Dataset Preview
Go to dataset viewer
sourceLang (string)targetlang (string)sourceString (string)targetString (string)
"afr"
"deu"
"Aanvaar u Visa kaarte?"
"Akzeptieren Sie Visa-Karten?"
"afr"
"deu"
"Aan wie het jy die brief geskryf?"
"Wem hast du den Brief geschrieben?"
"afr"
"deu"
"Alle banke is beset."
"Alle Bänke sind besetzt."
"afr"
"deu"
"Alle menslike wesens word vry, met gelyke waardigheid en regte, gebore. Hulle het rede en gewete en behoort in die gees van broederskap teenoor mekaar op te tree."
"Alle Menschen sind frei und gleich an Würde und Rechten geboren. Sie sind mit Vernunft und Gewissen begabt und sollen einander im Geist der Brüderlichkeit begegnen."
"afr"
"deu"
"Alles goed eindig goed."
"Ende gut, alles gut."
"afr"
"deu"
"Alles is normaal."
"Alles ist normal."
"afr"
"deu"
"Alles is onder beheer."
"Alles ist unter Kontrolle."
"afr"
"deu"
"Alles is verby tussen ons."
"Zwischen uns ist alles vorbei."
"afr"
"deu"
"Alles loop skeef."
"Alles läuft schief."
"afr"
"deu"
"Almal beweer dat hulle onskuldig is."
"Alle behaupten, sie seien unschuldig."
"afr"
"deu"
"Almal dans."
"Alle tanzen."
"afr"
"deu"
"Almal huil."
"Alle weinen."
"afr"
"deu"
"Almal kyk."
"Alle sehen zu."
"afr"
"deu"
"Almal lag vir hom."
"Alle lachten über ihn."
"afr"
"deu"
"Almal lieg."
"Alle lügen."
"afr"
"deu"
"Altyd vorentoe, nooit agtertoe nie."
"Vorwärts immer, rückwärts nimmer."
"afr"
"deu"
"Amerika is ryk aan olie."
"Amerika ist reich an Öl."
"afr"
"deu"
"Appels is gewoonlik groen, geel of rooi."
"Äpfel sind gewöhnlich grün, gelb oder rot."
"afr"
"deu"
"Arme kat."
"Arme Katze."
"afr"
"deu"
"As ek die pen in die hand het, skrik ek vir niks nie."
"Wenn ich die Feder in der Hand habe, schrecke ich vor nichts zurück."
"afr"
"deu"
"As ek na die partytjie toe gaan, sal ek 'n paar bottels wyn saamvat."
"Wenn ich auf die Party gehe, werde ich ein paar Flaschen Wein mitnehmen."
"afr"
"deu"
"As jy versigtig bestuur, kan jy ongelukke vermy."
"Wenn du vorsichtig fährst, kannst du Unfälle vermeiden."
"afr"
"deu"
"Baie dankie!"
"Danke schön."
"afr"
"deu"
"Baie dankie!"
"Danke vielmals!"
"afr"
"deu"
"Baie dankie!"
"Herzlichen Dank!"
"afr"
"deu"
"Baie dankie! Jy het sopas my lewe verwoes!"
"Vielen Dank! Du hast gerade mein Leben ruiniert!"
"afr"
"deu"
"Baie dankie!"
"Vielen Dank!"
"afr"
"deu"
"Baie raserig, hierdie seun."
"Sehr laut dieser Junge."
"afr"
"deu"
"Baie verjaarsdagkaartjies sal binnekort arriveer."
"Viele Geburtstagskarten werden in Kürze ankommen."
"afr"
"deu"
"Bedien asseblief die koffie."
"Servier bitte den Kaffee."
"afr"
"deu"
"Beheer jouself in alle omstandighede!"
"Beherrsche dich in jeder Lebenslage!"
"afr"
"deu"
"Beloftes maak skuld."
"Versprechen macht Schulden."
"afr"
"deu"
"Belowe u?"
"Versprechen Sie?"
"afr"
"deu"
"Berlyn is die hoofstad van Duitsland."
"Berlin ist die Hauptstadt Deutschlands."
"afr"
"deu"
"Berlyn is die hoofstad van Duitsland."
"Berlin ist die Hauptstadt von Deutschland."
"afr"
"deu"
"Besit julle 'n apteek?"
"Gehört euch eine Apotheke?"
"afr"
"deu"
"Besluit jy."
"Entscheide du."
"afr"
"deu"
"Besluit self!"
"Entscheide selbst!"
"afr"
"deu"
"Besoekers is welkom."
"Besucher sind willkommen."
"afr"
"deu"
"Betaal jy?"
"Zahlst du?"
"afr"
"deu"
"Beter as niks."
"Besser als nichts."
"afr"
"deu"
"Betty het toe laaste gekom."
"Betty kam als Letzte."
"afr"
"deu"
"Bevry jouself van bevooroordeling."
"Befreie dich von Vorurteilen!"
