Tatoeba-test.ltz-eng.ltz.eng 34.5 0.519
Tatoeba-test.mai-eng.mai.eng 74.0 0.769
Tatoeba-test.mar-eng.mar.eng 30.9 0.548
Tatoeba-test.mfe-eng.mfe.eng 63.1 0.805
Tatoeba-test.mkd-eng.mkd.eng 49.2 0.636
Tatoeba-test.msa-eng.msa.eng 38.5 0.565
Tatoeba-test.multi.eng 45.4 0.613
Tatoeba-test.mwl-eng.mwl.eng 18.7 0.473
Tatoeba-test.nds-eng.nds.eng 34.2 0.529
Tatoeba-test.nep-eng.nep.eng 3.5 0.192
Tatoeba-test.nld-eng.nld.eng 53.7 0.690
Tatoeba-test.non-eng.non.eng 39.4 0.534
Tatoeba-test.nor-eng.nor.eng 50.5 0.660
Tatoeba-test.oci-eng.oci.eng 18.6 0.384
Tatoeba-test.ori-eng.ori.eng 18.2 0.353
Tatoeba-test.orv-eng.orv.eng 11.6 0.271
Tatoeba-test.oss-eng.oss.eng 5.1 0.260
Tatoeba-test.pan-eng.pan.eng 18.6 0.395
Tatoeba-test.pap-eng.pap.eng 60.0 0.703
Tatoeba-test.pdc-eng.pdc.eng 40.7 0.512
Tatoeba-test.pms-eng.pms.eng 13.1 0.338
Tatoeba-test.pol-eng.pol.eng 46.2 0.630
Tatoeba-test.por-eng.por.eng 56.4 0.710
Tatoeba-test.prg-eng.prg.eng 3.4 0.198
Tatoeba-test.pus-eng.pus.eng 3.1 0.226
Tatoeba-test.roh-eng.roh.eng 20.0 0.414
Tatoeba-test.rom-eng.rom.eng 4.6 0.235
Tatoeba-test.ron-eng.ron.eng 51.1 0.668
Tatoeba-test.rue-eng.rue.eng 24.1 0.430
Tatoeba-test.rus-eng.rus.eng 47.4 0.630
Tatoeba-test.san-eng.san.eng 2.3 0.194
Tatoeba-test.scn-eng.scn.eng 46.1 0.434
Tatoeba-test.sco-eng.sco.eng 43.3 0.584
Tatoeba-test.sgs-eng.sgs.eng 29.5 0.395
Tatoeba-test.sin-eng.sin.eng 27.1 0.472
Tatoeba-test.slv-eng.slv.eng 38.2 0.564
Tatoeba-test.snd-eng.snd.eng 15.9 0.409
Tatoeba-test.spa-eng.spa.eng 52.8 0.685
Tatoeba-test.sqi-eng.sqi.eng 52.2 0.682
Tatoeba-test.stq-eng.stq.eng 21.3 0.389
Tatoeba-test.swe-eng.swe.eng 57.1 0.703
Tatoeba-test.swg-eng.swg.eng 17.0 0.322
Tatoeba-test.tgk-eng.tgk.eng 35.3 0.496
Tatoeba-test.tly-eng.tly.eng 0.9 0.094
Tatoeba-test.ukr-eng.ukr.eng 47.9 0.638
Tatoeba-test.urd-eng.urd.eng 27.0 0.465
Tatoeba-test.vec-eng.vec.eng 35.4 0.421
Tatoeba-test.wln-eng.wln.eng 11.9 0.278
Tatoeba-test.yid-eng.yid.eng 23.4 0.426
Tatoeba-test.zza-eng.zza.eng 1.4 0.127

opus2m-2020-08-01.zip

  • dataset: opus2m
  • model: transformer
