diff --git a/docs/tasks.md b/docs/tasks.md index 0cdd71a292..e4668a0824 100644 --- a/docs/tasks.md +++ b/docs/tasks.md @@ -5,287 +5,288 @@ The following tables gives you an overview of the tasks in MTEB. | Name | Languages | Type | Category | Domains | # Samples | Avg. Length (Char.) | |------|-----------|------|----------|---------|-----------|---------------------| -| [AFQMC](https://aclanthology.org/2021.emnlp-main.357) | {'cmn'} | STS | s2s | | | | -| [AILACasedocs](https://zenodo.org/records/4063986) | {'eng'} | Retrieval | p2p | [Legal] | | | -| [AILAStatutes](https://zenodo.org/records/4063986) | {'eng'} | Retrieval | p2p | [Legal] | | | -| [ATEC](https://aclanthology.org/2021.emnlp-main.357) | {'cmn'} | STS | s2s | | | | -| [AllegroReviews](https://aclanthology.org/2020.acl-main.111.pdf) | {'pol'} | Classification | s2s | | {'test': 1006} | {'test': 477.2} | -| [AlloProfClusteringP2P](https://huggingface.co/datasets/lyon-nlp/alloprof) | {'fra'} | Clustering | p2p | | | | -| [AlloProfClusteringS2S](https://huggingface.co/datasets/lyon-nlp/alloprof) | {'fra'} | Clustering | s2s | | | | -| [AlloprofReranking](https://huggingface.co/datasets/antoinelb7/alloprof) | {'fra'} | Reranking | s2s | | | | -| [AlloprofRetrieval](https://huggingface.co/datasets/antoinelb7/alloprof) | {'fra'} | Retrieval | s2p | | | | -| [AmazonCounterfactualClassification](https://arxiv.org/abs/2104.06893) | {'jpn', 'eng', 'deu'} | Classification | s2s | | {'validation': 335, 'test': 670} | {'validation': 109.2, 'test': 106.1} | -| [AmazonPolarityClassification](https://huggingface.co/datasets/amazon_polarity) | {'eng'} | Classification | s2s | | {'test': 400000} | {'test': 431.4} | -| [AmazonReviewsClassification](https://arxiv.org/abs/2010.02573) | {'spa', 'eng', 'fra', 'jpn', 'cmn', 'deu'} | Classification | s2s | | {'validation': 30000, 'test': 30000} | {'validation': 159.2, 'test': 160.4} | -| [AngryTweetsClassification](https://aclanthology.org/2021.nodalida-main.53/) | {'dan'} | Classification | s2s | | {'test': 1050} | {'test': 156.1} | -| [ArguAna](http://argumentation.bplaced.net/arguana/data) | {'eng'} | Retrieval | s2p | | | | -| [ArguAna-PL](https://huggingface.co/datasets/clarin-knext/arguana-pl) | {'pol'} | Retrieval | s2p | | | | -| [ArmenianParaphrasePC](https://github.com/ivannikov-lab/arpa-paraphrase-corpus) (Arthur Malajyan, 2020) | {'hye'} | PairClassification | s2s | [News] | {'train': 4023, 'test': 1470} | {'train': 243.81, 'test': 241.37} | -| [ArxivClassification](https://ieeexplore.ieee.org/document/8675939) (He et al., 2019) | {'eng'} | Classification | s2s | [Academic] | {'test': 2048} | | -| [ArxivClusteringP2P.v2](https://www.kaggle.com/Cornell-University/arxiv) | {'eng'} | Clustering | p2p | | {'test': 250000} | {'test': 1009.98} | -| [ArxivClusteringS2S](https://www.kaggle.com/Cornell-University/arxiv) | {'eng'} | Clustering | s2s | | {'test': 732723} | {'test': 74.0} | -| [AskUbuntuDupQuestions](https://github.com/taolei87/askubuntu) | {'eng'} | Reranking | s2s | | {'test': 2255} | {'test': 52.5} | -| [BIOSSES](https://tabilab.cmpe.boun.edu.tr/BIOSSES/DataSet.html) | {'eng'} | STS | s2s | | | | -| [BQ](https://aclanthology.org/2021.emnlp-main.357) | {'cmn'} | STS | s2s | | | | -| [BSARDRetrieval](https://huggingface.co/datasets/maastrichtlawtech/bsard) | {'fra'} | Retrieval | s2p | | | | -| [BUCC](https://comparable.limsi.fr/bucc2018/bucc2018-task.html) | {'eng', 'fra', 'rus', 'cmn', 'deu'} | BitextMining | s2s | | {'test': 641684} | {'test': 101.3} | -| [BambaraSentimentClassification](https://arxiv.org/abs/2009.08712) | {'mlt'} | Classification | s2s | [Reviews] | {'test': 673} | {'test': 29.4} | -| [Banking77Classification](https://arxiv.org/abs/2003.04807) | {'eng'} | Classification | s2s | | {'test': 3080} | {'test': 54.2} | -| [BengaliHateSpeechClassification](https://huggingface.co/datasets/bn_hate_speech) (Karim et al., 2020) | {'ben'} | Classification | s2s | [News] | {'train': 3418} | {'train': 103.42} | -| [BengaliSentimentAnalysis](https://data.mendeley.com/datasets/p6zc7krs37/4) (Sazzed et al., 2020) | {'ben'} | Classification | s2s | [Reviews] | {'train': 11807} | {'train': 69.66} | -| [BigPatentClustering](https://www.kaggle.com/datasets/big_patent) | {'eng'} | Clustering | s2s | | | | -| [BiorxivClusteringP2P](https://api.biorxiv.org/) | {'eng'} | Clustering | p2p | | {'test': 75000} | {'test': 1666.2} | -| [BiorxivClusteringS2S](https://api.biorxiv.org/) | {'eng'} | Clustering | s2s | | {'test': 75000} | {'test': 101.6} | -| [BlurbsClusteringP2P](https://www.inf.uni-hamburg.de/en/inst/ab/lt/resources/data/germeval-2019-hmc.html) | {'deu'} | Clustering | p2p | | {'test': 174637} | {'test': 664.09} | -| [BlurbsClusteringS2S](https://www.inf.uni-hamburg.de/en/inst/ab/lt/resources/data/germeval-2019-hmc.html) | {'deu'} | Clustering | s2s | | {'test': 174637} | {'test': 23.02} | -| [BornholmBitextMining](https://aclanthology.org/W19-6138/) | {'dan'} | BitextMining | s2s | | {'test': 500} | {'test': 89.7} | -| [BulgarianSentimentClassification](https://arxiv.org/abs/2009.08712) | {'bul'} | Classification | s2s | [Reviews] | {'validation': 838, 'test': 1673} | {'validation': 43.3, 'test': 46.3} | -| [BulgarianStoreReviewSentimentClassfication](https://doi.org/10.7910/DVN/TXIK9P) (Georgieva-Trifonova et al., 2018) | {'bul'} | Classification | s2s | [Reviews] | {'test': 182} | {'test': 316.7} | -| [CBD](http://2019.poleval.pl/files/poleval2019.pdf) | {'pol'} | Classification | s2s | | {'test': 1000} | {'test': 93.2} | -| [CDSC-E](https://aclanthology.org/P17-1073.pdf) | {'pol'} | PairClassification | s2s | | | | -| [CDSC-R](https://aclanthology.org/P17-1073.pdf) | {'pol'} | STS | s2s | | | | -| [CLSClusteringP2P](https://arxiv.org/abs/2209.05034) | {'cmn'} | Clustering | p2p | | | | -| [CLSClusteringS2S](https://arxiv.org/abs/2209.05034) | {'cmn'} | Clustering | s2s | | | | -| [CMedQAv1-reranking](https://github.com/zhangsheng93/cMedQA) | {'cmn'} | Reranking | s2s | | | | -| [CMedQAv2-reranking](https://github.com/zhangsheng93/cMedQA2) | {'cmn'} | Reranking | s2s | | | | -| [CQADupstackAndroidRetrieval](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/) | {'eng'} | Retrieval | s2p | | | | -| [CQADupstackEnglishRetrieval](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/) | {'eng'} | Retrieval | s2p | | | | -| [CQADupstackGamingRetrieval](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/) | {'eng'} | Retrieval | s2p | | | | -| [CQADupstackGisRetrieval](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/) | {'eng'} | Retrieval | s2p | | | | -| [CQADupstackMathematicaRetrieval](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/) | {'eng'} | Retrieval | s2p | | | | -| [CQADupstackPhysicsRetrieval](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/) | {'eng'} | Retrieval | s2p | | | | -| [CQADupstackProgrammersRetrieval](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/) | {'eng'} | Retrieval | s2p | | | | -| [CQADupstackStatsRetrieval](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/) | {'eng'} | Retrieval | s2p | | | | -| [CQADupstackTexRetrieval](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/) | {'eng'} | Retrieval | s2p | | | | -| [CQADupstackUnixRetrieval](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/) | {'eng'} | Retrieval | s2p | | | | -| [CQADupstackWebmastersRetrieval](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/) | {'eng'} | Retrieval | s2p | | | | -| [CQADupstackWordpressRetrieval](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/) | {'eng'} | Retrieval | s2p | | | | -| [ClimateFEVER](https://www.sustainablefinance.uzh.ch/en/research/climate-fever.html) | {'eng'} | Retrieval | s2p | | | | -| [CmedqaRetrieval](https://aclanthology.org/2022.emnlp-main.357.pdf) | {'cmn'} | Retrieval | s2p | | | | -| [Cmnli](https://huggingface.co/datasets/clue/viewer/cmnli) | {'cmn'} | PairClassification | s2s | | | | -| [CodeSearchNetRetrieval](https://huggingface.co/datasets/code_search_net/viewer) (Husain et al., 2019) | {'php', 'python', 'ruby', 'go', 'java', 'javascript'} | Retrieval | p2p | [Programming] | {'test': 1000} | {'test': 1196.4609} | -| [Core17InstructionRetrieval](https://arxiv.org/abs/2403.15246) (Orion Weller, 2024) | {'eng'} | InstructionRetrieval | s2p | [News] | {'eng': 39470} | {'eng': 2747.2883966244726} | -| [CovidRetrieval](https://arxiv.org/abs/2203.03367) | {'cmn'} | Retrieval | s2p | | | | -| [CroatianSentimentClassification](https://arxiv.org/abs/2009.08712) | {'hrv'} | Classification | s2s | [Reviews] | {'validation': 214, 'test': 437} | {'validation': 166.9, 'test': 151.4} | -| [CzechSubjectivityClassification](https://arxiv.org/abs/2009.08712) | {'ces'} | Classification | s2s | [Reviews] | {'validation': 500, 'test': 2000} | {'validation': 108.2, 'test': 108.3} | -| [DBPedia](https://github.com/iai-group/DBpedia-Entity/) | {'eng'} | Retrieval | s2p | | | | -| [DBPedia-PL](https://github.com/iai-group/DBpedia-Entity/) | {'pol'} | Retrieval | s2p | | | | -| [DKHateClassification](https://aclanthology.org/2020.lrec-1.430/) | {'dan'} | Classification | s2s | | {'test': 329} | {'test': 104.0} | -| [DalajClassification](https://spraakbanken.gu.se/en/resources/superlim) | {'dan'} | Classification | s2s | | {'test': 444} | {'test': 243.8} | -| [DanFEVER](https://aclanthology.org/2021.nodalida-main.47/) | {'dan'} | Retrieval | p2p | [Encyclopaedic, Non-fiction] | {'train': 8897} | {'train': 124.84} | -| [DanishPoliticalCommentsClassification](https://huggingface.co/datasets/danish_political_comments) | {'dan'} | Classification | s2s | | {'train': 9010} | {'train': 69.9} | -| [DiaBlaBitextMining](https://inria.hal.science/hal-03021633) | {'fra', 'eng'} | BitextMining | s2s | | | | -| [DuRetrieval](https://aclanthology.org/2022.emnlp-main.357.pdf) | {'cmn'} | Retrieval | s2p | | | | -| [DutchBookReviewSentimentClassification](https://github.com/benjaminvdb/DBRD) (Benjamin et al., 2019) | {'nld'} | Classification | s2s | [Reviews] | {'test': 2224} | {'test': 1443.0} | -| [EcomRetrieval](https://arxiv.org/abs/2203.03367) | {'cmn'} | Retrieval | s2p | | | | -| [EightTagsClustering](https://aclanthology.org/2020.lrec-1.207.pdf) | {'pol'} | Clustering | s2s | | | | -| [EmotionClassification](https://www.aclweb.org/anthology/D18-1404) | {'eng'} | Classification | s2s | | {'validation': 2000, 'test': 2000} | {'validation': 95.3, 'test': 95.6} | -| [EstQA](https://www.semanticscholar.org/paper/Extractive-Question-Answering-for-Estonian-Language-182912IAPM-Alum%C3%A4e/ea4f60ab36cadca059c880678bc4c51e293a85d6?utm_source=direct_link) | {'est'} | Retrieval | s2p | [Encyclopaedic] | {'test': 603} | {'test': 772.5331950207469} | -| [EstonianValenceClassification](https://figshare.com/articles/dataset/Estonian_Valence_Corpus_Eesti_valentsikorpus/24517054) | {'est'} | Classification | s2s | [News] | {'train': 3270, 'test': 818} | {'train': 226.70642201834863, 'test': 231.5085574572127} | -| [FEVER](https://fever.ai/) | {'eng'} | Retrieval | s2p | | | | -| [FQuADRetrieval](https://huggingface.co/datasets/manu/fquad2_test) | {'fra'} | Retrieval | s2p | [Encyclopaedic] | {'test': 400, 'validation': 100} | {'test': 937.0, 'validation': 930.0} | -| [FalseFriendsGermanEnglish](https://drive.google.com/file/d/1jgq0nBnV-UiYNxbKNrrr2gxDEHm-DMKH/view?usp=share_link) | {'deu'} | PairClassification | s2s | | {'test': 1524} | {'test': 40.3} | -| [FiQA-PL](https://sites.google.com/view/fiqa/) | {'pol'} | Retrieval | s2p | | | | -| [FiQA2018](https://sites.google.com/view/fiqa/) | {'eng'} | Retrieval | s2p | | | | -| [FilipinoHateSpeechClassification](https://pcj.csp.org.ph/index.php/pcj/issue/download/29/PCJ%20V14%20N1%20pp1-14%202019) (Neil Vicente Cabasag et al., 2019) | {'fil'} | Classification | s2s | [Social] | {'validation': 2048, 'test': 2048} | {'validation': 88.1, 'test': 87.4} | -| [FinParaSTS](https://huggingface.co/datasets/TurkuNLP/turku_paraphrase_corpus) | {'fin'} | STS | s2s | [News, Subtitles] | {'test': 1000, 'validation': 1000} | {'test': 59.0, 'validation': 58.8} | -| [FloresBitextMining](https://huggingface.co/datasets/facebook/flores) | {'lmo', 'twi', 'mri', 'san', 'tgl', 'khk', 'kir', 'mag', 'vie', 'azj', 'slv', 'heb', 'tur', 'knc', 'bjn', 'ban', 'ydd', 'ckb', 'pes', 'hau', 'ilo', 'tha', 'mar', 'arb', 'scn', 'eng', 'yue', 'pag', 'luo', 'bam', 'lua', 'tso', 'asm', 'nus', 'oci', 'ltg', 'kab', 'mai', 'mos', 'gla', 'apc', 'kat', 'kam', 'khm', 'mal', 'sat', 'kik', 'ace', 'ben', 'uzn', 'nob', 'ewe', 'epo', 'nso', 'smo', 'tum', 'min', 'vec', 'gle', 'kon', 'kmr', 'hun', 'fur', 'ell', 'tsn', 'wol', 'xho', 'bel', 'ceb', 'bem', 'srp', 'lao', 'lim', 'npi', 'awa', 'cym', 'pbt', 'ssw', 'ind', 'ron', 'eus', 'lit', 'sin', 'ajp', 'hin', 'bod', 'lug', 'plt', 'dik', 'nno', 'sun', 'mni', 'ita', 'sag', 'glg', 'swh', 'pan', 'crh', 'urd', 'mya', 'mkd', 'jpn', 'pap', 'tam', 'est', 'ary', 'spa', 'zul', 'war', 'bug', 'azb', 'ory', 'fon', 'quy', 'lus', 'umb', 'kbp', 'kan', 'fra', 'tgk', 'zho', 'ast', 'lin', 'cjk', 'grn', 'prs', 'hrv', 'shn', 'tuk', 'slk', 'bak', 'ces', 'fin', 'acm', 'ibo', 'tpi', 'arz', 'bho', 'sot', 'gaz', 'sna', 'afr', 'als', 'snd', 'szl', 'aeb', 'hat', 'som', 'aka', 'cat', 'uig', 'tat', 'taq', 'kac', 'hye', 'kea', 'amh', 'tzm', 'fij', 'ukr', 'kor', 'rus', 'nya', 'jav', 'bos', 'run', 'tel', 'ltz', 'mlt', 'dzo', 'hne', 'lij', 'dyu', 'kas', 'kmb', 'srd', 'fao', 'lvs', 'isl', 'fuv', 'dan', 'pol', 'nld', 'kin', 'tir', 'ayr', 'zsm', 'swe', 'acq', 'yor', 'por', 'ars', 'bul', 'kaz', 'deu', 'guj'} | BitextMining | s2s | | {'dev': 997, 'devtest': 1012} | | -| [FloresClusteringS2S](https://huggingface.co/datasets/facebook/flores) | {'spa'} | Clustering | s2s | | | | -| [GerDaLIR](https://github.com/lavis-nlp/GerDaLIR) | {'deu'} | Retrieval | s2p | | | | -| [GerDaLIRSmall](https://github.com/lavis-nlp/GerDaLIR) | {'deu'} | Retrieval | p2p | [Legal] | | | -| [GermanDPR](https://huggingface.co/datasets/deepset/germandpr) | {'deu'} | Retrieval | s2p | | | | -| [GermanQuAD-Retrieval](https://www.kaggle.com/datasets/GermanQuAD) | {'deu'} | Retrieval | s2p | | | | -| [GermanSTSBenchmark](https://github.com/t-systems-on-site-services-gmbh/german-STSbenchmark) | {'deu'} | STS | s2s | | | | -| [GreekLegalCodeClassification](https://arxiv.org/abs/2109.15298) | {'ell'} | Classification | s2s | [Legal] | {'validation': 2048, 'test': 2048} | {'validation': 4046.8, 'test': 4200.8} | -| [GreekSentimentClassification](https://huggingface.co/datasets/sepidmnorozy/Greek_sentiment) | {'ell'} | Classification | s2s | [Reviews] | {'validation': 383, 'test': 767} | {'validation': 208.0, 'test': 206.0} | -| [GujaratiNewsClassification](https://github.com/goru001/nlp-for-gujarati) | {'guj'} | Classification | s2s | [News] | {'train': 5269, 'test': 1318} | {'train': 61.95, 'test': 61.91} | -| [HALClusteringS2S](https://huggingface.co/datasets/lyon-nlp/clustering-hal-s2s) | {'fra'} | Clustering | s2s | | | | -| [HagridRetrieval](https://github.com/project-miracl/hagrid) | {'eng'} | Retrieval | s2p | | | | -| [HindiDiscourseClassification](https://aclanthology.org/2020.lrec-1.149/) | {'hin'} | Classification | s2s | [Fiction, Social] | {'train': 2048} | {'train': 79.23828125} | -| [HotelReviewSentimentClassification](https://link.springer.com/chapter/10.1007/978-3-319-67056-0_3) (Elnagar et al., 2018) | {'ara'} | Classification | s2s | [Reviews] | {'train': 2048} | {'train': 137.2} | -| [HotpotQA](https://hotpotqa.github.io/) | {'eng'} | Retrieval | s2p | | | | -| [HotpotQA-PL](https://hotpotqa.github.io/) | {'pol'} | Retrieval | s2p | | | | -| [HunSum2AbstractiveRetrieval](https://arxiv.org/abs/2404.03555) (Botond Barta, 2024) | {'hun'} | Retrieval | s2p | [News] | {'test': 1998} | {'test': 2462.2177177177177} | -| [IFlyTek](https://www.cluebenchmarks.com/introduce.html) | {'cmn'} | Classification | s2s | | | | -| [IN22ConvBitextMining](https://huggingface.co/datasets/ai4bharat/IN22-Conv) (Jay Gala, 2023) | {'san', 'ory', 'kan', 'doi', 'gom', 'mar', 'npi', 'brx', 'eng', 'tel', 'asm', 'hin', 'kas', 'mai', 'mni', 'mal', 'sat', 'pan', 'ben', 'snd', 'urd', 'tam', 'guj'} | BitextMining | s2s | [Social, Spoken, Fiction] | {'conv': 1503} | {'conv': 54.3} | -| [IN22GenBitextMining](https://huggingface.co/datasets/ai4bharat/IN22-Gen) (Jay Gala, 2023) | {'san', 'ory', 'kan', 'doi', 'gom', 'mar', 'npi', 'brx', 'eng', 'tel', 'asm', 'hin', 'kas', 'mai', 'mni', 'mal', 'sat', 'pan', 'ben', 'snd', 'urd', 'tam', 'guj'} | BitextMining | s2s | [Web, Legal, Government, News, Religious, Non-fiction] | {'gen': 1024} | {'gen': 156.7} | -| [ImdbClassification](http://www.aclweb.org/anthology/P11-1015) | {'eng'} | Classification | p2p | | {'test': 25000} | {'test': 1293.8} | -| [IndicCrosslingualSTS](https://huggingface.co/datasets/jaygala24/indic_sts) (Ramesh et al., 2022) | {'kan', 'eng', 'tel', 'mal', 'pan', 'asm', 'ory', 'ben', 'hin', 'urd', 'mar', 'tam', 'guj'} | STS | s2s | [News, Non-fiction, Web, Spoken, Government] | {'test': 10020} | {'test': 76.22} | -| [IndicLangClassification](https://arxiv.org/abs/2305.15814) | {'san', 'ory', 'kan', 'doi', 'gom', 'mar', 'npi', 'brx', 'tel', 'asm', 'hin', 'kas', 'mai', 'mni', 'mal', 'sat', 'pan', 'ben', 'snd', 'urd', 'tam', 'guj'} | Classification | s2s | [Web, Non-fiction] | {'test': 30418} | {'test': 106.5} | -| [IndicReviewsClusteringP2P](https://arxiv.org/abs/2212.05409) (Sumanth Doddapaneni, 2022) | {'kan', 'brx', 'tel', 'mal', 'pan', 'asm', 'ory', 'ben', 'hin', 'urd', 'mar', 'tam', 'guj'} | Clustering | p2p | [Reviews] | {'test': 1000} | {'test': 137.6} | -| [IndicSentimentClassification](https://arxiv.org/abs/2212.05409) (Sumanth Doddapaneni, 2022) | {'kan', 'brx', 'tel', 'mal', 'pan', 'asm', 'ory', 'ben', 'hin', 'urd', 'mar', 'tam', 'guj'} | Classification | s2s | [Reviews] | {'test': 1000} | {'test': 137.6} | -| [IndonesianIdClickbaitClassification](http://www.sciencedirect.com/science/article/pii/S2352340920311252) | {'ind'} | Classification | s2s | [News] | {'train': 2048} | {'train': 64.28} | -| [IsiZuluNewsClassification](https://huggingface.co/datasets/dsfsi/za-isizulu-siswati-news) (Madodonga et al., 2023) | {'zul'} | Classification | s2s | [News] | {'train': 752} | {'train': 43.1} | -| [Itacola](https://aclanthology.org/2021.findings-emnlp.250/) | {'ita'} | Classification | s2s | [Non-fiction, Spoken] | {'train': 7801, 'test': 975} | {'train': 35.95, 'test': 36.67} | -| [JDReview](https://aclanthology.org/2023.nodalida-1.20/) | {'cmn'} | Classification | s2s | | | | -| [JSTS](https://aclanthology.org/2022.lrec-1.317.pdf#page=2.00) | {'jpn'} | STS | s2s | [Web] | {'valudtion': 1457} | {'valudtion': 46.34} | -| [JaQuADRetrieval](https://arxiv.org/abs/2202.01764) (ByungHoon So, 2022) | {'jpn'} | Retrieval | p2p | [Encyclopaedic, Non-fiction] | {'validation': 2048} | {'validation': 400.75} | -| [KLUE-STS](https://arxiv.org/abs/2105.09680) (Sungjoon Park, 2021) | {'kor'} | STS | s2s | [Reviews, News, Spoken] | {'validation': 519} | {'validation': 33.178227360308284} | -| [KannadaNewsClassification](https://github.com/goru001/nlp-for-kannada) | {'kan'} | Classification | s2s | [News] | {'train': 6460} | {'train': 65.88} | -| Ko-StrategyQA | {'kor'} | Retrieval | s2p | | | | -| Ko-miracl | {'kor'} | Retrieval | s2p | | | | -| [KorSTS](https://arxiv.org/abs/2004.03289) (Ham et al., 2020) | {'kor'} | STS | s2s | [News, Web] | {'test': 1379} | {'test': 29.279433139534884} | -| [KurdishSentimentClassification](https://link.springer.com/article/10.1007/s10579-023-09716-6) (Badawi et al., 2024) | {'kur'} | Classification | s2s | [Web] | {'train': 6000, 'test': 1987} | {'train': 59.38, 'test': 56.11} | -| [LCQMC](https://aclanthology.org/2021.emnlp-main.357) | {'cmn'} | STS | s2s | | | | -| [LEMBNarrativeQARetrieval](https://huggingface.co/datasets/dwzhu/LongEmbed) | {'eng'} | Retrieval | s2p | [Fiction, Non-fiction] | {'test': 10804} | {'test': 326399.3} | -| [LEMBNeedleRetrieval](https://huggingface.co/datasets/dwzhu/LongEmbed) | {'eng'} | Retrieval | s2p | [Academic, Blog] | {'test_256': 150, 'test_512': 150, 'test_1024': 150, 'test_2048': 150, 'test_4096': 150, 'test_8192': 150, 'test_16384': 150, 'test_32768': 150} | {'test_256': 1074.4, 'test_512': 2067.0, 'test_1024': 4129.5, 'test_2048': 8513.4, 'test_4096': 17452.7, 'test_8192': 35261.6, 'test_16384': 72113.7, 'test_32768': 141829.0} | -| [LEMBPasskeyRetrieval](https://huggingface.co/datasets/dwzhu/LongEmbed) | {'eng'} | Retrieval | s2p | [Fiction] | {'test_256': 150, 'test_512': 150, 'test_1024': 150, 'test_2048': 150, 'test_4096': 150, 'test_8192': 150, 'test_16384': 150, 'test_32768': 150} | {'test_256': 914.9, 'test_512': 1823.0, 'test_1024': 3644.7, 'test_2048': 7280.0, 'test_4096': 14555.5, 'test_8192': 29108.1, 'test_16384': 58213.9, 'test_32768': 116417.9} | -| [LEMBQMSumRetrieval](https://huggingface.co/datasets/dwzhu/LongEmbed) | {'eng'} | Retrieval | s2p | [Spoken] | {'test': 1724} | {'test': 56136.4} | -| [LEMBSummScreenFDRetrieval](https://huggingface.co/datasets/dwzhu/LongEmbed) | {'eng'} | Retrieval | s2p | [Spoken] | {'validation': 672} | {'validation': 31445.8} | -| [LEMBWikimQARetrieval](https://huggingface.co/datasets/dwzhu/LongEmbed) (Ho et al., 2020) | {'eng'} | Retrieval | s2p | [Encyclopaedic] | {'test': 500} | {'test': 37513.0} | -| [LccSentimentClassification](https://github.com/fnielsen/lcc-sentiment) | {'dan'} | Classification | s2s | | {'test': 150} | {'test': 118.7} | -| [LeCaRDv2](https://github.com/THUIR/LeCaRDv2) | {'zho'} | Retrieval | p2p | [Legal] | | | -| [LegalBenchConsumerContractsQA](https://huggingface.co/datasets/nguha/legalbench/viewer/consumer_contracts_qa) | {'eng'} | Retrieval | s2p | [Legal] | | | -| [LegalBenchCorporateLobbying](https://huggingface.co/datasets/nguha/legalbench/viewer/corporate_lobbying) | {'eng'} | Retrieval | s2p | [Legal] | | | -| [LegalQuAD](https://github.com/Christoph911/AIKE2021_Appendix) | {'deu'} | Retrieval | s2p | [Legal] | | | -| [LegalSummarization](https://github.com/lauramanor/legal_summarization) | {'eng'} | Retrieval | s2p | [Legal] | | | -| [MIRACLReranking](https://project-miracl.github.io/) | {'spa', 'deu'} | Reranking | s2s | | | | -| MIRACLRetrieval | {'spa', 'deu'} | Retrieval | s2p | | | | -| [MLSUMClusteringP2P](https://huggingface.co/datasets/mlsum) | {'fra'} | Clustering | p2p | | | | -| [MLSUMClusteringS2S](https://huggingface.co/datasets/mlsum) | {'fra'} | Clustering | s2s | | | | -| [MMarcoReranking](https://github.com/unicamp-dl/mMARCO) | {'cmn'} | Reranking | s2s | | | | -| [MMarcoRetrieval](https://arxiv.org/abs/2309.07597) | {'cmn'} | Retrieval | s2p | | | | -| [MSMARCO](https://microsoft.github.io/msmarco/) | {'eng'} | Retrieval | s2p | | | | -| [MSMARCO-PL](https://microsoft.github.io/msmarco/) | {'pol'} | Retrieval | s2p | | | | -| [MSMARCOv2](https://microsoft.github.io/msmarco/TREC-Deep-Learning.html) | {'eng'} | Retrieval | s2p | | | | -| [MTOPDomainClassification](https://arxiv.org/pdf/2008.09335.pdf) | {'spa', 'eng', 'fra', 'hin', 'tha', 'deu'} | Classification | s2s | | {'validation': 2235, 'test': 4386} | {'validation': 36.5, 'test': 36.8} | -| [MTOPIntentClassification](https://arxiv.org/pdf/2008.09335.pdf) | {'spa', 'eng', 'fra', 'hin', 'tha', 'deu'} | Classification | s2s | | {'validation': 2235, 'test': 4386} | {'validation': 36.5, 'test': 36.8} | -| [MacedonianTweetSentimentClassification](https://aclanthology.org/R15-1034/) | {'mkd'} | Classification | s2s | [Social] | {'test': 1139} | {'test': 67.6} | -| [MalteseSentimentClassification](https://arxiv.org/abs/2009.08712) | {'mlt'} | Classification | s2s | [Reviews] | {'validation': 85, 'test': 171} | {'validation': 119.7, 'test': 132.4} | -| [MasakhaNEWSClassification](https://arxiv.org/abs/2304.09972) | {'eng', 'xho', 'swa', 'run', 'amh', 'fra', 'sna', 'pcm', 'tir', 'lin', 'ibo', 'hau', 'yor', 'lug', 'orm', 'som'} | Classification | s2s | | {'test': 422} | {'test': 5116.6} | -| [MasakhaNEWSClusteringP2P](https://huggingface.co/datasets/masakhane/masakhanews) | {'eng', 'xho', 'swa', 'run', 'amh', 'fra', 'sna', 'pcm', 'tir', 'lin', 'ibo', 'hau', 'yor', 'lug', 'orm', 'som'} | Clustering | p2p | | | | -| [MasakhaNEWSClusteringS2S](https://huggingface.co/datasets/masakhane/masakhanews) | {'eng', 'xho', 'swa', 'run', 'amh', 'fra', 'sna', 'pcm', 'tir', 'lin', 'ibo', 'hau', 'yor', 'lug', 'orm', 'som'} | Clustering | s2s | | | | -| [MassiveIntentClassification](https://arxiv.org/abs/2204.08582#:~:text=MASSIVE%20contains%201M%20realistic%2C%20parallel,diverse%20languages%20from%2029%20genera.) | {'spa', 'swa', 'tgl', 'sqi', 'msa', 'hun', 'vie', 'slv', 'ell', 'heb', 'hye', 'kan', 'tur', 'amh', 'fra', 'kor', 'tha', 'rus', 'cym', 'eng', 'jav', 'ind', 'ron', 'por', 'tam', 'tel', 'fin', 'jpn', 'hin', 'fas', 'kat', 'ita', 'isl', 'dan', 'pol', 'nld', 'khm', 'mal', 'mon', 'afr', 'ben', 'lav', 'swe', 'mya', 'urd', 'nob', 'ara', 'cmo', 'aze', 'deu'} | Classification | s2s | | {'validation': 2033, 'test': 2974} | {'validation': 34.8, 'test': 34.6} | -| [MassiveScenarioClassification](https://arxiv.org/abs/2204.08582#:~:text=MASSIVE%20contains%201M%20realistic%2C%20parallel,diverse%20languages%20from%2029%20genera.) | {'spa', 'swa', 'tgl', 'sqi', 'msa', 'hun', 'vie', 'slv', 'ell', 'heb', 'hye', 'kan', 'tur', 'amh', 'fra', 'kor', 'tha', 'rus', 'cym', 'eng', 'jav', 'ind', 'ron', 'por', 'tam', 'tel', 'fin', 'jpn', 'hin', 'fas', 'kat', 'ita', 'isl', 'dan', 'pol', 'nld', 'khm', 'mal', 'mon', 'afr', 'ben', 'lav', 'swe', 'mya', 'urd', 'nob', 'ara', 'cmo', 'aze', 'deu'} | Classification | s2s | | {'validation': 2033, 'test': 2974} | {'validation': 34.8, 'test': 34.6} | -| [MedicalQARetrieval](https://bmcbioinformatics.biomedcentral.com/articles/10.1186/s12859-019-3119-4) (Asma et al., 2019) | {'eng'} | Retrieval | s2s | [Medical] | {'test': 2048} | {'test': 1205.9619140625} | -| [MedicalRetrieval](https://arxiv.org/abs/2203.03367) | {'cmn'} | Retrieval | s2p | | | | -| [MedrxivClusteringP2P](https://api.medrxiv.org/) | {'eng'} | Clustering | p2p | | {'test': 375000} | {'test': 1981.2} | -| [MedrxivClusteringS2S](https://api.medrxiv.org/) | {'eng'} | Clustering | s2s | | {'test': 375000} | {'test': 114.7} | -| [MindSmallReranking](https://msnews.github.io/assets/doc/ACL2020_MIND.pdf) | {'eng'} | Reranking | s2s | | {'test': 107968} | {'test': 70.9} | -| MintakaRetrieval | {'ita', 'spa', 'fra', 'por', 'jpn', 'hin', 'ara', 'deu'} | Retrieval | s2p | | | | -| [MovieReviewSentimentClassification](https://github.com/TheophileBlard/french-sentiment-analysis-with-bert) (Théophile Blard, 2020) | {'fra'} | Classification | s2s | [Reviews] | {'validation': 1024, 'test': 1024} | {'validation': 550.3, 'test': 558.1} | -| [MultiHateClassification](https://aclanthology.org/2022.woah-1.15/) | {'ita', 'spa', 'eng', 'pol', 'nld', 'fra', 'por', 'hin', 'ara', 'cmn', 'deu'} | Classification | s2s | [Constructed] | {'test': 10000} | {'test': 45.9} | -| [MultiLongDocRetrieval](https://arxiv.org/abs/2402.03216) (Jianlv Chen, 2024) | {'ita', 'spa', 'eng', 'rus', 'fra', 'por', 'jpn', 'hin', 'kor', 'tha', 'ara', 'cmn', 'deu'} | Retrieval | s2p | | | | -| [MultilingualSentiment](https://github.com/tyqiangz/multilingual-sentiment-datasets) | {'cmn'} | Classification | s2s | | | | -| [NFCorpus](https://www.cl.uni-heidelberg.de/statnlpgroup/nfcorpus/) | {'eng'} | Retrieval | s2p | | | | -| [NFCorpus-PL](https://www.cl.uni-heidelberg.de/statnlpgroup/nfcorpus/) | {'pol'} | Retrieval | s2p | | | | -| [NQ](https://ai.google.com/research/NaturalQuestions/) | {'eng'} | Retrieval | s2p | | | | -| [NQ-PL](https://ai.google.com/research/NaturalQuestions/) | {'pol'} | Retrieval | s2p | | | | -| [NTREXBitextMining](https://huggingface.co/datasets/xianf/NTREX) | {'ita', 'spa', 'eng', 'tur', 'ind', 'rus', 'fra', 'por', 'zho', 'jpn', 'hin', 'kor', 'tha', 'ara', 'vie', 'deu'} | BitextMining | s2s | [News] | {'train': 1997} | {'train': 120.0} | -| [NarrativeQARetrieval](https://metatext.io/datasets/narrativeqa) | {'eng'} | Retrieval | s2p | | | | -| [NepaliNewsClassification](https://github.com/goru001/nlp-for-nepali) | {'nep'} | Classification | s2s | [News] | {'train': 5975, 'test': 1495} | {'train': 196.61, 'test': 196.017} | -| [NeuCLIR2022Retrieval](https://neuclir.github.io/) (Lawrie et al., 2023) | {'rus', 'fas', 'zho'} | Retrieval | s2p | [News] | {'fas': 2232130, 'zho': 3179323, 'rus': 4627657} | {'fas': 3500.5143969099317, 'zho': 2543.1140667919617, 'rus': 3214.755239654659} | -| [NeuCLIR2023Retrieval](https://neuclir.github.io/) (Dawn Lawrie, 2024) | {'rus', 'fas', 'zho'} | Retrieval | s2p | [News] | {'fas': 2232092, 'zho': 3179285, 'rus': 4627619} | {'fas': 3579.508213937439, 'zho': 2704.44834488453, 'rus': 3466.8192213553616} | -| [News21InstructionRetrieval](https://arxiv.org/abs/2403.15246) (Orion Weller, 2024) | {'eng'} | InstructionRetrieval | s2p | [News] | {'eng': 60258} | {'eng': 2331.381203215969} | -| [NewsClassification](https://arxiv.org/abs/1509.01626) | {'eng'} | Classification | s2s | [News] | {'test': 7600} | {'test': 235.29} | -| [NoRecClassification](https://aclanthology.org/L18-1661/) | {'nob'} | Classification | s2s | | {'test': 2050} | {'test': 82.0} | -| [NorQuadRetrieval](https://aclanthology.org/2023.nodalida-1.17/) | {'nob'} | Retrieval | p2p | [Encyclopaedic, Non-fiction] | {'test': 2602} | {'test': 502.19} | -| [NordicLangClassification](https://aclanthology.org/2021.vardial-1.8/) | {'fao', 'isl', 'dan', 'swe', 'nob', 'nno'} | Classification | s2s | | {'test': 3000} | {'test': 78.2} | -| [NorwegianCourtsBitextMining](https://opus.nlpl.eu/ELRC-Courts_Norway-v1.php) | {'nno', 'nob'} | BitextMining | s2s | [Spoken, Legal] | {'test': 456} | {'test': 82.11} | -| [NorwegianCourtsBitextMining](https://opus.nlpl.eu/index.php) | {'nno', 'nob'} | BitextMining | s2s | | {'test': 2050} | {'test': 1884.0} | -| [NorwegianParliamentClassification](https://huggingface.co/datasets/NbAiLab/norwegian_parliament) | {'nob'} | Classification | s2s | | {'test': 1200, 'validation': 1200} | {'test': 1884.0, 'validation': 1911.0} | -| [Ocnli](https://arxiv.org/abs/2010.05444) | {'cmn'} | PairClassification | s2s | | | | -| [OnlineShopping](https://aclanthology.org/2023.nodalida-1.20/) | {'cmn'} | Classification | s2s | | | | -| [OpusparcusPC](https://gem-benchmark.com/data_cards/opusparcus) | {'eng', 'fra', 'fin', 'swe', 'rus', 'deu'} | PairClassification | s2s | | | | -| [PAC](https://arxiv.org/pdf/2211.13112.pdf) | {'pol'} | Classification | p2p | | {'test': 3453} | {'test': 185.3} | -| [PAWSX](https://aclanthology.org/2021.emnlp-main.357) | {'cmn'} | STS | s2s | | | | -| [PSC](http://www.lrec-conf.org/proceedings/lrec2014/pdf/1211_Paper.pdf) | {'pol'} | PairClassification | s2s | | | | -| [PawsX](https://arxiv.org/abs/1908.11828) | {'spa', 'eng', 'fra', 'kor', 'jpn', 'cmn', 'deu'} | PairClassification | s2s | | | | -| [PersianFoodSentimentClassification](https://hooshvare.github.io/docs/datasets/sa) (Mehrdad Farahani et al., 2020) | {'fas'} | Classification | s2s | [Reviews] | {'validation': 2048, 'test': 2048} | {'validation': 90.37, 'test': 90.58} | -| [PolEmo2.0-IN](https://aclanthology.org/K19-1092.pdf) | {'pol'} | Classification | s2s | | | | -| [PolEmo2.0-OUT](https://aclanthology.org/K19-1092.pdf) | {'pol'} | Classification | s2s | | {'test': 722} | {'test': 756.2} | -| [PpcPC](https://arxiv.org/pdf/2207.12759.pdf) | {'pol'} | PairClassification | s2s | | | | -| [PunjabiNewsClassification](https://github.com/goru001/nlp-for-punjabi/) | {'pan'} | Classification | s2s | [News] | {'train': 627, 'test': 157} | {'train': 4222.22, 'test': 4115.14} | -| [QBQTC](https://github.com/CLUEbenchmark/QBQTC/tree/main/dataset) | {'cmn'} | STS | s2s | | | | -| [Quora-PL](https://quoradata.quora.com/First-Quora-Dataset-Release-Question-Pairs) | {'pol'} | Retrieval | s2s | | | | -| [QuoraRetrieval](https://quoradata.quora.com/First-Quora-Dataset-Release-Question-Pairs) | {'eng'} | Retrieval | s2s | | | | -| [RedditClustering](https://arxiv.org/abs/2104.07081) | {'eng'} | Clustering | s2s | | {'test': 420464} | {'test': 64.7} | -| [RedditClusteringP2P](https://arxiv.org/abs/2104.07081) | {'eng'} | Clustering | p2p | | {'test': 459399} | {'test': 727.7} | -| [RestaurantReviewSentimentClassification](https://link.springer.com/chapter/10.1007/978-3-319-18117-2_2) (ElSahar et al., 2015) | {'ara'} | Classification | s2s | [Reviews] | {'train': 2048} | {'train': 231.4} | -| [Robust04InstructionRetrieval](https://arxiv.org/abs/2403.15246) (Orion Weller, 2024) | {'eng'} | InstructionRetrieval | s2p | [News] | {'eng': 85290} | {'eng': 2680.043891349965} | -| [RomaTalesBitextMining](https://idoc.pub/documents/idocpub-zpnxm9g35ylv) | {'hun', 'rom'} | BitextMining | s2s | [Fiction] | {'test': 215} | {'test': 316.8046511627907} | -| [RomaniBibleClustering](https://romani.global.bible/info) | {'rom'} | Clustering | p2p | [Religious] | {'test': 2048} | {'test': 132.2} | -| [RomanianSentimentClassification](https://arxiv.org/abs/2009.08712) (Dumitrescu et al., 2020) | {'ron'} | Classification | s2s | [Reviews] | {'test': 2048} | {'test': 67.6} | -| [RonSTS](https://openreview.net/forum?id=JH61CD7afTv) (Dumitrescu et al., 2021) | {'ron'} | STS | s2s | [News, Social, Web] | {'test': 1379} | {'test': 60.5} | -| [SCIDOCS](https://allenai.org/data/scidocs) | {'eng'} | Retrieval | s2p | | | | -| [SCIDOCS-PL](https://allenai.org/data/scidocs) | {'pol'} | Retrieval | s2p | | | | -| [SICK-E-PL](https://aclanthology.org/2020.lrec-1.207) | {'pol'} | PairClassification | s2s | | | | -| [SICK-R](https://aclanthology.org/2020.lrec-1.207) | {'eng'} | STS | s2s | | | | -| [SICK-R-PL](https://aclanthology.org/2020.lrec-1.207) | {'pol'} | STS | s2s | | {'test': 9812} | {'test': 42.8} | -| [SICKFr](https://huggingface.co/datasets/Lajavaness/SICK-fr) | {'fra'} | STS | s2s | | | | -| [SNLClustering](https://huggingface.co/datasets/navjordj/SNL_summarization) (Navjord et al., 2023) | {'nob'} | Clustering | p2p | [Encyclopaedic, Non-fiction] | {'test': 2048} | {'test': 1101.3} | -| [SNLRetrieval](https://huggingface.co/datasets/navjordj/SNL_summarization) (Navjord