"afr"
"deu"
"Bill, maak die deur oop."
"Bill, mach die Tür auf."
"afr"
"deu"
"Bloed is 'n besondere sap."
"Blut ist ein ganz besondrer Saft."
"afr"
"deu"
"Bly by ons."
"Bleib bei uns."
"afr"
"deu"
"Bly by Tom!"
"Bleib bei Tom!"
"afr"
"deu"
"Blykbaar is die ontvangs nie in hierdie vertrek nie."
"Anscheinend ist die Rezeption nicht in diesem Raum."
"afr"
"deu"
"Bly sterk!"
"Bleib stark!"
"afr"
"deu"
"Bly stil, het hy gefluister."
"„Schweige“, flüsterte er."
"afr"
"deu"
"Bly te kenne."
"Schön, dich kennenzulernen!"
"afr"
"deu"
"Boere kla konstant oor die weer."
"Bauern beschweren sich immer über das Wetter."
"afr"
"deu"
"Boere skei goeie appels van slegtes."
"Bauern trennen gute Äpfel von schlechten."
"afr"
"deu"
"Borsel jou tande."
"Putze deine Zähne."
"afr"
"deu"
"Botter is sag."
"Butter ist weich."
"afr"
"deu"
"By my suster."
"Bei meiner Schwester."
"afr"
"deu"
"Daar is ateïste wat hulself verbeel dat hul God is."
"Es gibt Atheisten, die sich einbilden, Gott zu sein."
"afr"
"deu"
"Daar is baie besienswaardighede in die omgewing."
"In der Umgebung gibt es viele Sehenswürdigkeiten."
"afr"
"deu"
"Daar is baie voordele."
"Es gibt viele Vorteile."
"afr"
"deu"
"Daar is elke tien minute 'n bus."
"Alle zehn Minuten fährt ein Bus."
"afr"
"deu"
"Daar is geen god."
"Es gibt keinen Gott."
"afr"
"deu"
"Daar is niemand oor nie."
"Es ist niemand übrig."
"afr"
"deu"
"Daar is 'n kaart op die muur."
"An der Wand hängt eine Karte."
"afr"
"deu"
"Daar is 'n opvallende onderskeid tussen die twee."
"Es gibt einen auffallenden Unterschied zwischen den beiden."
"afr"
"deu"
"Daar is uitsonderings."
"Es gibt Ausnahmen."
"afr"
"deu"
"Daarom is ek vandag hier."
"Deshalb bin ich heute hier."
"afr"
"deu"
"Daar, reguit aan."
"Dort, geradeaus."
"afr"
"deu"
"Daar was blomme orals."
"Überall waren Blumen."
"afr"
"deu"
"Daar was oral blomme."
"Überall waren Blumen."
"afr"
"deu"
"Daar word gesê hy lewe nog."
"Er soll noch am Leben sein."
"afr"
"deu"
"Dankie!"
"Danke!"
"afr"
"deu"
"Dankie dat julle gekom het."
"Danke fürs Kommen."
"afr"
"deu"
"Dankie dat jy vra!"
"Danke, dass du fragst!"
"afr"
"deu"
"Dankie dat U gekom het."
"Danke fürs Kommen."
"afr"
"deu"
"Dankie. "Jy is welkom.""
"„Danke!“ – „Bitte!“"
"afr"
"deu"
"Dankie. "Jy is welkom.""
"„Danke!“ – „Keine Ursache!“"
"afr"
"deu"
"Dankie. "Jy is welkom.""
"„Danke!“ – „Nichts zu danken!“"
"afr"
"deu"
"Dankie vir alles."
"Danke für alles."
"afr"
"deu"
"Dankie vir die inligting."
"Danke für die Auskunft."
"afr"
"deu"
"Dankie vir die kom, om my te ontmoet."
"Danke fürs Kommen, um mich zu treffen."
"afr"
"deu"
"Dankie vir jou geskenk."
"Danke für dein Geschenk."
"afr"
"deu"
"Dans u graag?"
"Tanzen Sie gerne?"
"afr"
"deu"
"Die aanpaskamer is beset."
"Die Umkleidekabine ist besetzt."
"afr"
"deu"
"Die aantrek kamer daar anderkant is oop."
"Die Umkleidekabine da drüben ist frei."
"afr"
"deu"
"Die Aarde is rond, nie plat nie."
"Die Erde ist rund, nicht flach."
"afr"
"deu"
"Die aarde se bevolking gaan gou verdubbel."
"Die Bevölkerung der Erde wird sich bald verdoppeln."
"afr"
"deu"
"Die aartappel is 'n lekker groente."
"Die Kartoffel ist ein leckeres Gemüse."
"afr"
"deu"
"Die aartappel was so warm, dat dit my mond verbrand het."
"Die Kartoffel war so heiß, dass sie mir den Mund verbrannte."
"afr"
"deu"
"Die appels is ryp."
"Die Äpfel sind reif."
"afr"
"deu"
"Die bagasie is verseker."
"Das Gepäck ist versichert."
"afr"