  • source language(s): afr aln ang_Latn arg asm ast awa bel bel_Latn ben bho bos_Latn bre bul bul_Latn cat ces cor cos csb_Latn cym dan deu dsb egl ell enm_Latn ext fao fra frm_Latn frr fry gcf_Latn gla gle glg glv gom gos got_Goth grc_Grek gsw guj hat hif_Latn hin hrv hsb hye ind isl ita jdt_Cyrl ksh kur_Arab kur_Latn lad lad_Latn lat_Latn lav lij lit lld_Latn lmo ltg ltz mai mar max_Latn mfe min mkd mwl nds nld nno nob nob_Hebr non_Latn npi oci ori orv_Cyrl oss pan_Guru pap pdc pes pes_Latn pes_Thaa pms pnb pol por prg_Latn pus roh rom ron rue rus san_Deva scn sco sgs sin slv snd_Arab spa sqi srp_Cyrl srp_Latn stq swe swg tgk_Cyrl tly_Latn tmw_Latn ukr urd vec wln yid zlm_Latn zsm_Latn zza
  • target language(s): eng
  • model: transformer
  • pre-processing: normalization + SentencePiece (spm32k,spm32k)
  • download: opus2m-2020-08-01.zip
  • test set translations: opus2m-2020-08-01.test.txt
  • test set scores: opus2m-2020-08-01.eval.txt

Benchmarks

testset BLEU chr-F
newsdev2014-hineng.hin.eng 11.2 0.375
newsdev2016-enro-roneng.ron.eng 35.5 0.614
newsdev2017-enlv-laveng.lav.eng 25.1 0.542
newsdev2019-engu-gujeng.guj.eng 16.0 0.420
newsdev2019-enlt-liteng.lit.eng 24.0 0.522
newsdiscussdev2015-enfr-fraeng.fra.eng 30.1 0.550
newsdiscusstest2015-enfr-fraeng.fra.eng 33.4 0.572
newssyscomb2009-ceseng.ces.eng 24.0 0.520
newssyscomb2009-deueng.deu.eng 25.7 0.526
newssyscomb2009-fraeng.fra.eng 27.9 0.550
newssyscomb2009-itaeng.ita.eng 31.4 0.574
newssyscomb2009-spaeng.spa.eng 28.3 0.555
news-test2008-deueng.deu.eng 24.0 0.515
news-test2008-fraeng.fra.eng 24.5 0.524
news-test2008-spaeng.spa.eng 25.5 0.533
newstest2009-ceseng.ces.eng 23.3 0.516
newstest2009-deueng.deu.eng 23.2 0.512
newstest2009-fraeng.fra.eng 27.3 0.545
newstest2009-itaeng.ita.eng 30.3 0.567
newstest2009-spaeng.spa.eng 27.9 0.549
newstest2010-ceseng.ces.eng 23.8 0.523
newstest2010-deueng.deu.eng 26.2 0.545
newstest2010-fraeng.fra.eng 28.6 0.562
newstest2010-spaeng.spa.eng 31.4 0.581
newstest2011-ceseng.ces.eng 24.2 0.521
newstest2011-deueng.deu.eng 23.9 0.522
newstest2011-fraeng.fra.eng 29.5 0.570
newstest2011-spaeng.spa.eng 30.3 0.570
newstest2012-ceseng.ces.eng 23.5 0.516
newstest2012-deueng.deu.eng 24.9 0.529
newstest2012-fraeng.fra.eng 30.0 0.568
newstest2012-ruseng.rus.eng 29.9 0.565
newstest2012-spaeng.spa.eng 33.3 0.593
newstest2013-ceseng.ces.eng 25.6 0.531
newstest2013-deueng.deu.eng 27.7 0.545
newstest2013-fraeng.fra.eng 30.0 0.561