et al., 2023) | {'nob'} | Retrieval | p2p | [Encyclopaedic, Non-fiction] | {'test': 2048} | {'test': 1101.3} | -| [STS12](https://www.aclweb.org/anthology/S12-1051.pdf) | {'eng'} | STS | s2s | | | | -| [STS13](https://www.aclweb.org/anthology/S13-1004/) | {'eng'} | STS | s2s | | | | -| [STS14](https://www.aclweb.org/anthology/S14-1002) | {'eng'} | STS | s2s | | | | -| [STS15](https://www.aclweb.org/anthology/S15-2010) | {'eng'} | STS | s2s | | | | -| [STS16](https://www.aclweb.org/anthology/S16-1001) | {'eng'} | STS | s2s | | | | -| [STS17](http://alt.qcri.org/semeval2016/task1/) | {'ita', 'spa', 'eng', 'tur', 'nld', 'fra', 'kor', 'ara', 'deu'} | STS | s2s | | {'test': 500} | {'test': 43.3} | -| [STS22](https://competitions.codalab.org/competitions/33835) | {'ita', 'spa', 'eng', 'tur', 'pol', 'rus', 'fra', 'ara', 'cmn', 'deu'} | STS | p2p | | {'test': 8060} | {'train': 1992.8} | -| [STSB](https://aclanthology.org/2021.emnlp-main.357) | {'cmn'} | STS | s2s | | | | -| [STSBenchmark](https://github.com/PhilipMay/stsb-multi-mt/) | {'eng'} | STS | s2s | | | | -| [STSBenchmarkMultilingualSTS](https://github.com/PhilipMay/stsb-multi-mt/) | {'ita', 'spa', 'eng', 'pol', 'nld', 'rus', 'fra', 'por', 'cmn', 'deu'} | STS | s2s | | | | -| [STSES](https://huggingface.co/datasets/PlanTL-GOB-ES/sts-es) | {'spa'} | STS | s2s | | | | -| [ScalaDaClassification](https://aclanthology.org/2023.nodalida-1.20/) | {'dan'} | Classification | s2s | | {'test': 1024} | {'test': 109.4} | -| [ScalaNbClassification](https://aclanthology.org/2023.nodalida-1.20/) | {'nob'} | Classification | s2s | | {'test': 1024} | {'test': 98.4} | -| [ScalaNnClassification](https://aclanthology.org/2023.nodalida-1.20/) | {'nno'} | Classification | s2s | | {'test': 1024} | {'test': 104.8} | -| [ScalaSvClassification](https://aclanthology.org/2023.nodalida-1.20/) | {'swe'} | Classification | s2s | | {'test': 1024} | {'test': 98.3} | -| [SciDocsRR](https://allenai.org/data/scidocs) | {'eng'} | Reranking | s2s | [Academic, Non-fiction] | {'test': 19599} | {'test': 69.0} | -| [SciFact](https://github.com/allenai/scifact) | {'eng'} | Retrieval | s2p | | | | -| [SciFact-PL](https://github.com/allenai/scifact) | {'pol'} | Retrieval | s2p | | | | -| [SentimentAnalysisHindi](https://huggingface.co/datasets/OdiaGenAI/sentiment_analysis_hindi) (Shantipriya Parida, 2023) | {'hin'} | Classification | s2s | [Reviews] | {'train': 2497} | {'train': 81.29} | -| [SiswatiNewsClassification](https://huggingface.co/datasets/dsfsi/za-isizulu-siswati-news) (Madodonga et al., 2023) | {'ssw'} | Classification | s2s | [News] | {'train': 80} | {'train': 354.2} | -| [SlovakSentimentClassification](https://huggingface.co/datasets/sepidmnorozy/Slovak_sentiment) | {'slk'} | Classification | s2s | [Reviews] | {'validation': 522, 'test': 1040} | {'validation': 84.96, 'test': 91.95} | -| [SpanishNewsClusteringP2P](https://www.kaggle.com/datasets/kevinmorgado/spanish-news-classification) | {'spa'} | Clustering | p2p | | | | -| [SpanishPassageRetrievalS2P](https://mklab.iti.gr/results/spanish-passage-retrieval-dataset/) | {'spa'} | Retrieval | s2p | | | | -| [SpanishPassageRetrievalS2S](https://mklab.iti.gr/results/spanish-passage-retrieval-dataset/) | {'spa'} | Retrieval | s2s | | | | -| [SprintDuplicateQuestions](https://www.aclweb.org/anthology/D18-1131/) | {'eng'} | PairClassification | s2s | | {'validation': 101000, 'test': 101000} | {'validation': 65.2, 'test': 67.9} | -| [StackExchangeClustering](https://arxiv.org/abs/2104.07081) | {'eng'} | Clustering | s2s | | {'test': 373850} | {'test': 57.0} | -| [StackExchangeClusteringP2P](https://arxiv.org/abs/2104.07081) | {'eng'} | Clustering | p2p | | {'test': 75000} | {'test': 1090.7} | -| [StackOverflowDupQuestions](https://www.microsoft.com/en-us/research/uploads/prod/2019/03/nl4se18LinkSO.pdf) | {'eng'} | Reranking | s2s | | {'test': 3467} | {'test': 49.8} | -| [SummEval](https://tabilab.cmpe.boun.edu.tr/BIOSSES/DataSet.html) | {'eng'} | Summarization | p2p | | {'test': 2800} | {'test': 359.8} | -| [SummEvalFr](https://github.com/Yale-LILY/SummEval) | {'fra'} | Summarization | p2p | | | | -| [SweFaqRetrieval](https://spraakbanken.gu.se/en/resources/superlim) | {'swe'} | Retrieval | s2s | [Government, Non-fiction] | {'test': 1024} | {'test': 195.44} | -| [SweRecClassification](https://aclanthology.org/2023.nodalida-1.20/) | {'swe'} | Classification | s2s | | {'test': 1024} | {'test': 318.8} | -| [SwednClusteringP2P](https://spraakbanken.gu.se/en/resources/swedn) (Monsen et al., 2021) | {'swe'} | Clustering | p2p | [News, Non-fiction] | {'all': 2048} | {'all': 1619.71} | -| [SwednClusteringS2S](https://spraakbanken.gu.se/en/resources/swedn) (Monsen et al., 2021) | {'swe'} | Clustering | s2s | [News, Non-fiction] | {'all': 2048} | {'all': 1619.71} | -| [SwednRetrieval](https://spraakbanken.gu.se/en/resources/swedn) (Monsen et al., 2021) | {'swe'} | Retrieval | p2p | [News, Non-fiction] | {'test': 2048} | {'test': 1946.35} | -| [SyntecReranking](https://huggingface.co/datasets/lyon-nlp/mteb-fr-reranking-syntec-s2p) | {'fra'} | Reranking | s2p | | | | -| [SyntecRetrieval](https://huggingface.co/datasets/lyon-nlp/mteb-fr-retrieval-syntec-s2p) | {'fra'} | Retrieval | s2p | | {'test': 90} | {'test': 62.0} | -| [T2Reranking](https://arxiv.org/abs/2304.03679) | {'cmn'} | Reranking | s2s | | | | -| [T2Retrieval](https://arxiv.org/abs/2304.03679) | {'cmn'} | Retrieval | s2p | | | | -| [TNews](https://www.cluebenchmarks.com/introduce.html) | {'cmn'} | Classification | s2s | | | | -| [TRECCOVID](https://ir.nist.gov/covidSubmit/index.html) | {'eng'} | Retrieval | s2p | | | | -| [TRECCOVID-PL](https://ir.nist.gov/covidSubmit/index.html) | {'pol'} | Retrieval | s2p | | | | -| [TV2Nordretrieval](https://huggingface.co/datasets/alexandrainst/nordjylland-news-summarization) | {'dan'} | Retrieval | p2p | [News, Non-fiction] | {'test': 4096} | {'test': 784.11} | -| [Tatoeba](https://github.com/facebookresearch/LASER/tree/main/data/tatoeba/v1) | {'pam', 'tgl', 'sqi', 'kzj', 'uzb', 'vie', 'slv', 'heb', 'yid', 'tur', 'pes', 'bre', 'tha', 'mar', 'arq', 'eng', 'yue', 'wuu', 'oci', 'ina', 'kab', 'nds', 'gla', 'cmn', 'ang', 'kat', 'khm', 'mal', 'cor', 'ben', 'nob', 'epo', 'max', 'ile', 'gle', 'hun', 'ell', 'xho', 'swg', 'bel', 'ceb', 'srp', 'ido', 'awa', 'cym', 'ron', 'ind', 'lit', 'eus', 'hin', 'lfn', 'nno', 'ita', 'glg', 'swh', 'pms', 'urd', 'csb', 'jpn', 'mkd', 'cha', 'ara', 'tam', 'est', 'spa', 'war', 'fra', 'ast', 'cbk', 'hrv', 'tuk', 'slk', 'ces', 'fin', 'dtp', 'arz', 'kur', 'lat', 'mhr', 'mon', 'afr', 'orv', 'cat', 'nov', 'uig', 'ber', 'tat', 'hye', 'amh', 'hsb', 'ukr', 'kor', 'gsw', 'tzl', 'rus', 'jav', 'tel', 'bos', 'fry', 'fao', 'lvs', 'isl', 'dan', 'pol', 'nld', 'zsm', 'swe', 'dsb', 'por', 'bul', 'kaz', 'aze', 'deu'} | BitextMining | s2s | | {'test': 2000} | {'test': 39.4} | -| [TenKGnadClusteringP2P](https://tblock.github.io/10kGNAD/) | {'deu'} | Clustering | p2p | | {'test': 45914} | {'test': 2641.03} | -| [TenKGnadClusteringS2S](https://tblock.github.io/10kGNAD/) | {'deu'} | Clustering | s2s | | {'test': 45914} | {'test': 50.96} | -| [ThuNewsClusteringP2P](http://thuctc.thunlp.org/) | {'cmn'} | Clustering | p2p | | | | -| [ThuNewsClusteringS2S](http://thuctc.thunlp.org/) | {'cmn'} | Clustering | s2s | | | | -| [Touche2020](https://webis.de/events/touche-20/shared-task-1.html) | {'eng'} | Retrieval | s2p | | | | -| [ToxicConversationsClassification](https://www.kaggle.com/competitions/jigsaw-unintended-bias-in-toxicity-classification/overview) | {'eng'} | Classification | s2s | | {'test': 50000} | {'test': 296.6} | -| [TurHistQuadRetrieval](https://github.com/okanvk/Turkish-Reading-Comprehension-Question-Answering-Dataset) (Soygazi et al., 2021) | {'tur'} | Retrieval | p2p | [Encyclopaedic, Non-fiction, Academic] | {'test': 1330} | {'test': 1513.83} | -| [TurkishMovieSentimentClassification](https://www.win.tue.nl/~mpechen/publications/pubs/MT_WISDOM2013.pdf) (Erkin Demirtas, 2013) | {'tur'} | Classification | s2s | [Reviews] | {'train': 7972, 'test': 2644} | {'train': 141.03, 'test': 141.5} | -| [TurkishProductSentimentClassification](https://www.win.tue.nl/~mpechen/publications/pubs/MT_WISDOM2013.pdf) (Erkin Demirtas, 2013) | {'tur'} | Classification | s2s | [Reviews] | {'train': 4800, 'test': 800} | {'train': 247.75, 'test': 246.85} | -| [TweetEmotionClassification](https://link.springer.com/chapter/10.1007/978-3-319-77116-8_8) (Al-Khatib et al., 2018) | {'ara'} | Classification | s2s | [Social] | {'train': 2048} | {'train': 78.8} | -| [TweetSarcasmClassification](https://aclanthology.org/2020.osact-1.5/) | {'ara'} | Classification | s2s | [Social] | {'test': 2110} | {'test': 102.1} | -| [TweetSentimentExtractionClassification](https://www.kaggle.com/competitions/tweet-sentiment-extraction/overview) | {'eng'} | Classification | s2s | | {'test': 3534} | {'test': 67.8} | -| [TwentyNewsgroupsClustering](https://scikit-learn.org/0.19/datasets/twenty_newsgroups.html) | {'eng'} | Clustering | s2s | | {'test': 59545} | {'test': 32.0} | -| [TwitterHjerneRetrieval](https://huggingface.co/datasets/sorenmulli/da-hashtag-twitterhjerne) (Holm et al., 2024) | {'dan'} | Retrieval | p2p | [Social] | {'train': 340} | {'train': 138.23} | -| [TwitterSemEval2015](https://alt.qcri.org/semeval2015/task1/) | {'eng'} | PairClassification | s2s | | {'test': 16777} | {'test': 38.3} | -| [TwitterURLCorpus](https://languagenet.github.io/) | {'eng'} | PairClassification | s2s | | {'test': 51534} | {'test': 79.5} | -| [UyghurSentimentClassification](https://arxiv.org/abs/2009.08712) | {'uig'} | Classification | s2s | [Reviews] | {'test': 841} | {'test': 245.2} | -| [VGClustering](https://huggingface.co/datasets/navjordj/VG_summarization) (Navjord et al., 2023) | {'nob'} | Clustering | p2p | [News, Non-fiction] | {'test': 2048} | {'test': 1009.65} | -| [VideoRetrieval](https://arxiv.org/abs/2203.03367) | {'cmn'} | Retrieval | s2p | | | | -| [VieMedEVBitextMining](https://aclanthology.org/2015.iwslt-evaluation.11/) (Nhu Vo, 2024) | {'vie', 'eng'} | BitextMining | s2s | [Medical] | {'test': 2048} | {'test': 139.23} | -| [VieQuADRetrieval](https://aclanthology.org/2020.coling-main.233.pdf) | {'vie'} | Retrieval | s2p | [Encyclopaedic, Non-fiction] | {'validation': 2048} | {'validation': 790.24} | -| [VieStudentFeedbackClassification](https://ieeexplore.ieee.org/document/8573337) (Nguyen et al., 2018) | {'vie'} | Classification | s2s | [Reviews] | {'test': 2048} | {'test': 14.22} | -| [WRIMEClassification](https://aclanthology.org/2021.naacl-main.169/) | {'jpn'} | Classification | s2s | [Social] | {'test': 2048} | {'test': 47.78} | -| [Waimai](https://aclanthology.org/2023.nodalida-1.20/) | {'cmn'} | Classification | s2s | | | | -| [WikiCitiesClustering](https://huggingface.co/datasets/wikipedia) | {'eng'} | Clustering | p2p | | | | -| XMarket | {'spa', 'eng', 'deu'} | Retrieval | s2p | | | | -| [XPQARetrieval](https://arxiv.org/abs/2305.09249) | {'ita', 'spa', 'pol', 'fra', 'por', 'tam', 'jpn', 'hin', 'kor', 'ara', 'cmn', 'deu'} | Retrieval | s2p | | | | -| [YelpReviewFullClassification](https://arxiv.org/abs/1509.01626) (Zhang et al., 2015) | {'eng'} | Classification | s2s | [Reviews] | {'test': 50000} | | -| [YueOpenriceReviewClassification](https://github.com/Christainx/Dataset_Cantonese_Openrice) (Xiang