"deu"
"Die bakkery is langsaam die slaghuis."
"Die Bäckerei liegt neben der Metzgerei."
"afr"
"deu"
"Die bediener was af."
"Der Server war ausgefallen."
"afr"
"deu"
"Die beker is vol."
"Die Tasse ist voll."
"afr"
"deu"
"Die blare is geel."
"Die Blätter sind gelb."
"afr"
"deu"
"Die blare val stadig."
"Langsam fallen die Blätter."
"afr"
"deu"
"Die blom het 'n sterk geur."
"Die Blume hat einen starken Duft."
"afr"
"deu"
"Die boek gaan oor die koning wat sy kroon verloor."
"Dieses Buch handelt von einem König, der seine Krone verlor."
"afr"
"deu"
"Die bokant van die heuwel is plat."
"Der Gipfel des Hügels ist flach."
"afr"
"deu"
"Die boks is amper leeg."
"Die Schachtel ist fast leer."
"afr"
"deu"
"Die bome is groen."
"Die Bäume sind grün."
End of preview (truncated to 100 rows)
YAML Metadata Warning: The task_categories "conditional-text-generation" is not in the official list: text-classification, token-classification, table-question-answering, question-answering, zero-shot-classification, translation, summarization, conversational, feature-extraction, text-generation, text2text-generation, fill-mask, sentence-similarity, text-to-speech, automatic-speech-recognition, audio-to-audio, audio-classification, voice-activity-detection, depth-estimation, image-classification, object-detection, image-segmentation, text-to-image, image-to-text, image-to-image, unconditional-image-generation, video-classification, reinforcement-learning, robotics, tabular-classification, tabular-regression, tabular-to-text, table-to-text, multiple-choice, text-retrieval, time-series-forecasting, visual-question-answering, document-question-answering, zero-shot-image-classification, other
YAML Metadata Warning: The task_ids "machine-translation" is not in the official list: acceptability-classification, entity-linking-classification, fact-checking, intent-classification, multi-class-classification, multi-label-classification, multi-input-text-classification, natural-language-inference, semantic-similarity-classification, sentiment-classification, topic-classification, semantic-similarity-scoring, sentiment-scoring, sentiment-analysis, hate-speech-detection, text-scoring, named-entity-recognition, part-of-speech, parsing, lemmatization, word-sense-disambiguation, coreference-resolution, extractive-qa, open-domain-qa, closed-domain-qa, news-articles-summarization, news-articles-headline-generation, dialogue-generation, dialogue-modeling, language-modeling, text-simplification, explanation-generation, abstractive-qa, open-domain-abstractive-qa, closed-domain-qa, open-book-qa, closed-book-qa, slot-filling, masked-language-modeling, keyword-spotting, speaker-identification, audio-intent-classification, audio-emotion-recognition, audio-language-identification, multi-label-image-classification, multi-class-image-classification, face-detection, vehicle-detection, instance-segmentation, semantic-segmentation, panoptic-segmentation, image-captioning, grasping, task-planning, tabular-multi-class-classification, tabular-multi-label-classification, tabular-single-column-regression, rdf-to-text, multiple-choice-qa, multiple-choice-coreference-resolution, document-retrieval, utterance-retrieval, entity-linking-retrieval, fact-checking-retrieval, univariate-time-series-forecasting, multivariate-time-series-forecasting, visual-question-answering, document-question-answering