newstest2013-ruseng.rus.eng 24.4 0.514
newstest2013-spaeng.spa.eng 30.8 0.577
newstest2014-csen-ceseng.ces.eng 27.7 0.558
newstest2014-deen-deueng.deu.eng 27.7 0.545
newstest2014-fren-fraeng.fra.eng 32.2 0.592
newstest2014-hien-hineng.hin.eng 16.7 0.450
newstest2014-ruen-ruseng.rus.eng 27.2 0.552
newstest2015-encs-ceseng.ces.eng 25.4 0.518
newstest2015-ende-deueng.deu.eng 28.8 0.552
newstest2015-enru-ruseng.rus.eng 25.6 0.527
newstest2016-encs-ceseng.ces.eng 27.0 0.540
newstest2016-ende-deueng.deu.eng 33.5 0.592
newstest2016-enro-roneng.ron.eng 32.8 0.591
newstest2016-enru-ruseng.rus.eng 24.8 0.523
newstest2017-encs-ceseng.ces.eng 23.7 0.510
newstest2017-ende-deueng.deu.eng 29.3 0.556
newstest2017-enlv-laveng.lav.eng 18.9 0.486
newstest2017-enru-ruseng.rus.eng 28.0 0.546
newstest2018-encs-ceseng.ces.eng 24.9 0.521
newstest2018-ende-deueng.deu.eng 36.0 0.604
newstest2018-enru-ruseng.rus.eng 23.8 0.517
newstest2019-deen-deueng.deu.eng 31.5 0.570
newstest2019-guen-gujeng.guj.eng 12.1 0.377
newstest2019-lten-liteng.lit.eng 26.6 0.555
newstest2019-ruen-ruseng.rus.eng 27.5 0.541
Tatoeba-test.afr-eng.afr.eng 59.0 0.724
Tatoeba-test.ang-eng.ang.eng 9.9 0.254
Tatoeba-test.arg-eng.arg.eng 41.6 0.487
Tatoeba-test.asm-eng.asm.eng 22.8 0.392
Tatoeba-test.ast-eng.ast.eng 36.1 0.521
Tatoeba-test.awa-eng.awa.eng 11.6 0.280
Tatoeba-test.bel-eng.bel.eng 42.2 0.597
Tatoeba-test.ben-eng.ben.eng 45.8 0.598
Tatoeba-test.bho-eng.bho.eng 34.4 0.518
Tatoeba-test.bre-eng.bre.eng 24.4 0.405
Tatoeba-test.bul-eng.bul.eng 50.8 0.660
Tatoeba-test.cat-eng.cat.eng 51.2 0.677
Tatoeba-test.ces-eng.ces.eng 47.6 0.641
Tatoeba-test.cor-eng.cor.eng 5.4 0.214
Tatoeba-test.cos-eng.cos.eng 61.0 0.675
Tatoeba-test.csb-eng.csb.eng 22.5 0.394
Tatoeba-test.cym-eng.cym.eng 34.7 0.522
Tatoeba-test.dan-eng.dan.eng 56.2 0.708
Tatoeba-test.deu-eng.deu.eng 44.9 0.625
Tatoeba-test.dsb-eng.dsb.eng 21.0 0.383
Tatoeba-test.egl-eng.egl.eng 6.9 0.221
Tatoeba-test.ell-eng.ell.eng 62.1 0.741
Tatoeba-test.enm-eng.enm.eng 22.6 0.466
Tatoeba-test.ext-eng.ext.eng 33.2 0.496
Tatoeba-test.fao-eng.fao.eng 28.1 0.460
Tatoeba-test.fas-eng.fas.eng 9.6 0.306
Tatoeba-test.fra-eng.fra.eng 50.3 0.661
Tatoeba-test.frm-eng.frm.eng 30.0 0.457
Tatoeba-test.frr-eng.frr.eng 15.2 0.301
