Home / Query / WordAlign / Wiki    [ada83] [bible] [bianet] [books] [CCAligned] [CCMatrix] [CAPES] [DGT] [DOGC] [ECB] [EhuHac] [EiTB] [Elhuyar] [ELITR_ECA] [ELRC] [EMEA] [EUbooks] [EU] [Europarl] [EuroPat] [finlex] [fiskmö] [giga] [GNOME] [GlobalVoices] [hren] [infopankki] [JRC] [KDE4/doc] [liv4ever] [MBS] [memat] [MontenegrinSubs] [MultiUN] [MultiParaCrawl] [MultiCCAligned] [MT560] [NC] [Ofis] [OO/OO3] [subs/16/18] [Opus100] [ParaCrawl] [ParCor] [PHP] [QED] [sardware] [SciELO] [SETIMES] [SPC] [Tatoeba] [Tanzil] [TEP] [TED] [tico19] [Tilde] [Ubuntu] [UN] [UNPC] [WikiMatrix] [Wikimedia] [Wikipedia] [WikiSource] [WMT] [XhosaNavy]

HPLT v2

Parallel corpora from Web Crawls collected in the HPLT project

51 languages, 50 bitexts
total number of files: 810
total number of tokens: 17.53G
total number of sentence fragments: 841.78M

Please, acknowledge the HPLT project at https://hplt-project.org/. This version is derived from the original release at their website adjusted for redistribution via the OPUS corpus collection. Please, acknowledge OPUS as well for this service.

Download

Below you can download data files for all language pairs in different formats and with different kind of annotation (if available). You can click on the various links as explained below. In addition to the files shown on this webpage, OPUS also provides pre-compiled word alignments and phrase tables, bilingual dictionaries, frequency counts, and these files can be found through the resources search form on the top-level website of OPUS.

The data packaging is released under the Creative Commons CC0 license ("no rights reserved"), see https://hplt-project.org/ for more details

Release history:

Bottom-left triangle: download files
  • ces = sentence alignments in XCES format
  • leftmost column language IDs = tokenized corpus files in XML
  • TMX and plain text files (Moses): see "Statistics" below
  • lower row language IDs = parsed corpus files (if they exist)
Upper-right triangle: sample files
  • view = bilingual XML file samples
  • upper row language IDs = monolingual XML file samples
  • rightmost column language IDs = untokenized corpus files

af ar az be bg bn bs ca cy en eo et eu fa fi ga gl gu he hi hr is ja kk kn ko lt lv mk ml mr ms mt nb ne nn si sk sl sq sr sw ta te th tr uk ur uz vi xh
af viewaf af af
ar viewar ar ar
az viewaz az az
be viewbe be be
bg viewbg bg bg
bn viewbn bn bn
bs viewbs bs bs
ca viewca ca ca
cy viewcy cy cy
en ces ces ces ces ces ces ces ces ces viewviewviewviewviewviewviewen viewviewviewviewviewviewviewviewviewviewviewviewviewviewviewviewviewviewen viewviewviewviewviewviewviewviewviewviewviewviewviewviewviewviewen
eo ces eo eo eo
et ces et et et
eu ces eu eu eu
fa ces fa fa fa
fi ces fi fi fi
ga ces ga ga ga
gl ces gl gl gl
af ar az be bg bn bs ca cy en eo et eu fa fi ga gl gu he hi hr is ja kk kn ko lt lv mk ml mr ms mt nb ne nn si sk sl sq sr sw ta te th tr uk ur uz vi xh
gu ces gu gu gu
he ces he he he
hi ces hi hi hi
hr ces hr hr hr
is ces is is is
ja ces ja ja ja
kk ces kk kk kk
kn ces kn kn kn
ko ces ko ko ko
lt ces lt lt lt
lv ces lv lv lv
mk ces mk mk mk
ml ces ml ml ml
mr ces mr mr mr
ms ces ms ms ms
mt ces mt mt mt
nb ces nb nb nb
ne ces ne ne ne
af ar az be bg bn bs ca cy en eo et eu fa fi ga gl gu he hi hr is ja kk kn ko lt lv mk ml mr ms mt nb ne nn si sk sl sq sr sw ta te th tr uk ur uz vi xh
nn ces nn nn nn
si ces si si si
sk ces sk sk sk
sl ces sl sl sl
sq ces sq sq sq
sr ces sr sr sr
sw ces sw sw sw
ta ces ta ta ta
te ces te te te
th ces th th th
tr ces tr tr tr
uk ces uk uk uk
ur ces ur ur ur
uz ces uz uz uz
vi ces vi vi vi
xh ces xh xh xh
af ar az be bg bn bs ca cy en eo et eu fa fi ga gl gu he hi hr is ja kk kn ko lt lv mk ml mr ms mt nb ne nn si sk sl sq sr sw ta te th tr uk ur uz vi xh