Dataset Card for [Dataset Name]

Dataset Summary

The Tatoeba Translation Challenge is a multilingual data set of machine translation benchmarks derived from user-contributed translations collected by Tatoeba.org and provided as parallel corpus from OPUS. This dataset includes test and development data sorted by language pair. It includes test sets for hundreds of language pairs and is continuously updated. Please, check the version number tag to refer to the release that your are using.

Supported Tasks and Leaderboards

The translation task is described in detail in the Tatoeba-Challenge repository and covers various sub-tasks with different data coverage and resources. Training data is also available from the same repository and results are published and collected as well. Models are also released for public use and are also partially available from the huggingface model hub.

Languages

The data set covers hundreds of languages and language pairs and are organized by ISO-639-3 languages. The current release covers the following language: Afrikaans, Arabic, Azerbaijani, Belarusian, Bulgarian, Bengali, Breton, Bosnian, Catalan, Chamorro, Czech, Chuvash, Welsh, Danish, German, Modern Greek, English, Esperanto, Spanish, Estonian, Basque, Persian, Finnish, Faroese, French, Western Frisian, Irish, Scottish Gaelic, Galician, Guarani, Hebrew, Hindi, Croatian, Hungarian, Armenian, Interlingua, Indonesian, Interlingue, Ido, Icelandic, Italian, Japanese, Javanese, Georgian, Kazakh, Khmer, Korean, Kurdish, Cornish, Latin, Luxembourgish, Lithuanian, Latvian, Maori, Macedonian, Malayalam, Mongolian, Marathi, Malay, Maltese, Burmese, Norwegian Bokmål, Dutch, Norwegian Nynorsk, Norwegian, Occitan, Polish, Portuguese, Quechua, Rundi, Romanian, Russian, Serbo-Croatian, Slovenian, Albanian, Serbian, Swedish, Swahili, Tamil, Telugu, Thai, Turkmen, Tagalog, Turkish, Tatar, Uighur, Ukrainian, Urdu, Uzbek, Vietnamese, Volapük, Yiddish, Chinese

Dataset Structure

Data Instances

Data instances are given as translation units in TAB-separated files with four columns: source and target language ISO-639-3 codes, source language text and target language text. Note that we do not imply a translation direction and consider the data set to be symmetric and to be used as a test set in both directions. Language-pair-specific subsets are only provided under the label of one direction using sorted ISO-639-3 language IDs.

Some subsets contain several sub-languages or language variants. They may refer to macro-languages such as Serbo-Croatian languages that are covered by the ISO code hbs. Language variants may also include different writing systems and in that case the ISO15924 script codes are attached to the language codes. Here are a few examples from the English to Serbo-Croation test set including examples for Bosnian, Croatian and Serbian in Cyrillic and in Latin scripts:

eng	bos_Latn	Children are the flowers of our lives.	Djeca su cvijeće našeg života.
eng	hrv	A bird was flying high up in the sky.	Ptica je visoko letjela nebom.
eng	srp_Cyrl	A bird in the hand is worth two in the bush.	Боље врабац у руци, него голуб на грани.
eng	srp_Latn	Canada is the motherland of ice hockey.	Kanada je zemlja-majka hokeja na ledu.