Tatoeba-test.fry-eng.fry.eng 34.4 0.525
Tatoeba-test.gcf-eng.gcf.eng 18.4 0.317
Tatoeba-test.gla-eng.gla.eng 24.1 0.400
Tatoeba-test.gle-eng.gle.eng 52.2 0.671
Tatoeba-test.glg-eng.glg.eng 50.5 0.669
Tatoeba-test.glv-eng.glv.eng 5.7 0.189
Tatoeba-test.gos-eng.gos.eng 19.2 0.378
Tatoeba-test.got-eng.got.eng 0.1 0.022
Tatoeba-test.grc-eng.grc.eng 0.9 0.095
Tatoeba-test.gsw-eng.gsw.eng 23.9 0.390
Tatoeba-test.guj-eng.guj.eng 28.0 0.428
Tatoeba-test.hat-eng.hat.eng 44.2 0.567
Tatoeba-test.hbs-eng.hbs.eng 51.6 0.666
Tatoeba-test.hif-eng.hif.eng 22.3 0.451
Tatoeba-test.hin-eng.hin.eng 41.7 0.585
Tatoeba-test.hsb-eng.hsb.eng 46.4 0.590
Tatoeba-test.hye-eng.hye.eng 40.4 0.564
Tatoeba-test.isl-eng.isl.eng 43.8 0.605
Tatoeba-test.ita-eng.ita.eng 60.7 0.735
Tatoeba-test.jdt-eng.jdt.eng 5.5 0.091
Tatoeba-test.kok-eng.kok.eng 7.8 0.205
Tatoeba-test.ksh-eng.ksh.eng 15.8 0.284
Tatoeba-test.kur-eng.kur.eng 11.6 0.232
Tatoeba-test.lad-eng.lad.eng 30.7 0.484
Tatoeba-test.lah-eng.lah.eng 11.0 0.286
Tatoeba-test.lat-eng.lat.eng 24.4 0.432
Tatoeba-test.lav-eng.lav.eng 47.2 0.646
Tatoeba-test.lij-eng.lij.eng 9.0 0.287
Tatoeba-test.lit-eng.lit.eng 51.7 0.670
Tatoeba-test.lld-eng.lld.eng 22.4 0.369
Tatoeba-test.lmo-eng.lmo.eng 26.1 0.381
Tatoeba-test.ltz-eng.ltz.eng 39.8 0.536
Tatoeba-test.mai-eng.mai.eng 72.3 0.758
Tatoeba-test.mar-eng.mar.eng 32.0 0.554
Tatoeba-test.mfe-eng.mfe.eng 63.1 0.822
Tatoeba-test.mkd-eng.mkd.eng 49.5 0.638
Tatoeba-test.msa-eng.msa.eng 38.6 0.566
Tatoeba-test.multi.eng 45.6 0.615
Tatoeba-test.mwl-eng.mwl.eng 40.4 0.767
Tatoeba-test.nds-eng.nds.eng 35.5 0.538
Tatoeba-test.nep-eng.nep.eng 4.9 0.209
Tatoeba-test.nld-eng.nld.eng 54.2 0.694
Tatoeba-test.non-eng.non.eng 39.3 0.573
Tatoeba-test.nor-eng.nor.eng 50.9 0.663
Tatoeba-test.oci-eng.oci.eng 19.6 0.386
Tatoeba-test.ori-eng.ori.eng 16.2 0.364
Tatoeba-test.orv-eng.orv.eng 13.6 0.288
Tatoeba-test.oss-eng.oss.eng 9.4 0.301
Tatoeba-test.pan-eng.pan.eng 17.1 0.389
Tatoeba-test.pap-eng.pap.eng 57.0 0.680
Tatoeba-test.pdc-eng.pdc.eng 41.6 0.526
Tatoeba-test.pms-eng.pms.eng 13.7 0.333
Tatoeba-test.pol-eng.pol.eng 46.5 0.632
Tatoeba-test.por-eng.por.eng 56.4 0.710
Tatoeba-test.prg-eng.prg.eng 2.3 0.193
Tatoeba-test.pus-eng.pus.eng 3.2 0.194