et al., 2019) | {'yue'} | Classification | s2s | [Reviews] | {'test': 6161} | {'test': 173.0} | +| [AFQMC](https://aclanthology.org/2021.emnlp-main.357) | ['cmn'] | STS | s2s | | | | +| [AILACasedocs](https://zenodo.org/records/4063986) | ['eng'] | Retrieval | p2p | [Legal] | | | +| [AILAStatutes](https://zenodo.org/records/4063986) | ['eng'] | Retrieval | p2p | [Legal] | | | +| [ATEC](https://aclanthology.org/2021.emnlp-main.357) | ['cmn'] | STS | s2s | | | | +| [AllegroReviews](https://aclanthology.org/2020.acl-main.111.pdf) | ['pol'] | Classification | s2s | | {'test': 1006} | {'test': 477.2} | +| [AlloProfClusteringP2P](https://huggingface.co/datasets/lyon-nlp/alloprof) | ['fra'] | Clustering | p2p | | | | +| [AlloProfClusteringS2S](https://huggingface.co/datasets/lyon-nlp/alloprof) | ['fra'] | Clustering | s2s | | | | +| [AlloprofReranking](https://huggingface.co/datasets/antoinelb7/alloprof) | ['fra'] | Reranking | s2s | | | | +| [AlloprofRetrieval](https://huggingface.co/datasets/antoinelb7/alloprof) | ['fra'] | Retrieval | s2p | | | | +| [AmazonCounterfactualClassification](https://arxiv.org/abs/2104.06893) | ['deu', 'eng', 'jpn'] | Classification | s2s | | {'validation': 335, 'test': 670} | {'validation': 109.2, 'test': 106.1} | +| [AmazonPolarityClassification](https://huggingface.co/datasets/amazon_polarity) | ['eng'] | Classification | s2s | | {'test': 400000} | {'test': 431.4} | +| [AmazonReviewsClassification](https://arxiv.org/abs/2010.02573) | ['cmn', 'deu', 'eng', 'fra', 'jpn', 'spa'] | Classification | s2s | | {'validation': 30000, 'test': 30000} | {'validation': 159.2, 'test': 160.4} | +| [AngryTweetsClassification](https://aclanthology.org/2021.nodalida-main.53/) | ['dan'] | Classification | s2s | | {'test': 1050} | {'test': 156.1} | +| [ArguAna](http://argumentation.bplaced.net/arguana/data) | ['eng'] | Retrieval | s2p | | | | +| [ArguAna-PL](https://huggingface.co/datasets/clarin-knext/arguana-pl) | ['pol'] | Retrieval | s2p | | | | +| [ArmenianParaphrasePC](https://github.com/ivannikov-lab/arpa-paraphrase-corpus) (Arthur Malajyan, 2020) | ['hye'] | PairClassification | s2s | [News] | {'train': 4023, 'test': 1470} | {'train': 243.81, 'test': 241.37} | +| [ArxivClassification](https://ieeexplore.ieee.org/document/8675939) (He et al., 2019) | ['eng'] | Classification | s2s | [Academic] | {'test': 2048} | | +| [ArxivClusteringP2P.v2](https://www.kaggle.com/Cornell-University/arxiv) | ['eng'] | Clustering | p2p | | {'test': 250000} | {'test': 1009.98} | +| [ArxivClusteringS2S](https://www.kaggle.com/Cornell-University/arxiv) | ['eng'] | Clustering | s2s | | {'test': 732723} | {'test': 74.0} | +| [AskUbuntuDupQuestions](https://github.com/taolei87/askubuntu) | ['eng'] | Reranking | s2s | | {'test': 2255} | {'test': 52.5} | +| [BIOSSES](https://tabilab.cmpe.boun.edu.tr/BIOSSES/DataSet.html) | ['eng'] | STS | s2s | | | | +| [BQ](https://aclanthology.org/2021.emnlp-main.357) | ['cmn'] | STS | s2s | | | | +| [BSARDRetrieval](https://huggingface.co/datasets/maastrichtlawtech/bsard) | ['fra'] | Retrieval | s2p | | | | +| [BUCC](https://comparable.limsi.fr/bucc2018/bucc2018-task.html) | ['cmn', 'deu', 'eng', 'fra', 'rus'] | BitextMining | s2s | | {'test': 641684} | {'test': 101.3} | +| [BambaraSentimentClassification](https://arxiv.org/abs/2009.08712) | ['mlt'] | Classification | s2s | [Reviews] | {'test': 673} | {'test': 29.4} | +| [Banking77Classification](https://arxiv.org/abs/2003.04807) | ['eng'] | Classification | s2s | | {'test': 3080} | {'test': 54.2} | +| [BengaliHateSpeechClassification](https://huggingface.co/datasets/bn_hate_speech) (Karim et al., 2020) | ['ben'] | Classification | s2s | [News] | {'train': 3418} | {'train': 103.42} | +| [BengaliSentimentAnalysis](https://data.mendeley.com/datasets/p6zc7krs37/4) (Sazzed et al., 2020) | ['ben'] | Classification | s2s | [Reviews] | {'train': 11807} | {'train': 69.66} | +| [BigPatentClustering](https://www.kaggle.com/datasets/big_patent) | ['eng'] | Clustering | s2s | | | | +| [BiorxivClusteringP2P](https://api.biorxiv.org/) | ['eng'] | Clustering | p2p | | {'test': 75000} | {'test': 1666.2} | +| [BiorxivClusteringS2S](https://api.biorxiv.org/) | ['eng'] | Clustering | s2s | | {'test': 75000} | {'test': 101.6} | +| [BlurbsClusteringP2P](https://www.inf.uni-hamburg.de/en/inst/ab/lt/resources/data/germeval-2019-hmc.html) | ['deu'] | Clustering | p2p | | {'test': 174637} | {'test': 664.09} | +| [BlurbsClusteringS2S](https://www.inf.uni-hamburg.de/en/inst/ab/lt/resources/data/germeval-2019-hmc.html) | ['deu'] | Clustering | s2s | | {'test': 174637} | {'test': 23.02} | +| [BornholmBitextMining](https://aclanthology.org/W19-6138/) | ['dan'] | BitextMining | s2s | | {'test': 500} | {'test': 89.7} | +| [BulgarianSentimentClassification](https://arxiv.org/abs/2009.08712) | ['bul'] | Classification | s2s | [Reviews] | {'validation': 838, 'test': 1673} | {'validation': 43.3, 'test': 46.3} | +| [BulgarianStoreReviewSentimentClassfication](https://doi.org/10.7910/DVN/TXIK9P) (Georgieva-Trifonova et al., 2018) | ['bul'] | Classification | s2s | [Reviews] | {'test': 182} | {'test': 316.7} | +| [CBD](http://2019.poleval.pl/files/poleval2019.pdf) | ['pol'] | Classification | s2s | | {'test': 1000} | {'test': 93.2} | +| [CDSC-E](https://aclanthology.org/P17-1073.pdf) | ['pol'] | PairClassification | s2s | | | | +| [CDSC-R](https://aclanthology.org/P17-1073.pdf) | ['pol'] | STS | s2s | | | | +| [CLSClusteringP2P](https://arxiv.org/abs/2209.05034) | ['cmn'] | Clustering | p2p | | | | +| [CLSClusteringS2S](https://arxiv.org/abs/2209.05034) | ['cmn'] | Clustering | s2s | | | | +| [CMedQAv1-reranking](https://github.com/zhangsheng93/cMedQA) | ['cmn'] | Reranking | s2s | | | | +| [CMedQAv2-reranking](https://github.com/zhangsheng93/cMedQA2) | ['cmn'] | Reranking | s2s | | | | +| [CQADupstackAndroidRetrieval](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/) | ['eng'] | Retrieval | s2p | | | | +| [CQADupstackEnglishRetrieval](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/) | ['eng'] | Retrieval | s2p | | | | +| [CQADupstackGamingRetrieval](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/) | ['eng'] | Retrieval | s2p | | | | +| [CQADupstackGisRetrieval](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/) | ['eng'] | Retrieval | s2p | | | | +| [CQADupstackMathematicaRetrieval](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/) | ['eng'] | Retrieval | s2p | | | | +| [CQADupstackPhysicsRetrieval](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/) | ['eng'] | Retrieval | s2p | | | | +| [CQADupstackProgrammersRetrieval](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/) | ['eng'] | Retrieval | s2p | | | | +| [CQADupstackStatsRetrieval](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/) | ['eng'] | Retrieval | s2p | | | | +| [CQADupstackTexRetrieval](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/) | ['eng'] | Retrieval | s2p | | | | +| [CQADupstackUnixRetrieval](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/) | ['eng'] | Retrieval | s2p | | | | +| [CQADupstackWebmastersRetrieval](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/) | ['eng'] | Retrieval | s2p | | | | +| [CQADupstackWordpressRetrieval](http://nlp.cis.unimelb.edu.au/resources/cqadupstack/) | ['eng'] | Retrieval | s2p | | | | +| [ClimateFEVER](https://www.sustainablefinance.uzh.ch/en/research/climate-fever.html) | ['eng'] | Retrieval | s2p | | | | +| [CmedqaRetrieval](https://aclanthology.org/2022.emnlp-main.357.pdf) | ['cmn'] | Retrieval | s2p | | | | +| [Cmnli](https://huggingface.co/datasets/clue/viewer/cmnli) | ['cmn'] | PairClassification | s2s | | | | +| [CodeSearchNetRetrieval](https://huggingface.co/datasets/code_search_net/viewer) (Husain et al., 2019) | ['go', 'java', 'javascript', 'php', 'python', 'ruby'] | Retrieval | p2p | [Programming] | {'test': 1000} | {'test': 1196.4609} | +| [Core17InstructionRetrieval](https://arxiv.org/abs/2403.15246) (Orion Weller, 2024) | ['eng'] | InstructionRetrieval | s2p | [News] | {'eng': 39470} | {'eng': 2747.2883966244726} | +| [CovidRetrieval](https://arxiv.org/abs/2203.03367) | ['cmn'] | Retrieval | s2p | | | | +| [CroatianSentimentClassification](https://arxiv.org/abs/2009.08712) | ['hrv'] | Classification | s2s | [Reviews] | {'validation': 214, 'test': 437} | {'validation': 166.9, 'test': 151.4} | +| [CzechSubjectivityClassification](https://arxiv.org/abs/2009.08712) | ['ces'] | Classification | s2s | [Reviews] | {'validation': 500, 'test': 2000} | {'validation': 108.2, 'test': 108.3} | +| [DBPedia](https://github.com/iai-group/DBpedia-Entity/) | ['eng'] | Retrieval | s2p | | | | +| [DBPedia-PL](https://github.com/iai-group/DBpedia-Entity/) | ['pol'] | Retrieval | s2p | | | | +| [DKHateClassification](https://aclanthology.org/2020.lrec-1.430/) | ['dan'] | Classification | s2s | | {'test': 329} | {'test': 104.0} | +| [DalajClassification](https://spraakbanken.gu.se/en/resources/superlim) | ['dan'] | Classification | s2s | | {'test': 444} | {'test': 243.8} | +| [DanFEVER](https://aclanthology.org/2021.nodalida-main.47/) | ['dan'] | Retrieval | p2p | [Encyclopaedic, Non-fiction] | {'train': 8897} | {'train': 124.84} | +| [DanishPoliticalCommentsClassification](https://huggingface.co/datasets/danish_political_comments) | ['dan'] | Classification | s2s | | {'train': 9010} | {'train': 69.9} | +| [DiaBlaBitextMining](https://inria.hal.science/hal-03021633) | ['eng', 'fra'] | BitextMining | s2s | | | | +| [DuRetrieval](https://aclanthology.org/2022.emnlp-main.357.pdf) | ['cmn'] | Retrieval | s2p | | | | +| [DutchBookReviewSentimentClassification](https://github.com/benjaminvdb/DBRD) (Benjamin et al., 2019) | ['nld'] | Classification | s2s | [Reviews] | {'test': 2224} | {'test': 1443.0} | +| [EcomRetrieval](https://arxiv.org/abs/2203.03367) | ['cmn'] | Retrieval | s2p | | | | +| [EightTagsClustering](https://aclanthology.org/2020.lrec-1.207.pdf) | ['pol'] | Clustering | s2s | | | | +| [EmotionClassification](https://www.aclweb.org/anthology/D18-1404) | ['eng'] | Classification | s2s | | {'validation': 2000, 'test': 2000} | {'validation': 95.3, 'test': 95.6} | +| [EstQA](https://www.semanticscholar.org/paper/Extractive-Question-Answering-for-Estonian-Language-182912IAPM-Alum%C3%A4e/ea4f60ab36cadca059c880678bc4c51e293a85d6?utm_source=direct_link) | ['est'] | Retrieval | s2p | [Encyclopaedic] | {'test': 603} | {'test': 772.5331950207469} | +| [EstonianValenceClassification](https://figshare.com/articles/dataset/Estonian_Valence_Corpus_Eesti_valentsikorpus/24517054) | ['est'] | Classification | s2s | [News] | {'train': 3270, 'test': 818} | {'train': 226.70642201834863, 'test': 231.5085574572127} | +| [FEVER](https://fever.ai/) | ['eng'] | Retrieval | s2p | | | | +| [FQuADRetrieval](https://huggingface.co/datasets/manu/fquad2_test) | ['fra'] | Retrieval | s2p | [Encyclopaedic] | {'test': 400, 'validation': 100} | {'test': 937.0, 'validation': 930.0} | +| [FalseFriendsGermanEnglish](https://drive.google.com/file/d/1jgq0nBnV-UiYNxbKNrrr2gxDEHm-DMKH/view?usp=share_link) | ['deu'] | PairClassification | s2s | | {'test': 1524} | {'test': 40.3} | +| [FiQA-PL](https://sites.google.com/view/fiqa/) | ['pol'] | Retrieval | s2p | | | | +| [FiQA2018](https://sites.google.com/view/fiqa/) | ['eng'] | Retrieval | s2p | | | | +| [FilipinoHateSpeechClassification](https://pcj.csp.org.ph/index.php/pcj/issue/download/29/PCJ%20V14%20N1%20pp1-14%202019) (Neil Vicente Cabasag et al., 2019) | ['fil'] | Classification | s2s | [Social] | {'validation': 2048, 'test': 2048} | {'validation': 88.1, 'test': 87.4} | +| [FinParaSTS](https://huggingface.co/datasets/TurkuNLP/turku_paraphrase_corpus) | ['fin'] | STS | s2s | [News, Subtitles] | {'test': 1000, 'validation': 1000} | {'test': 59.0, 'validation': 58.8} | +| [FloresBitextMining](https://huggingface.co/datasets/facebook/flores) | ['ace', 'acm', 'acq', 'aeb', 'afr', 'ajp', 'aka', 'als', 'amh', 'apc', 'arb', 'ars', 'ary', 'arz', 'asm', 'ast', 'awa', 'ayr', 'azb', 'azj', 'bak', 'bam', 'ban', 'bel', 'bem', 'ben', 'bho', 'bjn', 'bod', 'bos', 'bug', 'bul', 'cat', 'ceb', 'ces', 'cjk', 'ckb', 'crh', 'cym', 'dan', 'deu', 'dik', 'dyu', 'dzo', 'ell', 'eng', 'epo', 'est', 'eus', 'ewe', 'fao', 'fij', 'fin', 'fon', 'fra', 'fur', 'fuv', 'gaz', 'gla', 'gle', 'glg', 'grn', 'guj', 'hat', 'hau', 'heb', 'hin', 'hne', 'hrv', 'hun', 'hye', 'ibo', 'ilo', 'ind', 'isl', 'ita', 'jav', 'jpn', 'kab', 'kac', 'kam', 'kan', 'kas', 'kat', 'kaz', 'kbp', 'kea', 'khk', 'khm', 'kik', 'kin', 'kir', 'kmb', 'kmr', 'knc', 'kon', 'kor', 'lao', 'lij', 'lim', 'lin', 'lit', 'lmo', 'ltg', 'ltz', 'lua', 'lug', 'luo', 'lus', 'lvs', 'mag', 'mai', 'mal', 'mar', 'min', 'mkd', 'mlt', 'mni', 'mos', 'mri', 'mya', 'nld', 'nno', 'nob', 'npi', 'nso', 'nus', 'nya', 'oci', 'ory', 'pag', 'pan', 'pap', 'pbt', 'pes', 'plt', 'pol', 'por', 'prs', 'quy', 'ron', 'run', 'rus', 'sag', 'san', 'sat', 'scn', 'shn', 'sin', 'slk', 'slv', 'smo', 'sna', 'snd', 'som', 'sot', 'spa', 'srd', 'srp', 'ssw', 'sun', 'swe', 'swh', 'szl', 'tam', 'taq', 'tat', 'tel', 'tgk', 'tgl', 'tha', 'tir', 'tpi', 'tsn', 'tso', 'tuk', 'tum', 'tur', 'twi', 'tzm', 'uig', 'ukr', 'umb', 'urd', 'uzn', 'vec', 'vie', 'war', 'wol', 'xho', 'ydd', 'yor', 'yue', 'zho', 'zsm', 'zul'] | BitextMining | s2s | | {'dev': 997, 'devtest': 1012} | | +| [FloresClusteringS2S](https://huggingface.co/datasets/facebook/flores) | ['spa'] | Clustering | s2s | | | | +| [GerDaLIR](https://github.com/lavis-nlp/GerDaLIR) | ['deu'] | Retrieval | s2p | | | | +| [GerDaLIRSmall](https://github.com/lavis-nlp/GerDaLIR) | ['deu'] | Retrieval | p2p | [Legal] | | | +| [GermanDPR](https://huggingface.co/datasets/deepset/germandpr) | ['deu'] | Retrieval | s2p | | | | +| [GermanQuAD-Retrieval](https://www.kaggle.com/datasets/GermanQuAD) | ['deu'] | Retrieval | s2p | | | | +| [GermanSTSBenchmark](https://github.com/t-systems-on-site-services-gmbh/german-STSbenchmark) | ['deu'] | STS | s2s | | | | +| [GreekLegalCodeClassification](https://arxiv.org/abs/2109.15298) | ['ell'] | Classification | s2s | [Legal] | {'validation': 2048, 'test': 2048} | {'validation': 4046.8, 'test': 4200.8} | +| [GreekSentimentClassification](https://huggingface.co/datasets/sepidmnorozy/Greek_sentiment) | ['ell'] | Classification | s2s | [Reviews] | {'validation': 383, 'test': 767} | {'validation': 208.0, 'test': 206.0} | +| [GujaratiNewsClassification](https://github.com/goru001/nlp-for-gujarati) | ['guj'] | Classification | s2s | [News] | {'train': 5269, 'test': 1318} | {'train': 61.95, 'test': 61.91} | +| [HALClusteringS2S](https://huggingface.co/datasets/lyon-nlp/clustering-hal-s2s) | ['fra'] | Clustering | s2s | | | | +| [HagridRetrieval](https://github.com/project-miracl/hagrid) | ['eng'] | Retrieval | s2p | | | | +| [HindiDiscourseClassification](https://aclanthology.org/2020.lrec-1.149/) | ['hin'] | Classification | s2s | [Fiction, Social] | {'train': 2048} | {'train': 79.23828125} | +| [HotelReviewSentimentClassification](https://link.springer.com/chapter/10.1007/978-3-319-67056-0_3) (Elnagar et al., 2018) | ['ara'] | Classification | s2s | [Reviews] | {'train': 2048} | {'train': 137.2} | +| [HotpotQA](https://hotpotqa.github.io/) | ['eng'] | Retrieval | s2p | | | | +| [HotpotQA-PL](https://hotpotqa.github.io/) | ['pol'] | Retrieval | s2p | | | | +| [HunSum2AbstractiveRetrieval](https://arxiv.org/abs/2404.03555) (Botond Barta, 2024) | ['hun'] | Retrieval | s2p | [News] | {'test': 1998} | {'test': 2462.2177177177177} | +| [IFlyTek](https://www.cluebenchmarks.com/introduce.html) | ['cmn'] | Classification | s2s | | | | +| [IN22ConvBitextMining](https://huggingface.co/datasets/ai4bharat/IN22-Conv) (Jay Gala, 2023) | ['asm', 'ben', 'brx', 'doi', 'eng', 'gom', 'guj', 'hin', 'kan', 'kas', 'mai', 'mal', 'mar', 'mni', 'npi', 'ory', 'pan', 'san', 'sat', 'snd', 'tam', 'tel', 'urd'] | BitextMining | s2s | [Social, Spoken, Fiction] | {'conv': 1503} | {'conv': 54.3} | +| [IN22GenBitextMining](https://huggingface.co/datasets/ai4bharat/IN22-Gen) (Jay Gala, 2023) | ['asm', 'ben', 'brx', 'doi', 'eng', 'gom', 'guj', 'hin', 'kan', 'kas', 'mai', 'mal', 'mar', 'mni', 'npi', 'ory', 'pan', 'san', 'sat', 'snd', 'tam', 'tel', 'urd'] | BitextMining | s2s | [Web, Legal, Government, News, Religious, Non-fiction] | {'gen': 1024} | {'gen': 156.7} | +| [ImdbClassification](http://www.aclweb.org/anthology/P11-1015) | ['eng'] | Classification | p2p | | {'test': 25000} | {'test': 1293.8} | +| [IndicCrosslingualSTS](https://huggingface.co/datasets/jaygala24/indic_sts) (Ramesh et al., 2022) | ['asm', 'ben', 'eng', 'guj', 'hin', 'kan', 'mal', 'mar', 'ory', 'pan', 'tam', 'tel', 'urd'] | STS | s2s | [News, Non-fiction, Web, Spoken, Government] | {'test': 10020} | {'test': 76.22} | +| [IndicLangClassification](https://arxiv.org/abs/2305.15814) | ['asm', 'ben', 'brx', 'doi', 'gom', 'guj', 'hin', 'kan', 'kas', 'mai', 'mal', 'mar', 'mni', 'npi', 'ory', 'pan', 'san', 'sat', 'snd', 'tam', 'tel', 'urd'] | Classification | s2s | [Web, Non-fiction] | {'test': 30418} | {'test': 106.5} | +| [IndicReviewsClusteringP2P](https://arxiv.org/abs/2212.05409) (Sumanth Doddapaneni, 2022) | ['asm', 'ben', 'brx', 'guj', 'hin', 'kan', 'mal', 'mar', 'ory', 'pan', 'tam', 'tel', 'urd'] | Clustering | p2p | [Reviews] | {'test': 1000} | {'test': 137.6} | +| [IndicSentimentClassification](https://arxiv.org/abs/2212.05409) (Sumanth Doddapaneni, 2022) | ['asm', 'ben', 'brx', 'guj', 'hin', 'kan', 'mal', 'mar', 'ory', 'pan', 'tam', 'tel', 'urd'] | Classification | s2s | [Reviews] | {'test': 1000} | {'test': 137.6} | +| [IndonesianIdClickbaitClassification](http://www.sciencedirect.com/science/article/pii/S2352340920311252) | ['ind'] | Classification | s2s | [News] | {'train': 2048} | {'train': 64.28} | +| [IsiZuluNewsClassification](https://huggingface.co/datasets/dsfsi/za-isizulu-siswati-news) (Madodonga et al., 2023) | ['zul'] | Classification | s2s | [News] | {'train': 752} | {'train': 43.1} | +| [Itacola](https://aclanthology.org/2021.findings-emnlp.250/) | ['ita'] | Classification | s2s | [Non-fiction, Spoken] | {'train': 7801, 'test': 975} | {'train': 35.95, 'test': 36.67} | +| [JDReview](https://aclanthology.org/2023.nodalida-1.20/) | ['cmn'] | Classification | s2s | | | | +| [JSTS](https://aclanthology.org/2022.lrec-1.317.pdf#page=2.00) | ['jpn'] | STS | s2s | [Web] | {'valudtion': 1457} | {'valudtion': 46.34} | +| [JaQuADRetrieval](https://arxiv.org/abs/2202.01764) (ByungHoon So, 2022) | ['jpn'] | Retrieval | p2p | [Encyclopaedic, Non-fiction] | {'validation': 2048} | {'validation': 400.75} | +| [KLUE-STS](https://arxiv.org/abs/2105.09680) (Sungjoon Park, 2021) | ['kor'] | STS | s2s | [Reviews, News, Spoken] | {'validation': 519} | {'validation': 33.178227360308284} | +| [KannadaNewsClassification](https://github.com/goru001/nlp-for-kannada) | ['kan'] | Classification | s2s | [News] | {'train': 6460} | {'train': 65.88} | +| Ko-StrategyQA | ['kor'] | Retrieval | s2p | | | | +| Ko-miracl | ['kor'] | Retrieval | s2p | | | | +| [KorSTS](https://arxiv.org/abs/2004.03289) (Ham et al., 2020) | ['kor'] | STS | s2s | [News, Web] | {'test': 1379} | {'test': 29.279433139534884} | +| [KurdishSentimentClassification](https://link.springer.com/article/10.1007/s10579-023-09716-6) (Badawi et al., 2024) | ['kur'] | Classification | s2s | [Web] | {'train': 6000, 'test': 1987} | {'train': 59.38, 'test': 56.11} | +| [LCQMC](https://aclanthology.org/2021.emnlp-main.357) | ['cmn'] | STS | s2s | | | | +| [LEMBNarrativeQARetrieval](https://huggingface.co/datasets/dwzhu/LongEmbed) | ['eng'] | Retrieval | s2p | [Fiction, Non-fiction] | {'test': 10804} | {'test': 326399.3} | +| [LEMBNeedleRetrieval](https://huggingface.co/datasets/dwzhu/LongEmbed) | ['eng'] | Retrieval | s2p | [Academic, Blog] | {'test_256': 150, 'test_512': 150, 'test_1024': 150, 'test_2048': 150, 'test_4096': 150, 'test_8192': 150, 'test_16384': 150, 'test_32768': 150} | {'test_256': 1074.4, 'test_512': 2067.0, 'test_1024': 4129.5, 'test_2048': 8513.4, 'test_4096': 17452.7, 'test_8192': 35261.6, 'test_16384': 72113.7, 'test_32768': 141829.0} | +| [LEMBPasskeyRetrieval](https://huggingface.co/datasets/dwzhu/LongEmbed) | ['eng'] | Retrieval | s2p | [Fiction] | {'test_256': 150, 'test_512': 150, 'test_1024': 150, 'test_2048': 150, 'test_4096': 150, 'test_8192': 150, 'test_16384': 150, 'test_32768': 150} | {'test_256': 914.9, 'test_512': 1823.0, 'test_1024': 3644.7, 'test_2048': 7280.0, 'test_4096': 14555.5, 'test_8192': 29108.1, 'test_16384': 58213.9, 'test_32768': 116417.9} | +| [LEMBQMSumRetrieval](https://huggingface.co/datasets/dwzhu/LongEmbed) | ['eng'] | Retrieval | s2p | [Spoken] | {'test': 1724} | {'test': 56136.4} | +| [LEMBSummScreenFDRetrieval](https://huggingface.co/datasets/dwzhu/LongEmbed) | ['eng'] | Retrieval | s2p | [Spoken] | {'validation': 672} | {'validation': 31445.8} | +| [LEMBWikimQARetrieval](https://huggingface.co/datasets/dwzhu/LongEmbed) (Ho et al., 2020) | ['eng'] | Retrieval | s2p | [Encyclopaedic] | {'test': 500} | {'test': 37513.0} | +| [LccSentimentClassification](https://github.com/fnielsen/lcc-sentiment) | ['dan'] | Classification | s2s | | {'test': 150} | {'test': 118.7} | +| [LeCaRDv2](https://github.com/THUIR/LeCaRDv2) | ['zho'] | Retrieval | p2p | [Legal] | | | +| [LegalBenchConsumerContractsQA](https://huggingface.co/datasets/nguha/legalbench/viewer/consumer_contracts_qa) | ['eng'] | Retrieval | s2p | [Legal] | | | +| [LegalBenchCorporateLobbying](https://huggingface.co/datasets/nguha/legalbench/viewer/corporate_lobbying) | ['eng'] | Retrieval | s2p | [Legal] | | | +| [LegalQuAD](https://github.com/Christoph911/AIKE2021_Appendix) | ['deu'] | Retrieval | s2p | [Legal] | | | +| [LegalSummarization](https://github.com/lauramanor/legal_summarization) | ['eng'] | Retrieval | s2p | [Legal] | | | +| [MIRACLReranking](https://project-miracl.github.io/) | ['deu', 'spa'] | Reranking | s2s | | | | +| MIRACLRetrieval | ['deu', 'spa'] | Retrieval | s2p | | | | +| [MLSUMClusteringP2P](https://huggingface.co/datasets/mlsum) | ['fra'] | Clustering | p2p | | | | +| [MLSUMClusteringS2S](https://huggingface.co/datasets/mlsum) | ['fra'] | Clustering | s2s | | | | +| [MMarcoReranking](https://github.com/unicamp-dl/mMARCO) | ['cmn'] | Reranking | s2s | | | | +| [MMarcoRetrieval](https://arxiv.org/abs/2309.07597) | ['cmn'] | Retrieval | s2p | | | | +| [MSMARCO](https://microsoft.github.io/msmarco/) | ['eng'] | Retrieval | s2p | | | | +| [MSMARCO-PL](https://microsoft.github.io/msmarco/) | ['pol'] | Retrieval | s2p | | | | +| [MSMARCOv2](https://microsoft.github.io/msmarco/TREC-Deep-Learning.html) | ['eng'] | Retrieval | s2p | | | | +| [MTOPDomainClassification](https://arxiv.org/pdf/2008.09335.pdf) | ['deu', 