There are also data sets with sentence pairs in the same language. In most cases, those are variants with minor spelling differences but they also include rephrased sentences. Here are a few examples from the English test set:

eng     eng     All of us got into the car.     We all got in the car.
eng     eng     All of us hope that doesn't happen.     All of us hope that that doesn't happen.
eng     eng     All the seats are booked.       The seats are all sold out.

Data Splits

Test and development data sets are disjoint with respect to sentence pairs but may include overlaps in individual source or target language sentences. Development data should not be used in training directly. The goal of the data splits is to create test sets of reasonable size with a large language coverage. Test sets include at most 10,000 instances. Development data do not exist for all language pairs.

To be comparable with other results, models should use the training data distributed from the Tatoeba MT Challenge Repository including monolingual data sets also listed there.

Dataset Creation

Curation Rationale

The Tatoeba MT data set will be updated continuously and the data preparation procedures are also public and released on github. High language coverage is the main goal of the project and data sets are prepared to be consistent and systematic with standardized language labels and distribution formats.

Source Data

Initial Data Collection and Normalization

The Tatoeba data sets are collected from user-contributed translations submitted to Tatoeba.org and compiled into a multi-parallel corpus in OPUS. The test and development sets are incrementally updated with new releases of the Tatoeba data collection at OPUS. New releases extend the existing data sets. Test sets should not overlap with any of the released development data sets.

Who are the source language producers?

The data sets come from Tatoeba.org, which provides a large database of sentences and their translations into a wide varity of languages. Its content is constantly growing as a result of voluntary contributions of thousands of users. The original project was founded by Trang Ho in 2006, hosted on Sourceforge under the codename of multilangdict.

Annotations

Annotation process

Sentences are translated by volunteers and the Tatoeba database also provides additional metadata about each record including user ratings etc. However, the metadata is currently not used in any way for the compilation of the MT benchmark. Language skills of contributors naturally vary quite a bit and not all translations are done by native speakers of the target language. More information about the contributions can be found at Tatoeba.org.

Who are the annotators?

Personal and Sensitive Information

For information about handling personal and sensitive information we refer to the original provider of the data. This data set has not been processed in any way to detect or remove potentially sensitve or personal information.

Considerations for Using the Data

Social Impact of Dataset

The language coverage is high and with that it represents a highly valuable resource for machine translation development especially for lesser resourced languages and language pairs. The constantly growing database also represents a dynamic resource and its value wil grow further.

Discussion of Biases

The original source lives from its contributors and there interest and background will to certain subjective and cultural biases. Language coverage and translation quality is also biased by the skills of the contributors.

Other Known Limitations

The sentences are typically quite short and, therefore, rather easy to translate. For high-resource languages, this leads to results that will be less useful than more challenging benchmarks. For lesser resource language pairs, the limited complexity of the examples is actually a good thing to measure progress even in very challenging setups.

Additional Information

Dataset Curators

The data set is curated by the University of Helsinki and its language technology research group. Data and tools used for creating and using the resource are open source and will be maintained as part of the OPUS ecosystem for parallel data and machine translation research.

Licensing Information

The data sets are distributed under the same licence agreement as the original Tatoeba database using a CC-BY 2.0 license. More information about the terms of use of the original data sets is listed here.

Citation Information

If you use the data sets then, please, cite the following paper: The Tatoeba Translation Challenge – Realistic Data Sets for Low Resource and Multilingual MT

@inproceedings{tiedemann-2020-tatoeba,
    title = "The Tatoeba Translation Challenge {--} Realistic Data Sets for Low Resource and Multilingual {MT}",
    author = {Tiedemann, J{\"o}rg},
    booktitle = "Proceedings of the Fifth Conference on Machine Translation",
    month = nov,
    year = "2020",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2020.wmt-1.139",
    pages = "1174--1182",
}

Contributions

Thanks to @jorgtied and @Helsinki-NLP for adding this dataset. Thanks also to CSC Finland for providing computational resources and storage space for the work on OPUS and other MT projects.

Downloads last month
88,902
Edit dataset card
Evaluate models HF Leaderboard