Tatoeba-test.roh-eng.roh.eng 17.5 0.420
Tatoeba-test.rom-eng.rom.eng 5.0 0.237
Tatoeba-test.ron-eng.ron.eng 51.4 0.670
Tatoeba-test.rue-eng.rue.eng 26.0 0.447
Tatoeba-test.rus-eng.rus.eng 47.8 0.634
Tatoeba-test.san-eng.san.eng 4.0 0.195
Tatoeba-test.scn-eng.scn.eng 45.1 0.440
Tatoeba-test.sco-eng.sco.eng 41.9 0.582
Tatoeba-test.sgs-eng.sgs.eng 38.7 0.498
Tatoeba-test.sin-eng.sin.eng 29.7 0.499
Tatoeba-test.slv-eng.slv.eng 38.2 0.564
Tatoeba-test.snd-eng.snd.eng 12.7 0.342
Tatoeba-test.spa-eng.spa.eng 53.2 0.687
Tatoeba-test.sqi-eng.sqi.eng 51.9 0.679
Tatoeba-test.stq-eng.stq.eng 9.0 0.391
Tatoeba-test.swe-eng.swe.eng 57.4 0.705
Tatoeba-test.swg-eng.swg.eng 18.0 0.338
Tatoeba-test.tgk-eng.tgk.eng 24.3 0.413
Tatoeba-test.tly-eng.tly.eng 1.1 0.094
Tatoeba-test.ukr-eng.ukr.eng 48.0 0.639
Tatoeba-test.urd-eng.urd.eng 27.2 0.471
Tatoeba-test.vec-eng.vec.eng 28.0 0.398
Tatoeba-test.wln-eng.wln.eng 17.5 0.320
Tatoeba-test.yid-eng.yid.eng 26.9 0.457
Tatoeba-test.zza-eng.zza.eng 1.7 0.131

opus4m-2020-08-12.zip

  • dataset: opus4m
  • model: transformer
  • source language(s): afr aln ang_Latn arg asm ast awa bel bel_Latn ben bho bos_Latn bre bul bul_Latn cat ces cor cos csb_Latn cym dan deu dsb egl ell enm_Latn ext fao fra frm_Latn frr fry gcf_Latn gla gle glg glv gom gos got_Goth grc_Grek gsw guj hat hif_Latn hin hrv hsb hye ind isl ita jdt_Cyrl ksh kur_Arab kur_Latn lad lad_Latn lat_Latn lav lij lit lld_Latn lmo ltg ltz mai mar max_Latn mfe min mkd mwl nds nld nno nob nob_Hebr non_Latn npi oci ori orv_Cyrl oss pan_Guru pap pdc pes pes_Latn pes_Thaa pms pnb pol por prg_Latn pus roh rom ron rue rus san_Deva scn sco sgs sin slv snd_Arab spa sqi srp_Cyrl srp_Latn stq swe swg tgk_Cyrl tly_Latn tmw_Latn ukr urd vec wln yid zlm_Latn zsm_Latn zza
  • target language(s): eng
  • model: transformer
  • pre-processing: normalization + SentencePiece (spm32k,spm32k)
  • download: opus4m-2020-08-12.zip
  • test set translations: opus4m-2020-08-12.test.txt
  • test set scores: opus4m-2020-08-12.eval.txt

Benchmarks

testset BLEU chr-F
newsdev2014-hineng.hin.eng 11.4 0.380
newsdev2016-enro-roneng.ron.eng 35.6 0.616
newsdev2017-enlv-laveng.lav.eng 25.3 0.545
newsdev2019-engu-gujeng.guj.eng 16.2 0.429
newsdev2019-enlt-liteng.lit.eng 24.7 0.527