Statistics and TMX/Moses Downloads

Number of files, tokens, and sentences per language (including non-parallel ones if they exist)
Number of sentence alignment units per language pair

Upper-right triangle: download translation memory files (TMX)
Bottom-left triangle: download plain text files (MOSES/GIZA++)
Language ID's, first row: monolingual plain text files (tokenized)
Language ID's, first column: monolingual plain text files (untokenized)

language files tokens sentencesaf ar az be bg bn bs ca cy en eo et eu fa fi ga gl gu he hi hr is ja kk kn ko lt lv mk ml mr ms mt nb ne nn si sk sl sq sr sw ta te th tr uk ur uz vi xh
af 4 95.7M 4.3M4.0M
ar 18 481.2M 17.7M17.5M
az 4 62.3M 3.5M3.2M
be 4 60.9M 3.3M3.1M
bg 23 501.5M 24.4M22.7M
bn 3 106.0M 2.3M2.3M
bs 5 108.8M 4.9M4.6M
ca 14 346.7M 13.9M13.1M
cy 4 93.1M 4.1M3.9M
en 405 8.7G 429.7M4.0M 17.5M 3.2M 3.1M 22.7M 2.3M 4.6M 13.1M 3.9M 1.5M 8.8M 1.5M 3.4M 29.1M 2.7M 2.8M 0.7M 8.7M 9.9M 14.3M 2.7M 18.9M 1.9M 0.7M 18.4M 12.9M 11.3M 4.0M 0.5M 0.7M 8.4M 1.5M 22.9M 0.3M 0.6M 0.3M 20.1M 10.3M 4.2M 5.3M 2.0M 1.1M 0.9M 4.1M 21.6M 25.1M 1.4M 1.2M 19.2M 0.4M
eo 2 37.9M 1.7M1.5M
et 9 160.0M 9.4M8.8M
eu 2 30.5M 1.7M1.5M
fa 4 107.7M 3.5M3.4M
fi 30 460.4M 31.5M29.1M
ga 3 66.3M 2.8M2.7M
gl 3 67.6M 3.0M2.8M
gu 1 29.5M 0.7M0.7M
he 9 198.1M 8.8M8.7M
hi 10 422.8M 10.0M9.9M
hr 15 292.8M 15.4M14.3M
is 3 53.9M 2.9M2.7M
ja 19 550.9M 19.0M18.9M
kk 2 36.9M 2.1M1.9M
kn 1 28.7M 0.7M0.7M
ko 19 368.4M 19.0M18.4M
lt 13 248.5M 13.9M12.9M
lv 12 222.4M 12.0M11.3M
mk 4 91.2M 4.2M4.0M
ml 1 23.3M 0.6M0.5M
mr 1 28.8M 0.7M0.7M
ms 9 173.6M 8.9M8.4M
mt 2 53.9M 1.6M1.5M
nb 23 458.4M 24.4M22.9M
ne 1 8.4M 0.3M0.3M
nn 1 12.4M 0.6M0.6M
si 1 6.8M 0.3M0.3M
sk 21 388.3M 21.4M20.1M
sl 11 222.3M 11.0M10.3M
sq 5 108.2M 4.4M4.2M
sr 6 106.5M 5.6M5.3M
sw 2 46.3M 2.2M2.0M
ta 2 54.6M 1.2M1.1M
te 1 28.4M 0.9M0.9M
th 5 148.5M 4.2M4.1M
tr 22 485.1M 31.8M21.6M
uk 26 490.5M 27.2M25.1M
ur 2 52.4M 1.4M1.4M
uz 2 23.6M 1.2M1.2M
vi 20 564.9M 20.7M19.2M
xh 1 8.1M 0.4M0.4M

Note that TMX files only contain unique translation units and, therefore, the number of aligned units is smaller than for the distributions in Moses and XML format. Moses downloads include all non-empty alignment units including duplicates. Token counts for each language also include duplicate sentences and documents.

Disclaimer

Notice and take down policy

Notice: Should you consider that our data contains material that is owned by you and should therefore not be reproduced here, please: Take down: We will comply to legitimate requests by removing the affected sources from the next release of the corpus.