'eng', 'fra', 'hin', 'spa', 'tha'] | Classification | s2s | | {'validation': 2235, 'test': 4386} | {'validation': 36.5, 'test': 36.8} | +| [MTOPIntentClassification](https://arxiv.org/pdf/2008.09335.pdf) | ['deu', 'eng', 'fra', 'hin', 'spa', 'tha'] | Classification | s2s | | {'validation': 2235, 'test': 4386} | {'validation': 36.5, 'test': 36.8} | +| [MacedonianTweetSentimentClassification](https://aclanthology.org/R15-1034/) | ['mkd'] | Classification | s2s | [Social] | {'test': 1139} | {'test': 67.6} | +| [MalteseSentimentClassification](https://arxiv.org/abs/2009.08712) | ['mlt'] | Classification | s2s | [Reviews] | {'validation': 85, 'test': 171} | {'validation': 119.7, 'test': 132.4} | +| [MasakhaNEWSClassification](https://arxiv.org/abs/2304.09972) | ['amh', 'eng', 'fra', 'hau', 'ibo', 'lin', 'lug', 'orm', 'pcm', 'run', 'sna', 'som', 'swa', 'tir', 'xho', 'yor'] | Classification | s2s | | {'test': 422} | {'test': 5116.6} | +| [MasakhaNEWSClusteringP2P](https://huggingface.co/datasets/masakhane/masakhanews) | ['amh', 'eng', 'fra', 'hau', 'ibo', 'lin', 'lug', 'orm', 'pcm', 'run', 'sna', 'som', 'swa', 'tir', 'xho', 'yor'] | Clustering | p2p | | | | +| [MasakhaNEWSClusteringS2S](https://huggingface.co/datasets/masakhane/masakhanews) | ['amh', 'eng', 'fra', 'hau', 'ibo', 'lin', 'lug', 'orm', 'pcm', 'run', 'sna', 'som', 'swa', 'tir', 'xho', 'yor'] | Clustering | s2s | | | | +| [MassiveIntentClassification](https://arxiv.org/abs/2204.08582#:~:text=MASSIVE%20contains%201M%20realistic%2C%20parallel,diverse%20languages%20from%2029%20genera.) | ['afr', 'amh', 'ara', 'aze', 'ben', 'cmo', 'cym', 'dan', 'deu', 'ell', 'eng', 'fas', 'fin', 'fra', 'heb', 'hin', 'hun', 'hye', 'ind', 'isl', 'ita', 'jav', 'jpn', 'kan', 'kat', 'khm', 'kor', 'lav', 'mal', 'mon', 'msa', 'mya', 'nld', 'nob', 'pol', 'por', 'ron', 'rus', 'slv', 'spa', 'sqi', 'swa', 'swe', 'tam', 'tel', 'tgl', 'tha', 'tur', 'urd', 'vie'] | Classification | s2s | | {'validation': 2033, 'test': 2974} | {'validation': 34.8, 'test': 34.6} | +| [MassiveScenarioClassification](https://arxiv.org/abs/2204.08582#:~:text=MASSIVE%20contains%201M%20realistic%2C%20parallel,diverse%20languages%20from%2029%20genera.) | ['afr', 'amh', 'ara', 'aze', 'ben', 'cmo', 'cym', 'dan', 'deu', 'ell', 'eng', 'fas', 'fin', 'fra', 'heb', 'hin', 'hun', 'hye', 'ind', 'isl', 'ita', 'jav', 'jpn', 'kan', 'kat', 'khm', 'kor', 'lav', 'mal', 'mon', 'msa', 'mya', 'nld', 'nob', 'pol', 'por', 'ron', 'rus', 'slv', 'spa', 'sqi', 'swa', 'swe', 'tam', 'tel', 'tgl', 'tha', 'tur', 'urd', 'vie'] | Classification | s2s | | {'validation': 2033, 'test': 2974} | {'validation': 34.8, 'test': 34.6} | +| [MedicalQARetrieval](https://bmcbioinformatics.biomedcentral.com/articles/10.1186/s12859-019-3119-4) (Asma et al., 2019) | ['eng'] | Retrieval | s2s | [Medical] | {'test': 2048} | {'test': 1205.9619140625} | +| [MedicalRetrieval](https://arxiv.org/abs/2203.03367) | ['cmn'] | Retrieval | s2p | | | | +| [MedrxivClusteringP2P](https://api.medrxiv.org/) | ['eng'] | Clustering | p2p | | {'test': 375000} | {'test': 1981.2} | +| [MedrxivClusteringS2S](https://api.medrxiv.org/) | ['eng'] | Clustering | s2s | | {'test': 375000} | {'test': 114.7} | +| [MindSmallReranking](https://msnews.github.io/assets/doc/ACL2020_MIND.pdf) | ['eng'] | Reranking | s2s | | {'test': 107968} | {'test': 70.9} | +| MintakaRetrieval | ['ara', 'deu', 'fra', 'hin', 'ita', 'jpn', 'por', 'spa'] | Retrieval | s2p | | | | +| [MovieReviewSentimentClassification](https://github.com/TheophileBlard/french-sentiment-analysis-with-bert) (Théophile Blard, 2020) | ['fra'] | Classification | s2s | [Reviews] | {'validation': 1024, 'test': 1024} | {'validation': 550.3, 'test': 558.1} | +| [MultiHateClassification](https://aclanthology.org/2022.woah-1.15/) | ['ara', 'cmn', 'deu', 'eng', 'fra', 'hin', 'ita', 'nld', 'pol', 'por', 'spa'] | Classification | s2s | [Constructed] | {'test': 10000} | {'test': 45.9} | +| [MultiLongDocRetrieval](https://arxiv.org/abs/2402.03216) (Jianlv Chen, 2024) | ['ara', 'cmn', 'deu', 'eng', 'fra', 'hin', 'ita', 'jpn', 'kor', 'por', 'rus', 'spa', 'tha'] | Retrieval | s2p | | | | +| [MultilingualSentiment](https://github.com/tyqiangz/multilingual-sentiment-datasets) | ['cmn'] | Classification | s2s | | | | +| [NFCorpus](https://www.cl.uni-heidelberg.de/statnlpgroup/nfcorpus/) | ['eng'] | Retrieval | s2p | | | | +| [NFCorpus-PL](https://www.cl.uni-heidelberg.de/statnlpgroup/nfcorpus/) | ['pol'] | Retrieval | s2p | | | | +| [NQ](https://ai.google.com/research/NaturalQuestions/) | ['eng'] | Retrieval | s2p | | | | +| [NQ-PL](https://ai.google.com/research/NaturalQuestions/) | ['pol'] | Retrieval | s2p | | | | +| [NTREXBitextMining](https://huggingface.co/datasets/xianf/NTREX) | ['ara', 'deu', 'eng', 'fra', 'hin', 'ind', 'ita', 'jpn', 'kor', 'por', 'rus', 'spa', 'tha', 'tur', 'vie', 'zho'] | BitextMining | s2s | [News] | {'train': 1997} | {'train': 120.0} | +| [NarrativeQARetrieval](https://metatext.io/datasets/narrativeqa) | ['eng'] | Retrieval | s2p | | | | +| [NepaliNewsClassification](https://github.com/goru001/nlp-for-nepali) | ['nep'] | Classification | s2s | [News] | {'train': 5975, 'test': 1495} | {'train': 196.61, 'test': 196.017} | +| [NeuCLIR2022Retrieval](https://neuclir.github.io/) (Lawrie et al., 2023) | ['fas', 'rus', 'zho'] | Retrieval | s2p | [News] | {'fas': 2232130, 'zho': 3179323, 'rus': 4627657} | {'fas': 3500.5143969099317, 'zho': 2543.1140667919617, 'rus': 3214.755239654659} | +| [NeuCLIR2023Retrieval](https://neuclir.github.io/) (Dawn Lawrie, 2024) | ['fas', 'rus', 'zho'] | Retrieval | s2p | [News] | {'fas': 2232092, 'zho': 3179285, 'rus': 4627619} | {'fas': 3579.508213937439, 'zho': 2704.44834488453, 'rus': 3466.8192213553616} | +| [News21InstructionRetrieval](https://arxiv.org/abs/2403.15246) (Orion Weller, 2024) | ['eng'] | InstructionRetrieval | s2p | [News] | {'eng': 60258} | {'eng': 2331.381203215969} | +| [NewsClassification](https://arxiv.org/abs/1509.01626) | ['eng'] | Classification | s2s | [News] | {'test': 7600} | {'test': 235.29} | +| [NoRecClassification](https://aclanthology.org/L18-1661/) | ['nob'] | Classification | s2s | | {'test': 2050} | {'test': 82.0} | +| [NorQuadRetrieval](https://aclanthology.org/2023.nodalida-1.17/) | ['nob'] | Retrieval | p2p | [Encyclopaedic, Non-fiction] | {'test': 2602} | {'test': 502.19} | +| [NordicLangClassification](https://aclanthology.org/2021.vardial-1.8/) | ['dan', 'fao', 'isl', 'nno', 'nob', 'swe'] | Classification | s2s | | {'test': 3000} | {'test': 78.2} | +| [NorwegianCourtsBitextMining](https://opus.nlpl.eu/ELRC-Courts_Norway-v1.php) | ['nno', 'nob'] | BitextMining | s2s | [Spoken, Legal] | {'test': 456} | {'test': 82.11} | +| [NorwegianCourtsBitextMining](https://opus.nlpl.eu/index.php) | ['nno', 'nob'] | BitextMining | s2s | | {'test': 2050} | {'test': 1884.0} | +| [NorwegianParliamentClassification](https://huggingface.co/datasets/NbAiLab/norwegian_parliament) | ['nob'] | Classification | s2s | | {'test': 1200, 'validation': 1200} | {'test': 1884.0, 'validation': 1911.0} | +| [Ocnli](https://arxiv.org/abs/2010.05444) | ['cmn'] | PairClassification | s2s | | | | +| [OnlineShopping](https://aclanthology.org/2023.nodalida-1.20/) | ['cmn'] | Classification | s2s | | | | +| [OpusparcusPC](https://gem-benchmark.com/data_cards/opusparcus) | ['deu', 'eng', 'fin', 'fra', 'rus', 'swe'] | PairClassification | s2s | | | | +| [PAC](https://arxiv.org/pdf/2211.13112.pdf) | ['pol'] | Classification | p2p | | {'test': 3453} | {'test': 185.3} | +| [PAWSX](https://aclanthology.org/2021.emnlp-main.357) | ['cmn'] | STS | s2s | | | | +| [PSC](http://www.lrec-conf.org/proceedings/lrec2014/pdf/1211_Paper.pdf) | ['pol'] | PairClassification | s2s | | | | +| [PawsX](https://arxiv.org/abs/1908.11828) | ['cmn', 'deu', 'eng', 'fra', 'jpn', 'kor', 'spa'] | PairClassification | s2s | | | | +| [PersianFoodSentimentClassification](https://hooshvare.github.io/docs/datasets/sa) (Mehrdad Farahani et al., 2020) | ['fas'] | Classification | s2s | [Reviews] | {'validation': 2048, 'test': 2048} | {'validation': 90.37, 'test': 90.58} | +| [PolEmo2.0-IN](https://aclanthology.org/K19-1092.pdf) | ['pol'] | Classification | s2s | | | | +| [PolEmo2.0-OUT](https://aclanthology.org/K19-1092.pdf) | ['pol'] | Classification | s2s | | {'test': 722} | {'test': 756.2} | +| [PpcPC](https://arxiv.org/pdf/2207.12759.pdf) | ['pol'] | PairClassification | s2s | | | | +| [PunjabiNewsClassification](https://github.com/goru001/nlp-for-punjabi/) | ['pan'] | Classification | s2s | [News] | {'train': 627, 'test': 157} | {'train': 4222.22, 'test': 4115.14} | +| [QBQTC](https://github.com/CLUEbenchmark/QBQTC/tree/main/dataset) | ['cmn'] | STS | s2s | | | | +| [Quora-PL](https://quoradata.quora.com/First-Quora-Dataset-Release-Question-Pairs) | ['pol'] | Retrieval | s2s | | | | +| [QuoraRetrieval](https://quoradata.quora.com/First-Quora-Dataset-Release-Question-Pairs) | ['eng'] | Retrieval | s2s | | | | +| [RedditClustering](https://arxiv.org/abs/2104.07081) | ['eng'] | Clustering | s2s | | {'test': 420464} | {'test': 64.7} | +| [RedditClusteringP2P](https://arxiv.org/abs/2104.07081) | ['eng'] | Clustering | p2p | | {'test': 459399} | {'test': 727.7} | +| [RestaurantReviewSentimentClassification](https://link.springer.com/chapter/10.1007/978-3-319-18117-2_2) (ElSahar et al., 2015) | ['ara'] | Classification | s2s | [Reviews] | {'train': 2048} | {'train': 231.4} | +| [Robust04InstructionRetrieval](https://arxiv.org/abs/2403.15246) (Orion Weller, 2024) | ['eng'] | InstructionRetrieval | s2p | [News] | {'eng': 85290} | {'eng': 2680.043891349965} | +| [RomaTalesBitextMining](https://idoc.pub/documents/idocpub-zpnxm9g35ylv) | ['hun', 'rom'] | BitextMining | s2s | [Fiction] | {'test': 215} | {'test': 316.8046511627907} | +| [RomaniBibleClustering](https://romani.global.bible/info) | ['rom'] | Clustering | p2p | [Religious] | {'test': 2048} | {'test': 132.2} | +| [RomanianSentimentClassification](https://arxiv.org/abs/2009.08712) (Dumitrescu et al., 2020) | ['ron'] | Classification | s2s | [Reviews] | {'test': 2048} | {'test': 67.6} | +| [RonSTS](https://openreview.net/forum?id=JH61CD7afTv) (Dumitrescu et al., 2021) | ['ron'] | STS | s2s | [News, Social, Web] | {'test': 1379} | {'test': 60.5} | +| [SCIDOCS](https://allenai.org/data/scidocs) | ['eng'] | Retrieval | s2p | | | | +| [SCIDOCS-PL](https://allenai.org/data/scidocs) | ['pol'] | Retrieval | s2p | | | | +| [SICK-E-PL](https://aclanthology.org/2020.lrec-1.207) | ['pol'] | PairClassification | s2s | | | | +| [SICK-R](https://aclanthology.org/2020.lrec-1.207) | ['eng'] | STS | s2s | | | | +| [SICK-R-PL](https://aclanthology.org/2020.lrec-1.207) | ['pol'] | STS | s2s | | {'test': 9812} | {'test': 42.8} | +| [SICKFr](https://huggingface.co/datasets/Lajavaness/SICK-fr) | ['fra'] | STS | s2s | | | | +| [SNLClustering](https://huggingface.co/datasets/navjordj/SNL_summarization) (Navjord et al., 2023) | ['nob'] | Clustering | p2p | [Encyclopaedic, Non-fiction] | {'test': 2048} | {'test': 1101.3} | +| [SNLRetrieval](https://huggingface.co/datasets/navjordj/SNL_summarization) (Navjord et al., 2023) | ['nob'] | Retrieval | p2p | [Encyclopaedic, Non-fiction] | {'test': 2048} | {'test': 1101.3} | +| [STS12](https://www.aclweb.org/anthology/S12-1051.pdf) | ['eng'] | STS | s2s | | | | +| [STS13](https://www.aclweb.org/anthology/S13-1004/) | ['eng'] | STS | s2s | | | | +| [STS14](https://www.aclweb.org/anthology/S14-1002) | ['eng'] | STS | s2s | | | | +| [STS15](https://www.aclweb.org/anthology/S15-2010) | ['eng'] | STS | s2s | | | | +| [STS16](https://www.aclweb.org/anthology/S16-1001) | ['eng'] | STS | s2s | | | | +| [STS17](http://alt.qcri.org/semeval2016/task1/) | ['ara', 'deu', 'eng', 'fra', 'ita', 'kor', 'nld', 'spa', 'tur'] | STS | s2s | | {'test': 500} | {'test': 43.3} | +| [STS22](https://competitions.codalab.org/competitions/33835) | ['ara', 'cmn', 'deu', 'eng', 'fra', 'ita', 'pol', 'rus', 'spa', 'tur'] | STS | p2p | | {'test': 8060} | {'train': 1992.8} | +| [STSB](https://aclanthology.org/2021.emnlp-main.357) | ['cmn'] | STS | s2s | | | | +| [STSBenchmark](https://github.com/PhilipMay/stsb-multi-mt/) | ['eng'] | STS | s2s | | | | +| [STSBenchmarkMultilingualSTS](https://github.com/PhilipMay/stsb-multi-mt/) | ['cmn', 'deu', 'eng', 'fra', 'ita', 'nld', 'pol', 'por', 'rus', 'spa'] | STS | s2s | | | | +| [STSES](https://huggingface.co/datasets/PlanTL-GOB-ES/sts-es) | ['spa'] | STS | s2s | | | | +| [ScalaDaClassification](https://aclanthology.org/2023.nodalida-1.20/) | ['dan'] | Classification | s2s | | {'test': 1024} | {'test': 109.4} | +| [ScalaNbClassification](https://aclanthology.org/2023.nodalida-1.20/) | ['nob'] | Classification | s2s | | {'test': 1024} | {'test': 98.4} | +| [ScalaNnClassification](https://aclanthology.org/2023.nodalida-1.20/) | ['nno'] | Classification | s2s | | {'test': 1024} | {'test': 104.8} | +| [ScalaSvClassification](https://aclanthology.org/2023.nodalida-1.20/) | ['swe'] | Classification | s2s | | {'test': 1024} | {'test': 98.3} | +| [SciDocsRR](https://allenai.org/data/scidocs) | ['eng'] | Reranking | s2s | [Academic, Non-fiction] | {'test': 19599} | {'test': 69.0} | +| [SciFact](https://github.com/allenai/scifact) | ['eng'] | Retrieval | s2p | | | | +| [SciFact-PL](https://github.com/allenai/scifact) | ['pol'] | Retrieval | s2p | | | | +| [SentimentAnalysisHindi](https://huggingface.co/datasets/OdiaGenAI/sentiment_analysis_hindi) (Shantipriya Parida, 2023) | ['hin'] | Classification | s2s | [Reviews] | {'train': 2497} | {'train': 81.29} | +| [SiswatiNewsClassification](https://huggingface.co/datasets/dsfsi/za-isizulu-siswati-news) (Madodonga et al., 2023) | ['ssw'] | Classification | s2s | [News] | {'train': 80} | {'train': 354.2} | +| [SlovakSentimentClassification](https://huggingface.co/datasets/sepidmnorozy/Slovak_sentiment) | ['slk'] | Classification | s2s | [Reviews] | {'validation': 522, 'test': 1040} | {'validation': 84.96, 'test': 91.95} | +| [SpanishNewsClusteringP2P](https://www.kaggle.com/datasets/kevinmorgado/spanish-news-classification) | ['spa'] | Clustering | p2p | | | | +| [SpanishPassageRetrievalS2P](https://mklab.iti.gr/results/spanish-passage-retrieval-dataset/) | ['spa'] | Retrieval | s2p | | | | +| [SpanishPassageRetrievalS2S](https://mklab.iti.gr/results/spanish-passage-retrieval-dataset/) | ['spa'] | Retrieval | s2s | | | | +| [SprintDuplicateQuestions](https://www.aclweb.org/anthology/D18-1131/) | ['eng'] | PairClassification | s2s | | {'validation': 101000, 'test': 101000} | {'validation': 65.2, 'test': 67.9} | +| [StackExchangeClustering](https://arxiv.org/abs/2104.07081) | ['eng'] | Clustering | s2s | | {'test': 373850} | {'test': 57.0} | +| [StackExchangeClusteringP2P](https://arxiv.org/abs/2104.07081) | ['eng'] | Clustering | p2p | | {'test': 75000} | {'test': 1090.7} | +| [StackOverflowDupQuestions](https://www.microsoft.com/en-us/research/uploads/prod/2019/03/nl4se18LinkSO.pdf) | ['eng'] | Reranking | s2s | | {'test': 3467} | {'test': 49.8} | +| [SummEval](https://tabilab.cmpe.boun.edu.tr/BIOSSES/DataSet.html) | ['eng'] | Summarization | p2p | | {'test': 2800} | {'test': 359.8} | +| [SummEvalFr](https://github.com/Yale-LILY/SummEval) | ['fra'] | Summarization | p2p | | | | +| [SweFaqRetrieval](https://spraakbanken.gu.se/en/resources/superlim) | ['swe'] | Retrieval | s2s | [Government, Non-fiction] | {'test': 1024} | {'test': 195.44} | +| [SweRecClassification](https://aclanthology.org/2023.nodalida-1.20/) | ['swe'] | Classification | s2s | | {'test': 1024} | {'test': 318.8} | +| [SwednClusteringP2P](https://spraakbanken.gu.se/en/resources/swedn) (Monsen et al., 2021) | ['swe'] | Clustering | p2p | [News, Non-fiction] | {'all': 2048} | {'all': 1619.71} | +| [SwednClusteringS2S](https://spraakbanken.gu.se/en/resources/swedn) (Monsen et al., 2021) | ['swe'] | Clustering | s2s | [News, Non-fiction] | {'all': 2048} | {'all': 1619.71} | +| [SwednRetrieval](https://spraakbanken.gu.se/en/resources/swedn) (Monsen et al., 2021) | ['swe'] | Retrieval | p2p | [News, Non-fiction] | {'test': 2048} | {'test': 1946.35} | +| [SyntecReranking](https://huggingface.co/datasets/lyon-nlp/mteb-fr-reranking-syntec-s2p) | ['fra'] | Reranking | s2p | | | | +| [SyntecRetrieval](https://huggingface.co/datasets/lyon-nlp/mteb-fr-retrieval-syntec-s2p) | ['fra'] | Retrieval | s2p | | {'test': 90} | {'test': 62.0} | +| [T2Reranking](https://arxiv.org/abs/2304.03679) | ['cmn'] | Reranking | s2s | | | | +| [T2Retrieval](https://arxiv.org/abs/2304.03679) | ['cmn'] | Retrieval | s2p | | | | +| [TNews](https://www.cluebenchmarks.com/introduce.html) | ['cmn'] | Classification | s2s | | | | +| [TRECCOVID](https://ir.nist.gov/covidSubmit/index.html) | ['eng'] | Retrieval | s2p | | | | +| [TRECCOVID-PL](https://ir.nist.gov/covidSubmit/index.html) | ['pol'] | Retrieval | s2p | | | | +| [TV2Nordretrieval](https://huggingface.co/datasets/alexandrainst/nordjylland-news-summarization) | ['dan'] | Retrieval | p2p | [News, Non-fiction] | {'test': 4096} | {'test': 784.11} | +| [Tatoeba](https://github.com/facebookresearch/LASER/tree/main/data/tatoeba/v1) | ['afr', 'amh', 'ang', 'ara', 'arq', 'arz', 'ast', 'awa', 'aze', 'bel', 'ben', 'ber', 'bos', 'bre', 'bul', 'cat', 'cbk', 'ceb', 'ces', 'cha', 'cmn', 'cor', 'csb', 'cym', 'dan', 'deu', 'dsb', 'dtp', 'ell', 'eng', 'epo', 'est', 'eus', 'fao', 'fin', 'fra', 'fry', 'gla', 'gle', 'glg', 'gsw', 'heb', 'hin', 'hrv', 'hsb', 'hun', 'hye', 'ido', 'ile', 'ina', 'ind', 'isl', 'ita', 'jav', 'jpn', 'kab', 'kat', 'kaz', 'khm', 'kor', 'kur', 'kzj', 'lat', 'lfn', 'lit', 'lvs', 'mal', 'mar', 'max', 'mhr', 'mkd', 'mon', 'nds', 'nld', 'nno', 'nob', 'nov', 'oci', 'orv', 'pam', 'pes', 'pms', 'pol', 'por', 'ron', 'rus', 'slk', 'slv', 'spa', 'sqi', 'srp', 'swe', 'swg', 'swh', 'tam', 'tat', 'tel', 'tgl', 'tha', 'tuk', 'tur', 'tzl', 'uig', 'ukr', 'urd', 'uzb', 'vie', 'war', 'wuu', 'xho', 'yid', 'yue', 'zsm'] | BitextMining | s2s | | {'test': 2000} | {'test': 39.4} | +| [TenKGnadClusteringP2P](https://tblock.github.io/10kGNAD/) | ['deu'] | Clustering | p2p | | {'test': 45914} | {'test': 2641.03} | +| [TenKGnadClusteringS2S](https://tblock.github.io/10kGNAD/) | ['deu'] | Clustering | s2s | | {'test': 45914} | {'test': 50.96} | +| [ThuNewsClusteringP2P](http://thuctc.thunlp.org/) | ['cmn'] | Clustering | p2p | | | | +| [ThuNewsClusteringS2S](http://thuctc.thunlp.org/) | ['cmn'] | Clustering | s2s | | | | +| [Touche2020](https://webis.de/events/touche-20/shared-task-1.html) | ['eng'] | Retrieval | s2p | | | | +| [ToxicChatClassification](https://aclanthology.org/2023.findings-emnlp.311/) (Zi Lin, 2023) | ['eng'] | Classification | s2s | [Constructed] | {'test': 1427} | {'test': 189.4} | +| [ToxicConversationsClassification](https://www.kaggle.com/competitions/jigsaw-unintended-bias-in-toxicity-classification/overview) | ['eng'] | Classification | s2s | | {'test': 50000} | {'test': 296.6} | +| [TurHistQuadRetrieval](https://github.com/okanvk/Turkish-Reading-Comprehension-Question-Answering-Dataset) (Soygazi et al., 2021) | ['tur'] | Retrieval | p2p | [Encyclopaedic, Non-fiction, Academic] | {'test': 1330} | {'test': 1513.83} | +| [TurkishMovieSentimentClassification](https://www.win.tue.nl/~mpechen/publications/pubs/MT_WISDOM2013.pdf) (Erkin Demirtas, 2013) | ['tur'] | Classification | s2s | [Reviews] | {'test': 2644} | {'test': 141.5} | +| [TurkishProductSentimentClassification](https://www.win.tue.nl/~mpechen/publications/pubs/MT_WISDOM2013.pdf) (Erkin Demirtas, 2013) | ['tur'] | Classification | s2s | [Reviews] | {'test': 800} | {'test': 246.85} | +| [TweetEmotionClassification](https://link.springer.com/chapter/10.1007/978-3-319-77116-8_8) (Al-Khatib et al., 2018) | ['ara'] | Classification | s2s | [Social] | {'train': 2048} | {'train': 78.8} | +| [TweetSarcasmClassification](https://aclanthology.org/2020.osact-1.5/) | ['ara'] | Classification | s2s | [Social] | {'test': 2110} | {'test': 102.1} | +| [TweetSentimentExtractionClassification](https://www.kaggle.com/competitions/tweet-sentiment-extraction/overview) | ['eng'] | Classification | s2s | | {'test': 3534} | {'test': 67.8} | +| [TwentyNewsgroupsClustering](https://scikit-learn.org/0.19/datasets/twenty_newsgroups.html) | ['eng'] | Clustering | s2s | | {'test': 59545} | {'test': 32.0} | +| [TwitterHjerneRetrieval](https://huggingface.co/datasets/sorenmulli/da-hashtag-twitterhjerne) (Holm et al., 2024) | ['dan'] | Retrieval | p2p | [Social] | {'train': 340} | {'train': 138.23} | +| [TwitterSemEval2015](https://alt.qcri.org/semeval2015/task1/) | ['eng'] | PairClassification | s2s | | {'test': 16777} | {'test': 38.3} | +| [TwitterURLCorpus](https://languagenet.github.io/) | ['eng'] | PairClassification | s2s | | {'test': 51534} | {'test': 79.5} | +| [UyghurSentimentClassification](https://arxiv.org/abs/2009.08712) | ['uig'] | Classification | s2s | [Reviews] | {'test': 841} | {'test': 245.2} | +| [VGClustering](https://huggingface.co/datasets/navjordj/VG_summarization) (Navjord et al., 2023) | ['nob'] | Clustering | p2p | [News, Non-fiction] | {'test': 2048} | {'test': 1009.65} | +| [VideoRetrieval](https://arxiv.org/abs/2203.03367) | ['cmn'] | Retrieval | s2p | | | | +| [VieMedEVBitextMining](https://aclanthology.org/2015.iwslt-evaluation.11/) (Nhu Vo, 2024) | ['eng', 'vie'] | BitextMining | s2s | [Medical] | {'test': 2048} | {'test': 139.23} | +| [VieQuADRetrieval](https://aclanthology.org/2020.coling-main.233.pdf) | ['vie'] | Retrieval | s2p | [Encyclopaedic, Non-fiction] | {'validation': 2048} | {'validation': 790.24} | +| [VieStudentFeedbackClassification](https://ieeexplore.ieee.org/document/8573337) (Nguyen et al., 2018) | ['vie'] | Classification | s2s | [Reviews] | {'test': 2048} | {'test': 14.22} | +| [WRIMEClassification](https://aclanthology.org/2021.naacl-main.169/) | ['jpn'] | Classification | s2s | [Social] | {'test': 2048} | {'test': 47.78} | +| [Waimai](https://aclanthology.org/2023.nodalida-1.20/) | ['cmn'] | Classification | s2s | | | | +| [WikiCitiesClustering](https://huggingface.co/datasets/wikipedia) | ['eng'] | Clustering | p2p | | | | +| XMarket | ['deu', 'eng', 'spa'] | Retrieval | s2p | | | | +| [XPQARetrieval](https://arxiv.org/abs/2305.09249) | ['ara', 'cmn', 'deu', 'fra', 'hin', 'ita', 'jpn', 'kor', 'pol', 'por', 'spa', 'tam'] | Retrieval | s2p | | | | +| [YelpReviewFullClassification](https://arxiv.org/abs/1509.01626) (Zhang et al., 2015) | ['eng'] | Classification | s2s | [Reviews] | {'test': 50000} | | +| [YueOpenriceReviewClassification](https://github.com/Christainx/Dataset_Cantonese_Openrice) (Xiang et al., 2019) | ['yue'] | Classification | s2s | [Reviews] | {'test': 6161} | {'test': 173.0} | \ No newline at end of file