newsdiscussdev2015-enfr-fraeng.fra.eng 30.0 0.550
newsdiscusstest2015-enfr-fraeng.fra.eng 33.4 0.574
newssyscomb2009-ceseng.ces.eng 24.1 0.521
newssyscomb2009-deueng.deu.eng 25.4 0.526
newssyscomb2009-fraeng.fra.eng 28.2 0.552
newssyscomb2009-itaeng.ita.eng 31.5 0.575
newssyscomb2009-spaeng.spa.eng 28.1 0.556
news-test2008-deueng.deu.eng 24.6 0.518
news-test2008-fraeng.fra.eng 24.7 0.525
news-test2008-spaeng.spa.eng 25.8 0.534
newstest2009-ceseng.ces.eng 23.8 0.519
newstest2009-deueng.deu.eng 23.3 0.513
newstest2009-fraeng.fra.eng 27.5 0.546
newstest2009-itaeng.ita.eng 30.3 0.568
newstest2009-spaeng.spa.eng 28.2 0.552
newstest2010-ceseng.ces.eng 24.0 0.527
newstest2010-deueng.deu.eng 26.3 0.546
newstest2010-fraeng.fra.eng 28.7 0.563
newstest2010-spaeng.spa.eng 31.9 0.584
newstest2011-ceseng.ces.eng 24.6 0.524
newstest2011-deueng.deu.eng 23.9 0.523
newstest2011-fraeng.fra.eng 29.7 0.572
newstest2011-spaeng.spa.eng 30.4 0.571
newstest2012-ceseng.ces.eng 23.8 0.519
newstest2012-deueng.deu.eng 25.2 0.531
newstest2012-fraeng.fra.eng 30.2 0.569
newstest2012-ruseng.rus.eng 30.2 0.568
newstest2012-spaeng.spa.eng 33.4 0.593
newstest2013-ceseng.ces.eng 26.2 0.535
newstest2013-deueng.deu.eng 27.9 0.547
newstest2013-fraeng.fra.eng 30.1 0.562
newstest2013-ruseng.rus.eng 24.0 0.514
newstest2013-spaeng.spa.eng 31.0 0.579
newstest2014-csen-ceseng.ces.eng 28.3 0.563
newstest2014-deen-deueng.deu.eng 28.3 0.548
newstest2014-fren-fraeng.fra.eng 32.5 0.593
newstest2014-hien-hineng.hin.eng 16.9 0.457
newstest2014-ruen-ruseng.rus.eng 27.5 0.555
newstest2015-encs-ceseng.ces.eng 25.5 0.522
newstest2015-ende-deueng.deu.eng 29.3 0.556
newstest2015-enru-ruseng.rus.eng 25.9 0.530
newstest2016-encs-ceseng.ces.eng 27.3 0.543
newstest2016-ende-deueng.deu.eng 33.8 0.596
newstest2016-enro-roneng.ron.eng 33.2 0.593
newstest2016-enru-ruseng.rus.eng 25.2 0.528
newstest2017-encs-ceseng.ces.eng 23.9 0.515
newstest2017-ende-deueng.deu.eng 29.7 0.560
newstest2017-enlv-laveng.lav.eng 19.0 0.489
newstest2017-enru-ruseng.rus.eng 28.2 0.549
newstest2018-encs-ceseng.ces.eng 25.1 0.524
newstest2018-ende-deueng.deu.eng 36.3 0.607
newstest2018-enru-ruseng.rus.eng 23.9 0.519
newstest2019-deen-deueng.deu.eng 31.7 0.571
newstest2019-guen-gujeng.guj.eng 13.0 0.391
newstest2019-lten-liteng.lit.eng 27.3 0.562
newstest2019-ruen-ruseng.rus.eng 27.6 0.543
Tatoeba-test.afr-eng.afr.eng 59.3 0.725
Tatoeba-test.ang-eng.ang.eng 7.5 0.240
Tatoeba-test.arg-eng.arg.eng 38.9 0.494
Tatoeba-test.asm-eng.asm.eng 21.8 0.403
Tatoeba-test.ast-eng.ast.eng 37.7 0.533
Tatoeba-test.awa-eng.awa.eng 10.2 0.271
Tatoeba-test.bel-eng.bel.eng 41.7 0.597
Tatoeba-test.ben-eng.ben.eng 45.7 0.599
Tatoeba-test.bho-eng.bho.eng 31.0 0.527
Tatoeba-test.bre-eng.bre.eng 24.3 0.412
Tatoeba-test.bul-eng.bul.eng 51.2 0.663
Tatoeba-test.cat-eng.cat.eng 51.7 0.680
Tatoeba-test.ces-eng.ces.eng 48.2 0.648
Tatoeba-test.cor-eng.cor.eng 5.6 0.213
Tatoeba-test.cos-eng.cos.eng 54.8 0.633
Tatoeba-test.csb-eng.csb.eng 24.3 0.410
Tatoeba-test.cym-eng.cym.eng 37.5 0.541
Tatoeba-test.dan-eng.dan.eng 56.7 0.711
Tatoeba-test.deu-eng.deu.eng 45.3 0.628
Tatoeba-test.dsb-eng.dsb.eng 22.4 0.431
Tatoeba-test.egl-eng.egl.eng 3.6 0.185
Tatoeba-test.ell-eng.ell.eng 62.3 0.743
Tatoeba-test.enm-eng.enm.eng 21.5 0.476
Tatoeba-test.ext-eng.ext.eng 30.9 0.491
Tatoeba-test.fao-eng.fao.eng 28.6 0.463
Tatoeba-test.fas-eng.fas.eng 9.7 0.309
Tatoeba-test.fra-eng.fra.eng 50.6 0.664
Tatoeba-test.frm-eng.frm.eng 30.6 0.459
Tatoeba-test.frr-eng.frr.eng 9.3 0.111
Tatoeba-test.fry-eng.fry.eng 33.1 0.514
Tatoeba-test.gcf-eng.gcf.eng 15.3 0.294
Tatoeba-test.gla-eng.gla.eng 24.7 0.412
Tatoeba-test.gle-eng.gle.eng 53.0 0.676
Tatoeba-test.glg-eng.glg.eng 50.1 0.667
Tatoeba-test.glv-eng.glv.eng 32.1 0.357
Tatoeba-test.gos-eng.gos.eng 21.9 0.389
Tatoeba-test.got-eng.got.eng 0.1 0.030
Tatoeba-test.grc-eng.grc.eng 2.1 0.144
Tatoeba-test.gsw-eng.gsw.eng 21.2 0.376
Tatoeba-test.guj-eng.guj.eng 28.1 0.421
Tatoeba-test.hat-eng.hat.eng 42.1 0.556
Tatoeba-test.hbs-eng.hbs.eng 51.9 0.668
Tatoeba-test.hif-eng.hif.eng 12.8 0.418
Tatoeba-test.hin-eng.hin.eng 42.0 0.589
Tatoeba-test.hsb-eng.hsb.eng 47.8 0.609
Tatoeba-test.hye-eng.hye.eng 39.8 0.559
Tatoeba-test.isl-eng.isl.eng 44.2 0.610
Tatoeba-test.ita-eng.ita.eng 61.0 0.738
Tatoeba-test.jdt-eng.jdt.eng 1.2 0.035
Tatoeba-test.kok-eng.kok.eng 3.7 0.167
Tatoeba-test.ksh-eng.ksh.eng 12.0 0.288
Tatoeba-test.kur-eng.kur.eng 11.7 0.221
Tatoeba-test.lad-eng.lad.eng 32.2 0.493
Tatoeba-test.lah-eng.lah.eng 11.5 0.309
Tatoeba-test.lat-eng.lat.eng 24.8 0.436
Tatoeba-test.lav-eng.lav.eng 48.7 0.654
Tatoeba-test.lij-eng.lij.eng 7.2 0.284
Tatoeba-test.lit-eng.lit.eng 51.7 0.671
Tatoeba-test.lld-eng.lld.eng 19.4 0.348
Tatoeba-test.lmo-eng.lmo.eng 19.5 0.362
Tatoeba-test.ltz-eng.ltz.eng 41.0 0.547
Tatoeba-test.mai-eng.mai.eng 67.8 0.740
Tatoeba-test.mar-eng.mar.eng 33.6 0.566
Tatoeba-test.mfe-eng.mfe.eng 63.1 0.822
Tatoeba-test.mkd-eng.mkd.eng 49.6 0.640
Tatoeba-test.msa-eng.msa.eng 38.5 0.568
Tatoeba-test.multi.eng 46.0 0.619
Tatoeba-test.mwl-eng.mwl.eng 38.4 0.653
Tatoeba-test.nds-eng.nds.eng 36.2 0.544
Tatoeba-test.nep-eng.nep.eng 5.4 0.220
Tatoeba-test.nld-eng.nld.eng 54.6 0.697
Tatoeba-test.non-eng.non.eng 47.5 0.650
Tatoeba-test.nor-eng.nor.eng 51.5 0.668
Tatoeba-test.oci-eng.oci.eng 20.9 0.401
Tatoeba-test.ori-eng.ori.eng 15.4 0.361
Tatoeba-test.orv-eng.orv.eng 14.3 0.310
Tatoeba-test.oss-eng.oss.eng 10.2 0.292
Tatoeba-test.pan-eng.pan.eng 17.4 0.409
Tatoeba-test.pap-eng.pap.eng 56.9 0.658
Tatoeba-test.pdc-eng.pdc.eng 41.6 0.518
Tatoeba-test.pms-eng.pms.eng 15.5 0.353
Tatoeba-test.pol-eng.pol.eng 47.0 0.636
Tatoeba-test.por-eng.por.eng 56.9 0.714
Tatoeba-test.prg-eng.prg.eng 2.5 0.182
Tatoeba-test.pus-eng.pus.eng 2.6 0.202
Tatoeba-test.roh-eng.roh.eng 20.4 0.435
Tatoeba-test.rom-eng.rom.eng 5.0 0.232
Tatoeba-test.ron-eng.ron.eng 51.8 0.674
Tatoeba-test.rue-eng.rue.eng 26.0 0.460
Tatoeba-test.rus-eng.rus.eng 48.1 0.637
Tatoeba-test.san-eng.san.eng 4.2 0.193
Tatoeba-test.scn-eng.scn.eng 39.9 0.414
Tatoeba-test.sco-eng.sco.eng 35.5 0.533
Tatoeba-test.sgs-eng.sgs.eng 31.8 0.483
Tatoeba-test.sin-eng.sin.eng 28.9 0.484
Tatoeba-test.slv-eng.slv.eng 40.4 0.581
Tatoeba-test.snd-eng.snd.eng 21.3 0.420
Tatoeba-test.spa-eng.spa.eng 53.1 0.688
Tatoeba-test.sqi-eng.sqi.eng 52.7 0.686
Tatoeba-test.stq-eng.stq.eng 8.5 0.387
Tatoeba-test.swe-eng.swe.eng 57.5 0.707
Tatoeba-test.swg-eng.swg.eng 18.5 0.355
Tatoeba-test.tgk-eng.tgk.eng 25.7 0.435
Tatoeba-test.tly-eng.tly.eng 1.4 0.122
Tatoeba-test.ukr-eng.ukr.eng 48.7 0.645
Tatoeba-test.urd-eng.urd.eng 28.1 0.481
Tatoeba-test.vec-eng.vec.eng 30.3 0.395
Tatoeba-test.wln-eng.wln.eng 18.1 0.338
Tatoeba-test.yid-eng.yid.eng 28.1 0.465
Tatoeba-test.zza-eng.zza.eng 2.0 0.134